UPF1 plays critical roles in early B cell development

The ATP-dependent RNA helicase UPF1 plays a crucial role in various mRNA degradation pathways, most importantly in nonsense-mediated mRNA decay (NMD). Here, we show that UPF1 is upregulated during the early stages of B cell development and is important for early B cell development in the bone marrow. B-cell-specific Upf1 deletion in mice severely impedes the early to late LPre-B cell transition, in which VH-DHJH recombination occurs at the Igh gene. Furthermore, UPF1 is indispensable for VH-DHJH recombination, without affecting DH-JH recombination. Intriguingly, the genetic pre-arrangement of the Igh gene rescues the differentiation defect in early LPre-B cells under Upf1 deficient conditions. However, differentiation is blocked again following Ig light chain recombination, leading to a failure in development into immature B cells. Notably, UPF1 interacts with and regulates the expression of genes involved in immune responses, cell cycle control, NMD, and the unfolded protein response in B cells. Collectively, our findings underscore the critical roles of UPF1 during the early LPre-B cell stage and beyond, thus orchestrating B cell development.

In addition to transcriptional control of gene expression, posttranscriptional mRNA decay programs emerge as indispensable players in sculpting the transcriptomic landscape, a critical role during B cell development 6 .A dynamic interplay of RNA-binding proteins (RBPs) is initiated by their recognition of cis-elements within target mRNAs.This recognition paves the way for the recruitment of ribonucleases that execute mRNA degradation through processes such as deadenylation or endoribonuclease activities.Among the array of RBPs, the Zinc finger protein 36 (ZFP36) family members, ZFP36L1 and ZFP36L2, emerge as key players in promoting V(D)J recombination.They achieve this by preserving cell quiescence before pre-BCR expression, primarily through their recognition of AU-rich elementcontaining mRNAs 7 .The splicing regulators TIA1 and TIAL1 play a pivotal role in B cell development by inducing the expression of the DNA damage repair machinery 8 .In addition, post-transcriptional mRNA N 6 -methyladenosine (m 6 A) modification catalyzed by is required for the large-pre-B-to-small-pre-B transition, emphasizing the multi-tiered significance of post-transcriptional regulation during the B cell development 4 .The components of RNA decay machinery, such as CNOT3 within the carbon catabolite repression 4-negative on TATAless (CCR4-NOT) deadenylation complex, are essential for V(D)J recombination via the suppression of p53 expression 9,10 .Furthermore, elements of RNA exosome, another ribonuclease complex, contribute by regulating proper B lymphopoiesis during the pro-B cell to large pre-B cell transition, primarily through the degradation of long noncoding RNAs 11,12 .With the accumulating knowledge in this field, B cell differentiation serves as an ideal system for the in vivo assessment of the roles played by post-transcriptional regulation.
While the in vitro mechanisms of UPF1-mediated mRNA metabolism are well-characterized, its in vivo role has remained less known.In this study, we generate B cell-specific Upf1 knockout mice, allowing us to explore the cell-intrinsic function of UPF1 in B cell development.Thus, we follow this process from early to late LPre-B cells and through the sPre-to the immature B stage.UPF1 interacts with and controls the expression of genes involved in immune responses, cell cycle regulation, and unfolded protein response (UPR) in B cells.Collectively, this study provides mechanistic insight into the complex in vivo role of UPF1 in B cell differentiation.

UPF1 expression increases in early LPre-B cells and late LPre-B cells
We first examined the expression level of UPF1 at distinct stages of B cell development using mouse BM cells (Fig. 1a and Supplementary Fig. 1a).Interestingly, the UPF1 protein expression was remarkably upregulated in early LPre-and late LPre-B cells compared to pre-pro-, pro-, sPre-, Imm-, and recirculating (Rec)-B cells (Fig. 1b).Furthermore, phosphorylation of UPF1 at S1111 residue, which is known to be mediated by SMG1 during initiation of NMD 22 , was highly increased in early LPre and late LPre-B cells.Conversely, the mRNA expression levels of Upf1 remained relatively consistent across the various stages of B cell differentiation (Fig. 1c), suggesting that the UPF1 protein is controlled translationally and/or post-translationally during B cell development.These data indicate that both the expression and activity of UPF1 dynamically change during B cell development.

