Deficient TACI Expression on B Lymphocytes of Newborn Mice Leads to Defective Ig Secretion in Response to BAFF or APRIL¹

Infants, particularly newborns, are highly susceptible to infections (1). The predisposition of newborns to infections has been ascribed to defects in both the humoral and cellular arms of the adaptive immune responses. For example, infants have a diminished pool of B and T lymphocytes both in peripheral and secondary lymphoid organs, lower serum complement levels, impaired Ag-presenting ability, low expression of costimulatory molecules, and reduced IL-12 production after infection or vaccination (2, 3, 4, 5). The ability of newborns to produce Igs is also restricted (5). More importantly, Ab responses of newborns to T cell-independent type II (TI-II)³ Ags, such as bacterial capsular polysaccharides (CPS), are deficient (5, 6). As a result, newborns are especially susceptible to infections with encapsulated bacteria such as Neisseria meningitidis, Streptococcus pneumoniae, and Haemophilus influenzae (1). Once developed, Abs against CPS Ags mediate complement-dependent opsonophagocytic bactericidal activity against encapsulated bacteria (7). However, due to its poor immunogenicity in newborns and infants, vaccines composed of CPS Ags are not protective in this age group (4, 5). The inability of newborns and infants to mount Ab against CPS Ags is attributed to the immaturity of the splenic marginal zone (MZ) (8). Newborn MZ contain decreased numbers of CD21^highCD23^low B lymphocytes, metallophilic macrophages, MZ macrophages, and immature dendritic cells (DCs). In contrast, B1 cells are found in increased numbers in newborns (9). Activation of CPS-specific B lymphocytes requires binding of C3d to CPS and cross-linking of membrane Ig and CD21/CR2 on B lymphocytes with the C3d-CPS complex (10). In newborns and infants, the deficiency of CD21/CR2 in the splenic MZ and low serum C3 levels has been implicated for the poor Ab response to CPS Ags (5, 11).

Recently discovered, the so-called “BAFF system” is comprised of two ligands, B cell-activating factor of the TNF family (BAFF) and a proliferation-inducing ligand (APRIL), and three receptors: BAFFR, transmembrane activator and calcium-modulator and cytophilin ligand interactor (TACI), and B cell maturation Ag (BCMA) (12). TACI and BCMA interact with APRIL, while BAFF binds to all three receptors. Engagement of BAFF and APRIL with their receptors on B lymphocytes has been shown to be very important for the development of the humoral immune response. For example, BAFF binding to BAFFR is required for the developmental maturation of B lymphocytes from T1 to T2 stage by providing survival signals (13, 14, 15). TACI appears to be essential for the development of an Ab response against TI-II Ags (16, 17) and may be balancing the B cell survival signals mediated by BAFFR (16). In addition, TACI and BAFFR transduce signals for isotype switching to IgG and IgA without the help of T lymphocytes (18, 19). Ligation of BCMA, in contrast, promotes plasma cell survival (20) and stimulates Ag presentation by B lymphocytes (21). These recent discoveries have helped understanding of the role of myeloid cells and the BAFF system in maintaining B lymphocyte homeostasis. However, the expression levels of these molecules in newborns have not yet been reported, and it is not yet known whether the biological functions attributed to these molecules in adults are similar in newborns. We therefore conducted this study to investigate the expression of the molecules belonging to the BAFF system in newborn mice. We determined that BAFF expression is increased and TACI expression is significantly decreased in newborn mice compared with those of adult mice. In addition, unlike those of adults, newborn murine B lymphocytes do not secrete Igs in response to BAFF or APRIL stimulation. Importantly, this unresponsiveness of B lymphocytes can be overcome by up-regulating TACI expression with CpG oligodeoxynucleotides (ODN). These results suggest that decreased TACI expression in newborns is responsible for impaired Ab response against TI-II Ags, and adjuvants such as TLRs may improve polysaccharide vaccines by increasing TACI expression.

Adult BALB/cAnNCr mice, 6–8 wk old, were purchased from Charles River Laboratories. Pregnant BALB/cAnNCr mice in late stages of pregnancy were purchased from Charles River Laboratories. TACI knockout (KO) mice on a C57BL/6 background were described previously (22). All B cell analysis was done on 5-day-old mice (newborn mice). Resident peritoneal cells were analyzed either on 5-day-old mice or on 8-day-old mice, 3 days after thioglycolate injection. Care and handling of animals was performed according to guidelines provided by the Animal Research Advisory Committee, National Institutes of Health, and Mayo Foundation Institutional Animal Care and Use Committee.

Abs against mouse cell markers and the isotype controls used in flow cytometry assay were: TACI-PE, BCMA-FITC, BAFF-PE, rat IgG1-FITC, rat IgG2a-PE (R&D Systems); CD21-FITC, CD21-PE, CD23-biotin, CD5-allophycocyanin, IgM-allophycocyanin, B220/CD45R-PE-Cy5, CD138-allophycocyanin, CD11b-FITC, rat IgG2b-FITC, rat IgG2b-PE, rat IgG2a-biotin, rat IgG2a-allophycocyanin, rat IgG2a-PE-Cy5, streptavidin-allophycocyanin-CY7 (BD Biosciences/BD Pharmingen); BAFFR-FITC, IgM-PECy7, rat IgG1-FITC, rat IgG2a-PE-Cy7 (eBioscience); rat B220/CD45R Alexa Flour 405, and IgG2a-Alexa Flour 405 (Invitrogen). Propidium iodide (50 μg/ml) was used to identify dead cells (BD Biosciences/BD Pharmingen). The following reagents were used to stimulate cells, Escherichia coli LPS (Sigma-Aldrich), F(ab′)₂ goat anti-mouse IgM (Jackson ImmunoResearch Laboratories), IL-4 (eBioscience), and TGF-β (R&D Systems). CpG ODN 1555 (sequence: GCTAGACGTTAGCGT) was synthesized at the Center for Biologics Evaluation and Research Core facility. Detection of intracellular BAFF was done by using the Fix and Perm kit (Caltag Laboratories) as described by the manufacturer. Recombinant BAFF and APRIL were purchased from Axxora. Anti-TACI polyclonal Ab and purified goat Igs were obtained from R&D Systems. The B cell isolation kit was purchased from Miltenyi Biotec.

Constitutive BAFF and APRIL expression in mouse peritoneal cells was compared in adult and newborns after thioglycolate injection. Neonatal mice were i.p. administered 0.5 ml of 3% thioglycolate while adult mice received 2.5 ml of 3% thioglycolate. Peritoneal cells were harvested 72 h after injection and adherent cells were collected after a 30-min incubation on petri dishes. Cells pooled from three to five newborns were designated as one “set.” To assess a possible effect of thioglycolate on BAFF or APRIL expression, peritoneal cells were also collected before thioglycolate injection. Expression of BAFF or APRIL in sets of newborn cells was compared with adult cells. To determine the effect of CpG ODN on BAFF or APRIL expression, cells (3 × 10⁶/ml) were also stimulated with CpG ODN for 24 h. Cells were subjected to flow cytometry for intracellular BAFF expression by using anti-BAFF-PE Abs. Abs against IgM and CD11b were also used to gate B lymphocytes and macrophages, respectively. Flow cytometry data were analyzed by using FlowJo Software (Tree Star). Macrophage BAFF and APRIL mRNA expression were analyzed in real-time PCR as described below.

The expression of BAFF receptors TACI, BAFFR, and BCMA was analyzed by flow cytometry and real-time PCR assays. Cell surface expression of receptors was assessed on purified B lymphocytes. B lymphocytes were purified from adult or neonatal mice splenocytes by using the B cell purification kit according to the manufacturer’s instructions. Two rounds of cell purification were performed to obtain >96% B cell purity of neonatal B lymphocytes. Newborn B lymphocytes, pooled from three to five newborns are named as one set. Abs against B220 were used to gate B lymphocytes. TACI and BAFFR expression was also determined on B lymphocyte subsets, newly formed (NF) (B220⁺CD23^negCD21^neg), MZ (B220⁺CD23^lowCD21^high), follicular (FO; B220⁺CD23^highCD21^int), and B1 (IgM⁺B220^lowCD5⁺) cells, which were gated according to published phenotypic descriptions (23, 24). In experiments requiring prestimulation of cells, purified B lymphocytes were resuspended in complete RPMI 1640 (RPMI 1640 (cellgro) supplemented with 10% heat-inactivated FBS, 2 mM l-glutamine, 10,000 U/ml penicillin and streptomycin, 10 mM HEPES, 1 mM sodium pyruvate, and 1 mM nonessential amino acids (Invitrogen)) medium at 1–2 × 10⁶ cells/well density in 24-well plates and stimulated with medium, CpG ODN (1 μg/ml), or anti-IgM (100 μg/ml) plus IL-4 (50 ng/ml). Cells were incubated in a CO₂ (5%) incubator at 37°C for 24 h, after which they were harvested, labeled, and analyzed for the respective cell surface markers. Receptor expression was also analyzed on CpG- or PBS-injected neonatal mice splenocytes. For this purpose, neonatal mice were i.p. injected with 100 μl of CpG ODN (200 μg/ml) or 100 μl of PBS. Twenty-four hours later, spleens were harvested and TACI, BAFFR, and BCMA expression was measured on B220⁺-gated splenocytes. Quantitative analysis of B lymphocyte receptor mRNA expression in real-time RT-PCR was performed as described below.

