Association of apoptosis-related microsatellite polymorphisms on chromosome 1q in Taiwanese systemic lupus erythematosus patients

Abstract

Apoptosis is important in the pathogenesis of systemic lupus erythematosus (SLE). Several genome-wide scan studies have suggested chromosome 1q as a genetic susceptibility locus for SLE. This study investigated the association of apoptosis-related genes on chromosome 1q, Fas ligand (FasL), interleukin (IL)-10 and poly(ADP-ribose) polymerase (PARP), promoter microsatellite multi-allelic polymorphisms with SLE susceptibility and clinical characteristics in Taiwan. This study recruited 237 SLE patients and 304 healthy controls. FasL, IL-10 and PARP promoter microsatellite polymorphisms were genotyped employing gene scan. IL-10, located on 1q31–32, emerged as a significant susceptibility gene locus in Taiwanese SLE (T4 statistic = 0·01). IL-10 CA21 allele was the most common allele of 15 identified in Taiwanese, displaying skewed distribution of susceptibility in Taiwanese SLE patients. Conversely, the IL-10 CA20 allele showed a protective effect of SLE susceptibility. Additionally, the IL-10 CA26 allele displayed a negative significant association with ascites and IL-10 CA25 allele increased the occurrence of the anti-cardiolipin IgM antibody. This study identified five alleles of FasL and nine alleles of PARP of microsatellite polymorphisms in Taiwanese patients. FasL and PARP alleles displayed no skewing distribution between Taiwanese SLE patients and controls. However, FasL GT15 and PARP CA17 allele demonstrated a high discoid rash presentation (T4 statistic 0·01 and 0·03, respectively) and PARP CA12 allele displayed a significant association with anti-cardiolipin IgM antibody production (T4 statistic 0·02). IL-10, FasL and PARP microsatellite polymorphisms exhibited significant associations with SLE susceptibility and/or clinical characteristics in Taiwanese patients. Thus, SLE is a complex and multiple genetics determined autoimmune disease. Chromosome 1q23–42 is an important genetic locus for further SLE subphenotype susceptibility study.

Introduction

Systemic lupus erythematosus (SLE) is a prototypic systemic autoimmune disease characterized by dysregulation of cellular immune responses and reactivity of various endogenous antigens. The precise pathogenic mechanisms to clarify autoimmune phenomena remain partly unexplained. Both environmental initiating elements and genetic background play crucial roles in the pathogenesis of SLE. Functional dissecting studies of lupus with congenic mice models demonstrated that the individual gene locus on different chromosome regions affects specific immunological functions and contributes to pathogenic autoimmunity [1]. Additionally, mice with apoptosis-related gene defects exhibit a number of characteristics that mimic SLE features and serological abnormality [2–4].

Human SLE heritable work also demonstrated family clustering of disease and higher concordance rates of 24% in monozygotic over 2% in dizygotic twins [5,6]. In recent years, genome-wide scan studies with SLE multi-case families have mapped many significant lupus predisposing gene loci including chromosomes 1q22–23, 1q31–32 and 1q41–42 [7–14]. The multigenic entities of SLE led to this investigation of potential function-related genes located in these regions.

Apoptosis is a physiological process leading to the ordered elimination of cells and particularly the removal of superfluous or autoreactive danger cells, averting inflammatory consequences by liberating intracellular contents [15]. SLE patients display accelerated apoptosis of circulating lymphocytes [16–18]. The surplus of apoptotic cells and inadequate clearance of apoptotic cells mediated by phagocytes may be a source of autoantigens [19]. Finally, the dissolution of self-tolerance by dysfunctional apoptotic pathways initiates B cell hyperactivity and consequent autoantibody overproduction [19,20].

