I don't think anyone has suggested this yet, but as long as your buffer
does not contain sulfur and your protein does contain a known number of
S atoms, then you can get an accurate molar ratio of any metal to sulfur
via PIXE (proton-induced x-ray emission). Elspeth Garman is one of the
few people in the world who can do this, but last I heard she was still
interested in having people send her samples. Very tiny samples (such
as protein crystals) are fine for this technique.
http://www.bioch.ox.ac.uk/aspsite/index.asp?pageid=578
-James Holton
MAD Scientist
Roger Rowlett wrote:
As others have implied, quantifying metals in metalloproteins is very
challenging. In my experience, the principal problems are (1)
adventitious metal contamination, (2) accurate measurement of protein
concentration, and (3) weak metal binding.
Zinc and iron are ubiquitous microcontaminants, and crop up in some
pretty unusual places. Zinc will get into everything, and plastic
bottles and microcentrifuge tubes seem to be a pretty rich source of
iron contamination. In addition, we have found that some plasticware
will also bind low concentrations of zinc and other metal ions from
solution, making standardization of instruments very frustrating.
Having said all this, ICP-OES or ICP-MS are absolutely the best
methods of quantifying metals in protein solutions. ICP methods are
linear over many orders of magnitude, and sample prep can be minimal.
We simply dilute protein solution in high-quality deionized water to
approximately 0.1-1.0 ppm in glass (acid washed and dried if
possible), NOT plastic, and aspirate into an ICP-OES. The protein must
be diluted enough to reduce viscosity issues in nebulization.
Standards can be prepared in deionized water. We dilute our protein
solutions with the same dI water we use to prepare the blank, and
there is always some detectable Zn background. (Fe is usually very low
in glass containers). The key to accurate measurements is Zn
concentrations >0.1 ppm. This has worked very well for us for
quantifying native and metal-substituted zinc-metalloproteins. If you
are really up against contamination, you can extract solutions with
Chelex resin, but it is possible that this treatment could also remove
loosely bound metals from proteins. If your putative zinc protein has
one or more sulfur ligands, this is less likely to be a problem.
Measuring protein concentrations accurately is also challenging. Every
protein quantification method is subject to idiosyncrasies. The least
idiosyncratic chemical methods are the microbiuret method and a
variety of UV methods (I like one based on A230 in 0.1% Tween-100 or
another detergent.) Measuring protein concentration will be the major
source of error in metalloprotein stoichiometry measurements. Protein
has to be really homogeneous for any method to be accurate.
Also keep in mind that His-tagged proteins (and some proteins with
vicinial His residues or other ligands on the surface) may
non-specifically bind metal ions, clouding metal stoichiometry.
Finally, if your metal is not in a tight binding site, it will be
difficult to prepare solutions with the saturated stoichiometry. In
these cases, it is possible that microcalorimetry might be useful.
Xuan Yang wrote:
Dear Mr. Fritz,
Yes, the protein is not an E.coli protein! Instead, it was cloned
from a virus. And since it was a nonstructural viral protein, I
thought it might be appropiate to treat it as eukaryotic proteins.
E.coli system was quite different from eukaryotic ones, hence I was
quite cautious about the ICP-ES result and trying to confirm it via
alternative method. Thanks very much for mentioning the examples
which suggested that Fe might be contaminants. Indeed, when I cut the
protein in two parts (still with MBP) and test them via ICP-ES again,
Fe became negligible in both and Zn stoichiometry increaed to 1:1 in
the C-terminal part. The result lead me to focus on Zn instead of Fe.
But I still want to confirm the idea.
Matallo biochemistry was exactly what I dreamed to do.
Sincerely,
Xuan Yang
2007/8/6 Guenter Fritz <guenter.fr...@uni-konstanz.de
<mailto:guenter.fr...@uni-konstanz.de>>
Hi Xuan,
I guess your protein is not an E.coli protein. There are several
examples that eukaryotic Zn-proteins expressed in E.coli contain
Fe instead of Zn. I am sceptic whether IMAC with different metal
ions will give the solution of the problem. If you really want to
get information on the metal ion binding properties you will have
to do some matallo biochemistry: preparing apo protein,
reconstitution with metal ions, UV-Vis spectroscopy, EPR would be
great, ...
Dear Sir or Madam,
The ICP-ES results indicated that 1 molar my protein
purified from E.coli Origami(DE3) contained about a half
molar Zinc and nearly a quarter molar Iron (whether II or III
was not available). The protein carried a MBP tag on the
N-terminal and the situation was similar with or without His
tag at the C terminal. I want to determine whether my protein
really bind Zinc or Iron. Does anyone have any experience
about such problems?
Specifically, now I want to compare the binding efficiency
on various IMAC, i.e. 50mM ZnSO4, FeSO4, Fe2(SO4)3,
NiSO4(control), or CuSO4(control). However, considering the
instability of Fe(II) in solution, the design still seemed
problematic.
Sincerely,
Xuan Yang
National Laboratory of Biomacromolecules and
Center for Infection and Immunity,
Institute of Biophysics,
Chinese Academy of Sciences,
Room 1617, 15 DaTun Road,Chaoyang District,
Beijing, China, 100101
Tel: 86-10-64884329
Academic email: ya...@moon.ibp.ac.cn
<mailto:ya...@moon.ibp.ac.cn> <mailto:ya...@moon.ibp.ac.cn
<mailto:ya...@moon.ibp.ac.cn>>
We will either find a way or make one.
--
------------------------------------------------------------------------
Roger S. Rowlett
Professor
Department of Chemistry
Colgate University
13 Oak Drive
Hamilton, NY 13346
tel: (315)-228-7245
ofc: (315)-228-7395
fax: (315)-228-7935
email: rrowl...@mail.colgate.edu
