Hepatotoxin-Induced Changes in the Adult Murine Liver Promote MYC-Induced Tumorigenesis

Background Overexpression of the human c-MYC (MYC) oncogene is one of the most frequently implicated events in the pathogenesis of hepatocellular carcinoma (HCC). Previously, we have shown in a conditional transgenic mouse model that MYC overexpression is restrained from inducing mitotic cellular division and tumorigenesis in the adult liver; whereas, in marked contrast, MYC induces robust proliferation associated with the very rapid onset of tumorigenesis in embryonic and neonatal mice. Methodology/Principal Findings Here, we show that non-genotoxic hepatotoxins induce changes in the liver cellular context associated with increased cellular proliferation and enhanced tumorigenesis. Both 5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) and carbon tetrachloride (CCl4) cooperate with MYC to greatly accelerate the onset of liver cancer in an adult host to less than 7 days versus a mean latency of onset of over 35 weeks for MYC alone. These hepatotoxin-enhanced liver tumors grossly and histologically resemble embryonic and neonatal liver tumors. Importantly, we found that MYC overexpression is only capable of inducing expression of the mitotic Cyclin B1 in embryonic/neonatal hosts or adult hosts that were treated with either carcinogen. Conclusion/Significance Our results suggest a model whereby oncogenes can remain latently activated, but exposure of the adult liver to hepatotoxins that promote hepatocyte proliferation can rapidly uncover their malignant potential.


Introduction
MYC is a proto-oncogene that regulates normal cellular growth, proliferation, apoptosis, and differentiation [1]. MYC is frequently overexpressed or mutated in human cancers and is thought to contribute to tumorigenesis by inducing autonomous cellular growth and proliferation, blocking differentiation, and inducing genomic instability [1][2][3][4][5]. Aberrant MYC oncogene expression has been frequently observed in primary human and rodent liver tumors [6][7][8]. Importantly, MYC expression is sufficient to induce tumor formation in murine models, with liver-specific overexpression directed by the albumin enhancer/promoter or the alpha-1antitrypsin promoter resulting in HCC [9,10]. However, the tumor incidence in these mice is relatively low and the latency is long. Notably, when MYC is co-expressed in the murine liver with either Transforming growth factor alpha (TGF-alpha) or E2F transcription factor 1 (E2F1), the onset of tumorigenesis is significantly accelerated [11][12][13]. These data suggest that MYC activation can be significantly augmented by various complementary stimuli, providing a system by which modifiers of the tumor phenotype may be revealed.
Epidemiological studies in humans, as well as data from animal models, support the idea that the liver is more susceptible to neoplastic transformation during states of liver growth and regeneration [14,15]. Patients with alcohol-induced cirrhosis of the liver were found to be at greater risk of developing hepatocellular carcinoma than those without cirrhosis [16]. Similarly, various hepatotoxins have been reported to enhance liver tumorigenesis in many contexts, including mice that are transgenic for hepatitis b virus (HBV) protein or TGF-alpha overexpression [17][18][19]. Together, these data support the model of accelerated liver tumorigenesis as a result of changes in the liver. Notably, different hepatotoxins damage distinct regions of the liver lobule and have been shown to elicit different cells in the hepatic lineage to restore hepatocyte loss during tissue regeneration, thereby influencing the cellular origin of HCC ( Figure 1A) [20,21]. Based on the differential effect of various toxins, some chemical carcinogenesis models have favored the hepatocyte as the precursor to HCC, while others have implicated liver progenitor cells (oval cells) as the cell of origin for HCC [22][23][24].
Recently, we have described the use of the tetracyclineregulated (Tet) system to develop a conditional transgenic model of MYC-induced HCC [25,26]. Using this system, we found that MYC's ability to induce mitotic division and tumorigenesis in the liver is developmentally regulated [26]. In the permissive context of an embryonic or neonatal liver, in which hepatocytes are actively undergoing mitotic cellular division, MYC overexpression further increased cellular proliferation and resulted in the almost immediate onset of liver tumorigenesis. In contrast, when overexpressed in mitotically quiescent adult murine hepatocytes, MYC was capable of inducing cellular growth and DNA replication, but was prohibited from inducing mitotic division and resulted in HCC only after a prolonged latency. Hence, the cellular and developmental context of the host appears to play a critical role in defining the ability of MYC to induce tumorigenesis in the liver.
Non-genotoxic hepatotoxins are examples of agents that could be responsible for changes in the liver that would enhance the ability of oncogenes such as MYC to induce liver tumorigenesis. CCl 4 is a well characterized carcinogen that causes centrilobular destruction of hepatocytes and triggers mature hepatocytes to proliferate and reconstitute the lost liver mass ( Figure 1A) [27,28].
Other agents such as DDC have also been reported to cause liver damage associated with the proliferation of oval cells in the periportal area of the hepatic lobule, however, as of yet have not been reported to be carcinogens ( Figure 1A) [29][30][31]. Here we show that both DDC and CCl 4 similarly facilitate MYC-induced initiation of hepatocyte proliferation and tumorigenesis in the adult liver. Our findings suggest a model whereby external stimuli may unveil latent oncogene activation in the mature liver.