The Fas ligand (FasL) colligates with the Fas death receptor, triggering apoptosis by a signalling pathway to caspase 8 and 10 activation [21]. Abnormal expression of FasL on lymphocytes and triggering apoptosis has been demonstrated in SLE patients [22]. Interleukin (IL)-10 promotes the activation of Fas ligand-mediated cell death but prevents the spontaneous cell death of germinal centre B cells by induction of bcl-2 [23,24]. Dysregulation of IL-10 production may contribute to lupus pathogenesis and disease activity [25,26]. Poly(ADP-ribose) polymerase 1 (PARP-1) is a key nuclear enzyme mediating the poly(ADP-ribosylation) reaction that detects and binds DNA produced by genotoxic agents [27]. SLE patients reveal a diminution in PARP synthesis that appears to be an inherited trait [28]. FasL (1q23), IL-10 (1q31–32) and PARP (1q42) genes, positioned on chromosome 1q 23–42, are considered to be the susceptible locus for lupus. Regarding the positional and functional candidate gene approach, the genetic effects of these apoptosis-associated genes with relevant immunological functions are pertinent to the pathogenesis of lupus.

This study investigated the consequences of promoter microsatellite genotypes variants among three apoptosis-associated genes in Taiwan lupus patients.

Materials and methods

Characteristics of study populations

A total of 237 patients with SLE in Taiwan was sampled. Additionally, 304 healthy controls, comprising 270 female and 34 male blood donors with a mean age of 33·1 ± 10·7 years (range 18–58), were selected following a questionnaire survey to exclude autoimmune diseases. SLE patients in this study were followed-up in the rheumatology clinics of Chang Gung Memorial Hospital. Rheumatology specialists assessed SLE patients to confirm that they fulfilled the 1982 and 1997 American College of Rhematology (ACR) criteria for SLE. Clinical manifestations of lupus and related serological findings were based on the definition of ACR classification criteria. Patients were grouped as positive or negative according to presentations within the first year of lupus diagnosis. Associated infections during the disease course were recorded, and were supported by positive culture.

Nucleic acid extraction and microsatellite genotype analysis

Genomic DNA was extracted from ethylenediamine tetraacetic acid (EDTA) anti-coagulated peripheral blood using the Purgene DNA isolation kit. Briefly, 3 ml of blood was lysed in 9 ml of red blood cell lysis solution. The resuspended cells were then added to 3 ml of cell lysis solution and pelleted. The RNase solutions were added to the nuclear lysate followed by protein precipitation solution. DNA in the supernatant was precipitated with isopropanol and washed with 70% ethanol. Unique oligonucleotide sequence flanking each microsatellite was designed as primer, one of which was labelled with either FAM or NED fluorescent dye, applied to the polymerase chan reaction (PCR) reaction. PCR reactions were executed in a GeneAmp PCR system with 10 ng of genomic DNA in a 20 µl reaction. The PCR primers and cycling conditions were as follows: IL-10 forward primer sequences were 5′-CCCAACTGGCTCCCCT TAC-3′ and the 5′ primer tagged with the fluorescent dye 6FAM; IL-10 reverse primer sequences were 5′-TGGGGTG GAAGAAGTTGAAATAA-3′. Cycling conditions were 95°C for 5 min followed by 30 cycles of denaturation at 95°C for 1 min, annealing at 56°C for 1 min and extension at 72°C for 1 min. The cycle was finished off with a final extension at 72°C for 7 min. FasL forward primer sequences were 5′-ATCTTGACCAAATGCAACCCATA-3′; FasL reverse primer sequences were 5′-GGGATGGAAAAGGAAGACTAGAGG-3′ and the 5′ primer tagged with the fluorescent dye 6FAM. PARP forward primer sequences were 5′-GATTCCCCA TCTCTTTCTTTACACA-3′; PARP reverse primer sequences were 5′-GGGCAATAGTCATCACAGACGTT-3′ and the 5′ primer tagged with the fluorescent dye NED. Cycling conditions resembled IL-10, except FasL annealing at 54·3°C and PARP annealing at 52°C. Aliquots of the PCR product were electrophoresed on a prism ABI sequencer. Varying fluorescent peaks estimated in base pairs by reference to the lane size standard Ros500. The fluorescent signals recorded and analysed by genescan software. The dinucleotide tandem repeat numbers were verified by sequencing of homozygous samples.

Statistical analysis

Allele frequencies of the microsatellite markers were compared between patients and controls. A paired t-test and a χ² test were employed in determining significance of each allele. Data were analysed using the SPSS statistical package for Windows and Fisher's exact test applied. The allocation of variant genotypes between lupus patients with and without clinical manifestations and autoantibodies was also compared. The allelic distribution was compared by 2 × N tables using the T4 statistic of the clump software [29]. An empirical estimate of the P-value was evaluated by performing 100 simulations using the Monte Carlo method. A P-value < 0·05 was considered significant.

Results

Clinical characteristic of SLE patients

This study enrolled 237 SLE patients (19 male and 218 female; mean age at onset, 29·7 ± 11·5, range 11–77 years). Malar rash (159 patients; 67·1%) was the most common manifestation, followed by arthritis (153 patients; 64·6%) and haematological involvement (150 patients; 63·3%). Altogether, 116 patients (48·9%) developed nephritis. The incidence of other manifestations was as follows: oral ulcer (77 patients; 32·5%), vasculitis (72 patients; 27·7%), photosensitivity (61 patients; 25·7%), serositis (58 patients; 24·5%; including 40 patients with pleural effusion, 37 with pericardial effusion and 10 patients with ascites), discoid rash (50 patients; 21·1%), Raynaud's phenomenon (37 patients; 15·7%) and neuropsychiatric manifestations (23 patients; 9·7%). The age at onset was most frequently 20–29 years (105 patients), followed by 30–39 years (61 patients), less than 19 years (50 patients) and older than 40 years (44 patients). Infections were encountered in 105 patients. Most SLE patients had anti-dsDNA antibody (176 of 215) and depressed complement (208 of 237). According to the available data of autoantibodies determined by enzyme-linked immunosorbent assay, 178 SLE patients included 82 with anti-Sm and 80 with anti-RNP antibody (66 had both positive), 103 SLE patients included 69 with anti-SSA and 31 with anti-SSB antibody (29 had both positive) and the 110 SLE patients included 52 with anti-cardiolipin IgG and 14 with anti-cardiolipin IgM antibody (nine had both positive).

Associations of FasL, IL-10 and PARP microsatellite polymorphisms with SLE susceptibility

Fifteen alleles of IL-10G microsatellites with CA tandem repeat ranged from 16 to 30 detected in Taiwanese patients. The allele of CA21 repeats was the most common and appeared as a susceptible allele in Taiwanese SLE. In contrast, IL-10 CA20 allele demonstrated a protective role to lupus susceptibility (Table 1). As a biallelic marker, an IL-10 long allele (IL-10 CA repeats ≥21) demonstrated an association with SLE disease susceptibility (P= 0·001; OD: 1·898; 95% CI 1·898–2·749). Five alleles of FasL with GT tandem repeat range from 14 to 18 were detected in Taiwanese SLE, while nine alleles of PARP with CA tandem repeats ranging from 12 to 20 were disclosed. No statistically significant differences appeared in the distribution of FasL and PARP alleles between normal controls and SLE patients (Tables 2 and 3). Overall, IL-10 is a significant SLE susceptibility genetic locus for Taiwanese patients with a T4 statistic P-value = 0·01 using the Monte Carlo method.

Table 1.

Interleukin (IL)-10 proximal CA dinucleotide microsatellite polymorphism of Taiwan systemic lupus erythematosus (SLE) patients and healthy controls.

Genotypes	SLE (2n = 474)	Controls (2 n = 608)	OR	95% CI	P-values
192 (CA)16	8 (0·017)	12 (0·02)	0·853	0·346–2·103	0·729
194 (CA)17	1 (0·002)	1 (0·002)	1·285	0·08–20·604	1·0
196 (CA)18	0	3 (0·005)	–	–	0·261
198 (CA)19	7 (0·015)	9 (0·015)	0·998	0·369–2·699	0·996
200 (CA)20	30 (0·063)	78 (0·128)	0·459	0·296–0·713	< 0·0001*
202 (CA)21	231 (0·487)	240 (0·395)	1·458	1·114–1·858	0·002
204 (CA)22	63 (0·133)	85 (0·140)	0·943	0·664–1·339	0·743
206 (CA)23	5 (0·011)	15 (0·025)	0·421	0·152–1·168	0·087
208 (CA)24	7 (0·015)	19 (0·031)	0·465	0·194–1·115	0·079
210 (CA)25	49 (0·103)	52 (0·086)	1·233	0·818–1·858	0·317
212 (CA)26	49 (0·103)	64 (0·105)	1·980	0·661–1·452	0·920
214 (CA)27	11 (0·023)	13 (0·021)	1·087	0·483–2·449	0·84
216 (CA)28	10 (0·021)	13 (0·021)	0·986	0·429–2·270	0·974
218 (CA)29	2 (0·004)	2 (0·003)	1·276	0·179–9·090	1·0
220 (CA)30	1 (0·002)	2 (0·003)	0·637	0·058–7·041	1·0

^*T4 statistic of Monte Carlo test = 0·01 in IL-10 proximal CA dinucleotide microsatellite polymorphism.

Table 2.

Fas ligand (FasL) GT dinucleotide microsatellite polymorphism of Taiwan systemic lupus erythematosus (SLE) patients and healthy controls.

Genotypes	SLE (2 n = 474)	Controls (2 n = 608)	OR	95% CI	P-values
293 (GT)14	2 (0·004)	4 (0·007)	0·640	0·117–3·508	0·604
295 (GT)15	15 (0·032)	11 (0·012)	1·774	0·807–3·898	0·149
297 (GT)16	436 (0·920)	563 (0·926)	0·917	0·585–1·438	0·706
299 (GT)17	20 (0·042)	30 (0·049)	0·851	0·477–1·518	0·584
301 (GT)18	1 (0·002)	0	–	–	0·257

Table 3.

ADP dinucleotide microsatellite polymorphism of Taiwan systemic lupus erythematosus (SLE) patients and healthy controls.

Genotypes	SLE (2 n = 474)	Controls (2 n = 608)	OR	95% CI	P-values
171 (GT)12	94 (0·198)	127(0·209)	0·937	0·695–1·263	0·669
173 (GT)13	4 (0·008)	3 (0·005)	1·716	0·382–7·706	0·476
175 (GT)14	1 (0·002)	0	–	–	0·257
177 (GT)15	4 (0·008)	2 (0·003)	2·579	0·470–14·139	0·258
179 (GT)16	155 (0·327)	174 (0·286)	1·212	0·934–1·572	0·148
181 (GT)17	37 (0·078)	60 (0·099)	0·773	0·504–1·187	0·239
183 (GT)18	148 (0·312)	189 (0·311)	1·006	0·777–1·304	0·961
185 (GT)19	33 (0·070)	54 (0·089)	0·768	0·489–1·205	0·249
187 (GT)20	2 (0·004)	1 (0·002)	2·572	0·233–28·450	0·424

Associations of FasL, IL-10 and PARP microsatellite polymorphisms with SLE clinical characteristics

The genetic association of FasL, IL-10 and PARP microsatellite polymorphisms was compared in clinical manifestations and serologic positive versus negative patients (Table 4). Patients carrying the IL-10 CA20 allele had a high incidence of central nervous system (CNS) involvement, IL-10 CA26 allele with ascites. With regard to serological findings, the IL-10 CA25 allele increased the prevalence of anti-cardiolipin IgM antibody. FasL microsatellite polymorphism revealed two clinical relationships to lupus patients. The FasL GT15 allele demonstrated a high discoid rash presentation (T4 statistic 0·01) and FasL GT17 allele reduced haematological involvement (T4 statistic 0·168). PARP microsatellite polymorphisms associated with three clinical manifestations and autoantibody production. PARP CA12 allele and PARP CA18 demonstrated a protective roles of arthritis ascites, respectively, and PARP CA17 showed an increased incidence of discoid rash. The PARP CA12 allele displayed a significant negative association with anti-cardiolipin IgG antibody production (T4 statistic 0·02). Additionally, the PARP CA12 allele revealed a negative association and the PARP CA18 allele displayed a positive association with ds-DNA antibody production (T1 statistic 0·001; T4 statistic 0·099).

