This Article Abstract Full Text (PDF) All Versions of this Article: 28/2/165    most recent Final Manuscript Author Manuscript Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vlotides, G. Articles by Melmed, S. Search for Related Content PubMed PubMed Citation Articles by Vlotides, G. Articles by Melmed, S.

Pituitary Tumor-Transforming Gene: Physiology and Implications for Tumorigenesis

Department of Medicine, Cedars-Sinai Medical Center, University of California School of Medicine, Los Angeles, California 90048

Correspondence: Address all correspondence and requests for reprints to: Shlomo Melmed, Academic Affairs, Room 2015, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, California 90048. E-mail: melmed{at}csmc.edu

	Abstract

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
Pituitary tumor-transforming gene-1 (PTTG1) is overexpressed in a variety of endocrine-related tumors, especially pituitary, thyroid, breast, ovarian, and uterine tumors, as well as nonendocrine-related cancers involving the central nervous, pulmonary, and gastrointestinal systems. Forced PTTG1 expression induces cell transformation in vitro and tumor formation in nude mice. In some tumors, high PTTG1 levels correlate with invasiveness, and PTTG1 has been identified as a key signature gene associated with tumor metastasis. Increasing evidence supports a multifunctional role of PTTG1 in cell physiology and tumorigenesis. Physiological PTTG1 properties include securin activity, DNA damage/repair regulation and involvement in organ development and metabolism. Tumorigenic mechanisms for PTTG1 action involve cell transformation and aneuploidy, apoptosis, and tumorigenic microenvironment feedback. This paper reviews recent advances in our understanding of PTTG1 structure and regulation and addresses known mechanisms of PTTG1 action. Recent knowledge gained from PTTG1-null mouse models and transgenic animals and their potential application to subcellular therapeutic targeting PTTG1 are discussed.

I. Introduction

II. PTTG1 Gene Structure and Regulation

A. PTTG1 gene structure

B. PTTG1 mRNA expression profile

C. Regulation of gene expression

III. PTTG Protein

A. PTTG1 protein structure

B. PTTG1 protein subcellular localization and complex formation

C. PTTG1 protein posttranscriptional modification

IV. PTTG Family Members

A. Cloning and characterization of PTTG2 and -3

B. PTTG family member expression profile

V. PTTG1: Physiological Functions

A. Securin function/replication

B. DNA damage/repair

C. Interaction partners and transactivation activity

VI. PTTG1: Tumorigenic Mechanisms

A. PTTG1 and cell proliferation

B. Cell transformation and aneuploidy

C. Apoptosis

D. Tumorigenic microenvironment

VII. Pttg1-Null Mouse Model

VIII. PTTG1 and Cancer

A. Endocrine-related cancer

B. Nonendocrine-related cancer

C. Potential therapeutic applications

IX. Summary

	I. Introduction

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
PITUITARY TUMOR-transforming gene (PTTG) was isolated from rat pituitary tumor cells in 1997 and identified as a pituitary-derived transforming gene (1). Structural homology led to the identification of PTTG1 protein as a vertebrate securin critical in regulating sister chromatid separation during mitosis (2). PTTG1 overexpression has been reported in a variety of endocrine-related tumors, especially pituitary, thyroid, breast, ovarian, and uterine tumors, as well as nonendocrine-related cancers involving the central nervous system, pulmonary system, and gastrointestinal system (3, 4, 5, 6, 7, 8, 9, 10). PTTG1 levels correlate with tumor invasiveness (10, 11), and PTTG1 has been identified as a key signature gene associated with tumor metastasis (12). PTTG1 functions in cell replication (2), DNA damage/repair (13), organ development, and metabolism (14, 15, 16). Partially elucidated mechanisms of PTTG1 action include protein-protein regulation, transactivation activity, and paracrine/autocrine regulation.

In this review, we describe the structure and regulation of the PTTG1 gene and protein and address the known PTTG1 physiological functions. We review current information on PTTG1-mediated tumorigenic mechanisms, including cell transformation and aneuploidy, apoptosis, and tumorigenic microenvironment feedback. We describe the knowledge gained from Pttg1-null mouse models and transgenic animals and review the known impact of PTTG1 on endocrine and nonendocrine-related neoplasms, focusing on potential subcellular therapeutic applications for these tumors.

	II. PTTG1 Gene Structure and Regulation

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
A. PTTG1 gene structure
Pttg1 was isolated from rat GH-secreting pituitary cell lines by differential mRNA display PCR (1), yielding two pituitary-specific mRNAs, one of which was a differentially expressed, 396-bp mRNA that showed no homology to current GenBank entries. Using this PCR fragment to probe a rat pituitary tumor cDNA library enabled the cloning and characterization of the full-length (974 bp) rat Pttg1 cDNA clone (GenBank accession no. U73030) (1). Rat Pttg1 gene is composed of five exons and four introns (17), sized about 0.8, 1.8, 1.1, and 0.6 kb, respectively (17) (Fig. 1). Northern analysis revealed two mRNA isoforms of this gene; a 1.3-kb isoform strongly expressed in pituitary tumor cells, whereas rat testis and embryonic liver expressed a truncated Pttg1 mRNA isoform of approximately 1 kb (1). In both pituitary and testicular cells, the transcription start site for rat Pttg1 was mapped to a thymidine residue 44 bp upstream from the ATG initiation codon (17), suggesting that the different testis transcript size is not a result of alternative promotion. A minimal 745 bp of the 5'-flanking sequences (Fig. 2) are required for Pttg1 transcriptional activation in rat testicular cell lines, and the sequence between –509 and –624 bp was identified as the core enhancer sequence critical for Pttg1 transcriptional activation (17).

View larger version (8K): [in this window] [in a new window]   FIG. 1. Schematic illustration of mammalian PTTG1 gene structure. Exons are indicated as boxes (translated sequences in black and untranslated regions in white), whereas the connecting lines represent the introns. Exon and intron sizes (bp) are indicated. To simplify comparison between species, the translation initiation sites (ATG) have been set at +1. Major TSS are indicated relative to the ATG. The lower panel represents a scale from –1000 to +7000 bp. This combined information was obtained from original published sequences (1 17 18 23 25 ) and published sequences in the University of California, Santa Cruz, genome browser (http://genome.ucsc.edu/). It should be noted that, due to discrepancies in the information available from the aforementioned sources, the depicted sizes should be viewed as a guide.

View larger version (47K): [in this window] [in a new window]   FIG. 2. Sequence alignment of human, mouse, and rat PTTG1 5'-flanking region. 5'-Flanking region sequences of human (h), mouse (m), and rat (r) PTTG1 were obtained from GenBank (accession no. F200719, AF060887/AF071209, and AF021802, respectively). Nucleotide sequence alignment was obtained with MegaAlign (DNAStar, Madison, WI) by the clustal method. Differing nucleotides are highlighted.

