The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sørensen, V.
Right arrow Articles by Sandlie, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sørensen, V.
Right arrow Articles by Sandlie, I.
The Journal of Immunology, 1999, 162: 3448-3455.
Copyright © 1999 by The American Association of Immunologists

Polymerization of IgA and IgM: Roles of Cys309/Cys414 and the Secretory Tailpiece1

Vigdis Sørensen*, Vibeke Sundvold{dagger}, Terje E. Michaelsen{ddagger} and Inger Sandlie2,*

* Department of Molecular Cell Biology, Institute of Biology, University of Oslo, Olso, Norway; {dagger} Institute of Immunology and Rheumatology, The National Hospital, Oslo, Norway; and {ddagger} Department of Vaccinology, National Institute of Public Health, Oslo, Norway


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We have investigated how the secretory tailpiece (tp), Cys414 and the amino acids flanking Cys414 or Cys309 are involved in regulating the different polymerization of IgM and IgA to pentamers and dimers/monomers, respectively. Whereas changing the tp of IgM to that of IgA has little effect on IgM polymerization, introducing the µtp to IgA leads to the formation of larger than wild-type IgA polymers, including pentamers and hexamer. This shows that the secretory tp can differentially regulate polymerization depending on the heavy chain context. Cys414, which is engaged in intermonomeric disulfide bonds in IgM, is not crucial for the difference in IgM and IgA polymerization; IgM with a C414S mutation forms more large polymers than IgA. Also, IgA with IgM-like mutations in the five amino acids flanking Cys309, which is homologous to Cys414, oligomerize similarly as IgA wild type. Thus, IgA appears to have an inherent tendency to form monomers and dimers that is partially regulated by the tp, while the Cys309 region has only a minor effect. We also show that complement activation by IgM is sensitive to alterations in the polymeric structure, while IgA is inactive in classical complement activation even for polymers such as pentamers and hexamers.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
IgM and IgA occur as covalent polymers of the H2L23 monomeric subunit. IgM forms mainly pentamers, while IgA forms mainly monomers and dimers. A polymeric structure is thought to increase the Abs avidity for Ag and is required for functions such as complement activation by IgM and transepithelial secretion of IgA and IgM. Knowledge about the structural basis for the polymerization will thus give information on the regulation of essential functions of IgM and IgA and may also be useful for the construction of multimers of other members of the Ig superfamily or engineered Ab-derived molecules as a means to increase their ligand avidity and/or modulate features such as tissue penetrance and serum half-life.

The polymerizing ability of IgM and IgA depends on cysteine residues in the heavy chains that form intermonomeric disulfide bonds. Three cysteine residues, Cys337, Cys414, and Cys575 are available for inter-heavy chain disulfide bonds in IgM. Cys337 probably forms the intramonomeric bond between Cµ2 domains while Cys414 in Cµ3 and Cys575 in Cµ4 are involved in intermonomeric bonds 1, 2, 3, 4 . However, the precise arrangement of the interchain disulfide bonds in the IgM polymer is not clear. Cys575 is located in the secretory tailpiece (tp) of the µ heavy chain, which is the extra 18 amino acids (µtp) found in the C-terminal domain of the secretory form of µ heavy chain and has its homologue in the {alpha} heavy chain ({alpha}tp). Eleven of eighteen amino acids are identical in the µtp and the {alpha}tp, including the penultimate cysteine numbered 575 in µ and 495 in {alpha}. The tp cysteine is critical for polymerization of both IgM and IgA 5, 6, 7 and is also the site for incorporation of J chain to IgM and IgA polymers 8, 9, 10 . The presence of J chain is required for binding of IgM and IgA to the poly-Ig receptor on epithelial cells, which mediates transcytosis of the Abs 11 . The J chain also influences IgM and IgA polymerization; in the absence of J chain, IgM will form more hexamers 12 , while IgA form fewer dimers 13 . Whereas Cys414 also participates in polymerization of IgM, a homologous cysteine residue, Cys309 (sometimes denoted Cys311), in the {alpha}-chain does not seem to participate in IgA dimerization, but instead forms a disulfide bridge to the poly-Ig receptor 14 .

Cys575 in the µtp has also been implicated as a key motif for regulating intracellular retention and degradation of unpolymerized IgM monomers in B cells and plasma cells 15 . Introduction of the murine {alpha}tp to IgM led to increased secretion of monomeric IgM, and it was proposed that secretion of unpolymerized IgA from plasma cells is a result of inefficient intracellular retention due to the amino acid composition of the {alpha}tp 16 . However, we have previously shown that human IgM molecules engineered to have a human {alpha}tp (IgM{alpha}tp) polymerize similarly to IgMwt, i.e., the same amount and the same type (pentamers and hexamers) of polymers are formed 17 . The µtp and {alpha}tp sequences also similarly induced formation of polymers when introduced to IgG, including pentamers and hexamers. Polymerization of IgG was most efficient when a Cys414/309 homologue was introduced in C{gamma}2 in combination with either µtp or {alpha}tp 17, 18 . Formation of large polymers has also been reported for other IgG subclasses with an engineered C-terminal µtp 19 . Based on these observations, we concluded that although the tp strongly favors polymerization, the µtp and {alpha}tp do not carry sufficient structural information to differentially regulate the pattern of polymerization. Rather, both IgM and IgG seem to have an inherent ability to form high polymers such as pentamers and hexamers, regardless of whether the tp is µ or {alpha}, while special features of the {alpha} heavy chain normally restrict assembly of IgA to monomers and dimers.

