Role of haptoglobin 2-2 genotype on disease progression and mortality among South Indian chronic kidney disease patients

Haptoglobin (HP), a plasma glycoprotein, binds to free hemoglobin and prevents the loss of iron and kidney damage. The variations of HP gene affect its enzyme activity, resulting in varied antioxidant, angiogenic and anti-inflammatory properties. HP 2-2 genotype showed 3.84 fold increased risk for the development of CKD in Taiwan population. With this background, the present work focused to conduct a prospective case-control study in South Indian population to evaluate whether the HP variants are associated to nondialysis (ND) (CKD stages 1–4) and ESRD (CKD stage 5) conditions. Totally 392 CKD patients (nondialysis, ND; n = 170, end-stage renal disease, ESRD; n = 222) and 202 healthy individuals were enrolled. The blood samples collected from the patients were used to determine biochemical parameters and HP genotyping. Gene frequency and biochemical parameters were statistically analyzed for disease association. Results showed that HP 2-2 genotypes were significantly associated with ND and ESRD disease development compared to controls. Higher HP2-2 genotype frequency showed an increased hazard ratio for overall disease progression among ND patients (hazard ratio = 3.86; 95% CI 1.88 to 7.93; P = 0.0002). Survival analysis also showed that non-HP2-2 patients have a statistically significantly decreased risk for mortality compared to patients with the HP2-2 genotype (ESRD patients hazard ratio = 4.05; P = 0.04). The present study confirms that HP2-2 polymorphism is statistically associated with the risk of CKD incidence, progression, and mortality among South Indians. Concluding our results, the HP2-2 genotype could be an independent predictor of all-cause mortality and disease progression in patients with CKD.


Introduction
Chronic kidney disease (CKD) is becoming a leading health problem prevalent globally, resulting in adverse effects such as the progressive decline in kidney function, development of cardiovascular disease (CVD), and premature death even in early stages [1]. CKD is commonly interrelated with permanent pathological changes within the kidney. This pathological change has an intricate correlation with other diseases, including hypertension (HT) and diabetes mellitus (DM) [2,3]. The higher incidence rate of CKD was recorded in South Asia [4] and 33% of global CKD patients belong to two countries, namely China (132.3 million) and India (115.1 million) [5].
Over the past several years, many studies reported that the genetic variants of haptoglobin (HP) predict the early onset of renal [6,7] and cardiovascular diseases (CVD) [8] in diabetic patients. HP is a plasma glycoprotein mainly produced by the liver. The primary biological function of HP is to bind free haemoglobin (Hb), preventing the loss of iron and subsequent kidney injury. Genetic polymorphism of HP gene resulted in three variants (HP 1-1, HP 2-1, HP  which are determined by two alleles (HP1&HP2). Compared to the HP 2-2 variant, HP 1-1 variant has higher antioxidant and anti-inflammatory properties, whereas HP 2-2 has higher angiogenic potential than HP 1-1 [9]. HP 2-1 exhibits relatively an average activity to both genotypes. The HP2 alleletranslated protein product has diminished binding capacity to hemoglobin [9], that may lead to iron overload during hemolysis and is linked with kidney injury [10][11][12].
The rate of HP allele occurrence differs globally, and a higher HP2 allelic frequency was observed in Asia compared to the West. HP1 allelic frequency differs from seven percentages in parts of India to more than seventy percentages in parts of West Africa and South America [10,13]. Chen et al. reported that HP 2-2 genotype is linked with susceptibility to CKD [9]. Recent study showed a diminished plasma haptoglobin levels in patients with acute kidney injury (AKI) when compared to patients without AKI in acute respiratory distress syndrome (ARDS) [14]. From this background, we hypothesized that HP2 allele has a role in CKD progression due to inferior ability to bind with hemoglobin and lesser antioxidant property. The current prospective case-control study is focused on the association of HP genotype to nondialysis (ND) and end stage renal disease (ESRD) among the South Indian population. This study followed up 392 CKD patients and analyzed disease progression and mortality using survival analysis. Correlation studies were done with genotypes and biochemical parameters. Additionally, multivariate Cox proportional hazard regression analysis was employed to identify a leading cause of mortality and disease progression.

