Graft-Extrinsic Cells Predominate in Vein Graft Arterialization
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Duke University Department of Medicine (Cardiology), Duke University Medical Center, Durham, NC.
Correspondence to Karsten Peppel, Box 3187, Duke University Medical Center, 447 CARL Building, Science Drive, Durham, NC 27710. E-mail karsten.peppel@duke.edu
Abstract
Objective— Vein graft disease involves neointimal smooth muscle cells, the origins of which are unclear. This study sought to characterize and quantitate vein graft infiltration by cells extrinsic to the graft in a mouse model of vein graft disease.
Methods and Results— Inferior vena cava-to-carotid artery interposition grafting between C57Bl/6 and congenic ?-galactosidase–expressing ROSA26 mice was performed. Vein grafts were harvested 6 weeks postoperatively and stained with X-gal. More than 60% of neointimal cells derived from the recipient, and 50% of these cells expressed smooth muscle -actin. The distribution of donor and recipient-derived cells within this vein graft wall layer was distinctly focal, consistent with focal infiltration and expansion of progenitor cells. When bone marrow transplantation with congenic green fluorescent protein (GFP)-expressing cells was used in vein graft recipients 1 month before surgery, abundant GFP-expressing cells appeared in the media, but not the neointima, of mature grafts. Endothelial cells in mature grafts derived from graft-intrinsic and graft-extrinsic sources and were, in part, of bone marrow origin.
Conclusions— Cells extrinsic to the graft, including bone marrow-derived cells, predominate during vein graft remodeling.
Key Words: smooth muscle cells ? vein graft neointimal hyperplasia ? mouse models ? bone marrow transplantation ? endothelium
Introduction
Autologous vein grafting ranks among the most common of surgical procedures, but its long-term success remains compromised by vein graft disease.1 Up to 30% of grafts become stenotic and require intervention within 2 years because of hemodynamically significant neointimal hyperplasia.2 Ten years after surgery, only approximately 50% of venous grafts are still patent, and only half of these are free of significant stenosis.1,3–5 Vein graft failure in humans has been intimately linked to atherosclerosis-predisposing neointimal hyperplasia, involving smooth muscle cells (SMCs) and extracellular matrix.1,3
The origin of neointimal SMCs remains incompletely characterized, despite its therapeutic implications. SMCs proliferating in the vein graft media have been shown to migrate into the neointima.6,7 In addition, arterial myofibroblasts contribute to neointimal hyperplasia in a pig endoluminal coronary artery injury model,8 and bone marrow (BM)-derived cells contribute neointimal SMCs in models of atherosclerosis9 or transplant arteriopathy.10,11 To date, however, there has been no quantitation of vein graft neointimal cell precursors.
To address this issue, we recently developed a unique murine model of vein grafting that uses sutured arteriovenous anastomoses, and neointimal hyperplasia mimics early human vein graft disease by reaching steady-state thickness without engendering significant luminal stenosis.12 In this model, we used ?-galactosidase–expressing ROSA26 mice as vein graft donors and/or recipients to quantitate the extent to which the mature vein graft wall and neointima comprise cells derived from precursors residing either within or outside of the vein graft at the time of its implantation. To further differentiate the origins of graft-extrinsic cells, we replaced the bone marrow of vein graft recipient mice with that of green fluorescent protein (GFP)-expressing animals.
Methods
Animals
Adult male C57BL/6 (wild-type [WT]) mice, and congenic mice that either express Lac Z (?gal, ROSA26 strain) or GFP, were purchased from Jackson Labs. All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee and complied with the Guide for the Care and Use of Laboratory Animals.
Vein Graft Surgery
Interposition vein graft surgery was performed as described previously.12 Inferior vena cavae (IVC) from C57Bl/6 or ROSA26 donor mice were anastomosed to the right common carotid artery of recipient mice, yielding four different donor/graft recipient groups, as follows: (1) WT/WT; (2) WT/?gal; (3) ?gal/WT; and (4) ?gal/?gal. All mice were between 12 and 20 weeks old.
IVC Graft Harvest and Analysis
Grafts were harvested 2 or 6 weeks postoperatively and prepared for histochemical and immunofluorescent analysis as described previously.12 For analysis of ?-galactosidase expression, the grafts were excised without fixation and incubated at 37°C (16 hours) in x-gal staining solution (10 mmol/L K4Fe(CN)6, 10 mmol/L K3Fe(CN)6, 2 mmol/L MgCl2 and 500 μg/mL x-gal (5-bromo-4-chloro-3-indoyl-?-D-galactopyranoside) in phosphate-buffered saline (PBS). Grafts were subsequently fixed in 10% formalin for 24 hours and photographed en bloc using RT SPOT 2 CCD camera (Diagnostic Instruments). Grafts were then embedded in optimal cutting temperature (OCT) compound to facilitate cryosectioning for light and immunofluorescence microscopy.
Bone Marrow Transplantation
C57Bl/6 mice received 950 rads of gamma irradiation at age 14 weeks and were infused with 107 BM cells procured from femurs of GFP mice. Successful BM engraftment was confirmed 3 weeks later by examining blood smears. Mice thus made chimeric for BM GFP expression received WT IVC grafts 4 weeks after bone marrow transplantation (BMT). Grafts were harvested 6 weeks after operation, perfused briefly with saline, and frozen in OCT compound.
Vein Graft Microscopic Analysis
Immunofluorescence and immunoperoxidase staining and microscopy were performed as previously described.12,13 T cells and macrophages were identified with rabbit anti-CD3- antibody (DAKO) or BM8 (Research Diagnostic), respectively. Endothelial cells were identified with rabbit anti-factor VIII (DAKO). Grafts from GFP BM chimeric mice were OCT-embedded, sectioned at 5 to 10 μm, and imaged for GFP fluorescence. Immediately thereafter, sections were processed for immunofluorescence. Positive controls for CD3 and macrophage staining were murine spleen and raw 264.7 macrophages, respectively. Negative controls were incubated with cognate non-immune IgG in lieu of primary antibody. The nuclear DNA dye Hoechst 33342 (10 μg/mL) was added to secondary antibody incubations. Sequential images of single microscopic fields were taken as described.13 In overlays of fluorescent and bright-field images, the black background of fluorescence images was removed to facilitate identification of non-SMCs in the vein grafts.
Quantitation of Graft Donor-Derived and Recipient-Derived Cells
Vein graft sections from 5 graft regions were each imaged at 4 to 8 different clock hours, at x440 magnification. Five different locations taken over the length of the graft were analyzed for each specimen. Both WT IVC grafts placed into ROSA26 recipients (WT/?gal) and ROSA26 IVC grafts placed into WT recipients (?gal/WT) were analyzed. Consequently, percent recipient-derived cells was calculated in either of 2 ways: (1) 100 x (number of x-gal–stained cells)÷(number of total [ie, Hoechst-stained] cells) for ROSA26/WT specimens; and (2) 100 x [1 - (number of x-gal–stained cells)÷(number of total cells)] for WT/ROSA26 specimens. More than 100 cells were counted per x440 field. Cells were counted only within the vein graft neointima.
Results
To address whether vein graft SMCs derive from progenitors extrinsic to the vein graft, we placed IVC segments from ROSA26 (?gal) or C57Bl/6 (WT) donor mice as interposition grafts into the right common carotid artery of ?gal or WT recipient mice (Figure 1). The ROSA26 strain is congenic to C57Bl/6, and thus transplantation of vascular grafts between these two strains does not engender graft rejection and confounding transplant arteriopathy. Therefore, this model system is well suited to assess infiltration of vein grafts by graft-extrinsic, recipient-derived cells. Grafts were harvested 6 weeks postoperatively, at a time when vein graft wall thickness and neointimal hyperplasia have reached steady-state in this model.12 To determine the relative contribution of ?gal-expressing cells to the vein graft wall, intact vein grafts were stained en bloc for ?gal expression (upper photographs, Figure 1A through 1D). WT IVCs grafted into WT recipient mice did not display any blue staining, either in whole explants (Figure 1A, bottom) or in cross-sections (Figure 1A, top), demonstrating the absence of activated macrophages in these specimens.14 All vein grafts involving ?gal mice as either donors or recipients showed some degree of blue staining, which traversed the entire length of the graft (Figure 1B through 1D, arrowheads), and varied by specimen group (Figure 1B through 1D, lower photos). Blue cellular staining evident in the WT IVC-to-?gal recipient specimens (Figure 1C) clearly demonstrated that vein graft wall cells may derive from sources that are extrinsic to the vein graft at the time of implantation. Similarly, the prevalence of blue cellular staining of the ?gal IVC-to-WT recipient specimens (Figure 1B) was substantially less than that obtained in ?gal IVC-to-?gal recipient specimens (Figure 1D). Thus, the contributions of graft recipient (ie, graft-extrinsic) cells to the vein graft wall were observed in WT IVC-to-?gal and ?gal IVC-to-WT specimens.
Figure 1. Heterogeneous origin of vein graft wall cells. Interposition vein grafting was performed with the indicated donor and recipient mice, and 6-week-old vein grafts were harvested and stained with x-gal. Explanted specimens were imaged before embedding in OCT compound (upper images in each panel) and after sectioning and counterstaining with eosin (lower image in each panel). The explanted specimens comprise (from left to right) the recipient’s right common carotid artery caudal to the bifurcation, the vein graft segment (between the arrowheads), the proximal common carotid, and the aortic arch with remnants of the left common carotid. Representative explant specimens are shown (n=3 for WT/WT and ?gal/?gal; 5 for WT/?gal and ?gal/WT). Scale bar in (A) is 5 mm. Photomicrographs (original magnification x440) are representative of 20 per vein graft. Scale bar is 100 μm.
In patients, the developing vein graft neointima predominately comprises cells expressing smooth muscle -actin and represents an atherosclerosis prone area that forms the foundation for the later development of graft atheroma.1 Similarly, smooth muscle -actin–expressing cells predominate in the neointima of murine grafts (Figure 2). Indeed, smooth muscle -actin–expressing cells reside almost exclusively in the collagen-poor, cytoplasm-rich neointima of mouse vein isografts (Figure 2C) and rabbit jugular vein autografts, which have long been used to model human vein graft disease (data not shown).3,15 To quantitate the percentage of graft-intrinsic and graft-extrinsic cells in the neointima of these murine vein grafts, we superimposed x-gal stained bright-field and red fluorescence SM -actin–stained images of identical WT-to-?gal vein graft cross-sections (Figure 2G). Total cellularity of the sections was assessed by quantitating nuclei after incubation with Hoechst 33342 (Figure 2D). Analysis of ?gal-to-WT and WT-to-?gal grafts revealed that approximately 60% of cells in the vein graft neointima develop from graft-extrinsic precursors (Table). Moreover, smooth muscle -actin expression was evident in 66% of graft donor-recipient and 56% of graft recipient-derived neointimal cells (Table). Thus, graft-intrinsic and graft-extrinsic precursors give rise predominantly to SMCs of the mature vein graft neointima. There was only scant evidence of inflammatory cell infiltration at 6 weeks (Table). ?gal IVC-to-WT grafts were analyzed in parallel and yielded congruent results.
Figure 2. Prevalence of SMCs among cells derived from vein graft-extrinsic progenitors. A, Vein graft paraffin section was stained with a modified connective tissue stain (see Methods) and imaged at 110x original magnification. B, C, The area boxed in (A) was imaged at 440x original magnification, and a serial section was stained with Hoechst 33342 and Cy3-conjugated anti-SM -actin. The fluorescence and light images are aligned, with their luminal surfaces upward. Images are representative of 7 different vein graft samples. D–G, Six-week-old WT IVC-to-?gal and ?gal IVC-to-WT specimens were processed as in Figure 1, with the omission of eosin counterstain. Each specimen was then imaged sequentially under fluorescence (D, F) and bright-field conditions (E) to identify nuclei (D), SM -actin–positive cells (F) and ?gal-positive cells (E) within the same microscopic field. G, Images in (E) and (F) were digitally merged, as described in the Methods section, to allow identification of SMCs (red staining cells) that are graft extrinsic (blue) or graft intrinsic (non-blue). Arrow indicates graft intrinsic SMCs; arrowhead, graft extrinsic SMCs; open arrow, graft extrinsic non-SMCs. Shown is a single microscopic field from one WT IVC-to-?gal specimen, representative of 4 samples analyzed from n 5 mice in each group. Original magnification x440; scale bar in (G) 100 μm. H, The percent graft-extrinsic cells was calculated as in Methods for WT IVC-to-?gal and ?gal IVC-to-WT specimens (n=5 for each group) at 1-mm intervals across the length of the grafts, as indicated. Plotted at each location are the mean±SE. Scale bar is 200 μm (A) and 100 μm (B, G).
Characterization of Vein Graft Neointimal Cells
Multiple possibilities exist for the origin of these prevalent graft-extrinsic precursor cells. It is conceivable that some of the graft-infiltrating cells derived by inward migration from the media of the adjacent carotid artery. To examine this possibility, we quantitated the percentage of graft-intrinsic and graft-extrinsic cells over the entire length of the vein graft (Figure 2H). If vascular cell migration into the graft from the anastomoses were a predominant source of graft-extrinsic precursor cells, then we would expect to see a higher prevalence of recipient cells in graft sections near the anastomoses, as compared with sections obtained from the center of the grafts (because graft-extrinsic cells constituted only 60% of total graft wall cells). The center of the vein grafts (segment 3) showed only 20% to 25% fewer recipient-derived cells than the peri-anastomotic portions of the graft (segments 1 and 5). Thus, although not excluding a contributing role of lateral-migrating arterial SMCs to the formation of vein graft lesions, this analysis does not support the notion of a predominance of anastomotic vascular cells in the population of cells originating outside of the graft.
To address the origin of graft-extrinsic cells further, we examined serial sections of 2-week-old vein grafts, in which the media and neointima are still enlarging.12 (Indeed, steady-state vein graft wall thickness is reached only by 4 weeks postoperatively in this animal model.) In these 2-week-old specimens, isolated foci of cells are evident in many sections—even in sections remote from the anastomoses (Figure 3A). These foci show numerous actively proliferating cells, as assessed by staining for proliferating cell nuclear antigen (PCNA) (Figure 3C, red arrow). This cell proliferation distinguishes vein graft proliferative foci from adjacent cells that constitute part of postsurgical adhesions (Figure 3, yellow arrow). Staining for smooth muscle -actin reveals that these foci are composed of SMCs and non-SMCs. These focal, cellular expansions of the graft wall could, of course, originate from either graft-extrinsic or graft-intrinsic cells. If these foci represent expansions of recipient-derived cells, however (please see later), then this finding suggests that graft-extrinsic precursor cells do infiltrate the graft and originate from sources distinct from the peri-anastomotic carotid artery—either the circulating blood or adventitial adhesions. Unfortunately, x-gal staining of these foci was inadequate to determine whether their progenitor cells originated in the graft donor or recipient—perhaps because insufficient ?gal is present in these cells undergoing relatively rapid mitosis. Therefore, to examine the origin of the focal cellular expansions observed in these immature grafts, we returned again to an analysis of mature, 6-week-old vein grafts in which remodeling has ceased.12 We reasoned that if focal cellular expansions in the arterializing vein graft enlarge to constitute the mature graft wall, then these mature vein grafts should evince focal staining for donor-derived and recipient-derived cells. This expectation was fulfilled in our 6-week-old vein grafts stained with x-gal (Figure 3D). Quantitatively, graft-extrinsic neointimal foci demonstrated a 60% relative increase in graft-extrinsic cells, compared with the neointima taken as a whole (Figure 3E). Thus, focal cellular vein graft wall expansions observed in 2-week-old specimens almost certainly did develop from both graft-intrinsic and graft-extrinsic precursor. The focal cellular nodules (remote from anastomoses) in 2-week-old vein grafts and focal expansions of graft-intrinsic and graft-extrinsic cell populations in mature vein grafts together support a model in which graft-extrinsic cells originate from precursors entering the graft via a transendothelial and/or transadventitial route in addition to cells entering the vein graft via migration from the anastomoses. What is the derivation of the graft extrinsic cells that do not originate from the adjacent carotid artery? Recent interest and controversy has focused on the role of BM-derived cells in vascular repair.9,11,16–18 To examine the potential contribution of BM-derived circulating precursor cells to vein graft remodeling, we performed vein graft surgeries in C57Bl/6 WT recipient mice that had their BM replaced with that of congenic GFP-expressing mice. Six-week-old mature grafts harvested from these chimeric mice showed prominent, albeit inhomogeneous, infiltration of BM-derived cells into the media and adventitia, but notably not the neointima, of the grafts (Figure 4C). These BM-derived cells comprised primarily cells of the macrophage lineage and did not express SMC -actin (Figure 4D).
Figure 3. Development of vein graft neointima involves focal cellular expansions of both graft-intrinsic and graft-extrinsic precursor cells. A–C, Two-week-old vein grafts (n=3) were formalin-fixed, paraffin-embedded, sectioned at 200-μm intervals, and stained either with a modified connective tissue stain ("Masson") or with immunofluorescence for SM -actin and PCNA, along with Hoechst 33342. A, Isolated areas of focal cellular infiltration can be seen (red arrow) in otherwise uniformly thin vein graft walls (arrowheads). Adjacent graft-extrinsic tissue is indicated by the yellow arrow. Serial sections 500 μm proximal and distal to this section demonstrated no focal proliferation of cells (not shown). The boxed area is enlarged in (B) and (C). B, The focal expansion contains SMC -actin–expressing and -actin–nonexpressing cells. A serial section of (A) is shown, with Hoechst and Cy3-anti-SM -actin images superimposed. C, The same section shown in (B) is portrayed, imaged only for anti-PCNA immunofluorescence. D, Six-week-old vein grafts, with the indicated donors and recipients, were stained, sliced, and counterstained as described for Figure 1. Foci of x-gal–stained neointimal cells have been circled, demonstrating progeny of graft-extrinsic and graft-intrinsic cells, as indicated. E, In foci like those identified in (D), the prevalence of cells derived from graft-extrinsic precursors and the prevalence of SMCs were calculated as in Figure 2 and plotted as mean±SE from 3 foci from each of 5 specimens. Representative specimens (n 5 for each graft type) are shown. Original magnification x110 (A, D) and x440 (B, C); scale bar is 200 μm (A, D).
Figure 4. Bone marrow derived cells participate in vein graft remodeling. GFP BM chimeric mice were generated as described in Methods and served as recipients of WT vein grafts. Grafts were harvested after 6 weeks and embedded in OCT. Serial sections were stained with H&E (A) or imaged for GFP fluorescence. Specimens imaged for GFP were subsequently processed for either SM -actin expression (B) or macrophage presence (D). C, E, The GFP fluorescent and corresponding immunofluorescent images were merged. Neointimal (arrowheads) and medial plus adventitial (open arrows) boundaries are shown.
Vein graft remodeling is initiated, in part, by hemodynamic and ischemic injury to the endothelium of the graft. Re-endothelialization is thought to play an important part in limiting vascular remodeling,1 We therefore sought to determine the origin of the endothelial cells that participate in vein graft remodeling. In WT IVC grafts that were placed into ?gal recipients and stained with x-gal, we found evidence of graft-intrinsic (non-stained) and graft-extrinsic (blue-stained) endothelial cells (Figure 5 A and 5B). In addition, when we examined WT grafts that were placed into GFP BM chimeric recipients (Figure 5C through 5F), we found a small prevalence (10%) of GFP-positive cells among the vein graft endothelial cells. This indicates that re-endothelialization in this model derives from multiple sources, including cells of the graft donor and endothelial progenitor cells that originate from BM intrinsic and extrinsic compartments of the graft recipient. These findings thus contrast with those of Xu et al, who recently used Tie2-driven ?gal expression to show that, in their vein grafts, endothelium is essentially replaced by recipient cells.19 Hemodynamic differences between vein graft model systems may underlie this discrepancy in endothelial data.
Figure 5. Vein graft endothelium derives from graft-intrinsic and graft-extrinsic cells, including bone marrow precursors. Isogenic vein grafting was performed using IVCs from WT mice placed into recipients that were either (A, B) ROSA26 or chimeric WT mice previously transplanted with whole BM from GFP-expressing mice (C–F). Grafts were harvested 6 weeks postoperatively. Serial sections of grafts placed in ROSA26 mice were stained either with x-gal to identify ?gal-expressing cells (A) or with anti-factor VIII antibody (immunoperoxidase) to identify endothelial cells (B). Endothelial cells expressing ?gal (closed arrows) or not (open arrows) are indicated. (C–F). Grafts placed into WT mice with GFP BM were frozen in OCT compound, sliced at 5 μm, and then imaged for GFP fluorescence (C, D). Subsequently, these sections were subjected to immunoperoxidase staining for factor VIII to identify endothelial cells (E, F). GFP fluorescence identifies BM-derived cells that contribute to endothelial layer formation (D, F) or that do not (C, E). The lumen of each specimen is oriented upward. Original magnification x1100 (A, B) or x440 (C–F).
Discussion
We have shown for the first time to our knowledge that even in the absence of atherosclerosis, the majority of vein graft neointimal and medial cells derive from precursors that are extrinsic to the vein at the time of grafting. We have also provided the first evidence that the arterializing vein graft wall develops largely from focal expansions of progenitor cells, which originate within and outside of the vein graft at the time of implantation. Finally, we have shown that endothelial remodeling of vein grafts is accomplished by the expansion of graft-intrinsic as well as graft-extrinsic cells, and that BM-derived cells participate in this process. Importantly, these novel insights into vein graft biology derive from a murine vein graft system that we have shown to model human vein grafts, with regard to their surgical anastomoses as well as the distribution, composition and extent of neointimal hyperplasia within the grafts.12
Sources of graft-extrinsic neointimal cells in our murine model remain to be determined, but potential sources include migrating vascular cells from the adjacent carotid artery,20 as well as progenitor cells infiltrating the graft through either circulating blood (vasa vasora21 or transendothelial diapedesis22) or postsurgical adventitial adhesions. Surgical trauma at the anastomoses activates SMCs of the adjacent artery to proliferate and ultimately migrate into vein grafts—a process clearly delineated in rat iliac vein grafts.20 Nevertheless, we can infer that activated SMCs of the adjacent artery do not account for all of the graft-extrinsic cells in the media and neointima of our murine model, from two complementary sources of data: (1) 2-week-old vein grafts demonstrated focal expansions of smooth muscle -actin expressing cells remote from the anastomoses (Figure 3); and (2) our lengthwise analysis of 6-week-old grafts found no significant decline in the prevalence of vein graft recipient cells as a function of distance from the graft anastomoses (Figure 2H). The relative contribution of adjacent arterial SMCs to vein graft neointima doubtless depends heavily on the graft system used.
Data from our GFP-chimeric vein graft recipients clearly demonstrate that BM-derived cells contribute to the genesis of the vein graft wall—principally in the media, adventitia, and (to a much smaller degree) the endothelium. However, the non-endothelial progeny of these BM-derived cells appeared to be largely macrophages and did not express SM -actin. It is possible that these cells represented fibrocytes, which had not yet differentiated into a SM -actin–expressing phenotype.23 Nevertheless, the absence of SM -actin expression in these cells may explain why Hu et al found no BM contribution to their mouse vein graft SMC populations.24 These authors replaced the BM of their vein graft recipient mice with marrow from SM22-lacZ transgenic mice, which express ?gal only in SMC-like cells. Neointimal SMCs have been shown to derive from BM in atherosclerosis;9 however, different combinations of growth factors present in atheromata may influence the differentiation of BM-derived precursor cells.23,24
The finding that vein graft neointima formation proceeds without bone marrow contribution to a large part from graft extrinsic cells distinguishes vein graft remodeling from other forms of arterial remodeling. In arterial remodeling associated with transplant arteriopathy11,25,26 and mechanical injury,18 BM-derived cells contribute to neointimal hyperplasia. Vascular progenitor cells have been isolated from other organs as well.27,28 Majka et al demonstrated the potential for skeletal muscle-derived progenitor cells to differentiate into endothelial cells as well as SMCs.28 This study also demonstrated the slow replacement of skeletal progenitor cells by cells derived from the bone marrow. Thus, our data support the notion that vascular injury elicits divergent repair responses, dependent on the type of injury inflicted.
Approximately 30% of our vein graft neointimal cells did not express SM -actin and may comprise fibroblasts or fibrocytes.23 Adventitial fibroblasts have been reported to migrate into the neointima of balloon-injured rat and pig arteries, and ultimately transdifferentiate into myofibroblasts that express SM -actin.8,29,30 Because adventitial fibroblasts in our model could originate from the vein graft adventitia itself or the peri-graft (postsurgical) adhesions, these fibroblasts could derive from either graft-intrinsic or graft-extrinsic cells. Moreover, graft-extrinsic fibroblasts originating from surgical adhesions need not originate from the BM—consistent with the absence of neointimal GFP fluorescence in grafts from BM-chimeric mice (Figure 4).
The novel finding of focal proliferation in arterializing vein grafts (Figure 3) recalls data regarding the clonal origins of fibrous plaques in atherosclerosis.31,32 Proliferative cellular foci in vein grafts were demonstrable by congruent data from 2- and 6-week-old vein grafts (Figure 3), which evince foci originating from either graft-intrinsic or graft-extrinsic precursor cells. Whether these distinct foci of cells represent clonal, or rather colonial, cell growth currently transcends the scope of this investigation.
Because most vein graft neointimal cells derive from precursors extrinsic to the vein graft, we can make novel mechanistic inferences about successful therapies for vein graft neointimal hyperplasia, like the E2F decoy.33 Successful therapies must, of course, target proliferation, migration, and extracellular matrix secretion by cells intrinsic to the implanted vein graft, because these cells ultimately constitute 40% of neointimal cells. In addition, however, these therapies must target recruitment to the graft of cells initially extrinsic to the graft, perhaps by reducing the inflammatory reaction associated with vein graft endothelial damage22 and/or the chemokine/growth factor secretion by vein graft cells.34
Acknowledgments
This work was supported by grants HL63288 (to N.J.F.) and HL64744 (to K.P.) and American Heart Association grants-in-aid (to N.J.F., K.P.).
References
Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation. 1998; 97: 916–931.
Mills JL, Fujitani RM, Taylor SM. The characteristics and anatomic distribution of lesions that cause reversed vein graft failure: a five-year prospective study. J Vasc Surg. 1993; 17: 195–204.
Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br J Surg. 1994; 81: 1254–1269.
Nwasokwa ON. Coronary artery bypass graft disease. Ann Intern Med. 1995; 123: 528–545.
Mehta D, Izzat MB, Bryan AJ, Angelini GD. Towards the prevention of vein graft failure. Int J Cardiol. 1997; 62: S55–S63.
Hart CE, Kraiss LW, Vergel S, Gilbertson D, Kenagy R, Kirkman T, Crandall DL, Tickle S, Finney H, Yarranton G, Clowes AW. PDGFbeta receptor blockade inhibits intimal hyperplasia in the baboon. Circulation. 1999; 99: 564–569.
Rectenwald JE, Moldawer LL, Huber TS, Seeger JM, Ozaki CK. Direct evidence for cytokine involvement in neointimal hyperplasia. Circulation. 2000; 102: 1697–1702.
Shi Y, O’Brien JE, Fard A, Mannion JD, Wang D, Zalewski A. Adventitial myofibroblasts contribute to neointimal formation in injured porcine coronary arteries. Circulation. 1996; 94: 1655–1664.
Sata M, Saiura A, Kunisato A, Tojo A, Okada S, Tokuhisa T, Hirai H, Makuuchi M, Hirata Y, Nagai R. Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis. Nat Med. 2002; 8: 403–409.
Han CI, Campbell GR, Campbell JH. Circulating bone marrow cells can contribute to neointimal formation. J Vasc Res. 2001; 38: 113–119.
Shimizu K, Sugiyama S, Aikawa M, Fukumoto Y, Rabkin E, Libby P, Mitchell RN. Host bone-marrow cells are a source of donor intimal smooth- muscle-like cells in murine aortic transplant arteriopathy. Nat Med. 2001; 7: 738–741.
Zhang L, Hagen PO, Kisslo J, Peppel K, Freedman NJ. Neointimal hyperplasia rapidly reaches steady state in a novel murine vein graft model. J Vasc Surg. 2002; 36: 824–832.
Peppel K, Jacobson A, Huang X, Murray JP, Oppermann M, Freedman NJ. Overexpression of G protein-coupled receptor kinase-2 in smooth muscle cells attenuates mitogenic signaling via G protein-coupled and platelet-derived growth factor receptors. Circulation. 2000; 102: 793–799.
Dannenberg A, Burstone M, Walter P, Kinsley J. A histochemical study of phagocytic and enzymatic functions of rabbit mononuclear and polymorphonuclear exudate cells and alveolar macrophages. I. Survey and quantitation of enzymes and states of cellular activation. J. Cell Biol. 1963; 17: 465–486.
Peppel K, Zhang L, Huynh TT, Huang X, Jacobson A, Brian L, Exum ST, Hagen PO, Freedman NJ. Overexpression of G protein-coupled receptor kinase-2 in smooth muscle cells reduces neointimal hyperplasia. J Mol Cell Cardiol. 2002; 34: 1399–1409.
Sata M, Tanaka K, Nagai R. Origin of smooth muscle progenitor cells: different conclusions from different models. Circulation. 2003; 107: e106–e107.
Hu Y, Davison F, Ludewig B, Erdel M, Mayr M, Url M, Dietrich H, Xu Q. Smooth muscle cells in transplant atherosclerotic lesions are originated from recipients, but not bone marrow progenitor cells. Circulation. 2002; 106: 1834–1839.
Tanaka K, Sata M, Hirata Y, Nagai R. Diverse Contribution of Bone Marrow Cells to Neointimal Hyperplasia After Mechanical Vascular Injuries. Circ Res. 2003; 93: 783–790.
Xu Q, Zhang Z, Davison F, Hu Y. Circulating progenitor cells regenerate endothelium of vein graft atherosclerosis, which is diminished in ApoE-deficient mice. Circ Res. 2003; 93: e76–e86.
Dilley RJ, McGeachie JK, Tennant M. The role of cell proliferation and migration in the development of a neo-intimal layer in veins grafted into arteries, in rats. Cell Tissue Res. 1992; 269: 281–287.
Davies MG, Hagen PO. Influence of perioperative storage solutions on long-term vein graft function and morphology. Ann Vasc Surg. 1994; 8: 150–157.
Westerband A, Mills JL, Marek JM, Heimark RL, Hunter GC, Williams SK. Immunocytochemical determination of cell type and proliferation rate in human vein graft stenoses. J Vasc Surg. 1997; 25: 64–73.
Abe R, Donnelly SC, Peng T, Bucala R, Metz CN. Peripheral blood fibrocytes: differentiation pathway and migration to wound sites. J Immunol. 2001; 166: 7556–7562.
Hu Y, Mayr M, Metzler B, Erdel M, Davison F, Xu Q. Both donor and recipient origins of smooth muscle cells in vein graft atherosclerotic lesions. Circ Res. 2002; 91: e13–e20.
Hillebrands JL, Klatter FA, Rozing J. Origin of vascular smooth muscle cells and the role of circulating stem cells in transplant arteriosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23: 380–387.
Li J, Han X, Jiang J, Zhong R, Williams GM, Pickering JG, Chow LH. Vascular smooth muscle cells of recipient origin mediate intimal expansion after aortic allotransplantation in mice. Am J Pathol. 2001; 158: 1943–1947.
Werner N, Junk S, Laufs U, Link A, Walenta K, Bohm M, Nickenig G. Intravenous transfusion of endothelial progenitor cells reduces neointima formation after vascular injury. Circ Res. 2003; 93: e17–e24.
Majka SM, Jackson KA, Kienstra KA, Majesky MW, Goodell MA, Hirschi KK. Distinct progenitor populations in skeletal muscle are bone marrow derived and exhibit different cell fates during vascular regeneration. J Clin Invest. 2003; 111: 71–79.
Li G, Chen SJ, Oparil S, Chen YF, Thompson JA. Direct in vivo evidence demonstrating neointimal migration of adventitial fibroblasts after balloon injury of rat carotid arteries. Circulation. 2000; 101: 1362–1365.
Bayes-Genis A, Campbell JH, Carlson PJ, Holmes DR, Jr., Schwartz RS. Macrophages, myofibroblasts and neointimal hyperplasia after coronary artery injury and repair. Atherosclerosis. 2002; 163: 89–98.
Benditt EP, Benditt JM. Evidence for a monoclonal origin of human atherosclerotic plaques. Proc Natl Acad Sci U S A. 1973; 70: 1753–1756.
Pearson TA, Dillman JM, Solex K, Heptinstall RH. Clonal markers in the study of the origin and growth of human atherosclerotic lesions. Circ Res. 1978; 43: 10–18.
Ehsan A, Mann MJ, Dell’Acqua G, Dzau VJ. Long-term stabilization of vein graft wall architecture and prolonged resistance to experimental atherosclerosis after E2F decoy oligonucleotide gene therapy. J Thorac Cardiovasc Surg. 2001; 121: 714–722.
