AZU-1: a candidate breast tumor suppressor and biomarker for tumor progression

Abstract

To identify genes misregulated in the final stages of breast carcinogenesis, we performed differential display to compare the gene expression patterns of the human tumorigenic mammary epithelial cells, HMT-3522-T4-2, with those of their immediate premalignant progenitors, HMT-3522-S2. We identified a novel gene, called anti-zuai-1 (AZU-1), that was abundantly expressed in non- and premalignant cells and tissues but was appreciably reduced in breast tumor cell types and in primary tumors. The AZU-1 gene encodes an acidic 571-amino-acid protein containing at least two structurally distinct domains with potential protein-binding functions: an N-terminal serine and proline-rich domain with a predicted immunoglobulin-like fold and a C-terminal coiled-coil domain. In HMT-3522 cells, the bulk of AZU-1 protein resided in a detergent-extractable cytoplasmic pool and was present at much lower levels in tumorigenic T4-2 cells than in their nonmalignant counterparts. Reversion of the tumorigenic phenotype of T4-2 cells, by means described previously, was accompanied by the up-regulation of AZU-1. In addition, reexpression of AZU-1 in T4-2 cells, using viral vectors, was sufficient to reduce their malignant phenotype substantially, both in culture and in vivo. These results indicate that AZU-1 is a candidate breast tumor suppressor that may exert its effects by promoting correct tissue morphogenesis.

Figure 1

The AZU-1 gene is differentially expressed in nonmalignant and tumorigenic human breast cells. Northern blot analysis was performed on total RNA (20 μg/lane) from breast cell and tissue extracts using ³²P-labeled AZU-1-specific probes. (A) Comparison of AZU-1 expression in S2 and T4-2 cells detected with the 180-bp differential display cDNA probe. Two lower-abundance transcripts are indicated by small arrows; the presence of these bands was not always reproducible. (B) AZU-1 expression in normal primary luminal epithelial and myoepithelial cells and in nonmalignant breast cell lines HMT-3522-S1 and MCF10A. (C) Compared with S2 cells (lane 1), AZU-1 expression is reduced in a number of breast carcinoma cell lines: lane 2, T4-2; lane 3, HMT-3909; lane 4, MCF-7; lane 5, CAMA-1; lane 6, BT-20; lane 7, MDA-MB-468; lane 8, SKBR-3; lane 9, T47D; lane 10, MDA-MB-231; lane 11, Hs578T; and lane 12, BT549. *, HMT-3909 cells display partial myoepithelial differentiation (O.W. Petersen, unpublished result). (D) AZU-1 expression in tissues derived from normal breast (lane 1) and three carcinomas in situ (lanes 2–4). For B and C, an AZU-1 coding region probe was used; in all cases, a GAPDH probe was used as a loading control.

Figure 2

Sequence and structure of AZU-1. (A) Deduced amino acid sequence of the AZU-1 571-amino-acid open reading frame. Four structural domains, labeled SPAZ, region I, region II, and CCD, are boxed, and two predicted NLS motifs are underlined. The N-terminal peptide used to generate the AZU-1 antibody is highlighted in gray. The HATDEEKLA sequence, a peptide conserved between AZU-1 and TACC1, appears in black. (B) Domain organization of AZU-1 and two AZU-1-related genes, TACC1 and TACC3. Based on its similarity with TACC1 and TACC3, AZU-1 can be partitioned into four domains: 1) the N-terminal SPAZ domain, 2) region I, a region that shares a moderate sequence similarity with TACC1 and to a lesser extent with TACC3, 3) region II, which is totally absent in TACC1 and partially removed from TACC3, and 4) the C-terminal coiled-coil domain. (C) Sequence alignments of SPAZ domains from AZU-1, TACC1 (2 copies, a and b), TACC3, and BCK1 from S. cerevisiae. Residues that are conserved in three or more of these sequences appear in black; the corresponding columns are marked with open circles. Two invariant serine residues are indicated by filled circles. Fold recognition analyses predict that SPAZ domains adopt Ig-like folds. (D) CCD sequence alignments of AZU-1, TACC1, TACC3, and SB1.8/DXS423E. Amino acid identities observed in two or more of the aligned sequences are indicated in black; in cases in which two pairs of identical amino acids are observed in the alignment, AZU-1-like sequences are preferentially highlighted. The CCD heptad repeat positions, a–g, are indicated in brackets above the three regions where all four proteins fall into register. Positions a and d, often occupied by hydrophobic residues, are indicated in capital letters. Sequence identities among all four proteins in this region are most notable in the second half of the CCD.

