| Ornithine decarboxylase: Expression and regulation in rat brain and in transgenic mice | ||
|---|---|---|
| Prev | Chapter 2. Review of the literature | Next |
Antizyme was originally identified from rat hepatoma cells cultures as a protein that inhibits ODC activity and is induced after addition of putrescine to cells (Fong et al. 1976). Soon it was demonstrated to be present in several other cell lines and to be induced also by spermidine and spermine (Heller et al. 1976) – which are actually more efficient inducers (Matsufuji et al. 1995). Since that antizyme has been observed to exists in various organisms from yeast and fungi to nematodes, insects and vertebrates. Its presence in bacteria has been suggested (Canellakis et al. 1993, Pantazaki et al. 1999), but also called into question (Ivanov et al. 1998a). In addition to inhibiting ODC activity, antizyme was demostrated to target ODC for degradation (Murakami et al. 1992a), to regulate polyamine uptake to cells (He et al. 1994, Mitchell et al. 1994) and potentially excretion out from the cells (Sakata et al. 2000). Within vertebrate species antizyme has multiple isoforms which are discussed in the next chapter. This chapter focuses on antizyme 1 that was the first observed and is the most studied and most ubiquitous. (For reviews, see Ivanov et al. 2000a, Coffino 2001a, Coffino 2001b).
Antizyme 1 has a molecular weight of 26 kDa. Its polypeptide chain consists of 227 amino acids in mouse (Kankare et al. 1997) and rat (Miyazaki et al. 1992), and 228 amino acids in human (Tewari et al. 1994, Hayashi et al. 1997). It has very high affinity to ODC monomer, dissociation constant Kd being ≈ 10 –11 M-1 (Kitani & Fujisawa 1984). The functional domains of antizyme 1 responsible for binding to ODC, tagging ODC for degradation and inhibition of polyamine transport have been defined. The carboxyterminal half (amino acids 121 - 227) is sufficient for binding to ODC (Ichiba et al. 1994, Li & Coffino 1994) and is essential for inhibition of the polyamine transport function (Sakata et al. 1997). Amino acids 69 – 112 are necessary for antizyme-mediated destabilization of ODC and are sufficient to confer accelerated degradation to unstable heterologous proteins when linked to them covalently (Li et al. 1996).
Antizyme 1 plays an important role in the regulation of cellular polyamine levels. Alterations in antizyme expression can have an effect on cellular functions and be of physiological significance. In addition to ODC and AdoMetDC, antizyme expression can be altered in cancer, either decreased (Tsuji et al. 1998) or increased (Saverio et al. 2000). The latter is likely to be the normal response to elevated polyamine concentrations and may, together with increased SSAT activity, prevent accumulation of polyamines to toxic levels. On the other hand, high concentrations of polyamines in primary cultures of prostate carcinoma cells suppressed cell growth of poorly metastatic cells that were able to induce antizyme expression, but not growth of highly metastatic cells unable to induce antizyme (Koike et al. 1999). Furthermore, forced antizyme overexpression in hamster malignant oral keratinocyte cell line resulted in the reversion of malignat phenotype and induction of epithelial differentiation and DNA demethylation (Tsuji et al. 2001) whereas targeted antizyme expression in the skin of transgenic mice reduced tumor promotor induction of ODC and decreased sensitivity to chemical carcinogenesis (Feith et al. 2001). When antizyme was overexpressed in immortalized human prostatic epithelial cells spermidine and spermine levels decreased only slightly, but putrescine levels decreased by 3-fold (Scorcioni et al. 2001). This led to basicly similar accumulation of cells in the S phase of the cell cycle as detected when cells suffer from polyamine deficiency (Fredlund & Oredsson 1996a, Fredlund & Oredsson 1996b). However, expression of antizyme 1 is not required for cell viability. According to unpublished observations (Matsufuji et al. unpublished data), antizyme knock-out mice are viable, morphologically normal and fertile, but they have high perinatal mortality with about one-third dying in the days before and after full term (Coffino 2001a, Coffino 2001b).
