Abstract
The complement protein C4, encoded by two genes (C4A and C4B) on chromosome 6p, is the most polymorphic among the MHC III gene products. We investigated the molecular basis of C4 deficiency in a Finnish woman with systemic lupus erythematosus. C4-specific mRNA was present at low concentrations in C4-deficient (C4D) patient fibroblasts, but no pro-C4 protein was detected. This defect in C4 expression was specific in that synthesis of two other complement proteins was normal. Analysis of genomic DNA showed that the proposita had both deleted and nonexpressed C4 genes. Each of her nonexpressed genes, a C4A null gene inherited from the mother, a C4A null gene, and a C4B null gene inherited from the father, all contained an identical 2-bp insertion (TC) after nucleotide 5880 in exon 29, providing the first confirmatory proof of the C4B pseudogene. This mutation has been previously found only in C4A null genes. Although the exon 29/30 junction is spliced accurately, this frameshift mutation generates a premature stop at codon 3 in exon 30. These truncated C4A and C4B gene products were confirmed through RT-PCR and sequence analysis. Among the possible genetic mechanisms that produce identical mutations in both genes, the most likely is a mutation in C4A followed by a gene conversion to generate the mutated C4B allele.
The two most polymorphic genes within the class III region of the human MHC, C4A and C4B, code for the complement protein C4, which functions in the classical pathway of complement activation 1 . Located on chromosome 6p, these genes produce single chain precursors (pro-C4) 2 that give rise to native C4, a three-chain glycoprotein of about 185 kDa Mr. The C4A and C4B genes specify functionally distinct C4 proteins; C4A is 100 times more reactive with targets containing free amino groups and 10 times less reactive with hydroxyl groups than C4B 3, 4 . The functional differences in these C4 proteins are due to four amino acid variations between residues 1101 and 1106 5, 6 .
The genetics of human complement C4 are complex. There is a frequent variation in the number and size of the C4 genes present in an individual (reviewed in 7 . The majority of the population has two C4 loci coding for C4A and C4B proteins, respectively. The C4 genes are either 21 or 14.6 kb in size due to the presence of an endogenous retrovirus HERV-K(C4) in intron 9 of the long C4 gene 8, 9, 10, 11 . Deletion or duplication of the C4 genes is always accompanied by neighboring genes RP at the 5′ region, steroid hydroxylase gene CYP21, and extracellular matrix protein gene tenascin TNX at the 3′ region. Along with C4, these genes make up a genetic module known as the RCCX4, 12, 13, 14 . There can be one, two, or three RCCX (RP-C4-CYP21-TNX) modules on one copy of chromosome 6 in the HLA class III region. Variation in RCCX module numbers between chromosomes may favor unequal cross-over events resulting in deletions and homoduplications of C4 and its neighboring genes. Indeed, such haplotypes have been recognized with a reasonably high frequency 15, 16 , accounting for a major fraction (∼60%) of the known C4 null (Q0) alleles, although other molecular mechanisms that can generate C4A or C4B null alleles have been identified 17 .
Approximately 35% of individuals among all races surveyed 18 have either a C4AQ0 or a C4BQ0 allele; about 8–10% do not express two of the four C4 alleles, and about 1% express only a single C4 allele. C4A and C4B null alleles have been associated with systemic lupus erythematosus (SLE), insulin-dependent diabetes mellitus, IgA nephropathy, Henoch-Schönlein purpura, subacute sclerosing panencephalitis, autoimmune chronic active hepatitis, membranoproliferative glomerulonephritis, rapid progression of HIV infection, and several other disorders 19, 20, 21, 22 . These associations may be due to linkage with other MHC genes, the C4 deficiency (C4D) itself, or both. Total deficiency of C4B (homozygous C4BQ0) is a risk factor for bacterial meningitis 23 . Homozygous deficiency of C4A occurs in the general population at a frequency of about 2%, but 10–15% of whites with SLE are homozygous C4AQ0, and >50% of whites with SLE are heterozygous for C4AQ0. Studies of other racial groups 24 suggest that the C4A null alleles are important variables, independent of or in addition to the closely linked MHC class II genes in the disease expression of SLE.
The apparent importance of C4D in the pathophysiology of SLE and perhaps some of the other associated diseases together with relatively few studies that have explored the molecular mechanisms of complete C4D prompted this study of a C4-deficient Finnish woman with SLE. The molecular basis of her complete C4D was elucidated by investigating the biosynthesis of C4 protein, transcription of C4 mRNA, detection of C4A and C4B genes, and nucleotide sequencing to pinpoint the mutations leading to nonexpression of C4A and C4B proteins from both chromosomes.