The hPTTG1 gene is located on chromosome 5 [5q33 (18, 22), 5q35.1 (23)] spans about 10 kb and is composed of five exons and four introns. The exons are sized about 130, 185, 94, 160, and 140 bp, respectively (18, 23). The four introns are about 0.3, 1.2, 2.4, and 0.69 kb, respectively (18, 23). Recently, the existence of an additional exon (159 bp) was reported within the 5' untranslated region, followed by a 147-bp intron (24) (Fig. 1). Northern analysis using a specific hPTTG1 probe reveals a single approximately 1.3-kb transcript (18, 20); however in human fetal liver the detected size is approximately 0.8 kb (18). The major transcription start site localizes at an adenine residue at 37 bases, and a second transcription initiation site (TSS) 317 bp upstream of the ATG translation start site (23, 24). Serial deletions of an active approximately 1.3 kb upstream region showed sequences –126 to +34 bp sufficient for hPTTG1 promoter activity and that the sequence between nucleotide –706 and –407 contains an enhancer element (23).

By testing several pairs of degenerate primers designed from the rat Pttg1 cDNA sequence, one pair produced a prominent RT-PCR band in murine cell lines (25). Cloning and sequencing revealed that PCR products comprised identical sequences and contained an entire coding region (GenBank accession no. AF069051), which at the cDNA level showed 88 and 78% identity to rat and human PTTG1 cDNA, respectively. The entire murine Pttg1 gene spans about 7 kb and comprises five exons of 391, 179, 94, 150, and 131 bp, respectively, and four introns of 0.9, 2.1, 2, and 0.8 kb, respectively (25) (Fig. 1). As determined by Northern analysis, the major mRNA transcript is sized approximately 1 kb; in several samples, such as whole embryo, heart, and F9 and AtT20 cells, an approximately 1.7 kb band was detected, whereas in whole embryo and embryonic F9 cells, a third, approximately 3.0 kb band was also detected, indicating the existence of alternatively spliced variants (25). Indeed, a murine Pttg1 cDNA variant has been identified (GenBank accession no. AF071209), which encodes a shorter Pttg1 with differences in the C-terminal tail and altered activity (26). A single transcription start site was localized 303 bp upstream of the Pttg1 ATG translation start site (25), and serial deletions of an active 4.3 kb upstream region showed sequences –313 to –150 bp to be critical for promoter activity. Three elements (A-C) within this region contribute to promoter activity, with element A (–313/–293) and particularly subregion –305 to –293 bp, most critical for murine Pttg1 promoter activity (25).

B. PTTG1 mRNA expression profile
Initially, studies of rat Pttg1 mRNA expression using Northern blotting revealed detectable levels exclusively in adult testis and embryonic liver (1). Subsequently, examination of rat Pttg1 mRNA in testicular cell lines, showed high levels of Leydig, Sertoli, as well as germ (GC2) cell expression (17). In murine tissues, however, detectable mRNA levels (by Northern blot) were reported in thymus, spleen, testis, and ovary, as well as intact embryo, but not in adult liver, heart, lung, brain, and kidney (25). The discrepancy between rat and mouse tissue is likely due to differences in Northern blot sensitivity.

In normal adult human tissues (Fig. 3), abundant PTTG1 mRNA expression is evident in the testis (18, 19, 20). Strong expression is also observed in thymus, and weak colon, small intestine, placenta, and spleen expression are seen (18, 19). In some cases, weak signals have been reported in brain, pancreas (18), and lung (19), whereas expression has not been detected in leukocytes, heart, liver, skeletal muscle, kidney (18, 19), or ovary (18).

View larger version (27K): [in this window] [in a new window]   FIG. 3. Tissue distribution of hPTTG1 mRNA, as determined by Northern blot. a, Normal human adult tissues. b, Normal human fetal tissues. c, Human cancer cells, left to right: promyelocytic leukemia HL-60; HeLa cell S3; chronic myelogenous leukemia K-562; lymphoblastic leukemia MOLT-4; Burkitt’s lymphoma Raji; colorectal adenocarcinoma SW480; lung carcinoma A549; melanoma G361. d, Pituitary tumors: NF, nonfunctioning tumor; PRL, PRL-secreting tumor; ACTH, ACTH-secreting tumor. [Reproduced with permission from Zhang et al. (18 ). Copyright 1999, The Endocrine Society.]

PTTG1 is abundantly expressed in several human cancer cell lines, including hematopoetic cell lines [promyelocytic leukemia HL-60, HeLa cell S3, chronic myelogenous leukemia K-562 (18, 21), lymphoblastic leukemia MOLT-4, and Burkitt’s lymphoma Raji (18)], colorectal adenocarcinoma SW480 (18, 21), lung carcinoma (A549), melanoma (G361) (18), hepatoma (HepG2) (21), breast tumor (MCF-7), endometrium (HEC-101), and ovarian tumor cell lines (CaO4, PA1, SKOV3, and VOI101) (20).

In addition to high PTTG1 mRNA levels reported in most pituitary tumors (3, 27, 28, 29), PTTG1 overexpression has been confirmed in colorectal (10) and esophageal (30) cancer, and hepatocellular carcinoma (31), testicular, ovarian, and breast tumors (6), in uterine leiomyomas (7), thyroid (11) and lung cancer (32).

C. Regulation of gene expression
Estrogen has been shown to be an important regulator of pituitary Pttg1 mRNA expression (33, 34, 35) (Fig. 4). Estrogen treatment of rat pituitary somato-lactotroph GH3 cells dose-dependently induced Pttg1 mRNA expression at 12 h, peaking at 24 h (33). In vivo, estrogen administration to Fischer 344 rats by osmotic minipump increased Pttg1 mRNA levels within 24 h, peaked at 48 h, and remained elevated throughout the 14-wk treatment period (33). In vivo estrogen-induced Pttg1 was blocked by antiestrogen coinfusion (34).

View larger version (56K): [in this window] [in a new window]   FIG. 4. Estrogen and Pttg1. a, Representative normal rat pituitary (NI) and rat pituitary tumor (E2). b, Serum PRL and pituitary wet weight. OVx, Ovariectomized controls. *, P < 0.05; **, P < 0.01. c, Northern blot analysis of pituitary tissue extracts derived from estrogen-treated rats. ß-Actin was used as loading control. M, marker lane. d and e, Reticulin fiber staining of rat anterior pituitary tissue at 24 h (d) and 1 wk (e) (broken circle) after initiation of estrogen treatment. f and g, Reticulin stain (arrows) (f) and hematoxylin and eosin stain (g) of rat anterior pituitary tissue 4 wk after commencement of estrogen infusion. Widespread vacuolation, vascular lakes (g, arrow), nuclear pleomorphism, and frequent mitosis (g, arrowhead) are visible. h, FGF-2 immunoreactivity after 4 wk of estrogen treatment. Original magnification, x200. [Reproduced with permission from S. Melmed (130 ). Copyright 2003, The American Society for Clinical Investigation.]