In this study we further investigate the structural basis for the different polymerization of IgM and IgA by altering the tp of IgA to that of IgM. We also study the contribution of the Cys414 and Cys309 residues to polymerization. We find that the {alpha}tp is an important determinant for dimer formation in the context of an IgA molecule, whereas the Cys309/Cys414 region has a minor effect. We also show that complement activation by IgM varies with its polymeric structure, while both low and high m.w. polymers of IgA are inactive in complement activation by the classical pathway.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Plasmids and mutagenesis

The cloning of the human µ and {alpha}1 constant genes into the vector pUC19 has been described previously, as has the IgM{alpha}tp mutant 17 . Mutagenesis on µ and {alpha} genes was performed in the M13 mp19 vector according to the method of Kunkel 20 , using reagents purchased from Bio-Rad (Richmond, CA).

Construction of IgMC414S and IgMC414S{alpha}tp A 1.9-kb SmaI fragment of the µ gene, containing part of the Cµ1 exon, the Cµ2 and Cµ3 exons, and part of the Cµ4 exon, was cloned in pUC19 on HincII sites. This 1.9-kb fragment was further cloned in M13 mp19 for mutagenesis. The C414S mutation in Cµ3 was introduced using the oligonucleotide 5'-atcctcgctgatgctgg-3'. Thereafter, a 1.3-kb DraIII-SacII fragment containing the mutation was sequenced and substituted with the corresponding fragment in the IgMwt and the IgM{alpha}tp heavy chain genes in pUC19 to produce the IgMC414S and IgMC414S{alpha}tp heavy chain genes.

Construction of IgAµtp, IgAalm, IgAalma, and IgAalm-µtp An 1.8-kb fragment containing the entire {alpha}1 constant region gene was amplified by PCR and cloned in M13 mp19 on HindIII-BamHI sites. The coding sequence of the {alpha}tp was changed to the sequence of the µtp using the oligonucleotide 5'-tagcaggtgccggccgtgtccgacatgacaagagacacattgtaaagggtgggtttacc-3'. The five mutations, P307S, G308I, A310E, E311D, and P312D, were introduced in the IgAwt heavy chain gene and to an already mutated IgAµtp heavy chain gene using the oligonucleotide 5'-cttcccatggttccaatcatcttcacatatgtccaggac-3'. These mutants are called IgAalm and IgAalm-µtp, respectively. The IgAalma mutant differs from IgAalm by the E311A mutation and was the result of a single nucleotide misincorporation during mutagenesis using the alm-oligonucleotide (see Fig. 1Go for overview of mutants). For all the IgA mutants an 1.3-kb XhoI-BamHI fragment was sequenced and substituted for the corresponding fragment in the IgA heavy chain gene in pUC19. The wild-type and mutated gene constructs were cloned on HindIII/BamHI sites in the expression vector pSV2gptVNP (a gift from Dr. M. S. Neuberger, Medical Research Council Laboratory of Molecular Biology, Cambridge, U.K.) or pLNOH2VNP 21 downstream of a VH gene segment specific for the hapten 4-hydroxy-3-nitrophenacetyl (NP).



View larger version (33K):
[in this window]
[in a new window]
  		FIGURE 1. Amino acid sequence of the secretory tp and the Cys414/Cys309 region of wild-type and mutated IgM and IgA. Amino acids are depicted by single letter symbols. The Cµ3, Cµ4, C{alpha}2, and C{alpha}3 domains are illustrated by boxes. Mutated amino acids/tp sequences are indicated in bold letters.

 
Cell culture and transfections

The mutant heavy chain genes were transfected by electroporation into the murine myeloma cell line J558L (a gift from Dr. S. L. Morrison, Department of Microbiology, Molecular Biology Institute, University of California-Los Angeles). The cells were grown and transfected as previously described 17 . J558L constitutively produces a mouse {lambda}1 light chain and J chain, but no Ig heavy chain. Stable transfectants were selected in a medium containing 250 µg/ml xanthine (Sigma, St. Louis, MO) and 1 µg/ml mycophenolic acid (Life Technologies, Grand Island, NY) for pSV2gptVNP plasmids or in medium containing 0.8 mg/ml G418 (Duchefa, Haarlem, The Netherlands) for pLNOH2VNP plasmids. (Using either of these expression vectors gave the same results.) Ab-producing colonies were identified by analyzing cell supernatants by ELISA using NIP-BSA-coated microtiter wells as previously described 22 . Three to five Ab-secreting colonies from each transfection were expanded for further analysis.

Metabolic labeling, immunoprecipitation, and gel electrophoresis

For analysis of secreted Abs, approximately 2 x 106/ml cells were labeled in methionine and cysteine-free DMEM (BioWhitaker, Walkersville, MD) containing 100 µCi of [35S]methionine and cysteine for 6 h before the supernatant was harvested. For pulse-chase analysis, approximately 10 x 106 cells were starved in DMEM without methionine and cysteine for 45 min before labeling with 250 µCi of [35S]methionine and cysteine for 5 min in the presence of 5 µg/ml brefeldin A (Sigma). The cells were then immediately washed and resuspended in prewarmed complete medium containing a 10-fold excess of unlabeled methionine and cysteine (Life Technologies) and 5 µg/ml brefeldin A. At the times indicated, aliquots of the culture were collected and held on ice. The cells were washed in cold PBS and lysed in 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 1% Nonidet P-40 supplemented with 60 mM iodoacetamide and protease inhibitors to final concentrations of 4 µg/ml PMSF, 2 µg/ml leupeptin, 1 µg/ml antipain, and 1 µg/ml pepstatin. IgM/IgA was immunoprecipitated from cell lysates or supernatants by rabbit anti-human IgM (µ-chain specific) or anti-human IgA ({alpha}-chain specific) Abs (Sigma) followed by 0.25 mg of Dynabeads sheep anti-rabbit IgG (Dynal, Oslo, Norway). The precipitate was washed twice in cold PBS and 1% Nonidet P-40 and eluted from the beads by incubation in gel loading buffer at 95°C for 3 min. The samples were analyzed nonreduced in a 4% SDS acrylamide-agarose composite gel as previously described 17 . Dried gels were exposed to BIOMAX-MR film (Eastman Kodak, Rochester, NY) and for quantitation of bands in the gels, the gels were analyzed by a phosphorimager (GS-250 Bio-Rad molecular imager).

