Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jul 1.
Published in final edited form as: J Invest Dermatol. 2021 Dec 20;142(7):1912–1922.e7. doi: 10.1016/j.jid.2021.11.035

Expression differences in BCL2 family members between uveal and cutaneous melanomas account for varying sensitivity to BH3 mimetics

Nabanita Mukherjee 1, Chiara R Dart 1,2, Carol M Amato 2, Adam Honig-Frand 2, James R Lambert 3, Karoline A Lambert 1, William A Robinson 2, Richard P Tobin 4, Martin D McCarter 4, Kasey L Couts 2, Mayumi Fujita 1,5,6, David A Norris 1,5, Yiqun G Shellman 1,6
PMCID: PMC9635014  NIHMSID: NIHMS1842928  PMID: 34942200

Abstract

Uveal Melanoma (UM) is a subtype of melanoma. Although they share a melanocytic origin with cutaneous melanoma (CM), patients with UM have few treatment options. BH3 mimetics are small molecule drugs that mimic pro-apoptotic BCL2 family members. We compared BCL2 family member expression between UM and CM using immunoblot and TCGA transcriptomic analysis. UM has a unique signature of low BFL1 and high PUMA compared to CM and 30 other cancer types, making them an attractive candidate for BH3 mimetics. We tested the efficacy of a BCL2 inhibitor (BCL2i) and MCL1 inhibitor (MCL1i) in UM, with viability assays, live-cell imaging, sphere assays, and mouse xenograft models. UM had a higher sensitivity to MCL1i compared with CM. Overexpression of BFL1 or knockdown of PUMA made the UM more resistant to MCL1i. In contrast, MEKi treatment in CM made them more sensitive to MCL1i. However, MCL1i-alone treatment was not very effective to reduce the uveal melanoma initiating cells (UMICs); to overcome, we employed a combination of MCL1i with BCL2i that synergistically inhibited UMIC’s capacity to expand. Overall, we identify a distinct expression profile of BCL2 family members for UM that makes them susceptible to BH3 mimetics.

INTRODUCTION

Uveal melanoma (UM) is a rare subtype of melanoma arising from the melanocytes of the eye (Jager et al., 2020, Kaliki and Shields, 2017, Khoja et al., 2019, Richards et al., 2020), and accounts for more than 85% of intraocular malignancies (Mahendraraj et al., 2016). Although UM and cutaneous melanoma (CM) both originate from melanocyte transformation, they differ in their incidence rate, genetic etiology, and oncogenic driver signaling pathways (Hoefsmit et al., 2020, Iwamoto et al., 2002, McLaughlin et al., 2005, Neale et al., 2001). Subsequently, targeted therapies and immunotherapies utilized in CM are not effective in UM (Croce et al., 2019, Schank and Hassel, 2019). For decades, UM lacked FDA-approved drugs, and treatment options were limited to surgery, radiation therapy, laser, and photodynamic therapy (Jager et al., 2020, Krantz et al., 2017). In 2021, the FDA granted a breakthrough therapy designation to tebentafusp (Goodman, 2021, Piperno-Neumann et al., 2021). However, this immunotherapy is limited to patients harboring HLA-A*02:01 (Goodman, 2021, Piperno-Neumann et al., 2021). Approximately 50% of the patients lack this specific HLA-type, thus there is still an extreme need for new, effective UM treatments.

The BCL2 family of proteins plays a crucial role in regulating cell death or survival, and dysregulation can trigger cancer development and affect sensitivity to anti-cancer therapies (D’Aguanno and Del Bufalo, 2020, Hata et al., 2015, Zhang et al., 2020). The BCL2 family of proteins includes three groups based on their functions: 1) the anti-apoptotic proteins (BCL2, BCLXL, BCLW, MCL1 and BFL1) keep the effectors in check and inhibit cell death, 2) the BH3-only pro-apoptotic proteins (PUMA, BAD, BMF, NOXA and BIM) are initiators of cell death that neutralize pro-survival proteins, and 3) multi-BCL2 homologous (BH) domain proteins (BAX, BAK, and BOK) are effectors of apoptosis (Kale et al., 2018, Warren et al., 2019). Interactions between different members are not mutually exclusive or equal, and various combinations of interactions control the initiation of apoptosis. BH3 mimetics are small-molecules that mimic the function of the BH3-only proteins by binding and inhibiting the anti-apoptotic members of the BCL2 family. They directly activate apoptosis, bypassing the requirement for upstream initiators, such as p53. BH3 mimetics have recently generated tremendous excitement as cancer treatments, due to the remarkable efficacy of the recent FDA approved BH3 mimetic BCL2 inhibitor (BCL2i) ABT-199 (venetoclax) in treating hematological malignancies (Chonghaile, 2019, Hird and Tron, 2019, Kehr and Vogler, 2021, Merino et al., 2018).

Here, we evaluated the in vitro and in vivo efficacy of several BH3 mimetics in UM, including MCL1 inhibitors (MCL1i) and BCL2i, as single drugs and in combination. We also compared BCL2 family proteins expression in UM and CM to determine how they contribute to the increased sensitivity to BH3 mimetics. Our results indicate that the unique expression pattern of low BFL1 and high PUMA in UM makes it sensitive to MCL1i treatment. Additionally, the combination MCL1i with ABT-199 effectively debulks (eradicates majority of tumor cells) and eliminates uveal melanoma-initiating cells, in vitro and in vivo. Thus, this work demonstrates that the expressions of BCL2 family members lead to difference in sensitivity to BH3 mimetics in UM vs CM. This study also highlights the potential of BH3 mimetics as effective and durable treatments for UM, particularly MCL1i in combination with ABT-199.