Materials and Methods
HLA and complement typing
HLA-A, -B, -C, and -DR Ags were assigned by the standard microlymphocytotoxicity test 25 . HLA class II typings DRB1, DRB3, DRB5, and DQB1 were also performed by the PCR-based chemiluminescent reverse dot blot method 26 . Complement factor B and C4 allotypes were determined electrophoretically as described by Marcus and Alper 27 . Standard methods were used for radial immunodiffusion 28 and total hemolytic complement in gels (Quantiiplate, Kallestad, Beaumont, TX).
Cells
Skin fibroblast cell lines were established from the C4-deficient patient, her HLA-identical brother, and a normal individual according to published methods 29, 30 . Cells were maintained at 37°C in 5% CO2 in DMEM supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin.
EBV-infected B cell lines were established from the father and the mother of the C4D patient and maintained at 37°C in 5% CO2 in RPMI 1640 medium (Life Technologies, Gaithersburg, MD) supplemented with 10% bovine calf serum, 5 mM sodium pyruvate, 2 mM glutamine, 10 U/ml penicillin, 10 μg/ml streptomycin, and 1 mM nonessential amino acids. Anticoagulated peripheral blood samples from all the family members were obtained for DNA extraction.
Biosynthetic labeling
Biosynthetic labeling experiments were performed using C4D and normal fibroblasts grown to confluence in 24-multiwell tissue culture plates (Corning, Corning, NY) in DMEM containing 10% FCS. Before each experiment the cells were washed twice with warm HBSS (Life Technologies) and incubated for 2 h in DMEM containing [35S]methionine at 250 μg/ml (ICN Radiochemical, Irvine, CA; sp. act., 1100 Ci/mmol). DMEM without serum but with 1 mg/ml BSA was used for the labeling period. In some experiments cells were incubated before labeling with mediators (IFN-γ (1000 U/ml; Amgen, Thousand Oaks, CA) and TNF-α (30 ng/ml; Genentech, South San Francisco, CA)) for 24 h to maximize C4 expression. At the end of the pulse period, the extracellular medium was recovered. The cells were washed twice with warm PBS and lysed by one freeze-thaw cycle in PBS containing 0.5% sodium deoxycholate (Fisher Scientific, Fairlawn, NJ), 1% (w/v) Triton X-100 (Sigma), 10 mM EDTA (Sigma), 2 mM PMSF (Sigma), and 100 μg/ml leupeptin (Boehringer Diagnostics, Somerville, NJ). The lysed cells and extracellular media were clarified by centrifugation, and total protein synthesis was measured by TCA (Sigma) precipitation of 5-μl aliquots of the cell lysate or extracellular medium 31 .
Immunoprecipitation and SDS-PAGE
The samples were precleared with Staphylococcus aureus protein A (Immunoprecipitin, Life Technologies) and then incubated in the presence of goat polyclonal Abs (IgG fraction) to C4, to C3, and to C1INH (Incstar, Scarborough, ME) overnight at 4°C with excess Ab 32 . After incubation, excess Staphylococcus protein A was added to capture Ag-Ab complexes. Immunoprecipitates were washed once with PBS containing 0.5% SDS, 1% (w/v) Triton X-100, 0.5% sodium deoxycholate, and 0.5% BSA (ICN Immunologicals, Costa Mesa, CA) and three times with the same buffer without BSA. The immunoprecipitates were subjected to SDS-PAGE under reducing conditions (5% 2-ME) according to the method of Laemmli 33 . The m.w. markers were analyzed in parallel lanes and visualized by Coomassie Brilliant Blue R-250 staining. The gels were fixed in a solution of 46% methanol and 8% acetic acid in water for 30 min, rinsed in water, treated with Fluoro-Hance (RPI, Mount Prospect, IL) for 30 min, dried, and exposed with an intensifier screen at −70°C to Kodak XAR-5 film (Eastman Kodak, Rochester, NY).
RNA isolation and Northern blot analysis
C4D and normal fibroblasts were grown to confluence in 100-mm culture dishes and stimulated with 500 U/ml of IFN-γ and with 30 ng/ml of TNF-α for 24 h before RNA harvest. Cells were lysed with guanidium isothiocyanate; total cellular RNA was purified by CsCl density gradient ultracentrifugation and quantified by absorbance at 260 nm 34 . RNA samples (20 and 40 μg) were denatured, subjected to electrophoresis in 1% agarose/formaldehyde gel, transferred to a nylon membrane (Amersham, Arlington Heights, IL), washed twice for 5 min each time in 2× SSC (1× SSC = 0.015 M sodium citrate and 0.15 M NaCl, pH 7.0), air-dried, and UV cross-linked before prehybridization at 65°C for 4 h in the following buffer: 50 mM PIPES, 100 mM NaCl, 50 mM sodium phosphate, and 1 mM EDTA 35 . The C4, factor B, and actin cDNA probes were radiolabeled with 100 μCi of [α-32P]dCTP (ICN, Irvine, CA) using the Random Prime DNA labeling kit (Promega, Madison, WI) according to the manufacturer’s protocol. After hybridization overnight at 65°C, the filters were washed at room temperature for 30 min, then three times at 65°C for 15 min for C4, and twice for factor B in 5× SSC/1% SDS and autoradiographed.
