Background
The mammalian spleen functions to remove old and damaged erythrocytes, participates in the immune response, particularly against blood-borne pathogens, and is the major site for peripheral B cell development [
1]. Immature surface immunoglobulin-expressing B cells that reach the spleen from the bone marrow (referred to as 'transitional' B cells) represent developmental precursors to mature follicular (FO) B lymphocytes, the major mature B cell population in the spleen [
2,
3]. The other two mature B cell populations consist of non-circulating splenic marginal zone (MZ) B cells that are believed to be derived from transitional cells, and B-1 cells, that are controversial in origin and are enriched in peritoneal and pleural cavities [
4]. FO B cells contribute to most T-cell dependent (TD) responses that induce germinal center (GC) development of affinity-matured long lived plasma cells and memory B cells [
4,
5]. In contrast, MZ cells and B-1 cells are predominately responsible for the initial rapid T-cell independent (TI) IgM antibody response and form an important line of defense against antigens and pathogens in the blood and mucosal sites [
4,
6]. More recent analysis of B cell populations suggests that there exist two distinguishable, long-lived, recirculating post-transitional follicular B cell populations, the mature FO B cell subset described above and a new subset referred to as FO-II B cells. The FO-II B cells differ from FO B cells in that they develop in an antigen-independent manner and may serve as a follicular precursor to both MZ B cells and FO B cells [
7].
Although the precise mechanisms are still unclear, maintenance of each peripheral B cell subset is affected by the availability of resources, the local environment, and interactions with other cell types [
8]. It is known that signaling through the B cell receptor (BCR) is required for the development and maintenance of mature splenic B cells [
9]. Genetic studies utilizing knockout mouse models for various BCR complex components and downstream effectors clearly demonstrate that an adequate BCR signal is indispensable for the development of transitional B cells and further differentiation into mature B cells [
2,
10‐
14]. Furthermore, an ongoing or 'tonic' BCR signal is required for B cells to survive [
11,
15]. This was elegantly shown using an inducible disruption of the BCR in mature B cells, which resulted in the absence of all three mature subsets [
16]. In addition to BCR signaling, both the splenic microenvironment and the ability to respond to locally produced growth factors play equally important roles in the development and maintenance of mature B cells [
17,
18].
Many short-lived mRNAs that encode cytokines and cell survival factors contain an adenosine/uridine-rich element (ARE) in the 3' non-coding region that allows for their post-transcriptional regulation by ARE-binding proteins. AUF1, also known as heterogeneous nuclear ribonucleoprotein D, is important in promoting the decay of ARE-containing mRNAs [
19‐
21]. In addition to mRNA turnover, AUF1 has more recently been implicated in a number of other cellular processes, including mRNA translation [
22] and chromatin remodeling [
23]. The physiological role of AUF1 in controlling the decay of key pro-inflammatory cytokine mRNAs, interleukin (IL)-1β and tumor necrosis factor (TNF)-α, was demonstrated by the increased sensitivity of AUF1
-/- mice to endotoxic shock and lethality [
20]. AUF1 also regulates inflammation within the skin, and AUF1
-/- mice develop atopic-like dermatitis [
24].
Given the high expression of AUF1 in lymphoid organs [
25], we characterized the lymphoid compartment of AUF1-deficient mice. Here we show that AUF1
-/- mice develop spleens of reduced size that support a roughly two-fold reduction in the number of lymphocytes. FO B cells exhibit a two-fold decrease in average half-life in AUF1
-/- mice that correlates with reduced expression of Bcl-2, A1, and Bcl-X
L, as well as prominent apoptosis within splenic follicles. The increased turnover of FO B cells corresponds to their decreased frequency in AUF1
-/- mice. Our studies indicate that AUF1 plays an important role in regulating splenic lymphocyte cellularity and FO B cell maintenance.
Methods
Mice
The AUF1
-/- mice were derived as previously described [
20]. RAG1
-/- mice (B6.129S7-Rag1
tm1Mom) and CD45.1 mice (B6.SJL-
Ptprc
a
Pep3
b
/BoyJ) were purchased from Jackson Labs. For adoptive transfer studies, 5 × 10
6 donor-derived bone marrow cells were injected intravenously into sub-lethally irradiated (500 rad) RAG1
-/- or into lethally irradiated (950 rad) AUF1
+/+ and AUF1
-/- recipient mice. Chimera mice were analyzed 10 weeks post transfer. CD45 allele expression was used to distinguish donor and recipient populations. All mice are kept under specific pathogen-free conditions. All animal protocols were approved by the NYU Institutional Animal Care and Use Committee.
Flow Cytometry
Single cell suspensions were prepared from spleen, inguinal and mesenteric lymph nodes, peritoneum lavage, bone marrow, and thymus. Cells were stained and analyzed as previously described [
20]. Cell sorting was performed on a BD FACSVantage cell sorter. The following Abs were used: from BD PharMingen: anti-B220-PerCP (RA3-6B2), anti-CD3-PerCP (17A2), anti-CD4-FITC (GK1.5), anti-CD8-APC (53-6.72), anti-IgM-FITC (R6-60.2), anti-CD94-APC (18d3), Streptavidin-APC, Streptavidin-FITC; from Caltag: anti-CD62L-PE (MEL-14), anti-CD23-PE (B3B4); from eBioScience: anti-CD44-biotin (IM7), anti-IgM-biotin (II/41), anti-CD5-FITC (53-7.3), anti-CD24-PE (M1/69), anti-IgD-biotin (11-26), anti-CD21/CD35-FITC,-APC (4E3), anti-F4/80-biotin (BM8), anti-CD86-biotin (GL1), anti-CD45.1-biotin (A20), and anti-CD45.2-biotin (104). Bcl-2 levels were detected using the Bcl-2 kit (BD Pharmingen), following manufacturer's instructions. The set contains anti-Bcl-2-PE (3F11) and PE-conjugated isotype control.
Immunohistochemistry
Hematoxylin and eosin stained sections were processed as previously described [
20]. Germinal center staining was performed at 8 and 22 days post-immunization with TNP-Ficoll and TNP-KLH, respectively. The reagents used were from eBioscience: anti-mouse B220 (RA3-6B2); Cell Signaling: rabbit anti-cleaved caspase-3 (5A1); BD Pharmigen: CR1 (8C12); Vector Laboratory: biotinylated peanut agglutinin, biotinylated anti-rabbit IgG, Vectastain ABC-AP, Vectastain ABC, Vector blue AP kit, and Vector NovaRED peroxidase kit.
BrdU labeling
Mice were injected intraperitoneally with 1 mg of BrdU (Sigma-Aldrich) and fed with drinking water containing 1 mg/ml BrdU for 7 days. At 0, 7, and 21 days post BrdU administration, spleens from 3 mice per group were analyzed by FACS. Cells were stained with anti-CD21-FITC and anti-CD24-PE or anti-CD23-PE, and then were fixed for 16 h at 4°C with 3.4% paraformaldehyde and 0.01% Tween-20 in PBS. Cells were washed, incubated for 30 min at 37°C with 5 mM MgCl2, 1% BSA, and DNase I (100U) in PBS, then labeled with biotinylated anti-BrdU (Br-3, Caltag) or biotinylated anti-IgG control (eBioscience), and subsequently stained with Streptavidin-APC. The mean BrdU labeling at 7 days was used to calculate the labeling rate per day; 2.5% and 1.5% per day in KO and WT mice, respectively. To calculate average life span, the mean values were plotted as a function of time in the chase period, and linear regression analysis was performed to calculate time to lose the BrdU labeled population, assuming a constant pool size during the analysis.
B cell isolation, Ca2+ mobilization response, in vitro proliferation and survival assays
Splenic B cells were purified with the MACS CD45R microbead system and were verified by FACS to be ≥ 98% B220+. Splenic B cells or sorted MZ B cells were loaded with 2 μM Fluo-4/AM (Molecular Probes), stimulated with 10 μg/ml F(ab')2 goat anti-IgM Ab and Ca2+ mobilization was recorded on live gated cells. For proliferation assays, 2 × 104 MZ B cells were plated in triplicate in 96 well plates and stimulated with 10 μg/ml F(ab')2 goat anti-IgM Ab or with 10 μg/ml LPS. For the last 8 h of the 48 h culture period the cultures were pulsed with 1 μCi/well [3H]-thymidine. Incorporated radioactivity was quantified by scintillation counting. For CD86 up-regulation studies, CD86 expression was analyzed by flow cytometry before and after 24 h treatment with F(ab')2 goat anti-IgM Ab (10 μg/ml). For cell viability assays, sorted FO B cells were cultured as stated and then analyzed by flow cytometry after the addition of 7-amino-actinomycin (7-AAD) (5 μg/ml).
Quantitative real-time RT-PCR
FACS-sorted splenic B220+CD21intCD24lo FO B cells were homogenized in Trizol (Invitrogen) and mRNA was extracted according to manufacturer's instructions. Bcl-2, A1, and Bcl-XL mRNA levels were determined using the Roche LightCycler system. CT values were used to calculate relative values and were normalized to CT values for cyclophilin A. The following primer pairs were used:
Bcl-2, CCTGTGGATGACTGAGTACC and GAGACAGCCAGGAGAAAT
A1, CTTCAGTATGTGCTACAGGTACCCG and TGGAAACTTGTTTGTAAGCACGTCCAT
Bcl-XL, AGAAGAAACTGAAGCAGAG and TCCGACTCACCAATACCTGCGTCCAT
BAFF, CTGTGGTCACTTACTCCAAAGG and GGATCAGATTCAACGGGTCACG
BAFF-R, GCCCAGACTCGGAACTGTCCCA and GCCCAGTAGAGATCCCTGGGTTCC
CypA, TATCTGCACTGCTAAGACTGAATG and CTTCTTGCTGGTCTTGCCATTCC.
Immunoblot analysis
Immunoblot analysis was performed according to standard protocols using ECL detection (PerkinElmer). Polyclonal antibodies to AUF1 (995) and to eIF4E (Sigma) were used.
Humoral response assay
Total pre-immune serum and TNP-specific titers were determined using the SBA Clonotyping ELISA kit from SouthernBiotech according to manufacturer instructions. To evaluate TI-II responses, mice were immunized with 10 μg TNP-Ficoll (Biosearch Technologies) and bled 7 days post immunization. To evaluate TD responses, mice were immunized with 20 μg TNP-KLH (Biosearch Technologies) with adjuvant (Imject Alum, Pierce) and bled 14 days post immunization. For
in vitro studies, 10
6 purified B cells were stimulated with murine IL-4 (20 ng/ml) and LPS (25 μg/ml) for 96 h. Immunoglobulin secretion in the supernatant was measured as described in immunization studies. RT-PCR looking for post-switch transcripts was performed as previously described [
26].
Statistical analysis
Data are presented as mean ± standard deviation. For statistical comparison of two samples, the two-tailed Student t-test was used for evaluation.
Discussion
Our results show that AUF1 deficiency leads to a disturbance in mature splenic B cell populations and in humoral immune responses. Past studies of mice with defective BCR signaling have demonstrated that the strength of the BCR signal is important in both maturation and determination of the cell-fate of the immature B cells [
2,
11,
15]. Furthermore, these studies collectively indicate that weak BCR signals favor MZ B cell development, whereas relatively strong signals favor the development of FO B cells, and perhaps even stronger signals favor the generation of B1 cells [
11]. Our results indicate that the reduction of B cell numbers is specific to FO B cells, as normal numbers of T1, T2, FO-II, MZP, and MZ B cells are seen in spleens of AUF1
-/- mice, as well as in irradiated RAG1
-/- mice reconstituted with AUF1
-/- bone marrow. These data suggest that there is no developmental block, nor are transitional B cells diverted to the mature MZ B cell population at the expense of FO B cells, in contrast to mice with impaired BCR signaling [
11]. In agreement with this hypothesis, there is no impairment in production of FO B cells
in vitro as shown by maturation assays, or
in vivo as demonstrated by BrdU labeling studies. Overall, we conclude that development of FO B cells is unaffected in AUF1
-/- mice. In addition to its role in development, signals processed by the BCR complex are required for mature B cells to survive in the periphery. The specific loss of FO B cells in AUF1
-/- mice does not seem to result from abnormal BCR signaling, as downstream events of BCR engagement, such as calcium mobilization are unaffected in AUF1
-/- B cells.
The phenotype seen in AUF1
-/- mice does not seem to involve BAFF signaling, which has been shown to be important in both B cell development and homeostasis [
34,
35]. BAFF, a member of the TNF superfamily, and its receptors, BR3 and TACI, are not encoded by ARE-mRNAs and are therefore unlikely to be direct targets of AUF1. As predicted, no difference was observed in the expression of BAFF or its receptor in AUF1-deficient mice. Furthermore, the phenotype of BAFF-deficient mice differs from that of AUF1
-/- mice. BAFF-deficient mice exhibit a loss of both MZ and FO B cells, as well as severely reduced antibody titers in response to both TD and TI antigens [
34,
41]. In contrast, AUF1
-/- mice exhibit a selective loss of the FO B cell subset and a moderate attenuation of the TI humoral response. BAFF signaling is important in CD21 expression in B cells independent of its role in cell survival [
41], and its regulation of CD23 expression is less clear. There are conflicting reports on the role of BAFF on CD23 expression from BAFF-deficient, BAFF-R-deficient, and transgenic BAFF mice [
41‐
43]. In contrast, AUF1-deficient B cells show normal CD21 surface expression, but exhibit a reduction in CD23 surface expression (Figure
1). Although direct functions of AUF1 cannot be ruled out, alterations in local factors such as interleukin-4 in part may explain the decrease in CD23 expression in AUF1
-/- B cells [
44]. CD23 expression on B cells is involved in antigen presentation and IgE response, as demonstrated by the enhanced IgE response elicited in CD23-deficent mice [
45]. We have previously shown that AUF1
-/- mice have increased circulating IgE levels [
36] that may be due to the reduced CD23 expression on AUF1-deficient B cells.
In the periphery, a lymphocyte acquires survival signals through receptors for cytokines, antigens, and hormones [
8,
46]. Failure to adequately acquire such signals results in cell death and thus determines the size of the peripheral lymphocyte pools [
8,
46,
47]. Although signaling through the BCR is required for the maintenance of mature splenic B cells [
9], it is not the only requirement. The splenic microenvironment also plays an essential role in the maintenance of mature B cells [
17]. However, knowledge regarding the nature and regulation of the splenic microenvironment is limited, particularly as it pertains to the accumulation and maintenance of splenic lymphocytes. Our work suggests that AUF1 is important in the regulation of survival signals needed for the maintenance of FO B cells within the spleen, as indicated by decreased Bcl-2, A1, and Bcl-X
L levels. Spleens from AUF1
-/- mice bear a strong resemblance to those of Bcl-2-deficient mice in that they are smaller, show a loss of mature B and T cells, and show many apoptotic cells [
32]. Mature recirculating IgM
loIgD
hi FO B cells from lymph nodes of AUF1
-/- mice exhibit no significant differences in Bcl-2 expression (see Additional file
1), suggesting that the survival signal provided in lymph nodes is unaffected by the absence of AUF1. These findings highlight differences in survival signals necessary for recirculating mature B cells in different lymphoid organs. In support of this hypothesis, it has been recently shown that macrophage migration inhibitory factor (MIF), through the activation of the CD74-CD44 complex, is necessary in the maintenance of mature B cells in the bone marrow, but not in the spleen or lymph nodes [
48]. The spleen-specific alterations in lymphocyte populations in AUF1
-/- mice suggests that AUF1 acts in regulating the local microenvironment, as opposed to regulating global factors in lymphocyte development, such as BCR signaling that appears normal in AUF1
-/- B lymphocytes. Future studies with a more comprehensive screening approach, such as cDNA micro-arrays, may prove valuable in identifying mRNAs regulated by AUF1 that play a role in FO B cell lymphocyte survival within the spleen. CD40 engagement on mature B cells results in induction of anti-apoptotic Bcl-2 family members, which protect these cells from antigen receptor-mediated apoptosis [
33] and are required for GC formation, progression [
49] and TD response [
50]. AUF1
-/- FO B cells were able to appropriately induce the expression of Bcl-2 and A1 in response to CD40 engagement, which supports their ability to form germinal centers in AUF1-deficient mice.
