(A) A diverse population of lymphoma cells is retained in tumors at the time of disease presentation. Cells were infected with a vector control or a retrovirus co-expressing GFP and an shRNA targeting Top2A. Partially transduced cell populations were injected into recipient mice. Palpable tumors were treated with doxorubicin, and the percentage of GFP positive lymphoma cells was assayed at tumor relapse. (B) In vivo RNAi screening strategy. Lymphoma cells were infected with a pool of retroviruses containing 2250 distinct hairpins, and transduced lymphoma cell populations were injected into three recipient mice or passaged in three separate culture dishes. Lymphomas were harvested from tumor-bearing mice and shRNAs were PCR amplified from genomic DNA derived from tumors or from cultured cells. (C) The number of unique hairpins present in each sample after two weeks proliferation. (D) Clustering of samples based on shRNA enrichment or depletion. Color scale represents the mean normalized log₂ of the fold change in shRNA read number relative to input cells. (E) Hairpins that scored as enriched or depleted in vivo are largely distinct those that enriched or depleted in vitro. Diagrams show the number of scoring genes in the in vivo and in vitro settings.

(A) shRNA-mediated stable knockdown of Rac2, CrkL, and Twf1. Target protein/gene expression was measured by immunoblotting or qPCR. For qPCR samples, n=2 and bar graphs represent mean and standard deviation. (B) In vivo GFP competition assay to functionally validate candidate hits. Lymphoma cell cultures, partially transduced with a vector coexpressing GFP and the indicated shRNA, were maintained in culture for two weeks or injected into recipient mice. The fold change in the percentage of GFP positive lymphoma cells, relative to cells at injection, is shown. p-values were determined by a two-tailed Student's t-test. (n=7 for vector control and shRac2-1, n=8 for shRac2-2 in left panel; n=4 for vector control, n=3 for shCrkL-1 and shCrkL-2 in middle panel; n=4 for vector control and shTwf-2, n=3 for shTwf-1 in right panel). (C and D) Rac2, CrkL, or Twf1 suppression causes chemotaxis defects in transwell migration assays. Cells expressing a control vector, shRac2 (C), shTwf1, or shCrkL (D) were stimulated with SDF-1α. The number of cells that migrated is displayed as a fold change relative to control wells. p-values were determined by one-way ANOVA or a two-tailed Student's t-test. Bar graphs represent mean and standard deviation. n=2 for each experimental group.

(A) Rac2 suppression causes defects in short term engraftment. Lymphoma cells expressing dsCherry or coexpressing GFP and shRac2 were mixed, injected into recipient mice, and assessed after 2 or 24 hours. p-values were determined by one-way ANOVA or a one-sample t-test. Bar graphs represent mean and standard deviation. n=1 for injected cells, n=3 for all other experimental groups. (B) Lymphoma cells suppressing Wave2 were depleted in an in vivo GFP competition assay. p-values were determined by a two-tailed Student's t-test. (n=7 for vector control, n=4 for shWave2-1). (C) Suppression of Rac2 impairs lymphoma cell migration to the lymph nodes and liver. 14 days after transplantation, shRac2 recipients show markedly reduced tumor dissemination. (D) Partially-transduced lymphoma cells were harvested from the liver at the time of disease presentation and the percentage of GFP positive cells was assessed. p-values were determined by one-way ANOVA. (n=7 for vector control, n=4 for shRac2-1 and shRac2-2) (E and F) Suppression of Rac2 in lymphoma cells delays disease progression. GFP sorted lymphoma cells were injected into recipient mice. Survival is displayed in Kaplan-Meier format (n=5 mice per group for tumor free survival and n=10 mice per group for overall survival).

(A) Pharmacological inhibition of Rac2 extends animal lifespan. Mice were treated with the Rac inhibitor NSC23776 every 12 hours starting nine days after injection of lymphoma cells. Results are shown in Kaplan-Meier format (n=5 for each experimental group). (B) Suppression of Rac2 activity extends animal survival following vincristine treatment. Mice bearing vector control or shRac2 tumors were treated with vincristine 11 days after injection of lymphoma cells, and overall survival was monitored (n=11 for each experimental group. Data was compiled from three independent experiments). (C) A model for the role of Rac2 in relapse following vincristine treatment. The appearance of terminal disease following vincristine treatment may require tumor cell migration from sites of residual disease to metastatic sites, including liver, lung, and brain.