Vaccine 23 (2005) 4754–4764

Expression of theNecator americanushookworm larval antigenNa-ASP-2 inPichia pastorisand purification of the recombinant

protein for use in human clinical trials

Gaddam Narsa Gouda,∗,1, Maria Elena Bottazzia,∗,1, Bin Zhana, Susana Mendeza,Vehid Deumica, Jordan Plieskatta, Sen Liua, Yan Wanga, Lilian Buenoa,Ricardo Fujiwaraa, Andre Samuela, So Yeong Ahna, Maneesha Solankie,Oluwatoyin A. Asojoc, Jin Wangb, Jeffrey M. Bethonya, Alex Loukasa,d,

Michael Roye, Peter J. Hoteza,∗a Department of Microbiology and Tropical Medicine, The George Washington University Medical Center,

Ross Hall 736, 2300 Eye Street NW, Washington, DC 20037, USAb Malaria Vaccine Development Branch, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA

c Eppley Institute for Research in Cancer and Allied Diseases, University of Nebraska Medical Center, Omaha, NE 68198-7696, USAd Division of Infectious Diseases and Immunology, Queensland Institute of Medical Research, Brisbane, Qld, Australia

e

A

antv 3 kDar icalpa tested fort©

K

1

cdwA

m

,000

or-zoleove

entsellingost-s of

sh

0d

Science Applications International Corporation, Frederick, MD, USA

Received 3 March 2005; received in revised form 26 April 2005; accepted 29 April 2005Available online 6 June 2005

bstract

The ASP-2 protein secreted by infective larvae of the human hookworm,Necator americanus, is under development as a recombinaccine. RecombinantNa-ASP-2 was expressed inPichia pastoris, and the purified protein was characterized. At the 60 L scale, the 21.ecombinant protein was produced at a yield of 0.4 g/L. When formulated with Alhydrogel® and injected into rats to determine immunologotency, three 50�g doses of the formulated recombinant protein elicited geometric mean antibody titers up to 1:234,881. Rat anti-Na-ASP-2ntibody recognized larval-derived ASP-2 and also inhibited larval migration through skin in vitro. The processes developed and

he high yield production of recombinantNa-ASP-2 provide a foundation for clinical vaccine development.2005 Elsevier Ltd. All rights reserved.

eywords:Hookworm; Vaccine;Necator americanus; Na-ASP-2

. Introduction

Human hookworm infection caused byNecator ameri-anusor Ancylostoma duodenaleis a major cause of ironeficiency anemia and protein malnutrition in the developingorld and a leading cause of parasitic disease burden[1,2].n estimated 740 million people are infected in areas of rural

∗ Corresponding author. Tel.: +1 202 994 3532; fax: +1 202 994 2913.E-mail addresses:mtmgng@gwumc.edu (G.N. Goud),

tmmeb@gwumc.edu (M.E. Bottazzi), Photez@gwu.edu (P.J. Hotez).1 Both GNG and MEB contributed equally to the manuscript.

poverty in the tropics and subtropics resulting in up to 65deaths and 22 million disability-adjusted life years[3,4]. Themajor approach to hookworm control relies on reducing mbidity through the frequent and periodic use of benzimidaanthelminthic (BZAs) drugs, which can temporarily remadult parasites in the human gastrointestinal tract[4,5]. Suchtreatment for school-aged children leads to improvemin health and educational achievement[4]. However, therare reasons why exclusive reliance on BZAs for controhookworm infection may not be effective. Among them, ptreatment hookworm re-infection rates are high in areahigh transmission[6], and the efficacy of BZAs can dimini

264-410X/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2005.04.040

G.N. Goud et al. / Vaccine 23 (2005) 4754–4764 4755

with increasing and frequent use[7]. It has been further sug-gested that the emergence of BZA drug resistance is possible[7]. School-based programs are also not expected to interrupthookworm transmission[8].

In order to develop new tools for the control of hook-worm infection, studies were undertaken to develop recom-binant hookworm vaccines. Based on previous success in thedevelopment of canine hookworm vaccines from third-stageinfective larvae (L3) that were attenuated by X-irradiation[9], our antigen discovery efforts focused on the identifi-cation and isolation of critical L3-secreted molecules thatmediate protective immunity[10]. These included a novelclass of cysteine-rich secretory proteins known as theAncy-lostomasecreted proteins or ASPs[10–14]. The ASPs belongto the pathogenesis-related (PR) protein superfamily whosemembers are found in a wide variety of animals, plants andfungi where they are typically produced in response to stressor injury, including invasion by pathogens[15]. Althoughthe function of the hookworm larval ASPs is not known,the observation that they are released by L3 upon receivinghost-specific stimuli suggests that these proteins are releasedduring host entry and function in the transition from the envi-ronment to a parasitic existence[13].

ASP-2 was selected for evaluation as a recombinant vac-cine candidate based on human immunoepidemiologic stud-ies and laboratory animal vaccine trials[16–19]. These datal ormN ibet es-s ey gi-c dya site-d anceo laytH

2

2t

3o hatar t wasgd xi-ms sw th thes posi-

tive clones from secondary screening were excised in vivoas pBluescript phagemids according to the manufacturer’sinstruction (Stratagene). Phagemid DNA was extracted usingthe alkaline lysis method (Qiagen), and both strands weresequenced using vector primers, T3 and T7.

The complete 5′ end of Na-asp-2cDNA was isolatedfrom first strand cDNA ofN. americanusL3 by a modifiedRNA ligase-mediated rapid amplification technique (GeneR-acer; Invitrogen) as described previously[21]. TheNa-asp-2gene specific primers used for isolating the 5′ end were Na-ASP2-R2 (TGT CTA GAT CAA GCA CTG CAG AGT CCCTTC TC) and Na-ASP2-R1 (TGT CTA GAG CAC TGCAGA GTC CCT TCT C). The PCR products were ligatedinto the T-ended vector pGEM-T (Promega) according to themanufacturer’s instructions. Recombinant plasmids contain-ing the correct insert were extracted using a Qiaprep SpinMiniprep kit (Qiagen) and submitted for dideoxy terminatorcycle sequencing. DNA and predicted protein sequences wereanalyzed using ESEE version 3.1[22]. Amino acid sequenceswere analyzed for putative signal peptides using signalP(http://www.cbs.dtu.dk/services/SignalP-2.0). Subsequently,an additional round of screening of theN. americanus-L3cDNA library was performed to confirm that there were nootherAc-asp-2cDNA orthologues.

