Functional screens for cancer-modulating microRNAs

MicroRNAs (miRNAs) are small non-coding RNAs that control various cellular processes by inhibiting messenger RNAs (mRNAs) with partially complementary target sites. Many molecular pathways that are commonly perturbed in cancer, are subject to miRNA regulation. Introducing or reconstituting specific miRNAs in cancer cells may disrupt cancer progression, and ameliorate disease outcome. To identify cancer-modulating miRNAs, I created a lentiviral expression library containing the majority of all known human miRNAs and over 400 candidate miRNA loci, comprising a total of 1120 lentiviral constructs. This library allows the expression of miRNAs from their native backbone in a wide variety of mammalian cell types, including cancer cells, primary cells, senescent cells, stem cells, and whole tissues. After infection, the target cells will swiftly and stably express the introduced miRNA. One particularly valuable aspect of the library is its arrayed layout in 96-well format. This greatly facilitates high-throughput, clonal screening for miRNA-induced phenotypes of interest. I have employed the expression library to assess phenotypes that can be exploited for cancer treatment. In one chapter, I describe how I used the library to identify miRNAs with the potential to hamper melanoma cell growth. In another chapter, I screened for miRNAs that attenuate canonical Wnt signaling, the molecular driver of many colorectal tumors. The identification of miRNAs that cause beneficial phenotypes is the first step towards miRNA-based therapeutics for cancer. In the last two scientific chapters of my thesis, I explore two alternative uses of the library. In chapter 5, I’m pioneering a pooled approach that enables monitoring long-term growth effects of miRNAs. Given the necessary optimization, this approach can provide an assay to assess miRNA-induced growth effects of all miRNAs simultaneously in an in vivo setting. In chapter 6, I apply massively parallel RNA sequencing to determine (over)expression of miRNAs after lentiviral infection. The results of this experiment suggests highly variable processing efficiencies with which different miRNAs are produced from their precursor RNA molecules. It highlights the need for further examination of both structural conditions of precursor RNAs and determination of additional cofactors for efficient miRNA processing. In conclusion, with the lentiviral miRNA expression library I present a resource that proved valuable to systematically unravel miRNA functions, and will hopefully continue to do so in the future. I have applied the library to identify miRNAs that inhibit melanoma cell growth and canonical Wnt signaling. These findings constitute a first step towards miRNA-based therapy for cance