The FASEB Journal • Research Communication

Retooling Leishmania metabolism: from sand fly gut tohuman macrophage

Doron Rosenzweig,* Derek Smith,‡ Fred Opperdoes,§ Shay Stern,† Robert W. Olafson†,and Dan Zilberstein*1

*Faculty of Biology and †Faculty of Chemical Engineering, Technion—Israel Institute of Technology,Haifa, Israel; ‡UVic Proteome Center, University of Victoria, British Columbia, Canada; and§Research Unit for Tropical Diseases, Christian de Duve Institute of Cellular Pathology and CatholicUniversity of Louvain, Brussels, Belgium

ABSTRACT To survive extremely different environ-ments, intracellular parasites require highly adaptablephysiological and metabolic systems. Leishmania dono-vani extracellular promastigotes reside in a glucose-rich, slightly alkaline environment in the sand fly vectoralimentary tract. On entry into human macrophagephagolysosomes, promastigotes differentiate into intra-cellular amastigotes. These cope with an acidic milieu,where glucose is scarce while amino acids are abundant.Here, we use an axenic differentiation model and anovel high-coverage, comparative proteomic methodol-ogy to analyze in detail protein expression changesthroughout the differentiation process. The analysisidentified and quantified 21% of the parasite proteomeacross 7 time points during differentiation. The datareveal a delayed increase in gluconeogenesis enzymes,coinciding with a decrease in glycolytic capacity. At thesame time, �-oxidation, amino acid catabolism, tricar-boxylic acid cycle, mitochondrial respiration chain, andoxidative phosphorylation capacities are all up-regu-lated. The results indicate that the differentiating par-asite shifts from glucose to fatty acids and amino acidsas its main energy source. Furthermore, glycerol andamino acids are used as precursors for sugar synthesis,compensating for lack of exogenous sugars. Thesechanges occur while promastigotes undergo morpho-logical transformation. Our findings provide new in-sight into changes occurring in single-cell organismsduring a developmental process.—Rosenzweig D.,Smith, D., Opperdoes, F., Stern, S., Olafson, R. W.,Zilberstein D. Retooling Leishmania metabolism: fromsand fly gut to human macrophage. FASEB J. 22,590–602 (2008)

Key Words: proteomics � metabolomics � intracellular differen-tiation � gene expression

Intracellular parasitism is a process in which anorganism invades and proliferates inside a specific hostcell. To succeed, parasites have developed mechanismsto cope with the abrupt environmental changes theyencounter during invasion. These mechanisms result inthe parasite differentiating from an extracellular to anintracellular form adapted to life within the phagolyso-

some, a process that involves changes in gene expres-sion (1, 2). Our understanding of these adaptationmechanisms is limited. Recent sequencing of severalparasite genomes and advances in high throughputanalytical techniques are enabling new insights intothese processes. In the present communication we haveemployed a novel proteomics approach to identify geneproducts that enable parasite adaptation to a new hostduring differentiation of the intracellular parasitic pro-tist Leishmania donovani.

L. donovani are the causative agents of kala azar, afatal disease in humans. These organisms cycle betweenphagolysosomes of mammalian macrophages and thealimentary tract of sand flies. In the insect vector, theygrow as flagellated extracellular promastigotes, whereasin the mammalian host, they proliferate as aflagellatedintracellular amastigotes. Promastigotes are introducedinto the host during a blood meal taken by the fly andare subsequently phagocytosed by macrophages, wherethey differentiate into amastigotes (3). This process ismimicked in vitro by shifting cultured promastigotes(grown at 26°C, pH 7) to a lysosome-like environment(37°C and pH 5.5; differentiation signal) (4–6). Al-though axenic amastigotes are not identical to animal-derived amastigotes (7), they closely resemble themand are the best available model for the study ofparasite intracellular development.

To date, most studies on Leishmania differentiationhave focused on comparing promastigote and amasti-gote gene and protein expression. These studies haveshown that promastigotes and amastigotes are highlyadapted to their respective environments. For example,amastigote respiration, catabolism of energy substrates,and synthesis of macromolecules are carried out opti-mally at acidic pH whereas promastigotes performthese activities optimally at neutral pH. In addition,metabolic pathways such as glycolysis are more active inpromastigotes whereas fatty acid oxidation is moreactive in amastigotes (8). Also, gluconeogenesis is es-

1 Correspondence: Faculty of Biology, Technion—Israel Insti-tute of Technology, Haifa 32000, Israel. E-mail: danz@tx.technion.ac.il

doi: 10.1096/fj.07-9254com

590 0892-6638/08/0022-0590 © FASEB

sential for the virulence of the amastigotes and theirproliferation inside macrophages, but promastigotescan survive without this pathway (9). Indeed, numerousstudies have identified stage-specific genes, highlight-ing the importance of gene expression changes duringdifferentiation (10–19). Despite a wealth of knowledgeconcerning the different gene and protein expressionprofiles of promastigotes vs. amastigotes, it is not yetpossible to envisage the series of molecular events thatunderlie the differentiation process connecting theseparasite forms. Recently, our laboratory initiated time-course analysis of L. donovani differentiation andshowed that it is a regulated process involving changesin morphology that parallel discrete stages of proteinand gene expression (1, 4, 5; Lahav et al., unpublishedresults). These studies showed that differentiation isdivided into four stages: I) 0–4 h when promastigotesreceive and process the differentiation signal; II) 5–9 h,when they cease movement and aggregate; III) 10–24h, when promastigotes change morphologically intoamastigote-shaped cells; and IV) 25–120 h, when matu-ration into amastigotes is completed. The transitionbetween the first and second stages occurs synchro-nously, while cells are arrested at G1 (5).

Transcriptome analysis revealed that the majority ofpromastigote-specific gene transcripts were down-regu-lated early in differentiation, while amastigote-specifictranscripts were up-regulated only at the fourth stage ofdifferentiation (i.e., �24 h). However, gene expressionin Leishmania is regulated mostly posttranscriptionally(20), which means that changes in mRNA abundanceare not necessarily reflected faithfully in altered proteinprofiles. Moreover, it has been shown recently, both invitro and in vivo, that protein remodeling via autophagyis an essential process for promastigote to amastigotedifferentiation (21). Therefore, in order to form abetter picture of the molecular events that occur dur-ing the four differentiation stages, we elected to studythese stages at the proteomic level.

To follow proteomic changes during Leishmania dif-ferentiation with high protein coverage, we employed anovel isobaric tagging methodology, isobaric tags forrelative and absolute quantification (iTRAQ; 22). Thismethod utilizes amine-reactive isobaric tags to label allpeptides in a particular protein digest. An advantage ofthis method is that 4 protein digest samples can each betagged differently, allowing direct comparison of up to4 samples. Samples are combined at an equal ratio andsubjected to LC-MS/MS, where, on fragmentation,every fragmented peptide tag produces distinct signa-ture ions differing by an m/z value of 114–117. Therelative intensities of these signals represent the relativeabundance of the analyzed peptide in each sample.Relative abundance values of all peptides attributed toeach specific protein are averaged to represent therelative abundance of the entire protein.

Approximately 21% of the entire theoretical pro-teome was detected at �95% confidence. We observedmajor changes in the abundance of metabolic enzymesas well as altered levels of translation machinery and

DNA condensation proteins. These observations shedlight on how promastigotes retool themselves for life asamastigotes during differentiation. Remarkably, only afew proteins displayed significant changes in theirabundance during the initial 10 h of differentiation.

