E V A L U A T I O N O F P O S T M O R T E M D N A D E G R A D A T I O N U S I N G T H E C O M E T A S S A Y by L A U R A A L E X A N D R A J O H N S O N B . S c , Simon Fraser University, 1998 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R T M E N T O F P A T H O L O G Y A N D L A B O R A T O R Y M E D I C I N E ; E X P E R I M E N T A L P A T H O L O G Y We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A A p r i l 11 t h , 2002 © Laura Alexandra Johnson, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my writ ten permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract One of the most important longstanding problems in the field of forensic medicine is the determination of time of death upon the discovery of a possible homicide victim. With a majority of homicide victims discovered within the first 48 hours, it is critically important to be able to determine time of death quickly and with accuracy and precision. Current methods of determining postmortem interval vary, but none can provide better than an 8-hour window time estimate. In this paper, the potential application of single-cell gel electrophoresis (also known as the comet assay) to evaluate postmortem cell death processes, specifically nuclear D N A fragmentation, is assessed. Upon the death of an organism, internal nucleases contained within the cells should cause chromosomal D N A to degrade into increasingly smaller fragments over time, and i f these fragments can be isolated and visualized, the fragmentation should prove to be measurable and quantifiable. A n original study providing proof of the concept of postmortem D N A fragmentation between early and late time periods was conducted using human leukocytes. With an established trend of D N A degradation seen in the leukocyte results, this study was then expanded using a porcine animal model, over a longer time period, with more frequent timepoints evaluated. D N A degradation in all samples was revealed by single-cell gel electrophoresis (also known as the comet assay) and quantified by the use of the DNA-specif ic quantitative stains propidium iodide and silver nitrate, as measured by digital camera affixed to a microscope. The comet 'tail-i i moment' gave a measure of the proportion of fragmented to non-fragmented DNA, while the 'tail-length' provided the relative size of degraded D N A fragments. In both models, an increase in D N A fragmentation was found to correlate with increased postmortem interval from 0-56 hours postmortem as evaluated by comet tail-moment and by comet tail-length, with tail-length providing the strongest statistical correlation, based on regression analysis. The postmortem D N A fragmentation observed in this study reveals a sequential, time-dependant process with the potential for use as a predictor of postmortem interval in homicide cases. iii TABLE OF CONTENTS Abstract i i Table of Contents iv List of Figures v Acknowledgments vi CHAPTER I OVERVIEW 1 1.1 Introduction 2 1.2 Postmortem Techniques Review 4 1.3 Legal Issues 16 1.4 Conclusions 18 CHAPTER I I EXPERIMENTAL DESIGN 19 2.1 Cellular Biology Introduction 20 2.2 Experimental Hypothesis 22 2.3 Quantitative Staining of D N A for Light Microscopy 24 2.4 Single Cell Gel Electrophoresis (Comet Assay) 43 2.5 Conclusions 44 CHAPTER I I I EXPERIMENTAL PATHOLOGY 45 3.1 Introduction 46 3.2 Materials and Methods 49 3.3 Results 52 3.4 Discussion and Conclusions 59 3.5 Future Experiments 62 3.6 Alternative Applications 63 BIBLIOGRAPHY 64 iv L I S T O F F I G U R E S C H A P T E R I Figure 1. D N A Double Helix Structure 21 C H A P T E R II Figure 2.1. Eosin & Hematoxylin Nuclear Staining 27 Figure 2.2. Acridine Orange Nuclear Staining 29 Figure 2.3. Giemsa stained slide of bone marrow 31 Figure 2.4. Masson's Trichrome Staining 33 Figure 2.5. Methyl green-pyronin staining of medullary cords and sinuses 35 Figure 2.6. Silver nitrate staining of D N A suspended in an agarose gel. 38 C H A P T E R III Figure 3.1. Single cell gel electrophoresis of human leukocytes 53 Figure 3.2. Single cell gel electrophoresis of porcine skeletal muscle cells 57 Figure 3.3. Regression analysis of D N A fragmentation 58 v A C K N O W L E D G E M E N T S I would like to thank Dr. J.A.J. 'Rex' Ferris for his ongoing support and advice as my senior thesis supervisor. I would also like to thank Drs. Peggy Olive and Judit Banath for their patience and advice regarding the comet assay. Deanna Haskins (R.T.) was of tremendous help in conducting the 'pig experiments'. I am appreciative to Dr. Herbert Geller for generously providing his comet scoring macro software for use with NIH Image, which was integral to the completion of this project. I would also like to acknowledge the Science Council of British Columbia for providing support in the form of a Graduate Research, Engineering and Applied Technology scholarship. Last but not least I would like to thank my husband, Ken Johnson for his patience and support through all this time. vi C H A P T E R I O V E R V I E W 1 C H A P T E R 1 O V E R V I E W 1.1 Introduction When a dead body is found, one of the first questions asked is, "When did this person die?" The answer to this question can be vital in the solving of a homicide case. Determining time since death has long posed a significant problem to forensic professionals, prompting a search for a method which is both precise and accurate. Recently, the U S Appeals Court has set guidelines on precisely what is and is not acceptable in court as scientific testimony, and a number of the time-withstood methods of determining postmortem interval may well fall short of this new yardstick. The utilization of technical advances over the years has produced new experimental methods which are intriguing, yet generally still lacking in some aspect. In this review of methods to determine time since death, the aspects of precision, error, and legal admissibility in the evaluation of the early postmortem interval will be addressed. One of the most important questions to answer in the field of forensic medicine is the determination of time elapsed since death upon the discovery of a possible homicide victim. Pinpointing the time of death is accurate only in detective novels or on television shows like the X Files or Quincy. In reality a window of the time of death is the best that can currently be offered. Determining the time since death or postmortem interval (PMI) can contribute to reconstruction of the crime scene, differentiate between homicide and suicide, pinpoint a suspect, make or break an alibi, and it is an important step in the mourning process of the bereaved relatives of the victim. 2 The physician who determines a time of death may well have to stand up in court at a homicide trial at some much later date and be able to describe how they determined this time of death, and how accurate and precise the timeframe is. This testimony must also be able to stand up to cross-examination and it may be necessary to refute claims to the contrary put forth by expert forensic witnesses for the defense. At the same time, it is important to allow for some imprecision by incorporating a larger time frame which will include the time of death, rather than deciding on a time point that is too restrictive, and which may actually exclude the precise time point of the murder event. In the latter case, the result may remove the actual perpetrator from the suspect list. Some of the methods used, such as rigor mortis or livor mortis, which will be discussed, are also useful in other aspects of the forensic investigation such as determining whether the body was moved after death. There are a number of different methods currently or historically used to determine postmortem interval, however they usually fall within one of two categories; those used in the early post-mortem period and those in the late postmortem period. For the purposes of this review, the early postmortem period (which shall be defined as the first 48 hours following death) will be investigated. 3 1.2 Postmortem Techniques Review This review will focus on determining time since death in the early postmortem period, which is also the timeframe in which the majority of homicide victims are found. The different methods will be described and compared as to their ability to pinpoint time since death and margin of error. Other factors will be considered in determining relative value, such as environmental variability and court admissibility. Probably the most widely used estimate of postmortem interval is based on body temperature. Algor mortis or body cooling estimates time since death by measuring the body temperature of the corpse, usually by comparing rectal temperature to the ambient temperature and referring to a predetermined nomogram of cooling rates to obtain the postmortem interval. The human body cools according to a sigmoidal curve, which can be expressed by a double exponential equation (Henssge C, 1988; Marshall T K & Hoare FE, 1962). Initially the body retains its heat, causing a plateau which can last up to 6 hours, then loses it to the external environment resulting in a lower body temperature (Baccino E et al, 1996). To obtain a relatively accurate estimate, the ambient temperature over the entire postmortem interval must be known, not just the ambient temperature upon discovery of the body. For a corpse found out of doors, this can pose a problem, as the average outdoor temperature will have to be determined. To determine the average outdoor temperature will require a nearby weather station, and paradoxically, the time the body has been present must be known before the average ambient temperature can be calculated. This sort of paradox adds to the variability of this method. The rate of cooling decreases with increasing body size and amount of clothing worn and 4 increases with lowered ambient temperature. One drawback to this method is that it is limited to the very short term, as the body will eventually slow the cooling trend, or even reverse the cooling rate, due to heat production by putrefactive bacteria originating within the gut of the body. An even larger problem however, is the large deviation provided for any time estimate. In the best case scenario, when ambient temperature is known and body weight and clothing factors are corrected for, still only a very basic wide window of time since death can be determined. For example, using Henssge's nomogram method which provides a readily usable table for reference (Henssge C, 1988), in the first 24 hours following death, body cooling provides a time window estimate of approximately 9 hours. Another drawback to this method is that it cannot be applied in cases where the body has been moved after death, which may not be readily apparent thereby leading to an incorrect estimate. Another method has been proposed by Nokes et al. (1992), which utilizes the body temperature at the outer ear or nose, rather than the core temperature which is obtained using rectal measurements. These authors postulate that the ear/nose method may be preferable, since it does not depend on environmental temperatures during cooling or the size of the body, and that it eliminates the need for complex mathematics required in applying the algorithms utilized in the former method. They also state that it eliminates the possibility of disturbing a site of possible sexual assault on the body in suspicious circumstances. In an earlier paper (Nokes L D M , Hicks B, and Knight B, 1986), these authors attempted to determine the time since death based on a cooling profile of a point within the trachea, but found a major