These authors contributed equally to this work.
Recurrence of HUS Due to CD46/MCP Mutation After Renal Transplantation: A Role for Endothelial Microchimerism
Article first published online: 5 JUL 2007
American Journal of Transplantation
Volume 7, Issue 8, pages 2047–2051, August 2007
How to Cite
Frémeaux-Bacchi, V., Arzouk, N., Ferlicot, S., Charpentier, B., Snanoudj, R. and Dürrbach, A. (2007), Recurrence of HUS Due to CD46/MCP Mutation After Renal Transplantation: A Role for Endothelial Microchimerism. American Journal of Transplantation, 7: 2047–2051. doi: 10.1111/j.1600-6143.2007.01888.x
- Issue published online: 5 JUL 2007
- Article first published online: 5 JUL 2007
- Received 10 February 2007, revised 01 May 2007 and accepted for publication 07 May 2007
- hemolytic uremic syndrome;
- renal transplantation
Mutations in the gene of the membrane cofactor protein (MCP/CD46), a complement regulatory protein, were recently described as a cause of hemolytic uremic syndrome (HUS). MCP is a transmembrane glycoprotein expressed in kidneys; therefore, the transplantation of a normal kidney should not be complicated by HUS recurrence. However, we report the case of a 32-year-old woman with an MCP mutation who developed a recurrence of HUS after renal transplantation. We found that she had vascular microchimerism of endothelial cells. We suggest that recurrence may be favored by vascular microchimerism, in which the mutated protein is produced in the in the kidney graft by endothelial cells originating from recipient.
Mutation in membrane cofactor protein (MCP/CD46), a protein regulating complement activation, is now a well-recognized cause of familial and sporadic atypical hemolytic uremic syndrome (HUS), unrelated to HUS associated with shigatoxin-secreting bacteria (1–3). Other atypical nonshigatoxin-associated HUS cases have also been described. These cases present defects in the regulation of the complement activation pathway, for example in Factor H, a soluble protein that inhibits the complement activation cascade. Factor H deficiency is complicated by a high recurrence rate of HUS—almost 50%—after kidney transplantation (4). MCP, in contrast to Factor H, is a transmembrane glycoprotein expressed mostly in kidneys acting locally as a cofactor with Factor I to activate complement (C3b and C4b fractions) catabolism; MCP also regulates the complement activation pathway but is produced and exerts its effects locally. Therefore, complement dysregulation related to MCP mutation should be corrected by the transplantation of a normal kidney. Only three siblings with an identified MCP mutation and who have been treated by kidney transplant have been reported: clinical evolution was favorable without a recurrence of HUS in the allograft (5). Here, we report the case of a patient with a novel MCP mutation. The patient suffered a recurrence after renal transplantation, indicating that some MCP mutations may be at risk factors for HUS recurrence. Analyses of graft biopsies revealed endothelial cells of recipient origin, indicating vascular microchimerism. This also suggests that the production of mutated MCP protein in the kidney graft by endothelial cells originating from recipient may explain the recurrence of HUS.
A 32-year-old woman had with HUS 10 years ago, which lead to severe acute renal failure, requiring dialysis. She had had two pregnancies, each complicated by pre-eclampsia, and gave birth to two healthy children. The patient had no infection related to shigatoxin-producing bacteria or other infectious agents at the onset of HUS, and was not taking any drugs. Plasma infusions were unsuccessful and she was treated by periodic hemodialysis. She had three brothers and three sisters. One of her sisters was also on dialysis therapy from the age of 29 years as a result of HUS; she died suddenly 4 years later of unknown causes. None of the patient's other siblings or children had HUS.
Biological investigations were performed to identify predisposing genetic or immunological factors for HUS. Complement C3 and C4 levels were within the normal ranges. Tests for anti-nuclear, anti-double strand DNA and anti-phospholipid antibodies were negative. The Von Willebrand factor-cleaving protease (ADAMTS-13) activity was within the normal range of a normal population (80% of the activity of control patients). Factor H and Factor I genes were sequenced, but no mutations were detected. We then analyzed her MCP/CD46 gene. We identified a heterozygous T to G mutation at position +2 of exon 2 (IVS2 + 2T > G) (Figure 1, panel B), which involved the donor splice site in intron 2. RT-PCR, using primers from exon 1 (forward primer) and exon 4 (reverse primer), of the MCP transcript resulted in a smaller fragment of 235 bp (Figure 1, panel C). Direct sequencing of this fragment after gel purification revealed that the first 45 bp from exon 2 was spliced onto exon 3 (Figure 1, panel D). Thus, this nucleotide change results in the selection of an alternative splice site (TTG gtaaac). This splice site is located in exon 2, and its use results in the deletion of 144 bp and 48 amino acids, which are in phase with the wild type sequence of the protein.
FACS analysis of the patient's peripheral blood mononuclear cells (PBMCs) showed that surface expression of CD46 was lower in patients than normal subjects (Figure 1A). Only one of the patient's sisters, who had never developed HUS, was available for testing; no mutation was detected in her MCP gene.
The patient received a kidney transplant from a woman who suffered brain death following cardiac arrest after a pulmonary embolism. The immunosuppressive regimen consisted of an induction therapy with anti-thymoglobulin and solumedrol. The maintenance therapy consisted of steroids, mycophenolate mofetil and sirolimus. The patient recovered renal function immediately after transplantation. Her serum creatinine concentration was 99 μmol/L on day 7, and there was no significant proteinuria. Sirolimus through levels were maintained between 12 and 16 ng/mL. A graft biopsy was performed 2 months later, because of sudden nephrotic-range proteinuria without graft dysfunction or signs of hemolysis. Lesions corresponding to thrombotic microangiopathy were observed. Optical microscopy revealed intraluminal fibrin thrombi in two arterioles and capillaries of one glomerulus (Figure 2). Other glomeruli were ischemic. The samples tested negative for the C3 fraction and all immunoglobulins by immunofluorescent staining. Donor-specific antibodies were absent as assessed by ELISA on sera from the pre-engraftment period or at the date of HUS and the samples were negative for C4d staining, ruling out the possibility of antibody-mediated rejection. HUS recurrence in the engraft kidney was diagnosed and the patient was treated by plasma exchange, followed by intravenous globulins. The immunosuppressive treatment was not modified. Proteinuria progressively regressed, and two further kidney biopsies were performed (after 1 and after 3 months): no signs of thrombotic microangiopathy were detected.
The patient experienced a second relapse of HUS 2 years later. She had refractory hypertension, mild acute renal failure with an increase in serum creatinine concentrations from 100 to 180 μmol/L, but no significant proteinuria. Kidney biopsies revealed glomerular capillary thrombosis, mesangiolysis and intimal swelling in the arteries. No donor-specific antibodies were detected by ELISA. She received a second course of plasma exchange and the sirolimus dose was halved. Sirolimus trough levels decreased from 12 to 6 ng/mL. Renal function returned to normal baseline levels and hypertension was controlled by anti-hypertensive therapy involving angiotensin converting enzyme inhibitors.
MCP is produced in the kidney and exerts its effects locally; therefore, HUS recurrence may have been due to endothelial microchimerism of the transplanted kidney. However, the demonstration of microchimerism by detecting chromosome Y (in case of male recipient transplanted with a female transplant) or AB blood group antigens would not have been informative, as both the donor and recipient were women with blood group A. Therefore we took advantage of the presence of an HLA class I mismatch to identify microchimerism. The A10 HLA antigen was expressed by the recipient, but not by the donor. Staining with a human anti-HLA A10 antibody was sporadically positive in the biopsy (Figure 3A). The same frozen sections were stained with a mouse IgG1 anti-human PECAM/CD31 monoclonal antibody, specific for human endothelial cells, to confirm that A10-positive cells were endothelial cells and not leucocytes adhering to the endothelium (Figure 3B). Overlaying the two immunofluorescent staining images confirmed that cells expressing recipient-type HLA antigens were endothelial cells (Figure 3C).
Various teams have demonstrated that there is a genetic predisposition to atypical HUS involving complement-regulating components: Factor H, CD46 (or MCP for membrane cofactor Protein) and Factor I (1,2,4,6–12). These three proteins are involved in the regulation of the alternative pathway of the complement activation pathway. Their inactivation or decreases in their concentration favor complement activation and a predisposition to the development of HUS. Membrane cofactor protein (MCP, CD46) is a transmembrane complement regulator that is widely produced, particularly in endothelial cells of the kidney (13). It acts as a local cofactor to Factor I, which cleaves C3b and C4b deposited on target membranes, and thereby inhibits complement activation. Mutated MCP has a decreased capacity to bind C3b and, therefore, a decreased ability to regulate complement activation. Upon exposure to a trigger of the complement cascade, the reduced cofactor activity of mutated MCP may result in insufficient local protection of renal endothelial cells and may consequently lead to HUS (6,12).