The percentages of adult and newborn mice plasma cells were determined by measuring CD138-expressing B220-positive cells in a flow cytometry assay as previously described (25, 26). Purified B lymphocytes, pooled from five newborns, were designated as one set. The development of plasma cells in adult and sets of newborn B lymphocytes were assessed by incubating B lymphocytes with BAFF (1 μg/ml) plus anti-IgM (100 μg/ml), APRIL (1 μg/ml) plus anti-IgM (100 μg/ml), or LPS (10 μg/ml). After 72 h, the percentage of B220⁺CD138⁺ was measured on cells that did not uptake propidium iodide (live cells). The effect of CpG ODN prestimulation in sensitization of B lymphocytes to BAFF- or APRIL-induced plasma cell development was also determined. Adult or newborn B lymphocytes (2 × 10⁶/ml) were incubated with CpG ODN for 24 h. Subsequently, cells were counted and equal number of B lymphocytes (0.5 × 10⁶/ml) were stimulated for 72 h with BAFF (1 μg/ml) plus anti-IgM (100 μg/ml), APRIL (1 μg/ml) plus anti-IgM (100 μg/ml) or LPS (10 μg/ml). Plasma cell development was assessed by measuring the percentage of B220⁺CD138⁺ on propidium iodide-negative cells.

BAFF- and APRIL-induced Ig secretion was compared in adult and neonatal mice. Purified B lymphocytes, pooled from three to five newborns, were designated as one set. Naive, CpG ODN-injected, or PBS-injected neonatal mice B lymphocytes were seeded into plates at a density of 1 × 10⁶/ml for stimulation. For the stimulation of B lymphocytes from naive mice, IL-4 (50 ng/ml) was included in BAFF (1 μg/ml) or APRIL (1 μg/ml) containing wells. When stimulating B lymphocytes from CpG ODN- or PBS-injected mice, IL-4 was excluded from BAFF or APRIL stimulation conditions. In separate experiments, each set of newborn B lymphocytes from naive mice was first incubated in vitro with medium, CpG ODN (1 μg/ml), or anti-IgM (100 μg/ml) plus IL-4 (50 ng/ml) for 24 h after which equal numbers of B lymphocytes (0.5 × 10⁶/ml) were restimulated with BAFF (1 μg/ml) or APRIL (1 μg/ml). All B lymphocyte stimulation studies also included stimulation conditions composed of LPS (10 μg/ml) and IL-4 (50 ng/ml), or LPS (10 μg/ml) and TGF-β (50 ng/ml), as control stimuli. After 6 days of stimulation, culture supernatant Ig isotype levels were determined in ELISA. Blocking goat anti-TACI Abs (27) were also used when stimulating CpG ODN pretreated newborn B lymphocytes with BAFF or APRIL. Newborn B lymphocyte sets were incubated with increasing concentrations of polyclonal anti-TACI Abs or control goat Igs together with BAFF (1 μg/ml) or APRIL (1 μg/ml). After 6 days of incubation, culture supernatant total IgG, IgA, and IgM concentrations were determined in ELISA.

Newborn and adult mice peritoneal macrophage BAFF and APRIL mRNA expression, as well as splenic B cell TACI, BCMA, and BAFFR mRNA expression were determined in real-time RT-PCR. Newborn mice cells from 15 to 18 mice were randomly divided into three groups, with each group containing 5–6 mice, were designated as one set. Macrophages or purified B cells were pooled within each group. Cells from three adult mice were used as control. Total RNA was extracted using the RNeasy Mini kit (Qiagen) according to the manufacturer’s instructions, and 1 μg/sample of RNA was reverse-transcribed into cDNA with random hexamers using the TaqMan Reverse Transcription Reagents kit (Applied Biosystems). Predesigned primer and probe sets (FAM labeled) from Applied Biosystems (TaqMan Gene Expression Assay) were used to amplify mRNA for BAFF, APRIL, and the three receptors. 18S RNA was also amplified as a reference gene using primer and probe sets from Applied Biosystems. Triplicate samples run on the Cepheid Smart Cycler System were quantified as fold difference (expression ratio) between adult and newborn samples as described by Pfaffl (28). Expression ratio was calculated as (E_target)^{ΔCt target (newborn − adult)}/(E_18S)^{ΔCt 18S (newborn − adult)}, where Ct is the cycle threshold. Target gene or reference gene (18S RNA) primer efficiencies (E) were calculated by using the E = 10^(−1/slope) formula after amplifying 5-log dilutions of concentrated mouse macrophage (for BAFF and APRIL) or B cell (for TACI, BCMA, and BAFFR) cDNA. Fold differences between newborn or adult mouse macrophage BAFF and APRIL mRNA compared with BAFF or APRIL expression in the murine macrophage cell line RAW 264.7 (American Type Culture Collection) were used to calculate statistical significance between adult and newborn BAFF and APRIL mRNA expression. Similarly, the murine B cell lymphoma line, A20 cell (a gift of Dr. D. Klinman, National Cancer Institute, Frederick, Maryland) TACI, BCMA, and BAFFR mRNA expression ratios were used to determine the statistical significance of the fold difference between adult and newborn B lymphocyte TACI, BCMA, or BAFFR expression.

Groups of five BALB/c mice (5-day old) were i.p. injected with CpG ODN (20 μg) plus NP-Ficoll (40 μg) or PBS plus (4-hydroxy-3-nitrophenyl)acetyl-Ficoll (NP-Ficoll; 40 μg) or PBS alone. The injection volumes were 100 μl. Groups of five newborn TACI KO mice or the control C57BL/6 littermates were also immunized i.p. with CpG ODN (20 μg) plus NP-Ficoll (40 μg) or PBS. After 7 days, mice were tail bled and serum anti-NP IgG Abs were determined in ELISA. ELISA plates were coated with NP-chicken gammaglobulin (10 μg/ml NP). One to 10 diluted mouse sera were added to NP- chicken gammaglobulin containing wells and bound Abs were detected by using peroxidase-conjugated goat anti-mouse IgG Ab and peroxidase developing solution TMB (KPL).

