Molecular and Biochemi~ al Parasttoloey. 45 ( 1991 ~ 137- I-1-6

137

Elsevier MOLBIO I)1476

Molecular cloning of Taenia taeniaeformis oncosphere antigen genes Wendy G. Cougle *t, Marshall W. Lightowlers, Henrik O. Bogh, Michael D. Rickard*" and Kevin S. Johnson .3 L,'mver,~ity o] Melbourne, ~etettnarv Clint~ al Centre. Il errthee, I ictoria. Australia IRecei~ed 17 April 1990; accepted 3 October 1990)

Infection of mice ~ ith the ceqode Taeniu taemaeformis exhibits se~ eral ~mportant features common to other eestode infections. including the abdit.v to ,.acemate ',~lth crude antigen mixtures [11. Partml purification of the protective oneosphere antigens has been reported ~ith a cutout from deoxycholate [DOCI acr3.1amide gels: this cutout ~as called fraction II (FII), and comprises approximatel~ 10% of total DOC-soluble oncosphere antigen. Western blots of DOC gels probed ~ith anti-Fl[ antisera revealed a series of 3-5 discrete bands ~ithin the FII region [21. Further fractionation of the FII antigens on DOC gels v,as impractical due to limitations in suppl 3 of oncospheres, so a eDNA hbrary ~as constructed from 150 ng of oneosphere mRNA and screened v, ith ~t-Fll antlsera. T~o distinct clone famihes were identified, omA and omB. Antibodies affinity-purified on either of two representative members, oncAI and o m B I . recognised all the FII bands. Individual FII bands e~clsed from a DOC gel resolved into an overlapping series of molecules when re-run on SDS-PAGE, indicating that each FII band consisted of several polypeptides of diffenng molecular weight. Immunoprecip~tates resolved on SDS-PAGE revealed that o-FIl recognised 3 major oncosphere antigens, of 62. 34 and 25 kDa: antisera against omB precipitated both the 34- and 25-kDa antigens. ,.~hereas (~-oncA antisera precipitated the 62-kDa antigen. We conclude that oncA and omB encode the maior antigens in the FII complex. The 62-kDa antigen encoded b~ oncA I was the onl~ common anugen precipitated by anti-Fll and tv.o other antisera raised against different protective extracts, suggesting that it may be a protective component in all three. Southern blot results indicate that oncA and omB are distinct genes present at Io~ cop', number in the genome. E~idence i,,, also presented suggesting that some cestode mRNAs. including oncA, ma.~ u~e ~ariant polyadenylatmn ,qgnal~. Ke.~ ~ord~: Taemu taemae/iwmt6: Oncosphere antigen: Fraction II: eDNA; DNA ~equence; Polyadenylation

Introduction

Taenia taenia~'ormis, the naturally occurring parasite of mice and rats, is a good model system for studying immunity to larval cestode infection C~,rrespondeme address." K.S. Johnson, Cambridge Antibody Technolog.,, Ltd.. The Dal} Research Laboratories, Babraham Hall, Babraham, Cambridge CB2 4AT, LI.K. *Pre~ent addresses. IA.M.R,A.D. Laboratories, 3 Guest Street. Ha~thom, Vic. 3122. Australia: -'C.S.I.R.O. Division of Animal Health, Park~ ille, Vic. 31)52. Australia. and ~Cambndge Antibod.,, Technology Ltd.. The Daly Research Laboratories. Babraham Hall. Babraham, Cambridge. CB2 4AT. U.K. Ahhre~'iations: DOC, deox.~cholate: FII, fraction It: PAGE. polyacrylamide gel electrophoresis. Note Nucleotide sequence~ described in this paper ha~e been submitted to the GenBank TM data base ~ith the accession number~: M38397-ONC A and M38398-ONC B.

in man and his domestic animals [1]. Immunity in the intermediate host is central to control of transmission, and considerable effort has been directed towards production of defined antigen vaccines based on oncosphere and metacestode material. Recently a recombinant antigen has been used to successfully vaccinate sheep against Taenia or'is infection [3]. While this represents a major breakthrough in development of defined vaccines against taeniid cestodes, it is not a host/parasite system which lends itself readily to the identification of protective immune mechanisms or optimising antigen delivery. The T. taeniaeformis mouse system would be more suitable for these purposes. Bowte[l et al. [4] used sera from animals vaccinated with homogenised oncospheres to isolate expressing clones from a metacestode eDNA libra,'. However. there was little evi-

11166-6851/91/$1)3.50 ,'.cS, 1991 Elsevier Science Publishers B.V. (Biomedical Division)

13s dence to implicate these antigem, as being hostprotectixe, and none of the recombinant antigens protected nlice. Recent e~ idence ~,uggests that oncosphere and metacestode extracts produce different types of immunity: whereas an oncosphere vaccine prevented establishment, the metacestode extract stimulated immunit~ that acted against survival of established larxae [5]. Oncospheres are a source of potent vaccinating antigens in T. taenia@~rmi.s [6,7,9] as well as in other Taeniid cestodes [I,3]. and are therefore a good starting point fcw developing defined antigen x accines. One aspect that has hindered progress in this area has been the difficult~ in purif 5 ing hostprotective antigens from complex mixtures, part[ 5 because of shortage of material, and, in the case of T. taeniaeformis, because the host-protective antigenl s~are denaturation-sensitive [9]. The most x~ell-defined oncosphere antigen ~accine against T. taeniaeformis infection is a cutout of a deox.~cholate ~DOCI-PAGE gel tenned "fraction II" FII~. which comprises approx. 10°/- of total DOCsoluble antigen and confers 70-80c? protection in mice [2]. Sera from animals vaccinated ~ith FII unique b recognise a series of 3-5 bands from the FII region on DOC-PAGE Western blots. These same bands are also uniquely recognised by sera from mice xaccinated x~ith more complex antigen mixture.,, such as total DOC-solubilised oncosphere antigen ITtO-DOCI and homogenised oncospheres ITtOl. suggesting that FII could be enriched for the ,,ame protective antigemsl found in more complex vaccine preparations. Attempts to further subdivide the FII region on DOC-PAGE haxe been unsucce.,,sful. Here x~e report construction of a cDNA librar.s using mRNA from hatched and activated oncospheres. The library x~as screened uith an antiserum against FII. and the positive clones used to investigate the makeup of the FII complex and the nature of the FII antigens. Materials and Methods

E.v~erimental animals and parasites.

The parasite life cycle was maintained in cats and mice as previously described [2].

Antigen preparation. Eggs from freshly-purged

tapex~orms were NaOCI hatched within 72 h of collection, and centrifuged over a Percoll cushion to separate oncospheres from the ,,hell blocks ~ref. 9. and references therein~. Oncosphere, ~ere disrupted b.s sonication in 21/mM Tri,,-HCI ~pH g.()) containing I% x~/v DOC and protease inhibitors exacth as previousl.\ described [91.

Iodination. Homogenates ~ere centrifuged at 100000 x .~ for 30 rain at 4¢C and the soluble extract diabsed against 1000 volume.,, of mouse tonicity PBS for 24 h at 4':'C prior to iodination with lodo-Gen exactly as described by Box~tell et al. [8]. Briefly. 50 111 of DOC-soluble antigen 250000 oncosphere equixalents/mlJ ~as added to 150 itl PBS/10 pl 0.25 mM KI and 200 pCi Na~-'~l. This ~a~ transferred to a glass tube that had been coated with 50 Itg lodo-Gen ¢Pierce Chemical Co.I. After 10 rain on ice. the reactants were transferred to a fresh tube containing 3¢)0 pl PBS and the reaction hahed b.~ addition of 10 Itl of 5 nlM KI. Labelled antigen was desalted by passage over a PDI0 column equilibrated with 0.125% gelatin and used immediateh.

ImmunOlm'Cilfitation.~.

These ~ere peH'om~ed using protein A-Sepharose as described b~ Kee et al. [10].

PAGE aml Western blots. DOC-PAGE ~as carried out as described b s Lightowlers et al. [2]. Stacking and resolving gel buft'ers were prepared as for SDS-PAGE, but ~ithout detergent. DOC 110% ~/~ in distilled ~ater~ ~as added to 0.1c~final immediatel~ before use. Electrode buffer ~ as the same as that used for SDS-PAGE except that DOC ~as substituted for SDS. SDS-E~GE and Western blotting ~ere performed using standard techniques [2].

Genomic DNA preparation. Genomic DNA ~as prepared from 5-month-old larvae. These ~ere ground to a paste in a pestle and mortar in liquid nitrogen and the powder transferred to a 50-ml pobpropylene tube IFalconl. Ten volumes of homogenisation buffer t l0 mM Tris-HCI 20 mM EDTA, pH 8.01 was added, and the tube ,,ortexed brieflx to disperse the tissue. After adding 1/10 vol. of 20% Sarkosyl ~Sigma, L 5125~ the tube

139 was inverted several times until the suspension became viscous. 0.94 g dr3' CsCI (BRL Ultrapure) and 25 ltl of 10 mg ml -~ ethidium bromide (Boehringer) was added per ml of extract, and centrifuged at 130000 × g for40 h at 15°C. The genomic DNA band was harvested with a widebore needle, dialysed against several changes of TE ~10 mM Tris-HCI, pH 8.0, 1 mM EDTA) and stored at 4°C. A second band was always observed after CsCI centrifugation of larval or adult worm extracts. This had a white opaque appearance and was denser than gDNA. Although this material precipitated with ethanol, it did not absorb light at ultraviolet wavelengths. From its intense staining with iodine we believe it is largely carbohydrate and probably glycogen.