DDC and CCl 4 treatment accelerate MYC-induced HCC in adult hosts
To examine the effects of DDC on MYC's ability to induce tumorigenesis in the adult liver, 6-week-old adult mice were fed a diet containing 0.1% DDC [29]. MYC was induced in one cohort of mice (MYC ON/DDC) and kept inactive in controls (MYC OFF/DDC) ( Figure 1B). To distinguish between DDC-specific and other hepatotoxin effects, a separate cohort of transgenic mice were injected with CCl 4 twice weekly for the duration of the experiment (4 months), either in the presence (MYC ON/CCl4) or absence (MYC OFF/CCl4) of MYC overexpression ( Figure 1B). As an additional control, adult mice in which MYC was overexpressed by itself (MYC ON) or kept inactive (MYC OFF) were also monitored ( Figure 1B). Remarkably, mice that overexpressed MYC in conjunction with either DDC or CCl 4 treatment (MYC ON/DDC and MYC ON/CCl4) developed liver cancers at a substantially accelerated onset as compared to the control MYC ON mice ( Figure 2). Specifically, the rate at which all MYC ON/DDC and MYC ON/CCl4 animals were moribund with tumor burden was 31 and 40 days, respectively, compared to 183 days for MYC ON by itself. Indeed, by the time hepatotoxinassociated mortality was at 100%, the MYC ON cohort had yet to demonstrate morbidity. Interestingly, the rapid onset of liver tumorigenesis in MYC ON/DDC and MYC ON/CCl4 mice was  similar to that observed when MYC was overexpressed in neonatal livers ( Figure 2). Mice in which MYC was inactive (MYC OFF/ DDC, MYC OFF/CCl4 and MYC OFF) remained tumor free for the duration of the experiment as expected (4 months, Figure 2).

DDC and CCl 4 cooperate with MYC to induce diffuse HCC
Since tumor latency was dramatically reduced upon hepatotoxin treatment, it was necessary to investigate whether the resulting disease appeared phenotypically similar. Upon gross examination, MYC ON/DDC and MYC ON/CCl4 mice developed tumors associated with three-fold enlarged livers with a multitude of coalescing tumor nodules ( Figure 3F, H). As observed with tumor latency, hepatotoxin-accelerated HCC appeared similar to the MYC-induced neonatal livers, demonstrating a diffuse tumor phenotype ( Figure 3D) [26]. In contrast, adult mice that overexpressed MYC by itself developed enlarged livers with multiple discrete tumor nodules, as previously described ( Figure 3B) [26]. Although MYC OFF/DDC mice did not develop tumors, they exhibited progressive abdominal enlargement during the duration of DDC feeding, reaching a peak at about 4 weeks. Grossly, this abdominal growth was associated with 2-fold enlarged livers ( Figure 3A, E). In contrast, MYC OFF/CCl4 livers were either equal to or 20% smaller than a normal liver and associated with a nodular appearance ( Figure 3A, G).

MYC expression does not enhance hepatotoxin-induced oval cell expansion
To further characterize the effects of DDC and CCl 4 on MYCinduced HCC formation in adult liver, histological analysis was performed. In MYC OFF/DDC livers, we found evidence for liver injury associated with the expansion of oval cells near the periportal areas ( Figure 4D). This oval cell expansion reached a peak three-to-four weeks after DDC diet initiation, consistent with previous findings [29], and may have contributed to the marked liver growth. MYC OFF/CCl4 mice developed liver injury associated with post-necrotic lobular collapse ( Figure 4G). This hepatocyte death and necrosis may explain the decrease in size of MYC OFF/CCl4 livers relative to normal livers ( Figure 3A, G).
In addition, MYC ON/DDC and MYC ON/CCl4 livers exhibited evidence for microcarcinomas as early as ten days after MYC activation ( Figure 4E, H), prior to developing diffuse HCC ( Figure 4F, I). Eighty-six percent of microcarcinomas found in MYC ON/DDC, and 50% of those found in MYC ON/CCl4 livers were located adjacent to periportal areas of the liver lobule ( Figure 4E, H). MYC ON adult mice also developed 93% of microcarcinomas near the periportal areas, although after a much longer latency period ( Figure 4B). These results are consistent with the previously suggested notion that many HCC's derive from the malignant expansion of less mature hepatocytes or oval cells [23].
In order to address whether the reduced tumor latency observed in DDC-treated mice may have been due to an increase in the oval cell population, immunohistochemistry (IHC) was employed. Accordingly, an antibody specific for the A6 antigen, known to be expressed on oval cells, was utilized [32]. Prior to induction, few cells were reactive for the A6 antigen, as expected ( Figure 5A). Upon initiation of DDC treatment, oval cells emerged in the periportal areas by 7 days ( Figure 5C) and continued to expand, reaching a peak between 3 and 4 weeks ( Figure 5E, G). It was expected that, if oval cell expansion was critical for increased   Figure 5B). One possible explanation for these data is that some oval cells were undergoing neoplastic transformation, or, alternatively, differentiating into immature hepatocytes and thus were no longer expressing the A6 antigen.
DDC and CCl 4 facilitate MYC-induced mitotic cellular division and proliferation in the adult liver.
In the adult murine liver, the ability of MYC to initiate tumorigenesis is associated with the ability to induce the cellular growth and DNA replication but not mitotic division of hepatocytes [26]. It has been demonstrated that continuous exposure to hepatotoxins such as DDC and CCl 4 causes liver injury that is accompanied by tissue regeneration. We speculated that DDC and CCl 4 were contributing to accelerated MYCinduced HCC by altering the liver context, thereby enhancing the mitotic division of hepatocytes. As such, liver cell proliferation was measured by Ki67 immunofluorescence and IHC for phosphorylated Histone H3 ( Figure 6A-C). DDC and CCl 4 treatments both resulted in a significant increase in the proliferative index, regardless of MYC expression. Indeed, either DDC or CCl 4 treatment for 17 days alone was sufficient to increase cellular proliferation and mitosis above that resulting from MYC overexpression for the same duration. However, this effect was exaggerated when combined with oncogene activation. As measured by Ki67, MYC ON/DDC and MYC ON/CCl4 livers exhibited a 30 and 40 percent increase in hepatocyte proliferation, respectively, as compared to DDC and CCl4-treated MYC OFF livers ( Figure 6A). In addition, a 300 and 250 percent increase in hepatocyte mitotic division as measured by phospho-Histone H3 was observed in cohorts with MYC on and exposure to either DDC or CCl 4 ( Figure 6C). Notably, the proliferation observed in MYC ON/DDC samples was diffuse, and not localized to the periportal areas of the liver lobule, further suggesting that these HCC's may not originate from the oval cells ( Figure 6 and data not shown). Importantly, MYC induction alone had only a modest impact on hepatocyte proliferation within 17 days ( Figure 6A-C). Thus, hepatotoxins can similarly promote continuous and diffuse hepatocyte proliferation irrespective of the targeted region of the liver lobule.
We previously reported that MYC is restrained from inducing the mitotic division of adult hepatocytes, and consequently for HCC formation, due to a p53-dependent cell cycle arrest [26]. To determine if loss of p53 is similarly necessary for MYC to induce HCC in the context of CCl 4 or DDC, we examined p53 protein expression by Western analysis ( Figure 6D). In contrast to our previous results, we found that adult tumor onset in MYC ON/ CCl4 and MYC ON/DDC mice was not associated with a loss of p53 protein expression ( Figure 6D). Thus, it appears that treatment with hepatotoxins allows for evasion of p53-cell cycle arrest upon induction of MYC overexpression, thereby allowing MYC to induce increased mitotic division and HCC. Interestingly, this is in agreement with our results demonstrating that p53 loss is not required for tumorigenicity in the neonatal liver [26].

