The genetic architecture of NAFLD among inbred strains of mice

  1. Simon T Hui  Is a corresponding author
  2. Brian W Parks
  3. Elin Org
  4. Frode Norheim
  5. Nam Che
  6. Calvin Pan
  7. Lawrence W Castellani
  8. Sarada Charugundla
  9. Darwin L Dirks
  10. Nikolaos Psychogios
  11. Isaac Neuhaus
  12. Robert E Gerszten
  13. Todd Kirchgessner
  14. Peter S Gargalovic
  15. Aldons J Lusis  Is a corresponding author
  1. University of California, Los Angeles, United States
  2. University of Oslo, Norway
  3. Harvard Medical School, United States
  4. Bristol-Myers Squibb, United States
11 figures, 11 tables and 2 additional files

Figures

Effects of genetic background on hepatic TG accumulation.

(A) Hepatic TG levels in male mice after 8 weeks of HF/HS feeding. Results are presented as mean + SD. (BD) Correlation of hepatic TG with liver weight (B), hepatic total cholesterol (TC) (C), and hepatic phospholipid (D). r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.003
Plasma ALT activities and immune cell marker expression among inbred and recombinant inbred strains of mice.

(A) Plasma alanine aminotransferase (ALT) activities in male mice after 8 weeks of HF/HS feeding. Results are presented as mean + SD. (BD) Correlation of plasma ALT activities with hepatic triglyceride (B), hepatic Ptprc (Cd45r) expression (C) and hepatic. Cd68 expression. r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.004
Correlation of hepatic TG content with plasma metabolites and HOMA-IR.

(AF) Correlation of hepatic TG with plasma TG (A), plasma cholesterol (B), plasma glycerol (C), plasma insulin (D), plasma glucose (E), and HOMA-IR (F). r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.005
Correlation of hepatic TG content with adiposity and fat mass.

(AE) Correlation of hepatic TG with adiposity (A), subcutaneous fat (B), gonadal fat (C), mesenteric fat (D), and retroperitoneal fat (E). r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.006
Association mapping of hepatic TG.

(A) Manhattan plot showing the significance (−log of p) of all SNPs and hepatic TG. Genome-wide significance cut-off at 1% false discovery rate (FDR) is shown by the red line and cut-off at 5% FDR is shown in green. (B) Locus plot for genome-wide significant locus on chromosome 7 with approximate linkage disequilibrium block and candidate genes.

https://doi.org/10.7554/eLife.05607.012
Association mapping of hepatic TG with adiposity or insulin resistance as covariates.

Manhattan plot showing the significance −log10 (p value) of all SNPs and hepatic TG conditioned for percentage body fat (A) or HOMA-IR (B) as covariates. Genome-wide significance cut-off at 1% FDR is shown by the red line.

https://doi.org/10.7554/eLife.05607.013
Correlation of candidate gene expression with hepatic TG content.

(AB) Correlation of hepatic TG with expression levels of Gde1 (A) and Knop1 (B) in the liver. (C) Correlation of hepatic TG with expression levels of Gde1 in the white adipose tissue. r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.017
Effects of Gde1 overexpression in mice by adenoviral transduction.

C57BL/6 mice were injected with Ad-Gde1 (1 × 109 pfu per mouse, i.v.) and fed with a HF/HS diet for 7 days. Control group received the same dose of Ad-LacZ. (A) Western-blot of liver homogenate using anti-GDE1 or anti-tubulin antibody. (B) Comparison of liver weight between Gde1-overexpressing mice and the control mice. (C) Differences in plasma triglyceride (TG), TC, and free fatty acids levels between Gde1-overexpressing mice (filled bars) and the control mice (empty bars). (D) Hepatic fat percentage in the two groups of mice was determined by MRI. (EG) Liver lipids were extracted and quantified: triglyceride (TG), TC, and phospholipids (PL). (H) Expression of lipogenic genes was measured by qPCR and normalized to the level of the housekeeping gene 36B4. Ad-LacZ (empty bars) and Ad-Gde1 (filled bars) Results are presented as mean + SD (n = 7–8) * denotes p < 0.05 and ** denotes p < 0.01.

https://doi.org/10.7554/eLife.05607.019
Effects of Gde1 knockdown in mice by adenoviral transduction.

C57BL/6 mice were injected with Ad-shGde1 (1 × 109 pfu per mouse, i.v.) and fed with a HF/HS diet for 7 days. Control group received the same dose of Ad-Ctl. (A) Equal amounts of liver protein were loaded in each lane and Western-blotted using anti-GDE1 or anti-actin antibody. (B) Comparison of Gde1 mRNA levels between Gde1-knockdown mice and the control mice. (CE) Liver lipids were extracted and quantified: triglyceride (TG), TC, and phospholipids (PL). (F) Expression of lipogenic genes was measured by qPCR and normalized to the level of the housekeeping gene 36B4. Results are presented as mean + SD (n = 11–12) * denotes p < 0.05 and ** denotes p < 0.01.

https://doi.org/10.7554/eLife.05607.020
Correlation of hepatic TG and polar metabolites in the plasma.

Correlation of hepatic TG with plasma levels of arginine (A), ornithine (B), citrulline (C), TMANO (D), creatine (E), and creatinine (F). r, biweight midcorrelation; p, p-value.

https://doi.org/10.7554/eLife.05607.021
Author response image 1
Variation in candidate gene expression among strains.

The expression level of Gde1 (A), Knop1 (B) and Coq7 (C) among HMDP strains are shown. Results are presented as mean ± SD in log2 scale.

https://doi.org/10.7554/eLife.05607.028

Tables

Table 1

Top 50 liver genes correlated with hepatic TG levels

https://doi.org/10.7554/eLife.05607.007
RankGene symbolrP
1Cd360.6951.85E-17
2Mrpl160.6222.57E-13
3Enc10.6118.12E-13
42010003K11Rik0.6002.84E-12
5Tceal80.5993.00E-12
6Cmpk10.5983.32E-12
7Avpr1a−0.5954.45E-12
8Hmgcl0.5907.60E-12
9Akap20.5811.86E-11
10C1ra−0.5724.34E-11
11Reep30.5621.16E-10
12Skp1a0.5542.33E-10
13Esd0.5513.11E-10
14Hadh0.5503.48E-10
15Syap10.5493.66E-10
16Ermp10.5455.44E-10
17Ang−0.5445.53E-10
18Dak0.5436.24E-10
19Matr3−0.5417.12E-10
20Nudt90.5389.35E-10
21Srsf5−0.5389.89E-10
22Vps290.5371.04E-09
23Ttc230.5371.05E-09
24Entpd50.5361.14E-09
25Chchd60.5351.27E-09
26Plekha10.5341.32E-09
27Mogat10.5311.76E-09
28S100a100.5282.11E-09
29Plin40.5272.36E-09
30Anxa20.5252.73E-09
31Srxn10.5242.96E-09
32Cstb0.5233.36E-09
33Cml1−0.5223.70E-09
34Tpp10.5214.00E-09
35Apoc20.5185.06E-09
36F7−0.5156.32E-09
37Wfdc20.5146.72E-09
38Bche0.5146.72E-09
39Mms19−0.5146.74E-09
40Jun0.5137.05E-09
41Lifr−0.5137.33E-09
42Gjb1−0.5127.63E-09
43Fabp20.5118.56E-09
44Morc40.5109.22E-09
45Rnf110.5109.33E-09
46Egfr−0.5091.03E-08
47Slc16a70.5071.12E-08
48Gfm10.5051.32E-08
49Chpt10.5051.33E-08
50Rbbp4−0.5051.33E-08
Table 2

