In addition, the treatment of PDE5A1-trans-fected Cos-7 and PC12 cells with staurosporine, an apoptotic agent that activates endogenous caspase-3, also induced proteolysis and inactivati
Trang 1Interaction of caspase-3 with the cyclic GMP binding cyclic GMP
specific phosphodiesterase (PDE5a1)
Mhairi J Frame1, Rothwelle Tate1, David R Adams2, Keith M Morgan3, M D Houslay4,
Peter Vandenabeele5and Nigel J Pyne1
1
Department of Physiology and Pharmacology, Strathclyde Institute for Biomedical Sciences, University of Strathclyde, Glasgow, Scotland;2Department of Chemistry, Heriot-Watt University, Riccarton, Edinburgh, Scotland;3School of Textiles,
Heriot-Watt University, Scottish Borders Campus, Galashiels, Scotland;4Molecular Pharmacology Group,
Division of Biochemistry & Molecular Biology, Institute of Biological and Life Sciences, University of Glasgow, Scotland;
5
Department of Molecular Biology, Institute of Biotechnology, Flanders Interuniversity, University of Ghent, Belgium
Here, we show that recombinant bovine PDE5A1 is
pro-teolysed by recombinant caspase-3 in in vitro and transfected
Cos-7 cells In addition, the treatment of
PDE5A1-trans-fected Cos-7 and PC12 cells with staurosporine, an apoptotic
agent that activates endogenous caspase-3, also induced
proteolysis and inactivation of PDE5A1 These findings
suggest that there is specificity in the interaction between
caspase-3 and PDE5A1 that requires application of an
apoptotic stimulus The potential proteolysis of the
[778]DQGD[781] site in PDE5A1 by caspase-3 might affect
cGMP’s hydrolyzing activity as this is within the boundary
of the active site We therefore created a truncated D781
mutant corresponding exactly to the potential cleavage
product This mutant was expressed equally well compared
with the wild-type enzyme in transfected Cos-7 cells and was
inactive Inactivity of the truncated mutant was not due
to potential misfolding of the enzyme as it eluted from gel filtration chromatography in the same fraction as the wild-type enzyme Homology model comparison with the catalytic domain of PDE4B2 was used to probe a func-tional role for the region in PDE5A1 that might be cleaved
by caspase-3 From this, we can predict that a caspase-3-mediated cleavage of the [778]DQGD[781] motif would result in removal of the C-terminal tail containing Q807 and F810, which are potentially important amino acids required for substrate binding
Keywords: apoptosis; caspases; cyclic GMP; phospho-diesterase; proteases
Members of the phosphodiesterase (PDE) family catalyze
the hydrolysis of cyclic nucleotides to inactive 5¢ nucleotides
Therefore, they terminate the action of agents, such as
b-adrenergic agonists and nitric oxide, which use cAMP and
cGMP as second-messengers, respectively, to initiate
cellular responses
There are at least 11 members of the PDE family
(PDE1-11) that are encoded by different genes These isoforms have
different specificities for cAMP and cGMP, are regulated by
several different protein kinases, e.g protein kinase A,
protein kinase B (Akt pro-oncogene), extracellular
signal-regulated kinase (ERK) and CAM kinase, and allosteric
molecules (e.g cyclic nucleotides, Ca2+) and display distinct
tissue distribution [1–3] PDE5A1 is a major cGMP-binding
protein expressed in lung [4] where it is believed to have a key
role in regulating nitric oxide signaling There are at least two
isoforms (termed PDE5A1 and 2) [4] The enzyme has a high-affinity for cGMP at both noncatalytic (GAF domains) and catalytic sites, is a dimeric protein with a subunit molecular mass of 93–98 kDa [5] The enzyme is phosphorylated at S92 and activated by both protein kinase A and protein kinase G [6–7] Here we explore the possibility that PDE5A1 may be regulated by caspase-3 as sequence inspection shows that the bovine enzyme contains five putative caspase consensus sites: DHWD(26–29), DEGD(134–137), DEKD(289–292), DCSD(365–368) and DQGD(778–781) (Fig 1) Of these sites, only two show strong consensus for caspase-3: DHWD(26–29) and DQGD(778–781) Indeed, we have shown that in the presence of the inhibitory protein (PDEc)
of the rod photoreceptor PDE6, PDE5A1 is a substrate for a low activity preparation of purified caspase-3 [8] Site-directed mutagenesis studies have defined the position of the GAF domains [9–11] and key amino acid residues involved in the metal ion coordination and catalytic activity [12–15] These are shown in Fig 1 to define their relative position of the putative caspase sites
The caspase family is composed of 13 distinct gene products, each with different substrate preferences and inhibitor sensitivities [16] The first caspase was identified as ICE (caspase-1), which converts pro-interleukin-1b into bioactive interleukin-1b [17] Subsequently, several human and murine caspases have been cloned These enzymes show sequence homology with CED-3 from the nematode, Caenorrhabdidtis elegans [18] The overexpression of
Correspondence to N J Pyne, Department of Physiology and
Pharmacology, Strathclyde Institute for Biomedical Sciences,
University of Strathclyde, 27 Taylor Street, Glasgow,
G4 ONR, Scotland, UK.
Fax: + 141 5522562, Tel.: + 141 5524400 ext 2659,
E-mail: n.j.pyne@strath.ac.uk
Abbreviations: DMEM, Dulbecco’s modified Eagle’s medium; PDE,
phosphodiesterase; PARP, poly (ADP-ribose) polymerase.
