cGCN2 inhibition sensitizes arginine deprived hepatocellular carcinoma cells to senolytic treatment

The urea cycle is suppressed in HCC and correlated with severe disease

RNA-seq data from The Cancer Genome Atlas (TCGA) HCC project identified the urea cycle (Fig. 1B) as one of the most under-expressed metabolic gene sets in HCC tumors (Fig. 1A), suggesting that reduced urea cycle activity may be a common feature. In contrast to other tumor types where only ASS1 expression is reduced, all five principal urea cycle enzymes, argininosuccinate synthetase 1 (ASS1), argininosuccinate lyase (ASL), arginase 1 (ARG1), carbamoyl phosphate synthetase 1 (CPS1), and ornithine transcarbamoylase (OTC) are downregulated in HCC (Fig. 1C). Similarly, mRNA and protein levels were reduced in primary HCC tumors compared to healthy liver samples (Fig. 1D-F) and tissue arrays containing both HCC and normal liver tissue sections (Suppl. Fig. S1A, B). Next, we evaluated urea cycle expression in HCC cell lines comprising a variety of driver mutations: SNU-182 (TP53), SNU-398 (CTNNB1/ TP53), SNU-423 (TP53), SNU-449 (TP53), SNU-475 (TP53), HepG2 (NRAS), Hep3B (AXIN1), and Huh7 (TP53). Essentially all exhibited very low or undetectable urea cycle protein and mRNA levels (Fig. 1G, H; Suppl. Fig. S1C, D). Grade segregation (grade 1-4) showed an inverse correlation between urea cycle gene expression and HCC progression, indicating that reduced expression is associated with more severe disease (Suppl. Fig. S1E). Kaplan - Meier analysis demonstrated a significantly worse prognosis for patients with reduced urea cycle mRNA levels (Fig. 1I, Suppl. Fig. S1F), with the most significant effects reflecting repression of the entire pathway.

To verify that urea cycle loss has a functional consequence for HCC progression, we re-expressed ASS1 in SNU-398 cells and showed that proliferation declined based on their capacity to synthesize DNA as shown by decreased EdU incorporation (Fig. 1J). Of note, ASS1’s ability to suppress growth was catalytic activity dependent, as two naturally occurring mutants rendering the enzyme catalytic activity dead (Khare et al., 2021) had no impact (Fig. 1J, Suppl. S1G). Moreover, as demonstrated for other cancers (Hajaj et al., 2021), urea cycle enzyme loss in HCC protects aspartate pools and pyrimidine synthesis needed for DNA replication: decreased growth upon ASS1 re-expression in SNU-398 cells was reversed by providing exogenous aspartate and the human excitatory amino acid transporter- (1 hEAAT1) (Fig. 1K), nucleosides (Fig. 1L), and pyrimidines (Fig. 1M).

Consistent urea cycle loss in HCC would be expected to result in ammonia buildup in tumors and cell lines (see Fig. 1B), as detected in primary human HCCs (n=7) and SNU-398, SNU-449, and Huh7 cells (Suppl. Fig. 1H, I). Given the findings of Spinelli et al. (Spinelli et al., 2017), HCC cells may assimilate ammonia through reductive amination catalyzed by glutamate dehydrogenase (GDH) where amino acids like proline and aspartate directly acquire this nitrogen to support biomass expansion. Fumarate and ornithine levels were lower in primary HCCs (Suppl. Fig. 1J, K) and ornithine essentially undetectable in three HCC cell lines (Suppl. Fig. 1L), consistent with downregulation of the urea cycle pathway that generates these products.

