Separable cell cycle arrest and immune response elicited through pharmacological CDK4/6 and MEK inhibition in RAS^mut disease models

Abstract

The combination of CDK4/6 and MEK inhibition as a therapeutic strategy has shown promise in various cancer models, particularly in those harboring RAS mutations. An initial high-throughput drug screen identified a high synergy between the CDK4/6 inhibitor palbociclib and the MEK inhibitor trametinib when used in combination in soft tissue sarcomas. In RAS mutant models, combination treatment with palbociclib and trametinib induced significant G1 cell cycle arrest, resulting in a marked reduction in cell proliferation and growth. CRISPR-mediated RB1 depletion resulted in a decreased response to CDK4/6 and MEK inhibition, which was validated in both cell culture and xenograft models. Beyond its cell cycle inhibitory effects, pathway enrichment analysis revealed the robust activation of interferon pathways upon CDK4/6 and MEK inhibition. This induction of gene expression was associated with the upregulation of retroviral elements. The TBK1(TANK-binding kinase 1) inhibitor GSK8612 selectively blocked the induction of interferon-related genes induced by palbociclib and trametinib treatment, and highlighted the separable epigenetic responses elicited by combined CDK4/6 and MEK inhibition. Together, these findings provide key mechanistic insights into the therapeutic potential of CDK4/6 and MEK inhibition in soft tissue sarcoma.

Introduction

Normal cell cycle regulation involves a delicate balance between proliferative and antiproliferative mechanisms. This balance is disrupted in cancer, leading to uncontrolled cell division driven by excessive mitogenic signaling and failure of cell cycle inhibitory checkpoints. Over the past ten years, cyclin-dependent kinases (CDKs) have emerged as promising druggable targets that can effectively restrict uncontrolled cell cycle progression; however, challenges persist [1–5]. The success of CDK4/6 inhibitors has spurred extensive research into their effectiveness in various cancer types [6–9]. However, inhibiting individual CDKs often proves to be ineffective and not curative because of the redundancy and compensatory capacity of various cyclins and CDKs, in addition to different cell type-specific dependencies and de novo or acquired resistance to CDK inhibition in certain cases [10–14]. A critical determinant of the response to CDK4/6 inhibitors is the presence of the RB tumor suppressor, which is infrequently lost in HR+/HER2- breast cancer, wherein these agents are the standard of care [15].

RAS-driven tumors, frequently observed in pancreatic ductal adenocarcinoma (PDAC), lung cancer, melanoma and soft-tissue sarcomas, have emerged as potential targets for CDK4/6 inhibitors due to the presence of RB loss or RAS mutations. Although preclinical studies have shown some success in using CDK4/6 inhibitors in various cancers beyond their FDA-approved application in hormone receptor-positive, HER2-negative breast cancer, such as PDAC [16], lung cancer [17], and sarcoma [18], clinical trials have consistently demonstrated the ineffectiveness of CDK4/6 inhibitors when used as monotherapy [19, 20]. Several clinical trials (NCT04162301, NCT02255461, NCT02607124, and NCT01747876) were terminated due to the ineffectiveness of CDK4/6 inhibitor monotherapy. Considering the crucial role of the mitogen-activated protein kinase (MAPK) and extracellular signal-regulated kinase (ERK) pathways in modulating tumor progression and drug resistance [21], MAPK/ERK kinase (MEK) inhibitors like binimetinib and trametinib have been developed and tested in approximately 300 phase I and phase II clinical studies. However, the clinical efficacy of MEK inhibitors as single agents in KRAS and NRAS-mutated tumors has been disappointing [22–26]. Several clinical trials (NCT01849874, NCT00174369, NCT01453387, NCT01278615, NCT00147550, NCT00957580, and NCT01328106) have been withdrawn or terminated due to unsatisfactory results. Additionally, a recent study concluded that, at the current time, there is no evidence for effectiveness of MEK inhibitors as monotherapy [27]. These underwhelming clinical results with CDK4/6 inhibitors and MEK inhibitors as monotherapies have driven researchers to explore combination therapy strategies. Combining CDK4/6 inhibitors with MEK inhibitors, DNA-damaging chemotherapeutic agents, or mTOR inhibitors has shown substantial improvements in effectiveness across various cancer models, including PDAC, lung cancer, breast cancer, and soft-tissue sarcomas [28–37]. Despite the promising synergistic effects observed in preclinical models, translating these findings into clinical practice while managing toxicity remains a significant challenge [38, 39].

The most extensively studied outcome of CDK4/6 inhibitor-induced suppression of E2F transcriptional activity is RB-dependent inhibition of proliferation [40]. Notably, beyond S-phase entry, E2F targets also govern replication origin licensing, DNA repair, DNA methylation, chromatin condensation, cellular metabolism, cellular differentiation, and apoptosis [14]. Beyond cell cycle inhibition, upregulated immunological gene expression signatures, including interferon (IFN) pathways, are commonly observed in response to CDK4/6 inhibitor-based combination treatment strategies [28, 34, 41, 42]. However, the precise mechanisms governing the activation of IFN response genes have yet to be fully elucidated. One encouraging finding suggests that CDK4/6 inhibition may trigger IFN responses and antitumor immunity by inducing tumor-intrinsic retroviral genes [40, 41].

Here, we explored the strategy of combining CDK4/6 and MEK inhibition in several RAS-mutated cancer models. The rationale for the synergistic potential of combining the CDK4/6 inhibitor palbociclib with the MEK inhibitor trametinib was discovered using an unbiased high-throughput drug screening assay. Using in vitro isogenic and in vivo xenograft models, our data demonstrated that combinatorial CDK4/6 and MEK inhibition induced robust cell cycle arrest and IFN-related genes. Additionally, our study revealed the indispensable role of RB1 for complete response as well as consistency of response between fibrosarcoma models and previous findings in lung and pancreatic carcinomas [28, 34].