(Received 5 November 2002, revised 8 January 2003,
accepted 16 January 2003)
Trang 2different caspases in cells induces apoptosis and/or
inflam-matory mediator production [19,20] Caspases are
synthes-ized in the cell as inactive proenzymes These are activated
by proteolysis at internal sites and are subdivided into
initiators and effector enzymes Initiator caspases (e.g
caspase-8 and -9) are activated by proximity induced
proteolysis by adaptor-dependent recruitment in the
recep-tosome or apoprecep-tosome complex Once activated, these will
further propagate the cascade by activating the downstream
effector caspases The effector (executioner) enzymes
include caspase-3, and proteolyse a number of substrates
resulting in structural changes, such as gelsolin, nuclear
changes such as ICAD and signal transduction such as
MEK kinase [21], Mst-1 [22], PAK-2 [23], PI3K/Akt [24],
PKCf [25] and FAK [26] Caspase-3 and -7 cleave proteins
at a4DX3X2D1consensus site, where apolar amino acids at
position 2 are preferred The cyclic nucleotides, cGMP and
cAMP have been shown to promote apoptosis of certain
mammalian cells For instance, nitric oxide stimulates
apoptosis in cardiomyocytes and endothelial cells via a
cGMP-dependent pathway [27–29] cGMP is also required
for nerve cell death caused by glutathione depletion, via
modulation of calcium channel activity [30] In addition,
Huston and colleagues have shown that PDE4A5, which
specifically hydrolyses cAMP, is proteolysed by caspase-3
This removes the N-terminal tail that contains specific
binding sites for the lyn kinase [31] These findings provide a
rationale for investigating whether caspase-mediated
path-ways can interact with PDE5 in intact cells
In this article, we show that caspase-3 either directly or
indirectly via caspase-3 activated proteases results in
cleavage and inactivation of PDE5A1 Homology model
comparison with the catalytic domain of PDE4B2 was
used to probe a functional role for the region in
PDE5A1 that might be cleaved by caspase-3 Residues
in PDE5 identified by Turko and colleagues [13,14]
H603, H607, H643, D644, E762, H675, T713, D754,
Q765 and Q779 were used for the modeling Mutations
of T713 and H675 that are cognate residues of those that
orientate the magnesium ion via H-bonds to water
ligands in PDE4B produce comparatively little impact on
catalytic activity in PDE5 From the modeling, it is
possible that inactivation of PDE5A1 by caspase-3 might
occur via removal of key regions which constitute part of
the wall of the catalytic site of PDE5A1 We also suggest
a possible interaction between PDE5A1 and an
uniden-tified caspase-3-initiated protease(s) that may constitute a
novel signaling event
Experimental procedures
Materials
All biochemicals were from Boehringer Mannheim
(Mann-heim, Germany), while general chemicals and snake venom
were from Sigma Chemical Co (Poole, UK) [3H]cGMP and [35S]methionine were from Amersham International (Amersham, Buckinghamshire, UK) Cell culture supplies were from Life Technologies (Paisley, UK) Ac-DEVD-CHO and anti-PDE5 IgG was from Calbiochem (UK) The pCAGGS-Casp-3 plasmid construct was kindly provided by
T Miyazaki, The Burnham Institute, La Jolla, CA, USA Protein purification
Recombinant murine caspases were purified according to [19] The proteolytic activity of these enzymes on procaspase substrates has been described previously [32]
Sub-cloning Bovine PDE5A1 cDNA (GenBank accession number L16545) in pBacPac9 (Clontech, CA, USA) was a gift from
J Corbin (Vanderbilt University, USA) It was subcloned into pcDNA3.1/Zeo(–) (Invitrogen, the Netherlands) by amplifying the ORF using primers ApaI-Koz-PDE5A1-FOR (AAG GGC C C G C C A C C A TGG AGA GGG CCG GCC CCG GCT) and XbaI-PDE5A1REV (GCT
TC T AGA C TC AGT TC C GC T TGG TC T GGC TGC TTT CAC), digesting the product and the vector with ApaI and XbaI (Promega, UK), ligating, and then transforming into TOP10 Escherichia coli (Invitrogen) Positive clones were selected and sequenced by BigDye terminator cycle sequencing (PE Biosystems, UK) using a PE373A auto-mated DNA sequencer
Site-directed mutagenesis of D781 (DfiA) in the caspase-3 consensus site was carried out using Stratagene’s QuikChange Mutagenesis kit (Stratagene, UK) This was achieved using pcDNA3.1-PDE5 Zeo(–) plasmid with
125 ng of the forward primer, PDE5MUTF (GACCAA GGA GCT AGA GAG AGG AAA GAA CTC) and the reverse primer, PDE5MUTR (GAG TTC TTT CCT CTC TCT AGC TCC TTG GTC) in a 50-lL PCR containing
50 ng of pcDNA3.1-PDE5 Zeo(–) plasmid, 10 mM KCl,
10 mM (NH4)2SO4, 20 mM Tris/HCl (pH 8.8), 2 mM MgSO4, 0.1% Triton X-100, 0.5 lg bovine serum albumin, 0.4 mMdNTPs and 2.5 U PfuTurbo DNA polymerase The cycling conditions were 95Cfor 30 s then 12 cycles of
95Cfor 30 s, 55 Cfor 1 min and 68 Cfor 15 min Ten units of DpnI restriction enzyme was added to the reaction following PCR Two microliters of reaction mixture was used in a transformation reaction with TOP10 competent
E coli Positive clones were selected and sequenced to confirm the mutagenesis
Truncation of PDE5A1 at D781 involved the use of primers that carry a single point mutation to introduce a premature stop codon at 782R The forward primer has the sequence GACCAA GGA GAT TGA GAG AGG AAA GAA CTC, while the reverse primer was GAG TTC TTT CCT CTC TCA ATC TCC TTG GTC Twelve cycles of
Fig 1 Schematic showing the positions of
caspase motifs in PDE5A1.
Trang 395Cfor 30 s, 55 Cfor 1 min and 68 Cfor 16 min were
used for the PCR
Transfection
Cos-7 cells or PC12 cells were grown to 50–70% confluence
in Dulbecco’s modified Eagle’s medium (DMEM)
contain-ing 10% (v/v) fetal bovine serum Five micrograms of
pcDNA-3.1-PDE5 or 0.1–1 lg of pC AGGS-C asp-3 was
added to DMEM and DEAE-dextran (10 mgÆmL)1), mixed
thoroughly and incubated at room temperature for 15 min
Cells were incubated at 37Cfor 1 h with this medium,
before this was removed and 1 mL of 10% (v/v)
dimethyl-sulfoxide added for 30 s The medium was then aspirated
and the cells washed twice with DMEM containing 10%
(v/v) fetal bovine serum The cells were then placed in DMEM
containing 10% (v/v) fetal bovine serum, grown to
conflu-ence and harvested 48 h after transfection Alternatively,
cells were transfected with the plasmid construct following
complex formation with LipofectAMINETM2000,
accord-ing to the manufacturer’s instructions The cDNA
contain-ing media was then removed followcontain-ing incubation for 24 h
at 37C, and the cells incubated for a further 24 h
Cell lysates
Cos-7 and PC12 cell lysates were prepared by adding 0.25M
sucrose, 1 mM EDTA, 10 mM Tris/HCl, pH 7.4, 2 mM
benzamidine and 0.1 mM phenylmethanesulfonyl fluoride
Cells were scraped into this buffer and homogenized by
passing through a 0.24-mm gauge syringe needle The lysates
were either used for caspase activity assays or combined with
boiling electrophoresis sample buffer for SDS/PAGE
Immunoblotting
Nitrocellulose membranes were blocked for 1 h at 4Cin
10 mMphosphate-buffered saline (NaCl/Pi) and 0.1% (v/v)
Tween-20 containing 5% (w/v) non fat dried milk and
0.001% (w/v) thimerisol The nitrocellulose sheets were then
incubated overnight at 4Cwith antibodies in blocking
solution The sheets were then washed with NaCl/Pi and
0.1%(v/v) Tween-20 prior to incubation with horseradish
peroxidase-linked anti-rabbit IgGs in blocking solution for
2 h at room temperature After washing the blots as above,
the immunoreactive bands were detected using an enhanced
chemiluminescence kit
PDE assay
Unless otherwise stated, the assay of PDE activity was by the
two-step radiotracer method [33] using 0.5 lM[3H]cGMP
PDE activity measurements were performed under
condi-tions of linear rate product formation and where less than
10% of the substrate was utilized during the assay
[35S]Methionine-labeled PDE5A1 and poly (ADP-ribose)
polymerase (PARP)
One microgram of pcDNA-3.1-PDE5 or pGEM-PARP
was combined with an in vitro transcription/translation kit
reaction (Promega, UK) to produce [35S]methionine-labeled proteins
Purified caspases [35S]Methionine-labeled PDE5A1 was combined with an incubation mix (25 lL) containing 50 mMHepes, pH 7.4,
1 mMEDTA and 10 mMdithiothreitol with 60 ng per assay purified recombinant murine caspases Incubations were for
2 h at 30Cand were terminated by addition of boiling sample buffer for SDS/PAGE
Caspase-3 activity assays Caspase-3 activity in cell lysates was measured using [35S]methionine-labeled PARP as a substrate In each assay, equal amounts of cell lysate protein ( 5 lg/assay) were used and incubated with PARP for 2 h at 30C Incubations were terminated by addition of boiling electro-phoresis sample buffer for SDS/PAGE Inhibition of caspase-3 activity was achieved using the reversible inhibitor acetyl-Asp(OMe)-Glu(OMe)-Val-Asp(OMe)-aldehyde (Ac-DEVD-CHO)
Molecular modeling Modeling studies were performed on an SGI Octane workstation using the automated homology-modeling program,MODELER, withinINSIGHTII(Accelerys Inc., San Diego, CA, USA) This program generates an all-atom model based on a specified sequence alignment and reference protein structure The crystal structure of the PDE4B2B core catalytic domain published by Xu et al [34] was used as a reference model (Protein Data Bank accession code 1FOJ, chain B) A truncated PDE5A1 sequence, corresponding to the region spanning helices 5–16 of the 1FOJ structure, was matched to the PDE4B2 sequence This region embodies the metal ion and substrate-binding pocket together with flanking helices and exhibits close homology to the PDE5A1 sequence (27% identity) Thirty models were generated using the program’s highest optimization level of molecular dynamics simulated annealing, and the model with the lowest overall probability density function violation was taken forward for further energy minimi-zation Key residues in the PDE5A1 model that contribute to the metal binding environment (H603, H607, H643, D644, E672, H675, T713, D754) overlaid the corresponding residues in the PDE4B2 reference structure (H234, H238, H274, D275, E304, H307, T345 and D392) with little deviation The coordinates of these residues and the a-carbon centers of Q765, Q779 and Q807 were frozen during subsequent minimization, which was carried out using the cvff forcefield imple-mented in the DISCOVER module of INSIGHTII The dielectric constant was set to 4.00 and the model refined through 3000 steps of steepest descent energy minimiza-tion followed by conjugate gradient energy minimizaminimiza-tion
