| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Pediatrics and
Surgery, and
The Howard Hughes Medical Institute, The Johns Hopkins University School of Medicine, Baltimore, MD 21287;
§
Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892;
¶
Department of Comparative Medicine, Stanford University, CA 94305; and
||
Animal Health Trust, Lanwades Park, Newmarket, Suffolk, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During the course of line breeding a colony of Brittany spaniels with hereditary canine spinal muscular atrophy (HCSMA) (3), a number of the dogs were noted to be unusually susceptible to recurrent bacterial infections. Further investigations revealed that the animals had a genetically determined deficiency of C3 (4). The disorder is inherited as an autosomal recessive trait and segregates independently of the HCSMA (4, 5).
Affected animals have less than 0.0003% of the normal amount of C3 in their serum and markedly reduced levels of serum opsonic and chemotactic activities (6). Like their human counterparts (2, 7, 8), C3-deficient dogs have an increased susceptibility to bacterial infections and develop type 1 membranoproliferative glomerulonephritis (9, 10).
Although previous studies have shown that C3 deficiency in the dog is the consequence of a null allele of the structural gene for C3 (5), the molecular basis of the defect has not been defined. The current studies were performed to characterize the cDNA for wild-type C3 in the dog and identify the mutation responsible for C3 deficiency. In this study, we demonstrate that canine C3 deficiency is the result of a deletion of a cytosine at position 2136, which results in a frameshift at codon 712 and the generation of a premature stop codon 11 amino acids downstream.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals for these studies were obtained from the colony of C3-deficient dogs housed at The Johns Hopkins Medical Institutions (Baltimore, MD). Animals were confirmed as C3 deficient by testing their serum for canine C3 by double immunodiffusion in gels using anti-canine C3 Ab (4). Tissues were obtained at the time of euthanasia and were cut into fragments weighing 10 to 50 g, snap frozen in liquid nitrogen, and stored at -70°C. Serum was obtained from clotted whole blood and was frozen and stored at -70°C.
Preparation of DNA and RNA
DNA was prepared from liver and peripheral blood leukocytes, as previously described (11). Total cellular RNA was obtained from liver, as previously described (11). Poly(A)+ mRNA was extracted from total cellular RNA by oligo(dT) cellulose chromatography (12).
Northern blot analysis
Total cellular RNA or poly(A)+ mRNA was electrophoresed in 1.5% agarose containing 2.2 M formaldehyde and transferred in 10x SSC to a Hybond nylon filter (Amersham, Arlington Heights, IL). The transferred RNA was hybridized to the appropriate cDNA probe in 50% formamide, 5x SSPE (1x = 0.18 M NaCl, 0.01 M Na phosphate, pH 6.8, and 0.001 M EDTA), 4x Denhart’s solution (1x Ficoll, BSA, and polyvinyl pyrrolidone, each at 0.2 mg/ml) containing 2% SDS, and 100 µg/ml of sheared salmon sperm DNA at 42°C. The filters were then washed twice in 1x SSC containing 1% SDS for 30 min at room temperature and once in 0.1x SDS for 30 min at 50°C, and autoradiography was performed.
Southern blot analysis
Genomic DNA was digested with restriction endonucleases, electrophoresed in 1.5% agarose, and transferred in 10x SSC (1x SSC = 150 mM NaCl and 15 mM sodium citrate) to GeneScreenPlus nylon membranes (DuPont NEN, Boston, MA) (12). The transferred DNA was hybridized to the appropriate cDNA probe in a solution of 50% formamide, 10% dextran sulfate, 1 mM NaCl, and 1% SDS. The filters were washed twice in 2x SSC with 1% SDS for 30 min at 65°C and twice in 0.1x SSC for 30 min at room temperature, and autoradiography was performed.
cDNA clones
Two cDNA clones for human C3 (pC3.11 and pC3.49) were obtained from American Type Culture Collection (Rockville, MD) (13). pC3.11 encompasses 4342 bp and represents approximately 90% of the 3' coding region of human C3. pC3.49 encompasses 2.9 kb of the 5' coding region of human C3 and overlaps the sequence of the pC3.11 clone.
A canine oligo(dT)-primed cDNA canine liver library in lamba gt11 (Clontech Laboratories) was screened with pC3.11. A 1.5-kb clone was isolated and subcloned into pGEM-4 before sequencing. The canine C3 cDNA clone (dC3.2) encompasses a region that corresponds to nucleotides 3540 to 5049 and includes 16 bp of the 3' untranslated region and the poly(A)+ tail.
