Genetic variation in the CRP promoter: association
with systemic lupus erythematosus
Jeffrey C. Edberg1,?, Jianming Wu1, Carl D. Langefeld4, Elizabeth E. Brown2, Miranda C.
Marion4, Gerald McGwin Jr3, Michelle Petri5, Rosalind Ramsey-Goldman6, John D. Reveille7,
Summer G. Frank8, Kenneth M. Kaufman8,9,10, John B. Harley8,9,10, Graciela S. Alarco ´n1
and Robert P. Kimberly1
1Division of Clinical Immunology and Rheumatology, Department of Medicine,2Department of Epidemiology and
3Section of Trauma, Burns, and Critical Care, Department of Surgery, University of Alabama at Birmingham,
Birmingham, AL, USA,4Section on Statistical Genetics and Bioinformatics, Division of Public Health Sciences,
Department of Biostatistical Sciences, Wake Forest University, Winston-Salem, NC, USA,5Division of Rheumatology,
Johns Hopkins University School of Medicine, Baltimore, MD, USA,6Division of Rheumatology, Northwestern
University Feinberg School of Medicine, Chicago, IL, USA,7Division of Rheumatology, Department of Medicine,
University of Texas Health Science Center at Houston, Houston, TX, USA,8Oklahoma Medical Research Foundation,
Oklahoma City, OK, USA,9US Department of Veteran Affairs, Oklahoma City, OK, USA and10Department
of Medicine, University of Oklahoma, Oklahoma City, OK, USA
Received October 9, 2007; Revised December 22, 2007; Accepted January 4, 2008
The pentraxin C-reactive protein (CRP), an innate immune system opsonin which binds nuclear debris and
apoptotic bodies, may protect against autoimmunity. A relative deficiency of CRP levels in patients with
systemic lupus erythematosus (SLE) might contribute to altered handling of self-antigens. We report that
the proximal 50promoter region of CRP contains several polymorphisms that exhibit association with SLE in
multiple populations. Strongest association was observed at the proximal promoter single nucleotide poly-
morphism (SNP) rs3093061 (CRP-707) (P 5 6.41 3 1027and P 5 2.13 3 1026in African-American and
Caucasian case–control samples respectively). This association remains after adjustment for admixture.
containing rs3091244/rs3093062 (CRP-409/-390) appear to be driven by the rs3093061 (CRP-707) association.
These data demonstrate that rs3093061 at the -707 site within the CRP gene is an SLE susceptibility locus.
Systemic lupus erythematosus (SLE) is a systemic autoimmune
disease characterized by the presence of autoantibodies pri-
marily directed against nuclear autoantigens as well as
cytoplasmic and circulating proteins (1,2). These autoanti-
bodies, together with alterations in cells of both the acquired
and innate immune system, lead to the development of a
chronic and severe clinical phenotype that can affect
many different organs and tissues including kidneys, skin
and brain. The prevalence of SLE in the general population
is on the order of 1:2000 individuals, but varies by gender, eth-
nicity, socioeconomic status, genetic and environmental back-
A significant amount of data support a role for host genetics
as both predisposing factors and as important determinants of
patient outcomes (4–13). SLE shows strong familial aggrega-
pensity for clustering of other autoimmune phenotypes (11). In
?To whom correspondence should be addressed at: Division of Clinical Immunology and Rheumatology, Department of Medicine, University of
Alabama at Birmingham, SHEL207, 1825 University Blvd, Birmingham, AL 35294-2182, USA. Tel: þ1 2059340894; Fax: þ1 2059966734;
# 2008 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/
licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is
Human Molecular Genetics, 2008, Vol. 17, No. 8
Advance Access published on January 8, 2008
by guest on June 12, 2013
for multiple genetic factors influencing disease susceptibility
ponents and TNFSF4 (4–9,13). Likewise, both linkage and
association studies have shown the importance of genetic
factors in disease outcomes. For example, linkage at 5q14.3–
15 in multiplex SLE pedigrees with autoimmune thyroid
disease has been observed (12). Similarly, we have recently
shown a strong association of a polymorphism in FCGR3A
(CD16A) with development of end-stage renal disease
(ESRD) in patients with SLE (10).
The pathogenesis of SLE is typically characterized by
autoantibody-mediated tissue damage. Work in both humans
and in model systems has suggested that autoantigen proces-
sing and presentation is important in the development of auto-
antibodies and systemic autoimmune disease (14–17). In
particular, the handling of apoptotic cell debris is thought to
be important in the promotion and development of autoanti-
bodies in patients with SLE.
C-reactive protein (CRP), a pentraxin, is an important innate
immune modulator that facilitates the clearance and handling
of cellular debris and apoptotic bodies (18,19). CRP is an
acute phase protein whose levels rise with inflammatory
responses. The functions of CRP are not necessarily limited
to the opsonization of cell debris and apoptotic bodies. CRP
can also activate complement and it can bind to certain Fc
receptors (20). Thus, CRP can have important immunoregula-
tory functions and has been shown to have protective effects in
mouse models of autoimmune disease. Interestingly, some
studies have suggested that SLE is characterized by lower
CRP levels than would be predicted (21,22). Heritability esti-
mates suggest that as much as 60% of the variance in base line
CRP levels is attributable to genetic variation (23–25). Recent
studies by us and others have shown that CRP levels are influ-
enced by genetic variation in the CRP promoter (24–30).
A functionally important tri-allelic single nucleotide poly-
morphism (SNP) (rs3091244) at position -390 relative to the
start codon, a second SNP (rs3093062) at position -409 and
an intronic microsatellite are associated with CRP levels (29).
