SEARCH

SEARCH BY CITATION

Keywords:

  • Alloantibodies;
  • antibody-mediated rejection;
  • C4d;
  • complement;
  • macrophages

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Antibody-mediated rejection of human cardiac transplants is correlated with C4d deposits and macrophage infiltrates in capillaries of endomyocardial biopsies. We produced an antibody to rat C4d to study C4d deposition and clearance in Lewis rats that were sensitized with a blood transfusion from DA rats 7, 14 or 21 days before cardiac transplantation. Cyclosporin A (CsA) immunosuppression was initiated after transplantation at a dose that inhibited graft rejection, antibody production and C4d deposition in unsensitized recipients. Blood transfusion elicited high levels of circulating IgG alloantibodies, predominantly of the complement-activating IgG2b subclass, that peaked 14 days after transplantation. At this time, macrophages accumulated in capillaries, and C4d deposits were diffuse and intense on arteries, capillaries and veins. Grafts that survived 90 days in sensitized recipients still had deposits of C4d that were associated with increased interstitial fibrosis and vasculopathy in arteries. Clearance of C4d was determined by retransplanting DA cardiac allografts from Lewis recipients back to DA recipients. C4d deposits were decreased to minimal levels within 5 days after retransplantation. Thus, C4d deposition is not limited to the capillaries, but extends throughout the arterial tree, and despite formation of a covalent bond, C4d is cleared within days.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The use of new markers of complement activation such as C4d has increased awareness of frequency of complement activation in transplants (1–8). C4d deposition is particularly prevalent in patients sensitized by previous transplants, pregnancies or blood transfusions. Although C4d has been widely studied in renal transplants, relatively few studies have examined the value of C4d as a correlate of antibody-mediated rejection in human cardiac transplants (9–11). In addition to C4d, Fishbein and coworkers (12) have emphasized an association between antibody-mediated rejection and accumulations of macrophages in capillaries of endomyocardial biopsies. However, criteria for the diagnosis of antibody-mediated rejection are still not established for clinical biopsies (13) and animal models have not tested C4d or macrophages as markers for antibody-mediated rejection.

Clinical endomyocardial biopsies are limited to sampling a small area on the right side of the interventricular septum. These biopsies capture primarily a network of capillaries with occasional small arterioles. As a result, the extent of complement deposition on larger arteries and the coronary arteries that are involved in chronic vasculopathy cannot be determined in endomyocardial biopsies. Arterial lesions have been studied in transplants to experimental animals, but appropriate reagents for studying C4d deposition in experimental animal models have not been available. Therefore, we have produced a polyclonal antibody to rat C4d.

C4d is highly homologous to C3d with over 35% shared amino acid sequence. One significant difference is an extra loop of 21 to 22 amino acids that is present in C4d but absent in C3d (14). This loop of amino acids is an ideal target for antibodies with specificity for C4d. Regele et al. (1) produced an antibody to a 15 amino acid long peptide from the carboxy end of the loop of human C4d. However, this loop has critical amino acid substitutions among species. For example, there are five amino acid differences between rat and human. Therefore, antibodies produced to peptide sequences of one species would not be expected to cross-react with the homologous peptide from other species. Indeed, we have found that the rabbit polyclonal antibody to human C4d is not useful for staining rat or mouse tissues. This is in distinction to rabbit polyclonal antibody to human C3d that we found cross-reacts strongly with deposits of rat C3d in organ allografts (15–18).

In this study, we characterize a rabbit polyclonal antibody to a 14 amino acid long peptide from the carboxy end of the loop of rat C4d. In addition, we have applied this antibody to study C4d deposition and clearance in cardiac transplants to rats presensitized with a blood transfusion. Finally, macrophage infiltration, fibrosis and arterial pathology were examined.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Animals

Male DA (RT1a) rats and Lewis (RT1l) rats weighing 200–225 g were used as donors and recipients, respectively. All rats were obtained from Harlan (Indianapolis, IN). All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86–23, revised 1985) and the Public Health Service Policy on Humane Care and Use of Laboratory Animals (1996).

