MXD3 Promotes Obesity and the Androgen Receptor Signaling Pathway in Gender-Disparity Hepatocarcinogenesis

Obesity is closely linked to metabolic diseases, particularly non-alcoholic steatohepatitis (NASH) or non-alcoholic fatty liver disease (NAFLD), ultimately leading to hepatocellular carcinoma (HCC). However, the molecular mechanisms of NASH-associated HCC (NAHCC) remain elusive. To explore the impact of Max dimerization protein 3 (MXD3), a transcription factor that regulates several cellular functions in disorders associated with metabolic diseases, we conditionally expressed Mxd3 proteins using Tet-on mxd3 transgenic zebrafish (MXs) with doxycycline (MXs + Dox) or without doxycycline (MXs − Dox) treatment. Overexpression of global MXD3 (gMX) or hepatic Mxd3 (hMX) was associated with obesity-related NAFLD pathophysiology in gMX + Dox, and liver fibrosis and HCC in hMX + Dox. Oil Red O (ORO)-stained signals were seen in intravascular blood vessels and liver buds of larval gMX + Dox, indicating that Mxd3 functionally promotes lipogenesis. The gMX + Dox-treated young adults exhibited an increase in body weight and visceral fat accumulation. The hMX + Dox-treated young adults showed normal body characteristics but exhibited liver steatosis and NASH-like phenotypes. Subsequently, steatohepatitis, liver fibrosis, and NAHCC were found in 6-month-old gMX + Dox adults compared with gMX − Dox adults at the same stage. Overexpression of Mxd3 also enhanced AR expression accompanied by the increase of AR-signaling pathways resulting in hepatocarcinogenesis in males. Our results demonstrate that global actions of Mxd3 are central to the initiation of obesity in the gMX zebrafish through their effects on adipogenesis and that MXD3 could serve as a therapeutic target for obesity-associated liver diseases.


Introduction
MAX dimerization protein 3 (MXD3) is a basic helix-loop-helix transcription factor associated with the MYC/MAX/MAD transcriptional network [1]. The MYC/MAX/MAD network was shown to be important for the regulation of cell proliferation, differentiation, and apoptosis [2]. MXD3 is a unique MAD family member protein, which competes with MYC for heterodimerization with the cofactor MAX, and is a functional MYC antagonist [1,3]. MYC has been shown to promote proliferation in several cell types, and the MAD proteins function as a "transcriptional repressor" to strongly antagonize MYC activity, resulting in inhibition of cellular proliferation and transformation [3][4][5][6]. Other members of the MXD family (MXD1, MXI1, and MXD4) are abundantly present in post-mitotic cells, while MXD3 is discovered in mitotic cells during the S-phase of the cell cycle [7,8].
MXD3 deficiency resulted in an enlarged sensitivity to apoptosis in return for DNA impairment [9]; however, MXD3 transcripts and proteins have also been recognized in proliferating cells [5,7]. These findings suggest that MXD3 proteins may function not only in differentiation, but also in processes involved in other aspects of cellular growth control. Aberrant expression levels of MXD3 have been seen in numerous cancers. Previously, MXD3 overexpression was detected to elevate in glioblastoma [10], medulloblastoma [11], acute lymphoblastic leukemia cell proliferation [12], renal cell carcinoma [13], and hepatocellular carcinoma (HCC) [14]. These varying results relate to the functions of MXD3 imply tissue-specific or pathophysiological roles of MXD3, depending on different tissues and developmental stages.
Recent studies have also demonstrated that prolonged MXD3 expression could increase lipogenesis and adipogenesis and have an effect on oxidative stress [15], which also drive lipid deposition and boost obesity in zebrafish [16]. Depletion of mxd3 conquers visceral fat in diet-induced obese (DIO) zebrafish and Mxd3 ameliorates adipocyte differentiation in the DIO zebrafish. A molecular analysis revealed that the expression of adipogenic transcriptional factors, including peroxisome proliferator activated receptor gamma (ppar-γ), C/EBP-α/β/γ, caveolin 1 (cav1), and collagen domain containing b (adipoqb) decrease by mxd3 depletion in 3T3-L1 cells and DIO zebrafish [17]. Obesity is defined as unusual fat addition in adipose tissue, and the liver closely communicates with adipose tissue [18,19]. MXD3 appears to be linked with various human cancers and adipogenesis. Several large-scale clinical studies have shown a connection between individuals with "plain" obesity and a larger risk of HCC, when compared to "non-obese" people [20][21][22]. MXD3 is also suggested as a diagnostic biomarker for HCC [14]. Considering these data, the role of MXD3 in liver cancers is interesting, and its role in obesity and HCC is worth investigating.
Zebrafish has recently become a significant model organism for studying obesityrelated human diseases [16,[23][24][25][26][27][28]. The aim of this study was to discover the Mxd3 relative to signaling pathways, which influence metabolic physiology in zebrafish models. We produced Mxd3 transgenic zebrafish to analyze the functional roles of the zebrafish Mxd3 protein during development. Here, we present evidence that Mxd3 promotes an early onset of adipocyte proliferation. Moreover, these findings shed light on the pivotal role of Mxd3 in obesity-related HCC, and could serve as a guide for further preclinical research.

Treatment with Doxycycline (Dox)
Zebrafish embryos and juvenile adults were treated by immersion in Dox at 25 µg/mL (Selleck Chemicals, Houston, TX, USA) in six-well plates and 3-L tanks, individually, and the water was exchanged daily.

Alamar Blue Metabolic Activity Assay
The datasets of metabolic activity assays were conducted as previously described [36].

Blood Analysis
Blood was collected, as reported by previously published research by Renquist [37]. The collected blood samples were pooled and used for quantification of leptin (Fish Leptin ELISA Kit, MBS021271, MyBioSource, San Diego, CA, USA) and adiponectin (Fish Adiponectin ELISA Kit, MBS098337, MyBioSource San Diego, CA, USA) by ELISA, and followed the manufacturer's instructions. Data were expressed as percent variances (∆%) of fish weight gain that occurred during 8 weeks of normal feeding.

Growth Rate
Growth rate studies were performed" as described previously [16,28]. The growth curve was measured monthly from 2 months post fertilization (mpf) to 12 mpf. All groups containing 25 fish (mixed sex) were picked randomly. Body weights (BW) were measured to 0.01 g monthly, a standard metric ruler was used to measured body lengths (BL) to 1 mm, and the length from the tip of the snout to the caudal peduncle was recorded. Body weights and body lengths of anesthetized juveniles and adults were recorded for the calculation of the body mass index (BMI).

Western Blot Analysis
The protein lysate samples from zebrafish liver tissue were measured by western blotting, as described previously [30,31]. The incubation of PVDF membrane was in antibodies to the following antibodies

Sex Hormone Treatment
These experiments were performed on histological sections according to a method described by Li et al. [38,39]. For sex hormone treatment, 5 µg/mL 17β-estradiol (E2) (E4389, Sigma-Aldrich, St. Louis, MO, USA) or 11-ketotestosterone (11-KT) (Steraloids, Newport, RI, USA) was used for four mpf zebrafish. During Dox induction, E2 or 11-KT was used to treat zebrafish for 2 weeks. After removing Dox, male zebrafish were treated by E2 or 11-KT for 1 week. Histological analysis was accomplished in the first, third, and ninth month post-treatment.

