DEAR1 is a dominant regulator of acinar morphogenesis and an independent predictor of local recurrence-free survival in early-onset breast cancer

Abstract

Background: Breast cancer in young women tends to have a natural history of aggressive disease for which rates of recurrence are higher than in breast cancers detected later in life. Little is known about the genetic pathways that underlie early-onset breast cancer. Here we report the discovery of DEAR1 (ductal epithelium-associated RING Chromosome 1), a novel gene encoding a member of the TRIM (tripartite motif) subfamily of RING finger proteins, and provide evidence for its role as a dominant regulator of acinar morphogenesis in the mammary gland and as an independent predictor of local recurrence-free survival in early-onset breast cancer.

Methods and findings: Suppression subtractive hybridization identified DEAR1 as a novel gene mapping to a region of high-frequency loss of heterozygosity (LOH) in a number of histologically diverse human cancers within Chromosome 1p35.1. In the breast epithelium, DEAR1 expression is limited to the ductal and glandular epithelium and is down-regulated in transition to ductal carcinoma in situ (DCIS), an early histologic stage in breast tumorigenesis. DEAR1 missense mutations and homozygous deletion (HD) were discovered in breast cancer cell lines and tumor samples. Introduction of the DEAR1 wild type and not the missense mutant alleles to complement a mutation in a breast cancer cell line, derived from a 36-year-old female with invasive breast cancer, initiated acinar morphogenesis in three-dimensional (3D) basement membrane culture and restored tissue architecture reminiscent of normal acinar structures in the mammary gland in vivo. Stable knockdown of DEAR1 in immortalized human mammary epithelial cells (HMECs) recapitulated the growth in 3D culture of breast cancer cell lines containing mutated DEAR1, in that shDEAR1 clones demonstrated disruption of tissue architecture, loss of apical basal polarity, diffuse apoptosis, and failure of lumen formation. Furthermore, immunohistochemical staining of a tissue microarray from a cohort of 123 young female breast cancer patients with a 20-year follow-up indicated that in early-onset breast cancer, DEAR1 expression serves as an independent predictor of local recurrence-free survival and correlates significantly with strong family history of breast cancer and the triple-negative phenotype (ER(-), PR(-), HER-2(-)) of breast cancers with poor prognosis.

Conclusions: Our data provide compelling evidence for the genetic alteration and loss of expression of DEAR1 in breast cancer, for the functional role of DEAR1 in the dominant regulation of acinar morphogenesis in 3D culture, and for the potential utility of an immunohistochemical assay for DEAR1 expression as an independent prognostic marker for stratification of early-onset disease.

Figure 1. DEAR1 structure, mapping, and expression in normal tissues.

(A) Chromosomal localization of DEAR1 as determined by FISH analysis using the DEAR1 P1-derived artificial chromosome (PAC) clone. (B) Graphical representation of DEAR1 exonic and protein structure. (C) DEAR1 multiple tissue Northern analysis detects a predominant 4.4 kb band in all tissues examined. Additional, lower molecular weight bands were observed in a number of tissues, including heart, placenta, skeletal muscle and brain. (D) DEAR1 peptide competition with 5× peptide specifically detects the predicted 54 kDa full-length protein in the immortalized HMEC line 76N-E6. (E) Transient transfection of HA-tagged DEAR1 into 293T cells (which do not express endogenous DEAR1) detects the appropriate sized protein. (F) Western blot analysis of normal tissue protein lysates using the α-N DEAR1 antibody identifies a strong band of approximately 54 kDa corresponding to the predicted full-length DEAR1 protein molecular weight. (G) Localization of DEAR1 protein in normal tissue assessed by immunohistochemistry using the α-N DEAR1 antibody on a multiple tissue microarray. Staining (dark brown, identified by arrow) is plainly visible in epithelial cells found in a wide range of tissues, including (i) bladder, (ii) gall bladder, (iii) kidney, (iv) prostate, (v) pancreas, and (vi) salivary gland.

Figure 2. Down-regulation of DEAR1 in breast cancer cell lines and in transition to DCIS in the breast epithelium.