Materials and methods

Cell lines, culture, and therapeutic agents

MIA PaCa-2 human pancreatic cancer cells (CRL1420) and HT-1080 human fibrosarcoma cells (CCL-121) [43] were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Both cell lines were authenticated by short tandem repeat (STR) DNA profiling analysis and mycoplasma testing was negative and performed on 8/31/2022. The isogenic pancreatic cancer cell line MIA PaCa-2-sgCtrl/sgRB was cultured in Dulbecco’s Modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 50 μg/mL penicillin/streptomycin. PDAC PDX cell lines 519 and 3226 were generated by our lab and maintained in Keratinocyte SFM medium as previously described but were not tested for mycoplasma [34]. HT1080 cells were stably infected with ISRE-mCherry lentivirus, as previously described [43]. Transduced HT1080 cells were cultured in DMEM supplemented with 10% FBS and 50 μg/ml penicillin/streptomycin. All above cell lines were selected from passage six to seven and were maintained at 37°C and 5% CO₂. Palbociclib, trametinib, and pimasertib were purchased from MedChemExpress (Monmouth Junction, NJ, USA) and were dissolved in DMSO to a final stock concentration of 10 mM. Additionally, a customized drug library comprising 311 compounds was obtained from SelleckChem (Houston, TX, USA).

Plasmids and infection procedures

Lentiviral vectors, including pL-CRISPR-sgCtrl (GACCGGAACGATCTCGCGTAG)-GFP and pL-CRISPR-sgRB (GGTTCTTTGAGCAACATGGG)-GFP, were sourced from Dr. Rod Bremner at the Lunenfeld Tanenbaum Research Institute (Toronto, ON, CAN). The pLenti0.3UbCGWH2BC1-PatGFP vector was obtained from Dr. Ethan Abel (RPCCC, Buffalo, NY, USA). Lentiviral infections were carried out in exponentially growing HT1080 or MIA PaCa-2 cells in the presence of 4 μg/mL polybrene (Sigma-Aldrich, St. Louis, MO, USA). Isogenic HT1080-RB-del and MIA PaCa-2-RB-del cell lines were generated through CRISPR-mediated deletion with guide sequences designed to target exon 7 of the human RB1 gene.

Drug screen

Cytation 5 (Agilent BioTek, Santa Clara, CA, USA) and IncuCyte (Sartorius, Ann Arbor, MI, USA) live cell imaging systems were used for drug screening in the HT1080 cell line. Cells were initially seeded in 384-well plates and subjected to a 24-hour pretreatment with either DMSO or 100 nM palbociclib. Subsequently, the compounds from the drug library were introduced into the plates at a dose of 100 nM. The growth of cells in each well was continuously monitored, and the growth rates were calculated over a period of 5 days. To evaluate drug efficacy, the growth rate of each compound in the drug library was normalized to the mean of DMSO-treated wells within the same drug plate.

Immunoblot analysis

Whole-cell lysates were extracted from the indicated cell lines using RIPA lysis buffer (10 mM Tris HCl pH 8.0, 1 mM EDTA, 150 mM NaCl, 1% Triton-X-100, 0.1% sodium deoxycholate, and 0.1% sodium dodecyl sulfate [SDS]) supplemented with Halt^™ Protease Inhibitor Cocktail (Thermo Fisher, Carlsbad, CA, USA). Subsequently, 20 μg of the resultant proteins were separated on a 10% SDS-polyacrylamide gel electrophoresis (PAGE) gel and then transferred to a nitrocellulose membrane for immunoblotting. Immunoblotting was performed by incubating the membranes with protein-specific primary antibodies overnight at 4°C, followed by incubation with horseradish peroxidase-conjugated anti-mouse, anti-goat, or anti-rabbit secondary antibodies for 1 h at RT. Immunoreactive bands were detected using an enhanced chemiluminescent substrate (National Diagnostics, CL-300, Atlanta, GA, USA). Primary antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA), including pRB S807/811 (8516S), pERK (T202/Y204) (4377S), ERK (4695S), pTBK1(S172) (5483S), TBK1 (3504S), EZH2 (5246S), RB (9309 L), and Cyclin B1 (4138S). GAPDH (SC-47724) antibody was procured from Santa Cruz Biotech (Dallas, TX, USA), and Cyclin A2 (AF5999) was purchased from R&D Systems (Minneapolis, MN, USA). Goat-anti-rabbit-HRP (31430), rabbit-anti-goat-HRP (31402), and goat-anti-rabbit-HRP (31460) were used as secondary antibodies (Thermo Fisher, Carlsbad, CA, USA). All primary and secondary antibodies were used at 1:1000 and 1:2000 dilutions, respectively.

Immunofluorescence analysis

Cells were seeded on glass coverslips and exposed to the experimental drugs for 48 h. After treatment, the cells were washed twice with cold phosphate-buffered saline (PBS), fixed, and permeabilized with cold methanol at −20°C for 10 min. The fixed cells were then blocked using IF buffer (1X PBS, 5% BSA, 0.4% NP40) and incubated with the primary antibody RB (9309 L; 1;50; Cell Signaling Technology, Danvers, MA, USA) at RT for 1 h. Following the primary antibody incubation, the coverslips were washed in PBS and then incubated with secondary antibody (1:200) in the presence of DAPI (1:1000) for nuclear staining. After incubation with the secondary antibody, the coverslips were washed again with PBS and mounted on glass slides. Fluorescence microscopy images were captured using an EVOS microscope (Thermo Fisher, Carlsbad, CA, USA) at 40X magnification.

Cell proliferation assay

Cell proliferation in response to various treatments was determined using a chemiluminescent BrdU ELISA kit (11669915001; Roche, Indianapolis, IN, USA), following the manufacturer’s instructions. Luminescence readings were obtained using a Biotek Synergy 2 plate reader (Agilent Technologies, Santa Clara, CA, USA).

Flow cytometry cell cycle analysis

To determine the cell cycle profile based on DNA content, cells were trypsinized, fixed in ice-cold 70% ethanol overnight at −20°C, and subsequently washed with PBS. The fixed cells were pelleted and washed with PBS. Prior to analysis using a BD LSR FORTESSA flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA), the cells were stained with propidium iodide (PI) (40 μg/mL) in the presence of RNase A (200 μg/mL) [15].

qRT-PCR analysis

Total RNA was extracted from cells after 48 h of culture or tumor tissues using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA) and subjected to reverse transcription into cDNA using SuperScript^™ VILO^™ Master Mix (11755050; Thermo Fisher, Carlsbad, CA, USA). Primers used for qPCR are listed in Supplementary Table S1. The primers were diluted in nuclease-free water with PowerUp^™ SYBR^™ Green Master Mix (A25779; Thermo Fisher, Carlsbad, CA, USA), and qPCR assays were performed using the QuantStudio 6 Pro Real-Time PCR System (Thermo Fisher, Carlsbad, CA, USA). Relative RNA levels were quantified using the 2ˆ(−ΔΔCT) method and normalized to the reference gene, GAPDH.

Gene expression analysis from public datasets

RNA-seq raw fragments per kilobase of transcript per million reads mapped (FPKM) data for lung cancer, PDAC, and fibrosarcoma models (accession numbers GSE110397 and GSE180265) were downloaded from the Gene Expression Omnibus (GEO). The fastq files were processed using the nf-core/rnaseq pipeline (version 3.3) [44], with GRCh38 serving as the reference genome. Computational analysis was conducted using a high-performance computing cluster hosted at the Center for Computational Research, University at Buffalo. Raw read counts were generated using the RSEM package [45]. These raw read counts were then utilized as inputs for differential gene expression analysis using DESeq2 [46]. A cutoff of Ilog2FoldChangeI ≥ 2 and an adjusted p-value < 0.05 were applied to define and identify differentially expressed genes.

TEtranscripts analysis:

BAM files generated for gene expression analysis were used as inputs for endogenous retroviral sequence/transposable element (TE) identification and differential expression analysis. The TEtranscripts python package [47] developed by Hammell Lab (https://github.com/mhammell-laboratory/TEtranscripts) was used for this purpose. We used a singularity container (pulled with this command “singularity pull tetranscripts.sif docker://mhammelllab/tetranscripts:latest” on October 26, 2023) to perform data processing and analysis on the high-performance computing cluster hosted at the Center for Computational Research, University at Buffalo. The input bam files were aligned using STAR2 in the nf-core/rnaseq pipeline. Two annotated gtf files were used. The gtf for gene annotation (gencode.v44.annotation.gtf) was downloaded from the gencode website (https://www.gencodegenes.org/human). The TE specific gtf (GRCh38_GENCODE_rmsk_TE.gtf) was downloaded on October 26, 2023, from https://www.dropbox.com/sh/1ppg2e0fbc64bqw/AACUXf-TA1rnBIjvykMH2Lcia?dl=0. For each cell line treatment and control comparison, the *_sigdiff_gene-TE.txt files contained filtered, differentially expressed TEs that had passed a predefined threshold (Ilog2FoldChangeI > 0 and adjusted p-value < 0.05).

Transcriptome analysis

RNA was extracted using the Qiagen RNeasyplus kit (Qiagen, Germantown, MD, USA), with subsequent quality assessment using the RNA6000 Nano assay and Agilent 2200 TapeStation (Agilent Technologies, Santa Clara, CA, USA) to ensure a minimum RNA Integrity Number (RIN) of 7.0 for inclusion in the study. cDNA synthesis employed random hexamers for the full-length, strand-specific representation of non-ribosomal RNA transcripts. Targeted RNA sequencing libraries were prepared using the DriverMap Human Genome-Wide Gene Expression Profiling Sample Prep Kit hDM18Kv3 (Cellecta, Mountain View, CA, USA), utilizing predesigned multiplex PCR primer sets for known protein-coding genes. The anchor PCR step mitigated primer dimer formation. Purification and quantification of PCR products was conducted using SPRI and Qubit (Thermo Fisher, Carlsbad, CA, USA) fluorescence assays, respectively. RNA-seq libraries were analyzed on an Illumina NextSeq 500 sequencer (Illumina, San Diego, CA, USA). Alignment, gene-level read counts, and transcript abundance estimates were obtained using STAR [48], Salmon [49], and EdgeR [50], respectively. Data normalization, systematic bias correction, and identification of differentially expressed genes were performed using the Bioconductor package EdgeR. Genes with an average read count of less than ten read counts across all samples were excluded from further analysis. Following data integration, systematic bias was corrected using ComBat, as described previously [51]. Differentially expressed gene (DEG) analysis was performed with EdgeR and selected based on a log2 fold change > 1 and a p-value < 0.05.

Knockdown experiments

HT1080 and 519 cells underwent reverse transfection using Dharmacon Human ON-TARGETplus Human siRNA: TBK1 (J-003788–08/09/10/11) and non-targeting siRNA (D-001810-10-05). Transfection was conducted using Lipofectamine RNAiMax Transfection Reagent (13778150; Invitrogen, Carlsbad, CA, USA), following the manufacturer’s protocol. After 24 h of transfection, the cells were exposed to different drugs for 48 h. Parallel experiments were performed using immunoblot analysis to confirm gene silencing.

Mice and xenografts

NOD scid gamma (NSG) mice were ethically and responsibly maintained at animal care facilities at RPCCC. All aspects of animal care, drug treatment, and sacrifice were approved by the Institutional Animal Care and Use Committee (IACUC) in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Male NSG mice, aged 8–10 weeks, were subcutaneously implanted with early passage HT1080-sgCtrl/sgRB cells (1.5 × 10⁶ cells/mouse). Once the tumor volume reached 200–250 mm³, the mice were subjected to treatment by gastric gavage with either vehicle or a combination of palbociclib and trametinib. Palbociclib (PD-0332991, 100 mg/kg) was diluted in 50 mM lactate buffer at pH 4.0, whereas trametinib (0.5 mg/kg) was diluted in 0.5% hydroxypropyl cellulose and 0.2% Tween 80. The tumor size was meticulously monitored by daily measurements using digital calipers. The tumor volume was computed using the following formula: (greatest diameter × (shortest diameter²))/2. The study was concluded when tumors reached a volume of 2000 mm³ or when mice were humanely sacrificed at the termination of the treatment period. Any mice displaying signs of illness or discovered deceased during treatment were systematically excluded from the analysis.

