Download Protective Anti-Helicobacter Immunity Is Induced with Aluminum

308
Protective Anti-Helicobacter Immunity Is Induced with Aluminum Hydroxide
or Complete Freund’s Adjuvant by Systemic Immunization
Judith M. Gottwein,1 Thomas G. Blanchard,2
Oleg S. Targoni,1 Julia C. Eisenberg,1
Brandon M. Zagorski,1 Raymond W. Redline,1
John G. Nedrud,1 Magdalena Tary-Lehmann,1
Paul V. Lehmann,1,a and Steven J. Czinn2
Departments of 1Pathology and 2Pediatrics, Case Western
Reserve University, Cleveland, Ohio
To determine whether systemic immunization against Helicobacter pylori could be achieved
with an adjuvant approved for human use, the efficacy of vaccination with Helicobacter
antigen in combination with aluminum hydroxide (AlOH) was evaluated in a murine model
of Helicobacter infection. Immunization with antigen and AlOH induced interleukin-5–
secreting, antigen-specific T cells, and immunization with antigen and complete Freund’s
adjuvant induced interferon-g–secreting, antigen-specific T cells, as determined by ELISPOT
assay. Both immune responses conferred protection after challenge with either H. pylori or
H. felis, as confirmed by the complete absence of any bacteria, as assessed by both histology
and culture of gastric biopsy samples. Protection was antibody independent, as demonstrated
with antibody-deficient mMT mice (immunoglobulin-gene knockout mice), and CD4+ spleen
T cells from immunized mice were sufficient to transfer protective immunity to otherwise
immunodeficient rag1⫺/⫺ recipients. These results suggest an alternative and potentially more
expeditious strategy for development of a human-use H. pylori vaccine.
Helicobacter pylori, one of the world’s most prevalent pathogens, is an extracellular bacterium that infects the gastric mucosa and plays an etiologic role in gastritis and peptic ulcer
disease [1, 2]. It is generally believed that infection occurs primarily in young children. Therefore, a prophylactic vaccine
administered to infants might prevent H. pylori infection and
the long-term consequences that occur in adults. Vaccination
strategies have focused primarily on orally and intranasally
administered immunizations to induce mucosal immunity [3].
These immunizations require bacterial exotoxin adjuvants that
are unsafe for use in humans. Systemic immunization has recently been reported as a possible means of inducing protective
Received 5 January 2001; revised 16 April 2001; electronically published
10 July 2001.
Presented in part: 100th annual meeting of the American Gastroenterology
Association, Orlando, Florida, April 1999 (abstract A695).
All animal handling and experimentation was reviewed by, and performed
according to the guidelines of, the Institutional Animal Use and Care Committee of Case Western Reserve University. The Case Western Reserve University animal facility is fully accredited by the American Association for
Accreditation of Animal Care.
Financial support: National Institutes of Health (research grants DK46461 and AI-36359 to S.J.C., DK-48799 and AI-42635-01 to P.V.L., and
AI-40701 to J.G.N.); National Multiple Sclerosis Society (grant 2470 A-1/
2 to P.V.L.). J.M.G. is a fellow of the Studienstifung des Deutschen Volkes.
a
P.V.L. is president of Cellular Technologies, whose product was used in
these studies to determine the number of cytokine-producing cells.
Reprints or correspondence: Dr. Steven J. Czinn, Dept. of Pediatrics,
Rainbow Babies & Children’s Hospital, 2101 Adelbert Rd., Cleveland, OH
44106 ([email protected]).
The Journal of Infectious Diseases 2001; 184:308–14
䉷 2001 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2001/18403-0008$02.00
immunity against H. pylori in mice [4]. The authors promoted
the idea that contribution of Th1-mediated responses, the response associated with H. pylori–related gastric pathology, are
required to induce immunity [4]. This suggestion was based on
immunoglobulin (Ig) G subclass analysis only; no direct measurement of cytokine profiles was performed [4, 5].
Because several reports have suggested that protection might
be mediated by type 2 immunity [6, 7], we investigated this in
a more direct fashion by use of parenteral immunization with
adjuvants that yield highly polarized Th1 or Th2 responses, as
measured by ELISPOT assay. In the present study, we investigated whether systemic immunization with the adjuvant aluminum hydroxide (AlOH) can induce protective immunity that
is mediated by Th2 CD4⫹ T cells in the absence of antibody.
AlOH-induced immunity would represent a significant advance
for the development of an H. pylori vaccine, because AlOH is
already approved for use in humans. Its safety, stability, and
low cost could expedite development of a human H. pylori
vaccine suited for large-scale application.
Materials and Methods
Mice. Mouse strains were purchased from Jackson Laboratories and were housed under specific-pathogen-free conditions in
microisolator units. The mice used were either C57BL/6 or C57BL/
6J-Rag1tm1Mom (rag1⫺/⫺), deficient in mature T and B lymphocytes,
or C57BL/6-Igh-6tm1Cgn (mMT), deficient in mature B cells, neither
of which can make antibodies.
Bacteria. H. felis was isolated in our laboratory from a feline
gastric biopsy sample [8]. H. pylori strain HpM6 was isolated from
JID 2001;184 (1 August)
H. pylori Immunity with a Human-Use Adjuvant
a human gastric biopsy sample and was adapted in our laboratory
to mice via long-term in vivo passage. HpM6 was determined to
be cagA positive by polymerase chain reaction. Inoculation of
C57BL/6 mice with HpM6 results in chronic infection that persists
for 112 months in 100% of infected mice. Both Helicobacter species
were identified on the basis of colony morphology, bacterial morphology, Gram stain, and the production of urease, catalase, and
oxidase. Bacteria were grown on solid media (Columbia blood
agar) and were resuspended in brucella broth.
Immunization and challenge of mice. Helicobacter lysate, which
we and others have demonstrated to be an effective experimental
vaccine antigen by the oral route [6, 8–11], was generated in our
laboratory, as described elsewhere [9]. Ovalbumin (OVA) was purchased from Sigma Chemical Co., AlOH (Imject Alum) was purchased from Pierce, and complete Freund’s adjuvant (CFA) was
made by mixing Mycobacterium tuberculosis H37RA (Difco Laboratories) at 1 mg/mL into incomplete Freund’s adjuvant (Gibco
BRL). Antigens in aqueous solution were mixed 1:1 with the adjuvant, and 100 mg of antigen in 100 mL of emulsion was injected
intraperitoneally into each mouse on day 0. On day 28, bacterial
challenge with 1 ⫻ 10 7 cfu of bacteria in 0.5 mL of nutrient broth
was given by gastric intubation with flexible tubing on an 18-gauge
needle. Bacterial numbers for challenge were determined by optical
density at 450 nm by use of a previously established growth curve.
Although H. felis is difficult to grow as isolated colonies, we use
a standard absorbance value that is based on the H. pylori growth
curve for both organisms, as described elsewhere [9].
Determination of vaccine efficacy. Twenty-eight days after challenge, the mice were examined for infection by Helicobacter organisms by silver staining of tissue (Steiner stain) [9, 12]. The animals were killed by CO2 asphyxiation, and a narrow strip of tissue
was surgically removed from the greater curvature of the stomach,
from the duodenum to the gastric cardia. Tissues were fixed in 10%
buffered formalin and were processed for histologic examination
at the University Hospitals of Cleveland Histology Laboratory.
Several sections of each mouse sample were stained by the Steiner
method, to facilitate the identification of H. pylori and H. felis on
the basis of bacterial location and morphology. The mice were
considered to be protected if there was a complete absence of detectable Helicobacter organisms in silver-stained sections.
Additionally, cultures were performed on gastric biopsy samples
from H. pylori–challenged mice to confirm the presence or absence
of bacteria. Two biopsy samples (2 ⫻ 2 mm) from mice challenged
with H. pylori were surgically removed from the gastric antrum and
were homogenized in 200 mL of Columbia broth. Homogenates were
plated on Columbia agar supplemented with 7% horse blood in 100mL aliquots. After 96 h in microaerophilic conditions, confirmation
of protection was determined by the absence of any culturable bacteria from immunized mice. When present, bacteria were confirmed
as H. pylori on the basis of colony morphology, Gram stain, and
the production of urease, catalase, and oxidase. Cultures for H. felis
were not performed because culture is less reliable, and H. felis tends
to grow as a bacterial lawn instead of as isolated colonies.
Evaluation of pathology. Longitudinal sections of the greater
curvature of the mouse stomach were evaluated for overall intensity
of inflammation, as described elsewhere [9, 13]. Sections included
the entire length of both the antral and fundic glandular mucosa.
Antral inflammation was graded on a 0–3 scale, and fundic in-
309
flammation on a 0–10 scale. A global score for each mouse was
determined on the basis of the linear extent (focal, multifocal,
patchy, or diffuse), depth (superficial and/or basal, panmucosal, or
extending to submucosa or muscular layers), and character of the
inflammatory infiltrate, as defined by the types of infiltrating cells
and tissue architecture changes.
Adoptive transfer. CD4⫹ T cells were purified with the mouse
T cell CD4 subset column kit (R&D Systems). Ten million cells
were transferred to each recipient mouse via injection into the tail
vein. These numbers have been established and are customarily
used in the field of autoimmune research [14]
ELISPOT assay. Single-cell suspensions were prepared from
the spleen, and 1 ⫻ 10 6 cells were plated per well in serum-free HL1 medium (BioWhittaker) supplemented with L-glutamine at 1 mM,
with or without Helicobacter antigens added, at a final concentration of 5 mg/mL. These cultures were added to ELISPOT plates
(ImmunoSpot; Cellular Technologies) that were precoated overnight with capture antibodies specific for interferon (IFN)–g or
interleukin (IL)–5, R46-A2 (4 mg/mL), or TRFK5 (5 mg/mL), respectively, in PBS. The plates were blocked with 1% bovine serum
albumin in PBS for 1 h at room temperature and were washed 4
times with PBS before the beginning of the 24-h cell culture. Next,
the cells were removed by washing, the detection antibody
(XMG1.2-HRP at 1 mg/mL for IFN-g or TRFK4 at 4 mg/mL for
IL-5) was added, and the plates were incubated overnight. For IL5, anti–IgG2a-HRP (Zymed) was added and was incubated for 2
h. The plate-bound final antibody then was visualized by adding
3-amino-9-ethyl-carbazole. To evaluate the results, we used a Series
1 ImmunoSpot Image Analyzer (Cellular Technologies).
Statistical analysis. Statistical significance between experimental groups for ELISPOT analysis was determined by analysis of
variance. The presence or absence of experimental infection after
challenge of immunized mice was evaluated for significance by
Fisher’s exact test.
Results
Murine models of Helicobacter infection and immunity have
used both the human pathogen itself or a related organism
isolated from cats, H. felis [15]. Unlike H. pylori, H. felis induces
inflammatory lesions in the gastric mucosa, reminiscent of human gastritis, when the mouse strain C57BL/6 is used [13, 16].
Additionally, H. felis appears to yield a heavier infection, as
gauged by histology, in the gastric tissue of mice, and unlike
H. pylori, which primarily resides at the antral-fundic junction,
its colonization pattern includes most of the antrum and fundus. We systemically immunized C57BL/6 mice with H. felis
or H. pylori antigens emulsified in either AlOH or CFA; we
have recently shown that these adjuvants can be used to induce
polarized type 2 and type 1 immunity, respectively [17]. Fourteen days later, the spleen cells of these mice were tested for
Helicobacter antigen–specific recall response by measuring a
type 1 and type 2 cytokine (IFN-g and IL-5, respectively) by
ELISPOT assay. Cells from mice immunized with antigen in
AlOH produced IL-5 in the virtual absence of IFN-g (figure
1A and 1C). The number of IL-5–producing cells was signifi-
310
Gottwein et al.
Figure 1. Induction of systemic, adjuvant-guided type 1 and type
2 immunity to Helicobacter antigens. Mice were injected with H. pylori
lysate in aluminum hydroxide (AlOH) (A) or complete Freund’s adjuvant (CFA) (B), and 14 days later their spleen cells were tested by
ELISPOT assays for interferon (IFN)–g (⽧) or interleukin (IL)–5 (䡬)
production in the presence of medium or H. pylori antigen, as indicated.
Each symbol represents the mean no. (of triplicate wells) of spot-forming cells per million spleen cells per individual mouse. AlOH induced
significantly more IL-5–producing cells than IFN-g–producing cells
(P ! .0001), whereas CFA induced significantly more IFN-g–producing
cells than IL-5–producing cells (P ! .0001). Mice immunized with H.
felis antigen in AlOH (C) or CFA (D) were tested under identical
conditions. Again, AlOH induced significantly more IL-5–producing
cells than IFN-g–producing cells (P ! .003 ), whereas CFA induced significantly more IFN-g–producing cells than IL-5–producing cells
(P ! .0001). The data are from 1 experiment, representative of the 3
experiments performed. Statistical significance of differences between
experimental groups was determined by analysis of variance.
cantly greater than that of antigen-specific IFN-g–producing
cells for both H. pylori antigen immunization (P ! .0001) and
H. felis antigen immunization (P ! .003 ). The reverse cytokine
profile was seen after immunization with antigen and CFA (figure 1B and 1D). The number of IFN-g–producing cells induced
by the recall response was significantly greater than that of IL5–producing cells (P ! .0001 for both H. pylori and H. felis
antigens). Additionally, serum samples from the CFA-immunized mice contained high titers of specific IgG2a antibodies,
whereas AlOH-immunized mice had an IgG1 antibody response that prevailed over IgG2a (data not shown). Thus, on
the basis of both the relative numbers of cells producing the
respective cytokines and the antibody isotype profiles of these
immune responses, immunization with AlOH induced polarized
JID 2001;184 (1 August)
type 2 immunity, whereas immunization with CFA triggered
type 1 immunity against the coinjected Helicobacter antigens.
We then tested whether the type 1 or type 2 immunity induced
by systemic immunization with Helicobacter antigens would
confer protection against Helicobacter infection of the gastric
mucosa. C57BL/6 mice were immunized with H. pylori or H.
felis antigens emulsified with AlOH or CFA; control mice were
injected with these adjuvants containing an irrelevant control
protein, OVA, or remained unimmunized. Mice were challenged
orally with 1 ⫻ 10 7 bacteria of the respective Helicobacter species on day 28, and 28 days later, the bacteria load was assessed
by direct visualization of silver-stained histologic sections of
the gastric mucosa (Steiner staining). The presence or absence
of Helicobacter organisms was confirmed by culture (see Materials and Methods). These studies of fully immunocompetent,
wild-type mice demonstrated that systemic immunization with
either adjuvant afforded profound protection, as determined
by the complete absence of Helicobacter organisms in silverstained histologic sections (table 1). Although 12 (86%) of 14
control mice became infected with H. pylori, 7 (88%) of 8 mice
injected with H. pylori:AlOH were free of infection. This protection was significant for both AlOH (P p .005) and CFA
(P p .049) adjuvants.
Results with H. felis, which causes a more aggressive infection
in mice than does H. pylori, were even more striking. Ninetysix percent (28 of 29) of the control mice became infected and
exhibited a high bacteria load, averaging 86 infected glands per
Table 1. CD4⫹ T cell–mediated protection against Helicobacter infection induced by systemic vaccination in mice.
C57BL/6
a
strain
WT
mMT
Experimental
group
Immunization
Challenge
No.
d
infected
A
B
C
D
E
F
G
H
I
J
K
Hp:AlOH
Hp:CFA
OVA:AlOH
OVA:CFA
Hf:AlOH
Hf:CFA
OVA:AlOH
OVA:CFA
None
Hf:AlOH
OVA:AlOH
Hp
Hp
Hp
Hp
Hf
Hf
Hf
Hf
Hf
Hf
Hf
1/8
2/8
7/8
5/6
0/5
5/21
5/5
16/17
7/7
1/8
8/8
b
c
e
P
.005
.049
.004
!.001
!.001
NOTE. AlOH, aluminum hydroxide; CFA, complete Freund’s adjuvant; Hf,
Helicobacter felis antigen; Hp, Helicobacter pylori antigen; OVA, ovalbumin; WT,
wild type.
a
Strains included C57BL/6 (WT) mice (groups A–I) and immunoglobulingene knockout (mMT) mice on the same background (groups J and K).
b
A total of 100 mg of Hp, Hf, or OVA, with AlOH or CFA, as specified.
c
Mice were challenged orally with 1 ⫻ 107 cfu of the respective Helicobacter strain.
d
Protection is defined as the complete absence of detectable bacteria. The
bacteria load in the gastric mucosa was assessed in all groups of mice 4 weeks
after infectious challenge, by counting silver stain–positive bacteria and scoring
as positive glands per longitudinal section. For experiments performed with Hp,
cultures also were performed on biopsy samples, to confirm the absence of bacteria in protected mice.
e
The presence or absence of experimental infection after challenge of immunized mice was evaluated by Fisher’s exact test for significance vs. corresponding OVA-immunized mice.
JID 2001;184 (1 August)
H. pylori Immunity with a Human-Use Adjuvant
stomach section. No bacteria could be found in any of the 5
H. felis:AlOH–vaccinated mice. Immunizations with Helicobacter antigens in CFA were also protective (5 of 21 mice were
infected). This level of protection was significant for both AlOH
(P p .004) and CFA (P ! .001) adjuvants and is similar to what
we and others have observed when the most effective known
mucosal vaccination strategies that require cholera toxin as the
adjuvant are used (reviewed in [3]). Figure 2 demonstrates the
variation in bacteria load observed in silver-stained gastric sections among groups of mice. No bacteria were observed in H.
felis:AlOH–vaccinated mice, versus large numbers of infected
glands in sham-immunized mice and small numbers of infected
glands in H. felis:CFA–vaccinated mice.
