Elongation of very long chain fatty acids-3 (Elovl3) is activated by ZHX2 and is a regulator of cell cycle progression

Abstract

Zinc fingers and homeoboxes 2 (Zhx2) are transcriptional regulators of liver gene expression with key functions in embryonic development as well as tissue regeneration in response to damage and disease, presumably through its control of target genes. Previous microarray data suggested that elongation of very long chain fatty acids-3 (Elovl3), a member of the ELOVL family of enzymes that synthesize very long chain fatty acids (VLCFAs), is a putative Zhx2 target gene. VLCFAs are core component of ceramides and other bioactive sphingolipids that are often dysregulated in diseases and regulate key cellular processes including proliferation. Since several previously identified Zhx2 targets become dysregulated in liver damage, we investigated the relationship between Zhx2 and Elovl3 in liver development, damage, and regeneration. Here, using mouse and cell models, we demonstrate that Zhx2 positively regulates Elovl3 expression in the liver and that male-biased hepatic Elovl3 expression is established between 4 and 8 wk of age in mice. Elovl3 is dramatically repressed in mouse models of liver regeneration, and the reduced Elovl3 levels in the regenerating liver are associated with changes in hepatic VLCFAs. Human hepatoma cell lines with forced Elovl3 expression have lower rates of cell growth; analysis of synchronized cells indicates that this reduced proliferation correlates with cells stalling in S-phase and lower mRNA levels of cell cyclins. Taken together, these data indicate that Elovl3 expression helps regulate cellular proliferation during liver development and regeneration, possibly through control of VLCFAs.

NEW & NOTEWORTHY Numerous targets of the transcription factor Zhx2 are dysregulated in liver disease. We show that the elongase Elovl3 is a novel Zhx2 target. Elovl3 and Zhx2 expression change during liver regeneration, which is associated with changes in very long chain fatty acids. Forced Elovl3 expression reduces cell growth and blocks cell cycle progression. This suggests that Elovl3 may account, at least in part, for the relationship between Zhx2 and proliferation during liver development and disease.

INTRODUCTION

Zinc fingers and homeoboxes 2 (Zhx2) is emerging as a critical transcriptional regulator of hepatic gene expression during development and disease. BALB/cJ mice, which have a natural hypomorphic mutation in the Zhx2 gene (1, 2), and engineered Zhx2 knockout mice have altered expression of numerous liver genes. In the absence of Zhx2, genes that are normally silenced in the postnatal liver, including α-fetoprotein (AFP), H19, glypican 3 (Gpc3), and lipoprotein lipase (Lpl), continue to be expressed (1, 3, 4). These putative Zhx2 target genes are frequently dysregulated during liver regeneration and in hepatocellular carcinoma (HCC) (5). Zhx2 also influences the expression of sex-biased target genes, including male-biased major urinary protein (Mup) and female-biased cytochrome P450 (Cyp) genes (6, 7); these genes are also dysregulated in regenerating liver. Furthermore, Zhx2 regulates hepatic lipid metabolism genes with notable effects on plasma lipid levels and consequent cardiovascular disease risk (8).

Liver microarray data from BALB/cJ and Zhx2-positive congenic mice showed a positive correlation between Zhx2 and elongation of very long chain fatty acids-3 (Elovl3) expression levels (8). Elovl3 belongs to a family of genes (Elovl 1–7) encoding enzymes that synthesize very long chain fatty acids (VLCFAs), complex lipids comprising fatty acids with 20 or more carbons (≥C20) (9). VLCFAs are commonly incorporated into ceramides or metabolized to lipid mediators that can regulate cellular growth, differentiation, proliferation, and other physiological functions (10). Elovl3, which is expressed primarily in brown and white adipose tissue, skin sebaceous glands, and the liver, synthesizes C20-C24 saturated and monounsaturated fatty acids (11–13). Dysregulated lipid metabolism is a key feature of numerous liver pathologies and is tightly regulated during regeneration for proper cellular proliferation and differentiation (14, 15). We therefore investigated a potential link between Zhx2 and Elovl3 during liver development, damage, and regeneration. Using several mouse models and tissue culture assays, we demonstrate that Zhx2 is a transcriptional activator of hepatic Elovl3 expression. Furthermore, we find that Elovl3 mRNA levels decrease dramatically in mouse models of liver regeneration. Consistent with this, hepatic concentrations of ELOVL3-synthesized VLCFAs are concomitantly lower in regenerating livers. We also investigated whether the changes in Elovl3 expression might impact cell proliferation. Our studies reveal that forced Elovl3 expression slows the growth and expansion of human hepatoma cells. In addition, studies in synchronized cells indicate that Elovl3 expression stalls the cell cycle in S-phase with significant reductions in the expression of Cyclins D, A, and E, key regulators of cell cycle progression. These data suggest Zhx2 regulation of Elovl3 expression and VLCFA synthesis are key components of cellular proliferation during liver development and regeneration.

MATERIALS AND METHODS

Mouse Strains

All mice were housed in the University of Kentucky Division of Laboratory Animal Research facility and maintained in accordance with Institutional Animal Care and Use Committee (IACUC) approved protocols. All mice had ad libitum access to food and water and were maintained on a 14/10-h light/dark cycle. BALB/cJ, BALB/cByJ, C57BL/6 (BL/6), and C3H/HeJ (C3H) mice were obtained from The Jackson Laboratory. Hepatocyte-specific Zhx2 knockout (Zhx2^Δhep) and whole body Zhx2 knockout (Zhx2^KO) mice have been previously reported (6, 7). Briefly, BL/6 mice homozygous with floxed Zhx2 alleles (Zhx2^f/f) were bred to BL/6 mice expressing albumin promoter-driven Cre recombinase (Alb-Cre, Jackson Labs no. 003574) or BL/6 mice expressing E2a promoter-driven Cre recombinase (E2a-Cre, Jackson Labs no. 003724) and then back-crossed to produce hepatocyte-specific Zhx2 knockout mice (Zhx2^Δhep) or whole-body Zhx2^KO mice, respectively. Littermates without Cre (Zhx2^f/f), which express Zhx2 at the same levels as wild-type mice, were used as controls.