SLC7A1 plays a dominant role in arginine uptake to ensure HCC cell proliferation

Urea cycle loss shunts metabolites away from Arg production towards pyrimidine biosynthesis as part of cancer metabolic reprogramming (Cheng et al., 2018, Rabinovich et al., 2015), making Arg an essential amino acid. Based on previous studies (Cheng et al, 2007; Lam et al., 2009; Xu et al., 2018), we predicted that HCC cells are dependent on Arg import for proliferation and/or viability. As expected, Arg removal from culture media for 72 hrs strongly impaired DNA synthesis as shown by nearly total inhibition of EdU incorporation (Fig. 2A, Suppl. Fig. S2A). Of note, HCC cells take up significantly more exogenous labeled Arg than primary hepatocytes at baseline (Fig. 2B), and prior Arg restriction for 72 hrs increased Arg uptake capacity, suggesting compensatory increases in cell surface expression of Arg transporter(s) (Fig 2C). In agreement, Arg levels were elevated in primary HCC tumors (Suppl. Fig. 2B) and SNU-398, SNU-449, and Huh7 HCC cells (Suppl. Fig. 2C).

Arg is imported by solute carrier (SLC) transporters (Closs et al., 2004). TCGA data indicated that SLC3A1, SLC7A2, SLC7A4, and SLC7A7 mRNAs are downregulated in HCC compared to normal liver, while SLC7A9 and SLC3A2 remain unchanged (Suppl. Fig. S2D, E), suggesting these are not responsible for increased Arg uptake. Whereas SLC7A1, SLC7A6, and SLC6A14 expression is increased, only SLC7A1 and SLC7A6 were consistently over-expressed across HCC tumor tissue and cell lines (Fig. 2D-F, Suppl. Fig S2F-H). Of these, SLC7A1 was the most upregulated and the only transporter induced during Arg deprivation (Fig. 2F, G, Suppl. Fig S2G, I), implicating SLC7A1 as a key Arg transporter in nutrient limiting conditions. Silencing SLC7A1 expression with shRNAs in SNU-398 cells decreased SLC7A1 expression by ~80-90% (Fig. 2H, Suppl. Fig S2J), reduced basal Arg uptake by ~60%, and blocked increased Arg uptake after Arg starvation (Fig. 2I, J). Moreover, SLC7A1 silencing did not induce compensatory expression of SLC7A6 or SLC6A14 (Suppl. Fig. S2K, L) or ASS1 or ASL (Suppl. Fig. S2M-O). SLC7A1 depletion also reduced EdU incorporation, cell proliferation, and intracellular Arg levels (Fig. 2K, L, Suppl. Fig. 2P). In parallel, treating HCC cells with the SLC7A1-3 inhibitor N-ethylmaleimide (NEM) reduced Arg uptake, EdU incorporation, and cell growth (Fig. 2M-O) in a dose-dependent fashion (Suppl. Fig. S2Q). We did not observe significant additive or synergistic effects of combined SLC7A1 shRNA and NEM treatment (Suppl. Fig. S2R, S) demonstrating that potential off-target effects of NEM are likely negligible.

To examine the role of SLC7A1 in vivo, SLC7A1 shRNA and vector control cells were used to generate subcutaneous HCC xenografts. SLC7A1 loss impaired SNU-398 tumor growth, as measured by volume (Fig. 2P) and end-point tumor weight (Fig. 2Q), and was associated with decreased proliferation (Fig. 2R, S). Physical health was monitored by regular body weight measurements (Suppl. Fig. S2T) and SLC7A1 shRNA maintained reduced SLC7A1 levels and lower endpoint intratumoral Arg levels (Suppl. Fig. S2U-W). Similarly, treatment of SNU-398 xenografts with 5mg/kg NEM imhibited tumor growth (Fig. 2T, U) and reduced HCC cell proliferation (Fig. 2V, W) without affecting mouse bodyweights (Suppl. Fig. S2X). Overall, these results show that HCC cells rely, surprisingly, on a single amino acid transporter (SLC7A1) to ensure sufficient Arg uptake to compensate for the loss of intracellular Arg due to urea cycle suppression.