Sensitization and immunosuppression

DA blood was collected from the tail vein into a heparinized syringe, and diluted 1:1 with sterile saline. One mL of this diluted blood was transfused into Lewis rats intravenously 7, 14 or 21 prior to transplantation. Control rats were transfused with 1 mL of saline. Cyclosporine (CsA) (Cyclosporine Injection; Ben Venue Laboratories, Inc., Bedford, OH) was injected subcutaneously three times per week. CsA was initiated at a dose of 10 mg/kg in the first week after transplantation and continued at a dose of 5 mg/kg until allograft hearts were recovered.

Heterotopic heart transplantation

DA hearts were transplanted into Lewis recipients under isoflurane (Abbott Laboratories, North Chicago, IL) inhalational anesthesia. Using a modification of the technique of Ono and Lindsey (19), the donor aorta and pulmonary artery were anastomosed to the abdominal aorta and inferior vena cava of the recipient, respectively, in an end-to-side fashion with 8–0 nylon suture (Ethicon, Inc., Somerville, NJ). Cardiac graft function was evaluated by abdominal palpation daily until rejection, which was defined as total cessation of contractions and was confirmed by direct visualization and histological examination of the allograft.

Measurement of alloantibodies

Alloantibodies were measured by flow cytometry on single-cell suspensions from cervical lymph nodes of donor strain rats as described previously (20). Briefly, 50 μL of aliquots containing 1.5 × 105 lymphocytes were incubated with 50 μL of diluted sera (1:4, 1:16, 1:64, 1:256) for 45 min in 4°C. The washed cells were reacted with 50 μL of phosphate buffered saline containing 1% bovine serum albumin and 0.02% NaN3 (PBA) containing a mixture of fluorescein isothiocynate-conjugated goat antibody specific for the Fc portion of rat IgG and PE-conjugated goat antibody specific for the mu chain of rat IgM (Jackson Immunoresearch Laboratories, West Grove, PA). After staining, the cells were washed, fixed and analyzed using FACScan flow cytometer (Becton Dickinson, Mountain View, CA). To analyze the IgG subclass of alloantibodies, cells were washed after the first incubation with diluted sera, and then reacted with 50 μL of PBA containing an optimal dilution of FITC-conjugated rat mAb against rat IgG1, IgG2a, IgG2b or biotin-conjugated rat mAb against rat IgG2c and streptavidin-fluorescein isothiocyanate (Pharmingen, San Diego, CA). After two washes, the cells were fixed in 1% formaldehyde in PBS, and analyzed using FACScan. The level of IgM, IgG and IgG subclass antibodies to class I MHC antigens on DA target cells is expressed as the mode channel fluorescent channel.

Development of polyclonal anti-rat C4d antibody

To develop a polyclonal peptide-specific anti-rat C4d antibody, a 14-mer peptide (STPAPRNPSEPVPQ) corresponding to amino acids 1223–1237 of rat C4 was synthesized. A cystein was added to the C-terminal to conjugate to keyhole limpet haemocyanin for immunization. Rabbits were immunized with 200 μg of the peptide dissolved in complete Freund's adjuvant and boosted five times in two-weekly intervals with 100 μg of the peptide in incomplete Freund's adjuvant. The rabbits were bled 7 days after the fifth booster and peptide-specific antibodies were isolated by affinity purification. .

Western blot of serum

The peptide-specific polyclonal rabbit antibody to rat C4d was tested for specificity by Western blot. Normal serum from a BN rat was diluted to 2% in reducing sample buffer, boiled and run on a 4–15% precast gradient Tris-HCl SDS-PAGE (Bio-Rad) for 1 h at 100V. The separated proteins were transferred to a Hybond nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ) and blocked for 3 h at 4°C in 1% BSA. The membrane was then incubated in affinity-purified rabbit antibodies to rat C4d at a 1:10 000 dilution overnight at 4°C. After extensive washes, the membrane was incubated with HRP-conjugated donkey anti-rabbit IgG (Jackson Immunoresearch Laboratories, West Grove, PA) diluted to 1:50 000. After washing, the membrane was incubated in ECL detection reagent (Amersham Biosciences, Piscataway, NJ) and exposed to Hyperfilm (Amersham Biosciences). The affinity-purified rabbit antibodies to rat C4d stained one major band of 93 kD in reduced serum on these blots. This band corresponds to the molecular weight of the C4 alpha chain, of which C4d is a segment (21).