Developmental Timepoint Studies

Female C3H mice were bred to male BL/6 mice to generate B6C3F1 offspring; these mice were used since they have the wild-type Zhx2 gene and were used for studies to compare Zhx2 and Elovl3 mRNA levels. Pregnant mice were monitored for vaginal plugs and then euthanized after 17 days to collect livers for embryonic timepoints. Neonatal pups and mature mice were euthanized by age-appropriate methods at the indicated timepoints. Livers were excised, snap-frozen, and stored at −80°C until further analysis.

Liver Regeneration

Liver regeneration was induced by a single intraperitoneal injection of carbon tetrachloride (CCl₄). Adult (2 mo old) male C3H, BL/6, and B6C3F1 mice were used since C3H and B6C3F1 mice have higher AFP activation after CCl₄ treatment than BL/6 mice, and we wanted to test whether changes in Elovl3 expression also exhibited strain-specific differences. Mice were administered either 0.05 mL mineral oil containing 10% CCl₄ (vol/vol; n = 5) or 0.05 mL mineral oil (MO) as control (n = 5). After 3 days, animals were euthanized, and livers were collected for future analysis as described (3).

RNA Extraction, cDNA Synthesis, and Quantitative Real-Time PCR

Samples stored at −80°C were thawed on ice. Approximately 100 mg of tissue or 10⁶ cells were homogenized in 1 mL RNAzol RT (Sigma no. R4533), and mRNA was extracted according to the product technical bulletin. cDNA was synthesized from 1 µg of RNA using the iScript cDNA Synthesis Kit (BioRad No. 170-8891). Reverse Transcriptase Quantitative PCR (RT-qPCR) reactions using SYBR Green PCR Supermix (BioRad No. 172-5275) were performed with a CFX96 Touch Real-Time PCR Detection System, and results were analyzed with the CFX Manager software (BioRad). Oligonucleotide primer sequences are listed in Table 1. RT-qPCR C_T values for mouse studies were normalized to the indicated reference gene to evaluate relative tissue expression values. Reference genes were validated for stable expression in both control and experimental groups in each experiment. The 2^−ΔΔCT method was used to calculate expression fold changes (16).

Table 1.

Primer sequences for RT-qPCR

Gene	Forward	Reverse
hAFP	GCACACAAAAAGCCCACTCCAGCAT	GCGAGCAGCCCAAAGAAGAATTGTA
hELOVL3	GCCCTATAACTTCGAGCTGTC	CTTCATGTAGTTCTGCCCCAC
hCCNA2 (Cyclin A2)	CTGCATCTCTGGGCGTCTTT	GTGCAACCCGTCTCGTCTTC
hCCNB (Cyclin B)	TGGTGAATGGACACCAACTCT	TAGCATGCTTCGATGTGGCA
hCCND1 (Cyclin D1)	GGATGCTGGAGGTCTGCGA	TAGAGGCCACGAACATGCAAGT
hCCNE1 (Cyclin E1)	CCCCATCATGCCGAGGGAG	TTATTGTCCCAAGGCTGGCT
hB2M	GACTTTGTCACAGCCCAAGATAG	TCCATTCCAAATGCGGCATCTTC
hGAPDH	TGCACCACCAACTGCTTAG	AGAGGCAGGGATGATGTT
mAFP	CCGGAAGCCACCGAGGAGGA	TGGGACAGAGGCCGGAGCAG
mElovl3	CCTCTGGTCCTTCCTGGCA	CGGCGTCATCCGTGTAGATGGC
mGpc3	CTCCCAAGCAACGCCAATATAGAT	CTGATTCTTCATCCCGTTCCTTGC
mL30	ATGGTGGCCGCAAAGAAGACGAA	CCTCAAAGCTGGACAGTTGTTGGCA
mSfrs4	CTCGCACAGAGTACAGACTTAT	TTGCGTCCCTTGTGAGCATCT
mZhx2	AGGCCGGCCAAGCCTAGACA	TGAGGTGGCCCACAGCCACT
mScd1	TTCACGACCCCACCTATCAG	GAGCGCTGGTCATGTAGTAG
mScd2	CTGCTGCATTTGGGAGCCT	AGGGCGCTGATTACATAGTAC
mScd4	GCTTCTAAACTTCACTTGGCTG	ACGTGTGATGGTAGTTGTGG

Western Blots

Western blots were prepared as described (6, 7). Briefly, protein lysates (30 µg/sample) were applied to SDS-PAGE gels, transferred to PVDF membranes, and probed with primary antibodies against either ZHX2 (Bethyl Laboratories) or the FLAG epitope (Sigma Cat. No. F1804-1MG) and reprobed with a rabbit polyclonal anti-actin (Sigma Cat. No. A2066) and species-specified HRP-conjugated secondary antibodies [Santa Cruz goat anti-mouse IgG-HRP (Cat No. sc-2031) and anti-rabbit IgG-HRP (Cat No. sc-2004)], followed by detection with the SuperSignal West Pico-Plus ECL kit (Thermo Scientific Cat. No. 34580).

Analysis of Hepatic VLCFA

Hepatic VLCFAs were analyzed by chemical derivatization and ultra-high-performance liquid chromatography-tandem mass spectrometry (LC-MS/MS) using adaptations of the methods of Bollinger et al. (17) and carried out at the University of Kentucky Small Molecular Mass Spectrometry Core Laboratory. Liver samples (∼10 mg) were homogenized in 0.5 mL 0.1 M HCl, and then 10 µl homogenate was mixed with 50 µl of internal standard (250 ng/mL Octadecanoic acid-d35), 0.5 mL 0.1 M HCl, 2 mL methanol, and 1 mL chloroform. After vortex-mixing was processed for 5 min, an additional 1 mL chloroform and 1.3 mL 0.1 M HCl were added, followed by an additional 5 min of vortexing, and the two phases were separated by centrifugation at 3,000 g for 10 min. The lower phase was removed and evaporated to dryness under filtered nitrogen. Dried lipids were incubated with 1 mL 1 N KOH in ethanol at 90°C for 2 h. After saponification, the solution was transferred to a 12-mL glass tube, the vial was rinsed with 1 mL ethanol, and the combined ethanol solution was extracted with 2 mL chloroform and 1.8 mL PBS. After phase separation, the lower layer was transferred to a 4-mL vial and dried under nitrogen. The dried material was derivatized with N-(4-aminomethylphenyl) pyridinium as described (17), followed by LC-MS/MS on the same day using an AB Sciex 4000 Q Trap coupled with an Exion LC system. The Analyst software package was used for data collection and analysis. Chromatography was carried out with a C₈ reverse-phase column (Waters ACQUITY UPLC BEH C8, 2.1 × 50 mm, 1.7 µm) maintained at 40°C with the flow rate set to 0.4 mL/min. Solvent A was 100% water/0.1% formic acid, and solvent B was acetonitrile/0.1% formic acid. A gradient program was used as follows (T min/%A): 0/90, 0.5/90, 0.51/80, 10.0/30, 10.1/0, 12.0/0, 12.1/90, and 15.0/90. The injection volume was 2.0 µL. The mass spectrometer was equipped with an electrospray ionization (ESI) source and operated in positive mode under the following operating parameters: IonSpray Voltage 5.5 kV, desolvation temperature 50°C ion source gas 1 40 psi, ion source gas 2 50 psi, curtain gas 30 psi, collision gas medium, declustering potential 160 V, entrance potential 10.0 V, and collision energy 55.0 V. Quantitative analysis was conducted by monitoring the precursor ion to production ion transition of each analyte.

Expression Plasmids

A full-length expression vector for the 271 amino acid mouse ELOVL3 protein was generated by PCR amplification of mouse liver cDNA (Forward primer: GCCACCATGGACACATCCATGAATTTCTCAC; Reverse primer: GGATCCTTGGCTCTTGGATGCAACTTTG). Amplicons were cloned into the pGEM-T Easy vector (Promega No. A1360), sequenced, excised using EcoRI and BamHI restriction enzymes, and cloned into the pcDNA3.1 expression vector (Invitrogen V790-20). Elovl3 and empty vector pcDNA3.1 expression plasmids were transformed into competent E. coli cells using a standard cell transformation protocol. Plasmid preparations were performed using a Plasmid Max Kit (Qiagen no. 12165). The Zhx2 expression plasmid was described previously (6).

Cell Culture and Transfection of Expression Plasmids

HepG2, Huh7, and HeLa cells were cultured in Dulbecco’s modified Eagle medium (DMEM, Corning Cellgro No. 10–017-CV) supplemented with 10% fetal bovine serum (FBS, Fisher no. 03–600-511) and maintained at 37°C and 5% CO₂. Cells were seeded onto 10-cm (2) plates and cultured for 24 h to obtain 70–80% confluence. Transfections using Turbofect (Thermo Scientific no. R0531) were performed according to the manufacturer’s protocol. In some experiments, cells were cotransfected with a GFP expression plasmid to visualize transfection efficiency before further experimentation or for cell sorting as described below.

Chromatin Immunoprecipitation

Chromatin immunoprecipitation (ChIP) analysis was performed as previously described (6, 7). Briefly, adult male mouse livers from Zhx2^WT and Zhx2^KO mice were harvested and processed for ChIP with rabbit anti-ZHX2 antibodies (Bethyl laboratories) or control IgG using the Magna ChIP HiSense kit (Millipore). Primers were used to generate amplicons from the Elovl3 promoter and DNase hypersensitive sites (DHSs) located roughly 7 kb (3′ DHS) and 21 kb (5′ DHS) upstream of the Elovl3 promoter (oligonucleotide primers shown in Table 2). ChIP DNA samples were analyzed by quantitative PCR using Universal SYBR Green Supermix (Bio-Rad) and a Bio-Rad CFX96 real-time PCR system.

Table 2.

Primer sequences for ChIP

Elovl3 Region	Forward	Reverse
Promoter	GATGTCCCTTGGCTAAGATTGG	CCGTGCCTGCCGTATATATATC
3′ DHS (−7 kb)	GTGGCTGACTTGGAACTTTTG	TGGTTTCCTTTGTGAGTACTTGC
5′ DHS (−21 kb)	AGGTTTGGAAAGAGCCACTG	CTCTGAGATGCTAATCCACTCAC

ChIP, chromatin immunoprecipitation; DHS, DNase hypersensitive site.

Growth in Monolayers

Huh7 cells were seeded in 6-well plates and cultured for 24 h and then transfected with either human ELOVL3 or pcDNA expression plasmids. After 72 h, cells were trypsinized, collected, and counted by hemocytometer with Trypan Blue dead cell exclusion.

Growth in Soft Agar

Transfected Huh7 cells were cultured for 24 h and then treated with trypsin to collect cells. Washed and resuspended cells (5 x 10 (3)) were suspended in growth media and 0.8% agar, seeded between two layers of 1.2% agar mixed with growth media and then cultured for one week in 96-well plates. Cell viability was measured by CellGlo Titer Luminescent Cell Viability Assay (Promega No. G7570) according to the product protocol and results were read on a luminometer. Sample luminescent values were normalized to wells containing media and agar layers without cells to account for background. Luminescent values for control cells were set to 100% and compared with values of ELOVL3 transfected cells in three separate experiments, with each sample assayed in technical triplicates.

