The RNA-binding protein hnRNP F is required for the germinal center B cell response

Huang, Hengjun; Li, Yuxing; Zhang, Gaopu; Ruan, Gui-Xin; Zhu, Zhijian; Chen, Wenjing; Zou, Jia; Zhang, Rui; Wang, Jing; Ouyang, Yu; Xu, Shengli; Ou, Xijun

doi:10.1038/s41467-023-37308-z

Download PDF

Article
Open access
Published: 30 March 2023

The RNA-binding protein hnRNP F is required for the germinal center B cell response

Hengjun Huang ORCID: orcid.org/0000-0001-9842-9312¹,
Yuxing Li¹,
Gaopu Zhang¹,
Gui-Xin Ruan²,
Zhijian Zhu¹,
Wenjing Chen¹,
Jia Zou³,
Rui Zhang¹,
Jing Wang¹,
Yu Ouyang¹,
Shengli Xu ORCID: orcid.org/0000-0002-2541-3608^4,5 &
…
Xijun Ou ORCID: orcid.org/0000-0002-4525-5205¹

Nature Communications volume 14, Article number: 1731 (2023) Cite this article

3339 Accesses
2 Citations
2 Altmetric
Metrics details

Subjects

Abstract

The T cell-dependent (TD) antibody response involves the generation of high affinity, immunoglobulin heavy chain class-switched antibodies that are generated through germinal center (GC) response. This process is controlled by coordinated transcriptional and post-transcriptional gene regulatory mechanisms. RNA-binding proteins (RBPs) have emerged as critical players in post-transcriptional gene regulation. Here we demonstrate that B cell-specific deletion of RBP hnRNP F leads to diminished production of class-switched antibodies with high affinities in response to a TD antigen challenge. B cells deficient in hnRNP F are characterized by defective proliferation and c-Myc upregulation upon antigenic stimulation. Mechanistically, hnRNP F directly binds to the G-tracts of Cd40 pre-mRNA to promote the inclusion of Cd40 exon 6 that encodes its transmembrane domain, thus enabling appropriate CD40 cell surface expression. Furthermore, we find that hnRNP A1 and A2B1 can bind to the same region of Cd40 pre-mRNA but suppress exon 6 inclusion, suggesting that these hnRNPs and hnRNP F might antagonize each-other’s effects on Cd40 splicing. In summary, our study uncovers an important posttranscriptional mechanism regulating the GC response.

FOXO1 is a master regulator of memory programming in CAR T cells

Article Open access 10 April 2024

FOXO1 enhances CAR T cell stemness, metabolic fitness and efficacy

Article Open access 10 April 2024

Depleting myeloid-biased haematopoietic stem cells rejuvenates aged immunity

Article 27 March 2024

Introduction

RNA-binding proteins (RBP) have emerged as critical players in the post-transcriptional regulation of gene expression¹. Heterogeneous nuclear ribonucleoproteins (hnRNP) are a large family of typical RBPs expressed abundantly in mammalian cells and comprising about 30 members, including 20 major types of hnRNPs A to U and several minor types of hnRNPs². Four different types of RNA-binding domains (RBD) are found in hnRNP proteins, including RNA recognition motif (RRM), non-classical quasi-RRM (q-RRM), glycine-rich domain harboring an RGG box, and K homology (KH) domain³.

HnRNP F belongs to the hnRNP F/H subfamily and contains three non-classical q-RRMs⁴. The best-known function of hnRNP F/H proteins is to regulate RNA alternative splicing (AS). HnRNP F was reported to regulate AS of several important genes, such as c-Src, Bcl-xl, Mcl-1, and Tcf3, primarily via exon-related AS events^5,6,7,8. In vitro studies have demonstrated that hnRNP F participates in AS regulation by directly binding to the G-rich pre-mRNA sequence (G-tract) through its q-RRMs^9,10,11. Furthermore, early studies showed that hnRNP F could bind to the adjacent regions of the target pre-mRNA exons and recruit splicing factors U1 or U2 to promote splicing of the target exons^12,13. In addition to regulating RNA AS, previous studies also found that hnRNP F could regulate mRNA stability, translation, and alternative polyadenylation (APA)^{14,15,16,17,18}. It was documented that hnRNP F could modulate the APA of immunoglobulin heavy (IgH) chain mRNA, promoting the expression of membrane-anchored B cell receptors (BCR) and repressing the expression of secreted Ig proteins in mouse B cell lines¹⁴. However, the physiological role of hnRNP F in B cell development and activation remains unknown.

B cell development starts from the hematopoietic stem cells in the bone marrow (BM) or fetal liver¹⁹. Through a tightly controlled multi-step process, progenitor B cells develop into mature B cells in the BM and then populate secondary lymphoid tissues. In the periphery, naïve B cells become activated upon encountering foreign antigens and differentiate into long-lived antibody-secreting plasma cells (PC) and memory B cells, establishing lifelong humoral immunity. Most long-lived PCs and memory B cells are generated in response to T cell-dependent (TD) proteinous antigens in germinal centers (GC), specialized microstructures in the secondary lymphoid tissues, such as the spleen and lymph nodes²⁰. Upon engagement by TD antigens, antigen-specific follicular B (FOB) cells upregulate chemokine receptor CCR7 that draws them towards the boundary of T and B cell zones in the spleen and lymph nodes, where they interact with CD4⁺ T cells that detect the cognate antigens presented by dendritic cells²¹. Subsequently, some activated FOB cells travel back to the center of B cell follicles and become GC B cells. GCs can be divided into the dark zone (DZ) and the light zone (LZ). In the DZ, B cells undergo robust proliferation and extensive somatic hypermutation. In contrast, in the LZ, B cells with high-affinity BCRs are positively selected through competing for the scarce antigens displayed by follicular dendritic cells and for the help of the limited number of T follicular helper cells²². The positively selected LZ B cells can return to DZ for further proliferation and mutation of their immunoglobulin variable region genes. After multiple rounds of shuttling, GC B cells whose BCRs are of high affinity for the specific antigens eventually differentiate into long-lived PCs and memory B cells and exit GCs²³.

Besides the primary signal transduced by BCR, secondary signals propagated by costimulatory receptors are also essential for GC response²⁴. CD40 is one of the best-characterized costimulatory molecules expressed on antigen-presenting cells, including B cells. The engagement of CD40 by CD40 ligand (CD40L) on T cells triggers multiple signaling cascades, leading to downstream immune and inflammatory responses. CD40-mediated signaling is essential for TD antibody response in almost all the stages, ranging from the interaction of cognate B and T cells, IgH chain class-switching, initiation and maturation of GCs, and antibody affinity maturation^25,26. A recent study using CRISPR/Cas9 whole-genome screening of a human B cell line unveiled multiple previously unappreciated mechanisms for regulating the CD40 pathway²⁷. Interestingly, at the post-transcriptional level, it was found that CUGBP and Elav-like family member 1 (CELF1) protein could promote Cd40 exon 6 inclusion, leading to the expression of signaling-competent membrane-bound CD40.

Here we specifically ablate hnRNP F in B cell lineage and find that hnRNP F is required for class-switched high-affinity antibody production during TD antibody response. In the absence of hnRNP F, GC B cell formation is profoundly impaired. Furthermore, we uncover that hnRNP F is required for proper cell surface expression of CD40 in B cells by regulating the AS of Cd40 pre-mRNA.

Results

B cell development is largely normal in hnRNP F-deficient mice

First, we examined if Hnrnpf gene was differentially expressed during B cell development by analyzing the data available in the Immunological Genome Project (ImmGen) database (https://www.immgen.org/). We found that Hnrnpf gene expression was comparable in all the B cell subsets, including B220⁺CD43⁺CD24⁻BP-1⁻ pre-pro-B (Hardy fraction (Fr.) A), B220⁺CD43⁺CD24⁺BP-1⁻ early pro-B (Fr. B), B220⁺CD43⁺CD24^low(lo)BP-1⁺ late pro-B (Fr. C) and B220⁺CD43⁻IgM⁺IgD⁻ immature B (Fr. E) cells, based on the Hardy classification, in the BM, and transitional 1 (T1), T2, T3 B, marginal zone B (MZB), GC B, memory B, and plasma cells in the spleen (Supplementary Fig. 1a). We further validated the result with a real-time quantitative reverse transcription PCR (qRT-PCR) analysis to determine the Hnrnpf mRNA levels in the various B cell subsets sorted from wild-type mice. Consistent with the data obtained from the ImmGen database, the expression of Hnrnpf did not differ much amongst most of the B cell subsets, except for GC B cells that exhibited a slightly, but not statistically significant, increased expression level of Hnrnpf (Supplementary Fig. 1b).

To determine the role of hnRNP F in B cell physiology, we crossed mice harboring loxP-flanked (floxed) Hnrnpf alleles with Cd19^Cre/+ mice, which specifically express Cre recombinase in B cells, to generate Hnrnpf ^fl/flCd19^Cre/+ (Hnrnpf bKO) mice. We first investigated B cell development in the BM of Hnrnpf bKO mice by flow cytometry. Using a gating strategy based on Hardy classification (Supplementary Fig. 1c), we detected indistinguishable pre-pro-B (Fr. A), early pro-B (Fr. B), late pro-B (Fr. C) and large pre-B (Fr. C’, CD24^high(hi) BP-1⁺) subsets in the B220⁺CD43⁺ early B cell compartment, as well as IgM⁻IgD⁻ small pre-B (Fr. D), IgM⁺IgD⁻ immature (Fr. E) and IgM⁺IgD⁺ recirculating mature (Fr. F) B cells in the B220⁺CD43⁻ B cell compartment in the BM between Hnrnpf ^fl/flCd19^+/+ (WT) and Hnrnpf bKO mice (Fig. 1a, b). These results suggest that B cell development in the BM is unaffected in the Hnrnpf bKO mice. As previous studies showed that CD19-Cre mediated deletion of some floxed genes was inefficient in early B cell populations in the BM compared to mature B cells in the spleen^28,29, we next asked if the undisturbed early B lymphopoiesis phenotype in the Hnrnpf bKO mice was due to the incomplete ablation of the floxed Hnrnpf alleles. We purified various B cell subsets from the BM and spleen of WT and Hnrnpf bKO mice and determined Hnrnpf mRNA levels by qRT-PCR analysis. We found that CD19-Cre-mediated deletion of Hnrnpf gene was only about 20% in Fr. B (early pro-B) cells and the deletion was increased to 50% in the more mature Fr. C (late pro-B), C’ (large pre-B) and Fr. D (small pre-B) cells, and was further improved to 85 and 90% in the Fr. E (immature B) and splenic B cells, respectively (Supplementary Fig. 1d). Immunoblotting further confirmed that hnRNP F protein was barely detected in the Hnrnpf bKO splenic B cells (Supplementary Fig. 1e). These results suggest that a more explicit determination of hnRNP F’s role in early B lymphopoiesis requires a more efficient Hnrnpf gene deletion in B cells mediated by other Cre-recombinases, such as Mb1-Cre. Next, we examined B cell maturation in the periphery by flow cytometry (Supplementary Fig. 1f). WT and Hnrnpf bKO mice manifested indistinguishable B220⁺CD93⁺ transitional B cell subsets (stage I to III) in their spleens (Fig. 1c, d). Interestingly, the Hnrnpf bKO mice exhibited an increase in the percentage and absolute number of B220⁺CD93⁻CD23⁻CD21⁺ MZB cells in their spleen compared with the WT mice. Conversely, the percentage of B220⁺CD93⁻CD23⁺CD21⁻ FOB cells was lower than in the WT mice, though the difference in the absolute cell number was not statistically significant (Fig. 1c, d). Thus, these data demonstrate that the hnRNP F-deficient mice have largely unperturbed B cell development except for the slightly increased MZB cell population.

