INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cytomegalovirus (CMV) infection is an important cause of morbidity and mortality among transplant recipients (1). In addition, CMV infection may elicit indirect effects on allograft dysfunction (2). Acute CMV infection may accelerate the appearance of early episodes of acute rejection (3, 4), and it may subject the allograft to late rejection (5). On the other hand, various cytokine responses during alloimmune activation, sepsis, and treatment of rejection with T cell antibodies may activate latent CMV infection (6). Years later, these early indirect effects of CMV may lead to reduced graft survival and the development of chronic rejection (9). The detrimental effects of CMV on various forms of allograft dysfunction emphasize the need for precise diagnostics and preemptive or prophylactic antiviral therapy.

The heterotopic rat tracheal allograft model is well established for the investigation of experimental obliterative bronchiolitis (10). In this model, the donor trachea is excised and transplanted in the greater omentum of the recipient. In syngeneic grafts, epithelium undergoes minor damage but recovers thereafter. The tracheal lumen remains completely open and the trachea is lined with normal, mucus-secreting epithelium at 30 d after transplantation (12). On the other hand, in untreated allografts, epithelium undergoes progressive damage leading to near total necrosis 10 d after transplantation. Allografts develop a strong inflammatory response, which is associated with increased expression of cytokines, chemokines, and growth factors, culminating in the development of a fibroproliferative lesion obliterating the tracheal lumen, which closely resembles obliterative changes detected in small bronchioles (OB) in humans (12).

Our observations using this model demonstrate that rat cytomegalovirus (RCMV) infection enhances smooth muscle cell proliferation and changes of obliterative bronchiolitis, which is a form of chronic rejection after lung transplantation (13). So far, it has not been possible to recover any viral antigen or genome from smooth muscle cells (SMC) by immunohistochemistry or in situ hybridization in this and other experimental models of chronic rejection (13, 14). Only occasionally have RCMV antigens been located in media SMC in highly immunosuppressed rats (15, 16). Rather, early and late RCMV antigen expression has been found only in scattered mononuclear inflammatory cells in the allograft. On the other hand, an alloimmune response must be associated with viral infection to accelerate chronic rejection, as RCMV infection does not promote chronic rejection in syngeneic grafts (17).

We hypothesized that RCMV infection enhances the development of chronic rejection by augmenting the alloimmune response leading indirectly to accelerated development of chronic rejection, rather than through a direct virus-mediated effect on SMC migration and proliferation. The purpose of this study was to test our hypothesis by investigating how immune modulation and antiviral therapy regulate RCMV infection-enhanced chronic rejection and to test the effect of the different treatment regimens in the prevention of RCMV infection-enhanced experimental OB.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Rats

Male DA (AG-B4, RT1^a) and WF (AG-B2, RT1^u) rats weighing 200- 300 g (Laboratory Animal Center, University of Helsinki, Helsinki, Finland) were used. Tracheal allografts were transplanted from DA donors to WF recipients as described previously (12). Permission for animal experimentation was obtained from the Provincial State Office of Southern Finland. Rats received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" (NIH Pub. No. 80-23, revised 1978).

Virus Stocks, Infections, Titration, and Nested PCR

Stocks of Maastricht strain RCMV salivary gland extracts were prepared as previously described (18). Acute RCMV infection was initiated intraperitoneally 3 h after transplantation with 5 × 10⁵ plaque-forming units (PFU) of RCMV. Chronic infection was established by inoculating RCMV 8 wk before transplantation. Titers in whole organ biopsies of salivary glands were measured as PFU, and graded from 0 to 4 as follows: 0, no cytopathic effect; 1, < 25% cytopathic effect; 2, 25-50% cytopathic effect; 3, 50-75% cytopathic effect; 4, > 75% cytopathic effect. The expression of the major immediate early (MIE) DNA of RCMV from biopsies of tracheal allografts was determined by a sensitive, single-tube, nested polymerase chain reaction (PCR) (19).

