| Objective In the first place, to reveal non-obstructive azoospermia(NOA) associated genetic risk factors, targeted resequencing technology, case-control studies and primary genotype-phenotype studies were performed to identify NOA related SNVs. Meanwhile, NOA associated SNV library was constructed for the next GWAS stage as well as application as genetic method in clinical NOA diagnoses.Methods NOA patients diagnosed in “Center for Reproductive Medicine of the First Hospital of Jilin University” from September, 2013 to December, 2014 were included into NOA group. Sperm donators registered in the Sperm Bank of Jilin Province from September, 2013 to December, 2014 were included into control group. G banding karyotype analysis and detection of azoospermia factor(AZF) microdeletion in Y chromosome has to be performed before grouping to ensure each subject carries normal karyotype and no AZF microdeletions. Subjects were also selected by personal information such as race, and medical investigation such as varicocele history. Targeted NOA associated genes were collected by referring to databases, such as OMIM, GENCODE, Ref Seq, Vega Genome Browser and Pubmed. Animal models, RNA/protein profile studies and statistical analysis have already applied and published to demonstrate their association with NOA. Capture probes were also prepared based on those NOA associated genes. Basic information such as ages were collected by professional questionnaires. Semen samples were collected by masturbation for semen analysis and biochemical assays of seminal plasma. Forearm superficial vein blood was collected for peripheral blood reproductive hormone level measurement. The exonic regions of target genes were resequenced by targeted high-throughput sequencing technology which consists of three stages including preparation of whole genome library, targeted sequencing and bioinformatic analysis. Case-control study was performed to select NOA associated SNVs and construct SNV library for next GWAS stage. The workflow of case-control study consists of primary selection of candidate SNVs, MAF comparisons between NOA group and control group, association study of SNVs, genotype analysis, haplotype analysis and association study of genotype and phenotypes. The construction of SNV library consists of selection of SNMs by Baylor’s pipeline, selection of SNMs with MAF as 0 in control group and description of the library.Results 1. 34 and 40 cases were included in NOA group and control group, respectively. 2. 467 NOA associated genes were included referring to the publications and databases. Capture probes were prepared based on these genes. 3. Significant differences were found in average ages, BMI index, sperm concentrations, serum FSH levels, serum LH levels, serum PRL levels, serum T levels, seminal fructose levels, seminal α-glucosidase activities, seminal zinc levels and seminal inhibin B levels between NOA group and control group(p<0.05). No significant differences were found in varicocele rates, semen volume and serum E2 levels between groups(p>0.05). 4. Large quantity of high quality outcomes were produced by targeted resequencing. 100% align rates are over 95%. The coverage rate with at least 20× sequencing depth are between 99.9-92%. 100% duplication rates are less than 20%. 5. Total 178966 variants were detected by sequencing of 74 cases. 65% variants are in intronic regions, 24% are within exonic regions, 10% are in regulatory regions and less than 1% locate in non-regulatory intergenic regions. The 178966 variants are divided into two types, SNV and Indel. Nonsynonymous, synonymous, stopgain, stoploss, splicing and unknown types are included in SNV. frameshift, non-frameshift, stoploss and unknown types are included in Indel. The rate of unknown types show significant difference between NOA group and control group(p<0.001). 6. 2391 exonic SNVs with at least 20× sequencing depth were selected from 178966 variants for case-control study. 7. Minor allele frequencies(MAF) of 2391 SNVs were compared between NOA group and control group. Meanwhile, Hardy-Weinberg equilibrium(HWE) of NOA group and control group were calculated. Significant differences of MAFs of 65 SNVs were found between groups(p<0.05). The distributions of 38 SNVs are in agreement with HWE(p>0.05). 