A Fission Yeast-Based Platform for Phosphodiesterase Inhibitor HTSs and Analyses of Phosphodiesterase Activity - Phosphodiesterases as Drug Targets - page 139

Negative regulators of PKA were also identified using fbp1-ura4 reporter-based
screens. Mutations in the cgs1 gene, which encodes the PKA regulatory subunit, were
identified among suppressors of an adenylyl cyclase deletion (Stiefel et al. 2004 ),
while mutations in the cgs2 gene, which encodes the only S. pombe PDE, were
identified among suppressors of an activation-defective form of adenylyl cyclase
(Wang et al. 2005 ). In both selections, the cgs1 and cgs2 mutations restore 5FOA-
resistant (5FOA R ) growth by repressing fbp1 - ura4 transcription. Thus, this system
for studying fbp1 transcriptional regulation is able to identify mutations that either
reduce PKA activity to stimulate growth in medium lacking uracil or increase PKA
activity to stimulate growth in 5FOA medium. The ability to use positive growth
selections for mutations that either decrease or increase cAMP levels greatly
enhances the robustness of these genetic selections. Although this system was
initially used to identify mutations that alter PKA activity, we have recently deployed
it as a HTS to detect small molecule inhibitors of heterologously expressed PDEs
(Alaamery et al. 2010 ; Ivey et al. 2008 ). The screen capitalizes on the conversion of a
PDE-expressing strain from 5FOA S into 5FOA R in the presence of a PDE inhibitor.
4 First Generation Screens
The cell growth-based screen to detect PDE inhibitors was developed by exploiting
mutations that reduce glucose detection and cAMP synthesis. Glucose is sensed by
the Git3 GPCR, which activates the Gpa2 G a subunit of the Gpa2-(G a ) Git5-(G b )
Git11-(G g ) heterotrimeric G protein. Gpa2-mediated activation of adenylyl cyclase
produces a cAMP signal to activate PKA, which represses fbp1 transcription by
multiple mechanisms (Fig. 1 ). Deletion of any of the git3 , gpa2 , git5 ,or git11 genes
confers a 5FOA S phenotype due to increased expression of the fbp1-ura4 reporter
(Landry and Hoffman 2001 ; Landry et al. 2000 ; Nocero et al. 1994 ; Welton and
Hoffman 2000 ). Assays of fbp1-lacZ expression in these strains reveal that the
relative importance of these proteins in glucose repression is Gpa2, Git3, Git5, and
Git11. Loss of Cgs2 PDE activity restores 5FOA R growth and repression of
fbp1-lacZ expression in all of these mutants, providing the conceptual basis for
PDE inhibitor screens.
The key to developing a strain suitable for a PDE inhibitor HTS lies in the
interaction between the heterologous PDE and the endogenous cAMP signaling
pathway. The PDE must lower the intracellular cAMP level sufficiently to create a
5FOA S growth phenotype that would be converted to 5FOA R upon inhibition of the
PDE. For highly active PDEs, little or no additional engineering is necessary to
attain 5FOA sensitivity; for example, strains expressing either of two isoforms of
the highly active human PDE4D enzyme (Conti and Beavo 2007 ; Houslay et al.
2007 ) display a leaky 5FOA S in cells with an intact cAMP signaling pathway and
require the loss of only the Git11 G g to produce a tight 5FOA S phenotype. Strains
expressing PDE4A or PDE4B enzymes encoded by two other genes of the PDE4
family, or PDE7A or PDE7B high-affinity cAMP-specific enzymes (Bender and