| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory for Immunohistochemistry and Immunopathology, Institute of Pathology, University of Oslo, Rikshospitalet, Oslo, Norway
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
SIgA is generated via the active transport of pIgA to the intestinal lumen by SC/pIgR-expressing epithelial cells; this common transport model for pIgA and pentameric IgM was first proposed by Brandtzaeg (6, 7) in 1974. In adult humans, more SIgA is generated per day than the daily production of IgG (8). Furthermore, SIgA production by lactating mammary glands ensures that SIgA is the dominant Ab in human breast milk (9). Such SIgA is nature’s way of passive immunization of infants against Ags in the mother’s environment, and numerous studies have documented a protective effect of exclusive breastfeeding against diarrhea (9). The proven efficacy of SIgA in preventing and fighting mucosal infections makes it an attractive candidate for therapeutic purposes (10, 11), and recombinant production in various expression systems has been described (12, 13, 14, 15). However, its complex nature with four different polypeptides produced by two distinct cell types imposes restrictions on potential expression systems for efficient production of recombinant SIgA.
Mucosal IgA-producing plasma cells typically also express the J chain
(16). This small polypeptide of 15 kDa promotes formation
of pIgA and pentameric IgM and is exclusively incorporated into these
two types of Ig polymers (17, 18, 19). The human J chain gene
is encoded by four exons (20). Exon 1 encodes the leader
peptide, while exons 2–4 encode the mature protein of 137 amino acids
(14, 20). The human J chain contains eight cysteine
residues; two (Cys15 and
Cys69) are involved in disulfide bridges with the
- or µ-chains, and six are involved in intrachain disulfide
bridges (Cys13:Cys101,
Cys72:Cys92,
Cys109:Cys134) (21, 22). The J chain is well conserved throughout phylogeny, and a
homologous peptide has been reported even in the earthworm (19, 23).
The three-dimensional structure of neither the J chain nor pIgA has
been reported, but experiments to determine the positioning of J chain
in pIgA have suggested that it bridges two -chains of opposing IgA
monomers which are dimerized tail-to-tail (21, 24, 25). It
is often assumed that pIgA and pentameric IgM contain only one J chain
molecule (26), but immunochemical studies have suggested
that dimeric IgA contains two and pentameric IgM three to four J chains
(27, 28). Characterization of J chain released from
dimeric, trimeric, and tetrameric IgA purified from the same myeloma
serum did indeed suggest that the molar J chain ratio increases with
the size of the polymer (29).
The J chain is a key protein in the generation of SIgA because it
promotes polymerization of IgA and because its presence in these
polymers is believed to be required for their affinity to SC/pIgR
(reviewed in Ref. 19). To better understand these two
functions of the J chain, we have used a mutational approach to
generate deletion mutants and mutants that abolish J chain intrachain
disulfide bonds or disulfide bonds between the J chain and the
-chain of the IgA. In this report, we demonstrate that the C
terminus of the J chain is dispensable for pIgA formation but
nevertheless is required for the affinity of such polymers to SC. We
also show that the same holds true for two of the three intrachain
disulfide bridges. Finally, we document that either one of the two
disulfide bonds normally present between the J chain and the -chain
is sufficient for polymer formation. Although these polymers are mostly
noncovalently stabilized, their affinity to SC is nearly the same as
that of the recombinant wild-type pIgA, but they appear to be
inferior ligands in pIgR-mediated transport compared with covalently
stabilized wild-type J chain-containing pIgA.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The plasmid for episomal expression of wild-type J chain, pCEP-wtJ chain, was based on pMEP4 (Invitrogen, De Schelp, The Netherlands) and pCH-J chain (14). A BglII-BamHI fragment from pCH-J chain that contained the CMV promoter and the human J chain open reading frame was subcloned into BglII- and BamHI-digested pMEP4. C-terminal deletion mutants were constructed by PCR with J chain-forward (14) and different downstream primers that introduced a stop codon followed by a BamHI recognition site into the J chain sequence with pCH-J chain as template. The resulting PCR products were digested with HindIII and BamHI and subcloned into HindIII- and BamHI-digested pCEP-wtJ chain. Cysteine-to-serine mutations (except C134S) were made by PCR splice overlap extension with J chain-forward and J chain-reverse (14) as outer primers and different inner primers. PCR products were subcloned as above. C134S was made by PCR with J chain-forward and a reverse primer that mutated the last cysteine codon to serine and introduced a BamHI recognition site immediately after the stop codon, and subcloned as above. For stable expression of C15S and C69S, the inserts were subcloned into pCH. Vectors without inserts were used as negative control (mock) in all transfections. Details of all plasmids will be provided upon request. The complete open reading frame of all expression plasmids was verified by sequencing (Medigenomics, Martinried, Germany).
