| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institut Suisse de Recherches Expérimentales sur le Cancer, Epalinges, Switzerland; and
Division d’Immunologie et d’Allergie, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
10 nM SC competitor. Western blot
analysis of IgAd digested with intestinal washes indicated
that the 
-chain in IgAd was primarily split into a
40-kDa species, a phenomenon delayed in rSC- or
SCm-IgAd complexes. In the same assay, either
of the SCs was resistant to degradation only if complexed with
IgAd. In contrast, the 
light chain was not digested at
all, suggesting that the F(ab')2 region was left intact.
Accordingly, IgAd and SC-IgAd digestion
products retained functionality as indicated by Ag reactivity in ELISA.
Size exclusion chromatography under native conditions of digested
IgAd and rSC-IgAd demonstrates that SC exerts
its protective role in secretory IgA by delaying cleavage in the
hinge/Fc region of the -chain, not by holding together degraded
fragments. The function of integral secretory IgA and
F(ab')2 is discussed in terms of mucosal immune
defenses. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mechanisms whereby sIgA Abs mediate protection are poorly understood. Aggregation, immobilization, and neutralization of pathogens at the mucosal surface are facilitated by the multivalency of sIgA (2). Such immune exclusion prevents the adhesion of micro-organisms to the luminal plasma membrane of the epithelial cells (3, 4). Polyclonal colostral Abs have been shown to protect humans against infection caused by enteropathogenic Escherichia coli (5), rotaviruses (6), Clostridium difficile (7), or Campylobacter jejuni (8). Less information is available as to the therapeutic potential of IgA/sIgA for passive mucosal protection. IgA Abs have been shown to be efficient in protecting against influenza virus infection (9). Only recently have mAbs been produced for analyzing the specificity and the mechanisms of protection against viruses and bacteria in the intestine (10, 11) or the airways (12); this is mostly due to the difficulties encountered in producing and purifying biologically active IgA and to the lack of enough SC for the reconstitution of sIgA in the laboratory.
Although future directions for producing IgAd/SC-IgAd Ab molecules might include recombinant plants (13, 14), transgenic animals (15, 16), or more conventionally engineered cell lines (17, 18, 19), the current methodology still relies on hybridoma technology. We have recently established a rapid and scalable procedure to purify IgAd from the mixture of molecular forms secreted by a hybridoma cell line (20), as well as expression systems suitable for production of human recombinant SC (rSC) in mammalian and insect cell hosts, which specifically associate with IgAd in vitro (21).
We demonstrate that the specificity, kinetics, and nature of
association of recombinant human SC (rSC) or SC isolated from human
colostral milk (SCm) to IgAd are the same by
all criteria tested. To our knowledge, no previous comparative study
has characterized the structure-function properties of purified
IgAd and reconstituted SC-IgAd after exposure
to intestinal washes. Upon association, rSC and SCm
increase the resistance of IgAd against proteases in vitro,
as shown by the delayed cleavage of the -chain. IgAd
also confers increased stability to SC, indicating mutual protection of
the partners in the SC-IgAd complex. Ag binding properties
of the native and digested SC-IgAd complex remain
undistinguishable, an observation in agreement with the preserved
structure of both the -chain and the Ab F(ab')2
fragment. Our results also suggest that the molecular properties of
SC-IgAd complex constitute a substantial advantage over
IgAd in terms of exogenous delivery for therapeutic passive
protection of mucosal surfaces.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
rSC was produced in recombinant vaccinia virus-infected HeLa S3
cells and purified as described (21). SCm was kindly
provided by Dr. Robert Jones (Department of Medical Biochemistry,
University of Geneva, Switzerland). The purification of
IgAd and monomeric IgA (IgAm) produced by the
mouse hybridoma ZAC3 (anti-Vibrio cholerae LPS) has been
previously described (20). Human sIgA was purchased from Sigma, and the
mAb M2 (a murine IgG1 recognizing the FLAG epitope) was from Eastman
Kodak (New Haven, CT).
In vitro reassociation of IgAd and SC
Purified IgAd from mouse hybridoma ZAC3 were combined with purified rSC or SCm, respectively, in PBS and incubated for the indicated time at ambient temperature, or in the case of reassociation, at different temperatures, for 4 h in a cold room (4°C) or in a water bath set at the appropriate temperature (15, 22, or 37°C). Reconstitution was usually performed with 10 µg of IgAd and 2 µg of purified SC in 100 µl final volume. When required, stable SC-IgAd complexes were purified by fast protein liquid chromatography (Pharmacia, Uppsala, Sweden), on a Superose 12 HR 10/30 column (21). Reactions were terminated by the addition of 1 volume of SDS-PAGE sample buffer (100 mM Tris-HCl, pH 6.8, 4% (w/v) SDS, 0.2% (w/v) bromophenol blue, 20% (v/v) glycerol). Covalently and noncovalently bound SC was subjected to SDS-PAGE under nonreducing conditions.
Measurement of SC-IgAd covalent association
The degree of covalent binding of SC to IgAd was monitored by immunoblotting using a rabbit antiserum recognizing both the free and covalently bound forms, followed by densitometric scanning. The percentage of free SC was assayed by a gel filtration procedure in which SC and IgAd mixed in vitro were applied onto a Superose 12 HR 10/30 column (30 cm x 1.0 cm) equilibrated and run in 50% (v/v) acetic acid. Under these conditions, SC noncovalently associated with IgAd elutes later as free SC. Fractions containing free SC were pooled, diluted in water to 1% (v/v) acetic acid, lyophilized, and resuspended in bicarbonate buffer, and the amount of protein was quantitated by ELISA (21). Subtraction from the total amount of SC initially mixed with IgAd reflects the percentage of covalently bound SC.
Sialidase treatment of human SC
One microgram of rSC and SCm was digested in 20 µl of 0.1 M sodium acetate, pH 4.5, for 5 h at 37°C with 2 milliunits of Arthrobacter ureafasciens sialidase (Calbiochem, San Diego, CA) or was incubated without enzyme. Protein samples were neutralized with 1 µl of 2 M sodium hydroxide before treatment for SDS-PAGE under reducing conditions.
Biotinylation of rSC and SCm
PBS buffer containing the protein samples was exchanged for 0.1 M sodium acetate (pH 5.5) by three cycles of dilution/concentration using a Centricon-50 filtration unit (Amicon, Beverly, CA). SC proteins at a concentration of 0.5 mg/ml were incubated on ice with 1 mM sodium periodate (final concentration) in the dark for 30 min, a condition in which sialic acids are selectively oxidized (22). Oxidation of sugar moieties was stopped by the addition of 1:10 (v:v) of 160 mM glycerol, followed by dilution/concentration as above. SC proteins in 300 µl were then reacted at ambient temperature for 2 h with 30 µl of 50 mM biotin-long chain-hydrazide (Pierce, Rockford, IL) dissolved in DMSO. The biotinylated proteins were desalted by buffer exchange against PBS containing 0.02% (w/v) sodium azide using a Centricon-50 filtration unit (Amicon), quantitated, and used within 2 wk of storage at 4°C.
Competition binding of human SC preparations to immobilized IgA
A competitive binding assay was employed to measure the relative
affinities of rSC and SCm for IgAd. The wells
of Nunc (Raskilde, Denmark MaxiSorp ELISA plates were coated
with 50 µl of purified ZAC3 IgAd (5 µg/ml (20))
dissolved in PBS. Purified ZAC3 IgAm (2.5 µg/ml) or mouse
IgG (2.0 µg/ml; Sigma, St. Louis, MO) were used in control
experiments. Nonspecific binding sites were blocked with Tris-buffered
saline (TBS; 25 mM Tris-HCl, 137 mM NaCl, 2.7 mM KCl, pH 7.5)
containing 5% (w/v) nonfat dry milk and 0.05% (v/v) Tween-20
(Bio-Rad, Hercules, CA). For the experiments described in Fig. 2
, biotinylated rSC or SCm in 100 µl of PBS (100 ng/ml)
was incubated at room temperature (23°C) for no more than 5 min, and
competition was initiated by the subsequent addition of unlabeled
SCm/rSC at concentrations ranging from 1.2 ng/ml to 40
µg/ml. The concentration of biotinylated SC used was in the linear
range of the binding curve. The competitor was incubated for 1 h
before washing, and detection of the undisplaced biotinylated SC was
performed with horseradish peroxidase-coupled ExtrAvidin (Sigma)
diluted at 1:1000. Control wells were incubated without competitor or
without biotinylated SC, respectively. Results were expressed as the
percentage of maximal biotinylated SC binding, which was obtained in
the absence of competitor. The IC50 for SCm and
rSC was defined as the concentration of competitor that inhibited by
50% the binding of biotinylated SC. Experiments shown in Fig. 3 were
performed under the same conditions with the following modifications:
biotinylated rSC was incubated for 5 min or 100 min; incubations were
conducted at 4°C or 23°C during the binding/competition
procedure.
|
|
Mouse intestinal washes were obtained from Dr. Irène
Corthésy-Theulaz (Division of Gastroenterology, Lausanne
Hospital, Switzerland). Briefly, mice fasted overnight were
anesthetized, and a laparotomy was performed. The gut was cut at the
stomach-duodenal junction and near the anus, and the intestinal lumen
was washed with 300 µl of PBS using a 26-gauge needle. Final
aspiration resulted in the recovery of 200 µl of intestinal wash,
which was immediately aliquoted and frozen in liquid nitrogen. For in
vitro digestion under gentle conditions, 2 µg of IgA or 2.5 µg of
sIgA in 20 µl were mixed with 2 µl of intestinal washes,
either undiluted or diluted at 1:3, 1:10, or 1:50, or with PBS alone,
and incubated at 37°C for 16 h. Digestions under harsher
conditions were performed as above, with a final Ab-intestinal fluid
ratio of 1:1, for 20 h at 37°C. Reactions were stopped by the
addition of 2 µl of Complete (protease inhibitor mixture;
Boehringer-Mannheim, Mannheim, Germany). Digests were kept at
-70°C until further use. Analysis of the digestion products
originating from the -chain, the -chain, and SC in
IgAd/SC-IgAd-containing samples was conducted
by SDS-PAGE and immunoblotting according to the methods of Lüllau
et al. (20).
Binding of digested IgAd and SC-IgAd to immobilized Ag
The Ag binding capacity of IgAd and reconstituted
SC-IgAd samples exposed to intestinal proteases was
assessed using LPS-specific ELISA. Wells of Nunc MaxiSorp plates were
coated with 40 µg/ml of LPS from V. cholerae,
serotype Inaba 569B (Sigma), and nonspecific binding sites were blocked
as described (20). Wells were incubated overnight at 4°C with 100
µl of IgAd and SC-IgAd digests or with
undigested preparations, all diluted in blocking buffer to 40 ng/ml of
IgA. After washing with PBS/0.05% Tween-20, the bound molecules were
directly detected with either biotinylated polyclonal Ab to mouse
-chain (Sigma; diluted 1:1000) or rabbit polyclonal Ab to mouse
-chain (Cappel, Turnhout, Belgium; diluted 1:500), both
incubated in blocking buffer for 1 h at room temperature.
Appropriate secondary reagents coupled to horseradish peroxidase
(Sigma) were used at a 1:1000 dilution under the same conditions.
Sandwich ELISAs were performed to correct for small variations between
protein samples, which could have otherwise affected the amount of
LPS-bound -chain and -chain. Capture was conducted using a goat
polyclonal Ab to mouse -chain (Sigma; diluted 1:500) in 50 µl of
bicarbonate buffer (pH 9.6); mouse -chain and -chain were
detected as above. LPS binding of undigested samples was taken as 100%
ELISA values.
Size exclusion separation under native conditions of digested IgAd and SC-IgAd
Twenty micrograms of IgAd or 25 µg of
SC-IgAd in 200 µl were mixed with 20 µl of intestinal
washes and incubated for 20 h at 37°C. Reactions were stopped by
the addition of 10 µl of Complete (protease inhibitor mixture;
Boehringer-Mannheim). 200 µl of the sample was loaded onto a 30
x 1-cm Superose 12 HR 10/30 column coupled to a fast protein liquid
chromatography system (Pharmacia). The column was pre-equilibrated and
run in PBS at 4°C using a flow rate of 0.5 ml/min. Fractions (250
µl) were collected into siliconized tubes and 1/10 was mixed with
SDS-PAGE sample buffer containing 100 mM DTT, and boiled for 3 min.
Every other fraction was analyzed by immunodetection with antisera
against mouse -chain (Sigma), mouse -chain (Cappel), and human SC
(21). Calibration of the column was performed with purified cytochrome
c (Sigma), carbonic anhydrase (Sigma), BSA (Sigma), rabbit
IgG F(ab')2 (Dako, Copenhagen, Denmark), murine IgG
(Nordic, Tilburg, The Netherlands), human sIgA (Sigma), and dextran
blue (Sigma).
Analytical methods
Proteins were quantitated with the bicinchoninic acid assay (Pierce), using BSA as a standard. Capture ELISA for human SC quantitation was performed as previously described (21). Silver staining of SDS-PAGE gels was performed according to Morrissey (23).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have shown that rSC bound covalently to IgAd following overnight incubation at ambient temperature (20). However, no information exists as to the optimal time and temperature required, nor to the behavior of rSC as compared with native SCm.
The kinetics of formation of covalent SC-IgAd complexes was
therefore determined by incubating equimolar amounts of purified
IgAd and rSC or SCm for increasing times,
followed by resolution of the heterodimer by SDS-PAGE under nonreducing
conditions. Covalent complexes were identified by immunoblotting with
an anti-SC serum (Fig. 1A). Disulfide links had
already formed after 5 min of incubation, and a plateau was
reached after 1 h at ambient temperature (Fig. 1B).
Very similar kinetics of association were obtained with rSC and
SCm, indirectly validating recombinant vaccinia
virus-infected mammalian cells as a convenient alternative source of SC
polypeptide. After 1 h of incubation, the proportion of covalent
complexes was estimated by size exclusion chromatography in 50% (v/v)
acetic acid (Fig. 1C). Under these denaturing conditions,
20% of total SC eluted as free rSC, indicating that 80% was
associated as covalent rSC-IgAd. When the same column was
run in PBS, 98–100% of rSC comigrated with IgAd as
covalent and noncovalent complexes (21).
|
D, no difference could be seen from 15–37°C, while
incubation at 4°C resulted in reduced formation of covalent
complexes. Gel filtration experiments following incubation of rSC and
IgAd in PBS at 4°C also revealed a lower level of
noncovalent association (Fig. 1E), i.e., up to 65% of SC
was recovered in free form as measured by ELISA of the pooled late
eluting fractions. Consistently, at other temperatures of association,
covalent and noncovalent SC-IgAd complexes formed to
completion (see above). When evaluating the effect on in vitro covalent
assembly of redox agents such as DTT and glutathione, we found no
visible improvement, with the proportion of covalently linked SC again
representing 80% (B.C, unpublished observations). Peptidyl disulfide
isomerase (PDI), a disulfide-interchange enzyme catalyzing covalent
bond formation in vitro, did not modify the intensity of the recovered
covalent SC-IgAd complex either (B.C., unpublished
observations). rSC and SCm bind IgAd with a similar affinity
Since the integrity of the sugar moieties in SC is not required
for binding to IgAd (24, 25), we chose to biotinylate the
sialic acid residues of the N-linked carbohydrate side chains to
produce labeled SC for competition studies. The effect of sialidase
treatment on SC is illustrated in Fig.
2A. rSC migrated as a single
broad band of 75 kDa in SDS-PAGE, whereas SCm appeared
as a doublet of 80 and 75 kDa. This difference probably reflects a
distinct glycosylation pattern, although partial proteolysis at the
carboxy-terminus cannot be ruled out (26). Sialidase treatment
increased the migration of rSC, but not SCm, indicating
that sialylation of rSC differs from that of the natural product, which
displays predominantly monosialylated glycans with some disialylated
and neutral N-linked chains (27). The same migration patterns were
observed in different preparations. Despite differences in sialylation,
both proteins could be biotinylated after mild oxidation of the sialic
acids.
The ability of SCm and rSC to competitively block their own
or reciprocal binding to immobilized IgAd is shown in
Figure 2, B and C. These experiments performed at
room temperature clearly indicate that rSC binds as efficiently as
SCm, irrespective of its use as a biotinylated ligand or as
an unlabeled competitor. The binding results provide more evidence that
rSC is undistinguishable from SCm in most respects.
Cross-competition experiments with a mixture of biotinylated and native
human SC demonstrated that chemical modification of SC did not affect
the binding properties to IgAd, and hence measurements of
IC50 (50% inhibiting concentration) in the nanomolar range
reflect the actual binding capacity of rSC and SCm (Table I). Binding was specific since no native
or biotinylated SC bound to immobilized IgAm or IgG (P.C.,
unpublished observations).
|
The similar properties of rSC and SCm allowed us to
address the unsolved issue of protein exchange in covalent
SC-IgAd complexes. When the competition experiments were
performed at 4°C within 5 min postassociation (Fig.
3B, filled circles), the
IC50 of 20 nM was almost identical to the 15 nM measured at
23°C (Fig. 3A, filled circles; see also Fig. 2C
and Table I). This result implies that noncovalent interactions
contribute extensively to the initial stability of SC-IgAd
complexes, given that the covalency is strongly reduced at 4°C (Fig. 1D). Since formation of covalent SC-IgAd takes
place within 5 min at room temperature (Fig. 1, A and
B), it is possible that newly formed covalent bonds are
labile. However, when the competitor was added after equilibrium had
been reached (100 min), only 35–40% of the bound biotinylated SC
could be displaced in experiments performed at 23°C (Fig. 3A, open circles), as compared with 60–65% in competitions
conducted at 4°C (Fig. 3B, open circles). These data
corroborate the weaker degree of covalency at low temperature,
indicating that a detectable degree of exchange occurs between SC in
SC-IgAd complexes and free SC used as a competitor.
Accordingly, equilibrium binding at 4°C followed by competition at
23°C resulted in a 67% displacement of the bound SC (P.C.,
unpublished observations). The curves for the portion that could be
displaced after equilibrium (Fig. 3, open circles) yielded
IC50 values of 55–60 nM at both temperatures, reflecting
the improved stability of the SC-IgAd complexes after
longer incubation times. These observations validate the use of ELISA
in quantitating both the overall binding of SC to immobilized IgA and
in establishing the fraction that binds reversibly. Furthermore,
the data suggest a two-step temperature-dependent mechanism of
association, with the first step being reversible (see
Discussion).
Mutual protection of SC and IgAd against protease digestion
The susceptibility of IgAd and SC-IgAd
complexes to digestion with intestinal washes was determined in vitro
at Ig concentrations reflecting those present in intestinal secretions
(28, 29, 30, 31). First, we analyzed the effect of proteases on the -chain
in IgAd and rSC-IgAd reconstituted as in Fig. 1. In the absence of intestinal washes, no difference could be observed
between IgA and rSC-IgAd heterodimers (Fig.
4A, lanes 0), indicating that
the purified Igs in the reaction mixture were stable during the period
of incubation (16 h). Increasing amounts of intestinal washes led to a
progressive appearance in immunoblots of low m.w. products reactive to
the anti--chain antiserum. Comparison of the lanes showed that
10-fold more concentrated washes were necessary to achieve a similar
degree of degradation between IgAd and
rSC-IgAd, demonstrating that the presence of SC in the
rSC-IgAd molecule delayed the action of the many
proteolytic activities present within the washes. Similar digestion
patterns of the -chain were obtained with
SCm-IgAd (B.C., unpublished observations).
Under the same incubation conditions, free SC was already completely
digested after 30 min of incubation (Fig. 4B). In contrast,
when associated with IgAd, SC was found to be stable after
16 h of digestion (Fig. 4C), indicating that the
association of SC and IgAd mutually protects each
polypeptide from proteolytic degradation.
|
While the intestinal dilutions used above were appropriate to
study the differential sensitivity of IgA and sIgA, the large amount of
-chain left undigested made it difficult to measure the Ag binding
properties of digested vs undigested molecules. Therefore, more
stringent conditions were applied: intestinal washes were diluted only
twice, and digestion was pursued for 20 h. Digestion caused the
conversion of the 62-kDa -chain into a major 40-kDa band in
IgAd as compared with rSC-IgAd. (Fig.
5A). Densitometric scanning of
the blots indicated that the signal for -chain in IgAd
decreased to 10% of the value for rSC-IgAd complexes.
Bands of discrete sizes were detectable, as already seen in digestion
mixtures performed under less stringent conditions. The complexed rSC
and SCm were partially resistant and yielded
undistinguishable digestion patterns. Interestingly, the -chain
remained uncleaved in both IgAd and rSC-IgAd
(Fig. 5A), suggesting that the F(ab')2 fragment
of both Ig forms is resistant to proteolysis.
|
A) can still bind to the V. cholerae
LPS Ag in ELISA. Binding of digested Ab molecules to the Ag was
detected using antisera directed against the - or the -chain
(Fig. 5B). The Ag binding capacity of the Ab was marginally
reduced when measured using the anti--chain antiserum. Similar
results were obtained when using the anti--chain antiserum,
indicating that no or very few epitopes were lost after digestion of
the -chain. This suggests that the overall Ab structure remains
assembled, yet exhibits sites of cleavage in the polypeptide backbone.
Further, the preserved competence of degraded IgAd to
associate with the Ag suggests that the band at 40 kDa is part of the
F(ab')2 fragment and/or the truncated IgAm (see
below). In contrast, a mouse IgG1 mAb (M2) treated under identical
conditions was completely degraded and therefore could no longer bind
to its cognate Ag, i.e., a protein tagged with the FLAG epitope (P.C.,
unpublished observations). Degradation of IgAd is delayed but not abrogated in the presence of SC
To determine more directly the structural features of digested
rSC-IgAd, the Igs treated as above were loaded onto a
sieving column equilibrated and run in PBS. Three peaks were obtained
(Fig. 6A), with peak 1 more
prominent with the digested rSC-IgAd material and peak 2
more intense with IgAd digests used as a control. When
analyzed by Western blot assay under reducing conditions, peak 1 was
shown to contain both the full length -chain (62 kDa) and the 40-kDa
band, indicating that a portion of the Ig variable and constant regions
remained assembled after digestion (Fig. 6B). In peak 1, the
similar intensity of bands at 62 and 40 kDa (Fig. 6B,
panel -chain), together with the detection of bands at 230 and
320 kDa (Fig. 6B, nonreduced, lane 39), is
consistent with a model in which the majority of the molecules consist
of dimers with two of the four -chains having lost the C3 domain
and a portion of the C2 domain (Fig. 6C). Since the same
pattern was obtained with IgAd (B.C., unpublished
observations), this excludes SC in a role of holding together the
rSC-IgAd complex, but demonstrates its protective function
at the molecular level. Protection is further reflected by the
detection of J chain in SC-IgAd; J chain stability in
SC-IgAd might indirectly reflect the close proximity of SC
in the complex. Conversely, J chain is no longer detectable in
IgAd, yet a significant portion of the digested
IgAd molecules keeps eluting as a dimer (Fig. 6A) owing to the direct covalent link existing between two
monomers (32). In addition to the vanishing band at 62 kDa, peak 2
comprised bands at 60 and 58 kDa, most likely resulting from the
C-terminal degradation of C3 domain in IgAm, together
with the 40-kDa band being part of F(ab')2 and the
C3/C2 truncated IgAm (Fig. 6B, panel
-chain). The presence of bands at 80 and 100 kDa in fraction
48, analyzed under nonreducing conditions (Fig. 6B),
confirms this model (Fig. 6C) and indicates that the
F(ab')2 is the major degradation product. Thus, degradation
of IgAd appears to start in the J chain region, and this
process is delayed in the presence of SC. The mere presence in
fractions 54–63 of the -chain-reactive 36-kDa species (most likely
fragment Fd) demonstrates that Fc is being degraded into small
fragments after release from the molecule. Given the relative m.w. of
proteins in peak 3, Fd and the -chain (the latter being a covalent
dimer in ZAC3 (20)) have dissociated into independent entities. In
conclusion, using a combination of calibration and immunodetection data
(Fig. 6, A and B), we show that peak 1 contains
truncated rSC-IgAd molecules, respectively,
IgAd partly lacking the C3/C2 domains, peak 2
comprises the F(ab')2 fragment and truncated
IgAm, and peak 3 contains the Fd fragment, as well as
undegraded -chain dimers and monomers and SC fragments as seen in
Fig. 5A. The biologic implication of these findings is
discussed below.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 domain of mouse
IgA—a residue well positioned to participate in covalency to SC—was
found in the reduced form in 65–90% of the chains (40), setting an
upper limit to the covalent association between SC and
IgAd. The identical behavior of rSC as compared with
SCm thus represents a strong indication of the correct
structure of the recombinant product. We observed that covalent association of rSC or SCm with IgAd reached a plateau after 1 h of incubation at 25°C. Incubation of rSC and IgAd at 4°C for 4 h reduced the degree of covalent association from 80% down to 30–35%, a result that can be explained in part by the limited production of noncovalent complexes. This suggests that mutual recognition of the two partners involves dynamic conformational changes that are inhibited at low temperature. Conversely, Lindh and Björk (34) linked such a temperature effect only to the formation of disulfide bonds, even while using a 2:1 SC to IgAd ratio. We found that DTT and glutathione did not enhance the formation of covalent association occurring spontaneously through the sulfhydryl-disulfide exchange reaction between SC and IgAd. This finding is in agreement with the lack of effect of alkylating agents upon covalent association of SCm with IgAd and with the restricted accessibility of free thiols in IgAd under native conditions (37). While the addition of PDI was without effect in our assay, Murkofsky and Lamm (41) reported that incubating SCm and polymeric IgA in the presence of PDI promoted the formation of noncovalent complexes, whereas the ratio of covalent vs noncovalent species remained the same. However, the low percentage of binding to IgA that Murkofsky and Lamm were able to see suggests that their SC preparation contained a significant amount of denatured protein, which during exposure to PDI might have acquired IgA binding capacity.
Using competition ELISA, we determined the relative affinities
(IC50) of rSC and SCm for mouse
IgAd as 5–15 nM at room temperature and 20 nM at 4°C.
Using the same assay at 4°C, an IC50 of 3–30 nM (25, 42, 43) has been reported for the interaction of human SCm with
human polymeric IgA. The IC50 for the recombinant domain I
of human SC was only 160 nM (43), indicating that significant binding
forces are contributed by domains II–V in rSC. A dynamic exchange
could apparently take place between covalently bound partners, because
unlabeled SC was able to completely displace biotinylated SC when added
within 5 min following the initial mixing of SC and IgAd,
i.e., after covalency has already been detected. Consistently, a
Kd of 10 nM has been determined by Weicker
and Underdown (36) using SCm and myeloma IgA under
conditions allowing covalency; displacement by competition was
suggested, although not strictly demonstrated. However, when we
modified the assay by adding the competitor after the equilibrium had
been reached, only a fraction of the bound SC could be competed out,
amounting to 35–40% at ambient temperature and 60–65% at 4°C.
This finding is consistent with covalent association being temperature
dependent and becoming irreversible with time. Further, we could
determine an IC50 of 55–60 nM for the reversible
interaction taking place at room temperature and at 4°C, indicating
that the noncovalent complexes had improved their stability with time.
Thus, our competition data strongly suggest that the primary
determinant for high affinity binding is of a noncovalent nature,
whereas a temperature-sensitive conformational change is required
before the sulfhydryl-disulfide interchange reaction can take place.
Some conformational changes have been detected in the second and third
Ig-like domains of SCm upon binding to IgA (44). In
addition, a local conformational change in IgAd preceding
and facilitating the covalent bond formation was suggested from the
inaccessibility of its critical thiol(s) to alkylating agents (37). In
early stages of the reaction (<5 min) involving the first Ig domain of
SC (25, 42, 43, 44, 45), SC molecules can still be exchanged, due to complete
reversibility. We propose that within 100 min of reaction, about
two-thirds of the bound SC molecules (one-third at 4°C) have
undergone structural changes leading to a "locked" conformation.
This hypothesis is reinforced by the observation that, in the absence
of denaturation, exposure of colostral sIgA complexes to 0.2 M 2-ME
results in only a partial release of SC (Ref. 37; and B.C., unpublished
observations). The irreversible nature of some interactions could be
due to the presence of a second disulfide bond between SC and
IgAd, as proposed by some authors (37, 46). However, a
direct analysis of sIgA isolated from milk has demonstrated that a
single bond linked Cys467 of SC to Cys311 of
one -chain, whereas Cys501 of SC and Cys311
of the second -chain in the same monomer were blocked with cysteine
(32). It is therefore conceivable that the reaction milieu could
influence the formation of disulfide bonds vs mixed disulfides. While
the Cys307 of mouse IgA is reduced in its majority and is
well positioned to initiate the disulfide interchange (40), it is not
clear which Cys of IgA could form a second bond by classical oxidation.
Previous studies conducted with sIgA isolated from mucosal secretions
have demonstrated that exocrine sIgA was more resistant to proteolytic
digestion than was serum IgA or IgG (47, 48). Furthermore, IgA combined
in vitro with SC had an increased resistance to proteases (49).
Importantly, in the present study (as in 49 , free
IgAd and SC-bound IgAd were directly
comparable. We have now reappraised the protease sensitivity of both
IgAd and SC-IgAd using intestinal washes
reflecting more closely the complex environment encountered by sIgA in
vivo. The patterns of -chain digestion in IgAd and
SC-IgAd were remarkably similar, yet fragmentation of
SC-IgAd was delayed in time when compared with
IgAd. This suggests that the same sensitive sites of IgA
remain exposed in SC-IgAd. It is thus likely that small
structural changes in bound SC could temporarily give access to
proteases. SC in SC-IgAd was also fragmented into discrete
fragments, whereas free SC was rapidly and completely degraded.
Following exposure to intestinal washes, both IgAd and
SC-IgAd exhibited Ag binding properties close to the native
Ig, whereas a control IgG was degraded. This suggests that the
F(ab')2 portions remained resistant to digestion, in
agreement with the observed absence of -chain degradation. In a
very recent study, the virus-neutralizing activity of an sIgA was
shown to resist mild trypsination, whereas both IgAp and
IgAm were degraded, the latter more significantly (50).
Many less F(ab')2 fragments were produced in the
IgAd associated with SC than in naked IgAd. The
appearance of degraded -chain at the position of elution of the
whole Ig (peak 1 in Fig. 6A) indicates intramolecular
stabilization resembling that seen in nicked DNA molecules. Since this
effect is seen for both IgAd and SC-IgAd, it is
not a consequence of SC stabilizing the overall structure. The
protection, rather than stabilization, provided by SC is further
demonstrated by the higher abundance of the 62-kDa band in
SC-IgAd digests as compared with IgAd (B.C.,
unpublished observations); this indicates that SC reduces the
sensitivity of the -chain hinge region. The role of SC in masking
sensitive stretches in the molecule is also reflected by the detection
of J chain in SC-IgAd digests, but not in IgAd
digests. Further, the production of IgAm that were digested
at their very carboxyl terminus is reminiscent of Pronase digests of
human IgAd (48). Thus, the sequential intermediate products
we detected suggest that progressive "peeling," starting from the
-chain C terminus, and proteolytic clipping in the hinge—an open,
extended area—lead to the production of F(ab')2, truncated
IgAm, Fd, and free dimers. Furthermore, only one-half
of the -chains seemed to be originally exposed to proteolysis. The
validity of these results is confirmed by the in vivo susceptibility of
the hinge region of human rIgAd and SC-IgAd,
the latter being more long lived (19), and by the presence in normal
human stool of sIgA fragments of various sizes, some smaller than
F(ab')2 and larger ones (200 kDa) containing SC (51).
sIgA in mucosal secretions combines with micro-organisms to reduce their motility, growth, and adhesive properties within the mucosal lumen and at its surface. Such a coating function could be fulfilled by either integral sIgA or F(ab')2 fragments following stabilization with protein Fv (51). Thus, one can postulate that cleavage of sIgA into F(ab')2 and Fc fragments in the gastrointestinal tract may not be totally deleterious to their function, but could advantageously permit rapid clearance by peristalsis. In contrast, Fab fragments were less potent than the native sIgA in inhibiting Streptococcus mutans adherence on saliva-coated hydroxy-apatite (52, 53, 54). This could be explained either by F(ab')2 being endowed with avidity or by the necessity to maintain the effector function of the Ab. It has been speculated that sIgA might form particular associations with mucins via their hinge/Fc regions and thus entrap the Ag-Ab complex within the mucus layer (55, 56), permitting limited Ag absorption for controlled presentation to the immune cells. Such a mechanism is plausible, as we have demonstrated that recombinant antigenized SC-IgAd bearing a B epitope in SC can successfully reach intestinal Ag sampling sites and thus function as a delivery vehicle when given orally (57). The ability of partially degraded sIgA to prevent Ag uptake might also reflect the role of this particular class of Abs in the avoidance of hypersensitivity to food allergens (58, 59, 60).
The extensive characterization of rSC conducted in this study provides convincing evidence that the protein can substitute its natural counterpart in in vitro reconstituted SC-IgAd to be used for passive immunization and active vaccination (33). It has now become possible to use biochemically well-defined material to address issues such as the mucosal mode of action (Peyer’s patches adsorption/absorption, peristalsis, epithelial cell binding) and the duration of protection of IgA vs sIgA following exogenous delivery.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Biozentrum, University of Basel, CH-4056 Basel, Switzerland.
3 Address correspondence and reprint requests to Dr. Blaise Corthésy, Division d’immunologie et d’allergie, Centre Hospitalier Universitaire Vaudois, BH18-701, Rue du Bugnon, CH-1011 Lausanne, Switzerland. E-mail address:
4 Abbreviations used in this paper: sIgA, secretory IgA; IgAd, dimeric IgA; IgAm, monomeric IgA; PDI, protein disulfide isomerase; SC, secretory component; SCm, human milk SC; rSC, recombinant human SC; IC50 (for SCm and rSC), the concentration of competitor that inhibited by 50% the binding of biotinylated SC.
Received for publication March 3, 1998. Accepted for publication July 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-1-antitrypsin by a flock of sheep. Biotechnology 11:1263.[Medline]
1-domain disulphide bridge. Biochem. J. 225:113.[Medline]
domains. J. Immunol. 151:1346.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  S. Boullier, M. Tanguy, K. A. Kadaoui, C. Caubet, P. Sansonetti, B. Corthesy, and A. Phalipon Secretory IgA-Mediated Neutralization of Shigella flexneri Prevents Intestinal Tissue Destruction by Down-Regulating Inflammatory Circuits J. Immunol., November 1, 2009; 183(9): 5879 - 5885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bonner, A. Almogren, P. B. Furtado, M. A. Kerr, and S. J. Perkins The Nonplanar Secretory IgA2 and Near Planar Secretory IgA1 Solution Structures Rationalize Their Different Mucosal Immune Responses J. Biol. Chem., February 20, 2009; 284(8): 5077 - 5087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E.J. Helmerhorst and F.G. Oppenheim Saliva: a Dynamic Proteome Journal of Dental Research, August 1, 2007; 86(8): 680 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bonner, C. Perrier, B. Corthesy, and S. J. Perkins Solution Structure of Human Secretory Component and Implications for Biological Function J. Biol. Chem., June 8, 2007; 282(23): 16969 - 16980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Fasching, T. Grossman, B. Corthesy, A. G. Plaut, J. N. Weiser, and E. N. Janoff Impact of the Molecular Form of Immunoglobulin A on Functional Activity in Defense against Streptococcus pneumoniae Infect. Immun., April 1, 2007; 75(4): 1801 - 1810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Perrier, N. Sprenger, and B. Corthesy Glycans on Secretory Component Participate in Innate Protection against Mucosal Pathogens J. Biol. Chem., May 19, 2006; 281(20): 14280 - 14287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Tjarnlund, A. Rodriguez, P.-J. Cardona, E. Guirado, J. Ivanyi, M. Singh, M. Troye-Blomberg, and C. Fernandez Polymeric IgR knockout mice are more susceptible to mycobacterial infections in the respiratory tract than wild-type mice Int. Immunol., May 1, 2006; 18(5): 807 - 816. [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]
|
|
|
|
|
|
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]
|
|
|
|
|
|
L. Favre, F. Spertini, and B. Corthesy Secretory IgA Possesses Intrinsic Modulatory Properties Stimulating Mucosal and Systemic Immune Responses J. Immunol., September 1, 2005; 175(5): 2793 - 2800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. R. Peters, E. L. Calvert, E. J. Hall, and M. J. Day Measurement of Immunoglobulin Concentrations in the Feces of Healthy Dogs Clin. Vaccine Immunol., September 1, 2004; 11(5): 841 - 848. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Rey, N. Garin, F. Spertini, and B. Corthesy Targeting of Secretory IgA to Peyer's Patch Dendritic and T Cells after Transport by Intestinal M Cells J. Immunol., March 1, 2004; 172(5): 3026 - 3033. [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]
|
|
|
|
 |
|
 R. S. Solorzano-Vargas, J. Wang, L. Jiang, H. V. Tsai, L. O. Ontiveros, M. A. Vazir, R. J. Aguilera, and M. G. Martin Multiple transcription factors in 5'-flanking region of human polymeric Ig receptor control its basal expression Am J Physiol Gastrointest Liver Physiol, August 1, 2002; 283(2): G415 - G425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F.-E. Johansen, R. Braathen, and P. Brandtzaeg The J Chain Is Essential for Polymeric Ig Receptor-Mediated Epithelial Transport of IgA J. Immunol., November 1, 2001; 167(9): 5185 - 5192. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Kirkeby, T. T. Rasmussen, J. Reinholdt, and M. Kilian Immunoglobulins in Nasal Secretions of Healthy Humans: Structural Integrity of Secretory Immunoglobulin A1 (IgA1) and Occurrence of Neutralizing Antibodies to IgA1 Proteases of Nasal Bacteria Clin. Vaccine Immunol., January 1, 2000; 7(1): 31 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Crottet, M. C. Peitsch, C. Servis, and B. Corthesy Covalent Homodimers of Murine Secretory Component Induced by Epitope Substitution Unravel the Capacity of the Polymeric Ig Receptor to Dimerize Noncovalently in the Absence of IgA Ligand J. Biol. Chem., October 29, 1999; 274(44): 31445 - 31455. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Crottet and B. Corthesy Mapping the Interaction Between Murine IgA and Murine Secretory Component Carrying Epitope Substitutions Reveals a Role of Domains II and III in Covalent Binding to IgA J. Biol. Chem., October 29, 1999; 274(44): 31456 - 31462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F.-E. Johansen, M. Pekna, I. N. Norderhaug, B. Haneberg, M. A. Hietala, P. Krajci, C. Betsholtz, and P. Brandtzaeg Absence of Epithelial Immunoglobulin a Transport, with Increased Mucosal Leakiness, in Polymeric Immunoglobulin Receptor/Secretory Component-Deficient Mice J. Exp. Med., October 4, 1999; 190(7): 915 - 922. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Roe, I. N. Norderhaug, P. Brandtzaeg, and F.-E. Johansen Fine Specificity of Ligand-Binding Domain 1 in the Polymeric Ig Receptor: Importance of the CDR2-Containing Region for IgM Interaction J. Immunol., May 15, 1999; 162(10): 6046 - 6052. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. W. Ackermann, L. A. Wollenweber, and G. M. Denning IL-4 and IFN-{gamma} Increase Steady State Levels of Polymeric Ig Receptor mRNA in Human Airway and Intestinal Epithelial Cells J. Immunol., May 1, 1999; 162(9): 5112 - 5118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Berdoz, C. T. Blanc, M. Reinhardt, J.-P. Kraehenbuhl, and B. Corthesy In vitro comparison of the antigen-binding and stability properties of the various molecular forms of IgA antibodies assembled and produced in CHO cells PNAS, March 16, 1999; 96(6): 3029 - 3034. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
);
rSC mixed with IgAd (
). C, Gel filtration
of complexes through a Superose 12 column (30 x 1 cm) run in 50%
acetic acid. Under these denaturing conditions, the position of elution
of noncovalently associated SC is shifted from SC-IgAd to
free SC. D, Formation of covalent SC-IgAd
complexes as a function of temperature. The 320-kDa cross-reacting band
was not observed in Figs. 1
