Genetic analysis of spermidine synthase from Leishmania donovani

Molecular & Biochemical Parasitology 115 (2001) 217–226

Genetic analysis of spermidine synthase from Leishmaniadono�ani�

Sigrid C. Roberts a, Yuqui Jiang a,1, Armando Jardim a,2, Nicola S. Carter a,Olle Heby b, Buddy Ullman a,*

a Department of Biochemistry and Molecular Biology, Oregon Health Sciences Uni�ersity, Portland, OR 97201-3098, USAb Department of Cell and Molecular Biology, Umea Uni�ersity, SE-901 87 Umea, Sweden

Received 12 January 2001; accepted in revised form 19 April 2001

Abstract

The polyamine biosynthetic pathway of protozoan parasites has been validated as a target in antiparasitic chemotherapy. Toinvestigate this pathway at the biochemical and genetic level in a model parasite, the gene encoding spermidine synthase(SPDSYN), a key polyamine biosynthetic enzyme, has been cloned and sequenced from Leishmania dono�ani. The L. dono�aniSPDSYN gene encodes a polypeptide of 300 amino acids that exhibits 56% amino acid identity with the human counterpart.SPDSYN is present as a single copy gene in the leishmanial genome and encodes a 1.6 kb transcript. Employing SPDSYNflanking sequences to construct drug resistance cassettes, a �spdsyn knockout strain of L. dono�ani was created by double targetedgene replacement. This �spdsyn line could not convert putrescine to spermidine and was auxotrophic for polyamines. Thepolyamine auxotrophy could be circumvented by exogenous spermidine but not by putrescine (1,4-diaminobutane), cadaverine(1,5-diaminopentane), 1,3-diaminopropane, or spermine. Incubation of the null mutant in polyamine-deficient medium resulted ina rapid depletion in the intracellular spermidine level with a concomitant elevation of the putrescine pool. In addition, the levelof trypanothione, a spermidine-containing thiol, was reduced, whereas the glutathione pool increased 3–4-fold. These dataestablish that SPDSYN is an essential enzyme in L. dono�ani promastigotes. The molecular and cellular reagents created in thisinvestigation provide a foundation for subsequent structure-function and inhibitor design studies on this key polyaminebiosynthetic enzyme. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Leishmania dono�ani ; Spermidine synthase; Gene knockouts; Polyamines

www.parasitology-online.com.

Abbre�iations: ADOMETDC, S-adenosylmethionine decarboxy-lase; bp, base pair; dAdoMet, decarboxylated S-adenosylmethionine;DFMO, D,L-�-difluoromethylornithine; DME-L, Dulbecco’smodified Eagle-Leishmania medium; G418, Geneticin; HPLC, highperformance liquid chromatography; HYG, hygromycin; MTA, 5�-methylthioadenosine; NEO, neomycin; ODC, ornithine decarboxy-lase; ORF, open reading frame; PCR, polymerase chain reaction;SPDSYN, spermidine synthase; UTR, untranslated region.

� Note: The nucleotide sequence for the Leishmania dono�ani sper-midine synthase gene has been deposited in GenBank™ under acces-sion number AF298195.

* Corresponding author. Tel.: +1-503-4948437; fax: +1-503-4948393.

E-mail address: [email protected] (B. Ullman).1 Present address: Corixa Corporation, 1124 Columbia St., Seattle,

WA 98104, USA.2 Present address: McGill University, MacDonald Campus, 21,111

Lakeshore Road, Ste. Anne Bellevue, Quebec, Canada, H9X 3V9.

1. Introduction

The polyamine biosynthetic pathway has been suc-cessfully targeted in anti-trypanosomal regimens. D,L-�-difluoromethylornithine (DFMO), an irreversibleinhibitor of ornithine decarboxylase (ODC), the initialand rate limiting enzyme in the polyamine biosyntheticpathway, exhibits efficacy against Trypanosoma bruceiinfections in both mice [1] and patients with late stageAfrican sleeping sickness [2–4]. DFMO is also activeagainst a variety of other protozoan parasites [5–7].Parasites of the genus Leishmania have served as aparticularly useful model to investigate polyaminemetabolism in parasites, as this genus can be grownaxenically as the promastigote (insect vector form) in

0166-6851/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 6 -6851 (01 )00293 -6

S.C. Roberts et al. / Molecular & Biochemical Parasitology 115 (2001) 217–226218

completely defined medium and is amenable to genetic,pharmacologic and biochemical manipulation [7,8].Promastigotes of L. dono�ani, the causative agent ofvisceral leishmaniasis, are sensitive to DFMO and lackan endogenous spermine pool [7]. Genetic and bio-chemical studies on L. dono�ani ODC [8] have revealedthat the entire polyamine biosynthetic pathway in thisparasite consists of three enzymes, ODC, S-adenosyl-methionine decarboxylase (ADOMETDC), and sper-midine synthase (SPDSYN). Furthermore, theseparasites cannot synthesize or utilize spermine[8]. Although parasite ODC and ADOMETDC en-zymes have been investigated in detail at both themolecular and biochemical level [9–14], essentially noinformation on SPDSYN, the enzyme that convertsputrescine and dAdoMet into spermidine and 5�-methylthioadenosine (MTA), is available from any pro-tozoan parasite.