Construction (~'oncosphere cDNA library. Oncospheres were activated in artificial intestinal fluid [12]. These were sterilised by sedimentation in 5 × 50 ml of 37~C RPMI, and cultured for 4 h at 37~C in a 5% CO, atmosphere prior to washing in sterile saline and storage in liquid nitrogen. 7 x 106 pooled oncospheres were ground to a paste in liquid nitrogen and RNA extracted by the GuHCI method [13], Total RNA (20 ,¢tg) was pudried on a 0.5 ml oligo-dT column IBRL) [14] rather than being used directly for cDNA synthesis. This purification step was used to reduce the number of clones arising from self-priming of ribosomal RNA and because we have found that this procedure is biased towards purification of intact mRNAs (unpublished observations). PolylA) + RNA (150 ng) was converted into dscDNA (70 ng) using the Amersham cDNA synthesis system (RPN 1256) and methylated with 20 U EcoRI meth.vlase + S-adenosyl methionine (Ne~ England BioLabs) for 3 h. After phenol/chloroform extraction and ethanol precipitation, cDNA was ligated to an equal mass of phosphorylated linkers (pGGAATTCC, New England BioLabs) seeded with unphosphorylated linkers that had been treated with Polynucleotide Kinase (New England BioLabs) and [32p]ATP (Amersham). After incubation for 3 days at 12°C, the ligation mix was heated to 65°C for 20 min and made up to 150 itl in high salt buffer (50 mM Tris-HCI, pH 7.5/10 mM MgCI~/100 mM NaCI/I mM DTT/100 mg

ml -~ BSA) followed by incubation with 100 U of EcoRl (New England BioLabsl at 35°C for 16 h. The restriction digest was then phenol/chloroform and chloroform extracted and cDNA purified by passage over Sepharose 4B (Pharmacia, 10-ml bed volume) in a 10-ml disposable pipette plugged with siliconised glass wool. Fractions were analysed on alkaline agarose gels [15] and those containing cDNA >400 bp pooled and ethanol precipitated. This was resuspended in 4 pl of 1 × ligation buffer containing 1.5 Itg of EcoRldigested dephosphorylated Agt 10 DNA I Promega Biotec), 400 U of T4 DNA ligase added (New England BioLabs), and incubated at 16°C for 16 h. Aliquots (I-ILl) were packaged in vitro (Promega Biotec) and dilutions plated on Escherichia coli C600 hfl. Bulk insert DNA was prepared from these phage [13] and cloned into EcoRl-digested dephosphorylated Agt I I IPromega Biotec).

Antibody screening amt affiniO'-purified antibodies. Antibody screening was performed as previously described [16]. Rabbit antisera were preadsorbed on filters of non-recombinant Agtll plaques prior to use for immunoscreening and affinity purification of antibody on individual clones [ 17]. Arrays of clones were made by placing 100-1~l volumes of phage lysate in triplicate in microtitre trays. These were replicated using a pin-board onto lawns of bacteria that had been plated in top agar and allowed to grow at 37':'C f o r l h. Southern blotting, Genomic DNA was digested according to the manufacturer's instructions, prior to resolution on 0.8% TBE agarose gels. DNA was transferred to nitrocellulose (Schleicher and Schuell) and hybridised as previously described [18]. Probes were made by random priming (Amersham multiprime RPN 16011 of gel-purified insert DNA. DNA sequencing strategy and sequence analysis. All sequencing was by the chain-termination method [19] with modifications described by Biggin et al. [20]. BAL 31 nested deletions were cloned into Smal-cut dephosphorylated M 13mp 19 and used to compile a provisional sequence and restriction map. Appropriate restriction fragments

14(,

~ere then chosen to obtain complete sequence data from both strands. For oncA I, a combination of Ddel and Avail cleax age ~ as u~,ed. EcoRl/Rsal fragments of oncBl completed the sequence of both strands. Computer analyse~ ~ere carried out u~ing a ~ariety of VAX-based software, particularh the "Staden Plus" package and FASTA. Hxdropath} x~as calculated using the algorithm of K~te and Doolittle [21].

Results

Con,gtruction and screenin,P, of the e.~pression libraries. Approximately 7 x I0" hatched and acti;ated T. taenia~:/-'ormi~ oncospheres >ielded approx. 150 ng of polvlA) ~ RNA, which ~as then taken through a series of carefulb optiraised steps (Materials and Methods) to produce 3 x 105 independent recombinants in ,\gtl0. Bulk insert DNA ~as prepared and subcloned into ,\gtl 1. Approximately 10 -~ recombinant plaques ~ere ~creened under non-denaturing conditions with a rabbit antiserum against FII that had been pre~,iously sho~+n to haxe the same specificit5 as sera from mice xaccinated with FII [2]. Nine positive clones were identified and their relationship ~ith one another determined b+ reacting arra.~s of the clones with antibody affinitypurified on individual clones. This analysis revealed that all the FII-positi+e clone,~ ~ere divisible into just 2 families, although not all member.,, ~ithin each family were equally immunoreactixe ~data not shown). The hierarch~ of immunoreactivit.,, ~ ithin each famil.~ ~ as maintained irre.,,pectire of which member ~as used to purify the antibod.~, and in all ca.,,es the least immunoreactive clones had the longest inserts and produced unstable fusion proteins (data not >,hown). Subsequent experiments x~ere performed on the most immunoreacti~e member of each family, which we called +mc.41 and +mcB I.

