Abstract
B cell tolerance is maintained by active deletion and functional anergy of self-reactive B cells depending on the time, amount, and site of the self-antigen expression. To study B cell tolerance toward a transplacentally transmitted viral Ag, we crossed transgenic mice expressing the μ heavy and the κ light chain of the lymphocytic choriomeningitis virus (LCMV)-neutralizing mAb KL25 (HL25-transgenic mice) with persistently infected LCMV carrier mice. Although HL25-transgenic LCMV carrier mice exhibited the same high virus titers as nontransgenic LCMV carrier mice, no evidence for B cell tolerance was found. In contrast, enhanced LCMV-neutralizing Ab titers were measured that, however, did not clear the virus. Instead, LCMV isolates from different tissues turned out to be neutralization resistant Ab escape variants expressing different substitutions of amino acid Asn119 of the LCMV-glycoprotein 1 that displays the neutralizing B cell epitope. Virus variants with the same mutations were also selected in vitro in the presence of the transgenic mAb KL25 confirming that substitutions of Asn119 have been selected by LCMV-neutralizing Abs. Thus, despite abundant expression of viral neo-self-antigen in HL25-transgenic LCMV carrier mice, transgenic B cells expressing LCMV-neutralizing Abs were rather stimulated than tolerized and neutralization resistant Ab escape variants were selected in vivo.
Activation of self-reactive B cells is precluded in many cases by the lack of self-reactive T cell help (1, 2). Furthermore, B cell tolerance can be maintained by active deletion and by functional anergization of self-reactive B cells depending on whether the recognized Ag is membrane bound (3, 4, 5, 6), or soluble (7) or whether it is organized multimerically (8, 9, 10). In mice persistently infected with the noncytopathic lymphocytic choriomeningitis virus (LCMV)6 the viral glycoproteins LCMV-GP1 and -GP2 are expressed in membrane bound form on the surface of virus infected cells and as free virus particles (11, 12, 13, 14). In transplacentally infected LCMV carrier mice, persisting LCMV is already detectable in the embryo beginning from week 2 after gestation (11, 12, 13, 15, 16). After birth the surface expression of LCMV-GP on infected cells is reduced when compared with carrier mice derived by neonatal infection (15), which might indicate a reduced capacity to induce specific B cell tolerance. In both transplacentally and neonatally induced LCMV carrier mice, the persisting virus leads to thymic deletion of specific CTL and T helper cells (17, 18, 19). Induction of B cell tolerance in LCMV carrier mice is less well understood. LCMV-neutralizing Abs specific for LCMV-GP1 have not been detected in the sera of LCMV carrier mice (20).
To analyze B cell tolerance induction toward persisting LCMV, we crossed HL25-transgenic mice (21) expressing the μ heavy chain and the κ light chain of the LCMV-neutralizing mAb KL25 with transplacentally and neonatally induced LCMV carrier mice. Despite similar virus expression in HL25-transgenic LCMV carrier mice and nontransgenic LCMV carrier mice, HL25-transgenic B cells were not rendered tolerant. Instead, HL25-transgenic LCMV carrier mice showed enhanced neutralizing Ab levels when compared with noninfected HL25-transgenic mice and neutralization-resistant LCMV Ab escape variants (LCMV-AEv) were selected in vivo. LCMV variants isolated from various tissues expressed different substitutions of the amino acid Asn119 of LCMV-GP1. LCMV variants selected in vitro with the parental transgenic mAb KL25 showed the same substitutions of Asn119 confirming that these variations were caused by the Ab selection pressure. Our results indicate, that LCMV escaped the virus neutralizing B cell response more rapidly than inducing B cell tolerance. Apparently, viral Ags are more variable than most autoantigens and therefore less likely meet criteria for induction of specific B cell tolerance.
Materials and Methods
Viruses
The LCMV isolates WE-T7a and WE were originally provided by Dr. F. Lehmann-Grube (Hamburg, Germany), and were grown on L-929 cells for 48 h in MEM/5% FCS after initial infection with a multiplicity of infection (MOI) of 0.01.
AEv from LCMV isolated from tissue homogenates and blood of HL25-transgenic LCMV carrier mice (LCMV-AEv) were grown on MC57G mouse fibroblasts for 48 h and subsequently plaque-purified three times in vitro by infecting MC57G mouse fibroblasts, and incubating them under a 1% methyl cellulose-overlay at 37°C/5% CO2 atmosphere. After 3 days the overlay was replaced by 1% agarose/0.01% neutral red, and after one other day of incubation plaques became visible in the MC57G monolayer. To pick plaques the agarose above the plaques was taken using aerosol resistant tips (Molecular Bio-Products, San Diego, CA), the virus was eluted in MEM/5% FCS, and grown on MC57G cells for 48 h.
For in vitro selection of LCMV-AEv (LCMV-ivAEv), 105 plaque-forming units (PFU) LCMV-WE, and 80 μg of purified mAb KL25 were coincubated for 90 min at 37°C/5% CO2 in a total volume of 2 ml MEM/5% FCS in one well of a six-well-plate. Then, 8 × 105 MC57G cells in 1 ml MEM/5% FCS were added and the culture was grown for 48 h at 37°C/5% CO2. Virus in the supernatant was harvested and was plaque-purified three times in vitro.
Mice
HL25-transgenic mice expressing the Igμ heavy and Igκ light chain of the LCMV-neutralizing mAb KL25 were generated as described previously (21). HL25-transgenic mice were bred under specific pathogen-free conditions at the Institut für Zuchthygiene (University of Zurich, Switzerland). For the generation of transplacentally induced HL25-transgenic LCMV carrier mice, C57BL/6 LCMV-WE-T7a carrier mice were crossed with HL25-transgenic mice under conventional housing conditions. For the generation of neonatally induced HL25-transgenic LCMV carrier mice, 105 PFU of LCMV-WE-T7a were injected i.c. within the first 24 h after birth. All experimental mice were kept under conventional conditions.
FACS analysis
FACS analysis was performed on a FACScan (Becton Dickinson, San Diego, CA) according to standard procedures. The binding of monoclonal LCMV-neutralizing Abs KL25 (22) and WEN1 (23) to LCMV-AEv was tested on MC57G mouse fibroblasts infected at an initial MOI of 0.01 40 h before analysis. mAbs KL25 and WEN1 were purified on Staphylococcus aureus protein G (Sepharose fast flow Protein G, Pharmacia, Uppsala, Sweden) and used at a concentration of 10 μg/ml. Binding of KL25 and WEN1 to infected MC57G cells was detected using goat anti-mouse IgG1-FITC (Southern Biotechnology Associates, Birmingham, AL), and goat anti-mouse IgG2a-FITC (Southern Biotechnology), respectively.
LCMV titer and neutralization assay
LCMV titers from organs of infected mice and from infected tissue cultures were determined in an infectious focus formation assay as described (24). Anti-LCMV-neutralizing Ab titers were determined in an in vitro infectious focus reduction assay as described (24). Total Ig values were determined in nonreducing conditions.
DNA-sequence analysis of LCMV-GP1
Total RNA of MC57G cells infected with either LCMV wild-type or LCMV-AEv for 48 h at an initial MOI of 0.01 was isolated according to the method of Chomczynski and Sacchi (25). RT-PCR was performed using the LCMV-GP1-specific primer R1 (5′-1037TCG TAG CAT GTC ACA GAA TTC TTC1014-3′; base pair numbering according to Romanowski et al. (26) for reverse transcription and the primer pairs 001/R1 and 001/RC1 for Hot Start PCR amplification (001, 5′-1CGC ACC GGG GAT CCT AGG CTT21-3′; RC1, 5′-965GAG CTC TGG AGC AA GGA TCA TCC942-3′). PCR products were sequenced by automated Taq cycle sequencing (Taq Dye Deoxy Terminator Cycle Sequencing kit, Applied Biosystems, Foster City, CA; Bio-Rad, Hercules, CA) using the primers 001, 333 (5′-333CTG ACG ATG CCC AAT GC349-3′) and RCM (5′-756GGT ACT GAT AGC TTG TTT GGC TGC ACC733-3′).
Results
HL25-transgenic mice (21) expressing the LCMV-neutralizing mAb KL25 specific for the LCMV-GP1 were crossed with LCMV carrier mice. HL25-transgenic LCMV carrier offsprings exhibited LCMV-neutralizing serum Ab titers as early as 9 days after birth, reached a maximum titer of 7–8 by day 18, and remained stable at this level for the rest of life (Fig. 1⇓). In contrast, noninfected HL25-transgenic mice developed spontaneous LCMV-neutralizing serum Abs ∼35 days after birth that remained stable at a titer of 4. Thus, in HL25-transgenic LCMV carrier mice spontaneous LCMV-neutralizing serum titers emerged earlier and reached higher levels than in noninfected HL25-transgenic mice indicating that transgenic B cells were rather activated than rendered tolerant.
LCMV-neutralizing Ab response in HL25-transgenic LCMV carrier mice. Serum from transplacentally induced HL25-transgenic LCMV carrier mice (▪), noninfected HL25-transgenic mice (▴), and transgene-negative LCMV carrier mice (○) were tested for LCMV-neutralizing Ab titers in an infectious focus reduction assay (24). Sera were 10× prediluted and titrated in 2-fold dilution steps. Shown are means ± SD of five mice per group and time point of one representative out of two similar experiments.
Because the induction of B cell tolerance depends on the level and the site of self-antigen expression (3, 4, 5, 6, 7) and in transplacentally induced LCMV carrier mice the expression of LCMV-GP1 on the surface of infected cells is reduced compared with neonatally infected carrier mice (15), HL25-transgenic mice were also infected neonatally with 105 PFU LCMV i.c. to establish a persistent infection. Again, no B cell tolerance was induced in neonatally infected HL25-transgenic LCMV carrier mice and again the LCMV-neutralizing Ab response was accelerated and enhanced by the persistent infection (data not shown).
Because B cells producing neutralizing Abs have been shown to contribute to the clearance of a persistent LCMV infection (27), we tested whether the high titers of neutralizing Abs cleared the LCMV infection in HL25-transgenic LCMV carrier mice. Viral burden was measured in the blood, spleen, kidney, liver, brain, and bone marrow of transplacentally and neonatally induced HL25-transgenic LCMV carrier mice as well as transgene-negative LCMV carrier mice. Virus was not reduced in HL25-transgenic LCMV carrier mice when compared with transgene-negative LCMV carrier mice over the entire observation period of 80 days (Fig. 2⇓; data from neonatally induced HL25-transgenic LCMV carrier mice are not shown).
LCMV is not eliminated in HL25-transgenic LCMV carrier mice. Blood samples and organ homogenates from transplacentally induced HL25-transgenic LCMV carrier mice (•) and transgene-negative LCMV carrier mice (○) were tested for infectious virus titer in an infectious focus formation assay (24). Values are means ± SD of log virus titers per ml blood from five mice per group and time point (A) or log virus titer per organ of individual mice (B). Dashed lines indicate the detection limits of the assay. Shown is one representative out of two similar experiments.
To test, whether in this case LCMV isolated from HL25-transgenic LCMV carrier mice had escaped from neutralization by the transgenic Ab, MC57G fibroblasts were infected with different LCMV isolates and binding of the parental transgenic mAb KL25 and the GP1-specific control mAb WEN1 (23) to LCMV-GP expressed on the cell surface was analyzed by FACS. Both mAbs bound cells infected with virus isolated from spleen of HL25-transgenic LCMV carrier mice 9 days after birth (Fig. 3⇓A, panels A and B). However, only the control mAb WEN1 but not the parental transgenic mAb KL25 bound cells infected with virus isolated 56 days after birth (Fig. 3⇓A, panels E and F). Cells infected with virus isolates from transgene-negative LCMV carrier mice were always bound by both mAbs (Fig. 3⇓A, panels C and D and G and H). Similar results were obtained for virus isolates from blood, kidney, liver, brain, and bone marrow: virus isolated at day 9 from HL25-transgenic LCMV carrier mice was bound by the transgenic mAb KL25, but isolates from day 18 or later were not (Fig. 3⇓B, panels A–D), whereas virus isolated from nontransgenic LCMV carrier mice was always bound (Fig. 3⇓B, panels E–H; data only shown from blood). Similar results were obtained from neonatally induced LCMV carrier mice.
Virus isolated from HL25-transgenic LCMV carrier mice is not bound by the transgenic mAb KL25. LCMV-GP expressed on infected MC57G cells was tested by FACS analysis for binding of the transgenic mAb KL25. (A) LCMV isolated at days 9 and 56 from spleen of HL25-transgenic LCMV carrier mice (tg; panels A, B, E,and F) and LCMV isolated from transgene-negative littermate LCMV carrier mice (ctrl; panels C, D, G, and H) was tested for binding of the transgenic mAb KL25 (panels A, C, E, and G) and of the control mAb WEN1 (panels B, D, F, and H). (B) MC57G mouse fibroblasts were infected with LCMV isolated at days 9, 18, 32, and 56 from blood of HL25-transgenic LCMV carrier mice (tg; panels A–D) and LCMV isolated from transgene-negative littermate LCMV carrier mice (ctrl; panels E–H).
To further support the finding of in vivo selected LCMV-AEv, the virus isolates were tested in a virus neutralization assay in vitro. As shown in Table I⇓, the transgenic Ab KL25 was not able to neutralize virus isolated from blood of either transplacentally or neonatally induced HL25-transgenic LCMV carrier mice at day 80 after birth confirming that these LCMV isolates were AEv. Three other LCMV-neutralizing control mAb, WEN1, WEN3, and WEN4 (23), were still able to neutralize the escape variants. A polyclonal LCMV-neutralizing hyperimmune serum also retained neutralizing capacity against the Ab escape mutants although it was reduced by two to four titer steps. LCMV isolated from transgene-negative LCMV carrier mice was neutralized by all mAbs and the polyclonal LCMV-neutralizing hyperimmune serum.
Neutralization of in vivo-selected LCMV-AEv
To examine the escape mutations at a molecular level, the cDNA sequence of the GP1 of LCMV-WE (22, 28) was analyzed from LCMV-AEv. MC57G fibroblasts were infected with LCMV-AEv isolated from blood of transplacentally and neonatally infected HL25-transgenic LCMV carrier mice and from nontransgenic LCMV carrier mice. The total RNA was isolated and subjected to RT-PCR. PCR products were purified and directly sequenced. Consistently, all analyzed LCMV-AEv showed a substitution of Asn119 of LCMV-GP1 (Fig. 4⇓). Asn119 was either replaced by lysine, serine, histidine, tyrosine, or threonine.
In vivo Ab escape by substitution of the Asn119 of LCMV-GP1 by lysine, serine, tyrosine, threonine, and histidine. Sequences of different LCMV-AEv are aligned with the LCMV-WE-T7a wild-type sequence (LCMV-WE-T7a wt, top row). Shown are sequences of LCMV-AEv isolated from transplacentally (AEv) and neonatally induced HL25-transgenic LCMV carrier mice [AEv(ic)]. Amino acids are numbered according to Romanowski et al. (26). Dashes indicate sequence identity.
To prove that the LCMV-GP1 Asn119 substitutions in LCMV-AEv were conferring the loss of a neutralizing determinant, LCMV-AEv were selected in vitro by the mAb KL25: LCMV-WE (105 PFU) and mAb KL25 (80 μg) were coincubated for 90 min at 37°C and grown for 48 h on MC57G mouse fibroblasts in the presence of the Ab. Culture supernatant was plaque-purified, and virus was tested for neutralization by mAb KL25 (Table II⇓), and for mutations in LCMV-GP1 (Fig. 5⇓). LCMV-ivAEv showed the same characteristic Asn119 substitutions as LCMV-AEv selected in vivo in HL25-transgenic mice. Thus, Ab must have selected LCMV variants in vitro as well as in vivo.
In vitro Ab escape by substitution of the Asn119 of LCMV-GP1 by lysine and serine after culture of virus in the presence of mAb KL25. Sequences of four different LCMV-ivAEv are aligned with the LCMV-WE wild-type sequence (LCMV-WE wt, top row). For notation and numbering, see Fig. 4⇑.
Neutralization of in vitro-selected LCMV-AEv
Discussion
No specific B cell tolerance was observed in HL25-transgenic LCMV carrier mice expressing the viral Ag and the transgenic antiviral Ab. Instead, transgenic B cells were stimulated and secreted virus neutralizing Abs. Such transgene-encoded virus-neutralizing Abs, however, did not clear the virus from the carrier mice but selected virus variants resistant to neutralization by the transgenic Ab.
The fact that HL25-transgenic B cells are not deleted or anergized may be due to several reasons. 1) As indicated by enhanced LCMV-neutralizing serum titers in HL25-transgenic LCMV carrier mice the contact between the B cell receptor and the viral Ag is long enough to stimulate B cells. However, wild-type LCMV may have escaped too fast to tolerize transgenic B cells. Furthermore, due to the fast elimination of wild-type virus and the fast kinetics of the appearance of escape virus and due to the high B cell renewal rate it is highly unlikely that a transient B cell tolerance state could have been detected. Furthermore, it is unlikely that a coincident inflammatory response interferes with tolerance induction because HL25-transgenic LCMV carrier mice got into first contact with LCMV at a prenatal or neonatal stage when mice are not yet immunocompetent. 2) B cells, binding GP on viral particles in a highly organized multimeric form may only be induced and cannot be anergized as suggested by previous studies (10). 3) The B cell response in HL25-transgenic mice is virtually monoclonal and it cannot completely be ruled out that the chosen Ab specificity is rather exceptional than representative for the LCMV-neutralizing Ab response in nontransgenic mice. However, there is convincing evidence that neutralizing Ab responses against viruses and bacteria may use a restricted oligoclonal Ig-VHVL repertoire (29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39), particularly early in life (40). Because of these factors, it may be difficult to compare the present conditions and results with previous reports indicating that transgenically expressed monomeric neo-self-antigen leads to deletion or anergy of self-reactive B cells (3, 4, 5, 6, 7, 41, 42).
In contrast to the observations in HL25-transgenic mice, the absence of GP-specific-neutralizing Abs in nontransgenic LCMV carrier mice was taken as an evidence for B cell tolerance (20). However, B cells expressing LCMV-neutralizing Abs are of low frequency in nontransgenic mice and therefore LCMV-neutralizing Abs in the serum may be below detection limits. Furthermore, titers below detection limits may also be due to complete consumption of neutralizing Abs by the virus in LCMV carrier mice.
Virus variants selected in vivo in HL25-transgenic LCMV carrier mice and virus variants selected in vitro in the presence of mAb KL25 exhibited similar substitutions of Asn119 in LCMV-GP1 that was replaced either by lysine, serine, tyrosine, threonine, or histidine. These amino acid substitutions did not affect the surface expression of LCMV-GP on infected cells, as indicated by FACS analysis of GP1, which is not affected by the substitution of Asn119 (Fig. 3⇑A). Therefore Asn119 of GP1 is either directly involved in the determinant bound by Ab KL25 or it indirectly affects the conformation of the viral epitope. Virus variants emerging under the selective pressure in vivo in HL25-transgenic LCMV carrier mice and in vitro in the presence of mAb KL25 differ in many aspects from influenza virus variants. In contrast to influenza AEv, LCMV variants are still neutralized by complex antisera and by some other neutralizing mAbs. It therefore may not surprise that the concept of original antigenic sin that was developed based on the observations on influenza virus epidemics (43, 44, 45, 46) does not apply to the presented situation with LCMV. An influenza virus infection induces a protective Ab response that cross-reacts in an evolutionary backwards direction, but not in a forward direction. On the molecular level this can be explained by exchanges of less bulky amino acids by more bulky ones in the epitopes of variant viruses. In the case of the LCMV-GP1 Asn119 replacements, both more bulky substitutions (lysine, tyrosine, and histidine) and less bulky ones (serine and threonine) have been found.
Ab escape variation is a prominent problem in widely distributed virus infections. In vitro experiments have been used to investigate induction of AEv (47, 48, 49, 50). In humans, AEv from the cytopathic influenza virus have been described at the population level (51, 52). In the case of the noncytopathic HIV, AEv have been isolated from single infected individuals (53, 54). Likewise, in HL25-transgenic LCMV carrier mice the noncytopathic LCMV escaped the neutralizing Ab response in vivo within single individuals. The present model of HL25-transgenic mice therefore may serve as an animal model to investigate conditions of Ab escape variation of noncytopathic viruses in vivo. It may be helpful for the development of humoral immunotherapy or vaccination strategies against comparable viruses in humans, e.g., hepatitis virus B and C and probably HIV.
Acknowledgments
We thank R. Städeli and D. Zimmermann for excellent help in the generation of DNA sequence data and J. Brecher for technical assistance.
Footnotes
-
↵1 This work was supported by Swiss National Science Foundation Grants 31-32195.91, 31-32179.91, 31-50884.97, and 31-50900.97.
-
↵2 All sequences reported have been deposited in the GenBank/EMBL database and assigned accession numbers AJ233160–AJ233202.
-
↵3 Address correspondence and reprint requests to Dr. Peter Seiler, Max-Planck-Institut für Infektionsbiologie, Monbijoustrasse 2, D-10117 Berlin, Germany. E-mail address: seiler_p{at}mpiib-berlin.mpg.de
-
↵4 Current address: Division de Pathologie Clinique, Centre Medical Universitaire, Genève, Switzerland.
-
↵5 Current address: European Molecular Biology Laboratory Mouse Biology Programme, Adriano Buzzati-Traverso Campus, Monterotondo (Rome), Italy.
-
↵6 Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; AEv, Ab escape variants; LCMV-AEv, LCMV AEv selected in vivo in transplacentally infected HL25-transgenic lymphocytic choriomeningitis carrier mice; LCMV-ivAev, LCMV AEv selected in vitro; GP, glycoprotein; PFU, plaque forming units; MOI, multiplicity of infection.
- Received November 2, 1998.
- Accepted January 28, 1999.
- Copyright © 1999 by The American Association of Immunologists