TbPDE1, a novel class I phosphodiesterase of Trypanosoma brucei Stefan Kunz1, Thomas Kloeckner2, Lars-Oliver Essen3,*, Thomas Seebeck1and Michael Boshart2 1 Institute of Cell Biology, Un
Trang 1TbPDE1, a novel class I phosphodiesterase of Trypanosoma brucei Stefan Kunz1, Thomas Kloeckner2, Lars-Oliver Essen3,*, Thomas Seebeck1and Michael Boshart2
1 Institute of Cell Biology, University of Bern, Switzerland; 2 Department of Biology I, University of Munich, Germany;
3 MaxPlanck Institute for Biochemistry, Martinsried, Germany
Cyclic nucleotide specific phosphodiesterases (PDEs) are
important components of all cAMP signalling networks
In humans, 11 different PDE families have been identified to
date, all of which belong to the class I PDEs
Pharmaco-logically, they have become of great interest as targets for the
development of drugs for a large variety of clinical
condi-tions PDEs in parasitic protozoa have not yet been
exten-sively investigated, despite their potential as antiparasitic
drug targets The current study presents the identification
and characterization of a novel class I PDE from the
para-sitic protozoon Trypanosoma brucei, the causative agent of
human sleeping sickness This enzyme, TbPDE1, is encoded
by a single-copy gene located on chromosome 10, and it functionally complements PDE-deficient strains of Sac-charomyces cerevisiae Its C-terminal catalytic domain shares about 30% amino acid identity, including all functionally important residues, with the catalytic domains of human PDEs A fragment of TbPDE1 containing the catalytic domain could be expressed in active form in Escherichia coli The recombinant enzyme is specific for cAMP, but exhibits
a remarkably high Kmof > 600 lMfor this substrate Keywords: African trypanosomes; cAMP signaling; class I phosphodiesterase; sleeping sickness
Cyclic AMP is involved in the regulation of numerous
biological functions, such as the control of metabolic
pathways in eubacteria [1], differentiation and virulence in
fungi [2], cell aggregation in Dictyostelium [3], transduction
of gustatory and olfactory signals [4], the control of
rhythmic oscillations in heart and brain [5] and learning
and long-term memory formation [6] in multicellular
organisms In eukaryotic cells, hydrolysis of cAMP by
cyclic nucleotide specific phosphodiesterases (PDEs) is the
only means of rapidly inactivating the cAMP signal PDEs
represent a large and divergent group of enzymes, and two
distinct PDE classes have been identified [7,8] Class I
enzymes include all currently known families of mammalian
PDEs, as well as a number of PDEs from lower euk aryotes,
such as PDE2 from the yeast Saccharomyces cerevisiae [8] or
the product of the regA gene of Dictyostelium discoideum [9]
In mammals, 11 distinct class I PDE families have been
identified, based on DNA sequence analysis and on the
pharmacological profiles of the enzymes [10,11] At the
amino acid level, family members exhibit > 50% sequence
identity within a conserved catalytic core of about 250
amino acids Between families, the sequence identity drops
to 30–40% in the same region [12], and no significant similarity is found outside the catalytic domain
Considering the importance of the PDEs for signal transduction, it is not unexpected that mutations in PDE genes have been recognized as the underlying cause of several genetic diseases [13–15] In clinical pharmacology, the PDEs have also become highly attractive targets for drug development, and a large number of highly family-specific inhibitors have been developed PDE inhibitors are under exploration, or already in clinical use, for ailments as diverse as autoimmune diseases, arthritis, asthma, impo-tency and as anti-inflammatory agents (reviewed in [16–18])
In view of the spectacular success of PDE inhibitors as chemotherapeutics, it is surprising how little effort has been made so far to explore the PDEs of parasites as potential targets for antiparasitic drugs The African trypanosome Trypanosoma bruceiis the protozoon that causes the fatal human sleeping sickness, as well as Nagana, a devastating disease of domestic animals in large parts of sub-Saharan Africa While many aspects of trypanosome cell biology have been extensively studied, very little is still known about cAMP signalling [19–22] Early workhas shown that the steady-state concentration of cAMP varies during the life cycle of the parasite in its mammalian host [23] Vassella
et al have provided evidence for a crucial role of cAMP
in triggering population-density induced differentiation of long-slender to short-stumpy bloodstream forms in culture [24] An early study on PDEs demonstrated PDE activity in cell lysates of the bloodstream form of T brucei [25] Recently, a small gene family coding for class I PDEs (TbPDE2) was identified in T brucei, and their gene products were characterized as cAMP-specific PDEs [26– 28] The current study describes the identification of a novel class I PDE from T brucei, TbPDE1 This enzyme bears no sequence similarity to any of the other class I PDE families
Correspondence to T Seebeck, Institute of Cell Biology,
Baltzerstrasse 4, CH-3012 Bern, Switzerland.
Fax: + 41 31 631 46 84, Tel.: + 41 31 631 46 49,
E-mail: thomas.seebeck@izb.unibe.ch
Abbreviations: PDE, cyclic-nucleotide specific phosphodiesterase;
IBMX, isobutyl-methyl-xanthine; IC 50 , 50% inhibitory
concentrations.
Note: A web site is available at http://www.izb.unibe.ch
*Present address: Department of Chemistry, Hans Meerwein-Strasse,
Philipps University, D-35032 Marburg, Germany.
(Received 16 October 2003, revised 10 December 2003,
accepted 16 December 2003)
Trang 2outside of the catalytic domain Sequence comparisons
indicate that TbPDE1 of T brucei is different from all PDE
families of its potential mammalian hosts In agreement
with these sequence data, TbPDE1 is also
pharmacologi-cally quite distinct from its mammalian counterparts, as
judged from its sensitivity to a number of established PDE
inhibitors Finally, TbPDE1 is a nonessential enzyme under
culture conditions or during the midgut infection of tsetse
flies, as was demonstrated earlier with deletion mutants for
this gene [29]
Materials and methods
Materials
5-Fluoroorotic acid monohydrate was from American
Bioorganics SuperTaq polymerase was from Anglia
Bio-tech Benzamidine, antipain, leupeptin,
phenylmethane-sulfonyl fluoride, and Ba(OH)2 solution (Cat number
14-3) were from Sigma Adenosine-3,5¢-cyclic
monophos-phate and adenosine-5¢-monophosmonophos-phate were from Roche
Molecular The radiochemicals [2,8-3H]adenosine-3¢5¢-cyclic
monophosphate (25–40· 1010BqÆmmol)1) and [3
H]adeno-sine-5¢-monophosphate (15–30 · 1010BqÆmmol)1) were
from NEN PDE inhibitors were from the following sources:
isobutyl-methyl-xanthine (IBMX), Sigma; etazolate,
Calbi-ochem; IBMQ and rolipram were generous gifts from Glaxo
Wellcome and Smith Kline Beecham, respectively
Trypanosomes
Procyclic trypanosomes (stock427) were grown in SDM-79
medium containing 5% fetal bovine serum [30] A
mono-morphic variant of AnTat1.1 [31] was cultivated as
described by Hesse et al [32]
Yeast strains
Strain PP5-12 (MATa leu2-3 leu2-112 ura3-52 his3-532 his4
cam pde1::ura3FOA–Respde2::HIS3) was derived from strain
PP5 [33]; a gift of J Colicelli (UCLA) by selection on
5-fluoroorotic acid [34] Strain YMS5 (MATa leu2 ura3 his4
lys2 pde1::LYS2 pde2::LEU2 pep4::ura3FOA–Res35 was
kindly provided by P Engels (Novartis Ltd)
Complementation screening
The phosphodiesterase-deficient, uracil auxotroph yeast
strain PP5-12 was transformed with a trypanosome
expression library The selectable phenotype of PP5-12
is heat-shocksensitivity The library (a kind gift of
R Schwartz, University of Marburg) contained cDNA
from bloodstream form trypanosomes of stock427, clone
221 in the yeast expression vector p426MET [35], which is a
2l plasmid with the repressible MET25 promotor and the
URA3 selection marker The cloning site of this plasmid
was derived from pBS SK(–) (Stratagene) The cDNA
library was inserted via the XhoI and EcoRI sites, and the
MET25 promotor (381 bp) was introduced between the
XbaI and the SacI sites Yeast transformation was carried
out exactly as described [36] Transformants were grown
for 3 days on selective medium lacking methionine and
uracil (SC–met–ura) to maintain the plasmid and to derepress the expression of the cDNA In order to select for complementation, the transformants were replica-plated onto plates prewarmed to 55C and incubated at this temperature for 15 min Plates were then cooled and incubated at 30C for 3 days Heat-shockresistant colon-ies were rescreened for heat-shocksensitivity Patches were replica-plated onto YPD plates prewarmed to 55C, and the heat shockwas continued for 15 min After cooling the plates to room temperature, they were incubated for 2–3 days at 30C Candidate clones were subjected to segregation analysis, and positive plasmids were finally used
to retransform PP5-12 in order to confirm the phenotype carried by the plasmid
Direct PCR screening of plasmids Screening of large numbers of yeast colonies for the presence of a plasmid insert was done by a rapid PCR procedure Colonies were picked and grown at 30C in
5 mL selective medium to high density (18–24 h) Cell culture (1.5 mL) was pelleted and resuspended in 100 lL
H2O The suspension was boiled for 5 min and then centrifuged for 30 s at 8400 g Five microlitres of the supernatant were taken as input into 50 lL PCR reac-tions Plasmid inserts were amplified using primers derived from the pBS SK(–) multicloning site: primer BS(+) forward: 5¢-GTTTTCCCAGTCACGACGTTG-3¢; and primer BS(+) back: 5¢-ACCATGATTACGCCAAGC GCG-3¢ Amplification was performed in a Perkin-Elmer thermal cycler using the following conditions: One cycle of
5 min at 94C, 1 min at 50 C, 2 min at 72 C, followed
by 30 cycles of 1 min at 94C, 1 min at 55 C, 2 min at
72C, followed by a final extension step of 5 min at
72C
Cloning and expression of the TbPDE1 locus
A genomic DNA fragment containing TbPDE1 was isolated from a k-DASHTM library constructed from genomic DNA of strain AnTat 1.1 which had been partially digested with Sau3A and packaged with the Gigapack II kit (Stratagene) (R Kraemer, unpublished results) The restriction map of subclones pCK16-1 and pCK59-1 matched the map of the genomic TbPDE1 locus derived from Southern blot analysis (T Kloeckner, unpublished results) Genomic Southern blots were hybridized with a PCR-amplified subfragment of plasmid pCK16-1 repre-senting amino acids 177–602 of TbPDE1
Identification of 5¢ and 3¢ termini of the TbPDE1 mRNA The mini-exon addition site was mapped by RT/PCR using primer 16-SP13 (5¢-ATTCGCTCGTTGATTTC-3¢) for reverse transcription (RT), and a mini-exon primer M4 (5¢-GGGAATTCCGCTATTATTAGAACAGTTTCT-3¢, added EcoRI site shown in bold) together with the TbPDE1-specific antisense primer 16-SP14 (5¢-AGC AGTTTGAAGCATTG-3¢) for amplification The prod-ucts were cloned via the EcoRI site in the M4 primer and
an internal XbaI site, and they were analysed by sequencing
Trang 3Expression of TbPDE1 inS cerevisiae
The ORF of TbPDE1 was cloned into the pLT1 expression
vector This vector was derived from p425CYC1 by
replacing the CYC1 promotor by the much stronger
TEF2 promotor [35] followed by the original Kozak
sequence 5¢-CTAAAC-3¢ and a start codon The complete
TbPDE1ORF was expressed either containing a His6tag
at its N terminus, or a His6 tag followed by a
haemag-glutinin tag to facilitate detection of the recombinant
