Rheumatol Int DOI 10.1007/s00296-015-3298-x
Rheumatology INTERNATIONAL
ORIGINAL ARTICLE - GENES AND DISEASE
Beta‑adducin and sodium–calcium exchanger 1 gene variants are associated with systemic lupus erythematosus and lupus nephritis Giuseppe A. Ramirez1,2 · Chiara Lanzani3 · Enrica P. Bozzolo1 · Laura Zagato3 · Lorena Citterio3 · Nunzia Casamassima3 · Valentina Canti1,2 · Maria Grazia Sabbadini1,2 · Patrizia Rovere‑Querini1,2 · Paolo Manunta2,3 · Angelo A. Manfredi1,2
Received: 19 January 2015 / Accepted: 25 May 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Genetic research in systemic lupus erythematosus (SLE) is rapidly developing, and numerous sets of genes are being associated with specific clinical subphenotypes in the setting of SLE. On the other hand, basic science studies are revealing strong connections between salt–water balance and inflammation. The aim of this study was to evaluate whether variants of genes known to influence the individual susceptibility to hypertension also influence the renal function in a cohort of SLE patients with and without lupus nephritis (LN). This study is a case–control study with candidate gene approach. A total of 111 patients with SLE (50 with SLE without nephritis, 55 with LN and 6 with simple urinary sediment abnormalities) and 62 healthy controls (HC) were genotyped for NCX1 rs11893826 (NCX1a) and rs434082 (NCX1b) and ADD2 rs4984 SNPs. Patients with ADD2 CT genotype were protected from LN and skin involvement; ADD2 CC | NCX1a AA/AG genotypes were associated with the presence of anti-cardiolipin antibodies; NCX1a AA genotype was slightly more frequent in lupus patients than in HC and associated with relapse risk and higher creatinine in patients with LN. NCX1b GG patients with LN had increased chances to reach complete remission. NCX1b GG | NCX1a GG genotype is associated with joint involvement. ADD2 and NCX1 variants influence the risk and the clinical features of SLE and LN, highlighting
* Giuseppe A. Ramirez [email protected] 1
Unit of Internal Medicine and Immunology, IRCCS Ospedale San Raffaele, Via Olgettina 60, 20132 Milan, Italy
2
Università Vita-Salute San Raffaele, Milan, Italy
3
Unit of Nephrology, IRCCS Ospedale San Raffaele, Milan, Italy
their potential role in regulating systemic inflammation and/or the local response to immune-mediated injury. Keywords Systemic lupus erythematosus · Sodium– calcium exchanger 1 · Beta-adducin · Nephritis · Skin · Anti-phospholipid antibodies · Arthritis Abbreviations ADD2 Beta-adducin gene CR Complete remission HC Healthy controls LLAC Lupus-like anticoagulants LN Lupus nephritis NCX1 Solute carrier family 8 (sodium–calcium exchanger) member 1 gene OR Odds ratio SLE Systemic lupus erythematosus SLEnonLN SLE without LN SNP Single-nucleotide polymorphism SUA Simple urinary sediment abnormalities WLC Whole lupus cohort
Introduction Systemic lupus erythematosus (SLE) is a prototypic multiorgan autoimmune disease, characterized by polymorphic clinical features, arising from a complex and only partially known genetic background [1–3]. Renal disease in SLE occurs in 40–70 % of lupus patients and is responsible for a high burden of morbidity and mortality [4]. A first set of genes within the so-called MHC region in chromosome 6 (which includes TNF and complement factors besides HLA variants) has the strongest association with the risk of developing SLE in large cohorts [5], together
13
with non-MHC genes such as interferon regulatory factor 5 (IRF5), signal transducer activator of transcription 4 (STAT4) and integrin alpha-M (ITGAM) [6, 7]. A recent meta-analysis from the International Consortium for SLE Genetics provided evidence for a role of a wide number of genes in the development of lupus nephritis (LN) in Caucasian women with SLE. The authors found that only a small fraction of classical SLE susceptibility loci were also specifically associated with the development of LN and concluded that LN loci probably influence the local (kidney) response to the thrust posed by the immune system, rather than acting systemically on the immune response [2]. Thus, structural genes with no evident immune function could be implicated in the susceptibility to SLE and to specific SLE phenotypes. Blood pressure control and inflammation are composite and apparently unrelated physiological functions that depend on the multi-organ cooperation of several genes. Cellular machineries devoted to ion transport and cytoskeletal assembly might be implicated in both processes: They