Arg deprived HCC cells induce cell cycle arrest without overt apoptosis

Environmental stresses such as prolonged nutrient deprivation can initiate specific cellular responses to ensure survival (Anda et al., 2017; Fulda et al., 2010). Six day proliferation assays indicated that HCC cell growth was severely impaired when cells were deprived of exogenous Arg (Fig. 3A). Interestingly, reduced population expansion did not result from significantly increased cell death based on numbers of apoptotic (AnnexinV⁺/PI⁺) cells (Fig. 3B, Suppl. Fig. S3A) and nonapoptotic but dying cells (AnnexinV⁻/PI⁺) following 72 hrs of Arg withdrawal (Suppl. Fig S3B). Accordingly, no detectable cleaved caspase 3 was observed (Suppl. Fig. 3C).

Cell cycle analysis revealed that Arg restriction reduced the fraction of cells in the S and G₂/M phases, while cells accumulated in G₁/G₀ (Fig. 3C, Suppl. Fig. S3D), accompanied by decreased cyclin D1 and cyclin E1 accumulation (Fig. 3D). Moreover, levels of Proliferating Cell Nuclear Antigen (PCNA), an essential factor for DNA polymerase δ (Strzalka and Ziemienowicz, 2011) were lower (Fig. 3D). We concluded that Arg restriction is sufficient to induce a stress response that blocks replication and forces HCC cells into a quiescent state (G₀ phase) (Fig. 3E).

Restricting the availability of Arg likely alters multiple metabolic pathways, especially since the urea cycle provides fumarate for the TCA cycle needed for mitochondrial function and energy production. Moreover, Arg restriction can result in mitochondrial stress due to disruption of the malate-aspartate shuttle via asparagine synthetase (ASNS) mediated depletion of aspartate (Cheng et al., 2018). To determine if Arg depletion affected cellular bioenergetics we conducted assays of oxygen consumption rates [OCR], intracellular ATP levels, and ADP/ATP ratios. Arg restricted SNU-398 cells exhibited decreased ATP levels, ADP/ATP ratios, and OCRs (Fig. 3F-I). Decreased ADP/ATP ratios suggest a buildup of AMP, consistent with the activation of AMPK as demonstrated below (see Fig. 6I). These findings were consistent with OCR measurements for Arg deprived SNU-449 and Huh7 cells (Suppl. Fig. S3E-H), indicating that reduced extracellular Arg results in bioenergetic stress more generally in HCC. Nevertheless, failure to proliferate in Arg restricted conditions was not entirely due to reduced ATP levels, as exogenous supplementation failed to enhance EdU incorporation (Fig. 3J, Suppl. S3I). Moreover, providing exogenous aspartate to Arg deprived SNU-398 cells also failed to restore cell growth (Fig. 3K). Therefore, other consequences of Arg deprivation were interrogated as described below.

Arg restriction induces a GCN2-p21 mediated cell cycle arrest

Arg availability provides a nexus for the synthesis of polyamines, nitric oxide (NO), creatine, and proline, which are essential for cell survival and proliferation (Casero Jr. et al., 2018; Keshet et al., 2018; Morris, 2004). Polyamine levels are frequently dysregulated in tumors (Casero Jr. et al., 2018; Erdman et al., 1999; Gamble et al., 2019). However, supplementation of ornithine, an upstream intermediate involved in polyamine synthesis, or putrescine, spermine, spermidine, or all three together in combination with aminoguanidine (AG), an inhibitor of copper-containing amine oxidases, did not rescue HCC proliferation (Suppl. Fig 4 A-E). Arg also contributes to NO generation via nitric oxide synthases (NOS) (Keshet and Erez, 2018). NO can be a stimulator and inhibitor of cancer development and progression (Ciani et al., 2004; Du et al., 2013; Lopez-Rivera et al., 2014; Vahora et al., 2016). While treatment with L-NAME, a pan-NOS inhibitor, slightly increased HCC proliferation in Arg replete conditions, (Suppl. Fig. S4F), supplementation of a NO-donor (SNAP) failed to rescue proliferation of Arg deprived HCC cells (Suppl. Fig. S4G), suggesting that NO production is not responsible for the observed growth arrest.