source of error involving the 5 misplacement of the temperature probe within the esophagus. In their ear/nose experiment, a sample of bodies with known postmortem intervals were fitted with thermocouples on their outer ear and within the nose, and temperature readings were taken every 15 minutes. With their initial results (5 corpses), the authors devised a hypothetical equation to be applied to determine time since death. This equation interestingly does not include any input of the ambient temperature, and assumes the initial body temperature at time of death is 35.5 degrees Celsius. Using both ear and nose data obtained at 10 hours known postmortem time, the authors attempted to apply this equation to their experimental set, hoping to find a predictor of approximately 10 hours postmortem interval. The actual predicted times for the ear data varied from 10 to 22 hours for a mean error of 5.2 hours, while the nose data predicted from 14 to 36 hours postmortem interval with a mean error of 15 hours. It was concluded that this error was unacceptable for use in determining PMI and suggested that a more complicated mathematical model may be needed to be determined before this method can become more accurate. The effects of different ambient temperatures on the cooling rates were not addressed, although one would expect them to have a significant effect. Baccino et al. (1996) together with Frammery et al. (1990) have devised an equation to determine time since death using outer ear temperature, and hypothesize that tympanic temperature measurement is similar to hypothalamus temperature which is a better representation of body cooling than using the rectum. They have found that using the outer ear, a more accurate estimate can be obtained, and that this method eliminates the problematic postmortem plateau observed with rectal temperature methods. Using an 6 experimental base of 138 deceased individuals, the authors applied both the rectal temperature method and the outer ear temperature method, as well as biochemical methods such as potassium concentration in the vitreous humor and CSF at 4 different stable ambient temperatures, ranging from 0 to 23 degrees Celsius. They found that outer ear temperature resulted in a better correlation coefficient in all cases followed by rectal temperature and then lastly by biochemical methods. The experimental samples in the outer ear temperature group had an average mean error window of time of 479 minutes, or approximately 8 hours. Some limitations to use of the outer ear temperature method are that it cannot be used when there is severe damage to the head, and that the error increases significantly past the 15 hour postmortem period. Postmortem electrical excitability is based on postmortem decay in muscle response when subjected to electrical stimulation. The muscle response is seen to decay with increasing time since death. A number of different authors have reported on this method, including Jones (1995) and Madea (1995). This method employs the insertion of a stimulating electrode, which releases an electric charge into the muscle of the corpse. The stimulus causes the muscle to contract, and the strength of this contraction is then measured and expressed as a subjective amount, or as a percentage of maximum force. The results of an experiment by Jones carried out on 7 rats in this manner resulted in a strong correlation of loss of muscle force with time, which was similar between animals tested and was reproducible. While this appears to be a very precise method for determining time since death, there are a number of limitations and confounding factors to be considered. The method works only within the first 100 minutes, or for about an hour and a half after death, as after this time the muscle response becomes negligible. It 7 should also be noted that the duration of muscle excitability for humans may vary greatly from that of rats. The scale used in the experiment was a relative scale, useful when comparing a number of different known specimens, but not as a predictor on a single unknown specimen. Also, the difference in muscle force between individuals in a human population will likely be much more varied than an animal model test group, with some individuals having less developed muscle fibres than others, resulting in a much different contractile pull, even when all are functioning at maximal force. Thus, any attempt at quantifying percent force exerted on an unknown human specimen would have to be extremely subjective. Variations reported to affect response include body temperature, ambient temperature, individual variations in nutrition, length of agonal period, cause of death (sudden and traumatic versus terminal and lingering), and exercise level or violent struggle (Jones M D et al, 1995). This method experimentally produced correlation with time since death to within about a half-hour period, however it is only viable up until approximately an hour and a half post-mortem. Another method for looking at postmortem interval using electrical response is proposed by Querido (1994), involving changes in the electrical resistance of the intact rat abdomen over time in dead rats. In his experiment, 8 rats were examined for resistance of the abdomen when a constant current was run through 2 electrodes attached to the anterior abdomen at 1, 5, and 24 hours postmortem, then at 24 hour intervals for a total of 504 hours (21 days). Generally, resistance was seen to decrease continually over time, however the variation between animals was so great as to preclude this method from useful practice. For example, one rat measured approximately 60 ohms resistance 8 at 24 hours postmortem, while another measured about the same resistance at 504 hours after death. One must also question the usefulness of the duration of this experiment, since at 21 days after death, the animals would be expected to be extremely decayed and desiccated or too decomposed to be useful for most biochemical forms of analysis. Postmortem mechanical excitation of skeletal muscle is also known as Zsako's phenomenon or idiomuscular contraction (Madea B, 1995). This phenomenon of mechanically induced muscle contraction after death can be observed up until about 10 hours postmortem. It occurs when a muscle is induced to contract by hitting, pinching or tapping a particular area, such as the face, knee, gastrocnemius, foot, forearm, and the back between the shoulder blades. Although the contractions were seen to lessen over time, this method produced extremely variable results and would be difficult to quantify for use as a predictor of time since death. Examination of hypoxanthine (Hx) and potassium levels in vitreous humor and cerebrospinal fluid (CSF) is another method used to estimate PMI. Adenosine is a chemical substance normally produced by the body. In cases of hypoxia (lack of oxygen) such as occurs following death, adenosine degrades to produce hypoxanthine. Following death, the levels of potassium in the body fluids also increase as the adenosine tri-phosphate (ATP) activated physiological sodium/potassium pumps in the cell membranes cease actively pumping potassium into the cell (Alberts et al, 1994). The vitreous humor fluid in the eye is commonly used as a source of obtaining Hx and potassium levels in the body's fluids for estimation of the postmortem interval. Cerebrospinal fluid is also used, 9 although less often. Rognum et al. (1991) conducted a study on the concentrations of Hx and potassium in the vitreous humor postmortem, which showed a strong correlation between their relative biochemical levels and time since death. The hypoxanthine results were seen to provide a better predictor (less variation) than the potassium, and both were found to positively correlate with increased temperature. These experiments were conducted on 87 human specimens and each eye was sampled at two different postmortem intervals, unfortunately introducing the possibility of environmental variability due to previous sampling and exposure to outside air. This experiment was repeated (Madea B et al, 1994;) in an attempt to independently evaluate the results, using 92 bodies but sampling only once from each eye to eliminate possible contamination of later samples. An additional 43 bodies were used to sample at time intervals between 2 and 20 hours postmortem. The vitreous humor was sampled, as well as the CSF for biochemical levels. Potassium was measured using ion sensitive electrodes (described in Madea B et al, 1989), while Hx concentration was determined using high pressure liquid chromatography (as per Rognum TO et al, 1991). Both Hx and potassium levels were found to increase with time, however Madea et al. found that the potassium levels provided a better correlation with less scatter of data points, allowing for a more specific estimation of time since death, which was the opposite of the Rognum et al. study. It should be noted that the authors used the mean values as data points, which eliminates much variation between actual individual samples, and as a result reduced the apparent scatter. Over an approximately 100 hour period postmortem, the hypoxanthine data produced a correlation coefficient of r 2 = 0.714 and a window of 64 hours for any given point. The potassium data, although slightly better, with a correlation coefficient of r = 1 0 0.925, still had a time window of 34 hours for any particular point. It was noted that there was in fact large variation between individuals for the same time period, as well as a variation between samples taken from different eyes of the same individual at the same time, which greatly reduces the usefulness of this technique. The Hx and potassium levels in cerebrospinal fluid also increase with postmortem interval, but this increase is exponential rather than linear. Published results on CSF biochemical levels are confounding, with excessively wide confidence intervals, so as to make this method undesirable for estimating postmortem interval (Madea B et al, 1994; Harkness RA, 1988). Livor mortis, also known as lividity or hypostasis, refers to the pooling of blood due to gravity, in the human body after death. This is observable as a dark red color produced on the skin. The blood pools slowly at first, but becomes thicker and viscous, eventually appearing as a permanent mark that will not leave, even if the body is moved into another position (Knight B, 1995). In places where pressure was applied to the skin during pooling, there will be pale markings, showing the locations where the capillaries were compressed, forcing the blood out. One such example of this would be the soles of the feet and the back of the thighs and buttocks in the case of someone who has died in a sitting position. The property of lividity can be used to determine the position in which the person died, and often whether or not the body has been moved after death. It has also been suggested that lividity can also be used experimentally to determine PMI. One method initiated by Vanezis (1991) utilizes simple colorimetry, whereby the "lightness" value of the skin is determined, and measured at repeated intervals after death. The 11 measurements are taken from a section where the blood has pooled, with the body left in a position opposite of that which the lividity originally occurred, in order to subject the blood to the pull of gravity away from the area. An approximately linear lightening effect was seen to occur over time (Vanezis P, 1991; Vanezis P & Trujillo O, 1996). Unfortunately, the rates of change were vastly different for each individual case, and this method seems unlikely to provide a good indicator of postmortem interval. Another method of estimating time since death using lividity has been proposed by Kaatsch, Schmidtke, and Nietsch (1994). This method uses photometric measurement of pressure-induced blanching of livor mortis. Earlier attempts at using this method have relied mostly on impressions of blanching caused by pressure applied by thumb or forceps, which are highly subjective. The Kaatsch et al. method includes the use of a computer-aided system for the photometric measurement of blanching as a function of pressure and time. It was found experimentally that increased pressure caused increased blanching, and the amount of blanching decreased from 10 hours postmortem until approximately 30 hours postmortem, at which time the effect of applying pressure became negligible. There is no blanching data for the bodies prior to the 10 hour postmortem mark, since no tests were conducted in that period. The data from 10 hours onward approximate a negative exponential plot, but the overlap in values between time intervals is very large, for example in measurements taken at 10 hours and then at 20 hours postmortem, there is approximately 85% overlap of data points. This would correlate to a 95% confidence interval of estimating a time since death within a 30 hour window, which would not be very precise. 