Cell culture and transfections
The 5-iodo-4-hydroxy-2-nitrophenylacetyl (NIP)-specific
IgA-producing Chinese hamster ovary (CHO) cell line 4.2 that has
been described previously (14) was cultured in Ham’s-F12
supplemented with 10% FCS, 2 mM glutamine, and 50 µg/ml gentamicin.
For semistable transfections (episomal expression), cells were seeded
at 10–15% confluence in six-well trays on day 1. On day 2, cells were
transfected with 1.4 µg of DNA and 3 µl of FUGENE (Roche
Diagnostics, Indianapolis, IN) according to the manufacturer’s
protocol. On day 3 the cells received fresh medium, and on day 4 they
were trypsinized and seeded into 10-cm plates in medium supplemented
with 300 µg/ml Hygromycin B (Roche Diagnostics). Cells received fresh
medium with Hygromycin B every 3–4 days, and after 10–12 days each
10-cm plate contained >500 Hygromycin B-resistant colonies. The
colonies were pooled and seeded into six-well trays for metabolic
labeling and IgA production. Stable transfectants were made essentially
the same way, except that on day 4 they were seeded at different
dilutions into 10-cm plates. On day 12, single colonies were isolated
by use of cloning cylinders and expanded for analysis of J chain
expression.
Transcytosis assays
Transcytosis by human pIgR-expressing Madin-Darby canine kidney (MDCK) cells was performed with 8 µg/ml IgA and an equal amount of IgG in the basal chamber overnight as described previously (30).
Metabolic labeling and SDS-PAGE
Approximately 80%-confluent six-well trays of semistable transfectants were starved in DMEM without cysteine, methionine, and FCS for 30 min. Radiolabeled Cys/Met was added and the cells were incubated for 6 h. Supernatants were immunoprecipitated with rabbit antiserum to human IgA (diluted 1/100; DAKO, Glostrup, Denmark) and magnetic beads coated with goat anti-rabbit Ig (Dynal Biotech, Oslo, Norway). The precipitate was washed in PBS with 1% Nonidet P-40 (Sigma-Aldrich, St. Louis, MO) and eluted in 1x sample buffer (1% SDS, 30% glycerol, 0.02 M phosphate buffer, bromophenol blue). Samples were heated to 95°C for 3 min and resolved by SDS-PAGE on a 4% acrylamide agarose gel as previously described (31). The gels were dried under vacuum, and radiolabeled IgA was visualized by exposure to x-ray film.
ELISA
ELISA specific for human IgG was performed by coating microtiter
plates with goat anti-human IgG (
-chain-specific) (diluted
1/1,000; Sigma-Aldrich). After sample incubation, IgG was detected with
alkaline phosphatase-conjugated goat anti-IgG. For all Ag-specific
ELISAs, microtiter plates were coated with 3 µg/ml NIP-BSA and
blocked with 1% (w/v) BSA in PBS for the J chain ELISA and the
SC-affinity ELISA. For IgA quantification, the primary Ab was mAb
against human IgA (diluted 1/30,000; a gift from T. Lea, Institute of
Immunology, Rikshospitalet, Oslo, Norway) and the secondary Ab
was alkaline phosphatase-conjugated rabbit anti-mouse Ig Ab (DAKO;
diluted 1/1,000). For J chain-specific ELISA, a dilution series of CHO
cell supernatants or purified IgA was incubated overnight. The
microtiter plates were washed and fixed with 2% glutaraldehyde for 30
min then incubated with 6 M urea (pH 3.0) for an additional 30 min.
After washing, the plates were incubated with rabbit antiserum against
human J chain (diluted 1/300) (32) and alkaline
phosphatase-conjugated goat anti-rabbit Ig (DAKO; 1/2,000). For SC
affinity measurement, a dilution series of IgA was incubated overnight.
The microtiter plates were washed and incubated with 8 µg/ml free SC
(from colostrum), then rabbit antiserum against human SC (DAKO; diluted
1/1,000), and finally goat anti-rabbit Ig as above. For the J chain
and the SC-affinity ELISA, the values were expressed as the linear
regression of OD405 vs IgA concentration. Pilot
experiments demonstrated that this method of calculation yielded the
same values for a given IgA sample over at least a four-fold different
initial concentration; thus minor errors in determining the IgA
concentrations of the different preparations were not carried over to
the J chain and SC affinity measurements. However, the relative values
of J chain content and SC affinity measured this way could be compared
only within a single experiment.
Purification and fractionation of NIP-specific IgA
NIP-specific IgA was purified from CHO cell supernatants on
NIP-Sepharose as described previously (14). For
fractionation by fast protein liquid chromatography (FPLC), the
eluted IgA was dialyzed against 20 mM Tris·HCl (pH 7.1) and loaded
onto a Mono Q column connected to an
ÄKTAFPLC system (Amersham Pharmacia
Biotech, Little Chalfont, Buckinghamshire, U.K.). The IgA was
eluted by a linear salt gradient from 2.5 to 500 mM NaCl in 20 mM
Tris·HCl (pH 7.1) and collected in 300-µl fractions which were
analyzed for IgA concentration by ELISA. Peak fractions were analyzed
for molecular size by nonreducing SDS-PAGE followed by immunoblotting,
and for SC affinity by ELISA. Several fractions (as indicated in Fig. 4
) were pooled before analysis by native PAGE and immunoblotting as
well as for transcytosis assays.
|
IgA (30 ng) was resolved by nonreducing SDS-PAGE on a 5% (w/v)
acrylamide gel with the Criterion system (Bio-Rad, Hercules, CA) and
transferred onto polyvinylidene fluoride membrane (Millipore,
Bedford, MA). The membrane was blocked by first air drying and then
incubation in 10% skimmed milk in PBS plus 0.05% Tween 20 (PBST).
Subsequent incubations were all performed in PBST. The blocked membrane
was first incubated with a rabbit antiserum to human IgA (diluted
1/3000; DAKO) for 90 min, washed three times for 5 min, and then
incubated with HRP-conjugated donkey anti-rabbit Ig (diluted
1/3000; Amersham Pharmacia Biotech) for 90 min. After washing three
times for 5 min, the membrane was incubated with SuperSignal (Pierce,
Rockford, IL) for 10 min, and the substrate was detected with a
light-sensitive camera (Chemidoc; Bio-Rad). For IgA immunoblotting, the
membrane was stripped in 0.1% SDS and 100 mM 
-mercaptoethanol for
30 min at 37°C, washed in PBST, and incubated with rabbit antiserum
to human J chain (previously adsorbed against monomeric IgA) diluted
1/600. Secondary Ab and revealing reaction were as above. For native
immunoblots, 60 ng of IgA was resolved on a 5% (w/v) acrylamide gel
(acrylamide:bis ratio of 37.5:1), pH 8.5, without stacking gel. SDS was
omitted from all buffers and the pH of the loading buffer was 8.5. The
gel was transferred to polyvinylidene difluoride membrane and processed
as above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There are no three-dimensional structure data available for the J
chain of any species. Therefore, we used computer-assisted secondary
structure predictions as a first approach to delineate the structure
and hence better understand the function of the J chain. A schematic
depiction of human J chain that combines the known pattern of disulfide
bonding with the predicted secondary structure is shown in Fig. 1. The Jpred program
(http://jura.ebi.ac.uk:8888/submit.html) predicted that the J chain
mostly consists of -strands and extended coils with two short
-helices (Fig. 1). Intrachain disulfide bonds between
Cys13 and Cys101,
Cys72 and Cys92, and
Cys109 and Cys134 further
restrict the possible conformations of the J chain.
|
The J chain performs two essential functions for SIgA formation:
it enhances dimerization (and some formation of larger polymers) of
IgA; and it is required for the affinity of pIgA to SC/pIgR. To
delineate which regions of the J chain are required for these
functions, we first made sequential C-terminal deletions in the J chain
sequence. We found that deleting the J chain to amino acid position
113, 105, or 96 appeared to reduce the level of J chain incorporation
into IgA (Fig. 2A), but this
could not be definitely concluded because the level of incorporation
was measured with a polyclonal antiserum to the J chain. Deletion of
this C-terminal region might therefore also remove epitopes detected by
the antiserum. Deletion from amino acid 76 or more appeared to preclude
incorporation of J chain into the coexpressed IgA (Fig. 2A).
The properties of IgA produced by CHO cells cotransfected with
these J chain mutants were therefore the same as IgA produced by CHO
cells transfected with empty vector (Fig. 2, A–C; mock, 
76 and 73).
|
B;
113). Further deletion of eight or 17 more amino acids still allowed
polymer formation, although at slightly reduced efficiency and
occasionally giving rise to different polymer sizes (Fig. 2B; 105 and 96, respectively). In agreement with our
inability to detect 76- and 73-mutated J chain associated with
IgA, these mutants did not contribute to enhanced polymerization of
IgA. Although the smallest C-terminal deletion in the J chain (25 aa)
did not significantly affect polymerization of coexpressed IgA (Fig. 2A; 113), it was clear that this deletion was sufficient
to abolish all affinity of the produced pIgA to SC (Fig. 2C;
113). Further deletions of the J chain also resulted in pIgA without
affinity to SC (Fig. 2C; 105 and 96). Role of J chain disulfide bridges for IgA polymerization and SC binding
The human J chain forms three intrachain disulfide bridges as well
as two bridges to the tailpieces of two opposing IgA -chains. To
assess the importance of these bonds, we mutated each cysteine
individually as well as Cys15 and
Cys69 in combination (C15+ 69S). We found that
mutating any one of the J chain cysteines reduced the level of J chain
incorporation to between 30 and 70% compared with the level of
wild-type J chain incorporated into coexpressed IgA (Fig. 3A). However, mutation of
Cys15 and Cys69 in
conjunction appeared to abolish incorporation of the mutated J chain
into IgA (Fig. 3A; C15+69S). Mutation of
Cys13, Cys101,
Cys109, or Cys134 did not
appear to affect polymer formation (Fig. 3B). Surprisingly,
all of these mutations abolished the affinity of the J chain-positive
pIgA to SC (Fig. 3C). Mutation of
Cys15, Cys72, and
Cys92 reduced the level of J chain-mediated pIgA
formation, while mutation of Cys69 and the double
mutant (C15 + 69S) did not enhance pIgA formation above the level seen
without coexpressed J chain (Fig. 3B). Mutation of
Cys15, Cys69,
Cys72, or Cys92 still
resulted in J chain-positive IgA with significant affinity for SC (Fig. 3C), while the C15+69S mutant did not affect the property of
the coexpressed IgA.
|
5% transport compared with pIgA containing
wild-type J chain (data not shown). Single CHO cell clones produce different molecular forms of IgA
Molecular size analysis revealed that each pool of IgA and J chain
transfectants produced more than one molecular form of IgA (Figs. 2B and 3B). Furthermore, Western blotting of
C15S- and C69S-containing IgA demonstrated that the J chain could be
found in both monomers and polymers of IgA (data not shown). To
determine whether the discrepancy between the good affinity to SC and
the poor ability to be transcytosed by pIgR-expressing polarized
epithelial cells of C15S- and C69S-containing IgA might be due to a
differential contribution of the various molecular forms in the two
assays, we decided to purify the different molecular forms of IgA
produced by each transfectant. First, we made stable transfectants of
C15S, C69S, and wild-type J chain-expressing, IgA-producing CHO cells
and purified IgA from supernatants of each cell line by affinity
chromatography. To further fractionate each sample of recombinant IgA
we chose Mono Q anion exchange chromatography because gel filtration
would presumably be inefficient at separating monomeric IgA with J
chain from monomers without. For IgA expressed without J chain, Mono Q
fractionation revealed three distinct peaks centered around fractions
36, 48, and 65 (Fig. 4A).
Surprisingly, fraction 48 showed low but significant affinity for SC
(Fig. 4A), while the other two peaks had no affinity for SC.
Nonreducing SDS-PAGE followed by Western blotting revealed that
only fraction 48 contained covalently stabilized dimers (Fig. 4B).
IgA with wild-type J chain eluted with a larger peak centered at
fraction 45 (with a shoulder extending to fraction 54) from the Mono Q
column. Fraction 45 was nearly pure J chain-positive dimeric IgA, while
fraction 53 was a mixture of dimeric J chain-positive IgA and monomeric
IgA (Fig. 4, C and D). Not surprisingly, fraction
45 demonstrated the highest affinity to free SC (Fig. 4C).
IgA with C15S-mutated J chain eluted with an additional peak centered
at fraction 51, and IgA with C69S-mutated J chain eluted with two
additional peaks centered at fractions 46 and 53 (Fig. 4, E
and G). The additional peak from C15S-containing IgA and the
two additional peaks from C69S-containing IgA all demonstrated
significant affinity to SC, i.e., 72–94% of wild-type fraction 45
(Fig. 4, E and G). Nonreducing SDS-PAGE revealed
that these fractions contained a mixture of J chain-positive dimeric
IgA, IgA monomers, and J chain-positive monomeric IgA migrating
slightly above the ordinary monomers (Fig. 4, F and
H).
The Mono Q elution profile of each preparation of IgA prompted us to
investigate whether there might be noncovalently stabilized pIgA
produced by the C15S and C69S transfectants. Therefore, we pooled the
fractions that made up the SC-binding peak (indicated in Fig. 4) and
analyzed each pool by native gel electrophoresis (Fig. 5). Indeed, the pooled fractions from the
C15S and C69S transfectants revealed that the distribution of IgA
dimers and larger polymers in these pools were nearly identical with
the distribution of polymers in the wild-type J chain pool.
|
IgA from supernatants of CHO clones stably expressing C15S- or
C69S-mutated J chain was relatively inefficiently transcytosed by human
pIgR-expressing MDCK cells compared with pIgA from wild-type J
chain-expressing clones (data not shown). Because these experiments
used crude IgA preparations, this result could be due to a less
efficient transcytosis of pIgA with mutated Cys15
or Cys69 in the J chain. Alternatively, these
mutations could result in less efficient assembly of pIgA, but the pIgA
formed could be an equally good substrate for epithelial transcytosis
as pIgA with wild-type J chain. To test these possibilities, we pooled
the fractions from the Mono Q purification of each IgA variant that
contained pIgA and tested each pool in our transcytosis assay. We found
that polymers containing C15S- or C69S-mutated J chain were
transcytosed to a level of 19 and 50%, respectively, of wild-type IgA
polymers (Fig. 6). The level of
transcytosis of each pool correlated well with the level of SC binding
of each peak fraction, although reduction in efficiency compared with
wild-type IgA caused by the mutations was greater for transcytosis than
for SC binding. Finally, we did not find that the fraction of IgA
lacking J chain, which contained some pIgA (Fig. 6, mock-A), was
transcytosed to a level above that of monomeric IgA (Fig. 6, mock-B) or
coincubated IgG (Fig. 6, dotted line), although it demonstrated some
binding to SC in the ELISA (Fig. 4A).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain was sufficient to promote the formation
of noncovalently stabilized IgA polymers. Such polymers retained high
affinity to SC but were poor ligands for pIgR-mediated transcytosis.
Finally, we demonstrated that IgA polymers devoid of J chain showed
some affinity to free SC but were not actively transcytosed by
pIgR-expressing MDCK cells to a level above the paracellular transport
of monomeric IgA and IgG. Role of J chain in IgA polymerization
The J chain is not absolutely required for polymerization of IgA,
but its coexpression significantly enhances the production ratio of
pIgA to monomers (14, 33). The three intrachain disulfide
bridges of the human J chain divide the polypeptide into a large loop
(Cys13 to Cys101) with a
smaller loop within this loop (Cys72 to
Cys92) (Fig. 1). Additionally, the C-terminal 29
amino acids form a small loop bridged by Cys109
to Cys134. We found that the C-terminal loop was
dispensable for polymerization of IgA. Furthermore, none of the
intrachain bridges was absolutely required for pIgA formation. These
data suggested that the three-dimensional structure of the J chain is
not so important for IgA polymerization as long as there are cysteines
in the appropriate places that can be linked to the cysteines of the
-chain tailpieces. Alternatively, the J chain could fold into two
distinct domains, where the N-terminal domain structure is intact in
our mutants and sufficient for polymerization of IgA.
The C15S and the C69S mutants of the J chain are precluded from
covalently linking two IgA monomers (25) but may still
enhance pIgA formation by bringing two monomers together to promote
direct covalent bonding between two opposing -chain tailpieces. In
agreement with this possibility, we found a small increase in covalent
pIgA production when IgA was coexpressed with either of these mutants.
Surprisingly, we found that J chain-positive IgA containing either the
C15S or the C69S mutant had only slightly reduced affinity to free SC
compared with IgA containing the wild-type J chain. This result was
probably due to noncovalently stabilized dimers present in these IgA
preparations. In fact, the third cysteine (equivalent to human
Cys69) is lacking in the J chain from
Xenopus laevis (34), suggesting that disulfide
bonding between the J chain at this position and IgA or IgM is not
required for function in that species. The fact that either
Cys15 or Cys69 could be
mutated without affecting the formation of noncovalently stabilized
polymers suggested that covalent bond formation in J chain-mediated
polymerization does not occur in a specific sequence. Rather, a complex
between two IgA monomers and the J chain is first formed, and then
either Cys15 or Cys69 can
form a disulfide bridge with the adjacent IgA monomer. Presence of this
complex also increases the likelihood of disulfide bond formation
between the two IgA monomers. However, in the absence of such covalent
stabilization, a single bond between the J chain and one of the two IgA
monomers appears to be sufficient to form stable polymers. Notably,
these polymers disintegrate in SDS because one of the IgA monomers
interacts only noncovalently with the covalent IgA monomer-J chain
complex. We concluded that covalent bonding between one monomeric IgA
molecule and the J chain is required for IgA polymerization, because
the double mutant (C15+69S) did not increase polymerization of IgA at
all, although its intracellular expression level was similar to that of
the wild-type J chain (data not shown). Thus, while the J chain may
function as a scaffold for IgA polymerization, it also has a more
active role as its covalent interaction with IgA greatly enhances the
efficiency of pIgA formation.
Role of J chain in SC binding and pIgR-mediated epithelial transcytosis
Knockout mice deficient in J chain or pIgR have demonstrated that both molecules are required for the generation of SIgA in vivo (33, 35, 36, 37). Furthermore, rabbit IgG Abs to the J chain and Fab of such Abs could block pIgA-SC interaction in vitro and inhibit pIgR-mediated transcytosis of IgA, suggesting a direct role of the J chain in the binding of pIgA to pIgR (38, 39).
We used a mutational approach to address the role of J chain structural
elements in creating the pIgA docking site for SC/pIgR. Interestingly,
the requirements of the J chain structure that allowed for binding of
pIgA to SC was much more stringent than those determining IgA
polymerization. The C-terminal structure of the J chain was essential
to create a binding site for SC on pIgA because even the smallest
C-terminal deletion we tested abolished SC interaction. Furthermore,
disruption of the Cys109 to
Cys134 disulfide bond (either C109S or C134S)
abolished binding to SC, indicating that this C-terminal loop was
required for SC binding, perhaps via direct interaction with the pIgR.
Interestingly, the amino acid sequence between
Cys101 and Cys109 is very
highly conserved among different species, suggesting a particular
functional importance of this region of the J chain. Alternatively, the
C-terminal loop could be critical for a structural feature of the J
chain required for pIgA binding to pIgR. Ablation of the disulfide bond
between Cys13 and Cys101
also had a dramatic effect on the affinity of pIgA to SC, although in
some experiments we could detect significant, although marginal,
binding activity even with these incorporated mutants of the J chain.
The last intrachain bond, between Cys72 and
Cys92, appeared to be less required for binding
to SC, perhaps because J chain without this bond is still able to
maintain its overall three-dimensional structure. The apparent
reduction in IgA dimers formed when this bond was prohibited (Fig. 3B) could be a result of Cys69 (or
Cys15) being less accessible for disulfide bond
formation with the -chain. The relative high affinity of these IgA
molecules to SC suggested that the supernatants from C72S- and
C92S-transfected CHO cells contained noncovalently stabilized IgA
dimers in addition to the covalent dimers seen by SDS-PAGE, although we
did not attempt to verify this possibility.
When we fractionated CHO cell-produced IgA without J chain by anion
exchange chromatography, we found that the IgA eluted in three distinct
peaks, indicating three different species of NIP-specific IgA. The
first peak appeared to be slightly smaller than monomers by SDS-PAGE,
perhaps reflecting partially assembled IgA lacking a single L chain.
The second peak contained some dimers and larger polymers (seen by
native acrylamide gel electrophoresis) and, surprisingly, demonstrated
marginal affinity to free SC. This result showed that the J chain is
not absolutely required for IgA polymerization or for some binding to
free SC. Binding of J chain-negative pIgA to SC has also been observed
previously, indicating that the -chain contains an interaction site
for SC (40). The third peak contained only IgA monomers
that demonstrated no affinity for free SC. Thus, the -chain
interaction site for SC apparently requires covalent attachment between
two opposing tailpieces to create a novel surface constituting a
functional binding site for SC/pIgR. Alternatively, this site is
present on monomeric IgA, but polymerization results in higher affinity
measurable in our SC binding assays. However, the latter explanation is
unlikely, because we failed to detect any affinity between IgA monomers
and SC by increasing the concentration of monomers (data not shown). In
accord with our findings in this report, J chain-negative myeloma pIgA
and human hexameric IgM (without J chain) failed to bind SC properly
(7). Also, J chain-deficient myeloma pIgA and murine pIgA
isolated from J chain knockout mice were not transcytosed by the pIgR
(35, 41).
When we analyzed the chromatography-purified IgA from CHO cells by
native gel electrophoresis, we found that mutating
Cys15 or Cys69 in the J
chain yielded a fraction of pIgA that was indistinguishable from the
pIgA fraction produced with wild-type J chain (Fig. 5). Furthermore,
the peak pIgA fraction from these two mutants showed 72 (C15S) and
94% (C69S) efficiency of SC binding compared with wild-type pIgA.
However, the level of pIgR-mediated transcytosis was reduced to only 19
and 50% of that of wild-type pIgA, respectively. These findings
suggested that a binding site for SC/pIgR is not sufficient for
efficient transcytosis by the pIgR-expressing epithelium, but that
covalent stabilization of pIgA is also required. One possible
explanation could be that noncovalently stabilized pIgA is less
resistant to the reduced pH in the endosomal compartment of the MDCK
cells. Thus, in vivo, there might be an additional level of quality
control of pIgA that is to be become luminal SIgA.
In conclusion, we have demonstrated that the integrity of the J chain structure is more important for receptor-mediated epithelial transcytosis of pIgA than for polymerization per se. Thus, the striking J chain expression by mucosal plasma cells seems to have production of SIgA and SIgM as its biological goal. Indeed, no other functional role has been identified for this polypeptide (42). Furthermore, we showed that a single disulfide bond between the J chain and one IgA monomer was sufficient to enhance polymerization of IgA, although the resulting polymers were mostly noncovalently stabilized. Surprisingly, these polymers retained high binding affinity for free SC but were relatively poorly transcytosed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Finn-Eirik Johansen at the current address: Children’s Hospital, Enders Building 1220, 300 Longwood Avenue, Boston, MA 02115. E-mail address: f.e.johansen{at}labmed.uio.no
3 Abbreviations used in this paper: S, secretory; p, polymeric; SC, secretory component; MDCK, Madin-Darby canine kidney; CHO, Chinese hamster ovary; FPLC, fast protein liquid chromatography; NIP, 5-iodo-4-hydroxy-2-nitrophenylacetyl.
Received for publication June 20, 2001. Accepted for publication September 4, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain CH2 domain. J. Immunol. 159:244.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  N. M. Pedersen, K. Breen, M. S. Rodland, C. Haslekas, E. Stang, and I. H. Madshus Expression of Epidermal Growth Factor Receptor or ErbB3 Facilitates Geldanamycin-Induced Down-Regulation of ErbB2 Mol. Cancer Res., February 1, 2009; 7(2): 275 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bonner, P. B. Furtado, A. Almogren, M. A. Kerr, and S. J. Perkins Implications of the Near-Planar Solution Structure of Human Myeloma Dimeric IgA1 for Mucosal Immunity and IgA Nephropathy J. Immunol., January 15, 2008; 180(2): 1008 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. J. Mantis, J. Palaia, A. J. Hessell, S. Mehta, Z. Zhu, B. Corthesy, M. R. Neutra, D. R. Burton, and E. N. Janoff Inhibition of HIV-1 Infectivity and Epithelial Cell Transfer by Human Monoclonal IgG and IgA Antibodies Carrying the b12 V Region J. Immunol., September 1, 2007; 179(5): 3144 - 3152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Froyland, K. M. Thompson, S. J. Thorpe, S. S. Sahota, T. Gedde-Dahl, and B. Bogen A VH4-34+ myeloma protein with weak autoreactivity Haematologica, May 1, 2007; 92(5): 690 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Hamuro, H. Suetake, N. R. Saha, K. Kikuchi, and Y. Suzuki A Teleost Polymeric Ig Receptor Exhibiting Two Ig-Like Domains Transports Tetrameric IgM into the Skin J. Immunol., May 1, 2007; 178(9): 5682 - 5689. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Braathen, V. S. Hohman, P. Brandtzaeg, and F.-E. Johansen Secretory Antibody Formation: Conserved Binding Interactions between J Chain and Polymeric Ig Receptor from Humans and Amphibians J. Immunol., February 1, 2007; 178(3): 1589 - 1597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Davids, J. E. D. Palm, M. P. Housley, J. R. Smith, Y. S. Andersen, M. G. Martin, B. A. Hendrickson, F.-E. Johansen, S. G. Svard, F. D. Gillin, et al. Polymeric Immunoglobulin Receptor in Intestinal Immune Defense against the Lumen-Dwelling Protozoan Parasite Giardia J. Immunol., November 1, 2006; 177(9): 6281 - 6290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-H. Lim, S.-J. Cho, S.-K. Park, J. Kim, D. Cho, W. J. Lee, and C.-J. Kang Stage-Specific Expression of Two Neighboring Crlz1 and IgJ Genes during B Cell Development Is Regulated by Their Chromatin Accessibility and Histone Acetylation J. Immunol., October 15, 2006; 177(8): 5420 - 5429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Braathen, A. Sandvik, G. Berntzen, S. Hammerschmidt, B. Fleckenstein, I. Sandlie, P. Brandtzaeg, F.-E. Johansen, and V. Lauvrak Identification of a Polymeric Ig Receptor Binding Phage-displayed Peptide That Exploits Epithelial Transcytosis without Dimeric IgA Competition J. Biol. Chem., March 17, 2006; 281(11): 7075 - 7081. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Haslekas, K. Breen, K. W. Pedersen, L. E. Johannessen, E. Stang, and I. H. Madshus The Inhibitory Effect of ErbB2 on Epidermal Growth Factor-induced Formation of Clathrin-coated Pits Correlates with Retention of Epidermal Growth Factor Receptor-ErbB2 Oligomeric Complexes at the Plasma Membrane Mol. Biol. Cell, December 1, 2005; 16(12): 5832 - 5842. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Lewis, R. J. Pleass, M. R. Batten, J. D. Atkin, and J. M. Woof Structural Requirements for the Interaction of Human IgA with the Human Polymeric Ig Receptor J. Immunol., November 15, 2005; 175(10): 6694 - 6701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. C. Morton, R. J. Pleass, J. M. Woof, and P. Brandtzaeg Characterization of the Ligand Binding Site of the Bovine IgA Fc Receptor (bFc{alpha}R) J. Biol. Chem., December 24, 2004; 279(52): 54018 - 54022. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. B. Renegar, P. A. Small Jr, L. G. Boykins, and P. F. Wright Role of IgA versus IgG in the Control of Influenza Viral Infection in the Murine Respiratory Tract J. Immunol., August 1, 2004; 173(3): 1978 - 1986. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Lu, M. E. Lamm, H. Li, B. Corthesy, and J.-R. Zhang The Human Polymeric Immunoglobulin Receptor Binds to Streptococcus pneumoniae via Domains 3 and 4 J. Biol. Chem., November 28, 2003; 278(48): 48178 - 48187. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. S. Hohman, S. E. Stewart, L. L. Rumfelt, A. S. Greenberg, D. W. Avila, M. F. Flajnik, and L. A. Steiner J Chain in the Nurse Shark: Implications for Function in a Lower Vertebrate J. Immunol., June 15, 2003; 170(12): 6016 - 6023. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Royle, A. Roos, D. J. Harvey, M. R. Wormald, D. Van Gijlswijk-Janssen, E.-R. M. Redwan, I. A. Wilson, M. R. Daha, R. A. Dwek, and P. M. Rudd Secretory IgA N- and O-Glycans Provide a Link between the Innate and Adaptive Immune Systems J. Biol. Chem., May 23, 2003; 278(22): 20140 - 20153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Braathen, V. Sorensen, P. Brandtzaeg, I. Sandlie, and F.-E. Johansen The Carboxyl-terminal Domains of IgA and IgM Direct Isotype-specific Polymerization and Interaction with the Polymeric Immunoglobulin Receptor J. Biol. Chem., November 1, 2002; 277(45): 42755 - 42762. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Hempen, K. M. Phillips, P. S. Conway, K. H. Sandoval, T. A. Schneeman, H.-J. Wu, and C. S. Kaetzel Transcriptional Regulation of the Human Polymeric Ig Receptor Gene: Analysis of Basal Promoter Elements J. Immunol., August 15, 2002; 169(4): 1912 - 1921. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
