of June 2, 2013.
This information is current as
Intestinal Antitoxic Protection
Gut IgA and Abrogates the Development of
Lack of J Chain Inhibits the Transport of
Nils Lycke, Lena Erlandsson, Lena Ekman, Karin Schön and
1999; 163:913-919; ;
, 12 of which you can access for free at:
cites 25 articles
is online at:
The Journal of Immunology
Information about subscribing to
Submit copyright permission requests at:
Receive free email-alerts when new articles cite this article. Sign up at:
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Immunologists All rights reserved.
Copyright © 1999 by The American Association of
9650 Rockville Pike, Bethesda, MD 20814-3994.
The American Association of Immunologists, Inc.,
is published twice each month by
The Journal of Immunology
by guest on June 2, 2013
Lack of J Chain Inhibits the Transport of Gut IgA and
Abrogates the Development of Intestinal Antitoxic Protection1
Nils Lycke,2* Lena Erlandsson,†Lena Ekman,* Karin Scho ¨n,* and Tomas Leanderson†
Recent publications have provided confusing information on the importance of the J chain for secretion of dimeric IgA at mucosal
surfaces. Using J chain-deficient (J chain?/?) mice, we addressed whether a lack of J chain had any functional consequence for
the ability to resist challenge with cholera toxin (CT) in intestinal loops. J chain?/?mice had normal levels of IgA plasma cells
in the gut mucosa, and the Peyer’s patches exhibited normal IgA B cell differentiation and germinal center reactions. The total
IgA levels in gut lavage were reduced by roughly 90% as compared with that in wild-type controls, while concomitantly serum
IgA levels were significantly increased. Total serum IgM levels were depressed, whereas IgG concentrations were normal. Fol-
lowing oral immunizations with CT, J chain?/?mice developed 10-fold increased serum antitoxin IgA titers, but gut lavage
anti-CT IgA levels were substantially reduced. However, anti-CT IgA spot-forming cell frequencies in the gut lamina propria were
normal. Anti-CT IgM concentrations were low in serum and gut lavage, whereas anti-CT IgG titers were unaltered. Challenge of
small intestinal ligated loops with CT caused dramatic fluid accumulation in immunized J chain?/?mice, and only 20% protection
was detected compared with unimmunized controls. In contrast, wild-type mice demonstrated 80% protection against CT chal-
lenge. Mice heterozygous for the J chain deletion exhibited intermediate gut lavage anti-CT IgA and intestinal protection levels,
arguing for a J chain gene-dosage effect on the transport of secretory IgA. This study unequivocally demonstrates a direct
relationship between mucosal transport of secretory SIgA and intestinal immune protection. The Journal of Immunology, 1999,
duction and transport of Abs, which are formed by plasma cells in
the lamina propria (LP)3. The predominant isotype is IgA, whereas
IgM and IgG Ab-producing cells are infrequently found at most
mucosal surfaces (1). Of these three isotypes, IgA and IgM, but not
IgG, are actively transported via the epithelial cell into the lumen
(1, 2). Therefore, intestinal secretions contain over 80% polymeric
IgA, some polymeric IgM, but almost no IgG Abs (1, 2). In re-
sponse to infection or to immunization with Ag, specific Abs are
formed, although even in the noninfected naive animal intestinal
IgA formation is high, probably reflecting the constant stimulation
of local immune responses by the commensal gut flora (3, 4). Thus,
the barrier functions of the mucosa appear to critically depend on
an efficient transport system of dimeric IgA, and perhaps also of
polymeric IgM (4).
ucosal surfaces constitute a first line of defense against
microbial invasion of the body. Apart from the mucus
layer, the barrier is maintained by the continuous pro-
This transport is mediated by up-take of dimeric IgA and poly-
meric IgM containing a J chain, which binds to the polymeric Ig
receptor (pIgR) expressed on the basolateral side of the epithelial
cell (1, 2). Thus, the plasma cell produces the J chain, which is a
15-kDa glycoprotein that covalently associates by disulfide link-
ages with polymeric IgA and IgM, but not to IgG, before secretion
(1, 2, 5). After binding of the J chain Ig polymer to the pIgR, the
complex is transported through the epithelial cell to the luminal
side, where the receptor is proteolytically cleaved off (2, 5). This
results in the formation of secretory IgA (sIgA), shown to be par-
ticularly resistant to proteolytic degradation (1, 2). Whether poly-
meric IgM at mucosal surfaces also has increased stability com-
pared with serum IgM is less well known (1, 4).
To investigate in vivo the role of the pIgR pathway for the
transport of IgA into the lumen, Hendrickson et al. recently de-
veloped a J chain-deficient mouse (6, 7). In their first report, the
concentration of IgA in feces was dramatically lowered compared
with wild-type (WT) mice (6). However, in a follow-up study,
using a different detection system, IgA Abs, albeit monomeric
forms, were present in normal concentrations at the mucosal sur-
faces in the intestine, mammary gland, and lung in J chain-defi-
cient mice, suggesting that the J chain was not absolutely required
for IgA transport into the lumen (7). This unexpected finding con-
trasted with the requirement for J chain in the hepatic transport of
IgA found in these same mice (6). Although most previous studies
have demonstrated a clear association between the presence of J
chain and transport of IgA into the lumen (1, 2, 4, 5, 8), there are
a few studies that support the view that IgA may reach the lumen
through a J chain-independent pIgR-mediated transport (2, 9, 10).
The mechanism or the functional relevance of such a system has
not been investigated.
Cholera toxin (CT) induces severe diarrhoeal disease (11). Fol-
lowing active oral immunization in humans and experimental an-
imals, a high degree of antitoxic protection is induced (11, 12).