play a role in renal and vascular regulation of salt–water balance and blood pressure, and in innate and adaptive inflammatory immune responses [8, 9]. On the other hand, the phenotypic spectrum of subtle genetic variations in known hypertension-related genes [10, 11] is far from being completely elucidated. In this study, we focused our attention upon three single-nucleotide polymorphisms (SNPs) involving β-adducin (ADD2 gene) and the solute carrier family 8 (sodium–calcium exchanger) member 1 (SLC8A1 gene, from now on NCX1), namely rs4984 (ADD2), rs11893826 (NCX1a) and rs434082 (NCX1b). The adducin family comprises a set of three proteins (α-, β- and γ-adducin) that form α/β or α/γ dimers or tetramers. Adducins are expressed in a large number of tissues, where they play a fundamental role in cytoskeleton assembly and cell membrane stability. In the setting of the kidney, adducins have been linked: (a) functionally, to the control of the permeability of the glomerular barrier at the level of the podocyte and to the modulation of the Na+/K+ ATPase at the tubular level [12, 13]; (b) clinically, to impaired response to sodium load, increased hypertensive risk and decline of renal function [10, 12, 14]. The rs4984 variants in the ADD2 gene do not differ in terms of amino acid composition of beta-adducin, but can diverge in their susceptibility to alternative splicing [15]. The rs4984 SNP has been reported to affect systolic blood pressure and hypertension risk, possibly with a sexual dimorphism [14, 15]. Furthermore, it has been shown to affect the decline of renal function in a setting of immunemediated glomerular damage, such as IgA nephropathy, and
13
Rheumatol Int
to influence the response to angiotensin-converting enzyme inhibitors on the amount of proteinuria [12]. Accordingly, it seemed an ideal variable to be investigated in the context of an inflammatory disease with known female/male imbalance and frequent kidney involvement, such as SLE. In a similar way, we focused on NCX1. Sodium–calcium exchangers, such as NCX1, are involved in the modulation of the concentration of calcium in the intracellular space and depend on the sodium gradient provided by the Na+/ K+ ATPase [11]. In particular, NCX1 has a known role in: (a) affecting systolic blood pressure after acute sodium load [11]; (b) promoting inflammation through TNF release from monocytes and macrophages after enhanced calcium influx, while operating in the reverse mode [16]. The rs11893826 and rs434082 NCX1 SNP were previously linked to the response to sodium load (although their precise impact at the molecular level is not known), whereas their role in inflammation has not yet been explored. With these premises, we aimed at evaluating possible correlations between these known hypertension-related polymorphisms and development of SLE, LN and their complications.
Patients and methods Blood samples and clinical data were collected upon informed consent from 111 patients (100 females, 11 males) diagnosed with SLE without nephritis (SLEnonLN; n = 50), LN (n = 55) or SLE associated with isolated simple urinary sediment abnormalities in the absence of clinical signs of nephritis (SUA; n = 6) and 62 healthy controls (HC). All patients with SLEnonLN/LN/SUA met the 1997 ACR criteria for the classification of SLE. Clinical and laboratory information (including arthritis, skin involvement, hematological alterations, central nervous system involvement, hypertension, anti-cardiolipin, antiβ2GPI, anti-nuclear, anti-dsDNA, anti-Ro, anti-La, antiSm, anti-centromere, anti-Scl70 and anti-neutrophil cytoplasm autoantibodies, lupus-like anticoagulants (LLAC), complement levels, urinary sediment abnormalities, serum creatinine values, proteinuria, glomerular filtration rate) was retrospectively recorded upon informed consent. Data regarding renal function and inflammation profile in LN patients were recorded at LN onset, at 6 and 18 months and at the end of the follow-up. Outcome measures and in particular complete remission (CR) of patients with LN were calculated according to the EULAR/ACR guidelines for the diagnosis and treatment of LN [17]. The median
Rheumatol Int
