INTRODUCTION
Cytomegalovirus (CMV) infection is associated with adverse effects in renal transplantation.1 The “direct effects” of CMV are caused by viral reactivation, replication (DNAemia), and CMV end-organ disease. The “indirect effects” of CMV include associations with acute rejection (AR) and late graft loss; virus-induced systemic immunomodulation conferring increased susceptibility to bacterial, fungal, and other viral infections; posttransplant vasculopathy; new-onset diabetes after transplantation; and other transplant comorbidities.1-5 Human CMV (HCMV)–positive serostatus is associated with inferior graft outcome, with the highest graft loss observed among HCMV-seropositive patients with AR episodes.6-9 Patients who develop HCMV DNAemia are also more likely to experience graft dysfunction and loss compared with those without DNAemia, suggesting that viral reactivation or replication may contribute to graft dysregulation.10-14 HCMV antigens can be identified in acutely rejecting allografts, as well as in those explanted due to graft failure.15-17 Antiviral treatment with ganciclovir (GCV) or valganciclovir to prevent HCMV disease (prophylaxis or preemptive therapy) is associated with improved graft survival and reduced interstitial fibrosis and tubular atrophy, suggesting that inhibition of viral replication may be beneficial for graft outcome.18-21 However, the exact mechanisms by which HCMV might contribute to renal allograft dysfunction remain incompletely understood.
Animal models for CMV pathogenesis have been used to demonstrate essential immunologic and host-pathogen interactions. In rodent renal transplant models, rat CMV (RCMV) or murine CMV (MCMV) infection is associated with increased inflammation and accelerated graft injury compared with uninfected allografts.22-26 In rat kidney transplants, RCMV infection interferes with tolerance induced by anti-CD4 monoclonal antibodies and increases both antiviral and alloreactive T-cell responses resulting in chronic allograft damage.27 In the murine model, MCMV reactivation from latency in the donor kidney is induced after allogeneic transplantation via cytokines such as tumor necrosis factor-α and by ischemia-reperfusion injury (IRI).28-31 MCMV infection of the donor allograft exacerbates intragraft infiltration of CD8+ T cells, macrophages, neutrophils, and natural killer (NK) cells.26,32 GCV administration ameliorates MCMV-associated allograft injury, supporting a role for viral reactivation in the pathogenesis of allograft damage.26
AR can be precipitated by T cell–mediated rejection or antibody-mediated rejection (AMR).33-37 AMR is known to be mediated by donor-specific antibodies directed against HLA antigens. However, circulating donor-specific antibodies are sometimes not found during episodes of AMR, raising the possibility that non-HLA antibodies might contribute to AMR.38-40 Recent literature suggests that patients who have antibodies directed against self-proteins, such as angiotensin-II type 1 receptor, MHC class I-related chain A, and endothelin type A receptor, may also have adverse kidney transplant outcomes.38,41-48 The role of pathogen-induced antibodies in AMR has not been explored. Although pathogen-induced antibodies serve important functions in host protection from infectious diseases, deposition of these antibodies could conceivably occur in the pathological setting of allograft rejection, particularly for viruses infecting the transplant organ such as HCMV. Anti-HCMV antibody titers vary among kidney transplant recipients and correlate with CMV DNAemia, indicating that antiviral antibody quantities differ among individuals and increase after viral reactivation, even in the immunocompromised host.49 The aim of the present study was to examine whether pathogen-induced antibodies can be detected in acutely rejecting allografts, using MCMV as a model pathogen that infects donor kidneys.
MATERIALS AND METHODS
Animals and Viruses
BALB/cJ (“BALB/c”), C57BL/6J (“B6”), and C57BL/6-Igh-6tm1Cgn (“µMT”) mice were purchased from Jackson Laboratory (Bar Harbor, ME). The B cell–deficient (µMT) mice have a targeted gene deletion in the µ-chain resulting in the deficiency of mature B cells and the absence of detectable immunoglobulin G (IgG) in serum.50 Mice were maintained in animal facilities approved by the Association for Assessment and Accreditation of Laboratory Animal Care under the Animal Resources Program of the University of Alabama at Birmingham (UAB, Birmingham, AL) or the Animal Resources Core of The Abigail Wexner Research Institute at Nationwide Children’s Hospital (AWRI, Columbus, OH) under specific pathogen-free conditions. Murine experimental protocols were approved by the Institutional Animal Care and Use Committees at UAB and AWRI.
MCMV strain Smith deleted for open reading frame m157 (“MCMVΔ157”), which infects BALB/c and B6 mice similarly, and green fluorescent protein-expressing MCMV (MCMV-GFP) viruses were provided by Ulrich Koszinowski (Maz von Pettenkofer-Institute, Ludwig Maximilians-University, Munich, Germany).51,52 Virus stocks were propagated using murine embryonic fibroblasts (MEFs) as described, frozen in aliquots at –80°C, and discarded after single use.26
Murine Kidney Transplantation
Donor and recipient mice were infected by intraperitoneal (IP) injection with 104 plaque-forming units (PFUs) MCMVΔ157 at least 12 weeks before renal transplantation (D+ or R+ transplants).26,32,53 Kidney transplantation was performed using vessel and bladder cuffs according to published techniques, with 1 or both native kidneys retained to maintain renal function posttransplant.26,54 BALB/c (allogeneic) or C57BL/6 (syngeneic) kidneys were transplanted into either B6 or µMT recipients. For some experiments, recipients were treated with cyclosporine (Novartis Pharmaceuticals, Cambridge, MA) at 10 mg/kg/d subcutaneously (SC), once daily until terminal euthanization. Each experimental transplant cohort consisted of 3–6 mice per group.
Tissues stored from a previously published GCV-treated cohort were also analyzed for this study.26 This group was treated with cyclosporine and with GCV (Genentech, Inc., South San Francisco, CA) at 15 mg/kg/d SC once daily for 14 days and euthanized at either day 14 or day 21 (n = 4/group).
Immunofluorescence Analysis of Kidney Sections
Recipients were euthanized, perfused to organ pallor with ice-cold saline, and organs were processed for formalin fixation and paraffin embedding (FFPE). Kidney sections were deparaffinized, blocked with goat serum, and incubated with either AlexaFluor-488 (AF488)-conjugated goat anti-mouse IgG antibodies (Invitrogen, Carlsbad, CA) or fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse complement C3 antibodies (MP Biomedicals LLC, Solon, OH), counterstained with 4′, 6-diamidino-2-phenylindole, and mounted using ProLong anti-fade reagent (Invitrogen, Carlsbad, CA). For each experiment, images were collected under identical exposure times and gained using an Olympus BX51 fluorescence microscope and Olympus image processing software. AF488- or FITC-positive immunofluorescence signals per high-power field (HPF) were quantified using ImageJ (NIH, https://imagej.net). For large, confluent areas of AF488- or FITC-positive staining, positive signals were quantitated by counting the number of 4′, 6-diamidino-2-phenylindole-positive nuclei within or immediately adjacent to each AF488 or FITC-positive area. For each slide, 10 HPFs were quantified (5 cortex and 5 medullae) and the average number of positive signals per HPF calculated for each sample. Kidneys from nontransplant B6 mice were procured and processed in an identical fashion as staining controls.
Passive Transfer of MCMV Antisera in µMT Transplants
To generate anti-MCMV antisera, B6 mice were infected with 106 PFU of MCMVΔ157 IP on day 1; boosted with 105 PFU MCMVΔ157 at days 14 and 21; euthanized at day 24; and MCMV immune sera collected, pooled, and analyzed by ELISA (“MCMV+ sera”). Sera from uninfected B6 mice were also collected and pooled (“MCMV– sera”). Recipient µMT mice were either untreated (no sera) or injected at day +1 posttransplant with 200 µL of MCMV– or MCMV+ sera IP and euthanized at posttransplant day 3 or 14.
ELISA for Anti-MCMV Antibodies
Ninety-six-well enzyme immunoassay/radioimmunoassay plates (Costar, Corning, NY) were coated with MCMV-infected MEF cell lysate, blocked with goat serum, and incubated with 2-fold serial dilutions of B6 MCMV– sera, MCMV+ sera, or transplant sera (D–/R–, D+/R–, D+/R+). The plates were developed using horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibodies and 3, 3′, 5, 5′-tetramethylbenzidine substrate solution (Pierce Biotechnology, Rockford, IL). The optical density was measured with an ELX808 Ultra Microplate Reader (BioTek Instruments, Winooski, VT) at 450 nm wavelength.
Western Blotting of Kidney Lysates
Portions of kidneys were snap-frozen at –80°C on the day of euthanization. Tissues were homogenized in radioimmunoprecipitation assay lysis buffer (50 mmol/L Tris-HCl pH 7.4, 150 mmol/L NaCl, 2 mmol/L EDTA, 1% NP-40, 0.1% SDS) containing protease inhibitor cocktail (Pierce Biotechnology). For each sample, 20-μg protein was separated by 10% SDS-PAGE and Western blotting was performed using HRP-conjugated goat anti-mouse IgG (H + L chain) and West Pico reagent (Pierce Biotechnology). Membranes were stripped and reprobed using a rabbit anti-actin antibody (Sigma-Aldrich, St. Louis, MO), HRP-conjugated goat anti-rabbit-IgG secondary antibody, and West Pico reagent. Band intensities were quantitated by densitometry using ImageJ.
Histopathology
FFPE transplant and native kidney sections were stained with hematoxylin and eosin. Histopathology was scored in a blinded fashion by veterinary pathologists (T.R.S. or R.C.), using a published scale of 8 pathological criteria (range, 0–3; maximum score, 24) reflecting criteria used for clinical allograft rejection.32,53,55 Concordance of scoring between the pathologists was confirmed using pathological samples previously used in other studies.32,53
Renal Epithelial Cell Isolation
Murine primary renal tubular epithelial cells (TECs) were isolated using selective media as described, with minor modifications.56 In brief, B6 or BALB/c kidneys were minced in Dulbecco’s Modification of Eagle’s Medium (Mediatech Inc., Manassas, VA) with Ham’s F12 (Cellgro, Herndon, VA) at 1:1 ratio, dissociated using collagenase at 1 mg/mL (Worthington Biochemical Corp., Lakewood, NJ), homogenized by needle disruption, and propagated on plates coated with bovine collagen I at 1 µg/mL (Advanced BioMatrix, Carlsbad, CA) using complete epithelial cell media (Cell Biologics, Chicago, IL).
Antibody-binding Assay
B6 TECs, BALB/c TECs, or BALB/c 3T3 cells (American Type Culture Collection, Manassas, VA) were infected with MCMV-GFP virus at a multiplicity of infection of 1 for each cell type. At 72 hours postinfection, the cells were detached and incubated with either MCMV– or MCMV+ B6 sera or with transplant sera (D–/R– or D+/R+) at 1:20 dilution for 1 hour on ice. MCMV-uninfected cells were also incubated with sera. The cells were washed and stained for viability (Ghost Dye UV450, TONBO Biosciences, San Diego, CA), EpCAM-PE (CD326, Clone G8.8, Biolegend, San Diego, CA), and anti-mouse IgG1-allophycocyanin (Clone A85-1, BD Biosciences, San Jose, CA). Samples were analyzed using a LSRFortessa (BD Biosciences) and FlowJo V.10 software (TreeStar Inc., Ashland, OR). Experiments were performed 3 times independently.
Complement-dependent Cytotoxicity Assay
BALB/c TECs or 3T3 cells were infected with MCMVΔ157 at multiplicity of infection of 1 for 72 hours, then incubated with B6 MCMV– sera, MCMV+ sera, or transplant sera at 1:20 dilution, with or without rabbit complement at 1:10 dilution (Cedarlane, Burlington, ON, Canada). Control wells were treated with no sera/no complement or with complement alone. After incubation at 37°C for 3 hours, media were replaced with 50-µL FluoroBrite Dulbecco’s Modification of Eagle’s Medium (Gibco Inc., Gaithersburg, MD) containing 7-AAD at 1:50 dilution (BD Biosciences, San Jose, CA) for 15 minutes. Images were acquired and quantitated using the EVOS FL Auto 2-cell imaging system and software (Thermo Fisher Scientific, Waltham, MA). Experiments were performed 3 times independently.
MCMV Quantitative DNA Polymerase Chain Reaction
MCMV viral loads were quantified from D+/R+ transplant and native kidneys as previously described.26,57 Briefly, total DNA from snap-frozen tissue was isolated using the Qiagen QiaAMP DNA Kit (Qiagen, Germantown, MD); quantitative DNA polymerase chain reaction (PCR) performed in triplicate for each sample using an ABI StepOnePlus Real-Time PCR cycler (Applied Biosystems, Waltham, MA); and copy number calculated by comparison with plasmid-derived standards. Viral loads were depicted as copies per gram tissue.
CD45+ Cellular Infiltrates by Flow Cytometry
At day 14 posttransplant, D+/R+ allografts and native kidneys were mechanically disrupted and digested with 0.05 µg/mL collagenase A (Roche, Indianapolis, IN) for 30 minutes at 37°C. After blocking with anti-mouse CD16/CD32 (clone 93, eBiosciences, San Diego, CA), the cells were stained with propidium iodide and CD45-FITC (clone 30-F11, eBiosciences) and analyzed by flow cytometry using a dual-laser FACSCalibur (BD Biosciences) and FlowJo software.
