Construction Of ShRNA Expression Libraries

The thesis consists three part of work concerning the construction of shRNA(short hairpin RNA) expression library. First, conventional protocol in whichshRNA expression library was generated from enzyme-digested cDNA wasimproved by introducing restriction enzyme CviKI-1 and employing ethanolprecipitation to recover DNA fragment. Second, a new kind of shRNA librarycalled randomized shRNA libraries was genetrated by employing a new protocoldeveloped by us, providing another choice for whole-genome phenotypic screeningof genes. Third, we investgated the possibility of constructing randomized or cDNAderived shRNA-mir expression library.Two techniques has been applied to improve the enzyme-digestion protocol. Thefirst technique was the introduction of one more restricted enzyme—CviKⅠ-1 into theenzyme-digestion phase. The enzymes adapted in the original method (HinpI(GCGC),BsaHI(GPuCGPyC), AciI(AACGTT), HpalI(CCGG) HpyCH4IV (ACGT)andTaqaI(TCGA)) averagely has only 5.4 sites/Kb as been anlyzed with nine cDNAfragments randomly chosen from the cDNA of human and mouse, while the distributionof CviKⅠ-1 site was about 20 sites/Kb. According to the results of our experiments, thevariety of the digestion fragments of EGFP eDNA by CviKⅠ-1 was consistent with theprediction and a library with considerable complexity may be acquired, while therelatively more restriction sites created by the original enzyme-digestion protocol was afalse appearance. Besides, direct ethonol precipitation might recover more digestionfragments at a higher concentration. The recoverd fragments were used afterward andno detrimental effects had been oberserved in the linking and PCR expriments. Thesetwo improvements ensured a library with relatively higher complexity while alleviatedthe difficulties involved in the original method.To construct a new kind of shRNA library called randomized shRNA library, ashRNA expression vector pLENTI-hU6m was constructed which contains a modifiedhuman U6 promoter sequence into which Xho I and Mly I was introduced. To testwhether our modifications alter the activity of promoter, a shRNA targeting EGFP with a modified human U6 promoter sequence was made. The "loop" we created in theshRNA library is a palindrome loop. we tested whether this change could effect theprocessing of shRNA to a functional siRNA, chemically synthesized randomized19-mers DNA were efficiently converted to double-stranded DNA fragments containingshRNA templates by a protocol developed by us. We inserted these random fragmentsinto shRNA expression vector pLENTI-hU6m. After transformation we isolated andsequenced 20 independent constructs. Of these, 19 constructs contained inserts with theappropriate structures and all were unique, This means that a high-complexitylibrary(1.1×10~7) was made and for any average mRNA of 1kb from any organism, therewill be 110 shRNA against it in the library. Compared with generation of shRNAlibrary by enzymatic engineering of cDNA, a randomized shRNA library has severaladvantages. It could be used to screen all possible genes in different cell types anddifferent species where shRNA is applicable.siRNA sequences can be engrafted into a microRNA structure under the control ofpolⅡpromoters. This shRNA-mir structure will have higher penetrance in knockingdown than simple hairpin designs. The current protocols of constructing a eDNAderived shRNA-mir expression library introduces modifications of flanking sequenceand stem length of miRNA and a randomized shRNA-mir expression library introducesmodifications of flanking sequence of miRNA. The effects of these two changes onpre-miRNA processing were unclear, and they were investigated by constructing twoshRNA-mir mutations targeting EGFP, one has a modified flanking sequence, the othera longer stem length. Changing flanking sequence without perturbing the secondarystructure appeared to adversely affect shRNA-mir processing. But the longer stemlength, even 3bp, may affect the proper processing of pre-miRNA resulting in no RNAinterference. This means that the current enzymatic protocols could not be used toconstruct shRNA-mir expression library. A randomized shRNA-mir expression librarycan be constructed using the current protocols.