Sie sind auf Seite 1von 6

THEJOURNAI OF BIOLOGICAL

CHEMISTRY Vol. 269,No. 45,Issue of November 11, pp. 2849626499, 1994


0 1994 by The American Society for’ Biochemistry and Molecular Biology, h e . Printed in U.S.A.

Inherited Human Complement C3 Deficiency


AN AMINO ACID SUBSTITUTION IN THE @-CHAIN(ASP549
TO ASN) IMPAIRS C3 SECRETION*

(Received for publication, June 29, 1994, and in revised form, September 9, 1994)

Lori Singer$§, William T.Whitehead$, Hideto AkamaS, Yitzhak Katzn, Zvi Fishelsonil,and
Rick A. WetselSll**$$
From the $Edward Mallinckrodt Department of Pediatrics and Il**Department of Molecular Microbiology, Washington
University School of Medicine, St. Louis, Missouri 63110, the nunit of Allergy and Immunology and Department of
Pediatrics, Afiliated to the Sackler Faculty of Medicine, IL.1 Aviv University School of Medicine, Israel, and lllepartment
of Cell Biology and Histology, Sackler School of Medicine, Tel Aviv University, Israel

We recently described a case of hereditary comple- deficient individual, providing additional evidence
ment C3 deficiency (C3D) in a New Zealand male who that multiple defects cause inherited C3 deficiency in
has a small amount of serum C3 (7 pg/ml), a normal size humans.
6.2-kilobase C3 mRNA that is present in normal quanti-
ties, and a normal size M, 180,000 proC3 molecule that
is synthesized in normal amounts. Secretion of C3 from
this patient’scellswas greatly diminished,however, Inherited complete human C3l deficiency, first described in
and an aberrant C3 trypsin cleavage profile indicated 1972 (l),is characterized by recurrent infections with pyogenic
an abnormality in the proC3 structure. To determine encapsulatedbacteriaandcertain immune-complex related
the primary structure of the C3D proC3 molecule, the disorders including membranoproliferative glomerulonephri-
corresponding cDNA was cloned and sequenced in the tis, systemic lupus erythematosus, and vasculitis(reviewed in
present study, revealing a normal signal peptide, tetra- Ref. 2). These clinical manifestations reflect the biologic func-
arginine linker, and thiolester domain. One nucleotide tions mediated by C3 and C3 peptides including immune cy-
substitution in exon13 (G”O6AC to AAC) wasfound, tolysis, phagocytosis, and non-cytotoxic enzyme release(re-
however, that resulted in an amino acid change in a viewed in Refs. 2,3), inhibition of immune-complex lattice
highly conserved region of the C3 Pchain (Aspas to formation, (41, and B cell activation and proliferation (5, 6).
Asn). This substitution has not been described in any C3 is a two-chain glycoprotein (7, 8) (a-chain, M , 110,000;
individual with either C3 Fast or C3 Slow phenotypes. P-chain, M , 70,000) that is synthesized as a single chain pre-
Immunoprecipitation ofC3 fromL-cells transfected proprotein (9) that undergoes co- and post-synthetic modifica-
with full-length normal and C3DcDNAs demonstrated tions involving cleavage of a signal peptide, excision of an in-
that C3 was secreted by the cells transfected with the terchain linking peptide(9, lo), andglycosylation (11, 12). The
normal C3 cDNA; however, onlya C3 precursor was de- C3 protein is translated from a 5.2-kb mature mRNA that is
tected in the intracellular compartment of the cells transcribed from a 42-kb gene of 41 separateexons (13, 14) on
transfected with the C3DcDNA and none detected ex- chromosome 19 (15). C3 is synthesized predominately by hepa-
tracellularly.Immunofluorescence studies revealed a tocytes (16);it is also expressed at extrahepatic sites (reviewed
perinuclear localization of C3 in the C3D transfectants, in Ref. 17), including skin fibroblasts (18). In humans, two
suggesting that transportof the mutant precursor C3 is
major C3 protein polymorphisms have beenidentified, based on
arrested early in the secretory pathway. Allele-specific
polymerase chain reaction analysis demonstrated that the agarose gel electrophoretic mobility of the allotypes (19).
this New Zealand family is a compound heterozygous The C3 slow (C3S) and C3 fast (C3F) variants exhibit very
C3D kindred, with the Asnm9point mutation being in- similar specific hemolytic activities (20) and have been de-
herited from the mother and a yet undescribed C3 de- scribed in manypopulations; C3S is the more common allotype
fect being inherited from the father. Taken together, with a frequency of 0.79, 0.95, 0.97, and 0.99 in Caucasian,
these data indicate that 1) C3 deficiency is caused in a African American, South American Indian, and Asianpopula-
New Zealand kindred by two distinct molecular genetic tions, respectively.
mutations, one being an amino acid substitution in a Since 1972, inherited C3deficiency has been described in 16
highly conserved region of the P-chain that results in families representing a variety of ethnic and national origins
impaired C3 secretion, and 2) the molecular basis of (1,21-35). Recently, the molecular basis of C3 deficiency has
this deficiency has not been described in any other C3- been reported in four of these kindreds:two 5‘-donor splice site
mutations (introns10 (36) and 18 (3111, a gene deletion (includ-
* This work was supported by United States Public Health Service ing exons 22 and 23) (371, and a point mutation (affecting a
Grants AI25011,AI24739,andHL17461,and Research Grant 1557 Factor I cleavage site) (35). We recently described the initial
from the Chief Scientist, Ministry of Health, Israel. Thecosts of publi- investigation of inherited C3deficiency in a New Zealand male
cation of this article were defrayed in part by the payment of page (33, 38). Our studies demonstrated that this individual has a
charges. This article must thereforebe hereby marked “advertisement” small amount of serum C3 (7yg/ml), a normal size 5.2-kb C3
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
5 This investigation was performed in partial fulfillment of the re- mRNA that is present in normal quantities, anda normal size
quirements for the degree of Doctor of Philosophy in Molecular Genet- M, 180,000proC3 molecule that is synthesizedinnormal
ics, Washington University,St. Louis, MO.
$$ Recipient of Research Career Development Award AI00919 from
the National Institutes of Health. To whom correspondence should be The abbreviations used are: C3, the third complement component;
addressed:Dept. of Pediatrics, Box 8116, Washington University School BSA, bovine serum albumin; C3D, C3 deficiency; IL-lp, interleukin-lp;
of Medicine, 400 S. KingshighwayBlvd., St. Louis, MO 63110.Tel.: kb, kilobase pair(s);PBS, phosphate-buffered saline, PCR, polymerase
314-454-2285;Fax: 314-454-2476. chain reaction; vWD, von Willebrand.

