CR1 exon variants are associated with lowered CR1 expression and increased susceptibility to SLE in a Plasmodium falciparum endemic population

Subjects

Patients with SLE admitted to the Department of Medicine, under the unit of Clinical Immunology and Rheumatology of SCB Medical College, Cuttack, Odisha, were enrolled in the present study. The diagnosis of SLE was based on revised American College of Rheumatology classification criteria.27 Since, SLE affects mostly women28 only female patients from coastal districts of Odisha (Cuttack, Puri, Anugul, Khurdha and Dhenkanal), endemic to P. falciparum infection,18 ,29 ,30 were included in the study. The average annual parasite index of malaria in these districts is 6.67.31 After a detailed clinical examination and laboratory investigations, patients were categorised into various clinical phenotypes. The clinical profiles of 297 patients with SLE are summarised in table 1. Three hundred age-matched, healthy female subjects from similar geographical areas (costal districts) were included as healthy controls (HCs). None of the controls reported history of autoimmune disorder or malaria. Both, patients with SLE and HCs were free of active P. falciparum infection as shown by immunochromatography test and P. falciparum-specific PCR.18 About 5 mL of blood was collected in EDTA tubes from each participant. The methods were carried out in accordance with approved guidelines. Informed consent was obtained from each patient and HC.

DNA isolation and genotyping of CR1 variants

Genomic DNA was isolated from whole blood by the GenElute Blood Genomic DNA Kit (SIGMA). As described earlier,32 samples were genotyped for the intron 27 HindIII A>T, exon 22 3650A>G and exon 33 5007 G>C genes by PCR-restriction length polymorphism (RFLP). For intron 27 and exon 22 the restriction fragment length polymorphism (RFLP) products were separated on 3.5% agarose and visualised by ethidium bromide staining under trans-UV illumination. To resolve restriction fragments of exon 33 polymorphism, 16% non-denaturing polyacrylamide gel was used. For quality control of genotyping results, about 25% of the randomly selected samples were directly sequenced for exon 22, exon 33 and about 15% were sequenced for intron 27 and the results were observed to be 100% concordant. Electropherograms of sequencing analysis are shown in online supplementary figure S1.

Quantification of CR1 expression on monocytes

CR1 expression on monocytes was determined by flow cytometry using two-colour fluorescence in 35 patients with SLE and 33 HCs. Briefly, 50 μL of whole blood were directly stained for 30 min at 4°C with phycoerythrin-labelled anti-CR1 (eBiosciences) and fluorescein-isothiocyanate labelled anti-CD14 (eBiosciences). Appropriate isotype controls (eBiosciences) were also used. Erythrocytes were lysed using fluorescence-activated cell sorting (FACS) lysis solution (BD Biosciences, USA) and the cells were washed with washing buffer followed by centrifugation at 300×g for 5 min. The cell pellet was then resuspended in 500 μL of sheath fluid and analysed by flow cytometry in FACS calibre using CellQuest Pro software.

Statistical analysis

The genotype and allele frequency were calculated by direct counting. Deviations from the Hardy-Weinberg equilibrium (HWE) were tested using the web-based site http://www.oege.org/software/hwe-mr-calc.shtml.33 Linkage disequilibrium (LD) between CR1 polymorphisms (intron 27, exon 22 and exon 33) were calculated by SNPAlyze software. Fisher's exact test was used for comparison of genotype, allele frequencies distribution among patients with SLE and HCs, and also among various clinical phenotypes. Ors and 95% CIs were calculated by Graphpad prism V.5.01. A p value <0.003 (Bonferroni correction for three markers 0.01/3=0.003) was considered statistically significant. Haplotype analysis was performed by the SNAPStats online tool and an LD plot was constructed in SNPAlyze software (Dynacom, Japan). Student's t-test was employed to compare the mean CR1 expression on monocytes in patients with SLE versus HCs and among patients with SLE with LN (LN+) versus those without kidney involvement (LN−). Association of CR1 genotypes (intron 27, exon 22 and exon 33) with CR1 expression on monocytes was analysed by analysis of variance followed by Tukey's post test and a p value <0.05 was considered statistically significant.

Prevalence of CR1 polymorphisms in the studied population

Three hundred healthy women were genotyped for the CR1 polymorphisms intron 27, exon 22 and exon 33. As shown in table 2, and similar to our earlier observation,18 heterozygous and homozygous alleles were more frequent than major alleles (intron 27: AT=51%, TT=27%; exon 22: AG=41%, GG=44%; exon 33: CG=50%, GG=35%). Distributions of CR1 genotypes in HCs were in HWE (intron 27: χ²=0.25, p=0.61; exon 22: χ²=3.18, p=0.07; exon 33: χ²=0.45, p=0.50). The pairwise LD analysis of all the three loci showed that exon 22 (A>G) and exon 33 (G>C) were in LD with |D′|=0.93, r²=0.72, Akaike's information criterion (AIC)=509.02. The other combinations were not in LD in the studied population (for intron 27 (A>T) and exon 22 (A>G) pair: |D′|=0.79, r²=0.38, AIC=256.82; for intron 27 (A>T) and exon 33 (G>C) pair |D′|=0.72, r²=0.39, AIC=250.77).