RESULTS

Uveal melanomas, but not cutaneous melanomas, are sensitive to single drug MCL1 inhibitors in vitro and in vivo

We tested the efficacy of single drug BH3 mimetics in a panel of UM and CM lines (Figure 1 and Supplementary Figure S1). These included ABT-199 (BCL2i) and S63845/S64315 (MIK665) (MCL1i) with doses ranging from 0.156 to 10 μM. We did not find significant differences in BCL2i sensitivity between the UM vs CM lines (Supplementary Figure S1a), but MCL1i sensitivity differences were dramatic (Figure 1a and Supplementary Figure S1b). UM lines were more sensitive to S64315/S63845 at nM doses or low μM doses, while CM lines were resistant to doses below 10 μM. The S63845/S64315 IC50 value in UM was approximately 6-fold lower than CM (p<0.001) (Figure 1b). Results held true for lines regardless of BAP1 expression (Supplementary Figure S1c). All UM lines responded to single drug MCL1i treatments, suggesting MCL1i alone is a promising treatment option for UM.

Figure 1. Single drug treatment of MCL1i inhibits UM in vitro and in vivo.

Figure 1.

Open in a new tab

(a) ATP assay of single drug treatments of S63845 on CM and UM cell lines. Dashed line represents the 50% viability. (b) Dot-Scatter plot with IC50 values for MCL1i treatment in CM and UM. (c) IncuCyte live cell imaging on MP41 with active caspase 3/7 analyses to assess proliferation and apoptosis. (d) Representative images from Figure 1c with caspase 3/7 staining (green) at 48 h post-treatment. Scale bar = 400 μm. (e) Tumor growth in a mouse xenograft model with MP41 line. (f) Mice weight across days during the treatment. *** indicates p < 0.001, ns indicates not significant. Within each significant treatment, the least significant p-value of the comparisons is displayed.

To evaluate whether MCL1i induces apoptosis and inhibits proliferation in UM cell lines, we performed IncuCyte live cell imaging to analyze proliferation and active caspase 3/7 in two representative UM lines, MP41 (Figure 1c–d and Supplementary Figure S2) and MP46 (Supplementary Figure S3). Single drug MCL1i significantly decreased proliferation and increased caspase 3/7 activation (p <0.001) compared to vehicle in MP41 and MP46, at the dose of 156 or 625 nM. In vivo, S63845 treatment significantly reduced tumor growth (p<0.001) (Figure 1e), without showing obvious toxicity as monitored by mouse weight (Figure 1f). Overall, these studies indicate that MCL1i is effective against UM in vitro and in vivo.

A lower expression of anti-apoptotic protein BFL1 in UM contributes to high MCL1i sensitivity

We hypothesized that differences in the levels of endogenous BCL2 family proteins between UM and CM may account for varying sensitivity to MCL1i. To test this mechanism, we performed immunoblotting (Figure 2). Of the eleven pro- and anti-apoptotic proteins, BFL1 was significantly lower, and BCL2 was significantly higher, in UM compared to CM (Figure 2a and b). Based on the mean value of the quantifications, UM also had higher PUMA (2.7 fold; p=0.17), lower BIM (3 fold; p=0.10), lower NOXA (2.1 fold; p = 0.07), and lower MCL1 (1.8 fold; p=0.24). However, these differences were not statistically significant, likely due to a limited sample size (Figure 2c).

Figure 2. BFL1 contribute to UM’s sensitivity to MCL1i.

Figure 2.

Open in a new tab

(a and b) Immunoblot showing the endogenous expression of anti-apoptotic (a) and BH3-only pro-apoptotic (b) BCL2 family of proteins in a panel of CM and UM cell lines. (c) Quantification of the immunoblot data of Figure A and B above. (d) ATP assay data of single drug treatments of S63845 on control and BFL1 Over-Expressed (OE) UM cell lines. Y-axis shows percentage of relative viability and X-axis indicates the dosages of drug in μM. (e) Immunoblots confirming the over expression. Error bars represent +/− SEM for all figures. * indicates p < 0.05; ** indicates p < 0.01.

Since we found higher levels of BFL1 in MCL1i-resitant CM cell lines, we next tested if overexpression (OE) of BFL1 in UM cell lines made them less sensitive to MCL1i (Figure 2d and e and Supplementary Figure S4). We chose the three UM lines with different sensitivity to MCL1i. Overexpressing BFL1 reduced sensitivity to MCL1i by 1.5 to 2.4 fold, compared to control cells. However, overexpressing BFL1 did not alter their sensitivity to BCL2i. These results were consistent across all three UM cell lines tested and indicate lower BFL1 expression in UM contributes to increased sensitivity to MCL1i.

Differences in the oncogenic MAPK signaling pathways contributes to the variable MCL1i sensitivity between UM and CM

CM and UM have different oncogenic driver mutations (Jager et al., 2020). The MAPK pathway is often constitutively activated in CM due to activating mutations in the BRAF and RAS genes (Maldonado et al., 2003). We hypothesized that the lack of MAPK activating mutations in UM contributes to their high MCL1i sensitivity, and that inhibition of the MAPK pathway in CM would confer MCL1i sensitivity, similar to UM. To test this hypothesis, we used the MEK inhibitor (MEKi) trametinib to treat the BRAF-mutated CM lines with MCL1i S64315 at sub-μM doses, either by itself or in combination with the MEKi (30 nM). MEKi treatment drastically enhanced MCL1i sensitivity in CM lines (p<0.001), indicating that the MAPK signaling pathway plays a role in determining MCL1i sensitivity (Figure 3a).

