Clinical And Genetic Study Of Autoimmune Diabetes Mellitus Complicated With Osteoporosis

Background:Autoimmune diabetes mellitus(ADM)is a type of diabetes mellitus in which insulin deficiency and hyperglycemia occur due to the destruction of isletβcells caused by autoimmune dysfunction.ADM mainly includes type 1 diabetes(T1D)and latent autoimmune diabetes in adults(LADA).Osteoporosis(OP)is a metabolic disease that increases bone fragility and fracture risk due to decreased bone mass and microstructural damage of bone tissue.OP is a common comorbidity of T1D.The pathogenesis of ADM combined with OP has not been fully elucidated.T1D,LADA and OP are all polygenic diseases.There is clinical evidence that genetic factors may play an important role in the pathogenesis of ADM combined with OP.On the one hand,ADM and OP may share a common genetic background,such as pleiotropic genetic variants may affect the pathogenesis of ADM and OP at the same time.On the other hand,some genetic variants independently associated with OP can also affect the risk of OP in ADM patients.Genome-wide associate study(GWAS)is a quantitative analysis method to detect disease-or phenotype-associated single nucleotide polymorphism(SNP)on a genome-wide scale.This method provides a condition for comprehensive screening of SNPs and genes affecting the risk of OP in ADM patients.Conditional false discovery rate(c FDR)is a model strategy for obtaining common disease and phenotype shared genes by performing bivariate integrated analysis of GWAS discovery data associated with phenotypes.This method enables the systematic identification of ADM and OP pleiotropic SNPs and genes.Objective:To clarify the similarities and differences of bone mass characteristics among LADA of ADM,T1D and T2D patients,and to explore the factors related to bone mass changes in diabetic patients.To explore the genetic mechanism of ADM patients with OP from two aspects.On the one hand,to identify the pleiotropic SNPs and genes which are associated with ADM and OP from the perspective of genetic pleiotropy,and to reveal the genetic basis of the comorbidity of the two diseases.On the other hand,to determine whether there are independent OP-related SNPs and genes which may affects the risk of ADM patients with OP,and to further understand the influence of genetic factors on the occurrence of OP in ADM patients.Methods:(1)The subjects of this study were from the inpatient department,outpatient clinic and health examination centers of the Second Xiangya Hospital.Using the propensity score matching method,108 LADA,T1D and T2D patients were matched according to age,sex,disease duration,and proportion of postmenopausal women.216 non-diabetic healthy controls were matched by age,sex,and proportion of postmenopausal women.The differences in femoral neck bone mineral density(BMD),total hip BMD,lumbar spine BMD and clinical indicators were compared among the four groups.Multiple linear stepwise regression and logistic stepwise regression methods were used to explore the associated factors of BMD and OP in diabetic patients.(2)The public GWAS summary statistics such as Caucasian T1D(n=18942),LADA(n=2634)and femoral neck BMD which representing OP traits(n=53236)were used.Hierarchical Q-Q plot was used to assess the pleiotropic enrichment of SNPs associated with T1D/LADA and femoral neck BMD.The c FDR method was used to identify pleiotropic SNPs and genes in T1D/LADA and femoral neck BMD.Differential expression analysis of pleiotropic genes was carried out using the public whole gene expression profiling chip database of women in high and low BMD groups,as well as patients with T1D and healthy controls.Functional linkages between pleiotropic genes and known OP/ADM susceptibility genes were determined by gene interaction analysis.Gene Ontology(GO)enrichment analysis was used to annotate and describe the biological information and functions,and to find the biological pathways,molecular functions and cellular components involved in pleiotropic genes.