Alternatively, amino acid restriction can induce specific stress responses such as GCN2-dependent ISR activation that promotes enhanced translation of specific mRNAs, like those encoding ATF4 and p21, which contain an ORF cluster within their 5’ untranslated regions (Fig. 4A) (Schmidt et al., 2020). Arg depletion induced GCN2 phosphorylation along with eIF2α inhibition and selective ATF4 translation in HCC cells (Fig. 4B-D), reversed by exogenous Arg (Suppl. Fig. S4H). To further explore GCN2 function, we employed the ATP competitive GCN2 inhibitor GCN2iB (Nakamura et al., 2018), which reduced GCN2 signaling (Suppl. Fig. S4I) and prevented GCN2 activation upon Arg limitation (Fig. 4E-G). More importantly, EdU incorporation was largely rescued in Arg restricted HCC cells after GCN2iB treatment (Fig. 4H). Arg deprivation resulted in a reversible accumulation of p21 (Fig. 4 I, J), which was reduced by GCN2iB (Fig. 4K, Suppl. Fig. S4J). To assess the contribution of p21 to cell cycle arrest, we silenced p21 using shRNAs capable of lowering mRNA and protein levels >80% (Fig. 4L, M). p21 loss partially rescued EdU incorporation (Fig. 4N), while having no effect on Arg replete cells, demonstrating that a GCN2 mediated p21 induction contributes to HCC cell cycle arrest under nutrient scarcity

Interestingly, during Toxoplasma gondii infection, Toxoplasma depletes host cell Arg, as it is auxotrophic for Arg, tryptophan, and purines (Augusto et al., 2019). In response infected host cells phosphorylate eIF2α via GCN2 leading to increased ATF4 which enhances SLC7A1 expression, resulting in increased Arg uptake. We therefore determined that elevated SLC7A1 levels were decreased by GCN2iB treatment of SNU-398 cells (Fig. 4O, P), suggesting that GCN2 promotes SLC7A1 accumulation in HCC cells as well. Moreover, SLC7A1 loss in HCC cells results in a modest but detectable activation of GCN2 (Fig. 4Q, R) likely due to decreased intracellular Arg levels (see Suppl. Fig. S2P). These findings implicate an interesting regulatory loop exists between SLC7A1 and GCN2 based on Arg availability.

GNC2 inhibition overrides mTORC1 activity and blocks autophagy in Arg restricted HCC cells

Because inhibition of the GCN2-p21 pathway only partially rescued HCC proliferation defects, other outputs of GCN2 likely contribute to consequences of Arg limitation (Fig. 5A). Increased eIF2α phosphorylation inhibits general mRNA translation, consistent with puromycin pulse chase data (Fig. 5B, C). eIF2α phosphorylation also promotes ATF4 translation and consequent induction of the ATF4 target SESTRIN2, which inhibits mTORC1 as part of the ISR (Anda et al., 2017; Ye et al., 2015). Arg restriction increased SESTRIN2/SESN2 mRNA and protein levels in HCC cells (Fig. 5D, E), and this response was blocked by GCN2iB treatment (Fig. 5F, G). HCC cells also exhibited mTORC1 inhibition (Fig. 5H-L). However, treating Arg deprived HCC cells with GCN2iB largely restored mTORC1 activity (Fig. 5M-P). These data demonstrate that cell cycle arrest and inhibition of general protein synthesis are completely GCN2 dependent and mediated through the GCN2-eIF2α and GCN2-SESTRIN2-mTORC1 pathways. Surprisingly, direct sensing of Arg scarcity by mTORC1 did not play a critical role in these adaptive responses (see Discussion).

