A Novel Atypical Hemolytic Uremic Syndrome–Associated Hybrid CFHR1/CFH Gene Encoding a Fusion Protein That Antagonizes Factor H–Dependent Complement Regulation

Genomic CFH-CFHRs Rearrangements

Among 154 aHUS patients (67 patients with complement gene mutations) screened by multiplex ligation-dependent probe amplification (MLPA), we found genomic rearrangements involving CFH-CFHRs in seven patients (4.5%), of whom only one patient also had a heterozygous mutation (c.1429+1G>C in CFI). Two cousins with disease onset at the ages of 6 months and 20 years (the former also carried the CFI mutation)³ and two patients with sporadic aHUS (onset: 1 and 22 years) had a previously described heterozygous CFH/CFHR1 hybrid gene including exons 1–21 of CFH and exons 5 and 6 of CFHR1.¹³ In a sporadic patient (onset at 5 months), we found another previously reported hybrid CFH/CFHR1¹⁴ including exons 1–22 of CFH and exon 6 of CFHR1. Four of five carriers of CFH/CFHR1 hybrid genes developed ESRD early after disease onset. Three of them manifested disease recurrence in the graft, whereas the fourth received post-transplant eculizumab prophylaxis and had good allograft function at 10 months follow-up.

In addition, in two affected patients from a family (the proband and his daughter in Figure 1), MLPA revealed three copies from probes 14–25 (Table 1), suggesting a large heterozygous duplication starting from exon 23 of CFH, including CFHR3, and ending after exon 5 of CFHR1 (Figure 2). The duplication was confirmed by comparative genomic hybridization (CGH) array (Figure 3). The proband’s wife was normal, whereas the unaffected son carried the heterozygous duplication like his affected relatives (Figures 1 and 2).

Breakpoint Mapping and Identification of a CFHR1/CFH Hybrid Gene

A 1955-bp PCR product was obtained from the DNA of the proband, his affected daughter, and his unaffected son using specific primers annealing CFHR1 intron 4 (forward) and CFH intron 23 (reverse) (Figure 2), whereas no PCR product was obtained with control DNA (Supplemental Figure 1A). Bidirectional sequencing of the long-range PCR product (primers are shown in Supplemental Table 1) revealed that the breakpoint is within a 193-bp region starting 999 bp after CFHR1 exon 5 (Figure 4). These results imply that an NAHR event occurred between the last exon of CFH and the last exon of CFHR1 (Figure 4B), resulting in a CFHR1/CFH hybrid and an extra copy each of CFHR3 and CFHR1 (Figure 4B). The CFHR1/CFH hybrid encodes for a protein with the first four SCRs of FHR1 and the last SCR20 of FH (Figure 5A). The rearrangement was confirmed by multiplex PCR-amplifying fragments with different sizes for the CFHR1/CFH hybrid and the wild-type CFHR1 in carriers of the duplication and only the wild-type CFHR1 fragment in noncarriers (Supplemental Figure 1B).

Clinical, Laboratory, and Genetic Data of the Family with the CFHR1/CFH Hybrid Gene

The proband (Figure 1) was hospitalized at the Molinette Hospital, Turin, Italy at the age of 49 years with malignant hypertension, severe ARF (serum creatinine=7.1 mg/dl), anemia (hemoglobin [Hb]=7.9 g/dl), and thrombocytopenia (platelets=109,000/μl) that manifested 1 week after an upper respiratory tract infection. His past medical history was uneventful. Laboratory examinations showed lactate dehydrogenase=2704 IU/L, serum haptoglobin<3 mg/dl, total bilirubin=2.4 mg/dl, and hematuria. Tests for autoantibodies (ANA, anti-DNA, ANCA, anticardiolipin, and lupus anticoagulant) were negative. The presence of Shiga toxin–producing Escherichia coli in stool culture was excluded, and ADAMTS13 activity was normal. A diagnosis of aHUS was made. Antihypertensive polytherapy with minoxidil, clonidine, doxazosin, and nifedipine was started. The patient was also treated with steroids, packed red blood cell transfusions, plasma infusions, and plasma exchanges. Despite treatment, the patient became anuric, and hemodialysis was started without recovering renal function. After 2 months, the patient was discharged under antihypertensive therapy in hematologic remission (Hb=11 g/dl, platelets=234,000/μl) but on chronic hemodialysis. At the last follow-up, 3 years after aHUS onset, the patient was in stable condition and waiting for renal transplantation. Serum C3, C4, and FH and plasma SC5b-9 levels were in the normal range; the test for anti-FH antibodies was negative (Table 2).