Complement-mediated lysis (CML)

The ability of the Abs to activate complement by the classical pathway was measured by a CML assay as previously described 22 . Briefly, 51Cr-labeled SRBC were coated with NIP-conjugated rabbit anti-SRBC Fab at various concentrations and patchiness (NIP-4-Fab', 80 ng; NIP-15-Fab', 400 ng; NIP-60-Fab', 2000 ng). Serial dilutions of the Abs (supernatant from transfected J558L) were added to the target cells, followed by addition of human serum. Cell lysis was recorded as released chromium. The cytotoxic index (CI) was calculated according to the formula: % CI = (cpm test - cpm spontaneous/cpm total (dH2O/0.1% Tween) - cpm spontaneous) x 100.

Isolation of polymer fractions

The different polymer fractions of selected IgA mutants were isolated from outgrown cell culture supernatant by affinity chromatography on NP-coupled to AH-Sepharose (Pharmacia, Uppsala, Sweden) as previously described 23 . The isolated IgA proteins were further separated in different polymer fractions by gel filtration on a Superdex-200 (Pharmacia) resulting in higher polymer, trimer, dimer, and monomer fractions. Although the fractions were somewhat overlapping, they were judged to be at least 80% pure, and monomer contamination in the oligomer/polymer fractions was <5%. The fractions were separately tested for CML activity.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Involvement of the tp in polymerization of IgM and IgA

Mutations in the human µ heavy chain that alter the µtp sequence into the corresponding human {alpha}tp sequence (Fig. 1Go) do not affect the secretion of mainly pentameric and hexameric IgM molecules from transfected J558L cells (Fig. 2GoA, IgMwt and IgM{alpha}tp) 17 . We did, however, observe a change in the hexamer/pentamer ratio toward more hexamers for IgM{alpha}tp (Table IGo).



View larger version (46K):
[in this window]
[in a new window]
  		FIGURE 2. Effects of mutations on polymerization of IgM and IgA. Biosynthetically labeled Abs were immunoprecipitated from the supernatants of transfected cells and analyzed in a 4% gel as described in Materials and Methods. The proteins run in the individual lanes are indicated on the top of the gel. The identity of each assembly product, as indicated by arrows, is based on comparison to migration of wild-type Abs, m.w. standards, and analysis of migration length vs estimated m.w. The relative amounts of the different assembly products were measured and are presented in Table IGo. A, IgM mutants; B, IgA mutants.

 

View this table:
[in this window]
[in a new window]
  		Table I. Distribution of assembly products1

 
To investigate whether the {alpha}tp structure is involved in determining the formation of monomeric and dimeric IgA, we constructed a human {alpha} heavy chain where the {alpha}tp was mutated to the µtp (Fig. 1Go). Expression of this heavy chain in J558L cells led to the secretion of monomeric and polymeric IgAµtp molecules as shown in Fig. 2GoB. The relative amounts of the different assembly products were quantitated and are reported in Table IGo. The monomeric fraction of IgAµtp was relatively large compared with that of IgAwt, indicating that the µtp does not actually increase the polymerization efficiency of IgA monomers. However, there was a significant difference in the polymeric structures formed by IgAwt and IgAµtp; IgAµtp formed relatively few dimers, but formed more large polymers such as trimers, tetramers, and even pentamers and hexamers, which were not observed for IgAwt. Thus, the {alpha}tp is necessary for polymerization of IgA into mainly dimers. However, since the {alpha}tp does not induce dimer formation of IgM, other regions in the {alpha} heavy chain must contribute to this type of polymerization, just as regions in addition to the µtp contribute to the formation and secretion of mainly pentameric IgM.

Involvement of Cys414/309 in polymerization of IgM and IgA

Possibly, an important difference in IgM and IgA polymerization is that IgM uses two cysteine residues, the Cys414 residue in addition to the tp cysteine, while IgA polymerization apparently only involves the tp cysteine. To study the contribution of the Cys414 residue for polymerization, we mutated this residue to serine in both IgM and IgM{alpha}tp. As shown in Fig. 2GoA, IgMC414S and IgMC414S{alpha}tp secreted from J558L cells have a similar distribution of assembly products, ranging in size from monomers to pentamers. The major polymers of IgMC414S and IgMC414S{alpha}tp are tetramers and pentamers, not pentamers and hexamers as is the case for IgMwt, i.e., apparently one monomer less is incorporated in the major polymeric structures. Thus, the Cys414 residue to some extent determines the number of monomers incorporated in IgM polymers. We found no great difference in the distribution of assembly products between IgMC414S and IgMC414S{alpha}tp; thus, even in the absence of Cys414-Cys414 bonds, the {alpha}tp does not regulate the formation of other IgM structures than the µtp. However, quantitation of the different assembly products of IgMC414S and IgMC414S{alpha}tp for three to five clones indicated that in the presence of the {alpha}tp, more of the small polymers were formed than in the presence of the µtp (Table IGo). Furthermore, IgMC414S and IgMC414S{alpha}tp formed considerably more large polymers than IgAµtp and IgA, indicating that motifs in addition to the tp and Cys414 contribute to the regulation of polymerization.