protein Transformants were selected on synthetic minimal
medium containing 0.67% (w/v) yeast nitrogen base
without amino acids (DIFCO) and 2% (w/v) glucose,
supplemented with an amino acid mixture lacking leucine
(SC–leu)
For the preparation of lysates from cells expressing
TbPDE1, yeast cells grown to mid- to end-log phase in
SC–leu medium were collected, resuspended quickly in the
original volume of prewarmed YPD medium and
incuba-ted for an additional 3 h at 30C in order to maximize
protein expression Cells were then harvested, washed once
in H2O and once in HHB buffer (Hank’s balanced salt
solution, containing 50 mM Hepes, pH 7.5) The washed
cell pellet was suspended in an equal volume of HHB
containing a protease inhibitor cocktail (CompleteTM,
Roche Molecular Biochemicals) Cells were lysed by
grinding with glass beads (425–600 lm; Sigma) in 2 mL
Sarstedt tubes using a FastPrep FP120 cell disruptor
(3· 45 s at setting 4) After cell breakage, a hole was
punched in the bottom of the tube with a needle, the tube
was placed on top of a 5 mL plastic tube and was
centrifuged in an SS34 rotor for 6 min at 4340 g and 4C
This step left the glass beads in the Sarsted tube while the
cell lysate was collected in the plastic tube, where unbroken
cells and large cell fragments formed a pellet The
supernatant was transferred to a fresh tube, clarified by
centrifugation for 15 min at 15 000 g and the clarified
supernatant was used for the assays
Expression of TbPDE1 inE coli
The gene encoding full-length TbPDE1 (residues Met1–
Thr620) was amplified from of T brucei 927 genomic DNA
(kindly provided by S Melville, Cambridge University)
using Takara Taq polymerase (BioWhittaker) and 30 cycles
of 30 s at 94C, 2 min at 58 C and 5 min at 72 C For
amplification, the primer pairs 5¢-GGGAATTCCATA
TGCTTGAGGCTTTGCGAAAGTGCCCGACCATGT
TTG-3¢ (NdeI site in bold) and 5¢-CCGCTCGAGT
CATTACTAGGTTCCCTGTCCAGTGTTACC-3¢ (XhoI
site in bold) were used The resulting 1.86-kbp fragment was
subsequently cloned into the NdeI/XhoI-cut expression
vector pET28a (Novagen; kanamycin-resistance marker),
resulting in plasmid pET-PDE1 Two gene fragments
coding for N-terminally truncated fragments of TbPDE1
were also amplified using the same protocol and pET-PDE1
as template PDE1(Arg189–Thr620) was amplified using the
primer pairs 5¢-GGGAATTCCATATGAGAGACAATA
TTTCCCGTTTATCAAATC-3¢ and 5¢-CCGCTCGAGT
CATTACTAGGTTCCCTGTCCAGTGTTACC-3¢, and
PDE1(Lys321–Thr620) was amplified with primers 5¢-GGG
AATTCCATATGAAGAATGATCAATCTGGCTGCG
GCGCAC-3¢ and 5¢-CCGCTCGAGTCATTACTAGG TTCCCTGTCCAGTGTTACC-3¢ The resulting DNA fragments (1.29 and 0.90 kbp) were digested with NdeI and XhoI and cloned into 28a The constructs pET-PDE1, pET-PDE1(R189–T620) and pET-PDE1(K321– T620) were verified by DNA sequencing
Expression and purification of full-length and truncated PDE1/His6-fusion proteins The plasmid constructs pET-PDE1, pET-PDE1(R189– T620) and pET-PDE1(K321–-T620) were transformed into E coli BL21(DE3) cells For protein expression, overnight cultures were grown at 37C in Luria–Bertani medium containing 50 lgÆmL)1 kanamycin Fresh over-night cultures were inoculated at a dilution of 1 : 50 into
TB medium [1.2% (w/v) Bacto-Tryptone, 2.4% (w/v) Bacto yeast extract, 0.4% (v/v) glycerol, 0.017M
KH2PO4, 0.072M K2HPO4, pH 7.5] containing
50 lgÆmL)1 kanamycin Cultures were incubated on a rotary shaker at 25C at 220 r.p.m until D595of 0.6–0.9 was reached (about 4 h) The cultures were induced by the addition of 0.5 mMisopropyl thio-b-D-galactoside and were shaken at 25C for a further 4 h Cells were harvested by centrifugation and washed once in NaCl/Pi The washed cell pellet was frozen in liquid nitrogen and stored at )70 C For protein purification, the frozen cell pellet was suspended in 1/40–1/30 of culture volume in extraction buffer [50 mM Na/phosphate buffer, pH 7.0,
300 mM NaCl, 5 mM MgCl2, 0.1% (v/v) Tween-20] containing a protease inhibitor cocktail (Complete, Roche Molecular Biochemicals) Cells were lysed by sonication (four pulses of 15 s with intermittant cooling
in an ice/water bath) The lysate was clarified by centrifugation at 16 000 g for 20 min at 4C Of the supernatant, 1.2 mL were added to a tube containing
250 lL bed volume of Talon resin (Clontech) preequili-brated with extraction buffer The tube was rotated for
30 min on a rotary shaker at 4C The resin was then washed once with 1.5 mL of wash buffer 1 (extraction buffer) and twice with 1.5 mL wash buffer 2 (extraction buffer containing 5 mM imidazole) The washed resin was then packed by gravity flow into an 8 mm diameter column, washed with 2.5 mL wash buffer 2, followed by elution of bound protein with elution buffer (extraction buffer containing 150 mM imidazole) Fractions (250 lL) were collected and 1 lL of each fraction were spotted onto nitrocellulose and stained with amido blackto visualize the protein Fractions containing the recombin-ant protein were pooled (750–1000 lL total) and fractionated over a Sephadex G-25 column (NAP, Pharmacia Biotech) preequilibrated with 15 mL NSP buffer (50 mM sodium phosphate buffer, pH 7.5,
300 mM NaCl, 5 mM MgCl2) The fractions containing the eluted protein were analysed spectrophotometrically to ascertain that their imidazole concentration was below
1 mM and were then pooled Finally, the purified protein was mixed with an equal volume of 50% (v/v) glycerol, 0.2% (v/v) Tween-20, 5 mM MgCl2, and aliquots were shock-frozen in liquid nitrogen and stored at )70 C Protein concentrations were determined using the Brad-ford reagent (Bio-Rad) and BSA as a standard
Trang 4PDE assays
PDE activity was assayed by a modification of the
procedure of Schilling et al [37] as described [26] All assays
were performed at 30C in 25 mM Tris/HCl, pH 7.4,
0.5 mMEDTA, 0.5 mMEGTA, 10 mMMgCl2and using
1 lM [3H]cAMP as the substrate After addition of
Ba(OH)2, the samples were allowed to precipitate on ice