Statistical Analysis
All data are represented as the mean ± SEM. GraphPad Prism 8.0 software (GraphPad, San Diego, CA, USA) was used for statistical analyses. A p-value of < 0.05 was considered statistically significant.
Three gMX (gMX1, gMX2, and gMX3) and four hMX transgenic lines (hMX1, hMX2, hMX3, and hMX4) were selected on the basis of their mxd3 expression using qRT-PCR ( Figure 1B,E). Compared with that in the wild type (WT), mxd3 was significantly overexpressed, 5.8-, 28.7-, and 14.1-fold in gMX1, gMX2, and gMX3 with doxycycline treatment (+Dox), respectively, and no mxd3 was detected in control groups without Dox (−Dox). gMX2 displayed no noticeable phenotypic variations in other lines at larval stages compare to WT ± Dox ( Figure 1C, panel 1); GFP fluorescence was seen in gMXs ± Dox larvae ( Figure 1C, panel 2). On the other hand, no mCherry fluorescence was seen in gMX2-Dox larvae ( Figure 1C, panel3), whereas the strong global mCherry as well as GFP fluorescence was present in gMX2+Dox larvae ( Figure 1C, panel 4). For hMXs, 11.7-, 39.6-, 28.2-, and 9.8-fold mxd3 expression were detected in the livers of hMX1, hMX2, hMX3, and hMX4, respectively, compared with that of WT controls. However, there was slight leakage (−Dox treatment for 4.7-fold expression) in the hMX2. hMX3 displayed no noticeable phenotypic variations in other founders at larval stages compare to WT ( Figure 1F, panel 1); only GFP fluorescence was observed in the livers of hMX3−Dox larvae ( Figure 1F, panel 2). The clear mCherry and GFP fluorescence was observed in hMX3 + Dox larvae, but no mCherry fluorescence was seen in hMX3 − Dox larvae ( Figure 1F, panels 3 and 4). These results suggest that the gMX2 and hMX3 lines expressed the highest levels of mxd3 expression without leakage in induction. Thus, in the present study, a majority of experiments were performed using gMX2 and hMX3 zebrafish. (F) Liver-specific inducible Mxd3 expression in the hMX3 at 7 dpf. Transgenic larvae were treated with 25 µg/mL Dox from 2 to 7 dpf. Scale bar: 100 µm.
We assumed that an early onset of adipogenesis occurred in the gMXs larvae and NAFLD phenotypes in the hMXs larvae. Molecular analysis of the gMXs larvae showed upregulated genes involved in both lipogenesis and adipogenesis; however, upregulation of only the lipogenic genes was observed in hMXs larvae ( Figure 2C,D). Furthermore, to evaluate the probable roles of Mxd3 in the metabolism of gMXs or hMXs in further examinations on energy expenses that might influence obesity, we carried out the Alamar Blue metabolic assay in 6 dpf larvae. gMXs+Dox larvae displayed significantly lower metabolic rates (56-72%), and hMXs exhibited metabolic rates of 68-84% compared with control groups ( Figure 2E). The assay showed increasing signals of the MXs+Dox corresponding to incubation times compared with the controls, and the results confirmed that lipids were accumulated as an energy reserve, as reflected by the occurrence of hepatic steatosis and adipogenesis. These data indicate that mxd3 overexpression can provoke lipid accumulation in zebrafish larvae.

Adult gMXs+Dox Are Overweight and Have Increased Adiposity
To investigate whether overexpression of Mxd3 has a physiological effect on obesity, we inspected the appearance of fat tissues of gMX adults and revealed that they displayed a vivid "reply" to weight gain or obesity. gMXs + Dox adults were overweight and larger compared with WT ( Figure 3A). After 9 months, we discovered that gMXs + Dox zebrafish showed a drastically keen response to the normal diet ( Figure 3B). In line with the growth rate, the body mass index (BMI) of gMX2 + Dox and gMX3 + Dox adults was intensely amplified within the 9 months of feeding ( Figure 3C). This intensification of weight gain corresponded with a manifest increase in the internal organs and visceral fat in the gMX2 + Dox adults ( Figure 3D-F). We also noticed an enlarged cell mass of subcutaneous and visceral adipocytes ( Figure 3G,H, left) in the gMX2 + Dox group; an increased percentage of gMX2 + Dox exhibited visceral adipocyte hyperplasia compared with the control groups ( Figure 3G,H, right). qRT-PCR analyses showed that mxd3 overexpression increased in mature adipocyte marker genes ( Figure 3I). In addition, we examined whether Mxd3 modulates adipokines in the adipose tissue of the gMX2 group and primarily accompanies obesity effects. We observed that the adiposity of gMX2 also accompanied the upregulated expression of leptin and downregulated expression of adiponectin ( Figure 3J). Thus, Mxd3 overexpression led to hyperplasia of adipose tissue in zebrafish, thereby involving an impact of Mxd3 on growth and obesity.
We next checked whether the early liver steatosis would develop to NAFLD in adult hMXs + Dox. Nearly 60% of the hMXs + Dox at 4 mpf started to display pale yellow livers compared with the control groups ( Figure 4B). Furthermore, H&E staining analysis discovered that liver sections of hMXs + Dox adults displayed hepatocytes with various grades of lipid cytoplasmic vacuolation ( Figure 4C; panels 2, 4, 6, and 8). On the other hand, regular liver sinusoids and hepatocytes displayed a vigorous cytoplasm and clear nucleus in control groups ( Figure 4C; panels 1, 3, 5, and 7). Results of the H&E staining were verified by ORO staining, which apparently uncovered macrovesicular steatosis and massive lipid deposits in the hepatocytes of hMXs + Dox adults ( Figure 4D; panels 2, 4, 6, and 8) compared with those in the control groups ( Figure 4D; panels 1, 3, 5, and 7). hMXs + Dox adults displayed upregulated expressions of the selected liver lipogenic genes compared with that in the control groups ( Figure 4E). As expected, the levels of hepatic triglycerides, cholesterol, and oxidative stress observed in hMXs + Dox were significantly greater than those in the controls ( Figure 4F). These results indicate that Mxd3 overexpression can develop various grades of liver steatosis in the hMXs + Dox adults.
As expected, the livers of gMX2 + Dox and hMX3 + Dox adults displayed upregulated expression of the selected inflammatory genes ( Figure 5D) and fibrotic genes ( Figure 5E) compared with their controls ( Figure 5D). Taken together, these results show that hepatic Mxd3 overexpression gives rise to the progression of NASH phenotypes.

Chronic Effects of Hepatic Mxd3 Expression on NAHCC in Male MXs + Dox
As hMX3 + Dox fish showed marked activation via inflammation (or oncogenesis) and accelerated NASH, we analyzed whether the livers of hMX3+Dox group were inclined to cancer progression from the beginning. By observing the MXs at 10 mpf, development of HCC was seen to occur earlier than 9 mpf in a fraction of male fish of hMX3s (1 and 2) + Dox ( Figure 6A, panels 3 and 4). Distinct NASH phenotypes (such as steatosis, hep-atitis/lymphocytic infiltration, portal fibrosis, and cholestasis) were seen in both males and females of gMX2 + Dox ( Figure 6A, panels 2 and 6), as well as females of hMX3s (1 & 2) + Dox ( Figure 6A, panels 7 and 8) than in WT controls ( Figure 6A, panels 1 and 5). Furthermore, main hepatic vessels were seen in both hMX3 ± Dox ( Figure 6B, panels 1-4), while only reticular veins were witnessed in the hMX3+Dox liver as a result of hepatic angiogenesis ( Figure 6B; panels 2 and 4, insets). Masson's trichrome stain revealed that clear HCC ( Figure 6C, panel 2), steatohepatitis ( Figure 6C, panel 3), and fibrosis concomitant with HCC (NAHCC) ( Figure 6C, panel 4), and cirrhosis concomitant with HCC ( Figure 6C, panel 4) was observed in male hMX3s (1-3) +Dox livers than in hMX3 − Dox ( Figure 6C, panel 1). Furthermore, 60% male and 20% female hMX3 + Dox fish showed evidence of NAHCC. Quantification of histological inspection of the livers exposed that all the MXs + Dox had developed varying grades of HCC (from 8% to 55%). A total of 55% male hMX3 + Dox and 37% male gMX2 + Dox fish had advanced HCC, only 18% hMX3 + Dox and 8% gMX3 + Dox females had clear HCC, and the rest had NAFL or NASH compared with MXs − Dox fish ( Figure 6D). Therefore, there was a dominant NAHCC development during HCC progression in male hMXs than in female hMXs.

Mxd3 Enhances Androgen Receptor (AR) Expression of NAHCC Progression in hMXs + Dox
There is evidence with respect to the expression of AR in HCC [40][41][42][43]. AR-dependent gender differences in liver cancers have also been studied in mice [44,45] and zebrafish [38,46]. Therefore, we sought to discover the connection between Mxd3 and AR in hMXs + Dox. Because the predominance of HCC in male fish was more pronounced in hMXs + Dox than in WT controls ( Figure 6D), we proposed that zebrafish Mxd3 might be involved in a similar androgen-signaling pathway to accelerate hepatocarcinogenesis in male hMXs. To determine the possible biological functions underlying Mxd3-regulated AR expression, we investigated the influence of Mxd3 on the regulation of AR activities during NAHCC progression in hMXs + Dox. As shown in Figure 7, a positive indicator between Mxd3 and AR expression was seen in the immunohistochemical analysis; Mxd3 and AR protein levels in both the male and female hMX3+Dox fish were significantly higher than in both the sexes of hMX3 − Dox ( Figure 7A). We also observed transcriptional upregulation of the selected androgen-responsive genes (AREs) involved in the AR/androgen-signaling pathway ( Figure 7B). Moreover, consistent with western blot analyses, we observed an affirmative correlation between Mxd3 and AR protein expression and the selected ARE proteins, accompanied by higher contents of AR protein in both the male and female hMX3 + Dox fish than in both the sexes of hMX3 − Dox ( Figure 7C). Thus, Mxd3 overexpression upregulated the AR at both the mRNA and protein levels in hMX3 + Dox. These data suggest that Mxd3 is positively correlated with AR in the hMX3 + Dox and further confirm that Mxd3 plays a conspicuous role in AR-mediated NAHCC ( Figure 7D).

Sex Hormone Treatment Affected AR-Mediated NAHCC in Male MX3 + Dox Fish
It has been reported that sex hormones affect HCC progression in the zebrafish [38,39,46] and rat [47] models as well as in humans [48][49][50]. In the current study, it was witnessed that male hMX3+Dox fish developed significantly aggressive HCC than female HCC hMX3 + Dox and controls fish ( Figure 6D) and that Mxd3 played an important role in AR-mediated NAHCC ( Figure 7D).
Hence, studies on treatment with sex hormones were carried out to evaluate their roles in NAHCC progression. Male MXs ± Dox and WT ± Dox adults were treated with either 17-estradiol (E2) or 11-ketotestosterone (11-KT). Representative gross observations of normal liver and HCC samples are presented in Figure 8A. After sex hormone treatment, the relative tumor sizes of male MXs adult ( Figure 8A, panels 4 and 6) showed significant decreases with E2 treatment (Figure 8A, panels 10 and 12) compared with those of vehicle controls ( Figure 8A, panels 1-3 and 5). As expected, the tumor sizes significantly increased only in the MXs + Dox after 11-KT treatment ( Figure 8A, panels 16 and 18) compared with those of vehicle controls ( Figure 8A, panels 13-15 and 17). Meanwhile, H&E staining also revealed similarly decreased tumor sizes in male MXs + Dox fish after E2 treatment and increased tumor sizes in these fish after 11-KT treatment ( Figure 8B). A total of 38.7% male gMX2 + Dox and 53.3% male hMX2 + Dox fish had advanced HCC with control treatment, only 23.1% gMX2 + Dox and 30.3% hMX2 + Dox males had clear HCC with E2 treatment, the dramatic growth of HCC population in gMX2 + Dox (71.2%) and hMX2 + Dox (86.4%) after 11-KT treatment, and the rest had fatty liver disease (FLD) compared with MXs−Dox fish ( Figure 8C). PCNA staining was also carried out to survey the hepatocyte proliferation in MXs after sex hormone treatments ( Figure 8D,E). After E2 treatment, the PCNA-positive staining reduced markedly in male MXs + Dox, indicating restrained cell growth in the liver tumors by the E2 treatment. After 11-KT treatment, the PCNA-positive cells increased significantly in male MXs + Dox, indicating that 11-KT boosted tumor growth. However, in vehicle control fish, both E2 and 11-KT showed no effect or had much less effect on cell proliferation. The protein expression level of PCNA showed a significant change in male hMX3+Dox livers with E2 or 11-KT treatment, which indicates a positive correlation with AR signaling ( Figure 8F). Overall, the above assessments reveal that the E2 treatment decelerated whereas 11-KT treatment enhanced the HCC progression in male MX3 + Dox.

Discussion
We analyzed the MXD3 expression of the MYC/MAX/MAD network during adipogenic differentiation and the process of NAFLD in zebrafish. We discovered that, although MXD3 expression was initially linked to both lipogenesis and adipogenesis (Figure 2), it significantly induced adiposity and NAFLD (Figures 2 and 4) during zebrafish development at juvenile (< 30 dpf), and young adult (< 4 mpf) stages. Most notable was the MXD3 overexpression, which induced strong expression of AR signaling ( Figure 7) and dominant NAHCC development during HCC formation in male rather than female MXs ( Figure 6). Our findings confirm and extend findings from previous studies that describe the expression of mxd3 network genes linked to lipogenesis and carcinogenesis in NAHCC.
The expression levels of MXD3 were decreased three-fold in adult skeletal muscle tissues than in such tissues in the fetal period in Qinchuan cattle [51]. Depletion of mxd3 led to a decrease in cell numbers, indicating that MXD3 is necessitated by cell proliferation [11,52]. By contrast, MXD3 overexpression is adequate to stimulate proliferation in mouse granule neuron precursors(GNPs) [52]. Surprisingly, MXD3 overexpression negatively regulated differentiation of mouse B cells [53]. Constant MXD3 overexpression, however, in mouse GNPs and human medulloblastoma cells, resulted in restrained cell proliferation as a result of the activated apoptosis [11,52]. Shimada et al. revealed that MXD3 is a new regulatory gene for adipogenesis in obese people. MXD3 expression is markedly increased in visceral fat in obese people as well as in DIO zebrafish model [17]. In the current study, Mxd3 overexpression increased both early ( Figure 2D) and mature adipocyte marker genes ( Figure 3J). Mxd3 overexpression can increase the adipogenic and lipogenic function (Figure 2). These results agreed with those reported previously that MXD3 could disrupt the MAX/MYC heterodimer, and reduce MYC activation to promote adipogenesis [15,17]. Thus, adipose tissue hyperplasia of gMXs fish (Figure 3) mainly occurred due to lipogenesis and adipogenesis of preadipocytes into adipocytes.
Low levels of serum adiponectin have been revealed to connect with many cancers, including breast, prostate, kidney, pancreatic, gastric, and colon cancers [54,55]. Adiponectin deficient mice treated with choline-deficient L-amino acid-defined diet [56] or high fat diets acquired liver cirrhosis and tumors [57]. Adipokines appear to have an imperative role, not only in the NAFLD progression, but also in NAHCC [58][59][60]. In our study, we found that MXD3 modulates adipokines in the adipose tissue of gMX2+Dox-treated zebrafish, and is accompanied primarily by obesity effects. The adiposity of gMX2+Dox correlated with increased expression of leptin and a decrease of adiponectin ( Figure 3J). These results may have been due to oncometabolic stress related to NAFL ( Figure 4) and NASH-related carcinogenesis in gMX2+Dox (Figures 5 and 6).
Barisone et al., using an online bioinformatics platform to inquire cancer datasets, revealed that MXD3 is expressed in various cancers [61]. Xu et al. found that 10 hub genes, including MXD3, were aligned with HCC progression based on a gene co-expression network analysis [14]. Moreover, Ngo et al. also found that, across the datasets, MXD3 was highly overexpressed in HCC with a 2.88-fold change in relation to normal tissues [10]. The MXD3 promoter sequence region has a tendency to be hypomethylated in liver cancer in respect to adjacent tissue samples [10]. In several other studies across several models, it has been revealed that both depletion and MXD3 overexpression causes reduced cell proliferation [11,12,61]. These data suggest that it is important for MXD3 to maintain its cellular concentrations to function optimally. In this study, we discovered the possibility of upregulation of the mxd3 through transgenic overexpression of mxd3 in zebrafish and its liver. Specifically, Mxd3 overexpression leads to the progression of NAFLD phenotypes, including NASH, fibrosis, and HCC (Figures 6 and 7).
