Dear Rick,

I have calculated and explored the mMCP-6 model and its
electrostatics. These are my observations:

The simplest and most straighforward explanation would be in terms of
electrostatic properties, Johnson & Barton suggestions
notwithstanding. If electrostatics were involved, one would expect a
region in mMCP-6 that concentrates many Lys & Arg residues, and not
many His nor Asp/Glu residues. In addition, there should be no such
region in mMCP-7 for consistency with our mMCP-7 modeling and your
experiments. Such a region on mMCP-6 would remain positively charged
at pH > 7 and would thus force mMCP-6 to remain bound to proteoglycans
outside cells. So I went serching for such a region in mMCP-6 and I
found it! The result is again very clearcut. The following points
summarize my observations about various candidate regions (only point
4 refers to the predicted mMCP-6 -- proteoglycan binding site):

1) The top region (region 1) in mMCP-6 is quite negative; it has 2 His
residues and 3 Lys/Arg residues __and__ 5 Asp/Glu residues, so the net
charge is 0 at best.

2) The bottom region (region 2) in mMCP-6 does not have any of the His
residues we found important in mMCP-7 (H8, H68, H70). In addition, the
distribution of the positive and negative charges in and around this
region is such that the region does not have a strong positive
potential even below pH 6.5 (the GRASP figure looks much red-er than
Fig.3b in Matsumoto et al.).

3) There is a slight apparent concentration of positive charges
involving R19, K21, K51, R76, H27, H49, and H244, located behind the
bottom region, but there are also many negative charges in the
vicinity and the shape of the region is such that again there is no
pronounced positive potential, according to GRASP, around this region
below or above pH6.5.

4) However, there is another region in mMCP-6, which we have not seen
in any of the chymases or tryptases yet, that has a very strong
positive potential below and above pH 6.5. It consists mainly of K203,
K201, K49, R186, R187, and K223. It does not have any His residues so
it cannot be pH/His dependent. This region is located on the front
face of the molecule, below region 1 (the five charges run
horizontally from left to right, somewhat above the center of the
molecule on Fig 2A in Matsumoto et al). The region does not cover the
active site cleft, although it passes close to it on the right
side. This region is NOT nearly so positively charged in mMCP-7 even
below pH 6.5 (it has one His residue, H187). Therefore, this region
has all the properties expected from the behavior of mMCP-6 and
mMCP-7, and thus provides the simplest explanation for your new
experiments. The prediction from the model is that a mutation of any
of the more centrally located positively charged residues (201, 49,
187) into a Glu or Asp would remove the heparin binding capacity of
mMCP-6 at all pH's.

A possibly related point is that the active site cleft in mMCP-6 is
extremely negatively charged, ensuring that it is not blocked by
heparin and is available for productive binding of a substrate. The
cleft is much less negatively charged in mMCP-7 (possibly because
mMCP-7 is supposed to be active in the absence of heparin anyway).

I have a comment on the Introduction: The description of the
positively charged regions ('Although the tryptase designated mMCP-6
...') in chymases/tryptases is perhaps a little too unprecise; some of
the proteases have two regions, etc. Perhaps you could copy the
concise summary of the regions from the Discussion in the Matsumoto et
al paper (last paragraph of column 1 on p.19530).

Perhaps you want to include a figure showing the positive
electrostatic potential of mMCP-6 (equivalent to Fig 3a but in the
orientation of Fig 2A because this is more appropriate for the new
region in mMCP-6) or the model itself (equivalent to Fig 2A, the same
orientation). In either case, I do not believe that a section in
Methods is necessary; a reference in the figure legend to our earlier
papers would be sufficient. Please let me know if you want the
figure(s)?

I am still amazed by your ability to pick interesting questions that
have straightforward 'modeling' explanations! Did you think further
about the opportunities for a common grant proposal?  I would like to
re-iterate that I am very very interested in expanding our
collaboration provided the manpower is available.

In this regard, I have a possible 'dirty' solution for getting some
support (very very hypothetical at this stage), which does not exclude
a grant proposal. There is a Japanese company, Teijin Ltd., that seems
to be interested in collaborating with me on modeling chymases and/or
tryptases. I do not know any details yet, but they could potentially
fund at least part of the work and/or provide a postdoc for modeling
and/or experimental work.  Would you be interested in a setup like
this? Should I suggest something like that to them and see how they
react?

Regards, Andrej

