Stromal Cell–Derived Factor-1 and CXCR4 Interaction Is Critical for Development of Transplant Arteriosclerosis
Background— Posttransplant chronic allograft deterioration associated with development of transplant arteriosclerosis (TA) remains an unresolved problem. Recent studies suggest that the smooth muscle cells (SMCs) constituting the neointima are derived from recipient hematopoietic stem cells (HSCs). However, the underlying mechanisms of the process are not yet fully elucidated.
Methods and Results— We examined the genes expressed in allografts at different stages of TA development using a mice aortic transplantation model. Genes were analyzed by a differential mRNA display technique. We show that stromal cell–derived factor-1α (SDF-1α) is a critical molecular target for the treatment of TA. During the course of TA, intragraft SDF-1α expression was upregulated with time, and the circulating HSCs expressing its counterreceptor CXCR4 increased in the recipients receiving allografts. CXCR4-positive HSCs, derived from transplant recipients, migrated into allografts via microvessels in the adventitia and then toward the luminal side. The HSCs differentiated into SMC-like cells, contributing to the in situ formation of the neointima. In support of a functional role for these molecules, in vivo neutralization of SDF-1α inhibited HSC mobilization and significantly attenuated neointimal formation.
Conclusions— Interaction between SDF-1α and CXCR4 plays a key role in TA development. Blockade of SDF-1α may become a new therapeutic modality for TA.
Received November 18, 2003; de novo received March 15, 2004; revision received June 4, 2004; accepted June 7, 2004.
Although advances in immunosuppressive regimens have largely overcome acute cellular rejection in clinical organ transplantation, long-term outcomes of vascularized allografts remain unsatisfactory. Graft loss occurring in the late period is predominantly due to the development of transplant arteriosclerosis (TA) in the graft arteries.1,2 TA is histopathologically characterized by formation of a diffuse neointima in the graft arteries that consists mainly of smooth muscle cells (SMCs).1,2 It is known that multifactorial events are involved in the pathogenesis of TA, both immunologic and nonimmunologic factors such as ischemia/reperfusion injury.3 Hence, neointima is formed through diverse events occurring in transplants, and proliferation of SMCs is considered a key event for TA development.2 Conventionally, the medial vascular SMCs in grafts (donor-derived) have been thought to proliferate and migrate into the intima during the process of TA.4 Refuting this wisdom, recent studies have demonstrated that the majority of neointimal SMCs are of host but not donor origin.5–8 Such SMCs differentiating from recipient precursor cells originate as hematopoietic stem cells (HSCs) from bone marrow (BM)6,7 or other unidentified tissues/organs.8 However, understanding the pathophysiology of the neointimal formation at a cellular level is currently at a relatively immature stage. Moreover, molecular-based mechanisms of TA development are not yet fully elucidated.
In an attempt to clarify the underlying molecular mechanism of TA and to search for a novel therapeutic target, we applied differential mRNA display gene analyses using a murine aorta transplant model. Here we show that the stromal cell–derived factor-1α (SDF-1α) and its receptor CXCR4 play a crucial role in TA development.
C57BL/6 (H-2b), BALB/c (H-2d), ICR, and C3H (H-2k) mice (Japan SLC Inc) were used at the age of 6 to 12 weeks. Animals were treated under institutional guidelines of animal use and care. The institutional animal care committee approved all experiments.
Murine aortic transplantation was performed with the use of a modified technique described by Sun et al.9 Briefly, a segment of donor thoracic aorta was anastomosed (end to side) to the infrarenal portion of the recipient aorta with interrupted sutures. The native aorta was ligated between the anastomotic sites. Aortic grafts were removed under anesthesia at 2, 4, 6, or 8 weeks after transplantation.
In some experiments, recipient mice received 10 μg/d per body of either neutralizing anti–SDF-1α antibody (R&D Systems) or control IgG (Jackson ImmunoReseach) intraperitoneally for 3 weeks after transplantation.
Aortic grafts were harvested and frozen at 6 weeks after transplantation. Six sections per each graft were cut from different layers through the grafts, stained with hematoxylin-eosin, and examined. Thickness of intimal and intimal+medial layers was measured from 10 sites per graft section with the use of NIH Image 1.62, and intima/intima+media ratios were calculated.
Frozen graft sections were incubated with anti–SDF-1α (Torrey Pines Biolabs), anti-CXCR4 (Santa Cruz), or anti-CD31 antibodies (PharMingen), visualized with a secondary goat anti-rabbit antibody (Dako), and examined under light microscopy.
Double-immunofluorescence staining was performed by using first antibodies against Cy3-conjugated α-SMA (Sigma), phycoerythrin-conjugated CD31 (PharMingen), MOMA-2 (Serotec), and biotin-conjugated H-2Kk antigen for C3H (PharMingen). Primary antibody was followed by incubation with fluorescein isothiocyanate (FITC)–conjugated anti-rabbit antibody (Jackson ImmunoReseach) or streptavidin-phycoerythrin (PharMingen). Sections were examined under a confocal microscope (Olympus).
RNA Isolation and Differential mRNA Display
Total RNA was extracted from aortic graft tissues, and cDNA was generated as follows: total RNA (0.2 μg) was reverse transcribed in a 50-μL reaction mixture with reverse transcriptase and 3′oligo(dT) primer. Control reactions were performed without reverse transcriptase. Differential display analysis was performed as previously described.10 The cDNAs were amplified by polymerase chain reaction (PCR) with 35S-dATP, 5′ primers that were arbitrary 12-mers, and 3′ primers that matched those used in the cDNA synthesis (Differential Display Kit; Takara). The reaction was performed at 94°C for 30 seconds, 40°C for 1 minute, and 72°C for 1 minute with 40 cycles. Radiolabeled PCR products were electrophoresed with the use of denaturing 4% polyacrylamide gels. All reactions were duplicated. Differentially upregulated bands were harvested from the sequencing gels and reamplified by PCR for cloning.
TA Cloning and Sequence Analysis
Reamplified cDNA fragments were cloned into the plasmid vector pCRII with the use of the TA cloning kit (Invitrogen) and sequenced (BLAST sequence analysis program).
cDNA of aortic graft was analyzed by a LightCycler (Roche Diagnostic) with the SDF-1α–specific primers 5′-CAAGTGGAAAAATACACCGT-3′ (forward) and 5′-CTGGTGGTTTTTGGTAACTA-3′ (reverse) and with 2 hybridization probes: 5′-TATTTGAAGTGGAGCCATAGTAATGCC-fluorescein-3′ and 5′-LCred705-GTAGATATCTCTATGATCTTGAGCTACTGGCA-3′. cDNA (1 μL) was amplified with the use of the LightCycler DNA Master Hybridization Probes Kit (Roche Diagnostic). SDF-1 transcripts were normalized with β-actin transcripts.
Polymerase Chain Reaction
Total RNA was extracted from the allograft neointima, which was mechanically peeled under a microscope. cDNAs were amplified by PCR with the CXCR4-specific primers 5′-AAGTGGATCTCCAT-CACAGA-3′ (forward) and 5′-AGGAGGCACAGAGATTGAA-3′ (reverse) and with the SRY antigen primers 5′-AGGAGGCACAGAGATTGAA-3′ (forward) and 5′-TGCAGGCTGTAAAATGCCA-3′ (reverse). PCR was performed at 94°C for 1 minute, 50°C for 1 minute, and 72°C for 1 minute for 35 cycles.
To enumerate the circulating HSCs, peripheral blood cells were incubated at 4°C with biotin-conjugated monoclonal antibodies against lineage markers (CD3, CD11b, CD45R/B220, Ly-6C/G, TER119), phycoerythrin-conjugated Sca-1, and allophycocyanin-conjugated c-Kit (CD117) (all PharMingen), followed by streptavidin-FITC (GIBCO-BRL). The c-Kit+/Sca-1+/Lin− cells were counted with an FACSCalibur system (Becton Dickinson). Mononuclear cells were isolated with Ficoll (Sigma). Lin− cells were further purified by negative selection against lineage markers with the use of the MACS (Miltenyi Biotech). Cells were stained with FITC/anti-CXCR4, phycoerythrin/anti–Sca-1, allophycocyanin/anti–c-Kit, and streptavidin/allophycocyanin antibodies (all PharMingen). CXCR4 expression was analyzed with the use of an FACSCalibur system (Becton Dickinson).
HSC and SMC Coculture
BM cells were collected from the femurs and tibias of C3H mice. Sca-1+/Lin− HSCs were purified with MACS (Miltenyi Biotech). Aortic SMCs were isolated from C3H mice as described previously.11 SMCs (1×105) were seeded on a 0.4-μm microporous membrane, and HSCs (5×104 per well) were seeded in a 24-well lower chamber with the use of the transwell system (Corning-Coster). Cells were cultured in 10% fetal bovine serum supplemented with Dulbecco’s modified Eagle’s medium at 37°C with 5% CO2.
All data are expressed as mean±SEM. Data were statistically analyzed by 1-way ANOVA with the use of the Fisher protected least significant difference post hoc test. A probability value <0.05 was considered significant.
