GPCRs, Drugs, Sequences and other things we dont understand

1
GPCRs, Drugs, Sequencesand other things we dont
understand

• G Vriend TIP GPCRforum

2
Some numbers (give or take the same factor 10
that GPCR people are used to give or take in Kd
values). We spend 25.000.000.000 Euro per
year, each year, on GPCR research. Even with the
overhead charged by my institute that would pay
for five hundred thousand PhD students. This has
gotten us 25.000 GPCR drugs, 90 of which are
redundant or useless. Assuming all involved
listened to Lipinski, the score is roughly one
PhD student for ever for every atom in a useful
drug. So we are doing something wrong! Industry
only earns money on one out of every ten drugs,
but the other nine also cost 500.000.000 Euro
each to develop. Most drugs fail in one stage of
the development or another. Often the reason for
drug failure tends to be that we dont understand
how GPCRs work, but also that we never understand
how the human body works. Drug design stays
trial and error or we need something radically
new. Lets try to make some suggestions.

3
Cysteines
We got the B2-adrenoceptor structure recently.
The big surprise was the extra-cellular helix in
loop IV-V. That one is really different (in
structure, not in function) from rhodopsin

4
Cysteines
This cysteine bridge is the second most conserved
aspect of all GPCRs (Class A,B,C,D,etc). The only
thing more conserved is that they have seven
TMs. That reminds me of Vriends first rule of
sequence/structure analysis If it is conserved,
it is important However, despite that they
nearly all have the bridge, this bridge is never
conserved in location in the sequence. So the
conservation has a functional role that is só
important that nature almost cannot do without
it, and has invented this bridge in many
different ways. Assembly of functional rhodopsin
requires a disulfide bond between cysteine
residues 110 and 187. Karnik, Khorana J Biol Chem
1990 Disulfide bond exchange in rhodopsin. Kono,
Oprian Biochemistry 1998.

5
Constitutively active

6
The pocket

7
The pocket

8
The pocket
So, there is plenty of space in the pocket to
make this inhibitor bigger. But even a highly
similar molecule will touch some of the seven
helices in a different way and thus do something
else to the signalling. I guess one can add or
remove just one atom to a drug and make an
agonist into an inverse agonist, or cause
increased turn-over, decrease cooperativity,
alter dimerisation, but probably not the
preferred G protein (?).

9
What next
The beta-2 adrenoceptor contradicted some of our
old hypotheses, and confirmed many others. But it
also allows us to see how we can learn from our
errors. When the rhodopsin structure came out I
told you in a Dutch GPCR meeting The modellers
are in deep CMA still will work. Rational
drug design still will not work on GPCRs. Many
GPCR structures will show surprises. There is a
common mechanism. Dimers are important. So,
what is that CMA?

10
What is CMA?
QWERTYASDFGRGH QWERTYASDTHRPM QWERTNMKDFGRKC QWERT
NMKDTHRVW Red conserved Green variable Blue
correlated
11
Thanks to
In random order TIPharma, LACDR, NBIC, EU,
Organon, Simon, Florence, Wilma, Bob, Friedrich,
TinyGRAP et al, CMBI, and, of course