Every allograft needs a silver lining

Abstract

The development of chronic allograft rejection is based on the hypothesis that cumulative, time-dependent tissue injury eventually leads to a fibrotic response. In this issue of the JCI, Babu and colleagues found that alloimmune-mediated microvascular loss precedes tissue damage in murine orthotopic tracheal allografts (see the related article beginning on page 3774). The concept that injury to the endothelium may precede airway fibrosis suggests that interventions to maintain vascular integrity may be important, especially in the case of lung transplantation. Further, for all solid organ allografts, it is possible that the key to long-term allograft survival is physiological vascular repair at early times following transplantation.

Clues to mechanisms underlying long-term allograft survival

The major obstacle to the long-term survival of lung transplant recipients is the development of the bronchiolitis obliterans syndrome (BOS), which typically occurs in up to 60% of patients who survive five years (1). BOS is considered a major posttransplant complication and is characterized by a progressive luminal airway narrowing and histological evidence of fibrosis. Pathological features of BOS suggest that injury and inflammation of epithelial cells and subepithelial structures within small airways lead to the fibrosis response. Moreover, it is thought that repetitive injury in lung allografts, as in other solid organ transplants (e.g., kidney) (2, 3), may elicit cumulative, time-dependent damage and result in cellular atrophy and chronic interstitial fibrosis (1). According to this paradigm, prevention or inhibition of injury (using time-dependent therapies) should result in long-term graft survival (4).

Nevertheless, for all solid organ allografts, an underappreciated aspect of the injury process is its association with repair. It has been proposed that augmenting physiological repair, even in the face of injury, will ensure that the end result of injury is limited (5, 6). On the other hand, if the repair process is inhibited, then even milder forms of injury have the potential to result in extensive tissue damage. Therefore, it is possible that an understanding of repair is key to long-term allograft survival.

In this issue of the JCI, Babu et al. set out to address whether the maintenance of microvasculature integrity is sufficient to sustain lung allograft function and to prevent the development of fibrosis/chronic rejection (7). These authors used a murine orthotopic tracheal transplantation model and found that alloimmune-mediated microvascular loss results in hypoxia, which precedes tissue damage and the development of intragraft fibrosis in the recipient trachea. If endothelial cell injury was inhibited (by limiting ongoing inflammation) and the microvasculature was maintained intact, physiological remodeling occurred and allograft tissue morphology in these animals returned to normal (Figure 1). Therefore, in general, once an allograft is revascularized and is functioning, protecting the vasculature and/or enabling physiological homeostatic repair of the microvasculature will prevent tissue fibrosis.

Figure 1. Microvascular injury and repair following tracheal transplantation.

Following orthotopic transplantation of trachea, revascularization occurs within four days and is associated with perfusion and return of blood flow. This revascularization response involves efficient physiological anastomoses between donor and recipient microvessels. This early efficient repair and/or homeostatic angiogenesis is required for normal graft function. After reperfusion, in the absence of inflammation (such as occurs in syngrafts or immunosuppressed recipients of allografts), physiological homeostatic vascular remodeling occurs, microvascular integrity is maintained, and tissue morphology remains normal. In contrast, as leukocytes infiltrate allografts in association with rejection, pathophysiological inflammatory angiogenesis occurs and is only sufficient to sustain graft function minimally. This inflammatory angiogenesis reaction likely facilitates ongoing leukocyte recruitment and endothelial damage, eventually leading to ischemia and a cycle of microvascular injury and hypoxia that cannot sustain tissue function. The end result of this cycle is tissue fibrosis and chronic rejection. In this issue of the JCI, Babu et al. (7) demonstrated that early physiological homeostatic repair and the absence of inflammation will sustain long-term tissue function and morphology.

Role of microvascular repair in the preservation of allografts

In the context of transplantation, it should be emphasized that microvascular endothelial cells are very susceptible to injury, including changes in oxygen tension (hypoxic injury), reperfusion injury, and oxidative stress, as well as persistent episodes of silent rejection (8). These repetitive insults clearly target the microvasculature and create a circumstance whereby endothelial loss and/or damage will typically occur (9, 10). Within allografts, the loss of the microvasculature results in impaired delivery of oxygen and nutrients to key organ-specific cells, such as columnar epithelial cells in the lung or tubular epithelial cells in the kidney, which in turn results in chronic ischemia and cell death (5, 6, 8, 11). If vascular integrity is coincidently compromised, recovery may not be complete and/or may not occur, and cellular atrophy will lead to progressive functional tissue loss. The implication of these observations is that the limiting factor in the allograft injury process is inefficient microvascular repair. This concept may be particularly relevant following lung transplantation, where bronchial artery vascular reanastomosis is often not done, and thus vascular integrity may be easily compromised at early times as a result of alloimmune attack. Indeed, microvascular injury has been documented in association with rejection (12), and replacement of endothelial cells correlates with degrees of injury in human allografts (8, 13, 14). Therefore, one must assume that the health of an allograft can only be maintained if early physiological turnover of endothelial cells is efficient.

The leukocyte-induced neovascular response: road to repair or a means to destruction?

