6 August 1998 blast: no hits, best is 1ps1[AB] at +1.93 double-blast: no hits t78.t98_6: some good hits: 2fbjL -7.950 1btn -7.210 1mph -7.210 1avo[BFFHJLN] -5.730 1bfvL -4.920 1cfvL -4.920 2bfvL -4.920 1cjl -4.590 library models: 1chkA -4.960 1ps1A -4.830 1nbaA -4.530 1nqbA -4.230 summing both ways: FSSP rep 2fbjL -7.950 ? immunoglobulin? 1btn -7.51 1btn 1mph -7.210 1btn 1jrhI -6.54 1jrhI antibody 1avo[DFHJLN] -5.730 1avoB 1itbB -5.02 1itbB 1avoB -5 1avoB 1chkA -4.960 1chkA 1bfvL -4.920 1nqbA 1cfvL -4.920 1nqbA 2bfvL -4.920 1nqbA 1ps1A -4.830 1ps1A 1lktA -4.73 1lktA 1cjl -4.56 1ppn t78.remote_4 brings in lots of light chains of immunoglobulins. 10 August 1998 Christian T98 is TESB_ECOLI. PMID:1645722 finds that Histidine 58 is involved in catalysis. I'm not sure what numbering this is relative to, as t78 may not start as the Swissprot sequence does. 1tht is the only other thioesterase whose structure has been solved. It is not in our model library. 11 August 1998 Christian I built 1thtA. The alignments 1thtA-t78-{global,pos} look promising in that they are nearly full length. I haven't looked at them with sae yet, though. 14 August 1998 Christian I think the sum scores are misleading for this target because t78.t98_6-varh50-pdb.rdb seems too have gone too far out in sequence space and grabbed immunoglobulins. So, going by the scores in t78-t98.rdb, I find the 3rd-highest structure, 1nbaA, interesting because many of its FSSP neighbors' functions are very suggestive of a link between them and t78. T78 is an acyl-CoA thioesterase. Acyl-CoA is an adenosine nucleotide linked to a sulfur, which is linked to a hydrocarbon chain. T78 removes this chain through hydrolysis. T78 is fairly non-specific in that it can deal with a range of hydrocarbon chain lengths. Many of the 1nbaA FSSP family members do things with CoA and fatty things linked to it. The alignments are rather full-length, especially t78-1nbaA-global.pw.a2m.gz. t78 scoring 1nbaA isn't encouraging, but the other direction is: 1nbaA-t78-global.pw.dist:T0078 288 -2.05 -6.90 1nbaA-t78-post.pw.dist:T0078 288 -2.05 -6.90 1nbaA-t78-vit.pw.dist:T0078 288 -4.01 -5.20 17 August 1998 Christian 1nba and T78 both exist as a homotetramer. I have created the alignment 1nbaA/1nbaA-t78-global.cbarrett-hand1.a2m. The active site, His57 ("VHS"), does occur in a surface position at the end of a helix could be a location for catalysis if the other subunits of the homotetramer are involved. The active site for 1nba is around Cys177, which is not very close to where His57 from T78 is superimposed on the structure. I'm not sure, though, that this is really an issue, since the two enzymatic reactions have different substrates. For this structure, the preferred of the two alignments is 1nbaA/1nbaA-t78-global.cbarrett-hand2.a2m. For this I essentially unaligned the long loop at the end. I took another look at the top-scoring library structure 1chkA, which for some reason on the first pass I didn't give much attention. I was probably distracted by the functional similarity in the 1nba FSSP family. The t78 target model doesn't much like the structure, but the converse is not especially true: 1chkA/1chkA-t78-fssp-global.pw.dist:T0078 -0.08 -1.74 1chkA/1chkA-t78-global.pw.dist:T0078 -3.56 -8.16 1chkA/1chkA-t78-post.pw.dist:T0078 -3.56 -8.16 1chkA/1chkA-t78-vit.pw.dist:T0078 -7.98 -8.51 1chkA/t78-1chkA-global.pw.dist:1chkA 5.41 1000000.00 1chkA/t78-1chkA-post.pw.dist:1chkA 5.41 1000000.00 1chkA/t78-1chkA-vit.pw.dist:1chkA -0.50 -1.01 The global alignment puts the active-site residue of t78 right in the bottom of the active-site pocket of 1chkA. I pushed the end of t78 back onto the structure to get a better alignment. I also unaligned a few bits around 220 because there is probably some rearrangement that is sure not to match that local part of the structure. The alignment is 1chkA-t78-global.cbarrett1.a2m. Based on the more sensible placement of T78's active-site residue, I favor the 1chkA alignment over those of 1nbaA. I still need to check 1jdbK, 1dapA, and 2isdB, which aren't done building. 18 August 1998 Christian Based on scores, 2isdB does not look promising. I have looked at 1dapA, which is the top hit in t78-fssp-vit.rdb with a score of -7.04, and the conservation pattern is much more suggestive than anything I've looked at so far. What's more, FSSP for 1dapA lists structures that have a functional connectedness with a T78. I suspect my alignment may need some fiddling. 19 August 1998 Christian I currently favor 1dapA-t78.cbarrett2.a2m.gz. Wed Aug 19 16:07:07 PDT 1998 Kevin I made some small changes to get 1dapA-t78-hand2.a2m It isn't a great deal better than Christian's---either is probably OK. Christian: I made an adjustment to 1dapA-t78-hand2.a2m to create 1dapA-t78.cbarrett3.a2m.gz. This alignment closes a big insert and a big delete at the expense of two identities. But is does result in better proximity of prolines and glycines between the two alignments. From cbarrett@cse.ucsc.edu Wed Aug 19 17:48:11 1998 Mail-from: From cbarrett@cse.ucsc.edu Wed Aug 19 17:48:11 1998 Return-Path: cbarrett@cse.ucsc.edu Received: from moo.cse.ucsc.edu (moo.cse.ucsc.edu [128.114.134.35]) by services.cse.ucsc.edu (8.8.4/8.6.12) with SMTP id RAA19766 for ; Wed, 19 Aug 1998 17:48:11 -0700 (PDT) Received: from localhost (cbarrett@localhost) by moo.cse.ucsc.edu (8.6.10/8.6.12) with SMTP id RAA19294 for ; Wed, 19 Aug 1998 17:48:10 -0700 X-Authentication-Warning: moo.cse.ucsc.edu: cbarrett owned process doing -bs Date: Wed, 19 Aug 1998 17:48:10 -0700 (PDT) From: Christian Barrett X-Sender: cbarrett@moo To: karplus@cse.ucsc.edu Subject: t78 fold Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Before I submit 1dapA-t78.cbarrett3.a2m (which is an adjustment of the alignment we just did in your office) for t78, I'd like you to take a look at my competing alignment. It is 1chkA/1chkA-t78.cbarrett2.a2m. Its scores are a somewhat encouraging. The known active site residue for t78 (the H in VHS) is located in the bottom of the obvious binding pocket. 1chkA/1chkA-t78-global.pw.dist:T0078 -3.56 -8.16 1chkA/1chkA-t78-post.pw.dist:T0078 -3.56 -8.16 1chkA/1chkA-t78-vit.pw.dist:T0078 -7.98 -8.51 Christian From karplus@cse.ucsc.edu Thu Aug 20 11:02:38 1998 Return-Path: karplus@cse.ucsc.edu Date: Thu, 20 Aug 1998 11:02:37 -0700 From: Kevin Karplus To: cbarrett@cse.ucsc.edu Cc: karplus@cse.ucsc.edu Subject: t78 After looking at them both, I think I like the 1dapA-t78.cbarrett3.a2m prediction best. Let's go with that one. The secondary structure match on 1chkA seems poorer, and conserved residues don't cluster like they do in the 1dapA alignment.