Faries PL, Marin ML, Veith FJ, Ramirez JA, Suggs WD, Parsons RE, Sanchez LA, Lyon RT. Immunolocalization and temporal distribution of cytokine expression during the development of vein graft intimal hyperplasia in an experimental model. J Vasc Surg. 1996; 24: 463–471.(Lisheng Zhang; Neil J. Fr)
Correspondence to Karsten Peppel, Box 3187, Duke University Medical Center, 447 CARL Building, Science Drive, Durham, NC 27710. E-mail karsten.peppel@duke.edu
Abstract
Objective— Vein graft disease involves neointimal smooth muscle cells, the origins of which are unclear. This study sought to characterize and quantitate vein graft infiltration by cells extrinsic to the graft in a mouse model of vein graft disease.
Methods and Results— Inferior vena cava-to-carotid artery interposition grafting between C57Bl/6 and congenic ?-galactosidase–expressing ROSA26 mice was performed. Vein grafts were harvested 6 weeks postoperatively and stained with X-gal. More than 60% of neointimal cells derived from the recipient, and 50% of these cells expressed smooth muscle -actin. The distribution of donor and recipient-derived cells within this vein graft wall layer was distinctly focal, consistent with focal infiltration and expansion of progenitor cells. When bone marrow transplantation with congenic green fluorescent protein (GFP)-expressing cells was used in vein graft recipients 1 month before surgery, abundant GFP-expressing cells appeared in the media, but not the neointima, of mature grafts. Endothelial cells in mature grafts derived from graft-intrinsic and graft-extrinsic sources and were, in part, of bone marrow origin.
Conclusions— Cells extrinsic to the graft, including bone marrow-derived cells, predominate during vein graft remodeling.
Key Words: smooth muscle cells ? vein graft neointimal hyperplasia ? mouse models ? bone marrow transplantation ? endothelium
Introduction
Autologous vein grafting ranks among the most common of surgical procedures, but its long-term success remains compromised by vein graft disease.1 Up to 30% of grafts become stenotic and require intervention within 2 years because of hemodynamically significant neointimal hyperplasia.2 Ten years after surgery, only approximately 50% of venous grafts are still patent, and only half of these are free of significant stenosis.1,3–5 Vein graft failure in humans has been intimately linked to atherosclerosis-predisposing neointimal hyperplasia, involving smooth muscle cells (SMCs) and extracellular matrix.1,3
The origin of neointimal SMCs remains incompletely characterized, despite its therapeutic implications. SMCs proliferating in the vein graft media have been shown to migrate into the neointima.6,7 In addition, arterial myofibroblasts contribute to neointimal hyperplasia in a pig endoluminal coronary artery injury model,8 and bone marrow (BM)-derived cells contribute neointimal SMCs in models of atherosclerosis9 or transplant arteriopathy.10,11 To date, however, there has been no quantitation of vein graft neointimal cell precursors.
To address this issue, we recently developed a unique murine model of vein grafting that uses sutured arteriovenous anastomoses, and neointimal hyperplasia mimics early human vein graft disease by reaching steady-state thickness without engendering significant luminal stenosis.12 In this model, we used ?-galactosidase–expressing ROSA26 mice as vein graft donors and/or recipients to quantitate the extent to which the mature vein graft wall and neointima comprise cells derived from precursors residing either within or outside of the vein graft at the time of its implantation. To further differentiate the origins of graft-extrinsic cells, we replaced the bone marrow of vein graft recipient mice with that of green fluorescent protein (GFP)-expressing animals.
Methods
Animals
Adult male C57BL/6 (wild-type [WT]) mice, and congenic mice that either express Lac Z (?gal, ROSA26 strain) or GFP, were purchased from Jackson Labs. All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee and complied with the Guide for the Care and Use of Laboratory Animals.
Vein Graft Surgery
Interposition vein graft surgery was performed as described previously.12 Inferior vena cavae (IVC) from C57Bl/6 or ROSA26 donor mice were anastomosed to the right common carotid artery of recipient mice, yielding four different donor/graft recipient groups, as follows: (1) WT/WT; (2) WT/?gal; (3) ?gal/WT; and (4) ?gal/?gal. All mice were between 12 and 20 weeks old.
IVC Graft Harvest and Analysis
Grafts were harvested 2 or 6 weeks postoperatively and prepared for histochemical and immunofluorescent analysis as described previously.12 For analysis of ?-galactosidase expression, the grafts were excised without fixation and incubated at 37°C (16 hours) in x-gal staining solution (10 mmol/L K4Fe(CN)6, 10 mmol/L K3Fe(CN)6, 2 mmol/L MgCl2 and 500 μg/mL x-gal (5-bromo-4-chloro-3-indoyl-?-D-galactopyranoside) in phosphate-buffered saline (PBS). Grafts were subsequently fixed in 10% formalin for 24 hours and photographed en bloc using RT SPOT 2 CCD camera (Diagnostic Instruments). Grafts were then embedded in optimal cutting temperature (OCT) compound to facilitate cryosectioning for light and immunofluorescence microscopy.
Bone Marrow Transplantation
C57Bl/6 mice received 950 rads of gamma irradiation at age 14 weeks and were infused with 107 BM cells procured from femurs of GFP mice. Successful BM engraftment was confirmed 3 weeks later by examining blood smears. Mice thus made chimeric for BM GFP expression received WT IVC grafts 4 weeks after bone marrow transplantation (BMT). Grafts were harvested 6 weeks after operation, perfused briefly with saline, and frozen in OCT compound.
Vein Graft Microscopic Analysis
Immunofluorescence and immunoperoxidase staining and microscopy were performed as previously described.12,13 T cells and macrophages were identified with rabbit anti-CD3- antibody (DAKO) or BM8 (Research Diagnostic), respectively. Endothelial cells were identified with rabbit anti-factor VIII (DAKO). Grafts from GFP BM chimeric mice were OCT-embedded, sectioned at 5 to 10 μm, and imaged for GFP fluorescence. Immediately thereafter, sections were processed for immunofluorescence. Positive controls for CD3 and macrophage staining were murine spleen and raw 264.7 macrophages, respectively. Negative controls were incubated with cognate non-immune IgG in lieu of primary antibody. The nuclear DNA dye Hoechst 33342 (10 μg/mL) was added to secondary antibody incubations. Sequential images of single microscopic fields were taken as described.13 In overlays of fluorescent and bright-field images, the black background of fluorescence images was removed to facilitate identification of non-SMCs in the vein grafts.
Quantitation of Graft Donor-Derived and Recipient-Derived Cells
Vein graft sections from 5 graft regions were each imaged at 4 to 8 different clock hours, at x440 magnification. Five different locations taken over the length of the graft were analyzed for each specimen. Both WT IVC grafts placed into ROSA26 recipients (WT/?gal) and ROSA26 IVC grafts placed into WT recipients (?gal/WT) were analyzed. Consequently, percent recipient-derived cells was calculated in either of 2 ways: (1) 100 x (number of x-gal–stained cells)÷(number of total [ie, Hoechst-stained] cells) for ROSA26/WT specimens; and (2) 100 x [1 - (number of x-gal–stained cells)÷(number of total cells)] for WT/ROSA26 specimens. More than 100 cells were counted per x440 field. Cells were counted only within the vein graft neointima.
Results
To address whether vein graft SMCs derive from progenitors extrinsic to the vein graft, we placed IVC segments from ROSA26 (?gal) or C57Bl/6 (WT) donor mice as interposition grafts into the right common carotid artery of ?gal or WT recipient mice (Figure 1). The ROSA26 strain is congenic to C57Bl/6, and thus transplantation of vascular grafts between these two strains does not engender graft rejection and confounding transplant arteriopathy. Therefore, this model system is well suited to assess infiltration of vein grafts by graft-extrinsic, recipient-derived cells. Grafts were harvested 6 weeks postoperatively, at a time when vein graft wall thickness and neointimal hyperplasia have reached steady-state in this model.12 To determine the relative contribution of ?gal-expressing cells to the vein graft wall, intact vein grafts were stained en bloc for ?gal expression (upper photographs, Figure 1A through 1D). WT IVCs grafted into WT recipient mice did not display any blue staining, either in whole explants (Figure 1A, bottom) or in cross-sections (Figure 1A, top), demonstrating the absence of activated macrophages in these specimens.14 All vein grafts involving ?gal mice as either donors or recipients showed some degree of blue staining, which traversed the entire length of the graft (Figure 1B through 1D, arrowheads), and varied by specimen group (Figure 1B through 1D, lower photos). Blue cellular staining evident in the WT IVC-to-?gal recipient specimens (Figure 1C) clearly demonstrated that vein graft wall cells may derive from sources that are extrinsic to the vein graft at the time of implantation. Similarly, the prevalence of blue cellular staining of the ?gal IVC-to-WT recipient specimens (Figure 1B) was substantially less than that obtained in ?gal IVC-to-?gal recipient specimens (Figure 1D). Thus, the contributions of graft recipient (ie, graft-extrinsic) cells to the vein graft wall were observed in WT IVC-to-?gal and ?gal IVC-to-WT specimens.
