I've already tried by using a comparison of untreated and treated sodium borohydride. Treatment with NaBH4 reduces shiff base to amine, a step that makes trypsine "jump" the catalytic lysine cut and get, once run in the MALDI-TOF, a full piece comprising the previous + the following (peptides) + PLP. Compare to the untreated tryptic digest spectrum, and you'll get the differential peptide which ends with the catalytic residue.
Very easy to say but actually there are a lot of issues because of the high number of lysine residues (42!), other impurities and eventually fragmentations at source (if so, it's too bad!).
Can anyone give me an advice or suggest another technique (which is not alanine scanning)?
Don't need to scan: you need to know which PLP enzyme fold your protein has. If the sequence alone doesn't tell you that, run a threading program (fold prediction), which will. That will probably tell you, from the structure-based alignment, which lysine it is, or at least narrow it down to a couple. Mutagenesis will confirm.
You could try to do a lysine modification in the presence of excess PLP to see which one is protected.
so you mean to take my enzyme, resuspend it in a free-PLP excess buffer, add something to modify lysines (acetic anhydride?), cut the protein, wash peptides, run them into maldi-tof and finally check for the only peptide which had remained of "standard" size?
Don't need to scan: you need to know which PLP enzyme fold your protein has. If the sequence alone doesn't tell you that, run a threading program (fold prediction), which will. That will probably tell you, from the structure-based alignment, which lysine it is, or at least narrow it down to a couple. Mutagenesis will confirm.
Something like that. Hopefully the catalytic lysine will be in the Schiff base in the presence of excess PLP and would not react like all the other lysines.
In the absence of PLP the catalytic lysine will be modified like all the others.
It is actually not so easy as it looks, because PLP is pretty reactive and does modify almost any accessible lysine. Thus, depending on the concentration, you may end with many modifications. In fact, before the upcoming of site-directed mutations, PLP was used as a chemical modifier to study ligand affinities, through the protection afforded by the ligand against the PLP-dependent chemical inactivation of an enzyme.
I may suggest starting with a bioinformatic approach, you may use HHpred (http://toolkit.tuebingen.mpg.de/hhpred) to search for distant homologues of your protein in the PDB. This may give a hint of the lysines close to the active site.
In addition, you may try to build a 3D-model of your protein in I-TASSER, SAM-T08, or other modeling servers and compare the structure to the templates.
Then, assuming you can clone your protein and express it in some accessible system (bacteria, yeast, insect cells, or other), you may try to use site-directed mutagenesis to identify essential lysines.
I don't know what enzyme you are working with. For the gram negative bacterial lysine decarboxylases such as E. coli, the PLP-reactive lysine is in a consensus sequence [see Levine M, Progulske-Fox A, Denslow ND, Farmerie WG, Smith DM, Swearingen
WT, et al. Identification of lysine decarboxylase as a mammalian cell growth
inhibitor in Eikenella corrodens: possible role in periodontal disease. Microb
Pathog 2001;30(April (4)):179–92]. . Consenusus sequences around the reactive lysine have also been identified in other PLP enzymes. If you can identify the consensus sequence you might be able to develop a method based on that of Dr Sotres.
Martin Levine
The fact that a protein has 42 lysines does not mean that they will all react with PLP. Some may be buried, others on the surface will be fully charged at neutral pH and will not react because it is the uncharged amine that reacts. So choosing a suitably low pH (7?) for reaction may help to limit the number of residues that react (see bovine glutamate dehydrogenase story from 1960s and 70s). It is likely that your Lys of interest may be in a more hydrophobic cleft and have a depressed pKa. If you find there are still several residues reacting you could resort to the old trick of differential labelling before and after protecting with substrate analogues. Or could you label with substrate and NaBH4 ?
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Enzymology