Subunit composition ot'FIl.

To directly establish the relationship between the clones and the FII bands on Western blots, antibody was affinitypurified on phage plaques and used to probe Westem blots of DOC-solubilised oncospheres. Antibody purified on both clones recognised the same bands as intact ~-FI[ (data not shown). As these

~ere distinct clone type.,,, thts suggested that the discrete FII bands ~ere serologicall.~ related and comprised at least t ~ o distinct antigenic componellt~.

Our findings prompted attempts to investigate the .,,ubunit nature of FII. DOC is not full 5 denaturing and the antigen+ are not recognised b~ FII antisera on Western blot>, of SDS gels 6(IkDa region was increased, suggesting tha~ the 34kDa antigen is disulphide-bonded to other components, and could in fact exist as a disulphidebonded dimer in its native state. It i,, a fcmnal possibility that the antigens, migrating at 34 and 25 kDa are different fom~.,, of the same molecule. The samples in Fig. 2A could be incompletely reduced, or ha~e refolded during analysis, such that the 34-kDa pol.vpeptide i~

A) ~ c AI !

~ A ITC CGT ~ T A ~ ATA CCT CAG AAC GTA C ~ AIT ~ G GCG ATC ~ C CCT CAC ACC ~ A CGC ATG ACA TGG GAC CCA CCA ~C A ~ TCC A S I P Q N V Q I E A I D P H T A R M T w D P P A K

S

2

/

TAC G~ ICC ATC All G~ TAC ACC AIC CAG I&G AGI ACG ~T GAI CTG TGG CTA GAA CAG GTC A~ GIG GCC TCA ~C AAC TCC TAT ~C Y G S | I G Y T [ Q W S T D D L W L E Q V K V A S D N S Y

D

S

I

1811TC ~ G ~ T TIG CAG TCT C~ CAG ACC ATT GII GCC TCC ATC fIG GCC CAT CAT CGA CCC GAI ACA TCG GIG A~ TIT ~G IAC ATC G~ F K D L Q S Q Q T I V A S I L A H H R P D T S V K F E Y I

G

B

/

I

7

9|

211 ACA CGC AGT ACT CCC GTC A~ GIA ACC ACA CCT CIG CCA ICI CGA GIC ATG A~ A~ CCG ICG IIC AGG GCA ACA IAC TGC GAG ~C CTC T R S T P V K V T T P L P S R V M K K P S F R A T Y C E U L

I

361 GAG ~G TTG ~C ATG CTG ATA CAC GAC CCC GAG ~G Grl GTA G~ ATG TTI G~ GGI TIC ~G GTT CIG AIG AGG GCA GGT ~T GCA ~A E E L O M L I H D P E E V V G M F G G F E V L M R A G D A E

I

451 CTA CTC AAG TCA TGG C~ TCA GTG GTT AAC CTI ACT ~A ~A ~G CGA CGA TAT AAG TIG AAG G~ CTG GTG CCA ICC TTA CCA TAC ~G L L K S W Q S V V N L T A E E R R Y K L K G L V P S L P Y A 541GIC ACAGTG CGT GGA GIA GCA CIG CCG AGC AGG AGG TTC AGT GAA CIT GCT GAG CCC GIG CAT TIT A~ ATA ICA AGG GCA ~G C~ AGI W I ~ R G Y A L P S R R F S E L A E P V H F K [ S R A E Q S

4

/

I

/

I 2

0

7

63!

GTT CCI C~ AAC GIG GAT AIG C~ ~A ACA ~T CCT CAI ACA ATA GAA ATG ACG TGG GAC IGC EGG CEG AAC CCA ACG GTC GCA TTA CTG V P Q N V D M Q A T D P H T I E M T W D C R P N P T V A L L

lZl

GET ACA £CA TCA~T GGA GTI GCA ATA ACG AGA A~ AGC CGA AGT CTC CAT CTC ~C TEA GTT CAG CIC IAC ACI ITC ACI GGT ITG~G A T P S A G V A I T R S S R S L H L A S V Q L Y T F T G L E

2

6

1

CCA C~ GTC ICG CTT ICT ~C TCC All TGT GTG CAC IAC ~ CCC C~ A~ TCA CCC AAT TTT ~G TAC ATC AGC TCC TIC AGT ~G GTT P Q V S L S A S I C V H Y K P Q K S P N F E Y | S S F S E V

2

9

1

81! 90|

GIG ACA ~A ACT ACA CCA TCC GIG GAG C~ GTC GAC TTC ~A GAA GGA GAG GAG TAG ACG CAC TGG GAA ATC GTA ~G TCG GCA AAG I ~ V T A T T P S V E Q V D F A E G E E " *

2

318

991 ATT GTT CGC TTT CTC AGT GCA AAT TAG AGA AAT ACA TTA C~ TTG GAA GGA A~ A~ A~ G~ ATT C

B) ONC BX 1

GAA TIC CGC AAG AGC GAC ATT GGG TCI' TCC CTT CCG CAA CAC TIT AGG 1GG AC,C CAA GIG GGT TCG CAG TCC ATC CAG CTA AGT TGG CAT K S O ! G S S L P O H F R W S O v G S O S ! Q L S W O ?l

91

GAC CAC AAG GIG AAC GGC CAC ACA CCA CAT CAA GII CAC CTC TTT GTC ATA CCA AGA AGI TCC ICC ITA CAP.AM GTT GAA AAA CAG GTT D

H

K

V

N

G

H

T

P

O

Q

v

H

L

F

V

I

P

R

S

S

S

L

Q

R

V

E

K

Q

V

51

181 GCC TIC AGC GCC A~ ACA GTC ATT CTT C&A GGA CTG CAT GGG AAC ACA CIG TAC TAC GTG ATA CTG ACI GIA AGC GCT GGT GAA GAG GAA A

?11

F

S

A

K

T

V

I

L

E

G

L

H

G

N

T

L

Y

Y

V

I

L

I

V

S

A

G

E

E

E

81

TGC ICA ATC ACA TCG GCA CAJ~ ICA GAA CTA AGA AGT ATG TGT ATA AA.~ IGA CGC CTC TIC CGC AAC ACT TIC f~AT GGA GTA ACG TGG GTC C S I T S A Q S E L R $ M C I K ''' ]Z3

361 CIC AGT CCG TCC GGC TAJ~GCT GGG ATA AAC ACA CCT TGG ACG C,~T ACA CTC CA.~ CTC TAG TGG ITA TGA CAC TCA AGT CA.~C,~A AGC CAG 45|

ATG AAI TC

Fig. 4. DNA and predicted amino acid ,~equences of on~AI and on(BI. Actual and gel estunate ,,ize~ (m bracket~ m kDa are: oncAI. 35.(i)74 136.0K o m B I . 11,354 I I3.0K

3

/

143 a reduced version of the 25-kDa moiety, having a more unfolded conformation, thereby migrating more slowly.

Southern bloning. The Southern blots shown in Fig. 3A and B suggest that oncAI and oncBI are distinct genes that are present at low copy number in the genome. Some of the digests yielded more than one hybridising fragment in the absence of internal restriction sites in the probe sequence: for example, Fig. 3A lane 1 shows three genomic EcoRI fragments hybridising to the oncAl insert, which does not contain internal EcoRl sites. This is often observed [22] and presumably reflects the existence of more than one copy of the gene (perhaps alleles) and/or the presence of introns. The present data does not enable us to distinguish between these possibilities. Sequence analysis. The DNA sequences and derived amino acid sequences are shown in Fig. 4. Each has a single long open reading frame, maintaining the reading frame of the EcoRI sites in Agtll and pGEX-I, predicting peptides in good agreement with size estimates of fusion proteins on SDS-PAGE (legend to Fig. 4). Neither of the derived amino acid sequences was significantly homologous to any other in the current data bases, so the biochemical function of these gene products is unknown. However, a common feature of the derived amino acid sequences is their hydrophobicity. Hydropathy profiles of the oncAl and oncBI peptides are shown in Fig. 5, where it can be seen that most of the sequence lies above the median line in both cases, indicating net hydrophobicity [211. One other point arising from the sequence data is the lack of a consensus "AATAAA" polyadenylation signal upstream of the poly(A) tail in oncA 1. This sequence was also absent from cDNAs encoding a 32-kDa T. ovis antigen (KSJ, unpublished) and Echinococcus granulosus cyclophilin [23]. While the majority of higher eukaryote mRNAs contain the AATAAA sequence 10-30 nucleotides upstream of the poly(A) tail [24], natural variants do occur, most often differing at not more than one position in the 6nucleotide motif [25,26]. To our knowledge, only 4 cestode cDNAs containing a poly(A) tail have

A) OncA1

'°L

[

i 40[ c

I 0

I

I I OO

I rescue

i 200

i

i 300

nurn~r

B) OncB1

~

-40 I 0

I

I 100

res~lde number

Fig. 5. Hydrophoblcity plots of IA) omAI and (B) omBI amino acid sequences. Degree of hydrophobicity ~ithin a ~indov, of 13 residues calculated as described [21]. been sequenced to date. Fig. 6 shows the 3'untranslated regions of these, revealing that single nucleotide variants of the consensus sequence do occur at the correct distance from the start of the poly(A) tail. The AATACA motif upstream of the poly(A) tail in both oncA 1 and To32 occurs in approx. 2% of eukaryotic cDNA sequences [25]. The AATGAA motif in E. granulosus cyclophilin is exceedingly rare, and has been associated with defective 3' end processing in other eukaryotes [25]. Many more sequences will be necessary to determine whether variant poly(A) addition sequences in cestode mRNAs will prove the rule rather than the exception.

Discussion We describe here the construction and use of cDNA clones to analyse antigens in complex mixtures that are not readily amenable to analysis by conventional protein chemical techniques. This is significant because whilst T. taeniaeftnwlis has been extensively characterised with respect to immunobiology of resistance to infection, little is known about the biochemical nature and complexity of the host-protective oncosphere antigens. oncA and oncB appear to encode the antigens that are the primary, targets of the protective antibody response to the FII complex, so it is now pos-

144

Eukar~,ot e

......................

~TAlkA

< ....

(i0-30)

oncAl

CTTTCTCAGTGCAAATTAGAGA~dkT~ILC.~TTACAATTGGAAGG

TO32

ATGACGGGTAGATACGCACGACAlkTA~CGAACACGGCACCTTTT'(A)~

EA21

TGTGGCCGTAACGTGTTTCAACA2kTG~GTCGTTGTGCGTATTTTG(A)

EgPS 3

TGACTAGTCATAGATTTTGCTC~TA/kACCATCTACTTTATCTCTT(A)44

. . . . . > {A) n

....

(A) I0

8

Fig. 6. Comparison of 3' untranslated sequences of four cestode cDNAs. The top line sho~s the consensu~ 3' structure of eukaryonc mRNAs ~[I read as TL The AATAAA polyaden>lauon signal is underlined in bold type. and the poly(Ab tail of ~ariable length is designated I Am. oncAI I thls stud)'l. To32 ~/ ovis antigen, KS J, unpublished ~, Ea21 [cc eranulosus c~clophihm [231 and EgPS3 t E eramd~,sus antigen B: Shepherd, J.C. 11988~ Ph.D. Thesis. Imperial College of Science and Technology, Londom. 3' untranslated regions are lined up on their putau~e polyaden.~lanon signals. Gaps introduced into the sequence are .ndicated ~ t h a " ,,}mbol.

sible to discuss the properties of the oncosphere antigens in light of the results presented here. The FII region comprises approx. 10% of total DOC-soluble antigen [2], and would therefore be expected to be complex. This was found to be true when each of the FII bands recognised by antibody on DOC gels resolved into an overlapping series of polypeptides on SDS-PAGE, confimfing and extending an earlier observation by Lightowlers et al. [2] who found thai the FII region contained polypeptides in a broad molecular weight range. Such complexity within each discrete FII band most likely results from the formation of stable micelles bet~.een DOC and oncosphere antigens [2,27,28]: this could have some bearing on the vaccination efficacy of FII, as such micellar structures would resemble liposomes. The FII antigens could be among the most immunodonfinant in oncospheres, since a variety of antisera raised against different crude vaccinating preparations exclusively recognise the FII bands on DOC Western blots. The correlation between protective capacity of an extract and recognition of the FII bands suggests that FII is enriched for protective antigens present in more complex mixtures. This correlation extends to the identification of a common 62-kDa antigen in immunoprecipitates using sera raised against three different vaccine preparations: homogenised oncospheres (a-TtO), c~-Fll and ~-Iarval ES products. CDNA clone oncAI appears to encode at least pan of this antigen and is therefore a promising

candidate for a protective antigen gene. It may be related to the Tt55/60-kDa antigen pair previously described by Boo, tell et al. [8] as potentially host-protective, even though no trace of a 55-kDa antigen ~as ever seen on our gels. Tt55/60 was almost uniquely precipitated from biosynthetically labelled ES antigens of both oncospheres and larvae by t~-TtO and ~-Iarval ES. Both antigens yielded identical peptide maps [8]. suggesting that the 55-kDa moiety could have resulted from proteolysis of the 60-kDa molecule in the culture prior to harvesting. Antisera to the oncB clone type immunoprecipitated a pair of antigens of 25 and 34 kDa which also appeared to be recognised by antisera to oncosphere but not la~'al extracts. These appear to be oncosphere-specific [8]. We are uncertain as to why two antigens were precipitated but could rule out disulphide bonding between the two. Antibodies to the two cDNA clones reconstruct the major antibody specificities of the FII antiserum when analysed by immunoprecipitation techniques, accounting for the fact that only two clone types were isolated from a theoretically representative cDNA libraD,. Earlier biochemical studies have indicated that the native antigens are sensitive to denaturing agents and sparingly soluble in aqueous buffers in the absence of detergent+ [2.6.7]. This. together with their solubility in DOC and the fact that they can be pelleted by high-speed centrifugation has led to the suggestion that they could

145

be membrane-associated [2,27,28]. The sequences of the two cDNA clones predict abundant hydrophobic amino acids. Taken at face value, this could account for insolubility of the native antigens in aqueous buffers, but does not necessarily mean that they are membrane-associated. Both cDNA sequences were analysed for ~:~-helices having the necessal3' characteristics for interacting with membranes (hydrophobic moment plots) [29]. but no suitable structures were found (data not shown). The stretches of hydrophobic residues could participate in intermolecular bonding with other hydrophobic interfaces, but perhaps more likely represent the interior of the folded molecule. A similar net hydrophobicity is often found in globular proteins with solved crystal structures, in which case the hydrophobic regions are most often found in the interior of the protein [21]. Complete sequences of the two antigens may clarify this point, though it is difficult to prove membrane association from sequence data alone. More cDNA sequence data is also needed to investigate the occurrence of variant polyadenylation signals, seen in 3/4 cestode sequences compiled to date. Apart from the sequence, location of this motif is important, enabling it to be identified in the overwhelming majority of mRNAs from eukaryotes [24-26]. The existence of single nucleotide variants of this sequence at the expected distance from the poly(A) tail (Fig. 6) does not prove that these sequences are functional, although it is somewhat unlikely that cestodes have evolved a unique method for 3' processing of mRNAs. Vaccination trials have been been performed with the oncAl and oncBI clones expressed as fusion proteins with glutathione S-transferase following subcloning into pGEX-I [30]. Both antigens induced strong antibody responses in mice, but did not stimulate significant protection against challenge infection with T. taeniaeformis eggs (data not shown). Vaccination experiments with native FII antigens strongly implicates antibodies to the native equivalents of oncA and oncB as being host-protective [2]. One possibility is that the oncAi and oncB1 clones, both incomplete, do not contain the relevant host-protective epitopes, oncAl and oncB! had the shortest inserts in the oncA (longest insert 2-kb) and oncB (longest

insert 950-bp) families, despite being the most immunoreactive. The tess reactive clones produced unstable fusion proteins (data not shown) which probably accounts for the reduced signal in immunoassay. A similar inverse relationship between insert size and immunoreactivity/fusion protein stability was obse~'ed ~ith T. ovis antigens, in which the shorter version of a protective antigen gene, producing a smaller and much more stable fusion protein, was not protective (ref. 3. and unpublished obse~'ations). Fusion protein instability in Escherichia ~'oli was a problem with the majority of oncosphere antigen clones we have isolated so far (8/12), and this problem must be solved, possibly through the use of alternative vector/host systems, betbre longer oncA and oncB clones can be tested in vaccination experiments. The approach described here has enabled us to narrow down the FII complex to two polypeptides, and the genes for these have been cloned. Though the recognition of one.4 I-encoded antigen by c~ES sera may be an important clue, it is likely that a combination of approaches will now be needed to elucidate the biochemical function of these gene products.

Acknowledgements This work was funded by research grants from the Melbourne and Metropolitan Board of Works. Coopers Animal Health New Zealand Ltd., and the National Health and Medical Research Council of Australia.

References I Rickard. M.D. and Williams. J.F. (1982t Hydatidosis/c.,,sticercosis: immune mechanisms and immunization against infection. Adv. Parasitol. 21, 229-296. 2 Lightov, lers, M.W.. Rickard, M.D. and Mitchell, G.F. ( 1986t Immunization against Taenia taeniaeformis in mice: identification of oncospheral antigens m pol~acrylamide gels by Western blotting and enzyme ~mmunoassa.~. Int. J. Parasitol. 16, 297-306. 3 Johnson, K.S., Harrison, G.B.L., lightowlers, M.W.. O'Ho~. K.L., Cougle. W.G., Dempster. R.P., Lawrence. S.B.. "Vinton, J.G.. Heath, D.D. and Rickard, M.D. (1989) Vaccination against ox ine c)sticercosls using a defined recombinant antigen. Nature 338. 585-587. 4 Bowtell, D.D.L, Saint. R.B., Rickard, M.D. and Mitchell, G.F. tl9841 Immunochemical analysis of Taenia taeniaeforints antigens expressed in Escherichia t'oh. Parasltolog) 93, 599-610.

