Effects of Increased Polyamine Biosynthesis in Mouse Skin

To examine the effects of elevated ODC and polyamines on histone acetylation in a more relevant physiological system in which epithelial cells can still both proliferate and differentiate, unlike the immortalized cell lines studied in the earlier experiments, primary keratinocytes were isolated and cultured from K6/ODC transgenic and normal littermate skin (34). After labeling the keratinocytes with [3H]acetic acid in the presence of sodium butyrate, histones were isolated and examined by fluorography and by immunoblot analyses. In contrast to what was observed in immortalized cell lines, there was reduced overall acetylation of histones in K6/ODC keratinocytes as compared with control keratinocytes. This might be attributed to intrinsic differences between normal primary keratinocytes, which are programmed to terminally differentiate, and immortalized epidermal cell lines, which have undergone genetic changes that circumvent their ability to undergo differentiation, thus permitting continuous proliferation.

Elevated levels of ODC and polyamines at least partially block terminal differentiation in primary keratinocytes (38). However, culturing under conditions normally sufficient to induce differentiation (24 h in 0.12 mM calcium medium) had no effect on the reduced level of acetylation mediated by elevated levels of ODC. Moreover, DFMO treatment restored both the normal level of acetylated histones and the induction of the differentiation marker keratin 1 in K6/ODC transgenic keratinocytes triggered to differentiate by the addition of calcium. Our labeling experiments were done in the presence of sodium butyrate, effectively inhibiting HDAC enzymes, implying that polyamines must directly or indirectly inhibit the enzymatic activity of HAT enzymes in normal diploid keratinocytes.

The extent of histone acetylation is the net result of the action of multiple enzymes with acetyltransferase and deacetylase activity. Therefore, we assayed the HAT activity present in homogenates of skin from K6/ODC transgenic mice and their normal litter-mates, as well as spontaneous skin tumors from ODC/Ras double transgenic mice. Interestingly, increased HAT activity was observed in both the epidermis and dermis, as well as in intact skin of both K6/ODC and ODC/Ras mice (34). Significant, dramatically higher levels of HAT activity were measured in spontaneous tumors from ODC/Ras bigenic mice as compared with normal and K6/ODC skin, as well as non-tumor-bearing skin from the same ODC/Ras mice. Furthermore, as a result of DFMO treatment, HAT activity in ODC/Ras skin was reduced, and the aberrantly high HAT activity in the tumors was reduced to a level comparable to that detected in nontreated skin. Thus elevated HAT activity in both the skin and tumors of ODC transgenic mice is dependent on polyamine levels.

Similar to the situation for HAT enzymatic activity, HDAC activity was generally found to be increased in the skin of K6/ODC mice relative to normal littermate controls (34). Although there was no obvious change in the level of the predominant HDAC-1 protein in ODC transgenic mouse skin, the possibility that an altered level of one or more of the many other known HDAC enzymes is responsible for the elevated deacetylase activity in ODC skin cannot be ruled out. Interestingly, in contrast to HAT activity, HDAC activity in ODC/Ras tumors was typically lower than that of adjacent non-tumor-bearing skin. Treatment of ODC transgenic mice with DFMO led to decreased intrinsic HDAC activity in non-tumor-bearing skin. Thus, like HAT activity, HDAC activity is also susceptible to regulation by polyamines. However, HDAC activity in tumors was not found to be responsive to DFMO treatment.

