OUP user menu

Kinetics of the mucosal antibody secreting cell response and evidence of specific lymphocyte migration to the lung after oral immunisation with attenuated S. enterica var. typhimurium

J.S. Allen, G. Dougan, R.A. Strugnell
DOI: http://dx.doi.org/10.1111/j.1574-695X.2000.tb01440.x 275-281 First published online: 1 April 2000


The kinetic of mucosal secretory responses elicited by the vaccine vector Salmonella enterica var. typhimurium (S. typhimurium) was examined by enzyme linked immunospot (ELISPOT) and compared with serum responses. Mice immunised orally with BRD509, the aroA, aroD mutant of virulent S. typhimurium SL1344 expressing the C Fragment of tetanus toxin (TT), simultaneously developed an IgA antibody secreting cells (ASC) response in the gastrointestinal lamina propria, the spleen and the lung, against both S. typhimurium lipopolysaccharide (LPS) and TT. The magnitude of the ASC response was greatest in the gut, was boosted by a secondary immunisation at day 25, and the kinetic of the response did not correlate with the appearance of serum antibodies. This study suggests that S. typhimurium can engage the common mucosal immune system to effect mucosal secretory responses at distal sites, however, the magnitude of the responses is both greatest in the gut and antigen-specific. The ASC origin of the serum antibodies specific for S. typhimurium and antigens expressed by the bacterium is yet to be elucidated.

  • Salmonella
  • Mucosa
  • Antibody
  • IgA

1 Introduction

The ability of attenuated Salmonella enterica var. typhimurium (S. typhimurium) to deliver heterologous immune responses against protective antigens from a wide range of bacteria, viruses and parasites to the immune systems of mice and other animals has been widely reported (reviews: [1,2]). Under some circumstances it may be important for S. typhimurium carrying heterologous antigens to elicit mucosal humoral responses. Such responses could limit the replication of pathogens within the mucosal lumen, the adhesin of mucosal pathogens to the mucosal epithelium, and/or neutralise toxins liberated within the gastrointestinal lumen.

The aim of this study was to examine the kinetic of the humoral mucosal immune responses to S. typhimurium lipopolysaccharide (LPS) and a model heterologous antigen, the fragment C of tetanus toxin (TT) [36]. The ability of attenuated salmonellae to induce mucosal responses against both antigens at local and distal sites was assessed by enumerating antigen-specific antibody secreting cells (ASC) in the gut and lung. The relationship between bacterial persistence and the development of immune responses, and the correlation between serum antibody responses and ASC was also investigated.

2 Materials and methods

BRD509, the aroA, aroD mutant of S. typhimurium SL1344 carrying plasmid pTETtac4 which encodes fragment C of TT elicits protective levels of antibody against TT challenge [3]. Female BALB/c mice were orally gavaged with approximately 1010 BRD509 (pTETtac4) in 200 µl, 30 min after gavage with 100 µl 10% (w/v) NaHCO3 in water. At 5-day intervals, three animals were killed and single cell preparations were made from the lamina propria [7], spleen and lungs. These cell suspensions were used in enzyme linked immunospot (ELISPOT) assays for ASC specific for TT and LPS as described [8,9]. The isotype (IgA, IgG and IgM) of specific ASC was determined using heavy chain-specific second antibodies. The number of bacteria present in the Peyer's patches (PP) of gavaged mice was determined by viable count and TT- and LPS-specific serum antibody levels measured by ELISA. Briefly, Immulon (Nunc) ELISA plates were coated with 10 µg ml−1 LPS (Sigma) or two Lfunits ml−1 Tetanus Toxoid (CSL Ltd, Parkville, Vic., Australia) in PBS or carbonate buffer respectively, washed once, blocked with 5% bovine serum albumin (Sigma) then washed prior to the addition of sera. The sera were diluted serially (three-fold) from 1/100 to 1/72900 in PBS containing 0.05% Tween 20. After incubation at 37°C for 2 h, the plates were washed and anti-mouse Ig HRP conjugate (Silenus Labs, Hawthorn, Vic., Australia), or anti-mouse IgA HRP conjugate (ICN Biologicals), diluted 1/1000 in PBS containing 0.05% Tween 20 was added to each well. After 1.5 h at 37°C, the plates were washed and antibody binding detected by the addition of Immunopure OPD substrate (Pierce), according to the manufacturer's instructions, and absorbance was read at 492 nm. The titre was defined at the dilution of antibody which gave an OD reading five-fold above the background OD obtained from control wells containing antigen, conjugated antibodies and substrate.

The in vivo stability of bacterial constructs was determined by plating of organ homogenates onto media with or without antibiotics.

