Adrian Bot, Stefan Antohi and Constantin Bona

The Department of Microbiology, Mount Sinai Medical School, New York, USA

Received 4/8/97; Accepted 4/15/97; On-line 5/1/97

3. IMMUNE RESPONSIVENESS OF THE NEONATAL B CELLS

Susceptibility of the neonatal B cells to high-dose tolerance was previously documented in various experimental systems. Neonatal injection with antibodies specific for immunoglobulin (Ig) antigenic determinants caused a long lasting suppression of the B cells that express the corresponding determinant (6-8). Furthermore, injection of large amounts of bacterial polysaccharides (4,5) or T-dependent antigens (9) impaired the subsequent synthesis of specific antibodies.

However, Howard and Hale were among the first to show that in certain conditions, adult animals injected as newborns with lower amounts of bacterial polysaccharides can mount an antibody response (21). Studies carried out later demonstrated that some antibody responses and the expression of their corresponding idiotypes can be elicited in 1-14 day-old mice subsequent to immunization with various antigens (Table 1).

Notably, recent studies showed that some immunization protocols that led to T cell unresponsiveness in terms of proliferation, were, in fact, followed by humoral responses (15, 16). Due to the fact that low antigen doses are associated with Th1 responses while higher doses are followed by poor proliferative Th2 responses (18, 19), one should expect different isotype profiles in T-dependent humoral responses, depending on the antigen dose inoculated in newborn organisms. It becomes apparent that the mechanisms of neonatal B and T cell tolerance are different. Whereas the induction of neonatal B cell tolerance may require higher antigen doses that can directly turn off the antigen-specific B lymphocytes, the neonatal T cell tolerance is a modified response rather than a true unresponsiveness.

This hypothesis is supported by the predominance of immature B cells in the neonatal repertoire. Indeed, neonatal B cells differ phenotypically from their mature counterparts since they are B220^hi, sIgM^hi, sIgD^-, whereas most adult B cells are B220^hi, sIgM^lo and sIgD^hi (28). The predominance of immature B cells (sIgM⁺IgD^-) was considered the major cause to explain why the neonatal B cells are more susceptible to tolerance induction by various T-independent antigens (9, 29). An early study showed that in vitro tolerance induction required 10³ times more antigen in the case of adult B cells as compared to neonatal B cells (30).

Several hypotheses were entertained to explain the unresponsiveness of the immature B cells, which are dominant in the periphery of newborns. The most prevailing hypothesis considers that the differences in isotype expression of the Ig receptors are responsible for the susceptibility of the immature B cells to tolerance. Thus, while the binding of antigens to sIgM only causes anergy or deletion, the interaction with sIgD induces a positive signal (31). Negative signaling through sIgM may be related to the fact that sIgM is not coupled downstream to the inositol-phospholipid signaling pathway (32). Furthermore, a recent study showed that 3 day-old mice have marked deficiencies in lyn, fgr and src-family tyrosine kinase expression (33). This model is supported by the observation that mature B cells stripped of IgD become more sensitive to the induction of tolerance (31). However, this model cannot easily account for the fact that both sIgM and sIgD can mediate negative selection in the transgenic mice (34). A more detailed analysis carried out in B cell receptor transgenic mice by Carsetti et al. (35) demonstrated that the subpopulation directly affected by tolerance is the late-immature (sIgM^hi, sIgD^-, HSA^bright, B220^low, CD44^hi) subpopulation rather than the dominant (sIgM^hi, sIgD^-, B220^hi) subpopulation of immature B cells.

Another hypothesis aimed at explaining the poor humoral immune responsiveness of newborns considers that neonatal B cells are unable to receive second signals from T cells, independent on their stage of differentiation. This idea is supported by recent studies which showed that human neonatal T cells have decreased expression of CD40-L and the cord-blood B cells displayed an impaired ability of isotype switch following CD40 binding (36). It is well known that the interaction between these two molecules is important for the activation and differentiation of both B and T cells (37, 38). Thus, both the abilities of B cells to undergo isotype switching and of T cells to differentiate to Th1 cells could be affected. Furthermore, the lack of CD40 - CD40-L interaction may lead to a predominant nonprofessional presentation of antigens to T cells because the up-regulation of B7 molecules on APCs is consequently impaired (39).

The discrepancy between the in vivo data (Table 1) demonstrating the ability of neonates to mount humoral responses in certain instances and the in vitro observations indicating that signaling through IgM leads to apoptosis of the immature B cells, can be explained by the presence of a few mature IgM⁺IgD⁺ B cells in the periphery of newborns. Thus, injection of antigens in neonates is followed by deletion or anergy of the immature, antigen-specific B cells and priming of few differentiated B cells. Lower doses of antigens may be unable to delete most of the specific precursors but may still be able to prime some mature B cells, leading to a humoral response. If the antigen dose is above a certain threshold, most of the specific B cell precursors would be deleted and the mature B cells anergized, leading to a state of unresponsiveness with a variable duration. It is noteworthy to mention that exposure to lipopolysaccharide (LPS) which is a polyclonal stimulator for B cells, can trigger a more rapid recovery from the state of B cell unresponsiveness induced by injection of antigens into neonates (9).