UPF1 is essential for guiding B cell differentiation from early to late LPre-B stage
To investigate the role of UPF1 during B cell development, we generated Upf1 flox mice and crossed them with Mb1-Cre mice to achieve B cell-specific Upf1 depletion (Supplementary Fig. 1b-d).In Upf1 flox mice, two loxP sites are flanking exons 4-6 of the Upf1 gene, resulting in the suppression of the UPF1 helicase domain expression upon the expression of Cre recombinase (Supplementary Fig. 1b and c).The size of the spleen and the number of splenocytes in Upf1 Flox/Flox Mb1 Cre/+ (Upf1-cKO) mice were noticeably smaller than that of control (Ctrl) mice (Fig. 1d and e).B cells were rarely found in the spleen of Upf1-cKO mice (Fig. 1f and g), indicating that UPF1 is essential for maintaining a mature B cell population in secondary lymphoid organs.
We next examined the effect of UPF1 depletion on B cell differentiation in the BM.We confirmed that UPF1 protein expression was abolished at Pro-B and early LPre-B stages in Upf1-cKO mice (Fig. 1h).Flow cytometry analysis revealed that late LPre-, sPre-, Imm-, and Rec-B cell populations vanished in Upf1-cKO mice, while the numbers of pro-B and early LPre-B cells remained comparable between Ctrl and Upf1-cKO mice (Fig. 1i and j).These findings offer compelling evidence that UPF1 depletion obstructs the differentiation process from early to late LPre-B cells, a pattern that aligns with the observed upregulation of UPF1 expression in this developmental stage (Fig. 1b).The primary distinction between early LPre-and late LPre-B cells is that early LPre-B cells undergo V H -D H J H recombination, whereas late LPre-B cells bear the recombined Igh gene (Fig. 1a).To examine whether the interruption of B cell differentiation in Upf1-cKO mice occurs during the V H -D H J H recombination phase, we stained the BM B cells with cell surface markers, c-kit and CD25, which are highly expressed before and after V H -D H J H recombination, respectively.We rarely detected CD25 + B cells in the Upf1-cKO BM (Fig. 1k), ascertaining that UPF1 is essential to establish Igh-recombined cells during B cell differentiation.These results suggest that UPF1 is required for the transition from early to late LPre-B cell stage, characterized by V H -D H J H recombination.

Transcriptome-scale profiling of early LPre-B cells lacking Upf1
Considering the essential role of UPF1 in the differentiation from early to late LPre-B stage, we examined the gene expression profile of early LPre-B cells from Upf1-cKO and Ctrl mice through RNA sequencing (RNA-seq) analysis (Fig. 2a and Supplementary Data 1).Our analysis revealed that 540 genes exhibited significantly higher expression levels (Log 2 fold change (FC) > 2, adjusted P (adj.P) < 0.05), and 201 genes displayed reduced expression (Log 2 FC < −2, adj.P < 0.05) in Upf1-deficient early LPre-B cells (Fig. 2a and Supplementary Data 1).Prominently recognized NMD target genes, including Gas5, Gadd45b, and Ddit3, exhibited elevated expression in Upf1-deficient cells, suggesting a compromised NMD mechanism within Upf1-deficient early LPre-B cells (Fig. 2a).In addition, gene set enrichment analysis (GSEA) using hallmark gene sets revealed enrichment of gene sets associated with interferon (IFN) responses among highly expressed genes in Upf1deficient cells (Fig. 2b).Notably, IFN-stimulated genes (ISGs), such as Mx2, Ifi44, and Oasl1, were highly expressed in Upf1-deficient early LPre-B cells (Fig. 2c, d).Conversely, genes linked to E2F TARGETS, OXIDATIVE PHOSPHORYLATION, MYC TARGETS V1, and G2M CHECKPOINT showed decreased expression in Upf1-deficient cells (Fig. 2e, f and Supplementary Fig. 2a-f).Moreover, GSEA using all gene sets revealed that Upf1-deficient cells expressed diminished levels of the genes associated with adaptive immune responses and cell cycle, including Igh-derived transcripts and histone mRNAs, respectively (Fig. 2g-i and Supplementary Fig. 2g, h).These findings collectively underscore the protective role of UPF1 in preventing aberrant expression of IFN signature genes while emphasizing the dependency on UPF1 for the expression of Igh-derived transcripts and genes associated with the cell cycle.