The Student t test was used to compare groups. A value of p < 0.05 was considered as statistically significant.

The constitutive BAFF and APRIL expression levels of adult and newborn (5-day-old) BALB/c mice were assessed. Peritoneal cells harvested after thioglycolate injection were analyzed for intracellular BAFF expression in flow cytometry, and for BAFF and APRIL mRNA expression in a real-time RT-PCR assay. Flow cytometry assessments revealed that BAFF was significantly higher (p < 0.01) in newborns (48.38 ± 16.98%) as compared with adult mice (27.51 ± 7.02%) (Fig. 1,A). Evaluation of newborn (865 ± 141) and adult (390 ± 129) BAFF mean fluorescence intensity (MFI) values also showed a significantly higher expression of BAFF in newborns (p < 0.001). Amplification of BAFF and APRIL mRNA reiterated the increased BAFF protein expression in newborn mice (p < 0.05), while APRIL expression was not significantly different between the two age groups (Fig. 1,B). To determine whether thioglycolate injection altered BAFF expression, peritoneal cells were also harvested before thioglycolate injection. Pan-macrophage marker, CD11b Ab was used to gate macrophages and IgM Ab was used to gate B cells (29, 30). As expected, thioglycolate injection led to an increase in the number of macrophages in both adult and newborn mice (Fig. 1 A). Adult and newborn B cell (IgM⁺CD11b⁻) BAFF expression was comparable before and after thioglycolate injection. Nevertheless, newborn macrophage (IgM⁻CD11b⁺) BAFF expression was significantly higher than that of adult cells with or without thioglycolate injection.

It has been shown that the TLR agonist LPS stimulates BAFF expression on adult mouse macrophages (31) while CpG ODN and poly I:C stimulate APRIL expression on adult human DCs (32). We sought to determine whether CpG ODN would have an effect on BAFF or APRIL expression on newborn cells. Stimulation of newborn peritoneal cells with CpG ODN for 24 h induced a modest and statistically insignificant increase in adult (p < 0.2) and newborn (p < 0.3) BAFF percentage (Fig. 1,C). Similarly, BAFF or APRIL mRNA expression remained unchanged in adult and newborn cells after CpG ODN treatment (Fig. 1 D).

Adult and newborn B lymphocyte numbers were determined in flow cytometry following the gating of B220⁺ cells in splenocytes. As reported previously (2), newborn mouse spleen contained a significantly low (p < 0.001) proportion of B lymphocytes (14 ± 3%) compared with adult spleens (50 ± 1%) (data not shown). The constitutive expression of BAFF/APRIL receptors BAFFR, TACI, and BCMA was analyzed in purified B lymphocytes (purity ≥96%) from adult and newborn mice. In flow cytometry assessments, the B lymphocytes from newborns had a significantly lower percentage of TACI-expressing cells (p < 0.0001) compared with those of adult mice (15.3 ± 1.5% vs 47.9 ± 0.3%) (Fig. 2,A). Comparison of MFI values for the adult (MFI 492 ± 102) and newborn (MFI 299 ± 57) TACI expression also yielded a significantly lower (p < 0.01) measurement for newborn mice. Conversely, newborn B lymphocytes expressed slightly increased levels of BAFFR (57.7% ± 8.9%) although this difference was not statistically significantly higher (p < 0.3) than adult BAFFR values (54.9 ± 10%). Newborn BCMA expression levels (5.1 ± 0.8%) were comparable (p < 0.4) to those of adults (5.1 ± 0.3%). An analysis of mRNA expression of TACI, BCMA, and BAFFR by real-time RT-PCR reiterated the flow cytometry results (Fig. 2 B). Here also, TACI mRNA expression was lower (p < 0.05) in B lymphocytes from newborns as compared with adult mice, but no significant differences were observed in BCMA (p < 0.4) or BAFFR (p < 0.1) expression between the two groups of mice.

Next, we sought to determine the expression of TACI and BAFFR on subsets of B lymphocytes. Reiterating the previous reports (2), multicolor staining of newborn or adult splenocytes demonstrated that newborns had lower numbers of FO and MZ B lymphocytes while their B1 and NF B lymphocyte numbers were higher than those of adults (Fig. 3 and Table I). Analysis of receptors on gated NF (B220⁺CD23^negCD21^neg), FO (B220⁺CD23^highCD21^int), MZ (B220⁺CD23^lowCD21^high), and B1 (IgM⁺B220^lowCD5⁺) cells demonstrated that, as with the adult cells (33), MZ and B1 B lymphocytes expressed the highest levels of TACI. Nevertheless, when compared with adults, newborns expressed lower levels of TACI on all subsets (Fig. 3 and Table I). The expression of newborn BAFFR was only higher on MZ and B1 cells, while NF and FO B lymphocyte percentages were comparable between the two age groups (Fig. 3 and Table I).

