Новый Компонент Parafibromin/Hyrax

Parafibromin/Hyrax Activates Wnt/Wg Target Gene Transcription by Direct Association with β-catenin/Armadillo Christian Mosimann, George Hausmann, and Konrad Basler Cell. - 2006.- V.125, No 2, P. 324-341\| DOI10.1016/j.cell.2006.01.053
The Wnt pathway controls cell fates, tissue homeostasis, and cancer. Its activation entails the association of β-catenin with nuclear TCF/LEF proteins and results in transcriptional activation of target genes. The mechanism by which nuclear β-catenin controls transcription is largely unknown. Here we genetically identify a novel Wnt/Wg pathway component that mediates the transcriptional outputs of β-catenin/Armadillo. We show that Drosophila Hyrax and its human ortholog, Parafibromin, components of the Polymerase-Associated Factor 1 (PAF1) complex, are required for nuclear transduction of the Wnt/Wg signal and bind directly to the C-terminal region of β-catenin/Armadi!!o. Moreover, we find that the transactivation potential of Parafibromin/Hyrax depends on the recruitment of Pygopus to β-catenin/Armadillo. Our results assign to the tumor suppressor Parafibromin an unexpected role in Wnt signaling and provide a molecular mechanism for Wnt target gene control, in which the nuclear Wnt signaling complex directly engages the PAF1 complex, thereby controlling transcriptional initiation and elongation by RNA Polymerase II.	Развитие и гомеостаз многоклеточных организмов в основном зависит от межклеточных коммуникаций для установления и координации клеточных судеб и контроля за ростом органов. Определенные сигнальные молекулы формируют концентрационные или activity градаиенты, чтобы задать клеткам в ткани позиционную информацию. Члены семейств Hedgehog и Wnt секретируют сигнальные белки, напр., индуцируют субнабор генов мишеней на определенном расстоянии от их источника рутем активации законсервированных трансдукционных каскадов. Аберрантная активация этих путей может приводить к разного типа рака у людей (Taipale and Beachy, 2001; rev. Gregorieff and Cievers, 2005). Канонический Wnt/Wingless (Wg) путь контролирует транскрипцию генов мишеней посредством своего центрального компонента β-catenin (или его гомолога у Drosophila Armadillo, Arm), который служит в качестве центра (hub) взаимодействия для сложного набора из ассоциированных с мембраной, цитозольных и ядерных белков (rev. Daniels et al., 2001). На клеточной мембране, β-catenin ассоциирует с α-catenin и E-cadherin в слипчивых соединениях, чтобы обеспечить функцию клеточной адгезии. Оставшийся цитоплазматический белок β-catenin постоянно маркируется для протеосомной деградации с помощью Glycogen Synthase Kinase 3β (GSK-3β)-обусловленного фосфорилирования (re. Peifer and Polakis, 2000). GSK-3β является частью деградационного комплекса, включающего также Axin и APC, продукт гена опухолевого супрессора Adenomatous Polyposis Coli. После связывания внеклеточного Wnt диганда с членами Frizzled (Fz) и LRP/Arrow рецепторов, Dishevelled (Dsh) активируется и как предполагается нарушает комплекс деградации β-catenin. В результате уровни свободного цитоплазматического β-catenin возрастают и β-catenin проникает в ядро и взаимодействует с TCF/LEF ДНК-связывающими белками, чтобы инициировать транскрипцию чувствительных к Wnt генов (rev. Nelson and Nusse, 2004). Ключом к пониманию эволюции и функции передачи сигналов β-catenin, а значит и аберрантной активности β-catenin при раке становится вопрос, как компонент клеточной адгезии перемещается в ядро и может способствовать и поддерживать транскрипцию генов. Хотя выявлен ряд взаимодействующих белков, всё ещё недостаточно ясно, как они наделяют β-catenin способностью контролировать транскрипуию его мишеней. Напр., недавнее открытие и характеристика of BCL9/Legless (Lgs) and Pygopus (Pygo) показало, что Lgs связывает Pygo с β-catenin, чтобы сформировать стержень комплекса с TCF/LEF (Kramps et al., 2002; Hoffmans et al., 2005), но как этот комплекс использует Pygo, чтобы контролировать транскрипцию Wnt мишеней остается загадочным. β-catenin представлен центральным регионом, содержащим 12 составленных в ряд Arm повторов , фланкированных структурно не так хорошо определенными N- и C-терминальными доменами (NTD и CTD). Область Arm повторов обеспечивает связывание с cadherins, Axin/APC и TCF/LEF, что позволяет β-catenin ассоциировать со слипчивыми соединениями с их деструктивными комплексами или косвенно с ДНК, соотв. Различные линии экспериментальных доказательств указывают на то, что два главных исхода β-catenin, связанных с его транскрипционной активностью истекают из его наиболее N-терминально локализованных Arm повторов, а также из его С-терминальной области (Orsulic and Peifer, 1996; Hecht et al., 2000; Kramps et al., 2002). N-терминальная активность может быть связана с рекрутированием Lgs и Pygo (Hoffmans et al., 2005; Stadeli and Basler, 2005). Напротив, трансактивационная активность, обеспечиваемая С-терминальной частью, по-видимому, более сложна, т.