Staining of rat C4d on sensitized, opsonized lymphocytes

Mononuclear cells were isolated from cervical lymph nodes of DA and Lew rats and 5 × 105 cells were incubated with serum from recipients of cardiac allografts or an IgM monoclonal antibody to rat RT1.A class I MHC (clone WFL3C6.1, ATCC, Rockville, MD). After washing, the cells were then incubated for 45 min at 37°C in a Gelatin Veronal Buffer Solution with Mg2+ and Ca2+ (GVB++) containing 10% serum from a C6 deficient PVG 1U rat (22). The cells were washed, and then incubated with phycoerythrin-conjugated mouse monoclonal antibody to rat CD3 (clone G4.18, BD Pharmingen, San Diego, CA) and our affinity-purified rabbit antibody to rat C4d. A secondary FITC-conjugated donkey antibody to rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) was used to stain for the affinity-purified rabbit antibody to rat C4d. After staining, the cells were fixed in 1% formalin solution and analyzed on a flow cytometer (FACscan, Becton Dickinson). C4d deposition was assessed on the CD3+ population using CellQuest software (Becton Dickinson, Mountain View, CA).

Staining of tissue for infiltrates and rat C4d

Full transverse segments of cardiac allograft tissue from presensitized LEW rats were fixed in acidic methanol (60% methanol, 10% acetic acid, 30% water), embedded in paraffin and sectioned at 7 micron thickness. Rejection was assessed on sections that were stained with hematoxylin and eosin. Fibrosis and collagen deposition was demonstrated with a Mallory trichrome stain. Neutrophils and macrophages were demonstrated by immunoperoxidase staining for myeloperoxidase (polyclonal rabbit anti-human MyeloPeroxidase, Biomeda, Foster City, CA) and CD68 (ED-1; Serotec Inc., Raleigh, NC), respectively. C4d deposition was localized by immunoperoxidase staining with our affinity-purified polyclonal rabbit antibody to C4d.

Single color fluorescent stains were performed on frozen sections using our affinity-purified polyclonal rabbit antibody to detect C4d or a polyclonal rabbit antibody to rat IgG followed by an FITC-conjugated donkey antibody to rabbit IgG. Two color fluorescent stains for C4d and IgG were performed on frozen sections. Our affinity-purified polyclonal rabbit antibody was used to detect C4d with a Rhodamine Red-conjugated donkey antibody to rabbit IgG. An FITC-conjugated donkey antibody was used to detect rat IgG. All of the labeled antibodies were F(ab′)2 antibodies with minimal cross-reactivity to other species (Jackson ImmunoResearch Laboratories, West Grove, PA).

To maximize objectivity of the data, the histological sections were evaluated by an experienced pathologist without knowledge of the other findings.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Alloantibody production and graft survival in unsensitized recipients

Lewis rats (n = 5) rejected cardiac transplants from DA donors between 6 and 7 days after transplantation in the absence of immunosuppression (Table 1). The low doses of CsA used in this study prolonged cardiac graft function in all 17 control, unsensitized recipients to over 90 days when the experiment was terminated for histological evaluation of the transplants.

Table 1. Survival of DA cardiac allografts in presensitized Lewis recipients
PretreatmentCsASurvival in days
  1. *CsA administered after transplantation.

  2. ** >14 and >90 = sacrificed with beating cardiac transplants.