(A) and (B) show immunohistochemical staining of two examples from 14 cases for which normal ductal structures, DCIS, and invasive carcinoma from the same individual are located within the same histologic section. Normal ducts are indicated by solid arrows, and representative foci of DCIS are indicated by an open arrowhead. Immunohistochemical staining using the α-N DEAR1 antibody appears as a dark brown precipitate. Panel (A) indicates (i) intense staining of DEAR1 in normal mammary ducts; (ii) diffuse, low level staining of DEAR1 observed in this single focus of DCIS. Note the slight increase in DEAR1 staining toward the center of the focus; (iii) diffuse, low level staining of DEAR1 is observed throughout much of this region composed of invasive carcinoma. Panel (B) shows intense staining of DEAR1 noted in the normal duct, with a dramatic decrease in expression in adjacent foci of DCIS. (C) DEAR1 expression on Western blot analysis of HMEC cultures (normal HMECs and immortalized HMECs 76N-E6 and 76N-F2v) and breast carcinoma cell lines.

Figure 3. Mutation and microdeletion analysis of DEAR1.

(A) Direct genomic sequencing identified a codon 187 missense mutation (C→T) in exon 3 in the 21MT cell line but not in the cell line H16N-2, derived from the normal mammary epithelium from the same patient. (B) A missense mutation in codon 473 of exon 5 (GTC→ATC, V473I) detected in a breast tumor sample as well as adjacent normal tissue, but not in the normal lymph node from this individual, indicating that the sequence alteration in the tumor was a somatic mutation of the DEAR1 sequence. (C) Diagram of genomic structure and core promoter and exon 1 of DEAR1 indicating the location of assays and primers by which HD in tumor 9BT was identified (indicated by *) as well as those used for deletion mapping in DEAR1 and flanking genes. (D) Schematic of homozygous deletion in 9BT. (E) STS mapping analysis indicates retention of MS1, deletion of MS2, and retention of MS3 in primary tumor sample (9BT).

Figure 4. Introduction of DEAR1 mediates acinar morphogenesis in 3D culture.

21MT, control 21MT/Δ187, and wild-type transfectant 21MT/J and 21MT/L analyzed (A) by quantitative RT-PCR and (B) in 3D culture for the percentage of acinar structures. (C) Propidium (red)-staining structures were photographed by confocal microscopy after 11 d in 3D culture. The lumen can be clearly seen in the DIC photomicrograph to the right of the fluorescent image. (D) Confocal images of 21MT, 21MT/Δ, and wild-type transfectant 21MT/J and 21MT/L (i) at low magnification (bar = 200 µm) illustrating the dramatic size differences in acini from transfectants with and without wild-type DEAR1 and compared with 21MT cells; (ii) after staining with propidium (red), and E-cadherin (green) discriminated the basal orientation of nuclei and expression of E-cadherin at cell–cell contacts in wild-type transfectant structures propagated in 3D culture as compared with the large, disorganized apolar structures in 21MT and 21MT/Δ cells (bar = 100 µm); (iii) introduction of wild-type DEAR1 into 21MT cells resulted in acinar morphogenesis with epithelial cells surrounding a lumen illustrated by staining with propidium (blue), which denotes basal orientation of nuclei, basal orientation of alpha-6-integrin (red), and increased caspase 3 (green) staining in luminal structures in wild-type transfectants as opposed to 21MT and 21MT/Δ.

Figure 5. DEAR1 is a dominant regulator of acinar morphogenesis in HMECs.

(A) Western analysis of shRNA control clones (C1 and C2) and shRNA knockdown clones (sh1, sh2, and sh3). (B) Confocal images of 3D cultures of control clones (C1 and C2) and DEAR1-knockdown clones (sh1, sh2, and sh3) showing representative acinus stained with alpha6-integrin (red), caspase 3 (green), or DAPI (blue), which shows the clear lumen in controls as opposed to shRNA knockdown clones (B i, ii, and iii are results at day 16; iv is at day 22).

Figure 6. DEAR1 is an independent predictor of local recurrence free survival in early onset breast cancer.

Immunohistochemical staining of an early onset tissue array resulted in a significant correlation between the expression of DEAR1 and the probability of local recurrence free survival (p = 0.0334). At 15 y post diagnosis, recurrence-free survival in DEAR1-negative patients was 58% compared to 95% in those patients whose tumors were positive for DEAR1 expression.