Tumor suppressor BRCA1 epigenetically controls oncogenic microRNA-155

Abstract

BRCA1, a well-known tumor suppressor with multiple interacting partners, is predicted to have diverse biological functions. However, so far its only well-established role is in the repair of damaged DNA and cell cycle regulation. In this regard, the etiopathological study of low-penetrant variants of BRCA1 provides an opportunity to uncover its other physiologically important functions. Using this rationale, we studied the R1699Q variant of BRCA1, a potentially moderate-risk variant, and found that it does not impair DNA damage repair but abrogates the repression of microRNA-155 (miR-155), a bona fide oncomir. Mechanistically, we found that BRCA1 epigenetically represses miR-155 expression via its association with HDAC2, which deacetylates histones H2A and H3 on the miR-155 promoter. We show that overexpression of miR-155 accelerates but the knockdown of miR-155 attenuates the growth of tumor cell lines in vivo. Our findings demonstrate a new mode of tumor suppression by BRCA1 and suggest that miR-155 is a potential therapeutic target for BRCA1-deficient tumors.

Figure 1

R1699Q mutant ES cells show reduced survival and differentiation defects. (a) Schematics of generation of R1699Q ES cells using PL2F8 cells containing a null and a conditional allele of Brca1. Two halves of human HPRT1 minigenes (HP and RT) flanking the two loxP sites (shaded triangles) of the conditional allele. Cre recombinants are HAT resistant (HAT^R). (b) Southern hybridization of HAT^R colonies from experiments without BAC (NO BAC), wild-type (WT) BRCA1 BAC and R1699Q BRCA1 BAC. Bottom band, null allele (MT); top band, conditional allele (cko). Rescue rate, percentage Brca1^ko/ko clones. Asterisk, Brca1^ko/ko ES cell. (c) Whole mount of embryoid bodies generated from ES cells expressing wild-type (left) and R1699Q (right) BRCA1 at day 14 in culture. Bottom, higher magnification of embryoid bodies. Scale bar, 50 μm, top; 20 μm, bottom. (d) H&E staining of embryoid bodies generated from ES cells expressing wild-type (left) and R1699Q (right) BRCA1 at day 14 in culture. Scale bar, 100 μm, top; 50 μm, bottom. (e) TUNEL staining of embryoid bodies. Arrow, TUNEL⁺ cells. Scale bar, 50 μm. (f) Teratoma growth of one wild-type and two R1699Q clones were examined in mice (n = 5 for each group). Values are means ± s.e.m. (P = 0.007). (g) H&E staining of teratomas dissected 15 d after injection. Top, section of the whole teratoma; middle, magnified images of the regions indicated at top. Arrows, neurorosette structures. Bottom, neural cells immature in wild-type (left) and more differentiated in R1699Q (right) teratomas. Scale bar, 2 μm, top; 0.2 μm, middle; 50 μm, bottom.

Figure 2

Identification of miR-155 upregulation in R1699Q mutant cells and its effect on ES cell differentiation. (a) Quantification of miR-155 by rtPCR in wild-type (WT), R1699Q (RQ) and M1652I (MI) ES cells and embryoid bodies (EB cells) on day 7 of culture. U6 small nuclear RNA (snRNA) was used for normalization. Top, BRCA1 protein expression. (b) Representative pictures of miR-155 in situ hybridization in wild-type and R1699Q embryoid bodies using DIG-labeled antisense LNA of miR-155. Top, no-probe controls showing background signal (scale bar, 50 μm). Right, relative average signal. Error bar = s.d. (c) Schematic of miR-155 inducible system in ES cells. miR-155 is induced by tetracycline, which then represses the luciferase reporter (luc) containing miR-155-binding sites at the 3′ end. (d) miR-155 overexpression in ES cells after tetracycline (tet) induction, with U6 snRNA as control. Top, overexpression by northern hybridization; bottom, quantification by rtPCR. Error bar = s.d. (e) Representative pictures of embryoid bodies generated from wild-type ES cells with induced expression of miR-155. Top, whole mount of embryoid bodies; bottom, H&E-stained sections. Scale bar, 100 μm. (f) Teratoma growth of wild-type ES cells with (Tet1–Tet5) and without (Con1–Con5) miR-155 induction. Left, teratoma growth. Right, average tumor volume on day 18 after injection. Minimum value of each group was excluded (n = 4; values are means ± s.e.m.). (g) Representative picture of teratomas in mice injected with control cells (−Tet, right) and Tet-induced cells (+Tet, left). Bottom, enlarged view.

