Volume 43, Issue 9 p. 2338-2348
Regular Article
Free Access

Nucleoprotein-specific nonneutralizing antibodies speed up LCMV elimination independently of complement and FcγR

Tobias Straub

Tobias Straub

Department of Immunology, Institute of Medical Microbiology and Hygiene, University of Freiburg, Freiburg, Germany

Faculty of Biology, University of Freiburg, Freiburg, Germany

Search for more papers by this author
Oliver Schweier

Oliver Schweier

Department of Immunology, Institute of Medical Microbiology and Hygiene, University of Freiburg, Freiburg, Germany

Search for more papers by this author
Michael Bruns

Michael Bruns

Heinrich-Pette-Institut für Experimentelle Virologie und Immunologie an der Universität Hamburg, Hamburg, Germany

Search for more papers by this author
Falk Nimmerjahn

Falk Nimmerjahn

Department of Biology, Institute of Genetics, University of Erlangen-Nürnberg, Erlangen, Germany

Search for more papers by this author
Ari Waisman

Ari Waisman

Institute for Molecular Medicine, Johannes Gutenberg-University of Mainz, Mainz, Germany

Search for more papers by this author
Hanspeter Pircher

Corresponding Author

Hanspeter Pircher

Department of Immunology, Institute of Medical Microbiology and Hygiene, University of Freiburg, Freiburg, Germany

Full correspondence: Prof. Hanspeter Pircher, Department of Immunology, Institute of Medical Microbiology and Hygiene, University of Freiburg, Hermann-Herder-Str.11, 79104 Freiburg, Germany

Fax: +49-761-203-6577

e-mail: [email protected]

See accompanying article:

http://dx.doi.org/10.1002/eji.201343566

Search for more papers by this author
First published: 10 June 2013
Citations: 29

See accompanying article by Richter and Oxenius

Abstract

CD8+ T cells have an essential role in controlling lymphocytic choriomeningitis virus (LCMV) infection in mice. Here, we examined the contribution of humoral immunity, including nonneutralizing antibodies (Abs), in this infection induced by low virus inoculation doses. Mice with impaired humoral immunity readily terminated infection with the slowly replicating LCMV strain Armstrong but showed delayed virus elimination after inoculation with the faster replicating LCMV strain WE and failed to clear the rapidly replicating LCMV strain Docile, which is in contrast to the results obtained with wild-type mice. Thus, the requirement for adaptive humoral immunity to control the infection was dependent on the replication speed of the LCMV strains used. Ab transfers further showed that LCMV-specific IgG Abs isolated from LCMV immune serum accelerated virus elimination. These Abs were mainly directed against the viral nucleoprotein (NP) and completely lacked virus neutralizing activity. Moreover, mAbs specific for the LCMV NP were also able to decrease viral titers after transfer into infected hosts. Intriguingly, neither C3 nor Fcγ receptors were required for the antiviral activity of the transferred Abs. In conclusion, our study suggests that rapidly generated nonneutralizing Abs specific for the viral NP speed up virus elimination and thereby may counteract T-cell exhaustion.

Introduction

Chronic infections with non- or poorly cytopathic viruses like HCV and HIV affect several hundred million of people worldwide. To combat these infections, T cells are essential; however, the role of humoral immunity is less clear. Inoculation of mice with lymphocytic choriomeningitis virus (LCMV) is a well-established animal model to study immunological effector mechanisms in infection with a prototypic noncytopathic virus. To control LCMV infection in mice, CD8+ T cells are required. B-cell-deficient mice have been used by many groups to investigate the role of humoral immunity in the LCMV infection model. The first experiments performed with such mice showed that virus elimination and generation of memory CD8+ T cells were not altered in the absence of B cells 1. When higher virus infection doses and other viral strains were used, virus clearance was, however, impaired 2-4. In other studies, recrudescence of viremia after initial virus clearance was observed months after infection, and memory T cells from long-term LCMV-infected B-cell-deficient mice were reported to be less efficient in adoptive immunotherapy 5, 6.

The conclusions of these studies in B-cell-deficient mice were challenged as it was realized that B-cell deficiency also alters the splenic microarchitecture. In particular, B-cell-deficient mice have a defective splenic marginal zone 7 and LCMV injected systemically may quickly spread to peripheral organs. In addition, the production of type I IFN after LCMV infection is nearly absent in mice lacking B cells due to the aberrant cell composition of the splenic marginal zone 8. To overcome these limitations, Bergthaler et al. used B-cell-sufficient mouse models with impaired abilities to generate antigen-specific Abs 9. Their data suggested that LCMV envelope specific Abs facilitated virus clearance after high-dose LCMV WE infection. The authors further showed that treatment with a neutralizing LCMV glycoprotein (GP) specific mAb prevented viral persistence and T-cell exhaustion. These data fit well with recent reports demonstrating that IL-6-, OX40-, or TLR7-deficient mice that failed to control chronic infection with LCMV clone 13 were also hampered in the generation of LCMV-specific IgG Abs 10-12.

In all of the studies mentioned above, mice were infected with high doses of LCMV that lead to viremia for a prolonged time and to the production of virus envelop specific Abs. Inoculation of mice with low doses of LCMV results in the rapid production of virus-specific Abs that are mainly directed against the viral nucleo-protein (NP) 13, 14. It is, however, unclear whether these Abs have any impact on virus elimination. In the current study, we have addressed this question by infecting B-cell-sufficient mice with an impaired ability to produce antigen-specific Abs with low doses of LCMV strains that differ in their replication speed. The results revealed that the requirement for adaptive humoral immunity to control the infection is dependent on the replicative capacity of the viral strains used. Ab transfer experiments further demonstrated that nonneutralizing NP-specific IgG Abs were capable of accelerating virus elimination in vivo. Surprisingly, these Abs functioned in an Fcγ receptor (FcγR) and C3 complement-independent manner.

Results

Normal responses in mice with impaired adaptive humoral immunity after LCMV Armstrong infection

To overcome the caveats of mice lacking B cells, B-cell-sufficient MD4 mice were used. MD4 mice express a transgenic B-cell receptor specific for hen egg lysozyme and due to allelic exclusion, their B-cell repertoire is compromised 15. For our experiments, we used the LCMV strains Armstrong, WE, and Docile, which differ in their replication speed (Docile > WE > Armstrong) 16. MD4 mice were first infected with the slowly replicating LCMV strain Armstrong using a low virus infection dose (200 PFU). This induced a strong GP33- and NP396-specific CD8+ T-cell response and marked upregulation of the effector cell marker killer lectin-like receptor G1 (KLRG1) on CD8+ T cells similar as in B6 wild-type mice (Fig. 1A). As in wild-type mice, virus was completely cleared in spleen, liver, and lungs of MD4 mice at day 8 postinfection (p.i.) (Fig. 1B). The same result was obtained with IgMi mice, which are severely impaired in the production of soluble Abs due to a mutated IgH gene locus 17 (Supporting Information Fig. 1). These data demonstrate that MD4 and IgMi mice were not inherently impaired in mounting a potent LCMV-specific CD8+ T-cell response and that an adaptive Ab response was not required to control LCMV Armstrong infection.

Details are in the caption following the image
Immune response in MD4 and IgMi mice after infection with LCMV Armstrong and LCMV WE. (A, B) Mice of the indicated genotypes were infected (i.v.) with 200 PFU LCMV Armstrong. At day 8 p.i., (A) splenocytes were analyzed by flow cytometry and (B) viral titers in the indicated organs were determined. Symbols in the graphs represent values from individual mice. Pooled data from three independent experiments are shown. Dashed horizontal lines represent mean values and the horizontal line indicates the detection limit (B). (C, D) Mice of the indicated genotypes were infected (i.v.) with 200 PFU LCMV WE. (C) At day 14 p.i., splenocytes were analyzed by flow cytometry. Pooled data from four experiments are shown. Dashed horizontal lines represent mean values. *p < 0.05; **p < 0.01; n.s.: not significant, Mann–Whitney-U-test. (D) Viral titers in spleens of MD4 (open circles), IgMi (open triangles), and B6 (filled circles) mice were determined at the indicated time points. Data are shown as mean ± SEM of 2–7 mice per time point pooled from one to four independent experiments performed. Horizontal lines indicate the detection limit.

Impaired CD8+ T-cell response and delayed viral clearance after LCMV WE infection

When the faster replicating LCMV strain WE was used, we observed a decrease in KLRG1 induction and fewer GP33-specific CD8+ T cells in MD4 compared with B6 wild-type mice at day 14 p.i. (Fig. 1C). Virus elimination in the spleen was delayed, nevertheless, virus was cleared in these mice as well (Fig. 1D, left). Similar to MD4 mice, virus clearance was also delayed in IgMi mice (Fig. 1D, right). Thus, after LCMV WE infection, the virus-specific CD8+ T-cell response and virus elimination were delayed in the absence of an Ab response.

T-cell exhaustion and virus persistence after LCMV Docile infection

Most strikingly, infection of MD4 mice with the fast replicating LCMV strain Docile led to classical signs of CD8+ T-cell exhaustion indicated by low KLRG1 expression, strongly decreased IFN-γ production and significant expression of the exhaustion markers, PD-1 and 2B4 (Fig. 2A and B). LCMV Docile infected B6 wild-type mice mounted a vigorous CD8+ T-cell response characterized by high-KLRG1 expression and potent IFN-γ production. T-cell exhaustion in LCMV Docile infected MD4 mice was paralleled by the inability to clear the infection leading to high viral titers in spleen, liver, and lungs in comparison with wild-type mice (Fig. 2C, top). The same results were obtained when viral titers in IgMi mice after LCMV Docile infection were analyzed (Fig. 2C, bottom). Taken together, these data suggested that Abs induced in the early phase of an LCMV Docile infection were required to prevent T-cell exhaustion and viral persistence. Due to the phenotype of Ab-deficient mice after LCMV Docile infection, we used this viral strain for all subsequent experiments of this study.

Details are in the caption following the image
T-cell exhaustion and virus persistence in MD4 and IgMi mice after LCMV Docile infection. Mice of the indicated genotypes were infected (i.v.) with 200 PFU LCMV Docile. (A) At day 14 p.i., splenocytes were analyzed by flow cytometry. Symbols in the graphs represent values from individual mice. Pooled data from two independent experiments are shown. Dashed horizontal lines represent mean values. (B) At day 33 p.i., the expression of PD-1 and 2B4 gated on splenic CD8+ T cells was determined. Data shown are geometric mean values of fluorescence intensity +SEM derived from three independent experiments each performed with 1–4 mice per group. (C) Viral titers in the indicated organs of LCMV Docile-infected MD4 (open circles), IgMi (open triangles), and B6 (filled circles) mice at the indicated time points. Data are shown as mean ± SEM of 3–6 mice per time point pooled from one to four independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 n.s.: not significant, Mann–Whitney-U-test. Horizontal lines indicate the detection limit.

Rapid production of nonneutralizing LCMV-specific Abs

Next, we determined the kinetics of the LCMV-specific Ab response in B6 mice using a newly established sensitive sandwich ELISA as detailed in the Material and methods. LCMV-specific IgG titers in serum of LCMV Docile infected mice strongly increased between days 6 and 8 and reached maximal levels 2 weeks p.i. (Fig. 3A, filled circles). The IgG response was T-cell help dependent since Ab titers were strongly decreased in CD4+ T-cell-depleted mice (Fig. 3A, open circles). The viral antigen specificity of immune serum taken from LCMV Docile infected mice at d20 p.i. was analyzed by immunoprecipitation and immunoblotting. The results revealed that LCMV immune serum predominantly contained Abs specific for LCMV NP (Fig. 3B) confirming previous data 14. Importantly, virus neutralizing activity was never observed in these LCMV immune sera even when used at a high concentration (Fig. 3C). To provide additional evidence for the lack of virus neutralizing activity, virus serum mixes (90% serum) were incubated overnight before inoculation into mice. Two days after inoculation, LCMV titers in spleens were enumerated. The neutralizing LCMV GP specific mAb KL25 was used as a positive control in these assays. As shown in Fig. 3D, treatment with mAb KL25 completed prevented infection whereas preincubation with LCMV immune serum did not affect initial viral replication.

Details are in the caption following the image
Ab response in B6 mice after LCMV infection. (A) B6 (filled circles) and CD4+ T-cell-depleted B6 mice (open circles) were infected (i.v.) with 200 PFU LCMV Docile and LCMV-specific IgG levels in serum were determined by ELISA. Endpoint titers are shown as mean ± SEM of 2–6 mice per group pooled from one to three independent experiments performed. (B) Reactivity of LCMV immune serum (IS). Cell lysates from LCMV infected (inf.) and control (ctrl.) MC57G cells were immunoprecipitated with mouse LCMV immune serum (IS) and precipitated proteins were visualized by Western blotting using a poly-reactive anti-LCMV rabbit serum. A representative blot from two independent experiments is shown. (C) Lack of virus neutralizing of activity of LCMV immune serum. LCMV was incubated with the indicated dilution of LCMV immune serum (IS, circles) or neutralizing mAb KL25 (triangles) followed by virus titration in focus-forming assays. The percent reduction in virus foci formation compared with that of controls is shown as mean ± SEM of three to ten samples pooled from three independent experiments. (D) B6 mice were infected (i.v.) with mixes containing 200 PFU LCMV Docile in 90% normal (NS, open circles) or LCMV immune serum (IS, filled circles). The LCMV neutralizing mAb KL25 (triangles) was used as positive control. Viral titers in the spleen were determined at day 2 p.i. Symbols in the graphs represent values from individual mice. Horizontal dashed lines represent median values, the horizontal line indicates the detection limit, n.s.: not significant. Data shown are pooled from two independent experiments.

Accelerated virus elimination by immune serum and Ab transfer

Having shown that mice with impaired humoral immunity were unable to control LCMV Docile infection, we next wondered whether transfer of LCMV immune serum could accelerate virus clearance. First, LCMV Docile infected MD4 and IgMi mice were treated with LCMV immune sera free of infectious virus that were obtained from infected wild-type mice at day 20 p.i. Viral titers in spleen, liver, and lungs were determined 14 days later. This treatment was able to lower viral titers in some mice but the antiviral effects were variable, particularly when using MD4 mice (Fig. 4). To obtain a more robust read-out for the potential antiviral activity of LCMV-specific Abs, we next tested B6 wild-type mice as hosts. Mice were infected with LCMV Docile and at day 1 serum from healthy uninfected mice (= normal serum) or LCMV immune serum was injected i.p. and the kinetics of viral elimination was followed. At day 2 and day 4 p.i., viral load between the two groups did not significantly differ (Fig. 5). At day 6, we observed a slight reduction in viral titers in spleen, liver, and lungs of immune compared with normal serum-treated mice. This difference became more prominent at day 8 p.i. At this time point, viral titers in spleen, liver, and lungs were 100–1000-fold lower in immune serum-treated mice. Further experiments in CD8+ T-cell-depleted recipients showed that accelerated virus clearance by immune serum transfer was only effective in the presence of CD8+ T cells. To provide direct evidence that the antiviral activity of the transferred immune serum was mediated by Abs, the experiments were repeated using protein-G-purified IgG Abs. As depicted in Fig. 6, viral titers in mice treated with purified IgG Abs from LCMV immune serum were significantly decreased compared to mice that received the same amounts of IgG from normal serum. Of note, purified IgG from immune serum lacked activity in virus neutralization assays in vitro up to a concentration of 100 μg/mL (data not shown). Hence, nonneutralizing IgG Abs from LCMV immune serum possessed antiviral activity in vivo.

Details are in the caption following the image
Transfer of LCMV immune serum into infected MD4 and IgMi mice. (A) MD4 mice were infected (i.v.) with 200 PFU LCMV Docile. At days 1 and 6 p.i., mice received (i.p.) 0.5 mL of normal serum (NS, open circles) or LCMV immune serum (IS, filled circles). (B) IgMi mice were infected (i.v.) with 200 PFU LCMV Docile. At day 1, mice received (i.p.) 0.5 mL of normal serum (NS, open triangles) or LCMV immune serum (IS, closed triangles). (A and B) Viral titers in the indicated organs were determined at day 14. Symbols in the graphs represent values from individual mice. Pooled data from three to four independent experiments are shown. Horizontal dashed lines represent median values, horizontal lines indicate the detection limit. *p < 0.05; **p < 0.01; n.s.: not significant, Mann–Whitney-U-test.
Details are in the caption following the image
Accelerated virus elimination in B6 mice after transfer of LCMV immune serum. B6 (circles) and CD8+ T-cell-depleted B6 mice (triangles) were infected (i.v.) with 200 PFU LCMV Docile. At day 1 p.i., mice received (i.p.) 0.5 mL of normal serum (NS, open circles) or LCMV immune serum (IS, filled circles and triangles). Viral titers in the indicated organs were determined at the indicated time points. Symbols in the graphs represent values from individual mice. Pooled data from two to three independent experiments are shown. Horizontal dashed lines represent median values, horizontal lines indicate the detection limit. *p < 0.05; **p < 0.01; ***p < 0.001; n.s.: not significant, Mann–Whitney-U-test.
Details are in the caption following the image
Accelerated virus elimination by transfer of IgG from LCMV immune serum. B6 mice were infected (i.v.) with 200 PFU LCMV Docile. At day 1 p.i., mice received i.p. 3.3 mg purified IgG from normal (NS, open circles) or immune (IS, filled circles) serum. Viral titers in the indicated organs were determined at day 8. Symbols in the graphs represent values from individual mice. Pooled data from two independent experiments are shown. Horizontal dashed lines represent the median values, horizontal lines indicate the detection limit. *p< 0.05, Mann–Whitney-U-test.

Ab transfer does not improve LCMV-specific CD8+ T-cell priming

Virus-specific Abs have been demonstrated to improve antiviral T-cell priming through the formation of immune complexes that enhance antigen presentation 18-20. We therefore compared the LCMV-specific CD8+ T-cell responses in B6 mice treated with normal or LCMV immune serum. Since viral load is known to inversely affect the magnitude of the LCMV-specific T-cell response 21, virus-specific T-cell responses were analyzed at day 6 p.i. At this time point, viral loads in both groups of mice differed only slightly. As shown in Fig. 7A, LCMV-specific CD8+ T-cell reactivity as determined by intracellular IFN-γ staining did not differ between the two groups. The same conclusion was reached when NK-cell and LCMV-specific CTL activity was examined in 51Cr release assays (Fig. 7B). Thus, transfer of LCMV immune serum did neither enhance NK-cell reactivity nor the LCMV-specific CTL response in the recipient mice.

Details are in the caption following the image
Transfer of LCMV immune serum does not enhance LCMV-specific T-cell priming. (A and B) B6 mice were infected (i.v.) with 200 PFU LCMV Docile. At day 1 p.i., mice received (i.p.) 0.5 mL of normal (NS, open circles) or LCMV immune (IS, close circles) serum. (A) At day 6 p.i., the LCMV-specific T-cell response of splenocytes was determined by KLRG1 staining and by intracellular IFN-γ staining. Symbols in the graphs represent values from individual mice. Pooled data from four independent experiments are shown. (B) LCMV-specific CTL and NK-cell activity from spleens of normal (open symbols) and immune serum (filled symbols) treated LCMV Docile infected B6 mice at day 6 p.i. CTL activity was determined at 5 hours 51Cr release assays using LCMV-infected (circles) and noninfected MC57G target cells (diamonds). YAC-1 target cells were used to determine NK-cell activity. The graphs show percent-specific lysis at the indicated effector-to-target cell ratios. Data are pooled from two independent experiments with two mice per group per experiment. n.s.: not significant, Mann–Whitney-U-test.

Antiviral activity of LCMV NP specific mAbs

The observation that the LCMV immune sera used in our experiments predominantly contained Abs specific for LCMV NP prompted us to ask whether NP-specific Abs per se show anti-viral activity. To address this point, LCMV Docile infected B6 mice were treated 1 day after infection with LCMV NP specific mAbs and viral titers were determined at day 8 p.i. Indeed, treatment of mice with these Abs significantly decreased viral titers compared with controls (Fig. 8A). Viral titer reduction was most prominent in liver followed by that in the lungs and spleen. Importantly, reduction of viral titers was observed with two different LCMV NP specific mAbs of mouse (KL53, IgG2a) and rat (VL-4, IgG2b) origin. As expected, both NP-specific mAbs did not exhibit virus neutralizing activity (data not shown) confirming previous findings 13, 22, 23.

Details are in the caption following the image
Accelerated virus elimination by transfer of LCMV NP-specific mAbs. (A) B6 mice were infected (i.v.) with 200 PFU LCMV Docile. At day 1 p.i., mice received (i.p.) NP-specific mAb KL53 (500 μg), NP-specific mAb VL4 (700 μg) or remained untreated (ctrl). Viral titers in the indicated organs were determined at day 8. Symbols in the graphs represent values from individual mice. Pooled data from two to three independent experiments with each mAb are shown. Horizontal dashed lines represent the median values, horizontal lines indicate the detection limit. *p < 0.05; **p < 0.01; Mann–Whitney-U-test. (B) Detection of LCMV NP epitopes on the cell surface of LCMV-infected MC57G fibrosarcoma cells. Cells were stained with mAb KL53 and KL25 specific for LCMV NP and GP, respectively. Isotype controls are shown in gray. Representative flow cytometry profiles from one experiment representative of three independent experiments performed are depicted.

LCMV NP represents the most abundant internal viral protein present in both infected cells and virions. The finding that Abs against this viral protein were capable to enhance viral clearance in vivo was therefore unexpected. However, epitopes of LCMV NP could be detected on the cell surface of LCMV-infected MC57G fibrosarcoma cells by flow cytometry using the LCMV NP specific mAb KL53 (Fig. 8B, left). The same result was obtained with the LCMV NP specific mAb VL4 (data not shown). The NP staining intensity was lower compared with staining with the LCMV GP-specific mAb KL25 (Fig. 8B, right) but nonetheless, it was clearly evident. Hence, epitopes of LCMV NP were present on the cell surface of infected cells and Abs specific for these epitopes enhanced virus clearance in vivo although they lacked virus neutralizing activity in vitro.

Role of FcγRs and C3 complement

To determine whether activating FcγR or complement were required for the antiviral effect of LCMV-specific Abs, mice deficient in the FcRγ chain or the complement component C3 were used. Similar to the findings described above with B6 mice, treatment of LCMV-infected FcRγ−/− or C3−/− mice with LCMV immune serum or the NP-specific mAb KL53 considerably lowered viral load in spleen, lungs, and liver compared with that in mice treated with normal serum (Fig. 9A and B). The overall reductions in viral titers by the Ab transfers were comparable in FcRγ−/−, C3−/−, and B6 wild-type mice (Fig. 9A and B versus Fig. 5 and 8). To exclude compensatory mechanisms between these two innate pathways, we repeated the anti-NP mAb transfer experiments with mice deficient for both C3 and FcRγ. As shown in Fig. 9C, the transfer of LCMV NP specific Ab also accelerated LCMV clearance in FcRγ−/−C3−/− double-deficient mice. Moreover, transfer of LCMV NP specific mAb also decreased viral titers in LCMV-infected FcRγ−/−FcγRIIB−/− double-deficient mice indicating that FcγRIIB was also dispensable for the antiviral activity of these Abs (Fig. 9D). Taken together, these data indicated that neither FcγR nor complement component C3 were required for the antiviral activity of the transferred LCMV NP-specific Abs.

Details are in the caption following the image
LCMV-specific Abs exert antiviral activity in the absence of C3 and FcγR. (A) FcRγ-chain−/−, (B) C3−/−, (C) FcRγ−/−C3−/−, (D) FcRγ−/−FcγRIIB−/− mice were infected (i.v.) with 200 PFU LCMV Docile. At day 1 p.i., mice received i.p. 0.5 mL of normal (NS, open circles), 0.5 mL LCMV immune serum (IS, filled circles) or 500 μg NP-specific mAb KL53 (diamonds). Viral titers in the indicated organs were determined at day 8. Symbols in the graphs represent values from individual mice and data shown are pooled from two to three independent experiments performed with 2–3 mice per genotype and treatment. Horizontal dashed lines represent the median values, horizontal lines indicate the detection limit. *p < 0.05; **p < 0.01; ***p < 0.001; n.s.: not significant, Mann–Whitney-U-test.

Discussion

Here, we demonstrate in the LCMV infection model that the requirement for adaptive humoral immunity in addition to CD8+ T cells is strongly dependent on the replication speed of the viral strains used for inoculation. An adaptive Ab response was required to control infection with the rapidly replicating Docile strain but was dispensable for other strains with lower replication speed. To provide direct evidence that LCMV-specific Abs assisted virus elimination, Ab transfer experiments were performed. The experiments showed that IgG Abs isolated from LCMV immune serum possessed antiviral activity in vivo. These Abs were mainly directed against LCMV NP and completely lacked virus neutralizing activity. The antiviral activity of NP-specific Abs could be further demonstrated using mAbs with single antigen specificity. The mechanism by which LCMV NP specific Abs accelerate virus elimination is not yet known. The data shown here using FcRγ−/−, C3−/−, double-deficient FcRγ−/− C3−/−, and FcRγ−/−FcγRIIB−/− mice nonetheless exclude mechanisms that involve FcγR and complement.

The strong LCMV NP specific Ab response after low-dose infection is likely due to potent LCMV-specific CTL response that leads to lysis of infected cells and release of cell internal viral proteins 14. We are not aware of any previous data on the biological role of LCMV NP specific Ab in infection but our findings in the LCMV model are reminiscent of previous work in the influenza virus system. Similar to our observations, influenza NP specific Abs have been shown to decrease viral titers in the lungs after adoptive transfer 24, 25. The underlying mechanisms, however, appear to be distinct. In contrast to our data, the antiviral activity of the transferred influenza NP-specific Abs was dependent on host FcγR expression and injection of NP-specific Abs also enhanced the NP-specific CTL response in the influenza system 25.

Remarkably, we could detect LCMV NP epitopes on the cell surface of intact LCMV-infected MC57G fibrosarcoma cells with NP-specific mAbs. Similar positive staining results were also obtained with LCMV-infected L929 cells and with other viral strains such as WE or clone 13 (data not shown). Moreover, we used two different LCMV NP specific mAbs rendering the possibility that this result was due to a peculiar cross-reactivity of the reagents very unlikely. Of note, the presence of LCMV NP epitopes on the surface of infected cells and virions has been described more than 20 years ago by Lehmann-Grube and colleagues 23. However, follow-up studies based on this surprising observation were never published. Thus, it is not yet understood why NP or fragments of this protein can be detected on the surface of intact cells or virions. LCMV NP represents the most abundant internal viral protein in both infected cells and virions. Adsorption of NP released by necrotic or killed infected cells onto the cell surface of intact cells or virions may represent one possible explanation for these findings. Interestingly, presence of influenza virus NP epitopes on the surface of infected cells has also been described long time ago but the underlying mechanism is nonetheless still obscure 26, 27. Hence, in both viral systems, epitopes of internal proteins usually associated with the viral RNA can be found on the surface of infected cells and corresponding Abs facilitate viral elimination in vivo although they are unable to directly prevent virus entry into host cells. Bergthaler et al. showed previously that clearance of high-dose LCMV WE infection in B6 mice was dependent on the generation of antigen-specific Abs 9. Ab transfer experiments in this study were, however, only performed with the virus neutralizing mAb KL25 specific for LCMV GP.

Interestingly, we observed that neither complement component C3 nor FcγR were required for the antiviral activity of the transferred nonneutralizing LCMV-specific Ab. Even in FcRγ−/−C3−/− and FcRγ−/−FcγRIIB−/− double-deficient mice, LCMV NP specific Abs were capable to exert their antiviral activity. These data confirm and extend previous work showing that C3/C4- or FcγR-deficient mice cleared high-dose LCMV WE infection with the same kinetics as wild-type mice 9. In contrast to these findings, the antiviral activity of nonneutralizing LCMV GP specific Abs has been shown to be dependent on complement 28. These data were derived from a B-cell receptor transgenic model based on the “neutralizing” LCMV GP specific mAb KL25 and viral Ab escape variants.

Antiviral activities of nonneutralizing Abs are well known and have been demonstrated in many other infection models 29-39. Such Abs may function autonomously 40, 41 or in conjunction with host components such as the complement system or FcγR-bearing cells 42-48. In all of these studies, the Abs were directed against viral envelope proteins expressed at high levels on the surface of virions or infected cells. This is distinct from our conditions analyzing the role of Abs specific for an internal viral protein that is predominantly present inside of virions and infected cells.

Antigen-IgG immune complexes are known to enhance T-cell priming by induction of dendritic cell maturation and improved antigen presentation 49. Short passive immunotherapy with neutralizing Abs has further been shown to enhance the CTL responses in mice infected shortly after birth with an ecotropic retrovirus derived from Friend murine leukemia virus 19. In our experimental system, transfer of LCMV immune serum did not increase the LCMV-specific CTL response rendering it unlikely that that the accelerated virus elimination we observed was due to increased CD8+ T-cell priming.

There is no doubt that T cells are essential for immunity against non- or poorly cytopathic viruses such as HCV or HIV in humans or LCMV in mice and that Abs on their own are unable to combat these infection. Nonetheless, our study performed in a prototypic CD8+ T-cell-controlled virus infection model unravels a role for nonneutralizing Abs specific for an internal viral protein. As exemplified with our experiments, these Abs generated in the early phase of the infection may shift the delicate balance from insufficient virus elimination and T-cell exhaustion to virus control and memory T-cell formation. In the accompanying publication by Richter and Oxenius 50, LCMV binding but nonneutralizing Abs were also shown to protect mice from chronic LCMV infection independently of activating FcγR or C3 complement. In this context, it is noteworthy that Ab-dependent cell-mediated cytotoxicity and not broadly neutralizing Ab or T-cell responses correlated with protective activity in the HIV-1 vaccine trial RV144 51. Our study encourages attempts to examine the role of nonneutralizing Abs specific for internal viral proteins also in viral infections in humans that often lead to pathogen persistence and T-cell exhaustion.

Materials and methods

Mice

C57BL/6J (B6), SWISS, and NMRI mice were obtained from Janvier. B-cell receptor transgenic MD4 mice 15, IgMi mice 17, FcRγ−/− 52, FcRγ−/−FcγRIIB−/− double deficient mice 53, and C3−/− mice 54 all back-crossed to B6 mice have been described previously. FcRγ−/− C3−/− mice were generated by breeding in our animal facility. Breeding pairs of MD4 and C3−/− mice were obtained from Dr. Christian Kurts (Bonn) and from Dr. Admar Verschoor (Munich), respectively. Mice were bred and kept in our animal facility under specific pathogen-free conditions. Animal care and use was approved by the Regierungspräsidium Freiburg.

Viruses and infections

LCMV Armstrong, LCMV WE, and LCMV Docile were propagated on baby hamster kidney cells, L929, and Madin Darby canine kidney cells, respectively. Viral titers were determined by standard focus-forming assay using serial dilutions of tissue homogenate and MC57G fibrosarcoma cells as described 55. Mice were infected i.v. with 200 PFU of the respective virus strain. MC57G fibrosarcoma or B16 melanoma cells were infected with LCMV Docile in vitro with multiplicity of infection (m.o.i.) of 0.01. Cells were harvested after 48–72 hours.

Ab treatments

LCMV immune serum was collected from 8–10 weeks old SWISS or NMRI mice 20 days after infection with 200 PFU LCMV Docile using BD Microtainer SST Tubes (BD Bioscience). Sera were used as pools from 20–40 mice and tested for LCMV titers and virus neutralizing activity using focus-forming assay as described 55. Only LCMV immune sera free of infectious virus were used. Normal mouse serum was purchased from Harlan Laboratories. Mice were treated (i.p.) with 500 μL of immune or normal serum at day 1 after infection with 200 PFU LCMV-Docile. IgG from LCMV immune serum was purified using HiTrap Protein G HP 1 mL columns (GE Healthcare) with the Amersham Biosciences UPC-900 FPLC. Purified IgG from normal mouse serum was purchased from Innovative Research. Mice were treated (i.p.) with 3.3 mg purified IgG in 0.4 mL of PBS. LCMV NP specific mAbs were derived from the mouse IgG2a secreting hybridoma KL53 23 or from the rat IgG hybridoma VL-4 55. Mice were given (i.p.) 500 μg KL53 mAbs (ascites fluid or concentrated hybridoma supernatant) or 700 μg purified VL4 mAbs (BioXcell). For CD8+ T-cell depletion, mice were treated (i.p.) with 400 μg anti-CD8 mAbs (YTS169) at d1 and d2 before infection.

Flow cytometry

The following mAbs were obtained from BD Biosciences or eBiosience: anti-CD8α (53–6.7), anti-KLRG1 (2F1), anti-PD1 (J43), anti-2B4 (ebio244F4). LCMV GP and LCMV NP on the surface of infected cells were stained with primary mAb KL25 56 or mAb KL53 23 derived from hybridoma supernatant followed by anti-mouse IgG-Alexa647 (Invitrogen) as a secondary Ab. Samples were analyzed using FACSCalibur or LSRFortessa flow cytometer (both BD Biosciences) and FlowJo software (Tree star).

LCMV-specific Abs

For detection of LCMV-specific IgG, 96-well high-binding ELISA plates (Greiner bio-one) were coated with 100 μL per well rabbit anti-LCMV immune serum diluted 1:2000 in PBS at 4°C overnight. Afterwards, plates were incubated (1 hour) at room temperature with LCMV particles (∼105 PFU/well) present in the supernatant of LCMV-infected MC57G cells. Plates were then washed four times with PBS containing 0.05% Tween-20. Serum sample were diluted 1:300 in PBS and a threefold dilution series was performed. A total of 100 μL per well of the serum dilution was transferred to the LCMV-coated plates. After 1 hour of incubation at room temperature, plates were washed four times, followed by incubation with 100 μL per well of HRP-conjugated goat-anti-mouse IgG (Jackson ImmunoResearch) diluted 1:30 000 in PBS, followed by 1 hour incubation. Thereafter, plates were again washed four times and 50 μL per well of the peroxidase substrate OPD (SIGMA) were applied and the color reaction stopped after 10 min by adding 100 μL per well of 2 M sulfuric acid. OD was determined at a wavelength of 492 nm. LCMV-specific Ab titers were determined by an endpoint titer 0.1 OD over background. To determine the viral antigen specificity of these Abs, cell lysates of LCMV-infected and noninfected B16 melanoma cells were immunoprecipitated with IgG from LCMV immune serum that were bound to protein G-coupled sepharose (GE Healthcare). Samples were separated by 4–12% gradient SDS-PAGE (SERVA) and visualized with rabbit anti-LCMV serum (1:5000), followed by HRP-conjugated donkey anti-rabbit IgG (Dianova). The ECL plus detection system (GE Healthcare) was applied for visualization.

Functional T cell assays in vitro

Single-cell suspensions of splenocytes were obtained by mechanical disruption. IFN-γ production of CD8+ T cells was determined by intracellular IFN-γ staining (anti-IFN-γ; clone XMG 1.2, ebioscience) after restimulation of 106 splenocytes with 10−7M LCMV GP33 peptide or LCMV NP396 peptide in the presence of 10 μg/mL Brefeldin A (SIGMA). CTL- and NK-cell activity was determined in a 51Cr-release assay. Target cells were loaded with 51Cr for 2 hours at 37°C and then incubated for 5 hours at 37°C with splenocytes that were previously titrated in a threefold dilution series. Duplicate wells were assayed for each effector-to-target ratio and percentages of specific lysis were calculated.

Statistical analysis

Data were analyzed using SigmaPlot Version 9.0 software. Significant differences were evaluated with Mann–Whitney U-test using InStat3 software (GraphPad).

Acknowledgements

The authors thank Maike Hofmann for helpful discussions and critical comments on the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft DFG (Pi295/6-1 to H.P. and SFB490 to A.W.).

    Conflict of interest

    The authors declare no financial or commercial conflict of interest.

    Abbreviations

  1. FcγR
  2. Fcγ receptor
  3. GP
  4. glycoprotein
  5. KLRG1
  6. killer lectin-like receptor G1
  7. LCMV
  8. lymphocytic choriomeningitis virus
  9. NP
  10. nucleoprotein
  11. p.i.
  12. postinfection
    • The full text of this article hosted at iucr.org is unavailable due to technical difficulties.