Biology Reference
In-Depth Information
4 An Early Proposal for the Substrate Specificity:
“Glutamine Switch”
In the analysis of the first X-ray crystal structure of a PDE4B catalytic domain,
Xu et al. (
2000
) proposed that the orientation of the side chain of the invariant
glutamine in PDEs might play a strict role to select substrates cAMP and cGMP. In
2004, Zhang et al. advanced this concept and called it the “glutamine switch”. This
hypothesis was based on the crystal structures of PDE4 and PDE5 in complex with
their products 5
0
-AMP and 5
0
-GMP (Zhang et al.
2004
). The crystal structures
showed that the side chain of the invariant glutamine is fixed in opposite orienta-
tions in the nucleoside-binding pockets of PDE4 and PDE5 and forms two hydrogen
bonds with 5
0
-AMP and 5
0
-GMP, respectively (Fig.
3
). These workers assumed that
the binding of the products simulates binding of the guanine or adenine portion of
the substrates. A 180
o
rotation of the glutamine side chain in either enzyme would
allow for formation of only one hydrogen bond with the less preferred substrate. For
the PDE families with a dual-specificity of the substrates, according to the “gluta-
mine switch” hypothesis, the invariant glutamine side chain would be free to rotate
and therefore form two hydrogen bonds when either cAMP or cGMP binds (Fig.
3
).
Thus, the difference in the orientation of the glutamine side chain would make a
difference of one hydrogen bond and account for the substrate preference of the
enzymes. The “glutamine switch” mechanism was supported by the structures of
dual cAMP/cGMP-specific PDE1B (Zhang et al.
2004
) and PDE3B (Scapin et al.
2004
), in which the invariant glutamine was not bonded with other protein residues
and was therefore free to rotate.
5 Evidence Against the “Glutamine Switch” Mechanism
Several lines of evidence suggest that the glutamine switch mechanism, while being
conceptually inviting, may be “too good to be true”. The structure of PDE2A3,
which has similar affinity for cAMP and cGMP and hydrolyzes both substrates
equally well, shows that the side chain of the invariant glutamine (Gln859) is fixed
by a hydrogen bond with Tyr827 in the unliganded state of the enzyme (Iffland et al.
2005
). In addition, the recently published structure of PDE10A2 showed Gln726 is
locked in place by two hydrogen bonds; one bond involves Tyr693 and the other
occurs through a water that is bound to Tyr730 and Trp762 (Wang et al.
2007a
).
Thus, in order for the invariant glutamine in either PDE2 or PDE10 to switch
orientation, these pre-existing hydrogen bonds would have to be broken, a process
that would be energetically costly. On the contrary, the Q817A mutation of PDE5
reduced K
M
for cGMP by 60-fold, but did not significantly change K
M
for cAMP
(Zoraghi et al.
2006
). Since the Gln817 side chain should have been free to rotate in
the Q759A mutant, it would have been predicted to improve affinity for cAMP.
Furthermore, mutation of the amino acid (Gln759) that tethers the side chain of the
