Towards transcriptome-wide studies of mRNA translation in tissues from cancer patients

Abstract

Gene expression consists of multiple strictly regulated steps, including transcription, RNA modification, splicing, messenger RNA (mRNA) transport, mRNA degradation, mRNA translation and protein degradation. mRNA translation, the most energy consuming step, plays a critical role in gene expression via global and selective control of protein synthesis. Translation is a complex process that is commonly divided into initiation, elongation and termination. Among these, translation initiation is widely acknowledged as the rate-limiting step for mRNA translation. The mammalian/mechanistic target of rapamycin (mTOR) pathway, as one important regulator of translation initiation, delivers vital signaling by phosphorylating eIF4E binding proteins (4E-BPs) thereby facilitating eIF4F complex formation which participates in eukaryotic cap dependent translation. Increased mTOR activity and dysregulation of translation have been observed in many diseases, such as cancer as well as immune and metabolic disorders. Sequence and structure features of the mRNA, the translational apparatus and trans-acting proteins facilitate or restrict translation regulation of an mRNA. Moreover, these factors can potentially alter the translational efficiency of an mRNA thereby impacting protein levels without changes in mRNA levels. Accordingly, a well-established technique to study translatomes, polysome profiling, separates efficiently translated mRNA from total mRNA into multiple fractions based on the number of ribosomes bound on the mRNA. Extraction of these fractions is a time consuming and laborious process, which makes polysome profiling inconvenient for large experiments or samples with low RNA amounts. Until now, these shortcomings have prevented assessments of translatomes in patient tissue samples.

This thesis introduces an optimized non-linear sucrose gradient which consistently enriches the efficiently translated mRNA in merely one or two fractions, thus reducing sample handling 5-10 fold and saving time in the lab 10-20 fold. When combined with smart-seq2 RNA sequencing, translatomes can be obtained from samples with low amount of RNA and small bio-banked tissues. mRNA yields and translatomes acquired from the optimized gradients resemble those obtained from the standard linear gradients. Thus, this optimized polysome-profiling technique expands the usage of the methodology to small tissue samples and primary cells in large study designs.

Insulin sensitive mRNA translation has been observed in cancer cells derived from insulin insensitive organs, for instance breast. It is largely unknown that if this insulin sensitivity resembles that of cells from insulin sensitive organs or if cancer cells tailor a novel program. To this end, this thesis explored insulin’s effect on metabolomes and translatomes in human primary myotubes, human mammary epithelial cells immortalized with human telomerase (HMEC/hTERT) and the MCF7 breast cancer cells. The data indicates that MCF7 cells have developed pathological responses to insulin induction that differ from those observed in cells from insulin sensitive or insensitive organs. The exploration of mechanisms concealed behind this discrepancy would disclose a potential strategy for cancer treatment through annulment of cancer specific effects of insulin.

The role of mRNA translation during treatments with experimental anti-cancer drugs or those used in the clinic is largely unknown. We examined the effect on translation of one such experimental drug called “Reactivation of p53 and induction of tumor cell apoptosis” (RITA). The α subunit of eukaryotic initiation factor 2 (eIF2α) is a key regulator of translation initiation. We found eIF2α to be phosphorylated during RITA treatment and to be involved in RITA induced apoptosis and repression of mRNA translation. This activity of RITA is independent of TP53 and mTOR pathway. The inhibition of eIF2α phosphorylation counteracts the impact of RITA on apoptosis and clonogenicity.

Another aspect of this thesis explored regulation of translation in immune cells. Short post-infusion persistence restricts treatment of hematological malignancies via adoptive infusion of stimulated natural killer (NK) cells. Interleukin-15 (IL-15) was demonstrated to hold stronger ability than IL-2 to maintain antitumor functions of NK cells after cytokine deprivation. To explore the mechanism underlying these differences, a transcriptome wide study through polysome-profiling technique was applied. Further, the role of mTOR pathway in this superiority of IL-15 was also investigated. Coupled with clinical outcome of patients with B-cell lymphoma, IL-15 but not IL-2 is argued to be implemented in adoptive NK cell cancer therapy.

In conclusion, in order to facilitate studies of the translatome for samples with low amount of RNA and small bio-banked tissues, the optimized non-linear gradient was designed. Its performance in aforementioned samples for large experiment set and general applicability was verified to be satisfying. The study on cancer specific effects of insulin unraveled the prospect to selectively target insulin/IGF1 dependent effects on metabolomes and/or translatomes for cancer therapy. As two important pathways regulating translation initiation, the effect of mTOR in immune cell functions and eIF2α in RITA induced apoptosis were unveiled and explored.