Cell cycle regulation meets tumor immunosuppression

Abstract

Reciprocal interactions between tumor cells and immune cells shape the tumor microenvironment. Recent studies indicate that enhanced cell cycle activity in cancer cells suppresses anti-tumor immunity. Here, we discuss potential mechanisms by which cell cycle programs intrinsic to tumor cells are coupled to immune behavior, with consequences for immunotherapy.

Tumor cell intrinsic features modulate anti-tumor immunity and immunotherapy sensitivity

Immunotherapy has significantly improved clinical care for cancer patients with various types of malignancies. However, most patients remain refractory to immunotherapy as a result of local immunosuppression and lack of T cell infiltration in the tumor microenvironment (TME) [1]. tumor cells themselves play a central role in this phenomenon, either by changing their surroundings through the release of cytokines or by rendering themselves resistant to immune attack, downregulating elements of the antigen presentation machinery or upregulating immunosuppressive molecules [1]. Although additional mechanisms remain to be discovered, studies of tumor immunity have yielded another interesting observation. Across tumor types, immunotherapy-resistant “cold” tumors may exhibit elevated oncogenic signaling, including hyperactivation of MYC, KRAS, MTOR, and WNT [1]. Accumulating evidence has suggested that there is a correlation between “cold” tumors and increased tumor cell cycle progression. This Forum article builds on recent literature to consider the hypothesis that the hyperactivation of cell cycle programs in tumor cells confers protection from immune surveillance. This concept has strong implications for cancer therapy, as it suggests that existing drugs that exert effects on the cell cycle may potentially have the added (and untapped) benefit of sensitizing tumors to immune modulators.

Pharmacological inhibition of the cell cycle can promote anti-tumor immunity

Progression through the cell cycle depends on the coordinated expression and activity of cyclins and cyclin-dependent kinases. Cyclin-dependent kinases 4 and 6 (CDK4/6) represent particularly promising targets for therapeutic intervention, as their function – activating the E2F transcription factor by phosphorylating the Retinoblastoma (Rb) protein – is known to be required for the G1/S transition during the cell cycle. Consequently, researchers have developed potent kinase inhibitors of CDK4/6 with the goal of slowing tumor growth. But once preclinical studies got underway, a secondary unanticipated effect was observed: certain tumors including breast cancer, pancreatic cancer and melanoma, exhibited a change in the tumor immune microenvironment and increased sensitivity to immune checkpoint blockade (ICB) [2–6].

The shifts in immune phenotypes observed after treatment with CDK4/6 inhibitors have been attributed to several mechanisms. An early study [3] found that the CDK4/6 inhibitor abemaciclib suppressed expression of the E2F target DNA methyltransferase 1 (DNMT1) in an MMTV-rtTA/tetO-HER2 transgenic mouse model of breast cancer. This in turn, allowed previously silent endogenous retroviral elements (ERVs) to be expressed in the cancer cells, prompting increases in type III interferon (IFN) and MHC class I, resulting in an improved response to ICB as evidenced by suppressed tumor growth [3]. Subsequent studies involving a variety of tumor types (both syngeneic mouse models and human samples) have reported similar effects following treatment with abemaciclib [4,6]. These findings establish an association between cell cycle inhibition and immunotherapy response – a connection which depends (at least in part) on tumor cell intrinsic regulation of IFN signaling and the antigen presentation machinery.

Two recent papers suggest that additional mechanisms may trigger synergies between cell cycle inhibition and ICB. In one study, combined treatment with a MEK inhibitor (trametinib) and a CDK4/6 inhibitor (palbociclib) caused KRAS-mutant human and mouse lung cancer cells to undergo an irreversible cell cycle arrest (i.e. senescence) based on in vitro cell clonogenic assays [5]. Furthermore, the growth of KRAS-mutant mouse lung tumors treated with this regimen in vivo was controlled, an effect that depended on natural killer (NK) cells recruitment as part of a senescence associated secretory phenotype (SASP) program examined by flow cytometry and staining [5]. When the same regimen was applied to a KRAS-mutant mouse pancreatic cancer model, increased cellular senescence and anti-tumor immunity was again observed, but this time, the SASP prompted cytotoxic CD8 T cell infiltration through vascular remodeling, sensitizing tumors to ICB [2]. It is intriguing that the combination of CDK4/6 and MEK inhibition results in different anti-tumor immune responses in two cancer models (pancreas and lung) that are both driven by mutant Kras and Trp53. The difference may be caused by tissue-specific immune environments driven by the distinct embryonic origins of the two tissues. These results suggest that both cell autonomous (intrinsic) mechanisms (e.g. antigen presentation by tumor cells) and non-cell autonomous (extrinsic) mechanisms (e.g. NK cell or T cell recruitment) can contribute to enhanced anti-tumor immunity following CDK4/6 inhibition.

