| Ornithine decarboxylase: Expression and regulation in rat brain and in transgenic mice | ||
|---|---|---|
| Prev | Chapter 6. Discussion | Next |
We carried out this study in order to localize and compare ODC and antizyme expression in different brain regions. Polyamines have a plethora of demonstrated or suggested functions in CNS and polyamine metabolisms by CNS has characteristic features of its own, which is discussed in the previous chapter and in the chapters 1.1.3 and 1.1.4. Considering all interesting observations surprisingly little is known about the regulation of polyamine metabolism and the roles of ODC and antizyme in the normal adult brain. The amount of ODC protein in various tissues usually correlates well with enzyme activity, but in brains a noticeably high amount of immunoreactive ODC is present although enzyme activity is relatively low (Laitinen 1985). This may be due to the antizyme, which is present in the brain and could form a complex with the enzyme. The complex is formed efficiently at least when cells are lysed for ODC purification (Laitinen et al. 1986). However, in cell lysates and in cell lines of non-neural origin ODC in ODCantizyme complex is degraded extremely rapidly (Murakami et al. 1992b, Murakami et al. 1992c, Murakami et al. 1993). Brain might be exception to this rule or ODC and antizyme could be expressed in different locations in the CNS. Correlation of expression patterns of ODC and antizyme to known function of polyamines in CNS could as well be interesting and potentially informative. Actually we considered that the localization of ODC and antizyme expression is a definite prerequisite for the comprehensive understanding of the balanced regulation of polyamine metabolism in CNS.
Both immunohistochemistry and in situ hybridization experiments demonstrated that ODC and antizyme are widely expressed in the rat brain. Their expression was clearly confined to neurons, agreeing with earlier observations about ODC expression in rodent and human brains (Müller et al. 1991, Müller et al. 1993, Bernstein & Müller 1995, Ichikawa et al. 1997, Ichikawa et al. 1998, Gritli-Linde et al. 2001). Glial cells have reported to display ODC activity (Laube & Veh 1997) as expected due to the ubiquitous presence and essentiality of polyamines, but immunoreactivity for ODC is on the detectable level only in certain rare cases (Bernstein & Müller 1999).
The highest expression for both genes was detected in the cerebellar cortex, hippocampus, hypothalamic paraventricular and supraoptic nuclei, locus coeruleus, olfactory bulb, piriform cortex and pontine nuclei. Overall expression pattern matched well with those reported later by others (Gritli-Linde et al. 2001). There was no obvious correlation between ODC mRNA expression and enzyme activities determined in our previous study (I) apparently because the brains were dissected to five relatively roughly defined regions and because antizyme mRNA was nearly always expressed strongly concomitantly with ODC mRNA leaving the final controlling of protein expression levels to depend on translational and post-tranlational regulation.
Polyamines are responsible for the intrinsic gating and rectification of those subgroups of inward rectifying Kir channels which are strong or intermediate rectifyers such as Kir2 and Kir3 channels (Ficker et al. 1994, Lopatin et al. 1994). Distribution of Kir2 and Kir3 mRNAs in the adult rodent brain has been studied in details (Karschin et al. 1994, Kobayashi et al. 1995, Dissmann et al. 1996, Horio et al. 1996, Karschin et al. 1996, Töpert et al. 1998) and exhibits interesting similarities with the expression of ODC and antizyme mRNAs reported in the present study. We detected a very high expression of both ODC and antizyme mRNAs in nearly all those regions that have been reported to express Kir2 or Kir3 channels at the high level. The only major exception with Kir2 channels was thalamus, where relatively high levels of antizyme mRNA were observed, but ODC mRNA was expressed only moderately. Furthermore, expression of all Kir2 mRNAs, like ODC and antizyme mRNAs, is restricted to neurons, no signals for them have been detected in glial cells (Horio et al. 1996, Karschin et al. 1996). Except Kir3.4 mRNA which is expressed only at the low levels in the brain, the overall distribution of other Kir3 mRNAs is rather widespread and overlapping in many CNS neurons. There is some controversial data, but generally olfactory bulb, hippocampus, cortex, thalamus and cerebellum all exhibit a strong hybridization signal for Kir3.1, Kir3.2 and Kir3.3 mRNAs (Karschin et al. 1994, Kobayashi et al. 1995, Dissmann et al. 1996, Karschin et al. 1996). Again all these regions with the exception of thalamus showed high ODC and antizyme expression in our study. High and simultaneous ODC and antizyme mRNA levels suggest that polyamine concentrations are regulated strictly and rapidly and maybe in a wider range in these regions. This might imply that the controlling of polyamine levels is indeed used to regulate Kir channels in vivo. In the thalamus hybridization signal for antizyme mRNA concentration is strikingly higher than for ODC mRNA. Could this mean that in the thalamus a lower polyamine concentration and weaker inward rectifying is required or sufficient?
