The Molecular Basis of Familial Hemolytic Uremic Syndrome

Mutation Analysis of Factor H Gene Reveals a Hot Spot in Short Consensus Repeat 20

Abstract

The aim of the present study was to clarify whether factor H mutations were involved in genetic predisposition to hemolytic uremic syndrome, by performing linkage and mutation studies in a large number of patients from those referred to the Italian Registry for Recurrent and Familial HUS/TTP. PCR and Western blot analyses were conducted to characterize the biochemical consequences of the mutations. Five mutations in the factor H gene were identified. Three, identified in two families and in a sporadic case, are heterozygous point mutations within the most C-terminal short consensus repeat 20 (SCR20) of factor H, resulting in single amino acid substitutions. The other two mutations introduce premature stop codons that interrupt the translation of factor H. A heterozygous nonsense mutation was identified in SCR8 in one family, and a homozygous 24-bp deletion within SCR20 was identified in a Bedouin family with a recessive mode of inheritance. Reverse transcription-PCR analysis of cDNA from peripheral blood leukocytes from the Bedouin family showed that the deletion lowered factor H mRNA levels. Although heterozygous mutations were associated with normal factor H levels and incomplete penetrance of the disease, the homozygous mutation in the Bedouin family resulted in severe reduction of factor H levels accompanied by very early disease onset. These data provide compelling molecular evidence that genetically determined deficiencies in factor H are involved in both autosomal-dominant and autosomal-recessive hemolytic uremic syndrome and identify SCR20 as a hot spot for mutations in the disease. The mutations identified here give an important hint to the relevance of the C-terminus of factor H in the control of the alternative complement activation pathway.

Hemolytic uremic syndrome (HUS) is a disease of nonimmune hemolytic anemia, thrombocytopenia, and renal failure caused by platelet thrombi in the microcirculation of the kidney and other organs (^1,2,3). In its typical presentation, HUS manifests as an acute disease and 80 to 90% of cases recover without sequelae, either spontaneously (as in most cases of childhood HUS) or after plasma infusion or exchange (as in adult or severe forms of HUS) (^{1, 2, 4}). Typical HUS is triggered by environmental factors, drugs, or infective agents such as the shiga-like toxin-producing Escherichia coli; systemic immune disorders or cancer may also cause the disease. These forms of HUS may subside when the underlying condition has been treated or removed. However, there are rare forms, often occurring in families, that frequently relapse even after complete recovery of the presenting episode (^{1, 5}), with death or permanent neurologic or renal sequelae being the final outcome in the majority of cases. These “atypical” forms are believed to have a genetic background that predisposes to microvascular thrombosis (^{6, 7}). Both autosomal-recessive and autosomal-dominant modes of inheritance have been recognized (^{5, 6, 8,9,10,11}), with precipitating events such as pregnancy, virus-like disease, or sepsis being identified in some (¹²) but not all series (^{13, 14}). Evidence that some of these cases could be cured, at least transiently, with plasma infusion or exchange suggested that the underlying genetic defect(s) associated with familial HUS were the cause of one or more abnormalities in plasma component(s) essential to the integrity of the microvascular circulation and/or to the defense mechanism of the host endothelium against injurious agents. In this context, reduced serum levels of the third component (C3) of the complement system had been reported since 1974 in sporadic or familial forms of HUS (^{15,16,17,18,19,20,21}). This initially was attributed to an inherited defect in C3 synthesis (²²), but more convincing data are now available that low C3 in HUS derives from either a lack (^{21, 23, 24}) or altered function (²⁵) of factor H, a regulatory component of the alternative pathway of the complement system (^26,27,28).

In a recent study (¹²) in a consistent number of families, we provided epidemiologic evidence of an association between factor H deficiency and low C3 levels in familial forms of HUS. Another report (²⁵) documented that an area on chromosome 1q, where factor H is mapped, segregates with HUS. In one family, a mutation in factor H gene was described, consisting of a C to G transversion causing an arginine to glycine change in short consensus repeat 20 (SCR20) (²⁵).

The present study was designed to characterize the genetic defect underlying atypical HUS by genotyping a large number of patients from our Registry for Familial and Recurrent HUS/TTP. Here we describe five mutations; three were found in families with dominant inheritance, one was found in a family with recessive inheritance, and one was found in an adult case of sporadic HUS whose disease recurred in the transplanted kidney. These results provide molecular evidence that the primary abnormality in atypical HUS is overactivity of the alternative complement pathway as a result of impaired factor H function.

Materials and Methods

Patients

HUS was diagnosed in all cases reported to have one or more episodes of microangiopathic hemolytic anemia and thrombocytopenia, defined on the basis of hematocrit <30%, hemoglobin <10 mg/dl, serum lactate dehydrogenase >460 units/L, undetectable serum haptoglobin, fragmented erythrocytes in the peripheral blood smear, and platelet count <150,000/μl, associated with acute renal failure.

HUS was defined as familial when at least two members of the same family were affected by the disease at least 6 mo apart and exposure to a common environmental triggering agent (in particular a shiga-like toxin-producing strain of E. coli) could be reasonably excluded.

Plasma levels of the third complement component below the lower limit of normal ranges (calculated as mean ± SD), i.e., <83 mg/dl, were taken as indicating hypocomplementemia.

Microsatellite Polymorphism Genotyping and Linkage Analysis

Genomic DNA was extracted from whole blood according to standard protocols (Nucleon BACC2 kit, Amersham, UK).

Microsatellite polymorphisms flanking and within the gene for complement factor H (HF, chromosome 1q32) were studied in a candidate gene approach to linkage analysis. Markers analyzed in this study are D1S240, D1S202, D1S412, D1S2816, D1S413, D1S2738, and D1S2796; primers were synthesized by Life Technologies (Paisley, UK).

PCR reactions were done in a 20-μl volume containing 100 ng DNA, 17 pmol of each primer, 16 nmol dNTP, 1.5 mM MgCl₂, 1 U Taq polymerase (Taq Gold, PE Applied Biosystems, Foster City, CA), in the presence of ³²PdCTP, and PCR buffer. After an initial 10-min denaturation at 94°C, 35 cycles were performed (94°C for 45 s, 54°C for 30 s, and 72°C for 45 s), followed by a final 10-min extension at 72°C. Samples were mixed with 20 μl of loading buffer, denatured at 75°C for 5 min, and electrophoresed on a denaturing 6% (19:1 acryl:bis) acrylamide gel in Tris Borate EDTA buffer, at 55 W for 2 to 4 h. Gels were then exposed to x-ray films for 3 h to overnight.

Linkage analysis was performed using a FASTLINK package for two-point analysis and Genehunter package for multipoint analysis (http://www.linkage.rockefeller.edu). Autosomal-recessive and autosomal-dominant transmission, with incomplete penetrance (0.7), were taken into consideration; the disease gene frequency in the general population was assumed to be 0.0001.

Single-Strand Conformation Polymorphism Analysis and Sequencing of Factor H

Factor H exons were amplified using primers located in the flanking introns and analyzed by single-strand conformation polymorphism (SSCP) of DNA; primers were constructed to avoid coamplification of highly homologous SCR in factor H related genes (hFHR) (²⁶). For amplification of SCR20, it was necessary to divide the exon into two parts (called SCR20A and SCR20B) that were independently amplified using two couples of primers, constructed to obtain 3′ mismatches on the hFHR-1 gene sequence (see Figure 1). Primers were synthesized by Life Technologies.

Figure 1: Alignment between short consensus repeat 20 (SCR20) genomic sequences of factor H and factor H-related 1 (FHR-1). Vertical bars with arrows indicate beginning and ending of SCR20. All mismatched nucleotides are boxed in gray. SCR20 was divided into two segments that were independently amplified using two sets of primers (gray arrows for SCR20B and black arrows for SCR20A).

A total of 100 ng of genomic DNA were amplified as described above. PCR products were electrophoresed on nondenaturing 6% (62:1 acryl:bis) acrylamide gel in TAE buffer (pH 6.8) at 35 W for 3 to 5 h and at 4 W overnight. Gels were visualized by silver staining, and aberrant bands were sequenced using an ABI 377 sequencer (PE Applied Biosystems).

mRNA Extraction, cDNA Synthesis, and Analysis

mRNA was extracted from peripheral blood mononuclear cells (PBMC) isolated from whole blood (standard protocol); reverse transcription (RT) was conducted using the Life Technologies Brl protocol with oligo (dT) primers, in a total volume of 20 μl, and the cDNA obtained was used to investigate mRNA levels in family 29. RT-PCR products of SCR20A from heterozygous individuals of family 29 were electrophoresed on a nondenaturing 8% (37.5:1 acryl:bis) acrylamide gel in TBE buffer at 80 W for 2 to 4 h. Gels were then stained with ethidium bromide.