Figure 3

The AZU-1 protein migrates with an apparent molecular mass of 80 kDa on SDS-polyacrylamide gels. (A) In vitro transcription and translation reactions were performed with [³⁵S]methionine in the absence (lane 1) or presence of luciferase cDNA (lane 2, positive control at 61 kDa) or AZU-1 cDNA (lanes 3–5). In lanes 1–3, 5 μl of whole lysate were loaded in each lane. The remaining AZU-1 lysate was immunoprecipitated with either preimmune (lane 4) or AZU-1-specific (lane 5) rabbit sera, and the precipitated samples were loaded into adjacent wells. The resolved protein products were analyzed by autoradioagraphy. The AZU-1 cDNA gives rise to a single predominant protein with an apparent molecular mass of 80 kDa. (B) Protein extracts from S1 and T4-2 monolayer cultures (20 μg/lane) were analyzed by Western immunoblotting using preimmune (lanes 1 and 2) or anti-AZU-1 (lanes 1′ and 2′) rabbit sera. E-cadherin antibodies were used to control for protein loading. Like the in vitro-translated protein, cellular AZU-1 migrates with an apparent molecular mass of 80 kDa by SDS-PAGE. On average, T4-2 cells exhibit a threefold reduction in AZU-1 protein levels in comparison with nonmalignant S1 cells.

Figure 4

AZU-1 is a predominantly cytoplasmic protein in S1 and T4-2 cells. After 4 d in culture, cell monolayers were either directly fixed with 2% paraformaldehyde (A–D) or permeabilized with Triton X-100 before fixation (E and F). Cells were immunostained with affinity-purified anti-AZU-1 polyclonal antibody (A, D, E, and H) or with an equivalent amount of purified rabbit IgG (B and F). Primary antibodies were detected using an FITC-conjugated secondary antibody. F-actin was visualized in S1 cells using FITC-phalloidin. Confocal images in A, C–E, G, and H show a 0.4-μm optical section through the center of the cell nuclei. In both S1 and T4-2 cells, AZU-1 is found primarily in the cell cytoplasm, albeit at generally lower levels in the T4-2 cells (see arrow in D for typical T4-2 expression pattern). In both cells, the cytoplasmic pool of AZU-1 is detergent extractable, indicating that AZU-1 is not likely to be tightly associated with the insoluble cytoskeleton. (F-actin was monitored as a positive indicator of detergent resistance.) A minor, detergent-resistant pool of AZU-1 is found throughout nuclei in dim speckles as well as in distinct subnuclear foci. All images were recorded at 120× magnification.

Figure 5

Reexpression of AZU-1 in T4-2 cell reduces their tumorigenicity in vitro. Northern (A) and Western (B) blot analyses were performed to monitor AZU-1 levels in S1, T4-2 control cells, and AZU-1-infected T4-2 cells. AZU-1 expression is increased at both the RNA and protein levels upon introduction of the AZU-1 transgene into T4-2 cells (in both cases approximately two- to threefold). A GAPDH probe and an E-cadherin antibody were used as loading controls in Northern and Western blots, respectively. In vitro tumorigenicity of the various HMT-3522 cells was measured in soft agar assays (C) and in invasion assays (D). In both cases, overexpression of AZU-1 in T4-2 cells gave rise to reduced tumorigenic behavior (i.e., reduced anchorage-independent growth and reduced capacity to migrate through a basement membrane-like gel). The data presented here represent the averages of three independent experiments and correspond to the mean activity of triplicate measurements ± SE.

Figure 6

Increased AZU-1 expression levels correlate with phenotypic reversion in 3D rBM assays. (A) AZU-1 induces phenotypic reversion. S1, T4-2 (vector-infected), and T4-2-AZU-1 cells were embedded as single cells in 3D rBM assays. After 10 d in culture, the colonies were measured (expressed as colony diameter in micrometers ± SE) and imaged using phase microscopy (a, b, and e). Cultures were immunostained with antibodies specific for collagen IV (c and f) or β4 integrin (d and g). (B) AZU-1 is reexpressed upon EGFR- and β1 integrin-induced phenotypic reversion. (a) S1 and T4-2 cells were cultured in 3D rBM assays in the absence or presence of functional inhibitors of β1 integrin (T4-2β1) or EGFR (T4-2tyr; tyr, tyrphostin). Unlike control cells, inhibitor-treated T4-2 cells exhibit an S1-like, acinar phenotype in 3D cultures. (b) Total RNA harvested from these cultures was analyzed in Northern blots using an AZU-1-specific probe. GAPDH was used as a loading control. AZU-1 expression is restored to S1-like levels in T4-2 cells that have undergone phenotypic reversion in the 3D rBM assay. Bars, 50 μm.