**Fig. 1: hnRNP F-deficient mice have normal B cell development.**

HnRNP F-deficient mice have a defective TD antibody response

We next examined the antibody response in the Hnrnpf bKO mice by challenging them with TD antigen 4-hydroxy-3-nitrophenyl-acetyl (NP) hapten conjugated to chicken gamma globulin (CGG). NP-specific IgM and class-switched different isotypes of IgG antibodies in the sera were measured by ELISA at various time points post-immunization. We noticed that Hnrnpf bKO mice produced comparable, if not slightly elevated, titers of NP-specific IgM antibodies on par with WT mice across the time points examined (Fig. 2a). In contrast, Hnrnpf bKO mice had pronouncedly decreased NP-specific IgG1, IgG2b, and IgG3 antibody titers than the WT mice (Fig. 2b–d). Furthermore, consistent with their normal NP-specific IgM but significantly decreased NP-specific IgG1 antibody titers in the sera, we also detected markedly reduced NP-specific IgG1 antibody-secreting cells (ASC) by ELISPOT assay in the spleen and BM of Hnrnpf bKO mice compared to the WT mice (Fig. 2e, f). These results suggest Hnrnpf bKO mice have defective class-switched IgG antibody production in response to the TD antigen challenge.

**Fig. 2: hnRNPF is required for TD antibody immune response.**

We also performed ELISA to measure the relative changes in antigen-specific serum IgG1, IgG2b, and IgG3 antibody affinity by determining the ratios of antibody titers binding low (NP₂) to highly (NP₂₀) haptenated BSA. We found that the ratios of NP₂/NP₂₀ IgG1, IgG2b and IgG3 titers from Hnrnpf bKO mice were significantly lower than those from the WT mice (Fig. 2g–i). These results suggest that Hnrnpf bKO mice have compromised antibody affinity maturation during TD immune response.

Defective IgG1 antibody class-switching in hnRNP F-deficient B cells

The unchanged IgM but decreased class-switched IgG antibody titers suggest that hnRNP F-deficiency may affect antibody class-switch recombination (CSR). Thus, we first measured IgM, IgG1, IgG2b, IgG3, and IgA antibodies in sera from unimmunized WT and Hnrnpf bKO mice. Interestingly, we found a significantly reduced IgG1 antibody titer, and a slightly decreased IgG3 antibody level but normal IgG2b and IgA antibody levels in the Hnrnpf bKO mice (Fig. 3a). These results suggest that Hnrnpf bKO B cells might have intact CSR machinery but are defective in responding to signals that can induce class-switching of IgG1, and IgG3 to a lesser extent.

**Fig. 3: hnRNP F is required for IgG1 CSR.**

Next, we examined IgG1 and IgG3 antibody CSR in vitro. We first induced IgG3 antibody CSR by culturing WT and Hnrnpf bKO splenic B cells in the presence of LPS for three days, followed by flow cytometric analysis. Equivalent percentages of IgG3⁺ cells were detected in the WT and Hnrnpf bKO B cell culture (Fig. 3b, c). In contrast, the IgG1 antibody CSR was severely compromised in the absence of hnRNP F when B cells were cultured with LPS and interleukin 4 (IL4), anti-CD40 antibody and IL4, and anti-CD40 antibody, IL4, and IL21, though the addition of IL21 seemed to be able to partially improve IgG1 antibody CSR in the mutant B cells (Fig. 3b, c). These results suggest that hnRNP F is essential for IgG1 antibody CSR.

Defective GC B cell response in Hnrnpf bKO mice

During TD antibody response, the class-switched antigen-specific antibodies with high affinities are produced via GC response³⁰. Given the impaired production of antigen-specific antibodies and defective IgG1 antibody CSR in Hnrnpf bKO mice, we sought to examine the GC response in detail in the mutant mice. We first challenged WT and Hnrnpf bKO mice with NP-CGG followed by an assessment of the peaked GC B cell response 10 days post-challenge by flow cytometry (Supplementary Fig. 1g, h). We found that the frequency and number of B220⁺Fas⁺CD38⁻ GC B cells were severely reduced by more than 70% in the spleen of Hnrnpf bKO mice compared to the WT mice (Fig. 4a, b). We also noticed that the percentage and number of NP-specific IgG1⁺NIP⁺ B cells were also significantly lower in Hnrnpf bKO than in the WT mice (Fig. 4c, d). These results indicate that the GC B cell response is severely compromised in Hnrnpf bKO mice.

**Fig. 4: hnRNP F is essential for GC B cell formation.**

To gain more insights into how hnRNP F-deficiency affected GC B cell response, we further examined the distribution of CXCR4^high(hi) CD83^low(lo) DZ and CXCR4^loCD83^hi LZ B cells in the GC B cell compartment by flow cytometry. To this end, we first gated NP⁺ cells, which were genuine activated B cells upon the NP-CGG immunization, followed by the analysis of DZ and LZ B cells in WT and Hnrnpf bKO (Supplementary Fig. 2a). Surprisingly, we found that the ratios of NP-specific DZ to LZ B cells were equivalent between the WT and Hnrnpf bKO mice, albeit the percentage of the total NP⁺ GC B cells was largely reduced in the Hnrnpf bKO mice (Supplementary Fig. 2a, b). However, when we examined the Hnrnpf mRNA levels in the GC B cells from the Hnrnpf bKO and WT mice, we found that they were largely comparable (Supplementary Fig. 2c). We reasoned that the GC B cells in the Hnrnpf bKO mice were most likely from an expansion of the hnRNP F-proficient GC B cells differentiating from the residual FOB cells that have escaped from CD19-Cre-mediated Hnrnpf deletion.

Next, we attempted to understand how hnRNP F-deficiency caused the saliently diminished GC B cells by examining the early GC B cell response when the nascent GC B cells were just formed (Supplementary Fig. 1i). We noticed that Hnrnpf bKO mice manifested a drastically reduced B220⁺IgD⁻GL-7⁺CD38⁻ early GC B cell population in the spleen on day 5 post-immunization (Fig. 4e, f). Enumeration of these early GC B cells revealed that the early GC B cell population was dramatically diminished in the Hnrnpf bKO mice as their absolute number was decreased by about 80% compared to the WT mice (Fig. 4g). These data suggest that the generation of GC B cells is severely compromised in the absence of hnRNP F. We also set up an in vitro induction of GC-phenotype B (iGB) cells as described previously³¹. Consistent with the results of early GC B cells, the generation of iGB cells was reduced by approximately 50% compared to the WT cells (Fig. 4h, i). Intriguingly, the mRNA level of Hnrnpf in iGB culture using the mutant B cells was about 60% of that in the culture using the WT cells (Fig. 4j), much higher than that in the freshly purified and yet-to-be-differentiated Hnrnpf bKO B cells, which was only 10% of the level in the WT B cells (Supplementary Fig. 1d). One of the possible reasons for the drastically different mRNA levels of Hnrnpf in mutant B cells before and after iGB differentiation could be the hnRNP F-deficient early differentiating GC B cells were defective in proliferation, such that they were much less representative compared to the deletion-escaped hnRNP F-proficient differentiating GC B cells that can proliferate normally. To test this possibility, we cultured WT and mutant B cells labeled with cell tracer under the iGB differentiation condition and examined the proliferation by flow cytometry. It was revealed that Hnrnpf bKO B cells were defective in cell proliferation (Fig. 4k). It is known that transcriptional factor c-Myc is induced during and necessary for GC B cell response^32,33. We next examined the c-Myc expression by qRT-PCR and found that the upregulation of c-Myc was severely compromised in cultured Hnrnpf bKO B cells than their WT counterparts (Fig. 4l). Collectively, these results suggest that hnRNP F plays an essential role in regulating GC B cell formation by promoting early GC B cell proliferation, which requires c-Myc upregulation.

HnRNP F controls AS of Cd40 pre-mRNA in FOB cells

FOB cells are the prototypic B cell subset that responds to TD antigen mounting robust antibody response through GC reaction^34,35. Given that hnRNP F-deficient mice manifested a slightly decreased percentage of FOB cells and drastically compromised early GC B cell formation, we hypothesized that hnRNP F could be required to maintain FOB cells and their subsequent activation and differentiation into GC B cells. Thus, we performed RNA sequencing (RNA-Seq) on FACS-sorted WT and Hnrnpf bKO FOB cells. Analysis of RNA-Seq data revealed that mRNA abundances of 157 genes were significantly increased and that of 79 genes were decreased (FDR (false discovery rate) ≤ 0.05, fold change ≥ 2) in Hnrnpf bKO FOB cells compared to their WT counterparts (Supplementary Fig. 3a and Supplementary Data 1). We further performed a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the differentially expressed genes. Interestingly, PI3K-Akt and MAPK signaling pathways were identified as the top two pathways showing enrichment for genes differentially expressed in hnRNP F-deficient FOB cells (Supplementary Fig. 3b). These data imply that hnRNP F is essential for regulating the expression of genes encoding molecules crucially involved in PI3K-Akt and MAPK signaling pathways in FOB cells.