mRNA levels from three heterozygous individuals from family 29 and from healthy control subjects (n = 3) were also evaluated by real-time PCR quantification (²⁹) on the ABI Prism 7700 platform (PE Applied Biosystems). Five μl of cDNA were amplified with 2.5 μl of glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) PDAR Kit (including forward and reverse primers as well as a specific VIC labeled TaqMan probe, PE Applied Biosystems), and 25 μl of TaqMan Universal MasterMix (PE Applied Biosystems) in a final volume of 50 μl. The amplification profile consisted of 50°C for 2 min, 95°C for 10 min and 40 cycles of 95°C for 15 s, then 60°C for 1 min. In parallel 5 μl cDNA were amplified using 25 μl of SYBR Green I (³⁰), TaqMan Universal Master Mix (PE Applied Biosystems), primers for SCR20A, in a final volume of 50 μl with an amplification profile of 2 min at 50°C, 10 min at 95°C and then 40 cycles of 15 s at 95°C, 20 s at 55°C, and 30 s at 72°C. The amplification curve was obtained using fluorescence values collected at 72°C.

After GAPDH normalization, mRNA levels of SCR20A (target gene) in heterozygous individuals were quantified as a percentage of mRNA levels in healthy control subjects. We applied the ΔΔCycles threshold (ΔΔCt) method (User bulletin #2, PE Applied Biosystems and references ^31,32,33). Melting temperature analysis (³⁰) was also performed: at the end of amplification reaction, samples were denatured at 95°C and then slowly cooled to 60°C in 20 min. The amplification product was verified by acrylamide gel. No primerdimers or aspecific amplification products were evidenced by melting temperature analysis and acrylamide gel.

Factor H and C3 and C4 Quantification

Factor H was quantified by radial immunodiffusion assay using a sheep polyclonal anti-human factor H antiserum, which reacts with various epitopes of factor H molecule (The Binding Site, Birmingham, UK). C3 and C4 were quantified by kinetic nephelometric measurements.

Western Blot Analysis of Factor H

To search for possible qualitative and quantitative defects of the different factor H molecules, Western blot analysis of factor H, factor H-like and factor H-related proteins in serum samples was also performed. Serum (1.5 μl) was separated by 10 or 7% sodium dodecyl sulfate polyacrylamide gel electrophoresis according to Laemmli using prestained bench markers (Life Technologies) as standards. Proteins were electroblotted to nitrocellulose by semidry blotting. Membranes were blocked for 30 min using 5% (wt/vol) dried milk in phosphate-buffered saline (PBS). Incubations with polyclonal goat anti-factor H antiserum (Calbiochem, San Diego, CA; diluted 1:1000) or anti FHL-1 rabbit anti-SCR1 to 4 antiserum, that does not detect FHR-1 and FHR-2 proteins (dilution 1:1000 (³⁴)) were performed at 4°C overnight. After the membranes were washed five times in PBS, they were incubated with peroxidase-conjugated rabbit anti-goat or swine anti-rabbit antibody (Dako, Hamburg, Germany) for 2 to 3 h. Protein bands were visualized by the addition of 0.3% (wt/vol) 4-chloro-1-naphthol in 10% (vol/vol) methanol in PBS.

Results

Patients

Family Pedigrees and Sporadic Case Report. From the database of the Italian Registry for Recurrent and Familial HUS/TTP (¹²), four families (^{1, 3, 24}, and ²⁹) were selected for the present study on the basis of persistent hypocomplementemia in all cases. A female with sporadic HUS (R16) who experienced disease recurrence in the renal allograft was also studied.

As reported in Figure 2, in families 1, 24, and 29, the onset of the disease was during infancy (from a few weeks to 8 yr), whereas in family 3, the disease became manifest in adulthood.

Figure 2: Pedigrees of the four families studied (indicated by #…) and haplotypes on chromosome 1q32 markers (D1S240, D1S202, D1S412, D1S2816, D1S413, D1S2738, D1S2796) flanking factor H gene. The pedigree and the haplotype (markers D1S2816, D1S413) of the sporadic case also are presented. Patients studied are indicated by a number (F… for familial R… for sporadic). Age at onset is indicated in italics; p, perinatal onset. Chromosomes carrying mutations are boxed.

