Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Nov 1.
Published in final edited form as: J Invest Dermatol. 2020 May 7;140(11):2242–2252.e7. doi: 10.1016/j.jid.2020.02.047

Targeting PHGDH upregulation reduces glutathione levels and re-sensitizes resistant NRAS mutant melanoma to MEK inhibition

Mai Q Nguyen 1, Jessica LF Teh 1, Timothy J Purwin 1, Inna Chervoneva 2, Michael A Davies 3, Katherine L Nathanson 4, Phil F Cheng 5, Mitchell P Levesque 5, Reinhard Dummer 5, Andrew E Aplin 1,6
PMCID: PMC7606255  NIHMSID: NIHMS1591998  PMID: 32389536

Abstract

Melanomas frequently harbor activating NRAS mutations leading to activation of MEK-ERK1/2 signaling; however, clinical efficacy of inhibitors to this pathway are limited by resistance. Tumors re-wire metabolic pathways in response to stress signals such as targeted inhibitors and drug resistance, but most therapeutic resistant pre-clinical models are generated in conditions that lack physiological metabolism. We generated human NRAS mutant melanoma xenografts that were resistant to the MEK inhibitor (MEKi), PD0325901, in vivo. MEKi-resistant (MEKiR) cells showed cross-resistance to the structurally distinct MEKi, trametinib, and elevated ERK1/2 phosphorylation and downstream signaling. Additionally, we observed upregulation of the serine synthesis pathway and phosphoglycerate dehydrogenase (PHGDH), a key enzyme in this pathway. Suppressing PHGDH in MEKiR cells together with MEKi treatment decreased oxidative stress tolerance and cell proliferation. Together, our data suggest targeting PHGDH as a potential strategy in overcoming MEKi resistance.

Graphical Abstract

graphic file with name nihms-1591998-f0006.jpg

INTRODUCTION

Melanoma is the deadliest form of skin cancer and among cancers with the fastest rising incidence (Howe et al, 2006). 25-30% of melanomas harbor activating mutations in the GTPase neuroblastoma RAS viral oncogene homolog (NRAS). Efforts to target NRAS mutant melanoma have involved inhibiting downstream effector pathways of RAS such as the receptor tyrosine kinase effector Rapidly Accelerated Fibrosarcoma kinases (RAF), mitogen-activated protein kinase (MEK1/2), extracellular signal-regulated proteins (ERK1/2) (Morris et al, 2013, Rebecca et al, 2014), the phosphoinositide 3-kinase (PI3K)-v-akt-murine thyoma viral oncogene homolog (AKT) pathway (Posch et al, 2013, Roberts et al, 2012, Schicher et al, 2009), TANK-binding kinase 1 (TBK1) (Vu and Aplin, 2014), and Rho GTPase-activated serine/threonine kinases (ROCK1/2) (Vogel et al, 2015). The most well-studied of these effector pathways is the (MEK)-ERK1/2 cascade. Unlike patients with mutations in the v-Raf murine sarcoma viral oncogene homolog B (BRAF), for whom a combination of BRAF inhibitor and MEK inhibitor (MEKi) improves clinical outcomes (Flaherty et al, 2012, Flaherty et al., 2012, Gupta et al, 2014), there are currently no FDA-approved targeted therapies for NRAS mutant melanoma patients. Phase I/II trials of MEKi in NRAS mutant melanoma patients have shown limited clinical efficacy and invariably lead to resistance after a few months (Ascierto et al, 2013, Falchook et al, 2012, Martinez-Garcia et al, 2012, Zimmer et al, 2014). The mechanisms underlying MEKi resistance are poorly characterized and further analysis is warranted to provide the basis for therapeutic approaches.

As first described by Otto Warburg (Warburg, 1956), tumor cells preferentially use aerobic glycolysis instead of mitochondrial respiration to support their rapid growth (Hsu and Sabatini, 2008, Pavlova and Thompson, 2016, Vander Heidin et al, 2009). In addition, cancer cells can rewire amino acid and lipid metabolism to adapt to an elevated demand for metabolic precursors to drive macromolecular synthesis (Pavlova and Thompson, 2016). Although poorly understood in melanoma, it has been demonstrated in other cancers that metabolic pathways can be rewired as tumors progress and targeting molecular changes can be efficacious in cancer treatment. The gene encoding glucose transporter 1 (GLUT1) was overexpressed in colorectal cancer lines with Kirsten rat sarcoma viral oncogene homolog (KRAS) or BRAF mutations. High GLUT1-expressing cells exhibit enhanced glucose uptake and increased glycolysis. The glycolysis inhibitor, 3-bromopyruvate, was shown to preferentially suppress the growth of these cell lines (Yun et al, 2009). Similarly, inhibiting glycolysis by targeting lactate dehydrogenase A, hence, blocking the conversion of pyruvate to lactate, compromises the ability of breast cancer cells to proliferate under hypoxia (Fantin et al, 2006). Targeting mitochondrial function in breast cancer lines can also induce apoptosis and decrease proliferation (Bonnet et al, 2007). These studies provide evidence that targeting metabolic processes is a potential strategy to inhibit cancer growth.

In this study, we showed up-regulation of phosphoglycerate dehydrogenase (PHGDH), which catalyzes the first committed step in the serine biosynthesis pathway, in NRAS mutant tumor xenografts that have acquired resistance to the MEKi, PD0325901 (PD901). Upon PHGDH suppression in combination with PD901 treatment, we observed a decrease in the ratio of reduced to oxidized glutathione levels, which was indicative of cellular ability to combat oxidative stress. Inhibiting PHGDH molecularly re-sensitized MEKiR cells to PD901 treatment. These results suggest that metabolic enzymes such as PHGDH are potential targets to overcome MEKi resistance.