Figure 3. Oncogenic driver signaling pathway and pro-apoptotic proteins account for differences between UM vs CM.

Figure 3.

Open in a new tab

(a) ATP assay data of single drug treatments of S64315 combined with MEKi trametinib on CM cell lines. (b) Immunoblot with lysates collected after 24 h treatment with DMSO, or MEKi, and then probed for indicated proteins. Molecular weight markers are in kD. (c) ATP assay with control (sh control) or knockdown (KD) of PUMA (sh PUMA), BIM (sh BIM), BAD (sh BAD), and BMF (sh BMF) in Mel202 UM line to test response to MCL1i. (d) Immunoblots confirming the KD for Figure 3c. For Figure a and c, Y-axis shows percentage of relative viability and X-axis indicates the dosages of drug in μM. *** indicates p < 0.001.

We also found that the MEKi treatment of CM cell lines significantly increased the expression of BH3-only pro-apoptotic proteins such as PUMA, BIM, BAD, and BMF (Figure 3b). There were no consistent alterations of anti-apoptotic proteins in response to MEKi treatment in CM (Supplementary Figure S5). These results suggest low MAPK signaling in UM correlates with high expression of specific pro-apoptotic BCL2 family members. We next examined if knocking down the MAPK-regulated pro-apoptotic proteins in UM will make them less sensitive to MCL1i. As expected, knocking down PUMA, BMF, BIM and BAD in UM cells reduced sensitivity to MCL1i at least by 1.3 to 9 folds compared to the control (p<0.05, Figure 3c and Supplementary Figure S6), suggesting that pro-apoptotic proteins contribute to MCL1i efficacy. Taken together, these data indicate that low MAPK signaling in UM correlates with higher expression of certain pro-apoptotic proteins and contributes to its greater MCL1i sensitivity.

TCGA data reveal the low BCL2A1 (BFL1) and high BBC3 (PUMA) gene expression pattern is unique to UM

We analyzed transcriptomic data (RNA-seq) from UM and CM patient tumors for differential expression of pro- and anti-apoptosis genes, analyzing a total of 553 tumors, 80 UM and 473 CM. We focused on BCL2 family members that showed differential expression in our immunoblot or mechanistic studies (Figures 2 and 3). For clarity and consistency with the literature and our immunoblot analysis, we use the common aliases for the genes/mRNA (see Figure 4a). The mRNA expression of all eight genes were significantly different between the two subtypes (Figure 4a). In UM compared with CM, the anti-apoptotic members BFL1 and MCL1 were significantly lower, while BCL2 was higher; For the pro-apoptotic genes, PUMA, BMF and BAD were significantly higher, while BIM and NOXA were lower (Figure 4a). The difference in BFL1 was the most dramatic, with ~250 folds lower expression in UM compared to CM. These data were consistent with the protein expression data from our immunoblot analysis (Figure 2), thereby confirming the unique BFL1 and PUMA protein expression pattern seen in our UM cell lines are observed in patient tumors.

Figure 4. TCGA data indicate the clinical relevancy of differential expression of BCL2 proteins between CM and UM and lower BFL1 and higher PUMA in UM compared with other cancers.

Figure 4.

Open in a new tab

(a) The expression data for all the eight genes analyzed to compare the expressions between UM and CM. Out of all 8 genes listed, first 7 genes are significantly different between the two subtypes. Expression data of (b) BFL1 and (c) PUMA across various cancers. For b and c, the mRNA level was compared for UM vs each of the other types of cancers. The *** indicates the least significant p-value of the comparison at p<0.001. For b, Pheochromocytoma, Diffuse Glioma, and Prostate adenocarcinoma are additionally significant p<0.01 compared with UM.

The differential expression of BFL1 and PUMA were the most consistent between our immunoblot and TCGA data. We extended our analyses of BFL1 and PUMA to additional cancer types in the TCGA data, with a total of 10,071 tumors, comparing UM to 33 other cancer types. BFL1 expression was significantly lower in UM (p<0.01) compared with 30 out of 33 other cancers (Figure 4b). UM had the highest PUMA expression, which was significantly different from the 33 other cancers (p<0.001) (Figure 4c). Out of the 34 cancers analyzed, only UM showed the distinct trend of high PUMA and low BFL1 expression. These data suggest that UM has a unique BFL1 and PUMA expression profile, which we show leads to MCL1i sensitivity.

Combination treatment of MCL1i and ABT-199 kills and inhibits the self-renewability of UMICs in vitro and in vivo

Cancer stem cells (CSCs), a small population of cells with stem like features, are the key players responsible for relapse after drug treatment (Atashzar et al., 2020, Mukherjee et al., 2015b). CSC-like cells have been demonstrated in UM (Chen et al., 2020, Kalirai et al., 2011), and are referred to here as uveal melanoma initiating cells (UMICs). The promising results of our in vitro and in vivo studies of MCL1i prompted us to test the efficacy of the single drug MCL1i in UMICs (Figure 4). We used the primary sphere formation assay to enrich the UMIC population and the secondary sphere assay to assess the self-renewability (Mukherjee et al., 2016, Mukherjee et al., 2020a, Mukherjee et al., 2021, Mukherjee et al., 2017, Mukherjee et al., 2015a, Mukherjee et al., 2020b, Mukherjee et al., 2018). MCL1i alone disrupted primary spheres and prevented the formation of secondary spheres, but only at a high dose (10 μM) (Figure 5 a–c), suggesting that a combination treatment was necessary to kill UMICs.

Figure 5. The combination of MCL1i with BCL2i is more effective than MCL1i alone to inhibit the UMIC population in vitro and in vivo.

Figure 5.