Treatment Protocols

Allograft recipients received cyclosporine A (CsA) 1.5 mg/kg/d subcutaneously and one group of rats received CsA 2 mg/kg/d subcutaneously. In the prophylaxis group, the recipients received either 9-(1,3-dihydroxy-2-propoxymethyl)guanine (DHPG), ganciclovir, Cymevene (Roche, Palo Alto, CA) starting 12 h before transplantation, or a single dose of rat hyperimmune serum (HIS) perioperatively. Treatment with either DHPG, or a single dose of HIS, or the combination of these drugs, was initiated 5 d after transplantation at the time of acute RCMV infection. Controls received either saline or nonimmune Brown-Norway rat sera. To evaluate the direct effect of DHPG on SMC proliferation, a group of noninfected nonimmunosuppressed rats was treated with DHPG only.

Drugs Used

Cyclosporine A (Sandimmun; Novartis, Basle, Switzerland) whole blood 24-h trough levels were determined weekly using radioimmunoassay (Sandimmun-Kit; Novartis) and ranged between 250 and 350 µg/L in all groups except in the group receiving CsA 2 mg/kg, where the trough level was 450-550 µg/L. DHPG was diluted in saline and given 20 mg/kg/d intraperitoneally in two doses. HIS with a neutralization titer of 160 was diluted 1:4 in phosphate-buffered saline (PBS) and 1 ml of this dilution was given intravenously.

Histological Evaluation and Morphometry

For histological evaluation, tracheal frozen sections were stained with Mayer's hematoxylin-eosin (H&E). Epithelial necrosis was evaluated as percentage of the tracheal circumference not lined by epithelium. Luminal occlusion was evaluated by determining the reduction in luminal area using NIH Image program version 1.59 (National Technical Information Service, Springfield, VA).

Single Immunostaining and In Vivo Cell Proliferation

The grafts were removed 10 d after transplantation to examine inflammatory cell infiltration, cytokine production, and cell proliferation. Immunostaining was performed using specific rat antibodies and the Vectastain Elite ABC Kit. The results are expressed as number of positively staining cells per allograft cross section. In the case of tumor necrosis factor- (TNF-) and MHC class II, the intensity of staining was scored semiquantitatively from 0 to 3 as follows: 0, no visible staining; 1, few cells with faint staining; 2, moderate intensity with multifocal staining; and 3, intense diffuse staining of the cells analyzed. The myofibroproliferative lesion is defined as the tracheal airway wall not surrounded by cartilage as well as all the tissue occluding the lumen.

Statistical Analyses

All data are expressed as mean ± SEM. For two group comparisons of small sample size, nonparametric Mann-Whitney U-test (Statview 512+ program; Brain Power Inc., Calabasas, CA) was used. For multiple group comparisons, the results of nonparametric Kruskal-Wallis (Statview 512+ program) test were used for the Dunn test at the significance level of 5% and 1% (Medstat; Astra Group A/S, Copenhagen, Denmark). p < 0.05 was regarded as significant (see online data supplement).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

RCMV Infection

The natural history of RCMV infection has previously been described in detail (20). When native rats are exposed to 10⁵ PFU of RCMV, acute infection with systemic virus dissemination occurs within 3 to 7 d. Shortly thereafter, the infection progresses to chronic phase, and the virus can be recovered only from the salivary glands of the host. Approximately after 3 to 4 mo postinoculation, infectious virions are no longer present in any tissue or organ, and the rat is considered to be latently infected (20, 21).