12 Rigor mortis or cadaveric rigidity is caused by the cessation of production of ATP following death, which is required for release of the muscle fibres, actin and myosin, allowing relaxation of the binding which occurs in muscle contraction. As the ATP present in the muscle cells is used up, without constant replenishment the muscle fibres remain locked together, giving the body a petrified posture that is virtually immovable (Krompecher T, 1995). Rigor mortis consists of three periods, an initial period where the rigor increases until complete rigidity is obtained, a period comprising the duration of complete rigidity, and a resolution period where the rigor eventually becomes relaxed (most likely due to decomposition effects on the muscle fibres.) In a summary of literature data by Mallach et al. (1964), rigidity was found to be reached anywhere between 2 and 20 hours, could persist from 24 to 96 hours, and resolution could occur between 12 and 192 hours. The periods of time when these phases occur have been used to roughly estimate time since death, however the intervals of each overlap significantly, and are mostly non-plausible for estimating PMI. More recently, Krompecher (1994) has experimented with objective measurements of rigor mortis, whereby a sample group of dead rats were repeatedly tested for resistance by the muscles to successively higher applications of force against them. With tests between 2 and 53 hours postmortem, the 'intensity' values, as determined by the amount of force required to move the leg muscle a certain distance, generally increased to a point around 6 hours postmortem, then continued to decrease after that point. The data produced does seem to provide a correlation, however the variation between specimens and standard deviation within time frames is too large to be 13 practical in using as an estimator of PMI. Also, variation may be expected in human corpses, due to a much larger degree of variability dependant upon body size and muscle development between individuals. Rigor mortis is also known to be strongly affected by heat or cold, with an increase in the rate of rigor with an increase in ambient temperature (Ibid). D N A degradation measured with a flow cytometer is discussed in a paper by Cina (1994). This method hypothesizes that as autolysis and putrefaction take place following death, the D N A will degrade within the cells. The author proposes using a flow cytometer to sort cells according to whether they contain less than the 'normal' 2n amount of D N A or not. This discernment is made in the flow cytometer using a laser beam to illuminate the propidium iodide stained D N A within cells, then an automated cell sorter determines the relative amounts of D N A and sorts each cell accordingly. The idea behind this method is that as decomposition continues with PMI, the proportion of cells with intact D N A will become less, and that this proportion should be quantifiable and will be able to be used as a predictor in unknown specimens. One would expect, however, that D N A would degrade uniformly within any particular tissue, resulting in an 'all ' (2n) or 'none' (less than 2n) amount in all cells measured at the same time. This method may be conducive to a determinant of whether death occurred prior to or following a particular time point, but not very useful as a precise PMI estimate. An interesting idea, which is worth pursuing has been put forth by Baccino et al. (1996). These authors suggest using combinations of different methods to minimize error 14 and optimize the result. For example, it is suggested that in cases where the ambient temperature is high, body cooling based estimates will be low due to slower cooling rates, while a biochemical method such as potassium concentration would suggest a higher than actual rate of decomposition. Therefore, if both of these methods are used and the results averaged, a more accurate result can be obtained. In theory, this method should compensate for individual method errors, resulting in a more precise and accurate PMI estimation. 15 1.3 Legal Issues In 1993 the U.S. Court of Appeal gave trial judges more authority to disregard testimony from scientists that do not meet strict tests of scientific validity. This is the outcome of the decision of Daubert v.Merrel Dow Pharmaceuticals, in which the Supreme Court called for trial judges to act as "gatekeepers" and screen out unreliable scientific testimony. The court stated that judges should use four specific criteria to judge admissibility: empirical testability, peer review and publication, rate of error of-a technique, and its degree of acceptance (Kaiser J, 1998). This ruling has allowed for the inclusion of certain technologies, such as D N A evidence, but will more often exclude testimony which is deemed to lack scientific validity. This is a hot topic for debate within the scientific community, in that many consider it to be a step backward in the court treatment of scientific evidence, while others might say that it makes scientists accountable for their work and should increase the level of professionalism. Previously, expert scientific testimony had been subject to a ruling made in 1923 over Frye v. United States, in which it was decided in a Court of Appeals for the District of Columbia that scientific testimony to be entered into evidence must pass the Frye standard (Bohan TL & Heels EJ, 1995). This stated that the "scientific principle or discovery from which a deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs" and this has been the precedent until today. Thus with the Daubert decision, the expert testimony on such subjects as time of death may no longer simply be accepted at face value. It will have to be proven scientifically, quantified, and the method used will need to have high accuracy, low error and be reproducible. 16 In light of this, future methods for determining time since death should look toward quantifiable testing that has strong scientific merit, and is accurate, with measurable statistical qualities such as confidence intervals, standard deviation, standard error and levels of variance. It must also be reproducible and should be based on a fundamental hypothesis that makes a clear statement that by definition can be disproved if invalid, as such is the definition of science. 17 1.4 Conclusions There has yet to be developed a reliable method by which to pinpoint PMI within less than approximately a 9 hour time frame, within the first 48 hours postmortem. Some methods that have been described above show certain strengths. The visual estimation of lividity or rigor mortis, temperature recording via the outer ear or even the rectum are very simple to do at the scene, with minimal or no equipment. Advantages to other techniques, such as measurement of vitreous humor Hx and potassium levels, or D N A degradation levels are that they are quantifiable, therefore lend themselves better to statistical analysis, and may tend to stand up better in court. The drawback to all of these methods is that they are extremely imprecise with very wide deviations and/or large errors involved. Most are only reliable in determining whether a murder may have occurred during a particular day, or possibly differentiating between events that occurred during the day or at night. More research needs to be done to determine a scientifically sound method of determining time since death with more precision, accuracy, and less error. The recent advances in molecular biology and microscopic imaging techniques may be useful in evaluating more about the physical occurrences, and thus the properties of cells and tissues in response to death and the process of decomposition, and should be investigated further. 18 C H A P T E R II E X P E R I M E N T A L DESIGN 19 CHAPTER II EXPERIMENTAL DESIGN 2.1 Cellular Biology Introduction In order to explore the biophysical properties which occur at the cellular level during the early postmortem interval, one must first examine basic cell biology. Each cell in the body contains a nucleus, which holds the genetic blueprint for each organism in the form of long, wound strands of deoxyribonucleic acid (DNA). These strands of D N A are physically bound to each other in a helical manner, physically resembling a ladder, joined together at the rungs, as shown in figure 1 below. There are known to be inherent nucleases within cells, held in check during normal development, to be released in cases of induced cell death, such as apoptosis or during necrosis (Majno G & Joris I, 1995). 20 one helical turn • 3.4 nm major groove minor groove Figure 1. D N A Double Helix. In a D N A molecule two anti-parallel strands that are complementary in their nucleotide sequence are paired in a right-handed double helix with about 10 nucleotide pairs per helical turn. A schematic representation (top) and a space-filling model (bottom) are illustrated here (from Alberts etal., 1994). 21 2.2 Experimental Hypothesis It is reasonable to propose that following organismal death, during the breakdown of the intracellular membranes and organelles, the intracellular nucleases are released and become free to break down the DNA. Exonuclease will cause shortening of the D N A strand by cleaving nucleotides one by one from the end of the strand, while endonuclease will cause the D N A to fragment by cleavage of phosphodiester bonds with in the nucleotide chain. One could further predict that as the D N A becomes further digested by these enzymes, the strands of D N A will become shorter and shorter as the break points and cleaved terminal nucleotides become more numerous. It is not known in this case whether these enzymes act specifically in a sequence-directed manner, as do restriction enzymes, or i f the nuclease digestion they will cause is a random event. It should be possible to test this hypothesis, if a method can be found by which single-stranded D N A breaks can be recorded and measured and compared to double-stranded D N A breaks. If the above hypothesis holds true, the strands of nuclear D N A should become measurably smaller as time since death increases. If single-stranded D N A samples measure at the same size as double-stranded D N A samples when taken from the same post-mortem time point, one can assume that the endonuclease activity is sequence-specific, cutting the D N A at the same locations on both strands equally. If there is a large difference in size between the single and double-stranded fragments, one can assume a random D N A degradation, which cuts the D N A unevenly. 