The entire coding sequence minus the N-terminal sig-nal peptide of theNa-asp-2gene was amplified by PCRfae dp zedws e-p

2

tt

sks( redM o-g nd1 with2 tedam eat-s f afim lant(t atchg of theB mo-n rpma t thef ase

ed to the selection of ASP-2 from the human hookw. americanusfor further development. Here we descr

he cloning of theNa-asp-2gene and the scale-up exprion of the purified recombinantNa-ASP-2 protein in theast,Pichia pastoris. We further describe the immunoloal potency ofNa-ASP-2 and the cross-reactivity of antibogainst the recombinant protein with the native, paraerived protein. The studies here also confirm the importf Na-ASP-2 in hookworm tissue migration. These data

he foundation for the clinical development of theNa-ASP-2ookworm Vaccine.

. Materials and methods

.1. Molecular cloning of Na-asp-2 cDNA andransformation into Pichia pastoris

A cDNA library was constructed with mRNA from Lf a Chinese strain ofN. americanus[20], and probed witheterologous cDNA fragment ofAc-asp-2cloned from

he dog hookwormAncylostoma caninum[12]. The Ac-sp-2pBluescript plasmid was cut withXhoI andBamHI,eleasing a 739 bp fragment (28–767 bp). The fragmenel purified (Qiagen) and randomly labeled with�-32[P]-CTP using a Rediprime labeling kit (Amersham). Approately 5× 105 plaques ofN. americanuscDNA library were

creened as previously described[20]. The positive cloneere excised and subjected to secondary screening wiame reagents to isolate single positive colonies. The

rom the first strand cDNA ofN. americanusL3 with Na-sp-2gene-specific primers and subcloned into thePichiaxpression vector, pPICZ�A (Invitrogen, CA) as describereviously[15]. The recombinant plasmids were lineariith SacI digestion and transformed intoP. pastorisX33train as described previously[15–17]. Seed stocks were prared from a colony of the transformedP. pastoris.

.2. Expression of Na-ASP-2 in Pichia pastoris

Expression ofNa-ASP-2 inP. pastoriswas conducted ahe 60 L scale and at the 10 L scale.

For 60 L scale expression, four 2.5 L Tunair shake flaShelton Scientific, CT), each containing 1 L of Buffeinimal Glycerol (BMG) medium (1.34% Yeast Nitren Base, 0.00004% (w/v) d-biotin, 1% (w/v) glycerol a00 mM potassium phosphate, pH 6.0) were inoculatedmL of theP. pastorisseed stock. The flasks were incubat 30± 2◦C for 24–25 h to a final OD600 of 5–15. Approxi-ately 3 L of this culture was used to inoculate 30 L of h

terilized basal salt media (BSM) containing 3.5 mL/L olter-sterilized trace element (PTM4) solution[16,23]. Fer-entation was conducted in a BioFlo5000 Mobile Pilot P

New Brunswick Scientific Co. Inc.). The growth ofP. pas-orisat the 60 L scale was divided into two phases: a fed-blycerol phase and a methanol induction phase. The pHSM was adjusted to and maintained at 5.0 with 14% amium hydroxide feed. The agitation speed was set at 550nd the dissolved oxygen maintained at 30% throughou

ermentation. At approximately 18 h into the glycerol ph

4756 G.N. Goud et al. / Vaccine 23 (2005) 4754–4764

and after a sharp increase in the percentage of dissolved oxy-gen, 50% (v/v) glycerol was introduced into the cell culturemedia at a set flow rate of 15 g/(L h), for 6 h. The pH of the cellculture media was then increased linearly from 5.0 to 6.0 byadding 14% ammonium hydroxide over a 2-h period beforethe completion of the fed-batch glycerol phase. The temper-ature was decreased linearly from 30 to 26◦C also over a 2-hperiod before the completion of the fed-batch glycerol phase.Excessive foaming was controlled with a feed-on-demand of10% (v/v) antifoam KFO673 in DI water (KABO Chemi-cals Inc.). The methanol induction phase was initiated whenthe wet cell weight (WCW) reached approximately 200 g/L.Methanol was added at an initial flow rate of 1.5 mL/(L h)increasing to 11.0 mL/(L h) over an 8-h period, and thensubsequently maintained at a flow rate of 11.0 mL/(L h) foranother 57 h. The fermentation cell culture was diluted from54 to 60 L with 50 mM sodium acetate buffer, pH 4.8 and thenconcentrated to 36 L. The concentratedP. pastoriscells werewashed four times with 36 L of 50 mM sodium acetate buffer,pH 4.8 using a 0.075�m microfiltration hollow fiber cartridge(UFP-750-E-55, Amersham Biosciences) and a Masterflexpump (Cole-Parmer instrument Company). Approximately174 L of supernatant was collected. This supernatant wasconcentrated to 12 L by ultrafiltration using a ProFlux M30Filtration System (Millipore Corporation) and a 3 kDa hollowfiber cartridge (UFP-3C-55-Amersham Biosciences). Thefi 2m

10 LB th-o uper-n ing aFt es),c ands

2

ac witha ead( UV-1 andE cqui-s areI

2P

S FasF Mil-l eo hec Ms m/h.

One liter of the 12 L of total concentrated supernatant fromthe 60 L fermentation (one-twelfth scale) or 1 L of the concen-trated supernatant from the 10 L fermentation (one-half scale)was thawed at 4◦C for 24 h and then at room temperature foranother 6 h. The pH and conductivity were adjusted to 4.8and below 5.5 mS/cm, respectively, and the supernatant wasapplied to the SP Sepharose FF column at a linear flow rate of300 cm/h. The column was washed with 6 column volumesof equilibrating buffer (50 mM sodium acetate buffer, pH4.8) and 16 column volumes of wash buffer (50 mM sodiumacetate buffer, pH 4.8 containing 50 mM NaCl). TheNa-ASP-2 recombinant protein was eluted with 7 column volumesof 50 mM sodium acetate, pH 4.8 containing 200 mM NaCl.The SP Sepharose FF column fractions containingNa-ASP-2recombinant protein (as determined by sodium dodecyl sul-fate polyacrylamide gel electrophoresis [SDS-PAGE]) werepooled and concentrated to 100 mL using a tangential flowfiltration system consisting of masterflex pump, 5 L reservoirand 5 K cut-off Omega membrane (Pall corporation). Theconcentrated protein sample was buffer exchanged with 1 Lof 20 mM ethanolamine buffer, pH 9.5 by tangential flow fil-tration prior to further concentration to a volume of 260 mL.The pH was adjusted to 9.5 with 5N NaOH and conductivitywas measured. Buffer exchange continued until the conduc-tivity reached approximately 0.8–1.4 mS/cm.

