UBC Faculty Research and Publications

Transplantation of human neural stem cells transduced with Olig2 transcription factor improves locomotor… Hwang, Dong H; Kim, Byung G; Kim, Eun J; Lee, Seung I; Joo, In S; Suh-Kim, Haeyoung; Sohn, Seonghyang; Kim, Seung U Sep 22, 2009

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

Item Metadata


52383-12868_2009_Article_1121.pdf [ 2.38MB ]
JSON: 52383-1.0220711.json
JSON-LD: 52383-1.0220711-ld.json
RDF/XML (Pretty): 52383-1.0220711-rdf.xml
RDF/JSON: 52383-1.0220711-rdf.json
Turtle: 52383-1.0220711-turtle.txt
N-Triples: 52383-1.0220711-rdf-ntriples.txt
Original Record: 52383-1.0220711-source.json
Full Text

Full Text

ralssBioMed CentBMC NeuroscienceOpen AcceResearch articleTransplantation of human neural stem cells transduced with Olig2 transcription factor improves locomotor recovery and enhances myelination in the white matter of rat spinal cord following contusive injuryDong H Hwang1, Byung G Kim*1,2, Eun J Kim1, Seung I Lee1, In S Joo2, Haeyoung Suh-Kim3, Seonghyang Sohn4 and Seung U Kim*1,5,6Address: 1Brain Disease Research Center, Institute for Medical Sciences, Ajou University School of Medicine, Suwon, Korea, 2Department of Neurology, Ajou University School of Medicine, Suwon, Korea, 3Department of Anatomy, Ajou University School of Medicine, Suwon, Korea, 4Laboratory of Cell Biology, Institute for Medical Sciences, Ajou University School of Medicine, Suwon, Korea, 5Medical Research Institute, Chungang University School of Medicine, Seoul, Korea and 6Division of Neurology, Department of Medicine, UBC Hospital, University of British Columbia, Vancouver, CanadaEmail: Dong H Hwang - drhdh@ajou.ac.kr; Byung G Kim* - kimbg@ajou.ac.kr; Eun J Kim - neukbg@naver.com; Seung I Lee - silee02@gmail.com; In S Joo - isjoo@ajou.ac.kr; Haeyoung Suh-Kim - hysuh@ajou.ac.kr; Seonghyang Sohn - sohnsh@ajou.ac.kr; Seung U Kim* - sukim@interchange.ubc.ca* Corresponding authors    AbstractBackground: Contusive spinal cord injury is complicated by a delayed loss of oligodendrocytes, resulting inchronic progressive demyelination. Therefore, transplantation strategies to provide oligodendrocyte lineage cellsand to enhance the extent of myelination appear to be justified for spinal cord repair. The present studyinvestigated whether transplantation of human neural stem cells (NSCs) genetically modified to express Olig2transcription factor, an essential regulator of oligodendrocyte development, can improve locomotor recovery andenhance myelination in a rat contusive spinal cord injury model.Results: HB1.F3 (F3) immortalized human NSC line was transduced with a retroviral vector encoding Olig2, anessential regulator of oligodendrocyte development. Overexpression of Olig2 in human NSCs (F3.Olig2) inducedactivation of NKX2.2 and directed differentiation of NSCs into oligodendrocyte lineage cells in vitro. Introductionof Olig2 conferred higher proliferative activity, and a much larger number of F3.Olig2 NSCs were detected by 7weeks after transplantation into contused spinal cord than that of parental F3 NSCs. F3.Olig2 NSCs exhibitedfrequent migration towards the white matter, whereas F3 NSCs were mostly confined to the gray matter oraround the lesion cavities. Most of F3.Olig2 NSCs occupying the spared white matter differentiated into matureoligodendrocytes. Transplantation of F3.Olig2 NSCs increased the volume of spared white matter and reducedthe cavity volume. Moreover, F3.Olig2 grafts significantly increased the thickness of myelin sheath around theaxons in the spared white matter. Finally, animals with F3.Olig2 grafts showed an improvement in the quality ofhindlimbs locomotion.Conclusion: Transplantation of NSCs genetically modified to differentiate into an oligodendrocytic lineage maybe an effective strategy to improve functional outcomes following spinal cord trauma. The present study suggestsPublished: 22 September 2009BMC Neuroscience 2009, 10:117 doi:10.1186/1471-2202-10-117Received: 15 May 2009Accepted: 22 September 2009This article is available from: http://www.biomedcentral.com/1471-2202/10/117© 2009 Hwang et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 16(page number not for citation purposes)that molecular factors governing cell fate decisions can be manipulated to enhance reparative potential of the cell-based therapy.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117BackgroundTraumatic spinal cord injury (SCI) results in severe andpermanent neurological deficits. However, there is no sin-gle effective therapeutic option to improve functional out-comes. Intense research efforts, employing a rodentmodel of contusive injury which closely mimics humanSCI, have identified that the pathology in the white matterincurred by injury is closely associated with the degree offunctional deficits [1-3]. One of the important pathologi-cal processes in the white matter is a chronic and progres-sive demyelination of the spared axons [4-7], whichoccurs primarily due to delayed and widespread apoptosisof the oligodendrocytes [8,9]. Absence of myelin sheathand resultant exposure of potassium channels lead to afailure of electrical conduction through spared axons,contributing to chronic functional deficits following SCI[10]From these observations, transplantation strategies to pro-vide cells capable of myelinating axons and enhanceremyelination seem to be well justified for spinal cordrepair. Transplantation of oligodendrocyte progenitorcells or glial progenitor cells derived from embryonic orneural stem cells promoted functional recovery in SCI ani-mal models [11-13]. Furthermore, a large part of behavio-ral gains following grafts of murine neural stem cellswithout lineage restriction has recently been attributed toan enhanced myelination in the spared white matter[14,15]. These studies indicate that enhancing myelina-tion by transplantation of stem/progenitor cells is a prom-ising approach to improve functional outcomes forpatients suffering from SCI.An increasing number of molecular factors that govern thefate determination of neural cells during developmenthave been identified [16-18]. Manipulation of appropri-ate factors may facilitate differentiation of transplantedcells to a desired lineage. The basic HLH transcription fac-tor Olig is a key regulator for the differentiation of oli-godendrocyte lineage cells during development [19-21].Olig2, one of the Olig family, is more highly expressed inthe ventral spinal cord during early developmental periodof human fetus and may play a crucial role in the differen-tiation of oligodendrocytes in the spinal cord [22]. In thepresent study, we overexpressed Olig2 gene in stableimmortalized human neural stem cells (NSCs), whichhave been widely employed to repair the CNS in variousexperimental models of neurological disorders [23-28].Here we show that overexpression of Olig2 transcriptionfactor directed differentiation of human NSCs exclusivelyinto an oligodendrocyte lineage in vitro. We also reportthat transplantation of Olig2 overexpressing human NSCsimproved locomotor function and increased the extent ofResultsCharacterization of NSCs transduced with Olig2 transcription factorHuman NSC line overexpressing Olig2 (F3.Olig2) wasgenerated by retroviral transduction of parental F3 humanNSC line with the full length coding region of bHLH tran-scription factor Olig2. Introduction of Olig2 gene resultedin a change of cellular morphology (Figure 1A, B).F3.Olig2 cells exhibited multiple thin-branched cytoplas-mic processes in a phase-contrast image, whereas F3 NSCsshowed polygonal shape without the branched processes.RT-PCR analysis confirmed expression of Olig2 mRNA inF3.Olig2 cells. Nkx2.2, a transcription factor whichdirectly regulates the differentiation and maturation ofoligodendrocytes [29], was not expressed in F3 parentalNSCs but newly expressed after introduction Olig2 gene(Figure 1C). Olig2 expression at the protein level was alsoconfirmed by immunocytochemistry (Figure 1D, E).Immunocytochemical analysis of phenotypic expressionshowed that essentially all of F3.Olig2 cells expressed oli-godendrocytic lineage markers such as O4 and GalC (Fig-ure 1G, I), whereas the expression of those markers wasrarely observed in F3 NSCs (Figure 1F, H). F3.Olig2 cellsalso expressed myelin basic protein (MBP) (Figure 1J, K),indicating that F3.Olig2 cells could differentiate into mye-linating mature oligodendrocytes. These results indicatethat ectopic overexpression of Olig2 transcription factorin NSCs turns on transcriptional machinery for the oli-godendrocytic cell fate and forces F3 human NSCs to dif-ferentiate into oligodendrocytic lineage path. However,F3.Olig2 cells did not show differentiation into matureastrocytes or neurons (See Additional file 1).Transplantation of NSCs transduced with Olig2 transcription factor into contused rat spinal cordAnimals received vehicle injection or cellular grafts 7 daysafter contusive SCI. At 2 weeks after transplantation,grafted cells that were immunoreactive to human specificmitochondria (either F3 or F3.Olig2) were found in clus-ters mostly around the lesion sites (Figure 2A-D). At 7weeks, both F3 and F3.Olig2 cells were still observed inthe contused spinal cord (Figure 2E, F). Unexpectedly, thenumber of F3.Olig2 cells was markedly larger than that ofF3 NSCs along the rostrocaudal extent of the lesioned spi-nal cord at this time point. The number of F3.Olig2 cellswas highest around the rostral injection site (2 mm rostralto the epicenter) (Figure 2G). Differences in the numberof surviving grafted cells between F3 and F3.Olig2 cellswere statistically significant over the different rostrocau-dal regions from the epicenter (p < 0.001 by repeatedmeasures two-way ANOVA). At this time point, distribu-tion of grafted F3.Olig2 cells was different from that of F3Page 2 of 16(page number not for citation purposes)myelination of spared white matter in a rat SCI model. cells. Many of grafted F3.Olig2 cells seemed to migratefrom the injection site and occupied the white matter. InBMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117contrast, most F3 cells without Olig2 overexpression werestill positioned inside the gray matter or around the lesioncavities, with only about 10% of them observed in theventrolateral white matter (Figure 2H). Differences in thepattern of distribution were most apparent at the rostralregions, where the proportions of grafted cells within thewhite matter were 5 fold higher in F3.Olig2 NSCs (Figure2F). The percent grafted cells within the white matter ver-sus total human mitochondria positive cells in the entirecross section was significantly higher in F3.Olig2 than F3NSCs (p < 0.001, by repeated measures two-way ANOVA).Taken together, these results suggest that overexpressionof Olig2 transcription factor leads to an increase in thenumber of surviving grafted cells in the contused spinalcord. In addition, they also indicate that grafted cells withOlig2 transcription factor possess propensity for migra-tion towards the white matter.A recent study indicated that Olig2 transcription factorplays a critical role in the regulation of NSC proliferation[30]. It is conceivable, therefore, that the presence ofhigher number of F3.Olig2 cells may be due to a height-ened proliferative activity induced by ectopic expressionof Olig2 transcription factor. Consistent with this idea,growth of F3.Olig2 cells in vitro was significantly acceler-ated after 24 hours in culture (Fig 3A). Differences in cellproliferation were highly significant at both 48 and 72hours (p < 0.001 by unpaired T test). BrdU incorporationindex was also obtained as a marker of cellular prolifera-tion at 36 hours when the slope of proliferation curve wassteepest (Fig 3B). F3.Olig2 cells incorporated proliferationmarker BrdU far more frequently than F3 cells (p < 0.001).More importantly, grafted F3.Olig2 NSCs at 2 weeks aftertransplantation were positive for proliferation markerKi67 more frequently (> 2.5 fold) than F3 (Figure 3C-I) (p< 0.01). These results indicate that Olig2 overexpressioncan increase the size of proliferative progenitor pool andthe higher proliferative potential may explain the largernumber of F3.Olig2 cells in the injured spinal cord at 7weeks after transplantation.Differentiation of F3.Olig2 cells in the contused spinal cordDifferentiation of grafted F3 or F3.Olig2 NSCs was exam-ined at 7 weeks after transplantation. F3.Olig2 cells iden-tified by human specific mitochondria that migrated tothe spared white matter were colocalized with oli-godendrocyte marker APC-CC1 (Figure 4A-F). In contrast,as shown in Figure 2C, F3 NSCs without Olig2 transcrip-tion factor were rarely found in the spared white matter,and even a majority of F3 cells that migrated to the whitematter did not express APC-CC1 (Figure 4G). Interest-ingly, F3.Olig2 cells that remained in the gray matter oraround the lesion cavity showed different behaviors (Fig-Characterization of human neural stem cells (NSCs) trans-duced with Olig2 transcription f ctorFigure 1Characterization of human neural stem cells (NSCs) transduced with Olig2 transcription factor. (A, B) Phase contrast images of the parental NSCs (F3) and F3.Olig2 NSCs. (C) Comparison of gene expression by RT-PCR analysis between F3 and F3.Olig2 NSCs. (D, E) Confir-mation of Olig2 transcription factor expression in F3.Olig2 cells (E). Olig2 expresssion was not observed in F3 cells (D). (F-K) In vitro immunocytochemical detection of oligodendro-cyte lineage markers as indicated. Cells were grown on cov-erslips in DMEM with 2% FBS for 5 days and then fixed with 4% paraformaldehyde. (F, H, J) F3 cells, (G, I, K) F3.Olig2 cells. GalC = galactocerebrosidase, MBP = myelin basic pro-Page 3 of 16(page number not for citation purposes)ure 4H). They were usually found in aggregation and dif-ferentiation into APC-CC1 positive oligodendrocytes wastein. Scale bar = 50 μm.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Page 4 of 16(page number not for citation purposes)Transplantation of Olig2 overexpressing human NSCs after contusive spinal cord injuryFigure 2Transplantation of Olig2 overexpressing human NSCs after contusive spinal cord injury. (A-D) Immunofluores-cence staining of the transverse sections close to the epicenter at 2 weeks after transplantation of F3 (A, B) and F3.Olig2 (C, D) cells. Grafted human NSCs were identified by immunoreactivity against human specific mitochondria (red). GFAP staining (green) was performed to depict the appearance of spinal cord lesions. Photomicrographs in B and D are magnified images of the boxed regions in A and C, respectively. (E, F) Bright field microscopic images of the transverse sections located at 2 mm rostral to the epicenter at 7 weeks after transplantation of F3 (E) and F3.Olig2 (F) NSCs. Many F3.Olig2 cells were observed in the white matter at this time point. Dotted lines indicate boundaries of the gray matter. Scale bar = 500 μm. (G) The number of human mitochondria positive cells was stereologically counted in the spinal cord sections from the epicenter to 4 mm ros-tral and caudal regions at 1 mm interval. (H) The percentage grafted cells in the ventrolateral white matter versus total number of grafted cells in the entire transverse sections. * = p < 0.05, ** = p < 0.01, and *** = p < 0.001 by unpaired T test at each dis-tance from the epicenter. E represents epicenter. Error bars indicate mean ± SEM. N = 4 animals per group.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Page 5 of 16(page number not for citation purposes)Comparison of proliferative capacity between F3 and F3Figure 3Comparison of proliferative capacity between F3 and F3.Olig2 human NSCs. (A) Cell proliferation assay measured by Cell Counting Kit-8. Cells were grown in a 96-well plate and the viability was measured at different time points after initial culture. ** = p < 0.01, *** = p < 0.001 by unpaired T test. Error bars indicate mean ± SD. N = 4 replicate experiments. (B) BrdU incorporation assay. Cells were grown on a 9 mm coverslip for 36 hours and BrdU was added for 2 hours. *** = p < 0.001 by unpaired T test. Error bars indicate mean ± SD. N = 3 coverslips for each condition. (C-H) Representative images of transverse spinal cord sections doubly stained with human mitochondria (red) and Ki67 (green) at 2 weeks after transplanta-tion. A majority of grafted F3 cells (C-E) did not express Ki67, whereas the nuclei of F3.Olig2 grafted cells (F-H) were fre-quently colocalized with Ki67 (arrows). The nuclei were visualized by DAPI (blue) (E, H). Scale bar = 20 μm. (I) Quantification of the percent grafted cells containing Ki67 positive nuclei. Error bars indicate mean ± SD. ** = p < 0.01 by unpaired T test. N = 3 and 4 animals for F3 and F3.Olig2 groups, respectively.