RESULTS

NRAS mutant melanoma tumors acquire resistance to MEK inhibition in vivo.

To generate MEKiR melanomas in conditions with physiological relevance, WM1366 xenografts were treated continuously with PD901 in athymic mice. Growth of WM1366 xenografts was delayed before resistance developed, defined as achieving polynomial or exponential growth rate in the presence of drug (Fig. 1a). The length of time to resistance was variable, ranging from 20 to 60 days. We isolated cell lines from tumors and tested them in 2D crystal violet proliferation assays. Cell lines generated from vehicle-treated tumors served as controls (Ctl). In vitro, MEKiR cells showed resistance to PD901 (Fig. 1b), as well as to the structurally distinct MEKi, trametinib (Fig. 1c). IC50 values ranged between 65-420 nM for PD901 and 27-100 nM for trametinib. By contrast, IC50 values for Ctl cells were <10 nM for both drugs (Fig. 1b–c and Table 1). We combined data for both Ctl lines, calculated the average percent survival at different drug doses for either PD901 or trametinib, and generated overall dose curves for Ctl lines. We generated similar overall dose curves for MEKiR lines. The dose curves were found to be significantly different between MEKiR cells and Ctl cells treated with either PD901 or trametinib (Fig. 1d). Since PD901 and trametinib inhibit MEK1/2 via different mechanisms, these data suggest that the MEKiR cell lines are broadly resistant to MEK inhibition.

Fig 1. Characterization of MEKiR progressive tumors and cell lines.

Fig 1.

Open in a new tab

a, Tumor-bearing mice were fed chow containing PD901 or Ctl chow and tumor volumes were measured using digital calipers. Resistant tumors were isolated once they achieved polynomial or exponential growth rate and before they reached 1500 mm3 (MEKiR #1, 2, 4). In the case of MEKiR #3, the tumor achieved polynomial growth rate and was isolated at the experiment endpoint of day 73. b & c, Cell lines were treated with DMSO or different concentrations of PD901 (b) or trametinib (c) (0.1, 0.5, 1, 5, 10, 50, 100, or 500 nM) for 6 days of total treatment. Medium and inhibitor were replenished every 2 days. Plates were then PBS washed, and fixed/stained with 0.25% crystal violet in buffered formalin. Quantification of percent area coverage is shown with each cell line being normalized to the DMSO condition (n=3; error bars indicate standard error). d, Mean proportion surviving (%) of combined Ctl cells and combined MEKiR cells for PD901 or trametinib were plotted against log10dose to create overall dose curves and statistical analysis was carried out as described in the Materials and Methods section (*** P<0.001).

Table 1. IC50 values of Ctl and MEKiR cell lines treated with PD901 and trametinib.

These were generated using a log fit curve equation from crystal violet proliferation assay data with different concentrations of PD901 or trametinib.

IC50 PD901 (nM) IC50 Trametinib (nM)
Ctl #2 8.4 7.4
Ctl #4 9.3 5.3
MEKiR #1 122.8 59.7
MEKiR #2 420.0 99.8
MEKiR #3 65.0 27.4
MEKiR #4 74.3 41.8
Open in a new tab

Targeting MEK-ERK1/2 pathway re-activation is not sufficient to overcome the MEKiR phenotype.

To examine phospho-proteomic alterations in MEKiR cells, we performed high-throughput antibody-based reverse-phase protein array (RPPA). Our results showed that downstream targets of ERK1/2, p-p90RSK (ribosomal protein S6 kinase α-1) and FRA1 (Fos-regulated antigen 1), were upregulated in MEKiR cells compared to Ctl cells at basal condition. The same effects were observed at increasing doses of PD901 (Fig. 2a). By Western blot, levels of p-ERK1/2, FRA1, and p-p90RSK were consistently higher in MEKiR #1 compared to Ctl #2 at basal level and following PD901 treatment (Fig. 2b). We obtained similar findings comparing the other MEKiR cell lines with Ctl #2 (Supplementary Fig. 1a). These data validate the RPPA findings for p-p90RSK and FRA1 in MEKiR lines.

Fig 2. Targeting MEK-ERK1/2 pathway re-activation is not sufficient to overcome the MEKiR phenotype.

Fig 2.

Open in a new tab

a, A heatmap showing RPPA data from Ctl and MEKiR cell lines treated with DMSO or PD901 (1, 10, or 100 nM) for 24 hours in triplicate for MEK-ERK1/2 pathway and cell cycle proteins. b, Cell lines were treated with DMSO or PD901 (0.1, 0.5, 1, 5, 10, or 100 nM) for 24 hours. Lysates were collected and analyzed by Western blot. Protein levels were quantitated and normalized to actin and Ctl #2 DMSO condition using densitometry (n=3; error bars indicate standard error; * P<0.05). c, Cell lines were treated with DMSO or SCH772984 (1 or 10 μM) for 12 hours. Lysates were collected and analyzed by Western blot. d, Cell lines were treated with DMSO or SCH772984 (1 or 10 μM) for 6 days of total treatment. Medium and inhibitor were changed every 2 days. Plates were then PBS washed, and fixed/stained with 0.25% crystal violet in buffered formalin.

We next examined the effect of an ERK1/2 inhibitor on MEKiR cell growth. We utilized SCH772984 and showed that p-ERK1/2, p-p90RSK, and FRA1 levels were markedly reduced after 12 hours of treatment in parental WM1366, Ctl #2, and MEKiR cell lines (Fig. 2c). By colony growth assay, there was a dose-dependent reduction in cell growth in all cell lines (Fig. 2d). However, there was no difference between the sensitivity of Ctl and MEKiR cell lines to SCH772984. These results suggest that even though MEKiR cells have elevated MEK-ERK1/2 activity, this pathway is likely not the sole reason for observed resistance to MEK inhibitors.