Further, we constructed LD plots for patients with SLE and HCs (see online supplementary figure S2). LD plots of codon 22 and codon 33 were similar in patients with SLE (D′=0.94) and HCs (D′=0.93). However, D′ values for intron 27 and codon 22 or between intron 27 and codon 33 were different in patients with SLE (intron 27-exon22: D′=0.74; intron 27-exon33: D′=0.63) and HCs (intron 27-exon22: D′=0.79; intron 27-exon33: D′=0.72).

Exon 22 and exon 33 polymorphisms are associated with susceptibility to SLE

We used a hospital based case control study to investigate the association between common polymorphism of the CR1 gene and SLE. Three functional CR1 polymorphisms (intron 27, exon 22 and exon 33) were typed in patients with SLE and HCs and results are shown in table 2.

For the CR1 exon 22 polymorphism, the frequency of the GG genotype and variant allele G were significantly higher in patients with SLE compared with HCs (GG: p<0.0001, OR=4.27; G: p<0.0001, OR=1.79). Similarly, distributions of homozygous allele (GG) and minor allele (G) for CR1 exon 33 polymorphism were more frequent in patients than controls (GG: p<0.0001, OR=3.82; G: p<0.0001, OR=2.11). Furthermore, both dominant and recessive genetic models revealed significant association of the CR1 codon (22 and 33) polymorphism with SLE (codon 22 (dominant model: p=0.00001, OR=3.66; recessive: p=0.001, OR=1.72); codon 33 (dominant model: p=0.001, OR=2.50; recessive: p=0.0001, OR=2.62). For CR1 intron 27, the genotype and allele distribution in patients with SLE and HCs were comparable.

Since codon 22 and codon 33 were associated with SLE, we further analysed haplotype association by the SNAPStat online tool and results are shown in table 3. The GG haplotype was more frequent in patients with SLE compared with HCs (p=0.0004, OR=1.94).

CR1 expressions on monocytes are lower in SLE and in LN

CR1 expression on monocytes was measured by flow cytometry in patients with SLE (n=35) and HCs (n=33). As shown in figure 1A, HCs displayed significantly higher levels of CR1 expression compared with patients with SLE (p=0.005). Patients with SLE were further categorised into two broad groups: (1) LN+, and (2) LN−. As shown in figure 1B, LN+ patients had lower CR1 expression on monocytes compared to LN- (p=0.01).

Request permission

Association of polymorphisms with monocyte expression of CR1

Association of common CR1 polymorphisms (intron 27, exon 22 and exon 33) with expression of CR1 on erythrocytes has been demonstrated earlier.19 ,25 ,26 ,32 ,34 ,35 We hypothesised that common CR1 polymorphisms may lead to its altered expression on monocytes which may be a possible trigger in the development of SLE. In patients with SLE, major alleles for exon 22 (AA) and exon 33 (CC) polymorphisms displayed significantly higher levels of CR1 on monocytes compared with their respective heterozygotes (exon 22: AG; exon 33: CG) and homozygous (GG) alleles (figure 2D, F). Furthermore, CR1 expression on monocytes of the CG genotype (exon 33) was observed to be higher than the homozygous mutants (GG). Such difference between heterozygous (AG) and homozygous minor alleles (GG) was not observed in CR1 exon 22 polymorphism (figure 2D, F). Expression of CR1 on monocytes of subjects with various genotypes of intron 27 polymorphisms was comparable (figure 2B). Interestingly, similar genotype-phenotype association results were also observed in HCs (figure 2A, C, E).

Request permission

Distribution of CR1 polymorphisms in LN

Patients with LN display lowered levels of CR1 on monocytes and its expression was associated with common CR1 variants (exon 22 and exon 33), We analysed for possible association of CR1 polymorphisms with susceptibility/resistance to LN. As depicted in table 4, the GG genotype of CR1 exon 33 polymorphism was more frequent in LN+ patients compared with LN− patients. However, the difference was not statistically significant (p=0.03, OR=3.12). Distributions of other polymorphisms (intron 27 and exon 22) in clinical phenotypes (LN+ vs LN−) were comparable (table 4).

Association between CR1 expression and autoantibodies

Diminished CR1 expression is functionally associated with defective clearance of ICs which could be a possible trigger for autoantibody formation in SLE. We analysed the association between CR1 polymorphisms and levels of anti-dsDNA and the results are depicted in figure 3. The GG genotype of CR1 exon 22 and exon 33 polymorphisms displayed higher anti-dsDNA titres than their respective reference genotypes (exon 22: AA; exon 33: CC). However, the difference was not statistically significant. Anti-dsDNA titres were comparable among genotypes of the CR1 intron 27 polymorphism. Furthermore, we analysed correlation between CR1 expression on monocytes and levels of anti-dsDNA (figure 3). A negative correlation was observed between CR1 expression and anti-dsDNA (p=0.0007, Spearman's r=−0.54), suggesting a possible role of CR1 in facilitating increased production of anti-ds DNA.

Request permission