DNA isolation and Southern blot analysis
Genomic DNA was extracted from fibroblasts or EBV-transformed lymphoblasts by proteinase K digestion followed by phenol-chloroform extraction and ethanol precipitation 36 . Buffy coats of peripheral white blood cells were digested with proteinase K. After digestion, cellular proteins were salted out by dehydration and precipitated with a saturated NaCl solution 37 . Individual DNA samples of 10 or 15 μg were digested with the appropriate restriction enzymes. The DNA fragments were separated by electrophoresis in an 0.8% agarose gel (ultrapure agarose, Life Technologies) and transferred to a nylon membrane (Hybond-N+, Amersham, Arlington Heights, IL). Prehybridization and hybridization were conducted at 65°C in 5× SSC/0.2% SDS. For hybridization, 60 ng of purified DNA insert was labeled with [α-32P]dCTP (ICN, Irvine, CA) by the random priming method (Promega, Madison, WI). Hybridization was performed overnight followed by two subsequent washes of 20 min each in 0.2× SSC/1% SDS at 65°C, and the blots were visualized by autoradiography.
DNA probes
The full-length C4 cDNA clone pAT-A and its 476-bp BamHI/KpnI fragment specific for the 5′ ends of both C4 genes were used for C4 probing 15, 38 . The CYP21-specific probe was a 500-bp SstI/PstI fragment from the 3′ end of the 2.4-kb genomic C4.5 CYP21 clone, provided by Dr. David Chaplin (Washington University School of Medicine, St. Louis, MO). The RP probe was a 1.1-kb fragment corresponding to nucleotides 522-1620 of the RP1 cDNA 13 . The TNX probe is a 500-bp fragment corresponding to exons 35–37 of the human TNXA gene 12 . The factor B cDNA probe was a 1761-bp PstI fragment isolated from the previously characterized pBfA-28 clone 16 . The α-actin probe was an 800-bp PstI-BamHI fragment of a cDNA isolated from chick skeletal muscle 39 .
Oligonucleotide synthesis and DNA sequence analysis
All primers used for RT, DNA amplification, and sequencing were synthesized using an automated DNA synthesizer PCR-Mate (model 391, Applied Biosystems, Foster City, CA) and are shown below. Genomic sequencing was performed using double-stranded templates and a model 373A automated DNA sequencer from Applied Biosystems according to the standard protocol of the taqDyeDeoxy Terminator Cycle Sequencing Kit. Before direct sequencing, all PCR-amplified DNA products were purified on a 1% agarose gel using Whatman DE81 paper (Clifton, NJ) and ethanol precipitation. The region covering genomic DNA from exons 19–29 except for three gaps was sequenced in the father, mother, and proposita as well as in an unrelated control. Genomic DNA from two other siblings was also used for sequencing some parts of this region. All oligonucleotides were identical with the published C4 genomic sequence 40 , except for substitutions that were introduced (oligonucleotides 9 and 13) to generate restriction sites to facilitate cloning. Numbering of nucleotides in the C4 sequence is given in Ref. 40.
Oligonucleotides
The following oligonucleotides were used: 1) 5′-GACACTGTGGCTCCCCGACTCTCTG-3′, 2) 5′-CGAGGGTCCTTCGAATTCCCTGTGG-3′, 3) 5′-GGGGTGCTCCATTCACCTCAATCTG-3′, 4) 5′-GACAAGGCCCCCTCAGAGCCTAAAG-3′, 5) 5′-TGCGGATCCAGCAGTTTCGGAAG-3′, 6) 5′-CCTGCTCCTGGGCCAAACTCAG-3′, 7) 5′-TCAGTGGCGTTTCTGCCCTCTG-3′, 8) 5′-ACCCTCCTCCCGTTTTCTTCCAG-3′, 9) 5′-GTGCTGGGATCCTGGGTGCCCACGCAG-3′, 10) 5′-GAGCCCAGCAGGGGGTGGCTAAG-3′, 11) 5′-CTCAGGATCCTAAGTCCCCTGGGCCT-3′, 12) 5′-CGCAGCCTGGTCTGCCATCTCTG-3′, 13) 5′-GTTGTTAAGCTTCAGCGCGTGGGACTTG-3′, and 14) 5′-CCCACCTTCACATTGATCTTG-3′.