In SLE, two studies have found evidence for an association
between CRP variants and SLE or SLE nephritis in Caucasians
notype has no known functional role in CRP production (basal
or induced) and it remains unknown which if any promoter
SNPs associate with disease. The tri-allelic rs3091244 variant
was the only promoter SNP examined and was found not to
associate with disease (31). Thus, it remains undetermined if
there are other proximal CRP promoter variants that associate
with SLE and it remains unknown what role genetic variants
in the CRP locus have in SLE in African-Americans. In this
study, we show in two independent African-American popu-
lations that a CRP promoter variant rs3093061 at position
-707, or apromoter SNP haplotype containing this SNP, associ-
a Caucasian population with SLE.
We initially determined the extent of genetic variation in the
1kb proximal promoter of the CRP locus through a
combination of direct sequencing of genomic DNA from 50
Caucasian and 50 African-American healthy control donors
and through assessment of prior re-sequencing data (23,25).
A total of six promoter SNPs were identified (Table 1). We
also included a coding region synonymous SNP and an SNP
in the 30-untranslated region (30-UTR) that have been pre-
viously examined by Russell et al. (31) in patients with
SLE. For analysis of genetic association with SLE, patients
were derived from the case–control study (CASSLE) consti-
tuted at the University of Alabama at Birmingham (UAB)
that comprises Caucasians and African-Americans, healthy
control participants and the African-American families
enrolled in the Lupus Multiplex Research Repository
(LMRR) based at Oklahoma Medical Research Foundation
(12,33). Clinical characteristics of patients have been pub-
lished (3,10,12,33,34). Baseline demographic features for the
CASSLE study participants are summarized in Table 2.
All markers were initially genotyped in the CASSLE case–
control study. Because of the differences in the minor allele
frequency (MAF) of the SNPs among the African-American
and Caucasian controls, these groups were analyzed separ-
ately. Summary statistics are shown in Table 3. All markers
were in Hardy–Weinberg equilibrium except for slight devi-
ations at the uncommon -861 and -390 among Caucasians
(P ¼ 0.01). The CRP-860 SNP (rs3093060) was invariant in
the Caucasian population and was not included in further
analysis of this group. Patterns of linkage disequilibrium
(LD) between the SNPs in the controls for each ethnic group
were determined (Fig. 1). As is evident, there are significantly
different LD patterns between the two ethnic groups. In par-
ticular, the LD pattern of -707 was distinct with high LD
across the locus among African-Americans that was largely
absent in the Caucasians. Of note is the high LD between
the functionally important SNP -390 and -409 with all
markers except -821 in Caucasians. Despite the strong LD
between these SNPs, the predictive power between SNPs
was low except for -707/-490 in African-Americans (r2¼
0.71). For initial analysis of the -390 tri-allelic variant
(rs3091244), we collapsed the T and A alleles because of
their similar functional properties (29).
Analysis of genotypic distribution in the African-American
participants in the CASSLE study showed strong association
of the CRP-707 SNP (rs3093061) with the SLE phenotype
(P ¼ 3.32 ? 1026, Table 3). Indeed, in this population, the
level of significance was greater under a dominant model
of inheritance (P ¼ 6.41 ? 1027) where CRP-707GG homo-
zygotes had an approximately 2-fold reduction in risk of
SLE compared with wild-type homozygotes [odds ratio
(OR) ¼ 0.49, 95% confidence interval (CI) 0.37–0.65]. In
addition, suggestive association was observed at CRP-409,
CRP-390 and CRP4 among African-American enrolled in
CASSLE. The CRP-409, CRP-390 and CRP4 variant allele
associations were slightly enhanced under the additive
model (Table 3). However, because of the higher LD
between CRP-707 and these other CRP SNPs, it is highly
plausible that these weaker associations were due to the
strong association at the CRP-707 site with SLE suscepti-
bility. Because of the multi-center enrollment of the patients
and controls, we tested for center effects but none were found
(data not shown).
1148Human Molecular Genetics, 2008, Vol. 17, No. 8
by guest on June 12, 2013
We replicated the association between the CRP-707 variant
allele (rs3093061) and SLE among African-Americans in Cau-
casians with SLE in the same CASSLE population (P ¼
1.25 ? 1025, Table 3). The association with CRP-707GG
among Caucasians showed a similar protective effect (OR ¼
0.37, 95% C.I. 0.24–0.56; P ¼ 2.13 ? 1026under the domi-
nant model). Suggestive association at CRP2 (rs1800947)
with the variant allele and SLE was also observed. This
effect is unlikely to be due to the association of variant
alleles at CRP-707 and CRP2 because of the low LD
between these two sites in the Caucasian population.
Bonferroni multiple comparisons adjustment for the number
of genetic markers does not dampen the evidence for -707
P-value ¼ 0.0320). We also considered the issue of population
stratification in our cases and controls. A total of 909 Cauca-
sian samples in this study were also part of the recent Lupus
Large Association Study (LLAS), a large case–control repli-
cation sample for our genome wide association study in SLE
P-value ¼ 8.75 ? 1025;
P-value ¼ 2.66?1025).