Expression levels for ODC mRNA were clearly lower than ones for antizyme mRNA in all the areas investigated, and the appropriate signal for ODC mRNA was obtained by using twice as long exposure time as for antizyme. Although ODC is regulated at the levels of transcription, translation and the degradation of enzyme protein (Davis et al. 1992, Shantz & Pegg 1999), the regulation of transcription plays nearly always a predominant role. When ODC expression is not stimulated the levels of ODC mRNA could be expected to be low. This is supported by the fact that ODC mRNA is relatively rapidly degraded with the half-life of 2.5 to 5 hours (Berger & Porter 1986, Weiner & Dias 1992). The principal regulation of antizyme expression takes place at the translational level where polyamines induce a programmed ribosomal frameshifting in antizyme mRNA (Coffino 2001a, Coffino 2001b). The synthesis of antizyme as a rapid response to increasing polyamine concentration can occur only if cells maintain a certain level of antizyme mRNA continuously and this has been shown to be the case, a large amount of antizyme mRNA is constitutively expressed in rat tissues and its half-life after actinomycin D treatment is as long as 12 hours (Matsufuji et al. 1990).
Consistently with others (Karschin et al. 1994, Kobayashi et al. 1995, Dissmann et al. 1996, Karschin et al. 1996) we detected immunoreactivity for ODC mainly in the cytoplasm. In most areas of brain antizyme staining was also localized in the cytoplasm, but interestingly in other areas including the cerebellar cortex, frontal cortex, cingulate cortex, cohlear nucleus and some nuclei in the thalamus, hypothalamus, and amygdala; antizyme protein was predominantly expressed in the nerve cell nuclei. Nuclear localization of antizyme was later confirmed by others employing different antibody (Gritli-Linde et al. 2001). They reported nuclear staining for antizyme in Purkinje cells of cerebellar cortex and, disagreeing with our observations, in the CA1-3 fields of hippocampus, but did not present a detailed report of overall expression pattern.
It is possible that in the regions where antizyme was localized to nuclei, it is rapidly degraded in the cytoplasm and the nuclear staining is predominant only because antizyme in the nuclei is not degraded with similar efficiency. Another possibility is that antizyme might have some unknown special function in the nucleus. It has appeared that the regulation and functions of antizyme are more complicated than previously thought. Antizyme has been shown to interact with signaling protein Smad1 and to target it together with, or potentially independently on ubiquitination for proteasomal degradation (Gruendler et al. 2001). Smad proteins mediate signaling induced by transforming growth factor β superfamily from cell mebrane receptors to nucleus where they translocate (Itoh et al. 2000). Furthermore, unpublished, but reviewed, results suggest that antizyme interacts also with the components regulating cell cycling, i.e. with cyclins and cyclindependent kinases (Coffino 2001b), which are nuclear proteins (Jordan et al. 2000, Ino & Chiba 2001, Small et al. 2001) Antizyme is devoid of a defined nuclear localization signal, but in the computerized analysis the similarity of antizyme amino acid composition to that of nuclear proteins gave evidence for a first choice location of antizyme to the nucleus with an expected prediction accuracy of 82.5 % (Gritli-Linde et al. 2001).
The recent studies have demonstrated that antizymes form a gene family. In the CNS in addition to antizyme 1, also antizyme 2 is expressed (Ivanov et al. 1998b). The oligonucleotide we used as a probe in in situ hybridization experiments corresponded to nucleotides 71 – 109 of rat antizyme mRNA (GenBank accession number D10706). This area is not homologous between mouse antizyme 1 and 2, and generally the nucleotide sequence of antizyme 1 is significantly different from that of antizyme 2. Apparently a probe specific for one antizyme mRNA would not hybridize with those of the other family members under stringent conditions. We cannot rule out the possibility that antibody against antizyme 1 recognises also antizyme 2, although similarity between mammalian antizymes 1 and 2 appears to be less than 70 %. Moreover, the fact that antizyme mRNA was found to be about 16 times more abundant in human tissues than that of antizyme 2 (Ivanov et al. 1998b), may suggest that antizyme 2 protein is unlikely to significantly contribute to the signal observed with the anti-antizyme 1 antibody.