Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Chemical and spectroscopic studies of complex organic molecules Bernstein, Michael Alec 1984

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1984_A1 B46.pdf [ 11.17MB ]
Metadata
JSON: 831-1.0059365.json
JSON-LD: 831-1.0059365-ld.json
RDF/XML (Pretty): 831-1.0059365-rdf.xml
RDF/JSON: 831-1.0059365-rdf.json
Turtle: 831-1.0059365-turtle.txt
N-Triples: 831-1.0059365-rdf-ntriples.txt
Original Record: 831-1.0059365-source.json
Full Text
831-1.0059365-fulltext.txt
Citation
831-1.0059365.ris

Full Text

c •<  CHEMICAL AND SPECTROSCOPIC STUDIES OF COMPLEX ORGANIC MOLECULES  by MICHAEL ALEC BERNSTEIN B.Sc.  (Hons.), University of Cape Town, 1977  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept  this thesis as conforming  to the required  standard  THE UNIVERSITY OF BRITISH COLUMBIA July, 1983 © Michael Alec Bernstein, 1983  In p r e s e n t i n g  t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of  requirements f o r an advanced degree at the  the  University  o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make it  f r e e l y a v a i l a b l e f o r reference  and  study.  I  further  agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying of t h i s t h e s i s f o r s c h o l a r l y purposes may  be granted by the head of  department or by h i s or her  representatives.  my  It is  understood t h a t copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain  s h a l l not be allowed without my  permission.  Department o f  C^^Ti^/  The U n i v e r s i t y of B r i t i s h 1956 Main Mall Vancouver, Canada V6T 1Y3 Date  IE-6  (3/81)  9e*T.;  Columbia  written  - i i-  ABSTRACT This thesis concerns i t s e l f mainly with  carbohydrates--  p a r t i c u l a r l y their physical properties and s t r u c t u r a l elucidation by spectroscopic methods. pursued.  Within this broad area a variety of topics are  The findings are divided into two sections; the f i r s t deals  with several spectroscopic approaches applicable to the study of carbohydrates attached to proteins (glycoproteins) and the second concentrates  on the use of recently devised NMR experiments which assist  in the determination  of molecular  structure and conformation.  The studies i n Section I depend on methods for "activating" s p e c i f i c l o c i i n glycoprotein glycans.  These materials were then  "tagged" with nitroxide spin-labels, which reported on the motional flexibility  of the glycans, which were found to be r e l a t i v e l y mobile, 2  and hydrated.  These data were corroborated  i n an analogous  H NMR  relaxation study, and this (superior) method reported very similar motional  behavior  to that derived from ESR.  F i n a l l y i n this section,  the u t i l i t y of synthetic glycoproteins i n the analysis of C  NMR  13  spectra of glycoproteins was studied. neoglycoprotein  F i r s t a useful method for  synthesis was devised, and proteins bearing  homogeneous 13  and known sugars were prepared.  It was determined that the  C chemical  s h i f t s of the pendant sugars are comparable with those of the corresponding  glycoside.  Enzymic oxidation studies suggested that this  approach may also greatly assist with the assignment of such carbohydrate molecules.  - iii -  The second section begins with a detailed survey of recent advances i n high-resolution NMR organic chemistry.  spectroscopy which may be useful i n  The experiments are explained i n a simple conceptual  fashion with i l l u s t r a t i v e data for trideuteriomethyl 2,3,4,6-tetra-0(trideuterioacetyl)-a-D-glucopyranoside given throughout.  This chapter  i s intended to be of assistance to the p r a c t i c i n g organic chemist has l i t t l e experience i n the areas of two-dimensional multi-pulse NMR  procedures.  who  (2D) and other  A limited number of these experiments are  used extensively i n later chapters. Chapter II.3 d e t a i l s a new procedure whereby an oligosaccharide may be characterized using NMR  spectroscopy, only.  The method i s "de  novo" i n that l i t t l e need be known about the molecule prior to analysis; i t r e l i e s on the characterization of a l l coupling pathways within each constituent monosaccharide and the determination of linkage order by inter-residue nuclear Overhauser  enhancements (nOe's).  To increase the  chemical s h i f t dispersion, the molecule i s derivatized - here as a per-O-acetate. 1  13  The then studied.  C and  H NMR  spectra of the plant a l k a l o i d , brucine, were  The emphasis here was on the choice and u t i l i z a t i o n of  selected 1- and 2D NMR  methods to determine the molecular structure and  conformation of a complex organic molecule i n the minimum of time. Elucidation of the f i n a l conformational d e t a i l required a strong emphasis to be placed on nOe  experiments.  - iv -  Chapter II.4 d e t a i l s an NMR digoxin.  study on a cardenolide glycoside,  Using the strategies devised e a r l i e r , the molecule was  at 500 MHz with the aim of complete spectral assignment.  studied  The spectrum  i s complicated by extensive signal overlap and the molecule's s e l f - a s s o c i a t i v e properties.  For this reason, the l i p o p h i l i c  (steroidal) moiety displays broader lines which r e s i s t analysis by spin-echo experiments.  The homonuclear and heteronuclear chemical s h i f t  c o r r e l a t i o n experiments performed well under these conditions, allowing the assignment of a l l sugar and some s t e r o i d a l protons/carbons.  The  genin resonances proved r e l a t i v e l y i n t r a c t a b l e , even with the powerful nOe  experiment.  - v TABLE OF CONTENTS Page ABSTRACT  i i  TABLE OF CONTENTS  v  LIST OF TABLES  viii  LIST OF FIGURES  x  ACKNOWLEDGEMENTS  xix  CHAPTER I.1 - MOTIONAL STUDIES ON GLYCOCONJUGATES 1.1.1  Background  2  References  12  CHAPTER 1 . 2 - ELECTRON SPIN RESONANCE STUDIES ON SPIN LABELLED GLYCOCONJUGATES 1.2.1 1.2.2 1.2.3 1.2.4  14  The ESR Spin Label Method Spin Labelling of Glycoconjugates Results and Discussion Information Derived from Spin Label Studies of Glycoconjugates •  14 20 25  References  32  CHAPTER 1 . 3 - DEUTERIUM MAGNETIC RESONANCE STUDIES ON GLYCOCONJUGATES References CHAPTER 1.4 - NEOGLYCOPROTEINS: 1.4.1 1.4.2 1.4.3 1.4.4 1.4.5  2  34 43  SYNTHESIS AND M4R STUDIES ...  Introduction Neoglycoprotein Synthesis C NMR of Neoglycoproteins *H NMR of Neoglycoproteins Conclusions References..  28  •  44 44 46 55 73 80 82  - vi Page CHAPTER 1.5 - EXPERIMENTAL FOR SECTION I 1.2 1.3 1.4 1.4.2 1.4.3  2  ESR Experiments H NMR Neoglycoproteins C NMR H SEAS Experiments  85 87 88 91 92  1 3 X  CHAPTER I I . 1 - INTRODUCTION  94  References  99  CHAPTER II.2 - HIGH RESOLUTION NMR METHODS 11.2.1 11.2.2  General Theory Spectral S i m p l i f i c a t i o n 11.2.2.1 11.2.2.2  11.2.3 11.2.4  I I . 2.5  Via Ti i n Via in  100 100 109  NMR Spectroscopy.... C NMR Spectroscopy  109 115  Spin Decoupling-Difference Spectra (SDDS) Nuclear Overhauser E f f e c t  119 125  11.2.4.1 11.2.4.2 11.2.4.3  130 135  Steady-State nOe Transient Overhauser Effect (TOE) Truncated Driven NOE Difference Spectroscopy  140  Two-Dimensional Experiments  142  11.2.5.1 11.2.5.2  Basic Concepts Homonuclear J-Modulated Spectroscopy (2D J-resolved) Jeener Experiment: J-Correlated 2D NMR...  142  11.2.5.3.1 11.2.5.3.2 11.2.5.3.3  168 175  11.2.5.3  11.2.5.3.4 11.2.5.4 11.2.5.5 11.2.5.6  II.2.6  85  COSY SECSY Delayed COSY - Detection of Long-Range Couplings "Decoupled" COSY  Heteronuclear Chemical Shift Correlation.. Relayed Coherence Transfer 2D Nuclear Overhauser Enhancement Spectroscopy  154 164  178 180 184 192 194  Conclusions  201  References  204  - vii Page CHAPTER II.3 - OLIGOSACCHARIDE CHARACTERIZATION DSING NHR ... References CHAPTER II.4 - NMR SPECTROSCOPIC ASSIGNMENT OF BRDCINE References CHAPTER I I . 5 - DIGOXIN References CHAPTER I I . 6 - EXPERIMENTAL FOR SECTION I I  209 229 232 257 258 284 285  - viii LIST OF TABLES Table  Page Chapter  1.2.1  1.2  Parameters characterizing spin label probes attached to various substrates. Percentage of possible s i t e s spin l a b e l l e d . N.D.: not determined due to low signal-to-noise Chapter  27  1.3 2  1.3.1  Comparison of T determined from H NMR relaxation measurements, and nitroxide ESR. a. T i n ns. b. Reductive amination of protein l y s y l 6-amine groups with hexadeuterioacetone, or 2,2,6,6-tetramethylpiperidine-4-keto-N-oxyl (TEMPONE) c  c  Chapter 1.4.1  41  1.4  Table of sugars coupled to BSA and their e f f i c i e n c i e s . BSA bears 55, l y s y l 6-amino groups per molecule.  Expressed as mole/mole  52  13 1.4.2  II.2.1  Comparison of C chemical s h i f t s of the carbohydrate resonances of neoglycoproteins with the methyl glycoside of the attached sugar. The chemical s h i f t s of neoglycorpteins are with an error of ± 0.1 ppm. A6 represents the difference in chemical s h i f t betwen a sugar resonance of a neoglycoprotein and i t s corresponding methyl glycoside... Chapter II.2 An indication of the r e l a t i v e number of publications r e l a t i n g to 2D NMR i n the l i t e r a t u r e , categorized by year  62  203  Chapter II.3 II.3.1  Comparison of expected and obtained v i c i n a l ring proton coupling constants for (A) a 8-glucopyranoside, and (B) 8-galactopyranoside ring  220  -  ix  -  Table  Page  Chapter II.4 11.4.1  11.4.2  Parameters  used  in  automated  collection J_-resolved  and p r o c e s s i n g experiments at  '^survey  of H COSY 360 MHz  S u m m a r i z e d *H NMR c h e m i c a l s h i f t c o u p l i n g constants of brucine i n  conditions" and  2D 240  assignments and CDCI3 ( c a . 0 . 0 2 M ) .  251  13 11.4.3  Table of assigned compared w i t h the are referenced to (6 7 7 . 0 ) . The s o l CDCl3:CD30D (10:1) WH-400, o p e r a t i n g  C chemical shifts for brucine, literature. Chemical shifts t h e c e n t r a l t r a n s i t i o n o f CDCI3 u t i o n was c a . 0 . 3 M i n and the i n s t r u m e n t a B r u k e r a t 1 0 0 . 6 MHz f o r C  254  Chapter II.5 II.5.1  Expected residue  3 J_ v a l u e s  for  a  8-digitoxose  sugar 263  x LIST OF FIGURES Figure  Page Chapter 1.2  1.2.1  Solution ESR spectrum of a rapidly tumbling nitroxide spin l a b e l i l l u s t r a t i n g the measurement of parameters (From Ref. 10)  16  Solution ESR spectra of a nitroxide spin l a b e l as a function of molecular motion, controlled by a l t e r i n g solvent v i s c o s i t y . (Most mobile, bottom spectrum; least mobile, top spectrum. From Ref. 3)  18  1.2.3  0- and N-glycosides  21  1.2.4  Oxidation of gactose termini by galactose oxidase  23  Reductive amination of an aldehyde with TEMPAMINE, using NaCNBH 3  24  1.2.2  1.2.5 1.2.6  of f e t u i n .  (Frm Ref. 17)  ESR spectra of n i t r o x i d e - l a b e l l e d materials. Concentration and r e l a t i v e gain are given i n parentheses. A. Periodate-activated fetuin (3.24 mg i n 50 p i H 0; gain 1). B. Galactose oxidaseactivated asialo-BSM (2.67 ul In 50 u l H 0; gain 6.4). C. Periodate-activated, intact erythrocytes (73 ul packed c e l l s i n phosphatebuffered saline; gain 187.5). D. Fetuin perfused with 1 and NaCNBH3 f o r 2 h (3.00 mg i n 50 u l H 0; 2  2  2  gain 62.5) (From Ref. 14)  26  Chapter 1.3 1.3.1  1.3.2  Procedure for the s p e c i f i c deuteration of s i a l i c acid termini of glycoproteins at the C-7 position  38  Procedure for the s p e c i f i c deuteration of galactose termini of a s i a l o - g l y c o p r o t e i n 6 at the C-6 position  38  - xi Figure 1.3.3  Page NMR spectra measurement at 61.4 MHz. Samples were dissolved i n «* 1.5 ml H2O. (A) BSM deuterated at C-7 on s i a l i c acid residues (145 mg; 2700 t r a n s i e n t s ) . (B) A s i a l o f e t u i n deuterated at C-6 position on terminal galactose residues (190 mg). (C) Fetuin reductively aminated with hexadeuterioacetone on lysine residues (240 mg; 550 t r a n s i e n t s ) . The HOD resonance was a r b i t r a r i l y assigned a s h i f t of 5 ppm. From Ref. 7  39  Chapter 1.4 1.4.1  Gray's procedure for the attachment of a reducing disaccharide to a protein (P) via i t s l y s y l 6-amino groups  46  1.4.2  Neoglycoprotein  48  1.4.3  Reaction mechanism for the reductive ozonolysis of an alkene i n methanol, using dimethyl sulphide..  50  C NMR spectra at 100.3 MHz of (A) native BSA in D2O, and (B) the denatured protein i n 8 M urea/D 0  56  1.4.4  synthesis protocol.  From Ref. 7...  13  2  1.4.5  C NMR spectra at 100.3 MHz, showing the "carbohydrate region" only. (A) Native BSA. (B).(A), superimposed with the spectrum of methyl 6-D-glucoside, with line-widths broadened  13  by 2 Hz.  D: i n t e r n a l dioxane reference ( 6 67.4)...  58  13  1.4.6  C NMR spectrum of a neglycoprotein (8-glucose attached to BSA), with the "carbohydrate region" expanded and assigned. Lines aboves the resonances indicate the l i t e r a t u r e chemical s h i f t s of methyl 6-D-glucoside  60  1.4.7  The "carbohydrate region" of the 100.3 MHz C NMR spectrum of three neoglycoproteins: (A) 6-glucose, (B) a-galactose, and (C) B-lactose attached to BSA...  61  1.4.8  13  The action of D-galactose a-D-galactopyranoside.  oxidase on methyl 64  - xli Figure 1.4.9  1.4.10  1.4.11  Page C NMR spectrum (100.3 MHz) of (A) methyl o-galactoside, and, (B), the same after oxidation at C-6 by g-galactose oxidate, at 20% e f f i c i e n c y . . .  65  The 100.3 MHz C NMR spectrum of B-lactose attached to BSA. (A) Unoxidized. (B) After oxidation by galactose oxidase overnight. (C) After reduction of (B) with NaBHi*  67  The 100.3 MHz C NMR spectrum of a mixture of BSA and a l l y l 6-lactoside, (A) after enzymic oxidation by galactose oxidase, and (B), a f t e r NaBH reduction of (A)  68  The 270 MHz *H spectra of BSA i n D 0 (14 mg/6.4 ml). (A) Single pulse experiment (B) -(F) SEAS experiments with the indicated T values. The HOD was supressed by 3 s presaturation  76  The 270 MHz H SEAS spectra with i = 40 ms. (A) BSA. (B) BSA plus a l l y l B-acetoglucosaraine. (C) 8-AcetoglcNAc attached to BSA. (D) a- and 8-glcNAc attached to BSA  77  1 3  1 3  1 3  4  1.4.12  1.4.13  2  X  Chapter II.2 11.2.1  The laboratory (A) and rotating-reference frame (B)  102  11.2.2  A s t y l i z e d AX spin system  102  11.2.3  The action of an AX spin system's vectors, (A) at equilibrium, (B) after a 90° pulse and (C) a f t e r some time to allow dephasing. (Solid l i n e denotes A spin vectors and broken l i n e s , X)...  104  Energy l e v e l diagrams of AX spin-system (A) showing t r a n s i t i o n p r o b a b i l i t i e s (B). 0 and ±6 represent r e l a t i v e equilibrium populations  104  The rotating reference frame vector model of the Tj inversion-recovery experiment. Vectors are shown f o r a short t^ value (A,B,C,E) and a r e l a t i v e l y long one (A,B,D,F)  110  Plot of z-magnetization vs time i n a Tj inversion recovery experiment, indicating the n u l l point.....  110  11.2.4  11.2.5  11.2.6  - xiii Figure II.2.7  Page 400 MHz T1IR experiment on 1 (0.1 M i n C6D6). The relaxation delay was 12 s, and t^ values noted next to each spectrum...  112  Spectral editing of 1 by T1IR. The control spectrum i s C. t 0.7s i s a "methine subspectrum", with methylene signals nulled. _t • 2.5s gives the "methylene sub-spectrum", with methine signals p a r t i a l l y nulled. Other parameters as In F i g . II.2.7. The asterisk marks an a r t i f a c t at the c a r r i e r position  113  11.2.9  DEPT performed on 2 (0.4 M; CDC1 ) at 100 MHz  118  11.2.10  SDDS on 1 at 400 MHz. The i r r a d i a t e d proton i s indicated i n each case. False responses a r i s i n g from Bloch-Ziegert effects are labelled "B.Z."  122  SDDS i r r a d i a t i n g H-l [ ( Y B / 2 H ) = 17 Hz], and varying the frequency difference ( A v ) between the decoupler on resonance (fixed) and the control, off-resonance (varied)  124  Energy l e v e l s , t r a n s i t i o n p r o b a b i l i t i e s and r e l a t i v e populations (A) for an IS system at equilibrium, and (B) for an l { s } s i t u a t i o n (where S i s saturated)  126  SSN0EDS on 1 (400 MHz). The i r r a d i a t e d proton i s indicated next to each difference spectrum. The i r r a d i a t i o n time was 6s  134  (A) The equilibrium z-magnetization of 1 (H-5). The z-magnetization immediately after (B) a single selective 180° pulse, and (C) a compositive 180° pulse  136  II.2.8  B  11.2.11  11.2.12  11.2.13  11.2.14  11.2.15  3  2  Difference inverting The time, inversion with each  TOE on 1 (400 MHz), s e l e c t i v e l y H-5 with a composite pulse (see t e x t ) . t^, between selective population and t o t a l signal a c q u i s i t i o n i s given spectrum ,  138  - xlvFigure 11.2.16 11.2.17  11.2.18  11.2.19  11.2.20  11.2.21  11.2.22  11.2.23  11.2.24  11.2.25  11.2.26 11.2.27 11.2.28  Page Data from F i g . II.2.15, showing the nOe build-ups of H-l and H-3 with time  139  Schematized data matrices i l l u s t r a t i n g the stages of processing of a 2D data-set  143  E f f e c t of apodization functions on line-shapes i n phase-sensitive (column 2) and magnitude mode (column 3)  146  Comparison of stacked (A) and contour plots (C). B. The F j projection of the t i l t e d data-set. The s t i p l e d peak arises through strong coupling between 6' and 6'. Data are from the same region A of a 2D J-resolved experiment of a disaccharide....  149  A. Stacked-plot of schematized 2D data set. B. Contour plot i l l u s t r a t i n g f u l l 0° and 90° projections (top and r i g h t ) , and " p a r t i a l " 90° projections (bottom)  150  Vector diagram model of the 2D J-resolved experiment. With the X transitions i n an AX spin system (A-E), both spins f e e l both pulses; i n A,B,C,F,G, only spin X i s subjected to the refocussing pulse (From Freeman )  157  Schematized 2D j-resolved contour plot output of AX spin system without (A) and with (B) tilting  159  270 MHz 2D J-resolved experiment on 1. for d e t a i l s  162  See text  Data taken from F i g . II.2.23B, now plotted i n the contour-plot mode. Quadrature images for the intense H-2 resonance are marked with an a s t e r i s k . .  163  Stylized jUF^.F^) contour-plot output of an AX spin system (A) with COSY and (B) with SECSY experiments •  165  Plot of s i n (TIJAXII) exp(-£ /T ) vs. t j f o r J - 4.0 and 0.5 Hz  167  Schematized density matrix formalism of COSY experiment for an AX spin-system  172  COSY spectrum of 1 (270 MHz)  174  1  2  - xv Figure 11.2.29  11.2.30  11.2.31 11.2.32  11.2.33  II.2.34  11.2.35  Page SECSY spectrum of 1 (400 MHz). The sweep-width i n J_i was s l i g h t l y small, causing a l i a s i n g of part of the J(4,5) connectivity; the aliased portion i s marked with an a s t e r i s k . . . . . . . . .  156  Delayed COSY experiment on 1 (270 MHz). A • 0.4 s. Responses from long-range couplings are shaded  179  Pulse scheme for the COSY experiment with homodecoupling i n Fj  181  Decoupled COSY experiment with 1 at 270 MHz. t^ • 182 ms and A ° 0. (A), contour-plot of 2D spectrum. (B), J j projection. (C), control spectrum  183  Energy levels of a CH fragment at equilibrium (A) and after a proton population inversion (B). (From a review by R. Freeman, Ref. 45)  186  The affect of a proton 90° - t_i - 90° pulse sequence on proton vectors, and their i n t e r pretation i n terms of proton spin state populations i s given i n (A) and (B). (C) maps the corresponding change i n population differences across the C t r a n s i t i o n s . (Adapted from Freeman and Morris. )  186  CSCM on a 0.5 M solution of 2, at 100.3 MHz f o r C i ~.  189  Steps of magnetization transfer In the relayed coherence transfer experiment  192  1 3  11.2.36 11.2.37  Relayed coherence transfer experiment performed on n-propanol ( H frequency 400 MHz). Peaks present i n the CSCM experiment are labeled with an asterisk  194  11.2.38  Schematized NOESY experiment and vector model  196  11.2.39  2D NOESY spectrum of 1 at 400 MHz. A mixing time of 0.3 s was used, randomized ± 10%  200  -  xvi  -  Figure  Page Chapter II.3  11.3.1  11.3.2  I I . 3.3  With a 6(1-4) glycosidic linkage, i r r a d i a t i o n of the g l y c o s i d i c proton (H-l') induces i n t r a - r i n g ( ) and a single i n t e r - r i n g ( ) nOe  215  270 MHz spectra of (A) B - a l l y l lactoside In D2O, and (B) 0 - a l l y l aceto(d_6) lactoside  215  Contour plot of 2D J_-resolved NMR spectrum of 2 (270 MHz; 0.2 M i n CsT> ). The assigned, resolution-enhanced spectrum i s plotted above e  11.3.4  11.3.5  11.3.6  11.3.7  217  400 MHz COSY spectrum of 2 (0.2 M i n C6D ). 6  Connectivities within the glycopyranosyl ring are shown ( ) i n the top l e f t h a l f , and the galactopyranosyl ( ) i n the bottom right h a l f . A l l y l connectivities are shown ( — • — • — ) . .  218  400 MHz SSN0EDS of 2 (0.2 M i n C 6 D ) . Intra-ring nOe's are shown with a s o l i d l i n e ( ) and i n t e r - r i n g nOe's with a broken line ( )  221  NOESY of 2 (0.2 M i n C6D6) at 400 MHz. The mixing time, = 0.75 s ± 15%. Intra-ring nOe's for the glucosyl ( ) and galactosyl ( ) rings are drawn i n the bottom right h a l f , and i n t e r - r i n g ( ) i n the top l e f t h a l f . . .  222  Combined NOESY (top l e f t ) and COSY (bottom right) data taken from Figs. II.3.4. and II.3.6, respectively  223  6  Chapter II.4 11.4.1  11.4.2  11.4.3  400 MHz *NMR spectrum of brucine (ca. 0.02 M i n C D C I 3 ) , showing the assignment of non-aromatic protons  237  360 MHz COSY spectrum of brucine i n the region between 6 4.5 and 1.0. Connectivities to the v i n y l i c proton are indicated, though they are not on the diagram  239  SSNOEDS experiments performed on brucine at 400 MHz  243  - xvii Figure 11.4.4  11.4.5  11.4.6  11.4.7  Page Conformations of rings IV and V (A), VI (B), VII (C) and II (D) i n brucine, indicating some of the results from the SSNOEDS experiments. The i r r a d i a t e d proton i s indicated with (•) and nOe's ( )  245  400 MHz NOESY spectrum of brucine. The 256 * IK data-matrix was collected i n ca. 8 h, with relaxation delay 3.0 s, mixing time 0.6 s ± 10% and 32 scans per t± i n t e r v a l . The data were processed and plotted using a Nicolet 1280 computer and Zeta 8 recorder. The f i n a l d i g i t i zation was 5.8 Hz pt~ i n both dimensions, a f t e r z e r o - f i l l i n g J j ; the " s i n e b e l l " m u l t i p l i e r was used i n both dimensions, and magnitude spectra calculated  248  The h i g h - f i e l d region of the NOESY plot i n Fig. II.4.5  249  360 MHz ( H) CSCM experiment with brucine (ca. 0.3 M i n C D /CD OD, 10:1). 6 22-80 and 6JJ 1.0-4.5 i s plotted i n the contour mode. Fj and F_£ projections are on the abcissa and ordinate, respectively  253  1  6  6  3  C  Chapter II.5 II.5.1  Solid-state conformations of digoxigenin (lower) and digoxin glycone (upper). (From Go et a l . )....  260  A. 500 MHz *H NMR spectrum of digoxin i n DMS0-d_6. B. Same molecule following a D 0 exchange. Hydroxy protons are marked (•) i n A  261  Expansions of the low-field region of the 500 MHz H NMR spectrum of digoxin i n DMS0-d_6i with a drop of D 0 added  264  500 MHz COSY spectrum of digoxin (1) i n DMS0-d , a f t e r deuterium exchange. The HOD signal was suppressed by p r e i r r a d i a t i o n , and the 512 * 1024 word data set symmetrized prior to display i n the contour mode  269  11.5.5  Expansion of region A i n F i g . II.5.4  270  11.5.6  Expansion of region B i n F i g . II.5.4  271  I I . 5.2  2  11.5.3  2  11.5.4  6  -  xviii  -  Figure 11.5.7  Page D r i v e n nOe e x p e r i m e n t s p e r f o r m e d o n d i g o x i n , showing nOe's induced i n t o the methylene/methine region  11.5.8  11.5.9  D r i v e n nOe e x p e r i m e n t s s h o w i n g i n t o d i g i t o x o s e H-5 p r o t o n s by or H - l " and H - l " » Driven peaks the  experiments  irradiated;  the  are  evident  to  with  nOe's induced irradiating H-l', 274  steroidal  displayed protons, lower  methyl  region although  is  that  of  other  field  276  CSCM e x p e r i m e n t ^ H , 3 6 0 M H z ) p e r f o r m e d o n digoxin. The l o w - f i e l d r e g i o n s a r e p r e s e n t e d as  11.5.11  nOe  methylene/methine  nOe's 11.5.10  273  a  stacked-plot  278  T h e h i g h - f i e l d r e g i o n o f t h e CSCM d a t a - s e t i n F i g . I I . 5 . 1 0 i s p l o t t e d between the frequency  limits  indicated  279  - xix ACKNOWLEDGEMENT  It i s a pleasure to acknowledge the guidance, encouragement and friendship of Dr. L.D. H a l l . It i s impossible to thank a l l those who contributed to my education, but a number of individuals stand out. Work i n neoglycoprotein synthesis was with the assistance of Ms. L. Darge (formerly, Evelyn), Dr. J.M. Berry and Dr. J.D. Stevens.  The  glycoconjugate work was assisted by Drs. P.E. Reid, J.D. Aplin, D.E. Brooks, the late Prof. C.F.A. C u l l i n g and Mr. R. Snoek.  Drs. J.C.  Waterton, S. Sukumar and G.A. Morris provided helpful discussions i n the i n i t i a l NMR studies.  It i s a pleasure to thank Drs. R.E. Hurd and L.F.  Johnson of Nicolet Magnetics for their contributions and use of equipment and Dr. G. Pouzard for discussions on NMR theory.  Dr. W.E.  H u l l recorded the H NMR spectra. F i n a l l y , any instrument-oriented study such as this r e l i e s on expert technical back-up, and Mr. T. Marcus and others i n the e l e c t r i c a l shop of this department are to be thanked. accurate and cheerful art-work.  Ms. E. Jensen provided f a s t ,  - 1 -  SECTION I  - 2 -  CHAPTER 1.1 MOTIONAL STUDIES ON GLYCOCONJUGATES I.I.I  Background It  has been known for some time that carbohydrates are found i n  Nature covalently linked to proteins and l i p i d s ;  1  commonly called "glycoproteins" and " g l y c o l i p i d s " , c o l l e c t i v e l y called glycoconjugates. play i n l i f e processes,  respectively and  Owing to the important part they  a tremendous e f f o r t has been placed i n the  characterization and determination molecules.  such molecules are  of the function and  In what follows here we w i l l concentrate  structure of such  on  glycoproteins;  spectroscopic studies have already played an important part i n the understanding of their function, and part of the purpose of the work described i n this chapter i s to develop methods for chemically  "tagging"  s p e c i f i c carbohydrates to act as "reporters" on their microenvironment. The  l i t e r a t u r e on glycoproteins w i l l not be reviewed i n any 2  detail.  Recent reviews by Montreuil,  recommended to the interested reader.  great  3  and Wagh and Bahl  are highly  Although somewhat dated now,  book by Sharon** covers most topics i n an "easy to read" s t y l e .  a  Membrane  glycoproteins' structure and function are i n s i g h t f u l l y reviewed by Hughes.  5  Glycoproteins  have been found i n v i r t u a l l y a l l forms of l i f e ,  from microorganisms to man.  They serve a wide range of functions, some  of which we w i l l soon high-light.  The  contribution of carbohydrate to  the t o t a l mass of a glycoprotein varies considerably;  thus, lysozyme has  no detectable carbohydrate, and some blood-group substances are more  - 3 -  than 80% carbohydrate. "proteoglycans",  6  An important  i s predominantly  proportion of protein.  class of structural molecules,  the  polysaccharide with a small  Three classes of monosaccharides are commonly  found i n mammalian glycoproteins:-  (1) neutral sugars, such as  g-galactose (gal), g-mannose (man), g-glucose  ( g l c ) , and L-fucose  (fuc),  (2) N a c e t y l amino sugars, N-acetyl glucosamine (glcNAc) and -  N-acetylgalactosaraine (galNAc), and (3) a c i d i c sugars such as the N_-acetylneuraminic acids (NANA). O-acetylated. sugars:  The neuraminic  acids may  Plant and microbial glycoproteins may  be  contain deoxy  g-xylose, L~arabinose and L-rhamnose. Although this diverse selection of monosaccharides i n  glycoproteins can, i n p r i n c i p l e , be joined i n an almost i n f i n i t e number of ways, consideration of the carbohydrate units and the nature of their linkage to the protein allows for a certain degree of c l a s s i f i c a t i o n . The glycans may  be conjugated to the protein i n two ways,  this i s the f i r s t basis for c l a s s i f i c a t i o n .  and  (1) The O-glycoproteins are  O - g l y c o s i d i c a l l y linked to L-serine or L-threonine units i n the protein back-bone; (2) the N-glycoproteins are N-glycosidated to L-asparagine. With O-glycoproteins, the sugar residue involved i n the linkage i s often galNAc, and the glycan can be detached treatment,  from the protein by a l k a l i  through a 8-elimination reaction. N-glycoproteins involve so  far only glcNAc attached to Jj-asparagine, and the glycan i s usually isolated as a glycopeptide, following extentensive proteolysis; they are more abundant than O-glycoproteins.  A proposed c l a s s i f i c a t i o n scheme  has been put forward by Montreuil,  where he draws comparisons between  - A -  glycoprotein glycans and the immunoglobulins; i n the glycoproteins he i d e n t i f i e s certain non-specific "invariable", or "core" arrangements to which are attached additional carbohydrates  coding the "variable"  fraction. The existance of certain "core" structures suggests a limited number of biosynthetic pathways.  A key step i n biosynthesis must be the  attachment of the glycan to the protein.  Thus, the amino-acid sequence  around the glycan attachment has been extensively studied. N-Glycoproteins  Invariably are linked to the peptide sequence Asn*-X-Thr  (Ser), where Asn i s asparagine, Thr, i s threonine and Ser i s serine. But the presence of this sequence does not guarantee glycosylation, nor does the p o l a r i t y of "X" determine the nature of carbohydrate attached.  units  Such an invariant amino-acid sequence i s not clear with  fj-glycoproteins.  Many d e t a i l s of the biosynthesis of glycoproteins are 8  9  known and the elegant results are reviewed. » Structure determination methods have developed considerably. C l a s s i c a l "wet" chemical methods are now augmented with GLC-MS (of 13  methylated d e r i v a t i v e s ) , enzymic (glycosidase) methods and NMR ( 1  H).  The interested reader i s referred to the l i t e r a t u r e  1 0  C and  for further  d e t a i l s , and discussions i n Ch. II.3. Whilst the fundamental aspects of glycoproteins, outlined above, have concerned some workers, others have asked why such molecules even exist.  Glycoproteins were i n i t i a l l y  playing no important major r o l e s .  biological role.  thought to be metabolic  accidents  In f a c t , glycoproteins play two  The f i r s t i s physico-chemical.  Glycoproteins such as the  - 5 -  "mucins" l u b r i c a t e the g a s t r o - i n t e s t i n a l t r a c t , protecting i t from mechanical abrasion and p r o t e o l y t i c attack.  Such viscous glycoproteins  also l u b r i c a t e the eye-socket and form a b a r r i e r at the cervix, preventing bacteria from entering the uterus (and abdominal c a v i t y ) . Upon ovulation, the mucin changes I t s physical properties so as to permit spermatazoa to enter the uterus and effect conception.  The  freezing point of c e r t a i n A r c t i c f i s h ' s serum i s depressed by special "anti-freeze" glycoproteins.*  1.  With a l l these glycoproteins, the  i n t e g r i t y of the glycan i s e s s e n t i a l for proper a c t i v i t y .  (With the  enzyme ribonuclease B, however, the glycoprotein has the same a c t i v i t y as the carbohydrate-free  counterpart,  ribonuclease A).  The second major role of glycoproteins concerns interactions.  molecular  Here the carbohydrate units play a very Important role.  For example, the ABO and Lewis blood-group c l a s s i f i c a t i o n results from 12  the constituent sugar residues of the glycan.  These glycans have  received wide interest and an understanding of their properties has benefitted greatly from synthetic and conformational by Leraieux  13  and co-workers.  studies performed  The influenza virus i s known to  agglutinate red blood-cells, unless s i a l i c acid i s removed from the c e l l membranes. The removal of the carbohydrate from the hormone, human chorionic gonadotropin, attenuates  i t s b i o l o g i c a l a c t i v i t y , but does not a f f e c t  i t s immunological a c t i v i t y .  The former implies a contribution of the  carbohydrate to the b i o l o g i c a l a c t i v i t y , i n either an active or passive way.  The l a t t e r i s i n accord with the well-known fact that glycan  - 6 -  moieties of glycoproteins are usually weakly immunogenic - the obvious exception being the blood-group  substances.  Glycan termini, often s i a l i c acids, serve as determinants of recognition for the clearance of some plasma glycoproteins.  and  *  Ik  Morell showed  Ashwell  that deslalized ceruloplasmin, now  bearing terminal  galactose residues, i s rapidly removed from c i r c u l a t i o n by the Although this clearance mechanism i s believed to be f a i r l y  liver.  general,  a s i a l o - t r a n s f e r r i n appears to have an alternative catabolic pathway. S i m i l a r l y , the presence of s i a l i c acid on c i r c u l a t i n g mammalian erythrocyte membranes Is e s s e n t i a l for their s u r v i v a l .  1 5  The glycoproteins on c e l l - s u r f a c e s are believed to play an important  role i n the way  i n i t s surroundings  the c e l l interacts with both small molecules  and, indeed, other c e l l s .  1 6  The importance of  c e l l - s u r f a c e glycoproteins i s perhaps exemplified by the changes occuring upon malignant or v i r a l transformation.  The evidence for the  Involvement of glycoproteins i n c e l l - c e l l Interactive phenomena i s far less d e t a i l e d , but this topic s t i l l must be regarded understanding  of how  as p i v o t a l to the  c e l l s grow to form tissues, and what processes  can  r e s u l t i n the loss of contact i n h i b i t i o n and the onset of uncontrolled cancerous growth. Most studies on glycoprotein functions have been phenomenological.  For example, Ashwell and Morell  labelled ceruloplasmin  a radionuclide and injected (a) the intact ( s i a l o - ) compound, and  with (b)  the asialo-compound into an animal, and measured the plasma h a l f - l i f e * *The time required to reduce the plasma concentration by  50%.  -  In each case.  7  -  In that the asialoceruloplasmin had a much shorter plasma  h a l f - l i f e , i t was  obviously being cleared from c i r c u l a t i o n faster than  the i n t a c t glycoprotein and this indicated that the s i a l i c acid guarantees the glycoprotein's c i r c u l a t o r y s u r v i v a l .  The r a d i o a c t i v i t y  accumulated i n the l i v e r , from whence the receptors for asialoceruloplasmin were l a t e r p u r i f i e d . The work of Ashwell and Morell i s now  considered  a "classic" in  studies on the role of the carbohydrate moiety of glycoproteins. Others, however, have focussed  their attention on the ways i n which  these phenomena occur at the molecular l e v e l ; the purpose of this section of the thesis i s to investigate some of the "tools" available for these studies, and the chemistry required to make them f e a s i b l e . Such studies involve spectroscopy, resonance) and NMR  17  and nitroxide ESR  (electron spin  (nuclear magnetic resonance) spectroscopy w i l l be  used here. A variety of methods and procedures have been investigated, and each chapter i n t h i s section w i l l be introduced discussion.  l a t e r i n this  F i r s t , we b r i e f l y discuss each spectroscopic method, to  define i t s p o t e n t i a l information content, and state the  underlying  prerequisites for i t s use. The p r i n c i p a l useful spectroscopic techniques available are fluorescence  studies, ESR  and NMR.  Fluorescence  and ESR  studies  i n v a r i a b l y require chemical attachment ("labelling") of the macromolecule with a reporter probe; NMR nuclides (e.g. H, 1  13  C  or  3 1  u t i l i z e naturally occuring  P ) , or i s o t o p i c incorporation (e.g. H, 2  13  or isotope enrichment of  may  C).  1 9  F,  -  8 -  How do these methods provide information with atomic detail? ESR  The  of spin-labels has been used to y i e l d d i f f e r e n t types of  information, but we concentrate on just two. F i r s t l y , the "g-value" of a nitroxide spin-label - analogous to chemical s h i f t i n NMR - i s an i n d i c a t i o n of the p o l a r i t y of the microenvironment i n which the probe i s resident.  For this information to be of any value, s i t e - s p e c i f i c  incorporation i s e s s e n t i a l .  This Introduces an important  underlying the studies i n this chapter: necessary  concept  chemical or enzymic methods are  to activate unique s i t e s i n a complex macromolecule f o r the  derived data to be of significance. Secondly, ESR, NMR relaxation and fluorescence studies share a common feature i n that they can p o t e n t i a l l y provide the "correlation time", T  £  - a measure of the motional l i b r a t i o n of the reporter group.  Roughly, the c o r r e l a t i o n time i s the time the probe takes to rotate i n solution through one radian, and t y p i c a l values encountered here range between 10  (fast rotation) and 10  s. (medium r o t a t i o n ) .  Rather than placing too much s i g n i f i c a n c e on actual derived numbers, most studies rely on comparisons of system.  upon perturbation of the  For example, preparation of a range of f a t t y acids bearing  reporter groups at different positions along the acyl chain, Incorporation into a model membrane system and measurement of  might  allow one to come to some conclusions regarding the r e l a t i v e mobility of d i f f e r e n t points of the chain.  Incorporating the previous concept of  -  9 -  s i t e - s p e c i f i c modification, a person wishing to study the interacion between a s p e c i f i c carbohydrate residue and a macromolecule to which i t i s bound might label that residue s p e c i f i c a l l y and monitor any changes In x  £  as the binding protein i s added. One advantage of attaching a spin-, fluorescent- or isotope-  probe to a molecule i s that the remainder of the molecule w i l l be 8pectroscopically simplified.  " s i l e n t " and interpretation of data may be greatly  S t i l l , a good case can be made for the observation of a  glycoprotein's NMR spectrum at natural abundance.  With proton spectra,  the problem l i e s i n the very large number of protein  resonances  overlapping with those of carbohydrate o r i g i n , and their  assignment.  Concerning the l a t t e r , a r i c h legacy of experimental data stems from the efforts  of Vliegenthart'8 group i n Holland, where the focus has been on  the sequencing of glycopeptide carbohydrates.  With carbon spectra,  the carbohydrate resonances f a l l into a "window" of the spectrum having very few protein resonances.  Thus, the resonances are e a s i l y detectable  (within the usual constraints of low inherent s e n s i t i v i t y ) , but assignment  i s s t i l l a problem.  Although the topics of this section e a s i l y  f a l l into three  d i s t i n c t categories, i t should be emphasized that the overlap i s considerable, at the conceptual l e v e l . In fact, many concepts share elements i n common with those described i n Section I I . Chapter 1.2 deals with nitroxide spin-label studies on Intact and Isolated glycoproteins. Methods were developed to s p e c i f i c a l l y  label  - 10 -  carbohydrate termini of s i a l o - and asiologlycoproteins,* and correlation time measurements made.  S i m i l a r l y , red blood c e l l surface  glycoproteins and g l y c o l i p i d s were tagged i n the same way and the data compared. Whilst the s p i n - l a b e l l i n g technique has many advantages, I t i s 20  also fraught with uncertainties.  Deuterium NMR i s recognized as a  r e l i a b l e source of c o r r e l a t i o n times, and i n Ch. 1.3 the synthetic scheme was altered for the s i t e - s p e c i f i c introduction of deuterons, and c o r r e l a t i o n time measurements were made.  This study was the f i r s t of  i t s kind i n this area and, i n addition to proving f e a s i b i l i t y of the technique,  provided motional  c o r r e l a t i o n times which compared favourably  with those from the ESR studies i n Ch. 1.2. Next, the question of the observation of ^•H and C NMR spectra 1 3  of glycoprotein glycans was addressed (Ch. 1.4). Since most i s o l a t e d glyco-proteins bear a variety of d i f f e r e n t glycans on a single protein 22  core, i t was decided that "synthetic glycoproteins" -  neoglycoproteins  - would be useful i n that one could attach any (predetermined) sugar and so simplify the analysis.  Several schemes for neoglycoprotein  synthesis  existed i n the l i t e r a t u r e which f a i l e d to meet a l l the c r i t e r i a we considered necessary f o r a useful protocol, and we therefore developed our own.  Some of these materials were analysed by H NMR, where a 1  multi-pulse experiment was used to assess the p o s s i b i l i t y of d i f f e r e n t i a l l y observing carbohydrate resonances under the broad protein * I t i s these carbohydrate residues which are responsible for the recognition phenomena involved i n the clearance of some plasma glycoproteins and c e l l s .  - 11 -  envelope.  With  C NMR, assignments were made by comparisons of  chemical s h i f t s of the attached and unattached sugar glycoside, and chemical s h i f t changes induced by enzymic modification.  This also  provided some interesting information on the requirements of the enzyme. It w i l l become evident i n the following chapters that a variety of methodologies exist for the study of glycoproteins by spectroscopic methods, each with i t s p a r t i c u l a r benefits and shortcomings.  The  handling of b i o l o g i c a l materials also provides new technical challenges for most spectroscopists and organic chemists, but the r e s t r i c t i o n to r e l a t i v e l y simple systems s t i l l allows conceptually important to be tested.  hypotheses  - 12 -  REFERENCES 1.  N. Sharon. S c i e n t i f i c American. (5), 90-116 (1980).  230 (5), 78-86 (1974); Ibid.  2.  J . Montreuil.  3.  P.V. Wagh and O.P. (1981).  4.  N. Sharon. Complex carbohydrates: their chemistry, biosynthesis and functions. Addison-Wesley Publishing Co., Mass. 1975.  5.  R.C. Hughes. Membrane glycoproteins: a review of structure and function. Butterworths, London, 1976.  6.  V.C. Hascall. Biology of carbohydrates, V o l . 1. Edited by V. Ginsburg and P. Robbins. John Wiley and Sons, New York. 1981. 1-49.  Adv. Carbohydr. Chem. Biochem. Bahl.  243  37, 157-223 (1980).  CRC C r i t . Rev. Biochem. 10, 307-377  7.  N. Parthasarathy and S.M. (1978) .  Bose.  J . Sclent. Ind. Res.  8.  A.P. Cornfield and R. Schauer. (1979) .  9.  H. Schachter. The glycoconjugates, V o l . 2. Edited by M.I. Horowitz and W. Pigman. Academic Press, New Y o r k . 1978. pp 87-181.  Biol. Cellulaire.  pp  37, 305-316  36, 213-226  3  10.  Relevant references may be found i n the review by Wagh and Bahl.  11.  R.E. Feeney.  12.  See Chaps 12 and 13 i n Ref. 4.  13.  R.U.  14. 15.  G. Ashwell and A.G. Morell. Adv. Enzymol. 41, 99-128 (1974). D. Aminoff, W.C. B e l l , and W.G. Vorder Bruegge. Prog. C l i n . B i o l . Res. 23, 569-581 (1978).  16.  K.M.  17.  See, for example, H.C. Anderson. (1978), and Ref. 3.  18.  See, for example, A.D. Keith, M. Sharnoff, and G.E. Cohn. Biochim. Biophys. Acta. 300, 379-419 (1973), and other reviews c i t e d i n Ch. 1.2.  Lemieux.  Amer. S c i e n t i s t .  Chem. Soc. Rev.  Yamada and J . Pouyssegur.  62, 712-719 (1974).  7, 423-452 (1978).  Biochimie.  60, 1221-1233 (1978).  Ann. Rev. Biochem.  47, 359-383  - 13 -  19.  J.F.G. Vliegenthart, H. van Halbeek, and L. Dorland. Chem. 53, 45-77 (1981).  Pure Appl.  20.  S. Schreler, C F . Polnaszek, and I . C P . Smith. Acta. 515, 375-436 (1978).  21.  A. Allerhand, K. D i l l , and W.J. Goux. NMR Biochem., Symp. Edited by S.J. Opella and P.Lu. Marcel Dekker, New York. 1978 (Pub. 1979). pp 31-50.  22.  J.D. Aplin and J.C. Wriston, J r . CRC C r i t . Rev. Biochem. 10, 259-306 (1981).  Biochim. Biophys.  - 14 -  CHAPTER 1.2 ELECTRON SPIN RESONANCE STUDIES ON SPIN LABELLED GLYCOCONJUGATES 1.2.1  The ESR Spin Label Method E l e c t r o n S p i n Resonance* (ESR) i s an i m p o r t a n t  for  spectroscopic  tool  t h e study o f b i o l o g i c a l systems; some systems n a t u r a l l y have  unpaired  e l e c t r o n s (e.g. those h a v i n g  some t r a n s i t i o n m e t a l s ) ,  while  o t h e r r e q u i r e i n c o r p o r a t i o n o f a paramagnetic r e p o r t e r group such as a "spin l a b e l " .  ESR s p i n l a b e l l i n g s t u d i e s hae c o n t r i b u t e d enormously t o  the u n d e r s t a n d i n g and  o f such b i o l o g i c a l systems as membranes and enzymes,  a number o f e x c e l l e n t r e v i e w s and monographs cover many a s p e c t s of  the t h e o r y , a p p l i c a t i o n s and p r a c t i c a l problems a s s o c i a t e d w i t h t h e technique.  1 - 1 1  S p i n l a b e l s can be c o n s i d e r e d  as p r o b e s , o r r e p o r t e r  groups d e s i g n e d t o be s p e c i f i c a l l y p l a c e d I n a m o l e c u l e , whose s p e c t r a would d e t a i l t h e environment o f t h e probe and, h o p e f u l l y , t h a t o f t h e  12 13 molecule t o which I t i s attached.  E a r l y work i n our l a b o r a t o r y  '  s e n s i t i z e d us t o t h e many problems and d i f f i c u l t i e s a s s o c i a t e d w i t h t h e approach.  Discussions  these i n the context  l a t e r i n t h e chapter w i l l h i g h - l i g h t some o f  of t h i s  study.  S p i n l a b e l s have the b a s i c s t r u c t u r e g i v e n below, where R j and R are chosen t o p r o t e c t t h e n i t r o x i d e f u n c t i o n a l i t y  from  d i s p r o p o r t i o n a t i o n , and p r o v i d e a means o f attachment t o t h e m o l e c u l e under  study. * A l s o c a l l e d E l e c t r o n P a r a m a g n e t i c Resonance (EPR).  2  -  -  15  o The basic resonance equation defines the large interaction of the electron spin with the laboratory magnetic f i e l d , and provides the central resonance  frequency:hv - g  0  B B  0  [1.2.1]  h i s Planck's constant and v the microwave frequency of the ESR spectrometer (of the order of G H z ) .  B i s the a constant for the  electron, the Bohr magneton, Bn i s the applied magnetic f i e l d (e.g.  0.34  T) and go i s the i s o t r o p i c g-value of the system. The unpaired electron i n a spin label i s largely confined to the N-0 group, and largely resides i n a it-orbital on the nitrogen.  In that  N has a nuclear spin quantum number, I_ » 1, i t w i l l have three nuclear spin quantum states (m • 1, 0, -1) with which the unpaired electron interact.  may  This i s called the electron-nuclear hyperfine i n t e r a c t i o n ,  and results i n a three-line spectrum being observed, with the lines having equal i n t e n s i t y and separated by the hyperfine coupling, a (ca.  0.17  mT).  F i g . 1.2.1  shows a t y p i c a l ESR spectrum of a nitroxide  tumbling rapidly In solution, and shows the measurements of g„, a and peak-to-peak l i n e widths,  WQ»  n  Hi * Ji-1 (corresponding to the three an  <  -  labelled transitions).  16  -  The recorded lines are f i r s t - d e r i v a t i v e and  require integration to y i e l d absorption spectra. The spectrum i n F i g . 1.2.1 values of a_ and g.  represents i s o t r o p i c a l l y averaged  By "doping" a c r y s t a l of the diamagnetic analogue  (reduced) of a spin label with the paramagnetic molecule and orienting i t i n the applied magnetic f i e l d , an angular dependence i s observed i n the g-value and coupling constant (a). The extremes i n g and ji l i e  Fig.  1.2.1.  •i  o  - - W.,  1 -Wo  -| -  -W  H  Solution ESR spectrum of a rapidly tumbling nitroxide spin label i l l u s t r a t i n g the measurement of parameters. (From Ref. 10).  along the p r i n c i p a l molecular x, y and z axes, where the x-axis l i e s p a r a l l e l to the N-0 orbital.  bond, and the z-axis along the nitrogen 2p  (ft)  For each nitroxide, the three values of g and a_ can be  measured; a > a , a , and a * ' —zz —xx —yy' —xx —yy The p r i n c i p a l uses of nitroxide spin labels pertinent to this  - 17 -  chapter l i e i n their a b i l i t y to report on (a) p o l a r i t y (by means of a_ ) and (b) molecular motion (by mean of l i n e widths and heights). u  s o l u t i o n , the nitroxide can be considered  In  to be made up of the following  resonance forms, with the zwitterion species favoured i n a polar environment:N  0*  «  •  +N*  0"  F i g . 1.2.2 i l l u s t r a t e s the basic changes i n line-shapes  as the motion of  a spin l a b e l i s r e s t r i c t e d . We have seen that a f r e e l y tumbling nitroxide spin l a b e l i n a non-viscous solvent displays a symmetrical pattern of three l i n e s of equal i n t e n s i t y ( F i g . 1.2.1).  As the motion  i s further r e s t r i c t e d (here, by increasing the solvent v i s c o s i t y ) i t i s seen that f i r s t the h i g h - f i e l d l i n e broadens, then the low-field and, f i n a l l y , the central l i n e . appear at frequencies  A d d i t i o n a l l y , spectral features begin to  distant from the o r i g i n a l ones ("outer extrema").  Ultimately, a "glass" spectrum i s observed, and any further r e s t r i c t i o n of motion w i l l not be apparent i n the spectrum:  the ESR spin l a b e l l i n g  technique s e l e c t i v e l y reports on motions ("correlation times") between 8  11  10"  1  and 10" s rad~  - a useful range for most b i o l o g i c a l studies.  The derivation of accurate c o r r e l a t i o n times from such spectra i s difficult,  and only t o t a l l y convincing when substantiated  simulations.  by spectral  However, an approximate value i s often a l l that i s  necessary when two spectra are to be compared where the introduction of a pertubatlon  induces large changes i n probe mobility.  For such cases,  12  s i m p l i f i e d equations may be derived chapter:-  and these have been used i n t h i s  - 18 -  *f<sk  */N^T  O  ,  - 1 0 0 ' C - ' ^  -36'C  0°C  ,  5 0 G  -  /  >  k  Mobility  T/S  - ^ _ _ _ ^ -  (rigid) Strongly immobilized  —S\j  8x 10"  i  — Weak! > immobi lized  —  -\\\——  W  -1  2xl0-«  -Mr—  Moderately immobilized  F i g . 1.2.2.  Approx.  • °  Freely tumbling  SxIO * -  8 x 10 "  1 0  5 x 10" "  Solution ESR spectra of a nitroxide spin label as a function of molecular motion, controlled by a l t e r i n g solvent v i s c o s i t y . (Most mobile, bottom spectrum; least mobile, top spectrum. From Ref. 3).  -10 * 6.6 * 10  1/2  P.]  - 1  [1.2.2]  - 19 -  - W*  i  c  -10 6.8 * 10  -.1/2 [1.2.3]  -0  -10 T - W~ * 6.7 * 10 c —0  - i 1/2  +  - 2  [1.2.4]  Generally, x was calculated by a l l three equations and the average c  reported as the c o r r e l a t i o n time. Aside from i t s high signal-to-noise s e n s i t i v i t y —5 between 10  (concentrations  —7 and 10  M are the lower l i m i t of d e t e c t a b i l i t y , depending  on the l i n e broadening) a r e f l e c t i o n of motion and p o l a r i t y , nitroxide spin labels have the further benefit that the spin concentration may be determined.  Double integration of the spectrum i s necessary, as i t i s  the area under the absorption peaks which varies l i n e a r l y with concentration.  By knowing the spin concentration,  the e f f i c a c y of a chemical  coupling procedure may be assessed. As mentioned previously, the spin label must provide the means for attachment to the b i o l o g i c a l substrate.  Although several basic  skeletons e x i s t , the studies i n this chapter involve the spin label based on the 2,2,6,6-tetramethylpiperidine-N-oxyl substituents at C-4.  structure, bearing  - 20 -  H  x  NH  2  O 1  With R representing a secondary amine, the molecule i s 2,2,6,6-tetramethylpiperidine-4-amino-N-oxyl (1)*. 1.2.2  Spin L a b e l l i n g of Glycoconjugates In this chapter a modest start i s made towards answering the  questions of how carbohydrates attached to proteins reorient i n solution. 14 To i l l u s t r a t e  the potential of s i t e - s p e c i f i c spin l a b e l l i n g of  the carbohydrate portion of glycoconjugates, we chose a plasma glycoprotein, f e t u i n  1 5  (isolated from foetal calf serum), a mucin,  16  bovine submaxillary mucin (BSM, isolated from boxine submaxillary glands), and human erythrocytes (red blood c e l l s = RBC's).  In a l l  cases, the goal was the l a b e l l i n g of the glycoconjugate sugar termini. A l l three materials are known to be r i c h i n s i a l i c acids (N-acetyl neuraminic acid = NANA), usually at the "non-reducing" terminus, attached to gal or galNAc.  Fetuin bears O-glycosides and  *This i s often called "TEMPAMINE" and indicated i n synthetic pathways as SL-NH2.  - 21 -  NANA  NANA  NANA  Gal  Gal  Gal  GlcNAc  GlcNAc GlcNAc j B l + 3 ( 4 ) JB1*2(4,6)  Ja2-»-3  ja2-»3  JB1-M  JB1+4  NANA  Ja2->-3  Gal  JB1+4  NANAa(2-*6) GalNAc Jo  Mana(l-»-2 (6) ) Mana(l-+3) Man (l->4) Gl cNAcB (l-*4) Gl cNAcBAsn  F i g . 1.2.3.  0- and N-glycosides of f e t u i n .  N-glycosides of the form depicted  i n F i g . 1.2.3;  17  Ser (Thr)  (From Ref. 17).  the outer  trisaccharide structure, NANA-a-(2->3)-gal-8-(l-»'4)-glcNAc, i s common i n animal glycoproteins.  The mucins are extensively s i a l a t e d , and the  following sequence occurs several hundred times i n each m o l e c u l e : NANA-a-(2->-6)-galNAc-a-(l-»- )-Ser-(or Thr). may  17  The exocyclic t r i o l of NANA  be fJ-acylated at one or more p o s i t i o n .  RBC's obviously present a  highly complex and varied carbohydrate code to i n t e r a c t i n g species.  A  18 major RBC membrane protein i s glycophorin A, s i a l a t e d trans-membrane protein bearing the MN blood-group c l a s s i f i c a t i o n .  which i s a highly  the carbohydrate determinant for  Although some doubt s t i l l exists as  to the detailed structure of the carbohydrate moieties,  the terminal  sequence of NANA-a-(2-*x)-gal i s known (x = 3 or 6 ) . I t has been mentioned that asialoglycoproteins* are an important sub-set of the sialoglycoproteins, as the terminal galactose marks a plasma protein for c i r c u l a t o r y clearance  and catabolism  by the l i v e r .  With glycophorin A, terminal s i a l i c acid i s e s s e n t i a l for the e x h i b i t i o n •Sialoglycoproteins having been desialated i n some way.  - 22 -  of MN blood-group a c t i v i t y .  Thus, we recognize the need for the  l a b e l l i n g of NANA termini of sialoglycoproteins, and galactose termini of the a s i a l o analogues, i f their roles at the molecular  l e v e l are to be  monitored. Precedent e x i s t s i n the l i t e r a t u r e for the generation of an aldehyde group either at NANA to be p o t e n t i a l l y useful.  19  or g a l  2 0  termini and we considered  The reductive amination  this  procedure had been  21 used to chemically modify protein l y s y l  6-NH2 groups  with methanal,  using NaBH^ to reduce the intermediate Schiff base; Borch et a l . described  22  a superior reductant for this purpose, sodium  cyanoborohydride (NaCNBH ). 3  It was determined by Van Lenten and Ashwell  that the exocyclic  t r i o l of NANA may be s e l e c t i v e l y oxidized by periodate under mild conditions.*  Strong oxidative conditions could result i n the oxidative  cleavage of a l l v i c i n a l d i o l s , thus destroying the i n t e g r i t y of the  •Periodate i s c l a s s i c a l l y used i n the "Smith degradation" to determine intersugar linkages i n a p o l y s a c c h a r i d e . Its r e a c t i v i t y towards v i c i n a l d i o l s varies i n the order: e x t r a c y c l i c > i n t r a c y c l i c c i s - > i n t r a c y c l i c trans23  carbohydrate chain and the s p e c i f i c i t y of the reaction.  The procedure  o r i g i n a l l y called for the addition of 10 moles of NalOi, per mole of s i a l i c acid, and reaction for 10 min at 0°C and pH 5.6. Later, Jourdian 24  et a l . quantified  the reaction and i t i s clear from their data that  reaction for 35 min at 0°C y i e l d s good s e l e c t i v i t y and oxidized NANA for isolated glycoproteins.  With RBC's, the t o t a l NANA content i s d i f f i c u l t  to estimate, and the reaction time i s accordingly reduced to 15 min. to l i m i t possible oxidation of terminal gal residues (which are the most susceptible to cleavage, a f t e r the NANA exocyclic t r i o l ) . The oxidation of galactose i s best performed enzymatlcally. Galactose  oxidase  (E.C. No. 1.1.3.9) e f f e c t s the oxidation of the  6-CH OH to an aldehyde ( F i g . 1.2.4). 2  gal or galNAc residues, and H 0 2  2  The enzyme w i l l act on terminal  i s produced.  Although not done i n this  spin-label study, i t i s advisable to remove the H 0 2  2  as i t i s formed  (with a second enzyme, catalase), as this i n h i b i t s the forward reaction of the enzyme by the c l a s s i c a l feed-back control mechanism.  F i g . 1.2.4.  Oxidation of galactose termini by galactose  oxidase.  - 24 -  To expose subtenninal galactose residues, i t i s necessary that NANA be removed; t h i s may be effected by mild acid hydrolysis, or enzymatically using an exoglycosidase s p e c i f i c to NANA - neuraminidase (B.C. No. 3.2.1.18). Having generated aldehyde f u n c t i o n a l i t i e s on NANA or gal termini, the spin l a b e l TEMPAMINE (1) was reductively aminated onto this point, using  NaCNBH3.  Unreacted, low molecular weight components were  separated from the glycoprotein by size-exclusion  chromatography  (Sephadex G-25), and by exhaustive d i a l y s i s i n the case of RBC's.  —C.  P  +  "  F i g . 1.2.5.  NH  =  2  SL  I" ~1 [-C=N—SLj +  H  2  NaCNBH,  • . • —CK2-NH-SL I  U  0  Reductive amination of an aldehyde with TEMPAMINE, using NaCNBH . 3  Major advantages of the use of  NaCNBH3  over NaBHi* include i t s resistance  to aqueous hydrolysis and s p e c i f i c i t y for the intermediate Schiff base over the s t a r t i n g aldehyde.* * NaBH^ w i l l also reduce the aldehyde to a primary alcohol.  c  i  - 25 -  1.2.3  Results and Discussion Fetuin, BSM  and RBC's were spin labeled at t h e i r NANA termini  using the above-mentioned procedure  based on s p e c i f i c activation under  mild periodate oxidative conditions, and coupling of the amine spin l a b e l , TEMPAMINE, by reductive araination using NaCNBH . 3  Subterminal  galactose residues of fetuin and BSM were exposed by neuraminidase treatment, and oxidized by galactose oxidase/catalase action. was  TEMPAMINE  then attached as before. Representative ESR  spectra are given i n F i g . 1.2.6.  the l e v e l of incorporation was  In a l l cases  high enough to e a s i l y detect an  ESR  spectrum with milligram quantities of glycoprotein, and a p a r t i a l l y immobilized  spin l a b e l was  indicated.  Amplification of the h i g h - f i e l d  region (outer extrema) did not reveal any signal from a r i g i d component to any spectrum.  Motional c o r r e l a t i o n times were determined from peak  heights and widths using Eqns. 1.2.2-4.  Isotropic hyperfine couplings,  a_ , were measured from spectra, and the percentage d e r i v a t i z a t i o n Q  determined by double integration of the ESR  signal.  The relevant data  are tabulated (Table 1.2.1). Control experiments involved the incubation of NH -SL and 2  NaCNBH3 with the unactivated glycoprotein, and Sephadex separation. Non-specific adsorption was signal and was  found to contribute < 0.5%  not pH dependent.  of the t o t a l  With the erythrocytes, spin l a b e l l i n g  In p h y s i o l o g i c a l saline resulted i n a minimum of c e l l l y s i s , and spin l a b e l s were found exclusively i n glycoconjugates the c e l l wall, a f t e r RBC  "ghosts" were prepared.  the  associated with  The coupling  reagents  - 26 -  Fig.  1.2.6.  ESR s p e c t r a of n i t r o x i d e - l a b e l l e d m a t e r i a l s . Concentration and r e l a t i v e g a i n a r e g i v e n i n p a r e n t h e s e s . A. P e r i o d a t e - a c t i v a t e d f e t u i n ( 3 . 2 4 mg i n 5 0 u l H 0 ; g a i n 1 ) . B. G a l a c t o s e o x i d a s e - a c t i v a t e d a s i a l o - B S M ( 2 . 6 7 u l i n 50 u l H2O; gain 6.4). C. P e r i o d a t e - a c t i v a t e d , intact e r y t h r o c y t e s (73 u l packed c e l l s i n phosphate-buffered saline; gain 187.5). D. F e t u i n p e r f u s e d w i t h 1 a n d NaCNBH3 2  for  2 h  (3.00  mg i n  50  ul  H 0; 2  gain  62.5).  (From Ref.  14.)  Table 1.2.1  Parameters characterizing spin label probes attached to various substrates. Percentage of possible sites spin l a b e l l e d . N.D.: not determined due to low signal-to-noise.  a  Activation Material  Procedure  T  c  (ns)  (±0.05 ns)  a^ (mT) (± 0.02 mT)  % Activation  Number spins  (±5%)  per Molecule  FETUIN  NalO^, 0°C, 35 min.  0.79  1.68  ca. 20  2.5  ASIALOFETUIN  Neuraminidase, galactose oxidase  0.52  1.68  50  6.0  BSM  NalCH, 0°C 35 min.  0.35  1.68  33  115.5  ASIALO-BSM  Neuraminidase, galactose oxidase  0.49  1.67  RBC  NalO^, 0°C 15 rain.  1.0  1.65  7.0  N.D.  N.D.  -  were  reacted  with  room-temperature extension  of  the -  this  Turning  no  to  indicate  a  what  might  expect,  to  highly  one  expected  be  label  correlation  times  interesting  to  and  BSM  quite the  label the  to  protein  T  the  pendant  c  in  labelling  for  2 h  at  efficiency  was  observed  by  ESR  a polar  as  the  sugar  a  the  amu)  from  Data  moderately  values,  labelled that  mobility  residue, are  obtained  by  This  sugar to  the  times at the  it  of  is  would  medium.  The  could  mobile H NMR  spin  of  the  label  be  be  and  it  000  termini  are  is  amu)  contribution joining  that,  in  respect  (Chapter  with  (MW 4 8  the  bonds  with  mT,  residues  sugar  bulk  1.67  consistent  fetuin  their  about  or  equally  around  immobilized  correlation  the  residues  exposed  suggest  sugar  0  environment.  and  spin  may  a_  the  (hydrophilic)  solvated  This  stems  data,  1.3)  the  both  to  to spin  cases,  the  support  the  Information Derived from Spin Label Studies of Glycoconjugates studies  procedures  and  additional  mobility or  indicate  may b e  discussed.  environment  the  6  terminal  glycoconjugates  bonds,  glycoconjugate  hypothesis.  Our  These  in  that  10  back-bone.  second  1.2.4  *  comparable.  observed  the  indicate  note  (MW 1 . 3  increase  -  period.  now  spin  activated  28  of  the  could  nitroxide  is  The  mobile  pendant  spin  NANA  efficiently  motional  be  that  spin  termini and  exists  with  the  in  sugars  to  by the  resulting motion sugar  from  about  of  spin  complex  labelled  could  terminus.  as  reflect  rotation the  using  a hydroxylated,  glycoprotein  degrees-of-freedom  dominated label  gal  activated  label  compared  or  a  chemical  polar  whole.  the  about  the  enhanced glycosidic  bonds  linking  - 29 -  Since the work described here was completed, others found application for this and related procedures to study the role of sugars In  binding events.  A discussion of these papers i s i n order as they  serve to highlight some of the d i f f i c u l t i e s with which the technique i s fraught. Grant and co-workers have performed an extensive study  •  of  the headgroup dynamics and carbohydrate role i n l e c t i n binding processes with the transmembrane glycoprotein, glycophorin A (vide supra). When terminal NANA residues are s p e c i f i c a l l y spin-labelled, the probe i s confined to a highly polar milieu, and the motion i s that of a moderately immobilized spin, even when constituted into a model membrane.  That which reduces NANA mobility (as reflected i n the ESR  c o r r e l a t i o n time) i s a s p e c i f i c binding event.  Wheat-germ l e c t i n (WGA)  binds to glycophorin A, and the terminal NANA has been implicated i n the binding process.  As WGA i s added, a dramatic increase i n  which i s a function of the amount of l e c t i n added.  i s observed  This study was  27  extended to gangliosides,  labelled i n an analogous fashion.  Here,  upon addition of WGA an i n i t i a l decrease i n T was observed, followed by £  (the  expected) increase as the WGA concentration was further raised.  Such observations were explained i n terms of randomizing an i n i t i a l patch-wise d i s t r i b u t i o n of gangliosides, although this was not p o s i t i v e l y determined as i t can be with spin l a b e l s .  28  29 30  Studies on a similar system by B u t t e r f i e l d and co-workers are  In contradiction to those of Grant.  These workers used  cyanoborohydride to couple spin label with periodate-activated  »  - 30 -  erythrocyte ghosts, which was found to be more e f f i c i e n t than NaBR\ (used by Grant).  Upon addition of l e c t i n s which bind glycophorin A (WGA  and Phaseolus vulgaris agglutinin, PHA) an increase i n mobility was 30  observed exposing binding.  which was interpreted i n terms of conformational  changes  the spin l a b e l to a less r e s t r i c t e d environment upon l e c t i n I t i s important  to note that these two sets of data are not  s t r i c t l y comparable, since Lee and Grant worked with a s p e c i f i c a l l y labelled glycophorin either i n solution or a model membrane, while B u t t e r f i e l d and co-workers used intact erythrocyte ghost membranes, where several questions a r i s i n g from the complexity raised.  A third study  31  of the system can be  of WGA binding to spin labeled erythrocytes i s  in accord with the findings of Lee and Grant; special pains were taken to purify the commercial l e c t i n s , and the a c t i v a t i o n was of the intact erythrocytes r e s u l t i n g i n very low degree of ganglioside l a b e l l i n g (cf. Butterfield). 32  F i n a l l y on t h i s subject, a French study  was concerned with the  f l e x i b i l i t y of isolated branched glycopeptides; NANA was spin labelled by periodate oxidation and reductive amination.  Concanavalin  A* was  added resulting i n a s l i g h t immobilization of the spin label - t h i s i s not altogether s u r p r i s i n g , as the binding s i t e for the l e c t i n i s distant from the labelled termini.  However, I t does suggest a certain r i g i d i t y  of the oligosaccharide chain i n i t s a b i l i t y to propagate such e f f e c t s . Since the inception of the studies presented  i n this chapter, two other  *A l e c t i n with a s p e c i f i c i t y for mannose, present i n the glycan.  - 31 -  groups have reported on the use of similar methods to spin label the carbohydrate moiety of immunoglobulins, and their f r a c t i o n s .  Both used  33  non-specific oxidation - 8 hours at room temperature, at the  room temperature or 0° C .  3 1 t  »  3 5  and 16-24 hours  One group e x p l i c i t l y states that  s e l e c t i v i t y for NANA was only 60%, and i t i s clear from the  pertinent l i t e r a t u r e reports that the l a b e l l i n g cannot be selective i n either case.  The Russian group i n i t i a l l y  reported a highly mobile spin  34  label  and then published revised data showing a more immobilized spin q c  label;  the i n i t i a l error was caused through carry over of  n o n - s p e c i f i c a l l y adsorbed l a b e l . The facts emerging from this chapter are that (a) NANA or galactose termini of glycoproteins and erythrocytes can be s p e c i f i c a l l y labelled (b) the spin labels are mobile r e l a t i v e to the protein as a whole, and i n a polar environment, and (c) extreme caution should be exercised when drawing conclusions from a complex system, when b i o l o g i c a l perturbants are introduced.  That i s not to say that the spin  l a b e l method i s without use, but the experiments should be very c a r e f u l l y designed and controlled.  - 32 -  REFERENCES 1.  I.CP. Smith. B i o l o g i c a l applications of electron spin resonance. Edited by H.M. Schwartz, J.R. Bolton, and D.C. Borg. John Wiley, New York. 1972. pp. 483-539.  2.  A.D. Keith, M. Sharnoff, and G.E. Cohn. 300, 379-419 (1973).  3.  R.A. Dwek. Nuclear magnetic resonance i n biochemistry: applications to enzyme systems. Clarendon Press, Oxford. Chapter 12.  Biochim. Biophys. Acta.  1975.  4.  G.I. Likhtenstein. Spin labeling methods i n molecular biology. John Wiley, New York. 1976.  5.  B.J. Gaffney and D.C. L i n . The enzymes of b i o l o g i c a l membranes. Vol. 1. Edited by A. Martonosi. Plenum Press, New York. 1976. pp 71-90.  6.  P.M. Vignais and P.F. Devaux.  7.  Spin labeling: theory and applications. Academic Press, New York. 1976.  8.  Spin labeling I I : theory and applications. B e r l i n e r . Academic Press, New York. 1979.  9.  S. Schreier, C. Polnaszek, and I.CP. Smith. Acta. 515, 375-436 (1978).  Ibid, pp. 91-117. Edited by L.J. Berliner. Edited by L . J .  Methods Enzymol.  Biochim. Biophys.  10.  P.C. Jost and O.H. G r i f f i t h .  11.  L.J. B e r l i n e r .  12.  J.D. Aplin. PhD. Thesis. and references therein.  13.  M. Yalpani. Ph.D. Thesis. and references therein.  14.  J.D. Aplin, M.A. Bernstein, C.F.A. C u l l i n g , L.D. H a l l , and P.E. Reld. Carbohydr. Res. 70, C9-C12 (1979).  15.  E.R.B. Graham. Glycoproteins: t h e i r composition, structure and function. Edited b y A. Gottschalk. E l s e v i e r , New York. 1972. pp. 716-731.  16.  W. Pigman and A. Gottschalk. Ibid, pp. 817-821.  Methods Enzymol.  49, 369-418 (1978),  49, 418-480 (1978).  University of B r i t i s h Columbia. University of B r i t i s h Columbia.  1979, 1980,  - 33 -  17.  N. Sharon. Complex carbohydrates: their chemistry, biosynthesis, and functions. Addison-Wesley, Reading, Mas. 1975.  18.  H. Furthmayr. Biology of carbohydrates, V o l . 1. Edited by V. Ginsburg and P. Robbins. John Wiley, New York. 1981. Chapter 4.  19.  L. Van Lenten and G. Ashwell. Methods Enzymol. (1972), and references therein.  20.  A.G. Morell and G. Ashwell. therein.  21.  G.E. Means and R.E. Feeney. Chemical Modification of Proteins. Holden-Day, San Francisco. 1971.  22.  R.F. Borch, M.D. Bernstein, and H.D. Durst. 93, 2879-2904 (1971).  23.  P.V. Wagh and O.P. Bahl. C.R.C. C r i t . Rev. Biochem. (1981), and references therein.  24.  G.W. Jourdian, L. Dean, and S. Roseman. 430-435 (1971).  25.  P.M. Lee and C.W.M. Grant. 856-863 (1979).  26.  P.M. Lee and C.W.M. Grant. Can. J . Biochem.  27.  P.M. Lee, N.V. K e t i s , K.R. Barber, and C.W.M. Grant. Biophys. Acta. 601, 302-314 (1980).  28.  L.D. H a l l and J.C. Waterton. (1979).  29.  J.B. Felx and D.A. B u t t e r f i e l d .  30.  J.B. Feix, L.L. Green, and D.A. B u t t e r f i e l d . 1001-1009 (1982).  31.  R. Snoek.  32.  J . Devoust, V. Michel, G. Spik, J . Montreuil, and P.F. Devaux. FEBS L e t t . 125, 271-276 (1981).  33.  K.J. Willan, B. Golding, D. G i v o l , and R.A. Dwek. 133-136 (1977).  34.  R.S. Nezlln, V.P. Timofeev, Y.K. Sykulev, and S.E. Zurabyan. Immunochem. 15, 143-144 (1978).  35.  V.P. Timofeev, I.V. Dudich, Y.K. Sykulev, and R.S. Nezlln. L e t t . 89, 191-195 (1978).  Personal  Ibid.  28, 209-211  205-208, and references  J . Amer. Chem. Soc.  10, 307-377  J . B i o l . Chem. 246,  Biochem. Biophys. Res. Comm. 90, 58, 1197-1205 (1980). Biochim.  J . Amer. Chem. Soc. 101, 3697-3698 FEBS L e t t .  115, 185-188 (1980). L i f e Sciences. 31,  communication.  FEBS L e t t . 80,  FEBS  CHAPTER 1.3 DEUTERIUM MAGNETIC RESONANCE STUDIES ON GLYCOCONJUGATES  Chapter 1.2 i n this thesis detailed ESR stdies on the carbohydrate moiety of glycoproteins, which provided information on the r e l a t i v e mobility of the glycan.  Although the ESR method has advantages  - c h i e f l y , s e n s i t i v i t y and the absence of "background"  signals from the  molecule i t s e l f - the physical bulk of the nitroxide probe i t s e l f may perturb the system and lead to erroneous conclusions.  In addition,  where one i s interested i n the motional correlation time ( T ) of a c  sugar to which the spin label i s attached, the measured value may largely r e f l e c t a high degree of mobility of the probe about the bonds attaching i t to the sugar, rather than the sugar per se. Alternatives exist for the extraction of motional information from such systems - preferably, these should be "non-invasive" i n that they do not introduce an extraneous bulky "reporter" group. l i g h t , NMR relaxation studies are the most promising.  In this  To simplify  matters, i t i s desirable to attach a "probe" i s o t o p i c a l l y enriched i n a nuclide with low natural abundance, as such a procedure w i l l obviate complications a r i s i n g from overlap with signals from the system ("back-ground"). values for T  c  Enrichment i n C i s one possible approach; numerical 1 3  may be calculated once the relaxation behaviour and the  chemical system are known. An alternative approach i s *H NMR; substitution of a proton with a deuteron i s c l e a r l y a minor perturbation to the system.  Deuterium,  having nuclear spin, I_ • 1, i s quadrupolar, meaning that i t s nuclear  - 35 -  charge i s asymmetrically  distributed about the nucleus.  This leads to  some broadening of the resonances but, most importantly to this discussion, the relaxation mechanism for the deuteron i s dominated by this quadrupole e f f e c t , and a s i m p l i f i c a t i o n of the determination of motional  information r e s u l t s .  Working i n i t s disfavour i s the low  2  1  inherent s e n s i t i v i t y of H (< 1% that of H):  this factor can be  p a r t i a l l y offset by the use of high magnetic f i e l d s ( B 9.4 T i n this 0  case), and the Fourier transform method, now so ubiquitous i n heteronuclear NMR studies. o  The application of high-resolution H NMR (and some broad-line studies) was exhaustively reviewed by Mantsch et a l . i n 1977, and 1  2  up-dated i n 1982.  These two treatises are mandatory reading for the  2  applicant of H NMR, and the reader wishing a broader view on the subject should consult these.  In the carbohydrate  area, a l l related  reports have been with i s o t o p i c a l l y enriched monosaccharides. 3  In 1973,  13  Brewer et^ a l . studied  the binding of  C-enriched  oc-methyl-D-glucopyranoside to the l e c t i n , * Concanavalin  A.  Their  mapping of the binding s i t e was found to d i f f e r considerably from previous X-ray studies, where an iodophenyl  glucoside was used.  Few  challenge the v a l i d i t y of the NMR data, and this example amply i l l u s trates the d i f f i c u l t i e s which may arise when large reporter groups are i  used.  A l t e r and Magnuson used  Q  l»  F NMR to study the same system, and  Neurohr et a l . s t u d i e d the binding of i s o t o p i c a l l y labelled methyl o*A l e c t i n i s a protein which binds s p e c i f i c carbohydrate(s). 5  - 36 -  8-D-galactopyranosides to peanut agglutinin (another l e c t i n ) by 13C  and  NMR relaxation.  The same group  6  studied the molecular dynamics of the  sugar binding process to the l e c t i n , wheat germ agglutinin.  In this  case, the sugar was i s o t o p i c a l l y enriched i n H and this s a t i s f y i n g relaxation study includes an analysis for the case where a contribution to the motion from i n t e r n a l rotation i s present i n the bound state. In the case of rapid (where to T < 1) and i s o t r o p i c motion, the o c dominant quadrupole relaxation mechanism i s given by: 1  (Ii)  -1  (T r  =  —  2  2  -J 8  2 2 (1 + U_) (i-39.) 3 h  2  [1.3.1]  T  c  n i s the asymmetry parameter of the f i e l d gradient (and may be neglected), and e qQ/h i s the quadrupole coupling constant.  Although T_i  or T_2 could be measured by multi-pulse methods, i t i s j u s t i f i a b l e to neglect contributions to the line-width from magnetic inhomogeneities and H- H couplings, and assume Tj, " l / i  2  peak width at half-height).  1 1  ^ 0.5 V  (where  AVQ 5 #  i s the  Equation 1.3.1 reduces to:  |(4^) x 2  (T2)"  By measuring  1  AVQ.S  - , Av  Q > 5  -  c  H.3.2]  and using a reasonable value for the quadrupole  coupling constant, x  c  may be e a s i l y calculated.  In Chapter 1.2 we described two methods by which aldehyde f u n c t i o n a l i t i e s may be introduced at s p e c i f i c l o c i i n a glycoprotein glycan.  Our approach to the s p e c i f i c incorporation of H into  glycoprotein rests on the use of these same procedures to generate the  -  37  -  s p e c i f i c aldehyde, and the regeneration of an alcohol using the ^ - l a b e l l e d reductant, sodium borodeuteride. The s p e c i f i c deuteration of terminal s i a l i c acid or galactosyl units i s i l l u s t r a t e d i n Figs. 1.3.1  and 1 . 3 . 2 , respectively.  With s i a l i c acid, the sugar i s modified  i n that the exocyclic t r i o l i s shortened by two carbons; with galactose, the labelled sugar matches the native sugar almost exactly. exocyclic -CHDOH unit  The  i s racemic, and t h i s , too, w i l l contribute to the  observed line-width. In l i g h t of our previous successes with the reductive amination procedure using sodium cyanoborohydride, i t was apparent that proteins could be l a b e l l e d with deuterons by performing such a reaction with hexadeuterioacetone and l y s y l 6-amino groups:  6-NHU+ 0=C  / C D  3  NaCNBH  3  6-N1C1H \ | or 2\  CD*  The above-mentioned  H  CD  3  procedures were used to l a b e l the s i a l i c  acid termini of bovine submaxillary mucin ( B S M ) , galactose termini of a s i a l o f e t u i n and l y s y l groups of f e t u i n . recorded at 6 1 . 4 MHz  The  H NMR spectra were  ( B Q 9.4 T) as solutions i n d i s t i l l e d water, and  are shown i n F i g . I . 3 . 3 . *  The p a r t i a l overlap of the sugar signal(s)  *Spectra were recorded by Dr. W.E.  H u l l of Bruker, West Germany.  - 38 -  F i g . 1.3.1  F i g . 1.3.2.  Procedure for the s p e c i f i c deuteration of s i a l i c acid termini of glycoproteins at the C-7 p o s i t i o n .  Procedure for the s p e c i f i c deuteration of galactose termini of asialo-glycoproteins at the C-6 p o s i t i o n .  - 39 -  F i g . 1.3.3.  H NMR spectra measurement at 61.4 MHz. Samples were dissolved i n = 1.5 ml H 2 O . (A) BSM deuterated'at C-7 on s i a l i c acid residues (145 mg; 2700 transients). (B) A s i a l o f e t u i n deuterated at C-6 position on terminal galactose residues (190 mg). (C) Fetuin reductively aminated with hexadeuterioacetone on lysine residues (240 mg; 550 transients). The HOD resonance was a r b i t r a r i l y assigned a s h i f t of 5 ppm. From r e f . 7.  - 40 -  with the residual HOD was  i n d i s t i l l e d water i s undesirable and the problem  largely a l l e v i a t e d by using deuterium-depleted  water as a solvent  (data not shown). For the determination of T » line-widths were measured and c  inserted i n Equation [1.3.2].  A typical  (165 kHz) was used throughout.  1  quadrupole  coupling constant  The correlation times from  H NMR  are  compared with those for the analogous molecule from ESR measurements (Ch. 1.2), in Table 1.3.1.  The quoted errors are that i n measuring peak  widths - especially d i f f i c u l t when the signal-to-noise i s low and overlap with the HOD  resonance i s s i g n i f i c a n t (as i n the case with F i g .  I.3.3.B). The c o r r e l a t i o n between the measurements i s good, and these data suggest that x  c  values derived from nitroxide ESR studies are  representative i n such systems. from the ESR,  H NMR  BSM  labelled at s i a l i c acid  "appears"  measurements to be roughly twice as mobile as from the  and this should serve to caution the reader not to take the figures  too l i t e r a l l y .  Indeed, this may  be a real e f f e c t , r e f l e c t i n g a somewhat  r e s t r i c t e d motion of the spin l a b e l with BSM,  where the glycan i s short  (a disaccharide), and the carbohydrate substitution l e v e l high.  As  mentioned i n the previous chapter, useful results may arise from  ESR  studies where changes i n T  C  resulting from the perturbation of the  system are considered, and this conviction i s supported by the literature. We conclude that, given ca. 100 milligrams of material, one  may  successfully label s p e c i f i c s i t e s on a glycoprotein with deuterons, and  - 41 -  Table 1.3.1  Substrate  Comparison of x determined from H NMR relaxation measurements, and nitroxide ESR. a. x i n ns. b. Reductive amination of protein l y s y l 6-amine groups with hexadeuterioacetone, or 2,2,6,6-tetramethylpiperidine -4-keto-N-oxyl (TEMPONE).  Activation Procedure  x  &  from H NMR 2  c  x  from ESR c  FETDIN  PERIODATE  0.73 ± 0.13  0.79  ASIALOFETUIN  GALACTOSE OXIDASE  0.57 ± 0.03  0.52  BSM  PERIODATE  0.12 ± 0.07  0.35  ASIALO-BSM  GALACTOSE OXIDASE  Not Determined  0.49  FETUIN  REDUCTIVE AMINATION  0.5  0.77  b  - 42 -  observe the  H NMR resonances on a h i g h - f i e l d spectrometer.  Measurement of line-widths  yields a correlation  time for the sugar very  s i m i l a r to that derived from an analogous nitroxide spin-label study. While this i s Interesting and, perhaps, useful Information, i t should be cautioned that i n neither study have we perturbed the system with, for example, a molecule which should bind the sugar ( i . e . a l e c t i n ) to evaluate the possible effect of the bulky spin l a b e l on the process.  It i s our b e l i e f that ^  binding  NMR studies of the kind detailed here  w i l l be useful i n understanding molecular interactions and properties of b i o l o g i c a l macromolecules.  physical  - 43 -  REFERENCES FOR CHAPTER 1.3 1.  H.H. Mantsch, H. Saito, and I.CP. Smith. Reson. Spectrosc. 11, 211-271 (1977).  Prog. Nuclear Magn.  2.  I.CP. Smith and H.H. Mantsch. NMR spectroscopy: new methods and applications. Edited by G.C Levy. ACS Symposium Series 191. 1982. pp. 97-117.  3.  C F . Brewer, H. S t e r n l i c h t , D.M. Marcus, and A.P. Grollman. Nat. Acad. S c i . USA. 70, 1007-1011 (1973).  4.  CM. A t l e r and J.A. Magnuson.  5.  K.J. Neurohr, N.M. Young, I.CP. Smith, and H.H. Mantsch. Biochemistry. 20, 3499-3504 (1981).  6.  K.J. Neurohr, N. L a c e l l e , H.H. Mantsch, and I.CP. Smith. Biophysical J . 32, 931-938 (1980).  7.  M.A. Bernstein, L.D. H a l l , and W.E. H u l l . 2744-2746 (1979).  Proc.  Biochemistry. 13, 4038-4045 (1974).  J . Amer. Chem. Soc.  101,  - 44 -  CHAPTER 1.4 NEOGLYCOPROTEINS: SYNTHESIS AND NMR STUDIES 1.4.1  Introduction At the time when the work described i n this chapter was  performed (1979/1980), much use had been made of C and *H NMR i n the 1 3  study of o l i g o - and polysaccharides (see Ch. II.3), but l i t t l e i n a second important  class of glycoconjugates,  glycoproteins (Ch. 1.2). At  natural abundance the scarcity of studies can, i n part, be attributed to the complexity  of the material and the lack of model compounds. In  addition to contending with protein resonances (overlapping with carbohydrate  peaks to a greater or lesser extent), glycoproteins often  display microheterogeneity  i n their glycans, making i t impossible to  write a single chemical structure for the carbohydrate added complexity  moiety.*  This  obviously compounds the d i f f i c u l t y of analysis.  Glycoproteins cannot always be isolated i n appreciable quantities, and t h i s could preclude One  1 3  C NMR studies.  approach to the development of simple model systems for NMR  studies i s to l i b e r a t e the glycans from the protein (or perform an enzymic protein d i g e s t ) , separate the fragments, and perform NMR experiments on these p u r i f i e d materials. microheterogeneity  This solves the problem of  and eliminates the bulk of the protein resonances,  and could be a worthwhile f i r s t approach.  We considered this to be  somewhat undesirable, as the low molecular weight fractions were too f a r  *An obvious exception to the rule Is the A r c t i c f i s h "antifreeze" glycoproteins, which are extremely homogeneous.  -  from the " r e a l " s i t u a t i o n .  45  -  An alternative exists and we have exploited  t h i s : known, characterized sugars can be covalently attached to a (hitherto sugarless) protein to create a "neoglycoprotein". protein resonances  The bulk of  can p o t e n t i a l l y be i d e n t i f i e d by observing the  protein before glycan attachment. the protein and carbohydrate  In essence, one can characterize both  independently of one another, and then  observe changes when the neoglycoprotein i s studied.  This allows one to  make assignments of neoglycoprotein glycans and evaluate such as the potential use of chemical s h i f t data of carbohydrate  concepts  standards.  With the purpose of spectral assignment we evaluated the use of enzymic (and, i n p r i n c i p a l , chemical) perturbations. We  f e l t that the neoglycoprotein approach had the further  advantage that the material i s of similar physical form to a glycoprotein, and this was  c r u c i a l i n our H NMR A  studies where use  was  made of the d i f f e r e n t i a l i n mobility between the glycan and protein back-bone.  The structure and complexity of the glycone would be  r e s t r i c t e d by the lengths an investigator i s prepared to go i n oligosaccharide synthesis*.  The neoglycoprotein synthetic protocol  allows also for f l e x i b i l i t y i n the nature and linkage length of the glycone to protein, and this could be useful to one interested i n d i f f e r e n t i a l mobility or enzymic studies (vide i n f r a ) . At this time, other groups have proceeded to successfully employ A3  C  NMR  to study a variety of aspects of the structure and form of  glycoprotein glycans, and these w i l l be reviewed  l a t e r i n the chapter.  *This i s , by no means, a t r i v i a l task! In this study, monosaccharides and purchasable disaccharides were considered.  -  1.4.2  46  -  Neoglycoprotein Synthesis The f i r s t neoglycoproteins were synthesized by Goebel and Avery  i n 1929, and their procedure i s s t i l l popular.  1  A sugar i s 0_ -  glycosylated with para-aminophenol, and the p-aminophenyl  glycoside  converted into the diazonium s a l t (nitrous acid, 0°, 15 min.).  This i s  then mixed with the protein, i n basic solution for another 15 min.; substitution primarily occurs at h i s t i d i n e , lysine and tyrosine amino acid l o c i .  This procedure has a number of major draw-backs:- the  s p e c i f i c i t y i s low, which can cause problems i n the interpretation of r e s u l t s , and the introduction of the bulky aromatic group can  itself  e l i c i t an antibody response, reducing the s p e c i f i c i t y of the antibody to the  sugar hapten. Literature precedent existed for the selective amination of  methanal by l y s y l amino acids, at the 6-amine position. »  Reductive  amination had been described by Gray and co-workers between l y s y l amines and reducing sugars reductant.  ( F i g . 1.4.1), using sodium cyanoborohydride as the  Although the reaction i s controllable and mild, i t has the  disadvantages that i t introduces a polyol (to which antibodies have been shown to d i r e c t ) , destroys one sugar ring (which i s disadvantageous when the  attached product i s the result of a long synthetic route) and the  R - Sugar F i g . 1.4.1.  Gray's procedure for the attachment of a reducing disaccharide to a protein (P) v i a i t s l y s y l 6-amino groups.  - 47  An encyclopaedic  -  review of methods described for neoglycoprotein  synthesis i s not i n order here - two excellent reviews have appeared recently. » b  6  Our survey of these methods for neoglycoprotein  synthesis did not point to an ideal procedure, compared with the c r i t e r i a we considered 1.  necessary:-  The method should be d i r e c t l y compatible with the reagents and protecting groups used i n oligosaccharide synthesis.  2.  It should cause no destruction of the i n t e g r i t y of the reducing (terminal) sugar ring.  3.  It should have i n - b u i l t v a r i a b i l i t y of the  "spacer"  between the terminal sugar and the point of attachment on the protein, i n terms of length and hydrophilicity. 4.  The coupling reaction should be convenient,  and  rapid at, or near, physiological pH i n aqueous media, using chemically stable 5.  reagents.  The reaction should give high-yielding coupling to a s p e c i f i c amino acid v i a a h y d r o l y t l c a l l y stable, covalent linkage.  6.  The surface charge of the protein should not  be  s i g n i f i c a n t l y altered by carbohydrate attachment. In consideration of the above-mentioned c r i t e r i a , and the protocols i l l u s t r a t e d above, i t seemed that the most obvious route  two was  the glycosidatlon of a sugar such that an aldehyde could be generated at a s p e c i f i c point on the aglycone.  Our  (see Ch. 1.2)  and Gray's  - 48 -  experience with the reductive amination reaction using sodium cyanoborohydride encouraged us to consider this method for the coupling reaction.  Further, we wanted a protocol which r e l i e d on previously  characterized reactions. Our p r o t o c o l  7  i s i l l u s t r a t e d i n F i g . 1.4.2  a-g-glucose to bovine serum albumin (BSA). steps:-  for the attachment of  The procedure involves three  glycosidation of the parent sugar with an alkenyl alcohol,  reductive ozonolysis i n methanol to generate an aglycone aldehyde (not i s o l a t e d ) , and reductive amination with l y s y l 6-amino groups on the protein using sodium cyanoborohydride. characterized i n the  F i g . 1.4.2.  A l l steps have been well  literature.  Neoglycoprotein synthesis protocol.  From Ref. 7.  - 49 -  The a l l y l group i s well known for i t s use as a protecting group 8  in carbohydrate synthesis,  9  and Lee and Lee have described  glycosidation of sugars with a l l y l alcohol.  the  The glycosidation  procedures we used were either those of Lee, or v a r i a t i o n s .  1 0  o-Glycosides were generally prepared by heating the free sugar i n a l l y l alcohol under reflux, and B-glycosides prepared from the a-bromo sugar per-acetate and alkenyl alcohol v i a a Koenigs-Knorr procedure.  11  related  To i l l u s t r a t e the v e r s a t i l i t y of spacer arms, glucose was  also glycosidated with undecenyl alcohol - H0-(CH2)9CH=CH2, thereby introducing a long ( r e l a t i v e l y hydrophobic) spacer arm between the "nascent aldehyde" ( i . e .  the alkene) and the sugar.  The prepared  allyl  glycosides a l l had melting points i n close agreement with the l i t e r a t u r e , and their NMR structures.  12  spectra were consistent with their  They are a l l stable molecules which can, as far as we  know, be stored i n d e f i n i t e l y at room temperature. Although alternate routes exist for the oxidative cleavage of the alkene aglycone (e.g. periodate-ruthenium dioxide o x i d a t i o n ) , we 13  14  found the reductive ozonolysis reaction  to be fast, s p e c i f i c , and  high-yielding ( e s s e n t i a l l y quantitative i n a t o t a l of ca. 2 hours).  The  sugar i s dissolved i n a methanolic solvent (pure methanol i n the case of a free sugar or 4:1 MeOH-CH Cl with protected sugars) and purged with 2  ozone at reduced temperature  2  (5-10 min./mmole).  The reaction i s  " s e l f - i n d i c a t i n g " , as excess ozone Is manifested by the clear solution turning blue.  This step produces the hydroperoxide (8) which i s reduced  i n s i t u with dimethyl sulphide (DMS). mechanism  15  The generally accepted  for the reaction i s given i n F i g . 1.4.3.  - 50 -  F i g . 1.4.3.  Reaction mechanism for the reductive ozonolysis of an alkene i n methanol, using dimethyl sulphide.  After adding an excess of DMS, temperature DMS  the solution i s allowed to reach room  and l e f t for ca. 2 hours.  Methanol, methanal and unreacted  can be removed by rotary evaporation under reduced pressure and, i n  the case of the protected sugar, DMSO removed by repeated washing of a  - 51 -  chloroform solution with water.  The aldehydo sugar (3) was only  i s o l a t e d for the purpose of characterization, and the reaction routinely monitored for completion by TLC. The aldehydo sugar was then dissolved with BSA at a molar r a t i o of ca. 5 (sugar:lysyl 6-amine).  The pH was adjusted to 9.0, and a ca.  20-fold molar excess of NaCNBH3 added. blocked sugar, a heterogeneous  In the case of the coupling of a  reaction was performed i n a  water/methanol solvent pair (85:15).  The mixture was shaken and  coupling allowed to proceed for 18-48  hours, at which time the  neoglycoprotein was isolated. Sephadex G-25 this was  The homogeneous solution was applied to a  column and the excluded volume c o l l e c t e d .  In some cases  further p u r i f i e d by d i a l y s i s against running water, and then  the solution was  lyophilized.  The s o l i d neoglycoprotein was  stored at -4°C.  Sugar analyses  were performed on milligram quantities using the standard phenolsulphuric acid a s s a y ,  16  with a c a l i b r a t i o n curve prepared from freshly  prepared standard solutions of the appropriate sugar.  The protein was  not found to i n t e r f e r e with the assay, and accuracies are within 10%. Using the above-mentioned procedures, neoglycoproteins (10-17) were prepared.  The coupling conditions chosen were, largely, those of 17  Schwartz and Gray.  The use of ratios of sugar/amine (5:1), should, 18  based on other workers' experience,  result predominantly i n 2° amine  formation at the ~35 easily accessible 6-amino groups on BSA.  The  higher levels of incorporation observed with the attachment of acetylated sugars to BSA, i n methanollc/aqueous  solutions, could either  -  Table  1.4.1  -  52  -  T a b l e o f s u g a r s c o u p l e d t o BSA and BSA b e a r s 5 5 , l y s y l 6 - a m i n o g r o u p s a  Expressed  as  their efficiencies. per molecule.  mole/mole  Coupling Sugar  Glycoside  Aglycone  Efficiency  Sugars BSA  Molecule  10  a-galactose  allyl  36  20  11  o-glucose  allyl  36  20  12  8-N-acetyl  allyl  5  3  13  6-lactose  allyl  24  13  14  ct-acetoglucose  allyl  36  20  15  8-acetoglucose  allyl  16  8-aceto-N-acetyl-  allyl  24  13  undecenyl  15  8  glucosamine  69-82  38-45  glucosamine 17  8-acetoglucose  per 3  - 53 -  result from the "exposure" of formerly inaccessible amines, or 3° amine formation ( i . e . attachment of >1 sugars to each amine) r e s u l t i n g , perhaps, from a decrease i n aldehydrol concentration caused by the reduced water concentration. Although extensive pH p r o f i l e studies were not performed here, the reaction was found to proceed with greater e f f i c i e n c y i n non-buffered solutions.  This could result from the anionic form of the  buffer associating with the cationic l y s y l 6-amino group (pH < 10) i n an ion pair, thereby "shielding" i t from c o l l i s i o n with the aldehyde, and possible aldimine formation. Coupling reactions were often performed overnight (18 hrs) but s a t i s f a c t o r y levels of d e r i v a t i z a t i o n were possible i n 2-4 hrs.  Rather  than extensively characterize an already well-studied reaction, we were content with the levels of substitution and made no attempts to "fine-tune" the coupling procedure.  S i m i l a r l y , we make no attempt to  interpret the varying coupling levels too c l o s e l y . Neutral free sugars appear to give a substitution l e v e l of coupling of ca. 20 sugars per BSA*, and N-acetyl glucosamine appears to give lower l e v e l s .  