Insulin and IGF-I activate PTTG1 mRNA transcription in breast (MCF-7) tumor cells and malignant astrocytes, mediated in both cases by phosphotidylinositol 3-kinase (PI3K) and MAPK signaling (8, 38). In U87 astrocytoma cells, PTTG1 mRNA expression was also induced by the epidermal growth factor (EGF) receptor ligands, EGF and TGF, and hepatocyte growth factor (HGF) (37).

In most of these models it was unclear, however, whether induction of PTTG1 transcription occurred as a direct effect of the specific treatment agent or secondary to induced proliferation and associated activation of the cell cycle machinery. Using pituitary folliculostellate cells (TtT/GF), induction of Pttg1 mRNA levels by EGF required cells to progress through the cell cycle, because blockade of cells in early S phase suppressed EGF-mediated Pttg1 induction (39). PI3K, protein kinase C and MAPK pathways signal for EGF-induced Pttg1 expression in folliculostellate cells, and signaling suppression of Pttg1 was associated with inability of cells to proceed through S phase (39).

Indeed, PTTG1, as a mammalian securin protein (2), exhibits a cell cycle-dependent expression pattern (2) and is subject to ubiquitin-mediated degradation at the end of metaphase (2, 40) (see Section III.B). Pttg1 mRNA levels are also cell cycle-dependent; Pttg1 mRNA increases during S phase and peaks at the S-G₂ transition to decline thereafter (39, 41) (Fig. 5). Transcription factors Sp1 and, to a lesser extent, nuclear factor-Y are important for basal transcription activity of the PTTG1 promoter (24, 42), particularly in cancer cells. The nuclear factor-Y site appears to be essential for DNA damage-induced and p53-mediated inhibition of PTTG1 transcription (42). Although the ß-catenin/T cell factor (TCF) pathway has been implicated in regulation of hPTTG1 expression in esophageal and colorectal cancer (43, 44), mechanisms controlling cell cycle-dependent PTTG1 transcription and patterns of dysregulation in transformed cells remain to be elucidated.

View larger version (38K): [in this window] [in a new window]   FIG. 5. Cell cycle-dependent Pttg1 mRNA expression. a, Time-dependent effects on Pttg1 mRNA expression. Pituitary folliculostellate TtT/GF cells were serum-starved for 20 h and subsequently treated with EGF (5 nM) for indicated times. b, Effect of early S phase blockade on Pttg1 mRNA expression. TtT/GF cells were serum-starved for 20 h and subsequently treated with aphidicolin for 1 h before stimulation with EGF (5 nM). a and b, Pttg1 mRNA expression was determined by Northern blot. Subsequently, membranes were stripped and reblotted with specific probes for PBF and PCNA, respectively (middle panels). The ratios of Pttg1 and PCNA mRNA vs. ß-actin mRNA were calculated by densitometric analysis of each treatment group. The Pttg1/ß-actin ratio or PCNA/ß-actin ratio of the control group was set as 1.0. Relative mRNA expression levels of other groups were normalized to these control groups. Relative Pttg1 or PCNA mRNA expression (mean ± SEM) of three independently performed experiments (upper panel). *, P < 0.05; **, P < 0.01; ***, P < 0.001. At the indicated times, cells were fixed, and cell cycle analysis was performed by flow cytometry. Percentage of cells in G0/G1 phase is depicted in black bars, cells in S phase in white bars, and cells in the G2/M phase of the cell cycle in gray bars (lower panel). [Reproduced with permission from Vlotides et al. (39 ). Copyright 1999, The Endocrine Society.]

	III. PTTG Protein

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

Figure 6 is a schematic representation of mammalian PTTG1 protein, comprising an N-terminal regulatory domain and a C-terminal functional domain (19, 45).

View larger version (6K): [in this window] [in a new window]   FIG. 6. Schematic illustration of mammalian PTTG1 protein. Mammalian PTTG1 protein comprises a mainly regulatory N-terminal domain (left panel) and functional C-terminal domain (right panel). Regulatory and functional domains, as well as SH3-binding sites and important serine amino acid residues are indicated. Positions of the elements are shown by amino acid residue numbers. Italics indicate information obtained from rodent Pttg1 protein data.

View larger version (32K): [in this window] [in a new window]   FIG. 7. Sequence alignment of vertebrate securins. The hSecurin (hSec) sequence is identical to that of hPTTG1. The sequences of mouse (mSec) and rat securin (rSec) were obtained from GenBank (accession no. AF069051 and U73030, respectively) (xSec, Xenopus securin). Amino acid sequence alignment was obtained with MegaAlign (DNAStar) by the clustal method. Identical or conserved residues are shaded in black. The conserved D box is boxed. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr. [Reproduced with permission from Zou et al. (2 ). Copyright 1999, American Association for the Advancement of Science.]

C-terminal PXXP motifs are critical for hPTTG1 action (18, 46, 47). Similar to its human counterpart, mutations of C-terminal key amino acids (P¹³⁹, S¹⁵⁹) or in the PPXP (P¹⁵⁷-P-S-P¹⁶⁰) motif in murine Pttg1 disrupt its functionality (26). Interestingly, a murine Pttg1 188-amino acid variant containing a different C-terminal tail (GKGVRSNGCKDLVT) is devoid of primary functions of Pttg1, and postulated as an endogenous competitor to wild-type Pttg1 (26). Studies with rodent Pttg1 have identified DNA-binding (approximately from residue 60 to 118) (48) and transactivating domains (approximately from residue 119 to 164), important for Pttg1 transcriptional activity (49).

The N terminus contains a destruction (D) box (RKALGTVN; amino acid residues 61 to 68), which is conserved in human, rat, and mouse PTTG1, as well as a KEN or KDN box in human or mouse and rat, respectively (2), both involved in PTTG1 degradation (see Section III.C).

B. PTTG1 protein subcellular localization and complex formation
Reported differences in cytoplasmic vs. nuclear localization may be due to different cellular systems and techniques employed; although the physiological role of cytoplasmic PTTG1 expression remains unclear, nuclear localization of PTTG1 is consistent with the biological activity of PTTG1 as a transcriptional activator and securin protein.