MATERIALS AND METHODS

Leishmania cell culture

A cloned line of L. donovani 1SR was used in all experiments(4). Promastigotes and amastigotes were grown as described(5). Differentiation of promastigotes to amastigotes in axenicculture was carried out as in Barak et al. (5). Briefly, late-logpromastigotes were transferred from promastigote medium at26°C to amastigote medium at 37°C and 5% CO2. Three daysafter initiation of differentiation, cells were split 1:2 inprewarmed amastigote medium.

Protein extraction

Cells were washed once in ice-cold 0.9% NaCl, then twice inice-cold PBS and finally suspended in PBS. Subsequently, cellsuspension was sonicated (3 cycles of 1 min sonication and 1min incubation on ice). Resulting lysate was centrifuged(12,000 g, 4°C, 10 min), and supernatant was precipitated incold acetone.

Peptide labeling

Labeling of peptides with the iTRAQ reagent was carried outessentially as suggested by the manufacturer (Applied Biosys-tems, Inc., Foster City, CA, USA). Briefly, from the commer-cial kit, denaturant was added to each of up to 4 sample tubescontaining 100 �g of protein, followed by reducing reagentand incubation at 60°C for 1 h. Cysteine residues were thenS-alkylated for 10 min followed by tryptic digestion overnightat 37°C. This was followed by N-alkylation of peptides with theappropriate iTRAQ reagents for 10 min at 37°C and combin-ing of differently labeled samples for analysis. Two labeledpeptide mixes were created; one mix included equal amountsof protein from promastigote, 2.5, 5, and 10 h of differenti-ation (early differentiation), and the other included promas-tigotes, 15 and 24 h of differentiation and mature amastigotes(late differentiation).

Multidimensional chromatography and mass-spectrometry

Labeled and multiplexed sample peptides were fractionatedusing 2-D LC employing strong-cation exchange (SCX) chro-matography for the first dimension separation. The latter wascarried out on a Vision Workstation (Applied Biosystems)equipped with a Polysulfoethyl A (Poly LC, Columbia, MD,USA) 100- � 4.6-mm, 5-�m, 300-Å SCX column. Buffer A was10 mM KPO4 (pH 2.7) in 25% ACN, while buffer B was 10mM KH2PO4 with 25% ACN in 0.5 M KCl. The column wasdeveloped with a linear gradient of 0–35% B in 30 min at aflow rate of 0.5 ml/min. SCX fractions were collected, andeach fraction subsequently was analyzed using an integratedonline RP LC/MS/MS system. The latter consisted of anautosampler, switching pump, and micropump (LC Packings,Amsterdam) interfaced to a hybrid Quadrupole-TOF LC/MS/MS Mass Spectrometer (QStar Pulsar I, Applied Biosys-tems). The latter was equipped with a nanoelectrosprayionization source (Proxeon, Odense, Denmark) and fitted

591PARASITE RETOOLING FOR LIFE IN A NEW HOST

with a 10 �m fused silica emitter tip (New Objective, Woburn,MA). RP LC was performed on a 75-�m � 15-cm C18 PepMapNano LC column (LC Packings, Amsterdam, The Nether-lands) with a 300-�m � 5-mm C18 PepMap guard column(LC Packings) in place. Samples were loaded onto a trap orguard column in a volume of 25–50 �l and were equilibratedfor 10 min in 95% solvent A/5% solvent B at a flow rate of 100�l/min. On switching inline with the MS, a linear gradient at200 nL/min from 95–40% solvent A was developed for 40min, and in the following 5 min the composition of mobilephase was increased to 20% A before decreasing to 95% A fora 15-min equilibration before the next sample injection.Fractions were eluted with a linear 60-min (low peptideconcentration) or 120-min (high peptide concentration)gradient from 5–60% solvent B. MS data were acquiredautomatically using Analyst QS 1.0 software Service Pack 8(Applied Biosystems MDS SCIEX, Concord, Canada). A 1-sTOFMS survey scan was conducted over 400–1200 amu,followed by two 2.5-s product ion scans over mass range100–1500 amu. The two most intense peaks showing 20counts/s with charge state of 2–5 were selected for fragmen-tation. A precursor ion within a 6-amu window, once selectedfor fragmentation, was excluded from detection for 180 s.Curtain gas was set at 23 (gas flow 1.22 L/min), nitrogen wasused as the collision gas, and the ionization tip voltage usedwas 2700 V.

Results-dependent data acquisition and peptide analysis

Using additional aliquots of the SCX fractions, exclusionanalyses were carried out. For this procedure, data files wereprocessed using the Protein Pilot software (Ver. 1.0) (AppliedBiosystems). A list of peptides (i.e., m/z values) that weredetected with confidence greater than 95% was generated byProtein Pilot, saved as an Excel file, and imported into theAnalyst method as a text file. The same data acquisitionmethod (.dam file) was used to acquire the first and secondLC/MS/MS data. Duplicate aliquots of SCX fractions wereanalyzed by the first and second methods. Peptides wereexcluded throughout the entire run of a specific SCX fractionbased on their m/z value. This procedure allowed identifica-tion of significantly more peptides than a single LC/MS/MSanalytical cycle.

Data processing and analysis

Data files were processed using the Protein Pilot software.The LC/MS/MS data were used to search GeneDB L. infan-tum V3.0 database. Proteins were considered to be confidentlyidentified if they had an unused Protscore � 1.3(Conf.�95%). The Protein Pilot software then calculated theratios for the reporter ion fragment masses for each peptide.

Data from early and late differentiation iTRAQ analysiswere united, using the promastigote time points as referencevalues, and log2-transformed. For clustering analysis, onlyproteins quantified at all time points were used. Proteins weregrouped according to self-organizing maps algorithm, withparameters set at 2 � 2 grid and 100,000 iterations usingExpander software (23).

Protein coexpression significance of each functional groupwas calculated using the Pearson correlation coefficient. Eachpairwise correlation average was then compared with a distri-bution of 1000 random groups with the same number ofproteins. Western blot analysis was done as in Barak et al. (5).Protein loading was based on tubulin load but confirmedusing Ponceau-red stain densitometry and NanoDrop® spec-trophotometry.

RESULTS

Identification of soluble proteins by iTRAQ

To date, studies from this laboratory have identifiedthree parameters that can be utilized as characteristicmarkers of L. donovani differentiation: morphologicalchanges (5), a gradual increase in the abundance ofmRNA coding for ascorbate-dependent peroxidase (1),and a transient increase in spliced leader RNA abun-dance (Lahav et al., unpublished results). In the presentstudy, promastigotes were differentiated into amasti-gotes on two separate occasions, representing indepen-dent biological repeats of complete axenic differentia-tion. All three time-course markers were exhibitedsimilarly in both repeats, and thus, according to thesedifferentiation criteria, were complete and consistentin both repeats.

L. donovani promastigotes were subjected to thedifferentiation signal (pH 5.5 and 37°C; time zero) andsoluble proteins collected at 2.5, 5, 10, 15, and 24 h, aswell as at 6 days postsignal when the parasites aremature, fully differentiated amastigotes. Protein sam-ples were divided into two groups of four for labeling.The first labeling group was unstimulated promasti-gotes (no differentiation signal) and promastigotes 2.5,5, and 10 h postsignal; the second labeling group wasunstimulated promastigotes, promastigotes 15 and 24 hpostsignal, and mature amastigotes. Each sample wasdigested and labeled with one of the four iTRAQ tags.Equal amounts of peptides from each sample weremixed together (according to the aforementioned la-beling groups) and subjected to multidimensionalLC-MS analysis (see Materials and Methods). In addi-tion, a control experiment was performed where twoindependent soluble protein samples derived fromunstimulated promastigotes were mixed and subjectedto multidimensional LC-MS analysis. The mean differ-ence in protein abundance for the control experimentwas �0.1-fold, indicating the iTRAQ method accuracy.Peptide masses obtained from mass-spectrometry anal-ysis were characterized using the theoretical L. infan-tum version 3.0 proteome (http://www.genedb.org/genedb/linfantum/). This genome was used becauseL. infantum and L. donovani are closely related OldWorld species.