8. Association studies of 38 SNVs and NOA were performed and corrected with age as the covariant. 18 SNVs show significant correlation with NOA(p<0.05). 5 have positive correlations while the other 13 have negative correlations with NOA. 9. The dominant and recessive model analysis of 18 NOA associated SNVs were performed and corrected with age as the covariant. The results show that significantly increased risks were found in dominant models of 3 SNVs. Significantly decreased risks were found in dominant models of 7 SNVs. Significantly increased risks were found in recessive models of 2 SNVs. Significantly decreased risks were found in recessive models of 5 SNVs. 10. The association assay was performed to study the correlation between SNV genotypes and NOA associated phenotypes. It is indicated that significant differences were found in the frequencies of genotypes of 8 SNVs between phenotypically normal and abnormal cases. Among these, dominant genotypes of MTRR c.537T>C(rs161870, OR: 3.231), MTRR c.1049A>G(rs162036, OR: 3.231) and PIWIL1 c.1580G>A(rs1106042, OR:3.592) have positive correlations with abnormally increased FSH levels; Recessive genotypes of ACE c.81C>T(rs4316, OR: 2.858), ACE c.471A>G(rs4331, OR: 2.858), ACE c.606G>A(rs4343, OR: 2.858) and ACE c.1665T>C(rs4362, OR:3.771) have positive correlations with abnormally increased PRL levels; Dominant genotype of PIWIL1 c.1580G>A(rs1106042) has positive correlations with abnormally decreased T and Inhibin B levels; Recessive genotype of TAF4 B c.1815T>C(rs1677016, OR:3.871) has positive correlations with abnormally decreased seminal α-glucosidase levels. 11. Haplotypic association results show that there are 2 haplotypic blocks in chromosome 1. Block1 consists of 3 SNVs and 2 haplotypes including AAG and TCA. The frequency of haplotype AAG in NOA group is significantly higher than that in control group(83.8% vs. 68.8%, p=0.033). The frequency of haplotype TCA in NOA group is significantly lower than that in control group(16.2% vs. 31.2%, p=0.033). Block 2 consists of 2 SNVs and 3 haplotypes including TT, CG and CT. The frequency of haplotype TT in NOA group is significantly higher than that in control group(86.8% vs. 68.8%, p=0.009). The frequency of haplotype CG in NOA group is significantly lower than that in control group(13.2% vs. 28.7%, p=0.022). No significant differences were found in the frequencies of haploytype CT between NOA group and control group(p>0.05). One haplotype block was identified in chromosome 5. The block consists of 2 SNVs and 2 haplotypes including TA and CG. The frequency of haplotype CG in NOA group is significantly higher than that in control group(19.1% vs. 7.5%, p=0.035). The frequency of haplotype TA in NOA group is significantly lower than that in control group(80.9% vs. 92.5%, p=0.035). One haplotype block was identified in chromosome 6. The block consists of 2 SNVs and 3 haplotypes including AG, GA and AA. The frequency of haplotype AG in NOA group is significantly higher than that in control group(94.1% vs. 77.5%, p=0.005). The frequency of haplotype GA in NOA group is significantly lower than that in control group(4.4% vs. 15%, p=0.033). No significant differences were found in the frequencies of haploytype AA between NOA group and control group(p>0.05). One haplotype block was identified in chromosome 17. The block consists of 4 SNVs and 3 haplotypes including TGAC、CAGT and TGAT. The frequency of haplotype CAGT in NOA group is significantly higher than that in control group(41.2% vs. 21.2%, p=0.009). The frequency of haplotype TGAC in NOA group is significantly lower than that in control group(51.5% vs. 76.2%, p=0.002). No significant differences were found in the frequencies of haploytype TGAT between NOA group and control group(p>0.05). 12. SNVs without significant differences in MAF between NOA group and control groups have to be further selected by Baylor’s pipeline. 73 SNMs were finally selected from 62376 candidate SNVs by Baylor’s pipeline. 13. 73 SNMs underwent further selection based on the MAF in control group. 