3.1. Humoral responses of mice immunized as neonates with naked DNA

There are numerous studies demonstrating that immunization of adult mice (40), chickens (41), ferrets and monkeys (42, 43) with plasmids bearing HA or NP genes of various strains of type A influenza virus, can induce protective immune responses.

In order to study the neonatal humoral response to mammalian expression vectors, we immunized newborn mice with a plasmid bearing the HA gene cDNA from A/WSN/33 virus (pHA), driven by the initial-early cytomegalovirus (CMV) promoter (Fig.1A). Newborn and adult BALB/c mice were immunized three times with 100 µg of plasmid and the immune response was studied at various intervals (Fig.1B).

Figure 1. The structure of the pHA plasmid (A) and the immunization schedule with naked DNA of neonatal and adult mice (B). The cDNA of the HA was inserted into a plasmid bearing the regulatory elements (enhancer + promoter) of the initial-early genes of CMV and the polyadenylation signal of the bovine growth-hormone gene. Mice were immunized three times in the muscle with 100 µg of plasmid/dose. Blood or tissue samples were harvested at various intervals after the completion of the immunization. Some of the mice were boosted with live-virus 7 days prior to the sample harvest.

Antibodies specific for WSN virus were detected by both hemagglutination-inhibition (HI) and radio-immunoassay analysis (RIA) (44). The data presented in Fig. 2 show the presence of WSN specific antibodies as detected by HI assay at 1 and 3 months after the completion of immunization, in mice inoculated with pHA as adults or as neonates.

Serum HI titers dramatically declined between 3 and 6 months after the immunization of adult mice. Immunization of adult mice with live WSN virus resulted in HI titers that were 10-20 higher than those obtained following inoculation with pHA. In contrast, inoculation of neonates or adults with a control plasmid that did not express HA was not followed by humoral response. The mice immunized with pHA were able to mount significant secondary humoral responses following the virus boost, indicating that the B cell repertoire was not affected by the prolongued exposure to the antigen (44). The plasmid could be detected even at 9 months after the completion of immunization, in approximately 25% of the immunized mice (Fig. 2 and data not shown).

Figure 2. The HI titer of serum antibodies obtained from animals immunized with pHA as adults or newborns, at various intervals after the completion of immunization. HI titers of WSN specific antibodies, which are good correlates of the neutralization ability, were estimated in groups of 5-10 mice. The results were expressed as means of log₂ of the individual HI titers at various intervals after the immunization. Control mice were injected with the plasmid lacking the HA open reading frame (CP) or with live WSN virus, 1 week previous to the blood harvest. PCR results showing the persistence of pHA plasmid at the site of injection were considered positive (+) if at least one mouse in a given group was positive. ND - not done.

Two major differences were noted between the humoral responses of mice immunized with pHA as adults and neonates: first, the isotype profile of the WSN specific antibodies was distinct in mice immunized as neonates versus adults, after the virus boost (ref. 44 and below). Whereas IgG₂ antibodies predominated in the mice immunized as adults, the newborn immunized mice displayed increased titers of IgG₁ and IgG₃ antibodies after the virus boost. Second, the reactivity analysis of clonotypes obtained from mice immunized as adults or newborns with pHA, showed that in contrast to mice immunized as neoantes, those inoculated as adults displayed more frequent cross-reactive clonotypes (44). These differences can be accounted for by the particular T cell responsiveness of neonates and by the more restricted B cell repertoire of the young mice (45).

Mice immunized with pHA as newborns or adults and challenged with one 100% lethal dose (1LD₁₀₀) of WSN virus one month after the completion of immunization, displayed 50-80% long-time survivors (44). However, the cross-protection against influenza virus PR8 drift-variant and the pattern of virus lung-titer decrease between day 3 and 7 after the infection, suggested that T cell immunity is an important protective factor following the immunization with pHA (44).

It appears that plasmid immunization of newborn mice leads to protective humoral responses rather than unresponsiveness. Continuous exposure to small doses of antigen following plasmid immunization did not lead to deletion of the specific immature B cells or anergy of the more mature, virus-reactive B cells. In contrast, the mature B cells present during the early stages of the postnatal development were effectively primed to antibody-forming and memory B cells. Thus, whereas antibodies are continuously generated as long as the antigen persists in the organism, the animals are able to mount strong protective responses following the virus infection.