Inhibition of glutaminase elicits senolysis in therapy-induced senescent melanoma cells

The cyclin D1-Cyclin-Dependent Kinases 4 and 6 (CDK4/6) complex is crucial for the development of melanoma. We previously demonstrated that targeting CDK4/6 using small molecule inhibitors (CDK4/6i) suppresses BrafV600E melanoma growth in vitro and in vivo through induction of cellular senescence. Clinical trials investigating CDK4/6i in melanoma have not yielded successful outcomes, underscoring the necessity to enhance the therapeutic efficacy of CDK4/6i. Accumulated research has shown that while senescence initially suppresses cell proliferation, a prolonged state of senescence eventually leads to tumor relapse by altering the tumor microenvironment, suggesting that removal of those senescent cells (in a process referred to as senolysis) is of clinical necessity to facilitate clinical response. We demonstrate that glutaminase 1 (GLS1) expression is specifically upregulated in CDK4/6i-induced senescent BrafV600E melanoma cells. Upregulated GLS1 expression renders BrafV600E melanoma senescent cells vulnerable to GLS1 inhibitor (GLS1i). Furthermore, we demonstrate that this senolytic approach targeting upregulated GLS1 expression is applicable even though those cells developed resistance to the BrafV600E inhibitor vemurafenib, a frequently encountered substantial clinical challenge to treating patients. Thus, this novel senolytic approach may revolutionize current CDK4/6i mediated melanoma treatment if melanoma cells undergo senescence prior to developing resistance to CDK4/6i. Given that we demonstrate that a low dose of vemurafenib induced senescence, which renders BrafV600E melanoma cells susceptible to GLS1i and recent accumulated research shows many cancer cells undergo senescence in response to chemotherapy, radiation, and immunotherapy, this senolytic therapy approach may prove applicable to a wide range of cancer types once senescence and GLS1 expression are induced. Key points Upregulated GLS1 expression renders senescent BrafV600E melanoma cells induced by CDK4/6 inhibitor (CDK4/6i) or vemurafenib susceptible to GLS1 inhibitor (GLS1i) even with BrafV600E inhibitor resistance. This senolytic therapy combining CDK4/6i and GLS1i provides insights into potential novel therapeutic strategies for metastatic melanoma and may be applicable to various types of cancers providing alternative therapy options.


Introduction
Senescence is a state of stable and irreversible cell cycle arrest that occurs as a natural part of aging, as well as in response to various cellular stressors [1][2][3][4] .It is an essential cellular process that plays a crucial role in maintaining tissue homeostasis and preventing the proliferation of damaged or dysfunctional cells.In recent years, senescence has emerged as a mechanistic doubleedged sword for cancer.While the loss of proliferative capacity of senescent cells appears to prevent the accumulation of aberrant cells, thus acting as a barrier to tumorigenesis, senescent cells secrete a variety of inflammatory cytokines and form chronic microinflammation in the surrounding tumor microenvironment, promoting carcinogenesis, resistance to tumor immunity, and metastasis [5][6][7][8] .Furthermore, senescent cancer cells can stimulate the growth of neighboring non-cancerous cells such as stromal fibroblasts, potentially contributing to therapy resistance and tumor relapse [9][10][11] .Senolytic therapy is an emerging therapeutic approach that specifically targets and eliminates senescent cells, which prevents cancer recurrence [12][13][14][15] as well as extending a healthy lifespan in murine models 16,17 .Incorporating senolytic therapy into existing cancer treatment regimens has the potential to significantly improve patient outcomes and transform cancer management 18,19 .
Previous research demonstrated that cancer cells exhibit increased glutamine metabolism, which fuels the tricarboxylic acid (TCA) cycle, nucleotide and fatty acid biosynthesis, and redox balance in cancer cells 30,31 .This glutamine addiction is frequently observed in several types of human cancers and is driven by multiple oncogenes such as cyclin D1 or Myc [32][33][34] .Therefore, targeting glutamine metabolism is a novel strategy for cancer treatment.The antitumor activity of CB-839, a glutaminase inhibitor, has been the subject of several clinical trials (clinicaltrials.gov-NCT02071862,NCT02771626), but these trials have been unsuccessful, suggesting that additional strategies that enhance the treatment efficacy are of clinical necessity.Recent studies demonstrated that CDK4/6i reprograms mitochondria metabolism, which elicits unique vulnerabilities 35,36 .
Glutaminase 1 (GLS1) is an essential gene for the survival of human senescent cells and inhibition of GLS1 eliminated senescent cells resulting in improved age-associated organ dysfunction 37 .
Importantly, senolysis following cancer therapies has been postulated as an attractive approach to eliminating cancer cells as well as extending a healthy lifespan.However, it is unclear whether or not GLS1 is essential for the survival of senescent cancer cells induced by therapies such as CDK4/6i.
In this study, we demonstrate that prolonged treatment of Braf V600E melanoma cells with the CDK4/6i induces the expression of KGA, a major isoform of GLS1, in Braf V600E melanoma cells, but not in Nras Q61 melanoma cells.CDK4/6i-induced senescent cells become sensitive to GLS1 inhibition and undergo cell death in vitro and in vivo.We further demonstrate that the senolytic therapy is applicable for Braf V600E inhibitor vemurafenib-resistant melanoma cells, suggesting this strategy can be used after patients develop resistance to Braf V600E targeted therapy.
Together, these results indicate that this sequential combinatorial therapy approach using CDK4/6i and GLS1i may improve the current landscape of treating patients with Braf V600E melanoma.