Open in a new tab

(a-c), Effects of MCL1i alone on UMICs. (a) Bright field microscopy images of primary (upper panel) and secondary (lower panel) spheres; Relative viability of the primary (b), and the secondary spheres (c). (d-g) Effects of the combination of MCL1i with BCL2i (0.156 μM). (d) Images of primary (upper panel) and secondary (lower panel) spheres. (e) Relative viability of the primary spheres. (f) The combination synergistically inhibited the UMICs. CI values < 0.6 indicate strong synergism. (g) Relative viability of the secondary spheres. (h) The combination of MCL1i and BCL2i inhibited the UMIC-mediated tumor formation in vivo. Tumor-free survival curve shows a significantly longer tumor–free time in the combination group, compared to other groups. For visual clarity, the * is not shown in the figure. Scale bar = 100 μm.

The bulk of UM is sensitive to MCL1i, and our previous work has shown that CM MICs carrying BRAF wild type are sensitive to the combination of MCL1i+BCL2i (Mukherjee et al., 2020a). Further, UM displayed a high expression of BCL2 (Figure 2c and Figure 4a). Therefore, we examined if combining BCL2i (ABT-199) with MCL1i (S63845/S64315) would target UMICs. In vitro, this combination significantly decreased the number of primary spheres in all cell lines tested (p<0.001) (Figure 5d and Supplementary Figure S7). The ATP assay indicated that the combination significantly inhibited the viability of the UMIC population (Figure 5e) in a highly synergistic manner (Figure 5f). The secondary sphere assay results showed that the combination of each drug at 156 nM was effective in inhibiting the self-renewability of multiple UM cell lines (Figure 5d; Figure 5g and Supplementary Figure S8).

We also examined if this combination could inhibit UMIC-mediated tumor growth in vivo (Figure 5h). We enriched UMICs using the primary sphere assay, treated the spheres in vitro (vehicle, single drug or combined), and implanted the surviving, viable, single cell suspensions in mice (Supplementary Figure S9) (Mukherjee et al., 2016, Mukherjee et al., 2017). As assessed by tumor-free survival time (Figure 5h), the combination group was the most effective (p<0.05). Taken together, the results suggest that the combination of MCL1i and BCL2i reduced the UMIC population.

MCL1i in combination with ABT-199 induces apoptosis and reduces viability of UM in vitro and inhibits xenograft tumor growth in vivo

We next examined the effects of the combination treatment of S63845/S64315 and ABT-199 at the nM dose range on both UMIC and non-UMIC populations (Figure 6, Supplementary Figures S10, S11). IncuCyte live-cell imaging quantification of active caspase 3/7 and proliferation showed that the combination treatment of either S63845 or S64315 with ABT-199 significantly decreased proliferation and increased caspase 3/7 activation (p<0.01) (Figure 6a–b, Supplementary Figure S10). Additionally, the combination treatment, but not single drugs, significantly reduced cell viability (p<0.001) at both 156 nM and 625 nM (Supplementary Figure S11). The combination was highly synergistic (Figure 6c), and the IC50 of the combination was in the nM range for all four cell lines.

Figure 6. The combination has a synergistic effect on the bulk of UM cells in vitro and inhibit tumor growth in vivo in conventional mouse xenograft model.

Figure 6.

Open in a new tab

(a and b), IncuCyte live cell imaging with active caspase 3/7 analyses to study proliferation (a) and apoptosis (b) in response to vehicle, single drug or combination (156 nM) in MP41 line. (c) The combination synergistically inhibited the UM cells. CI values < 0.9 indicate synergism. Smaller CI values indicate stronger synergy. (d) Tumor growth in a mouse xenograft model with MP41. ** indicates p < 0.01; *** indicates p < 0.001, ns indicates not significant. (e) A simplified model to illustrate the unique distribution of BCL2 family proteins BFL1 and PUMA, which sensitizes UM to MCL1i.

Finally, we evaluated whether MCL1i and BCL2i combination therapy is effective in vivo using a standard mouse xenograft model. The combination treatment significantly inhibited tumor growth when compared to the control or single drug treatments (p<0.001) (Figure 6d). In the single drug study, the S63845 was administered on five consecutive days during week 1, followed by three times weekly for rest of the study. In the combination study, both S63845 and ABT-199 were administered twice per week at 12.5 mg/kg and 50 mg/kg respectively. There were no significant changes in mouse weight or cytotoxicity, as measured by complete blood count (Supplementary Figure S12), indicating no adverse effects. Altogether, these data indicate that MCL1i in combination with BCL2i is likely to be a highly effective and promising new treatment option for UM patients.

DISCUSSION

This study aimed to identify new treatment options for UM, and discovered a unique expression pattern of BCL2 family members in UM, supporting further exploration of the therapeutic potential of BH3-mimetics in UM. We examined the efficacy of BH3 mimetics in UM both in vitro and in vivo, and our results demonstrate the promising therapeutic potential of MCL1i for UM patients, especially in combination with the BCL2i.

Our data points to UM being particularly sensitive to MCL1i compared to CM (Figure 1). Our recent publication also show UM’s sensitivity to MCL1i plus azacytidine, supporting UM patients as good candidates for treatment regimens including MCL1i (Dart et al., 2021). Multiple assays identified differential pro- and anti-apoptotic protein expression as a contributing factor to MCL1i sensitivity, including low BFL1 and high PUMA (Figure 2). Based on our findings, we propose that the low BFL1 and high PUMA expression makes the UM “primed for death” upon the MCL1i treatment, tilting the balance between the pro-and anti-apoptotic protein ratio and inducing apoptosis (Figure 6e). However, our studies do not exclude other factors in UM may also modulate sensitivity to MCL1i.