mTORC1 responds to multiple environmental stimuli, including nutrient limitation, ER stress, hypoxia, and oxidative stress (Saxton and Sabatini, 2017), and induces autophagy to promote amino acid recycling and cell survival (Fig. 6A). In agreement, Arg deprivation elevated autophagy, as indicated by lipid modification of LC3BI to form LC3BII, reduced p62 levels (Fig. 6B, C), and increased autophagosome numbers (Fig. 6D, Suppl Fig. S5A). Interestingly, the induction of autophagy upon Arg deprivation was fully reversed by GCN2iB (Fig. 6E-H, Suppl. Fig. S5B), indicating that GCN2 inhibition alone is sufficient to block mTORC1 regulation of autophagy. Finally, Arg limitation increased AMPK phosphorylation in HCC cells (Fig. 6I, J), consistent with previous reports that SESTRIN2-deficient mice exhibit metabolic pathologies related to altered AMPK signaling (Jin et al., 2019). These data demonstrate that autophagy induction in Arg restricted HCC cells appears to be entirely GCN2 dependent.

Prolonged GCN2 inhibition induces senescence in Arg restricted HCC cells, rendering them sensitive to senolytic treatment

The cell proliferation and autophagy phenotypes induced by short term Arg restriction are reversible by GCN2 inhibition or Arg restoration (see above). We showed that short term Arg restriction (72 hrs) induces cellular quiescence by increasing cell fractions in G0. Excessive cellular stress such as nutrient starvation is linked to production of reactive oxygen species, mitochondrial abnormalities, and DNA damage (Fulda et al., 2010). Cell cycle arrested cells in these conditions favor a state of irreversible cell cycle arrest/ senescence to limit propagation of damaged cells (Pawlikowski et al., 2013). However, although 72 hr Arg restriction induces cell cycle arrest, we did not observe a significant number of β–galactosidase positive cells (Suppl. Fig. S5 C, D). Arg re-supplementation for 24 hrs reactivated HCC proliferation, and full recovery of EdU incorporation levels (Suppl. Fig. S5E). Although GCN2 dependent induction of p21 was observed, other markers of cellular senescence such as p16^INK4a and 27^kip1 were not at 72 hrs (Suppl. Fig. S5F-H). Moreover, expression of genes involved in the Senescence-Associated Secretory Phenotype (SASP), including immune modulators, inflammatory cytokines, and proteases (Faget et al., 2019), showed only a modest increase compared to HCC cells cultured in control conditions (Suppl. Fig. S5I).

While HCC cells “cope” with short term Arg shortage by inducing quiescence, prolonged Arg restriction (7 days) increased apoptosis suggesting these mechanisms were limited (Fig. 6K, Suppl. Fig. S5J, K). Moreover, sustained nutrient stress increased p21, p16, and p27 levels (Fig. 6L) as well as β-galactosidase⁺ cell numbers (Fig. 6M, Suppl Fig. S5L). Arg re-supplementation did not completely rescue proliferation defects (Fig. 6N), indicating many cells had now entered a senescent state. GCN2 inhibition during short term Arg restriction largely rescued EdU incorporation, possibly through forced re-initiation of protein synthesis and suppression of p21 (see Fig. 4I). Restoration of growth could force cells to use limited nutrient pools. In agreement, more prolonged (8 day) GCN2 inhibition in Arg restricted cells did not sustain cell growth (Fig. 6O). Cellular Arg depletion was observed by day 4 (see Suppl. Fig. S6G), indicating that any compensatory pathways engaged to maintain intracellular levels become exhausted. This is consistent with the data shown in Fig. 6O, where cells exhibit slowed growth in response to GCN2iB at that time.

Previous studies demonstrated that autophagy provides nutrients necessary to meet bioenergetic demands during the transition from quiescence to activation, and defective autophagy results in accumulation of toxic waste and senescence (García-Prat et al., 2016; Tang and Rando, 2014). As GCN2 inhibition impaired autophagy, cells were unable to re-initiate cell cycle progression, a feature characteristic for cellular senescence. Indeed, β-galactosidase staining showed that GCN2 inhibition during prolonged (7 day) Arg restriction markedly induced senescence (Fig. 6M, Suppl Fig. S5L), and SASP activation (Fig. 6P).