duration of the follow-up for SLEnonLN/LN patients was 87.24 months (range 15–519 months). After DNA extraction from peripheral blood with standard methods, patients and controls were genotyped for polymorphisms in NCX1 and ADD2 genes. In particular, we analyzed the relative frequencies of rs11893826 (NCX1a) and rs434082 (NCX1b) SNPs, both located among the first intronic sequence of SLC8A1, and of the rs4984 (ADD2) SNP, located in the coding sequence of the beta-adducin gene on the alternatively spliced exon 15. Genotyping was performed with the 5′ nuclease allelic discrimination assays, with allelespecific MGB probes (TaqMan SNP Genotyping Assay, Applied Biosystems, Foster City, CA, USA). Genotype frequencies were compared with reference Caucasian and Italian (Tuscan) populations [18] (Table 1). Other clinical characteristics of patients with SLEnonLN/LN are given in Table 2. The study conformed to the declaration of Helsinki and was approved by the research ethics committee of our hospital. Statistical analysis Given the small sample size, lupus patients with SUA were excluded from analysis when comparing lupus subgroups one with each other as well as with other groups, but were included in the analysis when the lupus cohort was considered as a whole (WLC). Chi-square test was employed to compare genotype frequencies and other qualitative variables among groups and lupus subgroups. Kaplan–Meier survival curves were built to evaluate the incidence of LN, nephritis relapse and the rates of CR among different genotypes in the lupus cohort during follow-up. Log-rank test was used to compare differences in survival rates among different groups of patients. Potential confounders (such as gender, age of SLE onset, hypertension and induction therapy for LN) were included as covariates during binary logistic regression analysis and Cox regression analysis in order to test for interaction among ADD2 and NCX1 polymorphisms and development of SLE, LN and their complications. Generalized linear models were employed to compare quantitative variables such as 24-h proteinuria and serum creatinine among genotype subgroups during follow-up.
Results ADD2 The relative frequencies of ADD2 genotypes in patients did not deviate from Hardy–Weinberg equilibrium (p = 0.24). They were in line with those reported in the reference populations (Table 1) and similar among the WLC and HC. No
subject carried the rare TT genotype. Lupus patients carrying the CT genotype were protected from the development of LN during follow-up (recessive model for C allele: logrank 5.502; p = 0.019; HR = 2.571 with 95 % CI 1.15– 5.88; Fig. 1). Patients with CT genotype had also significantly lower prevalence of skin manifestations (recessive model for C allele: χ2 = 5.488; p = 0.027; OR = 3.012 with 95 % CI 1.20–7.69; Table 3). These phenomena were apparently independent of gender and age at SLE onset. Nonsignificant differences among ADD2 genotypes were detected with respect to serum creatinine, 24-h proteinuria, complement factors C3 and C4, incidence of hypertension and clinical outcome in LN patients. Patients with the ADD2 CC | NCX1a AA/GA genotypes had a significantly higher prevalence of anti-cardiolipin antibodies, irrespectively of the presence of concomitant anti-phospholipid syndrome (χ2 = 7.211; p = 0.007; OR = 3.176 with 95 % CI 1.35–7.48; Table 3). Differences regarding anti-β2GPI antibodies and LLAC were not significant. In the LN subgroup, the ADD2 CC | NCX1a AA/GA genotype is also associated with higher levels of complement factor C4 between 6 and 18 months from the onset of nephritis, independently of the induction therapy (F = 5.733; p = 0.023; N = 33, Table 4). No additional phenotypic differences were observed with respect to the ADD2 SNP among SLEnonLN and LN patients. NCX1a and NCX1b Considering the whole studied population, the relative frequencies of NCX1a and NCX1b genotypes did not deviate from Hardy–Weinberg equilibrium (p = 0.23 and p = 0.70, respectively). NCX1b genotype frequencies were in line with those observed in the reference populations. NCX1a genotype frequencies were also in line with those observed in the reference populations, with the exception of the group of HC (Table 1). In general, no significant differences were observed among lupus patients and HC. The NCX1a AA genotype tended to be more represented in the group of patients with lupus and was more frequent in the WLC when compared to HC (χ2 = 6.266; p = 0.044), although this difference could be attributable to the unexpectedly low frequency of NCX1 AA in HC (as compared to the reference populations), possibly as a consequence of low sample size. Differences observed among NCX1a genotypes in terms of CR from LN were not significant. Patients with NCX1 AA and AG genotypes had increased risk of relapse during follow-up (dominant model for A allele: log-rank 16.082; p
Rheumatology INTERNATIONAL
ORIGINAL ARTICLE - GENES AND DISEASE
Beta‑adducin and sodium–calcium exchanger 1 gene variants are associated with systemic lupus erythematosus and lupus nephritis Giuseppe A. Ramirez1,2 · Chiara Lanzani3 · Enrica P. Bozzolo1 · Laura Zagato3 · Lorena Citterio3 · Nunzia Casamassima3 · Valentina Canti1,2 · Maria Grazia Sabbadini1,2 · Patrizia Rovere‑Querini1,2 · Paolo Manunta2,3 · Angelo A. Manfredi1,2
Received: 19 January 2015 / Accepted: 25 May 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Genetic research in systemic lupus erythematosus (SLE) is rapidly developing, and numerous sets of genes are being associated with specific clinical subphenotypes in the setting of SLE. On the other hand, basic science studies are revealing strong connections between salt–water balance and inflammation. The aim of this study was to evaluate whether variants of genes known to influence the individual susceptibility to hypertension also influence the renal function in a cohort of SLE patients with and without lupus nephritis (LN). This study is a case–control study with candidate gene approach. A total of 111 patients with SLE (50 with SLE without nephritis, 55 with LN and 6 with simple urinary sediment abnormalities) and 62 healthy controls (HC) were genotyped for NCX1 rs11893826 (NCX1a) and rs434082 (NCX1b) and ADD2 rs4984 SNPs. Patients with ADD2 CT genotype were protected from LN and skin involvement; ADD2 CC | NCX1a AA/AG genotypes were associated with the presence of anti-cardiolipin antibodies; NCX1a AA genotype was slightly more frequent in lupus patients than in HC and associated with relapse risk and higher creatinine in patients with LN. NCX1b GG patients with LN had increased chances to reach complete remission. NCX1b GG | NCX1a GG genotype is associated with joint involvement. ADD2 and NCX1 variants influence the risk and the clinical features of SLE and LN, highlighting
* Giuseppe A. Ramirez [email protected] 1
Unit of Internal Medicine and Immunology, IRCCS Ospedale San Raffaele, Via Olgettina 60, 20132 Milan, Italy
2
Università Vita-Salute San Raffaele, Milan, Italy
3
Unit of Nephrology, IRCCS Ospedale San Raffaele, Milan, Italy
their potential role in regulating systemic inflammation and/or the local response to immune-mediated injury. Keywords Systemic lupus erythematosus · Sodium– calcium exchanger 1 · Beta-adducin · Nephritis · Skin · Anti-phospholipid antibodies · Arthritis Abbreviations ADD2 Beta-adducin gene CR Complete remission HC Healthy controls LLAC Lupus-like anticoagulants LN Lupus nephritis NCX1 Solute carrier family 8 (sodium–calcium exchanger) member 1 gene OR Odds ratio SLE Systemic lupus erythematosus SLEnonLN SLE without LN SNP Single-nucleotide polymorphism SUA Simple urinary sediment abnormalities WLC Whole lupus cohort
Introduction Systemic lupus erythematosus (SLE) is a prototypic multiorgan autoimmune disease, characterized by polymorphic clinical features, arising from a complex and only partially known genetic background [1–3]. Renal disease in SLE occurs in 40–70 % of lupus patients and is responsible for a high burden of morbidity and mortality [4]. A first set of genes within the so-called MHC region in chromosome 6 (which includes TNF and complement factors besides HLA variants) has the strongest association with the risk of developing SLE in large cohorts [5], together
13
with non-MHC genes such as interferon regulatory factor 5 (IRF5), signal transducer activator of transcription 4 (STAT4) and integrin alpha-M (ITGAM) [6, 7]. A recent meta-analysis from the International Consortium for SLE Genetics provided evidence for a role of a wide number of genes in the development of lupus nephritis (LN) in Caucasian women with SLE. The authors found that only a small fraction of classical SLE susceptibility loci were also specifically associated with the development of LN and concluded that LN loci probably influence the local (kidney) response to the thrust posed by the immune system, rather than acting systemically on the immune response [2]. Thus, structural genes with no evident immune function could be implicated in the susceptibility to SLE and to specific SLE phenotypes. Blood pressure control and inflammation are composite and apparently unrelated physiological functions that depend on the multi-organ cooperation of several genes. Cellular machineries devoted to ion transport and cytoskeletal assembly might be implicated in both processes: They play a role in renal