Study population and serum parameters
Internal Research and Review Board (IRB), Ethical Clearance, Biosafety and Animal Welfare Committee of Madurai Kamaraj University, approved the study (IRBEC-2012/05/ HG09, approved on 04.05.2012). Informed consent was obtained from all study subjects before participating in the study. Venous blood from 392 CKD patients (ND; n = 170, ESRD; n = 222) was collected from Meenakshi Mission Hospital and Research Centre (MMHRC), Madurai, South India, between May 2012 and December 2012. Concurrently blood samples were collected from 202 control subjects, along with their family history in the absence of any diabetes, hypertension, and chronic kidney diseases. Glomerular filtration rate (GFR) was determined by the 'Modification of Diet in Renal Disease' (MDRD) calculator [12]. CKD stages were confirmed and classified into ND (CKD stages 1-4) and ESRD (CKD stage 5) using the calculated GFR. Disease progression in ND patients was measured by the number of patients moved from ND stage to ESRD stage during the follow up of 70 months. There were 36 deaths noted down among the ESRD patients during the follow up of 70 months. Basic demographic information such as age, gender, body mass index (BMI), and the duration of CKD as well as a history of diabetes and hypertensive illnesses were recorded. Serum biochemical parameters including the HbA1c, creatinine, urea, calcium, haemoglobin and phosphorous levels were estimated.

DNA isolation and determination of HP genotype
Extraction of DNA from the blood leukocytes was done using the standard phenol/chloroform method [13]. Isolated DNA was stored at − 20°C and HP genotyping was done as reported previously [14]. In brief, in protocol 1: The oligos A (5′-GAG GGG AGC TTG CCT TTC CATTG-3′) and B (5′-GAG ATT TTT GAG CCC TGG CTGGT-3′) were used to analyze both HP1-allele (1,757 bp) and HP2-allele sequences (3,481 bp). In protocol 2: The oligos C (5'CCT GCC TCG TAT TAA CTG CAC CAT -3′) and D (5′-CCG AGT GCT CCA CAT AGC CATGT-3′) were used to analyse the HP2-allele specific sequence of 349 bp. The PCR amplification was carried out for 35 cycles using primers A and B with steps of initial denaturation at 95 °C-5 min, cycle denaturation at 95 °C-1 min, annealing at 64 °C-45s extension at 72 °C-90s and final extension at 72 °C-10 min and finally stored at 4 °C for infinite time. Amplified products (HP1allele:1757 bp, HP2-allele: 3481 bp) were resolved on 2% agarose gel. HP1-1 appeared as a single band of 1,757 bp, HP2-2 as a single band of 3,481 bp, and HP2-1 by both of these bands ( Supplementary Fig. 1). The HP2 (3,481 bp) band generally appears lighter than the HP1 (1757 bp) band, may lead to misinterpretation in determining the HP2-1 genotype. To overcome this, all the samples were reconfirmed by performing PCR using protocol 2 with another set of oligonucleotide primers C & D, which have specific binding site for only the HP2 allele to confirm the presence of the HP2 band. For this protocol (protocol 2) the PCR amplification was carried out with steps, initial denaturation at 95 °C-5 min, cycle denaturation at 95 °C-45s, annealing at 69 °C-40s, extension at 72 °C-30s, final extension at 72 °C-10 min. The triplicates were performed, and every sample was undergone to both protocol 1 and protocol 2 and to confirm the results. The amplicons were assessed by 2% agarose gel electrophoresis with reference to 1 kb DNA ladder ( Supplementary Fig. 2).

Statistical analysis
Data of continuous variables were expressed as mean ± standard deviation (SD) and data of noncontinuous variables as frequency (N, %). Categorical variables were presented using frequency counts and compared by χ 2 -test. The genetic variants and their risk for disease were computed by logistic regression analysis using odds ratios (OR) and 95% confidence intervals (CI). The Kaplan-Meir method was employed to analyse survival curves and differences between patients with HP 2-2 group and non HP 2-2 group were compared by the log-rank test. The relationship between HP variants with all-cause mortality and all-cause disease progression were analyzed using Weibull regression analysis.