SDF-1 Gene Is Upregulated in Aortic Allografts
To investigate possible molecular mechanisms underlying TA development, we searched for genes upregulated at the transcriptional level during the course of TA with the use of differential mRNA display gene analysis. C57BL/6 mice aortas were transplanted to ICR recipients. In this model, a moderate degree of cellular infiltration into the graft adventitia was noted 2 weeks after transplantation, and neointima formation became prominent after 8 weeks (data not shown). We first examined differentially expressed mRNAs of 2- and 8-week aortic allografts and found 16 genes that were differentially upregulated in allografts after 8 weeks compared with those at 2 weeks (Table).
To strengthen the argument that these genes were related to TA development rather than other factors such as the animal strain combination, we analyzed histology of aortic grafts and evaluated the aforementioned genes in aortic allografts using a BALB/c-to-C3H mouse combination. As noted in the C57BL/6-to-ICR transplants, we found a strong cellular infiltrate around microvessels of the adventitia in the BALB/c allografts at 2 weeks after transplantation (Figure 1a). At this stage, the media was intact and covered by a single endothelium layer. After 6 weeks, neointima had developed in the BALB/c allografts, and predominantly macrophages and foam cells had infiltrated into the media (Figure 1b). These histological features were comparable to those of the 8-week allografts in a C57BL/6-to-ICR combination. In contrast, intimal thickening was not evident in syngeneic grafts studied after 6 weeks (Figure 1c).
Using the BALB/c-to-C3H combination, we further compared mRNA expression levels of the 16 genes among 6-week isografts and 2- and 6-week allografts by real-time reverse transcription–PCR12 (n=4 grafts per group). Of the 16 genes tested, the expression of 3 genes, SDF-1α, SH2-containing inositol phosphatase-1, and dexamethasone-induced product, were significantly upregulated in the allografts during the development of TA compared with the isografts (Table and Figure 1d). SDF-1 is known to be involved in the development of angiogenesis,13,14 pathogenesis of atherosclerosis,15 and vascularization in the gastrointestinal tract.16 We thus focused our attention on the role of SDF-1 in the development of TA.
CXCR4-Positive Cells Enter Allografts via Microvessels in the Adventitia
If the SDF-1 gene were to play a role in TA, it would also have to be translated. We therefore examined its protein expression by immunohistochemistry. In the 2-week allografts, SDF-1α+ cells were scattered in the adventitia (Figure 2a). Further detailed examination of the serial sections revealed that the SDF-1α+ cells were also located around newly formed CD31+ microvessels in the adventitia (Figure 2b). By 4 weeks after transplantation, SDF-1α+ cells were more frequent in the adventitia, and they also appeared in the media (Figure 2c). After 6 weeks, we found intense and diffuse expression of SDF-1α not only in the adventitia but also in the neointima (Figure 2d). Interestingly, the pattern and kinetics of CXCR4 expression were very similar to those of SDF-1α at both 2 (Figure 2e) and 6 weeks (Figure 2f). However, there was little if any expression of either SDF-1α or CXCR4 in the 6-week isografts (Figure 2g).
CXCR4-Positive Cells Originate From Recipient and Form Neointima
Confocal microscopic examination revealed that in the 6-week allografts, neointimal SMCs defined by positive α-actin staining coexpressed SDF-1α (Figure 3a). Neither endothelial cells (CD31+; Figure 3b) nor macrophages (MOMA-2+; Figure 3c) expressed SDF-1α. The expression pattern of CXCR4 was similar to that of SDF-1α (Figure 3d). To identify whether the CXCR4+ cells originate from transplant recipients, we costained 6-week allografts with anti-CXCR4 and anti-recipient MHC class I (H-2Kk) antibodies. The majority of neointima consisting of CXCR4+ cells (Figure 3e) were of recipient origin, as examined by confocal microscopy. In a sex-mismatched transplant combination (female-to-male), we further verified that the dissected allograft neointima expresses both CXCR4 and SRY antigen mRNA (Figure 3f). Taken together, these results suggest that the majority of CXCR4+ cells are of recipient origin, and they enter allografts mainly through microvessels in the adventitia, migrate toward the media, and form neointima.
CXCR4-Positive HSCs Increase in Recipient’s Circulation After Aortic Allografting
Recent studies show that most neointimal SMCs in allografts originate from host BM-derived HSCs.7,8 Therefore, we first enumerated the circulating HSCs (c-Kit+/Sca-1+/Lin−) in peripheral blood after transplantation. The numbers of circulating HSCs were very low in naïve C3H mice (1.2±0.6 cells) and were increased slightly in isograft recipients (3.6±1.0 cells). In contrast, the numbers were significantly higher in allograft recipients (15.6±5.1 cells; P<0.01 versus control; Figure 4a). We further hypothesized that these HSCs express CXCR4. Indeed, the majority of circulating HSCs expressed CXCR4 (Figure 4b).
BM-Derived HSCs Differentiate Into SMC-like Cells In Vitro
Given that CXCR4+ HSCs are mobilized into peripheral blood in allograft recipients, we examined whether BM-derived HSCs could differentiate into SMCs. We also asked whether such differentiation, if it occurred, required cell-cell contact by using a transwell cell culture system. Murine SMCs were seeded on the porous membrane, and Sca-1+/Lin− HSCs were placed in the lower chamber. After 3 days in culture, HSCs became spindle-shaped cells (data not shown). This appearance was prominent at 10 days (Figure 4c). These spindle-shaped cells expressed α-actin (Figure 4d), a marker of SMCs, showing that direct cell-cell interactions were not essential for this process. In contrast, HSCs cultured in the absence of SMCs did not show spindle-shaped morphology and remained α-actin negative at 10 days (data not shown). These results indicate that allografting elicits the migration of CXCR4+ HSCs from BM into the allografts that differentiate into SMC-like cells and contribute to the in situ formation of the neointima.
SDF-1α Neutralization Inhibits Mobilization of HSCs and Abrogates TA Formation
Finally, we investigated the potential in vivo role of SDF-1 and CXCR4 interaction in TA development. We injected a neutralizing anti–SDF-1α or control antibody into C3H recipient after transplantation of BALB/c aortic allografts. This approach for SDF-1 blockade was applied because gene mutation of SDF-1 and CXCR4 is fatal in mice.17 SDF-1 neutralization significantly ameliorated neointimal formation (0.17±0.01 versus 0.56±0.02, respectively; P<0.001; Figure 5a, 5b) and reduced infiltrating CXCR4+ cells in the neointima (Figure 5c) compared with control. Additionally, the average number of circulating HSCs in the anti–SDF-1α antibody–treated mice was significantly lower than that in control (5.1±1.8 versus 17.1±5.9 cells, respectively; P<0.01; Figure 5d). These results suggest that SDF-1/CXCR4 interaction contributes to HSC mobilization into allografts and to the development of TA.
SDF-1 is a chemoattractant that was originally identified as a BM stromal cell–secreting factor that supports the proliferation and development of B cells.18 Mice lacking SDF-1 gene expression are subjected to fetal death and present abnormalities in B lymphopoiesis, BM myelopoiesis, blood vessel formation, and ventricular septal development.17 It has been shown that SDF-1α is primarily involved in the trafficking of CD34+ HSCs19 and that elevation of plasma SDF-1α induces HSC mobilization.20,21 Indeed, we found that SDF-1α/CXCR4 interaction is involved in mobilization of HSCs (Figures 4a and 5⇑c) into aortic allografts and that presumably those cells migrate from the adventitia toward the intima (Figure 2a to 2f) and differentiate into SMCs expressing α-actin (Figures 3a and 4⇑d). More importantly, in vivo blockade of SDF-1α resulted in inhibition of both HSC mobilization and TA development (Figure 5a to 5d).
Several chemokine interactions through receptors such as CXCR322 and CCR523 have been shown to play an important role during acute cellular rejection, namely, alloreactive T cell–mediated immune responses. Because SDF-1 is expressed on lymphocytes, monocytes, and many other cell types,15,24,25 it is possible that SDF-1/CXCR4 interaction is also involved in T cell–mediated immune reactions. We do not exclude this possibility; however, expression of SDF-1 and CXCR4 was not necessarily associated with graft-infiltrating CD4+ and CD8+ T cells during the course of TA in the present study (data not shown). Furthermore, the majority of the circulating HSCs, which differentiate into SMC-like cells, expressed CXCR4 (Figure 4b) but not CCR5 (data not shown). These results suggest that interaction between SDF-1 and CXCR4 plays a major role in recruitment of HSCs but not alloreactive T cells into the allograft during TA development.
The origin of neointimal SMCs in TA lesions remains a point of controversy. Several recent studies show that most neointimal SMCs in the heart7 and aortic6 allografts originate from host BM-derived HSCs. Conversely, Hu et al8 demonstrated that the neointimal SMCs are of recipient origin but not from the BM. In this study we have shown that the frequency of circulating HSCs increased in allograft recipients (Figure 4a). In addition, BM-derived HSCs could differentiate into SMC-like cells in vitro when cocultured with SMCs (Figure 4c and 4d). Our results suggest that SMC progenitors stem from the recipient and that BM could be one of the major sources for the progenitors.