to convergence with a 0.001 kcalÆmol)1ÆA˚)1 root mean square energy gradient difference between successive minimization steps
Trang 4Results and discussion
Proteolysis of PDE5A1 by caspase-3
Recombinant PDE5A1 was produced from a
pcDNA-1-PDE5 plasmid construct (which has a T7 polymerase
initiation site) using an in vitro reticulocyte transcription/
translation kit A major 98 kDa [35S]methionine-labeled
protein corresponding to PDE5A1 was produced from the
plasmid construct and resolved on SDS/PAGE (Fig 2)
Several minor lower molecular mass [35S]methionine-labeled
proteins were also produced in the transcription and
translation reaction These are probably derived by
differ-ential internal translation initiation or proteolysis of
PDE5A1 by reticulocyte proteases Figure 2 shows that
caspase-3 cleaved PDE5A1 to a major 82 kDa fragment
This in vitro reaction showed specificity for caspase-3
because caspase-2, -12 and -14 did not significantly cleave
the enzyme Consensus sites for caspase-2, -12 and -14 are
not present in PDE5A1 Assays were deliberately designed
such that the final concentration of caspase-3 in the
incubation was 40 nM, which is equivalent with its
concen-tration in mammalian cells [32] These conditions were used
to best predict the extent and nature of the proteolysis that
might occur in intact cells Higher concentrations of
caspase-3 or extended incubation times cause extensive
proteolysis of the 82 kDa fragment into smaller
polypep-tides and is therefore, less stringent
Caspase-3 cleaved 50% of the PDE5A1 under the
assay conditions used Proteolysis is dependent upon both
the specific activity of the caspase-3, which might be
limiting, and the affinity of interaction, which in vitro may
reflect reduced efficiency compared with in vivo The
findings show that PDE5A1 is a substrate for caspase-3
in vitro, consistent with the presence of consensus caspase-3
sites in PDE5A1 They also support our previous results
showing that PDE5A1 is proteolysed by low activity
purified caspase-3 in the presence of PDEc [8]
PDE5A1 cleavage by caspase-3 and/or caspase-3 activated proteases in Cos-7 cell and PC12 cells Cos-7 cells were transiently transfected with PDE5 and/or caspase-3 plasmid constructs The main objective here was
to establish whether the overexpression of recombinant caspase-3 induces the proteolysis of PDE5A1 in an intact cell system
cGMP hydrolysing activity was increased 10- to 20-fold
in PDE5A1-transfected vs mock-transfected cells (n > 20), and was inhibited by > 90% by addition of the selective PDE5 inhibitor, zaprinast (10 lM) to the assay A major
98 kDa protein was detected on Western blots probed with specific anti-PDE5 IgGs in lysates from
PDE5A1-transfect-ed but not mock-transfectPDE5A1-transfect-ed cells and which comigratPDE5A1-transfect-ed with recombinant PDE5A1 (see later) These results are
Fig 2 The effect of recombinant caspases on PDE5A1
Autoradio-graph showing the effect of purified caspase-2, 3, 12 and 14
(60 ngÆassay)1) on [ 35 S]methionine PDE5A1 Control represents no
addition to PDE5A1 Radioactive-labeled molecular mass standards
are shown (M r ¼ 200–33 kDa) This is a representative result of three
separate experiments.
Fig 3 Caspase-3 in transfected Cos-7 cells Cells were transfected with pCAGGS-Casp-3 cDNA (1 lg) and/or wild-type pcDNA-3.1-PDE-5 (5 lg) (A) Western blot probed with anti-caspase-3 antibodies showing the expression of recombinant caspase-3 in pCAGGS-Casp-3-transfected cells (B) Autoradiograph showing the effect of Ac-DEVD-CHO (100 l M ) (added at the time of transfection) and recombinant PDE5A1 on caspase-3 activity in Cos-7 cells Caspase-3 activity was measured using [35S]methionine-labeled PARP as a sub-strate These are representative results of three experiments C3 denotes caspase-3.
