CXCR4 promotes differentiation of oligodendrocyte progenitors and remyelination Jigisha R. Patela, Erin E. McCandlessa,b, Denise Dorseya, and Robyn S. Kleina,b,c,1 Departments of aInternal Medicine, bPathology and Immunology, and cAnatomy and Neurobiology, Washington University School of Medicine, St. Louis, MO 63110 Communicated by Anthony A. James, University of California, Irvine, CA, May 7, 2010 (received for review December 21, 2009)
Multiple sclerosis is a neurodegenerative disease characterized by episodes of autoimmune attack of oligodendrocytes leading to demyelination and progressive functional deficits. Because many patients exhibit functional recovery in between demyelinating episodes, understanding mechanisms responsible for repair of damaged myelin is critical for developing therapies that promote remyelination and prevent disease progression. The chemokine CXCL12 is a developmental molecule known to orchestrate the migration, proliferation, and differentiation of neuronal precursor cells within the developing CNS. Although studies suggest a role for CXCL12 in oligodendroglia ontogeny in vitro, no studies have investigated the role of CXCL12 in remyelination in vivo in the adult CNS. Using an experimental murine model of demyelination mediated by the copper chelator cuprizone, we evaluated the expression of CXCL12 and its receptor, CXCR4, within the demyelinating and remyelinating corpus callosum (CC). CXCL12 was significantly up-regulated within activated astrocytes and microglia in the CC during demyelination, as were numbers of CXCR4+ NG2+ oligodendrocyte precursor cells (OPCs). Loss of CXCR4 signaling via either pharmacological blockade or in vivo RNA silencing led to decreased OPCs maturation and failure to remyelinate. These data indicate that CXCR4 activation, by promoting the differentiation of OPCs into oligodendrocytes, is critical for remyelination of the injured adult CNS. chemokine
| cuprizone | CXCL12 | central nervous system
ultiple sclerosis (MS), a progressive, neurodegenerative M disease of the CNS, occurs most often in a relapsing/remitting form, in which a period of demyelination is followed by a period of functional recovery (1). The recovery stage involves remyelination via the migration and maturation of oligodendrocyte precursor cells (OPCs) (2). However, as the disease progresses, remyelination fails with continuous loss of function (3). Possible explanations for remyelination failure of intact axons include defects in OPC recruitment to the site of demyelination or in OPC differentiation into myelinating oligodendrocytes. Although studies indicate that both aspects of OPC biology are altered in MS (4, 5), the molecular mechanisms that orchestrate these processes within the adult CNS are incompletely understood. Studies in mice indicate that neural precursors that give rise to cells of oligodendrocytes lineage can be identified within the ventral half of the ventricular zones of all CNS regions by embryonic days 12–14 (E12–E14) via their expression of NG2 chondroitin sulfate proteoglycan (6). In the final stage of oligodendrocyte differentiation, which occurs primarily during the postnatal period (P4–P12), OPCs begin to express mature markers of oligodendrocytes including 2′3′-cyclic nucleotide phosphohydrolase (CNPase), myelin basic protein (MBP), proteolipid protein (PLP) and myelin oligodendrocyte glycoprotein (MOG). Similar events occur during remyelination; NG2+ OPCs proliferate within subventricular zones, migrate to areas of demyelination, and differentiate into premyelinating oligodendrocytes that become mature oligodendrocytes as they begin wrapping demyelinated axons and forming new myelin, which is usually thinner and shorter than the original myelin (7). Thus, molecular events that direct myelination during development may be recapitulated in the adult CNS during white matter injury. 11062–11067 | PNAS | June 15, 2010 | vol. 107 | no. 24
The chemokine CXCL12 and its receptor CXCR4 have well known roles in the patterning and function of the immune and nervous systems where they localize various cell types to specific microenvironments (8). CXCL12 was originally identified as a bone marrow stromal cell-derived chemoattractant and proliferative factor for B-cell precursors and binds two receptors, CXCR4, which couples to the Gi family of signal transducers (9), and CXCR7, whose signaling capacity is unclear and may function predominantly to regulate CXCR4 activation via sequestration of CXCL12 (10). Transcripts for CXCL12, CXCR4, and CXCR7 are detected in the CNS of rodent embryos (11, 12), where these molecules function to promote proliferation and migration of neural precursors (13, 14). In the mature CNS, both CXCR4 and CXCR7 are expressed by subpopulations of neurons, whereas CXCR7 is additionally expressed by the endothelium and CXCR4 by neural stem cells (NSCs) and mediates their homing to damaged areas within the CNS (12, 15, 16). Although several in vitro studies suggest possible roles for CXCL12 and CXCR4 in the migration and maturation of OPCs (17–19), these questions have not been studied in vivo. To study the role of CXCL12 in regulating OPC recruitment and remyelination within the adult CNS, we used a well-established model of demyelination induced via exposure to the copper chelator, cuprizone (CPZ), in the diet. Ingestion of CPZ for 6 wk leads to apoptosis of mature oligodendrocytes within the corpus callosum (CC) with simultaneous accumulation of OPCs at this site (20). Cessation of CPZ after 6 wk is followed by OPC differentiation and remyelination, whereas continued ingestion leads to OPC apoptosis and chronic demyelination. Prior studies have demonstrated that demyelination and glial injury occurs most prominently in the caudal region of the CC and that continued ingestion of CPZ leads to extension of injury to OPCs and decreased remyelination even after cessation of toxin (21, 22). Thus, this model provides an opportunity to examine signals associated with both successful and failed remyelination. Herein, we show that CPZ-induced demyelination is associated with CXCL12 expression by activated astrocytes and microglia and the accumulation of CXCR4+NG2+ OPCs within the CC with the greatest numbers in the caudal region. Administration of a specific CXCR4 antagonist or lentiviral delivery of CXCR4 shRNA led to increased numbers of proliferating OPCs within subventricular areas and of NG2+ OPCs within the CC with decreased differentiation and remyelination at this site, highlighting the critical role of these molecules in recovery from injury within this brain region. Results CXCL12 Is Expressed by Activated Astrocytes and Microglia Within the Corpus Callosi of CPZ-Fed Mice. To define the specific roles of
CXCL12 and CXCR4 within the CNS during remyelination, we used the CPZ model of demyelinating disease (20). Studies were
Author contributions: J.R.P. and R.S.K. designed research; J.R.P. and D.D. performed research; E.E.M. contributed new reagents/analytic tools; J.R.P., D.D., and R.S.K. analyzed data; and J.R.P. and R.S.K. wrote the paper. The authors declare no conflict of interest. Freely available online through the PNAS open access option. 1
To whom correspondence should be addressed. E-mail:
[email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1006301107/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1006301107
Fig. 1. CXCL12 expression in the CC after CPZ exposure. (A) qRT-PCR analysis of CXCL12 expression in dissected whole CC derived from naive mice and from mice after 6 and 12 wk of CPZ ingestion. Each group represents three to five individual animals, *, P < 0.05. (B) Confocal IHC analysis of CC from naive mice and after 6 and 12 wk of CPZ ingestion. Images were stained for GFAP or CD11b (green, Upper and Lower, respectively), CXCL12 (red), and nuclei (blue). Higher power images of boxed areas in images from 6-wk time-point are shown. Representative images shown for three sections from three mice in two separate experiments. IC = isotype control. (Scale bars, 25 μm.) (C and D) Quantitation of GFAP+ (C) and CD11b+ (D) cells that express CXCL12 within rostral, central, and caudal sections of the CC in naive mice (white bars) and after 6 (gray bars) and 12 (black bars) wk of CPZ exposure; n = 6 images taken from three to five mice/group. *, P < 0.05 and **, P < 0.005.
levels of CXCL12 expression compared with other CC regions at both 6 and 12 wk postexposure (Fig. 1C). In contrast, CD11b+ microglia exhibited high levels of CXCL12 expression within all CC regions at 6 wk compared with CC microglia evaluated after 12 wk of CPZ exposure (Fig. 1D). The elevated levels of CXCL12 observed after 6 wk of CPZ ingestion are consistent with the notion that ongoing oligodendrocyte and OPC apoptosis at this timepoint promotes activation of astrocytes and microglia (20). After 12 wk of CPZ exposure, these populations are depleted and trigger less glial activation through exposure to cellular debris. Prior in vitro studies reported that OPCs express the CXCL12 receptor, CXCR4, whose activation leads to their migration, proliferation, and differentiation (17–19). Consistent with this, the increase in CXCL12 expression within the demyelinating CC at 6-wk postexposure to CPZ coincided with a significant increase in the levels of CXCR4 mRNA (P = 0.0312) at this time-point, compared with unexposed mice, as assessed by qRT-PCR (Fig. 2A). The vast majority of CXCR4+ cells were determined to be NG2+ OPCs via double-label IHC (Fig. 2B). Levels of CXCR4 mRNA within the CC of mice after 12 wk of CPZ ingestion were no different from those detected in naive CC, which was significantly lower than CXCR4 mRNA levels detected at the 6-wk time-point (P = 0.0178) (Fig. 2A). Consistent with the qRT-PCR results, quantitative confocal analysis of CXCR4+NG2+ cells at these
Fig. 2. CXCR4 expression in the CC after CPZ exposure. (A) QRT-PCR analysis of CXCR4 expression in dissected whole CC derived from naive mice and from mice after 6 and 12 wk of CPZ ingestion. Each group represents three to five individual animals, *, P < 0.05. (B) Confocal IHC analysis of CC from naive mice and after 6 and 12 wk of CPZ ingestion. Images were stained for NG2 (green), CXCL12 (red), and nuclei (blue). Higher power images of boxed area in image from 6wk time-point are shown. Representative images shown for three sections from three mice in two separate experiments. IC = isotype control. (Scale bars, 25 μm.) (C) Quantitation of NG2+ cells that express CXCL12 within rostral, central, and caudal sections of the CC in naive mice (white bars) and after 6 (gray bars) and 12 (black bars) wk of CPZ exposure; n = 6 images taken from three to five mice/ group. *, P < 0.05 and **, P < 0.005.
Patel et al.