Plasmids and viral production
HEK293T cells were co-transfected with either GLS shRNA-encoding or KGA overexpression plasmids along with packaging plasmids (pMD2.G and psPAX2 or Qψ) using Lipofectamine 2000 (Invitrogen).Viral supernatants were collected and filtered 72 hours post-transfection and used to infect melanoma cells in the presence of 1 μg/mL polybrene (Millipore).Selection of stably transduced cells was achieved by applying 0.5 to 1 μg/mL puromycin or 1 mg/mL G418 to the growth medium, with antibiotic treatments administered every 2 to 3 days over a total of 8 days.
The immunoreactive bands were captured using autoradiography.The obtained images were scanned and bands were quantified using ImageJ software.

Annexin V staining
Cultured melanoma cells with or without treatments were harvested and washed twice with PBS.
The cell pellet was resuspended in 1x binding buffer at a concentration of 1×10 6 cells/ml.Subsequently, 100 μl of the suspension (1×10 5 cells) was incubated with 5 μl of APC Annexin V (BD Biosciences) and incubated for 15 minutes at room temperature.Cells were analyzed by flow cytometry.

Immunohistochemistry (IHC)
Mouse xenograft tissue samples were fixed in 10% formalin, followed by a series of ethanol washes for deparaffinization and rehydration.Tissue sections on slides were subjected to blocking with 10% goat serum before incubation with primary antibodies.The IHC protocol included the application of biotinylated rabbit secondary antibodies, visualization with an ABC substrate kit, and subsequent development with a DAB chromogen, all from Vector Laboratories.
Counterstaining was performed using hematoxylin (Thermo Fisher Scientific), with slides mounted using Permount (Thermo Fisher Scientific).The antibodies used in the study is as follows: anti-glutaminase-1/GLS1 (E9H6H) antibody and anti-cleaved caspase-3 (Asp175) antibody obtained from Cell Signaling Technology and rabbit secondary antibody conjugated to biotin obtained from Sigma Aldrich.The staining intensity was assessed on a scale from 0 (no staining) to 3 (strong staining).The percentage of positive cells was estimated on a scale from 0 (no population) to 3 (100% positive).Each specimen was scored by multiplying the intensity by the percentage of positive cells, yielding a raw score.The H-score was calculated for each section using the formula: H = Σ (I × P), where I represents the intensity scores, and P represents the percentage of positive cells.

