Preprint Club
A cross-institutional Journal Club Initiative

Keywords

• Ribosomal heterogeneity

• Tumour immunity

• Tumour cell biology

Main Findings

In this preprint, the authors investigate the ways in which human melanoma tumour cells respond to pro-inflammatory cytokines common to the tumour microenvironment. Particularly, the authors wanted to investigate if a particular ribosomal subset within the tumour cell is responsible for driving the translation of cytokine-responsive mRNAs. To investigate whether pro-inflammatory cytokines selectively alter the ribosomal translation machinery, the authors performed LC-MS analysis on polysomes isolated from Mel624 and M026 human melanoma cell lines stimulated with or without a cocktail of TNFɑ/IFN𝛾. In their analysis, the ribosomal P-stalk protein – P1 – was upregulated in response to the cytokine mixture relative to controls. Importantly, although it was not present in their LC-MS analysis, the P2 protein, required to complete the P-stalk heterodimer equally increased in polysomes of TNFɑ/IFN𝛾–treated cells when examined via immunoblotting. To show that the P1/P2 “alert state” ribosome (ASR) is a unique cytokine-responsive population, the authors confirmed that the uL30 ribosomal protein was not induced in polysomes following cytokine treatment when examined via LC-MS or immunoblot. The authors further verify the presence of the ASR in SK-OV-3 ovarian adenocarcinoma, HCT116 colorectal carcinoma, and immortalized RPE-1 retinal pigment cells.

To validate the biological importance of the ASR, knockdown (KD) melanoma cell lines of P1 were generated along with scramble control, or eL28 (another ribosomal protein found to not be induced by TNFɑ/IFN𝛾) knockdown cells. These cell lines were then treated with the cytokine cocktail and the absolute protein abundance of HLA class I isoforms was determined by Western blot. The total protein and surface HLA was found to be diminished in the shP1 KD cells relative to the control cell lines. This effect was found to be independent of mRNA transcription as shP1 KD cells did not have a decrease in HLA class I transcript levels. Furthermore, bulk proteomic analysis comparing shP1 vs. scramble control KD cells indicated a broad downregulation of multiple class I antigen presentation machinery components in response to TNFɑ/IFN𝛾. Then, the authors asked whether TNFɑ/IFN𝛾-induced mRNAs were selectively bound by the P1-containing ASR. To answer this question, HA-P1 and control HA-eL22 constructs were incorporated into the melanoma cell lines, then exposed to TNFɑ/IFN𝛾 stimulation, followed by immunoprecipitation of HA-P1 and HA-eL22. Analysis of mRNAs precipitated with ASR revealed that P1-containing ribosomes associated with mRNAs constituting the response to cytokine stimulation (e.g ICAM1, CXCL9, CXCL10, CXCL11) and mRNAs of the class I antigen-presentation machinery. Their results therefore indicate that the ASR is responsible for translation of cytokine-responsive mRNAs.

To further validate the functional importance of the ASR in T cell mediated tumour killing, P1 KD cells were co-cultured with CD8+ T cells containing TCRs specific for antigens naturally presented by melanoma cell lines. Indeed, they found that the P1 KD melanoma cells had a lower propensity to activate the cognate CD8+ T cells and were more resistant to T cell mediated killing. This effect the authors argue is due to the blunted surface class I HLA expression due to the knockdown of the P1 ASR.

To determine the mechanisms involved in regulating selective cytokine responsive mRNA translation by ASR, the authors performed phosphoproteomics on TNFɑ/IFN𝛾-treated vs control tumor cells. Their analysis revealed that the pro-inflammatory cytokines lead to a de-phosphorylation of P1/P2. Furthermore, kinome analysis using published datasets revealed a possible role for TGFβ-receptor kinases in P1/P2 phosphorylation. Indeed, TGFβ-treated cells had a decrease in the P1/P2 stalk proteins in their polysomes. Collectively, these results indicate that the dephosphorylation and phosphorylation of the ASR lead to its activation and inhibition respectively.

Lastly, by parsing the published cancer genome atlas program (TCGA) database, the authors were able to find that an enrichment of P1 and P2 relative to other ribosomal proteins were positively correlated with intratumoral IFNγ and effector CD8+ T cell signatures.

Limitations

• The authors argue that the discovery of the ASR is a conserved mechanism across cancer and tissue types. Although the authors have done a very thorough investigation into the mechanisms of the ASR in various tumour cells in vitro, more studies employing non-tumour cell types representing various tissues are needed to strengthen this claim.

• In line with the limitation above, an outstanding question is the role of the ASR in typical antigen-presenting cells such as DCs, B cells, and Macrophages given their heightened antigen-presentation following microbial or cytokine exposure. Does the ASR mediate these processes?[AM1]  Furthermore, do other pro-inflammatory cytokines such as IL-1, IL-6, and all interferons mediate ASR-driven mRNA translation?

• It would be helpful to determine mechanistically how the ASR mediates preferential translation of cytokine-responsive mRNAs following cytokine treatment. For example, is there a particular affinity for the ASR in recognizing specific motifs present on mRNAs induced by pro-inflammatory cytokines?

• The bulk proteomics data and HA-P1 precipitation ribosomal profiling data is presented in the form of a mechanistic diagram. It would be better to show the data in the form of a volcano plot including adjusted p-values and fold change to appreciate the variety of transcripts and proteins that are upregulated and downregulated. This would provide a more holistic representation of the data.

• The authors claim that the ASR P1/P2 protein activity is regulated by phosphorylation and dephosphorylation states. Mechanistically it would be interesting to mutate potential phosphorylation sites to see whether this would lead to ASR hyperactivation.

• Furthermore, a phospho-blot of P1/P2 would be a good means to verify the effects of TGFβ on ASR phosphorylation.

• Further biological repeats are required to strengthen the manuscript. For example, certain Western blot images are described as being only an n = 1 independent experiment.

Significance/Novelty

The authors establish a new ribosome population – the “alert state ribosome” – that selectively translates cytokine-responsive mRNAs. Since the discovery of mRNA in 1961, ribosomes have been considered to be mostly homogeneous. Although there have been investigations into the existence of potential subsets of ribosomes, thorough approaches to investigate ribosomal heterogeneity remains sparse. Considering this framework, the preprint by Dopler et al. provides novel insights into ribosome activation, heterogeneity and selectivity. Their work is of great general importance, as it is one of the first to consider mechanisms in which immunological molecules such as HLA are regulated at the level of translation. Indeed, most studies on intra-cellular immunological processes focus mainly on their transcriptional or post-translational regulations. The authors’ work thus provides a new way of thinking about cytokine responsiveness by tumour and healthy cells, by positioning the ribosome as a nexus of such signal integration.

This work reveals a new mechanism by which tumour cells may respond to cytokines present in the tumour microenvironment. In fact, the authors show that the intratumoral levels of P1/P2 correlate positively with IFNγ and effector CD8+ T cell signatures. Therefore, the relative levels of P1/P2 may be used during tumour biopsies to potentially predict efficacy of immune checkpoint blockade and other therapies that make use of IFNγ and CD8+ effector T cell responses as treatment mechanisms.

Credit

Reviewed by Boyan Tsankov as part of a cross-institutional journal club between the Icahn School of Medicine at Mount Sinai, the University of Oxford, the Karolinska Institute and the University of Toronto.

The author declares no conflict of interests in relation to their involvement in the review.