京都大学 / 新制・課程博士 / 博士(医科学) / 甲第24971号 / 医科博第153号 / 新制||医科||10(附属図書館) / 京都大学大学院医学研究科医科学専攻 / (主査)教授 村川 泰裕, 教授 竹内 理, 教授 松田 道行 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Tumor suppressor p53, the guardian of the genome, has been intensely studied molecule owing to its central role in maintaining cellular integrity. While the level of p53 protein is maintained low in unstressed conditions, there is a rapid increase in the functional p53 protein levels during stress conditions. It is now well documented in literature that p53 protein accumulates in the cells following DNA damage by posttranslational modifications leading to increased stability and half life of protein. Additionally, recent studies have also highlighted the significance of increased p53 translation during stress conditions. Interestingly, an alternative initiation codon has been shown to be present within the coding region of p53 mRNA. Translation initiation from this internal AUG results in an N-terminally truncated p53 isoform, described as ΔN-p53. However, the mechanisms underlying co-translational regulation of p53 and ΔN-p53 are still poorly understood. Studies have suggested that synthesis of both p53 and its ΔN-p53 isoform is regulated during cell cycle and also stress and cell-type specific manner. Interestingly, reports also demonstrate continued synthesis of both p53 isoforms during stress conditions. In contrast, global rates of cap-dependent translation initiation are shown to be reduced during stress conditions. This translation attenuation is observed mainly due to restricted availability of critical initiation factors. Interestingly, preferential synthesis of a vital pool of survival factors persists even during these circumstances. Studies have suggested that this selective translation is mediated via alternative mechanisms of translation initiation. One of the important mechanisms used for protein synthesis during these conditions is internal initiation. In this mechanism, the ribosomes are recruited to a complex RNA structural element known as ‘Internal Ribosome Entry Site (IRES)’, generally present in the 5’ untranslated region (UTR) of mRNA. Therefore, it is possible that the translation of p53 and ΔN-p53 could also be regulated by IRES mediated translation, especially during stress conditions. In this thesis the role of internal initiation in translational control of p53 and ΔN-p53 has been investigated. Additionally, the putative secondary structure of p53 IRES RNA has been determined. Further, it has been shown that polypyrimidine tract binding (PTB) protein acts as an important regulator of p53 IRES activities. The probable mechanism of action of PTB protein has also been investigated. The results suggest that interaction with PTB alters the p53 IRES conformation which could facilitate translation initiation. Finally, the possible physiological significance of existence of p53 IRES elements has been addressed. In the first part of the thesis, the presence of internal ribosome entry site within p53 mRNA has been investigated. As a first step, the 5’UTRs mediating the translation of both p53 and ΔN-p53 were cloned in the intercistronic regions of bicistronic constructs. Results of in vivo transfection of these bicistronic constructs suggested the presence of two IRES elements within p53 mRNA, with activities comparable to known viral and cellular IRESs. The IRES directing the translation of p53 is in the 5'-untranslated region of the mRNA, whereas the IRES mediating the translation of ΔN-p53 extends further into the protein-coding region. To further validate, stringent assays were performed to rule out the possibility of any cryptic promoter activity, re-initiation/scanning or alternative splicing in the p53 mRNA. Transfection of in vitro synthesized bicistronic RNAs confirmed the presence of IRES elements within p53 mRNA. Incidentally, this constitutes the first report on translational control of p53 by internal initiation. In the second part of the thesis, the secondary structure of p53 IRES RNA has been investigated. Structural analysis of p53 RNA was performed using structure-specific nucleases and modifying chemicals. The results obtained from chemical modification and nuclease probing experiments were used to constrain Mfold predicted structures. Based on this, a putative secondary structure model for p53 IRES RNA has been derived. Sequence alignment suggested that the p53 IRES RNA showed significant sequence conservation across mammalian species. To study the effect of mutations on the IRES structure, mutant p53 IRESs were used that harbor silent mutations at critical locations within the p53 IRES element. Incidentally, one of the mutant constructs used in the study was observed to be a naturally occurring mutation in a chronic lymphocyte leukemia patient. RNA structure analyses of these two mutant p53 IRES RNAs were performed. The nuclease mapping data suggested conformational alteration in these mutant RNAs with respect to wild