Protection is thought to depend on the presence of intestinal sIgA
*Department of Medical Microbiology and Immunology, University of Go ¨teborg,
Go ¨teborg, Sweden; and†Immunology Group, Department of Cellular and Molecular
Immunology, Lund University, Lund, Sweden
Received for publication February 17, 1999. Accepted for publication April 26, 1999.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This study was supported by the Foundation for Strategic Research, European
Union-Biotechnology Grant (Bio4-CT-98-0505), the World Health Organization
GPV-Transdisease Program, the Swedish Medical Research Council, the National
Institutes of Health (Grant 1R01A140701-01), the Swedish Cancer Foundation, and
the Crafoord O¨sterlund and Kocks Foundations.
2Address correspondence and reprint requests to Dr. Nils Lycke, Department of
Medical Microbiology and Immunology, University of Go ¨teborg, S-413 46 Go ¨teborg,
Sweden. E-mail address: email@example.com
3Abbreviations used in this paper: LP, lamina propria; ASF, anti-secretory factor;
CT, cholera toxin; PP, Peyer’s patches; sIgA, secretory IgA; SFC, spot-forming cells;
GC, germinal centers; PNA, peanut hemagglutinin; WT, wild type; pIgR, polymeric
Ig receptor; CMF, Ca2?- and Mg2?-free; AP, alkaline phosphatase.
Copyright © 1999 by The American Association of Immunologists0022-1767/99/$02.00
by guest on June 2, 2013
Abs (13–15), although some investigators have argued for other
factors, such as IFN-?, neuropeptides, or anti-secretory factors
(ASF), to be responsible for the development of antitoxic protec-
We recently developed a mouse with an inactivated J chain and
have reported on the systemic effects on IgM secretion of this gene
deletion (19). In the present study, we addressed whether the J
chain plays an essential role in the secretion and transport of in-
testinal antitoxin IgA and if this process is associated with an in-
creased resistance to CT challenge of small intestinal ligated loops
in orally immunized mice.
Materials and Methods
WT (J chain?/?) mice or mice with homozygous (J chain?/?) or heterozy-
gous (J chain?/?) deletions of the J chain (19) were bred at the Department
of Cellular and Molecular Biology at Lund University (Lund, Sweden) and
housed under pathogen-free conditions at the Department of Medical Mi-
crobiology and Immunology at the University of Go ¨teborg (Go ¨teborg,
Sweden) before experimentation.
Age- and sex-matched mice were given three oral immunizations 10 days
apart with 10 ?g of CT (List Biological Laboratories, Campbell, CA) in
6% (w/v) NaHCO3in PBS through a baby feeding tube under light ether
anesthesia (13, 20). Control mice received PBS alone. Analysis of small
intestinal antitoxic protection and systemic and local antitoxin Ab produc-
tion were performed 4 or 7 days after the final immunization as indicated.
Intestinal lavage samples
Intestinal secretions for Ab determinations were collected by a method
adopted from Elson et al. (21). Briefly, the small intestines were removed,
rinsed in PBS to discard feces, and carefully injected with 3 ml of a pro-
tease inhibitor solution consisting of 0.1 mg/ml soybean trypsin inhibitor
(Sigma, St. Louis, MO), 50 mM EDTA (Merck, Darmstadt, Germany), and
1 mM PMSF (Boehringer Mannheim, Mannheim, Germany) in PBS. After
incubation for 10 min at room temperature, the intestinal content was trans-
ferred to a test tube, vigorously vortexed, sonicated, and centrifuged for 10
min at 1800 rpm at 4°C. The supernatant was transferred to a microfuge
tube, and PMSF (Boerhinger Mannheim) was added to a final concentra-
tion of 1 mM. After mixing the solution, it was again centrifuged at
13,000 ? g for 15 min at 4°C. The resulting supernatant was mixed with
PMSF and sodium azide to a final concentration of 1 mM and 0.001%,
respectively, and incubated at 4°C for 30 min. Thereafter, 50 ?l FCS was
added per ml solution, which was then centrifuged for an additional 15 min
at 13,000 ? g at 4°C. The supernatants were stored at ?70°C until
Preparation of Peyer’s patch (PP) and LP lymphocytes
PP were excised from the serosal side of the intestine and passed through
a nylon net as described (22). Intestinal LP lymphocytes were prepared as
previously described (22). Briefly, after thorough washing in Ca2?- and
Mg2?-free HBSS (CMF-HBSS) (Life Technologies, Paisley, U.K.), the
tissue pieces were incubated in CMF-HBSS containing 5 mM EDTA
(Merck) to remove epithelial cells and intraepithelial lymphocytes. The
intestinal pieces were then incubated in RPMI 1640 (Life Technologies)
containing collagenase type C-2139 (300 U/ml; Sigma) to enzymatically
extract the LP lymphocytes. The single-cell suspensions of PP and LP
lymphocytes were then washed three times in CMF-HBSS (Life Technol-
ogies) and diluted to a final concentration of 2 ? 106cells/ml in Iscove’s
medium (Life Technologies) containing 5% FCS (Life Technologies).