Correlating cell cycle progression with tumor immune suppression

Beyond evidence from drug studies, genomic and transcriptional profiling of human tumors also support the notion that cell cycle dysregulation in tumor cells promotes immune evasion. In one study, whole exome sequencing (WES) of 249 human tumors (spanning multiple tumor types) observed a strong correlation between resistance to ICB and genomic amplifications of cyclin D1 (an activator of CDK4/6) and CDK4 itself [7]. Likewise, a study comparing CD4 and CD8 T-cell-inflamed and non-T-cell-inflamed mouse pancreatic cancers found that the latter group exhibited a greater frequency of deletions and mutations in CDKN2A/B, encoding an endogenous inhibitor of CDK4/6 [8]. Together, these reports provide genomic evidence linking cell cycle regulation to modulation of anti-tumor immunity.

Analyses of tumor cell transcriptomes provide further evidence for such a connection. In a mouse model of pancreatic cancer driven by Kras^G12D and Trp53^R172H, for example, signatures of enhanced cell cycle progression were significantly enriched in immunologically “cold” tumors (having a paucity of T cells) compared to “hot” tumors [9]. Likewise, response to ICB therapy in a Braf^V600E-driven mouse model of melanoma was inversely correlated with the magnitude of a 7-gene “proliferation score” defined by key G2M checkpoint and E2F target genes [10]. Furthermore, single cell transcriptional analyses of human pancreatic cancers and melanomas demonstrated a clear association between the immune status of a tumor and heightened expression of cell cycle regulatory gens in the tumor cells [6,11]. Despite the correlative nature of these findings, they span multiple cancer types and are consistent with the results obtained with CDK4/6 inhibitors [6–11], suggesting that the inferred connection between tumor cell cycling and anti-tumor immunity is likely, in our view, a generalized feature of cancer biology.

Implications, mechanisms, and questions

From an immunological perspective, such an association seems counterintuitive. For many intracellular pathogens – principally viruses – propagation depends on the replicative activity of the host. One might therefore expect the immune system to be primed to recognize dividing cells as potential vectors of infection. An alternative possibility is that cell cycle activation induces an immunosuppressive response as a protective mechanism, allowing uninfected cells to divide and maintain tissue homeostasis in the face of an active infection. Under such a scenario, infected cells – whether proliferating or not – could still be detected and eliminated through standard mechanisms (antigen presentation, IFN signaling, etc.). Regardless of whether an evolutionarily conserved program is responsible for the inverse relationship between cell cycle regulation and anti-tumor immunity, these observations raise several important questions:

1. What molecular mechanisms connect tumor cell-intrinsic cell cycle progression with anti-tumor immunity? Cell cycle inhibition results in both cell autonomous (dysregulation of the antigen presentation machinery) and non-cell autonomous (e.g. release of SASP-associated factors and T cell recruiting chemokines) mechanisms of anti-tumor immunity [2–6]; however, the molecular events by which the cell cycle machinery regulates immune phenotypes remain largely unknown.

2. While most work in this area has focused on CDK4/6 inhibitors, which regulate the G1/S transition, would inhibitors acting at other phases of the cell cycle have similar immune effects? Of note, one recent study using both human and mouse small cell lung cancer cells reported that inhibiting another cell cycle regulator, CDK7, enhanced anti-tumor immunity program in both CD4 and CD8 T cells and ICB sensitivity by inducing genome instability [12]. Thus, there may be multiple opportunities for influencing the tumor immune microenvironment by targeting various components of the cell cycle machinery.

3. Does cell cycle progression interact with other cellular processes to affect anti-tumor immunity? Immune surveillance of tumor cells can be influenced by tumor cell-intrinsic DNA damage response, metabolic activity and autophagy. These cellular processes can modulate and be influenced by cell cycle progression. However, whether cell cycle regulators can affect anti-tumor immunity via cross-talk with these cellular processes remains to be explored [12].

4. Do such links between cell cycle progression and anti-tumor immunity affect metastasis? Metastatic competence is affected by the activity of various immune components – in both the primary tumor (invasion) and the metastatic site (colonization). Highly metastatic tumors might therefore reflect the combined influences of a hyperactivated cell cycle in the tumor cells and an associated reduction in immune constraints on spread. However, this remains to be assessed.

Concluding Remarks

Mounting evidence supports the hypothesis that tumor cell cycle programs impact anti-tumor immunity through cell autonomous and non-cell autonomous mechanisms; thus, inhibition of cell cycle progression might provide novel therapeutic opportunities (Figure 1). Of clinical relevance, inhibition of CDK4/6 activity has been shown to upregulate PD-L1 at either transcriptional or post-translational levels in human and mouse tumor cells which provides further rationale for combining CDK4/6 inhibitors with ICB to treating certain tumors [13,14]. This article focuses on the effect of tumor cell-intrinsic cell cycle progression on anti-tumor immunity. Nevertheless, there is increasing evidence showing that cell cycle inhibitors may have beneficial effects through the direct modulation of immune cells, including T cells. One recent study demonstrated that CDK4/6 inhibition could promote anti-tumor immunity by directly enhancing IL-2 secretion from CD8 T-cells through de-repression of NFAT activity as well as promoting abundance of T cells in a mouse lung cancer model [15]. This evidence further supports the potential use of cell cycle inhibition to improve the efficacy of immunotherapies.

Figure 1. Tumor cell-intrinsic cell cycle programs regulate anti-tumor immunity.

Diagram showing reported connections between cell cycle programs in tumor cells and the effects on immune recognition and infiltration. The left side of the figure shows a tumor cell with its cell cycle machinery fully intact, while the right side of the figure shows a tumor cell in which components of the cell cycle machinery have been perturbed. Cell cycle inhibition by CDK4/6 inhibitors results in both cell autonomous changes (e.g. reduced interferon signaling, decreased expression of antigen presentation machinery molecules, and increased expression of immunosuppressive molecules, for example PD-L1) and non-cell-autonomous changes (e.g. expression of secreted factors influencing immune cells or other components in the surrounding tumor microenvironment) that together lead to improved anti-tumor immunity.

Of note, clinical trials testing these combinations are currently ongoing in several types of malignancies, including non-small cell lung cancer or breast cancer (NCT02779751^I), head and neck cancer (NCT04169074^II, NCT03655444^III), hepatocellular carcinoma (NCT03781960^IV), prostate cancer (NCT04272645^V), glioma (NCT04118036^VI, NCT04220892^VII), and gastroesophageal adenocarcinoma (NCT03997448^VIII), As these trials “read out,” hopefully demonstrating patient benefits, it will be vital to have a better grasp of the molecular mechanisms connecting two of the most central processes in tumor biology: cell proliferation and immune surveillance.

Glossary

“Cold” tumors

non-T-cell-inflamed tumors with relative low abundance of infiltration of CD4 and CD8 T cells

“Hot” tumors

T-cell-inflamed tumors with relative high abundance of infiltration of CD4 and CD8 T cells

Immune checkpoint blockade

Immune therapies targeting inhibitory molecules expressed in T cells, including PD-1 blockade and CTLA-4 blockade

Endogenous retrovirus elements

Repetitive elements existing in the genome that resemble and can be derived from retroviruses

Senescence associated secretory phenotype

A phenotype wherein senescent cells secrete inflammatory cytokines, immune modulators, or proteases

DNA damage response

A cellular process that sense, signal and repair DNA damages through activation of a network of molecular machineries

Autophagy

A cellular process wherein cells remove cellular components within a cell through activation of a cascade of molecular events