Immunohistochemical analysis

Hematoxylin and eosin (H&E) staining and immunohistochemistry for pRB S807/811 (8516S-1:400) and Ki67 (RM-9106S1-1:200) were performed on tumor tissues following formalin fixation, using standard procedures [34]. The staining process was conducted using a Leica auto-stainer and subsequently imaged using Aperio (Leica Biosystems, Deer Park, IL, USA).

Statistical analysis

GraphPad Prism 9 software was used for all the statistical analyses. Student’s t-test, One-way and Two-way ANOVA with multiple comparisons were used, as indicated in each figure where appropriate. Statistical significance was defined as a p-value of < 0.05.

Data availability

The data generated in this study are available from the corresponding author upon reasonable request. Publicly available data analyzed in this study was obtained from Gene Expression Omnibus (GEO) at GSE110397 and GSE180265.

Results

CDK4/6 and MEK inhibition in RAS^mut models elicits potent cell cycle inhibition

Because of the known resistance of RAS^mut models to CDK4/6 inhibitor monotherapy [32, 35] and the particular sensitivity of RAS^mut preclinical models to MEK inhibitors [52, 53], several compounds were assessed for cooperativity in suppressing proliferation of the HT1080 fibrosarcoma cell line in culture. Previous studies have demonstrated the efficacy of MEK inhibitors in combination with CDK4/6 inhibitors in other cancer cell types [28, 31, 34]. Thus, multiple MEK inhibitors were screened in combination with palbociclib to assess synergistic inhibition of HT1080 cell growth in vitro (Fig. 1A, Supplementary Fig. S1). Trametinib (GSK1120212) was identified as one of the top hits in the live-cell drug screen and displayed a lower IC₅₀ than the other top hits and was thus selected for further validation (Supplementary Fig. S1, Supplementary Table S2). Trametinib, used in combination with palbociclib, reduced HT1080 cell proliferation (Fig. 1B). Using a BrdU incorporation assay, it was revealed that the cooperation between palbociclib and trametinib treatment of HT1080 cells was through cell cycle inhibition (Fig. 1C). Dose-response and Bliss analyses were conducted to further evaluate synergistic drug interactions, revealing significant drug synergy with a Bliss synergy score of 15.695 for palbociclib and trametinib (Fig. 1D). Additionally, validation was extended to the pancreatic cancer cell line MIA PaCa-2, where a similar reduction in cell proliferation was observed with combined treatment with palbociclib and trametinib (Fig. 1E). Furthermore, analyses illustrated the suppression of the cell cycle (Fig. 1F) and synergy (Fig. 1G). An additional MEK inhibitor identified from the original HT1080 growth inhibition screen, pimasertib, was also assessed for its efficacy in combination with palbociclib. While synergistic inhibition was observed to be greater than trametinib in the MIA PaCa-2 cell line at a higher dose (Bliss synergy score of 25.771, Supplementary Fig. S1), pimasertib underperformed trametinib in the HT1080 cell line, showing a smaller effect in the BrdU incorporation assay and a bliss synergy score of 6.201 (Supplementary Fig. S1). Thus, trametinib was used in subsequent experiments.

Figure 1. CDK4/6 and MEK inhibition in RAS^mut models elicits potent cell cycle inhibition.

A. Heat map representing the relative growth rate of HT1080 cells in response to MEK inhibitors from a drug library at a concentration of 100 nM. B. Live-cell imaging tracking HT1080 cell growth when treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination at indicated concentrations for 96 h. C. BrdU incorporation was assessed in HT1080 cells treated with CDK4/6i (Palbociclib, 100 nM) or MEKi (Trametinib, 25 nM) alone, or in combination at indicated concentrations for 72 h. D. Heatmap illustrating relative BrdU incorporation after 72 h of treatment. Synergy was determined using the BLISS method. E. Live-cell imaging tracking MIA PaCa-2 cell growth when treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 12.5 nM), or CDK4/6i and MEKi in combination at indicated concentrations for 96 h. F. BrdU incorporation in MIA PaCa-2 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 12.5 nM), or CDK4/6i and MEKi in combination for 72 h. G. Heatmap illustrating relative BrdU incorporation after 72 h of treatment. Synergy was determined using the BLISS method. H. qPCR analysis of E2F target genes (CCNA2 and EZH2) in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 48 h. I. Immunoblot analysis of cell cycle proteins in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 48 h. J. Representative propidium iodide-flow cytometry analysis of HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 48 h. Error bars represent standard deviation (SD) from triplicates. *p < 0.05, **p < 0.01, ***p < 0.001, as determined by one-way ANOVA with multiple comparisons.

To further address the impact of palbociclib-trametinib treatment on cell cycle inhibition, several E2F-regulated target genes were assessed. First, CCNA2 and EZH2 expression was quantified in HT1080 cells treated with DMSO, palbociclib, trametinib, or a combination of palbociclib-trametinib for 48 h. While palbociclib treatment alone reduced the expression of both CCNA2 and EZH2, the combined palbociclib-trametinib treatment led to further reduced levels of both transcripts (Fig. 1H). Immunoblot analysis validated the qPCR findings; combination treatment with palbociclib-trametinib robustly reduced the expression of cyclin A and EZH2, in addition to cyclin B1 and pRB, whereas trametinib selectively inhibited ERK activity (Fig. 1I). HT1080 cells treated with DMSO, palbociclib, trametinib, or a combination of palbociclib and trametinib were stained with propidium iodide (PI) and subjected to flow cytometry to assess the cell cycle distribution. Treatment with palbociclib, trametinib, or a combination of palbociclib and trametinib resulted in a significant increase in cells in G1 phase, although no significance was determined between combined treatment and either drug alone (Fig. 1J).