The amino acids flanking Cys414 and Cys309 in IgM and IgA, respectively, are quite different (Fig. 1Go). By comparison to IgG sequence and crystal structure, the Cys309/414 residue may well be positioned in or near the loop between ß-strands E and F in the C{alpha}2/Cµ3 domain (Fig. 3Go). This could be an exposed loop with residues important for both noncovalent and covalent monomer-monomer interactions. We altered the five residues flanking Cys309 in IgA to become identical with those found in the Cys414 region of IgM. This {alpha} loop to µ loop (alm) mutation was introduced into both IgAwt (IgAalm) and IgAµtp (IgAalm-µtp). However, only aminor effect on polymerization was observed for the IgAalm mutant (Fig. 2GoB); The dimeric fraction was similar to that of IgAwt, while there was a small increase in trimers and tetramers. The trimers and tetramers constituted approximately 6% of the total amount of IgAalm, while for IgAwt, trimers constituted approximately 2–3%, and tetramers were barely detectable. The IgAalm-µtp mutant had a polymer distribution very similar to that of IgAµtp, with the one exception that no hexamers were formed. In the mutagenesis reaction we also obtained a variant of the IgAalm mutant that has a D to A substitution (IgAalma). IgAalma formed more large polymers than either IgAwt or IgAalm. The tetrameric fraction of IgAalma constituted approximately 12%, and even pentamers and hexamers were observed, although these two species together constituted only approximately 3% of the total (Table IGo). The results show that making IgA more like IgM in the Cys309 region to some extent can facilitate participation of the Cys309 residue in polymerization, but the effect is quite small and does not significantly alter the formation of mainly monomers and dimers. Thus, neither the Cys414 nor the Cys309 region seems to play a major role in determining whether large or small polymers of IgM and IgA are formed.



View larger version (35K):
[in this window]
[in a new window]
  		FIGURE 3. Models of Fc structures. A, Fc region of an IgG1 molecule. The C{gamma}2 and C{gamma}3 domains of one heavy chain are depicted in a wire frame (back), and the other heavy chain is shown in ribbons (front). The Fc is oriented so that the two residues in position 309 are approximately overlapping. In this position the front 309 side chain is pointing out of the paper toward the reader. Residues 307–312 in the front C{gamma}2 domain and the C, F, and G ß strands (Y face) in the front C{gamma}3 domain are shown in gray. The model was made using RasMol and is based on file 1FC1 from the Brookhaven Database. B, Sketch of an Ig constant domain with a secretory tp. ß strands A–G and hypothetical ß strands formed by the secretory tp are depicted by arrows. In this model we suggest that the 18 amino acids of the tp form two extra ß strands next to the G strand in the Y face of the domain (white strands).

 
Intracellular polymerization rates

The observed differences in the sizes of the total polymeric fractions between different Ab constructs, as measured in the cell supernatant, may reflect differences in the polymerizing ability of the monomers. However, they could also reflect differences in how the monomers are retained or secreted by the cells, as has been suggested previously for IgM and IgA monomers 16 .

To compare the polymerizing abilities of IgM, IgMC414S, IgA, and IgAµtp monomers, we performed a pulse-chase experiment in which intracellular assembly was studied at several time points from 0–120 min (Fig. 4Go). Since any differences in monomer secretion efficiencies would differentially deplete the intracellular pool of substrate for polymerization, we performed the pulse-chase experiments in the presence of brefeldin A, which blocks transport of proteins out of the endoplasmic reticulum. Thus, all monomers and assembly intermediates were available for polymerization for the same length of time. As shown in Fig. 4GoA, assembly proceeded through H, HL, H2L, H2L2 (monomer), and larger multimeric species. The polymeric fraction, defined as the sum of all assembly products larger than the monomer, was calculated for each time point and is plotted against time in Fig. 4GoB.



View larger version (73K):
[in this window]
[in a new window]
  		FIGURE 4. Pulse-chase analysis of intracellular polymerization of IgM, IgMC414S, IgA, and IgAµtp. A, Transfected cells of IgM, IgMC414S, IgA, and IgAµtp were starved in cysteine-deficient medium, pulse labeled with [35S]methionine and cysteine, and chased at the indicated times in medium containing 5 µg/ml brefeldin A. At each time point an aliquot of the cells were lysed, and Ab was immunoprecipitated and resolved in a 4% gel as described in Materials and Methods. m, monomers; d, dimers; p, pentamers. B, Gels from three or four independent experiments performed on two different clones of each Ab were analyzed on a phosphorimager. The amounts of intracellular polymers (assembly products larger than the monomer) were quantitated at each time point and are plotted as an average percentage of the total intracellular Ab-associated radioactivity vs time. The SD at each point was ±10% or less.

 
We found that polymerization of IgM was quite efficient, showing 50% polymers after 60 min and approximately 65% after 120 min. Polymerization of IgMC414S, IgA, and IgAµtp was far less productive; maximum intracellular polymer fraction only reached 10–25%. Although this maximum level was reached within 20–30 min, prolonged incubation did not yield an increase in the polymeric fraction or the formation of larger polymers. We have also studied the intracellular assembly under normal conditions (in the absence of brefeldin A) and obtained similar results; the intracellular polymeric IgM fraction reached a higher level than the corresponding polymeric IgMC414S and IgA fractions (results not shown).

Thus, the IgMwt monomers have a better ability to form polymeric structures than the IgA monomers, and efficient intracellular polymerization of IgM depends on the Cys414 residue. It is therefore likely that the secretion of smaller assembly products of IgA is mainly a consequence of a limited ability to polymerize, although the results cannot rule out the possibility that there are also differences in intracellular retention of the molecules.

The Fc structure

At present there are no solved crystal structures for IgM or IgA Fc regions, but assumptions about the structure can be made based on the known structure of IgG. The C{gamma}2 and C{gamma}3 domains in the Fc region of IgG are homologous to the Cµ3 and Cµ4 domains in IgM and the C{alpha}2 and C{alpha}3 domains in IgA. Fig. 3GoA shows an IgG Fc with highlighting of residues 307–312, corresponding to the amino acids mutated in IgAalm, and the ß strands of the Y face (C, F, and G) in C{gamma}3. The X face of the C-terminal domain is at the interface in intramonomeric interactions; thus, the Y face could be the site for intermonomeric interactions (i.e., polymerization) in IgM/IgA. There is no homologue in the {gamma}-chain to the secretory tp; however, the overall homology between the C-terminal domains of the different classes suggests that the tp can be considered an extension to the very end of the C-terminal domain. It seems reasonable to suggest that the 18 amino acids of the tp form an extension to the ß sheet of the Y face in the C-terminal domain,possibly in the form of one or two ß strands next to the G strand (Fig. 3GoB). With this proposed structure of IgM and IgA Fcs, a surface for monomer interactions can be imagined that confines the Cys414/309 loop region as well as part of the Y face of the C-terminal domain, including the tp.