Lactose coupled to BSA at a lower l e v e l of  e f f i c i e n c y caused, perhaps, by i t s greater bulk.  The highly e f f i c i e n t  coupling of 8-acetoglucose (38-45 sugars/BSA) cannot be explained without further experiments.  The r e l a t i v e l y low e f f i c i e n c y of coupling  the long-chain B-acetoglucose molecule could result from Its reduced s o l u b i l i t y In aqueous media, and the p o s s i b i l i t y of i t s s e l f - a s s o c i a t i o n . *With much longer coupling times, up to 32 sugars may be attached.  - 54 -  In summary, we believe we have developed a useful protocol for neoglycoprotein synthesis.  By c a p i t a l i z i n g on prior detailed analyses  of a similar coupling reaction, such studies were obviated here. l i m i t a t i o n s of Gray's procedure are circumvented  The  by using a sugar  glycoside bearing a "nascent" aldehyde on the aglycone i n the form of an alkene.  Protected* or free sugars have been attached to l y s y l 6-amine  groups of a common test protein,  BSA.  It i s appropriate to mention related developments which have occurred since the inception of our studies. reported  Lee and Lee have  18 19 » the synthesis of 1-thioglycosides bearing co-aldehydo  groups, stored as a dimethyl acetal. (0.05 M HC1,  100°, 20 min.)  Acid hydrolysis of the acetal  i s required before the coupling stage, and  we draw the reader's attention to the p o s s i b i l i t y that acid-sensitive g l y c o s i d i c bonds (e.g. fucosyl) may these sugars be present.  be hydrolysed at this stage, should  Jentoft and Dearborn performed  a detailed  analysis of the reductive amination reaction with raethanal and BSA - a close comparison between their observations and ours i s dangerous, due to the very different nature of the aldehyde.  Lee and Lee recently des-  20  cribed  the use of 1-thioaldehydo sugars very similar to ours; they ob-  served that at very high sugar/protein molar ratios (ca. 1000:1) as many as 70 sugars can be attached to each BSA molecule.  Obviously, i n this  case, a large number of 3° amines are formed, and the authors also postulate the existence of -35 "accessible" l y s y l 6-amine groups, and the rest r e l a t i v e l y hidden.  One might postulate that the l a t t e r are perhaps  involved i n i o n i c bonds with a c i d i c amino acids, to s t a b i l i z e the *Neoglycoproteins of this kind can be de-O-acetylated, i f required.  - 55 -  protein's t e r t i a r y structure.  F i n a l l y , Friedman et a l . have reported 21  the reductive amination of a protein and an aromatic aldehyde;  the  reaction proceeds i n very good y i e l d (ca. 90% of l y s y l 6-amines modified), suggesting  that this non-polar aldehyde can penetrate  the  protein to the l y s y l amines previously determined "inaccessible". finding i s i n accord with ours for the neoglycoprotein very high y i e l d was  This  (15), where a  observed when coupling a protected sugar, which i s  r e l a t i v e l y apolar, to a methanol!c/aqueous solution of B S A . 1.4.3  C  13  NMR  of  Neoglycoproteins  L i t e r a t u r e precedent has established  13  C  NMR  spectroscopy  as a  useful tool i n the study of proteins i n solution, within the r e s t r a i n t s of the necessary large sample size and high s o l u b i l i t y (caused by the 13  low inherent s e n s i t i v i t y of  C).  Comparing the spectra of native and  denatured proteins, i t became clear i n the late 1970's that i t would be 22  possible to observe single carbon s i t e s i n the native protein. The relaxation behaviour of  13  C  nuclei i s of great importance as  i t has the potential to report on the mobility of the protein at a s p e c i f i c s i t e , once the relaxation mechanisms are understood. Therefore,  i t i s common to report T_i, nOe  and line-width (Avo.5)  measurements i n an attempt to derive conclusions on this fundamental aspect of protein behaviour i n solution (vide i n f r a ) . To simplify analysis the protein's spectrum i s divided into three regions:-  a l i p h a t i c region (6 10-75), aromatic region (6 105-160)  and the carbonyl region (6 170-185).  The  100.3  MHz  native and denatured B S A are shown i n F i g . 1.4.4.  13  C  NMR  spectra of  Fig.  1.4.4  1  3  C  and  NMR  spectra  (B)  the  at  100.3  denatured  MHz  protein  of in  (A)  native  8 M  BSA  urea/D20.  in  D 0, 2  - 57 -  It  i s well known that the  C NMR  spectra of common neutral  23  carbohydrates  i s d i v i s i b l e into two regions:-  the O-glycosidic  anomeric carbon region (6 95-105) and the region i n which the rest of the hydroxylated and N-glycosldic anomeric carbons resonate (6 60-80). C l e a r l y , the anomeric carbons f a l l into a spectroscopically s i l e n t "window" of the protein spectrum, and the other carbohydrate resonances overlap s l i g h t l y with a few protein a l i p h a t i c resonances (mainly, serine and threonine C P ) .  Hence i t would seem reasonable to postulate that  anomeric carbons w i l l be uniquely discernible and other carbohydrate 13  carbons somewhat less so, i n the  C NMR  spectrum of natural or  synthetic glycoproteins. A possible factor i n the assignment  of carbohydrate resonances  between 6 60-80 might be the r e l a t i v e mobility of the overlapping resonances.  Our previous studies (Chs 1.2 and 1.3) indicate that the  terminal sugar has at least an order of magnitude of rotational  freedom  greater than the protein and t h i s , reflected In their line-widths, could prove s u f f i c i e n t to d i f f e r e n t i a t e the two. Fig.  1.4.5  shows the result of superimposing the spectrum of native BSA  with that of B-O-methyl glucopyranoside. resonances are broadened Fig.  To i l l u s t r a t e this point,  In F i g . I.4.5B, the sugar  by 2 Hz with an Exponential f i l t e r , and i n  I.4.5C, by 15 Hz; only the "carbohydrate region" (6 60-110) i s 23  plotted, and the carbohydrate peaks assigned. of  C l e a r l y , a large number  r e l a t i v e l y mobile carbohydrate resonances can be readily  distinguished from the broader peaks originating from protein carbons, i n F i g . I.4.5B.  - 58 -  B.  100  F i g . 1.4.5  80  6(ppm)  C NMR spectra at 100.3 MHz, showing the "carbohydrate region" only. (A) Native BSA. (B).(A), superimposed with the spectrum of methyl B-D-glucoside, with  13  line-widths broadened by 2 Hz. (C).(B), with line-widths broadened by 15 Hz. D: internal dioxane reference (6 67.4).  -  These elementary  59  -  " f e a s i b i l i t y studies" bode well for the 13  p o s s i b i l i t y of observing  C sites on a (neo)glycoprotein, but they 13  cannot predict the possible effects the protein may have on the chemical s h i f t s of the sugar(s).  13  C  C  resonances are notoriously  d i f f i c u l t to assign and the most common resource i s to make  comparisons  between the chemical s h i f t s of known, assigned molecules and those of the "unknown".  Thus, we might consider the methyl glycosides of mono-,  d i - and oligosaccharides to be useful model compounds i n the  assignment  13  of (neo)glycoproteins' C NMR spectra, and we went about testing this theory. P f e f f e r et a l . have used deuterium isotope-induced s h i f t s (DIS) 13 23 to assign the C NMR spectra of a range of mono- and disaccharides, and their data was used throughout this study.  This was  the rationale  for studying neoglycoproteins as model glycoproteins: the carbohydrate moieties have been assigned independently of the protein and, by 13  recording the  C NMR  spectra of the (known) same sugar, now attached to  the protein, useful generalizations might be derived concerning assignment  i n the glycoconjugate.  F i g . 1.4.6  shows the 100.3 MHz  13  C  NMR  spectrum of a  neoglycoprotein comprising 8-glucose attached to BSA* (11, having 14 glucose molecules per BSA), comparing i t with the spectrum of BSA.  The  carbohydrate region i s expanded above the f u l l spectrum and the chemical s h i f t s of methyl 6-g-glucoside indicated above the peaks.  We  immediately note that the carbohydrate peaks are sharp i n comparison *A11 neoglycoproteins were prepared by the procedure detailed i n the previous section (1.4.2).  - 60 -  C-6  150 Fig. 1.4.6  100  50  5(ppm)  C NMR spectrum of a neoglycoprotein (B-glucose attached to BSA), with the "carbohydrate region" expanded and asigned. Lines aboves the resonances indicate the l i t e r a t u r e chemical s h i f t s of methyl 8-D-glucoside.  13  - 61 -  C-6'  C-2,-3'  C-1'  C-3 — 1 C-2' C-5 - ,  C-5'n'l C-1  C-6  C-4'  C-4  C-3,-4  C-1  C-6  C-5  C-2  B.  C-5  C-3  C-1  The "carbohydrate  80  6 (ppm)  region" of the 100.3 MHz  S ^ B-lactose T ^ attached * andd(C) 8 1 7  1  8  " to BSA. ( A )  C-6  C-4  C-2  100 F i g . 1.4.7.  V  P  g  l  u  c  o  -  1 3  C NMR  osT  (B) ^ a lgalactose, ac  Table 1.4.2  Comparison of C chemical s h i f t s of the carbohydrate resonances of neoglycoproteins with the methyl glycoside of the attached sugar. The chemical s h i f t s of neoglycoproteins are with an error of ± 0 . 1 ppm. A6 represents the difference i n chemical s h i f t between a sugar resonance of a neoglycoprotein and i t s corresponding methyl glycoside.  C-l  C-2  C-3  C-4  C-5  C-6  8-glc-BSA B-Me-glc  103.1 103.4 -0.3  74.0 74.1 -0.1  76.6 76.8 -0.2  70.6 70.7 -0.1  76.9 76.8 -0.1  61.7 61.8 -0.1  oc-gal-BSA o-Me-gal  99.6 100.1 -0.5  69.2 69.2 0.0  70.2 70.5 -0.3  70.2 70.4 -0.2  72.1 71.6 -0.5  62.1 62.2 -0.1  103.3 103.8 -0.5  73.8 73.8 0.0  75.4 75.3 0.1  79.7 79.3 0.4  76.0 75.6 0.4  61.4 61.1 0.3  A6  A6  B-lac-BSA B-Me-lac  A6  C-l'  C-2»  C-3'  C-4'  C-5'  C-6'  104.1 103.8 0.3  72.1 71.9 0.2  73.8 73.5 0.3  69.7 69.5 0.2  76.4 76.2 0.2  62.0 62.0 0.3  - 63 -  with those of the protein, and the chemical s h i f t correlation i s 13  excellent.  F i g . 1.4.7  shows the  C NMR  spectrum of the carbohydrate  region of the three neoglycoproteins indicated.  Sharp resonances are  seen i n every case and the measured chemical s h i f t s are compared with those of their parent methyl glycosides i n Table 1.4.2. Within the l i m i t s of ± 0.5 ppm,  the agreement i n chemical s h i f t s  between the methyl glycoside and the corresponding neoglycoprotein sugar resonances i s excellent, with the constraints of this s t a t i s t i c a l l y small survey. (=0.5  ppm)  Except for a trend for C-l to be s l i g h t l y deshielded  i n the neoglycoprotein compared with the methyl glycoside,  the differences i n chemical s h i f t (A6) are often within the experimental error of the measurement. of a sugar may  Hence, we conclude that the methyl glycoside  be used to give chemical s h i f t s which can be compared  closely (± 0.5 ppm) with the chemical s h i f t s of the sugar attached to a protein.  This i s an important finding and i s analogous to the use of  small saccharides' chemical s h i f t s to assign and determine the structure. Oh  of polysaccharides,  which i s now a r e l a t i v e l y common procedure.  We then decided to extend the study and determine whether some of the technologies used i n previous chapter could be applied to these systems to a s s i s t i n carbohydrate structure determination. S p e c i f i c a l l y , we considered D-galactose oxidase which catalyses the oxidation of, e.g., methyl o-D-galactose to the molecule having an aldehyde at C-6; i n aqueous solution, methyl 1,5-pyranoside NMR  (19), predominates.  by Maradufu and P e r l i n ,  2 6  q-D-galacto-hexodialdo-  This reaction has been studied by *H  and Whyte and E n g l a r .  2 7  The  13  C  chemical  - 64 -  F i g . 1.4.8.  The action of D-galactose oxidase on methyl a-D-galactopyranoside.  28  s h i f t of an aldehydrol may D2O; C-13  be t y p i f i e d by a study  on streptomycins i n  (aldehydrol) has a chemical s h i f t of 90-96 ppm.  The chemical  s h i f t of the hydrated aldehydic carbon of acetaldehyde i s 88.9 Hence, we might confidently anticipate that the  13  C  ppm.  chemical s h i f t of  C-6 of a terminal galactosyl residue on a glycoprotein might change by ~ 30 ppm after oxidation with galactose oxidase and the aldehydrol should be e a s i l y detectable, as i t resonates i n a spectroscopically " s i l e n t " region.  To test t h i s , and ascertain the effect of oxidation at  C-6 on the other ring carbons' chemical s h i f t s , we subjected a 0.2 M solution of methyl a-g-galactopyranose to such an oxidation, the  13  C  NMR  measuring  spectrum before and after the reaction ( F i g . 1.4.9).  I.4.9A shows the  13  C  NMR  Fig.  spectrum before, and F i g . I.4.9B after ca. 20%  galactose oxidase oxidation.  The s h i f t In resonant frequency (A6) i s  - 65 -  C-3-, C-5-  C-1  C-6  C-4 -C-4(0) 0.5 C-2  C-5(0) -,  B.  r- C6(0)  C-2(0)  -1.7  -27.4  0.1  C-5  C-1  r-C-3 rC-4 r-C-2  C-6  u 100  F i g . 1.4.9  80  5 (ppm)  C NMR spectrum (100.3 MHz) of (A) methyl a-galactoslde, and, (B), the same after oxidation at C-6 by D-galactose  1 3  oxidase, at 20% e f f i c i e n c y .  - 66 -  recorded f o r each carbon.  As might be expected, C-6 suffers the largest  s h i f t (-27.6 ppm); C-5 and C-4 are also shifted s i g n i f i c a n t l y . and C-3 did not s h i f t measurably.  Only C-1  I t i s interesting to note that the  r e l a t i v e peaks heights of C-6 and C-3 do not change s i g n i f i c a n t l y with oxidation, but the decreased C-2,  C-4, C-5 and C-6.  intensity i s measurable i n peak areas of  This p i l o t experiment bodes well for the use of  galactose oxidase i n the assignment of C-4, -5 and -6 terminal galactosyl residues, especially when an accurate integration i s possible. Galactose-containing neoglycoproteins were then reacted with galactose oxidase.  Compound 13 bears terminal galactose residues  (lactose bound to BSA) and, a f t e r treatment with the enzyme, the spectrum i n F i g . 1.4.10B was obtained.  The aldehydrol C-6' i s clear (6  98.8), but the l e v e l of oxidation i s low, as the large galactosyl methylene C-6' peak suggests.  This peak i s narrower, (indicating less  area), and so we may only conclude from this experiment that (a) the pendant sugar bears terminal galactose residues, and, (b) the galactosyl C-6 resonates at 6 62.0.  S h i f t s i n resonance frequency  of other carbon  centres are small, overlap with other peaks, or are lost i n the noise. Whether this would be the case with complete oxidation i s not possible to predict, but the model study with the a-galactoslde, above, suggests that s i g n i f i c a n t s h i f t s i n the resonance frequencies of C-5' and C-4' would occur, a s s i s t i n g i n their assignment, too. Two control experiments were performed.  F i r s t l y , the oxidized  sample which gave the spectrum of F i g . 1.4.10B was reduced with sodium  -  1.4.10  -  80  100  Fig.  67  8 (ppm)  T h e 1 0 0 . 3 MHz C NMR s p e c t r u m o f 6 - l a c t o s e a t t a c h e d t o BSA. (A). Unoxidized. (B) A f t e r o x i d a t i o n b y g a l a c t o s e oxidase overnight. ( C ) A f t e r r e d u c t i o n o f (B) w i t h 1 J  NaBHi*.  - 68 -  c-6'n  Fig.  1.4.11  The  100.3  allyl  MHz  1 3  C  8-lactoside,  galactose  (A).  8 (ppm)  80  100  oxidase,  NMR s p e c t r u m (A)  after  and  (B),  of  a mixture  enzymic after  of  BSA  oxidation  by  NaBHi  t  reduction  and of  - 69 -  borohydride  (Fig. I.4.IOC).  By this time the sample was  quite degraded,  and this i s evident i n the line-widths i n F i g . 1.4.IOC, which r e f l e c t the fact that some sample had precipitated by the end of the measurement.  However, the aldehydrol peak i s no longer present and  the  spectrum bears the features of that of the starting material, as might be predicted.  To determine whether the protein was  i n some way  i n h i b i t i n g the action of the enzyme, a l l y l B-g-lactoslde was mixed with BSA  i n a mole r a t i o comparable with the neoglycoprotein.  oxidase treatment almost completely  oxidized C-6'  as indicated by i t s  large down-field resonance s h i f t (Fig. I.4.11A,B). monosaccharide, C-5'  s h i f t s down-field  1.6 ppm,  Galactose  Similarly with the  C-4'  up-field 0.8  ppm  and the carbon involved i n the glycosidic linkage (C-4) down-field by a surprisingly large 0.9  ppm.  The last observation has  important  implications as i t suggests that the galactose oxidase treatment gives information on the point of attachment of the terminal galactose, too. It i s interesting to note the changes i n the spectrum between 6 65 and 80 - many extra peaks appear after oxidation; i t i s possible that some of these result from side-reactions galactose oxidase i s known to •t  produce.  26 Again, borohydride  reduction resulted i n the loss of the  aldehydrol peak, and a reappearance of the galactosyl methylene  C-6'.  This second experiment i s not a perfect control, but unambiguously shows that galactose oxidase can oxidize C-6' the presence of BSA,  galactose groups on lactose i n  and that d i a g n o s t i c a l l y useful chemical  changes occur with f u l l oxidation.  shift  - 70 -  Galactose oxidase f a i l e d to react with 10, a-galactose attached to BSA.  I t i s tempting to postulate that the free substrate ( a l l y l  8-D-lactopyranoside) can easily f i t i n the enzyme's active s i t e and i s e f f i c i e n t l y oxidized. Lactosyl-BSA (13) i s p a r t i a l l y oxidized due to somewhat r e s t r i c t e d " f i t " , or a c c e s s i b i l i t y of substrate to enzyme caused by the proximity of BSA.  With galactosyl-BSA (10), the substrate  cannot f i t into the active s i t e at a l l , and no reaction can occur.  To  test this hypothesis i t would be i n t e r e s t i n g to synthesize a series of alkenyl glycosides having spacer arms of varying length, attach each one to BSA and determine galactose oxidase.  the a c c e s s i b i l i t y of the sugar to oxidation by I f our reasoning i s correct, an increase i n l e v e l of  oxidation should occur as the spacer arm i s lengthened. I t i s interesting to speculate on the effect s p e c i f i c chemical 13  C NMR  modifications might have on the  spectra of (neo)glycoproteins.  I t i s probable that such approaches might usefully augment the enzymatic approach which we have evaluated and found p o t e n t i a l l y useful. 13  In summary, studying the  C spectra of neoglycoproteins might  prove useful i n the investigation of glycoproteins. The resonances  carbohydrate  are sharper than most protein resonances, r e f l e c t i n g the  influence of their greater mobility on T_2.  The corresponding methyl 13  glycoside of the attached sugar appears to be a good model for assignment.  C  That i s to say, i f a sugar glycoside Is characterized, and  13  its  C NMR  spectrum assigned, i t i s reasonable to expect that the  spectrum of the same sugar attached to a protein w i l l have peaks at the same frequencies (± 0.5 ppm).  F i n a l l y , except with only galactose  attached to a protein, terminal galactosyl residues present w i l l be  -  71  -  determinable from their s u s c e p t i b i l i t y to oxidation by galactose oxidase.  This procedure not only i d e n t i f i e s the presence of terminal  galactose, but also c l e a r l y assigns a number of i t s resonances, and the attaching carbon i n the sub-terminal  sugar residue.  F i n a l l y , we make mention of the relevant  13 C NMR studies of  glycoconjugates performed i n other laboratories a f t e r the inception of our work; much of these come from the laboratory of Allerhand.  Harris  29 and Thornton  studied carbohydrate head-group dynamics of g l y c o l i p i d s  i n various phases; only anomeric resonances were assigned  and I t was  concluded that the pentasaccharide had no d i f f e r e n t i a l i n mobility along i t s chain, i n d i c a t i n g a closely hydrogen-bonded e n t i t y . Allerhand  30  reported the f i r s t  glucoamylase.  1 3  D i l l and  C NMR spectrum of a glycoprotein,  Few s p e c i f i c assignments were possible due to the complex  nature of the carbohydrate, but certain generalizations were possible. 3 A study  1 on the highly regular and simple antifreeze glycoproteins  yielded b e a u t i f u l  1 3  C NMR spectra; a d i f f e r e n t i a l i n mobility i s quite  clear from the carbohydrate line-widths and e x p l i c i t assignments were possible by comparison of chemical s h i f t s of low molecular weight analogues. The anomeric carbons were assigned on the basis of *J_CH measurements. Jennings et_ a l . used chemical and enzymic methods to 13 a s s i s t i n the assignment of the  C NMR spectrum of a streptococcal  polysaccharide. Ribonuclease (RNA'se) Is found i n glycosylated (RNA'se B) and non-glycosylated  forms (RNA'se A), and both were studied by Allerhand 33  and co-workers.  Protein f o l d i n g was found to have small effects on  the chemical s h i f t s of close carbohydrate centres.  Comparison of model  - 72 -  oligosaccharides allowed the structure determination of the complex glycoprotein carbohydrate. Arm!tage and co-workers  characterized the glycopeptide of  glycophorin A, a red-blood-cell membrane glycoprotein.  Assignments were  largely based on procedures common i n polysaccharide studies - model compound chemical s h i f t s plus a d d i t i v i t y rules.  Calcium ion t i t r a t i o n s  13  induced s h i f t s i n N-acetylneuraminic acid  C resonances.  Several  unusual s h i f t s are explained by the involvement of hydrogen bonding i n the maintenance of a fixed secondary Berman and A l l e r h a n d  35  structure.  studied the hydrolytic action,  s p e c i f i c i t y , and k i n e t i c s of an a-mannosidase (endoglycosidase enzyme) on the glycopeptide of ovalbumin, where they found that most Information came from a study of the anomeric region. assignments has not, yet, been published.  The rationale for their  36  Goux et a l . perfomed  a detailed relaxation study on uniformly  13  C-enriched galactose, attached by enzymatic methods to the carbohydrate chain of hen ovalbumin.  The o v e r a l l motion of the terminal  residue was found to be anisotropic, comprising a slow (23 ns) i s o t r o p i c contribution ( a r i s i n g from the motion of the glycoprotein as a whole) and a fast (40-80 ps) contribution a r i s i n g from rotation of the carbohydrate C-H vector rotating at an angle of 30° about the e f f e c t i v e axis of rotation. 37 F i n a l l y , Bedford and co-workers have reported the spectra of canine mucins.  13  C  NMR  P u r i f i e d materials y i e l d reasonably sharp  resonances i n the carbohydrate region of the spectrum, which further sharpen when the material i s treated with 8-mercaptoethanol  (which  - 73 -  reduces  protein cysteinyl disulphide bridges).  T h e i r m a t e r i a l was found  t o have a v e r y low s i a l i c a c i d c o n t e n t , which was s u r p r i s i n g . 1.4.4  *H NMR of Neoglycoproteins 13  The  previous section i l l u s t r a t e d  C NMR as h a v i n g the p o t e n t i a l  t o be of c o n s i d e r a b l e u t i l i t y i n t h e study of g l y c o p r o t e i n s . Two major d i f f i c u l t i e s remain:-  the l a r g e amounts of sample r e q u i r e d (> 120 mg)  and the d i f f i c u l t y o f a s s i g n i n g complex g l y c a n s .  The f i r s t problem may  be p a r t i a l l y s o l v e d by o b s e r v i n g a more s e n s i t i v e n u c l i d e , such as H , 1  and i n Chapter  I I . 3 we d e s c r i b e how *H NMR may be used de_ novo t o a s s i g n  oligosaccharides. When s t u d y i n g the *H NMR spectrum o f g l y c o p r o t e i n s , one major d i f f i c u l t y l i e s i n the e x t e n s i v e o v e r l a p o f g l y c a n and p r o t e i n s i g n a l s . Our p r e v i o u s s t u d i e s have i n d i c a t e d t h a t the g l y c a n s a r e at l e a s t an o r d e r o f magnitude more mobile  than t h e p r o t e i n , and t h i s f a c t was seen  t o be p i v o t a l I n the r o u t e t o the " s p e c t r o s c o p i c f a c t o r i z a t i o n " of t h e sugar resonances.  L i t e r a t u r e precedent  r a p i d l y t u m b l i n g molecules  e x i s t s f o r the o b s e r v a t i o n o f  i n the presence  o f immobile ones, where a l l  methods r e l y on the d i f f e r e n t i a l i n t r a n s v e r s e r e l a x a t i o n times (T_ ) 2  between the two e n t i t i e s .  These w i l l be reviewed  a t t h e end o f the  chapter. A l l methods f o r such d i f f e r e n t i a t i o n come under the heading o f "spin-echo"  techniques.  E s s e n t i a l l y , a 90° p u l s e c r e a t i n g  m a g n e t i z a t i o n i n the x'-y' p l a n e i s f o l l o w e d by a d e l a y T , and a second p u l s e i s used t o induce t h e s p i n - e c h o , h a v i n g a maximum a t 2T. The c h o i c e o f t h e d e l a y I s c r u c i a l , s i n c e t h i s i s where the slow and f a s t tumbling molecules  are "sorted".  P r o t o n s h a v i n g l o n g T^'s ( r a p i d l y  - 74 -  tumbling) w i l l s t i l l have magnetization after the refocussing pulse (180°), and w i l l , therefore, be observed; those with short T_2's (slowly tumbling) w i l l have lost their transverse magnetization by the time the 180° pulse i s applied and, therefore, w i l l not be seen when the receiver i s turned on. The c l a s s i c a l experiment i s the Hahn spin-echo, comprising a (90°-T-180°-X-AQN) pulse-sequence.  A spin-echo maximum i s generated at  2x, and the following half-echo i s then acquired (see explanation of 2D J-resolved spectroscopy, Ch. II.2.5.2).  The amplitude of the refocussed  echo i s determined by a T_2 term, and i n some cases, d i f f u s i o n rates and magnetic f i e l d f i e l d gradients w i l l enhance the decay of the spin-echo amplitude.  With homonuclear spin-coupled spins, the peaks' phases w i l l  be "J-modulated", meaning that the peaks may be p a r t i a l l y or f u l l y Inverted.  In some instances, such as when J_ i s known, this may be of  u t i l i t y for assignment purposes, but this i s not often the case.  One  recourse i s to calculate the phase-insensitive spectrum (magnitude, or power mode), but contributing dispersive components lead to broad l i n e s being observed with this approach. I f a high-resolution spin-echo spectrum with no J-modulation i s required, one of two approaches may be taken.  A t r a i n of closely spaced  180° pulses can be used to refocus the magnetization  (rather than a  single 180° pulse) and this removes the dependencies on J_ and the 39  d i f f u s i o n and magnetic f i e l d gradient terms. i s used and narrow lines are observed.  Phase-sensitive display  Alternately, one can make use of  SEAS * (spin-echo absorption spectroscopy), which requires c o l l e c t i o n of 1  0  - 75 -  the f u l l Hahn-echo (90° - x - 180°^ - AQN). echo, Fourier transform  With a f u l l y  symmetrical  (FT) and phase-insensitive display removes  J_-modulation and gives absorption-like peaks. Although the two methods were not c r i t i c a l l y compared i n our laboratory, we chose to use SEAS f o r these studies. experiment  was modified  The basic  s l i g h t l y to include a solvent n u l l procedure -  3 s presaturation p r i o r to the 90° pulse.  A time-domain window function  which insured that there was zero i n t e n s i t y at the beginning and end of the a c q u i s i t i o n * improved spectral q u a l i t y . The mg/0.4 ml).  f i r s t test of the technique was on a D 0 solution of BSA (14 2  F i g . 1.4.12A shows the normal spectrum, with HOD solvent  n u l l i n g ; only broad spectral features are d i s c e r n i b l e .  The SEAS  experiment was then performed, with the indicated x delays ( F i g . I.4.12B-F).  Using x = 40 ms results In very l i t t l e signal i n t e n s i t y  a r i s i n g from the BSA protons, and so i t was decided  to perform a l l  subsequent experiments using t h i s delay time. Results f o r the SEAS experiment (x • 40 ms) for a variety of molecules are given i n F i g . I. 4.13.  The question was:  can SEAS be  used to y i e l d the spectrum of the glycan portion of a glycoconjugate with s u f f i c i e n t d e t a i l f o r this to be of u t i l i t y i n their assignment? F i r s t l y , BSA and the B - a l l y l glycoside of acetylated glucosamine were co-dissolved i n D 0. 2  SEAS (x = 40 ms) yielded spectrum 1.4.13B,  which compares favourably with the spectrum of the monosaccharide i n D 0 2  *The Nicolet (NTCFTB) "trapezoidal m u l t i p l i c a t i o n " (TM) command was used.  - 76 -  5  F i g . 1.4.12  6(ppm)  0  The 270 MHz H spectra of BSA i n D 0 (14 mg/6.4 ml). (A) Single pulse experiment. (B)-(F) SEAS experiments with the indicated x values. The HOD was supressed by 3 8 presaturation. A  2  - 77 -  OH  H O - ^ L ^ -NHAc ^ - o -BSA  ^OAc AcONHAc  C.  BSA  x  .OAc  +BSA  AcONHAc  B. 1-  LJVAAJHA  _J  S(ppm) F i g . 1.4.13  The 270 MHz H SEAS spectra with T»40 ms. (A) BSA. (B) BSA plus a l l y l 8-acetoglucosamine. (C) B-AcetoglcNAc attached to BSA. (D) o- and B-glcNAc attached to BSA. X  - 78 -  (data not shown).  The carbohydrate  using T up to 100 ms.  spectrum i s s t i l l clear even when  We conclude that SEAS can be used to y i e l d the  spectrum of a rapidly tumbling monosaccharide i n the presence of a protein macromolecule. Next, the neoglycoprotein having aceto-glucosamine attached to BSA  (12) was  studied and i t s SEAS spectrum i s given i n F i g . 1.4.13C*  A pattern of peaks similar to the monosaccharide i s evident, but much of the d e t a i l i n coupling constants has been lost to the increased i n t r i n s i c line-widths.  With x > 40 ms,  the carbohydrate  are rapidly l o s t , leaving only the acetyl peaks.  ring protons  The B-anomeric proton  i s , as with a l l these spectra, hidden under the incompletely HOD  suppressed  peak. Next, a neoglycoprotein was prepared, bearing a 1:1 mixture of  o- and 8-N-acetyl  glucosamine.  given i n F i g . 1.4.13D.  The SEAS spectrum of this material i s  Broad "humps" are v i s i b l e i n the  carbohydrate  region, but, again, l i t t l e d e t a i l i s evident upon close inspection, and the two anomeric proton signals cannot be detected, as hoped. Our data indicate that SEAS can be used to obtain a *H spectrum of the carbohydrate moiety of a neoglycoprotein.  The line-widths of the  signals appear to be broadened s l i g h t l y i n comparison with the monosaccharide, to the point where they obscure the d e t a i l s of the coupling constants i n the sugar spectrum. of the great strength of *H NMR l i e s i n this information.  This i s unfortunate, as part  as a method of structure determination  Whether this situation could be remedied by 41  the a p p l i c a t i o n of time-domain resolution enhancement functions •Neoglycoproteins  was  were synthesized as described i n Ch. 1.4.2.  - 79 -  not determined, but might be a p o s s i b i l i t y .  The solvent peak overlap  with the anomeric protons was a problem which could probably be p a r t i a l l y a l l e v i a t e d by elevation of temperature.  This would also  increase the mobility (and sharpen the lines of) the carbohydrate molecules, which would be an added bonus.  In short, the spin-echo  experiments described here bode well for the Incorporation of *H NMR i n the scheme of procedures available for the study of glycan components of glycoconjugates.  However, much work s t i l l has to be done before the  u t i l i t y of the procedure can be accurately assessed. In  1975, Campbell et^ a l . described  experiments to simplify protein spectra.  the use of spin-echo  Egmond et a l . studied the *H u3  NMR  spectrum of the sialo-glycoprotein, glycophorin.  The broad,  protein component was p a r t i a l l y removed by the use of a convolution-difference time-domain function.  Several aspects of the  molecule's macroscopic behaviour were studied, and some sialo-sugar assignments were possible i n the h i g h - f i e l d region; the sugars were mobile, r e l a t i v e to the protein. Several workers have used the Hahn spin-echo to study l i v i n g erythrocytes. Brown and co-workers assigned some of the unassociated, low molecular weight components of red blood c e l l s (RBC's), and l a t e r went on to study RBC membrane transport using NMR.  39  US  >  Rabenstein  *+6  i n Alberta has reported on several aspects of the chemistry of small molecules inside and out of RBC's. ••7  The SEAS experiment was f i r s t described by Bax jet a l .  and  l a t e r by H a l l and Sukumar who applied i t to a mixture of carbohydrates, lysozyme * 1  0  and RBC's. * 1  8  - 80  1.4.5  -  Conclusions To sum up, we have described an e f f i c i e n t route to the synthesis  of a r t i f i c i a l glycoproteins (neoglycoproteins). The method Is quick and high-yielding, and we anticipate i t could be of use, too, i n the synthesis of carbohydrate haptens. In NMR  Our use for this technology has been  studies of conjugated sugars. 13  The study of neoglycoproteins by  C NMR  i s f a c i l i t a t e d by the  fact that the sugar resonances resonate, largely, i n a spectroscopically " s i l e n t " region, and, through t h e i r enhanced mobility, may identified.  be  It was determined that the corresponding methyl glycoside  of the sugar(s) attached to the protein acts as a good source of reference for assigning C  resonances, with our neoglycoproteins.  1 3  No  s i g n i f i c a n t s h i f t s i n resonance frequency are observed when we compared three neoglycoproteins t h i s way.  Oligosaccharides bearing terminal  galactose residues can be i d e n t i f i e d by oxidation with D-galactose oxidase.  Significant s h i f t s occur i n galactosyl C-6, C-5 and C-4,  the carbon to which the galactosyl anomeric carbon i s attached.  and  With an  ot-galactoside and 8-lactoslde, the s h i f t s were very similar i n their d i r e c t i o n and magnitude.  This approach would appear to be of great  potential i n the assignment  of sugars bearing terminal galactose,  provided a reasonable l e v e l of oxidation can be achieved. *H NMR  studies of neoglycoproteins require that special steps be  taken to observe the sugar resonances; t h i s can be done by a spin-echo experiment, SEAS.  The method r e l i e s on d i f f e r e n t i a l s i n mobilities (and  T ) and can e a s i l y give the spectrum of a monosaccharide with a 2  - 81 -  protein present.  When the monosaccharide i s attached to the protein,  the d i f f e r e n t i a t i o n i s not as dramatic, and l i n e broadening of the sugar resonances occurs to the point where i t obscures many J^-couplings. Several p o s s i b i l i t i e s exist to remedy this and the other major l i m i t a t i o n , v i z . , the overlap of the solvent resonance with the anomeric proton, but these were not assessed. We f e e l that these studies are, i n their own r i g h t , of interest, but the hope i s that the information derived w i l l ultimately be of use i n the study of glycoproteins using NMR.  - 82 -  REFERENCES 1.  W.F. Goebel and O.T. Avery.  J . Exp. Med. 50, 521-531 (1929).  2.  G.E. Means and R.E. Feeney.  Biochemistry.  3.  N. Jentoft and D.G. Dearborn. (1979) .  4.  G.R. Gray. Methods Enzymol. 50, 155-160 (1978), and references c i t e d therein.  5.  J.D. Aplin and J.C. Wriston, J r . 259-306 (1981).  6.  C P . Lee and Y.C. Lee. Adv. Carbohydr. Chem. Biochem. 37, 225-281 (1980) .  7.  M.A. Bernstein and L.D. H a l l .  8.  P.A. Gent and R. Gigg. J . Chem. Soc. Perkln Trans. 1. (1975), and others i n the series.  9.  R.T. Lee and Y.C. Lee. Carbohydr. Res. 37, 193-201 (1974).  7, 2191-2201 (1968).  J . B i o l . Chem. 254, 4359-4365  CRC C r i t . Rev. Biochem. 10,  Carbohydr. Res. 78, C1-C3 (1980). 361-363  10.  From the s e r i e s : - Methods i n carbohydrate chemistry. Academic Press, New York, Vols. 2 and 5. 1963 and 1965, respectively.  11.  K. Igarashi.  Adv. Carbohydr Chem. Biochem. 34, 243-283 (1977).  12.  J . Balatoni.  B.Sc. Thesis.  13.  H.O. House. Modern synthetic reactions. Benjamin, New York. 1972. Chapter 7, and references cited therein.  14.  J . J . Pappas, W.P. Rlareney, E. Gancher, and M. Berger. Tetrahedron L e t t . 4273-4278 (1966). P.S. Bailey and R.E. Erickson. Org. Synthesis. 4 1 , 41-45 (1961).  15.  P.S. Bailey.  16.  M. Dubois, K.A. G i l l e s , J.K. Hamilton, P.A. Rebers, and F. Smith. Anal. Chem. 28, 350-356 (1956).  17.  B. Schwartz and G.R. Gray. (1977).  18.  R.T. Lee and Y.C. Lee. Carbohydr. Res. 77, 149-156 (1979).  19.  R.T. Lee and Y.C. Lee. Biochemistry.  20.  R.T. Lee and Y.C. Lee. Carbohydr. Res. 101, 49-55 (1982).  University of B r i t i s h Columia.  1981.  Chem. Rev. 58, 925-1010 (1958).  Arch. Biochem. Biophys. 181, 542-549  19, 156-163 (1980).  - 83 -  21.  M. Friedman, L.D. Williams, and M.S. M i s r i . Protein Res. 6, 183-185 (1974).  Int. J . Peptide  22.  A. Allerhand. Accts. Chem. Res. 11, 469-474 (1978), and references c i t e d therein.  23.  P.E. P f e f f e r , K.M. Valentine, and F.W. Parrish. J . Amer. Chem. Soc. 101, 1265-1274 (1979); Ibid., 7438 (correction).  24.  P.A.J. Gorin.  25.  See Chaps 1.2 and 1.3, and references cited therein.  26.  A. Maradufu and A.S. P e r l i n .  27.  J.N.C. Whyte and J.R. Englar.  28.  K. Bock and C. Pedersen.  29.  P.L. Harris and E.R. Thornton. (1979).  30.  K. D i l l and A. Allerhand.  31.  E. Berman, A. Allerhand, and A.L. De V r i e s . 4407-4410 (1980).  32.  H.J. Jennings, C. Lugowski, and D.L. Kasper. 4511-4518 (1981).  33.  E. Berman, D.E. Walters, and A. Allerhand. 3853-3857 (1981).  34.  R. Prohaska, T.A.W. Koerner, J r . , I.M. Armltage, and H. Furthmayr. J. B i o l . Chem. 256, 5781-5791 (1981).  35.  E. Berman and A. Allerhand.  36.  W.J. Goux, C. Perry and T.L. James. (1982).  37.  K. Barrett-Bee, G. Bedford, and P. Loftus. 257-263 (1982).  38.  R. Freeman and H.D.W. H i l l . Determination of spin-spin relaxation times i n high-resolution NMR. Edited by L.M. Jackman and F.A. Cotton. Academic Press, New York. 1975. pp. 131-162.  39.  K.M. Brindle, F.F. Brown, I.D. Campbell, C. Grathwohl, and P.W. Kuchel. Biochem. J . 180, 37-44 (1979), and references cited therein.  Adv. Carbohydr. Chem. Biochem. 38, 13-104 (1981).  Carbohydr. Res. 32, 93-99 (1974). Carbohydr. Res. 57, 273-280 (1977).  J . A n t i b i o t i c s . 27, 139-140 (1974). J . Amer. Chem. Soc. 100, 6738-6745  J . B i o l . Chem. 254, 4524-4531 (1979). J . B i o l . Chem. 255, Biochemistry. 20, J . B i o l . Chem. 256,  J . B i o l . Chem. 256, 6657-6662 (1981). J . B i o l . Chem. 257, 1829-1835 Bioscience Reports. 2,  - 84 -  40.  L.D. H a l l and S. Sukumar. . J . Magn. Reson. 38, 559-564 (1980).  41.  J.C. Lindon and A.G. Ferrige. 27-66, 1980.  42.  I.D. Campbell, CM. Dobson, R.J.P. Williams, and P.E. Wright. L e t t . 57, 96-99 (1975).  43.  M.R. Egmond, R.J.P. Williams, E.J. Welsh, and D.A. Biochem. 97, 73-83 (1979).  44.  F.F. Brown, I.D. Campbell, P.W. L e t t . 82, 12-16 (1977).  45.  F.F. Brown and I.D. Campbell. 395-406 (1980).  46.  D.L. Rabenstein, S.J. Backs, and A.A. Isab. J . Amer. Chem. Soc. 103, 2836-2841 (1981), and others i n the series.  47.  A. Bax, A.F. Mehlkopf, and J . Smidt. (1979).  48.  S. Sukumar.  Ph.D.  Thesis.  Prog, i n NMR  Spectroscopy. 14,  Rees.  FEBS  Eur. J .  Kuchel, and D.C. Rabenstein.  FEBS  P h i l . Trans. R. Soc. Land. B. 289,  J . Magn. Reson. 35, 373-377  University of B r i t i s h Columia.  1980.  -  85  -  CHAPTER 1.5 EXPERIMENTAL FOR SECTION I 1.2  ESR Experiments  1.  Reagents A l l  reagents  purification: sodium  fetuin  periodate  (Aldrich),  2.  from  IV,  Sigma),  sodium  G-25 medium  Centriflo  and used w i t h o u t  BSM ( B o e h r i n g e r  cyanoborohydride  (Pharmacia),  500 U / m l ) ,  CF 25 c o n e s  further  (Aldrich),  Vibrio  galactose  Mannheim),  cholera  oxidase  (retention  TEMPAMINE  limit  (Sigma,  25000  227  amu)  were  Amicon.  ESR Measurements ESR  spectra  spectrometer integrated  and  in  freshly  scan  rate  power  -  levels  placed ul).  In a  were flat,  X-band  absorption Co.  by c u t t i n g  a standardization  samples  settings were  at  Precision  calculated  with  prepared  Spectrometer  recorded  a Pacific  were  comparison  were  the d e r i v a t i v e  with  derivatives  73  (Type  (Behringwerke,  protein).  purchased  purchased  (Fisher),  Sephadex  neuraminidase U/mg  were  -  chosen  of  spin  in  each  non-saturating. high-quality  mode.  curve of  to  A l l  glass  avoid aqueous  cell  signals  integrator;  derived  known  E-3  Derivative  MP-1012A  amplitude,  case  a Varian  out and w e i g h i n g  label  modulation  using  were  second  the peak  areas,  f r o m ESR s p e c t r a  of  concentration.  filter  time  spectral samples  ( J . Scanlon  constant  and  distortions, (50-73 Co. -  ul  and  were  capacity  -  3.  86  -  Fetuin-SL 15 mg of fetuin (4.14 umoles NANA) was dissolved i n 2.7 ml  H 0, 2  to which was added 14.6 umoles NalO^ and the solution kept at 0° for 35 minutes.  The reaction was quenched with an excess of Na S 03 and KI. 2  The solution was concentrated using CF 25 membrane cones.  2  A 15 molar  excess of SL-NH and 100 molar excess of NaCNBH were added i n 1-2 ml of 2  3  solution at pH 8-9; this was kept at room temperature for 2 hours. solution was passed through a Sephadex G-25  column (eluant, 0.02%  The azide;  column dimensions, 2.2 cm (i.d.) * 20 cm) and the excluded volume collected.  The sample was freeze-dried and stored at -4°C u n t i l  spectroscopic measurements were made. 4. Asialofetuln-SL 15 mg of fetuin was dissolved i n 5 ml of buffer (pH 6.9, 0.2 M NaOAc, 0.15 M NaCl, 0.002 M C a C l ) , to which was added ca. 20 units of 2  galactose oxidase and 30 units of neuraminidase.  The l a t t e r was added  in equal aliquots, one at the start of the experiment, and the second a f t e r 24 hours.  The solution was maintained at 37° for 48 hours.  After  C e n t r i f l o CF 25 concentration, the coupling procedure was as for s i a l o - f e t u i n (3).  5. BSM-SL 15 mg of BSM (16 umoles s i a l i c acid) was f i r s t  saponified  (removal of 0_-acetyl substituents on NANA e x t r a c y c l i c t r i o l ) by exposure to 1.5 ml of 1 M NaOH f o r 30 min. at room temperature, and neutralized with 1 M HC1.  Desalting was affected by C e n t r i f l o CF 25 membranes, and  the material exposed to NalO^ (43 umoles) for 35 min. at 0°. procedures were as for fetuin-SL (3).  A l l other  - 87 -  6.  AslaloBSM-SL The native BSM was saponified as above (5) and the neuraminidase  and galactose oxidase treatment performed as i n (4). The coupling of TEMPAMINE proceeded as i n (3). 7.  Erythrocytes-SL Freshly drawn venous blood from a healthy donor was packed, and  to 7 ml of packed erythrocytes, 0.5 ml of 0.01 M NalOi^ was added. solution was kept at room temperature for 15 min.  The  The oxidation was  terminated by the removal of reagents - three washes with physiological saline solution, with RBC p r e c i p i t a t i o n by centrifugation (400 g, 10 min).  A 25 fold molar excess of  SL-NH2 and 100 fold molar excess of  NaCNBH were added and the saline solution (pH 7.5-8) kept at room 3  temperature f o r 2 hours.  With centrifugation and washing, some c e l l  l y s i s was indicated by a s l i g h t l y red supernatant.  The labelled  erythrocytes were dialysed against PBS buffer for 3 days at 4° - the buffer was changed d a i l y .  The c e l l s were spun down and ESR spectra run  on packed erythrocytes. 1.3  *H NMR 2  H NMR spectra were recorded at 61.4 MHz ( B 9.4 T) on a Bruker Q  WH-400 instrument (Karlsruhe, West Germany, and U.B.C.).  Samples were  either dissolved i n d i s t i l l e d water (spectra i n F i g . 1.3.3) or deuterium-depleted water ( A l d r i c h ) .  The ca. 1.5 ml sample, i n a 10 mm  tube, was run at room temperature without field-frequency lock and proton decoupling. A 25 us pulse was used and the acquisition time was 0.85 s. BSM and fetuin were activated by the procedure  detaile  in  - 88 -  Chapter 1.2.  T y p i c a l l y , 125 mg of material was used.  For the  deuteration, the aldehyde-containing compound was dissolved at pH 9, and ca. 2.5 mg NaBD  H  (MSD)  added to the solution at 0°.  allowed to reach room temperature  and l e f t for one hour, when the pH  lowered to ca. 5 to decompose unreacted NaBD^. either on a Sephadex G-25  The solution was  Desalting was  was  performed  (medium) column, or using C e n t r i f l o PM 10  size-exclusion membranes (Amicon).  With a s i a l o f e t u i n , catalase was not  used i n conjunction with galactose oxidase, and this probably accounts for the rather low signal-to-noise (low l e v e l of activation) i n this instance. Fetuin was labelled at lysine residues with deuterium by dissolving the glycoprotein i n a borate buffer solution (pH 9.0, 0.1 at a concentration of ca. 100 mg/200 ml.  Sodium cyanoborohydride  M)  (15  moles/mole lysine) was added, followed by a ca. 200 fold molar excess of hexadeuterioacetone, added drop-wise to prevent protein p r e c i p i t a t i o n . The solution was s t i r r e d overnight at room temperature, and unreacted hexadeuterioacetone removed c a r e f u l l y under reduced pressure.  Further  desalting procedures were as described previously. 1.4  Neoglycoproteins Chemicals were often used as supplied from the chemical  companies.  A l l y l alcohol (Aldrich) was dried over calcium sulphate and  stored under N . 2  Chloroform was rendered alcohol-free by washing with  cone. I^SOit, n e u t r a l i z i n g with a l k a l i washes and dried over 4 A molecular sieves.  Mercuric oxide and bromide were from BDH and mercuric  cyanide from ICN.  Acid resins were from Bio-Rad.  Melting points were  determined on a Fisher-Johns apparatus and are uncorected.  - 89 -  Ozone was generated by a Welsbach Ozonator (90 V with 2 p . s . i . Input 0  2  pressure), dimethyl sulphide was from Sigma and sodium  cyanoborohydride from A l d r i c h .  Bovine serum albumin was purchased  from  Miles and passed down a Sephadex G-25 column prior to use, to remove low molecular weight  impurities.  A l l a l l y l glycosides were prepared by standard methods and checked for purity by TLC, B. NMR and melting points. l  O-Acetylations  were performed with pyridine and acetic anhydride i n the cold. Acetoglycosyl bromides were prepared from the per-O-acetylated sugar using 30% HBr performed  i n acetic acid.  Koenigs-Knorr glycosidations were  either using s i l v e r carbonate  (undecenyl  8-D-acetoglucopyranoside), mercuric cyanide (2-acetamido-3,4,6-tri0_-acetyl allyl-8-D-glycopyranose), or mercuric oxide and mercuric cyanide ( a l l other a l l y l p-D-acetoglycosides). De-O-acetylations were with a methanolic solution of the sugar to which was added a freshly prepared solution of 0.1 M sodium methoxide i n methanol. For example, a l l y l a-D-galactopyranose  was prepared thus.  One  gram of Dowex 50W-X8 (H ) resin was prepared by washing several times +  with methanol (to remove high molecuar weight impurities), and then washed several times with a l l y l alcohol.  Galactose (1.8 g) was dried  (over NaOH i n a vacuum desiccator) and to this was added the resin and 20 ml of dry a l l y l alcohol.  The apparatus was f i t t e d with a reflux  condenser and heated under reflux (B.Pt. 98°) f o r 20 hours, by which time a l l the galactose had dissolved and the reaction judged  completed  - 90 -  by TLC (eluant, CHCl /MeOH/HOAc/H 0, 25:15:4:2). 3  2  The a l l y l alcohol was  removed by rotary evaporation at reduced pressure, and the a-glycoside precipitated p r e f e r e n t i a l l y from an ethanolic solution of the o i l at -4°.  The isolated y i e l d was 10%, and the melting point 146-148° ( L i t .  143-145°). The synthesis of a B-glycoside i s t y p i f i e d by a l l y l B-fi-glucopyranoside.  