Serum and lavage Ab determinations by ELISA
Flat-bottom 96-well microtiter plates (Nunc, Roskilde, Denmark) were
coated with 0.5 nmol/ml GM1 ganglioside (Sigma) in PBS at 4°C over-
night followed by 0.5 ?g/ml CT (List Biological Laboratories) in PBS at
4°C overnight as described (23). Total IgA, IgM, or IgG determinations in
serum or lavage were performed using highly isotype-specific unlabeled
goat anti-mouse IgA, IgM, or IgG Abs (Southern Biotechnology Associ-
ates, Birmingham, AL) at 5 ?g/ml in PBS at 4°C overnight. After washing
in PBS and blocking with 0.1% BSA/PBS for 30 min at 37°C, serial 3-fold
dilutions of sera at 1/200 or intestinal lavage at 1/20 was performed in
corresponding subwells in 0.1%BSA/PBS, and the plates were incubated at
4°C overnight. For antitoxin-specific IgA, IgM, or IgG determinations,
serum was diluted 1/100 and lavage was diluted 1/2 and subjected to 3-fold
serial dilutions in corresponding subwells. After washing in PBS/0.05%
Tween 20, alkaline phosphatase (AP) -conjugated or HRP-conjugated goat
anti-mouse IgA, IgM, or IgG Abs (Southern Biotechnology Associates) at
1/500 dilution in 0.1%BSA/PBS were added to total isotype- and CT-
specific wells, respectively, and the plates were incubated 2 h at room
temperature. Ag-specific HRP-labeled Abs bound to the plate were visu-
alized using ortho-paraphenylenediamine (1 mg/ml)-0.04% H2O2substrate
in citrate buffer, and the reaction was read at 450 nm using a Titertek
Multiscan spectrophotometer (Flow Laboratories, Irving, U.K.). AP-la-
beled Abs were visualized using phosphatase-substrate tablets nitrophenyl-
substrate (Sigma) at 1 mg/ml in ethanolamine buffer, pH 9.6, added to the
wells, and the reaction was read at 405 nm in the Titertek Multiscan spec-
trophotometer (Flow Laboratories). Antitoxin titers were defined as the
interpolated OD reading giving rise to an absorbance of 0.4 above back-
ground, which consistently gave OD readings on the linear part of the
curve. Titers were given in log10means ? SD of five mice per group. Total
IgA, IgM, or IgG in ?g/ml was calculated from standard curves generated
by serial dilutions of purified hybridoma IgA, IgM, or IgG (PharMingen,
San Diego, CA) of known concentrations. The isotype-specific antisera
were highly specific and did not cross-react with purified proteins of other
Individual cells secreting Igs, spot-forming cells (SFC), were determined
by the ELISPOT technique essentially as described (24). Briefly, polysty-
rene petri dishes (Nunclon; Nunc) were coated with 3 nmol/ml of GM1
ganglioside (Sigma) in PBS at 4°C overnight, followed by an additional
incubation at 4°C with 3 ?g/ml CT (List Biological Laboratories) in PBS
overnight. For isotype-specific SFC, 96-well flat-bottom microtiter plates
(Nunc) were coated with 5 ?g/ml goat anti-mouse IgA, IgG, or IgM Abs
(Southern Biotechnology Associates) in PBS at 4°C overnight. After wash-
ing with PBS, the petri dishes or the microtiter plates were blocked by
0.1% BSA in PBS for 30 min at 37°C. Lymphocytes at 4 ? 105in 400 ?l
medium were added to each petri dish or 105lymphocytes in 150 ?l me-
dium were added to the top row in 96-well plates and then serially diluted
1:3 in corresponding subwells and sedimented by centrifugation for 3 min
at 400 rpm. All dilutions were performed using Iscove’s medium (Se-
romed, Berlin, Germany) supplemented with 10% FCS (Seromed), 5 ?
10?5M 2-ME (Sigma), 1 mM L-glutamine, and 50 ?g/ml gentamicin. Cells
were incubated for 4 h in 37°C in 8% CO2. After incubation, each petri dish
or well was washed out by repeated rinsing in PBS containing 0.05%
Tween 20. Single LP anti-CT SFC were visualized by adding goat anti-
mouse IgA primary Ab at 1/300 dilution (Cappel, Durham, NC) followed
by HRP-conjugated rabbit anti-goat Ig-secondary Abs (Dako, Carpinteria,
CA) at 1/200. For the isotype-specific SFCs, AP-conjugated Abs (Southern
Biotechnology Associates) at 1/500 dilution were used. For the petri
dishes, (HRP substrate) paraphenylenediamine substrate (Sigma) in 1%
agar-in-PBS solution was used, and for the 96-well plates (AP substrate)
we used 5-bromo-4-chloro-3-indolylphosphate (Sigma) in agarose-in-wa-
ter as described (24). The SFCs were counted under low magnification.
Isolated PP or LP lymphocytes were pooled from two mice, and each pair
was analyzed in duplicate petri dishes or triplicate wells plus one uncoated
well as a control for unspecific binding. Values were expressed as mean
SFC/107lymphocytes ? SD of at least three pairs per group.
Ligated loop test
For evaluation of protection against CT-induced diarrhea/fluid loss, the
method described by Lange and Holmgren was used (25). The abdomen
was opened under light ether anesthesia, and a 6- to 8-cm loop was ligated
in the middle part of the small intestine. CT (List Biological Laboratories),
2,5 ?g in 0.2 ml of PBS, was injected into the loop, and the abdomen was
closed. After 4 h, the mice were sacrificed, whereafter the loop with its
fluid content was weighed and its length was determined. Unimmunized
mice of all strains were equally sensitive to CT injected in ligated small
intestinal loops. Values for fluid accumulation in the ligated loops, reflect-
ing the degree of immune protection, were expressed as the weight-per-
length ratio in mg/cm ? SD of five to seven mice per group.