Statistical Analysis
Groups were compared using Student’s t test or 1-way ANOVA with Tukey’s multiple comparison test. Continuous variables were compared using linear regression. Statistically significant differences were accepted at a P value of <0.05. Results were depicted as mean ± SD. All statistical analyses were performed using Prism 7.0 (GraphPad, San Diego, CA).
RESULTS
IgG and Complement Component C3 Deposit Into MCMV-infected Allografts
We have previously shown that MCMV infection accelerates renal allograft injury compared with uninfected grafts and is associated with T-cell and NK-cell infiltrates consistent with T cell–mediated rejection.26 AMR was not previously assessed in this model. To identify AMR, IgG deposition was examined by immunofluorescence staining of D–/R– and D+/R– allografts without immunosuppression at day 7 posttransplant. FFPE sections were stained with goat anti-mouse IgG-AF488 (Figure 1A), and positive staining quantitated using Image J (Figure 1B). IgG staining was detected in both D–/R– and D+/R– allografts and was more abundant in D+/R– grafts (157 ± 28 versus 36 ± 10 spots/HPF, P = 0.0426). D+/R– grafts had greater IgG immunostaining in the medulla (Figure 1A, lower panel) compared with the cortex (Figure 1A, upper panel). Nontransplant B6 kidney controls showed no IgG staining (Figure 1A, left panel; Figure 1B).
;)
FIGURE 1.: IgG and C3 staining of CMV-infected and uninfected renal allografts. (A, C) Allografts of MCMV D–/R– and D+/R– transplants, without immunosuppression, were fixed at posttransplant d 7 and stained with IgG-AF488 (A) or C3-FITC (C). Kidneys from nontransplant C57BL/6 mice (“No Tx”) served as staining controls (left panels). Representative images of cortex (upper panels) and medulla (lower panels) are shown, with brightfield overlay to orient tissue morphology (60×; white bar = 50 µm). (B, D) Fluorescence immunostaining for IgG (B) or C3 (D) was quantitated for 10 HPFs per kidney using Image J software. Average positive staining was calculated for each graft and compared among the groups (n = 3/group). *P < 0.05. AF488, AlexaFluor-488; CMV, cytomegalovirus; FITC, fluorescein isothiocyanate; HPF, high-power field; IgG, immunoglobulin G; MCMV, murine cytomegalovirus; Tx, transplant.
The detection of the complement split product C4d is a diagnostic criterion for the histopathologic assignment of AMR in clinical transplant biopsies.55,58 However, attempts to quantify C4 deposition in the murine allografts by Western blotting and immunofluorescent staining were not successful. Therefore, immunofluorescent staining for the complement split product C3 was performed (Figure 1C), which showed no staining in control (nontransplant) B6 kidneys, minimal staining in D–/R– grafts, and greater staining in the D+/R– grafts (5 ± 3 spots/HPF versus 45 ± 12 spots/HPF, P = 0.0457) (Figure 1D), with more abundant staining in the medulla (lower panel) compared with the cortex (upper panel). Together, results of immunostaining for IgG and C3 indicate that this model of AR includes findings consistent with AMR in the MCMV-infected allografts.
Passively Transferred Antibodies From MCMV Immune Sera Deposit Into MCMV-infected Allografts of µMT Mice
Although IgG deposition was evident in D+/R– transplants, immunofluorescent staining could not distinguish whether these were alloantibodies or antiviral antibodies. To investigate this question, µMT recipients were used to prevent alloantibody generation after transplant.50 Pooled sera from MCMV-uninfected (MCMV– Ig) or MCMV-infected (MCMV+ Ig) B6 mice were compared by ELISA against MCMV-infected MEF cell lysates and confirmed the presence of anti-MCMV antibodies in pooled B6 MCMV+ Ig (Figure 2A). Binding characteristics of these sera to syngeneic or allogeneic target cells ex vivo were shown to be similar in later experiments, indicating no significant differences in autoantibody or alloantibody quantities between these pooled sera. D+/R– transplants using µMT recipients without immunosuppression were treated with either no sera (“no Ig”), sera from uninfected B6 mice without MCMV antibodies (MCMV– Ig), or sera from MCMV-infected B6 mice (MCMV+ Ig) at posttransplant day 1 and euthanized at either posttransplant day 3 or 14.
FIGURE 2.: Adoptive transfer of MCMV immune or nonimmune sera into D
+/R
– µMT transplants. A, Sera from uninfected (MCMV
– Ig) or MCMV-infected (MCMV
+ Ig) C57BL/6 mice were pooled and anti-MCMV antibody titers measured by ELISA at 450 nm OD. (B–I) D
+/R
– transplants were performed using µMT recipients, which were treated posttransplant with no Ig, MCMV
– Ig, or MCMV
+ Ig (n = 3/experimental group). (B, C) Western blots of allograft lysates at posttransplant d 3 were probed using anti-mouse IgG and anti-actin antibodies (B), and the IgG:actin ratio quantitated (C) using Image J. (D–G) Allografts from µMT recipients at posttransplant d 3 were stained for IgG-AF488 (D) or C3-FITC (F). Representative images of cortex (upper panels) and medulla (lower panels) are shown, with brightfield overlay to orient tissue morphology (60×; white bar = 50 µm). Immunostaining for IgG (E) or C3 (G) was quantitated using Image J as in
Figure 1, and compared between groups receiving no Ig, MCMV
– Ig, or MCMV
+ Ig. (H, I) Allografts at d 14 posttransplant were stained using hematoxylin and eosin (H) and damage scores quantitated (I). *
P < 0.05; **
P < 0.01; ***
P < 0.001. AF488, AlexaFluor-488; FITC, fluorescein isothiocyanate; HPF, high-power field; IgG, immunoglobulin G; MCMV, murine cytomegalovirus; OD, optical density; µMT, B cell–deficient; TEC, tubular epithelial cell.