2low

( oreV f4 col-u Me /h.T con-d col-u asw ineb Me2 meso MN -n cen-t ions hos-pN erc

2er-

s anda h7 hec BSa ated

nal concentrated supernatant was filtered through a 0.2�membrane and frozen at−70◦C until purification.Fermentation at the 10 L scale was conducted in a

ioflo 3000 (New Brunswick Scientific Co. Inc.) using meds similar to those described above. The cell culture satant was harvested, processed by microfiltration uslex Stand Pilot Scale system and a 0.075�m microfil-

ration (UFP-750-E-6A) cartridge (Amersham Bioscienconcentrated by ultrafiltration (UFP-5C-9G) cartridge,tored in 1 L aliquots at−70◦C until purification.

.3. Protein purification

For purification of the recombinantNa-ASP-2 protein,hromatography instrumentation system was assembledMasterflex Digital standard drive, Easy-Load II pump h

Model# 77200-60-Cole-Parmer instrument Company),Control and Optical units (Amersham Biosciences),zStart Chromatography data system with SS420 data aition box and EZ Chrom Elite software (Scientific Softwnc.).

.3.1. SP-Sepharose FFTo capture theNa-ASP-2 recombinant protein with S

epharose FF, approximately 92.5 mL of SP Sepharoselow (Amersham Biosciences) resin was packed into a

ipore Vantage L column (VL 32× 250) at a linear flow ratf 410-cm/h, following the manufacturer’s instructions. Tolumn was equilibrated with 8.6 column volumes of 50 modium acetate buffer, pH 4.8 at a linear flow rate of 300 c

t

.3.2. Q Sepharose FFApproximately 46.25 mL of Q Sepharose Fast F

Amersham Biosciences) gel was packed into a Millipantage L column (VL 32× 250) at linear flow rate o10 cm/h, following the manufacturer’s instructions. Themn was equilibrated with 6.5 column volumes of 20 mthanolamine buffer, pH 9.5 at a linear flow rate of 300 cmhe eluate was concentrated and adjusted for pH anductivity and was then applied to the Q-Sepharose FFmn at a linear flow rate of 300 cm/h. The column washed with 11 column volumes of 20 mM ethanolamuffer, pH 9.5, followed by 13 column volumes of 20 mthanolamine buffer, pH 9.5 with 50 mM NaCl. TheNa-ASP-recombinant protein was eluted with 10 column volu

f 20 mM ethanolamine buffer pH 9.5, containing 150 maCl. The eluted fractions containingNa-ASP-2 recombiant protein (as determined by SDS-PAGE) were con

rated nine-fold to 50 mL using a tangential flow filtratystem. The concentrate was exchanged with 1 L of phate buffered saline (PBS: 1.04 mM KH2PO4, 2.97 mMa2HPO4·7H2O, 154 mM NaCl, pH 7.4) and then furthoncentrated to 175 mL.

.3.3. Sephadex G-25 FineApproximately 200 g of Sephadex G-25 Fine resin (Am

ham Biosciences) was swollen overnight in 2 L of PBSMillipore Vantage column (VL 44× 500) was packed wit45 mL of this medium at a linear flow rate of 315 cm/h. Tolumn was equilibrated with 2 column volumes of Pt the flow rate of 216 cm/h. The purified and concentr

G.N. Goud et al. / Vaccine 23 (2005) 4754–4764 4757

Na-ASP-2 was applied to a Sephadex G-25 column at alinear flow rate of 216 cm/h and then eluted with 0.4 columnvolumes of PBS. The eluted peak fractions were analyzed bySDS-PAGE. The protein concentration of the pooled frac-tions from the Sephadex G-25 Fine column was determinedby measuring the OD280 and by using the theoretical extinc-tion coefficient calculated for theNa-ASP-2 protein[24].

2.4. SDS-PAGE, N-terminal sequencing, massspectrometry and monosaccharide analysis

SDS-PAGE was performed on 4–20% gradient Tris-glycine polyacrylamide gels using an Xcell mini cell appara-tus (Invitrogen) under denaturing and reducing conditions.The gels were stained overnight with Commassie Bril-liant Blue (CBB) R-250 (Bio-Rad). N-terminal sequencingwas performed by Edman degradation as described previ-ously [16]. A Matrix Assisted Laser Desorption Ionization(MALDI) spectrum of recombinantNa-ASP-2 was recordedon a MALDI-Time of Flight (TOF) Mass SpectrometerModel # AXIMA-CFR (Kratos Analytical) using a linearmode for the detection of positive ions at the ProteomicsCore Facility of The George Washington University (GWU).The instrument was externally calibrated with horse myo-globin. The MALDI matrix was sinapinic acid (10 mg/mL inwater/acetonitrile 1:1 containing 0.1% trifluoroacetic acid).M angeca on,R

2 y(

ro-m and2 izeE XL,r io-s5 tedi f theR witht tions andc ularws witha

2p

toA ofa

solution of the protein. Aliquots (0.5 mL) were collected at0, 5, 10 and 30 min and 1, 2, 4, 7 and 24 h, and centrifugedfor 10 s at 16,000×g to sediment the adsorbed protein. Thesupernatants were retained. To confirm that the protein boundto Alhydrogel® the starting material and unbound super-natants were analyzed by SDS-PAGE and CBB R-250 stain-ing with densitometry using a desktop digital imaging system(Beckman Coulter, CA) and an optical scanner. Desorptionof the Na-ASP-2 recombinant protein from Alhydrogel®

was also performed by mixing equal amounts of formulatedmaterial and 20% sodium citrate (w:v in water). Aliquots(0.5 mL) were taken at 0, 10, 30 and 60 min and centrifugedat 16,000×g for 10 s to sediment the Alhydrogel®. Super-natants were separated by SDS-PAGE and the percent des-orption was calculated following CBB R-250 staining anddensitometry as described above.

