An aberrant form of CD59 derived from HeLa cells

We isolated a CD59 cDNA from a HeLa cell library which encoded a mutated form of CD59, having a single base substitution (G to T) that changed Arg55 to Met. Since this mutation occurred in the vicinity of the putative active site of CD59, we expressed the aberrant form of the protein in Chinese hamster ovary cells in order to test for effects upon function. We found that the mutation did not influence complement inhibitory activity of CD59. However, the epitopes recognised by the function-blocking CD59 monoclonal antibodies BRIC229 and YTH 53.1 were significantly affected. The G to T substitution caused loss of an Mnl I restriction site which permitted PCR-RFLP analysis. All of 52 human subjects studied, and our in-house HeLa cells, were homozygous for the normal CD59 sequence, indicating that the altered sequence was not due to normal variation in the general population. Therefore this mutation probably arose spontaneously in the HeLa cell line used to generate the commercially obtained cDNA library.

Reference typing report for complement components C6, C7 and C9 including mutations leading to deficiencies

The results of the present (VIIth Complement Genetics Workshop and Conference, Mainz, May 1998) and past reference typing workshops for the terminal complement components C6, C7 and C9 are compiled and discussed both on the protein level and on the DNA level. This report also focuses on the molecular bases of expressed and silent polymorphisms and reviews the molecular bases of subtotal and complete deficiencies of these proteins and their associations with protein and DNA markers. The results of the protein typing for C6 are published in the following paper of this issue.

Five new polymorphisms in the complement C7 gene and their association with C7 deficiency

Five new polymorphisms in the C7 gene are described: 2 in intron 1, and 1 each in introns 7, 8 and 15. Four of these are single nucleotide exchanges, while the fifth is a T insertion at 10 sequential Ts. Allele frequency data are presented for intervening sequence (IVS)1+ 55 in 6 normal population groups. We present new and updated data in these populations on a previously described C7 polymorphism in exon 13 (cDNA 1792 A/T). We also report the extended haplotypes associated with C7 deficiency for which marker investigation is a useful, and in some cases vital, adjunct to the identification of the gene defects. Almost without exception, a particular haplotype is associated with a particular mutation causing the deficiency state. Haplotyping is especially useful where polymerase chain reaction failure on one chromosome could be a cause for difficulties in detecting a molecular defect due to heterozygosity for large deletions or unidentified variations at the locations of the primers.

C7 deficiency in an Irish family: a deletion defect which is predominant in the Irish

Human deficiencies of terminal complement components are known to be associated with increased susceptibility to Neisseria meningitidis infection. Polymorphic DNA marker studies in complement deficient investigations allow identification of haplotypes associated with the deficiency and enable the possible identification of heterozygote carriers of the defect. We report studies of an Irish family in which the index case had suffered recurrent meningococcal disease and was found to be deficient in the seventh component of complement (C7). The availability of all family members enabled us to determine the segregating haplotypes. The defects in the family segregated with two very closely related C6 and C7 DNA haplotypes, one of which is known to be associated with the large Irish C7 DNA deletion defect. The index case and two C7 deficient siblings were found to be homozygous for this defect, a deletion that spans approx. 6.8 kbp and encompasses exons 7 and 8. The deletion defect of exons 7 and 8 of C7 has been found in homozygous form in another C7 deficient Irish individual, and is present in heterozygous form in C7 deficient members of a third Irish family. Therefore, this deletion defect occurs in five of the six deficient chromosomes of these three unrelated Irish families, raising the interesting question of how prevalent this defect may be within the Irish community.

Complement C7 deficiency: seven further molecular defects and their associated marker haplotypes

Seven further molecular bases of C7 deficiency are described. All these new molecular defects involve single-nucleotide events, deletions and substitutions, some of which alter splice sites, and others codons. They are distributed along the C7 gene, but predominantly towards the 3' end. All were found in compound heterozygous individuals. The C6/C7 marker haplotypes associated with most C7 defects are tabulated.