Cell Synchronization and Analysis of Cell Cycle

HeLa cells cotransfected with GFP and either ELOVL3 or pcDNA3.1 expression plasmids were cultured for 24 h and then synchronized by blocking cell growth using a double Thymidine block and serum starvation (18). After release into fresh growth media, cells were immediately collected (T₀) and at subsequent timepoints as designated. For cell cycle analysis, cells fixed overnight in 70% ethanol were stained with propidium iodide (Roche No. 11348639001) and filtered to remove cell aggregates immediately before analysis by flow cytometry. Cells were gated for GFP expression to focus on transfected cells and then analyzed for DNA content to determine the percentage of cells in each cycle phase. FACS was performed by the University of Kentucky Flow Cytometry & Cell Sorting core facility.

Statistical Analysis

Data are graphed as means ± SD. Statistical significance was determined by Student’s t test or ANOVA with a P value < 0.05 considered significantly different. Data were graphed and analyzed using GraphPad Prism software.

RESULTS

Liver Elovl3 mRNA Levels Are Reduced in the Absence of Zhx2

The BALB/cJ mouse substrain carries a natural hypomorphic mutation in Zhx2, providing a model to identify putative Zhx2 target genes. Microarray data indicated that adult male liver Elovl3 mRNA levels are reduced in BALB/cJ mice, suggesting that Zhx2 is required for normal Elovl3 expression (8). To test the influence of Zhx2 on hepatic Elovl3 expression, we analyzed Elovl3 mRNA levels by RT-qPCR in adult male livers from BALB/cJ mice and the highly related BALB/cByJ substrain, which has wild-type Zhx2 expression. Elovl3 mRNA levels are substantially lower in BALB/cJ than in BALB/cByJ mice (Fig. 1A), and this corresponds with very low ZHX2 protein levels in livers from BALB/cJ mice (Fig. 1E). We also attempted to assess ELOVL3 protein expression to determine if they also showed changes but were unable to obtain clear results with the limited number of commercially available antibodies (data not shown). Since Elovl3 is expressed in hepatocytes of the adult liver, male Zhx2^Δhep mice, in which Zhx2 is deleted solely in hepatocytes and control littermates (Zhx2^f/f) were examined for differences in Elovl3 mRNA levels. In the absence of hepatocyte Zhx2 expression (Fig. 1B), Elovl3 mRNA levels are significantly reduced in male Zhx2^Δhep mice (Fig. 1C), whereas AFP, which is repressed by Zhx2, is expressed at higher levels in Zhx2^Δhep mice (Fig. 1D). These results are consistent with Elovl3 regulation by Zhx2.

Figure 1.

Elovl3 mRNA levels are reduced in the absence of Zhx2. RNA was prepared from the livers of adult male (A) BALB/cByJ (cByJ; Zhx2+) and BALB/cJ (cJ; Zhx2-) mice and (B–D) BL/6 mice that express (Zhx2^f/f) or do not express Zhx2 in hepatocytes (Zhx2^Δhep). E: Western blot of protein lysates from BALB/cByJ (cByJ; Zhx2+) and BALB/cJ (cJ; Zhx2-) mouse livers confirming lack of ZHX2 protein in BALB/cJ mice. Elovl3 (A and C), Zhx2 (B), and AFP (D) levels were measured by RT-qPCR and normalized to Sfrs4 (A) or ribosomal protein L30 (B, C, and D) mRNA levels and are displayed as relative expression levels. Means (SD), analyzed by Student’s t test, (n = 3–5/group). *P < 0.05; **P < 0.01; ***P < 0.001.

Zhx2 Activates Elovl3 Expression in Transfected HepG2 Cells and Binds Putative Elovl3 Control Regions in Adult Liver

We next investigated whether Zhx2 controls Elovl3 expression. Transient transfections were performed in human hepatoma HepG2 cells, which normally exhibit very low ELOVL3 mRNA levels. Cells were transiently transfected with a Zhx2 expression vector or empty vector control. After 48 h, cells were harvested, and mRNA and protein were prepared. Western blot analysis confirmed increased Flag-ZHX2 protein levels in transfected cells (Fig. 2A). The Elovl3 mRNA levels were markedly higher in Zhx2-transfected cells compared with control cells Fig. 2B, supporting our mouse data indicating that Zhx2 is a positive regulator of Elovl3 expression. In contrast, AFP, which is expressed at high levels in HepG2 cells and is repressed by Zhx2, exhibited reduced expression in Zhx2-transfected cells (Fig. 2B).

Figure 2.

Zhx2 activates Elovl3 expression in transfected HepG2 cells and binds putative Elovl3 control regions in adult liver. A, B: human hepatoma HepG2 cells were transiently transfected with a pcDNA vector containing flag-tagged Zhx2 (Zhx2-flag) or pcDNA empty vector control (pcDNA). After 48 h, cells were harvested and RNA and protein extracts were prepared. A: Western blots were probed using anti-Flag antibodies and reprobed with anti-actin antibodies to confirm equal lysate loading. B: RT-qPCR analysis was performed to measure ELOVL3 and AFP mRNA levels normalized to GAPDH. Data are representative of three separate experiments and is displayed as relative expression levels. C: adult male liver samples from Zhx2^WT and Zhx2^KO mice were analyzed by chromatin immunoprecipitation (ChIP) using anti-ZHX2 antibodies or control IgG and primers that generate amplicons from the Elovl3 promoter, 3′ DNase hypersensitive site (DHS) or 5′ DHS site. For all three regions, levels of IgG in Zhx2^WT and Zhx2^KO mice were set to 1. Data are from three separate experiments. *P < 0/05; **P < 0.01 (D). DHS from adult mouse heart (with red peaks, top track) and liver (with yellow peaks, bottom track) flanking the mouse Elovl3 gene are shown from analysis using the University of California, Santa Cruz (UCSC) Genome Browser.

