journa l homepage: www.e lsev ier .com/ locate /vacc ine
D4+ T-cell responses to an oral inactivated cholera vaccine in young children incholera endemic country and the enhancing effect of zinc supplementation
anvir Ahmeda,b, Mohammad Arifuzzamana, Michael Lebensb, Firdausi Qadria, Anna Lundgrenb,∗
International Centre for Diarrhoeal Disease Research (ICDDR,B), Bangladesh, GPO Box 128, Dhaka 1000, BangladeshGothenburg University Vaccine Research Institute (GUVAX) and Department of Microbiology and Immunology, The Sahlgrenskacademy at the University of Gothenburg, Box 435 S-40530, Gothenburg, Sweden
r t i c l e i n f o
rticle history:eceived 12 May 2009eceived in revised form 26 August 2009ccepted 8 October 2009vailable online 17 October 2009
Immunization of young children with the oral inactivated whole cell cholera vaccine Dukoral® containingrecombinant cholera toxin B subunit (CTB) induces antibody responses which can be further enhanced byzinc supplementation. We have investigated if immunization with the cholera vaccine induces specificT-cell responses in young children and also whether zinc supplementation influences these responses.Bangladeshi children (10–18 months old) received vaccine alone, vaccine together with zinc supplemen-tation or only zinc. T-cell blast formation indicating a proliferative response was analyzed by the flowcytometric assay of cell-mediated immune response in activated whole blood (FASCIA) and cytokineswere measured by ELISA. Stronger T-cell responses were detected if a modified CTB molecule (mCTB)with reduced binding to GM1 ganglioside was used for cell stimulation compared to normal CTB. Aftervaccination, CD4+ T cells responded to mCTB with significantly increased blast formation (P < 0.01) and
oung children IFN-� production (P < 0.05) compared to before vaccination. No responses to mCTB were detected in chil-dren receiving zinc alone (P > 0.05). The IFN-� production was significantly higher (P < 0.01) but the blastformation comparable (P > 0.05) in children receiving zinc plus vaccine compared to in children receivingvaccine alone. The vibriocidal antibody responses induced by the vaccine were also significantly higher inchildren receiving zinc supplementation (P < 0.001). Our results thus show that oral cholera vaccinationinduces a Th1 T-cell response in young children, and that the IFN-� as well as the vibriocidal antibody
ed by
responses can be enhanc
. Introduction
The oral inactivated whole cell cholera vaccine (Dukoral®),hich contains recombinant cholera toxin B subunit (CTB), has
een extensively tested in field trials both in adults and children1,2]. The vaccine gives rise to mucosal IgA responses directedgainst CTB as well as to Vibrio cholerae lipopolysaccharide (LPS)nd other bacterial components [3–5]. The vaccine also inducesobust vibriocidal and anti-toxin antibody responses in serum5–7]. Less is known about the T-cell responses induced by thisaccine. In mice, B-cell responses to cholera toxin (CT) are strictlyependent on the presence of CD4+ T-cells [8]. CTB-specific T-cellesponses have also been detected in in vitro cultures of human
eripheral blood mononuclear cells after oral cholera vaccination9].
The oral cholera vaccine has previously mainly been used indults and older children, but we have recently shown that the
vaccine is also safe and immunogenic in young children downto 6 months of age [10]. However, the vaccine has been shownto confer more long lasting protection in adults than in childrenin Bangladesh [1]. Many vaccines are less immunogenic and pro-tective in children than in adults, which is thought to partly bea result of immature lymphocytes and antigen presenting cells(APCs) [11–13]. However, more distinct hyporesponsiveness hasbeen observed in children in developing compared to developedcountries to several oral vaccines, including rotavirus and polio vac-cines [14–16]. A number of factors may contribute to the reducedimmunogenicity in these children, including poor nutritional sta-tus, micronutrient deficiencies, breast-feeding habits or parasiticinfections [16]. Our recent studies have documented that theimmunogenicity of the oral cholera vaccine, as reflected by vibrioci-dal antibody responses, can be improved in young (10–18 months)[10] and in older children (2–5 years old) [3], with interventions
such as brief modification of the breast-feeding pattern [10] orzinc supplementation [6,10]. It is well established that zinc influ-ences multiple aspects of the immune system, including the normaldevelopment, differentiation, and function of immune cells [17,18],but the mechanisms responsible for the positive effects of zinc
Fig. 1. Schedule for vaccination, zinc supplementation and blood sampling. On theindicated days (D), samples were collected and/or vaccine was administered. T-cella(tw
teid
ietmoeco
2
2
DhaiheompTcaewtUp
2
O(mA
nd antibody responses as well as zinc levels were analyzed before (pre) and afterpost) vaccination and zinc supplementation on the indicated days. For determina-ion of cumulative responder frequencies for antibody responses, plasma samplesere also analyzed 7 days after administration of the first and second vaccine dose.
reatment observed after vaccination as well as in infectious dis-ases are still not well understood. Furthermore, it is still unclearf zinc supplementation only promotes immune responses in zinceficient individuals.
The aim of this study was to determine if cholera vaccinationnduces a specific T-cell response in young children in a cholerandemic country and to analyze the influence of zinc supplemen-ation on these responses. Since T-cell responses are difficult to
easure in infants and young children due to the limited amountf blood that can be obtained for testing, we modified a recentlystablished flow cytometric assay, FASCIA (flow cytometric assay ofell-mediated immune response in activated whole blood), whichnly requires small volumes of blood, for our studies.
. Materials and methods
.1. Study participants
The study was conducted in an urban slum area (Mirpur) inhaka, Bangladesh, between December 2007 and April 2008. Sixtyealthy male (n = 27) and female children (n = 33), 10–18 months ofge, living under similar socio-economic conditions, were enrolledn the study. Children were excluded from the study if they had aistory of gastrointestinal disorder, suffered from any diarrheal dis-ase in the past 2 weeks, had febrile illness in the preceding weekr had received antibiotic treatment at least 7 days prior to enroll-ent. Children ≤2 SD (weight/length as NCHS) or whose stool were
ositive for common enteric pathogens [19] were also excluded.he study physician assessed the general health condition of thehildren prior to enrollment in the study. Six healthy Swedishdults (mean age 34.7 ± 5.3 years, 2 males) were also recruited forstablishment of methods for analysis of T-cell responses. The studyas approved by the Research Review and Ethical Review Commit-
ees of ICDDR, B and the Ethical Committee for Human Research,niversity of Gothenburg. Informed consent was obtained from thearents or by the participants before enrollment.