The molecular basis of C6 deficiency in the western Cape, South Africa

Deficiency of the sixth component of human complement (C6) has been reported in a number of families from the western Cape, South Africa. Meningococcal disease is endemic in the Cape and almost all pedigrees of total C6 deficiency (C6Q0) have been ascertained because of recurrent disease. We have sequenced the expressed exons of the C6 gene from selected cases and have found three molecular defects leading to total deficiency: 879delG, which is the common defect in the Cape and hitherto unreported, and 1195delC and 1936delG, which have been previously reported in African-Americans. We also show that the 879delG and 1195delC defects are associated with characteristic C6/C7 region DNA marker haplotypes, although small variations were observed. The 1936delG defect was observed only once in the Cape, but its associated haplotype could be deduced. The data from the haplotypes indicate that these three molecular defects account for the defects in all the 38 unrelated C6Q0 individuals we have studied from the Cape. We have also observed the 879delG defect in two Dutch C6-deficient kindreds, but the 879delG defect in the Cape probably did not come from The Netherlands.

Molecular bases of C7 deficiency: three different defects

The molecular basis of C7 deficiency has been investigated in two Irish families and a number of Israeli families of Moroccan Sephardic Jewish origin. Exon PCR and sequencing revealed a heterozygous point mutation at the 3' splice acceptor site of intron 1 in one Irish family. In the other Irish family, exons 7 and 8 failed to amplify and they were shown to be deleted. Marker haplotype studies of the C6 and C7 gene region and Southern blots show that the Irish family with the splice defect also segregate for the deletion, which is not easily detected in heterozygotes. The Israeli C7-deficient cases all share a C7 haplotype and are homozygous for a mis-sense mutation in exon 9. However, one individual is heterozygous for markers at adjacent C6 loci, showing that there has been an intergenic recombination and suggesting that the deficiency mutation is of appreciable antiquity.

Molecular bases of C7 deficiency: three different defects

The molecular basis of C7 deficiency has been investigated in two Irish families and a number of Israeli families of Moroccan Sephardic Jewish origin. Exon PCR and sequencing revealed a heterozygous point mutation at the 3' splice acceptor site of intron 1 in one Irish family. In the other Irish family, exons 7 and 8 failed to amplify and they were shown to be deleted. Marker haplotype studies of the C6 and C7 gene region and Southern blots show that the Irish family with the splice defect also segregate for the deletion, which is not easily detected in heterozygotes. The Israeli C7-deficient cases all share a C7 haplotype and are homozygous for a mis-sense mutation in exon 9. However, one individual is heterozygous for markers at adjacent C6 loci, showing that there has been an intergenic recombination and suggesting that the deficiency mutation is of appreciable antiquity.

An unusual combined insertion/deletion polymorphism in intron 10 of the human complement C6 gene

Investigation of intron 10 of the human complement C6 gene revealed an unusual combined insertion/deletion polymorphism at position 493: the subsequent 6 bp is deleted and is substituted by a different 26 bp, giving a net gain of 20 bp. The variant shows autosomal co-dominant inheritance. The 26 bp insertion is homologous to a human endogenous retrovirus-type sequence and could tentatively be ascribed to a retroposon. Alternatively, the presence of three copies of a 5 bp direct repeat, an 8 bp palindrome and a 12 bp split symmetrical element could suggest an endogenous, sequence-mediated mutational process. Polymorphisms of this type are extremely rare, although there are several examples of such mutations causing disease.

DNA haplotypes of the complement C6 and C7 genes associated with deficiencies of the seventh component; and a new DNA polymorphism in C7 exon 13

Eight common DNA polymorphisms have been described for the linked C6 and C7 genes. We now describe a ninth polymorphism in C7 exon 13 which is located in a tight cluster with two previously reported markers. We have used all these markers to investigate the heterogeneity of C7 deficiency. Five of the nine C7 deficient probands (resident in Ireland, South Africa, Russia and Israel) are heterozygous for C6/C7 haplotypes. Seven different C7 deficient haplotypes were found for C7 markers alone, but all the four Israelis share one and three out of four Irish haplotypes share another. The markers appear to be a good guide to the heterogeneity of C7 deficiency and have been useful in choosing homozygous subjects for the investigation of molecular defects.

Difficulties in the ascertainment of C9 deficiency: lessons to be drawn from a compound heterozygote C9-deficient subject

A group of patients with long-surviving mismatched kidney allografts were investigated for complement function using haemolytic assays in agarose gels. One patient was found to have no alternative pathway activity but a low normal classical pathway. Surprisingly, investigation revealed that the patient's complement was normal for all components except C9, which was functionally absent. The patient was shown to be heterozygous for DNA markers in the C6, C7 and C9 region of chromosome 5 and therefore appears to be a compound heterozygote for two uncharacterized C9 deficiency genes. Serological analysis by ELISA revealed that he has trace concentrations of a non-functional C9 molecule. Western blot analysis was not sufficiently sensitive to permit detection of this molecule. We hypothesize that the patient is heterozygous for a complete deficiency of C9 and for a gene directing hyposynthesis of a defective C9. We also suggest that C9 deficiency may be more common among Caucasians than has been reported.