The transcriptional coactivator proteins, CBP and p300, play a pivotal role in facilitating synergistic crosstalk between different signaling pathways. They interact with a multitude of nuclear proteins, bridging transcription factors with the basal transcription machinery (39). Moreover, they possess (and recruit proteins with) intrinsic HAT activity (39), altering the local chromatin structure to ultimately influence transcription at individual promoters. It is conceivable that polyamine-mediated modulation of CBP or p300 function could account for many of the diverse and broad effects that polyamines exert on cell growth and differentiation. Indeed, we found that there is increased HAT activity associated with immunoprecipitated CBP/p300 in extracts of K6/ODC trans-genic mouse skin relative to normal littermate skin (34). Interestingly, there is no apparent increase in CBP/p300-associated HAT activity in tumors compared with K6/ODC skin or non-tumor-bearing ODC/Ras skin. Furthermore, DFMO treatment has little effect on the CBP/p300-associated HAT activity in the tumors. Thus the high level of HAT activity found in tumors from ODC/Ras transgenic mice does not appear to be primarily contributed by CBP/p300 (or CBP/p300-associated PCAF), but by other HAT enzymes.

As a first step in characterizing the enzyme(s) responsible for the high level of HAT activity in ODC/Ras tumor lysates, in vitro HAT assays were done revealing a striking specificity of lysine acetylation. Using a panel of peptides preacetylated at various lysine residues, it was determined that Lys-12 is a preferred site of acetylation in histone H4 by the HAT activity present in the tumor extracts (34). In this regard, it is interesting that Lys-12 is consistently underused in monoacetylated histone H4 in several mammalian cell types, yet is frequently used in the more highly acetylated H4 isoforms typically associated with actively transcribed DNA (40). Thus the increased HAT activity in ODC transgenic mouse skin and tumors may be associated with hyperacetylation of histones and enhanced transcription at localized regions of the genome. Alternatively, the distinct preference for Lys-12 exhibited by the HAT activity in ODC transgenic tissue may reflect localized remodeling of chromatin into a more heterochromatin-like structure, perhaps mediated by HAT1. Nuclear HAT1 enzyme has been found to be required for telomeric silencing, mediated solely through Lys-12 acetylation (41). It is thought that acetylation of Lys-12 in histone H4 facilitates binding of silencing proteins that propagate formation of a specialized transcriptionally repressive chromatin structure (41,42). Thus the distinct preference for Lys-12 that is characteristic of ODC transgenic skin and tumors may be indicative of remodeling of localized regions of chromatin into a conformation that is prohibitive to gene transcription.

It is not likely a coincidence that polyamine-mediated enhancement of HAT function occurs in both perpetually growing immortalized cells (33) and in the highly proliferative cells in ODC/Ras tumors (34). In contrast to immortalized cells, normal keratinocytes undergo terminal differentiation after only a few cell divisions. This intrinsic difference in cellular programming probably underlies the opposite effects on histone acetylation observed in radiolabeled primary K6/ODC keratinocytes (34) as compared with cell lines overexpressing ODC (33). In any case, these varied observations indicate that polyamines can both promote and inhibit the acetylation of histones in epidermal cells, presumably via different molecular mechanisms. This fact, coupled with the demonstration that polyamines modulate both HAT and HDAC enzymes, support the notion that polyamines exert influence on multiple mechanisms controlling histone acetylation (Fig. 1). Neither purified p300 nor PCAF, representatives of two of several distinct classes of nuclear HAT enzymes, were found to be directly stimulated by putrescine or spermidine in cell-free assays (34). Therefore, if the situation with p300 and PCAF is reflective of other HAT enzymes in general, then polyamines likely modulate intrinsic HAT catalytic activity by indirect mechanisms. For instance, polyamines might indirectly regulate enzymatic activity by influencing signaling pathways responsible for the posttranslational modification of HAT or HDAC enzymes. Notably, ODC overexpression leads to greater overall protein phosphorylation in epidermal cells (43), suggesting a major impact on signaling pathways. In this regard, it is intriguing to note that the enzymatic activity and nuclear translocation of casein kinase 2, that includes transcription factors among its substrates, has been found to be stimulated by increased polyamine levels in epidermal cells (43). In addition to potentially affecting posttranslational modification of HAT/HDAC enzymes, polyamines might positively or negatively interfere with interactions between the various components of HAT or HDAC protein complexes. In addition, by modulating subcellular localization or rates of synthesis or turnover of HAT and HDAC proteins, polyamines could effectively manipulate the delicate balance between these enzymes.