28494
Human C3 Deficiency 28495
amounts (38). Processing and secretion of C3 in this patient's for the C3 FasffSlowpolymorphism and the C3D mutation at nucleotide
cells was greatly diminished, however, and an aberrant C3 positions 364 and 1705, respectively. All reactions were performed by
initially denaturing 200 ng of genomic DNAat 95 "C for 5 min with 200
trypsin cleavage profile indicated an abnormality in the proC3 ng of each oligonucleotidein a25-pI solution containing 20 mM Tris, pH
structure (38). The exact molecular genetic mutation that 8.5, 16 mM ammonium sulfate, 2.5 m~ MgCI,, and 150 pg/ml BSA, 1.6
causes C3deficiency inthis individual was not defined, mM dNTPs, and 3 units of Klentaq I enzyme (from Dr. Wayne Barnes,
however. Washington University School of Medicine, St. Louis, MO). The follow-
Accordingly, in this report we have characterized the struc- ing oligonucleotidepairs were used in the allele-specificPCR reactions
ture of the C3 cDNA using RNA isolated from the skin fibro- (see below for sequences):Fast, W766 and W767; Slow, W766and W768;
blasts of the C3-deficient individual. This study, together with Normal, W775 and W776; C3-deficient, W775and W777.
transfection and allele-specific PCR experiments, demon- Oligonucleotide/Primer Synthesis-All primers used for(c)DNAam-
plification and sequencing were synthesized by the automated DNA
strated that C3 deficiency is caused in this kindred by two
synthesizer PCR-Mate, model 391 (Applied Bio-systems, Inc., Foster
distinct molecular genetic defects, one being a maternally in- City, CA). The primers used for reverse transcription, amplification of
herited point mutation in exon 13 resulting in a single amino the cDNA, genomic DNA,and allele specific PCR are detailed below. All
acid substitution in the P-chain that impairs C3 secretion. oligonucleotides are identical t o the cDNA sequence described previ-
ously (10)except where indicated by lowercase letters. Restriction en-
EXPERIMENTAL.PROCEDURES zyme recognition sites are underlined. Reverse transcription: W208,
Fibroblast Cultures-Fibroblast cell lines were established from the 5'-GCGGTCCA?TCGCAGGAGGAAGTTGACGTl"3'; cDNA amplifica-
C3-deficient New Zealand family members and from normal individuals tion:W519, 5'-CTGTCCCTCTGcggCcGCACTGTCCCAGCA-3' (NotI);
according to published methods (18). Fibroblasts were maintained at W207, 5'-CCCCGGGTCTGAGCTCTGTAGGTAGCACTG-3'; genomic
37 "C in 5%CO, in Dulbecco's modified Eagle's medium (Life Technolo- DNA W495,5'-GCGAGAGCCCGGCCAGGACCTGGTGGTGCT-3'; W346,
gies, Inc.) supplemented with 10%fetal calf serum, 2mM glutamine, 100 5'- CTGCTGCCCAGGTACAGGCT-3'; allele-PCR W766,5'-ACCCACAT-
unitdml penicillin, and 100 pg/ml streptomycin. GGGCAACGTCACCTI'CACG3'; W767,
5"GGCCTGCACGGTCACGA-
Isolation of RNA-Normal and C3-deficientfibroblasts were grownto AC'ITG'ITGCC-3';W768, 5"GGCCTGCACGGTCACGAAC'ITG"G€G
confluence in 162 cm2 flasks and stimulated with 10 ng/ml IL-lp (pro- 3'; W775, 5'-CGCACGAAGAAGCAGGAGCTCTCG3'; W776, 5"CTI'GA-
vided by Dr.John McKearn, Monsanto, St. Louis, MO) 20 h prior to RNA CGTCCACCCACACGGAGTC-3';W777,5"CTFGACGTCCACCCACACG
harvest. Approximately lo8 cells were lysed, and RNA was harvested GAG"-3'. All primers used for DNAsequencing were 20-mers, identical to
using the method described by Chirgwin et al. (39). RNA was quanti- the SP6 or T7 promoters orto the C3 cDNA sequence (10).
tated by absorbance a t 260 nm. Polyadenylated mRNA was separated Construction of Full-Length C3-deficient and Normal cDNA Mam-
on an oligo(dT)-cellulosecolumn (Collaborative Research Products, malian Expression Vectors-Full-length C3-deficientand normal cDNA
Bedford, M A ) using previously described methods (40). clones were constructed fromoverlapping, cDNA clones using the
Construction of cDNA Library and Amplification of C3D cDNA-Ten unique SnaBI and SphI sites in the C3 cDNA (10).The full-length C3
pg of poly(A)+ mRNAwas employed in the construction of the oligo(dT) cDNAs were subcloned into pRcfCMV (Invitrogen) using engineered
cDNA library using the cDNA synthesis method of Gubler and Hoffman NotI restriction sites. This vector providesthe human cytomegalo-virus
(41) and the reagents supplied in a cDNA synthesis kit (Invitrogen, San promoter and enhancer, bovine growth hormone polyadenylationsignal,
Diego,CAI. After addition of EcoRI-Not1 adapters (Invitrogen), the and neomycin resistance.
cDNAwas ligated to A-Zap I1vector arms (Stratagene, La Jolla, CA) and Dansfection and Biosynthetic Labeling and Immunoprecipitation-
in vitro packaged using the Gigapack Gold packaging extract (Strat- Murine L-cells (American TypeCulture Collection, Rockville,MD) were
agene). Over lo6 recombinants were prepared and plated. C3 cDNA- grown to 40-50% confluence (-5 x lo4 cells/60 mm2 dish) and trans-
containing phage were identified by hybridization with a nick-trans- fected with the cDNA constructs using the CaPO, method (46) using 15
lated human C3 cDNA fragment (BstEII fragment of pC3.59 or BglII pg of plasmid. Precipitates were removed after 5 h, the cells washed
fragment of pC3.11, (10)).Phagemids were prepared as described in the twice, and fresh medium was added and incubated 48 h. Cells werethen
Stratagene A-Zap I1 protocol (Stratagene), and the C3 inserts were incubated with complete medium containing 400 pg/ml G418 (Geneti-
characterized by EcoRI and NotI digestion and 1%agarose gel electro- cin, LifeTechnologies, Inc.). Cells transfected with normal and C3-