Table 1. C-reactive protein (CRP) variants
CRP locus SNP SNP identifier SNP location Position (gb18)Allele 1 (major)Allele 2 (minor) Allele 3 (minor)
Synonymous, codon 144
Table 2. Demographics of the CASSLE case–control study
Number of subjects
Age at enrollment
Table 3. Single site genetic association of C-reactive protein (CRP) single nucleotide polymorphisms (SNPs) with systemic lupus erythematosus (SLE) in the
CASSLE case–control study
SNP MAF casea
Genotype test Model P-valueModelOR [95% CI]
3.32 3 1026
6.41 3 1027
0.90 [0.69, 1.17]
2.91 [0.94, 8.99]
1.23 [0.91, 1.66]
0.49 [0.37, 0.65]
0.69 [0.52, 0.92]
1.43 [1.07, 1.90]
1.09 [0.50, 2.36]
1.42 [1.08, 1.88]
1.25 3 1025
2.13 3 1026
Additive0.95 [0.68, 1.33]
1.18 [0.94, 1.48]
0.37 [0.24, 0.56]
1.69 [0.49, 5.8]
0.99 [0.78, 1.26]
0.69 [0.50, 0.97]
1.00 [0.79, 1.25]
aMAF case: n ¼ 491 (African-American); n ¼ 627 (Caucasian).
bMAF control: n ¼ 428 (African-American); n ¼ 604 (Caucasian).
Human Molecular Genetics, 2008, Vol. 17, No. 81149
by guest on June 12, 2013
(see Materials and Methods) (35). As part of the LLAS a prin-
cipal component analysis on over 8000 SNPs were computed
to allow adjustment for potential admixture in tests of associ-
ation. In the subset of 909 Caucasians individuals that over-
lapped thesetwo studies
association tests with the principal component loadings as cov-
ariates to adjust for potential admixture. The significant -707
association remained [P ¼ 1.17 ? 1026, OR ¼ 0.30 (0.19–
0.49)]. Details of these analyses are reported in Supplementary
Material, Table S1.
We performed a logistic regression analysis to determine if
the CRP-707 SNP (rs3093061) was sufficient to explain all of
the association with SLE. Under a dominant model, we found
that the CRP-409 SNP (rs3093062) (presence/absence of the A
allele) was associated with SLE after adjusting for the -707
SNP (presence/absence of the G allele) in both the
(African-Americans, P ¼ 0.024, OR ¼ 1.91, 95% CI 1.09–
3.36; Caucasians, P ¼ 0.030, OR ¼ 5.04, 95% CI 1.18–
21.60). We note that the MAF for CRP-409 in Caucasians is
,1% resulting in an unreliably high OR in that group.
To provide further support for the association of the CRP
promoter with SLE, we genotyped the CRP promoter using
African-American multiplex SLE families in LMRR (12,33).
Pedigree disequilibrium test (PDT) analysis of the CRP-707
variant alone, while not reaching statistical significance, did
trend towards under transmission of the minor allele consistent
with the under representation of this allele in the case–control
study. However, in this independent family-based study,
(global P ¼ 0.046) (Table 4). Thus, we have provided evi-
dence of a genetic association between the CRP-707 promoter
SNP (rs3093061) and SLE in three independent studies utiliz-
ing both case–control and family-based approaches.
Promoter SNPs would not be expected to be functional in
isolation of other nearby SNPs, especially under situations in
which high LD is observed between closely arrayed SNPs.
Accordingly, haplotypes were determined and analyzed for
association with SLE in the CASSLE case–control study.
Using rolling 2- and 3-marker haplotypes, strong association
with SLE was observed around the CRP-707 site (Table 5).
Figure 1. Linkage disequilibrium (LD) patterns in the C-reactive protein (CRP) locus for the single nucleotide polymorphisms (SNPs) examined in this study.
Measurements of LD were calculated by the software Dprime. The statistics D0(top right triangle) and r2(lower left triangle) are shown for African-American
controls (A) and Caucasian controls (B). Dark shading: D0. 0.75/r2. 0.75, intermediate shading: 0.50,D0?0.75/0.502,r2,0.752, light shading: 0.25 , D0?
0.50/0.252, r2, 0.502.
Table 4. Replication of association of CRP(C-reactive protein)-707 haplotype
with systemic lupus erythematosus (SLE) in a family-based African-American
aData are derived from analysis of 830 individuals in 100 multiplex SLE
1150 Human Molecular Genetics, 2008, Vol. 17, No. 8
by guest on June 12, 2013
All 2- and 3-marker haplotypes containing the CRP-707G
allele in separate evaluations of the African-American and
the Caucasian populations demonstrated significant associ-
ation with levels of significance ranging from P ¼ 3.3 ?
1025to P ¼ 3.9 ? 1027. However, they do not provide
strong additional evidence of association beyond the individ-
ual -707 association.
To identify protective and risk haplotypes, we restricted our
analysis to the CRP-707 site together with the known function
sites, CRP-409/-390 (Table 6). Our prior work has shown that
the 11 and 22 -409/-390 haplotypes result in lower constitutive
and stimulated CRP levels than the 12 and 13 haplotypes in
healthy control donors and reporter construct studies respecti-
vely (29). Among African-Americans the -707/-409/-390
haplotype 111 was associated with increased risk of SLE
(OR ¼ 1.47, 95% CI: 1.22–1.77). Haplotype 122 was observed
in ,1% of cases and controls. In contrast, variation at the -707
site to the less common allele changed the risk haplotypes to
African-Americans. A further protective haplotype, 213 but
not 113, was also observed further extending the role of the
-707 site in generating protective CRP promoter haplotypes.
In the Caucasians, similar trends were found. The 111 haplo-
type trended towards increased risk but did not reach statistical
significance (OR ¼ 1.15, 95% CI 0.98–1.35). The 122 haplo-
type was not observed among Caucasians. The minor allele of
CRP-707 again resulted in haplotypes (122 and 212) associated
containing haplotypes were observed in ,1% of the cases and
controls limiting our ability to define protective haplotypes.