We now report the molecular cloning and geneticanalysis of SPDSYN from L. dono�ani, the firstSPDSYN gene to be isolated and characterized fromany parasite. The construction and characterization of�spdsyn knockouts reveals an essential role of this genein L. dono�ani promastigotes and suggests that theSPDSYN protein may be a promising target for thera-peutic validation.

2. Materials and methods

2.1. Materials, chemicals, and reagents

[2,3-3H]putrescine dihydrochloride (30 Ci mmol−1)was obtained from NEN Life Science Products, Inc.(Boston, MA). Spermidine, putrescine, other diamines,puromycin, MTA and Geneticin (G418) were boughtfrom Sigma Chemical Co. (St Louis, MO). Variousbatches of spermine were purchased from Sigma Chem-ical Co. (St Louis, MO), Aldrich (Milwaukee, WI), andAcros Organics (Acros Organics, Fischer Scientific,Pittsburgh, PA). Hygromycin was procured fromRoche (Roche Diagnostic Cooperation, Indianapolis,IN). Reverse transcriptase, restriction endonucleases,and Tfl polymerase were purchased from either Gibco-BRL (Gaithersburg, MD), New England Biolabs (Bev-erly, MA), or Epicentre Technologies (Madison, WI).The pX63-NEO, pX63-HYG, and pX63-PAC plasmidsused for cloning and parasite transfections were pro-vided by Dr Stephen M. Beverley (Washington Univer-sity, St. Louis, MO). A directionally clonedUni-ZapXR L. dono�ani cDNA library was obtainedfrom Drs. Peter Myler and Kenneth Stuart (SeattleBiomedical Research Institute, Seattle, WA). All othermaterials, chemicals, and reagents used in these experi-ments have been reported previously.

2.2. Cell lines and cell culture

The wild type L. dono�ani clone, DI700, originallyderived from the 1S Sudanese strain, was used forDNA isolation, library construction, and as a recipientstrain in the initial transfections. L. dono�ani pro-mastigotes were propagated routinely in DME-L, acompletely defined Dulbecco’s modified Eagle-basedmedium especially designed for Leishmania cultivation[15]. Transfected parasites and parasites used in growthexperiments were maintained in a modified medium,DME-L-CS, in which the bovine serum albumin com-ponent was replaced with 10% chicken serum to avoidpolyamine oxidase-mediated toxicity [7]. TheSPDSYN/spdsyn heterozygote and the �spdsyn nullmutants were routinely maintained in DME-L-CS sup-plemented with either 50 �g ml−1 hygromycin or 50 �gml−1 hygromycin, 20 �g ml−1 G418, and 100 �Mspermidine, respectively, unless otherwise specified.Growth rate experiments were conducted and parasitesenumerated on a Coulter Counter Model ZF as de-scribed [7,8].

2.3. DNA manipulations and library construction

L. dono�ani genomic DNA was isolated by standardprotocols [16,17]. The construction of the L. dono�anigenomic library in the Supercos 1 cosmid vector(Stratagene, La Jolla, CA) has been previously reported[17]. Southern blot analysis was performed as described[16,17] except that DNA was labeled with fluoresceinusing the Renaissance Random Primer Fluorescein La-beling Kit (NEN, Life Science Products, Boston, MA).Membranes were developed using the Renaissance En-hanced Luminol Nucleic Acid ChemiluminescenceReagent (NEN, Life Science Products, Boston, MA).Total RNA from L. dono�ani promastigotes was pre-pared and cDNA was generated using random hexameroligonucleotides and the Superscript reverse transcrip-tion kit (Life Technologies, Inc.).

2.4. Isolation of SPDSYN

A portion of the L. dono�ani SPDSYN was amplifiedfrom cDNA using the polymerase chain reaction(PCR). The nondegenerate sense primer correspondedto the L. dono�ani mini-exon gene sequence [17] andcontained a BamHI restriction site (underlined), 5�-CTCGGGATCCCAACGCTATATAAGTATCAGTT-TCTGTACTTTATTG-3�. The degenerate antisenseprimer contained an EcoRI site (underlined), 5�-TCT-GAATTCCC(A/G)TCNCCNCCNCC(A/T/G)AT(A/T/G)AT-3�, and was synthesized from the hepta-peptide, IIGGGDG, that is conserved among human,Saccharomyces cere�isiae, and Escherichia coliSPDSYNs [18–20]. The SPDSYN fragment was then

S.C. Roberts et al. / Molecular & Biochemical Parasitology 115 (2001) 217–226 219

amplified from L. dono�ani cDNA on a PTC-100 ther-mocycler (MJ Research, Watertown, MA) after 30 cy-cles of 94°C for 40 s, 55°C for 1 min, and 72°C for 1min followed by a 5 min final extension reaction at72°C. The approximately 500 bp PCR product wassequenced on a Perkin Elmer Applied Biosystems 377DNA sequencer to confirm that it was a SPDSYNfragment. The 500 bp fragment was then hybridized tothe L. dono�ani cDNA library, and a cDNA clone wasplaque purified. The cDNA insert was subsequentlyused as a probe to screen the cosmid library, andpositive cosmids were purified. All library screeningswere performed under high stringency conditions[16,17]. Southern blotting of cosmid DNA resulted inthe isolation of a 4.5 kb BamHI-HindIII fragment thatencompassed the entire SPDSYN open reading frame(ORF). This fragment was subcloned into pBluescriptKS+ and the ORF sequenced in both directions.Multiple sequence alignments were performed using theFeng-Doolittle algorithm [21].

2.5. Molecular constructs for the replacement of theSPDSYN alleles

To construct drug resistance cassettes for SPDSYNreplacement, 1.5 kb 5� (5�F) and 1.6 kb 3� (3�F) untrans-lated regions (UTRs) of SPDSYN were ligated into theHindIII-SalI and SmaI-BglII sites, respectively, of boththe pX63-HYG and pX63-NEO vectors. The 5�F and3�F flanking fragments were identified by Southernblotting and limited nucleotide sequencing of the 4.5 kbBamHI-HindIII fragment. The 5�F was amplified byPCR from the 4.5 kb BamHI-HindIII template usingthe T7 pBluescript primer that encompasses a HindIIIsite as the sense primer and an oligonucleotide, 5�-CTT-GTCATGTCGACCAGCTCACATCGCTC-3�, com-plementary to a sequence in the 5� UTR and containinga SalI site (underlined) as the antisense primer. ThePCR consisted of 30 cycles of 95°C for 1 min, 50°C for1 min, 72°C for 2 min, followed by final extension at72°C for 10 min. The pX63-HYG and pX63-NEOvectors containing 5�F were designated pX63-HYG-5�Fand pX63-NEO-5�F, respectively.

The 3�F was excised from the 4.5 kb BamHI-HindIIISPDSYN insert with XhoI and HindIII, the endsblunted with Mung Bean nuclease, and the fragmentsubcloned into the SmaI site of pX63-HYG-5�F tocreate the knockout construct pX63-HYG-�spdsyn. A3.0 kb BamHI-BglII fragment containing 3�F and 1.3kb of the vector (downstream from the HYG marker)was excised from pX63-HYG-�spdsyn and subclonedinto the BglII site of pX63-NEO-5�F to create thesecond knockout construct pX63-NEO-�spdsyn. Thecorrect orientation of the 3� flanking region in pX63-NEO-�spdsyn was then verified by restriction enzymedigestion and gel electrophoresis.

2.6. Molecular construct for generating SPDSYNo�erexpressors

The SPDSYN ORF, 1 kb of 5� flanking region, and2.6 kb of 3� flanking region were amplified togetherfrom the 4.5 kb BamHI-HindIII fragment by PCRusing the following conditions: 30 cycles of 94°C for 1min, 65°C for 3 min, and 72°C for 2 min followed by a5 min final extension reaction at 72°C. Sense (5�-AC-TAGTGGATCCCCCGCTGGC-3�) and antisense (5�-ACTAGTGGATCCCCCGCTGGC-3�) primerscontained BamHI restriction sites (underlined), and the4.5 kb PCR product was cloned into the BamHI site ofthe pX63-PAC vector to create pX63-PAC-SPDSYN.

2.7. Transfections

Parasites were transfected by electroporation usingconditions described previously [8,22]. The firstSPDSYN allele was targeted in DI700 wild type pro-mastigotes with HYG-�spdsyn and clonal transfectantsisolated on DME-L plates containing 50 �g ml−1 hy-gromycin. A SPDSYN/spdsyn heterozygote clone wasthen transfected with NEO-�spdsyn and selected inDME-L-CS supplemented with100 �M spermidine, 50�g ml−1 hygromycin and 20 �g ml−1 G418. The geno-type of �spdsyn clones was established by Southernblotting. The recommended genetic nomenclature forthe SPDSYN/spdsyn and �spdsyn strains is�spdsyn ::HYG and �spdsyn ::HYG/�spdsyn ::NEO, re-spectively [23].

To generate SPDSYN overexpressors, �spdsyn cellswere transfected with the pX63-PAC-SPDSYN con-struct. A bulk culture was then expanded in DME-L-CS containing 50 �g ml−1 hygromycin, 20 �g ml−1

G418, and 100 �M puromycin.

2.8. Antibody generation and immunoblotting

The L. dono�ani SPDSYN was overexpressed in E.coli and recombinant SPDSYN purified to homogeneity(data to be presented elsewhere). Polyclonal antiseraagainst SPDSYN were generated in rabbits by CocalicoBiologicals Inc. (Reamstown, PA) using purified recom-binant SPDSYN as immunogen and standard injectionprotocols. Immunoblots were performed according tostandard protocols [24,25].

2.9. Putrescine incorporation experiments

The 5×106 wild type or �spdsyn parasites wereincubated with 50 �Ci [3H]putrescine (30 Ci mmol−1)in 5.0 ml of DME-L-CS for 4 days. Polyamines werederivatized and analyzed as described [8]. A total of 1ml fractions were collected and radioactivity quanti-tated by liquid scintillation spectrometry. The identity


of the radiolabeled polyamines was substantiated byco-injection with known polyamine standards.

2.10. Polyamine, dAdoMet, and thiol pool analysis

Parasites at 1×107 were harvested and sonicated in100 �l of ice-cold 0.2 M perchloric acid and the acidsoluble extract analyzed for putrescine, spermidine,spermine, and dAdoMet content by reversed-phasehigh performance liquid chromatography (HPLC) us-ing 1-octanesulfonic acid as an ion pairing agent on aKromasil KR 100-5C18 (Eka Nobel; 15 cm×4.6 mminside diameter) column [26]. dAdoMet was detected bypassage of the HPLC eluate through a UV detector setat 254 nm. Subsequently, the eluate was mixed witho-phthalaldehyde in a reaction coil and passed througha fluorescence detector for polyamine analysis. Parasitesat 1×108 were harvested for each thiol pool analysisand thiol pools determined by HPLC as described [8].

3. Results

3.1. Molecular cloning and sequence analysis ofSPDSYN

The gene encoding the L. dono�ani SPDSYN enzymewas isolated from a 4.5 kb BamHI-HindIII fragmentusing a 500 bp PCR-amplified SPDSYN sequence as ahybridization probe. Nucleotide sequence analysis iden-tified 900 bp ORF predicting a polypeptide of 300amino acids with a molecular mass of 33 kDa and acalculated pI of 5.5. An in-frame termination codonwas located 55–57 nucleotides upstream from the as-signed initiation codon. Pairwise alignments indicatedthat the L. dono�ani SPDSYN exhibited 56, 55, and45% identity with the human [18], S. cere�isiae [19], andE. coli [20] enzymes, respectively (Fig. 1). Areas ofamino acid sequence identity among the alignedSPDSYNs were interspersed throughout the protein

Fig. 1. Multisequence alignment of phylogenetically diverse SPDSYN proteins. SPDSYN proteins were aligned according to the Feng-Doolittlealgorithm [21]. SPDSYN sequences shown include those from L. dono�ani (LdSPDSYN), human (HsSPDSYN) [18], S. cere�isiae (ScSPDSYN)[19], and E. coli (EcSPDSYN) [20]. Identical residues are highlighted in dark boxes, and similar residues are shown in shaded boxes.


with the longest contiguous identical stretch corre-sponding to the antisense degenerate primer used in thePCR amplification of the SPDSYN hybridization probe(amino acids 93–99). Southern blot analysis using L.dono�ani genomic DNA indicated that SPDSYN waspresent as a single copy gene within the L. dono�anigenome, and Northern blotting revealed a single 1.6 kbSPDSYN transcript (data not shown).

3.2. Genetic analysis of SPDSYN

To disrupt the SPDSYN locus each gene copyin wild type L. dono�ani was sequentially replaced witha drug resistance marker. The first SPDSYN copy wasdisplaced with HYG-�spdsyn, and SPDSYN/spdsyn heterozygotes were selected in DME-L supple-mented with 50 �g ml−1 hygromycin. A second roundof transfection with NEO-�spdsyn replaced theother SPDSYN copy, and clonal lines were selected inmedia containing 50 �g ml−1 hygromycin, 20 �g ml−1

G418, and 100 �M spermidine. Hygromycin was addedto the G418 selections to ensure that the second roundof gene targeting yielded colonies in which NEO-�spdsyn had supplanted the wild type allele in theSPDSYN/spdsyn heterozygote. Spermidine was addedto the selective medium to circumvent potentialpolyamine auxotrophy. A total of 16 colonies wereisolated from the second round of transfection, and twoclones, �spdsyn3 and �spdsyn6, were chosen for furtherstudy.

Southern blot analysis of genomic DNA from wildtype, SPDSYN/spdsyn, and �spdsyn3 and �spdsyn6parasites established that each of the two wild typeSPDSYN copies had been sequentially replaced byhomologous recombination of the drug resistance cas-settes (Fig. 2). The novel alleles could be effectivelydiscriminated from the wild type allele by distinct dif-ferences in the locations of XhoI and BglI sites. XhoI-digested DNA probed with a 1.6 kb SacI fragmentfrom the 5� UTR (see Fig. 2) revealed a novel 5 kbband in the SPDSYN/spdsyn heterozygote and thedisappearance of the wild type 7 kb fragment in the�spdsyn null mutants. The novel alleles could also beobserved after digestion with BglI and hybridizing witha 1.2 kb XhoI-HindIII fragment from the 3� UTR (seeFig. 2). The 9 and 11 kb BglI fragments were obtainedin the SPDSYN/spdsyn and �spdsyn lines as a conse-quence of the novel alleles created by homologousintegration of HYG-�spdsyn and NEO-�spdsyn, re-spectively, whereas the 1.8 kb BglI fragment from thewild type locus was missing in the null mutants (Fig. 2).The loss of SPDSYN coding sequences in the �spdsynknockouts was confirmed by probing with theSPDSYN ORF (Fig. 2). Restriction maps of the wildtype and mutant SPDSYN alleles are compared in Fig.2.