Conditions that potentially predispose to the disease were recognized in all cases from families 1 and 24 (upper respiratory tract infection) and in patient F39#3 (pregnancy). One or more disease relapses were reported in all cases but one (F40#3). Eleven participants had died before the study was conducted. Among the seven survivors, four are on chronic dialysis; of these, two (F34#1 and F39#3) received a kidney transplant that failed because of recurrence of disease.

The pedigrees of the four families and haplotypes on 1q32 are shown in Figure 2. C3 concentrations were below the normal range in all patients and in some of their relatives (Table 1), whereas C4 concentrations were normal, indicating selective activation of the alternative complement pathway.

Table 1: Complement profiles^a

The pattern of inheritance of HUS could not be established in family 1, because only one patient was alive at the time of the study. No consanguinity was inferred. The pedigree was scarcely informative, with either the dominant or the recessive hypothesis (Table 2). In families 3 and 24, no consanguinity was reported; either recessive or dominant mode of inheritance could fit the haplotype data (Figure 2) and linkage results (Table 2). In family 29, a recessive inheritance was hypothesized because of the high grade of consanguinity; this was confirmed by the haplotype data (Figure 2) and logarithm of odds score values (Table 2). Results of SLINK simulation are reported in Table 2 as an index of the informativeness of the pedigrees.

Table 2: LOD scores obtained by two-point analysis between hemolytic uremic syndrome and seven markers around factor H^a

Sporadic Case. In 1997, a 48-yr-old woman, who had been on chronic hemodialysis since 1995 because of progressive loss of renal function after an HUS episode in 1980, received in our Transplant Department a cadaveric renal allograft, under cover of steroid plus cyclosporine-based immunosuppression. Graft function deteriorated acutely on day 7 post-transplantation, and allograft biopsy showed recurrent HUS. A partial improvement was achieved with plasma exchange, but this was not sustained and the graft failed 3 mo later. She is currently on chronic peritoneal dialysis. No relevant clinical history was reported in her family.

SSCP and Sequencing Results

PCR fragments corresponding to the 20 SCR of the factor H gene were analyzed for mutation screening by SSCP followed by sequencing of aberrant bands. SSCP detected a band shift in all cases and carriers from the four families as identified by haplotype analysis. The aberrant bands were localized in SCR8 in family 3 and in SCR20 in families 1, 24, and 29. The sporadic case showed an aberrant band in SCR20, which was also found in the father and two of five healthy siblings. None of the aberrant bands was found in 100 normal chromosomes.

Mutation Description and Biochemical Characterization of Factor H

Family 1. Sequence analysis revealed a heterozygous 3717 G to A transition; this results in an amino acid change of arginine to glutamine (R1215Q). This mutation was present in the patient and his unaffected father and his unaffected grand-father (Figure 3, panel A). Factor H protein levels (Table 1) and the Western blot profile were normal both in the case and his relatives (Figure 4A).

Figure 3: Partial chromatograms of SCR8 (panels B and B1) and 20 (panels A and C to F) sequences describing the mutations in our series. (Panel A) DNA from subject F34#1 (familial hemolytic uremic syndrome [HUS]), heterozygous G to A transition. (Panel B) DNA from subject F39#3 (familial HUS), heterozygous frame shift caused by an A deletion. (Panel B1) cDNA from the same patient: the heterozygous A deletion is present also in the cDNA. (Panel C) DNA from patient F106#24 (familial HUS), heterozygous C to T transition. (Panel D) DNA from patient F86#29 (healthy carrier), heterozygous A to T transversion and frame shift caused by a 24-bp deletion. (Panel E) DNA from patient F29#29 (familial HUS), homozygous A to T transversion and frame shift caused by a 24-bp deletion. (Panel F) DNA from patient R16 (sporadic HUS), heterozygous C to T transition.

Figure 4: Western blot analysis of factor H in human serum. Sera were separated by 7% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and the Western blots were developed with polyclonal rabbit anti-factor H antiserum, specific for factor H and FHL-1/reconectin. Closed arrows indicate factor H, and open arrows point to the abnormal high-molecular-weight bands in C and D. Patients are marked by an asterisk. The mobility of the size markers (kD) is indicated. (A) Family 1, sera from the patient still alive (F34) and from a healthy control subject are shown. (B) Family 3, sera from the two patients (F39 and F40), the carrier brother (F43), and two healthy relatives (F42 and F41) are shown. (C) Family 24, sera from the patients still alive (F106 and F108), the carrier father (F104), the healthy mother (F105), and the healthy sister (F107) are shown. (D) Sera from the patient with sporadic HUS (R16), the healthy mother (R22), the two carrier healthy siblings (R17 and R21), and a healthy control subject are shown.

Family 3. A heterozygous bp deletion was found in the two patients and one unaffected brother involving one of the three adenines from position 1494 to 1496; this mutation causes a frame shift that results in the formation of a short sequence of five anomalous codons and a premature stop codon that interrupts the transcription within SCR8 (Figure 3, panel B). The same mutation was evident also on the RT-PCR product of RNA extracted from PBMC, indicating that the mutated DNA is transcribed (Figure 3, panel B1).

Radial immunodiffusion assay found that factor H levels in carriers of the mutation were in the normal range (Table 1). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blot analysis (Figure 4B) of sera in this family showed that the 150-kD factor H band was reduced in the mutation carriers. The truncated protein was not identified in these patients by Western blotting; however, the truncated protein could have been masked by factor H-like 1 (²⁶), which derives from an alternative splicing of factor H primary transcript, as a result of the very close molecular weight of the two proteins.

Family 24. A heterozygous 3701 C to T transition that determines an arginine to cysteine change (R1210C) was found (Figure 3, panel C). This mutation is present in the two patients and their unaffected father. Thus, the patients seem to have inherited the abnormal factor H allele from their father. Although factor H protein levels in the patients' plasma were normal, Western blot experiments revealed additional high-molecular-weight bands that reacted with anti-factor H antiserum in the two affected patients and their healthy father. These bands were absent in their healthy mother and sister and in control subjects (Figure 4, C and D). The nature of these bands with lower mobility is not clear. One possible explanation could be the formation of factor H dimers caused by intermolecular disulphide bonding with the additional cysteine in SCR20.

Family 29. An A to T transversion and a 24-bp deletion was found in homozygosity in affected members of family 29 (Figure 3, panel E). This mutation was also present in heterozygosity (Figure 3, panel D) in all of the other family members, as supposed by haplotype analysis (Figure 2). The deletion causes a stop codon within SCR20 and the loss of the four most C-terminal amino acids.

Ying et al. (³⁵) described an apparent exchange at position 3645 in SCR20 in this Bedouin family. However, they used a set of primers that are complementary to both HF and FHR-1. While amplifying DNA from three healthy control subjects using this set of primers, we found the same base exchange as Ying et al. (³⁵) and a further nucleotide mismatch that exactly corresponds to FHR-1.

The factor H concentration was abnormally low in this family: factor H serum levels were severely depressed in the two affected patients (Table 1 and reference ¹²) and below the normal range in the heterozygous subjects. The Western blot patterns of factor H in this family confirmed the low protein levels in the two homozygous patients, although some residual protein was detectable (¹²).

We used PBMC to conduct RT-PCR analysis on SCR20 to determine how this deletion affected mRNA levels. Upon separation of the cDNA PCR products from heterozygous carriers by acrylamide gel electrophoresis, we found two bands of different intensities: a 178-bp band corresponding to the wild-type cDNA—as evidenced by cDNA analysis from normal control subjects—and a second 154-bp band corresponding to the deleted cDNA (Figure 5A). The intensity of the aberrant band was much lower than the wild-type band (Figure 5A), suggesting a selective defect in transcription or stability of the mutated mRNA. Thus, quantitative real-time PCR showed approximately a 50% reduction of mRNA levels in heterozygous individuals (Figure 5B). Melting temperature analysis showed two peaks in heterozygous subjects corresponding to the wild-type and the deleted cDNA. Unfortunately, RNA from homozygous patients was not available.

Figure 5: Quantification of mRNA from patients of family 29. (A) Electrophoresis of PCR products: lane 1, DNA from F29 (homozygous for the 24-bp deletion); lane 2, DNA from F86 (heterozygous for the 24-bp deletion); lanes 3 and 4, cDNA from F86 and F84 (heterozygous for the 24-bp deletion); lane 5, cDNA from a healthy control subject. Levels of the deleted mRNA are clearly low in the heterozygous patients. (B) Real-time PCR results: ^*mRNA levels are expressed as percentage of levels in healthy control subjects (n = 3) after normalization for the housekeeping gene GAPDH. Each sample was run in triplicate; results are expressed as mean ± SD of the three runs. ⁺, mean of three controls.