22 There is a method by which cellular D N A can be examined biophysically, on a single-cell basis. This method is called single-cell gel electrophoresis, also known as the comet assay. This assay can determine proportions of fragmented versus non-fragmented DNA, as well as measure relative lengths of either single-stranded or double-stranded DNA. In order to utilize this method, a quantitative DNA-specific stain must be used. For the purposes of this study, a permanent stain would be desirable in order to store slides long-term for future medico-legal purposes. If no permanent stains are available, a temporary stain could be utilized as long as the slides are fixed to cease further D N A degradation, in which case they can be stored and re-stained at a later date. If a sequential process of D N A degradation can be shown and quantified, this method may provide an alternative way to determine time since death in unknown cases. Hypothesis: Nuclear D N A will degrade increasingly with time since death, the rate of degradation will be a measurable constant, which can provide a potential predictor of post-mortem interval in unknown cases. 23 2.3 Quantitative Staining of D N A for L igh t Microscopy In light microscopy, unstained tissue and cell samples are virtually invisible to the human eye. In order to visualize, different stains and dyes are applied to the sample which combine in specific ways to produce observable results. A number of stains bind tissues according to their biochemical make-up, attaching negatively charged dye molecules to positively charged cell components and vice-versa. Because of this, certain tissue types can be differentiated according to their physical charge properties, such as the binding of the positively charged stain, hematoxylin to negatively charged D N A to identify nuclei within cells. Other stains bind covalently to certain cell components, allowing for a more specific dye preparation, depending on molecular constitution. A number of structural stain features are important in determining the interaction between stain and substrate, these are i) the electric charge carried by a dye, ii) size, as represented by ionic weight, iii) number of covalent bonds involved in the conjugated systems of a dye, and iv) hydrophobic-hydrophilic nature of a dye (Stevens A & Bancroft JD, 1990). There are numerous methods and approaches with equally many varying results to histological staining, for identification of pathologies, physical examination of cell ultrastructure, and examination of foreign substances within the tissues amongst others. A number of stains have been seen to stain quantitatively, showing a gradation of specific colouration dependant on the amount of binding substrate available, thereby indicating an amount of substrate present in any given preparation. In this paper, methods for examining quanitative staining for D N A under the light microscope will be reviewed and evaluated. 24 D N A is deoxyribonucleic acid, consisting of a chain of nucleotides composed of a deoxyribose sugar, a base (adenine or guanine (purines), or thymine or cytosine (pyrimidines)), and a phosphate group which produces a net negative charge. These nucleotides bind together at their bases using hydrogen bonding, which results in a ladder-like structure with base rungs and sugar-phosphate sides which is referred to as a helix-formation. These different properties of D N A are the deciding factors of the way different dyes will or will not bind. D N A is found mostly within the nucleus, while RNA is found in the ribosomes (cytoplasm) and nucleoli of the cell. Ribonucleic acid (RNA) is composed similarly to DNA, except for a 2' hydroxyl group on the sugar, and in the base, uracil is substituted for thymine. Demonstration of D N A by staining is dependent on the interaction of dyes with the phosphate groups, intercalation of the dye into the duplex or production of aldehydes from the deoxyribose sugar (Stevens A & Bancroft JD, 1990). Pure nucleic acids can be stained with basic dyes, but nucleic acids in living cells are bound with protein are not available to react. Different fixation formulae will affect the D N A / cellular chromatin differently, for example, inactivation of the protein-bound amino groups by formalin will increase the basophilia of the nucleic acids by releasing increasing numbers of nucleic acid phosphoryl groups to bind a basic dye (Presnell JK & Schreibman MP, 1997). Sucrose-formalin will produce strong formation of nuclear fluid into chromatin threads, while alcohol-formalin will cause this less so, and acetic acid or Carnoy's fixative will result in the weakest precipitation of chromatin (McManus JFA & Mowry RW, 1965). Thus some sort of tissue fixation may be desirable in making a stained preparation. Nuclear chromatin stains well with a number of basic dyes as will be examined in this chapter. 2 5 Techniques and Methods Hematoxylin There are many varieties of hematoxylin stain, but all have certain things in common. A l l are originally derived from an extract of the heartwood tree, Haematoxylon campechianum. This extract, hematoxylin, itself is not a dye. It is the oxidation product of hematoxylin, hematein which produces the desired acidophilic staining property. This oxidation can be accomplished in one of two ways: "naturally", using a slow process of exposure to air for 3 to 6 weeks, (as in Heidenhain hematoxylin) or artificially by the use of mercuric oxide, hydrogen peroxide, or another oxidizing agent (as in Harris hematoxylin) (Presnell JK & Schreibman MP, 1997). On its own, hematein does not demonstrate good binding affinity for tissues, and must therefore be combined with a mordant, in the form of a metal salt such as those derived from aluminum, iron or tungsten, (Stevens A, 1990; Mallory FB, 1944). The resulting dye becomes basic in form and will bind the negatively charged acidic nuclear elements of D N A . Hematoylins are generally applied in conjunction with mildly basic solutions, resulting in a fairly specific D N A stain exhibiting a blue to black colour. R N A in cytoplasm and nucleoli also stain blue with this method. This method is commonly used in conjunction with eosin to provide a thorough staining of many structures within tissues and cells, since each binds to either positive (eosin) or negative (hematoxylin) charged particles indiscriminately of molecular size. See figure 2.1 for an example of slides prepared using hematoxylin & eosin. 26 Figure 2.1. Eosin & Hematoxylin Nuclear Staining. Example of hematoxylin & eosin stained connective tissue, note the darkened cell nuclei (from Burkitt H G , Young B and Heath JW, 1993). Original magnification 320X. 27 Feulgen Staining The Feulgen reaction originally described by Feulgen & Rossenbeck (1924) utilizes the properties of DNA's deoxyribose sugar to visually demonstrate DNA. In this technique, mild acid hydrolysis is used (HC1 at 60 degrees C), to break the purine base-deoxyribose bond, resulting in exposure of the deoxyribose portion of the D N A in an aldehyde form. These aldehydes are differentially visualized with a step including a "Schiff s reagent", an acid solution containing a sulphur dioxide with a variable dye such as pararosanilin, thionin, azure-A, or acridine orange. Acridine orange can be used as the Schiff s reagent for light microscopy, as well as being a fluorescent stain. The Schiff s reagent step discolours the original dye being added, but when exposed to bound aldehydes on the DNA, recombines to form a new colouration (Stevens A & Bancroft JD, 1990). Generally, the D N A appears from blue to red-purple with the cytoplasm colourless to light green (depending on possible counter-staining steps with varying dyes). When acridine orange is used under a fluorescence microscope, D N A appears yellow, while R N A is stained red-orange (as shown in figure 2.2 below). This method is useful in differentially dying D N A and not RNA, since in the hydrolysis step, the ribose-purine bond is unaffected and so R N A is not exposed to aldehyde formation. In the Feulgen reaction, the most important step is the acid hydrolysis, and as such the hydrolysis time should be determined according to which fixative has been used to preserve the specimen (Ibid.). One possible drawback to Fuelgen staining is in the event of excessive hydrolysis, degraded nucleic acids will diffuse out from the tissue. 28 Figure 2.2. Acridine Orange Nuclear Staining. Fluorescence micrograph of a bone marrow smear, stained with acridine orange. DNA (cell nuclei) appear green or yellow, while RNA (cytoplasm) appears red-orange (from Burkitt HG, Young B and Heath JW, 1993). Original magnification 320X. 29 Methylene Blue Methylene blue is a strongly basic dye, and therefore can be used as a nuclear stain. Composed of tetramethylthionin, and azure B trimethylthionin, it shares similar properties with Azure B (Boon M E & Kok LP, 1986), which will be discussed later for it's use in Masson's trichrome stain. At lower (i.e. room temp.) temperatures, methylene blue tends to bind multiple structures, although at high temperatures is found to bind specifically to nucleic acids, producing a blue colour (Ibid.). This is believed to be due to the increased accessibility of the dye molecules to the nucleic acids as the two strands become denatured and separate at increasing temperatures. Romanowsky-Giemsa (RG) Staining This technique utilizes a pair of stains, Azure B (cationic) and Eosin Y (anionic) as a general-purpose stain. One basophilic and one acidophilic component produce a thorough visualization of many cell structures at once, similar to the hematoxylin-eosin method. Nuclear chromatin is stained purple, which is fairly unusual amongst histological dyes (Boon M E & Kok LP, 1986). The acidophilic dye, Azure B is a sub-component of the stain methylene blue, which acts through similar mechanisms. R G staining is frequently used in blood film preparations, since it strongly shows the eosinophilic granules present in many leucocytes, as well as the nuclei. Azure A is also an acidophilic stain, but much less so and therefore is not as useful for general purposes. See figure 2.3 for an example of Giemsa staining. 30 Figure 2.3. Giemsa stained slide of bone marrow. Note the cell nuclei, which stain darkly (from Burkitt HG, Young B and Heath JW, 1993). Original magnification 640X. 3 1 Masson's Trichrome Another general cellular staining technique is Masson's trichrome. This technique utilizes phosphotungstic (PTA) or phosphomolybdic (PMA) acid. Pretreatment of a sample with PTA or P M A causes a differential binding and colouration effect in a standard dye (such as Methyl blue) added in a later step. When a section is first treated with a levelling dye or other suitable small molecule anionic dye and then with P M A or PTA solution, the P M A or PTA competes with the dye and gains access to collagen easily, expelling the dye in the process. If this treatment is stopped at any given moment, the collagen is free to stain with a different large molecule dye added later (Bradbury P & Gordon K C , 1990). Nuclei are seen to stain blue-black, as seen in figure 2.4 below. 32 Figure 2.4. Masson's Trichrome Staining. Masson's trichrome stained slide of connective tissue (collagen) note the darkly stained cell nuclei (from Burkitt HG, Young B and Heath JW, 1993). Original magnification 320X. 3 3 Methyl Green In methyl green staining, the D N A phosphate (negatively charged) group combines with this basic dye at an acid pH. In its original state, methyl green is an impure dye containing methyl violet. The methyl violet can be removed by washing with chloroform, resulting in a stain which when used at a slightly acid pH binds specifically to D N A (excluding RNA). The commonly accepted procedure for D N A staining with methyl green is the methyl green-pyronin method, where a mixture of the two dyes are used, the pyronin differentially staining the RNA. The D N A stains green-blue, and the RNA stains red (Stevens A & Bancroft JD, 1990). It has also been reported that the synthetic stain malachite green dyes nucleic acids similarly to methyl green, but is stable and free from contamination by metachromatic impurities (Gurr E, 1971). See an example of methyl green-pyronin staining in figure 2.5. 34 Figure 2.5. Methyl green-pyronin staining of medullary cords and sinuses, note darkened cell nuclei (from Burkitt HG, Young B and Heath JW, 1993). Original magnification 480X. 35 Gallocyanin chromalum This technique relies upon the combination of the nucleic acid phosphate group with gallocyanin at an acid pH. Unspecific to D N A or RNA, both are demonstrated with this method, and therefore to specifically identify DNA, the R N A must be extracted, usually with the use of RNAase (Stevens A & Bancroft ID, 1990). Both the D N A and RNA stain blue. Initially a time-intensive staining procedure, it has now been refined to be carried out within one hour and the dye is reported to bind quantitatively to the nucleic acids (Husain O A N & Watts K C , 1984). Gallocyanin chromalum has a relatively short shelf life compared to other stains, lasting only about 4 weeks. While this would be a fairly simple, quick staining procedure, it is a slight drawback that R N A must first be completely digested, or else it will contaminate results when quantifying DNA. Silver nitrate In silver staining, positively charged (AgN03) silver molecules are covalently attached to the negatively charged D N A molecules in a quantitative fashion (Gottlieb M and Chavko M , 1987). Silver staining of D N A is discussed in a paper by Gottlieb & Chavko (Ibid.). When complexed with sodium carbonate, silver nitrate, 1% tungstosilicic acid plus formalin plus ammonium nitrate and applied to D N A suspended in an agarose matrix, the silver stain bound quantitatively to the nucleic acids. This method has been shown to quantitatively bind even very small amounts of DNA, less than 5 ng total (Somerville L L & Wang K, 1981; Marshall T, 1983). Compared to the commonly used method of applying ethidium bromide (which is a fluorescent stain that intercalates into the D N A strands), the silver stain method was found to be more sensitive, 36 densitometrically quantitative, and preferable for long-term storage, as fluorescent stains can fade (Ibid.). Under a light microscope, silver-stained cell nuclei appear opaque based on density-dependence, with the areas containing more D N A staining more darkly than those which have less concentrated nuclear material. Figure 2.6 shows single-stranded D N A in an agarose gel, stained with silver nitrate. This method of staining should stain D N A specifically, quantitatively, and be useful for long-term storage of slides without change, as would be desirable for use in the comet assay. 37 IP Figure 2.6. Silver nitrate staining of D N A suspended in an agarose gel (from Gottlieb M and Chavko M , 1987). 3 8 Propidium iodide The use of propidium iodide as a nuclear stain is discussed in a paper by Yeh C-JG, Hsi B-L , and Faulk WP (1981). Propidium iodide is a DNA-specific stain, which fluoresces under 540nm wavelength light. PI stains quickly with minimal preparation, and therefore may be useful for fast determination of D N A quantification. However it requires the use of a fluorescent microscope to visualize, and once cell preparations are exposed to light, they 'quench' or fade. Due to this quenching effect, the same slide may not emit the same intensity of stain at different times. Also, cell preparations stained with PI must be kept moist. Once dried, the slides are unuseable, therefore not desirable for cases requiring long-term storage or repeated use. Propidium iodide is known to intercalate into the D N A of cells, providing a quantitative, nucleic-acid specific stain (Krishan A, 1975; van Rood JJ, van Leeuwen A and Ploem JS, 1976; Hudson B et al, 1969). In order for propidium iodide to access the nuclear chromatin, the cell membrane must be compromised. This stain has been suggested for use in differentiating between live and dead or apoptotic and necrotic cells, based on the ability to identify 'leaky' membranes (Allen RT, Hunter WJ & Agrawal DK, 1997; Darzynkiewicz Z et al, 1998; Lincz LF , 1998; Kroemer G et al, 1995; Majno G & Joris I, 1995). Since the proposed comet assay lyses the cellular membranes prior to staining, the PI method should intercalate all nucleic acids and provide a quantitative stain, regardless of the cause of cell death (ie. apoptotic or necrotic). 39 Discussion & Conclusions Techniques such as hematoxylin-eosin staining and Romanowski-Giemsa staining are widely used and are useful in that they provide a good view of many cell structures at the same time on a slide. However, for D N A quantification, these methods provide too much background colouration and usually stain the RNA indiscriminately in addition to the DNA, causing quantification difficulties. To be useful as a quantitative D N A stain, a dye should bind D N A solely, and the amount of stain taken up should correlate with the amount of D N A present. To quantify DNA, the results must first be reproducible with relative accuracy, so that the variation between samples is significantly less than the variation between different lots of the same sample or dye lot. The amount of stain bound must also be observable either as occupying a particular measurable area on the slide, and it should be quantifiable by intensity of colouration. Since human error can be significant when attempting to calculate microscopic areas, and colour densitometry is an extremely subjective task, an applicable computer scanning program would be the preferable method of data collection and analysis. Desirable quantitative D N A dyes should also exclude demonstration of other structures, since most computer analysis will be confounded by any other subject in the field of view, the program "assuming" that any artifact will be stained DNA, and attempting to analyze it as such, throwing off the results. In a paper by Mikel U V & Bakker RL (1991), a number of different staining procedures were tested for their quantitative ability as well as their reproducibility and statistical variations. The Feulgen reaction was used with four different Schiff s reagents: pararosaniline, thionin, azure-A and acridine orange. The gallocyanin 40 chromalum stain was also tested. A l l were tested by light microscopic procedures, while acridine orange and propidium iodide were also tested by fluorescence microscopy. A l l of the methods were found to quantitatively stain DNA, with reproducible results, however, the acridine orange was found to provide the lowest coefficient of variation, both in the light microscopic tests and in the fluorescent determination. While this study is certainly worth taking note of, the results should not simply be accepted without reproduction, since within each staining method utilized, there are literally dozens of variations on solution formulations, fixative treatments and different resolution times for each step, which can make any particular stain act more or less favourably, while only one variation was tested for each stain here. Probably most useful is the information that all of the above treatments did indeed quantitatively stain for DNA. Certain procedures that have been discussed here, such as hematoxylin-eosin staining, Romanowski-Giemsa and Masson's trichrome should be ruled out for use in D N A quantitation, simply because they are too general and bind all basophilic particles within the cells. It is this very property which makes them good all-purpose stains, however in D N A quantification, they would produce too much background, and may not be thoroughly quantitative in their staining. The same could be said for methyl green-pyronin on the basis of too much background production, with pyronin demonstrating RNA. One of the more promising methods for D N A quantification seems to be the Feulgen reaction, with any of the variations of the Schiff s reagent previously discussed, 41 however this method seems to be somewhat reagent intensive and time-consuming. A major benefit of this method is that the stain specifically binds only DNA, and it does so quantitatively. The gallocyanine chromalum staining method would also seem to be a likely candidate for good laboratory D N A quantification, since it is an easy preparation, and fairly non-time consuming, producing quantitative results. The drawback to this method is the requirement for prior removal of RNA, since the stain does not differentiate between the deoxyribonucleic acid and ribonucleic acid. Both methods are purportedly repeatable with reasonably high accuracy and simplicity to conduct in the laboratory. An alternative would be to use a fluorescent dye such as propidium iodide or Acridine orange, which would be a fast and easy preparation, however access to a fluorescent microscope with visual recording equipment is required. Silver staining, as described by Gottlieb M & Chavko M (1987) may also be a favourable choice, providing a D N A specific, quantitative, permanent stain. In theory there is no concern with background staining nor quenching, which can be a concern with fluorescent dyes, plus the single-stranded method is highly sensitive to a small amount of DNA. 42 2.4 Single Cell Gel Electrophoresis (Comet Assay) Single Cell Gel Electrophoresis (SCGE) also known as the Comet Assay is a sensitive and specific method which is used for biophysical analysis of D N A fragmentation at the single cell level (Fairbairn DW, Olive P & O'Neill K L , 1995). Historically, this method has been used to examine the extent of D N A fragmentation induced by apoptosis following various chemo- or radio-therapy techniques designed to initiate cell death (Olive P, Banath JP & Durand RE, 1990; Anderson D, Wright J & Ioannides C, 1998; Huang P & Durand RE, 1998). To utilize this method, a single-cell suspension must be made of the desired tissue sample. These cells are then suspended in an agarose matrix, which is coated over a microscope slide. The slide is sequentially immersed in a buffered solution designed to cause lysis of the cell membranes and then rinsed in electrophoresis gel running buffer. After being placed in a horizontal gel electrophoresis chamber, a current is applied, which causes the negatively charged D N A from the cell nucleus to migrate through the gel. Smaller fragments of D N A will move quickly through the gel, while larger fragments travel more slowly. The result, once stained with a D N A specific dye is visible with a compound microscope, in the shape of a comet. The 'head' of the comet contains the long, uncut nuclear DNA, while any fragmented D N A pieces will form the 'tail' of the comet. For full comet assay protocols, see chapter III. 43 2.5 Conclusions To determine the rate of D N A degradation in the postmortem interval, different tissues should be collected at specific times following death. These samples should be treated in such a way that further D N A degradation will not occur (ie. freezing or immediate evaluation of D N A damage). Chemical fixation (as with formalin) should be avoided, as the formalin crosslinking of the D N A fragments will hamper fragmentation analysis. The comet assay (single-cell gel electrophoresis) should be used to reveal the amount of D N A fragmentation which has occurred, when stained with a D N A specific, quantitative stain, such as propidium iodide (for convenience and rapid results) or silver-stain (for a permanent record). In silver staining of comet assay slides, protein contamination contributing to background staining is not a concern. In the membrane lysis step of the comet assay procedure, strong detergent is added, such as sodium dodecyl sulphate or n-lauroyl sarcosine. Contact of the exposed intracellular proteins with the detergent molecules causes them to be lifted out and into solution. The stained, electrophoresed D N A can then be imaged on a microscope using a digital camera. Once micrographs of each individual cell's D N A have been captured, they can be analyzed and quantified using image analysis software. In this way, the hypothesis that nuclear D N A will degrade increasingly with time since death, the rate of degradation providing a measurable constant with the potential use as a predictor of post-mortem interval, can be evaluated. 44 C H A P T E R I I I E X P E R I M E N T A L P A T H O L O G Y 45 C H A P T E R III E X P E R I M E N T A L P A T H O L O G Y 3.1 Introduction In the investigation of homicide, one of the persistent problems that has long plagued forensic pathologists has been the determination of time since death, or postmortem interval (PMI). Pinpointing the time of death in the immediate postmortem interval is currently so imprecise that only a wide (8-hour) window of estimation is the best that can be obtained (Henssge C, 1995). Determining the PMI can contribute to the reconstruction of a crime scene, differentiate between homicide and suicide, pinpoint a suspect, make or break an alibi, and is important for the mourning process of the bereaved relatives of the victim. Although much is known about cell death as caused by apoptosis and the mechanisms involved (Tang, D G & Porter A , 1996; Majno G & Joris I, 1995) little is known about the process of nuclear and cellular degradation following organismal death. It has recently been shown by evaluation through image cytometry (Payne PW et al., 1999) that following death, nuclear chromatin undergoes specific morphological changes as time progresses. The application of single-cell gel electrophoresis (SCGE), also known as the comet assay, has been well documented for it's use as a tool to evaluate the physical characteristics of D N A damage within cells, brought about by various chemotherapeutic and radiation-induced treatments (Fairbairn DW, Olive P, O'Neill K L , 1995; Olive PL, 46 Banath JP, & Durand RE, 1990; Anderson D, Wright J, Ioannides C, 1998; Gutierrez S, etal, 1998; Huang R & Durand RE, 1998; Ross G M , Wilcox P, Collins AR, 1995; McKelvey-Martin V J et al, 1998; Pjuhler S & Wolf H U , 1996). As the chromatin within the nucleus becomes damaged, breaks occur in the D N A strands, resulting in shorter D N A fragments. D N A in situ is formed of two anti-parellel strands of deoxyribose sugar, plus phosphate backbone. The two strands of D N A are bound together by hydrogen bonds between the attached base groups (adanine, cytosine, guanine and thymine). When these D N A strand breaks occur, they can result in 1 of 2 forms. Single-stranded breaks occur in the cases of relatively little damage. While each strand may be nicked, the overall DNA helical structure will be maintained. In order to visualize these D N A fragments, the two D N A strands must be separated first. In the case of increased D N A damage, nicks will become more numerous and may occur in the same location on both sides of the D N A helix. In this case, the helical structure itself will form separate, discrete fragments, with the bases still bonded together. Suspending cells in an agarose matrix, and running an electric current through them causes the smaller fragments of D N A to travel further through the gel, away from the larger strands of DNA, which remain within the vicinity of the cell nucleus. When stained with a quantitative D N A specific dye, the result of this electrophoresis is visible microscopically, in the shape of a comet. The head of the comet represents the long-stranded D N A nucleus of a single cell, while the comet-tail is composed of the smaller 47 pieces of fragmented D N A which have electrophoresed down the gel. Smaller pieces travel further, creating a longer tail, while increasingly numerous fragments result in a comet-tail which is more dense. There are two commonly used quantitative measures of D N A damage used in conjunction with the comet assay; tail length and the tail moment. The tail length is a relative measurement of fragment size, whereby the smaller the fragment, the larger the tail length, while the tail moment is a calculation based on the proportion of D N A in the tail of the comet relative to the proportion of D N A in the comet head [5,13]. We have applied this technique to the quantification of D N A degradation in two different models, a human blood (leukocyte) model, and a porcine animal model, in the early (0-72 hours) PMI. 48 3.2 Materials and Methods For sample collection of human leukocytes, whole blood was drawn fresh from a human subject in heparinized tubes. The blood was suspended in freezing solution (15% DMSO, 1.8% w/v NaCl, in Dulbecco's RPMI 1640 (Sigma)). Hypoxia induced by removal of the blood from the body was used to model the effects on white blood cells in the body following death. In vitro at 25 degrees Celsius, one blood sample was collected after 2 hours as an indication of early PMI effects, while a second sample was collected after 22 hours, comprising a late PMI period. A l l samples were then flash-frozen and stored in liquid nitrogen at -170 degrees Celsius to prevent any further nuclease activity. Blood samples suspended in freezing solution were thawed in 37 degrees Celsius water bath. For electrophoresis, leukocyte density was determined by staining with thionin and counting on a hemacytometer. Approximately 5,000 - 10,000 leukocytes were used per slide electrophoresed. To further the study, Sus domesticus , the domesticated pig, was chosen as an animal model due to its similarity to human body size and basal metabolic rate. 24 specimens of Sus domesticus with an average weight of 85 kg were obtained freshly sacrificed from an abattoir, and stored in a temperature-controlled environment at 15 degrees Celsius, the temperature continuously monitored using temperature data-loggers. Using standard dissection equipment, 1 cc samples of skeletal muscle, heart, liver and kidney were taken at approximately 8-hour Intervals, starting at 3 hours and continuing until 72 hours postmortem. The samples were placed into 2 mL cryovials containing 1 mL of freezing solution (plus 10% FBS) for 10 minutes, then flash-frozen and stored in 49 liquid nitrogen. Porcine tissue samples were thawed in a 37 degrees Celsius water bath and placed into 5 mL test-tubes in 2 mL IX PBS (Sigma). Using dissecting scissors the sample was sheared for approximately 10 minutes on ice, forming a homogenous cloudy solution. For muscle tissues, the cells were filtered sequentially through 160 and 100 micron nylon mesh, heart through 160, 100 and 37 micron mesh and liver and kidney through 160 micron mesh. A l l samples were treated identically to reduce the effects of any artefactual D N A damage induced by processing. For electrophoresis and staining, the alkaline comet assay was used to evaluate single-stranded breaks in DNA. In this method, the strands of D N A are separated prior to electrophoresis. This allows viewing of relatively few D N A damaging events. Approximately 10,000 cells were suspended in 1% liquid low temperature agarose in IX PBS and then pipetted onto a pre-coated microscope slide. The agarose was set at 4 degrees Celsius for 5 minutes. The slides were washed in lysis buffer (0.03M NaOH, 1.2M NaCI, 0.1M N-lauroylsarcosine) while covered for 1 hour and then rinsed in an alkali buffer solution (0.03M NaOH, 2mM EDTA) for 1 hour, using 3 changes of solution. Using a large (1.5L) horizontal submarine electrophoresis chamber, the slides were electrophoresed at 0.6V/cm for 25 minutes at 40mA in fresh alkali buffer solution. Slides were then rinsed in distilled water and stained with propidium iodide at 2ug/mL for 20 minutes. The neutral comet assay was used to evaluate double-strand breaks in D N A (modified from Cerda et al, 1997). In this assay the D N A is relaxed, but the strands 50 remain bound together, allowing for viewing of a large amount of D N A damage. Approximately 5000 cells were suspended in 200 uL 1% liquid low temperature agarose in I X PBS and pipetted onto a microscope slide, covered with a coverslip and chilled at 4 degrees Celsius for 10 minutes. The coverslips were removed and the slides were rocked in lysis solution (2.5% SDS in 0.5X TBE) for 30 minutes. The slides were then rinsed in a conditioning solution of 0.5X TBE for 30 minutes with 2 changes of solution. Using a large (1.5L) horizontal submarine electrophoresis chamber, the slides were electrophoresed at 2.0V/cm for 2 minutes in fresh 0.5X TBE, rinsed in distilled water for 5 minutes, and finally silver stained using the BioRad Silver Stain Plus kit as previously described (Cerda et al, 1997). For comet imaging and analysis, silver stained slides were viewed on a standard light microscope, while propidium iodide stained slides were viewed on a fluorescent microscope, using a 20X objective lens. Comet images were captured using a Pixera digital camera attached to the microscope, and analyzed on computer using NIH Image software with a comet scoring macro. With this software, size and density of the comet head and tail are measured, and assigned tail-length and tail-moment values accordingly, as previously described (Bocker W et al, 1997). Statistical analysis was conducted by evaluating the mean values of comet-tail length and comet-tail moment of 30 randomly chosen comets per sample. Experimental replicates were averaged for a given time point. 51 3.3 Results On any particular slide, there was an approximately normal distribution of comet sizes. If reproduced graphically, the comet tail-moments and tail-lengths measured on each slide would appear as a bell-curve, with the peak of the bell-curve representing the mean value. For any given sample, a few round comet-heads were visible without comet-tail. Likewise, a few fully 'apoptotic' looking comets were also visible, with very little D N A in the nucleus, and almost a full tail-moment. However, in each case the comet measurements were utilized by randomly selecting 30 comets per slide and taking and average of tail-moment and tail-length. The SCGE of human blood showed an increase in the D N A fragmentation from 2 to 22 hours after removal from the body, as shown by the change in comet shape seen in figure 3.1. In samples from the early period (2 hours), the comet appears mostly as a single circular shape. This represents the majority of long-stranded D N A remaining in the nucleus of the cell, with very little fragmentation. In samples from the later period (22 hours) an increase in the length and density of the comet-tail is seen, representing a decrease in fragment size and an increase in the number of fragmented pieces of DNA. As the D N A becomes increasingly degraded, the fragments travel out of the nucleus and along the gel, carried by the electrical current. At the same time, the nucleus is seen to decrease in size and density, due to the loss of long stranded D N A by fragmentation. If a quantitative D N A stain such as propidium iodide or silver nitrate is used, the quantity of fragmented D N A that has migrated can be compared to the quantity of long stranded D N A remaining in the nucleus as measured by the comet-tail moment. 52 Figure 3.1. Single cell gel electrophoresis of human leukocytes showing an increase in comet-tail size over time after induction of hypoxic conditions. Samples taken at A) 2 hours after removal from the body B) 22 hours after removal from the body. Slides stained with propidium iodide, original magnification 200X. 53 Figure 3.1. Single cell gel electrophoresis of human leukocytes showing an increase in comet-tail size over time after induction of hypoxic conditions. Samples taken at A) 2 hours after removal from the body B) 22 hours after removal from the body. Slides stained with propidium iodide, original magnification 200X. 5 3 The double-stranded (neutral) comet assay was also run on these blood samples, which also showed an increase in D N A fragmentation. As expected, when testing for double-strand versus single-strand breakage, the variation between comet sizes produced by the double-strand (neutral) assay for 22 hours was much less than that seen in the single-strand (alkaline) assay. This indicates that the single-strand method is a more sensitive method, which should be used when examining low amounts of D N A damage, such as those seen in very early time points, while for extensive D N A fragmentation (later time points), the double-strand assay is recommended. Samples of porcine skeletal muscle tissue taken from 3 to 72 hours postmortem were initially evaluated using the alkaline comet assay. The resulting comets formed showed full comet-tails with minimal D N A remaining in the nucleus, even in the earliest samples. This indicated a large amount of single-stranded D N A cleavage occurring early on postmortem. Since a higher level of degradation is required to result in double-stranded D N A breaks, these samples were then evaluated using the neutral double-stranded comet assay in order to assess extensive fragmentation events occurring over a longer PMI. As shown in figure 3.2, the muscle cells show a progressively larger comet-tail moment, suggesting an increase of D N A fragmentation from 3 to 56 hours postmortem. At time points after 56 hours, the comet image does not change significantly, indicating that no further detectable D N A fragmentation occurs. This may represent a maximum level of D N A fragmentation, or it may indicate the sensitivity limit of this particular method. Postmortem porcine liver and kidney tissues were also subjected to both the single- and double-stranded comet assay, however there was no 54 comet formation visible at any PMI period examined within the first 72 hours. This result had been reported earlier (Johnson L et al, 1999), and can now be explained as being due to the accelerated nuclear decomposition in these two tissues, caused by inherent enzyme activity. Due to fast-acting internal nucleases, all measurable D N A in the liver and kidney samples had already been thoroughly fragmented by the earliest time point, to the extent that separate cell nuclei were no longer discernable, therefore the comet assay was found to be not suitable for use on these tissues. It is not known whether this accelerated degradation was as a result of the freezing and subsequent comet assay preparation (i.e. mechanical dissection), or was inherent to these tissues. For the purposes of this experiment, all samples were subjected to the same treatment for control purposes. As demonstrated in figure 3.3, a quantitative analysis of D N A fragmentation related to PMI shows a strong correlation between increased fragmentation and increasing time since death. Both comet-tail length and comet-tail moment show an approximate linear relationship, which increases with PMI. Upon visual inspection, there is an inherent pattern whereby the comet-tail length and density increase with time postmortem. Regression analysis shows that comet-tail length provides a stronger correlation (r2 = 0.8934) than comet-tail moment (r2 = 0.5244). For the purposes of this experiment, the comet-tail length, which is determined by the smallest size of D N A fragments created, appears to provide a better indicator of PMI than the comet-tail moment, which is determined by the ratio of short fragmented D N A to long, non-migrated DNA. At time points after 56 hours, the D N A has generally degraded to a state 55 which cannot be recognized as a 'comet' and therefore does not provide a quantifiable image. 56 PMI 3 hrs 16 hrs 24 hrs 32 hrs 40 hrs 48 hrs 56 hrs Figure 3.2. Single cell gel electrophoresis of porcine skeletal muscle cells sampled at increasing PMI from 3-56 hours showing a pattern of increasing comet-tail formation with time. Slides stained with silver nitrate, original magnification 200X. c ca co I -CD E o o 600 550 500 450 400 350 300 250 200 1 50 FP = 0.8934 0 10 20 30 40 PMI (hours) 50 60 c CD E o CO I -E o o B 0 00 9 00 8 00 7 00 6 00 5 00 4 00 3 00 2 00 1 00 0 00 R2 = 0.5244 0 1 0 20 30 40 PMI (hours) 50 60 Figure 3.3. Regression analysis of D N A fragmentation demonstrated by single cell gel electrophoresis of porcine skeletal muscle from 3-56 hours postmortem. Fragmentation was quantified using A) comet-tail length and B) comet-tail moment, relative to PMI. Mean values given +/- standard deviation. 58 3.4 Discussion and Conclusions These results indicate that there is indeed a process whereby nuclear D N A is fragmented following death, which is organ and time dependent. In tissues such as kidney and liver, where enzymes tend to be more active, neither comet formation nor intact nuclei were observed, suggesting that the D N A within these tissue cells may have already been fully degraded by the first sampling time point. Proof of concept of PMI D N A fragmentation was demonstrated first in blood and then applied to porcine skeletal muscle. It was shown that fragmentation of nuclear D N A increased with PMI in the 3-56 hour postmortem period. In particular, much of this degradation took place early, in the 3-24 hour postmortem period. This method should be further pursued to determine whether this D N A fragmentation pattern holds true for all body tissues, at varying temperatures, and also over shorter sampling intervals, particularly in the 0-24 hour postmortem period. It is widely known that a number of factors can influence the rate of decomposition of a body. Foremost, increases in temperature have been shown to increase decomposition. It would be expected that an increase in temperature would cause an increase in the rate of reaction which causes the D N A to degrade within the cells postmortem. While 15 degrees Celsius was chosen for this experiments in order to approximate a room-temperature environment, further experiments in a variety of other environmental conditions (indoor, outdoor, warmer, cooler) must be undertaken to determine the relative stability of this method for practical use. 59 While the results of these experiments show a potential for use as a future method of estimating time since death, further studies must be conducted to determine the accuracy and precision which can be achieved. Initially, much shorter intervals should be examined, and then a positive-predictive analysis must be done whereby the experimentally determined postmortem interval must be compared to actual postmortem interval. Since the largest change in comet tail length/moment takes place by 24 hours, it is possible that this time period represents a faster rate of D N A degradation than later time points. With shorter sampling intervals, it should be revealed if there is a single fragmentation event early on, or if this large change in tail length/moment is brought about by a degradation rate which is constant and increases in the first 24 hours. It is suggested that this genomic D N A fragmentation may be caused by some process other than apoptosis, as the fragmentation continues well after energy in the form of ATP ceases to be made available to the body. The apparently sequential fragmentation seen here also contradicts the currently understood process of necrotic cell degradation, which is believed to be a random fragmentation process (Majno G & Joris I, 1995). Further steps should be taken to examine whether apoptosis or necrosis is the cause of the postmortem D N A fragmentation. If both of these pathways can be ruled out, then a novel subcellular destruction pathway may be at work. 60 These results suggest that internal nucleases do act upon a body following death, which contributes to the progressive fragmentation of nuclear D N A in the early postmortem period. This fragmentation can be quantified and appears to be a time-dependant process, which has the potential for use as a predictor of PMI in the field of forensic pathology. For use in a medico-legal setting, ideally greater than one sample would be taken at known points following discovery of a body. With three samples, a curve of tail-length or tail-moment could be created, which could then be compared to experimentally determined rates for approximation of a more specific time since death. 61 3.5 Future experiments In order to further evaluate the hypothesis that D N A degrades in a way causing fragmentation which increases with time since death, there are a number of further experiments that could be undertaken. In order to differentiate the mechanism of action that initiates the D N A degradation between apoptosis and necrosis, there are a number of methods which can be used. The D N A could be extracted and purified and then run on an agarose gel (pulsed-field gel electrophoresis for large fragments) to determine if there are discrete bands, representing a 'ladder' such as occurs with apoptosis, or a 'smear' as seen in the case of necrosis. There is also a commercial staining kit available, the Vybrant Apoptosis Assay Kit (Molecular probes, Eugene OR), which utilizes 2 stains, Yo-Pro-1 and propidium iodide, to stain a cell suspension which is then analyzed using a fluorescence activated cell-sorter (FACS). This method differentiates between apoptosis and necrosis based upon the morphological characteristic of cell membrane permeability to Yo-Pro induced by apoptosis (Majno G & Joris I, 1995; Kroemer et al, 1995; Lincz LF , 1998; Darzynkiewics, Z, 1998; Allen RT, Hunter WJ, Agrawal DK, 1997). This could be utilized to further differentiate between apoptosis and necrosis based on cell membrane permeability. More experiments need to be conducted over shorter intervals in the early time period, to see if the rate of degradation is indeed constant, and determine whether it follows a linear, exponential or logarithmic pattern. Should the pattern which appears here hold true, it could be possible to determine a rate of degradation correlated with time since death within a much narrower range of error than is currently available. 62 3.6 Alternative Applications Another potential application for this technology lies in the field of organ transplantation. Experimental cell samples from various organs could be evaluated by comet assay at known times after removal from the body. A portion of each of these samples could then be cultured in vivo and assayed for survival. It would be expected that a particular level of D N A degradation (as assayed by comet assay) would correspond with a particular survival rate (ie. percentage of surviving cells). This method could then be utilized to test potential donated organs for their viability prior to transplantation. For this procedure, the single-stranded (alkaline) comet assay would be recommended for two reasons. First, since the cells to be tested come from a presumably healthy organ, the amount of D N A degradation should be low. The alkaline assay provides the most sensitivity. Second, even one double-strand break in the D N A will result in unrecoverable cell death. Since the double-strand (neutral) comet assay is not sensitive below approximately 100 double-stranded breaks, this would not be desirable for conducting a cell survival study. 63 BIBLIOGRAPHY Alberts B, Bray D, Lewis J, Raff M , Roberts K, Watson JD. 1994. Molecular Biology of the Cell. 3 r d ed. London: Garland Publishing. Allen RT, Hunter WJ, Agrawal DK. 1997. Morphological and biochemical characterization and analysis of apoptosis. J Pharmacol Toxicol Methods. 37(4)215-28. Anderson D, J. Wright, C. Ioannides 1998. An examination of D N A strand breakage in the comet assay and antioxidant capacity in diabetic patients. Mutat. Res. 398(1-2): 151-161. Baccino E, DeSaint Martin L, Schuliar Y , Guilloteau P, LeRhun M , Morin JF. 1996. Outer ear temperature and time of death. Forensic Sci Int 83:133-46. Bocker W, T. Bauch, W.U. Muller, C. Streffer. 1997. Image analysis of comet assay measurements. Int. J. Radiat. Biol. 72(4):449-460. Boon, M E & Kok, LP. 1986. Standardization and Quantitation of Diagnostic Staining in Cytology. Coulomb Press, Ley den. Bohan TL, Heels EJ. 1995. The case against Daubert: The new scientific evidence "standard" and the standards of several states. J Forensic Sci Nov;40(6): 1030-44. 64 Bradbury, P. & Gordon, K . C 1990. Connective tissues and stains. In: Theory and Practice of Histological Techniques, Bancroft & Stevens Eds. 3 r d Ed. Churchill Livingstone Inc. New York, N Y . Burkitt, H.G., Young, B. & Heath, J.W. Eds. 1993. Wheater's Functional Histology: A Text and Colour Atlas. 3 r d Ed. Churchill Livingston Inc., New York. Cerda H, H. Delincee, H. Haine, H. Rupp. 1997. The D N A 'comet assay' as a rapid screening technique to control irradiated food. Mutat. Res. 375:167-181. Cina SJ. 1994. Flow cytometric evaluation of D N A degradation: A predictor of postmortem interval? Am J Forensic Med Pathol 15:300-2. Darzynkiewicz Z, Juan G, L i X , Gorczyca W, Murakami T, Traganos F. 1997. Cytometry in cell necrobiology: Analysis of apoptosis and accidental cell death (necrosis). Cytometry. 27(1)1-20. Fairbairn DW, P. Olive, K . L . O'Neill. 1995. The comet assay: A comprehensive review. Mutat. Res. 339:37-59. Feulgen, R. & Rossenbeck, H. 1924. Mikroskopischchemischer Nachweis einen Nucleinsaure von Typus der Thymonuclinsaure und die darauf berhende elektive Farbung 65 von Zellkernen in microskopischen Preparaten. Zeitschrift Physiology Chemistry, 135, 203. Frammery MF, Leglise D, Baccino E, Morin JF, Koulmann L, Frammery D. 1990. Time of death, a multiple-indicator study; Blood, vitreous humor, CSF electrolytes and liver temperature, Proceedings of the 12th International Association of Forensic Scientists meeting, Adelaide p. 481. Gottlieb, M . and Chavko, M . 1987. Silver staining of vative and denatureated eucaryotic D N A in agarose gels. Anal. Biochem., 165, 33-37. Gurr, E. 1971. Synthetic dyes in Biology, Medicine and Chemistry. Academic Press, London. P. 114. Gutierrez S, E. Carbonell, P. Galofre, A. Creus, R. Marcos. 1998. Application of the single cell gel electrophoresis (SCGE) assay to the detection of D N A damage induced by 1 3 1 I treatment in hyperthyroidism patients. Mutagenesis 13(l):95-98. Harkness RA. 1988. Hypoxanthine, xanthine and uridine in body fluids, indicators of ATP depletion. J Chromatogr 429:255-78. Henssge C. 1988. Death time estimation in casework. I. The rectal temperature time of death nomogram. Forensic Sci Int 38:209-36. 66 Henssge C, (Ed.) 1995. Estimation of the time since death in the early postmortem period. London, Edward Arnold. Horobin, R.W. 1990. An overview of the theory of staining. In: Theory and Practice of Histological Techniques, Bancroft & Stevens Eds. 3 r d Ed. Churchill Livingstone Inc. New York, N Y . Huang P, R.E. Durand. 1998. Use of the comet assay for assessment of drug resistance and it's modulation in vivo. Br. J. Cancer 77(3):412-416. Hudson B, Upholt WB, Devinny J, Vinograd J. 1969. The use of an ethidium analygur in the dye-buoyant density procedure for the isolation of closed circular DNA: the variation of the superhelix density of mitochondrial DNA. Proc Natl Acad Sci. 62(3)813-20. Husain, O.A.N. & Watts, K .C . 1984. Rapid demonstration of nucleic acids using oxidised gallocyanin and chromic potassium sulphate: methods and applications. J Clin Pathol 37:99-101. Johnson L, P.W. Payne, D. Haskins, E. Baccino, J.A.J. Ferris. 1999. Estimation of time since death by single-cell gel electrophoresis analysis of cell nuclei. Proceedings of the Fifth International Conference in Clinical Forensic Medicine of the World Police Medical Officers, Vancouver, Canada, August 1999. J. Clin. Forensic Med. 6(3): 197-198. 67 Jones M D , James WS, Barasi S, Nokes L D M . 1995. Postmortem electrical excitability of skeletal muscle: preliminary investigation of an animal model. Forensic Sci Int 76:91-6. Junquiera, L C , Carneiro, J. and Long, JA. 1986. Basic Histology. 5 Ed. Lange Medical Publications, Los Altos, CA. Kaatsch HJ, Schmidtke W, Nietsch W. 1994. Photometmic measurement of pressure-induced blanching of livor mortis as an aid to estimating time of death. International J Leg Med 106:209-14. Kaiser J. 1998. Should engineer witnesses meet same standards as scientists? Science 281:1578. Knight B. 1995. Hypostasis and timing of death. In: Henssge C, Knight B, editors. Estimation of the Time Since Death in the Early Postmortem Period. London: Edward Arnold. Krishan A. 1975. Rapid flow cytofluorometric analysis of mammalian cell cycle by propidium iodide staining. J Cell Biol. 66(1)188-93. Kroemer G, Petit P, Amazami N , Vayssiere LF , Mignotte B. The biochemistry of programmed cell death. FASEBJ. 9(13)1277-87. 68 Krompecher T. 1994. Experimental evaluation of rigor mortis. VIII. Estimation of time since death by repeated measurements of the intensity of rigor mortis on rats. Forensic Sci Int 68:149-59. Krompecher T. 1995. Rigor mortis: estimation of the time since death by evaluation of the cadaveric rigidity. In: Henssge C, Knight B, editors. Estimation of the Time Since Death in the Early Postmortem Period. London: Edward Arnold. Lincz LF. 1998. Deciphering the apoptotic pathway: A l l roads lead to death. Immunol Cell Biol. 76(1)1-19. McKelvey-Martin VJ , S.R McKeown, S.R. Johnston, P.J. McCarthy, N.F. Rajab, C S . Downes. 1998. Emerging applications of the single cell gel electrophoresis (comet) assay. I. Management of invasive transitional cell human bladder carcinoma. II. Fluorescent in situ hybridization comets for the identification of damaged and repaired D N A sequences in individual cells. Mutagenesis 13(l):l-8. McManus, J.F.A. & Mowry, R.W. 1965. Staining Methods: Histologic and histochemical. Harper & Row, New York. 69 Madea B. 1995. Muscle and tissue changes after death - Postmortem mechanical excitation of skeletal muscle. In: Henssge C, Knight B, editors. Estimation of the Time Since Death in the Early Postmortem Period. London: Edward Arnold. Madea B, Kaferstein H, Hermann N , Sticht G. 1994. Hypoxanthine in vitreous humor and cerebrospinal fluid - a marker of postmortem interval and prologned (vital) hypoxia? Remarks also on hypoxanthine in SIDS. Forensic Sci Int 65:19-31. Madea B, Henssge C, Honig W, Gerbracht A. 1989. References for determining the time of death by potassium in vitreous humor. Forensic Sci Int 40:231-43. Majno G, I. Joris. 1995. Apoptosis, oncosis and necrosis: An overview of cell death. Am. J. Pathol. Jan., 146(1):3-15. MallachHJ. 1964. Zur Frage des Todeszeitnes timmung. Berl Med 18:577-82. Mallory, F.B. 1944. Pathological Technique. W.B. Sanders, Philadelphia. Marshall TK, Hoare FE. 1962. Estimating the time of death. The rectal cooling after death and its mathematical expression. J Forensic Sci 7:56-81. Marshall T. 1983. Alternative approach to two-dimensional gel electrophoresis of serum. Clin Chern. 29(3)592-3. 70 Mikel, U.V. & Becker, R.L. (Jr.) 1991. A comparative study of quantitative stains for D N A in image cytometry. In Analytical and Quantitative Cytology and Histology. 1304:253-260. Nokes L D M , Flint T, Jaafar S, Knight B H . 1992. The use of either the nose or outer ear as a means of determining the postmortem period of a human corpse. Forensic Sci Int 54:153-8. Nokes L D M , Hicks B, Knight, B. 1986. The use of trachea temperature as means of determining the postmortem period. Med Sci Law 26:199-202. Olive PL, J.P. Banath, R.E. Durand. 1990. Heterogeneity in radiation-induced D N A damage and repair in tumor and normal cells measured using the "comet" assay. Radiat. Res. 122:86-94. Payne PW, L . A . Johnson, D. Haskins , E. Baccino, J.A.J. Ferris. 1999. Estimation of time since death by image cytometric analysis of cell nuclei. Proceedings of the Fifth International Conference in Clinical Forensic Medicine of the World Police Medical Officers. Vancouver, Canada. J. Clin. Forensic Med. 6(3): 197. Pfuhler S, H.U. Wolf. 1996. Detection of DNA-crosslinking agents with the alkaline comet assay. Environ. Mol. Mutagen. 27:196-201. 71 Presnell, J.K., & Schreibman, M.P. 1997. Humason's Animal Tissue Techniques. 5 m Ed. Johns Hopkins University Press, Baltimore, M D . Querido D. 1994. Time-dependent changes in electrical resistance of the intact abdomen during the 1-504 h postmortem period in rats. Forensic Sci Int 67:17-25. Rognum TO, Hauge S, Oyasaeter S, Saugstad OD. 1991. A new biochemical method for estimation of postmortem time. Forensic Sci Int 51:139-4. Ross G M , P. Wilcox, A.R. Collins. 1995. The single cell microgel electrophoresis assay (comet assay): technical aspects and applications. Report on the 5th L H Gray Trust Workshop, Institute of Cancer Research, 1994. Mutat. Res. 337:57-60. Somerville L L & Wang K. 1981. The ultrasensitive silver "protein" stain also detects nanograms of nucleic acids. Biochem Biophys Res Commun. 102(1)53-8. Stevens, A . 1990. The hematoxylins.. In: Theory and Practice of Histological Techniques, Bancroft & Stevens Eds. 3 r d Ed. Churchill Livingstone Inc. New York, N Y . 72 Stevens, A. , & Bancroft, J.D. 1990. Proteins and nucleic acids. In: Theory and Practice of Histological Techniques, Bancroft & Stevens Eds. 3 r d Ed. Churchill Livingstone Inc. New York, N Y . Tang DG, A . Porter. 1996. Apoptosis: A current molecular analysis. Pathol. Oncol. Res. 2(3):117-131. van Rood JJ, van Leeuwen A, Ploem JS. 1976. Simultaneous detection of two cell populations by two colour fluorescence and application to the recognition of B-cell determinants. Nature. 262(5571)795-7. Vanezis P. 1991. Assessing hypostasis by colorimetry. Forensic Sci Int 52:1-3. Vanezis P, Trujillo O. 1996. Evaluation of hypostasis using a colorimeter measuring system and its application to assessment of the post-mortem interval (time of death). Forensic Sci Int 78:19-28. Yeh C-JG, Hsi B-L and Faulk WP. 1981 Propidium iodide as a nuclear marker in immunofluorescence. II. Use with cellular identification and viability studies. J Immunol Methods, 43;269-275. 73