His daughter (Figure 1) was hospitalized at 20 years of age because of severe hemolytic anemia (Hb=5.1 g/dl, lactate dehydrogenase=2000 IU/L), thrombocytopenia (platelets=111,000/μl), schistocytes in the peripheral smear, hypertension, ARF (serum creatinine=6 mg/dl), microhematuria, and proteinuria (4.2 g/24 h). She was taking birth control pills. ADAMTS13 activity was normal, and Coombs test was negative. Shiga toxin–producing E. coli was not detected in stool culture. aHUS was suspected, and the patient was treated with daily plasma infusions, packed red blood cell transfusions, steroids, furosemide, and two Ig infusions. Hypertension was controlled by amlodipine and doxazosin. Because of persistence of hemolysis and severe renal dysfunction, seven daily plasma exchanges were performed without benefit. Serum C3 levels were slightly lower than normal, serum C4 and plasma SC5b-9 were normal, and anti-FH antibodies were absent (Table 2). On the thirteenth day of hospitalization, the patient started treatment with eculizumab (four 900-mg weekly doses and then 1200 mg every 2 weeks) after meningococcal vaccination and under prophylaxis with ciprofloxacin, obtaining hematologic remission and progressive renal function improvement. After 2 months, the patient was discharged, with serum creatinine=1.88 mg/dl, Hb=9 g/dl, platelets=270,000/μl, proteinuria=2.4 g/24 h, under 1200 mg eculizumab every 2 weeks, and antihypertensive therapy. At the last follow-up, 7 months after onset, the patient was in stable remission (serum creatinine=1.5 mg/dl, Hb=12.8 g/dl, platelets=294,000/μl, proteinuria=1.4 g/24 h) with eculizumab treatment every 2 weeks. Serum C3 was normal, whereas plasma SC5b-9 was slightly elevated (Table 2). Total complement CH50 activity was heavily depressed (Table 2) because of chronic eculizumab treatment. Family history did not disclose other relatives with aHUS.

Genotyping for the CFH single nucleotide polymorphisms rs3753394, rs800292, rs1061170, rs3753396, rs1065489, and rs7144 in the membrane cofactor protein (MCP) gene showed that the proband was heterozygous for the aHUS risk haplotype CFH-H3 (TGTGT) on the chromosome without the duplication and homozygous for the MCP allele c.*897 T>C (rs7144) that tags the MCPggaac risk haplotype. Neither the affected daughter nor the unaffected son, both carrying the duplication, showed the CFH-H3 risk haplotype, whereas both were heterozygous for the MCP c.*897 T>C risk allele (Figure 1).

FHR1 Western Blotting

Western blot analysis of the proband’s serum showed FHR1 bands with higher density than serum from his wife, consistent with the presence of an extra CFHR1 copy encoding the FHR1/FH fusion protein (Figure 5, B and C).

Hemolytic and Competition Assays

The FH C terminus is crucial for FH binding to cell surfaces. It is expected that, in the FHR1/FH hybrid protein reported here, the replacement of the FHR1 C-terminal region by the C terminus of FH will confer to the protein the capacity to compete with FH for binding to cell surfaces, thus antagonizing FH complement regulatory activity. To test this possibility, we evaluated whether the serum from heterozygous carriers of the CFHR1/CFH hybrid lyses sheep erythrocytes in an FH-dependent hemolysis assay. Serum from the proband lysed sheep cells in a dose-dependent manner, whereas normal serum and serum from the proband’s wife without the duplication did not (Figure 6A). A replica of this experiment, setting the lysis obtained with the proband serum as 100%, revealed a lower degree of lysis with serum from the affected daughter or the unaffected son, both carriers of the CFHR1/CFH hybrid (Figure 6B).

To evaluate whether the lysis obtained with the proband’s serum was a consequence of the competition between FH and the FHR1/FH hybrid protein, we purified the FHR1 fraction from the serum of hybrid protein carriers and tested whether purified FHR1 added to normal human serum competed with FH and caused lysis in a modified FH-dependent sheep hemolysis assay. FHR1, FHR2, and FHR5 do not exist as individual molecular entities in human plasma. They are present as a complex mix of heterooligomeric proteins (dimers and tetramers), in which the oligomers containing FHR1 are mostly the major components.¹⁷ We, therefore, refer to the purified mix of FHR1 (including the hybrid FHR1/FH), FHR2, and FHR5 heterooligomers here as purified FHR1 (Figure 6C). Addition of purified FHR1 from heterozygous CFHR1/CFH hybrid carriers to two different control sera resulted in dose-dependent sheep erythrocyte hemolysis, whereas equal amounts of purified FHR1 from normal serum had no effect (Figure 6D).

