Invariant chain (Ii)-deficient mice exhibit profound B cell defects that have remained poorly understood, because they could not be simply explained by impaired Ag presentation. We found that Ii deficiency induced cell autonomous defects of two distinct B cell lineages. The life span of mature follicular (FO) B cells was reduced, accounting for their markedly decreased frequency, whereas, in contrast, marginal zone (MZ) B cells accumulated. Other Ii-expressing lineages such as B1 B cells and dendritic cells were unaffected. Surprisingly, the life span of FO B cells was fully corrected in Ii/I-Aβ doubly deficient mice, revealing that Ii-free I-Aβ chains alter FO B cell survival. In contrast, the accumulation of MZ B cells was controlled by a separate mechanism independent of I-Aβ. Interestingly, in Ii-deficient mice lacking FO B cells, the MZ B cells invaded the FO zone, suggesting that intact follicules contribute to the retention of B cells in the MZ. These findings reveal unexpected consequences of Ii deficiency on the development and organization of B cell follicles.

B cell development proceeds through sequential checkpoints from hemopoietic stem cells in the bone marrow to the formation of mature B cells in the periphery (1, 2, 3). Immature B cells expressing surface Ig migrate from the bone marrow to the spleen where transition to the mature B cell stage is characterized by marked phenotypic changes. Although immature B cells are IgMhighIgDlowCD21lowCD23low, mature follicular (FO)4 B cells acquire a IgMlowIgDlowCD21intCD23high phenotype (4, 5). In addition, the spleen contains a distinct population of mature B cells called marginal zone (MZ) B lymphocytes, which accumulate over time in the splenic MZ to represent up to 10–15% of the B cell population and characteristically express IgMhighIgDlowCD21highCD23lowCD1high (6, 7, 8, 9, 10, 11). Unlike FO B cells, MZ B cells do not recirculate and they express peculiar functional properties, such as prompt activation and differentiation into Ab-secreting cells upon recognition of low concentrations of complement-coated Ags (6, 9, 12). Furthermore, MZ B cells include a population of B cells expressing germline-encoded B cell receptors with autoreactive or antimicrobial specificities (13, 14, 15).

Spontaneous or induced mutations have revealed many genes involved in the development of various B cell subsets (16). However, the report that invariant chain (Ii)-deficient mice exhibited marked B cell defects was particularly intriguing because these defects appeared to be independent of the main function attributed to Ii, which is to bind MHC class II molecules in the endoplasmic reticulum to prevent peptide binding and serve as a chaperone until they reach the endosome (17, 18). In fact, I-Aβ or class II transactivator (CIITA)-deficient mice with deficient MHC class II-mediated peptide presentation and CD4 T cell development had normal B cells (17, 19, 20). Recently, Ii was also shown to associate with DM and CD1d (21, 22, 23), which are highly expressed in B cells, prompting us to re-examine the B cell phenotype of Ii-deficient mice and investigate the possibility that potential defects in DM or CD1 functions might explain the B cell defects.

We found that Ii-deficient mice expressed a more complex B cell phenotype than previously appreciated. Thus, the frequency of FO B cells was decreased, not because of a block in maturation of immature precursors, but because of a markedly reduced life span. In contrast, a considerable, previously unrecognized accumulation of MZ B cells was detected in adult mice. These defects were cell autonomous and could not be explained by MHC class II or CD1 Ag presentation defects. Our studies further revealed that the presence of Ii-free I-Aβ chains was responsible for the reduced life span of FO B cells. Interestingly, in Ii-deficient mice with depleted FO B cells, the expanded MZ B cells invaded the FO zone, whereas upon correction of the FO defect in Ii/I-Aβ doubly deficient mice, they were retained in the MZ, suggesting that intact follicules contribute to the retention of B cells in the MZ.

C57BL/6, C57BL/6.CD45.2, C57BL/6.I-Aβb KO, and C57BL/6.Ii KO mice were obtained from Taconic Farms (Germantown, NY). CIITA-deficient mice (24) were used after eight backcrosses to C57BL/6. CD1d-deficient mice (25) were used after 12 backcrosses to C57BL/6. DM-deficient mice (26) were used after four backcrosses to C57BL/6. Ii/I-Aβb and Ii/CD1d doubly deficient mice were bred in our mouse facility.

FITC-conjugated mAbs against CD21, B220, and CD4, PE-conjugated anti-CD23, CD54.2, CD1d, and B220, allophycocyanin-conjugated anti-CD5, CD45.1, CyChrome-conjugated anti-B220, CD8, biotinylated anti-CD23, ICAM-1, CD11a, CD18, CD49d, CD29, CXCR-5 and streptavidin-CyChrome were obtained from BD PharMingen (San Diego, CA). PE-conjugated rat anti-mouse IgD and biotinylated goat anti-mouse IgM were from Southern Biotechnology Associates (Birmingham, AL). PE-conjugated goat anti-mouse IgM was obtained from Caltag Laboratories (Burlingame, CA). Biotinylated anti-CD9 was a kind gift from J. Kearney (University of Alabama, Birmingham, AL). Samples were analyzed using a four-color FACSort equipped with argon and 635-nm diode lasers (BD Biosciences, Mountain View, CA) and CellQuest software.

Drinking water containing 1 mg/ml BrdU (Sigma-Aldrich, St. Louis, MO) was administered for 2 wk, with fresh water preparations made every day. BrdU-positive cells were detected with the BrdU Flow kit (BD PharMingen) according to the manufacturer’s instructions.

Frozen splenic sections were prepared and stained as previously described (27). Ten-micrometer spleen sections were double stained using FITC-conjugated rat anti-mouse metallophilic macrophages MOMA-1 (Serotec, Oxford, U.K.) and biotinylated rat anti-B220 detected with streptavidin-Texas Red. In a second combination, sections were stained with rat anti-CD1d detected with biotinylated goat anti-rat IgG and streptavidin-Texas Red, then with FITC-conjugated MOMA-1.

C57BL/6 mice received a whole body gamma-irradiation (1000 rad) with a 137Ce source (Gammacell 40; MDS Nordion, Ontario, Canada). Six hours after irradiation, they were reconstituted with one i.v. injection of 5–10 × 106 liver cells from day 14 fetuses.

Three- to 4-mo-old mice were immunized by i.p. injection of 10 μg of trinitrophenyl (TNP)-Ficoll in PBS. Serum samples were collected before and at day 13 after immunization. TNP-specific Abs were measured by isotype-specific ELISA on plates coated with TNP-BSA as described previously (12).

Previous studies reported an accumulation of immature B cells and a loss of mature B cells in mice lacking Ii (17). However, these studies characterized immature (or newly formed (NF)) B cells based on their IgMhighIgDlow profile, which is shared with MZ B cells as well. We performed four-color FACS analysis to further examine the various B cell subsets based on their expression of CD23 and CD21, which distinguishes NF B cells (CD21lowCD23low) from MZ B cells (CD21highCD23low), as well as their expression of CD1d and CD9 which is characteristically high in MZ B cells. Fig. 1,A shows that, as previously reported, the frequency of FO B cells among B220+ splenic cells was markedly reduced, from 72% in wild-type (WT) to 46% in Ii KO mice. Because Ii-deficient spleens harbored slightly fewer cells than WT (shown later in Fig. 7,C), the decreased frequency reflected decreased absolute numbers of B cells. This defect was not the consequence of aberrant migration to other lymphoid organs because drastic decreases in B cells were found in lymph nodes, blood, and bone marrow as well (see gated populations of B220+CD5 cells or B220highIgMhigh cells in Fig. 1,B). However, in contrast with previous conclusions based on isolated IgM/IgD profiles, we found that cells with the CD21lowCD23lowIgMhighIgDlow profile that characteristically defines NF B cells had a normal frequency (Fig. 1,A). This conclusion was further supported by the finding that bone marrow B220intIgMhigh cells, which correspond to NF cells, were normal in frequency (Fig. 1,B). Finally, splenic CD21highCD23low MZ-like B cells were markedly increased in Ii KO over WT mice (Fig. 1,A). These cells had escaped previous recognition because they share the same CD23lowIgMhighIgDlow profile as NF cells, but they differ with respect to the expression of additional markers such as CD21, high in MZ but low in NF cells, and CD1d and CD9 characteristically high in MZ B cells. In summary, although the number of NF cells was normal, marked changes were observed in the two mature splenic B cell lineages, with decreased FO B cells but increased MZ-like B cells. These opposite changes, which resulted in a 5-fold increase in the splenic MZ:FO ratio, explained the overall modest diminution of B cells in the spleen (shown later in Fig. 7,C). Importantly, other MHC class II-expressing cell lineages were apparently conserved. For example, CD23high B1 B cells constituted a normal proportion of lymphocytes in the peritoneal cavity (Fig. 1,B), and splenic CD11c+MHC II+ dendritic cells (DC) were also conserved (Fig. 1 B, note that Ii-deficient DC express lower surface levels of MHC class II).

FIGURE 1.

B cell changes in Ii KO mice. A, Four-color FACS staining identify a normal immature (NF), a reduced FO, and an increased MZ B cell subset in the spleen of 4-mo-old Ii KO as compared with WT mice. CD21/CD23 dot plots are gated on B220+ cells and numbers represent percentage of B220+ cells in the indicated gates. Data are representative of >50 individual mice. B, FACS staining identifies B cells in the lymph node (LN) and blood compartments as B220+CD5 (numbers correspond to percent total cells in the gates). In the bone marrow (BM), B220lowIgM cells are pro/pre-B, B220lowIgM+ are NF, and B220highIgM+ are recirculating mature FO-type B cells. In cells obtained from peritoneal lavage (PerC), CD23IgM+ are B1 cells and CD23+IgM+ are B2 cells. In the spleen, boxed MHCII+/CD11c+ cells represent DC (note that DC from Ii KO mice express lower levels of MHCII).

FIGURE 1.

B cell changes in Ii KO mice. A, Four-color FACS staining identify a normal immature (NF), a reduced FO, and an increased MZ B cell subset in the spleen of 4-mo-old Ii KO as compared with WT mice. CD21/CD23 dot plots are gated on B220+ cells and numbers represent percentage of B220+ cells in the indicated gates. Data are representative of >50 individual mice. B, FACS staining identifies B cells in the lymph node (LN) and blood compartments as B220+CD5 (numbers correspond to percent total cells in the gates). In the bone marrow (BM), B220lowIgM cells are pro/pre-B, B220lowIgM+ are NF, and B220highIgM+ are recirculating mature FO-type B cells. In cells obtained from peritoneal lavage (PerC), CD23IgM+ are B1 cells and CD23+IgM+ are B2 cells. In the spleen, boxed MHCII+/CD11c+ cells represent DC (note that DC from Ii KO mice express lower levels of MHCII).

Close modal
FIGURE 7.

Selective correction of the FO B cell defect associated with Ii KO in Ii/IAβ double KO mice. A, FACS analysis, as in Fig. 1, of the spleen, lymph node (LN), and BM of 4-mo-old mice of the indicated genotypes. B, NF, FO, and MZ splenic B cell subsets plotted as percentage of B220+ cells. Dots represent individual 4-mo-old mice; mean values are indicated by bars. C, Total cell numbers (upper panels) and percentages of B220+ cells (lower panels) in the spleen, lymph node, and bone marrow of Ii and Ii/I-Aβ KO mice are represented relative to WT. In the bone marrow, only the mature B220highIgM+ cells are counted. Dots represent individual 4-mo-old mice.

FIGURE 7.

Selective correction of the FO B cell defect associated with Ii KO in Ii/IAβ double KO mice. A, FACS analysis, as in Fig. 1, of the spleen, lymph node (LN), and BM of 4-mo-old mice of the indicated genotypes. B, NF, FO, and MZ splenic B cell subsets plotted as percentage of B220+ cells. Dots represent individual 4-mo-old mice; mean values are indicated by bars. C, Total cell numbers (upper panels) and percentages of B220+ cells (lower panels) in the spleen, lymph node, and bone marrow of Ii and Ii/I-Aβ KO mice are represented relative to WT. In the bone marrow, only the mature B220highIgM+ cells are counted. Dots represent individual 4-mo-old mice.

Close modal

We performed immunohistological examination of the spleen to further define the nature of the B cell subset alterations. Surprisingly, in spleen sections stained with anti-B220 and MOMA-1 (Fig. 2,A), the overall architecture of the follicule seemed conserved. In particular, despite the 5-fold increase in the MZ:FO ratio determined by FACS analysis, Ii KO mice clearly failed to show a corresponding enlargement of the MZ and depletion of the FO area. We reasoned that the expanded MZ-like B cells might have invaded the FO area and be mixed with residual FO B cells. To test this hypothesis, we took advantage of the fact that MZ B cells express 7–10 times higher levels of CD1d than FO B cells and diluted the anti-CD1d staining mAb to selectively reveal CD1high cells. In this staining condition, only B cells located outside the area delimitated by MOMA-1 stained for CD1d in the WT control mouse (Fig. 2,B). In Ii KO mice, however, cells expressing high levels of CD1d were not only seen in the MZ but also within the FO area. In younger Ii KO mice, where MZ-like B cells have not yet accumulated, these CD1high cells were absent from the FO area (data not shown). We conclude that the expanded MZ-like B cell population of adult Ii KO mice invaded the FO area. To search for a molecular mechanism, we investigated the expression of various integrins previously shown to regulate retention to the MZ (28), but did not find differences in their expression pattern (Fig. 2,C). The expression of CXCR5, shown to be required for B cell migration to splenic follicles (29), was not affected either (Fig. 2 C).

FIGURE 2.

Splenic distribution and chemokine receptor pattern of B cells in Ii KO mice. A, Immunohistochemical analysis of frozen splenic sections from 4-mo-old WT and Ii KO stained for MOMA-1 (green) and B220 (red). Data are representative of six mice analyzed in each group. B, Immunohistochemical analysis of frozen splenic sections from 4-mo-old WT, Ii KO, and CD1 KO stained for MOMA-1 (green) and CD1 (red). Data are representative of six mice analyzed in each group. C, Integrin and chemokine receptor pattern expressed by B220+CD21highCD23low MZ B cells of WT and Ii KO cells. Data are representative of four separate experiments.

FIGURE 2.

Splenic distribution and chemokine receptor pattern of B cells in Ii KO mice. A, Immunohistochemical analysis of frozen splenic sections from 4-mo-old WT and Ii KO stained for MOMA-1 (green) and B220 (red). Data are representative of six mice analyzed in each group. B, Immunohistochemical analysis of frozen splenic sections from 4-mo-old WT, Ii KO, and CD1 KO stained for MOMA-1 (green) and CD1 (red). Data are representative of six mice analyzed in each group. C, Integrin and chemokine receptor pattern expressed by B220+CD21highCD23low MZ B cells of WT and Ii KO cells. Data are representative of four separate experiments.

Close modal

Because Ii KO mice lack most CD4 T cells, only T-independent B cell responses could be assessed in vivo. Of particular interest was the response to TI-2 Ags such as TNP-Ficoll, because it is largely restricted to the MZ B cells (12). The anti-TNP IgM and IgG3 responses were largely conserved in Ii KO mice, supporting the preservation of MZ functions (Fig. 3). However, given the actual increase in MZ B cell numbers found in Ii KO mice, one might have expected an augmented instead of a merely conserved response. This relative discrepancy may point to functional defects of Ii KO MZ-like B cells, perhaps as a result of their aberrant location. Alternatively, the anti-TNP response in our system might already be maximal and could not reveal increased MZ B cell frequencies.

FIGURE 3.

T-independent B cell response in Ii KO mice. Serum Abs to the TI-2 Ag TNP-Ficoll in preimmunized (pi) and 2 wk after i.p. injection of Ag.

FIGURE 3.

T-independent B cell response in Ii KO mice. Serum Abs to the TI-2 Ag TNP-Ficoll in preimmunized (pi) and 2 wk after i.p. injection of Ag.

Close modal

In 2-wk-old Ii KO mice, the splenic FO B cell population was already markedly reduced, as was the corresponding recirculating B cells in the bone marrow (Fig. 4, A and B). In contrast, the MZ B cells accumulated much later with a kinetics similar to that of normal MZ B cells which are absent at birth and reach steady state after a period of 3–4 mo. The marked differences between the kinetics of the FO B cell decrease and the MZ B cell accumulation suggested that the latter did not result from the former, i.e., that FO cells were unlikely to disappear as a result of an increased transition into MZ B cells.

FIGURE 4.

B cell ontogeny in MHC II and CD1 Ag presentation mutant mice. A, CD21/CD23 FACS dot plots gated on B220+ splenocytes at 2, 5, and 14 wk of age in WT and Ii KO mice (numbers are percentage of B220+ cells in indicated gates). B, Mature recirculating B220 highIgM+ cells in the BM of 2-wk-old WT and Ii KO. C, Age-dependent accumulation of CD21highCD23low MZ B cells (shown as percentage of splenic B220+ cells at 3, 4, and 6 mo) in WT, I-Aβ KO, CIITA KO, CD1KO, and Ii KO mice, all in the C57BL/6 background. D, Representative CD21/CD23 FACS dot plots of gated B220+ splenocytes in 3-mo-old WT and mutant mice.

FIGURE 4.

B cell ontogeny in MHC II and CD1 Ag presentation mutant mice. A, CD21/CD23 FACS dot plots gated on B220+ splenocytes at 2, 5, and 14 wk of age in WT and Ii KO mice (numbers are percentage of B220+ cells in indicated gates). B, Mature recirculating B220 highIgM+ cells in the BM of 2-wk-old WT and Ii KO. C, Age-dependent accumulation of CD21highCD23low MZ B cells (shown as percentage of splenic B220+ cells at 3, 4, and 6 mo) in WT, I-Aβ KO, CIITA KO, CD1KO, and Ii KO mice, all in the C57BL/6 background. D, Representative CD21/CD23 FACS dot plots of gated B220+ splenocytes in 3-mo-old WT and mutant mice.

Close modal

Because the Ii KO phenotype could be related to a loss of function of the various molecules that Ii reportedly binds to, including MHC class II, DM, and CD1d (18, 21, 22, 23), we analyzed B cell subsets in mice deficient for I-Aβ, CIITA KO, where the expression of both I-Aβ and I-Aα is defective, DM, or CD1d. Interestingly, all of these mutant mice exhibited normal B cell subsets, including FO and MZ B cells (Fig. 4, C and D). We conclude that the B cell defects are unlikely to be the consequence of loss of function of MHC class II, DM, or CD1d.

To test whether the accumulation of MZ-like B cells and the reduced number of FO B cells in Ii KO mice were intrinsic to the B cells or were due to the absence of Ii in other cell types, we created fetal liver chimeric mice by reconstituting lethally irradiated B6 mice with a mixture of B6 WT (CD45.2) and B6.Ii KO (CD45.1) fetal liver cells (2.5 × 106 cells of each). As shown in Fig. 5, the B, CD4, and CD8 T cell compartments originating from the WT and Ii KO precursors in the spleen of these adult chimeras were similar in size. However, the Ii KO compartment of these chimeras exhibited the same B cell defect as in Ii KO mice, i.e., a marked increase of the MZ-like B cell population and a decrease in the FO compartment of the spleen and a decrease in the recirculating B cell pool in the bone marrow. In contrast, the WT B cell compartment of these chimeras was normal. Similar results were observed in 15 individual chimeras analyzed at 4–6 mo of age.

FIGURE 5.

B cells in mixed WT:Ii fetal liver → WT radiation chimeras. A, B cells originating from WT and Ii KO progenitors are identified by their CD45 allotype staining and analyzed for CD21/CD23 expression in the spleen (after gating on B220+ cells) or identified by B220/IgM staining in the bone marrow. B, CD45 allotype staining in the CD8, CD4, and B220+ splenocytes of the mixed chimeras.

FIGURE 5.

B cells in mixed WT:Ii fetal liver → WT radiation chimeras. A, B cells originating from WT and Ii KO progenitors are identified by their CD45 allotype staining and analyzed for CD21/CD23 expression in the spleen (after gating on B220+ cells) or identified by B220/IgM staining in the bone marrow. B, CD45 allotype staining in the CD8, CD4, and B220+ splenocytes of the mixed chimeras.

Close modal

These results indicate that the B cell defects of Ii KO mice are B cell autonomous rather than the indirect consequence of defects of other lineages.

The decreased numbers of FO B cells and of corresponding recirculating B cells in the bone marrow could be the result of a block of differentiation from immature precursors, or, alternatively, of a decreased life span. As shown in Fig. 6, Ii KO mice under continuous BrdU exposure for 2 wk exhibited a markedly increased frequency of BrdU+ FO B cells (64% vs 34% in WT, on average). After 2 wk of chase, the frequency of BrdU+ FO cells decreased 2-fold (from 64 to 27%) in Ii KO mice, whereas it only slightly decreased (from 34 to 30%) in WT. Together with the decreased size of the FO compartment, these results suggest that the life span of FO B cells is considerably shortened as a consequence of Ii deficiency, resulting in a decreased FO pool with higher turnover rate. NF B cells exhibited marginal changes in their rate of BrdU incorporation and chase, suggesting that the decreased life span of mature FO cells must be the primary cause of B cell deficiency.

FIGURE 6.

Higher turnover and shorter life span of Ii KO FO B cells: BrdU staining of gated splenic FO B cells at 3 mo of age after 2 wk of continuous exposure to BrdU in drinking water (upper panels) then 2 wk without BrdU (lower panels). Histograms are representative of 10 mice analyzed for each group (see Table I for statistics).

FIGURE 6.

Higher turnover and shorter life span of Ii KO FO B cells: BrdU staining of gated splenic FO B cells at 3 mo of age after 2 wk of continuous exposure to BrdU in drinking water (upper panels) then 2 wk without BrdU (lower panels). Histograms are representative of 10 mice analyzed for each group (see Table I for statistics).

Close modal

A very different picture was observed for MZ-like B cells. Indeed, their rates of incorporation and loss of BrdU were essentially similar in WT and Ii KO cells (Table I). This lack of detectable alteration in the dynamics of the MZ B pool is consistent with the gradual accumulation of MZ-like cells over several months and supports the notion that the MZ B cell accumulation is independent of changes in the FO B cell compartment.

Table I.

Percentage of BrdU incorporation in B cell subsets of WT, li KO, and li/I-Aβ KO mice

% BrdU+
2-wk Labeling2-wk Labeling + 2-wk chase
NFFOMZNFFOMZ
Spleen       
 C57BL/6 89 ± 3a 34 ± 2 35 ± 2 8 ± 5 30 ± 1 32 ± 2 
 li KO 65 ± 5 64 ± 4 33 ± 2 18 ± 5 27 ± 6 29 ± 2 
 li/I-Aβ KO 62 ± 1 28 ± 1 30 ± 3 13 ± 4 27 ± 3 25 ± 2 
Lymph node       
 C57BL/6  31 ± 2   27 ± 3  
 li KO  61 ± 1   30 ± 1  
 li/I-Aβ KO  28 ± 1   28 ± 3  
% BrdU+
2-wk Labeling2-wk Labeling + 2-wk chase
NFFOMZNFFOMZ
Spleen       
 C57BL/6 89 ± 3a 34 ± 2 35 ± 2 8 ± 5 30 ± 1 32 ± 2 
 li KO 65 ± 5 64 ± 4 33 ± 2 18 ± 5 27 ± 6 29 ± 2 
 li/I-Aβ KO 62 ± 1 28 ± 1 30 ± 3 13 ± 4 27 ± 3 25 ± 2 
Lymph node       
 C57BL/6  31 ± 2   27 ± 3  
 li KO  61 ± 1   30 ± 1  
 li/I-Aβ KO  28 ± 1   28 ± 3  
a

Mean ± SD are calculated from 10 mice in each group.

Although the B cell defects did not appear to be a consequence of a loss of function of Ii-regulated, surface-expressed molecules such as MHC class II or CD1d, we considered the possibility that it might result from a dysregulation of these molecules in the absence of Ii. To test this hypothesis, we crossed Ii KO mice with mice deficient in I-Aβ or CD1d and examined the B cell subsets in Ii/I-Aβ and Ii/CD1d doubly deficient mice. Strikingly, the FO B cell defect was fully reversed in Ii/I-Aβ doubly deficient mice (Fig. 7), as were the corresponding defects in recirculating B cells in the lymph node and the bone marrow (see gated populations in middle and lower panels of Fig. 7,A). These cells belong to the same recirculating pool as the splenic FO B cell pool. We further measured the rate of BrdU incorporation and chase in Ii/I-Aβ KO mice and found that, as expected, the Ii KO-associated anomalies of FO B cells were corrected (Table I). In contrast, Ii/CD1d doubly deficient mice exhibited the same defect as Ii KO mice, demonstrating the specific role of I-Aβ in the absence of Ii (Fig. 7 A).

Interestingly, and in contrast with FO B cells, the MZ B cell pool of Ii KO mice remained increased, confirming that the FO and MZ B cell defects were independent and demonstrating that they were controlled by different mechanisms. However, closer in situ examination by immunohistochemistry revealed that, after repopulation of the FO B cell compartment in Ii/I-Aβ doubly deficient mice, the expanded MZ B cells reintegrated their normal MZ location (Fig. 8). Together with the normal integrin pattern shown in Fig. 2 C, this striking result suggests that the expanded MZ B cells of Ii KO mice invaded the FO zone because it was empty rather than because they lacked the integrins required to remain in the MZ.

FIGURE 8.

MZ reintegration of MZ-like B cells in Ii/I-Aβ double KO mice. Splenic frozen sections of 4-mo-old mice of indicated genotypes were stained for MOMA-1 (green) and CD1 (red). The histogram represents the mean ± SD of the width of the MZ measured on an average of 20 follicles in each of four individual spleens.

FIGURE 8.

MZ reintegration of MZ-like B cells in Ii/I-Aβ double KO mice. Splenic frozen sections of 4-mo-old mice of indicated genotypes were stained for MOMA-1 (green) and CD1 (red). The histogram represents the mean ± SD of the width of the MZ measured on an average of 20 follicles in each of four individual spleens.

Close modal

By combining cellular and genetic studies, we have examined the nature and mechanism of the B cell defects associated with Ii deficiency. We refined the analysis of the previously reported defects by extensive four-color FACS analysis with multiple lineage and subset markers, by in situ immunohistochemistry, and performed dynamic studies including BrdU incorporation and chase, ontogenic analysis, and competitive repopulation assays in mixed fetal liver chimeras. In contrast with earlier reports (17, 30, 31) describing an accumulation of immature NF B cells and a decrease in mature FO B cells, suggestive of a block of maturation, our studies define a more complex picture resulting from opposite, cell autonomous defects affecting two different B cell lineages. We confirmed the marked decrease in recirculating FO B cells, but our analysis revealed that the NF pool of immature B cells was actually normal both in steady state and in dynamic or competitive situation. Instead of the maturation block suggested previously, the FO B cell defect could be attributed to a decreased life span of FO B cells. Furthermore, based on a more extensive FACS analysis, we identified a new expanded subset which was likely included in the immature compartment by previous studies because of its CD23lowIgMhighIgDlowprofile, but in fact appeared to be a long-lived, gradually accumulating a population of cells that were indistinguishable from MZ B cells based on the expression of their CD21highCD23lowIgMhighIgDlowCD1dhighCD9high phenotype and their ICAM-1/αL412 integrin and CXCR5 chemokine receptor pattern. Based on immunohistochemical studies, we further demonstrated that these expanding MZ B-like cells were ectopically present in the FO area. Thus, instead of a block in peripheral B cell maturation, our results demonstrated two separate, seemingly independent defects, one sharply reducing the life span of the mature FO B cell compartment and the other leading to the gradual accumulation, over the course of several months, of an expanded population of MZ-like B cells invading the FO area.

Using a panoply of single and double mutant mice, our studies have also revealed novel mechanisms underlying the B cell defects. First, we examined several mutants affecting known partners of Ii molecules, including MHC class II, DM, and CD1d molecules to investigate whether a loss of function in these pathways might reproduce the Ii defects. We found that neither the I-Aβ KO, nor the DM KO, or the CD1d KO exhibited any of the B cell defects found in Ii KO mice. We also examined the CIITA KO mice where expression of both I-Aβ and I-Aα is undetectable and Ii is expressed at reduced levels and found normal B cell development. Second, we bred Ii/I-Aβ and Ii/CD1d double-mutant mice to test whether I-Aβ and CD1d might mediate the B cell defects when expressed as “orphans” without Ii. Whereas Ii/CD1d KO mice behaved like Ii KO, the Ii/I-Aβ double KO completely and selectively corrected the defect in FO B cells by normalizing their reduced life span. This striking observation suggests that it is the presence of Ii-free orphan I-Aβ chains that induces the marked reduction of the FO B cell life span. Interestingly, a B cell defect partially resembling that of Ii KO was recently reported in I-Aα KO mice (19). These mice exhibited a markedly decreased B cell life span, although no increase in MZ-like B cells was reported. This I-Aα KO-associated phenotype, along with the normal B cell development in I-Aβ KO mice, in CIITA KO mice, and in mice lacking the extended MHC class II locus, and with the B cell defects previously reported in mice transgenically expressing an excess of I-Aβ but not I-Aα chains, have suggested the hypothesis that orphan Ii-free or I-Aα-free I-Aβ chains might impair B cell development (17, 19, 20, 32, 33, 34, 35, 36). Our results in Ii/I-Aβ doubly deficient mice provide the first direct, formal demonstration of this hypothesis at the genetic level and identify mature FO B cells as the target of the life span-reducing effect of free I-Aβ chains. The biochemical basis of the reduced life span of FO B cells by orphan I-Aβ chains remains to be explored. However, the selectivity of this life span-reducing effect, which appears to affect a precise developmental stage of FO B cells, but spares the MZ and B1 B cell lineages, as well as other MHC II-expressing lineages such as DC, is remarkable, suggesting the exquisite involvement of an as yet unidentified molecule which might, for example, selectively alter FO B cell survival after binding orphan I-Aβ chains. Although, at present, these findings only apply to mutant mice, they make several interesting predictions with potentially important pathological implications. For example, similarly drastic B cell defects might be expected in mice and perhaps also humans exhibiting diverse mutations favoring the emergence of such orphan I-Aβ chains. Furthermore, the findings underscore the importance of a tightly coordinated control of expression of Ii, I-Aα, and I-Aβ and suggest that conditions creating an imbalance between these molecules might significantly and selectively modulate the life span of the FO B cell and therefore the dynamics of B cell responses.

Interestingly, the gradual, massive accumulation of MZ B cells persisted in Ii/I-Aβ doubly deficient mice, indicating that an independent mechanism was specifically involved in this aberrant expansion. However, instead of invading the FO zone, the expanding MZ B cells were now retained in the MZ. This observation suggested the intriguing possibility that a fully replenished FO zone might be essential to the formation of the MZ, independently of the pattern of integrins or chemokine receptor expressed by MZ B cells.

Some reports have suggested a diversity of functions of Ii inside or outside the immune system, raising the possibility that pleiotropic indirect defects might result from Ii deficiency or overexpression (37, 38, 39, 40). Therefore, the specific B cell development defects reported here might reflect nonphysiological consequences of Ii deficiency as discussed above. Alternatively Ii may regulate exquisite steps of B cell development through molecularly identifiable physiological mechanisms. Thus, Ii-deficient mice might prove useful to investigate various aspects of B cell development, including mechanisms regulating the life span of FO B cells and the development of the MZ.

We thank Mercedes Balazs, Hayet and Abdessattar Bellagha, Cindy Benedict, Jason Cyster, Ron Germain, Bana Jabri, John Kearney, Yijin Li, Flavius Martin, Mark Shlomchick, Martin Weigert, and Woong-Jai Won for helpful discussions, reagents, and technical advice; Shihong Li for assistance with immunohistochemical experiments; Shirley Bond for help with microscopy analysis; and Joel Veiga for technical assistance.

1

This work was supported by Grant AI50847 from the National Institutes of Health and by a Leukemia and Lymphoma Special Fellowship (to K.B).

4

Abbreviations used in this paper: FO, follicular; Ii, invariant chain; MZ, marginal zone; NF, newly formed; CIITA, class II transactivator; BrdU, 5-bromo-2′-deoxyuridine; WT, wild type; TNP, trinitrophenyl; KO, knockout; DC, dendritic cell.

1
Osmond, D. G..
1990
. B cell development in the bone marrow.
Semin. Immunol.
2
:
173
.
2
Hardy, R. R..
1990
. Development of murine B cell subpopulations.
Semin. Immunol.
2
:
197
.
3
Hardy, R. R., C. E. Carmack, S. A. Shinton, J. D. Kemp, K. Hayakawa.
1991
. Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow.
J. Exp. Med.
173
:
1213
.
4
Hardy, R. R., K. Hayakawa.
1995
. B-lineage differentiation stages resolved by multiparameter flow cytometry.
Ann. NY Acad. Sci.
764
:
19
.
5
Allman, D. M., S. E. Ferguson, V. M. Lentz, M. P. Cancro.
1993
. Peripheral B cell maturation. II. Heat-stable antigenhigh splenic B cells are an immature developmental intermediate in the production of long-lived marrow-derived B cells.
J. Immunol.
151
:
4431
.
6
Snapper, C. M., H. Yamada, D. Smoot, R. Sneed, A. Lees, J. J. Mond.
1993
. Comparative in vitro analysis of proliferation, Ig secretion, and Ig class switching by murine marginal zone and follicular B cells.
J. Immunol.
150
:
2737
.
7
Waldschmidt, T., K. Snapp, T. Foy, L. Tygrett, C. Carpenter.
1992
. B-cell subsets defined by the FcεR.
Ann. NY Acad. Sci.
651
:
84
.
8
Roark, J. H., S. H. Park, J. Jayawardena, U. Kavita, M. Shannon, A. Bendelac.
1998
. CD1.1 expression by mouse antigen-presenting cells and marginal zone B cells.
J. Immunol.
160
:
3121
.
9
Oliver, A. M., F. Martin, G. L. Gartland, R. H. Carter, J. F. Kearney.
1997
. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses.
Eur. J. Immunol.
27
:
2366
.
10
Kraal, G..
1992
. Cells in the marginal zone of the spleen.
Int. Rev. Cytol.
132
:
31
.
11
Gray, D., I. C. MacLennan, H. Bazin, M. Khan.
1982
. Migrant 32 μ+δ+ and static μ+δ B lymphocyte subsets.
Eur. J. Immunol.
12
:
564
.
12
Guinamard, R., M. Okigaki, J. Schlessinger, J. V. Ravetch.
2000
. Absence of marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral response.
Nat. Immunol.
1
:
31
.
13
Chen, X., F. Martin, K. A. Forbush, R. M. Perlmutter, J. F. Kearney.
1997
. Evidence for selection of a population of multi-reactive B cells into the splenic marginal zone.
Int. Immunol.
9
:
27
.
14
Martin, F., J. F. Kearney.
2000
. Positive selection from newly formed to marginal zone B cells depends on the rate of clonal production, CD19, and btk.
Immunity
12
:
39
.
15
Martin, F., A. M. Oliver, J. F. Kearney.
2001
. Marginal zone and B1 B cells unite in the early response against T-independent blood-borne particulate antigens.
Immunity
14
:
617
.
16
Martin, F., J. F. Kearney.
2002
. Marginal-zone B cells.
Nat. Rev. Immunol.
2
:
323
.
17
Shachar, I., R. A. Flavell.
1996
. Requirement for invariant chain in B cell maturation and function.
Science
274
:
106
.
18
Cresswell, P..
1996
. Invariant chain structure and MHC class II function.
Cell
84
:
505
.
19
Rolink, A. G., T. Brocker, H. Bluethmann, M. H. Kosco-Vilbois, J. Andersson, F. Melchers.
1999
. Mutations affecting either generation or survival of cells influence the pool size of mature B cells.
Immunity
10
:
619
.
20
Markowitz, J. S., P. R. Rogers, M. J. Grusby, D. C. Parker, L. H. Glimcher.
1993
. B lymphocyte development and activation independent of MHC class II expression.
J. Immunol.
150
:
1223
.
21
Kang, S. J., P. Cresswell.
2002
. Regulation of intracellular trafficking of human CD1d by association with MHC class II molecules.
EMBO J.
21
:
1650
.
22
Jayawardena-Wolf, J., K. Benlagha, Y. H. Chiu, R. Mehr, A. Bendelac.
2001
. CD1d endosomal trafficking is independently regulated by an intrinsic CD1d-encoded tyrosine motif and by the invariant chain.
Immunity
15
:
897
.
23
Lindstedt, R., M. Liljedahl, A. Peleraux, P. A. Peterson, L. Karlsson.
1995
. The MHC class II molecule H2-M is targeted to an endosomal compartment by a tyrosine-based targeting motif.
Immunity
3
:
561
.
24
Chang, C. H., S. Guerder, S. C. Hong, W. van Ewijk, R. A. Flavell.
1996
. Mice lacking the MHC class II transactivator (CIITA) show tissue-specific impairment of MHC class II expression.
Immunity
4
:
167
.
25
Park, S. H., D. Guy-Grand, F. A. Lemonnier, C. R. Wang, A. Bendelac, B. Jabri.
1999
. Selection and expansion of CD8α/α1 T cell receptor α/β1 intestinal intraepithelial lymphocytes in the absence of both classical major histocompatibility complex class I and nonclassical CD1 molecules.
J. Exp. Med.
190
:
885
.
26
Fung-Leung, W. P., C. D. Surh, M. Liljedahl, J. Pang, D. Leturcq, P. A. Peterson, S. R. Webb, L. Karlsson.
1996
. Antigen presentation and T cell development in H2-M-deficient mice.
Science
271
:
1278
.
27
Jacob, J., R. Kassir, G. Kelsoe.
1991
. In situ studies of the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl. I. The architecture and dynamics of responding cell populations.
J. Exp. Med.
173
:
1165
.
28
Lu, T. T., J. G. Cyster.
2002
. Integrin-mediated long-term B cell retention in the splenic marginal zone.
Science
297
:
409
.
29
Ansel, K. M., V. N. Ngo, P. L. Hyman, S. A. Luther, R. Forster, J. D. Sedgwick, J. L. Browning, M. Lipp, J. G. Cyster.
2000
. A chemokine-driven positive feedback loop organizes lymphoid follicles.
Nature
406
:
309
.
30
Kenty, G., E. K. Bikoff.
1999
. BALB/c invariant chain mutant mice display relatively efficient maturation of CD4+ T cells in the periphery and secondary proliferative responses elicited upon peptide challenge.
J. Immunol.
163
:
232
.
31
Kenty, G., W. D. Martin, L. Van Kaer, E. K. Bikoff.
1998
. MHC class II expression in double mutant mice lacking invariant chain and DM functions.
J. Immunol.
160
:
606
.
32
Gilfillan, S., S. Aiso, S. A. Michie, H. O. McDevitt.
1990
. Immune deficiency due to high copy numbers of an Ak β transgene.
Proc. Natl. Acad. Sci. USA
87
:
7319
.
33
Singer, S. M., D. T. Umetsu, H. O. McDevitt.
1996
. High copy number I-Ab transgenes induce production of IgE through an interleukin 4-dependent mechanism.
Proc. Natl. Acad. Sci. USA
93
:
2947
.
34
Cosgrove, D., D. Gray, A. Dierich, J. Kaufman, M. Lemeur, C. Benoist, D. Mathis.
1991
. Mice lacking MHC class II molecules.
Cell
66
:
1051
.
35
Tourne, S., T. Miyazaki, P. Wolf, H. Ploegh, C. Benoist, D. Mathis.
1997
. Functionality of major histocompatibility complex class II molecules in mice doubly deficient for invariant chain and H-2M complexes.
Proc. Natl. Acad. Sci. USA
94
:
9255
.
36
Labrecque, N., L. Madsen, L. Fugger, C. Benoist, D. Mathis.
1999
. Toxic MHC class II β chains.
Immunity
11
:
515
.
37
Romagnoli, P., C. Layet, J. Yewdell, O. Bakke, R. N. Germain.
1993
. Relationship between invariant chain expression and major histocompatibility complex class II transport into early and late endocytic compartments.
J. Exp. Med.
177
:
583
.
38
Bonnerot, C., M. S. Marks, P. Cosson, E. J. Robertson, E. K. Bikoff, R. N. Germain, J. S. Bonifacino.
1994
. Association with BiP and aggregation of class II MHC molecules synthesized in the absence of invariant chain.
EMBO J.
13
:
934
.
39
Matza, D., A. Kerem, H. Medvedovsky, F. Lantner, I. Shachar.
2002
. Invariant chain-induced B cell differentiation requires intramembrane proteolytic release of the cytosolic domain.
Immunity
17
:
549
.
40
Matza, D., F. Lantner, Y. Bogoch, L. Flaishon, R. Hershkoviz, I. Shachar.
2002
. Invariant chain induces B cell maturation in a process that is independent of its chaperonic activity.
Proc. Natl. Acad. Sci. USA
99
:
3018
.