42 SNMs with MAF as 0 in control group were finally selected. All 42 SNMs distribute among 39 SNV sites within 34 genes. 14. SNV library was constructed with 52 NOA non-negatively associated SNVs and 39 SNMs.(5/91) 5% of the library are positively associated with NOA.(20/91) 22% have significant differences in MAF between groups and no significant deviation from HWE, however no significant association with NOA.(27/91) 30% of the library have significant differences in MAF between groups, however have significant deviations from HWE.(39/91) 43% are SNMs.(1/91) 1% of the library could only be retrieved from HGMD database.(80/91) 88% of the library could only be retrieved from db SNP database.(6/91) 7% of the library could be retrieved from both HGMD and db SNP databases.(4/91) 4% of the library could be retrieved from neither HGMD nor db SNP databases.(57/91) 63% of the library are nonsynonymous variations.(34/91) 37% of the library are synonymous variations.Conclusions 1. 18 SNVs are significantly associated with NOA. 5 SNVs including MTRR c.537T>C(rs161870), MTRR c.1049A>G(rs162036), PIWIL1 c.1580G>A(rs1106042), TAF4 B c.1815T>C(rs1677016) and SOX10 c.927T>C(rs139884) increase the risk of NOA. KIF2 C c.531A>T(rs3795713), KIF2 C c.1345A>C(rs4342887), KIF2 C c.1500G>A(rs1140279), MAEL c.12T>C(rs2296837), MAEL c.121T>G(rs11578336), HLA-DRB1 c.227T>A(rs17884945), HLA-DPB1 c.292A>G(rs1042140), HLA-DPB1 c.313A>G(rs1042151), HLA-DPB1 c.315G>A(rs1042153), ACE c.81C>T(rs4316), ACE c.471A>G(rs4331), ACE c.606G>A(rs4343) and ACE c.1665T>C(rs4362) decrease the risk of NOA and are protective factors. 2. 3 SNVs, including MTRR c.537T>C(rs161870), MTRR c.1049A>G(rs162036) and PIWIL1 c.1580G>A(rs1106042) increase the risk of NOA in dominat models. 7 SNVs including KIF2 C c.531A>T(rs3795713), KIF2 C c.1345A>C(rs4342887), KIF2 C c.1500G>A(rs1140279), MAEL c.12T>C(rs2296837), MAEL c.121T>G(rs11578336) HLA-DPB1 c.313A>G(rs1042151) and HLA-DPB1 c.315G>A(rs1042153) decrease the risk of NOA in dominant models and are protective factors. TAF4 B c.1815T>C(rs1677016) and SOX10 c.927T>C(rs139884) increase the risk of NOA in recessive models. 5 SNVs including HLA-DPB1 c.292A>G(rs1042140), ACE c.81C>T(rs4316), ACE c.471A>G(rs4331), ACE c.606G>A(rs4343) and ACE c.1665T>C(rs4362) decrease the risk of NOA in recessive models and are protective factors. 3. The risks of NOA could be probably increased by abnormally increased FSH levels which mightbe induced by the dominant genotypes of MTRR c.537T>C(rs161870), MTRR c.1049A>G(rs162036) and PIWIL1 c.1580G>A(rs1106042). The risks of NOA could be probably decreased by abnormally increased PRL levels which mightbe induced by the recessive genotypes of ACE c.81C>T(rs4316), ACE c.471A>G(rs4331), ACE c.606G>A(rs4343) and ACE c.1665T>C(rs4362). The risks of NOA could be probably increased by abnormally decreased T and Inhibin B levels which mightbe induced by the dominant genotypes of PIWIL1 c.1580G>A(rs1106042). The risks of NOA could be probably increased by abnormally decreased seminal α-glucosidase levels which mightbe induced by the recessive genotype of TAF4 B c.1815T>C(rs1677016). 4. Two haplotype blocks were identified in chromosome 1. KIF2 C c.531A>T(rs3795713), KIF2 C c.1345A>C(rs4342887) and KIF2 C c.1500G>A(rs1140279) locate in block1, while MAEL c.12T>C(rs2296837) and MAEL c.121T>G(rs11578336) locate in block2. Genotypes of AAG in block1 and TT in block2 are risk factors of NOA. Genotypes of TCA in block1 and CG in block2 are protective factors of NOA. MTRR c.537T>C(rs161870) and MTRR c.1049A>G(rs162036) were identified with linkage disequilibrium in chromosome 5. Haplotype CG is a risk factor. Haplotype TA is a protective factor. HLA-DPB1 c.313A>G(rs1042151) and HLA-DPB1 c.315G>A(rs1042153) were identified with linkage disequilibrium in chromosome 6. Haplotype AG is a risk factor. Haplotype GA is a protective factor. ACE c.81C>T(rs4316), ACE c.471A>G(rs4331), ACE c.606G>A(rs4343) and ACE c.1665T>C(rs4362) were identified with linkage disequilibrium in chromosome 17. Haplotype CAGT is a risk factor. Haplotype TGAC is a protective factor. 5. SNV library consists of 91 SNVs including 52 NOA non-negatively associated SNVs and 39 SNMs. |
PDF Full Text Request