Our data also indicate that the lack of activating mutations in the BRAF/MAPK pathway contributes to higher MCL1i sensitivity in UM. In CM, inhibiting this pathway with MEKi sensitized CM to MCL1i, making them similar to UM in response to MCL1i. These results are consistent with a previous report with a different MCL1i (Sale et al., 2019). In addition, MEKi treatment in CM increased various pro-apoptotic BCL2 family members, including PUMA, BAD, BMF and BIM, and their knockdown in UM decreased their sensitivity to MCL1i. In short, higher expression of pro-apoptotic proteins in UM contributes to its higher sensitivity to MCL1i. This contrasts with CM, which have low expression of pro-apoptotic proteins due to MAPK pathway activation. Interestingly, there was a higher expression of PUMA in UM cells compared to CM (Figure 2 and 4), consistent with the hypothesis that the lack of activating MAPK mutations in UM, at least in part, contributes to its higher PUMA expression (Figure 3). Due to the different driving oncogenic pathways in UM and CM, we recognize the limitation of comparing the effect of MAPK pathway. Future investigations are needed to test this hypothesis, for example, determining the effects on MCL1i sensitivity of activating MAPK pathway in UM with commonly mutated CM mutations.

Importantly, transcriptomic analyses of TCGA data comparing UM and CM displayed similar overall trends for the expression of BCL2 family members as our immunoblots. For example, both TCGA and immunoblot data demonstrated the significant differences of low BFL1 and high PUMA expression in UM vs CM. Overall, these results suggest that our findings in UM cell lines are likely to translate into the clinic.

Our analyses of TCGA data from multiple cancer types revealed that UM has a unique expression pattern of low BFL1 and high PUMA in patient tumors (Figure 4b and 4c). To strengthen this finding, we further conducted similar analysis using DepMap portal (https://depmap.org/portal/). Overall, UM cell lines displayed lower BFL1 and higher PUMA expression compared with the majority of cancer types (Supplementary figure S13), consistent with the finding with TCGA data. However, the DepMap database only includes 1371 samples and 9 UM samples vs 10071 samples and 80 UM samples in the TCGA database. Thus, analyses with TCGA have higher statistical power, and the smaller sample size in DepMap likely contribute to fewer cancer cell types that displayed significantly different expression pattern from UM (Supplementary figure S13). Taken together, these data support the ideas that UM may be a good candidate for MCL1i treatment due to their low BFL1 and high PUMA expression.

Previous studies reported molecular differences between UM and CM, including dissimilar mutational burden and oncogenic signaling pathways (D’Aguanno et al., 2021, Liu-Smith and Lu, 2020, Pandiani et al., 2017, van den Bosch et al., 2010, van der Kooij et al., 2019). Our study provides clear evidence that BFL1 and PUMA are uniquely expressed in UM compared to many other cancer types (Figure 4). BFL1 is one of the least characterized members of the anti-apoptotic BCL2 family members. It is often upregulated in various cancers and plays a crucial role in oncogenesis and therapeutic resistance of CM (Ashkenazi et al., 2017, Hind et al., 2015). Out of all anti-apoptotic BCL2 family members, BFL1 is the most structurally and functionally similar to MCL1. Further, both MCL1 and BFL1 antagonize the pro-apoptotic function of the same BH3-only BCL2 family member, NOXA. Thus, this overlapping function between BFL1 and MCL1 is likely a main reason that lower expression of BFL1 contributes to MCL1i sensitivity in UM.

The pro-apoptotic protein PUMA can neutralize all five anti-apoptotic BCL2 family members (Li, 2021, Mustata et al., 2011, Nakano and Vousden, 2001, Yu and Zhang, 2008), promoting cell death. Our data indicate that the unique expression pattern of low BFL1 and high PUMA in UM contributes to their high sensitivity to MCL1i (Figure 6e). Due to their importance in regulating cell death/survival, their expression should be considered for future drug development for UM.

Our data also shows that the combination treatment of MCL1i with BCL2i was highly effective to de-bulk and kill UMICs at nM doses, in vitro and in vivo. Resistance to single drug BH3 mimetics often occur through upregulation of other anti-apoptotic family members. For example, upregulation of MCL1 is an acquired mechanism of resistance during BCL2i (ABT-199) treatment in hematological malignancies, and consequent MCL1 inhibition is an effective strategy to overcome this resistance (Parry et al., 2021, Tahir et al., 2017). Thus, the proposed combination treatment may prevent acquired resistance by thwarting escape through MCL1 or BCL2 upregulation. Further, the combination’s potent killing effect on UMICs is promising for preventing relapse. The effectiveness of the combination is consistent with our previous finding in other melanomas, mainly those not carrying the common BRAF-V600 mutation (Mukherjee et al., 2020a).

Our approach with BH3 mimetics utilizes a different mechanism of action than the current UM treatment and clinical trials. The only FDA-approved drug for UM, tebentafusp, works by redirecting T cells to kill tumor cells expressing specific target antigens (Goodman, 2021, Piperno-Neumann et al., 2021). Targeted therapy approaches, using inhibitors of Gαq or its downstream signaling protein FAK, have been tested in UM in vitro and in vivo as single drugs or in combination (Hitchman et al., 2021, Lapadula et al., 2019, Paradis et al., 2021), and some are currently in clinical trial (NCT04720417). Thus, BH3 mimetics likely provide alternatives for patients not responding to current treatments.