2.7. Immunological potency

Immunological potency studies were conducted under theapproval of a GWU Institutional Animal Care and Use Com-mittee (Protocol 24-8,3). Randomly bred Sprague DawleyRats (Harlan Laboratories) were received, acclimated for 7days and examined prior to study initiation. The rats (5 pergroup) were vaccinated by intramuscular injection on days 0,10 and 20. The vaccines consisted of 50�g of recombinantNa wasi sc ela omt hoodt thed ), ont daysf dyw

ectE rbFr pH9 sw n 20( to thep 7f hp ab-o dedt ture.A b-s rk for1 er th of4 singSA stern

onosaccharide analysis was performed by anion exchhromatography with pulsed amperometric detection[25]t the Biologics Testing Facility, BioReliance Corporatiockville, MD.

.5. Size exclusion high pressure liquid chromatographSE-HPLC)

SE-HPLC was performed using a Waters Liquid chatography system (600S Controller, 626 Pump,487 Dual Absorbance Detector) in combination with Sxclusion and Guard Columns—G2000SWXL and SW

espectively (TSK-GEL-HPLC Column, from TOSOH Bcience). Twenty micrograms of purifiedNa-ASP-2 protein in0�L of 50 mM sodium acetate buffer, pH 4.8 was injec

nto the size exclusion column through the aperture oheodine Injector of a Waters 616/626 Pump and eluted

he same buffer at a flow rate of 0.25 mL/min. Qualificatudies determined the range of detection (elution timeshromatograms of the Absorbance at OD280) of moleceight controls. The limit of detection was 0.2�g/50�L ofample, which corresponded to a chromatogram peakmplitude of 2 mV peaks.

.6. Formulation of purified recombinant Na-ASP-2rotein with Alhydrogel®

The recombinantNa-ASP-2 protein was adsorbedlhydrogel® (Superfos Biosector, NY) by adding 1.5 mL3 mg/mL solution of Alhydrogel® to 8.5 mL of a 0.4 mg/mL

a-ASP-2, either bound with Alhydrogel® (300�g) or non-djuvanted in a volume of 0.1 mL; an additional group

njected with Alhydrogel® alone. Adjuvant formulation waarried out as described above by mixing the Alhydrog®

nd the protein and stirring the mixture for 10 min at roemperature. The vaccine was prepared in a laminar flowo assure sterility of the product. Blood was collected onay prior to beginning the vaccination series (T1, day 1

he day of the second vaccination (T2, day 10) and at 10ollowing the final vaccination (T3, day 30), when the stuas terminated.Rat anti-Na-ASP-2 IgG titers were measured by indir

LISA. To coat 96-well microtiter plates (NUNC Maxiso96 Fisher Scientific) with antigen, 1�g/mL solution ofecombinantNa-ASP-2 in 50 mM carbonate bicarbonate,.6 was incubated overnight at 4◦C. After washing, the plateere blocked for 2 h at room temperature with PBS-Twee

0.5%) and washed. Experimental rat sera were addedlates and serially diluted. The plates were incubated at 3◦C

or 3 h and washed before 100�L of 1:1000 horseradiseroxidase (HRP)-conjugated anti-rat total IgG (Zymed Lratories, CA) diluted in PBS/Tween 20 (0.05%) was ad

o the plates prior to incubation for 1.5 h at room temperafter washing, 100�L per well ofo-phenylenediamine sutrate was added to each well and incubated in the da5 min, when 50�L of 2N H2SO4 was overlaid to stop theaction. Colorimetric reaction was read at a waveleng92 nm on a SpectraMax 340 PC (Molecular Devices) uOFTmax Pro for Windows for data capture. Rat anti-Na-SP-2 antibody responses were also analyzed by We

4758 G.N. Goud et al. / Vaccine 23 (2005) 4754–4764

blot using either recombinantNa-ASP-2 orN. americanusL3 lysates as antigen as described previously[14], but usingpooled rat anti-Na-ASP-2 antiserum at a dilution of 1:4000and goat anti-rat IgG-HRP (Zymed Laboratories) diluted to1:2000.

2.8. N. americanus larval migration inhibition assay

Larval migration inhibition assays were adapted from apreviously described method[26]. Briefly, fresh hamster skinwas shaved, and approximately 4 cm2 section of skin wasstretched and sandwiched between 2× 20 mL syringe bar-rels that were clamped together with bulldog clips. The lowersyringe was filled to the top with PBS so that the buffer wasin contact with the underside of the skin.N. americanusL3,obtained from infected hamsters, were incubated with pooledsera (neat) from either rats immunized withNa-ASP-2 for-mulated with Alhydrogel® or rats injected with Alhydrogel®.The sera were obtained from the rats used in the immunolog-ical potency study described above.

For these studies, the L3 (300 L3/group) were then incu-bated in 0.05 mL of undiluted serum pooled from either theNa-ASP-2-immunized rats or Alhydrogel®-injected rats for1 h at 37◦C. The L3 together with the antiserum were thenapplied to hamster skin to observe whether serum antibod-ies interfered with the migration in vitro. Each group of L3w ed tom d ont mov-i skint reed ltingm esst totaln liedb

3

3

eac ineda f 206aTn -tp thec p)c ofh nalc y att -

tial screening. Minor nucleotide mutations (eight changes)were found among the 14Na-asp-2sequences. Each was asilent substitution except for the nucleotides at position 119.Ten of the 14 different sequences exhibited a T at position119, while the other four exhibited an A; this resulted in anamino acid substitution from Leu-36 to Met-36 in the pre-dicted open reading frame.

The 5′ end ofNa-asp-2was isolated fromN. americanusL3 cDNA by rapid amplification of cDNA ends (RACE) usingthe GeneRacer kit and 5′ GeneRacer primer. The full-lengthNa-asp-2cDNA consisted of 753 bp encoding an ORF of209 amino acids and a poly-A tail at the 3′ end. There was ahydrophobic signal peptide sequence with a signal peptidasecleavage site between amino acids 18 and 19. The predi-cated ORF of matureNa-asp-2(without signal peptides) hada calculated molecular mass of 20,621 Da, and a predicted pIof 8.00. There were no putative N-linked glycosylation sitesdetected in the sequence. The nucleotide and deduced aminoacid sequence ofNa-asp-2is available in GenBank underaccession number AY288089.

