Transcription variants of the prostate-specific PrLZ gene and their interaction with 14-3-3 proteins

Abstract

We have reported isolation and characterization of the prostate-specific and androgen-regulated PrLZ gene abnormally expressed in prostate cancer. PrLZ is a potential biomarker for prostate cancer and a candidate oncogene promoting cell proliferation and survival in prostate cancer cells. A full delineation of the PrLZ gene and its gene products may provide clues to the mechanisms regulating its expression and function. In this report, we identified three additional exons in the PrLZ gene and recognized five transcript variants from alternative splicing that could be detected by RT-PCR and Western blotting. Structural comparison demonstrated that the PrLZ proteins are highly conserved among species. PrLZ contains multiple potential sites for interaction with other proteins. We used mammalian two-hybrid assays to demonstrate that PrLZ isoforms interact with 14-3-3 proteins, and multiple sites in the PrLZ may be involved in the interaction. Alternative splicing may contribute to abnormally enhanced PrLZ levels in prostate cancer, and interaction with 14-3-3 proteins may be a mechanism by which PrLZ promotes cell proliferation and survival during prostate cancer development and progression. This information is a valuable addition to the investigation of the oncogenic properties of the PrLZ gene.

Figure 1. Structure of the PrLZ gene and alternatively spliced isoforms

A, schematic presentation of PrLZ gene organization. Exons are shown as vertical bars and coding regions are shaded in gray. The translation start site is marked (ATG) and two in frame stop codons are marked with asterisks (*). Note that sizes of the exons are not in scale. B, schematic presentation of the PrLZ cDNA sequences due to alternative splicing. Two polyadenylation signals (AATAAA) are marked. On top of each scheme the location of primers used for RT-PCR detection of specific transcripts is marked with a horizontal bar. C, RT-PCR detection of the alternatively spliced transcripts in C4-2 prostate cancer cells. Specific transcripts detected are PrLZ-151 (lane 1), PrLZ 233 (lane 2), and PrLZ-238 (lane 3). GAPDH was used as control (lane 4). Commercial λHindIII/ΦX174HaeIII was used as a DNA marker (M). D, Western blot detection of PrLZ isoforms. In the upper panel, whole cell lysates (40 μg) of LNCaP and C4-2 cells treated with R1881 (+) were subjected to western blotting for PrLZ. Sizes (kD) of the bands were indicated on the right. The result is representative of two separate experiments. In the lower panel, the level of the β-actin was used as a loading control.

Figure 2. Amino acid sequences of the PrLZ isoforms and comparison with other TPD52 family members

The upper five lines show structural differences between the PrLZ isoforms. The lower nine lines sequence conservation among TPD52 family members. Compared to PrLZ-247, an identical residue is marked with a dot. A gap is introduced to ensure the comparison reaches maximal homology. Along the PrLZ-247 sequence, the two PEST domains are underlined and the leucine zipper domain is marked by underlining the repetitive leucine (L)/isoleucine (I) residues in bold. Also underlined and in bold are consensus sites for tyrosine (Y) phosphorylation, serine (S)/threonine (T) phosphorylation, and N-glycosylation (N).

Figure 3. Expression of 14-3-3 genes in human prostate cancer cell lines

Expression of the seven 14-3-3 genes was detected with gene-specific primer pairs in RT-PCR analysis. Commercial total RNA of a normal human prostate (Prostate) was used as control. Representative results from two experiments are shown.

Figure 4. Interaction between PrLZ and 14-3-3 proteins as detected with mammalian two-hybrid assay

A, the assay was calibrated with empty vectors of the two-hybrid system and with kit controls (pACT-MyoD and pBIND-Id). B, results of the PrLZ isoforms in pACT co-transfected with each 14-3-3 cDNA in pBIND. TPD52, the prototype of the TPD52 family, was included in the study. Similar results were obtained in a complementary experiment, when PrLZ isoforms in pBIND was examined for interaction with 14-3-3 in pACT (data not shown). When transfected individually, neither the PrLZ nor 14-3-3 constructs induced significant luciferase activity from the pG5-luc reporter (data not shown). Each data represent the mean of triplicate assays. Standard deviation was less than 5% of the mean and is not shown.