- Email: [email protected]
Human neural stem cell-derived cholinergic neurons innervate muscle in motoneuron deficient adult rats
Neuroscience 131 (2005) 257–262
RAPID REPORT HUMAN NEURAL STEM CELL-DERIVED CHOLINERGIC NEURONS INNERVATE MUSCLE IN MOTONEURON DEFICIENT ADULT RATS J. GAO, R. E. COGGESHALL, Y. I. TARASENKO AND P. WU*
do not form typical motoneurons in adult spinal cord presumably because of lack of sufficient environmental cues (Svendsen et al., 1997; Sheen et al., 1999; Shihabuddin et al., 2000; Cao et al., 2001; Fricker et al., 1999; Vroemen et al., 2003). For example, most human embryonic germ cells, when grafted into rats with diffuse motor neuron injury, remain undifferentiated and only a few become cholinergic, and these do not connect to muscle (Kerr et al., 2003). To obtain a higher proportion of stem cellderived cholinergic neurons, we developed an in vitro priming technology that resulted in human fetal NSCs successfully differentiating into cholinergic neurons in vitro, and many acquire a cholinergic phenotype when grafted into intact adult spinal cord (Wu et al., 2002a). In this study, we show that these primed NSCs, when grafted into adult rat spinal cord with motoneuron degeneration, develop into cholinergic neurons that innervate peripheral muscle with concomitant improvement of motor function.
Department of Neuroscience and Cell Biology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-1043, USA
Abstract—Motoneuron damage occurs in spinal cord injury and amyotrophic lateral sclerosis. Current advances offer hope that human embryonic stem cells [Science 282 (1998) 1145] or neural stem cells (NSC) [Exp Neurol 161 (2000) 67; Exp Neurol 158 (1999) 265; J Neurosci Methods 85 (1998) 141; Proc Natl Acad Sci USA 97 (2000) 14720; Exp Neurol 156 (1999) 156] may be donors to replace lost motoneurons. Previously, we developed a priming procedure that produced cholinergic cells that resemble motoneurons from human NSCs grafted into adult rat spinal cord [Nat Neurosci 5 (2002a) 1271]. However, effective replacement therapy will ultimately rely on successful connection of new motoneurons with their muscle targets. In this study, we examined the potential of human fetal NSC transplantation to replace lost motoneurons in an animal model of chronic motoneuron deficiency (newborn sciatic axotomy) [J Comp Neurol 224 (1984) 252; J Neurobiol 23 (1992) 1231]. We found, for the first time, that human neural stem cell-derived motoneurons send axons that pass through ventral root and sciatic nerve to form neuromuscular junctions with their peripheral muscle targets. Furthermore, this new cholinergic innervation correlates with partial improvement of motor function. © 2005 IBRO. Published by Elsevier Ltd. All rights reserved.
EXPERIMENTAL PROCEDURES Human NSCs and AAVegfp vectors The K048 line of hNSCs (Svendsen et al., 1998) were propagated as neurospheres in growth medium supplemented with epidermal growth factor, basic fibroblast growth factor (bFGF) and leukemia inhibitory factor (Wu et al., 2002a). For priming, 2-day neurospheres (passages 25–31) were treated with 20 ng/ml bFGF, 5 g/ml heparin, 1 g/ml laminin and 50 ng/ml mouse sonic hedgehog amino-terminal peptide (R&D Systems, Minneapolis, MN, USA) for 5 days and then cultured with 1⫻ B27 (Invitrogen, Carlsbad, CA, USA) diluted in DMEM/F12 (Invitrogen) for 2 days in vitro. Four days before transplantation, primed cells were labeled by transduction with a recombinant adeno-associated viral AAVegfp (Wu et al., 2002b). For unprimed cell transplants, proliferating neurospheres were treated with AAVegfp directly and maintained in medium containing growth factors for 4 days.
Key words: neuromuscular junction, axotomy, motoneuron disease, cell therapy.
To replace lost motoneurons, important issues are 1) to direct stem cells to differentiate into cholinergic motoneurons in the ventral horn of the spinal cord, 2) to have enough cues to allow new motoneurons to send axons through appropriate nerves to form synapses on muscle cells, and 3) to show concomitant improved motor function. Several groups generated cholinergic neurons from mouse embryonic stem (ES) cells in vitro by treating them with inductive factors such as retinoic acid and sonic hedgehog (Renoncourt et al., 1998; Wichterle et al., 2002; Barberi et al., 2003). However, these ES or neural stem cells (NSCs)
Neonatal sciatic axotomy and transplantation Left sciatic nerves of newborn Sprague–Dawley rats were crushed mid- thigh for 30 s using smooth-tipped forceps. Two months later, axotomized rats were divided into three groups to receive 1) primed cell grafts, 2) unprimed cell grafts or 3) vehicle. The spinal cord transplantation protocol was established according to the NIH guidelines for the care and use of laboratory animals, and approved by The University of Texas Medical Branch IACUC. Approximately 105 primed or unprimed hNSCs/2 l or 2 l vehicle were implanted into L4 motoneuron deficient spinal ventral horns (ML: ⫹0.7 mm; DV: ⫺1.5 mm from dura). All animals were treated with the immunosuppressor NEORAL cyclosporine (Novartis Pharmaceuticals Co., East Hanover, NJ, USA), 100 g/ml in drinking water 3 days before surgery and thereafter.
*Corresponding author. Tel: ⫹1-409-772-2748; fax: ⫹1-409-7721861. E-mail address: [email protected] (P. Wu). Abbreviations: bFGF, basic fibroblast growth factor; ChAT, choline acetyltransferase; CTB, cholera toxin subunit B; ES, embryonic stem; LAMP2, lysosomal-associated membrane protein 2; NSC, neural stem cells; SFI, sciatic function index; WGA, wheat germ agglutinin.
