Induction of long term mucosal immunological memory in humans by an oral inactivated multivalent enterotoxigenic Escherichia coli vaccine

Vaccine 34 (2016) 3132–3140

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Induction of long term mucosal immunological memory in humans by an oral inactivated multivalent enterotoxigenic Escherichia coli vaccine Anna Lundgren ∗ , Marianne Jertborn, Ann-Mari Svennerholm University of Gothenburg Vaccine Research Institute (GUVAX), Dept. of Microbiology and Immunology, University of Gothenburg, Box 435, 405 30 Gothenburg, Sweden

a r t i c l e

i n f o

Article history: Received 27 January 2016 Received in revised form 31 March 2016 Accepted 19 April 2016 Available online 28 April 2016 Keywords: ETEC Oral vaccine Immunological memory Mucosal immunity IgA

a b s t r a c t We have evaluated the capacity of an oral multivalent enterotoxigenic Escherichia coli (ETEC) vaccine (MEV) to induce mucosal immunological memory. MEV consists of four inactivated E. coli strains overexpressing the major colonization factors (CFs) CFA/I, CS3, CS5 and CS6 and the LTB-related toxoid LCTBA. Memory responses were analyzed by comparing the magnitudes and kinetics of intestine-derived antibody-secreting cell responses to a single dose of MEV in three groups of adult Swedish volunteers (n = 16–19 subjects per group) in a Phase I trial: non-immunized controls (I) and subjects who in a previous Phase I trial 13–23 months earlier had received two biweekly doses of MEV (II) or MEV + double mutant LT (dmLT) adjuvant (III). Responses against CFs and LTB were analyzed in antibodies in lymphocyte secretions (ALS) of blood mononuclear cells collected before (day 0) and 4/5 and 7 days after immunization. Specific circulating memory B cells present at the time of the single dose vaccination were also studied to determine if such cells may reflect mucosal memory. Considerably higher and significantly more frequent IgA ALS responses against all CFs and LTB were induced by the single vaccine dose in the previously immunized than in non-immunized volunteers. Furthermore, peak IgA ALS responses against all antigens were observed on days 4/5 in most of the previously immunized subjects whereas only a few previously non-vaccinated individuals responded before day 7. Priming with adjuvant did not influence memory responses. Circulating vaccine specific IgA memory B cells were not detected, whereas anti-toxin IgG memory B cells were identified 13–23 months after priming vaccination. We conclude that MEV induces functional mucosal immunological memory which remains at least 1–2 years. Furthermore, our results support that analysis of antibody-secreting cell responses after booster vaccination may be a useful approach to evaluate longstanding mucosal immunological memory in humans. Clinical trials registration: ISRCTN27096290. © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Abbreviations: AE, adverse event; ALS, antibodies in lymphocyte supernatants/secretions; ASC, antibody-secreting cell; BAFF, B cell activating factor; CF, colonization factor; CT, cholera toxin; CTB, cholera toxin binding subunit; dmLT, double mutant LT; ETEC, enterotoxigenic Escherichia coli; GM, geometric mean; LT, heat labile toxin; LTB, heat labile toxin binding subunit; LCTBA, CTB/LTB hybrid protein; LPS, lipopolysaccharide; MEV, multivalent ETEC vaccine; PBMCs, peripheral blood mononuclear cells; PPS, per protocol analysis set; PWM, pokeweed mitogen; SAC, Staphylococcus aureus protein A; SAS, safety analysis set; SIgA, secretory IgA; ST, heat stable toxin. ∗ Corresponding author. Tel.: +46 31 7866213; fax: +46 31 7866205. E-mail addresses: [email protected] (A. Lundgren), [email protected] (M. Jertborn), [email protected] (A.-M. Svennerholm).

Intense efforts are in progress to develop an effective vaccine against the most common bacterial cause of diarrhea, i.e. enterotoxigenic Escherichia coli (ETEC), in children in developing countries and in travelers to these areas [1,2]. ETEC colonize the intestinal mucosa by different specific colonization factors (CFs) on the bacterial surface. Following colonization, the bacteria produce a heat-labile toxin (LT) and/or a heat stable toxin (ST) that cause watery diarrhea. An effective vaccine, particularly for use in ETEC endemic countries, should not only induce acute, effective mucosal IgA antibody responses against the most prevalent types of ETEC, but also give rise to long-lasting immunological memory that can be boosted by single vaccine doses given several years after initial vaccination. However, in most clinical vaccine studies, immune

http://dx.doi.org/10.1016/j.vaccine.2016.04.055 0264-410X/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/ by-nc-nd/4.0/).

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

responses are only assessed shortly after immunization. Studies of long term immunological memory to ETEC and other enteric vaccines have also been limited by the lack of suitable methods for analysis of mucosal memory responses in humans. Recent studies suggest that mucosal infection and vaccination may induce IgA memory B cells that can be detected in peripheral blood [3–5], but the duration of these memory B cell responses appear to be short in the circulation, and the relation to functional mucosal memory has not been investigated. Hence, it is important to identify approaches to evaluate if oral enteric vaccines can induce long term mucosal memory in humans. We have recently tested the safety and immunogenicity of a novel oral inactivated multivalent ETEC vaccine (MEV; trade name ETVAX) in a large Phase I trial in adult Swedish volunteers [6]. MEV consists of four inactivated E. coli strains over-expressing the most common CFs, i.e. CFA/I, CS3, CS5 and CS6, and an LT binding subunit (LTB) related toxoid; LCTBA [7]. Two doses of the vaccine were shown to be safe, well tolerated and capable of inducing strong mucosal immune responses against the key vaccine antigens, i.e. the CFs and LTB, in a majority of vaccineess, as determined by analysis of intestine-derived antibody-secreting cell (ASC) responses as well as detection of antibodies in fecal extracts, when evaluated 1–6 weeks after initial immunization [6]. Addition of a promising mucosal adjuvant, double-mutant LT (dmLT) [8], in a dose of 10 ␮g enhanced immune response against the CFs present in the lowest amounts in the vaccine [6], supporting the dose-sparing effect of the adjuvant previously observed in animal studies [7]. Encouraged by these results we have now studied whether the previous vaccinations with MEV had induced an immunological memory that could be boosted by a single vaccine dose given one to two years after the initial vaccination and whether priming with dmLT influenced memory responses. We have previously shown that adult Swedish volunteers primed with two oral doses of an inactivated cholera vaccine (Dukoral® ) responded earlier than naïve subjects with peak intestine-derived circulating IgA ASC responses to a single oral booster dose given several years later [9]. Studies show that vaccine-specific ASC responses in peripheral blood reflect intestinal ASC responses [10] as well as antibody responses in fecal or lavage specimens [11] after oral ETEC or cholera vaccination, supporting the use of ASC assessment as a surrogate for mucosal IgA responses. We have now compared the kinetics and magnitudes of ASC as well as serum antibody responses to a single dose of MEV in subjects who 13–23 months earlier had received two oral doses of MEV alone or together with 10 ␮g of dmLT and in previously non-immunized controls. ASC responses were analyzed by the antibodies in lymphocyte supernatants (ALS) assay, which we and others have shown give comparable results as the more traditional ELISPOT assay for assessment of ASC responses to ETEC and cholera vaccines [6,9,12,13]. We also studied whether the previous priming immunizations with MEV had induced vaccine specific IgA and IgG memory B cells that were detectable in the circulation at the time for the booster vaccination. Furthermore, we analyzed whether presence of circulating vaccine specific memory B cells might predict functional immunological memory responses elicited by oral booster vaccination. 2. Materials and methods For additional information, see supplementary material.

3133

oral doses two weeks apart with vaccine buffer alone (previous Group A, now designated Group I; n = 13 of 34), MEV alone (previous Group B, now Group II; n = 17 of 35) or MEV plus 10 ␮g dmLT (previous Group C, now Group III; n = 18 of 30) were enrolled. In addition, 9 naïve controls, who had not previously participated in any ETEC vaccine trial, were recruited to the new Group I (total n = 22). All subjects received one oral dose of MEV alone (no adjuvant) in 150 ml bicarbonate buffer. The study was performed in accordance with the Declaration of Helsinki and written informed consent was obtained from all volunteers. 2.2. Immunological assays Blood samples for isolation of peripheral blood mononuclear cells (PBMCs) and serum were collected on the day of vaccination (day 0) and 4 or 5 and 7 days post vaccination. Mucosal immune responses were evaluated by measuring intestine-derived ASC responses in blood by the ALS assay and systemic immune responses by determining serum antibody levels by ELISA [6,13]. Vaccine specific memory B cells in blood were determined by stimulating PBMCs for 5–6 days with CpG DNA, pokeweed mitogen (PWM) and Staphylococcus aureus protein A (SAC), as described by Crotty et al. [14], or with CpG DNA, IL-15 and B cell activating factor (BAFF), as described by Tengvall et al. [15]. Numbers of specific ASCs were determined by ELISPOT using plates coated with CFA/I, CS3, CS5, CS6 or GM1 plus LTB [13,15,16]. 2.3. Safety assessment Safety was determined by evaluation of adverse events (AEs) via diary cards, physical examination, clinical chemistry and hematology tests. 2.4. Statistical analyses Statistical differences were evaluated using paired and unpaired t-tests and Fisher’s exact test. Correlations were evaluated by the Pearson test. P-values < 0.05 were considered significant. 3. Results 3.1. Study subjects Of 64 subjects screened, 57 were enrolled and immunized, resulting in 17–22 subjects in each of the three study groups (Fig. 1 and Table S1). In Groups II and III, the average time between the initial (primary) [6] and the third (late booster) vaccination was 18 months (range 13–23 months, Table S1). For the ALS and serum analyses, 52–54 subjects were included in the per protocol analysis sets (PPS) and for the circulating memory B cell analyses 43 subjects were included; reasons for exclusions are given in Figs. 1 and 6. 3.2. Safety The vaccine was safe and well tolerated with similar low frequencies of mainly mild AEs reported in all groups (Table S2). No serious AEs (SAEs) or severe AEs were reported. 3.3. Immune responses

2.1. Study design This was a three-armed, non-randomized, open-label, singlecenter Phase I trial. Random subsets of subjects who had previously, in a randomized, placebo controlled Phase I trial [6], received two

Subjects immunized with MEV alone (Group II) or MEV + 10 ␮g dmLT (Group III) 13–23 months earlier [6] were orally immunized with a single dose of MEV; their intestine-derived vaccine specific ALS responses in the circulation were compared with

3134

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

Group I

Group II

Group III

26 screened for eligibility

18 screened for eligibility

20 screened for eligibility

Enrollment

4 excluded; met excl.crit.