At day 3, allograft lysates were analyzed by Western blotting for IgG deposition (Figure 2B) and showed no intragraft IgG in µMT recipients receiving no Ig or MCMV– Ig, whereas MCMV+ Ig-treated µMT recipients had significantly greater intragraft IgG deposition by densitometry (Figure 2C, P = 0.0107). Intragraft IgG-AF488 immunostaining was absent in µMT recipients receiving no sera (Figure 2D, left column), sparsely detected in allografts treated with MCMV– Ig (Figure 2D, middle column), and detected in allografts from MCMV+ Ig-treated µMT recipients (Figure 2D, right column) at significantly higher levels (Figure 2E, P = 0.0016). C3-FITC immunostaining showed similar results (Figure 2F), with MCMV+ Ig-treated allografts showing greater staining compared with the other groups (Figure 2G, P = 0.0005). Allograft histopathology scores at day 14 posttransplant (Figure 2H and I) showed similar damage scores for allografts treated with no Ig or MCMV– Ig (8.75 ± 0.25 versus 9.17 ± 0.17; P = NS) but more severe damage in MCMV+ Ig-treated grafts (11.00 ± 0.41, P = 0.0081). Together, these findings indicate that the deposition of antibodies derived from B6 MCMV+ Ig is sufficient to result in complement component C3 fixation in MCMV-infected allografts and is associated with more severe histopathologic allograft injury.
IgG Deposition in Immunosuppressed Mice
To test whether immunosuppression influences the intragraft deposition of MCMV-induced antibodies, day 14 allografts from cyclosporine-treated D–/R– and D+/R– transplants were stained for IgG (Figure 3A) and showed less abundant IgG deposition compared with nonimmunosuppressed mice at day 7 (Figure 1), indicating that humoral responses were impaired by cyclosporine. Next, IgG immunostaining was examined in a cohort of immunosuppressed MCMV D+/R+ transplants (Figure 3A), so that recipients had pretransplant anti-MCMV immunity and showed significantly greater IgG deposition compared with D–/R– and D+/R– transplants (Figure 3B, P = 0.0002). Similar to transplants without immunosuppression, immunostaining was more abundant in the medulla (lower panels) compared with the cortex (upper panels). Staining controls of nontransplant B6 kidneys showed no IgG immunostaining (Figure 3A, left panel). Sera from immunosuppressed D–/R–, D+/R–, and D+/R+ recipients were analyzed by ELISA for anti-MCMV antibodies (Figure 3C), with pooled B6 MCMV– Ig and MCMV+ Ig included as negative and positive controls. Anti-MCMV antibodies were less abundant in D+/R– mice compared with both D+/R+ and MCMV+ Ig (P < 0.0001). A positive correlation was observed between the ELISA optical density of D+/R– or D+/R+ sera and the quantity of intragraft IgG immunostaining for D+/R– and D+/R+ transplants (R2 = 0.82) (Figure 3D), indicating that the intensity of antibody deposition in allografts correlated with quantities of anti-MCMV antibodies in recipient blood.
FIGURE 3.: IgG deposition in allografts with immunosuppression. A, Allografts of cyclosporine-immunosuppressed MCMV D–/R–, D+/R–, and D+/R+ recipients were stained with IgG-AF488 at d 14 posttransplant (n = 5–6/group). Kidneys from nontransplant C57BL/6 mice (“No Tx”) served as staining controls (left panels, n = 3). Representative images of cortex (upper panel) and medulla (lower panel) are shown (40×; white bar = 50 µm). B, Fluorescence immunostaining was quantitated for 10 HPF per kidney using Image J and compared between groups. C, MCMV– Ig, MCMV+ Ig, or sera from immunosuppressed D–/R–, D+/R–, or D+/R+ transplant recipients were analyzed by ELISA for anti-MCMV antibodies at OD450 (OD). D, Serological anti-MCMV antibody levels (OD) were correlated with allograft IgG immunostaining for D+/R– and D+/R+ transplant recipients. *P < 0.05; ***P < 0.001. AF488, AlexaFluor-488; HPF, high-power field; IgG, immunoglobulin G; MCMV, murine cytomegalovirus; OD, optical density; Tx, transplant.
Virus-induced Antibodies Bind to Infected Renal Tubular Epithelial Cells In Vitro
CMV can infect renal TECs.59-62 Because IgG deposition was observed by immunostaining in the medulla, IgG binding to MCMV-infected renal TECs was examined in vitro (Figure 4A). B6 or BALB/c TECs were either uninfected (“MCMV– TECs”) or infected with MCMV-GFP (“MCMV+ TECs”), then incubated with either MCMV– sera or MCMV+ sera from nontransplant B6 mice, or with pooled sera from cyclosporine-immunosuppressed D–/R– or D+/R+ transplant recipients. MCMV+ TECs were identified by flow cytometry via GFP+ expression among EpCAM+ cells, and IgG binding was identified by staining with anti-mouse IgG-allophycocyanin (Figure 4A; Figure S1, SDC, https://links.lww.com/TP/C62).
FIGURE 4.: Antibody binding and CDC of MCMV-infected target cells. A, Uninfected or MCMV-GFP–infected B6 or BALB/c renal TECs were incubated either with sera from MCMV nonimmune B6 mice (MCMV–, red lines) or from MCMV immune (MCMV+) B6 mice, D–/R– or D+/R+ recipients (blue lines). IgG binding to target cells was identified by staining with anti-mouse IgG-APC antibodies and detection by flow cytometry. Histogram subgating is shown with gray bars. (B, C) MFI of IgG binding was compared among groups for both uninfected (white bars) and MCMV-infected (gray bars) B6 (B) or BALB/c (C) TEC-target cells. (D, E) A CDC assay using uninfected (D) or MCMV-GFP–infected (E) BALB/c 3T3 target cells was performed by incubation of target cells with MCMV–/+ sera or transplant sera as in (A), in the presence of rabbit complement for 3 h. Cytotoxicity was quantitated using 7-AAD staining. Cells incubated without sera or complement (–sera/–compl) or with complement alone (+compl alone) were included as controls. (F, G) CDC assay was repeated using MCMV–/+ sera and BALB/c TEC-target cells, as described for (D, E). (A–G) All experiments were performed 3 times, and representative experiments are shown. *P < 0.05; **P < 0.01; ***P < 0.001. APC, allophycocyanin; CDC, complement-dependent cytotoxicity; GFP, green fluorescent protein; IgG, immunoglobulin G; MCMV, murine cytomegalovirus; MFI, mean fluorescence intensity; TEC, tubular epithelial cell.