Amino acid sequence comparisons revealed thatNa-ASP-2 exhibited 58–65% identity to other ASP-2 molecules fromother hookworms of the genusAncylostoma[16]. Comparedto ORFs ofAncylostomaspp. ASP-2 molecules,Na-ASP-2contains fewer amino acids including an eight amino acidsdeletion at the C-terminus, one amino acid deletion at posi-t 147.T inp2 aD PRPp

3

ec-t og2 a nots t cellw teiny icald ere,w romt

fp g/La litero eda s, andw ion-e g onS ice groups ial,t inor

as then placed on the upper side of the skin and allowigrate for 30 min at room temperature. L3 that remaine

he surface of the skin were collected and counted by reng the remaining liquid with a pipette and washing thehree times with 0.5 mL of PBS. The number of L3 in thifferent aliquots of 0.05 mL was counted, and the resuean was multiplied by a dilution factor of 30. To expr

he percentage of larval inhibition by the antiserum, theumber of L3 recovered was divided by 300 and multipy 100. Each assay was performed in triplicate.

. Results

.1. Identification and cloning of Na-asp-2

Two positive clones were obtained by screening thN.mericanusL3 cDNA library with the radiolabeledAc-asp-2DNA (28–767 bp) heterologous probe. Each clone conta731 bp cDNA encoding an open reading frame (ORF) omino acids with 65% amino acid identity withAc-ASP-2.heNa-asp-2cDNA sequence contained a 3′ poly(A) tail buto translation initiation codon (ATG) at the 5′-end. To iden

ify additionalNa-asp-2cDNAs, further screening of 5× 105

laques was undertaken using a probe designed fromonserved regions ofAc-asp-2(87–275 bp and 465–598 bompared with otherasp-2genes from different speciesookworm. This resulted in the identification of 14 additiolones, each of which was nearly identical (99% identithe nucleotide level) with theNa-asp-2cloned during the ini

ion 21 and two amino acids deletion at position 146–herefore, theNa-ASP-2 is approximately 1.0 kDa lessredicted molecular mass relative to theAncylostomaASP-molecules. In addition,Na-ASP-2 was noted to containSG signature sequence that was found previously in arotease from cone snails[27].

.2. Expression and purification in Pichia pastoris

TheNa-asp-2cDNA was cloned into the expression vor and transformed intoP. pastoris, which was used tenerate seed stocks for expressing recombinantNa-ASP-at the 10 L scale and the 60 L scale, respectively (dat

hown). Fermentation of the seed stocks resulted in weeights ranging from 408 to 466 g/L and purified proields of 0.24–0.40 g/L (data not shown). Only biochemata resulting from the 60 L fermentation is reported hhereas immunological potency data will be reported f

he 60 L fermentation and a 10 L fermentation.After 65 h of methanol induction the wet cell weight oP.astoriscells from the 60 L scale fermentation was 454nd yielded 0.4 g of purified recombinant protein perf BSM. The recombinantNa-ASP-2 protein was expresss a secreted recombinant protein in yeast supernatantas purified using three major steps comprised of twoxchange chromatography steps followed by desaltinephadex G-25 (Fig. 1). Chromatography with the cationxchange resin SP Sepharose FF was effective in theeparation ofNa-ASP-2. Relative to the starting materhe recombinant protein eluted together with only m

G.N. Goud et al. / Vaccine 23 (2005) 4754–4764 4759

Fig. 1. Column elution profiles during the purification of recombinantNa-ASP-2: (a) SP Sepharose FF; (b) Q Sepharose FF; (c) Sephadex G-25 Fine, duringthe purification ofNa-ASP-2. (d) SDS-PAGE analysis of the column eluates duringNa-ASP-2 purification: lane 1, SeeBlue® Plus2 Pre-Stained Standard(Invitrogen); lane 2, concentrated supernatant from the 60 L fermentation; lane 3, SP Sepharose FF elution with 50 mM sodium acetate, pH 4.8 with 200 mMNaCl; lane 4, Q-Sepharose FF elution with 20 mM ethanolamine buffer pH 9.5 with 150 mM NaCl; lane 5, Sephadex G-25 Fine elution with PBS, pH 7.4.

impurities from SP Sepharose using 200 mM NaCl in astepwise gradient (Fig. 1(a) and (d)). Further purification wasachieved with anionic chromatography on Q-Sepharose FF(Fig. 1(b) and (d)), which resulted in highly enriched recom-binant protein (Fig. 1(c) and (d)). Following purification,Na-ASP-2 was the only protein detected by SDS-PAGE and CBBstaining, migrating with an apparent molecular weight ofapproximately 21 kDa. In order to estimate its purity, the pro-tein was subjected to SE-HPLC that was qualified previouslyto a limit of detection of 0.2�g. This detection limit corre-sponds to the appearance of a chromatogram peak equivalentto 2 mV at OD 280. Based on the calculated limits of detec-tion it was determined that the recombinantNa-ASP-2 wasgreater than 95% pure (Fig. 2(a) and (b)). Additional studiesusing SDS-PAGE followed by silver staining confirmed theabsence of additional contaminants (data not shown).

3.3. Biochemical characterization

The recombinant protein was expressed with an addi-tional 6 amino acid EAEAEF N-terminal sequence, whichcorresponds to pPICZ� vector sequence when theEcoR1site is used for cloning. It was noted that the addition ofthe sequence improved the yield of the secreted recombinantprotein (data not shown). N-terminal amino acid sequencingo onr pre-d mid

vector (after cleavage of the�-mating factor signal pep-tide) and GCPD represents the start of theNa-ASP-2 matureprotein.

The purified recombinant protein with the 6 amino acidvector tag has a predicted molecular mass of 21.3 kDa(21,298 Da), a finding in agreement with its apparent migra-tion on SDS-PAGE as well as mass spectrometry. TheMALDI-TOF spectrum of the recombinantNa-ASP-2 pro-tein obtained revealed a single major species that corre-sponded to a molecular mass of 21.3 kDa and a minor specieswith an additional 161 Da of mass (Fig. 3). Since there are noN-linked glycosylation sites on theNa-ASP-2 ORF, the sec-ond species could correspond to a protein with an O-linkedmonosaccharide. This was confirmed by carbohydrate analy-sis that revealed the presence of 0.483 moles of mannose permole of recombinantNa-ASP-2 (Table 1).

Table 1Analysis of monosaccharides in purified recombinantNa-ASP-2

Sugar Quantity (mole/mole)a

Fucose 0Galactosamine 0Glucosamine 0Galactose 0Glucose 0Mannose 0.483b

sam-p

f the first 10 residues ofNa-ASP-2 by Edman degradatievealed the sequence EAEAEFGCPD, confirming theicted sequence where EAEAEF is derived from the plas

a Mole/mole = pmol of each monosaccharide observed/total pmol perle injected.b 113.3 pmol/234.8 pmol.

4760 G.N. Goud et al. / Vaccine 23 (2005) 4754–4764

Fig. 2. HPLC profile after purification of recombinantNa-ASP-2. Panel A shows the HPLC chromatogram of 50�g (total) of Gel Filtration Standards (Bio Radcat# 151-1901) comprised of five proteins, run at 0.25 mL/min. The profile denotes six peaks of which molecular weights can be determined. From left to right,peak 1: protein aggregates, peak 2: thyroglobulin (670 kD), peak 3: gamma globulin (158 kD), peak 4: ovalbumin (44 kD), peak 5: myoglobin (17 kD), peak 6:Vitamin B12 (1.35 kD). Panel B shows 20�g of Na-ASP-2 run on same size exclusion column at 0.25 mL/min for 55 min. The elution time was determinedto be between 38 and 42 min. This corresponds as expected between peak 4 (44 kD) and peak 5(17 kD) of the gel filtration standards shown in panel A. Thisqualified assay was shown to have a sensitivity of 0.2�g equivalent to a 2 mV peak (data not shown).

G.N. Goud et al. / Vaccine 23 (2005) 4754–4764 4761

Fig. 3. MALDI-TOF spectrum ofNa-ASP-2. The single ion species exhibited a molecular mass of 21,311.3. Two double ion species of 10,653.8 and 10,734.2 Dawere also noted which correspond to molecular mass of 21,307.6 and 21,468.4 Da, respectively, as well as a triple ion species of 7106.7, which corresponds toa molecular mass of 21,320.1 Da.

3.4. Rat immune responses to vaccination

The adsorption of recombinantNa-ASP-2 to Alhydrogel®

was greater than 75%. The protein remained adsorbed toAlhydrogel® for up to 24 h since no unboundNa-ASP-2 wasdetected in the supernatants of the formulation when analyzedby SDS-PAGE (data not shown). To determine immunolog-ical potency, Sprague Dawley rats (n= 5) were vaccinatedthree times at 10-day intervals with a dose of 50�g Na-ASP-2, with or without adjuvant (Alhydrogel®, 300�g). Thecontrol group was vaccinated with the same dose of adju-vant. As shown inTable 2, vaccination of rats with twodifferent batches of the Alhydrogel-formulated recombinantprotein induced a marked antibody production (total IgG)with geometric mean titers ranging from 1:9784 to 1:20,949after two doses, and 1:97,533 to 1:234,881 after three doses,

Table 2Mean total IgG titers (determined by ELISA) from sera of rats (n= 5) vacci-nated three times (T1, day 0; T2, day 10; T3, day 20) at 10-day intervals withthree different lots of theNa-ASP-2 Hookworm Vaccine containing 50�gNa-ASP-2 and 300�g Alhydrogel®

Fermentation (L) T1 (day 0) T2 (day 20) T3 (day 30)

10 <100 9,784 234,88160 <100 20,949 97,533

Fig. 4. Identification of nativeNa-ASP-2 by Western blot with pooled ratantiserum raised against recombinantNa-ASP-2. Twenty-five nanograms ofrecombinantNa-ASP-2 (lane 1) and approximately 10N. americanusL3 in4�L (lane 2) were homogenized in SDS-PAGE sample buffer, separated bySDS-PAGE and transferred to a PVDF membrane. The Western blot wasprobed with a 1:4000 dilution of pooled rat anti-Na-ASP-2 antiserum (A).The same rat antiserum absorbed with recombinantNa-ASP-2 was used asa control serum (B). Arrows point to the putativeNa-ASP-2.

4762 G.N. Goud et al. / Vaccine 23 (2005) 4754–4764

Fig. 5. Inhibition ofN. americanusL3 hamster skin migration by rat anti-Na-ASP-2 antiserum. For these studies, 300N. americanusL3 were appliedin buffer and the number of non-penetrating L3 was counted. To expressthe percentage of larval inhibition by the antiserum, the total number ofL3 recovered was divided by 300 and multiplied by 100. The percentageinhibition resulting from either rat-anti-Na-ASP-2 antiserum or antiserumobtained from Alhydrogel®-injected rats is expressed as the mean± standarddeviation of assays performed in triplicate.

respectively. In contrast, immunization with the recombi-nant protein alone (unadjuvanted protein) elicited IgG titersless than 1:100 (data not shown), suggesting that the pres-ence of Alhydrogel® is necessary for the hookworm vac-cine to be immunogenic. Antiserum from rats injected withAlhydrogel® alone also contained IgG titers less than 1:100(data not shown). On western blots the rat anti-Na-ASP-2antiserum recognized both the recombinant protein as well asa band of approximately 21 kDa inN.americanusL3 extracts,which corresponds to the native protein (Fig. 4). As shownin Fig. 5, the rat anti-Na-ASP-2 antiserum inhibited larvalskin migration 90± 7% in three different experiments com-pared with 17± 7% inhibition by serum from rats injectedwith Alhydrogel® (P= 0.019).

4. Discussion

Here we describe the expression, purification and charac-terization ofNa-ASP-2, a lead candidate antigen under devel-opment as a human hookworm vaccine. We also show thatanti-Na-ASP-2 antibody inhibits larval migration in vitro,providing evidence for the role of this protein in host tissuemigration byN. americanusL3.