Since hnRNP F’s primary function in the post-transcriptional gene regulation is to modulate AS of pre-mRNAs^5,36,37,38, we next focused on studying differential AS events between the WT and Hnrnpf bKO FOB cells using rMATS, a robust computational tool for detecting differential AS events from the RNA-Seq data³⁹. Among the five types of AS events, exon skipping (SE) is the most affected by hnRNP F-deficiency, with 657 events more skipped in Hnrnpf bKO than in the WT B cells and 360 events more included in Hnrnpf bKO than in the WT B cells (Fig. 5a and Supplementary Data 2). This result is consistent with previous reports showing that hnRNP F mainly regulates exon skipping or inclusion^6,8,13,37,40. Among the top 10 genes with the highest aberrant AS events, the Cd40 gene drew our attention (Fig. 5b), given that CD40 is critical for antibody CSR, c-Myc induction, and GC B cell proliferation^24,41. However, we noticed that the abundance of Cd40 mRNA was not significantly affected in the mutant FOB cells (Fig. 5c). As presented in sashimi plots, exon junction visualization revealed that hnRNP F predominantly controls AS of Cd40 pre-mRNA from exon 6 to 9 but not from exon 1 to 5 (Fig. 5d). Further analysis revealed multiple aberrant Cd40 AS events in Hnrnpf bKO FOB cells, including skipped exon (SE) 6, 7, and 8, mutually exclusive exon (MXE) 7 and 8, and retained intron (RI) 7 (Fig. 5e).

To validate Cd40 AS changes in hnRNP F-deficient FOB cells, we performed RT-PCR and Sanger Sequencing to examine the abundance of various Cd40 mRNA variants using previously reported primers (Fig. 5f), which can cover exons 5–9⁴². We showed that Cd40 pre-mRNA in WT FOB cells had an AS pattern consistent with the previous reports⁴². As the sizes of variant 1 (V1) and V4 PCR products were very similar and could not be distinguished readily by the electrophoresis, we performed Sanger sequencing and showed that V1, the membrane-bound and signaling-competent form, but not V4, the signaling-incompetent form of Cd40, was predominant in both WT and KO B cells (Supplementary Fig. 3c). Interestingly, the loss of hnRNP F resulted in a drastic decrease in Cd40 V1 accompanied by a significant increase in Cd40 V2, as evidenced by the markedly increased V2/V1 ratio (Fig. 5g). We also detected two previously unidentified variants, one with exon 6 skipped and intron 7 retention and the other one with exons 6, 7 and 8 all skipped. Interestingly, a small amount of the variant with exons 6, 7 and 8 skipping was also present in the WT B cells, but it had not been reported previously (Fig. 5g). Thus, we named it V6. As the intron 7 retention could not be well-discerned in the PCR products amplified with primers P1 and P2, we used primers P3 and P4 to further evaluate the proportion of intron 7 retention. A higher ratio of variants with intron 7 retention to that with no retention was detected in Hnrnpf bKO B cells compared with the WT B cells (Fig. 5h). Together these results suggest that hnRNP F regulates the AS of Cd40 pre-mRNA by promoting exons 6, 7, and 8 inclusion and intron 7 removal.

A previous study of B cell lines showed that hnRNP F promotes the membrane-bound IgM (mIgM) expression and inhibits the secretory type IgM (sIgM) production through regulating APA of IgH chain mRNA¹⁴. Thus, we also compared the expression of mIgM on the surface of total splenic B cells and FOB cells from the WT and Hnrnpf bKO mice by flow cytometry and found them to be comparable (Supplementary Fig. 3d, e). Moreover, the analysis of RNA sequencing data also indicated that mIgM-encoding transcript was predominantly and equivalently present in WT and Hnrnpf KO FOB cells, and they barely had any sIgM-encoding transcript (Supplementary Fig. 3f). These results suggest hnRNP F-deficiency does not affect APA of IgH chain mRNA in mouse splenic B cells.

HnRNP F regulates GC B cell formation by promoting signaling-competent CD40 expression

Given the dramatically altered AS of Cd40 mRNA in Hnrnpf bKO FOB cells (Fig. 5), we asked if the cell surface expression of CD40 was affected in the mutant cells. Indeed, flow cytometric analysis revealed that FOB cells from the unimmunized Hnrnpf bKO mice expressed much lower CD40 on their cell surface than the cells from the WT mice (Fig. 6a). Next, we examined the cell surface expression of CD40 in B cells from immunized mice. Similar to the situation of resting FOB cells, the CD38⁺Fas⁻ non-GC B cells, supposedly non-dividing cells, from Hnrnpf bKO mice expressed much lower levels of CD40 than their WT counterparts (Fig. 6b). However, consistent with the largely normal expression of Hnrnpf (Supplementary Fig. 2c), the residual CD38⁻Fas⁺ GC B cells from the Hnrnpf bKO mice had similar levels of CD40 compared to their counterparts from the WT mice, except for a small fraction (~15-20%) of GC B cells expressed lower levels of CD40 on their cell surface (Fig. 6b). We reasoned that the majority of residual GC B cells in immunized Hnrnpf bKO mice was from the B cells that escaped Hnrnpf-deletion and underwent expansion, thus having normal CD40 expression. Only a small fraction of GC B cells were the genuine hnRNP F-deficient cells, and hence, had lower CD40 expression. Thus, we sorted the CD40^hi and CD40^lo GC B cells from the immunized Hnrnpf bKO mice and examined the Hnrnpf expression by qRT-PCR. It was corroborated that the CD40^lo GC B had much lower levels of Hnrnpf mRNA than their CD40^hi counterparts and WT GC B cells (Fig. 6c).

**Fig. 6: hnRNP F regulates GC B cell formation by promoting signalling-competent CD40 expression.**

To further evaluate the cell surface expression of Cd40 V2 and V6, the two most dominant variants present in Hnrnpf bKO FOB cells, we transfected HEK 293 T cells with plasmids expressing these two variants. In contrast to the ready detection of CD40 on the 293 T cells transfected with a full-length Cd40 V1, CD40 was barely detected on the cell surface of 293 T cells over-expressing V2 and V6 (Fig. 6d). These results are consistent with the previous study reporting that only Cd40 V1 but not the exon 6-skipped variants, such as V2 and V6, can be translated into a membrane-bound CD40 receptor⁴². We also assessed the biochemical consequence of the decreased cell surface expression of CD40 in the mutant B cells by examining the activation of NF-κB, which was picked up by the gene enrichment analysis (Supplementary Fig. 3b) and is one of the most important signaling pathways downstream of CD40 signaling⁴³. Our immunoblotting showed that hnRNP F-deficient splenic B cells exhibited much reduced NF-κB activation upon stimulation with an agnostic anti-CD40 antibody, as evidenced by their significantly decreased phosphorylation of IKKα/β and their substrate IKBα compared to the WT B cells (Fig. 6e, f). These results also dovetailed with the previous data that hnRNP F-deficient B cells were defective in proliferation upon treatment with anti-CD40 antibody plus IL-4, and IL-21, and failed to upregulate c-Myc upon CD40 engagement (Fig. 4l). Our results suggest that the altered Cd40 AS due to hnRNP F-deficiency leads to significantly reduced cell surface expression of CD40, and consequently compromised CD40-mediated downstream signaling in B cells.

Next, we asked if the compromised GC B cell formation in Hnrnpf bKO mice was due to the defective cell surface expression of CD40 caused by the abnormal Cd40 mRNA AS in the mutant B cells. To this end, we performed a mixed BM chimera experiment by transferring 33.3% of Hnrnpf bKO BM cells transduced retrovirally with V1 or V2 Cd40 variant together with 66.6 % of BM cells from the μMT mice into the sub-lethally irradiated WT mice. The chimeric mice were challenged with NP-CGG 6 weeks after reconstitution, and GC B cells were analyzed 10 days after immunization (Fig. 6g and Supplementary Fig. 4a). Flow cytometric analysis showed that the defective cell surface expression of CD40 in the hnRNP F-deficient B cells was largely restored by Cd40 V1 but not by the V2 variant (Fig. 6h). More importantly, the reintroduction of Cd40 V1 variant into the hnRNP F-deficient B cells could largely rescue their defective GC B cell formation (Fig. 6i). In contrast, the Cd40 V2 variant failed to salvage the GC B cell defect of Hnrnpf bKO mice. Examination of Hnrnpf transcript further showed that the GC B cells differentiated from the Hnrnpf bKO B cells rescued by Cd40 V1 variant had lower levels of Hnrnpf mRNA (~ 50% of the levels in WT), suggesting a considerable fraction of GC B cells in these mice were from the Hnrnpf-deleted B cells, whereas the residual GC B cells from the Hnrnpf bKO cells rescued with the V2 variant had similar Hnrnpf mRNA levels compared to the unrescued control cells (~ 80 %), indicating they were most likely from the deletion-escaped cells (Supplementary Fig. 4b). These results suggest the V1 but not the V2 variant of Cd40 can partially rescue the GC B cell defects caused by hnRNP F-deficiency. Consistently, the abundance of NIP⁺IgG1⁺ cells also increased after the Cd40 V1 variant rescue, although the difference was not statistically significant (Supplementary Fig. 4c, d). Thus, our results indicate that hnRNP F regulates GC B cell formation, at least partially, by ensuring a proper cell surface expression of CD40 via regulating AS of Cd40 pre-mRNA.

HnRNP F directly binds to G-tracts of Cd40 pre-mRNA to modulate exon inclusion and intron removal

We subsequently explored the molecular mechanisms whereby hnRNP F regulates the AS of Cd40. We performed a native RNA immunoprecipitation (RIP) using a hnRNP F-specific antibody followed by qRT-PCR analysis to examine hnRNP F’s binding of Cd40 pre-mRNA in FOB cells. First, a pair of primers specific to Cd40 exon 6 was used to test the pull-down efficiency of the anti-hnRNP F antibody. A significantly higher amount of Cd40 mRNA was detected in the RNA precipitated by the anti-hnRNP F antibody than that pulled down by a control IgG antibody (Fig. 7a), suggesting that the antibody can specifically enrich hnRNP F-associated Cd40 mRNA in B cells.

Next, we set to identify hnRNP F’s binding sites within the Cd40 locus critical for the AS of Cd40. Previous studies have demonstrated that hnRNP F can bind to the G-rich motif within G-tracts in the vicinity of target exons, thus regulating their inclusion or skipping^9,36. By analyzing introns 6 and 7 sequences adjacent to exons 6 and 7, we found many G-tracts adjacent to the 5’ splicing sites of introns 6 and 7, implying that they might be the potential hnRNP F’s binding sites and required for the AS of Cd40. To validate if these putative sites are essential for the normal AS of Cd40, we first performed a minigene-based splicing assay. We constructed a WT Cd40 minigene containing partial sequences of Cd40 introns 5 and 6, with most of their central regions deleted, plus the intact introns 7 and 8 and exons 1 to 9 of Cd40 (Fig. 7b). We also made several mutant Cd40 minigenes with G mutated into A in the G-tract regions 1, 2, and 3 of intron 6 and region 1 of intron 7, namely I6-M1, I6-M2, I6-M3 and I7-M1 (Fig. 7b, shown in the boxes). The mouse 3T3-L1 cells were then transfected with the WT or various mutant minigenes, and RNAs were extracted after 48 h and subjected to qRT-PCR analysis using primers P1 and P2. We found that cells transfected with I6-M1 or I6-M2 minigene expressed considerably higher levels of Cd40 V2 transcript and increased V2 to V1 ratios (Fig. 7c, lanes 3 and 4). These results were in great contrast to that of cells transfected with I6-WT minigene, which predominantly expressed Cd40 V1 transcript and very little Cd40 V2 transcript and thus had a low V2 to V1 ratio (Fig. 7c, lane 2). Conversely, levels of Cd40 V2 transcript and V2 to V1 ratios were indistinguishable between I6-WT and I6-M3 transfectants (Fig. 7c, lanes 2 and 5). These results suggest that the G-tracts in regions 1 and 2 but not 3 of intron 6 are important for exon 6 inclusion. We also noticed that I7-M1 transfectants manifested a substantially enhanced level of Cd40 V3 transcript and a much higher V3 to V1 ratio than I7-WT transfectants (1.12 vs 0.16) (Fig. 7d), indicating that G-tracts in this region of intron 7 are critical for preventing I7 retention. Together, our results suggest that G-tracts within intron 6 regions 1 and 2 promote exon 6 inclusion, whereas the G-tracts in intron 7 region 1 facilitate its removal.