Sporadic Case. A heterozygous mutation was found in the patient who experienced a sporadic episode of HUS. The mutation, a 3701 C to T transition, results in the change of an arginine to cysteine (R1210C) and is identical to that found in family 24, (Figure 3, panel F). We were unable to find any familial relationship between the sporadic patient, who is from the north of Italy, and family 24, which is from the south. Analysis of markers D1S2816 and D1S413 showed that this patient does not share the haplotypes of family 24 (Figure 2); thus, the possibility of a common foundation event is unlikely.

The mutation was also found in the father and two of five healthy siblings. Serum factor H levels were within the normal range (Table 1). Western blot experiments revealed in the sporadic case the presence of the same additional high-molecular-weight bands as in family 24. These bands were also present in all relatives who carried the mutation (Figure 4D).

Discussion

In the present study, we identified five mutations in the factor H gene in four families and one sporadic case of HUS. Three are point mutations within the most C-terminus SCR of factor H (SCR20), resulting in single amino acid substitutions. Interestingly, an R1210C mutation was identified separately in one family and in one unrelated individual who developed a sporadic form of the disease. In another family, an R1215Q substitution was identified. A mutation in the same codon has been previously reported by Warwicker et al. (²⁵) in a family with HUS, which resulted, however, in a different amino acid exchange (R1215G).

We also identified two mutations that lead to the formation of truncated factor H proteins. Specifically, a nonsense mutation was found in one family, which introduces a premature stop codon. This mutation results in a truncated protein that consists of SCR1 to 8 and has the last twelve SCR deleted. A 24-bp deletion within SCR20 was found in a Bedouin family, resulting in a protein having the most C-terminal end truncated. The 24-bp deletion was present in homozygosity in patients and in heterozygosity in their relatives, consistent with a recessive mode of inheritance. This pattern distinguishes this family from all of the others, in which all affected members were heterozygous for factor H mutations, consistent with a dominant pattern. Two groups have already studied the Bedouin family for factor H mutations. Ying et al. (³⁵) described an apparent exchange of a single nucleotide at position 3645 in SCR20, and Buddles et al. (³⁶) reported the 24-bp deletion that we confirm here. Data reported by Ying are very likely artifacts due to co-amplification of the strongly homologous FHR-1 gene (³⁷), as the forward and reverse primers that they used are complementary to both factor H and FHR-1. This is confirmed by our findings in healthy control subjects that the same primers amplified a DNA fragment with the same base exchange described by Ying et al. and a further nucleotide mismatch that exactly corresponds to FHR-1.

Thus, the new mutations presented here, together with the previously published data (^25,36) (Figure 6), provide strong molecular evidence that factor H is involved in both the dominant and the recessive forms of HUS. None of the mutations were ever found in healthy control subjects, excluding the possibility that they are common polymorphisms in the normal population. The point mutations (reference ²⁵ and present data) all were inherited in an autosomal-dominant mode. By contrast, nonsense mutations gave rise to heterogeneous phenotypes; thus, a stop codon in SCR8 was associated with a dominant pattern of inheritance and late onset of the disease, whereas a stop codon in SCR20 produced a recessive pattern with a very early onset. Finally, a third nonsense mutation, located in SCR1, has already been found (²⁵) in a sporadic case of HUS with late onset.

Figure 6: Summary of all seven mutations reported in the factor H gene of HUS patients. Three mutations introduce premature stop codons within SCR1, SCR8, and SCR20; the other four mutations affect arginine residues (R1210 and R1215). With the exception of the deletion within SCR20 (Bedouin family), all mutations are heterozygous.

What seems to distinguish dominant and recessive forms are the biochemical and clinical consequences of the mutations: dominant inheritance is associated with normal factor H levels and an abnormal protein in the circulation. In these cases, an incomplete penetrance was observed; indeed, approximately 40% of carriers did not develop HUS. In most cases, the disease developed during late infancy or adulthood, often as a consequence of events such as infection or pregnancy, which are known to be triggering factors for HUS (^1,2). This suggests that in dominant forms, haploinsufficiency of factor H, which is compensated by the normal allele in basal conditions, requires intercurrent environmental or acquired factors to manifest fully. By contrast, the mutation that causes autosomal-recessive HUS in the Bedouin family resulted in a severe reduction of serum factor H levels, accompanied by very early onset of the disease (1 to 20 wk) and no recognized trigger. Factor H reduction in this family may be due either to impaired secretion of the mutant protein, consequent to the loss of a disulphide bridge essential for protein folding and secretion (³⁸), or to reduced mRNA levels as a result of premature arrest of translation (³⁹). The latter possibility is supported by our results of mRNA analysis in PBMC from heterozygous carriers, in which the expression of the mutated allele was lower than the normal allele.

A number of reports have claimed an association between factor H deficiency and recurrent (²²), atypical (²⁴), and familial (¹²) forms of HUS. However, factor H deficiency has also been associated with an increased tendency to other diseases, including systemic lupus erythematosus (⁴⁰), pyogenic infection susceptibility (⁴¹), and a patient with membranoproliferative glomerulonephritis with collagen III deposition in the mesangium (⁴²). In the latter patient, Ault et al. (⁴³) identified two point mutations in SCR9 and SCR16, which cause abnormal secretion and severe factor H deficiency.

Why different mutations in the same gene can cause different diseases cannot be fully clarified on the basis of the present and previous data, and further investigations on the genotype—phenotype correlations are required. However, the finding that five of the seven mutations described until now in HUS are located in SCR20 of complement factor H identifies this domain as a mutational hot spot for HUS. It also suggests the presence, in the very C-terminus of factor H, of highly relevant functional sites whose loss is involved in the pathogenesis of this disease.

Factor H binds C3b and controls the alternative pathways of complement activation in two ways. First, it accelerates the decay of the alternative pathway C3 convertase (C3bBb) by displacing factor B. Second, it acts as a cofactor for factor I—mediated inactivation of C3b (cofactor activity) (²⁶).

Factor H has at least three C3b binding domains, one of which resides within SCR 16 to 20 (⁴⁴). In addition, SCR7 and 19 to 20 contain two major binding sites for heparin and heparin-like proteins, such as sialic acid on glycoproteins and glycolipids (⁴⁵). All of these functions are very likely lost in mutated factor H in family 3, in which the entire 9 to 20 SCR segment was deleted. The three point mutations that cause arginine substitutions in SCR20, described here, might impair the heparin-binding capacity of factor H as this involves lysine and arginine residues (^46,47).

Interaction of factor H with sialic acids and other polyanions on human cells and tissues increases the affinity of factor H for C3b and enhances its inhibitory effect on the alternative pathway of complement activation (⁴⁸). This control step protects host cells from autolytic attack of the alternative complement pathway, because complement activation induces the deposition of C3b indiscriminately onto host and foreign particles. This applies particularly to endothelial cells, with their high surface density of heparin-like glycosaminoglycans (⁴⁸). Thus, when an environmental trigger activates the alternative complement pathway, modification of the heparin-binding capacity of mutated factor H might facilitate the occurrence of microvascular endothelial damage.

The present data represent a step forward in understanding the pathogenesis of the microangiopathic process in HUS. We propose that intercurrent exposure to agents that damage the vascular endothelium, such as certain viruses, bacteria, toxins, immunocomplexes and cytotoxic drugs (^1,2), may initiate local unrestricted complement activation within capillary vessels (^49,50). In normal conditions, however, by modulating C3bBb activity (^26,27,28), factor H may efficiently limit complement deposition, thus preventing any further enhancement of the activation. On the contrary, when the bioavailability or the activity of factor H is congenitally defective, C3bBb convertase formation and complement activation become uncontrolled and result in microangiopathic damage, which leads to full manifestation of the disease.

No specific therapy is effective in familial cases of HUS. Infusion of fresh-frozen plasma, or plasmapheresis, which is effective in recurrent and atypical adult cases, often is inactive in patients with the familial form of the disease (^1,5), possibly because of insufficient amounts of factor H in the plasma infused. Replacement with recombinant factor H protein or the use of specific complement inhibitors could represent future perspectives for the treatment of the disease as alternatives to whole plasma.

Appendix

Italian Registry for Recurrent and Familial HUS/TTP

Coordinators. P. Ruggenenti, MD, M. Noris, Chem. Pharm. D., G. Remuzzi, MD (Clinical Research Centre for Rare Diseases Aldo e Cele Dacco[Combining Grave Accent], Ranica).