Cellular senescence is associated with increased expression of BCL2 homology gene family members such as BCL2 and MCL1, which can be targeted for directed elimination of senescent cells (Ovadya and Krizhanovsky, 2018; Soto-Gamez et al., 2019). Accordingly, we found that long term (7 day) Arg restriction enhanced the expression of BCL2, but not MCL1 (Suppl Fig. S5M). Interestingly, combined treatment of Arg deprived HCC cells with GCN2iB and the pan-BCL2 inhibitor ABT-263 significantly increased the number of apoptotic cells, whereas single treatments of GCN2iB or ABT-263 alone had only modest effects (Fig. 6Q). To evaluate the impact of prolonged SLC7A1 loss (in Arg replete conditions) on growth arrested G0 cells, autophagy induction, senescence, and responses to GCN2iB plus ABT-263 treatment, we applied shSLC7A1 RNAs to SNU-398 cells. As shown in Suppl. Fig. S5O-R, reduced SLC7A1 expression resulted in increased G0 cell numbers, autophagy, and cell death upon GCN2 inhibition combined with senotherapy. Murine Hepa1-6 cells were also examined for their urea cycle status (Suppl. Fig. S6A), sensitivity to Arg restriction (Suppl. Fig. S6B), effects of GCN2 inhibition on autophagy (Suppl. Fig. S6C), and responses to combined Arg depletion and GCN2iB/ ABT-263 exposure. As shown in (Suppl. Fig. S6D), results indicate that urea cycle loss, Arg auxotrophy, and senotherapy responses are similar to human HCC cell lines. Together, these data demonstrate that prolonged GCN2 inhibition in Arg restricted (or SLC7A1 deficient) HCC cells induces cellular senescence, which in turn renders them sensitive to BCL2 inhibition.

Given the reported role of p53 in urea cycle inhibition (Li et al, 2019), we assessed whether TP53 mutational status influences Arg related phenotypes across HCC cell lines. SNU-398, SNU-449, and Huh 7 exhibit the following TP53 mutations: TP53 splice variant, TP53^A161T, and TP53^Y220C respectively. We therefore examined the effects of Arg restriction, GCN2iB, and ABT-263 on HepG2 (NRAS^Q61L) and Hep3B (AXIN1 mutant) cells which have intact TP53 loci. Based on previous reports, SNU-449 and Huh7 express relatively more p53 protein than HepG2 and Hep3B cells, and p53 is essentially undetectable in SNU-398 (Sayan et al., 2001). Of note, HepG2 and Hep3B cells responded similarly to the combination of Arg deprivation and GCN2iB/ABT-263 senotherapy as SNU-398 cells (Suppl. Fig. S6E, F). We concluded that all Arg related phenotypes described here are consistent across HCC cell lines regardless of TP53 status.

We also performed the combination of Arg restriction and GCN2iB/ ABT-263 treatment on three additional ASS1 deficient cell lines: PC3 prostate cancer, SKLMS leiomyosarcoma, and MDA-MB-231 breast cancer cells and demonstrated significant efficacy, although MDA-MB-231 cells exhibit responses to GCN2iB and ABT-263 even in the presence of Arg (see Figure S6H-J).