In Figure 4A (top row), B6 MCMV– sera (red line) and MCMV+ sera (blue line) bound similarly to MCMV-uninfected B6 and BALB/c TECs (“MCMV-TECs”), indicating that antibodies recognizing nonviral TEC antigens (autoantibodies and alloantibodies) were present at similar quantities in nontransplant B6 MCMV+/– sera. MCMV– sera (red) bound at a similar mean fluorescence intensity (MFI) to both MCMV-uninfected and MCMV-infected B6 and BALB/c TECs (Figure 4A, top row; Figure 4B), indicating that MCMV infection of TECs did not upregulate antigens that significantly altered binding by antibodies found in MCMV– sera. In contrast, MCMV+ sera bound to MCMV-infected B6 and BALB/c TECs at a significantly higher MFI than the MCMV– sera (Figure 4A, top row; Figure 4B, gray bars; P ≤ 0.0001). Because MCMV– and MCMV+ sera bound similarly to uninfected B6 and BALB/c TECs, this difference was unlikely to be due to increased autoantibodies or alloantibodies in MCMV+ sera. As these pooled sera were used for the adoptive transfer to µMT recipients (Figure 2), it is likely that the difference in antibody deposition observed in µMT recipients treated with MCMV immune sera was not due to autoantibodies or alloantibodies, which were similarly present in MCMV+/– sera.
Sera from immunosuppressed D–/R– transplants bound to MCMV-uninfected B6 and BALB/c TECs at similar MFI as MCMV– sera (Figure 4A, middle row; Figure 4B, white bars). Binding of D–/R– sera to MCMV-infected B6 and BALB/c TECs resembled that of MCMV– sera, except that D–/R– sera showed a second binding peak with an MFI significantly higher than MCMV– sera (P < 0.01). D+/R+ sera had significantly greater binding to MCMV-infected B6 and BALB/c TECs compared with D–/R– sera (Figure 4A, bottom row; Figure 4B, gray bars, P ≤ 0.01). D+/R+ sera bound to uninfected B6 and BALB/c TECs similarly to nontransplant MCMV– sera (Figure 4B, white bars), indicating no significant difference in autoreactive and alloreactive antibody quantities. However, D+/R+ binding to MCMV-infected BALB/c TECs did show a small peak at ≥103 fluorescence intensity, similar to D–/R– sera (Figure 4A, right column). Taken together, these results indicate that MCMV+ and D+/R+ sera contain virus-induced antibodies, not present in MCMV– or D–/R– sera, that bind to antigens expressed by MCMV-infected but not uninfected TECs.
Antibodies in MCMV+ and D+/R+ Sera Induce Complement-dependent Cytotoxicity
Next, CDC assays were performed using BALB 3T3 cells, with or without MCMV infection, by incubation with MCMV– or MCMV+ B6 sera in the presence or absence of rabbit complement, and cytotoxicity was quantified by 7-AAD staining. Control cells were incubated without sera or complement (–sera/–compl) or with complement alone (+compl alone). B6 MCMV– and MCMV+ sera without complement did not induce cytotoxicity above control wells (data not shown). In the presence of complement (Figure 4D and E), MCMV– sera did not induce cytotoxicity above that observed in control wells, whereas MCMV+ sera induced cytotoxicity of MCMV+ cells (Figure 4E) but not MCMV– cells (Figure 4D). Similarly, D–/R– sera did not induce cytotoxicity, whereas D+/R+ sera induced cytotoxicity only in MCMV+ cells (Figure 4D and E). The CDC assay was repeated using BALB/c TECs with similar results for MCMV+ and MCMV– sera (Figure 4F and G), but insufficient D–/R– and D+/R+ sera were available to perform the assay using BALB/c TECs. However, in aggregate, these results indicate that MCMV+ and D+/R+ sera can induce CDC of MCMV-infected 3T3 cells or TECs in vitro.
Native Kidneys of D+/R+ Transplant Recipients Lack IgG Deposition
The renal transplants in this model are implanted heterotopically, retaining 1 or both native kidneys. For R+ recipients, the native kidneys harbor MCMV before transplant. To examine whether MCMV antibodies deposit into the native kidneys of D+/R+ transplant recipients, IgG staining of R+ native kidneys was compared with the D+ allografts (Figure 5A and B) and showed minimal IgG deposition (Figure 5C, P < 0.0001). Histopathology (Figure 5D–F) showed significantly greater damage scores for allografts compared with native kidneys (12.00 ± 0.71 versus 0.40 ± 0.40, P < 0.0001). Similarly, CD45+ cell infiltrates were significantly greater in allografts compared with native kidneys (Figure 5G, P = 0.0002). In contrast, allografts and native kidneys had similar viral loads by quantitative DNA PCR (Figure 5H). Taken together, these results indicate that R+ native kidneys lack significant antibody deposition and cellular infiltrates despite MCMV viral loads comparable with those found in the contralateral allografts, suggesting that the presence of MCMV in the native kidney is not sufficient to elicit intrarenal IgG deposition.
FIGURE 5.: IgG immunostaining of D
+/R
+ native kidneys and D
+/R
– allografts treated with GCV. (A, B) Allografts and native kidneys of D
+/R
+ recipients were stained with IgG-AF488 as in
Figure 1. (C) IgG immunostaining was quantitated for 10 HPF and compared between Tx and native kidneys (n = 4/group). (D–F) Hematoxylin and eosin staining of allografts (D) and native kidneys (E) was scored for organ injury, and damage scores (F) compared between Tx and native kidneys. G, MCMV viral loads in Tx and native kidneys were compared by quantitative DNA PCR. H, CD45
+ cell infiltrates in Tx and native kidneys were quantified by flow cytometry. (I–K) D
+/R
– recipients were treated with GCV for 14 d (I), or for 14 d followed by 7 d without antiviral treatment (J). Allografts were stained for IgG-AF488 (I, J) and quantitated (K) using Image J in comparison with IgG staining of d 14 D
+/R
– allografts without GCV treatment (
Figure 3). (L–N) D
+/R
+ syngeneic grafts and native kidneys of immunosuppressed recipients were stained for IgG and quantitated using Image J (n = 3). (A–J) Representative images are shown (40×, bar = 50 µm). **
P < 0.01; ***
P < 0.001. AF488, AlexaFluor-488; GCV, ganciclovir; HPF, high-power field; IgG, immunoglobulin G; MCMV, murine cytomegalovirus; PCR, polymerase chain reaction; Tx, transplant.
Ganciclovir Reduces Intragraft IgG Deposition
GCV treatment of D+/R– transplant recipients inhibits viral replication and reduces intragraft leukocyte infiltrates.26 To determine whether GCV treatment affects intragraft IgG deposition, allografts from D+/R– recipients treated with GCV for 14 days were stained for IgG (Figure 5I), as well as grafts from mice euthanized 7 days after discontinuing GCV (Figure 5J). Allografts from GCV-treated mice showed lower IgG staining for both conditions, compared with untreated mice (Figure 5K, P = 0.0002). This result suggests that pharmacological inhibition of viral antigen expression may reduce IgG deposition within MCMV-infected allografts either via inhibition of primary antiviral antibody generation or by reduced binding of antiviral antibodies in the absence of viral antigen expression.