Additionally, it has been reported that mechanisms of resistance to BRAF/MEK pathway inhibitors can occur through acquired somatic genetic mutations (Kakadia et al, 2018). Therefore, we carried out targeted sequencing of tumor DNA. The variant calls detected one alteration, a deleterious mutation in the gene encoding AT-rich interactive domain-containing protein 2 (ARID2) in MEKiR #1 and #2 (Supplementary Fig. 1b). However, this result was not validated by Sanger sequencing (Supplementary Fig. 1c–d). No MEK1/2, secondary NRAS mutation, or alterations in other well-known hotspots were detected. Therefore, commonly observed genetic alterations to the MEK-ERK1/2 pathway are likely not the mechanism behind the MEKiR phenotype.

PHGDH is upregulated in MEKiR cells.

Next, we explored MEK-ERK1/2-independent pathways that may also contribute to resistance. Since our MEKiR tumors developed resistance in vivo, we considered effects of relevant physiological conditions and analyzed the expression of metabolic enzymes by RNA sequencing (RNA-seq) (Fig. 3a). We found that the serine and folic acid synthesis pathways were among the top ten enriched metabolic pathways in MEKiR tumors (Fig. 3a). Within these pathways, we found that PHGDH and methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) were among the top upregulated genes (Fig. 3b). PHGDH is the first and rate-limiting enzyme of the serine synthesis pathway while MTHFD2 is important for folic acid production. The serine synthesis pathway contributes two important precursors to the folate cycle, serine and glycine, and the two pathways are known to be closely connected (Locasale, 2013, Yang and Vousden, 2016) (Fig. 3c). The genes contained in each pathway when examined by RNA-seq were also found to highly overlap (Supplementary Fig. 2a). Thus, we decided to focus on these two pathways as potential therapeutic targets for MEKiR cells.

Fig 3. The serine synthesis pathway and PHGDH are upregulated in MEKiR tumors and cell lines.

Fig 3.

Open in a new tab

a, A bar plot showing the top ten enriched metabolic gene sets for MEKiR versus Ctl tumor samples. b, Heat-map showing unsupervised hierarchical clustering of samples based on the top 10 differentially expressed genes in the enriched pathways from a for normalized, log2-transformed RNA-seq data. c, Schematic of the serine synthesis pathway with PHGDH as the first and rate-limiting enzyme. The serine synthesis pathway is also connected to the folate cycle, the glutathione antioxidant system, and the methionine cycle. d, Lysates of the indicated tumors were collected and analyzed for PHGDH protein expression by Western blot. Black lines indicate lanes that were cropped out from the same blot. e, Cells were seeded in regular medium. Lysates were collected after 24 hours, analyzed by Western blot, and PHGDH protein levels normalized to actin and quantified using densitometry (n=3; error bars indicate standard error; * P<0.05, ** P<0.01).

The folate cycle is important for DNA synthesis and can be targeted using an FDA-approved, chemotherapeutic agent and immune system suppressant, methotrexate. We tested methotrexate to target MEKiR cell lines and found that parental WM1366, Ctl #2, MEKiR #1, and MEKiR #3 were all sensitive to increasing doses of methotrexate (Supplementary Fig. 2b). In fact, MEKiR cells seemed to be slightly less sensitive to folate cycle blockage than Ctl cells. Therefore, folic acid upregulation as an outcome of MTHFD2 and folic acid pathway upregulation and/or PHGDH upregulation is not likely an important metabolic adaptation for MEKiR cells.

PHGDH was previously identified as being frequently amplified and highly expressed in breast and melanoma cells (Possemato R, 2011) and reported to be upregulated in BRAF mutant melanomas resistant to BRAF inhibitors (Ross et al., 2017). This knowledge highlights the importance of the serine synthesis pathway, in general, and PHGDH, in particular, in the context of metabolic adaptation to targeted therapies. In agreement with the RNA-seq data, our tumor lysates showed an upregulation of PHGDH protein levels in MEKiR tumors compared to Ctl tumors (Fig. 3d). We found that the basal protein level of PHGDH was upregulated in the majority of MEKiR cells compared to Ctl cell lines with statistical significance in MEKiR #1 and #3 (Fig. 3e). Interestingly, PHGDH levels did not change with PD901 treatment (Supplementary Fig. 3a). Coupled with the upregulation of PHGDH in vivo that was sustained in vitro, these data suggest that increased PHGDH levels are not merely a short-term adaptive response to MEKi. Additionally, PHGDH upregulation was also identified in BRAF mutant xenografts resistant to MEK-ERK1/2 pathway inhibition (Supplementary Fig. 3b) (Ross et al., 2017). Therefore, PHGDH may play a broader role in the maintenance of tumor fitness in response to targeted therapy outside the context of NRAS mutant melanoma.

Suppression of PHGDH leads to decreased glutathione production and cell growth in MEKiR cells with PD901.

To determine requirement, we examined the effect of suppressing PHGDH expression by siRNA knockdown. We obtained highly efficient knockdown of PHGDH in both Ctl and MEKiR cell lines with two independent siRNAs (Fig. 4a and Supplementary Fig. 4a). Since we have determined that folic acid production is likely not essential for the resistant phenotype of MEKiR cells, we examined another output of the serine synthesis pathway, the glutathione antioxidant system (Fig. 3c). Glutathione levels determine cell capacity to neutralize reactive oxygen species and directly affect cell survival in oxidative stress, which is present in the tumor microenvironment and can also be induced by targeted inhibitors such as MEKi. Therefore, we performed reduced to oxidized glutathione (GSH/GSSG) assays with PHGDH siRNA knockdown and treatment with DMSO or PD901 to measure cell antioxidant capability. In the DMSO-treated condition, there was a slight decrease in GSH/GSSG ratio between siPHGDH and siCtl conditions for all cell lines. The addition of PD901 led to a significant decrease in GSH/GSSG ratio following PHGDH knockdown in PHGDH-overexpressing lines MEKiR #1 and #3 (Fig. 4b). This significant decrease, however, was not observed in Ctl #2. The reduction specific to PHGDH-overexpressing cells was further confirmed by similar results in Ctl #4, MEKiR #2, and MEKiR #4 (Supplementary Fig. 4b). Together, these data suggest that PHGDH upregulation and its subsequent impact on the glutathione antioxidant system might be beneficial in only PHGDH-overexpressing cells and especially when an additional stress is present, in this case, MEKi.

Fig 4. Suppression of PHGDH using siRNA knockdown reduces glutathione production and cell growth in the presence of MEK inhibitor.

Fig 4.

Open in a new tab

a, Cell lines were transfected with nontargeting or PHGDH-targeting siRNAs for 72 hours. Lysates were collected and analyzed by Western blot. b, Cell lines were transfected as described in a, for 24 hours and treated with DMSO or PD901 at 20 nM for 48 hours. Lysates were collected and analyzed for reduced and oxidized glutathione ratio using the Biovision, Inc. glutathione fluorometric assay kit. Results were normalized for each cell line to the siCtl DMSO condition (n=3; error bars indicate standard error; * P<0.05, ** P<0.01). c, Cell lines were transfected as described in a, for 24 hours and reseeded for treatment with DMSO or different concentrations of PD901 (5, 20, or 50 nM) for 6 days of total treatment. Medium and inhibitor were changed every 2 days. Plates were then PBS washed, and fixed/stained with 0.25% crystal violet in buffered formalin. d, Quantification of percent area coverage, each cell line normalized to DMSO condition (n=3; error bars indicate standard error; ns = not significant, * P<0.05, ** P<0.01, *** P<0.001).

To measure proliferative capacity, we employed a 2D crystal violet assay with PHGDH knockdown and treatment with DMSO or PD901. In the DMSO condition, we detected no change in cell proliferation between siCtl and siPHGDH conditions for any of the cell lines. However, the addition of PD901 resulted in a decrease in cell proliferation in both Ctl and MEKiR lines when PHGDH was knocked down compared to siCtl condition (Fig. 4c–d). This effect was more pronounced for PHGDH-overexpressing cells MEKiR #1 and MEKiR #3 compared to Ctl #2. There was statistical significance between knockdown and control conditions for MEKiR cell lines but not for Ctl cells. Thus, with the addition of PD901, cell proliferative capacity was significantly diminished with PHGDH knockdown in PHGDH-overexpressing cells. We further tested this effect in PHGDH CRISPR knockout cells generated from MEKiR #3, with loss of PHGDH expression confirmed by Western blot (Supplementary Fig. 4c). While PHGDH knockdown cells maintained similar growth capacity at basal conditions compared to parentals, complete knockout of PHGDH led to slower cell proliferation (Supplementary Fig. 4d). The addition of PD901, however, did not lead to a more significant reduction in cell growth.

To further investigate the role of PHGDH in response to MEKi, we overexpressed PHGDH in parental cells by lentiviral transduction. We confirmed PHGDH overexpression by Western blot (Supplementary Fig. 4e). By crystal violet assay, Ctl #2 WT PHGDH cells have compatible colony formation capability to their parental counterparts at basal level (DMSO). However, the addition of PD901 revealed a survival advantage in PHGDH-overexpressing cells versus to controls, although this observation did not reach statistical significance (Supplementary Fig. 4f). Overall, these results suggest that PHGDH plays a role in modulating glutathione levels and cell proliferation in the presence of MEKi and that knocking down PHGDH re-sensitizes MEKiR cells to MEK inhibition.

PHGDH-overexpression in MEKiR tumors may result from the selection of pre-existing high-PHGDH single cells.

Since increased PHGDH levels are apparently not a short-term adaptive response to MEKi, we examined potential mechanisms for PHGDH upregulation in MEKiR tumors and cell lines. A search for cross-cancer alterations of PHGDH using TCGA provisional data showed that melanoma is among cancers with the highest frequency of PHGDH DNA amplification (Fig. 5a). We investigated the effect of inhibiting key oncogenic pathways downstream of RAS on PHGDH levels to determine if any RAS player plays a role in regulating PHGDH levels. Treatment with either an ERK1/2 inhibitor (SCH772984), an Akt inhibitor (MK2206), a PI3K inhibitor (GDC0032), or a CDK4/6 inhibitor (Palbociclib) did not alter PHGDH levels in PHGDH-overexpressing cell lines MEKiR #3 and MEKiR #1 (Fig. 5b and Supplementary Fig. 5). The mTOR inhibitor, AZD2014, led to a partial reduction of PHGDH in MEKiR #3 and MEKiR #1, suggesting that mTOR may contribute to regulation of PHGDH levels. However, these data suggest that PHGDH upregulation in MEKiR cells is likely not driven primarily by the hyper-activation of key oncogenic pathways downstream of RAS.

Fig 5. PHGDH upregulation might arise from the enrichment of pre-existing high-PHGDH single cells.

Fig 5.

Open in a new tab

a, PHGDH amplification from TCGA provisional data across cancer types. Cancers (lung adeno = lung adenocarcinoma, pancreas = pancreatic adenocarcinoma, lung squ = lung squamous cell carcinoma) with the highest levels of PHGDH amplification are shown. b, Cell lines were treated with 1 μM of each of the indicated inhibitors or DMSO for 24 hours. Lysates were collected and analyzed by Western blot. c, Heterogeneity of PHGDH expression in immunotherapy treatment-naive tumors from a publicly available single-cell sequencing dataset.