Phylogenetic analysis of the four hookworm ASP-2molecules studied to date (one fromN. americanusand threeft y oft sH tri-cf tri-c nailP P14

plant PRP, which has antifungal properties[29], no func-tion has been assigned to any of the PR-1 family proteins.The cone snail protease contains a DTG signature sequence,which is believed to have a pivotal role in enzyme catalysis[27].Na-ASP-2 contains a DSG sequence, in which a serinemay possibly substitute for a catalytic threonine. Althoughno protease activity forNa-ASP-2 has been detected so far(data not shown), our finding that anti-Na-ASP-2 antibodyinhibits larval skin migration, suggests that the molecule hasan important role in larval host entry and in migration throughthe tissues and pulmonary capillaries before reaching theintestine. This could provide the basis by which anti-ASP-2antibodies are protective against larval challenge infections[16,17].

Hookworm infection is a disease of the world’s poorest,occurring almost exclusively in areas with poor sanitation andhigh levels of poverty[1,19]. Such a small commercial mar-ket demands that the production of recombinant hookwormantigens formulated for vaccines must be made in high yieldand at low cost[19]. Although employment ofEscherichiacoli for expression of hookworm antigens ordinarily wouldmeet these criteria, we have so far been unable to refold or sol-ubilize either ASP-1 or ASP-2 fromE. coli inclusion bodies[10]. Other investigators have also reported failure to expresssoluble PR-1 proteins[30], most likely because the high cys-teine content causes improper protein folding secondary toa s1

s fold-i ressN ter-m EFs d intot minoa byi ed aset

asa step.S iredo stepu hadexG sis-t en-t in ah pro-t nce,i rity( gni-t icalp erentp ortedh rredt ould

rom different members of the genusAncylostoma) revealshat they constitute a unique cluster of the PR-1 familhe PRP superfamily[12,14,16]. The cluster also includec24, a protective vaccine antigen from the ruminanthonstrongyle nematode,Haemonchus contortus[14,28]. Nounction has yet been assigned to any hookworm orhostrongyle ASP-2. With the exception of the cone-sRP[27], which was revealed to be a protease, and the

berrant disulfide bond formation[10]. Na-ASP-2 contain1 cysteines.

Expression ofNa-ASP-2 in P. pastoris resulted in aecreted product that was soluble and did not require reng. Codon optimization was not required in order to expa-ASP-2 in high yield (data not shown). It was also deined that the addition of an acidic N-terminal EAEA

equence increased the amount of ASP-2 protein secretehe fermentation broth. This sequence represented six acids of pPICZ�A vector sequence that was introduced

ncorporating anEcoRI restriction site into the cloning. Wo not know if the addition of this sequence would incre

he yield of otherP. pastoris-expressed proteins.Capture of theNa-ASP-2 recombinant protein w

chieved by an efficient SP Sepharose FF purificationubsequent elution of the highly enriched protein requnly an additional anion-exchange chromatographysing Q-Sepharose FF and a desalting step with Sep-25 Fine in order to obtain highly pure protein of con

ently high quality. Further studies revealed that the fermation and purification process was robust and resultedighly uniform, consistent, pure and stable recombinant

ein. Using qualified assays for pH, color and appearadentity (SDS-Silver Stain and Western Blots), and puSE-HPLC and SDS-PAGE Silver Stain), immune recoion (Indirect ELISA and Western Blots) and immunologotency, no major differences were detected among diffilot lots. The process development technologies repere are sufficiently simple so that they could be transfe

o manufacturers in middle-income countries. These w

G.N. Goud et al. / Vaccine 23 (2005) 4754–4764 4763

include countries where hookworm infection is endemic suchas Brazil, India, and China[19].

Technology transfer to a manufacturing facility willbe followed by toxicity and stability studies, and then adose-escalating Phase 1 human trial to evaluate safety,tolerability, and immunogenicity of theNa-ASP-2 Hook-worm Vaccine. To our knowledge this will be the firstclinical trial attempted for a vaccine against any humannematode. Ultimately, immunization with theNa-ASP-2Hookworm Vaccine would be indicated for the prevention ofhookworm disease caused by high intensity infection withN. americanus. Vaccination would be expected to reduce thenumber of infective hookworm L3 that enter into the humangastrointestinal tract and become adult worms. This wouldreduce both the number of hookworms and the fecal eggcounts in individuals exposed to L3 living in the soil.

Acknowledgements

We wish to thank Drs. Allan Saul and Louis Miller of theMalaria Vaccine Development Branch of the NIAID, NIH forallowing us to use their fermentation facilities. These studieswere supported by the Human Hookworm Vaccine Initiativeof the Bill and Melinda Gates Foundation and the Sabin Vac-cine Institute. JB is supported by an International ResearchS iss romt us-t keF nto,C .

R

the

A,ed

violiure.

his-HO

912.

D,forg-–9.ithazoleMed

y-om-

tar-rgan

[8] Chan MS, Bradley M, Bundy DA. Transmission patterns and theepidemiology of hookworm infection. Int J Parasitol 1997;26(6):1392–400.

[9] Miller TA. Industrial development and field use of the canine hook-worm vaccine. Adv Parasitol 1978;16:333–42.

[10] Hotez PJ, Zhan B, Bethony JM, et al. Progress in the developmentof a recombinant vaccine for human hookworm disease: the HumanHookworm Vaccine Initiative. Int J Parasitol 2003;33(11):1245–58.

[11] Hawdon JM, Jones BF, Hoffman DR, Hotez PJ. Cloning and charac-terization ofAncylostoma-secreted protein. A novel protein associ-ated with the transition to parasitism by infective hookworm larvae.J Biol Chem 1996;271(22):6672–8.

[12] Hawdon JM, Narasmihan S, Hotez PJ. Ancylostoma secreted protein2: cloning and characterization of a second member of a fam-ily of nematode secreted proteins fromAncylostoma caninum. MolBiochem Parasitol 1999;99(2):149–65.

[13] Hawdon JM, Hotez PJ. Hookworm: developmental biology of theinfectious process. Curr Opin Genet Dev 1996;6(5):618–23.

[14] Zhan B, Liu YY, Badamchian M, Williamson A, Feng JJ, Loukas A,et al. Molecular characterization of the Ancylostoma-secreted proteinfamily from the adult stage ofAncylostoma caninum. Int J Parasitol2003;33(9):897–907.