Upf1-cKO mice
Since type I IFNs are reported to impede pro-B cell development 23 , we hypothesized that dysregulated expression of ISGs might disrupt the proper differentiation of B cells in Upf1-cKO mice.Therefore, we sought to investigate whether the inhibition of IFN signaling could ameliorate the defects in B cell development in Upf1-cKO mice by crossing IFN receptor (Ifnar1)-KO mice with Upf1-cKO mice.However, the spleen size in Upf1/Ifnar1 double knockout (DKO) mice (Upf1 Flox/ diminished in DKO mice, comparable to the observations in Upf1-cKO mice (Fig. 3b).Notably, Upf1/Ifnar1 DKO mice showed B cell differentiation defects at early to late LPre-B transition, mirroring the patterns observed in Upf1-cKO mice (Fig. 3c and d).The analysis of early LPre-B cells in DKO mice revealed a reduction in ISG expression, confirming the suppression of IFN signaling (Fig. 3e).Noteworthy, we observed no apparent difference in the B cell development in the BM between control (Upf1 Flox/Flox Mb1 +/+ Ifnar1 +/-) and Ifnar1-KO (Upf1 Flox/Flox Mb1 +/+ Ifnar1 -/-) mice, suggesting that the loss of type I IFN signaling alone does not affect B cell development in the BM (Supplementary Fig. 3).Collectively, these data demonstrate that the differentiation defect from early LPre-B cells to late LPre-B cells observed in Upf1-cKO mice cannot be solely attributed to the heightened IFN signaling levels.)  UPF1 is necessary for Igh V H -D H J H recombination during early LPre-B stage The GSEA revealed a marked reduction in Igh transcripts in Upf1-deficient early LPre-B cells compared with Ctrl cells (Fig. 2g, h).Therefore, we further checked the expression of individual transcripts within the Igh locus.The majority of Ighv transcripts, which constitute the variable region, were found to be significantly diminished in Upf1-deficient early LPre-B cells (Fig. 4a).In contrast, transcript levels of Ighm (representing a constant region), Ighj (joining region), and Ighd (diversity region) remained unaltered following UPF1 depletion (Fig. 4b-d).
It is known that the expression of Ighv transcripts is induced during or after the V H -D H J H recombination 24 .Therefore, we next investigated whether UPF1 depletion has any impact on V H -D H J H recombination by performing genomic DNA PCR using three different primer sets: one for the detection of distal (V H J558 to D H J H ), another for proximal (V H 7183 to D H J H ), and the third for D H to J H rearrangement (Fig. 4e).In Ctrl early LPre-B cells, all PCR products corresponding to D H -J H , distal V H -D H J H , and proximal V H -D H J H rearranged genomic DNA were detected, indicating the occurrence of D H -J H and V H -D H J H recombination (Fig. 4f).In contrast, Upf1-deficient early LPre-B cells exhibited an absence of both proximal and distal V H -D H J H rearranged bands, while D H -J H rearrangement was still detectable under Upf1 deficiency (Fig. 4f).These findings indicate that Upf1 deficiency selectively inhibits V H -D H J H while leaving D H -J H recombination unaffected.Accordingly, the protein expression of the intracellular Igμ heavy chain was rarely detected in CD43 + , CD2 -B cells from Upf1-cKO mice (Fig. 4g).Collectively, these data demonstrate the requirement of UPF1 in facilitating V H -D H J H recombination at the early LPre-B stage, thereby enabling the proper expression of the Igμ heavy chain.
The aforementioned results motivated us to investigate whether introducing genetically pre-arranged Igh could rescue the impaired B cell development in Upf1-cKO mice.To address this, we crossbred Upf1-cKO mice with B1-8 hi Igh pre-arranged mice 25 (hereafter referred to as cKO/Igh B1-8hi mice).The number of late LPre-and sPre-B cells was significantly increased in cKO/Igh B1-8hi mice compared with Upf1-cKO mice with WT Igh gene (cKO/Igh WT ) mice, with the number of late LPre-B cells notably similar between cKO/Igh B1-8hi and Ctrl/Igh B1-8hi mice (Fig. 4h, I and Supplementary Fig. 4a).Furthermore, c-Kit − , CD25 + B cells in cKO/Igh B1-8hi mice exceeded those observed in cKO/Igh WT mice (Supplementary Fig. 4b, c).These findings suggest that the genetic prearrangement of the Igh locus facilitates early to late LPre-B cell differentiation, even in the absence of UPF1.