2,3,4,6-Tetra-O-acetyl-a-g-glucopyranosyl  bromide  (1.5 g) was added to a mixture of 10 ml dry a l l y l alcohol, 10 ml alcohol-free chloroform, 1.0 g " D r i e r i t e " , 0.65 g HgO and 0.05 g H g B r 2 , and s t i r r e d for 48 hours.  The mixture was f i l t e r e d through a layer of  C e l l t e , and evaporated to dryness.  The sugar was extracted into dry  chloroform; any remaining mercuric s a l t s were precipitated and removed by f i l t r a t i o n .  The f i l t r a t e was evaporated to dryness and the product  c r y s t a l l i z e d from an ethanolic solution at -4°, i n 70% overall y i e l d . M.Pt.  86° ( L i t .  86/88°).  The ozonolysis was performed as follows.  The sugar was dissolved  in methanol (unprotected) or MeOH/CH2Cl2, 4:1 (protected sugar) at a concentration of 1 mmole per 5 ml solvent.  The solution was cooled to  -60° (acetone/dry i c e ) and ozone-enriched oxygen bubbled through (5-10 min/mmole) u n t i l the methanolic solution went pale blue, i n d i c a t i n g an excess of ozone.  The solution was flushed with N2 to remove this ozone,  and excess dimethyl sulphide (DMS; > 2 mole/mole sugar) added. The solution was removed from the coolant and allowed to reach room temperature, where i t was s t i r r e d for 2 hours. methanol were removed under reduced pressure.  Excess DMS, methanal and With blocked sugars, the  - 91 -  sirup was dissolved i n methylene chloride and washed several times with water to remove DMSO. In a t y p i c a l reductive amination reaction, BSA (18 mg; 0.015 mmole l y s i n e ) , aldehydo sugar (16.2 mg; 0.074 mmole) and NaCNBH (22.8 3  mg; 0.38 mmole) were dissolved i n 6 ml water, and the pH adjusted to 9.0 with 0.1 M NaOH.  The solution was l e f t overnight, and then passed down  a Sephadex G-25 column (0.02 % sodium azide eluant), where the excluded volume was c o l l e c t e d .  This was freeze-dried and the neoglycoprotein  analyzed by the phenol-sulphuric acid method, measuring the absorption at 490 nm. With a blocked sugar, the procedure was i d e n t i c a l except that the reaction solution was H 0/MeOH, 3:1. The blocked aldehydo sugar (In 2  MeOH/H 0, 2:1) was added over 10 min. 2  Methanol was removed under  reduced pressure taking care to minimize frothing.  I f any solids were  suspended i n the solution, they were f i l t e r e d off prior to desalting, usually by d i a l y s i s (3 days against d i s t i l l e d H 0). 2  2-acetamido-3,4,6-trio-0_-acetyl  Note that  allyl-8-D-glucopyranose Is moderately  soluble i n water, and methanol was not necessary In the reaction solution. 1.4.2  1 3  C NMR  Most neoglycoprotein spectra were recorded on solutions (ca. 150 mg/1.8 ml D 0) at 308 K. 2  The spectrometer  equipped with a single-frequency probe f o r  used was a Bruker WH-400, 1 3  C (100.6 MHz). The  spectral width was 200 ppm and the recycle time ca. 0.8 s. An e x c i t a t i o n pulse (18 us for neoglycoproteins or 12 us f o r  - 92 -  monosaccharides; data-points,  90°  while  Signal-averaging For was  added  (Sigma; left  the  to  generally  U)  the  was  to  took  studies,  20 U / 1 5 0  60,000  us)  broad-banded  enzyme  (ca.  open  «* 2 2  followed  by  the  acquisition  *H d e c o u p l i n g  was  employed.  place  overnight  galactose  oxidase  mg n e o g l y c o p r o t e i n )  ca.  1 5 0 mg o f  atmosphere  (0 )  160 000  (Sigma  together  at  room  or  with  neoglycoprotein.  overnight  2  (ca.  of  The  16  K  scans).  Worthington) 5 ul  catalase  solution  temperature,  was  and  the  13 C NMR  s p e c t r u m was  reduction  1.4.3  was  recorded.  performed  (2  the  dissolving  in  exchangeable  * H NMR D 0  (SEAS)  and  2  from  sodium  borohydride  MCB).  protons.  The  spectrometer,  6.35  Bruker  T),  a  U.B.C.  units  (Dr.  with  was  deuterium  and Mr.  400  spectra  the  use  of  exponential recorded. typical  were a  3  acquired  trapezoidal  for  and  in  T.  90°  and  180°  values  are  10  us  were  time,  recorded  to at  1180  Marcus).  relaxation tau  (NTCFTB  delay  -  at  and  21  were us,  magnet and  widths  checked  1.2  spectra  before  respectively.  proton  were  HOD  used.  and  (BQ  293B  and  each  3012  resonance Typically,  Data-processing 55")  a  Chemistry,  the  was  Magnitude-calculation pulses  of  Spectral  value. "Tl  using  detection,  Presaturatlon  and  2  2 7 0 MHz  Department  by  remove H 0  computer  quadrature  the  prepared  superconducting  Nicolet  lock,  each  window  line-broadening. The  no  several were  collected.  seconds;  samples  an O x f o r d  manufactured  Morris  for  on  The  8K d a t a - p o i n t s  frequency  based  console,  were  G.A.  spectra  WP-60  pulse-programmer. decoupler  experiments,  freeze-drying  home-built  and  mg,  specified,  *H SEAS Experiments For  Hz,  Where  involved  Hz were experiment,  -  93  -  SECTION II  - 94 -  CHAPTER I I . 1 INTRODUCTION  High-resolution nuclear magnetic  resonance (NMR)  spectroscopy has  deservedly earned a place as one of the most universal and powerful tools i n organic s t r u c t u r a l determination.  The NMR  reported for bulk solids by Bloch and coworkers coworkers  2  1  phenomenon was  first  and P u r c e l l and  In 1945, who were j o i n t l y awarded i n 1952 the Nobel prize for  t h e i r discovery.  Improvements i n magnetic  field  homogeneity soon  resulted i n measurement of higher resolution spectra, which revealed f i r s t the chemical s h i f t , and then the spin coupling phenomena. Ernst and Anderson  3  demonstrated  In 1966  the application of Fourier  transformation (FT) of pulse responses, and along with signal averaging t h i s led to further increases i n s e n s i t i v i t y .  With the concomitant  integration of d i g i t a l computers, i n s e n s i t i v e but important nuclei such 13  as  15  C and  N could be routinely observed.  The next major land-mark i n the development of high-resolution NMR  was  the application of super-conducting materials to magnet  construction.  Starting In the late 1960's the then common 100  instruments were soon superceded by 200-360 MHz 1970'8, 400 MHz  devices, and by the late  instruments were common, and 500 and 600 MHz  were being described at NMR  congresses.  MHz  These high f i e l d  instruments  instruments  gave the anticipated increases In dispersion and inherent signal-to-nolse. At the conceptual l e v e l , perhaps the most s t a r t l i n g development came i n 1971, when Jeener suggested  the p o s s i b i l i t y of  two-dimensional  - 95 -  (2D) NMR  experiments.  I n t e r e s t i n g l y , although i t was  f i v e years before  the detailed quantum-mechanical theory of the experiment was  presented,  5  many other experiments were described i n the l i t e r a t u r e , most from the research groups of Ernst, i n Switzerland, and Freeman, i n England. Although the technological c a p a b i l i t y to perform most of these experiments has been commercially  available f o r at least five years  (since 1978), s u r p r i s i n g l y few reports on their applications have appeared.  In the b i o l g i c a l area, the collaborative e f f o r t s of Ernst and  Wuthrich have amply demonstrated the potential for studies of small proteins.  6  H a l l ' s group i n Canada had an early interest i n the  evaluation of the technology as applied to s t r u c t u r a l organic chemistry; t h i s thesis forms a part of that t r a d i t i o n . In t h i s work, the author has undertaken a systematic, objective appraisal of some of the available 2D NMR  experiments, i n the l i g h t of  t h e i r a p p l i c a t i o n to structural investigations i n organic and natural-products chemistry.  I t became evident early on that 2D  experiments were not a panacea to a l l the problems associated with the investigation of complex molecules, and modified forms of certain ID experiments found an important place i n the general protocols which emerged.  In a l l cases, the fundamental objective was  the e f f i c i e n t  integration of spectroscopic methods, the most expeditious route to the required answers, and the conditions to which each experiment was  best  suited. It was decided that before attempting  to apply the methods to  complex molecules, a thorough assessment of each experiment using a  -  96  -  single simple exemplar would prove b e n e f i c i a l .  This lengthy  investigation Is detailed i n Chapter II.2, where the data are along with t h e o r e t i c a l aspects. was  presented  A simple protected o-glucopyranoside  chosen, since i t i s conformationally r i g i d and has a  simple, f i r s t - o r d e r spectrum at 270 or 400 MHz.  reasonably  As w i l l be seen, these  experiments are an important, i n t e g r a l component of the research of the thesis, since an equivalent array of experiments has not previously been systematically applied to a single molecule.  Furthermore, because t h i s  i n i t i a l investigation of the procedures so successfully provided a broad base of expertise and experience,  a l l subsequent studies could be  performed rather e f f i c i e n t l y , and the descriptions of subsequent chapters are correspondingly  concise.  This was  a very welcome  development. Chapter II.3 describes a p r o t o c o l ' based on NMR sequencing of oligosaccharides.  Vliegenthart's  have already exposed some of the power of NMR  for the de novo  and Dabrowski's  groups  i n the s t r u c t u r a l analysis  of glycopeptldes and ceramlde oligosaccharides, respectively, where they found great u t i l i t y .  Dabrowski's more recent work has elements In  common with our procedures i n that reliance on chemical i n Vliegenthart's approach as a basis for assignment was intention was  s h i f t s Implicit avoided.  to develop a general procedure whereby the complete  primary structure of an oligosaccharide could be determined by NMR only reliance on t y p i c a l v i c i n a l coupling constants ring protons.  The  with  for the carbohydrate  As w i l l be seen, the combination of ID and 2D methods  found suitable proved more than adequate for the disaccharide glycoside, which was  used as a model.  A more complex b i o l o g i c a l material, digoxin,  -  was  97  -  studied i n Chapter II.5. Chapter II.4 deals with the assignment and  i n v e s t i g a t i o n of brucine.  The main objective here was  expeditious route to the assignment of i t s *H and conformational  structural  information on the molecule.  i3  C  to find the most NMR  spectra, and  Further, the author also  attempted to place himself i n the position of a professional applied spectroscopist to evaluate the performance of the required experiments under "survey conditions".  This required the minimum knowledge of the  molecule's spectroscopic c h a r a c t e r i s t i c s before performing under "routine" conditions where demands on instrument consideration.  experiments  time i s a major  This i s i n sharp contrast to the developmental situation  in Ch. II.2, which i s conceptually akin to a spectroscopy  research  study. The  f i n a l study i n Chapter II.5 i s on a s t e r o i d a l glycoside, 1  digoxin.  Here again, the object was  to assign the  spectra using the currently available techniques.  13  H and  C  NMR  The glycone i s  considered as a c r i t i c a l test of the oligosaccharide sequencing methodology described i n Ch. II.3. sugars, the s i t u a t i o n i s complicated  Although i t consists of only three by the fact that they are i d e n t i c a l  In nature and linkage, and are 2-deoxy sugars with methylene resonances overlapping with the bulk of the steroid protons.  The  cardenolide  (steroid genin) i s equally Intractable because of i t s low extent of substitution and conformation.  As w i l l be seen, these complexities are  compounded by the fact that the molecule self-associates i n a polar solvent, giving broad lines for the steroid moiety, especially  - 98 -  troublesome In the H NMR spectrum,  the molecule was studied at 500 MHz  and, f o r the f i r s t time i n this work, s i g n i f i c a n t limitations became evident In some 2D experiments. Anticipating somewhat the conclusions of Section I I , the author believes that this thesis points to an increasingly bright future f o r NMR spectroscopy.  It i s clear that certain protocols f o r structure  determination are extremely powerful under certain regimes; thus, rapidly tumbling, medium molecular weight molecules can most c e r t a i n l y be studied under "survey conditions", demanding comparatively l i t t l e instrument time.  However, although some larger molecules can d e f i n i t e l y  benefit from use of the highest obtainable s t a t i c f i e l d and a modified approach, some i n t r i n s i c limitations e x i s t .  - 99 -  REFERENCES 1.  F. Bloch, W.W. Hansen, and M. Packard.  Phys. Rev. 69, 127 (1946).  2.  E.M. P u r c e l l , H.C. Torrey, and R.V. Pound. (1946).  3.  R.R. Ernst and W.A. Anderson.  4.  J . Jeener. Ampere International Summer School, Basko P o l j e , Yugoslavia. 1971.  5.  W.P. Aue, E. Batholdi, and R.R. Ernst. 2229-2246 (1976).  6.  G. Wagner, A. Kumar, and K. Wuthrich. Eur. J . Biochem. 114, 375-384 (1981), and others i n the series.  7.  M.A. Bernstein and L.D. H a l l . (1982).  8.  J.F.G. Vliegenthart, H. van Halbeek, and L. Dorland. Chem. 53, 45-77 (1981), and references therein.  9.  J . Dabrowski, P. Hanfland, and H. Egge. (1981), and references therein.  Phys. Rev. 69, 37-38  Rev. S c i . Instr. 37, 93-102 (1966).  J . Chem. Phys. 64,  J . Amer. Chem. Soc. 104, 5553-5555  Pure Appl.  Methods Enzymol. 83, 69-86  - 100 -  CHAPTER II.2 HIGH RESOLUTION NMR METHODS  II.2.1  General Theory Although some understanding by the reader of pulsed Fourier  Transform (FT) high resolution NMR  w i l l be assumed, i t i s necessary  to b r i e f l y review a few relevant concepts to ensure that the reader has access to the relevant nomenclature.  In what follows below, we  Introduce three models, each of which gives d i f f e r e n t insight to the e f f e c t of a pulse of radlofrequency energy on an ensemble of nuclear spins. We consider an ensemble of Identical spin-1/2 nuclei i n a s t a t i c magnetic  f i e l d , J J . Two energy l e v e l s , o and B , w i l l e x i s t , d i f f e r i n g 0  in energy by AE: AE - yfc B Y i s the magnetogyric  [II.2.1]  0  r a t i o of the nuclide, and fi i s the reduced  Planck's constant (h/2n).  At thermal equilibrium, nuclei are  distributed between the two energy l e v e l s by a Boltzmann d i s t r i b u t i o n , with the resultant net macroscopic magnetization termed My. It i s convenient to define a frame of reference as shown i n F i g . II.2.1A.  The applied magnetic f i e l d , BJJ i s defined to l i e p a r a l l e l  to the z-axis. Magnetic moments (vectors), u, are required to precess about JB at the Larmor angular v e l o c i t y , co : 0  0  too " "Y B  0  [II.2.2]  Expressed i n frequency units, the Larmor precessional frequency, v , 0  - 101 -  i s given by: vo - -y Bo/2 n  [II.2.3]  There i s no phase coherence i n the x-y plane and consequently no resultant magnetization; the resultant magnetization, MQ, l i e s on the z-axis. To induce resonance, a second radio frequency (RF) f i e l d , j3\, must be applied i n the x-y plane.  Resonance i s attained when the  frequency of B>i exactly matches VQ. It i s now convenient to redefine the terms of reference i n F i g . II.2.1.A.; now the reference frame i t s e l f i s rotating about the z axis with the same sense and frequency, v, as the RF f i e l d , JJi.  When v » vo,  magnetic moments are e f f e c t i v e l y motionless, or moving slowly (vo - v ) . The x and y axes are r e - l a b e l l e d x' and y' to make the d i s t i n c t i o n , as in F i g . I I . 2 . 1 . B (where v «• vo).  This i s the "rotating reference  frame". Let  us now consider the effect of B_i on Mn> placed along the x'  axis, with the resonance condition met. tipping i t away from the z axis. (the  "pulse length"), M  0  Bj exerts a torque on MQ,  I f B^ i s applied for a time, tp*  rotates about x' by an angle G, defined by:  *A square pulse of length jtp (time domain) of frequency vo corresponds, i n frequency space, to a range of frequencies centred at vo - t " to vo t " . Hence, with t t y p i c a l l y of ~ P ~ P ~P the order of microseconds, a range of frequencies extending over tens or hundreds of KHz w i l l e x i s t . and extending from  1  V n  +  1  -  Fig.  II.2.2  A  102  stylized  -  AX  spin  system  - 103 -  9 - Bi t Y/2H — -p There i s now  a component of M along the y' axis, My.  [II.2.4]  This perturbs  the spin system from i t s thermal equilibrium with respect to i t s surrounding (the l a t t i c e ) . The Boltzmann equilibrium state i s re-established by leakage of magnetization into the l a t t i c e ; the f i r s t - o r d e r rate-constant for this process i s T j , the s p i n - l a t t i c e (longitudinal) relaxation rate.  A  second rate constant, T_ (the spin-spin, or transverse relaxation rate) 2  describes the exponential decay of transverse magnetization i n the x'-y' plane. Following application of the perturbing pulse, measurement of the magnetization i n the x'-y' plane as a function of time produces the free-induction decay (FID) - this measure of My vs. time, has the form  Mo cos(tot) exp(-£Ar ). 2  Fourier-transform (FT) of this time-dependent  function yields the frequency-space line.  spectrum as a single Lorentzian  This i s the key to pulsed FT NMR  spectroscopy.  Let us now b r i e f l y look at a homonuclear AX system, In the rotating reference frame, with the spectrometer frequency, to, placed close to that of the h i g h - f i e l d t r a n s i t i o n of spin A; the four l i n e , frequency-domain spectrum i s s t y l i z e d i n F i g . II.2.2. F i g . II.2.3 follows the behavior of the magnetization vectors i n the x-y plane, Ignoring T_j relaxation.  At thermal equilibrium, a l l four  vectors are aligned along the z-axis ( F i g . II.2.3A).  After application  - 104 -  c. Fig. II.2.3  The action of an AX spin system's vectors, (A) at equilibrium, (B) after a 90° pulse and (C) after some time to allow dephasing. (Solid l i n e denotes A spin vectors and broken l i n e s , X).  A.  F i g . II.2.4  B.  Energy Level diagrams of AX spin-system (A) showing t r a n s i t i o n p r o b a b i l i t i e s (B). 0 and ±6 represent r e l a t i v e equilibrium populations.  - 105 -  of  a perfect 90° pulse to a l l spins, the vectors l i e along y' (Fig.  II.2.3B).  Since a l l four vectors have a d i f f e r e n t angular v e l o c i t y than  that (to) of the rotating reference frame, with time these precess i n the x'-y' plane ( F i g . II.2.3C), and dephase with respect to each other. average of l i n e s  to  i2  and  to  31+  The  w i l l precess at angular v e l o c i t y , 0^  (the difference between to and 10*) and s i m i l a r l y , to,, and to„, precess slower at  Each doublet w i l l have a r e l a t i v e l y slow (s)  and fast ( f ) precessing component. provides a convenient  As we s h a l l see l a t e r , this model  description of the concepts of spin precession and  phase coherence. We now consider the same spin system using the "energy-level" model. and  The homonuclear AX system has four energy l e v e l s :  oca (aB, pa)  88 i n increasing energy,* as depicted i n F i g . II.2.4A.  In t h i s  model, the 90° pulse Induces the four single-quantum t r a n s i t i o n s , ( F i g . II.2.4B) and i t i s these which are detected. double quantum t r a n s i t i o n s (W  0  and W  2  Note that the zero- and  respectively) are not induced by  t h i s single pulse experiment. Although the rotating reference frame and energy l e v e l models discussed above are simple and provide reassuring, p i c t o r i a l explanations for many NMR experiments, they are often i l l - s u i t e d for explanations of many processes.  For that reason, the density matrix  formalism, which i s more complicated but exact i n a l l cases, w i l l now be  *In the heteronuclear case, the aP energy state does not have the same energy as 6a. In the homonuclear case their energies are almost equal, as indicated by their equal equilibrium populations.  - 106  b r i e f l y introduced.  -  For s i m p l i c i t y , i t i s worthwhile using a  diagrammatic formalism.  We w i l l ignore relaxation e f f e c t s , as t h i s  s i m p l i f i e s the discussion. An AX system i s generally represented by a four-by-four matrix, a, constituting a l l the possible products of the four energy states:  aa  aB  Ba  BB  oa  oil  a  °13  a  aB  °21  a ^  8a  °"31  a  BB  a m  o-»*2  1 2  32  0  23  a  3 3  \u,  a  2 1 +  [II.2.5] a  34  °"<+3  At thermal equilibrium, the density matrix, c(0), w i l l have unequal Boltzmann population terms (P°) along the diagonal, as i n Eq.  II.2.6a.  Inserting the r e l a t i v e populations (see F i g . II.2.4A) gives the matrix Eq. II.2.6b, i n which 6 i s the population difference between the a B or B a state, and a a or B B .  p°  0  0  0  0  pO  0  0  0  0  p°  0  0  0  0  p°  [II.2.6a]  o(0)  - 107 -  6  0  0  0  0  0  0  0  0  0  0  0  0  0  0  -6  [II.2.6b]  a(0)  Consider now new  the effect of the pulse (rotation) on o(0).  The  density matrix, a', i s given by: a' = R_ o(0)  where R  +  are rotational matrices.  R  [II.2.7]  +  For a 90° pulse, R  90 +  w i l l take on a  special form and, omitting the d e t a i l s , we obtain, e x p l i c i t l y , Eq. II.2.8a which may  be s t y l i z e d as i n Eq. II.2.8b.  216  2i6  0  -216  0  0  2i6  -216  0  0  0  [II.2.8a] 0  -216  -216  2i6 0  [II.2.8b]  - 108 -  If the reader compares Eq. II.2.8a with Eq. II.2.5, i t w i l l be clear that only single-quantum excited.  transitions have non-zero values, and hence are  The zero- (o- 3» ^32^  o r  2  double- ( o ^ . a^\) t r a n s i t i o n s have  zero values. Consider now the behaviour of these components as the matrix i s allowed to "evolve" with time, t .  The 012 term, for example, previously  2i6 now becomes 2i6 exp ( i (012 t ) . The exponential term contains s i n and cos contributions and we may  loosely represent the "detected"  matrix, o ' ( t ) , as:  o'(t)»  [II.2.9]  - 109 -  In 2D experiments at least 2 pulses are necessary and where necessary, these w i l l be discussed i n light of the density matrix formalism introduced i n i t s s t y l i s t i c form here.  II.2.2  Spectral S i m p l i f i c a t i o n The purpose of t h i s section i s to introduce some of the methods  which are available to "simplify" otherwise complex NMR spectra.  II.2.2.1  V i a T^ i n *H NMR  Spectroscopy * 5  6  Spectra can often by simplified by taking advantage of the fact that d i f f e r e n t types of protons i n a molecule frequently have widely d i f f e r e n t s p i n - l a t t i c e relaxation times.  For example methylene signals  can quite e a s i l y be "separated" from methine, when overlap causes complications and impedes spectral assignment.  The concept rests on the  fact that methylene protons w i l l relax faster than methine protons, and w i l l a l l relax at approximately  the same rate, since their mutual  relaxation i s so dominant. Consider the Inversion recovery  T_ experiment (T1IR); the pulse x  sequence i s : (RD - 180° - ^ - 90° - AQN) , n  where RD i s a relaxation delay which Is set >^ 5 * T_i (longest). According to the rotating reference frame model ( F i g . II.2.5), the 180° pulse inverts a l l spin populations, placing the Mo vector along the -z axis.  During the delay time, t, M^ returns towards i t s equilibrium  - 110 -  D. Fig. II.2.5  Fig. II.2.6  F.  The rotating reference frame vector model of the T_i inversion-recovery experiment. Vectors are shown for a short t_ value (A,B,C,E) and a r e l a t i v e l y long one (A,B,D,F).  Plot of z-magnetlzation vs time i n a T i inversion recovery experiment, indicating the n u l l point.  -  position at the rate, expC-t/T^).  Ill  -  Since only x'-y' magnetization can be  detected, a second 90° ("read") pulse i s applied to determine the magnitude of Mg, the z magnetization.  This two-pulse sequence i s  performed f o r a number of d i f f e r e n t t_ values. trldeuteriomethyl  Results for  2,3,4,6-tetra-0j-(trideuterioacetyl)-a-I>-  glucopyranoside (1) are given i n F i g . II.2.7.* t_ values the peaks are negative-going  1 : R - C0.CD , R* - CD  3  2  3  3  : R - CO.CH3, R' -  CH  We can see that f o r short  (Figs. II.2.5C, E) eventually  becoming positive-going (Figs. II.2.5D,F).  A t^ value exists where 1 ^  • 0 at the time when the "read" pulse Is applied, I.e. the " n u l l condition" ( F i g . II.2.6).  This i s the basis for this form of "spectral  *At this stage we are not concerned with the q u a n t i f i c a t i o n of J_i although i t i s worth mentioning that these data may be used to gain useful information concerning inter-proton distances for molecules i n solution (vide i n f r a ) .  -  Fig.  II.2.7  400  MHz  T1IR  relaxation each  112  -  experiment  delay  spectrum.  was  12  on  1 (0.1  s,  and  M in values  C D ). 6  6  noted  The next  to  - 113 -  METHYLENE  Fig. II.2.8  Spectral editing of 1 by T1IR. The control spectrum i s C. jt • 0.7s Is a "methine sub-spectrum", with methylene signals nulled. _t • 2.5s gives the "methylene sub-spectrum", with methine signals p a r t i a l l y nulled. Other parameters as i n F i g . II.2.7. The asterisk marks an a r t i f a c t at the c a r r i e r p o s i t i o n .  - 114 -  editing".  A preliminary T1IR experiment with a series of a r b i t r a r i l y  selected t_ values Is performed to determine the n u l l time for the - C H 2 resonances, f o r example.  One of the resultant spectra might  accidentally give a good n u l l i n g of the - C H 2 resonances under consideration.  I f not, a single experiment may be performed to achieve  good n u l l i n g ; under these conditions the more slowly relaxing -CH resonances w i l l then have negative amplitudes.  An attempt to edit  methine peaks i s not as easy, since t h e i r T_i's may vary considerably. F i g . I I . 2 . 8 i l l u s t r a t e s these statement. shown i n F i g . I I . 2 . 8 C .  The normal spectrum of 1 i s  Choosing _t « 0 . 7 s gives a "methine  sub-spectrum", with methylene signals nulled (Fig. I I . 2 . 8 A ) .  With t •  2 . 5 s , both methylene protons are almost f u l l y relaxed whilst methine protons are either nulled (H-4) p a r t l y inverted (H-l, H-3) or p a r t l y upright ( H - 2 ) ,  (H-5) ( F i g . I I . 2 . 8 B ) .  Thus, a suitable choice of _t allows the creation of a spectrum lacking methylene peaks.  This procedure has been used e f f e c t i v e l y i n  the assignment of the complex spectrum of a steroid.  Although i t i s  generally impossible to n u l l a l l methine signals simultaneously, I t Is possible to do a number of separate experiments and choose the _t values which n u l l overlapping resonances.  This method constitutes a simple and  easy, yet l i t t l e - u s e d solution to the hidden resonance problem. With  1 3  C NMR, t h i s approach may be used to determine the number  of attached protons to a carbon  9  (Chapter I I . 4 ) .  - 115 -  II.2.2.2  Via J  C  In  T  1 3  C  NMR  Spectroscopy  Although broad-band proton decoupled  13  C NMR  spectra are  a t t r a c t i v e i n their s i m p l i c i t y , a price i s paid for the removal of the 13  1  C- H coupling information i n that the spectra are d i f f i c u l t to 13  assign.  The f i r s t  steps towards assignment of  C NMR  spectrum involves  determining the chemical s h i f t s , and the number of protons attached to each carbon.  Single-frequency-off-resonance-decoupling  widely used to achieve the l a t t e r purposes;  10  (SFORD)* i s  however, i t has serious  l i m i t a t i o n s i n crowded regions of a spectrum, i s inherently i n s e n s i t i v e and can produce misleading r e s u l t s .  1 1  12— 18  Fortunately, a number of multi-pulse experiments exist ~ which are easy to use, are not as time consuming as SFORD and give a much clearer representation of the information.  The f i r s t of these i s  12  "refocussed INEPT  " ( i n s e n s i t i v e nuclei enhanced by p o l a r i z a t i o n  t r a n s f e r ) ; although this works well with decoupled  spectra, the  observed  t r a n s i t i o n s are distorted when the J TJH couplings are preserved.  A  1  J  number of other variants have been described; however, the author 13  had the opportunity to try only one of these, dubbed DEPT ( d i s t o r t i o n l e s s enhancement by p o l a r i z a t i o n _transfer).  19  »  has  20  »  This experiment  gives undistorted coupled spectra, as well as good e d i t i n g . " P o l a r i z a t i o n transfer" i s a process i n which magnetization from the protons, which are high natural abundance and sensitive nuclides, i s *SFORD employs a single *H decoupler frequency placed (usually) to h i g h - f i e l d of the H spectrum. This causes suppression of the one-bond C- !! couplings 1  -  116  -  transferred to the r e l a t i v e l y insensitive i n t e n s i t y enhancement i s  YH^YC  "  C nuclei.  The maximum  compared with the maximum nOe  attainable by broad-band H decoupling l  (1 +  ( 1 / 2 ) YH/YC  may not seem to be a s i g n i f i c a n t improvement.  "  ^)  T  N  I  S  However, a further  advantage l i e s i n the fact that the repetition rate for the experiment i s governed by the s p i n - l a t t i c e relaxation rates of the protons rather than of the  13  C spins; given that  1 H relaxation rates are t y p i c a l l y  13 faster than those of unit time.  C , 2 - 3 times as many scans can be obtained i n  Other advantages of DEPT over INEPT are that the former i s  unhindered by overlap of solute and solvent peaks (which i s p a r t i c u l a r l y important when working at high d i l u t i o n ) and that i t has a reduced dependence on the magnitude of ^JjZE' 1  H e n c e  »  a  "average" value f o r  n  J_CH may be chosen which produces good results for a l l carbon s i t e s i n  a molecule. The pulse sequence for the DEPT experiment i s presented below. Proton pulses are applied through the *H decoupler c o i l s of the probe, at the normal offset frequency for broadband ( i . e . near 6JJ = 5 ) .  1  3  C:  90°(«i) - ( 2 J )  3  C  (BB) decoupling  An average value of the * J coupling i s used i n  the c a l c u l a t i o n of the *H:  1  _ 1  ( 2 J )  -  -  1  value, giving  180°($ ) 2  90°($3)  -  (2J)  - 1  3.6  -  ms.  e($ > 3  A l l pulses, and the "  180 ($it) o  (2J)"  1  - (BB) AQN($ ) 5  [II.2.10]  -  -  117  Phase cycling: Cycle 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  *1  X  X  X  X  X  X  X  X  -x  -x  -x  -x  -x  -x  -x  -x  $2  X  -x  X  -x  X  -x  X  -x  X  -x  X  -x  X  -x  X  -x  y  y  y  y  y  y  y  y  y  y  y  y  y  y  y  y  *4  X  X  -x  -x  y  y  -y  -y  X  X  -x  -x  y  y  -y  -y  *5  X  X  -x  -x  y  y  -y  -y  X  X  -x  -x  y  y  -y  -y  $  3  [II.2.11]  receiver are phase-cycled.($i - $5) through sixteen steps for the best cancellation of a r t i f a c t s .  This appears complex on paper, but i s quite  simple to programme on most modern spectrometers. 13  C  Acquisition of the  FID ("AQN") i s often with broad-band proton decoupling ("BB").* Spectral editing i s achieved by acquiring three spectra, for  (1)  0i  -  45°  (TC/4), ( 2 )  0  2  -  90°(TI/2)  and  (3)  0  3  =  135°  (3it/4).  scans are acquired for 0 i and 0 3 , 2_n should be acquired for 0 . 2  If  n  The  free induction decay signals are Fourier transformed, phased, and the edited spectra obtained by l i n e a r combinations -CH -CH  :  2  -CH3  0  of the three spectra:  2  :  0 1 - 0 3  :  01 + 03 -  0.707  02  [II.2.12]  *The abbreviations "AQN" ( a c q u i s i t i o n ) , BB (broad-band proton decoupling) and $ (phase-cycling programme) w i l l be freely used i n pulse schemes i n the rest of this chapter. x  - 118 -  CH.  CH  2  CH  CH  100  90  Fig.  II.2.9  DEPT  •OCH  J  X  80  performed  on 2  3  70  (0.4  M;  CDC1 ) 3  60  at  100  MHz.  - 119 -  These additions/subtractions may the -CH2  subspectrum  varies between 1.00  not be t o t a l l y exact;  f o r example,  i s more correctly given by ©i - a. 63, where "a" and 1.10.  Most modern spectrometers have a "real  time" subtractive routine which allows the operator to "fine-tune" these values to achieve most e f f i c i e n t e d i t i n g .  One instrument  manufacturer  20 boasts an i t e r a t i v e routine which finds the best subtracted specta. Although the question of proton-pulse c a l i b r a t i o n i s deferred u n t i l Section II.2.5.4, It i s important to point out the -CH sub-spectrum  (62)  i s reasonably sensitive to mlsset i n the proton f l i p  angles (vide i n f r a ) and as a result can be used to determine lengths with reasonable accuracy.  *H pulse  In practice, the experiment i s  repeated for a number of pulse lengths close to the approximate  value,  and the transformed spectra inspected for the most e f f i c i e n t cancellation of a l l but methine carbon  peaks.  Results are presented i n F i g . II.2.9 for DEPT performed  for a  =0.4 M solution of methyl 2,3,4,6-tetra-O-acetyl-a-D-glucopyranoside, in  CDCI3.  The -OMe  peak appears as a weak, aliased peak at  However, we see that i t Is c l e a r l y v i s i b l e i n the -CH3  6  2  63.2.  sub-spectrum;  n u l l i n g i s excellent. II.2.3  Spin Decoupling-Difference S p e c t r a * 6  2 1 - 2 3  (SDDS)  This i s a variant of the well known spin decoupling experiment, i n which responses are observed only from those resonances  directly  - 120 -  involved i n spin coupling; a l l others are "nulled" by the "difference" method.  The advantages of this mode of presentation only become obvious  when one considers crowded spectra, which have overlapping resonances. In general, interpretation i s limited to observation of spectral changes; detailed analysis of the decoupled difference peaks may not be easy by inspection, but may  be computer-simulated  for v e r i f i c a t i o n .  In practice two spectra are acquired for each decoupling experiment,  one with the decoupler set to the precise resonance  frequency and the other, the "control", with i t set just off resonance. To minimize instrumental d r i f t over a long period of time, the data are collected i n blocks, which may be repeated i f long-term signal-averaging i s necessary to build up adequate signal-to-noise. termed "interleaving". performed  This process i s  I f a series of separate experiments  are to be  at one time, the appropriate decoupling frequencies are l i s t e d  i n the computer memory, together with decoupler power settings adequate for homonuclear decoupling. A t y p i c a l sequence of events would be: 1.  Where 11 resonances  require decoupling, create on disc 2n  blank f i l e s , specifying decoupler frequencies and power settings. 2.  Read into core memory f i l e number 1.  3.  Perform 2-4  "dummy" scans (discarded), to e s t a b l i s h a  steady-state. 4.  Acquire ^ scans (8 or 16) with the decoupler on  resonance.  - 121 -  5.  Save the (on-resonance) decoupled free induction decay on disc.  6.  Read into core memory f i l e 2.  7.  Acquire ^ scans with the decoupler  8.  Save the (off-resonance)  set just off resonance.  "control" free induction decay on  disc. 9.  Repeat steps 2-8 (n-2) times, reading i n each appropriate file.  10.  Repeat 2-9 m times, so as to build up s u f f i c i e n t signal-to-noise.  The t o t a l number of scans for each  difference spectrum i s 2(A * m). Note that the t o t a l number of scans (A * m) need only be equal or s l i g h t l y more than that required to obtain a normal spectrum with adequate signal-to-noise.  A l l free induction decays are then  exponentially weighted ( l i n e broadening, 1.5 - 2.0 Hz), Fourier transformed,  phased, and the frequency-domain spectra subtracted.  The  phase of the subtracted spectrum may require minor a l t e r a t i o n before plotting. Figure II.2.10 i l l u s t r a t e s SDDS on 1.  Several points a r i s e .  We  see that i n crowded regions ( H - l , -2, -4, and H-5, -6 and -6•) the r e s u l t s are not always clear-cut, due to Bloch-Slegert e f f e c t s . *  *In the presence of the decoupling f i e l d B 2 , a l l resonance frequencies are shifted from their "true" frequency by an amount which i s proportional to the decoupler power (yB^/ZTt), and inversely proportional to their frequency separation. This i s part of the Bloch-Slegert e f f e c t .  - 122 -  JI  H-6  H-6  4-  H-5  41—1*-  nli"  T  ~}\—  -AA  H-4  H-3  A-  B.S.  H-2 H-l H-3  H-4 H^2H-1  JL_JULL 6.0 II.2.10  5.0  H-6 H-6'  i±5L 4.0  SDDS on 1 at 400 MHz. The i r r a d i a t e d proton i s indicated i n each case. False responses a r i s i n g from Bloch-Siegert effects are labelled "B.S."  - 123 -  Computational solutions to these problems have been proposed.  With  {H-2}, for example, the calculated s h i f t i n H-4 i s 0.87 Hz with the decoupler on-resonance.  In general, the decoupler can be moved either  to high or low f i e l d for measurement of the control spectrum.  In t h i s  case I t was moved to low f i e l d as H-4 i s farther from H-2 than H-l and this minimized Bloch-Siegert d i s t o r t i o n s .  For the control experiment,  the decoupler was 285.4 Hz from H-4 and now the induced Bloch-Siegert s h i f t Is 0.93 Hz. The difference, 0.06 Hz between these two values, accounts for the false "response" at H-4 In the SDDS spectrum.  We note  that a false response due to this phenomenon i s very different from the genuine decoupling-difference response ( c f . H-3). point, consider now the |H-6'} experiment  Having made t h i s  (top trace).  Protons 6' and 5  are coupled (J • 2.3 Hz), yet the "response" at H-5 looks more l i k e a Bloch-Siegert a r t i f a c t than a decoupling difference peak; i n contrast, the {H-3} experiment i s e n t i r e l y unambiguous. We conclude then that SDDS i s a powerful t o o l , but that i t should be approached with caution i n s p e c t r a l l y crowded regions. I f t h i s method alone i s used to e s t a b l i s h J_-connectivity, several experiments may be necessary, varying decoupler pulse power and/or control spectrum frequency.  An i l l u s t r a t i o n i s given i n F i g . II.2.11  where, with a constant decoupler power (yB?/2n • 17 Hz), the control o f f s e t i s sequentially moved from 50 to 10 Hz from vn_i ( v Vfj.j i s termed Av).  c o n t r o  i -  With Av - 50 Hz, Bloch-Siegert e f f e c t s are seen  for every proton; i n the top trace (Av - 10 Hz), Bloch-Siegert s h i f t s are i n s i g n i f i c a n t .  The line-shape of H-2 i s poorer than with Av - 15  -  124  -  10 Hz  15 Hz  20 Hz T  25 Hz  30 Hz  40 Hz  -vvv-  Av  =50 Hz  H-2 H-l  Fig.  II.2.11  SDDS  4.0  5.0  6.0  irradiating  H-l  [(YB /2TI) 2  = 17 H z ] ,  and  varying  t h e f r e q u e n c y d i f f e r e n c e (Av) b e t w e e n t h e d e c o u p l e r r e s o n a n c e ( f i x e d ) and the c o n t r o l , off-resonance (varied).  on  - 125  Hz; H-3  -  i s long-range coupled to H-l and the response i s probably  genuine. II.2.4  Hudear Overhauser E f f e c t  2 5  ^  2 8  The nuclear Overhauser effect (nOe) 1  routine  i s of great u t i l i t y i n  13  H and  C NMR  spectroscopy.  C l a s s i c a l l y , one considers a  saturating RF f i e l d , B_2, and two spins, I and S.*  The nOe  specifies a  v a r i a t i o n i n Intensity (= integrated area) of spin I when spin S i s saturated.  In this section, discussion i s r e s t r i c t e d to a homonuclear  steady-state experiment involving intramolecular nOe's between weakly coupled  (or coupled) spins (such as ^H^H}  i n r i g i d molecules, i n the  absence of chemical exchange phenomena). The nOe,  defined as the f r a c t i o n a l change i n i n t e n s i t y of the  signal of spin 1 when spin j (or group of spins j ) i s saturated, i s given  by: f.(j>  <I ^> i s the average magnetization Z  <I .> - I , — oi  [II.2.13]  of i i n the presence of saturation  of spin(s) j , and I ^ the equilibrium magnetization 0  i n the absence of  I$2«  The nOe  phenomenon involves population transfer r e s u l t i n g from  relaxation e f f e c t s between spins. relaxation rate of spin 1, i t may  If  i s the t o t a l direct  be considered  as:  *In keeping with the l i t e r a t u r e nomenclature, the two designated I and S (with S i r r a d i a t e d ) , rather than A and X.  spins are  - 126  R  =  Z p i*j  -  + p*  [II.2.14]  2  where p ^ j i s the dipolar relaxation rate of spin i with j , and  pj*  i s the direct relaxation rate due to mechanisms other than dipolar (which for small molecules i s often n e g l i g i b l e i n comparison with p ) . At this stage a f e e l for the populations p r o b a b i l i t i e s i s useful.  and t r a n s i t i o n  Consider the energy l e v e l diagram for the  A. F i g . II.2.12  B.  Energy l e v e l s , t r a n s i t i o n p r o b a b i l i t i e s and r e l a t i v e populations (A) for an IS system at equilibrium, and (B) for an l { s } s i t u a t i o n (where S i s saturated).  weakly coupled IS spin system at equilibrium ( F i g . II.2.12A). F_ i s the equilibrium population of the x-state, and W the t r a n s i t i o n x  probability.  The  P_l) + (Pi^. - P_2).  i n t e n s i t y of the I spins w i l l be proportional to (P3 Measurement of single-quantum W_i t r a n s i t i o n s has  information on dipolar-relaxation and, hence, distances; information i s i n the zero - (Wo)  no  this  and double-quantum (W ) t r a n s i t i o n 2  - 127  -  p r o b a b i l i t i e s and can only be determined when the spin populations been prepared i n some suitable fashion (e.g. saturation or inversion).  When S i s saturated  2(P  1+  - P_i).  population  (Fig. II.2.12B), populations  1 and 2, 3 and 4 are equalized:  (P  - P^) - (P^ - P ) , and  3  2  This change i n the energy l e v e l populations  have  of levels  (P^  - P_i°) =  0  w i l l be negated  W_2 w i l l attempt to r e - e s t a b l i s h  by the "cross-relaxation" processes:  equilibrium by increasing Pj and decreasing  Pj,.  The  effect of WQ w i l l  be to increase P3 and decrease P_2« These cross-relaxation terms contribute to a cross-relaxation rate, o"ij :  o  For the t o t a l l y  - wJ  ± j  ( i j ) - W?  D  (ij)  D  dlpole-dipole (DD)  [II.2.15]  relaxation mechanism, p^j can  be  written i n terms of t r a n s i t i o n p r o b a b i l i t i e s :  P  = 1;J  H.0  D  (il)  + 2W.1  D  (ij)  +W?  Eq. II.2.13 can be expressed i n terms of a and  f  Expressions may  i J> (  * °U/  D  (ij)  [II.2.16]  p:  [H.2.17]  Pii  also be derived for pj^j and ojj which  i n t e r e s t us more, as they contain the dipolar distance information the homonuclear experiment (as mentioned).  for  Under extreme narrowing  conditions, the frequency terms i n the f u l l e r expressions can be Ignored  -  128  -  and i t can be shown that:  ij  " i  o  = (1/2) Y  P  ±i  Y  j  Y  h  ' c Ij  UL2.18]  r  T  h  2  T  2  c  r~J  [II.2.19]  where r ^ j i s the internuclear separation between spins i and j , z i s the motional c o r r e l a t i o n time* and n, Planck's constant  c  divided by  2 it. We now  return to the measurable f r a c t i o n a l nOe,  relate this to the concepts introduced p r a c t i c a l use for the experiment.  f^(j),  and  above as we work towards a  The measured f ^ ( j ) has (the f i r s t )  d i r e c t p o l a r i z a t i o n of spin i by spin j , minus the i n d i r e c t p o l a r i z a t i o n of spin I by spin j through the other spins, n, in a multispin system: f Yj o f  The  i  (  j  )  -  ^Tli  Z Y  ±j  ~  n  cr  in  J (j) n  7R  UI.2.20]  second term In Eq. 1.2.20 accounts for the "three-spin e f f e c t " which  i s the i n d i r e c t interaction between i and j, discussed  mediated by n.  This i s  later.  The next stage requires developing determination three-spin AMX  a working protocol for the  of internuclear distances involves consideration of a system.  From Eq. II.2.14 we  get:  *Very roughly, the motional (= rotational) c o r r e l a t i o n time i s the time taken for the molecule to tumble i n solution and rotate by one radian.  - 129 -  =  P  AM  +  P  AX  P  +  A  PAS PMX P +  =  * Assuming  * »  PMX  +  PAM  +  P  +  LI1  X  -2  211  M  * « p^, and g i v e n t h a t R^, R^ and  were measurable  q u a n t i t i e s , t h e t h r e e unknowns i n these t h r e e e q u a t i o n s can e x p l i c i t l y be determined,  and i n t e r n u c l e a r d i s t a n c e s c a l c u l a t e d from Eq. I I . 2 . 1 8 .  T h i s i n v o l v e s a f u r t h e r assumption t h a t t h e c o r r e l a t i o n times o f the AM, 13 AX and MX v e c t o r s (which c o u l d be q u a n t i f i e d from  29 C r e l a x a t i o n data  a r e e q u a l , Hence, we a r r i v e a t a master e q u a t i o n o f t h i s  form:  f,(M) + f , ( X ) f_(M) _2 £ ± f (X) + f (M) f (X) A  A  o 221 Ul.^.^J  [ T T  M  W i t h t h e t w o - s p i n a p p r o x i m a t i o n , o n l y t h e f i r s t terms and f ^ ( X ) ] a r e used.  experiments  [f^(M)  I f one of t h e two d i s t a n c e s i s e i t h e r known o r  can be assumed, t h e o t h e r can be c a l c u l a t e d . i s sought,  )  A l s o , i f the l o c a t i o n of A  i t s h o u l d not be i r r a d i a t e d , but the A{M} and Afx} s h o u l d be performed.  I n terras o f e x p e r i m e n t a l p r a c t i c e , t h e  power l e v e l s have t o be c a r e f u l l y s e t t o ensure t h e same s a t u r a t i o n o f each s p i n ( s ) , o r n o r m a l i z e d .  percentage  These a r e t h e fundamental  tenets of the process of determining i n t e r n u c l e a r d i s t a n c e s i n s m a l l organic  molecules.  3 0  31 As d e s c r i b e d i n some d e t a i l by Mersh and Sanders,  i ti s well  known t h a t t h e 3 - s p i n case can l e a d t o the o b s e r v a t i o n o f n e g a t i v e nOe's  - 130 -  "3 i n roughly l i n e a r systems  o  when long (2-3 T_T) i r r a d i a t i o n times are  used.  The authors explain the phenomenon and i t s potential diagnostic  uses.  In such cases, more e x p l i c i t equations (e.g. Eq. II.2.20) are  used which take into account the p o s i t i v e "direct e f f e c t s " and " i n d i r e c t " negative ones.  I t i s interesting to note that, under some  conditions, these two contributions can almost cancel, leading to a small induced nOe i n a proton which i s geometrically close i n space to the one being i r r a d i a t e d ; this i s an unpleasant p o s s i b i l i t y . II.2.4.1  Steady-State nOe This i s the simplest nOe experiment and has been alluded to i n  the preceding remarks.  With two spins I and S, radiofrequency f i e l d  ( B 2 ) i s applied to one resonance long enough (= 3 T_i) f o r the nOe to build up i n the other.  Immediately  prior to a c q u i s i t i o n , the decoupler  i s turned o f f so as to retain spin-spin coupling and preclude Bloch-Ziegert e f f e c t s .  Schematically: 5TJ (RD) - 3Tj - 90° - AQN  I 1 -  &2 •  The experiment  [II.2.22]  i s repeated to obtain a good signal-to-nolse r a t i o and i s  then repeated with B_2 placed o f f resonance, and the integrals of the two spectra compared. Several p r a c t i c a l considerations warrant expansion.  When the  - 131 -  frequencies  of I and S are close (say, < 1 ppm) I t Is d i f f i c u l t to  saturate one spin without radiation leakage into the other.  In order to  maintain the required frequency s e l e c t i v i t y the strength of the i r r a d i a t i n g f i e l d iy^/Zn) saturation.  