Prolactin, interleukin-2 and FGF-2 stimulate expression, nuclear distribution and DNA-binding of rat homolog of pombe Cdc5 in Nb2 T lymphoma cells

Prolactin signaling through the short isoform of the mouse prolactin receptor regulates DNA binding of specific transcription factors, often with opposite effects in different reproductive issues

BACKGROUND

It has been well established that prolactin (PRL) signals through the long form of its receptor (PRL-RL) and activates the Jak/Stat pathway for transcription of PRL target genes. However, signaling pathways mediated through the short PRL-R isoform (PRL-RS) remains controversial. Our recent finding that PRL signaling through PRL-RS represses two transcription factors critical for follicular development lead us to examine other putative PRL/PRL-RS target transcription factors in the decidua and ovary, two well-known target tissues of PRL action in reproduction.

METHODS

In this investigation we used mice expressing PRL-RS on a PRL-R knockout background and a combo protein/DNA array to study the transcription factors regulated by PRL through PRL-RS only.

RESULTS

We show that PRL activation of the PRL-RS receptor either stimulates or inhibits the DNA binding activity of a substantial number of transcription factors in the decidua as well as ovary. We found few transcription factors to be similarly regulated in both tissues, while most transcription factors are oppositely regulated by PRL in the decidua and ovary. In addition, some transcription factors are regulated by PRL only in the ovary or only in the decidua. Several of these transcription factors are involved in physiological pathways known to be regulated by PRL while others are novel.

CONCLUSION

Our results clearly indicate that PRL does signal through PRL-RS in the decidua as well as the ovary, independently of PRL-RL, and activates/represses transcription factors in a tissue specific manner. This is the first report showing PRL/PRL-RS regulation of specific transcription factors. Many of these transcription factors were not previously known to be PRL targets, suggesting novel physiological roles for this hormone.

Cell cycle-dependent phosphorylation of human CDC5 regulates RNA processing

CDC5 proteins are components of the pre-mRNA splicing complex and essential for cell cycle progression in yeast, plants and mammals. Human CDC5 is phosphorylated in a mitogen-dependent manner, and its association with the spliceosome is ATP-dependent. Examination of the amino acid sequence suggests that CDC5L may be phosphorylated at up to 28 potential consensus recognition sequences for known kinases, however, the identity of actual phosphorylation sites, their role in regulating CDC5L activity, and the kinases responsible for their phosphorylation have not previously been determined. Using two-dimensional phosphopeptide mapping and nanoelectrospray mass spectrometry, we now show that CDC5L is phosphorylated on at least nine sites in vivo. We demonstrate that while CDC5L is capable of forming homodimers in vitro and in vivo, neither homodimerization nor nuclear localization is dependent on phosphorylation at these sites. Using an in vitro splicing assay, we show that phosphorylation of CDC5L at threonines 411 and 438 within recognition sequences for CDKs are required for CDC5L-mediated pre-mRNA splicing. We also demonstrate that a specific inhibitor of CDK2, CVT-313, inhibits CDC5L phosphorylation in both in vitro kinase assays and in vivo radiolabeling experiments in cycling cells. These studies represent the first demonstration of a regulatory role for phosphorylation of CDC5L, and suggest that targeting these sites or the implicated kinases may provide novel strategies for treating disorders of unguarded cellular proliferation, such as cancer.

Characterization of a novel, cytokine-inducible carboxypeptidase D isoform in haematopoietic tumour cells