An important phenotype of AUF1-deficiency is a decrease in the number of splenic lymphocytes and a skewing of the B cell subpopulations, with an over-representation of immature and MZ B cells and a decrease in FO B cells. There are several possible mechanisms that may account for this complex phenotype. AUF1 may regulate a survival signal needed for the maintenance of FO B cells, the largest component of splenic B cells, and loss of this population could lead to a decrease in the accumulation of T cells within the spleen. Previous studies on B cell-deficient mice have reported a decrease in splenic T cell numbers [
30,
51]. BCR-deficient mice exhibit a three-fold reduction in the number of splenic CD4 and CD8 T cells, although no loss of T cells was seen in lymph nodes from these animals [
30]. These results suggest that B cells provide signals that promote CD4 and CD8 T cell accumulation and/or survival in the spleen [
30]. Notably, CD3ε- and CD4-deficient mice show a 40% reduction of splenic mature B cells, suggesting that T cells may also impact the survival of B cells within the spleen [
31]. Likewise, AUF1 may regulate survival signals required independently for the maintenance of FO B and T cells. At this time we cannot distinguish whether the loss of T and B cells within the spleen are related, or whether both populations, independent of each other, are deficient in survival signals, resulting in decreased Bcl-2 expression in both populations in AUF1
-/- mice (see Additional file
1). These mechanisms predict that the 50% decrease in splenic B cells in AUF1
-/- mice results from increased FO B cell turnover that outpaces the rate of FO B cell production by two-fold. However, BrdU studies indicate that there is a 2.1-fold decrease in splenic FO B cell average half-life and a 1.7-fold increase of B cell production in AUF1
-/- mice. This would result in a 30% decrease in FO B cells and not the 50% decrease observed in AUF1
-/- mice. Furthermore, if turnover did outpace production by two-fold, one would predict a greater decrease than the 20% seen in recirculating mature B cells in lymph nodes and bone marrow of AUF1
-/- mice.
The data from bone marrow adoptive transfer studies more likely indicate that two separate processes are involved in the observed phenotype: (1) AUF1 is important in the maintenance of FO B cells within the spleen, most likely through regulation of survival factors; and (2) AUF1 impacts the number of lymphocytes within the spleen through the regulation of yet unknown signals. Adoptive transfer studies support the involvement of two separate processes, as reconstitution of irradiated AUF1-/- mice with wild-type bone marrow corrects the shifts in B cell subpopulation proportions but not the overall number of splenocytes. The involvement of two separate processes also better explains the deficit in FO B cells. The 40% decrease in the number of splenocytes, in addition to the increased FO B cell turnover that would result in a 30% decrease in the proportion of FO B cells, more accurately approximates the 50% decrease in the number of FO B cells observed in AUF1-/- mice.
The complex phenotype seen in AUF1-deficient mice is not surprising, given that AUF1 regulates the expression of numerous immune-modulating cytokines and chemokines [
20,
36,
52,
53]. Although we have focused here on defects in the B cell compartment in AUF1-deficient mice, we have previously shown defects in the T cell and macrophage compartments [
36]. Moreover, despite the fact that the FDC network can be visualized, we cannot exclude that the dendritic subset may be affected in AUF1-deficient mice and may impact the function, size, and location of other immune cell compartments. Tristetraprolin, another ARE-binding protein, binds and regulates the expression of mRNAs encoding key regulators of human dendritic cell maturation [
54]. Future analysis of specific targets of AUF1 in different immune cell subsets will be important in dissecting the role of AUF1 on individual populations leading to the overall complex phenotype seen in its deficiency in the mouse model.
Authors' contributions
NS made substantial contributions to conception and design of the study, acquired, analyzed, and interpreted the data, and drafted the manuscript. RJS contributed in data analysis and writing of the manuscript. JL and MB made contributions to the study design and acquisition of data. All authors read and approved the final manuscript.