Complement activation by IgM, but not IgA

Whereas polymeric IgM can activate complement, IgA is inactive in classical complement activation. However, it is not known whether this is because IgA lacks a C1q binding site or because it does not form an appropriate polymeric structure. We have therefore investigated whether large polymers of IgA, such as those obtained for IgAµtp, can activate complement, and how complement activation by IgM is affected by variations in the polymeric structure. IgA, IgAµtp, IgM, IgM{alpha}tp, IgMC414, and IgMC414S{alpha}tp were tested in a CML assay as described in Materials and Methods.

We found that IgM{alpha}tp shows a higher maximum activity than IgMwt, while IgMC414S and IgMC414S{alpha}tp are inactive (Fig. 5GoA). This is in agreement with previous findings. Hexameric mouse IgM was reported to be 20 times more active in complement activation than pentamers 6, 24 , and we found that the IgM{alpha}tp mutant has a higher hexamer/pentamer ratio than IgMwt (Table IGo). Also, mouse IgMC414S has previously been reported to be inactive in complement activation 2 . IgAwt as well as IgAµtp are inactive in CML (Fig. 5GoB). As the amounts of polymers in the different IgA preparations were low, IgA monomers might block a possible activity of polymeric IgA. Therefore, we isolated three polymeric fractions (dimers, trimers, and higher polymers) of IgAµtp and IgAalma, and tested each fraction for CML activity at a minimum of 1 µg/ml. All IgA polymeric fractions were totally negative. Thus, even pentameric and hexameric structures found in the IgAµtp and IgAalma preparation are unable to activate complement.



View larger version (14K):
[in this window]
[in a new window]
  		FIGURE 5. Analysis of complement-mediated cell lysis induced by IgM, IgM{alpha}tp, IgMC414S, IgMC414S{alpha}tp, IgA, and IgAµtp. 51Cr-labeled target cells were hapten sensitized and incubated with dilutions of the NP-specific Abs IgM, IgM{alpha}tp, IgMC414S, or IgMC414S{alpha}tp (A) or with IgA or IgAµtp (B), followed by addition of human serum as described in Materials and Methods. Cell lysis was recorded as released chromium and plotted vs the Ab concentration. Similar results were obtained with different Ag concentrations and patchiness on the target cells.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We show here that an IgA molecule with a µtp (IgAµtp) forms larger polymers than IgAwt. Whereas IgAwt forms mainly monomers and dimers, IgAµtp forms monomers and a range of polymeric structures from dimers to hexamers. However, IgAµtp forms fewer polymers in total; the monomeric fraction of IgAµtp is larger than the corresponding monomeric fraction of IgAwt. Such a significant differential effect of the {alpha}tp and µtp is not observed for other sets of molecules differing only by {alpha}tp and µtp; IgMwt and IgM{alpha}tp, IgMC414S and IgMC414S{alpha}tp, IgGµtp and IgG{alpha}tp, IgGL309Cµtp and IgGL309C{alpha}tp (as reported here and in 17 . However, we have also earlier observed variable effects of a specific tp mutation in IgG and IgM, which suggested that the effect of a tp sequence on the polymeric structure depends on other parts of the heavy chain 17 . In the context of an IgA molecule the {alpha}tp can thus be considered an important motif for regulating dimer formation, whereas in the context of an IgM or IgG molecule an {alpha}tp is not instructive for dimer formation.

It has been proposed for murine IgM that the µtp contains an intracellular retention motif that prevents the secretion of monomeric and incompletely polymerized IgM molecules. It was shown that murine IgM with a murine {alpha}tp is secreted partially as monomers, and it was therefore suggested that monomeric IgA is secreted from plasma cells because the amino acids flanking the tp cysteine weaken intracellular retention 16 . However, this is at variance with our results for human IgM and IgA, where the µtp does not prevent the secretion of monomeric IgA, and the {alpha}tp does not increase the secretion of monomeric IgM. Rather, our results show that IgA polymerization is less efficient than IgM polymerization, and we therefore suggest that secretion of monomers and small polymers is mainly due to inefficient polymerization, and that both polymerization and monomer retention may be regulated by complex motifs comprising the tp as well as other parts of the molecules.

The distribution of IgAµtp polymers resembles the previously observed distribution of IgGµtp and IgG{alpha}tp polymers 17 . For the IgG mutants, formation of large polymers such as pentamers and hexamers was greatly enhanced by the introduction of a cysteine residue in position 309 (IgGL309Cµtp and IgGL309C{alpha}tp) 17, 18 . To what extent Cys309 is employed for formation of dimeric or higher polymeric IgA is not clear. Chapius and Koshland 25 reported that the tp cysteine forms the only disulfide bridges between the monomers in dimeric serum IgA, while, according to Yang et al. 26 , some, but not all, Cys309 residues form Cys309-Cys309 intermonomeric bonds in serum polymeric IgA. Thus, theoretically, Cys309 is either structurally restricted from participating in polymerization or is accessible but not used for other reasons. The observed overall homology between the Cµ3 and C{alpha}2 domains suggests that Cys414 and Cys309 may have similar positions in the two domains, although the five amino acids immediately flanking the cysteines are quite different (Fig. 1Go). Proline residues in positions 307 and 312 of IgA may affect the local structure around Cys309. Cys414 in IgM is flanked by three acidic residues (positions 415, 416, and 417), compared with one in IgA (position 311), which may be of relevance for cysteine reactivity 26 or conformation. However, we find that the effect of making IgA IgM-like in the five amino acids flanking Cys309 is modest. Only a small increase in the amount of larger polymers is observed, which could be somewhat increased by reducing the number of acidic residues (mutants IgAalm and IgAalma). Thus, the reactivity of the Cys309 residue apparently is improvable. However, monomers and dimers are still the most common forms also for the IgAalm/alma mutants, suggesting that the IgA monomers have an inherent tendency to form dimers, which is only weakly affected by the Cys309 region. The structure around Cys309, however, may be of importance for the disulfide bridge to the poly-Ig receptor that is made as the IgA-poly-Ig receptor complex is transported through epithelial cells 14, 27 . IgM, on the other hand, interacts with poly-Ig receptor only noncovalently.