As shown in Table 1, in immunosuppressed allograft recipients with acute RCMV infection, no infectious virus could be detected from the biopsies of liver, spleen, or salivary glands at 10 d. However, MIE-DNA of RCMV could be detected from either the spleen or liver of the rats at 10 d (not shown). At 30 d after transplantation, high numbers of infectious RCMV could be recovered by plaque assay from salivary glands in saline- or normal rat sera- treated controls. Neither prophylaxis with DHPG or HIS nor treatment with DHPG altered the amount of infectious RCMV in the salivary glands. Treatment with HIS or with the combination of DHPG and HIS significantly reduced the number of infectious RCMV compared with those of both normal rat sera-treated rats as well as rats receiving HIS prophylaxis (p < 0.05). In allograft recipients with chronic RCMV infection, no replicative virus could be identified from the liver, spleen, or salivary glands either 10 or 30 d after transplantation. As demonstrated by nested PCR, RCMV MIE-DNA was found in the tracheal allografts of all groups during acute or chronic RCMV infection at 10 and 30 d after transplantation (Table 1). RCMV early and late antigen expression was also analyzed immunohistochemically from the tracheal allografts of noninfected controls and saline and normal rat sera-treated RCMV-infected rats at 10 d after transplantation. As expected, no RCMV antigen expression could be detected from the noninfected controls. Only 1 of 10 tracheal allografts of the RCMV-infected rats demonstrated RCMV antigen expression and even then in only few mononuclear inflammatory cells (data not shown).

Histology

The extent of epithelial necrosis and degree of luminal occlusion were determined 30 d after transplantation, and the results are summarized in Figures 1 and 2. Noninfected allografts showed mild epithelial necrosis and the degree of luminal occlusion was 29% ± 10%. In allograft recipients receiving saline or normal rat sera, both acute and chronic RCMV infection led to near complete epithelial necrosis and induced a three-fold increase in the degree of luminal occlusion, when compared with noninfected controls (p < 0.01). In recipients with acute infection, prophylaxis with DHPG significantly reduced the development of epithelial necrosis (p < 0.05) and tracheal occlusion (p < 0.01), whereas HIS prophylaxis reduced only the degree of luminal occlusion (p < 0.05). Treatment with DHPG initiated 5 d after transplantation reduced luminal occlusion (p < 0.05), but markedly less than DHPG prophylaxis (p < 0.01). Treatment with HIS alone or in combination with DHPG had no effect on histology. When RCMV-infected allograft recipients received CsA 2.0 mg/kg, a highly significant reduction in both epithelial necrosis and tracheal occlusion was observed compared with RCMV-infected recipients receiving CsA 1.5 mg/kg (p < 0.01). Also, in recipients with chronic RCMV infection, DHPG prevented the development of luminal occlusion (p < 0.05).

View larger version (39K): [in this window] [in a new window]   Figure 1.   Effect of different treatment regimens on the percentage of epithelial necrosis (upper panel ) and the degree of luminal occlusion (lower panel ) in rat tracheal allograft recipients with acute (A, C ) and chronic (B, D) RCMV infection. Acute RCMV infection was initiated 3 h after transplantation with 5 × 105 PFU of Maastricht strain RCMV. Chronic infection was established by inoculating RCMV 8 wk before transplantation. All allograft recipients received cyclosporine 1.5 mg/ kg/d subcutaneously, except for one group that was given CsA 2.0 mg/ kg/d. Data are given as mean ± SEM. *p < 0.01 compared with noninfected controls, **p < 0.05 and §p < 0.01 compared with respective saline- or normal rat sera-treated controls by Mann-Whitney for two group comparison and Kruskal-Wallis and Dunn tests for multiple comparison. n = 10/group, and n = 5 in CsA 2.0 mg/kg/d. Nil indicates noninfected allograft recipients.

View larger version (68K): [in this window] [in a new window]   Figure 2.   Photomicrographs of tracheal allografts 30 d after transplantation. On the top panel: (A) noninfected allograft and (B) noninfected, nonimmunosuppressed DHPG-treated allograft. On the four middle panels RCMV-infected allografts are presented: (C ) saline-treated control, (D) normal rat sera control, (E ) DHPG prophylaxis, (F  ) HIS prophylaxis, (G) DHPG treatment, (H ) HIS treatment, (I ) DHPG and HIS treatment, (  J ) CsA 2.0 mg/kg. The lowest panel shows chronic RCMV- infected allografts: (K ) saline-treated control, (L) DHPG prophylaxis. (H&E; original magnification: ×40).