Figure 1. Heterogeneous origin of vein graft wall cells. Interposition vein grafting was performed with the indicated donor and recipient mice, and 6-week-old vein grafts were harvested and stained with x-gal. Explanted specimens were imaged before embedding in OCT compound (upper images in each panel) and after sectioning and counterstaining with eosin (lower image in each panel). The explanted specimens comprise (from left to right) the recipient’s right common carotid artery caudal to the bifurcation, the vein graft segment (between the arrowheads), the proximal common carotid, and the aortic arch with remnants of the left common carotid. Representative explant specimens are shown (n=3 for WT/WT and ?gal/?gal; 5 for WT/?gal and ?gal/WT). Scale bar in (A) is 5 mm. Photomicrographs (original magnification x440) are representative of 20 per vein graft. Scale bar is 100 μm.
In patients, the developing vein graft neointima predominately comprises cells expressing smooth muscle -actin and represents an atherosclerosis prone area that forms the foundation for the later development of graft atheroma.1 Similarly, smooth muscle -actin–expressing cells predominate in the neointima of murine grafts (Figure 2). Indeed, smooth muscle -actin–expressing cells reside almost exclusively in the collagen-poor, cytoplasm-rich neointima of mouse vein isografts (Figure 2C) and rabbit jugular vein autografts, which have long been used to model human vein graft disease (data not shown).3,15 To quantitate the percentage of graft-intrinsic and graft-extrinsic cells in the neointima of these murine vein grafts, we superimposed x-gal stained bright-field and red fluorescence SM -actin–stained images of identical WT-to-?gal vein graft cross-sections (Figure 2G). Total cellularity of the sections was assessed by quantitating nuclei after incubation with Hoechst 33342 (Figure 2D). Analysis of ?gal-to-WT and WT-to-?gal grafts revealed that approximately 60% of cells in the vein graft neointima develop from graft-extrinsic precursors (Table). Moreover, smooth muscle -actin expression was evident in 66% of graft donor-recipient and 56% of graft recipient-derived neointimal cells (Table). Thus, graft-intrinsic and graft-extrinsic precursors give rise predominantly to SMCs of the mature vein graft neointima. There was only scant evidence of inflammatory cell infiltration at 6 weeks (Table). ?gal IVC-to-WT grafts were analyzed in parallel and yielded congruent results.
Figure 2. Prevalence of SMCs among cells derived from vein graft-extrinsic progenitors. A, Vein graft paraffin section was stained with a modified connective tissue stain (see Methods) and imaged at 110x original magnification. B, C, The area boxed in (A) was imaged at 440x original magnification, and a serial section was stained with Hoechst 33342 and Cy3-conjugated anti-SM -actin. The fluorescence and light images are aligned, with their luminal surfaces upward. Images are representative of 7 different vein graft samples. D–G, Six-week-old WT IVC-to-?gal and ?gal IVC-to-WT specimens were processed as in Figure 1, with the omission of eosin counterstain. Each specimen was then imaged sequentially under fluorescence (D, F) and bright-field conditions (E) to identify nuclei (D), SM -actin–positive cells (F) and ?gal-positive cells (E) within the same microscopic field. G, Images in (E) and (F) were digitally merged, as described in the Methods section, to allow identification of SMCs (red staining cells) that are graft extrinsic (blue) or graft intrinsic (non-blue). Arrow indicates graft intrinsic SMCs; arrowhead, graft extrinsic SMCs; open arrow, graft extrinsic non-SMCs. Shown is a single microscopic field from one WT IVC-to-?gal specimen, representative of 4 samples analyzed from n 5 mice in each group. Original magnification x440; scale bar in (G) 100 μm. H, The percent graft-extrinsic cells was calculated as in Methods for WT IVC-to-?gal and ?gal IVC-to-WT specimens (n=5 for each group) at 1-mm intervals across the length of the grafts, as indicated. Plotted at each location are the mean±SE. Scale bar is 200 μm (A) and 100 μm (B, G).
Characterization of Vein Graft Neointimal Cells
Multiple possibilities exist for the origin of these prevalent graft-extrinsic precursor cells. It is conceivable that some of the graft-infiltrating cells derived by inward migration from the media of the adjacent carotid artery. To examine this possibility, we quantitated the percentage of graft-intrinsic and graft-extrinsic cells over the entire length of the vein graft (Figure 2H). If vascular cell migration into the graft from the anastomoses were a predominant source of graft-extrinsic precursor cells, then we would expect to see a higher prevalence of recipient cells in graft sections near the anastomoses, as compared with sections obtained from the center of the grafts (because graft-extrinsic cells constituted only 60% of total graft wall cells). The center of the vein grafts (segment 3) showed only 20% to 25% fewer recipient-derived cells than the peri-anastomotic portions of the graft (segments 1 and 5). Thus, although not excluding a contributing role of lateral-migrating arterial SMCs to the formation of vein graft lesions, this analysis does not support the notion of a predominance of anastomotic vascular cells in the population of cells originating outside of the graft.
To address the origin of graft-extrinsic cells further, we examined serial sections of 2-week-old vein grafts, in which the media and neointima are still enlarging.12 (Indeed, steady-state vein graft wall thickness is reached only by 4 weeks postoperatively in this animal model.) In these 2-week-old specimens, isolated foci of cells are evident in many sections—even in sections remote from the anastomoses (Figure 3A). These foci show numerous actively proliferating cells, as assessed by staining for proliferating cell nuclear antigen (PCNA) (Figure 3C, red arrow). This cell proliferation distinguishes vein graft proliferative foci from adjacent cells that constitute part of postsurgical adhesions (Figure 3, yellow arrow). Staining for smooth muscle -actin reveals that these foci are composed of SMCs and non-SMCs. These focal, cellular expansions of the graft wall could, of course, originate from either graft-extrinsic or graft-intrinsic cells. If these foci represent expansions of recipient-derived cells, however (please see later), then this finding suggests that graft-extrinsic precursor cells do infiltrate the graft and originate from sources distinct from the peri-anastomotic carotid artery—either the circulating blood or adventitial adhesions. Unfortunately, x-gal staining of these foci was inadequate to determine whether their progenitor cells originated in the graft donor or recipient—perhaps because insufficient ?gal is present in these cells undergoing relatively rapid mitosis. Therefore, to examine the origin of the focal cellular expansions observed in these immature grafts, we returned again to an analysis of mature, 6-week-old vein grafts in which remodeling has ceased.12 We reasoned that if focal cellular expansions in the arterializing vein graft enlarge to constitute the mature graft wall, then these mature vein grafts should evince focal staining for donor-derived and recipient-derived cells. This expectation was fulfilled in our 6-week-old vein grafts stained with x-gal (Figure 3D). Quantitatively, graft-extrinsic neointimal foci demonstrated a 60% relative increase in graft-extrinsic cells, compared with the neointima taken as a whole (Figure 3E). Thus, focal cellular vein graft wall expansions observed in 2-week-old specimens almost certainly did develop from both graft-intrinsic and graft-extrinsic precursor. The focal cellular nodules (remote from anastomoses) in 2-week-old vein grafts and focal expansions of graft-intrinsic and graft-extrinsic cell populations in mature vein grafts together support a model in which graft-extrinsic cells originate from precursors entering the graft via a transendothelial and/or transadventitial route in addition to cells entering the vein graft via migration from the anastomoses. What is the derivation of the graft extrinsic cells that do not originate from the adjacent carotid artery? Recent interest and controversy has focused on the role of BM-derived cells in vascular repair.9,11,16–18 To examine the potential contribution of BM-derived circulating precursor cells to vein graft remodeling, we performed vein graft surgeries in C57Bl/6 WT recipient mice that had their BM replaced with that of congenic GFP-expressing mice. Six-week-old mature grafts harvested from these chimeric mice showed prominent, albeit inhomogeneous, infiltration of BM-derived cells into the media and adventitia, but notably not the neointima, of the grafts (Figure 4C). These BM-derived cells comprised primarily cells of the macrophage lineage and did not express SMC -actin (Figure 4D).
Figure 3. Development of vein graft neointima involves focal cellular expansions of both graft-intrinsic and graft-extrinsic precursor cells. A–C, Two-week-old vein grafts (n=3) were formalin-fixed, paraffin-embedded, sectioned at 200-μm intervals, and stained either with a modified connective tissue stain ("Masson") or with immunofluorescence for SM -actin and PCNA, along with Hoechst 33342. A, Isolated areas of focal cellular infiltration can be seen (red arrow) in otherwise uniformly thin vein graft walls (arrowheads). Adjacent graft-extrinsic tissue is indicated by the yellow arrow. Serial sections 500 μm proximal and distal to this section demonstrated no focal proliferation of cells (not shown). The boxed area is enlarged in (B) and (C). B, The focal expansion contains SMC -actin–expressing and -actin–nonexpressing cells. A serial section of (A) is shown, with Hoechst and Cy3-anti-SM -actin images superimposed. C, The same section shown in (B) is portrayed, imaged only for anti-PCNA immunofluorescence. D, Six-week-old vein grafts, with the indicated donors and recipients, were stained, sliced, and counterstained as described for Figure 1. Foci of x-gal–stained neointimal cells have been circled, demonstrating progeny of graft-extrinsic and graft-intrinsic cells, as indicated. E, In foci like those identified in (D), the prevalence of cells derived from graft-extrinsic precursors and the prevalence of SMCs were calculated as in Figure 2 and plotted as mean±SE from 3 foci from each of 5 specimens. Representative specimens (n 5 for each graft type) are shown. Original magnification x110 (A, D) and x440 (B, C); scale bar is 200 μm (A, D).
Figure 4. Bone marrow derived cells participate in vein graft remodeling. GFP BM chimeric mice were generated as described in Methods and served as recipients of WT vein grafts. Grafts were harvested after 6 weeks and embedded in OCT. Serial sections were stained with H&E (A) or imaged for GFP fluorescence. Specimens imaged for GFP were subsequently processed for either SM -actin expression (B) or macrophage presence (D). C, E, The GFP fluorescent and corresponding immunofluorescent images were merged. Neointimal (arrowheads) and medial plus adventitial (open arrows) boundaries are shown.
Vein graft remodeling is initiated, in part, by hemodynamic and ischemic injury to the endothelium of the graft. Re-endothelialization is thought to play an important part in limiting vascular remodeling,1 We therefore sought to determine the origin of the endothelial cells that participate in vein graft remodeling. In WT IVC grafts that were placed into ?gal recipients and stained with x-gal, we found evidence of graft-intrinsic (non-stained) and graft-extrinsic (blue-stained) endothelial cells (Figure 5 A and 5B). In addition, when we examined WT grafts that were placed into GFP BM chimeric recipients (Figure 5C through 5F), we found a small prevalence (10%) of GFP-positive cells among the vein graft endothelial cells. This indicates that re-endothelialization in this model derives from multiple sources, including cells of the graft donor and endothelial progenitor cells that originate from BM intrinsic and extrinsic compartments of the graft recipient. These findings thus contrast with those of Xu et al, who recently used Tie2-driven ?gal expression to show that, in their vein grafts, endothelium is essentially replaced by recipient cells.19 Hemodynamic differences between vein graft model systems may underlie this discrepancy in endothelial data.
Figure 5. Vein graft endothelium derives from graft-intrinsic and graft-extrinsic cells, including bone marrow precursors. Isogenic vein grafting was performed using IVCs from WT mice placed into recipients that were either (A, B) ROSA26 or chimeric WT mice previously transplanted with whole BM from GFP-expressing mice (C–F). Grafts were harvested 6 weeks postoperatively. Serial sections of grafts placed in ROSA26 mice were stained either with x-gal to identify ?gal-expressing cells (A) or with anti-factor VIII antibody (immunoperoxidase) to identify endothelial cells (B). Endothelial cells expressing ?gal (closed arrows) or not (open arrows) are indicated. (C–F). Grafts placed into WT mice with GFP BM were frozen in OCT compound, sliced at 5 μm, and then imaged for GFP fluorescence (C, D). Subsequently, these sections were subjected to immunoperoxidase staining for factor VIII to identify endothelial cells (E, F). GFP fluorescence identifies BM-derived cells that contribute to endothelial layer formation (D, F) or that do not (C, E). The lumen of each specimen is oriented upward. Original magnification x1100 (A, B) or x440 (C–F).
Discussion
We have shown for the first time to our knowledge that even in the absence of atherosclerosis, the majority of vein graft neointimal and medial cells derive from precursors that are extrinsic to the vein at the time of grafting. We have also provided the first evidence that the arterializing vein graft wall develops largely from focal expansions of progenitor cells, which originate within and outside of the vein graft at the time of implantation. Finally, we have shown that endothelial remodeling of vein grafts is accomplished by the expansion of graft-intrinsic as well as graft-extrinsic cells, and that BM-derived cells participate in this process. Importantly, these novel insights into vein graft biology derive from a murine vein graft system that we have shown to model human vein grafts, with regard to their surgical anastomoses as well as the distribution, composition and extent of neointimal hyperplasia within the grafts.12
Sources of graft-extrinsic neointimal cells in our murine model remain to be determined, but potential sources include migrating vascular cells from the adjacent carotid artery,20 as well as progenitor cells infiltrating the graft through either circulating blood (vasa vasora21 or transendothelial diapedesis22) or postsurgical adventitial adhesions. Surgical trauma at the anastomoses activates SMCs of the adjacent artery to proliferate and ultimately migrate into vein grafts—a process clearly delineated in rat iliac vein grafts.20 Nevertheless, we can infer that activated SMCs of the adjacent artery do not account for all of the graft-extrinsic cells in the media and neointima of our murine model, from two complementary sources of data: (1) 2-week-old vein grafts demonstrated focal expansions of smooth muscle -actin expressing cells remote from the anastomoses (Figure 3); and (2) our lengthwise analysis of 6-week-old grafts found no significant decline in the prevalence of vein graft recipient cells as a function of distance from the graft anastomoses (Figure 2H). The relative contribution of adjacent arterial SMCs to vein graft neointima doubtless depends heavily on the graft system used.
Data from our GFP-chimeric vein graft recipients clearly demonstrate that BM-derived cells contribute to the genesis of the vein graft wall—principally in the media, adventitia, and (to a much smaller degree) the endothelium. However, the non-endothelial progeny of these BM-derived cells appeared to be largely macrophages and did not express SM -actin. It is possible that these cells represented fibrocytes, which had not yet differentiated into a SM -actin–expressing phenotype.23 Nevertheless, the absence of SM -actin expression in these cells may explain why Hu et al found no BM contribution to their mouse vein graft SMC populations.24 These authors replaced the BM of their vein graft recipient mice with marrow from SM22-lacZ transgenic mice, which express ?gal only in SMC-like cells. Neointimal SMCs have been shown to derive from BM in atherosclerosis;9 however, different combinations of growth factors present in atheromata may influence the differentiation of BM-derived precursor cells.23,24
The finding that vein graft neointima formation proceeds without bone marrow contribution to a large part from graft extrinsic cells distinguishes vein graft remodeling from other forms of arterial remodeling. In arterial remodeling associated with transplant arteriopathy11,25,26 and mechanical injury,18 BM-derived cells contribute to neointimal hyperplasia. Vascular progenitor cells have been isolated from other organs as well.27,28 Majka et al demonstrated the potential for skeletal muscle-derived progenitor cells to differentiate into endothelial cells as well as SMCs.28 This study also demonstrated the slow replacement of skeletal progenitor cells by cells derived from the bone marrow. Thus, our data support the notion that vascular injury elicits divergent repair responses, dependent on the type of injury inflicted.
Approximately 30% of our vein graft neointimal cells did not express SM -actin and may comprise fibroblasts or fibrocytes.23 Adventitial fibroblasts have been reported to migrate into the neointima of balloon-injured rat and pig arteries, and ultimately transdifferentiate into myofibroblasts that express SM -actin.8,29,30 Because adventitial fibroblasts in our model could originate from the vein graft adventitia itself or the peri-graft (postsurgical) adhesions, these fibroblasts could derive from either graft-intrinsic or graft-extrinsic cells. Moreover, graft-extrinsic fibroblasts originating from surgical adhesions need not originate from the BM—consistent with the absence of neointimal GFP fluorescence in grafts from BM-chimeric mice (Figure 4).