Systemic induction of type 1 or type 2 immunity may, therefore, be a viable alternative to mucosal immunization against
311
Helicobacter infection. Examination of hematoxylin-eosin–
stained sections from these mice revealed that immunization of
mice with H. felis antigens in combination with either AlOH
or CFA each resulted in equivalent levels of gastric inflammation at both the antrum and fundic mucosa. The degree and
character of the inflammation were not appreciably different
from those previously observed by us for orally administered
immunization with Helicobacter antigen and cholera toxin adjuvant [9]. Mice immunized with Helicobacter antigens had average fundic inflammation of 7.2 Ⳳ 0.9 and 7.4 Ⳳ 0.7 when
AlOH or CFA was used, respectively. Antral inflammation was
2.0 Ⳳ 0 and 2.1 Ⳳ 0.6 for AlOH and CFA, respectively. In all
cases, mice previously immunized with Helicobacter antigen
had a greater degree of inflammation after challenge than did
mice immunized with OVA. This is consistent with the post-
Figure 2. Histologic illustration that systemic immunization prevents Helicobacter infection. Sections of gastric mucosa from immunized mice
challenged with H. felis are shown. The most heavily colonized sections are shown for mice immunized with H. felis sonicate plus aluminum
hydroxide (AlOH; table 1, group E, no organisms visible) (A) and H. felis sonicate plus complete Freund’s adjuvant (CFA; table 1, group F) (B),
and typical sections are shown for mice that were sham immunized with ovalbumin (OVA) plus AlOH (table 1, group G) (C) and OVA plus
CFA (table 1, group H) (D). Arrows indicate an individual H. felis organism or a group of H. felis organisms in the tissue section (Steiner stain;
original magnification, ⫻200).
312
Gottwein et al.
immune gastritis observed in many laboratories and is most
likely due to the development of anti-Helicobacter immune
memory secondary to the immunization.
We next began to investigate the immune mechanism that
mediates anti-Helicobacter immunity after systemic immunization. Because CFA, like cholera toxin, is highly toxic and
unsuitable for use in humans, we focused on the use of AlOH
as an adjuvant. In these subsequent experiments, we used the
murine–H. felis model, because this organism exhibits more
pronounced inflammation in mice than does H. pylori and also
results in a heavier infection that can readily be detected in
histologic sections. We immunized congenic mMT mice (immunoglobulin-gene knockout mice) with H. felis antigen:
AlOH, or with OVA:AlOH (table 1, groups J and K) to determine whether the presence of antibody is necessary for protection after parental immunization. All 8 of the OVA-control
mMT mice became infected after challenge with H. felis, exhibiting a high bacteria load (n p 65 positive glands per longitudinal section of gastric mucosa). However, only one of the
H. felis antigen:AlOH–injected mMT mice became infected (2
positive glands per longitudinal section of gastric mucosa), indicating significant protection (P ! .001 ). Inflammation in these
mice was not appreciably different from that in the fully immunocompetent B6 mice, because immunization with H. felis
antigen and AlOH resulted in postchallenge fundic inflammation of 7.8 Ⳳ 0.6 and antral inflammation of 2.38 Ⳳ 0.5.
Immunization with OVA and AlOH resulted in average scores
of 3.7 Ⳳ 1.5 in the fundus and 1.5 Ⳳ 0.5 in the antrum. This is
consistent with our previous use of mMT mice for orally administered immunization [9].
Because these experiments suggested that antibodies are not
JID 2001;184 (1 August)
required for immunity after parenteral vaccination, we next performed an experiment to determine whether purified T cell populations from vaccinated mice can adoptively transfer protection.
CD4⫹ T cells were purified by affinity columns from the spleens
of C57BL/6 mice on day 28 after immunization with either H.
felis antigen:AlOH, OVA:AlOH, H. felis antigen:CFA, or OVA:
CFA. Ten million cells (194% CD4⫹ by flow cytometry) were
injected into each congenic, immunodeficient rag1⫺/⫺ recipient
mouse (table 2, groups L, M, P, and Q). The recipient mice were
challenged 3 days later with 1 ⫻ 10 7 H. felis organisms, and 28
days later, the bacteria load of the gastric mucosa was measured
by inspection of silver-stained histologic sections. All 3 rag1⫺/⫺
mice reconstituted with OVA:AlOH–primed CD4⫹ T cells became heavily infected, exhibiting an average of almost 30 positive
glands per section. Similarly, all 8 rag1⫺/⫺ mice reconstituted with
OVA:CFA–primed CD4⫹ T cells had average H. felis loads of
150 positive glands per section.
In contrast, 6 (75%) of the 8 rag1⫺/⫺ mice that were grafted
with H. felis antigen:AlOH–primed CD4⫹ T cells had undetectable levels of bacteria (table 2, group L). Furthermore, the 2 mice
that did become infected had a markedly reduced number of
total Helicobacter organisms, ∼4% of that observed in control
mice (data not shown). Therefore, the gastric immunity induced
by systemic immunization with H. felis antigen and AlOH was
mediated by CD4⫹ T cells. Although this protection did not quite
achieve statistical significance (P p .061 ), this was most likely
because of the reduced size of the control group (table 2, group
M), the result of accidental deaths after a cage flooded. Histologic
inflammation of these mice after challenge with H. felis was similar to the inflammation observed for challenge of immunized
mice. The mice receiving CD4⫹ T cells from Helicobacter-im-
Table 2. Helicobacter immunity in rag1⫺/⫺ mice by adoptive transfer of Agprimed CD4⫹ T cells.