Since it was previously shown that the gene encoding PHGDH resides in a region of chromosome 1p that is frequently amplified in melanoma (Possemato et al, 2011), it is possible that PHGDH upregulation in MEKiR cells might have come from pre-existing high-PHGDH-expressing single cells. To investigate PHGDH expression in melanoma tumors, we consulted a publicly available single-cell sequencing data set from a study looking at resistance to immune checkpoint inhibitors (Kumar et al, 2018). In immunotherapy treatment-naive malignant tumors, we observed heterogeneity in PHGDH expression between and within different melanoma tumors (Fig. 5c). This is consistent with the notion that PHGDH-overexpressing MEKiR tumors might represent a fraction of single cells with high PHGDH that were subsequently selected for during drug treatment. It might be useful, in the future with the emergence of more single-cell sequencing studies, to look at PHGDH levels pre- and post-treatment with targeted therapy to further test this hypothesis.

DISCUSSION

NRAS mutant melanoma patients make up a third of melanoma cases but have very limited treatment options. MEKi, while efficacious in BRAF mutant melanoma in combination with RAF inhibitors, have not yielded the same results in NRAS mutant melanomas. There have been evidence linking immunotherapy pre-treatment with an advantage in response to binimetinib (Dummer et al, 2017a, Dummer et al, 2017b) and MEKi-based combinations are in development (Roberts et al, 2012, Teh et al, 2016); however, no conclusive results have been reached with regards to their efficacy. To understand the molecular mechanisms for MEKi resistance in NRAS mutant melanoma and provide rationale for novel combinatorial approaches, we developed in vivo NRAS mutant melanoma models that display broad resistance to MEK inhibition. We found that besides MEK-ERK1/2 pathway reactivation, PHGDH was upregulated in MEKiR tumor xenografts and cell lines. Targeting PHGDH in combination with MEKi resulted in decreased oxidative stress tolerance and cell proliferation in PHGDH-overexpressing cells. We propose that this upregulation of PHGDH might be a result of clonal selection of pre-existing high-PHGDH cells during drug treatment. Together, our data suggest targeting a metabolic adaptation to overcome resistance to MEKi.

Although metabolic enzymes have been reported as potential targets for other cancers (Bonnet et al, 2007, Fantin et al, 2006, Yun et al, 2009), relatively little is known about metabolic alterations in melanomas undergoing therapeutic stress. BRAF mutant melanomas resistant to RAF inhibitors have been shown to induce an oxidative phosphorylation program and increased expression of PGC1α to support survival in therapeutic stress [38]. Treatment of resistant cells with inhibitors of oxidative phosphorylation re-sensitized cells to BRAF inhibition (Haq et al, 2013). In support of this study, our work showed for a different subset of MEKiR melanoma the utility of combining an inhibitor of the MEK-ERK1/2 pathway with depletion of a key metabolic enzyme. Future studies can further test the clinical relevance of these results with the emergence of better ways to modulate PHGDH activity in a dose-wise manner and with high specificity (Dong et al, 2018, Pacold et al, 2016).

The serine synthesis pathway has been shown to be essential in breast cancers and to directly contribute to oncogenesis (Locasale, 2013, Liu et al, 2013, Locasale et al, 2011, Ou et al, 2015). Upregulation of PHGDH resulted in increased flux through the serine synthesis pathway, flux through the TCA cycle, and increased glutathione production. Conversely, suppression of PHGDH in breast cancer cells with elevated levels caused a strong reduction in cell proliferation (Possemato et al, 2011). In agreement with previous data, we showed that cells with higher PHGDH levels were more sensitive to the effect of PHGDH knockdown in the presence of MEKi than cells without. It was previously shown that targeting the folic acid pathway was sufficient to reverse the effect of PHGDH overexpression and targeted therapy resistance (Ross et al, 2017). However, our data suggest that oxidative stress tolerance might be especially important for PHGDH-induced therapeutic resistance. Future work can evaluate the other outcomes of the serine synthesis pathway in contributing to the MEKiR phenotype.

A previous study identified PHGDH as being frequently amplified and highly expressed in breast and melanoma cells (Possemato et al, 2011). PHGDH upregulation has also been reported for the case of BRAF mutant melanoma resistant to BRAF inhibitor, the mechanism for which requires further elucidation. Another study in lung adenocarcinomas illustrated the role of PHGDH upregulation in erlotinib resistance (Dong et al, 2018). The mechanisms regulating PHGDH levels remain unclear. Some studies suggest a possible connection between enhanced MEK signaling and increased activity of ATF4 (Thiaville et al, 2008). ATF4 is a transcription factor responsible for transcribing amino acid starvation response genes such as those in the serine synthesis pathway, among others. Thus, reactivated MEK-ERK1/2 signaling might be responsible for the observed upregulation of PHGDH via ATF4. ATF4 can also be activated by the Akt-mTOR signaling pathway, which is parallel to the MEK-ERK1/2 pathway but is centrally regulated by RAS (Selvarajah et al, 2019). mTOR signaling has also been connected to NRF2, a transcription factor that has been described to regulate serine biosynthesis and the expression of anti-oxidant proteins (DeNicola et al, 2015). Blocking of growth signals by MEK inhibition might place additional burden on the parallel Akt-mTOR pathway to cope with stresses (nutrient, oxidative, etc.) by upregulating stress-responsive pathways.

Our data suggested a connection between PHGDH upregulation, as a possible result of DNA amplification, and its observed upregulation in MEKiR NRAS mutant melanoma tumors and together with previously published work, propose that the pre-existing PHGDH amplification might become especially important to maintain tumor fitness under therapeutic stress. Future work can look at PHGDH in other cases of drug resistance to expand our understanding of its functions and mechanisms of upregulation. It might also be worthwhile to examine PHGDH in BRAF, NF1, or triple WT melanoma background, as PHGDH amplification was also identified in these subsets.

