Abstract
Genome-wide association studies (GWAS) have identified hundreds of genetic variants associated with dyslipidemia. However, about 95% of of these variants are located in genome noncoding regions and cluster in different loci. The disease-causing variant for each locus and underline mechanism remain largely unknown. We systematically analyzed these noncoding variants and found that rs1997243 is the disease-causing variant in locus 7p22, which is strongly associated with hypercholesterolemia. The rs1997243 risk allele is associated with increased expression of GPR146 in human and targeted activation of the rs1997243 site specifically up regulates GPR146 expression in cultured cells. GPR146 is an orphan G-protein coupled receptor that is located on plasma membrane and responses to stimulation of heat-inactivated serum. Disrupting gpr146 specifically in the liver decreases the blood cholesterol level and prevents high-fat or high-fat high-cholesterol diets induced hypercholesterolemia in mice. Thus we uncovered a novel G-protein coupled receptor that regulates blood cholesterol level in both human and mouse. Our results also suggest that antagonizing GPR146 function will be an effective strategy to treat hypercholesterolemia.
Genome-wide association study (GWAS) is a powerful tool to ascertain the contribution of common genetic variants in population-wide diseases variability. Hundreds of GWAS studies have been applied to a variety of diseases or traits including dyslipidemia, diabetes, and hypertension et al1. More than 93% of these disease- and trait-associated variants are located in the non-coding regions, makes it difficult to evaluate their function2. Previous studies showed that these disease- and trait-associated variants are concentrated in regulatory DNA, with about 80% of all noncoding GWAS single nucleotide polymorphisms (SNPs) or linkage disequilibrium (LD) SNPs are located within DNase I hypersensitive sites (DHS)2, which suggest that most of these noncoding SNPs function through transcriptional regulation.
Hypercholesterolemia is the leading risk factor for cardiovascular diseases. Current evidence suggests that the heritability for blood cholesterol level is high, with 40-50% for low-density lipoprotein cholesterol (LDL-C) and 40-60% for high-density lipoprotein cholesterol (HDL-C)3,4. GWAS has been performed extensively on blood lipids traits and more than 300 risk loci were found in different populations5. These loci cover almost all the well-known genes that are important in lipid metabolism, such as LDLR6, ARH7, ABCA18, ABCG5/G89, PCSK910, NPC1L111,12, LIMA113, et al. Although GWAS is a powerful approach, we find that about 2/3 of these risk loci are located in noncoding regions and are not close to any gene known plays a role in lipid metabolism. The disease-causing variants in these loci and their underlying molecular mechanism remain largely unknown, which prevent the interpretation of the GWAS results and their application in precise medicine. On the other hand, these noncoding regions may harbor novel genes or signaling pathways involved in lipid metabolism and provides a valuable resource for further mechanistic studies.
We systematically analyzed these noncoding loci and primarily focused on loci that are not close to any gene known plays a role in lipid metabolism. One such locus 7p22 is strongly associated with increased level of blood cholesterol in multi cohorts14–16 (Fig S1a, S2b). The lead SNP rs1997243 is a common variant and has the highest frequency in European population but is absent in East Asian (Fig S1b). It is located in the noncoding region and has a strong linkage disequilibrium non-synonymous variant rs11761941 (GPR146 p. Gly11Glu) in some populations (Fig S1a, S2a, S2b). Both rs1997243 and rs11761941 are significantly associated with blood cholesterol level (Fig S2b)15. However, the GPR146 p. Gly11 is not conserved and has been substituted with Asp, Asn or Ala in many other species except Gray wolf (Fig S2c). Bioinformatics analysis also predicts that GPR146 p. Gly11Glu is benign and neutral (Fig S2d), rendering rs11761941 less likely to be the disease-causing variant.
We reasoned that any SNPs that have strong linkage disequilibrium with the lead SNP could be the real disease-causing variant. Since all these variants are located in the noncoding region, the real disease-causing variant is most likely located in the regulatory region, such as regions marked by DNase I hypersensitivity and/or histone modifications marker H3K27ac and H3K4me317–21. We systematically analyzed all SNPs that have strong linkage disequilibrium with the lead SNP rs1997243 in 7p22 locus. There are 125 SNPs were identified, with 28 of them are located in genome active regions (Fig S3a, S3i, Table S1). We then applied luciferase reporter assay to compare the transcriptional activities between the minor allele and the major allele for each of these SNPs. Promoter sequence of APOA1 was used as a positive control for the assay (Fig S3b). We found that across all SNPs tested, only rs1997243 shows increased promoter activity compared with its reference allele under similar transfection efficiency (Fig 1b, Fig S3c-k). The rs1997243 does not change enhancer activity in enhancer luciferase reporter assay (Fig 1c, Fig S3l), which is consistent with enriched promoter specific histone marker H3K4me3 at this position (Fig 1a)17,20,21.
Expression quantitative trait loci (eQTL) analysis showed that the rs1997243 minor allele (G-allele) is strongly associated with increased expression of GPR146 in human (Fig 1d). Targeted activation of rs1997243 site with enzymatic dead Cas9 (dCas9) system and a gRNA specifically to the rs1997243 position increases the expression level of GPR146 significantly, with no detectable impact on other genes in this region (Fig 1e, Fig S2e). These data suggest that the rs1997243 is the disease-causing variant and may increase blood cholesterol level through up regulating GPR146 expression.
GPR146 is an orphan G-protein coupled receptor that is highly expressed in liver and adipose tissue of both human and mouse (Fig S4a, Fig 2a). In the liver, it specifically expresses in hepatocytes (Fig 2b). By prediction, it contains typical seven transmembrane domains with N terminal facing extracellular compartment (Fig S4b, c). When expressed in cells, GPR146 is located in membrane fraction and is located on plasma membrane, which suggests that it may function as a receptor (Fig 2c, d). GPCRs typically signaling through Gαs, Gαi/o, Gαq/11, Gα12/13 or Gβ/γ and regulate cAMP production, intracellular Ca2+ mobilizations, ERK/MAPK activity or small G protein RhoA activity et al22. We found that GPR146 responses to serum filtered by 3 kDa cut-off filter and activates the transcriptional activity of cAMP response element (CRE) (Fig 2e). Moreover, this response is preserved when the serum was further heat-inactivated by boiling and can be fully blocked by PKA inhibitor H-89 (Fig 2f). Taken together, our data suggests that GPR146 is a cell signaling receptor that responses to serum stimulation and activates the PKA signaling pathway.
To further study its function in vivo, we generated gpr146 knockout mouse model with Cre-LoxP system (Fig S5a). Totally we got 6 lines of LoxP positive F1 mice and they were genotyped by genome sequencing (data not shown), southern blot analysis (Fig S5b) and PCR genotyping (Fig S5c). Line 92 was used for all experiments except otherwise indicated. We generated whole body, liver specific and adipose tissue specific gpr146 knockout mice by crossing the LoxP/LoxP mice with Cre recombinase driven by CMV, albumin and adiponectin promoters respectively. Whole body gpr146−/− mice have significantly decreased blood cholesterol level compared with their littermate controls (Fig S6a, b). In contrast, the adipose tissue specific gpr146−/− mice have no detectable difference of blood lipids levels compared with their littermate controls (Fig S6c, d). The liver specific gpr146−/− (Li-gpr146−/−) mice have significantly decreased blood cholesterol level as the whole body gpr146−/− mice and are protected from high-fat high-cholesterol diet induced hypercholesterolemia (Fig 3a-d), which suggest that gpr146 regulates blood cholesterol level mainly through the liver. Consistent with decreased plasma cholesterol level, both ApoB-100 and ApoB-48 protein levels are significantly decreased in the plasma, especially under high-fat high-cholesterol diet feeding (Fig 3e, f). ApoA1 is also slightly decreased in Li-gpr146−/− mice under high-fat high-cholesterol diet feeding (Fig 3e, f). Moreover, Li-gpr146−/− mice are protected from high-fat diet induced hypercholesterolemia (Fig 4a). To test whether acutely suppressing gpr146 will decrease blood cholesterol level, we knocked down gpr146 in livers of adult mice through adeno-associated virus mediated shRNA delivering. As shown in Figure 4b, knocking down gpr146 in the liver significantly decreases the blood cholesterol level, which indicates that blocking gpr146 function will be an effective strategy to decrease blood cholesterol in adults. These results were confirmed in Li-gpr146−/− mice derived from an independent F1 line (Fig 4c, d, Fig S5b). Taken together, our results clearly demonstrate that GPR146 positively regulates blood cholesterol level, which is consistent with increased blood cholesterol level in humans with rs1997243 minor allele.