Amplification of cDNA and genomic DNA
Ten micrograms of total RNA isolated from fibroblasts of C4D and normal individuals was incubated with 10 U of reverse transcriptase at 42°C for 1 h using the buffers and dNTPs supplied in a cDNA synthesis kit (Invitrogen, San Diego, CA). Oligonucleotide 14, antisense to the normal C4 cDNA sequence, was used as a primer in the first-strand synthesis. The cDNA was subsequently amplified by PCR using the first strand as template and oligonucleotide pairs of 9 and 13. These primers were constructed with restriction sites near the 5′ ends, but the PCR product was purified as described above and used for direct sequencing. The first-strand cDNA was initially denatured at 93°C for 1 min with 1 μg of each oligonucleotide in a 100-μl solution containing 10 mM Tris (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.1% gelatin, 20 mM dNTPs, and 0.5 U of KlenTaq 1 DNA polymerase 41 . Following initial denaturation, the cDNA was amplified by melting at 93°C for 1 min, annealing at 66°C for 1 min, and polymerization at 72°C for 2 min. Forty cycles of amplification were performed using a programmable Hybaid OmniGene thermal cycler (Labnet, Middlesex, U.K.) followed by a final elongation at 72°C for 5 min. The C4 cDNA was purified and sequenced as outlined above.
Genomic DNA was amplified using 400 ng of DNA and 200 ng of each primer in a primer pair with the other reagents as described above in a “touchdown” protocol 42 . The first six cycles were conducted at decreasing annealing temperatures in 2°C steps for two cycles each from 72 to 64°C, followed by 30 cycles using the following conditions: 10 s at 98°C, 1.5 min at 64°C, and 5 min at 72°C. The amplified genomic DNA was purified and sequenced as outlined above.
Cloning and sequencing of the entire C4d genomic regions for the C4B genes
DNA fragments corresponding to the C4d region were amplified with synthetic PCR primers C4E 22.5 and C4E 31.3 using the High Fidelity PCR system (Boehringer Mannheim, Indianapolis, IN). PCR conditions were one cycle at 94°C for 2 min; 10 cycles at 94°C for 15 s, 65°C for 30 s, and 72°C for 1 min; 20 cycles at 94°C for 15 s, 56°C for 30 s, and 72°C for 1 min plus 20-s time extensions; one cycle at 72°C for 7 min; followed by a 4°C dwell cycle. The PCR product was purified (Qiagen, Chatsworth, CA) and cloned into TA cloning vector (Invitrogen). The C4d insert of the clones was amplified again via PCR, and the purified PCR product was restriction digested with NlaIV to determine the presence of either C4A or C4B. A 467-bp fragment represents C4B, while two fragments (276 and 191 bp) represent C4A 15 . Clones representing C4B from both the proposita’s mother and father were sequenced (ABI Prism). Comparison of DNA sequences with national databases was performed by the GCG FASTA program from the Pittsburgh Supercomputing Center. DNA alignments were performed with the SeqManII program (DNASTAR). Synthetic DNA primers for PCR and for cycle sequencing were synthesized by Life Technologies (Grand Island, NY). Details of the primer sequences are as follows. For amplification of the C4d region of the C4 gene the primer sequences were: C4E 22.5, GAAGGGGCCATCCATAGAGA; and C4E 31.3, CTTCAGGGTTCCTTTGCTGT. For sequencing of the C4d region of the C4B gene the primer sequences were: C4E 25.3, CAGGTGCTGCTGTCCCGTGA; C4E 26.5, GCTCACAGCCTTTGTGTTGAA; C4E 27.3, CACTCTCTGCTTCAATGGCT; C4E 28.5, GAAGCCTCCATCTCAAAGGC; and C4E 29.3, TTGGGTACTGCGGAATCCCC.