MATERIALS AND METHODS

In vivo experiment

Animal experiments were performed in a facility that is accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. Studies were approved by the Institutional Animal Care and Use Committee. Female athymic mice (NU/J, Homozygous, Jackson, 6–8 weeks, 20–25 g) were used. Xenografts were allowed to reach 50–100 mm3 before mice were fed either Ctl (AIN-76A, n = 5) or MEKi (AIN-76A with a diet dose of 7 mg/kg PD0325901, n = 5) chow (Research Diets, Inc. (New Brunswick, NJ)). Digital caliper measurements were carried out as previously described (Teh et al.). Resistant tumors were isolated after achieving polynomial or exponential growth rate and before reaching 1500 mm3. For MEKiR #3, the tumor achieved polynomial growth rate and was isolated at day 73 (experiment endpoint).

Cell culture

WM1366 parental cells were cultured in MCDB153 medium containing 20% Leibovitz L-15 medium, 2% FBS, and 0.2% sodium bicarbonate (WM medium). WM1366 MEKiR cell lines were cultured in WM medium with 0.5 nmol/L PD901. Cell lines were validated as being NRAS mutant by STR analysis.

RPPA analysis

Cell lysates were prepared in triplicate and analyzed as previously described (Tibes et al, 2006) using 304 antibodies. Matlab was used to perform hierarchical clustering on median-centered, log2-transformed normalized data for a curated list of antibodies.

Western blot analysis

Proteins were extracted with Laemmli buffer, resolved by SDS-PAGE, and transferred to PVDF membranes. Immunoreactivity was detected using HRP-conjugated secondary antibodies (CalBioTech, Spring Valley, CA) and chemiluminescence HRP-recognizing substrates (ThermoScientific, Waltham, MA) on a VersaDoc Multi-Imager. Primary antibodies for p-ERK1/2 (9102), p90RSK (9333), p-p90RSK (9341), PHDGH (66350), SHMT2 (12762), p-S6 (2215S), S6 (2217), p-Akt (2965), Akt (9272), p-Rb (9307), and Hsp90 (4877) were obtained from Cell Signaling (Danvers, MA); ERK2 (sc-154), FRA1 (sc-605) from Santa Cruz Biotechnology (Santa Cruz, CA); and actin (A2066) from Sigma Aldrich (St. Louis, MO).

RNA interference

Cells (4 x 104) were plated per well of 12-well plates and transfected for 4 hours with siRNA at a final concentration of 25 nmol/L using Lipofectamine™ RNAiMax (Invitrogen). Cells were then fed WM medium with either PD901 at the final concentrations indicated in each experiment or an equivalent volume of DMSO. Medium was replenished at day 3 if the experiments lasted for more than 3 days. PHGDH siRNAs (#1: GAACUCACUUGUGGAAUGA; #2: GGAAAGACCCUGGGAAUUC) were purchased from Dharmacon Inc. (Lafayette, CO).

Colony formation assay

Cells (2 x 104) were plated per well of 6-well plates in complete medium with or without inhibitors, which were replenished every 2 days. After 6 days, cells were stained with 0.2% crystal violet in buffered formalin, and plates were imaged by a scanner. Cell density was quantified using ImageJ software and the intensity mean was calculated for each well. PD901 and trametinib (GSK11202212) were purchased from Selleck Chemicals LLC (Houston, TX).

RNA-Seq

RNA capture was performed with TruSeq RNA Library Prep Kit v2 (Illumina) and sequenced on a HiSeq4000. RNA-Seq data were aligned to the GRCh38 reference genome using Star aligner (Dobin et al, 2013). RSEM (Li and Dewey, 2011) was used to quantify gene and transcript expression levels. Gene differential expression analysis was performed using DESeq2 (Love et al, 2014). GSEA (Mootha et al, 2003, Subramanian et al, 2005) was used to determine enriched metabolic pathways in the MSigDB Gene Ontology biological process gene set collection (The Gene Ontology, 2019). Heatmaps were created using the ComplexHeatmap package (v1.20.0) (Gu et al, 2016). Statistical analyses were performed in R v3.5.1.

Reduced to oxidized glutathione ratio functional assay

Cells (3 x 105) were plated per well of 6-well plates in complete medium and treated with siRNA interference for 24 hours followed by PD901/DMSO for 48 hours. Lysates were collected and analyzed using BioVision, Inc. (Milpitas, CA) Glutathione (GSH/GSSG/Total) Fluorometric Assay Kit (Cat. No. K264).

Supplementary Material

1

ACKNOWLEGMENTS

We would like thank members of the Aplin lab for their help in offering suggestions, guidance, and technical expertise. The Sidney Kimmel Cancer Center laboratory animal core facility is supported by NIH/NCI (P30 CA056036). The RPPA studies were performed at the Functional Proteomics Core Facility at The University of Texas MD Anderson Cancer Center, which is supported by NCI Cancer Center Support Grant (CA16672). This work was supported by grants from National Institutes of Health (NIH) R01 CA196278, R01 CA182635, and the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation to A.E. Aplin.

CONFLICT OF INTEREST

Dr. Mike Davies has received grants from GlaxoSmithKline, AstraZeneca, Roche/Genentech, Oncothyreon, and Sanofi-Aventis and serves as a consultant on an advisory board for Array, Novartis, GlaxoSmithKline, Roche/Genentech, and Sanofi-Aventis and Vaccinex. Prof. Reinhard Dummer has intermittent, project focused consulting and/or advisory relationships with Novartis, Merck Sharp & Dhome (MSD), Bristol-Myers Squibb (BMS), Roche, Amgen, Takeda, Pierre Fabre, Sun Pharma, Sanofi, Catalym, Second Genome outside the submitted work. Dr. Andrew Aplin has received a grant from Pfizer Inc. and has ownership interest in patent number 9880150. No potential conflicts of interest were disclosed by the other authors.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Data and materials availability: All data are available in the main text or the supplementary materials.

REFERENCES

  1. Ascierto PA, Schadendorf D, Berking C, Agarwala SS, van Herpen CM, Queirolo P, et al. MEK162 for patients with advanced melanoma harbouring NRAS or Val600 BRAF mutations: a non-randomised, open-label phase 2 study. Lancet Oncol 2013;14:249–56. [DOI] [PubMed] [Google Scholar]
  2. Bonnet S, Archer SL, Allalunis-Turner J, Haromy A, Beaulieu C, Thompson R, et al. A mitochondria-K+ channel axis is suppressed in cancer and its normalization promotes apoptosis and inhibits cancer growth. Cancer Cell 2007;11(1):37–51. [DOI] [PubMed] [Google Scholar]
  3. DeNicola GM, Chen PH, Mullarky E, Sudderth JA, Hu Z, Wu D, et al. NRF2 regulates serine biosynthesis in non-small cell lung cancer. Nat Genet 2015;47(12):1475–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 2013;29(1):15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dong JK, Lei HM, Liang Q, Tang YB, Zhou Y, Wang Y, et al. Overcoming erlotinib resistance in EGFR mutation-positive lung adenocarcinomas through repression of phosphoglycerate dehydrogenase. Theranostics 2018;8(7):1808–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dummer R, Ramelyte E, Schindler S, Thurigen O, Levesque MP, Koelblinger P. MEK inhibition and immune responses in advanced melanoma. Oncoimmunology 2017a;6(8):e1335843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dummer R, Schadendorf D, Ascierto PA, Arance A, Dutriaux C, Di Giacomo AM, et al. Binimetinib versus dacarbazine in patients with advanced NRAS-mutant melanoma (NEMO): a multicentre, open-label, randomised, phase 3 trial. Lancet Oncol 2017b;18(4):435–45. [DOI] [PubMed] [Google Scholar]
  8. Falchook GS, Lewis KD, Infante JR, Gordon MS, Vogelzhang NJ, DeMarini DJ, et al. Activity of the oral MEK inhibitor trametinib in patients with advanced melanoma: a phase 1 dose-escalation trial. Lancet Oncol 2012;13:782–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fantin VR, St-Pierre J, Leder P. Attenuation of LDH-A expression uncovers a link between glycolysis, mitochondrial physiology, and tumor maintenance. Cancer Cell 2006;9(6):425–34. [DOI] [PubMed] [Google Scholar]
  10. Flaherty KT, Infante JR, Daud A, Gonzalez R, Kefford R, Sosman J et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012;367:1694–703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Flaherty KT, Robert C, Hersey P, Nathan P, Garbe C, Milhem M, et al. Improved survival with MEK inhibition in BRAF-mutated melanoma. N Engl J Med 2012;367:568–73. [DOI] [PubMed] [Google Scholar]
  12. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 2016;32(18):2847–9. [DOI] [PubMed] [Google Scholar]
  13. Gupta A, Love S, Schuh A, Shanyinde M, Larkin JM, Plummer R, et al. DOC-MEK: a double blind randomized phase 2 of docetaxel with or without selumetinib in wild-type BRAF advanced melanoma Ann Oncol 2014;25:968–74. [DOI] [PubMed] [Google Scholar]
  14. Haq R, Shoag J, Andreu-Perez P, Yokoyama S, Edelman H, Rowe GC, et al. Oncogenic BRAF regulates oxidative metabolism via PGC1alpha and MITF. Cancer Cell;23(3):302–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Howe HL, Wu X, Ries LA, Cokkinides V, Ahmed F, Jemal A, et al. Annual report to the nation on the status of cancer, 1975–2003, featuring cancer among U.S. Hispanic/Latino populations. Cancer 2006;107:1711–42. [DOI] [PubMed] [Google Scholar]
  16. Hsu PP, Sabatini DM. Cancer cell metabolism: Warburg and beyond. Cell 2008;134(5):703–7. [DOI] [PubMed] [Google Scholar]
  17. Locasale JW. Serine, glycine and one-arbon units: cancer metabolism in full circle. Nat Rev Cancer 2013;13:572–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kakadia S, Yarlagadda N, Awad R, Kundranda M, Niu J, Naraev B, et al. Mechanisms of resistance to BRAF and MEK inhibitors and clinical update of US Food and Drug Administration-approved targeted therapy in advanced melanoma. Onco Targets Ther 2018;11:7095–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kumar MP, Du J, Lagoudas G, Jiao Y, Sawyer A, Drummond DC, et al. Analysis of Single-Cell RNA-Seq Identifies Cell-Cell Communication Associated with Tumor Characteristics. Cell Rep 2018;25(6):1458–68 e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Li B, Dewey CN. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 2011;12:323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu J, Guo S, Li Q, Yang L, Xia Z, Zhang L, et al. Phosphoglycerate dehydrogenase induces glioma cells proliferation and invasion by stabilizing forkhead box M1 J Neurooncol 2013;111:245–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Locasale JW. Serine, glycine and one-carbon units: cancer metabolism in full circle. Nat Rev Cancer;13(8):572–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Locasale JW, Grassian AR, Melman T, Lyssiotis CA, Mattaini KR, Bass AJ et al. Phosphoglycerate dehyrogenase diverts glycolytic flux and contributes to oncogenesis. Nat Genet 2011;43:869–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014;15(12):550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Martinez-Garcia M, Banerji U, Albanell J, Bahleda R, Dolly S, Kraeber-Bodere F, et al. First-in-human, phase I dose-escalation study of the safety, pharmacokinetics, and pharmacodynamics of RO5126766, a first-in-class dual MEK/RAF inhibitor in patients with solid tumor. Clin Cancer Res 2012;18:4806–19. [DOI] [PubMed] [Google Scholar]
