Studio di Geni Candidati associati a Nefrolitiasi nella Coorte Incipe.

Kidney stones disease (Nephrolithiasis) is a complex problem of worldwide prevalence that is influenced by both genetic and environmental factors. Kidney stones (KS) are predominantly composed by calcium oxalate (∼65%) but may also contain calcium phosphate (∼10%), uric acid (∼15%), magnesium ammonium phosphate (∼10%), cystine (∼1%), 2,8-dihydroxyadenine (<1%), xanthine (<1%), or excreted drugs such as indinavir (<1%). Stones are formed in the urine as a result of an imbalance in the relative concentration of lithogenic substances such as calcium or oxalate, compared to stone formation inhibitors, such as citrate or magnesium, leading to the precipitation and aggregation of crystals. Therefore, nephrolithiasis is often associated with metabolic abnormalities in urinary solute concentration or decreased urinary solubility; these include hypercalciuria, hyperoxaluria, hypocitraturia, hyperuricosuria, cystinuria, low urinary volume and urinary acidification defects. The etiology of kidney stones and metabolic abnormalities is multifactorial involving genetic and environmental factors. The monogenetic disorders of nephrolithiasis can be classified into two categories which are those associated with calcium containing stones and those which do not contain calcium. We focused our attention on 10 genes associated with calcium-containing stones monogenetic disorders and responsible for the following clinical manifestations: Distal renal tubular acidosis (genes ATP6V1B1-chromosome 2; SLC4A1-chromosome 17; ATP6V0A4-chromosome 7); Hereditary hypophosphatemic rickets with hypercalciuria (genes: XPR1- chromosome 1; SLC34A1-chromosome 5; SLC20A2- chromosome 8; SLC34A3- chromosome 9; SLC9A3R1-chromosome 17); Infantile Hypercalcaemia (genes: SLC34A1; FGF23-chromosome 12; CYP24A1-chromosome 20); Fanconi Syndrome (XPR1). The Incipe cohort (January- 2006) consisted of ~4000 patients all Caucasians, ≥40-years old randomly chosen from the lists of patients of 62 randomly selected general practitioners based in four geographical areas in the Veneto region, NE Italy. Clinical and genetic information were collected from each participant. The clinic of the Incipe Cohort was evaluated with a preliminary descriptive analysis of the variables, considering their distribution, the degree of correlation for the quantitative variables (including glucose, urea, creatinine, uric acid, cholesterol, triglycerides, insulin, albumin, phosphate, potassium, calcium, sodium, chlorine, weight, height) and the chi-squared for the qualitative variables (among these we considered the Kidney Stones variable as our outcome. Answer: YES / NO). The non-parametric Mann-Whitney test for continuous variables allowed us to discriminate the statistically significant variables associated with the kidney stone event. Subsequently these variables have been also evaluated through GLM (Generalized Linear Model) including the most relevant covariates; a nominal p-value ≤ 0.05 was chosen as the cut off. For the candidate gene investigation, on the other hand, the patients were studied with two panels: a sub-cohort of 893 patients (Incipe 1) were analyzed with HumanOmniExpress-12 array (the included genomic probes are scattered across the entire genome sequence); while a sub-cohort of 2153 patients (Incipe 2) were analyzed with Humancoreexome-12v1 (most of the included genomic probes map on exonic regions); both are Illumina technology. Through imputation of the genotype, it was possible to have a dense genetic profile for each individual, that can then was used to search for risk variant associated to kidney stones. For the imputation we used TOPMed (Trans-Omics for Precision Medicine) reference panel that is useful for whole genome sequencing samples. It has the advantage of making full use of key genetic characteristics such as linkage patterns (or the ordering of genes on chromosomes), mutations and recombination hotspots. Since the data were imputed, the association analysis of the single marker was conducted by estimating the relative abundance (parameter DS - Estimate of the alternative dosage), for each genotype, of the alternative allele between cases and controls. DS is the probability of being heterozygous plus the probability of being homozygous dominant or recessive - P (0 | 1) + 2*P (1 | 1). The shrinked t-test was used to test the association of each genetic variant with KS and controlling for multiple testing. The significant risk variants (p- value ≤ 0.05) were also tested GLM including the covariates for KS found in the steps described above, one at a time. The dataset of clinical information of the Incipe cohort is organized in 55 quantitative variables, both discrete and continuous, and 49 qualitative. For the more correlated quantitative variables with KS, a non-parametric test (Mann-Whitney) was applied. The most significant ones were explored with a GLM model that showed a statistically significant p-value for uric acid (0.00209), insulin (0.03596), phosphate (0.02621) and potassium (0.03586). These results confirmed what is already known about the etiology of the pathology. As regards the candidate gene panels, a total of 59,302 variants (assayed or imputed) were studied for the 10 genes under study. Incipe 1 was used to discovery panel and Incipe 2 to replicate the genetic associations. Incipe 1 and Incipe 2 samples were therefore analyzed independently. In both cases, the shrinked t test highlighted variables differentially expressed between the cases (stone formers) and the controls (no stone formers). These significant variants were annotated (this consists in mapping the genes and identifying the biological characteristics of the variants) with Ensembl Vep (hg38) and studied with both simple and multivariate GLM models, thus associating uric acid, insulin, phosphate, and potassium; significant variables resulting from the analysis of the clinical dataset. In Incipe 1, some of the 69 variants found to be significant, have already been described in other studies: 6 variants for ATP6V0A4 gene (according to annotation, 4 variants are related to ammonium metabolism, 1 variant to acidification by the renal tubule and 1 to risk of alcohol addiction), 5 variants for the CYP24A1 gene (all of them already reported in the literature because associated with other phenotypes; 4 concerning the metabolism of vitamin D and 2 associated with hypertension). Also in Incipe 2, some of the 18 significant variants, have already been described in other studies: 1 variant of the SLC34A1 gene (associated to Idiopathic Infantile Hypercalcemia); 1 variant in SLC34A3 gene (associated to hypophosphatemic rickets); 3 variants of the CYP24A1 gene (related to the metabolism of vitamin D). The variants chr20:54171755:C:A and chr20:54173157:G:A, mapping on the gene CYP24A1 are the only two statistically significant in both Incipe 1 and Incipe 2 panels. CYP24A1 is one of the gene involved in the vitamin D pathway; it belongs to hydroxylases from CYP450 family. Both variants were found to be significant with multivariate tests (the risk alleles are the adenines - A - for both variants), associating them with the variables found to be significant in the clinic. Further investigations by genotyping analysis are needed to confirm the results obtained.