C5b-9 Deposition on Endothelial Cells

To determine whether the presence of the FHR1/FH hybrid protein in the circulation resulted in complement dysregulation at the endothelial cell level, the human microvascular endothelial cell line-1 (HMEC-1) preactivated with ADP was exposed to serum from the proband or his relatives, and the surface area covered by C5b-9 deposits was evaluated. We found more C5b-9 deposition on endothelial cells exposed to sera from the proband (in remission) and his affected daughter (with acute disease) versus cells incubated with control serum or serum from the proband’s wife (Figure 7, A and B). Interestingly, serum from the unaffected son who carries the FHR1/FH hybrid also induced higher than normal C5b-9 deposits on HMEC-1, although to a lesser extent than sera from his affected relatives (Figure 7, A and B). Addition of increasing amounts of purified FHR1 from normal serum to serum from the wild-type proband’s wife had no effect on C5b-9 deposits (Figure 7C).

Study Participants

aHUS was diagnosed in cases reported to have one or more episodes of nonimmune hemolytic anemia, thrombocytopenia, and renal impairment. One hundred fifty-four consecutive patients of the International Registry of Hemolytic-Uremic Syndrome/Thrombotic Thrombocytopenic Purpura analyzed for CFH, CFI, MCP, C3, CFB, and THBD genes were screened for genomic rearrangements affecting CFH, CFHR1, CFHR2, CFHR3, and CFHR5. The study was approved by the Ethics Committee of the Azienda Sanitaria Locale, Bergamo, Italy, and informed consent was obtained in accordance with the Declaration of Helsinki.

A detailed description of study participants, methods used for complement profile assessment, mutation screening and genotyping, MLPA, CGH arrays, breakpoint identification, and Western blot analysis of FHR1 is in Supplemental Material.

FH-Dependent Hemolytic Assay

Sheep erythrocytes (2% packed cell volume) in AP buffer (2.5 mM barbital, 1.5 mM sodium barbital, 144 mM NaCl, 7 mM MgCl₂ and 10 mM EGTA, pH 7.4) were incubated with increasing amounts of serum. After 30 minutes at 37°C, the reaction was stopped by adding AP buffer with 20 mM EDTA. After centrifugation, supernatants were read at 414 nm. Erythrocytes diluted in distilled water were taken as 100% lysis, and erythrocytes diluted in AP buffer were used as blank for spontaneous lysis.

Purification of FHR1 and FHR1/FH Hybrid Protein and Competition Assays

FHR1 and FHR1/FH hybrid heterooligomers were purified from human plasma/serum by immunoaffinity chromatography. Briefly, filtered plasma/serum was loaded into an immunoaffinity column coupled to an in-house anti-human FHR1 mouse mAb (MBC125) (Supplemental Material). Bound protein was eluted at low pH, and fractions containing the protein were pooled and purified using a heparin column (GE Healthcare). Eluted proteins were further polished in a gel filtration column (GE Healthcare) equilibrated with 10 mM Hepes buffer (pH 7.0) and 150 mM NaCl. Before use, FHR1 and FHR1/FH concentrations were normalized to obtain preparations that contained identical concentrations of FHR1 protein. The capacity of the FHR1/FH hybrid protein to compete with the activity of FH was assessed in a modified sheep hemolytic assay. The human serum used in these assays was previously deprived from endogenous FHR1 using an immunoaffinity column with the anti-FHR1, FHR2, and FHR5 MBC125 mAb,¹⁷ and the FH activity decreased by approximately 50% (20% lysis in the sheep erythrocyte assay) by adding the anti–N-terminal FH antibody OX24 (Supplemental Figure 2). Sheep erythrocytes in AP buffer (2% packed cell volume) were incubated with 10% human serum in AP buffer containing increasing amounts of purified FHR1 from controls and carriers of the CFHR1/CFH hybrid for 1.5 hour at 37°C. The reaction was stopped by adding AP buffer with 20 mM EDTA; after centrifugation, supernatants were read at 414 nm, and lysis was plotted versus concentration of FHR1.

C5b-9 Deposition on Endothelial Cells

HMEC-1, activated with ADP, was incubated with serum from patients or controls. The staining on the endothelial cell surface with rabbit anti-human complement C5b-9 was evaluated by confocal inverted laser microscope. The area occupied by the fluorescent staining was evaluated and expressed as pixels² per field analyzed. For each sample, the mean of 15 fields (excluding the lowest and the highest values) was calculated. Results are expressed as means±SEMs.

Data were analyzed by ANOVA. P values of <0.05 were considered to be statistically significant. Additional details are in Supplemental Material.