Presumably, the abnormally high HAT enzymatic activity characteristic of ODC transgenic mouse skin and tumors is able to target histones in vivo, facilitating enhanced gene expression. However, just as histone deacetylase inhibitors do not promote global changes in gene transcription (44,45), the positive and negative regulation of nucleosomal acetylation exerted by polyamines might be expected to have functional consequences only at localized gene promoters, affecting only a subset of genes. Given that HAT enzymes acetylate non-histone proteins including transactivating factors and components of the basal transcriptional machinery (37,39), aberrantly high HAT activity resulting from ODC overexpression may exert additional major influence

Fig. 1. Potential modes by which polyamines may modulate chromatin remodeling. Transcriptionally competent DNA (vertical bars) comprises a very small amount of the total eukaryotic genome. Whether a gene is transcriptionally silent (white bars) or activated (gray bars) depends largely on the composition of multicomponent histone acetyltransferase/histone deacetylase (HAT/HDAC) protein complexes bound to its regulatory region. Polyamines may influence the function of these protein complexes through direct interaction with HAT or HDAC enzymes. Alternatively, they may indirectly affect enzymatic function by modulating posttrans-lational modifications, interaction, or the availability of components of the HAT/HDAC complexes. Thus, whether polyamines exert a positive or negative influence on histone acetylation at a localized promoter is likely to be dependent on the specific composition of the regulatory protein complexes governing expression of the gene.

Fig. 1. Potential modes by which polyamines may modulate chromatin remodeling. Transcriptionally competent DNA (vertical bars) comprises a very small amount of the total eukaryotic genome. Whether a gene is transcriptionally silent (white bars) or activated (gray bars) depends largely on the composition of multicomponent histone acetyltransferase/histone deacetylase (HAT/HDAC) protein complexes bound to its regulatory region. Polyamines may influence the function of these protein complexes through direct interaction with HAT or HDAC enzymes. Alternatively, they may indirectly affect enzymatic function by modulating posttrans-lational modifications, interaction, or the availability of components of the HAT/HDAC complexes. Thus, whether polyamines exert a positive or negative influence on histone acetylation at a localized promoter is likely to be dependent on the specific composition of the regulatory protein complexes governing expression of the gene.

on transcription-related processes. Indeed, we have preliminary evidence suggesting that elevated levels of polyamines affect the acetylation status of multiple non-histone proteins (data not shown). One example is the tumor suppressor protein, p53-over-expression of ODC in skin tissue leads to higher levels of acetylated p53 with concomitant increased binding of p53 to promoters, and associated increases in transcription, of p53 target genes (manuscript in preparation). The future challenge will be to dissect the various regulatory mechanisms modulated by changes in intracellular concentrations of polyamines within the context of individually affected genes. Identification of specific

HAT and HDAC enzymes and gene promoters that are targets of polyamine modulation will provide greater insight into the complex manner by which polyamines regulate his-tone acetylation and gene expression, and promote tumor development.