Frozen sections (5 ?m) from unimmunized or CT-immunized mice were
prepared on microslides using a cryostat (model 1720; Leitz, Wetzlar, Ger-
many) and frozen at ?70°C. The slides were fixed in 50% acetone for 30 s
followed by 100% acetone for 5 min at 4°C. After washings in PBS, the
slides were treated with 5% horse serum in PBS for 15 min in a humid
chamber. For detection of germinal center (GC) reactions and IgA?and
914 J CHAIN-DEFICIENT MICE FAIL TO DEVELOP ANTITOXIC PROTECTION
by guest on June 2, 2013
IgM?cells in the gut-associated lymphoid tissue, fixed sections were dou-
ble labeled with FITC-conjugated peanut (Arachis hypogaea) hemaggluti-
nin (PNA; Sigma) and Texas Red-conjugated anti-mouse IgA or IgM
(Southern Biotechnology Associates), both at 1/100 dilution (26). For de-
tection of IgA plasma cells in LP, the sections were labeled with FITC-
conjugated anti-IgA (Southern Biotechnology Associates). The slides were
mounted with fluorescence mounting media (Dako), and sections were
evaluated and photographed using DAS Mikroskop, (Leica Microscope
System, Welzar, Germany).
We used Student’s t test for independent samples for analysis of
Distribution of total IgA and IgM in serum and intestinal lavage
in J chain-deficient mice
Conflicting information regarding the relative distribution of total
IgA and IgM in serum and gut mucosal secretions in J chain-
deficient mice have recently been published (6, 7, 19). However,
previous studies assessed mucosal IgA levels in fecal extracts or
by applying wicks tampons directly to the mucosal surface of the
small intestine (6, 7, 27), but a more reliable method is to collect
intestinal lavage fluid from the entire small intestine using a buffer
containing strong protease inhibitors (13, 14, 20, 21). Thus, we
collected serum and gut lavage from WT mice or mice with a
homozygous or heterozygous deletion of the J chain gene (19). We
found that both J chain?/?and J chain?/?mice had significantly
reduced levels of total IgA and IgM in gut lavage compared with
WT mice (Fig. 1). In contrast, and in agreement with previous
reports (6, 7), serum IgA levels were increased (Fig. 1) while se-
rum IgM levels were dramatically reduced in J chain?/?and J
chain?/?mice compared with WT mice (19). The mean IgM val-
ues of three experiments were 34 ? 16, 113 ? 74, and 236 ? 98
?g/ml, respectively. Intestinal total IgG levels were similar in J
chain?/?, J chain?/?, and WT mice and roughly 300-fold lower
than serum IgG levels in these mice (not shown). At the single-cell
level, IgM-producing cells (SFC) were significantly fewer in both
the PP and the LP of J chain?/?than in WT mice (Table I). In
contrast, the level of total IgA SFC were similar in PP and LP of
J chain?/?and WT mice, indicating that the ability to develop IgA
responses was unperturbed in J chain-deficient mice. This conclu-
sion was supported further by immunohistochemical analysis of
IgA B cell differentiation and maturation at inductive sites, i.e., PP,
as well as at the LP-effector site in the intestinal mucosa (Fig. 2).
We found no reduction in the number of IgA?cells in the LP and
PP and no alteration in the ability to develop GC in the PP of J
chain-deficient mice (Fig. 2). Thus, the induction, differentiation,
and maturation of IgA B cells into plasma cells occurred normally
in the absence of J chain, but there was a significant reduction in
luminal concentrations of IgA and IgM.
Oral immunizations with cholera toxin in J chain deficient mice
Next we addressed whether oral immunization of J chain-deficient
mice resulted in the development of specific systemic and mucosal
antitoxin Ab responses. Following three oral immunizations, we
detected 10-fold increased IgA antitoxin titers in the serum of J
chain?/?mice, while IgM antitoxin Ab levels were significantly
reduced compared with WT mice (Fig. 3). Mice with a homozy-
gous deficiency exhibited higher IgA and lower IgM responses,
compared with heterozygous and WT mice, suggesting a gene-
dosage effect. The serum anti-CT IgG responses were equivalent in
all mouse strains (Fig. 4). By contrast, the presence of antitoxin
IgA in gut lavage was reduced in a gene-dependent fashion with
only poor IgA antitoxin levels in J chain?/?, intermediate in J
chain?/?, and strong anti-CT IgA levels in WT mice (Fig. 4). Very
low, but similar, titers of IgG and no IgM anti-CT Abs were de-
tected in gut lavage in J chain?/?, J chain?/?, or WT mice (not
shown). Further, detection of anti-CT SFC responses in the LP of
immunized mice demonstrated that the development of specific
Ab-producing cells in the LP of J chain-deficient mice was unal-
tered compared with that in WT mice (Table II). Therefore, the J
chain deficiency was not associated with a defect in the induction
of antitoxic IgA immunity, but was clearly hampering the secretion
of specific IgA into the intestinal lumen.
Lack of antitoxic protection following oral immunization of J
Given that host resistance against CT-induced diarrhea requires the
neutralization of toxin with specific sIgA at the mucosal surface
(13–15, 20, 28), we investigated the consequence of an impaired
pIgR-mediated transport for the ability to develop gut antitoxic
protection following oral immunization with CT (13, 28). We
found that, whereas WT mice demonstrated strong protection
against CT challenge, the J chain-deficient mice were poorly pro-
tected (Fig. 5). Compared with unimmunized control animals, WT
mice exhibited 80% protection while J chain?/?mice were only
protected to 20%, indicating that the lack of IgA transport in J
chain-deficient mice dramatically reduced the ability to neutralize
luminal CT. Heterozygous J chain?/?mice showed intermediate
protection, consistent with a gene-dosage effect in the ability to
J chain-deficient mice. Serum and intestinal lavages from naive WT (gray
bars), J chain?/?(black bars), and J chain?/?(open bars) mice were an-
alyzed for content of total IgA and IgM Abs using a highly specific ELISA.