Trang 5consistent with previous reports showing expression of
functionally active recombinant PDE5 in Cos-7 cells [6]
Western blot analysis with anti-caspase-3 IgG confirmed
expression of recombinant caspase-3 in
pCAGGS-Casp-3-transfected cells Figure 3A shows that the antibody
reacted with five polypeptides of molecular mass
corres-ponding to 35, 30, 17, 12 and 9 kDa in lysates from
caspase-3-transfected cells These proteins were not detected
in lysates from mock-transfected cells These polypeptide
fragments are formed from auto-processing of the protease
Internal cleavage of native protein results in the formation
of p17 and p12, which are catalytically active toward
endogenous protein substrates The formation of p9 might
be due to extensive cleavage of intermediate fragments, as a
result of particularly good overexpression of the enzyme in
Cos-7 cells Cotransfection of PDE5A1 did not affect the
auto-activation of caspase-3 Caspase-3 activity in cell
lysates was also measured using [35S]methionine-labeled
PARP (Mr¼ 115 kDa) as a substrate Figure 3B shows
that there is substantial endogenous caspase-3 activity in
lysates from mock-transfected cell, possibility activated as a
consequence of stressing cells during the transfection
procedure In the current study, endogenous caspase-3
activity converted 70% of the 115 kDa PARP to an
85-kDa fragment (p85) The overexpression of recombinant
caspase-3 in Cos-7 cells resulted in more extensive
proteo-lysis of the exogenous 115 kDa PARP in the assay
(Fig 3B) Caspase-3 activity was completely abolished by
treatment of the cells with the caspase-3/7 inhibitor,
Ac-DEVD-CHO (added at the time of transfection with
caspase-3 plasmid construct) Overexpression of PDE5A1
did not inhibit the auto-activation of caspase-3 (Fig 3B)
This is in line with results showing that PDE5A1 did not affect auto-proteolysis of caspase-3 (Fig 3A)
We investigated the effect of overexpressing recombinant caspase-3 on PDE5A1 in transfected Cos-7 cells 98 kDa PDE5A1 levels were markedly reduced by 60–75% in lysates of cells cotransfected with caspase 3 and PDE5A1 plasmid constructs (Fig 4A,B) This is consistent with depletion of the enzyme via caspase-3-mediated cleavage
An 82-kDa fragment appeared only in lysates of cells overexpressing both enzymes (Fig 4A,B) No other frag-ments were detected on Western blots The accumulation of
82 kDa fragment was not correlated with a similar reduc-tion in the native 98 kDa PDE5A1 level The most likely hypothesis is that caspase-3 proteolyses PDE5A1 as it is expressed and that the 82 kDa fragment thus formed, is then immediately processed further In addition, caspase-3 may act on other proteases that cleave PDE5A1 This in itself is a potentially important and interesting finding as it might suggest a hitherto unidentified caspase-3 initiated protease cascade regulating PDE5A1 activity
Fig 4 The interaction of caspase-3 with PDE5A1 in Cos-7 cells Cells
were transfected with pCAGGS-Casp-3 cDNA (1 lg) and/or
wild-type or truncated D781 pcDNA-3.1-PDE5 (5 lg) plasmid constructs.
(A) Western blot probed with anti-PDE5 IgG showing the effect of
Ac-DEVD-CHO (100 l M ) on the cleavage of PDE5A1 by caspase-3 in
transfected Cos-7 cells The position of the truncated D781 mutant on
SDS/PAGE expressed in Cos-7 cells is also shown; (B) Western blot
probed with anti-PDE5 IgG showing the proteolysis of wild-type
PDE5A1 by recombinant caspase-3 in transfected Cos-7 cells to reveal
the faster migrating 82 kDa fragment These are representative results
of at least three separate experiments C3 denotes caspase-3.
Fig 5 Changes in activity of PDE5A1 upon cleavage by caspase-3 Cells were transfected with pCAGGS-Casp-3 cDNA (0.1–1 lg) and/or wild-type or truncated D781 pcDNA-3.1-PDE5 (5 lg) plasmid con-structs The histogram shows the effect of overexpressing recombinant caspase-3 and the treatment of cells with Ac-DEVD-CHO (100 l M ) on wild-type recombinant PDE5A1 activity in Cos-7 cells PDE5A1 activity was measured at 0.5 l M [ 3 H]cGMP Results are expressed as the fold increase over basal PDE activity in mock-transfected cells D781 truncated PDE5A1 was expressed as an inactive enzyme Inset is the corresponding Western blot showing 98 kDa PDE5A1 levels The
82 kDa fragment is not evident as the Western blot is underexposed to better demonstrate the increase in 98 kDa PDE5A1 in Ac-DEVD-CHO-treated cells In the latter case, cells were transfected with pCAGGS-Casp-3 cDNA (1 lg) and wild-type or truncated D781 pcDNA-3.1-PDE5 (5 lg) plasmid constructs These are representative results of at least three separate experiments C3 denotes caspase-3.
Trang 6The reduction in 98 kDa PDE5A1 levels was correlated
with a decrease in PDE5A1 activity (Fig 5) The remaining
PDE activity in caspase-3/PDE5A1 transfected cells was
recovered by gel filtration on Superose 12 with a similar
elution compared with PDE5A1 from cells overexpressing
this enzyme alone (Fig 6) Further evidence to support the
possibility that PDE5A1 interacts with caspase-3 and
indirectly with caspase-3-activated proteases was shown
by results showing that the caspase-3/7 inhibitor,
Ac-DEVD-CHO abolished the reduction in PDE5A1 levels
observed in cells cotransfected with PDE5A1 and caspase-3
(Fig 4A) This was correlated with the reversal of the
reduction in cGMP hydrolysing PDE activity (Fig 5) It is
interesting to note that the treatment of cells with
Ac-DEVD-CHO appeared to increase 98 kDa PDE5A1
levels and activity above controls, consistent with an action
of endogenous caspase-3/7 (Figs 4A and 5) It remains to
be determined which of the potential caspase sites is
cleaved to inactivate the enzyme However, only two sites
exhibit strong consensus for caspase-3 26DHWD29 and
778DQGD781 Cleavage at78DQGD781would produce an
82-kDa fragment We cannot ascertain at the moment
whether cleavage at 78DQGD781 causes inactivation, as
there is no correlation in the reduction in 98 kDa protein
levels with the appearance of the 82 kDa fragment
Importantly, as the overexpression of caspase-3 in Cos-7
cells induces cell death [32], we conclude from the current
findings that cleavage and inactivation of PDE5A1 medi-ated by caspase-3 may be associmedi-ated with this process However, further studies are necessary to establish whether the cleavage of PDE5A1 is a key event governing cell death
To demonstrate the robustness of the interaction between caspase-3 and PDE5A1, we repeated the experiments in PC12 cells In contrast with Cos-7 cells, the treatment of PC12 cells with Ac-DEVD-CHO did not modulate the expression level of recombinant PDE5A1 (Fig 7A), indi-cating that endogenous caspase-3 activity is not a factor that might influence the native state of recombinant PDE5A1
in this case However, in common with Cos-7 cells,
Fig 7 Effect of caspase-3 and staurosporine on PDE5A1 proteolysis Cells were transfected with pCAGGS-Casp-3 cDNA (1 lg) and/or wild-type pcDNA-3.1-PDE5 (5 lg) plasmid constructs Cells stimu-lated with and without staurosporine (10 l M , 24 h) were transfected only with wild-type pcDNA-3.1-PDE5 (5 lg) plasmid construct (A) Western blot probed with anti-PDE5 IgG showing the proteolysis of wild-type PDE5A1 by recombinant caspase-3 (and the effect of Ac-DEVD-CHO (100 l M ) added at the time of transfection) and in response to staurosporine in PC12 cells Also shown is a histogram of the corresponding reduction in PDE5A1 activity (B) Western blot probed with anti-PDE5 IgG showing the proteolysis of wild-type PDE5A1 in Cos-7 cells stimulated with staurosporine Also shown is a histogram of the corresponding reduction in PDE5A1 activity All activities were measured using samples equalized for protein PDE5A1 activity was measured at 0.5 l M [ 3 H]cGMP These are representative results of at least 2–4 separate experiments C3 denotes caspase-3.
Fig 6 Elution of PDE5A1 from Superose-12 Cells were transfected
with pCAGGS-Casp-3 cDNA (1 lg) and/or wild-type or truncated
D781 pcDNA-3.1-PDE5 (5 lg) plasmid constructs The figure shows
Western blots of chromatographic fractions eluted from Superose 12
probed with anti-PDE5 IgG and a PDE5A1 activity profile (taken
from high-speed supernatants of cells overexpressing caspase-3/
PDE5A1) Total elution volume was 35 mL, with 1-mL fractions.
These are representative results of at least three separate experiments.
Trang 7overexpression of recombinant caspase-3, results in the
reduction of 98 kDa PDE5A1 levels, concomitant with a
similar decrease in cGMP PDE activity (Fig 7A)
The effect of the apoptotic agent, staurosporine
We also investigated whether apoptotic agents induce
the cleavage of PDE5A1 For this purpose we used,
staurosporine (PKCinhibitor), which has been shown by
Brophy et al [35] to activate caspase-3 activity in Cos-7
cells Figure 7A,B shows that the treatment of
PDE5A1-transfected Cos-7 and PC12 cells, respectively, with
stauro-sporine caused a marked reduction in 98 kDa PDE5A1
levels and PDE activity These findings suggest that there is
specificity in the interaction between caspase-3 and
PDE5A1 that requires application of an apoptotic stimulus
DQGD(778–781) site
Proteolysis of the DQGD(778–781) by caspase-3 might
affect catalytic activity of PDE5A1 as the site is within the
boundary of the active site In addition, cleavage at this site
would produce an 82-kDa fragment To test whether a
potential cleavage of the DQGD(778–781) site might affect
catalytic activity of PDE5A1, we created a truncated D781
mutant corresponding exactly to the 82 kDa fragment This
mutant was expressed equally well compared with the
wild-type enzyme in transfected Cos-7 cells, comigrated with the
82 kDa fragment formed from the cleavage of PDE5A1 in
cells cotransfected with caspase-3 (Fig 4A) and was inactive
(Fig 5) Inactivation of the truncated mutant was no due to
potential misfolding of the enzyme This was shown by results showing that the truncated mutant eluted from Superose 12 at the same position compared with the wild-type enzyme (Fig 6), suggesting similar hydrodynamic properties
The inactivity of the truncated mutant provides indirect support for the possibility that cleavage of DQGD(778–781)
is one potential mechanism that might lead to inactivation
of PDE5A1 activity In this regard, we found that a more subtle change in PDE5A1 using a single point mutation at D781 (replaced with A) also results in a reduction of PDE activity The D781A mutant partial loss of PDE5A1 activity
to 70% of the wild type measured at 0.5 lMcGMP The reduction in PDE activity was due, in part, to an increase in the Kmfor cGMP The Kmfor the wild-type enzyme was 2.2 lMcompared with 8.4 lMfor the D781A mutant The kinetic constants were determined in samples where the expression level of the D781 mutant PDE5A1 mg)1 cell lysate protein was approximately twice that of the wild-type enzyme Assays were normalized for protein From this data, we calculated that the mutant PDE5A1 exhibits a
Vmaxthat is approximately 50% of the wild-type enzyme These findings are in agreement with studies by Turko et al [14], who reported identical changes in the kinetic constants
of the D781A mutant
The DQGD site is within the boundary of the catalytic domain of PDE5A1 (Fig 1) We have used the X-ray crystal structure of PDE4B2 [34] as a template to generate a homology model for PDE5A1 to rationalize the structural implications of a caspase-3-catalyzed proteolysis at the PDE5A1 DQGD(778–781) site From the PDE5
Fig 8 PDE5 homology model Homology model of PDE5A1 based on PDE4B2 crystal structure showing how the removal of the C-terminal tail containing Q807 and F810 by caspase-3 affects the architecture of the cata-lytic site, and in particular interaction with Q765.
Trang 8homology model, proteolytic cleavage at DQGD(778–781)
in PDE5A1 might be expected to remove the C-terminal tail
(highlighted in yellow in Fig 8) containing Q807 and F810,
which are potentially important amino acids required for
substrate binding Q807 is completely conserved across the
PDE superfamily and, in principle, might accept either the
guanine base of cGMP or the adenine base of cAMP F810
in PDE5A1 is conserved in PDE4B2 as F446, and this
residue has been shown by site-directed mutagenesis to be
essential for catalytic competence in PDE4 and to play a key
role in the binding of competitive PDE4 inhibitors [36] The
side chain of this residue may conceivably p-stack with the
purine base of the bound substrate and form hydrophobic
interactions with a number of inhibitors In PDE4B2 the
sequence QQGD(416–419), corresponding to the PDE5A1
caspase-3 site DQGD(778–781) is identical, except that the
site is disabled by replacement of D for Q at the P4 position
The site is located on the exposed C-terminal end of helix 14
in the PDE4B2 crystal structure In conclusion, the
caspase-3-catalyzed cleavage at DQGD(778–781) in PDE5A1 will
very likely remove a key wall from the catalytic site
containing Q807/F810 This might prevent potential
inter-action with critical adjacent amino acid residues present on
the other side of the catalytic pocket, identified by Turko
and colleagues, such as Q765 [13] The removal of part of
the catalytic pocket explains the inactivity of the
engine-ered protein truncated at D781 Potential cleavage at
DQGD(778–781) by caspase-3 could severely disrupt the
structure of PDE5A1 Interestingly, there is substantial
similarity between the amino acid sequence of the PDE5A1
DQGD(778–781) site and the corresponding region in
PDE2A3, PDE4C , PDE4D, PDE6ab and PDE11A1 D778
at P4 of the caspase-3 consensus site in PDE5A1 is replaced
with E in PDE11A1 and PDE6ab, Q in PDE4C, R in
PDE4D and S in PDE2A3 Therefore, the replacement of
the4D effectively disables the caspase-3 site in these PDE
isoforms
Summary
The results presented in this article are consistent with
PDE5A1 acting as a substrate for caspase-3 in intact cells In
addition, PDE5A1 may be subject to cleavage by a
caspase-3-initiated protease(s) event These results raise the
possi-bility of a role for PDE5A1 in apoptosis However, further
investigation is required to establish a causal linkage
between PDE5A1 cleavage and apoptosis
Acknowledgments
This study was supported by the BBSRC.
References
1 Beavo, J.A (1995) Cyclic nucleotide phosphodiesterases:
functional implications of multiple isoforms Physiol Rev 75,
725–748.
2 Beavo, J.A & Brunton, L.L (2002) Cyclic nucleotide research –
still expanding after half a century Nat Rev Mol Cell Biol 3 (9),
710–718.
3 Houslay, M.D & Adams, D.R (2003) PDE4 cAMP
phos-phodiesterases: modular enzymes that orchestrate signalling
cross-talk, desensitisation and compartmentalisation Biochem.
J 370, 1–18.
4 Corbin, J.D & Francis, S.H (1999) Cyclic GMP phosphodies-terase-5: target of sildenafil J Biol Chem 274, 13729–13732.
5 McAllister-Lucas, L.M., Sonnenburg, W.K., Kadlecek, A., Seger, D., Trong, H.L., Colbran, J.L., Thomas, M.K., Walsh, K.A., Francis, S.H., Corbin, J.D & Beavo, J (1993) The structure of a bovine lung cGMP-binding, cGMP-specific phos-phodiesterase deduced from a cDNA clone J Biol Chem 268, 22863–22873.
6 Thomas, M.K., Francis, S.H & Corbin, J (1990) Characteriza-tion of a purified bovine lung cGMP-binding cGMP phospho-diesterase J Biol Chem 265, 14964–14970.
7 Burns, F., Rodger, I.W & Pyne, N.J (1992) The catalytic subunit
of protein kinase A triggers activation of the type V cyclic GMP-specific phosphodiesterase from guinea-pig lung Biochem J 283, 487–491.
8 Frame, M., Wan, K.-F., Tate, R., Vandenabeele, P & Pyne, N.J (2001) The gamma subunit of the rod photoreceptor cGMP phosphodiesterase can modulate the proteolysis of two cGMP binding cGMP-specific phosphodiesterases (PDE6 and PDE5) by caspase-3 Cell Signal 13, 735–741.
9 Aravind, L & Ponting, C.P (1997) The GAF domain: an evolu-tionary link between diverse phototransducing proteins Trends Biochem Sci 22, 458–459.
10 Ho, Y.S., Burden, L.M & Hurley, J.H (2000) Structure of the GAF domain, a ubiquitous signaling motif and a new class of cyclic GMP receptor EMBO J 19, 5288–5299.
11 Martinez, S.E., Wu, A.Y., Glavas, N.A., Tang, X.B., Turley, S., Hol, W.G & Beavo, J.A (2002) The two GAF domains in phosphodiesterase 2A have distinct roles in dimerization and in cGMP binding Proc Natl Acad Sci USA 99, 13260–13265.
12 McAllister-Lucas, L.M., Haik, T.L., Colbran, J.L., Sonnenburg, W.K., Seger, D., Turko, I.V., Beavo, J.A., Francis, S.H & Corbin, J.D (1998) An essential aspartic acid at each of two allosteric cGMP-binding sites of a cGMP-specific phosphodiesterase.
J Biol Chem 270, 30671–30679.
13 Turko, I.V., Francis, S.H & Corbin, J.D (1998) Hydropathic analysis and mutagenesis of the catalytic domain of the cGMP-binding cGMP-specific phosphodiesterase (PDE5) cGMP versus cAMP substrate selectivity Biochemistry 37, 4200–4205.
14 Turko, I.V., Francis, S.H & Corbin, J.D (1998) Potential roles of conserved amino acids in the catalytic domain of the cGMP-binding cGMP-specific phosphodiesterase J Biol Chem 273, 6460–6466.
15 Francis, S.H., Turko, I.V., Grimes, K.A & C orbin, J.D (2000) Histidine-607 and histidine-643 provide important interactions for metal support of catalysis in phosphodiesterase-5 Biochemistry
39, 9591–9596.
16 Nunez, G., Benedict, M.A., Hu, Y.M & Inohara, N (1998) Caspases: the proteases of the apoptotic pathway Oncogene 17, 3237–3245.
17 Alnemri, E.S., Fernandes-Alnemri, T & Litwack, G (1995) Cloning and expression of four novel isoforms of human inter-leukin-1 beta converting enzyme with different apoptotic activities.
J Biol Chem 270, 4312–4317.
18 Hengartner, M.O., Ells, R.E & Horvitz, H.R (1992) Caenor-habditis elegans gene ced-9 protects cells from programmed cell death Nature 356, 494–499.
19 Van De Craen, M., Vandenabeele, P., Declercq, W., Van Den Brande, I., Van Loo, G., Molemans, F., Schotte, P., Van Criek-inge, W., Beyaert, R & Friers, W (1997) Characterization of seven murine caspase family members FEBS Lett 403, 61–69.
20 Van De Craen, M., Declercq, W., Van Den Brande, I., Friers, W.
& Vandenabeele, P (1999) The proteolytic procaspase activation network: an in vitro analysis Cell Death Differ 11, 1117–1124.
Trang 921 Cardone, M.H., Salvesen, G.S., Widmann, C., Johnson, G &
Frisch, S.M (1997) The regulation of anoikis: MEKK-1
activa-tion requires cleavage by caspases Cell 90, 315–323.
22 Graves, J.D., Gotoh, Y., Draves, K.E., Ambrose, D., Han, D.K.,
Wright, M., Chernoff J., Clark, E.A & Krebs, E.G (1998)
Cas-pase-mediated activation and induction of apoptosis by the
mammalian Ste20-like kinase Mst1 EMBO J 17, 2224–2234.
23 Rudel, T & Bokoch, G.M (1997) Membrane and morphological
changes in apoptotic cells regulated by caspase-mediated
activa-tion of PAK2 Science 276, 1571–1574.
24 Francois, F & Grimes, M.L (1999) Phosphorylation-dependent
Akt cleavage in neural cell in vitro reconstitution of apoptosis.
J Neurochem 73, 1773–1776.
25 Frutos, S., Moscat, J & Diaz-Meco, M.T (1999) Cleavage of
zetaPKCbut not lambda/iotaPKCby caspase-3 during
UV-induced apoptosis J Biol Chem 274, 10765–10770.
26 Van de Water, B., Nagelkerke, J.F & Stevens, J.L (1999)
Dephosphorylation of focal adhesion kinase (FAK) and loss of
focal contacts precede caspase-mediated cleavage of FAK
during apoptosis in renal epithelial cells J Biol Chem 274,
13328–13337.
27 Baixeras, E., Garcia-Lozano, E & Martinez, A.-C (1996)
Decrease in cAMP levels promoted by CD48–CD2 interaction
correlates with inhibition of apoptosis in B cells Scan J Immunol.
43, 406–412.
28 Tortorella, C., Piazzolla, G., Spaccavento, F & Antonaci, S.
(1998) Effects of granulocyte-macrophage colony-stimulating
factor and cyclic AMP interaction on human neutrophil
apopto-sis Med Inflamm 7, 391–396.
29 Shimojo, T., Hiroe, M., Ishiyama, S., Ito, H., Nishikawa, T &
Marumo, F (1999) Nitric oxide induces apoptotic death of
cardiomyocytes via a cyclic-GMP-dependent pathway Exp Cell Res 247, 38–47.
30 Li Y., Maher, P & Schubert, D (1997) Requirement for cGMP in nerve cell death caused by glutathione depletion J Cell Biol 139, 1317–1324.
31 Huston, E., Beard, M., McCallum, F., Pyne, N.J., Vandenabeele, P., Scotland, G & Houslay, M.D (2000) The cAMP-specific phosphodiesterase PDE4A5 is cleaved downstream of its SH3 interaction domain by caspase-3 Consequences for altered intracellular distribution J Biol Chem 275, 28063–28074.
32 Stennicke, H.R., Jurgansmeier, J.M., Shin, H., Deveraux, Q., Wolf, B.B., Yang, X., Zhou, Q., Ellerby, H.M., Bredesen, D., Green, D.R., Froelich, C.J & Salvesen, G.S (1998) Pro-caspase-3
is a major physiologic target of caspase-8 J Biol Chem 273, 27084–27090.
33 Thompson, J & Appleman, M (1971) Multiple cyclic nucleotide phosphodiesterase activities from rat brain Biochemistry 10, 311– 316.
34 Xu, R.Y., Hassell, A.M., Vanderwall, D., Lambert, M.H., Hol-mes, W.D., Luther, M.A., Rocque, W.J., Milburn, M.U., Zhao, Y., Ke, H & Nolte, R.T (2000) Atomic structure of PDE4: insights into phosphodiesterase mechanism and specificity Science
288, 1822–1825.
35 Brophy, V.A., Tavare, J.M & Rivett, A.J (2002) Treatment of COS-7 cells with proteasome inhibitors or gamma-interferon reduces the increase in caspase 3 activity associated with staurosporine-induced apoptosis Arch Biochem Biophys 397, 199–205.
36 Richter, W., Unciuleac, L., Hermsdorf, T., Kronbach, T & Dettmer, D (2001) Identification of inhibitor binding sites of the cAMP-specific phosphodiesterase 4 Cell Signall 13, 287–297.