We (27) and others (19) have previously shown that adult TACI KO mice do not secrete IgG and IgA when stimulated with BAFF or APRIL. We therefore sought to understand whether decreased TACI expression in newborns would also lead to impaired Ig secretion. For this purpose, B lymphocytes from adult and newborn BALB/c mice were cultivated with BAFF or APRIL in the presence of IL-4. Analysis of culture supernatants revealed that Ig secretion in response to BAFF or APRIL was virtually absent in newborns for all the isotypes tested (Fig. 4). There were significant differences between adult and newborn IgM, IgA, IgG1, IgG2a, IgG2b, and IgG3 concentrations following BAFF or APRIL stimulation (Fig. 4). Although B lymphocytes from newborns were poorly responsive to BAFF and APRIL even in the presence of IL-4 costimulation, the responses to TACI-independent stimuli (LPS along with IL-4 or TGF-β) were comparable to those of adults.

We have recently shown that in vitro prestimulation of adult B lymphocytes with CpG ODN up-regulates TACI and BAFFR expression, and leads to TACI-dependent augmentation of Ig secretion following BAFF or APRIL stimulation (27). We therefore asked whether CpG ODN enhances the expression of TACI on newborn B lymphocytes as well. The expression of BAFF and APRIL receptors on B lymphocytes from adult and newborn mice was determined by flow cytometry after stimulation with CpG ODN for 24 h. As expected, percentage of TACI and BAFFR expressing adult mice B lymphocytes were increased (p < 0.01 and p < 0.01, respectively) following CpG ODN stimulation (Fig. 5). Similarly, TACI levels on B lymphocytes from newborn mice increased (p < 0.01) from 15.8 ± 1.0% to 31.2 ± 0.4% after CpG ODN treatment. Enhancement (from 49.5 ± 1.4% to 56.2 ± 2.6%) in BAFFR expression also reached statistically significant levels (p < 0.01), while increases in BCMA levels (from 6.8 ± 1.1% to 7 ± 1.3%) were not statistically significant (p < 0.1). Measurement of MFI values for TACI and BAFFR also revealed a significant increase after CpG ODN stimulation (p < 0.001 and p < 0.02, respectively). Analysis of TACI expression on B lymphocyte subsets demonstrated that CpG ODN strongly up-regulated TACI in NF, MZ, FO, and B1 B lymphocytes of newborn mice (Fig. 6 and Table II). As with TACI expression, NF, MZ, and B1 B lymphocyte BAFFR levels were increased after CpG ODN incubation (Fig. 6 and Table II). Interestingly, CpG ODN treatment led to a decrease in FO cell BAFFR expression. The effect of CpG ODN on newborn B lymphocyte subset composition was variable. Similar to the results obtained with adult mice (27), CpG ODN stimulation led to an increase in NF (p < 0.001) B lymphocyte population and a decrease in FO (p < 0.001) B lymphocyte population, while MZ (p < 0.45) and B1 (p < 0.3) B lymphocyte percentages remained unchanged.

Having shown that CpG ODN stimulated expression of TACI in newborns, we then asked whether this would lead to increased Ig secretion in response to BAFF or APRIL stimulation. For this purpose, B lymphocytes pretreated with CpG ODN or medium for 24 h were washed and counted, and then equal numbers of cells were restimulated with medium alone, BAFF, APRIL, LPS and IL-4 or LPS and TGF-β for 6 days. Analysis of culture supernatant Ig isotype levels revealed that CpG ODN pretreated neonatal B lymphocytes became responsive to both BAFF and APRIL even in the absence of a costimulation with IL-4 (Fig. 7,A). BAFF- or APRIL-stimulated cells yielded a significant increase in culture supernatant IgM, IgA, IgG1, IgG2a, IgG2b, and IgG3 compared with B lymphocytes that had been incubated with medium. Similar to our previous report (27) CpG ODN pretreatment of adult B lymphocytes also led to an increase in Ig production after BAFF and APRIL stimulation (Fig. 7,A). It should be noted that incubation of both newborn and adult B lymphocytes with CpG ODN led to spontaneous induction of IgM secretion and neither BAFF nor APRIL could significantly enhance further IgM secretion. To determine whether up-regulation of TACI by CpG ODNs is responsible for the increased Ig secretion in response to BAFF or APRIL stimulation, we next added blocking anti-TACI Abs in CpG ODN-pretreated and BAFF- or APRIL-stimulated B lymphocytes. We have previously shown that the polyclonal anti-TACI Ab efficiently blocks BAFF- or APRIL-mediated Ig secretion without inhibiting LPS- or CD40L-mediated Ig secretion and does not impact B lymphocyte survival in adult mice (27). Measurement of culture supernatant Ig levels demonstrated that TACI-blocking Abs inhibited BAFF or APRIL (Fig. 7 B) induced IgG or IgA secretion in CpG ODN pretreated B lymphocytes in a concentration-dependent fashion, while IgM levels remained unchanged. These results clearly demonstrated that as with adult B lymphocytes, the CpG ODN-mediated increase in newborn TACI expression correlated with the augmentation of Ig secretion in response to BAFF or APRIL stimulation.