To test whether ZHX2 directly binds to regions that control Elovl3 expression, ChIP experiments were performed with adult liver samples from Zhx2^WT or Zhx2^KO mice with anti-ZHX2 antibodies or control IgG. Although studies have identified several genes that are directly regulated by ZHX2, including AFP, Cyclin D1 (Ccnd1), and Mup20, a consensus ZHX2 binding site has not been identified. Furthermore, Elovl3 regulatory elements have not been characterized. Therefore, to identify candidate ZHX2 binding regions controlling Elovl3 expression, we searched for DHSs surrounding the Elovl3 gene using the ENCODE database. This analysis identified DHSs at the Elovl3 promoter and regions that are ∼7 kb (3′ DHS) and ∼21 kb (5′ DHS) upstream of the Elovl3 exon 1 (Fig. 2D); the promoter DHS site was seen in both liver and heart (where Elovl3 is not expressed), whereas the 5′ DHS and 3′ DHS are only present in the liver. Using primers to amplify these three regions, ChIP analysis from Zhx2^WT mice indicated that ZHX2 was enriched in the 3' DHS and Elovl3 promoter, but not in the 5' DHS site (Fig. 2C). No enrichment of these regions was found with anti-ZHX2 antibodies when liver samples from Zhx2^KO mice were used. These data suggest that ZHX2 directly activates Elovl3 expression by binding to one or several regulatory regions.

Male-Biased Elovl3 Expression in the Liver is Established Postnatally

A number of previously described Zhx2 targets, including those that are sex-biased, exhibit changes in expression in the perinatal liver (1, 6, 7). We characterized developmental changes in liver Elovl3 expression in male and female mice. In male mice, Zhx2 mRNA levels increase gradually between embryonic day 17.5 (e17.5) and postnatal day 28 (d28) and then have a nearly sevenfold increase between d28 and d56 (Fig. 3A). A very similar pattern is seen with Elovl3, with mRNA levels barely detectable at e17.5, showing a modest gain between d14 and d21, followed by a greater than six-fold increase between d28 and d56 (Fig. 3B). ZHX2 proteins levels are also low at embryonic and early postnatal stages and increase during development (Fig. 3C). A different pattern of Zhx2 and Elovl3 occurs in female liver. Zhx2 mRNA levels increase slightly between d7 and d14, followed by a greater increase between d21 and d28 (Fig. 3D), and there is a similar increase in ZHX2 protein with development stages (Fig. 3F). At d56, Zhx2 mRNA levels in male and female livers are nearly the same. Female hepatic Elovl3 mRNA levels are low but detectable at e17.5 and remain relatively constant until day 21, followed by a slight increase at day 28, then a steep decline between d28 and d56 (Fig. 3E). In contrast, AFP expression exhibits a consistent, dramatic decline in the postnatal liver in both male and female mice (6). Thus, hepatic Elovl3 expression is fully activated in males and nearly completely repressed in females between 4 and 8 wk of age, establishing the male-biased expression pattern that persists in the healthy liver of adult mice, as previously reported (11, 19).

Figure 3.

Male-biased Elovl3 expression in the adult liver is established between 4 and 8 wk of age. Liver mRNA was prepared from male and female B6C3F1 mice at embryonic day 17.5 (e17.5) and postnatal day 1 (d1), d7, d14, d21, d28, and d56. RT-qPCR analysis was performed to assess Zhx2 (A and D) and Elovl3 (B and E) mRNA levels in male (A and B) and female (D and E) mice, which were normalized to L30 and are displayed as relative expression levels. Means (SD), analyzed by ANOVA with Dunnett’s test for multiple comparisons; each timepoint compared with e17.5 expression. n = 4–10 mice/group. *P < 0.05; ***P < 0.001; ****P < 0.0001. Protein levels of ZHX2 during developmental stages are indicated by Western blot for male (C) and female (F) mice. n = 2 mice/sex/timepoint.

Zhx2 and Elovl3 mRNA Levels and ELOVL3-Synthesized Lipids Decrease in Regenerating Liver

Expression of many Zhx2 target genes becomes dysregulated in liver disease. In general, genes that are repressed by Zhx2 in the postnatal liver, including AFP, Gpc3, and H19, are activated during liver regeneration (1, 3, 20, 21). Conversely, genes that are postnatally activated by Zhx2, including certain Mups and Cytochrome p450 4a12 (Cyp4a12), exhibit decreased levels in regenerating livers (6, 7). We therefore assessed Zhx2 and Elovl3 expression in response to liver damage. Liver regeneration in response to acute injury is a process in which normally quiescent hepatocytes reenter the cell cycle to replace damaged cells that are lost due to injury and restore normal liver mass and function (22). To monitor changes in Zhx2 and Elovl3 expression during liver regeneration, the CCl₄ model was used (23). Adult male C3H and BL/6 mice were administered a single intraperitoneal injection of either CCl₄ or mineral oil (MO) as control and euthanized after 72 h when regeneration is robust. In both mouse strains, hepatic Zhx2 and Elovl3 mRNA levels were reduced in CCl₄-treated compared with control mice, although the decrease in Elovl3 mRNA levels is more significant in C3H mice (Fig. 4, A and C). In contrast, AFP mRNA levels increased in both strains three days after CCl₄ treatment; this increase is greater in C3H than in BL/6 mice (Fig. 4B). We then assessed whether decreased Elovl3 expression during regeneration was reflected in hepatic lipid composition. In the liver, VLCFAs are synthesized by various elongase enzymes with some redundancy but also with a great deal of specificity for carbon chain length and saturation (Fig. 4D). We analyzed free fatty acids and saponified lipids extracted from B6C3F1 mouse livers, which have robust Elovl3 repression and AFP induction similarly to C3H mice (data not shown) and found that regenerating livers have significantly lower concentrations of C22:1 monounsaturated and C20:0-C22:0 saturated VLCFAs (Fig. 4, F and H–I). There were no differences in C20:1, C24:0, or C24:1 VLCFAs (Fig. 4, F and I). Furthermore, there was a notable difference in liver lipid concentration depending on carbon length and desaturation: multiple lipid saturated and monounsaturated species C22 and shorter were decreased during regeneration, whereas C22 and longer polyunsaturated VLCFAs were generally increased (Table 3). We assessed changes in the expression levels of desaturase enzymes Scd1, Scd2, and Scd4 to elucidate why C22:1 VLCFAs were lower during regeneration without changes in the other monounsaturated lipids in the ELOVL3 pathway. In C3H regenerating livers, Scd1 mRNA levels were significantly lower than controls (Supplemental Fig. S1A), whereas Scd2 levels were higher (Supplemental Fig. S1B), and Scd4, which was expressed at very low levels in the liver, was unchanged (Supplemental Fig. S1C). These changes were specific to the C3H strain, as no changes were detected in BL/6 mouse livers.