phoresis. The longest C3 cDNAclone (pAGI.11) of 4 kb was isolated and deficient cDNAs were subcloned to select C3-producing cells. Afterse-
sequenced. 5'-C3 cDNA clones that overlapped pAGI.11 were obtained lection of stable transfectants, biosynthetic labeling experiments were
using amplification of single-stranded cDNA. Single-stranded cDNA performed by incubating cells for 6 h in Dulbecco's modified Eagle's
was synthesized from 1pg of AGI total RNA using the "cDNA CycleKit" medium containing 10% dialyzed fetal calf serum and [35S]methionine
(Invitrogen). The C3-specific primer, W208, usedfor reverse tran- and [35Slcysteine(ICN, Costa Mesa, CAI, 250 pCi/ml (specific activity:
scriptase is described below. A 1.3-kb fragment was amplified by the approximately 1000 Ci/mmol). After the pulse period, the medium was
polymerase chain reaction (42) using the followingconditions. The removed, and the cells were washed twice with cold Hanks' balanced
single-stranded cDNA product was initially denatured at 95 "C for 3 salt solution; the cells were lysed by freeze-thawing in the presence of
min with 30 pmol of oligonucleotides W519 and W207 (see below) in a protease inhibitors. The supernatant and cell lysates were prepared
100-pl solution containing 10 m~ Tris-C1, pH 8.3, 50 mM KC1, 1.5 mM for immunoprecipitation, preabsorbed with Staphylococcal protein A
MgCI,, 0.1% gelatin, 200 p~ dNTPs, and 3 units of Ilhq DNApolymerase (Bethesda Research Laboratory), then incubated in the presence of the
(Perkin Elmer Cetus). Followingthe initialdenaturation, the cDNAwas IgG fraction of monospecific goat anti-human C3 antibody (Atlantic
amplified by melting at 95 "C for 1min, annealing at 62 "C for 2 min., Antibody, Scarborough,ME) overnight at 4 "C. StaphylococcusAprotein
and polymerizing at 72 "C for 3 min. Thirty-five cycles of amplification was added to capture antigen-antibody complexes. Immunoprecipitates
were performed using a Tempcycler(Coy Laboratory Products, Ann were washed, redissolved, and analyzed by 1%SDS, 7.5% polyacrylam-
Arbor,MI) followed bya final elongation at 72 "C for7 min. This product ide gel electrophoresisunder reducing conditions. The gel wasdried and
was digested with NotI (engineered site) and SmaI (endogenous site), exposed to Hyperfilm (Amersham Corp.).
gel purified, and subcloned into pBluescript I1 (Stratagene). Competent Indirect Immunofluorescence-Indirect immunofluorescence of
SURE cells (Stratagene) were transformed with the plasmids. These transfected cells was performed as described (47) with minor modifica-
clones were subjected to DNA sequence analysis as outlined below. tions. L-cells transfected with normal and AGI C3D cDNAand untrans-
DNA SequenceAnalysis-All cloned DNA sequencing was performed fected cells weretransferred onto glass coverslips. Cells were incubated
using double-stranded templates (43). Two pg of template were dena- at 37 "C for 48h, rinsed in cold water for 5 min, and permeabilized with
tured in 0.2 M NaOH, 0.2 m~ EDTA, neutralized, annealed with SP6, 0.2% Triton X-100 in PBS (5 mM phosphate, 150mM NaC1, pH 7.4) for20
T7, or C3-specificprimers, and sequenced employingthe dideoxy chain min at room temperature. The cells were incubated overnight at 37 "C
termination method (44) and the modified bacteriophage T7 DNA po- with goat anti-human C3 antibody (IgG fraction, Atlantic Antibody), or
lymerase, Sequenase, (United States BiochemicalCorp., Cleveland, purified goat IgG as a negative control (Pierce) (0.8 pg/ml, each). Cells
OH). Direct sequencing ofPCR fragments was performed using the were washed three times with PBS containing 1%Triton X-100, 0.2%
protocol and reagents of the Promega fmol sequencing system (Pro- Tween 20, and incubated with biotin-conjugated rabbit anti-goat IgG
mega, Madison, WI).All DNA was sequenced on both strands. antibody (Pierce) (diluted 1:5000 in PBS containing 3% BSA) for1 h at
Allele-specific PCR Reactions-Genomic DNA isolated as described room temperature. This incubation was followed by an incubation with
previously (45) from the New Zealand C3-deficient family members as avidin-fluorescein isothiocyanate (Boehringer Mannheim) (diluted
well as 70 unrelated individuals was examined by allele-specific PCR 1:300 in PBS with 3% BSA) for1h at room temperature, then washed
28496 Human C3 Deficiency
Procedures.” Both normal and AGI stable L-cell clones were
selected that contained equal numbers of transfected cDNA,
and C3 expression in these cells was analyzed by immunopre-
cipitation using a monospecific polyclonal goat anti-human C3
antibody (Fig. 3). Immunoprecipitation of C3 from the normal
C3 cDNA transfectants is shown in lanes 1 and 2. In lane 1,
proC3 is expressed in the intracellular compartment, and the
mature C3, seen as a- and P-chains, is detected in the extra-
cellular medium (lane 2). In contrast,proC3 is expressed in the
intracellular compartment of the L-cells transfected with the
AGI mutant(lane 31, but no secreted C3 is detected in
T C the extracellularmedium (lane 4 ) . In addition, higher quanti-
ties of p r o 4 3 a r edetected in theAGI transfectants compared
to the normal transfectants, suggesting that pro-C3 accumu-
lates inside the L-cells transfected with the AGI C3 cDNA as
FIG.1. Nucleotide sequence of the AGI C3DcDNA. Sequence has also been observed in the primary fibroblasts from AGI
analysis of cDNA clone (sense strand) showing the Gl7OSto A transition (38). These data indicate that the Asp549to Asn amino acid
of asparagine for aspartic
a t codon 549 which results in the substitution
acid. Codon positions are taken from the published cDNA sequence of substitution causes the impairment in C3 secretion in theAGI
proC3 (10). cells.
Intracellular Processing of proC3 Examined by Immuno-
a s above. The coverslips were immersed in a 0.1% solution of Evans fluorescence-Immunofluorescence studies of the transfected
Blue (48)a t room temperature for 20 min as a counterstain. The cov- L-cells were performed as outlined under “Experimental Pro-
erslips were washed gently for 5 min, rinsed in distilled water, dried
thoroughly, and mounted on glass slides with 50% glycerol in PBS. cedures.” Cells expressing the normal C3 showed a diffuse pat-
1,4-Diazabicyclo-(2.2.2.)-octane(Sigma) (2.5%)was included as an anti- tern of fluorescence throughout the cell (Fig. 4, a and 6). In
fade reagent in the mounting media. contrast, only perinuclear immunofluorescence was observed in
cells transfected with the Asd4’ mutant cDNA (Fig. 4,c and d 1.
RESULTS In addition, theoverall fluorescence signal was approximately
Characterization of the AGIC3D cDNA-From our previous 10 timesmore intense in the AGI transfectants compared with
studies, we determined that C3 secretion was impaired in the the normal transfectants (note exposure times in Fig. 4). The
proband’s (AGI) cells because of an apparent structuralabnor- negative controls of untransfected L-cells (Fig. 4, e and f and
mality in C3 (38).To determine if the impairment inC3 secre- L-cells mock-transfected with vector alone (data not shown)
tion was caused by a defect in the primary structureof proC3, showed only minimal background immunofluorescence, dem-
the sequence of the AGI C3Dmessage was determinedby cDNA onstrating theabsence of autofluorescence in theL-cells trans-
cloning strategies. An oligo(dT) cDNA library was constructed fected with the normal and AGI cDNA. The increased fluores-
employing mRNA isolated from stimulated (IL-lP) AGI fibro- cence observed in theAGI transfectansts is inaccord with the
blastsas described under“Experimental Procedures.” The immunoprecipitation data, indicating intracellular accumula-
longest cDNA clone, 4 kb, representing -80% of the entire C3 tion of the AS^"^^ proC3. Furthermore, theperinuclear localiza-
coding and 3”untranslated sequence was isolated and fully tion of the immunofluorescence in the L-cells transfected with
sequenced. In addition, mRNA-specific reverse transcription the AGI cDNA suggests that the mutant proC3 protein is ar-
and amplification (described under“Experimental Proce- rested early in the secretory pathway.
dures”) wasused to complete the 5’-end of the cDNA sequence. Determination that AGI is a Compound C3D Heterozygoteby
Comparison of the AGI cDNA sequence to the normal cDNA
Allele-specificP C R T o determine if both C3 null alleles of the
sequence (10) revealed a normal signal peptide, tetra-arginine
C3D New Zealandcontain theto A mutation, allele-spe-
linker, and thiolester domain in theC3D proC3 structure. Five
cific PCR reactions were performed using genomic DNA iso-
nucleotide differences were found in the C3D cDNA three of
lated from all family members as described under “Experimen-
the nucleotide sequence differences did not cause amino acid
tal Procedures” (Fig. 5). In addition, all family members were
changes (CI6l5to A, CZRo5 to G, and F956 to C). One nucleotide
substitution (C364to G) results in an amino acid change (ArgIo2 examined for the C3S/F polymorphism at nucleotide position
to Gly) that has been described previously as a polymorphic 364 by a similar allele-specific PCR strategy (“Experimental
variant associated withthe C3 proteinSlow and Fastallotypes, Procedures”). The results from this study reveal that: 1)AGI
respectively (491, indicating that theAGI C3D cDNA codes for and hishomozygous C3-deficient sister areC3D compound het-
the Fast allotype. The final nucleotide substitution, G”05 to A erozygotes, 2) the G1705to Amutation is maternally inherited, 3)
(Fig. 11, results in an amino acid change, Asp549to Asn, in the the father harborsa yet undescribed molecular mutation that
P-chain of the proC3 molecule that has not been previously causes C3 protein deficiency, and 4) both C3 alleles in this
described. The results from the cDNA analysis are presented C3-deficient kindred contain G364which is associated with the
schematically in Fig. 2. C3 Fast allotype. These results aresummarized in Fig. 6.
Expression of Full-length AGI C3D and Normal C3 cDNA in In addition to the New Zealand C3D kindred, DNA from 70
Murine L-cells-To determine if the aminoacid change, Asp549 unrelated Caucasian individuals was examined by allele-spe-
to Asn, causes the impairment in C3 secretion observed in the cific PCR for the t o A mutationandthe C3S/F polymor-
AGI cells, stable transfection experimentswere performed us- phism a t nucleotide position 364. The PCR results demon-
ing two full-length C3 cDNA that differed only a t nucleotide stratedthat seven (10%) individuals inthis group are
1705. This nucleotide difference resultsinthe amino acid homozygous for the C3F polymorphism, 12 (17%) individuals
change at residue 549. Both cDNAs encode the C3 Fast allo- are C3F/S heterozygotes, and that the remaining51 (73%) in-
type, with Gly at residue 102. dividuals are homozygous for the C3S polymorphism; however,
The cDNA expression vectors were constructed and trans- the to A mutationwas not presentinany individual ex-
fected into murine L-cells as described under “Experimental amined, including all 27 that contained the C3Fpolymorphism.
Human C3 Deficiency 28497