We next asked if the above-verified G-tracts are the bona fide binding sites for hnRNP F. To this end, we performed an RNA pull-down assay using biotin-tagged synthesized WT (I6-WT or I7-WT) or mutant RNAs covering regions 1 and 2 of intron 6 (I6-M1M2) or region 1 of intron 7 (I7-M1), respectively (Fig. 7e, g), followed by immunoblotting to detect the hnRNP F protein pulled down by the RNA. We noticed that hnRNP F protein was readily detected in the lysates pulled down by the biotinylated I6-WT or I7-WT RNAs but not by their reverse sequences (I6-reverse and I7-reverse) (Fig. 7f, h, lanes 2 and 4). However, the amount of hnRNP F protein was much lower in the lysate pulled down by the I6-M1M2 or I7-M1 RNA sequences (Fig. 7f, h, lane 3), suggesting that these G-tract regions are the authentic hnRNP F-binding sites.

To further confirm if hnRNP F could directly bind to these G-tract RNA sequences of Cd40, we performed an RNA electrophoretic mobility shift assay (EMSA). To this end, His-tagged hnRNP F protein was used to determine the interaction between protein and Cd40 I6 G-tract RNA. The addition of hnRNP F caused a pronounced retarded migration of the I6-WT-biotin RNA (Fig. 7i, lane 2), compared to the fast migration of the RNA-alone sample (Fig. 7i, lane 1). The retarded migration of I6-WT-biotin RNA by hnRNP F could be partially reversed by the addition of unlabeled I6-WT RNA, which could compete for binding with hnRNP F (Fig. 7i, lane 3) but not by the unlabeled mutant form of I6 RNA, I6-M1M2 (Fig. 7i, lane 4). These results substantiate that hnRNP F directly binds to Cd40 I6 at regions 1 and 2. We further performed a similar assay to examine the binding of hnRNP F to the Cd40 I7 region and corroborated that hnRNP F also directly binds Cd40 I7 at region 1 (Fig. 7j).

HnRNP A1/A2B1 and hnRNP F antagonistically regulate the AS of Cd40 RNA

More than half of the AS events are regulated by complexes of multiple hnRNP proteins, acting either synergistically or antagonistically^38,44. Having demonstrated the critical role of hnRNP F in Cd40 AS, we next asked if other hnRNPs also regulate the AS of Cd40 given that CD40 is crucial for B cell physiology. To this end, we performed an RNA pull-down assay followed by protein mass spectrometry to identify the potential hnRNP proteins involved in regulating the AS of Cd40. A sense RNA of 404 nucleotides in length, including the exon 6 and the adjacent region of intron 6 sequences of mouse Cd40 pre-mRNA, was transcribed in vitro by the T7 RNA polymerase and labeled with biotin. An antisense RNA of the same region was prepared and served as a control. Mass spectrometric analysis of proteins pulled down by the sense and antisense RNA identified nine proteins with fold-change (sense vs antisense pull-down) of label-free quantification (LFQ) more than 5. Six of the nine proteins are hnRNP proteins, including hnRNP H1, hnRNP H2, hnRNP H3, hnRNP F, hnRNP A1, and hnRNP A2B1(Fig. 8a). Among the six hnRNP proteins, hnRNP H1, hnRNP H2, hnRNP H3, and hnRNP F belong to the same hnRNP F/H subfamily and contain the q-RRMs for recognizing G-tracts and regulating AS^9,10,45. Our EMSA demonstrated that hnRNP H1 could bind to the same sites of Cd40 pre-mRNA as hnRNP F did (Supplementary Fig. 5a, b). Moreover, our pull-down assay also identified CELF2, a homolog of CELF1 that was previously shown to promote Cd40 exon 6 inclusion in human B lymphoma cell line Daudi^27,46, to bind the same region of Cd40 pre-mRNA, further proving the fidelity of our assay (Fig. 8a).

**Fig. 8: hnRNP A1 and A2B1 bind directly to *Cd40* pre-mRNA and regulate its AS.**

Interestingly, the most abundant protein identified in the pull-down experiment was hnRNP A2B1, with its homolog A1 also on the list. HnRNP A1 and A2B1 are members of hnRNPs A/B subfamily and can bind G/A-rich sequences to regulate AS of RNAs^38,47,48. Consistently, our immunoblotting assay detected the presence of hnRNP A1 and A2B1 in a sense but not antisense RNA pull-down lysates (Fig. 8b), corroborating that these two hnRNP proteins bind to Cd40 pre-mRNA sequences adjacent to exon 6 as hnRNP F does. To further explore the role of hnRNP A1 and A2B1 in regulating AS of Cd40, we retrovirally overexpressed Hnrnpa1 and Hnrnpa2b1, respectively, in mouse splenic B cells. Our RT-PCR analysis showed that hnRNP A1 overexpression led to significantly increased levels of Cd40 V2 transcript compared to the control group (Fig. 8c). These results suggest that hnRNP A1 promotes Cd40 RNA exon 6 skipping, consistent with previous data that hnRNPA1 mainly promotes exon skipping^38,44,47,49.

We also conditionally ablated hnRNP A1 and hnRNP A2B1 knockout mice by deleting the floxed Hnrnpa1 and Hnrnpa2b1 alleles using Mb1-Cre recombinase to study their physiological functions. As functional redundancy might exist due to the high homology between hnRNP A1 and hnRNP A2B1 proteins, we further generated conditional double knockout (DKO) mice with both Hnrnpa1 and Hnrnpa2b1 genes deleted specifically in B cells. Immunoblotting assay confirmed that the hnRNP A1 or hnRNP A2B1 proteins were absent in splenic B cells from the individual single and DKO mice (Fig. 8d). Analysis of Cd40 AS showed that the deletion of either Hnrnpa1 or Hnrnpa2b1 could only marginally inhibit exon 6 skipping, as evidenced by the slightly reduced Cd40 V2 transcript in the single KO compared to the WT B cells (Fig. 8e, lanes 2 and 3). In contrast, the DKO B cells manifested a significant reduction in Cd40 V2 transcript (Fig. 8e, lane 4). We hypothesized that the decrease in exon 6 skipping would, in turn, increase the Cd40 V1 transcript leading to an enhanced cell surface expression of CD40. Indeed, we detected a significantly increased CD40 expression on the cell surface of the DKO B cells (Fig. 8f, g).

The opposite phenotypes of CD40 cell surface expression observed in B cells deficient in hnRNP F or A1/A2B1 prompted us to delve into how Cd40 AS were regulated by these closed but different hnRNP family members. Given that both hnRNP F and A1/A2B1 bind Cd40 pre-mRNA near its exon 6 region, preferably at the G-tracts, we asked if they compete for the same or overlapping binding sites and exert opposing roles. We first analyzed the Cd40 pre-mRNA sequences around the exon 6 region to identify the putative hnRNP F- and A1/A2B1-binding sites using the RBPmap (https://rbpmap.technion.ac.il/index.html). We noticed that hnRNP F-binding sites in region 2 of intron 6 overlapped with A1/A2B1-binding sites (Fig. 8h). To prove further that hnRNP F and A1/A2B1 bind Cd40 pre-mRNA at these overlapping sites, we performed an RNA pull-down assay using biotin-labeled WT and mutant RNAs, with the RNA of reverse sequence (I6-R-biotin) serving as the negative control. One of the mutants was I6-M1M2-biotin RNA, which was used to study hnRNP F’s binding previously (Fig. 7e). We also synthesized another mutant, I6-CM-biotin RNA, in which G within the WT binding region was mutated into C (Fig. 8i). This was necessary because the predicted RNA binding motifs for hnRNP F and A2/A2B1 were slightly different, with the latter being G/A-riched sequences³⁸, and therefore G to A mutations might not be able to abolish the binding. The RNA pull-down assays showed that both hnRNP F/H and A1/A2B1 could bind to I6-WT RNA but not the negative control I6-R-biotin (Fig. 8j, lanes 2 and 5). Interestingly, the binding of hnRNP F/H, but not A1/A2B1, was significantly reduced when we used mutant RNA I6-M1M2-biotin, in which G was changed to A (Fig. 8k, lane 3). When we used mutant RNA I6-CM-biotin, in which the binding motifs for both hnRNP F and A1/A2B1 were abolished by G to C mutations, to pull down the proteins, we found that neither hnRNP F/H nor A1/A2B1 could be pulled down (Fig. 8k, lane 4). These data suggest that hnRNP A1/A2B1 and hnRNP F/H bind to the overlapping sites in the region 2 of intron 6 of Cd40 pre-mRNA and antagonistically regulate Cd40 AS, with hnRNP A1/A2B1 promoting its exon 6-skipping and thus downmodulating CD40 cell surface expression.

Discussion

An early study showed that hnRNP F could suppress secretory immunoglobulin expression but favour membrane-bound BCR expression in mouse B cell lines by regulating APA of IgH mRNA¹⁴. However, our study found a normal expression of membrane-bound BCR in hnRNP F-deficient B cells in mice. This discrepancy could be due to different regulatory mechanisms existing in native B cells versus transformed B cell lines. Additionally, hnRNP H, which also regulate APA of IgH mRNA¹⁴, could potentially compensate for hnRNP F’s role for APA in the mutant B cells. While B cell development was also largely normal in hnRNP F-deficient mice, the TD antibody response was profoundly compromised due to the lack of hnRNP F in B cells. The mutant mice failed to produce IgH class-switched high-affinity antibodies in response to TD antigen challenge. Concomitantly, hnRNP F-deficient mice had severely compromised GC B cell formation. Our mechanistic studies demonstrated that the mutant B cells exhibited an abnormal AS of Cd40 mRNA leading to a significantly decreased signaling-competent CD40 expression on the cell surface of B cells. This was supported by the defective proliferation and c-Myc induction in the mutant B cells. Moreover, only the signaling-competent V1 variant, but not the incompetent V2 variant, could partially rescue the CD40 expression and GC B cell defects. However, not all the GC B cells in the chimera mice were from the Hnrnpf-deleted B cells rescued with V1. We hypothesize that these rescued cells could not compete with the residual V1-transfected hnRNP F deletion-escaped B cells for expansion, or some other CD40-independent mechanisms may exist downstream of hnRNP F accounting for the incomplete rescue by V1. The defective IgG1 class switching of hnRNP F-deficient B cells induced by LPS and IL-4 also implies that hnRNP F possibly regulates IgG1 class switching beyond controlling CD40 expression. Further investigations are needed to determine if hnRNP F could control GC B cell response by additional mechanisms.

