**3.4 Low-frequency and rare variants**

Whole-genome and exome sequencing approaches as relatively new genetic analysis technologies are being used to pinpoint the effects of minor allele frequencies (MAF ≤ 5%) and rare variants (MAF ≤ 0.5%) on the heritability of metabolic disorders such as MetS and IR.

The genomes of 1092 individuals from 14 populations were analyzed by using both the whole-genome and exome sequencing methods to identify low-frequency and rare genetic variants across 14 populations in the 1000 Genome Project [60]. The reference panels gained from this project can capture up to 98% accessible SNPs at a frequency of 1% in related populations and also enable researchers to analyze common and low-frequency variants in each individual from various populations. The 38 million SNP panels from the 1000 Genomes Project gave near complete coverage of common and low-frequency genetic variation with MAF ≥0.5% across European ancestry populations.

The European Network for Genetic and Genomic Epidemiology (ENGAGE) consortium carried out 22 GWAS to examine associations of genetic variants with WHR, fasting glucose, BMI, and fasting insulin in 87,048 individuals of European ancestry. This study identified two new loci for BMI, and fasting glucose and new lead SNPs at 29 loci including the SNP (rs1260326) near *GCKR* for fasting insulin [61].

Whole exome sequencing in a Danish cohort of 1000 individuals with T2D, BMI >27.5 kg/m2 , and hypertension and of 1000 controls identified 70,182 SNPs with MAF > 1%. Subsequent exome sequencing was performed in a two-stage follow-up in 15,989 Danes and a further 63,896 Europeans. This study showed associations of two common SNPs in *COBLL1* (MAF = 12.5%, OR = 0.88, *p* = 1.2 × 10−11) and *MACF1* (MAF = 23.4%, OR = 1.10, *p* = 8.2 × 10−10) with T2D and a low frequency variant in *CD300LG* (MAF = 3.5%, *p* = 8.5 × 10−14) with fasting HDLcholesterol [62].

Although physiological functions of risk variants in *COBLL1* and *MACF1* remain still unclear, a risk variant rs72836561 at *CD300LG* was found to be associated with the decreased mRNA expression level of *CD300LG* in both skeletal muscle and adipose tissue, elevated intramyocellular lipid, and decreased insulin sensitivity, through a functional study. These results suggest an association between this variant and MetS traits [63].

Exome sequencing in an Icelandic population revealed that a low-frequency (1.47%) variant (rs76895963) in *CCND2* decreased the risk of T2D (OR = 0.53, *p* = 5.0 × 10−21) and was associated with elevated *CCND2* expression [64]. However, this variant was also associated with both greater height and higher BMI (1.17 cm per allele, *p* = 5.5 × 10−12 and 0.56 kg/m2 per allele, *p* = 6.5 × 10−7, respectively).

In 2733 individuals from the Greenlandic population that were historically isolated, combination analyses of Cardio-Metabochip based genotyping and exome sequencing revealed that a common variant in *TBC1D4* was associated with higher fasting glucose and decreased insulin sensitivity, resulting in decreased insulinstimulated glucose uptake due to the variant [65].

**219**

*Genetic Diversity of Insulin Resistance and Metabolic Syndrome*

Recently, whole-genome sequencing in a UK10K-cohort project consisting of 3781 healthy individuals with exome sequencing of 6000 individuals with either rare disease, severe obesity, or neurodevelopmental disorders has been performed to identify low-frequency and rare variants [66]. This project identified 24 million novel genetic variants including novel alleles associated with levels of TAG (*APOB*), adiponectin (*ADIPOQ*), and LDL-cholesterol (*LDLR* and *RGAG1*) from single-marker and rare variant aggregation tests and provided reference panels with increased coverage of low-frequency and rare variants. These panels are now being used to identify associations of low-frequency and rare variants with various traits related to

Fatty acid-binding proteins (FABPs) play important roles in lipid metabolism and signaling. Dyslipidemia often occurs along with IR, obesity, and hypertension in individuals with MetS. The methylation status of CpG islands of a key regulator of lipid homeostasis, *FABP3*, is known as a quantitative trait associated with MetS phenotypes in humans. To identify if CpG methylation of *FABP3* affects MetS traits in 517 Northern European family populations, the CpG islands in the *FABP3* gene were profiled by a quantitative methylation analysis method. In this study, regional methylation was found to be strongly associated with plasma total cholesterol (*p* = 0.00028) and associated with LDL-cholesterol (*p* = 0.00495) [67]. Methylation at individual units was significantly associated with MetS traits such as insulin sensitivity and diastolic BP (*p* < 0.0028). These results suggest that DNA methylation of *FABP3* strongly affects MetS and might have important implications for insulin,

Meanwhile, malnutrition in childhood, infancy, or fetus affects the prevalence of MetS in adults and their offspring [68], suggesting that maternal malnutrition

Although many GWA studies are widely used to identify genetic loci associated with IR, it remains challenging to identify the causal gene in each locus [69]. Recently, structural and functional connections between GWAS loci and vicinal or distal genes were identified by chromosome conformation capture (3C) technology and expression quantitative trait loci (eQTL) studies [70, 71]. However, the 3C experiments are expensive and the eQTL studies cannot identify all the causal genes for a locus. Moreover, the 2 methods cannot pinpoint the causal genes and mechanisms related to the risk loci of IR. More recently, clustered regularly interspaced short palindromic repeats (CRISPR) knockout screening platform as an alternative method has been applied to pinpoint functions of new candidate causal genes at IR-associated loci in human preadipocytes and adipocytes [72]. This screening platform successfully characterized the functions of 10 new candidate causal genes at IR-associated loci. The 10 candidate genes (*PPARG*, *IRS-1*, *FST*, *PEPD, PDGFC*, *MAP3K1*, *GRB14*, *ARL15*, *ANKRD55*, and *RSPO3*) showed diverse phenotypes in the 3 insulin-sensitizing mechanisms, including lipid metabolism, adipogenesis, and insulin signaling, and the first 7 of these genes could affect all the 3 mechanisms. Additionally, 5 of 6 eQTL genes were identified as the top candidate causal genes (*IRS-1*, *GRB14*, *FST*, *PEPD*,

To date, most studies examining epigenetic changes related to MetS or IR have been conducted in animals and few studies have been conducted in humans.

affects gene expression in offspring through epigenetic mechanisms.

Therefore, further studies in humans are needed in the future.

**4. CRISPR screen for genes affecting MetS or IR**

*DOI: http://dx.doi.org/10.5772/intechopen.93906*

health and disease.

**3.5 Epigenetic determinants**

lipids, and cardiovascular phenotypes of MetS.

*Genetic Diversity of Insulin Resistance and Metabolic Syndrome DOI: http://dx.doi.org/10.5772/intechopen.93906*

Recently, whole-genome sequencing in a UK10K-cohort project consisting of 3781 healthy individuals with exome sequencing of 6000 individuals with either rare disease, severe obesity, or neurodevelopmental disorders has been performed to identify low-frequency and rare variants [66]. This project identified 24 million novel genetic variants including novel alleles associated with levels of TAG (*APOB*), adiponectin (*ADIPOQ*), and LDL-cholesterol (*LDLR* and *RGAG1*) from single-marker and rare variant aggregation tests and provided reference panels with increased coverage of low-frequency and rare variants. These panels are now being used to identify associations of low-frequency and rare variants with various traits related to health and disease.

#### **3.5 Epigenetic determinants**

*Genetic Variation*

with the genetic link with IR [48, 49].

**3.4 Low-frequency and rare variants**

≥0.5% across European ancestry populations.

disorders such as MetS and IR.

insulin [61].

>27.5 kg/m2

cholesterol [62].

and MetS traits [63].

per allele, *p* = 5.5 × 10−12 and 0.56 kg/m2

stimulated glucose uptake due to the variant [65].

*RSPO3* (rs2745353) and *LYPLAL1* (rs4846565). The PRS including the risk alleles of the 10 loci was associated with the cardiometabolic phenotypes such as lower BMI, lower body fat percentage, smaller hip circumference, and decreased leg fat mass as well as the risk phenotypes such as higher fasting insulin and higher TAG levels. These results suggest that limited storage capacity of subcutaneous adipose tissue (SAT) and consequently the elevation of ectopic fat deposition may be associated

Whole-genome and exome sequencing approaches as relatively new genetic analysis technologies are being used to pinpoint the effects of minor allele frequencies (MAF ≤ 5%) and rare variants (MAF ≤ 0.5%) on the heritability of metabolic

The genomes of 1092 individuals from 14 populations were analyzed by using both the whole-genome and exome sequencing methods to identify low-frequency and rare genetic variants across 14 populations in the 1000 Genome Project [60]. The reference panels gained from this project can capture up to 98% accessible SNPs at a frequency of 1% in related populations and also enable researchers to analyze common and low-frequency variants in each individual from various populations. The 38 million SNP panels from the 1000 Genomes Project gave near complete coverage of common and low-frequency genetic variation with MAF

The European Network for Genetic and Genomic Epidemiology (ENGAGE) consortium carried out 22 GWAS to examine associations of genetic variants with WHR, fasting glucose, BMI, and fasting insulin in 87,048 individuals of European ancestry. This study identified two new loci for BMI, and fasting glucose and new lead SNPs at 29 loci including the SNP (rs1260326) near *GCKR* for fasting

Whole exome sequencing in a Danish cohort of 1000 individuals with T2D, BMI

Although physiological functions of risk variants in *COBLL1* and *MACF1* remain still unclear, a risk variant rs72836561 at *CD300LG* was found to be associated with the decreased mRNA expression level of *CD300LG* in both skeletal muscle and adipose tissue, elevated intramyocellular lipid, and decreased insulin sensitivity, through a functional study. These results suggest an association between this variant

Exome sequencing in an Icelandic population revealed that a low-frequency (1.47%) variant (rs76895963) in *CCND2* decreased the risk of T2D (OR = 0.53, *p* = 5.0 × 10−21) and was associated with elevated *CCND2* expression [64]. However, this variant was also associated with both greater height and higher BMI (1.17 cm

In 2733 individuals from the Greenlandic population that were historically isolated, combination analyses of Cardio-Metabochip based genotyping and exome sequencing revealed that a common variant in *TBC1D4* was associated with higher fasting glucose and decreased insulin sensitivity, resulting in decreased insulin-

per allele, *p* = 6.5 × 10−7, respectively).

MAF > 1%. Subsequent exome sequencing was performed in a two-stage follow-up in 15,989 Danes and a further 63,896 Europeans. This study showed associations of two common SNPs in *COBLL1* (MAF = 12.5%, OR = 0.88, *p* = 1.2 × 10−11) and *MACF1* (MAF = 23.4%, OR = 1.10, *p* = 8.2 × 10−10) with T2D and a low frequency variant in *CD300LG* (MAF = 3.5%, *p* = 8.5 × 10−14) with fasting HDL-

, and hypertension and of 1000 controls identified 70,182 SNPs with

**218**

Fatty acid-binding proteins (FABPs) play important roles in lipid metabolism and signaling. Dyslipidemia often occurs along with IR, obesity, and hypertension in individuals with MetS. The methylation status of CpG islands of a key regulator of lipid homeostasis, *FABP3*, is known as a quantitative trait associated with MetS phenotypes in humans. To identify if CpG methylation of *FABP3* affects MetS traits in 517 Northern European family populations, the CpG islands in the *FABP3* gene were profiled by a quantitative methylation analysis method. In this study, regional methylation was found to be strongly associated with plasma total cholesterol (*p* = 0.00028) and associated with LDL-cholesterol (*p* = 0.00495) [67]. Methylation at individual units was significantly associated with MetS traits such as insulin sensitivity and diastolic BP (*p* < 0.0028). These results suggest that DNA methylation of *FABP3* strongly affects MetS and might have important implications for insulin, lipids, and cardiovascular phenotypes of MetS.

Meanwhile, malnutrition in childhood, infancy, or fetus affects the prevalence of MetS in adults and their offspring [68], suggesting that maternal malnutrition affects gene expression in offspring through epigenetic mechanisms.

To date, most studies examining epigenetic changes related to MetS or IR have been conducted in animals and few studies have been conducted in humans. Therefore, further studies in humans are needed in the future.

### **4. CRISPR screen for genes affecting MetS or IR**

Although many GWA studies are widely used to identify genetic loci associated with IR, it remains challenging to identify the causal gene in each locus [69]. Recently, structural and functional connections between GWAS loci and vicinal or distal genes were identified by chromosome conformation capture (3C) technology and expression quantitative trait loci (eQTL) studies [70, 71]. However, the 3C experiments are expensive and the eQTL studies cannot identify all the causal genes for a locus. Moreover, the 2 methods cannot pinpoint the causal genes and mechanisms related to the risk loci of IR. More recently, clustered regularly interspaced short palindromic repeats (CRISPR) knockout screening platform as an alternative method has been applied to pinpoint functions of new candidate causal genes at IR-associated loci in human preadipocytes and adipocytes [72]. This screening platform successfully characterized the functions of 10 new candidate causal genes at IR-associated loci. The 10 candidate genes (*PPARG*, *IRS-1*, *FST*, *PEPD, PDGFC*, *MAP3K1*, *GRB14*, *ARL15*, *ANKRD55*, and *RSPO3*) showed diverse phenotypes in the 3 insulin-sensitizing mechanisms, including lipid metabolism, adipogenesis, and insulin signaling, and the first 7 of these genes could affect all the 3 mechanisms. Additionally, 5 of 6 eQTL genes were identified as the top candidate causal genes (*IRS-1*, *GRB14*, *FST*, *PEPD*,

and *PDGFC*), and expression levels of these 5 genes in human subcutaneous adipose tissue were found to be associated with increased risk of IR. Interestingly, it was first revealed in this study that the *FST, PEPD*, and *PDGFC* are involved in the functions of adipose in IR. Despite these findings, little is known about other functions of these 3 genes in adipose tissue, which may include novel molecular mechanisms for cardiometabolic disease. In this regard, studies will be needed to uncover new functions of these 3 genes in adipose tissue.
