Co-injection strategies to modify radiation sensitivity and tumor initiation in transgenic Zebrafish

Abstract

The zebrafish has emerged as a powerful genetic model of cancer, but has been limited by the use of stable transgenic approaches to induce disease. Here, a co-injection strategy is described that capitalizes on both the numbers of embryos that can be microinjected and the ability of transgenes to segregate together and exert synergistic effects in forming tumors. Using this mosaic transgenic approach, gene pathways involved in tumor initiation and radiation sensitivity have been identified.

Figure 1

Co-injection of multiple transgenes into one-cell stage zebrafish embryos leads to co-expression in tumors. Rhabdomyosarcoma (RMS) was induced by injection of rag2-kRASG12D (a–f), whereas T-cell acute lymphoblastic leukemia (T-ALL) was induced by injection of rag2-mMyc (g–l). Oncogene-containing transgenes were co-injected with rag2-GFP (a, b, g and h), rag2-dsRED2 (c, d, i and j) or both (e, f, k and l). Fluorescent microscopic images show expression of GFP (a and g), dsRED2 (c and i) or both (e and k). FACS analysis confirms expression patterns of tumor cells from animals shown in the corresponding fluorescent images (b, d, f, h, j and l). The rag2-GFP (Jessen et al., 2001), rag2-mouse cMyc (Langenau et al., 2003), rag2-dsRED2 and rag2-human kRASG12D (Langenau et al., 2007) constructs have been described previously. DNA constructs were linearized with either XhoI or Not1, phenol:chloroform extracted and ethanol precipitated as described previously (Langenau et al., 2003). For rag2-oncogene, rag2-dsRED2 co-injections (a, b, g and h), the oncogene containing construct was linearized with XhoI and the second linearized with NotI. For rag2-oncogene co-injections with rag2-GFP (c, d, i and j), both constructs were digested with the same enzyme, either NotI or XhoI. DNA was quantified by both spectrophotometric analysis and gel quantification. DNA was diluted in 0.5XTE + 100 mM KCl to 100 ng μl⁻¹, with co-injection of two transgenes having 50 ng of each construct per microliter or three transgenes having 33.3 ng of each construct per microliter. FACS, fluorescence activated cell sorting; GFP, green fluorescent protein.

Figure 2

Co-injection approaches can be used to modify radiation sensitivity in Myc-induced T-cell acute lymphoblastic leukemias (T-ALLs). Primary T-ALLs were generated by co-injecting rag2-mMyc along with other transgenes denoted to the left of figure (a, d and g). Primary tumor cells were transplanted into irradiated recipients and engrafted cells could be visualized 10 days later (b, e and h). Transplanted animals were irradiated (23 Gy) and assessed for tumor regression at 4 days post-treatment (c, f and i). Control tumors were generated in AB animals injected with rag2-mMyc + rag2-dsRED2 (a–c). T-ALLs from AB fish injected with rag2-mMyc + rag2-EGFP-bcl2 (d–f). p53 loss-of-function animals (p53-LOF) injected with rag2-mMyc + rag2-GFP (g–i). Fish oriented with head toward left and dorsal toward top. The rag2-EGFP-zebrafish bcl2 construct was described previously (Langenau et al., 2005b). Cell transplantation, FACS and irradiation protocols were completed essentially as described (Langenau et al., 2003, 2005b; Traver et al., 2003). 1 × 10⁶ unsorted cells were used in all transplantation experiments. FACS, fluorescence activated cell sorting.

Figure 3

Noxa is a potent genetic suppressor of RAS-induced rhabdomyosarcoma. Tumor onset in AB wild –type (WT) (a) or p53 loss-of-function animals (b) co-injected with rag2-kRASG12D and either rag2-noxa, rag2-dsRED2 or rag2-GFP. (c) Diagram of rag2-noxa transgene with accompanying primers used in analysis. (d) PCR of genomic DNA isolated from the externally-visible tumor portion of rag2-kRASG12D + rag2-noxa or rag2-kRASG12D (WT) injected AB-strain animals. The upper panel shows presence of the rag2-noxa transgene (primers used in this analysis −277F + R1). The lower panel shows presence of genomic DNA amplification of the endogenous rag2 locus. (e) Quantitative reverse-transcriptase PCR showing that the noxa transgene is expressed in RMS (RT +). No RT (RT −) controls confirm that amplification is not due to contaminating genomic DNA. The rag2-zebrafish noxa transgene was created by cloning a BamHI/NotI fragment that contains both the open reading frame and the SV40 polyadenylation site from the pcs2 + noxa expression vector (CJ and ATL, unpublished data) into the rag2-GFP construct (Jessen et al., 2001). Injection volumes were assessed by injecting DNA samples into oil. One nanoliter of DNA was injected per animal.

Figure 4

Co-injection approaches can be used to selectively induce gene expression within established RAS-induced rhabdomyosarcomas. The rag2-kRASG12D transgene was co-injected with rag2-GFP and hsp70-dsRED2 into AB strain, one-cell stage embryos. By 30 days of life, co-injected animals that contained GFP-labeled RMS (a–d) were placed into room temperature fish water (40 ml in a 50-ml Falcon tube) and allowed to acclimate for 10–20 min. Next, animals were transferred to 37 °C and incubated for 45 min. Fish were removed from the water bath and allowed to acclimate to room temperature for 10–15 min, after which time they were introduced back into individual tanks and raised at 28.5 °C. Two days post-heat-shock, RMS expressed both GFP and dsRED2 in surviving animals (e–h, n = 19 of 20). Confocal microscopic analysis of the same animal shown in a–h following heat-shock (×10, i–k) was performed essentially as described (Langenau et al., 2007). Brightfield (a and e), GFP (b, f and i), dsRED2 (c, g and j) and merged fluorescent (d, h and k) images. The hsp70-dsRED2 construct was created by amplifying the hsp70-4 promoter from genomic DNA using a forward primer that contains an XhoI digest site and a reverse primer containing a BamHI site (Supplementary Table 3). This PCR fragment was purified, digested and cloned into the rag2-dsRED2 construct, replacing the rag2 promoter sequence. GFP, green fluorescent protein.