IgG Deposits Into Syngeneic D+/R+ Transplants
Transplant IRI can induce MCMV reactivation in syngeneic D+ grafts of immunosuppressed mice.31 To induce MCMV reactivation while eliminating the contribution of alloantibodies in our model, syngeneic D+/R+ transplants were performed using B6 donors and recipients with cyclosporine immunosuppression. IgG immunostaining in the syngeneic grafts was significantly greater than that observed in the native kidneys (P = 0.0023). This result indicates that antiviral antibodies, in the absence of alloantibodies, can deposit into D+ kidneys of immunosuppressed R+ recipients after IRI.
DISCUSSION
In this study, a murine renal transplant model was used to determine whether pathogen-induced antibodies could deposit within allografts during AR, using MCMV as a model pathogen. MCMV naturally infects the kidney and reactivates after allogeneic transplantation.28,29,63 In this model, IgG and C3 got deposited into MCMV-infected allografts during AR, consistent with AMR, and were recapitulated by passive transfer of MCMV immune sera in µMT mice. In vitro, MCMV+ and D+/R+ sera could induce CDC of MCMV-infected renal TECs. These data suggest that MCMV-induced antibodies might deposit into infected allografts during AR and may be capable of triggering CDC. However, MCMV-induced antibodies did not deposit into MCMV-infected native kidneys, suggesting that intrarenal MCMV infection alone is not sufficient to induce antibody deposition in the kidney. Together, these data suggest the possibility that pathogen-induced antibodies might be capable of contributing to acute AMR.
HCMV positive serostatus and infection (DNAemia) are risk factors for late allograft dysfunction and loss via “indirect effects.” Although CMV-induced antibodies protect the host against CMV disease, their potential role in allograft injury has not been previously examined. Our prior work in this transplant model showed that other cell types that are important for control of MCMV infection, such as NK cells and CD8+ T cells, can also induce allograft injury during AR.26,32 The current study extends the evidence supporting a model that immune mechanisms necessary for the protection of the host against systemic infection might, conversely, contribute to injury of the CMV-infected allograft. As anti-CMV antibody titers vary in kidney transplant recipients depending on their history of viral reactivation,49 future work could examine the possibility that these varying antibody titers might perhaps contribute to varying degrees of allograft dysfunction among HCMV-infected patients.
In this study, MCMV-induced antibodies were shown to be capable of fixing complement and lysing MCMV-infected TECs in vitro, suggesting that virus-induced antibodies may participate in CDC during AMR. Further studies to investigate CDC in vivo could define the functional relevance of this pathway within rejecting allografts. In addition, pathogen-induced IgM antibodies were not analyzed in this work and deserve further study to define their role in complement fixation and CDC during AMR. Our prior work in this model also showed that NK cells are recruited to MCMV-infected allografts and manifest both cytolytic and cytotoxic phenotypes, raising the untested possibility that virus-induced antibodies might also contribute to NK cell–mediated antibody-dependent cellular cytotoxicity. Further studies in animal models and clinical populations could examine whether CDC or antibody-dependent cellular cytotoxicity triggered by pathogen-induced antibodies might constitute possible mechanisms contributing to the “indirect effects” of CMV in renal transplantation.
In this animal model, MCMV-induced antibodies were identified in allografts but not native kidneys of D+/R+ transplants, despite the presence of similar quantities of CMV DNA in both organs. A possible explanation for this observation may be that CMV reactivation in allografts during AR results in the expression of viral antigens recognized by virus-induced antibodies. Consistent with this interpretation, inhibition of viral replication via GCV treatment was associated with a lack of intragraft antibody deposition, although this result could also have been due to inhibition of primary antiviral antibody generation in R– recipients by suppressing viral antigen expression. These observations align with clinical studies showing that HCMV antigens can be identified in acutely rejecting allografts and that antiviral drugs used to prevent viral reactivation are associated with improved graft longevity among HCMV-infected renal transplant patients.17,19,20 Other groups have also shown that reactivation of latent CMV from murine renal allografts can be induced by inflammatory cytokines such as tumor necrosis factor-α that are expressed after allogeneic transplantation, as well as by IRI after syngeneic transplantation.29,31 Zhang et al31 also showed that immunosuppression alone failed to induce reactivation from latency in native kidneys of nontransplant mice, supporting the possibility that viral antigens might not be expressed in the native kidney of the immunosuppressed transplant recipient, and is consistent with our observations in the R+ native kidneys. However, a limitation of this study is that we were not able to characterize MCMV antigen expression in allografts and native kidneys. Unfortunately, attempts to quantify MCMV antigen expression in this renal transplant model using commercially available mouse anti-MCMV antibodies (Center for Proteomics, Rijeka Croatia) were unsuccessful because of high background staining (data not shown).
Another limitation to the interpretation of this study is the possibility that the antibodies observed in the MCMV-infected allografts were directed against nonviral antigens. In this work, the MCMV immune and nonimmune sera used for adoptive transfer in the µMT mice showed similar quantities of autoreactive and alloreactive antibodies (measured by ex vivo binding to syngeneic [B6] and allogeneic (BALB/c) TECs), which would be predicted to deposit similarly into allografts. Because intragraft IgG staining was more intense after adoptive transfer of MCMV immune sera compared with nonimmune sera, these data indirectly support the likelihood that virus-directed antibodies, rather than autoantibodies or alloantibodies, comprised the observed differential staining. However, an alternative explanation might be that MCMV-infected cells within rejecting allografts could upregulate the expression of antigens that bind yet-uncharacterized nonviral antibodies that might be present only in MCMV immune sera. In vitro, MCMV infection does alter numerous transcriptional pathways that may affect host cellular protein expression by infected cells.64 MCMV infection also induces expression of interferon-regulated genes and other cell stress–response genes.65 It is, therefore, also possible that intragraft MCMV-infected cells might express stress-induced molecules that promote nonviral antibody binding and complement activation; however, this pathway would be predicted to induce deposition of nonviral antibodies similarly by MCMV nonimmune and immune sera in µMT mice. Finally, in a syngeneic model with intragraft MCMV reactivation induced by IRI, IgG deposition was observed in the absence of alloantigens, supporting that the intragraft IgGs are antiviral antibodies and not alloantibodies.
In summary, in this murine model, complement-fixing antibodies are localized to MCMV-infected acutely rejecting allografts. In vitro, MCMV-induced antibodies could induce CDC of infected renal TECs. The relevance of these findings may merit further exploration in the animal model and clinical populations to elucidate the “indirect effects” of CMV in renal transplantation.
REFERENCES
1. Freeman RB Jr. The ‘indirect’ effects of cytomegalovirus infection. Am J Transplant. 2009;9:2453–2458.
2. Tong CY, Bakran A, Peiris JS, et al. The association of viral infection and chronic allograft nephropathy with graft dysfunction after renal transplantation. Transplantation. 2002;74:576–578.