Five animals were included in each group and the values are expressed in
mg/ml of IgA and ?g/ml of IgM and represent the mean ? SD. The total
IgA or IgM concentrations were calculated from a standard curve gener-
ated with purified mouse IgA or IgM. Values for J chain?/?and J chain?/?
mice were significantly different from values obtained with WT mice (p ?
0.05). This is one representative experiment of three.
Total IgA and IgM content in serum and intestinal lavage in
Table I. Total number of isotype-specific SFCs in freshly isolated
lymphocytes from PP or LP of naive J chain-deficient or WT mice
73 ? 18
29 ? 10
33 ? 9
82 ? 17
31 ? 746
6 ? 3
2996 ? 902a
84 ? 21
28 ? 2
2827 ? 488
92 ? 18
5 ? 3
aValues for single Ab-producing cells using the ELISPOT technique are given as
mean numbers of SFC/106isolated lymphocytes ? SD of three pairs of J chain-
deficient (J chain?/?) or WT mice. Values for IgM responses in J chain-deficient and
WT mice are statistically different (p ? 0.05). This is one representative experiment
of two giving similar results.
915The Journal of Immunology
by guest on June 2, 2013
develop intestinal antitoxic protection. These statistically signifi-
cant results (p ? 0.05) argue for a strong dependence on J chain-
pIgR transport of sIgA across the intestinal epithelium for local
protective immunity against CT.
Here, we have demonstrated that functional mucosal immunity is
highly dependent on an active transport of specific IgA Abs to the
mucosal surfaces. The present investigation unequivocally sup-
ports the notion that the J chain is involved in this process and that
in its absence strong Ag-specific IgA immunity in the LP of the
mucosa cannot mediate protection against a toxin-induced diar-
rheal challenge. Also, we observed a gene-dosage effect of the J
chain gene for intestinal antitoxic protection and IgA immunity in
the heterozygous and WT mice, strongly supporting the idea that
the J chain is directly linked to the mucosal secretion of Igs.
Furthermore, the presence of gut lavage IgG Abs, the isotype
that does not bind J chain and, therefore, cannot associate with the
demonstrating the presence of IgA?cells in the LP and GC reactions in the PP from a naive J chain?/?(left) and a WT (right) mouse. Frozen sections
were prepared and labeled with FITC-conjugated anti-IgA (upper panel) demonstrating IgA?cells in the gut LP, FITC-conjugated PNA plus TXRD-
conjugated anti-IgA (middle panels) demonstrating double staining (yellow/orange) showing the colocalization of IgA?B cells to the PNA?GC in J
chain?/?and WT mice in the PP, and FITC-conjugated PNA plus Texas Red-conjugated anti-IgM PNA (bottom panels) demonstrating GC (green) and
membrane IgM?cells in the PP. These slides are representative sections from three separate experiments.
IgA B cell differentiation and maturation in the gut mucosa of J chain-deficient mice. Light level micrographs (magnification, ?20)
916 J CHAIN-DEFICIENT MICE FAIL TO DEVELOP ANTITOXIC PROTECTION
by guest on June 2, 2013
pIgR, did not differ between the three strains, arguing that passive
diffusion is unlikely to account for the translocation of protective
Igs into the lumen. If this had been the case, then the 10-fold
increased serum IgA antitoxin levels detected in J chain?/?mice
following immunization should also have augmented the represen-
tation of gut lavage IgA and protection in mice lacking the J chain
relative to those that have intact J chain production. Whether IgM
antitoxin Abs have a protective function against CT cannot be
answered by the present study, because, consistent with our pre-
vious findings, J chain-deficient mice appeared to have a perturbed
IgM secretion (19). However, in a parallel study using IgA-defi-
cient mice (29), we observed a similarly low level of protection,
around 20%, despite a 10-fold compensatory increase in gut lavage
anti-CT IgM levels compared with WT mice (N.Y.L., unpublished
observation). Thus, the latter observation would argue against a
functional protective role of gut polymeric IgM antitoxin Abs in
host resistance against CT.
The findings reported here are at variance with a previous study
by Hendrickson et al. using a different strain of J chain-deficient
mice (6, 7). Although we targeted exon 1, whereas Hendrickson et
al., targeted exon 2, of the J chain gene, we have no reason to
believe that the targeting strategy can explain the conflicting ob-
servations, because both approaches led to a failure to detect J
chain gene transcripts (6, 19). Rather, we think that differences in
the methods for detection of mucosal Abs and perhaps to some
extent environmental factors or the recently discussed polymor-
phism in I129 mice (30) could have influenced the results. Al-
though the neutralizing ability of monomeric IgA and resistance to
proteolytic degradation is poor compared with that of sIgA (1, 2, 4),
we do not think that the J chain-deficient mice that we have studied
secreted monomeric or dimeric IgA (lacking J chain) into the lumen,
as was suggested by the findings of Hendrickson et al. (7).
First, we detected very little IgA in the gut lavage, despite the
fact that our ELISA was sufficiently sensitive to detect high levels
of the predominantly monomeric IgA (6, 19) in serum of the J
chain?/?mice. It should be noted that J chain-deficient mice also
have substantial levels of dimeric IgA (6, 7, 19), supporting our
observation that dimerization of IgA in the absence of J chain is
not sufficient for luminal transport. Secondly, our lavage method of
collecting intestinal IgA is more representative and sensitive than
the wicks tampon method used previously. Although both methods
IgM concentrations were reduced in perorally immunized J chain deficient
mice. J chain?/?, J chain?/?, or WT mice were given three oral doses with
CT (10 ?g/dose) and sacrificed 8 days after the final immunization. Serum
Ab levels were determined by ELISA. Values represent log10titers ? SD
of five mice per group with serum anti-CT IgA (open bars) and serum
anti-CT IgM (filled bars). The higher concentration of anti-CT IgA in se-
rum of J chain?/?mice was statistically significant compared with het-
erozygous J chain?/?and WT mice (p ? 0.05).