Senescence-Associated β-Galactosidase (SA-β-gal) Staining
SA-β-gal activity was assessed in both in vitro cultured melanoma cells and in vivo melanoma tumor sections embedded in Tissue-Tek O.C.T. compound (Sakura Finetek).Following treatment with drugs, both in vitro and in vivo samples were fixed and stained using the X-gal staining solution (Sigma) following the manufacturer's guidelines.

Xenograft experiments
A total of 1 × 10 6 melanoma cells were subcutaneously injected into 6-week-old SCID mice with Matrigel (BD Biosciences).Mice were treated with vehicle, palbociclib (90 mg/kg), or CB-839 (200 mg/kg) by oral gavage once a day for 17 days after tumors had developed 5 mm in diameter following inoculation.For the combination group, mice were treated with palbociclib for 8 days and then treated with palbociclib and CB-839 for an additional 9 days.Tumor volumes were measured by caliper every 2 days and calculated by the following formula: V = length × width × width / 2. Mice were euthanized, and the tumor size was measured at day 17.Experimental animal maintenance was conducted in compliance with the institutional guidelines of Case Western Reserve University.

Quantification of mitotic index and necrotic cells
The hematoxylin and eosin (H&E)-stained slides of all tumors were scanned using Aperio AT2 imager (Leica) and utilized for analysis of mitotic index and necrotic proportion.The mitotic cells were enumerated within a representative 1 mm 2 area, and the proportion of necrotic cells was quantified in each scanned H&E image using QuPath software.

Statistical analysis
Parametric data were expressed as mean ± standard deviation (SD), and non-parametric data were expressed as median and interquartile range (IQR).To compare the two independent groups, Student's t-test was used for parametric data, and the Mann-Whitney U test was employed for nonparametric data.Experiments were performed at least three times independently.Statistical significance was set at a p-value of less than 0.01.All analyses were performed using GraphPad Prism.

GLS1 specifically KGA is upregulated in senescent Nras WT melanoma cells.
Glutaminase (GLS1) is an enzyme known to convert glutamine to glutamate in the Tricarboxylic acid (TCA) cycle.GLS1 consists of two splice variants, Kidney-type glutaminase (KGA) and glutaminase C (GAC) 38 .KGA and GAC are identical except for amino acids in the cterminus.GLS2, referred to as liver-type glutamine, is derived from a gene distinct from GLS1.
To determine the expression of KGA, GAC and GLS2 in senescent cells induced by prolonged treatment of CDK4/6i, we treated a variety of melanoma cell lines harboring different oncogenic drivers (Braf WT Nras WT , Braf V600E Nras WT (Braf V600E ) and Braf WT Nras Q61R/K (Nras Q61 )) with CDK4/6i palbociclib for 8 days, compared to normal human melanocyte (ATCC-HEMa) as a control.Importantly, we previously demonstrated that while CDK4/6i treatment for 1 day induces reversible cell cycle arrest (quiescence), prolonged CDK4/6i treatment for 8 days induces senescence in Braf V600E melanoma cell lines 26,27 .We demonstrated that palbociclib induced senescence specifically in Braf WT Nras WT and Braf V600E melanoma senescent cells but not in Nras Q61R/L cell lines nor primary human melanocytes as assessed by SA-βgal assay, an assay to Consistent with a previous study 39 , primary human melanocytes did not undergo senescence following CDK4/6i treatment because normal human melanocytes are not dependent on CDK4/6 to proliferate compared with melanoma cell lines (Fig 1A -B).Upregulation of KGA after palbociclib treatment for 8 days in Braf V600E melanoma cell lines was verified by western blot analysis (Fig 1E).We previously established four independent clones of palbociclib and ribociclibresistant (PR) cells from 1205Lu Braf V600E melanoma cells 26 .We used those PR cells to assess the expression of KGA and demonstrated that KGA is not upregulated (Fig 1F) compared to control cells while KGA is upregulated only in palbociclib-induced senescent cells, underscoring the specificity of KGA upregulation in CDK4/6i-induced Braf V600E senescent cells.Mechanistically, it has shown that the GLS1 transcript contains an AU-rich element (ARE) within the 3′ untranslated region (3′UTR) 40 , which increases mRNA stability through RNA binding protein HuC binding under acidic conditions 41,42 .The GLS1 transcript was stable in senescent cells but unstable in non-senescent cells and HuC depletion in senescent cells reduced survival of senescent cells 37 .Therefore, we assessed the mRNA stability of KGA in palbociclib-induced senescent cells using the transcription inhibitor actinomycin D. We treated Braf V600E 1205Lu and skmel28 cells and Nras Q61R skmel2 with or without palbociclib for 8 days followed by actinomycin D treatment for