3 Results

The ability of the BRD509 (pTETtac4) to persist in the PP was investigated by viable count over the course of 50 days. The in vivo stability of pTETtac4 in BRD509 was also assessed. The effect of a second inoculation on colonisation was also measured, with a group of mice receiving a boost immunisation on day 28. The viable counts were maximal in the PP at day 10, thereafter declining, with some mice having cleared all organisms by day 35 and other animals having low counts after this time (Fig. 1). The boost infection resulted in only a small, transient increase in the viable count suggesting that the immune mechanisms responsible for clearance of the original inoculum eliminated the secondary infection. The numbers of bacteria recovered from tissues cultured on media with and without ampicillin were not statistically different, supporting previous studies which showed that pTETtac4 was stable in vivo in the absence of antibiotic selection.

Figure 1

PP counts from mice infected with BRD509(pTETtac4). PP were isolated from each mouse, homogenised, serially diluted and plated onto media selecting for streptomycin resistant organisms ○ and media with ampicillin ●, selecting for organisms carrying pTETtac4. PP were also isolated from boosted mice, with the homogenate being plated onto streptomycin △ and ampicillin+streptomycin-containing media ▲. Arrow denotes time of inoculation of mice, with all mice being inoculated at day 0 and boosted mice receiving a second inoculum on day 25. No organisms were isolated from the PP of any mice after 45 days post-primary inoculation.

Serum antibody responses to S. typhimurium LPS and the plasmid encoded fragment C of TT were assessed over a period of 50 days in mice immunised orally with BRD509 (pTETtac4). Eight mice were orally immunised with BRD509 (pTETtac4), with four of the eight mice receiving a further inoculation on day 28. All mice were bled at 5-day intervals and the serum antibody responses assessed by ELISA against S. typhimurium LPS and tetanus toxoid (TT). LPS- and TT-specific Ig (Fig. 2) and IgA responses (Fig. 3) were detected in all mice. The serum IgA LPS-specific response was not routinely detected until 30 days after inoculation, after which time a high level was maintained until the experiment was terminated. Boosting resulted in a transient increase in responses which decreased to the level present in unboosted mice by day 60 post-infection. The kinetic of the total immunoglobulin response to LPS was different to that of IgA, in that titres were elevated by day 15. As with IgA LPS responses, maximal titres were obtained by day 30, and were maintained for the following 30 days. Boosting had little effect on the titres of LPS-specific immunoglobulins and total Ig titres were significantly elevated over the levels of IgA.

Figure 2

Total serum immunoglobulin LPS- and TT-specific responses in Salmonella infected mice. Mice were infected with BRD509/pTETtac4 and were bled every 15 days post infection. Serum was pooled from four mice, initially diluted 1/100 and then titrated and analysed by ELISA. □ denotes mice receiving a single inoculum, △ denotes mice which were inoculated on days 0 and 28. Arrows indicate when the inoculations occurred.

Figure 3

Serum IgA LPS- and TT-specific responses in Salmonella infected mice. Mice were infected with BRD509/pTETtac4 and were bled every 15 days post infection. Serum was pooled from four mice, diluted 1/100, titrated, and assayed by ELISA. □ denotes mice receiving a single inoculum, △ denotes mice which were inoculated on days 0 and 25. Arrows indicate when the inoculations occurred.

Serum IgA levels specific for TT were reduced compared with LPS (Fig. 3) and were initially detected 15 days post-immunisation, but failed to increase with either time or following a second immunisation. Total serum Ig responses specific for TT were higher than IgA levels, but as with IgA, failed to increase after day 15. There was a poor correlation between persistence of the bacterium and antibody responses and the two antigens elicited different response kinetics, which might their reflect their relative T cell dependence (C Fragment) or independence (LPS).

LPS- and TT-specific ASC in spleen, gut and lung cell populations were enumerated with the ELISPOT assay technique. In this technique, the number of lymphocytes which secrete specific antibodies are detected using antigen-coated plates. The secreted antibodies react with the relevant antigen in the immediate vicinity of the secreting cell, and can be detected with enzyme conjugated antisera and substrate which results in an insoluble spot, with each spot being representative of a single ASC. The isotype of the ASC was determined by use of a second antibody which specifically reacted with IgA, IgM and IgG, allowing the presence of each isotype to be ascertained for the gut-, spleen- and lung-derived ASC. Thirty-nine BALB/c mice were orally immunised with BRD509 (pTETtac4), with 12 mice receiving a second inoculation on day 25. Three mice were killed on days 5, 10, 15, 20 and 25, with three singly inoculated and three boosted mice being killed on days 30, 35, 40 and 50. Spleen, gut and lung single cell suspensions from each animal were plated in duplicate in 96-well tissue culture plates previously coated with LPS and TT. Following enzyme linked detection of IgA, IgG and IgM-reacting immunoglobulins, spots were counted and the mean number of counts from each mouse organ against LPS and TT was calculated and expressed as number of ASC per 106cells in the organ cell suspension.

The predominant ASC responses toward LPS were of IgA and IgM isotypes, with low numbers of IgG-producing ASC. IgA-producing LPS-specific ASC were detected in all organs of singly inoculated and boosted mice, with large numbers isolated from the gut (Fig. 4). Whilst large numbers of IgM-producing ASC specific for LPS were detected in the gut and spleen of mice, the boost did not elevate numbers above the peak in singly inoculated mice (Fig. 5). Only low numbers of IgG-producing ASC specific for LPS were very limited in the gut, spleen and lungs of mice, with the boost resulting in a small increase in numbers (Fig. 6).

Figure 4

IgA-producing ASC responses to LPS and TT in the gut, spleen and lungs of BRD509(pTETtac4) immunised mice. Three mice per group were killed at 5-day intervals, with the number of ASC enumerated by ELISPOT assay. Each symbol represents the mean number of spots from three mice, with the error calculated as S.E.M. □ denotes mice immunised once, △ denotes mice receiving a second inoculum on day 25. Arrows denote the time of inoculation/boosting.

Figure 5

IgM-producing ASC responses to LPS and TT in the gut, spleen and lungs of BRD509(pTETtac4) immunised mice. Three mice per group were killed at 5-day intervals, with the number of ASC enumerated by ELISPOT assay. Each symbol represents the mean number of spots from three mice, with the error calculated as S.E.M. □ denotes mice immunised once, △ denotes mice receiving a second inoculum on day 25. Arrows denote the time of inoculation/boosting

Figure 6

IgG-producing ASC responses to LPS and TT in the gut, spleen and lungs of BRD509(pTETtac4) immunised mice. Three mice per group were killed at 5-day intervals, with the number of ASC enumerated by ELISPOT assay. Each symbol represents the mean number of spots from three mice, with the error calculated as S.E.M.

As with LPS-specific ASC, the greatest number of TT-specific ASC were IgA producing and isolated from the gut and spleen (Fig. 4), although the numbers of ASC detected were much lower than for LPS. No IgA-producing TT-specific ASC were detected in the lungs of mice receiving one or two inocula of BRD509 (pTETtac4). Few IgM ASC specific for TT were detected in any of the animals, with the greatest number being detected in the lungs of boosted animals (Fig. 5). As with ASC responses to LPS, IgG-producing TT-specific ASC were only detected in the spleens and lungs of boosted mice (Fig. 6). IgG ASC were the only isotype where numbers of ASC were greater against TT than LPS.

Irrespective of isotype or antigen specificity, ASC were only detectable for a brief period of approximately 10 days, with a peak seen in unboosted mice on day 25 and in boosted mice 5 days following the boost on day 30. This ASC response did not correlate with serum antibody specific for LPS or TT, which was first observed on day 15, and which increased or remained constant for the next 45 days. These observations suggests that ASC measured by ELISPOT in the gut, spleen or lungs are not the major contributors to the early serum antibody responses, or to the maintenance of serum antibody responses over the course of 60 days.

4 Discussion

This study was performed to gain insight into the kinetics of both the serum titres and ASC numbers in different organs during the course of infection with Salmonella expressing the foreign antigen, fragment C. Whilst previous studies have investigated the humoral responses to both Salmonella and carried, foreign antigens, the kinetics of increases in ASC numbers and trafficking to different organs following infection has not been widely investigated [1,2,6,1012]. The results from this study should provide insight into the use of salmonellae to deliver mucosal antibody responses, in particular to the timing and persistence of mucosal humoral immunity generated by this bacterium.

Investigation of the kinetics of serum antibody responses to LPS and TT has suggested that these S. typhimurium LPS and fragment C are processed differently by the immune system, resulting in a different timepoint for onset on maximal titres. Whilst the serum antibody response argues for different handling of the two antigens, the kinetics of ASC numbers were similar between the two antigens, with a transient peak in ASC numbers around day 25 in singly inoculated, and day 30 in re-immunised mice. The magnitude of the responses to LPS were greater than to TT, with the greatest differences observed in IgA-producing ASC. The presence of serum antibody titres prior to, and after the transient peak in ASC numbers suggests that the spleen, gut or lungs are not the major source of serum antibody during the 60 day experimental period. The detection of antigen-specific ASC in the gut and lung illustrates the ability of Salmonella to induce immune responses at both local and distal mucosal sites.