PNAS | June 15, 2010 | vol. 107 | no. 24 | 11063
NEUROSCIENCE
performed with 8-wk-old C57BL/6 mice, which were fed CPZsupplemented chow for 6 and 12 wk and then examined for expression of CXCL12 and CXCR4 within the CC. These timepoints were based on prior studies, indicating that accumulated OPCs observed within the CC after 6 wk of CPZ undergo apoptosis and are depleted in mice following 12 wk of exposure (20). Dissected whole CC derived from mice after 6 wk of CPZ ingestion exhibited significantly elevated levels (P = 0.0174) of CXCL12 mRNA compared with CC derived from naive animals (Fig. 1A), as assessed by quantitative RT-PCR (qRT-PCR). Levels of CXCL12 mRNA within CC derived from mice after 12 wk of CPZ ingestion were elevated compared with CC from naive animals but this did not reach statistical significance. Consistent with the RNA analysis, evaluation of CXCL12 protein expression within the CC of mice after 6 and 12 wk of CPZ ingestion revealed an increase in expression compared with naive controls (Fig. 1B). Double-label immunohistochemistry (IHC) revealed that cellular sources of CXCL12 within the CC of CPZexposed mice included both GFAP+ astrocytes and CD11b+ microglia. Assessment of rostral-caudal patterns of CXCL12 expression within the CC of mice after 6 and 12 wk of CPZ ingestion revealed significant increases over baseline within both cell types and in all regions (Fig. 1 C and D). GFAP+ astrocytes within the caudal CC of mice at 6 wk of CPZ ingestion exhibited the highest
time-points revealed significantly higher numbers within all regions of the CC at the 6-wk time-point only (Fig. 2C). The expression patterns of CXCL12 and CXCR4 in the CC of mice after CPZ ingestion suggest that these molecules regulate the responses of OPCs recruited into this CNS region in response to demyelination. CXCR4 Antagonism Prevents Remyelination Within the CC After Cessation of CPZ Exposure. Because higher numbers of CXCR4+NG2+ cells
were detected within the CC of mice after 6 wk of CPZ ingestion compared with control mice, a time-point when accumulated OPCs will commence remyelination if CPZ feeding ceases, we hypothesized that CXCL12 mediates the differentiation of OPCs into mature oligodendrocytes. To test this, we treated CPZexposed mice with the CXCR4 antagonist AMD3100, which specifically inhibits binding of CXCL12 to CXCR4 (23). AMD 3100, which has a plasma half-life of 0.9 h in rodents after i.p. injection (24), was continually dosed via the s.c. implantation of drug-infused osmotic pumps, as previously described (25). Continuous administration of AMD3100 for 2 wk, begun after 6 wk of CPZ ingestion, at the time of refeeding with normal chow, led to increased numbers of CXCR4+NG2+ cells within the CC compared with mice that received vehicle (PBS) alone (Fig. 3A). In addition, oligodendrocytes were in characteristic linear alignment, indicating normal myelination, within the CC of mice that received PBS, whereas the CC of AMD3100-treated mice did not exhibit this pattern (Fig. 3A, arrowheads). Quantitative IHC of CXCR4+NG2+ cells within central sections of the CC were significantly increased (P = 0.01031). Rostral (P = 0.09423) and caudal (P = 0.06576) regions were also increased in AMD3100treated mice compared with PBS-treated controls; however, these differences did not reach significance (Fig. 3B). Analysis of CNPase, which is expressed by more mature OPCs, revealed decreased numbers of CNPase+ OPCs within the CC of AMD3100treated mice compared with those that received PBS (Fig. 3C), which reached significance (P = 0.0376) within the caudal CC (Fig. 3D). Remyelination, as assessed by both Luxol-fast blue (LFB)-staining at 2 wk (Fig. 3E) and IHC detection of MBP at 4 wk (Fig. 3F) after PBS versus AMD3100 treatment, were dramatically abrogated in the context of CXCR4 antagonism. Quantitative assessment of MBP staining throughout the CC revealed a significant decrease (P = 0.0147) in AMD3100-treated versus PBS-treated animals (Fig. 3G). Taken altogether, these data support the notion that CXCR4 activation is required for OPC-mediated recovery from CPZ-induced demyelination. In addition to the recruitment of neural precursors, CXCL12 has been reported to affect both their proliferation and differentiation (14, 17, 19). Studies in mice exposed to CPZ indicate that NG2+ precursors proliferate within areas surrounding the lateral ventricle before migrating into the CC, where they differentiate into mature oligodendrocytes (26, 27). To test whether CXCR4 antagonism during remyelination affects the proliferation of OPCs, we performed in vivo bromodeoxyuridine (BrDU) incorporation studies in mice treated with PBS versus AMD3100 after 6 wk of CPZ exposure. A significant increase in the number of NG2+BrDU+ cells within rostral subventricular zones (P = 0.0132), but not the CC, was observed in AMD3100-treated animals compared with PBS-treated controls (Fig. 4). These data suggest that CXCR4 antagonism prevents cell-cycle exit of NG2+ cells within the SVZ of CPZ-exposed mice but does not affect the proliferation of cells present within the CC during remyelination. Taken altogether, these data support the notion that CXCR4 antagonism primarily blocks the maturation of OPCs into mature oligodendrocytes within the CC. In Vivo CXCR4 RNA Silencing Inhibits Remyelination After CPZMediated Demyelination. Because pharmacological agents may
