Are all Histidines in Foldit the same?

Case number:845799-2005025
Opened by:jeff101
Opened on:Tuesday, April 10, 2018 - 23:33
Last modified:Friday, April 13, 2018 - 02:44

In Puzzle 1500 (,
"Aflatoxin Challenge: Round 6 with Insertions",
in a structure I called Je4 (made in a file I called 1500k9),
I mutated residues 42 43 52 & 74 to Histidine and then tried
to hydrogen bond their N-H groups to various O's on the
aflatoxin (residue 229). Oddly, the N-H groups on
His42 & His74 were atoms 10 & 17 while the N-H groups on
His43 & His52 were atoms 7 & 15. Is it normal in Foldit
puzzles to have different types of Histidines in the protein?
Do these reflect the N1-H or N3-H tautomers of Histidine as
described below? How many different types of Histidine does
Foldit use? From the following, it seems like there are
3 possible types (sometimes called HSD HSE & HSP):

If Foldit has more than one type of Histidine, does it let the
different types interconvert like it does with pairs of cysteines
that get close enough together to form disulfide bonds?


(Tue, 04/10/2018 - 23:33  |  3 comments)

jeff101's picture
User offline. Last seen 1 hour 42 min ago. Offline
Joined: 04/20/2012
Groups: Go Science

When we mutate Histidines, is there a way
to control which type of Histidine is used?
If we select amino acid 42, for example, and
keep mutating it to Histidine, will we sometimes
get one type and sometimes another? Will switching
the rotamer of a particular Histidine sidechain
let it convert from one type to another? Will
shaking a particular Histidine sidechain let
it convert from one type to another?

rmoretti's picture
User offline. Last seen 1 hour 52 min ago. Offline
Joined: 01/15/2010
Groups: None

Foldit does have different types of histidines, and a single structure with multiple histidines can have different versions in different positions at the same time. It doesn't account for all three types you mention, but it does have both of the neutrally-charged tautomers (HSD & HSE - the difference being which nitrogen the sidechain hydrogen is attached to, the one nearer to the backbone or the one further from the backbone). Foldit doesn't consider the two-hydrogen protonated form, as predicting when that's relevant can get tricky.

As you indicate, this difference is important for scoring and hydrogen bonding. Even when not displayed, Foldit keeps track of where that hydrogen is, and will only count hydrogen bonds going the "right" way. (The nitrogen with the hydrogen being a hbond donor, and the one without being a hbond acceptor.)

Foldit can freely interconvert between the two forms. Whenever you do a Shake or Mutate which involved histidine, Foldit will consider both tautomers, and pick whichever gives the best score in that context. The tautomers aren't considered, though, when Pulling or Wiggling. Those two tools work only on whatever tautomer currently exists in the structure. (Other tools may or may not switch between the two, depending on if they have a shake-like component.)

There isn't, though, currently a way to manually control which histidine tautomer you get with Shake/Mutate. Foldit will give you whichever one gives the best score. If you do want to change which tautomer, you would have to arrange the environment around the histidine to be favorable for that tautomer, and then shake that residue. If everything works out, Foldit should put in the desired form.

I haven't tested this, so I may be wrong, but I think the rotamer cycling tool (the left and right arrows) in the Selection Mode interface will also consider both tautomers. If you're interested in a particular tautomer, you may be able to use those tools to pick out the specific tautomer/rotamer you want. (Note, though, that as soon as you Shake/Mutate you'll go back to whichever tautomer scores best, so be sure to stabilize that particular tautomer prior to shaking.)

brow42's picture
User offline. Last seen 3 days 10 hours ago. Offline
Joined: 09/19/2011
Groups: None

Thanks a lot for pointing this out, Jeff101. Back in 2015 I updated my atom tables to move the ring polar hydrogen. I couldn't understand how I had made such a mistake. It didn't know about the tautomers. I guess they are approx. equiprobable at physiological pH.

I can confirm that both tautomers MIGHT be selected by rotamer.SetRotamer(). Also, the polar hydrogen does move around in selection interface. N1-H is indicated by a 1.0 A bond between atoms 7 and 15. N2-H is indicated by a 1.01 A bond between atoms 10 and 17.

I say MIGHT because the available rotomers is subject to a probability cutoff, so in certain local environments you might not have both tautomers. Also, to clarify what moretti said, wiggle sidechains does not seem to change the tautomer, but pulling the sidechain DOES.

The fact that histidine reconfigures during game play could be a problem. We've seen bad behavior while mutating banded residues. I always remove bands before mutating to avoid crashes. It looks like band points to atoms with an integer index, and the atom count isn't changing, so this isn't a crasher. But having the polar hydrogen bounce around makes bonding harder. It might be a source of frustration.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons