Before memories fade, here are some brief remarks on the science presented at the EMBO Conference “Ubiquitin and Ubiquitin-like modifiers in cellular regulation“, which was organized by EMBO and the RUBICON network in Riva del Garda. There were three and a half days worth of presentations (long ones and short ones), with three smallish poster sessions in the evening. I cannot provide details on every single talk – for some of them I didn’t take notes (and as I said in my previous post, nobody stuck to their abstracts). Most presenters showed at least some unpublished data, which is always good, but makes me unsure how appropriate it is to mention other people’s unpublished data on my blog. To be on the safe side, I will try to avoid revealing too much.
First session: Conjugation and Deconjugation
The first plenary speaker, Alex Varshavsky, did not show up, which gave us more time for breakfast. In a funny coincidence, the first three speakers in the Sunday morning session were all engaged in a collaborative project and noticed just before the meeting that they were going to give severely overlapping talks. Some discussions and last-minute changes during breakfast took care of that. First speaker was Dan Finley (Harvard), who was mainly talking about a new ubiquitin-binding subunit of the proteasome. He also discussed briefly some published work on ubiquitin chain editing by Hul5 and UBP6. Second speaker was yours truly, talking about – guess what – ubiquitin binding domains (with some emphasis on UBA and the new UBM and UBZ classes). Third speaker was Ivan Dikic (Frankfurt), talking about the same proteasome subunit as Dan, and also about UBZ domains in the NFkB regulator TAX1BP1. Fourth and last speaker in the morning session was Keith Wilkinson (Emory), talking about the sumo-1 vs. sumo-2 specificities of various sumo-proteases. It appears that most SENP proteases prefer Sumo-2, in particular the SENPs with the high numbers.
After the lunch break, Christopher Lima (Sloan-Kettering) talked about a structural and functional dissection of the domains in the PIAS sumo-ligases. He also discussed which sumoylation processes require an E3 and which ones don’t. Next speaker was Titia Sixma (Amsterdam) who talked about structural data on non-covalent interactions of E2 and E3 enzymes, which regulate ubiquitination. After another coffee, Frauke Melchior (Göttingen) talked about USP25, a known deubiquitinating protease that she found to be sumoylated. The first day was concluded by two short talks on the Cdc48/p97 ATPase system. Hemmo Meyer (Zürich) talked about the role of VCP/p97 (the human version) in regulating the re-formation of the nucleus after mitosis. The ubiquitin-dependent inactivation of the kinase Aurora-B is a central player. Finally, Thorsten Hoppe (Hamburg) talked about the role of Cdc-48 (the C.elegans version) in the turnover of the myosin-directed chaperone unc-45. This process apparently plays a role in human myopathy diseases.
The second day continued the session with a talk by Yoshinori Ohsumi (Okazaki). He discussed the two ubiquitin-like conjugation systems that are involved in autophagy: Atg12 and Atg8. It has been known for some time that Atg8 is special, as it is not conjugated to proteins but rather to lipids. A recent paper has shown that this conjugation step drives hemifusion of autophagic vesicles (see blog entry on the Daily Transcript). The role of the Atg12 conjugation onto Atg5 has been less clear, but there is evidence that the Atg12-Atg5 conjugate serves as an E3 for Atg8 conjugation. Next speaker was Ron Hay (Dundee), talking about different properties in sumo-1 and sumo-2 isoforms. He also discussed the ubiquitin ligase Rnf4, which contains sumo-binding motifs (SIMs) and specifically targets sumoylated proteins for degradation. I was not too happy to see all this data, as I am involved in a competing project studying this mechanism in yeast. It appears that Ron Hay has scooped us in several respects. Next were two short talks, one by Stefan Müller (MPI Martinsried) on sumoylation and desumoylation of nucleophosmin, and another one by Alexander Buchberger (same place) on the role of UBX and PUB domains in recognizing the Cdc48/p97 protein.
After the coffee break, William Tansey (CSH) talked about transcriptional control by the ubiquitin system. I head several of his talks and find the system extremely interesting: apparently, many transcription factors become ubiquitinated within their transactivation domain, and this process is intimately linked with transactivation. On one hand, ubiquitination makes the transcription factors short-lived, on the other hand it is absolutely required for their activity. The likely reason for the latter is the ubiquitin-dependent recruitment of coactivators. Another transcription-related process regulated by ubiquitination is the subunit composition of the Pol II complex. The final talk in the session was by James Chen (Dallas), who was talking about the role of ubiquitination in the NFkappaB pathway. There is some new data suggesting that the UBC13/UEV1 complex (a two-component E2) is essential only for NFkB activation by IL-1 signaling, while in the TNF pathway is can be partially replaced by UBC5.
Second Session: UPS and Regulatory Processes
First speaker in the second session was Alan D’Andrea (Dana-Farber), who talked about two recently published results on the Fanconi Anemia pathway. Most genes mutated in this DNA-repair disease encode components of an enigmatic ubiquitin ligase with only two known targets: the ‘classical’ FANCD2 and the novel FANCI proteins. The DNA damage response by the Fanconi pathway also requires a USP1, a deubiquitinating enzyme that can deactivate itself by auto-cleavage. Alan also talked about a novel USP1-associated factor UAF1. The next talk was by Jasper Svejstrup (CR-UK, London) on ubiquitination and degradation of RNA Polymerase II. Unfortunately, I don’t have notes on this talk. The following two short presentations had a strong focus on overthrowing current dogmata. First, Kazuhiro Iwai (Osaka) talked about a ligase that forms linear head-to-tail ubiquitin chains on the Nemo protein. I had some acoustical problems following this talk (poorly adjusted mic) but I found this idea mildly irritating. The only thing I remember is that the gels and blots of the linear chains looked spectacularly good – maybe these chains are better behaved than others on SDS gels. Next, Michael Glickman (Technion, Haifa) presented his rather unconventional idea on how the 19 S particle of the proteasome is organized. The data are unpublished, so I won’t go and spoil his story. I heard this talk three times within a few months, and must say that I like this idea (and the atomic force microscopy data that support it). The last talk of day2 was given by Stephanie Carter (Glasgow) on p53 ubiquitination by MDM2. Again, I don’t have notes for this talk – sorry.
The third day continued the ‘regulatory’ session by a talk from Stefan Jentsch (MPI Martinsried). He was talking about recently published work on the sumoylation of PCNA and Rad52 in response to DNA double-strand breaks. It is known from previous work that PCNA ubiquitinatin regulates the switch from processive DNA polymerases to translesion repair polymerases, and the PCNA sumoylation competes with ubiquitination. According to Stefan, PCNA sumoylation acts as a ‘reset button’ that facilitates new PCNA interactions. The sumoylation of Rad52 apparently plays a role in starting the ‘error free’ repair pathway rather than the error-prone translesion synthesis. In the next talk, Maria Masucci (Karolinska, Stockholm) discussed the modulation of the ubiquitin pathway in Epstein-Barr virus infection. The GA-repeats in the EBNA1 proteins interfere with the proteasome and prevent EBNA1 degradation. In addition, the EBV and some other viruses appear to encode deubiquitinating enzymes that can be expected to mess up the cellular ubiquitin system. The last talk in this session was by Keiji Tanaka (Tokyo), who summarized his recently published data on a thymus-specific form of the proteasome. If you want to know more, read this blog entry on Mystery Rays.
Third Session: UPS and Subcellular Compartments
The third and last session was started by Pier Paolo Di Fiore (Milan) who did not talk about endocytosis but rather on the tumor-suppressing role of the Numb protein. Numb is strongly down-regulated in breast cancers with a poor prognosis but normal p53 status. Apparently, Numb interacts with MDM2 and p53 and regulates MDM2 activity for ubiquitinating p53. Pier Paolo also showed some data supporting his notion that the Numb/p53 axis is important for the maintenance of the stem cell compartment. The next talk was by Scott Emr (Cornell), and he did talk about endocytosis. As usual, his talk was a very entertaining walk through the whole endocytosis/ESCRT business. Scott also presented some new data on a novel class of adapters that seem to mediate selective internalization of integral membrane proteins.
The first talk on day 4 was by Simona Polo (Milan). She discussed the choice between clathrin-dependent and independent pathways in the internalization of the EGF-receptor. If I understood the data correctly, a loss of the clathrin-mediated pathway affects EGFR signaling much more than EGFR levels (i.e. degradation). The next talk was by Mark Hochstrasser (Yale), who presented data on two new assembly chaperones of the 20S proteasome. While the proteins themselves (Pac3 and Pac4 in humans, Pba3 and Pba4 in yeast) have been published recently by others, Mark apparently has some unique functional data showing how the Pba3/4 complex sees that every 20S component finds its proper place in the proteasome. Dieter Wolf (Stuttgart) talked about a the Gid-complex, a novel multimeric ubiquitin ligase responsible for the degradation of fructose 1,6 bisphosphatase in yeast. I find this complex very interesting (admitted, I am a minor collaborator) because the human genome encodes a lot of orphan proteins that resemble Gid complex components. It is well possible that these proteins form the basis of a novel class of complex ubiquitin ligases.
After the last coffe break, the session was concluded by two talks on protein quality control by the ERAD (ER-associated degradation) pathway. In this process, misfolded proteins in the ER lumen are recognized and retro-translocated into the cytoplasm. Upon translocation, the proteins become ubiquitinated and degraded by the proteasome. First, Randy Hampton (UCSD) presented a very entertaining overview of the system, culminating in the crucial question on the nature of the elusive retro-translocation channel. The original candidate Sec61 (the channel for cytoplasm-to-ER secretion) has been discounted a while ago. Now it seems that the next two candidates (Der1 and Dfm1) don’t fit the bill either. Apparently, Randy now favors Hrd1 as a novel candidate pore. The final talk was by Thomas Sommer (MDC Berlin), who presented his data on the transmembrane ERAD ligase complex (Der1, Usa1, Hrd3, Hrd1, Ubx2) and the associated lumenal Yos9 protein that helps to recognize misfolded proteins within the ER.
All in all, a very interesting meeting. Most of the people I have talked to appeared quite satisfied, with only very few critical voices. I overheard some people complain on either too much or too little biology – this is probably hard to avoid. Other people complained that it doesn’t make sense anymore to have meetings on the ubiquitin system (like there are no meetings dedicated to kinase signaling anymore). According to them, ubiquitin topics should be dealt with on a session-basis and find their place in meetings on Endocytosis or DNA repair or Cancer. I am not convinced. For once, we already have UPS sessions as part of several biology-oriented meetings. Moreover, ubiquitination is far more complicated than phosphorylation. There are many components and paradigms shared by completely different pathways, the majority of which are still poorly understood. In my opinion, there is still plenty of overlap between the different areas where ubiquitin is important. Actually, it becomes increasingly hard to find an area in (eukaryotic) biology, where ubiquitin and its cousins are not important.
Another question that has been posed is: was this EMBO conference a representative overview on ubiquitin research in Europe? I am not sure. There clearly was a certain country bias. Judging by the speakers affiliations, most of European ubiquitin research seems to happen in the US, followed by Germany, the UK and Japan. Some other countries (e.g. France, Spain) were not represented by speakers (although there were a number of posters). The strong presence of Germans might be partially due to some organisatorial bias (three of the six organizers are from Germany), but a big part of the uneven distribution is probably caused by the presence of ‘national clusters’ in the ubiquitin field. It will be interesting to watch the field develop during the years to come.