Hepatocyte-specific loss of LAP2α reduces hepatic steatosis in male mice by enhancing LMNA-mediated transcriptional regulation

There is increasing evidence for the importance of the nuclear envelope in lipid metabolism, nonalcoholic fatty liver disease (NAFLD), and nonalcoholic steatohepatitis (NASH). Human mutations in LMNA, encoding A-type nuclear lamins, cause early-onset insulin resistance and NASH, while hepatocyte-specific deletion of Lmna predisposes to NASH with fibrosis in male mice. Given that variants in the gene encoding LAP2α, a nuclear protein that regulates LMNA, were previously identified in patients with NAFLD, we sought to determine the role of LAP2α in NAFLD using a mouse genetic model. Hepatocyte-specific Lap2α-knockout (HKO) mice and littermate controls were fed normal chow or high-fat diet (HFD) for 8 weeks or 6 months. In contrast to what was observed with hepatocyte-specific Lmna deletion, male HKO mice showed no increase in hepatic steatosis or NASH compared to controls. Rather, HKO mice demonstrated reduced hepatic steatosis, particularly after long-term HFD, with decreased susceptibility to diet-induced NASH. Accordingly, whereas pro-steatotic genes Cidea, Mogat1, and Cd36 were upregulated in Lmna-KO mice, they were downregulated in HKO mice, as were pro-inflammatory and pro-fibrotic genes. These data indicate that Lap2α deletion enhances hepatoprotective LMNA-mediated regulation of gene expression in mouse hepatocytes; therefore, LAP2α might represent a potential therapeutic target in human NASH. Brief Summary Loss of mouse LAP2α protected against diet-induced hepatic steatosis and NASH via enhancing protective regulatory functions of LMNA.

Similarly, a study of a small cohort of twins and siblings with NAFLD identified multiple variants in the gene encoding lamina-associated polypeptide-2α (LAP2α) (25), which is known to bind to LMNA and regulate its solubility and distribution within the nucleus, between the lamina and the nucleoplasm (26,27). In mice harboring global Lap2α deletion, cardiac and skeletal muscle abnormalities were noted, but liver development and early (4 week) liver histology were normal (28). However, the role of LAP2α in the protection from, or susceptibility to, NAFLD has not been tested directly.
Thus, despite strong evidence implicating nuclear lamins and their associated proteins in lipid metabolism and human disease, the in vivo function and physiologic relevance of LAP2α in hepatocytes remain unknown. To directly test the effect of LAP2α on susceptibility to NAFLD, we have generated mice harboring hepatocyte specific Lap2α deletion. Here we report that, 5 unlike Lmna and Lap1 deletion, Lap2α deletion protects against HFD-induced steatosis in mice, with reversal of the transcriptional signature seen in Lmna-KO mice, including genes associated with lipid metabolism, inflammation, and fibrosis. These data suggest that targeting of laminassociated proteins, particularly LAP2α, might offer a potential therapeutic strategy in human NASH. 6

Results
Normal baseline histology, with no predisposition to NAFLD, in Lap2α HKO mice.
Hepatocyte-specific deletion of either Lmna or Lap1 led to spontaneous NAFLD in mice (23,24), which was male-selective in the former and in both sexes in the latter. We previously reported variants in TMPO, encoding LAP2, in twins and siblings with NAFLD (25). Given that the α-isoform of LAP2 is known to regulate LMNA, we hypothesized that loss of LAP2α might affect susceptibility to NAFLD in mice. To test this hypothesis, we generated mice with hepatocyte-specific deletion of Lap2α (HKO mice) as described in Methods. Real-time qPCR (RT-qPCR) of whole liver RNA confirmed reduction in Lap2α transcript levels to <5% of WT level in HKO mice ( Figure 1A). HKO mice fed chow diet exhibited normal body mass as well as liver morphology and histology, consistent with a prior report (Figure 1B and C). Serum ALT and TG levels were normal in both HKO and WT mice ( Figure 1D). These results are in stark contrast to hepatocyte-specific Lmna and Lap1 deletion, which both led to spontaneous hepatic steatosis on chow diet (23,24).

Loss of Lap2α protected against diet-induced hepatic steatosis
To systematically test the role of LAP2α in HFD-induced NAFLD, we subjected HKO and control mice to high-fat diet with supplemental sucrose (HFD) for short-term (8 weeks) and long-term (6 months) treatment conditions. As the phenotype in mice with hepatocyte-specific Lmna deletion (Lmna-KO mice) was most prominent in male mice and we observed only modest hepatic steatosis in female WT and HKO mice after 6 months of HFD (Supplementary Figure   1), male mice were used for all further testing. After eight weeks of HFD, no increased susceptibility of male HKO mice to NAFLD or NASH was observed. Rather, we observed 7 decreased lipid deposition in HKO livers compared to WT as determined by oil red O (ORO) staining ( Figure 2A); blinded semi-quantitative steatosis scoring of hematoxylin and eosin stained sections from the same mice by an expert pathologist showed a trend toward less steatosis in HKO mice, though this was not statistically significant. However, after 6 months of HFD, HKO livers were found to have significantly decreased hepatic steatosis as determined by ORO staining or by blinded semi-quantitative steatosis scoring ( Figure 2B).