To exclude the direct inhibitory effect on SMC proliferation, a group of noninfected rats was treated with DHPG only. The results demonstrated that DHPG had no inhibitory effect on the development of epithelial necrosis or tracheal allograft obliteration (not shown).

In Vivo Cell Proliferation

The rate of in vivo cell proliferation at 10 d was determined by bromodeoxyuridine (BrdU)-immunoreactivity in the airway wall and myofibroproliferative lesion (Figure 3). Both acute and chronic RCMV infection led to a significant increase in the number of proliferating cells in the myofibroproliferative lesion (Figure 3C and 3D). In recipients with acute RCMV infection, DHPG prophylaxis was associated with a significant decrease in the number of proliferating cells in the airway wall (p < 0.01) as well as proliferating cells in the myofibroproliferative lesion (p < 0.01), compared with saline-treated controls. HIS prophylaxis proved insufficient in down-regulating cell proliferation, whereas HIS treatment either alone or in combination with DHPG reduced the proliferative response in the allograft (p < 0.05) during acute RCMV infection, but did not result in decreased luminal occlusion. Treatment with CsA 2 mg/kg did not affect the number of proliferative cells. During chronic RCMV infection, DHPG prophylaxis halved the number of proliferating cells in the airway wall (p = NS) but did not affect BrdU immunoreactivity in the myofibroproliferative lesion.

View larger version (34K): [in this window] [in a new window]   Figure 3.   Effect of different treatment regimens on in vivo cell proliferation in the airway wall (upper panel ) and in the myofibroproliferative lesion (lower panel ) in rat tracheal allograft recipients with acute (A, C) and chronic (B, D) RCMV infection 10 d after transplantation. The number of BrdU-labeled cells per allograft cross section was counted. Values are mean ± SEM. n = 10/group, and n = 5 in CsA 2.0 mg/ kg/d, HIS prophylaxis, and HIS treatment. For details, see the legend of Figure 1.

Immunohistochemical Analyses of Tracheal Allografts

In noninfected immunosuppressed allografts, a moderate inflammatory response consisting of CD4⁺ and CD8⁺ T cells and ED1⁺ macrophages was observed (Table 2). RCMV infection increased the number of graft-infiltrating ED1⁺ macrophages (p = NS), increased the immunoreactivity of TNF- (p < 0.05) and interleukin (IL)-2 (p < 0.05), and decreased the immunoreactivity of IL-10 (p < 0.05), compared with noninfected controls (Figures 4 and 5). The beneficial effect of DHPG prophylaxis was associated with decreased numbers of cells expressing interferon (IFN)- (p < 0.05) and TNF- (p = NS), and with an increase in cells staining positive for IL-10 (p = NS), compared with RCMV-infected controls or antiviral treatment groups (Figures 4 and 5). HIS prophylaxis significantly reduced the number of graft-infiltrating ED1⁺ cells (p < 0.05) and IFN- (p < 0.05) immunoreactivity but was associated with an increase in TNF- immunoreactivity (p < 0.05). Immunosuppression with high dose CsA decreased IL-2 (p < 0.05) and TNF- (p < 0.05) immunoreactivity and increased that of IL-10 (p = NS) in comparison with RCMV-infected controls. Chronic RCMV infection was associated with an increased number of cells expressing IL-2 (p < 0.05) and TNF- (p < 0.05) as well as a decrease in IL-10 immunoreactivity (p < 0.05). DHPG prophylaxis attenuated these changes.

View larger version (116K): [in this window] [in a new window]   Figure 4.   Photomicrographs of interleukin-2 (IL-2), interferon- (-IFN), interleukin-4 (IL-4), interleukin-10 (IL-10), and tumor necrosis factor- (TNF-) immunoreactivity in noninfected allografts (Nil), saline-treated RCMV-infected allografts (Saline), RCMV-infected allografts receiving ganciclovir prohylaxis (DHPG), and RCMV-infected allografts receiving CsA 2.0 mg/kg (CsA 2 mg/kg) 10 d after transplantation (original magnification: ×200).