Summing all the signals from both repeat experi-ments, a total of 12,021 distinct peptides were detected.These were grouped to identify 1713 proteins at �95%confidence out of 8184 theoretical proteins (Supple-mental Table S1). The detected proteins represent20.9% of the entire L. infantum proteome. Assumingthat at any given time point about half of the proteomeis expressed, this analysis detected almost half of theexpressed proteome. On average, 7 distinct peptideswere detected for each identified protein. Of theidentified proteins, 934 have a known or predictedfunction, whereas 779 are proteins with no knownfunction. The analysis revealed 969 proteins at all 7time points (Supplemental Table S1a), while the rest

592 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal

were detected only in the first (i.e., promastigote, 2.5, 5,10 h, Table S1b) or second labeling group (promasti-gote, 15, 24 h, amastigote, Table S1b). Of note, out ofthe 1713 proteins identified, only 931 were detected inboth biological repeats. The reason for the limitedoverlap is that iTRAQ is a shotgun-proteomic method,and, as such, the assortment of peptides identified ineach independent assay is random. Focusing on the 931proteins detected in both repeats, a correlation plot wasconstructed (Fig. 1A), demonstrating high correlationbetween the experimental repeats, ranging from 0.38in promastigotes to 0.85 in mature amastigotes. Thisplot indicated relatively small variations in a givenprotein’s abundance between repeats and corrobo-rated the reproducibility of data generated usingiTRAQ methodology. Since we were interested in char-acterizing the entire time course of differentiation, wedecided to include in our data analysis proteins identi-fied in only one of the repeats.

The protein expression profiles began to divergeslightly during the first two differentiation stages (up to9 h), but diverged more prominently during the thirdand fourth stages (Fig. 1B). Analysis of only proteinsdetected at all 7 time points, using a self-organizingmap algorithm, yielded four expression clusters (Fig.1C). The amastigote-specific cluster I contains 289proteins (16.9% of the proteome), while the promas-tigote-specific cluster II contains 310 proteins (18.1% ofthe proteome). The other proteins (370) do notchange expression significantly during differentiation(clusters III and IV). Both amastigote- and promastig-ote-specific expression gradually increased and de-creased, respectively, with the largest changes occur-ring during the third and fourth differentiation stages.

Expression pattern of metabolic enzymes

iTRAQ analysis detected many enzymes. Most majormetabolic pathways were represented in the data, in-cluding glycolysis, gluconeogenesis, �-oxidation, the

tricarboxylic acid cycle, oxidative phosphorylation, thepentose phosphate pathway, mitochondrial respiration,and amino acid catabolism (Table S2; Fig. 2).

Glycolysis is down-regulated during latedifferentiation

The glycolytic pathway between glucose and 3-phospho-glycerate occurs in the glycosomes, whereas the finalsteps that lead to formation of pyruvate are cytosolic(8). In trypanosomatids, the regulation of glycolysistakes place in the cytosol, with pyruvate kinase beingthe regulatory enzyme (24). We observed that theexpression of glycosomal glycolytic enzymes was up-regulated gradually during differentiation, such that itwas �1.4-fold higher in mature amastigotes (Fig. 2A).In contrast, the expression of cytosolic glycolytic en-zymes was down-regulated significantly and coordi-nately from 10 h postsignal, with phosphoglyceratemutase, enolase and pyruvate kinase expression re-duced 2-fold, on average, in amastigotes (Fig. 2B). Thisconcerted down-regulation of three consecutive glyco-lytic enzymes, one of which is the rate-limiting step,suggests active, comprehensive down-regulation of gly-colysis during the late stages of differentiation.

�-oxidation is up-regulated during late differentiation

iTRAQ analysis detected 13 proteins directly involvedin �-oxidation (Fig. 2C; Table S2). Except for oneacyl-CoA dehydrogenase isoform (LinJ28_V3.2700),which exhibited increased expression as early as 2.5 hpostsignal, the expression of these proteins did notchange significantly during the initial 15 h of differen-tiation. However, during the late third and fourthdifferentiation stages, 11 of the 13 proteins displayedsignificantly increased expression. iTRAQ also identi-fied 7 proteins that participate in �-oxidation indi-rectly, including lipases, fatty acyl-CoA synthetases/ligases and acetyl-CoA carboxylase. The amount of long

Figure 1. Time point correlation between biological repeats and protein expression pattern divergence as a function ofdifferentiation progression. A) Pairwise comparison of protein log2-transformed fold change from promastigotes (t�0 h) duringdifferentiation time points: I) t � 2.5 h; II) t � 5 h; III) t � 10 h; IV) t � 15 h; V) t � 24 h; VI) t � 144 h (amastigotes). B)Histogram of number of proteins as a function of log2-transformed expression values of proteins at each experimental timepoint during differentiation: (F) 2.5 h, (E) 5 h, (�) 10 h, (ƒ) 15 h, (f) 24 h (�) mature amastigotes. C) Expanderself-organizing maps algorithm clustering of proteins quantified at all 7 time points, using a 2 � 2 grid at 100,000 iterations.

593PARASITE RETOOLING FOR LIFE IN A NEW HOST

chain fatty acyl CoA synthetase, the rate-limiting en-zyme of �-oxidation responsible for associating long-chain free fatty acids with coenzyme A, increased 3-foldafter 15 h of differentiation and �13-fold in matureamastigotes, while the other ligases were down-regu-lated. The large increase in abundance of this enzymewas accompanied by a significant increase in the ex-pression of three acyl-CoA dehydrogenase isoenzymes,predicted by homology to be specific for very-long-,long-, and short/branched-chain fatty acids. In addi-tion, the level of 2,4-dienoyl-CoA reductase, an enzyme

responsible for processing unsaturated fatty acids priorto oxidation, was increased 1.8-fold at 15 h and 7.5-foldin the mature amastigote. These data substantiate thehypothesis that fatty acid oxidation is more active inamastigotes (25) and is adapted primarily to very-long-and long-chained and/or unsaturated fatty acids.

More support for this premise is provided by twoother observations. First, acetyl-CoA carboxylase, whichsynthesizes a key inhibitor of �-oxidation that is aprecursor of fatty acid synthesis (malonyl-CoA) (26),exhibited �1.6-fold down-regulated expression after

Figure 2. Expression patterns of metabolic en-zymes during differentiation. Log2-transformedexpression values of: A) glycosomal glycolyticenzymes; B) cytosolic glycolytic enzymes; C)�-oxidation enzymes; D) tricarboxylic acid cycleenzymes; E) mitochondrial respiration chainproteins; F) oxidative-phosphorylation proteins;G) amino acid catabolism enzymes; H) pentosephosphate pathway enzymes; I) glycosomal glu-coneogenesis enzymes; and J) cytosolic glucone-ogenesis enzymes. Coexpression P values offunctional groups A–F and H–I are �0.01, andof groups G and J �0.05.

594 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal

15 h of differentiation, retaining this reduced level inamastigotes. Second, �2-fold increases were detectedin the abundance of lipolytic enzymes, namely phos-pholipase A2 (PLA2) and monoglyceride lipase (Sup-plemental Table S2). PLA2 releases free fatty acids fromphospholipids and diacylglycerols, while monoglycer-ide lipase releases fatty acids from monoacylglycerols.Monoglyceride lipase increases gradually throughoutdifferentiation, while expression values of PLA2 onlyare available for late differentiation. Incidentally, the1.8-fold up-regulated level of glycerol kinase in matureamastigotes suggests strongly that the glycerol pro-duced by the lipases is utilized for gluconeogenesis (8).

Tricarboxylic acid cycle enzymes are up-regulatedduring differentiation

All enzymes of the tricarboxylic acid (TCA) cycle wereidentified at all differentiation time points (Fig. 2D;Supplemental Table S2). Although there were no ma-jor changes in abundance detected for most TCAenzymes during the first 15 h of differentiation, thelevels of these enzymes increased on average �1.7-foldduring the fourth differentiation stage. While two fu-marate hydratase isoenzymes displaying opposite ex-pression trends were detected, sequence alignment andin silico subcellular localization by TargetP indicatedthat LinJ24_V3.0310, the up-regulated isoenzyme, isprobably mitochondrial, whereas LinJ29_V3.2080 isprobably glycosomal. Thus, the generally increasedabundance of TCA cycle enzymes by the fourth differ-entiation stage suggests that TCA activity is higher inamastigotes.

Mitochondrial respiration and oxidativephosphorylation are up-regulated throughoutdifferentiation

Fourteen proteins involved in mitochondrial respira-tion were detected (Fig. 2E; Supplemental Table S2). Aslow but steady increase in expression of these proteinsreached a final �1.7-fold change, on average, in amas-tigotes. The coordinated increased expression of respi-ratory chain proteins implies that mitochondrial respi-ratory activity increases during differentiation.

Two proteins of the F0/F1 ATPase complex, an F1 subunit and a � subunit, were detected. The abundanceof these proteins increased slightly over the timecourse, the rise beginning at 15 h of differentiation andreaching 1.5-fold in mature amastigotes (Fig. 2F).Though modest, this increase was consistent acrossseveral time points, and we therefore consider it signif-icant. The ATP/ADP translocase, a carrier that couplesATP export from mitochondrion with ADP import, anda mitochondrial phosphate transporter responsible forphosphate transport from the cytosol to the mitochon-drion exhibited the same expression pattern. Theconcurrently increased expression of F1 ATPase sub-units, the ATP/ADP translocase, and a mitochondrialphosphate transporter suggests that oxidative phos-

phorylation increases during differentiation. Moreover,the similarity between the differentiation expressionpatterns of oxidative phosphorylation and mitochon-drial respiratory proteins suggests that expression ofthese proteins is linked functionally.

Amino acid catabolic enzymes are up-regulated duringlate differentiation

Five enzymes that catabolyze amino acids were detected(Fig. 2G): two glutamate dehydrogenase isoforms, analanine aminotransferase, a branched chain amino-transferase and a glycine cleavage system H protein. Allof these enzymes displayed an up-regulated differenti-ation expression pattern, reaching 3.3-fold expressionfor glycine cleavage system H protein, 1.9- and 3.8-foldfor the two glutamate dehydrogenase isoforms, 3.2-foldfor branched-chain aminotransferase and 2-fold foralanine aminotransferase in mature amastigotes.

Pentose phosphate pathway

iTRAQ detected five enzymes involved in the pentosephosphate pathway (Fig. 2H). The abundance of theseproteins decreased slightly, but significantly (P�0.01),during the fourth differentiation stage.

Expression pattern of gluconeogenic enzymes

Sixteen proteins of the gluconeogenesis pathway weredetected (Fig. 2I, J; Supplemental Table S2). Two keyenzymes of this pathway were up-regulated more than2-fold in amastigotes. The first is phosphoenolpyruvatecarboxykinase (PEPCK), which converts oxaloacetateto phosphoenolpyruvate (PEP), and the second isfructose 1,6-bis-phosphatase (F16BP), which dephos-phorylates fructose 1,6-bis-phosphate to fructose-6-phos-phate. The levels of PEPCK and F16BP increasedsteadily throughout differentiation, reaching 2.1- and2.9-fold in amastigotes, respectively (Fig. 2I). Othergluconeogenic enzymes exhibited moderate up-regula-tion, with the exception of the cytosolic enzymes,enolase, and two phosphoglycerate mutase isoenzymes,which also are part of the glycolytic pathway and weredown-regulated (Fig. 2J).

The observed decrease in glycolysis results in insuffi-cient pyruvate for gluconeogenesis in the maturingamastigote. A possible alternative source of pyruvatecould be the transamination of alanine; indeed, alanineaminotransferase expression was up-regulated 2-fold inthe mature amastigotes. Pyruvate is likely to be furtherphosphorylated by pyruvate phosphate dikinase (PPDK),the abundance of which began increasing early in thetime course, finally reaching 4.4-fold in amastigotes(Supplemental Table S1).

Proteins of the translation machinery

iTRAQ identified 90 ribosomal protein subunits, ofwhich 78 were detected at all differentiation time

595PARASITE RETOOLING FOR LIFE IN A NEW HOST

points (Fig. 3A, Supplemental Table S3). These pro-teins remained unchanged during the initial 10 h ofdifferentiation and subsequently gradually decreased inabundance, reaching �2.4-fold decrease in amasti-gotes. Interestingly, we observed an almost uniformexpression pattern of these proteins, suggesting strictregulation of the translation machinery. This is furtheremphasized by the similar expression pattern of trans-lation elongation and initiation factors detected (Fig.3B, Supplemental Table S3) as well as tRNA synthetases(Fig. 3C, Supplemental Table S3). Twenty-seven trans-lation factors and 22 tRNA synthetases were detectedand quantified. On average, both groups remainedunchanged during initial differentiation and were co-ordinately down-regulated from 15 h postsignal, to halfthe promastigote expression level in mature amastigotes.

Other expression trends

Twelve histones were detected. The expression ofthese proteins did not change during the first 10 h ofdifferentiation, and then increased, reaching �2-foldin mature amastigotes (Fig. 3D). One isoform ofhistone H2B displayed higher expression throughoutdifferentiation; its abundance increased 5.2-fold inamastigotes. The elevated expression of histones inthe amastigote stage may indicate transcriptionalchanges.

Calpains are cysteine peptidases that have been dem-onstrated to play a role in calcium-regulated functionssuch as signal transduction and cell differentiation inother organisms (27). iTRAQ analysis detected 10calpains (Fig. 3E) out of 33 putative calpains in the L.infantum genome. Five of these displayed 2- to 4-folddecreased expression in amastigotes, while threeshowed 1.4-fold to 1.92-fold increased expression inamastigotes. These observations indicate a develop-ment-dependent switch in the repertoire of calpainsexpressed.

Nineteen heat shock proteins and chaperones weredetected by iTRAQ. These displayed a very small in-crease during the first and second differentiationstages, then returned to promastigote or lower levels(Fig. 3F, Supplemental Table S4). One exception wasHSP100, which was previously reported to be amastig-ote-specific (28). Since HSP83 is encoded by a multi-copy gene family (29), it is not clear whether theiTRAQ results represented the products of all theHSP83 genes. Using polyclonal antibodies specific toHSP83 (5), Western blot analysis (Fig. 4A) demon-strated that the overall HSP83 expression pattern wassimilar but not identical to the iTRAQ data. On theother hand, Western blot analysis of HSP100, which issingle gene (28), displayed an expression pattern iden-tical to that reported by iTRAQ (Fig. 4B).

Figure 3. Expression patterns of translation ma-chinery proteins, histones, calpains and HSPsduring differentiation. Log2-transformed expres-sion values of: A) ribosomal protein subunits; B)translation factors; C) tRNA synthetases; D) his-tones; E) calpains; and F) heat shock proteinsand chaperones. Coexpression P values of func-tional groups A–D are �0.01.

596 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal

Proteins up-regulated at the beginning ofdifferentiation

Although most of the changes in protein abundanceoccurred during the third and fourth stages of differ-entiation, the expression of 20 proteins was up-regu-lated more than 2-fold within 2.5 h of differentiation(Table 1). Of these, 13 are proteins with unknownfunction, and the other 7 are as follows: D1 myo inositoltransporter (up-regulated 7.2-fold at 2.5 h), universalminicircle sequence binding protein (UMSBP) (up-regu-lated 6.7- and 10.4-fold at 2.5 and 5 h, respectively), aflavoprotein subunit, a methionine synthase, a histoneH1, a kinetoplast-associated protein, and ascorbate-dependent peroxidase. In the case of ascorbate-depen-dent peroxidase, the increase in protein abundance is

mirrored by the increase in its encoding mRNA (1).Notably, UMSBP was detected only in one repeat andonly at 2.5 and 5 h postsignal, in both cases at marginalconfidence. However, as its expression increased dra-matically, we used specific antibodies raised againstCrithidia fasciculata UMSBP to verify the differentiationexpression pattern by an alternative protocol (30). Inagreement with the iTRAQ data, Western blot analysisshows that a low level of UMSBP expression is detectedin unstimulated promastigotes, but expression subse-quently increases dramatically, remaining highthroughout the first 24 h of differentiation, only de-creasing in mature amastigotes back to unstimulatedpromastigote levels (Fig. 4C). These Western blot datanot only corroborate the developmental regulation ofUMSBP expression but also validate the reliability ofiTRAQ data.

Correlation analysis of functional group expression

To analyze and quantify the relationship between path-ways, a dendogram was created using Expander soft-ware whereby the average expression values for eachgroup of proteins with related functions were calcu-lated, correlated, and depicted (Fig. 5) (23). Thisdendogram analysis highlights several features of dif-ferentiation. Catabolic pathways that lead to energyproduction were up-regulated concurrently during thethird and fourth differentiation stages. Specifically, thecatabolic pathways TCA cycle, mitochondrial respira-tion, and oxidative phosphorylation exhibited highlycorrelated expression. The expression pattern of onebiosynthetic pathway, glycosomal gluconeogenesis, was

Figure 4. Validation of expression patterns. Western blotanalysis of: A) HSP83 expression pattern; B) HSP100 expres-sion pattern; and C) UMSBP expression pattern.

TABLE 1. Proteins up-regulated significantly at early hours of differentiation

GeneDB ID Protein name

Time in differentiation (h)a

0 2.5 5 10 15 24 144

LinJ28_V3.2020 Hypothetical protein, conserved 0.0 3.9 3.1 4.6 ND ND NDLinJ24_V3.0690 D1 myo inositol transporter 0.0 2.8 1.7 1.9 ND ND ND

LinJ36_V3.1680UMSBP1 universal minicircle sequence binding protein,

putative 0.0 2.8 3.4 ND ND ND NDLinJ26_V3.2190 Hypothetical protein, conserved 0.0 2.7 2.9 2.3 ND ND NDLinJ07_V3.0910 Flavoprotein subunit-like protein 0.0 1.9 1.5 2.4 1.0 1.0 2.0LinJ07_V3.0240 Methionine synthase, putative 0.0 1.5 0.9 1.8 ND ND NDLinJ33_V3.3390 h1 histone-like protein 0.0 1.5 0.1 0.0 0.2 0.0 0.5LinJ29_V3.0870 Hypothetical protein, conserved 0.0 1.5 0.6 0.2 ND ND NDLinJ03_V3.0510 Hypothetical protein, conserved 0.0 1.4 1.4 1.1 ND ND NDLinJ13_V3.0700 Hypothetical protein, conserved 0.0 1.4 1.3 1.3 ND ND NDLinJ21_V3.1380 Hypothetical protein, conserved 0.0 1.3 0.1 0.2 0.7 ND NDLinJ33_V3.1460 Hypothetical protein, conserved 0.0 1.2 0.9 0.8 0.4 0.3 1.1LinJ02_V3.0210 Hypothetical protein, conserved 0.0 1.2 0.6 0.3 ND ND NDLinJ30_V3.1760 Hypothetical protein, conserved 0.0 1.2 0.3 0.7 ND ND NDLinJ27_V3.0250 Kinetoplast-associated protein-like protein 0.0 1.1 1.4 2.0 2.0 1.7 NDLinJ34_V3.0070 Ascorbate-dependent peroxidase, putative 0.0 1.1 1.4 1.5 0.9 1.1 2.1LinJ24_V3.1040 Hypothetical protein, unknown function 0.0 1.1 0.8 0.9 ND ND NDLinJ18_V3.0610 Hypothetical protein, conserved 0.0 1.1 0.5 0.9 0.9 0.9 0.3LinJ22_V3.0520 Hypothetical protein, conserved 0.0 1.0 0.4 1.5 0.0 0.0 0.2LinJ15_V3.1000 Hypothetical protein, conserved 0.0 1.0 0.0 0.2 ND ND ND

aValues are log2-fold from promastigotes. ND � not detected.

597PARASITE RETOOLING FOR LIFE IN A NEW HOST

found to correlate with the expression patterns of�-oxidation and amino acid catabolism enzymes. Ribo-somal protein subunits and translation factors dis-played a coordinated pattern of down-regulation. Thus,in general, anabolic functions were down-regulated,whereas catabolic functions were up-regulated.

The dendogram analysis confirms that, as expected,when considering all the time points, amastigote pro-tein expression (6 days postsignal) is the most divergedfrom promastigote expression (unstimulated). Theprotein expression patterns detected at 15 h and 24 hpostsignal are significantly different from those ob-served at 2.5, 5, and 10 h, which reflects the increasingmagnitude of expression changes during the thirddifferentiation stage.

DISCUSSION

The present study followed changes in the abundanceof proteins during L. donovani differentiation in orderto understand how the extracellular promastigote pre-pares for a life as an intracellular amastigote. Atpresent, the continuous monitoring of L. donovanidifferentiation using animal models is not possible;therefore, we used an established, host-free axenicsystem that simulates the in vivo differentiation process(5). Using high-coverage, comparative proteomic anal-ysis of soluble parasite proteins, we collected expressioninformation on 21% of the L. donovani proteome. Ourdata reveal changes in the expression level of mostproteins involved in core metabolic pathways and ofmany molecules involved in DNA condensation andtranslation. This study, for the first time, enables acomprehensive view of the biogenesis processes under-lying differentiation from promastigote to amastigote.

This is summarized in the dynamic metabolome of L.donovani differentiation (Fig. 6). Our major observa-tions include: a) during differentiation, the parasitesshift from glucose to fatty acid oxidation as the mainsource of metabolic energy; b) in line with the increasein �-oxidation capacity, TCA cycle enzymes, mitochon-drial respiratory chain, and oxidative phosphorylationproteins are up-regulated; c) concomitantly, differenti-ating parasites up-regulate gluconeogenesis, producingsugars from glycerol and amino acids; d) translationslows down in maturing amastigotes, and DNA conden-sation increases; and e) most changes in proteins ofphysiological significance occurred during the thirdand fourth differentiation stages, when promastigotesundergo a morphological transition to amastigotes.

Our investigations revealed that during late differen-tiation, the abundance of glycosomal glycolytic enzymesincreases slightly, whereas the expression of three con-secutive cytosolic enzymes is down-regulated signifi-cantly. Therefore, we propose that the decrease inabundance of cytosolic glycolytic enzymes must resultin the formation of a bottleneck, which slows theglycolytic flux. Our conclusion concurs with and vali-dates early in vivo studies by Hart and Coombs (25),who reported low activities for several amastigote gly-colytic enzymes, particularly pyruvate kinase.

In contrast to the down-regulated expression ob-served for glycolytic enzymes, the expression of two keyregulatory gluconeogenic enzymes, PEPCK and F16BP,is gradually but significantly up-regulated throughoutdifferentiation. Since gluconeogenic flux is one orderof magnitude smaller than glycolytic flux (8), theup-regulation of key enzymes likely indicates a signifi-cant increase in gluconeogenic flux. These data fit withrecent in vivo studies demonstrating that gluconeogen-esis is essential for amastigote proliferation inside mu-rine macrophages (9). Naderer et al. (9) also indicatedthat the parasitophorous vacuole is poor in glucose,requiring parasites to synthesize their own sugars fromother sources, such as amino acids and glycerol. In linewith this, we observed that the expression of glycerolkinase and NAD-dependent glycerol-3-phosphate dehy-drogenase is increased significantly in amastigotes, ac-celerating the formation of glycosomal triosephos-phate, two molecules of which are used as substrate forgluconeogenesis. The reduced NAD required for thisprocess is provided by the observed increased amountof PEPCK and malate dehydrogenase, which also wasreported by others (31, 32). Notably, consumption ofglycerol occurs in the glycosomes, which ensures that itis not affected by the cytosolic bottleneck likely createdby the observed low expression of enolase and phos-phoglycerate mutase. Under glycerol limitation, PEPalone or in combination with glycerol may be used assubstrate for glucose formation. In the latter case theglycosomal PPDK reaction (together with adenylatekinase) is responsible for the production of the addi-tional ATP required. Indeed, PPDK and adenylatekinase also are up-regulated in the amastigote stage.Based on our observation that the abundance of cyto-

Figure 5. Function group differentiation time course coex-pression analysis. Hierarchical analysis of protein functiongroups expression pattern during differentiation was per-formed using proteins that were detected at all 7 time points.Protein expression values for each time point were averagedfor each function group. Average expression patterns wereanalyzed for correlation using Expander complete-linkagehierarchical clustering.

598 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal

Figure 6. Model of the dynamic metabolome of L. donovani differentiation. This model is an expansion of the scheme describedin (8). Red arrows represent amastigote up-regulated enzymes, green arrows represent amastigote down-regulated enzymes, grayarrows indicate undetected enzymes, and black arrows represent no significant change in abundance. Dashed arrows indicatesubstrates common to several pathways. Enzymes indicated: 1) hexokinase; 2) glucose-6-phosphate isomerase; 3) 6-phospho-1-fructokinase; 4) fructose-1,6-bis-phosphate aldolase; 5) triosephosphate isomerase; 6) glyceraldehyde 3-phosphate dehydroge-nase; 7) phosphoglycerate kinase C; 8) phosphoglycerate kinase B; 9) phosphoglycerate mutase; 10) enolase; 11) pyruvatekinase; 12) fructose-1,6-bis-phosphatase; 13) glucosamine-6-phosphate isomerase; 14) phospholipase C; 15) phospholipase A2;16) monoglyceride lipase; 17) glycerol kinase; 18) glycerol-3-phosphate dehydrogenase; 19) acetyl-CoA synthetase; 20)acetyl-CoA carboxylase; 21) long-chain fatty Acyl CoA synthetase; 22) acetate-succinate CoA transferase; 23) pyruvatedehydrogenase; 24) aconitase; 25) isocitrate dehydrogenase; 26) 2-oxoglutarate dehydrogenase; 27) succinyl-CoA ligase; 28)succinate dehydrogenase; 29) fumarate hydratase; 30) malate dehydrogenase; 31) citrate synthase; 32) glutamate dehydroge-nase; 33) alanine aminotransferase; 34) citrate lyase; 35) pyruvate phosphate dikinase; 36) phosphoenolpyruvate carboxykinase;37) fumarate reductase; and 38) adenylate kinase.

599PARASITE RETOOLING FOR LIFE IN A NEW HOST

solic enzymes common to both glycolysis and glucone-ogenesis is decreased significantly during amastigotematuration, we hypothesize that residual enzymaticactivity must be dedicated mainly to gluconeogenesis.

In this case, as differentiation progresses, an alterna-tive to glucose must be utilized as the main energysource. A logical candidate for such a source during thelate stages of differentiation is fatty acids. Early in vivostudies showed that enzymatic activities associated with�-oxidation of fatty acids were significantly higher in L.mexicana amastigotes than in promastigotes (25, 33). Inagreement with this, we found in the present study thatthe expression of L. donovani �-oxidation pathwayenzymes is up-regulated gradually from 10 h postsignal,suggesting an increase in �-oxidation activity. Up-regu-lated expression of long-chain fatty acid acyl-CoA syn-thetase and lipolytic enzymes on the one hand, anddown-regulated expression of acetyl-CoA carboxylaseon the other, further support a strong up-regulation of�-oxidation flux in amastigotes. Moreover, the in-creased abundance of three different chain-length spe-cific acyl-CoA dehydrogenases and a 2,4-dienoyl-CoAreductase indicates that maturing amastigotes increasetheir capacity for utilizing a wide range of fatty acidchain lengths, including very-long chain and unsatur-ated fatty acids. The multiple acetyl-CoA units pro-duced by �-oxidation are handled by the TCA cycle inthe mitochondrion, and the reducing equivalents sogenerated are oxidized by the mitochondrial respira-tory chain,

Dendogram analysis revealed highly correlated ex-pression patterns for �-oxidation, TCA cycle, mitochon-drial respiration, and oxidative phosphorylationthroughout differentiation. As this correlation reflectsdata from multiple time points and since �-oxidationhas been demonstrated to be up-regulated in vivo, weconclude that likely all these pathways are up-regulatedduring differentiation. Moreover, we propose that theincrease in �-oxidation is linked functionally to in-creased TCA cycle activity, mitochondrial respiration,and oxidative phosphorylation.

In addition to �-oxidation of fatty acids, amastigotesalso are geared toward the catabolism of certain aminoacids. A branched chain aminotransferase and alanineaminotransferase, two glutamate dehydrogenase isoen-zymes, and at least one component of the glycine-cleavage system are all up-regulated, suggesting thatisoleucine, valine, alanine, glutamate, and glycine allcan be used as substrates by amastigotes.

We observed a differentiation-associated, significantincrease in the abundance of two enzymes that synthe-size PEP, PPDK, and PEPCK, which utilize pyruvate andoxaloacetate, respectively. In line with this, our analysisindicates that levels of these substrates should rise asdifferentiation progresses. Elevated oxaloacetate levelsare expected from the deduced up-regulation of TCAcycle activity, and increased pyruvate levels can beassumed in light of the observed up-regulation ofamino acid catabolism. In addition, increased expres-sion of both alanine aminotransferase and glutamate

dehydrogenase during differentiation enhances pro-duction not only of pyruvate but also NADH.