Our data strongly implicates BH3-mimetics’ therapeutic potential for UM. The BH3 mimetic ABT-263 was previously studied in UM and showed effects by itself and in combination with inhibitors of mTOR, MEK and MDM2 in vitro and in vivo (Bellini et al., 2020, Némati et al., 2014). However, because ABT-263 is a pan inhibitor of BCL2/BCLXL/BCLW, it can cause serious side-effects, including thrombocytopenia, which limits its clinical use. Our results also indicate MCL1i was more effective than ABT-263 as a single drug (data not shown). Currently, Phase I/II/III clinical trials with MCL1i in combination with BCL2i are ongoing (NCT03672695; NCT03218683) in patients with hematological malignancies, and early results suggest that these combinations are relatively well tolerated. Together, these clinical trials and our preclinical data support future testing of BH3 mimetic combinations in UM patients.

In summary, this study shows that the unique expression pattern of low BFL1 and high PUMA in UM make them susceptible to MCL1i-mediated apoptosis. Combination of MCL1i and BCL2i may prevent relapse by targeting UM cancer-initiating cells. Therefore, this combination is a promising treatment that may be both highly efficacious and durable for patients with this difficult-to-treat form of melanoma.

MATERIALS AND METHODS

Creation of BFL1 overexpressed Cell Lines

UM cells were transduced with BFL1 (NM_004049) Human Tagged ORF Clone Lentiviral Particles (#RC201965L4V) or Lenti-ORF Control Particles (#PS100093V) (Origene Technologies, Inc., Rockville, MD) per the manufacturer’s protocol. Cells were puromycin selected (0.25–1 μg/ml).

Analysis of the TCGA dataset

Gene expression data and clinical information for UM and CM were downloaded from the FireBrowse website (http://firebrowse.org/), DepMap portal (https://depmap.org/portal/) and cbioportal (Cerami et al., 2012) (http://cbioportal.org). The databases were accessed from May 2020 to October 2021 and analyzed with GraphPad Prism V8 software (https://www.graphpad.com/). For comparing UM vs CM samples, statistical significance was determined using multiple t-tests, corrected with multiple comparisons using the Holm-Sidak method. Each gene was analyzed individually, without assuming a consistent standard deviation. To determine whether the expression in UM was significantly different from other cancers, one-way ANOVA was used followed by Dunnett post-hoc test. The raw data was converted to log2 scale with the “value+1” method used by TCGA and DepMap.

Other Methods

Additional materials and methods are in the Supplementary Materials.

Supplementary Material

Supplemental materials and methods
Supplemental Figures

ACKNOWLEDGEMENTS

This work was supported in part by a Veterans Administration merit grant (BX000141) from the Department of Veterans Affairs (Veterans Health Administration, Office of Research and Development, Biomedical Laboratory Research and Development) to DAN, and a pilot grant (P30AR057212) from University of Colorado Skin Disease Research Center grant to YGS. CD was supported by Gates Center Summer Internship Program, which is funded by generous charitable gifts from Rhondda and Peter Grant, Monty and Frank Kugeler, and the Walter S. Rosenberry III Charitable Trust. AHF was supported by the Cancer Research Experience for Undergraduates (CREU) funded by the University of Colorado Cancer Center. We thank the CU Office of Laboratory Animal Resources (OLAR) Veterinary technicians for their help with the animal experiments. We are grateful to CU Comparative Pathology Shared Resource for the analysis of mouse blood. We also express our gratitude to the CU Cell Technologies Shared Resource (Supported by Grant P30CA046934) core facility for their technical assistance in use of the IncuCyte S3 Live-Cell Analysis System.

Abbreviations:

UM

uveal melanoma

CM

cutaneous melanoma

MCL1i

MCL1 inhibitor

BCL2i

BCL2 inhibitor

UMIC

uveal melanoma initiating cells

Footnotes

CONFLICTS OF INTEREST

The authors state no conflicts of interest.

Data availability statement

No datasets were generated or analyzed during the current study.