View larger version (17K): [in this window] [in a new window]   Figure 5.   Effect of different treatment regimens on cytokine expression in the allografts at 10 d after transplantation. TNF- expression was graded semiquantitatively from 0 to 3, whereas other cytokines are shown as positive cells/cross section. Values are mean ± SEM. n = 10/group, and n = 5 in CsA 2.0 mg/ kg/d, HIS prophylaxis, and HIS treatment. For details, see the legend of Figure 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, RCMV infection was established either at the time of transplantation, mimicking primary CMV infection, or 2 mo before transplantation, yielding chronic RCMV infection. Both acute and chronic RCMV infection markedly enhanced the development of tracheal occlusion compared with noninfected CsA-immunosuppressed controls. Prophylaxis with either DHPG or HIS completely prevented the deleterious effects of acute RCMV infection, whereas treatment with DHPG, HIS, or the combination of these drugs had only a slight beneficial effect. The positive effect of DHPG prophylaxis was evident also in the chronic infection group. Furthermore, increased CsA immunosuppression resulted in diminished tracheal occlusion during acute RCMV infection, demonstrating that immune activation is one of the requirements for RCMV infection-enhanced lesion development.

In recipients with acute RCMV infection, replicative RCMV could be isolated from the salivary glands of nearly all allograft recipients regardless of the antiviral treatment strategy. Prophylaxis with neither DHPG nor HIS affected the amount of infectious RCMV recovered from rat salivary glands, whereas treatment with HIS or the combination of DHPG and HIS significantly reduced the number of infectious virus. RCMV MIE-DNA expression was uniformly present in allografts. These results indicate that CMV infection-enhanced cell proliferation develops independently of the viral load. In the recipients with chronic RCMV infection, no evidence of systemic RCMV activation was observed despite the presence of RCMV infection-enhanced chronic rejection. However, RCMV MIE-DNA could be extracted from the allografts in both the control and DHPG groups. It is plausible that latent RCMV residing in the graft-infiltrating recipient monocytes or macrophages may be activated by the allograft environment, for example by IFN- or TNF- (6, 7). Thus, either the activated virus or a viral epitope on the cell surface would evoke a local immune response. On the other hand, CMV is able to encode -chemokines and an -chemokine recruiting macrophages and neutrophils to the graft (22, 23). Of these cells, macrophages are capable of producing cytokines and growth factors regulating SMC migration and proliferation, and intragraft PDGF-AA and -BB expression has been linked to the development of chronic rejection (12, 24).

The results of the present study argue for the role of early and late gene products of RCMV as a triggerer and/or enhancer of immune responses leading to cytokine, chemokine, and growth factor production and SMC migration and proliferation, as DHPG affects RCMV nucleic acid synthesis only after the production of MIE-DNA gene products (25). Synthesis of this and previous studies also support the hypothesis that RCMV-enhanced chronic rejection may be an immunopathological event rather than a direct virus-mediated effect on SMC migration and proliferation (12, 24). However, in vitro studies suggest that CMV may also directly regulate SMC migration and proliferation. HCMV IE84 protein has been shown to increase SMC proliferation through p53 tumor suppressor gene inactivation (26) and HCMV IE gene products have been reported to up-regulate PDGF- receptor expression in rat SMC culture (27). Recent in vitro evidence suggests that HCMV-encoded chemokine receptor US28 mediates SMC migration in the presence of macrophage-derived MCP-1 or RANTES (28). Of notice is that animal models do not support direct effects of the virus on SMC, for no viral antigen or genome has been recovered from SMC by immunohistochemistry or in situ hybridization in this and other experimental models of chronic rejection (13, 16). Only occasionally have RCMV antigens been located in media SMC in strongly immunosuppressed rats (15, 16). Instead, early and late RCMV antigen expression has been located only in scattered mononuclear inflammatory cells in the graft.