There are some very interesting unanswered questions about antizymes and their functions. Unpublished results suggesting that antizyme 1 targets cyclins and cyclindependent kinases for degradation have already been reviewed (Coffino 2001b) and it is possible that antizymes are not involved only in the regulation of polyamine metabolism, but more largely regulating various growth or cell cycle-related processes. This is further supported by the observation that Smad1 protein is bound and targeted for degradation by antizyme 1 (Gruendler et al. 2001). Smad proteins mediate signaling induced by the members of transforming growth factor β superfamily and related to cell proliferation, differentiation and apoptosis (Itoh et al. 2000). The more exact picture about the roles of antizymes will certainly emerge in the near future. It would also be interesting to find out: how antizyme controls polyamine transport? Is it the degradation of some transport system component that antizyme is initiating? Could it be possible to find still elusive mammalian polyamine transport proteins by searching proteins interacting with antizyme?
The antizymes form an ancient gene family. Within vertebrate species, multiple isoforms are found; humans have at least four antizymes encoded by different genes (Coffino 2001a). These antizymes share three common features. They display structural homology, that is strongest in the carboxy-terminal half required for the binding of ODC. They appear to be able to associate with ODC, reducing its activity and, depending on isoform, potentially enhancing its degradation. Finally, their synthesis is stimulated by polyamines via ribosomal frameshifting that requires conserved motifs in the mRNA near the site of frameshifting.
Antizyme 1 was the first described and it has wide tissue distribution (Matsufuji et al. 1990). Antizyme 2 has a similar wide tissue distribution, but it is expressed less abundantly (Ivanov et al. 1998b). Antizyme 3 is expressed only in male germ cells in a post-meiotic stage of their differentiation to mature sperm (Tosaka et al. 2000, Ivanov et al. 2000c). Antizyme 4 is presently known only as an EST (Coffino 2001a). Human antizyme 2 is 54 % identical to human antizyme 1, but 99.5 % identical to mouse antizyme 2 indicating unexpectedly high selection pressure against changes in antizyme 2 proteins (Ivanov et al. 1998b). Antizyme 3 is more divergent. Human and mouse antizyme 3s show 29 – 38 % identity with antizymes 1 and 2 (Tosaka et al. 2000, Ivanov et al. 2000c).
Antizymes 1 and 2 both bind and inhibit ODC, and they are approximately equipotent as inhibitors of polyamine uptake (Zhu et al. 1999). Antizyme 2 enhanced ODC degradation in insect cells strongly overexpressing antizyme, but it had little or no ability to drive proteosomal degradation of ODC in vitro. Antizyme 2 could be suitable to act as a reversible inhibitor of ODC activity (Zhu et al. 1999). In this way, it might store the enzyme for future use. Antizyme 3 binds and inactivates ODC (Ivanov et al. 2000c), but no further information about its biochemical properties has yet emerged. It is apparently needed in late spermatogenesis for temporarily restricted control of polyamine production.
The most important factor regulating antizyme synthesis is cellular polyamine concentration. Polyamines induce antizyme by a rare mechanism; by programmed ribosomal frameshifting (Matsufuji et al. 1995). The antizyme mRNA contains two overlapping open reading frames. The second of these encodes most of the protein, but lacks an initiation codon. Translation initiates in reading frame 1, must shift to reading frame 2 for production of functionally active antizyme. Polyamines increase the efficiency of antizyme mRNA frameshifting just before translation would otherwise terminate in reading frame 1. The site of frameshifting in mammalian antizyme 1, as well as antizyme 2, is UCCUGA, where quadruplet translocation occurs at UCCU to shift reading to +1 frame before the UGA stop codon (Matsufuji et al. 1995, Ivanov et al. 1998b). For the frameshifting to occur efficiently, it is important that the 3’base of the quadruplet is the first base of the stop codon. Other important features are a pseudoknot just 3’ of the shift site and a specific sequence 5’ of the shift site (Matsufuji et al. 1995, Ivanov et al. 1998b, Ivanov et al. 2000a). A pseudoknot 3’ of shift site is a common stimulator for eukaryotic –1 frameshifting, but the synthesis of antizyme is the only known case utilizing it in +1 frameshifting (Ivanov et al. 2000b). None of these cis-acting sequences appears to mediate polyamine-specific induction. Instead, the polyamine induction of frameshifting seems to be mediated directly by the translational machinery itself, i.e. the ribosome or some component of it (Matsufuji et al. 1995, Ivanov et al. 2000a).