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Page 6 of 16(page number not for citation purposes)Phenotypic differentiation of human NSCs transduced with Olig2 transcription factor at 7 weeks following transplantation into contused spinal cordFigure 4Phenotypic differentiation of human NSCs transduced with Olig2 transcription factor at 7 weeks following transplantation into contused spinal cord. Grafted cells were detected by immunoreactivity against human mitochondrial antigen (red). (A-C) F3.Olig2 human NSCs positioned in the spared white matter (A) were colocalized with APC-CC1 (CC-1) (B). Merged cells were shown in yellow (C). Arrows indicate the same cells in the three images. (D-E) Magnified images of a grafted F3.Olig2-derived CC1 positive cell. (Inset) a low magnification merged image showing the location of the magnified cell close to the outer rim of the spared white matter. (F) A 3D reconstruction image of the double positive cells in (D, E). (G) F3 NSCs in the white matter. In contrast to F3.Olig2 cells, F3 cells were very infrequently observed in the white matter and very few of them were positive for CC1. (H) In contrast to F3.Olig2 NSCs in the spared white matter, most of F3.Olig2 cells that remained in the gray were not positive for CC1. (I-J) F3.Olig2 NSCs in the white matter expressed myelin basic protein (MBP). Arrows indicate the same cells. (K) Most of the F3.Olig2 cells in the white matter did not differentiated into astrocytes. (L) Only a small proportion of F3.Olig2 or F3 NSCs in the gray matter were positive for GFAP (arrow) (M) Many F3.Olig2 or F3 cells around the lesion cavity still expressed NSC marker nestin (arrows). Dotted lines in (L-M) indicate the margins of lesion cavities. All scale bars = 20 μm. (N) Quantification of percent NSCs that differentiated into oligodendrocytes (CC1) or astro-cytes (GFAP). Error bars indicate mean ± SD. *** = p < 0.001 by unpaired T test. N = 4 animals for each group.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117very infrequently observed when compared to those cellspositioned in the white matter. Quantification showedthat 52.6 ± 9.8% of grafted F3.Olig2 NSCs were colocal-ized with APC-CC1 as compared to 7.7 ± 5.9% for paren-tal F3 cells (Figure 4N) (p < 0.001). When themeasurement was confined to the white matter, 84.9 ±2.2% of F3.Olig2 NSCs were doubly positive for humanspecific mitochondria and APC-CC1. F3.Olig2 NSCs inthe spared white matter also express MBP (Figure 4I-J), acell type specific marker for myelinating oligodendro-cytes, suggesting that F3.Olig2 cells in the spared whitematter differentiated into mature oligodendrocytes thatwere capable of producing myelin. Consistent with theseresults, most F3.Olig2 NSCs in the white matter did notcolocalize with astroglial marker GFAP (Figure 4K). In thegray matter, a small proportion of F3.Olig2 NSCsexpressed GFAP. F3 NSCs in the gray matter showed a verysimilar extent of differentiation into astrocytes (Figure4L). Overall, approximately 10% of both F3 and F3.Olig2cells differentiated into astrocytes (Figure 4N). Neuronaldifferentiation was rarely observed in either F3 orF3.Olig2 NSCs in the gray matter (data not shown). Thesefindings suggested that the majority of F3.Olig2 or F3NSCs in the gray matter remained undifferentiated.Indeed, many of F3 or F3.Olig2 NSCs in the gray matter,especially around the lesion cavities, were still positive forimmature NSC marker nestin (Figure 4M).Transplantation of F3.Olig2 NSCs increased the volume of spared white matter and enhanced tissue sparingThe areas of myelinated white matter were measured inthe eriochrome-stained transverse spinal cord sections(Fig 5A-C), and the volume of spared white matter wascalculated using Cavalieri's Principle. Transplantation ofF3 NSCs did not affect the mean volume of spared whitematter compared to Vehicle group (11.4 ± 2.6 mm3 inVehicle and 10.9 ± 2.0 mm3 in F3 groups). In contrast,F3.Olig2 grafts significantly increased the volume ofspared white matter compared to Vehicle and F3 groups(13.8 ± 2.7 mm3 in F3.Olig2 group; p < 0.05 by one-wayANOVA followed by Tukey's posthoc analysis) (Figure 5D).We also measured the volume of cystic cavities, and foundthat transplantation of F3.Olig2 NSCs significantlyreduced the volume of cystic cavities (2.2 ± 1.8 mm3 inVolumes of the spared white matter and lesion cavitiesFig re 5Volumes of the spared white matter and lesion cavities. (A-C) Representative images of erichrome-stained spinal cord sections at 1 mm caudal to the epicenter. (A) Vehicle (Veh), (B) F3, (C) F3.Olig2 groups. Scale bar = 500 μm. (D) Comparison of volumes of the spared white matter. (E) Comparison of volumes of lesion cavities. Error bars indicate mean ± SD. N = 11, 12, and 12 for Vehicle, F3, and F3.Olig2 groups, respectively. * = p < 0.05, ** = p < 0.01 by one-way ANOVA followed by Page 7 of 16(page number not for citation purposes)Tukey's posthoc analysis.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Vehicle, 1.4 ± 0.9 mm3 in F3, and 0.5 ± 0.3 mm3 inF3.Olig2 group) (Figure 5E). One-way ANOVA revealedsignificant differences in the volume of cavities betweenthe three groups (p < 0.01), and Tukey's posthoc analysisshowed a significant difference only between Vehicle andF3.Olig2 groups (p < 0.01).Extent of myelination in the spared white matterThe observation that F3.Olig2 NSCs in the white matterdifferentiated into mature oligodendrocytes and thattransplantation of F3.Olig2 NSCs increased the volume ofspared white matter led us to examine the extent of mye-lination in the white matter. Thickness of individual mye-lin sheaths and diameter of the axons were measured insemithin (1 μm) spinal cord sections stained with toluid-ine blue (3 rats and 692 axons in Vehicle, 6 and 1311 inF3, 7 and 1348 in F3.Olig2 groups, respectively). In tolui-dine blue stained transverse sections, regions devoid ofmyelinated axons were frequently observed in the whitematter after SCI (Figure 6A). Such demyelinated regionswere less frequent in animals with F3.Olig2 grafts (Figure6A-C). Degenerating axons were also less frequentlyobserved in F3.Olig2 group. We found that the mean mye-lin ratio was higher in animals with F3 cells compared toVehicle group (Figure 6E). Transplantation of F3.Olig2cells further increased the mean myelin ratio. One-wayANOVA revealed a significant group effect on the thick-ness of myelin sheath (p < 0.001). Tukey's posthoc analysisshowed significant differences between F3.Olig2 andVehicle or F3 groups (p < 0.001 or p < 0.05, respectively),and the difference between F3 and Vehicle groups wasalso significant (p < 0.05). Distribution of the myelin ratioin F3.Olig2 group was most shifted to the right comparedto the other groups in the cumulative frequency histogram(Figure 6E), indicating that a higher proportion of axonswere ensheathed by thicker myelin in animals withF3.Olig2 grafts.Measurement of myelin thickness in the spared white matterFigure 6Measurement of myelin thickness in the spared white matter. (A-C) Representative images of the ventral white mat-ter in toluidine-blue stained semithin sections from (A) Vehicle, (B) F3, and (C) F3.Olig2 groups. Asterisks indicate degenerat-ing axons. Scale bar = 10 μm. (D) Comparison of the mean myelin ratio. N = 3, 6, and 7 animals for Vehicle, F3, and F3.Olig2 groups, respectively. Error bars represent mean ± SD. * = p < 0.05, *** = p < 0.001, by one-way ANOVA followed by Tukey's posthoc analysis. (E) Cumulative frequency histogram of the myelin ratio. The number of axons analyzed was 692, 1311, and Page 8 of 16(page number not for citation purposes)1348 in Vehicle, F3, and F3.Olig2 groups, respectively.