Figure 4.

Zhx2 and Elovl3 mRNA levels and ELOVL3-synthesized lipids decrease in regenerating liver. A–C: adult male C3H or BL/6 mice were injected with vehicle (mineral oil, MO; black bars) or CCl₄ (gray bars) and euthanized after 3 days to assess mRNA changes during liver regeneration (n = 5/group). RT-qPCR was performed to analyze hepatic Elovl3 (A), AFP (B), and Zhx2 (C), and mRNA levels in both strains and are displayed as relative expression levels. D: schematic of VLCFA de novo synthesis pathway highlighting ELOVL3 metabolites, modified from Guillou et al. (24) Those in blue and green represent monounsaturated and saturated VLCFA, respectively. E–J: liver lipids were extracted and analyzed for lipid composition by LC-MS/MS and normalized to the weight of the tissue mass. Data are shown for VLCFAs synthesized by ELOVL3; C20:1 (D), C22:1 (E), C24:1 (F), C20:0 (G), C22:0 (H), and C24:0 (I). Data for all analyzed VLCFAs are shown in Table 3. Means (SD), A–C analyzed by two-way ANOVA with Tukey’s correction for multiple comparisons, E–J analyzed by Student’s t test; *P < 0.05; **P < 0.01, ***P < 0.001, ****P < 0.0001.

Table 3.

Saponified lipids in MO and CCl₄-treated livers

Lipids	MO, µg/mg Liver	CCl4, µg/mg Liver	Effect in CCl4	Significance (P < 0.05)
C14:0	0.7097	0.6340	Down
C16:0	5.6696	4.8819	Down
C18:0	5.3628	4.0986	Down	*
C18:1	4.2443	3.6730	Down
C18:2	5.7281	5.4713	Down
C19:0	0.3068	0.2354	Down	*
C20:0	0.3565	0.0978	Down	*
C20:1	1.6373	1.1540	Down
C20:2	1.6404	1.7800	Up
C20:3	2.0457	1.9632	Down
C20:4	5.4518	5.9013	Up
C20:5	0.6718	0.7168	Up
C21:0	0.0103	0.0061	Down	*
C22:0	0.0827	0.0453	Down	*
C22:1	0.2925	0.1498	Down	*
C22:2	0.1133	0.1159	No change
C22:3	0.2253	0.2889	Up
C22:4	1.2871	2.3613	Up	*
C22:5	1.0816	1.5071	Up	*
C22:6	2.2403	2.7369	Up	*
C23:0	0.0116	0.0112	No change
C24:0	0.0200	0.0182	Down
C24:1	0.2870	0.3540	Up
C24:2	0.0647	0.1594	Up	*
C24:3	0.0146	0.0455	Up	*
C24:4	0.2159	0.5545	Up	*
C24:5	0.1565	0.2751	Up	*
C24:6	0.1358	0.1806	Up
C26:1	0.0017	0.0036	Up	*
C26:2	0.0023	0.0070	Up	*
C26:3	0.0061	0.0058	No change
C26:4	0.0075	0.0279	Up	*
C26:5	0.0114	0.0437	Up	*
C26:6	0.0147	0.0274	Up	*
C34:0	0.0035	0.0019	Down
C36:2	0.0025	0.0015	Down

Increased Elovl3 Expression Reduces Cell Proliferation and Inhibits the S-G₂ Cell Cycle Transition

Previous studies demonstrated that Zhx2 inhibits the growth of liver tumor cell lines and blocks cell proliferation through direct binding to the promoters of Cyclins A and E, inhibiting their expression (25). Furthermore, Zhx2 inhibits the expression of Lpl and activates miR-24-3P to reduce SREBP1c levels; these activities block both lipid uptake and de novo lipogenesis and subsequent tumor cell expansion in vitro (26). Given the effects of Zhx2 and its other target genes on cell proliferation, and because of the dramatic decrease in Elovl3 mRNA levels we observed in development and regeneration, we next evaluated whether increased Elovl3 expression would influence cell growth and proliferation. Human HCC Huh7 cells, which have negligible endogenous ELOVL3 expression, were plated as monolayers and transfected with an ELOVL3 expression vector or an empty vector control (pcDNA), and viable cells were counted after 72 h. In four independent experiments, ELOVL3 overexpression reduced the number of cells by roughly 45% (Fig. 5A). Anchorage-independent growth is a signature feature of oncogenesis that can be gauged by cellular growth and expansion in soft agar. Huh7 cells transfected with either the ELOVL3 expression vector or pcDNA empty vector were seeded in media-supplemented soft agar, cultured for one week, and measured using a luminescent viability assay. In three independent experiments, ELOVL3 transfected cells exhibited a 34% reduction in cell viability compared with pcDNA3.1-transfected control cells (Fig. 5B). Taken together, these data indicate that elevated ELOVL3 levels diminish Huh7 cell viability.

Figure 5.