NoneNone6 (×2), 7 (×3)
SalineYes*>14 (×8)**, >90 (×9)**
Blood 1 weekYes*>14 (×5), >90 (×3)
Blood 2 weeksYes*6, 12, 12, >14 (×3), >90 (×2)
Blood 3 weeksYes*>14 (×4), >90 (×3)

The low dose CsA treatment also limited circulating donor-specific alloantibodies (Figure 1). Both IgM and IgG alloantibodies were only transiently detected at very low levels by a sensitive flow cytometry assay on donor leukocytes.

image

Figure 1. Donor-specific IgM and IgG alloantibodies measured at a 1:64 dilution in the circulation of sensitized and nonsensitized recipients of cardiac allografts. Alloantibodies were measured at serial dilutions by flow cytometry on donor strain mononuclear cells (see Materials and Methods for details). A single blood transfusion from a DA rat administered 1, 2 or 3 weeks (bottom, middle and top panels, respectively) before transplantation elicited a transient IgM response (solid circles) followed by a persistent IgG response (solid squares). The donor-specific IgG alloantibody response was sustained in sensitized recipients after transplantation and CsA treatment were instituted (upward arrows indicate time of transplantation). In contrast, the same CsA treatment was sufficient to limit the IgM and IgG responses (open circles and squares, respectively) to extremely transient low levels in unsensitized recipients of allogeneic cardiac transplants. Vertical bars indicate 1 standard deviation.

Download figure to PowerPoint

Alloantibody production and graft survival in recipients sensitized with a blood transfusion

A single blood transfusion from male DA rats to male Lewis rats resulted in high levels of donor-specific alloantibodies in the circulation of all 23 experimental recipients (Figure 1). The seven recipients that received a blood transfusion 3 weeks prior to transplantation produced a transient IgM response that peaked 7 days after the transfusion and returned to near baseline by 3 weeks. The IgM response switched to IgG alloantibodies that peaked in the circulation 2 weeks after blood transfusion and were undiminished at 3 weeks. When CsA immunosuppression and a cardiac transplant from a DA donor were introduced 3 weeks after presensitization with a blood transfusion, circulating donor-specific IgG alloantibodies decreased transiently (Figure 1). Two weeks after transplantation, the donor-specific antibodies rebounded to pre-transplantation levels and then gradually waned but remained readily detectable over the next 9 weeks.

The pattern of circulating antibodies was not altered greatly when the interval between sensitization by blood transfusion and the implementation of immunosuppression and transplantation was shortened to 2 weeks. High levels of IgG alloantibodies were present at the time of transplantation 2 weeks after transfusion (n = 8). Circulating donor-specific IgG alloantibodies were again decreased transiently 7 days after transplantation and then rebounded to pre-transplantation levels before gradually waning to a plateau level (Figure 1).

When the interval between sensitization by blood transfusion and the implementation of immunosuppression and transplantation was shortened to 1 week, the maximal level of circulating donor-specific alloantibodies was truncated, but IgG alloantibody levels remained close to the pre-transplantation levels until the experiment was terminated at 90 days (Figure 1).

IgG2b was the predominant subclass of donor-specific antibodies in all groups of sensitized recipients both at 14 and 90 days after transplantation (Figure 2). As verified in our subsequent experiments, these sera containing IgG2b alloantibodies activate complement effectively.

image

Figure 2. Donor-specific IgG alloantibody subclasses measured at a 1:64 dilution in the circulation of Lewis rats sensitized with a DA blood transfusion 3 weeks (top), 2 weeks (middle) and 1 week (bottom) before transplantation. IgG subclasses were measured at serial dilutions by flow cytometry on donor strain mononuclear cells at the time of heart transplantation (Htx) and 14 and 90 days after transplantation (see Materials and Methods for details). IgG2b alloantibodies (solid circles) were present in the highest levels in each of the sensitized cardiac allograft recipients.

Download figure to PowerPoint

Alloantibodies from transplant recipients cause complement activation and C4d deposition

The capacity of the alloantibodies to activate complement and cause deposition of C4d on cells was confirmed in flow cytometry studies. The affinity-purified polyclonal antibody to rat C4d did not react with control leukocytes from normal Lewis rats. It also did not react with cells coated with monoclonal IgM antibodies to Lewis MHC (WFL3C6.1) in the absence of a source of complement. As expected, the antibody to rat C4d bound to cells coated with monoclonal IgM antibodies and incubated with serum from a C6 deficient rat, but not to cells with heat inactivated serum from the same source. This experiment confirmed that the affinity-purified antibody to C4d reacted with complement deposited on cells after activation by antibody.