Greatly augmented serum anti-CT IgA levels while anti-CT
lumen of perorally immunized J chain-deficient mice. J chain?/?, heterozy-
gous J chain?/?, or WT mice were given three oral doses with CT (10
?g/dose) and sacrificed 8 days after the final immunization. Serum anti-CT
IgG and gut lavage anti-CT IgA Ab levels were determined by ELISA.
Values represent log10titers ? SD of five mice per group with serum
anti-CT IgG (filled bars) and gut lavage anti-CT IgA (open bars). The
lower concentration of anti-CT IgA in gut lavage of J chain?/?mice was
statistically significant compared with heterozygous J chain?/?and
WT mice (p ? 0.05). The anti-CT IgA values for J chain?/?and heterozy-
gous J chain?/?mice were also significantly different from each other
(p ? 0.05).
Impaired transport of specific IgA antitoxin into the gut
Table II. Unaltered LP antitoxin IgA responses to oral immunization
with CT in J chain-deficient micea
Mice Antitoxin IgA SFC/107LP Cells
6050 ? 302
5916 ? 589
aMice were given three oral doses of CT (10 ?g/dose) and sacrificed 6 days after
the final immunization. Specific anti-CT-producing cells (SFC) were assayed by the
ELISPOT method, and values are given as means ? SD. Three pairs of mice were
analyzed in each group. This is one representative experiment of three.
immunization with CT in J chain-deficient mice. We determined functional
antitoxic protective immunity in gut-ligated loops of perorally CT-immu-
nized J chain?/?, J chain?/?, or WT mice. Three oral immunizations with
10 ?g/dose of CT were given, and 7 days following the final dose mice
were analyzed for antitoxic protection against CT-induced diarrhea by the
ligated loop test. Fluid accumulation in response to challenge with CT, 2.5
?g/loop, was determined and compared with that observed in CT-chal-
lenged loops of unimmunized control mice. Values in mg/cm ? SD of five
to seven mice in each group with immunized mice (open bars) and unim-
munized control mice (filled bars). The results of J chain?/?, J chain?/?,
and WT mice are all statistically significantly different from each other
(p ? 0.05) The data is representative of three identical experiments giving
Impaired antitoxic protection in gut mucosa following oral
917The Journal of Immunology
by guest on June 2, 2013
exclude feces, the wicks detect Abs in only a limited region, while
the lavage method collects IgA from the entire length of the small
intestine (20, 21, 27). Moreover, it is difficult to rule out the pos-
sibility that the tampons establish a fluid diffusion gradient that
may affect the barrier functions of the intestinal mucosa, thereby
allowing the diffusion of IgA from the LP and serum into the
tampons. Therefore, we believe that the lavage method is more
accurate and provides highly reproducible measurements of lumi-
nal IgA. Although Hendrikson et al. used the lavage method for
detection of IgA in bronchoalveolar fluid of J chain?/?mice (7),
we think that the increased levels of IgA that they reported reflect
the transudation of tissue and serum IgA into the alveolar lumen,
similar to that seen in the genital tract (1, 4). Third, CT causes fluid
loss after binding to the GM1 receptor on the epithelial cell (11,
12). If normal levels of surface dimeric anti-CT IgA had been
present in immunized J chain?/?mice (7), we would have ex-
pected these Abs to confer a higher degree of protection, because
the ability to neutralize toxin should not be a function of the pres-
ence of J chain or secretory component (1, 4, 13–15, 28).
Our study clearly demonstrates a strong association between the
presence of specific IgA Abs in the gut lumen and protection
against CT-induced fluid loss. This finding agrees well with pre-
vious studies using other models to show a direct relationship be-
tween gut mucosal transport of IgA and protection, as, e.g., the
back-pack IgA-hybridoma model developed by Neutra and co-
workers (15, 28) However, evidence has been presented to suggest
that antitoxic protection may not be mediated solely by antitoxin
Abs and may involve additional factors such as IFN-?, or other
cytokines, neuropeptides, or ASF, which were recently cloned and
purified (16–18). However, the present results are difficult to rec-
oncile with any major antitoxic protective function of these alter-
native factors in vivo, a conclusion supported further by our dem-
onstration that antitoxic protection develops normally in IFN-?
receptor-deficient mice (31) and that also IgA-deficient and B cell-
deficient mice fail to develop significant antitoxic protection
(N.Y.L., unpublished observation). Of course, our results with the
J chain-deficient mice do not exclude a protective function of ASF
or other factors against CT-induced diarrhea, although we do not
know why the absence of the J chain in our model would have
affected the level or function of, e.g., ASF (16). We did observe
20% protection in immunized J chain?/?mice. Whether this level
of protection was conferred by the presence of low levels of lu-
minal anti-CT IgA, lacking J chain, or these alternative factors is
not known. However, orally immunized B cell-deficient mice ex-
hibit no protection at all against CT challenge (N.Y.L., unpub-
lished observation). Also, we find it highly unlikely that there
would be any connection between the neuropeptides or ASF and
the J chain, or for that matter a dependence on epithelial transcy-
tosis of local dimeric IgA Abs and these alternative protective
factors. Further analyses are required to delineate the role of these
factors in intestinal antitoxic protection.