C57BL/6
a
strain
WT
WT
WT
WT
WT
WT
WT
WT
CD4rrag1⫺/⫺
CD4rrag1⫺/⫺
CD4rrag1⫺/⫺
CD4rrag1⫺/⫺
Experimental
group
Immunization
Challenge
No.
d
infected
L
M
N
O
P
Q
R
S
Hf:AlOH
OVA:AlOH
Hf:AlOH
OVA:AlOH
Hf:CFA
OVA:CFA
Hf:CFA
OVA:CFA
Hf
Hf
Hf
Hf
Hf
Hf
Hf
Hf
2/8
f
3/3
5/15
7/7
7/7
8/8
3/10
8/9
b
c
e
P
.061
.023
1.0
.014
NOTE. AlOH, aluminum hydroxide; CFA, complete Freund’s adjuvant; Hf, Helicobacter
felis antigen; OVA, ovalbumin; WT, wild type.
a
Strains included C57BL/6 (WT) mice (groups N, O, R, and S) and rag1⫺/⫺ mice on the
same background (groups L, M, P, and Q).
b
A total of 100 mg of Hf or OVA with AlOH or CFA, as specified. Naive rag1⫺/⫺ mice were
injected with 1 ⫻ 107 CD4⫹ T cells isolated from spleens of WT mice immunized with either Hf
or OVA, in combination with either AlOH or CFA, and challenged with Hf 3 days later.
c
Mice were challenged orally with 1 ⫻ 107 cfu of H. felis.
d
Protection is defined as the complete absence of detectable bacteria. The bacteria load in
the gastric mucosa was assessed in all groups of mice 4 weeks after infectious challenge by
counting silver stain–positive bacteria and scoring as positive glands per longitudinal section.
e
The presence or absence of experimental infection after challenge of immunized mice
was evaluated for significance vs. corresponding OVA-immunized mice by Fisher’s exact test.
f
The small size of the group of mice receiving CD4⫹ T cells from OVA:AlOH-immunized
mice was the result of the premature deaths of some of the recipient mice by drowning when
their cage was accidentally flooded.
JID 2001;184 (1 August)
H. pylori Immunity with a Human-Use Adjuvant
munized mice had average fundic inflammation of 7.3 Ⳳ 0.9 and
antral inflammation of 2.3 Ⳳ 0.5. Inflammation in mice receiving
OVA-primed T cells was slightly lower, with fundic scores averaging 6.7 Ⳳ 0.6 and antral scores of 1.7 Ⳳ 0.6.
Mice engrafted with H. felis antigen:CFA–primed CD4⫹ T
cells and then challenged with H. felis became uniformly infected, with average bacteria loads of 90 infected glands per
section (table 2). Apparently an equal number of CFA-primed
Th1 cells was less effective in transferring protection than was
the same number of AlOH-primed Th2 cells. This is consistent
with the reduced efficacy in CFA-immunized animals, compared with AlOH-immunized animals, observed in our other
experiment (table 1, groups B and F). Alternatively, in the absence of regulatory networks present in wild-type mice, polarized Th2 CD4⫹ T cells alone (but not polarized Th1 cells alone)
may be sufficient to mediate reduction in bacteria loads in the
gastric mucosa.
Discussion
In summary, we have demonstrated that systemic immunization that uses AlOH as an adjuvant can induce protective
immunity against H. pylori and H. felis in mice that is mediated
by CD4⫹ type 2 cells. These findings may have a direct impact
on the development of a Helicobacter vaccine for humans, for
several reasons. First, and perhaps most importantly, the present report strengthens the concept that systemic immunization
with an adjuvant approved for use in humans may be a viable
alternative to experimental orally administered immunization
for the induction of mucosal immunity against H. pylori. In
further support of this, in a separate series of experiments designed to investigate the efficacy of parental immunization
against H. pylori in neonates, we successfully and reproducibly
used subcutaneous immunization to obtain protective immunity (authors’ unpublished data). This concept, backed by a
recent report by Guy et al. [4], introduces a new strategy for
H. pylori vaccination that is contrary to long-held principles
of mucosal immunity.