In summary, through bioinformatics analysis and experimental verification we found a noncoding disease-causing variant rs1997243 in locus 7p22. The risk allele of rs1997243 up regulates an orphan G-protein coupled receptor GPR146 that plays an important role in regulating blood cholesterol level. We believe that the increased expression level of GPR146 can at least partially explain the disease-causing effect of rs1997243 in human.
In contrast to the causal variants in Mendelian disease, which typically confer large effect, the common variants from GWAS usually have modest effects for each of them. This is especially true for GWAS SNPs that are located in genome noncoding regions. However, variants that explain a small proportion of the traits may provide substantial biological or therapeutic insights. The rs1997243 confers a modest effect on total blood cholesterol level with effect size of 0.03314. However, combining bioinformatics analysis and functional studies we found that the downstream target gene GPR146 has a large impact on blood cholesterol level. Our data also reveal that GPR146 responses to an endogenous ligand in the serum and actives the PKA signaling pathway, which suggest that GPR146 is a functional GPCR and has therapeutic potential. Thus our study provides an example that the common noncoding variant with modest effect may provide important biological or therapeutics insights. The strategy we developed here can be applied to other noncoding loci with unknown mechanisms as well.
Our study should be interpreted within the context of its limitations. First, we systematically analyzed all SNPs in 200 Kb window across the locus and found that rs1997243 is the only variant that changes promoter activity of the genome sequence and increases the expression level of GPR146. We cannot exclude the possibility that there exist other variants that extremely far away from the lead SNP and mediate the disease-causing effect together with rs1997243. However, our study provides compelling evidences that rs1997243 is the disease-causing variant and increases GPR146 expression level, which at least contributes to the increased blood cholesterol level in human. Second, our animal models strongly suggest that Gpr146 regulates blood cholesterol level mainly through the liver. However, eQTL analysis showed that the strongest association for GPR146 expression level and the rs1997243 risk allele is in human whole blood cells. Thus we cannot exclude the possibility that GPR146 may regulate blood cholesterol level through other tissues together with liver in human. Third, we found that gpr146 knockout mice have decreased blood cholesterol level, however the underline mechanism needs further investigation. Our preliminary data (not shown) indicate that the gpr146 knockout mice have normal food intake and fecal cholesterol excretion rate, which suggest that the decreased blood cholesterol could be caused by decreased cholesterol secretion into circulation or increased cholesterol clearance from the circulation.
During preparation of this manuscript, Dr. Cowan’s group reported the phenotypic characterization of gpr146 knockout mice23. They reported that gpr146 knockout mice have decreased level of blood cholesterol, which is consistent with our results. However, our results provide genetic evidence that GPR146 regulates blood cholesterol level not only in mice but also in human. First, although the 7p22 locus is strongly associated with hypercholesterolemia, we are the first to show that the rs1997243 is the disease-causing variant in this locus. Second, we provide multiple evidences that the rs1997243 risk allele specifically up regulates the expression level of GPR146 in this locus. Third, by generating the gpr146 knockout mouse models, we provide strong evidences that Gpr146 positively regulates blood cholesterol level mainly through the liver. Altogether, our results indicate that GPR146 is an important regulator of blood cholesterol level in both human and mouse. Together with the decreased atherosclerosis in gpr146 knockout mice in Dr. Cowan’s report23, we believe GPR146 will be an attractive drug target for hypercholesterolemia and atherosclerotic cardiovascular diseases.
Methods
Mice
Mice were housed in the temperature-controlled specific pathogen-free animal facility on a 12 h light-dark daily cycle with free access to water and normal chow diet. All animal care and experimental procedures were approved by the Institutional Animal Use and Care Committee of College of Life Sciences, Wuhan University. Gpr146 LoxP mice were generated with CRISPR/Cas9 technology on C57BL/6J background by Nanjing Biomedical Research Institute of Nanjing University. Two gRNA spanning the gpr146 genome locus are used: 5’-CCAGCAATGCTGGGAGACGT-3’ and 5’-GGCTCCGGGCTCATGTGGGA-3’. Donor vector containing the gpr146 genome sequence and LoxP sites are co-injected with Cas9/gRNA complex (Fig S5a). F0 mice were crossed with wild-type C57BL/6J mice and F1 mice were further genotyped by DNA sequencing, southern blot and PCR analysis. CMV-Cre (The Jackson Laboratory: 006054), Albumin-Cre (The Jackson Laboratory: 003574) and Adiponectin-Cre (The Jackson Laboratory: 010803, gift from Dr. Shengzhong Duan, Shanghai Jiao Tong University) mice were used to generate whole body, liver specific and adipose tissue specific gpr146 knockout mice respectively. High-fat diet (60%, catalog: D12492) and high-fat high-cholesterol diet (40% Cal and 1.25% cholesterol, catalog: D12108C) were obtained from Research diets.