induce nonspecific effects, we also used genetic approaches to block CXCR4 signaling via lentivirus delivery of CXCR4 shRNA directly into the CC of mice after 6 wk of CPZ exposure. Two lentiviral constructs were generated for this study, one expressing CXCR4 shRNA and one expressing a nonsense shRNA (Fig. 5A). The CXCR4 shRNA used was selected for best performance after an initial in vitro screen of five different constructs (SI Materials and Methods and Fig. S1). Glial infection with lentivirus expressing the CXCR4 shRNA revealed up to a 50% decrease in 11064 | www.pnas.org/cgi/doi/10.1073/pnas.1006301107
Fig. 3. CXCR4 antagonism prevents maturation of OPCs and remyelination. (A) Confocal IHC analysis of CC from mice after 6 wk of CPZ ingestion then 2 wk of treatment with PBS (Left) or AMD3100 (Right, AMD). Sections were stained for CXCR4 (green), NG2 (red), and nuclei (blue). Note linear organization of NG2-negative, oligodendrocyte nuclei in PBS-treated specimen (arrowheads). (Inset) Isotype control. (Scale bar, 25 μm.) (B) Quantitation of CXCR4+NG2+ cells within rostral, central and caudal sections of the CC in naive mice (white bars) and after 6 (gray bars) and 12 (black bars) wk of CPZ exposure; n = 6 images taken from four mice/group. *, P < 0.05. (C) Sections from CC of PBS- (Left) or AMD3100- (Right, AMD) treated mice were stained for CNPase (green) and nuclei (blue). (Inset) Isotype control. (Scale bar, 25 μm.) (D) Quantitation of CNPase+ cells within rostral, central, and caudal sections of the CC in naive mice (white bars) and after 6 (gray bars) and 12 (black bars) wk of CPZ exposure; n = 6 images taken from four mice/group. *, P < 0.05. (E) LFB myelin staining of CC in mice at 0 and 6 wk of CPZ treatment and after 6 wk of CPZ treatment plus 2 wk of treatment with PBS or AMD3100 (AMD). (F) IHC analysis of MBP expression (red) within CC of naive mice (Upper Left) and in mice treated with 4 wk of PBS (Lower Left) or AMD3100 (Lower Right, AMD) after 6 wk of CPZ ingestion. Nuclei counterstained with DAPI (blue). (Scale bar, 50 μm.) (G) Quantitation of MBP expression within all regions of the CC in mice treated with 4 wk of PBS or AMD3100 (AMD) after 6 wk of CPZ ingestion; n = 6 images taken from four mice/group. *, P < 0.05.
CXCR4 expression, as assessed by both qRT-PCR in primary astrocytes (Fig. 5B) and quantitative confocal IHC of primary oligodendroglia (Fig. 5C), compared with infection with lentivirus expressing the nonsense shRNA. After stereotactic injection of lentivirus expressing either CXCR4 or nonsense shRNAs into the CC of naive mice, GFP-expressing cells could be identified throughout the CC, in all directions examined (Fig. 5D). To determine whether silencing CXCR4 expression in accumulated OPCs within the demyelinated CC prevents remyelination, we stereotactically injected lentivirus-expressing CXCR4 versus nonsense shRNAs into the CC of mice after 6 wk of CPZ ingestion. Evaluation of CXCR4 protein expression within 2 wk of lentiviral infection, during the period of remyelination determined that delivery of CXCR4 shRNA led to a significant Patel et al.
decrease in the numbers of CXCR4+ cells within the CC compared with delivery of nonsense shRNA (P = 0.0388) (Fig. 6A), although NG2+ cell numbers were increased (Fig. 6B) and CNPase+ cell numbers were decreased (Fig. 6C), similar to effects observed during treatment with AMD3100 (Fig. 3B). Myelin staining with LFB (Fig. 6D) or IHC for MBP expression (Fig. 6E) in mice with silenced CXCR4 RNA was decreased compared with mice injected with nonsense shRNA bearing lentivirus. CXCR4 silencing significantly decreased MBP expression (P = 0.0124) when areas of GFP-expression were quantitatively compared with control lentivirus-infected mice (Fig. 6E). These data are consistent with those obtained using pharmacological antagonism of CXCR4 and indicate that CXCR4 is critical for remyelination of the CC after CPZ-mediated demyelination. Discussion Repair of demyelinated lesions requires the migration, proliferation, and differentiation of OPCs, which originate in subventricular zones distant from white matter areas within the CNS. The expression of chemoattractant molecules, which regulate these processes during development, is therefore a critical component of the CNS injury response, ensuring that neural precursor cells appropriately migrate and mature to replace damaged cells. Our data indicate that up-regulation of the chemokine CXCL12 is essential for the differentiation of CXCR4-expressing OPCs into mature oligodendrocytes within the demyelinated CC in vivo. We show that the pattern of CXCL12 expression mirrors rostral-
Fig. 5. CXCR4 RNA silencing in glial cells in vitro. (A) Schematic of lentiviral construct used to silence CXCR4 expression. pLKO.1 plasmids carrying shRNAs (CXCR4 and nonsense) driven by the human U6 promoter and HIV packaging signal were reengineered to contain sequence encoding GFP with a polyA tail driven by the CMV promoter. CXCR4 shRNA and nonsense sequences are shown. (B) qPCR analysis of CXCR4 mRNA levels in primary astrocytes infected with CXCR4 shRNA lentivirus at multiplicity of infection (MOI) of 5, 10, and 20. Data are presented as the fold-change in CXCR4 mRNA levels normalized to those observed in astrocytes infected with nonsense shRNA lentivirus. (C) Confocal IHC analysis of GFP (green), CXCR4 (red), and NG2 (blue) expression by primary oligodendroglia infected with nonsense (Left) versus CXCR4 (Middle) shRNA lentiviruses at 20 MOIs reveals significant knock-down of CXCR4 as assessed by confocal IHC (Right); n = 5–6 images taken from three replicates/group, *, P < 0.05. (D) Low power images depicting stereotactic infection of CC with GFP-expressing lentiviruses expressing nonsense (Left) and CXCR4 (Right) shRNAs. Arrow = site of injection. (Scale bar, 50 μm.)
Patel et al.