In their seminal studies evaluating the process of leukocyte-induced angiogenesis, Sidky and Auerbach injected either allogeneic or syngeneic spleen cells into the skin of nude mice (15, 16). They found that allogeneic, but not syngeneic, leukocytes mediated an angiogenic response. These classical studies clearly demonstrated that the alloactivated leukocyte produces factors that facilitate neovascularization. Moreover, Auerbach found that allogeneic leukocyte–induced angiogenesis occurred within 3–6 days following injection and was dose-dependent inasmuch as it was enhanced with an increase in the number of injected cells (15). Therefore, it should be no surprise that angiogenesis is a component of allograft rejection. Interestingly, in this issue of the JCI, Babu et al. (7) observed a neovascularization reaction in their analysis of tracheal allografts undergoing rejection. Although they did not characterize this response from the time of transplantation (day 0), their studies clearly illustrate a prominent neovascularization reaction between days 6 and 8 after transplantation; but with ongoing inflammation, this neovascularization reaction did not persist, and it was not sufficient to sustain tissue oxygenation. Therefore, in the tracheal allograft model, local tissue hypoxia and the production of hypoxia-inducible growth factors are likely the stimuli for the initial revascularization response (until day 4); but, one might conclude that it is unlikely that hypoxia mediates the later angiogenenic response. Rather, the second phase of angiogenesis is likely mediated by infiltrating allogeneic leukocytes and inflammation, analogous to Auerbach’s original observations. The local tissue hypoxia may be a consequence of sluggish blood flow associated with the inflammatory reaction.

So why does this later neoangiogenic response fail to sustain allograft function? This is an important question, as this response likely occurs in association with chronic graft rejection. One possible answer to this question relates to the proinflammatory nature of the leukocyte-inducible angiogenic reaction (Figure 1). Neovessels are in themselves “sticky” and facilitate the recruitment of leukocytes, in part via their expression of adhesion molecules and chemokines. Therefore, in allografts, once inflammatory angiogenesis is established, it may serve to facilitate alloimmune leukocyte recruitment at a time when clonal expansion of destructive lymphocytes is maximized. This will result in further injury. This concept is characteristic of chronic inflammatory disease processes, which have been defined as repetitive bouts of acute inflammation and tissue destruction that proceed simultaneously with attempts at repair (17). The study by Babu et al. (7) has provided some clues as to the functional role of this type of angiogenesis. The authors found that removal of allografts with intact microvessels from the recipients on day 6 and retransplantation into immunosuppressed allogeneic recipients resulted in rescue, with normal airway morphology at day 28. If, on the other hand, microvessels within the removed and retransplanted allografts were not intact, then rescue was not achieved. Therefore, one might conclude that so long as early microvascular repair occurs efficiently, then protection of homeostatic angiogenesis/microvascular repair and the interruption of leukocyte recruitment and leukocyte-induced angiogenesis within grafts at critical times after transplant will sustain allograft function.

A balancing act: therapeutic considerations

In vascularized solid organ allografts, such as kidney or heart, the early ischemia/reperfusion response can result in profound injury to the microvasculature (6, 8). Clearly, in part, early repair can be augmented by limiting the degree of injury (hypoxia, oxidative stress, ischemia/reperfusion, etc.), but it could also be accomplished via specific therapies, for instance, erythropoietin treatment (18). Erythropoietin can induce protective genes within vascular endothelial cells (phosphorylated Akt, Stat5, and downstream substrates such as members of the FoxO family, eNOS, and Bcl2, etc.) (19) and has the potential to facilitate microvascular repair following ischemia/reperfusion (20).

Growth factors such as VEGF have multiple biological properties that can facilitate both protection of the microvasculature and angiogenesis (5, 21). However, VEGF also facilitates leukocyte recruitment and in this manner may promote tissue injury (22). VEGF may be a prototype growth factor that facilitates the overlap between angiogenesis and inflammation (5). Consistent with studies in vascularized allografts (22), Babu et al. (7) found that blockade of VEGF did not inhibit initial revascularization, but it markedly inhibited the later leukocyte-induced response (after day 4–6). We suspect that VEGF antagonism may have had some effect on the early response (albeit subtle) to disrupt the degree of protection or cell survival pathways within endothelial cells, such that any inflammation response at later times is more destructive. Clearly there are times when VEGF may be protective and there are also circumstances, such as in inflammation, when VEGF-induced angiogenesis is not protective and may become a means to destruction. As discussed by the authors (7), perhaps VEGF expression within the tracheal allograft serves to promote leukocyte recruitment, but this inflammatory angiogenesis may only be temporarily protective. Moreover, we suggest that leukocyte-induced angiogenesis cannot promote effective repair and thus, we like to consider it “pathophysiological angiogenesis.” This is the type of angiogenesis that occurs in association with chronic rejection. If VEGF/VEGF function is inhibited or if inflammation is inhibited, the cycle of leukocyte recruitment and angiogenesis will be reduced, and thus, tissue destruction will be attenuated (Figure 1). Indeed, we and others have found that graft morphology is remarkably improved and infiltrates are reduced when recipients with early chronic rejection are treated with angiogenesis inhibitors or VEGF antagonists (5, 23–25).

In conclusion, the study by Babu et al. (7) reported in this issue of the JCI has emphasized the importance of microvascular integrity for long-term allograft function. These data suggest that therapy that protects vascular endothelial cells from injury and/or optimizes physiological noninflammatory angiogenesis within allografts may serve to facilitate long-term allograft function. In contrast, the later pathophysiological inflammatory angiogenesis may facilitate ongoing leukocyte recruitment and injury and does not serve a protective function. Clearly, this is an area that deserves further study, as chronic rejection is the most common cause of solid organ allograft failure and to date no therapy has been successful in interrupting this process in humans. Therefore, a robust vasculature appears to be the “silver lining” that is necessary to sustain long-term allograft function.