Sera from sensitized recipients that were sampled at the peak of circulating donor-specific IgG alloantibodies 14 days after transplantation also caused C4d deposition on DA cells that was easily detectable with the affinity-purified antibody to C4d (Figure 3). No C4d deposition was detected when heat-inactivated serum was used as a source of complement, or when serum from unsensitized cardiac allograft recipients was used as a source of alloantibody (Figure 3). Similar results were obtained when complement activation was measured with an affinity-purified rabbit antibody to C3d (DAKO, Glostrup, Denmark). This experiment demonstrated that alloantibodies elicited by cardiac transplants were capable of activating complement.

image

Figure 3. Circulating alloantibodies from sensitized cardiac recipients bind to donor strain leukocytes and cause C4d deposition. Upper panel illustrates that IgG in a representative serum sample from a sensitized, but not an unsensitized, allograft recipient binds to DA leukocytes as measured by flow cytometry. Lower panel shows that this same serum from a sensitized (transfused), but not an unsensitized, allograft recipient caused measurable C4d deposition on the leukocytes when normal, but not heat-inactivated, serum was used as a source of complement (see Materials and Methods for details). These experiments confirm that the donor-specific alloantibodies elicited by transfusion and transplantation were capable of activating complement as measured by our polyclonal antibody to rat C4d.

Download figure to PowerPoint

C4d deposition in cardiac transplants to sensitized recipients

The affinity-purified antibody to rat C4d did not stain nontransplanted or isografted hearts. The intensity and distribution of staining of cardiac allografts with the anti-C4d correlated with the amount of circulating donor-specific antibodies. As an additional control, it was demonstrated that pre-incubating the affinity-purified antibody with the C4d peptide blocked the stain for C4d in tissue sections.

There was an obvious contrast in the intensity of staining for C4d in all of the hearts transplanted to sensitized recipients compared to unsensitized recipients. In the majority of unsensitized recipients, only trace amounts of C4d was discernible focally on endothelium of arteries, capillaries or veins of hearts sampled at 14 or 90 days after transplantation (Figures 4 and 5A). Minor amounts of C4d were evident on the endothelium of capillaries in two of seven hearts sampled at 14 days and four of nine hearts sampled at 90 days after transplantation. These results correlated with the lack of readily detectable antibodies in the circulation of unsensitized cardiac allograft recipients (Figure 1).

image

Figure 4. Scattergrams of C4d deposition in arteries (circles) and capillaries (plus signs) of sensitized and nonsensitized recipients of cardiac allografts. Cardiac allografts recovered 14 days after transplantation to sensitized recipients had intense, diffuse, linear deposits on both arteries and capillaries (top panel). In contrast, allografts to unsensitized control recipients had little staining of arteries and only weak staining of capillaries. At 90 days after transplantation, C4d deposits were still strong in the sensitized recipients, but not in the control recipients (bottom panel).

Download figure to PowerPoint

image

Figure 5. Immunohistology of cardiac allografts. Minimal C4d was deposited on capillary endothelium and no C4d on arterial endothelium of a heart transplant removed 14 days after transplantation to a control recipient (A). A serial section from the same transplant, stained for CD68 demonstrates moderate focal infiltrates of macrophages (B). In contrast, C4d was deposited in a diffuse, strong (3+) pattern on capillary and arterial vascular endothelium of a heart transplant removed 14 days after transplantation to a recipient sensitized with a blood transfusion (C). A serial section from the same transplant, stained for CD68 demonstrates diffuse, intense, intra- and perivascular macrophages (D). Diffuse, strong deposits of C4d were retained in a DA heart 1 day after retransplantation from a sensitized Lewis recipient back to a DA recipient as demonstrated by immunofluorescence (inset) and immunoperoxidase techniques (E). C4d deposits were almost completely cleared by 5 days after retransplantation from a sensitized Lewis recipient back to a DA recipient (F). Diffuse, strong deposits of C4d were present on the vascular endothelium of capillaries and arteries in a heart transplant removed 90 days after transplantation to a recipient sensitized with a blood transfusion (G). The artery has significant neointimal thickening and some C4d is evident on the elastic layer (arrows) between the neointima and media. Although the transplant was still beating, a Mallory trichrome stain demonstrates substantial fibrosis (blue) by 90 days after transplantation (H).