Organ counts of the PP were performed to investigate if a correlation between bacterial load and immune response existed. The peak number of organisms in the PP was apparent 10 days after inoculation, with organisms detectable in some mice 40 days after inoculation. Analysis of spleen colonisation (data not shown) supported previous studies which revealed a correlation between PP colonisation and humoral immune responses, and not between splenic invasion and antibody levels [6,13].

Serum immune responses were detected against both LPS and TT, with maximal responses detected against LPS on day 30 and TT on day 15. While the timing of the maximal responses was equivalent, the relative titres of LPS-specific antibody were greater than those against TT. LPS is a potent inducer of humoral immune responses, and can directly activate B cells by a mechanism distinct from the pathway used by antigens binding to membrane Ig [1417]. Polyclonal B cell responses induced by B cell mitogens including LPS are characterised by the rapid production of predominantly IgM Ab which binds to the antigen with low affinity [18]. Whilst all B cells have the potential to respond to LPS, the CD5+ (B-1a) B cell subpopulation plays a major role in the in vivo IgM responses to polyclonal activators [16,19,20]. Our data suggested that a significant IgM response was produced against LPS, on day 15 after immunisation; this response may have derived from B-1a immunisation.

The numbers of LPS- and TT-specific ASC were investigated in three organs, the lamina propria, the spleen and the lung. The PP are inductive sites within the GALT, where the immune responses are initiated, whereas the lamina propria of the intestinal wall forms part of the effector component of the GALT [21]. As IgA-positive B cells have been shown to migrate from the PP to the lamina propria [21,22], it was of interest to sample lamina propria lymphocytes (LPL) which were most probably derived from the site of invasion and residence of S. typhimurium, i.e. the PP. Spleen and lung lymphocytes were also assayed for the presence of antigen-specific ASC, forming part of the systemic immune system, and as a distant mucosal site, respectively.

Important to the context of mucosal immunisation by recombinant salmonellae, this study has shown that BRD509(pTETtac4) is an effective inducer of IgA-producing ASC against both LPS and TT in the gut and spleen, and against LPS in the lung. Whilst IgA-producing ASC were the predominant isotype against TT, both IgA and IgM-producing LPS-specific ASC were detected. The differences between the ASC isotypes directed against LPS and TT may be due to differences in the antigens and processing. As stated, LPS-stimulated polyclonal B cell responses are characterised by a predominance of IgM, much of this antibody derived from CD5+ (B-1a) B cells [16,19,20].

The brief period of detectable ASC compared with the longer maintenance of LPS- and TT-specific antibody responses for the 60 days of the study suggests that the specific serum antibody is not derived from ASCs within the organs tested. As serum immunoglobulins have a short half life, in the order of days to, at most, a few weeks, maintenance of serum antibody levels requires a constant pool of plasma cells synthesising antibody [23,24]. Whilst ASC in the blood may be potential candidates for maintenance of serum antibody titres, studies in humans with S. typhi Ty21a have shown that there is only a transient population of ASCs in the blood, consistent with migration between induction and effector sites [11,12,25]. Bone marrow has been identified as a major site of long-term antibody production in murine infections with lymphocytic choriomeningitis virus (LCMV) [24]. This tissue may also be the source of ASC maintaining prolonged antibody titres in the blood of mice infected with Salmonella.

The low level of ASC detectable in the lungs of mice may be a reflection of the compartmentalisation of the common mucosal immune system, where the site of induction can affect the location of immune responses [25,26]. Whilst oral immunisation with Salmonella may not be the optimal route for induction of lung responses, S. typhimurium carrying foreign antigens have been shown to be capable of inducing antibodies detectable in lung washes [27,28]. Walker et al. [27] investigated responses after three immunisations and found detectable ASC numbers mainly after boosting, suggesting that multiple inoculations may be required for detectable responses, consistent with detection of lung ASC predominantly in boosted mice.

These studies support previous investigations into the mucosal immunisation which showed that S. typhimurium has the capacity to elicit mucosal humoral responses. In an era when vaccine development against complex pathogens such as HIV is seeking novel means of controlling viral entry, mucosal immunisation appears to offer considerable potential. As the prototypic S. typhi vaccine strains complete early clinical trials, and initial vaccine vector studies based on these bacteria are completed, it will become opportune to analyse the ability of these organisms to deliver mucosal immune responses against carried antigens.


R.A.S. is a member of the CRC for Vaccine Technology. This work was supported in part by the NH and MRC (Australia). J.S.A. was the recipient of an Australian Postgraduate Research Award.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
View Abstract