Table 4.

Interleukin (IL)-10, Fas ligand (FasL) and ADP dinucleotide microsatellite polymorphisms display skewed distribution between positive and negative clinical characteristics of Taiwan systemic lupus erythematosus (SLE) patients.

Characteristics	Genotypes	Positive n (%)	Negative n (%)	OR	95%CI	P-value	T4†
IL-10
″CNS‡	200 CA20	6/46 (13·0)	24/404 (5·6)	2·525	0·975–6·540	0·049
″Ascites	212 CA26	6/20 (30·0)	43/454 (9·5)	4·126	1·508–11·291	0·003
″ACA IgM	210 CA25	5/28 (17·9)	11/192 (5·7)	3·577	1·141–11·215	0·021
Fas L
″Discoid rash	295 GT15	9/100 (9·0)	6/374 (1·6)	6·066	2·105–17·477	< 0·001	0·01
″Haematology	299 GT17	8/300 (2·7)	12/174 (6·9)	0·370	0·148–0·923	0·027	0·168
ADP
″Arthritis	171 CA12	51/306 (16·7)	43/168 (25·6)	0·581	0·367–0·920	0·02	0·079
″Discoid rash	181 CA17	13/100 (13)	20/374 (5·3)	2·645	1·266–5·525	0·008	0·03
″Ascites	183 CA18	2/20 (10·0)	146/454 (32·2)	0·227	0·052–0·991	0·046	0·178
″ACA IgG	171 CA12	15/104 (14·4)	34/116 (29·3)	0·406	0·206–0·800	0·008	0·02
″Ds DNA	171 CA12	65/352 (18·5)	21/78 (26·9)	0·615	0·348–1·085	0·091	0·099
	183 CA18	120/352 (34·1)	18/78 (23·1)	1·724	0·974–3·052	0·059

^†Monte Carlo test.

^‡CNS: central nervous system.

Discussion

Searching for potential lupus susceptibility loci by using genome-wide scans studies has produced disparate results. These erratic results may due to different applications of statistical methods, stratification and the admixture of populations with a heterogeneous disease entity of SLE as well as ethnic differences of population. Another possible explanation of the disparities could be the genetic markers and genetic maps used for analysis. Although no genome-wide scan was accomplished in the Taiwanese population, chromosome 1q22–42 has been mapped as a significant SLE susceptibility locus by many studies in different populations [7–14]. Additionally, apoptosis defects involving multiple genes may contribute to SLE disease development. Thus, three apoptosis-associated genes on chromosomes 1q23, 1q31–32 and 1q42 were selected for this controlled study on association. The present study disclosed that chromosome 1q31–32 is a potential lupus susceptibility locus in Taiwanese patients.

Abnormal expressions of Fas and FasL on lymphocytes and keratinocytes contribute to apoptosis defect and tissue injury in SLE patients [22,30,31]. Anti-FasL antibody was detected in SLE patients, inhibiting the apoptotic cell death mediated by Fas/Fas ligand pathway [32]. The FasL gene is positioned on chromosome 1q23, which is a putative susceptible gene locus of SLE, including cutaneous manifestations [8–10,33]. However, gene mutation of FasL over coding regions is observed infrequently among human lupus patients [34]. Thus, gene variations of promoter regions influencing FasL expression may participate in lupus disease pathogenesis. Mehrian et al. demonstrated a skewed distribution of FasL GT15 alleles between American Mexican SLE patients and controls [35]. FasL microsatellite polymorphisms, with a relatively high homogeneous distribution of up to 90% of FasL GT16 allele in Taiwanese SLE, indicated that the FasL microsatellite polymorphism is not a susceptibility marker to lupus in Taiwanese SLE. Nevertheless, FasL GT15 alleles established a significant relationship with discoid rash in Taiwanese SLE patients. This implies that the chromosome 1q23 gene locus may relate to the pathogenesis of SLE dermatological manifestations such as discoid rash.