Download figure to PowerPoint

In sensitized recipients, the strongest C4d deposits were found in hearts that were removed 14 days after transplantation (Figure 4). In these hearts, intense linear staining for C4d was demonstrated diffusely on the endothelium of arteries, capillaries and veins (Figure 5C). Diffuse, linear deposits of C4d were still detected in the cardiac allografts 90 days after transplantation (Figure 5G). The staining for C4d at this later interval was less intense than at the early sampling time. The decreased C4d staining correlated with the abated levels of IgG alloantibodies in the circulation (Figure 1).

Correlation of C4d deposits with injury and rejection of cardiac transplants

Sensitization increased the extent and degree of acute and chronic injury to the vessels and myocardium. Hearts transplanted to unsenstized recipients had little or no myocardial injury and limited infiltrates of mononuclear cells at either 14 or 90 days after transplantation. Mallory trichrome stains demonstrated negligible interstitial fibrosis at 90, 150 or 180 days.

In sensitized recipients, the degree and extent of vascular and myocardial injury were correlated with the time between sensitization and transplantation. A 1-week interval between sensitization, precluded the development of a complete IgG response before transplantation. Histology demonstrated a variable degree of interstitial and periarterial cellular infiltration. Only local infiltrates with little myocyte injury were found in two of five transplants removed 14 days after transplantation. In the three other transplants in this group, diffuse infiltrates expanded the interstitial spaces. The capillary endothelium remained largely intact and little intravascular platelet aggregation was noted. Immunohistology for CD68 disclosed intra- and pericapillary accumulations of macrophages. All of the transplants in this group survived 90 days, at which time there was moderate fibrosis in the interstitium and myointimal proliferation in arteries.

Intervals of 2 or 3 weeks between sensitization and transplantation allowed a significant donor-specific IgG alloantibody response to develop before transplantation. The high levels of circulating donor-specific alloantibodies were accompanied by an intense, diffuse interstitial and periarterial infiltration of macrophages (Figure 5D) that was associated with substantial myocytolysis and initial collagen formation in transplants removed 14 days after transplantation. Capillary networks were completely effaced in the areas of most intense infiltration and platelet aggregates occluded numerous capillaries and arteries. Three of the eight hearts that were transplanted 2 weeks after sensitization ceased functioning within 6 to 12 days after transplantation. In addition to large numbers of macrophages, the infiltrates in these rejected hearts contained neutrophils that stained for cytoplasmic myeloperoxidase. Although the majority of transplants continued to beat in spite of high levels of circulating alloantibodies, these transplants sustained extensive injury. The transplants that survived 90 days had large areas of fibrosis that were densest in the subepicardium and subendocardium. These fibrotic areas encased the remaining myocardial bundles and were associated with neointimal expansion in arteries (Figure 5H).

Clearance of C4d deposits in cardiac transplants

Unlike antibodies, C4b and its subsequent split products ending with C4d have the unusual ability to bind covalently to proteins or carbohydrates in tissues. The capacity for covalent binding to tissues is often cited as an advantage for C4d as a marker of antibody-mediated rejection (3). It has also raised questions about the length of time that C4d remains after the deposition of antibodies has ceased. To answer this question, the in vivo clearance of C4d was determined by retransplanting DA cardiac allografts from Lewis recipients back to DA recipients. Intense C4d deposits were established by transplanting hearts from DA donors to Lewis recipients that had been presensitized by a transfusion of blood from DA donors. One week after transplantation, the transplanted hearts were explanted. Three of these hearts were preserved at the time of explantation for immunohistology to verify the degree of C4d deposits, and the remaining nine explanted DA hearts were retransplanted into normal DA recipients. These retransplanted hearts were removed at 1, 3, 5 or 7 days after retransplantation to document the clearance of C4d from the transplant.

Diffuse, continuous deposits of C4d were readily detectable on arterial, capillary and venous endothelium 1 and 3 days after retransplantation (Figures 5E and 6). The intensity of these deposits was not decreased from the time of transplantation. C4d deposits decreased to minimal levels 5 and 7 days after retransplantation (Figures 5F and 6). Thus, in spite of covalent binding, C4d was cleared within days after contact with antibodies is terminated.

image

Figure 6. Scattergram of C4d deposition in arteries (circles) and capillaries (plus signs) of retransplanted cardiac allografts. C4d deposits were cleared from the arterial and capillary endothelium of DA hearts 5 to 7 days after retransplantation from a sensitized Lewis recipient back to a DA recipient.

Download figure to PowerPoint

Comparison of IgG and C4d deposits in cardiac transplants

In order to compare directly the sensitivity of C4d and IgG as markers in graft rejection, frozen sections were stained by indirect fluorescence using rabbit polyclonal antibodies to C4d or IgG as primary reagents and the same FITC-conjugated secondary antibody. The fluorescent and immunoperoxidase techniques both demonstrated that C4d was deposited on vascular endothelial cells in arteries, capillaries and veins (Figures 5 and 7). Little staining of matrix was evident in immunoperoxidase stains for C4d, but the matrix surrounding the capillaries was stained by immunofluorescence for C4d (Figure 7). The fluorescent and immunoperoxidase techniques demonstrated a similar clearance of C4d from the endothelium. The strong capillary deposition of C4d that was demonstrated immediately after retransplantation converted to weak, focal staining by 5 days after retransplantation.

image

Figure 7. Immunofluorescence stains for IgG (left panel) and C4d (right panel) in a cardiac allograft removed from a sensitized recipient prior to retransplantation. Fluorescence signal for IgG is evident at low power (upper panel) on the arterial endothelium and in the matrix surrounding the artery (A), capillaries and myocytes. At higher power (lower left panel), the signal on capillary endothelium (arrows) is not distinguishable from the matrix surrounding myocytes (plus symbol). In contrast, the fluorescence signal for C4d is significantly stronger on the endothelium of the artery (A) and capillaries than on the matrix. At higher power (lower right panel), some signal for C4d is evident on the matrix surrounding the individual capillaries (arrows).

Download figure to PowerPoint

In these same transplants, IgG deposits were detected on arterial endothelium. This was accompanied by an extensive distribution of IgG in the matrix. Immunoflourescence extended around arteries, capillaries and myocytes (Figure 7). This immunofluorescence in the matrix eclipsed the signal for IgG on the capillary endothelium in both intensity and persistence. The signal for IgG on capillary endothelium was never distinct from the underlying matrix and it diminished within 3 days after retransplantation. The differences in staining patterns for IgG and C4d are not idiosyncratic to this experiment because we have similar findings in other models of antibody-mediated rejection (see Supplemental Figure 1).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We describe the production and application of a polyclonal antibody to rat C4d. We found that this polyclonal antibody to rat C4d yielded reliable results in Western blots, flow cytometry and immunohistology. Immunohistology with this polyclonal antibody allowed important extensions of the clinical data on C4d in a rat model of cardiac transplantation. Clinically, only limited data are obtainable on the distribution and clearance of C4d, as well as the progression of graft injury in the presence of extensive deposition of C4d.

In cardiac transplants, immunohistology is primarily limited to endomyocardial biopsies from the right side of the heart. These biopsies capture a circumscribed sample of capillaries and occasional arterioles. Therefore, the diagnosis of antibody-mediated rejection in cardiac transplants has been limited to changes in the small vessels. Even in the more extensively reported experience with C4d in renal biopsies, the focus has been on the staining pattern of peritubular capillaries with less consensus about the incidence and significance of arterial or venous staining (4).