Analysis of extended promoter haplotypes, including all
analyzed promoter variants, was performed (Table 7). While
African-American group, definition of at-risk and protective
haplotypes generally aligned with the common and minor
alleles of CRP-707 respectively. The most frequent CRP-707
common allele containing haplotype (111111, Table 7)
defined an at-risk haplotype (P ¼ 0.039, OR ¼ 1.36, 95%
C.I. 1.11–1.68) whereasthe
minor allele haplotype(111222)
haplotype (P ¼ 0.073, OR ¼ 0.73, 95% CI 0.58–0.93) for
and 222) amongthe
were definedin the
Analysis of the extended promoter haplotypes in CASSLE
among Caucasians revealed a potential role for CRP-821 in
defining both at-risk and protective haplotypes (Table 7). An
at-risk haplotype containing the minor allele of CRP-821
together with the CRP-707/-409/-390 111 haplotype was
associated with increased risk of SLE (OR ¼ 1.29, 95% CI:
1.07–1.54). In contrast, the same minor allele of CRP-821
together with the CRP-707/-409/-390 211 haplotype, while
rare, resulted in decreased risk of SLE (OR ¼ 0.15, 95% CI:
0.06–0.41). Thus, the CRP promoter in Caucasians appears
to have additional influences beyond the putative core
CRP-707/-409/-390 SNPs that appear most important for
SLE susceptibility among African-Americans.
Given SLE susceptibility, we examined the influence of
common variation in CRP with clinical manifestations associ-
ated with SLE progression [renal involvement, cardiovascular
disease (CVD), stroke, hypertension and diabetes] in the
subset of patients enrolled in the longitudinal portion of
CASSLE (PROFILE). At the single locus marker level, the
tri-allelic variant at position -390TA (alleles 2 and 3) was
associated with a 1.3 to 2.3-fold decreased risk of CVD,
stroke, hypertension and diabetes among African-American
SLE patients compared with the -390C allele following addi-
tive or dominant models (data not shown). A similar trend
was observed among Caucasians, albeit not significantly.
However, at the individual level, putative CRP-390TA-
represented among Caucasian SLE patients with hypertension
compared with 390C/409G/707A carriers (OR ¼ 0.66, 95%
CI 0.49–0.89) suggesting that the -390TA variant may be
more important for disease progression, as determined by
these indices,thanfor SLEsusceptibility. Nosignificantassoci-
ations were observed among SLE patients with renal involve-
ment (data not shown).
This is the first study to comprehensively analyze CRP prox-
imal promoter SNPs in relation to the SLE phenotype and
we present evidence for a role of CRP promoter genetic var-
iants in susceptibility to SLE. In multiple independent study
populations, including both African-Americans and Cauca-
sians, the variation in the CRP promoter at CRP-707
(rs3093061), and haplotypes containing CRP-707, strongly
and reproducibly associates with the SLE phenotype. Principal
component analysis indicates that this association is not due to
population stratification. In our African-American population,
we do observe association between CRP4 and SLE and a sug-
gestive association between CRP2 and SLE in the Caucasian
population was observed. We note the high degree of LD
between the promoter SNPs and CRP2 and CRP4 (Fig. 1)
which may explain prior reported association between CRP4
and SLE and of CRP2 and CRP levels in which only 1 promo-
ter SNP was analyzed (31).
Our results provide a biological link between SLE and
genetic variants in the CRP promoter. Our prior work, and
the work of others, has provided evidence of association
between the CRP-409/-390 variants which alter functional
E-box transcription factor binding sites (29) and CRP levels
Table 5. Haplotypic genetic association of C-reactive protein (CRP) single
nucleotide polymorphisms (SNPs) with systemic lupus erythematosus (SLE)
in the CASSLE case–control study
2.15 3 1025
6.51 3 1026
3.31 3 1025
9.28 3 1026
4.44 3 1026
2.16 3 1026
3.87 3 1027
1.82 3 1026
5.35 3 1026
aThe P-value corresponds to the haplotype with the listed SNP as the first
SNP in the haplotype. In the Caucasian group, -860 did not show variation
and was not included in the analyses.
Human Molecular Genetics, 2008, Vol. 17, No. 81151
by guest on June 12, 2013
in healthy individuals (24–29). Additional sites in the promo-
ter predicted to be functional are the -821 site that alters a glu-
cocorticoid receptor GR transcription factor and the -861 site
which alters another E-box site. The most strongly associating
SNP, CRP-707, is not predicted to directly lie in a transcrip-
tion factor binding site (TRANSFAC 7.0 Web Site, http://
www.gene-regulation.com/pub/databases.html). However, it
is important to note that while the regression analysis suggests
a modest independent genetic effect at CRP-409, our haplo-
type analysis did not identify any specific haplotypes that
associated with disease beyond the effects seen at CRP-707
The identification of LD blocks in the CRP promoter con-
taining known functionally important transcription factor
binding sites with SLE suggests a role for genetic variation
in the promoter in SLE and in CRP transcriptional efficiency.
While we have examined all SNPs within 1kb of the transcrip-
tional start site, it is likely that there are additional variants
that will be important determinants of both basal and
inducedCRP levels. For
30-untranslated region of the CRP gene and could alter
message stability (31). CRP2, while a synonymous SNP,
could also alter translational efficiency and protein folding
as recently reported in MDR (36). Thus, there are multiple
levels at which CRP levels can be affected by host genetics.
example,CRP4 is inthe
Numerous studies have identified variation in CRP levels
and CPR genetic variants as powerful and specific predictor
of CVD risk in healthy individuals (26,27,30). We were able
to demonstrate that alleles at the functionally important
CRP-390 site were associated with an increased risk of CVD
in African-Americans and hypertension in Caucasians. While
these results are consistent with the established associations
between CRP and CVD, further work in phenotypically
characterized populations of patients with SLE will be necess-
ary to fully explore the importance of CRP variants in CVD in
patients with SLE as well as other clinical manifestations of
SLE. Our inability to replicate the association between CRP4
and SLE nephritis (32) highlights the difficulty and importance
of replication of genetic findings in multiple studies.