ELOVL3 overexpression reduces Huh7 cell growth in vitro. A, B: cells were transfected with a pcDNA empty vector control (pcDNA) or a human ELOVL3 expression vector (ELOVL3). A: Huh7 cells were maintained in culture as monolayers for 72 h then washed, trypsinized, and viable cells were counted. B: 24 h after transfection, Huh7 cells were trypsinized, seeded in media-supplemented agar and cultured for 1 wk. Cell viability was measured using a luminescence assay. In both sets of experiments, pcDNA-transfected cell counts were set to 1. Monolayer cultures were performed in four independent experiments; soft agar growth repeated in three independent experiments. Means (SD), analyzed by Student’s t test; **P < 0.01.

To further explore the effect of ELOVL3 expression on cell cycle progression, highly proliferative HeLa cells were cotransfected with a GFP expression plasmid along with either the ELOVL3 expression vector or control pcDNA and then synchronized in G₀. After release into growth media, cells were collected immediately or at multiple timepoints for up to 12 h and stained with propidium iodide. Using flow cytometry, GFP-positive cells were analyzed for DNA content. As early as 2 h after release, there was a notable increase in the proportion of ELOVL3-transfected cells in S-phase and a concomitant decrease in the G₂ cell population compared with control (pcDNA) transfected cells (Fig. 6, A–C). This trend continued to hold for the first 8 h after release. The cell cycle is a tightly regulated process that requires the coordinated expression of several cyclins at varying points in the cycle, as well as the association with the appropriate cyclin-dependent kinase (CDK) to permit progress to the subsequent phase. Based on the ELOVL3-mediated delay in the S phase-G₂ transition, we examined cyclin expression at timepoints after cell synchronization. In cells overexpressing ELOVL3, mRNA levels of cyclins D, E, and A are considerably reduced at early timepoints after release whereas levels of cyclin B, which is activated later in the cell cycle to promote mitosis, do not differ between ELOVL3 and control transfected cells (Fig. 6, D–G). Cyclin A expression and its association with CDK2 is a key regulator of S-phase transition. Control cells have a sharp increase in cyclin A expression 2 h after release, which is greatly blunted in ELOVLl3-transfected cells (Fig. 6F). This coincides with the detectable difference in the proportion of cells in S-phase and G₂ (Fig. 6, B–C).

Figure 6.

ELOVL3 overexpression stalls Hela cells at the S-G₂ transition. A–C: Hela cells cotransfected with GFP and either human ELOVL3 expression plasmid or empty vector were synchronized in G₀ by double thymidine block in serum-free media, then released into growth media. Cells gated for GFP expression were analyzed by FACS for cell cycle stage (G₀/G₁; S-phase; G₂) by propidium iodide staining at the designated timepoints. Synchronized cell cycle assays were performed in three independent experiments. Mean (SD), analyzed by two-way ANOVA with repeated measure and Bonferroni’s correction for multiple testing. D–G: RNA was extracted from cell lysates of the synchronized cells and RT-qPCR was performed to analyze changes in expression of Cyclin D (D), Cyclin E (E), Cyclin A (F), and Cyclin B (G) at the indicated timepoints after release. Expression of each cyclin is normalized to B2M and are displayed as the fold change compared with control cells at T₀. Means (SD) analyzed by Student’s t test at each timepoint; *P < 0.05.

We next queried publicly available RNA microarray datasets through the NCBI Gene Expression Omnibus (GEO) repository to further assess the impact of Elovl3 expression on cell proliferation and found two relevant datasets that included Elovl3 expression data. The first study investigated the role of suppressor of cytokine signaling 3 (Socs3) on liver regeneration after two-thirds partial hepatectomy in mice in which hepatocyte-specific deletion of Socs3 resulted in increased hepatocyte proliferation (27); liver Elovl3 expression was significantly reduced in these Socs3-deficient livers compared with controls (Supplemental Fig. S2A). The second study used GSK3 inhibitors to block proliferation of human leukemia RS4;11 cells (28); this study found ELOVL3 expression to be significantly higher in treated cells compared with controls (Supplemental Fig. S2B). These data corroborate our findings and support a role of Elovl3 in regulating cell cycle progression and cellular proliferation.

DISCUSSION

Zhx2 expression is required for the complete postnatal repression of numerous fetal liver genes, including AFP, H19, and Gpc3; the continued expression of these genes in the adult liver of BALB/cJ mice is due to a natural mutation in the Zhx2 gene in this strain (1, 3). In contrast to this repression, Zhx2 is required for the developmental activation of Mup genes and certain Cyp genes, both of which also exhibit sex-specific regulation (6, 7). Here, we demonstrate that Zhx2 is a positive regulator of Elovl3 in the adult male mouse liver. Elovl3 mRNA levels are lower in BALB/cJ mice compared with the BALB/cByJ substrain, which has a wild-type Zhx2 gene, and in adult BL/6 mice with a hepatocyte-specific Zhx2 deletion compared with wild-type littermate controls (Fig. 1). Consistent with these data, Zhx2 transfections in HepG2 cells significantly increase endogenous Elovl3 mRNA levels (Fig. 2). We also show that Zhx2 and Elovl3 mRNA levels are significantly reduced in regenerating liver (Fig. 4, A and C) with a concordant decrease in lipids synthesized by ELOVL3 (Fig. 4, E–J). Our data indicate that elevated Elovl3 expression diminishes the growth of Huh7 cells and stalls HeLa cell in S-phase (Figs. 5 and 6), suggesting that decreased Elovl3 expression during development and regeneration may enable increased cellular growth and proliferation.