Five Cd40 mRNA transcript variants can be generated in mouse B cells through AS⁴². Amongst the five variants, only variant 1 encodes a full-length signaling-competent membrane-bound CD40. In contrast, variant 2 has membrane domain-encoding exon 6 skipped, resulting in a truncated secretory form of CD40. The other three variants are also incompetent in transducing CD40 signals, as they either lack intracellular signaling domains of CD40 (variants 3 and 4) or the transmembrane domain (variant 5)⁴². Thus, the aberrant AS of Cd40 mRNA would lead to impaired CD40 signalling. A recent study reported that CELF1 could suppress Cd40 exon 6 skipping in human B lymphoma cell line Daudi²⁷. Here we showed that hnRNP F could regulate AS of Cd40 mRNA in mouse B cells and thus control TD antibody response in vivo. These findings highlight the importance of hnRNP F in controlling AS of Cd40 pre-mRNA in B cells under physiological settings.

Despite the normal B cell development phenotype of hnRNP F-deficient mice, we can not rule out the possibility that hnRNP F plays a role in B cell development. Firstly, CD19-Cre-mediated deletion of floxed Hnrnpf alleles was inefficient in the Fr. B to D subsets (Supplementary Fig. 1d), making it difficult to determine hnRNP F’s function in early B development using Hnrnpf bKO mice. Secondly, hnRNP H, the homologous proteins of hnRNP F, could compensate for hnRNP F’s deficiency in B cell lineage, making it challenging to explicitly elucidate hnRNP F’s role in early B cell development. HnRNP H1 and hnRNP H2 are 96% identical at the amino acid level, and the amino acid identicalness between hnRNP F and hnRNP H1 is 68%, with that in the third q-RRM being up to 80%⁵⁰.In fact, homologous hnRNP members commonly exist in the same cells and exert overlapping functions. For example, PTBP2 and PTBP3 were upregulated and compensated for the PTBP1 deficit in B cell development in PTBP1-deficient mice⁵¹. However, PTBP1 conditional knockout mice exhibited a pronounced defect in the positive selection of GC B cells, and PTBP2 could only partially offset the loss of PTBP1. Similarly, our hnRNP F conditional knockout mice also showed very severe defects in GC B cell formation and antibody response, and apparently, the hnRNP H proteins could not compensate for the hnRNP F’s deficiency. These findings indicate that individual hnRNP proteins might play more dominant roles in certain stages or context.

HnRNP proteins primarily regulate AS by binding to a sequence near the target exon³⁸. Consistent with this, we found that hnRNP F was also bound to G-tracts adjacent to the 5’ splicing site of the intron 6 of Cd40 pre-mRNA to promote the exon 6 inclusion. It was reported previously that hnRNP F regulated AS by binding to pre-mRNA G-quadruplex (G4) structures³⁷. Therefore, it will be interesting to identify the putative G4 structures within the whole Cd40 pre-mRNA fragment using the G4 analysis tool QGRS Mapper⁵², and to determine if these G4 structures are necessary for hnRNP F-mediated regulation of Cd40 AS. Earlier studies also demonstrated that hnRNP F could recruit components of splicing factors U1 or U2 to promote the target exon splicing^12,13. We speculate that hnRNP F might promote intron 6 removal and exon 6 inclusion by recruiting the U1 splicing factors because hnRNP F could bind to the region adjacent to the 5’ splicing site of intron 6. Furthermore, G-tracts were not found within exon 6 and at the 3’ splicing site of intron 6. The same model might also apply to explain the removal of Cd40 intron 7, as hnRNP F could bind to the G-tracts adjacent to its 5’ splicing site as well.

Another interesting finding from our study was that hnRNP A1 and A2B1 could suppress Cd40 exon 6 skipping. It is known that AS are mainly regulated by the hnRNP protein complexes, and some combinations of hnRNPs act synergistically while others exert antagonistic functions. Earlier studies have shown that hnRNP A1 and hnRNP F can regulate AS of the same pre-mRNA but oppositely^38,44. Here we showed that hnRNP A1/A2B1 and hnRNP F also had opposing roles in regulating the AS of Cd40, with hnRNP A1/A2B1 suppressing and hnRNP F promoting exon 6 inclusion. Interestingly, both hnRNP A1/A2B1 and hnRNP F have been found to bind to G-tracts for AS regulation^9,38, suggesting that hnRNP A1 and A2B1 may compete with hnRNP F for the same binding site on Cd40 pre-mRNA. Without the competition of hnRNP F, as in the case of hnRNP F-deficient B cells, hnRNP A1 and A2B1 could bind even more readily to the 5’ splice site of Cd40 pre-mRNA intron 6, thus promoting exon 6 skipping. Therefore, the availability of different Cd40 variants in B cells is determined by the balanced effect of the competitive and co-regulatory hnRNP proteins. Future study of hnRNP A1/A2B1 DKO mice and mice deficient in all three hnRNPs might shed more light on how these molecules regulate Cd40 AS in B cells.

In summary, our study showed that hnRNP F is required to produce high affinity IgH class-switched antibodies during TD antibody response and promote GC B cell formation. Mechanistically, hnRNP F controlled the appropriate cell surface expression of CD40 in B cells by regulating the AS of Cd40 pre-mRNA. Furthermore, we discovered that hnRNP A1 and A2B1 are also involved in regulating AS of Cd40 but have an opposite effect to hnRNP F, suggesting they work in concert to regulate AS of Cd40 to ensure optimal GC B cell formation and antibody response.

Methods

Mice and immunization

All experimental animal operations were approved by the Institutional Animal Care and Use Committee of the Southern University of Science and Technology. Cd19^Cre/+ (B6.129P2(C)-Cd19^tm1(cre)Cgn/J, strain #: 007770) and Mb1^Cre/+ (B6.C(Cg)-Cd79a^tm1(cre)Reth/EhobJ, strain #: 020505) mice were obtained from The Jackson Laboratory. Hnrnpf floxed mice were obtained from Nanjing Biomedical Research Institute of Nanjing University (Cat #: T000629). To generate conditional knockout mice, the Hnrnpf floxed mice were crossed to Flp expressing mice (Cyagen, Cat #: C001090) to remove the neo-cassette and then crossed to Cd19^Cre/+ mice to delete exon 4 of Hnrnpf gene (Supplementary Fig. 6a). Hnrnpa2b1 floxed mice (Cat #: S-CKO-11602) were obtained from Cyagen and crossed with Mb1^Cre/+ to generate conditional knockout mice to delete its exons 2 to 6 (Supplementary Fig. 6b). Hnrnpa1 floxed mice (Cat #: T013507) were obtained from Gempharmatech and crossed with Mb1^Cre/+ to generate conditional knockout mice to delete its exons 2 to 11 (Supplementary Fig. 6c). Male and female mice of 6–12 weeks old were used for breeding and experiments. For chimera assay, muMT (B6.129S2-Ighm^tm1Cgn/J, strain #: 002288) mice were obtained from The Jackson Laboratory and the recipients were wild-type C57BL/6 J mice (The Jackson Laboratory, strain #: 000664). All mice were on a C57BL/6 background and feeded with SPF grade Rats & Mice Growing and Breeding feed (Beijing Keao Xieli Feed Co.,Ltd., Cat. 1016706476803973120). All mice were housed under specific pathogen-free conditions and on a 12 h on/12 h off light/dark cycle with temperature 20–26 °C and humidity ~70% in the Laboratory Animal Center of the Southern University of Science and Technology. For experiments, the mice were euthanized in a CO₂ chamber.

For T-cell dependent antibody immune response, mice were immunized with 100 μg NP₃₀-CGG or NP₃₀-KLH (Biosearch Technologies) in alum (Thermo) intraperitoneally.

Cell culture and B cell stimulation in vitro

3T3-L1 fibroblast cells were a gift from the Life Science Experimental Teaching Demonstration Center of Southern University of Science and Technology, and were cultured with DMEM high glucose liquid media (Hyclone) containing 10% fetal bovine serum (ExCell Bio), 100 U/ml penicillin and 100 μg/ml streptomycin (Hyclone) in 24-well plates or 100 mm dishes (NEST Biotechnology) at 37°C with 5% (v/v) CO2. HEK 293 T cells were cultured under the same conditions as 3T3-L1 fibroblast cells.

For B cell stimulation in vitro, anti-CD43 beads (Miltenyi Biotec) and LS columns (Miltenyi Biotec) were used for the purification of splenic B cells. Two million purified splenic B cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum (Gibco), 100 U/ml penicillin and 100 μg/ml streptomycin (Hyclone). LPS (20 µg/ml, Sigma) was used to stimulate B cells for retroviral transduction. Anti-CD40 (1 µg/ml, BD Biosciences) antibody was used for immunoblot analysis of NF-κB pathway in splenic B cells upon CD40 engagement.

CSR and cell proliferation in vitro

Mouse splenic B cells were isolated using EasySep™ Mouse B Cell Isolation Kit (Stem Cell) and stimulated with 20 μg/ml LPS (Sigma-Aldrich) for CSR to IgG3. For CSR to IgG1, B cells were stimulated with 20 μg/ml LPS (Sigma-Aldrich) and 10 ng/ml of mouse recombinant IL-4 (Bio-Techne, # 404-ML), or 1 μg/ml anti-CD40 (BD Biosciences) and 10 ng/ml of mouse recombinant IL-4 (Bio-Techne, # 404-ML), or 1 μg/ml anti-CD40 (BD Biosciences), 10 ng/ml of mouse recombinant IL-4 (Bio-Techne, # 404-ML) and 10 ng/ml of mouse recombinant IL-21 (Bio-Techne, # 594-ML).