The results presented so far suggested that decreased TACI expression contributes to impaired Ig secretion in newborns and up-regulation of TACI by CpG ODN restores BAFF- and APRIL-mediated Ig secretion from newborn B lymphocytes. We next sought to determine how low TACI expression leads to decreased Ig secretion in response to BAFF or APRIL stimulation. We have recently shown that TACI expression is crucial in the development of TI-II Ag-specific plasma cells (22). To understand whether decreased expression of TACI in newborn B lymphocytes would have an impact on plasma cells, we compared adult and newborn plasma cell numbers and development. Determination of B220⁺CD138⁺ cells demonstrated that plasma cells constituted a smaller percentage (p < 0.01) of newborn B lymphocytes (3.5 ± 0.3%) compared with those of adult (9.9 ± 2.4%) mice (Fig. 8). Incubation of adult B lymphocytes with BAFF plus anti-IgM (21.9 ± 1.3%, p < 0.01), APRIL plus anti-IgM (17 ± 0.6%, p < 0.005), or LPS (22.3 ± 1.3%, p < 0.001) for 3 days resulted in a significant increase in plasma cells in all three conditions (Fig. 8). In newborns, however, BAFF plus anti-IgM (5.6 ± 1.6%, p < 0.06) or APRIL plus anti-IgM (3.3 ± 1%, p < 0.3) stimulation did not have an effect on CD138 expression, while LPS stimulated cells (19.1 ± 0.9%) manifested a significant increase (p < 4.5E-07) (Fig. 8). These results demonstrated that BAFF or APRIL promote plasma cell development in adult B lymphocytes but they cannot in low TACI-expressing newborn B lymphocytes. In contrast, the percentage of plasma cells developed after LPS stimulation of newborn B lymphocytes was comparable to that of adult mouse cells.

Because CpG ODN stimulation of newborn B lymphocytes increases TACI expression, we reasoned that CpG ODN-induced TACI up-regulation may promote BAFF- or APRIL-mediated plasma cell development in newborns. For this purpose, we preincubated newborn B lymphocytes with CpG ODN for 24 h and then restimulated with medium, BAFF plus anti-IgM, APRIL plus anti-IgM or LPS. After 3 days of incubation, both BAFF (18.4 ± 3.3%, p < 0.0002) and APRIL (24 ± 5.4%, p < 0.0001) induced a significant increase in the percentage of plasma cells (Fig. 8). Thus, CpG ODN-mediated TACI up-regulation correlated with BAFF- or APRIL-mediated plasma cell development in newborn mouse B lymphocytes. Similarly, CpG ODN pretreatment enhanced adult plasma cell development in response to BAFF (p < 0.0001) and APRIL (p < 0.001) stimulation.

Activation of B cells by cross-linking of BCR without the help of T cell is the hallmark of T cell-independent type II Ags such as polysaccharide vaccines (34). Multivalent Ag-mediated activation of B lymphocytes can be mimicked in in vitro assays by using F(ab′)₂ anti-IgM Abs (35, 36). In newborns, however, the anti-IgM mediated signal typically induces apoptosis in B lymphocytes (37, 38) unless a second signal, such as LPS or T lymphocyte cytokines are added to culture medium (39, 40, 41). Previous reports demonstrated a moderate but significant enhancement of TACI expression in anti-IgM-stimulated adult B lymphocytes (27, 42). To determine whether membrane IgM engagement modulates BAFF/APRIL receptor expression in newborns, we stimulated purified B lymphocytes with anti-IgM plus IL-4 for 24 h and compared their TACI, BCMA, and BAFFR levels with medium-treated cells. As previously shown (39, 40), anti-IgM plus IL-4 stimulation led to proliferation of newborn B lymphocytes (data not shown). Flow cytometry results demonstrated a modest but significant increase (p < 0.01) in BAFFR (from 39.2 ± 1.8% to 44.9 ± 0.81%) expression whereas TACI (from 13.23 ± 1.45% to 17.2 ± 5%) and BCMA (from 5.2 ± 0.2% to 5.3 ± 0.6%) levels remained unchanged (p < 0.13 and p < 0.35, respectively) (Fig. 9,A). Furthermore, underscoring the role of TACI expression levels, but not BAFFR levels, in BAFF- or APRIL-mediated Ig secretion, stimulation of anti-IgM plus IL-4 pretreated newborn B lymphocytes with BAFF or APRIL did not lead to production of Igs (Fig. 9 B).

The adjuvant effect of CpG ODN in vaccines has been well documented (43, 44). Immunization with CpG ODN is also known to successfully increase Ab production in neonatal mice (45, 46). Because we detected an up-regulation of TACI in newborn B lymphocytes after in vitro CpG ODN stimulation, we next sought to understand whether CpG ODN would have a similar effect on in vivo TACI expression after injection of 5-day-old mice. Newborns were i.p. injected with 20 μg of CpG ODN or PBS (control mice), and the receptor expression on gated spleen cells was measured in flow cytometry 24 h later. A significant increase in the percentage of TACI (p < 0.0001) and BAFFR (p < 0.01) expressing B220⁺ cells was observed in newborns injected with CpG ODN (TACI: 15.8 ± 4.5% and BAFFR: 9.16 ± 1.2%) as compared with PBS injected (TACI: 4.3 ± 0.7% and BAFFR: 17.7% ± 4%) mice, while BCMA levels remained unchanged (p < 0.31) (Fig. 10,A). Re-evaluation of the changes in receptor expression by comparing the MFI values of PBS (254.2 ± 153) injected mice with CpG ODN (748 ± 95) injected mice also yielded a significant increase in TACI expression (p < 0.0001) with CpG ODN. There was a statistically significant difference (p < 0.01) in the BAFFR MFI values between PBS-injected (103.5 ± 7.9) and CpG ODN-injected (143.3 ± 13.2) mice also. BCMA expression (76 ± 10 vs 64 ± 8) was not altered (p < 0.14) after CpG ODN injection. Similar to results obtained in in vitro stimulation experiments, in vivo administration of CpG ODN also up-regulated TACI on all B lymphocytes subsets tested (Fig. 10,B, Table III). BAFFR levels were also increased on all subsets except FO cells where a significant decrease was detected after CpG ODN injection (Fig. 10,B, Table III).

We next examined whether CpG ODN-injected mice B lymphocytes respond to BAFF or APRIL with increased Ig secretion as seen in in vitro CpG ODN pretreated B lymphocytes. B lymphocytes isolated from newborns that had been injected with CpG ODN or PBS were subsequently cultured in the presence of BAFF, APRIL, LPS/IL-4, or LPS/TGF-β for 6 days. B lymphocytes from CpG ODN-injected neonatal mice secreted profoundly higher levels of IgM, IgA, IgG1, IgG2a, IgG2b, and IgG3 in response to BAFF or APRIL stimulation, while PBS-injected mice B lymphocytes did not respond to BAFF or APRIL stimulation (Fig. 11).

Having shown that CpG ODN injection renders murine B lymphocytes sensitive to BAFF- or APRIL-induced Ig secretion, we next investigated whether CpG ODN would help augment Ab response against the TI-II Ag NP-Ficoll. Indeed, immunization of 5-day-old mice (n = 5 in each group) with CpG ODN plus NP-Ficoll, PBS plus NP-Ficoll, or PBS alone demonstrated that CpG ODN containing vaccine yielded significantly higher levels of serum anti-NP IgG Abs (p < 0.0001) compared with those immunized with NP-Ficoll in PBS or PBS alone (Fig. 12,A). To determine the role of TACI in CpG ODN-mediated augmentation of the Ab response against NP, we next immunized TACI KO mice and the control C57BL/6 mice with CpG ODN plus NP-Ficoll or PBS (n = 5 in each group). Measurement of serum anti-NP IgG levels 7 days after immunization demonstrated that while the control mice developed significantly increased levels (p < 0.0001) of anti-NP Abs, the TACI KO mice response was comparable to the PBS-injected mice response (Fig. 12 B).

In this study, we demonstrate that newborn mice B lymphocyte TACI levels are severely reduced compared with those of adult mice, and newborn B lymphocytes do not secrete Igs when stimulated with BAFF or APRIL. Reduced TACI levels appear to be responsible for the unresponsiveness of newborn B lymphocytes to TI-II Ags such as bacterial CPS. The requirement for TACI in response to TI-II Ags is well documented in TACI KO mice (16, 17). Also, common variable immunodeficiency patients with mutations in the gene encoding TACI (TNFRSF13B) exhibit reduced Ab response against TI-II Ags and manifest frequent and recurrent infections with encapsulated bacteria (18).

It is not exactly known why TACI is critical for the development of Abs against TI-II Ags. However, the fact that TACI is mostly expressed on MZ B lymphocytes (20, 27), the B lymphocyte population that is responsible for the generation of Ab response against TI-II Ags (47), suggests that BAFF and APRIL expressed by macrophages and DCs located in MZ provide an essential T cell-independent second signal for CPS-specific B lymphocytes (48, 49, 50). Indeed, supporting this hypothesis, recent findings suggested a crucial role for TACI in the development of TI-II Ag-specific plasma cells (22). Additionally, TACI is shown to enhance plasma cell differentiation in CD40-stimulated B lymphocytes (51). In this study, we provide further evidence for the key role of TACI in plasma cell development. For example, reduced expression of TACI correlated with significantly lower CD138 expression in newborns compared with those in adults. In addition, despite a robust up-regulation of CD138 expression in adult B lymphocytes with BAFF plus anti-IgM or APRIL plus anti-IgM stimulation, newborn CD138 levels remained unchanged when stimulated with these TACI ligands. More importantly, up-regulation of TACI expression by CpG ODN resulted with a significant increase in plasma cell development in BAFF plus anti-IgM or APRIL plus anti-IgM stimulated newborn B lymphocytes.

In adults, Ab development against TI-II Ags depend on cross-linking of BCR with multivalent TI-II determinants (34), which can be mimicked by anti-IgM Ab stimulation of B lymphocytes. Anti-IgM stimulation not only results in proliferation of B lymphocytes and secretion of Igs with the help of cytokines such as IL-4, but also stimulates TACI expression (35, 36). Thus, BCR-mediated up-regulation of TACI expression may be instrumental in the production of Ab response against TI-II Ags. Here, we show that, in contrast to adults, anti-IgM plus IL-4 stimulation of newborn B lymphocytes not only failed to enhance TACI expression but also did not sensitize B lymphocytes to BAFF- or APRIL-mediated Ig secretion. Hence, in addition to decreased expression of TACI in newborns, their inability to regulate TACI expression in response to TI-II Ags may be the reason for the poor antigenicity of TI-II vaccines in newborns.

We have recently shown that the level of TACI expression in adult mice correlates with the amount of Igs secreted in response to stimulation of B lymphocytes with BAFF or APRIL (27). In those experiments, CpG ODN-induced up-regulation of TACI correlated with BAFF- and APRIL-mediated Ig secretion. Here, we show that newborns constitutively express decreased TACI, and low TACI expression correlates with impaired Ig secretion in BAFF- and APRIL-stimulated B lymphocytes. These findings are reminiscent of those from TACI KO mice (19, 27) or common variable immunodeficiency patients (52, 53) who also respond poorly to BAFF and APRIL stimulation. The fact that in vitro incubation of newborn B lymphocytes with CpG ODN up-regulates TACI expression and CpG ODN injection can stimulate in vivo TACI expression in newborn B lymphocytes, and CpG ODN pretreated B lymphocytes are able to respond to BAFF and APRIL by secreting high levels of Igs, underscores the role of low TACI expression in decreased Ig secretion in response to BAFF and APRIL stimulation. Detailed analysis of B lymphocyte subsets demonstrated that, as seen in adults, MZ B lymphocytes express the highest levels of TACI followed by B1 cells. Upon CpG ODN stimulation, however, TACI expression was increased on all B lymphocyte subsets tested. Interestingly, although CpG ODN led to an increase in splenic NF and B1 cell numbers, numbers of MZ B lymphocyte, the subset that is responsible for the production of Ab response against TI-II Ags, did not change. Nevertheless, strong and widely distributed up-regulation of TACI was responsible for the adjuvant effect of CpG ODN on TI-II Ags in newborns. BALB/c mice immunized with CpG ODN containing NP-Ficoll vaccine induced significantly higher levels of anti-NP IgG Abs compared with those immunized with NP-Ficoll in PBS, whereas TACI KO mice did not respond to NP-Ficoll despite the inclusion of CpG ODN to NP-Ficoll vaccine.

Although it has been previously shown that CpG ODNs can help boost the immune response against TI-II Ags in neonates (46, 54), the underlying mechanism for the increased immunogenicity of TI-II Ags remained poorly understood. The studies presented here establish decreased TACI expression as a likely cause of unresponsiveness of newborns to TI-II Ags (Fig. 13). Moreover, the fact that CpG ODN-mediated up-regulation of TACI expression in neonatal B lymphocytes leads to increased plasma cell development and secretion of Igs in BAFF- or APRIL-stimulated B lymphocytes suggests a model for how CpG ODN-containing CPS vaccines increase the immunogenicity of TI-II Ags (Fig. 13). We also found that, in contrast to decreased TACI expression, newborn BAFF levels are markedly higher than those of adults. We do not know the reason for the increased BAFF expression in newborns. One possible explanation may be related to low TACI expression, which in turn may lead to decreased Ig secretion. Low levels of circulating Igs may be sending a feedback signal to myeloid cells for further production of BAFF.

The reasons for the decreased expression of TACI at birth are currently unknown. The transition of newborns from the mostly sterile intrauterine environment to colonization with commensal bacteria is shown to provide signals that shape several aspects of mammalian physiology (55, 56, 57). Commensal bacteria are also important in the development of mucosal and systemic immune system. For example, in the absence of microbial flora, germfree mice manifest reduced numbers of IgA-secreting plasma cells and decreased numbers of CD4⁺ T cells in the lamina propria (55). These mice also have altered splenic structure and reduced levels of circulating IgGs (55). The results of our previous (27) and current studies have shown that adult and newborn TACI expression can be regulated by TLR agonists. Others have shown an increase in systemic Ab response mediated by TLR induced up-regulation of BAFF and APRIL expression (58, 59). Therefore, it is likely that the establishment of microbial flora in newborns may be providing TLR-mediated signals to maintain functional levels of TACI, and its ligands BAFF and APRIL.

Taken together, these new insights may help design novel vaccines that can selectively modulate TACI expression for the development of Ab response against CPS of encapsulated bacteria and against viruses where CD40-independent, but BAFF/APRIL-mediated Ig isotype switch (60, 61) plays a role in the development of Ab response.

We thank Karen Elkins and Ronald L. Rabin (Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD) for critical reading of the manuscript.

The authors have no financial conflict of interest.