for 30 min The precipitate was filtered onto GF/C glass
fibre filters, and radioactivity was measured by scintillation
spectrometry For each set of experiments, control
precipi-tations with [3H]AMP and [3H]cAMP were performed in
order to determine the efficiency of AMP capture in the
precipitate, and the extent of cAMP trapping in the
precipitate Both values were reproducible from experiment
to experiment, and over the whole concentration range used
in our assays AMP was precipitated with an efficiency in
the range of 60% of the input, and cAMP contamination of
the precipitate corresponded to about 0.7% in the input
The amount of AMP produced by PDE activity was
calculated according to:
CAMP¼ CcAMPðaprobe qcAMP acAMPÞ
=ðacAMPðqAMP qcAMPÞÞ
where CcAMPis the cAMP substrate concentration, acAMPis
the total activity used in the enymatic reaction, aprobeis the
total radioactivity on the filter, and qcAMPand qAMPare the
precipitation efficiencies of cAMP and AMP, respectively
In all experiments, < 20% of the substrate was hydrolysed,
and all data points were taken in triplicate or quadruplicate
For inhibitor studies, the test compounds were dissolved in
H2O or dimethylsulfoxide The dimethylsulfoxide
concen-tration in the final assay solutions never exceeded 2%, and
appropriate controls were always included Data were
evaluated using the PRISM 3.0 software package from
GraphPad
Results
Isolation of a PDE gene fromT brucei by functional
complementation inS cerevisiae
At the onset of this project, no sequences with similarities to
PDEs were available in the T brucei genome databases,
and the gene was isolated by complementation screening in
S cerevisiae Yeast strains deficient in PDE activity are
heat-shocksensitive and do not survive exposure to elevated
temperatures [7] This phenotype provided a convenient
screening system for the search for PDE genes In the PP5
yeast strain used for the screening, both endogenous PDE
genes (PDE1 and PDE2) have been disrupted by URA3 and
HIS3 marker genes, respectively [33] Since the
trypano-somal cDNA library to be used was constructed in a vector
carrying the URA3 selection marker for uracil auxotrophy,
the PP5 strain, which is Ura+, first had to be selected on
5-fluoroorotic acid for spontaneous ura–mutants Several
such mutants were isolated and analysed for their genetic
stability The clone with the lowest reversion frequency,
PP5-12, was used for further experiments
PP5-12 was transformed with the cDNA library, and
24 000 transformants were recovered after Ura+
selec-tion Transformants were replica-plated onto SC–met–ura
plates preheated to 55C and were incubated at 55 C for
15 min Plates were then cooled and incubated at 30C for
3 days An exploratory screen had revealed a high frequency (0.5%) of spontaneous heat-shockresistant revertants The
120 heat-shockresistant colonies were thus individually retested, and 109 of them that still proved heat-resistant were analysed further Segregation analysis and retransfor-mation of individual plasmids into PP5-12 resulted in a single plasmid, pBa46, which confered heat-shockresistance upon back-transformation into PP5-12 pBa46 was found
to contain a cDNA fragment representing most of the ORF (amino acids Met159 through the stop codon after Thr620) and 210 bp of the 3¢-untranslated region of a novel PDE gene of T brucei, TbPDE1
Cloning and genomic organization of TbPDE1 While the complementation screen was ongoing, a DNA fragment coding for a protein kinase A-related gene (TbPKAC3) was isolated from a genomic phage library of
T brucei(N Wild and M Boshart, unpublished results) Upon sequencing beyond the 3¢-untranslated region of TbPKAC3, an ORF of 620 amino acids (TbPDE1) was identified that encompassed the cDNA sequence contained
in pBA46 TbPDE1 is a single-copy gene, as several restriction enzymes produced single bands with different mobilities upon Southern blot analysis (Fig 1) The gene is located on chromosome 10 The two hybridizing bands
Fig 1 Genomic organization of TbPDE1 Southern blot analysis of genomic T brucei DNA of strain AnTat1.1 digested with BamHI (B), EcoRI (E), HindIII (H) and XhoI (X) Plasmid controls representing 0.5, 1 and 2 genome equivalents are included in the right hand part of the blot The hybridization probe was a PCR fragment representing amino acids 177–602 of the TbPDE1 open reading frame Molecular mass markers are given on the left.
Trang 5observed with BamHI-digested DNA reflect a polymorphic
BamHI site This was confirmed by restriction mapping of
independent genomic phage clones (data not shown) and
also by independent knockout experiments of TbPDE1 [29]
The observation that TbPDE1 is a single-copy gene was
further supported by quantification of the gene copy
number by using internal plasmid standards equivalent to
0.5, 1 and 2 haploid genome copies of TbPDE1
Hybrid-ization of genomic blots at low stringency (Tm¼ 45 C) as
well as library screening under similar conditions failed to
detect related PDE genes or other putative TbPDE1 family
members Complete sequencing of the cDNA clone pBa46
and of genomic clones revealed a small number of
nucleotide sequence differences despite careful verification
by resequencing This was not unexpected as the genomic
and the cDNA sequences were derived from different strains
of T brucei (see Materials and methods) and because an
allelic polymorphism in the TbPDE1 locus was also detected
by BamHI restriction enzyme analysis (Fig 1)
The trans-splicing site at the 5¢-end of the TbPDE1
transcript was mapped by RT-PCR using two nested
gene-specific primers and a primer directed to the conserved
mini-exon sequence present at the 5¢-end of all trypanosomal
mRNAs Seven out of eight such cDNA clones contained
the mini-exon splice site at position )159 relative to the
translational start, and one clone at position)155 Both
sites were preceded by an AG dinucleotide and a long
polypyrimidine stretch immediately upstream (Fig 2A)
These results demonstrated that the intergenic region
between TbPKAC3 and TbPDE1 is only 117–135 bp long,
as the 3¢-end of theTbPKAC3-transcript had previously
been mapped by RT/PCR (T Kloeckner, unpublished
results) The oligo-A stretch at the 3¢-end of cDNA clone
pBa46 most likely represents the beginning of the polyA tail
of the TbPDE1 mRNA since no corresponding oligoA
stretch is found in the genomic sequence, and since poly
pyrimidine-rich stretches which are typically located
upstream from the polyadenylation sites of other
trypano-somal mRNAs [38] were found upstream of this site
(Fig 2B)
TbPDE1 mRNA is expressed in the bloodstream
and in the procyclic life cycles stages
According to the mapped transcript ends, TbPDE1 should
give rise to an mRNA of approximately 2.5 kb This was
detected in Northern blots using RNA from three different
life cycle stages of T brucei, long-slender and short-stumpy
forms isolated from infected rats, and cultured procyclic
forms (Fig 2C) In good agreement with the results from
Northern blotting experiments, TbPDE1 mRNA was also
detected in cultured bloodstream and procyclic forms by
using real-time RT/PCR (data not shown)
Organization of the predicted amino acid sequence
The ORF of TbPDE1 encodes a protein of 620 amino acids,
with a calculated molecular mass of 70 336 (Fig 3) Since
TbPDE1 was identified by complementation of a PDE
deletion strain of S cerevisiae, its function as a PDE had
already been established Analysis of the predicted amino
acid sequence fully supported this initial assumption The
amino acid sequence unambiguously identified TbPDE1 as
a class I PDE [12], with amino acids His413–Phe424 representing the signature sequence for class I cyclic nucleotide PDE This motif, Pdease_1 [39], displays the consensus sequence HisAsp(LeuIleValMetPheTyr)XHis-X(Ala,Gly)XXAsnX(LeuIleValMetPheTyr) Based on the three-dimensional structures of two isoenzymes of human PDE4 and one of human PDE5 [40–42], the active site is well conserved between these human PDEs and TbPDE1 (Fig 4)
A comparison of the core region of the catalytic domain (Phe347–Phe578) of TbPDE1 with those of other class I PDEs indicates that it is about equidistant from all other class I families, including the dunce gene of Drosophila, the
Fig 2 TbPDE1 mRNA (A) 5¢-Upstream region of TbPDE1 Pol-ypyrimidine-tracts are indicated by blackdots The two alternative trans-splicing acceptor sites are indicated with arrows, and the AG sequences preceding them are underlined The N-terminal part of the ORF is underlined and shown in bold type (B) 3¢-Untranslated region
of TbPDE1 The end of the ORF is underlined and shown in bold type.
A polypyrimidine tract upstream of the poly(A) addition site is indi-cated by blackdots The poly(A) addition site is mark ed with an asterisk The complete DNA sequence of the TbPDE1 locus was submitted to GenBankunder the accession number AF253418 (C) Northern blot hybridized with a TbPDE1-specific riboprobe using
10 lg total RNA per lane LS, Long-slender forms purified from rodent blood; SS, short-stumpy forms; PC, procyclic culture forms derived from short-stumpy forms by in vitro differentiation RNA size mark ers in are indicated on the left (k b).
Trang 6regAgene of Dictyostelium or the trypanosomal TbPDE2
family (Fig 5) The lowest degree of sequence identity was
found with the mammalian PDEs 2 and 6 (< 30% identity),
while the highest degree of sequence identity was found with
the PDEs 1, 3 and 4 (> 40% identity) Using the standard
sequence homology criteria to define PDE families [12],
TbPDE1 clearly represents a new family of the class I of
PDEs The status of a new family is also supported by the
observation that no sequence similarity with other PDEs is
detected outside the catalytic domain, either with
mamma-lian PDEs or with the trypanosomal TbPDE2 family
Outside of the catalytic domain, sequence similarity
decrea-ses, within 10–40 amino acids at the N-terminal side of the
domain, and within 15 amino acids at its C-terminal side
Expression of TbPDE1 inS cerevisiae
The successful complementation screening in yeast indicated
that recombinant TbPDE1 is enzymatically active In
addition to the cDNA plasmid pBa46 (encoding amino
acids Met159–Thr620), the full-length TbPDE1 construct
(pLTHisPDE1) and an N-terminally truncated TbPDE1
construct (pLTBoris, amino acids Arg189–Thr620) also
restored the wild-type phenotype of the yeast mutant (Fig 6A) In contrast, a construct expressing only the core of the catalytic domain (pHisPDEcat1; amino acids Phe347–Phe578) did not (Fig 6A) Very similar results were also obtained in a genetically different PDE-deletion strain
of S cerevisiae, YMS5 ([43]; data not shown)
In addition to conferring a heat-shockresistance pheno-type to the two yeast PDE deletion strains, the introduction
of functional TbPDE1 also significantly changed the phenotype during growth as suspension cultures The
Fig 3 Amino acid sequence of TbPDE1 Arrows indicate the starting
point of various recombinant TbPDE1 polypeptides referred to in the
text: 1, pBa46 (original cDNA clone recovered by complementation
screening); 2, (R189–T620) and pLTBoris; 3,
pET-PDE-(K321–T620) Underlined, (Phe347–Phe578) core of the catalytic
region that was used to calculate amino acid identities between
different PDEs [12] The Gene Bankaccession number of the TbPDE1
polypeptide is AAL580095.
Fig 4 Alignment of the catalytic regions of human PDE4B2B, human PDE5A4 and TbPDE1 A, TbPDE1; B, human PDE4B2B; C, human PDE4D2; D, human PDE5A4 Open bars show the approximate location of alpha-helices Helices predicted from the TbPDE1 sequence correspond reasonably well with those (a4 through a18) found in the three-dimensional structures of the human PDEs 4B2B, 4D2 and 5A4 Blackbars and shaded regions show the signature motifs for class I PDEs Blackdots indicate conserved residues.
Trang 7PDE deletion strain PP5, and its Ura–derivative PP5-12,
exhibit extensive clumping when grown in SC medium
(Fig 6B) When complemented by a heterologous PDE
(TbPDE1 or human PDE4A), clumping was significantly
reduced (Fig 6B) The overall experience with expressing
different fragments of several different PDEs (TbPDE1, the
TbPDE2 family [26], and human PDE4A) suggested that
the extent of clumping of the S cerevisiae PP5 strain
correlates inversely with the extent of recombinant PDE
activity (unpublished results)
Despite the functional complementation observed in
intact yeast cells, no significant PDE catalytic activity could
be detected in yeast cell extracts In contrast, control lysates
from yeast cells that expressed either human PDE4A or
trypanosomal TbPDE2A [26] from the same yeast vector
plasmid pLT1 showed high levels of PDE catalytic activity
To determine if the very low level of TbPDE1 activity might
be due to instability of the recombinant protein, a full-size
TbPDE1 construct was expressed which contained a
haemagglutinin-tag at its N terminus This tagged protein
fully rescued the heat-shockphenotype, was detectable by
immunoblotting and was stable throughout cell breakage
and PDE assay Nevertheless, no enzyme activity could
be detected These observations indicate that TbPDE1 is
expressed in S cerevisiae at levels that are sufficient to
produce a clear phenotype (heat-shockresistance, growth as
a smooth suspension), but that are too low to be detectable
in PDE assays of cell lysates
Expression of TbPDE1 inE coli
Recombinant full-length TbPDE1 was expressed from
plasmid pET-PDE1 and purified from the cytosolic fraction
of E coli cell lysates with high yields after 4 h of expression
at 25C However, the purified protein exhibited only low levels of catalytic activity Consequently, N-terminally truncated variants were designed on the basis of sequence alignments (see above) The C-terminal 250 residues were predicted to comprise the catalytic domain of TbPDE1 The N-terminally truncated construct pET-PDE1-(Arg189– Thr620) was expressed in E coli as an active enzyme which could be purified from the soluble fraction In contrast, the more extensively truncated construct pET-PDE1-(Lys312–
Fig 5 Amino acid sequence identities between the catalytic cores of
class I PDEs The following PDE sequences were used for comparison
(GCG Pileup, using default parameters): 1, human PDE1C (access
number AAC50437); 2, human PDE2A (O00408); 3, human PDE3A
(AAA35912); 4, human PDE4A (AAC35012); 5, human PDE5A
(NM_001074); 6, human PDE6B (NP_000274); 7, human PDE7A
(Q13946); 8, human PDE8A (O60658); 9, human PDE9A
(AAO34689); 10, human PDE10A (AAD32595); 11, human PDE11A
(CAB82573); a, T brucei TbPDE2C (AAK33016); b, D melanogaster
dunce (NP_726859); c, S cerevisiae PDE2 (AAA34846); d, D
dis-coideum regA (AAB03508).
Fig 6 Functional complementation of PDE-deficient S cerevisiae by TbPDE1 (A) Restoration of heat-shockresistance Duplicate patches
of recombinant yeast strains were exposed to a 55 C heat shockfor
15 min and were then grown at 30 C for 2 days pLTHisPDE1, Full length TbPDE1 containing an N-terminal His 6 tag; pLTBoris, amino acids Arg189–Thr620 of TbPDE1, containing an N-terminal His 6 tag; pHisPDEcat1, catalytic core (Phe347–Phe578) containing an N-terminal His 6 tag; pLT1, empty expression vector pLT1; pLC-h6.1, full length human PDE4 (B) Clumping of the PDE-deletion strain PP5-12 and suppression of clumping by the expression of a PDE Yeast cultures were grown for 24 h at 30 C on a rotary wheel and were photographed immediately after removal from the wheel 1, His 6 tagged full-length TbPDE1; 2, pLTBoris; 3, pBa46; 4, pHisPDEcat1;
5, empty expression vector pLT1; 6, pLC-h6.1 (full-length human PDE4A4B) All cultures grew to approximately the same cell density (C) Map of TbPDE1 regions expressed by the various constructs Numbers indicate first and last amino acid expressed by each con-struct Grey box: catalytic core region of TbPDE1.
Trang 8Thr620) produced an inactive protein which was found in
inclusion bodies exclusively This was not unexpected since
this construct most likely lacks a considerable part of the
catalytic domain and thus may be unable to fold correctly
In the initial experiments, the specific activity of
recom-binant TbPDE1 (Arg189–Thr620) was consistently very
low, and the enzyme was highly unstable While the net yield
of soluble enzyme could be considerably improved by
growing the cells in TB medium instead of Luria–Bertani
medium (see Materials and methods), low activity and high
instability remained unsatisfactory Inclusion of 5 mM
Mg2+during cell breakage and in all subsequent
purifica-tion steps greatly stabilized the enzyme and increased its
activity Incubation of the enzyme with the cation-chelator
EDTA led to rapid inactivation (Fig 7) Once the enzyme
was inactivated, removal of the EDTA and the addition of
Mg2+did not restore its activity Gel filtration analysis of
recombinant TbPDE1 demonstrated a marked difference in
migration of the enzyme in the presence or absence of
Mg2+, suggesting that conformational changes were
induced by the cations (data not shown) In addition to
the continuous presence of Mg2+, inclusion of low
concen-trations of detergent [0.1% (v/v) Tween-20] further
activa-ted the enzyme about fourfold and improved the
preservation of activity upon freezing
Kinetic analysis
Enzyme activity was stimulated by either Mg2+or Mn2+
ions, but enzyme preparations inactivated by the prior
removal of Mg2+could not be reactivated by either cation
(see above) Although Mn2+ stimulated the activity
more strongly, 10 mM Mg2+ was used as the cation in
all subsequent experiments The recombinant enzyme
(Arg189–Thr620) displayed standard Michaelis–Menten kinetics, as observed for other PDEs (Fig 8A) An unex-pected finding was the high Kmof TbPDE1 for its substrate cAMP (Fig 8A) An exact Km value was difficult to determine as the assay procedure became unreliable at substrate concentrations beyond 1 mM, and thus did not allow measurement at substrate concentrations far beyond
Km Nevertheless, the combined results of many independ-ent determinations place the Kmfor cAMP at > 600 lM This high Kmfor cAMP is most probably not due to the extraction and purification conditions of the enzyme, as similar values were obtained with enzyme batches prepared
in the presence or absence of Mg2+and detergent Also, the high Kmis unlikely to be an artefact of expression in E coli because the catalytic domains of several human class 1 PDEs have been successfully expressed in the same E coli strain and have exhibited the expected low Kmvalues for their respective substrates [44,45] When reactions were performed in the presence of a 100-fold excess of cGMP, cAMP hydrolysis was not affected (data not shown), indicating that TbPDE1 is cAMP specific, and that its activity is not influenced by allosteric binding of cGMP Inhibitor screening
The potency of a series of known PDE inhibitors against the recombinant enzyme was determined In of all these assays, cAMP concentration was kept at 1 lM, i.e., far below Km,
so that the 50% inhibitory concentrations (IC50) approxi-mate Ki Most of the inhibitors tested showed essentially no effect on the activity of TbPDE1 (Table 1 and Fig 9) The few that exhibited significant potency were sildenafil, a highly specific inhibitor of human PDE5, trequinsin, an inhibitor of human PDE3, ethaverine and dipyridamole However, their IC50values were rather high (1 and 2.5 lM
for sildenafil and trequinsin, respectively) when compared with the IC50 values against their specific targets (human PDE5 for sildenafil, 0.0039 l ; human PDE3 for
trequin-Fig 7 Mg2+dependence of TbPDE1 stability Aliquots of purified
TbPDE1 were incubated for 60 min at 30 C At different times during
this preincubation, EDTA was added to 10 m M final concentration.
After the 60 min preincubation, the enzyme solutions were diluted
1250· into standard reaction buffer (25 m M Tris/HCl pH 7.4, 0.5 m M
EDTA, 0.5 m M EGTA, 10 m M MgCl 2 , 1 l M cAMP), and PDE
activity was determined (A) Preincubation on ice (60 min), no EDTA.
(B) Preincubation at 30 C (60 min), no EDTA (defined as 100%
activity) (C) EDTA added after 60 min preincubation at 30 C,
sample then immediately diluted and assayed (D) EDTA for 20 min.
(E) EDTA for 40 min (F) EDTA for 60 min.
Fig 8 Analysis of TbPDE1 activity Michaelis–Menten kinetics of TbPDE1 indicates a K m of > 600 l M for cAMP Insert: Hanes plot.
Trang 9sin, 0.0003 lM) The inhibitor profile of TbPDE1, including
the four most potent inhibitors, sildenafil, trequinsin,
ethaverine and dipyridamole, is very similar to that
determined for TbPDE2A [26], and it is entirely different
from that of mammalian PDEs No correlation was
observed between the selectivity and potency of inhibitors
for their respective human target PDE family, and their
potency against TbPDE1 Interestingly, the broad-spectrum
PDE inhibitor IBMX, which is widely used in cell biological
experimentation, is essentially inactive on TbPDE1, with an
IC50value of > 1 mM
Discussion
This study reports the identification and characterization of
a novel cyclic-nucleotide-specific PDE, TbPDE1, from
T brucei TbPDE1 is a member of the class I PDEs and
represents a new family within this class The amino acid
sequence of its catalytic domain is approximately
equidis-tant from those of all mammalian class I PDEs, the class I
PDEs dunce of D melanogaster, regA of D discoideum and
PDE2 of S cerevisiae, as well as from another class I PDE
of T brucei, TbPDE2A TbPDE1 is camp specific, its
activity is not modulated by cGMP, and it exhibits an
unusually high Kmfor its substrate cAMP (> 600 lM) It
can functionally complement PDE-deficient yeast strains
and is not inhibited by the broad-spectrum PDE-inhibitor
IBMX, even at high concentrations The latter observation
is reminiscent of the human PDE9, that is similarly resistent
to IBMX Initial functional studies in T brucei have
demonstrated that TbPDE1 is not an essential enzyme,
either for proliferation in culture or for tsetse fly infection by
TbPDE1 knockout trypanosomes [29] This is radically
different from the situation of the TbPDE2 family The
members of the TbPDE2 family are essential Heterozygous
TbPDE2 knockout strains exhibit haploid insufficiency, and
homozygous knockouts are nonviable Downregulation of
TbPDE2 activities either by inhibitors or via RNA
inter-ference leads to a disruption of nuclear division and to rapid cell death ([27] and T Seebeckand M Linder, unpublished results)
Expression of recombinant TbPDE1 proved to depend crucially on the selection of the correct gene fragment A fragment containing the entire catalytic domain (Met159– Thr620) exhibited only minimal activity, which was sufficient for the phenotypic conversion of yeast PDE-deletion mutants in vivo, but was not detectable by in vitro assays in either S cerevisiae or E coli cell lysates A shorter fragment comprising Arg189–Thr620 was also able to convert the phenotype of PDE-deficient yeast, and it could
be expressed as a soluble and catalytically active protein in
E coli A still shorter fragment lacking the N-terminal
Table 1 Potency of selected PDE inhibitors IC 50 values were
deter-mined using 1 l M cAMP as a substrate n.d., Not determined.
Inhibitor
Specificity for human PDE family
IC 50 (l M ) for human PDE family
IC 50 (l M ) for TbPDE1
Trequinsin 3 0.0003 2.5
Ethaverine n.d n.d 8
Dipyridamole 5, 6 0.38 13
Papaverine Nonselective 5–25 30
IBMX Nonselective > 1000
8-Methoxymethyl-IBMX 1 4 > 100
Vinpocetine 1 20 > 100
Milrinone 3 0.3 > 100
Cilostamide 3 0.005 > 100
Zardaverine 3, 4 0.5 > 100
Zaprinast 5, 6 0.5 > 100
Theophylline Nonselective 50–300 > 100
Fig 9 Potency of selected PDE inhibitors against TbPDE1.
Trang 10moiety of the catalytic domain, but still containing its core
part (Lys321–Thr620) was inactive in yeast, and was
expressed in E coli as an inactive polypeptide in the form
of inclusion bodies which could not be refolded into an
active form Stability and activity of the recombinant
enzyme Arg189–Thr620 proved to be extremely sensitive to
the presence of Mg2+ When extracted in the absence of
Mg2+, the resulting protein was poorly active, and activity
could not be restored by the addition of Mg2+ to the
reaction buffer Extraction in the presence of Mg2+ions
produced a very active and stable enzyme preparation,
suggesting that Mg2+is needed not only as a catalyst during
the enzymatic reaction, but also for stabilizing the enzyme
structure This is in agreement with structural workon
human PDE4B2B [41,42] that had demonstrated the
presence of two divalent cations in the active centre In
addition to the continuous presence of Mg2+, inclusion of
low concentrations of detergent, 0.1% (v/v) of Tween-20,
further activated the enzyme about fourfold and improved
the preservation of activity upon freezing Current data do
not allow us to determine if the N-terminal moiety of
TbPDE1 is involved in maintaining the stability of the
enzyme, or if it modulates the activity of the catalytic
domain Similar experiments with another trypanosomal
PDE, TbPDE2A, have demonstrated that the N-terminal
domain exerts no direct effect on either stability or activity
of the catalytic domain [26]
The recombinant TbPDE1 Arg189–Thr620 represents a
cAMP-specific PDE cGMP neither competes as a substrate
nor does it modulate enzyme activity This specificity is in
good agreement with the conservation of Asp378 and
Gln575 that are predicted to confer cAMP-specificity to
human PDE4 [41,42] In contrast to all other members of
the class I PDEs, the Kmof TbPDE1 Arg189–Thr620 for
cAMP is very high (> 600 lM) Available data cannot
formally exclude that this high Kmmight reflect an artefact
of expression in E coli However, this seems unlikely as very
similar fragments of mammalian class I PDEs have been
expressed in the same strain of E coli and were shown to
exhibit the characteristic specificities and low Kmvalues for
their respective substrates [42–44] If indeed genuine, the
high Kmof TbPDE1 for cAMP is particularly remark able as
the intracellular steady-state cAMP concentration in T
bru-cei is in the range of 1–10 lM [24,27] This situation is
reminiscent of PDE1 from S cerevisiae, a high-Kmclass II
PDE In this organism PDE1 appears to play a major role in
the quenching of short-term cAMP peaks upon metabolic
stimulation Deletion of PDE1 in S cerevisiae does not
confer a significant phenotype, but the addition of glucose
to cells grown in glucose-free medium induces cAMP peaks
of much longer duration than seen in wild-type cells [45]
Similarly, TbPDE1, a nonessential enzyme in T brucei [29],
may represent a modulatory element of the cAMP signalling
pathways of T brucei, and its physiological roles remain to
be explored
Acknowledgements
We thankRalph Schwarz, University of Marburg for his cDNA
library, John Colicelli and Peter Engel for their PP5 and YMS5 strains,
respectively, Miles Houslay for providing a human PDE4 clone, Sara
Melville for a gift 927 genomic DNA, and Boris Bieger for providing
some of the constructs We are indebted to Miriam van der Bogaard and Markus Linder for expert technical assistance, and to Min Ku for converting our text into palatable English We gratefully acknowledge the generosity of Glaxo-Wellcome, Smith Kline Beecham and Pfizer for providing samples of PDE inhibitors This workwas supported by grants 31-058927.99 and 3100-067225/1 of the Swiss National Science Foundation, grant C98.0060 of COST program B9 of the European Union, and by the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (to T.S.), the Max-Planck Gesellschaft, grant BEO21/0316211A from the German Federal Ministry for Science and Technology (BMFT), and by grant Bo1100
of the Deutsche Forschungsgemeinschaft (to M.B.).
References
1 Botsford, J.L & Harman, J.G (1992) Cyclic AMP in prokaryotes Microbiol Rev 56, 100–122.
2 D’Souza, C.A & Heitman, J (2001) Conserved cAMP signaling cascades regulate fungal development and virulence FEMS Microbiol Rev 25, 349–364.
3 Meili, R & Firtel, R.A (2003) Follow the leader Dev Cell 4, 291–293.
4 Zufall, F & Munger, S.D (2001) From odor and pheromone transduction to the organization of the sense of smell Trends Neurosci 24, 191–193.
5 Wainger, B.J., DeGennaro, M., Santoro, B., Siegelbaum, S.A & Tibbs, G.R (2001) Molecular mechanism of cAMP modulation of HCN pacemaker channels Nature 411, 805–810.
6 Yu, T.P., McKinney, S., Lester, H.A & Davidson, N (2001) Gamma-aminobutyric acid type A receptors modulate cAMP-mediated long-term potentiation and long-term depression at monosynaptic CA3-CA1 synapses Proc Natl Acad Sci USA 98, 5264–5269.
7 Nikawa, J., Sass, P & Wigler, M (1987) Cloning and character-ization of the low-affinity cyclic AMP phosphodiesterase gene of Saccharomyces cerevisiae Mol Cell Biol 7, 3629–3636.
8 Sass, P., Field, J., Nikawa, J., Toda, T & Wigler, M (1986) Cloning and characterization of the high-affinity cAMP phos-phodiesterase of Saccharomyces cerevisiae Proc Natl Acad Sci USA 83, 9303–9307.
9 Thomason, P.A., Taynor, D., Stock, J.B & Kay, R.R (1999) The RdeA-RegA system, a eukaryotic phospho-relay controllinh cAMP breakdown J Biol Chem 274, 27379–27384.
10 Francis, S.H., Turko, I.V & Corbin, J.D (2001) Cyclic nucleotide phosphodiesterases: relating structure and function Prog Nucleic Acid Res Mol Biol 65, 1–52.
11 Soderling, S.H & Beavo, J.A (2000) Regulation of cAMP and cGMP signaling new phosphodiesterases and new functions Curr Opin Cell Biol 12, 174–179.
12 Beavo, J.A (1995) Cyclic nucleotide phosphodiesterases: func-tional implications of multiple isoforms Physiol Rev 75, 725– 748.
13 Muradov, K.G., Granovsky, A.E & Artemyev, N.O (2003) Mutation in rod PDE6 linked to congenital stationary night blindness impairs the enzyme inhibition by its gamma-subunit Biochemistry 42, 3305–3310.
14 Strettoi, E & Pignatelli, V (2000) Modifications of retinal neurons
in a mouse model of retinitis pigmnentosa Proc Natl Acad Sci USA 97, 11020–11025.
15 Secchiero, P., Zella, D., Curreli, S., Mirandola, P., Capitani, S., Gallo, R.C & Zauli, G (2000) Pivotal role of cyclic nucleotide phosphodiesterase 4 in Tat-mediated CD4+T cell hyperactivation and HIV type I replication Proc Natl Acad Sci USA 97, 14620– 14625.
16 Dean, S.M (2002) Pharmacologic treatment for intermittent claudication Vasc Med 7, 301–309.