GCN2 inhibition sensitizes Arg restricted HCC cells to senolytic treatment in vivo

To examine effects of Arg restriction on HCC progression in vivo, we measured the growth of subcutaneous SNU-398 xenograft tumors in mice subjected to an Arg deficient diet. This regimen significantly impaired tumor growth based on tumor volume (Fig. 7A), and end point tumor weights (Fig. 7B), but had no obvious effect on body weight (Suppl. Fig. S7A). Slowed tumor growth was the result of cell cycle arrest based on significantly decreased EdU incorporation (Fig. 7C). Arg restriction via dietary means was sufficient to induce GCN2 signaling and expression of downstream factors ATF4 and SESTRIN2 (Fig. 7D), consistent with our in vitro results. GCN2iB treatment of Arg deprived mice alone had minor effects on xenograft tumor growth, which failed to reach statistical significance, and had no effect on tumor growth in mice maintained on a control Arg replete diet (Fig. 7E, F, Suppl. Fig S7B). However, GCN2iB treatment of Arg deprived mice induced the expression of BCL2 and SASP associated genes (Fig. 7G, H) and β-galactosidase staining in xenograft tumors, reflecting the induction of senescence in vivo (Fig. 7I, J). Importantly, combined treatment with GCN2iB and daily doses of 50 mg/kg ABT-263 strongly decreased SNU-398 tumor growth in Arg restricted mice (Fig. 7E, F), also without significant weight loss (Suppl. Fig. S7A). GCN2iB/ABT-263 treatment enhanced apoptosis in this regimen, as indicated by increased cleaved caspase 3 positive cells, but not in tumors derived from mice receiving a control diet (Fig. 7K, L).

A negative impact of dietary Arg restriction on tumor growth was also observed in Huh7 subcutaneous xenografts (Suppl. Fig. S7C), murine HepaMP-4 and Hepa55.1c syngeneic tumors in C57/Bl6 recipients (Suppl. Fig. 7D-F), and a HepaMP-9-1 orthotopic model (Suppl. Fig. S7F-H); HepaMP-4 and HepaMP-9-1 cells were generated from Myc^OE;Trp53^KO autochthonous HCC tumors (see Methods). GCN2iB treatment combined with ABT-263 further decreased Hepa55.1c tumor growth in Arg restricted mice (Suppl. Fig. S7E, F). Arg deprived mice harboring HepaMP-9-1 orthotopic tumors treated with GCN2 inhibitor also developed senescent cells (Suppl. Fig. S7I, J) as shown for SNU-398 xenografts (Fig. 7I, J). Interestingly, HepaMP-9-1 orthotopic tumors in Arg deprived mice treated with GCN2iB exhibited increased levels of VEGF (Suppl. Fig. S7K), a phenotype not shared with Huh7 flank tumors (Suppl. Fig. S7L). As such, some (but not all) HCCs might respond with increased efficacy to anti-VEGF agents like Bevacizumab when combined with GCN2 inhibition. Overall, these results demonstrate that dietary Arg restriction sensitizes tumors to GCN2iB and ABT-263 combination treatment, suggesting a novel therapeutic strategy for treating genetically heterogeneous human HCCs.

Limitations of Study

Subcutaneous HCC tumors do not fully represent metabolic exchanges between tumor and typical local hepatic microenvironments, especially since HCC occurs in an organ with unique metabolic properties. To address this, we determined that Arg deprivation combined with GCN2iB treatment reduced tumor growth and elicited significant HCC cellular senescence in an Myc^OE;Trp53^KO orthotopic model. Unfortunately, because of postoperative recovery needed from survival surgery, mice harboring orthotopic HCCs did not tolerate oral gavage as a route of drug delivery, even with vehicle. Our regimen was dependent on ABT-263 delivery within a week of orthotopic tumor initiation, while most treatments of mice with orthotopic liver tumors must wait approximately one month. As such, it was impossible to assess the efficacy of Arg restriction with combined GCN2iB/ABT-263 treatment in this setting. Future studies testing combinations of Arg deprivation, GCN2iB exposure, and ABT-263 oral gavage will deploy spontaneous murine HCC models under development in the laboratory at this time.

RESOURCE AVAILABILITY

EXPERIMENTAL MODEL AND SUBJECT DETAILS

METHOD DETAILS

ADDITIONAL RESOURCES

This study has not generated or contributed to a new website/forum and is not part of a clinical trial.