Frameshifting in antizyme mRNA is remarkably conserved during evolution in yeasts, molds, fungi, nematodes, insects and vertebrates (Ivanov et al. 2000a). The frameshifting sequence is also relatively well conserved. When putative antizyme sequence was analysed from 26 organism, in three somewhat distantly related nematodes the frameshifting sequence was UUU UGA instead of the usual UCC UGA. In the two molds and one fungus the shift site was CCC UGA. Putrescine, spermidine and spermine are all able to induce frameshift when rat antizyme mRNA was translated in vitro in rabbit reticulocyte lysate (Matsufuji et al. 1995). The optimal concentration was lowest for spermine, 0.12 mM, spermidine induced frameshifting optimally at concetration 0.8 mM and putrescine at 4 mM. The optimal concentrations of spermidine and spermine are in physiological ranges. Polyamines can bring the efficiency of frameshifting up to about 30 % and stimulate it over 10-fold compared to the level with no added exogenous polyamines in the reticulocyte lysate.
Otherwise not so much is known about the regulation of antizyme. Interleukin-1 induced up-regulation of antizyme mRNA has been detected in a human melanoma cell line and was shown to be due to elevated antizyme gene transcription (Yang et al. 1997). Polyamine depletion has been observed to markedly decrease transcription of antizyme 1 gene, but high polyamine concentrations did not have an effect on the steady-state levels of antizyme mRNA (Nilsson et al. 1997). Osmotic stress has been reported to induce changes in cellular levels of antizyme, hypotonic growth medium decreased and hypertonic increased the amount of antizyme (Mitchell et al. 1998a). This regulation appears to be mediated via alterations in antizyme stability. In one study, antizyme expression was reported to be controlled according to cell cycle (Bettuzzi et al. 1999).
A very interesting, but still poorly known, phenomenon in the regulation of antizyme is the antizyme inhibitor protein. It was detected at first from rat liver extracts as a protein that re-activated antizyme-inactivated ODC by replacing ODC in an ODC-antizyme complex (Fujita et al. 1982). It has higher affinity for ODC than antizyme (Kitani & Fujisawa 1989, Murakami et al. 1989). Its molecular weight is similar to ODC and both rat (Murakami et al. 1996) and human (Nilsson et al. 2000a) antizyme inhibitors are nearly 50 % identical to corresponding ODC, but display no enzyme activity. Expression of antizyme inhibitor seems to be growth-related, since inhibitor was rapidly induced in growth-stimulated mouse fibroblasts (Nilsson et al. 2000a) and its gene was identified by differential display as one of those genes showing enhanced expression in gastric tumors compared to normal gastric tissue (Jung et al. 2000).
Very recently a cloning of a novel ODC-like protein expressed in brain and testis was reported (Pitkänen et al. 2001). This human protein is 54 % identical to ODC and 45 % identical to antizyme inhibitor. Its putative antizyme-binding region is more similar to the antizyme-binding region of antizyme inhibitor than to that of ODC. This together with the facts that its expression in CHO moderately but clearly increases ODC activity in cell lysates, but not in reticulocyte lysates when expressed in vitro , suggests that the novel protein either is able to release ODC from the ODC-antizyme complex or possess weak ODC activity itself. It might be another antizyme inhibitor or possess some other still unknown activities or functions, which are regulated by antizyme.