CPD-N is a cytokine-inducible CPD (carboxypeptidase-D) isoform identified in rat Nb2 T-lymphoma cells. The prototypic CPD (180 kDa) has three CP domains, whereas CPD-N (160 kDa) has an incomplete N-terminal domain I but intact domains II and III. CPD processes polypeptides in the TGN (trans-Golgi network) but the Nb2 CPD-N is nuclear. The present study identified a cryptic exon 1', downstream of exon 1 of the rat CPD gene, as an alternative transcription start site that encodes the N-terminus of CPD-N. Western-blot analysis showed exclusive synthesis of the 160 kDa CPD-N in rat Nb2 and Nb2-Sp lymphoma cells. Several haematopoietic cell lines including human K562 myeloma, Jurkat T-lymphoma and murine CTLL-2 cytotoxic T-cells express a 160 kDa CPD-immunoreactive protein, whereas mEL4 T-lymphoma cells express the 180 kDa CPD. The CPD-immunoreactive protein in hK562 cells is also nuclear and cytokine-inducible. In contrast, MCF-7 breast cancer cells express only the 180 kDa CPD, which is mainly in the TGN. CPD/CPD-N assays using substrate dansyl-L-alanyl-L-arginine show approx. 98% of CPD-N activity in the Nb2 nucleus, whereas MCF-7 CPD activity is enriched in the post-nuclear 10000 g pellet. The K(m) for CPD-N and CPD are 132+/-30 and 63+/-9 microM respectively. Specific activity/K(m) ratios show that dansyl-L-alanyl-L-arginine is a better substrate for CPD-N than for CPD. CPD-N has an optimal pH of 5.6 (due to domain II), whereas CPD has activity peaks at pH 5.6 (domain II) and pH 6.5-7.0 (domain I). CPD and CPD-N are inhibited non-competitively by zinc chelator 1,10-phenanthroline and competitively by peptidomimetic inhibitor DL-2-mercaptomethyl-3-guanidinoethylthiopropanoic acid. The Nb2 CPD-N co-immunoprecipitated with phosphatase PP2A (protein phosphatase 2A) and alpha4 phosphoprotein. In summary, a cytokine-inducible CPD-N is selectively expressed in several haematopoietic tumour cells. Nuclear CPD-N is enzymatically active and interacts with known partners of CPD.

Role of O-linked beta-N-acetylglucosamine modification in the subcellular distribution of alpha4 phosphoprotein and Sp1 in rat lymphoma cells

The mTOR alpha4 phosphoprotein is a prolactin (PRL)-downregulated gene product that is found in the nucleus of PRL-dependent rat Nb2 lymphoma cells. Alpha4 lacks a nuclear localization signal (NLS) and the mechanism of its nuclear targeting is unknown. Post-translational modification by O-linked beta-N-acetylglucosamine (O-GlcNAc) moieties has been implicated in the nuclear transport of some proteins, including transcription factor Sp1. The nucleocytoplasmic enzymes O-beta-N-acetylglucosaminyltransferase (OGT) and O-beta-N-acetylglucosaminidase (O-GlcNAcase) adds or remove O-GlcNAc moieties, respectively. If O-GlcNac moieties contribute to the nuclear targeting of alpha4, a decrease in O-GlcNAcylation (e.g., by inhibition of OGT) may redistribute alpha4 to the cytosol. The present study showed that alpha4 and Sp1 were both O-GlcNAcylated in quiescent and PRL-treated Nb2 cells. PRL alone or PRL + streptozotocin (STZ; an O-GlcNAcase inhibitor) significantly (P <or=.05) increased the O-GlcNAc/alpha4 ratio above that in control quiescent cells. However, PRL + alloxan (ALX; an OGT inhibitor) or ALX alone did not decrease O-GlcNAcylation of alpha4 below that of controls and alpha4 remained nuclear. In comparison, PRL (+/-ALX/STZ) greatly increased Sp1 protein levels, caused a significant decrease in the GlcNAc/Sp1 ratio (P <or=0.05, n = 3) as compared to controls and partially redistributed Sp1 to the cytosol. Finally, a 50% downregulation of OGT gene expression by small interfering RNA (i.e., siOGT) partially redistributed both alpha4 and Sp1 to the cytosol. The alpha4 protein partner PP2Ac had no detectable O-GlcNAc moieties and its nuclear distribution was not affected by siOGT. In summary, alpha4 and Sp1 contained O-GlcNAc moieties, which contributed to their nuclear targeting in Nb2 cells.