0306-4522/05$30.00⫹0.00 © 2005 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.10.033
257
258
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 1. Confocal images of primed human NSCs grafted in adult rats with motoneuron deficiency. (a) A view of the ventral horn with many GFP-labeled cells that resemble motoneurons; inset, a low power view showing the whole hemi-cord at the graft site. Grafted hNSCs send numerous large GFP fibers into the ventral root (VR) (b) and sciatic nerve at mid-thigh (c). * Ventral white matter of the spinal cord. Scale bars⫽200 m.
Gait analysis Motor function of rats with neonatal sciatic axotomy was evaluated before transplantation and then 1 or 3 months after lesion by a sciatic function index (SFI; de Medinaceli et al., 1982). SFI of ⫺100% represents a complete lost of gait, SFI of 0⫾11% is a normal gait.
Retrograde tracing Animals received injections of 7.5 l of Alexa 594-wheat germ agglutinin (WGA; 10 g/l; Molecular Probes, Eugene, OR, USA) into the sciatic nerve, or 10 l of Alexa 594-cholera toxin subunit B (CTB; 1 g/l; Molecular Probes) into the left medial and lateral gastrocnemius muscles. Three days post-WGA injection or 10 days after CTB administration, animals were perfused with 4% paraformaldehyde.
Immunohistochemistry Five animals in each group were perfused with 4% paraformaldehyde 3 months after grafting. L4 –L6 segments of spinal cords and ventral roots plus the sciatic nerves were cryo-sectioned at 45 m. Primary antibodies included choline acetyltransferase (ChAT; Chemicon) at 1:100, Hb9 (Chemicon) at 1:100, human specific lysosomal-associated membrane protein 2 (LAMP2) at 1:1600 (Developmental Studies Hybridoma Bank, Iowa City, IA, USA; Chen et al., 1985), human specific neurofilament 70 (Chemicon) at 1:200 and rat specific neurofilament 200 (Chemicon) at 1:200.
Neuromuscular junction labeling Gastrocnemius muscles from two animals in each group were sectioned at approximately 500 m with a razor blade. The muscle slices were incubated with 1 g/ml of rhodamine-conjugated
␣-bungarotoxin (Molecular Probes) for 1 h at room temperature. Some of the slices were double-labeled with either GFP or human specific neurofilament 70.
Quantitative analyses Sections were imaged confocally. Numbers of GFP-labeled hNSCs in the ventral horn were determined stereologically using a fractionator analysis (Coggeshall, 1992). Percentages of double- or triple-labeled cells or neuromuscular junctions with and without an attached GFP or human neurofilament axon were also determined. Motor function (SFI) was analyzed using repeated measures ANOVA (GraphPad Software, Inc., San Diego, CA, USA).
RESULTS Grafted human NSCs survived and differentiated in motoneuron deficient spinal cords Neonatal sciatic axotomy resulted in an approximately 40% loss of myelinated axons in the L4 ventral root, which is equated to a 40% motoneuron loss, and an average SFI of ⫺50%. Axotomized animals were screened by gait analysis just before implantation, and those with SFI values of ⫺40 to ⫺70% were randomly divided into three groups to receive primed or unprimed hNSC grafts or vehicle. Three months after grafting with primed cells, approximately 500 GFP-labeled cells (514⫾64) appeared in each transplanted ventral horn (Fig. 1a). Many of these cells had radiating dendrites and sent axons into ventral roots (Fig. 1b) and sciatic nerves (Fig. 1c). In contrast, most trans-
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 2. GFP-labeled cells are of human NSC origin. (a– c) The ventral horn showing large cells (arrows) that co-localize GFP and a human specific antibody, LAMP2; d–f, similar cells labeled by Alexa 594WGA, a retrograde tracer injected into sciatic nerves 17 days post transplantation (before grafted cells project into the nerve). Note that GFP and retrogradely labeled cells are in close proximity with no co-localization of label. Scale bar⫽200 m. (g–i) Longitudinal sections of sciatic nerves showing GFP axons (green, g), endogenous rat specific neurofilament stained axons (red, h) and a combined image showing that the two labels in separate axon populations (i) as evidence that neither cell fusion or leaking of label explain our results. Scale bar⫽100 m.
planted unprimed hNSCs remained small and round, with occasional neuron-like GFP cells and fibers in the ventral white matter. Only one GFP fiber was detected in the sciatic nerves from all five unprimed hNSC animals and none in vehicle animals (data not shown). To verify the GFP cells were of human (transplanted) origin, immunostaining with a human specific antibody, LAMP2, showed that all green GFP-labeled cells colocalized with LAMP2 (Fig. 2a– c). We injected the retrograde tracer WGA into the sciatic nerve on the transplanted side at 17 days post-grafting and animals were killed 3 days later, 20 days after transplant and before regenerating axons could get into the sciatic nerve. Thus these WGA⫹ cells are endogenous, and the green GFP (exogenous) and red WGA (endogenous) cells are in close proximity without co-localization of labels (Fig. 2d–f). Further analysis on GFP axons in sciatic nerves, as shown in Fig. 2g–i, revealed no sign of fusion or overlapping of GFP markers with host axons stained by rat specific neurofilament antibodies (red). Human neural stem cell (NSC)-derived motoneuron innervated target muscle Many grafted hNSCs acquired motoneuron morphology (Fig. 2). At 3 months post-transplant, 51% of the GFP cells were cholinergic (Fig. 3a, b) and Hb9 positive (Fig. 3e– g),
259
19% were labeled retrogradely from muscle (Fig. 3a, c) and 14% were both cholinergic and retrogradely labeled (Fig. 3a– d). This indicates that some grafted hNSCs become cholinergic motoneurons that send axons to peripheral muscle targets. Numbers of grafted primed cells with various labeling as well as numbers of peripheral axons from grafted neurons are shown in Table 1. In the unprimed graft group, cell survival rates were similar to those of their primed counterparts, but only a few differentiated into cholinergic neurons, and none were retrogradely labeled (data not shown). To determine whether the terminals of hNSC-derived motoneurons attached to muscle, gastrocnemius slices were stained with rhodamine-conjugated ␣-bungarotoxin that specifically binds to the nicotinic acetylcholine receptor to identify neuromuscular junctions. Human NSCderived motoneuron axons were visualized either by GFP or human specific neurofilament 70 antibody, and 6.4% of the ␣-bungarotoxin⫹ neuromuscular junctions were contacted by GFP- (data not shown) or human specific neurofilament 70-labeled axons (Fig. 3h–j). No neuromuscular junctions were co-labeled with human specific neurofilament antibody in unprimed NSC-grafted or vehicle control rats. Human NSC grafts improved motor function in motoneuron deficient rats Neonatal sciatic nerve crushed rats (n⫽10 per group) showed a marked gait deficiency for the left hind limb (Fig. 4a) with SFI scores of approximately ⫺50% before implantation (Fig. 4b and Table 2). One month after transplantation, a slight but not significant improvement was observed in all groups. At 3 months post-grafting, only rats that received primed hNSC grafts showed statistically significant improvement (Fig. 4b). Such behavioral improvement (Table 2) correlates to some extent with the total numbers of grafted cholinergic neurons that are retrogradely labeled and to a lesser degree with the numbers of GFP axons in peripheral nerves, but not with the total numbers of survived cells (Table 1), although the number of animals sampled is too low for a statistical correlation.
DISCUSSION The present study shows for the first time that motoneurons generated from primed human fetal NSC emit axons that innervate peripheral muscle targets. These findings correlate with partial improvement of motor function and indicate that appropriate environmental cues for forming new connections still exist in axotomized adult rats with chronic motoneuron degeneration. This advances recent work showing that mouse ES cell-derived motoneurons can extend axons into ventral roots only after dibutyryl cAMP treatment (Harper et al., 2004). However, even with this treatment, those axons were confined to proximal ventral roots, and no neuromuscular junctions were detected. These different findings may be due to different injury models, acute vs. chronic injury, differences in cell type (mouse ES cell vs. human NSC), or different treat-
260
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 3. Human NSC-derived motoneurons innervate muscle in adult motoneuron deficient rats. (a– d) hNSCs triple labeled with GFP (a), ChAT (b) and CTB retrograde tracer (c) and a merged picture for all three labels (d). Note the cells labeled for GFP and ChAT (red arrows), for ChAT and CTB (yellow arrowheads) and for all three labels (yellow arrows). (e– g) hNSCs double labeled with GFP and a motoneuron marker Hb9 (yellow arrows). (h–j) Axon terminals labeled with a human specific neurofilament (hNF) antibody (h) and ␣-bungratoxin (␣-BT)-labeled neuromuscular junctions (i) with an enlarged merged picture showing hNSC-derived nerve terminals attaching to junctions (j). Scale bars⫽100 m.
ment of stem cells prior to grafting (retinoic acid/sonic hedgehog vs. primed). Rats grafted with unprimed hNSCs also show an improving behavioral trend but considerably less than after primed-cell transplants. These rats have very few cholinergic neurons differentiated from grafted hNSCs, with only one grafted axon detectable in sciatic nerves and no human derived axons attaching to neuromuscular junctions. Thus, the improvement is most likely attributable to trophic factors released from grafted hNSCs as shown by others (Kerr et al., 2003). In contrast, animals grafted with primed hNSCs show statistically significant improvement of motor function and also have grafted neurons send axons to host muscle, supporting our hypothesis that hNSC-derived neu-
ron-muscle attachments contribute partially to functional improvement. Thus, our findings suggest that primed hNSCs adapt to an adult environment by cholinergic differentiation and at least some integration into peripheral host circuitry since they attach to an appropriate muscle and seem to form neuromuscular junctions. It will be necessary to confirm that the stem cell-derived neuromuscular junctions are functionally normal, but the fact that grafted neurons innervate muscles together with behavioral improvement allows optimism. Next, even though some transplanted cells contact muscle, there is no proof that they can integrate into the existing circuitry in the ventral horn. However, many transplanted cells develop dendrites and resemble normal
Table 1. Anatomical analyses per individual animal grafted with primed hNSCs)a
Rat number
G
G⫹Ch
(G⫹Ch)/G
G⫹CTB
(G⫹CTB)/G
G⫹Ch⫹CTB
(G⫹Ch⫹CTB)/G
Number of GFP neurites per 40 m section in sciatic nerve
1 2 3 4 5
657 359 509 614 356
351 280 296 150 204
54% 80% 59% 24% 59%
193 95 80 50 54
20% 27% 16% 8% 16%
158 66 54 33 46
24% 19% 11% 5% 13%
14 13 14 7 5
a Single and multiple labeling of cells and axons presented in relation to individual animals. G-GFP, Ch-cholinergic, CTB-cholera toxin B. ⬙⫹⬙, double or triple labeled; ⬙G⬙ numbers of double or triple labeled cells divided by total numbers of grafted GFP cells in each animal.
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 4. Gait analysis. (a) Footprints from a representative rat before and 3 months after primed hNSC-graft. L, left injured; R, right normal. (b) SFI, or right versus left comparison of gaits, changes before and 1 and 3 months after transplants. Note the significant improvement only in the 3 month primed cell grafted animals. Symbols P, UP and V⫽grafts of primed hNSCs, unprimed and vehicle, respectively. * P⬍0.001.
Table 2. SFI per individual animal grafted with primed hNSCs Rat no.
Before graft
3 Months after graft
1 2 3 4 5 6 7 8 9 10
⫺40% ⫺50% ⫺46% ⫺42% ⫺61% ⫺41% ⫺45% ⫺57% ⫺51% ⫺51%
2% ⫺14% ⫺49% ⫺40% ⫺56% ⫺32% ⫺35% ⫺31% ⫺32% ⫺13%
motoneurons, which combined with the functional improvement, presumably gives indication of at least some integration of these new motoneurons with previous circuitry. These findings, in our opinion, represent a conceptual advance, and may shed light on the development of stem cell replacement techniques for neural disorders characterized by motoneuron loss. Acknowledgments—We thank Dr. L. A. Vergara, S. G. Scarbrough, and T. Dunn for technical assistance; P. Gazzoli for manuscript preparation; Dr. E. F. Epstein for critical review; and support from the TIRR Foundation and NS046025 and NS10161 from NIH.
REFERENCES Barberi T, Klivenyi P, Calingasan NY, Lee H, Kawamata H, Loonam K, Perrier AL, Bruses J, Rubio ME, Topf N, Tabar V, Harrison NL, Beal MF, Moore MAS, Studer L (2003) Neural subtype specification of fertilization and nuclear transfer embryonic stem cells and application in parkinsonian mice. Nat Biotech 21:1200 –1207. Cao QL, Zhang YP, Howard RM, Walters WM, Tsoulfas P, Whittemore SR (2001) Pluripotent stem cells engrafted into the normal or lesioned adult rat spinal cord are restricted to a glial lineage. Exp Neurol 167:48 –58.
261
Carpenter MK, Cui X, Hu ZY, Jackson J, Sherman S, Seiger A, Wahlberg LU (1999) In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 158:265–278. Chen JW, Murphy TL, Willingham MC, Pastan I, August JT (1985) Identification of 2 lysosomal membrane-glycoproteins. J Cell Biol 101:85–95. Coggeshall RE (1992) A consideration of neural counting methods. Trends Neurosci 15:9 –13. de Medinaceli L, Freed WJ, Wyatt RJ (1982) An index of the functionalcondition of rat sciatic-nerve based on measurements made from walking tracks. Exp Neurol 77:634–643. Fricker RA, Carpenter MK, Winkler C, Greco C, Gates MA, Bjorklund A (1999) Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 19:5990 – 6005. Harper JM, Krishnan C, Darman JS, Deshpande DM, Peck S, Shats I, Backovic S, Rothstein JD, Kerr DA (2004) Axonal growth of embryonic stem cell-derived motoneurons in vitro and in motoneuroninjured adult rats. Proc Natl Acad Sci USA 101:7123–7128. Kerr DA, Llado J, Shamblott MJ, Maragakis NJ, Irani DN, Crawford TO, Krishnan C, Dike S, Gearhart JD, Rothstein JD (2003) Human embryonic germ cell derivatives facilitate motor recovery of rats with diffuse motor neuron injury. J Neurosci 23:5131–5140. Renoncourt Y, Carroll P, Filippi P, Arce V, Alonso S (1998) Neurons derived in vitro from ES cells express homeoproteins characteristic of motoneurons and interneurons. Mech Dev 79:185–197. Schmalbruch H (1984) Motoneuron death after sciatic-nerve section in newborn rats. J Comp Neurol 224:252–258. Sheen VL, Arnold MW, Wang Y, Macklis JD (1999) Neural precursor differentiation following transplantation into neocortex is dependent on intrinsic developmental state and receptor competence. Exp Neurol 158:47– 62. Shihabuddin LS, Horner PJ, Ray J, Gage FH (2000) Adult spinal cord stem cells generate neurons after transplantation in the adult dentate gyrus. J Neurosci 20:8727– 8735. Snider WD, Elliott JL, Yan Q (1992) Axotomy-induced neuronal death during development. J Neurobiol 23:1231–1246. Svendsen CN, Caldwell MA, Shen J, ter Borg MG, Rosser AE, Tyers P, Karmiol S, Dunnett SB (1997) Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson’s disease. Exp Neurol 148:135–146. Svendsen CN, ter Borg MG, Armstrong RJ, Rosser AE, Chandran S, Ostenfeld T, Caldwell MA (1998) A new method for the rapid and long term growth of human neural precursor cells. J Neurosci Methods 85:141–152. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS, Jones JM (1998) Embryonic stem cell lines derived from human blastocysts. Science 282:1145–1147. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, Phan TV, Tsukamoto AS, Gage FH, Weissman IL (2000) Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci USA 97:14720 –14725. Vescovi AL, Parati EA, Gritti A, Poulin P, Ferrario M, Wanke E, Frolichsthal-Schoeller P, Cova L, Arcellana-Panlilio M, Colombo A, Galli R (1999) Isolation and cloning of multipotential stem cells from the embryonic human CNS and establishment of transplantable human neural stem cell lines by epigenetic stimulation. Exp Neurol 156:71– 83. Villa A, Snyder EY, Vescovi A, Martinez-Serrano A (2000) Establishment and properties of a growth factor-dependent, perpetual neural stem cell line from the human CNS. Exp Neurol 161:67– 84. Vroemen M, Aigner L, Winkler J, Weidner N (2003) Adult neural progenitor cell grafts survive after acute spinal cord injury and integrate along axonal pathways. Eur J Neurosci 18: 743–751.
262
J. Gao et al. / Neuroscience 131 (2005) 257–262
Wichterle H, Lieberam I, Porter JA, Jessell TM (2002) Directed differentiation of embryonic stem cells into motor neurons. Cell 110:385–397. Wu P, Tarasenko YI, Gu YP, Huang LYM, Coggeshall RE, Yu YJ (2002a) Region-specific generation of cholinergic neurons from
fetal human neural stem cells grafted in adult rat. Nat Neurosci 5:1271–1278. Wu P, Ye Y, Svendsen CN (2002b) Transduction of human neural progenitor cells using recombinant adeno-associated viral vectors. Gene Ther 9:245–255.
(Accepted 23 October 2004)
RAPID REPORT HUMAN NEURAL STEM CELL-DERIVED CHOLINERGIC NEURONS INNERVATE MUSCLE IN MOTONEURON DEFICIENT ADULT RATS J. GAO, R. E. COGGESHALL, Y. I. TARASENKO AND P. WU*
do not form typical motoneurons in adult spinal cord presumably because of lack of sufficient environmental cues (Svendsen et al., 1997; Sheen et al., 1999; Shihabuddin et al., 2000; Cao et al., 2001; Fricker et al., 1999; Vroemen et al., 2003). For example, most human embryonic germ cells, when grafted into rats with diffuse motor neuron injury, remain undifferentiated and only a few become cholinergic, and these do not connect to muscle (Kerr et al., 2003). To obtain a higher proportion of stem cellderived cholinergic neurons, we developed an in vitro priming technology that resulted in human fetal NSCs successfully differentiating into cholinergic neurons in vitro, and many acquire a cholinergic phenotype when grafted into intact adult spinal cord (Wu et al., 2002a). In this study, we show that these primed NSCs, when grafted into adult rat spinal cord with motoneuron degeneration, develop into cholinergic neurons that innervate peripheral muscle with concomitant improvement of motor function.
Department of Neuroscience and Cell Biology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-1043, USA
Abstract—Motoneuron damage occurs in spinal cord injury and amyotrophic lateral sclerosis. Current advances offer hope that human embryonic stem cells [Science 282 (1998) 1145] or neural stem cells (NSC) [Exp Neurol 161 (2000) 67; Exp Neurol 158 (1999) 265; J Neurosci Methods 85 (1998) 141; Proc Natl Acad Sci USA 97 (2000) 14720; Exp Neurol 156 (1999) 156] may be donors to replace lost motoneurons. Previously, we developed a priming procedure that produced cholinergic cells that resemble motoneurons from human NSCs grafted into adult rat spinal cord [Nat Neurosci 5 (2002a) 1271]. However, effective replacement therapy will ultimately rely on successful connection of new motoneurons with their muscle targets. In this study, we examined the potential of human fetal NSC transplantation to replace lost motoneurons in an animal model of chronic motoneuron deficiency (newborn sciatic axotomy) [J Comp Neurol 224 (1984) 252; J Neurobiol 23 (1992) 1231]. We found, for the first time, that human neural stem cell-derived motoneurons send axons that pass through ventral root and sciatic nerve to form neuromuscular junctions with their peripheral muscle targets. Furthermore, this new cholinergic innervation correlates with partial improvement of motor function. © 2005 IBRO. Published by Elsevier Ltd. All rights reserved.
EXPERIMENTAL PROCEDURES Human NSCs and AAVegfp vectors The K048 line of hNSCs (Svendsen et al., 1998) were propagated as neurospheres in growth medium supplemented with epidermal growth factor, basic fibroblast growth factor (bFGF) and leukemia inhibitory factor (Wu et al., 2002a). For priming, 2-day neurospheres (passages 25–31) were treated with 20 ng/ml bFGF, 5 g/ml heparin, 1 g/ml laminin and 50 ng/ml mouse sonic hedgehog amino-terminal peptide (R&D Systems, Minneapolis, MN, USA) for 5 days and then cultured with 1⫻ B27 (Invitrogen, Carlsbad, CA, USA) diluted in DMEM/F12 (Invitrogen) for 2 days in vitro. Four days before transplantation, primed cells were labeled by transduction with a recombinant adeno-associated viral AAVegfp (Wu et al., 2002b). For unprimed cell transplants, proliferating neurospheres were treated with AAVegfp directly and maintained in medium containing growth factors for 4 days.
Key words: neuromuscular junction, axotomy, motoneuron disease, cell therapy.
To replace lost motoneurons, important issues are 1) to direct stem cells to differentiate into cholinergic motoneurons in the ventral horn of the spinal cord, 2) to have enough cues to allow new motoneurons to send axons through appropriate nerves to form synapses on muscle cells, and 3) to show concomitant improved motor function. Several groups generated cholinergic neurons from mouse embryonic stem (ES) cells in vitro by treating them with inductive factors such as retinoic acid and sonic hedgehog (Renoncourt et al., 1998; Wichterle et al., 2002; Barberi et al., 2003). However, these ES or neural stem cells (NSCs)
Neonatal sciatic axotomy and transplantation Left sciatic nerves of newborn Sprague–Dawley rats were crushed mid- thigh for 30 s using smooth-tipped forceps. Two months later, axotomized rats were divided into three groups to receive 1) primed cell grafts, 2) unprimed cell grafts or 3) vehicle. The spinal cord transplantation protocol was established according to the NIH guidelines for the care and use of laboratory animals, and approved by The University of Texas Medical Branch IACUC. Approximately 105 primed or unprimed hNSCs/2 l or 2 l vehicle were implanted into L4 motoneuron deficient spinal ventral horns (ML: ⫹0.7 mm; DV: ⫺1.5 mm from dura). All animals were treated with the immunosuppressor NEORAL cyclosporine (Novartis Pharmaceuticals Co., East Hanover, NJ, USA), 100 g/ml in drinking water 3 days before surgery and thereafter.
*Corresponding author. Tel: ⫹1-409-772-2748; fax: ⫹1-409-7721861. E-mail address: [email protected] (P. Wu). Abbreviations: bFGF, basic fibroblast growth factor; ChAT, choline acetyltransferase; CTB, cholera toxin subunit B; ES, embryonic stem; LAMP2, lysosomal-associated membrane protein 2; NSC, neural stem cells; SFI, sciatic function index; WGA, wheat germ agglutinin.
0306-4522/05$30.00⫹0.00 © 2005 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.10.033
257
258
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 1. Confocal images of primed human NSCs grafted in adult rats with motoneuron deficiency. (a) A view of the ventral horn with many GFP-labeled cells that resemble motoneurons; inset, a low power view showing the whole hemi-cord at the graft site. Grafted hNSCs send numerous large GFP fibers into the ventral root (VR) (b) and sciatic nerve at mid-thigh (c). * Ventral white matter of the spinal cord. Scale bars⫽200 m.
Gait analysis Motor function of rats with neonatal sciatic axotomy was evaluated before transplantation and then 1 or 3 months after lesion by a sciatic function index (SFI; de Medinaceli et al., 1982). SFI of ⫺100% represents a complete lost of gait, SFI of 0⫾11% is a normal gait.
Retrograde tracing Animals received injections of 7.5 l of Alexa 594-wheat germ agglutinin (WGA; 10 g/l; Molecular Probes, Eugene, OR, USA) into the sciatic nerve, or 10 l of Alexa 594-cholera toxin subunit B (CTB; 1 g/l; Molecular Probes) into the left medial and lateral gastrocnemius muscles. Three days post-WGA injection or 10 days after CTB administration, animals were perfused with 4% paraformaldehyde.
Immunohistochemistry Five animals in each group were perfused with 4% paraformaldehyde 3 months after grafting. L4 –L6 segments of spinal cords and ventral roots plus the sciatic nerves were cryo-sectioned at 45 m. Primary antibodies included choline acetyltransferase (ChAT; Chemicon) at 1:100, Hb9 (Chemicon) at 1:100, human specific lysosomal-associated membrane protein 2 (LAMP2) at 1:1600 (Developmental Studies Hybridoma Bank, Iowa City, IA, USA; Chen et al., 1985), human specific neurofilament 70 (Chemicon) at 1:200 and rat specific neurofilament 200 (Chemicon) at 1:200.
Neuromuscular junction labeling Gastrocnemius muscles from two animals in each group were sectioned at approximately 500 m with a razor blade. The muscle slices were incubated with 1 g/ml of rhodamine-conjugated
␣-bungarotoxin (Molecular Probes) for 1 h at room temperature. Some of the slices were double-labeled with either GFP or human specific neurofilament 70.
Quantitative analyses Sections were imaged confocally. Numbers of GFP-labeled hNSCs in the ventral horn were determined stereologically using a fractionator analysis (Coggeshall, 1992). Percentages of double- or triple-labeled cells or neuromuscular junctions with and without an attached GFP or human neurofilament axon were also determined. Motor function (SFI) was analyzed using repeated measures ANOVA (GraphPad Software, Inc., San Diego, CA, USA).
RESULTS Grafted human NSCs survived and differentiated in motoneuron deficient spinal cords Neonatal sciatic axotomy resulted in an approximately 40% loss of myelinated axons in the L4 ventral root, which is equated to a 40% motoneuron loss, and an average SFI of ⫺50%. Axotomized animals were screened by gait analysis just before implantation, and those with SFI values of ⫺40 to ⫺70% were randomly divided into three groups to receive primed or unprimed hNSC grafts or vehicle. Three months after grafting with primed cells, approximately 500 GFP-labeled cells (514⫾64) appeared in each transplanted ventral horn (Fig. 1a). Many of these cells had radiating dendrites and sent axons into ventral roots (Fig. 1b) and sciatic nerves (Fig. 1c). In contrast, most trans-
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 2. GFP-labeled cells are of human NSC origin. (a– c) The ventral horn showing large cells (arrows) that co-localize GFP and a human specific antibody, LAMP2; d–f, similar cells labeled by Alexa 594WGA, a retrograde tracer injected into sciatic nerves 17 days post transplantation (before grafted cells project into the nerve). Note that GFP and retrogradely labeled cells are in close proximity with no co-localization of label. Scale bar⫽200 m. (g–i) Longitudinal sections of sciatic nerves showing GFP axons (green, g), endogenous rat specific neurofilament stained axons (red, h) and a combined image showing that the two labels in separate axon populations (i) as evidence that neither cell fusion or leaking of label explain our results. Scale bar⫽100 m.
planted unprimed hNSCs remained small and round, with occasional neuron-like GFP cells and fibers in the ventral white matter. Only one GFP fiber was detected in the sciatic nerves from all five unprimed hNSC animals and none in vehicle animals (data not shown). To verify the GFP cells were of human (transplanted) origin, immunostaining with a human specific antibody, LAMP2, showed that all green GFP-labeled cells colocalized with LAMP2 (Fig. 2a– c). We injected the retrograde tracer WGA into the sciatic nerve on the transplanted side at 17 days post-grafting and animals were killed 3 days later, 20 days after transplant and before regenerating axons could get into the sciatic nerve. Thus these WGA⫹ cells are endogenous, and the green GFP (exogenous) and red WGA (endogenous) cells are in close proximity without co-localization of labels (Fig. 2d–f). Further analysis on GFP axons in sciatic nerves, as shown in Fig. 2g–i, revealed no sign of fusion or overlapping of GFP markers with host axons stained by rat specific neurofilament antibodies (red). Human neural stem cell (NSC)-derived motoneuron innervated target muscle Many grafted hNSCs acquired motoneuron morphology (Fig. 2). At 3 months post-transplant, 51% of the GFP cells were cholinergic (Fig. 3a, b) and Hb9 positive (Fig. 3e– g),
259
19% were labeled retrogradely from muscle (Fig. 3a, c) and 14% were both cholinergic and retrogradely labeled (Fig. 3a– d). This indicates that some grafted hNSCs become cholinergic motoneurons that send axons to peripheral muscle targets. Numbers of grafted primed cells with various labeling as well as numbers of peripheral axons from grafted neurons are shown in Table 1. In the unprimed graft group, cell survival rates were similar to those of their primed counterparts, but only a few differentiated into cholinergic neurons, and none were retrogradely labeled (data not shown). To determine whether the terminals of hNSC-derived motoneurons attached to muscle, gastrocnemius slices were stained with rhodamine-conjugated ␣-bungarotoxin that specifically binds to the nicotinic acetylcholine receptor to identify neuromuscular junctions. Human NSCderived motoneuron axons were visualized either by GFP or human specific neurofilament 70 antibody, and 6.4% of the ␣-bungarotoxin⫹ neuromuscular junctions were contacted by GFP- (data not shown) or human specific neurofilament 70-labeled axons (Fig. 3h–j). No neuromuscular junctions were co-labeled with human specific neurofilament antibody in unprimed NSC-grafted or vehicle control rats. Human NSC grafts improved motor function in motoneuron deficient rats Neonatal sciatic nerve crushed rats (n⫽10 per group) showed a marked gait deficiency for the left hind limb (Fig. 4a) with SFI scores of approximately ⫺50% before implantation (Fig. 4b and Table 2). One month after transplantation, a slight but not significant improvement was observed in all groups. At 3 months post-grafting, only rats that received primed hNSC grafts showed statistically significant improvement (Fig. 4b). Such behavioral improvement (Table 2) correlates to some extent with the total numbers of grafted cholinergic neurons that are retrogradely labeled and to a lesser degree with the numbers of GFP axons in peripheral nerves, but not with the total numbers of survived cells (Table 1), although the number of animals sampled is too low for a statistical correlation.
DISCUSSION The present study shows for the first time that motoneurons generated from primed human fetal NSC emit axons that innervate peripheral muscle targets. These findings correlate with partial improvement of motor function and indicate that appropriate environmental cues for forming new connections still exist in axotomized adult rats with chronic motoneuron degeneration. This advances recent work showing that mouse ES cell-derived motoneurons can extend axons into ventral roots only after dibutyryl cAMP treatment (Harper et al., 2004). However, even with this treatment, those axons were confined to proximal ventral roots, and no neuromuscular junctions were detected. These different findings may be due to different injury models, acute vs. chronic injury, differences in cell type (mouse ES cell vs. human NSC), or different treat-
260
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 3. Human NSC-derived motoneurons innervate muscle in adult motoneuron deficient rats. (a– d) hNSCs triple labeled with GFP (a), ChAT (b) and CTB retrograde tracer (c) and a merged picture for all three labels (d). Note the cells labeled for GFP and ChAT (red arrows), for ChAT and CTB (yellow arrowheads) and for all three labels (yellow arrows). (e– g) hNSCs double labeled with GFP and a motoneuron marker Hb9 (yellow arrows). (h–j) Axon terminals labeled with a human specific neurofilament (hNF) antibody (h) and ␣-bungratoxin (␣-BT)-labeled neuromuscular junctions (i) with an enlarged merged picture showing hNSC-derived nerve terminals attaching to junctions (j). Scale bars⫽100 m.
ment of stem cells prior to grafting (retinoic acid/sonic hedgehog vs. primed). Rats grafted with unprimed hNSCs also show an improving behavioral trend but considerably less than after primed-cell transplants. These rats have very few cholinergic neurons differentiated from grafted hNSCs, with only one grafted axon detectable in sciatic nerves and no human derived axons attaching to neuromuscular junctions. Thus, the improvement is most likely attributable to trophic factors released from grafted hNSCs as shown by others (Kerr et al., 2003). In contrast, animals grafted with primed hNSCs show statistically significant improvement of motor function and also have grafted neurons send axons to host muscle, supporting our hypothesis that hNSC-derived neu-
ron-muscle attachments contribute partially to functional improvement. Thus, our findings suggest that primed hNSCs adapt to an adult environment by cholinergic differentiation and at least some integration into peripheral host circuitry since they attach to an appropriate muscle and seem to form neuromuscular junctions. It will be necessary to confirm that the stem cell-derived neuromuscular junctions are functionally normal, but the fact that grafted neurons innervate muscles together with behavioral improvement allows optimism. Next, even though some transplanted cells contact muscle, there is no proof that they can integrate into the existing circuitry in the ventral horn. However, many transplanted cells develop dendrites and resemble normal
Table 1. Anatomical analyses per individual animal grafted with primed hNSCs)a
Rat number
G
G⫹Ch
(G⫹Ch)/G
G⫹CTB
(G⫹CTB)/G
G⫹Ch⫹CTB
(G⫹Ch⫹CTB)/G
Number of GFP neurites per 40 m section in sciatic nerve
1 2 3 4 5
657 359 509 614 356
351 280 296 150 204
54% 80% 59% 24% 59%
193 95 80 50 54
20% 27% 16% 8% 16%
158 66 54 33 46
24% 19% 11% 5% 13%
14 13 14 7 5
a Single and multiple labeling of cells and axons presented in relation to individual animals. G-GFP, Ch-cholinergic, CTB-cholera toxin B. ⬙⫹⬙, double or triple labeled; ⬙G⬙ numbers of double or triple labeled cells divided by total numbers of grafted GFP cells in each animal.
J. Gao et al. / Neuroscience 131 (2005) 257–262
Fig. 4. Gait analysis. (a) Footprints from a representative rat before and 3 months after primed hNSC-graft. L, left injured; R, right normal. (b) SFI, or right versus left comparison of gaits, changes before and 1 and 3 months after transplants. Note the significant improvement only in the 3 month primed cell grafted animals. Symbols P, UP and V⫽grafts of primed hNSCs, unprimed and vehicle, respectively. * P⬍0.001.
Table 2. SFI per individual animal grafted with primed hNSCs Rat no.
Before graft
3 Months after graft
1 2 3 4 5 6 7 8 9 10
⫺40% ⫺50% ⫺46% ⫺42% ⫺61% ⫺41% ⫺45% ⫺57% ⫺51% ⫺51%
2% ⫺14% ⫺49% ⫺40% ⫺56% ⫺32% ⫺35% ⫺31% ⫺32% ⫺13%
motoneurons, which combined with the functional improvement, presumably gives indication of at least some integration of these new motoneurons with previous circuitry. These findings, in our opinion, represent a conceptual advance, and may shed light on the development of stem cell replacement techniques for neural disorders characterized by motoneuron loss. Acknowledgments—We thank Dr. L. A. Vergara, S. G. Scarbrough, and T. Dunn for technical assistance; P. Gazzoli for manuscript preparation; Dr. E. F. Epstein for critical review; and support from the TIRR Foundation and NS046025 and NS10161 from NIH.
REFERENCES Barberi T, Klivenyi P, Calingasan NY, Lee H, Kawamata H, Loonam K, Perrier AL, Bruses J, Rubio ME, Topf N, Tabar V, Harrison NL, Beal MF, Moore MAS, Studer L (2003) Neural subtype specification of fertilization and nuclear transfer embryonic stem cells and application in parkinsonian mice. Nat Biotech 21:1200 –1207. Cao QL, Zhang YP, Howard RM, Walters WM, Tsoulfas P, Whittemore SR (2001) Pluripotent stem cells engrafted into the normal or lesioned adult rat spinal cord are restricted to a glial lineage. Exp Neurol 167:48 –58.
261
Carpenter MK, Cui X, Hu ZY, Jackson J, Sherman S, Seiger A, Wahlberg LU (1999) In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 158:265–278. Chen JW, Murphy TL, Willingham MC, Pastan I, August JT (1985) Identification of 2 lysosomal membrane-glycoproteins. J Cell Biol 101:85–95. Coggeshall RE (1992) A consideration of neural counting methods. Trends Neurosci 15:9 –13. de Medinaceli L, Freed WJ, Wyatt RJ (1982) An index of the functionalcondition of rat sciatic-nerve based on measurements made from walking tracks. Exp Neurol 77:634–643. Fricker RA, Carpenter MK, Winkler C, Greco C, Gates MA, Bjorklund A (1999) Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 19:5990 – 6005. Harper JM, Krishnan C, Darman JS, Deshpande DM, Peck S, Shats I, Backovic S, Rothstein JD, Kerr DA (2004) Axonal growth of embryonic stem cell-derived motoneurons in vitro and in motoneuroninjured adult rats. Proc Natl Acad Sci USA 101:7123–7128. Kerr DA, Llado J, Shamblott MJ, Maragakis NJ, Irani DN, Crawford TO, Krishnan C, Dike S, Gearhart JD, Rothstein JD (2003) Human embryonic germ cell derivatives facilitate motor recovery of rats with diffuse motor neuron injury. J Neurosci 23:5131–5140. Renoncourt Y, Carroll P, Filippi P, Arce V, Alonso S (1998) Neurons derived in vitro from ES cells express homeoproteins characteristic of motoneurons and interneurons. Mech Dev 79:185–197. Schmalbruch H (1984) Motoneuron death after sciatic-nerve section in newborn rats. J Comp Neurol 224:252–258. Sheen VL, Arnold MW, Wang Y, Macklis JD (1999) Neural precursor differentiation following transplantation into neocortex is dependent on intrinsic developmental state and receptor competence. Exp Neurol 158:47– 62. Shihabuddin LS, Horner PJ, Ray J, Gage FH (2000) Adult spinal cord stem cells generate neurons after transplantation in the adult dentate gyrus. J Neurosci 20:8727– 8735. Snider WD, Elliott JL, Yan Q (1992) Axotomy-induced neuronal death during development. J Neurobiol 23:1231–1246. Svendsen CN, Caldwell MA, Shen J, ter Borg MG, Rosser AE, Tyers P, Karmiol S, Dunnett SB (1997) Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson’s disease. Exp Neurol 148:135–146. Svendsen CN, ter Borg MG, Armstrong RJ, Rosser AE, Chandran S, Ostenfeld T, Caldwell MA (1998) A new method for the rapid and long term growth of human neural precursor cells. J Neurosci Methods 85:141–152. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS, Jones JM (1998) Embryonic stem cell lines derived from human blastocysts. Science 282:1145–1147. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, Phan TV, Tsukamoto AS, Gage FH, Weissman IL (2000) Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci USA 97:14720 –14725. Vescovi AL, Parati EA, Gritti A, Poulin P, Ferrario M, Wanke E, Frolichsthal-Schoeller P, Cova L, Arcellana-Panlilio M, Colombo A, Galli R (1999) Isolation and cloning of multipotential stem cells from the embryonic human CNS and establishment of transplantable human neural stem cell lines by epigenetic stimulation. Exp Neurol 156:71– 83. Villa A, Snyder EY, Vescovi A, Martinez-Serrano A (2000) Establishment and properties of a growth factor-dependent, perpetual neural stem cell line from the human CNS. Exp Neurol 161:67– 84. Vroemen M, Aigner L, Winkler J, Weidner N (2003) Adult neural progenitor cell grafts survive after acute spinal cord injury and integrate along axonal pathways. Eur J Neurosci 18: 743–751.
262
J. Gao et al. / Neuroscience 131 (2005) 257–262
Wichterle H, Lieberam I, Porter JA, Jessell TM (2002) Directed differentiation of embryonic stem cells into motor neurons. Cell 110:385–397. Wu P, Tarasenko YI, Gu YP, Huang LYM, Coggeshall RE, Yu YJ (2002a) Region-specific generation of cholinergic neurons from
fetal human neural stem cells grafted in adult rat. Nat Neurosci 5:1271–1278. Wu P, Ye Y, Svendsen CN (2002b) Transduction of human neural progenitor cells using recombinant adeno-associated viral vectors. Gene Ther 9:245–255.
(Accepted 23 October 2004)