Elovl3, like other putative Zhx2 targets, is developmentally regulated, and as with other known sex-specific Zhx2 target genes, male-biased Elovl3 expression becomes apparent between 4 and 8 wk of age (Fig. 3). In fact, Elovl3 is expressed higher in female liver during earlier developmental stages (Fig. 3, B and E). The mechanism by which Elovl3 decreases in female and increases in male livers is not known, although these changes occur concurrently with sexual maturation in mice that occurs 4–8 wk after birth. Previous studies in castrated mice demonstrated that androgens are required for male-biased hepatic Elovl3 expression and that female mice injected with 5α-dihydrotestosterone exhibited increased Elovl3 expression with circadian oscillations that mimicked male expression patterns (11, 19). Although the effects of estrogen on Elovl3 repression have not been tested directly, additional studies suggest estrogen receptor signaling, either through agonist treatment or compared with knockdown or deletion, maintains lower mRNA levels of Elovl3 in female mice and cells (29, 30). It is not clear whether crosstalk between Zhx2 and sex hormones is required for Elovl3 regulation, but our data support that ZHX2 directly binds the Elovl3 promoter and an upstream DHS to regulate transcription. Additional studies are needed to clarify the relationship between ZHX2, steroid hormone metabolism, and male-biased Elovl3 expression.

Zhx2 putative target genes AFP, Gpc3, and H19 are silenced at birth but transiently reactivated in regenerating liver after a single CCl₄ treatment (3, 20, 21, 31). The extent of this induction is less in BL/6 mice than in other strains, including C3H (21). This strain-specific difference is a monogenic trait that has been mapped to a locus on mouse chromosome 2 called α-fetoprotein regulator 2 (Afr2) (31). In an opposite expression pattern compared with these other Zhx2 targets, Elovl3 levels increase postnatally in male mice (Fig. 3B) and are reduced in regenerating liver (Fig. 4A). Interestingly, this Elovl3 silencing is less robust in BL/6 mice than in C3H liver (Fig. 4A), suggesting Afr2 also impacts Elovl3 repression in regenerating liver. The product of the Afr2 locus has not been identified, but the fact that Zhx2 targets also appear to be controlled by Afr2 during liver regeneration suggests an interaction between Zhx2 and Afr2 in the adult liver.

Hepatocyte proliferation is robust during embryonic development, slows in the perinatal period, and remains low in the healthy adult liver; liver regeneration reestablishes transient and rapid hepatocyte proliferation to replace damaged and dead cells (32). Elovl3 mRNA levels are extremely low in embryonic livers, increase postnatally (Fig. 3B) and decrease dramatically during liver regeneration (Fig. 4A). This expression pattern supports the possibility that Elovl3 may inhibit cellular proliferation and that its downregulation permits cellular expansion. Consistent with this, we find that ELOVL3 overexpression slows the in vitro growth of Huh7 cells (Fig. 5, A and B) and stalls HeLa cell cycle progression in S-phase (Fig. 6, B and C). In agreement with these primary data, analysis of published datasets supports that Elovl3 expression is lower during liver regeneration after partial hepatectomy in mice (Supplemental Fig. S2A), and that ELOVL3 expression is higher in leukemia cells with GSK3-inhibition to block proliferation (Supplemental Fig. S2B). The notion that Elovl3 influences cell cycle regulation is further supported by a study demonstrating that Elovl3 is transcriptionally activated by the tumor suppressor p53, which has a well-established role in cell cycle arrest (33). Our data suggest that VLCFAs synthesized by ELOVL3 contribute to its regulation of cell proliferation. The C20 to C24 saturated and monounsaturated VLCFA synthesized by ELOVL3 are often incorporated into ceramides and other classes of bioactive lipids. Ceramides are effective tumor suppressors by inducing apoptosis and reducing proliferation through cell cycle arrest (34–36). Reduced Elovl3 expression (Fig. 4A), resulting in lower hepatic C20-C22 VLCFAs levels, is seen in regenerating liver (Fig. 4, E–J) which may reduce certain bioactive lipids and allow increased cellular proliferation. There were also changes in mRNA levels of Scd1 and Scd2 in regenerating livers from C3H mice but not BL/6 mice (Supplemental Fig. S1). Interestingly, differential expression of these two genes in mouse liver has been noted during development: Scd1 is low in embryonic livers and is developmentally activated with highest expression occurring between d21 and d56, whereas Scd2 expression is highest at e18.5 and decreases postnatally (37). These expression patterns are consistent with what we would expect if they had a role in cellular proliferation, as Scd1 expression was lower in CCl₄-treated livers and Scd2 was increased compared with controls (Supplemental Fig. S1). More detailed metabolic tracing will be required to fully elucidate the lipid species and direct effects of ELOVL3-synthesized VLCFAs on proliferation and to investigate the potential role of other elongases in these processes. For example, ELOVL7 was found to contribute to prostate cancer by increasing production of saturated VLCFA that were synthesized into androgen steroids, serving as a potent prostate cancer cell growth signal (38). Future studies characterizing and tracing the lipid species produced by ELOVL3 and other elongases and desaturase enzymes will provide insight into this family of enzymes, VLCFAs, and their effects on cell proliferation.

DATA AVAILABILITY

Source data for this study are openly available at https://doi.org/10.6084/m9.figshare.23691063.

SUPPLEMENTAL DATA

Supplemental Figs. S1 and S2: https://doi.org/10.6084/m9.figshare.23691045.v5.

GRANTS

This work was supported by the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases (NIH/NIDDK) Training Grant T32DK007778 (to K.T.C.), American Heart Association predoctoral fellowship 10PRE4250000 (to H.R.), NIH/NIDDK Grant R01DK074816 and pilot project from the University of Kentucky Center of Research in Obesity & Cardiovascular Disease (P20RR021954 to B.T.S.), and the Shared Resource Facilities of the University of Kentucky Markey Cancer Center (P30CA177558 to B.T.S.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

K.T.C., H.R., and B.T.S. conceived and designed research; K.T.C., H.R., and J.J. performed experiments; K.T.C., H.R., J.J., M.L.P., and B.T.S. analyzed data; K.T.C., J.J., M.L.P., and B.T.S. interpreted results of experiments; K.T.C., H.R., and B.T.S. prepared figures; K.T.C. and B.T.S. drafted manuscript; K.T.C., J.J., M.L.P., and B.T.S. edited and revised manuscript; K.T.C., J.J., H.R., M.L.P., and B.T.S. approved final version of manuscript.