Previous observations in rat tracheal and aortic allograft models demonstrate that acute alloimmune response must be associated with RCMV infection to induce accelerated allograft arteriosclerosis (13, 17). When RCMV infection was given 2 mo after transplantation, at the time of subsided alloimmune response, the effect on the development of chronic rejection was nonexistent (17). These experimental findings are in line with recent clinical observations in human heart transplant recipients, where lack of DHPG prophylaxis was a major risk factor for the development of chronic rejection in CMV seronegative patients receiving seropositive allografts, although the incidence of CMV illness was not altered (9). Although the incidence of CMV infection was not reduced, it is most likely that CMV infection postponed from the time of acute alloimmune response to a later time point blocks the deleterious effects of CMV infection on chronic rejection. In addition, the activation of latent CMV by acute alloimmune response-associated cytokines, in particular by interferon- and TNF-, is also inhibited by DHPG prophylaxis (6, 7).

Lately, a wealth of evidence indicates an indirect proinflammatory role for CMV in the pathogenesis of chronic rejection. CMV infection has been associated with increased serum levels of Th1-type cytokines (29), as well as TNF- production by infected macrophages (30). Furthermore, CMV infection- induced NF-B-mediated IL-6 production in lung fibroblasts (31). In addition to enhancing the production of host proteins, the CMV genome may encode its own cytokine homologs, such as the viral IL-10 homolog (32). In our study, prevention of tracheal occlusion was associated with a reduction in the number of ED1⁺ cells and cells staining positive for Th1 cytokines. Up-regulation of IL-10 during RCMV infection was not observed, but it is possible that the viral IL-10 homolog does not cross react with the antibody used in this study. Also, prophylaxis with DHPG-reduced TNF- immunoreactivity during both acute and chronic RCMV infection. It is likely that the proinflammatory effects of CMV are mediated indirectly mainly via up-regulation of these cytokines. Thus, acute RCMV infection would provide increased means for both activation and migration for host inflammatory cells, leading to an enhanced alloimmune response and local cytokine and mesenchymal cell growth factor production, resulting in myofibroproliferation and finally obliteration of the airway lumen. On the other hand, allograft transplantation is associated with a surge in proinflammatory cytokine production as part of the early alloimmune response, which in turn may cause the activation of latent CMV (6, 7). In this study, increased immunosuppression prevented the development of RCMV-enhanced obliterative bronchiolitis by down-regulation of the RCMV infection-associated immune activation and/or by inhibition of alloimmune response-induced viral activation.

Of interest was the finding that DHPG prophylaxis reduced tracheal occlusion below the level of noninfected controls. A plausible explanation is offered by the "bystander effect" in which the conversion of DHPG leads to 100-fold local concentrations of its active cytotoxic triphosphate form (33). These concentrations could then induce DNA chain termination in dividing cells and thereby inhibit cell proliferation not only in the CMV-infected cells but also in those surrounding them.

Taken together, our results suggest that CMV infection-enhanced experimental obliterative bronchiolitis develops independently of the viral load and results from simultaneous interaction between RCMV infection and acute allommune response. Antiviral prophylaxis, either with DHPG or HIS, entirely abolishes the accelerating effect of RCMV on the development of luminal occlusion associated with down-regulation of ED1⁺ macrophages, TNF-, and Th1 cytokines. Furthermore, we show that blocking the IL-2 activation pathway with high dose CsA completely prevents RCMV infection-enhanced obliterative bronchiolitis and is associated with a shift toward Th2 cytokine expression. Because DHPG inhibits RCMV nucleic acid synthesis after IE gene expression, IE gene expression does not suffice to cause RCMV infection-enhanced obliterative bronchiolitis, but rather, gene expression beyond immediate early genes, together with adequate immune response, is required for lesion development. However, as we have so far recovered only minimal RCMV early and late antigen expression from aortic and tracheal allografts, our results imply that RCMV infection-enhanced chronic rejection in vivo is an immunopathological disorder and not a result of a direct virus-mediated effect on SMC migration and proliferation.