| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Instituto de Investigaciones Bioquímicas, Fundación Campomar, and
Cátedra de Immunología, Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Argentina;
Center for Advanced Research in Biotechnology, University of Maryland Biotechnology Institute, Rockville, MD 20850; and
§
Département d’Immunologie, Institut Pasteur, Paris, France,
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
More recent studies using protein Ags indicated a different situation. For example, early mAbs to the vesicular stomatitis virus (VSV) coat glycoprotein displayed high affinities and fast association rates that did not increase with prolonged immunization or change of dose (10). Similarly, mAbs against hen egg lysozyme (HEL), obtained after prolonged immunizations did not show an increase in average avidity (11).
We present the binding kinetics and equilibrium constants of the Fab
derived from 23 specific anti-HEL IgG1 
mAbs representative of
short-term and long-term responses. We have also analyzed Fab' derived
from specific anti-HEL IgM mAbs obtained late in the response. We
report two independently derived IgG mAbs that are the result of
identical or nearly identical somatic recombination events
(VH+D+JH, VL+JL) in
which mutations contribute to affinity increase. No systematic overall
change is observed between mAbs of short- and long-term responses
either in the equilibrium binding constants or in kinetic
association constants. The results are consistent with the view that
affinity maturation is a relatively fast process, closely related to
the early IgM-IgG class switch.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The short-term and long-term response mAbs (all IgG1 ) of the
D and F series were produced from ascites in BALB/c mice, as previously
described (12, 13). Briefly, 14- to 15-wk-old BALB/c mice were first
immunized s.c. or i.p. with 100 µg of HEL in CFA. Subsequent
injections (except for D10.4-D74.3, see Table I
) of Ag (50 or 100 µg) were in IFA.
Three days after an i.v. booster injection of 50 µg of HEL in PBS,
splenic lymphocytes were used for cell fusion. The immunization
schedules are summarized in Table I.
|
Preparation of Fabs from IgG1 and IgM anti-HEL mAbs
mAbs were purified from ascites, and their Fabs were prepared by papain hydrolysis and purified as previously reported (12). An additional step of purification after DEAE chromatography was included to ensure that only univalent Fab were present; Fabs were applied to a Superose 12 column (Pharmacia, Uppsala, Sweden), and the peaks corresponding to the 50-kDa Fab were collected, concentrated, and used for biosensor analysis.
Anti-HEL IgMs were purified from ascites by dialyzing against water (14). After centrifugation, the pellet was redissolved in 0.2 M NaCl and 0.1 M Na acetate (pH 4.6), to a final concentration of 10 mg/ml. Pepsin was added to a final molar ratio of 1/25, and the IgM was digested for 24 h at 37°C. The sample was diluted 1/10 in 50 mM 2-morpholine ethanesulfonic acid (pH 5.5) and applied to a MonoS (Pharmacia) column. A linear gradient was created using 50 mM 2-morpholine ethanesulfonic acid and 1 M NaCl (pH 5.5). The Fab' eluted at 0.2–0.3 M NaCl. Pooled peaks from several runs were concentrated, reduced with 10 mM 2-ME, and alkylated with 50 mM iodoacetamide at room temperature. The iodoacetamide-treated Fab' was purified by gel exclusion chromatography in a Superose 12 column, in which it eluted as a sharp peak of about 80 kDa. The Fab' was concentrated to 0.5 mg/ml and analyzed by SDS-PAGE (12.5% acrylamide concentration) with and without 2-ME in the sample buffer.
Measurement of equilibrium and kinetic association constants
Affinities were measured using a BIAcore instrument (BIAcore, Piscataway, NJ) by the detection of surface plasmon response from the interacting molecules adsorbed onto a specially prepared surface (15). All experiments were performed at 25°C. Samples were dissolved in HBS buffer (150 mM NaCl, 3.4 mM EDTA, 0.005% (w/v) surfactant P-20 (BIAcore), and 10 mM HEPES, pH 7.4). The buffer flow rate was 5 µl/min. HEL was coupled to a CM5 sensor chip using the amine coupling kit (BIAcore) at a concentration of 30 µg/ml in 10 mM Na acetate (pH 4.8). After a binding assay, the chip was regenerated by washing twice with 5 µl of 10 mM HCl. The Fabs were diluted in running buffer to concentrations ranging from 10-8 to 10-6 M. Runs at different concentrations were performed to obtain data for the association and dissociation rates. Dissociation rates were measured washing the chip surface with a 10 µg/ml solution of HEL in HBS to avoid rebinding of dissociated Fab. Data were analyzed as previously described (16, 17). The Ka was calculated from kinetic association (kon) and dissociation (koff) constants as: Ka = kon/koff.
Affinity analysis by ultracentrifugation
Equilibrium sedimentation of the Fab' from IgM HEL-2 and its complex with HEL was performed in a Beckman XL-A Optima analytical ultracentrifuge (Spinco Division, Beckman Instruments, Palo Alto, CA) using a four-hole An-55 rotor. All experiments were performed at 25°C at a rotor speed of 18,000 rpm. The molar extinction coefficients used were: Fab'-HEL-2, 112,500 M-1 cm-1; and HEL, 37,200 M-1 cm-1. The concentration distributions of the samples at sedimentation equilibrium were acquired as an average of 25 absorbance measurements (280 nm) at each radial position, with nominal spacing of 0.001 cm between radial positions. Samples were prepared by dialysis against 50 mM Tris-HCl (pH 7.5) at a concentration of 4 µM for each species, either alone or as a 1/1 molar mixture. Partial specific volumes for the Fab' and HEL were assumed to be 0.73 ml/mg. The data obtained were analyzed as previously described (18). Errors in the equilibrium dissociation constants (Kd) were estimated to be <12% based on the fit of parameter values.
Cloning and sequencing of VH and VL F10.6.6.
mRNA was extracted and purified from approximately 5 x 107 hybridoma cells using a Fast-track kit (Invitrogen, Carlsbad, CA). cDNA was synthesized using the Copykit (Invitrogen) with an oligo(dT) primer. The cDNA was used in PCR cycles as a template with degenerate primers for conserved regions in the N-terminal and the J regions of VH and VL, as previously described (19): VH1BACK and VH1FOR-2 were used for VH, VK2BACK, and VK4FOR-1 and VK4FOR-2 were used for VL. The PCR products were cloned using the TA cloning kit (Invitrogen). Several clones for each chain were sequenced to verify the absence of artifacts arising from PCR amplification.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To study the affinity of Abs during the immune response to HEL, 23
specific IgG1 mAbs were analyzed. Twelve of the mAbs were from shorter
responses, and 11 were from longer term responses (see Table I
).
Affinity measurements by surface plasmon response using whole IgG
produce, in some cases, a significant increase in the observed
Ka, which correlates with a slower kinetic
dissociation constant (koff). This effect
appears to be different for each mAb (see Table II), a fact that could distort the values
observed for different mAbs. To circumvent this problem, the intrinsic
association constants were measured using purified Fab. It should be
noted that the affinity constants of many of the mAbs reported here
have been previously determined (20, 21) by different techniques, such
as titration calorimetry, fluorescence quenching, and ELISA. Comparable
Ka were obtained with all these techniques, thus
validating the values obtained in this work. Fig. 1 shows the results obtained with Fab
from the 23 IgG1 anti-HEL mAbs. The equilibrium affinity constants
(Ka) are shown in Fig. 1a for the
shorter term and the longer term response groups. Both groups had
similar distributions of affinities, ranging from 107 to
1010 M-1. Affinities of mAbs from the
shorter term responses ranged from 1.1 x 107
M-1 (mAb D136.11) to 3.2 x 109
M-1 (mAb F2.22.1). Long- term response affinities ranged
from 6.5 x 107 M-1 (mAb H17) to 1.4
x 1010 M-1 (mAb F10.6.6). The mean
affinity values for the mAbs of the shorter term and the
long-term responses were 5.7 x 108 and 1.6 x
109 M-1, respectively. Thus, there was no
significant difference in the affinities between these two populations.
|
|
b). For example, mAb F2.22.1, whose
Ka was 3.2 x 109
M-1, had a lower kon than mAb
D74.3, whose Ka was 1.1 x 108
M-1 (kon of 1.4 x
105 M-1 s-1 for F222 vs 4.3
x 105 M-1 s-1 for D74). The
lower value for both populations was 1.0 x 104
M-1 s-1 (mAb D44.1), and the higher values
were 4.4 x 106 and 3.4 x 106
M-1 s-1 for mAbs D11.15 (short-term response)
and F10.6.6 (longer term response), respectively. Detailed analysis of two closely related mAbs
We decided to study the mAb F10.6.6 (Ka = 1.4 x 1010 M-1) derived from a long-term immunization to compare it to D44.1 (Ka = 1.0 x 107 M-1), derived from a short-term immunization. F10.6.6 bound an epitope of HEL that, by immunochemical tests, partially or totally overlapped that bound by mAb D44.1 (12, 13). Other properties, such as the close crystallization conditions for Fabs from these Abs, suggested that they may be related (22).
The rearranged genes of mAb F10.6.6 were cloned and sequenced. Indeed,
F10.6.6 proved to be very close to D44.1. Their VL domains
were about 96% identical in sequence, with differences at only seven
positions: two in CDR2, one in CDR3, and 4 non-CDR (data not shown).
The VH domains of both mAbs seemed to derive from the same
VH and JH gene segments
and, very likely, the same D segment and were more
diversified by somatic mutations. Fig. 2
shows the aligned sequences of VHD44.1,
VHF10.6.6, and VHHyHEL5 compared with that of
the closest germline V gene (mouse VH gene
J00530) found using the BLAST algorithm (23).
|
Analysis of IgM mAbs
The appearance and characteristics of anti-HEL IgMs along the
immune response were also studied. Anti-HEL IgM mAbs were obtained
after secondary and long-term immunizations. We analyzed one of these
clones, HEL-2. Competitive ELISA showed that this IgM reacts with
an epitope that overlaps with that recognized by D44.1 and F10.6.6. The
monovalent HEL-2 Fab' had an apparent molecular mass of 94 kDa by
ultracentrifugal analysis (see Materials and Methods).
Analysis by SDS-PAGE (Fig. 3, inset) gave
an apparent molecular mass of 85,000 kDa and under reducing conditions
showed the presence of an L chain and a fragment of the H chain of
approximately 50 kDa. Fab' HEL-2 binds HEL with an affinity constant of
7 x 104 M-1, as determined by
equilibrium sedimentation analysis (Fig. 3).
|
IgM mAbs F2.4.6, F3.10.8, and F8.6.1, derived from a long- term
response (Table I), were also analyzed. Their Fab' were similar to
HEL-2 Fab' by SDS-PAGE and had anti-HEL activity by indirect ELISA
at high HEL concentrations (results not shown). However, these Fab' did
not show any signal when assayed with immobilized HEL by BIAcore
analysis, indicating that their affinity constants are even lower than
that of HEL-2 (< 7 x 104 M-1).
The results suggest that the affinities of IgM Abs elicited during the anti-HEL response are in the 105 M-1 range or lower. This very low intrinsic affinity is probably compensated by the great avidity resulting from the multimeric assembly of IgMs.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
indicates a problem that can affect the determination of affinities
when using whole IgGs instead of Fabs. The use of Ag coupled to a solid
phase (as in the BIAcore technique) can yield affinities and
association rates comparable to those obtained in solution studies.
Indeed, the values of Ka and
kon presented in this work are in agreement with
previous measurements (for mAbs D1.3, D11.15, F9.13.7, D44.1, and
F10.6.6) (20, 21) that were made using different experimental
techniques, such as fluorescence quenching, titration microcalorimetry,
and stop-flow kinetics, all in solution, and with solid phase ELISA.
Thus, our work differs from that on VSV (10) and earlier work on HEL
(11) in that we have measured directly the association rates and
equilibrium constants, and we have used Fabs. HEL is a good model Ag for this work, since its three-dimensional structure and those of several complexes with specific mAbs are known (reviewed in Refs. 24 and 25). These structural studies and epitope mapping by immunochemical techniques (11, 12, 26) have contributed much information about its antigenic properties. The three-dimensional structures of Ag complexes with the mAbs D1.3, D44.1, D11.15, and F9.13.7 analyzed in this report have been determined (27, 28, 29, 30) defining the epitopes bound by the Abs. These epitopes are unique for each of these mAbs, with only a small overlap between those of D1.3 and D11.15. The epitopes defined by D44.1 and F9.13.7 overlap almost completely with those of HyHEL-5 and HyHEL-10, respectively (11, 31, 32). The affinity of HyHEL-5 (32) is very close to that of F10.6.6. Similarly, HyHEL-10 reacts with an epitope closely overlapping that defined by F9.13.7 but with a somewhat higher affinity: Ka = 4.5 x 1010 M-1 (33) vs Ka = 1.4 x 109 M-1 for F9.13.7.
The affinity values observed in the anti-HEL mAbs reported here go from a minimum Ka = 1.1 x 107 M-1 to a maximum Ka = 1.4 x 1010 M-1. In structural terms this is a small difference that could, in principle, be explained by a single amino acid replacement capable of making an additional hydrogen bond, with a free energy of association of about 4 kcal M-1. Somatic mutations have been shown to generate Abs of higher affinity in a number of experimental systems (2, 3, 4, 5, 6, 7, 8, 9). However, as in the anti-HEL mAbs, the differences in the free energies of association are small. Thus, it is not clear that all the amino acid substitutions generated by the mutations are necessary for the observed increase in affinity.
We detected a close structural and functional similarity between mAbs
D44.1 and F10.6.6, which are the products of short- and long-term
immunizations, respectively. F10.6.6 is very close to D44.1. Their
VL domains are about 96% identical in sequence, with
differences in four framework residues, two CDR2 residues, and one CDR3
residue (data not shown). The VH domains of both Abs derive
from the same VH and JH
gene segments (Fig. 2). There are two amino acid residue differences in
the part of the CDR3 normally derived from the D segment.
These differences could be due to somatic mutations or to N
junction diversity. A model of the combining site of F10.6.6 based on
the structure of the highly homologous D44.1 fails to give clues to the
structural determinants of the higher affinity for F10.6.6. Since a
structure of the F10.6.6 complex with HEL is not available, we cannot
judge whether the amino acid changes in CDR3 relative to those in D44.1
are responsible for the observed difference in affinity between the two
Abs. As stated above, this difference could, for example, arise from a
hydrogen bond resulting from a single amino acid replacement. If we
accept that the VH genes of F10.6.6 and D44.1
derive from the germline gene sequence J00530 (Fig. 2), then we
conclude that F10.6.6 shows more mutations that D44.1, interspersed in
CDR and non-CDR regions, consistent with the idea that somatic
mutations confer increased affinity to the Ag.
For the study of the physicochemical parameters of affinity maturation, hapten systems present the apparent advantage that Abs directed always to the same ligand, the hapten, can be selected during the course of immune responses. In contrast, Abs raised against a protein Ag are directed to different epitopes, and thus mAbs from early and late responses may not be strictly comparable. However, working with haptens presents other limitations. One of them is that the response tends to be dominated by a single germline gene or a few germline genes, whereas in the response to Ags many genes are expressed to provide Abs reacting with different epitopes. Another limitation is that a conjugated hapten is only part of an antigenic determinant. For example, in the hapten phOx, the area of contact between the hapten and an Ab combining site is of about 200 Å2 (34). In most protein-Ab complexes that area is 650-1200 Å2 (24, 25), providing for more extensive interactions with the Ab combining site. In addition, haptens conjugated to a protein carrier will be part of a protruding structure that will be bound by many Abs as a central feature in an environment closed to solvent, thus with a favorable dielectric constant and high affinity. Other Abs binding the protein carrier and all or part of the hapten at the edge of the interface will have a much lower affinity for the hapten and may be missed by experimental studies focusing on haptens. Moreover, in both types of anti-hapten Abs it is not the change in the total free energy of binding to the antigenic determinant that is measured but, rather, that of a selected part of it. Any conclusion derived for hapten binding is thus limited, in that the affinity for it may increase with prolonged immunizations while that for the total antigenic determinant may not.
Physicochemical factors such as diffusion coefficients suggest a limit
for the rates of association between Ab and Ag. Although less precise,
a limit can also be envisaged for the dissociation rates (35) leading
to an upper limit of about 1010 M-1 for the
Ka of their reactions. A study of cellular
stimulation and Ab induction (36) concludes that Abs will have upper
and lower limits of affinity, 1010 M-1 and
107 M-1, respectively. While lower affinities
may not be sufficient to stimulate clones of Ab-producing cells,
affinities >1010 M-1 may not confer any
selective advantage in terms of cell proliferation. There is also a
structural interpretation for the upper limit of
Ka. The packing of side chains in the Ag-Ab
interfaces is not as tight as in the interior of a folded protein, and
although in principle a tighter packing leading to optimal noncovalent
binding could be achieved, in practice this may be infrequent. The
upper and lower limits suggested by physicochemical (35) and cellular
stimulation studies (36) are indeed observed in the anti-HEL mAbs
(Fig. 1).
The study of mAbs to the VSV protein (10) showed that neutralizing Abs
of high affinity against a pathogenic virus arise as early as day 6,
with no overall change during a long-term response. Since no affinity
maturation could be observed, and many of the Abs were in germline
conformation, it was postulated that the lack of a significant
difference in affinities (and in association rates) in early and late
Abs may be due to the fact that the VSV infections are lethal, and
consequently, germline gene information is tuned to a rapid high
affinity response to this Ag. Although the anti-HEL mAbs reported
here were obtained after longer immunizations (
15 days), and for at
least two of the anti-HEL mAbs there were many mutations, the final
result is similar to that with VSV Abs even though the Ag HEL is not of
viral or bacterial origin. In agreement with Roost et al. (10) and
Newman et al. (11), our observations indicate that obtaining high
affinities does not depend on prolonged immunizations. The observation
that IgMs occur even after prolonged Ag challenge suggests that each
immunization can produce a de novo response with the recruitment of new
clones of Ab-producing cells and the implied repertoire shift, rather
than by mutations in memory B cells. Furthermore, the observation that
different immunizations give rise to the expression of the same
VH gene (D44.1, F10.6.6, and HyHEL5), binding a
nearly identical epitope, suggests defined paths of gene expression and
recombination of genetic segments for the production of Ag-specific
Abs.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Roberto J. Poljak, Center for Advanced Research in Biotechnology, 9600 Gudelsky Dr., Rockville, MD 20850-3479. E-mail address:
3 Abbreviations used in this paper: phOx, 2-phenyl-5-oxazolone; kon, association rate constant; VSV, vesicular stomatitis virus; HEL, hen egg lysozyme; Ka, equilibrium association constant or affinity; koff, dissociation rate constant; CDR, complementarity-determining region.
Received for publication January 7, 1999. Accepted for publication March 3, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Mol. Immunol. 32:147.[Medline]
2a variable region. Cell 24:625.[Medline]
This article has been cited by other articles:
|
|
 |
|
  T. R. Poulsen, P.-J. Meijer, A. Jensen, L. S. Nielsen, and P. S. Andersen Kinetic, Affinity, and Diversity Limits of Human Polyclonal Antibody Responses against Tetanus Toxoid J. Immunol., September 15, 2007; 179(6): 3841 - 3850. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. De Genst, K. Silence, M. A. Ghahroudi, K. Decanniere, R. Loris, J. Kinne, L. Wyns, and S. Muyldermans Strong in Vivo Maturation Compensates for Structurally Restricted H3 Loops in Antibody Repertoires J. Biol. Chem., April 8, 2005; 280(14): 14114 - 14121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Cauerhff, F. A. Goldbaum, and B. C. Braden Structural mechanism for affinity maturation of an anti-lysozyme antibody PNAS, March 9, 2004; 101(10): 3539 - 3544. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Cabrera, M. Muniz, J. Hidalgo, L. Vega, M. E. Martin, and A. Velasco The Retrieval Function of the KDEL Receptor Requires PKA Phosphorylation of Its C-Terminus Mol. Biol. Cell, October 1, 2003; 14(10): 4114 - 4125. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-F. Lu, M. Singh, and J. Cerny Canonical germinal center B cells may not dominate the memory response to antigenic challenge Int. Immunol., May 1, 2001; 13(5): 643 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Jirholt, L. Strandberg, B. Jansson, E. Krambovitis, E. Soderlind, C. A.K. Borrebaeck, R. Carlsson, L. Danielsson, and M. Ohlin A central core structure in an antibody variable domain determines antigen specificity Protein Eng. Des. Sel., January 1, 2001; 14(1): 67 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Brown, M. A. Schumacher, G. D. Wiens, R. G. Brennan, and M. B. Rittenberg The Structural Basis of Repertoire Shift in an Immune Response to Phosphocholine J. Exp. Med., June 19, 2000; 191(12): 2101 - 2112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Cerutti, J. M. Centeno, F. A. Goldbaum, and G. de Prat-Gay Generation of Sequence-specific, High Affinity Anti-DNA Antibodies J. Biol. Chem., April 13, 2001; 276(16): 12769 - 12773. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. F. Ochsenbein, D. D. Pinschewer, S. Sierro, E. Horvath, H. Hengartner, and R. M. Zinkernagel Protective long-term antibody memory by antigen-driven and T help-dependent differentiation of long-lived memory B cells to short-lived plasma cells independent of secondary lymphoid organs PNAS, November 21, 2000; 97(24): 13263 - 13268. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) immune responses (see
Materials and Methods and Table I