Several mutations of MCP have been reported to be associated with HUS (3). The most common mutation of CD46 reported involves the second intron and is associated with a decrease in its production at the cell surface. This mutation was not present in our patient. However, the patient had a heterozygotic mutation in the second exon, which is thought to involve the splice site: the thymidine (T) in position +2 is replaced by guanosine (G). We confirmed the presence of a transcript using RT-PCR and primers for exon 1 and 4. The transcript was indeed shorter than the wild-type (minus 144 bp) due to alternative splicing in exon 2. This transcript is associated with a weaker expression of CD46 at the cell surface, as demonstrated by the FACS analysis of PBMCs. Thus, this alternative splicing appears to impair the production or the trafficking of CD46 at the cell surface. This might explain why under normal conditions the patient did not suffer HUS, but under certain particular conditions there were triggers that may have overwhelmed the weakened regulatory properties of the affected CD46 leading to the development of HUS.
MCP is produced in the kidney and acts locally. Therefore, transplanting a normal kidney (with normal MCP production) would be expected to prevent a recurrence of HUS (1,2). By contrast, mutation factor I, which is produced in the liver, is not cured by renal transplantation and is associated with HUS recurrence. We surprisingly observed HUS recurrence in our patient. The very low incidence of MCP mutations in the general population, the absence of HUS history in the donor, and the absence of HUS episodes involving the controlateral kidney of the recipient largely exclude the possibility that the donor also carried a MCP mutation, which have been responsible for the recurrence of HUS. HUS is also observed in cases of vascular rejection, mostly antibody-mediated. No histological lesions compatible with humoral rejection (conventional analysis and C4d immunostaining), or donor-specific antibodies were detected using ELISA in our patient, indicating that HUS was not associated with an acute humoral rejection. HUS has also been reported as a consequence of the use of immunosuppressive drugs. As the most commonly described molecules are calcineurin inhibitors (cyclosporine A or tacrolimus) (14). Then, we decided to use an alternate treatment: sirolimus. Sirolimus belongs to the macrolide antibiotic family and binds to the same intracellular FKB12 protein than tacrolimus. Despite these similarities, sirolimus does not display the same nephrotoxicity profile as calcineurin inhibitors. Some authors have reported the resolution of HUS in transplant patients due to calcineurin inhibitors after switching to sirolimus (14). However, more recently few cases of HUS associated with sirolimus treatment have been reported. However, no study of predisposing factors of HUS involving the regulatory pathway of complement activation has been performed (15–17). We cannot exclude the possibility that sirolimus was directly or indirectly associated with HUS. However, sirolimus trough levels were strictly within the normal range and HUS disappeared after the first course of plasma exchange, even though the dose and trough levels of sirolimus were unchanged. Additionally, no HUS relapse was observed over a period of 2 years with similar sirolimus trough levels.
Another possible explanation for the recurrence of HUS is the replacement of some graft endothelial cells by recipient endothelial cells; these cells would have the patient's genotypic background and produce the mutated MCP protein. This endothelial microchimerism has been suggested by Medawar as a mechanism for graft adaptation. Such endothelial microchimerism has previously been recognized and described in human organ transplantation (18). Microchimerism, as observed by Lagaaij and colleagues, appears to be a mechanism for repair, occurring after the injury of endothelial cells. Chimerism occurred significantly more frequently in female than in male recipients, suggesting a role for hormonal factors (19). Consequently, we looked for endothelial microchimerism in the kidney graft as an explanation for the production of the mutated MCP and for HUS recurrence in our case. As donor and recipient were both women sharing the same blood group, we took advantage of the HLA class 1 mismatch between donor and recipient, and tested whether endothelial cells from the graft expressed the restricted recipient antigen (A10). Some vessels in the graft—essentially peritubular capillaries—expressed the HLA A10 antigen; this confirms the occurrence of microchimerism in our patient and provides an explanation for the recurrence of HUS. Therefore, microchimerism may constitute a susceptibility factor for HUS recurrence. Indeed, the HUS associated with a MCP mutation was detected in our patient after childhood and evolved with relapses. The triggers generally remain unknown; however, patients with genetic predisposition to developing HUS may require a trigger to develop a relapse. In our patient, sirolimus (despite trough levels in the normal range) may have been a potential trigger for HUS due to renal microchimerism of endothelial cells from the recipient (with mutated MCP) invading the graft. We, therefore, decided to decrease the dose after the second relapse.
In conclusion, our observation indicates that HUS due to MCP mutation may relapse following renal transplantation and we suggest that it could be favored by endothelial microchimerism of the kidney by recipient cells.
We thank Dr. Ketty Lee, Etablissement Français du Sang, Hôpital Henri Mondor, Créteil, France, for her technical assistance.