For the proliferation assay, cells were stained with CellTrace™ Far Red Cell Proliferation Kit (Thermofisher) for 5 min at 37 °C and cultured with indicated stimulation. Cells were analyzed at day 0 and day 3 using flow cytometry.

iGB cell culture

Splenic B cells were purified using EasySep™ Mouse B Cell Isolation Kit (Stem Cell) and seeded with 10⁶ cells per well of 96-well plate in 200 ul RPMI-1640 medium (Gibico) supplemented with 15 % FBS (ExCell Bio), 55 μM 2-mercaptoethanol (Sigma), 1 X Non-Essential Amino Acids (Gbico), 10 mM HEPES (Gibico), 1 mM sodium pyruvate, 100 U/ml penicillin and 100 μg/ml streptomycin (Hyclone). B cells were stimulated with 1 μg/ml of anti-CD40 (BD Biosciences), 5 ng/ml of mouse recombinant IL-4 (Bio-Techne, # 404-ML), and 5 ng/ml of mouse recombinant IL-21 (Bio-Techne, # 594-ML) for primary culture. After 1 day, Cells from each well of 96-well plate were transferred to 10 cm plates containing 10 ml medium and then cultured with the same stimulation for another 3 days. iGB cells were analyzed at day 4 using flow cytometry.

Plasmid construction

All plasmids used in this study were constructed in a MIGR1-based vector as described previously⁵³. For the construction of MIGR1-CD40-V1, MIGR1-CD40-V2, MIGR1-CD40-V6 and MIGR1-Hnrnpa1, target DNAs were amplified using KOD-Plus-neo (TOYOBO) from cDNA library of mouse splenic B cells and then inserted into the MIGR1 vector using ClonExpress MultiS one Step Cloning Kit (Vazyme). For MIGR1-CD40-minigene construction, we used a minigene composed of mouse CD40 coding sequence and introns 5 (excluding the region spanning 5′-CATGTGCATGCATGC…ATCTGATCATCTTTT-3′), 6 (excluding the regions spanning 5′- CAGTGTCACCTGGTA…GCACAGATGCACACA-3′), 7 and 8 (Fig. 6b). Sanger sequencing was used for the confirmation of all inserted sequences.

Transfection and retrovirus transduction

3T3-L1 cells were transiently transfected with WT or mutant MIGR1-CD40-minigene plasmids coated by Liposome Transfection Reagent (Yeasen) in OPTI-MEM (Gibco), cultured for 48 h, and then collected for following experiments. For the transfection in 24-well plates, 8 × 10⁴ 3T3-L1 cells were seeded, and 1 μg of plasmids mixed with 2 uL of transfection reagent was added for each well. HEK 293 T cells were transiently transfected with MIGR1-CD40-V1, -V2 or/-V6 using the same reagents as 3T3-L1 cells.

For retrovirus transduction, HEK 293 T cells were co-transfected with equal amounts of MIGR1 or MIGR1-based plasmids and pCL-ECO (retrovirus packaging vector) plasmids using Lipofectamine™ 3000 Transfection Reagent (Thermo). The virus supernatant was collected at 48 h and 72 h after transfection and added to target cells in a fresh culture medium with 6 μg/ml polybrene (Sigma) and cytokines or stimulators. Then target cells were centrifuged at 600 × g at 37 °C for 1.5 h. After two rounds of transductions, cells were cultured for a further 24 h or 48 h followed by the subsequent analysis or experiments.

Flow cytometry

Single-cell suspensions were prepared from mouse spleens or BM by removal of red blood cells using RBC lysis buffer (BioLegend) and passing through 70-µm nylon mesh. Cells were then stained with indicated antibodies in PBS containing 5% (v/v) FBS at 4 °C. NP-specific antibodies were detected by NIP-BSA (Biosearch Technologies) conjugated with PE as described previously⁵⁴. Dead cells were excluded using DAPI staining. The data were collected using CytoFLEX S flow cytometry analyzer with CytExpert software (Beckman Coulter, V2.3). The geometric mean fluorescence intensity was calculated by Flowjo V10.8.1.

For FOB cell sorting, single-cell suspensions were prepared and stained as described above and the cells (B220⁺CD93⁻CD23^highCD21^low) were sorted on a BD FACSAiraII (BD bioscience).

ELISPOT assay

ELISPOT assay was performed as described previously⁵⁵. In brief, MultiScreen filter plates with 0.45 μm PVDF membrane (Milipore) were coated with 1 μg per well NP₂₀-BSA (Biosearch Technologies) in PBS overnight at 4°C and blocked with 2% BSA in PBS for 2 h at room temperature. One million splenocytes or three million BM cells were seeded in each well, and then the plate was incubated at 37°C for 2 h. After washing, the plate was incubated with Biotinylated anti-IgM (1: 1000, SouthernBiotech, 1140-08) or anti-IgG1 (1: 1000, SouthernBiotech, 1144-08) antibodies for 1 h at room temperature, followed by streptavidin-AP (1:1000, SouthernBiotech, 7105-04) incubation for 0.5 h at room temperature. The substrates (MABTECH) were added to each well for visualization of antibody-secreting cells.

ELISA

NP-specific antibodies levels were detected using ELISA as described previously⁵⁶. Briefly, sera were collected before and after NP-CGG immunization on the indicated days. NP₂₀-BSA or NP₂-BSA (Biosearch Technologies) were used for capturing NP-specific antibodies and total immunoglobulins were captured by Goat Anti-Mouse Immunoglobulins (SouthernBiotech, 1010-01). HRP-conjugated anti-IgG1 (1:1000, SouthernBiotech, 1071-05), -IgG2b (1:1000, SouthernBiotech, 1091-05), -IgG3 (1:1000, SouthernBiotech, 1101-05), -IgA (1:1000, SouthernBiotech, 1040-05) and -IgM (1:1000, SouthernBiotech, 1021-05) antibodies conjugated with HRP were used as secondary antibodies, followed by TMB substrate (BioLegend) development. OD450 values were measured using a microplate reader with Gene5 Software (BioTek, V2.08).

RNA isolation, RT-PCR and qRT-PCR

RT-PCR was performed as described previously⁴². Briefly, total RNA was purified using RNAiso Plus (Takara) from measured cells. cDNA was reversely synthesized with RT SuperMix (Vazyme) and amplified with appropriate primers using 2× Taq Master Mix (Vazyme), and the PCR products were analyzed through agarose gel electrophoresis. Image J (V1.53e) was used for RT-PCR gel gray scale analysis. RT-PCR primers used for AS of Cd40 are as follows: P1, 5′-CTGCCCAGTCGGCTTCTTCTC-3′; P2, 5′-CCTGTGTGACAGGCTGACAC-3′; P3, 5′-GTTTAAAGTCCCGGATGCGA-3′; P4, 5′-CTCATTATCCTTTGGTTTCTTGACC-3′. Primers for semiquantitative analysis: Actin forward, 5′-CGTGAAAAGATGACCCAGATCA-3′; Actin reverse, 5′-CACAGCCTGGATGGCTACGT-3′; Cd40 forward, 5′-GCAGTGTGTTACGTGCAGTG-3′; Cd40 reverse, 5′-TGTGCAGTGGCTTGTCAGTC-3′; Hnrnpa1 forward, 5′-AACCGACGAGAGTCTGAGGA-3′; Hnrnpa1 reverse, 5′-CAAACCCAAAGCCCCTGGAT-3′.

For qRT-PCR, RNA extraction and cDNA synthesis were performed according to the above mentioned methods. Quantitative real-time PCR was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme). The primer sequences for qRT-PCR used in this study were as follows: Myc forward, 5′-GTGCTGCATGAGGAGACACC-3′, Myc reverse, 5′-GACCTCTTGGCAGGGGTTTG-3′; Hnrnpf forward, 5′- AGAAGGCATCTGTGGTGGTT-3′, Hnrnpf reverse, 5′- GAAATGAACACCTGCGACCC −3′.

Immunoblotting

Immunoblotting was performed as described previously⁵⁷. Briefly, whole-cell lysates were prepared using RIPA buffer (Sigma-Aldrich) containing protease inhibitor cocktail (Roche). Proteins were separated by running on 10% SDS-PAGE gels and transferred to PVDF membranes. After blocking with 5% skim milk in TBST, membranes were probed with appropriate primary antibodies overnight and further incubated with goat anti-rabbit IgG-HRP (Proteintech). Blots were developed using Western chemiluminescent HRP substrate (Millipore) and visualized using an imaging system (Tanon, 5200). The primary antibodies used in immunoblotting are as follows: β-actin (1:2000, Proteintech, 20536-1-AP), hnRNP F (1:1000, Thermo, PA522341), hnRNPA1 (1:1000, Cell Signaling Technology, #8443), hnRNPA2B1 (1:1000, ABclonal, A1162), p-IKKα/IKKβ (1:1000, Cell Signaling Technology, #2078), IKKβ (1:1000, Cell Signaling Technology, #8943), p-IKBα (1:1000, Cell Signaling Technology, #2859) and IKBα (1:1000, Cell Signaling Technology, #9242).

Chimera assay

Chimera assay was performed as described previously⁵⁸. In brief, donor mice were intraperitoneally injected with 0.15 mg/g 5-FU (Sigma) according to their body weight for 3 days. BM cells were then collected and cultured in a fresh medium containing 20 ng/ml of IL-3 (Bio-Techne, # 403-ML), 50 ng/ml of IL-6 (Bio-Techne, # 406-ML) and 50 ng/ml of SCF (Bio-Techne, # 455-MC). After 24 h, cells were collected with a cell scraper (NEST Biotechnology) and transduced by retrovirus using the method described above. The cells were then injected intravenously into irradiated recipient mice (900 rads) to generate chimera mice.

RIP-qPCR

Native RNA immunoprecipitation (RIP) was performed as described previously⁵⁹. Briefly, mouse splenic B cells were lysed and centrifuged to isolate the supernatant cell lysate. After taking 10 % of the supernatant as input, the cell lysate (100% to input) was incubated with immobilized hnRNP F antibody (Thermo, PA522341) to Protein G Magnetic Beads (Cell Signaling Technology, #9006) overnight at 4 °C. After extensive washing, immunoprecipitation complexes were digested using Proteinase K (Beyotime) and processed for RNA purification using RNAiso Plus (Takara). IP and input RNAs were reversely transcribed with RT SuperMix (Vazyme) to produce cDNA. Quantitative real-time PCR was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme) using primers CD40-E6-F (5′-CTGTGAGGATAAGAACTTGGAG-3′) and CD40-E6-R (5′-CACAGATGACATTAGTCTGACT-3′) to amplify Cd40 RNA. Fold enrichment was calculated as described previously⁵⁹.