RB1 is required for full response to CDK4/6 and MEK inhibition in both cell culture and xenograft fibrosarcoma models

RB1 deficiency is associated with resistance to CDK4/6 inhibitor therapy [14, 54]. CRISPR-mediated RB1 deletion was performed to assess the efficacy of the combination of palbociclib and trametinib in an RB-depleted setting. In both HT1080 and MIA PaCa-2 RB-depleted cell lines, RB expression was blunted, whereas the suppression of cyclin A expression was largely recovered in cells treated with palbociclib and trametinib (Fig. 2A, Supplementary Fig. S2). Cell cycle inhibition and cell proliferation were measured using a BrdU incorporation assay and GFP live-cell imaging assay, respectively. Compared to RB-replete HT1080 and MIA PaCa-2 cells, RB-depleted cells displayed a blunted response to palbociclib and trametinib combination treatment in terms of both BrdU incorporation (Fig. 2B) and overall cell count (Fig. 2C). The role of RB in the combination of palbociclib- and trametinib-treated HT1080 cells was further assessed by performing a clonogenic cell growth assay. Although combination treatment with palbociclib and trametinib significantly reduced the growth of control cells, RB-depleted cells displayed modest colony inhibition (Fig. 2D). RB-deplete and replete HT1080 cells were injected into mice to assess whether the reduced response of RB1-depleted cells to the combination palbociclib-trametinib treatment in vitro was also observed in vivo. Control HT1080-injected mice treated with a combination of palbociclib (100 mg/kg) and trametinib (0.5 mg/kg) displayed significantly reduced relative tumor volume and tumor mass compared to mice treated with vehicle, while mice injected with RB-depleted HT1080 cells displayed a blunted response to combination palbociclib-trametinib treatment, as evidenced by significantly larger relative tumor volume and tumor mass (Fig. 2E, Supplementary Fig. S2). Importantly, the dose of palbociclib used in the present study was established based on its regular use in the clinic as standard-of-care in estrogen receptor-positive breast cancer (125 mg/day palbociclib) in combination with hormone therapy [55]. Additionally, the dose of trametinib used in the present study was established as the minimum effective dose according to our previous studies that identified 0.5–1.5 mg/kg to be effective at restricting tumor growth when used as combination therapy with various agents with minimal observed toxicity, while lower doses were shown to be ineffective [34, 56, 57]. Immunohistochemical staining of excised tumors revealed that Ki67 staining in control HT1080 cell-injected mice treated with a combination of palbociclib and trametinib was absent, whereas RB-depleted HT1080 cell-injected mice treated with a combination of palbociclib and trametinib displayed robust Ki67 staining (Fig. 2F).

Figure 2. RB1 is required for full response to CDK4/6 and MEK inhibition in both cell culture and xenograft fibrosarcoma models.

A. Immunoblot analysis for E2F target proteins in HT1080 and MIA PaCa-2-sgCtrl/sgRB cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) in combination for 48 h. B. BrdU incorporation was assessed in HT1080 and MIA PaCa-2-sgCtrl/sgRB cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) in combination at indicated concentrations for 72 h. Data shown as mean ± standard deviation (SD). C. Live-cell imaging tracking HT1080 or MIA PaCa-2-sgCtrl/sgRB cell growth when treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) in combination at indicated concentrations for 96 h. Data displayed as mean ± SD. D. Clonogenic assay in HT1080 or MIA PaCa-2-sgCtrl/sgRB cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) in combination for 14 days. E. HT1080-sgCtrl/sgRB-injected xenograft models treated with vehicle or CDK4/6i (Palbociclib, 100 mg/kg) and MEKi (Trametinib, 0.5 mg/kg) in combination once tumor volume reached 200 mm³. Relative tumor volume was determined every 24 h. Data shown as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 as determined by two-way ANOVA. F. HT1080-sgCtrl/sgRB xenograft tumors from vehicle and CDK4/6i (Palbociclib, 100 mg/kg) and MEKi (Trametinib, 0.5 mg/kg)-treated mice stained for pRB (S807/811), Ki67 and by H&E. Representative images are displayed (scale bar = 100 μm).

CDK4/6 and MEK inhibition in fibrosarcoma, PDAC and lung cancer models leads to inhibition of E2F and induction of IFN-like response

To assess the impact of CDK4/6 and MEK inhibition across multiple RAS^mut cancer types (Supplementary Table S3), gene expression analysis was conducted using RNA-seq data collected from publicly available datasets, including HT1080, three pancreatic cancer, and three lung cancer cell models. HT1080 cells were treated with either palbociclib or the MEK inhibitor pimasertib, whereas lung and pancreatic cancer cell models were treated with a combination of palbociclib and trametinib. Both palbociclib and pimasertib treatment of HT1080 cells resulted in significant downregulation of genes within “E2F target” and “G2M checkpoint” pathways, consistent with previous observations of cell cycle inhibition (Fig. 3A–B, Supplementary Fig. S3). Additionally, enrichment of type I IFNs was observed following pimasertib treatment and, to a lesser extent, palbociclib treatment of HT1080 cells (Fig. 3C–D, Supplementary Fig. S3). Type I and II IFNs were associated with upregulated genes in all three pancreatic ductal adenocarcinoma (PDAC) models (MIA PaCa-2, PANC-1, and PaTu-8988T cell lines) (Fig. 3E, Supplementary Fig. S4), and in two of the three lung cancer models assessed (A549 and H2030 but not H460 cell lines) (Fig. 3F, Supplementary Fig. S4) in response to the combined palbociclib and trametinib treatment. IFN signaling in HT1080 cells treated with DMSO, palbociclib alone, trametinib alone, or a combination of palbociclib and trametinib were probed by assessing the expression of STAT2 and IRF9. Both transcripts were significantly increased by trametinib alone and, to a greater extent, in combination with palbociclib (Fig. 3G). HT1080 cells infected with a lentivirus encoding an ISRE-mCherry reporter [43] largely expressed ISREs following trametinib treatment as a single agent or in combination (Fig. 3H). These results validated the gene expression analyses and demonstrated that trametinib substantially drives IFN signaling in HT1080 cells (Fig. 3H).

Figure 3. CDK4/6 and MEK inhibition in fibrosarcoma, PDAC and lung cancer models leads to inhibition of E2F and induction of IFN-like response.