REFERENCES

  1. Ashkenazi A, Fairbrother WJ, Leverson JD, Souers AJ. From basic apoptosis discoveries to advanced selective BCL-2 family inhibitors. Nat Rev Drug Discov 2017;16(4):273–84. [DOI] [PubMed] [Google Scholar]
  2. Atashzar MR, Baharlou R, Karami J, Abdollahi H, Rezaei R, Pourramezan F, et al. Cancer stem cells: A review from origin to therapeutic implications. J Cell Physiol 2020;235(2):790–803. [DOI] [PubMed] [Google Scholar]
  3. Bellini L, Strub T, Habel N, Pandiani C, Marchetti S, Martel A, et al. Endoplasmic reticulum stress mediates resistance to BCL-2 inhibitor in uveal melanoma cells. Cell Death Discovery 2020;6(1):22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer discovery 2012;2(5):401–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen YN, Li Y, Wei WB. Research Progress of Cancer Stem Cells in Uveal Melanoma. Onco Targets Ther 2020;13:12243–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chonghaile TN. BH3 mimetics: Weapons of cancer cell destruction. Science Translational Medicine 2019;11(475):eaaw5311. [Google Scholar]
  7. Croce M, Ferrini S, Pfeffer U, Gangemi R. Targeted Therapy of Uveal Melanoma: Recent Failures and New Perspectives. Cancers (Basel) 2019;11(6):846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. D’Aguanno S, Mallone F, Marenco M, Del Bufalo D, Moramarco A. Hypoxia-dependent drivers of melanoma progression. J Exp Clin Cancer Res 2021;40(1):159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. D’Aguanno S, Del Bufalo D. Inhibition of Anti-Apoptotic Bcl-2 Proteins in Preclinical and Clinical Studies: Current Overview in Cancer. Cells 2020;9(5):1287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dart CR, Mukherjee N, Amato CM, Goulding A, MacBeth M, Van Gulick R, et al. A Novel Regimen for Treating Melanoma: MCL1 Inhibitors and Azacitidine. Pharmaceuticals (Basel) 2021;14(8):749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goodman A. Tebentafusp: First Treatment to Improve Overall Survival in Metastatic Uveal Melanoma. 2021.
  12. Hata AN, Engelman JA, Faber AC. The BCL2 Family: Key Mediators of the Apoptotic Response to Targeted Anticancer Therapeutics. Cancer discovery 2015;5(5):475–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hind CK, Carter MJ, Harris CL, Chan HT, James S, Cragg MS. Role of the pro-survival molecule Bfl-1 in melanoma. Int J Biochem Cell Biol 2015;59:94–102. [DOI] [PubMed] [Google Scholar]
  14. Hird AW, Tron AE. Recent advances in the development of Mcl-1 inhibitors for cancer therapy. Pharmacology & Therapeutics 2019;198:59–67. [DOI] [PubMed] [Google Scholar]
  15. Hitchman TD, Bayshtok G, Ceraudo E, Moore AR, Lee C, Jia R, et al. Combined Inhibition of Gαq and MEK Enhances Therapeutic Efficacy in Uveal Melanoma. Clinical Cancer Research 2021;27(5):1476–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hoefsmit EP, Rozeman EA, Van TM, Dimitriadis P, Krijgsman O, Conway JW, et al. Comprehensive analysis of cutaneous and uveal melanoma liver metastases. Journal for ImmunoTherapy of Cancer 2020;8(2):e001501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Iwamoto S, Burrows RC, Kalina RE, George D, Boehm M, Bothwell MA, et al. Immunophenotypic Differences Between Uveal and Cutaneous Melanomas. Archives of Ophthalmology 2002;120(4):466–70. [DOI] [PubMed] [Google Scholar]
  18. Jager MJ, Shields CL, Cebulla CM, Abdel-Rahman MH, Grossniklaus HE, Stern M-H, et al. Uveal melanoma. Nature Reviews Disease Primers 2020;6(1):24. [DOI] [PubMed] [Google Scholar]
  19. Kale J, Osterlund EJ, Andrews DW. BCL-2 family proteins: changing partners in the dance towards death. Cell Death & Differentiation 2018;25(1):65–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kaliki S, Shields CL. Uveal melanoma: relatively rare but deadly cancer. Eye (Lond) 2017;31(2):241–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kalirai H, Damato BE, Coupland SE. Uveal melanoma cell lines contain stem-like cells that self-renew, produce differentiated progeny, and survive chemotherapy. Invest Ophthalmol Vis Sci 2011;52(11):8458–66. [DOI] [PubMed] [Google Scholar]
  22. Kehr S, Vogler M. It’s time to die: BH3 mimetics in solid tumors. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research 2021;1868(5):118987. [DOI] [PubMed] [Google Scholar]
  23. Khoja L, Atenafu EG, Suciu S, Leyvraz S, Sato T, Marshall E, et al. Meta-analysis in metastatic uveal melanoma to determine progression free and overall survival benchmarks: an international rare cancers initiative (IRCI) ocular melanoma study. Ann Oncol 2019;30(8):1370–80. [DOI] [PubMed] [Google Scholar]
  24. Krantz BA, Dave N, Komatsubara KM, Marr BP, Carvajal RD. Uveal melanoma: epidemiology, etiology, and treatment of primary disease. Clin Ophthalmol 2017;11:279–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lapadula D, Farias E, Randolph CE, Purwin TJ, McGrath D, Charpentier TH, et al. Effects of Oncogenic Gαq and Gα11 Inhibition by FR900359 in Uveal Melanoma. Molecular Cancer Research 2019;17(4):963–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Li M. The role of P53 up-regulated modulator of apoptosis (PUMA) in ovarian development, cardiovascular and neurodegenerative diseases. Apoptosis 2021;26(5):235–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Liu-Smith F, Lu Y. Opposite Roles of BAP1 in Overall Survival of Uveal Melanoma and Cutaneous Melanoma. J Clin Med 2020;9(2):411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mahendraraj K, Lau CS, Lee I, Chamberlain RS. Trends in incidence, survival, and management of uveal melanoma: a population-based study of 7,516 patients from the Surveillance, Epidemiology, and End Results database (1973–2012). Clinical ophthalmology (Auckland, NZ) 2016;10:2113–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Maldonado JL, Fridlyand J, Patel H, Jain AN, Busam K, Kageshita T, et al. Determinants of BRAF mutations in primary melanomas. J Natl Cancer Inst 2003;95(24):1878–90. [DOI] [PubMed] [Google Scholar]
  30. McLaughlin CC, Wu XC, Jemal A, Martin HJ, Roche LM, Chen VW. Incidence of noncutaneous melanomas in the U.S. Cancer 2005;103(5):1000–7. [DOI] [PubMed] [Google Scholar]
  31. Merino D, Kelly GL, Lessene G, Wei AH, Roberts AW, Strasser A. BH3-Mimetic Drugs: Blazing the Trail for New Cancer Medicines. Cancer Cell 2018;34(6):879–91. [DOI] [PubMed] [Google Scholar]
  32. Mukherjee N, Almeida A, Partyka KA, Lu Y, Schwan JV, Lambert K, et al. Combining a GSI and BCL-2 inhibitor to overcome melanoma’s resistance to current treatments. Oncotarget 2016;7(51):84594–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mukherjee N, Amato CM, Skees J, Todd KJ, Lambert KA, Robinson WA, et al. Simultaneously Inhibiting BCL2 and MCL1 Is a Therapeutic Option for Patients with Advanced Melanoma. Cancers (Basel) 2020a;12(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mukherjee N, Lambert KA, Norris DA, Shellman YG. Enrichment of Melanoma Stem-Like Cells via Sphere Assays. Methods Mol Biol. 2265. 2021/03/12 ed; 2021. p. 185–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mukherjee N, Lu Y, Almeida A, Lambert K, Shiau CW, Su JC, et al. Use of a MCL-1 inhibitor alone to de-bulk melanoma and in combination to kill melanoma initiating cells. Oncotarget 2017;8(29):46801–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Mukherjee N, Reuland SN, Lu Y, Luo Y, Lambert K, Fujita M, et al. Combining a BCL2 inhibitor with the retinoid derivative fenretinide targets melanoma cells including melanoma initiating cells. J Invest Dermatol 2015a;135(3):842–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mukherjee N, Schwan JV, Fujita M, Norris DA, Shellman YG. Alternative Treatments For Melanoma: Targeting BCL-2 Family Members to De-Bulk and Kill Cancer Stem Cells. J Invest Dermatol 2015b;135(9):2155–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mukherjee N, Skees J, Todd KJ, West DA, Lambert KA, Robinson WA, et al. MCL1 inhibitors S63845/MIK665 plus Navitoclax synergistically kill difficult-to-treat melanoma cells. Cell Death Dis 2020b;11(6):443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mukherjee N, Strosnider A, Vagher B, Lambert KA, Slaven S, Robinson WA, et al. BH3 mimetics induce apoptosis independent of DRP-1 in melanoma. Cell death & disease 2018;9(9):907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Mustata G, Li M, Zevola N, Bakan A, Zhang L, Epperly M, et al. Development of small-molecule PUMA inhibitors for mitigating radiation-induced cell death. Curr Top Med Chem 2011;11(3):281–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Nakano K, Vousden KH. PUMA, a novel proapoptotic gene, is induced by p53. Mol Cell 2001;7(3):683–94. [DOI] [PubMed] [Google Scholar]
  42. Neale MH, Myatt NE, Khoury GG, Weaver P, Lamont A, Hungerford JL, et al. Comparison of the ex vivo chemosensitivity of uveal and cutaneous melanoma. Melanoma Research 2001;11(6). [DOI] [PubMed] [Google Scholar]
  43. Némati F, de Montrion C, Lang G, Kraus-Berthier L, Carita G, Sastre-Garau X, et al. Targeting Bcl-2/Bcl-XL Induces Antitumor Activity in Uveal Melanoma Patient-Derived Xenografts. PLoS One 2014;9(1):e80836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pandiani C, Béranger GE, Leclerc J, Ballotti R, Bertolotto C. Focus on cutaneous and uveal melanoma specificities. Genes Dev 2017;31(8):724–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Paradis JS, Acosta M, Saddawi-Konefka R, Kishore A, Lubrano S, Gomes F, et al. Synthetic Lethal Screens Reveal Cotargeting FAK and MEK as a Multimodal Precision Therapy for GNAQ-Driven Uveal Melanoma. Clinical Cancer Research 2021;27(11):3190–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Parry N, Wheadon H, Copland M. The application of BH3 mimetics in myeloid leukemias. Cell death & disease 2021;12(2):222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Piperno-Neumann S, Hassel JC, Rutkowski P, Baurain J-F, Butler MO, Schlaak M, et al. Abstract CT002: Phase 3 randomized trial comparing tebentafusp with investigator’s choice in first line metastatic uveal melanoma. Cancer research 2021;81(13 Supplement):CT002–CT. [Google Scholar]
  48. Richards JR, Yoo JH, Shin D, Odelberg SJ. Mouse models of uveal melanoma: Strengths, weaknesses, and future directions. Pigment cell & melanoma research 2020;33(2):264–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sale MJ, Minihane E, Monks NR, Gilley R, Richards FM, Schifferli KP, et al. Targeting melanoma’s MCL1 bias unleashes the apoptotic potential of BRAF and ERK1/2 pathway inhibitors. Nature Communications 2019;10(1):5167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Schank TE, Hassel JC. Immunotherapies for the Treatment of Uveal Melanoma-History and Future. Cancers (Basel) 2019;11(8):1048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Tahir SK, Smith ML, Hessler P, Rapp LR, Idler KB, Park CH, et al. Potential mechanisms of resistance to venetoclax and strategies to circumvent it. BMC Cancer 2017;17(1):399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. van den Bosch T, Kilic E, Paridaens D, de Klein A. Genetics of Uveal Melanoma and Cutaneous Melanoma: Two of a Kind? Dermatology Research and Practice 2010;2010:360136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. van der Kooij MK, Speetjens FM, van der Burg SH, Kapiteijn E. Uveal Versus Cutaneous Melanoma; Same Origin, Very Distinct Tumor Types. Cancers (Basel) 2019;11(6):845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Warren CFA, Wong-Brown MW, Bowden NA. BCL-2 family isoforms in apoptosis and cancer. Cell death & disease 2019;10(3):177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Yu J, Zhang L. PUMA, a potent killer with or without p53. Oncogene 2008;27 Suppl 1(Suppl 1):S71–S83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang X, Liu X, Zhou D, Zheng G. Targeting anti-apoptotic BCL-2 family proteins for cancer treatment. Future Medicinal Chemistry 2020;12(7):563–5. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental materials and methods
NIHMS1842928-supplement-Supplemental_materials_and_methods.docx (28.6KB, docx)
Supplemental Figures
NIHMS1842928-supplement-Supplemental_Figures.pdf (1.8MB, pdf)

Data Availability Statement

No datasets were generated or analyzed during the current study.

RESOURCES