HCC mainly influences the males, with a higher prevalence in male participants than in females [62,63]. The causes for the gender differences are complex in metabolic cytokines, such as adipokines and hepatokines that may cooperate with each other for maintaining liver health [64,65]. The zebrafish liver cancer also has statistically significant sexual dimorphism with male dominance; this has been confirmed by comparison of the male and female kras V12 [38,46] and Myc/xmrk transgenic zebrafish [39]. In this study, we showed a similar sexual dimorphism in the liver cancer progression of hMX3 ( Figure 6). Specifically, male MXs+Dox transgenic zebrafish developed significantly severe HCC than female MXs+Dox transgenic zebrafish ( Figure 6D). We should note that the AR activation is aligned with human malignancy, such as prostate cancer, HCC, and pancreatic cancer, and that AREs, such as TGF-beta, CCRK, GRP78, and VEGF, play indispensable roles in the AR-mediated carcinogenesis [66][67][68]. Li et al. specifically depleted the ar gene in the liver of kras V12 transgenic zebrafish and observed alleviated liver tumor progression in these zebrafish [38]. In agreement with these findings, we found that overexpression of MXD3 enhances AR expression and is accompanied by an increase in the AR-signaling pathways, resulting in hepatocarcinogenesis. (Figure 7C,D). In line with these results, we showed that sex hormones show important roles in HCC development. In particular, E2 could retard HCC (also NASH and cirrhosis) progression, particularly in both the male gMXs + Dox zebrafish ( Figure 8A,B), while 11-KT could promote HCC progression predominantly in male hMX3+Dox zebrafish ( Figure 8A-C). Additionally, cell proliferation in gMXs + Dox increased notably after 11-KT treatment and decreased markedly after E2 treatment, compared with those in the control groups ( Figure 8D, E). In line with our findings, the response level of ARE proteins was accompanied by sex hormone treatment ( Figure 8F). Generally, our results suggest that sex hormones affect NAHCC progression in the MXs zebrafish.

Conclusions
We showed that MXD3 is one of the factors that contributes towards an increase in visceral fat via the proliferation of preadipocytes and early adipogenesis. Overexpression of mxd3 caused somatic growth and obesity, which then deteriorated the systemic lipid metabolic dysfunction (Mxd3 modulated adipokines in the adipose tissue of gMX2), ultimately resulting in dyslipidemia, liver steatosis, and NASH in MXs+Dox-treated zebrafish. Our results further show that mxd3 expression increases both adipogenesis and carcinogenesis in AR-signaling associated NAHCC, indicating that the function of MXD3 on HCC growth is most closely linked to oncogenic AR signaling in the zebrafish liver. Our results also indicate that MXD3 plays a significant role in the regulation of adipokines and enhancement of the AREs, and then oncogenic AR signaling to join in the NAHCC progression. Thus, the functions of MXD3 on the modulation of adipokines and regulation of AR signaling could provide a novel therapeutic strategy for HCC.

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets used in the current study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare that they have no competing financial interests.