Dr. Bansal, it seems like during initial injections as the conc. of metal is increasing the protein affinity of protein for metal ion progressively increases and the binding sites gets saturated eventually. Probably, the metal ion is inducing certain structural changes in the protein. You may try repeating the experiment, if you have not done already. Check buffer pH, protein and ligand must be in the same buffer, repeatby varying conc. of protein and ligand.

regards

Thanks Arun, I have repeated the experiment but still got the same thermogram. And my query is that big dip why there is sudden change in increase or decrease of heat observed.

Regards

Ruby

Try do add increasing concentrations of NaCl  while keeping the  protein and

metal concentrations constant.The kind of curve you get  is indicative of more than one

protein conformations which bind the metal with different affinities.

Please, could you show the corresponding thermogram?And could you also show the control experiment (injecting the metal solution into buffer)?

I have appended both the thermograms. First thermogram is related to protein titration followed by control experiment (Metal ion in to buffer solution). Please suggest 

I do not understand... In the first binding isotherm the normalized heats were between -19 and -22 kcal/mol. In the one you showed later, heats up to -200 kcal/mol are observed. Unless there are is a considerable amount of protons exchanged upon complex formation and the ionization enthalpy of the buffer is large, a ligand binding enthalpy of -200 kcal/mol is quite unexpected. In addition, the control experiment shows significant concentration-dependent heat effects.

Which are the experimental conditions? Which is the metal you are assaying? Some metals are more difficult to employ than others (solubility, redox states...).

Before suggesting different posible explanations, I suggest to try different experimental conditions (different buffer, different pH, different ionic strength...), and see what happens.

I agree with Adrian that we need more  details  about your protein and the metal in

order to help with an interpretation.However a believe you have a protein  structural

transition between structures  which bind  the ion completely  differently and accompanying large charge transfers to  and from the solvent   the 200 kcal/mol are

not unusual in hydration dehydaration ionic processes. and of cource the instrument

measures all the enthalpies  entering  not  protein ,protons,ions,solvent separately..

 add prefferably simple salts  to, see the effects of competition between binding

and collective screening effects?

Thanks Adrian and Dikeos Mario

I have used BSA protein and titrated it against Cu(II) which is well reported in literature the buffer i have used is tris (pH 9.18). I have purified the BSA first and after that titrated it against Cu(II).

@Dikeos, Although dehydration events in metal ions can be expected to be associated with a large enthalpy, in fact, the binding process (which includes metal and protein dehydration, protein conformational change, creation of protein-metal interactions, proton exchange as a result of changes in pKa's...) is usually accompanied by small-to-moderate overall binding enthalpies (usually less than +-20 kcal/mol).

@Ruby, Working with metal can be quite tricky. Due to their low solubility, high density charge, redox properties, interactions with buffers and other solution componets, etc., it is difficult to work with them. Therefore, many control/optimization experiments are required.

In the following article:

https://www.ncbi.nlm.nih.gov/pubmed/11196901

you can find a calorimetric titration for BSA interacting with Cu(II) in 100 mM Tris buffer at pH 9.1.

Thanks Adrian, But I already gone through this paper and with BSA itself I got nice sigmoidal curves. In this one only which I shared, I have purified commercialized BSA and then titrated it with Cu and got this big dip in fitted curve.

@Adrian.What about  salt bridged   aggregates?or destruction of  disulfide   bridges

I do not think  one can   base an interpretation on "usual values" thats why  I suggested  competition experiments and additional parallel spectroscopic monitoring

of the protein  moeity  something most people do  as can be seen  in related papers

section of the paper you have suggested.

@ Dikeos, I do not say that the main criterion sholud be comparing with usual values, but if you get a -200 kcal/mol binding enthalpy, then you should be cautious, check the experimental conditions, and test other conditions. In the case in question, the large enthalpy observed in the titration could be a real observación, but in the mentioned paper the same titration provides a much smaller enthalpy

@Adrian.Part of our problem is that Ruby does not clearly say what  kind of sample

he  has  and what ph what other ions except Cu, for example is Zn present in the

"commercialized sample" etc.anyway I have said what I  suspect on the basis of my

experience with biomolecules   water and ions let us leave it there.

@Dikeos I have clearly mentioned that protein is BSA (bovine serum albumin) Ph 9.18, titrated with Cu(II) and commercialized sample meant by commercial BSA procured from sigma and I have purified this BSA on column, and this purified sample is only giving this dip. What else i should mention, which is required kindly let me know.

@RubyObviously the difference between the thermograms in the study @Adrian

mentioned  and your purified form   is the  purification since  the other values (ph,

protein molarities )are similar.It is possible  that strongly  binding ions( Zn binds about  10 times stronger than  Cu)  remain bound in the dirty sample and stabilize  one of the possible conformations i.e the dirty sample ITC looks "normal"  sigmoidal

I t  is  possible that the purification changes things and  you get  several binders of different affinities which can be interpreted as at least  2 different structures being

involved accompanied by all the other exotic stuff ph changes ,hydration changes

etc.If you are really interested in the dip  send me  all conditions and a file with the

data  of the ITC and I will try to model the binding isotherm with up to 3 binders to

see  the partial enthalpy changes involved  and whether they correlate with known

values.

A reasonable interpretation for the titration is the existence of two populations of BSA that are binding Cu(II), and they are not interconverting between each other. In addition the percentage of each fraction is much smaller than 0.5.

But again, I do not see any correspondence between the first and second titrations. I do not understand how the firts one derives from the first.

@Dikeos, I think that analyzing right away the output of a titration with a battery of models without thinking critically on thta output is a bad practice. I have seen fittings of titrations with a sequential binding model for 4 binding sites that could be analyzed with a single ligand binding site. I do not think you need ther types of binders for the firts titration shown.

Anyway, I do not want to enter into a non-constructive discussion.

@ Adrian I was seeking explanation for only two graphs which i attached as file in subsequent comments. I hope you are not getting confused by my first one  from which i have started because for that I have not performed control experiment.

Thanks for your replies, As per your suggestion of two populations I do not agree because I have purified BSA just to get monomer only and then only I got this dip.

@Dikeos Thanks I will upload my files but I already tried two site and four site model fitting.

@Adrian I am ready to bet a beer that with a single site on a single protein conformation you never get  a dip like this.when I speak of several binders I don't mean sequential binding on sites on the same structure (this should give a staircase

not a dip)  but what you also say now several populations of structures with different

affinities.Or in other words structural transition induced by ionic binding plus accompanying enthalpy changes.

@Ruby so if you have already done that fitting did the program deliver partial

enthalpy changes and  binding constants or not?

@Dikeos, I did not say that Ruby's titration can be analyzed with a single site on a single conformation. In addition, with the "sequential binding" model you can fit a binding isotherm like that. In fact, the sequential binding model is nothing else that the general binding model for n identical or non-identical, independent or cooperative binding sites, expressed in terms of the step-wise association binding parameters (association constants and association enthalpies). The sequential binding sites does not always result in a "staircase"; it depends on the relative binding affinites of the binding sites and their binding enthalpies. Please, see the following publication:

https://www.ncbi.nlm.nih.gov/pubmed/19228114

where a "dip" similar to that of Ruby's titration can be found, and it corresponds to positive binding cooperativity.

The best way to proceed is to analyze the titration with a general model considering either overall binding parameters or step-wise binding parameters. From that, some conclusions about the type of binding sites (identical/nonidentical, independent/cooperative) can be extracted.

Ruby's titration could be analyzed considering there are two different and independent binding sites, or two different conformations, but it could also be analyzed considering there are two cooperative binding sites. But in order to be cooperative, they should be within the same molecule. The analysis using the general binding model will provide valuable information on that.

Still, I consider that a binding enthalpy of (at least) -200 kcal/mol is too large, meaning that not only binding is observed, but also additional (artifactual) processes are ocurring in the calorimetric cell.