A. ENRICHR analysis identifying downregulated pathways in HT1080 cells treated with CDK4/6i (Palbociclib, 2 μM) or MEKi (Pimasertib, 0.5 μM). B. GSEA from CDK4/6i (Palbociclib, 2 μM) or MEKi (Pimasertib, 0.5 μM)-treated HT1080 cells displaying enrichment of downregulated genes in cell cycle pathways, including E2F_Targets and G2M_checkpoint. C. ENRICHR analysis identifying upregulated pathways in HT1080 cells treated with CDK4/6i (Palbociclib, 2 μM) or MEKi (Pimasertib, 0.5 μM). D. GSEA from CDK4/6i (Palbociclib, 2 μM) or MEKi (Pimasertib, 0.5 μM)-treated HT1080 cells displaying enrichment of upregulated genes in IFN alpha response pathway. E. ENRICHR analysis identifying upregulated pathways in response to CDK4/6i (Palbociclib, 0.5 μM) and MEKi (Trametinib, 25 nM) treatment in three PDAC models: MIA PaCa-2, PANC-1 and PA-TU 8988T. F. ENRICHR analysis identifying upregulated pathways in response to CDK4/6i (Palbociclib, 0.5 μM) and MEKi (Trametinib, 25 nM) treatment in three lung cancer models: A549 (left), H2030 (middle) and H460 (right). G. qPCR analysis for STAT2 and IRF9 in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 48 h. Data displayed as mean ± SD in triplicate. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 as determined by one-way ANOVA with multiple comparisons. H. ISRE-mCherry IFN reporter assay in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 0 – 120 h. Scale bar = 400 μm.

RB1 deficiency blunts both cell cycle suppression and IFN response to combination palbociclib-trametinib treatment

Based on previous observations that RB1 is required for full response of HT1080 cells to combination palbociclib-trametinib treatment and that the response consisted of both downregulated E2F signaling and upregulated IFN signaling, targets from both pathways were assessed for RB1 dependence. Indeed, HT1080-sgCtrl cells treated with a combination of palbociclib-trametinib displayed significantly downregulated CCNA2, CCNB1, and EZH2 expression, whereas HT1080-sgRB cells displayed elevated expression of all three transcripts in the combined treatment cohort (Fig. 4A). This RB1-dependent effect was also observed in vivo in the HT1080 xenograft model when assessing the expression of the same three cell cycle transcripts in the combined treatment cohort (Fig. 4B). Further qPCR analysis confirmed increased STAT2, IRF9, and HLA-A expression with in vitro HT1080-sgCtrl and in vivo HT1080-sgCtrl-derived xenograft model systems, whereas HT1080-sgRB models displayed blunted induction of all IFN-related transcripts (Fig. 4C–D). RB1-dependent IFN signaling was further assessed using the ISRE-mCherry reporter system as previously described [43]. ISRE induction following palbociclib and trametinib treatment was inhibited in HT1080 cells lacking RB1 (Fig. 4E).

Figure 4. RB1 deficiency blunts cell cycle suppression and IFN response to combination palbociclib-trametinib treatment.

A. qPCR analysis of CCNA2, CCNB1, and EZH2 cell cycle genes in HT1080-sgCtrl/sgRB cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) in combination. B. qPCR analysis of CCNA2, CCNB1, and EZH2 cell cycle genes in HT1080-sgCtrl/sgRB tumor tissues treated with vehicle or CDK4/6i (Palbociclib, 100 mg/kg) and MEKi (Trametinib, 0.5 mg/kg) in combination. C. qPCR analysis of STAT2, IRF9 and HLA-A immune-related genes in HT1080-sgCtrl/sgRB cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) for 48 h. D. qPCR analysis of STAT2, IRF9, and HLA-A immune-related genes in HT1080-sgCtrl/sgRB tumor tissues treated with vehicle or CDK4/6i (Palbociclib, 100 mg/kg) and MEKi (Trametinib, 0.5 mg/kg) in combination. E. ISRE-mCherry IFN reporter assay in HT1080-sgCtrl/sgRB cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM), MEKi (Trametinib, 25 nM), or CDK4/6i and MEKi in combination for 72 h. Scale bar = 100 μm. Data displayed as mean ± SD in triplicate. * p < 0.05, ** p < 0.01, *** p < 0.001 **** p < 0.0001 as determined by two-way ANOVA.

CDK4/6 and MEK inhibition triggers transposable element upregulation in lung and PDAC models

Detection of IFN-like responses to palbociclib-trametinib treatment supported further investigation of nucleotide sensing as a common stimulant of IFN pathways [40]. We assessed transposable elements (TE) that might be upregulated following CDK4/6 inhibition and induce immune responses via double-stranded RNA (dsRNA) [41, 58]. Publicly available RNA sequencing data from the PDAC model cell lines 519, PA-TU-8988T, MIA PaCa-2, and 3226, and the lung cancer cell line A549 treated with a combination of palbociclib and trametinib (100 nM and 25 nM, respectively) were analyzed, and differentially expressed TEs are displayed as a volcano plot. Significant and broad TE upregulation resulted from the combined treatment across all models (Fig. 5A, Supplementary Fig. S5) and revealed six shared upregulated TEs (Fig. 5B–D). Relative expression of the TE ERV3–1 transcript gradually increased during the 48 hours following combination palbociclib-trametinib treatment in 519 and 3226 cell lines (Fig. 5E).

Figure 5. CDK4/6 and MEK inhibition triggers transposable element upregulation in lung and PDAC models.