hPTTG1 localizes both to the cytoplasm and nucleus (3, 19, 27, 41, 50, 51); however, the ratio of cytoplasmic vs. nuclear localization remains controversial. Endogenous Jurkat cell hPTTG1, as determined by subcellular fractionation, was reported to be mainly cytoplasmic (85%) (19). Predominantly cytoplasmic PTTG1 expression was observed by in situ hybridization (3) and immunohistochemistry (27) in a large series of human pituitary adenomas (27). Nuclear PTTG1 staining was detected in some tumor cells, whereas PTTG1 was not detected in normal pituitary tissue (27). In the same study, 14 of 18 breast and 9 of 10 lung adenocarcinomas showed mainly cytoplasmic immunoreactivity; however, in these more aggressive tumors, stained nuclei were also frequently detected (27). Similarly, predominantly cytoplasmic localization was reported in HCT116 cells transfected with enhanced green fluorescent protein (EGFP)-tagged PTTG1 (51). However, JEG-3 cells, with low constitutive PTTG1 expression, transfected with three different PTTG1 constructs (wild-type PTTG1, a FLAG epitope-tagged PTTG1, or a PTTG1-EGFP construct), all demonstrated predominantly nuclear PTTG1 localization during interphase (41). Interestingly, in some cells (<5%) PTTG1 also localized to the plasma membrane (41). Predominantly nuclear expression was observed regardless of the respective expression level of PTTG1-EGFP and was confirmed in a variety of other cell lines (3T3 murine fibroblast, rat GH3 and mouse AtT20 pituitary tumor, SKOV-3 human ovarian cancer, COS-7 monkey kidney cells). In the same study, live imaging of PTTG1-EGFP during mitosis revealed colocalization with microtubule asters in prophase and prometaphase, aggregation into distinct granules during anaphase, and diminished telophase expression (41). Mu et al. (52) reported cell type-dependent PTTG1 subcellular distribution, which was predominantly nuclear in HeLa, Cos-7, and DU145 cells, but diffuse nuclear and cytoplasmic localization in A549, DLD-1, and NIH3T3 cells.

Although the PTTG1 molecule does not possess an obvious nuclear localization signal (NLS), it may enter the nucleus due to its small molecular size. However, PTTG-binding factor (PBF) was identified, which facilitates PTTG1 nuclear translocation (50). PBF is a 22-kD protein ubiquitously expressed in various tissues and contains a bipartite NLS between amino acids 149 and 166 at the C terminus (50). COS-7 cells transfected with a hemagglutinin-tagged full-length PBF construct exhibited mainly nuclear expression, whereas a NLS deletion mutant yielded a predominantly perinuclear and cytoplasmic pattern (50). In vitro glutathione-S-transferase fusion protein binding assays and immunoprecipitation studies showed that PTTG1 specifically interacts with PBF and that the region between the C-terminal amino acids 123 and 154 of PTTG1, as well as the C-terminal region of PBF, are essential for this interaction (50). The mainly cytoplasmic PTTG1-EGFP localization (overexpression of PTTG1 alone) became predominantly nuclear with simultaneous PBF overexpression (50). Furthermore, PBF was shown to be important for PTTG1 transcriptional activity (50). However, in HCT116 cells transfected with PTTG1-EGFP, which was primarily cytoplasmic, cotransfection with hemagglutinin-tagged PBF did not result in different subcellular localization, and colocalization of PTTG1 and PBF was mainly cytoplasmic (51). In addition to the PTTG1 interaction with PBF, activation of the MAPK cascade was proposed as another mechanism for PTTG1 nuclear translocation and required the presence of MAPK phosphorylation and MAPK kinase 1 (MEK1) interaction sites (49).

C. PTTG1 protein posttranscriptional modification
1. PTTG1 phosphorylation.
PTTG1 sometimes migrates as a doublet, possibly due to cell cycle-dependent phosphorylation (1, 2, 13, 53) particularly during mitosis (53) (Fig. 8). Cdc2 is responsible for PTTG1 phosphorylation, mainly at S¹⁶⁵, which is also conserved in rat and mouse Pttg1 (53). Because this site is located between the two C-terminal proline-rich motifs, important for PTTG1 transcriptional activity, PTTG1 phosphorylation may be required for its functional activity (53). In rat Pttg1, S¹⁶² is phosphorylated by MAPK, and this site plays a critical role in Pttg1 transactivation function (49). Furthermore, at least six potential motifs (S/T-Q or P-S/T) for DNA-dependent protein kinase (DNA-PK) phosphorylation have been identified in hPTTG1, and in vitro, the catalytic subunit of the DNA-PK complex does indeed phosphorylate hPTTG1. The exact location as well as significance of DNA-PK-mediated phosphorylation remains to be elucidated, because it had no demonstrable effect on hPTTG1-Ku-70 association (13).

View larger version (15K): [in this window] [in a new window]   FIG. 8. PTTG1 phosphorylation during mitosis. HeLa cell PTTG1 protein expression determined by Western blot arrested at the indicated cell cycle phases. Shown is one representative Western blot (lower panel). The level of hPTTG1 was adjusted relative to the amount of an unrelated noncycling protein that was detected with a polyclonal serum (data not shown). The highest level, observed in M phase, was set at 100%. Shown are relative values of three independent experiments (upper panel). [Adapted by permission from Macmillan Publishers Ltd.: Oncogene (53 ). Copyright 2000.]

PTTG1 degradation ensues as a result of ubiquitination by anaphase-promoting complex (APC) (40), a ubiquitin ligase complex that contains several different subunits and acts to ubiquitinate several cell cycle proteins (57). In contrast to budding yeast pds1, hPTTG1 degradation is catalyzed by both fzy (fizzy/cdc20) and fzr (fizzy-related/cdh1/hct1) proteins (40). hPTTG1 degradation is both D box and KEN box-dependent (40), in contrast to Xenopus securin, where D box mutations alone prevent degradation (2). Mutations of the KEN or D box alone did not, or only partially, inhibited hPTTG1 degradation, respectively (40), whereas double mutations of the KEN box and the D box abolished PTTG1 degradation (40).

	IV. PTTG Family Members

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
A. Cloning and characterization of PTTG2 and -3
Evidence for the existence of more than one hPTTG isoform was reported in 1999 (22); PCR of genomic DNA led to the identification of an intronless homolog of hPTTG1, mapped to chromosome 4p12 and termed hPTTG2 (22). Sequence analysis (GenBank accession no. AF116538) revealed 94% nucleotide homology to hPTTG1 cDNA and 91% identity at the amino acid level (22). A subsequent study of hPTTG family members confirmed the existence of hPTTG2 (GenBank accession no. AF200719), which mapped to chromosome 4p15.1 by fluorescence in situ hybridization (58). hPTTG2 contains a 606-bp coding sequence, encoding a 202-amino acid protein, with a 91% identity to hPTTG1 at the amino acid level (58). hPTTG2 also contains two conserved motifs at the C terminus (58).

The observation of multiple polymorphisms in a region of hPTTG2 in the normal thymus suggested the existence of yet a third PTTG isoform (22). Indeed, Chen et al. (58) confirmed the existence of a second intronless homolog of hPTTG1, which was termed hPTTG3 and mapped to chromosome 8q13.1. hPTTG3 also contains a 606-bp coding sequence and encodes a 202-amino acid protein with 89 and 84% identity to hPTTG1 and hPTTG2, respectively (58). Similar to hPTTG1 and -2, hPTTG3 contains two conserved proline-rich motifs at the C terminus (58).

B. PTTG family member expression profile
With the use of specific probes, which did not appear to cross-hybridize between the two family members, Northern analysis revealed low-level hPTTG2 mRNA expression in spleen, prostate, testis, ovary, small intestine, and colon, but not in the thymus or peripheral blood leukocytes (58). With the use of more sensitive RT-PCR, hPTTG2 mRNA expression was detected in pituitary and pituitary tumors (22), liver and liver tumors (58), testis and testicular tumor, and ovary and ovarian tumor cell lines (22, 58) (Table 1).

	V. PTTG1: Physiological Functions

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
A. Securin function/replication
During metaphase, replicated paired sister chromatids are held together by the cohesin complex, a process essential for chromosome biorientation on mitotic spindles (59). This multisubunit "cohesin" requires activity of four proteins, i.e., Smc1, Smc3, Scc1, and Scc3 (extensively reviewed by Nasmyth in Ref. 60 and Yanagida in Ref. 61), and several variants of a large cohesin ring model entrapping sister chromatids have been proposed (60, 61, 62, 63, 64). In contrast to yeast, two distinct processes induce cohesin dissociation from chromosomes during mitosis in mammalian cells. The "prophase pathway" begins during prophase and involves phosphorylation of Scc3-like subunits by the Polo-like kinase PLK1 (65, 66). The second process takes place at the onset of anaphase and involves cleavage of Scc1 by a protease termed separase (as in yeast) (67). Separases are, in most species, large proteins (150–230 kDa), initially described as Esp1 in budding yeast (68) and Cut1 in fission yeast (69). A C-terminal (ca. 50 kDa) "separase domain" is highly conserved and harbors the proteolytic site (70, 71), whereas the N-terminal regions are not obviously conserved between species and likely harbor sites important for cellular trafficking of the protein and its interaction with securin (72, 73, 74, 75).

During most of the cell cycle, securin binds separase and thus inhibits its proteolytic activity (67, 74, 76, 77), in addition to inhibition of vertebrate but not yeast separase by Cdk1/cyclin B protein kinase-dependent phosphorylation (78, 79). At the metaphase to anaphase transition, once chromosome biorientation is complete, securin is targeted for proteasomal degradation by the ubiquitin ligase APC (see Section III.C); securin destruction releases tonic separase inhibition, which in turn mediates degradation of the cohesin complex (80), and equal separation of sister chromatids proceeds to diploid daughter cells. Interestingly, experiments in yeast have shown both an activating and inhibitory role of securin on separase, suggesting a dual mechanism of separase regulation (76). Binding of securin to separase (both N- and C-terminal binding sites) not only blocks interaction of separase with its substrates, but also prevents the N terminus from interacting with and possibly inducing an activating conformational change at the C-terminal active site, thus inhibiting separase activity (76). On the other hand, securin is important for nuclear accumulation of (inactive) separase and ensures that separase gains full proteolytic activity in anaphase after its own destruction (76).

In addition to PTTG1 securin function, recent reports indicate a role for PTTG1 in regulation of the G₁/S phase transition (81). PTTG1 interacts with the transcription factor Sp1, and a PTTG1-Sp1 complex colocalizes on the cyclin D3 (CCND3) promoter in JEG-3 and HCT116 cells. Suppression of Sp1 or PTTG1 [small interfering RNA (siRNA)] resulted in attenuated G₁/S transition in JEG3 cells, with increased G₁ and decreased S phase (81). Cotransfection of Sp1 siRNA and PTTG1 plasmid or PTTG1 siRNA and Sp1 plasmid reversed the effect of the other, indicating a coordinate role of PTTG1 and Sp1 on modulation of G₁/S phase transition (81). Sp1 transfection induced CCND3 mRNA and protein expression, whereas cotransfection of Sp1 and PTTG1 siRNA neutralized CCND3 induction. Furthermore, using p21 –/– HCT116 cells, PTTG1-mediated G₁/S phase transition was shown to be p21-independent (81).

B. DNA damage/repair
In budding yeast, securin is required to prevent anaphase in response to spindle and DNA damage (82, 83). Securin (Pds1) is directly phosphorylated by Chk1 (83), a conserved kinase critical in the DNA checkpoint pathways (84). Because DNA damage had no effect on APC activity (83), it was suggested that Chk1-dependent securin phosphorylation may increase its stability and thus block cell cycle progression. In fission yeast, securin (Cut2) is essential for proper repair of DNA damaged by UV, x-ray, and -ray irradiation, possibly by removal of local cohesin in interphase cells by the securin/separase complex (85).

In mammalian cells, hPTTG1 binds to Ku-70, the regulatory subunit of the DNA-PK, an enzyme involved in repairing DNA double-strand breaks (13). DNA-PKs phosphorylate hPTTG1 in vitro, and the hPTTG1-Ku association is disrupted by genome damaging events, such as DNA double-strand breaks, implicating a role for hPTTG1 in the linkage of DNA damage-response pathways with sister chromatid segregation, delaying the onset of mitosis while DNA repair proceeds (13). hPTTG1 is required for cell proliferation arrest after UV treatment (86). However, unlike in yeast cells, x-rays and UV light rapidly reduce hPTTG1 protein levels in mammalian cells, mediated by specific protein synthesis inhibition as well as proteasome-dependent degradation (86). Treatment of human cancer cell lines with DNA-damaging drugs doxorubicin and bleomycin induces p53-dependent suppression of hPTTG1 expression (42). hPTTG1 overexpression induced genetic instability in colorectal cancer cells, mediated at least in part by inhibition of DNA damage repair activity and repressed Ku heterodimer function (87). Thus, PTTG1-mediated inhibition of Ku-70 activity may be critical in DNA-damage repair regulation and induction of genetic instability in this human cancer model.

C. Interaction partners and transactivation activity
With the use of the yeast two-hybrid system in testicular cells, Pttg1 was shown to specifically interact with a novel human homolog of the bacterial heat-shock protein DnaJ (HSJ2) and with the ribosomal large subunit protein S10 (88). Pttg1 may therefore be targeted to the ribosome through interaction with S10 and association with HSJ2 required for subsequent dissociation of the Pttg1-S10 complex (88). However, this hypothesis and its potential significance in testicular physiology has not yet been further explored (for overview of major PTTG1 interaction partners, see Table 2).

Pttg1 contains a consensus MAPK phosphorylation site (P-X-S/T-P) within the transactivation domain and is phosphorylated by MAPK at S¹⁶² in vitro, a site critical for Pttg1 transactivation function (49). Furthermore, Pttg1 directly interacts with MEK1 through the N-terminal SH3 domain-binding motif (located between amino acids 51 and 54), and this interaction is required to mediate effects of MAPK on Pttg1 transactivation activity (49).

By generating cell lines with tightly regulated inducible Pttg1 expression and profiling using DNA arrays (investigation of 84 genes), five genes showed increased expression (>2-fold) after Pttg1 induction. These included c-myc oncogene, protein kinase C ß-1, MEK1, MEK3, and heat shock protein HSP70 (48). Detailed examination of the effect of Pttg1 on c-myc expression revealed that Pttg1 activates c-myc transcription in transfected cells and that Pttg1 binds to the c-myc promoter near the TSS in a protein complex containing the upstream stimulatory factor (USF1) (48). In this study, a DNA binding region was identified from amino acid residue 60 to 118, and Pttg1 DNA binding is necessary for its transcriptional activity (48).

