OUP user menu

A mouse model utilising human transferrin to study protection against Neisseria meningitidis serogroup B induced by outer membrane vesicle vaccination

Fredrik Oftung, Martinus Løvik, Svein Rune Andersen, L. Oddvar Frøholm, Gunnar Bjune
DOI: http://dx.doi.org/10.1111/j.1574-695X.1999.tb01374.x 75-82 First published online: 1 October 1999

Abstract

We have previously developed a mouse model based on transient bacteraemia in normal B10.M mice to evaluate the protective efficacy of outer membrane vesicle vaccines against serogroup B meningococci. To obtain a course of infection similar to that observed in man, we have in this work modified the mouse model by administration of human holo-transferrin upon bacterial challenge. Co-challenge with holo-transferrin induced increasing bacteraemia and subsequent death in normal non-immune mice, but not in vaccinated animals. The model system is dependent on challenge with meningococci expressing the transferrin receptor which is obtained by culturing the bacteria under iron restriction. The modified model system for protection against meningococcal infection presented here makes it possible to measure outer membrane vesicle vaccine induced protection by using bacteraemia as well as survival as parameters.

Keywords
  • Neisseria meningitidis
  • Outer membrane vesicle vaccine
  • Mouse protection model

1 Introduction

Infection with Neisseria meningitidis is a common cause of septic shock and meningitis world-wide [1]. While protective vaccines based on capsular polysaccharides against serogroup A and C meningococci have been developed [2,3], this immunisation principle can not be applied to the B serogroup of meningococci, since its polysaccharide is not immunogenic in humans [4]. Primarily to cope with serogroup B epidemics [5], the National Institute of Public Health in Norway has developed an outer membrane vesicle (OMV) vaccine [6], which has been shown to induce protection against meningococcal B disease among teenagers in Norway [7].

Although the OMV vaccine will be included in the Norwegian vaccination programme, efforts to improve its protective efficacy are made with regard to both antigenic composition [8] and route of administration [9,10]. In this context the use of in vitro correlates to protection (serum bactericidal activity and phagocytosis) established in humans [11,12] is important, but the possibility of conducting corresponding human protection trials is not feasible. As a supplement to the laboratory correlates, we have recently developed a mouse model for the evaluation of vaccine-induced protection based on transient bacteraemia in normal animals [13]. This model system was used to demonstrate that both a primary infection and OMV vaccination induced a significant degree of protection in B10.M mice [13]. However, the model differs significantly from systemic meningococcal infection in humans, since normal mice are able to control and eliminate a high level of bacteraemia without clinical disease and death. The protection observed is thus evident only by a more rapid reduction in bacteraemia compared to control experiments. The low growth potential of meningococci in mice is mainly due to the fact that meningococcal transferrin receptors are specific for human transferrin and not able to bind and utilise mouse transferrin as a source of iron, which is required for bacterial growth [1416]. In addition, mice are highly resistant to lipopolysaccharide (LPS)-mediated pathogenesis.

To obtain increasing bacteraemia and subsequent death in unprotected animals, we have in this work modified the established mouse model by the administration of human transferrin upon challenge with bacteria expressing the transferrin receptor induced by iron starvation. In addition, we have investigated the possibility of reducing the bacterial challenge dose by increasing the LPS sensitivity of the mice with galactosamine treatment. The mouse model described here allows evaluation of OMV vaccine-induced protection by using survival as well as bacteraemia as parameters.

2 Materials and methods

2.1 Animals

Immunologically mature (6–8 weeks) female mice of the congenic B10.M strain were obtained from Harlan Olac, Shaws Farm, Blackthorn, Bicester OX6 OTP, UK. The H-2 haplotype expressed by this strain is f [17].

2.2 Culture of meningococci

N. meningitidis serogroup B (subtype 15:P1.7.16, strain 44/76) was stored frozen (−70°C) in small aliquots in Greaves medium. The strain is identical to the vaccine strain used in protection experiments. Bacteria were grown on BHI and horse blood agar plates to assess purity before being expanded as a fluid culture in BHI medium overnight. The number of bacteria used in experimental infections (immunisations) was estimated by optical density (OD600) and controlled by plating of serial dilutions of the suspension used for inoculation of mice.

2.3 Determination of conditions inducing expression of transferrin receptor

The experimental conditions required to obtain expression of the meningococcal receptor for human transferrin during bacterial growth were determined by culturing meningococci under iron restriction. This was performed by titrating the BHI culture medium with respect to the iron chelator ethylene diamino di-ortho-hydroxyphenyl acetic acid (EDDA). Concentrations of EDDA used in these experiments ranged from 0 to 200 µM.

2.4 Experimental infection

Meningococci for experimental infections (primary infection) were prepared fresh for each experiment as described above. Mice were infected with a sub-lethal dose of 107 bacteria, prepared as described above, by intravenous (i.v.) injection in a volume of 0.2 ml. The course of the primary infection was controlled by monitoring bacteraemia 3 and 8 h post-inoculation.

2.5 Vaccination

Vaccination and re-vaccination (4 weeks later) of mice were performed by a subcutaneous (s.c.) injection of 8 µg (protein content) of meningococcal B strain 44/76 subtype 15:P1.7.16 OMV adsorbed to aluminium hydroxide on the chest of the animals [12].

2.6 Bacterial challenge

Challenge of mice treated with different doses of human transferrin (0–20 mg) was performed by intraperitoneal (i.p.) injection of 107 bacteria (0.2 ml) grown in the presence of 60 µM EDDA (iron restriction). In the protection experiments, all mice were challenged with 107 iron restricted bacteria together with 20 mg transferrin. Co-challenge of normal mice with 15 mg galactosamine (Sigma) i.p. was combined with 20 mg transferrin and 105–104 bacteria grown under iron restriction.

2.7 Design of protection experiments

Three experimental groups of 8–10 animals were included in protection experiments: (1) untreated normal control mice, (2) mice vaccinated and re-vaccinated with the same dose (8 µg MenB OMV) 4 weeks after the first vaccination and (3) mice infected i.v. with 107 viable bacteria at the same time as the first vaccination. After 7 weeks all groups were challenged i.p. by 107 meningococci (iron-restricted growth) co-administered with 20 mg human transferrin.

2.8 Bacteraemia

To determine bacteraemia, all groups of mice were bled (25 µl) from the lateral femoral vein into heparinised capillary tubes at 3, 8 and 24 h post-challenge. The level of bacteraemia of individual mice was determined by plating of serial 10-fold dilutions of blood on blood agar plates. The number of bacterial colonies was determined after 24 h incubation at 33°C in a CO2 (5%) enriched atmosphere.

2.9 Survival

The condition of the animals was assessed at the time of blood collection 3, 8 and 24 h post-challenge. In addition, the animals were inspected after 48 and 72 h. The survival status of the groups was determined 72 h post-challenge.

2.10 Statistics

Wilcoxon signed rank test (non-parametric) was used in the statistical analysis of the results (GraphPad Prism for Windows).

3 Results

3.1 Culture of meningococci under iron restriction

A prerequisite for a relevant effect of human transferrin in the model system is that the challenge bacteria can utilise this source of iron through specific binding and uptake by the transferrin receptor [1416]. Since the surface expression of the transferrin receptor is up-regulated in the absence of iron [15,16,18], we first determined the culture conditions required to obtain bacterial growth under iron restriction. Parallel bacterial cultures were inoculated in BHI medium in the presence of increasing concentrations of EDDA (Fe2+ chelator) and the growth was monitored continuously by measurement of optical density (OD600) for 11 h. The results, shown in Fig. 1, demonstrated that meningococcal growth was clearly reduced as a function of increasing EDDA concentrations. EDDA concentrations in the lower range (0–35 µM) did not significantly reduce the final bacterial density reached, whereas high concentrations of the chelator added completely blocked meningococcal growth. However, in the presence of intermediate levels of EDDA (60 µM), it was still possible to obtain bacterial growth, but at a significantly reduced rate compared to normal culture conditions. On the basis of these results, we chose 60 µM EDDA as standard concentration for obtaining transferrin receptor positive challenge bacteria.

Figure 1

Bacterial growth (44/76), measured as optical density (OD600), studied as a function of increasing concentration of the iron chelator EDDA (ranging from 0 to 200 µM EDDA).

3.2 Determination of transferrin dose to obtain increasing bacteraemia and death in non-immunised B10.M mice

To determine the dose of human transferrin necessary to obtain progressively increasing bacteraemia and death (monitored after 72 h), we challenged normal non-immune B10.M mice with 107 bacteria grown in the presence of 60 µM EDDA. In addition, human transferrin was co-administered in increasing doses from 0 to 20 mg peranimal. Bacteraemia was monitored at 3, 8 and 24 h after challenge and total mortality was determined after 72 h.

The results, shown in Fig. 2, demonstrated that it was possible to induce a shift in the course of bacteraemia from decreasing to increasing kinetics as a function of the transferrin dose given upon challenge. By co-administration of 20 mg transferrin with iron-restricted challenge bacteria, we were able to obtain a stable increasing bacteraemia during the infection which after 24 h reached a level approximately 500 times higher than observed in control mice not receiving transferrin. With respect to mortality, we also observed that the percentage of dead animals monitored after 72 h increased as a function of the transferrin dose given. Co-challenge of B10.M mice with 0, 1, 5 and 20 mg human holo-transferrin resulted in a mortality of 0, 25, 88 and 100%, respectively (Table 1).

Figure 2

Bacteraemia (CFU ml−1 blood) studied as a function of co-administration of different doses of human transferrin upon challenge of non-immune B10.M mice with 107 meningococci grown under iron restriction (60 µM EDDA).

View this table:
Table 1

Survival of non-protected B10.M mice challenged with 107 meningococci grown under iron restriction (60 µM EDDA) together with increasing doses of human transferrin (1, 5 and 20 mg/animal)

TreatmentSurvivals/tested% Survival
107 meningococci, without human transferrin8/8100
107 meningococci, 1 mg human transferrin2/825
107 meningococci, 5 mg human transferrin1/812.5
107 meningococci, 20 mg human transferrin0/80
  • Eight mice were used in each experimental group and survival was recorded 72 h post-challenge. Meningococcal challenge without transferrin was included as a control experiment.

Experiments in which the number of challenge bacteria was reduced (106, 105 and 104 bacteria and 20 mg transferrin) did not result in increasing bacteraemia and death of the animals (results not shown). On the basis of these results, a challenge dose of 107 bacteria (iron-restricted growth) co-administered with 20 mg human transferrin was used in the protection experiments described below.

3.3 Co-challenge with galactosamine

To investigate the possibility of reducing the bacterial challenge dose and still obtain a lethal course of bacteraemia, we also performed co-challenge experiments with galactosamine (15 mg per animal), which is reported to increase the sensitivity to the pathological effects of LPS in mice [19,20]. As expected, co-challenge with galactosamine increased the mortality of normal B10.M mice, however, these experiments also showed that the galactosamine-induced increase in mortality systematically was associated with a shift in bacteraemia to significantly higher levels compared to control experiments. This effect made it difficult to demonstrate that a relevant increase in LPS sensitivity was achieved in our model system and co-challenge with galactosamine was not used in the protection experiments.

3.4 Protection by OMV vaccination and primary infection monitored in the transferrin mouse model by bacteraemia and survival

B10.M mice, in groups of 10 animals, were s.c. vaccinated with 8 µg OMV given in two doses or naturally immunised with living meningococci given in a sub-lethal dose as previously described [13]. Normal non-immune mice (control group), OMV vaccinated mice and infected mice were then challenged after 7 weeks by iron-restricted meningococci (107 bacteria) co-administered with 20 mg human transferrin. As demonstrated in Fig. 3, the non-immunised control mice showed increasing bacteraemia during the observation period (five-fold increase at 24 h) and 100% mortality at 72 h (Table 2). In contrast, mice protected by OMV vaccination or primary infection showed a progressive decrease in bacteraemia during the first 24 h and finally a mortality of 44% and 0%, respectively (Table 2). At 24 h post-challenge, the concentration of meningococci in the blood of the control group, vaccinated group and the primary infected group of mice were 2×107, 3×105 and 5×104 CFU ml−1, respectively (Fig. 3). The difference in bacteraemia between control -and OMV vaccinated mice was significant after both 8 and 24 h (P<0.01). The results obtained represent a 450-fold difference between the control and maximally protected group studied here. Hence, both with respect to bacteraemia and mortality, the observed range between maximally protected animals (primary infected) and the non-immune control group should allow comparison of the protective effect of different OMV vaccines.

Figure 3

Bacteraemia (CFU ml−1 blood) of non-treated, OMV vaccinated and primary infected B10.M mice after challenge with 107 meningococci grown under iron restriction (60 µM EDDA) and with 20 mg human transferrin.

View this table:
Table 2

Survival of non-immunised, OMV vaccinated and primary infected B10.M mice challenged with 107 meningococci grown under iron restriction (60 µM EDDA) and 20 mg human transferrin

TreatmentSurvivals/tested% Survival
Normal control mice0/100
OMV vaccinated mice4/944
Primary infected mice10/10100
  • Ten mice were used in each experimental group and survival was recorded 72 h post-challenge.

We also compared the protective effect induced by a primary infection with meningococci cultured under iron-restricted versus normal conditions and found no significant difference neither with respect to bacteraemia nor to survival (results not shown).

4 Discussion

To develop improved vaccines against meningococcal disease, there is a general need for animal model systems which allow comparison of the protection induced by active immunisation with different candidate vaccines. Although the interpretation of results from protection studies in animal models is difficult, such results nevertheless represent valuable information in addition to assays for bactericidal and opsonophagocytic activity in human sera in vitro [11,12].

To evaluate the protective efficacy of OMV vaccines against infection with serogroup B meningococci, we have previously developed a protection model in normal B10.M mice based on transient bacteraemia [13]. This mouse model was used to demonstrate a significant degree of protection induced by OMV vaccination or a sub-lethal primary infection [13] when compared to non-immune control animals. However, this model system differs significantly from systemic meningococcal infection in humans. Normal non-immune mice only show transient bacteraemia and recover even when challenged with high doses of bacteria. One of several reasons is that the meningococcal strategy for capturing iron in vivo is not adopted to this species. The meningococcal transferrin receptor is specific for human transferrin and not able to bind and utilise mouse transferrin [1416]. To obtain a more physiological course of infection, we have in this work modified the present animal model by administration of human holo-transferrin upon bacterial challenge, which induced increasing bacteraemia and subsequent death in normal non-immune mice, but not in OMV vaccine protected or primary infected animals.

Meningococci express an array of iron-regulated surface accessible receptors that mediate specific binding and uptake of different iron containing proteins from the host of which transferrin is the most important [18]. A prerequisite for a relevant effect of exogenous human transferrin in the model system was, therefore, to challenge with meningococci grown under iron restriction to up-regulate the expression of the transferrin receptors [15,16,18]. By studying the in vitro growth of meningococci in the presence of increasing amounts of the Fe2+ chelator EDDA, we were able to determine the optimal concentration of the chelator (60 µM) that should be used to obtain challenge bacteria which fully utilise human transferrin for their growth. The subsequent titration of the human transferrin dose co-administered with bacteria expressing the transferrin receptor showed that 20 mg transferrin was necessary to induce increasing bacteraemia and 100% mortality in non-immune mice. Application of these challenge conditions to protection experiments showed increasing bacteraemia during the observation period and 100% mortality in non-immunised control mice, whereas the protective effect in mice subjected to OMV vaccination or primary infection could be measured as a progressive decrease in bacteraemia as well as lower mortality. Important to the comparative application of the modified mouse model system, was the large range observed between fully protected animals (primary infected) and the non-immune control group both with respect to bacteraemia (4×104–2×107 CFU ml−1 blood) as well as mortality (0–100%). This situation combined with the observation that the present OMV vaccine against meningococcal B disease protects at intermediate levels, should allow comparison of the protective effect of different OMV vaccines at least in relation to the present vaccine as a standard. In addition, the question of serotype-specific protection versus cross-protection could also be addressed.

Co-administration of galactosamine in this work was used to increase the sensitivity to the endotoxin effect of LPS [19,20], which potentially would allow us to reduce the meningococcal challenge dose and bacteraemia to a level approaching the human situation. Our observation that the anticipated increased mortality induced by galactosamine was directly associated with elevated bacteraemia, made it difficult to apply this approach in our model system. This effect can be explained by the known hepatotoxic effect of galactosamine [20], which may influence the ability of the mice to control the infection.

In conclusion, the modified model system for protection against meningococcal infection presented here makes it possible to compare the efficacy of different OMV vaccines by using bacteraemia as well as survival as parameters of protection. Importantly, the transferrin approach should also allow specific evaluation of the protective effect of antibodies against the transferrin receptors themselves which have been suggested to represent critical targets of the immune response to block systemic bacterial growth. The ability of purified meningococcal transferrin receptors (Tbp1 and Tpb2) to induce protection and bactericidal anti-Tbp antibodies in mice is indeed demonstrated [2125] and the model system established here could also be used to evaluate the importance of Tbps in protection induced by OMV vaccines.

Acknowledgements

This work was supported financially by Ninas Minnefond, Oslo, Norway. We also want to thank Kari G. Løken, Trude Olsen and Mona Iren Skullerud for excellent technical assistance.

References

  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].
View Abstract