GLS1 inhibitor induces senolysis in Nras WT senescent melanoma cells.
GLS1 upregulation is essential for the survival of senescent cells because glutaminolysis produces ammonia, which neutralizes the acidic conditions caused by the leaking of intralysosomal H+ from damaged lysosomes in these senescent cells.However, when GLS1 is inhibited, senescent cells cannot produce ammonia to neutralize the acidic pH, leading to cellular acidosis 37 .We next assessed whether the specific upregulation of KGA expression in palbociclib-induced senescent cells renders Braf V600E melanoma cells vulnerable to GLS1 inhibitor (GLS1i).We treated a variety of melanoma cell lines (Braf WT Nras WT , Braf V600E , and Nras Q61R/L ) with palbociclib for 1 day or 8 days, followed by a specific GLS1i (CB-839) for an additional 3 days to assess cell death by To align with current clinical practice in melanoma management, we subsequently extended the Braf V600E inhibitor (Brafi)-resistant population and assessed the efficacy of senolytic therapy using CDK4/6i and GLS1i in Brafi vemurafenib-resistant Braf V600E melanoma cells.Brafi has been used as front-line therapy for Braf V600E melanoma, but resistance to Braf V600E inhibitor has been frequently observed clinically and is problematic for treating melanoma 43 .We used two vemurafenib-resistant Braf V600E melanoma cell lines derived from 1205Lu and 983B that we established previously 27 .We first assessed the expression of KGA after treatment with palbociclib in vemurafenib-resistant (VR) Braf V600E melanoma cells.We treated VR cells with palbociclib for 1 day or 8 days and harvested cells to assess KGA expression.We demonstrated that KGA is significantly induced after treatment of palbociclib for 8 days while 1-day treatment of palbociclib  6).Together these findings suggest that the senolytic treatment combining CDK4/6i and GLS1i is effective for Braf V600E melanoma cells, including those that have developed resistance to vemurafenib.This data demonstrates in vitro proof-of-concept that combination senolytic therapy may be a viable treatment option for patients with resistant melanoma and warrants further clinical trials to evaluate its efficacy.
The above results have shown that induction of GLS1 expression (specifically KGA expression) by palbociclib in senescent cells is critical for successful targeting by GLS1i.
Consistently, CDK4/6i increases mitochondrial metabolism in Braf V600E melanoma cells 36 and pancreatic cancer 35 .We asked whether palbociclib-induced senescence represents glutamine dependence, a state where cancer cells rely on glutamine for survival.We treated VR melanoma cells with palbociclib for 8 days and cultured them with medium lacking glutamine for an additional 3 days to assess cell death by Annexin V. We demonstrated that, unlike CB-839, glutamine depletion did not provoke senolysis in palbociclib-induced senescent cells (Fig 4D).This suggests that CDK4/6i-induced senescent cells rely on the upregulation of KGA expression rather than on glutamine itself for survival in Braf V600E melanoma cells.Moreover, this indicates that KGA induction in senescent cells is critical target for GLS1i to provoke senolysis.