Top 50 adipose genes correlated with hepatic TG levels

https://doi.org/10.7554/eLife.05607.008
RankGene symbolrp
1Nrbp2−0.6242.42E-13
2Cp−0.5984.15E-12
3Hoxa7−0.5918.42E-12
4Prkcb0.5851.62E-11
5Cstb0.5822.03E-11
6Cd530.5822.06E-11
7Smap20.5802.62E-11
8Sft2d10.5783.04E-11
9Btk0.5763.56E-11
10Was0.5763.62E-11
11Il7r0.5753.94E-11
12Tmem53−0.5734.77E-11
13Rgs100.5725.70E-11
14Srp190.5716.27E-11
15Gpc3−0.5706.53E-11
16Bcl100.5706.72E-11
17Gpr650.5706.82E-11
18Tlr70.5697.18E-11
19Efhd20.5697.26E-11
20Actr30.5687.77E-11
21Cd720.5688.02E-11
22Dera0.5678.89E-11
23Pip4k2a0.5669.63E-11
24Hcst0.5669.64E-11
25Tyms0.5669.78E-11
26Cenpv−0.5669.79E-11
27Plxnc10.5651.05E-10
28Birc50.5651.07E-10
29Ptpn180.5631.27E-10
30Hoxa5−0.5621.37E-10
31Fam105a0.5621.43E-10
32Capza10.5621.44E-10
33Nap1l3−0.5611.50E-10
34Rgs180.5611.51E-10
35Phtf20.5601.59E-10
36Nckap1l0.5601.65E-10
37Coro1c0.5601.67E-10
38Coro1a0.5591.78E-10
39Lrrfip10.5591.89E-10
40C1qb0.5572.13E-10
41Taok30.5562.28E-10
42Ms4a6c0.5562.29E-10
43Bco2−0.5562.33E-10
44Adrb3−0.5562.34E-10
45Arhgap90.5562.37E-10
46Lrmp0.5562.47E-10
47Fyb0.5552.50E-10
48Lipa0.5552.63E-10
49Cdt10.5552.65E-10
50S100a40.5542.80E-10
Table 3

Pathway-enrichment analysis of the top 1000 hepatic genes correlated with hepatic TG levels, assessed with the DAVID database, and presented as total genes meeting that criterion in each pathway (Count), along with Benjamini corrected p values (Adj. p)

https://doi.org/10.7554/eLife.05607.009
CategoryTermCountAdj. pFold enrichment
GOTERM_CC_FATGO:0005739 ∼ mitochondrion1271.99E-051.57
KEGG_PATHWAYmmu04610:Complement and coagulation cascades183.93E-033.03
GOTERM_CC_FATGO:0044420 ∼ extracellular matrix part175.89E-032.91
Table 4

Pathway-enrichment analysis of the top 1000 adipose genes correlated with hepatic TG levels, assessed with the DAVID database, and presented as total genes meeting that criterion in each pathway (Count), along with Benjamini corrected p values (Adj. p)

https://doi.org/10.7554/eLife.05607.010
CategoryTermCountAdj. pFold enrichment
GOTERM_BP_FATGO:0000087 ∼ M phase of mitotic cell cycle369.69E-062.87
KEGG_PATHWAYmmu04142:Lysosome251.73E-032.50
GOTERM_BP_FATGO:0008064 ∼ regulation of actin polymerization or depolymerization195.97E-065.43
GOTERM_BP_FATGO:0007010 ∼ cytoskeleton organization391.89E-021.87
GOTERM_BP_FATGO:0006955 ∼ immune response501.80E-021.72
GOTERM_BP_FATGO:0009611 ∼ response to wounding422.64E-021.78
GOTERM_BP_FATGO:0045321 ∼ leukocyte activation347.64E-042.36
KEGG_PATHWAYmmu04662:B cell receptor signaling pathway208.62E-043.08
GOTERM_BP_FATGO:0001819 ∼ positive regulation of cytokine production131.85E-023.56
Table 5

Correlation between human GWAS candidate gene expression in mouse liver and adipose tissue with hepatic TG level

https://doi.org/10.7554/eLife.05607.011
LiverAdipose
rprp
Pnpla30.070.423−0.040.645
Gckr0.190.044*0.140.156
Ncan−0.100.3160.376.1 × 10-5*
Tm6sf20.150.123−0.230.012*
Lyplal10.270.003*−0.120.228
Trib1−0.100.3130.240.012*
  1. *

    Denotes p < 0.05.

Table 6

Association p-values of candidate genes on chromosome 7 with the peak SNP for hepatic triglyceride levels

https://doi.org/10.7554/eLife.05607.014
GeneLiver p-valueLiver expressionAdipose p-valueAdipose expression
Rps15aNAYesNAYes
Arl6ip10.011Yes0.294Yes
Smg1NANoNAYes
4930583K01RikNANoNAYes
Syt70.495No0.748Yes
Itpripl2NAYesNAYes
Coq77.18E-13Yes7.96E-14Yes
Tmc7NAYesNAYes
Tmc5NANoNAYes
Gde13.21E-05Yes3.98E-08Yes
Ccp110NANoNANo
9030624J02RikNAYesNAYes
Knop13.85E-07Yes0.043Yes
IqckNANoNANo
Gprc5b0.002No0.661Yes
Gpr139NAYesNANo
Gp20.910No0.209No
Table 7