Cell Culture and Reagents
293T cells, HEK293 cells and HepG2 cells were purchased from China Center for Type Culture Collection (CCTCC) and cultured in high glucose DMEM with 10% FBS and 100 units/mL penicillin G/streptomycin (Gibco, 15140-122). Cells were incubated at 37°C with 5% CO2. Cell culture medium was obtained from Life Technologies (12800-082), FBS was obtained from Pan Seratech (ST30-3302). EDTA-free protein inhibitor cocktail was purchased from Bimake (B14001). Polyethylenimine (PEI) was obtained from Polysciences (catalog: 24765, Warrington, PA).
Lentivirus was produced in 293T cells by co-transfecting the packaging plasmids pVSVg (AddGene, 8454) and psPAX2 (AddGene, 12260) with Cas9 (Addgene, 52962) or gRNA (Addgene, 52963) expressing plasmids using PEI according to the instruction. 48 hours after transfection, condition medium was harvest for further experiment.
CRISPR/dCas9 mediated genome activation
CRISPR/dCas9 mediated genome activation was performed as described24 with following modification. The dCas9-GCN4-scFc-p65-HSF1 coding motifs (gift from Dr. Hui Yang, Institute of Neuroscience, Chinese Academy of Sciences) were splitted and expressed separately in pLVX-IRES-Puro vector (Clontech, 632183). pLVX-dCas9-GCN4, PLVX-scFv-p65-HSF1 are packed into lentivirus separately and co-infected with lentivirus expressing gRNA targeting the rs1997243 site in HepG2 cells. 48 hours after infection, cells were collected and RNA was extracted for gene expression analysis.
Luciferase reporter assay
For promoter activity assay, about 2 kb genome sequence covering indicated SNPs were amplified from HepG2 genome using primers listed in Table S2. The DNA fragments were cloned into upstream of firefly luciferase through Xho1 and Kpn1 in pGL3-basic vector (Promega, E1751). For enhancer activity assay, a TATA mini promoter was first cloned into upstream of firefly luciferase and then DNA fragments were cloned into upstream of TATA box through Xho1 and Kpn1 in pGL3-mini promoter vector. Corresponding minor alleles for each SNPs were introduced by recombination with primers listed in Table S2. The firefly luciferase reporter plasmids and renilla luciferase control plasmid (Promega, E2241) were co-transfected into HepG2 cells with Lipoplus (Sagecreation, Q03003) according to the instruction. 24 hours after transfection, luciferase activity was determined using Dual-luciferase reporter assay system (Promega, E1960). Firefly luciferase activity was normalized with renilla luciferase activity and the vector transfected group was set to one. Transfection efficiency for each reporters were measured by real-time PCR using DNA extracted from cells in parallel experiments. Specific primers recognizing firefly luciferase and renilla luciferase are listed in Table S2.
The CRE-luciferase reporter was generated by putting the cAMP regulatory elements (CRE) in front of firefly luciferase in pGL4 basic vector (Promega, E134A). hRenilla luciferase (Promega, E692A) was used as internal control. To increase plasma membrane localization of GPR146, a 39 amino acids of bovine rhodopsin was fused to the N terminal of GPR146 as described previously25. GPR146 expressing plasmid or vector control plasmid was co-transfected with luciferase reporter plasmids by transient transfection using PEI. Eight hours later, cells were changed with serum free medium containing 0.1% fatty acid free BSA (Sangon Biotech, A602448). About 16 hours later, cells were treated with or without indicated amount of fetal bovine serum (FBS) for 6 hours. Luciferase activity was assayed with Dual-luciferase reporter assay system (Promega, E1960). Filtered FBS was made by passing through a 3 kD cut-off filter with centrifugation (Millipore, UFC9003). Heat-inactivated FBS was made by boiling the 3 kD filtered FBS at 95oC for 10 minutes. Then the serum was span down and the supernatant was collected for experiments. In some experiments, cells were pre-incubated with 10 μM of PKA inhibitor H-89 (Selleck, S1582) for 60 minutes before FBS treatment.