GLS1 inhibitor induces senolysis in vemurafenib-resistant melanoma cells in vivo.
To assess the efficacy of the combination of palbociclib and CB-839 in vivo, we performed a melanoma cell xenograft experiment.1205Lu Braf V600E VR melanoma cells were injected subcutaneously into 8-week-old SCID mice.We treated mice with either palbociclib alone for 17 days, CB-839 alone for 17 days, or pre-treatment of palbociclib for 8 days followed by the combination for an additional 9 days (total 17 days) after tumors had developed 5 mm in diameter (Fig 5A).While a single treatment of palbociclib suppressed melanoma cell proliferation through senescence induction as demonstrated previously 27 , the combined treatment of palbociclib and CB-839 reduced tumor burden further compared to the palbociclib-alone treatment group (Fig 5A ).
Mouse body weights after treatment were unchanged (Sup Fig 7).Treatment of CB-839 alone is less effective in suppressing cell proliferation compared with palbociclib alone or a combination group (Fig. 5A).H&E staining revealed that the mitotic index, indicative of cell proliferation, was markedly reduced in palbociclib-alone and combination-treatment groups compared to the control group (Fig 5B-C).Despite the mitotic index of tumors treated with CB-839 being comparable to that of the control group, an increase in necrotic cell occurrence was observed exclusively in the CB-839-alone treated tumors (Fig 5B and D).The simultaneous presence of proliferating and necrotic cells provides insight into the modest inhibition of tumor growth by CB-839 treatment, even when the number of mitotic cells aligns with that of the control group.We further characterized the tumors following the various treatments.Using the SA-βgal assay we found that an attractive target for proliferating cancer cells, and currently being evaluated in many clinical trials for different types of cancers.However, senolytic therapy induced by the combination of palbociclib and CB-839 significantly reduces tumor burdens, which may be a more attractive strategy compared with targeting proliferating cancer cells by CB-839.Taken together these in vivo results strongly support our concept that induction of senolysis by targeting GLS1 in palbociclib-induced senescent cells is highly effective.Furthermore, these data indicate that senolytic therapy may be a potential therapeutic solution to address the clinical challenge of Braf V600E inhibitor resistance in treating Braf V600E melanoma.Further optimization of drug dosages and treatment schedules will be needed to enhance its potential efficacy.

GLS1 inhibitor induces senolysis in vemurafenib-induced senescent cells.
The results thus far demonstrated that targeting upregulated GLS1 by a GLS1 inhibitor in CDK4/6i-induced senescent cells is an attractive strategy for treating Braf V600E melanoma.While we demonstrated that CDK4/6i induces senescence in Braf V600E melanoma, the previous study has shown that a low concentration of vemurafenib induces senescent features in Braf V600E melanoma cells 44 .We next tested whether senescent cells induced by other therapeutics like vemurafenib are susceptible to GLS1i.We treated Braf V600E 1205Lu melanoma cells with vemurafenib for 8 days Additionally, the phosphorylation of S6, a target of ERK, was also diminished (Fig 6C).This suggests that a low dose of vemurafenib suppresses downstream signaling.We next administered vemurafenib to 1205Lu melanoma cells for 8 days and then treated them with CB-839 for an additional 3 days.Cell death was then assessed using Annexin V, as detailed in Figure 2A.We demonstrated that this sequential treatment effectively induced the death of vemurafenib-triggered senescent cells (Fig 6D).Together, these data suggest that GLS1 represents a strategic molecular target in senescent cells with elevated KGA, induced not only by CDK4/6i but also by other cytostatic drugs such as a low concentration of vemurafenib.The potential to optimize senolytic therapeutic strategies through the use of GLS1i is significant, contingent upon the identification of pharmacological combinations that elicit a more robust and enduring senescent state, concomitant with KGA expression.