[15] Asojo OA, Goud G, Dhar K, Loukas A, Zhan B, Deumic V, et al.Novel X-ray structure ofNa-ASP-2, a pathogenesis related-1 proteinfrom the nematode parasite,Necator americanus, and a vaccine anti-gen for human hookworm infection. J Mol Biol 2005;346:801–14.

[16] Goud GN, Zhan B, Ghosh K, et al. Cloning, yeast expression,isolation, and vaccine testing of recombinant Ancylostoma-secretedprotein (ASP)-1 and ASP-2 fromAncylostoma ceylanicum. J InfectDis 2004;189(5):919–29.

[17] Mendez S, Zhan B, Goud G, Ghosh K, Dobardzic A, Wu WH,etedam-

[ otezs in

t Rev

[ m—

[ on-e of–9.

[ ,iated

[ leiciosci

[ ion.

[ sureSci

[ s ofulsed

[ C,con-and

[ so-logyily. J

cientist Development Award (1K01 TW00009) and ALupported by an RD Wright Career Development Award fhe National Health and Medical Research Council of Aralia. We would like to acknowledge Andrew Murdin, Miletcher and Annie Dookie from Aventis Pasteur, ToroA for helping optimize theNa-ASP-2 vaccine formulation

eferences

[1] Brooker S, Bethony J, Hotez PJ. Human hookworm infection in21st century. Adv Parasitol 2004;58:197–288.

[2] Hotez PJ, Brooker S, Bethony JM, Bottazzi ME, LoukasXiao SH. Current concepts: hookworm infection. N Engl J M2004;351(8):799–807.

[3] deSilva NR, Brooker S, Hotez PJ, Montresor A, Engels D, SaL. Soil-transmitted helminth infections: updating the global pictTrends Parasitol 2003;19:547–51.

[4] WHO (World Health Organization). Prevention and control of sctosomiasis and soil-transmitted helminthiasis. Report of a WExpert Technical Committee. WHO Technical Report SeriesGeneva: WHO; 2002. p. 33–5.

[5] Savioli L, Stansfield S, Bundy DA, Mitchell A, Bhatia R, Engelset al. Schistosomiasis and soil-transmitted helminth infections:ing control efforts. Trans R Soc Trop Med Hyg 2002;96(6):577

[6] Albonico M, Smith PG, Ercole E, et al. Rate of reinfection wintestinal nematodes after treatment of children with mebendor albendazole in a highly endemic area. Trans R Soc TropHyg 1995;89(5):538–41.

[7] Albonico M, Bickle Q, Ramsan M, Montresor A, Savioli L, Talor M. Efficacy of mebendazole and levamisole alone or in cbination against intestinal nematode infections after repeatedgeted mebendazole treatment in Zanzibar. Bull World Health O2003;81(5):343–52.

et al. Effect of combining the larval antigens Ancylostoma secrprotein 2 (ASP-2) and metalloprotease 1 (MTP-1) in protecting hsters against hookworm infection and disease caused byAncylostomaceylanicum. Vaccine 2005;23(24):3123–30.

18] Brooker S, Bethony JM, Rodrigues L, Alexander N, Geiger S, HPJ. Epidemiological, immunological and practical considerationdeveloping and evaluating a human hookworm vaccine. ExperVaccine 2005;4(1):35–50.

19] Hotez PJ, Bethony J, Bottazzi ME, Brooker S, Buss P. Hookworthe great infection of mankind. PLoS Med 2005;2(3):67.

20] Zhan B, Hawdon J, Shan Q, Ren H, Qiang H, Xiao SH, et al. Cstruction and analysis of cDNA Library of the third stage larvaNecator americanus. Chin J Parasitol Parasitic Dis 2000;18(1):26

21] Zhan B, Wang Y, Liu Y, Williamson A, Loukas A, Hawdon JMet al. Ac-SAA-1, an immunodominant 16 kDa surface-assocantigen of infective larvae and adults ofAncylostoma caninum. IntJ Parasitol 2004;34(9):1037–45.

22] Cabot EL, Beckenbach AT. Simultaneous editing of multiple nucacid and protein sequences with ESEE. Comput Appl B1989;5(3):233–4.

23] Stratton J, Chiruvolu V, Meagher M. High cell-density fermentatMethods Mol Biol 1998;103:107–20.

24] Pace CN, Vajdos F, Fee L, Grimsley G, Gray T. How to meaand predict the molar absorption coefficient of a protein. Protein1995;4(11):2411–23.

25] Hardy MR, Townsend RR, Lee YC. Monosaccharide analysiglycoconjugates by anion exchange chromatography with pamperometric detection. Anal Biochem 1988;170(1):54–62.

26] Williamson AL, Brindley PJ, Abbenante G, Prociv P, BerryGirdwood K, et al. Hookworm cathepsin D aspartic proteases:tributing roles in the host-specific degradation of serum proteinsskin macromolecules. Parasitology 2003;126:179–85.

27] Milne TJ, Abbenante G, Tyndall JDA, Halliday J, Lewis RJ. Ilation and characterization of a cone snail protease with homoto CRISP proteins of the pathogenesis-related protein superfamBiol Chem 2003;278(33):31105–10.

4764 G.N. Goud et al. / Vaccine 23 (2005) 4754–4764

[28] Kooyman FN, Schallig HD, Van Leeuwen MA, MacKellar A,Huntley JF, Cornelissen AW, et al. Protection in lambs vaccinatedwith Haemonchus contortus antigens is age related, and correlateswith IgE rather than IgG1 antibody. Parasite Immunol 2000;22(1):13–20.

[29] Niderman T, Genetet I, Bruyere T, Gees R, Stintzi A, Legrand M, etal. Pathogenesis-related PR-1 proteins are antifungal. Isolation and

characterization of three 14-kilodalton proteins of tomato and of abasic PR-1 of tobacco with inhibitory activity against Phytophthorainfestans. Plant Physiol 1995;108(1):17–27.

[30] Monsalve RI, Lu G, King TP. Expressions of recombinant venomallergen, antigen 5 of yellowjacket (Vespula vulgaris) and paperwasp (Polistes annularis), in bacteria or yeast. Protein Expr Purif1999;16(3):410–6.