and vascular regulation of salt–water balance and blood pressure, and in innate and adaptive inflammatory immune responses [8, 9]. On the other hand, the phenotypic spectrum of subtle genetic variations in known hypertension-related genes [10, 11] is far from being completely elucidated. In this study, we focused our attention upon three single-nucleotide polymorphisms (SNPs) involving β-adducin (ADD2 gene) and the solute carrier family 8 (sodium–calcium exchanger) member 1 (SLC8A1 gene, from now on NCX1), namely rs4984 (ADD2), rs11893826 (NCX1a) and rs434082 (NCX1b). The adducin family comprises a set of three proteins (α-, β- and γ-adducin) that form α/β or α/γ dimers or tetramers. Adducins are expressed in a large number of tissues, where they play a fundamental role in cytoskeleton assembly and cell membrane stability. In the setting of the kidney, adducins have been linked: (a) functionally, to the control of the permeability of the glomerular barrier at the level of the podocyte and to the modulation of the Na+/K+ ATPase at the tubular level [12, 13]; (b) clinically, to impaired response to sodium load, increased hypertensive risk and decline of renal function [10, 12, 14]. The rs4984 variants in the ADD2 gene do not differ in terms of amino acid composition of beta-adducin, but can diverge in their susceptibility to alternative splicing [15]. The rs4984 SNP has been reported to affect systolic blood pressure and hypertension risk, possibly with a sexual dimorphism [14, 15]. Furthermore, it has been shown to affect the decline of renal function in a setting of immunemediated glomerular damage, such as IgA nephropathy, and
13
Rheumatol Int
to influence the response to angiotensin-converting enzyme inhibitors on the amount of proteinuria [12]. Accordingly, it seemed an ideal variable to be investigated in the context of an inflammatory disease with known female/male imbalance and frequent kidney involvement, such as SLE. In a similar way, we focused on NCX1. Sodium–calcium exchangers, such as NCX1, are involved in the modulation of the concentration of calcium in the intracellular space and depend on the sodium gradient provided by the Na+/ K+ ATPase [11]. In particular, NCX1 has a known role in: (a) affecting systolic blood pressure after acute sodium load [11]; (b) promoting inflammation through TNF release from monocytes and macrophages after enhanced calcium influx, while operating in the reverse mode [16]. The rs11893826 and rs434082 NCX1 SNP were previously linked to the response to sodium load (although their precise impact at the molecular level is not known), whereas their role in inflammation has not yet been explored. With these premises, we aimed at evaluating possible correlations between these known hypertension-related polymorphisms and development of SLE, LN and their complications.
Patients and methods Blood samples and clinical data were collected upon informed consent from 111 patients (100 females, 11 males) diagnosed with SLE without nephritis (SLEnonLN; n = 50), LN (n = 55) or SLE associated with isolated simple urinary sediment abnormalities in the absence of clinical signs of nephritis (SUA; n = 6) and 62 healthy controls (HC). All patients with SLEnonLN/LN/SUA met the 1997 ACR criteria for the classification of SLE. Clinical and laboratory information (including arthritis, skin involvement, hematological alterations, central nervous system involvement, hypertension, anti-cardiolipin, antiβ2GPI, anti-nuclear, anti-dsDNA, anti-Ro, anti-La, antiSm, anti-centromere, anti-Scl70 and anti-neutrophil cytoplasm autoantibodies, lupus-like anticoagulants (LLAC), complement levels, urinary sediment abnormalities, serum creatinine values, proteinuria, glomerular filtration rate) was retrospectively recorded upon informed consent. Data regarding renal function and inflammation profile in LN patients were recorded at LN onset, at 6 and 18 months and at the end of the follow-up. Outcome measures and in particular complete remission (CR) of patients with LN were calculated according to the EULAR/ACR guidelines for the diagnosis and treatment of LN [17]. The median
Rheumatol Int