Discussion
Therapeutic inhibition of CDK4/6 has been recognized as a potent anticancer strategy.
Selective CDK4/6 inhibitors (CDK4/6i), such as palbociclib, abemaciclib and ribociclib have received FDA approval for the treatment of estrogen receptor-positive (ER+)/HER2-negative breast cancer.While CDK4/6 inhibition has been shown to effectively suppress tumor proliferation by inducing senescence in both in vitro and in vivo preclinical melanoma models, its efficacy has not been replicated in clinical trials for melanoma.Although the development of acquired resistance to CDK4/6i presents a challenge, recent evidence suggests that senescence can paradoxically promote tumorigenesis and recurrence by secreting various inflammatory cytokines [5][6][7][8] .This underscores the necessity to innovate and develop novel senolytic agents capable of eradicating senescent cells induced by current therapeutic interventions, thereby circumventing the onset of resistance and mitigating the risk of neoplastic recurrence.
In this study, we assessed the expression of Glutaminase 1 (GLS1) in CDK4/6i-induced senescent melanoma cells and tested the applicability of targeting GLS1 to eliminate those senescent cells by GLS1 inhibition-mediated acidosis induction, given a recent study describing GLS1 as an essential gene supporting senescent cells for survival by neutralizing acidic conditions through ammonia production induced by upregulated GLS1 expression 37 .We demonstrated that prolonged treatment of CDK4/6i upregulates GLS1 expression in Braf V600E melanoma cells, which is consistent with recent studies demonstrating that CDK4/6i reprograms mitochondria metabolism 35, 36 .This mitochondrial reprogramming may occur through GLS1 regulation since the expression of glutamine transporters such as SLC1A5 (ASCT2) is unchanged after treatment of palbociclib (Sup Fig 8).We further demonstrated that targeting GLS1 using a specific glutaminase inhibitor (CB-839) eliminates Braf V600E senescent cells induced by either CDK4/6i or vemurafenib.
Importantly, the synergistic pharmacological intervention comprising palbociclib and CB-839 retains its efficacy against Braf V600E melanoma cells exhibiting acquired resistance to vemurafenib, as evidenced by both in vitro and in vivo studies.Although we demonstrated that tumor growth suppression by senolysis induction using a combination of CDK4/6i and GLS1i is durable using long-term colony formation assay (Fig 2E), subsequent research evaluating the lifespan of mice bearing Braf V600E melanoma following treatment with CDK4/6i and GLS1i could yield a more accurate assessment of this therapeutic regimen.This approach provides a promising therapeutic alternative for the management of Braf V600E and Braf WT Nras WT and melanoma treatment, particularly for patients who have developed resistance to MAPK-directed therapies and consequently face a paucity of viable treatment options.
Statistically, Braf and Nras mutations are almost mutually exclusive according to cBioPortal.Our research has established that treatment with CDK4/6i does not upregulate KGA expression in melanoma cells harboring Nras mutations.Combined treatment with CB-839 in these cells did not exhibit synergistic effects.These observations may partially elucidate the clinical ineffectiveness of the combined CDK4/6i and GLS1i in treating solid tumors with Ras mutations (clinicaltrials.gov-NCT03965845).These are relevant to developing a clinical trial as we need to test Braf mutations only.Interestingly, recent work revealed that the combination of CDK4/6i and MEKi induces senescence in pancreatic ductal adenocarcinoma 45,46 and suppressed tumor growth in Braf/MEK inhibitor-resistant melanoma 47 .Furthermore, the triple combination of Braf, MEK and CDK4/6 has shown promising results 48,49 and has been evaluated in a clinical trial (clinicaltrials.gov-NCT04720768).We also previously demonstrated that an mTORC1 inhibitor (everolimus) enhanced colony outgrowth after co-culture with palbociclib compared to palbociclib alone 26 .Recent cumulative research suggests that further investigation to induce a more durable senescent state through such cytostatic drug combinations could advance senolytic therapy targeting upregulated GLS1.Moreover, the co-administration of GLS1i to existing treatment regimens may permit dosage reduction, potentially diminishing adverse side effects and enhancing patient tolerance in clinical trials.CB-839 demonstrated limited efficacy as an antineoplastic drug, which was notably evident in human liver cancer cell lines with significant glutamine dependence.Consequently, its combined use with other pharmacological agents has been explored across various cancer types.