Interestingly, we observed that during the third stageof differentiation, at the time when promastigotestransform morphologically into amastigotes (15 h), ahighly synchronized decrease occurred in the abun-dance of translation machinery proteins, includingribosomal protein subunits, translation factors, andtRNA synthetases. This synchronous decrease was asso-ciated with an estimated 5-fold decrease in cell volumeand a 2.5-fold decrease in growth rate when comparingamastigotes to promastigotes. The fact that thesechanges occur during the third differentiation stageand not before suggests strongly that they are part ofthe parasite differentiation program and underlieadapting to a generally lower protein synthesis rate.Indeed, previous studies indicate that protein synthesisis lower in cultured amastigotes than in promastigotes(4) and is low also in animal-derived amastigotes (34).Of note, the observed decreased expression of transla-tion machinery proteins correlates inversely with theobserved increase in histone abundance. The increasedexpression of histones suggests transcriptional changesduring late differentiation.

In general, the present study reveals that most met-abolic and translational changes take place synchro-nously during the third and fourth differentiationstages. Therefore, it stands out that the expression ofmajor heat shock proteins is up-regulated transientlyduring the first two differentiation stages. This tran-sient expression may allow the parasite time for signalperception and processing before morphogenesis.Taken together, our data support strongly the notionthat amastigotes are highly adapted to life within thephagolysosomes, which are likely to be poor in carbo-hydrates but rich in amino acids, triglycerides, phos-pholipids, and their lipolytic byproducts such as fattyacids and glycerol.

The results also indicate the high level of promasti-gote adaptation to the sand fly midgut. The sand flydiet is based on fruit nectar, which is sucrose rich. Theycontinuously release sucrose from their crop into themidgut, where it is converted to glucose (35). Occa-sional bloodmeals supplement the nectar diet of femalesand flies. Thus, midgut promastigotes are continu-ously exposed to high sugar levels. Evidently, late logand stationary promastigotes that were used in thiswork to initiate differentiation were shown to expressenhanced glycolytic activity relative to amastigotes (36).

We did not expect to observe so few (only 20)significant changes in protein abundance during theearly stages of differentiation. These proteins mayinitiate or reflect the initiation of differentiation. Un-like the expression changes observed in the third andfourth differentiation stages, the changes in proteinabundance observed during the first two stages do notappear coordinated. Based on these observations, wesurmise that the mechanisms underlying the observedmetabolic retooling of amastigotes must be mostlyposttranslational. Given that metabolic retooling oc-

600 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal

curs late in differentiation, the following question isprompted: how do incompletely differentiated para-sites manage to reside in the same host environment asmature amastigotes? We speculate that the parasitescarry into the new environment endogenous substratepools, which are utilized during the period of transitionand enable survival during this time. Future researchcan address this speculation directly.

To summarize, the present study suggests that aprominent feature of the promastigote-to-amastigotedifferentiation process is a switch to producing moremetabolic energy. We propose that the energy is impor-tant for driving primary proton pumps and secondarycarriers, required to battle the large proton gradientbetween the acidic phagolysosome and the amastigotecytosol. Thus the energy enables pH homeostasis to bemaintained. The significantly increased expression of anumber of proton-translocating ATPases and carriersthat we observed provides further support for thismodel.

The present study focused on changes in proteinabundance occurring during differentiation from pro-mastigote into amastigote, the latter being the highlyinfectious parasitic form observed in diseased patients.As such, this study is the first proteome-wide time-course analysis of the metabolic and physiological ad-justments that take place during the developmentalcycle of parasitic protozoa.

We thank Drs. N. Holland and T. Lahav for helping withthe differentiation culture and Prof. D. Cassel for his fruitfuldiscussions. This work was supported by UNDP/World Bank/WHO Special Programme for Research and Training inTropical Diseases and grant 2003237 from the U.S.-IsraelBinational Science Foundation.

REFERENCES

1. Saxena, A., Lahav T, Holland, N., Anupama, A., Huang, G.,Volpin, H., Myler, P. J., and Zilberstein, D. (2007) Analysis of theLeishmania donovani transcriptome reveals an ordered progres-sion of transient and permanent changes in gene expressionduring differentiation. Mol. Biochem.Parasitol. 152, 53–65

2. Bozdech, Z., Llinas, M., Pulliam, B. L., Wong, E. D., Zhu, J., andDeRisi, J. L. (2003) The transcriptome of the intraerythrocyticdevelopmental cycle of Plasmodium falciparum. PLoS Biol. 1, E5

3. Herwaldt, B. L. (1999) Leishmaniasis. Lancet 354, 1191–11994. Saar, Y., Ransford, A., Waldman, E., Mazareb, S., Amin-Spector,

S., Plumblee, J., Turco, S. J., and Zilberstein, D. (1998) Charac-terization of developmentally-regulated activities in axenicamastigotes of Leishmania donovani. Mol. Biochem. Parasitol. 95,9–20

5. Barak, E., Amin-Spector, S., Gerliak, E., Goyard, S., Holland, N.,and Zilberstein, D. (2005) Differentiation of Leishmania dono-vani in host-free system: analysis of signal perception andresponse. Mol. Biochem. Parasitol. 141, 99–108

6. Debrabant, A., Joshi, M. B., Pimenta, P. F., and Dwyer, D. M.(2004) Generation of Leishmania donovani axenic amastigotes:their growth and biological characteristics. Int. J. Parasitol. 34,205–217

7. Holzer, T. R., McMaster, W. R., and Forney, J. D. (2006)Expression profiling by whole-genome interspecies microarrayhybridization reveals differential gene expression in procyclicpromastigotes, lesion-derived amastigotes, and axenic amasti-gotes in Leishmania mexicana. Mol. Biochem. Parasitol. 146, 198–218

8. Opperdoes, F. R., and Coombs, G. H. (2007) Metabolism ofLeishmania: proven and predicted. Trends Parasitol.

9. Naderer, T., Ellis, M. A., Sernee, M. F., De Souza, D. P., Curtis,J., Handman, E., and McConville, M. J. (2006) Virulence ofLeishmania major in macrophages and mice requires the glu-coneogenic enzyme fructose-1,6-bisphosphatase. Proc. Natl.Acad. Sci. U. S. A. 103, 5502–5507

10. Zhang, W. W., Charest, H., Ghedin, E., Matlashewski, G.(1996) Identification and overexpression of the A2 amasti-gote-specific protein in Leishmania donovani. Mol. Biochem.Parasitol. 78, 79 –90

11. Bates, P. A. (1993) Characterization of developmentally-regu-lated nucleases in promastigotes and amastigotes of Leishmaniamexicana. FEMS Microbiol. Lett. 107, 53–58

12. Hubel, A., Brandau, S., Dresel, A., and Clos, J. (1995) A memberof the ClpB family of stress proteins is expressed during heatshock in Leishmania spp. Mol. Biochem. Parasitol. 70, 107–118

13. Wiese, M. (1998) A mitogen-activated protein (MAP) kinasehomologue of Leishmania mexicana is essential for parasitesurvival in the infected host. EMBO J. 17, 2619–2628

14. Walker, J., Vasquez, J. J., Gomez, M. A., Drummelsmith, J.,Burchmore, R., Girard, I., and Ouellette, M. (2006) Identifica-tion of developmentally-regulated proteins in Leishmania pana-mensis by proteome profiling of promastigotes and axenicamastigotes. Mol. Biochem. Parasitol. 147, 64–73