A well-established target of PTTG1 is the mitogenic and angiogenic factor fibroblast growth factor (FGF)-2. Both PTTG1 and FGF-2 are overexpressed in pituitary (3, 29, 33), thyroid (11), ovarian tumors (6), and uterine leiomyoma (7). In vitro experiments have shown FGF-2 mRNA induction by PTTG1 overexpression in several cell types, including NIH3T3 cells (18, 90), primary thyroid cells (47), human fetal neuronal NT2 cells, JEG-3 and MCF-7 cells (46). In primary thyroid and PTTG1-null cell lines, transactivation of FGF-2 by hPTTG1 was not affected by PTTG1 phosphorylation but was dependent on the integrity of C-terminal PXXP motifs (47). Furthermore, direct transcriptional control of FGF-2 by PTTG1 has been reported (50). COS-7 cells transiently transfected with the FGF-2 promoter linked to luciferase were cotransfected with PTTG1 and/or PBF (50). Coexpression of PBF and PTTG1 significantly enhanced reporter gene activity, indicating the requirement of PBF for activation of FGF-2 transcription by PTTG1 (50).

In addition to FGF-2 regulation, hPTTG1, and particularly its C terminus (amino acids 147–202), regulates pituitary hormone expression (91). Pituitary lacto-somatotroph GH3 cells transfected with wild-type PTTG1 C-terminus expression vector decreased (10-fold) prolactin (PRL) mRNA expression and PRL hormone secretion, compared with cells overexpressing a mutated PTTG1 C-terminus SH3-binding motif (91). In contrast to PRL, GH mRNA expression and GH secretion in GH3 cells was enhanced (4-fold) (91). Direct effects of PTTG1 C terminus on PRL gene transcription were confirmed, because cells overexpressing wild-type PTTG1 C terminus demonstrated approximately 10-fold decreased PRL promoter-driven luciferase activity compared with mutant transfectants (91).

Recently, ChIP-on-chip technology was used to survey PTTG1 action on 20,000 genes (81). Of 700 gene promoters significantly enriched by immunoprecipitation with PTTG1 antibody, about 400 were identified and categorized into three major functional groups, involved in cell cycle, metabolic control, or signal transduction pathways (81). The use of an algorithm designed to specifically unravel transcriptional patterns yielded by the ChIP-on-chip results revealed enriched transcription factor binding sites in PTTG1-targeted genes, with a significant overrepresentation of Sp1 (81). PTTG1 was shown to physically interact with the Sp1 complex, and the PTTG1-Sp1 complex was shown to regulate CCND3 expression and G₁/S phase transition (see Section V.A). These results support a role for PTTG1 in transcriptional regulation of genes involved in a variety of cellular processes.

	VI. PTTG1: Tumorigenic Mechanisms

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
A. PTTG1 and cell proliferation
The effect of PTTG1 on cell proliferation as tested in transfection studies remains unclear. Because it is an oncogenic protein, one would expect a pro-proliferative effect; on the other hand, as a securin protein, which normally inhibits cell division unless degraded, high PTTG1 expression levels would be expected to inhibit cell proliferation. Indeed, results using cell lines overexpressing PTTG1 have been inconclusive. Stable transfection of NIH3T3 cells with rat Pttg1 (1) or HeLa and A549 with hPTTG1 (52) inhibited cell proliferation, as determined by nonradioactive cell proliferation assays. Cell cycle analysis of JEG-3, MG-63, or H1299 cells overexpressing EGFP-tagged or pcDNA3-PTTG1 revealed blockade of the cell cycle and accumulation of cells in G₂M (41, 92, 93). On the other hand, NIH3T3 cells (20) and human embryonic kidney (HEK) 293 cells (94) stably transfected with hPTTG1 exhibited increased cell proliferation rates compared with control vector-transfected cells. In HeLa S3 cells engineered to exhibit tetracycline-regulated PTTG1 expression, induction of PTTG1 also stimulated cell growth (48).

The observed differences could be ascribed to different cellular systems tested, varying PTTG1 expression levels depending on transfection efficiency or other factors that may interfere with PTTG1 physiology, such as the ability of a particular cell to translocate to the nucleus or degrade high levels of exogenous PTTG1 expression.

Targeting endogenous PTTG1 expression might constitute a more physiological approach for understanding PTTG1 regulation; however, very low levels of PTTG1 may also inhibit cell cycle progression due to the dual effect of PTTG1 on separase regulation and the requirement of PTTG1 abundance for nuclear accumulation and separase activation (76). Inhibition of cell proliferation by hPTTG1 siRNA has been reported in U87 glioma cells (37), HeLa S3 (95), and SH-J1 hepatoma cells (31).

The impact of transfected PTTG1 expression levels on cell proliferation was demonstrated in NT-2 cells. Relatively low PTTG1 up-regulation (1.7-fold) stimulated cell proliferation, whereas high PTTG1 up-regulation (6-fold) inhibited cell turnover (96).

Interestingly, the phosphorylation status of PTTG1 is also involved in pro- or antiproliferative effect of PTTG1 overexpression. The use of stable NIH3T3 cells overexpressing mutants preventing Pttg1 phosphorylation (Phos–) or those mimicking constitutive phosphorylation of Pttg1 (phos+) revealed that cells expressing Phos– demonstrated increased, and Phos+ decreased, proliferation compared with wild-type Pttg1 (47). Thus, the ability of a particular cell system for posttranslational modification (e.g., phosphorylation) of exogenous PTTG1 may influence the outcome of PTTG1 overexpression on cell proliferation.

The significance of PTTG1 regulation of cell cycle control gene promoters and the impact on cell proliferation is currently under active investigation.

B. Cell transformation and aneuploidy
In contrast to the effect of PTTG1 overexpression on cell proliferation, the ability of PTTG1 to transform cells is well established both in vitro and in vivo. As assessed by anchorage-independent growth in soft agar, PTTG1-transfected NIH3T3 and HEK293 cells formed much larger colonies compared with control vector-transfected cells (1, 18, 47, 94). In HeLa S3 cells engineered to exhibit tetracycline-regulated PTTG1 expression, induction of PTTG1 with tetracycline withdrawal led to 2-fold increased colony formation (48). Consistent with these results, H1299 lung carcinoma cells targeted with hPTTG1 siRNA formed smaller colonies in soft agar (9). In vivo, sc injection of PTTG1-transfected NIH3T3 or HEK293 cells into nude mice led to formation of large tumors within 2 to 4 wk (1, 18, 20, 94).

