Protection against Bordetella pertussis in mice in the absence of detectable circulating antibody: implications for long-term immunity in children

2087 CONCISE COMMUNICATION P o ec ion agains Bo de ella pe ussis in Mice in he Absence o De ec able Ci cula ing An ibody: Implica ions o Long-Te m Immuni y in Child en Be na d P. Mahon, Mi iam T. B ady, and Kings on H. G. Mills In ec ion and Immuni y G oup, Depa men o Biology, Na ional Uni e si y o I eland, Maynoo h, Co. Kilda e Mos accines used o humans wo k h ough humo al immuni y, ye many appea o be p o ec i e e en a e speci ic ci cula ing an ibody le els ha e waned o unde ec able le els. Fu he mo e, i has been di icul o de ine a se ologic co ela e o p o ec ion agains a numbe o in ec ious diseases, including hose caused by Bo de ella pe ussis. B. pe ussis clea ance in immunized mice has been shown o co ela e wi h pe ussis accine e icacy in child en. This mu ine espi a o y challenge model was used o demons a e pe sis en accine-induced p o- ec ion agains B. pe ussis in he absence o ci cula ing an ibody a he ime o challenge. Whole-cell and acellula pe ussis accines induced pe sis en memo y T and B cells and anamnes ic an ibody esponses a e challenge. The indings sugges ha immunologic mem- o y is mo e signi ican in p o ec ion han is he induc ion o immedia e an ibody esponses and imply ha accina ed child en s ill may be p o ec ed agains disease ollowing he dis- appea ance o speci ic se um IgG. Whole-cell pe ussis accines appea o induce long-las ing immuni y agains se e e whooping cough in child en and young adul s. Howe e , conce n abou he sa e y o whole-cell ac- cines has esul ed in he de elopmen o acellula pe ussis ac- cines p epa ed om pu i ied componen s o Bo de ella pe ussis. The new accines, which ha e a highe sa e y p o ile, ha e eplaced whole-cell accines in many de eloped coun ies. Re- sul s om clinical ials ha e indica ed ha acellula pe ussis accines con e ela i ely high le els o p o ec ion agains se e e disease [1–4], bu hese s udies ha e no p o ided de ini i e in o ma ion abou a ela ionship be ween an ibody le els agains p o ec i e an igens and immune p o ec ion. Follow-up s udies ha e indica ed ha an ibody esponses in immunized child en decline o low o unde ec able le els 2 yea s a e a comple e cou se o accina ion [5, 6]. These child en do no appea o be de eloping o e disease [7], bu , wi hou a his o y o exposu e and he applica ion o sensi i e indica o s o in- ec ion in la ge pos immuniza ion su eillance p og ams, i is impossible o de e mine whe he hey a e immune o in ec ion and o es ablish he mechanisms ha media e p o ec ion. Al- Recei ed 17 Decembe 1999; e ised 15 Feb ua y 2000; elec onically published 5 June 2000. P esen ed in pa : B i ish Socie y o Immunology Annual Cong ess, Ha - oga e, Uni ed Kingdom, Decembe 1999 (abs ac W2.2). Financial suppo : Wellcome T us (“New Blood” ellow o B.P.M. and p ojec g an 052317 o K.H.G.M.); Heal h Resea ch Boa d o I eland ( o M.T.B.). Rep in s o co espondence: D . Kings on Mills, In ec ion and Immuni y G oup, Dep . o Biology, Na ional Uni e si y o I eland, Maynoo h, Co. Kilda e, I eland (kings [email protected]). The Jou nal o In ec ious Diseases 2000;181:2087–91 q2000 by he In ec ious Diseases Socie y o Ame ica. All igh s ese ed. 0022-1899/2000/18106-0036$02.00 hough hese undamen al ques ions a e di icul o answe in a clinical se ing, we ha e been able o add ess hem by use o a well-es ablished animal model. By use o a mu ine espi a o y challenge model o B. pe ussis in ec ion, in which p o ec ion co ela es wi h accine e icacy in human clinical ials, oge he wi h gene knockou mice, we ha e iden i ied complemen a y oles o cellula and humo al immuni y in p o ec ion [8–11]. In ou s udy, we used his model o examine he ela ionship be ween he pe sis ence o he im- mune esponse and he de elopmen o immunologic memo y and p o ec ion agains espi a o y in ec ion. Ma e ials and Me hods Immuniza ion. BALB/c mice we e immunized in ape i oneally (a 0 and 4 weeks) wi h 0.04 human dose o ei he Wellcome whole- cell pe ussis accine (B i ish Re e ence eagen 88/522; Na ional Ins i u e o Biological S anda ds and Con ol, Po e s Ba , He s- o dshi e, UK) o Chi on acellula pe ussis accine (Chi on Vac- cines, Siena, I aly), equi alen o 0.2 mg o ecombinan pe ussis oxin (PT), 0.1 mg o ilamen ous hemagglu inin (FHA), and 0.1 mg o pe ac in (PRN). Con ol mice we e immunized wi h adju an (alum) only. Challenge and cou se o B. pe ussis in ec ion. Respi a o y in- ec ion o mice wi h B. pe ussis was ini ia ed by ae osol challenge as desc ibed elsewhe e [8]. The kine ics o bac e ial clea ance we e de e mined by coun ing colony- o ming uni s (c u) in he lungs on days 0, 3, 7, 14, and 21 a e challenge [8]. Resul s a e epo ed as he mean numbe o iable B. pe ussis o indi idual lungs om a leas 4 mice pe ime poin pe expe imen al g oup. An ibody esponses. The le els o se um an ibody o B. pe - ussis and bac e ial componen s we e de e mined by ELISA as desc ibed elsewhe e [9]. B. pe ussis sonica e (5.0 mg/mL) o PT, 2088 Mahon e al. JID 2000;181 (June) PRN, o FHA (1.0 mg/mL o each) we e used o coa pla es. Bound an ibodies we e de ec ed by use o alkaline phospha ase–conju- ga ed an i–mouse IgG. An ibody le els a e exp essed as he mean end-poin i e (5SE), de e mined by ex apola ion o he s aigh pa o he i a ion cu e o 2 SE abo e he backg ound alue (OD <0.2) ob ained wi h nonimmune mouse se um. Assays we e s anda dized wi h e e ence an ise um p epa ed om mice im- munized wi h he B i ish e e ence p epa a ion o whole-cell pe - ussis accines [9]. Ti e s >1/20 (1.3 log 10 ) we e conside ed posi i e. T cell cy okine assays. Spleen cells om indi idual mice we e es ed a di e en ime poin s a e immuniza ion and a e chal- lenge o in i o cy okine p oduc ion in esponse o hea -killed B. pe ussis (10 6 /mL), PT (5.0 mg/mL; hea -inac i a ed a 807C o 20 min), FHA (5.0 mg/mL), PRN (5.0 mg/mL), concana alin A (2 mg/ mL), o medium alone as con ol. Supe na an s we e emo ed om iplica e cul u es a e 72 h, and in e e on-gand in e leukin-4 concen a ions we e de e mined by immunoassay [7]. Assessmen o memo y B cells. An ibody- o ming cells we e quan i ied by an insoluble subs a e B. pe ussis–speci ic ELISPOT. The wells o pla es (Mul isc een HA; Millipo e, Bed o d, MA) we e coa ed o e nigh wi h 100 mL o sonica ed B. pe ussis o inac i- a ed PT (5 mg/mL). Bo h su aces o he memb ane we e washed 5 imes asep ically wi h PBS and hen we e blocked o 1 h wi h 8% ( ol/ ol) e al cal se um in RPMI medium. A e blocking, spleen o pe iphe al blood mononuclea cell p epa a ions we e added o quad uplica e wells a concen a ions om o 4 1310 /mL and we e incuba ed a 377C wi h 5% CO 2 o 2 days. 7 1310 A e being washed 6 imes in PBS, pla es we e incuba ed o 2 h wi h ho se adish pe oxidase–conjuga ed goa an i–mouse IgG (Sigma, Do se , UK). Pla es we e washed 3 imes mo e be o e he addi ion o 200 mL/well 3-amino-9-e hylca bazole subs a e bu e . A e 45 min, his was emo ed, and he pla es we e washed once wi h wa e and allowed o ai d y. Spo s ep esen ing indi idual an ibody- o ming cells we e coun ed by use o a zoom s e eo mic oscope. Resul s Acellula and whole-cell pe ussis accines p o ec agains B. pe ussis a e he disappea ance o se um IgG esponses. Im- muniza ion o mice wi h 2 low doses o ei he a whole-cell o an acellula pe ussis accine induced s ong se um IgG e- sponses agains PT, FHA, and PRN, bu an ibody le els de- clined apidly a e 3 mon hs and we e unde ec able by 6–9 mon hs ( igu e 1A). Immunized mice we e challenged wi h B. pe ussis ei he a he peak o he an ibody esponse 6 weeks a e p ima y immuniza ion o once ci cula ing an ibody i e s had waned, 44 weeks a e p ima y immuniza ion. Despi e he decline in B. pe ussis–speci ic se um IgG o unde ec able le els, bo h accines s ill con e ed p o ec ion agains B. pe ussis when challenged 44 weeks a e immuniza ion ( igu e 1B). The numbe s o bac e ia in he lungs we e signi ican ly highe in con ol han in ei he whole-cell accine– o acellula accine– immunized mice a all ime poin s es ed ( ). In pa -P!.01–.001 icula , mice immunized wi h whole-cell pe ussis accines dis- played a high le el o p o ec ion ha was equi alen o ha obse ed a he peak o he an ibody esponse 6 weeks a e immuniza ion; coun s o c u we e signi ican ly g ea e in whole- cell accine– han in acellula accine–immunized mice 3, 7, and 10 days a e challenge a 44 weeks. Immuniza ion induces memo y T and B cells. Recall T cell esponses o B. pe ussis an igens pe sis ed in whole-cell ac- cine– and acellula accine–immunized mice, and, on he day o challenge a 44 weeks a e p ima y immuniza ion, B. pe - ussis–speci ic T cell esponses clea ly we e de ec able in im- munized mice ( igu e 2A). Whole-cell pe ussis accine induced a Th1-domina ed esponse, whe eas acellula pe ussis accine induced a Th2- ype esponse, and hese pa e ns o cy okine sec e ion pe sis ed o a leas 44 weeks. Howe e , 14 days a e challenge, he cy okine p o ile o B. pe ussis an igen–s imula ed spleen cells om mice immunized wi h acellula pe ussis ac- cine had swi ched o a mixed Th1/Th2- ype esponse (18–33 ng/mL in e e on-gand 100–217 pg/mL in e leukin-4), whe eas he Th1 esponse was main ained in mice immunized wi h whole-cell pe ussis accine (58–160 ng/mL in e e on-gand unde ec able in e leukin-4). Memo y B cells speci ic o B. pe ussis also we e induced by immuniza ion wi h whole-cell and acellula pe ussis ac- cines and pe sis ed o a leas 44 weeks a e accina ion. By use o an ELISPOT echnique, speci ic an ibody- o ming cells we e de ec ed in pe iphe al blood mononuclea cells and spleen cells a e s imula ion wi h PT o B. pe ussis sonica e ( igu e 2B). Signi ican ly g ea e numbe s o B. pe ussis–speci ic B cells could be de ec ed in he blood and spleens o immunized mice han in hose o con ols. Memo y induc ion was con i med by he obse a ion o an anamnes ic an ibody esponse o B. pe - ussis an igens in immunized mice a e challenge ( igu e 1A). IgG esponses o B. pe ussis could no be de ec ed in unim- munized con ol mice o a leas 21 days a e challenge. In con as , 7 days a e challenge, ci cula ing an ibodies we e de- ec ed agains PRN and B. pe ussis sonica e in whole-cell ac- cine–immunized mice and agains PRN and PT in acellula accine–immunized mice. Discussion Ou s udy demons a es ha ci cula ing an ibody, conside ed o be he majo e ec o mechanism agains i uses and bac e ia, need no be de ec able a he ime o exposu e o he main- enance o accine-induced p o ec i e immuni y agains a bac- e ial pa hogen o humans. P o ec i e le els o ci cula ing an- ibody ha e been iden i ied o ce ain pa hogens, such as polio i us, he agen s o diph he ia and e anus, and Haemo- philus in luenzae. Howe e , ce ain pe sons who ha e an ibody le els below hese h eshold alues appea o be p o ec ed. Fu - he mo e, i has been di icul o es ablish p o ec i e le els o an ibody induced wi h ce ain accines, including hose agains pe ussis [1–4]. Recen epo s om household con ac s udies ha e indica ed ha an ibody esponses o PT, PRN, and im- JID 2000;181 (June) Pe ussis and Immunologic Memo y 2089 Figu e 1. Pe sis ence o p o ec ion agains Bo de ella pe ussis espi a o y in ec ion a e decline in speci ic se um IgG o unde ec able le els. Mice we e immunized a weeks 0 and 4 wi h 0.04 human dose o whole-cell pe ussis accine (Pw) o acellula pe ussis accine (Pa) o wi h adju an only as con ol. A, Se um an ibody esponses o B. pe ussis (BP) sonica e, pe ussis oxin (PT), ilamen ous hemagglu inin (FHA), and pe ac in (PRN), de ec ed by ELISA a in e als a e immuniza ion and 1, 2, and 3 weeks a e espi a o y challenge. B, Immunized and con ol mice a e exposu e o ae osol challenge wi h B. pe ussis 6 o 44 weeks a e p ima y immuniza ion. Resul s a e mean (5SE) colony- o ming uni s (c u) in lungs, es ima ed o 4 mice/g oup a each ime poin (e o ba s a e smalle han symbols a ce ain ime poin s o con ol mice). * ; ** , whole-cell s. acellula accine (Mann-Whi ney U es ).P!.05 P!.01 b iae may co ela e wi h absence o se e e disease [7, 12]. Al- hough hese s udies sugges ed ha an ibodies o PRN and imb iae a e he mos impo an , a monocomponen pe ussis oxoid accine con e ed 71% p o ec ion agains se e e disease in a clinical ial [4]. Fu he mo e, an ibody le els in child en decline apidly a e accina ion [5, 6, 13]. The esul s o he p esen s udy sugges ha , i exposed o B. pe ussis, hese chil- d en s ill would be p o ec ed agains de eloping whooping cough. The e a e a numbe o explana ions o he main enance o p o ec ion a e he decline in ci cula ing an i–B. pe ussis an- ibody. I is possible ha an ibody does no play a ole in p o ec ion and ha immuni y is con e ed exclusi ely by cel- lula mechanisms. Analysis o cell-media ed immuni y in mice demons a ed ha speci ic T cell esponses we e main ained up o he ime o challenge, 44 weeks a e immuniza ion. Whole- cell pe ussis accines selec i ely induced B. pe ussis–speci ic Th1 cells, whe eas acellula pe ussis accines induced T cells mo e pola ized o he Th2 sub ype, which is consis en wi h epo s on sho - e m immuniza ion [8, 9]. We ha e sugges ed elsewhe e ha he Th1 subpopula ion con e s op imal p o ec- ion agains B. pe ussis [8–11] and ha hese cells may unc ion by p omo ing he an ibac e ial ac i i y o mac ophages and neu ophils and by s imula ing opsonizing an ibodies.Howe e , acellula pe ussis accines also can con e a high le el o p o- ec ion agains se e e disease, and his appea s o be media ed solely by an ibody wi h help om Th2 cells. The sligh ly lowe le els o bac e ia in he lungs o whole-cell accine–immunized mice, compa ed wi h acellula accine–immunized mice a e challenge a 44 weeks, is consis en wi h p e ious s udies on sho - e m p o ec ion wi h a ange o whole-cell and acellula pe ussis accines [8] and may e lec he con ibu ion o cell- media ed immuni y o p o ec ion. Ne e heless, he esul s sug- ges ha he Th1/Th2 dicho omy does no impinge on memo y 2090 Mahon e al. JID 2000;181 (June) Figu e 2. Recall T and B cell esponses o Bo de ella pe ussis an igens a e immuniza ion wi h whole-cell pe ussis accine (Pw) o acellula pe ussis accine (Pa). A, 44 Weeks a e immuniza ion, spleen cells ( /mL) we e incuba ed wi h hea -inac i a ed pe ussis oxin (PT), 6 2310 ilamen ous hemagglu inin (FHA), pe ac in (PRN), hea -killed B. pe ussis (BP), concana alin A (Con A) o medium only as posi i e and nega i e con ols, espec i ely. Supe na an s we e emo ed a e 72 h and assessed o in e e on-g(IFN-g) and in e leukin-4 (IL-4). Resul s a e mean (5SE) cy okine concen a ions o 4 mice/g oup, assayed in iplica e. Cy okines we e unde ec able in supe na an s om cells s imula ed wi h medium alone. B, 44 Weeks a e immuniza ion, pe iphe al blood mononuclea cells and spleen cells om immunized o con ol mice we e incuba ed in pla es coa ed wi h B. pe ussis sonica e (BP) o inac i a ed PT, and nos. o speci ic an ibody- o ming cells (AFC) we e de e mined by ELISPOT. Resul s a e mean alues (5SE) o iplica e de e mina ions om 4–6 mice/g oup. Cy okine and AFC da a a e ep esen a i e o 4 di e en ime poin s a e immuniza ion in 2 independen expe imen s, wi h simila le els o s a is ical signi icance a each de e mina ion. * ; ** ; *** s. con ol (S uden ’s es ).P!.05 P!.01 P!.001 induc ion. Fu he mo e, al hough hese pa e ns o cy okine sec e ion a e s able, hey a e no immu able o “locked in.” A e bac e ial challenge, he e is a b oadening o he pa e n o cy okine esponses obse ed in acellula pe ussis accine– immunized mice, om a Th2 esponse 7 days a e challenge o a mixed Th1/Th2 o Th0- ype esponse he ea e . In a s udy o child en immunized wi h acellula pe ussis accines, he mixed Th1/Th2 p o ile obse ed ea ly a e accina ion [14, 15] had swi ched o a Th1- ype esponse 3–4 yea s la e [13]. I was sugges ed ha exposu e o subclinical in ec ions, which would enhance Th1 esponses, may explain he pe sis en p o ec ion agains ypical pe ussis, despi e a subs an ial waning o bo h an ibody and T cell esponses induced by he p ima y immu- niza ion [13]. An al e na i e, bu no mu ually exclusi e, hypo hesis o ex- plain long- e m p o ec ion in child en a e declining an ibody esponses is ha immune e ec o mechanisms gene a ed om memo y B and T cells a e bac e ial exposu e a e capable o clea ing he bac e ia om he espi a o y ac . This is consis- en wi h ou s udy and wi h ecen sugges ions ha an ibodies agains mul iple an igens can media e p o ec ion agains B. pe - ussis [3, 8]. Gi en he long incuba ion pe iod o B. pe ussis in ec ion, anamnes ic an ibody esponses a e exposu e may be su icien o p o ec agains se e e whooping cough. How- e e , i appea s ha cu en pe ussis accines do no con e s e ilizing immuni y agains B. pe ussis. These accines a e capable o p e en ing se e e disease bu a e less e icien when judged on he c i e ion o p e en ing in ec ion o mild disease, e en in he p esence o high le els o ci cula ing an ibody a he ime o in ec ion. Al hough he induc ion o a gi en con- cen a ion o ci cula ing an ibody ea ly a e accina ion may s ill p o ide a use ul co ela e o p o ec ion, in ha i may p edic he le el o B cell p iming, i is no clea ha i will p edic he ex en o memo y T and B cell induc ion. Al hough we do no ye ha e de ini i e in o ma ion on he ela i e im- po ance o T o B cell memo y, we ha e demons a ed else- JID 2000;181 (June) Pe ussis and Immunologic Memo y 2091 whe e ha sho - e m memo y B and T cells bo h con ibu e o p o ec ion induced by na u al in ec ion [10]. Ou indings ha e impo an implica ions o pe ussis ac- cina ion policy in ela ion o he le el o p o ec ion in he ace o waning an ibody le els and he need o boos e immuni- za ion. The da a om he mouse model sugges ha pe sis ence o ci cula ing an ibody is no a p e equisi e o he main enance o p o ec i e immuni y. These obse a ions also ha e impli- ca ions o he e alua ion and de elopmen o accines agains a ange o in ec ious diseases. Al hough ci cula ing neu alizing an ibodies clea ly a e impo an o con e s e ilizing immuni y, especially agains i al in ec ions, i also appea s ha ce ain accines can p o ec by ecall o immunologic memo y a he T and B cell le el. Acknowledgmen s We hank Ge aldine Mu phy and Helen S ewa o echnical assis- ance and Chi on Vaccines (Siena, I aly) o supplying pe ussis ac- cines and an igens. Re e ences 1. G eco DD, Salmaso S, Mas an onio P, e al. A con olled ial o wo acellula accines and one whole-cell accine agains pe ussis. N Engl J Med 1996;334:341–8. 2. Gus a sson L, Hallande HO, Olin P, Reizens ein O, S o sae e J. A con olled ial o a wo-componen acellula , a i e-componen acellula , and a whole-cell pe ussis accine. N Engl J Med 1996;334:349–55. 3. Olin P, Rasmussen F, Gus a sson L, Hallande HO, Heijbel H. Randomised con olled ial o wo-componen , h ee-componen , and i e-componen acellula accines compa ed wi h whole-cell pe ussis accine. Lance 1997;350:1569–77. 4. T oll o s B, Ta ange J, Lage ga˚ d T, e al. A placebo-con olled ial o a pe ussis- oxoid accine. N Engl J Med 1995;333:1045–50. 5. Giuliano M, Mas an onio P, Giammanco A, Pisci elli A, Salmaso S, Was- silak SG. An ibody esponses and pe sis ence in he wo yea s a e im- muniza ion wi h wo acellula accines and one whole-cell accine agains pe ussis. J Pedia 1998;132:983–8. 6. Salmaso S, Mas an onio P, Wassilak SG, e al. Pe sis ence o p o ec ion h ough 33 mon hs o age p o ided by immuniza ion in in ancy wi h wo h ee-componen acellula pe ussis accines. Vaccine 1998;16:1270–5. 7. S o sae e J, Hallande HO, Gus a sson L, Olin P. Le els o an i-pe ussis an ibodies ela ed o p o ec ion a e household exposu e o Bo de ella pe ussis. Vaccine 1998;16:1907–16. 8. Mills KHG, Ryan M, Ryan E, Mahon BP. A p edic i e model o pe ussis accine e icacy in child en e eals complemen a y oles o humo al and cell media ed immuni y in p o ec ion. In ec Immun 1998;66:594–602. 9. Redhead K, Wa kins J, Ba na d A, Mills KHG. E ec i e immuniza ion agains Bo de ella pe ussis espi a o y in ec ion in mice is dependen on induc ion o cell-media ed immuni y. In ec Immun 1993;61:3190–8. 10. Mills KHG, Ba na d A, Wa kins J, Redhead K. Cell media ed immuni y o Bo de ella pe ussis: ole o Th1 cells in bac e ial clea ance in a mu ine espi a o y in ec ion model. In ec Immun 1993;61:399–410. 11. Mahon BP, Sheahan BJ, G i in F, Mu phy G, Mills KHG. A ypical disease a e Bo de ella pe ussis espi a o y in ec ion o mice wi h a ge ed dis- up ions in IFN-g ecep o o immunoglobulin mchain genes. J Exp Med 1997;186:1843–51. 12. Che y JD, Go nbein J, Heininge U, S eh K. A sea ch o se ologic co - ela es o immuni y o Bo de ella pe ussis cough illnesses. Vaccine 1998;16: 1901–6. 13. Ausiello CM, Lande R, U bani F, e al. Cell-media ed immune esponses in ou -yea -old child en a e p ima y immuniza ion wi h acellula pe ussis accines. In ec Immun 1999;67:4064–71. 14. Ausiello CM, U bani F, La Sala A, Lande R, Cassone A. Vaccine- and an igen-dependan ype 1 and ype 2 cy okine induc ion a e p ima y accina ion o in an s wi h whole-cell o acellula pe ussis accines.In ec Immun 1997;65:2168–74. 15. Ryan M, Mu phy G, Ryan E, e al. Dis inc Th cell sub ypes induced wi h whole-cell and acellula pe ussis accines in child en. Immunology 1998;93:1–10.