к., по крайней мере, для Arm было установлено происхождение (to stem) нескольких субрегионов, из которых область ближайшая к Arm повтору 12 играет наиболее важную роль для пути выхода in vivo (Orsulic et al., 1996; Cox et al., 1999). Белки, участвующие в связывании этой С-терминальной области β-catenin/Arm включаютe histone acetyltransferases CBP и p300 и АТФ-зависимый гистон ремоделирующй фактор Brg1/Brahma (Brm) (Hecht et al., 2000; Takemaru and Moon, 2000; Barker et al., 2001), представляющие компоненты общего инструментального ящика для ремоделирования хроматина (rev. Narlikar et al., 2002). Помимо реорганизации хроматина, транскрипционная активация также зависит от рекрутирования и обработки RNA Polymerase II (RNAPII) для инициации и элонгации синтеза мРНК. Однако, для многих сигнальных каскадов, включая Wnt путь, механизмы, эти последние ступени транскрипционного контроля обеспечиваются, изучены плохо. В данном сообщении мы описываем генетическую идентификацию и молекулярную характеристику Hyrax (Hyx), нового компонента ядерного сигнального пути Wnt/Wg у Drosophila. Hyx является гомологом дрожжевого Cdc73p, компонента Polymerase-Associated Factor 1 (PAF1)комплекса, а ортологом Drosophila у человека является Parafibromin, белковый продукт Hyperparathyrotdism-Jaw Tumor syndrome (HTP-JT) гена опухолевого супрессора HRPT2. Избыточная экспрессия Hyx способна противодействовать скомпроментированной активности Lgs/Pygo. Напротив, снижение функции Hyx in vivo, как и knock-down hyx или HRPT2 экспрессии в культивируемых клетках, устраняет Arm/β-catenin-обусловленную передачу сигналов Wg и Wnt, соотв. Эти результаты выявляют новую и неожиданную роль для Parafibromin как важного компонента ядерной передачи сигналов Wnt. Более того, мы показали, что N-терминальная область Parafibromin/Hyx непосредственно взаимодействует с С-терминальной областью β-catenin/Arm. Полученные результаты указывают на механистическую модель для контроля Wnt/Wg генов мишеней, в которой Pygo-обусловленный и C-терминальный исходы β-catenin/Arm конвергенции в рекрутировании PAF1 комплекса, активируя тем самым инициацию и ээлонгацию транскрипции с помощью RNAPII. DISCUSSION Wnt signaling proteins govern many important cellular events during development, tissue homeostasis, and tu-morigenesis. In most cases, the effects of these signals are brought about by transcriptional changes of target genes. It is thought that Wnt targets are activated by β-catenin-mediated recruitment of auxiliary factors to TCF/LEF DNA binding proteins. Despite the identification of both dedicated and more general factors influencing the transcriptional output of β-catenin activity, little is known about how these proteins act together to control Wnt target gene expression. Here we extend the current understanding of Wnt/Wg signaling by the genetic identification and molecular characterization of Hyrax (Hyx) and its human ortholog Parafibromin, a novel transcriptional mediator of β-catenin/Arm. Parafibromin/Hyx is homologous to Cdc73p, a component of the Saccharomyces cerevisiae PAF1 complex, which Is an evolutionary conserved RNA Polymerase II interacting complex involved in regulating transcriptional initiation and elongation. The results Dresented here suggest a model for Wnt/Wg target gene :ontrol in which the C-terminal region of P-catenin recruits Darafibromin/Hyx, thereby engaging the PAF1 complex :or the transcriptional upregulation of target genes. Hyx As a Novel Wg Pathway Component Three lines of evidence argue for the notion that Hyx represents a component of the Wg pathway. (1) Our initial observation that increased expression of hyx can overcome the dominant-negative effect of overexpressed lgs^17E provides a first indication that Hyx positively influences Wg signaling outputs in vivo. lgs^17Eencodes an altered form of Lgs which contains a mutation in its Arm-interacting domain that severely decreases binding of Lgs to Arm and consequently the recruitment of Pygo to Arm. When provided in excess, Lgs^17E protein likely impairs the function of nuclear Arm by outcompeting endogenous Lgs and thus disturbs the sensitive balance and/or sequence of factors normally recruited at Wg-responsive enhancers. Elevating the levels of a positively acting nuctear factor involved in Wg signaling, in this case Hyx, could readily explain the reversion of the Lgs^17E phenotype in our genetic assays. (2) Our subsequent observation that genetic reduction of hyx function in imaginal discs as well as the RNAi-mediated knock-down of hyx expression in S2 cells caused a severe decrease in Wg pathway activity is a strong argument for a requirement of Hyx in Wg signaling. (3) Ultimate confirmation of the above genetic claims was our discovery of Hyx as a direct binding partner of Arm. Together these observations provide a solid basis for a model in which Hyx plays a key role in mediating the transcriptional output of Arm in response to Wg pathway activation. In contrast to the Arm partners Lgs and Pygo, Hyx is most likely not a component dedicated solely to the Wg pathway. The phenotypes associated with hyx loss-of-function mutations indicate that Hyx is involved in other developmental processes, possibly in the transcriptional output of some other signal transduction pathways). A Conserved PAF1-Like Complex Involved in Wnt Signaling The high degree of homology between Hyx and its single human ortholog suggested that Parafrbromin serves the same function in Wnt signaling as Hyx wi Wg signaling. Indeed, with the exception of genetic evidence for an in vivo requirement, equivalent lines of reasoning as those outlined above for Hyx argue for an important role of Parafi-bromin in human β-catenin signaling. What could this role be? It has recently been shown that Parafibromin/Hyx represents the Cdc73 subunit of a metazoan PAF1 complex (Rozenblatt-Rosen et al., 2005; Yart et al., 2005; Adelman et al., 2006). The yeast PAF1 complex has originally been found associated with initiating and elongating forms of RNAPII (reviewed by Sims etal., 2004). Moreover, the PAF1 complex interacts genetically and physically with the histone H2B ubiquitination complex, the Set1 methylase-containing COMPASS complex, and Set2, thus conferring control over a number of distinct histone modifications on RNAPII (reviewed by Hampsey and Reinberg, 2003; Krogan et al., 2003; Wood et al., 2003). Together, these findings suggest important conserved functions of the PAF1 complex in coordinating histone modifications "downstream" of chromatin preparation on target promoters to ensure proper initiation, elongation, and memory of transcription (reviewed by Sims et al., 2004). To date, Cdc73p has not been reported to interact directly or indirectly with a sequence-specific DMA binding transcription factor, and it is not clear how theP&F1 complex is recruited to its target genes. However, the meta-zoan homologs Parafibromin and Hyx share an extended N-terminal region, not present in Cdc73p, which we have found to physically interact with the core WntflVg component β-catenin/Arm. It is thus tempting to speculate that during metazoan evolution, Cdc73 homology proteins evolved in their N-terminal sequences interaction domains for certain signal transduction pathways, suchas the Wnt/ Wg pathway, while conserving C-terminal sequences for PAF1 complex and/or RNAPII association. Pygo As a Potential Coordination Module for Transcriptional Activators at the C Terminus of β-catenin/Arm β-catenin/Arm has two "branches" of transcriptional output, an N-terminal and a C-terminal branch, «Wch can be separated experimentally (Stadeli and Basler, 2005). The N-terminal activity maps to Arm repeat 1 and can be attributed to the recruitment of Lgs and Pygo (HoUnans et al., 2005; Stadeli and Basler, 2005). Our results here suggest that Parafibromin/Hyx mediates an important aspect of the C-terminal output of β-catenin/Arm. The significance of any transcriptional activity mapping to C-terminal sequences of β-catenin/Arm is seemingly undermined by the finding that C-terminally truncated forms of Arm (such as the product of the allele arm^XM19) are able to drive Wg target gene expression under certain experimental conditions (Cox et al., 1999; Tolwinski and Wieschaus, 2004). However, the arm^XM19 allele only erii&fts robust signaling activity when its product is "forced" into the nucleus by overexpression of a membrane-tethered form of Arm and most likely uses the N-terminal Lgs/Pygo-dependent branch for this activity. Under physiological conditions, Arm^XM19 is severely impaired for Wg signaling (Peifer et al., 1994; Cox et al., 1999). Arm^H8-6, which only lacks a distal portion of the CTD, retains residual transactivation potential at 18°C (Cox et al., 1999). This apparent correlation between signaling activity and the extent of C-terminal integrity of Arm might reflect the capacity of Arm to recruit Hyx, a view consistent with our protein-protein interaction results. Recent advances in the understanding of how transcriptional activators modulate gene transcription suggest a sequential recruitment of histone acetylases (such as CBP/p300) and chromatin-remodeling complexes (like SWI/SNF) to target genes before RNAPII is contacted to initiate transcription on the prepared chromatin (reviewed by Narlikar et al., 2002; Roeder, 2005). The β-catenin region encompassing Arm repeat 11 to the C terminus has previously been implicated in being necessary for chromatin remodeling using in vitro assays (Tutteretal., 2001). Parafibromin/Hyx interacts with a region of β-catenin/Arm (repeat 12-C) (Figure 6G) that overlaps with the CBP/p300 binding site (repeat 10/11-C) and the Brg-1/Brm binding region (repeat 7-12) (Hecht et al., 2000; Takemaru and Moon, 2000: Barker et al.. 2001). This raises the intriguing possibility of a concerted or sequential recruitment of chromatin remodeling factors during the control of Wnt/Wg-responsive genes to the C-terminal portion of β-catenin/Arm, as is being reported for an increasing number of transcription factors (reviewed by Narlikar et al.. 2002; Roeder, 2005). In such a scenario, CBP/p300 and Brg-1/ Brm would, in sequential or arbitrary order, mediate chromatin remodeling steps at β-catenin/Arm-dependent target genes before the Parafibromin/Hyx-mediated recruitment of a PAF1 -like complex orchestrates iater trans-activation steps involving the preparation of RNAPII with histone methylase complexes. In a final step, the PAF1 complex, including Parafibromin/Hyx, may be transferred from β-catenrn/Arm to RNAPII to travel with it through the actively transcribed gene. What role does the Wnt/Wg pathway component Pygo play in such a model? We found in several readouts that the Parafibromin/Hyx-enhanced transactivation activity of β-catenin is dependent on Pygo (Figures 1A-1F. 7C, and 7D). We interpret the Pygo-Parafibromin/Hyx dependence as an indication for a more general cross talk between Pygo and proteins interacting with the C-terminal region of β-catenin/Arm (Figure 7F). Thus, Pygo could act as a flexible recruitment module to facilitate the exchange or stabilization of transactivating complexes that sequentially bind to the β-catenin C terminus. We therefore propose that Wnt/Wg target gene activation might be a concerted, Pygo-guided process, which dynamically coordinates the sequential action of transcriptional modulators at the central scaffold protein β-catenin/Arm. Specificity The yeast PAF1 complex shows cotranscriptional association with a wide range of genes and has therefore been considered a general transcription cofactor complex. However, deletion of individual components of this complex does not have a global effect on mRNA transcription but instead has a more selective impact on the transcription of only a subset of genes (Shi et al., 1997; Porter et al., 2002). Currently, aside from our findings of an involvement in Wnt signaling, little is known about the target gene spectrum of metazoan PAF1-like complexes. Recently published data indicate that, as in yeast, the Drosophila PAF1-like complex is broadly associated with active genes but, functionally, Cdc73/Hyx seems only necessary for a subset of PAF1 complex targets (Adelman et al., 2006). This would be consistent with a view that Parafibromin and Hyx provide an adaptor function only to certain transcription factors, such as, for example, β-catenin and Arm. Indeed, our in vivo and in vitro assays indicate that in contrast to a cohort of other genes, whose expression is constitutive or controlled by other pathways, Wg targets are remarkably sensitive to reduction of Hyx levels. However, since our assays severely reduced but never abolished hyx expression, we can currently not evaluate the extent to which Hyx activity is also required for the transcription of targets of other pathways, which potentially are more resilient to reductions in Hyx levels. Hyperparathyroidism, Parafibromin, and Wnt Signaling Originally, HRPT2 has been cloned as the tumor suppressor gene responsible for HPT-JT in both familial and sporadic cases (reviewed by Wang et al., 2005). Little is known about the seemingly complex molecular pathology of parathyroid tumors and HPT-JT in particular (reviewed by Hendy, 2000). Constitutive Wnt signaling is causally involved in many different tumor types. However, our findings clearly indicate a positive role for Parafibromin in the Wnt pathway; loss-of-heterozygosity of the HRPT2 tumor suppressor gene is thus expected to compromise rather than enhance pathway output. One way to reconcile this apparent discrepancy, between a tumor suppressor function of HRPT2 and a positive sign of action of its product in Wnt signaling is to postulate that the transcriptional target(s) of Parafibromin relevant for HPT-JT encodes a negative growth regulator. It will be interesting to find out whether loss-of-function mutations in other positive Wnt pathway components might also cause hyperparathyroidism. It is also possible that the expression of the critical Parafibromin target(s) depends on another pathway that adopted the PAF1 complex for efficient target gene control. Clarification of these important issues will have to await the identification and analysis of the presumptive Parafibromin target(s) whose product(s) normally prevent parathyroid cancer.