should be reduced so as to e f f e c t 60-80%  A less than saturating decoupler f i e l d does not invalidate  the measurement since the magnitude of the induced nOe for a p a r t i c u l a r spin i s proportional to the extent of saturation of i t s neighbour. ** 3  Use of a weaker B_2 f i e l d reduces the magnitude of induced nOe's and t h i s could demand a s i g n i f i c a n t l y larger signal-to-noise r a t i o i f these (small) changes i n i n t e n s i t y are to be r e l i a b l y measured. In practice, the frequency s e l e c t i v i t y can be checked by performing a number of experiments varying the decoupler power. minimize nOe build-up, time (ca. 0.01s).  To  the decoupler i s l e f t on for a short period of  Comparison of peak I n t e n s i t i e s of resonances close to  the decoupler frequency with those i n a "control" experiment  (decoupler  off-resonance) w i l l ascertain whether the desired frequency s e l e c t i v i t y i s achieved. It i s invariably convenient to perform the experiment difference mode** (nOe difference spectroscopy = NOEDS). 3  i n the  For n  i r r a d i a t i o n experiments, (n_ + 1) experiments are required, including one "control" where the decoupler frequency i s chosen distant from the spectral region - say, 5 or 10 p.p.m. to high f i e l d of the h i g h e s t - f i e l d resonance. 1.  Again, the signal a c q u i s i t i o n i s interleaved: 6  Create on disc (n_ + 1) blank f i l e s , with the decoupler frequencies  and power settings s p e c i f i e d .  - 132 -  2.  Read into core memory f i l e number 1.  3.  With the decoupler set to u>i, obtain 2-4 "dummy" scans (discard), to establish a steady-state.  4.  Acquire  scans (8 or 16).  5.  Save the spectrum on disc.  6.  Repeat 2-5, now with f i l e number 2 ((1)2).  7.  Continue u n t i l (n_ + 1) spectra have been saved on disc, reading i n the appropriate f i l e each time.  8.  Repeat 2-7, m times so as to build up a good signal-to-noise.  The t o t a l number scans f o r each experiment *  * m.  Again, the  FID's are exponentially weighted (using 1.5 - 2.0 Hz line-broadening), Fourier transformed and phase corrected.  Each spectrum i s then  subtracted from the "control", and plotted.* are  Overhauser enhancements  best measured by d i g i t a l integration of signals (and not by  measuring peak heights).  Subtraction errors, usually obvious with  intense s i n g l e t s , may be reduced by small (0.001 Hz) s h i f t s of the 24  spectra. When plotted with the i r r a d i a t e d peak negative-going i n i n t e n s i t y , positive going peaks indicate a positive nOe (fast molecular tumbling) and negative peaks indicate a negative nOe (usually indicating slow tumbling).  Care should be exercised i n applying t h i s rule, since  radiation leakage into a nearby signal may f a l s e l y indicate a negative  its  *Small phase adjustments i n the difference spectrum may Improve quality.  -  133  -  nOe, and, as has been discussed previously, certain linear spin-systems can produce negative nOe's i n the extreme narrowing l i m i t . We r e i t e r a t e that, i n practice, excellent signal-to-noise r a t i o s are required for each spectrum, e s p e c i a l l y when (a) one i s using s i g n i f i c a n t l y less than a saturating decoupling f i e l d , and (b) the expected nOe's are small.  With care, enhancements down to 0.5% are  measurable, as seen i n l a t e r chapters. It i s not always necessary to attempt a l l the nOe experiments which are possible for the molecule of i n t e r e s t . signals to saturate i s important.  Choice of which  I t i s d i f f i c u l t (but not impossible)  to induce and detect nOe's into a methyl resonance because the methyl protons relax one another e f f i c i e n t l y without requiring s i g n i f i c a n t relaxation pathways from other protons i n the molecule; furthermore, induced nOe's are d i f f i c u l t to detect since methyl signals are often sharp s i n g l e t s and, therefore, r e a d i l y show small frequency o f f s e t s when displayed i n the difference mode.  In short, for q u a l i t a t i v e data one  should i r r a d i a t e the proton having the fewest alternative relaxation pathways, since this w i l l be dominated by dlpole-dipole relaxation from the smallest number of protons and therefore show maximal nOe's. 6  been noted >  I t has  31  that i r r a d i a t i o n of steroid methyl groups i s useful i n  providing information on 6-axial protons In the v i c i n i t y . note that t a i l o r e d e x c i t a t i o n s e q u e n c e s * 35  30  F i n a l l y , we  could prove useful i n (a)  t h e i r frequency s e l e c t i v i t y and (b) the a b i l i t y to " i r r a d i a t e " simultaneously serveral chosen multlplets i n a spectrum. F i g . II.2.13 i l l u s t r a t e s a series of NOEDS(SS) experiments  F i g . II.2.13  SSNOEDS on 1 (400 MHz). The i r r a d i a t e d proton Is indicated next to each difference spectrum. The i r r a d i a t i o n time was 6s.  - 135 -  performed on 1.  It i s clear that I t i s d i f f i c u l t  to get the required  frequency s e l e c t i v i t y given the small frequency separation (Av) H-l and H-2  (75% saturation; Av - 64 Hz).  between  {5} i s an interesting  experiment as i t i l l u s t r a t e s both the strong nOe induced i n i t s 1,3 trans d i a x i a l neighbour  (H-3), and the comparatively small nOe  into the  methylene protons (which have an e f f i c i e n t , mutual relaxtion patheway). II.2.4.2  Transient Overhauser E f f e c t  2 8  .  3 6  *  3 7  (TOE)  This experiment i s similar to the steady-state nOe experiment i n that the populations of one spin multiplet are s e l e c t i v e l y inverted (using a selective 180° pulse) and the recovery of the spin system i s followed.  To obtain frequency s e l e c t i v i t y , the duration of the  decoupler pulse i s , t y p i c a l l y , 0.01  to 0.03  s.  The experiment i s  depicted below:  1_JL 5T (RD)X  - t - 90° -  AQN [II.2.23]  a)  n An i n i t i a l relaxation delay i s ended by the selective 180° pulse, which i s applied through the decoupler channel.  A variable time-delay (t)  follows ( c f . T1IR experiment), with a non-selective 90° "read" pulse and data a c q u i s i t i o n .  For each jt value, ri scans (8-16) are  initially  - 136 -  c o l l e c t e d , and the cycle repeated i n a fashion analogous to the SSNOEDS and SDDS experiments. composite pulse 240°+  y  90°  To achieve a more selective 180° pulse, a  comprising a "sandwich" of three pulses, 90°+  x  i s highly recommended, as this gives excellent  ± x  population inversion and removes errors caused by resonance o f f s e t . comparison of  A  the effect of using a single selective 180° pulse vs. the  composite pulse i s given i n F i g . II.2.14.  A.  F i g . II.2.14  B.  C.  (A) The equilibrium z-magnetizatlon of 1 (H-5). The z-magnetizatlon immediately after (B) a single selective 180° pulse, and (C) a compositive 180° pulse.  Conceptually, the TOE experiment can be explained by invoking similar arguments to those proposed i n II.2.4.1. magnetization  In general  of the inverted resonance w i l l follow a multiexponential  recovery, while the unperturbed  spin suffers a transient change i n  i n t e n s i t y , which depends on the extent of i t s relaxation with the one perturbed.  Attention i s directed to the growth i n intensity of the  -  137  -  unperturbed spin (displayed i n the difference mode) as a function of time, _t. As we s h a l l see, comparisons of absolute nOe's f o r d i f f e r e n t protons do  not necessarily give the correct distance information; a 26 28  more precise evaluation comes from the i n i t i a l slope build-up  »  of the TOE  curve, which i s approximately equal to a. Using a two-spin  approximation and assuming p^j ™ p j ^ , 0"AM/a^X  =  ^ AX^ AM) r  r  6,  Although not absolutely precise, this gives us a maximum figure for r , the internuclear Fig.  separation.  I I . 2 . 1 5 i l l u s t r a t e s a difference TOE experiment conducted  on 1; H-5 i s s e l e c t i v e l y inverted by a composite pulse ( 1 8 0 ° = 2 0 ms). t_ values are In seconds.  A r e l a t i v e l y fast build-up  of nOe i n H-3 i s  observed, while, f o r example, H-l builds up a smaller nOe at a slower rate.  F i g . I I . 2 . 1 6 shows plots of the nOe build-ups Consider now the nOe build-ups  for H-3 and H - l .  i n H-3 and H-l.  The ratios  fH_l(H-5)/fH-3(H-5) vary from 0 . 2 5 - 0 . 4 1 and cannot be used f o r distance c a l c u l a t i o n s .  From the i n i t i a l slopes', a\/03 " 0 . 2 6 . Knowing  that ( 0 1 / 0 3 ) = ( r _ / r _ ) 5  3  5  internuclear separation  1  + 6  ,  and assuming a t y p i c a l H-3, H-5  of 2 . 6 4 A , we calculate r  = 3.32 A ( ± . 1 A ) .  - 138 -  H-3  H-l 8.00  X  1  4.75  3.50 k.  JJ^  **•—f^-  ^2.50  175  -^075  10.25  SEL.180  CONTROL  F i g . II.2.15  0  JLOL  Difference TOE on 1 (400 MHz), s e l e c t i v e l y inverting  H-5  with a composite pulse (see text). The time, j t , between selective population inversion and t o t a l signal a c q u i s i t i o n i s given with each spectrum.  - 139 -  F i g . II.2.16  Data from F i g . II.2.15, showing the nOe build-ups of H-l and H-3 with time.  Considering the small size of OR-I  and the resultant error In Its  c a l c u l a t i o n , we note a close agreement  to the distance obtained by g  neutron-diffraction for a similar system, of 3.33 A.  Note also that  errors i n the measurement of a translate to sixth-root of errors i n distances.  This example i l l u s t r a t e s the ease with which approximate  distance measurements may be obtained for r i g i d systems.  The interested  g  reader i s referred to a thesis on this Distance information  subject.  derived from dynamic methods such as TOE i s  much more useful than SSNOEDS, since  the method automatically  gives a  - 140 -  f e e l f o r the rate at which these population transfer processes occur, and t h i s gives a better picture of the system.  TOEDS eliminates the  guess-work involved with choice of the i r r a d i a t i o n time i n the SSNOE experiment. Integrals.  Results are found i n build-up rates, rather than SSNOE peak We also draw attention to the potential frequency  s e l e c t i v i t y through the use of a selective 180° pulse.  If further  s e l e c t i v i t y Is demanded, one may make use of a t a i l o r e d excitation pulse sequence  35  or DANTE** . 1  For the author's purposes, the composite 180°  pulse has proved s a t i s f a c t o r y . In conclusion we note the potential of the experiment with 37  slowly tumbling molecules  •*()  39  »  »  ( s p i n - d i f f u s i o n l i m i t ; vide  infra). II.2.4.3  Truncated Driven HOE Difference Spectroscopy For studies of molecules outside of the extreme narrowing  l i m i t , the results from steady-state NOE measurments no longer bear a d i r e c t r e l a t i o n to interproton distances; Instead, the behaviour depends on "spin d i f f u s i o n " . It i s general practice to study b i o l o g i c a l l y important molecules - even as small as an octapeptlde - dissolved i n viscous solvents (e.g. DMS0-d^ ) and observed at as high a f i e l d - s t r e n g t h as possible (to obtain 6  maximum dispersion).  These two factors combine to place the system i n  the spin d i f f u s i o n regime where a maximum nOe of -1 i s attainable.  Two  a l t e r n a t i v e , one-dlmensional experiments may be used to study molecules i n this regime:  transient nOe experiment (vide supra) and the more  -  useful  truncated The  driven  pulse  sequence  near-saturating followed  by  experiment presented  a  in  to  "read"  is  nOe  be  collected the  experiment for  driven  applied  pulse  and  in  -  2 9  • ** **  requires  cog f o r  each  -  with  nOe").  selective,  times,  acquisition.  At  -90°  a  varying  together  mode.  ("driven  2-4  nOe  signal  -  1  at  blocks,  difference  RD(5T )  141  jt.*  Again,  a  blank,  frequency,  This  is  the and  the  a number  data  of  AQN  B,_ri  [II.2.24]  £2  t  experiments  is  For each  spin  each  curve  irradiated slope  of  information If that  for  linear,  and  the  gives  the  nOe  curve using  one  is  "rising" with  *t_ I s  0.5  then  •  - 0  the  s).  levels the  short  t^.  induced nOe  with  irradiation of  the  be  with  spin  discussed  system  time  one  time  the  spin  i  to  slope  of  the  i n i t i a l  distance  earlier.  is  the one  compare  irradiation  a  is  investigation  in  for  up  under  curve,  and  i n i t i a l  into  Jt,  builds  s plotted,  translated  build-up  irradiation  compared  If  for  it  against  proton  information.  methods  the  each  the  can  comparative  plotted  is  curve  This  is  for  Again,  distance  ~ij'  nOe  curve  off.  build-up  portion this  the The  familiar  a particular  experiment,  «  an nOe.  and  which  varying  experiment  showing  maximum v a l u e each  performed,  and  knows  approximately may p e r f o r m  single  times  nOe's  used  in  just  one  rather  SSNOEDS  - 142 -  than the slopes, to obtain the distance information. Examples of this experiment appear i n Chapter II.5.  II.2.5 II.2.5.1  Two-Dimenslonal  Experiments'* "' 5  17  Basic Concepts Most readers are f a m i l i a r with the concept of normal pulse FT  NMR.  A pulse i s applied to the system and a signal i s detected as a  function of time, js(jt). this time-dependent  The process of Fourier transformation converts  function into a frequency-dependent  one, S(F) -  which i s the familiar display of absorption jvs frequency. dimensional ("2D") NMR  With two  spectroscopy the magnetization i s detected as a  function of two time Intervals; one a l i n e a r l y incremented time delay, j t i , and detection occurring during the a c q u i s i t i o n time, labelled tjj, r e s u l t i n g i n a two dimensional data array, J5(t_i, Jt^).  Usually, ii  experiments are performed, each with an Increasing value of t_i, and the ii data sets saved on a mass data storage device, such as a hard disc. Each signal detected i n t2 corresponds b a s i c a l l y to a normal single-quantum spectrum of the molecule.  However, each resonance has a  "memory" of processes occurring during t_i and this w i l l result i n a modulation of the signals - either by phase or amplitude. has been drawn between 2D NMR  and two-dimensional paper  The analogy  chromatography.  Chromatographic separation of similar compounds i s improved by using one solvent system to disperse components along one axis of the paper which i s then turned 90° and developed with a d i f f e r e n t solvent system i n the second dimension to provide enhanced dispersion.  If the paper i s  -  143  -  2m i  -A A.  s(\ \ ) ] t  C. S(F ,ti)  5(t,Jc ) 2  D. S(F Fi)  2  2/  .2m.  i n  JL_L  A  E. S ( F i , F ) 2  Fig.  II.2.17  Schematized processing  data of  a  matrices 2D  illustrating  data-set.  n  ^  /B.  2  1  the  stage  - 144 -  sprayed to v i s u a l i s e the compounds, a plot of colour intensity over the eluted area would produce diagrams similar to those presented i n this chapter!  With 2D NMR, the NMR responses are spread over two frequency  axes to improve dispersion.  The "solvent" used i n J_2 ^  s  almost always  chemical s h i f t , but F j may characterise a variety of parameters  (e.g.  homonuclear J_'s), depending on the pulse-sequence used to e l i c i t the NMR responses. As mentioned, ti experiments are performed and the FID's (2m words) sequentially stored on disc ( F i g . II.2.17.A.). of n_ rows and 2m columns Is denoted sCti.t^).  This data matrix  Next, a l l FID's are  Fourier transformed, y i e l d i n g ( F i g . II.2.17.B.) the sCtj.F^) matrix; this comprises a set of spectra, the signals of which are modulated i n either their amplitude or phase.  Next, a transposition i s performed  giving a data-set which indicates the dependence of the magnetization at a p a r t i c u l a r point i n T^, as a function of t ^ ; j5(F2»t.l) * II.2.17.C).  n  (Fig*  This generates a series of interferograms* - one for each  absorption frequency i n F^.  These are traces of columns taken from the  s(t.l»F.2) data set and indicate the "spectrum" of amplitude or phase modulation i n tj.  The interferograms are Fourier transformed, to give  jS(F_2,j?i) i n F i g . II.2.17.D.; these signals are often displayed i n a phase-insensitive mode (vide i n f r a ) since phasing i n both frequency dimensions i s d i f f i c u l t .  F i n a l l y , this matrix i s transposed again to  give a S(Fj ,F ) data set ( F i g . II.2.17.E.). ?  This process i s simpler i n  *An interferogram i s e s s e n t i a l l y an "FID" of magnetization as a function of t ^ , given another name to d i f f e r e n t i a t e i t from _t2 signals.  - 145 -  practice than i t may  appear here, and can be performed r e l a t i v e l y e a s i l y  with standard soft-ware packages included i n most modern  spectrometers.  Some programmes require the operator f i r s t to specify a l l parameters and then with a single command, the above steps are performed automatically, transparent to the operator, while others require operator intervention at  each step. As mentioned previously, 2D spectra are normally presented i n  either power- or magnitude-mode ( i . e . phase i n s e n s i t i v e , * as this obviates the necessary complex phase adjustments required over dimensions).  two  Unfortunately this procedure creates a problem, since the  phase-insensitive display of a Lorentzlan l i n e has very wide "wings" i n t e n s i t y distant from the resonance frequency.  These dramatically  reduce the base-line separation between i n d i v i d u a l signals, and can give r i s e to a r t i f a c t s - false signals a r i s i n g from interference between the wings of two peaks.  F i g . II.2.18 i l l u s t r a t e s the problem and some of  the solutions. The f i r s t column shows the appearance of several weighting  (apodisation) functions,** the second shows the derived  phase-sensitive frequency-domain spectra and the last i s the magnitude c a l c u l a t i o n of the absorptions i n the second column on the same frequency  scale.  F i g . II.2.18A shows the exponential  filter  *Fourier transformation of a time-domain signal y i e l d a " r e a l " (u) and imaginary (v) component. Usual (phase-sensitive) detection requires observation of the " r e a l " component, with some phase adjustments. The "magnitude" (or "absolute") display calculates the modulus of the signal, (u + v ) * ' , and the "power" mode i s i t s square ( u + v ) . 2  2  2  - 146 -  "FID"  . II.2.18  •PHASE-SENSITIVE  MAGNITUDE SPECTRUM  Effect of apodization functions on line-shapes i n phase-sensitive (column 2) and magnitude mode (column 3).  - 147 -  (line-broadening of 0.75  Hz) applied to a sine-wave.  FT y i e l d s a  Lorentzian absorption line-shape which, when subjected to magnitude c a l c u l a t i o n yields the very broad "wings" (^''tailing"), indicated with arrows.  Converting the sine-wave into a Gaussian function, as i n F i g .  II.2.18B, y i e l d s an improved s i t u a t i o n i n both regimes. i l l u s t r a t e s the use of the " s i n e b e l l " function.  An  F i g . II.2.18C  exponential-shaped  FID ( F i g . II.2.18A) i s apodised with a sine-function of period twice the a c q u i s i t i o n time, and zero phase s h i f t ( i . e . zero at jt=0 and _t=AT).  Its  phase-sensitive FT shows decreased line-width accompanied by s i g n i f i c a n t d i s t o r t i o n at the base of the peak.  A s i g n i f i c a n t loss i n  signal-to-noise i s incurred, not obvious from the second frame. Magnitude c a l c u l a t i o n produces a line-shape which approximates  to an  absorption l i n e and i s comparable with the phase-sensitive Lorentzian signal.  This form of modulation  roughly symmetical  about AT/2  pseudo-echo shaping may  i s termed a "pseudo-echo", as i t i s  and zero at both start and f i n i s h .  The  be further improved by f i r s t multiplying the  decaying exponential function by a r i s i n g exponential to  artificially  eliminate the decay, and then applying the s i n e b e l l function (or a similar symmetrical Guassian envelope). A similar protocol, less drastic i n i t s attenuation of signal-to-noise i s the "double exponential" f i l t e r , shown i n F i g . II.2.18D.  Here, the FID i s weighted with a r i s i n g exponential f i l t e r  and a f a l l i n g function f o r Gaussian l i n e shape.  The result i s a  compromise - the resolution enhancement i s poorer, but there i s less peak-base d i s t o r t i o n and degradation of slgnal-to-noise.  The magnitude  - 148 -  c a l c u l a t i o n spectrum shows less t a i l i n g . The display of 2D data sets i s a n o n - t r i v i a l problem.  Early  experiments were invariably displayed as a "stacked plot"; t h i s consists of a series of conventional of J _ i (or, vice versa).  S(J_2) spectra, one for each regular i n t e r v a l  The three-dimensional  impression  i s further  enhanced by use of the "white-wash" routine, whereby peaks from previous traces are not overwritten  (the pen l i f t s up and skips over them).  Although the three-dimensional  effect Is excellent and the plots are  a e s t h e t i c a l l y very pleasing, i t i s often d i f f i c u l t to extract the s c i e n t i f i c information.  Furthermore, the process i s unecessarily  time-consuming because the majority of the traces only display noise.  A  far more e f f i c i e n t routine employs the "contour p l o t " display which i s s i m i l a r to that used for geographical say, equal elevation.  With 2D NMR,  maps; contour lines l i n k l o c i of,  the F_i/J_2 plane Is viewed from  above, along the amplitude axis; contours j o i n points with the same amplitude.  Data are easily extracted from such displays.  Problems  a r i s e i n choosing the contour levels when a large dynamic-range exists between d i f f e r e n t peaks; In fact, a series of peaks could be e n t i r e l y missed i f the threshold l e v e l were set too high - hence, caution be exercised.  should  F i g . II.2.19 i l l u s t r a t e s the benefit of the contour plot  over the stacked p l o t .  The diagrams are from the 2D homonuclear  J_-correlated spectrum of a derivatized disaccharide. The  relevant parameters can often be extracted from the data set  without the need for the 3D representation. projections * 5 0  5 1  Here, use i s made of  (or sums) either taken over the entire frequency  - 149 -  4.20  4.15  4.10 8 (ppm) J(Hz)  F i g . II.2.19  Comparison of stacked (A) and contour plots (C). B. The J_i projection of the t i l t e d data-set. The s t l p l e d peak arises through strong coupling between 6' and 6^. Data are from the same region of a 2D J-resolved experiment of a disaccharide.  - 150 -  Fig. II.2.20  A. Stacked-plot of schematized 2D data set. B. Contour plot i l l u s t r a t i n g f u l l 0° and 90° projections (top and r i g h t ) , and " p a r t i a l " 90° projections (bottom).  I  - 151 -  range i n £2 or Fj (0° or 90° projection, respectively) or a single trace, or " s l i c e " may  be taken out at a p a r t i c u l a r J_2 value and viewed  at 90° ( F i g . II.2.20).  The l a t t e r are called JFJ traces and generally  contain a l l the information needed. An almost overwhelming variety of 2D experiments has already been published and a l l indications are that this rapid p r o l i f e r a t i o n w i l l continue.  An objective inspection of the recent  literature  indicates that the major experiments have already been developed; what i s now  evident Is a second generation of pulse sequences i n which  e x i s t i n g ones are either refined (see, for example, r e f . 46, pp. 50-65) or two experiments are combined to either y i e l d a d d i t i o n a l information (e.g. Section II.2.5.3.4) or improve the experiment (e.g. r e f . 52).  It  i s useful at this juncture to broadly c l a s s i f y the available 2D experiments: A.  Resolved  NMR  In this case, J_2 Is the chemical s h i f t axis of the observed nucleus and Fj may  represent a variety of "resolving"  parameters such as homonuclear J-coupling (vide i n f r a ) , heteronuclear J_-coupling, heteronuclear  chemical s h i f t (vide  i n f r a ) or dipolar coupling. B.  Correlated  NMR  Also termed "autocorrelation" the same chemical s h i f t axis i s plotted along both dimensions and spins are correlated either through J_-coupling (vide i n f r a ) or spin exchange or chemical exchange).  (nOe  - 152  C.  -  Indirect Detection Multiple quantum t r a n s i t i o n s (MQT)  cannot be observed  d i r e c t l y , but experiments have been developed to allow for their i n d i r e c t detection. also f a l l s In t h i s D.  Indirect detection of rare n u c l e i  category.  Combination of A, B and C, and Others  Before commencing with more detailed comments on some selected experiments, a few general points are appropriate.  A major  experimental  r e s t r i c t i o n i s often d i g i t i z a t i o n , since one i s often limited by available data storage space and time. sweep-widths i n  As a r e s u l t , the minimum  and J_2 should be selected.  Any a l i a s i n g i n F_ ,  due  2  to i n s u f f i c i e n t sweep-width can usually be chosen so that these peaks f a l l into a blank region of the spectrum.  They can be attenuated  of the frequency f i l t e r i f they have no information.* possible to perform frequency f i l t e r i n g i n  by use  It Is not  and the experimenter i s  advised either to avoid a l i a s i n g by choosing a suitable sweep-width, or exercise caution by being aware of the way  i n which t h i s occurs.  The number of experiments performed, n, i s usually a binary number (6A, 128, 256, e t c ) .  The smaller n,  be required to obtain _ s ( t i , t ^ ) *  the less machine time w i l l  D i g i t i z a t i o n i n F j can usually be  improved by z e r o - f i l l i n g , but caution should be exercised i n t r y i n g to  •Solvent signals, impurities and acetate s i n g l e t s can often be treated i n t h i s way.  - 153 -  l e t a very small ri s u f f i c e , since this may lead to severe truncation of the interferograms, and associated line-shape d i s t o r t i o n s . chosen a suitable block-size and sweep width ( i n Yj),  Having  one can calculate  the acquisition time and set up the experiment such that the e f f e c t i v e a c q u i s i t i o n time i n r j , (ti * A t i ) i s of similar duration. The incremental delay-time, A t i i s easily calculated from the required SW i n F j , 0.5 Ajti used i n spin-echo experiments i s given by expression II.2.24, and A_ti Is given by II.2.25:  0.5 A t i - (4 * ± SW„ )  _ 1  [II.2.24]  - 1  [II.2.25]  £.1  —  A t i - (2 * ± SW,, ) £.1  —  F i n a l l y the t o t a l experimental time may be estimated from equation II.2.26. Total time - n.NA[l/2 ( t  - 1  )  m a x  + AT + RD]  [II.2.26]  n i s the number of increments i n j t i , AT i s the acquisition time (_t_2^» i s the relaxation delay between a c q u i s i t i o n and the f i r s t pulse i n the sequence, and NA i s the number of acquisitions for each jti value (including dummy scans). F i n a l l y , we address the topic of signal-to-noise i n 2D experiments, which has been e x p l i c i t l y formulated by Ernst and 53 co-workers.  I t happens that the o v e r a l l demands of the method are  reasonable, since both the signals and the noise are spread over the (F_l ,F_2)-plane.  Obviously a loss i n signal-to-noise i s incurred as a  ^  - 154 -  r e s u l t of signal decay during j t j , but t h i s only becomes c r i t i c a l when dealing with molecules with very short T^'s (see Chapter II.5).  Some  autocorrelation experiments suffer a further degradation i n s e n s i t i v i t y because each conventional resonance generates -2E~1  components, where  m i s the number of weakly-coupled non-equivalent spins.  This i s offset  by an increase i n s e n s i t i v i t y r e s u l t i n g from symmetrization procedures (discussed i n II.2.5.3.1).  The author's "working rule" i s that i f  signals can be seen upon FT of the f i r s t of the n_ i n t e r v a l s (blocks) of the jsCti.t^) data-set, the o v e r a l l experiment succeeding.  stands a good chance of  Of course, i f resolution enhancement i s necessary for the  r e s o l u t i o n of closely spaced peaks, the slgnal-to-noise requirements of the o r i g i n a l data set w i l l be correspondingly higher.  II.2.5.2  Homonuclear J-Modulated Spectroscopy * ( 2 D JHresolved) 51  The key to this experiment i s the basic Hahn spin-echo, which has been documented  55  f o r over 30 years.  Hahn used a 90°-T-180°-T-AQN  pulse-sequence; he noted the narrower l i n e widths attainable and used t h i s very early on i n making some of the f i r s t  scalar J_ measurements.  He analysed plots of echo-amplitude vs i i n the frequency domain and came to some anticipatory conclusions. A modulation of the echo envelope which depends on both observed.  and the chemical s h i f t separation was  Further, he noted that " . . . i f only one of two coupled nuclei 19  Is  subjected to resonance (e.g. between  1  F and  H) no echo envelope  modulation w i l l appear for either of such coupled n u c l e i , although the steady state resonance w i l l s t i l l reveal the J_ s p l i t t i n g . "  - 155 -  Herein l i e the concepts of 2D J-resolved spectroscopy, albeit without the complex two-dimensional  FT.  The above quotation has also  been documented i n the l i g h t of 2D J-resolved spectroscopy.  56  The basic 2D J-resolved pulse-sequence i s : [II.2.27]  RD - 90°($i) - 0.5ti - 180°($ ) - 0.5ti - AQN($ ) 2  3  Phase cycling of the pulses and receiver i s important. scheme proposed by Bodenhausen, Freeman and T u r n e r  The EXORCYCLE  (II.2.28)  57  58 eliminates so-called "ghosts" and "phantoms" a r i s i n g  from pulse  Imperfections and residual transverse magnetization. Further improvements i n the quality of the f i n a l spectrum may be achieved by (a) using a composite 180° refocussing pulse (180°^ = 90 o ±X  180°^ 9 0 ° ; 180+ = 9 0 ° ± x  y  v  180°- 90°. and (b) extending the phase-cycling x  y  to 16 steps by incorporating the well known CYCLOPS scheme (II.2.29) to reduce quadrature Images i n F 2 . Cycle 1  2  3  4  *1  X  X  X  X  $2  X  y  -x  -y  *3  X  -x  X  -x  Cycle 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  *1  X  X  X  X  y  y  y  y  -x  -x  -x  -x  -y  -y  -y  -y  $2  X  y  -x  -y  y  -x  -y  X  X  -y  x  y  -y  X  y  -x  $3  X  -x  X  -x  y  -y  y  -y  X  -x  X  -x  y  -y  y  -y  [11.2.  [II.2.29]  156 -  The 16-cycle scheme i s preferable should one require much signal-averaging, especially for studies of dilute solutions.  If only  four pulses y i e l d s u f f i c i e n t signal-to-noise, then phase-cycling according to II.2.28 with a composite 180° pulse should be s u f f i c i e n t , but one should be aware of very strong signals which may quadrature  images i n F _ 2  &/  (vide i n f r a ) .  The 2D J-resolved experiment may diagrams.  s t i l l show  be explained i n terms of vector  Consider the X nucleus of an AX spin system.  The  relaxation-delay (RD) i s terminated by the application of a 90^ pulse, which t i p s the equilibrium magnetization onto the y_'-axis ( F i g . II.2.21A).  The X magnetization  corresponding coupled.  has a fast (f) and slow (s) vector*,  to the two possible spin states of A to which I t i s  With time, the vectors w i l l precess i n the  (x'-y_')-plane  ( F i g . II.2.21B) (as depicted previously i n F i g . II.2.3) mainly under the influences of chemical s h i f t and J . _  A 180° (refocussing) pulse i s now y  applied which takes the vectors to t h e i r mirror-image positions about the y'-axis i n the rotating reference frame.  Consider the s i t u a t i o n  where both A and X experience the 180° pulse, which gives the s i t u a t i o n i n F i g . II.2.21C.  As well as changing the position of the vectors  *Here, we are using the terms " f a s t " and "slow" to describe two d i f f e r e n t t r a n s i t i o n s , as i n F i g . II.2.3. Note that i n discussions on the removal of magnetic inhomogenelty e f f e c t s by this experiment, these terms pertain to a single resonance and have a d i f f e r e n t meaning.  -  F i g . II.2.21  157  -  Vector diagram model of the 2D J-resolved experiment. With the X t r a n s i t i o n s i n an AX spin system (A,B,E,F), both spins f e e l both pulses; i n A-D, only spin X i s subjected to the refocussing pulse (From Freeman**' ). 5  - 158  -  representing X, the labels of the two X t r a n s i t i o n s are changed (F  <->  jS) as A changes i t s spin state (o <->  Thus, during the second  0.5  _ti period the vectors continue to diverge and the echo consists of  two  B).  components ( F i g . II.2.21F); that i s , only the chemical s h i f t e f f e c t s are refocussed  and not spin-spin coupling.  The o v e r a l l r e s u l t i s a  modulation i n phase of the signals by J_, as a function of j t i .  The phase  of a signal w i l l be a function of i t J t ^ . A second s i t u a t i o n i s possible when A i s a heteronuclear  species  19  (  F, for example), for the 180°  pulse a f f e c t s the X spins but not  A.  Under these conditions the X vectors are f l i p p e d through 180°, but t h e i r " l a b e l s " remain unchanged ( F i g . II.2.21C). the X t r a n s i t i o n s w i l l be exactly refocussed  Therefore,  a f t e r O.St^  both  ( F i g . II.2.21D); the echo  i s modulated by neither chemical s h i f t nor J_ e f f e c t s , and would be seen to decay i n amplitude due We  to transverse  relaxation e f f e c t s (T^) only.  have seen that the usual 2D J_ experiment (Figs.  II.2.21A,B,E,F) leads to a Jj-modulated spin echo.  With the step-wise  incrementation of t_i one builds up a data matrix which can be handled i n the usual way J.  (Section II.2.1) and the FT i n j t w i l l be a function of  Hence, t h i s i s c a l l e d the "J_-diraensIon".  x  One  further point not  apparent from the above discussion i s that line-widths i n the  2D  * J_-resolved experiment w i l l approach the natural line-width (T_2 •*• T_2'» i . e . the contribution of magnetic inhomogeneities to line-widths w i l l very small, compared with a normal FT experiment.  This arises from the  fact that magnetic f i e l d inhomogeneity e f f e c t s also refocus at t ^ , the time of the echo.  be  - 159 -  A generalized out-put i s given i n F i g . II.2.22A, depicted i n the contour mode.  Well resolved 2D peaks hold a one-to-one  correspondence  to the resonance lines observed i n ID NMR, i n the absence of strong coupling e f f e c t s .  One notes that a line drawn through a multlplet Is at  45° to J_2, when J_i and F_ are drawn to the same scale.  I t i s mandatory  2  to remove the 6 contribution from J_2 by "shearing", or " t i l t i n g " the data matrix according to (0)1,0.12') = (coi, 012 ~ " l ) , shown i n F i g . II.2.2.22B.  to give the plot  Now J_2' i s a function of 6 only and J_i (the  J_-dimension) encodes only the scalar *H - *H couplings.  5  8  A  8  X  F (5J)  F (6)  A.  B.  2  Fig.  8  A  II.2.22  X  2  Schematized 2D J-resolved contour plot output of AX spin system without (A) and with (B) t i l t i n g .  The u t i l i t y of this experiment becomes apparent when one 51  considers projections and s l i c e s taken  as outlined In F i g . II.2.20.  The 0° projection (onto J^') of the complete data matrix w i l l given one l i n e for each spin-multiplet - this has the appearance of a "broad-band decoupled" proton spectrum and gives the *H s h i f t s .  " S l i c e s " taken at  - 160 -  the frequency corresponding  to each s h i f t , and viewed at 90°, w i l l give  the "J-spectrum" for each resonance. J_" spectra.  These are referred to as " p a r t i a l  In the absence of either strong coupling or very short  T_2's, homonuclear 2D J-resolved spectroscopy  must be the ultimate answer  to the "hidden resonance" problem.  Resonances need only be removed from  one another by a few Hz (= 0.01  at high f i e l d ) for their J-spectra  to be completely  ppm  resolvable.  The question of line-shapes i n the 2D J-resolved experiment has received much attention.  Because the method i s based on phase  modulation, the resonance lines have a phase-twisted very d i f f i c u l t to phase c o r r e c t  6 0  i n two dimensions.  shape  which i s  With a  phase-sensitive S(F_i,F_2) matrix one can almost always phase correct i n d i v i d u a l p a r t i a l J_-spectra using the conventional A and B phase-corrections.  ,FJ>  A phase-sensitive t i l t e d  ) matrix w i l l  always y i e l d J_-spectra with d i s t o r t i o n s at the base of resonance l i n e s ; however, the resultant absorption-like spectrum i s o b t a i n a b l e significantly  61  with  better line-width than those from the more commonly used  phase-insensitive spectrum (power- or magnitude-spectrum). Homonuclear 2D J_-resolved spectroscopy  i s known to produce  62  a d d i t i o n a l t r a n s i t i o n s for strongly coupled increases the complexity  63  »  spin systems, which  of the i n t e r p r e t a t i o n . Fortunately, these  effects are generally predictable; a d d i t i o n a l t r a n s i t i o n s a r i s i n g from strong coupling w i l l often l i e midway between the coupled  resonances and  usually (but not always) appear as a broad hump i n the 0° projection of  - 161  the t i l t e d data-set.  -  As such, they a c t u a l l y have diagnostic u t i l i t y i n  the sense that they provide warning of the presence of strong coupling. A note of caution should be sounded i n the i n t e r p r e t a t i o n of such strongly-coupled systems, since the observable t r u l y depicted i n _Fi.  resonances may  not be  Of course, one can make use of a two-dimensional 6 3  spectral simulation program routine procedure.  as a check, but this i s u n l i k e l y to be a  Strong coupling effects i n the J_-resolved spectra of  amino acids have been studied i n d e t a i l , * * as have other spin systems. 6  63  The question of d i g i t a l f i l t e r i n g (apodization) deserves special mention.  The use of a pseudo-echo f i l t e r  6 5  w i l l produce absolute  value  l i n e s with absorptive Gaussian character; that i s , narrower lines than what one would obtain using an exponential envelope (giving Lorentzian l i n e s i n the frequency attenuates  domain).  However, this i s not ideal as i t  signal-to-noise, causes severe intensity d i s t o r t i o n s and i s  p a r t i c u l a r l y destructive when a many-lined multiplet resonates close to a strong s i g n a l .  In a situation where signal-to-noise i s at a premium,  the usual exponential f i l t e r may  be used i n t_2' * f possible, i t i s  recommended to use a pseudo-echo f i l t e r i n the j t Line-narrowing  i n _tj may  then be effected by a  transformation (a double-exponential  2  dimension.  Lorentzian-to-Guassian  function).  If one i s optimising  the experiment for resolution i n F j , a large number of experiments (n) should be performed to y i e l d an acquisition time i n F i of several T_ 's. 2  T i l t e d proton 2D J_-resolved spectra should obviously be symmetrical proposed  66  about F i • 0 ,  and a symmetrization  procedure has been  which improves spectral quality and signal-to-noise by  H-4  Figure II.2.23  H-2  H-1  270 MHz 2D J-resolved experiment on 1. details.  H-6  See text for  -  4I  6.0  1  1  1  I  163  UM.W '  1  1  I  1  1  1  -  • > • M«r«- •  I  i  1  1  1  i  1  *  (<5  5.0  4.0 *  k  1  •  <  ft  i  j l  i  I  •  *  1  ft  ii  !)  i  i i  1 i  I  Fig.  t  *  II.2.24  Data  taken  from  contour-plot H-2  I  resonance  Fig.  mode. are  II.2.23B, Quadrature  marked  with  now  plotted  images an  *  for  asterisk.  in the  the intense  ppm) J(Hz)  - 164 -  elminating much of the " t a i l i n g " i n F_ , and i s especially useful i n the case of strong singlet resonances. F i g . II.2.23 contains phase-insensitive displays of the homonuclear 2D J-resolved spectrum of 1. spectrum, showing seven multiplets.  F i g . II.23A i s the ID  F i g . II.2.23B i s the stacked plot  representation of the 2D J-resolved experiment, performed with Exorcycle phase cycling and a single 180° pulse. was used i n both dimensions.  Sinebell resolution enhancement  F i g . II.2.23C i s the 0° projection of the  t i l t e d data matrix - the "proton-decoupled" proton spectrum.  From this  the chemical-shifts are measured and the seven corresponding p a r t i a l J-spectra are displayed.  Their enhanced resolution i s evident,  e s p e c i a l l y for the H-l and H-5 resonances which are long-range coupled to one another, and other protons. Alternately, one may use a contour mode of display, as demonstrated i n F i g . II.2.24.  The resolution-enhanced ("sinebell") ID  spectrum i s above the contour plot.  Most of the s p l i t t i n g s are c l e a r l y  discernable i n the contour-plot and J_ measurements may be made d i r e c t l y from i t with adequate accuracy. II.2.5.3  Jeener Experiment: J-Correlated 2 D NMR The "autocorrelation" 2D experiment was one of the f i r s t to 67  be mathematically explained  and documented, yet did not gain wide  acceptance u n t i l r e l a t i v e l y recently when i t was improved to allow 68  quadrature detection i n both dimensions.  Now i t i s perceived as a  very useful experiment, which i s quick and easy to perform, and whose  -  Fig.  II.2.25  Stylized (A) w i t h  165  -  S_(F_i,F_2) c o n t o u r - p l o t o u t p u t o f a n A X s p i n COSY a n d ( B ) w i t h S E C S Y e x p e r i m e n t s .  system  - 166 -  informational content i s high and predictably extractable. Two major variants on the experiments  e x i s t , which w i l l be  discussed separately. The 2D c o r r e l a t i o n spectroscopy (COSY) experiment involves a (90° — Jt out-put  A  — 90° - j t ) pulse sequence and the idealized 2  (AX spin-system) i s given i n Pig. II.2.25A.  The spin-echo  c o r r e l a t i o n spectroscopy (SECSY) variant involves a (90° - 0.5 t_i - 90° 69  - 0.5tj - _t ) pulse sequence, 2  F i g . II.2.25B.  and i t s s t y l i z e d output i s shown i n  The COSY experiment  displays both chemical s h i f t and  couplings along both frequency axes, and gives r i s e to two types of peaks:  "diagonal" peaks (8) along the l i n e J _ i - J_2 ( l a b e l l e d "D") and  " c o r r e l a t i o n " , or "cross" peaks (8) between pairs of spins which are scalar coupled ( l a b e l l e d "C").  The l a t t e r are symmetrically disposed  with respect to the p r i n c i p a l diagonal.  The SECSY experiment  displays  chemical s h i f t and scalar couplings along F^, and A6/2 along F^. Thus, the informational content i s the same as COSY i n that diagnoal- and cross peaks are both present - but the display i s s l i g h t l y d i f f e r e n t . For both experiments, optimum magnetization transfer i n t^i between coupled spins (AX) occurs at the maximum of s i n ( i t J j ^ t,i) exp(-jt /T ), 1  2  while optimum detection of the transferred magnetization  occurs at the maximum of the corresponding expression for jt . 2  The  expression i s plotted for a t y p i c a l T^ (1.5s) and J_'s of 0.5 and 4 Hz i n F i g . II.2.26.  This topic w i l l be raised l a t e r i n Section II.2.5.3.3.  A f i n a l general note concerns apodisatlon. With the Jeener experiment  a pseudo-echo shaping function (e.g. " s i n e b e l l " ) i s  recommended for use i n both dimensions.  As usual, absolute value mode  - 167 -  - 168 -  display i s chosen and the reader interested i n phase-sensitive line-shapes i s referred to the l i t e r a t u r e . II.2.5.3.1  67 68 •  COSY The basic COSY experiment uses the pulse sequence given  below, and the phase-cycling of II.2.31.  The l a t t e r i s required to  d i f f e r e n t i a t e between the so-called N-type (echo) and P-type (anti-echo) signals which must be d i f f e r e n t i a t e d when quadrature RD - 90°(x) - Jtj - 90°($ ) - AQN 1  Cycle  1 2  $1  3  4  x  y  —x  -y  (»>  x  -x  x  -x  $3 (P)  x  x  x  x  *  2  (_t ; $ 2  2  detection i s used.  or $ )  [II.2.30]  3  [II.2.31]  The f i r s t preparatory pulse (90£) i s followed by the evolution period ( t i ) and the second mixing pulse (90°), and detection ( t ) . 2  The  preparatory pulse has constant phase and the phase of the mixing pulse Is Incremented In 90° steps ( $ i ) . The receiver phase chosen, d> or $ 3 , 2  selects the coherence transfer echo or anti-echo, respectively, and cancels the a x i a l peaks at F  t  • 0 (which have Tj information and are  r e l a t i v e l y uninteresting). This phase-cycling scheme may  be expanded to  incorporate CYCLOPS, which removes Images r e s u l t i n g from  quadrature  detection.  The basic four-cycles are repeated three times,  incrementing  - 169 -  a l l phases by 90° to give a 16-cycle scheme. A satisfactory explanation of the experiment i s not possible without  a density-matrix formalism, and even this becomes rather  unwieldy when considering more than two or three spins.  We l i m i t  discussion here to a highly s t y l i z e d density-matrix formalism (see F i g . II.2.27) for an AX spin-system, with relaxation effects ignored.  States  A, B and C were described at the beginning of the chapter, i n equations II.2.5 - II.2.9.  A rotational matrix changes the i n i t i a l Boltzmann  state density matrix, OQ into o"i which has eight non-zero terms from single-quantum t r a n s i t i o n s .  With time, the terms i n the density matrix  evolve, and after j t may be represented by o"2» x  90° x  Preparation  A.  90° x  Evolution Ct.i>  Detection ( t ^ )  Boltzmann State P  00  Application of a  0  0 0  0  p  0  0  0  P  0  0  0  c  0  - 170  B.  After  Preparatory  Pulse  0  0  0  0  «1  C.  After  Evolution  - 171 -  D.  Mixing  Pulse  o  0  A^  o  0  0  0  0  A/>  A^  Mil  H i  0  -  Fig.  II.2.27  Schematized for  an AX  density  172  -  matrix  spin-system  formalism  of  COSY  experiment  - 173 -  second r o t a t i o n a l operator ( i . e . a 90° pulse) on o  2  has the e f f e c t of  mixing the elements i n the two areas denoted by broken l i n e s * element Is a sum of four frequencies. The observable  Now each  (single-quantum)  coherences are depicted whilst generated zero- and multiple- quantum coherencies are ignored and indicated by parentheses. resonance frequency w i l l be detected during  Each observable  modulated  by events i n  jti which are due to scalar coupling. The experiment Is performed with incremental delays i n Jtj i n the usual way, and subjected to 2D complex FT. are  symmetrical about the F_j • j ? axis. 2  The diagonal and cross peaks  Noise i n the J j  e s p e c i a l l y from strong solvent peaks, i s not symmetric  dimension,  and may be  removed by "symmetrization" procedures described by Wuthrich and co-workers; * 70  For  71  an increase i n signal-to-noise automatically r e s u l t s .  the symmetrization, each of the pairs of points i n the 2D matrix  which should be Identical are compared i n turn; either the geometric average of the two points i s calculated and used i n place of both values, or the smaller of the two i s taken as correct and i s substituted for the larger. Fig.  A much "cleaner" and unambiguous display r e s u l t s .  II.2.28 i s the 2D COSY spectrum of 1.  f a i r l y good (2.7 Hz p t  - 1  Digitization i s  i n F_j and F ^ ) , and considerable d e t a i l i s  obvious within each response.  In general practice one may often have to  tolerate much coarser d i g i t i z a t i o n (« 10 Hz p t ) but this does not - 1  appear to s i g n i f i c a n t l y a f f e c t the information content;* d e t a i l within  *The antiphase nature of peaks i n a c o r r e l a t i o n suggests that course d i g i t i z a t i o n w i l l lead to their cancellation and poor i n t e n s i t y seen; this conviction does not appear to born out i n the experiments performed i n t h i s thesis.  Fig.  II.2.28  COSY  spectrum  of  1 (270  MHz).  -  cross-peaks s t i l l  175  may n o t b e i d e n t i f i a b l e  be c l e a r .  This  displayed  data-set  anti-echo  selection.  point  will  was c o l l e c t e d  but  -  "the cross  peak"  itself  later  sections.  be i l l u s t r a t e d  in  using  cycle  T h e jS(F_2,F_i)  the four  display  was n o t  will  sequence  The  with  symmetrized.  3 Correlations easily  mapped A  all  are clear  that  for a l l  short  necessary,  II.2.5.3.2  and the c o u p l i n g  pattern  delay  s is  is  out.  surprisingly is  J_ p a i r s  relaxation  making  the data  of  c a . 0.1  collection  time  quite  usually short.  SECSY The  pulse  sequence  used  for  this  experiment  is  given  by E q .  II.2.32.  RD -  Phase  90°(x)  cycling  is  as  -  0 . 5 ^ -  for  t h e COSY  experiment  has the p o t e n t i a l  represents  differences  itself.  As a r e s u l t ,  sweep-width resonances  in Fi  translates  means into  acquisition many two  organic  should  these  it  at  it  to  so as  over  shift,  those data  at  is  non-viable extremes  of  echo  $2)  [ I I . 2.32]  selection.  COSY  that  the Fj  rather  than  chemical  to  substantially that  the  The  dimension shift  reduce  the  highest-field This  smaller  ( n ) i n F_i a r e n e c e s s a r y ,  time  both  this  is  since  in  terms  aliasing,  of  an a t t r a c t i v e i t  is  quite  the spectrum  the theoretical  to avoid  2  lowest-field.  points  Although  AQNU ;  with  provided  experimental  different  circumstances,  be c h o s e n ,  advantage  aliasing,  fewer  a smaller  molecules  0.5_t_i -  may b e p o s s i b l e  without  that  -  experiment,  chemical  and computation.  resonances  Under  in  are not coupled  sweep-width  90°(d>i)  maximum  data option,  feasible  may b e  for that  coupled.  sweep-width  and the p o t e n t i a l  which  i n j?i  advantages  -  of  SECSY  over  case  with  class  of  is  SECSY  discussed  performed  compounds  covered  in  Fig. sweep  COSY  width  that  detail  in  II.2.29 in  Fj  H-3  earlier  -  are  on p r o t e i n s ,  the  experiment  the  review  shows  was  176  the  possible  of  SECSY in  inapplicable.  however, shows  and  most  This  it  is  is  with  promise.  SECSY was  J_(4,5) an  were  already  known;  Data-retrieval T h e J_i  •  0 line  correlations plotted  for  to  spectrum  this  case,  of  1.  since  A  slightly  the  at  same  reduced  connectivities  H-6 H-6' H - 5  of  1  (400  small,  MHz).  causing  connectivity;  The  aliasing  the  aliased  the  case,  2  sweep-width of  part  portion  is  of  i n _F_i the  marked  with  asterisk.  if  this  SECSY  is  corresponds  lie  the  spectrum  slightly  topic  Nagayaraa.  H - 4 H-2 H-l  II.2.29  the  this  This  E Fig.  not  a  were  marginally to  constant  scale);  not  the  diagonal  angle  this  less  with  process  would  convenient in  the is  it  COSY. F_i  = 0  be  inadvisable.  than Lines  line  illustrated  in  with  COSY.  joining  (135°, Fig.  if II.2.29.  - 177 -  Two disadvantages of the experiment merit note.  Firstly,  since  correlations i n J_i occur at half the difference i n chemical s h i f t s (vs the f u l l chemical s h i f t difference with COSY) the spectral resolution i n F_l i s halved.  As molecular complexity increases, the number of  cross-peaks increase too and with SECSY the "frequency area" (Fi*_F ) i s 2  half that of COSY, Increasing the l i k e l i h o o d of cross-peak overlap, and the resultant  ambiguity.  The second point arises from considerations of transverse relaxation times, and signal-to-noise. experiment  When performing any echo  on a large molecule with broader lines (shorter  JJJ'S),  s i g n i f i c a n t intensity loss can occur during the refocussing period and the detected signal may be severely attenuated.  This i s less of a  l i m i t a t i o n with SECSY than with horaonuclear 2D J-correlated spectroscopy, as the following sample figures i l l u s t r a t e .  Assume a  half-height line-width (Avo.5) of 5 Hz with n e g l i g i b l e contribution from magnetic  inhomogeneity;  from _T " 1/HAVQ 5 we get T_2 » 64 ms. 2  The amplitude of a signal w i l l be attenuated by 75% after T_2*An 4 = 88 ms.  Now, assuming a sweep-width i n ¥_\ of ± 1000 Hz (10 ppm f o r a 4.7  Tesla magnet), the incremental delay required w i l l be 0.25 ms and the evolution period w i l l reach 88 ms after about 350 experiments.  With a  2D J_-resolved experiment, the sweep-width i n F_i might be ± 20 Hz, giving an incremental delay of 12.8 ms. experiments.  One T_2 period w i l l have passed after 5  The signal w i l l be 75% attenuated i n i n t e n s i t y a f t e r < 7  increments which i s a small number compared with the number (32) required to give a d i g i t i z a t i o n of 1.25 Hz per point.  Of course, this  disadvantage would be offset by the fact that resonances with broad  - 178 -  line-widths do not require very fine d i g i t i z a t i o n , and 1.25 Hz per point could be more than adequate.  These arguments hold i n practice, as we  8ha11 see i n l a t e r chapters.  This question of s e n s i t i v i t y losses i n the  spin-echo experiment i s addressed b r i e f l y by Morris i n his discussion on SECSY.  72  SECSY appears to be a viable alternative to COSY when cross-peak overlap and broad line-widths are not evident.  Whilst i t i s true that,  in p r i n c i p l e , SECSY should be faster to perform than COSY, this may not always be the case since signal degradation during the second 0.5 _ti period may demand more scans to be collected than would be the case with COSY. II.2.5.3.3  Delayed COSY - Detection of Long-Range Couplings As mentioned  i n the introduction to this section, the  observed J_'s i n a COSY experiment are a function of t^ and t ^ .  I f one  i s Interested i n detecting small J_'s (< 1 Hz), such as those which arise from long-range coupling, the data for a COSY experiment sampled around t%  m  0.5 - 1.0s a f t e r the mixing pulse.  should be  Using the pulse  scheme i n Section II.2.5.3.1, to have an acquisition time <*> Is i n both dimensions would require the a c q u i s i t i o n and storage of prohibitive amounts of data and one resorts to a variant on the COSY experiment, 68  where a fixed delay, A, i s inserted a f t e r each pulse. RD - 90* - A -  - 90°(*i) - A - A0N(t_ J *2 or » ) 2  3  [II.2.33]  P r a c t i c a l l y , A i s set to approximately 0.3 s, which offers the best compromise between observation of correlations between protons coupled by small J_'s, and loss i n signal amplitude during A.  -  Fig.  II.2.30  Delayed  COSY  Responses  179  experiment  from  -  on  long-range  1  (270  MHz).  couplings  are  A = 0.4 shaded.  s.  - 180 -  F i g . II.2.30 shows this version of the COSY experiment, again using 1 as the exemplar.  The delay, A, was 0.4 s.  Responses  corresponding to a l l J_ couplings are v i s i b l e , i n addition to those from a wealth of long-range couplings (which are shaded i n the diagram). H-l appears to be the most r i c h l y coupled proton, showing long-range responses to H-3, H-5 and H-6.  The long-range coupling, J(H-1, H-6), i s  quite interesting and presumably indicates a reasonably " r i g i d orientation" of the a c y c l i c C-6 moiety. coupling - only to H-l and H-3.  H-2 i s conservative i n i t s  We draw the reader's attention to the  2D J-resolved experiment, F i g . II.2.24, where i n the 0° projection (a 73  sky-line silhouette and H-l the weakest.  ) i t i s clear that H-2 i s the most intense s i g n a l , We attribute this to the fact that H-2 has only  four, well d i g i t i z e d l i n e s , whilst H-l must have at least sixteen, poorly d i g i t i z e d . II.2.5.3.4  "Decoupled" COSY An interesting variant on the basic COSY experiment has been  48 suggested  which has the net result that a l l couplings i n F j are  collapsed, and the autocorrelation map i s considerably s i m p l i f i e d . I t s potential merits are obvious. The experimental scheme i s given i n F i g . II.2.31.  The 90°  (preparatory) and 45° (mixing) pulses represent a variant on the 2-pulse COSY experiment and are phase-cycled i n exactly the same way.  Now,  however, the time between these pulses Is held constant ( t ^ ) and a reduced mixing pulse of 45° i s used.  Increments of 0.5t± separate the  preparatory pulse from the refocussing 180° pulse.* •Again, a composite p u l s e  3 8  i s used: 180°.  The period between = 90°  180° 90° x  - 181 -  90°  180°  1$,)  F i g . II.2.31  1^3)  45°  (<I>) 2  Pulse scheme for the COSY experiment with in Fj.  1  2  3  4  5  6  7  8  *1  X  X  X  X  X  X  X  X  $2  X  y  -x  -y  X  y  -x  -y  y  y  y  y  -y  -y  -y  -y  X  -x  X  -x  X  -x  -x  -x  Cycle  * 3  $1+  homodecoupling  [II.2.34]  the refocusing and mixing pulses, t ^ - 0.5_tj has to be decremented. At this time a decremental t\ i s not available i n an instrument 74 manufacturer's standard soft-ware, and must be programmed.  If  - 182 -  long-range couplings are of i n t e r e s t , a delay, A, i s Inserted prior to acquisition.  The phase-cycling i s according to Eq. II.2.34.  Usually the l a s t value of 0.5t± has the composite 180° pulse j u s t before the 45° pulse; that i s , the value of the sweep-width and d i g i t i z a t i o n i n P j .  Now,  i s determined by  the magnitude of t  w i l l favour cross-peaks a r i s i n g from a certain J .  / 5  d  ** 1 s, J_  With  - 0.25 Hz i s favoured, and with Jt^ - 200 ms, £ - 1.25 Hz i s favoured. A i s non-zero only when long-range couplings are of i n t e r e s t . F i g . II.2.32A shows a contour plot of the decoupled COSY spectrum of 1. dimensions.  Sinebell resolution enhancement was u t i l i z e d i n both  A sweep-width i n J_i of ± 350 Hz, and the 256 x 512 complex  point data matrix resulted i n t ^ - 182 ms.  Decoupling i s seen to be  good; the only v i s i b l e imperfections i n this plot i s a set of spurious responses at the mid-point between the geminal protons i n F_2, which arises through strong coupling.  Other smaller imperfections are not  v i s i b l e i n the contour plot, but are seen i n the F j projection - they are signals aliased i n J_2 about J_2 - 0 and are c a l l e d "quadrature images".  These are presumably due to imperfect phase settings for the  pulses and residual transverse magnetization, and could possibly be eliminated by extending the phase cycling to include CYCLOPS.  Once  again the J(4,5) c o r r e l a t i o n i s weak i n this experiment, as i t was i n a l l other previous Jeener experiments. •Again, a composite p u l s e  3 8  i s used: 180°  = 90°  180°  x  90°  -  Fig.  II.2.32  Decoupled ms  and  A •  COSY 0.  projection.(C),  183  experiment (A),  -  with  1 at  contour-plot  control  spectrum.  of  270 MHz. 2D  tj  spectrum.  -  182 (B),  Fj  - 184 -  The experiment c l e a r l y has elements of COSY and homonuclear 2D correlated spectra, and i f the d i g i t i z a t i o n i n F_ were fine enough, this 2  single experiment could replace both these experiments, with the added advantage of the s i m p l i f i c a t i o n of the COSY content.  Of course,  caution  would have to be exercised with the detailed interpretation of line-shapes i n F . 7  II.2.5.4  Heteronuclear  Chemical S h i f t Correlation  This very powerful 2D experiment correlates the chemical s h i f t s of heteronuclei.  Usually F_j i s chosen as the chemical s h i f t axis 13  for protons, and such as B , 11  15  as  C.  Although i t B uses with other heteronuclei  N and P have been reported,** 3 1  5  we s h a l l r e s t r i c t our  discussion to the C - *H chemical s h i f t c o r r e l a t i o n map (CSCM) 1 3  7 6 77  experiment  »  which correlates carbon resonances with the d i r e c t l y  bound proton(s), as this i s the most common and diagnostically useful one.  The pulse-sequence i s best designed such that  C resonances (J_2)  are broad-band decoupled, whilst H resonances (F^) show homonuclear 1  coupling.  These couplings turn out to be useful i n interpretation of  data, even though (a) a detailed examination of the proton m u l t i p l i c i t i e s can lead to errors a r i s i n g from i n f i d e l i t i e s inherent to the experiment, and (b) the d i g i t i z a t i o n i n F^ i s seldom good enough to permit the observation of fine d e t a i l i n homonuclear proton The experiment involves a magnetization-transfer  couplings.  step (see DEPT,  II.2.2.2) and i s therefore inherently quite s e n s i t i v e . T y p i c a l l y , 512 experiments are performed, c o l l e c t i n g 2K complex data points i n _t » The 2  minimum sweep-width i n J_  2  i s chosen and one can a l i a s quaternary signals  since they w i l l give no c o r r e l a t i o n s . peaks w i l l be absent).  ( S i m i l a r l y , deuterated  solvent  - 185 -  The pulse-sequence  has been refined to one having  quadrature  detection i n both dimensions and the c h a r a c t e r i s t i c s l i s t e d above.  A  minimum of four phase-cycles are mandatory, but this i s often extended to eight (Eq. II.2.35).  X  13  H:  RD - 90°(.^  - 0.5tj -  C:  - 0.5tj - A - 90°($ ) - A x  2  90°(*O  180°($ ) 3  2  - BB - AON  (t ;*5> 2  [II.2.34] Cycle 1  2  3  4  5  6  7  8  $1  X  X  X  X  -x  -x  -x  -x  $2  X  y  -x  -y  -x  -y  X  y  $  3  X  X  X  X  -x  -x  -x  -x  $  4  X  X  X  X  X  X  X  X  X  y  -x  -y  X  y  -x  -y  &5  [II.2.35] The experiment may be understood  by considering the populations  for an isolated C - H fragment, as given i n F i g . II.2.33A; the r e l a t i v e spin populations r e f l e c t the four-fold s e n s i t i v i t y difference between * H and  C , v i z . YJJ = 4y^. Now, we may ignore the 1 8 0 ° ,  C pulse at  this stage, and concentrate on the proton pulses: 9 0 ° - t^ - 9 0 ° .  A  population inversion across one of the proton transitions results i n the s i t u a t i o n i n F i g . II.2.33B. The affect of two 9 0 ° ( * H ) pulses, separated by tl to produce "population inversion" i s shown i n F i g . TI.2.34A and B.  The i n i t i a l 9 0 °  * H pulse creates transverse magnetization which precesses i n the x'-y' plane according to jJ_c-H*  Depending on i t s position a f t e r t_i, the  second 9 0 ° pulse may, i n the l i m i t s , cause population inversion,  - 186 -  B. F i g . II.2.33  Energy levels of a CH fragment at equilibrium (A) and after a proton population inversion (B). (From a review by R. Freeman, Ref. 45).  PROTON F R E E PRECESSION (U)  Wtio! 90* proton pulse  Second 9 0 ' proton pulse  Proton populations  diverted  Saturated  Boftzmam  Saturated  n i verted  Change in ° C moanetization  F i g . II.2.34  The affect of a proton 90° - _ti - 90° pulse sequence on proton vectors, and their interpretation i n terms of proton spin state populations i s given i n (A) and (B). ( C ) maps the corresponding change i n population differences across the C transitions. (Adapted from Freeman and Morris. )  -  187  -  saturation, or regeneration of Boltzmann state populations.  The  proton  populations follow a cosine modulation after the second 90° pulse. Since the  i3  C  transitions share an energy l e v e l In common with t h i s 13  modulated proton t r a n s i t i o n , the population differences between  C  levels are necessarily affected and this can be monitored with a 90°  C  "read" pulse. Four antiphase signals would be detected from such a pulse-sequence (90° - t^ - 90° 90° - AQN),  showing heteronuclear  coupling i n both frequency dimensions. It would be desirable to remove the heteronuclear coupling; t h i s i s accomplished i n by employing 1 13 broad-band H decoupling during C a c q u i s i t i o n , and i n J_i by i n s e r t i n g  13 a  C, 180° pulse halfway through the _tj period.  be expected  This experiment might  to result i n one t r a n s i t i o n for a C-H  fragment,* but would  actually result i n no signal at a l l , due to the exact cancellation of the four antiphase signals.  Two  delays, A} and A  2  are Inserted to allow  180° r e l a t i v e phase rotation between these components (which d i f f e r i n frequency by Jp_ ) and this provides a p r a c t i c a l solution to the problem described above. This i s the experiment described i n Eq. II.2.34. For a value of (2 * ^J^VE) effects the required phase 13 H  -1  rotation of the proton vectors.  Considering the  C vectors, the  situation i s more complex and c a l l s for a compromise, since the time for the phase rotation w i l l depend on the number of attached protons. has shown** that a good compromise i s to select A 6  Further, two refinements e x i s t .  2  • 0,3/^  To improve the  i 3  C  Jj^* decoupling,  a composite 180° p u l s e (180° - 90° 180° 90° ) i s recommended. •This would be located at 6 ( i n F 2 ) and fig ( i n _F_i). 3 8  C  Bax  It  - 188 -  has also been demonstrated that broad-band  C decoupling during t i may  78 be effected with an MLEV-16-type procedure  giving good results without  7 9  the problem of excessive sample heating. Let us now pause to consider the merits and shortcomings of the experiment.  I t s strengths l i e i n the r e l a t i v e ease with which i t i s  performed with a l l variables being determinable before the 2D experiment i s performed.  The s e n s i t i v i t y i s good, since one benefits from the nOe  as well as the polarization transfer.  Further, the r e p e t i t i o n rate i s  based on proton relaxation parameters as opposed to the (generally 13 longer)  C values.  A relaxation delay of ca. 1.3 Tj i s optimal.  One  may find spurious peaks at 6Q when the protons attached to the carbon form a tightly-coupled AB p a i r .  Secondly, since the experiment r e l i e s  on p o l a r i z a t i o n transfer, t h i s must be e f f i c i e n t .  Hence the condition  (1/2J) < T_2 i s necessary, which may r e s t r i c t work with very large, conformationally r i g i d macromolecules.  S t i l l , one major factor makes  the experiment very a t t r a c t i v e ; one uses the very high dispersion of the 13  C  spectrum as a lever for the *H spectrum. F i g . II.2.35 shows the result of an experiment performed on a  0.5 M solution of 2 i n C D , displayed i n the contour plot mode. 6  6  Details of the proton spectrum are not clear even when these are taken out as s l i c e s (not shown).  In that the assignment of the proton  1 ^  spectrum i s known, the  C assignment immediately follows (going from  low- to h i g h - f i e l d : C-1, C-2, C-3, C-4, C-5, C-6). We conclude with a few remarks addressed to some p r a c t i c a l considerations and suggest a general protocol. The experiment may be performed with a proton spectrum.  13  C probe, using the  1 H decoupler c o i l to pulse the  -  1.  Tune the C and ^ 1 3  189  -  c o i l s and shim the magnet  s i n c e p r o t o n s p i n n i n g side-bands 2.  Record  the  1 3  C spectrum  carefully,  can be bothersome.  (quadrature d e t e c t i o n ; broad-band  1  H  d e c o u p l i n g , w i t h the minimum sweep-width n e c e s s a r y q u a t e r n a r y peaks may be f o l d e d o v e r ) .  Select a block size  which g i v e s an a c q u i s i t i o n time (AT) > 1T_2 and adequate d i g i t i z a t i o n f o r c l o s e l y spaced sample spectrum  1 3  C resonances.  Save a  on d i s c , w i t h v e r y good d i g i t i z a t i o n and  signal-to-noise.  C-1  C-2  C-3 C-5 c-4  C-6  J L 435-  100  90  80  60 6  70  1  H-5 •  E,  C  1  1  H-6'  •  -  CH -H-4  1  1 -435-  -H-3  E (6c) 2  F i g . II.2.35  CSCM on a 0.5 M s o l u t i o n o f 2, a t 100.3 MHz f o r C . 1 3  - 190 -  3.  Determine  C pulse-lengths.*  I f these measurements are  made on the sample I t s e l f , i t may  be best to observe  the  fastest relaxing carbon (e.g. -CH ). 3  4.  Set up the spectrometer  to observe protons, pulsing and  observing with the decoupler c o i l .  Set the decoupler offset  to the middle of the *H spectrum and choose a minimum sweep-width without introducing a l i a s i n g . spectrum on disc.  Note the dwell-time  (DW)  Save a sample for the  p a r t i c u l a r sweep-width. 5.  Measure the *H t_go (90° pulse-length), observing  6.  Change back to  7.  Determine an accurate *H 90° pulse length: a.  13  C  observe.  By residual coupling i n a single frequency decoupled 2  (YB ) 2  off-resonance  (SFORD) experiment. ^  - AVFT —RES = 1/4( B ) Y  Av  *H.  2  1) Hz  s  [II. 2.36]  [II.2.37]  = distance (Hz) the decoupler i s placed from the proton resonance frequency.  J_0  = J_(C-H) for the attached carbon i n the absence of continuous-wave decoupling. - residual J_(C-H) for the attached carbon with continuous-wave decoupling during acqusition.  *For determination of a l l pulse lengths, one may choose to use a simpler sample at higher concentration, such as methanol.  -  191  -  With this experiment i t i s necessary to know at least one  1  3  C - *H assignment and determine the *H frequencies 13  C observe.  before returning to  Two experiments are  performed, one with the decoupler off during acqusition (to give Jfl) and one with i t on at i t s highest power JRES^*  l e v e l during acqusition (to give too i s now easily calculated. b.  (Eqns.  Knowing Av, 11.2.36,37).  By a simple coherence-transfer experiment. 1  RD -  C :  3  X  90°  -  1/2J  -  H:  - AQN 9  [II.2.38]  - BB  This simple pulse scheme i s performed after any J ( C - H ) has been measured.  When 8  M  9 0 ° , no signal i s seen for  13  C resonance.  this 8.  The  CSCM  9.  Ai,  A2  10.  experiment i s now set up.  are selected for a t y p i c a l  J£H»  E  * 8 *  135  Hz.*  The incremental delay i n _ti i s set to half the proton dwell time (see  4)  or calculated from  II.2.24.  The number of  experiments to be performed ( t y p i c a l l y , 5 1 2 ) should give an acqusition time i n _ti of ca. 11.  1 0 0 ms.  The number of scans at each jti value should be set to the value which gives "reasonable" signal-to-noise i n the ID  1  experiment. 12.  A relaxation delay of ca. 1 . 3 * T_i ( H) i s chosen.  *Ai °  1  3.7  ms and A  2  ™  2.2  ms.  3  C  -  13.  Data a r e now c o l l e c t e d  192 -  (minimum time f o r experiment « c a . 1  hour). 14.  The d a t a a r e p r o c e s s e d w i t h as much resolution-enhancement i n b o t h dimensions as the s i g n a l - t o - n o i s e p e r m i t s . double-exponential or s h i f t e d  A  s i n e b e l l m u l t i p l i e r should be  used i n the data m a n i p u l a t i o n .  II.Z.5.5  Relayed Coherence Transfer  80 T h i s 2D experiment was suggested by B o l t o n  as an assignment  a i d , h a v i n g elements o f CSCM and homonuclear c o r r e l a t i o n experiment.  A  CSCM-like d i s p l a y h a v i n g , i n a d d i t i o n , c o r r e l a t i o n s when the a t t a c h e d p r o t o n s are s c a l a r c o u p l e d . F i g . II.2.36A.  The f i r s t  coherence between H heteronucleus,  A  stage o f the experiment i n v o l v e s t r a n s f e r o f  and H^.  T h i s may then be r e l a y e d  t o the  In a CSCM experiment.  C  C  C  X  A  F i g . II.2.36  C o n s i d e r an AMX fragment as i n  B  X  C  X  C  C  C  Steps o f m a g n e t i z a t i o n t r a n s f e r i n the r e l a y e d t r a n s f e r experiment.  The p u l s e sequence II.2.40, r e s p e c t i v e l y :  coherence  and p h a s e - c y c l i n g are as per II.2.39 and  - 193 -  X  H: RD-90°-0. 51^-  1 3  C:  - 0 . 5 _ t i ~ 9 0 ( * i ) - 0 . 5 t -l80°-0.5t^-Ai-90 ( $ ) - A - B B 0  0  ffi  2  180°  2  90°($ )-  AQN <t )  2  2  [II.2.39] Cycle 1  2  3  4  5  6  7  8  9  10  *1  X  -x  y  -y  X  -x  y  -y  X  -x  y  $2  X  X  X  X  -x  -x  -x  -x  y  y  y  13  14  15  16  -y  X  -x  y  -y  y  -y  -y  -y  -y  12  n  [II.2.40] The  f r e e i n d u c t i o n decays are co-added and ( t m + A i ) ™ c a . 33  ms, f a v o u r i n g  JAM = 7.7 Hz (1/4JAM).  The experiment  was f i r s t performed  on a c o n c e n t r a t e d sample of  n - p r o p a n o l , d i s p l a y e d as a s t a c k e d - p l o t i n F i g . I I . 2 . 3 7 .  Peaks a r i s i n g  from h e t e r o n u c l e a r c o r r e l a t i o n s a r e l a b e l l e d (*) and the o t h e r symmetrical c o r r e l a t i o n s i d e n t i f y neighboring carbons/protons. t h e one experiment  Hence,  has the p o t e n t i a l t o a c t as a t o o l i n d e t e r m i n i n g  h e t e r o n u c l e a r c o r r e l a t i o n s and c o u p l i n g pathways w i t h i n a m o l e c u l e .  The  1  13  C and  H assignments The experiment  are o b v i o u s . w i l l not be d e a l t w i t h at any l e n g t h , s i n c e i t  was observed i n t h e a u t h o r ' s hands t o be l i m i t e d t o s i m p l e m o l e c u l e s having weakly-coupled  proton spectra.  When s o l u t i o n s of s u c r o s e and 2  were t r i e d , t h e r e s u l t s were l e s s than s a t i s f a c t o r y . homonuclear c o r r e l a t i o n experiment  An  which has a COSY-llke  analogous display  81  identifies  a f a m i l y of s c a l a r coupled s p i n s .  The output i s q u i t e  -  194  -  messy even for the single AMQ X system i l l u s t r a t e d , and one i s led to 3  speculate that even two such overlapping systems would be d i f f i c u l t to decipher.  F i g . II.2.37  Relayed coherence transfer experiment performed on n-propanol (*H frequency 400 MHz). Peaks present i n the CSCM experiment are labelled with an asterisk. 82 83  II.2.5.6  2D Nuclear Overhauser Enhancement Spectroscopy > A 2D experiment exists which explores magnetization transfer  (via chemical exchange or nOe) and, i n the nOe sense, i s called 2D nuclear Overhauser enhancement spectroscopy or "NOESY".  The experiment  bears many s i m i l a r i t i e s to COSY/SECSY; i t involves a three-pulse  - 195 -  sequence and i s an autocorrelation experiment, correlating spins involved i n magnetization transfer, such as nOe. The display of the experimental output may be chosen as COSY-like (II.2.47A) or SECSY-like (II.2.41B). RD - 90°($ ) - t_i - 9 0 ( $ ) ~ ^ i x " 8  1  9  0  2  -  RD - 90°($i)  0.5t\ - 90°($ ) - f 2  mlx  ^ ^  5  " A0N(_t ;^)  [II.2.41A]  2  -90°($3) - 0.5t^ - AQN(t ;5>0 2  [II.2.41B]  For sequence II.2.41A the display shows a plot of 6 vs 6 with the normal spectrum along the diagonal, and cross-peaks indicating magnetization transfer ( c f . COSY).  With sequence II.2.41B, Fj. represents A6/2  and  the normal spectrum l i e s at F_ - 0 ( c f . SECSY). x  Phase-cycling Is important i n this experiment and achieves two purposes:  (a) quadrature detection i n Fj with quadrature image  supression and (b) elimination of multiple quantum coherence contributions to cross-peaks (see l a t e r ) .  (MQC)  H u l l proposed  the following  scheme. Cycle 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  *l  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  $2  X  y  -x  -y  X  y  -x  -y  -x  -y  X  y  -x  -y  X  y  $3  X  -y  X  -y  y  X  y  X  -x  y  -x  y  -y  -x  -y  -x  *4  X  X  -x  -x  y  y  -y  X  X  -x  -x  y  y  -y  -y  -y  tH.2.42]  -  196 -  Since the experiment was f i r s t suggested, several detailed papers have emerged from the same group concentrating on t h e o r e t i c a l aspects, * ' 8 2  8 3  8 5  refinements  and a p p l i c a t i o n s .  8 6 - 8 9  9 0 - 9 2  A brief  description i s given i n F i g . II.2.38.  90  Preparation  90  X  Evolution •A  F i g . II.2.38  90  X  Mixing B'  •C  ( T  mix  )  X  Detection ( t ^ ) •E  Schematized NOESY experiment and vector model.  - 197 -  We consider two (homonuclear) spin-1/2 n u c l e i , close i n space and not scalar coupled.  The "preparation" period i s terminated by the f i r s t 90°  "preparatory" pulse, which creates transverse magnetization (Fig. II.2.38A).  The "evolution" period (t^i)* serves to frequency-label the  vectors ( c f . COSY) and at the end of this period they may be represented by F i g . II.2.38B. z-magnetization  A second "mixing" pulse now generates  (Fig. II.2.38C);  transverse magnetization  also exists at  this stage but i s undesirable and removed by phase-cycling and, therefore, omitted from the diagram. %ix>  magnetization  During the "mixing" period,  transfer (e.g. dipole-dipole or cross relaxation)  processes occur where possible.  The end of the mixing period (Fig.  II.2.38D) i s marked by a f i n a l 90° "detection" pulse, which recreates detectable transverse magnetization  (Fig. II.2.38E) which i s sampled  during the "detection" period, t_2> Normal 2D FT yields a COSY-like plot with a diagonal bearing the ID spectrum and cross-peaks magnetization  a r i s i n g where  transfer processes occurred during t ^ x ( i d e a l l y ) .  We now look at the experiment i n the light of the motion of the molecule.  A rapidly tumbling, low molecular weight molecule  (extreme-narrowing l i m i t ) w i l l behave d i f f e r e n t l y from a large, high molecular weight molecule at h i g h - f i e l d (spin d i f f u s i o n l i m i t ) .  In the  extreme-narrowing l i m i t , the dipole-dipole relaxation mechanism often dominates and NOESY cross-peaks would be negative going compared to the diagonal peaks ( i f phase-sensitive display were used) and small - a maximum of ca. 10% of diagonal peak height ( t ^ x case of uin^c ° 1*12 leads to zero cross-peak  =  T  intensity.  n  e  critical  In the spin  *_ti i s systematically incremented i n the usual fashion.  - 198 -  d i f f u s i o n regime, cross relaxation dominates and cross peaks have the same phase as diagonal peaks.  Moreover, cross- and diagonal-vpeak  i n t e n s i t i e s can be quite comparable under favourable conditions.  The  degree of intragroup relaxation i s an important factor i n the two regimes.  I f one were considering two close methyl groups i n a molecule,  these protons would have strong intragroup relaxation and under extreme narrowing conditions the cross-peaks would amost c e r t a i n l y be undetectably small; i n the spin d i f f u s i o n l i m i t , however, strong cross-peaks could be expected. When attempting a NOESY experiment, a decision has to be made regarding the choice of T^^* i n t e n s i t y as a function of T m i extreme-narrowing  The v a r i a t i o n of nOe cross-peak x  i s well understood.  In the  l i m i t * , the cross-peak build up occurs as a function  of Tmix i s roughly analogous to those seen i n the TOE experiment (II.2.4.2) and nOe builds up l i n e a r l y with  to a maximum value,  and then tapers o f f . In such a case an optimal i i m  x  would be » 1T_J.  In the spin d i f f u s i o n l i m i t , * the build up curve i s e n t i r e l y analogous 93  to that obtained from a driven nOe experiment  (1.2.4.3).  Generally, a  mixing time can be selected by doing a t r i a l driven nOe experiment; an i r r a d i a t i o n time which produces a strong negative nOe can be used as %ix'  *In the extreme narrowing l i m i t the spin d i f f u s i o n l i m i t , ca. 100 ms.  w i l l be ca. Is, and i n  - 199 -  A complication with the NOESY experiment, alluded to e a r l i e r , i s that more than one mechanism can give r i s e to cross-peaks. and MQC COSY.  Zero (Z-)  give r i s e to cross-peaks with scalar coupled spins, analogous to These are referred to as J_ cross-peaks and arise from coherent  magnetization transfer (vs. incoherent magnetization transfer i n the case of nOe). contributions.  Several schemes have been put forward to supress these F i r s t - , second- and third order MQC 8 3  removed by phase-cyling.  »  86  may  be easily  The problem remains with ZQC.  Several  87  8 5  ideas have been suggested  »  i n this respect but have not been  systematically evaluated: a.  D i g i t a l apodization can be used to discriminate against J_ cross-peaks.  b.  Randomized  T  m  Instead of using exactly the same Tmix  j y «  i n each experiment, (e.g. 0.9 < T 1.0 s ) .  NOe  m  T ^ X i s randomly varied 10-20%  <. 1.1 i s a 10% randomization of x ^ x  •  cross-peaks are hardly affected by t h i s , but J_  cross-peak i n t e n s i t y behaves i n an o s c i l l a t o r y fashion during Xmix  an(  *  t n  e  randomization causes a smearing of J_  cross-peaks i n the Fj axis, resulting i n an additional noise band s i g n i f i c a n t l y reduced In intensity and, hopefully, below the contour l e v e l threshold.  The noise band could,  however, obscure a small nOe response of Interest. c.  P a r t i a l refocussing. Introduction of a refocussing 180° pulse placed at random positions during T ^ X  n a s  suggested to result i n a similar effect to (b).  been  - 200 -  II.2.39  2D NOESY spectrum of 1 at 400 MHz. was used, randomized ± 10%.  A mixing time of 0.3  - 201 -  d.  Sequential v a r i a t i o n of  Y  .  In this experiment,  x  m i x  i s progressively incremented i n p a r a l l e l with the evolution time, j t j .  The mixing time i s set  - x ^ ^ + X t.i,  where x° Is the i n i t i a l value and x * mix T  mix ^  ( 3 * 'ac  u  s:  t:|l  - -^ on t  we  i - £.].)• n  s  chosen such that  ^ symmetrization  procedure i s imperative to eliminate ZQC and MQC are not symmetrical about the l i n e Fj = J_2.  peaks which  Here, JT  cross-peaks are not smeared out as i n the case of (b), but i t was advised that the experiment be repeated for a second T  mix  t o  r  e  m  o  v  e  a n  Y  doubts of ambiguities a r i s i n g from  accidental symmetrical d i s p o s i t i o n of Z- or MQC  peaks.  In the author's hands, the simplest procedure has been the u t i l i z a t i o n of phase-cycling, the randomized apodization i n both jt  x  x^x  and s i n e b e l l  and _t « 2  F i g . II.2.39 shows a contour plot of the NOESY experiment  for 1,  which should be compared with the J_-correlated spectrum ( F i g . I I . 2.28). II.2.6  Conclusions The reader has been introduced to a large number of the  NMR  experiments which we believe w i l l find greater u t i l i t y i n the years to come.  Where possible, the theory has been explained at as p i c t o r i a l a  l e v e l as possible since the assumption has been made that the reader w i l l not have had extensive experience i n these areas. Except where an experiment has f a i l e d (relayed  coherence  transfer), since  the  these  have  intention  been viewed  with  illustrate  potentially  several  these  of  evaluation  effective  protocol  NMR  graph,  2D NMR r e l a t e d steady  workers  of  "real-life"  focus source  will for the  on  used  structure  s t i l l the  results. from  to  as  the  we w o r k of  a  light  identify, In  possible, and  following  our  elucidation  as  explain  way  experiments  specifically,  organic  considering  of  the  of  chapters,  towards to  a  yield  organic  to  a  In  the  further  rapid  an  molecules  the  This  serve  chemist  a particular  thesis new  as  or  a  air  their  of  rather will of  with  the  application and  ID  should  subsequent  have  their  drawing  as  This  in  as  a  the  chapter,  2D NMR  find  on  their  evaluation  assist  and  1981.  to  articles  will  confidence.  spectroscoptist  experiment.  involved  than  of  community to  papers  by  standard  1979  1983  years  of  from  hopefully  compilation  4  of  incorporation  scientific  literature  of  3 or  handful  into  phase  were  a  The  the  opinion,  numbers  experiments,  a  groups  explosion  In  publications first  by  soft-ware  growth  and  of  The  development.  authors's  Increasing  1975.  generated  few  number  dominated  necessary  experiments,  2D NMR w i t h should  relative  conception  fundamental  relatively  information. of  the  Its  led  first  a  shows  excitement  years,  problems.  data  of  an  is  2D NMR p u b l i c a t i o n ,  the  witness  acceptance  which  of  and  preceding  assessment  2 D NMR  impartial  combination  since  manufacturers  Interestingly,  years  papers  rate  spectrometers, the  the  below,  concentrating  instrument  in  for  as  experiments.  are  the  in  -  spectroscopy.  The  saw a  of  202  chapter  useful  experients  practical  using  this  -  useful  experiments when  -  Year  Relative  1975  *  1976  *  1977  ***  1978  ***  1979  *********  1930  ************  1981  ********************  1982  **********************  Table  II.2.1  An  Number  indication  relating  to  of  of  203  -  Publications  the  2D NMR i n  relative the  number  literature,  of  publications  categorized  by  year.  - 204 -  REFERENCES  1.  M.L. Martin, J . - J . Delpeuch, and G.J. Martin. P r a c t i c a l spectroscopy. Heyden and Son Ltd., Philadelphia. 1980.  2.  T.C. Farrar and E.D. Becker. Pulse and Fourier transform Introduction to theory and methods. Academic press, N.Y.  3.  D. Shaw. Fourier transform N.M.R. spectroscopy. York. 1976.  4.  J.C. Lindon and A.G. Ferrige. 27-66 (1980).  Progress i n NMR  5.  L.D. Colebrook and L.D. H a l l .  Can. J . Chem. 58, 2016-2023 (1980).  6.  L.D. H a l l and J.K.M. Sanders. (1980).  J . Amer. Chem. Soc. 102, 5703-5711  7.  Ref. 1, pp. 254-255.  8.  K.F. Wong.  9.  G.E. Martin. Ernst.  Ph.D.  NMR NMR. 1971.  Elsevier,  Specroscopy, 14,  thesis, University of B r i t i s h Columbia,  J . Pharm. S c i . 70, 81-84  New  1980.  (1981).  10.  R.R.  J . Chem. Phys. 45, 3845-3861 (1966).  11.  H.J. Osten and R. Radeglia. J . Magn. Reson. 49, 8-21  12.  D.P.  13.  D.T. Pegg, D.M. Doddrell, W.M. Reson. 44, 32-40 (1981).  Brooks, and M.R.  14.  S.L. Patt and J.N. Schoolery. (1982).  J . Magn. Reson. 46, 535 - 539  15.  J.-C. B e l o e i l , C. Le Cocq, and Y.-V Lallemand. Reson. 19, 112-115.  16.  D.M. Doddrell, D.T. Pegg, and M.R. 323-327 (1982).  Bendall.  17.  H.J. Jakobsen, O.W. Sorensen, W.S. Reson. 48, 328-335 (1982).  Brey, and P. Kanyha.  18.  F.-K.  19.  M.R. Bendall, D.M. Doddrell, D.T. Pegg, and W.E. H u l l . High resolution multipulse NMR spectrum editing and DEPT. Bruker Spectrospin note. 1982.  Burum and R.R.  Ernst.  Pel and R. Freeman.  (1982).  J . Magn. Reson. 39, 163-168 (1980). Bendall.  J . Magn.  Org. Magn.  J . Magn Reson. 48, J . Magn.  J . Magn. Reson. 48, 318-322 (1982).  - 205 -  20.  R. Richarz, W. Amman, and T. W i r t h l i n . XL-200/300 i n DEPT: A new experiment for the ADEPT spectroscopist! Varian Associates note Z-15. 1982.  21.  C.D. Barry, CM. Dobson, D.A. Swiegart, L.E. Ford, and R.J.P. Williams. NMR s h i f t reagents. Edited by R.E. Sievers. Academic Press, New York. 1973. pp 173-195.  22.  M. Kuo and W.A.  23.  L.D. H a l l and J.K.M. Sanders. (1980).  24.  J.D. Mersh and J.K.M. Sanders.  25.  J.H. Noggle and R.E. Schirmer. The nuclear Overhauser e f f e c t . Chemical applications. Academic Press, New York. 1971.  26.  B.P. Roques, R. Rao, and D. Marion.  27.  A.A. Bothner-By. B i o l o g i c a l applications of magnetic resonance. Edited by R.G. Shulman. Academic Press, New York. 1979. pp 177-219.  28.  I.D. Campbell, CM. Dobson, R.G. R a t c l l f f e , and R.J.P. Williams. J . Magn. Reson. 19, 397-417 (1978).  29.  Ref. 26, pp. 26-31; r e f . 1, p. 18.  30.  R. Freeman, H.D.W. H i l l , B.L. Tomlinson, and L.D. H a l l . Phys. 61, 4466-4473 (1974).  31.  J.D. Mersh and J.K.M. Sanders. (1982).  32.  Ref. 26. pp.  33.  H. Thogersen, R.U. Lemieux, K. Bock, and B. Meyer. 60, 44-57 (1982).  34.  R. Richarz and K. Wuthrich.  35.  B.L. Tomlinson and H.D.W. H i l l . (1973).  36.  Ref. 26, pp. 116-117.  37.  A.A. Bothner-By and J.H. Noggle. 5152-5155, (1979).  38.  M.H. L e v i t t . J . Magn. Reson. 48, 234-264 (1982), and references c i t e d therein.  Gibbons.  J . B i o l . Chem. 254, 6278-6287 (1979). J . Chem. Soc. Chem. Commun., 368-370 J . Magn. Reson. 50, 289-298 (1982).  Biochimle 62, 753-773 (1980).  J . Chem.  Org. Magn. Reson. 18, 122-124  57-69. Can. J . Chem.  J . Magn. Reson. 30, 147-150 (1978). J . Chem. Phys. 59, 1775-1784  J . Amer. Chem. Soc. 101,  -  206  -  39.  N.M. Szeverenyl, A.A. Bothner-By, and R. Bittner. 84, 2880-2883 (1980).  J . Phys. Chem.  40.  S.L. Gordon and K. Wiithrich. (1978) .  J . Amer. Chem. Soc. 100, 7094-7096  41.  G.A. Morris and R. Freeman.  J . Magn. Reson. 29, 433-462 (1978).  42.  G. Wagner and K. Wiithrich.  43.  F.M. Poulsen, J.C. Hoch, and CM. Dobson. (1980).  44.  CM. Dobson, E.T. Olejniczak, F.M. Poulsen, and R.G. R a t c l i f f e . J . Magn. Reson. 48, 97-110 (1982).  45.  Recent reviews: G.A. Morris. Fourier, Hadamard, and Hilbert transforms i n chemistry. Edited by A.G. Marshall. Plenum Press, New York. 1982. pp. 271-305. K. Nagayama. Adv. Biophys. 14, 139-204 (1981). R. Freeman. Proc. R. Soc. Lond. A. 373, 149-178 (1980). R. Freeman and G.A. Morris. B u l l . Magn. Reson. 1, 5-26 (1979) .  46.  A. Bax. Two-dimensional nuclear magnetic resonance i n l i q u i d s . D. Reidel, Holland. 1982.  47.  S. Sukumar.  48.  G. Bodenhausen, R. Freeman, R. Niedermeyer, and D.L. Turner. J . Magn. Reson. 26, 133-164 (1977).  49.  Ref. 46, pp. 41-43.  50.  S. Brownstein.  51.  K. Nagayama, P. Bachman, K. Wiithrich, and R.R. Ernst. Reson. 31, 133-148 (1978).  52.  D.L. Turner.  53.  W.P. Aue, P. Bachmann, A. Wokaun, and R.R. Ernst. 29, 523-533 (1978).  54.  Ref. 46, pp. 99-101.  55.  E.L. Hahn and D.E. Maxwell.  56.  L.D. H a l l and S. Sukumar. (1979).  57.  G. Bodenhausen, R.Freeman, and D.L. Turner. 511-514 (1977).  J . Magn. Reson. 33, 675-680 (1979). Biochem. 19, 2597-2607  Ph.D. Thesis, University of B r i t i s h Columbia.  1981.  J . Magn. Reson. 42, 150-154 (1981). J . Magn.  J . Magn. Reson. 49, 175-178 (1982). J . Magn. Reson.  Phys. Rev. 88, 1070-1084 (1952).  J . Amer. Chem. Soc. 101, 3120-3121 J . Magn. Reson. 27,  - 207 -  58.  G. Bodenhausen and D.L. Turner. (1980) .  J . Magn. Reson. 41, 200-206  59.  L.D. H a l l , G.A. Morris, and S. Sukumar. (1979).  60.  M. L e v i t t and R. Freeman.  J . Magn. Reson. 34, 675-678 (1979).  61.  L.D. H a l l and S. Sukumar.  J . Magn. Reson. 38, 555-558 (1980).  62.  A. Kumar.  63.  G. Bodenhausen, R. Freeman, G.A. Morris, and D.L. Turner. Magn. Reson. 31, 75-95 (1978).  64.  G. Wider, R. Bauraann, K. Nagayama, R.R. Ernst, and K. Wiithrich. J . Magn. Reson. 42, 73-87 (1981).  65.  A. Bax. R. Freeman, and G.A. Morris. (1981) .  66.  J.D. Mersh and J.K.M. Sanders.  67.  W.P. Aue, E. Bartholdi, and R.R. Ernst. 2229-2246 (1976).  68.  A. Bax and R. Freeman.  69.  K. Nagayama, A. Kumar, K. Wiithrich, and R.R. Ernst. Reson. 40, 321-334 (1980).  70.  R. Bauman, G. Wilder, R.R. Ernst, and K. Wiithrich. J . Magn. Reson. 44, 402-406 (1981).  71.  R. Bauman, A. Kumar, R.R. Ernst, and K. Wiithrich. 44, 76-83 (1981).  72.  Morris' review, r e f . 45.  73.  B. Bliimich and D. Ziessow.  74.  S. Sukumar, personal communication.  75.  A. Bax, personal communication.  76.  A. Bax and G.A. Morris. others i n the series.  77.  Ref. 46, pp. 50-65, and references cited therein.  78.  J . S. Waugh.  Carbohydr. Res. 76, C7-C9  J . Magn. Reson. 30, 227-249 (1978). J.  J . Magn. Reson. 43, 333-338  J . Magn. Reson. 50, 171-174 (1982). J . Chem. Phys. 64,  J . Magn. Reson. 44, 542-561 (1981). J . Magn.  J . Magn. Reson.  J . Magn. Reson. 49, 151-154 (1982).  J . Magn. Reson. 42, 501-505 (1981), and  J . Magn. Reson. 49, 517-521 (1982).  - 208 -  79.  P.H. Bolton.  J . Magn. Reson. 51, 134-137 (1983).  80.  P.H. Bolton.  J . Magn. Reson. 48, 336-340 (1982).  81.  G. E i c h , G. Bodenhausen, and R.R. 3731-3732 (1982).  82.  J . Jeener, B.H. Meier, P. Bachman, and R.R. Ernst. 71, 4546-4553 (1979).  83.  S. Macura and R.R. Ernst, Molec. Phys. 41, 95-117 (1980).  84.  W.E. H u l l . (1982).  85.  S. Macura, Y. Huang, D. Suter, and R.R. Ernst. 259-281 (1981).  86.  G. Bodenhausen and R.R.  87.  S. Macura, K. Wiithrich, and R.R. Ernst. 351-357 (1982).  J . Magn. Reson. 47,  88.  S. Macura, K. Wiithrich, and R.R. Ernst. 269-282 (1982).  J . Magn. Reson. 46,  89.  G. Bodenhausen and R.R. Ernst. (1982).  90.  A. Kumar, R.R. Ernst, and K. Wiithrich. Comm. 95, 1-6 (1980).  91.  A. Kumar, G. Wagner, R.R. Ernst, and K. Wiithrich. Biophys. Res. Comm. 96, 1156-1163 (1980).  92.  G. Wagner, A. Kumar, and K. Wiithrich. 375-384 (1981).  93.  A. Kumar, G. Wagner, R.R. Ernst, and K. Wiithrich. Soc. 103, 3654-3658 (1981).  Two-dimensional NMR.  Ernst.  Ernst.  J . Amer. Chem. Soc. 104,  Aspect 2000.  J . Chem. Phys.  Bruker Report J . Magn. Reson. 43  Molec. Phys. 47, 319-328 (1982).  J . Amer. Chem. Soc. 104, 1304-1309 Biochem. Biophys. Res. Biochem  Eur. J . Biochem. 114, J . Amer. Chem.  - 209 -  CHAPTER I I . 3 OLIGOSACCHARIDE CHARACTERIZATION USING NMR Complex carbohydrates are widely dispersed i n Nature where they play a variety of roles.  Often i n polymerized form, they constitute  s t r u c t u r a l materials (e.g. c e l l u l o s e , b a c t e r i a l c e l l - w a l l s ) , or energy storage molecules (e.g. starch, glycogen).  With b a c t e r i a l c e l l - w a l l s  they can serve the additional purpose of coding the immunogenicity organism presents.  the  An equally important role i s played by carbohydrates  conjugated to protein or l i p i d - glycoproteins or g l y c o l i p i d s respectively.  Again these serve the gamut of functions, as structural  molecules (e.g. collagen), the " l e t t e r s " of the blood-group  substance's  "alphabet", lubricants (e.g. g a s t r o - i n t e s t i n a l tract mucins), "anti-freeze" agents ( i n A r c t i c f i s h ) and bound to a host of proteins (e.g. some enzymes and the immunoglobulin speculated  proteins).  It has even been  that i n time, a l l proteins w i l l be found to contain some  carbohydrate! The ubiquity and importance of complex carbohydrates i n b i o l o g i c a l systems has, not s u r p r i s i n g l y , attracted much attention. I n i t i a l investigations tended to focus on primary structure - which sugars are present and the d e t a i l s of the ordering. However, i n the l a s t few years people have begun to try to relate structure with function, and to this end knowledge of the secondary, three-dimensional picture i s sought. Prior to this study, a wide range of technologies existed for sequencing complex carbohydrates, which w i l l be b r i e f l y reviewed below.  - 210 -  The l a s t few years has seen a demonstration  of the tremendous potential  of NMR i n this f i e l d and this chapter discusses the development of one such protocol which we believe to be of some potential. procedure i s the use of recently developed  P i v o t a l to the  techniques detailed i n the  preceding chapter; an understanding of these w i l l henceforth be assumed.  NMR has the broad advantages of being a non-destructive  technique with several extractable parameters, information content.  each with i t s own  As a counterpoint, the method i s inherently  i n s e n s i t i v e and demands on sample quantity may t o t a l l y preclude i t s use for many b i o l o g i c a l l y important systems. Before the 1970's, almost a l l structural studies hinged on c l a s s i c a l "wet" chemical analyses.  3  These demand r e l a t i v e l y large  amounts of sample, are labour intensive and have even been found to be incorrect when re-examined.  The early seventies saw the Infusion of  spectroscopic techniques and at this time, such methods must be considered indispensable.  Although the focus here i s primarily on NMR,  the u t i l i t y of mass-spectroscopy  cannot be overlooked, especially i n  l i g h t of i t s signal-to-noise s e n s i t i v i t y . 13  C NMR spectroscopy has been of great u t i l i t y i n the area, (often i n conjunction with "wet" methods), and recent reviews have been presented by Gorin  and Barker jat a l .  Nevertheless  C NMR spectra  remain d i f f i c u l t to f u l l y assign and structural assignments are t o t a l l y 7 8 convincing only when a series of molecules i s synthesized ' (or produced  by degradation) and a systematic investigation i s pursued.  Anomeric configuration may confidently be assigned on the basis of  -  either  C  chemical  shifts  or  211  -  JJCI-HI*  T  n  potential  e  exists  for  the  1 extraction  of  conformational  information,  9 couplings. However, these are often may h a v e t o s y n t h e s i z e the molecule has  limited  potential.  on i n t e r - r i n g  13 H-  C  s m a l l (<_ 6 H z ) , u n r e s o l v e d a n d o n e enriched in C ; as a r e s u l t this  1 0  approach  based  1 3  Nevertheless,  the  the  is  overall  simplicity  of  13 C NMR s p e c t r a as  superior  is  appealing  experiments  and  technique  and p r o t o c o l s  are  finding  designed.  In  a larger  the  place  limit  13 though,  the  quantities  fact may  serve  A number some and  that  of  considerable  C NMR i s to  limit  European effort  cerebrosides.  The  impractical  its  the  sub-milligram  utility.  and C a n a d i a n  to  with  groups  application  Franco-Dutch  group  of  of  have  already  ^H NMR t o  dedicated  oligopeptides  Vliegenthart  and  Montreuil 2  has  focussed  on  the  isolated  from  anomeric  and H-2  protons  protons.  Their  chemical  and, 11 made.  the  structurally  urine  of  patients  resonate shifts  by  reference  to  were  In  a broader  study  conclusions  were  drawn and  the  is  wholly  appropriate  attention  to  its  "library"  of  reference  complicate  lack  anomeric  concentrated  of  data by  for  proton  on g l y c o s p h i n g o l l p i d s  from  the  The bulk  constants  mannose of  were  the  ring  accurately  from known Carver  compounds, assignments 12 and Grey , useful  extended.  molecules because  Furthermore,  signals.  glycopeptides,  mannosidosis.  coupling  method  generality  data.  with  oligomannoside  down-field  and  measured  probably  related  13 and  of of  Although this  type,  approach  we  draw  its  reliance  on  virtual  coupling  may  group  Germany  Dabrowski's his  this  work  has  in  recently  a  has  been  14 reviewed.  Compounds  were  dissolved  I n DMSO-d^ and p r o t o n  measurements  - 212 -  made on sub-milligram quantities.  Extensive ID NMR  measurements,  together with catalogue data, allowed almost complete assignment cerebrosides having up to ten sugar residues. performed, no 2D NMR  of  At the time our study was  had been used In such structural studies, and the  developments are presented chronologically. 15 18  H a l l and co-workers have put some considerable e f f o r t the use of proton Jj's  ~  into  as a measure of inter-proton distances i n  carbohydrate molecules.  These, together with nOe methods described  19  earlier,  have been capitalized on by Lemieux and co-workers i n the  complex carbohydrate area.  An elegant study  7  used such NMR  techniques  to validate carbohydrate secondary structures determined by hard-sphere molecular modelling calculations, taking into account the exo-anomeric e f f e c t (HSEA). A series of blood-group determinant oligosaccharides 1  were synthesized and systematically studied by  13  H and  paper represents a land-mark i n both spectral assignment  C NMR.  This  and  conformational analysis of complex oligosaccharides. Cerebrosides have been derivatised and studied i n organic solvents by NMR.  Martin-Lomas and Chapman made per-O-acetylated 20  galactocerebrosldes  and noted large chemical s h i f t changes upon  varying solvents; this allowed them to make p a r t i a l f i r s t - o r d e r assignments  and confirm the anomeric configuration, and conformation of 21  the pyranose ring.  Falk et a l . studied  cerebrosides as their (>-methyl ethers. including a heptaglycosylcerebroslde.  more complex Lewis-like Six compounds were considered,  The fact that the O-CH3  resonances resonate at roughly the same chemical s h i f t as the other ring  - 213 -  protons r e s t r i c t e d their study mainly to anomeric protons at 6 > 4 . 0 .  A  method was developed for d i f f e r e n t i a t i n g Type I and Type II saccharide chains.  1  R =H  2  R=CO.CD  3  We now propose a protocol for the structure elucidation of 22  complex oligosaccharides, using NMR.  As a model compound we chose  a l l y l B-D-galactopyranosyl-B-(l->-4)-D-glucopyranoside ("8 a l l y l 23  l a c t o s i d e " ) , 1, which we prepared using standard procedures. The oligomer i s i n i t i a l l y considered as a mixture of glycosides of  the constitutent monosaccharides,  of  carbohydrate ring protons, and the determination of linkage order by  Inter-ring nOe measurements. 1.  and we aim for the f u l l  assignment  Our protocol i s as follows:  Maximize signal dispersion.  Use of a very h i g h - f i e l d  spectrometer with the parent compound may be adequate, but most times d e r i v a t i z a t i o n w i l l be necessary.  Preferably the  - 214 -  d e r i v a t i z a t i o n reagents should be f u l l y deuterated and, hence, spectroscopically s i l e n t , to preclude introduction of additional resonances which may overlap with those of the parent solute resonances. 2.  A solvent, pure or mixed, should be found which minimizes signal overlap.  3.  The assignment process i s begun with ID techniques, such as p a r t i a l l y T j relaxed spectra,  SDDS  and NOEDS.  Should  these prove s u f f i c i e n t to complete the assignment, steps 4 and 5 are omitted. 27  4.  The 2D J-resolved experiment resonance  i s used to locate the  frequencies of a l l ring protons.  P a r t i a l J_ spectra  allow the measurement of a l l coupling constants. 5.  Next, the J_ connectivities are established by the 2D J_ c o r r e l a t i o n experiment,  COSY (or SECSY); each ring acts as an  isolated spin system oblivious to i t s partners, since i n t e r - r i n g coupling w i l l be indetectably small or zero. 6.  A few very s p e c i f i c ID experiments  (3) may be necessary to  resolve any outstanding ambiguities i n assignment. 7.  Next, each unit i s i d e n t i f i e d (e.g. 8-galactopyranosyl) by  29 comparison of the measured J_-values with reference values. This i d e n t i f i e s the sugars, their anomeric  configurations  and ring sizes. 8.  The f i n a l task i s the determination of the sequence along the chain; t h i s i s done by determining the s p e c i f i c i n t e r - r i n g relaxation e f f e c t s .  With I D nOe experiments  i t i s necessary  - 215 -  Fig. II.3.1  With a B(l-4) glycosidic linkage, i r r a d i a t i o n of the glycosidic proton ( H - l ) induces i n t r a - r i n g ( ) and a single i n t e r - r i n g ( ) nOe. 1  that a l l the anomeric protons be separately resolved may be possible by changing solvent).  (which  Aside from known  i n t r a - r i n g nOe's (1,3 d i a x i a l protons, e.g.), each  F i g . II.3.2.  270 MHz spectra of (A) 8 - a l l y l lactoside p - a l l y l aceto(d_a) lactoside.  i n D2O, and (B)  - 216 -  irradiation  should  Fig. II.3.1).  induce  a unique i n t e r - r i n g nOe (see  Since a l l protons  l i n k a g e can be e a s i l y deduced. obtained  have been a s s i g n e d , the The same i n f o r m a t i o n may be  from a 2D NOESY e x p e r i m e n t .  3 0  In t h i s case t h e  advantage i s t h a t the anomeric s i g n a l s may be hidden; may be o f g r e a t e s t u t i l i t y  w i t h more complex  structures.  We i l l u s t r a t e t h i s approach u s i n g B - a l l y l l a c t o s i d e , exemplar. is  The 270 MHz H NMR spectrum o f a D 0 s o l u t i o n 1  2  shown i n F i g . II.3.2.A.  as the d o u b l e t s  The anomeric protons  near 6 4.5.  this  1, as the  o f the m o l e c u l e  are c l e a r l y  The bulk o f the r i n g protons  discernable  a l l resonate  between 6 3.9 and 3.5 and o n l y a few are s e p a r a t e l y d i s c e r n a b l e . corresponding synthesized  per-O-trideuterlo-acetylated derivative,  by s t a n d a r d m e t h o d s  31  2, was  using perdeuterio-acetic  I t s spectrum (0.2 M i n CgD *) i s shown i n F i g . II.3.2B. 6  many s i g n a l s ( a r i s i n g from O - a c e t y l a t e d spectrum i s c o n s i d e r a b l y s i m p l i f i e d . spin-decoupling  The  anhydride.  As expected,  c e n t r e s ) a r e d e s h i e l d e d and the  I n t h i s simple  experiments almost s u f f i c e d  instance,  t o determine the f u l l  assignment. 2D NMR i n v e s t i g a t i o n s began w i t h the 2D J - r e s o l v e d experiment which allowed five a l l y l  the d e t e c t i o n o f t e n carbohydrate  protons.  given i n F i g . II.3.3; recorded.  A contour  plot  r i n g protons  and the  o f the 2D J - r e s o l v e d experiment i s  t h e i r chemical  s h i f t s and c o u p l i n g c o n s t a n t s  were  Next, t h e COSY experiment was performed a t 400 MHz; the data  are shown i n F i g . I I . 3 . 4 ,  i n the c o n t o u r - p l o t mode a g a i n .  *In t h i s case, t h i s balance found t o minimize s i g n a l o v e r l a p .  It i s  o f c o n c e n t r a t i o n and s o l v e n t was  F i g . II.3.3  Contour plot of 2D J-resolved NMR spectrum of 2 (270 MHz; 0.2 M i n CgDg). The assigned, resolution-enhanced spectrum i s plotted above.  -  II.3.4  4 0 0 MHz within left right  COSY the  half, half.  spectrum  218  of  glucopyranosyl and  the  Allyl  -  2  (0.2 ring  M in are  6  6  shown  galactopyranosyl connectivities  C D ). (  are  ( )  Connectivities )  in  shown  in  the  the  top  bottom  (—•—•—).  - 219 -  convenient to begin tracing connectivities with the anomeric protons. For example, starting with H-l, the connectivities to the other glucosyl ring protons are clear from an inspection of F i g . II.3.4. In this way, i t was now possible to trace out a l l connectivities.  The only ambiguity  arose from the galactosyl H-2', H-3' and H-4 where cross-peak contour 1  overlap resulted from H-2' and H-4' resonating very close to one another and both being coupled with H-3'.  The assignment  was made after careful  Inspection of an expansion of this region of the COSY plot, and confirmed by matching coupling constants from the 2D J-resolved data. Comparison of these coupling constants with standard l i t e r a t u r e values i n Table II.3.1, confirmed the existence of a 6-galactopyranoside and a 8-glucopyranoside moiety. monosaccharides  The i d e n t i f i c a t i o n of the constituents  and their anomeric configuration was complete.  Next, the linkage sequence was determined from the nOe experiments  involving the anomeric protons.  The SSNOEDS experiment was  just feasible at 400 MHz with adequate decoupler-frequency s e l e c t i v i t y . The r e s u l t s , shown In F i g . II.3.5, c l e a r l y indicate a single " i n t e r - r i n g " nOe In each case.  As expected, i r r a d i a t i o n of the  galactosyl H-l' resonance induces an nOe into the glycosyl H-4 proton (as well as some l n t r a - r l n g nOe's), and i r r a d i a t i o n of the glucosyl H-l resonance induces an nOe into the a l l y l methylene pair, H - l ^ and H-l". Similar information i s obtained from the 2D NOESY experiment, the contour plot i n F i g . II.3.6. and H-4  (nji_^)  shown as  The unique i n t e r - r i n g nOe between H - l '  i s clear, but the nOe into the methylene pair i s small  and below the contour l e v e l threshold.  - 220 -  Table II.3.1 - Comparison of expected and obtained v i c i n a l ring proton coupling constants for (A) a 8-glucopyranoside, and ( B ) 8-galactopyranoside ring.  Expected  A:  B:  J(Hz)  2 9  Measured J(Hz)  (1,2)  7.8  8.0  (2,3)  9.6  9.9  (3,4)  8.8  8.9  (4,5)  9.5  10.4  (l'-2')  7.8  8.1  (2'-3')  9.8  11.0  (3'-4')  3.4  3.7  (4'-5«)  1.0  1.7  - 221 -  6.0  Fig. II.3.5  4.0  5.0  400 MHz SSNOEDS of 2 (0.2 M i n C D ) . 6  shown with a s o l i d l i n e ( broken l i n e ( ).  6  8  (ppm)  Intra-ring nOe's are  ) and l n t e r - r l n g  nOe's with a  - 222 -  F i g . II'.3.6  NOESY of 2 (0.2 M i n C D ) at 400 MHz. The mixing time, mix "0.75 8 ± 15%. Intra-ring nOe's for the glucosyl ( ) and galactosyl ( ) rings are drawn i n the bottom right half, and i n t e r - r i n g ( ) i n the top l e f t half. 6  T  6  Fig.  II.3.7  Combined taken  NOESY  (top left)  from F i g s .  II.3.4.  a n d COSY  (bottom  and II.3.6,  right)  respectively.  dat  - 224 -  Since symmetrical  both  the  COSY  a n d NOESY  about  the  principal  displays  diagonal,  are  they  similar  can  be  and  combined  into  is  in  a  32 single  display  which  has  all  the  information.  This  shown  Fig.  II.3.7. A number does  not  provides  use an  molecules has  a  wide  of  any  points  which w i l l  biological  minimal  reliance molecules.  knowledge,  no  instances  occur, extra  unusual  it  steps By  should  prove  be  experiments  etc.)  and  experiments  requires  position,  and has  set  has  up  high  over  to  in  that  what  to  been  note  If  from  such a the  such  chosen  more m a c h i n e  demands be  a  on  of  of  up  on  are  and  sample are  (delays, the neither  prerequisites. The  because  to  to  relatively  furthermore,  instrument  period  were  demands  decoupler  demands;  rings  couplings,  performing  time.  the  a  anomaly.  parameters  small;  sugar  and O - a c e t y l a t i o n  before  to  author's  situation  spectroscopy  signal-to-noise  calibration  may  is  approach  the  where  *H e x p e r i m e n t s ,  most  complex  the  to  vicinal an  scheme  applicable  that  reported  J-resolved  repetition  the  many  it  resolve  2D  with  make  signal-to-noise  technique,  this  data w i l l  obvious  and  considerably careful  into  that  however,  reliably  of  integration  fact  have  to  protocol  Both O-raethylation  demanding  high  -  this  the  important  protocol  COSY  need  demand  technique  "difference"  the  c a n be  the  experiments  unusually  necessary  excessive.  to  We f e e l  Nature  immediately  reactions.  experiment these  be  could  sweep-widths  in  Is  conformations  restricting  not  efficient easy  would  It  Although  application  on r e f e r e n c e  of  in  their  find  interest.  array  exist  discussion.  n e w NMR e x p e r i m e n t s ,  approach  of  warrant  The  of nOe  SSNOEDS power  and  it  a  is  stability 2 days,  if  are one  were  -  working  with  addition  to  Tjnix.  As  on  rate  the  Although  extremely  signal-to-noise  mentioned at  not  which  venturing  strong  nOe  methylene  since  the  of  respect,  shown)  There means  to  "external"  did  is  the  arises  dispersed  resonances  structure, desirable have  to  that  the  Tj  or  secondary  be  obtained  back  In  this,  of  to  in  instead  the  for  our  experiments experiments and  a the  quite  efficiently  NOESY at  appeared  the  nOe  of  for  the  of  In  the  "derivatized"  material  derivatization  event  introduces  H-l".  or  sharp, on  not  well  the such it  about  is  data fact  the  uncertainties. a  to  primary  that  molecule,  s t i l l  of  chemical  yield  Information  is  shift  ideal  gives  to  in  whether An  to  fail  and  using  information  potential  to  H-l"  factors. itself  expected,  chemical  into  question  conformational  unprotected  case  the  the  another  information.  from  any  to  encountered,  be  molecule  addition  occurs.  between H-l  literature The  build-up  might  small  number  have  in  depends  x  what  taken a  m  in  time,  t±  ID nOe  one  with  one  the  "natural"  structure  of  for  on a  experiments  the  relax  dispersion.  since,  the  accord  In  is  has,  mixing  of  nOe  exploratory  contributions.  rests  the  extrapolation  derivatized spite  when  nOe  will  and  a weaker  NOESY  the  choice  We o b s e r v e d  but  in  the  -  new m o l e c u l e  some  evidence  spectral  situation  is  precedent  molecule  a  cross-section  show  ample  improve  derivatize  protons  variable  tumbling  2D f i e l d .  This  a  is  time  and H-4,  although  a  quantities.  II.2.5.6,  perform  the  pair.  methylene  exclusion  (not  every  between H - l '  allyl  H-l  into  demands,  molecule  to  -  (microgram)  Section  the  may w i s h  before  this  in  necessary  experimenter  the  small  225  strong  one  - 226 -  because, In addition to the increased spectral dispersion, d e r i v a t i z a t i o n of amphipathic molecules can often reduce the tendency of the unprotected  molecules to associate i n polar solvents to form  micelles, thereby decreasing the e f f e c t i v e molecular  weight and  narrowing the l i n e s observed (lengthening T ) . ?  In the case of oligosaccharides, i t i s d i f f i c u l t to decide whether the O-acetylated study.  or C^-methylated derivative should be chosen for  I f mass spectrometrlc analyses are planned, the 0_-methyl  derivative w i l l probably have to be made i n any case.  However, i f the  parent material bears 0_-acetate groups, d i f f i c u l t i e s with 0_-acetyl migration may arise during the d e r i v a t i z a t i o n procedure. since 0_-methyl groups resonate  Furthermore,  i n the region of Interest, the deuterated  methyl iodide used f o r the d e r i v a t i z a t i o n should be of the highest lsotopic p u r i t y obtainable.  Experience i n this laboratory has shown 33  that f u l l (>-acetylation of polysaccharides  i s n o n - t r i v i a l and demands 34  that conditions be optimised  f o r each substrate.  O-methylatlon of polysaccharides may be a more this i s not substantiated with experimental  On this basis, general procedure,  35  but  evidence.  Since the completion of t h i s study, other groups have found u t i l i t y f o r 2D NMR i n s t r u c t u r a l studies on complex oligosaccharides. More recently two dimensional J_-resolved spectroscopy  and SECSY have  been integrated into Dabrowski's studies (vide supra) to allow the 36  extension of the work to a ceraraide pentadecasaccharide. Conceptually,  resonance frequencies are determined from the 2D  J-resolved experiment, and Jj-connectivlties established by taking  -  sections to  the  no  means  through  utility a  with  broadness  of in  An 2D  of  some  a l l  signal  convictions. of  the  shift  -  -  Whilst he  overlap  (short  by  Prestegard  a n d nOe  correlated  they  problems...". we  draw  attention are  This  attention we k n o w  2  experiments  draws  cautions  T_ ) w h i c h  2D  study  Dabrowski  also  Finally,  signals  echo-type  Independent  chemical  data-set.  experiments,  2D  for  our  spin  SECSY  the  panacea  agreement  problems  the  227  "...by  is  to  to  in the  present  II.5).  (see  et_ a l .  found  experiments  utility  for  and  (SECSY  the in  NOESY)  37 the is  elucidation known  that  of  primary  cellulose  structure  can  be  of  "fully"  a  glycolipid  O-acetylated  at and  MHz.  500  yield  a  It spectrum  38 in  CDCI3 w i t h  discernable  vicinal  couplings.  Gagnaire's  group  in  39 France  has  reported  derivatized due  to  the  polysaccharides. rapid  suggest  that  partial  or  such  could  2  2D  effectively the  This  relaxation  spectroscopy experiment  during  NMR a n a l y s e s  narrow  line-widths,  of  the the  intact  not  easy  refocussing  either  procedures.  disregard  analysis  the  is  on  on p o l y s a c c h a r i d e s  depolymerization k1  degradative  adequately  complicate  T_  complete  bacteriophage yields  J-resolved  2D  this  terminal single  If  by  mild  partial would  be  reducing  repeating  and to  perform,  period. may  acid  be  We  improved  hydrolysis  by or  depolymerization preferable  sugar  unit  since  which  oligosaccharide  one  would  unit.  42 Bruch J-resolved  and  Bruch  spectroscopy  responded  have to  also  reported  glycopeptides;  the  application  several  broad  of  anomeric  p o o r l y to the a n a l y s i s . P a t t has d e s c r i b e d the use of 43 13 44 INADEQUATE to assign the C spectra of a monosaccharide and disaccharide, * b u t t h e p u n i t i v e demands on sample c o n c e n t r a t i o n 1  5  2D signals 2D a make  - 228 -  the  experiment  of  little  more  than  academic  interest  in  the  present  context. One that  we  believe  consisting bond one  is can  final  of  point it  units  measurable speculate  oligopeptides.  to  be  of at  on  on  the  generality  extendable  coupled 500 MHz * 1  the  to  spins. 6  it  is  application  of  sequencing  almost  any  Whilst  coupling  s t i l l of  our  such  small  complex  0.5  a  technology  to  is  molecule  through  ( J =ca. 5  protocol  a Hz)  peptide and  - 229 -  REFERENCES 1.  N. Sharon. Complex carbohydrates - their chemistry, biosynthesis and functions. Addison-Wesley Reading, Mass. 1975.  2.  J . Montreuil. Adv. Carb. Chem. Biochem. 37, 157-223 (1980), and references cited therein.  3.  G.O. Aspinall and A.M. Stephen. In MTP International reviews of science. Organic chemistry, series one. Edited by G.O. A s p i n a l l . University Park Press, Baltimore, Maryland. 1973. V o l . 7, pp. 285-317.  4.  G.G.S. Dutton and E.H. M e r r i f i e l d . (1982).  5.  P.A.J. Gorin.  6.  R. Barker, H.A. Nunez, P. Rosevear, and A.S. Serianni. Enzymol. 83, 58-69 (1982).  7.  R.V. Lemieux, K. Bock, L.T.J. Delbaere, S. Koto, and V.S. Rao. Can. J . Chem. 58, 631-653 (1980).  8.  L.O. S i l l e r u d , R.K. Yu, and D.E. Schafer. (1982).  9.  J.A. Schwarcz and A.S. P e r l i n .  Carbohydr. Res.  Adv. Carbohydr. Chem. Biochem. 38,  105, 189-203  13-104  (1981).  Methods  Biochem. 21, 1260-1271  Can. J . Chem. 50, 3667-3676 (1972).  10.  P.R. Rosevear, H.A. Nunez, and R. Barker. (1982).  Biochem. 21, 1421-1431  11.  J.F.G. Vliegenthart, H. van Halbeek, and L. Dorland. Chem. 53, 45-77 (1981), and references therein.  12.  J.P. Carver and A.A. Grey.  13.  J-R. Brisson and J.P. Carver. (1982).  14.  J . Dabrowski, P. Hanfland, and H. Egge. (1982), and references therein.  15.  J.M. Berry, L.D. H a l l , D.G. Welder, and K.F. Wong In Anomeric e f f e c t , origins and consequences. Edited by W.A. Szarek and D. Horton. ACS Symposium Series, No. 87. 1979. pp. 30-49.  16.  K.F. Wong, Ph.D. Thesis, University of B r i t i s h Columbia references therein.  17.  L.D. H a l l and CM. Preston.  Pure Appl.  Biochem. 20, 6607-6616 (1981). J . B i o l . Chem. 257, 11207-11209 Methods Enzymol. 83, 69-86  (1979), and  Carbohydr. Res. 29, 522-524 (1973).  - 230 -  18.  L.D. H a l l and H.D.W. H i l l . (1976).  19.  See section II.2.A.  20.  M. Martin-Lomas and D. Chapman. (1973) .  21.  K-E. Falk, K-A. Karlsson, and B.E. Samuelsson. Biophys. 192, 191-202 (1979).  22.  M.A. Bernstein and L.D. H a l l . (1982).  23.  V . Horejsi and J . Kocourek. (1974) ; H.M. Flowers.  J . Amer. Chem. Soc. 98, 1269-1270  Chem. Phys. L i p i d s .  10, 152-164  Arch. Biochem.  J . Amer. Chem. Soc. 104, 5553-5555 Biochim. Biophys. Acta. 336, 338-343  Methods Carbohydr. Chem. 6, 474-480 (1972).  24.  See section II.2.2.1.  25.  See section II.2.3.  26.  See section II.2.4.  27.  See section II.2.5.2.  28. 29.  See section II.2.5.3. C. Altona and C.A.G. Haasnoot. (1980) .  30.  See section II.2.5.6.  31.  M.L. Wolfrom and A. Thompson. (1973).  32.  K. Nagayama and K. Wuthrich. (1981) .  33.  J . Hoffman and B. Lindberg. 117-122 (1980).  34.  K.R. Holme, private  35.  H.E. Conrad. Methods Carbohydr. Chem. 7, 361-364 (1972), and references cited therein.  36.  J . Dabrowski and P. Hanfland.  37.  J.H. Prestegard, T.A.W. Koerner, J r . , P.C. Detnou, and R.K. Yu. J . Amer. Chem. Soc. 104, 4993-4995 (1982).  38.  H. F r i e b o l i n , G. K e i l i c h , and E. S i e f e r t . 766-767 (1969).  Org. Magn. Reson. 13, 417-429  Methods Carbohydr. Chem. 2, 211-215 Eur. J . Biochem. 114, 375-384  Methods Carbohydr. Chem. Biochem. 8,  communication,  FEBS L e t t . 142, 138-142 (1982).  Angew. Chem. Int. Ed. 8,  - 231 -  39.  D.Y. Gagnaire, F.R. Taravel, and M.R. Vignon. 126-129 (1982).  Macromol. 15,  40.  M.L. Wolfram and M.E. Franks. Methods Carbohydr. Chem. 5, 276-280 (1965), and others i n the same section.  41.  G.G.S. Dutton, J.L. DiFabio, D.M. Leek, E.H. M e r r i f i e l d , J.R. Nunn, and A.M. Stephen. Carbohydr. Res. 97, 127-138 (1981).  42.  R.C. Bruch and M.D. Bruch.  43.  A. Bax, T.A. Frenkiel, R. Freeman, and M.H. L e v i t t . Reson. 43, 478-483 (1981).  44.  S.L. Patt,  45.  S.L. Patt, F. Sauriol, and A.S. P e r l i n . (1982).  46.  H. Kessler, W. Bermel, A. F r i e d r i c h , G. Krack, and W.E. H u l l . Amer. Chem. Soc. 104, 6297-6304 (1982).  J . B i o l . Chem. 257, 3409-3414 (1982). J . Magn.  Carbohydr. Res., i n press. Carbohydr. Res. 197, C1-C4 J.  - 232 -  CHAPTER II.4 NMR SPECTROSCOPIC ASSIGNMENT OF BRUCINE Interest i n the Strychnos a l k a l o i d s , of which strychnine (1) and brucine  (2) are members, dates from their i s o l a t i o n i n 1819, and 1946  when the correct skeletal structure was f i n a l l y proposed.  Although of  "medium" molecular weight, the skeleton Is disproportionately complex, having s i x c h i r a l centres i n a 24 atom skeleton.  NMR spectroscopists  have also been Interested i n these alkaloids f o r some time, and the relevant l i t e r a t u r e i s reviewed below.  1 :R - H 2 : R - OCH  3  -  233  -  The object of the exercise detailed i n this chapter i s to determine the generality of the NMR methods presented i n Chapter I I . 2, and to determine which i s the most expeditious route to the f i n a l assignment of the H and  C spectra of complex molecules.  An important  aspect of this study was an evaluation of the time required to perform the experiments, and optimum ways to integrate data-acquisition and processing.  I t should be stressed that the experiments were  i n t e n t i o n a l l y carried out using the minimum of l i t e r a t u r e data; only brucine*s  skeletal structure and configuration, and an early 13  (erroneous) table of reference data.  thus,  1  C chemical-shift assignments  was used as  No data that could lead to subtle chemical s h i f t  arguments were sought, since the Intention was to keep the reliance on such tacks to a minimum.  Wherever possible, spectra were acquired under  "survey conditions", that i s , using the f u l l e s t possible sweep-widths In both dimensions f o r the 2D experiments, with " t y p i c a l " delays.  In t h i s  way, an attempt was made to simulate conditions which an "applied" spectroscopist might encounter when analyzing the structure of a molecule for which l i t t l e or no prior spectroscopic knowledge was available.  The questions are:  which experiments should be performed,  In what order and how much instrument time w i l l they demand? The considerable body of c l a s s i c a l chemistry performed on the Strychnos alkaloids has been reviewed. degradative  After uncountable man-hours of  analyses by numerous groups the correct s k e l e t a l structure  was proposed i n 1946 by R. Robinson and colleagues.  3  The absolute  configuration at the asymmetric centres was determined by X-ray  - 234 -  d i f f r a c t i o n analysis,  and Woodward succeeded  i n synthesizing  strychnine i n 1954. The i n i t i a l NMR spectroscopic interest focussed on the C 1 3  spectrum, probably because of the complexity of the H spectrum at the 1  field  strength of the then common magnets (< 100 MHz f o r H ) . Since the 1  6  7  i n i t i a l report by Wehrli, • some considerable controversy has 13  occurred i n the l i t e r a t u r e over the  C assignments, which has only  recently reached f u l l consistency. In  1973, Wehrli described  quaternary carbons i n brucine. a d d i t i v i t y rules and T  A  6  the assignment of the seven  This was based on chemical s h i f t  relaxation data.  In that the relaxation pathway  i s dipolar, quaternary carbons are relaxed by protons on a carbon atoms.  The Tj's vary from 2.62 to 11.75 sec.  Shortly thereafter, /  13  Wehrli published the complete  C assignment of brucine.  Again, much  13  attention was focussed on  C T j relaxation rates.  Single frequency  off-resonance decoupling (SFORD) was used to determine  1 3  C  m u l t i p l i c i t i e s and further assignments were based on chemical s h i f t and residual s p l i t t i n g ( XH) arguments, and comparisons with related molecules. This assignment was correct, but numerous "reassignments" 1J  1  were made over the following years. et  Srlnivasan and L i c h t e r ,  8  and Singh  a l . , followed with some erroneous "reassignments,"' l a t e r summarized  and c l a r i f i e d by Verpoorte and coworkers. 9  10  Leung and Jones r e p o r t e d  on the C NMR spectrum of strychnine, which needed only one 1 3  reassignment.  Wenkert et a l . , r e p o r t e d  12  the CNMR (and some % 13  NMR)  assignments of a number of related alkaloids of this genre, which were  11  - 235 -  i n agreement with Verpoorte.  The above workers r e l i e d on  SFORD  to  provide the information on the number of attached protons, various d e r i v a t i v e s , chemical s h i f t arguments, and (reduced) ^^JJCH  V A  their  *  U E 8  f°  T  assignments. Many of the discrepancies between the various assignments can be  traced to the data from the  SFORD experiment.  In the case of an {AB}X  experiment with a methylene carbon, complexities may arise and this i s well documented by Roth and Bauer,  13  using strychnine as an exemplar.  Depending on the decoupler position and power, C-15 i s seen to show 3 or 4 l i n e s (6 t r a n s i t i o n s , maximum). Careful analysis can y i e l d 1  J  and J^B. 2  £X  This was extended i n Roth's description ** of 1  (noise-modulated off-resonance  published the  Martin suggested ideas.  D i f f i c u l t i e s can arise i n the  SFORD spectrum  SFORD spectrum 15  NORD  decoupling) which i s similar i n concept  and potential information-yield. i n t e r p r e t a t i o n of the  J^X»  1  where l i n e s are crowded; Wehrli has  of b r u c i n e , and the complexity Is evident. 7  an extension of Wehrli*s T j relaxation  In that the relaxation mechanism i s predominantly dipole-dipole  (even at 1.3 M) and mediated by the d i r e c t l y attached protons, carbons bearing two protons w i l l relax at about twice the rate of those bearing one.  Quaternary carbons, as mentioned, relax at a rate proportional to  the number of a protons.  Methyl carbons do not always conform  because  they can have additional i n t e r n a l reorientation, but s t i l l relax much faster than methine or methylene carbons. those presented i n II.2.2.1.  Additionally,  These concepts are similar to  SFORD experiments  were  - 236 -  performed around way,  with  the  sometimes  Martin  carbon  in  selective  was  the  excitation  {AB}X  misleading  able  to  correctly  molecule,  and  of  the  C signals.  system  determine  assign  discussed the  them u s i n g  This  is  above.  In  multiplicity  the  tactics  one  of  way  this  each  outlined  previously. The Carter  et  250 MHz  al.  in  1974.  prepared  and  spectral  simulation  assignment.  *H NMR s p e c t r u m  this,  in  1 6  strychnine  was  Specifically  deuterated  analogues  conjunction  and  Wenkert  of  solvent  et^ a l .  with  fully  double-resonance  induced  shifts,  substantiated  the  assigned  by  were  experiments,  resulted  in  assignments  the  with  total some  12 small  modifications  geminal  protons  conformational were  not  made  these  C-15,  with  assignments study  were  reversal C-18  arguments  compared  relaxation brucine  at  -  by  and  using  based  assignment  C-20.*  coupling  solid-state on  Colebrook  assigned.  of  work  constants;  solvent  a  Their  conformational  and H a l l ,  Finally,  within  induced 17  the  *H n O e  each was  data.  on  the  results  Carter  shifts.  by  of  based  however  aromatic  study  pair  In  a  protons  Bothner-By  et^ a l . , proton of  and  18 co-workers  used  compound  the  for  Before number  of  literature  brucine driven  nOe  presenting  points  should  detailed  dissolved  be  above  in  experiment  the  viscous  (See  spectroscopic  reiterated. was  a  The  deliberately  solvent  as  a  model  II.2.4.2). studies  of  majority consulted  of  this the  post  thesis,  a  considerable  facto,  and  21 * A r e c e n t s t u d y by C o l e b r o o k and c o - w o r k e r s confirms r e a s s i g n m e n t , and c o r r e c t s t h a t of H - 1 7 a and 8. NOEDS a n d 7\ measurements, a l o n g w i t h s p e c t r a l s i m u l a t i o n were used.  this  in  - 237 -  our reasoning, a minimal reliance was made on chemical s h i f t arguments. As the emphasis was on the spectroscopic aspects, access to derivatized materials and data on analogues were also excluded. The o v e r a l l strategy chosen was f i r s t l y to f u l l y assign the proton NMR spectrum and then use this to assign the non-quaternary signals i n the carbon-13 spectrum.  The former was achieved primarily  using COSY (II.2.5.3.2), with some assistance from 2D J-resolved spectroscopy (II.2.5.2), and 1- and 2-D nOe experiments.  With this  completed, the C NMR spectrum was easily assigned using the CSCM 1 3  experiment (II.2.5.4), with spectral editing performed using DEPT (II.2.2.2).  Fig.  II.4.1  400 MHz NMR spectrum of brucine (ca. 0.02 M i n CDC1 ), showing the assignment of non-aromatic protons. 3  - 238 -  Portions of the F i g . II.4.1.  Two  H  NMR  spectrum of brucine are presented in  singlets in the aromatic region are omitted, and the  multiplet i n the v i n y l region i s shown i n an inset. began with the COSY spectrum. coherence  The experiment  Analysis at 360  was performed  MHz  with  transfer echo selection, on a ca. 0.01 M solution i n CDC1 . 3  A  relaxation delay of only 0.2s, and the minimum of 4 acquisitions per t ^ i n t e r v a l meant that 20 min was required for the c o l l e c t i o n of a 512 * IK data-set; the sweep-width was J-resolved experiment  10 ppm i n both dimensions.  was set up; this experiment  for c o l l e c t i o n of the data - this was  Next, a 2D  required ca. 45 min  somewhat longer than that of the  COSY due to the necessarily longer relaxation delay and more scans. Whilst these J_-resolved data were being c o l l e c t e d , the COSY data-set was processed and plotted.  At t h i s stage, a c q u i s i t i o n of the 2D J-resolved  data set was complete, and data processing was performed. operations were performed Nlcolet system.*  A l l the above  automatically i n a MACRO program on the  Minimal operator intervention was required, except at  the p l o t t i n g stage, where decisions were made to expand certain spectral regions.  Details of the "survey-condition" parameters are i n Table  II.4.1. The COSY spectrum of brucine Is shown i n F i g . II.4.2.  Only the  region between 64.5 and 1.0 i s plotted here, f o r the sake of c l a r i t y . The analysis began with the r e a l i z a t i o n that four "pockets" of coupled spins exist i n the molecule, with l i t t l e or no coupling between them  *A11 360 MHz experiments were performed Corp., Fremont, Ca. USA.  at Nicolet  Magnetics  -  Fig.  II.4.2  239  -  3 6 0 MHz C O S Y s p e c t r u m o f b r u c i n e i n t h e r e g i o n b e t w e e n 6 4.5 a n d 1.0. C o n n e c t i v i t i e s to the v i n y l i c proton are i n d i c a t e d , though they are not on the d i a g r a m .  Table II.4.1.  Parameters used i n automated "survey conditions" c o l l e c t i o n and processing of H COSY and 2D J_-resolved experiments at 360 MHz. F u l l sweep-width (Hz). ^ B i o c ^ - g f (words). R e l a x a t i o n delay (s). Apodization function: EM = exponential m u l t i p l i e r ; DM = double exponential multiplier; MS » sinebell m u l t i p l i e r . With the J-resolved experiment, the strongest resolution enhancement function i s chosen for t_2 which the signal-to-noise permits. Symmetrization was not possible with the 2D J-resolved data, but would be recommended. a  z e s  d  e  Sweep Width  Total Time (min)  r  F  2  Fi  3  Block S i z e F_  2  £i  b  R  J Apodization Scans D e l a y Per _t i t_i tj, e  l  a  0  x  ,  d  0  -  Calibrate pulse widths  0  Begin collecting COSY data  3,600  3,600  1,024  512  0.2  4  20  Begin collecting 2D  3,600  29  4,096  64  1.0  16  120  J ° 7 Symm. fill tj?  e  16K  Collect ID spectrum  65  Z e  Process and plot COSY  MS  MS  No  Yes  Commence processing and plotting of 2D J-resolved experiment  EM DM MS  DM MS  Yes  No  END  - 2A1 -  ( c f . II.3):namely  (a) H-20o and H-20B, (b) H-17a, H-17B, H-18a and  H-18B, (c) H-22, H-23o and H-238, and (d) H - l l a , H-11B, H-12, H-13, H-8, H-14, H-15o, H-15B and H-16.  COSY succeeded i n assigning a l l 18 of  these protons, although i t was not possible at this stage to d i f f e r e n t i a t e a from 8 geminal protons.* indicated along with the assignments without ambiguity.  The connectivities are  i n the Figure; these were made  The large -OCH peaks which resonate very close to 3  H-8 and H-16 could have hindered the assignment  had H-8 and H-16 been  scalar coupled as they might then have obscured a set of cross-peaks; this was not the case.  Special considerations were required for the  t i g h t l y coupled geminal protons, which were assigned to the H-23 and H-17 pairs.  Distortion of peak i n t e n s i t i e s indicated tight coupling,  e s p e c i a l l y i n the case of the H-17 p a i r .  To assess the r e l i a b i l i t y of  the COSY spectrum, i t i s i n t e r e s t i n g to consider H-15B (ca. 6 1.4), which shows one large s p l i t t i n g and several smaller (<_ 1.5 Hz) ones. In the COSY spectrum, the expected three connectivities are clear, even though the d i g i t i z a t i o n i s less than 7 Hz per point. The 2D J_-resolved experiment data are not included here. the -OCH3 signal region. remained singlets.  did not prove very useful and these  I t would have been p a r t i c u l a r l y useful i n  However, the overlapping H-16 resonance  obscured by the f a m i l i a r " t a i l i n g " i n F j from the intense -OCH  3  The p a r t i a l J_-spectrum of the hidden H-l8 resonance was not  *The molecule i s not planar and hence, an upper (p) and lower (a) face i s not c l e a r . For t h i s reason, geminal protons are defined (See 1 and 2) for the purpose of t h i s study.  - 242 -  u s e f u l , since i t suffered from the same v i r t u a l coupling effects as the other H-18 resonance at 6 2.83, making the extraction of accurate coupling constant data impossible without spectral simulation. (The geminal H-17 protons, to which they are v i c i n a l , are t i g h t l y coupled). The assignment now sought.  of the aromatic, geminal and ^-methyl protons was  The solid-state conformation as determined by X-ray  crystallography was used to build a molecular model having the correct configuration.  This was found to be a r i g i d structure, suggesting that  the solution conformation might be very similar to that i n the solid-state.  The experiments  required to complete  these *H assignments  also served to corroborate this postulated solution conformation. In the determination of conformation, techniques relying on "through-space"  phenomena (e.g. T_i, nOe) prove very useful, often i n  conjunction with (through-bond) coupling-constant information. protons take on unusual importance i n an nOe experiment.  Axial  The 1,3  d i a x i a l i n t e r a c t i o n provides important information i n a six-membered ring, and with a boat conformation, the "flagpole" protons are close to one another i n spite of their being 5 bonds removed; an nOe experiment w i l l also detect t h i s . The SSNOEDS results are presented In F i g . II.4.3. e f f o r t was made to optimise the experiment In approximately three hours.  No special  and the data were collected  A single decoupler power setting was used  (which resulted i n 40-80% saturation, depending on the t o t a l multiplet width).  In a l l cases a check was made for decoupler power leakage into  - 243  F i g . II.4.3.  -  SSNOEDS experiments performed on brucine at 400  MHz.  - 244 -  neighbouring resonances ( " s p i l l o v e r " ) , and was found to be absent. i r r a d i a t i o n time was  set at 0.5 s*, and the experiments  The  "interleaved" by  c o l l e c t i n g scans i n blocks of 16, repeated 30 times. Let us consider ring IV, which Is i n a chair conformation in the s o l i d - s t a t e ( F i g . II.4.4.A).  Evidence for this also being so i n  solution comes from a consideration of coupling-constants, and experiments.  nOe  J_ (8, 13) = 10.5 Hz, which indicates a trans or c i s  arrangement of the two protons.  In keeping with the chair conformation,  H-l4 and H-l6 are judged to be roughly co-planar since J (16, 15a), and J (14, 158) « J (16, 156).**  (14, 15a) «  Strong confirmatory  evidence for the proposed conformation comes from the SSNOEDS experiment where H-15  (6 1.46) was i r r a d i a t e d , inducing an nOe i n H-13;  the two  protons must assume a x i a l orientations In a ring having a chair conformation, and the irradiated proton i s accordingly assigned to H-l5p.  I r r a d i a t i o n of the h i g h e r - f i e l d of the two H-l8 protons induced  an nOe i n H-8, H-8  thereby i d e n t i f y i n g the proton i r r a d i a t e d as H-18B  and  to be a x i a l , as depicted i n F i g . II.4.4A. Consider now ring VI.  Since the H-20  protons have no v i c i n a l  neighbours, no conformational information can be sought from arguments based on v i c i n a l coupling-constnts. The nOe experiment where H-15a  was  *This i s far less than the time usually used for nOe build-up (ca. 3* Tj)» and was chosen to minimize the experimental acquisition time and "three spin e f f e c t s " , without jeopardizing the observation of " d i r e c t " interactions. **The H-15 protons are assigned here to allow the reader to check the statement with the reported data i n Table II.4.2, but the absolute assignment of the H-15 protons Is not necessary at this stage.  - 245 -  A.  B.  C.  D.  o  F i g . II.4.4  Conformations of rings IV and V (A), VI (B), VII (C) and I I I (D) i n brucine, indicating some of the results from the SSNOEDS experiments. The i r r a d i a t e d proton i s indicated with (•) and nOe's ( ).  - 246 -  i r r a d i a t e d induced a strong nOe i n the lower-field resonances.  This confirms the approximate  of the two  H-20  "boat" conformation of t h i s  r i n g ( F i g . II.4.4.B), which places the "flagpole"  protons, H-15o  and  H-20a, close to one another; t h i s i s a "bowsprit-flagpole" i n t e r a c t i o n . Hence, t h i s single experiment  confirms the expected conformation of ring  VI, and i d e n t i f i e s H-20oc (and H-20B, by The analysis  default).  of ring VII was complicated somewhat by the fact  that i t i s a seven-membered ring, with a double bond and ring hetero-atom.*  Two nOe experiments were performed  structure shown i n F i g . II.4.4C.  to support the  Irradiation of the v i n y l i c  induced an nOe into the pseudo-equatorial proton H-238, the of the two H-23  protons.  Additionally,  based on the i r r a d i a t i o n of H-l5a.  of the H-23  protons i s now  The conformational analysis  and H-20B are close to one  of the geminal H-20  H-14  conformation.  and H-23a.**  The  confirmed.  of ring III i s more complex.  the H - l l a and H-lIB protons must be assigned. favours a boat-like  protons  I r r a d i a t i o n of H-l2 induced an nOe  into other a x i a l or a x i a l - l i k e protons: assignment  lower-field  an nOe i s induced i n H-208.  Although not depicted i n t h i s diagram, H-22 another and this confirms the assignment  H-22  Here  The molecular model  A pure boat conformation would make  H-l2 and H-lip trans d i a x i a l , and one would anticipate  a large v i c i n a l  coupling constant here on the basis of the Karplus relationship;  *This i s a tetrahydrooxepin r i n g . **Here, a negative nOe i s Induced i n H-23 "three-spin e f f e c t " .  p due to the  using  - 247 -  t h i s argument, the lower f i e l d of the two H - l l protons would be assigned to H-11B.  The geminal coupling constant J_ (11a, 116) i s , at 17.4 Hz,  unusually large.  This suggests that the carbonyl on C-10 roughly 19  bisects the angle between the geminal protons and C - l l ,  and this would  not be the case i f ring III were to be i n the pure boat conformation considered above.  Satisfying the condition of the relationship between  the carbonyl and geminal protons " f l a t t e n s " one end of the boat, placing N-9, C-10, C - l l , C-12 and C-13 approximately co-planar.  In this  s i t u a t i o n H-l2 and H - l l a have a small dihedral angle (ca. 10°), and, hence, a large v i c i n a l coupling constant would be anticipated.  This  argument assigns the lower f i e l d of the two H - l l protons as H-l16 - a reversal of assignment. are  I t i s clear that i n both cases H-8 and H-l16  close i n space, and an nOe experiment would not suffer from these  ambiguities.  The H - l l protons could be assigned, and on this basis the  ring conformation determined from the coupling-constant arguments. The SSNOEDS experiment  (vide supra) was not possible here, since  a l l three protons - H-8, H - l l a and H-lIB - overlap with other resonances, precluding i r r a d i a t i o n s e l e c t i v i t y .  A 2D NOESY experiment  was performed, and the data are given i n Figs. II.4.5 and .6.*  Fig.  II.4.5 shows the NOESY spectrum over the entire spectral width, and Fig.II.4.6 shows the h i g h - f i e l d portion of this data i n more d e t a i l ; we concentrate on the l a t t e r , for the moment.  Before searching for the  *Here, the p r i n c i p a l diagonal runs from the top left-hand corner to the bottom right; a mirror image of the displays previously shown, but e n t i r e l y analogous.  F i g . II.4.5  400 MHz NOESY spectrum of brucine. The 256 * IK data-matrix was collected i n ca. 8 h, with relaxation delay 3.0 s, mixing time 0.6 s ± 10% and 32 scans per t_i i n t e r v a l . The data were processed and plotted using a Nicolet 1280 computer and Zeta 8 recorder. The f i n a l d i g i t i z a t i o n was 5.8 Hz pt i n both dimensions, after z e r o - f i l l i n g Jj; the " s i n e b e l l " m u l t i p l i e r was used i n both dimensions, and magnitude spectra calculated.  - 249 -  Fig.  II.4.6  The h i g h - f i e l d region of the NOESY plot In F i g . II.4.5  - 250 -  8-1 i p nOe, i t i s worth checking the experiment by looking for the nOe's determined previously using SSNOEDS.  These are a l l c l e a r l y discernable  and are indicated i n the upper half of F i g . II.4.6. with H-16  A number of nOe's  are clear here which were not i n the ID experiments, as a  result of being obscured by subtraction errors with the overlapping sharp 0-CH_3 s i n g l e t s . The 8-1lp nOe correlation i s discernable, assigning H-lip as the higher f i e l d of the H - l l resonances.  This i s i n accord with the second  argument above using v i c i n a l and geminal J_-coupling magnitudes and suggests a boat-like conformation, flattened at C - l l II.4.4.D). H-12  (See F i g .  This i s confirmed by the observation of a larger nOe between  and H - l l o than H-12  and H - l l p .  The NOESY spectrum also provides the means for the assignment of the aromatic and O-methyl signals.  The relaxation of H-4  must be  dominated by the 0-CH protons on C-3, and H-l by those on C-2. 3  additional relaxation contributions from H-16  and H-17a.  H-l has  F i g . II.4.5  shows the nOe between the aromatic and the O-methyl protons. Correlations between an aromatic proton and protons on ring V are below the contour threshold; s l i c e s taken at each aromatic proton's chemical s h i f t (not shown) i d e n t i f y the higher f i e l d of the two aromatic resonances as that of H-l, by the above reasoning.  The nOe with H-17 i s  not c l e a r , as i t overlaps with the strong correlation to the O-methyl resonance.  Further, the O-methyl resonances are e a s i l y assigned; the  higher f i e l d of the two singlets corresponds to the O-methyl on C-2 the lower f i e l d to that on C-3.  and  -  Table  II.4.2  Summarized  251  H NMR  chemical  6  Proton  shift  (ppm)  J  -  4  7.80  -  8  3.83  (8-18)  11a  3.09  (11a,  118)  118  2.64  (lip,  12)  12  4.27  (12,13)  13  1.26  (13,  14)  14  3.14  (14,  15a)  15a  2.34  (15a,  15p)  156  1.46  (15p,  16)  16  3.9 (17a,  17p)  -  (20a,  20p)  = 14.8  = 13.8  18a  18B  1.9,  a  b  3.15  b  2.83  1.8  =  =  23a  4.05  (23a,  23p)  238  4.13  (23p,  22)  (C-3)-OCH  3  3.89  (11a,  12)  =  8.5  3.3  •= 3 . 2  2.70  -  17.4  3.3  20 8  3.85  -  3.68  3  by  10.5  20a  (C-2)-OCH  M).  (Hz)  6.66  p  and  coupling.  1  17a,  assignments  of brucine i n CDCI3 (ca. 0.02 b Coupling pattern complicated  coupling constants a strong coupling. virtual  -  -  4.0 = 14.2  -  =  (14,  15  (15a,  p)  16)  = -  1.0 ca.  4.9  2.1  ca.  6.9  12.0  (20a,  (23a,  22)  22)  =  1.6  -  6.3  - 252 -  The f i n a l assignments of the H NMR spectrum of brucine are compiled i n Table II.4.2.  These data are i n complete accord with those 12  of Wenkert e>t a l . , for strychnine. We have seen that the COSY experiment was s u f f i c i e n t to perform the bulk of the analysis of the *H NMR spectrum of brucine. The ID and 2D nOe experiments assigned individual geminal protons and provided valuable conformational information. 13 Now, the  C spectrum was analysed. A ca. 0.3 M solution of  brucine i n CDCl3:CD OD (10:1) was prepared. 3  F i r s t , the number of  protons attached to each carbon was determined using DEPT.  This might  have been redundant because a CSCM experiment was planned; however i n this case not a l l methylene protons had very d i f f e r e n t chemical s h i f t s (vide i n f r a ) . collection.  The CSCM experiment required 45-60 min. f o r data The CSCM spectrum of the i n t e r e s t i n g regions of 6 c 22-80  and 6H 1.0-4.5 i s shown i n F i g . II.4.7. region are unambiguous).  ( A l l assignments outside this  The DEPT information i s included.  Knowing the  proton assignments, the indicated carbon assignments were made. two cases was there any uncertainty.  In only  H-8 and H-16 resonate very close  to one another making the assignment of the carbon signals around 6 59 difficult.  The carbon signals around 6 41 are very close and  i n s u f f i c i e n t l y d i g i t i z e d i n this case.  Hence, C - l l and C-16 cannot be  distinguished. No attempt was made to assign quaternary carbons since this would not f a l l within the scope of the present design protocol.  The  data from the CSCM i s compared with that of Srinivasan and L i c h t e r  1  -  - 253 -  -OCH.  «5H  f PROJECTION 2  C-8 C-16 C-12 \  DEPT-CH  C-18  C-11 C-17  C-20 C-13  C-23 A  A/A.  iV  CH , CH CH, CH, CH CH.  E C-14 1 C-15 -j\ A — CH  £  CH„  s 8  ^  <  5  l  o CN  CN  O  CO  CO  8  o I  80  I  I—I—I—I—I—I—I—I—I—I  70 60  -i—i—I—r  50  40  I  I—i—r  30 ppm 5  Fig.  II.4.7  C  360 MHz ( H) CSCM experiment with brucine (ca. 0.3 M i n C6D /CD OD, 10:1). 6 22-80 and 6 1.0-775 i s plotted i n the contour mode. Fj and F2 projections are on the abcissa and ordinate, respectively. i  6  3  C  H  - 254 -  Table II.4.3  Table of assigned C chemical s h i f t s for brucine, compared with the l i t e r a t u r e . Chemical s h i f t s are referenced to the central t r a n s i t i o n of CDCI3 (6 77.0). The solution was ca. 0.3 M i n CDCl3:CD30D (10:1) and the instrument a Bruker WH-400, operating at 100.6 MHz for C . ' may be interchanged. 1 3  a  b  Carbon  Ref. 1  Ref. 10  This Study  1  105.78  105.7  106.4  4  101.00  101.1  101.5  8  59.84  60.3  60.2  a  11  50.09  42.3  42.2  b  12  77.64  77.8  77.5  13  48.18  48.3  48.1  14  31.50  31.5  31.4  15  26.73  26.8  26.5  16  60.28  59.9  59.6  17  42.28  42.3  41.9  18  52.57  50.1  49.9  20  42.28  52.7  52.4  22  127.14  127.2  127.6  23  64.47  64.6  64.4  56.43; 56.08  56.4; 56.2  56.6; 56.0  OCH3  a  b  - 255 -  that referred to whilst the work was i n progress - and the now accepted a  assignments of Verpoorte.  Our data are i n complete agreement with the  latter. Several conclusions can be drawn from this study.  F i r s t , only  modest amounts of spectrometer time are necessary f o r the p a r t i a l assignment of the *H NMR spectrum of a medium sized molecule such as brucine; i n this case a t o t a l of ca. 2 hours for the 2D J-resolved and COSY experiments and an additional 3-8 hours for the complete spectroscopic and conformational assignment.  Additional conformational  information, plus the assignment of Individual geminal protons results from nOe experiments.  Where the selective i r r a d i a t i o n experiments can  be performed, the SSNOEDS experiment i s probably the most expeditious route.  In the case of signal overlap, the 2D NOESY experiment w i l l  provide a l l the necessary Information i n a single experiment, but with greater demands on spectrometer time.  This investment of time i s offset  by the fact that a l l possible nOe experiments are "simultaneously" performed, resulting i n a high return being realized i n terms of information y i e l d . With the proton spectrum assigned at least to the point before 13 nOe experiments are performed, non-quaternary sites i n the  C NMR  spectrum can be routinely assigned providing their resonances are not too close (ca. > lppm) and the attached proton resonances are not badly overlapping. We believe that the 2D J_-resolved and COSY experimental data can be collected and processed In a highly e f f i c i e n t , integrated fashion,  - 256 -  with the prior knowledge of the chemical system.  NOe data might require  some educated guess-work i n the selection of an i r r a d i a t i o n time (SSNOEDS experiment) or the mixing time (NOESY), or a series of exploratory experiments could be performed.  A l l the experiments  described In this chapter provided the required information from the data collected the f i r s t time i t was performed. Hurd  20  has independently reported on the u t i l i t y of the  combination of COSY with the CSCM experiment. a X  1 3  I t i s advantageous to use  C probe to measure both the C and *H spectra, the l a t t e r v i a the 1 3  H decoupler c o i l .  In this way the COSY assignment (and NOESY, i f  required) i s f o r the molecule at the same concentration as f o r the CSCM experiment and no time i s expended changing probes. the  This approach has  further merit that i f i t were possible to automate changeover of the  probe c i r c u i t s , i t would be possible to automate the acquisition and processing of a l l *H and C data into a simple (overnight) run. 1 3  i t s high e f f i c i e n c y , i t seems probable that the mode of analysis described i n this chapter w i l l f i n d a place i n organic s t r u c t u r a l investigations which warrant more than a few, unrelated NMR measurements.  Given  - 257 -  REFERENCES 1.  P.R. Srinlvasan and R.L. Lichter. (1976).  Org. Magn. Reson. 8, 198-201  2.  G.F. Smith. The alkaloids. Edited by R.H.F. Manske. Press, New York, Vol. 8, 1965, pp. 591-671.  3.  L.H. Briggs, H.T. Openshaw, and R. Robinson. 903-908 (1946), and others i n the series.  4.  A.F. Peerdeman.  5.  Ref. 2, pp. 642-647, and references therein.  6.  F.W.  Wehrli.  J . Chem. Soc. Chem. Commun.  7.  F.W.  Wehrli.  Adv. Mol. Relaxation Processes.  8.  S.P. Singh, V.I. Stenberg, S.S. Parmar, and S.A. Farnum. Pharm. S c i . 68, 89-92 (1979).  9.  R. Verpoorte.  Academic  J . Chem. Soc.  Acta Cryst. 9, 842 (1956).  379-380 (1973). 6, 139-151 (1974). J.  J . Pharm. S c i . 69, 865-867 (1980).  10.  R. Verpoorte, P.J. Hylands, and N.G. 567-571 (1977).  Bisset.  11.  J . Leung and A.J. Jones.  12.  E. Wenkert, H.T.A. Cheung, H.E. Gottieb, M.C. Koch, A. Rabaron, and M.M. Plat. J . Org. Chem. 43, 1099-1105 (1978).  13.  K. Roth and H. Bauer.  14.  K. Roth.  15.  G.E. Martin.  16.  J.C. Carter, G.W. 122-131 (1974).  17.  L.D. Colebrook and L.D. H a l l .  18.  N.M. Szeverenyl, A.A. Bothner-By, and R. B i t t n e r . 84, 2880-2883 (1980).  19.  R.R. Fraser and B.F. Raby. (1972).  20.  R.E. Hurd. Relaxation Times, January 1982. An in-house publication of Nicolet Magnetic Corp., Fremont, C a l i f o r n i a , U.S.A.  21.  W.J. Chazin, J.T. Edward, and L.D. Colebrook. publication.  Org. Magn. Reson.  Org. Magn. Reson.  J . Magn. Reson.  Org. Magn. Reson.  9,  9, 333-337 (1977).  15, 331-332 (1981).  42, 132-140 (1981).  J . Pharm. S c i . 70, 81-84  (1981).  Luther, I I I , and T.C. Long. Can. J . Chem.  J . Magn. Reson. 15, 58, 2016-2023 (1980).  J . Amer. Chem. Soc.  J . Phys. Chem. 94, 3458-3463  Submitted for  - 258  CHAPTER  -  II.5  DIGOXIN The  studies on oligosaccharide sequencing (Ch. II.3) and  (Ch. II.4) suggested to us that the NMR  brucine  spectroscopic methods detailed  i n Ch. II.2 are e f f e c t i v e for molecules which tumble rapidly i n solution*  These encouraging results prompted us to see i f s p e c i f i c  problems might be associated with the analysis of NMR complex, higher molecular  spectra of more  weight molecules.  Our laboratory's experience  with carbohydrates and steroids led  us to the naturally occurring s t e r o i d a l glycosides. group of such plant natural products,  2  Of the  1  extensive  we chose digoxin (1) - f i r s t  isolated from the leaves of D i g i t a l i s lanata by Smith  3  i n 1930.  molecule i s of more than academic Interest due to i t s important pharmacological  place as the most frequently prescribed d i g i t a l i s  steroid for heart congestion  therapy.  1  The  - 259 -  The molecule i s highly complex.  The glycone consists of three  B-D-digitoxose units, i d e n t i c a l l y linked (1 •* 4).  From the NMR  viewpoint one could anticipate i n advance of any experiment, that this molecule poses some stringent tests.  Thus, the steroid moiety ( v i z .  digoxigenin) i s substituted by hetero-atoms  at only three positions  (C-3, C-12 and C-14), which implies that only a few steroid protons w i l l resonate outside the "methylene hump".  (It i s advantageous to have as  many steroid protons deshielded from this region as possible, since they are of tremendous assistance i n assignment).  Furthermore, digoxin i s  amphlpathic and would be expected to aggregate (form micelles) i n polar solvents such as DMSO (which i s one of the few good solvents f o r digoxin).  This would increase the e f f e c t i v e molecular weight, adding to  the already broader l i n e s resulting from the high monomer molecular mass of 780 ( C  1|1  H » 0 ). 6  t  Owing to the higher v i s c o s i t y of DMSO (cf. CDC1 ),  14  3  molecules dissolved therein are expected to display broader l i n e s , even i n the absence of molecular aggregation. A l l H NMR investigations* were performed at 500 MHz, and C at J  1 3  90 MHz (360 MHz for H ) , using NIcolet spectrometers. 1  As In previous  chapters, reliance on chemical s h i f t arguments was kept to a minimum, and l i t t l e chemical d e r i v a t i z a t i o n was performed.  Our goals were to  assign as many of the H and C resonances as possible, draw 1  1 3  conformational conclusions and, most importantly, determine the l i m i t s  *The findings of t h i s chapter resulted from a collaborative study with Dr. R.E. Hurd of NIcolet Magnetics Corp., Fremont, Ca. U.S.A.  -  260  -  of the procedures hitherto extensively r e l i e d upon.  I f an experiment  were to f a i l , could a s h i f t i n emphasis make up for this? The proton NMR  spectrum was investigated f i r s t using the  procedures found useful i n Chaps. II.3 and II.4.  As the assignments  progressed, the emphasis was allowed to find i t s balance; the chapter presented largely chronologically. The digitoxose rings have the substituents equatorial.  Cj conformation, with the - C H 3  The conformation of the glycoside and genin  are depicted below.  Fig. II.5.1  Solid-state conformations of digoxigenin (lower) and the digoxin glycone (upper). (From Go et a l . )  - 261 -  F i g . II.5.2  A . 500 MHz H NMR spectrum of digoxin i n DMS0-d . B. Same molecule following a D 0 exchange. Hydroxy protons are marked (•) In A. A  6  2  - 262 -  A solution of 19 mg digoxin i n 0.4 ml DMSO-dg was used throughout  the *H NMR studies.  Some line-narrowing was induced by  heating the sample to ca. 70°C and allowing to cool to room temperature.  We assume that this treatment results In the molecules  forming a more homogeneous solution.  The spectrum reveals a low-field  region with r e l a t i v e l y sharp, overlapping l i n e s , whereas those i n the the methylene region appear broader and l i t t l e d e t a i l i s evident.  This  line-width (T2) phenomenon can be postulated to result from micelle formation, with the l i p o p h i l i c steroid moieties i n the aggregated phase, and solvated, more mobile, pendant sugars.  I f this were the case, of  the sugar rings one would expect ring III to have the greatest mobility, and ring I the least; conclusions to support this hypothesis appear later. The o r i g i n a l spectrum i n DMS0-d_6 required some s i m p l i f i c a t i o n , and a D2O exchange was performed.  The spectrum of the parent compound  i s shown i n F i g . II.5.2 A with the -OH resonances marked (•), and that of the -0D compound shown i n F i g . II.5.2 B; the l a t t e r was used for a l l subsequent  studies.  At the outset, the probable v i c i n a l coupling constants for the digitoxose groups were estimated,  5  and these are l i s t e d i n Table II.5.1.  - 263 -  Protons  Table II.5.1  Expected J_ (Hz)  (1,2a)  10.0  (l,2e)  2.5  (2a,3)  2.6  (2e,3)  3.1 - 3.6  (3.4)  2.6 - 3.1  (4.5)  9.3 - 9.6  (5.6)  6.2  Expected  J_ values for a B-digitoxose sugar residue.  These data enabled a surprising number of assignments to be made by direct inspection of the spectrum.  We concentrate f i r s t on the  region corresponding to 6 > 2, where twelve carbohydrate and six s t e r o i d a l protons are expected to resonate.  Five groups of resonances  can be distinguished In this region, and these are shown expanded i n Fig.  II.5.3 (after resolution enhancement with a double-exponential  time-domain f i l t e r ) . require  "Ladder" assignments indicate the resonances which  assignment.* On the basis of i t s chemical s h i f t , we assign the most low-field  proton (6 5.81) to H-22 on ring E of the steroid, which has a small (1.83 Hz) long-range coupling to the geminal H-21 protons. Inspection of the region 6 4.92 - 4.72 ( F i g . II.5.3. B) reveals, amongst other resonances, the AB portion of an ABX system; a large (geminal) coupling *The f i n a l assignments are included, for reference l a t e r i n the chapter.  - 264 -  II.5.3  Expansions  of  the  H NMR s p e c t r u m of  D2O  added.  of  low-field digoxin  region in  of  DMS0-d_6,  the  500  with  a  MHz drop  -  T  | I  4.06  I 1  | >~ 1 I J ' T T T j - T T T p l T T T I  1 | 1 I  I | 1 I" I ~[ I  265 -  T I T"T"I  ' *[ fT  I |  4.02 3.98 3.94 3 . 9 0 3.86  Fig.  II.5.3  cont.  ppm  5  - 266 -  constant J_ (A,B) i s indicated by the intensity build-up and direct measurement.  The smaller coupling constant matches that i n H-22.  Accordingly, these protons are assigned to the geminal H-21 protons. Further "general" assignments of digitoxose protons may be made on the basis of a comparison of v i c i n a l coupling constants (Table II.5.1) and the s i m p l i f i c a t i o n effected by O-deuteration.  Based on the  magnitude of their two coupling constants to the geminal H-2 protons, three sets of anomeric protons are distinguishable i n F i g . II.5.3 B. Between 6 A.07 and 3.84 ( F i g . II.5.3 C) the almost i d e n t i c a l digitoxose H-3 resonances occur.  This i s based on their three very similar  coupling constants (giving r i s e to a coupling pattern resembling a quartet) and their s i m p l i f i c a t i o n upon O-deuteration (data not shown). By virtue of their unique coupling to the methyl groups, the digitoxose H-5 protons can be i d e n t i f i e d to resonate between 6 3.74 and 3.57 ( F i g . II.5.3 D)* F i g . II.5.3 E and F show the digitoxose H-4 protons, i d e n t i f i e d by matching coupling constants and comparison with data i n Table II.5.1.  Two other protons resonate i n this region and, by  default, must be s t e r o i d a l i n o r i g i n .  The proton at 63.01 can be  assigned to H-4"', as ring I I I i s the only digitoxose residue having an exchangeable hydroxyl proton at position 4 and t h i s agrees with the r e s u l t s of the deuterium exchange experiment.  Thus, the simple  procedure of matching spectral s p l i t t i n g s and a deuterium exchange  *The HOD peak at 6 3.66 has been eliminated by presaturation.  - 267 -  experiment led to s i g n i f i c a n t steps being taken i n the assignment of the "low-field" region of the spectrum (6 > 2). The above assignments were made on the underivatized material, and encouraged us to proceed with a f u l l assignment of the spectrum. Although this was i n apparent c o n f l i c t with the conclusions of Chap. II.3,  the temptation to proceed with this assignment was strong since  any conformational deductions would be of considerable Interest. A proton 2D J_-resolved experiment was performed i n an attempt to unravel the h i g h - f i e l d methylene/methlne  resonances (24 protons  extensively overlapping between 61 and 2.1).  Unfortunately this f a i l e d  to produce useful information for the h i g h - f i e l d region due to the protons' short J_2's (See Discussion i n II.2.5.2). With this experiment, most of the transverse magnetization had relaxed beyond the point of detection after only s i x or seven experimental increments ( t i / 2 ) had been performed. low-field  Some data were extractable for the resonances of the  region, but these provided no additional information over what  had been gained by straight inspection of the 500 MHz *H NMR spectrum.  We believe this to be the f i r s t report of the f a i l u r e of the  2D J-resolved experiment.  Attempts to improve the efficacy of this  experiment included the use of a much larger than necessary sweep-width i n the J-dimension, to shorten At_i/2 and, hence, improve the d i g i t i z a t i o n of the i n i t i a l portion of the interferograra.  Also, the  experiment was attempted at elevated temperature (60°C) i n an effort to lengthen J_2's; both attempts were to no a v a i l .  - 268 -  Next the COSY experiment  (II.2.5.3.1) was performed  proton coupling connectivities. experiment  to establish  (SECSY was avoided as i t Is a spin-echo  and would be expected to suffer similar limitations to those  described for the 2D J_-resolved experiment).  A contour plot of the COSY  spectrum of the deuterium exchanged digoxin i s shown i n F i g . II.5.4. Aside from a confirmation of assignment  for the lactone, ring E, not  much i s clear u n t i l we consider expansions of the two regions indicated.  Between 6 4.2 and 2.8 the digitoxose 3,4 and 5 protons  resonate, and the COSY plot for this region i s given i n F i g . II.5.5. With H-4''' previously assigned, H-3''  1  and H - 5 " are readily assigned 1  as the most h i g h - f i e l d resonances In each group of resonances.  The  lowest f i e l d of the carbohydrate H-4 resonances i s coupled to the lowest f i e l d of the H-5 resonances, but these cannot yet be unambiguously assigned to H-4' and H-5', or H-4" and H-5".  The connectivities do  not distinguish H-3' from H-3". Next, the connectivities into, and out of, the h i g h - f i e l d region are considered i n F i g . II.5.6.  With H-3'" assigned, H-2'" and H-2'"  and H-2"' can be i d e n t i f i e d i n the overlapping high f i e l d region, and these lead i n turn to the assignment and H-l''.  of H-l' " , which l i e s between H - l '  The highest f i e l d anomeric proton (which we l a t e r assign to  H-l') has two clear correlations into the methylene region, and those protons are coupled to the highest f i e l d of the carbohydrate H-3 protons (H-3').  S i m i l a r l y , the anomeric proton at lowest-field i s on the same  ring as the low-field carbohydrate 3 proton ( H - 3 " ) .  Hence, we have now  -  r-—i—i—i—i  i  r  1  1  269  I  I T  1  T  -  I  1  1  1  1  4  a  A  r  1  !  !  I  1  JUL I  1 ppm  •  - - y /  • • c£7  Fig. II. 5 . 4  u  B  i  I  22  I  I  22 5 0 0 MHz COSY spectrum of digoxin ( 1 ) i n D M S 0 - c [ , 6  after deuterium exchange. The HOD signal was suppressed by p r e i r r a d i a t i o n , and the 5 1 2 * 1 0 2 4 word data set symmetrized prior to display i n the contour mode.  - 270 -  Fig.  II.5.5  Expansion  of  region  A in  Fig.  II.5.A.  Fig.  II.5.6.  Expansion  of region  B i n F i g . II.5.4.  - 272 -  determined a l l connectivities of the ring protons of the digitoxose units I and I I , but we cannot assign a particular set of resonances to either  unit. Connectivities  H-5  between the carbohydrate methyl-groups  and the  protons are present, but unassignable due to extensive cross-peak  overlap.  S i m i l a r l y , couplings between the steroid protons (H-3,  and H-17)  and the methylene/methine region are evident, but not  assignable.  The H-12  and H-17  correlations, and H-3, H-12  or H-17  correlations  four.  H-12 fully  resonances are each expected to have two The resonance at 6 3.88 cannot be either  and must, by default, be H-3 .  However, only three  e  are clear, and the fourth must either be below the contour  threshold or overlap with another cross-peak. The next stage of the assignment nOe"  experiments  involved a series of "driven  (II.2.A.3); these proved p i v o t a l i n the d i f f e r e n t i a t i o n  of ring I and ring II protons, and several s t e r o i d a l protons.  It turns  out that the molecule i s i n the spin-diffusion regime and a l l observed nOe's were strong and negative.  The highest f i e l d anomeric proton i s  s u f f i c i e n t l y far from the other two that i t could be irradiated with reasonable s e l e c t i v i t y .  This Induced a strong i n t e r - r i n g nOe into the  steroid proton at 6 3.88  (data not shown) which was thereby assigned as  H-3.  Since only H - l ' i s expected to induce an nOe into a steroid  proton, t h i s assignment  provided confirmation of the assignment  of  H-3;  numerous other nOe's were induced i n the a l i p h a t i c region - presumably into the H-2  and H-4  protons.  With t h i s second c r i t i c a l assignment  f i r s t being H-4'"), H-2',  H-2',  6  8  H-3',  H-4'  i d e n t i f i e d from the COSY connectivities.  and and H-5' may  (the  be  Since these resonances, and  -  273  -  {H-3',3"}  i'  I ' I  2.1  2.1  1 1 1 ' '  1 1 1 1 1  1.9  1.9  • i'  1 1 1 1 1 '  1.7  1.7  >  1 1 1  i ' ' ' ' i • '' i ' ' ' ' i ' ' ' ' i 1  1.5  1.5  1.3  1.1  ppm  1.3  1.1  ppm CONTROL  2.1  Fig.  II.5.7  1.9  1.7  1.5  1.3  1.1  ppm  D r i v e n nOe e x p e r i m e n t s p e r f o r m e d o n d i g o x i n , induced i n t o the methylene/methine region.  showing  nOe's  - 274 -  hence, their c r o 6 S - p e a k s are so close, v e r i f i c a t i o n for these assignments was sought from the nOe experiments. Confirmation of the digitoxose H-2 protons' assignments comes from the three nOe experiments, i l l u s t r a t e d i n F i g . II.5.7. of (equatorial) H-3'" clearly shows H-2'" and H-2"'. e a  Irradiation  I r r a d i a t i o n of  H-3' and -3" induces nOe's i n the indicated geminal protons; H-2^ and H-2", and H-2' and H-2" are d i f f e r e n t i a t e d by the COSY experiment 6  3  9.  (Fig. II.5.6) which shows the connectivities from H - l to H-2 1  g  to high  f i e l d of H-2 ' and H-2^" (which are coincident) and H-2^ to h i g h - f i e l d e  of H-2", based on the connectivities from H-3', 3 " .  The experiment  involving i r r a d i a t i n g of H-l' i s included (vide supra). The carbohydrate H-5 region assignment was confirmed by the nOe experiments indicated i n F i g . II.5.8.  By i r r a d i a t i n g H-l', or H - l "  r > <i l i i i I i i i I i i i I i « • l • i i l • • i I « >  3.74 Fig. II.5.8  3.70  3.66  3.62  ppm  Driven nOe experiments showing nOe's induced into digitoxose H-5 protons by i r r a d i a t i n g H-l', or H-l'' and H - l ' "  - 275 -  Table II.5.2  Average interproton distances between -Me(18) and -Me(19), and steroid ring protons. (A)  C(18)-H  3  H-ll  3  2.868  H-8  3.193  H-17  3.398  H-l 2  3.680  H-23  C(19)-H  a  a.e  3.088, 4.410  H-22  3.982  H-l  2.621, 2.825  a.e  H-8  2.732  H-5  2.762  H-ll  2.887 i  H-6  2.999  - 276 -  plus H-l''* an nOe i s induced i n the corresponding H-5 with which i t has a 1,3-diaxial relationship. A number of further checks were performed which confirmed the previous assignments. H-4'"  I r r a d i a t i o n of H-5' and - 5 ' " induced nOe's into  and the lower-field of the H-4'/H-4'* overlapping resonances, and  confirmed assignments of H - l ' and H - l ' " (data not shown). complemented the experiment where H-5"  was i r r a d i a t e d .  This  I t i s clear  that a carefully selected array of i r r a d i a t i o n experiments has, i n addition to providing assignments, the added bonus of enabling previous assignments to be checked or confirmed; hence, such a series of experiments provides a "network" of assignment information. These nOe experiments shed some light on the steroid assignment, but far less than one might have anticipated.  I t i s well known  10  that  the a x i a l methyl groups (C-18 and C-19) can act as "transmitter beacons" to induce responses from a x i a l protons which are near them i n space. When the C(18)-H3 protons are i r r a d i a t e d , one might expect nOe's with H - l l , H-8 a  on the p-face of the steroid, and H-12  and H-17.  Further,  one might expect an nOe into ring E, depending on i t s orientation. c i s C,D  The  ring-fusion results i n the protons on C-15 and C-16 being too  far from -Me(18) to have an nOe induced.  I r r a d i a t i o n of C(19)-H i s 3  unlikely to be of much use i n assigning protons on ring A, as a result of the c i s A/B ring junction.  Protons referred to i n the following  discussion are l i s t e d i n Table II.5.2, with the internuclear separations calculated from the X-ray data . I r r a d i a t i o n of -Me(18) induces the expected nOe's. we note two nOe's - one into H-8  At 6 < 2.1,  (6 1.48) and the other, H-lip (6  - 277  I  i  2.1  I I  i  l  i i  i  i  I i  1.9  F i g . II.5.9  i  i i  I i  i i  i  l  i i  1.7  i i I i  i  i  i  -  l  i i  1.5  i i  I i  i i  i  l  i  1.3  i  i i  l  i i  i  i  l  i i  i i  I i i  1.1  Driven nOe experiments with steroidal methyl peaks irradiated; the displayed region 1B that of the methylene/methine protons, although other nOe's are evident to lower f i e l d .  - 278 -  1.12).  This assignment  i s consistent with the COSY data, where a  correlation exists between H-17 and H-16, but none between H-17/H-12 and H-8.  Since both proton multiplets are overlapping with other signals,  the enhancement factor cannot be r e l i a b l y quantified.  An nOe i s seen at  6 3.2 where H-12 and H-17 overlap, but this i s l i k e l y to include nOe's into both protons.  (Based on their solid-state interproton distances,  H-12 would be expected to show 70% the nOe of H-17).  Significant nOe's  are seen i n a l l three protons on ring E. I r r a d i a t i o n of -Me(19) also reveals s i g n i f i c a n t nOe's i n H-8 and H-ll8.  A t o t a l of six protons i n the methylene/methine region display  nOe's, consistent with the six protons having interproton separations < 3.0 A, l i s t e d i n Table II.5.2.  No cross checks can be made, since the  multiplets overlap with the signals from their closest neighbours. I r r a d i a t i o n of H-22 results In an nOe i n H-17, allowing i t to be distinguished from H-12 (data not shown).  The proton assignment  from complete, s t i l l much information has been gleaned. assignments  i s far  With the  resulting from the nOe experiments, l i t t l e i s to be gained  by returning to the COSY map f o r the crowded methylene/methine region. Far fewer cross-peaks are d i s t i n c t than anticipated; t h i s w i l l be discussed l a t e r In the chapter, along with other observations r e l a t i n g to the performance  of a l l the experiments described.  Before leaving the *H NMR data It i s i n t e r e s t i n g to look at the dlgltoxose sugar assignments  i n the l i g h t of predictions on the s e l f  association of the molecule i n solution.  Such a system would have  r e l a t i v e l y mobile, solvated sugars extending from the aggregated phase.  - 279 -  3.0  Fig.  II.5.10  CSCM e x p e r i m e n t  N  ppm 6  H  8 c  Fig.  II.5.11  The  high-field  II.5.10 is indicated.  region  plotted  of  the  between  CSCM d a t a - s e t  the  frequency  in  fig.  limits  - 281 -  Hence, one might expect the terminal digitoxose ring III to have more mobility (and, therefore, narrower l i n e s ) than ring I I , and likewise even more mobility than ring I.  Since the coupling pattern for each  proton at a p a r t i c u l a r position on a l l rings i s the same, we can consider peak height to be a r e f l e c t i o n of peak width (and T_2).  With  the eye of f a i t h , i t i s clear that this i s so for a l l the observable resonances to lower f i e l d ; ring III appears to have greater mobility than ring II and s t i l l greater than ring I. 13  We now turn our attention to the  C data.  The broad-band  13  decoupled  spectrum of digoxin i n DMS0-d_6 i s well dispersed with  C NMR  few overlapping l i n e s .  The CSCM experiment was performed and parts of  the resultant data-table are presented i n Figs. II.5.10 and 11. Those 1 13 H assignments made were used to assign the corresponding C signals. C signals between 6 40 and 120 produced correlations i n the H spectrum between o„ 2.9 and 6.0; a l l protons i n this region were a.  13  assigned.  The  C NMR  assignments followed smoothly provided i t was not  the case that both the *H signals and the  13  C  signals were very close.  13  The  C assignments are given i n Figs. II.5.10 and 11.  In the proton  methylene/methine region, eleven assignments were made and the signals were readily correlated.  The steroid methyl carbons were  r e a d i l y i d e n t i f i e d (data not shown). checked with the l i t e r a t u r e ,  5  C  A l l the above assignments were  and were i n agreement.  Whilst the exercise of extracting an F^ ( H) trace for each A  1 3  C  signal was performed, no proton assignments were made on this basis. Our spectra were run with the molecule dissolved in DMSO-dj, whilst Brown et a l .  6  used CDCI3:DMS0-d_e (1:2); small but sometimes s i g n i f i c a n t  282  -  differences  in  chemical  shifts  were  -  apparent,  making  such  comparisons  13 dangerous. fully  It  Is  assigned,  significant  the  to  note  that  *H NMR a s s i g n m e n t  if  would  the  C NMR s p e c t r u m  easily  follow  from  were  the  CSCM  in  this  experiment. The chapter  large  invites  complicated higher wider  a  few  information  which,  in  of  broadened  been  trans,  since  result  from  is  in  the  nOe  In is  other  equally line  hand,  have  broadening  difficulty have NMR  seen  in  assignments  and  comparisons  longer  an  are  known.  is one  of  a  not is  to  and  of  to  groups  spectrum  pose lie  a  in  the  same  for  they  have  the  represented,  have  a  limit  of  Tj>-  *H NMR ring  a  spectrum  junctions  been  more  information  to  were  irradiated.  consider  the  is  simpler,  different  be  true *H -  used here,  1 3  C  On  is  but  T_2  now  procedures. to  assign  where  pairs  experimentation procedures,  related  it  reasons.  problem,  available  can  other  structurally  much  serious  those  further  is  for  H NMR s p e c t r u m  available,  series  to  However,  driven  C/D  considerably  more  overlap  The  and  methyl  are  these  area.  A/B  are  dispersion)  extensive  isolation,  seems  II.4)  NMR d a t a  *H s p e c t r a  however,  the  C NMR in  1  had  C  parameters  the  where  seems  assigned  (Chapter  Hi a s s i g n m e n t limited,  no  by  1 3  3  the  assign  assignment  that  spectrum  to  more  permitted  experiments  although  difficult  that  assign  the  intrinsic  possible;  1  the  the  methylene/methine  to  would  *H NMR v s .  Although  determined  the  easier  this  the  J_; s m a l l e r  content.  case,  signals  have  of  experiments  this  on  remarks.  (inclusion  range  would  discussion  13  We  a  C  *H  for is  the  which  the  somewhat  such  as  molecules.  13 Although NMR a s s i g n m e n t s  it will  is  probable  emerge,  that  there  is  further only  spectroscopic  one  contender  at  aids  to  present.  C  -  The  2D  INADEQUATE  problem  for  solution  to  period.  A  digoxin  the  less  -  8  40  narrow  that is  the  the  not  and  ranges  to  of  be be  small  occur.  3',  It may  In  a  one  resonances  from  3 " ,  5',  second  been  time  order  shift  range  compensation  similar  5*"  with for  2 13 J_ ( C  that  centres  and  In  demonstrated  considers  5 "  obvious  concentrated  chemical  has  The  realistic  with  require  when  3 * " ,  useful.  sufficiently  associated  experiment  carbon  prove  performed  unreasonable  (e.g.  a  relatively  resonances  3  Hz,  preparation  from  -  undoubtedly  p r o b l e m may  ^ C  this  can  experiment  obvious  spectra  = ca.  fairly  the  (coupled)  effects; C)  the  resulting  carbohydrate such  is  enable  distortions which  experiment  283  fall  within  into  4  p.p.m.). In  concluding  this  chapter  limits  of  the  various  experiments  limits  on  the  general  strategy  The resonances to  2D J - r e s o l v e d having  jT  resonances  experiment,  as  methylene/methine 1.0  to  earlier  2.1  had  were  2  experiences,  values the  in  and  and  were  2D J _ - r e s o l v e d  effects  for  protons  these  distinguish  the  out  were  in of  the  Thus,  experiment  of  those  those attached precisely ID  the  resonances  short.  the  chapters.  mostly,  resonated  too  the  earlier  were,  to  consider  satisfactorily  unfortunately  which  to  Unfortunately,  difficult  And  used  worked  which  deshielded  region. T_  values  not  important  postulated  carbons.  w h i c h were they  2  is  actually  experiment  heteroatom-substituted  the  6  longer  it  in  proved  to  in  the  contrast  range to  be  disappointing. The  COSY  Notwithstanding  experiment the  coarse  was  very  useful  digitization  (ca.  for  the  1 0 Hz  low-field p t  -  1  ) ,  region.  correlations  - 284 -  to protons only a few Hz apart could be r e l i a b l y d i f f e r e n t i a t e d . c r i t i c a l to note that this was only possible with l i k e protons  It i s  (e.g.  H-3 * and H-3 ') where the two c o r r e l a t i o n peaks have very similar 1  patterns.  By comparison, i t was not possible on a f i r s t inspection to  discern correlations between H-12 or H-17 and the h i g h - f i e l d region. Also, coupling pathways to (broader) proton resonances to high f i e l d i n the methylene/methine region were clear (e.g. H-3 ). e  f i e l d region, f a r fewer cross-peaks  In the high-  appeared than expected.  This i s  probably the result of two factors, cross-peak overlap and the attenuation of the (broad) h i g h - f i e l d signals by the necessary apodization function. "tailing"  Unfortunately without  sinebell  the l a t t e r , extensive  occurs which can lead to false "cross-peaks"  where they  Q  interfere.  The question of attenuation by the apodization function may  be partly remedied by using a less severe function (e.g. the double-exponential  function), and by using a smaller block-size and  z e r o - f i l l i n g to give adequate d i g i t i z a t i o n . The driven nOe experiment proved quite useful as a q u a l i t a t i v e rather than quantitative t o o l . unfortunate  Quantitations were limited by  signal overlap i n the low-field region even at 500 MHz which  made selective i r r a d i a t i o n d i f f i c u l t , and overlap i n the h i g h - f i e l d region which made the determination of f r a c t i o n a l enhancements nigh on impossible.  The procedure of assigning, e.g. the Irradiated methyl  protons an area of 300 "units", i n the difference mode, and induced i n t e n s i t y enhancement calculated by area r e l a t i v e to t h i s , did not prove successful when compared with expected ratios based on known  -  distances  from  break-down  of  motion,  the  the  or  the the  methyl  performed  non-overlapping The suited the  to  simplified  difficulty  digitoxose  experiment  solid-state.  resonances  system  of  *H l i n e - s h a p e s  this  (Fj)  either  the  in  ID  the  worked  be  resulting a n nOe  of  Nevertheless,  within  usual  result the  the  anisotropic  a proton  directly  qualitatively,  restrictions  of  under  the  of  experiment.  Within  easily  a  from  extraordinarily  nature.  were  could  measuring  signals.  CSCM e x p e r i m e n t  a  This  -  equations,  in  well,  285  the  well  limits  recognizable.  and  of  It  seems  well  digitization,  is  without  doubt  13  that  almost  credibly emerge  all  proton  assigned.  from  found  most  There  the  this  assignments  In  some  digoxin  effective initial  ways  this  study.  with  could  It  was  on  is  the  calls  derivatized  emphasis  be made  the  if  most  for  a  the  C  signals  important  change  in  the  oligosaccharides  and  total  of  assignment  fact  were  to  protocol  brucine. the  1  H  NMR  13 spectrum, the  followed  by  assignment  CSCM e x p e r i m e n t .  difficult would  to  fully  probably  possible  using  be  With  digoxin,  assign to  the  of  even  first  at  assign  methods  the  C NMR  where  the  500 MHz, as  described  1  spectrum  H NMR  a more  much o f above,  the and  on  the  spectrum  effective  of  is approach  *H NMR s p e c t r u m then  basis  concentrate  as on  the  13 total  assignment  INADEQUATE time  In procedures  the  experiment  prevented  completion  of  of  this the  C NMR  should  in  a  be  spectrum.  To  performed,  but  reasonable  *H NMR a s s i g n m e n t  conclusion, described  in  this  study  previous  on  time would  be  can  end,  the  a paucity  period.  digoxin  chapters  this  With  of  this  2D instrument information,  trivial. indicates be  relied  that upon  not for  all  the  studies  -  of  molecules  particular, optimal  in  which the  this  have  broad,  286  strongly  -  overlapping  homonuclear  2D e x p e r i m e n t s  regard.  the  Of  appear  alternative  to  resonances. be  less  experiments,  CSCM  In  than operates 13  best  in  this  spectrum should  be  must  milieu, be  with  the  assignable.  performed  to  fundamental To  complete  this this  end,  limitation the  chapter.  2D  that  the  INADEOUATE  C NMR experiment  -  287  -  REFERENCES 1.  L.F.  Fieser  and M.  Corporation, 2.  J.H.  Hoch.  of  South  3.  S.  Smith.  4.  K.  Go,  A  survey  Carolina  G.  Fieser.  New Y o r k .  J .  of  Kartha,  Reinhold  Publishing  752-754.  glycosides  and  genins.  University  1961.  Soc.  and  pp.  cardiac  Press.  Chem.  Steroids.  1959.  508-510  J.P.  (1930).  Chen.  Acta  Cryst.  B36,  1811-1819  (1980). 5.  C.  Altona  and  C.A.G.  Haasnoot.  Org.  Magn.  Reson.  13,  417-429  (1980). 6.  L.  Brown,  Nemorin. 7.  A.  Bax,  Reson. 8.  S.L.  9.  A.  43,  Cheung,  Chem.  T.A.  Patt.  Bax  H.T.A. J .  Frenkiel, 478-483  Thomas,  Perkin  R.  T.R.  Trans.  Freeman,  Watson,  1.  and  1779-1781  and M.H.  Levitt.  J.L.E. (1981). J .  Magn.  (1981).  Carbohydr.  a n d R.  R.  S o c ,  Freeman.  Res. J .  In Magn.  press. Reson.  44,  542-561  (1981).  -  II.6  288 -  EXPERIMENTAL FOR SECTION I I  In general, a l l spectra reported at 270 MHz were recorded on a home-built spectrometer based on an Oxford Instruments  superconducting  magnet (B^j 6.35 T), a Bruker WP-60 console, and NIcolet 1180 computer and 293B pulse-programmer.  Unless stated otherwise, standard NTCFTB  soft-ware was used with this instrument.  Spectra at 400 MHz ( H) were A  recorded on a Bruker WH-400 instrument, having a 9.4 T magnet. The operating program i n routine use was DISNMRP.  Five hundred and 360 MHz  spectra were performed using Nicolet spectrometers at the factory In California. It i s the author's opinion that a detailed documentation parameters  of a l l  for each experiment would not be useful, since these w i l l  vary widely depending on the instrument being used.  I t ±e_ useful,  however, to document any extra pieces of information which are relevant to a p a r t i c u l a r experiment, and not mentioned i n the text. For ^H experiments, the 180° pulse-length was determined (zero i n t e n s i t y cross-over), and a l l pulse-lengths calculated from t h i s . For 13  C observe the procedure was the same, often using a reasonably concentrated "dummy" sample i n a similar solvent to that i n which the molecule of interest was dissolved.  Proton pulses through the decoupler  channel on the C probe were determined by the methods outlined i n the 1 3  text. When attempting an experiment  for the f i r s t time, i t i s common to  begin with a very simple molecule at high concentration. Knowing the molecule's geometry, T  A  parameters, e t c . can greatly a s s i s t i n the  - 289 -  I n i t i a l evaluation, or "debugging"  of an experiment.  With heteronuclear  c o r r e l a t i o n experiments i n this thesis, for example, a concentrated solution of n-propanol was used to f a m i l i a r i z e the operator with important parameters and experimental set-up. For DEPT (II.2.2.2), the C parameters were selected i n the 1 3  usual manner, and H pulse-lengths "fine-tuned" by performing the A  experiment with 6  a  90°.  When perfect cancellation of - C H and - C H 2  3  peaks was achieved, the proton pulse lengths were correctly set. With a l l homonuclear decoupling experiments (including nOe), a c r i t i c a l factor i s the choice of decoupler power and position. with s u f f i c i e n t decoupler power, Incomplete  Even  decoupling may result from  the resonance frequency of the saturated spin changing while the decoupler i s on. In most cases t h i s 1B not too important a consideration, but i t may be necessary to slowly move the decoupler frequency ( i n 0.5 Hz steps) to find the optimal position to e f f e c t perfect decoupling.  This Is p a r t i c u l a r l y a consideration i n s p e c t r a l l y  crowded regions where decoupler s e l e c t i v i t y i s d i f f i c u l t to achieve.  To  achieve this s e l e c t i v i t y Is most d i f f i c u l t with the SDDS experiment (II.2.3) and i t i s fortunate that r e l i a b l e two-dimensional  experiments  having none of these problems (e.g. COSY) may be resorted to. With nOe experiments i t i s important to come as close to 100% saturation as possible, without leakage (or " s p i l l - o v e r " ) of decoupler power Into spectrally close spins. experiment  i s done:  This i s e a s i l y checked before the  the i r r a d i a t i o n of the spin i n the most crowded  spectral region i s used to determine the maximum permissible power,  - 290  without  spill-over.  -  No excessive time-averaging  i s necessary here, as  one i s not looking for very small (< 1%) changes in peak height.  Such  checks not only give the experimenter a degree of confidence, but the percentage saturation In each experiment may  be useful information i f  c e r t a i n comparisons are to be made. This practice of c o l l e c t i n g only the FID comprising  the  difference spectrum, and not both spectra i n d i v i d u a l l y , i s not recommended, since (a) information i s needlessly discarded, and  (b)  where multiple nOe experiments are necessary, more machine time w i l l be required to accumulate the same data.  With the procedure for SSNOEDS  ( I I . 2.4.1) i n the text (ri + 1) experiments are necessary for ii nOe i r r a d i a t i o n s ; i f only the difference FID i s collected, 2n_ experiments would be necessary  to derive the same information.  When the sample concentration i s low, demanding a very large number of scans to build up signal-to-noise, the author has found i t useful to increase the r e p e t i t i o n rate by (a) using no relaxation delay (b) decreasing the period of i r r a d i a t i o n to ca. 1 or 2 T j , and, to compensate for these a l t e r a t i o n s i n the experimental decreasing the tip-angle (from ca. 80° to 60°).  scheme, (c)  In such a way,  four  nOe  experiments were possible i n an overnight run with ca. 2 mg of sample on a Bruker WH-400, allowing the detection of < 1% nOe's with  confidence.  With the one-dimensional experiments discussed above, data storage space on hard disc i s seldom l i m i t i n g , and spectra can be acquired with good d i g i t i z a t i o n . NMR  In the case of two-dimensional  (2D)  experiments the reverse i s often the case: long "acquisition times"  -  291 -  i n t^i may require large amounts of instrument time to acquire, large blocks of data to store, and long computation times for data-processing.  I t i s therefore common to r e s t r i c t the sweep-widths i n  both dimensions to the minimum possible value.  I f , for example, acetate  peaks to high f i e l d have no information, they may be "folded" into the spectrum where they do not overlap with any other peaks.  With t h i s  approach, the minimum block-size can be chosen i n both dimensions and experimental times reduced. With the 2D J-resolved experiment (II.2.5.2), fine d i g i t i z a t i o n Is not necessary i n F_2, and the experiment should be optimized for d i g i t i z a t i o n i n F j , the "J_-dimenslon".  In the interest of minimizing  instrument time, i t i s common to c o l l e c t data for 32 t_i increments and z e r o - f i l l i n JFj_ to give the necessary d i g i t i z a t i o n to resolve small couplings.  Although the author has not had opportunity to assess  symmetrization procedures i n this experiment, i t i s quite l i k e l y they w i l l be extremely useful since second-order artefacts are seldom symmetrical about F_| • 0 and such a procedure should prove useful i n i d e n t i f y i n g such peaks and simplifying the spectrum.  Other  advantages  of this procedure are mentioned i n the text. We have had cause to use the COSY experiment (II.2.5.3) quite extensively i n t h i s study, and have found i t remarkably resistant to abuse.  Pulse-angles need only be set to ± 10%, and coarse d i g i t i z a t i o n  (e.g. 5 Hz/pt) i s often s u f f i c i e n t .  The data may be acquired quite  rapidly (ca. 0.2 sec. relaxation delay) and the s i n e b e l l apodization function i s almost always appropriate for both dimensions.  The Jeener  - 292 -  experiment  i s so powerful that, given a state-of-the-art  spectrometer,  i t i s arguable whether SDDS methods are j u s t i f i a b l e . The important heteronuclear chemical s h i f t correlation (CSCM: II.2.5.4) i s described i n some d e t a i l i n the text and a discussion on a " t y p i c a l " set-up procedure i s included. The author has not had time to thoroughly evaluate the 2D nOe experiment  (NOESY: II.2.5.6).  t h i s experiment  Probably the most d i f f i c u l t decision with  i s choosing an optimal mixing time.  tumbling regime, a "crude" driven nOe experiment  In the slow  (II.2.4.3) should  indicate the rate of nOe build-up and allow an optimal x ^ n  chosen.  x  to be  In the fast tumbling regime, a v a l i d approach might be to f i r s t  perform a "crude" transient nOe experiment  (II.2.4.2) i n order to  characterize a " t y p i c a l " nOe bulld-up curve and f a c i l i t a t e the choice of a useful " % £ . x  Cross-peaks are small i n this l a t t e r experiment, and  symmetrlzatlon of the data-set w i l l almost certainly  be necessary.  As  i s the case with COSY, coarse d i g i t i z a t i o n w i l l often s u f f i c e , thereby reducing the time required to c o l l e c t the data f o r this somewhat long experiment.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059365/manifest

Comment

Related Items