In our model of sensitization induced by a blood transfusion, the deposition of C4d correlated with the level of circulating donor-specific IgG alloantibodies. The alloantibodies were measured in a clinically relevant assay using flow cytometry to detect IgM and IgG antibodies bound to donor leukocytes. A single intravenous blood transfusion elicited a transient IgM alloantibody that switched to a strong IgG response. The IgG response continued after treatment with CsA was initiated and no additional therapy was administered to remove antibodies or modify B-cell responses. These conditions resulted in diffuse, dense deposits of C4d on the capillaries throughout the myocardium. The staining of C4d in the capillaries was similar to that found in clinical endomyocardial biopsies (7,9–12). In addition, C4d deposition was found on the large arteries and veins. The deposition of C4d on arteries was strong in the acute phases of rejection before any pathological effects were evident in the arteries on routine histology. However, the staining was more intense in the capillaries both in the acute and chronic phases of rejection. As in clinical biopsies (9–12), C4d deposition in capillaries was associated with intra- and pericapillary accumulations of macrophages that stained with monoclonal antibodies to CD68. Macrophages can respond directly to many split products of complement. First, macrophages are chemoattracted and activated by receptors for C3a and C5a. Second, C4b and C3b can act as ligands for complement receptor type 1 (CR1; CD35) on macrophages. Finally, CR3 (CD 11b/CD18) on macrophages recognizes iC3b, the enzymatically inactive fragment of C3b. Together, these interactions of macrophages with complement activation products could be responsible for the accumulation of macrophages in sites of complement activation.

Although the majority of transplants to sensitized recipients in this study continued to function in spite of extensive C4d deposition, most of the grafts demonstrated a degree of cellular infiltration that would be associated with graft dysfunction and that would prompt therapeutic intervention clinically. Indeed, within 90 days after transplantation, there was extensive replacement of myocardium by fibrosis. This fibrosis was concentrated around the larger arteries, the subepicardium and the subendocardium. These are locations that had the most intense C4d deposition and macrophage infiltrates in the acute phases of rejection. Sensitization also was associated with increased neointima formation in large arteries of long-term transplants. C4d deposition in these arteries was most intense on the endothelial cells, but in addition, C4d was evident on the remnants of the elastica that demarcates the expanded neointima from the media.

Our experimental model also allowed the clearance of C4d to be determined. It has been suggested that deposits of C4d may have a long half-life because of the capacity of C4b and its subsequent split product C4d to bind covalently to tissues. Anecdotal reports of sequential biopsies from cardiac or renal transplants in humans have suggested that C4d can be cleared within weeks (6,7,23). These clinical observations are complicated by many variables including incomplete removal of circulating antibodies by plasmapheresis or other treatments, redistribution of antibodies from interstitial compartments and continued antibody production. All of these variables were eliminated experimentally by removing the graft from the allogeneic recipients and retransplanting it into an isogeneic recipient. Under these simplified conditions, immunohistological evidence of C4d deposition was cleared within 5 days, and only small amounts of C4d were detectable in foci of inflammation. This rapid clearance of C4d is similar to the rapid clearance of C3d from endothelial cells in culture (24). Although the mechanism for this rapid clearance is under investigation, one possible mechanism is that sublytic quantities of the terminal complement components (C5b-C9) cause exocytosis and endocytosis of portions of the cell membranes that contain C4d.

Immunofluorescence stains demonstrated a much more extensive distribution of IgG than C4d in the matrix. Although distinct deposits of IgG were detectable on the arterial endothelium, the staining for IgG in the matrix prevented an accurate assessment of IgG on the capillary endothelium. It is not clear whether the IgG associated with the matrix is functional because it is not associated with C4d staining. A similar difficulty in interpreting stains for IgG has been noted in clinical endomyocardial biopsies (9,10).

In summary, with the aid of a polyclonal rabbit antibody specific for rat C4d, we were able to define the extent of C4d deposition in cardiac transplants to sensitized recipients and to determine the clearance of C4d in vivo.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This research was supported by NIH grants R01-AI42387, R01-HL63948 and P01-HL56091.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References