duration of the follow-up for SLEnonLN/LN patients was 87.24 months (range 15–519 months). After DNA extraction from peripheral blood with standard methods, patients and controls were genotyped for polymorphisms in NCX1 and ADD2 genes. In particular, we analyzed the relative frequencies of rs11893826 (NCX1a) and rs434082 (NCX1b) SNPs, both located among the first intronic sequence of SLC8A1, and of the rs4984 (ADD2) SNP, located in the coding sequence of the beta-adducin gene on the alternatively spliced exon 15. Genotyping was performed with the 5′ nuclease allelic discrimination assays, with allelespecific MGB probes (TaqMan SNP Genotyping Assay, Applied Biosystems, Foster City, CA, USA). Genotype frequencies were compared with reference Caucasian and Italian (Tuscan) populations [18] (Table 1). Other clinical characteristics of patients with SLEnonLN/LN are given in Table 2. The study conformed to the declaration of Helsinki and was approved by the research ethics committee of our hospital. Statistical analysis Given the small sample size, lupus patients with SUA were excluded from analysis when comparing lupus subgroups one with each other as well as with other groups, but were included in the analysis when the lupus cohort was considered as a whole (WLC). Chi-square test was employed to compare genotype frequencies and other qualitative variables among groups and lupus subgroups. Kaplan–Meier survival curves were built to evaluate the incidence of LN, nephritis relapse and the rates of CR among different genotypes in the lupus cohort during follow-up. Log-rank test was used to compare differences in survival rates among different groups of patients. Potential confounders (such as gender, age of SLE onset, hypertension and induction therapy for LN) were included as covariates during binary logistic regression analysis and Cox regression analysis in order to test for interaction among ADD2 and NCX1 polymorphisms and development of SLE, LN and their complications. Generalized linear models were employed to compare quantitative variables such as 24-h proteinuria and serum creatinine among genotype subgroups during follow-up.
Results ADD2 The relative frequencies of ADD2 genotypes in patients did not deviate from Hardy–Weinberg equilibrium (p = 0.24). They were in line with those reported in the reference populations (Table 1) and similar among the WLC and HC. No
subject carried the rare TT genotype. Lupus patients carrying the CT genotype were protected from the development of LN during follow-up (recessive model for C allele: logrank 5.502; p = 0.019; HR = 2.571 with 95 % CI 1.15– 5.88; Fig. 1). Patients with CT genotype had also significantly lower prevalence of skin manifestations (recessive model for C allele: χ2 = 5.488; p = 0.027; OR = 3.012 with 95 % CI 1.20–7.69; Table 3). These phenomena were apparently independent of gender and age at SLE onset. Nonsignificant differences among ADD2 genotypes were detected with respect to serum creatinine, 24-h proteinuria, complement factors C3 and C4, incidence of hypertension and clinical outcome in LN patients. Patients with the ADD2 CC | NCX1a AA/GA genotypes had a significantly higher prevalence of anti-cardiolipin antibodies, irrespectively of the presence of concomitant anti-phospholipid syndrome (χ2 = 7.211; p = 0.007; OR = 3.176 with 95 % CI 1.35–7.48; Table 3). Differences regarding anti-β2GPI antibodies and LLAC were not significant. In the LN subgroup, the ADD2 CC | NCX1a AA/GA genotype is also associated with higher levels of complement factor C4 between 6 and 18 months from the onset of nephritis, independently of the induction therapy (F = 5.733; p = 0.023; N = 33, Table 4). No additional phenotypic differences were observed with respect to the ADD2 SNP among SLEnonLN and LN patients. NCX1a and NCX1b Considering the whole studied population, the relative frequencies of NCX1a and NCX1b genotypes did not deviate from Hardy–Weinberg equilibrium (p = 0.23 and p = 0.70, respectively). NCX1b genotype frequencies were in line with those observed in the reference populations. NCX1a genotype frequencies were also in line with those observed in the reference populations, with the exception of the group of HC (Table 1). In general, no significant differences were observed among lupus patients and HC. The NCX1a AA genotype tended to be more represented in the group of patients with lupus and was more frequent in the WLC when compared to HC (χ2 = 6.266; p = 0.044), although this difference could be attributable to the unexpectedly low frequency of NCX1 AA in HC (as compared to the reference populations), possibly as a consequence of low sample size. Differences observed among NCX1a genotypes in terms of CR from LN were not significant. Patients with NCX1 AA and AG genotypes had increased risk of relapse during follow-up (dominant model for A allele: log-rank 16.082; p