146 5 Bogh, HO.. Rickard, M . D and Lightox~lers, M.\~,.L Iggg~ Studle~ on ,,rage-specific imrnunlty to [uellttt ttl~'Ilttl~'./Ot't~ttS metacestodes m mice. Para,,ite Immunol. 10, 255-264. 6 Raja~ekanah, G.R.. Mitchell, G.F. and Rickard, M.D I It)g0) ~dt'/titl loe/ltttefi;/r~lt3 in mice: protectl',e inm3unisalion ~.,.ith onco~phere~, and their products. Int. J. Para,qtol. 10. 155-160. 7 Rajasekarmh, G.R.. M~tchell, G.F. and R~ckard. M D It)g0) Immum,,ation of mice against infection ~ ith Taema tdcnt(telormts u>mg ~ariou'~ antigens prepared from egg,,. oncospheres, de~elopine larxae and ~lrobilocerci. Int. J. ParasltOl I(I. 315-324. 8 Boo, tell. D.D.L, Mitchell. G.F.. Anders, R.F.. Lightowlers. M W. and Rickard. M.D. t l t)83~ ~lent~t ttlem~leJ~)t'mts: immunoprecipitatlon analysis of the protein antigen,~ of oncosphere~ and larvae. [~xp. Parasitol. 56, 416--1-27. g Lighto~lers, M.\V.. Mitchell. G.F.. Bowtell. D.D.L.. Ander,,, R.F. and Rickard, M . D ~It)841 lmmunisation against Tdenta Identael~)rmts in mice: studie,~ on the characlen•,ation of antigen,~ from oncospheres. [nt. J. Parasitol. 14. 321-333. I0 Kee, K,C.. Ta~lor. D.W., Cordingle~, J.S.. Butter~orth, A.E. and Munro, A.J. ~It)86) Genetic .nfluence of antibod.~ re,,ponses to antigens of Schtstosonta mansoni in chronicall.v infected mice. Para,~ite [mmunol. g. 565-574. II Bowtell. D.D.L., Saint. R.B.. Rickard, M . D and Mitchell, G.F. (I t)84) Expression of Taenia taenia(Jbrmts anugens in E~t hertchta cHt. Mol. Biochem. Parasltol. 13, 173-185. [2 Heath, D.D. and Sm3th, J.D. (1t)70~ In vmo cultivation Of Et'hllh)t'OL t ItS ~,ttlttlllostt~, ]-tl('tlltl h)datieena. F. orig. T. ptstl~)tmi~ and T setialt,~ front oncosphere to c.,,StlC larva. ParasltOlog 3 61, 32t)-343. 13 Kemp, D.J., Coppel. R.L., Cowman, A.F.. Saint. R.B., Bro~n. G.V. and Anders, R.F. 11983~ Expression of Plaslilt)till#It .ftlhlptIl'tltlt blood stage antigens in E.scherlchia ,,It: detection ~ ~th anlibodies from immune humans. Proc. Natl. Acad. Sci. LISA 80, 3787-3791. 14 Av~, H. and Leder. P ( 1972t Purificauon of biologicall) active globin messenger RNA b.,, chromatographs on oligoth~mid.~lic acid-cellulose. Proc. Natl. Acad. Sci. USA 69, 1408-1412. 15 Manlatl~. T., Fritsch, E.F. and Sambrook. J. 119821 Molecular Cloning. A LaboraloD' Manual. Cold Spring Harbor LaboratoD. Cold Spring Harbor. NY. 16 Hu,,nh, T.V., Xtbung. R.A. and Davis. R.W. (19851 Construtting and ~creenlng eDNA libraries in AgtlO and ),gtl I. In: DNA Cloning. A Practical Approach. tD. GIo,,er. ed.L pp. 49-78. IRL Press. Oxford.

17 Beall. J.A. and Ikhtchell. G.F. q 1'4~6~ Identification ol a particular parasite anugen from a para,qte eDNA hbrar) uelected portionuel. M.L. Busslinger, M. and Strub. K. ( It;85~ Tran,,cription termination and 3' processing: the end is in s~ght! Cell 41, 349-359. 26 Wicken~, M. and Stephenson. P ( It)84~ The role of the com, e~'ed AALIAAA sequence: four AALIAAA point mutation,, pre~ent me>senger RNA 3' end formation. Science 226. 1045-1051. 27 Helemus, A., McCa~lin, D.R., Fries, E. and Tanford, C. (197t)) Propertle~ of detergents. Methods Enz)mol. 56. 734-749. 28 Heleniu,,, A. and Simons, K. (1975b Solubilisation of membranes ,x ith detergent,~. Biochmt. Bioph)s. Acta 415. 29-79. 29 Eisenberg, D., Schv, arz, E.. Komarom~, M. and Wall, R. I lt)84t Analysis of membrane and surface protein sequences ~ith the h.~drophobic moment plot. J. Mol. Biol. 179, 125-142. 30 Smith, D.B. and Johnson, K.S. 11988~ Single-step purdication of polypeptides expressed in Eschert('hia colt as fu•,~ons ~ ith glutath~one S-transferase. Gene 67. 31--I-0