Despite the fact that neointimal SMCs mainly differentiate from recipient progenitor cells, their route of entry and process of mobilization are unclear. Our detailed observation indicated that progenitor cells, ie, CXCR4+ cells, enter from newly formed microvessels of the adventitia and migrate toward the luminal side in allografts (Figure 2). This finding is also consistent with the notion that adventitial neovascularization in the allografts precedes neointima formation,26 which facilitates infiltration of recipient-derived cells, including SMC progenitor cells as well as alloreactive and proinflammatory leukocytes.
Interestingly, our present data demonstrated that the CXCR4+ HSCs increased in the circulation and their mobilization into the grafts occurred when the transplants were allogeneic but not syngeneic (Figure 4a). Because neutralization of SDF-1α inhibited mobilization of these CXCR4+ HSCs in the peripheral blood and neointima formation (Figure 5a, 5b), SDF-1/CXCR4 interaction seems to be critical for recruiting SMC progenitor HSCs into the allografts. Indeed, a recent study demonstrated that SDF-1 expression is highly induced in the damaged liver and that expressed SDF-1 elicits recruitment and migration of CXCR4+ HSCs into the injured site, which contributes to tissue repairing.27 Thus, SDF-1 seems to be a key molecule for regulation of HSC recruitment that leads to physiological remodeling of tissues/organs; however, such process could be beneficial or vice versa under different circumstances. In future studies, it is important to determine whether mobilized HSCs express both CXCR4 and its ligand SDF-1α or whether CXCR4+ HSCs are recruited by SDF-1α, which is secreted from graft-composing cells such as vascular adventitial cells.
In conclusion, SDF-1α/CXCR4 interaction contributes to formation of neointima in aortic allografts. In vivo neutralization of SDF-1α significantly inhibited HSC mobilization and attenuated neointimal formation. Blockade of the SDF-1α/CXCR4 interaction may become a new therapeutic modality for TA.
This study was supported by the grants from the Ministry of Education, Culture, Sports, Science, and Technology (13470248) and the Ministry of Health, Labor, and Welfare of Japan. We thank Dr Fritz H. Bach (Lewis Thomas Professor, Beth Israel Deaconess Medical Center, Harvard Medical School) for comments and review of the manuscript.
Hu Y, Davison F, Ludewig B, et al. Smooth muscle cells in transplant atherosclerotic lesions are originated from recipients, but not bone marrow progenitor cells. Circulation. 2002; 106: 1834–1839.
Liang P, Pardee AB. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science. 1992; 257: 967–971.
Chamley-Campbell J, Campbell GR, Ross R. The smooth muscle cell in culture. Physiol Rev. 1979; 59: 1–61.
Abi-Younes S, Sauty A, Mach F, et al. The stromal cell-derived factor-1 chemokine is a potent platelet agonist highly expressed in atherosclerotic plaques. Circ Res. 2000; 86: 131–138.
Nagasawa T, Kikutani H, Kishimoto T. Molecular cloning and structure of a pre-B-cell growth-stimulating factor. Proc Natl Acad Sci U S A. 1994; 91: 2305–2309.
Aiuti A, Webb IJ, Bleul C, et al. The chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med. 1997; 185: 111–120.
Hattori K, Heissig B, Tashiro K, et al. Plasma elevation of stromal cell-derived factor-1 induces mobilization of mature and immature hematopoietic progenitor and stem cells. Blood. 2001; 97: 3354–3360.
Shen H, Cheng T, Olszak I, et al. CXCR-4 desensitization is associated with tissue localization of hemopoietic progenitor cells. J Immunol. 2001; 166: 5027–5033.
Melter M, Exeni A, Reinders ME, et al. Expression of the chemokine receptor CXCR3 and its ligand IP-10 during human cardiac allograft rejection. Circulation. 2001; 104: 2558–2564.
Bleul CC, Fuhlbrigge RC, Casasnovas JM, et al. A highly efficacious lymphocyte chemoattractant, stromal cell-derived factor 1 (SDF-1). J Exp Med. 1996; 184: 1101–1109.
Gupta SK, Pillarisetti K, Lysko PG. Modulation of CXCR4 expression and SDF-1alpha functional activity during differentiation of human monocytes and macrophages. J Leukocyte Biol. 1999; 66: 135–143.
Bhattacharya V, McSweeney PA, Shi Q, et al. Enhanced endothelialization and microvessel formation in polyester grafts seeded with CD34(+) bone marrow cells. Blood. 2000; 95: 581–585.