Investigators that referred cases to the Registry. F. Casucci, MD, F. Cazzato, MD (“Miulli” Hospital, Acquaviva delle Fonti, Bari); P. Ponce, MD (Garcia de Orta Hospital, Bairro do Motadoro, Portugal); F. Della Grotta, MD (Polyspecialized Hospital, Anzio, Roma); R. Bellantuono, MD, T. De Palo, MD (“Giovanni XXIII” Pediatric Hospital, Bari); D. Landau, MD (Soroka Medical Center, Beer-Sheba, Israel); T. Barbui, MD (Riuniti Hospitals, Bergamo); R. Wens, MD (CHU Brugmann, Bruxelles, Belgium); J. Ferraris, MD (Italian Hospital, Buenos Aires, Argentina); A.Cao, MD (“Universita[Combining Grave Accent] degli Studi,” Cagliari); C. Cascone, MD, G. Delfino, MD (“S. Giacomo” Hospital, Castelfranco Veneto, Treviso); T. Ring, MD (Aalborg Hospital, Denmark); C. Cappelletti, MD, C. Ciccarelli, MD (“S. Giovanni di Dio” Hospital, Firenze); G.C. Barbano, MD (“G. Gaslini” Institute, Genova); G.B. Haycock, MD (Guy's Hospital, London, United Kingdom); A. Bettinelli, MD (“G. e D. De Marchi” Pediatric Clinic, Milano); Edefonti, MD (“G. e D. De Marchi” Pediatric Clinic, Milano); E. Rossi, MD (“L. Sacco” Hospital, Milano); V. Toschi, MD (“San Carlo Borromeo” Hospital, Milano); S. Bassi, MD (“Umberto I” Hospital, Brescia); D. Belotti, Biol. Sci. D., E. Pogliani, MD (“San Gerardo” Hospital, Monza, Milano); A. Indovina, MD, R. Marceno[Combining Grave Accent], MD (“V. Cervello” Hospital, Palermo); G. Enia, MD (Division of Nephrology, Fisiology Clinic Centre, CNR, Reggio Calabria); F. Mallamaci, MD, C. Zoccali, MD (“Bianchi-Melacrino-Morelli” Hospital, Reggio Calabria); A. Amendola, MD, F. Mandelli, MD (“Universita[Combining Grave Accent] degli Studi La Sapienza,” Roma); A. Gianviti, MD, G.F. Rizzoni, MD (“Bambino Gesu[Combining Grave Accent]” Pediatric Hospital, Roma); T. Cicchetti, MD, G. Putorti[Combining Grave Accent], MD (“N. Giannettasio” Hospital, Rossano Calabro); A. Pinto, MD (“San Giovanni di Dio e Ruggi D' Aragona” Hospital, Salerno); E. Nesti, MD (S. Miniato Hospital, Firenze); M. Sanna, MD (Hospital of Sassari); A. Amore, MD, R. Coppo, MD (“Regina Margherita” Pediatric Hospital, Torino); A. Khaled, MD (“S. Chiara” Hospital, Trento); O. Amatruda, MD (“Fondazione Macchi” Hospital, Varese).

Laboratory Analysis. A. Crippa, MD, A. Vernocchi, MD (Riuniti Hospitals, Bergamo); F. Gaspari, Chem. D. (Clinical Research Centre for Rare Diseases Aldo e Cele Dacco[Combining Grave Accent], Ranica, Bergamo).

We thank Dr. Piero Ruggenenti, Dr. Beatrice Vasile and Dr. Elena Bresin for clinical advice and Dr. Silvia Orisio for her contribution in setting up the methods. Fabio Sangalli helped us for statistical analysis of linkage data. Dr. Judy Baggot edited the manuscript. We deeply are indebted to all participating members of the families and to their referring clinicians: Dr. C. Cascone, Dr. D. Landau, Dr A. Edefonti, Dr. G. Barbano, and Dr. S. Bassi.

We are grateful to the Comitato 30 ore per la vita and to Credito Bergamasco for supporting the study.

Dr. Jessica Caprioli is a recipient of the fellowship in memory of Maria Paola Casirati. Dr. Paola Bettinaglio is a recipient of the fellowship from Associazione Ricerca Malattie Rare (ARMR, Delegazione di Milano) in memory of Prof. Cosimo Conterno. Dr. Barbara Amadei received a fellowship from Associazione Ricerca Malattie Rare (ARMR, Bergamo) through the generosity of Vin Service Srl, Zanica, Bergamo. Dr. Peter F. Zipfel's work is supported by the Deutsche Forschungsgemeinschaft (DFG) Zi 432.

Portions of this work were presented at the XVIIIth International Complement Workshop, July 23-27, 2000, and at the 33rd Annual Meeting of the American Society of Nephrology, October 13-18, 2000.