type. Consistently, a comparative Circular-Dichroism spectroscopy of the Wt and mutant RNAs also validated the conformational alteration of the mutant RNAs. This also suggested that the presence of mutations in p53 IRES might result in decreased induction of p53 protein following DNA damage due to altered RNA structure. This might constitute as one of the mechanisms leading to tumor development in some types of cancers. In the third part of the thesis, the role of important cellular proteins that might modulate p53 IRES mediated translation has been studied. These cellular proteins act as IRES interacting trans-acting factors (ITAFs). Polypyrimidine tract binding (PTB) protein is an important ITAF implicated in regulating IRES mediated gene expression during apoptosis. It was observed that PTB protein specifically interacts with both the IRES elements within p53 mRNA. Interestingly, the affinity of interaction of PTB protein with both p53 IRES RNAs was observed to be significantly different. In order to determine the contact points of PTB on p53 IRES, a foot-printing assay using structure specific nuclease and recombinant-PTB protein was performed on p53 RNA. The data from foot-printing as well as primer extension inhibition assay (toe-printing analysis) suggested the presence of multiple PTB binding sites on p53 IRES RNA. Based on these results, a deletion mutant was generated that showed reduced PTB binding and also reduced IRES activity as compared to wild type. Further, to study the role of PTB in mediating p53 translation, the expression of PTB gene was partially silenced by using PTB specific siRNA. Partial depletion of endogenous PTB protein showed a significant decrease in the p53 IRES activities. These results suggest that PTB protein is essential for the p53 IRES activities. To understand the probable mechanism by which PTB regulates p53 IRES mediated translation, CD spectroscopy analysis of p53 IRES RNA was performed in the absence and presence of PTB protein. Interestingly, CD spectra analysis of the p53 RNA in the presence of PTB suggested a specific conformational change in p53 IRES, which might probably facilitate ribosome loading during internal initiation. This also suggests that abnormal expression of p53 ITAFs might lead to reduced p53 induction following DNA damage conditions. It could also be another event leading to malignant transformation of cells bearing wild type p53. It is highly tempting to speculate that the levels of p53 ITAFs could also be used as tumor biomarkers. In the fourth part of the thesis, the physiological relevance of existence of IRES elements within p53 mRNA has been investigated. The levels of p53 and ΔN-p53 proteins are known to be regulated in a cell cycle phase-dependent manner. The IRES activities of both p53 IRES elements were investigated at different phases of cell cycle. The activity of the IRES responsible for translation of p53 protein was found to be highest at G2-M transition and the maximum IRES activity corresponding to ΔN-p53 synthesis was observed at G1-S transition. These results suggested that the p53 IRES activities are regulated in a cell-cycle phase-dependent manner. Next, the regulation of p53 IRES mediated translation during stress conditions was studied. Human lung carcinoma cell line, A549 cells (that endogenously express both the p53 isoforms), were exposed to DNA damaging drug, doxorubicin. The level of p53 protein was observed to increase in a time-dependent manner. Interestingly, PTB protein, which is predominantly nuclear, was found to translocate to the cytoplasm during stress condition in a time-dependent manner. Under similar conditions, p53 protein was observed to reverse translocate from the cytoplasm to nucleus, probably to function as a transcription factor. Next, the influence of partial PTB silencing on p53 isoforms in the presence of cell stress (mediated by doxorubicin) was investigated. The data indicated reduced levels of both p53 and ΔN-p53 when PTB gene expression was partially silenced. These observations constitute “the proof of concept” that relative abundance of an ITAF, such as PTB protein, might contribute to regulating the coordinated expression of the p53 isoforms. The thesis reveals the presence as well as the physiological relevance of existence of IRES elements within p53 mRNA. The novel discovery of p53 IRES elements may provide new insights into the underlying mechanism of translational regulation. The modulation of the p53 IRES activities by PTB protein suggests that the regulated expression of p53 isoforms depends on the integrity of IRES elements and availability of cellular proteins that can serve as p53 ITAFs. Thus, studies pertaining to the identification of mutations within p53 IRES region as well as abnormal expression of p53 ITAFs such as PTB in cancer cells may have far reaching implications. These studies might lead to further advances in the field of cancer detection, prognosis and design of novel therapeutic strategies.