DDC and CCl 4 treatment permit MYC-induced Cyclin B1 expression
MYC is thought to promote unrestrained proliferation and tumorigenesis in part by directly inducing the transcription of genes involved in cell cycle progression and mitosis [1,[33][34][35]. However, while MYC can rapidly induce mitotic entry in neonatal livers, oncogene activation initially results in hepatocyte hypertrophy in adult livers and only results in mitosis in this context upon neoplasia [26].
To determine if the accelerated tumorigenesis in hepatotoxintreated adult livers is due to differences in MYC's ability to induce the expression of genes involved in mitotic division, we measured the expression of known transcriptional targets of MYC by Quantitative Real-Time PCR (qRT-PCR). As a control, MYC expression was shown to increase in a temporal manner in all samples, with no significant difference observed in hepatotoxintreated livers ( Figure 7A). Similarly, no differences were observed in ODC and Nucleolin mRNA expression levels, two canonical targets of MYC's transcriptional activity ( Figure 7B, C). However, when Cyclin B1 expression was examined, we discovered that MYC ON/DDC and MYC ON/CCl4 livers exhibited a 250 and 150 percent increase in Cyclin B1 mRNA, respectively, compared to MYC OFF/DDC and MYC OFF/CCl4 livers ( Figure 7D). Cyclin B1 mRNA was also elevated in MYC OFF/DDC and MYC OFF/CCl4 livers relative to MYC ON and normal adult livers ( Figure 7D). Interestingly, the pattern of Cyclin B1 expression in MYC ON/DDC and MYC ON/CCl4 is strikingly similar to the expression pattern in neonatal livers in which MYC expression was induced.
Our results suggest that when MYC is overexpressed in conjunction with DDC or CCl 4 treatment, it may be more capable of transactivating genes involved in mitotic regulation in the adult liver. Moreover, the difference in Cyclin B1 expression could, in part, be responsible for the dramatic difference in MYC's ability to induce the mitotic division of hepatocytes in the context of DDC or CCl 4 .

