Deficiency of Human Complement C4 Due to Identical Frameshift Mutations in the C4A and C4B

This information is current as Marja-Liisa Lokki, Antonella Circolo, Pirkko Ahokas, Kristi of September 27, 2021. L. Rupert, C. Yung Yu and Harvey R. Colten J Immunol 1999; 162:3687-3693; ; http://www.jimmunol.org/content/162/6/3687 Downloaded from

References This article cites 45 articles, 10 of which you can access for free at: http://www.jimmunol.org/content/162/6/3687.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 27, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Deficiency of Human Complement Protein C4 Due to Identical Frameshift Mutations in the C4A and C4B Genes1,2

Marja-Liisa Lokki,*† Antonella Circolo,*‡ Pirkko Ahokas,§ Kristi L. Rupert,¶ C. Yung Yu,¶ and Harvey R. Colten3*

The complement protein C4, encoded by two genes (C4A and C4B) on 6p, is the most polymorphic among the MHC III products. We investigated the molecular basis of C4 deficiency in a Finnish woman with systemic 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 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 Downloaded from 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 Journal of Immunology, 1999, 162: 3687–3693. http://www.jimmunol.org/ he two most polymorphic genes within the class III region is always accompanied by neighboring genes RP at the 5Ј region, of the human MHC, C4A and C4B, code for the comple- steroid hydroxylase gene CYP21, and extracellular matrix protein ment protein C4, which functions in the classical pathway gene tenascin TNX at the 3Ј region. Along with C4, these genes T 4 of complement activation (1). Located on chromosome 6p, these make up a genetic module known as the RCCX (12–14). There genes produce single chain precursors (pro-C4) (2) that give rise to can be one, two, or three RCCX (RP-C4-CYP21-TNX) modules native C4, a three-chain glycoprotein of about 185 kDa Mr. The on one copy of in the HLA class III region. Vari- C4A and C4B genes specify functionally distinct C4 proteins; C4A ation in RCCX module numbers between may favor is 100 times more reactive with targets containing free amino unequal cross-over events resulting in deletions and homoduplica- by guest on September 27, 2021 groups and 10 times less reactive with hydroxyl groups than C4B tions of C4 and its neighboring genes. Indeed, such (3, 4). The functional differences in these C4 proteins are due to have been recognized with a reasonably high frequency (15, 16), four amino acid variations between residues 1101 and 1106 (5, 6). accounting for a major fraction (ϳ60%) of the known C4 null (Q0) The genetics of human complement C4 are complex. There is a alleles, although other molecular mechanisms that can generate frequent variation in the number and size of the C4 genes present C4A or C4B null alleles have been identified (17). in an individual (reviewed in Ref. 7). The majority of the popu- Approximately 35% of individuals among all races surveyed lation has two C4 loci coding for C4A and C4B proteins, respec- (18) have either a C4AQ0 or a C4BQ0 allele; about 8–10% do not tively. The C4 genes are either 21 or 14.6 kb in size due to the express two of the four C4 alleles, and about 1% express only a presence of an endogenous retrovirus HERV-K(C4) in intron 9 of single C4 allele. C4A and C4B null alleles have been associated the long C4 gene (8–11). Deletion or duplication of the C4 genes with systemic lupus erythematosus (SLE), insulin-dependent diabetes mellitus, IgA nephropathy, Henoch-Scho¨nlein purpura,

*Edward Mallinckrodt Department of Pediatrics, Washington University School of subacute sclerosing panencephalitis, autoimmune chronic active Medicine, St. Louis, MO 63110; †Blood Transfusion Service, Finnish Red Cross, hepatitis, membranoproliferative glomerulonephritis, rapid pro- Helsinki, Finland; ‡Division of Clinical Immunology and Rheumatology, University gression of HIV infection, and several other disorders (19–22). of Alabama, Birmingham, AL 35294; §Raahe District Hospital, Raahe, Finland; and ¶Children’s Hospital Research Foundation, Department of Pediatrics, Ohio State Bio- These associations may be due to linkage with other MHC genes, chemistry Program, and Department of Medical Microbiology and Immunology, Ohio the C4 deficiency (C4D) itself, or both. Total deficiency of C4B State University, Columbus, OH 43205 (homozygous C4BQ0) is a risk factor for bacterial meningitis (23). Received for publication October 2, 1998. Accepted for publication December Homozygous deficiency of C4A occurs in the general population 10, 1998. at a frequency of about 2%, but 10–15% of whites with SLE are The costs of publication of this article were defrayed in part by the payment of page Ͼ charges. This article must therefore be hereby marked advertisement in accordance homozygous C4AQ0, and 50% of whites with SLE are heterozy- with 18 U.S.C. Section 1734 solely to indicate this fact. gous for C4AQ0. Studies of other racial groups (24) suggest that 1 The genomic DNA sequence for the C4d region of the C4BQO gene is available the C4A null alleles are important variables, independent of or in from the GenBank database under accession number AF092085. addition to the closely linked MHC class II genes in the disease 2 This work was supported by grants from the National Institutes of Health (AI24739 expression of SLE. 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 Co- operative 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, 4 Abbreviations used in this paper: RCCX, RP, C4, CYP21, and TNX; SLE, systemic IL 60611. E-mail address: [email protected] lupus erythematosus; C4D, C4 deficiency.

Copyright © 1999 by The American Association of Immunologists 0022-1767/99/$02.00 3688 COMPLEMENT C4 DEFICIENCY