p53 is a nodal tumor suppressor protein that acts as a major defense against cancers. Approximately 50% of human tumours have mutations in p53 gene. Among its myriad features, the most distinctive is the ability to elicit both apoptotic death and cell cycle arrest. p53 has several isoforms. Most of them are produced by either internal promoter activity of the gene or alternate splicing of the pre-mRNA. Apart from these mechanisms, p53 mRNA has also been shown to be translated into two isoforms, the full-length p53 (FL-p53) and a truncated isoform ΔN-p53, which acts as a dominant-negative inhibitor of FL-p53. Under conditions of cellular stress, the canonical mode of translation initiation is compromised. To maintain the synthesis of proteins important for cell survival and cell-fate decisions, a subset of cellular mRNAs utilizes a non-canonical mode of translation initiation. The 5’ untranslated region of these mRNAs are highly structured and function as Internal Ribosome Entry Site (IRES). Previously, from our laboratory it has been shown that translation of p53 and its N-terminally truncated isoform ΔN-p53 can be initiated by IRES mediated mechanism. IRES mediated translation of ΔNp53 was maximum at G1-S phase but that of FL-p53 was maximum at the G2-M phase. Interestingly in case of a human genetic disorder X-linked dyskeratosis congenita (X-DC), aberrant IRES mediated p53 translation has been reported. It has also been reported that during oncogenic induced senescence (OIS) a switch between cap-dependent to IRES meditated translation occurs in p53 mRNA. From our laboratory, we have also demonstrated that polypyrimidine tract binding protein (PTB) positively regulates the IRES activities of both the p53 isoforms by shuttling from nucleus to the cytoplasm during genotoxic stress conditions. It is very important to understand how these two isoforms are regulated and in turn control the cellular functions. In the first part of the thesis, to investigate the importance of the structural integrity of the cis acting elements within p53 RNA, we have compared the secondary structure of the wild-type RNA with cancer-derived silent mutant p53 RNAs having mutations in the IRES elements such as L22L (CTA to CTG) a natural cancer mutation and Triple Silent Mutation (mutations were present at the wobble position of codon 17, 18, 19). These mutations result in the conformational alterations of p53 IRES RNA that abrogates the IRES function ex vivo significantly. It appears that these mutant RNAs failed to bind some trans-acting factors (p37, p41/44 etc) which might be critical for the IRES function. By super-shift assay using anti hnRNPC1/C2 antibody, we have demonstrated that the TSM mutant showed reduced binding to this protein factor. Partial knockdown of hnRNP C1/C2 showed significant decrease in p53 IRES activity and reduced synthesis of ΔN-p53. Also we have showed that introducing compensatory mutations in TSM mutant RNA rescued the secondary structure as well as function of p53 IRES. Further, the role of another silent point mutation in the coding sequence of p53 was investigated. Silent mutation (CCG to CCA) at codon 36 (P36P) showed decreased IRES activity. The mutation also resulted in differential binding of cellular proteins. Taken together, our observations suggest pivotal role of some specific trans acting factors in regulating the p53-IRES function, which in turn influences the synthesis of different p53 isoforms. In the second part of the thesis, p53 IRES RNA interacting proteins were identified using RNA affinity approach. Annexin A2 and PTB associated Splicing Factor (PSF/SFPQ) were identified and their interaction with p53 IRES RNA in vitro and ex vivo was studied. Interestingly, in the presence of Ca2+ ions Annexin A2 showed increased binding with p53 IRES. By competition UV crosslinking we have showed Annexin A2 and PSF interact specifically with p53 IRES. Toe printing assay results showed the putative contact points of Annexin A2 and PSF proteins on p53 IRES RNA. Interestingly, both proteins showed extensive toe-prints in the neighbourhood of the initiator AUG region of p53. Further, competition UV-crosslinking reveals the interplay of these two proteins. Annexin A2 and PSF appear to compete each other for binding with p53 IRES. PSF is known to interact with PTB protein. Since PTB also interacts with p53 IRES and positively regulates the translation, we wanted to study the interplay between PTB and PSF proteins binding with p53 IRES. To address this, we have performed competition UV crosslinking experiment and showed that increasing concentrations of PTB decreases PSF and p53 IRES interaction. However, increasing concentrations of PSF does not decrease or increase in PTB p53 IRES interaction. Results suggest that both Annexin A2 and PSF proteins play important role in regulation of p53 IRES activity. To address the physiological role of Annexin A2 and PSF proteins on p53 IRES activity, these proteins were partially knocked down in cellulo. This in turn showed decrease in p53 IRES activity in dual luciferase assays as well as in the steady state levels of both the p53 isoforms in transient transfection experiments. Heightened or continued expression of p53 protein is very important under stress where IRES-dependent translation supersedes normal cap-dependent translation. Results showed that expression of Annexin A2 under doxorubicin and thapsigargin induced stress are important for maintenance of both p53 IRES activity and steady state levels of p53 isoforms. Earlier from our laboratory we have showed that the IRES responsible for ∆N-p53 translation is active at G1/S phase while the IRES responsible for full length p53 translation is active at G2/M phase. Subcellular localization of the trans-acting factors plays a pivotal role in regulation of IRES activity of cellular mRNA. In this context we wanted to study the nuclear and cytoplasm localization of Annexin A2 under different cell cycle stages. We have seen Annexin A2 protein is dispersed in nucleus and cytoplasm at G1/S boundary, but post-G2 phase it moved from nucleus to cytoplasm. Further we wanted to investigate the effect of Annexin A2 and PSF on expression of p53 transactivated genes. Partial knock down of Annexin A2 and PSF proteins showed decrease in p21 luciferase activity. By real-time PCR analysis, we have also showed decrease in expression of different p53 targets upon silencing of Annexin A2 protein. Taken together, our observations suggest pivotal role of cis acting and trans-acting factors in regulating the p53-IRES function, which in turn influences the synthesis of p53 isoforms.