Our studies provide support and rationale for more detailed
functional characterization of the CRP promoter region. The
influence of multiple promoter SNPs, based on our haplotype
determinations, should be tested for differences in promoter
determined lower levels of CRP could allow for alterations
in the presentation of autoantigens to the immune system
resulting in the generation of autoantibody production and
autoimmunity. Indeed, either infusion of CRP or introduction
of a CRP transgene into mice with experimental autoimmune
disease results in amelioration of disease (37–40). However, it
Table 6. Analysis of CRP Promoter haplotypes in the CASSLE case–control study
P-value OR [95% CI] Freq-control
aMarkers included in the haplotype were -707/-409/-390. At each site, the common allele is coded as 1 and the minor allele is coded as 2. For CRP-390,
the two minor alleles were coded 2 and 3.
bIn African-Americans, haplotype 122 is not shown because the observed frequency was less than 1% in both cases and controls. This haplotype was not observed
Table 7. Analysis of all C-reactive protein (CRP) promoter haplotypes in the CASSLE case–control study
P-valueOR [95% CI] Freq-control
P-valueOR [95% CI]
1.36 [1.11, 1.68]
0.87 [0.65, 1.17]
1.31 [1.01, 1.69]
0.94 [0.76, 1.17]
0.63 [0.32, 1.26]
1.57 [0.41, 6.09]
0.73 [0.58, 0.93]
0.21 [0.09, 0.50]
0.94 [0.79, 1.11]
1.08 [0.91, 1.29]
1.29 [1.07, 1.54]
0.94 [0.68, 1.29]
0.87 [0.40, 1.87]
0.29 [0.14, 0.59]
1.85 [0.49, 7.02]
0.15 [0.06, 0.41]
0.37 [0.12, 1.15]
aMarkers included in the haplotype were -861/-860/-821/-707/-409/-390. At each site, the common allele is coded as 1 and the minor allele is coded as
2. For CRP-390, the two minor alleles were coded 2 and 3. In the Caucasian study population, the -860 site did not show variation and is coded as 1.
bIn African-Americans, haplotypes not shown included 111113, 111122, 111213, 221112, 221113, 221212 and 221213 with observed frequencies less than 1% in
both cases and controls. These haplotypes were not observed in Caucasians.
cIn Caucasians, haplotypes not shown included 112112, 112222 and 212113 with observed frequencies less than 1% in both cases and controls. These haplotypes
were not observed in African-Americans.
1152 Human Molecular Genetics, 2008, Vol. 17, No. 8
by guest on June 12, 2013
is important to emphasize that analysis to date of proximal
promoter SNPs and extended CRP SNP haplotypes has been
shown to explain ,5% of the total variance for CRP
expression (24). Clearly, with the high heritability of basal
CRP expression, additional genetic factors involved in regulat-
ing CRP expression are yet to be described. These factors
could include additional CRP variants and variants in other
genes such as the transcriptional regulators IL6 and IL1B
that are associated with CRP levels (24,41,42).
The CRP gene is encoded on human chromosome 1q21-23,
a region that has shown multiple linkage and association
signals with the SLE phenotype (33,35). Further genetic
characterization of this region, including the known CRP
receptors FCGR1 and FCGR2, will likely yield additional
insights into the genetic component of SLE susceptibility.
MATERIALS AND METHODS
Patients with SLE were derived from the case–control study
(CASSLE) constituted at the UAB that is comprised of
follow-up data available on a subset of participants [the longi-
tudinal PROFILE study (10,34)] and healthy control partici-
pants. Patients and controls were from UAB, Johns Hopkins
University, Northwestern University and University of Texas
Health Science Center. The second study population included
830 individuals in 100 multiplex SLE African-American
families enrolled in the LMRR based at Oklahoma Medical
Research Foundation (12,33). All participants provided
written informed consent and these studies were approved
by the Institutional Review Board of each participating insti-
tution. Clinical characteristics of patients in the PROFILE
and LMRR studies have been published (3,10,12,33,34). Base-
line demographic features for the CASSLE study participants
are summarized in Table 2. All patients enrolled into the
CASSLE case–control study were seen by a study physician
and had the diagnosis of SLE based on the presence of at
least four ACR criteria (43,44). Ethnicity was self-reported
and verified by the ethnicity of the participants’ four grandpar-
ents, when known. All participants had the cumulative pre-
sence/absence of the revised ACR criteria documented at the
time of enrollment. Characteristics of the LMRR participants
have been previously published (12,33).
Genotypingof the CRP variants was performed byPyrosequen-
cing. Using a nested polymerase chain reaction (PCR) strategy,
we initially amplified two regions of the CRP gene. One reac-
tion encompassed the proximal promoter and the second
spanned from the CRP coding region to the 30-UTR. Second
round PCR reactions were then performed around groups of
closely spaced SNPs (Table 8). Pyrosequencing primers used
for each SNP (or group of SNPs) are also shown in Table 8.
First round PCR reactions contained 25–50 ng of template
DNA, 1.5U Taq polymerase, 0.10 mM of each primer,
0.2 mM dNTP, 1.5 mM of MgCl2and 20 mM of Tris–HCl
(pH 8.4)and 50 mMKCl ina25 ml volume.Cycling conditions
were 5 min at 958C, 35 cycles of denaturation at 948C for 30 s,
annealing at 568C for 30 s, and extension at 728C for 45 s, and
then a final extension step at 728C for 7 min. Second round
PCR conditions were identical except 0.25ml of first round
Table 8. PCR and Pyrosequencing primers used in the analysis of the CRP variants
aAll primers are shown 5’ – 3’ with the forward primer shown first followed by the reverse primer. Biotinylated primers in the second round reactions are
Human Molecular Genetics, 2008, Vol. 17, No. 81153
by guest on June 12, 2013
PCR product was used as template and 40 cycles of amplifica-
tion were performed. All PCR reactions were run in ABI9700
PCR machines (Applied Biosystems). Following PCR, single
stranded PCR products were purified from 10 ml of second
round PCR reaction using a biotinylated primer and immobiliz-
ation to streptavidin beads with the PyroMark Vacuum Prep
Workstation (Biotage, Charlottesville, VA, USA), denatured
with NaOH and annealed to the sequencing primer by heating
to 808C for 2 min. Pyrosequencing reactions were performed
PSQ-HS96A system (Biotage). Genotyping of CRP variants
was successful in .98% of samples. 1% of samples failed at
multiple sites and were omitted from analyses. Quality
control measures included replication of one DNA plate (95
samples) with identical genotypes at each variant from two
independent PCR amplifications/Pyrosequencing reactions.
Within the family-based study, five families with unexplained
Mendelian errors were omitted from the study. Genotyping
reliability was confirmed by comparison of Pyrosequencing
results to direct sequence based analysis in 100 healthy controls
(50 Caucasian and 50 African-American) and 97 patients with
SLE (47 Caucasian and 50 African-American) that had been
previously sequenced (29).
African-American samples. Each of the eight genetic markers
was tested for departures from Hardy–Weinberg equilibrium
expectations using the exact test (45) in the control sample.
Measures of LD (D0and r2) were computed to determine the
block structure with the expectation-maximization algorithm-
based program Dprime (http://www.phs.wfubmc.edu/web/pub-
lic_bios/sec_gene/downloads.cfm). In the CASSLE case–
control study, each marker was tested for association with
SLE using the software SNPGWA and Dandelion (http://
cfm). Specifically, for each SNP we computed the following
tests: overall genotypic association test, dominant, additive
and recessive genetic model tests, departures from additively,
and two- and three-marker haplotype analyses. The OR and
corresponding 95% CI were computed relative to the major
allele for each of the above tests.
Although multiple tests were computed for each SNP to
provide context for any association, the primary statistical
inference was based on the overall genotypic test of associ-
ation. Specifically, if the overall genotypic test of association
was significant the three a priori genetic models were exam-
ined for the most likely source of the departure from the
null hypothesis. If the lack-of-fit test is not significant then
the additive model is reported. If the lack-of-fit test is signifi-
cant then the minimum of the dominant, additive and recessive
genetic models is reported. The approach is consistent with
the Fisher’s least significant difference (LSD) multiple
comparisons approach within an individual SNP. It maintains
the correct experiment-wise type 1 error rate since if
the overall genotypic test is not statistically significant no
individual genetic model is deemed statistically significant.
Finally, across the SNPs a simple Bonferroni correction is
were done separatelyfor Caucasianand
To test for association with SLE in the pedigree data a gen-
eralized estimating equations (GEE1) was computed using the
sandwich estimator to account for the within family corre-
lation. In addition, the PDT was used to test for LD (i.e.
association and linkage).
To account for potential confounding substructure or
admixture in these samples, principal component analyses
(PCA) were computed using 8230 SNPs in 1840 cases and
1819 controls of self-reported European descent from the
LLAS (35). A total of 906 individuals from LLAS overlapped
the Caucasian samples reported here. The first principal com-
ponent explained 85% of the genetic variation. The associated
first principal component scores measuring the genetic
variation were used as a covariate in a logistic regression
Supplementary Material is available at HMG Online.
We thank Debbie McDuffie and LiFeng Zhang for assistance
with genotyping and Dr Alex Szalai for helpful discussions
about CRP in autoimmune diseases. We also thank the
SLEGEN Consortium [JB Harley (Director), ME Alarco ´n-
Riquelme, LA Criswell, CO Jacob, RP Kimberly, KL Moser,
BP Tsao, TJ Vyse, and CD Langefeld (Co-Director)]
(https://www.phsapps.wfubmc.edu/SLEGEN) for access to
data for the Principal Component analysis. Materials and
data from the Lupus Family Registry and Repository
(NO1-AR6-22771) is acknowledged and appreciated.
Conflict of Interest statement. None declared.
This work was supported by NIH PO1 AR49084 Program
Project in the Genetics of SLE, AR43727, AR42460,
AI24717, AR048940, RR020143, a Mary Kirkland Award,
the US Department of Veteran Affairs and by General Clinical
Research Centers: M01-RR00032 (UAB), M01-RR00052
(JHU), M01-RR00048 (NU), and M01-RR02558 (UTH).
1. Kyttaris, V., Krishnan, S. and Tsokos, G. (2006) Systems biology in
systemic lupus erythematosus: integrating genes, biology and immune
function. Autoimmunity, 39, 705–709.
2. Petri, M. (2006) Systemic lupus erythematosus: 2006 update. J. Clin.
Rheumatol., 12, 37–40.
3. Ferna ´ndez, M., Alarco ´n, G., Calvo-Ale ´n, J., Andrade, R., McGwin, G.J.,
Vila ´, L. and Reveille, J. (2007) A multiethnic, multicenter cohort of
patients with systemic lupus erythematosus (SLE) as a model for the study
of ethnic disparities in SLE. Arthritis Rheum., 57, 576–584.
4. Alarco ´n-Riquelme, M. (2006) The genetics of systemic lupus
erythematosus: understanding how SNPs confer disease susceptibility.
Springer Semin. Immunopathol., 28, 109–117.
