Toolkit for evaluating genes required for proliferation and survival using tetracycline-regulated RNAi

Abstract

Short hairpin RNAs (shRNAs) are versatile tools for analyzing loss-of-function phenotypes in vitro and in vivo. However, their use for studying genes involved in proliferation and survival, which are potential therapeutic targets in cancer and other diseases, is confounded by the strong selective advantage of cells in which shRNA expression is inefficient. We therefore developed a toolkit that combines Tet-regulated miR30-shRNA technology, robust transactivator expression and two fluorescent reporters to track and isolate cells with potent target knockdown. We demonstrated that this system improves the study of essential genes and was sufficiently robust to eradicate aggressive cancer in mice by suppressing a single gene. Further, we applied this system for in vivo negative-selection screening with pooled shRNAs and propose a streamlined, inexpensive workflow that will facilitate the use of RNA interference (RNAi) for the identification and evaluation of essential therapeutic targets.

Figure 1

Dual-color TRMPV vectors enable Tet-regulated shRNA expression for suppression of genes involved in cell proliferation and survival. (a) Vector schematic of TRMPV, which was constructed in the pQCXIX self-inactivating (SIN) retroviral backbone. Ψ+, extended retroviral packaging signal. (b) Western blot of immortalized Rosa26-rtTA-M2 MEFs transduced with TRMPV harboring different Rpa3 shRNAs (numbered by start position) or a Renilla luciferase (Ren) shRNA. After selection, cells were cultured in the presence or absence of doxycycline for 4 d before collection. β-actin served as loading control. Uncropped blots are shown in Supplementary Figure 2d. (c) Representative flow cytometry plots of Rosa26-rtTA-M2 MEFs transduced with TRMPV.shRen.713 and TRMPV. shRpa3.455 in competitive proliferation assays. Cells selected for TRMPV were mixed with untransduced cells and passaged in doxycycline for 16 d. (d) Quantification of fluorescent cells in representative competitive proliferation assays. Each series of bars is a time course from left to right: day 0, 4, 8, 12, 16, 20. In all series, day 0 bar represents percentage Venus-positive cells before doxycycline treatment. In the presence of doxycycline, transduced cells were gated either on Venus or on both Venus and dsRed. (e) Vector schematic of TRMPVIR showing constitutive and inducible transcripts produced by the vector. IRES-dependent rtTA3 expression from the inducible transcript creates a positive feedback loop of TRE induction. (f) Representative flow cytometry plots of Eμ-myc;Trp53^−/− lymphoma cells transduced with TRMPVIR.shRen.713 and TRMPVIR.shRpa3.455 in competitive proliferation assays over 10 d. Cells were incompletely transduced with TRMPVIR before day 0 (rather than selected and admixed with untransduced cells). (g) Quantification of fluorescent cells in representative competitive proliferation assays. Each series of bars is a time course from left to right: day 0, 2, 4, 6, 8, 10; see d for details.

Figure 2

TRMPV enables RNAi-based evaluation of genes involved in tumor maintenance in vivo. (a) Schematic of the generation and application of a Tet-On-competent mouse model of leukemia. Hematopoietic stem and progenitor cells (HSPC, Cd45.2⁺) are transduced with retroviruses that coexpress oncogenes, rtTA3 and firefly luciferase and subsequently transplanted into recipient mice. Resulting Tet-On leukemias are collected, transduced with TRMPV, selected and retransplanted into secondary Cd45.1⁺ recipients. After leukemia onset, shRNAs are induced by doxycycline and effects analyzed using different readouts. (b) Representative bioluminescence imaging of recipient mice transplanted with TRMPV-transduced AML cells. Mice were treated with doxycycline at disease onset (day 0). (c) Kaplan-Meier survival curve of recipient mice of AML cells transduced with indicated TRMPV shRNAs. Mice were treated with doxycycline at disease onset as assayed by imaging (7 d after transplantation). (d) Representative flow cytometry plots of donor-derived (Cd45.2⁺) cells in bone marrow of moribund doxycycline-treated mice in c. (e) Quantification of Venus⁺dsRed⁺ cells in Cd45.2⁺ bone marrow cells collected from doxycycline-treated recipient mice (n = 4) at a terminal disease stage. Mean and s.e.m. are plotted.

Figure 3

TRMPV-induced suppression of Rpa3 cures clonal MLL-AF9;Nras^G12D AML. (a) Bioluminescence imaging of recipient mice of clonal MLL-AF9;Nras^G12D AML harboring TRMPV.shRpa3.455. After leukemia onset as assayed by imaging (day 0) mice were either left untreated (no dox), treated with doxycycline (early dox) or treated at a more advanced disease stage (late dox). (b) Bone marrow and liver histology of untreated and doxycycline- treated mice 4 d after leukemia onset. Scale bars: 20 μm for bone marrow, 100 μm for liver. (c) Kaplan-Meier survival curve of recipient mice of clonal TRMPV.shRpa3.455 or TRMPV. shRen.713 leukemias. After disease onset as assayed by bioluminescent imaging (day 7 after transplantation), mice were either left untreated (off dox) or treated with doxycycline for 40 d (on dox).

Figure 4

Pooled negative selection RNAi screening in vivo detects shRpa3 depletion in MLL-AF9;Nras^G12D AML. (a) Scatter plot illustrating the correlation of normalized reads per shRNA between the plasmid library and transduced selected leukemia cells before transplantation (T₀); r, nonparametric (Spearman) correlation coefficient. (b) Scatter plot of normalized reads per shRNA in T₀ cells compared to an untreated leukemic recipient mouse. (c) Scatter plot of normalized reads per shRNA in T₀ cells compared to average reads in three untreated recipient mice. (d) Relative abundance of Rpa3 and Renilla luciferase shRNAs in leukemias isolated from untreated (off dox) and doxycycline-treated (on dox) recipient mice, each compared to the initial representation before transplantation (T₀). Leukemias from doxycycline-treated mice were analyzed both without and with purification of shRNA-expressing cells (Venus⁺dsRed⁺) before DNA isolation. (e) Relative abundance of all 824 shRNAs in Venus⁺dsRed⁺-sorted leukemia cells from doxycycline-treated mice compared to T₀ cells. The mean of normalized reads in doxycycline-treated mice (n = 3) was divided by normalized reads in T₀ cells; shRNAs are plotted according to the resulting ratios in ascending order. All three shRNAs targeting Rpa3 were among the 25 most depleted shRNAs, whereas neutral shRen.713 was not altered.