(3)The patients of T1D/LADA combined with OP(n=69/151)or with normal bone mass(n=58/127)of Chinese Han population were included.Whole blood was collected for DNA extraction,and Illumina ASA gene chip was used to detect the whole genome DNA sequences.Then GWAS was performed to identify SNPs associated with OP in ADM patients by using bioinformatics software.Differential expression analysis of the annotated genes of associated SNPs was performed using the publicly available whole gene expression profiling microarray database of women in the high and low BMD groups.The GENEMIANIA software was used to explore the interaction between the annotated genes of associated SNPs and known OP-associated genes.Results:(1)Femoral neck BMD,total hip BMD and lumbar spine BMD were all gradually increased in T1D,LADA,T2D and the control groups(P<0.05 for all comparisons).The mean femoral neck BMD of the control,T1D,LADA and T2D groups were:0.80±0.13 g/cm~2,0.68±0.12 g/cm~2,0.72±0.14 g/cm~2and 0.75±0.12 g/cm~2,respectively.The mean total hip BMD were:0.91±0.13 g/cm~2,0.81±0.12 g/cm~2,0.86±0.15 g/cm~2,and 0.89±0.13 g/cm~2,respectively.The mean lumber spine BMD were:0.95±0.14 g/cm~2,0.86±0.14 g/cm~2,0.90±0.14 g/cm~2and0.96±0.14 g/cm~2,respectively.Compared with LADA,T2D and control group,femoral neck BMD,total hip BMD and lumbar spine BMD in T1D group were significantly lower(P<0.05 for all comparisons).BMD in each site of LADA group was significantly lower than that in control group(P<0.05 for all comparisons),and lumbar spine BMD in LADA group was significantly lower than that in T2D group(P<0.05).Femoral neck BMD in T2D group was significantly lower than that in control group(P<0.001),but there was no significant difference in the total hip and lumbar spine BMD between two groups.The proportions of OP patients in controls,T1D,LADA and T2D were 6.9%,27.8%,22.2%and2.8%,respectively.The proportions of OP patients in T1D and LADA groups were significantly higher than that in T2D and control groups(all P<0.001).However,there was no significant difference in the proportion of OP between either the T1D and LADA groups or the T2D and control groups.Multiple linear stepwise regression analysis showed that BMD in diabetic patients was positively correlated with body weight(P<0.001)and C-peptide(P<0.05),but negatively correlated with age(P<0.001).Logistic regression analysis revealed that Hb A1c(OR 1.18,95%CI 1.03-1.35,P=0.016)and diabetes type(P=0.002)were both independent risk factors of diabetes complicated with OP.Compared with T2D patients,the risk of complicated with OP in LADA and T1D patients was significantly increased(LADA:OR 14.12,95%CI 3.12-64.05,P<0.001;T1D:OR 15.04,95%CI 3.31-68.35,P<0.001).The disease duration,blood lipids,25-hydroxyvitamin D,blood calcium and phosphorus,chronic complications of diabetes were not found to be related to BMD.(2)There was significant pleiotropic enrichment of associated SNPs between T1D and femoral neck BMD.Taking femoral neck BMD as the condition,2981 SNPs were found to be significantly associated with T1D(c FDR<0.05),of which 1573 SNPs were newly discovered.Taking T1D as the condition,138 SNPs were found to be significantly associated with femoral neck BMD(c FDR<0.05),of which 107 SNPs were newly discovered.Forty-seven pleiotropic SNPs of T1D and femoral neck BMD(cc FDR<0.05)were identified,corresponding to a total of 30 pleiotropic genes.GO gene function enrichment analysis showed that pleiotropic genes were mainly involved in the interferon-γeffect-related signaling pathways,antigen processing and presentation signaling pathways,immune response activation signaling pathways and T cell receptor signaling pathway.Significant expressive differences of nine pleiotropic genes including AIF1,BTNL2,GAL,HLA-DQA1,HLA-DRA,HLA-DRB1,KDM6B,PLEKHA1 and PPP6R3 were found in peripheral blood mononuclear cells or blood cells in T1D and non-diabetic control groups,as well as in peripheral blood monocytes in female with high and low BMD.These genes were important potential pleiotropic genes for T1D and OP.(3)There was significant pleiotropic enrichment of associated SNPs between LADA and femoral neck BMD.A total of 779 SNPs were found to be significantly associated with LADA under the condition of femoral neck BMD(c FDR<0.05),of which 336 SNPs were newly discovered.In addition,118 SNPs(c FDR<0.05)were found to be significantly associated with femoral neck BMD under LADA conditions,of which 86SNPs were newly discovered.Eleven pleiotropic SNPs in LADA and femoral neck BMD were identified,corresponding to 9 pleiotropic genes,respectively as follows:rs11513729(ALDH2,MAPKAPK5-AS1),rs7953257(HECTD4),rs482194(TSBP1),rs537757(TSBP1),rs544358(TSBP1),rs9368716(TSBP1),rs7454557(NOTCH4,TSBP1),rs9268082(NOTCH4,TSBP1),rs6920486(NCR3,AIF1),rs10955908(SAMD 12-AS1,TNFRSF11B),rs2055101(SAMD12-AS1,TNFRSF11B).These above were also pleiotropic SNPs and genes for T1D and femoral neck BMD.GO enrichment analysis showed that pleiotropic genes performed multiple molecular functions such as aldehyde dehydrogenase activity,Notch binding,oxidoreductase activity,and NAD binding.The interaction network analysis showed that the pleiotropic genes such as NOTCH4,ALDH2,TSBP1,NCR3,AIF1,TNFRSF11B and HETD4 were co-expressed,co-localized or genetically interacted with multiple associated genes of LADA and BMD.(4)Through GWAS analysis,40 associated SNPs of OP in T1D were found(P<1.0E-05),corresponding to six genes including KCNK2,RNU7-74P,RBM47,LINC01411,SUMO2P6 and HIGD1AP3.The expression difference analysis found that KCNK2 and RBM47 genes expression was significantly different in peripheral blood monocytes of premenopausal women with high and low BMD groups(P=0.013 and0.006).The interaction network analysis found that KCNK2 and RBM47genes were co-expressed,co-localized or genetically interacted with multiple associated genes of OP.(5)12 SNPs associated with OP in LADA reaching genome-wide significance(P<1.0E-05)were identified through GWAS method,corresponding to 9 genes in total,including PALM2AKAP2,GEAMD1B,LOC105377814,LINCO1048,NRXN3,FHOD3,LINC00113,LINC00314.The expression difference analysis found that there was a significant difference in the expression of NRXN3 and FHOD3 genes in the peripheral B lymphocytes of postmenopausal women with high and low BMD groups(P=0.037 and 0.016).The interaction network analysis revealed that PALM2AKAP2,NRXN3 and FHOD3 genes were co-expressed,co-localized or genetically interacted with multiple associated genes of OP.Conclusion:In the clinical study,the BMD level of patients with different types of diabetes is lower than that of healthy people with the same age,sex,menopause status and disease duration.The BMD level of LADA patients is significantly higher than that of T1D patients and lower than that of T2D patients.The proportion of patients with ADM(LADA and T1D)combined with OP is significantly higher than that of T2D and non-diabetic healthy controls.Patients with ADM are at high risk of OP and fracture.This study further confirmed that genetic factor is one of the reasons for ADM combined with OP.On the one hand,pleiotropic genetic variations are involved in the pathogenesis of ADM and OP in Caucasian population.A total of 47 pleiotropic SNPs associated with T1D and OP and 11 pleiotropic SNPs associated with LADA and OP are identified through c FDR method.Multiple important potential pleiotropic genes are associated with the pathogenesis of ADM and OP.On the other hand,40 SNPs associated with T1D and 12 SNPs associated with OP in LADA patients are identified in Chinese Han population through GWAS study.Multiple important potential candidate genes are found to associate with OP risk of ADM patients.Several important candidate genes associated with the pathogenesis of ADM combined with OP are identified.This study not only provides new evidence for the genetic mechanism of ADM combined with OP,but also provides a new direction for further biological function experiments and disease prevention and treatment.