Loss of Lap2α increased LMNA nuclear rim staining
Given that Lmna deletion resulted in NAFLD (24) and that loss of Lap2α appeared to protect against NAFLD (Figure 2), and the known role of LAP2α in the regulation of LMNA, we asked whether LMNA distribution within the nucleus might be altered in HKO livers. It was previously reported that nucleoplasmic LAP2α binds to LMNA and affects the ratio of nuclear rim to nucleoplasmic LMNA staining by regulating the mobility and assembly state of LMNA (27,29).
To examine the distribution of lamin A/C in the absence of hepatocyte LAP2α, we performed immunofluorescence staining of mouse liver tissue. Consistent with prior reports (27,29), livers of HKO mice showed stronger LMNA staining at the nuclear rim compared to WT, with particularly prominent LMNA staining in HKO mice after 8 weeks of HFD ( Figure 3A); under chow diet conditions, LMNA staining was similar in WT and HKO mouse livers (Supplementary Figure 2). Immunoblot analysis of whole livers did not show significantly different overall LMNA levels in livers from HFD-fed HKO and WT mice ( Figure 3B). 8

Pro-steatotic genes were downregulated in HKO mice
Given that HKO mice were protected against HFD-induced NAFLD, as well as our observation of increased nuclear rim LMNA staining in HKO livers under HFD conditions, in the context of the known role of LAP2α in regulating LMNA distribution and association with chromatin (27), we hypothesized that the pro-steatotic transcriptional changes seen in Lmna-KO mice might be reversed in the setting of Lap2α deletion. To evaluate this hypothesis, we selected some of the most highly up-regulated pro-steatotic genes in Lmna-KO mice, including Cidea, Cd36, and Mogat1, for analysis. Cidea encodes a member of the CIDE family of proteins, which regulate lipogenesis and lipolysis (30,31). In contrast to Lmna-KO mice, HKO mice showed significantly reduced levels of Cidea transcript compared to WT mice after 8 weeks and 6 months of HFD ( Figure 4A, 4B). Notably, Cidea was also highly downregulated in HKO mice compared to WT under chow diet conditions (Figure 4C), suggesting that this transcriptional difference was a direct result of loss of LAP2α and a contributor to protection from NAFLD, rather than a consequence of decreased steatosis in HKO mice. Consistent with these results, levels of Cd36, which encodes a fatty acid translocase (32,33), were decreased in HKO mice at baseline and after long-term HFD. Similarly, expression of Mogat1, encoding the enzyme responsible for conversion of monoacylglycerol to diacylglycerol (34,35), was also decreased in HKO mice after long-term HFD and at baseline (Figure 4B, 4C). Taken together, these data indicate that loss of LAP2α in hepatocytes protects mice against HFD-induced steatosis via enhancing LMNA-mediated downregulation of pro-steatotic genes. 9

Loss of Lap2α protected against NASH and decreased expression of pro-inflammatory and pro-fibrotic genes in long-term HFD-fed mice
In human NAFLD/NASH, hepatocyte injury, steatohepatitis, and progressive fibrosis are thought to be the primary mediators of long-term sequelae including cirrhosis and hepatocellular carcinoma (3,5). Notably, Lmna-KO mice were more susceptible to both inflammation and fibrosis compared to control mice (24). Given that HKO mice were protected against hepatic steatosis and that biochemical and molecular testing indicated enhanced hepatoprotective transcriptional regulation by LMNA in the absence of LAP2α, we hypothesized that HKO mice might be protected against NASH, inflammation, and early fibrosis compared to control mice.
Among all mice fed chow diet or short-term (8 weeks) HFD, no fully developed NASH was seen in either HKO or WT mice ( Figure 5A). However, among mice fed HFD for 6 months, although serum ALT and TG levels did not differ, NAFLD activity scores (NAS, as previously described (36), with slight modification as per Methods section) were significantly lower in HKO mice compared to controls ( Figure 5B), suggesting that loss of LAP2α protected against lipidmediated hepatocyte injury and inflammation in addition to hepatic fat deposition. Consistent with this, HKO mice showed reduced expression of pro-inflammatory genes including Ubd, Irf7, Stat1, Themis, and Tnfa, though for some of these genes the difference did not reach statistical significance ( Figure 6A); all of these genes had been highly upregulated in Lmna-KO mice (24).
As expected with HFD lacking high cholesterol and fructose content (37, 38), we did not observe any significant fibrosis in any mice under any dietary condition. However, in livers from HKO mice after 6 months of HFD, as compared to control livers, there was a trend toward decreased expression of several pro-fibrotic genes, including Tgfb, Col1a1, Timp1, and Acta2, which did not reach statistical significance ( Figure 6B). 10

Discussion
Mutations in genes encoding nuclear envelope proteins are known to cause lipodystrophy syndromes with hepatic steatosis and progression to NASH (13,21,(39)(40)(41). Evidence is now accumulating from genetic animal models for the functional importance of the hepatocyte nuclear envelope in NAFLD (23,24,42). Hepatocyte-specific deletion of either Lmna or Lap1 leads to hepatic steatosis in mice, though the phenotypes of the two models appear to be distinct, with a male-predominant phenotype in the case of Lmna deficiency and a prominence of nuclear lipid droplets in the case of Lap1 deficiency.
Herein we show that, in contrast to Lmna or Lap1 deletion, hepatocyte-specific Lap2α deletion protected male mice from high fat diet-induced NAFLD and NASH. It was previously shown that LAP2α binds to LMNA via the proteins' respective carboxy-terminal tails, with consequent regulation and maintenance of LMNA in the nuclear interior in a mobile and low assembly state (26,27). These protein-protein interactions influence several physiological functions, including proliferation and differentiation, as well as chromatin organization and gene expression (43)(44)(45).
Our data, in agreement with prior reports (27,29), show that loss of LAP2α enhanced LMNA staining at the nuclear periphery without changing its overall expression level, which was accompanied by pro-steatotic, pro-inflammatory, and pro-fibrotic transcriptional changes that were opposite to those seen in Lmna-KO mice. In particular, whereas male mice lacking lamin A/C in hepatocytes were predisposed to steatohepatitis via a dramatic increase in transcriptional expression of pro-steatotic genes including Cidea, Cidec, Mogat1, and Cd36 (24), in this study we observed opposite changes in the expression of all of these genes in the absence of LAP2α.
While additional LMNA-independent mechanism(s) of protection via loss of LAP2α cannot be ruled out, these data suggest that the observed protection in LAP2α-deficient mice is due to 11 enhanced hepatoprotective LMNA-mediated regulation of gene expression in hepatocytes. This is in alignment with previous findings and supports the idea that lamin A/C has different properties at the nuclear periphery than within the nucleoplasm (27,29,46). Additionally, our data are supportive of a model in which there are mechanistic differences between LAP1-related steatosis (male equal to female, prominent nuclear lipid droplets) and LMNA/LAP2α-related steatosis (male > female, nuclear lipid droplets not prominent). This may reflect a predominance of ER dysfunction and defective lipid secretion in the case of loss of LAP1 (23, 42) versus a predominantly gene regulation-based mechanism of susceptibility to NAFLD in the case of LMNA and LAP2α (24).
It is important to note that variants in TMPO, encoding the six LAP2 isoforms including LAP2α, associated with increased risk of NAFLD in a twin and sibling cohort (25), rather than with protection from NAFLD. However, some of these variants were predicted to impact multiple LAP2 isoforms, rather than the α-isoform specifically, and thus their impacts cannot be directly compared to the effects of Lap2α deletion in the current study. Additionally, the consequences of TMPO variants that impact LAP2α function outside of its interaction with LMNA may be difficult to predict, whereas variants that enhance the LAP2α-LMNA interaction would be predicted to increase the risk of NAFLD. Finally, such germline genetic variants could impact liver development in ways that hepatocyte-specific deletion of the α-isoform of LAP2 via an albumin-Cre transgene (47), as in the current study, did not. Together, these differences likely account for the seemingly contradictory effects of TMPO variants in humans and Lap2α deletion in mice. 12 Collectively, our findings advance the current understanding of the physiological roles of LAP2α, LMNA, and the nuclear envelope in the onset and progression of steatosis and NASH.
Notably, whereas Lmna-KO mice were more susceptible to NASH and fibrosis compared to controls (24), here we observed significant protection from steatohepatitis in Lap2α-deficient (HKO) mice compared to WT and heterozygous control mice, with decreased expression of proinflammatory and pro-fibrotic genes. This apparent protection provided by loss of LAP2α from not only hepatic steatosis, but also NASH and susceptibility to fibrosis, is important given that morbidity and mortality of NAFLD are highly correlated with steatohepatitis and fibrosis (3,5).
Taken together, our data suggest that hepatocyte LAP2α may be a potential therapeutic target to enhance LMNA-mediated repression of lipogenic, pro-inflammatory, and pro-fibrotic gene expression in humans to prevent the development and/or progression of NASH. 13  14 For the 8-week HFD challenge, whole blood was collected by intracardiac puncture, and the liver was harvested under isoflurane anesthesia. For normal chow diet and 6-month HFD, mice were euthanized by CO2 asphyxiation prior to cardiac puncture and harvesting of the liver. Whole blood was centrifuged at 4°C and 3000 RPM for 10 min for serum collection. Liver tissue was stored in 10% formalin (for histology), optimum cutting temperature compound (OCT) for immunofluorescence staining, RNAlater (for gene expression studies) or snap-frozen and stored at -80°C (for protein analysis). Mice and whole livers were weighed before liver processing to calculate the liver percentage of body mass (% liver weight).

Biochemical Parameters and Liver histology
Serum alanine aminotransferase (ALT) and serum triglyceride (TG) values were determined by the Unit for Laboratory Animal Medicine at the University of Michigan. Paraffin-embedded livers were cut into 6 µm sections and stained with hematoxylin and eosin (H&E); images were captured using a Leica DM 5000B microscope. An expert liver pathologist (E.K.C.) scored the stained sections in a blinded fashion for steatosis and NASH activity, with the latter according to the method of Kleiner et al. (36), with slight modification. Briefly, microvesicular steatosis was included together with macrovesicular steatosis in the histologic scoring, and slightly lower thresholds were used to assign lobular hepatitis scores compared to the original description.

Oil Red O Staining (ORO) and Quantitation
Six-micron thick frozen liver sections were fixed in ice-cold 10% formalin and washed three times with water, followed by 5 mins in absolute propylene glycol (Sigma Aldrich) and stained with 0.5% ORO (Sigma Aldrich) for 10 min at 60°C. Stained slides were washed and 15 counterstained for 30-45 seconds with Gill's 3 Hematoxylin (Sigma Aldrich), then rinsed thoroughly with water and mounted with glycerol gelatin (Sigma Aldrich). Eight to ten fields per liver section (10X objective) were photographed using a Leica DM 5000B microscope for analysis. Fiji ImageJ Analysis Software version 1.51j8 (National Institutes of Health, Bethesda, MD) was used to calculate the proportion of each section with positive ORO staining.

Immunofluorescence Staining
OCT-embedded frozen liver tissue was cut into 6-micron sections and fixed in methanol at -20 °C for 10 min, followed by washed, permeabilization (0.1% Triton X-100 in PBS), and blocking (5% bovine serum albumin in PBS). Primary antibody was incubated overnight at 4 °C followed by a 1-hour incubation with Alexa Fluor-488 goat anti-mouse IgG. Slides were mounted using Prolong Gold Anti-Fade Reagent with DAPI, and stained sections were visualized with a Leica DM 5000B fluorescence microscope.

Quantitative real-time polymerase chain reaction (qPCR) analysis
Total RNA from WT and HKO mice was isolated using RNeasy Mini Kit (Qiagen), and cDNA was synthesized using iScript cDNA Synthesis kit (BIO-RAD CA, USA). Transcript levels of genes of interest were quantified using StepOne Real time PCR system (Thermo Scientific, CA, USA); qPCR primer sequences are shown in Supplementary Table 1. Relative expression was determined after normalizing to 18S RNA and use of 2 -ΔΔCT method. 16

Immunoblotting
Liver samples were homogenized using T-PER tissue protein extraction reagent buffer (Thermo Scientific, CA) containing protease and phosphatase inhibitors (Sigma). Total protein was quantified using BCA Kit (Thermo Scientific), and equal amount of protein was separated on 4-12% Novex tris-glycine gels (Thermo Scientific). Proteins were transferred to PVDF membrane (Bio-Rad, USA) and analyzed with anti-lamin A/C antibody. Blots were stripped using stripping buffer (Thermo Scientific) and probed with β-actin antibody (1:1500) to confirm equivalent protein loading.

Statistics
The data is expressed as mean ± SEM and analyzed by unpaired t test (2-tailed) or one-way analysis of variance (ANOVA), followed by the Mann-Whitney test using Graph Pad Prism 9.0 (CA, USA). *P<0.05, **P<0.01 and ***P<0.001 were considered to be significant.       .s • -f-

300
WT/Het HKO WT/Het HKO Figure 5. Decreased steatohepatitis in HKO livers. H&E staining of liver section of control and HKO mice was performed, and images were acquired via Leica DMRB 5000B microscope, with representative images shown. NAFLD activity scores (NAS) were determined by an expert liver pathologist in blinded fashion, and serum ALT and TG were measured. (A) Mice were fed with short term HFD (8 weeks; n>9 mice per group); (B) Mice were fed with long-term HFD (6 months; n=5 mice per group). Data represented as mean ± S.E.M. *P < 0.05, **P < 0.01 or ***P < 0.001, HKO versus control; where indicated, heterozygous mice were included as control mice with WT group. Scale bar, 100 µm.