.2. Vaccination and zinc treatment
The children were randomly assigned in three groups (Fig. 1).ne group received 2 doses of the oral cholera vaccine Dukoral®
SBL Vaccines, Sweden) at a 2-week interval without zinc supple-entation (vaccine only group; no zinc placebo was administered).second group received both zinc supplementation and vac-
8 (2010) 422–429 423
cine (zinc plus vaccine group). Individuals in this group weresupplemented with zinc (20 mg/day elemental zinc syrup contain-ing zinc acetate; two teaspoons; ACME Laboratories Ltd., Dhaka,Bangladesh) every day for 42 days starting 3 weeks before adminis-tration of the first dose of vaccine until 1 week after the second dose.A third group received only zinc supplementation for 42 days (zinconly group, no vaccine placebo was administered). The vaccine con-sists of 1 × 1011 inactivated V. cholerae O1 bacteria, and 1 mg ofrecombinantly produced CTB. Immediately before use, each vaccinedose (3 ml) was mixed with 20 ml of standard bicarbonate buffer.Food or drink was not allowed for 1 h before and 1 h after vacci-nation. Each vaccinee served as his/her own control based on thepreimmune responses. Adult Swedish volunteers were vaccinatedwith the Dukoral® vaccine according to the standard procedure [7].
2.3. Follow up
Study children were monitored for side effects such as diarrhea,vomiting and fever for 1 h after vaccination by the study physicianand the following 3 consecutive days after each vaccine dose byhealth assistants. Study staff carried out active surveillance eachday by home visits. For the zinc intervention studies, mothers wereinstructed to feed their children with zinc syrup daily at home. Thereliability of feeding was verified weekly by home visits by healthassistants.
2.4. Collection of specimens
Three ml of venous blood were collected in lithium-heparinvacutainer tubes from children prior to (pre-immune) immu-nization and/or initiation of zinc supplementation (Fig. 1). Fromthe immunized children, blood/plasma samples were also col-lected 7 days after the second vaccine dose (post-immune). Thepre- and post-immunization samples were used for analysis of T-cell responses. Plasma samples were also analyzed for antibodyresponses 7 days after administration of the first as well as thesecond vaccine dose. From children in the zinc only group, bloodsamples were collected on day 0 and day 42 of the zinc treatment,corresponding to the sampling time points in the zinc plus vac-cine group, and these samples were used for anlaysis of both T-and B-cell responses. From the adult Swedish volunteers, 50 ml ofblood was collected before and 7 and 14 days after the second vac-cine dose. Serum and plasma separated from blood were stored at−70 ◦C for zinc and antibody analyses, respectively.
2.5. Analysis of T-cell proliferation
In the current study, we adopted and optimized the FASCIAtechnique [20] for analysis of T-cell responses to cholera vaccina-tion. This technique allows analysis of T-cell blast formation, whichindicates a proliferative response, after stimulation with specificantigens using 50 �l of whole blood per antigen tested. Briefly,heparinized blood was diluted to a final dilution of 1:10 in DMEMF12 medium (Invitrogen, USA) supplemented with 10% fetal calfserum (HyClone, PerBios Science, Belgium) and 0.5 mg/ml gentam-icin (Sigma). The cells were cultured at 37 ◦C and 5% CO2 in 5 mlpolystyrene tubes (BD, USA) in the presence or absence of antigens(see below). After 6 days, cell culture supernatants were collectedand the cells were stained with anti-CD3-APC, anti-CD4-PerCP andanti-CD8-FITC antibodies (BD, USA), the red blood cells were lysedwith ammonium chloride and the samples were washed and fixed
in paraformaldehyde (BD, USA). The samples were analyzed usinga FACSCalibur machine (BD, USA) and the FlowJo analysis software(Tree Star Inc., USA). The numbers of blast forming CD3+ CD4+ T-cells acquired in each sample during 120 s were determined. Theflow rate of the flow cytometer was determined using Truecount
4 ccine 28 (2010) 422–429
bC
trbd
2
m(itfmiawsai[pgwp2tpaPa
Sdf[raSUwCc
2
c[CuaufCmwc
2
l
Table 1Baseline characteristics of study children.
eads (BD, USA) and the results were expressed as the numbers ofD4+ T-cell blasts/100 �l of sample.
For initial setup experiments on vaccinated Swedish volunteers,he FASCIA assay was compared to a standard thymidine incorpo-ation assay using peripheral mononuclear cells (PBMCs) isolatedy Ficoll-paque. The thymidine assay was performed as previouslyescribed [21].
.6. Antigens
For initial setup experiments, the responses to recombinant nor-al CTB were compared to those induced by a modified CTB antigen
mCTB), which has a single amino acid substitution (Gly-33 → Asp)n the binding site for GM1 ganglioside [22]. The FASCIA responseso mCTB were stronger than to unaltered CTB and mCTB was there-ore used in subsequent T-cell stimulation experiments. The mCTB
olecule has previously been shown to have considerably dimin-shed affinity for GM1 [22]. We have generated a similar CTB mutantnd expressed it in a V. cholerae-based expression system. The geneas expressed constitutively from the tac promoter in the O1 clas-
ical strain JS1569 in which the ctxA gene has been deleted [23]. Thessembled protein is secreted into the growth medium from whicht was precipitated using sodium hexametaphosphate at low pH24]. The precipitated protein was re-dissolved in 20 mM sodiumhosphate pH 7.4. Non-dissolved debris was removed by centrifu-ation followed by filtration through a 0.45 �m filter. The proteinas then purified by ion exchange chromatography. The elutedrotein was finally subjected to gel filtration using a Superdex00 column and dialysed against PBS before being concentratedo a final concentration of 1 mg protein per ml. The mCTB proteinreparation gives a single peak in analytical FPLC (gel filtration) andppears homogenous when visualized by coomassie blue in SDS-AGE. The protein preparation contains less than 0.01% LPS (w/w),s determined by a Limulus test.
Recombinant unaltered CTB was provided by SBL (SBL Vaccines,tockholm, Sweden). The cholera membrane protein (MP) was pro-uced by sonication of V. cholerae (O1, El Tor, Ogawa strain X25049)ollowed by differential centrifugation, as previously described25]. Mass spectrometry analyses of the MP show that this prepa-ation contains a mixture of a large number of different bacterialntigens, but CTB could not be detected. Tetanus toxoid (TT, Statenserum Institute, Denmark) and phytohemagglutinin (PHA, Remel,SA) were used as positive controls in the T-cell assay. T cellsere stimulated with the antigens at the following concentrations;TB and mCTB; 10 �g/ml (in setup experiments also 100 �g/ml),holera MP; 1 and 10 �g/ml, TT 10 �g/ml and PHA; 1 �g/ml.
.7. Antibody analysis
Plasma samples were tested for vibriocidal antibodies using V.holerae O1, El Tor, Ogawa (strain X25049) as the target organism4]. Plasma specimens were also analyzed for antibodies specific forTB (unaltered CTB, provided by SBL Vaccines, Stockholm, Sweden)sing GM1 ELISA in all individuals [7]. Antibody responses to CTBnd mCTB were also analyzed and compared in some individualssing plates directly coated with the two different molecules. Weound that the responses to mCTB were only slightly lower than toTB and that individuals who responded to CTB also responded toCTB. Antibody responses to LPS (prepared from strain X 25049)ere also determined by ELISA [26,27]. Pooled human sera from
holera patients were used as control for interassay variations [4].
.8. Cytokine analysis
The levels of IFN-� [21] and IL-13 (R & D systems Inc., Minneapo-is, USA) were measured in culture supernatants using ELISA and
a Data given for the 18 volunteers who completed the study.b Zinc deficiency is defined as ≤0.7 mg/l zinc in serum.c Data given as mean ± SD.
the levels of IL-4, IL-5, IL-2, IL-10 and TNF-� were measured bythe cytometric bead array procedure (BD Pharmingen). Cytokineanalyses were performed on supernatants collected after 6 daysof incubation, since our previous results suggest that both Th1and Th2 cytokine responses to cholera and control antigens canbe measured at this time point ([28] and A. Lundgren, Unpublishedobservations).
2.9. Analysis of zinc levels
For analysis of serum zinc levels, serum was collected in zincfree tubes and stored at −70 ◦C until analysis. Serum zinc levelswere determined at baseline for all vaccinated children and at theend of the study period for the zinc only and the zinc plus vaccinegroups. The levels of zinc (divalent cations) were measured by flameatomic absorption spectrophotometry [29,30] after digestion andextraction with zinc deficiency being defined as levels ≤0.7 mg/l[31].
2.10. Data analysis
Statistical analyses were carried out using the SigmaStat 3.1 pro-gram (SPSS Systat Software, Inc.). Children with ≥2-fold antibodyresponses to CTB or LPS in ELISA and ≥4-fold increase of vibriocidalantibodies compared to the prevaccination levels were consideredas responders [7]. The cumulative responder frequency for antibodyresponses was defined as the percentage of individuals respondingat any of the time points studied.
Paired samples were assessed by the Wilcoxon signed rank test,non-paired samples by the Mann–Whitney U-test and the propor-tion of responses using the �2 or the Fisher exact test. Correlationanalyses were performed using the Spearman test. P values <0.05were considered to be statistically significant.
3. Results
3.1. Study children
Among the 60 study children enrolled in the study, 58 com-pleted the follow up visits. In accordance with previous studies ofDukoral® in this age group [10], none of the children in any of thedifferent groups had any adverse events after vaccination.
Children in the three different intervention groups were com-parable with respect to age, weight and length (Table 1). Theaverage baseline zinc levels did not differ significantly betweenthe study groups (P > 0.05 for all comparisons) (Table 2), althoughbaseline zinc deficiency tended to be somewhat more prevalentin the vaccine only group (65%) compared to vaccine plus zinc
(50%) and the zinc only (40%) groups (Table 1). All children whoreceived zinc were zinc sufficient at the end of the supplementa-tion period (Table 2). Children who were zinc sufficient before thestudy also increased their serum zinc levels during the supplemen-tation period, although to a lesser extent than the zinc deficient
T. Ahmed et al. / Vaccine 28 (2010) 422–429 425
Table 2Serum zinc status before (pre) and after (post) zinc supplementation in zinc deficient and zinc sufficient children.
a Data is expressed as mean ± SD.b Significantly different (P < 0.05) when comparing zinc levels before and after zin
hildren. After zinc supplementation the serum zinc levels wereomparable in previously zinc deficient and zinc sufficient childrenTable 2).
.2. Establishment of the FASCIA assay for analysis of T-cellesponses to CTB
Initial establishment of techniques for analysis of T-cellesponses to the oral cholera vaccine was performed using bloodamples collected from adult Swedish volunteers before and afteraccination with the Dukoral® vaccine. Results from these experi-ents showed that PBMCs obtained 1 week after administration of
he second vaccine dose responded with increased proliferation totimulation with normal CTB compared to PBMCs obtained beforeaccination, as determined by a traditional thymidine incorpora-ion assay (Data not shown). However, only small blood volumesan be collected from children, which made it difficult to use tra-itional methods based on stimulation of PBMCs for analysis of-cell responses in this study. Therefore, we used the FASCIA assaynstead, in which T cells are stimulated in small volumes of wholelood [20]. Using the FASCIA assay, increased numbers of CD4+-cell blasts were detected in samples stimulated with CTB afteraccination of the adult Swedish volunteers, but the number oflasts were very low (Fig. 2A) and close to the detection limit.ince one major difference between whole blood and PBMCs ishe lack of granulocytes in PBMCs, we hypothesized that the lowesponses to CTB detected in whole blood compared to PBMCs maye a result of bining of CTB to GM1 ganglioside on granulocytes,hich could prevent the CTB from being taken up and presented
y APCs, thereby limiting the responses detected in the whole bloodulture system. To determine if we could overcome this poten-
ial binding effect, we tested a CTB molecule with reduced bindingapacity to GM1 in our culture systems. The mCTB molecule hadeen constructed and produced previously in our laboratory fortudies of binding of enterotoxins to different receptors [32]. Using
ig. 2. Establishment of the FASCIA assay for evaluation of T-cell responses to oralholera vaccination. Diluted whole blood collected from adult Swedish volunteersas stimulated with 10 or 100 �g/ml unaltered CTB (A) or mCTB (B). T-cell responsesere analyzed before immunization (open bars) and 7 (filled bars) and 14 days
hatched/gray bars) after administration of the second dose of the Dukoral® vacciney counting the numbers of CD3+ CD4+ T-cell blasts. The data shown are withoutubtraction of the background responses to medium alone. The bars indicate theeans ± SEM (n = 3).
1.11 ± 0.20b 0.75 ± 0.16 1.11 ± 0.24b
plementation, as determined by the Wilcoxon signed rank test.
this molecule in the FASCIA assay, more than 15-fold higher vac-cine induced T-cell responses could be detected compared to whenunaltered CTB was used for the in vitro testing (Fig. 2B). Our pre-liminary results also indicate that the mCTB molecule gave similarresponses as the unaltered CTB molecule in PBMC assays (Datanot shown). A significant correlation was also found between theresponses to mCTB molecule detected by FASCIA and by a tradi-tional thymidine incorporation assay (R = 0.92, P = 0.01, n = 6). ThemCTB molecule was therefore used to analyze T-cell responses inthe FASCIA assay.
3.3. Vaccine specific T-cell responses
Using the FASCIA assay to analyze T-cell responses to mCTB inBangladeshi children, we found that immunization with two dosesof Dukoral® gave rise to significant, i.e. 3- to 4-fold increased num-bers of CD4+ T-cell blasts in the vaccine only (P = 0.01) as well asthe zinc plus vaccine (P = 0.005) groups compared to the prevac-cination responses (Fig. 3A). No significant blast formation wasobserved among CD8+ T cells in response to immunization in any ofstudy groups (Data not shown). Vaccination induced significantlyincreased production of IFN-� in response to mCTB stimulation inboth the vaccine only (P = 0.03) and the zinc plus vaccine groups(P < 0.01) compared to the secretion detected before vaccination(Fig. 3B). The IFN-� levels and numbers of blasts formed in responseto vaccination correlated significantly with each other in both vac-cination groups (Vaccine only; r2 = 0.62, P = 0.003, and zinc plusvaccine; r2 = 0.80, P < 0.001). No differences in blast responses tomedium alone were detected between pre- and post-vaccinationsamples in the vaccine plus zinc or the zinc only groups (P > 0.05),but unstimulated cells in the vaccine only group showed a smallbut significant increase in the numbers of CD4+ T-cell blasts aftercompared to before vaccination (median 220 and 90 CD4+ T-cellblasts/100 �l sample, respectively; P = 0.02). However, the vaccine-induced responses to mCTB were significantly larger than theincreases in background responses in this (P < 0.001) as well as inthe two other study groups. The IFN-� levels produced by unstimu-lated cells were also low and comparable (P > 0.05) before and aftervaccination in all study groups. A small but significant decrease inthe proliferative responses to mCTB was found in the zinc only con-trol group during the study period, whereas the IFN-� response wascomparable in this group over time (Fig. 3A and B). No detectablelevels of IL-13 were found in culture supernatants in any of the vol-unteers. The levels of IL-4, IL-5, IL-2, IL-10 and TNF-� were analyzedin supernatants collected from 4 individuals from each group andwere found to be close to the detection limits in all samples (Datanot shown).
The proliferative T-cell response to cholera MP did not differsignificantly from the baseline responses in any of the vaccination
groups or in the zinc control group (P > 0.05, Data not shown). Cellsfrom almost all children responded well to stimulation with thecontrol antigen TT and comparable (P > 0.05) TT-responses weredetected before and after vaccination in all study groups (Medianresponses 860 and 650 CD4+ blasts/100 �l of sample in the zinc
426 T. Ahmed et al. / Vaccine 28 (2010) 422–429
Fig. 3. T-cell responses to mCTB (10 �g/ml) before (pre) and after (post) administration of two doses of oral cholera vaccine with or without zinc supplementation inB d usino medib ant din used
gtwp(1PPni
3
citrpdsn
TV
t
angladeshi children. T-cell proliferation (A) and IFN-� responses (B) were detectenly (Vacc) and zinc plus vaccine (Zn + Vacc) groups. The horizontal bars indicateackground responses to medium alone. The asterisks indicate statistically significot significant). The Wilcoxon signed rank and Mann–Whitney rank sum tests were
roup, 1000 and 910 in the vaccine group and 870 and 1600 inhe zinc plus vaccine group). Comparable levels of blast formationere also detected in response to the mitogen PHA in pre- andost-vaccination samples in the zinc and zinc plus vaccine groups7720 and 12,250 CD4+ blasts/100 �l of sample and 12,480 and4,800, respectively), but a significant increase was noted in theHA response in the zinc plus vaccine group (5890 and 14,070,= 0.03), possibly as a result of difficulties in gating the very largeumber of blasts with varying forward and side scatter character-
stics which were formed in response to the mitogen.
.4. The influence of zinc supplementation on T-cell responses
Supplementation with 20 mg of zinc before and during the vac-ination period did not affect the levels of T-cell blast formationnduced by mCTB (P > 0.05, Fig. 3A), but enhanced the produc-ion of IFN-� about 6-fold (P = 0.008, Fig. 3B). Although the IFN-�
esponses were significantly higher in the zinc plus vaccine com-ared to the vaccine only group, we could not detect any clearifferences in T-cell responses between children who were zincufficient or zinc deficient at the start of the study (P > 0.05, Dataot shown).
able 3accine specific plasma antibody responses in the different intervention groups.
a Pre = day 0 and post = 7 days after the 2nd dose of vaccine or at the end of the zinc supb % = cumulative responder frequency.c Significantly different (P < 0.05) when comparing pre- and post-immunization titers,d Significantly different (P < 0.05) when comparing the titers in the vaccine only group
est.
g FASCIA and ELISA assays, respectively in children in the zinc only (Zn), vaccinean responses at each time point. The data shown are without subtraction of thefferences between pre- and post-immunization responses (*P < 0.05; **P < 0.01; ns:for statistical evaluation.
3.5. Vaccine specific antibody responses and the influence of zincsupplementation
Immunization with two doses of the Dukoral® vaccine alsoresulted in significant increases in vibriocidal as well as CTB-specific IgG and IgA responses in both the vaccine only and thezinc plus vaccine groups (Table 3). No vaccine specific antibodyresponses were observed in the zinc only control group over time.
Zinc supplementation resulted in as a mean about 3-fold highervibriocidal antibody levels and a 50% increased responder fre-quency compared to the vibriocidal responses observed in thevaccine only group. In contrast, the vast majority (94–100%) of thevaccinated individuals responded with CTB-specific IgA and IgGantibodies even in the absence of zinc supplementation (Table 3).Although the post-immunization levels of CTB IgG were 2-foldhigher in the zinc plus vaccine group compared to the vaccine group(Table 3), the prevaccination IgG titers differed in a similar way
and there was no difference in the fold increases in antibody levelswhen the two groups were compared. No effects of zinc supple-mentation were observed on the CTB-specific IgA responses. Theeffect of zinc on the vibriocidal responses was primarily seen inthe zinc deficient children, since zinc supplementation significantly
as determined by the Wilcoxon signed rank test.with the zinc plus vaccine group, as determined by the Mann–Whitney rank sum
T. Ahmed et al. / Vaccine 2
Fig. 4. Vibriocidal antibody responses in plasma before (pre) and after (post) admin-istration of two doses of oral cholera vaccine with or without zinc supplementation.Children in the zinc only (Zn), vaccine only (Vacc) and zinc plus vaccine (Zn + Vacc)study groups were divided into zinc deficient (Zn Def) and zinc sufficient (Zn Suf)subgroups based on their baseline serum zinc levels. The asterisks indicate sta-tmir
eazttwz
ordeL
ntr
4
vi1ctab1stihtfpsC
istically significant differences between pre- and post-immunization levels, i.e.agnitude of responses in the respective vaccine group (*P < 0.05, ns; not signif-
cant). The horizontal lines indicate the geometric mean titers. The Wilcoxon signedank and Mann–Whitney rank sum tests were used for statistical evaluation.
nhanced the responses in this subgroup, whereas the responsesmong zinc sufficient children were comparable with and withoutinc supplementation (Fig. 4). Among the zinc deficient children,he responder frequencies increased from 54% to 89% as a result ofhe supplementation (P < 0.05), whereas the responder frequenciesere comparable in the zinc sufficient children with and without
inc supplementation (Fig. 4).Although significant LPS-specific IgA responses were only
bserved in the vaccine plus zinc group, the difference in anti-LPSesponses between the vaccine only and the zinc plus vaccine groupid not reach statistical significance (Table 3). No significant influ-nces of baseline zinc status or zinc supplementation on the CTB- orPS-specific antibody responses were observed (Data not shown).
The antibody responses (vibriocidal or anti-toxin antibodies) didot correlate significantly with the T-cell responses (blast forma-ion or IFN-� production) (P > 0.05) in any of the study groups whenesponses were compared on an individual level.
. Discussion
In this study, we show that immunization with the oral inacti-ated whole cell cholera vaccine, Dukoral®, does not only inducencreased levels of anti-toxin and vibriocidal antibodies in children0–18 months of age, but also induces a distinct response amongirculating CD4+ T-cells. The responding T cells primarily secretedhe Th1 type cytokine IFN-�, but little Th2 cytokines. These findingsre consistent with the previous demonstration of increased num-ers of IFN-� producing cells in the small intestine of adult Swedesweek after administration of Dukoral® [33]. However, the present
tudy for the first time reveal the capacity of the Dukoral® vaccineo induce a vaccine specific Th1 T-cell response in young childrenn an endemic country. In general, helper T-cell responses oftenave low magnitudes in children [34] and Th1 responses are par-
icularly weak, probably as a result of reduced production of IL-12rom APCs in children than in adults [34]. However, under appro-riate priming conditions, children can respond to vaccination withtrong Th1 responses, as illustrated by the responses to the bacillusalmette-Guerin vaccine in infants [35].
8 (2010) 422–429 427
Our demonstration of a T-cell response in children vaccinatedwith the oral cholera vaccine suggests that T cells can influencethe B-cell responses to the vaccine. Effects of IFN-� on the immuneresponse include enhanced tissue recruitment of lymphocytes andother immune cells via increased expression of chemokines andadhesion molecules, elevated production of the poly immunoglob-ulin receptor that transports IgA across the intestinal epitheliumand increased permeability of the epithelial barrier [36–38]. How-ever, in contrast to the Th1 response detected in vaccinated childrenin this study, we have recently observed that both IFN-� andIL-13 producing T cells contribute to the immune responses inadults with cholera disease [28]. It will therefore be importantto determine the relative importance of Th1 and Th2 responsesfor short-term protection as well as development of memoryresponses to cholera in both children and adults in continuedstudies.
Compared to antibody responses, cell-mediated immunity isoften more difficult to measure and these assays often requirerelatively large numbers of cells, which may not be available inpaediatric studies. To enable studies of T-cell responses in chil-dren in this study, we used the recently described FASCIA methodwhich only requires small volumes of whole blood. To be able toclearly detect T-cell responses using this assay, a modified mCTBmolecule with reduced capacity to bind to the GM1 receptor wasused [22]. In the FASCIA assay, mCTB gave about 15-fold strongerresponses than unaltered CTB. We believe that this difference maybe due to the high frequency of granulocytes in whole blood whichmay bind the unaltered CTB molecule and prevent it from beingtaken up by APCs. This idea is supported by recent data show-ing that PBMCs, which lack granulocytes, respond similarly tomCTB and CTB (A. Lundgren, Unpublished results). Ongoing stud-ies aim to fully determine the capacity of the two molecules tobind and stimulate human blood cells. Nevertheless, since we aswell as others [20] have shown that responses to many other anti-gens, including tetanus toxoid, staphylococcal enterotoxins andPHA, induce comparable T-cell proliferation and cytokine profileexpression in whole blood and PBMC assays, we believe that theFASCIA assay can be a useful tool for evaluation of T-cell responsesto many different infections and vaccines, particularly in youngchildren.
In contrast to the significant T-cell responses observed to mCTB,no increase in responsiveness could be detected to cholera MP aftervaccination. We have recently shown that adult patients with natu-ral cholera disease respond to MP with increased proliferation andcytokine production [28], suggesting that natural disease inducesstronger T-cell responses to the MP antigen than oral vaccination.However, adult patients may be more primed by earlier choleraexposure than young children. The children participating in thisstudy are however likely to have experienced many other entericinfections, including enterotoxigenic E. coli (ETEC). Children livingin the same area of Dhaka have recently been shown to experi-ence 2.3 episodes of ETEC diarrhoea per year during their two firstyears of life [39]. It is possible that infection with ETEC strains thatexpress the heat labile toxin (LT), which is highly homologous withCT, may also have primed some of the children to respond betterto mCTB/CTB after vaccination in the present study. This may alsoexplain why some of the volunteers exhibited relatively strong anti-body and T-cell responses to mCTB/CTB even before vaccination.However, even in children with relatively high baseline responsesto mCTB/CTB, the vaccination enhanced the responses further, sug-gesting that immune responses were promoted by the vaccination.
Both antibacterial and anti-toxin antibodies have been shown tosynergistically contribute to cholera protection in animal models[5]. However, field studies have suggested that killed cholera vac-cines not containing CTB also are protective [40]. Further studiesare required to determine if T-cell responses to different fractions
4 ccine 2
oCc
vdaptswtrtwttdmtiIwdapptt
dsrmolafiimidl
msrszdcdgt
enbtrtdatos
[
[
[
[
[
[
[
[
[
[
[
28 T. Ahmed et al. / Va
f the cholera bacteria and/or purified cholera antigens other thanTB may be important for and correlate with protection againstholera infection.
About 50% of the children in this study were zinc deficient. Pre-ious studies have shown that zinc deficiency is associated withecreased primary as well as memory T-cell and B-cell responsesfter vaccination [41,42]. We have recently shown that zinc sup-lementation can enhance the vibriocidal antibody responses tohe oral cholera vaccine in 10–18 months old children [10]. In thistudy, we found that the IFN-� responses to mCTB after vaccinationere significantly stronger in children receiving zinc supplementa-
ion compared to children not receiving zinc. Consistent with theseesults, Th1 cells and IFN-� responses have previously been showno be influenced by differences in zinc levels, whereas Th2 cellsere not affected [43,44]. Zinc has also been reported to influence
he levels of T-cell proliferation in some studies [45,46], whereashis effect has not been observed by others [47]. In our study, weid not detect any influence of zinc on the proliferative responses toCTB. Similar to the results from our previous study [10], we found
hat vibriocidal but not anti-toxin antibody responses were signif-cantly enhanced by zinc supplementation in the present study.t is possible that the effect of zinc is more easily detectable for
eaker responses such as production of cytokines and vibrioci-al antibodies, compared to the antibody responses to CTB, wherelmost all individuals respond even in the absence of zinc sup-lementation. In agreement with this notion, a recent study of aneumococcal conjugate vaccine in Bangladeshi infants showedhat zinc supplementation only enhanced antibody responses tohe least immunogenic pneumococcal serotypes [48].
We also analyzed whether zinc sufficient and deficient chil-ren responded differently to the vaccine with and without zincupplementation, but could not see any clear differences in T-cellesponses between the different study groups. Since zinc supple-entation increased zinc levels in most of the children, regardless
f whether they were defined as zinc deficient or not, we specu-ate that zinc supplementation may enhance T-cell responses over
wide range of zinc levels and that intake of zinc may be bene-cial even in children with better micronutrient status. However,
n contrast to the T-cell responses, we observed that zinc supple-entation primarily enhanced the vibriocidal antibody responses
n zinc deficient children. This may suggest that T and B cells haveifferent zinc requirements, but this needs to be confirmed in a
arger study.Previous studies suggest that under some circumstances, zinc
ay also have inhibitory effects on T cells [18,49], and that zincupplementation during vaccination may hinder efficient T-cellesponses in zinc sufficient children [49,50]. Studies have alsohown that the antibody responses to CTB can be suppressed byinc intake in adults or 2–5-year-old children [3,51]. However, weid not observe any tendencies for zinc to suppress either B- or T-ell responses in the 10–18-month-old children studied here. Ourata therefore support the use of zinc supplementation in this ageroup, which is at high risk for micronutrient deficiency, withouthe need for prior assessment of the zinc status of each child.
In conclusion, this study shows that young children in a cholerandemic area develop a Th1 T-cell response to the CTB compo-ent of an oral cholera vaccine and that the IFN-� productiony the responding T cells can be enhanced by zinc supplemen-ation. Our data suggest that T cells may influence the antibodyesponses induced by the vaccine, but further studies are neededo determine the role of T cells for protection against cholera and for
evelopment of immunological memory after cholera vaccinationnd infection in this age group. Such studies may be facilitated byhe possibility to use the whole blood FASCIA assay for evaluationf T-cell responses, as illustrated by the results presented in thistudy.
[
8 (2010) 422–429
Acknowledgements
This work was supported by the Swedish Agency for Interna-tional Development and Corporation (Sida/SAREC), the Marianneand Markus Wallenberg Foundation through the support to GUVAXand the International Centre for Diarrhoeal Disease Research,Bangladesh (ICDDR, B). The Centre is supported by agencies andcountries that share its concern for the health problems of devel-oping countries.
References
[1] Clemens JD, Sack DA, Harris JR, Van Loon F, Chakraborty J, Ahmed F, et al. Fieldtrial of oral cholera vaccines in Bangladesh: results from three-year follow-up.Lancet 1990 Feb 3;335(8684):270–3.
[2] Lopez AL, Clemens JD, Deen J, Jodar L. Cholera vaccines for the developing world.Human vaccines 2008;4(March–April (2)):165–9.
[3] Qadri F, Ahmed T, Wahed MA, Ahmed F, Bhuiyan NA, Rahman AS, et al. Sup-pressive effect of zinc on antibody response to cholera toxin in children giventhe killed, B subunit-whole cell, oral cholera vaccine. Vaccine 2004;22(January(3–4)):416–21.
[4] Qadri F, Mohi G, Hossain J, Azim T, Khan AM, Salam MA, et al. Comparison ofthe vibriocidal antibody response in cholera due to Vibrio cholerae O139 Bengalwith the response in cholera due to Vibrio cholerae O1. Clin Diagn Lab Immunol1995;2(November (6)):685–8.
[5] Svennerholm AM, Jertborn M, Gothefors L, Karim AM, Sack DA, Holmgren J.Mucosal antitoxic and antibacterial immunity after cholera disease and afterimmunization with a combined B subunit-whole cell vaccine. J Infect Dis1984;149(June (6)):884–93.
[6] Albert MJ, Qadri F, Wahed MA, Ahmed T, Rahman AS, Ahmed F, et al. Supple-mentation with zinc, but not vitamin A, improves seroconversion to vibriocidalantibody in children given an oral cholera vaccine. J Infect Dis 2003;187(March(6)):909–13.
[7] Jertborn M, Svennerholm AM, Holmgren J. Safety and immunogenicity of anoral recombinant cholera B subunit-whole cell vaccine in Swedish volunteers.Vaccine 1992;10(2):130–2.
[8] Hornqvist E, Goldschmidt TJ, Holmdahl R, Lycke N. Host defense againstcholera toxin is strongly CD4+ T cell dependent. Infect Immun 1991;59(October(10)):3630–8.
[9] Castello-Branco LR, Griffin GE, Poulton TA, Dougan G, Lewis DJ. Characterizationof the circulating T-cell response after oral immunization of human volunteerswith cholera toxin B subunit. Vaccine 1994;12(January (1)):65–72.
10] Ahmed T, Svennerholm AM, Tarique AA, Sultana GN, Qadri F. Enhancedimmunogenicity of an oral inactivated cholera vaccine in infants in Bangladeshobtained by zinc supplementation and by temporary withholding breast-feeding. Vaccine 2009;27(February (9)):1433–9.
11] Gans HA, Arvin AM, Galinus J, Logan L, DeHovitz R, Maldonado Y. Deficiency ofthe humoral immune response to measles vaccine in infants immunized at age6 months. JAMA 1998;280(August (6)):527–32.
12] Upham JW, Rate A, Rowe J, Kusel M, Sly PD, Holt PG. Dendritic cell immaturityduring infancy restricts the capacity to express vaccine-specific T-cell memory.Infect Immun 2006;74(February (2)):1106–12.
13] Zinkernagel RM, Hengartner H. Regulation of the immune response by antigen.Science 2001;293(July (5528)):251–3.
14] Linhares AC, Lanata CF, Hausdorff WP, Gabbay YB, Black RE. Reappraisal ofthe Peruvian and Brazilian lower titer tetravalent rhesus-human reassortantrotavirus vaccine efficacy trials: analysis by severity of diarrhea. Pediatr InfectDis J 1999;18(November (11)):1001–6.
15] Rennels MB. Influence of breast-feeding and oral poliovirus vaccineon the immunogenicity and efficacy of rotavirus vaccines. J Infect Dis1996;174(September (Suppl. 1)):S107–11.
16] Sack D, Qadri F, Svennerholm AM. Determinants of responses to oral vaccinesin developing countries. In: Hernell O, editor. Annales Nestle. Umea: Karger;2008. p. 71–9.
17] Rink L, Haase H. Zinc homeostasis and immunity. Trends Immunol 2007;28(Jan-uary (1)):1–4.
18] Wellinghausen N, Kirchner H, Rink L. The immunobiology of zinc. ImmunolToday 1997;18(November (11)):519–21.
19] Qadri F, Ahmed T, Ahmed F, Begum YA, Sack DA, Svennerholm AM. Reduceddoses of oral killed enterotoxigenic Escherichia coli plus cholera toxin B subunitvaccine is safe and immunogenic in Bangladeshi infants 6–17 months of age:dosing studies in different age groups. Vaccine 2006;24(March (10)):1726–33.
20] Svahn A, Linde A, Thorstensson R, Karlen K, Andersson L, Gaines H. Devel-opment and evaluation of a flow-cytometric assay of specific cell-mediatedimmune response in activated whole blood for the detection of cell-mediated
immunity against varicella-zoster virus. J Immunol Methods 2003;277(June(1–2)):17–25.
21] Lundin BS, Johansson C, Svennerholm AM. Oral immunization with aSalmonella enterica serovar typhi vaccine induces specific circulating mucosa-homing CD4+ and CD8+ T cells in humans. Infect Immun 2002;70(October(10)):5622–7.
ccine 2
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
T. Ahmed et al. / Va
22] Jobling MG, Holmes RK. Analysis of structure and function of the B sub-unit of cholera toxin by the use of site-directed mutagenesis. Mol Microbiol1991;5(July (7)):1755–67.
23] Sanchez J, Holmgren J. Recombinant system for overexpression of cholera toxinB subunit in Vibrio cholerae as a basis for vaccine development. Proc Natl AcadSci USA 1989;86(January (2)):481–5.
24] Lebens M, Johansson S, Osek J, Lindblad M, Holmgren J. Large-scale productionof Vibrio cholerae toxin B subunit for use in oral vaccines. Biotechnology (NaturePublishing Company) 1993;11(December (13)):1574–8.
25] Helmuth R, Achtman M. Cell-cell interactions in conjugating Escherichia coli:purification of F pili with biological activity. Proc Natl Acad Sci USA 1978Mar;75(3):1237–41.
26] Qadri F, Wenneras C, Albert MJ, Hossain J, Mannoor K, Begum YA, et al. Compar-ison of immune responses in patients infected with Vibrio cholerae O139 andO1. Infect Immun 1997;65(September (9)):3571–6.
27] Westphal O, Jann K, Himmelspach K. Chemistry and immunochemistry of bac-terial lipopolysaccharides as cell wall antigens and endotoxins. Prog Allergy1983;33:9–39.
28] Bhuiyan TR, Lundin SB, Khan AI, Lundgren A, Harris JB, Calderwood SB, et al.Cholera caused by Vibrio cholerae O1 induces T-cell responses in the circulation.Infect Immun 2009;77(May (5)):1888–93.
29] Hotz C, Lowe NM, Araya M, Brown KH. Assessment of the trace element sta-tus of individuals and populations: the example of zinc and copper. J Nutr2003;133(May (5 Suppl. 1)):1563S–8S.
30] Smith Jr JC, Butrimovitz GP, Purdy WC. Direct measurement of zinc in plasmaby atomic absorption spectroscopy. Clin Chem 1979;25(August (8)):1487–91.
31] Gibson RS, Hess SY, Hotz C, Brown KH. Indicators of zinc status at the populationlevel: a review of the evidence. Br J Nutr 2008;99(June (Suppl. 3)):S14–23.
32] Backstrom M, Shahabi V, Johansson S, Teneberg S, Kjellberg A, Miller-Podraza H,et al. Structural basis for differential receptor binding of cholera and Escherichiacoli heat-labile toxins: influence of heterologous amino acid substitutions in thecholera B-subunit. Mol Microbiol 1997;24(May (3)):489–97.
33] Quiding M, Nordstrom I, Kilander A, Andersson G, Hanson LA, HolmgrenJ, et al. Intestinal immune responses in humans. Oral cholera vaccinationinduces strong intestinal antibody responses and interferon-gamma produc-tion and evokes local immunological memory. J Clin Invest 1991;88(July (1)):143–8.
34] Willems F, Vollstedt S, Suter M. Phenotype and function of neonatal DC. Eur JImmunol 2009;39(January (1)):26–35.
35] Vekemans J, Amedei A, Ota MO, D’Elios MM, Goetghebuer T, Ismaili J, etal. Neonatal bacillus Calmette-Guerin vaccination induces adult-like IFN-gamma production by CD4+ T lymphocytes. Eur J Immunol 2001;31(May (5)):1531–5.
36] Amin PB, Diebel LN, Liberati DM. T-cell cytokines affect mucosal immunoglob-ulin A transport. Am J Surg 2007;194(July (1)):128–33.
[
8 (2010) 422–429 429
37] Sollid LM, Kvale D, Brandtzaeg P, Markussen G, Thorsby E. Interferon-gammaenhances expression of secretory component, the epithelial receptor for poly-meric immunoglobulins. J Immunol 1987;138(June (12)):4303–6.
38] Brandtzaeg P, Bjerke K, Kett K, Kvale D, Rognum TO, Scott H, et al. Productionand secretion of immunoglobulins in the gastrointestinal tract. Ann Allergy1987;59(November (5 Pt 2)):21–39.
39] Qadri F, Saha A, Ahmed T, Al Tarique A, Begum YA, Svennerholm AM. Diseaseburden due to enterotoxigenic Escherichia coli in the first 2 years of life in anurban community in Bangladesh. Infect Immun 2007;75(August (8)):3961–8.
40] Thiem VD, Deen JL, von Seidlein L, Canh do G, Anh DD, Park JK, et al. Long-term effectiveness against cholera of oral killed whole-cell vaccine producedin Vietnam. Vaccine 2006;24(May (20)):4297–303.
41] Fraker PJ, Jardieu P, Cook J. Zinc deficiency and immune function. Arch Dermatol1987;123(December (12)):1699–701.
43] Beck FW, Prasad AS, Kaplan J, Fitzgerald JT, Brewer GJ. Changes in cytokineproduction and T cell subpopulations in experimentally induced zinc-deficienthumans. Am J Physiol 1997;272(June (6 Pt 1)):E1002–7.
44] Sandstead HH, Prasad AS, Penland JG, Beck FW, Kaplan J, Egger NG, et al. Zincdeficiency in Mexican American children: influence of zinc and other micronu-trients on T cells, cytokines, and antiinflammatory plasma proteins. Am J ClinNutr 2008;88(October (4)):1067–73.
45] Raqib R, Roy SK, Rahman MJ, Azim T, Ameer SS, Chisti J, et al. Effect of zincsupplementation on immune and inflammatory responses in pediatric patientswith shigellosis. Am J Clin Nutr 2004;79(March (3)):444–50.
46] Schlesinger L, Arevalo M, Arredondo S, Diaz M, Lonnerdal B, Stekel A. Effect ofa zinc-fortified formula on immunocompetence and growth of malnourishedinfants. Am J Clin Nutr 1992;56(September (3)):491–8.
47] Kramer TR, Udomkesmalee E, Dhanamitta S, Sirisinha S, Charoenkiatkul S, Tun-tipopipat S, et al. Lymphocyte responsiveness of children supplemented withvitamin A and zinc. Am J Clin Nutr 1993;58(October (4)):566–70.
48] Osendarp SJ, Prabhakar H, Fuchs GJ, van Raaij JM, Mahmud H, Tofail F, etal. Immunization with the heptavalent pneumococcal conjugate vaccine inBangladeshi infants and effects of zinc supplementation. Vaccine 2007;25(April(17)):3347–54.
49] Meydani A, Ahmed T, Meydani SN. Aging, nutritional status, and infection inthe developing world. Nutr Rev 2005;63(July (7)):233–46.
50] Overbeck S, Rink L, Haase H. Modulating the immune response by oral zinc
supplementation: a single approach for multiple diseases. Arch Immunol TherExp (Warsz) 2008;56(January–February (1)):15–30.
51] Karlsen TH, Sommerfelt H, Klomstad S, Andersen PK, Strand TA, Ulvik RJ, et al.Intestinal and systemic immune responses to an oral cholera toxoid B subunitwhole-cell vaccine administered during zinc supplementation. Infect Immun2003;71(July (7)):3909–13.