Discussion
Here, we show that two non-genotoxic hepatotoxins, DDC and CCl 4 , can cooperate with MYC to induce robust mitotic cellular division and the near immediate onset of liver tumorigenesis. Our results have implications for the mechanisms by which oncogenes such as MYC induce and are restrained from causing HCC.
Previously, we have shown that MYC overexpression induces robust cellular proliferation in embryonic and neonatal hepatocytes, but, in marked contrast, induced cellular growth and DNA replication without mitotic cellular division in the adult liver [26]. We speculated that for MYC to induce liver cancer in adult hosts, the liver context must be changed, presumably through factors that stimulate hepatocyte proliferation. Indeed, both DDC and CCl 4 , toxins that target distinct regions of the liver [28][29][30] were found to markedly increase the ability of MYC to induce mitotic cellular division and permit the almost immediate onset of HCC. Our results suggest that oncogenes such as MYC can become latently active in the liver but only upon subsequent exposure to a hepatotoxin is their malignant potential manifested.
Our data are consistent with the general notion that the liver is most permissive to malignant transformation under circumstances when hepatocytes are undergoing cellular proliferation [26,36]. Moreover, our observations further validate previous reports that MYC can cooperate with agents that modulate the liver and thereby stimulate hepatocyte proliferation. Most notably, MYC has been shown to cooperate with TGF-alpha and phenobarbital to induce accelerated tumorigenesis in the adult murine liver [37]. Our model recapitulates the latency of tumor formation associated with adult onset HCC. Hence, we should be able to define factors that either enhance or repress tumorigenesis. To our knowledge, our report is the first to describe that hepatotoxins cooperate with MYC overexpression to induce accelerated HCC specifically in the context of an adult murine liver.
A possible explanation for our findings is that MYC appears to be only capable of inducing the activation of Cyclin B1-Cyclin-dependent kinase 1 (Cdk1) in embryonic/neonatal hepatocytes or in adult livers treated with hepatotoxins. Elevated Cyclin B1 levels often precede the onset of tumor cell immortalization and aneuploidy [38,39]. When we compared Cyclin B1 expression levels in the different cohorts of mice, we found that treatment with DDC or CCl 4 was associated with elevated Cyclin B1 expression relative to a normal or MYC ON liver. Furthermore, when MYC was overexpressed in conjunction with either DDC or CCl 4 we unveiled evidence for a further increase in Cyclin B1 expression levels. Notably, Cyclin B1 has been shown to be a direct transcriptional target of MYC [35,41]. Hence, one likely explanation for these differences is that the Cyclin B1 promoter region may be more accessible to MYC in hepatotoxin-treated livers, possibly by regulation of chromatin structure [40]. Alternatively, the increased expression of Cyclin B1 could be attributed to mitotic signals from the regenerative response initiated in the liver by CCl 4 or DDC damage. Future experiments will be directed at determining if MYC is directly activating the transcription of Cyclin B1 in hepatotoxin-treated livers.
Previously, we were able to show that a loss of p53 is associated with MYC-induced tumorigenesis in the normal adult liver [26]. In contrast, there was no evidence for p53 protein loss in MYC ON/DDC and MYC ON/CCl4 liver tumors, consistent with our observations in embryonic or neonatal MYC-induced tumors. Interestingly, Yin and colleagues recently showed that loss of p53 function cooperates with MYC to induce increased Cyclin B1 expression [41]. Our earlier report supports this mechanism in that loss of p53 in MYC-induced adult tumors was associated with elevated Cyclin B1 mRNA. Interestingly, we did not observe a correlation between p53 protein status and Cyclin B1 expression in hepatotoxin-treated livers. One possible explanation for this discordance is that, after exposure of the liver to DDC or CCl 4 , the already elevated levels of Cyclin B1 enable MYC to bypass the p53 checkpoint and further drive hepatocyte proliferation.
We hypothesized that one possible mechanism by which DDC facilitates MYC-induced HCC is by inducing oval cell expansion. Since these cells more closely resemble embryonic hepatoblasts, we posited that they may be more susceptible to MYC-induced tumorigenesis [42,43]. However, when assayed by A6 immunohistochemistry, there was no evidence for a difference in oval cell number between MYC OFF/DDC and MYC ON/DDC livers. One possible explanation for a lack of change in oval cell number between MYC ON/DDC and MYC OFF/DDC livers is that some oval cells were undergoing neoplastic transformation, thereby losing their oval cell properties and were no longer expressing the A6 antigen. This idea is further supported by the fact that the first appearance of neoplastic foci and microcarcinomas in the MYC ON/DDC livers occurs as early as 10 days after MYC activation, and is predominantly located in the periportal area of the liver. Alternatively, these oval cells could have also lost their oval cell properties by quickly differentiating to immature hepatocytes. Our model should be useful in further investigating the role of oval cells in MYC-induced tumorigenesis.
In contrast to DDC, continuous exposure to CCl 4 damages the liver by causing massive centrilobular necrosis and stimulating an inflammatory response through the activation of kupffer cells. Activated kupffer cells produce pro-inflammatory cytokines, such as Tumor necrosis factor-alpha (TNF-alpha), that can exacerbate chemical-induced liver injury as well as functioning as hepatic mitogens [44,45]. TNF-alpha is commonly induced almost immediately after CCl 4 exposure and is responsible for the activation of nuclear transcription factors that promote the expression of genes involved in cell cycle progression and mitotic division [44,46]. It will be interesting to subsequently investigate the role of CCl 4 -induced TNF-alpha signaling in MYC-induced liver tumorigenesis.
In this report, we demonstrate that toxin-mediated liver damage is sufficient to accelerate MYC-induced HCC formation in adults. Interestingly, this effect is irrespective of the target of the hepatotoxin or whether it functions independently as a carcinogen. Our results are consistent with a model whereby oncogenic events, such as MYC overexpression, may be generally silent in adult somatic hepatocytes unless the host is exposed to a liver toxin that promotes a permissive environment for tumorigenesis.

Transgenic mice
The TRE-MYC transgenic line generated for these experiments was described previously [2]. The LAP-tTA transgenic line was kindly provided by H. Bujard [47]. MYC expression was activated by removing doxycycline treatment (100 mg/ml) from the drinking water of mice transgenic for both TRE-MYC and LAP-tTA.

Tumorigenicity assays
MYC was activated in the liver by removing doxycycline treatment from the water. For toxin treatment, 6-week-old mice were either continuously fed a diet containing 0.1% DDC (Sigma-Aldrich, St. Louis, Missouri, United States) or injected intraperitoneally with carbon tetrachloride (Sigma-Aldrich) two times a week every week until sacrifice. CCl 4 was diluted in mineral oil (Sigma-Aldrich) and administered at a dose of 1 mL per gram of mouse. Mice were monitored daily and were sacrificed when moribund with tumor burden. During necropsy, liver tissues were saved by snap freezing in liquid nitrogen or were prepared for histology by fixing in 10% buffered formalin for 24 h and then transferring to 70% ethanol until paraffin embedding. All procedures were approved by the Animal Care Committee at Stanford University.

Immunohistochemistry
Tissue sections 4 mm thick were cut from paraffin-embedded blocks and placed on glass slides. Hematoxylin and eosin (H&E) staining was performed using standard procedures. The Stanford Histology Core laboratory prepared paraffin sections and performed H&E staining. A6 IHC was performed on liver tissue frozen in O.C.T. (280uC), sectioned at 7 mm, and fixed in 100% acetone at 220uC for 10 minutes. Phospho-Histone H3 IHC was performed on formalin-fixed paraffin embedded tissue. Endogenous peroxidase activity was blocked in 0.3% H 2 O 2 in methanol for 30 minutes, followed by blocking in rabbit or goat serum for 1 hour. The A6 rat monoclonal antibody was applied at a 1:20 dilution for 1.5 hours at room temperature. The A6 antibody was a generous gift from Dr. Valentina Factor (National Cancer Institute, National Institutes of Health, Bethesda, Maryland, United States). The phospho-Histone H3 antibody (#9701, Cell Signaling Technology, Danvers, Massachusetts, United States) was applied at a 1:100 dilution overnight at 4uC. Primary antibody incubation was followed by detection using either the rabbit antirat ABC Elite Kit or the goat anti-rabbit ABC Elite Kit (Vector Laboratories, Burlingame, California, United States) according to the manufacturer's protocol. Color was obtained using diaminobenzadine/nickel chloride (Vector Laboratories).

RNA Isolation and cDNA Preparation
Total cellular RNA was isolated from snap-frozen liver tissue using the RNeasy Kit (Qiagen, Valencia, California, United States) following the manufacturer's protocol. The amount of total RNA was quantified using spectrophotometric OD 260 measurements. RNA was reverse transcribed using Oligo(dT) 12-18 primers (Invitrogen, Carlsbad, California, United States) and SuperScript II Reverse Transcriptase (Invitrogen).
Western Blot. Western analysis was performed using conventional techniques. Liver tissues were disrupted and protein was isolated using a tissue homogenizer in NP-40 lysis buffer. Proteins were electrophoresed on 7.5% Tris-HCl polyacrylamide gels and transferred onto PVDF membranes. The membrane was blocked in 5% nonfat dry milk solution in TBST at 4uC overnight. p53 protein expression was detected using the NCL-p53-CM5p rabbit polyclonal antibody (Novacastra, VisionBiosystems, Inc., Norwell, Massachusetts, United States) at a 1:500 dilution. As a positive control, we used liver tissue from mice injected with a p53 expressing vector.