Second, contrary to the findings of Guy et al. [4], which
indicated a requirement for type 1 responses in Helicobacter
immunity, our data demonstrate that type 2 responses are at
least equal to type 1 responses and are perhaps more efficacious
in their ability to confer protective immunity against Helicobacter infection. AlOH, the only adjuvant available for human
use, is a type 2 adjuvant. Some recent reports have suggested
a role for mucosally induced type 2 immunity in protecting
hosts from Helicobacter, including orally administered immunizations with cholera toxin and other experimental adjuvants
that have elicited type 2 or mixed type 2–type 1 responses [6,
7]. Recently, Escherichia coli heat-labile toxin was used as an
adjuvant to parentally immunize mice against H. pylori [5]. The
heat-labile toxin was found to significantly increase the serum
IgG1:IgG2a ratio, a possible indicator of type 2 immunity. Our
data now show, at the cellular level, that a human-use adjuvant,
313
AlOH, directs a polarized type 2 systemic immune response
that confers protection from Helicobacter infection in mice.
Presumably, local release of either type 1 or type 2 cytokines
in the stomach after parenteral immunization may be effective
in recruiting or activating the effector cells that eliminate the
bacteria. As such, it is possible that experimental Th1 adjuvants, such as CpG oligonucleotides, may also eventually prove
to be effective at inducing protective H. pylori immunity. However, AlOH is already approved for human use and is widely
used in humans, and the present study is encouraging in that
it demonstrates that AlOH-induced type 2 immunity can induce
protective immunity in mice.
Third, we found that the Helicobacter immunity induced by
systemic immunization can be conveyed by CD4⫹ T cells and
can occur in the absence of antibodies. The gastric mucosa must,
therefore, be included in the trafficking pathway of systemically
primed CD4⫹ T cells providing immune surveillance [18]. This
is consistent with recent reports by us and others demonstrating
that protective immunity against Helicobacter, induced by mucosal immunization, is also antibody independent [9, 11, 19].
Induction of IgA-independent protective immunity against a
noninvasive gastrointestinal bacterium suggests a novel CD4⫹ T
cell–mediated immune effector mechanism at the gastric mucosa.
Fourth, we have shown here that the type 1 bias of C57BL/
6 mice [20] can be overridden by the intrinsic type 2 polarizing
effects of AlOH [21]. Because humans are also biased toward
either type 1 or type 2 responses [22–24], this finding suggests
that uniform success might be accomplished with minimal side
effects even in the mixed human population. The primary candidate for preventive vaccination will be young children, because they typically become infected within the first several
years of life. Newborns are type 2 biased; therefore, the induction of type 2 immunity will be even further facilitated. Our
ongoing work with neonatal mice has confirmed this concept
(authors’ unpublished data). In humans, ulcers and gastritis are
associated with a type 1 IFN-g response to H. pylori in the
gastric mucosa [25–27], whereas AlOH induces Th2-type responses. Although AlOH-containing vaccines can induce antigen-specific IgE responses [17], AlOH has been widely used
for human immunizations, including the common diphtheriatetanus–pertussis vaccinations. The notion put forth in this report—that a systemic immunization with Helicobacter antigens
in AlOH is suited to induce a degree of immune protection that
has been accomplished previously only by the use of toxic mucosal adjuvants or proprietary experimental adjuvants—should
provide a novel rationale for, and expedite the development of,
an H. pylori vaccine.
References
1. Warren JR, Marshall BJ. Unidentified curved bacilli on gastric epithelium
in active chronic gastritis. Lancet 1983; 1:1273–5.
2. NIH Consensus Conference. Helicobacter pylori in peptic ulcer disease.
JAMA 1994; 272:65–9.
3. Blanchard TG, Czinn SJ, Nedrud JG. Host response and vaccine development.
314
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Gottwein et al.
In: Westblom TU, Czinn SJ, Nedrud JG, eds. Gastroduodenal disease and
Helicobacter pylori. Vol 241. Berlin: Springer-Verlag, 1999:181–213.
Guy B, Hessler C, Fourage S, et al. Systemic immunization with urease
protects mice against Helicobacter pylori infection. Vaccine 1998; 16:850–6.
Weltzin R, Guy B, Thomas J, WD, Gianasca PJ, Monath TP. Parental adjuvant activaties of Escherichia coli heat-labile toxin and its B subunit for
immunization of mice against gastric Helicobacter pylori infection. Infect
Immun 2000; 68:2775–82.
Mohammadi M, Nedrud J, Redline R, Lycke N, Czinn S. Murine CD4 T
cell responses to Helicobacter infection: TH1 cells enhance gastritis and
TH2 cells reduce bacterial load. Gastroenterology 1997; 113:1848–57.
Saldinger PF, Porta N, Launois P, et al. Immunization of BALB/c mice with
Helicobacter urease B induces a T helper 2 response absent in Helicobacter
infection. Gastroenterology 1998; 115:891–7.
Czinn SJ, Cai A, Nedrud JG. Protection of germ-free mice from infection
by Helicobacter felis after active oral or passive IgA immunization. Vaccine 1993; 11:637–42.
Blanchard TG, Czinn SJ, Redline RW, Sigmund N, Harriman G, Nedrud
JG. Antibody-independent protective mucosal immunity to gastric Helicobacter infection in mice. Cell Immunol 1999; 191:74–80.
Mohammadi M, Czinn S, Redline R, Nedrud J. Helicobacter-specific cellmediated immune responses display a predominant TH1 phenotype and
promote a DTH response in the stomachs of mice. J Immunol 1996; 156:
4729–38.
Sutton P, Wilson J, Kosaka T, Wolowczuk I, Lee A. Therapeutic immunization against Helicobacter pylori infection in the absence of antibodies.
Immunol Cell Biol 2000; 78:28–30.
Blanchard TG, Lycke N, Czinn SJ, Nedrud JG. Recombinant cholera toxin
B subunit is not an effective mucosal adjuvant for oral immunization of
mice against H. felis. Immunology 1998; 94:22–7.
Mohammadi M, Redline R, Nedrud J, Czinn S. Role of the host in pathogenesis of Helicobacter-associated gastritis: H. felis infection of inbred and
congenic mouse strains. Infect Immun 1996; 64:238–45.
Heeger PS, Forsthuber T, Shive C, et al. Revisiting tolerance induced by autoantigen in incomplete Freund’s adjuvant. J Immunol 2000;164:5771–81.
Lee A, Fox JG, Otto G, Murphy J. A small animal model of human Helicobacter pylori active chronic gastritis. Gastroenterology 1990; 99:1315–23.
JID 2001;184 (1 August)
16. Sakagami T, Dixon M, O’Rourke J, et al. Atrophic gastric changes in both
H. felis and H. pylori infected mice are host dependent and seperate from
antral gastritis. Gut 1996; 39:639–48.
17. Yip HC, Karulin AY, Tary-Lehmann M, et al. Adjuvant-guided type 1 and
type 2 immunity: infectious/noninfectious dichotomy defines the class of
response. J Immunol 1999; 162:3942.
18. Michetti M, Kelly CP, Kraehenbuhl J-P, Bouzourene H, Michetti P. Gastric
mucosal a4b7-integrin–positive CD4 T lymphocytes and immune protection
against Helicobacter infection in mice. Gastroenterology 2000;119:109–18.
19. Ermak TH, Giannasca PJ, Nichols R, et al. Immunization of mice with urease
vaccine affords protection against Helicobacter pylori infection in the absence of antibodies and is mediated by MHC class II–restricted responses.
J Exp Med 1998; 188:2277–88.
20. Hsieh C-S, Macatonia SE, O’Garra A, Murphy KM. T cell genetic background determines default T helper phenotype development in vitro. J
Exp Med 1995; 181:713–21.
21. Forsthuber T, Yip HC, Lehmann PV. Induction of TH1 and TH2 immunity
in neonatal mice. Science 1996; 271:1728–30.
22. Shirakawa T, Enomoto T, Shimazu S-I, Hopkin JM. The inverse association
between tuberculin responses and atopic disorder. Science 1997; 275:77–9.
23. Scott P, Natovitz RL, Coffman RL, Pearce E, Sher A. Immunoregulation
of cutaneous leishmaniasis: T cell lines that transfer protective immunity
or exacerbation belong to different T helper subsets and respond to distinct
parasite antigens. J Exp Med 1988; 168:1675–84.
24. Reiner SL, Locksley RM. The regulation of immunity to Leishmania major.
Annu Rev Immunol 1995; 13:151–77.
25. D’Elios MM, Manghetti M, De Carli M, et al. T helper 1 effector cells specific
for Helicobacter pylori in the gastric antrum of patients with peptic ulcer
disease. J Immunol 1997; 158:962–7.
26. Karttunen RA, Karttunen TJ, Yousfi MM, El-Zimaity H, Graham DY, ElZaatari F. Expression of mRNA for interferon-gamma, interleukin-10,
and interleukin-12 (p40) in normal gastric mucosa and in mucosa infected
with Helicobacter pylori. Scand J Gastroenterol 1997; 32:22–7.
27. Bamford KB, Fan X, Crowe SE, et al. Lymphocytes in the human gastric
mucosa during Helicobacter pylori have a T helper cell 1 phenotype.
Gastroenterology 1998; 114:482–92.