15. Nugent, P. G., Karsani, S. A., Wait, R., Tempero, J., and Smith,D. F. (2004) Proteomic analysis of Leishmania mexicana differen-tiation. Mol. Biochem. Parasitol. 136, 51–62

16. Almeida, R., Gilmartin, B. J., McCann, S. H., Norrish, A., Ivens,A. C., Lawson, D., Levick, M. P., Smith, D. F., Dyall, S. D., Vetrie,D., Freeman, T. C., Coulson, R. M., Sampaio, I., Schneider, H.,and Blackwell, J. M. (2004) Expression profiling of the Leishma-nia life cycle: cDNA arrays identify developmentally regulatedgenes present but not annotated in the genome. Mol. Biochem.Parasitol. 136, 87–100

17. Akopyants, N. S., Matlib, R. S., Bukanova, E. N., Smeds, M. R.,Brownstein, B. H., Stormo, G. D., and Beverley, S. M. (2004)Expression profiling using random genomic DNA microarraysidentifies differentially expressed genes associated with threemajor developmental stages of the protozoan parasite Leishma-nia major. Mol. Biochem. Parasitol. 136, 71–86

18. Bente, M., Harder, S., Wiesgigl, M., Heukeshoven, J., Gelhaus,C., Krause, E., Clos, J., and Bruchhaus, I. (2003) Developmen-tally induced changes of the proteome in the protozoan parasiteLeishmania donovani. Proteomics 3, 1811–1829

19. Huynh, C., Sacks, D. L., and Andrews, N. W. (2006) A Leishma-nia amazonensis ZIP family iron transporter is essential forparasite replication within macrophage phagolysosomes. J. Exp.Med. 203, 2363–2375

20. Clayton, C. E. (2002) Life without transcriptional control? Fromfly to man and back again. EMBO J. 21, 1881–1888

21. Besteiro, S., Williams, R. A., Morrison, L. S., Coombs, G. H., andMottram, J. C. (2006) Endosome sorting and autophagy areessential for differentiation and virulence of Leishmania major.J. Biol. Chem. 281, 11384–11392

22. Melanson, J. E., Avery, S. L., and Pinto, D. M. (2006) High-coverage quantitative proteomics using amine-specific isotopiclabeling. Proteomics 6, 4466–4474

23. Shamir, R., Maron-Katz, A., Tanay, A., Linhart, C., Steinfeld, I.,Sharan, R., Shiloh, Y., and Elkon, R. (2005) EXPANDER–anintegrative program suite for microarray data analysis. BMCBioinformatics 6, 232

24. Van Schaftingen, E., Opperdoes, F. R., and Hers, H. G. (1985)Stimulation of Trypanosoma brucei pyruvate kinase by fructose2,6-bisphosphate. Eur. J. Biochem. 153, 403–406

25. Hart, D. T., and Coombs, G. H. (1982) Leishmania mexicana:energy metabolism of amastigotes and promastigotes. Exp. Para-sitol. 54, 397–409

26. McGarry, J. D., Mannaerts, G. P., and Foster, D. W. (1977) Apossible role for malonyl-CoA in the regulation of hepatic fattyacid oxidation and ketogenesis. J. Clin. Invest. 60, 265–270

27. Besteiro, S., Williams, R. A., Coombs, G. H., and Mottram, J. C.(2007) Protein turnover and differentiation in Leishmania. Int. J.Parasitol. 37, 1063–1075

28. Hubel, A., Krobitsch, S., Horauf, A., and Clos, J. (1997) Leish-mania major Hsp100 is required chiefly in the mammalian stageof the parasite. Mol. Cell. Biol. 17, 5987–5995

601PARASITE RETOOLING FOR LIFE IN A NEW HOST

29. Ivens, A. C., Peacock, C. S., Worthey, E. A., Murphy, L.,Aggarwal, G., Berriman, M., Sisk, E., Rajandream, M. A., Adlem,E., Aert, R., Anupama, A., Apostolou, Z., Attipoe, P., Bason, N.,Bauser, C., Beck, A., Beverley, S. M., Bianchettin, G., Borzym, K.,Bothe, G., Bruschi, C. V., Collins, M., Cadag, E., Ciarloni, L.,Clayton, C., Coulson, R. M., Cronin, A., Cruz, A. K., Davies,R. M., De, G. J., Dobson, D. E., Duesterhoeft, A., Fazelina, G.,Fosker, N., Frasch, A. C., Fraser, A., Fuchs, M., Gabel, C., Goble,A., Goffeau, A., Harris, D., Hertz-Fowler, C., Hilbert, H., Horn,D., Huang, Y., Klages, S., Knights, A., Kube, M., Larke, N., Litvin,L., Lord, A., Louie, T., Marra, M., Masuy, D., Matthews, K.,Michaeli, S., Mottram, J. C., Muller-Auer, S., Munden, H.,Nelson, S., Norbertczak, H., Oliver, K., O’Neil, S., Pentony, M.,Pohl, T. M., Price, C., Purnelle, B., Quail, M. A., Rabbinowitsch,E., Reinhardt, R., Rieger, M., Rinta, J., Robben, J., Robertson, L.,Ruiz, J. C., Rutter, S., Saunders, D., Schafer, M., Schein, J.,Schwartz, D. C., Seeger, K., Seyler, A., Sharp, S., Shin, H., Sivam,D., Squares, R., Squares, S., Tosato, V., Vogt, C., Volckaert, G.,Wambutt, R., Warren, T., Wedler, H., Woodward, J., Zhou, S.,Zimmermann, W., Smith, D. F., Blackwell, J. M., Stuart, K. D.,Barrell, B., and Myler, P. J. (2005) The genome of the kineto-plastid parasite, Leishmania major. Science 309, 436–442

30. Onn, I., Kapeller, I., bu-Elneel, K., and Shlomai, J. (2006)Binding of the universal minicircle sequence binding protein atthe kinetoplast DNA replication origin. J. Biol. Chem. 281,37468–37476

31. Mottram, J. C., and Coombs, G. H. (1985) Leishmania mexicana:enzyme activities of amastigotes and promastigotes and theirinhibition by antimonials and arsenicals. Exp. Parasitol. 59,151–160

32. Hassan, H. F., Mottram, J. C., and Coombs, G. H. (1985)Subcellular localisation of purine-metabolising enzymes in Leish-mania mexicana mexicana. Comp. Biochem. Physiol. B. 81, 1037–1040

33. Coombs, G. H., Craft, J. A., and Hart, D. T. (1982) A compar-ative study of Leishmania mexicana amastigotes and promastig-otes. Enzyme activities and subcellular locations. Mol. Biochem.Parasitol. 5, 199–211

34. Mukkada, A. J., Meade, J. C., Glaser, T. A., Bonventre, P. F.(1985) Enhanced metabolism of Leishmania donovani amasti-gotes at acid pH: an adaptation for intracellular growth. Science229, 1099–1101

35. Dillon, R. J., and el-Kordy, E. (1997) Carbohydrate digestionin sandflies: alpha-glucosidase activity in the midgut of Phle-botomus langeroni. Comp. Biochem.Physiol. B Biochem. Mol. Biol.116, 35– 40

36. Mukkada, A. J., Schaefer, F. W. I., Simon, M. W., and Neu, C.(1974) Delayed in vitro utilization of glucose by Leishmaniatropica promastigotes. J. Protozool. 21, 393–397

Received for publication June 18, 2007.Accepted for publication August 23, 2007.

602 Vol. 22 February 2008 ROSENZWEIG ET AL.The FASEB Journal