IL-10, an important immunoregulatory cytokine, is elevated in SLE patients as well as their relatives [4,25]. Inhibition of IL-10 production could delay the onset of systemic lupus erthematosus [36]. IL-10G and IL-10R microsatellite haplotypes influence the production of IL-10 [37]. IL-10 promoter polymorphisms were correlated with the clinical manifestation of SLE [38,39]. These correlations indicate that the genetic variation of IL-10 genotypes is crucial to IL-10 production and lupus disease entity. Alfonso et al. verified that IL-10G is the only distinct IL-10 marker of this important lupus susceptibility gene in an Italian population [40,41]. They concluded that IL-10 long alleles (CA repeats ≥22) are associated with high IL-10 production and correlate with SLE susceptibility [41]. Recently, Chong et al. reported that a short allele with IL-10 CA repeats ≤21 had a dose-dependent effect on SLE susceptibility in Hong Kong Chinese patients [42]. The present study displayed a similar distribution of IL-10 alleles to Hong Kong Chinese patients, but revealed a protective role with short alleles of ≤ CA20 repeats, especially the IL-10 CA20 allele and susceptibility to IL-10 CA21 allele. The possible differences may due to sample selection bias such as the high percentage of female controls in this study. Thus, IL-10G is a potential susceptibility gene marker in Taiwanese SLE patients. However, different alleles have been linked with SLE susceptibility in different studies, including the IL-10 CA21 allele in the present study, the IL-10 CA22 allele in a Mexican American population, the IL-10 CA25 allele in a British population and the IL-10 CA23 allele in an Italian population [35,39–41]. Taken together, these studies showed that the IL-10 gene locus is a significant marker with linkage disequilibria to lupus susceptibility for different populations.

PARP-1 involved several nuclear processes that related to post-translational modification of proteins, including DNA damage signalling and base excision repair, regulation of genomic stability, modulation of transcription and cell death [27]. The deficiencies of DNA repair were observed in SLE and defects in PARP synthesis has been documented in SLE patients and their relatives [27,28]. Additionally, anti-PARP antibody was detected in SLE patients that may interfere with caspase-3 mediated cleavage of PARP during early apoptosis [43]. Tsao et al. demonstrated a skewed transmission of PARP alleles in a family study with the PARP CA8 allele as susceptible and the PARP CA14 allele as a protector of lupus transmission [44]. They suggested that PARP was linked to the lupus susceptibility gene located in chromosome 1q 41–42, although other groups cannot confirm it [44–46]. Additionally, Delrieu et al. concluded that PARP did not influence disease susceptibility in SLE or primary anti-phospholipid syndrome in French Caucasians [47]. In contrast, Tan et al. observed that PARP alleles might exist in linkage disequilibria with a lupus susceptibility locus in an African American cohort association study [48]. However, the small size of cohort samples of these studies necessitates cautious interpretation. This present study encountered a different distribution of PARP alleles to previous different ethnic groups. Taiwanese displayed a lesser prevalence of the PARP CA12 allele and a greater prevalence of PARP CA16 and PARP CA18 alleles. Although there was no statistically significant association with SLE susceptibility, PARP microsatellite polymorphisms demonstrate associations with clinical subphenotypes such as discoid rash and arthritis, anti-cardiolipin IgG and anti-ds-DNA antibody production. These indicate that PARP CA repeats may play a key role in lupus pathogenesis involving DNA repair of cell damage and consequent autoantibody production.

In conclusion, SLE patients demonstrate multi-organ and multi-system involvement with heterogeneous manifestations. Lupus pathogenesis is complex, and multiple genetic defects might be essential for the development of full-blown disease. The present study suggested that chromosome 1q23–42 is an important locus for lupus. Further detailed mapping of putative lupus susceptibility and clinically associated genes in these regions are indicated in the future.