Clinical and demographic data
The basic demographic and clinical characteristics of the patients and control subjects are presented in Supplementary  Table 1. Creatinine and urea, the key indicators for CKD, increased in both patients' populations compared to the control population. Level of glycosylated hemoglobin (HbA1c, a test to determine diabetes) was also elevated in both patients' populations. Table 1 shows the comparison of ND patients' clinical characteristics between Non-HP2-2 and HP2-2 groups. Results showed that serum haemoglobin level significantly (P = 0.03) differs between both groups. Established risk factors for CKD, including diabetes (P = 0.04) and hypertension (P = 0.02), also showed an association with the HP2-2 genotype.

HP genetic variants and genotype frequencies
The HP genotype frequency of the study participants is presented in Supplementary Table 2. χ 2 analysis was employed to compare the genotype frequency differences among the three groups. For further analysis, study participants (Control, ND, and ESRD) were separated into two groups based on HP genotype: participants with HP2-2 genotype (HP2-2 group) and participants with a non-HP2-2 genotype (HP 1-1 + HP 2-1 genotype; non-HP2-2 group). The distribution of HP variants in ND and control subjects (ND:χ 2 = 1.03, P = 0.307; control:χ 2 = 3.45, P = 0.06) were consistent with Hardy-Weinberg equilibrium. Whereas, in ESRD patients the distribution of HP variants was inconsistent with Hardy-Weinberg equilibrium (HP: χ 2 = 5.01, P = 0.03) (Supplementary Table 2).

Relationship of HP variants with ND and ESRD
To analyse the relationship of HP variants with CKD, we carried out a comparison analysis between control and CKD population (ND + ESRD) ( Table 3). Logistic regression analysis revealed no individual association between the genotypes HP1-1 vs. HP2-1(OR 1.33; 95% CI 0.50 to 3.52; P = 0.56) and HP1-1 vs. HP2-2 (OR 1.93; 95% CI 0.76 to 4.85; P = 0.16) for CKD development. However, statistically significant association obtained for CKD development between HP2-2 group and non HP2-2 group (OR 1.51; 95% CI 1.02 to 2.23; P = 0.04). Table 2. represents the regression analysis of HP genotype distribution of ND and ESRD population to analyse the risk of ESRD. Results indicated that there was no association existed between the genotypes HP 1-1 vs. HP 2 − 1 (OR 0.86; 95% CI 0.23 to 3.26; P = 0.84) and HP1-1 vs. HP2-2 (OR 1.45; 95% CI 0.41 to  Table 5). This data suggests that HP 2-2 genotype could be a risk factor for CKD and ESRD development.

Kaplan-Meir analysis of ND and ESRD population
Kaplan-Meir analysis was employed to analyse the relationship between outcome of the study (mortality and disease progression) and HP 2-2 genotype. During the study period with the mean follow-up of 36.67 months, 36 deaths happened among the ESRD population. Non-HP2-2 and HP2-2 groups showed 5.26% and 18.47% mortality rates, respectively. Survival analysis of ESRD Non-HP2-2 patients showed a significantly decreased risk for mortality compared to patients with the HP2-2 genotype with a hazard ratio of 4.05 (P = 0.04) (Fig. 1a). We also employed Kaplan-Meier analysis to assess if HP2-2 genotype could impact the disease progression from ND to ESRD. Across the study period, 38 ND subjects progressed from ND to ESRD. Out of these 38 patients, 26.77% of the people with HP2-2 genotypes and 9.30% were with non-HP2-2 genotype which could confirm that the HP2-2 genotype is a critical risk factor for disease progression than Non-HP2-2 group (P = 0.04). Log-rank test results revealed 2.86 fold increased risk for HP2-2 group than Non-HP2-2 group for disease progression (Fig. 1b). In overall, the data suggested that the HP 2-2 genotype might cause mortality and disease progression in ESRD and ND patients, respectively.

Weibull regression analysis
The sole influence of HP polymorphism on the leading cause for mortality in the ESRD population and the disease progression in the ND population were estimated using Weibull regression analysis with six more parameters (

Discussion
Under certain circumstances ROS are primarily formed in the kidneys by the action of mitochondrial respiratory chain and upregulated expression of enzymes such as NADPH oxidases which are primarily responsible for poor vascular function, and fibrosis by mediating oxidative stress [15]. As haptoglobin is an innate antioxidant involves in the scavenging of free haemoglobin and prevents iron-mediated formation of free reactive oxygen species, numerous studies were conducted to reveal the association of HP genotype to CVD [8] and diabetic nephropathy in different ethnicity [16,17]. Chen et al. reported that HP 2-2 genotype is a modifier for CKD in patients with diabetes, hypertension, and dyslipidemia in the Taiwanese population [9]. In our study, to delineate specifically how the HP 2-2 genotype is associated to the CKD, we followed up the patients with different stages of kidney diseases (non-dialysis, ND; CKD stages 1-4 and end stage renal disease ESRD; CKD stage 5) for 70 months. The study found that HP2-2 variant frequency was higher in ESRD and ND patients than in control. The logistic regression analysis between ESRD and controls revealed that the HP2-2 genotype has a 2-fold increased hazard to develop ESRD. Nevertheless, our study did not find a significant correlation between ND and the control population. Multivariate Weibull regression revealed that HP 2-2 genotype has 4.42 fold increased hazard for all-cause mortality in ESRD population. HP2-2 genotype also showed a strong association with mortality among ESRD patients and disease progression among ND patients. In addition to these findings, Weibull regression analysis denoted that HP2-2 genotype has a 4-fold increased hazard for all-cause mortality in ESRD population. Thus, our study found a strong relationship to the ND and ESRD development for HP 2-2 genotype, which is consistent with the HP2-2 genotype that showed an association to the mortality among AKI patients in the US population [18] and association with CKD incidence in Taiwan population [9] [19].
Since the HP polymorphism was reported to be correlated with renal complications in different ethnicity and geographic locations, it is essential to validate the HP polymorphism in the Indian population due to the lowest  incidence of HP1 allele frequency as reported previously [10]. The result of the current study is the first report from India concerning the role of the HP gene in the ND disease progression and mortality among ESRD patients. HP polymorphism in control subjects, ND, and ESRD were genotyped in this study. Our results show that apart from the association of HP mutants with CKD, it also acts as an inducer in disease progression. The results also revealed a significant association of HP2-2 variant with CKD occurrence. Genome wide association study (GWAS) among 805 people (Chinese and Malay) showed that rs75444904/ kgp16506790 variant was robustly associated with urinary HP level. Furthermore, elevated urinary HP level is associated with increased risk of diabetic kidney disease progression. This study also reported that urinary HP level is higher in patients with HP1-1 and HP2-1 as compared to HP2-2 genotype [20]. Our study suggested that HP2-2 genotype showed a significant association with disease progression and mortality among South Indian population. This increased urinary Haptoglobin levels in HP1-1 and HP2-1 as compared to HP2-2 genotypes may probably be due to increase in in-situ expression in renal tissues. Diabetes and proteinuria are the established risk factors for the occurrence and progression of CKD. Several serum parameters are abnormally elevated in CKD patients. Hence, we investigated the correlation between genotypes and several clinical observations. Decreased haemoglobin levels significantly correlated with the HP2-2 variant in ND patients (Supplementary Table 2), which indicated a diminished activity of HP2-2 genotype involved in pathomechanisms underlying disease development and progression. Diabetes and hypertension also significantly correlated with HP2-2 variant denoting the relationship with established risk factors for CKD. Though, compared to other ethnicities, the lesser distribution of HP1 allelic frequency was found in the Indian population (Control vs. CKD: 0.16 vs. 0.12), the current study obtained a statistically significant HP 2-2 genotype association to the ESRD development. This study used multiple parameters to associate the HP 2-2 genotype with risk factors such as diabetes, hypertension, and hemoglobin during a long-term follow-up of up to 70 months. The study's limitations include the smaller size of the study samples and the possibility of a cryptic population substructure of the subjects. Besides, it could have been precise if the current study carried out the haptoglobin functional activity from each sample of participants. Though this study represents only CKD patients of the South Indian population, the study results still can be used as a guideline with different racial and ethnic categories for future studies. We conclude that HP 2-2 genotype may play a pivotal role in determining the CKD disease development in South Indian population. To further explore, the study results can be elaborated under different environmental conditional factors to understand the CKD pathology better.