Eugene Gussakovsky gussak at ias.agri.gov.il
Tue Mar 23 09:54:52 EST 1999

Eugene Gussakovsky <gussak at ias.agri.gov.il> wrote:
[base-64 encoded Word 8 snipped; here is the contents. Please note
 forged From: and Reply-To: header. I did not write this. - CK]

Dear Mr. Bhupest,

On your questions about thermal unfolding measurements with tryptophan

First, as regular, thermal unfolding don't reach full unfolding. Be
careful with this. Trp probes a very local conformation. In this respect,
CD measurements may be better. But Tris buffer is not the best choice for
both CD and temperature scans: Tris has significant absorption in far UV
and its pH strongly depends on temperature. The both problems can be solved
with phosphate buffer (low UV absorption) or MOPS buffer (lowest pH shift
with temperature). 

KI can not be bound to Trp residue itself but to other charged groups in Trp
vicinity.  You can try to check whether you have a collision quenching or
static quenching analyzing the dependence as a Stern-Volmer plot. Don't foget
that KI strongly absorbs the excitation light producing inner filter effect.
May be quenching by neutral acryl amide would be better.

Unfortunately, there is a mistake in the Frank Fuerst reply concerning the
baseline in the temperature dependence of quantum yield (q): you are correct
that there is a trivial temperature quenching, not connected with the protein
unfolding. Try to plot 1/q versus absolute temperature/viscosity of protein
solution (buffer or water in fact), 1/q-vs-T/v. You can find at least two
regions with linearity: one for native state (prior to the unfolding started)
and second for unfolded state (after the unfolding was over). If you didn't
find such two regions, it may mean that you started heating the protein when
the unfolding started yet or finished heating when the unfolding is not
finished yet. So you'll have two linear equations y1=a1+b1*T/v (native) and
y1=a2+b2*T/v. Using these equations you can construct logK-vs-1/T plot in
temperature range between these two linear regions, where K=(1/q-y1)/(y2-1/q)
for each temperature in this range. If you have a two-state transition, this
plot will be linear and gives you thermodynamic parameters of unfolding:
slope=-enthalpy/R and intercept=entropy/R. If you'll find the plot is
non-linear, you have more complicate transition.

Best wishes.

More information about the Proteins mailing list