Развитие и гомеостаз многоклеточных организмов в основном зависит от межклеточных коммуникаций для установления и координации клеточных судеб и контроля за ростом органов. Определенные сигнальные молекулы формируют концентрационные или activity градаиенты, чтобы задать клеткам в ткани позиционную информацию. Члены семейств Hedgehog и Wnt секретируют сигнальные белки, напр., индуцируют субнабор генов мишеней на определенном расстоянии от их источника рутем активации законсервированных трансдукционных каскадов. Аберрантная активация этих путей может приводить к разного типа рака у людей (Taipale and Beachy, 2001; rev. Gregorieff and Cievers, 2005).

Канонический Wnt/Wingless (Wg) путь контролирует транскрипцию генов мишеней посредством своего центрального компонента β-catenin (или его гомолога у Drosophila Armadillo, Arm), который служит в качестве центра (hub) взаимодействия для сложного набора из ассоциированных с мембраной, цитозольных и ядерных белков (rev. Daniels et al., 2001). На клеточной мембране, β-catenin ассоциирует с α-catenin и E-cadherin в слипчивых соединениях, чтобы обеспечить функцию клеточной адгезии. Оставшийся цитоплазматический белок β-catenin постоянно маркируется для протеосомной деградации с помощью Glycogen Synthase Kinase 3β (GSK-3β)-обусловленного фосфорилирования (re. Peifer and Polakis, 2000). GSK-3β является частью деградационного комплекса, включающего также Axin и APC, продукт гена опухолевого супрессора Adenomatous Polyposis Coli. После связывания внеклеточного Wnt диганда с членами Frizzled (Fz) и LRP/Arrow рецепторов, Dishevelled (Dsh) активируется и как предполагается нарушает комплекс деградации β-catenin. В результате уровни свободного цитоплазматического β-catenin возрастают и β-catenin проникает в ядро и взаимодействует с TCF/LEF ДНК-связывающими белками, чтобы инициировать транскрипцию чувствительных к Wnt генов (rev. Nelson and Nusse, 2004).

Ключом к пониманию эволюции и функции передачи сигналов β-catenin, а значит и аберрантной активности β-catenin при раке становится вопрос, как компонент клеточной адгезии перемещается в ядро и может способствовать и поддерживать транскрипцию генов. Хотя выявлен ряд взаимодействующих белков, всё ещё недостаточно ясно, как они наделяют β-catenin способностью контролировать транскрипуию его мишеней. Напр., недавнее открытие и характеристика of BCL9/Legless (Lgs) and Pygopus (Pygo) показало, что Lgs связывает Pygo с β-catenin, чтобы сформировать стержень комплекса с TCF/LEF (Kramps et al., 2002; Hoffmans et al., 2005), но как этот комплекс использует Pygo, чтобы контролировать транскрипцию Wnt мишеней остается загадочным.

β-catenin представлен центральным регионом, содержащим 12 составленных в ряд Arm повторов , фланкированных структурно не так хорошо определенными N- и C-терминальными доменами (NTD и CTD). Область Arm повторов обеспечивает связывание с cadherins, Axin/APC и TCF/LEF, что позволяет β-catenin ассоциировать со слипчивыми соединениями с их деструктивными комплексами или косвенно с ДНК, соотв. Различные линии экспериментальных доказательств указывают на то, что два главных исхода β-catenin, связанных с его транскрипционной активностью истекают из его наиболее N-терминально локализованных Arm повторов, а также из его С-терминальной области (Orsulic and Peifer, 1996; Hecht et al., 2000; Kramps et al., 2002). N-терминальная активность может быть связана с рекрутированием Lgs и Pygo (Hoffmans et al., 2005; Stadeli and Basler, 2005). Напротив, трансактивационная активность, обеспечиваемая С-терминальной частью, по-видимому, более сложна, т.к., по крайней мере, для Arm было установлено происхождение (to stem) нескольких субрегионов, из которых область ближайшая к Arm повтору 12 играет наиболее важную роль для пути выхода in vivo (Orsulic et al., 1996; Cox et al., 1999). Белки, участвующие в связывании этой С-терминальной области β-catenin/Arm включаютe histone acetyltransferases CBP и p300 и АТФ-зависимый гистон ремоделирующй фактор Brg1/Brahma (Brm) (Hecht et al., 2000; Takemaru and Moon, 2000; Barker et al., 2001), представляющие компоненты общего инструментального ящика для ремоделирования хроматина (rev. Narlikar et al., 2002). Помимо реорганизации хроматина, транскрипционная активация также зависит от рекрутирования и обработки RNA Polymerase II (RNAPII) для инициации и элонгации синтеза мРНК. Однако, для многих сигнальных каскадов, включая Wnt путь, механизмы, эти последние ступени транскрипционного контроля обеспечиваются, изучены плохо.

В данном сообщении мы описываем генетическую идентификацию и молекулярную характеристику Hyrax (Hyx), нового компонента ядерного сигнального пути Wnt/Wg у Drosophila. Hyx является гомологом дрожжевого Cdc73p, компонента Polymerase-Associated Factor 1 (PAF1)комплекса, а ортологом Drosophila у человека является Parafibromin, белковый продукт Hyperparathyrotdism-Jaw Tumor syndrome (HTP-JT) гена опухолевого супрессора HRPT2. Избыточная экспрессия Hyx способна противодействовать скомпроментированной активности Lgs/Pygo. Напротив, снижение функции Hyx in vivo, как и knock-down hyx или HRPT2 экспрессии в культивируемых клетках, устраняет Arm/β-catenin-обусловленную передачу сигналов Wg и Wnt, соотв. Эти результаты выявляют новую и неожиданную роль для Parafibromin как важного компонента ядерной передачи сигналов Wnt. Более того, мы показали, что N-терминальная область Parafibromin/Hyx непосредственно взаимодействует с С-терминальной областью β-catenin/Arm. Полученные результаты указывают на механистическую модель для контроля Wnt/Wg генов мишеней, в которой Pygo-обусловленный и C-терминальный исходы β-catenin/Arm конвергенции в рекрутировании PAF1 комплекса, активируя тем самым инициацию и ээлонгацию транскрипции с помощью RNAPII.

DISCUSSION

Wnt signaling proteins govern many important cellular events during development, tissue homeostasis, and tu-morigenesis. In most cases, the effects of these signals are brought about by transcriptional changes of target genes. It is thought that Wnt targets are activated by β-catenin-mediated recruitment of auxiliary factors to TCF/LEF DNA binding proteins. Despite the identification of both dedicated and more general factors influencing the transcriptional output of β-catenin activity, little is known about how these proteins act together to control Wnt target gene expression. Here we extend the current understanding of Wnt/Wg signaling by the genetic identification and molecular characterization of Hyrax (Hyx) and its human ortholog Parafibromin, a novel transcriptional mediator of β-catenin/Arm. Parafibromin/Hyx is homologous to Cdc73p, a component of the Saccharomyces cerevisiae PAF1 complex, which Is an evolutionary conserved RNA Polymerase II interacting complex involved in regulating transcriptional initiation and elongation. The results Dresented here suggest a model for Wnt/Wg target gene :ontrol in which the C-terminal region of P-catenin recruits Darafibromin/Hyx, thereby engaging the PAF1 complex :or the transcriptional upregulation of target genes.

Hyx As a Novel Wg Pathway Component Three lines of evidence argue for the notion that Hyx represents a component of the Wg pathway. (1) Our initial observation that increased expression of hyx can overcome the dominant-negative effect of overexpressed lgs^17E provides a first indication that Hyx positively influences Wg signaling outputs in vivo. lgs^17Eencodes an altered form of Lgs which contains a mutation in its Arm-interacting domain that severely decreases binding of Lgs to Arm and consequently the recruitment of Pygo to Arm. When provided in excess, Lgs^17E protein likely impairs the function of nuclear Arm by outcompeting endogenous Lgs and thus disturbs the sensitive balance and/or sequence of factors normally recruited at Wg-responsive enhancers. Elevating the levels of a positively acting nuctear factor involved in Wg signaling, in this case Hyx, could readily explain the reversion of the Lgs^17E phenotype in our genetic assays. (2) Our subsequent observation that genetic reduction of hyx function in imaginal discs as well as the RNAi-mediated knock-down of hyx expression in S2 cells caused a severe decrease in Wg pathway activity is a strong argument for a requirement of Hyx in Wg signaling. (3) Ultimate confirmation of the above genetic claims was our discovery of Hyx as a direct binding partner of Arm. Together these observations provide a solid basis for a model in which Hyx plays a key role in mediating the transcriptional output of Arm in response to Wg pathway activation. In contrast to the Arm partners Lgs and Pygo, Hyx is most likely not a component dedicated solely to the Wg pathway. The phenotypes associated with hyx loss-of-function mutations indicate that Hyx is involved in other developmental processes, possibly in the transcriptional output of some other signal transduction pathways).

A Conserved PAF1-Like Complex Involved in Wnt Signaling The high degree of homology between Hyx and its single human ortholog suggested that Parafrbromin serves the same function in Wnt signaling as Hyx wi Wg signaling. Indeed, with the exception of genetic evidence for an in vivo requirement, equivalent lines of reasoning as those outlined above for Hyx argue for an important role of Parafi-bromin in human β-catenin signaling. What could this role be? It has recently been shown that Parafibromin/Hyx represents the Cdc73 subunit of a metazoan PAF1 complex (Rozenblatt-Rosen et al., 2005; Yart et al., 2005; Adelman et al., 2006). The yeast PAF1 complex has originally been found associated with initiating and elongating forms of RNAPII (reviewed by Sims etal., 2004). Moreover, the PAF1 complex interacts genetically and physically with the histone H2B ubiquitination complex, the Set1 methylase-containing COMPASS complex, and Set2, thus conferring control over a number of distinct histone modifications on RNAPII (reviewed by Hampsey and Reinberg, 2003; Krogan et al., 2003; Wood et al., 2003). Together, these findings suggest important conserved functions of the PAF1 complex in coordinating histone modifications "downstream" of chromatin preparation on target promoters to ensure proper initiation, elongation, and memory of transcription (reviewed by Sims et al., 2004).

To date, Cdc73p has not been reported to interact directly or indirectly with a sequence-specific DMA binding transcription factor, and it is not clear how theP&F1 complex is recruited to its target genes. However, the meta-zoan homologs Parafibromin and Hyx share an extended N-terminal region, not present in Cdc73p, which we have found to physically interact with the core WntflVg component β-catenin/Arm. It is thus tempting to speculate that during metazoan evolution, Cdc73 homology proteins evolved in their N-terminal sequences interaction domains for certain signal transduction pathways, suchas the Wnt/ Wg pathway, while conserving C-terminal sequences for PAF1 complex and/or RNAPII association.

Pygo As a Potential Coordination Module for Transcriptional Activators at the C Terminus of β-catenin/Arm β-catenin/Arm has two "branches" of transcriptional output, an N-terminal and a C-terminal branch, «Wch can be separated experimentally (Stadeli and Basler, 2005). The N-terminal activity maps to Arm repeat 1 and can be attributed to the recruitment of Lgs and Pygo (HoUnans et al., 2005; Stadeli and Basler, 2005). Our results here suggest that Parafibromin/Hyx mediates an important aspect of the C-terminal output of β-catenin/Arm. The significance of any transcriptional activity mapping to C-terminal sequences of β-catenin/Arm is seemingly undermined by the finding that C-terminally truncated forms of Arm (such as the product of the allele arm^XM19) are able to drive Wg target gene expression under certain experimental conditions (Cox et al., 1999; Tolwinski and Wieschaus, 2004). However, the arm^XM19 allele only erii&fts robust signaling activity when its product is "forced" into the nucleus by overexpression of a membrane-tethered form of Arm and most likely uses the N-terminal Lgs/Pygo-dependent branch for this activity. Under physiological conditions, Arm^XM19 is severely impaired for Wg signaling (Peifer et al., 1994; Cox et al., 1999). Arm^H8-6, which only lacks a distal portion of the CTD, retains residual transactivation potential at 18°C (Cox et al., 1999). This apparent correlation between signaling activity and the extent of C-terminal integrity of Arm might reflect the capacity of Arm to recruit Hyx, a view consistent with our protein-protein interaction results.

Recent advances in the understanding of how transcriptional activators modulate gene transcription suggest a sequential recruitment of histone acetylases (such as CBP/p300) and chromatin-remodeling complexes (like SWI/SNF) to target genes before RNAPII is contacted to initiate transcription on the prepared chromatin (reviewed by Narlikar et al., 2002; Roeder, 2005). The β-catenin region encompassing Arm repeat 11 to the C terminus has previously been implicated in being necessary for chromatin remodeling using in vitro assays (Tutteretal., 2001).

Parafibromin/Hyx interacts with a region of β-catenin/Arm (repeat 12-C) (Figure 6G) that overlaps with the CBP/p300 binding site (repeat 10/11-C) and the Brg-1/Brm binding region (repeat 7-12) (Hecht et al., 2000; Takemaru and Moon, 2000: Barker et al.. 2001). This raises the intriguing possibility of a concerted or sequential recruitment of chromatin remodeling factors during the control of Wnt/Wg-responsive genes to the C-terminal portion of β-catenin/Arm, as is being reported for an increasing number of transcription factors (reviewed by Narlikar et al.. 2002; Roeder, 2005). In such a scenario, CBP/p300 and Brg-1/ Brm would, in sequential or arbitrary order, mediate chromatin remodeling steps at β-catenin/Arm-dependent target genes before the Parafibromin/Hyx-mediated recruitment of a PAF1 -like complex orchestrates iater trans-activation steps involving the preparation of RNAPII with histone methylase complexes. In a final step, the PAF1 complex, including Parafibromin/Hyx, may be transferred from β-catenrn/Arm to RNAPII to travel with it through the actively transcribed gene.

What role does the Wnt/Wg pathway component Pygo play in such a model? We found in several readouts that the Parafibromin/Hyx-enhanced transactivation activity of β-catenin is dependent on Pygo (Figures 1A-1F. 7C, and 7D). We interpret the Pygo-Parafibromin/Hyx dependence as an indication for a more general cross talk between Pygo and proteins interacting with the C-terminal region of β-catenin/Arm (Figure 7F). Thus, Pygo could act as a flexible recruitment module to facilitate the exchange or stabilization of transactivating complexes that sequentially bind to the β-catenin C terminus. We therefore propose that Wnt/Wg target gene activation might be a concerted, Pygo-guided process, which dynamically coordinates the sequential action of transcriptional modulators at the central scaffold protein β-catenin/Arm.

Specificity The yeast PAF1 complex shows cotranscriptional association with a wide range of genes and has therefore been considered a general transcription cofactor complex. However, deletion of individual components of this complex does not have a global effect on mRNA transcription but instead has a more selective impact on the transcription of only a subset of genes (Shi et al., 1997; Porter et al., 2002). Currently, aside from our findings of an involvement in Wnt signaling, little is known about the target gene spectrum of metazoan PAF1-like complexes. Recently published data indicate that, as in yeast, the Drosophila PAF1-like complex is broadly associated with active genes but, functionally, Cdc73/Hyx seems only necessary for a subset of PAF1 complex targets (Adelman et al., 2006). This would be consistent with a view that Parafibromin and Hyx provide an adaptor function only to certain transcription factors, such as, for example, β-catenin and Arm. Indeed, our in vivo and in vitro assays indicate that in contrast to a cohort of other genes, whose expression is constitutive or controlled by other pathways, Wg targets are remarkably sensitive to reduction of Hyx levels.

However, since our assays severely reduced but never abolished hyx expression, we can currently not evaluate the extent to which Hyx activity is also required for the transcription of targets of other pathways, which potentially are more resilient to reductions in Hyx levels.

Hyperparathyroidism, Parafibromin, and Wnt Signaling Originally, HRPT2 has been cloned as the tumor suppressor gene responsible for HPT-JT in both familial and sporadic cases (reviewed by Wang et al., 2005). Little is known about the seemingly complex molecular pathology of parathyroid tumors and HPT-JT in particular (reviewed by Hendy, 2000). Constitutive Wnt signaling is causally involved in many different tumor types. However, our findings clearly indicate a positive role for Parafibromin in the Wnt pathway; loss-of-heterozygosity of the HRPT2 tumor suppressor gene is thus expected to compromise rather than enhance pathway output. One way to reconcile this apparent discrepancy, between a tumor suppressor function of HRPT2 and a positive sign of action of its product in Wnt signaling is to postulate that the transcriptional target(s) of Parafibromin relevant for HPT-JT encodes a negative growth regulator. It will be interesting to find out whether loss-of-function mutations in other positive Wnt pathway components might also cause hyperparathyroidism. It is also possible that the expression of the critical Parafibromin target(s) depends on another pathway that adopted the PAF1 complex for efficient target gene control. Clarification of these important issues will have to await the identification and analysis of the presumptive Parafibromin target(s) whose product(s) normally prevent parathyroid cancer.

Parafibromin/Hyrax Activates Wnt/Wg Target Gene Transcription by Direct Association with β-catenin/ArmadilloChristian Mosimann, George Hausmann, and Konrad Basler Cell. - 2006.- V.125, No 2, P. 324-341| DOI10.1016/j.cell.2006.01.053

Parafibromin/Hyrax Activates Wnt/Wg Target Gene Transcription by Direct Association with β-catenin/Armadillo
Christian Mosimann, George Hausmann, and Konrad Basler
Cell. - 2006.- V.125, No 2, P. 324-341| DOI10.1016/j.cell.2006.01.053