RNA pull-down and mass spectrometry

Biotin-labelled RNAs were synthesized by GENEWIZ or transcribed using T7 RNA polymerase (Beyotime) combined with Biotin RNA Labeling Mix (Sigma) in vitro. RNA pull-down assay was performed according to a previous report with some modifications³⁷. Two μg biotinylated RNA in vitro or 250 pmol biotinylated RNA oligos were immobilized to 30 μl Streptavidin Magnetic Beads (MedChemExpress) according to the manufacturer’s instruction. Then RNA-beads complexes were incubated with 100 μg nuclear extracts from mouse splenic B cells in 250 μl pull-down buffer (20 mM Tris-HCl pH = 7.5, 200 mM NaCl, 6 mM EDTA, 5 mM potassium fluoride, 5 mM β-Glycerophosphate, 2 mg/mL protease inhibitor cocktail) for 4 h at 4 °C. The beads were washed five times with pull-down buffer and then dissolved in western blot loading buffer for loading onto 10% SDS-PAGE gels.

For immunoblotting, proteins in gels were transferred to PVDF membranes for further detection. For protein mass spectrometry analysis, proteins in gels were sliced for label-free protein quantification. MaxQuant software (version 1.6.10.43) was used for MS/MS analysis using Mus musculus database (55470 sequences) with the followed parameters: oxidation of methionine protein N-terminal acetylation, variable modifications; carbamidomethylation, fixed modifications; digestion enzyme, trypsin/P; a maximum of peptide charge, 6 and 2 missed cleavages/peptide allowed; mass tolerance for precursor ions, 20 p.p.m for the first search and 7 p.p.m for the main search; FDRs of peptide and protein, 1%.

EMSA

The mouse recombinant hnRNP F and hnRNP H1 proteins were purified as previously described³⁷. In brief, hnRNP F and H1 proteins were expressed in E. coli Rosetta induced by 0.5 mM IPTG, and purified from supernatant using BeaverBeads™ His-tag Protein Purification beads (Beaver) after sonication. Binding of Cd40 RNA and hnRNP F/H was detected using the BersinBio^TM RNA-EMSA Kit (BersinBio) according to the operating manual. For each reaction, 12.5 nM biotin-labeled RNAs were incubated with 200 ng purified protein.

RNA-seq and data processing

Total RNAs were exacted from FACS-sorted FOB cells from WT and Hnrnpf KO mice using RNeasy Mini Kit (Qiagen). After quantification and poly(A) mRNA isolation, the mRNA was fragmented and subjected to library preparation using NEBNext Ultra RNA Library Prep Kit for Illumina. The library DNA was loaded on an Illumina NovaSeq 6000 instrument 150 bp paired-end (PE150) configuration carried out by GENEWIZ. The software FastQC (v0.10.1) was used to detect the sequencing quality of the raw data, and Cutadapt (v1.9.1) was used to filter the raw data to get clean reads. Clean reads were aligned to GRCm38 (Ensembl) using Hisat2 (v2.0.1), and gene expression levels were estimated by HTSeq (v0.6.1). Genes with differential expression were obtained using DESeq2 (v1.6.3) Bioconductor package. KEGG analysis of differentially expressed (fold change ≥ 2, adjusted p ≤ 0 .05) genes was performed using R (v3.6.3) package ClusterProfiler (v3.14.3). rMATS turbo (v4.1.0) was used for analyzing differential AS events with mapped BAM files from above and with parameter: -variable-read-length.

Sequencing data visualization and statistical analysis

The volcano plot of sequencing data was obtained using the OmicStudio tools at https://www.omicstudio.cn/tool. Sashimi plot of differential exons junction in WT or Hnrnpf KO B cells was generated using the IGV (http://www.igv.org/). Statistical analysis was performed using GraphPad Prism 7, and the significance tests we used were indicated in the figure legends. P value <0.05 was considered as the threshold for statistical significance.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

RNA-Seq data of Hnrnpf expression in B cell subsets were obtained from the Immunological Genome Project (ImmGen) database (https://www.immgen.org/). RNA-Seq data (including raw data, gene expression profiling, and changed alternative splicing events) produced in this study have been deposited in the GEO database (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE188538) with the accession number GSE188538. Data for LC/MS/MS of CD40 mRNA pull-down have been deposited to Figshare (https://doi.org/10.6084/m9.figshare.19750645.v1). Source data are provided with this paper.

