We congratulate Ying et al.(1999) on their study demonstrating failure of secretion of factor H in a family with autosomal recessive hemolytic uremic syndrome (HUS)(MIM 235400). Their study indicates factor H (HF) involvement in the recessive form of the disease as well as the autosomal dominant form (Warwicker et al. 1998). HF is a member of the regulators of complement activity gene cluster that includes four factor H–related proteins (FHR1-4)(Zipfel and Skerka 1994). All secreted proteins encoded by this gene family are organized into repetitive elements known as short consensus repeats (SCRs) or complement control protein modules, and, generally, each SCR is encoded by one exon. We have found it difficult to design HF exon 20 specific primers because of the sequence similarity between HF exons 18–20 and FHR1 exons 3–5. We have now designed an intronic forward primer and a reverse primer in the 3 untranslated region and have confirmed their specificity by sequencing PCR products. Ying et al. state that they have identified a point mutation in HF exon 20, the last coding exon of the gene. We were also sent DNA samples from two affected individuals from this Bedouin pedigree and found them to be homozygous for microsatellites spanning HF. We amplified exon 20 by means of the primers shown in figure 1. Electrophoresis of the product through a 2% agarose gel indicated a single smaller product, and sequencing revealed a T substitution and a 24-bp deletion. The effect of this is to replace the last seven amino acids of the protein with three different amino acids, as shown in figure 2. Crucially, this deletes the final cysteine residue. Each SCR has four cysteine residues at conserved positions that form two disulphide bridges essential for protein folding (Janatova et al. 1989). Mutations that disrupted these disulphide bridges led to a profound secretion block in a child with HF deficiency and membranoproliferative glomerulonephritis (Ault et al. 1997; Schmidt et al. 1999.

The change reported by Ying et al. is present in FHR1 exon 5, although these investigators did not observe the second nucleotide substitution in FHR1 exon 5. The forward and reverse primers used by Ying et al. are complementary to both HF and FHR1 (fig. 1). The reverse primer overlaps the deletion in the Bedouin family (fig. 1) and, hence, these primers will amplify FHR1 exon 5 but not HF exon 20 in the affected individuals. In the discussion, the authors suggest that the recessive form of HUS results from HF deficiency and the dominant form results from the presence of abnormal HF protein in serum. However, in the isolated case that we reported, with an exon 1 mutation, the HF levels were half the normal level. Why did our patient develop HUS whereas the heterozygotes in the Bedouin family did not, and why did the child with abnormal HF secretion described by Ault develop membranoproliferative glomer-