Several lines of evidence suggest that an important transforming mechanism underlying PTTG1 overexpression is the induction of chromosomal instability and aneuploidy.

p53-deficient MG-63 osteosarcoma cells transiently or stably transfected with PTTG1-EGFP were observed for signs of aneuploidy, such as the presence of micronuclei, macronuclei, or chromosomal bridges (92). Indeed, 34% of transiently transfected cells (1% in control cells) and 10% of stably transfected cells (5% in control cells) showed signs of aneuploidy (92). Live imaging of single human lung cancer H1299 cells, which are p53 deficient and exhibit undetectable levels of endogenous PTTG1 expression, transfected with EGFP-tagged PTTG1, revealed that high PTTG1-EGFP levels blocked progression of mitosis to anaphase (97). In all PTTG1-EGFP-expressing cells that underwent apparent normal mitosis, PTTG1-EGFP was degraded about 1 min before anaphase onset, consistent with the securin function of PTTG1 (97). Cells that failed to degrade PTTG1-EGFP demonstrated asymmetrical cytokinesis without chromosome segregation, or chromosome decondensation without cytokinesis, resulting in the appearance of macronuclei (97). Most cells expressing a nondegradable PTTG1 mutant exhibited asymmetrical cytokinesis without chromosome segregation, whereas few cells had decondensed chromosomes, and both resulted in formation of macronuclei (97). Thus, this study performed in single live human cells, demonstrated that PTTG1 accumulation (overexpression or failure of degradation) inhibits mitosis progression and chromosome segregation, but does not directly affect cytokinesis, resulting in aneuploidy.

The importance of hPTTG1 for chromosomal stability was confirmed in human colorectal cancer HCT116 cells (98), a cell line with a stable karyotype and intact DNA damage and mitotic spindle checkpoints (99, 100). hPTTG1 inactivation by homologous recombination in HCT116 cells resulted in frequent chromosome loss, linked to abnormal anaphases during which cells underwent repetitive unsuccessful attempts at chromosomal segregation (98). Abnormal mitosis due to lack of PTTG1 activity was associated with biochemical defects in separin activation and reduced cohesin subunit Scc1 cleavage efficiency (98).

Interestingly, with the use of a similar technique to inactivate hPTTG1 in HCT116 cells, chromosome loss was observed to be transient (101). Although approximately 60% of metaphase spreads indeed showed loss of at least one chromosome during the first three passages, the rate of chromosome loss progressively decreased, and by passage 12 karyotyping of HCT116 cells was identical to parental cells (101). However, PTTG1 loss-associated reduction of separase activity and cleavage of the cohesin subunit Scc1 persisted. hPTTG1-null HCT116 cells were chromosomally stable and mitosis was normal, suggesting additional compensatory mechanisms for chromosome segregation in human cells (101).

C. Apoptosis
In a study of placental JEG-3 cells, most cells overexpressing hPTTG1 underwent apoptosis (41). With the use of MCF-7 breast cancer cells expressing wild-type p53 and p53-deficient MG-63 osteosarcoma cells, Yu et al. (92) demonstrated that PTTG1 overexpression may cause both p53-dependent and p53-independent apoptosis. PTTG1 induced p53 in MCF-7 cells (92, 102) and its translocation to the nucleus (92). Furthermore, induction of p53 promoter activity by PTTG1 was shown to be mediated through regulation of c-myc, which then interacted with the p53 promoter (102). PTTG1 overexpression also stimulated Bax (encodes a pro-apoptotic member of the BCL2 gene family), a known downstream target of p53 (102). The nature of the p53-independent mechanism remains unclear. Thus, PTTG1-induced apoptosis may be a protective mechanism for clearing cells becoming aneuploid. When both apoptotic systems fail, PTTG1-induced aneuploid cells may continue to divide, thereby supporting the transformed phenotype.

In contrast, Bernal et al. (93) reported that hPTTG1 specifically interacts with p53 in vitro and in vivo and that this interaction blocks specific binding of p53 to DNA and inhibits its transcriptional activity (93). Interestingly, Bax promoter activity was diminished by the p53-securin interaction, leading to decreased apoptosis. In hPTTG1-deficient human tumor cells, both apoptotic and transactivating functions of p53 were potentiated (93). In addition, overexpressed PTTG1 in hepatoma cell lines attenuated p53 induction of apoptosis (31). Thus, these results propose a tumorigenic mechanism for PTTG1, because inhibition of p53-mediated apoptosis by high securin expression could explain the survival of tumor cells harboring functional p53 (93, 103).

The discrepancy between findings in these studies may depend on transfection systems used, because the amount of PTTG1 protein expressed in the study of Bernal et al. (93) was relatively low compared with expression levels in other studies (92, 102). Thus, induction of p53 promoter activity and associated apoptosis induction may require a threshold intracellular PTTG1 protein level.

D. Tumorigenic microenvironment
The microenvironment surrounding transformed cells is permissive for accelerated growth, which may be altered by paracrine/autocrine feedback mechanisms (104, 105, 106). In this context, an important element of tumor growth is angiogenesis. In colon cancers, PTTG1 expression and tumor vascularity correlate strongly (10), and in vivo pituitary FGF-2 and vascular endothelial growth factor (VEGF)-A induction coincides with estrogen-stimulated PTTG1 expression and angiogenesis (33). Induction of angiogenesis by PTTG1 was demonstrated in vitro by activation of proliferation, migration, and tube formation of human umbilical vein endothelial cells treated with conditioned medium derived from NIH3T3 cells overexpressing hPTTG1 (90). In vivo, concentrated hPTTG1-conditioned medium induced chick chorioallantoic membrane spoke-wheel-like appearances, and in both cases, FGF-2 antibody suppressed PTTG1-mediated angiogenic induction (90). Indeed, overexpression of PTTG1 induces FGF-2 mRNA and/or protein in NIH3T3 (18, 47, 90), HEK293 (94), NT-2 (46, 96), JEG-3, MCF-7 (46), and uterine leiomyoma cells (7). In some cases, PTTG1 overexpression also induces VEGF-A, another potent angiogenic growth factor, in uterine leiomyoma cells (7), FTC133 thyroid cells (107), NT-2, JEG-3, and MCF-7 cells (46), and induces VEGF-A as well as IL-8 in HEK293 cells (94). A recent study with angiogenesis-specific cDNA arrays after PTTG1 transfection in thyroid cells revealed regulation of multiple downstream angiogenic genes, including inhibitor of DNA binding-3 (ID3) and thrombospondin-1 (TSP-1) (108). Although the regulatory mechanism of PTTG1-mediated growth factor induction, including that of VEGF-A, has not yet been elucidated, PTTG1 in complex with its binding factor PBF has been shown to act as a direct transcriptional activator of FGF-2 (50) (see Section V.C).

Thus, overexpressed PTTG1 may act as a paracrine/autocrine activator, while overexpressed, enhancing expression of growth factors that in turn further sustain tumor growth and contribute to the tumorigenic microenvironment. The stimulatory effect of growth factors such as FGF-2 or EGF on PTTG1 expression per se (see Section II.C) further enhances this positive autofeedback mechanism.