A. Volcano plot displaying differentially expressed transposable elements in combined CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM)-treated compared to DMSO-treated 519 and 3226 cell lines. B. Transposable element analysis using TEtranscripts to analyze the differential expression of transposable elements in different PDAC and lung cancer cell lines treated with CDK4/6i (Palbociclib, 0.5 μM) and MEKi (Trametinib, 25 nM) in combination. C. Venn diagram showing distribution of significantly upregulated transposable elements in different lung and PDAC cell models treated with CDK4/6i (Palbociclib, 0.5 μM) and MEKi (Trametinib, 25 nM) in combination. D. Heatmap displaying log fold change of the 6 conserved upregulated transposable elements from the five indicated cell lines treated with CDK4/6i (Palbociclib, 100 nM for 519 and 3226 cells; 0.5 μM for PA-TU 8988T, MIA PaCa-2 and A549 cells) and MEKi (Trametinib, 25 nM) in combination. E. qPCR analysis of ERV3–1 expression in CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM)-treated 3226 and 519 cells over 48 h post-treatment. Data displayed as mean ± SD in triplicate. * p < 0.05, **** p < 0.0001 as determined by students t test with multiple comparisons. TE, transposable element.

TBK1 inhibitor uncouples the cell cycle inhibitory response from the IFN response

TANK-binding kinase 1 (TBK1) is a kinase downstream of dsRNA sensing mediated by RIG1/MAVS and promotes the signaling of IFN regulatory factors or NF-κB for innate immune activation [59]. We then assessed whether the TBK1 inhibitor GSK8612, in combination with palbociclib-trametinib treatment, reduced cell cycle signaling and inactivated the IFN response. Poly (I:C) treatment was used to introduce dsRNA into the culture and induce IFN signaling as a positive control. GSK8612 effectively suppressed the IFN response in palbociclib-trametinib and poly (I:C)-treated HT1080 cells grown in vitro, as assessed using the ISRE-mCherry reporter system (Fig. 6A), and reduced STAT2 and IRF9 expression (Fig. 6B–C). The specificity of GSK8612 for TBK1 was confirmed using siTBK1-treated HT1080 cells, as repression of STAT2 and IRF9 occurred similarly in palbociclib-trametinib treated HT1080 cells lacking TBK1 (Supplementary Fig. S6). GSK8612 treatment did not alter immune gene expression independent of the type I IFN pathway, further highlighting the specificity of the TBK1 inhibitor (Figure S6C). Biochemical analysis revealed that the combination of palbociclib-trametinib treatment of HT1080 cells repressed RB phosphorylation and cyclin A expression but induced phosphorylation of TBK1 (S172). However, GSK8612 used in conjunction with palbociclib-trametinib treatment inhibited TBK1 activation, as measured by TBK1 S172 phosphorylation, but did not affect its repressive effects on pRB or cyclin A (Fig. 6D). Similar selective effects of either TBK1 inhibitor GSK8612 (Fig. 6E) or siTBK1 (Supplementary Fig. S6) treatment of 519 cells on repressed TBK1 expression were observed without affecting the palbociclib-trametinib treatment-induced repression of cyclin A or pRB. Finally, RNA sequencing was performed on 519 cells treated with the combination of palbociclib-trametinib with or without concurrent GSK8612 treatment. As expected from our biochemical analyses, palbociclib-trametinib treatment alone resulted in significant induction of the IFN response, in addition to the suppression of cell cycle genes. However, the addition of GSK8612 suppressed the palbociclib-trametinib-induced upregulation of IFN-related genes without affecting the downregulation of cell cycle genes (Fig. 6F–H).

Figure 6. TBK1 inhibitor uncouples the cell cycle inhibitory response from the IFN response.

A. ISRE-mCherry reporter assay representative images of HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM), TBK1i (GSK8612, 5 μM) or in combination. Poly (I:C) was used as a positive control for dsRNA-activated IFN response as measured by the ISRE-mCherry reporter system. Scale bar = 400 μm. B. qPCR analysis of STAT2 and IRF9 IFN-related gene expression in HT1080 cells treated with vehicle, Poly (I:C) alone or in combination with TBK1i (GSK8612, 5 μM). C. qPCR analysis of STAT2 and IRF9 IFN-related gene expression in HT1080 cells treated with DMSO or CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM), with and without TBK1i (GSK8612, 5 μM). D. Immunoblot analysis for TBK1, pTBK1, and cell cycle proteins in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM), TBK1i (GSK8612, 5 μM) or in combination. E. qPCR analysis of STAT2 and IRF9 IFN-related gene expression in HT1080 cells treated with DMSO, CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) with and without TBK1i (GSK8612, 5 μM). F. Bubble plot of gene ontology analysis for both CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) and triple-treatment (Palbociclib + Trametinib + GSK8612) groups using GSEA and KEGG database (biological process) of the top 25 downregulated (blue) and upregulated (red) pathways. G. Venn diagram displaying distribution of significantly differentially expressed genes in 519 cells treated with CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) or triple treatment with TBK1i (GSK8612, 5 μM) included. ENRICHR analysis identifying shared enriched pathways downregulated in 519 cells treated with CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) and triple treatment with TBK1i (GSK8612, 5 μM) included or uniquely expressed genes in 519 cells treated with CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM). H. Heatmap of select cell cycle and IFN-related gene expression in HT1080 cells treated with CDK4/6i (Palbociclib, 100 nM) and MEKi (Trametinib, 25 nM) ± TBK1i (GSK8612, 5 μM). Data displayed as mean ± SD in triplicate. **** p < 0.0001 as determined by one-way (B) and two-way ANOVA (C, E).

Discussion

CDKs intricately orchestrate critical cellular processes such as the cell cycle (CDK1/2/3/4/5/6/10), transcription (CDK7/8/9), and RNA splicing (CDK11) [60]. Their dysregulated activation and expression have been extensively documented in association with various tumors, prompting the exploration of CDKs as compelling therapeutic targets. Given the impactful modulation of CDK4/6 inhibitors on cell cycle via the RB-E2F pathway, several clinical trials spanning various tumor types have been conducted [17, 61–63]. Notably, hormone receptor-positive breast cancer cells exhibit increased sensitivity to CDK4/6 inhibition [64]. However, therapeutic challenges have emerged, including the complex biological redundancy inherent in cell cycle regulation, disease heterogeneity, and acquired resistance, impeding broader adoption and limiting the efficacy of CDK inhibitors as monotherapy [13, 14, 65]. To address these challenges, numerous clinical trials have investigated CDK4/6 inhibitor combination therapies using diverse pharmacological agents across various cancer types [32, 35, 36, 66, 67]. Of note is the potent synergy between CDK4/6 inhibitors and the inhibition of upstream kinases within specific tumor types. In the context of PDAC and lung cancer, both characterized by a high prevalence of KRAS mutations, findings from our group and others have demonstrated that the combined inhibition of MEK and CDK4/6 yields significant and therapeutically relevant downregulation of E2F target gene expression and induction of G1 arrest [34]. However, to our knowledge, CDK4/6 and MEK inhibition has yet to be investigated in sarcoma models.

HT-1080 cells, derived from the connective tissue of a patient with fibrosarcoma, served as the sarcoma model for our study. Sarcomas, characterized by over 100 subtypes with unique genomic alterations, frequently exhibit dysregulation of the cyclin D-CDK4/6-RB1 pathway, making them potential candidates for CDK4/6 inhibitor therapy [54]. In our genetic analysis of 20 soft-tissue sarcoma cell lines, we observed mutated NRAS (p.Q61K) in HT1080 cells as well as other NRAS mutations in different sarcoma models (RD, p.Q61H; and MFH-ino, p.Q61H/K) (Supplementary Table S4). Our high-throughput drug screening assay on HT1080 cells revealed promising synergistic inhibition of cell growth when MEK inhibitors were combined with palbociclib. The synergistic effect was validated through growth curves and BrdU proliferation assays in both HT1080 and MIA PaCa-2 models. Trametinib (GSK1120212) consistently demonstrated superior effects in HT1080 cells compared to other MEK inhibitors identified in the initial drug screen, even when used at lower doses. The ability to use trametinib effectively at a lower dose than other MEK inhibitors is significant due to toxicity concerns when combining MEK inhibitors with CDK4/6 inhibitors, which has limited clinical implementation and remains a significant concern. These initial findings were substantiated by several molecular, biochemical, and flow cytometric assays that confirmed the synergistic effect of combined MEK and CDK4/6 inhibition on the downregulation of E2F target gene expression and induction of G1 arrest.

Genomic alterations leading to loss of RB1 function are unsurprisingly associated with both de novo and acquired resistance to CDK4/6 inhibitors in breast cancer [13, 14, 68]. In our experimental models, deletion of RB1 enabled resistance to combined treatment with CDK4/6 and MEK inhibitors through unaffected cyclin A expression and a decreased response in cell growth, colony formation, and BrdU incorporation assays. This resistance phenomenon was also evident in vivo, where the treated RB-depleted group exhibited a larger relative tumor volume and mass that was associated with increased Ki67 staining compared to the control group treated with CDK4/6 and MEK inhibitors. It is worth noting that the combination of CDK4/6 and MEK inhibitors not only reduced phosphorylated RB (pRB) levels but also total RB. This observation aligns with earlier findings indicating that MEK inhibitor monotherapy causes a modest reduction in both pRB and total RB in several PDAC models [67].

Upregulated immunological gene expression signatures in response to CDK4/6 inhibitor-based combination treatment strategies have been observed in various models by our group and others [34, 41, 58]. In this study, we conducted gene expression analyses on RNA-seq data obtained from publicly available datasets of HT1080 cells, three lung cancers, and three pancreatic cancer cell models. The most significantly downregulated pathways were related to cell cycle progression (E2F target and G2-M checkpoint), whereas the most significantly upregulated pathways were associated with IFN signaling. Additionally, we demonstrated that RB1 depletion attenuates cell cycle suppression and IFN response during palbociclib-trametinib combination treatment in vitro and in vivo.

The precise mechanisms underlying IFN response gene activation mediated by CDK4/6 inhibition have yet to be fully elucidated. IFN responses are commonly upregulated upon sensing intracellular DNA or RNA [69]. A recent study demonstrated that DNA damage within tumor cells deficient in CDK4/6 promotes type I IFN expression through cGAS-STING signaling [70]. However, the immunostimulatory interaction of CDK4/6 inhibitors with the STING pathway remains unclear, as evidence elsewhere suggests that palbociclib directly inhibits STING [71]. Other studies have suggested a role for RB1 activation in transcriptionally repressing epigenetic modifiers, such as DNMT1 or EZH2, which suppress endogenous retroviruses and subsequent IFN responses mediated by dsRNA [41, 72]. RB1-mediated repression of epigenetic methylation of retroviral elements may, thus, serve as a mechanism to upregulate immunological responses within tumors [68]. Indeed, we observed the upregulation of several TEs in multiple PDAC and lung cancer models treated with a combination of palbociclib and trametinib. TE ERV dsRNA may be directly sensed in the cytosol or reverse-transcribed into DNA for convergence with the STING pathway [72, 73]. TBK1 is instrumental for transducing signaling cascades upon STING-independent recognition of cytosolic dsRNA and STING-dependent recognition of cytosolic DNA [59, 70–72]. Through pharmacological inhibition and knockdown, we demonstrated that TBK1 enhances type I IFN responses during combined CDK4/6 and MEK inhibition in the HT1080 fibrosarcoma and 519 PDAC models without affecting cell cycle repression. However, effector immune responses activated by upregulated type I IFN response genes, including ISREs and HLA-A remain to be elucidated. Notably, type I IFNs may facilitate T cell priming and proliferation for anti-tumor activity [69]. Our previous study demonstrated that CDK4/6 and MEK inhibition potently upregulate granzyme B-expressing CD8+ T cells within an orthotopic pancreatic tumor model [34]. Further combining CDK4/6 and MEK inhibition with anti-PD-L1 significantly reduced tumor burden via a CD8-dependent mechanism. Utilizing other modes of immune checkpoint inhibition that target CD8+ T cells, such as anti-PD-1, is of future interest for enhancing antitumor CD8+ T cell activity during CDK4/6 and MEK inhibition.

Taken together, this study highlights the efficacy of combined CDK4/6 and MEK inhibition in fibrosarcoma, PDAC, lung, and other RB-proficient, RAS^mut cancers to induce stable cell cycle arrest. Additionally, TBK1 promotes an IFN response during CDK4/6 and MEK inhibition independent of the desired downregulated cell cycle response.