5. Graham, R., Kyogoku, C., Sigurdsson, S., Vlasova, I., Davies, L.,
Baechler, E., Plenge, R., Koeuth, T., Ortmann, W., Hom, G. et al. (2007)
Three functional variants of IFN regulatory factor 5 (IRF5) define risk and
1154 Human Molecular Genetics, 2008, Vol. 17, No. 8
by guest on June 12, 2013
protective haplotypes for human lupus. Proc. Natl Acad. Sci. USA, 104,
6. Gregersen, P. and Behrens, T. (2006) Genetics of autoimmune diseases–
disorders of immune homeostasis. Nat. Rev. Genet., 7, 917–928.
7. Harley, J., Kelly, J. and Kaufman, K. (2006) Unraveling the genetics of
systemic lupus erythematosus. Springer Semin. Immunopathol., 28,
8. Wu, H., Boackle, S., Hanvivadhanakul, P., Ulgiati, D., Grossman, J., Lee,
Y., Shen, N., Abraham, L., Mercer, T., Park, E. et al. (2007) Association
of a common complement receptor 2 haplotype with increased risk of
systemic lupus erythematosus. Proc. Natl Acad. Sci. USA, 104,
9. Yang, Y., Chung, E., Wu, Y., Savelli, S., Nagaraja, H., Zhou, B., Hebert,
H., Jones, K., Shu, Y., Kitzmiller, K. et al. (2007) Gene copy-number
variation and associated polymorphisms of complement component C4 in
human systemic lupus erythematosus (SLE): low copy number is a risk
factor for and high copy number is a protective factor against SLE
susceptibility in European Americans. Am. J. Hum. Genet., 80,
10. Alarco ´n, G., McGwin, G.J., Petri, M., Ramsey-Goldman, R., Fessler, B.,
Vila ´, L., Edberg, J., Reveille, J. and Kimberly, R. (2006) Time to renal
disease and end-stage renal disease in PROFILE: a multiethnic lupus
cohort. PLoS Med., 3, e396.
11. Criswell, L., Pfeiffer, K., Lum, R., Gonzales, B., Novitzke, J., Kern, M.,
Moser, K., Begovich, A., Carlton, V., Li, W. et al. (2005) Analysis of
families in the multiple autoimmune disease genetics consortium
(MADGC) collection: the PTPN22 620W allele associates with multiple
autoimmune phenotypes. Am. J. Hum. Genet., 76, 561–571.
12. Namjou, B., Kelly, J., Kilpatrick, J., Kaufman, K., Nath, S., Scofield, R.
and Harley, J. (2005) Linkage at 5q14.3-15 in multiplex systemic lupus
erythematosus pedigrees stratified by autoimmune thyroid disease.
Arthritis Rheum., 52, 3646–3650.
13. Graham, D., Graham, R., Manku, H., Wong, A., Whittaker, J., Gaffney,
P., Moser, K., Rioux, J., Altshuler, D., Behrens, T. et al. (2007)
Polymorphism at the TNF superfamily gene TNFSF4 confers
susceptibility to systemic lupus erythematosus. Nat. Genet, doi:10.1038/
14. Mahoney, J. and Rosen, A. (2005) Apoptosis and autoimmunity. Curr.
Opin. Immunol., 17, 583–588.
15. Mevorach, D. (2003) Systemic lupus erythematosus and apoptosis: a
question of balance. Clin. Rev. Allergy Immunol., 25, 49–60.
16. Navratil, J., Liu, C. and Ahearn, J. (2006) Apoptosis and autoimmunity.
Immunol. Res., 36, 3–12.
17. Ogden, C. and Elkon, K. (2006) Role of complement and other innate
immune mechanisms in the removal of apoptotic cells. Curr. Dir.
Autoimmun., 9, 120–142.
18. Carroll, M. (2004) The complement system in regulation of adaptive
immunity. Nat. Immunol., 5, 981–986.
19. Du Clos, T. (2003) C-reactive protein as a regulator of autoimmunity and
inflammation. Arthritis Rheum., 48, 1475–1477.
20. Marnell, L., Mold, C. and Du Clos, T. (2005) C-reactive protein: ligands,
receptors and role in inflammation. Clin. Immunol., 117, 104–111.
21. Pepys, M., Lanham, J. and De Beer, F. (1982) C-reactive protein in SLE.
Clin. Rheum . Dis., 8, 91–103.
22. Rordorf, C., Schnebli, H., Baltz, M., Tennent, G. and Pepys, M. (1982)
The acute-phase response in (NZB X NZW)F1 and MRL/l mice. J. Exp.
Med., 156, 1268–1273.
23. Crawford, D., Yi, Q., Smith, J., Shephard, C., Wong, M., Witrak, L.,
Livingston, R., Rieder, M. and Nickerson, D. (2006) Allelic spectrum of
the natural variation in CRP. Hum. Genet., 119, 496–504.
24. Crawford, D., Sanders, C., Qin, X., Smith, J., Shephard, C., Wong, M.,
Witrak, L., Rieder, M. and Nickerson, D. (2006) Genetic variation is
associated with C-reactive protein levels in the Third National Health and
Nutrition Examination Survey. Circulation, 114, 2458–2465.
25. Carlson, C., Aldred, S., Lee, P., Tracy, R., Schwartz, S., Rieder, M., Liu,
K., Williams, O., Iribarren, C., Lewis, E. et al. (2005) Polymorphisms
within the C-reactive protein (CRP) promoter region are associated with
plasma CRP levels. Am. J. Hum. Genet., 77, 64–77.
26. Lange, L., Carlson, C., Hindorff, L., Lange, E., Walston, J., Durda, J.,
Cushman, M., Bis, J., Zeng, D., Lin, D. et al. (2006) Association of
polymorphisms in the CRP gene with circulating C-reactive protein levels
and cardiovascular events. JAMA, 296, 2703–2711.
