Our data clearly demonstrate that the R2TP complex plays an essential role in stabilizing Nop58 in the process of box C/D snoRNP maturation (Figures 1 and 2). Intriguingly, this R2TP function is influenced by growth phases and nutrient conditions and is regulated by the subcellular localization of its protein components (Figures 5, 6, and 7).
The depletion or deletion of the R2TP components, except for Tah1, results in reduced Nop58 levels (Figure 2A and B). This result suggests that the main functional components of the R2TP complex for snoRNP biogenesis are Rvb1, Rvb2, and Pih1, while Tah1 could be an accessory protein. This is consistent with our previous result showing that Tah1 effect on snoRNP biogenesis is subtle, and that Tah1 functions with Hsp90 to maintain Pih1 protein stability . Although we observed a direct interaction of Rvb1/2 with Nop58 C-terminal domain in vitro (Figure 4C, left panel), the Nop58-FLAG pulldown experiment in pih1Δ showed a significant reduction of Rvb1/2 binding with Nop58, indicating that the interaction of the R2TP complex with Nop58 is largely mediated by Pih1 in vivo (Figure 1C). More specifically, we found that the N-terminal domain of Pih1, residues 1–230, binds to the C-terminal Nop domain of Nop58 (Figure 1D). Furthermore, Pih1(1–230) was also found to interact with Rvb1/2, but the interaction was significantly enhanced for Pih1(1–248) . We consider Pih1 to define the R2TP complex in yeast.
We found that the R2TP complex significantly interacts with the C-terminal domain of Nop58 spanning residues 285–447 (Figure 1E). No binding of snoRNP core proteins to this region of Nop58 was detected suggesting that R2TP interacts with Nop58 that has not been assembled with other snoRNP core factors. Unexpectedly, it was also observed that deletion of the KKE/D region from Nop58 C-terminus (residues 448–511) significantly reduced Nop58 interaction with the other snoRNP core proteins, suggesting that this charged region of Nop58 might contribute to the association between Nop58 and other snoRNP core factors. In yeast, the deletion of KKE/D does not significantly affect cell growth and box C/D snoRNA binding but causes a loss of the compaction of the nucleolus [47-49]. Also, it has been found that this highly charged domain of Nop58 interacts with Tgs1 in vitro; Tgs1 is a tri-methylguanosine synthase and is essential for hypermethylation of the 5' m7G cap of snRNAs and snoRNAs . Furthermore, in mammalian cells, the C-terminal domain of Nop58 containing the charged region is essential for the nucleolar localization of Nop58 and functions as a nucleolar localization signal (NoLS) . Taken together, these observations suggest an important role for the KKE/D region of Nop58 in snoRNP biogenesis. It would be important to further analyze the physiological role of this domain.
If the yeast and archaeal box C/D complexes have the same general arrangement [33,34], then the Nop domain of Nop58 has multiple binding partners including the snoRNA, Snu13, and Pih1. Furthermore, based on the recent X-ray crystal structure of the archaeal box C/D RNA-protein complex , it is clear that Nop58 is a key component of the complex providing an interaction platform for the other core proteins and for the snoRNA.
Recently, it has been reported that Pih1-Tah1 heterodimer interacts with Snu13-U14 snoRNA through Rsa1 in vitro . However, we were not able to identify an interaction between Snu13 and Pih1-Tah1 in vivo (Figure 1A). We also performed Rsa1-FLAG pulldowns using yeast soluble cell lysate but no interaction with R2TP components was detected (data not shown), suggesting that the Pih1-Tah1-Snu13-snoRNA-Rsa1 interaction might be weak or transient under physiological condition.
Intriguingly, the R2TP complex interacts with the unassembled form of Nop58 with high affinity (Figure 1A,B,E), but does not interact with either mature or premature box C/D snoRNA (Figure 3). These results suggest that R2TP is involved in a very early stage of the box C/D snoRNP biogenesis before Nop58 assembles with the other snoRNP components: Snu13, Nop1, Nop56, and snoRNAs. In a previous study, we showed that the deletion or depletion of R2TP components affects the accumulation of mature box C/D snoRNAs . This phenotype could be the result of the destabilization of Nop58. The interaction between R2TP and Nop58 is evident in log phase cells that require high production rate of snoRNPs for efficient ribosome biogenesis but not in stationary phase cells. Therefore, it is likely that the total box C/D snoRNP levels in the cell, and consequently pre-rRNA processing, could be effectively regulated by modulating Nop58 levels rather than other box C/D snoRNP core components. This would explain why the R2TP complex, by affecting the stability of Nop58, plays a critical role in cell physiology. Previously, we reported that R2TP is involved in snoRNP biogenesis in both log and stationary phases . As mentioned above, we observed a strong interaction between R2TP and Nop58 in log phase whereas it was weaker in stationary phase. However, this low level of binding could be sufficient in stationary phase cells to maintain the basal levels of Nop58 and, subsequently, conserve snoRNA production.
In mammalian cells, it has been reported that the treatment of human cells by proteasome inhibitor MG-132 leads to improper localization of Nop58 . Furthermore, post-translational modifications of mammalian Nop58 have also been reported [53,54]. The protein has been found to be phosphorylated by casein kinase II (CK2), and, subsequently, SUMOylated to increase its stability. These observations further highlight the importance of regulating Nop58 stability. It will be interesting to identify such modifications for yeast Nop58 and to determine how the modifications affect Nop58 stability, function, and Nop58-R2TP interaction in response to different growth and nutrient conditions.
Based on the localization studies (Figures 5C, 6A, and 7A,B,D), we unexpectedly found that the ability of the R2TP complex to modulate the assembly of box C/D snoRNPs is regulated by a nucleo-cytoplasmic shuttling mechanism that is dependent on the Kap121 and Crm1 karyopherins, which are also known to be involved in nucleo-cytoplasmic trafficking of ribosomal proteins [41,43]. The shuttling of the R2TP proteins between the nucleus and the cytoplasm is growth phase and nutrient dependent. In stationary phase cells (Figure 5C) or in the absence of carbon or nitrogen sources (Additional file 4: Figure S4), or in the presence of rapamycin (Figure 6A), we observed that some of the R2TP complex shuttles out of the nucleus, all core box C/D snoRNP protein levels, except for Snu13, are reduced (Figure 5A and 6B) and a larger proportion of the snoRNP proteins are present in the nucleoplasm (Figures 5C, 6A, and Additional file 4: Figure S4). In contrast, when nutrients are available, then the R2TP complex is enriched in the nucleus (Figure 5C), all core box C/D snoRNP protein levels were recovered (Figures 5A and 6B) and localized in the nucleolus (Figure 5C). As shown in Figure 8, this shuttling of the R2TP complex actively influences cell growth. It is known that ribosome synthesis is the major energy consuming process in the cell , and that ribosome activity is tightly coupled to growth phase and nutrient availability . Hence, the relocalization of the R2TP complex allows the cells to rapidly respond and control snoRNP biogenesis, and, subsequently, regulate pre-rRNA processing and ribosome biogenesis in response to different growth conditions.
It is interesting to note that Rix7, which is an essential AAA + ATPase required for the biogenesis and nuclear export of 60S ribosomal subunits, also undergoes growth phase redistribution . It localizes throughout the nucleus in exponentially growing cells, but concentrates in the nucleolus in stationary phase cells. Hence, the dynamic relocalization of protein complexes involved directly or indirectly in ribosome biogenesis might be a generally conserved mechanism designed to allow the cell to easily correlate the number of ribosomes to nutrient availability and growth phase.
It is known that ribosome biogenesis is regulated by the TOR signaling pathway , which balances the production of ribosome components to nutrient availability. We observed that the R2TP complex is regulated by the TOR signaling pathway since the localization of the R2TP proteins is affected by the specific inhibitor for TOR, rapamycin. At this stage, we do not know how the TOR pathway effects the nucleo-cytoplasmic translocation of the R2TP complex. However, one intriguing finding from our study points to the presence of a signaling-like pathway that links chaperone activity to ribosome biogenesis, which is based on stabilizing protein complexes (Additional file 6: Figure S6). In our earlier study , we had shown that the stability of Pih1 depends on the activity of Hsp90 chaperone together with its co-factor Tah1. This activity then results in the proper assembly of the R2TP complex. In this study, we demonstrated that the proper function and localization of the R2TP complex is required for the stability of Nop58 and assembly of the box C/D snoRNP complexes. The critical players in this pathway are Hsp90, Pih1, and Nop58; both Pih1 and Nop58 are unstable proteins. Importantly, there seems to be a directionality to this signaling-like pathway since functionally deficient Nop58 does not affect Pih1 (Figure 2C), while the destabilization or deletion of Pih1 does affect Nop58 (Figure 2A,B). The presence of such signaling-like pathways that are based on the stabilization of protein complexes rather than protein modification might be widespread in the cell and warrants further investigation.