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
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 arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Feng, J. Q.
Right arrow Articles by Gerhard, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Feng, J. Q.
Right arrow Articles by Gerhard, W.

 Previous Article  |  Next Article 

Journal of Virology, February 2002, p. 1369-1378, Vol. 76, No. 3
0022-538X/01/$04.00+0     DOI: 10.1128/JVI.76.3.1369-1378.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Complement Component C1q Enhances the Biological Activity of Influenza Virus Hemagglutinin-Specific Antibodies Depending on Their Fine Antigen Specificity and Heavy-Chain Isotype

Jing Qi Feng, Krystyna Mozdzanowska, and Walter Gerhard*

The Wistar Institute, Philadelphia, Pennsylvania 19104-4268

Received 2 August 2001/ Accepted 24 October 2001


arrow
ABSTRACT
 
We have previously observed that selected influenza virus hemagglutinin (HA)-specific monoclonal antibodies (MAbs) with poor virus-neutralizing (VN) activity in vitro exhibited greatly enhanced VN activity in vivo after administration to SCID mice. The same Abs displayed improved VN activity also when tested in vitro in the presence of noninactivated serum from SCID mice. To identify Ab-dependent properties and serum components that contributed to enhancement of Ab activity, we screened a large panel of HA-specific MAbs for hemagglutination inhibition (HI) in the presence of noninactivated serum from naive mice (NMS). We found that HI activity was enhanced by NMS depending on the Ab’s fine specificity (antigenic region Cb/E > Ca/A,D > Sa,Sb/B), its heavy-chain isotype (immunoglobulin G2 [IgG2] > IgG3; IgG1 and IgM negative), and to some extent also on its derivation (primary response > memory response). On average, the HI activity of Cb/E-specific MAbs of the IgG2 isotype isolated from the primary response was enhanced by 20-fold. VN activity was enhanced significantly but less strongly than HI activity. Enhancement (i) was destroyed by heat inactivation (30 min, 56°C); (ii) did not require C3, the central complement component; (iii) was abolished by treatment of serum with anti-C1q; and (iv) could be reproduced with purified C1q, the binding moiety of C1, the first complement component. We believe that this is the first description of a direct C1q-mediated enhancement of antiviral Ab activities.


arrow
INTRODUCTION
 
It is a frequently encountered phenomenon that the therapeutic activity of passively administered antibodies (Abs) in vivo deviates significantly from predictions made on the basis of their activity measured in vitro (2, 20, 35). An example is our previous finding that a group of influenza virus hemagglutinin (HA)-specific monoclonal Abs (MAbs) which lacked substantial virus-neutralizing (VN) activity in vitro nevertheless were quite effective in protecting against infection in vivo when given prophylactically to SCID mice (24). For instance, one Ab from this group, H35-C12, exhibited a VN activity (Ab concentration at which 50% of Madin-Darby canine kidney cell [MDCK cell] microcultures were protected from infection by {approx}30 50% tissue culture infective doses [TCID50] of PR8) in vitro of 2 µg/ml and protected 50% of SCID mice against infection by a similar dose of virus at a concentration in serum of {approx}8 µg/ml. The prophylactic protective activity in vivo (referred to as VN in vivo because it operates similarly to VN in vitro by preventing initiation of infection by the virus inoculum) was unexpectedly high compared to that of another HA-specific MAb (H36-4), which exhibited an {approx}500-fold-higher VN activity in vitro (0.004 µg/ml [combined data from a previous study {24} and the present study]) but only an {approx}7-fold-higher VN activity in vivo ({approx}1.2 µg/ml of serum). As both of these MAbs were of the same isotype (immunoglobulin G2a [IgG2a]), exhibited similar half-lives in vivo, and were expected to transude at the same rate from serum into the respiratory tract lining fluid, the different ratios of VN activity in vivo to VN activity in vitro ({approx}4 for H35-C12 and {approx}300 for H36-4) indicated that some factors in vivo either enhanced VN activity of H35-C12 or inhibited VN activity of H36-4. The former possibility was supported by the finding that the VN activity of H35-C12, but not that of H36-4, was strongly enhanced when tested in vitro in the presence of noninactivated serum from SCID mice (24).

Here, we describe additional investigations of this serum-dependent enhancement of antiviral Ab activity measured in hemagglutination inhibition (HI) and VN assays. We found that enhancement of HI activity by noninactivated naive mouse serum (NMS) (i) was dependent both on the Ab’s heavy-chain isotype and its specificity for certain regions on the HA molecule; (ii) was mediated by C1q, the protein that provides specificity to the first complement component; and (iii) did not require the presence of C3, the central component of the complement system. VN activity was less strongly enhanced by C1q than HI activity and appeared to be modulated by additional serum factors, as NMS from SCID mice was significantly more effective in enhancing VN but not HI activity than NMS from immunocompetent BALB/c and C57BL/6 mice.


arrow
MATERIALS AND METHODS
 
Virus. The influenza virus strain A/PR/8/34(H1N1) was originally obtained from Mt. Sinai Hospital (New York, N.Y.) and is referred to as PR8. B/Lee is an influenza virus type B strain. Influenza virus types A and B are immunologically not cross-reactive. The viruses were propagated by inoculation of {approx}5 x 103 TCID50 (measured in MDCK cell culture) into the allantoic cavities of 10-day-old embryonated hen’s eggs, and allantoic fluid was harvested after 3 days of incubation at 35°C. Aliquots of infectious allantoic fluid were stored at -60 to -70°C. Infectious stocks typically contained approximately 109 TCID50/ml. Virus used in enzyme-linked immunosorbent assays (ELISAs) and HI assays was purified from allantoic fluid by differential centrifugation and banding in a sucrose gradient. Purified virus was quantitated by protein content and HA activity (see below). It typically contained 7 ng of protein per HA unit (HAU).

Media and solutions. ISC-CM consists of Iscove’s Dulbecco’s medium (Life Technologies, Gaithersburg, Md.) supplemented with 0.05 mM 2-mercaptoethanol, 0.005 mg of transferrin (Sigma, St. Louis, Mo.) per ml, 2 mM glutamine (JRH Biosciences, Lenexa, Kans.), and 0.05 mg of gentamicin (Mediatech, Herndon, Va.) per ml. ISC-CM was further supplemented with fetal calf serum (FCS) (HyClone Laboratories, Logan, Utah), bovine serum albumin (BSA) (Sigma), trypsin (Whittaker Bioproducts Inc.), or NMS at specified concentrations. PBSN is phosphate-buffered saline containing 3 mM NaN3. PBSN(Ca,Mg) additionally contained 0.9 mM CaCl2 and 0.5 mM MgCl2.

Antibodies and biologic reagents. Hybridomas were grown in ISC-CM-5% FCS, and Abs were purified by adsorption to and elution from protein A-agarose (Pierce, Rockford, Ill.) using ImmunoPure(A) IgG binding buffer and ImmunoPure IgG elution buffer, respectively (both from Pierce). Goat anti-human C1q (anti-huC1q) was obtained from Sigma, and purified huC1q was obtained from Calbiochem (La Jolla, Calif.). NMS was prepared by exsanguination of anesthetized mice (70 mg of ketamine and 7 mg of xylazine per kg, given intraperitoneally). Blood was allowed to clot at room temperature (45 to 60 min) and then kept for 60 min on ice before serum was harvested. Serum was usually stored at 4°C and used within the next few weeks. For longer storage it was kept frozen. Sera were heat inactivated for 30 min at 56°C.

Determination of protein concentration. Protein concentrations in purified virus and Ab preparations were determined by protein assay (Bio-Rad Laboratories, Hercules, Calif.) using bovine Ig (Pierce) as a standard. Protein concentrations in purified Ab preparations were additionally estimated by UV light absorption at wavelengths of 260, 280, and 320 nm (GeneQuant; Pharmacia, Cambridge, United Kingdom) and computed according to the formula 1.55 x (A280 - A320) - 0.76 x (A260 - A320), expressed in milligrams per milliliter. The means of both determinations were used as the protein concentrations in Ab preparations.

HA assay. The HA assay was done as described previously (30). In brief, virus-containing samples (25 µl) were serially diluted in PBSN (25 µl) in round-bottom microtiter plates. Twenty-five microliters of a 1% (vol/vol) suspension of chicken red blood cells (CRBC) in PBSN was added to each well and mixed, and the pattern of CRBC sedimentation was recorded after letting the plates sit for 35 min at room temperature. The sample dilution showing partial agglutination (ring-like sedimentation pattern) was read as the end point, and the inverse of the corresponding dilution was recorded as the HA titer. The latter, multiplied by 2, was taken as the sample’s virus concentration in HAU per milliliter.

HI assay. The HI assay was done essentially as described previously (30) except for performing the assay in distinct diluents. In brief, Abs were serially diluted in 25 µl of PBSN(Ca,Mg) in round-bottom microtiter plates. Twenty-five microliters of PBSN(Ca,Mg), unmodified or supplemented as indicated with NMS or C1q and containing a previously determined amount of HA corresponding to four agglutinating doses of PR8 under the specific assay conditions, was added to each well and mixed, and the plates were incubated for 60 min at room temperature. Fifty microliters of 1% CRBC was then added to each well and mixed, and the patterns of CRBC sedimentation were recorded after letting the plates sit for 40 min at room temperature. Partial agglutination, at which the Ab inhibited three out of the four agglutinating doses, was read as the titration end point. Some batches of NMS, usually only after heat inactivation, modified the agglutination pattern such that the standard type of end point reading became difficult and unreliable. In those instances, end points were read as the last Ab dilution at which the CRBC pellet flowed down freely in a tear-like pattern when the assay plate was held at a 45 to 60° angle for {approx}0.5 min. All diluents were also tested for HI activity in the absence of Ab.

VN assay. The standard VN assay was initially done as described previously (30). It was subsequently modified as follows (nonstandard VN assay). A freshly trypsinized MDCK cell suspension at 3 x 105 cells/ml in ISC-CM-5% FCS was seeded (100 µl/well) into flat-bottom microtiter plates. Plates were incubated overnight at 37°C in air-6% CO2 to allow formation of a cell monolayer. Serial Ab dilutions in ISC-CM-0.1% BSA and additionally supplemented, as indicated, with NMS or huC1q were prepared in separate 96-well microtiter plates. An equal volume of ISC-CM-0.1% BSA, containing on average 30 (range, 15 to 45) TCID50 of PR8 per 50 µl, was added to each Ab dilution and incubated for 60 min at 37°C. One-hundred-microliter portions of the various Ab-virus mixtures (12 replicates per sample) were then transferred to the washed MDCK monolayers, and the plates were incubated for 60 min at 37°C. The Ab-virus inocula were then removed, and 150 µl of ISC-CM-0.1% BSA was added per well. The plates were incubated for another 4 h at 37°C before addition of 50 µl of ISC-CM-0.1% BSA containing 32 µg of trypsin (Whittaker Bioproducts) per ml to give a final trypsin concentration of {approx}8 µg/ml. Plates were incubated for another 2.5 days and then screened for infection by testing the medium for HA activity. The Ab concentration that protected 50% of the MDCK microcultures from infection was computed and defined as 1 U of VN activity. Each assay included a titration of the virus in the various media without Ab to confirm usage of an appropriate virus dose and absence of nonspecific inhibition by the various assay diluents.

Reaction of Ab with virus in solution. Purified virus was diluted in PBSN-0.01% BSA to give 1,200 HAU/ml (8.7 and 11.7 µg of viral protein/ml for PR8 and B/Lee, respectively). Virus samples were then mixed with equal volumes of Ab samples, typically at 20 µg/ml, and incubated for 60 min at room temperature. Replicate samples of the mixture were then serially diluted in PBSN and tested for HA titer as described above. In addition, 300 µl was overlaid onto a cushion (300 µl) of 10% sucrose in PBSN-0.01% BSA in ultraclear centrifuge tubes (5 by 41 mm) (Beckman, Palo Alto, Calif.). The samples were then centrifuged in an SW50 rotor (10 min, 35,000 rpm, {approx}100,000 x g, at {approx}20°C), and the supernatant was sucked off. The top half of the tube was cut off (to reduce contamination by residual free Ab adhering to the tube wall). The virus pellet was resuspended and partially dissolved in 50 µl of NP-40 (0.1% in H2O) and transferred into 250 µl of NaCl (0.02 M). Serial dilutions of this sample were made in 0.02 M NaCl, and replicate 25-µl samples of each dilution were transferred into wells of polyvinyl plates and dried overnight at room temperature into the plastic well. In parallel, 150 µl of the starting Ab sample was diluted with 50 µl of 0.1% NP-40 and 100 µl of 0.02 M NaCl to obtain the Ab concentration present in the original virus-Ab mix and was then diluted similarly to the resuspended virus-Ab pellet for preparation of immunoadsorbent. For assay, the plates were blocked by incubation with PBSN-1% BSA ({approx}1 h at room temperature) and washed, and the amount of plate-bound IgG was determined by standard ELISA, using biotinylated Ab 187.1 (anti-CK) for detection, followed by avidin-alkaline phosphatase and p-nitrophenyl phosphate (all from Sigma). The amount of bound Ab in the test samples was quantitated by comparison to a dilution series of purified H36-4, using an E max ELISA reader and Softmax software (both from Molecular Devices Corp., Sunnyvale, Calif.).


arrow
RESULTS
 
Noninactivated NMS enhances the HI activity of HA-specific Abs depending on their fine specificity and heavy-chain isotype. In a previous study (24), we had observed that noninactivated NMS from SCID mice differentially affected the VN activity of a selected group of HA-specific Abs, enhancing it 10- to 70-fold in the case of some Abs (exemplified in the following by H35-C12) while having no significant effect in the case of others (exemplified by H36-4). To explore this phenomenon further, we tested the effect of NMS from BALB/c and C57BL/6 mice on Ab-mediated HI activity. The HI assay, which measures an Ab’s ability to prevent virus from agglutinating CRBC, was chosen rather than the VN assay, first because we wanted to test a large panel of MAb samples that contained sodium azide as a preservative, which would have interfered in the VN assay, and second because the HI assay is more reproducible and less labor-intensive than the VN assay. In agreement with the previous findings with VN assays, noninactivated NMS enhanced the HI activity of purified MAb H35-C12 approximately 100-fold but had a minimal effect (2- to 3-fold enhancement) on the HI activity of H36-4 (Fig. 1). Heat inactivation (30 min at 56°C) destroyed the enhancing activity of NMS. Enhancement reached maximum levels at a concentration of 0.5% NMS and declined sharply at concentrations of below 0.25%. At these concentrations, noninactivated NMS had no measurable effect on its own on viral HA activity, but at a concentration of >1%, NMS increasingly interfered in the assay by reducing viral HA activity (Fig. 1C), by agglutinating CRBC or by modifying the agglutination pattern of CRBC (data not shown). Sera from many donor animals of both the BALB/c and C57BL/6 strains were tested and found to result in comparable enhancement.



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 1. Noninactivated NMS enhances the HI activity of H35-C12 much more than the activity of H36-4. (A and B) Purified MAbs H35-C12 (A) and H36-4 (B) were tested for HI activity against PR8 in the presence of the indicated concentrations of noninactivated (closed symbols) and heat-inactivated (open symbols) NMS from C57BL/6 and BALB/c mice. HI activity is the Ab concentration at which three of four agglutinating doses of PR8 are prevented from agglutinating CRBC. The data show the means and standard deviations from three to five independent determinations. The dashed line in panel A indicates the fact that all but one assay were performed at a starting Ab concentration of 20 µg/ml. (C) Effect of NMS alone on viral HA activity. At concentrations of above 1%, noninactivated NMS interfered with the HI assay by reducing viral HA activity on its own and by altering the agglutination patterns of CRBC.

To investigate the possible relationship between Ab specificity, heavy-chain isotype, and susceptibility for enhancement, culture fluids from a panel of 140 hybridomas were tested for HI activity in the presence of 1% noninactivated or heat-inactivated NMS. A >=3-fold increase in HI titer in the presence of noninactivated serum was considered significant (dashed line in Fig. 2). These assays indicated that enhancement was dependent on both the Ab’s heavy-chain isotype and specificity. Accordingly, it was seen predominantly with IgG2a and IgG2b, rarely with IgG3, and never with IgG1 and IgM isotypes. The role of Ab specificity was evidenced by the fact that among Abs of the IgG2a and IgG2b isotypes, 100% of those directed to antigenic region Cb/E (designation of antigenic region according to the H1 model [reference 7 and see Fig. 7] followed after the slash by designation used in the H3 model [37, 38]), 50% of those directed to region Ca/D,A, and fewer than 10% of those directed to region Sa,Sb/B were enhanced. In some of the culture fluids, no HI activity was measurable in the absence of noninactivated serum; in these cases, the enhancements shown are minimum estimates. The relationship between enhancement, heavy-chain isotype, and specificity is also demonstrated in Table 1, which shows the geometric mean enhancement of individual Ab groups.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 2. Susceptibility of Abs to enhancement of their HI activity by NMS is dependent of their heavy-chain isotype and fine specificity. Culture fluids from PR8 HA-specific hybridomas were tested for HI activity against PR8 in the presence of 1% noninactivated or heat-inactivated NMS from C57BL/6 mice. A threefold increase in HI activity in the presence of noninactivated NMS is considered significant, and this level is indicated by the dashed line. The geometric mean HI titer ratio (noninactivated to inactivated) from at least two independent assays is shown and plotted for individual MAbs. The MAbs are grouped according to their heavy-chain isotype and the region of HA to which they bind as deduced from reduced reaction with a panel of single point virus mutants (7). Each circle represents an individual MAb.



View larger version (75K):
[in this window]
[in a new window]
  		FIG. 7. Locations of antigenic regions of the HA trimer. A model of the HA trimer of the H3 subtype (38, 39) is shown in side view with one monomer stained in light blue and the other stained in darker blue. Amino acids that were present in mutated form on individual mutants of PR8 (H1 subtype) and used to assign MAb specificity (7) are colored in yellow for the Cb/E, green for the Ca/A,D, red for the Sa, and dark blue for the Sb regions, with the latter making up the Sa,Sb/B region comprising mostly the tip of the trimer (7). MAbs that displayed reduced reaction with mutations belonging to distinct regions were assigned in this analysis to the region that had a dominant effect. In this side view, most of the mutated amino acids are visible only on one of the monomers, but in some cases, analogous mutations are visible on more than one monomer. The location of the sialic acid binding site, which is a shallow depression in the globular region of the HA monomer, is indicated by a yellow circle. Most mutations other than those marking the Cb/E region cluster around the binding site.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Enhancement of HI activity of MAbs by NMS in relation to their specificity and heavy-chain isotypea

The enhancing factor in noninactivated serum is C1q. The heat lability and IgG isotype dependency (enhancement of IgG2 > IgG3 > IgG1) suggested that the complement system may be involved (26, 32). In the classic pathway of complement activation, C1q, the binding moiety of the first complement component C1, binds to the heavy chains of antigen-complexed Abs. This interaction may then result in the activation of the associated serine esterases (C1r and C1s); the subsequent activation and local deposition of C4 and C2 (C4b2a), which then promote activation and deposition of C3, the central and most abundant protein of the system; and eventually the activation of additional downstream components. However, since noninactivated serum from C3-/- mice (36) was as effective as NMS from BALB/c and C57BL/6 mice (data not shown), C3 and subsequent activities of the complement system were not involved in the enhancement. We next tested purified huC1q at 0.5 µg/ml, which is the approximate concentration of mouse C1q (msC1q) present in 0.5% NMS, and found that it enhanced the HI activity of H35-C12 to roughly the same degree as 0.5% NMS (Table 2; Fig. 1A and 3A). By contrast, the HI activity of H36-4 was only minimally enhanced ({approx}2-fold) by huC1q at 0.5 µg/ml (Fig. 3B). A more substantial enhancement was seen at C1q concentrations of >=10 µg/ml. However, at these higher concentrations, huC1q had an inhibitory effect on its own on viral HA activity (Fig. 3B, inset), suggesting that these further increases in enhancement may have been due in large part to direct antiviral activities of huC1q. To verify that the enhancement by NMS was due to C1q, NMS was incubated for 60 min at room temperature with goat anti-huC1q antiserum prior to testing its enhancing activity. This treatment strongly suppressed the enhancement activity of NMS (Table 2). The observation that goat anti-huC1q was slightly less effective at inhibiting the enhancing activity of NMS than that of purified huC1q is probably due to a less effective interaction of this antiserum with the heterologous msC1q.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Enhancement of HI activity by noninactivated NMS is mediated by C1qa



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 3. Enhancement of HI activity by purified huC1q. HI assays were performed in the presence of the indicated concentrations of huC1q. (A) HI activities of H35-C12 (circles) and of two other HA Cb/E-specific Abs of the IgG2a isotype (triangles and inverted triangles). Results from a representative assay of several repeat assays is shown. Each symbol shows the mean of three replicate values per dilution. (B) HI activity of H36-4. Inset, HA activity of PR8 incubated with the indicated concentrations of huC1q in the absence of Ab, expressed as percentage of the original activity.

Taken together, these findings indicated that msC1q present in noninactivated NMS was responsible for enhancing the HI activity of H35-C12.

How does C1q enhance the HI activity of H35-C12? C1q consists of six heterotrimers, each containing a C-terminal globular domain (approximately 5 by 7 nm) and a collagenous N-terminal stem (approximately 23 nm long) (21). The latter is bundled in its N-terminal half, giving the molecule its characteristic shape of a bouquet of flowers. With its globular domains, C1q binds to the CH3 domain of IgM and the CH2 domains of certain IgG isotypes when aggregated or complexed with antigen. In view of its large size (460,000 Da) and high avidity for antigen-bound Ig (8), one could envision two principle mechanisms by which C1q enhanced the HI activity of certain HA-specific Abs. First, by cross-linking virus-bound Abs, C1q may stabilize virus-Ab complexes of low avidity. Second, in the case of virus-bound Abs that are ineffective, on their own, in preventing attachment of virus to CRBC, C1q may increase Ab-mediated activity by enhancing steric interference due to the increase in protein mass deposited onto the virus.

The first possibility was tested by comparing the binding of Abs in ELISA in the presence or absence of NMS or huC1q. As shown in Fig. 4A and B, noninactivated NMS and huC1q had no substantial effect on the binding activity of H35-C12 or H36-4. Furthermore, both Abs fixed msC1q and huC1q to comparable degrees (Fig. 4C and D). Thus, the differential enhancement of the HI activity of these MAbs by C1q could not be attributed unequivocally to a differential improvement of their binding to virus or to a difference in interaction with C1q. However, a stabilizing effect of C1q on virus-bound Ab, which may become relevant only in the HI assay, in which Ab has to compete with sialylated glycoconjugates of CRBC for reaction with HA, is indirectly supported by the finding that most of the strongly enhanced MAbs were isolated from the primary response. For instance, in the case of the Cb/E-specific MAbs of the IgG2a isotype, 12 were isolated from the primary response and 6 were isolated after various secondary immunization protocols. The former showed an average enhancement, on a log10 basis, of 1.28 ± 0.13, while the latter showed a significantly smaller (t test, P < 0.05) enhancement of 0.8 ± 0.11. The basis for this difference between primary and memory MAbs may be a difference in avidity (28), but this has not been verified experimentally and we cannot exclude a difference in fine specificity.



View larger version (23K):
[in this window]
[in a new window]
  		FIG. 4. NMS and huC1q do not substantially modify the reactions of H35-C12 and H36-4 to PR8 in ELISA, and both Abs fix C1q equally well. H35-C12 (A and C) and H36-4 (B and D) were tested for binding to PR8 immunoadsorbent, prepared by adsorption of virus to plastic from a solution as used in the HI test. Ab dilutions were tested in the following diluents: PBSN(Ca,Mg) containing 0.1% BSA (circles) and the same diluent supplemented with (i) 0.5 µg of huC1q per ml (triangles), (ii) 0.5% noninactivated NMS (squares), or (iii) 0.5% inactivated NMS (diamonds). (A and B) Ab bound to the viral immunoadsorbent as detected by developing the assay with the biotinylated anti-CK MAb 187. (C and D) Bound C1q as detected by developing the assay with goat anti-huC1q antiserum followed by biotinylated mouse anti-goat MAb. Each point shows the mean optical density at 405 nm minus optical density at 750 nm (OD405–750) for triplicate samples above background (diluent alone). The data are from one of three repeat assays.

To test the second possibility, we investigated whether Cb/E-specific MAbs were capable of binding to intact virus in solution without preventing it from attaching to CRBC. To this end, purified PR8 virus at a protein concentration of 4.4 µg/ml ({approx}20 nM HA monomer, assuming that HA constitutes {approx}35% of viral protein mass [9]) or B/Lee at a similar concentration was incubated for 1 h at room temperature with a threefold molar excess of purified MAb (10 µg/ml; 70 nM). A sample of the virus-Ab mixture was then tested for residual HA activity, while the rest was layered on top of a 10% sucrose cushion and centrifuged for 10 min at 110,000 x g to pellet the virus and virus-bound Ab. The pellet was resuspended in 0.1% NP-40, and the Ab concentration was determined by an IgG-specific ELISA. The difference in Ab concentration between the PR8 and B/Lee pellets was taken as the PR8-specific binding value. At 70 nM, both H35-C12 (Fig. 5A) and H36-4 (Fig. 5B) saturated the PR8 virus sample and showed only minimal (<1%) nonspecific binding to B/Lee. Under these conditions, 19 ± 3 nM H35-C12 (average ± standard error of the mean [SEM], n = 7) and 49 ± 9 nM H36-4 (n = 7) bound to PR8 (Fig. 5C). (The excessive amount of virus-bound Ab relative to viral HA could be due to inaccuracies in the determination of viral and/or antibody protein concentrations, the former being based entirely on a chemical assay and the latter being based on both chemical assay and UV light absorption, and/or to secondary interactions between Abs that enhanced the amount of virus-bound Ab.) Importantly, saturation of the virus with H35-C12 had a negligible effect on the HA activity of the virus (<2-fold reduction), while saturation with H36-4 reduced HA activity by >200-fold to a nondetectable level (Fig. 5D). Four additional MAbs, two specific for region Cb/E and two specific for region Sa,Sb/B, were tested in the same manner. The two Cb/E-specific Abs showed two- to threefold-higher binding to PR8 than did H35-C12, yet the Ab-opsonized virus still agglutinated CRBC to a similar titer as control virus incubated with diluent only. By contrast, the two Sa,Sb/B-specific Abs saturated PR8 at 2- to 3-times-lower densities than H36-4 yet still prevented the virus from agglutinating CRBC. Taken together, the data clearly showed that Cb/E-specific Abs could be bound at high density to PR8 without significantly reducing the ability of the opsonized virus to agglutinate CRBC. Thus, it is conceivable that deposition of C1q to such Ab-opsonized virus may then result in inhibition of viral HA activity through steric interference.



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 5. Relationship between density of virus-bound Abs and inhibition of viral HA activity. (A and B) A constant dose of PR8 (open symbols) or Lee virus (closed symbols) ({approx}20 nM viral HA monomer) was incubated for 1 h at room temperature with increasing concentrations (3 to 330 nM) of purified MAb H35-C12 (A) or H36-4 (B). The virus was then pelleted through a sucrose cushion, and the amount of MAb associated with the virus pellet was determined by ELISA. Mean values from two independent determinations are shown. (C) Virus ({approx}20 nM HA) was incubated for 1 h at room temperature with the indicated MAbs ({approx}70 nM): 1, H36-4, IgG2a; 2, H37-77, IgG2a; 3, H36-12, IgG2b; 4, H35-C12, IgG2a; 5, H2-4C2, IgG2a; and 6, L2-10C1, IgG2a. MAbs 1 to 3 are specific for Sa,Sb/B, and MAbs 4 to 6 are specific for Cb/E. The virus-Ab mixture was pelleted though a sucrose cushion, and the concentration of copelleted Ab was determined by ELISA. The data are means and SEMs from three to seven independent determinations. (D) Prior to centrifugation, a sample of the virus-Ab mixture was tested for HA activity, which is expressed as percentage of the original HA activity contained in the sample.

Enhancement of HI activity of serum Abs in the presence of noninactivated NMS. In view of the strong enhancement seen with Cb/E-specific hybridoma Abs, we were interested in knowing what degree of enhancement, if any, would be seen with sera from immunized mice. To this end, serum samples obtained from BALB/c mice undergoing a primary or memory response to PR8 virus were heat inactivated and then tested for HI activity in the presence of 0.5% noninactivated or inactivated NMS. Figure 6A shows that the HI titers of all sera obtained 7 to 14 days after primary intravenous (i.v.) immunization were significantly (paired t test) enhanced, on average by 2.5-fold, in the presence of 0.5% noninactivated NMS. The enhancement was less pronounced and not significant (paired t test) with sera from the memory response (on average, 1.3-fold). The difference in enhancement between primary and secondary sera was significant. Furthermore, enhancement was more pronounced in sera obtained after i.v. immunization (Fig. 6B) than in those obtained after pulmonary infection (Fig. 6C). Overall, these findings are consistent with previous studies indicating that Cb/E-specific Abs make up roughly 50% of the primary response after i.v. immunization and that a dominant and structurally restricted group of Abs in this response, of which H35-C12 is a member, fail to participate in the memory response (18, 19). Note that a twofold enhancement in HI titer would be consistent with 50% of the serum Abs exhibiting NMS-enhanced HI activity. These findings indicate that the enhancement is not an idiosyncratic behavior of some peculiar hybridoma antibodies but is a property of a significant fraction of the primary anti-HA Ab response.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 6. Enhancement of HI titer of immune sera by noninactivated NMS. (A) Several immune sera from BALB/c mice obtained 10 to 14 days after primary immunization and 13 to 25 days after secondary immunization with PR8 were heat inactivated and then tested for HI titer against PR8 in the presence of 0.5% noninactivated or heat-inactivated NMS. The primary-response sera showed an enhancement of 10(0.38 ± 0.04) (mean ± SEM) (paired t test, P < 0.05), and the secondary-response sera showed an enhancement of 10(0.11 ± 0.06) (paired t test, P > 0.05), by noninactivated NMS compared to inactivated NMS. Primary sera were more strongly enhanced than secondary sera (t test, P < 0.05). (B) Heat-inactivated plasma samples from BALB/c mice obtained at various time intervals after i.v. immunization with 500 HAU ({approx}3.5 µg of viral protein) of purified PR8 were tested for HI activity in the presence of 0.5% noninactivated (filled circles) or inactivated (open circles) NMS. Pooled plasma samples from 10 mice were tested at each time point. (C) Heat-inactivated serum samples from BALB/c mice obtained at various time points after primary pulmonary infection were tested for HI activity as for panel B. Each time point represents the data obtained with a serum pool from two to four mice.

Effects of NMS and huC1q on VN activity. We reported previously that the VN activity of H35-C12 was enhanced on average by 70-fold in the presence of 1.65% noninactivated serum from SCID mice (24). To test whether enhancement of VN activity was also attributable to C1q, we measured VN activity of MAbs diluted in the standard culture medium containing 0.1% BSA or in the same medium supplemented with 0.5% noninactivated or heat-inactivated NMS or 0.5 µg of huC1q per ml. The supplemented media were present during the 1-h preincubation of virus with Ab dilutions and the subsequent 1-h incubation of the virus-Ab mixture with MDCK cell microcultures, both at 37°C. The MDCK cultures were then washed, and all were incubated for 3 days with the same standard medium containing 0.1% BSA and trypsin. As shown in Table 3, the VN activity of H35-C12 was significantly, though on average only approximately fourfold, enhanced in the presence of 0.5% noninactivated NMS, and a similar degree of enhancement was seen in the presence of 0.5 µg of huC1q per ml. As in HI, the enhancing activity was abolished by heat treatment or preincubation of serum with goat anti-huC1q (Table 3). Thus, enhancement of VN activity by NMS, though smaller in degree than that seen with HI activity, was also due to C1q. Noninactivated serum and huC1q had no significant effect on the VN activity of H36-4 (Table 3), and none of the media had a measurable effect on the infectivity titer of the virus inoculum in the absence of MAb (data not shown).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Effect of serum on VN activitya


arrow
DISCUSSION
 
This study showed that noninactivated NMS substantially enhanced the HI and, to a lesser degree, VN activities of certain HA-specific Abs. The enhancement was mediated by C1q, the recognition protein of the first component of complement, and did not require the participation of downstream components of the complement system.

C1q consists of six heterotrimers, each containing a C-terminal globular domain (approximately 5 by 7 nm) and a collagenous N-terminal stem (approximately 23 nm long) (21). With its globular domains, it can bind to the CH3 domain of IgM and the CH2 domains of certain IgG isotypes when these are aggregated or complexed with antigen. The avidity of this interaction depends largely on its valency, ranging from a poor monovalent binding with a Kd of {approx}40 µM to a strong hexavalent binding with a Kd of {approx}5 nM (8). Apart from the Fc domains, C1q can also bind to many other determinants, including various proteins, most notably from retroviruses, lipids, carbohydrates, cellular membrane components, and many others (8, 33). However, as shown by the highly Ab isotype- and specificity-dependent enhancement, the C1q-mediated activity studied here is clearly due to its reaction with the Fc region of virus-bound Abs. Ab-independent activities, due to C1q or contaminants in the C1q preparation, became detectable only at C1q concentrations that exceeded the one used here (0.5 µg/ml) by roughly 10-fold.

Our data are consistent with the notion that several mechanisms contributed to the C1q-mediated enhancement. The principal mechanism appears to be an increase in steric inhibition due to deposition of the large C1q (460 kDa) onto virus-bound Ab. We conclude this from the observed relationship between Ab specificity and susceptibility for C1q-mediated enhancement of HI activity. To mediate HI activity, Abs must inhibit a fraction of the 1,000 to 1,500 sialic acid binding sites on a virus particle from interacting with sialylated cellular ligands (6, 12). They may achieve this by direct obstruction of ligand access to the sialic acid binding site if they bind to an epitope that encompasses part of this site (3) or by global steric interference that prevents sufficient proximity between viral and cellular membranes if they bind to an epitope that is remote from the sialic acid binding site (16). The efficacy of steric inhibition would be expected to depend greatly on the topographical relationship between the sialic acid binding site and the Ab’s epitope. Thus, Abs bound to the Cb/E region, which comprises residues located membrane proximal of the sialic acid binding site (Fig. 7), would be expected to provide poor global steric inhibition, while Abs bound to the Sa,Sb/B region, which comprises residues on the tip of the HA, would be expected to provide effective steric inhibition and to inhibit HA activity even if their epitopes did not encompass part of the sialic acid binding site. The Ca/A,D region occupies the middle segment of the globular head of the HA trimer, and Abs binding to it would be expected to provide intermediate global steric interference. Accordingly, if C1q enhanced Ab-mediated HI by improving its global steric interference, it would be expected to be most effective with Cb/E-specific Abs, less effective with Ca/A,D-specific Abs, and least effective with Sa,Sb/B-specific Abs, as was found (Fig. 2 and Table 1). The role of steric inhibition in anti-Cb/E Ab-mediated HI is also supported by the finding that four MAbs of the IgM isotype directed to this site displayed substantial HI activity on their own (0.01 to 0.25 µg/ml) (data not shown). This makes sense in view of the large mass of the IgM pentamer (950 kDa) or hexamer (1,150 kDa), which is in the same range or larger than the mass of the complex of three molecules of IgG and one of C1q (910 kDa). Although C1q is likely to have interacted well also with IgM bound to the Cb/E region, the additional increase in mass apparently failed to further increase steric interference. The second mechanism by which C1q enhanced Ab-mediated HI activity appears to be through stabilization of IgG-virus complexes of low avidity. Although C1q had no substantial effect on the binding of Ab in ELISA (Fig. 3), evidence that this mechanism nevertheless contributed to enhancement came from the finding that Abs from the primary response were on average more strongly enhanced than Abs from the memory response. We assume that the former are, on average, of lower avidity than the latter (28). Finally, the possibility that aggregation of Ab-opsonized virus by C1q contributed to enhancement of HI activity appears unlikely, because this mechanism would be expected to be equally effective with IgM- and IgG2-opsonized virus, which was not the case (Fig. 2 and Table 1). Taken together, the data indicate that enhancement resulted principally from increased steric inhibition of HA activity by deposition of the bulky C1q onto Ab-opsonized virus and to a lesser degree from stabilization of virus-bound Abs of low avidity.

The relatively large range of enhancement seen within groups of MAbs of the same isotype and binding to the same region may in part be due to the fact that, as argued above, more than one mechanism is thought to contribute to enhancement. In addition, one must keep in mind that epitopes recognized by individual MAbs have been assigned to certain regions of HA sometimes on the basis of only a few mutated amino acid positions, often clustered closely together, that affected the MAb’s binding activity (7). As crystallographic studies have shown that a typical epitope comprises between 15 and 20 amino acids (3, 16), it is quite possible that Ab epitopes that have been mapped to the same region actually differ by more contact residues than they share, and such differences in fine specificity may contribute to the differences in enhancement seen within Ab specificity groups. In addition, differences in glycosylation and possibly V regions of Abs may contribute by affecting the binding of C1q (5). Large clonal differences in the ability of murine Abs of the same isotype to activate human complement have also been reported by other investigators (32, 39).

The isotype dependency of the enhancement described here cannot be directly compared to results of previous studies which investigated the role of murine heavy-chain isotypes in complement activation or cell lysis, because C1q binding does not invariably result in the subsequent activation of the complement system (8, 34). Furthermore, the source of complement affects the extent of fixation (10, 22), and murine Abs have been tested mainly with heterologous complement; in addition, as mentioned above, V region differences, affinity, and antigen specificity also appear to contribute to the efficacy of complement activation. Nevertheless, it is noteworthy that IgG2a, IgG2b, and IgG3 often scored as the best activators (10, 22, 26, 32), while IgG1 has been shown to comprise two subtypes, one with and the other without complement-activating capability (15, 26). IgM, which is generally considered to be a good activator of complement (26), may be a poor activator of murine complement (22), and its activation activity appears to be dependent on its degree of polymerization, with IgM hexamers being {approx}20 times more effective than pentamers (11, 29). The proportions of penta- and hexamers in our IgM Ab preparations have not been determined.

The finding that MAbs could bind at high density to HA Cb sites of virus particles without significantly reducing their ability to attach to CRBC (Fig. 5) and MDCK cells (our unpublished observation) was surprising in view of the comparable masses and sizes of the IgG molecule (150 kDa, comprising two 8-nm-long and 4- to 5-nm-wide Fab arms attached to one end of the 7-nm-long and 6-nm-wide Fc fragment [27]) and the HA trimer (225 kDa, 13.5 nm long, and 6 to 7 nm wide) and the narrow intertrimer space (average interspike distance of 10 nm minus the width of the trimer [16]), which makes it unlikely that the Ab could be tucked into this space and thus leave the membrane-distal sialic acid binding site relatively unobstructed. One possible explanation for the persisting HA activity of virus particles covered with HA Cb-specific MAb is that the Ab itself mediated the agglutination, e.g., if it bound monovalently to HA and displayed cross-reactivity with a determinant present on CRBC. However, we have no evidence for such a cross-reaction, and H35-C12 neither agglutinates nor can be absorbed with CRBC. Another possibility is that HA Cb-bound Ab becomes dislodged from the virus in the presence of CRBC (or MDCK cells) and in competition with cellular ligands for reaction with HA. Of note in this regard is a study (23) indicating that HA Sb/B-specific Abs of low affinity ({approx}6 x 107 M-1) may saturate the HA trimers yet neutralize only 50% of the virus particles. Those authors proposed that during competition of the Ab-opsonized virus with cellular ligands, roughly 90% of the bound Ab would be displaced at this stage and permit 50% of the virus to initiate infection (23). Although H35-C12 also has a relatively low affinity (0.8 x 107 M-1 versus solid-phase bromelain-released HA [24] and 2 x 108 M-1 versus infected cells [25]), a substantial displacement of the Ab during competition with cellular ligands would be difficult to reconcile with the finding that the infectivity of the virus remained neutralized (our unpublished observation; see also below). The possibility that the HA Cb region on PR8 displays antigenic heterogeneity, possibly due to differences in the size of a nearby carbohydrate chain, which could leave a sufficiently large fraction of HA monomers nonreactive with certain HA Cb-specific MAbs, is inconsistent with the strong binding of these MAbs to virus (Fig. 5C). Thus, the most likely explanation is that terminal sialic acid ligands are present on sufficiently long and narrow membrane components so that they can access the sialic acid binding site of an HA monomer even if Ab is bound to its Cb region.

Our finding that virus which is densely coated with certain types of Abs can nevertheless agglutinate CRBC is in disagreement with the concept that mere occupancy of the virus surface by Ab above a certain threshold density determines Ab-mediated antiviral activity (6). In the case of influenza virus, Ab fine specificity (reference 31 and the present study) and affinity (23) are additional important determinants of its activity. For instance, occupancy of 8 to 9% of viral HA monomers with the HA Sb/B-specific MAb H36-4 resulted in a 50% reduction of HA activity (our unpublished observation), while a much higher occupancy by some HA Cb/E-specific MAbs reduced HA activity by less than 50%.

Another surprising finding was that H35-C12, at 20 µg/ml, failed to reduce significantly the ability of the virus to attach to CRBC yet prevented it from infecting MDCK cells, even at a 10-fold-lower concentration (Table 3). Although agglutination of CRBC and attachment to MCDK cells are different processes, the above finding is consistent with the possibility that H35-C12 mediated VN activity not by inhibiting virus attachment to MDCK cells but rather by inhibiting a subsequent step of infection, such as endocytosis, intraendosomal fusion, or release of replication-competent ribonucleoprotein complexes into the cytoplasm. The importance of neutralization at a stage subsequent to attachment has long been postulated by Dimmock (12). By measuring virus attachment and VN in the same ELISA system, Edwards and Dimmock have recently been able to show clearly that some Abs are more effective in inhibiting virus infection than virus attachment (13, 14). This could be demonstrated most convincingly with an HA Cb-specific IgG MAb (H9-D3), most notably its Fab (14), and showed unequivocally that the Fab of this MAb neutralized virus at a postattachment stage, apparently by inhibiting intracellular fusion (14). We have used the same method (14) and found that H35-C12 was more than 10 times more effective in inhibiting virus infection than virus attachment to MDCK cells (our unpublished observation), thus providing another example of a MAb that neutralizes virus at a postattachment stage. It is noteworthy that H9-D3 and H35-C12 recognize overlapping but not identical epitopes in the HA Cb region: both MAbs show reduced reaction with virus mutants with the replacement mutations L78P, L79P, and S83P (position according to H3 numbering, preceded by the parental amino acid and followed by replacement mutation) but differ in binding to V81E and E122K (reduced for H35-C12) or R82G (reduced for H9-D3) (18).

The rather modest enhancement (4-fold) of VN activity of H35-C12 seen here in the presence of 0.5% NMS was surprising in view of the 70-fold enhancement seen previously in the presence of 1.65% NMS from SCID mice (24). Repeat assays confirmed that noninactivated NMS from SCID mice promoted a significantly stronger enhancement of VN activity, but not HI activity, than NMS from immunocompetent BALB/c mice, even when both were tested at the same concentration. The reason for this difference is presently unknown.

Complement has been shown previously to enhance Ab-mediated VN activity in various virus systems (17), including influenza type A virus (1). However, in contrast to the findings in the present study, C3 and in some cases components of the membrane attack complex were involved in all cases. This includes the anti-HA Abs in human sera, which neutralized A/WSN influenza virus, yet only in the presence of C1,C4, C2, and C3 (1). The independence from the residual complement system of the C1q-mediated enhancement seen here may be due to an ineffective activation of the classic pathway by C1q upon reaction with virus-bound anti-Cb/E Abs, as it is well established that binding of C1 does not always result in its activation (8, 34). Another possibility is that the sera were used at a suboptimal concentration for effective activation of the classic pathway. However, higher concentrations of NMS (>1.65%) were difficult to assess because they displayed inhibitory activities in the absence of exogenous MAb. The role of the C1q-mediated enhancement of Ab activities in vivo remains to be determined by comparing the activities of passive MAbs in C1q-/- (4) and C4-/- (36) mice.


arrow
ACKNOWLEDGMENTS
 
We thank Mike Carroll for serum samples from C3-/- mice and Laszlo Otvos for reviewing the manuscript.

This work was supported by NIH grant AI 13989.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: The Wistar Institute, 3601 Spruce St., Philadelphia, PA 19104. Phone: (215) 898-3840. Fax: (215) 898-3868. Email: gerhard{at}mail.wistar.upenn.edu. Back


arrow
REFERENCES
 
    1
  1. Beebe, D. P., R. D. Schreiber, and N. R. Cooper. 1983. Neutralization of influenza virus by normal human sera: mechanisms involving antibody and complement. J. Immunol. 130:1317–1322.[Abstract]
    2. 2
  2. Bindon, C. I., G. Hale, and H. Waldmann. 1988. Importance of antigen specificity for complement-mediated lysis by monoclonal antibodies. Eur. J. Immunol. 18:1507–1514.[Medline]
    3. 3
  3. Bizebard, T., B. Gigant, P. Rigolet, B. Rasmussen, O. Diat, P. Boeseccke, S. A. Wharton, J. J. Skehel, and M. Knossow. 1995. Structure of influenza virus haemagglutinin complexed with a neutralizing antibody. Nature 376:92–95.[CrossRef][Medline]
    4. 4
  4. Botto, M., C. Dell’Agnola, A. E. Bygrave, E. M. Thompson, H. T. Cook, F. Petry, M. Loos, P. P. Pandolfi, and M. J. Walport. 1998. Homozygous C1q deficiency causes glomerulonephritis associated with multiple apoptotic bodies. Nat. Genet. 19:56–59.[CrossRef][Medline]
    5. 5
  5. Brekke, O. H., T. E. Michaelsen, and I. Sandlie. 1995. The structural requirements for complement activation by IgG: does it hinge on the hinge. Immunol. Today 16:85–89.[CrossRef][Medline]
    6. 6
  6. Burton, D. R., E. O. Saphire, and P. W. H. I. Parren. 2001. A model for neutralization of viruses based on antibody coating of the virion surface. Curr. Top. Microbiol. Immunol. 260:109–143.[Medline]
    7. 7
  7. Caton, A. J., G. G. Brownlee, J. W. Yewdell, and W. Gerhard. 1982. The antigenic structure of the influenza virus A/PR/8/34 hemagglutinin (H1 subtype). Cell 31:417–427.[CrossRef][Medline]
    8. 8
  8. Cooper, N. R. 1985. The classical complement pathway: activation and regulation of the first complement component. Adv. Immunol. 37:151–207.[Medline]
    9. 9
  9. Cusack, S., R. W. H. Ruigrok, P. C. J. Krygsman, and J. E. Mellema. 1985. Structure and composition of influenza virus. A small-angle neutron scattering study. J. Mol. Biol. 186:565–585.[CrossRef][Medline]
    10. 10
  10. Dangl, J. L., T. G. Wensel, S. L. Morrison, L. Stryer, L. A. Herzenberg, and V. T. Oi. 1988. Segmental flexibility and complement fixation of genetically engineered chimeric human, rabbit and mouse antibodies. EMBO J. 7:1989–1994.[Medline]
    11. 11
  11. Davis, A. C., K. H. Roux, and M. J. Shulman. 1988. On the structure of polymeric IgM. Eur. J. Immunol. 18:1001–1008.[Medline]
    12. 12
  12. Dimmock, N. J. 1993. Neutralization of animal viruses. Curr. Top. Microbiol. Immunol. 183:1–150.[Medline]
    13. 13
  13. Edwards, M. J., and N. J. Dimmock. 2000. Two influenza A virus-specific Fabs neutralize by inhibiting virus attachment to target cells, while neutralization by their immunoglobulin Gs is complex and occurs simultaneously through fusion inhibition and attachment inhibition. J. Virol. 278:423–435.
    14. 14
  14. Edwards, M. J., and N. J. Dimmock. 2001. A haemagglutinin (HA1)-specific Fab neutralizes influenza A virus by inhibiting fusion activity. J. Gen. Virol. 82:1387–1395.[Abstract/Free Full Text]
    15. 15
  15. Ey, P. L., S. J. Prowse, and C. R. Jenkin. 1979. Complement-fixing IgG1 constitutes a new subclass of mouse IgG. Nature 281:492–493.[CrossRef][Medline]
    16. 16
  16. Fleury, D., B. Barriere, T. Bizebard, R. S. Daniel, J. J. Skehel, and M. Knossow. 1999. A complex of influenza hemagglutinin with a neutralizing antibody that binds outside the virus receptor binding site. Nat. Struct. Biol. 6:530–534.[CrossRef][Medline]
    17. 17
  17. Hirsch, R. L. 1982. The complement system: its importance in the host response to viral infection. Microbiol. Rev. 46:71–85.[Free Full Text]
    18. 18
  18. Kavaler, J., A. J. Caton, L. M. Staudt, D. Schwartz, and W. Gerhard. 1990. A set of closely related antibodies dominate the primary antibody response to the antigenic site Cb of the A/PR/8/34 influenza virus hemagglutinin. J. Immunol. 145:2312–2321.[Abstract]
    19. 19
  19. Kavaler, J., A. J. Caton, L. M. Staudt, and W. Gerhard. 1991. A B cell population that dominates the primary response to influenza virus hemagglutinin does not participate in the memory response. Eur. J. Immunol. 21:2687–2695.[Medline]
    20. 20
  20. Kelley, V. E., G. N. Gaulton, and T. B. Strom. 1987. Inhibitory effects of anti-interleukin 2 receptor and anti-L3T4 antibodies on delayed type hypersensitivity: the role of complement and epitope. J. Immunol. 138:2771–2775.[Abstract]
    21. 21
  21. Kishore, U., and K. B. M. Reid. 1999. Modular organization of proteins containing C1q-like globular domain. Immunopharmacology 42:15–21.[CrossRef][Medline]
    22. 22
  22. Klaus, G. G. B., M. B. Pepys, K. Kitajima, and B. A. Askonas. 1979. Activation of mouse complement by different classes of mouse antibody. Immunology 38:687–695.[Medline]
    23. 23
  23. Kostolansky, F., E. Vareckova, T. Betakova, V. Mucha, G. Russ, and S. A. Wharton. 2000. The strong positive correlation between effective affinity and infectivity neutralization of highly cross-reactive monoclonal antibody IIB4, which recognizes antigenic site B on influenza A virus haemagglutinin. J. Gen. Virol. 81:1727–1735.[Abstract/Free Full Text]
    24. 24
  24. Mozdzanowska, K., M. Furchner, G. Washko, J. Mozdzanowski, and W. Gerhard. 1997. A pulmonary influenza virus infection in SCID mice can be cured by treatment with hemagglutinin-specific antibodies that display very low virus-neutralizing activity in vitro. J. Virol. 71:4347–4355.[Abstract]
    25. 25
  25. Mozdzanowska, K., K. Maiese, M. Furchner, and W. Gerhard. 1999. Treatment of influenza virus-infected SCID mice with nonneutralizing antibodies specific for the transmembrane proteins matrix 2 and neuraminidase reduces the pulmonary virus titer but fails to clear the infection. Virology 254:138–146.[CrossRef][Medline]
    26. 26
  26. Neuberger, M. S., and K. Rajewsky. 1981. Activation of mouse complement by monoclonal mouse antibodies. Eur. J. Immunol. 11:1012–1016.[Medline]
    27. 27
  27. Padlan, E. A. 1996. X-ray crystallography of antibodies. Adv. Protein Chem. 49:57–128.[Medline]
    28. 28
  28. Rajewsky, K. 1996. Clonal selection and learning in the antibody system. Nature 381:751–757.[CrossRef][Medline]
    29. 29
  29. Randall, R. D., L. B. King, and R. B. Corley. 1990. The biological effects of IgM hexamer formation. Eur. J. Immunol. 20:1971–1979.[Medline]
    30. 30
  30. Scherle, P. A., G. Palladino, and W. Gerhard. 1992. Mice can recover from pulmonary influenza virus infection in the absence of class I-restricted cytotoxic T cells. J. Immunol. 148:212–217.[Abstract]
    31. 31
  31. Schofield, D. J., J. R. Stephenson, and N. J. Dimmock. 1997. Variations in the neutralizing and haemagglutination-inhibiting activities of five influenza A virus-specific IgGs and their antibody fragments. J. Gen. Virol. 78:2431–2439.[Abstract]
    32. 32
  32. Seino, J., P. Eveleigh, S. Warnaar, L. J. M. van Haarlem, L. A. van Es, and M. R. Daha. 1993. Activation of human complement by mouse and mouse/human chimeric monoclonal antibodies. Clin. Exp. Med. 94:291–296.
    33. 33
  33. Sim, R. B., and K. B. M. Reid. 1991. C1: molecular interactions with activating systems. Immunol. Today 12:307–311.[CrossRef][Medline]
    34. 34
  34. Tao, M.-H., R. I. F. Smith, and S. L. Morrison. 1993. Structural features of human immunoglobulin G that determine isotype-specific differences in complement activation. J. Exp. Med. 178:661–667.[Abstract/Free Full Text]
    35. 35
  35. Tutt, A. L., R. R. French, T. M. Illidge, J. Honeychrch, H. M. McBride, C. A. Penfold, D. F. Fearon, R. M. E. Parkhouse, G. G. B. Klaus, and M. J. Glennie. 1998. Monoclonal antibody therapy of B cell lymphoma: signaling activity on tumor cells appears more important than recruitment of effectors. J. Immunol. 161:3176–3185.[Abstract/Free Full Text]
    36. 36
  36. Wessels, M. R., P. Butko, M. Ma, H. B. Warren, A. L. Lage, and M. C. Carroll. 1995. Studies of group B streptococcal infection in mice deficient in complement components C3 or C4 demonstrate an essential role for complement in both innate and acquired immunity. Proc. Natl. Acad. Sci. USA 92:11490–11494.[Abstract/Free Full Text]
    37. 37
  37. Wiley, D. C., and J. J. Skehel. 1987. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annu. Rev. Biochem. 56:366–394.
    38. 38
  38. Wilson, I. A., and N. J. Cox. 1990. Structural basis of immune recognition of influenza virus hemagglutinin. Annu. Rev. Immunol. 8:737–771.[CrossRef][Medline]
    39. 39
  39. Yokoyama, I., and F. Waxman. 1993. Clonal variations in complement activation and deposition of C3b and C4b on model immune complexes. Immunology 80:168–176.[Medline]

Journal of Virology, February 2002, p. 1369-1378, Vol. 76, No. 3
0022-538X/01/$04.00+0     DOI: 10.1128/JVI.76.3.1369-1378.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Jayasekera, J. P., Moseman, E. A., Carroll, M. C. (2007). Natural Antibody and Complement Mediate Neutralization of Influenza Virus in the Absence of Prior Immunity. J. Virol. 81: 3487-3494 [Abstract] [Full Text]  
  • Jayasekera, J. P., Vinuesa, C. G., Karupiah, G., King, N. J. C. (2006). Enhanced antiviral antibody secretion and attenuated immunopathology during influenza virus infection in nitric oxide synthase-2-deficient mice.. J. Gen. Virol. 87: 3361-3371 [Abstract] [Full Text]  
  • Huber, V. C., McKeon, R. M., Brackin, M. N., Miller, L. A., Keating, R., Brown, S. A., Makarova, N., Perez, D. R., MacDonald, G. H., McCullers, J. A. (2006). Distinct Contributions of Vaccine-Induced Immunoglobulin G1 (IgG1) and IgG2a Antibodies to Protective Immunity against Influenza.. CVI 13: 981-990 [Abstract] [Full Text]  
  • Zharikova, D., Mozdzanowska, K., Feng, J., Zhang, M., Gerhard, W. (2005). Influenza Type A Virus Escape Mutants Emerge In Vivo in the Presence of Antibodies to the Ectodomain of Matrix Protein 2. J. Virol. 79: 6644-6654 [Abstract] [Full Text]  
  • Torres, M., May, R., Scharff, M. D., Casadevall, A. (2005). Variable-Region-Identical Antibodies Differing in Isotype Demonstrate Differences in Fine Specificity and Idiotype. J. Immunol. 174: 2132-2142 [Abstract] [Full Text]  
  • Li, J. S.-y., Chu, F., Reilly, A., Winslow, G. M. (2002). Antibodies Highly Effective in SCID Mice During Infection by the Intracellular Bacterium Ehrlichia chaffeensis Are of Picomolar Affinity and Exhibit Preferential Epitope and Isotype Utilization. J. Immunol. 169: 1419-1425 [Abstract] [Full Text]  

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
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 arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Feng, J. Q.
Right arrow Articles by Gerhard, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Feng, J. Q.
Right arrow Articles by Gerhard, W.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»