27. Miller, D., Zee, R., Suk Danik, J., Kozlowski, P., Chasman, D., Lazarus,
R., Cook, N., Ridker, P. and Kwiatkowski, D. (2005) Association of
common CRP gene variants with CRP levels and cardiovascular events.
Ann. Hum. Genet., 69, 623–638.
28. Pankow, J., Folsom, A., Cushman, M., Borecki, I., Hopkins, P., Eckfeldt,
J. and Tracy, R. (2001) Familial and genetic determinants of systemic
markers of inflammation: the NHLBI family heart study. Atherosclerosis,
29. Szalai, A., Wu, J., Lange, E., McCrory, M., Langefeld, C., Williams, A.,
Zakharkin, S., George, V., Allison, D., Cooper, G. et al. (2005)
Single-nucleotide polymorphisms in the C-reactive protein (CRP) gene
promoter that affect transcription factor binding, alter transcriptional
activity, and associate with differences in baseline serum CRP level.
J. Mol. Med., 83, 440–447.
30. Hage, F. and Szalai, A. (2007) C-reactive protein gene polymorphisms,
C-reactive protein blood levels, and cardiovascular disease risk. J. Am.
Coll. Cardiol., 50, 1115–1122.
31. Russell, A., Cunninghame Graham, D., Shepherd, C., Roberton, C.,
Whittaker, J., Meeks, J., Powell, R., Isenberg, D., Walport, M. and Vyse,
T. (2004) Polymorphism at the C-reactive protein locus influences gene
expression and predisposes to systemic lupus erythematosus. Hum. Mol.
Genet., 13, 137–147.
32. Jo ¨nsen, A., Gunnarsson, I., Gullstrand, B., Svenungsson, E., Bengtsson,
A., Nived, O., Lundberg, I., Truedsson, L. and Sturfelt, G. (2007)
Association between SLE nephritis and polymorphic variants of the CRP
and FcgammaRIIIa genes. Rheumatology (Oxford), 46, 1417–1421.
33. Edberg, J., Langefeld, C., Wu, J., Moser, K., Kaufman, K., Kelly, J.,
Bansal, V., Brown, W., Salmon, J., Rich, S. et al. (2002) Genetic linkage
and association of Fcgamma receptor IIIA (CD16A) on chromosome 1q23
with human systemic lupus erythematosus. Arthritis Rheum., 46,
34. Alarco ´n, G., McGwin, G.J., Petri, M., Reveille, J., Ramsey-Goldman, R.
and Kimberly, R. (2002) Baseline characteristics of a multiethnic lupus
cohort: PROFILE. Lupus, 11, 95–101.
35. SLEGEN Consortium (2008) A genome-wide association scan in women
with systemic lupus erythematosus (SLE) identifies ITGAM, PXK,
KIAA1542 and rs10798269 and confirms HLA, STAT4, FCGR2A, IRF5
and PTPN22. Nat. Genet., in press.
36. Kimchi-Sarfaty, C., Oh, J., Kim, I., Sauna, Z., Calcagno, A., Ambudkar,
S. and Gottesman, M. (2007) A ‘silent’ polymorphism in the MDR1 gene
changes substrate specificity. Science, 315, 525–528.
37. Rodriguez, W., Mold, C., Kataranovski, M., Hutt, J., Marnell, L. and Du
Clos, T. (2005) Reversal of ongoing proteinuria in autoimmune mice by
treatment with C-reactive protein. Arthritis Rheum., 52, 642–650.
38. Rodriguez, W., Mold, C., Marnell, L., Hutt, J., Silverman, G., Tran, D.
and Du Clos, T. (2006) Prevention and reversal of nephritis in MRL/lpr
mice with a single injection of C-reactive protein. Arthritis Rheum., 54,
39. Szalai, A., Nataf, S., Hu, X. and Barnum, S. (2002) Experimental allergic
encephalomyelitis is inhibited in transgenic mice expressing human
C-reactive protein. J. Immunol., 168, 5792–5797.
40. Szalai, A., Weaver, C., McCrory, M., van Ginkel, F., Reiman, R.,
Kearney, J., Marion, T. and Volanakis, J. (2003) Delayed lupus onset in
(NZB x NZW)F1 mice expressing a human C-reactive protein transgene.
Arthritis Rheum., 48, 1602–1611.
41. Berger, P., McConnell, J., Nunn, M., Kornman, K., Sorrell, J.,
Stephenson, K. and Duff, G. (2002) C-reactive protein levels are
influenced by common IL-1 gene variations. Cytokine, 17, 171–174.
42. Vickers, M., Green, F., Terry, C., Mayosi, B., Julier, C., Lathrop, M.,
Ratcliffe, P., Watkins, H. and Keavney, B. (2002) Genotype at a promoter
polymorphism of the interleukin-6 gene is associated with baseline levels
of plasma C-reactive protein. Cardiovasc. Res., 53, 1029–1034.
43. Tan, E., Cohen, A., Fries, J., Masi, A., McShane, D., Rothfield, N.,
Schaller, J., Talal, N. and Winchester, R. (1982) The 1982 revised criteria
for the classification of systemic lupus erythematosus. Arthritis Rheum.,
44. Hochberg, M. (1997) Updating the American College of Rheumatology
revised criteria for the classification of systemic lupus erythematosus.
Arthritis Rheum., 40, 1725.
45. Wigginton, J., Cutler, D. and Abecasis, G. (2005) A note on exact
tests of Hardy-Weinberg equilibrium. Am. J. Hum. Genet., 76,
Human Molecular Genetics, 2008, Vol. 17, No. 8 1155
by guest on June 12, 2013