PNAS | June 15, 2010 | vol. 107 | no. 24 | 11065
NEUROSCIENCE
Fig. 4. CXCR4 activation differentially affects OPC proliferation within the SVZ and CC. (A) Confocal IHC analysis of rostral subventricular zones (SVZ) and CC from mice after 6 wk of CPZ ingestion plus 2 wk of treatment with PBS (Left) or AMD3100 (Right) with daily i.p. injection of BrDU. Images were stained for BrDU (green), NG2 (red), and nuclei (blue). Representative images shown from three sections from three individual mice per group. (Scale bars, 25 μm.) (B) Quantitation of BrDU+NG2+ cells within rostral, central, and caudal sections of the CC and SVZ of mice after 6 wk of CPZ ingestion and 2 wk of treatment with PBS (gray bars) or AMD3100 (black bars); n = 6 images taken from three mice/group, *, P < 0.05.
caudal patterns of pathology previously observed during CPZmediated demyelination (21, 22). Thus, highest levels of CXCL12 were expressed by activated astrocytes within caudal sections, which exhibit the most extensive demyelination and glial cell activation. Accordingly, CXCR4+NG2+ OPCs accumulated within all sections of the CC but were highest in the caudal sections. Loss of CXCR4 signaling, via pharmacologic blockade or RNA silencing, increased the numbers of NG2+ OPCs, decreased the numbers of CNPase+ OPCs, and abrogated expression of myelin proteins, preventing remyelination of the CC after cessation of CPZ exposure. CXCR4 blockade also increased the numbers of NG2+ cells proliferating within rostral subventricular areas. These data identify CXCL12 as a critical regulator of OPC remyelination within damaged white matter of the adult CNS. Studies examining repair of the CNS in pathological states indicate that endogenous mechanisms that recapitulate developmental processes are triggered during injury. For example, the Notch1 receptor and its ligand, Jagged1, which regulate myelination during CNS development, were recently shown to similarly affect OPC differentiation during remyelination after lysolecithin injection into the CC (28). Several in vitro studies suggest that CXCR4-expressing OPCs exhibit chemotactic and maturational responses to CXCL12, which promotes their expression of myelin proteins and differentiation into oligodendrocytes (17, 18). Additional studies have shown that exposure to myelin basic protein leads to up-regulation of CXCL12 within astrocytes in vitro (29), implicating oligodendrocyte injury in glial CXCL12 expression. Our data demonstrating that CXCL12-mediated CXCR4 activation is required for myelin production by differentiating OPCs is consistent with proposed roles for these molecules during developmental myelination. The CPZ model of demyelination exhibits characteristic patterns of demyelination and extent of remyelination that depend on the length of CPZ exposure. Thus, animals that ingest CPZ for shorter time periods, such as the 6-wk period used in our study, exhibit extensive accumulation of OPCs within the CC and efficient remyelination upon toxin removal, whereas longer periods of CPZ ingestion lead to OPC depletion and chronic demyelination (20). The lack of remyelination in these studies has been attributed to increased apoptosis of OPCs in mice kept on CPZsupplemented diets. We observed that levels of CXCL12 and CXCR4 after 6 wk of CPZ ingestion are higher than those observed after 12 wk of exposure and that this coincides with overall numbers of CXCR4+NG2+ cells. Thus, impaired remyelination during chronic CPZ ingestion may also be due to decreased levels of CXCL12 with consequent lack of CXCR4 activation in remaining OPCs. As activated microglia and astrocytes were both sources of CXCL12, it is possible that alterations in the inflammatory response of one or both of these cell types contributes to remyelination failure in the setting of chronic CPZ exposure.
OPCs within MS lesions (35) suggests that the recruitment of these precursors remains intact. Remyelination failure in the setting of CXCL12 expression therefore suggests defects in additional maturational cues or interference with CXCR4 signaling, which could be accomplished via aberrant expression of CXCR7. Studies in zebrafish development suggest CXCR7 is a nonsignaling, decoy receptor that sequesters CXCL12 from the microenvironment, thereby limiting CXCR4 activation (10). Up-regulation of CXCR7 within astrocytes and endothelial cells has been observed in damaged brain regions after focal ischemia in rodent models and in human CNS tissues derived from patients with metastatic brain tumors (33, 36). Increased expression of CXCR7 by these cells types within MS lesions could therefore decrease CXCR4 activation in OPCs, limiting their differentiation into mature oligodendrocytes. Studies using CXCR7 inhibitors in murine models of MS and analysis of CXCR7 expression in MS lesions will reveal essential information regarding these interactions. The elucidation of endogenous CNS repair mechanisms is critical for the development of biologically-based therapies that promote remyelination in patients with MS. We have shown that up-regulation of CXCL12 within injured white matter leads to the differentiation of CXCR4-expressing OPCs into mature oligodendrocytes that produce myelin. Our work suggests that inhibition of this process leads to remyelination failure and suggests that CXCL12 and/or its receptors may provide targets to enhance recovery from demyelinating episodes in MS. Materials and Methods Mouse Model of CPZ Demyelination. Eight-wk-old C57BL/6 mice were fed ad libitum 0.2% CPZ mixed in standard rodent chow for 6–12 wk to induce demyelination of the CC. Animals are taken off the CPZ-supplemented feed and returned to normal feed for 2–4 wk before sacrificing for CNS tissue. Fig. 6. In vivo CXCR4 silencing prevents OPC maturation and remyelination. (A and B) Confocal IHC analysis of CC from mice after 6 wk of CPZ ingestion followed by stereotactic injection of lentivirus bearing shRNAs (Nonsense, Left, versus CXCR4, Right). Images depict GFP expression (green) with staining for CXCR4 (A), NG2 (B), and CNPase (C). (Scale bar, 20 μm.) Quantitation of CXCR4+ (A), NG2+ (B), and CNPase+ (C) cells within CC of mice injected with lentivirus expressing Nonsense and CXCR4 shRNAs after 6 wk of CPZ exposure (Far Right); n = 6 images taken from three to five mice/group. *, P < 0.05. (D) LFB-stained sections of CC from mice injected with lentivirus expressing Nonsense (Upper) and CXCR4 (Lower) shRNAs after 6 wk of CPZ exposure. (A– D) Representative images shown for three sections from three to five mice in two separate experiments. Note linear area without remyelination in mouse injected with CXCR4 shRNA lentivirus (arrowheads). (E) IHC analysis of MBP expression (red) within CC of mice injected with lentivirus expressing GFP (green) plus Nonsense (Left) and CXCR4 (Right) shRNAs after 6 wk of CPZ exposure. Note area of demyelination in CC of mouse with CXCR4 RNA silencing that correspond to sites of GFP staining (arrowheads). Nuclei counterstained with DAPI (blue). Quantitation of MBP staining in areas of GFP expression was performed on two to three sections/mouse in three to five mice/group in two separate experiments, *, P < 0.05 (Far Right).
Prior in vitro studies reported that CXCL12 promotes the proliferation of neural progenitor cells but has minimal effects on OPCs (17, 30). In our study, CXCR4 antagonism led to increased proliferation in OPCs within SVZ but not within the remyelinating CC. Thus, the increase in numbers of NG2+ cells in the CC during CXCR4 antagonism is unlikely to be due to proliferative effects of CXCL12 at this site. Increased proliferation in subventricular areas, however, might be a direct effect of CXCR4 blockade or occur in response to signals triggered by the lack of OPC differentiation within the CC. Further studies examining additional molecules that promote OPC maturation are required to better define the connection between these CNS regions during remyelination. Sources of CXCL12 in the normal adult CNS include neurons, endothelial cells, and meninges (13, 15, 25), whereas astrocyte expression of CXCL12 has been observed only in several pathological states including MS, HIV-1 encephalitis, and malignancy (29, 31– 33). In studies of MS tissues, CXCL12 expression is observed within activated astrocytes in both silent and active MS lesions (34), suggesting its expression occurs in response to injury. The presence of 11066 | www.pnas.org/cgi/doi/10.1073/pnas.1006301107
Primary Oligodendroglial Cultures. Neonatal C57BL/6 mice were used for oligodendroglial cultures. Brains were extracted and the cortices were removed and incubated in 25 mg/mL trypsin (Sigma) for 30 min at 37 °C and plated in MEM (Invitrogen) supplemented with 10% horse serum, 10% FBS (FBS), 20 mM glucose, and 10 ng/mL recombinant mouse epidermal growth factor (EGF; Invitrogen). Mixed glia cultures were grown for 7–9 d and then oligodendrocytes were separated by shaking the flasks for 4–6 h at 37 °C at 100–150 rpm. Oligodendrocytes were obtained via differential adhesion on nontissue culture-treated plates for 30 min. Collected cells were plated on poly-d-lysine coated coverslips for 4–7 d before infection with lentivirus. Quantitative PCR. Total RNA and quantitative real-time (qRT)-PCR was performed using primers generated using Primer Express 2.0 software (Applied Biosystems) for CXCL12 and CXCR4 as published previously (25). Antibodies. The following antibodies were used in this study: CXCL12 rabbit polyclonal, (eBioscience); IgG isotype, (Jackson ImmunoResearch); CXCR4 polyclonal, (eBioscience); anti-CD11b (BD Pharmingen); anti-NG2 (Millipore), anti-MBP (Abcam), anti-BrDU (Sigma) and anti- GFAP (Invitrogen) antibodies. Immunohistochemistry and Quantitative Confocal Microscopy. Frozen sections were prepared as described (25). Detection of CXCL12, GFAP, CD11b, NG2, MBP, BrDU, and CXCR4 antibodies were all assayed as described (25). Primary antibodies were detected with goat anti-rabbit or anti-rat IgG conjugated to Alexa 555 or Alexa 488 (Molecular Probes) and nuclei were counterstained with ToPro3. Control sections were incubated with isotype-matched IgG. Sections were analyzed using a Zeiss LSM 510 laser scanning confocal microscope and mean area was quantified via Volocity image analysis software (Improvision). Cell morphology, the presence of nuclei and isotype control images were used to set threshold parameters to choose GFAP, CD11b, NG2, or CNPase positive cells. For single-labeling studies, positive cell numbers were determined via cell counting of cell marker positive, nucleated cells. For double-labeling studies, cells that were CXCL12 or CXCR4 positive in addition to cell markers were included in an area analysis. The criteria for CXCL12 or CXCR4 positive staining was based on comparisons to regions outside of the corpus callosum known to express either CXCL12 (i.e., venules) (25) or CXCR4 (i.e., ependymal cells) (15). The sum of the area of double positive cells was calculated per mouse and the mean area was calculated per treatment group. These set parameters were applied to every image, six images per region of corpus callosum for three to five mice per treatment group.
Patel et al.
Luxol-Fast Blue Analysis. Frozen sections were stained for myelin using 0.1% LFB and counter stained with 0.1% cresyl violet. The sections were differentiated with 0.5% lithum carbonate and counterstained with cresyl violet for 20 min and mounted for imaging using a Zeiss Axioskop 40 light microscope. In Vivo AMD3100 Treatment. Continuous dose treatment with AMD3100 or PBS was achieved using 14-d osmotic pumps with an infusion rate is 0.5 μL/h (Alzet) as described (25). Osmotic pumps were implanted s.c. at the time of CPZ cessation.
For in vivo lentiviral delivery of shRNAs, anesthetized mice (groups of three to five) were positioned in a stereotaxic apparatus and a small incision in the scalp, and the lateral CC was marked using the following stereotaxic coordinates, relative to Bregma: anterior-posterior −2, medial-lateral ±0.5, and dorsal-ventral −1.2 (37). After drilling through the skull, a 33-gauge needle attached to a 10-μL Hamilton syringe was lowered into the corpus callosum according to the dorsal-ventral coordinate. A nanoinjector pump (Stoelting) delivered 2 μL of lentivirus expressing CXCR4 or nonsense shRNAs at a rate of 0.2 μL/min, after which the needle was left in place for 5 min to ensure complete diffusion of virus. The mice were continued on CPZ feed for 5 d after injection to ensure viral infection and RNA silencing within accumulated OPCs. Mice were then switched to standard chow for 2 wk before evaluation of CNS tissues. In Vivo Bromodeoxyuridine Incorporation. Bromodeoxyuridine (BrDU; Sigma) in PBS (100 mg/kg) was injected into the peritoneal cavity of experimental mice starting 48 h after pump implantation (AMD3100 pumps or PBS pumps), then every 8 h for 4 d, and then killed for CNS tissue.
Lentiviral Delivery of CXCR4 shRNA. Plasmids carrying CXCR4 and nonsense shRNAs under the mouse U6 promoter (Sigma) were re-engineered to carry a cytomegalovirus promoter-EGFP (CMV–EGFP) expression cassette (provided by M. Sands, Washington University, St. Louis, MO) to create a vector that simultaneously produces CXCR4 shRNAs and an EGFP reporter. Plasmid was verified using restriction digest using PvuII (Sigma constructs) and SalI (lentiviral plasmids). 1 μg of shRNA construct-containing plasmid along with 1 μg of an 8:1 ratio mix of lentiviral plasmids pHR’8.2deltaR (packaging) and pCMVVSV-G (envelope) (provided by S. Stewart, Washington University, St. Louis, MO) were mixed with 6 μL of FuGENE (Roche) for 30 min at room temperature and used to transfect 293T cells. After 48 h, media with virus was collected, filtered and added to primary glial cultures for 72 h, which were then evaluated for CXCR4 mRNA and protein.
ACKNOWLEDGMENTS. The authors thank M. Goldberg for manuscript comments and D. Holman for experimental and R. Lewis for technical assistance. This study was supported by National Institutes of Health/ National Institute of Neurological Disorders and Stroke Grant NS059560 and National Multiple Sclerosis Society Grant RG3982 (to R.S.K.).
1. Weiner HL (2009) The challenge of multiple sclerosis: how do we cure a chronic heterogeneous disease? Ann Neurol 65:239–248. 2. Chari DM (2007) Remyelination in multiple sclerosis. Int Rev Neurobiol 79:589–620. 3. Blakemore WF, Keirstead HS (1999) The origin of remyelinating cells in the central nervous system. J Neuroimmunol 98:69–76. 4. Wilson HC, Scolding NJ, Raine CS (2006) Co-expression of PDGF alpha receptor and NG2 by oligodendrocyte precursors in human CNS and multiple sclerosis lesions. J Neuroimmunol 176:162–173. 5. Nakahara J, Kanekura K, Nawa M, Aiso S, Suzuki N (2009) Abnormal expression of TIP30 and arrested nucleocytoplasmic transport within oligodendrocyte precursor cells in multiple sclerosis. J Clin Invest 119:169–181. 6. Polito A, Reynolds R (2005) NG2-expressing cells as oligodendrocyte progenitors in the normal and demyelinated adult central nervous system. J Anat 207:707–716. 7. Chang A, Nishiyama A, Peterson J, Prineas J, Trapp BD (2000) NG2-positive oligodendrocyte progenitor cells in adult human brain and multiple sclerosis lesions. J Neurosci 20: 6404–6412. 8. Klein RS, Rubin JB (2004) Immune and nervous system CXCL12 and CXCR4: parallel roles in patterning and plasticity. Trends Immunol 25:306–314. 9. D’Apuzzo M, et al. (1997) The chemokine SDF-1, stromal cell-derived factor 1, attracts early stage B cell precursors via the chemokine receptor CXCR4. Eur J Immunol 27: 1788–1793. 10. Boldajipour B, et al. (2008) Control of chemokine-guided cell migration by ligand sequestration. (Translated from eng). Cell 132:463–473. 11. McGrath KE, Koniski AD, Maltby KM, McGann JK, Palis J (1999) Embryonic expression and function of the chemokine SDF-1 and its receptor, CXCR4. Dev Biol 213:442–456. 12. Schonemeier B, et al. (2008) Regional and cellular localization of the CXCl12/SDF-1 chemokine receptor CXCR7 in the developing and adult rat brain. J Comp Neurol 510: 207–220. 13. Lu M, Grove EA, Miller RJ (2002) Abnormal development of the hippocampal dentate gyrus in mice lacking the CXCR4 chemokine receptor. Proc Natl Acad Sci USA 99: 7090–7095. 14. Klein RS, et al. (2001) SDF-1 alpha induces chemotaxis and enhances Sonic hedgehoginduced proliferation of cerebellar granule cells. Development 128:1971–1981. 15. Stumm RK, et al. (2002) A dual role for the SDF-1/CXCR4 chemokine receptor system in adult brain: isoform-selective regulation of SDF-1 expression modulates CXCR4dependent neuronal plasticity and cerebral leukocyte recruitment after focal ischemia. J Neurosci 22:5865–5878. 16. Imitola J, et al. (2004) Directed migration of neural stem cells to sites of CNS injury by the stromal cell-derived factor 1alpha/CXC chemokine receptor 4 pathway. Proc Natl Acad Sci USA 101:18117–18122. 17. Maysami S, et al. (2006) Modulation of rat oligodendrocyte precursor cells by the chemokine CXCL12. Neuroreport 17:1187–1190. 18. Kadi L, et al. (2006) Differential effects of chemokines on oligodendrocyte precursor proliferation and myelin formation in vitro. J Neuroimmunol 174:133–146. 19. Dziembowska M, et al. (2005) A role for CXCR4 signaling in survival and migration of neural and oligodendrocyte precursors. Glia 50:258–269.