References

Mortimer, S. A., Kidwell, M. A. & Doudna, J. A. Insights into RNA structure and function from genome-wide studies. Nat. Rev. Genet. 15, 469–479 (2014).
Article CAS PubMed Google Scholar
Piñol-Roma, S. & Dreyfuss, G. hnRNP proteins: localization and transport between the nucleus and the cytoplasm. Trends Cell Biol. 3, 151–155 (1993).
Article PubMed Google Scholar
Geuens, T., Bouhy, D. & Timmerman, V. The hnRNP family: insights into their role in health and disease. Hum. Genet 135, 851–867 (2016).
Article CAS PubMed PubMed Central Google Scholar
Bent, H. et al. Heterogeneous nuclear ribonucleoproteins h, h′, and f are members of a ubiquitously expressed subfamily of related but distinct proteins encoded by genes mapping to different chromosomes*. J. Biol. Chem. 270, 28780–28789 (1995).
Article Google Scholar
Min, H. S., Chan, R. C. & Black, D. L. The generally expressed hnRNP F is involved in a neural-specific pre-mRNA splicing event. Genes Dev. 9, 2659–2671 (1995).
Article CAS PubMed Google Scholar
Tyson-Capper, A. & Gautrey, H. Regulation of Mcl-1 alternative splicing by hnRNP F, H1 and K in breast cancer cells. RNA Biol. 15, 1448–1457 (2018).
Article PubMed PubMed Central Google Scholar
Garneau, D., Revil, T., Fisette, J. F. & Chabot, B. Heterogeneous nuclear ribonucleoprotein F/H proteins modulate the alternative splicing of the apoptotic mediator Bcl-x. J. Biol. Chem. 280, 22641–22650 (2005).
Article CAS PubMed Google Scholar
Yamazaki, T. et al. TCF3 alternative splicing controlled by hnRNP H/F regulates E-cadherin expression and hESC pluripotency. Genes Dev. 32, 1161–1174 (2018).
Article CAS PubMed PubMed Central Google Scholar
Caputi, M. & Zahler, A. M. Determination of the RNA binding specificity of the heterogeneous nuclear ribonucleoprotein (hnRNP) H/H ‘/F/2H9 family. J. Biol. Chem. 276, 43850–43859 (2001).
Article CAS PubMed Google Scholar
Dominguez, C., Fisette, J. F., Chabot, B. & Allain, F. H. Structural basis of G-tract recognition and engaging by hnRNP F quasi-RRMs. Nat. Struct. Mol. Biol. 17, 853–861 (2010).
Article CAS PubMed Google Scholar
Wang, L. & Yan, F. Molecular insights into the specific recognition between the RNA binding domain qRRM2 of hnRNP F and G-tract RNA: A molecular dynamics study. Biochem. Biophys. Res. Commun. 494, 95–100 (2017).
Article CAS PubMed Google Scholar
Wang, E. & Cambi, F. Heterogeneous nuclear ribonucleoproteins H and F regulate the proteolipid protein/DM20 ratio by recruiting U1 small nuclear ribonucleoprotein through a complex array of G runs. J. Biol. Chem. 284, 11194–11204 (2009).
Article CAS PubMed PubMed Central Google Scholar
Fan, L. et al. Histone demethylase JMJD1A promotes alternative splicing of AR variant 7 (AR-V7) in prostate cancer cells. Proc. Natl Acad. Sci. USA 115, E4584–e4593 (2018).
Article CAS PubMed PubMed Central Google Scholar
Veraldi, K. L. et al. hnRNP F influences binding of a 64-kilodalton subunit of cleavage stimulation factor to mRNA precursors in mouse B cells. Mol. Cell Biol. 21, 1228–1238 (2001).
Article CAS PubMed PubMed Central Google Scholar
Decorsiere, A., Cayrel, A., Vagner, S. & Millevoi, S. Essential role for the interaction between hnRNP H/F and a G quadruplex in maintaining p53 pre-mRNA 3 ‘-end processing and function during DNA damage. Genes Dev. 25, 220–225 (2011).
Article CAS PubMed PubMed Central Google Scholar
Li, F. et al. HnRNP-F regulates EMT in bladder cancer by mediating the stabilization of Snail1 mRNA by binding to its 3’ UTR. EBioMedicine 45, 208–219 (2019).
Article PubMed PubMed Central Google Scholar
Herviou, P. et al. hnRNP H/F drive RNA G-quadruplex-mediated translation linked to genomic instability and therapy resistance in glioblastoma. Nat. Commun. 11, 17 (2020).
Article Google Scholar
Reznik, B., Clement, S. L. & Lykke-Andersen, J. hnRNP F Complexes with Tristetraprolin and Stimulates ARE-mRNA Decay. Plos One 9, e100992 (2014).
Article ADS PubMed PubMed Central Google Scholar
Carsetti, R. The development of B cells in the bone marrow is controlled by the balance between cell-autonomous mechanisms and signals from the microenvironment. J. Exp. Med. 191, 5–8 (2000).
Article CAS PubMed PubMed Central Google Scholar
Suan, D., Sundling, C. & Brink, R. Plasma cell and memory B cell differentiation from the germinal center. Curr. Opin. Immunol. 45, 97–102 (2017).
Article CAS PubMed Google Scholar
Corcoran, L. M. & Tarlinton, D. M. Regulation of germinal center responses, memory B cells and plasma cell formation—an update. Curr. Opin. Immunol. 39, 59–67 (2016).
Article CAS PubMed Google Scholar
Mesin, L., Ersching, J. & Victora, G. D. Germinal center B cell dynamics. Immunity 45, 471–482 (2016).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Y., Garcia-Ibanez, L. & Toellner, K.-M. Regulation of germinal center B-cell differentiation. Immunol. Rev. 270, 8–19 (2016).
Article CAS PubMed PubMed Central Google Scholar
Luo, W., Weisel, F. & Shlomchik, M. J. B. Cell receptor and cd40 signaling are rewired for synergistic induction of the c-myc transcription factor in germinal center b cells. Immunity 48, 313–326.e315 (2018).
Article CAS PubMed PubMed Central Google Scholar
Laman, J. D., Claassen, E. & Noelle, R. J. Functions of CD40 and its ligand, gp39 (CD40L). Crit. Rev. Immunol. 37, 371–420 (2017).
Article PubMed Google Scholar
Elgueta, R. et al. Molecular mechanism and function of CD40/CD40L engagement in the immune system. Immunol. Rev. 229, 152–172 (2009).
Article CAS PubMed Google Scholar
Jiang, C. et al. CRISPR/Cas9 Screens Reveal Multiple Layers of B cell CD40 Regulation. Cell Rep. 28, 1307–1322.e1308 (2019).
Article CAS PubMed PubMed Central Google Scholar
Schmidt-Supprian, M. & Rajewsky, K. Vagaries of conditional gene targeting. Nat. Immunol. 8, 665–668 (2007).
Article CAS PubMed Google Scholar
Xu, S. et al. Mir-17-92 regulates bone marrow homing of plasma cells and production of immunoglobulin G2c. Nat. Commun. 6, 6764 (2015).
Article ADS CAS PubMed Google Scholar
Lau, A. W. Y. & Brink, R. Selection in the germinal center. Curr. Opin. Immunol. 63, 29–34 (2020).
Article CAS PubMed Google Scholar
Robinson, M. J., Pitt, C., Brodie, E. J., Valk, A. M. & O’Donnell, K. Nitschke L, et al. BAFF, IL-4 and IL-21 separably program germinal center-like phenotype acquisition, BCL6 expression, proliferation and survival of CD40L-activated B cells in vitro. Immunol. Cell Biol. 97, 826–839 (2019).
Dominguez-Sola, D. et al. The proto-oncogene MYC is required for selection in the germinal center and cyclic reentry. Nat. Immunol. 13, 1083–1091 (2012).
Article CAS PubMed PubMed Central Google Scholar
Calado, D. P. et al. The cell-cycle regulator c-Myc is essential for the formation and maintenance of germinal centers. Nat. Immunol. 13, 1092–1100 (2012).
Article CAS PubMed PubMed Central Google Scholar
Allman, D. & Pillai, S. Peripheral B cell subsets. Curr. Opin. Immunol. 20, 149–157 (2008).
Article CAS PubMed PubMed Central Google Scholar
Oracki, S. A., Walker, J. A., Hibbs, M. L., Corcoran, L. M. & Tarlinton, D. M. Plasma cell development and survival. Immunol. Rev. 237, 140–159 (2010).
Article CAS PubMed Google Scholar
Martinez-Contreras, R. et al. Intronic binding sites for hnRNP A/B and hnRNP F/H proteins stimulate pre-mRNA splicing. Plos Biol. 4, 172–185. (2006).
Article CAS Google Scholar
Huang, H., Zhang, J., Harvey, S. E., Hu, X. & Cheng, C. RNA G-quadruplex secondary structure promotes alternative splicing via the RNA-binding protein hnRNPF. Genes Dev. 31, 2296–2309 (2017).
Article CAS PubMed PubMed Central Google Scholar
Huelga, S. C. et al. Integrative genome-wide analysis reveals cooperative regulation of alternative splicing by hnRNP proteins. Cell Rep. 1, 167–178 (2012).
Article CAS PubMed PubMed Central Google Scholar
Shen, S. & Park, J. W. et al. rMATS: robust and flexible detection of differential alternative splicing from replicate RNA-Seq data. Proc. Natl Acad. Sci. USA 111, E5593–E5601 (2014).
Article CAS PubMed PubMed Central Google Scholar
Du, J., Wang, Q., Ziegler, S. F. & Zhou, B. FOXP3 interacts with hnRNPF to modulate pre-mRNA alternative splicing. J. Biol. Chem. 293, 10235–10244 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kawabe, T. et al. The immune responses in CD40-deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity 1, 167–178 (1994).
Article CAS PubMed Google Scholar
Tone, M., Tone, Y., Fairchild, P. J., Wykes, M. & Waldmann, H. Regulation of CD40 function by its isoforms generated through alternative splicing. Proc. Natl Acad. Sci. USA 98, 1751 (2001).
Article ADS CAS PubMed PubMed Central Google Scholar
Hömig-Hölzel, C. et al. Constitutive CD40 signaling in B cells selectively activates the noncanonical NF-kappaB pathway and promotes lymphomagenesis. J. Exp. Med. 205, 1317–1329 (2008).
Article PubMed PubMed Central Google Scholar
Talukdar, I. et al. hnRNP A1 and hnRNP F Modulate the Alternative Splicing of Exon 11 of the Insulin Receptor Gene. Plos One 6, e27869 (2011).
Article ADS CAS PubMed PubMed Central Google Scholar
Dominguez, C. & Allain, F. H. T. NMR structure of the three quasi RNA recognition motifs (qRRMs) of human hnRNP F and interaction studies with Bcl-x G-tract RNA: a novel mode of RNA recognition. Nucleic acids Res. 34, 3634–3645 (2006).
Article CAS PubMed PubMed Central Google Scholar
Blech-Hermoni, Y., Stillwagon, S. J. & Ladd, A. N. Diversity and conservation of CELF1 and CELF2 RNA and protein expression patterns during embryonic development. Dev. Dyn. 242, 767–777 (2013).
Article CAS PubMed PubMed Central Google Scholar
Hutchison, S., LeBel, C., Blanchette, M. & Chabot, B. Distinct sets of adjacent heterogeneous nuclear ribonucleoprotein (hnRNP) A1/A2 binding sites control 5’ splice site selection in the hnRNP A1 mRNA precursor. J. Biol. Chem. 277, 29745–29752 (2002).
Article CAS PubMed Google Scholar
Alarcón, C. R. et al. HNRNPA2B1 is a mediator of m(6)a-dependent nuclear rna processing events. Cell 162, 1299–1308 (2015).
Article PubMed PubMed Central Google Scholar
Chiou, N.-T., Shankarling, G., Lynch, K. W. & hnRNP, L. and hnRNP A1 induce extended U1 snRNA interactions with an exon to repress spliceosome assembly. Mol. cell 49, 972–982 (2013).
Article CAS PubMed PubMed Central Google Scholar
Mauger, D. M., Lin, C., Garcia-Blanco, M. A. & HnRNP, H. and hnRNP F complex with Fox2 to silence fibroblast growth factor receptor 2 Exon IIIc. Mol. Cell Biol. 28, 5403–5419 (2008).
Article CAS PubMed PubMed Central Google Scholar
Monzón-Casanova, E. et al. The RNA-binding protein PTBP1 is necessary for B cell selection in germinal centers. Nat. Immunol. 19, 267–278 (2018).
Article PubMed PubMed Central Google Scholar
Kikin, O., D’Antonio, L. & Bagga, P. S. QGRS Mapper: a web-based server for predicting G-quadruplexes in nucleotide sequences. Nucleic acids Res. 34, W676–W682 (2006).
Article CAS PubMed PubMed Central Google Scholar
Ruan, G.-X. et al. The spliceosome component Usp39 controls B cell development by regulating immunoglobulin gene rearrangement. Cell Rep. 38, 110338 (2022).
Article CAS PubMed Google Scholar
Li, Y.-F., Xu, S., Ou, X. & Lam, K.-P. Shp1 signalling is required to establish the long-lived bone marrow plasma cell pool. Nat. Commun. 5, 4273–4273 (2014).
Article ADS CAS PubMed Google Scholar
Ou, X., Xu, S. & Lam, K. P. Deficiency in TNFRSF13B (TACI) expands T-follicular helper and germinal center B cells via increased ICOS-ligand expression but impairs plasma cell survival. Proc. Natl Acad. Sci. USA 109, 15401–15406 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Huang, H. et al. Mettl14-Mediated m6A modification is essential for germinal center B cell response. J. Immunol. 208, 1924–1936 (2022).
Article CAS PubMed Google Scholar
Huo, J., Xu, S., Lin, B., Chng, W. J. & Lam, K. P. Fas apoptosis inhibitory molecule is upregulated by IGF-1 signaling and modulates Akt activation and IRF4 expression in multiple myeloma. Leukemia 27, 1165–1171 (2013).
Article CAS PubMed Google Scholar
Ou, X., Xu, S., Li, Y. F. & Lam, K. P. Adaptor protein DOK3 promotes plasma cell differentiation by regulating the expression of programmed cell death 1 ligands. Proc. Natl Acad. Sci. USA 111, 11431–11436 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Gagliardi, M. & Matarazzo, M. R. RIP: RNA immunoprecipitation. Methods Mol. Biol. 1480, 73–86 (2016).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

This work was supported by the Shenzhen Science and Technology Innovation Commission (Grants No. JCYJ20220530115212028 and JCYJ20190809161807432, to X.O.), National Natural Science Foundation of China (Grant No. 32170882, to X.O.), Guangdong Basic and Applied Basic Research Foundation (Grant No. 2020A1515010262, to X.O.), and Singapore Agency for Science Technology and Research (to S.X.). We thank SUSTech Core Research Facilities and Laboratory Animal Research Center for their assistance with animal experiments.

Author information

Authors and Affiliations

School of Life Sciences, Southern University of Science and Technology, Shenzhen, 518055, China
Hengjun Huang, Yuxing Li, Gaopu Zhang, Zhijian Zhu, Wenjing Chen, Rui Zhang, Jing Wang, Yu Ouyang & Xijun Ou
Medical School, Taizhou University, Taizhou, 318000, China
Gui-Xin Ruan
Department of Computer Science and Engineering, College of Engineering, Southern University of Science and Technology, Shenzhen, 518055, China
Jia Zou
Singapore Immunology Network, Agency for Science, Technology and Research, Singapore, 138648, Singapore
Shengli Xu
Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, 117599, Singapore
Shengli Xu

Authors

Hengjun Huang
View author publications
You can also search for this author in PubMed Google Scholar
Yuxing Li
View author publications
You can also search for this author in PubMed Google Scholar
Gaopu Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Gui-Xin Ruan
View author publications
You can also search for this author in PubMed Google Scholar
Zhijian Zhu
View author publications
You can also search for this author in PubMed Google Scholar
Wenjing Chen
View author publications
You can also search for this author in PubMed Google Scholar
Jia Zou
View author publications
You can also search for this author in PubMed Google Scholar
Rui Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Jing Wang
View author publications
You can also search for this author in PubMed Google Scholar
Yu Ouyang
View author publications
You can also search for this author in PubMed Google Scholar
Shengli Xu
View author publications
You can also search for this author in PubMed Google Scholar
Xijun Ou
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

X.O. and S.X. conceived the study. H.H., X.O., and S.X. designed the research and wrote the paper. H.H. performed most of the experiments and data analysis. G.Z. performed CD40-minigene related experiments and plasmid construction. J.Z. carried out the bioinformatics analysis. W.C., Y.L., G.X.R., R.Z., and J.W. performed animal works. Z.Z. performed part of hnRNP A1/A2B1 related experiments, Y.O. performed part of qRT-PCR.

Corresponding authors

Correspondence to Shengli Xu or Xijun Ou.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Stefano Casola, Anselm Enders and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Huang, H., Li, Y., Zhang, G. et al. The RNA-binding protein hnRNP F is required for the germinal center B cell response. Nat Commun 14, 1731 (2023). https://doi.org/10.1038/s41467-023-37308-z

Download citation

Received: 28 April 2022
Accepted: 10 March 2023
Published: 30 March 2023
DOI: https://doi.org/10.1038/s41467-023-37308-z

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.