PCR amplification of canine C3 cDNA
Alignment of mammalian C3 cDNA sequences (human (13), guinea pig
(14), and murine (15)) indicated a high degree of sequence
conservation, particularly for regions predicted to be important in its
function (1). In addition, canine-specific sequence at the 5' terminus
was available (see Fig. 3
). PCR primers were designed to the known
canine sequence and to the mammalian consensus sequence. The full
length of the wild-type and mutant canine C3 genes was amplified in
overlapping fragments (see Fig. 3). All PCR amplifications were
performed with a 1:1 combination of Taq polymerase
(Boehringer Mannheim Corp., Indianapolis, IN; 2.5 U/tube) and
Pfu polymerase (Stratagene, La Jolla, CA; 2.5 U/tube) to
minimize Taq polymerase-induced errors. The PCR products
were extracted from gels, purified (Qiagen, Valencia, CA), and cloned
to the TA cloning vector (Invitrogen Corp., San Diego, CA). The
amplification efficiencies varied for each primer pair. Amplification
of the bands required between 35 and 40 cycles on a thermal cycler
(Cetus, Emeryville, CA).
|
Transformation of competent Escherichia coli DH5 
cells was undertaken after a 42°C heat shock for 45 s. White
colonies containing insert were identified in the presence of X-gal
(Sigma Chemical Co., St. Louis, MO). The presence of insert was further
confirmed by PCR analysis before automated sequencing (Applied
Biosystems, Foster, CA). A series of internal primers were designed to
sequence both strands of plasmids (Table I).
|
To confirm the mutation found in the cDNA of C3-deficient dogs, the intron-exon boundaries of the human C3 were compared with the corresponding canine sequence. The region containing the canine C3 mutation corresponds to exon 17 in humans (16). Amplification of canine DNA was undertaken with primers 27 and 28, which were designed to prime within exon 17. Direct sequencing of PCR fragments extracted from gels was undertaken to confirm the presence of the mutation in canine DNA.
Analysis of the inheritance of the mutation was undertaken with
allele-specific oligonucleotides ASO, as described (17). Fourteen-mer
oligonucleotides corresponding to the wild-type (PR33ASO) and the
mutant (PR35ASO) sequences were labeled with 
-32P.
PCR-amplified genomic products were slot blotted onto a filter and
probed with 32P-labeled oligonucleotides (Table I
).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Southern blot analysis was performed with DNA obtained from white blood cells from normal and C3-deficient dogs. DNA digested with a variety of enzymes (TaqI, HindIII, PstI, BamHI) showed the presence of common restriction fragments in both normal and C3-deficient dogs when probed with the cDNA for canine C3 (dC3.2, not shown). Since this probe covers only approximately 30% of the 3' terminus of the full-length cDNA, we also repeated the Southern analysis using two human probes (pC3.11 and pC3.49) that together span the full coding region of human C3 (13). No differences were found between normal and deficient dogs (not shown), indicating that there was no evidence for a gross deletion or rearrangement of the C3 gene in the C3-deficient animals.
Northern blot analysis of liver RNA
Studies were also performed to determine whether C3-deficient dogs
produced C3 mRNA. Liver was selected for study since it is the major
site of C3 synthesis in man (18). When liver RNA was examined by
Northern blot analysis using the probe for canine C3 (dC3.2),
C3-deficient dogs had markedly reduced amounts of C3 mRNA (Fig. 1). In contrast, the content of ß-actin
mRNA was equivalent in RNA isolated from the livers of C3-deficient and
normal dogs. The C3 mRNA from C3-deficient dogs appeared identical in
size to C3 mRNA from normal dogs (
5.2 kb). Semiquantitative
slot-blot analysis demonstrated that the livers of C3-deficient dogs
had approximately 10% of the normal amount of C3 mRNA (Fig. 2). Northern blot analysis of
poly(A)+ mRNA confirmed the presence of reduced
amounts of a normal sized mRNA in the livers of C3-deficient dogs (not
shown).
|
|
The full-length canine cDNA was amplified in overlapping fragments
with primers designed to the consensus sequence of mammalian C3 cDNA
and the canine-specific cDNA sequences (Fig. 3). Multiple products (>5 each) of
independent reverse-transcriptase PCR amplifications of both wild-type
and mutant canine C3 RNA were performed in both directions to confirm
each sequence. The PCR fragments corresponded to the size of human C3
cDNA, and the full-length cDNA is consistent with the 5.2-kb C3
transcript seen on Northern blot analysis. No size differences were
noted between the wild-type and mutant products.
The full-length canine C3 cDNA sequence was assembled with the
assistance of the overlapping regions of the PCR products (Fig. 3). The
composite sequence of both wild-type and mutant cDNAs was aligned, and
any sequence discrepancies were further investigated by sequencing
additional clones or by direct sequencing of PCR products. The
nucleotide sequence was numbered with the A of the initiation
methionine codon as +1. Three synonymous mutations were found at
positions 1590, 1727, and 2919. The wild-type sequence has been
submitted to GenBank (accession bankit #27404).
Identification of conserved domains in the canine C3 gene
Alignment of canine C3 cDNA with known mammalian C3 cDNA sequences
shows a high degree of conservation of regions thought to be critical
for post-translational modification and for C3 function (Fig. 4). The canine C3 gene is predicted to
produce a pre-pro-C3 protein of 1663 amino acids in length, which is
identical to that of man. However, in contrast to the 22-amino-acid
signal peptide in human C3, canine C3 has a 23-amino-acid signal
peptide (Fig. 4). As seen in Figure 4, the amino acid sequence in the
vicinity of the thioester bond is highly conserved in all of the
mammalian species characterized to date. The tetra-arginine linker,
which is cleaved to produce the ß- and -chains, is also highly
conserved. The C3a cleavage site and the arginine required for
mediating C3a activity are also seen in the canine C3 sequence
(Fig. 4).
|
Identification of the mutation responsible for canine C3 deficiency
The cytosine normally present at position 2136 is deleted in
C3-deficient animals (Fig. 5). This
deletion was present in four separate clones and also by direct
sequencing of PCR products. The deletion results in a frameshift
followed 10 codons downstream by a stop codon truncating the protein at
amino acid 721. The mutant cDNA is expected to transcribe the ß-chain
and a small fragment of the -chain. The mutation was seen in all six
C3-deficient animals examined, but not in any of four normal animals.
Direct PCR sequencing of cDNA from two heterozygotes showed the
expected double sequence at the point of the single base deletion.
|
To confirm the single base pair deletion, the intron-exon boundaries of the human C3 gene were aligned with the corresponding canine sequence. The region containing the mutation corresponds to exon 17 in humans. Amplification of this region with primers 27 and 28 produced a single band of equal size from either DNA or cDNA, indicating that the fragment was amplified from within a single exon (not shown). Direct sequencing of these bands in both directions showed the presence of the C2136 deletion in the C3-deficient animals.
Segregation analysis was undertaken with allele-specific
oligonucleotides (Fig. 6) (17).
Comparison of genotypic assignment based on C3 levels and ASO analysis
of the C2136 deletion is shown in Figure 6. There is perfect agreement
between genotype and C3 phenotype. They indicate that the wild-type
probe (PR33ASO) binds only to genomic DNA of wild-type animals.
Similarly, the mutant probe (PR35ASO) binds only to DNA of mutant
animals. Heterozygous animals give bands of intermediate intensity with
either the wild-type or mutant probes consistent with the presence of
two alleles.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The current studies were performed to determine the molecular basis for the C3 deficiency. Initial Northern blot analysis demonstrated that the C3-deficient dogs have markedly reduced (10% of normal) levels of a normally sized C3 mRNA. Southern blot analysis did not reveal evidence of a major rearrangement of the C3 structural region.
Amplification of the full-length canine C3 cDNA was possible because of the highly conserved nature of the gene. Previous studies have established that mammalian complement proteins are frequently able to interact functionally with complement components from other species. This is characteristic of highly conserved systems that play a vital role in homeostasis. Analysis of the predicted amino acid sequence of the canine C3 gene has confirmed that domains responsible for critical functions show a high degree of identity with human, murine, and guinea pig C3. The canine C3 thioester domain, for example, has 100% sequence identity with that of related species. Minor interspecies differences are, however, present. For example, canine C3 would appear to lack a third factor I cleavage site at position 955 when compared with human C3, but not when compared with mouse or guinea pig C3.
Sequence analysis showed that the cytosine at position 2136 is deleted,
resulting in a frameshift and premature termination 11 codons
downstream. This deletion was not present in normal dogs. The presence
of reduced amounts of a normal sized mRNA in the liver of the deficient
dogs is consistent with the fact that mutations that result in
premature termination of translation are associated with marked
reductions in steady state levels of mRNA (21). The mutant allele would
be expected to generate a protein containing the ß-chain and a small
part of -chain. Analysis of the sequences in the vicinity of the
deletion does not demonstrate any of the elements known to be
associated with single base deletions (22).
C3 deficiency has been identified previously in one strain of guinea pigs and in one strain of rabbits. A strain of guinea pigs was identified in the 1980s with <5% serum C3 activity (23). Extensive molecular analysis of the gene has failed to identify a mutation within the coding region (14). It is currently thought that these animals may have a specific defect of the enzyme(s) responsible for generating the thioester bond. A strain of rabbits with 10% of normal serum C3 activity has also been identified (24). The nature of the defect in these animals has not been identified. Recently, C3-deficient mice have been generated by gene targeting. These mice lack any detectable C3 in their serum and are susceptible to recurrent bacterial infections (25).
C3 deficiency has been identified in 20 humans from 15 kindreds (2, 7, 8). The responsible mutations have been identified in four patients.
Two patients have had splicing defects (26, 27). A G-A substitution of
the 5' splice donor site of intron 18 was found in the first patient,
which resulted in 61-bp deletion of exon 18 (26). The second patient
from Taiwan had G-T substitution of the splice donor site of intron 10.
This resulted in transcripts that skipped exon 10 and were predicted to
cause premature truncation of the C3 protein as a result of a
frameshift (27). A third patient from South Africa was shown to have an
800-bp deletion leading to loss of exons 22 and 23 of the -chain
(28). Premature truncation of the protein is again predicted to occur
as a result of a frameshift. The fourth patient characterized to date
has a critical amino acid substitution resulting in a secretory defect
(29). In this patient, an Asp549Asn substitution is thought to
interfere with secretion of C3.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. J. A. Winkelstein, CMSC 1103, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Received for publication August 18, 1997. Accepted for publication November 24, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. S. Swanson, L. B. Schook, and G. C. Fahey Jr. Nutritional Genomics: Implications for Companion Animals J. Nutr., October 1, 2003; 133(10): 3033 - 3040. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