3. Streblow DN, Orloff SL, Nelson JA. Acceleration of allograft failure by cytomegalovirus. Curr Opin Immunol. 2007;19:577–582.
4. Legendre C, Pascual M. Improving outcomes for solid organ transplant recipients at risk from cytomegalovirus infection: late onset disease and indirect consequences. Clin Infect Dis. 2008;46:732–740.
5. Einollahi B, Motalebi M, Salesi M, et al. The impact of cytomegalovirus infection on new-onset diabetes mellitus after kidney transplantation: a review on current findings. J Nephropathol. 2014;3:139–148.
6. Rubin RH, Tolkoff-Rubin NE, Oliver D, et al. Multicenter seroepidemiologic study of the impact of cytomegalovirus infection on renal transplantation. Transplantation. 1985;40:243–249.
7. Schnitzler MA, Woodward RS, Brennan DC, et al. The effects of cytomegalovirus serology on graft and recipient survival in cadaveric renal transplantation: implications for organ allocation. Am J Kidney Dis. 1997;29:428–434.
8. Schnitzler MA, Woodward RS, Brennan DC, et al. Impact of cytomegalovirus serology on graft survival in living related kidney transplantation: implications for donor selection. Surgery. 1997;121:563–568.
9. Humar A, Gillingham KJ, Payne WD, et al. Association between cytomegalovirus disease and chronic rejection in kidney transplant recipients. Transplantation. 1999;68:1879–1883.
10. Browne BJ, Young JA, Dunn TB, et al. The impact of cytomegalovirus infection ≥1 year after primary renal transplantation. Clin Transplant. 2010;24:572–577.
11. Erdbruegger U, Scheffner I, Mengel M, et al. Impact of CMV infection on acute rejection and long-term renal allograft function: a systematic analysis in patients with protocol biopsies and indicated biopsies. Nephrol Dial Transplant. 2012;27:435–443.
12. López-Oliva MO, Flores J, Madero R, et al. Cytomegalovirus infection after kidney transplantation and long-term graft loss. Nefrologia. 2017;37:515–525.
13. Selvey LA, Lim WH, Boan P, et al. Cytomegalovirus viraemia and mortality in renal transplant recipients in the era of antiviral prophylaxis. Lessons from the western Australian experience. BMC Infect Dis. 2017;17:501.
14. Blazquez-Navarro A, Dang-Heine C, Wittenbrink N, et al. BKV, CMV, and EBV interactions and their effect on graft function one year post-renal transplantation: results from a large multi-centre study. Ebiomedicine. 2018;34:113–121.
15. Markovic-Lipkovski J, Müller CA, Engler-Blum G, et al. Human cytomegalovirus in rejected kidney grafts: detection by polymerase chain reaction. Nephrol Dial Transplant. 1992;7:865–870.
16. Gerstenkorn C, Robertson H, Mohamed MA, et al. Detection of cytomegalovirus (CMV) antigens in kidney biopsies and transplant nephrectomies as a marker for renal graft dysfunction. Clin Chem Lab Med. 2000;38:1201–1203.
17. Holma K, Törnroth T, Grönhagen-Riska C, et al. Expression of the cytomegalovirus genome in kidney allografts during active and latent infection. Transpl Int. 2000;13(suppl 1):S363–S365.
18. Opelz G, Döhler B, Ruhenstroth A. Cytomegalovirus prophylaxis and graft outcome in solid organ transplantation: a collaborative transplant study report. Am J Transplant. 2004;4:928–936.
19. Reischig T, Hribova P, Jindra P, et al. Long-term outcomes of pre-emptive valganciclovir compared with valacyclovir prophylaxis for prevention of cytomegalovirus in renal transplantation. J Am Soc Nephrol. 2012;23:1588–1597.
20. Reischig T, Kacer M, Hruba P, et al. Less renal allograft fibrosis with valganciclovir prophylaxis for cytomegalovirus compared to high-dose valacyclovir: a parallel group, open-label, randomized controlled trial. BMC Infect Dis. 2018;18:573.
21. Witzke O, Nitschke M, Bartels M, et al. Valganciclovir prophylaxis versus preemptive therapy in cytomegalovirus-positive renal allograft recipients: long-term results after 7 years of a randomized clinical trial. Transplantation. 2018;102:876–882.
22. Koskinen PK, Yilmaz S, Kallio E, et al. Rat cytomegalovirus infection and chronic kidney allograft rejection. Transpl Int. 1996;9(suppl 1):S3–S4.
23. Lautenschlager I, Soots A, Krogerus L, et al. CMV increases inflammation and accelerates chronic rejection in rat kidney allografts. Transplant Proc. 1997;29:802–803.
24. van Dam JG, Li F, Yin M, et al. Effects of cytomegalovirus infection and prolonged cold ischemia on chronic rejection of rat renal allografts. Transpl Int. 2000;13:54–63.
25. Inkinen K, Holma K, Soots A, et al. Expression of TGF-beta and PDGF-AA antigens and corresponding mRNAs in cytomegalovirus-infected rat kidney allografts. Transplant Proc. 2003;35:804–805.
26. Shimamura M, Saunders U, Rha B, et al. Ganciclovir transiently attenuates murine cytomegalovirus-associated renal allograft inflammation. Transplantation. 2011;92:759–766.
27. Pascher A, Proesch S, Pratschke J, et al. Rat cytomegalovirus infection interferes with anti-CD4 mAb-(RIB 5/2) mediated tolerance and induces chronic allograft damage. Am J Transplant. 2006;6:2035–2045.
28. Klotman ME, Starnes D, Hamilton JD. The source of murine cytomegalovirus in mice receiving kidney allografts. J Infect Dis. 1985;152:1192–1196.
29. Hummel M, Zhang Z, Yan S, et al. Allogeneic transplantation induces expression of cytomegalovirus immediate-early genes in vivo: a model for reactivation from latency. J Virol. 2001;75:4814–4822.
30. Hummel M, Kurian SM, Lin S, et al. Intragraft TNF receptor signaling contributes to activation of innate and adaptive immunity in a renal allograft model. Transplantation. 2009;87:178–188.
31. Zhang Z, Qiu L, Yan S, et al. A clinically relevant murine model unmasks a “two-hit” mechanism for reactivation and dissemination of cytomegalovirus after kidney transplant. Am J Transplant. 2019;19:2421–2433.
32. Shimamura M, Seleme MC, Guo L, et al. Ganciclovir prophylaxis improves late murine cytomegalovirus-induced renal allograft damage. Transplantation. 2013;95:48–53.