20. Mason JL, et al. (2004) Oligodendrocytes and progenitors become progressively depleted within chronically demyelinated lesions. Am J Pathol 164:1673–1682. 21. Wu QZ, et al. (2008) MRI identification of the rostral-caudal pattern of pathology within the corpus callosum in the cuprizone mouse model. J Magn Reson Imaging 27: 446–453. 22. Stidworthy MF, Genoud S, Suter U, Mantei N, Franklin RJ (2003) Quantifying the early stages of remyelination following cuprizone-induced demyelination. Brain Pathol 13: 329–339. 23. Hatse S, Princen K, Bridger G, De Clercq E, Schols D (2002) Chemokine receptor inhibition by AMD3100 is strictly confined to CXCR4. FEBS Lett 527:255–262. 24. Datema R, et al. (1996) Antiviral efficacy in vivo of the anti-human immunodeficiency virus bicyclam SDZ SID 791 (JM 3100), an inhibitor of infectious cell entry. Antimicrob Agents Chemother 40:750–754. 25. McCandless EE, Wang Q, Woerner BM, Harper JM, Klein RS (2006) CXCL12 limits inflammation by localizing mononuclear infiltrates to the perivascular space during experimental autoimmune encephalomyelitis. J Immunol 177:8053–8064. 26. Kumar S, Biancotti JC, Yamaguchi M, de Vellis J (2007) Combination of growth factors enhances remyelination in a cuprizone-induced demyelination mouse model. Neurochem Res 32:783–797. 27. Arnett HA, et al. (2001) TNF alpha promotes proliferation of oligodendrocyte progenitors and remyelination. Nat Neurosci 4:1116–1122. 28. Zhang Y, et al. (2009) Notch1 signaling plays a role in regulating precursor differentiation during CNS remyelination. Proc Natl Acad Sci U S A . 29. Calderon TM, et al. (2006) A role for CXCL12 (SDF-1alpha) in the pathogenesis of multiple sclerosis: regulation of CXCL12 expression in astrocytes by soluble myelin basic protein. J Neuroimmunol 177:27–39. 30. Gong X, He X, Qi L, Zuo H, Xie Z (2006) Stromal cell derived factor-1 acutely promotes neural progenitor cell proliferation in vitro by a mechanism involving the ERK1/2 and PI-3K signal pathways. Cell Biol Int 30:466–471. 31. Zhang K, et al. (2003) HIV-induced metalloproteinase processing of the chemokine stromal cell derived factor-1 causes neurodegeneration. Nat Neurosci 6:1064–1071. 32. McCandless EE, et al. (2008) Pathologic expression of CXCL12 at the blood-brain barrier correlates with severity of multiple sclerosis. Am J Pathol, in press. 33. Salmaggi A, et al. (2009) CXCL12, CXCR4 and CXCR7 expression in brain metastases. Cancer Biol Ther 8:1608–1614. 34. Moll NM, et al. (2009) Imaging correlates of leukocyte accumulation and CXCR4/ CXCL12 in multiple sclerosis. Arch Neurol 66:44–53. 35. Reynolds R, et al. (2002) The response of NG2-expressing oligodendrocyte progenitors to demyelination in MOG-EAE and MS. J Neurocytol 31:523–536. 36. Schonemeier B, Schulz S, Hoellt V, Stumm R (2008) Enhanced expression of the CXCl12/SDF-1 chemokine receptor CXCR7 after cerebral ischemia in the rat brain. J Neuroimmunol. 37. Paxinos G, Franklin KBJ (2001) The Mouse Brain in Stereotaxic Coordinates (Academic, San Diego, CA).
Patel et al.
Statistical Analyses. Data were analyzed using Prism software (GraphPad Software). The Student t test was used to determine the statistical significance of qRT-PCR and histological analyses. In all experiments, P < 0.05 was considered to be significant.
PNAS | June 15, 2010 | vol. 107 | no. 24 | 11067
NEUROSCIENCE
Quantitative analyses of CXCL12- and CXCR4-expressing cells and those with BrDU incorporation within the CC of naive and CPZ-exposed animals was performed in a blinded fashion by counting cell numbers/high power field. Quantitative analysis of MBP staining in lentivirus-infected areas of the CC was performed using ImageJ (National Institutes of Health).