	VII. Pttg1-Null Mouse Model

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
Although loss of yeast Pds1p or Drosophila securin pimples is lethal (74, 109), disruption of murine Pttg1 by homologous recombination resulted in a viable and fertile phenotype with tissue-specific defects, including pituitary, testis, and pancreatic ß-cell hypoplasia (14, 15). Pttg1 deficiency did not enhance intrauterine mortality, but Pttg1 –/– mice demonstrated female subfertility with a reduced average litter size compared with Pttg1 –/+ or Pttg1 +/+ (14). Pttg1 –/– mice exhibited testicular and splenic hypoplasia, thymic hyperplasia, and thrombocytopenia (14). Testicular hypoplasia was more pronounced in sexually mature mice, whereas splenic hypoplasia was observed after weaning and continued for up to 8 months of observation (14). Thymic hyperplasia was apparent approximately 4 wk after birth and persisted for up to 8 months, but it was more pronounced at an early age (14). Furthermore, the distribution of CD4+CD8+, CD4+CD8–, and CD4–CD8– thymocytes differed in Pttg1 –/– mice, and hematological analysis revealed the presence of thrombocytopenia, despite normal bone marrow megakaryocyte numbers (14).

Although Pttg1 –/– and Pttg1 +/+ mouse embryonic fibroblasts (MEFs) exhibited similar doubling times (30 h), their cell cycle parameters differed markedly, because Pttg1 –/– MEFs had a shortened G₁ and prolonged G₂M phase, an effect partially reversed by introduction of Pttg1 into Pttg1 –/– MEFs via retroviral transfection (14). Pttg1 –/– MEFs demonstrated damaged nuclei and aberrant chromosome morphology, such as quadric-radials, tri-radials and breaks, binucleated or multinucleated nuclei, and enhanced signs of aneuploidy and premature centromere division. These abnormalities in Pttg1 –/– MEFs were not lethal for these cells (14).

This Pttg1 –/– mouse also revealed an important role for Pttg1 in pancreatic ß-cell function (15). Pttg1 disruption impaired glucose homeostasis and led to male-selective diabetes developing during late adulthood (15). Development of diabetes was associated with nonautoimmune islet damage, insulinopenia, and reversed -/ß-cell ratio (15). Reduction of islet ß-cell mass was evident before diabetes manifestation, and islet cell proliferation, as determined by 5-bromo-2'-deoxyuridine incorporation, was reduced. Pttg1 –/– ß-cells had pleiotropic nuclei, suggesting defects in cell division (15). The sex steroid milieu determined diabetes rescue in gonadectomized and estrogen-treated Pttg1 –/– male mice (110). Although gonadectomy and estradiol therapy had no direct observable effect on hypoplastic ß-cell morphology, altered sex steroid milieu prevented diabetes development by increasing insulin sensitivity, possibly mediated by serum adiponectin elevation (110).

The involvement of Pttg1 in diabetes is further supported by a recent application of a novel network-based approach elucidating genetic networks underlying complex traits in an extensive murine whole brain expression study (16). Application of these techniques identified and experimentally validated gene expression traits predicted to respond to a strong expression quantitative trait locus for Pttg1, and Pttg1 brain expression was linked to metabolic traits, including obesity and diabetes (16).

Mice bearing a single retinoblastoma (Rb) mutant allele develop pituitary tumors with almost complete penetrance (111, 112, 113). To examine the role of PTTG1 depletion on pituitary tumor development in Rb +/– animals, compound Rb X Pttg1 mutant mice were generated (114). This study confirmed spleen, pancreas, and testis hypoplasia in Pttg1-depleted mice and also showed decreased pituitary weights (114) in single Pttg1 mutants as well as Rb+/–Pttg1–/– mice (114). Hypoplastic pituitary glands exhibited decreased cell proliferation and induction of p21 expression in both Pttg1-depleted genotypes (114) (see Section VIII.A).

Mechanisms for the protective effect of Pttg1 deletion on pituitary tumorigenesis may include accelerated gland senescence (115), associated with elevated p21 levels.

Despite aberrant chromosome morphology observed in Pttg1 –/– cells, the nonlethal phenotype suggests the existence of additional compensatory mechanisms for sister chromatid separation in mammalian cells. Furthermore, tissue-specific phenotypic responses in Pttg1 –/– mice suggest a differential role of PTTG1 in different tissues and cell type-specific suppression of cell growth, with particular impact on slow-growing endocrine cells. The gender-specific effect of PTTG1 disruption on pancreatic ß-cells supports the role of additional factors, such as sex steroids, in PTTG1-target tissue regulation.

	VIII. PTTG1 and Cancer

Top Abstract I. Introduction II. PTTG1 Gene Structure... III. PTTG Protein IV. PTTG Family Members V. PTTG1: Physiological... VI. PTTG1: Tumorigenic... VII. Pttg1-Null Mouse Model VIII. PTTG1 and Cancer IX. Summary References

 
A. Endocrine-related cancer
1. Pituitary.
Increased PTTG1 mRNA in pituitary tumor tissue has been reported in several studies (3, 18, 20, 27, 28, 29, 33, 34, 46).

Although Northern analysis and in situ hybridization initially failed to detect PTTG1 expression in the normal pituitary (1, 3, 18), PTTG1 is indeed expressed in normal pituitary tissue at low levels (3, 27) and displays an estrous cycle-dependent expression pattern (34).

Examination of PTTG1 mRNA expression in 54 pituitary tumors by RT-PCR revealed more than 50% increased expression in 21 of 30 nonfunctioning, all 13 GH, nine of 10 PRL, and one ACTH-secreting tumor, and in some cases, more than 10-fold increases were reported (3, 27) (Fig. 9A and Table 3). PTTG1 expression was higher in hormone-secreting tumors that had invaded the sphenoid bone compared with tumors confined to the pituitary fossa (3). McCabe et al. (29) examined PTTG1 mRNA expression in 10 normal and 111 tumorous pituitaries and reported an approximately 5-fold increase overall, and Western blot analysis showed that PTTG1 protein expression was consistent with the mRNA findings (Fig. 9B and Table 3). In this study, PBF mRNA expression was increased in all pituitary tumor subtypes (5.7-fold increase) but particularly in nonfunctioning pituitary tumors (29). A significant correlation between PTTG1 and PBF expression was observed in pituitary tumors, but not in normal pituitary glands (29).

View larger version (15K): [in this window] [in a new window]   FIG. 9. PTTG1 expression in pituitary tumor subtypes. A, Normalized PTTG1 vs. glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression is shown as individual values for each tumor, and mean expression within each tumor subtype group is shown by the horizontal line. NFPA, Nonfunctioning pituitary adenoma (n = 18); somatotroph (n = 12); corticotroph (n = 5); lactotroph (n = 5). Differences between means were analyzed by ANOVA. [Reproduced with permission from Hunter et al. (28 ). Copyright 2003, Society of the European Journal of Endocrinology.] B, Representative Western blot analysis of PTTG1 protein expression in normal pituitary (NP) and nonfunctioning pituitary tumors (T). M, Size marker; +, JEG-3 cell extract (positive control). [Reproduced with