33. Singh N, Pirsch J, Samaniego M. Antibody-mediated rejection: treatment alternatives and outcomes. Transplant Rev (Orlando). 2009;23:34–46.
34. Pouliquen E, Koenig A, Chen CC, et al. Recent advances in renal transplantation: antibody-mediated rejection takes center stage. F1000Prime Rep. 2015;7:51.
35. Randhawa P. T-cell-mediated rejection of the kidney in the era of donor-specific antibodies: diagnostic challenges and clinical significance. Curr Opin Organ Transplant. 2015;20:325–332.
36. Halloran PF, Venner JM, Madill-Thomsen KS, et al. Review: the transcripts associated with organ allograft rejection. Am J Transplant. 2018;18:785–795.
37. Chong AS, Rothstein DM, Safa K, et al. Outstanding questions in transplantation: B cells, alloantibodies, and humoral rejection. Am J Transplant. 2019;19:2155–2163.
38. Delville M, Lamarthée B, Pagie S, et al. Early acute microvascular kidney transplant rejection in the absence of anti-HLA antibodies is associated with preformed IgG antibodies against diverse glomerular endothelial cell antigens. J Am Soc Nephrol. 2019;30:692–709.
39. Parajuli S, Redfield RR, Garg N, et al. Clinical significance of microvascular inflammation in the absence of anti-HLA DSA in kidney transplantation. Transplantation. 2019;103:1468–1476.
40. Senev A, Otten HG, Kamburova EG, et al. Antibodies against ARHGDIB and ARHGDIB gene expression associate with kidney allograft outcome. Transplantation. 2020;104:1462–1471.
41. Costa C, Touscoz GA, Bergallo M, et al. Non-organ-specific and anti-endothelial antibodies in relation to CMV infection and acute rejection in renal transplant recipients. Clin Transplant. 2010;24:488–492.
42. Philogene MC, Jackson AM. Non-HLA antibodies in transplantation: when do they matter? Curr Opin Organ Transplant. 2016;21:427–432.
43. Zhang Q, Reed EF. The importance of non-HLA antibodies in transplantation. Nat Rev Nephrol. 2016;12:484–495.
44. Cardinal H, Dieudé M, Hébert MJ. The emerging importance of non-HLA autoantibodies in kidney transplant complications. J Am Soc Nephrol. 2017;28:400–406.
45. Kamburova EG, Gruijters ML, Kardol-Hoefnagel T, et al. Antibodies against ARHGDIB are associated with long-term kidney graft loss. Am J Transplant. 2019;19:3335–3344.
46. Lefaucheur C, Viglietti D, Bouatou Y, et al. Non-HLA agonistic anti-angiotensin II type 1 receptor antibodies induce a distinctive phenotype of antibody-mediated rejection in kidney transplant recipients. Kidney Int. 2019;96:189–201.
47. Philogene MC, Johnson T, Vaught AJ, et al. Antibodies against angiotensin II type 1 and endothelin A receptors: relevance and pathogenicity. Hum Immunol. 2019;80:561–567.
48. Farouk S, Zhang Z, Menon MC. Non-HLA donor-recipient mismatches in kidney transplantation-A stone left unturned. Am J Transplant. 2020;20:19–24.
49. Iglesias-Escudero M, Moro-García MA, Marcos-Fernández R, et al. Levels of anti-CMV antibodies are modulated by the frequency and intensity of virus reactivations in kidney transplant patients. PLoS One. 2018;13:e0194789.
50. Kitamura D, Roes J, Kühn R, et al. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin mu chain gene. Nature. 1991;350:423–426.
51. Brune W, Ménard C, Heesemann J, et al. A ribonucleotide reductase homolog of cytomegalovirus and endothelial cell tropism. Science. 2001;291:303–305.
52. Bubić I, Wagner M, Krmpotić A, et al. Gain of virulence caused by loss of a gene in murine cytomegalovirus. J Virol. 2004;78:7536–7544.
53. Li M, Boddeda SR, Chen B, et al. NK cell and Th17 responses are differentially induced in murine cytomegalovirus infected renal allografts and vary according to recipient virus dose and strain. Am J Transplant. 2018;18:2647–2662.
54. Zhang Z, Schlachta C, Duff J, et al. Improved techniques for kidney transplantation in mice. Microsurgery. 1995;16:103–109.
55. Roufosse C, Simmonds N, Clahsen-van Groningen M, et al. A 2018 reference guide to the Banff classification of renal allograft pathology. Transplantation. 2018;102:1795–1814.
56. Zhao Y, Zhao H, Zhang Y, et al. Isolation and epithelial co-culture of mouse renal peritubular endothelial cells. BMC Cell Biol. 2014;15:40.
57. Bantug GR, Cekinovic D, Bradford R, et al. CD8+ T lymphocytes control murine cytomegalovirus replication in the central nervous system of newborn animals. J Immunol. 2008;181:2111–2123.
58. Haas M, Loupy A, Lefaucheur C, et al. The Banff 2017 Kidney Meeting Report: revised diagnostic criteria for chronic active T cell-mediated rejection, antibody-mediated rejection, and prospects for integrative endpoints for next-generation clinical trials. Am J Transplant. 2018;18:293–307.
59. Heieren MH, Kim YK, Balfour HH Jr. Human cytomegalovirus infection of kidney glomerular visceral epithelial and tubular epithelial cells in culture. Transplantation. 1988;46:426–432.
60. Ustinov JA, Loginov RJ, Mattila PM, et al. Cytomegalovirus infection of human kidney cells in vitro. Kidney Int. 1991;40:954–960.
61. Hendrix RM, Wagenaar M, Slobbe RL, et al. Widespread presence of cytomegalovirus DNA in tissues of healthy trauma victims. J Clin Pathol. 1997;50:59–63.
62. Shimamura M, Murphy-Ullrich JE, Britt WJ. Human cytomegalovirus induces TGF-β1 activation in renal tubular epithelial cells after epithelial-to-mesenchymal transition. PLoS Pathog. 2010;6:e1001170.
63. Pollock JL, Virgin HW IV. Latency, without persistence, of murine cytomegalovirus in the spleen and kidney. J Virol. 1995;69:1762–1768.
64. Juranic Lisnic V, Babic Cac M, Lisnic B, et al. Dual analysis of the murine cytomegalovirus and host cell transcriptomes reveal new aspects of the virus-host cell interface. PLoS Pathog. 2013;9:e1003611.
65. Marcinowski L, Lidschreiber M, Windhager L, et al. Real-time transcriptional profiling of cellular and viral gene expression during lytic cytomegalovirus infection. PLoS Pathog. 2012;8:e1002908.