  26. Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J, et al. PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 2003;34(3):267–73. [DOI] [PubMed] [Google Scholar]
  27. Morris EJ, Jha S, Restaino CR, Dayananth P, Zhu H, Cooper A, et al. Discovery of a novel ERK inhibitor with activity in models of acquired resistance to BRAF and MEK inhibitors. Cancer Discov 2013;3:742–50. [DOI] [PubMed] [Google Scholar]
  28. Ou Y, Wang SJ, Jiang L, Zheng B, Gu W. p53 protein-mediated regulation of phosphoglycerate dehydrogenase (PHGDH) is crucial for the apoptotic response upon serine starvation. J Biol Chem 2015;290:457–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pacold ME, Brimacombe KR, Chan SH, Rohde JM, Lewis CA, Swier LJ et al. A PHGDH inhibitor reveals coordination of serine synthesis and one-carbon unit fate. Nat Chem Biol 2016;12:452–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pavlova NN, Thompson CB. The emerging hallmarks of cancer metabolism. Cell Metab 2016;23(1):27–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Posch C, Moslehi H, Feeney L, Green GA, Ebaee A, Feichtenschlager V et al. Combined targeting of MEK and PI3K/mTOR effector pathways is necessary to effectively inhibit NRAS mutant melanoma in vitro and in vivo. Proc Natl Acad Sci U S A 2013;110:4015–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Possemato R, Marks KM, Shaul YD, Pacold ME, Kim D, Birsoy K, et al. Functional genomics reveal that the serine synthesis pathway is essential in breast cancer. Nature 2011;476:346–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rebecca VW, Alicea GM, Paraiso KH, Lawrence H, Gibney GT, Smalley KS. Vertical inhibition of the MAPK pathway enhances therapeutic responses in NRAS-mutant melanoma. Pigment Cell Melanoma es 2014;27:1154–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Roberts PJ, Usary JE, Darr DB, Dillon PM, Pfefferle AD, Whittle MC et al. Combined PI3K/mTOR and MEK inhibition provides broad antitumor activity in faithful murine cancer models. Clin Cancer Res 2012;18:5290–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ross KC, Andrews AJ, Marion CD, Yen TJ, Bhattacharjee V. Identification of the Serine Biosynthesis Pathway as a Critical Component of BRAF Inhibitor Resistance of Melanoma, Pancreatic, and Non-Small Cell Lung Cancer Cells. Mol Cancer Ther 2017;16(8):1596–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Schicher N, Paulitschke V, Swoboda A, Kunstfeld R, Loewe R, Pilarski P. Erlotinib and bevacizumab have synergistic activity against melanoma. Clin Cancer Res 2009;15:3495–502. [DOI] [PubMed] [Google Scholar]
  37. Selvarajah B, Azuelos I, Plate M, Guillotin D, Forty EJ, Contento G et al. mTORC1 amplifies the ATF4-dependent de novo serine-glycine pathway to supply glycine during TGF-beta1-induced collagen biosynthesis. Sci Signal 2019;12(582). [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102(43):15545–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Teh JL, Purwin TJ, Greenawalt EJ, Chervoneva I, Goldberg A, Davies MA et al. An In Vivo Reporter to Quantitatively and Temporally Analyze the Effects of CDK4/6 Inhibitor-Based Therapies in Melanoma. Cancer Res;76(18):5455–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. The Gene Ontology C The Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res 2019;47(D1):D330–D8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Thiaville MM, Pan YX, Gjymishka A, Zhong C, Kaufman RJ, Kilberg MS. MEK signaling is required for phosphorylation of eIF2alpha following amino acid limitation of HepG2 human hepatoma cells. J Biol Chem 2008;283(16):10848–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tibes R, Qiu Y, Lu Y, Hennessy B, Andreeff M, Mills GB, et al. Reverse phase protein array: validation of a novel proteomic technology and utility for analysis of primary leukemia specimens and hematopoietic stem cells. Mol Cancer Ther 2006;5(10):2512–21. [DOI] [PubMed] [Google Scholar]
  43. Vander Heidin MG, Cantley LC, Thompson CB. Understanding the Warburg effect: The metabolic requirements of cell proliferation. Science 2009;324:1029–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Vogel CJ, Smit MA, Maddalo G, Possik PA, Sparidans RW, van der Burg SH, et al. Cooperative induction of apoptosis in NRAS mutant melanoma by inhibition of MEK and ROCK. Pigment Cell Melanoma es 2015;28:307–17. [DOI] [PubMed] [Google Scholar]
  45. Vu HL, Aplin AE. Targeting TBK1 inhibits migration and resistance to MEK inhibitors in mutant NRAS melanoma. Mol Cancer Res 2014;12:1509–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Warburg O On the origin of cancer cells. Science 1956;123(3191):309–14. [DOI] [PubMed] [Google Scholar]
  47. Yang M, Vousden KH. Serine and one-carbon metabolism in cancer. Nat Rev Cancer;16(10):650–62. [DOI] [PubMed] [Google Scholar]
  48. Yun J, Rago C, Cheong I, Pagliarini R, Angenendt P, Rajagopalan H, et al. Glucose deprivation contributes to the development of KRAS pathway mutations in tumor cells. Science 2009;325(5947):1555–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zimmer L, Barlesi F, Martinez-Garcia M, Dieras V, Schellens JH, Spano JP, et al. . Phase I expansion and pharmacodynamic study of the oral MEK inhibitor RO4987655 (CH4987655) in selected patients with advanced cancer with RAS-RAF mutations. Clin Cancer Res 2014;20:4251–61. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS1591998-supplement-1.pdf (373.9KB, pdf)

RESOURCES