Intact p53-dependent responses in miR-34-deficient mice

Abstract

MicroRNAs belonging to the miR-34 family have been proposed as critical modulators of the p53 pathway and potential tumor suppressors in human cancers. To formally test these hypotheses, we have generated mice carrying targeted deletion of all three members of this microRNA family. We show that complete inactivation of miR-34 function is compatible with normal development in mice. Surprisingly, p53 function appears to be intact in miR-34-deficient cells and tissues. Although loss of miR-34 expression leads to a slight increase in cellular proliferation in vitro, it does not impair p53-induced cell cycle arrest or apoptosis. Furthermore, in contrast to p53-deficient mice, miR-34-deficient animals do not display increased susceptibility to spontaneous, irradiation-induced, or c-Myc-initiated tumorigenesis. We also show that expression of members of the miR-34 family is particularly high in the testes, lungs, and brains of mice and that it is largely p53-independent in these tissues. These findings indicate that miR-34 plays a redundant function in the p53 pathway and suggest additional p53-independent functions for this family of miRNAs.

Figure 1. MiR-34 expression in wild-type and p53^−/− mouse tissues.

(A) Sequence alignment of mouse miR-34a, miR-34b and miR-34c. Differing nucleotides are colored in blue. The seed sequences are in bold. (B–D) MiR-34a and miR-34c expression as detected by qPCR (B,C) and by Northern blotting (D) in tissues of wild-type and p53^−/− mice.

Figure 2. Targeted deletion of miR-34a and miR-34b∼c.

(A) Targeting and screening strategy for the generation of constitutive and conditional miR-34a KO alleles. The restriction sites used for the Southern blot screening are indicated (S = SphI, E = EcoRI). The gray bar with an asterisk represents a genomic region absent in the 129SvJae strain but present in the C57BL/6 strain, which results in two distinct sizes in digestions. (B) Targeting and screening strategy for the generation of miR-34b∼c KO allele (H = HindIII, S = SpeI). (C) Genotyping by tail genomic PCR showing germline transmission of the miR-34a deleted and floxed alleles (upper panel), and the miR-34b∼c deleted allele (lower panel). (D) Northern blotting (upper panel) on total RNA extracted from the testes of mice with the indicated genotypes. Probes specific for miR-34a and miR-34c were used. Complete loss of miR-34a and miR-34c expression was further confirmed in MEFs by qPCR (lower panel). Representative pictures of miR-34a^−/− (E), miR-34b∼c^−/− (F), and miR-34^TKO/TKO (G) males at 4 weeks of age. The table below each picture summarizes the expected and observed frequencies of mice of each genotype as obtained from heterozygous inter-crosses. For the miR-34^TKO allele (G), double heterozygous mice were inter-crossed.

Figure 3. Response to p53 activation in miR-34^TKO/TKO mouse embryonic fibroblasts (MEFs).

(A) MiR-34a and miR-34c expression in serially-passaged wild-type MEFs, as measured by qPCR. Error bars indicate 1 standard deviation (SD). (B) Cumulative population doublings of wild-type, miR-34^TKO/TKO and p53^−/− MEFs. Error bars indicate 1 SD. (C) Growth curves of wild-type and miR-34^TKO/TKO MEFs. Error bars indicate 1 SD. (D) Immunoblots of p53, p21 and Mdm2 in wild-type (W) and miR-34^TKO/TKO (K) MEFs treated with 0.2 µg/ml doxorubicin for the indicated time. (E) Expression of selected p53 targets in total RNA from doxorubicin-treated MEFs. Cells were treated with 0.2 µg/ml doxorubicin for 12 hours (Dox) or left untreated (U). Expression of the indicated genes was determined by qPCR. Error bars represent 1 SD. (F) Immunoblots showing p53 activation in three wild-type and three miR-34^TKO/TKO MEF lines. Cells were left untreated or treated with 0.2 µg/ml doxorubicin for 12 hours. (G) Time course of miR-34a and miR-34c expression in wild-type and p53^−/− cells treated with 0.2 µg/ml doxorubicin. MicroRNA expression was determined by qPCR. Error bars indicate 1 SD. (H, I) Cell cycle distribution of wild-type and miR-34^TKO/TKO MEFs. Asynchronously growing MEFs of the indicated genotype were treated with increasing doses of doxorubicin for 16 hours (H), or with 0.2 µg/ml doxorubicin for increasing time (I). Error bars indicate 1 SD. (J) Upper panel: cell cycle distribution of wild-type, miR-34^TKO/TKO, and p53^−/− MEFs after 72 hours in starvation medium (gray histogram). Starved cells were released in complete medium containing colcemid and mock-treated (light blue histogram) or exposed to 20 Gy irradiation (red histogram). Cells were analyzed by 7-AAD staining at the indicated time after release in complete medium. Lower panel: percentages of irradiated and untreated cells in G1 and G2-M phases after 24 hours in complete medium. Experiments were performed on three independent wild-type and three independent miR-34^TKO/TKO MEF lines. (K) Immunoblot detection of predicted miR-34 targets on three independent wild-type and three independent miR-34^TKO/TKO MEF lines.

Figure 4. p53-dependent apoptosis in thymocytes and in vivo.

(A) Percentages of viable wild-type, miR-34^TKO/TKO, and p53^−/− thymocytes 16 hours after treatment with increasing doses of irradiation (0, 2, 4, 6, 8, and 10 Gy). Error bars represent 1 SD. (B) Percentages of viable wild-type, miR-34^TKO/TKO, and p53^−/− thymocytes 4, 8, and 24 hours after irradiation (5 Gy). Error bars correspond to 1 SD. (C) Expression levels of p53 transcriptional targets in the thymi and spleens of untreated (U) and irradiated (IR, 10 Gy) wild-type, miR-34^TKO/TKO and p53^−/− mice by qPCR. (D, E) Representative cleaved caspase-3 immunohistochemistry of the thymus (D) and the small intestine (E) of untreated and irradiated (10 Gy) wild-type, miR-34^TKO/TKO and p53^−/− mice (n = 3 mice per group). Brown staining indicates cleaved caspase-3 (CC3). (F,G) Quantification of apoptosis in the thymus (F) and in the intestine (G) of control and irradiated animals. In panel F the relative staining intensity averaged over three microscopic fields per sample is plotted. In panel G, the average number of CC3-positive cells per crypt is plotted. At least 25 randomly selected crypts per sample were counted. Error bars correspond to 1 SD. P values were calculated using the two-tailed Student's t-test.

Figure 5. Oncogene-induced transformation in miR-34^TKO/TKO fibroblasts and mice.

(A) Representative focus formation assays of wild-type, miR-34^TKO/TKO, and p53^−/− MEFs. MEFS were infected with retroviruses expressing K-Ras^V12 alone or K-Ras^V12 and E1A. The results are representatitve of two independent experiments performed on a total of four wild-type and four miR-34^TKO/TKO MEF lines. (B) Bar plot showing the number of transformed foci. Error bars are 1 SD. (C) Survival curves of Eμ-Myc;miR-34^+/+ and Eμ-Myc;miR-34^TKO/TKO mice. P-value was calculated using the log-rank (Mantel-Cox) test. (D) Histopathology and cleaved caspase-3 (CC3) immunohistochemistry of representative lymphomas obtained from Eμ-Myc;miR-34^+/+ and Eμ-Myc;miR-34^TKO/TKO mice. (E) Bar plot showing the number of CC3-positive cells per low magnification field. Five Eμ-Myc;miR-34^+/+ tumors and and four Eμ-Myc;miR-34^TKO/TKO tumors were analyzed. Error bars indicate 1 SD.