Mutation of Cys414 in IgM also shows that employment of Cys414 for intermonomeric disulfide bonds is not of major importance for the polymerization pattern of IgM, since IgMC414S and IgMC414S{alpha}tp consist of more and larger polymers than IgA. The consequence of mutating Cys414 is, however, that one less monomer is incorporated in the main polymeric structures, i.e., instead of hexamers and pentamers, about equal amounts of pentamers and tetramers are formed for IgMC414S (as well as IgMC414S{alpha}tp). We can suggest two possible explanations for this effect of the C414S mutation. The absence of Cys414-Cys414 disulfide bonds may have a general effect on the stacking of the monomers in the polymer, such that one less monomer can be incorporated. Alternatively, Cys414 is the only cysteine used for covalent incorporation of one of the monomers in a polymer, while the other monomers can be incorporated by Cys575 only. Possibly, this is related to incorporation of the J chain to the polymer.

Formation of specific disulfide bonds must be preceded by noncovalent interactions between the polypeptides. Thus, the conformation of an initial noncovalent monomer-monomer interaction should direct the final polymeric structure of the Ab. As shown here and previously 17 , the amino acid composition of the tp can regulate the number of monomers incorporated in the polymer, but in a way that varies with the heavy chain context. Regions of the heavy chain that modulate the effect of the tp or independently contribute to regulation of polymerization are less well defined. Knowledge about the Fc structure of IgM and IgA would be required to identify a motif for monomer interactions, but a model based on the known structure of IgG can be useful for making predictions. Pumphrey 28 has previously suggested and modelled an extended structure for the µtp and IgM monomer interactions through Y faces of the Cµ4 domains. Considering the length of the tp and other ß strands in Ig domains and also the general folding patterns of ß-structure proteins 29 , we think it is reasonable to suggest that the tp might form two ß strands that are an integrated part of the Cµ4 domain (Fig. 3Go). An extended Y face in the C-terminal domain could comprise most of the monomer interface. In the computer-generated model of IgM by Perkins et al. 30 , the monomers are packed in a pentamer by close contact between Cµ4 domains. The closest contact is between the edges of the Y faces, e.g., the G strands of adjacent Cµ4 domains. However, this model, which is based on crystallographic data for IgG, does not include specific modelling of the 18 extra amino acids in the µtp. In the model by Perkins et al. the Cys414/309 residues of adjacent monomers make close contact, but there is less extensive contact between Cµ3 domains than between Cµ4 domains. The possibility for formation of Cys414-Cys414/Cys309-Cys309 bonds could therefore be dependent on the way the C-terminal domains interact. The structures of IgM and IgA as seen in electron microscopic images 31 indicate that the monomers interact in quite different relative orientations in the pentamer and the dimer, respectively. This suggests that despite the homology between Cµ4 and C{alpha}3 domains, a significant difference exists in the monomer interface for IgM and IgA, and thus, it is possible that the amino acid variations in the C-terminal Cµ4 and C{alpha}3 domains account for the different polymerizations of IgM and IgA.

IgA has presumably evolved from IgM by duplication and mutation of the µ gene. Many of the features of IgM, such as the tp and Cys414/309 and the abilities to form polymers, bind the J chain, and bind poly-Ig receptor, are conserved in IgA, while the differences include formation of lower polymers/monomers and inability of complement activation for IgA. The ability of IgM to activate complement is highly dependent on the polymeric conformation; loss of Cys414-Cys414 bonds makes the molecules inactive 2 , while hexamers are more active than pentamers 6, 24 , and monomers are inactive 32 . It is believed that the C1q binding site is disguised in uncomplexed IgM. A conformational change, dependent on Cys414-Cys414 bonds and apparently features of all the Cµ2, Cµ3, and Cµ4 domains, seems to be required to expose the C1q site after complexing with Ag 2, 33, 34 . Although specific amino acids in Cµ3 have been implicated as a C1q binding motif 35 , this is not fully elucidated. IgA does not have the C1q binding motifs suggested for IgG 36, 37 or the motif suggested for IgM and does not activate complement by the classical pathway 38 . It cannot, however, be ruled out that a C1q binding site exists in IgA and that the monomeric/dimeric conformation is preventing efficient C1q interaction. We investigated this by testing whether the IgAµtp mutant as well as enriched polymeric fractions of IgAµtp and IgAalma can activate complement in a CML assay. The mutant IgAs, however, are completely inactive, as is IgAwt. Although this argues for the lack of a C1q binding site in IgA, it cannot be ruled out that a failure to adopt a required polymeric conformation is sufficient to prevent complement activation by IgA. This may represent one of the special adaptions of IgA to accommodate its functions near the mucosal membranes where it might be beneficial to limit general tissue damage caused by complement activation. The smaller size of IgA than IgM may also be advantageous for mucosal protection by allowing improved diffusion through the extracellular matrix associated with mucosal epithelia 39 .


   Acknowledgments
 
We thank Randi Sandin for technical assistance.


