Can you make an educated guess as to which kinase may be responsible, thinking about the function of your transcription factor, when it is activated etc and look for a concensus sequence for a particular kinase. Each kinase has specific target. If you have no clues then have you considered mutating your protein to convert a serine to glutamate to mimic phosphorylation. Alternatively, if you just want it phosphorylated without knowing which kinase is responsible, you could try incubating your purified transcription factor on beads with a whole cell lysate and 32P in the chance that a kinase in your lysate will target your protein.

This really is not the appropriate question to be asking. Virtually all proteins have serene, threonine and tyrosine residues and each of these amino acids have the potential to be phosphorylated. However, most occurrences of Ser/Thr/Tyr are not phosphorylated because of their context within the protein. Simply phosphorylating a protein non-specifically in vitro will be relatively meaningless. You need to first determine whether it is phosphorylated in cells/tissues and then which residues are phosphorylated (most phosphoproteins are phosphorylated at multiple sites - some turn over, some are stable. Once this is achieved (and you can take often get clues from other members of the family), then you can determine which are the likely kinases responsible for the phosphorylation. This is also not trivial because specificities of kinases are overlapping in many cases. There are some online tools such as NetworKIN (http://networkin.info/search.php) that can assist in site/kinase predictions.

If you do want to do the oversimplified experiment, the most promiscuous kinase in vitro is Casein Kinase 1. However, there would be no guarantee that it targets sites with any relevance to biology as the reason it's promiscuous in vitro is that most of its physiological selectivity is lost (such as subcellular anchoring, binding partners, etc). I would not recommend this approach.

BTW, Amanda's suggested experiment is also worth a try but you will need to inhibit phosphatases in the extract, use high specific activity ATP and gel filter the extract to remove endogenous (unlabelled) ATP. This is similar to an approach developed in Dundee termed KESTREL (reviewed in: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1383659/).

Thank u both for your suggestions.

Dr. Amanda, thank you for the valuable suggestion, i will give it a try after inhibiting the phosphatases!

Dr.James, networkin.info have database for human, drosophila, yeast and C.elegans. I am sorry that i didn't reveal my protein's origin. It is a Mycobacterial protein and i have found using NetPhosK 1 server, the scores of threonine and serine residues which are higher than the base scores that qualifies them to get phosphorylated by phosphokinase C. Can i proceed with the prediction of NetPhosK 1 or what do you suggest?

It sounds like you have a lot of preliminary data to collect. You can try several in vitro kinase assays but remember that this is an artificial approach that will likely mislead regarding what is going on in vitro. Since the HLH protein is of mycobacterial origin, you should determine whether its phosphorylation (if phosphorylated) is via host kinases (which would be my bet). In this case, the prediction algorithms may still be helpful.