Opened 7 years ago

Closed 7 years ago

Last modified 7 years ago

#1293 closed defect (not a bug)

AddH fails to add hydrogen to specific peptide nitrogen

Reported by: Tristan Croll Owned by: Eric Pettersen
Priority: moderate Milestone:
Component: Structure Editing Version:
Keywords: Cc: Elaine Meng
Blocked By: Blocking:
Notify when closed: Platform: all
Project: ChimeraX

Description

The attached file is snipped out of a much larger structure (unpublished model from a colleague, so I can't share the whole thing). AddH fails to add a hydrogen to the N atom of Thr381.

Attachments (1)

test.pdb (37.1 KB ) - added by Tristan Croll 7 years ago.

Download all attachments as: .zip

Change History (8)

by Tristan Croll, 7 years ago

Attachment: test.pdb added

comment:1 by Eric Pettersen, 7 years ago

Must be an error in my port of the code from Chimera1, since Chimera1 protonates it correctly (and the IDATM type for that atom is the same).

comment:2 by Eric Pettersen, 7 years ago

Status: assignedaccepted

comment:3 by Eric Pettersen, 7 years ago

Cc: Elaine Meng added
Resolution: not a bug
Status: acceptedclosed

Man, I am bad at cross-checking things. Chimera 1 _also_ does not add this proton. Chimera1/X won't add protons pointing towards metal ions (in this case the MG ion) if the heavy atom to ion distance is 3.6 angstroms or less. The distance is 3.54 angstroms here.

--Eric

in reply to:  5 comment:4 by tic20@…, 7 years ago

Ah. Is there any scope to adjust this behaviour (e.g. as an optional switch)? I don’t know of any real case where a peptide nitrogen is actually stably deprotonated.

 
 
Tristan Croll
Research Fellow
Cambridge Institute for Medical Research
University of Cambridge CB2 0XY
 

 


in reply to:  6 ; comment:5 by Eric Pettersen, 7 years ago

Well, it would seem that deprotonated backbone amides do on occasion coordinate metals.  From the Metalloprotein page of Wikipedia:

Coordination chemistry principles[edit <https://en.wikipedia.org/w/index.php?title=Metalloprotein&action=edit&section=2>]
In metalloproteins, metal ions are usually coordinated by nitrogen <https://en.wikipedia.org/wiki/Nitrogen>, oxygen <https://en.wikipedia.org/wiki/Oxygen> or sulfur <https://en.wikipedia.org/wiki/Sulfur> centers belonging to amino acid <https://en.wikipedia.org/wiki/Amino_acid> residues of the protein. These donor groups are often provided by side-chains on the amino acid residues. Especially important are the imidazole <https://en.wikipedia.org/wiki/Imidazole> substituent in histidine <https://en.wikipedia.org/wiki/Histidine> residues, thiolate <https://en.wikipedia.org/wiki/Thiolate>substituents in cysteine <https://en.wikipedia.org/wiki/Cysteine> residues, and carboxylate <https://en.wikipedia.org/wiki/Carboxylate> groups provided by aspartate <https://en.wikipedia.org/wiki/Aspartate>. Given the diversity of the metalloproteome <https://en.wikipedia.org/wiki/Proteome>, virtually all amino acid residues have been shown to bind metal centers. The peptide backbone also provides donor groups; these include deprotonated amides <https://en.wikipedia.org/wiki/Amide> and the amide carbonyl <https://en.wikipedia.org/wiki/Carbonyl> oxygen centers. Lead(II) binding in natural and artificial proteins has been reviewed.[9] <https://en.wikipedia.org/wiki/Metalloprotein#cite_note-9>

Here’s one pub that features this kind of coordination: Aromatic Interactions in Unusual Backbone Nitrogen-Coordinated Zinc Peptide Complexes:  A Crystallographic and Spectroscopic Study - Inorganic Chemistry (ACS Publications) <https://pubs.acs.org/doi/abs/10.1021/ic0500053>

What I can do is offer control over the metal-coordination deprotonation distance.  Your situation is just below the cutoff of 3.6.  You could either tweak the protonation process for this structure by lowering the cutoff to 3.5, or if you know you are working with a structure that may be “wacky”, completely turn off such deprotonation by dropping the cutoff to zero. Would this be good enough?

—Eric

in reply to:  7 ; comment:6 by tic20@…, 7 years ago

Wow - you learn something new every day!

Yes, that would do it, I think.

 
 
Tristan Croll
Research Fellow
Cambridge Institute for Medical Research
University of Cambridge CB2 0XY
 

 


comment:7 by Eric Pettersen, 7 years ago

Okay, added a 'metalDist' keyword to control the heavy-atom-to-metal distance criteria (default 3.6). For Elaine's benefit, there are several other criteria for deprotonation to occur:

1) Heavy atom must be electronegative (more details upon request)
2) If heavy atom is planar, must be bonded to 2 other heavy atoms
3) The donor-proton-metal angle must be > 120°

Note: See TracTickets for help on using tickets.