Results
Family studies and clinical case report
The patient was a 30-yr-old Finnish woman who developed photosensitivity, a malar rash, polyarthritis, leukopenia, and positive anti-nuclear Ab (1/320), anti-Sm Ab (1/1280), and a weakly positive rheumatoid factor coincident with her first pregnancy and shortly after delivery. In her serum, complement protein C4 was undetectable, and C3 was modestly reduced or at the lower limit of normal (74–111 mg/100 ml). A skin biopsy revealed vasculitis with deposition of IgM and C3. Five years later she developed aphthous ulcers and increased joint symptoms, for which she was treated with prednisone. An exacerbation of mucositis 1 yr later led to an increase in prednisone therapy and subsequent treatment with azathioprine and methotrexate. An ectopic pregnancy and a pregnancy with her second child each precipitated exacerbations of her symptoms. She has had no renal or cardiac disease, no increased susceptibility to infection, hypersensitivity to medications, or central nervous system disease. Family history revealed a younger male sibling (HLA identical) who suffers from photosensitivity. Her children, her parents, and her MHC nonidentical brother are alive and well. Laboratory studies showed undetectable total hemolytic complement (CH100) and no C4 protein by radial immunodiffusion in the patient or her HLA identical brother. Quantitation of C4 and CH100 in the immediate family members and their HLA typing is shown in Fig. 1⇓. In brief, the proposita inherited the HLA A2 B39 Cw7 DRB1*1501 DRB5*0101 DQB1*0602, BF-S, C4AQ0 BQ0 from the paternal chromosome, and HLA A2 B40 Cw3 DRB1*1501, DRB5*0101 DQB1*0602, BF-S, C4AQ0 BQ0 from the maternal chromosome. The C4 haplotypes for the patient’s father are C4AQO BQO/C4AQ0 B2. The C4 haplotypes for the patient’s mother are C4AQ0 BQ0/C4A3 B3.
The nuclear family. The proposita with SLE and her MHC-identical C4D brother have severe and minimal symptoms, respectively. The serum C4 concentration is in grams/liter/total hemolytic complement activity (in units/ml/normal values, 0.2–0.69/>50). Numbers below haplotypes represent C4 concentration/total hemolytic activity, which are in grams per liter and units per milliliter, respectively. The normal values for C4 concentration are between 0.2 and 0.69; the normal values for hemolytic activity are >50. An arrow indicates the proposita. MHC haplotypes are: a) A2, B39, Cw7, DRB1*1501, DRB5*0101, DQB1*0602, BF S, C4AQ0 BQ0; b) A24, B40, Cw3, DRB1*1302, DRB3*0301, DQB1*0605, BF S, C4AQ0 B2; c) A2, B40, Cw3, DRB1*1501, DRB5*0101, DQB1*0602, BF S, C4AQ0 BQ0; and d) A3, B62, Cw3, DRB1*0401, DRB4*01, DQB1*0302, BF S, C4A3 B3.
Biosynthesis and secretion of C4 in cell culture
To determine whether a defect in regulation of C4 gene expression, C4 synthesis, or secretion accounted for the low to absent C4 protein in sera of the homozygous deficient patient, primary fibroblast cell cultures were established. SDS-PAGE of immunoprecipitates of cell lysates and culture media from [35S]methionine pulse-labeled normal and proposita’s C4 deficient (C4D) fibroblast cultures is shown in Fig. 2⇓. Fibroblasts were incubated with IFN-γ (500 U/ml) and TNF-α (30 ng/ml) to yield maximum C4 expression 24 . A polypeptide of approximately 185 kDa, corresponding in size to pro-C4, was detected in lysates, and native C4 subunits (α, ∼93 kDa; β, ∼78 kDa; γ, ∼33 kDa) plus the approximately 125-kDa processing intermediate α-γ fragment were recovered from the culture media in normal fibroblast cultures. Neither pro-C4 nor native C4 protein was detected in IFN-γ- and TNF-α-stimulated C4D fibroblast cultures even after prolonged exposure (10 days) of the autoradiograph. In contrast, both C4D and normal fibroblasts synthesized and secreted comparable amounts of C1 inhibitor, a highly IFN-γ-responsive complement protein, and C3, a protein similar to C4 in size and postsynthetic processing (Fig. 3⇓, A and B). C3 synthesis and secretion were more variable among separate replicate experiments, but the mean C3 expression in C4D cells was not significantly different from normal.
Biosynthesis of C4 protein in fibroblasts from a normal individual and from a C4D patient. Human primary skin fibroblasts were pulse labeled with [35S]methionine, stimulated with IFN-γ and TNF-α (even-numbered lanes) and with medium alone (odd-numbered lanes), and subjected to immunoprecipitation, SDS-PAGE, and autoradiography. Cell lysates (left panel) and culture media (right panel) from normal cells (lanes 1,2, 5, and 6) and from C4D (lanes 3, 4, 7, and8) are shown. Exposure time was 24 h (intracellular; left) and 72 h (extracellular; right).
Biosynthesis of C1 inhibitor (A) and C3 (B) in skin fibroblasts. The samples were processed and applied similar to that described in Fig. 2⇑. Media alone (odd-numbered lanes), cytokine-stimulated (even-numbered lanes), normal fibroblasts (lanes 1, 2, 5, and 6), and C4D (lanes 3, 4, 7, and 8) are shown. Exposure times were 16 h (A) and 8 h (B). These data are examples of one of our complete sets.
C4 mRNA expression
The size and amount of C4 mRNA in normal and C4D IFN-γ- and TNF-α-stimulated fibroblasts were estimated by Northern blot analysis. C4-specific mRNA (5.5 kb) in cytokine-stimulated cells was of similar size in normal and C4D fibroblasts, but the amount of C4 mRNA was markedly reduced in the C4D fibroblasts (Fig. 4⇓). Scanning of replicate Northern blots for C4, factor B, and actin confirmed a selective decrease in C4 mRNA (<3% of normal) in the deficient cells, with only a minor difference (1.5:1) in Bf mRNA compared with normal cells.
RNA blot analysis of C4 transcripts from normal (C4S) and C4D fibroblasts. RNAs were isolated from C4D and normal fibroblasts stimulated with IFN-γ and TNF-α. Twenty micrograms (lane 1) and 40 μg (lane 2) of RNA in replicate blots were probed for the presence of factor B, C4, and actin transcripts. The two upper panels were exposed for about 18 h, and the ∗ panel was exposed for 64 h to visualize C4 mRNA in the C4D cells (data are from one of three experiments showing similar results). Densitometric scanning revealed a ratio of C4 mRNA >33:1 and of Bf mRNA of 1.5:1.0 in normal and C4D samples, respectively, when normalized for loading by scanning for mRNA actin.
Three nonexpressing C4 genes in the patient
The presence of C4 and its neighboring genes for the RCCX modules in the HLA were investigated to determine the molecular basis of the C4D. Genomic DNAs from the patient and her parents were digested with restriction enzyme TaqI and hybridized to RP, C4, CYP21, and TNX probes. As shown in Fig. 5⇓, the father (lane 3) has the 7.0- and 5.4-kb restriction fragments corresponding to RP1-C4(L) and RP2-C4(S), respectively; the 3.7- and 3.2-kb fragments corresponding to CYP21B and CYP21A, respectively; and the 2.5- and 2.4-kb fragments corresponding to TNXB and TNXA genes, respectively. The band intensities for these fragments are similar; therefore, he is homozygous for RP1-C4(L)-CYP21A-TNXA-RP2-C4(S)-CYP21B-TNXB. The C4 allotyping results suggested that he has C4AQ0 C4BQ0 and C4AQ0 C4B2. Therefore, the father has RCCX bimodular (L-S) haplotypes in both chromosomes. Both C4A and C4B genes are present in the C4AQ0 BQ0 chromosome, but these genes are not producing C4 protein. Similarly, the C4AQ0 C4B2 chromosome with the RCCX bimodular (L-S) haplotype is likely to have a C4A pseudogene.
Southern blot analysis of the RCCX modular structures in the C4D patient and her parents. Genomic DNAs were digested with TaqI, processed according to Southern’s procedure, and hybridized with 32P-labeled RP1, CYP21, and TNX probes. Lane 1, Control DNA from the T cell line MOLT4; lane 2, mother; lane 3, father; lane 4, patient.
The mother (lane 2) has 7.0- and 6.0-kb TaqI fragments corresponding to RP1-C4(L) and RP2-C4(L), respectively. The 7.0-kb fragment is twice as intense as the 6.0-kb fragment. In addition, her 3.7-kb CYP21B fragment is twice as intense as the 3.2-kb CYP21A fragment, and the 2.5-kb TNXB fragment is more intense than the 2.4-kb TNXA fragment. These results suggest that she is heterozygous for one RCCX bimodular chromosome, RP1-C4(L)-CYP21A-TNXA-RP2-C4(L)-CYP21B-TNXB, and one RCCX monomodular chromosome, RP1-C4(L)-CYP21B-TNXB. Since she has C4A3 B3 on one chromosome and C4AQ0 BQ0 on the other, the monomodular (L) chromosome corresponds to the C4AQ0 C4BQ0 phenotype. One of the C4 genes on this chromosome has been deleted.
For the patient, the 7.0-kb TaqI fragment for RP1-C4(L) is more intense than the 5.4-kb fragment for RP2-C4(S) (lane 4). In addition, the 3.7-kb fragment for CYP21B is more intense than the 3.2-kb CYP21A fragment. She is heterozygous for a RCCX bimodular chromosome RP1-C4(L)-CYP21A-TNXA-RP2-C4(S)-CYP21B-TNXB and one monomodular RCCX chromosome RP1-C4(L)-CYP21B-TNXB. She has two long C4 genes and one short C4 gene. None of these three C4 genes expresses detectable C4 protein. Lane 1 shows the result of a control DNA (T cell line MOLT4) with bimodular RCCX (L-S) haplotypes.
C4 typing by restriction mapping and sequence analysis
To define and type the C4 genes inherited in this family, a 934-bp genomic fragment spanning from exon 25 to the 5′ end of intron 28 that permits differentiating among specific C4 allotypes 8 was amplified from genomic DNA using primers 5 and 11, subjected to digestion with NlaIV and EcoO109, and sequenced. The digests and the patterns characteristic for specific C4 isotypes at each of the C4 genes and sequence analysis are shown in Table I⇓.
Derived amino acid sequences at the C4d regiona
To ascertain the molecular mechanisms accounting for the nonexpressed C4A and C4B genes in the proposita, a 267-bp PCR product spanning the segment from nucleotide 5776 (intron 28) to nucleotide 6043 (exon 29) was generated, and the total PCR product was directly sequenced to search for a mutation previously found in the C4A gene 43 . The results (Fig. 6⇓) show a 2-bp (TC) insertion after nucleotide 5880. This insertion leads to a frameshift mutation, a predicted change in amino acid sequence, and a stop signal at codon 3 in exon 30. The HLA-identical brother’s amplified DNA showed the same sequence. Amplified DNA from her haploidentical brother, her father, and her mother showed both normal and mutant sequences. Based on the amplitude of the peaks, the father has the TC insertion in both C4A and C4B genes, whereas the mother has the 2-bp insertion only in her C4A gene. An unrelated control showed only the normal sequence. The proposita showed only the mutated sequence.
Nucleotide sequence and derived amino acid sequence of the 267-bp PCR products from nucleotide 5776 (intron 28) to 6043 (exon 30) in patient and normal control DNA. The first nucleotide indicated is nucleotide 5867, and first amino acid is codon 1209. Shaded nucleotides indicate the 2-bp insertion; the dotted line indicates nucleotides not shown.
To confirm the 2-bp insertion in the C4BQ0 gene, genomic sequences of the C4d region from the father and mother were amplified by PCR, cloned into TA vector, and sequenced. The C4B clones were identified by NlaIV digests 15 . Two C4B clones from the father and one from the mother were sequenced to completion. The paternal C4B clones contain the C4B isotypic amino acid sequence LSPVIH at positions 1101–1106, the C4B-associated amino acid sequence ADLR for the Ch1 Ag at positions 1188–1191, and S1157 that is associated with Ch6. Additionally, he has the D1054 that is associated with Ch5 6 . In intron 28, both paternal clones have the nucleotide sequence ggctct that is 14 nucleotides downstream of intron donor site. However, the most distinguishing feature for both paternal C4B clones is the presence of a TC dinucleotide insertion at codon 1213. This is the first definitive proof showing the presence of a 2-bp insertion into a C4B gene. Conversely, the C4B sequence for the maternal clone is identical with the C4B3 sequence published previously, which has the characteristic sequences G1054, LSPVIH 1101–1106, S1157, and ADLR 1188–1191 5 . As expected, the exon 29 sequence is intact, with no insertion.
This 2-bp insertion in exon 29 of C4A was previously reported 43 , and the stop codon in the next exon was predicted on the assumption that normal splicing (excision at the normal splice sites) of intron 29 would occur. To directly test this presumption, RNA from the patient’s and normal fibroblasts (stimulated with IFN-γ and TNF-α) was amplified by RT-PCR (primer 14 to generate cDNA and primers 9 and 13 for PCR amplification) and then subjected to sequence analysis. These results revealed an exonic sequence identical with that obtained from genomic sequencing and normal splicing between exons 28–30 (data not shown).
Discussion
The C4 genes are expressed in liver 32 and in extrahepatic tissues, including skin fibroblasts 31 . C4 mRNA, 5.5 kb in length, programs synthesis of a single-chain preprotein that undergoes extensive postsynthetic processing to yield the native three-chain C4 protein. The availability of primary skin fibroblast culture from the C4D patient reported here permitted for the first time an analysis of C4 protein synthesis and secretion in an individual with a genetic deficiency of this protein. Even under tissue culture conditions designed to maximize C4 expression 24 , no pro-C4 synthesis was observed in fibroblasts from the C4D patient, whereas synthesis of pro-C4 and secretion of native C4 were readily detectable in control fibroblasts. This defect is selective, since the C4D fibroblasts synthesized and secreted C1 inhibitor and C3 in amounts comparable to those produced by control cells and in published reports 44, 45 . The failure to detect pro-C4 could be due to a pretranslational mechanism, a selective defect in pro-C4 translation, or instability of the C4 translation product(s). In part, the decrease in C4 protein expression results from markedly reduced steady state C4 mRNA levels (<3% of normal).