ENCODE analysis
DNase I hypersensitive signal integrates display of DNase I hypersensitivity in multi cell lines by UCSC. DNase I Seq and H3K27ac, H3K4me3 histone modification Chip-seq were performed on human liver samples and the data was downloaded from ENCODE project data portal.
Links for these 4 datasets are:
DNase I hypersensitive signal:
http://genome.ucsc.edu/cgi-bin/hgTrackUi?db=hg38&g=wgEncodeRegDnase
DNase I Seq: https://www.encodeproject.org/experiments/ENCSR158YXM/
H3K27ac Chip-seq: https://www.encodeproject.org/experiments/ENCSR981UJA/
H3K4me3 Chip-seq:https://www.encodeproject.org/experiments/ENCSR344TLI/
Linkage disequilibrium analysis
Linkage disequilibrium (r2) was calculated with Ensemble LD calculator using European population from 1000 genome phase 3. The window size was set to 200 kb around rs1997243. r2>0.8 was considered as positive.
eQTL analysis
The eQTL data in whole blood cells was obtained from gTex and the eQTL figure was generated using gTex webtool.
Transmembrane domain and variant impact prediction
GPR146 Transmembrane domain was predicted by TMHMM server V.2.0 at http://www.cbs.dtu.dk/services/TMHMM/
GPR146 p. Gly11Glu impact prediction was performed with two different software.
PolyPhen-2: http://genetics.bwh.harvard.edu/pph2/
PROVEAN: http://provean.jcvi.org/index.php
Blood chemistry and lipoprotein profiling
Plasma total cholesterol and triglyceride levels were measured by enzymatic kits (Shanghai Kehua Bio-Engineering Co). Plasma lipoprotein particles were size fractionated by Fast Protein Liquid Chromatography (FPLC) using a Superose 6 column (GE Healthcare) and the cholesterol content in each fraction was measured accordingly26.
Immunoblot analysis
Biotinylation and immunoblot was performed as described previously27 except that the cells and liver tissue were lysated in RIPA buffer (50 mM Tris-HCl, pH=8.0, 150 mM NaCl, 0.1% SDS, 1.5% NP40, 0.5% deoxycholate, 2 mM MgCl2). For cell fractionation experiments, cells were homogenized by passing through #7 needle for 60 times on ice. Nuclear pellet was isolated by low speed centrifugation (750 g) for 20 minutes at 4°C. The supernatant was transferred for high speed (100,000 g) centrifugation for 60 minutes at 4°C. Then the supernatant was saved as cytosol fraction and the membrane pellet was re-suspended in SDS lysis buffer (10mM Tris-HCl, pH 6.8, 100mM NaCl, 1% SDS, 1mM EDTA, 1 mM EGTA) and incubated at 37 degree for 30 minutes. Then membrane suspension was span down again at 12,000 g for 5 minutes at room temperature and was saved as membrane fraction. The nuclear pellet was re-suspended in nuclear lysate buffer (20mM HEPES/KOH, pH7.6, 2.5% (v/v) glycerol, 1.5mM MgCl2, 0.42M NaCl, 1mM EDTA, 1mM EGTA)and rotates at 4 degree for 1 hour. Then the suspension was span down at 14,000 g for 20 minutes at 4oC and the supernatant was saved as nuclear fraction. Cytosol, membrane and nuclear fractions were added with sample buffer and incubated at 37 degree for 30 minutes and then subjected to immunoblot analysis. To detect apolipoproteins, the fresh blood was collected into EDTA coated tubules containing aprotinin (Sigma, A1153-25MG). Then plasma was isolated at 4 degree and was subjected to immunoblots analysis.
The following antibodies were used in this study: anti-Flag antibody (Medical & Biological Laboratories, PM020); anti-calnexin antibody (Proteintech, 10427-2-AP); anti-Actin antibody (Proteintech, 20536-1-AP), anti-Lamin B1 antibody (Proteintech, 20536-1-AP), anti-GPR146 antibody (CUSABIO,CSB-PA006863), anti-ATP1A1 antibody (Abclonal, A0643), anti-ApoB antibody26 and anti-ApoA1 antibody (Proteintech, 14427-1-AP).
Real-Time PCR analysis
Total RNA was extracted from cells or mouse tissues using TRI Reagent (Sigma-Aldrich, T9424). The quality and concentration of RNA were measured with NanoDrop ONEc (Thermo Scientific). cDNA was synthesized from 2 μg of total RNA using a cDNA Reverse Transcription Kit (Thermo Scientific, M1682). The cDNA was quantified by Real-Time PCR using SYBR Green master mix (YEASEN Biotech Co, 11201ES08). Reactions were running in technical duplicate on CFX96 or CFX384 wells plates. Relative quantification was completed using the ΔΔCT method. Gene expression was normalized to housekeeping gene Gapdh, 36B4 or Cyclophilin.
Primary hepatocytes and non-hepatocytes
Mouse liver was first ligated and a piece of liver was sliced and frozen in liquid N2 as whole liver sample. The left over liver was perfused with wash buffer followed by collagenases 1 digestion buffer exactly the same as described27. Then the liver was transferred to a 60 mm dish in cold digestion buffer and the particulate material was filtered through a 70 μm filter. The pass through was span down at 40 g*10 minutes for three times at 4°C. The pellets from each spin were pooled together and are the primary hepatocytes. The supernatant was span down again at 500 g for 10 minutes. The pellet was collected as non-hepatocytes. Total RNA was extracted and 2 μg of RNA was subjected to reverse transcription as described above followed by real-time PCR analysis.
Adeno-associated virus packaging and purification
Adeno-associated virus was produced in 293T cells by co-transfecting the shRNA expression AAV shuttle plasmid together with Delta F6 helper plasmid, Rev Cap 2/9 plasmids using PEI. 60 hours after transfection, cells were harvested and freeze-thaw in lipid nitrogen for five times. Then Benzonase Nuclease (Sigma, E1014) was added and incubated at 37 degree for 45 minutes. After centrifugation at 4000 rpm for 30 minutes, the supernatant was collected and added on top of the iodixanol gradient (15%, 25%, 40%, 58%). Then the sample was centrifuged at 48000 rpm in a Beckman type 70Ti rotor for 130 minutes at 18 degree. The fraction in the 40% iodixanol was collected and dialyzed with 1×PBS extensively. The titer of the virus was determined by qRT-PCR. To knock down gpr146 in the liver, each mouse was infused with 1X1011 virus particles through tail vein and samples were collected at one and two weeks after injection. The shRNA sequence against gpr146 was listed in Table S2.
Data analysis
All data are expressed as mean ± SEM and p values were calculated using Student’s test in GraphPad unless otherwise indicated.
Author contributions
F.F.H, X.L, C.F.C, Y.N.L and Y.W designed the study. F.F.H, X.L, C.F.C, Y.N.L, M.K.D, Y.Y.G, Y.L.Z, X.D.L developed experimental methods. F.F.H, X.L, C.F.C, Y.N.L, M.K.D, Y.Y.G, Y.L.Z, X.D.L, X.M.L performed experiments. D.H.W, Y.Z, M.A, Y.L, B.L.S, H.S.H.H, Y.W contributed to data analysis and interpretation. F.F.H, X.L, C.F.C, Y.N.L and Y.W wrote the manuscript with input from all authors. Y.W supervised the work and obtained the funding.
Competing interests
We have no conflict of interests to disclose.
Acknowledgements
The authors would like to thank Dr. Hui Yang (Institute of Neuroscience, Chinese Academy of Sciences) for the dCas9-GCN4-scFc-p65-HSF1 plasmid, the ENCODE Consortium and the ENCODE production laboratory(s) generating the datasets used in the manuscript. This work was supported by grants from the National Natural Science Foundation of China (31570807, 31771304, 91754101, 91857000), the National Key Research and Development Program of China (2016YFA0500100, 2018YFA0800700), the 111 Project of China (B16036), the Fundamental Research Funds for the Central Universities of China (2042017kf0240, 2042017kf0187), the Key Research and Development Program of Hubei province (2019CFA067) and state key laboratory of Natural Medicines (SKLNMKF201902).
Abbreviations
- GWAS
- Genome-wide association study
- SNPs
- Single nucleotide polymorphisms
- DHS
- DNase I hypersensitive signal
- eQTL
- Expression quantitative trait loci
- HDL
- High-density lipoprotein
- LDL
- Low-density lipoprotein
- CRISPR/Cas9
- Clustered regularly interspaced short palindromic repeats