Candidate genes under the chromosomes 3, 9, and 11 loci

https://doi.org/10.7554/eLife.05607.015
Chromosome 3Chromosome 9Chromosome 11
1110032F04RikNphp3Il12b
Ift80Uba5Ublcp1
Smc4Acad11Rnf145
Trim59Dnajc3Ebf1
Kpna4AcppGm12159
Gm1647Cpne4F630206G17Rik
Arl14Mrpl3Clint1
Ppm1lNudt16Lsm11
B3galnt11700080E11RikThg1l
Nmd3Nek11Sox30
SptssbAste1Adam19
Otol1Atp2c1Nipal4
Pik3r4Cyfip2
Itk
Fam71b
Table 8

Significant cis-eQTL at chromosome 7 locus

https://doi.org/10.7554/eLife.05607.016
GeneGene start positionSNP IDPositionLiver p-value
Arl6ip1118118891rs306680411182669692.79E-24
Coq7118509659rs324615101183500095.01E-24
Gde1118688545rs325114191190705211.32E-06
Knop1118842237rs322467451191608231.18E-07
GeneGene start positionSNP IDPositionAdipose p-value
Arl6ip1118118891rs306680411182669691.14E-36
Coq7118509659rs324308511179610921.65E-18
Gde1118688545rs315164251183727862.53E-12
Knop1118842237rs325323701192889743.47E-10
Gprc5b118972040rs486479261189184553.70E-09
  1. eQTL, expression Quantitative Trait Locus.

Table 9

Missense variants of candidate genes on chromosome 7

https://doi.org/10.7554/eLife.05607.018
GeneMissense variants
Rps15aNone
Arl6ip1None
Smg1None
4930583K01RikNone
Syt7Q117R
Itpripl2R204H, V240L, S420G
Coq7A290G
Tmc7C73S, T358S
Tmc5M11V, Q42L, V82D, A119T, V139I, P179S, A243G, R258K, V448I, E737D
Gde1None
Ccp110R180K, I199V, N248S, A326V, A332T, P427T, S439P, S445P, F624S, G746S
9030624J02RikS3A, D23G, V122A, G175A, T660M
Knop1I70M, V49A
IqckV49A, S148N
Gprc5bNone
Gpr139None
Gp2V79D, R483G
  1. The effect of missense mutation was assessed by PROVEAN software. Neutral amino acid substitutions, which do not affect protein stability and function are shown in blue whereas deleterious mutations are labeled in red.

Table 10

Correlation between hepatic lipids and gut microbiota

https://doi.org/10.7554/eLife.05607.022
TGTCUCPL
Family:
Clostridiaceae−0.1330.0350.1160.052
Erysipelotrichaceae0.1380.005−0.057−0.016
Lachnospiraceae0.066−0.1120.0000.045
Mogibacteriaceae0.033−0.055−0.092−0.089
Peptostreptococcaceae0.006−0.019−0.0030.096
Rikenellaceae0.0020.1420.125−0.047
Ruminococcaceae0.1380.1760.1690.030
S24-7−0.0920.003−0.111−0.166
Genus:
02d06−0.146−0.062−0.050.012
Adlercreutzia0.0140.0010.1330.084
Akkermansia−0.0310.0290.0800.074
Allobaculum−0.083−0.205*−0.178−0.009
Anaeroplasma−0.019−0.0470.0010.018
Bifidobacterium−0.014−0.082−0.131−0.084
Blautia0.003−0.217*−0.244−0.183
Clostridium−0.108−0.063−0.044−0.008
Clostridium.10.0810.044−0.0230.015
Coprobacillus−0.008−0.075−0.138−0.130
Coprococcus−0.0160.1740.2610.212*
Dehalobacterium0.0860.0700.0660.070
Dorea0.0780.0700.0560.081
Lactobacillus0.0050.021−0.012−0.073
Lactococcus0.0850.0010.0480.065
Oscillospira0.1400.1270.2610.206*
R.Gnavus0.103−0.024−0.037−0.016
Roseburia0.063−0.090−0.132−0.151
Ruminococcus0.099−0.131−0.160−0.218*
Sarcina−0.1090.0350.0670.014
SMB53−0.012−0.069−0.0290.052
Turicibacter−0.0350.0160.0190.013
  1. TG: triglyceride; TC: total cholesterol; UC: unesterified cholesterol; PL: phospholipids.

  2. *

    Denotes FDR < 0.05.

  3. Denotes FDR <0.01.

Table 11

Inbred and recombinant inbred strains used in this study

https://doi.org/10.7554/eLife.05607.023
Inbred strainsRecombinant inbred strains
StrainnStrainnStrainn
129X1/SvJ4AXB10/PgnJ3BXD1/TyJ3
A/J4AXB12/PgnJ5BXD11/TyJ2
AKR/J5AXB13/PgnJ5BXD12/TyJ5
BALB/cJ5AXB15/PgnJ4BXD13/TyJ2
BTBR T<+> tf/J4AXB19/PgnJ5BXD14/TyJ4
BUB/BnJ5AXB19a/PgJ4BXD15/TyJ5
C3H/HeJ4AXB19b/PgJ5BXD16/TyJ3
C57BL/6J4AXB2/PgnJ3BXD19/TyJ4
C57BLKS/J4AXB5/PgnJ4BXD20/TyJ7
C57L/J5AXB6/PgnJ3BXD21/TyJ5
C58/J4AXB8/PgnJ5BXD24/TyJ4
CBA/J4BXA1/PgnJ3BXD31/TyJ2
CE/J4BXA11/PgnJ4BXD32/TyJ5
DBA/2J4BXA12/PgnJ4BXD34/TyJ4
FVB/NJ5BXA13/PgnJ3BXD36/TyJ4
I/LnJ5BXA14/PgnJ4BXD38/TyJ4
KK/HlJ5BXA16/PgnJ3BXD39/TyJ4
LG/J4BXA2/PgnJ5BXD40/TyJ7
MA/MyJ4BXA24/PgnJ4BXD43/RwwJ5
NOD/ShiLtJ2BXA4/PgnJ4BXD44/RwwJ4
NON/ShiLtJ5BXA7/PgnJ6BXD45/RwwJ5
NZB/BlNJ4BXA8/PgnJ4BXD48/RwwJ4
NZW/LacJ1BXH19/TyJ6BXD49/RwwJ4
PL/J6BXH2/TyJ3BXD5/TyJ5
RIIIS/J4BXH20/KccJ3BXD50/RwwJ4
SEA/GnJ4BXH22/KccJ4BXD51/RwwJ4
SJL/J5BXH4/TyJ4BXD55/RwwJ4
SM/J4BXH6/TyJ5BXD56/RwwJ6
SWR/J4BXH8/TyJ3BXD6/TyJ5
BXH9/TyJ6BXD60/RwwJ2
CXB11/HiAJ5BXD61/RwwJ4
CXB12/HiAJ5BXD62/RwwJ6
CXB13/HiAJ5BXD64/RwwJ4
CXB3/ByJ5BXD66/RwwJ5
CXB4/ByJ5BXD68/RwwJ4
CXB6/ByJ6BXD70/RwwJ4
CXB7/ByJ6BXD71/RwwJ4
BXD73/RwwJ4
BXD74/RwwJ4
BXD75/RwwJ4
BXD79/RwwJ4
BXD8/TyJ2
BXD84/RwwJ4
BXD85/RwwJ4
BXD86/RwwJ5
BXD87/RwwJ3
BXD9/TyJ4

Additional files

Supplementary file 1

Strain average of hepatic lipids.

https://doi.org/10.7554/eLife.05607.024
Supplementary file 2

Correlation of hepatic TG and plasma metabolites.

https://doi.org/10.7554/eLife.05607.025

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Simon T Hui
  2. Brian W Parks
  3. Elin Org
  4. Frode Norheim
  5. Nam Che
  6. Calvin Pan
  7. Lawrence W Castellani
  8. Sarada Charugundla
  9. Darwin L Dirks
  10. Nikolaos Psychogios
  11. Isaac Neuhaus
  12. Robert E Gerszten
  13. Todd Kirchgessner
  14. Peter S Gargalovic
  15. Aldons J Lusis
(2015)
The genetic architecture of NAFLD among inbred strains of mice
eLife 4:e05607.
https://doi.org/10.7554/eLife.05607