Enrollment

1 eligible but not randomized, selected as alternate

Enrollment

22 enrolled and immunized

17 enrolled and immunized

22 completed last visit

17 completed last visit

18 completed last visit

SAS: 22 PPS: 19 1, 2

SAS: 17 PPS: 16-17 2

SAS: 18 PPS: 17-18 2

2 excluded; met excl.crit.

18 enrolled and immunized

Two subjects (one male, one female) from Group I were excluded from all PPS used for immunological analyses due to travel to ETEC endemic countries ≤3 years prior to vaccination (travel history revealed after screening and booster vaccination). 2 A few samples could not be evaluated due to unspecific binding in ELISA assays, resulting in exclusion of one male subject from Group I from all ALS and serum PPS and one female subject from Group II and one female subject from Group III from the CF ALS PPS. 1

Fig. 1. Trial profile. All safety analyses were performed on the safety analysis set (SAS), including all subjects that received one dose of MEV vaccination (all 57 enrolled subjects). All immunogenicity analyses were performed on the per protocol sets (PPS), which were defined per specimen (ALS, serum or circulating memory B cells) and antigen type (LTB, CFs or LPS). Only samples which fulfilled the quality criteria specified in the standard operating procedure for each assay type were included in the PPS. The number of subjects included in the ALS and serum PPS are indicated in the figure and reasons for exclusion are listed in the footnote. The circulating memory B cell PPS is defined in Fig. 6.

corresponding responses induced by a single dose of MEV in previously non-vaccinated subjects (Group I). 3.3.1. Intestine-derived anti-CF ALS responses A majority of subjects (53–88%) in Groups II and III responded to the single booster vaccine dose with significant ALS IgA responses to all four vaccine CFs (CFA/I, CS3, CS5 and CS6; Fig. 2), with peak responses detected already on day 4/5 and lower responses on day 7 (significant declines for all antigens except CS5). In contrast, only 11% of the non-vaccinated subjects in Group I developed ALS responses to CFA/I, CS5 or CS6, respectively, and 26% to CS3 after single dose vaccination (Fig. 2). The responses detected on day 4/5 in Groups II and III were significantly higher and/or more frequent than all anti-CF responses in Group I (Fig. 2). No differences in responses to any of the vaccine CFs were observed in subjects primed with MEV alone and MEV + dmLT (Group II versus III; P > 0.05). 3.3.2. Intestine-derived anti-LTB and anti-O78 LPS ALS responses All subjects in Groups II and III responded with significant LTB-specific IgA and IgG antibodies in ALS specimens after the booster vaccination; peak responses were detected on day 4/5 post vaccination in a majority of subjects and the responses declined significantly until day 7 (Fig. 3A and B). In contrast, only 11–16% of the subjects in Group I responded to LTB; for those responding, highest responses were detected on day 7, the latest time point studied. Both the frequencies and magnitudes of anti-LTB responses were significantly higher in the primed than in the non-primed subjects (Fig. 3A and B). Comparable anti-LTB ALS responses were observed in subjects primed with MEV alone and MEV + 10 ␮g dmLT (Group II versus III; P > 0.05). All subjects in Group II and III also responded to booster vaccination with anti-O78 lipopolysaccharide (LPS) IgA ALS responses that peaked on day 4/5 and then significantly declined (Fig. 3C). Subjects in Group I also developed significant anti-LPS IgA responses, but these responses were comparable on day 4/5 and day 7.

Comparable ALS responses to CFs, LTB and O78 LPS were induced when the booster vaccine dose was administered 13–18 months (n = 19–20) compared to 19–23 months (n = 14–15) after primary vaccination (not shown). 3.3.3. Comparison of ALS responses on day 4 and 5 To identify the optimal time point for assessing intestinederived immune responses after booster vaccination, ALS responses to the different CFs, LTB and O78 LPS were compared on day 4 and 5 in the previously vaccinated subjects. ALS responses to CFA/I, CS5, CS6 and LTB were significantly higher and/or more frequent on day 5 than day 4 (Fig. 4A and B), whereas responses against CS3 and O78 LPS did not differ significantly between the days (Fig. 4A and C). 3.3.4. Serum antibody responses The systemic memory responses were analyzed by measuring serum antibody responses to a single dose of MEV. Whereas antiLTB serum IgA titers were comparable in all three groups before vaccination, anti-LTB IgG titers were significantly higher in Group II and III than Group I (P < 0.001). The majority (88–94%) of subjects in Groups II and III developed significant anti-LTB serum IgA and IgG responses to the booster vaccination, whereas only 11% of the Group I subjects responded with significant anti-LTB titers (Fig. 5A and B). Strongest responses to LTB were detected on day 7 in all groups and these responses were significantly higher than on day 4/5. Significantly higher magnitudes as well as frequencies of responses were detected both on day 4/5 and day 7 in Group II and III compared to Group I subjects (Fig. 5A and B). No differences in magnitudes or frequencies of serum anti-LTB responses were recorded between Group II and III. Strongest anti-O78 LPS IgA serum responses were detected on day 7 and these responses were comparable in the three study group (Fig. 5C). None of the subjects in Group I developed serum IgA antibody responses to any of the CFs. The frequencies of IgA responders to

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

3135

Fig. 2. ALS responses to vaccine CFs after administration of a single oral dose of MEV. IgA ALS responses against CFA/I (A), CS3 (B), CS5 (C) and CS6 (D) were determined in previously non-vaccinated subjects (Group I) and in subjects previously immunized with two oral doses of MEV alone (Group II) or two doses of MEV plus 10 ␮g dmLT (Group III). Geometric mean (GM + SEM) levels of antibodies in samples collected before (day 0) and 4/5 and 7 days after administration of a single dose of MEV are shown as bars. Results of statistical comparisons between groups and time points are indicated in the graphs. Responder frequencies (RF; %) on day 4 or 5 post vaccination are shown below the graphs; results of statistical comparisons of responder frequencies in Groups II and III versus Group I, respectively, on day 4/5 are indicated. ns; not significant.

