A Novel Lipid Inhibitor of Protein Phosphatase-1

1
A Novel Lipid Inhibitor of Protein
Phosphatase-1 Kathleen R. Perreault, Brian
Dembinski, Jason T. Maynes, Michael N. G.
James, Elena Posse de Chaves, and Charles F.
B. Holmes  From the Canadian Institutes of
Health Research, Group in Protein Structure and
Function and the Signal Transduction Research
Group, the Departments of Biochemistry and
Pharmacology, Faculty of Medicine, University of
Alberta, Edmonton, Alberta T6G 2H7, Canada
Introduction Reversible protein phosphorylation
is an integral mechanism of signal transduction
in many important cellular processes, including,
but certainly not limited to, mitogenesis,
apoptosis, and regulation of gene expression. In
the human genome, the ratio of Ser/Thr protein
kinases to Ser/Thr protein phosphatases is
approximately 81. The corollary of this
unbalanced ratio is that an individual
phosphatase is responsible for dephosphorylating
a broad range of substrates, and therefore must
be promiscuous with respect to substrate
specificity. To compensate for this relative
lack of specificity, Ser/Thr phosphatases are
regulated by a large number of inhibitory and
targeting subunits, which serve to direct their
activity towards the appropriate
substrate. Protein phosphatase-1 (PP1) is a
Ser/Thr phosphatase of the PPP family, which is
also comprised of PP2A, PP2B (Calcineurin), PP4,
PP5, PP6 and PP7. PP1 activity is regulated by
many endogenous protein inhibiting/targeting
subunits. A number of structurally diverse
natural product toxins are also potent inhibitors
of PP1 activity (Figure 1). Despite the
structural diversity of these toxins, they have
several coarsely similar features that aid in
binding to PP1 hydrogen bonding to specific
conserved residues in close proximity to the site
of enzymatic activity, acidic groups that
interact with conserved basic amino acids within
the active site, and hydrophobic regions that
allow binding at the hydrophobic groove adjacent
to the active site (1-3). Ceramide is a
sphingolipid second messenger produced in
response to cellular stress via activation of
sphingomyelinases. Agonists that cause cellular
production of ceramide include cytokines (TNF,
Fas), agents of environmental stress (heat, UV
irradiation), and chemotherapeutic agents. The
accumulation of ceramide activates JNK/SAPK,
PKC?, caspases as well as PP1 and PP2A (6).
Substrates of PP1 and PP2A that are
dephosphorylated in response to either
ceramide-inducing agonists or addition of
exogenous ceramide include c-jun, SR proteins,
retinoblastoma protein, PKB/Akt1, protein kinase
Ca and Bcl-2. Glucosylceramide (GlcCer) is a
metabolite of ceramide produced by the
glycosylation of the 1-hydroxyl group of ceramide
by the enzyme Glucosylceramide Synthase (GCS)
(Figure 1). Given the similarities in structure
between the natural product inhibitors of PP1,
the clavosines, and the sphingolipid GlcCer, we
hypothesized that GlcCer may affect PP1c activity
by binding to the catalytic subunit in a similar
fashion. Using radiolabelled glycogen
phosphorylase a, a physiological substrate of
PP1, we found that GlcCer inhibited PP1 activity
in vitro. Using site-directed mutagenesis of the
PP1 catalytic subunit (PP1c), we determined that
the ß12-ß13 loop (Figures 2 and 4), an
unstructured chain of five non-conserved amino
acid residues present in PP1, PP2A and PP2B, is
important for binding of GlcCer to PP1c.
Ceramide activation of PP1c is unaffected by
mutations in this region. We also found that
mutation of Tyr-134, an amino acid residue
present at the interface between the hydrophobic
groove and the active site of PP1c, greatly
decreases the potency of GlcCer inhibition.
Finally, we used lysates of live cells with
accumulated GlcCer to show that endogenous PP1
activity is also decreased in the presence of
GlcCer.
Conclusions We have identified glucosylceramide
as a novel inhibitor of PP1c and PP2Ac.
Mutagenesis studies of PP1c have shown that
residues in both the ß12-ß13 loop and the
hydrophobic groove are important for the
inhibition of PP1c by glucosylceramide. Studies
using lysates from live cells show that this
inhibition is not purely an in vitro phenomenon,
as endogenous PP1 activity is also affected by an
increase in glucosylceramide.
Results
Future Directions We hope to carry out studies
on the effect of endogenous and exogenous
glucosylceramide on the phosphorylation states of
PP1 and PP2A substrates. Because
glucosylceramide has been shown to accumulate in
multidrug-resistant cancer cell lines like the KB
cell lines used in our study, we are particularly
interested in the phosphorylation state of
proteins involved in cell cycle arrest and
apoptosis. Previous studies have shown that
treatment of sympathetic neurons with ceramide
(PP1 activator) blocks hyperphosphorylation of
pRB (retinoblastoma gene product), and therefore
we hypothesize that we may see hyperphosphorylatio
n of pRB upon treatment of neurons with GlcCer
(5).
Figure 3 GlcCer inhibits both PP1c and PP2Ac.
Inhibition of PP1c (IC505 µM) is approximately 3
times more potent than inhibition of PP2Ac
(IC5015 µM).
Figure 4 (left) Differences in sequence in the
ß12-ß13 loop region of PP1, PP2A and PP2B
(Calcineurin). The ß12-ß13 loop corresponds to
residues 273-277 in PP1c.