The apparent importance of C4D in the pathophysiology of SLE RNA isolation and Northern blot analysis and perhaps some of the other associated diseases together with C4D and normal fibroblasts were grown to confluence in 100-mm cul- relatively few studies that have explored the molecular mecha- ture dishes and stimulated with 500 U/ml of IFN-␥ and with 30 ng/ml nisms of complete C4D prompted this study of a C4-deficient of TNF-␣ for 24 h before RNA harvest. Cells were lysed with gua- Finnish woman with SLE. The molecular basis of her complete nidium isothiocyanate; total cellular RNA was purified by CsCl density C4D was elucidated by investigating the biosynthesis of C4 pro- gradient ultracentrifugation and quantified by absorbance at 260 nm (34). RNA samples (20 and 40 ␮g) were denatured, subjected to elec- tein, transcription of C4 mRNA, detection of C4A and C4B genes, trophoresis in 1% agarose/formaldehyde gel, transferred to a nylon and nucleotide sequencing to pinpoint the mutations leading to membrane (Amersham, Arlington Heights, IL), washed twice for 5 min nonexpression of C4A and C4B proteins from both chromosomes. each time in 2ϫ SSC (1ϫ SSC ϭ 0.015 M sodium citrate and 0.15 M NaCl, pH 7.0), air-dried, and UV cross-linked before prehybridization at 65°C for4hinthefollowing buffer: 50 mM PIPES, 100 mM NaCl, Materials and Methods 50 mM sodium phosphate, and 1 mM EDTA (35). The C4, factor B, and ␮ ␣ 32 HLA and complement typing actin cDNA probes were radiolabeled with 100 Ci of [ - P]dCTP (ICN, Irvine, CA) using the Random Prime DNA labeling kit (Promega, HLA-A, -B, -C, and -DR Ags were assigned by the standard microlym- Madison, WI) according to the manufacturer’s protocol. After hybrid- phocytotoxicity test (25). HLA class II typings DRB1, DRB3, DRB5, and ization overnight at 65°C, the filters were washed at room temperature DQB1 were also performed by the PCR-based chemiluminescent reverse for 30 min, then three times at 65°C for 15 min for C4, and twice for dot blot method (26). and C4 allotypes were deter- factor B in 5ϫ SSC/1% SDS and autoradiographed. mined electrophoretically as described by Marcus and Alper (27). Standard methods were used for radial immunodiffusion (28) and total hemolytic DNA isolation and Southern blot analysis complement in gels (Quantiiplate, Kallestad, Beaumont, TX). Genomic DNA was extracted from fibroblasts or EBV-transformed Downloaded from lymphoblasts by proteinase K digestion followed by phenol-chloroform Cells extraction and ethanol precipitation (36). Buffy coats of peripheral white blood cells were digested with proteinase K. After digestion, Skin fibroblast cell lines were established from the C4-deficient patient, her cellular proteins were salted out by dehydration and precipitated with a HLA-identical brother, and a normal individual according to published saturated NaCl solution (37). Individual DNA samples of 10 or 15 ␮g methods (29, 30). Cells were maintained at 37°C in 5% CO2 in DMEM were digested with the appropriate restriction enzymes. The DNA frag- supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, and ments were separated by electrophoresis in an 0.8% agarose gel (ultra- http://www.jimmunol.org/ 100 mg/ml streptomycin. pure agarose, Life Technologies) and transferred to a nylon membrane EBV-infected B cell lines were established from the father and the (Hybond-Nϩ, Amersham, Arlington Heights, IL). Prehybridization and mother of the C4D patient and maintained at 37°C in 5% CO2 in RPMI hybridization were conducted at 65°C in 5ϫ SSC/0.2% SDS. For hy- 1640 medium (Life Technologies, Gaithersburg, MD) supplemented with bridization, 60 ng of purified DNA insert was labeled with [␣-32P]dCTP 10% bovine calf serum, 5 mM sodium pyruvate, 2 mM glutamine, 10 U/ml (ICN, Irvine, CA) by the random priming method (Promega, Madison, penicillin, 10 ␮g/ml streptomycin, and 1 mM nonessential amino acids. WI). Hybridization was performed overnight followed by two subse- Anticoagulated peripheral blood samples from all the family members quent washes of 20 min each in 0.2ϫ SSC/1% SDS at 65°C, and the were obtained for DNA extraction. blots were visualized by autoradiography. DNA probes Biosynthetic labeling The full-length C4 cDNA clone pAT-A and its 476-bp BamHI/KpnI by guest on September 27, 2021 Biosynthetic labeling experiments were performed using C4D and normal fi- fragment specific for the 5Ј ends of both C4 genes were used for C4 broblasts grown to confluence in 24-multiwell tissue culture plates (Corning, probing (15, 38). The CYP21-specific probe was a 500-bp SstI/PstI Corning, NY) in DMEM containing 10% FCS. Before each experiment the fragment from the 3Ј end of the 2.4-kb genomic C4.5 CYP21 clone, cells were washed twice with warm HBSS (Life Technologies) and incubated 35 ␮ provided by Dr. David Chaplin (Washington University School of Med- for2hinDMEM containing [ S]methionine at 250 g/ml (ICN Radiochem- icine, St. Louis, MO). The RP probe was a 1.1-kb fragment correspond- ical, Irvine, CA; sp. act., 1100 Ci/mmol). DMEM without serum but with 1 ing to nucleotides 522-1620 of the RP1 cDNA (13). The TNX probe is mg/ml BSA was used for the labeling period. In some experiments cells were a 500-bp fragment corresponding to exons 35–37 of the human TNXA incubated before labeling with mediators (IFN-␥ (1000 U/ml; Amgen, Thou- ␣ gene (12). The factor B cDNA probe was a 1761-bp PstI fragment sand Oaks, CA) and TNF- (30 ng/ml; Genentech, South San Francisco, CA)) isolated from the previously characterized pBfA-28 clone (16). The for 24 h to maximize C4 expression. At the end of the pulse period, the ex- ␣-actin probe was an 800-bp PstI-BamHI fragment of a cDNA isolated tracellular medium was recovered. The cells were washed twice with warm from chick skeletal muscle (39). 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 (Sig- Oligonucleotide synthesis and DNA sequence analysis ma), 10 mM EDTA (Sigma), 2 mM PMSF (Sigma), and 100 ␮g/ml leupeptin (Boehringer Diagnostics, Somerville, NJ). The lysed cells and extracellular All primers used for RT, DNA amplification, and sequencing were syn- media were clarified by centrifugation, and total protein synthesis was mea- thesized using an automated DNA synthesizer PCR-Mate (model 391, Ap- sured by TCA (Sigma) precipitation of 5-␮l aliquots of the cell lysate or ex- plied Biosystems, Foster City, CA) and are shown below. Genomic se- tracellular medium (31). quencing 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 Immunoprecipitation and SDS-PAGE Kit. Before direct sequencing, all PCR-amplified DNA products were pu- The samples were precleared with Staphylococcus aureus protein A (Im- rified on a 1% agarose gel using Whatman DE81 paper (Clifton, NJ) and munoprecipitin, Life Technologies) and then incubated in the presence of ethanol precipitation. The region covering genomic DNA from exons goat polyclonal Abs (IgG fraction) to C4, to C3, and to C1INH (Incstar, 19–29 except for three gaps was sequenced in the father, mother, and Scarborough, ME) overnight at 4°C with excess Ab (32). After incubation, proposita as well as in an unrelated control. Genomic DNA from two other excess Staphylococcus protein A was added to capture Ag-Ab complexes. siblings was also used for sequencing some parts of this region. All oli- Immunoprecipitates were washed once with PBS containing 0.5% SDS, gonucleotides were identical with the published C4 genomic sequence (40), 1% (w/v) Triton X-100, 0.5% sodium deoxycholate, and 0.5% BSA (ICN except for substitutions that were introduced (oligonucleotides 9 and 13) to Immunologicals, Costa Mesa, CA) and three times with the same buffer generate restriction sites to facilitate cloning. Numbering of nucleotides in without BSA. The immunoprecipitates were subjected to SDS-PAGE un- the C4 sequence is given in Ref. 40. der reducing conditions (5% 2-ME) according to the method of Laemmli Oligonucleotides (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 The following oligonucleotides were used: 1) 5Ј-GACACTGTGGCTC of 46% methanol and 8% acetic acid in water for 30 min, rinsed in water, CCCGACTCTCTG-3Ј,2)5Ј-CGAGGGTCCTTCGAATTCCCTGTGG-3Ј, treated with Fluoro-Hance (RPI, Mount Prospect, IL) for 30 min, dried, and 3) 5Ј-GGGGTGCTCCATTCACCTCAATCTG-3Ј,4)5Ј-GACAAGGC exposed with an intensifier screen at Ϫ70°C to Kodak XAR-5 film (East- CCCCTCAGAGCCTAAAG-3Ј,5)5Ј-TGCGGATCCAGCAGTTTCGGA man Kodak, Rochester, NY). AG-3Ј,6)5Ј-CCTGCTCCTGGGCCAAACTCAG-3Ј,7)5Ј-TCAGTGGC The Journal of Immunology 3689

GTTTCTGCCCTCTG-3Ј,8)5Ј-ACCCTCCTCCCGTTTTCTTCCAG-3Ј,9) 5Ј-GTGCTGGGATCCTGGGTGCCCACGCAG-3Ј, 10) 5Ј-GAGCCCAG CAGGGGGTGGCTAAG-3Ј, 11) 5Ј-CTCAGGATCCTAAGTCCCCTGGGC CT-3Ј, 12) 5Ј-CGCAGCCTGGTCTGCCATCTCTG-3Ј, 13) 5Ј-GTTGTTAA GCTTCAGCGCGTGGGACTTG-3Ј, and 14) 5Ј-CCCACCTTCACATTGA TCTTG-3Ј. 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 oli- gonucleotide pairs of 9 and 13. These primers were constructed with re- striction sites near the 5Ј ends, but the PCR product was purified as de- FIGURE 1. The nuclear family. The proposita with SLE and her MHC- scribed above and used for direct sequencing. The first-strand cDNA was identical C4D brother have severe and minimal symptoms, respectively. initially denatured at 93°C for 1 min with 1 ␮g of each oligonucleotide in The serum C4 concentration is in grams/liter/total hemolytic complement a 100-␮l solution containing 10 mM Tris (pH 8.3), 50 mM KCl, 1.5 mM activity (in units/ml/normal values, 0.2–0.69/Ͼ50). Numbers below hap-

MgCl2, 0.1% gelatin, 20 mM dNTPs, and 0.5 U of KlenTaq 1 DNA poly- lotypes represent C4 concentration/total hemolytic activity, which are in merase (41). Following initial denaturation, the cDNA was amplified by grams per liter and units per milliliter, respectively. The normal values for melting at 93°C for 1 min, annealing at 66°C for 1 min, and polymerization C4 concentration are between 0.2 and 0.69; the normal values for hemo- Downloaded from at 72°C for 2 min. Forty cycles of amplification were performed using a lytic activity are Ͼ50. An arrow indicates the proposita. MHC haplotypes programmable Hybaid OmniGene thermal cycler (Labnet, Middlesex, are: a) A2, B39, Cw7, DRB1*1501, DRB5*0101, DQB1*0602, BF S, U.K.) followed by a final elongation at 72°C for 5 min. The C4 cDNA was purified and sequenced as outlined above. C4AQ0 BQ0; b) A24, B40, Cw3, DRB1*1302, DRB3*0301, DQB1*0605, Genomic DNA was amplified using 400 ng of DNA and 200 ng of each BF S, C4AQ0 B2; c) A2, B40, Cw3, DRB1*1501, DRB5*0101, primer in a primer pair with the other reagents as described above in a DQB1*0602, BF S, C4AQ0 BQ0; and d) A3, B62, Cw3, DRB1*0401, “touchdown” protocol (42). The first six cycles were conducted at decreas- DRB4*01, DQB1*0302, BF S, C4A3 B3. ing annealing temperatures in 2°C steps for two cycles each from 72 to http://www.jimmunol.org/ 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 to an increase in prednisone therapy and subsequent treatment with purified and sequenced as outlined above. azathioprine and methotrexate. An ectopic pregnancy and a preg- Cloning and sequencing of the entire C4d genomic regions for nancy with her second child each precipitated exacerbations of her the C4B genes symptoms. She has had no renal or cardiac disease, no increased DNA fragments corresponding to the C4d region were amplified with syn- susceptibility to infection, hypersensitivity to medications, or cen- thetic PCR primers C4E 22.5 and C4E 31.3 using the High Fidelity PCR tral nervous system disease. Family history revealed a younger system (Boehringer Mannheim, Indianapolis, IN). PCR conditions were male sibling (HLA identical) who suffers from photosensitivity. one cycle at 94°C for 2 min; 10 cycles at 94°C for 15 s, 65°C for 30 s, and Her children, her parents, and her MHC nonidentical brother are by guest on September 27, 2021 72°C for 1 min; 20 cycles at 94°C for 15 s, 56°C for 30 s, and 72°C for 1 alive and well. Laboratory studies showed undetectable total he- 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) molytic complement (CH100) and no C4 protein by radial immu- and cloned into TA cloning vector (Invitrogen). The C4d insert of the nodiffusion in the patient or her HLA identical brother. Quantita- clones was amplified again via PCR, and the purified PCR product was tion of C4 and CH100 in the immediate family members and their restriction digested with NlaIV to determine the presence of either C4A or HLA typing is shown in Fig. 1. In brief, the proposita inherited the C4B. A 467-bp fragment represents C4B, while two fragments (276 and 191 bp) represent C4A (15). Clones representing C4B from both the HLA A2 B39 Cw7 DRB1*1501 DRB5*0101 DQB1*0602, BF-S, proposita’s mother and father were sequenced (ABI Prism). Comparison of C4AQ0 BQ0 from the paternal chromosome, and HLA A2 B40 DNA sequences with national databases was performed by the GCG Cw3 DRB1*1501, DRB5*0101 DQB1*0602, BF-S, C4AQ0 BQ0 FASTA program from the Pittsburgh Supercomputing Center. DNA align- from the maternal chromosome. The C4 haplotypes for the pa- ments were performed with the SeqManII program (DNASTAR). Syn- tient’s father are C4AQO BQO/C4AQ0 B2. The C4 haplotypes for thetic DNA primers for PCR and for cycle sequencing were synthesized by Life Technologies (Grand Island, NY). Details of the primer sequences are the patient’s mother are C4AQ0 BQ0/C4A3 B3. as follows. For amplification of the C4d region of the C4 gene the primer sequences were: C4E 22.5, GAAGGGGCCATCCATAGAGA; and C4E Biosynthesis and secretion of C4 in cell culture 31.3, CTTCAGGGTTCCTTTGCTGT. For sequencing of the C4d region To determine whether a defect in regulation of C4 , of the C4B gene the primer sequences were: C4E 25.3, CAGGTGCTGCT GTCCCGTGA; C4E 26.5, GCTCACAGCCTTTGTGTTGAA; C4E 27.3, C4 synthesis, or secretion accounted for the low to absent C4 pro- CACTCTCTGCTTCAATGGCT; C4E 28.5, GAAGCCTCCATCTCAAA tein in sera of the homozygous deficient patient, primary fibroblast GGC; and C4E 29.3, TTGGGTACTGCGGAATCCCC. cell cultures were established. SDS-PAGE of immunoprecipitates of cell lysates and culture media from [35S]methionine pulse-la- Results beled normal and proposita’s C4 deficient (C4D) fibroblast cul- Family studies and clinical case report tures is shown in Fig. 2. Fibroblasts were incubated with IFN-␥ The patient was a 30-yr-old Finnish woman who developed pho- (500 U/ml) and TNF-␣ (30 ng/ml) to yield maximum C4 expres- tosensitivity, a malar rash, polyarthritis, leukopenia, and positive sion (24). A polypeptide of approximately 185 kDa, corresponding anti-nuclear Ab (1/320), anti-Sm Ab (1/1280), and a weakly pos- in size to pro-C4, was detected in lysates, and native C4 subunits itive rheumatoid factor coincident with her first pregnancy and (␣, ϳ93 kDa; ␤, ϳ78 kDa; ␥, ϳ33 kDa) plus the approximately shortly after delivery. In her serum, complement protein C4 was 125-kDa processing intermediate ␣-␥ fragment were recovered undetectable, and C3 was modestly reduced or at the lower limit of from the culture media in normal fibroblast cultures. Neither normal (74–111 mg/100 ml). A skin biopsy revealed vasculitis pro-C4 nor native C4 protein was detected in IFN-␥- and TNF-␣- with deposition of IgM and C3. Five years later she developed stimulated C4D fibroblast cultures even after prolonged exposure aphthous ulcers and increased joint symptoms, for which she was (10 days) of the autoradiograph. In contrast, both C4D and normal treated with prednisone. An exacerbation of mucositis 1 yr later led fibroblasts synthesized and secreted comparable amounts of C1 3690 COMPLEMENT C4 DEFICIENCY

FIGURE 2. Biosynthesis of C4 protein in fibroblasts from a normal in- FIGURE 4. RNA blot analysis of C4 transcripts from normal (C4S) and dividual and from a C4D patient. Human primary skin fibroblasts were C4D fibroblasts. RNAs were isolated from C4D and normal fibroblasts pulse labeled with [35S]methionine, stimulated with IFN-␥ and TNF-␣ stimulated with IFN-␥ and TNF-␣. Twenty micrograms (lane 1) and 40 ␮g (even-numbered lanes) and with medium alone (odd-numbered lanes), and (lane 2) of RNA in replicate blots were probed for the presence of factor subjected to immunoprecipitation, SDS-PAGE, and autoradiography. Cell B, C4, and actin transcripts. The two upper panels were exposed for about lysates (left panel) and culture media (right panel) from normal cells (lanes panel was exposed for 64 h to visualize C4 mRNA in the Downloaded from ء h, and the 18 1, 2, 5, and 6) and from C4D (lanes 3, 4, 7, and 8) are shown. Exposure C4D cells (data are from one of three experiments showing similar results). time was 24 h (intracellular; left) and 72 h (extracellular; right). Densitometric scanning revealed a ratio of C4 mRNA Ͼ33:1 and of Bf mRNA of 1.5:1.0 in normal and C4D samples, respectively, when normal- inhibitor, a highly IFN-␥-responsive complement protein, and C3, ized for loading by scanning for mRNA actin. a protein similar to C4 in size and postsynthetic processing (Fig. 3,

A and B). C3 synthesis and secretion were more variable among analysis. C4-specific mRNA (5.5 kb) in cytokine-stimulated cells http://www.jimmunol.org/ separate replicate experiments, but the mean C3 expression in C4D was of similar size in normal and C4D fibroblasts, but the amount cells was not significantly different from normal. of C4 mRNA was markedly reduced in the C4D fibroblasts (Fig. 4). Scanning of replicate Northern blots for C4, factor B, and actin C4 mRNA expression confirmed a selective decrease in C4 mRNA (Ͻ3% of normal) in The size and amount of C4 mRNA in normal and C4D IFN-␥- and the deficient cells, with only a minor difference (1.5:1) in Bf TNF-␣-stimulated fibroblasts were estimated by Northern blot mRNA compared with normal cells.

Three nonexpressing C4 genes in the patient by guest on September 27, 2021 The presence of C4 and its neighboring genes for the RCCX mod- ules 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 frag- ments corresponding to CYP21B and CYP21A, respectively; and the 2.5- and 2.4-kb fragments corresponding to TNXB and TNXA

FIGURE 3. Biosynthesis of C1 inhibitor (A) and C3 (B) in skin fibro- blasts. The samples were processed and applied similar to that described in FIGURE 5. Southern blot analysis of the RCCX modular structures in Fig. 2. Media alone (odd-numbered lanes), cytokine-stimulated (even-num- the C4D patient and her parents. Genomic DNAs were digested with TaqI, bered lanes), normal fibroblasts (lanes 1, 2, 5, and 6), and C4D (lanes 3, processed according to Southern’s procedure, and hybridized with 32P- 4, 7, and 8) are shown. Exposure times were 16 h (A)and8h(B). These labeled RP1, CYP21, and TNX probes. Lane 1, Control DNA from the T data are examples of one of our complete sets. cell line MOLT4; lane 2, mother; lane 3, father; lane 4, patient. The Journal of Immunology 3691

Table I. Derived amino acid sequences at the C4d region a C4 typing by restriction mapping and sequence analysis To define and type the C4 genes inherited in this family, a 934-bp Amino Acid Number genomic fragment spanning from exon 25 to the 5Ј end of intron 1054 1101 1102 1105 1106 1157 1182 1188 1191 1213 28 that permits differentiating among specific C4 allotypes (8) was amplified from genomic DNA using primers 5 and 11, subjected to Father a C4AQ0 D P C L D NTVLtcb digestion with NlaIV and EcoO109, and sequenced. The digests a C4BQ0 D L S I H S T A R tcb and the patterns characteristic for specific C4 isotypes at each of b b C4AQ0 D P C L D NTVLtc the C4 genes and sequence analysis are shown in Table I. b C4B2 D L S I H S T A R Mother To ascertain the molecular mechanisms accounting for the non- c C4AQ0 D P C L D NTVLtcb expressed C4A and C4B genes in the proposita, a 267-bp PCR c C4BQ0 (deleted) product spanning the segment from nucleotide 5776 (intron 28) to d C4A3 D P C L D NTVL nucleotide 6043 (exon 29) was generated, and the total PCR prod- d C4B3 G L S I H S T A R Patient uct was directly sequenced to search for a mutation previously a C4AQ0 D P C L D NTVLtcb found in the C4A gene (43). The results (Fig. 6) show a 2-bp (TC) a C4BQ0 D L S I H S T A R tcb insertion after nucleotide 5880. This insertion leads to a frameshift b c C4AQ0 D P C L D NTVLtc mutation, a predicted change in amino acid sequence, and a stop c C4BQ0 (deleted) signal at codon 3 in exon 30. The HLA-identical brother’s ampli- a Complete sequences for this region were obtained but only amino acid residues Downloaded from that vary as a function of allotype are shown (bold, uppercase). fied DNA showed the same sequence. Amplified DNA from her b A2-bp nucleotide insertion at codon 1213 (lowercase). haploidentical brother, her father, and her mother showed both normal and mutant sequences. Based on the amplitude of the genes, respectively. The band intensities for these fragments are peaks, the father has the TC insertion in both C4A and C4B genes, similar; therefore, he is homozygous for RP1-C4(L)-CYP21A- whereas the mother has the 2-bp insertion only in her C4A gene. TNXA-RP2-C4(S)-CYP21B-TNXB. The C4 allotyping results An unrelated control showed only the normal sequence. The http://www.jimmunol.org/ suggested that he has C4AQ0 C4BQ0 and C4AQ0 C4B2. There- proposita showed only the mutated sequence. fore, the father has RCCX bimodular (L-S) haplotypes in both To confirm the 2-bp insertion in the C4BQ0 gene, genomic se- chromosomes. Both C4A and C4B genes are present in the C4AQ0 quences of the C4d region from the father and mother were am- BQ0 chromosome, but these genes are not producing C4 protein. plified by PCR, cloned into TA vector, and sequenced. The C4B Similarly, the C4AQ0 C4B2 chromosome with the RCCX bimo- clones were identified by NlaIV digests (15). Two C4B clones dular (L-S) haplotype is likely to have a C4A pseudogene. from the father and one from the mother were sequenced to com- The mother (lane 2) has 7.0- and 6.0-kb TaqI fragments corre- pletion. The paternal C4B clones contain the C4B isotypic amino sponding to RP1-C4(L) and RP2-C4(L), respectively. The 7.0-kb acid sequence LSPVIH at positions 1101–1106, the C4B-associ- fragment is twice as intense as the 6.0-kb fragment. In addition, her ated amino acid sequence ADLR for the Ch1 Ag at positions by guest on September 27, 2021 3.7-kb CYP21B fragment is twice as intense as the 3.2-kb 1188–1191, and S1157 that is associated with Ch6. Additionally, CYP21A fragment, and the 2.5-kb TNXB fragment is more intense he has the D1054 that is associated with Ch5 (6). In intron 28, both than the 2.4-kb TNXA fragment. These results suggest that she is paternal clones have the nucleotide sequence ggctct that is 14 nu- heterozygous for one RCCX bimodular chromosome, RP1-C4(L)- cleotides downstream of intron donor site. However, the most dis- CYP21A-TNXA-RP2-C4(L)-CYP21B-TNXB, and one RCCX tinguishing feature for both paternal C4B clones is the presence of monomodular chromosome, RP1-C4(L)-CYP21B-TNXB. Since a TC dinucleotide insertion at codon 1213. This is the first defin- she has C4A3 B3 on one chromosome and C4AQ0 BQ0 on the itive proof showing the presence of a 2-bp insertion into a C4B other, the monomodular (L) chromosome corresponds to the gene. Conversely, the C4B sequence for the maternal clone is iden- C4AQ0 C4BQ0 phenotype. One of the C4 genes on this chromo- tical with the C4B3 sequence published previously, which has the some has been deleted. For the patient, the 7.0-kb TaqI fragment for RP1-C4(L) is more characteristic sequences G1054, LSPVIH 1101–1106, S1157, and intense than the 5.4-kb fragment for RP2-C4(S) (lane 4). In addi- ADLR 1188–1191 (5). As expected, the exon 29 sequence is in- tion, the 3.7-kb fragment for CYP21B is more intense than the tact, with no insertion. 3.2-kb CYP21A fragment. She is heterozygous for a RCCX bi- This 2-bp insertion in exon 29 of C4A was previously reported modular chromosome RP1-C4(L)-CYP21A-TNXA-RP2-C4(S)- (43), and the stop codon in the next exon was predicted on the CYP21B-TNXB and one monomodular RCCX chromosome RP1- assumption that normal splicing (excision at the normal splice C4(L)-CYP21B-TNXB. She has two long C4 genes and one short sites) of intron 29 would occur. To directly test this presumption, C4 gene. None of these three C4 genes expresses detectable C4 RNA from the patient’s and normal fibroblasts (stimulated with protein. Lane 1 shows the result of a control DNA (T cell line IFN-␥ and TNF-␣) was amplified by RT-PCR (primer 14 to gen- MOLT4) with bimodular RCCX (L-S) haplotypes. erate cDNA and primers 9 and 13 for PCR amplification) and then

FIGURE 6. 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. 3692 COMPLEMENT C4 DEFICIENCY subjected to sequence analysis. These results revealed an exonic the earlier study that identified this mutation only in C4AQ0, the sequence identical with that obtained from genomic sequencing 2-bp (TC) insertion was surprisingly also present in the patient’s and normal splicing between exons 28–30 (data not shown). C4BQ0 nonexpressed genes, giving the first evidence of the mo- lecular basis for C4B pseudogenes. This 2-bp insertion in exon 29 of the nonexpressed C4BQ0 gene could have been acquired from Discussion the C4AQ0 gene of the HLA B40-positive haplotype by unequal The C4 genes are expressed in liver (32) and in extrahepatic tis- cross-over or gene conversion-like events. It has been noticed that sues, including skin fibroblasts (31). C4 mRNA, 5.5 kb in length, some of the C4BQ0 phenotypes in the population are attributed to programs synthesis of a single-chain preprotein that undergoes ex- the deletion of the C4B gene (as shown here for the HLA B40 DR2 tensive postsynthetic processing to yield the native three-chain C4 haplotype) or to the expression of C4A proteins from the two tan- protein. The availability of primary skin fibroblast culture from the dem C4 loci as documented in the HLA B44 DR6 haplotype (15). C4D patient reported here permitted for the first time an analysis This report shows that the third cause of a C4BQ0 phenotype is of C4 protein synthesis and secretion in an individual with a ge- due to a 2-bp insertion at codon 1213 of the C4B gene that causes netic deficiency of this protein. Even under tissue culture condi- a frameshift mutation. tions designed to maximize C4 expression (24), no pro-C4 syn- thesis was observed in fibroblasts from the C4D patient, whereas Acknowledgments synthesis of pro-C4 and secretion of native C4 were readily de- tectable in control fibroblasts. This defect is selective, since the We thank Barbara Hermann for secretarial assistance, Joie Haviland for C4D fibroblasts synthesized and secreted C1 inhibitor and C3 in expert technical assistance, Dr. Rick Wetsel for helpful review of the Downloaded from manuscript, Dr. Juha Kere for generating the EBV cell lines, and Dr. Seppo amounts comparable to those produced by control cells and in Meri for quantitating the hemolytic activities. 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, References the decrease in C4 protein expression results from markedly re- 1. Carroll, M. C., R. D. Campbell, D. R. Bentley, and R. R. Porter. 1984. A mo- duced steady state C4 mRNA levels (Ͻ3% of normal). lecular map of the human major histocompatibility complex class III region link- http://www.jimmunol.org/ ing complement genes C4, C2 and factor B. Nature 307:237. Analysis of the nuclear family established that the patient in- 2. Hall, R. E., and H. R. Colten. 1977. Cell-free synthesis of the fourth component herited a nonexpressed long C4A gene from her mother, a nonex- of guinea pig complement (C4): identification of a precursor of serum C4 (pro- pressed long C4A gene, and a nonexpressed short C4B gene from C4). Proc. Natl. Acad. Sci. USA 74:1707. 3. Law, S. K. A., A. W. Dodds, and R. R. Porter. 1984. A comparison of the her father. Her maternally derived haplotype, A2; B40; Cw3; DR2 properties of two classes, C4A and C4B, of the human complement component (DRB1*1501, DRB5*0101, DQB1*0602), also contains a deleted C4. EMBO J. 3:1819. C4B gene. C4B deletions have been recognized in association with 4. Isenman, D., and J. R. Young. 1984. The molecular basis for the differences in immune hemolysis activity of the Chido and Rodgers isotypes of human com- B40 (46, 47), while the linkage between B40 and DR2 is unusual. plement component C4. J. Immunol. 132:3019. Her paternally derived haplotype, A2;B39;Cw7;DR2 (DRB1* 5. Yu, C. Y., K. T. Belt, C. M. Giles, R. D. Campbell, and R. R. Porter. 1986. 1501, DRB5*0101, DQB1*0602), is identical at the DR/DQ loci Structural basis of the polymorphism of human complement component C4A and by guest on September 27, 2021 C4B: gene size, reactivity and antigenicity. EMBO J. 5:2873. and differs at the class I loci. A detailed analysis of the basis for 6. Yu, C. Y., R. D. Campbell, and R. R. Porter. 1988. A structural model for the lack of expression of the patient’s nondeleted C4 genes was un- location of the Rodgers and the Chido antigenic determinants and their correla- dertaken. This revealed an identical 2-bp insertion in her three tion with the human complement C4A/C4B isotypes. Immunogenetics 27:399. 7. Yu, C. Y. 1999. Molecular genetics of the human MHC complement gene cluster. nonexpressed C4 genes, which include two C4A genes and one Exp. Clin. Immunogenet. In press. C4B gene. This 2-bp insertion in exon 29 was previously observed 8. Dangel, A. W., A. R. Mendoza, B. J. Baker, C. M. Daniel, M. C. Carroll, L.-C. Wu, and C. Y. Yu. 1994. The dichotomous size variation of human complement by Schneider and colleagues (43) in several C4A null alleles; the C4 gene is mediated by a novel family of endogenous retroviruses which also majority of them were associated with HLA-B60 and DR6. In their establishes species-specific genomic patterns among Old World primates. Immu- study, based entirely on sequence analysis of genomic DNA (PCR nogenetics 40:425. 9. Dangel, A. W., B. J. Baker, A. R. Mendoza, and C. Y. Yu. 1995. Complement amplified), they predicted that this frameshift mutation would gen- component C4 gene intron 9 as a phylogenetic marker for primates: long terminal erate a stop codon in the next exon of the C4 transcript derived repeats of the endogenous retrovirus ERV-K(C4) are a molecular clock of evo- from this gene. Generation of the stop codon would occur only if lution. Immunogenetics 42:41. 10. Chu, X., C. Rittner, and P. M. Schneider. 1995. Length polymorphism of the splicing of intron 29 was executed at the normal splice sites. Ex- human complement component C4 gene is due to an ancient retroviral integra- amples of splicing abnormalities, particularly in the context of mu- tion. Exp. Clin. Immunogenet. 12:74. tations (including frame shifts) have been reported (48–52). Be- 11. Tassabehji, M., T. Strachan, M. Anderson, R. D. Campbell, S. Collier, and M. Lako. 1994. Identification of a novel family of human endogenous retrovi- cause C4 is expressed in a readily available tissue (skin ruses and characterization of one family member, HERV-K(C4), located in the fibroblasts), we were able to establish that splicing at the exon complement C4 gene cluster. Nucleic Acids Res. 22:5211. 29–30 junction was normal and that a truncated C4 translation 12. Yang, Z., A. R. Mendoza, W. Zipf, and C. Y. Yu. 1998. Modular Variations of HLA class III genes RP, complement C4, steroid 21-hydroxylase CYP21 and product would be generated. This truncated C4 polypeptide was tensacin TNX (RCCX): a mechanism for gene deletions and disease associations. not detected in cell culture, probably because the shortened pro-C4 GenBank accession no. AF86641. may be unstable. 13. Shen, L. M., L. C. Wu, S. Sanlioglu, R. Chen, A. R. Mendoza, A. Dangel, M. C. Carroll, W. Zipf, and C. Y. Yu. 1994. Structure and genetics of the partially Nonsense mutations have been associated with marked de- duplicated gene RP located immediately upstream of the complement C4A and creases in steady state levels of specific mRNA in other instances, C4B genes in the HLA class III region: molecular cloning, exon-intron structure, ␤ ␣ composite retroposon and breakpoint of gene duplication. J. Biol. Chem. 269: such as -globin (53), 1-antitrypsin (54), surfactant protein B 8466. (55), cystic fibrosis transmembrane regulator (56), and others (56, 14. Rupert, K. L., R. M. Rennebohm, and C. Y. Yu. 1999. An unequal crossover 57). Although the basis for this observation is not completely un- between the RCCX modules of the HLA leading to the presence of a CYP21B gene and a tenascin TNXB/TNXA-RP2 recombinant between C4A and C4B genes derstood (52), a similar mechanism could account for the de- in a patient with juvenile rheumatoid arthritis. Exp. Clin. Immunogenet. In press. creased levels of C4 mRNA observed in this patient (see Fig. 4). 15. Yu, C. Y., and R. D. Campbell. 1987. Definitive RFLPs to distinguish between Nonetheless, sufficient C4 mRNA was present in cytokine-stimu- the human complement C4A/C4B isotypes and the major Rodgers/Chido deter- minants: application to the study of C4 null alleles. Immunogenetics 25:384. lated fibroblasts to allow nucleotide sequence analysis of RT-PCR 16. Schneider, P. M., M. C. Carroll, C. A. Alper, C. Rittner, A. S. Whitehead, products from the homozygous deficient fibroblasts. In contrast to E. J. Yunis, and H. R. Colten. 1986. Polymorphism of human complement C4 and The Journal of Immunology 3693

steroid 21-hydroxylase genes: restriction fragment length polymorphisms reveal- 39. Schwartz, R. J., J. A. Haron, K. N. Rothblum, and A. Dugaiczyk. 1980. Regu- ing structural deletions, homoduplications, and size variants. J. Clin. Invest. 78: lation of muscle differentiation: cloning of sequences from ␣-actin messenger 650. ribonucleic acid. Biochemistry 19:5883. 17. Braun, L., P. M. Schneider, C. M. Giles, J. Bertrams, and C. Rittner. 1990. Null 40. Yu, C. Y. 1991. The complete exon-intron structure of a human complement alleles of human complement C4: evidence for pseudogenes at the C4A locus and component C4A gene: DNA sequences, polymorphism, and linkage to the 21- for gene conversion at the C4B locus. J. Exp. Med. 171:129. hydroxylase genes. J. Immunol. 146:1057. 18. Hauptmann, G., G. Tappeiner, and J. Schifferli. 1988. Inherited deficiency of the 41. Barnes, W. M. 1992. The fidelity of Taq polymerase catalyzing PCR is improved fourth component of human complement. Immunodefic. Rev. 1:3. by an N-terminal deletion. Gene 112:29. 19. Hauptmann, G. 1989. Frequency of complement deficiencies in man, disease 42. Don, R. H., P. T. Cox, B. J. Wainwright, K. Baker, and J. S. Mattick. 1991. associations and chromosome assignment of complement genes and linkage ‘Touchdown’ PCR to circumvent spurious priming during gene amplification. groups: a summary of the data from the literature. Complement Inflamm. 6:74. Nucleic Acids Res. 19:4008. 20. Liszewski, M. K., and J. P. Atkinson. 1991. The role of complement in autoim- munity. Immunol. Ser. 54:13. 43. Barba, G., C. Rittner, and P. M. Schneider. 1993. Genetic basis of human com- 21. Fan, Q., B. Uring-Lambert, B. Weill, C. Gautreau, C. J. Menkes, and M. Delpech. plement C4A deficiency: detection of a point mutation leading to nonexpression. 1993. Complement component C4 deficiencies and gene alterations in patients J. Clin. Invest. 91:1681. with systemic lupus erythematosus. Eur. J. Immunogenet. 20:11. 44. Katz, Y., and R. C. Strunk. 1989. Synthesis and regulation of C1 inhibitor in 22. Howard, P. F., M. C. Hochberg, W. B. Bias, F. C. J. Arnett, and R. H. McLean. human skin fibroblasts. J. Immunol. 142:2041. 1986. Relationship between C4 null genes, HLA-D region antigens, and genetic 45. Katz, Y., and R. C. Strunk. 1988. Synovial fibroblast-like cells synthesize seven susceptibility to systemic lupus erythematosus in Caucasian and black Ameri- proteins of the . Arthritis Rheum. 31:1365. cans. Am. J. Med. 81:187. 46. Partanen, J., and R. D. Campbell. 1989. Restriction fragment analysis of non- 23. Bishof, N. A., T. R. Welch, and L. S. Beischel. 1990. C4B deficiency: a risk deleted complement C4 null genes suggests point mutations in C4A null alleles, factor for bacteremia with encapsulated organisms. J. Infect. Dis. 162:248. but gene conversions in C4B null alleles. Immunogenetics 30:520. 24. Olsen, M. L., R. Goldstein, F. C. Arnett, M. Duvic, M. Pollack, and 47. Fredrikson, G. N., L. Truedsson, A. G. Sjoholm, and M. Kjellman. 1991. DNA J. D. Reveille. 1989. C4A gene deletion and HLA associations on black Amer- analysis in a MHC heterozygous patient with complete C4 deficiency: homozy- icans with systemic lupus erythematosus. Immunogenetics 30:27.

gosity for C4 gene deletion and C4 pseudogene. Exp. Clin. Immunogenet. 8:29. Downloaded from 25. Terasaki, P. I., D. Bernoco, M. S. Park, G. Ozturk, and Y. Iwaki. 1978. Micro- 48. Lin, C. Y., and J. L. Huang. 1993. Molecular study of a Chinese C3 deficient droplet testing for HLA-A, -B, -C, and -D antigens: the Phillip Levine Award patient. Mol. Immunol. 30:29. Lecture. Am. J. Clin. Pathol. 69:103. 26. Duffy, B. F., P. DeTogni, D. Chaplin, and T. Mohanakumar. 1993. Simultaneous 49. Higashi, Y., A. Tanae, H. Inoue, T. Hirosama, and Y. Fujii-Kuriyama. 1988. DRB1, DRB3 and DRB5 typing by chemiluminescent reverse dot blot. Hum. Aberrant splicing and missense mutations cause steroid 21-hydroxylase Immunol. 37S:140. [P-450(C2)] deficiency in humans: possible gene conversion products. Proc. 27. Marcus, D., and Alper, C. A. 1986. Manual of Clinical Laboratory Immunology. Natl. Acad. Sci. USA 85:7486. American Association for Microbiology, Washington, D.C. 50. Botto, M., K. Y. Fong, A. K. So, A. Rudge, and M. J. Walport. 1990. Molecular

28. Mancini, G., A. O. Carbonara, and J. F. Heremans. 1965. Immunochemical quan- basis of hereditary C3 deficiency. J. Clin. Invest. 86:1158. http://www.jimmunol.org/ titation of antigens by single radial immunodiffusion. Immunochemistry 2:235. 51. Gayther, S. A., W. Warren, S. Mazoyer, P. A. Russell, P. A. Harrington, 29. Rooney, D. E., and B. H. Zebulkowski. 1986. Cytogenetics: A Practical Ap- M. Chiano, S. Seal, R. Hamoudi, E. J. Vanrensburg, A. M. Dunning et al. 1995. proach. IRL Press, Oxford. Germline mutations of the BRCA1 gene in breast and ovarian cancer families 30. Johnson, C. A., P. Densen, R. Wetsel, F. S. Cole, N. E. Goeken, and H. R. Colten. provide evidence for a genotype-phenotype correlation. Nat. Genet. 11:428. 1992. Molecular heterogeneity of C2 deficiency. N. Engl. J. Med. 326:874. 52. Cheng, J., and L. E. Maquat. 1993. Nonsense codons can reduce the abundance 31. Kulics, J., A. Circolo, R. C. Strunk, and H. R. Colten. 1994. Regulation of syn- of nuclear mRNA without affecting the abundance of pre-mRNA or the half-life thesis of complement protein C4 in human fibroblasts: cell- and gene-specific of cytoplasmic mRNA. Mol. Cell. Biol. 13:1892. effects of cytokines and lipopolysaccharide. Immunology 82:509. 53. Baserga, S. J., and E. J. J. Benz. 1988. Nonsense mutations in the human ␤-globin 32. Morris, K. M., D. P. Aden, B. B. Knowles, and H. R. Colten. 1982. Complement gene affect mRNA metabolism. Proc. Natl. Acad. Sci. USA 85:2056. biosynthesis by the human hepatoma-derived cell line HepG2. J. Clin. Invest. 54. Satoh, K., T. Nukiwa, M. Brantly, R. I. J. Garver, M. Hofker, M. Courtney, and 70:906. R. G. Crystal. 1988. Emphysema associated with complete absence of ␣1-antit- 33. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the by guest on September 27, 2021 rypsin in serum and the homozygous inheritance [corrected] of a stop codon in an head of bacteriophage T4. Nature 227:680. ␣1-antitrypsin-coding exon. Am. J. Hum. Genet. 42:77. 34. Chirgwin, J. M., A. E. Przybyla, R. J. MacDonald, and W. J. Rutter. 1979. Isolation of biologically active ribonucleic acid from sources enriched in ribo- 55. Nogee, L. M., G. Garnier, H. C. Dietz, L. Singer, A. M. Murphy, D. E. deMello, nuclease. Biochemistry 18:5294. and H. R. Colten. 1994. A mutation in the surfactant protein B gene responsible 35. Virca, G. D., W. Northemann, B. R. Shiels, G. Widera, and S. Broome. 1990. for fatal neonatal respiratory disease in multiple kindreds. J. Clin. Invest. 93: Simplified Northern blot hybridization using 5% sodium dodecyl sulfate. Bio- 1860. techniques 8:370. 56. Hamosh, A., B. C. Trapnell, P. L. Zeitlin, C. Montrose-Rafizadeh, 36. Johnson, C. A., P. Densen, R. K. Hurford, Jr., H. R. Colten, and R. A. Wetsel. B. J. Rosenstein, R. G. Crystal, and G. R. Cutting. 1991. Severe deficiency of 1992. Type I human complement C2 deficiency. J. Biol. Chem. 267:9347. cystic fibrosis transmembrane conductance regulator messenger RNA carrying 37. Miller, S. A., D. D. Dykes, and H. F. Polesky. 1988. A simple salting out pro- nonsense mutations R553X and W1316X in respiratory epithelial cells of patients cedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16: with cystic fibrosis. J. Clin. Invest. 88:1880. 1215. 57. Kadowaki, T., H. Kadowaki, M. M. Rechler, M. Serrano-Rios, J. Roth, 38. Belt, K. T., M. C. Carroll, and R. R. Porter. 1984. The structural basis of the P. Gorden, and S. I. Taylor. 1990. Five mutant alleles of the insulin receptor gene multiple forms of human complement component C4. Cell 36:907. in patients with genetic forms of insulin resistance. J. Clin. Invest. 86:254.