Consistent with our previous report on the J chain-deficient
mouse, we observed low levels of IgM Abs (19). Not only did we
find decreased levels of IgM in gut lavage, we also had low serum
IgM concentrations and 5-fold reduced total IgM SFC in the gut
LP as compared with WT mice. Therefore, the low gut IgM levels
in the J chain?/?mice seem to be an effect of poor plasma cell
differentiation, rather than an effect of impaired pIgR transport.
Thus, this maturational defect appears to affect IgM B cells selec-
tively, as both IgG and IgA production, including gut anti-CT IgA
SFC, were normal. This is an intriguing finding, demonstrating that
the production of J chain is not required for terminal maturation of
IgA plasma cells in the gut. Because almost all IgA- or IgG-pro-
ducing cells in human gut LP express J chain and only few cells of
these isotypes produce J chains in peripheral lymph nodes,
Brandtzaeg and coworkers have speculated that coexpression of J
chain would constitute a maturational requirement for mucosal B
cell development (1). However, the data from J chain-deficient
mice does not appear to support such a notion (7, 19). At present,
we do not have an explanation for why coexpression of J chain is
selectively more important for IgM plasma cell differentiation. Po-
tentially, the J chain is critical in the assembly and secretion of
polymeric IgM (32). Alternatively, it plays no role in terminal
plasma cell maturation; rather, it may be that, in the absence of J
chain, the secreted form of IgM is highly unstable and is rapidly
degraded in serum and tissue. However, the work of many inves-
tigators has indicated that the J chain is not critical for production
of polymeric IgM, including hexameric IgM, which is the most
complement-activating form of IgM (1, 33). In our previous study,
we found lower concentrations of both hexameric and monomeric
forms of IgM in serum of J chain?/?mice, and serum from these
mice was a significantly poorer activator of complement compared
with serum from WT mice, indicating that there are also qualitative
differences (19). Further studies will be undertaken in this model to
address the role of the J chain for the maturation of IgM plasma
1. Brandtzaeg, P., and I. N. Farstad,. 1999. The human mucosal B-cell system. In
Mucosal Immunology, P. L. Ogra, W. Strober, J. Mestecky, J. R. McGhee,
M. E. Lamm, and J. Bienenstock, eds. Academic Press, San Diego, CA, p. 439.
2. Mostov, K., and C. S. Kaetzel, 1999. Immunoglobulin transport and the poly-
meric immunoglobulin receptor. In Mucosal Immunology, P. L. Ogra, W. Strober,
J. Mestecky, J. R. McGhee, M. E. Lamm, and J. Bienenstock, eds. Academic
Press, San Diego, CA, p. 319.
3. Cebra J. J., H.-Q. Jiang, J. Sterzl and H. Tlaskalova-Hogenova. 1999. The role of
the mucosal microbiota in the development and maintenance of the mucosal
immune system. In Mucosal Immunology, P. L. Ogra, W. Strober, J. Mestecky,
J. R. McGhee, M. E. Lamm, and J. Bienenstock, eds. Academic Press, San Diego,
CA, p. 267.
4. Russell, M. W., M. Kilian, and M. E. Lamm. 1999. Biological activities of IgA.
In Mucosal Immunology, P. L. Ogra, W. Strober, J. Mestecky, J. R. McGhee,
M. E. Lamm, and J. Bienenstock, eds. Academic Press, San Diego, CA, p. 225.
5. Brandtzaeg, P., and H. Prydz. 1984. Direct evidence for an integrated function of
J chain and secretory component in epithelial transport of immunoglobulins. Na-
6. Hendrickson B. A., D. A. Conner, D. J. Ladd, D. Kendall, J. E. Casanova,
B. Corthesy, E. E. Max, M. R. Neutra, C. E. Seidman, and J. G. Seidman. 1995.
Altered hepatic transport of immunoglobulin A in mice lacking the J chain.
J. Exp. Med. 182:1905.
7. Hendrickson B. A., L. Rindisbacher, B. Corthesy, D. Kendall, D. A. Waltz,
M. R. Neutra, and J. G. Seidman. 1996. Lack of association of secretory com-
ponent with IgA in J chain-deficient mice. J. Immunol. 157:750.
8. Vaerman, J. P., A. E. Langendries, D. A. Giffoy, C. S. Kaetzel, C. M. Fiani,
I. Moro, P. Brandtzaeg, and K. Kobayashi. 1988. Antibody against the human J
chain inhibits polymeric Ig-receptor-mediated biliary and epithelial transport of
human polymeric IgA. Eur. J. Immunol. 28:171.
9. Tomasi, T.B., and Czerwinski, D.S. 1976. Naturally occurring polymers of IgA
lacking J chain. Scand. J. Immunol. 5:647.
10. Jerry, L. M., H. G. Kunkel, and L. Adams, L. 1972. Stabilization of dissociable
IgA2 proteins by secretory component. J. Immunol. 109:275.
11. Holmgren, J., and A.-M. Svennerholm. 1996. Oral vaccines against Cholera and
enterotoxigenic Escherichia coli diarrhea. In Mucosal Vaccines. H. Kiyono,
P. L. Ogra, and J. McGhee, eds. Academic Press, San Diego, p. 241.
12. Svennerholm, A.-M., M. Jertborn, L. Gothefors, A. M. Karim, D. A. Sack, and
J. Holmgren. 1984. Mucosal antitoxic and antibacterial immunity after cholera
disease and after immunization with a combined B subunit-whole cell vaccine.
J. Infect. Dis. 149:884.
13. Lycke, N., L. Eriksen, and J. Holmgren 1987. Protection against cholera toxin
after oral immunization is thymus dependent and associated with intestinal pro-
duction of neutralizing IgA antitoxin. Scand. J. Immunol. 25:413.
14. Lycke, N., A. K. Bromander, and J. Holmgren. 1989. Role of local IgA antitoxin-
producing cells for intestinal protection against cholera toxin challenge. Int. Arch.
Allergy Appl. Immunol. 88:273.
15. Apter, F. M., W. I. Lancer, R. A. Finkelstein, J. J. Mekalanos, and M. R. Neutra.
1993. Monoclonal immunoglobulin A antibodies directed against cholera toxin
block binding to intestinal epithelial cells in vitro. Infect. Immun. 61:5271.
16. Johansson, E., I. Lo ¨nnroth, S. Lange, I. Jonson, E. Jennische, and C. Lo ¨nnroth.
1995. Molecular cloning and expression of a pituitary gland protein modulating
intestinal fluid secretion. J. Biol. Chem. 270:20615.
918 J CHAIN-DEFICIENT MICE FAIL TO DEVELOP ANTITOXIC PROTECTION
by guest on June 2, 2013
17. Eklund, S., A. Sjo ¨kvist, J. Fahrenkrug, M. Jodal, and O. Lundgren. 1988. Soma-
tostatin and methionine-enkephalin inhibit cholera toxin-induced jejunal net fluid
secretion and release of vasoactive intestinal polypeptide in the cat in vivo. Acta
Physiol. Scand. 135:551.
18. Holmgren, J., J. Fryklund, and H. Larsson. 1989. ?-Interferon-mediated down-
regulation of electrolyte secretion by intestinal epithelial cells: a local immune
mechanism? Scand. J. Immunol. 30:499.
19. Erlandsson L., K. Andersson, M. Sigvardsson, N. Lycke, and T. Leanderson.
1998. Mice with an inactivated joining chain locus have perturbed IgM secretion.
Eur. J. Immunol. 28:2355.
20. Ho ¨rnqvist, E., T. J. Goldschmidt, R. Holmdahl, and N. Y. Lycke. 1991. Host
defense against cholera toxin is strongly CD4?T cell dependent. Infect. Immun.
21. Elson, C. O., W. Ealding, and J. Lefkowitz. 1984. A lavage technique allowing
repeated measurment of IgA antibody in mouse intestinal secretions. J. Immunol.
22. Lycke, N., and L. Leofrancois. 1995. Isolation of mouse small intestinal intra-
epithelial lymphocytes, Peyer’s patch, and lamina propria cells, In Current Pro-
tocols in Immunology. J. Coligan, ed. Current Protocols, New York, p. 3.19.6.
23. Ho ¨rnquist E., and N. Lycke, 1993. Cholera toxin adjuvant greatly promotes an-
tigen priming by T cells. Eur. J. Immunol. 23:2136.
24. Lycke, N. 1991. Measurement of polyclonal immunoglobulin synthesis using the
ELISPOT assay, In Current Protocols in Immunology. J. Coligan, ed. Current
Protocols, New York, p. 7.14.1 .
25. Lange S., and J. Holmgren. Protective antitoxic cholera immunity in mice: in-
fluence of route and number of immunizations and mode of action of protective
antibodies. Acta Pathol. Microbiol. Scand. 86:145.
26. Vajdy, M., M. H. Kosco-Vilbois, M. Kopf, G. Ko ¨hler, and N. Y. Lycke. 1995.
Impaired mucosal immune responses in interleukin 4-targeted mice. J. Exp. Med.
27. Haneberg, B., D Kendall, H. M. Amerongen, F. M. Apter, J.-P. Kraenhenbuhl,
and M. R. Neutra. 1994. Induction of specific immunoglobulin A in small intes-
tine, colon-rectum, and vagina measured by a new method for collection of se-
cretions from local mucosal surfaces. Infect. Immun. 62:15.
28. Apter, F. M., P. Michetti, L. S. Winner III, J. A. Mack, J. J. Mekalanos, and
M. Neutra. 1993. Analysis of the roles of antilipopolysaccharide and anti-cholera
toxin immunoglobulin A (IgA) antibodies in protection against Vibrio cholerae
and cholera toxin by use of monoclonal IgA antibodies in vivo. Infect. Immun.
29. Harriman, G. R, M. Bogue, P. Rogers, M. Finegold, S. Pacheco, A. Bradley,
Y. Zhang, and I. N. Mbawuike. 1999. Targeted deletion of the IgA constant
region in mice leads to IgA deficiency with alterations in expression of other
immunoglobulin isotypes. J. Immunol. 162:2521.
30. Simpson, E. M., C. C. Linder, E. E. Sargent, M. T. Davisson, L. E. Mobraaten,
and J. J. Sharp. 1997. Genetic variation among 129 substrains and its importance
for targeted mutagenesis in mice. Nat. Genet. 16:19.
31. Kjerrulf, M., D. Grdic, L. Ekman, K. Scho ¨n, M. Vajdy, and N Lycke. 1997.
Interferon-? deficient mice exhibit impaired gut mucosal immune responses but
intact oral tolerance. Immunology. 92:60.
32. Carelli, S., A. Ceriotti, and R. Sitia, 1997. Stringent thiol-mediated retention in B
lymphocytes and Xenopus oocytes correlates with inefficient IgM polymerization.
Eur. J. Immunol. 27:1283.
33. Wiersma, E. J., C. Collins, S. Fazel, and M. J. Shulman. 1998. Structural and
functional analysis of J chain-deficient IgM. J. Immunol. 160:5979.
919 The Journal of Immunology
by guest on June 2, 2013