The novel finding of focal proliferation in arterializing vein grafts (Figure 3) recalls data regarding the clonal origins of fibrous plaques in atherosclerosis.31,32 Proliferative cellular foci in vein grafts were demonstrable by congruent data from 2- and 6-week-old vein grafts (Figure 3), which evince foci originating from either graft-intrinsic or graft-extrinsic precursor cells. Whether these distinct foci of cells represent clonal, or rather colonial, cell growth currently transcends the scope of this investigation.
Because most vein graft neointimal cells derive from precursors extrinsic to the vein graft, we can make novel mechanistic inferences about successful therapies for vein graft neointimal hyperplasia, like the E2F decoy.33 Successful therapies must, of course, target proliferation, migration, and extracellular matrix secretion by cells intrinsic to the implanted vein graft, because these cells ultimately constitute 40% of neointimal cells. In addition, however, these therapies must target recruitment to the graft of cells initially extrinsic to the graft, perhaps by reducing the inflammatory reaction associated with vein graft endothelial damage22 and/or the chemokine/growth factor secretion by vein graft cells.34
Acknowledgments
This work was supported by grants HL63288 (to N.J.F.) and HL64744 (to K.P.) and American Heart Association grants-in-aid (to N.J.F., K.P.).
References
Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation. 1998; 97: 916–931.
Mills JL, Fujitani RM, Taylor SM. The characteristics and anatomic distribution of lesions that cause reversed vein graft failure: a five-year prospective study. J Vasc Surg. 1993; 17: 195–204.
Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br J Surg. 1994; 81: 1254–1269.
Nwasokwa ON. Coronary artery bypass graft disease. Ann Intern Med. 1995; 123: 528–545.
Mehta D, Izzat MB, Bryan AJ, Angelini GD. Towards the prevention of vein graft failure. Int J Cardiol. 1997; 62: S55–S63.
Hart CE, Kraiss LW, Vergel S, Gilbertson D, Kenagy R, Kirkman T, Crandall DL, Tickle S, Finney H, Yarranton G, Clowes AW. PDGFbeta receptor blockade inhibits intimal hyperplasia in the baboon. Circulation. 1999; 99: 564–569.
Rectenwald JE, Moldawer LL, Huber TS, Seeger JM, Ozaki CK. Direct evidence for cytokine involvement in neointimal hyperplasia. Circulation. 2000; 102: 1697–1702.
Shi Y, O’Brien JE, Fard A, Mannion JD, Wang D, Zalewski A. Adventitial myofibroblasts contribute to neointimal formation in injured porcine coronary arteries. Circulation. 1996; 94: 1655–1664.
Sata M, Saiura A, Kunisato A, Tojo A, Okada S, Tokuhisa T, Hirai H, Makuuchi M, Hirata Y, Nagai R. Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis. Nat Med. 2002; 8: 403–409.
Han CI, Campbell GR, Campbell JH. Circulating bone marrow cells can contribute to neointimal formation. J Vasc Res. 2001; 38: 113–119.
Shimizu K, Sugiyama S, Aikawa M, Fukumoto Y, Rabkin E, Libby P, Mitchell RN. Host bone-marrow cells are a source of donor intimal smooth- muscle-like cells in murine aortic transplant arteriopathy. Nat Med. 2001; 7: 738–741.
Zhang L, Hagen PO, Kisslo J, Peppel K, Freedman NJ. Neointimal hyperplasia rapidly reaches steady state in a novel murine vein graft model. J Vasc Surg. 2002; 36: 824–832.
Peppel K, Jacobson A, Huang X, Murray JP, Oppermann M, Freedman NJ. Overexpression of G protein-coupled receptor kinase-2 in smooth muscle cells attenuates mitogenic signaling via G protein-coupled and platelet-derived growth factor receptors. Circulation. 2000; 102: 793–799.
Dannenberg A, Burstone M, Walter P, Kinsley J. A histochemical study of phagocytic and enzymatic functions of rabbit mononuclear and polymorphonuclear exudate cells and alveolar macrophages. I. Survey and quantitation of enzymes and states of cellular activation. J. Cell Biol. 1963; 17: 465–486.
Peppel K, Zhang L, Huynh TT, Huang X, Jacobson A, Brian L, Exum ST, Hagen PO, Freedman NJ. Overexpression of G protein-coupled receptor kinase-2 in smooth muscle cells reduces neointimal hyperplasia. J Mol Cell Cardiol. 2002; 34: 1399–1409.
Sata M, Tanaka K, Nagai R. Origin of smooth muscle progenitor cells: different conclusions from different models. Circulation. 2003; 107: e106–e107.
Hu Y, Davison F, Ludewig B, Erdel M, Mayr M, Url M, Dietrich H, Xu Q. Smooth muscle cells in transplant atherosclerotic lesions are originated from recipients, but not bone marrow progenitor cells. Circulation. 2002; 106: 1834–1839.
Tanaka K, Sata M, Hirata Y, Nagai R. Diverse Contribution of Bone Marrow Cells to Neointimal Hyperplasia After Mechanical Vascular Injuries. Circ Res. 2003; 93: 783–790.
Xu Q, Zhang Z, Davison F, Hu Y. Circulating progenitor cells regenerate endothelium of vein graft atherosclerosis, which is diminished in ApoE-deficient mice. Circ Res. 2003; 93: e76–e86.
Dilley RJ, McGeachie JK, Tennant M. The role of cell proliferation and migration in the development of a neo-intimal layer in veins grafted into arteries, in rats. Cell Tissue Res. 1992; 269: 281–287.
Davies MG, Hagen PO. Influence of perioperative storage solutions on long-term vein graft function and morphology. Ann Vasc Surg. 1994; 8: 150–157.
Westerband A, Mills JL, Marek JM, Heimark RL, Hunter GC, Williams SK. Immunocytochemical determination of cell type and proliferation rate in human vein graft stenoses. J Vasc Surg. 1997; 25: 64–73.
Abe R, Donnelly SC, Peng T, Bucala R, Metz CN. Peripheral blood fibrocytes: differentiation pathway and migration to wound sites. J Immunol. 2001; 166: 7556–7562.
Hu Y, Mayr M, Metzler B, Erdel M, Davison F, Xu Q. Both donor and recipient origins of smooth muscle cells in vein graft atherosclerotic lesions. Circ Res. 2002; 91: e13–e20.
Hillebrands JL, Klatter FA, Rozing J. Origin of vascular smooth muscle cells and the role of circulating stem cells in transplant arteriosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23: 380–387.
Li J, Han X, Jiang J, Zhong R, Williams GM, Pickering JG, Chow LH. Vascular smooth muscle cells of recipient origin mediate intimal expansion after aortic allotransplantation in mice. Am J Pathol. 2001; 158: 1943–1947.
Werner N, Junk S, Laufs U, Link A, Walenta K, Bohm M, Nickenig G. Intravenous transfusion of endothelial progenitor cells reduces neointima formation after vascular injury. Circ Res. 2003; 93: e17–e24.
Majka SM, Jackson KA, Kienstra KA, Majesky MW, Goodell MA, Hirschi KK. Distinct progenitor populations in skeletal muscle are bone marrow derived and exhibit different cell fates during vascular regeneration. J Clin Invest. 2003; 111: 71–79.
Li G, Chen SJ, Oparil S, Chen YF, Thompson JA. Direct in vivo evidence demonstrating neointimal migration of adventitial fibroblasts after balloon injury of rat carotid arteries. Circulation. 2000; 101: 1362–1365.
Bayes-Genis A, Campbell JH, Carlson PJ, Holmes DR, Jr., Schwartz RS. Macrophages, myofibroblasts and neointimal hyperplasia after coronary artery injury and repair. Atherosclerosis. 2002; 163: 89–98.
Benditt EP, Benditt JM. Evidence for a monoclonal origin of human atherosclerotic plaques. Proc Natl Acad Sci U S A. 1973; 70: 1753–1756.
Pearson TA, Dillman JM, Solex K, Heptinstall RH. Clonal markers in the study of the origin and growth of human atherosclerotic lesions. Circ Res. 1978; 43: 10–18.
Ehsan A, Mann MJ, Dell’Acqua G, Dzau VJ. Long-term stabilization of vein graft wall architecture and prolonged resistance to experimental atherosclerosis after E2F decoy oligonucleotide gene therapy. J Thorac Cardiovasc Surg. 2001; 121: 714–722.
Faries PL, Marin ML, Veith FJ, Ramirez JA, Suggs WD, Parsons RE, Sanchez LA, Lyon RT. Immunolocalization and temporal distribution of cytokine expression during the development of vein graft intimal hyperplasia in an experimental model. J Vasc Surg. 1996; 24: 463–471.(Lisheng Zhang; Neil J. Fr)