Fig. 2. Southern blot analysis of wild type and mutant parasites.Panel A: 2 �g of genomic DNA from wild type, SPDSYN/spdsyn,�spdsyn3, and �spdsyn6 parasites was digested with the indicatedrestriction enzymes and hybridized to SPDSYN probes from thecoding region (left panel), the 5� flank (middle panel), or the 3� flank(right panel). Molecular weight markers are indicated to the left.Panel B: restriction maps of the wild type and rearranged SPDSYNloci are depicted. A black box shows the SPDSYN ORF probe,hatched boxes indicate both flanking regions, a gray box depicts thelocus with the HYG marker integration, and the dotted box illus-trates the NEO marker insertion.

3.3. SPDSYN expression and acti�ity in parasites

Immunoblots of wild type and �spdsyn parasitesdemonstrated the lack of SPDSYN protein in theknockouts (Fig. 3). The same blot was reprobed withan antiserum against hypoxanthine-guanine phosphori-bosyltransferase to demonstrate that equivalentamounts of extract were loaded onto each lane of thegel (Fig. 3).

To prove that intact �spdsyn knockout parasiteslacked functional SPDSYN activity in vivo, the abilities


of wild type and �spdsyn3 promastigotes to convertradiolabeled putrescine to spermidine were compared(Fig. 4). These experiments demonstrated that wild typeparasites could synthesize spermidine from putrescine,whereas the �spdsyn3 cells were incapable of thismetabolic transformation (Fig. 4).

3.4. Nutritional requirements of �spdsyn parasites

The nutritional requirements of the �spdsyn parasiteswere evaluated after the genotypes were established. Asshown in Fig. 5A, wild type and SPDSYN/spdsynparasites grown in defined DME-L-CS withoutpolyamine supplementation grew normally, whereas the�spdsyn3 and �spdsyn6 cells did not. The null mutantsincubated under these conditions arrested after severalcell divisions, assumed a rounded morphology, andeventually died (Fig. 5A and see below). The additionof 100 �M spermidine to the culture medium bypassedthe lethality conferred by SPDSYN loss (Fig. 5B,closed diamonds). Indeed, spermidine addition rescued

�spdsyn cells starved for polyamines for 7 days (Fig.5C) and even up to 3 weeks (data not shown). After 3weeks of polyamine starvation, however, spermidinesupplementation did not revive the culture, indicatingthat the entire population was dead. The growth rate of�spdsyn parasites in the presence of spermidine waseffectively indistinguishable from that of the wild typeand heterozygotes in unsupplemented medium (Fig. 5Aand B). As expected, spermidine did not affect thegrowth rate of either wild type or SPDSYN/spdsynparasites grown in DME-L-CS (data not shown). Vari-ation of the exogenous spermidine level indicated that10 �M spermidine was sufficient to enable �spdsyn cellsto reach their maximum density (Fig. 5D). No otherdiamine, including putrescine (1,4-diaminobutane), 1,3-diaminopropane, or cadaverine (1,5-diaminopentane)could bypass the polyamine auxotrophy conferred byreplacement of the SPDSYN locus (Fig. 5B). In addi-tion, neither 100 �M 5�-methylthioadenosine, the otherproduct of the SPDSYN reaction, nor spermine en-abled growth of the �spdsyn promastigotes (data notshown).

3.5. Genetic rescue of �spdsyn mutants

To establish genetically that SPDSYN replacementconferred spermidine auxotrophy, an episomal copy ofthe SPDSYN gene was re-introduced into the �spdsynknockout. The episomal SPDSYN construct includedboth 5� and 3� flanking regions to provide sequences forcorrect processing of the SPDSYN transcript. Im-munoblot analysis showed that these transfectants ex-pressed SPDSYN protein (Fig. 6A). Moreover,episomal SPDSYN conferred polyamine prototrophyupon the �spdsyn lines (Fig. 6B).

3.6. Polyamine, thiol, and dAdoMet pool measurements

The consequences of polyamine starvation onpolyamine, reduced thiol, and dAdoMet pools in�spdsyn parasites were also evaluated. After removalfrom spermidine, intracellular spermidine levels in�spdsyn cells dropped precipitously (Fig. 7A). After 4days of polyamine starvation, the spermidine pool was�15% of the initial level of the polyamine supple-mented culture. Conversely, the putrescine pool in-creased 3–4-fold in the absence of polyaminesupplementation (Fig. 7A). The putrescine:spermidineratio increased steadily throughout the incubation froman initial ratio of 1.24 to a final ratio of 38 after 8 daysin the absence of exogenous spermidine. Concomitantwith the spermidine depletion was a diminution in thelevels of reduced trypanothione, the spermidine-con-taining thiol (Fig. 7B). Indeed, trypanothione levelswere undetectable in �spdsyn cells incubated in theabsence of spermidine for 8 days. Conversely, levels of

Fig. 3. Western blot analysis of expressed SPDSYN protein in wildtype and genetically altered L. dono�ani. Polyclonal antiserum againstL. dono�ani SPDSYN was generated in rabbits and used to detectSPDSYN protein in cell lysates prepared from 107 wild type,�spdsyn3, and �spdsyn6 parasites. The same blot was reprobed withan antibody against the L. dono�ani hypoxanthine-guanine phospho-ribosyltransferase protein in order to normalize the loading onto eachlane of the gel.


Fig. 4. Spermidine synthesis in intact wild type and �spdsyn parasites. Wild type (�) and �spdsyn (�) parasites were incubated with[3H]putrescine for 4 days, harvested, and polyamines were processed as described under Section 2. The derivatized polyamine samples werefractionated by HPLC, and 1 ml fractions were analyzed for radioactivity by liquid scintillation counting. The relative migration of knownputrescine and spermidine standards is indicated.

reduced glutathione, a trypanothione precursor, in-creased 3–4-fold in the �spdsyn parasites as a result ofextended incubation in the absence of spermidine. Thecellular concentrations of dAdoMet, the otherSPDSYN substrate, did not vary in �spdsyn cells incu-bated without spermidine (data not shown). Interest-ingly, dAdoMet was not detectable in wild type cellsindicating that the dAdoMet pool is dramatically ele-vated in the knockout line as a consequence of thegenetic lesion.

4. Discussion

The creation of a �spdsyn strain by double targetedgene replacement established that SPDSYN is an essen-tial gene in L. dono�ani. The �spdsyn promastigoteswere proven to be polyamine auxotrophs, and thispolyamine auxotrophy could be circumvented by am-plification and expression of an episomal SPDSYNconstruct or by supplementation of the culture mediumwith spermidine. Neither putrescine, 1,3-di-aminopropane, cadaverine, 5�-methylthioadenosine norspermine could restore the ability of �spdsyn cells toproliferate in DME-L-CS. The failure of these com-pounds to rescue �spdsyn parasites could not be as-cribed to their inherent toxicity as wild type parasiteswere unaffected by these agents. Previous genetic stud-ies with �odc L. dono�ani that were also generated bysequential gene displacement demonstrated that pu-trescine, 1,3-diaminopropane, cadaverine, and sper-midine but not spermine could overcome the geneticallyinduced polyamine auxotrophy [8]. That 1,3-di-aminopropane and cadaverine fail to fulfill thepolyamine requirements of the �spdsyn knockout im-

plies that the ability of these diamines to rescue the�odc mutant is dependent upon their conversion totheir spermidine equivalents, a process that is catalyzedby SPDSYN. Taken together, the growth phenotypesof the �odc and the �spdsyn parasites establish geneti-cally that spermidine is both essential and sufficient tosupport the polyamine requirements of L. dono�anipromastigotes.

Incubation of �spdsyn promastigotes in mediumlacking spermidine resulted in a marked reduction inspermidine and trypanothione pools with an accompa-nying increase in the cellular levels of putrescine andglutathione. In contrast, polyamine starvation of �odcL. dono�ani or DFMO treatment of wild type parasitesresults in a profound depletion of cellular putrescineconcentrations without particularly affecting sper-midine levels [7,8]. The decrease in spermidine in the�spdsyn parasites can be ascribed to their inability toconvert putrescine to spermidine (Fig. 4), whereas thediminution of trypanothione is presumably a down-stream consequence of the genetic lesion and the conse-quent lack of spermidine availability for trypanothionesynthesis. Whether the cessation of �spdsyn cell growthobserved in unsupplemented growth medium can beattributed specifically to the spermidine or to the try-panothione pool depletion cannot be discerned by thesedata. Spermidine depletion does not provoke an abruptgrowth failure, as �spdsyn cells undergo 2–3 cell divi-sions after removal of spermidine from the medium.

The accumulation of both putrescine and glutathioneunder polyamine starvation conditions could be theresult of cellular overproduction to compensate for thegenetically induced spermidine and trypanothione defi-ciencies. In support of this hypothesis is the observationthat �spdsyn parasites supplemented with spermidine


exhibit putrescine and glutathione levels that are equiv-alent to those of wild type cells. However, the increasesin putrescine and glutathione obviously do not ulti-mately compensate for the loss of spermidine and/ortrypanothione. Although it has been shown previouslythat L. dono�ani lack a polyamine catabolic pathway[8], and therefore the putrescine accumulation cannotbe due to back conversion from spermidine, it remainsa formal possibility that trypanothione cleavage cancontribute to the augmented glutathione pools. The factthat the dAdoMet pool in �spdsyn parasites remainsunvarying under conditions of spermidine starvation(data not shown) also supports the contention thatputrescine is overproduced by the polyamine-starvedparasites. These data raise the possibility that the regu-latory controls on the activities of ODC andADOMETDC, the enzymes that generate putrescineand dAdoMet, are disparate.

A multisequence alignment of SPDSYN proteinsfrom phylogenetically diverse organisms revealed sev-eral regions of significant homology. The most strik-ingly conserved region, the VLIIGGGDGGcorresponding to amino acids 90–100 of the L. dono-�ani SPDSYN, has been implicated in alternative mod-els as a binding motif for either putrescine [27] ordAdoMet [28]. A similar sequence to the SPDSYNdecapeptide is also found among spermine synthaseproteins, implicating this sequence in the binding ofdAdoMet, a common substrate for both polyaminesynthases [18,28]. There are also several other aminoacids in this region that are conserved among sper-midine and spermine synthases, although there are noother regions of significant amino acid homology be-tween these two enzymes. There are several other areasof limited amino acid sequence identity but unknownfunction that are found among the aligned SPDSYN

Fig. 5. Growth of wild type and mutant parasites in DME-L-CS medium in the absence and presence of diamine and polyamine supplements.Panel A: wild type (�), SPDSYN/spdsyn (�), and �spdsyn3 (�), and �spdsyn6 (�). All growth experiments with L. dono�ani were conductedin DME-L-CS medium, and parasites enumerated every 24 h. Panel B: the �spdsyn3 parasites were removed from DME-L-CS medium that hadbeen supplemented with 100 �M spermidine, washed twice in phosphate buffered saline, and then incubated in medium without polyamineaddition (�) or supplemented with either 100 �M spermidine (�), putrescine (�), cadaverine (�), or 1,3-diaminopropane (�). Panel C: The�spdsyn3 parasites were incubated for 7 days in polyamine-free media prior to resuspension in DME-L-CS without polyamine (�) or DME-L-CSto which either 100 �M spermidine (�), putrescine (�), cadaverine (�), or 1,3-diaminopropane (�) was added. Panel D: the �spdsyn3 (�), and�spdsyn6 (�) parasites were incubated in DME-L-CS supplemented with various concentrations of spermidine as indicated. Parasites werecounted on day 7.


Fig. 6. Genetic rescue of �spdsyn parasites. Panel A: Western blotanalysis of wild type, �spdsyn3, and �spdsyn3 transfected withpX63PAC-SPDSYN (transfectants 1and 3). Fractionated lysates wereprobed with polyclonal antisera against SPDSYN. Panel B: theability of wild type (�), �spdsyn3 (�), and �spdsyn3 harboringpX63PAC-SPDSYN, transfectant 1 (�) and transfectant 3 (�), toproliferate in DME-L-CS without spermidine addition was assessed.

Fig. 7. Polyamine and thiol pool measurements in �spdsyn cells.�spdsyn3 parasites were incubated in DME-L-CS without polyamineaddition, and aliquots were analyzed every 2 days for: Panel A:putrescine (�) and spermidine (�); and Panel B: glutathione (�)and trypanothione (�).

Acknowledgements

This work was supported by grants AI10096 (S.R.)and AI41622 (B.U.) from the National Institute ofAllergy and Infectious Disease and by the SwedishNatural Science Research Council (O.H.). Buddy Ull-man is a Burroughs Wellcome Fund Scholar in Molec-ular Parasitology, and this work was supported in partby a grant from The Burroughs Wellcome Fund. Wewould like to thank Francy Pesek-Jardim and BirgittaGrahn for their excellent help with the cell culture andHPLC analyses, respectively.

References

[1] Bacchi CJ, Nathan HC, Hutner SH, McCann PP, Sjoerdsma A.Polyamine metabolism: a potential therapeutic target in try-panosomes. Science 1980;210:332–4.

[2] Bacchi CJ, McCann PP. Parasitic protozoa and polyamines. In:McCann PP, Pegg AE, Sjoerdsma A, editors. Inhibition ofPolyamine Metabolism: Biological Significance and Basis forNew Therapies. Orlando: Academic Press, 1987:317–44.

[3] Schechter PJ, Barlow JLR, Sjoerdsma A. Clinical aspects ofinhibition of ornithine decarboxylase with empahsis on therapeu-tic trials of eflornithine (DFMO) in cancer and protozoan dis-eases. In: McCann PP, Pegg AE, Sjoerdsma A, editors.Inhibition of Polyamine Metabolism: Biological Significance andBasis for New Therapies. Orlando: Academic Press, 1987:345–64.

enzymes. One of these, the negatively charged TE/DR/YDEF hexapeptide corresponding to amino acids 66–71 of the L. dono�ani protein, has been proposed as abinding site for the protonated amino group of pu-trescine [27].

Overall, this genetic dissection demonstrates thatSPDSYN is an essential gene in L. dono�ani promastig-otes and that spermidine is the crucial polyamine forparasite proliferation. Since Leishmania promastigotesare capable of transporting polyamines [29,30], it isimportant to determine whether the mammalianamastigote stage of the parasite is also capable oftransporting polyamines and whether specificpolyamines have access to the phagolysosomal com-partment of the macrophage where the amastigoteresides. Thus, the SPDSYN and ODC genes and theircorresponding knockout lines that are polyamine aux-otrophs provide key molecular and cellular reagentsthat will serve as cornerstones for infectivity studies inmammalian cells.


[4] World Health Organization. Tenth Programme Report. WorldHealth Organization Geneva, Switzerland, 1987. pp. 59–68.

[5] Bitonti AJ, McCann PP, Sjoerdsma A. Plasmodium falciparumand Plasmodium berghei : effects of ornithine decarboxylase in-hibitors on erythrocytic schizogony. Exp Parasitol 1987;64:237–43.

[6] Gillin FD, Reiner DS, McCann PP. Inhibition of growth ofGiardia lamblia by difluoromethylornithine, a specific inhibitorof polyamine biosynthesis. J Protozool 1984;31:161–3.

[7] Kaur K, Emmett K, McCann PP, Sjoerdsma A, Ullman B.Effects of Dl-alpha difluoromethylornithine on Leishmania dono-�ani promastigotes. J Protozool 1986;33:518–21.

[8] Jiang Y, Roberts SC, Jardim A, Carter NS, Shih S,Ariyanayagam M, Fairlamb AH, Ullman B. Ornithine decar-boxylase gene deletion mutants of Leishmania dono�ani. J BiolChem 1999;274:3781–8.

[9] Phillips MA, Coffino P, Wang CC. Trypanosoma brucei or-nithine decarboxylase: enzyme purification, characterization, andexpression in Escherichia coli. J Biol Chem 1988;263:17933–41.

[10] Coons T, Hanson S, Bitonti AJ, McCann PP, Ullman B. Alpha-difluoromethylornithine resistance in Leishmania dono�ani is as-sociated with increased ornithine decarboxylase activity. MolBiochem Parasitol 1990;39:77–89.

[11] Rathaur S, Wittich RM, Walter RD. Ascaris suum andOnchocerca �ol�ulus : S-adenosylmethionine decarboxylase. ExpParasitol 1988;65:277–81.

[12] Tekwani BL, Bacchi CJ, Pegg AE. Putrescine activated S-adeno-sylmethionine decarboxylase from Trypanosoma brucei brucei.Mol Cell Biochem 1992;117:53–61.

[13] Kinch LN, Scott JR, Ullman B, Phillips MA. Cloning andkinetic characterization of the Trypanosoma cruzi S-adenosylme-thionine decarboxylase. Mol Biochem Parasitol 1999;101:1–11.

[14] Muller S, Da’dara A, Luersen K, Wrenger C, Das Gupta R,Madhubala R, Walter RD. In the human malaria parasitePlasmodium falciparum, polyamines are synthesized by a bifunc-tional ornithine decarboxylase, S-adenosylmethionine decar-boxylase. J Biol Chem 2000;275:8097–102.

[15] Iovannisci DM, Ullman B. High efficiency plating method forLeishmania promastigotes in semidefined or completely-definedmedium. J Parasitol 1983;69:633–6.

[16] Hanson S, Adelman J, Ullman B. Amplification and molecularcloning of the ornithine decarboxylase gene of Leishmania dono-�ani. J Biol Chem 1992;267:2350–9.

[17] Wilson K, Collart FR, Huberman E, Stringer JR, Ullman B.Amplification and molecular cloning of the IMP dehydrogenasegene of Leishmania dono�ani. J Biol Chem 1991;266:1665–71.

[18] Myohanen S, Kauppinen L, Wahlfors J, Alhonen L, Janne J.Human spermidine synthase gene: structure and chromosomallocalization. DNA Cell Biol 1991;10:467–74.

[19] Hamasaki-Katagiri N, Tabor CW, Tabor H. Spermidine biosyn-thesis in Saccharomyces cere�isae : polyamine requirement of anull mutant of the Spe3 gene (spermidine synthase). Gene1997;187:35–43.

[20] Xie QW, Tabor CW, Tabor H. Spermidine biosynthesis inEscherichia coli : promoter and termination regions of the speedoperon. J Bacteriol 1989;171:4457–65.

[21] Feng DF, Doolittle RF. Progressive sequence alignment as aprerequisite to correct phylogenetic trees. J Mol Evol1987;25:351–60.

[22] LeBowitz JH, Coburn CM, McMahon-Pratt D, Beverley SM.Development of a stable Leishmania expression vector and appli-cation to the study of parasite surface antigen genes. Proc NatlAcad Sci USA 1990;87:9736–40.

[23] Clayton C, Adams M, Almeida R, et al. Genetic nomenclaturefor Trypanosoma and Leishmania. Mol Biochem Parasitol1998;97:221–4.

[24] Laemmli UK. Cleavage of structural proteins during the assem-bly of the head of bacteriophage T4. Nature 1970;227:680–5.

[25] Sambrook J, Fritsch EF, Manniatis T. Molecular Cloning. ALaboratory Manual, 2nd ed. Cold Spring Harbor, NY: ColdSpring Harbor Laboratory, 1989.

[26] Seiler N, Knodgen B. Determination of amino acids by separa-tion of their ion pairs with dodecyl sulphate. J Chromatogr1985;341:11–21.

[27] Shirahata A, Morohohi T, Fukai M, Akatsu S, Samejima K.Putrescine or spermidine binding site of aminopropyltransferasesand competitive inhibitors. Biochem Pharmacol 1991;41:205–12.

[28] Korhonen VP, Halmekyto M, Kauppinen L, Myohanen S,Wahlfors J, Keinanen T, Hyvonen T, Alhonen L, Eloranta T,Janne J. Molecular cloning of a cDNA encoding human sper-mine synthase. DNA Cell Biol 1995;14:841–7.

[29] Kandpal M, Tekwani BL. Polyamine transport systems of Leish-mania dono�ani promastigotes. Life Sci 1997;60:1793–801.

[30] Basselin M, Coombs GH, Barrett MP. Putrescine and sper-midine transport in Leishmania. Mol Biochem Parasitol2000;109:37–46.

.

Genetic analysis of spermidine synthase from Leishmania donovani

Documents

Transcript of Genetic analysis of spermidine synthase from Leishmania donovani