Report

Control of Early Viral and Bacterial Distribution and Disease by Natural Antibodies

See allHide authors and affiliations

Science  10 Dec 1999:
Vol. 286, Issue 5447, pp. 2156-2159
DOI: 10.1126/science.286.5447.2156

Abstract

Natural antibodies are often dismissed from immunological analysis as “background,” but they may play an important role in conferring immunity against infections. In antibody-free mice infected with various viruses or with Listeria monocytogenes, viral or bacterial titers in peripheral organs, including the kidney and brain, were 10 to 100 times greater than in antibody-competent mice (and enhanced their susceptibility to some infections), and titers in secondary lymphoid organs were 10 to 100 times lower than in antibody-competent mice. Thus, natural antibodies play a crucial role by preventing pathogen dissemination to vital organs and by improving immunogenicity through enhanced antigen-trapping in secondary lymphoid organs.

The humoral immunity of naı̈ve animals consists of B cells secreting “natural” antibodies representing the spontaneous repertoire of circulating immunoglobulins (Ig's) that mainly belong to the IgM class, although IgG and IgA natural antibodies have also been described (1). A major source of natural antibodies seems to be the peritoneal CD5+ B-1 cell subset (2). Circulating antibodies against toxins, bacteria, and erythrocytes in the sera of naı̈ve animals were described very early in the history of immunology (3). Nevertheless, despite their abundance and early detection, their role, particularly in host protection, remains unclear. Natural antibodies may form part of innate immunity (4); may facilitate antigen uptake, processing, and presentation by B lymphocytes via complement and Fc-receptors (4); may induce or prevent autoimmune diseases (1, 5); may clear lipopolysaccharides (6) and protect against Streptococcus pneumoniae (7); or may be involved in a hypothetical immunoregulatory idiotype–anti- idiotype network (8).

In many virological and immunological assays it is a convenient practice to predilute sera to variable extents, from 1:10 up to 1:80, to avoid so called “nonspecific background” signals; the consequence of this practice is that naturally occurring antibodies are not normally detected or are not registered.

It is recognized that during infections with many cytopathic viruses (such as polio, influenza, and rabies viruses), the presence of early neutralizing antibodies is essential for protection against lethal disease, which often correlates with viral replication in neural tissues (9). The possible role of natural antibodies in viral infections has largely remained unexplored. We found a key role of natural antibodies during generalized infections with vesicular stomatitis virus (VSV), lymphocytic choriomeningitis virus (LCMV), vaccinia virus (vacc-WR), or Listeria monocytogenes.

Sera from naı̈ve mice assessed by enzyme-linked immunosorbent assays (ELISAs) on purified LCMV- or vaccinia virus–coated plates (10) revealed LCMV-specific natural IgM, but not IgG, at a titer of 1:16 (Fig. 1A) or vaccinia-specific natural IgM at a titer of 1:64 (Fig. 1B). LCMV-specific natural antibodies did not detectably neutralize the virus in vitro (11), and only LCMV-binding natural antibodies could be assessed (Table 1).Listeria-specific natural antibodies were assessed by fluorescence-activated cell sorting (FACS) analysis (Fig. 1C) (12); specific natural IgM, but not IgG, antibodies were detectable up to a dilution of 1:16. Sera from naı̈ve C57BL/6 mice were tested for antibodies specific for two different VSV serotypes: VSV–New Jersey (VSV-NJ) and VSV-Indiana (VSV-IND) (10). Because VSV glycoprotein (VSV-G) exhibits the neutralizing determinant at its tip, ELISAs with intact VSV measure essentially neutralizing antibodies (Fig. 1, D and E). In all cases, only IgM was detected. Natural G-binding (neutralizing) antibodies to intact VSV-IND did not cross-react with intact VSV-NJ, as shown by the fact that preabsorption of sera with VSV-IND reduced only VSV-IND–specific, not VSV-NJ–specific, natural antibody titers and vice versa (Fig. 1, D and E).

Figure 1

Detection of natural antibodies. Sera from unimmunized mice were tested in an ELISA (11) on purified LCMV-coated (A) or purified vacc-WR–coated (B) plates for the presence of specific IgM or IgG. (C) Listeria (5 × 106 CFU) was incubated with 50 μl of prediluted sera (in 1:2 dilution steps; only the 1:4 dilution is shown) from naı̈ve C57BL/6 or μMT mice, stained with fluorescein isothiocyanate–labeled goat anti-mouse IgMb (from PharMingen) and analyzed by FACS (12). Sera of SPF mice were preabsorbed with equal volumes of purified intact VSV-IND or VSV-NJ (≈1011PFU/ml) overnight (D and E) Thereafter, sera were tested either on plates coated with purified VSV-IND (D) or with VSV-NJ (E) virus particles. The mean of three animals per group is shown; variations in the groups were within plus or minus one titer step of 1:2.

Table 1

VSV- and LCMV-neutralizing antibody titers in naı̈ve mice. ND, not determined.

View this table:

Sera from naı̈ve, conventional, specific pathogen–free (SPF) and germ-free (GF) C57BL/6 mice protected in vitro against lytic infection with VSV-IND and VSV-NJ up to a dilution of 1:16 to 1:32 (Table 1), regardless of the hygiene status of the mice. To assess the role of self antigens on natural antibodies, neutralizing natural antibodies were measured in sera of naı̈ve mice that express the glycoprotein of VSV-IND as a transgene under the control of the H-2Kb promotor (KINDG mice) but nevertheless possess VSV-IND–specific B cells that can be induced by infection with VSV (13). KINDG mice exhibited no VSV-IND–specific antibodies but showed normal levels of natural VSV-NJ–specific antibodies (Table 1). Neutralization of the virus by natural antibodies was always mediated by IgM, because reduction with 0.1 mM β-mercaptoethanol almost completely inhibited neutralization. Thus, (i) VSV-specific natural antibodies are serotype-specific in vivo, and (ii) low levels of antibodies are produced by VSV-specific B cells independently of intentional stimulation by external or self antigens.

The influence of natural antibodies on early infection kinetics was tested either in mice totally lacking antibody production (μMT and RAG-1−/− mice) (14) or in mice that lacked only IgM but did produce all other Ig subclasses (IgM−/−) (15). We studied virus dissemination between 2 and 8 hours after infection (with the exception of the slowly replicating LCMV, because virus was not detectable in peripheral organs before 24 hours after infection), because this time interval was too short for a specific IgM antibody response to be induced. After infection with 2 × 106 plaque-forming units (PFU) of LCMV-WE, the early trapping of virus particles to the spleen in mice producing antibody was more efficient than in μMT animals (Fig. 2A). The reduced early recruitment of virus to the spleen led to virus titers in the blood and peripheral organs at 24 hours after infection that were up to 10 to 50 times greater than those in antibody-producing mice (Fig. 2B). Infection with 5 × 107 PFU of vaccinia virus or 2 × 107 colony-forming units (CFU) of Listeriarevealed a comparable effect of natural antibodies on the early dissemination of these pathogens (Table 2).

Figure 2

Viral and bacterial distribution in antibody-deficient and antibody-competent mice. μMT, RAG-1−/−, and C57BL/6 (control) mice were infected with 2 × 106 PFU of LCMV-WE i.v., and virus titers in the spleen were determined per organ at the time points indicated (A) (10). LCMV titers in peripheral organs and in the blood were detectable 24 hours after infection with 2 × 106 PFU and are shown in (B). Antibody-deficient mice and C57BL/6 control mice were infected with 2 × 108 PFU of VSV-IND i.v. [subcutaneous injection did not yield detectable virus titers (9)]. VSV titers were determined per spleen (C) and per kidney (D). In (A) through (D), means ± SD of three mice per group are shown, and one of three to five comparable experiments is shown.

Table 2

Viral and bacterial distribution in antibody-deficient and antibody-competent mice. ND, not determined.

View this table:

The role of natural antibodies in virus dissemination and disease was analyzed in detail after infection with VSV. Mice deficient in antibody production showed elevated titers in peripheral nonlymphatic organs such as the liver, kidney, and brain but showed reduced titers in the spleen when compared with antibody-competent mice (Fig. 2, C and D, and Table 2). The replication of VSV in vivo is stringently controlled by interferon-α (IFN-α) (16); however, comparison of IFN-α/β receptor−/−[A129 (16)] mice with IFN-α/β receptor−/−/RAG-1−/− (AR129) mice revealed that IFN-α/β was not responsible for the observed titer reduction in the lymphatic organs of antibody-deficient mice. IgM−/− mice (15) exhibited higher titers in peripheral nonlymphoid organs but had a titer in the spleen that was two to three times lower as compared to that in control mice, verifying that IgM was largely responsible for the findings (Table 2).

Reconstitution experiments with μMT and RAG-1−/−mice were performed with 0.25 to 1 ml of naı̈ve serum (Fig. 3A). This substitution, as compared with untreated controls, prevented VSV infection of peripheral organs in a dose-dependent manner and increased viral titers in the spleen. Recruitment to the spleen did not reach the levels found in control spleens, however, for at least three possible reasons: First, the spleens of μMT and RAG-1−/− mice are about two to three times smaller than C57BL/6 spleens; second, the disrupted splenic architecture of B cell–deficient mice (17) may contribute to a reduced filtration capacity; third, 1 ml of naı̈ve serum reconstitutes natural antibodies only to about 30 to 50% of control levels. Additional experiments showed that complement inactivation with cobra venom factor revealed no obvious influence on natural antibody–dependent virus distribution (Fig. 3B).

Figure 3

Analysis of virus distribution after reconstitution with naı̈ve serum. (A) μMT mice were reconstituted with 1 ml, 0.5 ml, or 0.25 ml of pooled sera from naı̈ve SPF C57BL/6 mice or with 1 ml of pooled sera from μMT mice as controls. Thirty minutes later, these mice and C57BL/6 mice were infected with 2 × 108 PFU of VSV i.v. VSV titers were determined 2 hours after infection. Bars indicate the mean ± SD of three mice per group. (B) Complement inactivation was performed with 1.2 μg of purified cobra venom factor (Sigma number 233550) injected twice on day 1 (at 16 hours and at 24 hours); this protocol was checked in a complement-mediated hemolysis assay (28). μMT mice with or without complement inactivation were then reconstituted with 1 ml of heat-inactivated pooled sera from C57BL/6 mice 30 min before infection. (C) μMT mice were reconstituted once with 2 ml of pooled naı̈ve serum (1 ml i.v. and 1 ml i.p), with 0.5 ml of pooled serum daily, or left untreated. Thirty minutes later, these mice were infected i.v. with 104 PFU of VSV-IND (about 10 times LD50). One of three to four experiments is shown.

Natural antibodies protected against paralytic disease caused by VSV; the median lethal dose (LD50) in μMT mice was found to be around 103 PFU of VSV-IND as compared with 108 PFU in controls. Enhanced susceptibility to VSV of B cell–depleted (3) or of μMT mice versus controls has been observed previously (18). To test the effect of natural antibodies on the course of a VSV infection, μMT mice were injected once with 1 ml of naı̈ve serum intravenously (i.v.) and 1 ml of naı̈ve serum intraperitoneally (i.p.), or alternatively were injected daily with 0.5 ml of naı̈ve serum for 5 days [to compensate for the short in vivo half-life of IgM antibodies of about 20 hours (19)] and then challenged with 104 PFU of VSV (about 10 times the LD50). All control μMT animals died within 12 days after infection, whereas seven out of eight of those repetitively supplemented and three out of four of the mice given one large dose of normal serum remained healthy during the observation period of 60 days. In addition, the in vivo specificity of natural antibodies was revealed by the heightened susceptibility, by a factor of about 100, of KINDG mice when compared to controls. All KINDG mice died when infected with 108 or 107 PFU, and one out of three died when infected with 106 PFU of VSV, as compared to only one out of three C57BL/6 control mice infected with 108 PFU.

Natural antibodies enhance the early recruitment of virus to secondary lymphoid organs after infection with viruses that are known to be controlled by antibodies, such as VSV (9), or partially controlled, such as vaccinia virus (20), but they also enhance recruitment after infection with pathogens for which antibodies are not thought to be of importance in a primary infection, such as LCMV (21) and Listeria (22). Because no specific Fc receptors for IgM are known, and a direct role of complement was not demonstrable here (Fig. 3B), the formation of virus-antibody complexes, increasing particle size, may be an important factor enhancing phagocytosis and hence uptake from the blood by the spleen. This has been suggested earlier by in vivo studies (23) and by in vitro experiments showing that uptake by macrophages is more efficient for large antigen particles or complexes than for small particles (24).

A specific B or T cell response is induced in secondary lymphatic organs (25). An important protective function of natural antibodies may therefore be not only to influence initial virus distribution and early dissemination to avoid the infection of vital target organs, but also to direct the virus to secondary lymphatic organs in order to accelerate and enhance immune responses. The delayed neutralizing IgG response in IgM−/− mice (15) supports this notion. In antibody-deficient μMT mice, the reduction in early filtration of LCMV to the spleen delayed cytolytic T lymphocyte (CTL) activation and allowed the virus to spread to peripheral organs, with the consequence that μMT mice could not control high-dose infections and the CTL response in these mice was exhausted (26). In addition, it was shown that μMT mice are resistant to intraperitoneal prion infections (17) and that therefore natural or induced antibodies (or B) cells may play a role in prion disease. Collectively, the data show that natural antibodies may well be involved in the recruitment of many pathogens to secondary lymphoid organs and, together with earlier observations (27), illustrate an interesting link between innate immunity and the acquired immune responses.

REFERENCES AND NOTES

View Abstract

Navigate This Article