Analysis of the nuclear family established that the patient inherited a nonexpressed long C4A gene from her mother, a nonexpressed long C4A gene, and a nonexpressed short C4B gene from her father. Her maternally derived haplotype, A2; B40; Cw3; DR2 (DRB1*1501, DRB5*0101, DQB1*0602), also contains a deleted C4B gene. C4B deletions have been recognized in association with B40 46, 47 , while the linkage between B40 and DR2 is unusual. Her paternally derived haplotype, A2;B39;Cw7;DR2 (DRB1*1501, DRB5*0101, DQB1*0602), is identical at the DR/DQ loci and differs at the class I loci. A detailed analysis of the basis for lack of expression of the patient’s nondeleted C4 genes was undertaken. This revealed an identical 2-bp insertion in her three nonexpressed C4 genes, which include two C4A genes and one C4B gene. This 2-bp insertion in exon 29 was previously observed by Schneider and colleagues 43 in several C4A null alleles; the majority of them were associated with HLA-B60 and DR6. In their study, based entirely on sequence analysis of genomic DNA (PCR amplified), they predicted that this frameshift mutation would generate a stop codon in the next exon of the C4 transcript derived from this gene. Generation of the stop codon would occur only if splicing of intron 29 was executed at the normal splice sites. Examples of splicing abnormalities, particularly in the context of mutations (including frame shifts) have been reported 48, 49, 50, 51, 52 . Because C4 is expressed in a readily available tissue (skin fibroblasts), we were able to establish that splicing at the exon 29–30 junction was normal and that a truncated C4 translation product would be generated. This truncated C4 polypeptide was not detected in cell culture, probably because the shortened pro-C4 may be unstable.
Nonsense mutations have been associated with marked decreases in steady state levels of specific mRNA in other instances, such as β-globin 53 , α1-antitrypsin 54 , surfactant protein B 55 , cystic fibrosis transmembrane regulator 56 , and others 56, 57 . Although the basis for this observation is not completely understood 52 , a similar mechanism could account for the decreased levels of C4 mRNA observed in this patient (see Fig. 4⇑). Nonetheless, sufficient C4 mRNA was present in cytokine-stimulated fibroblasts to allow nucleotide sequence analysis of RT-PCR products from the homozygous deficient fibroblasts. In contrast to the earlier study that identified this mutation only in C4AQ0, the 2-bp (TC) insertion was surprisingly also present in the patient’s C4BQ0 nonexpressed genes, giving the first evidence of the molecular basis for C4B pseudogenes. This 2-bp insertion in exon 29 of the nonexpressed C4BQ0 gene could have been acquired from the C4AQ0 gene of the HLA B40-positive haplotype by unequal cross-over or gene conversion-like events. It has been noticed that some of the C4BQ0 phenotypes in the population are attributed to the deletion of the C4B gene (as shown here for the HLA B40 DR2 haplotype) or to the expression of C4A proteins from the two tandem C4 loci as documented in the HLA B44 DR6 haplotype 15 . This report shows that the third cause of a C4BQ0 phenotype is due to a 2-bp insertion at codon 1213 of the C4B gene that causes a frameshift mutation.
Acknowledgments
We thank Barbara Hermann for secretarial assistance, Joie Haviland for expert technical assistance, Dr. Rick Wetsel for helpful review of the manuscript, Dr. Juha Kere for generating the EBV cell lines, and Dr. Seppo Meri for quantitating the hemolytic activities.
Footnotes
↵1 The genomic DNA sequence for the C4d region of the C4BQO gene is available from the GenBank database under accession number AF092085.
↵2 This work was supported by grants from the National Institutes of Health (AI24739 and HD17461 to H.R.C.; AR43969 to C.Y.Y.) and the Pittsburgh Supercomputing Center through the National Institutes of Health Center for Research Resources Cooperative Agreement (1P41RR06009 to C.Y.Y.).
↵3 Address correspondence and reprint requests to Dr. Harvey R. Colten, Northwestern University Medical School, 303 East Chicago Ave., Morton Building 4-656, Chicago, IL 60611. E-mail address: colten{at}nwu.edu
↵4 Abbreviations used in this paper: RCCX, RP, C4, CYP21, and TNX; SLE, systemic lupus erythematosus; C4D, C4 deficiency.
- Received October 2, 1998.
- Accepted December 10, 1998.
- Copyright © 1999 by The American Association of Immunologists
References
In this issue