• References
• 1. Kathleen R. Perreault, Jason T. Maynes, Maia
  M. Cherney, Hue Anh Luu, Michael N. G. James, and
  Charles F. B. Holmes. Crystal Structure and
  Mutagenesis of a Protein Phosphatase-1Calcineurin
  Hybrid Elucidate the Role of the ß12-ß13 Loop in
  Inhibitor Binding. J. Biol. Chem. 279
  43198-43206 (October 2004).
• 2. Jason T. Maynes, Katherine S. Bateman, Maia
  M. Cherney, Amit K. Das, Hue Anh Luu, Charles F.
  B. Holmes, and Michael N. G. James. Crystal
  Structure of the Tumor-promoter Okadaic Acid
  Bound to Protein Phosphatase-1. J. Biol. Chem.
  276 44078-44082 (November 2001).
• 3. Charles F. B. Holmes, Jason T. Maynes,
  Kathleen R. Perreault, and Michael N. G. James.
  Molecular Enzymology Underlying Regulation of
  Protein Phosphatase-1 by Natural Toxins. Curr.
  Med. Chem. 9 1981-1989 (November 2002).
• 4. Yaakov Lavie, Hui-ting Cao, Stuart L.
  Bursten, Armando E. Giuliano, and Myles C. Cabot.
  Accumulation of glucosylceramides in
  multidrug-resistant cancer cells. J. Biol. Chem.
  27119530-6 (August 1996).
• Greg Plummer, Kathleen R. Perreault, Charles F.
  B. Holmes, and Elena I. Posse de Chaves.
  Activation of Serine/Threonine Protein
  Phosphatase-1 is Required for Ceramide-Induced
  Survival of Sympathetic Neurons. Biochem. J.
  385 685-693 (February 2005).
• Yusuf A. Hannun and Chiara Luberto. Ceramide in
  the Eukaryotic Stress Response. Trends Cell
  Biol. 10 73-80 (February 2000).

Figure 1 The PP1 inhibitor clavosine (left),
and the sphingolipid glucosylceramide (right).
Figure 5 The ß12-ß13 loop as well as residue
Tyr-134 of the PP1c hydrophobic groove, are
important in binding of GlcCer to PP1c.
Figure 6 Glucosylceramide has been shown to
accumulate in several multidrug-resistant (MDR)
cancer cell lines (4). We examined PP1 activity
and GlcCer content in multidrug-resistant human
epidermoid carcinoma cells. We found these cells
have increased glucosylceramide content and
decreased PP1 activity. ( refers to plt0.005).
Cell Type
KB-3-1
KB-V1
KB-V.1
KB-V.01
PP1 Activity (x107 ?U/?g protein)
4.60.3
4.00.2
3.40.1
Figure 2 Crystal structure of PP1c bound to
clavosine (Maynes, JT et al, unpublished results.
The residues of the ß12-ß13 loop as well as the
residue Y134 of the hydrophobic groove of the
enzyme are indicated in green.
3.60.1
GluCer content (TLC)
Figure 7 GlcCer in rat sympathetic neurons
causes a decrease in PP1 activity in cell lysates
from rat sympathetic neurons.
Figure 8 A close-up view of the proximity of
bound clavosine to the ß12-ß13 loop (left) and
Y134 residue (bottom middle) of PP1c (Maynes, JT
et al, unpublished data).