   Footnotes
 
1 This work was supported by Grants 107257/310 and 107256/310 from the Research Council of Norway and Grant B95078/001 from the Norwegian Cancer Association. Back

2 Address correspondence and reprint requests to Dr. Inger Sandlie, Department of Biology, Division of Molecular and Cell Biology, University of Oslo, P.O. Box 1050, 0316 Oslo, Norway, E-mail address: Back

3 Abbreviations used in this paper: H, heavy chain of Ig; L, light chain of Ig; tp, tailpiece; wt, wild type; NP, 4-hydroxy-3-nitrophenacetyl; NIP, 5-iodo-NP; CML, complement-mediated lysis. Back

Received for publication March 2, 1998. Accepted for publication December 4, 1998.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Beale, D., A. Feinstein. 1969. Studies on the reduction of a human 19S immunoglobulin M. Biochem. J. 112:187.[Medline]
  2. Davis, A. C., K. H. Roux, J. Pursey, M. J. Shulman. 1989. Intermolecular disulfide bonding in IgM: effects of replacing cysteine residues in the µ heavy chain. EMBO J. 8:2519.[Medline]
  3. Milstein, C. P., N. E. Richardson, E. V. Dieverson, A. Feinstein. 1975. Interchain disulphide bridges of mouse immunoglobulin M. Biochem. J. 151:615.[Medline]
  4. Wiersma, E. J., M. J. Shulman. 1995. Assembly of IgM. Role of disulfide bonding and noncovalent interactions. J. Immunol. 154:5265.[Abstract]
  5. Atkin, J. D., R. J. Pleass, R. J. Owens, J. M. Woof. 1996. Mutagenesis of the human IgA1 heavy chain tailpiece that prevents dimer assembly. J. Immunol. 157:156.[Abstract]
  6. Davis, A. C., K. H. Roux, M. J. Shulman. 1988. On the structure of polymeric IgM. Eur. J. Immunol. 18:1001.[Medline]
  7. Sitia, R., M. Neuberger, C. Alberini, P. Bet, A. Fra, C. Valetti, G. Williams, C. Milstein. 1990. Developmental regulation of IgM secretion: the role of the carboxy-terminal cysteine. Cell 60:781.[Medline]
  8. Bastian, A., H. Kratzin, K. Eckart, N. Hilschmann. 1992. Intra- and interchain disulfide bridges of the human J chain in secretory immunoglobulin A. Biol. Chem. Hoppe-Seyler 373:1255.[Medline]
  9. Mestecky, J., R. E. Schrohenloher, R. Kulhavy, G. P. Wright, M. Tomana. 1974. Site of J chain attachment to human polymeric IgA. Proc. Natl. Acad. Sci. USA 71:544.[Abstract/Free Full Text]
  10. Mestecky, J., R. E. Schrohenloher. 1974. Site of attachment of J chain to human immunoglobulin M. Nature 249:650.[Medline]
  11. Brandtzaeg, P., H. Prydz. 1984. Direct evidence for an integrated function of J chain and secretory component in epithelial transport of immunoglobulins. Nature 311:71.[Medline]
  12. Randall, T. D., J. W. Brewer, R. B. Corley. 1992. Direct evidence that J chain regulates the polymeric structure of IgM in antibody-secreting B cells. J. Biol. Chem. 267:18002.[Abstract/Free Full Text]
  13. 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, J. G. Seidman. 1995. Altered hepatic transport of immunoglobulin A in mice lacking the J chain. J. Exp. Med. 182:1905.[Abstract/Free Full Text]
  14. Fallgreen-Gebauer, E., W. Gebauer, A. Bastian, H. D. Kratzin, H. Eiffert, B. Zimmermann, M. Karas, N. Hilschmann. 1993. The covalent linkage of secretory component to IgA: structure of sIgA. Biol. Chem. Hoppe-Seyler 374:1023.[Medline]
  15. Fra, A. M., C. Fagioli, D. Finazzi, R. Sitia, C. M. Alberini. 1993. Quality control of ER synthesized proteins: an exposed thiol group as a three-way switch mediating assembly, retention and degradation. EMBO J. 12:4755.[Medline]
  16. Guenzi, S., A. M. Fra, A. Sparvoli, P. Bet, M. Rocco, R. Sitia. 1994. The efficiency of cysteine-mediated intracellular retention determines the differential fate of secretory IgA and IgM in B and plasma cells. Eur. J. Immunol. 24:2477.[Medline]
  17. Sørensen, V., I. B. Rasmussen, L. Norderhaug, I. Natvig, T. E. Michaelsen, I. Sandlie. 1996. Effect of the IgM and IgA secretory tailpieces on polymerization and secretion of IgM and IgG. J. Immunol. 156:2858.[Abstract]
  18. Smith, R. I., S. L. Morrison. 1994. Recombinant polymeric IgG: an approach to engineering more potent antibodies. BioTechnology 12:683.[Medline]
  19. Smith, R. I., M. J. Coloma, S. L. Morrison. 1995. Addition of a µ-tailpiece to IgG results in polymeric antibodies with enhanced effector functions including complement-mediated cytolysis by IgG4. J. Immunol. 154:2226.[Abstract]
  20. Kunkel, T. A., J. D. Roberts, R. A. Zakour. 1987. Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154:367.[Medline]
  21. Norderhaug, L., T. Olafsen, T. E. Michaelsen, I. Sandlie. 1997. Versatile vectors for transient and stable expression of recombinant antibody molecules in mammalian cells. J. Immunol. Methods 204:77.[Medline]
  22. Michaelsen, T. E., A. Aase, C. Westby, I. Sandlie. 1990. Enhancement of complement activation and cytolysis of human IgG3 by deletion of hinge exons. Scand. J. Immunol. 32:517.[Medline]
  23. Michaelsen, T. E., P. Garred, A. Aase. 1991. Human IgG subclass pattern of inducing complement-mediated cytolysis depends on antigen concentration and to a lesser extent on epitope patchiness, antibody affinity and complement concentration. Eur. J. Immunol. 21:11.[Medline]
  24. Randall, T. D., L. B. King, R. B. Corley. 1990. The biological effects of IgM hexamer formation. Eur. J. Immunol. 20:1971.[Medline]
  25. Chapuis, R. M., M. E. Koshland. 1975. Linkage and assembly of polymeric IgA immunoglobulins. Biochemistry 14:1320.[Medline]
  26. Yang, C., H. Kratzin, H. Gotz, N. Hilschmann. 1979. Rule of antibody structure: primary structure of a human monoclonal IgA-immunoglobulin (myeloma protein Tro). VII. Purification and characterization of the disulfide bridges. Hoppe-Seyler’s Z. Physiol. Chem. 360:1919.[Medline]
  27. Chintalacharuvu, K. R., A. S. Tavill, L. N. Louis, J. P. Vaerman, M. E. Lamm, C. S. Kaetzel. 1994. Disulfide bond formation between dimeric immunoglobulin A and the polymeric immunoglobulin receptor during hepatic transcytosis. Hepatology 19:162.[Medline]
  28. Pumphrey, R.. 1986. Computer models of the human immunoglobulins: binding sites and molecular interactions. Immunol. Today 7:206. (Abstr.).
  29. Efimov, A. V.. 1997. Structural trees for protein superfamilies. Proteins 28:241.[Medline]
  30. Perkins, S. J., A. S. Nealis, B. J. Sutton, A. Feinstein. 1991. Solution structure of human and mouse immunoglobulin M by synchrotron x-ray scattering and molecular graphics modelling: a possible mechanism for complement activation. J. Mol. Biol. 221:1345.[Medline]
  31. Feinstein, A., E. A. Munn, N. E. Richardson. 1971. The three-dimensional conformation of M and A globulin molecules. Ann. NY Acad. Sci. 190:104.[Medline]
  32. Taylor, B., J. F. Wright, S. Arya, D. E. Isenman, M. J. Shulman, R. H. Painter. 1994. C1q binding properties of monomer and polymer forms of mouse IgM µ-chain variants, Pro544Gly and Pro434Ala. J. Immunol. 153:5303.[Abstract]
  33. Chen, F. H., S. K. Arya, A. Rinfret, D. E. Isenman, M. J. Shulman, R. H. Painter. 1997. Domain-switched mouse IgM/IgG2b hybrids indicate individual roles for Cµ2, Cµ3, and Cµ4 domains in the regulation of the interaction of IgM with complement C1q. J. Immunol. 159:3354.[Abstract]
  34. Feinstein, A., N. Richardson, M. J. Taussig. 1986. Immunoglobulin flexibility in complement activation. Immunol. Today 7:169. (Abstr.).
  35. Arya, S., F. Chen, S. Spycher, D. E. Isenman, M. J. Shulman, R. H. Painter. 1994. Mapping of amino acid residues in the Cµ3 domain of mouse IgM important in macromolecular assembly and complement-dependent cytolysis. J. Immunol. 152:1206.[Abstract]
  36. Duncan, A. R., G. Winter. 1988. The binding site for C1q on IgG. Nature 332:738.[Medline]
  37. Boackle, R. J., B. J. Johnson, G. B. Caughman. 1979. An IgG primary sequence exposure theory for complement activation using synthetic peptides. Nature 282:742.[Medline]
  38. Kerr, M. A.. 1990. The structure and function of human IgA. Biochem. J. 271:285.[Medline]
  39. Natvig, I. B., F. E. Johansen, T. W. Nordeng, G. Haraldsen, P. Brandtzaeg. 1997. Mechanism for enhanced external transfer of dimeric IgA over pentameric IgM: studies of diffusion, binding to the human polymeric Ig receptor, and epithelial transcytosis. J. Immunol. 159:4330.[Abstract]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
Z. Wei, Q. Wu, L. Ren, X. Hu, Y. Guo, G. W. Warr, L. Hammarstrom, N. Li, and Y. Zhao
Expression of IgM, IgD, and IgY in a Reptile, Anolis carolinensis
J. Immunol., September 15, 2009; 183(6): 3858 - 3864.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
D. M. Czajkowsky and Z. Shao
The human IgM pentamer is a mushroom-shaped molecule with a flexural bias
PNAS, September 1, 2009; 106(35): 14960 - 14965.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
A. Ghumra, J.-P. Semblat, R. S. McIntosh, A. Raza, I. B. Rasmussen, R. Braathen, F.-E. Johansen, I. Sandlie, P. K. Mongini, J. A. Rowe, et al.
Identification of Residues in the C{micro}4 Domain of Polymeric IgM Essential for Interaction with Plasmodium falciparum Erythrocyte Membrane Protein 1 (PfEMP1)
J. Immunol., August 1, 2008; 181(3): 1988 - 2000.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
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]

Home page
 J. Biol. Chem.Home page
J. Arthos, C. Cicala, T. D. Steenbeke, T.-W. Chun, C. D. Cruz, D. B. Hanback, P. Khazanie, D. Nam, P. Schuck, S. M. Selig, et al.
Biochemical and Biological Characterization of a Dodecameric CD4-Ig Fusion Protein. IMPLICATIONS FOR THERAPEUTIC AND VACCINE STRATEGIES
J. Biol. Chem., March 22, 2002; 277(13): 11456 - 11464.
[Abstract] [Full Text] [PDF]

Home page
 Int ImmunolHome page
V. Sorensen, I. B. Rasmussen, V. Sundvold, T. E. Michaelsen, and I. Sandlie
Structural requirements for incorporation of J chain into human IgM and IgA
Int. Immunol., January 1, 2000; 12(1): 19 - 27.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
E. M. Yoo, M. J. Coloma, K. R. Trinh, T. Q. Nguyen, L.-U. C. Vuong, S. L. Morrison, and K. R. Chintalacharuvu
Structural Requirements for Polymeric Immunoglobulin Assembly and Association with J Chain
J. Biol. Chem., November 19, 1999; 274(47): 33771 - 33777.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sørensen, V.
Right arrow Articles by Sandlie, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sørensen, V.
Right arrow Articles by Sandlie, I.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS