- Email: [email protected]
Connecting protein family resourcesusing the proWeb network
COMPUTECORNER R database to ensure that the gene has not already been named; (2) review the terminology used previously for related genes among all organisms; (3) include in your discussions the leading research groups working on the gene or gene product; and (4) check that the proposed name of the gene family does not conflict with a previouslyassigned name. We would also request that we be contacted early in this process so that we can advise and ensure that your proposed gene name conforms to the CPGN guidelines. We would also be
TIBS 21 - NOVEMBER 1996
happy to help establish a working group consisting of others in the field, and, recover, align and assign tentative member numbers to the sequences from the DNA databases, and assist in having the agreed nomenclature entered into the sequence databases.
References 1 Hallick, R. N. and Bottomley, W. (1983) Plant MoL Biol. Rep. 1, 38-43 2 Hallick, R. B. (1989) Plant Mol, Biol. Rep. 7, 226-271 3 Lonsdale, D. M. and Leaver, C. J. (1988) Plant Mol. Biol. Rep. 6, 14-21
4 Pfanner, N. et af. (!996) Trends Biochem. ScL 21, 51-52 5 Stewart, A., ed. (1995) Trends in Genetics: Genetic Nomenclature Guide, Elsevier Science
DAVID LONSDALE Co-Chairman CPGN,John Innes Centre, Norwich, UK NR4 7UH. Email: [email protected]
CARL PRICE Co-Chairman CPGN,Wakesman Institute, Rutgers University,Piscataway, NJ 08855-0759, USA. Email: [email protected]
been identified, and have been implicated through mutant analysis and cellular localization in diverse microtubulebased processes, including spindle and chromosome motility, kinetochore function, and vesicle/organelle transport. More than 75 members of the kinesin family have been identified so far, grouping into at least eight subfamilies features of a protein family using a based on phylogenetic analysis of the documentation entry, proWeb uses a -340-residue motor domain6. X-ray crysdedicated World Wide Web (WWW) site tallography of both kinesin and ncd, a on a given protein family, and links this kinesin-related protein, confirm that the to home pages of people who work on motor domains have nearly identical individual family members. In this way, structures, even though these motors proWeb takes advantage of existing re- move in opposite directions along the sources maintained by the researchers microtubule. The kinesin WWW site lists themselves. We intend that a proWeb some 50 laboratories that currently study site be initially designed in collaboration the kinesins, with links to researchers' with an expert on the protein family,who home pages where individual interests along with others in the field, will pro- are described. Kinesin researchers invide input for maintaining and updating clude biochemists, biophysicists, cell the site once it is launched. We expect that biologists, geneticists, structural biolothe primary use of a proWeb site will be gists and theoretical biologists. to present information at a level of detail The minimum information for a protypical of written reviews. In addition, tein family Web site should include a links from searchable databases such list of known family members and a seas Blocks Oattp://www.blocks.lhcrc.org) quence alignment showing the region and Pfam (http://www.sanger.ac.uk/Pfam) of sequence similarity. The kinesin to a proWeb site and links from there WWW site provides links from the proto relevant individual sites provide a teins to sequence database entries, ilsmooth transition from a database hit lustrated subfamily descriptions, short to biological information. Moreover, a synopses of major topics written by proWeb site could be updated, which in kinesin researchers, and links to bibliosome cases could be automatic using graphic databases. The site comprises database links, and include animation multiple pages to make viewing the inand sound features. formation easier for those using computers with limited memory. The abilKinesins as a prototype ity of the kinesin motors to move on We have chosen the kinesin motor microtubules and their involvement in proteins as our prototype proWeb site dynamic cellular processes is docu(http://www.blocks.fhcrc.org/-kinesin). mented in a movie page depicting Many proteins containing a domain re- microtubule and cellular motility of the lated to the kinesin motor domain have kinesin proteins.
Connecting protein family resources using the proWeb network Not long ago, a significant hit to your amino acid sequence in a homology search was a cause for celebration. Nowadays, there are usually multiple hits to contend with, often to sequences from large-scale sequencing projects. Trying to make the most out of a list of database hits to infer function can be a daunting task requiring specialized biological knowledge. Here, protein family classification systems, such as Prosite 1, Blocks2, SBase3, ProDom4 and Prints 5 can be useful. In some cases, documentation is provided that summarizes salient features of the family in question, often with references to the literature. Unfortunately, a summary written by an expert on protein families cannot provide the sort of biological depth that one typically finds in a review written by an expert on a particular protein family. Furthermore, the question of subfamily classification is usually beyond the scope of a summary, even though such classifications can be essential for drawing functional inferences.
proWebstructure proWeb is our name for an alternative approach to the problem of protein family documentation. As with previous protein family classification systems, proWeb categorizes families based on significant sequence homology; however, rather than summarizing biologica!
444
9 1996, Elsevier Science Ltd
PII: S0968-0004(96)30039-X
REFLECTIONS
T[BS 21 - NOVEMBER 1996 References
Table I. Protein family World Wide Web sites Protein family
Uniform resource Iocator
AAA Chaperonins Cytokines and receptors Esterases G protein-coupledreceptors Glucoamylases Glucocorticoid receptors Glycotransferases Histocompatibility proteins Homeoboxproteins Kinesins Mutator transposons Olfactory receptors
http://yeamob.pci.chemie.uni-tuebingen.de/AAA/Description.html http://biocO9.uthscsa.edu/~seale/Chap/chap.html http://www.ocms.ox.ac.uk/~smb/cyt_web/ http://ensam.inra.fr/cholinesterase/ http://receptor.mgh.harvard.edu/GCRDBHOME.html http://www.public.iastate.edu/~pedro/glase/glase.html http://biochem l.basic-sci.georgetown.edu/GRR/GRR.HTML http://bellatrix.pcl.ox.ac.uk/people/iain/glycosyltransferase.html http://histo.cryst.bbk.ac.uk/ http://copan.bioz.unibas.ch/homeo.html http://www.blocks.fhcrc.org/~kinesin/ http://www-leland.stanford.edu/~jeisen/Mutator/Mutator.html http://paella.med.yale.edu/cgi-bin/receptor_top/DB_CGl.p/ query?FORM%3bHOME http://base.icgeb.trieste.it/p450/ http://www.sdsc.edu/projects/kinases/kinase_intro.html http://www-leland.stanford.edu/~jeisen/RecA/RecA.html http://www-leland.stanford.edu/~jeisen/SNF2/snf2.html http://xanadu.mgh.harvard.edu/receptor/trr front.html http://ellington.pharm.arizona.edu/~bear/top/topo.html
P450-containingsystems Protein kinases RecA proteins SNF2 Thyroid hormone receptors Topoisomerases
At this time, we are aware of about 20 WWW sites dedicated to specific protein families and that provide up4o-date documentation and useful links (Table I). Protein family databases, such as the Blocks database, now have links to these
sites to provide access for interpretation of database search results. Any biologist interested in setting up a site for other protein families can contact us (kinesin@ sparky.fhcrc.org) for further information on the proWeb network.
Lessons from the discovery of the ubiquitin system It is now well known and widely accepted that, in eukaryotic cells, many proteins are targeted for degradation by covalent ligation to ubiquitin. It is also recognized that the ubiquitin system carries out the selective degradation of many important regulatory proteins. Notable examples are cell-cycle regulators, such as mitotic cyclins ~'2, G1 cyclins 3,4, some inhibitors of cyclindependent kinases 5,6and proteins whose degradation is required for the onset of anaphase 7,8. Other important regulatory proteins whose levels are controlled by ubiquitin-mediated degradation include the p53 tumor suppressor 9, the transcriptional regulator NF-KB and its inhibitor IKBcx(Refs 10, 11), the m o s protooncogene ~2 and many transcription factors (reviewed in Refs 13, 14). It has become increasingly evident that the specific and programmed degradation of short-lived regulatory 9 1996, Elsevier Science Ltd
proteins is a recurrent theme in temporally controlled processes. This seemingly wasteful way of destroying protein regulators might be essential to ensure irreversibility. Thus, ubiquitinmediated protein degradation is of considerable current interest. Yet, not many are aware of how obscure this field was less than 20 years ago. Some imaginative models have been proposed to account for the selectivity of intracellular protein breakdown, such as one suggesting that all cellular proteins are rapidly engulfed by the lysosome, but only short-lived proteins are degraded in the lysosome, while long-lived proteins escape back to the cytoso115. The story of the discovery of the ubiquitin system might be instructive. t became interested in the problem of how proteins are degraded in cells when I was a postdoctoral fellow in the