Figure 3

BRCA1 negatively controls miR-155 expression. (a) Quantification of miR-155 in tumors from the Brca1^ko/+;Trp53^ko/+;Tg^R1699Q mice (RQ, n = 9). Normal liver (RQ93 nor), mammary gland (RQ515 nor) and tumors from Brca1^ko/+;Trp53^ko/+;Tg^M1652I mice (MI, n = 4) used as control. (b) Quantification of miR-155 in a breast tumor from an R1699Q mutation carrier. Normal, normal breast tissue. (c) Expression analysis of miR-155 by northern hybridization in four human breast cancer cell lines of different BRCA1 genotypes. Positive control, HEK293 cell line overexpressing miR-155; loading control, U6 snRNA; asterisk, miR-155 signal. Right, quantification of signal, normalized by U6 snRNA. (d) rtPCR of miR-155 in four Brca1 deficient tumors (42, 572, 907 and M161) from Brca1^cko/cko;p53^ko/+;MMTV Cre mice and four tumors (Con 112X, Con 172X, Con I107 and Con I224) from Her2/Neu transgenic mice. MCF7 and HCC1937 were negative and positive control, respectively. (e) rtPCR of miR-155 in MECs from two Brca1^cko/cko;K14 Cre mice. Control, MECs from K14 Cre only and Brca^cko/cko. (f) Knockdown of BRCA1 in two clones (75 and 80) of HEK293 cells stably expressing BRCA1 short hairpin RNA (top). Loading control, β-actin (middle). Bottom, rtPCR quantification of miR-155 in the BRCA1 knockdown clones (75 and 80). Con, control. (g) Real-time quantification of miR-155 after ectopic expression of wild-type (WT) or R1699Q BRCA1 in BRCA1-deficient MDA-MB-436 cells. Top, expression of wild-type and R1699Q BRCA1. (h) miR-155 reporter assay in BRCA1-deficient HCC1937 cells using increasing amounts of wild-type BRCA1 cDNA (Con, untransfected cells and BRCA50, 50 ng; BRCA100, 100 ng. Error bar = s.d.

Figure 4

Mechanism of miR-155 repression by BRCA1. (a) Binding of BRCA1 to miR-155 promoter at two potential regions (BRCA1-1 and BRCA1-2). ES cells and embryoid bodies (EB) at day 7 in culture with wild-type and R1699Q (RQ). BRCA1 by ChIP assay. Input, 5% of total chromatin. (b) Effect of mutation of putative BRCA1-binding sites (Mut1, Mut2 or double mutant) on miR-155 promoter activation measured by luciferase assay (*P < 0.05, **P < 0.01, ***P < 0.005). (c) Effect of HDAC inhibitors on miR-155 expression measured by rtPCR in BRCA1⁺ MDA-MB-468 (left) and BRCA1-deficient MDA-MB-436 cells (right). Con, no treatment; Aph, apicidin. (d) Effect of HDAC inhibitors on wild-type miR-155 promoter (left) and mutant promoter (right) activation in MDA-MB-468 cells measured by luciferase assay (*P < 0.02, **P < 0.01, ***P < 0.005). Con, no treatment; Aph, apicidin. (e) ChIP assay to quantify acetylation level of histones H2A and H3 on miR-155 promoter, in wild-type and R1699Q EB cells by rtPCR. (f) Effect of wild-type, RQ and wild-type + RQ BRCA1 on acetylation of histones H2A and H3 on miR-155 promoter in MDA-MB-436 cells measured by ChIP assay and rtPCR. Con, transfection with vector only (*P < 0.02, **P < 0.01). Western blot (WB) analysis of ectopic protein expression using antibody to hemagglutinin (HA). (g) Association of HDAC2 on miR-155 promoter in wild-type and R1699Q EB cells analyzed by ChIP assay and rtPCR. IgG, negative control. (h) Co-immunoprecipitation (Co-IP) between HDAC2 and wild-type or M1652I or R1699Q BRCA1 in MECs from mice with wild-type or R1699Q BAC transgenes (left) or M1652I transgene (right). Top panels, IP with HDAC2 antibody; bottom panels IP with human-specific BRCA1 antibody. (i) MDA-MB 468 cells transfected with luciferase reporter plasmid with mouse wild-type or double mutant (MT) miR-155 promoter (pro). ChIP assay was carried out with indicated antibodies (Ab). Association of BRCA1-HDAC2 complex and acetylation of histones H2A and H3 on transfected mouse miR-155 promoter (mir155, top) and endogenous miR-155 promoter (MIR155 to test competition effect. BRCA1 binding on ESRRG, CCNB1 and STAT5A promoters with putative binding sites. Error bar = s.d.

Figure 5

Physiological relevance of miR-155 upregulation in BRCA1-deficient tumors. (a) Xenograft tumor growth of MDA-MB-468 cells with stable expression of miR-155 (P = 4.12 × 10⁻⁴, ANCOVA test). Values are mean ± s.e.m. (b) Growth of two clones (C9 and D6) with miR-155 stable knockdown and the parental cells (Con) orthotropically transplantated in mice (values are means ± s.e.m., *P = 0.031, **P = 0.025). (c) Average mass of tumors in b (values are means ± s.e.m.). (d) Representative pictures of 66 human breast tumors in tumor tissue microarray probed with antibody to BRCA1 (Ab-1) or mouse IgG (negative control) and DIG-labeled antibody to miR-155. Scale bar, 50 μm. (e) rtPCR quantification of miR-155 in 28 tumors from non-BRCA1 controls and 14 tumors from BRCA1 mutation (mut) carriers (see Supplementary Table 6 for mutation description). RNU5A was used for normalization. Shaded area, low miR-155 based on two-fold cut-off (dashed horizontal line). Text box, number of miR-155 high and low tumors in each group. (f) Schematic of role for BRCA1 in epigenetic control of miR-155 promoter. In wild-type BRCA1–containing cells, miR-155 is silenced by the BRCA1-HDAC2 complex, which deacetylates H2A and H3 histones. Without functional BRCA1, the interaction of BRCA1-HDAC2 complex is disrupted, which in turn increases acetylated (Ac) H2A and H3, which activates the miR-155 promoter.