For instance, 5-fluorouracil (5-FU) has been noted to augment the antitumor activity of CB-839 in PIK3CA mutant colorectal cancers 50 , and the proteasome inhibitor carfilzomib has shown synergistic effects when used with CB-839 in multiple myeloma 51 .Given that a range of treatments, including chemotherapy and radiotherapy, are known to induce cellular senescence, employing CB-839 could strategically target cells rendered senescent following these therapies.Indeed, senescence induction has been documented in response to 5-FU 52 .
Collectively, our study elucidates the scientific underpinnings for employing GLS1i in oncology and opens avenues for repurposing CB-839, which has hitherto shown limited success in clinical trials.Strategizing to identify drug combinations that elicit senescence accompanied by KGA upregulation -targets amenable to GLS1i -represents an innovative therapeutic approach for cancer management.This strategy diverges from traditional methods that primarily aim to exploit the glutamine dependency of cancer cells to directly induce apoptosis.Further research to advance this novel senolytic strategy using CDK4/6i and GLS1i into clinical trials is anticipated, with a focus on targeting Braf V600E melanoma, inclusive of variants with acquired resistance to Braf and/or MEK inhibitors.
determine senescence induction (Fig 1A and Sup Fig 1).Senescence was not induced in Nras Q61 melanoma cell lines by palbociclib as Nras Q61 mutant activates both MAPK signaling and PI3K-mTOR pathway and we previously demonstrated that mTORC1 signaling activation is the resistance mechanism to CDK4/6i 26 .Therefore, cells harboring Nras Q61 are refractory to CDK4/6iinduced senescence.Quantitative PCR (QPCR) gene expression analysis revealed that KGA expression is induced only in 8 days palbociclib treatment-induced senescent cells but not 1-day palbociclib treatment-induced cell cycle arrested cells (Fig 1B).The expression of GAC and GLS2 are marginally changed, but these changes are not as robust as KGA upregulation (Fig 1C-D).
2 hours to assess the expression of KGA by QPCR.While KGA mRNA is reduced by approximately 50% in control cells after 2 hours of actinomycin D treatment, KGA mRNA is stabilized in palbociclib-induced Braf V600E melanoma senescent cells (Fig 1G).In contrast, KGA expression is unchanged by actinomycin D before and after treatment of CDK4/6i in Nras Q61R skmel2 cells (Fig 1F), demonstrating the specificity of KGA upregulation in Braf V600E melanoma cells.
Fig 2B), adding rigor and reproducibility to our results.We also assessed the remaining senescent marginally induced KGA expression (Fig 4A).Upregulated KGA was further verifiedd by western blot (Fig 4B; the top band represents KGA expression).We next treated VR Braf V600E melanoma cells with palbociclib for 1 day or 8 days, followed by CB-839 for an additional 3 days to assess apoptosis (as described in Fig 2A).We demonstrated that GLS1i induced senolysis in palbociclibinduced senescent cells, but not in cells without palbociclib treatment or palbociclib-induced cell cycle arrested cells (Fig 4C), consistent with results observed using vemurafenib sensitive Braf V600E melanoma cells.Morphological analysis of 1205 VR cells showed that palbociclib induced senescence-like morphology, whereas the combination effectively damaged cells (Sup Fig tumors treated with palbociclib-alone exhibited senescent cells and tumors treated with the combination of palbociclib and CB-839 showed less SA-βgal positivity than palbociclib-alone treated cells (Fig 5E top panel), suggesting CB-839 treatment effectively eliminated palbociclibinduced senescent cells in vivo.Further investigation using IHC staining revealed that GLS1 expression is upregulated in palbociclib-induced senescent cells, however, this upregulation was negated when tumors were concurrently treated with palbociclib and CB-839 (Fig 5E middle panel).Consistent with our in vitro results, we observed apoptosis by staining for cleaved Caspase-3 expression in tumors treated with palbociclib and CB-839 (Fig 5E bottom panel).This staining indicated that although CB-839 itself has an impact on the induction of tissue necrosis, reflecting and assessed senescence induction by SA-βgal assay and the expression of KGA by QPCR.Vemurafenib treatment for 8 days induced senescence (Fig 6A) and KGA expression (Fig 6B).The inhibition of the MAPK pathway was confirmed by the reduced phosphorylation of ERK (Fig 6B).