1 Bairoch, A. (1992) Nucleic Acids Res. 20, 2013-2018 2 Henikoff, J. G. and Henikoff, S. (1996) Methods Enzymol. 266, 88-105 3 Pongor,S. et al. (1993) Nucleic Acids Res. 21, 3111-3115 4 Sonnhammer,E. L. L. and Kahn, D. (1994) Protein Sci. 3, 482-492 5 Attwood, T. K. and Beck, M. E. (1994) Protein Eng. 7,841-848 6 Moore, J. D. and Endow,S. A. (1996) Bioessays 18, 207-219
STEVEN HENIKOFF Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, 1124 Columbia Street, Seattle, WA 98104, USA. Email: [email protected] SHARYN A. ENDOW Department of Microbiology, Duke University Medical Center, Durham, NC 27710, USA. Email: [email protected] ELIZABETH A. GREENE Laboratoire de Biologie Mol6culaire, BP 27, Chemin de Borde Rouge, 31326 Castanet Tolosan, France. Email: [email protected]
laboratory of Gordon Tomkins at the University of California, San Francisco, between 1969 and 1971. At that time, Gordon's main pre-occupation was the mechanism by which steroid hormones cause the increased synthesis of specific enzymes. This problem was studied in cultured hepatoma cells, in which the enzyme tyrosine aminotransferase (TAT) is markedly induced by corticosteroids. It was a large laboratory, with many post-docs working on different aspects of TAT induction. So, I chose to study a different process that also regulates TAT levels: the degradation of the enzyme. I found that the degradation of TAT in hepatoma cells was completely inhibited by inhibitors of cellular ATP production 16. These results confirmed and extended previous observations of Simpson 17 on the energy-dependence of protein degradation in liver slices. Similar energy requirement for the degradation of other proteins was subsequently observed in a variety of organisms (reviewed in Ref. 18). Following my return to Israel, I continued to work on the mechanisms of intracetlular protein degradation, t was very much intrigued by its energydependence, because proteolysis per
PII: S0968-0004(96) 10054-2
445
TIBS 21 - NOVEMBER 1996
happy to help establish a working group consisting of others in the field, and, recover, align and assign tentative member numbers to the sequences from the DNA databases, and assist in having the agreed nomenclature entered into the sequence databases.
References 1 Hallick, R. N. and Bottomley, W. (1983) Plant MoL Biol. Rep. 1, 38-43 2 Hallick, R. B. (1989) Plant Mol, Biol. Rep. 7, 226-271 3 Lonsdale, D. M. and Leaver, C. J. (1988) Plant Mol. Biol. Rep. 6, 14-21
4 Pfanner, N. et af. (!996) Trends Biochem. ScL 21, 51-52 5 Stewart, A., ed. (1995) Trends in Genetics: Genetic Nomenclature Guide, Elsevier Science
DAVID LONSDALE Co-Chairman CPGN,John Innes Centre, Norwich, UK NR4 7UH. Email: [email protected]
CARL PRICE Co-Chairman CPGN,Wakesman Institute, Rutgers University,Piscataway, NJ 08855-0759, USA. Email: [email protected]
been identified, and have been implicated through mutant analysis and cellular localization in diverse microtubulebased processes, including spindle and chromosome motility, kinetochore function, and vesicle/organelle transport. More than 75 members of the kinesin family have been identified so far, grouping into at least eight subfamilies features of a protein family using a based on phylogenetic analysis of the documentation entry, proWeb uses a -340-residue motor domain6. X-ray crysdedicated World Wide Web (WWW) site tallography of both kinesin and ncd, a on a given protein family, and links this kinesin-related protein, confirm that the to home pages of people who work on motor domains have nearly identical individual family members. In this way, structures, even though these motors proWeb takes advantage of existing re- move in opposite directions along the sources maintained by the researchers microtubule. The kinesin WWW site lists themselves. We intend that a proWeb some 50 laboratories that currently study site be initially designed in collaboration the kinesins, with links to researchers' with an expert on the protein family,who home pages where individual interests along with others in the field, will pro- are described. Kinesin researchers invide input for maintaining and updating clude biochemists, biophysicists, cell the site once it is launched. We expect that biologists, geneticists, structural biolothe primary use of a proWeb site will be gists and theoretical biologists. to present information at a level of detail The minimum information for a protypical of written reviews. In addition, tein family Web site should include a links from searchable databases such list of known family members and a seas Blocks Oattp://www.blocks.lhcrc.org) quence alignment showing the region and Pfam (http://www.sanger.ac.uk/Pfam) of sequence similarity. The kinesin to a proWeb site and links from there WWW site provides links from the proto relevant individual sites provide a teins to sequence database entries, ilsmooth transition from a database hit lustrated subfamily descriptions, short to biological information. Moreover, a synopses of major topics written by proWeb site could be updated, which in kinesin researchers, and links to bibliosome cases could be automatic using graphic databases. The site comprises database links, and include animation multiple pages to make viewing the inand sound features. formation easier for those using computers with limited memory. The abilKinesins as a prototype ity of the kinesin motors to move on We have chosen the kinesin motor microtubules and their involvement in proteins as our prototype proWeb site dynamic cellular processes is docu(http://www.blocks.fhcrc.org/-kinesin). mented in a movie page depicting Many proteins containing a domain re- microtubule and cellular motility of the lated to the kinesin motor domain have kinesin proteins.
Connecting protein family resources using the proWeb network Not long ago, a significant hit to your amino acid sequence in a homology search was a cause for celebration. Nowadays, there are usually multiple hits to contend with, often to sequences from large-scale sequencing projects. Trying to make the most out of a list of database hits to infer function can be a daunting task requiring specialized biological knowledge. Here, protein family classification systems, such as Prosite 1, Blocks2, SBase3, ProDom4 and Prints 5 can be useful. In some cases, documentation is provided that summarizes salient features of the family in question, often with references to the literature. Unfortunately, a summary written by an expert on protein families cannot provide the sort of biological depth that one typically finds in a review written by an expert on a particular protein family. Furthermore, the question of subfamily classification is usually beyond the scope of a summary, even though such classifications can be essential for drawing functional inferences.
proWebstructure proWeb is our name for an alternative approach to the problem of protein family documentation. As with previous protein family classification systems, proWeb categorizes families based on significant sequence homology; however, rather than summarizing biologica!
444
9 1996, Elsevier Science Ltd
PII: S0968-0004(96)30039-X
REFLECTIONS
T[BS 21 - NOVEMBER 1996 References
Table I. Protein family World Wide Web sites Protein family
Uniform resource Iocator
AAA Chaperonins Cytokines and receptors Esterases G protein-coupledreceptors Glucoamylases Glucocorticoid receptors Glycotransferases Histocompatibility proteins Homeoboxproteins Kinesins Mutator transposons Olfactory receptors
http://yeamob.pci.chemie.uni-tuebingen.de/AAA/Description.html http://biocO9.uthscsa.edu/~seale/Chap/chap.html http://www.ocms.ox.ac.uk/~smb/cyt_web/ http://ensam.inra.fr/cholinesterase/ http://receptor.mgh.harvard.edu/GCRDBHOME.html http://www.public.iastate.edu/~pedro/glase/glase.html http://biochem l.basic-sci.georgetown.edu/GRR/GRR.HTML http://bellatrix.pcl.ox.ac.uk/people/iain/glycosyltransferase.html http://histo.cryst.bbk.ac.uk/ http://copan.bioz.unibas.ch/homeo.html http://www.blocks.fhcrc.org/~kinesin/ http://www-leland.stanford.edu/~jeisen/Mutator/Mutator.html http://paella.med.yale.edu/cgi-bin/receptor_top/DB_CGl.p/ query?FORM%3bHOME http://base.icgeb.trieste.it/p450/ http://www.sdsc.edu/projects/kinases/kinase_intro.html http://www-leland.stanford.edu/~jeisen/RecA/RecA.html http://www-leland.stanford.edu/~jeisen/SNF2/snf2.html http://xanadu.mgh.harvard.edu/receptor/trr front.html http://ellington.pharm.arizona.edu/~bear/top/topo.html
P450-containingsystems Protein kinases RecA proteins SNF2 Thyroid hormone receptors Topoisomerases
At this time, we are aware of about 20 WWW sites dedicated to specific protein families and that provide up4o-date documentation and useful links (Table I). Protein family databases, such as the Blocks database, now have links to these
sites to provide access for interpretation of database search results. Any biologist interested in setting up a site for other protein families can contact us (kinesin@ sparky.fhcrc.org) for further information on the proWeb network.
Lessons from the discovery of the ubiquitin system It is now well known and widely accepted that, in eukaryotic cells, many proteins are targeted for degradation by covalent ligation to ubiquitin. It is also recognized that the ubiquitin system carries out the selective degradation of many important regulatory proteins. Notable examples are cell-cycle regulators, such as mitotic cyclins ~'2, G1 cyclins 3,4, some inhibitors of cyclindependent kinases 5,6and proteins whose degradation is required for the onset of anaphase 7,8. Other important regulatory proteins whose levels are controlled by ubiquitin-mediated degradation include the p53 tumor suppressor 9, the transcriptional regulator NF-KB and its inhibitor IKBcx(Refs 10, 11), the m o s protooncogene ~2 and many transcription factors (reviewed in Refs 13, 14). It has become increasingly evident that the specific and programmed degradation of short-lived regulatory 9 1996, Elsevier Science Ltd
proteins is a recurrent theme in temporally controlled processes. This seemingly wasteful way of destroying protein regulators might be essential to ensure irreversibility. Thus, ubiquitinmediated protein degradation is of considerable current interest. Yet, not many are aware of how obscure this field was less than 20 years ago. Some imaginative models have been proposed to account for the selectivity of intracellular protein breakdown, such as one suggesting that all cellular proteins are rapidly engulfed by the lysosome, but only short-lived proteins are degraded in the lysosome, while long-lived proteins escape back to the cytoso115. The story of the discovery of the ubiquitin system might be instructive. t became interested in the problem of how proteins are degraded in cells when I was a postdoctoral fellow in the
1 Bairoch, A. (1992) Nucleic Acids Res. 20, 2013-2018 2 Henikoff, J. G. and Henikoff, S. (1996) Methods Enzymol. 266, 88-105 3 Pongor,S. et al. (1993) Nucleic Acids Res. 21, 3111-3115 4 Sonnhammer,E. L. L. and Kahn, D. (1994) Protein Sci. 3, 482-492 5 Attwood, T. K. and Beck, M. E. (1994) Protein Eng. 7,841-848 6 Moore, J. D. and Endow,S. A. (1996) Bioessays 18, 207-219
STEVEN HENIKOFF Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, 1124 Columbia Street, Seattle, WA 98104, USA. Email: [email protected] SHARYN A. ENDOW Department of Microbiology, Duke University Medical Center, Durham, NC 27710, USA. Email: [email protected] ELIZABETH A. GREENE Laboratoire de Biologie Mol6culaire, BP 27, Chemin de Borde Rouge, 31326 Castanet Tolosan, France. Email: [email protected]
laboratory of Gordon Tomkins at the University of California, San Francisco, between 1969 and 1971. At that time, Gordon's main pre-occupation was the mechanism by which steroid hormones cause the increased synthesis of specific enzymes. This problem was studied in cultured hepatoma cells, in which the enzyme tyrosine aminotransferase (TAT) is markedly induced by corticosteroids. It was a large laboratory, with many post-docs working on different aspects of TAT induction. So, I chose to study a different process that also regulates TAT levels: the degradation of the enzyme. I found that the degradation of TAT in hepatoma cells was completely inhibited by inhibitors of cellular ATP production 16. These results confirmed and extended previous observations of Simpson 17 on the energy-dependence of protein degradation in liver slices. Similar energy requirement for the degradation of other proteins was subsequently observed in a variety of organisms (reviewed in Ref. 18). Following my return to Israel, I continued to work on the mechanisms of intracetlular protein degradation, t was very much intrigued by its energydependence, because proteolysis per
PII: S0968-0004(96) 10054-2
445