CFs were also low in Groups II and III; 24% responded to CS3 and about 10% to the other CFs (not shown). 3.3.5. Circulating vaccine specific memory B cells We also evaluated whether the primary vaccinations had induced circulating memory B cells that remained at the time for the late booster vaccination. Circulating CF- and LTB-specific memory B cells were analyzed in blood samples collected before giving the booster dose to Groups II and III, and for comparison also before vaccination of Group I. Memory B cells were transformed into cells actively secreting antibodies by polyclonal stimulation with CpG DNA, PWM and SAC [14]. The frequencies of CF-specific IgA and IgG memory B cells were low and comparable in all three groups (not shown). However, we detected significantly higher LTB-specific IgG secreting cells in Groups II and III compared to in the non-vaccinated controls in Group I (Fig. 6A). The frequencies

of pre-existing LTB-specific IgG memory B cells before vaccination correlated significantly with the magnitudes of ALS IgG responses induced by the booster vaccination (Fig. 6B). In contrast, no significant differences in LTB-specific IgA secreting cells were detected between the previously vaccinated and non-vaccinated groups (Fig. 6C). Similar results were obtained using an alternative stimulatory cocktail [15] consisting of CpG DNA, IL-15 and BAFF (not shown). 3.3.6. Comparison of responses after primary and booster vaccinations We compared IgA ALS responses induced by the initial two oral doses of MEV determined in the previous primary vaccination study [6] with the responses induced by the single booster dose in the same subset of subjects in the present study. The magnitudes of ALS responses 5 days after booster vaccination were at least

3136

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

Fig. 3. ALS responses to LTB and O78 LPS after administration of a single oral dose of MEV. IgA (A) and IgG (B) ALS responses against LTB and IgA responses against LPS (C) were determined in previously non-vaccinated subjects (Group I) and in subjects previously immunized with two oral doses of MEV alone (Group II), or two doses of MEV plus 10 ␮g dmLT (Group III). Geometric mean (GM + SEM) levels of antibodies in samples collected before (day 0) and 4/5 and 7 days after administration of a single dose of MEV are shown as bars. Results of statistical comparisons between groups and time points are indicated in the graphs. Responder frequencies (RF; %) on day 4/5 are shown below the graphs; results of statistical comparisons of responder frequencies in Groups II and III versus Group I, respectively, on day 4/5 are indicated. ns; not significant.

comparable, and for some antigens (CS3, LTB and O78 LPS) significantly higher compared to the responses detected 5 days after the second vaccine dose in the previous study (Table S3). For additional information about this comparison, see Supplementary material. 4. Discussion We show that two oral doses of an oral inactivated whole cell ETEC vaccine, MEV [6,7], is capable of inducing significant functional mucosal immunological IgA memory IgA to the primary vaccine antigens, i.e. LTB and four different CFs. Thus, memory B cells induced by the initial two dose vaccination could quickly be reactivated by a single booster dose given 13–23 months later. This was shown by the appearance of considerably higher, more frequent and earlier appearing intestine-derived IgA ALS responses in the circulation after administration of a single dose of MEV in the previously vaccinated than in previously non-vaccinated subjects. Indeed, maximal ALS responses were detected already on day 5 in the primed individuals, whereas the highest responses

in previously non-vaccinated volunteers were detected on day 7 among the time points evaluated. Our results are consistent with previous findings that specific IgA memory responses against cholera toxin binding (CTB) subunit could be induced locally in the intestine by the oral CTB-containing cholera vaccine, Dukoral. Thus, presence of vaccine specific memory IgA cells in the intestinal mucosa, that could be reactivated by administration of a single oral vaccine dose 5 months after primary vaccination, has been reported [10]. Furthermore, we have detected rapid anamnestic responses in intestinal lavages already 3 days after a late booster vaccination given 15 months after primary vaccinations [17] and in ALS samples on day 4 or 5 after a late booster given up to 14 years after priming [9]. Since CTB is highly homologous to the LCTBA molecule in MEV [18], it is likely that mucosal memory responses to LTB, and potentially also to CFs and LPS, may be even more long-lived than shown in the present study. ASC responses in blood induced by the oral typhoid vaccine Ty21a also peak slightly earlier after booster compared to primary vaccine doses, administered 1–2 years previously, indicating persistent

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

Fig. 4. Comparison of ALS responses on day 4 and 5 after booster vaccination. IgA ALS responses to CFA/I, CS3, CS5 and CS6 (A), IgA and IgG ALS responses to LTB (B) and IgA ALS responses to O78 LPS (C) on day 4 (striped bars) and day 5 (black bars), respectively, after administration of a single oral booster dose of MEV in subjects previously immunized with two oral doses of MEV alone or MEV plus 10 ␮g dmLT (Groups II and III combined). Magnitudes of responses (GM + SEM) are shown as bars and responder frequencies (RF; %) are indicated below the graphs. The results of statistical comparisons of magnitudes of responses and responder frequencies on day 4 versus day 5 are shown. ns; not significant.

mucosal memory [19]. These results support that analysis of blood ASC/ALS responses may be a simple approach to assess intestinederived memory responses after oral immunization in humans and that the method can be applied to different types of vaccines and antigens. Almost 100% of specific ASCs which transiently migrate through peripheral blood after oral cholera vaccination express the mucosal homing marker ␣4␤7 [20], supporting the mucosal origin and destination of these cells. High expression of ␣4␤7 on antibody secreting plasmablasts has also been observed after oral typhoid and Shigella vaccination [21,22]. Vaccine specific ASC responses in the circulation have been shown to reflect intestinal ASC responses [10] as well as antibody responses in fecal or lavage specimens [11] after oral cholera and ETEC vaccinations. These findings strongly support that determination of vaccine-specific blood ASC responses after mucosal booster vaccination can be used to assess intestinal immune responses. However, the migration of specific ASCs from the mucosa into peripheral blood after vaccination seems to be more transient than previously realized. Thus, ASC responses have

3137

been recommended to be optimally analyzed 6–8 days after oral vaccination [23] and have usually been assessed 7–10 days after each vaccine dose [24–30]. In the present study, however, we found that the optimal time point to analyze ALS responses in primed individuals was 5 days after the booster and that responses declined markedly from day 5 to 7. This response kinetics differs from that in the non-primed individuals in whom highest ALS responses were detected on day 7. In a previous study of ALS responses to the oral cholera vaccine Dukoral, responses remained high or even increased slightly in a few subjects on day 9 compared to day 7 after the first dose [9], supporting that ALS responses to a first oral vaccine dose are more prolonged than responses to later doses, which we have observed to be very transient. These findings suggest that ASC responses against CFs and toxin antigens to repeated oral vaccine doses may have been grossly underestimated in several previous studies, which often have shown similar, or even lower, responses to a second than a first dose of cholera or ETEC vaccine when measured 7 days after vaccination [13,16,24–28]. Previous studies of memory B cell responses in humans have primarily been based on polyclonal stimulation and transformation of memory B cells into ASCs which can be detected using ELISPOT. This method was initially developed for analysis of memory B cell responses induced by parenteral vaccination [14], but has more recently also been used to analyze circulating memory B cells specific for ETEC, Shigella and cholera antigens after oral vaccination or natural infection [3–5,31–33]. In most studies memory B cells have only been analyzed 1–3 months after mucosal vaccination or infection. E.g. in a recent study, ETEC infection was shown to induce IgA and IgG memory B cells specific for LTB, CFA/I or CS6 which could be detected in the circulation one month after infection [5]. In a few studies of longer duration, CTB-specific IgG memory B cells have been identified in the circulation 3–6 months after oral cholera vaccination and 6–12 month after cholera infection, whereas IgA memory B cells only remained for one month after vaccination and 3–9 months after infection [3,4,32]. In the present study, we analyzed the occurrence of circulating vaccine specific memory B cells 13–23 months after primary ETEC vaccination. Consistent with previous cholera vaccine studies, significant frequencies of vaccine specific circulating IgA memory B cells were not detected at this late time point. Thus, although IgA memory B cells may remain for prolonged periods in intestinal mucosa and can be rapidly reactivated upon challenge, such cells only circulate for a few months after vaccination. In contrast, we found significant frequencies of circulating LTB-specific IgG memory B cells up to 23 months post vaccination. These frequencies correlated with the LTB-specific IgG ALS responses induced by the single booster dose of MEV. Collectively, our results suggest that long-term mucosal memory IgA responses, at variance with memory IgG responses, cannot be assessed by analysis of circulating memory B cells. One possibility to improve the immunogenicity of an oral whole cell ETEC vaccine, particularly for use in children, who may require reduced doses of vaccine to avoid gastro-intestinal side-effects [1,34], is to administer the vaccine together with a mucosal adjuvant. In the primary vaccination study, MEV was given alone or together with dmLT adjuvant [6]. Since the ALS responses to the CFs present in the lowest amounts in MEV were increased by coadministration of the vaccine with dmLT in that study, we evaluated whether the adjuvant might have affected the development of immunological memory. However, the strong memory responses induced against all primary vaccine antigens in the previously vaccinated subjects, irrespective of whether they had been primed with or without dmLT, does not support a beneficial effect of the adjuvant on memory development. Further studies will be undertaken to evaluate the effect of dmLT on both primary and memory B cell responses in young children and infants, who may require lower doses of vaccine.

3138

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

Fig. 5. Serum antibody responses to LTB and O78 LPS after administration of a single oral dose of MEV. IgA (A) and IgG (B) serum antibody responses against LTB and IgA serum antibody responses against LPS (C) in previously non-vaccinated subjects (Group I) and in subjects previously immunized with two oral doses of MEV alone (Group II) or two doses of MEV plus 10 ␮g dmLT (Group III). Geometric mean (GM + SEM) levels of antibodies in samples collected before (day 0) and 4/5 and 7 days after administration of a single dose of MEV are shown as bars. Results of statistical comparisons between groups and time points are indicated in the graphs. Responder frequencies (RF; %) on day 7 are shown below the graphs; results of statistical comparisons of responder frequencies in Groups II and III versus Group I on day 7 are indicated. ns; not significant.

Our study also suggests that MEV induces memory responses to O78 LPS. In contrast to the responses to the protein antigens, strong anti-LPS ALS responses were detected already on day 4/5 also in the previously non-vaccinated individuals. However, one of the main distinguishing features of a recall response, i.e. a decline in ALS antibody levels from day 4/5 to day 7, was also observed for anti-O78 LPS responses in the previously vaccinated subjects. Given that O78 LPS is one of the most common types of LPS expressed by ETEC, and is also prevalent among other E. coli colonizing humans [35], the rapid LPS response in previously non-vaccinated subjects may at least partly be explained by natural exposure to O78 E. coli. We also demonstrated systemic memory responses to LTB, since significantly higher and more frequent anti-LTB IgA and IgG serum responses were detected in response to a single dose of MEV in previously vaccinated compared to non-vaccinated subjects. However, at variance with the ALS responses, higher and more frequent serum antibody responses were detected on day 7 compared to day 4/5. In contrast, serum responses to O78 LPS were comparable in previously non-vaccinated and vaccinated subjects. Serum responses to CFs were infrequent after both primary and booster vaccination. Thus, memory responses to CFs or LPS

could be detected using analysis of kinetics and magnitudes of ALS responses, but not by analysis of serum antibodies. Comparison of ALS responses to the initial two doses of MEV and to a late booster revealed that the booster induced comparable, or even higher, responses to the primary vaccine antigens. However, responses to the late booster dose are unlikely to reflect responses which may be elicited by administration of a third dose of MEV shortly after the two first doses. Comparisons of intestinal antibody responses to CFs and CTB in intestinal lavage specimens after oral administration of two and three doses of a previous oral inactivated whole cell ETEC vaccine candidate demonstrated that maximal responses were achieved already after the second dose and were even lower after the third dose administered two weeks later [36]. The strong responses observed to the late booster in the present study may suggest efficient reactivation of long-lived high affinity memory B cells. Thus, a single booster dose of vaccine given ≥2 years after the initial priming may be sufficient to induce comparable, or even higher, levels of immune responses than induced by the initial vaccinations. In summary, our results show that the oral ETEC vaccine studied induces memory IgA responses to all primary vaccine antigens in a

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

A 0.10

3139

B

LTB IgG P=0.004 P<0.001

C LTB IgA

0.10

ns % LTB specific IgA ASCs per tIgA ASCs (GM + SEM)

1000 r=0.65

P<0.001

ALS IgG (fold rise)

% LTB specific IgG ASCs per tIgG ASCs (GM + SEM)

LTB IgG

0.03

0.01

Group I Group II Group III

100

10

1

0.1 0.001

ns

0.03

0.01

0.01

0.1

1

Group I Group II Group III

Memory IgG (% of tIgG) Fig. 6. LTB-specific memory B cells in the circulation 13–23 months after primary MEV vaccination. LTB-specific IgG (A and B) and IgA (C) memory B cells in non-vaccinated subjects (Group I) and in subjects immunized with two oral doses of MEV alone (Group II) or MEV plus 10 ␮g dmLT (Group III) 13–23 months earlier, as detected by polyclonal stimulation with CpG DNA, PWM and SAC followed by ELISPOT analysis. Frequencies (%) of LTB specific IgG or IgA ASCs per total numbers of IgG (tIgG)or IgA (tIgA) secreting ASCs prior to vaccination (on day 0) in the present trial are indicated. (B) Correlations between frequencies of LTB-specific IgG circulating memory cells before vaccination and magnitudes of LTB-specific ALS IgG responses induced by administration of a single oral dose of MEV to Group I (open circles) and Groups II and III (closed circles) measured on day 4/5 post vaccination. The circulating memory B cell PPS included 43 subjects (Group I; n = 14, Group II; n = 14 and Group III; n = 15). Reasons for exclusion from this PPS were travel to ETEC endemic countries (see Fig. 1) or low numbers of cells.

majority of subjects and that these responses persist for at least 1–2 years after vaccination. The results also suggest that a single vaccine dose may be sufficient to boost mucosal immune responses several years after priming vaccination. Our results support that analysis of the kinetics of ALS responses following immunization may be a useful approach to assess long-lived functional mucosal IgA memory to different types of vaccines in humans. The long lasting mucosal memory induced by MEV is encouraging for the ongoing efforts to develop this vaccine for use in young children and infants in ETEC endemic countries and in travelers.

PCT/EP2011/065784-PCT pending. The other authors declare that they have no conflicts of interest. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.vaccine.2016.04. 055. References

Acknowledgements We thank Scandinavian Biopharma, Sweden for supplying the vaccine and some of the antigens used in the study. We also thank Joanna Kaim, Madeleine Löfstrand, Gudrun Wiklund, Anneli Ekman and Patricia Loayza Fryklund for excellent technical assistance and Therese Schagerlind, Dan Curiac and the staff at the Clinical Trial Center and Gothia Forum at the Sahlgrenska University Hospital for valuable clinical support, Niklas Svensson for data management and all volunteers who participated in the trial. This work was supported by the European Union Seventh Framework Programme (FP7/2007-2013, grant agreement no. 261472-STOPENTERICS), the Sahlgrenska University Hospital (LUA-ALF, grant number 144411), the Swedish Research Council (grant numbers VR 521-2012-3464 and VR 348-2013-6615) and the Swedish Foundation for Strategic Research (grant number SA120072). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Author’s contribution: AL and AMS were the principal investigators. All authors participated in the design of the studies and interpretation of results. AL and AMS were responsible for the immunological analyses and wrote the manuscript. All authors contributed to the critical review and revision of the manuscript and have approved the final version. Conflict of interest statement: A.-M. Svennerholm is shareholder of the biotech company Gotovax AB that may receive a small royalty on sales of the ETEC vaccine if it becomes a commercial product. A.-M. Svennerholm has patent PCT/EP2012/067598-PCT and

[1] Svennerholm AM, Lundgren A. Recent progress toward an enterotoxigenic Escherichia coli vaccine. Expert Rev Vaccines 2012;11(4):495–507. [2] Walker RI. An assessment of enterotoxigenic Escherichia coli and Shigella vaccine candidates for infants and children. Vaccine 2015;33(8):954–65. [3] Harris AM, Bhuiyan MS, Chowdhury F, Khan AI, Hossain A, Kendall EA, et al. Antigen-specific memory B-cell responses to Vibrio cholerae O1 infection in Bangladesh. Infect Immun 2009;77(9):3850–6. [4] Alam MM, Riyadh MA, Fatema K, Rahman MA, Akhtar N, Ahmed T, et al. Antigenspecific memory B-cell responses in Bangladeshi adults after one- or two-dose oral killed cholera vaccination and comparison with responses in patients with naturally acquired cholera. Clin Vaccine Immunol 2011;18(5):844–50. [5] Alam MM, Aktar A, Afrin S, Rahman MA, Aktar S, Uddin T, et al. Antigenspecific memory B-cell responses to enterotoxigenic Escherichia coli infection in Bangladeshi adults. PLoS Negl Trop Dis 2014;8(4):e2822. [6] Lundgren A, Bourgeois L, Carlin N, Clements J, Gustafsson B, Hartford M, et al. Safety and immunogenicity of an improved oral inactivated multivalent enterotoxigenic Escherichia coli (ETEC) vaccine administered alone and together with dmLT adjuvant in a double-blind, randomized, placebo-controlled Phase I study. Vaccine 2014;32(52):7077–84. [7] Holmgren J, Bourgeois L, Carlin N, Clements J, Gustafsson B, Lundgren A, et al. Development and preclinical evaluation of safety and immunogenicity of an oral ETEC vaccine containing inactivated E. coli bacteria overexpressing colonization factors CFA/I, CS3, CS5 and CS6 combined with a hybrid LT/CT B subunit antigen, administered alone and together with dmLT adjuvant. Vaccine 2013;31(20):2457–64. [8] Norton EB, Lawson LB, Freytag LC, Clements JD. Characterization of a mutant Escherichia coli heat-labile toxin, LT(R192G/L211A), as a safe and effective oral adjuvant. Clin Vaccine Immunol 2011;18(4):546–51. [9] Leach S, Lundgren A, Svennerholm AM. Different kinetics of circulating antibody-secreting cell responses after primary and booster oral immunizations: a tool for assessing immunological memory. Vaccine 2013;31(30):3035–8. [10] Quiding M, Nordstrom I, Kilander A, Andersson G, Hanson LA, Holmgren J, et al. Intestinal immune responses in humans. Oral cholera vaccination induces strong intestinal antibody responses and interferon-gamma production and evokes local immunological memory. J Clin Invest 1991;88(1):143–8.

3140

A. Lundgren et al. / Vaccine 34 (2016) 3132–3140

[11] Ahren C, Jertborn M, Svennerholm AM. Intestinal immune responses to an inactivated oral enterotoxigenic Escherichia coli vaccine and associated immunoglobulin A responses in blood. Infect Immun 1998;66(7):3311–6. [12] Carpenter CM, Hall ER, Randall R, McKenzie R, Cassels F, Diaz N, et al. Comparison of the antibody in lymphocyte supernatant (ALS) and ELISPOT assays for detection of mucosal immune responses to antigens of enterotoxigenic Escherichia coli in challenged and vaccinated volunteers. Vaccine 2006;24(18):3709–18. [13] Lundgren A, Leach S, Tobias J, Carlin N, Gustafsson B, Jertborn M, et al. Clinical trial to evaluate safety and immunogenicity of an oral inactivated enterotoxigenic Escherichia coli prototype vaccine containing CFA/I overexpressing bacteria and recombinantly produced LTB/CTB hybrid protein. Vaccine 2013;31(8):1163–70. [14] Crotty S, Aubert RD, Glidewell J, Ahmed R. Tracking human antigen-specific memory B cells: a sensitive and generalized ELISPOT system. J Immunol Methods 2004;286(1–2):111–22. [15] Tengvall S, Lundgren A, Quiding-Jarbrink M, Svennerholm AM. BAFF, stimulatory DNA and IL-15 stimulates IgA+ memory B cells and provides a novel approach for analysis of memory responses to mucosal vaccines. Vaccine 2010;28(33):5445–50. [16] Jertborn M, Ahren C, Svennerholm AM. Dose-dependent circulating immunoglobulin A antibody-secreting cell and serum antibody responses in Swedish volunteers to an oral inactivated enterotoxigenic Escherichia coli vaccine. Clin Diagn Lab Immunol 2001;8(2):424–8. [17] Svennerholm AM, Gothefors L, Sack DA, Bardhan PK, Holmgren J. Local and systemic antibody responses and immunological memory in humans after immunization with cholera B subunit by different routes. Bull World Health Organ 1984;62(6):909–18. [18] Lebens M, Shahabi V, Backstrom M, Houze T, Lindblad N, Holmgren J. Synthesis of hybrid molecules between heat-labile enterotoxin and cholera toxin B subunits: potential for use in a broad-spectrum vaccine. Infect Immun 1996;64(6):2144–50. [19] Kantele A, Makela PH. Different profiles of the human immune response to primary and secondary immunization with an oral Salmonella typhi Ty21a vaccine. Vaccine 1991;9(6):423–7. [20] Quiding-Jarbrink M, Nordstrom I, Granstrom G, Kilander A, Jertborn M, Butcher EC, et al. Differential expression of tissue-specific adhesion molecules on human circulating antibody-forming cells after systemic, enteric, and nasal immunizations. A molecular basis for the compartmentalization of effector B cell responses. J Clin Invest 1997;99(6):1281–6. [21] Toapanta FR, Simon JK, Barry EM, Pasetti MF, Levine MM, Kotloff KL, et al. Guthoming conventional plasmablasts and CD27 – plasmablasts elicited after a short time of exposure to an oral live-attenuated Shigella vaccine candidate in humans. Front Immunol 2014;5:374. [22] Kantele A, Arvilommi H, Iikkanen K, Savilahti E, Makela HP, Herzog C, et al. Unique characteristics of the intestinal immune system as an inductive site after antigen reencounter. J Infect Dis 2005;191(2):312–7. [23] Saletti G, Cuburu N, Yang JS, Dey A, Czerkinsky C. Enzyme-linked immunospot assays for direct ex vivo measurement of vaccine-induced human humoral immune responses in blood. Nat Protoc 2013;8(6):1073–87.

[24] Wenneras C, Svennerholm AM, Ahren C, Czerkinsky C. Antibody-secreting cells in human peripheral blood after oral immunization with an inactivated enterotoxigenic Escherichia coli vaccine. Infect Immun 1992;60(7):2605–11. [25] Savarino SJ, Brown FM, Hall E, Bassily S, Youssef F, Wierzba T, et al. Safety and immunogenicity of an oral, killed enterotoxigenic Escherichia coli-cholera toxin B subunit vaccine in Egyptian adults. J Infect Dis 1998;177(3):796–9. [26] Savarino SJ, Hall ER, Bassily S, Brown FM, Youssef F, Wierzba TF, et al. Oral, inactivated, whole cell enterotoxigenic Escherichia coli plus cholera toxin B subunit vaccine: results of the initial evaluation in children. PRIDE Study Group. J Infect Dis 1999;179(1):107–14. [27] Qadri F, Wenneras C, Ahmed F, Asaduzzaman M, Saha D, Albert MJ, et al. Safety and immunogenicity of an oral, inactivated enterotoxigenic Escherichia coli plus cholera toxin B subunit vaccine in Bangladeshi adults and children. Vaccine 2000;18(24):2704–12. [28] Qadri F, Ahmed T, Ahmed F, Bradley Sack R, Sack DA, Svennerholm AM. Safety and immunogenicity of an oral, inactivated enterotoxigenic Escherichia coli plus cholera toxin B subunit vaccine in Bangladeshi children 18–36 months of age. Vaccine 2003;21(19–20):2394–403. [29] McKenzie R, Venkatesan MM, Wolf MK, Islam D, Grahek S, Jones AM, et al. Safety and immunogenicity of WRSd1, a live attenuated Shigella dysenteriae type 1 vaccine candidate. Vaccine 2008;26(26):3291–6. [30] Harro C, Sack D, Bourgeois AL, Walker R, DeNearing B, Feller A, et al. A combination vaccine consisting of three live attenuated enterotoxigenic Escherichia coli strains expressing a range of colonization factors and heat-labile toxin subunit B is well tolerated and immunogenic in a placebo-controlled double-blind phase I trial in healthy adults. Clin Vaccine Immunol 2011;18(12):2118–27. [31] Simon JK, Wahid R, Maciel Jr M, Picking WL, Kotloff KL, Levine MM, et al. Antigen-specific B memory cell responses to lipopolysaccharide (LPS) and invasion plasmid antigen (Ipa) B elicited in volunteers vaccinated with live-attenuated Shigella flexneri 2a vaccine candidates. Vaccine 2009;27(4): 565–72. [32] Alam MM, Arifuzzaman M, Ahmad SM, Hosen MI, Rahman MA, Rashu R, et al. Study of avidity of antigen-specific antibody as a means of understanding development of long-term immunological memory after Vibrio cholerae O1 infection. Clin Vaccine Immunol 2013;20(1):17–23. [33] Leung DT, Rahman MA, Mohasin M, Patel SM, Aktar A, Khanam F, et al. Memory B cell and other immune responses in children receiving two doses of an oral killed cholera vaccine compared to responses following natural cholera infection in Bangladesh. Clin Vaccine Immunol 2012;19(5):690–8. [34] Qadri F, Ahmed T, Ahmed F, Begum YA, Sack DA, Svennerholm AM. Reduced doses of oral killed enterotoxigenic Escherichia coli plus cholera toxin B subunit vaccine is safe and immunogenic in Bangladeshi infants 6–17 months of age: dosing studies in different age groups. Vaccine 2006;24(10):1726–33. [35] Wolf MK. Occurrence, distribution, and associations of O and H serogroups, colonization factor antigens, and toxins of enterotoxigenic Escherichia coli. Clin Microbiol Rev 1997;10(4):569–84. [36] Ahren C, Wenneras C, Holmgren J, Svennerholm AM. Intestinal antibody response after oral immunization with a prototype cholera B subunitcolonization factor antigen enterotoxigenic Escherichia coli vaccine. Vaccine 1993;11(9):929–34.