1 4
s-c
p
10
2 954, NORMAL
8-CHAIN IC381 .-CHAIN
I I I I I’ C3 cDNA
CS64 G1705 T

1 ?
s-c
9102 9 5 4 9 AGI
8-CHAIN IC381 .-CHAIN
I I I I I‘ C3D cDNA
0304 A C

FIG.2. Comparison of the normal and AGI C3D cDNAsequences. Schematic diagram of normal C3 cDNAsequence (10)and AGI C3D cDNA
sequence (5.2 kb). Nucleotide differencesare shown below the figures (see textfor nucleotide positions).Amino acid differences are shown circled,
above the figures. kg’’* to Gly is the previously described polymorphismthat results in the Slow and Fast allotypes, respectively (49); Aspi4’to

.%-3
200 -
--
Asn is the mutationof interest. Also shown above the figure is the relativeposition of the region coding for thiolester in the a-chain.

Normal C3D

f Pro-C3

f a-chain
100-
97 -

+ P-chain
68 -

Lane 1 2 3 4
FIG.3. C 3 synthesis in L cells transfected with AGI C3D and
normal cDNA. Shown is an autoradiogram of a SDS-polyacrylamide
gel (7.5% in reducing conditions)of immunoprecipitated C3 [%]methi-
onine-labeled fibroblasts treated as described under “Experimental Pro-
cedures.” Intracellular C3 was immunoprecipitated from cell lysates
(lanes 1 and 3)and extracellular C3 was immunoprecipitatedfrom cell
supernatants (lanes 2 and 4 ) . Lanes 1 and 2 are from L cells transfected
with normal C3 cDNA lanes 3 and 4 are from the L cells transfected
with AGI C3D cDNA.

I I
DISCUSSION
FIG.4. Indirect immunofluorescence microscopy of m u r i n e L
We previously demonstrated that C3 deficiency in the pro- cells transfected with normal and m u t a n t h u m a n C 3 cDNAs.
band (AGI) of a New Zealand familywas the result of impaired Protocol is described under “Experimental Procedures.” The patterns
secretion of C3 most likely causedby a structural defect in the shown in the differentpanels wereobserved for L cells transfected with:
a and b, normal C3 cDNA;panelsc and d,AGIC3D cDNA;panels
precursor proC3 (38). Fibroblasts and monocytes of this indi- panels
e and f, untransfected L-cells; Panels a, c, and e, x100 magnification;
vidual synthesize a normal amount of precursor C3, however, panels b, d, and f,x1000 magnification. The photomicrographs shown in
only a limited amount issecreted at a slower rate thannormal. panels c andd of the AGI transfectants were obtainedfrom a 1-s expo-
Trypsin cleavage data of the intracellular C3 in the deficient sure. The photomicrographs shown ain, b, e, and fof the untransfected
L-cells and AGI cDNA-transfected L-cells were obtained from a 10-s
fibroblasts indicated an abnormal precursor C3 structure. In exposure.
this study, to examine the primary structure of the C3D pro-
tein, we cloned a C3DcDNA from the deficient fibroblasts, teinase inhibitor (Alp,) deficiency (52), impaired secretion of
analyzed the nucleotide sequence, and examined the secretion the AI light chain(mouse) (531, and othercomplement deficien-
of the C3D molecule using L-cells transfected with the full- cies, including type I1 C2 deficiency (54) and certain Cl-inhib-
length C3D cDNA. The results from these investigations dem- itor deficiencies (55).The structuralmechanism by which these
onstrate that the impairment in C3 secretion is caused by a amino acid changes cause the secretory defects has been deter-
single amino acid substitution (Asp549to Asn) in the P-chain of mined in some of these cases. Group I mutations in Type IIA
the proC3 molecule. Immunofluorescence studies indicate that vWD occur predominately in a domain that is thought to be
the deficient proC3 molecule is retained early in thesecretory highly exposed on the surface of vWD, causing delayed trans-
pathway. Studies are presently underway in our laboratory to port through thesecretory pathway. Disruption of a salt bridge
localize the intracellular compartment in which the mutant caused by a single amino acid substitution in Alp, appears to
proC3 accumulates. alter a highly conserved sequence of the protein leading to
Secretory defects due tosingle amino acid substitutions have increased intracellular degradation. A mutation within a con-
been reported in several protein deficiencies, including Type served region of the variable domain of the murine AI light
IIA von Willebrand disease (Type IIA vWD) (501, defective se- chain blocks secretion of the protein without significantly al-
cretion of high molecular weight kininogen (rat) (51), a-l-pro- tering the normal conformation of the protein.
28498 Human C3 Deficiency
A B

-
5'

" 5
OFG
3
n n n n n
9
KEG AVG
bp - N
c
D
c

A G I Normol
N D N D N D N D N D P C R
-
-*
OFG KEG
5'

AVG
n n n n n
F S F S F S F S F S
870bp-
c

L
F
C

A G I Normol

Lane 1 2 3 4 5 6 7 8 9 1 0 Lane I 2 3 A 5 6 7 8 910

NormaI/Deficient Fast/Slow
FIG.5.Allele-specific PCR reactions that test for the C3D mutation and C3 FastISlow polymorphism.Genomic DNA from the New
Zealand C3-deficient kindred was subjected toallele-specific PCR using pairsof oligonucleotide primers to detect the normal ( N , G"", Asp549)or
mutated ( D ,AI7", AS^'^'') C3 sequences (panelA ) a s well as the Fast(F,G""', Gly'"') or Slow (S, C""", Arg'") associated C3 polymorphisms(panel
B ) . The strategiesfor the PCR reactions are illustratedat thetop of each panel. TheC3D family members are indicateda s follows: OFG, mother
with half-normal levelsof serum C3;KEG, father with half-normal levelsof serum C3; AVG, sister of proband who is deficient in serumC3; AGI,
the proband who is deficient in serumC3. The normal sample isfrom an unrelated Caucasian male with normal levelsof serum C3comprised of
both C3F and C3S allotypes. For experimental details see "Experimental Procedures."