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Transplantation of F3.Olig2 NSCs improves locomotor recoveryBBB locomotor rating scale was used to assess the extentof locomotor recovery. Most animals were not able tomove their hindlimbs or showed only slight movement oftheir hip and/or knee joints immediately after contusiveinjury at the T9 segment. The locomotor deficits were rap-idly recovered during the first two weeks regardless oftreatment (Figure 7A). The average locomotor scale of theanimals with vehicle injection or transplantation of F3hNSCs began to plateau at 3 weeks and did not showremarkable changes until 7 weeks after injury, the lasttime point measured. Animals with F3.Olig2 grafts con-tinuously improved locomotor score even after 3 weeks ofinjury to the extent where they could regain coordinationbetween the fore- and hindlimbs with almost consistentweight-supported plantar steps (average BBB scale 12.8).Repeated measures two-way ANOVA revealed a signifi-cant treatment effect over time (p < 0.05). Tukey's posthocanalysis at each time point showed significant differencesin the mean BBB score from 4 weeks after injury (p < 0.05at 4 and 5 weeks; p < 0.01 at 6 and 7 weeks). We also meas-ured the number of paw placement errors in grid walk,which is generally regarded as more challenging to injuredanimals than open field locomotion. Although shamoperated animals crossed the grid runway with very fewerrors, injured animals made more than ten errors pereach run (Figure 7B), suggesting that spinal injury severelycompromised the sensorimotor integration of the hind-limbs. Transplantation of either F3 or F3.Olig2 cells didnot significantly reduce the number errors on grid.Transplantation of F3Figure 7Transplantation of F3.Olig2 NSCs improves locomotor recovery. (A) Comparison of BBB locomotor scale. An arrow indicates the time point when transplantation was performed. N = 11, 12, and 12 for Vehicle (Veh), F3, and F3.Olig2 groups, respectively. Error bars represent mean ± SD. * = p < 0.05, ** = p < 0.01 by repeated measures two-way ANOVA followed by posthoc Tukey's analysis. (B) Grid walk. The number of hindpaw placement errors on grid per run was counted at 4 and 7 weeks (4w and 7w, respectively) after transplantation. N = 9, 8, and 9 for Vehicle, F3, and F3.Olig2 groups, respectively. Error bars represent mean ± SD. (C-E) Foot print analysis. (C) Interlimb distance (ILD) or base of support. (D) Stride length (SL) of the left hindlimb. (E) SL of the right hindlimb. N = 5, 4, 5, and 5 for sham, Vehicle, F3, F3.Olig2 groups, respectively. Error bars represent mean ± SD. * = p < 0.05 compared to sham. For B - E, stippled, white, gray, and black bars represent sham, Vehicle, Page 9 of 16(page number not for citation purposes)F3, and F3.Olig2 groups, respectively.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117The quality of hindlimbs movement during locomotionwas also measured using footprint analysis. Spinal injuryresulted in an increase of the distance between the twohindlimbs (interlimb distance) by about 40% (Figure7C). Transplantation of F3 cells slightly reduced the dis-tance, and the animals with F3.Olig2 grafts almost com-pletely normalized the interlimb distance (p < 0.05). Theanimals with spinal injury also exhibited shorter stridelength in both limbs than sham operated rats. Transplan-tation of F3.Olig2 cells tended to restore normal stridelength, although the difference was not statistically signif-icant (Figure 7D, E). These results suggest that graftinghNSCs transduced with Olig2 transcription factor into thecontused spinal cord enhances recovery of open fieldlocomotion and improves quality of the hindlimbs move-ment during locomotion.DiscussionIn the present study, we demonstrated that transductionof human NSCs with Olig2 transcription factor inducesthe expression of various phenotypic markers of the oli-godendrocyte in vitro. The Olig transcription factors(Olig1 and Olig2) are required for the phenotypic deter-mination of oligodendrocytes during development.Genetic ablation of Olig transcription factors resulted inthe absence of oligodendrocytes and motor neurons [31-33]. In the spinal cord, Olig2 seems to play a more crucialrole for the specification of oligodendrocyte phenotype,since Olig2 expression in the early spinal cord is higherthan that of Olig1 [19,20,22]. In the present study, over-expression of Olig2 in the human NSCs led to an activa-tion of NKX2.2, which is known to initiate the expressionof oligodendrocyte-specific proteins in collaboration withOlig2 [29,34]. This finding is consistent with the recentreports that overexpression of Olig2 transcription factor issufficient to induce the expression of Nkx2.2 and differen-tiation of neural stem/progenitor cells into oligodendro-cytes [35,36]. Taken together, these results indicate thatforced expression of Olig2 is sufficient to activate the cel-lular machinery in NSCs that favors phenotypic differen-tiation into oligodendrocytes.Although F3.Olig2 cells in the white matter differentiatedinto APC-CC1 positive oligodendrocytes, the same cellsthat remained in the gray matter exhibited different bio-logical behaviors. F3.Olig2 cells in the gray matter veryinfrequently differentiated into mature oligodendrocytesor any other neural cells. F3 NSCs, which are mostly con-fined around the lesioned areas, also showed very poordifferentiation potential. This phenomenon indicates thatlesion environment of the spinal cord (especially grayregions around the lesion cavities in this study) inhibitsdifferentiation of transplanted cells as previously reportedproperties may not be sufficient to override the influenceof lesion environment. Alternatively, adequate environ-mental factors, which are produced by the white matter,might be required for proper differentiation of F3.Olig2cells. F3.Olig2 cells also exhibited much higher capacityfor migration towards the white matter than parental F3cells. It is possible that demyelinating but intact axonsmay produce certain molecular cues to recruit oli-godendrocyte lineage cells from the transplantation sites[39]. Most of the parental F3 NSCs without Olig2 overex-pression, however, might not possess capacity to respondto such a signal. These observations may exemplify theimportance of an adequate crosstalk between cell-autono-mous traits of transplanted cells and host environmentalfactors for the successful integration of exogenous cellulargrafts.Transplantation of F3.Olig2 NSCs significantly increasedthe volume of myelinated white matter. Furthermore, ani-mals with F3.Olig2 grafts showed significantly highermyelin ratio in the spared white matter. These data suggesta possibility that F3.Olig2 cells differentiated into func-tional oligodendrocytes and participated in remyelinatingdenuded axons in the residual ventrolateral white matter.It is also possible that F3.Olig2 NSCs in the spared whitematter exerted trophic effects [40], which could preventfurther demyelination and promote endogenous remyeli-nation process [41]. Our finding that F3 NSC grafts, whichrarely differentiated into oligodendrocytes, also increasedmyelin ratio to some extent supports the notion that thetrophic mechanisms were indeed in operation. Therefore,the higher myelin ratio in animals with transplantation ofF3.Olig2 NSCs could be accounted for by both directremyelination and trophic effects by the grafted cells.We found in the present study that the overall number ofF3.Olig2 cells detected at 7 weeks after transplantationwas larger than that of NSCs across the rostrocaudal extentof the spinal cord examined. A recent study has reportedthat Olig2 critically regulates replication competence inneural stem cells and malignant glioma [30]: for example,loss of Olig2 resulted in a dramatic reduction of neuralstem cell proliferation. Our data also demonstrated thatOlig2 overexpressing human NSCs possess higher prolif-erative capacity in vitro and in vivo. Therefore, the largernumber of F3.Olig2 cells detected after the transplanta-tion could be explained by the influence of Olig2 gene onproliferation potential of NSCs. Despite the proliferation-promoting effect of Olig2 gene, we did not observe a grosstumor formation. This is consistent with the fact thatOlig2 is not sufficient for brain cancer formation [30]. Itis conceivable that the increase in the number of graftedcells may have contributed to a better preservation of spi-Page 10 of 16(page number not for citation purposes)[37,38]. In F3.Olig2 cells that do not migrate towards thewhite matter, genetic modification of differentiationnal cord tissue in animals with F3.Olig2 grafts. Many ofF3.Olig2 or F3 cells that stayed in the gray matter orBMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117around the lesion cavity did not acquire any mature phe-notype. These undifferentiated progenitor cells can exertneuroprotective effects by paracrine actions from secretedmolecules or via an immunomodulatory mechanism[42,43]. Furthermore, undifferentiated NSCs can promoteaxonal sprouting/regeneration by producing neuro-trophic factors or acting as a kind of cellular guidance[44,45]. F3 human NSCs have been shown to producevarious neurotrophic factors [26,28]. The presence of amuch larger number of grafted NSCs in F3.Olig2 groupcould promote tissue sparing or axonal growth of a muchlarger magnitude, which then resulted in the reduction ofcavity volume. It is likely that the neuroprotective and/orneuroregenerative mechanism by grafted F3.Olig2 cells,together with enhanced myelination in the white matter,contributed to improvement in the quality of locomotion.The animals grafted with F3.Olig2 cells improved qualityof hindlimb locomotion as assessed by BBB score andfootprint analysis. They exhibited better coordinationbetween hind- and forelimbs. The smaller distancebetween the two hindlimbs (interlimbs distance) in ani-mals with F3.Olig2 grafts suggests an improved balanceduring locomotion. However, there was no difference inthe number of errors in grid walk test between the treat-ment groups. Walking on the grid runway requires precisemotor control (hindlimbs placement) integrated withsensory information on the grid and thus depends on theintegrity of supraspinal projections more heavily thanspontaneous locomotor behavior [46,47]. Regainingsupraspinal control may not be achieved merely byenhancement of myelination without regeneration of sev-ered axons. In this regard, it may be necessary to combinestrategies that are aimed at inducing axonal regenerationto achieve more meaningful functional improvementafter SCI.Several studies have highlighted the therapeutic benefitsof supplying new oligodendrocytes by transplantation ofstem/progenitor cells following SCI [11,12,15,48]. Trans-plantation of retinoic acid-treated or predifferentiatedembryonic stem cells resulted in differentiation into oli-godendrocyte lineage and improvement of behavioralrecovery after SCI [12,48]. Although no tumor formationwas reported, the possibility of teratocarcinoma forma-tion in a longer term still raises a safety issue in usingembryonic stem cells. Transplantation of multipotent orglial lineage-restricted neural stem/progenitor cells alsosuccessfully promoted myelination and enhanced func-tional recovery [11,15]. However, the efficiency of oli-godendrocyte differentiation might be unsatisfactory oradditional measures would be needed to improve the dif-ferentiation efficiency [49]. The present study is the first toferentiation following SCI. The beneficial outcomes of thecurrent approach would provide a promising alternativeto supply myelinating oligodendrocytes for spinal cordrepair.ConclusionThe present study showed that transplantation of humanNSCs genetically modified to express Olig2 transcriptionfactor into the contused spinal cord enhances the extent ofmyelination in the spared white matter and improvedlocomotor recovery. Transplantation of neural stem/pro-genitor cells genetically modified to differentiate into oli-godendrocytic lineage may be an effective strategy toimprove functional outcomes following traumatic inju-ries to the spinal cord. Our study further suggests thatmanipulation of molecular factors governing cell fatedecisions during development can influence the fate ofgrafted neural stem/progenitor cells and positively affectthe reparative potential of the transplantation therapy.MethodsCulture of human neural stem cellsPrimary dissociated cell cultures of fetal human telen-cephalon tissues of 14 weeks gestation were prepared asdescribed previously [23,50,51]. The cells were grown inT25 flasks in Dulbecco's modified Eagle medium (DMEM;HyClone, Logan, UT), supplemented with high glucose,5% fetal bovine serum (FBS), 20 μg/ml gentamicin(Sigma, St Louis, MO), and 2.5 μg/ml amphotericin B(Sigma). The medium was changed twice a week. The per-mission to use the fetal tissues was granted by the ClinicalResearch Screening Committee involving Human Subjectsof the University of British Columbia, and the fetal tissueswere obtained from the Anatomical Pathology Depart-ment of Vancouver General Hospital.PA317 amphotropic packaging cell line was infected withthe recombinant replication-incompetent retroviral vec-tor pLNX.v-myc, and the supernatants from the packagingcells were used to infect NSCs in human fetal telen-cephalon cultures. Stably transfected colonies wereselected by neomycine resistance. Several stable clones ofhuman NSCs were isolated, and one of them, HB1.F3 (F3hereafter), was expanded for the present study. F3 humanNSCs express ABCG2, nestin, and Musashi1, which arecell type specific markers for NSCs [28,52,53].To generate Olig2 overexpressing human NSC line(F3.Olig2), Olig2 cDNA (a generous gift from Dr. Takeba-yashi, Okazaki, Japan) was ligated into multiple cloningsites of the retroviral vector pLPCX. PA317 amphotropicpackaging cell line was infected with the recombinant ret-roviral vector, and the supernatants from the packagingPage 11 of 16(page number not for citation purposes)utilize a molecular factor governing the fate determina-tion of oligodendrocytes to enforce oligodendrocytic dif-cells were added to the F3 cells. Stably transfected colonieswere selected by puromycine resistance.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117RT-PCR and immunocytochemistryFor RT-PCR analysis, NSCs were grown on poly-L-lysinecoated Petri dishes in DMEM with 2% FBS for 3 days.Total RNA was extracted from cultured cells using Trizol(GIBCO-BRL). One μg of total RNA was reverse-tran-scribed into first-strand cDNA using oligo-dT primer(Promega, Madison, WI). For PCR amplification, specificprimer pairs were incubated with 1 μl of cDNA in a 20 μlreaction mixture containing Taq polymerase. Thesequences of the primers were as follows; AAATCGCATC-CAGATTTTC/CACTGCCTCCTAGCTTGTC for Olig2,TCTACGACAGCAGCGACAAC/CTTGGAGCTTGAGTC-CTGAG for Nkx2.2, CATGACCACAGTCCATGCCAT-CACT/TGAGGTCCACCACCCTGTTGCTGTA for GAPDH.To examine phenotypic differentiation, human NSCswere grown on poly-L-lysine coated 9 mm Aclar fluorocar-bon plastic coverslips in DMEM with 2% FBS for 5 days.Cells were washed three times with phosphate bufferedsaline (PBS) and then fixed with 4% paraformaldehydefor 10 minutes. After blocking with 5% normal goatserum, cells were incubated with primary antibodies for 2hours at room temperature (RT). The following primaryantibodies were used to examine the expression of neuralcell phenotypic markers; anti-O4 (1:5; mouse mono-clonal; Kim Lab), anti-galactocerebrosidase (GalC) (1:5;mouse monoclonal; Kim Lab), anti-myelin basic protein(MBP) (1:500; rabbit polyclonal; Chemicon, Temecula,CA), anti-glial fibrillary acidic protein (GFAP) (1:500; rab-bit polyclonal; Chemicon), and anti-MAP2 (1:200; rabbitpolyclonal; Chemicon). The coverslips were then incu-bated with appropriate secondary antibodies tagged withAlex-488 or Alexa-594 fluorophores (Molecular probes,Eugene, OR) for an hour. The nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI; Sigma) beforemounting on the slides.Proliferation assayGrowth rate of NSCs was determined by measuring viablecell numbers at different time points after plating. Thenumber of cells was measured by Cell Counting Kit-8(Dojindo Laboratories, Kumamoto, Japan) according tothe manufacturer's instructions. Briefly, cells were dis-pensed as 100 μl of cell suspension (1000 cells/well) in a96-well plate. Ten μl of CCK-8 solutions were added toeach well at different time points (0, 6, 12, 18, 24, 48, and72 hours), and the plate was incubated at 37°C for 4 h.The absorbance value at 450 nm wavelength was meas-ured with a dual-beam microtiter plate reader. The exper-iment was replicated four times with triplicate samplesincluded in each experiment.For 5-bromo-2'-deoxyuridine (BrdU) incorporationhours. BrdU (final concentration 2 μM, Sigma) was addedto each well for 2 hours followed by immunocytochemis-try as described above using anti-BrdU (1:500; rat mono-clonal; Oxford, UK). Counting was performed in threerandomly selected fields (×200) and the average value wasobtained from three coverslips for each condition. Thepercent of BrdU positive cells out of all DAPI stainednuclei was obtained as BrdU incorporation index.Animals and surgeryAdult female Sprague Dawley rats (250 - 300 gram, OrientBio Inc. Seongnam, Korea) were used in this study. All ratswere maintained on a 12:12 hour dark and light cyclewith food and water provided ad libitum. After being anes-thetized with 4% chloral hydrate (10 ml/kg, injectedintraperitoneally), rats were subjected to a dorsal laminec-tomy at the 9th thoracic vertebral level (T9-10) to exposethe dorsal surface of the spinal cord. The NYU spinal cordimpactor was used to inflict a standardized contusion; a10 g impactor head was dropped from a height of 12.5mm onto the exposed T9-10 spinal cord. Muscles and sub-cutaneous tissues were sutured in layer, and the skin wasstapled. Sham operation involved only laminectomywithout contusion on the spinal cord. The bladder wasmanually expressed twice daily until the animals resumedself voiding. Seven days after spinal injury, the spinal cordwas reexposed for cellular transplantation. Animals wererandomly divided into three groups: 1) vehicle (PBS)injected (Vehicle group), 2) transplantation of F3 cells (F3group), and 3) transplantation of F3.Olig2 cells (F3.Olig2group). Two injections were made at 2 mm rostral andcaudal to the epicenter using a glass micropipette (tipdiameter < 70 μm) configured with Hamilton syringe. Thepipette pierced the dorsal spinal cord slightly off the dor-sal median sulcus, avoiding blood vessels at the midline.We advanced the pipette with a depth of 1.2 mm from thedorsal surface and kept it for 3 minutes during the injec-tion which was controlled by Nanoliter syringe pump (KDscientific; Holliston, MA, USA). Each injection consistedof 1 × 105 of the dissociated cells in 2 μl of PBS. Thus, atotal of 2 × 105 cells were transplanted for each animal. Toprevent leakage from the injection site, the pipette wasmaintained for three more minutes and then withdrawnslowly. Animals were assigned with new identificationcodes after transplantation to ensure blind evaluation ofbehavioral performance. All animals received daily intra-peritoneal injection of cyclosporine (Sandimmun;Novartis, Bern, Switzerland) at a dosage of 10 mg/kg,beginning from one day prior to transplantation to threeweeks after transplantation. After that, cyclosporine (50μg/ml) was administered through drinking water untilanimals were sacrificed. Prophylactic antibiotics wereintraperitoneally injected on the next day after each sur-Page 12 of 16(page number not for citation purposes)experiment, 100 μl of cell suspension (1000 cells/well)was plated on a 9 mm coverslip and incubated for 36gery.BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117Assessment of locomotor recoveryA total of 35 animals (N = 11, 12, and 12 for Vehicle, F3,and F3.Olig2 groups, respectively), divided in three series,underwent behavioral tests to assess locomotor recovery.The BBB (Basso, Beattie, and Bresnahan) locomotor ratingscale was used to assess the extent of locomotor recoveryduring open field locomotion. Two experimenters whowere blind to experimental conditions scored BBB scaleseparately and the average of the two scores was obtained.The grid walk test was conducted at 4 and 7 weeks afterinitial injury. We performed grid walk test for the secondand third series of animals, so the animal in the first serieswere excluded (total 5; Vehicle = 1, F3 = 3, F3.Olig2 = 2).For grid walk, rats were trained to cross a grid runway (30cm × 140 cm with 50 × 50 mm holes) for a water reward.Two animals in Vehicle group and one animal in F3.Olig2group could not be trained adequately enough to undergogrid walk test. Therefore, grid walk data were obtainedfrom 27 rats (Vehicle = 8, F3 = 9, F3.Olig2 = 9). On the dayof test, four runs were recorded using a three-CCD digitalvideo camera (NV-GS250, Panasonic, Japan), and lateranalyzed frame by frame in slow motion. The averagenumber of limb placement errors per run was obtained foreach animal. For footprint analysis, rats were trained towalk on a runway with a narrow alley made of transparentPlexiglas (7 cm × 170 cm) on top for a water reward dur-ing the same training session for grid walk. The animals'hindpaws were inked and footprints were obtained onwhite paper covering the floor of the runway. The base ofsupport, which is a distance between the two hindlimbs,was determined by measuring the distance between thecentral pads of both hindpaws. Right and left stridelengths were measured between two consecutive prints oneach side. Footprint analysis was performed only in thethird series of animals (total 14; Vehicle = 4, F3 = 5,F3.Olig2 = 5). Normal foot print data were obtained from5 rats with sham operation.Histological processing and immunohistochemistryFor histological analysis, animals were anesthetized withan overdose of chloral hydrate and perfused withheparinized saline (0.9%), followed by 4% paraformalde-hyde in 0.1 M phosphate buffer, pH 7.4. The spinal cordwas dissected and post-fixed in 4% paraformaldehyde atRT for 2 hours, followed by cryoprotection in a gradedseries of sucrose solutions (10%-30%) in 0.1 M phos-phate buffer at 4°C. Transverse sections (20 μm) of thespinal cord were cut using cryostat (Leica CM 1900; Wet-zlar, Germany) in a 1:10 series and thaw-mounted ontosilane-coated glass slides. To quantify the area of myeli-nated white matter, transverse spinal cord sections werestained with eriochrome cyanine that stains myelinatedwhite matter [54]. The transverse sections were immersed10% FeCl3·6H2O (Sigma) in 3% HCl. The sections werethen washed with running tap water, followed by differen-tiation in 1% aqueous NH4OH.For immunohistochemistry, transverse spinal cord sec-tions were incubated overnight at 4°C with anti-humanmitochondria (1:400; mouse monoclonal; Chemicon),anti-Ki-67 (1:500; rabbit polyclonal; Novocastra, Newcas-tle, UK), anti-myelin basic protein (MBP) (1:400; rabbitpolyclonal; Chemicon), anti-APC-CC1 (1:200; mousemonoclonal; Calbiochem, La Jolla, CA), anti-GFAP(1:500; rabbit polyclonal; Chemicon), anti-nestin (1:500;rabbit polyclonal; Chemicon), and anti-MAP2 (1:200:rabbit polyclonal; Chemicon). The spinal cord sectionswere washed and then incubated with biotinylated orfluorophore-tagged secondary antibodies. For chromoge-nic detection of antigen-antibody reaction, preformed avi-din-biotinylated peroxidase complexes were applied for30 minutes, followed by incubation with peroxidase sub-strate (DAB) until desired intensity developed. For fluo-rescence staining, coverslips were mounted onto glassslides using Gelvatol and examined under Olympus con-focal laser scanning microscope (Model FV 300, Tokyo,Japan).To analyze the thickness of myelin sheath wrapping indi-vidual axons, a separate series of animals (16 rats; Vehicle= 3, F3 = 6, F3.Olig2 = 7) were perfused with modifiedKarnovsky's fixative solution (2% glutaraldehyde and 1%paraformaldehyde in 0.1 M cacodylate buffer). Dissectedspinal cord was divided into smaller blocks with about 5mm in length. The tissue block containing caudal regionsto the epicenter was transversely sectioned into 200 μmslices with a vibratome (model Vibratome Series 1000;Technical Products Int'l. Inc., O'Fallon, MO). The slicescontaining the regions around 1 mm caudal to the epi-center were immersed in 1% OsO4 solution, and thenembedded in Poly/Bed 812 embedding media (Poly-sciences Inc., Warrington, PA). Transverse semithin (1μm) sections were cut from the rostral surface with glassknife and then stained with toluidine blue.Quantitative cell counting and image analysisHuman mitochondria positive cells (developed with DABsubstrates) were counted using unbiased stereology. Twotransverse sections 200 μm apart from each other at theepicenter, ± 1 mm, ± 2 mm, ± 3 mm, and ± 4 mm rostraland caudal from the epicenter were chosen. Cell countingwas performed on an Olympus BX51 Microscope with aMAC 6000 Motorized Stage Encoder System that was cou-pled with a computer running Stereo Investigator 8 soft-ware (MBF Bioscience, Williston, VT). Humanmitochondria positive cells within optical dissectors ran-Page 13 of 16(page number not for citation purposes)for 8 minutes in the staining solution consisting of 240 mlof 0.2% eriochrome cyanine RC (Sigma) and 10 ml ofdomly placed in regions of interest were counted usingstandard stereological criteria for inclusion. StereologicalBMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117estimates were done in a systematic way using the formulain the software. The average of the two sections at eachlevel was obtained for each animal. The percentage of thecell number in the spared ventrolateral white matter wascalculated by dividing the number of cells in the ventrola-teral white matter by the total number in the entire sec-tion. To determine the percentage of grafted human NSCsthat were colocalized with proliferation marker Ki67 orneural cell specific markers, two transverse sections withthe highest graft survival were chosen for each animal. Thenumber of human NSCs colocalized with these markerswas counted in the entire section and divided by the totalnumber of human mitochondria positive cells. The aver-age from the two sections was calculated as the final per-centage value for each animal.The volume of myelinated white matter and cystic cavitieswere determined for the animals that completed behavio-ral tests. Every 10th eriochrome-stained transverse spinalcord sections were viewed on an optical microscope(Olympus BX41). The section containing injury epicenterwas defined visually as the one with a smallest visible rimof spared myelin. Then, serial sections with an equal dis-tance (400 μm) spanning ± 2 mm from epicenter wereimaged at a 40× magnification and captured with CCDcamera (Olympus DP11). Myelinated spared ventrola-teral white matter anterior to the dorsal horn was drawnwith the Pen Tablet input device (Bamboo MTE-450K;Wacom Co., Tokyo, Japan) on each image and the cross-sectional areas were measured using publicly availableImage J software (NIH, Bethesada, MD). The total esti-mated volume was calculated using the Cavalieri's Princi-ple. The individual subvolumes were obtained bymultiplying the cross-sectional area by the distancebetween sections, and the subvolumes were summed togenerate the total volume of spared white matter (∑n[cross sectional areas × intersection distance], n = numberof sections analyzed). The volume of cystic cavities wascalculated by the same equation.Thickness of myelin sheath in individual axons was deter-mined on toluidine blue stained transverse semithin sec-tions. Two sections with adequate preparation wereselected for each animal. In order to have correspondingregions imaged across all the animals, a single image field(120 μm × 90 μm) per section was located in the ventralwhite matter just anterior to the ventral gray horn by anindependent experimenter who was blind to the experi-mental conditions. Images were taken at 1000× magnifi-cation using a Zeiss Axiophot upright microscopyequipped with Axiocam HR digital camera. Horizontalgrid lines with a 10 μm interval were drawn using Pho-toshop software and the same grid lines were applied toof myelin sheath and the shortest axon diameter weremeasured using the Image J software. According to thepreviously published report by Karimi-Abdolrezaee et al.(2006) [15], the myelin ratio was obtained by dividingthe total axonal diameter including thickness of myelinsheath by the diameter of axonal fiber excluding myelinsheath.Statistical methodsStatistical analysis was performed with SPSS version 12.0(Chicago, IL, USA) or GraphPad Prism software version4.0 (San Diego, CA, USA). The unpaired Student T test orone-way ANOVA followed by Tukey's post hoc test wasused for statistical comparison of group means. Repeatedmeasures two-way ANOVA was used to compare thenumber of cells or the percent cells in the white matter atdifferent regions and BBB locomotor scores at multipletime points.List of abbreviationsSCI: (spinal cord injury), NSC: (neural stem cell); CNS:(central nervous system); PBS: (phosphate bufferedsaline)Authors' contributionsDHH carried out in vivo studies, participated in experi-mental design, helped in drafting the manuscript. BGKconceived the study, helped in acquisition and interpreta-tion of data, wrote the manuscript, and gave finalapproval of the version to be published. EJK carried out invitro studies and participated in acquisition of in vitrodata. SIL helped in conduction of experiments and partic-ipated in acquisition and interpretation of data. ISJ partic-ipated in study design and participated in criticalreviewing of the manuscript. HSK participated in studydesign and interpretation of data. SS helped in acquisitionand interpretation of data on myelination. SUK conceivedthe study, helped in writing and reviewing the manu-script, and gave final approval of the version to be pub-lished. All authors read and approved the final version ofthe manuscript.Additional materialAdditional file 1Immunocytochemical detection of differentiation into mature astro-cytes or neurons. F3 (A, C) or F3.Olig2 (B, D) cells were grown on cov-erslips in DMEM with 2% FBS for 5 days and then fixed with 4% paraformaldehyde. Then the cells were stained with anti-GFAP (A, B) or anti-MAP2 (C, D) antibodies. Some of F3 cells showed differentiation into astrocyte or neurons, but virtually no F3.Olig2 cells were immunore-active against GFAP or MAP2.Click here for file[http://www.biomedcentral.com/content/supplementary/1471-Page 14 of 16(page number not for citation purposes)all images. Only the axons that were intersected by thegrid lines were included for quantification. The thickness2202-10-117-S1.TIFF]BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/117AcknowledgementsThe study was supported by research funds from the Guwon scholarship foundation, by the Korean Science and Engineering Foundation (KOSEF) through Chronic Inflammatory Disease Research Center Ajou University Grant (R13-2003-019), by Ministry of Commerce, Industry and Energy (10024156) (to BGK), and by Canadian Myelin Research Initiative (to SUK).References1. Basso DM, Beattie MS, Bresnahan JC: Graded histological andlocomotor outcomes after spinal cord contusion using theNYU weight-drop device versus transection.  Exp Neurol 1996,139:244-256.2. Fehlings MG, Tator CH: The relationships among the severityof spinal cord injury, residual neurological function, axoncounts, and counts of retrogradely labeled neurons afterexperimental spinal cord injury.  Exp Neurol 1995, 132:220-228.3. Noble LJ, Wrathall JR: Correlative analyses of lesion develop-ment and functional status after graded spinal cord contu-sive injuries in the rat.  Exp Neurol 1989, 103:34-40.4. Reier PJ: Cellular transplantation strategies for spinal cordinjury and translational neurobiology.  NeuroRx 2004,1:424-451.5. Bunge RP, Puckett WR, Becerra JL, Marcillo A, Quencer RM: Obser-vations on the pathology of human spinal cord injury. Areview and classification of 22 new cases with details from acase of chronic cord compression with extensive focal demy-elination.  Adv Neurol 1993, 59:75-89.6. Guest JD, Hiester ED, Bunge RP: Demyelination and Schwanncell responses adjacent to injury epicenter cavities followingchronic human spinal cord injury.  Exp Neurol 2005,192:384-393.7. Totoiu MO, Keirstead HS: Spinal cord injury is accompanied bychronic progressive demyelination.  J Comp Neurol 2005,486:373-383.8. Crowe MJ, Bresnahan JC, Shuman SL, Masters JN, Beattie MS: Apop-tosis and delayed degeneration after spinal cord injury in ratsand monkeys.  Nat Med 1997, 3:73-76.9. Springer JE, Azbill RD, Knapp PE: Activation of the caspase-3apoptotic cascade in traumatic spinal cord injury.  Nat Med1999, 5:943-946.10. Nashmi R, Fehlings MG: Mechanisms of axonal dysfunction afterspinal cord injury: with an emphasis on the role of voltage-gated potassium channels.  Brain Res Brain Res Rev 2001,38:165-191.11. Cao Q, Xu XM, Devries WH, Enzmann GU, Ping P, Tsoulfas P, WoodPM, Bunge MB, Whittemore SR: Functional recovery in trau-matic spinal cord injury after transplantation of multineuro-trophin-expressing glial-restricted precursor cells.  J Neurosci2005, 25:6947-6957.12. Keirstead HS, Nistor G, Bernal G, Totoiu M, Cloutier F, Sharp K,Steward O: Human embryonic stem cell-derived oligodendro-cyte progenitor cell transplants remyelinate and restorelocomotion after spinal cord injury.  J Neurosci 2005,25:4694-4705.13. Bambakidis NC, Miller RH: Transplantation of oligodendrocyteprecursors and sonic hedgehog results in improved functionand white matter sparing in the spinal cords of adult ratsafter contusion.  Spine J 2004, 4:16-26.14. Hofstetter CP, Holmstrom NA, Lilja JA, Schweinhardt P, Hao J,Spenger C, Wiesenfeld-Hallin Z, Kurpad SN, Frisen J, Olson L: Allo-dynia limits the usefulness of intraspinal neural stem cellgrafts; directed differentiation improves outcome.  Nat Neu-rosci 2005, 8:346-353.15. Karimi-Abdolrezaee S, Eftekharpour E, Wang J, Morshead CM, Feh-lings MG: Delayed transplantation of adult neural precursorcells promotes remyelination and functional neurologicalrecovery after spinal cord injury.  J Neurosci 2006, 26:3377-3389.16. Shirasaki R, Pfaff SL: Transcriptional codes and the control ofneuronal identity.  Annu Rev Neurosci 2002, 25:251-281.17. Rowitch DH: Glial specification in the vertebrate neural tube.Nat Rev Neurosci 2004, 5:409-419.and gliogenesis in the developing spinal cord.  Development2007, 134:1617-1629.19. Zhou Q, Wang S, Anderson DJ: Identification of a novel family ofoligodendrocyte lineage-specific basic helix-loop-helix tran-scription factors.  Neuron 2000, 25:331-343.20. Lu QR, Yuk D, Alberta JA, Zhu Z, Pawlitzky I, Chan J, McMahon AP,Stiles CD, Rowitch DH: Sonic hedgehog--regulated oli-godendrocyte lineage genes encoding bHLH proteins in themammalian central nervous system.  Neuron 2000, 25:317-329.21. Takebayashi H, Yoshida S, Sugimori M, Kosako H, Kominami R,Nakafuku M, Nabeshima Y: Dynamic expression of basic helix-loop-helix Olig family members: implication of Olig2 in neu-ron and oligodendrocyte differentiation and identification ofa new member, Olig3.  Mech Dev 2000, 99:143-148.22. Jakovcevski I, Zecevic N: Olig Transcription Factors AreExpressed in Oligodendrocyte and Neuronal Cells in HumanFetal CNS.  J Neurosci 2005, 25:10064-10073.23. Ryu JK, Kim J, Cho SJ, Hatori K, Nagai A, Choi HB, Lee MC, McLarnonJG, Kim SU: Proactive transplantation of human neural stemcells prevents degeneration of striatal neurons in a rat modelof Huntington disease.  Neurobiol Dis 2004, 16:68-77.24. Kim SU, Park IH, Kim TH, Kim KS, Choi HB, Hong SH, Bang JH, LeeMA, Joo IS, Lee CS, et al.: Brain transplantation of human neuralstem cells transduced with tyrosine hydroxylase and GTPcyclohydrolase 1 provides functional improvement in animalmodels of Parkinson disease.  Neuropathology 2006, 26:129-140.25. Jeong S-W, Chu K, Jung K-H, Kim SU, Kim M, Roh J-K: Human Neu-ral Stem Cell Transplantation Promotes Functional Recov-ery in Rats With Experimental Intracerebral Hemorrhage.Stroke 2003, 34:2258-2263.26. Yasuhara T, Matsukawa N, Hara K, Yu G, Xu L, Maki M, Kim S, Bor-longan C: Transplantation of Human Neural Stem CellsExerts Neuroprotection in a Rat Model of Parkinson's Dis-ease.  J Neurosci 2006, 26:12497-12511.27. Meng XL, Shen JS, Ohashi T, Maeda H, Kim SU, Eto Y: Brain trans-plantation of genetically engineered human neural stemcells globally corrects brain lesions in the mucopolysacchari-dosis type VII mouse.  J Neurosci Res 2003, 74:266-277.28. Lee HJ, Kim KS, Kim EJ, Choi HB, Lee KH, Park IH, Ko Y, Jeong SW,Kim SU: Brain transplantation of immortalized human neuralstem cells promotes functional recovery in mouse intracer-ebral hemorrhage stroke model.  Stem Cells 2007, 25:1204-1212.29. Qi Y, Cai J, Wu Y, Wu R, Lee J, Fu H, Rao M, Sussel L, Rubenstein J,Qiu M: Control of oligodendrocyte differentiation by theNkx2.2 homeodomain transcription factor.  Development 2001,128:2723-2733.30. Ligon KL, Huillard E, Mehta S, Kesari S, Liu H, Alberta JA, Bachoo RM,Kane M, Louis DN, Depinho RA, et al.: Olig2-regulated lineage-restricted pathway controls replication competence in neu-ral stem cells and malignant glioma.  Neuron 2007, 53:503-517.31. Lu QR, Sun T, Zhu Z, Ma N, Garcia M, Stiles CD, Rowitch DH: Com-mon developmental requirement for Olig function indicatesa motor neuron/oligodendrocyte connection.  Cell 2002,109:75-86.32. Takebayashi H, Nabeshima Y, Yoshida S, Chisaka O, Ikenaka K,Nabeshima Y: The basic helix-loop-helix factor olig2 is essen-tial for the development of motoneuron and oligodendro-cyte lineages.  Curr Biol 2002, 12:1157-1163.33. Zhou Q, Anderson DJ: The bHLH transcription factors OLIG2and OLIG1 couple neuronal and glial subtype specification.Cell 2002, 109:61-73.34. Zhou Q, Choi G, Anderson DJ: The bHLH transcription factorOlig2 promotes oligodendrocyte differentiation in collabora-tion with Nkx2.2.  Neuron 2001, 31:791-807.35. Liu Z, Hu X, Cai J, Liu B, Peng X, Wegner M, Qiu M: Induction ofoligodendrocyte differentiation by Olig2 and Sox10: evi-dence for reciprocal interactions and dosage-dependentmechanisms.  Dev Biol 2007, 302:683-693.36. Copray S, Balasubramaniyan V, Levenga J, de Bruijn J, Liem R, BoddekeE: Olig2 overexpression induces the in vitro differentiation ofneural stem cells into mature oligodendrocytes.  Stem Cells2006, 24:1001-1010.37. Cao QL, Zhang YP, Howard RM, Walters WM, Tsoulfas P, Whitte-more SR: Pluripotent stem cells engrafted into the normal orPage 15 of 16(page number not for citation purposes)18. Sugimori M, Nagao M, Bertrand N, Parras CM, Guillemot F, NakafukuM: Combinatorial actions of patterning and HLH transcrip-tion factors in the spatiotemporal control of neurogenesislesioned adult rat spinal cord are restricted to a glial lineage.Exp Neurol 2001, 167:48-58.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Neuroscience 2009, 10:117 http://www.biomedcentral.com/1471-2202/10/11738. Cao QL, Howard RM, Dennison JB, Whittemore SR: Differentia-tion of engrafted neuronal-restricted precursor cells is inhib-ited in the traumatically injured spinal cord.  Exp Neurol 2002,177:349-359.39. Di Bello IC, Dawson MR, Levine JM, Reynolds R: Generation of oli-godendroglial progenitors in acute inflammatory demyeli-nating lesions of the rat brain stem is associated withdemyelination rather than inflammation.  J Neurocytol 1999,28:365-381.40. Zhang YW, Denham J, Thies RS: Oligodendrocyte progenitorcells derived from human embryonic stem cells express neu-rotrophic factors.  Stem Cells Dev 2006, 15:943-952.41. Yang H, Lu P, McKay HM, Bernot T, Keirstead H, Steward O, GageFH, Edgerton VR, Tuszynski MH: Endogenous neurogenesisreplaces oligodendrocytes and astrocytes after primate spi-nal cord injury.  J Neurosci 2006, 26:2157-2166.42. Martino G, Pluchino S: The therapeutic potential of neural stemcells.  Nat Rev Neurosci 2006, 7:395-406.43. Pluchino S, Quattrini A, Brambilla E, Gritti A, Salani G, Dina G, GalliR, Del Carro U, Amadio S, Bergami A, et al.: Injection of adult neu-rospheres induces recovery in a chronic model of multiplesclerosis.  Nature 2003, 422:688-694.44. Pfeifer K, Vroemen M, Blesch A, Weidner N: Adult neural progen-itor cells provide a permissive guiding substrate for corticos-pinal axon growth following spinal cord injury.  Eur J Neurosci2004, 20:1695-1704.45. Lu P, Jones LL, Snyder EY, Tuszynski MH: Neural stem cells con-stitutively secrete neurotrophic factors and promote exten-sive host axonal growth after spinal cord injury.  Exp Neurol2003, 181:115-129.46. Kunkel-Bagden E, Dai HN, Bregman BS: Methods to assess thedevelopment and recovery of locomotor function after spi-nal cord injury in rats.  Exp Neurol 1993, 119:153-164.47. Schucht P, Raineteau O, Schwab ME, Fouad K: Anatomical corre-lates of locomotor recovery following dorsal and ventrallesions of the rat spinal cord.  Exp Neurol 2002, 176:143-153.48. McDonald JW, Liu XZ, Qu Y, Liu S, Mickey SK, Turetsky D, GottliebDI, Choi DW: Transplanted embryonic stem cells survive, dif-ferentiate and promote recovery in injured rat spinal cord.Nat Med 1999, 5:1410-1412.49. Kim BG, Hwang DH, Lee SI, Kim EJ, Kim SU: Stem cell-based celltherapy for spinal cord injury.  Cell Transplant 2007,16(4):355-364.50. Flax JD, Aurora S, Yang C, Simonin C, Wills AM, Billinghurst LL, Jen-doubi M, Sidman RL, Wolfe JH, Kim SU, et al.: Engraftable humanneural stem cells respond to developmental cues, replaceneurons, and express foreign genes.  Nat Biotechnol 1998,16:1033-1039.51. Kim SU: Antigen expression by glial cells grown in culture.  JNeuroimmunol 1985, 8:255-282.52. Ryu JK, Choi HB, Hatori K, Heisel RL, Pelech SL, McLarnon JG, KimSU: Adenosine triphosphate induces proliferation of humanneural stem cells: Role of calcium and p70 ribosomal proteinS6 kinase.  J Neurosci Res 2003, 72:352-362.53. Kim SU: Human neural stem cells genetically modified forbrain repair in neurological disorders.  Neuropathology 2004,24:159-171.54. Rabchevsky AG, Fugaccia I, Turner AF, Blades DA, Mattson MP, ScheffSW: Basic fibroblast growth factor (bFGF) enhances func-tional recovery following severe spinal cord injury to the rat.Exp Neurol 2000, 164:280-291.yours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralPage 16 of 16(page number not for citation purposes)


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items