The human complement C9 gene: identification of two mutations causing deficiency and revision of the gene structure

The ninth component of human complement (C9) is the last of the terminal complement components creating the membrane attack complex. C9 is a single-chain serum protein that is encoded by a gene located on chromosome 5p. Deficiency of terminal complement components is generally associated with recurrent neisseria infections. We studied a previously described Swiss family with inherited C9 deficiency. To identify the genetic basis of C9 deficiency, we developed an approach using exon-specific PCR and direct DNA sequencing. As a cause of C9 deficiency, we found two different point mutations, both generating TGA stop codons in the coding sequence. One mutation, a C to A exchange, was detected in exon 2 at cDNA position 166, the other, a C to T exchange, was located in exon 4 (cDNA position 464). In family studies of three first-degree relatives with heterozygous C9 deficiency, we demonstrated that the two mutations are segregating independently. Therefore, these mutations are sufficient to explain the complete deficiency of both the probands studied. DNA sequencing of the exon-intron junctions revealed a number of revisions regarding the boundaries between exons 4, 5, and 6 as well as between exons 10 and 11. No additional introns were detected in exons 6 and 10. Furthermore, DNA marker studies were conducted using known polymorphisms of the C6, C7, and C9 genes, confirming the linkage of the observed C9 mutations with defined haplotypes.

The human complement C9 gene: identification of two mutations causing deficiency and revision of the gene structure

The ninth component of human complement (C9) is the last of the terminal complement components creating the membrane attack complex. C9 is a single-chain serum protein that is encoded by a gene located on chromosome 5p. Deficiency of terminal complement components is generally associated with recurrent neisseria infections. We studied a previously described Swiss family with inherited C9 deficiency. To identify the genetic basis of C9 deficiency, we developed an approach using exon-specific PCR and direct DNA sequencing. As a cause of C9 deficiency, we found two different point mutations, both generating TGA stop codons in the coding sequence. One mutation, a C to A exchange, was detected in exon 2 at cDNA position 166, the other, a C to T exchange, was located in exon 4 (cDNA position 464). In family studies of three first-degree relatives with heterozygous C9 deficiency, we demonstrated that the two mutations are segregating independently. Therefore, these mutations are sufficient to explain the complete deficiency of both the probands studied. DNA sequencing of the exon-intron junctions revealed a number of revisions regarding the boundaries between exons 4, 5, and 6 as well as between exons 10 and 11. No additional introns were detected in exons 6 and 10. Furthermore, DNA marker studies were conducted using known polymorphisms of the C6, C7, and C9 genes, confirming the linkage of the observed C9 mutations with defined haplotypes.

Molecular bases of combined subtotal deficiencies of C6 and C7: their effects in combination with other C6 and C7 deficiencies

Combined subtotal deficiency of C6 and C7, in which both proteins are expressed at very low levels, has been observed in homozygous form in two families. A defect at the 5' splice donor site of intron 15 of the C6 gene explains the low molecular weight of the C6 protein and is probably responsible for its low expressed concentration. The C7 defect is more enigmatic: the protein is of normal molecular weight, low circulating concentration, and altered isoelectric point. An Arg > Ser codon substitution in exon 11 is the only molecular alteration within the mature C7 protein. These defects are associated with a characteristic set of polymorphic DNA markers in the C6/C7 region, forming a distinct haplotype. The haplotype has been found in combination with a number of other haplotypes containing defective genes that lead either to C6 or C7 deficiency, but with different consequences. Where it is combined with a C6-deficient gene, the serum C7 levels can be surprisingly high, possibly because there is no C6 generating C56 to consume the C7. In contrast, where the C7 genes are both defective (but still partially functional), there may be a profound deficit of circulating C7 because there is ample C6 to produce C56 and consume the already small amount of C7. Each molecular defect has also been found in isolation and has the expected effect.

Molecular bases of combined subtotal deficiencies of C6 and C7: their effects in combination with other C6 and C7 deficiencies

Combined subtotal deficiency of C6 and C7, in which both proteins are expressed at very low levels, has been observed in homozygous form in two families. A defect at the 5' splice donor site of intron 15 of the C6 gene explains the low molecular weight of the C6 protein and is probably responsible for its low expressed concentration. The C7 defect is more enigmatic: the protein is of normal molecular weight, low circulating concentration, and altered isoelectric point. An Arg > Ser codon substitution in exon 11 is the only molecular alteration within the mature C7 protein. These defects are associated with a characteristic set of polymorphic DNA markers in the C6/C7 region, forming a distinct haplotype. The haplotype has been found in combination with a number of other haplotypes containing defective genes that lead either to C6 or C7 deficiency, but with different consequences. Where it is combined with a C6-deficient gene, the serum C7 levels can be surprisingly high, possibly because there is no C6 generating C56 to consume the C7. In contrast, where the C7 genes are both defective (but still partially functional), there may be a profound deficit of circulating C7 because there is ample C6 to produce C56 and consume the already small amount of C7. Each molecular defect has also been found in isolation and has the expected effect.

How partial C7 deficiency with chronic and recurrent bacterial infections can mimic total C7 deficiency: temporary restoration of host C7 levels following plasma transfusion

An apparently completely complement C7-deficient patient with refractory otitis media and two episodes of meningococcal disease was given therapeutic plasma transfusions in 1992 and 1994. Following these transfusions unexpected changes were found in C7 levels. Immediately after transfusion the serum C7 levels failed to rise to the expected levels but then rose to 5-10% of the normal mean during the next 5 days and remained at that level for more than 2 weeks before eventually returning to zero. The patient's DNA genotyped C7 M, and therefore C7 N donor plasma was selected for the second transfusion to allow identification of the source of the C7 circulating post-transfusion. This C7 phenotyped C7 M, demonstrating it to be of recipient origin. Therefore, the apparently completely C7-deficient patient was able to secrete some C7. By a combination of DNA typing and isoelectric focusing of the C7 appearing after transfusion, it was demonstrated that the patient was heterozygous for combined subtotal C6/C7 deficiency (inherited from his father) and a different, so far uncharacterized, subtotal C7 deficiency (inherited from his mother). The low amount of C7 secreted appeared to be constantly consumed, probably by generation of C5b6 as a result of his chronic infection. He had been shown to have circulating C5b6 most of the time, and thus only when sufficient exogenous C7 was given to consume the free C5b6 did his own C7 appear in circulation.

A new intronic polymorphism in the C7 gene 36 bp from the common expressed C7 M/N polymorphism

We report a new polymorphism in the complement C7 gene that results from an A-C transversion in intron 12, 27 bp upstream of exon 13 (C712.-27) and 36 bp upstream of the point mutation that underlies the C7 M/N antigenic polymorphism. The C7 12.-27 polymorphism subdivides C7 M haplotypes, but not C7 N. It also sheds light on the evolution of the various types of deficiency genes at the adjacent C6 locus.

Complement C6 and C7 DNA polymorphisms analysed by PCR in seven ethnic groups and characterisation of the C6 MspI RFLP

Five polymorphisms in the C6 and C7 genes have been investigated in seven ethnic groups. The allele frequencies are broadly similar in most groups except C7 M/N which is monomorphic in our group of Africans, and C6 MspI and C7 S367T where the allele frequencies in African and Cape Coloured subjects are very different from the other ethnic groups. There is very little allelic association except between C6 A/B and C6 MspI. Seventeen of the 32 possible haplotypes have been observed, suggesting that much recombination has taken place. We describe a new method for the investigation of the MspI RFLP located in intron 3 of C6 (approximately 3 kbp 3' from exon 3 and 1.5 kbp 5' from exon 4) and its molecular basis, together with an improved method for the isolation of DNA from stored serum.

Importance of the third thrombospondin repeat of C6 for terminal complement complex assembly

The anti-C6 monoclonal antibody WU 6-4 was shown to be unequivocally native restricted since it neither binds to the terminal complement complex (TCC) nor to C5b6. In addition, it was shown to inhibit TCC formation by interfering with C5b6 generation. Using the pUEX expression system and C6 cDNA the WU 6-4 epitope was mapped to the third thrombospondin repeat of C6. This protein domain may therefore contribute to the C5 binding site of C6 and be involved in terminal complement complex assembly. The presence of the epitope in rabbit C6 indicates a useful model for studying inhibition of TCC formation in vivo.

Structure of the human C7 gene and comparison with the C6, C8A, C8B, and C9 genes

The seventh component of complement is a single chain plasma glycoprotein that is involved in the cytolytic phase of complement activation. We have determined the structure of the C7 gene, which is encoded by 18 exons whose sizes vary from 56 to 244 bp. For the most part, the exons do not correspond to the protein homology units. However, two intron/exon boundaries occur at junctions between different functional parts of the protein. The first is at a site between the end of the C9 homology unit and the carboxyl-terminal extension which is also a feature of C6. The second of these boundaries occurs between the regions encoding two pairs of cysteine-rich modules (the short consensus repeats and the factor I modules) located in the carboxyl-terminal part of C7. In contrast to the exons, the introns range considerably in size from 0.5 to 8.5 kbp. The complete analysis indicates that the gene encoding C7 is approximately 80 kbp in length. We show here that the C7 gene is highly homologous to that for C6, and also to C8A, C8B, and C9, confirming and extending the published data. With the exception of exon 1, all intron/exon boundaries are preserved with respect to phase when compared with C6.

Structure of the human C7 gene and comparison with the C6, C8A, C8B, and C9 genes

The seventh component of complement is a single chain plasma glycoprotein that is involved in the cytolytic phase of complement activation. We have determined the structure of the C7 gene, which is encoded by 18 exons whose sizes vary from 56 to 244 bp. For the most part, the exons do not correspond to the protein homology units. However, two intron/exon boundaries occur at junctions between different functional parts of the protein. The first is at a site between the end of the C9 homology unit and the carboxyl-terminal extension which is also a feature of C6. The second of these boundaries occurs between the regions encoding two pairs of cysteine-rich modules (the short consensus repeats and the factor I modules) located in the carboxyl-terminal part of C7. In contrast to the exons, the introns range considerably in size from 0.5 to 8.5 kbp. The complete analysis indicates that the gene encoding C7 is approximately 80 kbp in length. We show here that the C7 gene is highly homologous to that for C6, and also to C8A, C8B, and C9, confirming and extending the published data. With the exception of exon 1, all intron/exon boundaries are preserved with respect to phase when compared with C6.

Molecular basis of the complement C7 M/N polymorphism. A neutral amino acid substitution outside the epitope of the allospecific monoclonal antibody WU 4-15

The allotypes of the C7 M/N polymorphism are determined by comparing the ELISA reaction pattern of the allospecific mAb WU 4-15 with polyclonal anti-C7 IgG. To characterize the molecular basis of this polymorphism, the WU 4-15 epitope was mapped by expression of cDNA fragments. It was found to be located within the boundary region of the two short consensus repeats of C7 (amino acid residues 595-612). Coincidentally an A or C substitution was found in the course of investigation of the gene structure at nucleotide 1759. This leads to an electrically neutral Pro/Thr substitution at residue 565 and alternative Sau961 or RsaI restriction sites, respectively. Enzymatic digestion or sequencing of PCR-amplified genomic DNA from C7 M/N-typed individuals showed a perfect correspondence of the genotype with the phenotype of the C7 M/N polymorphism. In contrast, no association was found with the genotypes of a second polymorphic site at residue 367. Thus, a threonine at amino acid residue 565 allows access of WU 4-15 to its epitope but its substitution with proline almost completely inhibits access in the native molecule, although the epitope is about 40 amino acid residues distant from the polymorphic site in the primary structure. Furthermore, the different conformational structure is probably responsible for the hypomorphic appearance of C7*N.

Molecular basis of the complement C7 M/N polymorphism. A neutral amino acid substitution outside the epitope of the allospecific monoclonal antibody WU 4-15

The allotypes of the C7 M/N polymorphism are determined by comparing the ELISA reaction pattern of the allospecific mAb WU 4-15 with polyclonal anti-C7 IgG. To characterize the molecular basis of this polymorphism, the WU 4-15 epitope was mapped by expression of cDNA fragments. It was found to be located within the boundary region of the two short consensus repeats of C7 (amino acid residues 595-612). Coincidentally an A or C substitution was found in the course of investigation of the gene structure at nucleotide 1759. This leads to an electrically neutral Pro/Thr substitution at residue 565 and alternative Sau961 or RsaI restriction sites, respectively. Enzymatic digestion or sequencing of PCR-amplified genomic DNA from C7 M/N-typed individuals showed a perfect correspondence of the genotype with the phenotype of the C7 M/N polymorphism. In contrast, no association was found with the genotypes of a second polymorphic site at residue 367. Thus, a threonine at amino acid residue 565 allows access of WU 4-15 to its epitope but its substitution with proline almost completely inhibits access in the native molecule, although the epitope is about 40 amino acid residues distant from the polymorphic site in the primary structure. Furthermore, the different conformational structure is probably responsible for the hypomorphic appearance of C7*N.

Complement component C6 and C7 haplotypes associated with deficiencies of C6

Both complete C6-deficiency (C6*Q0) and subtotal C6-deficiency (C6*SD) have been described as simple recessive traits and C6*SD has been described in combination with subtotal deficiency of the C7 coded at an adjacent locus. The trace of C6 protein found in both C6*SD traits is phenotypically indistinguishable, being smaller than normal C6 and having different isoelectric properties. A defect has been found in the C6 gene which plausibly explains the C6*SD phenotype, and this defect is also common to both C6*SD traits. We present data from seven DNA markers of the C6 and C7 genes which show that although at least four haplotypes are associated with C6*Q0, most South African C6*Q0 patients carry a common defective haplotype. The most common haplotype associated with C6*Q0 has been observed only once among unaffected haplotypes of relatives. In one family, the cases of C6*SD share a complete haplotype with both cases of combined deficiency and are probably heterozygous for this condition and complete deficiency of C6. In another family, the C6*SD is on a slightly different haplotype and C7 is normally expressed. Thus, the C6 defect is not sufficient on its own to explain the C7 deficiency in the combined deficient haplotype. The haplotype associated with the combined deficiency is found not only in normal control subjects, but also in one case of complete C6 deficiency. In this case the molecular defect seen in combined or C6*SD cases is absent.

Molecular basis of subtotal complement C6 deficiency. A carboxy-terminally truncated but functionally active C6

Individuals with subtotal complement C6 deficiency possess a C6 molecule that is 14% shorter than normal C6 and present in low but detectable concentrations (1-2% of the normal mean). We now show that this dysmorphic C6 is bactericidally active and lacks an epitope that was mapped to the most carboxy-terminal part of C6 using C6 cDNA fragments expressed as fusion proteins in the pUEX expression system. We thus predicted that the abnormal C6 molecule might be carboxy-terminally truncated and sought a mutation in an area approximately 14% from the carboxy-terminal end of the coding sequence. By sequencing PCR-amplified products from this region, we found, in three individuals from two families, a mutation that might plausibly be responsible for the defect. All three have an abnormal 5' splice donor site of intron 15, which would probably prevent splicing. An in-frame stop codon is found 17 codons downstream from the intron boundary, which would lead to a truncated polypeptide 13.5% smaller than normal C6. This result was unexpected, as earlier studies mapped the C5b binding site, or a putative enzymatic region, to this part of C6. Interestingly, all three subjects were probably heterozygous for both subtotal C6 and complete C6 deficiency.

DNA polymorphisms of the complement C6 and C7 genes

The linked C6 and C7 loci are rich in genetic markers, both at the protein and DNA levels. There are now seven common DNA polymorphisms distributed over about 300 kbp of chromosome 5p12-14. We report a new TaqI RFLP for C7 and a method for typing a C7 variant (T368S) hitherto known only from cDNA clones. We have re-investigated the published RFLPs to provide information on their frequency in North European Caucasian (predominantly British and Irish) subjects and have revised some of the published parameters, especially the sizes of polymorphic restriction fragments. Their precise locations within the genes are also reported: the three markers for C6 are in exon 3, intron 3 and adjacent to exon 17 and the four markers for C7 are in introns 15 and 13 and in exons 13 and 9. The gene frequencies of the second commonest allele of all seven markers lie in the range 0.2 to 0.37, except C6 A/B in the Japanese, where the frequencies of both common alleles are about 0.45. We have estimated the gene frequencies for the DNA polymorphisms which correlate with C7 M/N phenotype and for the C6 A/B phenotype and find them to be the same as the phenotypic estimates in Caucasians and in the Japanese respectively. The markers provide the possibility of defining 128 haplotypes, many (28) of which have been observed. Allelic associations in these genes are generally surprisingly weak.

C6 haplotypes: associations of a Dde I site polymorphism to complement deficiency genes and the Msp I restriction fragment length polymorphism (RFLP)

Complement C6 has a common charge polymorphism designated A and B with gene frequencies of 0.65 and 0.35. The probable molecular basis for this is a Glu (C6A) for Ala (C6B) substitution at amino acid position 98, and is detected by digestion with the restriction enzyme Dde I of a polymerase chain reaction (PCR)-amplified fragment of genomic DNA. C6A was found to be Dde I-positive and C6B corresponds to Dde I-negative. We have applied our Dde I A/B polymorphism genotyping method to the investigation of C6-deficient individuals with complete (C6Q0) and sub-total deficiency (C6SD) protein phenotypes, including members of four families. We have also investigated the RFLP detected by digestion of genomic DNA with the enzyme Msp I, which is due to a polymorphic site located in the 5' section of the gene, the variable sequence of which has yet to be determined. Sixteen out of seventeen unrelated C6Q0 subjects were found to be genotypically Dde I B/Msp I-negative; the remaining subject was heterozygous at both the loci under investigation. The C6SD phenotype was found to be associated with the Dde I A/Msp I-positive genotype in two families with combined C6/C7 subtotal deficiency and two with C6SD. It can be concluded that the two forms of C6 deficiency, C6Q0 and C6SD, arose independently on two different C6 allelic backgrounds. These associations have allowed the genotyping of the rare families that contain both types of deficiency. We have also defined a number of normal C6 Dde I/Msp I haplotypes in Caucasians and Cape Coloured populations.

C6 epitope expression by an unrelated antisense cDNA clone: an inadvertent surface-simulation mimotope

A cDNA clone which directs the expression of a fusion protein reacting with anti-C6 antibodies has been isolated and sequenced. A synthetic peptide corresponding to the 14 C-terminal residues of the expressed protein elicited the production of antibodies which are specific for native C6, confirming the presence of a C6 epitope on the expressed protein. However, analysis of the intron-exon boundaries of a corresponding genomic clone revealed that the expression clone is in antisense orientation, and is therefore not C6 cDNA. Comparison of the sequences of the expression clone and expressed protein with those for C6 have not demonstrated any significant sequence homology. It is therefore apparent that what has been cloned is a mimotope for C6 which includes in its continuous sequence an epitope that is conformational in C6 and not represented as a continuous sequence in the C6 molecule. Although this was not the purpose of the investigation, these results confirm that screening random expression libraries with antibodies may be an alternative to the synthetic peptide approach to obtain mimotopes reacting with particular antibodies.

A cleavable biotinylating agent and its use in protein electroblots

A direct method for the synthesis of N-biotinyl penicillamine is described. It has been shown to be a convenient biotinylating agent for antibodies which have been previously coupled with SPDP. The biotinylated antibodies can be used to detect antigens on protein electroblots using 125I-labelled streptavidin and radioautography. The biotin and its attached streptavidin and radiolabel can be removed under mild conditions and the blot reprobed with a different antibody using an identical protocol.

Properties of a low molecular weight complement component C6 found in human subjects with subtotal C6 deficiency

A sensitive ELISA assay was used to quantitate serum complement component C6 concentrations. Levels in the range 0.3-3 micrograms/ml were measured in samples from eight individuals (four separate pedigrees) and two subjects with subtotal combined C6/C7 deficiency who have been reported previously. We defined C6 levels in this range as subtotal C6 deficiency (C6SD). In contrast, C6 deficiency with levels below 0.03 micrograms/ml was defined as C6Q0. C6Q0 has been found in 29 unrelated cases which have already been reported. Investigations of the properties of the C6 found in the C6SD subjects showed it to be haemolytically active and able to incorporate into the terminal complement complex. The protein had a relative molecular weight (Mr) of approximately 86% of normal C6 and this Mr was identical to that of the C6 of one combined deficient subject. The Mr of the C6 of the other combined deficient subject was previously estimated as 79% of the Mr of normal C6. Isoelectric focusing (IEF) analysis with band development by haemolytic overlay revealed that all C6SD samples produced an identical weak C6 band pattern anodal to normal C6A bands. The C7 IEF patterns of the two combined deficient subjects were identical, and the C6 IEF patterns of both were identical to those of the C6SD subjects. Thus the C6 of the combined deficient subjects is probably the same abnormal protein found in the C6SD individuals. None of the C6SD or combined deficient subjects have had meningococcal disease and it may be that low C6 levels afford some protection.