KEG OFG TARLE


I
Phylogenetic comparison of region spanning residue549

Human (IO) 100/100 EWADSVWVD 100/100


(60) Rat 78/86 EVVADSVWVD 1001100
(61)Guinea
pig 78/86 EWADSVWAD 90/100
Mouse
63)(62, 77/85 EWAPSVWVD 100/100
AVG AGI (64)
Cobra 5 1/66 EIVAPSVWVD 90/100
(65) Trout 44/59 DLVADSVWVD 80/100
0C3S, Normal (66)
Hagfish 31/47 ELVADSIIID 60190
(67)
Lamprey 32/48 EIVADSVTVE 70190
C3F, Deficient, mutation unknown
C3F, Deficient, Asp549+A~n549 Similarity based on charge, polarity, and biochemical properties of
amino acids. Sequences compared to the normal human sequence.
Diagram of the C3-deficientNew Zealandkindred sum-
FIG. 6. Residue positions are of the proC3 molecule (10).
marizing the results of the allele-specific PCR study. The open box
indicates a normal C3 gene that contains the C3Slow associated poly-
morphism. The closed box indicates a C3 null gene containing a n un- In addition to theAsp"'" to Asn mutation that is maternally
knownmutationandtheC3Fastassociatedpolymorphism.The inherited, allele-specific PCR results demonstrated thatC3 de-
AS^"^^ point muta-
hatched box indicates a C3 null gene containing the ficiency in this New Zealand family is caused by another dis-
tion and the C3 Fastassociated polymorphism. tinct molecular genetic defect that is paternally inherited. The
paternal molecular genetic defect was not found during the
Although the three-dimensional structure of C3 has notbeen cDNA cloning and sequencing studies. Several reverse tran-
determined, binding studies and epitope mapping of C3 have scription-PCR cDNA clones were sequenced that encoded the
provided insight about the regions of C3 necessary for protein 5'-end of the AGI C3 mRNA, all were normalin sequence. Only
interactions andfunction (56-59). The region of the C3 p-chain one clone (pAGI.ll) wassequenced that encoded the remaining
affected by the amino acid substitution at residue 549 has not 4.0 kb of C3 mRNA, however. This clone contains the
been associated with C3 protein interactions orfunctions. How- maternal mutation. Hence, the paternal mutation is possibly
ever, this residue is identical in several vertebratespecies, and containedwithin the corresponding 4.0 kb ofmRNA tran-
the surrounding sequence also is highly conserved (Table I), scribed from the father'sC3 gene. Since the father's cells prod-
suggesting the importance of this region in thenormal expres- uce half normal levels of C3 protein, but a normal size C3
sion of a functional C3 protein. The substitution of Asn for mRNA in normal quantities (data not shown), the paternal
Asp549may result in the disruptionof a salt bridge as in Alpi defect could possibly be a nonsensemutation that abrogates C3
and/or this region may be on the surface of C3 as is theregion protein synthesis without changing the stability of the C3
mutated in Type I1vWD. Either of these possibilities could mRNA. Of course, it isalso possible that the paternaldefect is
affect interactions of the mutant C3 protein with resident en- not contained in the C3 gene but in some other gene whose
doplasmic reticulum proteins and cause retention in thiscom- product is required for C3 protein synthesis. Future cloning
partment. Additionally, the relationship, if any, between the and sequencing studies usingmRNA isolated from the father's
amino acidsa t residues 102 (FasVSlow polymorphism) and 549 fibroblast cells shouldhelp determine which hypothesis is
remains to be determined. Future studiesof the mechanism of correct.
this block in secretion and of the interactionsbetween distant Sixteen families with hereditary C3 deficiency have been
amino acids in the primary structure will provide new infor- reported. Five of these families, including the one in the pre-
mation regardingthetertiarystructure of C3 as well as sent study, have been examined for molecular genetic defects
the processing and secretory pathway of this complement causing C3 protein deficiency. All five families have been found
component. to contain distinct molecular genetic mutations. In a Japanese
Human C3 Deficiency 28499
family, the defect is a point mutation that does not prevent C3 and Stoop, J. W. (1983)Pediatrics 71, 81-87
29. Borzy, M. S., Gewurz, A,, Wolff, L., Houghton, D., and Lovrien, E.(1988)Am.
secretion, but causes hypercatabolism of serum C3 due to an J . Dis. Child. 142, 79-83
amino acid substitution at one of the cleavage sites of factor I, 30. Grumach, A. S., Vilela, M. M. S., Gonzalez, C. H., Starobinas, N., Pereira,A. B.,
a regulator of C3 activation (35). The other remaining muta- Dias-da-Silva, W., and Carneiro-Sampaio, M.M. S. (1988)Braz. J . Med.
Biol. Res. 21, 247-257
tions include two splice site mutations(31,361 and an 800-base 31. Botto, M., Fong, K. Y., So, A. K., Rudge, A., and Walport, M. J . (1990)J . Clin.
pair partial gene deletion (37). Collectively, all theseinvestiga- Znuest. 86, 1158-1163
tions indicate that inherited human C3 deficiency is caused by 32. Imai, K., Nakajima, K., Eguchi, K., Miyazaki, M., Endoh, M., Tomino, Y.,
Nomoto, Y., Sakai, H., and Hyodo, Y. (1991)Nephron 59, 148-152
an array of molecular genetic mutations and suggests addi- 33. Peleg, D., Harit-Bustan, H., Katz,Y., Peller, S., Schlesinger, M., and Schonfeld,
tional causes of C3 deficiency may be found in other C3-defi- S . (1992)Pediutr. Infect. Dis. J. 11, 401-404
cient individuals. Continued investigation of the molecular ba- 34. Sanal, 0.. Loos, M., Ersoy, F., Kanra, G., Secmeer, G., and Tezcan, I. (1992)Eur.
J . Pedintr. 151,67&679
sis of C3 deficiency will
provide information about the 35. Watanabe, Y., Matsui, N., Yan, K., Nishimukai, H., Tokunage, K., Juji, T.,
biosynthesis, processing, and structure of the C3 protein as Kobayashi, N., and Kohsaka, T. (1993)Mol. Immunol. 30,62
36. Lin, C.-Y., and Huang, J.-L. (1993)Mol. Zmmunol. 30, 29
well as insight into the function of C3 and C3 fragments inthe 37. Botto, M., Fong, K.Y., So, A.K., Barlow, R., Routier, R., Morley, B. J., and
immune response. Walport, M. J. (1992)Proc. Natl. Acad. Sci. U. S. A . 89,4957-4961
38. Katz, Y., Singer, L., Wetsel, R. A., Schlesinger, M., and Fishelson,2. (1994)Eur.
Acknowledgments-We thank the members of Dr. Singer's thesis J. Zmunnol. 24,1517-1522
committee, Drs. Matthew L. Thomas, John P. Atkinson, V. Michael 39. Chirgwin, J. M., Pryzbyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979)
Holers, Arnold W.Straws, Susan Church, and R. Paul Levine for their Biochemistry 18,5294-5299
help and insightful comments. We also thank Drs. Harvey R. Colten and 40. Ausubel, F. M., Brent, R., Kingston,R. E., Moore, D. D., Seidman, J. G., Smith,
J. A,, and Struhl, K. (1993)Current Protocols in Molecular Biology, pp.
Robert C. Strunk for support and critical evaluation of the data, and 4.5.1-4.5.3, Green Publishingkssociates andWiley-Interscience, New York
Joie Haviland's excellent technical assistance in the immunoprecipita- 41. Gubler, U., and Hoffman, B. J. (1983)Gene (Amst. 26, 263-269
tion experiments. 42. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Horn, G. T.,
Mullis, K. B., and Erlich, H. A. (1988)Science 239, 487-491
REFERENCES 43. Tabor, S., and Richardson, C. C. (1987)Proc.Natl.Acad.Sci. U. S . A . 84,
1. Alper, C. A,, Colten, H.R., Rosen, F. S., Rabson, A. R., Macnab, G . M., and Gear, 4767-4771
J. S. S. (1972b)Lancet 2, 1179-1181 44. Sanger, F., Nickles, S., and Coulson,A. R. (1977)Proc. Natl. Acad. Sci.U. S. A .
2. Singer, L., Colten, H. R., and Wetsel, R. A. (1994)Pathobiology 62, 14-28 74,54634467
3. Volanakis, J. E. (1989)Current Topics in Microbiology and Immunology: The 45. Johnson, C. A., Densen, P., Hurford, R. K., Jr., Colten, H. R., and Wetsel, R. A.
Third Component of Complement (Lambris, J. D., ed) pp. 1-21, Springer- (1992)J. Biol. Chem. 267,9347-9353
Verlag, Berlin 46. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G.,Smith,
4. Schifferli, J. A., and Ng, Y. C. (1988)Clin. Immunol. Allergy 2, 319-334 J. A,, and St&, K (1990)Current Protocols in Molecular Biology, pp.
5. Fearon, D.T. (1993)Curr. Opin. Zmmunol. 5,341-348 9.1.1-9.1.9,Green Publishing Associates and Wiley-Interscience, New York
6. van Noesel, C.J. M., Lankester, A. C., and van Lier, R. A.W. (1993)Zmmunol. 47. Harlow, E., and Lane, D. (1988)Antibodies: a Laboratory Manual, pp. 359-
Today 14.8-11 420,Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
7. Bokisch, V. A,, Dierich, M. P., and Muller-Eberhard, H. J. (1975)Proc. Natl. 48. Nichols, R. L., and McComb, D. E. (1962)J . Immunol. 89,545454
Acad. Sci. U. S. A. 72, 1989-1993 49. Botto, M., Lissandrini, D., Sorio, C.. and Walport, M. J. (1992)J. Zmmunol.
8. Tack, B. F., and Prahl, J. W. (1976)Biochemistry 15,4513-4521 149,1348-1355
9. Brade, V., Hall, R. E., and Colten, H. R. (1977)J. Exp. Med. 146, 759-765 50. Lyons, S. E., Bruck, M. E., Bowie, E. J. W., and Ginsberg, D. (1992)J. Biol.
10. deBruijn, M. H. L., and Fey, G. H. (1985)Proc. Natl. Acad. Sci. U. S . A . 82, Chem. 267,4424-4430
708-712 51. Hayashi, I., Hoshiko, S., Makabe, O.,andoh-ishi, S. (1993)J.Biol. Chem.268,
11. Tomana, M., Niemann, M., Garner, C., and Volanakis, J. E. (1985)Mol. 17219-17224
Immunol. 22, 107-111 52. Brodbeck, R. M., Samandari, T., and Brown, J. L. (1993)J. Biol. Chem. 268,
12. Hase, S., Kikuchi, N., Ikenaka, T., and Inoue, K. (1985)J. Biochem. (Tokyo)98, 6771-6776
863-874 53. Dul, J. L., and Argon, Y. (1990)Proc. Natl. Acad. Sci. U. S. A. 87, 8135-8139
13. Fong, K. Y., Botto, M., Walport, M. J.,and So,A. K. (1990)Genomics 7,579586 54. Kulics, J., Kiepiela,P., Johnson, C. A,, Densen,P., Singer, L., Colten, H.R., and
14. Vik, D., Amiguet, P., Moffat, G., Fey, M., Amiguet, F., Wetsel, R., and Tack, B. Wetsel, R. A., (1993)Mol. Zmmunol. 30, Suppl. 1, 26
(1991)Biochemistry 30, 1080-1085 55. Verpy, E.,Eldering, E., Biasotto, M., Couture-Tosi, E., Meo, T., and %si, M
15. Whitehead, A. S., Solomon, E., Chambers, S., Bodmer, W. F., Povey, S., and Fey, (1993)Mol. Immunol. 30, Suppl. 1, 59
G . (1982)Proc. Natl. Acad. Sci. U. S. A. 79, 5021-5025 56. Fishelson, Z. (1991)Mol. Zmmunol. 28, 545-552
16. Alper, C. A., Johnson, A. M., Birtch, A. G., and Moore, F. D. (1969)Science 163, 57. Becherer, J. D., Alsenz, J., and Lambris, J. D.(1989)The Third Component
286-288 of Complement:Chemistryand Biology (Lambris, J. D., ed) pp. 45-72,
17. Colten, H. R. (1993)Progressive Immunology (Gergely, J., ed) Vol. VIII, pp. Springer-Verlag, Berlin
483493,Springer-Verlag, Budapest, Hungary 58. Esparza, I., Becherer, J. D., Alsenz, J., De la Hera, A., Lao, Z., Tsoukas, C. D.,
18. Katz, Y., and Strunk, R. C. (1989)J. Immunol. 142, 3862-3867 and Lambris, J. D. (1991)Eur. J. Zmmunol. 21,2829-2838
19. Alper, C. A., and Propp, R. P. (1968)J. Clin. Inuest. 47,2181-2191 59. Alsenz, J., A d a , D., Huemer, H. P., Esparza, I., Becherer, J. D., Kinoshita, T.,
20. Colten, H. R., and Alper, C. A. (1972)J. Immunol. 108, 1184-1187 Wang, Y., Oppermann, S., and Lambris, J. D. (1992)Deu. Comp. Immunol.
21. Ballow, M., Shira, J. E., Harden, L., Yang, S. Y., and Day, N.K. (1975)J. Clin. 16,63-76
Inmost. 56.
~
", 7w-710 60. Misumi, Y., Sohda, M., and Ikehara, Y. (1990)Nucleic Acids Res. 18, 2178
22. Grace, H. J., Brereton-Stiles, G. G., Vos, G. H., and Schonland,M. (1976)S . Afr. 61. Auerbach, H. S., Burger, R., Dodds,A., and Colten, H. R. (199O)J.Clin. Inuest.
Med. J . 50. 139-140 86,96106
23. Osofsky, S. G.; Thompson, B. H., Lint, T. F., and Gewurz, H. (1977)J . Pediatr. 62. Wetsel, R. A., Lundwall, A., Davidson, F., Gibson, T., Tack, B. F., and Fey, G . H.
90,180-186 (1984)J. Biol. Chem. 259, 13857-13862
24. Davis, A. E., Davis, J. S., Rabson, A. R., Osofsky, S., G., Colten, H. R., Rosen, 63. Lundwall, A., Wetsel, R. A,, Domdey, H., Tack, B. F., and Fey, G. H. (1984)J.
F. S., and Alper, C. A. (1977)Clin. Immunol. Zmmunopathol.8, 543-550 Biol. Chem. 259, 13851-13856
25. Pussell, B. A., Nayef, M., Bourke, E., Morris, S., and Peters, D.K. (1980) 64. Fritzinger, D. C., Petrella, E. C., Connelly, M. B., Bredehorst, R., andVogel, C.
Lancet 1, 675-677 W. (1992)J. Zmmunol. 149, 3554-3562
26. Hsieh, K.-H., Lin, C.-Y., and Lee, T.-C. (1981)Clin. Immunol. Zmmunopathol. 65. Lambris, J . D., Lao, Z., Pang, J., and Alsenz, J. (1993)J. Immunol. 151,
20,305-312 6123-6134
27. Sano, Y., Nishimukai, H., Kitamura, H., Nagaki, K, Inai, S., Hamasaki, Y., 66. Ishiguro,H.,Kobayashi, K., Suzuki, M., Titani, K., Tomonaga, S., and
Maruyama, I., and Igata, A. (1981)Arthritis Rheum. 24, 1255-1260 Kurosawa, Y.(1992)EMBO J. 11, 829-837
28. Roord, J . J., Daha, M., Kuis, W., Verbrugh, H. A., Verhoef, J., Zegers, B. J . M., 67. Nonaka, M., and Takahashi, M. (1992)J. Zmmunol. 148,32903295