It is logical to speculate that the high intrinsic HAT and HDAC activities detected in ODC-overexpressing skin might influence the overall level of acetylated histones in the tissue chromatin. Western analyses using antibodies specific for acetylated histones revealed some modest differences in the acetylation of histones isolated from K6/ODC transgenic skin as compared with normal littermate skin (34). Although noticeable differences in acetylation were not always detected, whenever observed, the differences were consistent between independent preparations of isolated histones. In those cases, fewer acetylated histones H3 and H4 were detected in K6/ODC epidermis, whereas acetylation of histones H3 and H4 was increased in K6/ODC dermis and total skin as compared with histones in normal skin. Moreover, changes in the extent of acetylation of Lys-12 in histone H4 paralleled those observed for hyperacetylated histone H4. These modest changes in overall acetylation suggest that elevated intracellular levels of polyamines do not have a major effect on the acetylation of nucleosomal histones in bulk chromatin. That ODC overexpression induces effects on a small proportion of total histones is consistent with the notion that polyamines promote only very localized changes in the acetylation status of histones in chromatin, perhaps potentiating altered transcription of a small fraction of genes. Moreover, transcriptionally active chromatin actually comprises a very small proportion of the total genome. Thus, any localized effects of polyamines on the acetylation of histones in nucleosomes bound to gene promoters may not be detected within the context of bulk chromatin. This problem is compounded by the fact that recruitment of histone-modifying activities resulting in perturbation of chromatin structure is selectively targeted to specific regions of DNA that may span only one or two nucleosomes (17). Ultimately, the effect of polyamines on histone acetylation will be determined by the composition of the specific regulatory complexes that are recruited (2), as well as the subnuclear compartment in which the gene resides (46,47). Because polyamines apparently influence histone acetylation by multiple mechanisms, changes in cellular polyamine levels can potentially mediate decreased acetylation of nucleosomal histones at one gene promoter, while stimulating increased acetylation at another promoter, resulting in different transcriptional outcomes for those genes (Fig. 1). The ramification for tumor development is that increased polyamine biosynthesis might lead to transcriptional repression of one or more genes controlling cellular proliferation, while simultaneously activating transcription of genes involved in sustaining tumor growth and invasiveness.

Other transgenic mouse models have been developed that feature perturbations of polyamine levels. These include a mouse in which expression of the enzyme spermi-dine/spermine A^-acetyltransferase (SSAT) is under control of the natural SSAT promoter (48) and another in which the keratin 6 promoter is used to target SSAT overexpression to hair follicles (49). SSAT regulates the catabolism and export of intracel-lular polyamines. Activation of polyamine catabolism by overexpression of SSAT leads to the production of extremely high levels of acetylated spermidine in both of these models, whereas acetylated spermine remains undetectable. Huge amounts of putrescine also accumulate in relevant tissues in these mice, presumably because of the back conversion of spermidine as a consequence of SSAT overexpression, combined with the forward conversion of ornithine resulting from a compensatory increase in ODC enzyme activity. Because of this compensatory activation of polyamine biosynthesis, the pools of spermidine and spermine remain relatively unaffected in these SSAT mice, despite the production of large quantities of acetylated spermidine. Significantly, in SSAT mouse prostate, the heightened metabolic flux through the polyamine pathway leads to depletion of pools of acetyl-CoA, which is not only a cofactor for SSAT, but is also critically required for both fatty acid metabolism and the acetylation of histones and other nonhistone proteins. As such, SSAT transgenic mice might be useful for investigating the effects of competition for rate-limiting quantities of acetyl-CoA between enzymes that acetylate proteins and those that acetylate polyamines. Furthermore, given the similarities and differences between ODC and SSAT transgenic mice, comparison of the effect of their respective polyamine pathways on HAT and HDAC enzyme function in tissue might help elucidate the contribution of individual polyamines and their acetylated isoforms to regulating histone acetylation.

Early studies using crude tissue extracts or several "purified" acetyltransferases have reported overlapping specificities of these enzymes for histone and polyamine substrates (50,51). However, in addition to SSAT, the subsequent identification of several different classes of nuclear HAT enzymes, each comprising multiple distinct family members, calls the interpretation of these early studies into question. Notably, the existence of overlapping specificity could conceivably have implications for prolifer-ative diseases that feature revved up production of polyamines, should the excess polyamines successfully compete with protein substrates for rate-limiting HAT enzyme. Likewise, the altered expression of HAT enzymes often observed in tumors (36,37) could potentially disrupt the normal balance between nonacetylated and acetylated polyamines. The cloning and expression of individual acetyltransferases should now permit the unequivocal determination of substrate specificities of the various acetylating enzymes.

Sultry Skin

Sultry Skin

Get All The Support And Guidance You Need To Be A Success At Sultry Skin. This Book Is One Of The Most Valuable Resources In The World When It Comes To Having Beautiful Skin.

Get My Free Ebook


Post a comment