Commonalities and distinctions in the gene expression profiles of B cell progenitors lacking RBPs
Previous studies have demonstrated the critical role of RBPs, specifically ZFP36L1/L2 and CNOT3, in B cell development 7,10 .These studies showed that Zfp36l1 Flox/Flox Zfp36l2 Flox/Flox Mb1 cre/+ (Zfp36l1/l2-double conditional knockout (DCKO)) mice and Cnot3 Flox/Flox Mb1 cre/+ (Cnot3-cKO) mice exhibited B cell differentiation defects just prior to V H -D H J H recombination at the Igh locus, a phenotype quite similar to that of Upf1-cKO mice.Consequently, we sought to compare the gene expression profiles of B cell progenitors lacking Upf1, Zfp36l1/l2, and Cnot3.The genes with low expression in Upf1-deficient early LPre-B cells showed enrichment in genes with low expression in Zfp36l1/l2-DCKO and Cnot3-cKO B cells (Fig. 5a, blue lines).In contrast, the highly expressed genes in Upf1-deficient early LPre-B cells exhibited enrichment solely in genes highly expressed in Zfp36l1/l2-DCKO B cells rather than in Cnot3-cKO B cells (Fig. 5a, red lines).This suggests that the RNA expression profile in Upf1-cKO B cells is closer to that of Zfp36l1/l2-DCKO B cells than Cnot3-cKO B cells.
The compromised V H -D H J H recombination observed in Zfp36l1/l2-DCKO B cells is attributed to the loss of cell cycle quiescence, a critical step for promoting V(D)J recombination 7,26 .Therefore, we delved into the regulatory role of UPF1 in the cell cycle of early LPre-B cells.We conducted BrdU assays to analyze the cell cycle status of B cells undergoing V H -D H J H recombination (CD19 + , B220 + , sIgM − , CD43 + , CD2 -).The results revealed an increase in the number of cells in the S-phase among Upf1-deficient cells (Fig. 5b and c).In our RNA-seq analysis, E2F target genes and other cell cycle-related genes showed lower expression in Upf1-deficient early LPre B cells (Fig. 2b and i), probably because Upf1-cKO early LPre B cells lack pre-BCR, which induces potent cell cycles 2,27 .Nevertheless, even in the absence of pre-BCR expression, a set of cell cycle-related genes exhibited higher expression in Upf1-deficient early LPre B cells (Fig. 5d), implying that UPF1 is involved in suppressing some cell cycle-related genes.These results suggest that Upf1-deficiency promotes the transition from the G0/G1-phase to the S-phase, and the loss of cell cycle quiescence may emerge as a plausible cause for the hindrance of V H -D H J H recombination in Upf1-deficient early LPre-B cells.
Further GSEA revealed the enrichment of gene sets associated with cell cycle progression, such as E2F TARGETS (adj.P = 0.087), as well as UNFOLDED PROTEIN RESPONSE (UPR) (adj.P = 0.37, P = 0.021) in Upf1-cKO/Igh B1-8hi sPre-B cells compared to Ctrl/Igh B1-8hi cells (Fig. 6e).We next investigated transcriptome transition from early LPre-to sPre-B cells in Ctrl/B1-8 hi and cKO/B1-8 hi (Supplementary Fig. 6a, b).In control mice, during the differentiation from early LPre-to sPre-B cells, the gene sets related to cell cycling, such as the E2F target, G2M_Chcekpoint, and MYC target, exhibited downregulation    (Supplementary Fig. 6a).These alterations align with the transition from actively proliferating large Pre-B cells to quiescent small Pre-B cells.The enrichment of the E2F target gene set in Upf1-deficient sPre-B cells (cKO/Igh B1-8hi ), but not early LPre-B cells (Fig. 6e), suggests a potential requirement of UPF1 in driving the downregulation of cell cycle-related gene expression during the transition from early LPre-to sPre-B stage.
In contrast, consistent with the results of PCR assay for the Ig light chain, the expression levels of v-region transcripts from Ig light chains (Igkv and Iglv) were largely similar between UPF1-cKO/Igh B1-8hi and Ctrl/ Igh B1-8hi sPre-B cells (Fig. 6f).This finding suggests that UPF1 is dispensable for the recombination of Ig light chain DNA.
Notably, the gene sets "INTERFERON ALPHA RESPONSE" and "INTERFERON GAMMA RESPONSE", which were significantly enriched in cKO early LPre-B cells, did not show enrichment in early LPre-B cells in the absence of UPF1 with the pre-rearranged Igh allele (cKO/Igh B1-8hi ) (Fig. 6g).Consistently, genes related to the IFN response were only modestly elevated in cKO/Igh B1-8hi early LPre-B cells, whereas these genes were more robustly increased in cKO/Igh WT early LPre-B cells (Fig. 6h).These data suggest that the Igh pre-rearrangement prevented the aberrant expression of IFN response-related genes in Upf1-deficient early LPre-B cells.In contrast, IFN response-related genes were upregulated in sPre-B cells from cKO/Igh B1-8hi compared with Ctrl/Igh B1- 8hi mice (Fig. 6h).For instance, ISGs, such as Ifi44 and Ddx60, started to elevate in cKO/Igh B1-8hi sPre-B cells (Supplementary Fig. 6c), suggesting that the Igh pre-rearrangement failed to suppress the expression of IFN response-related genes in sPre-B cells lacking UPF1.These results imply that UPF1 plays a role in preventing the abnormal expression of IFN response-related genes not only in the early LPre-B stage but also in the sPre-B stage.
The UPR acts as a critical checkpoint in B cell development by regulating the maintenance of BCR expression and proper transport to the cell surface [28][29][30] .Genes associated with the UPR were highly enriched in both early LPre-B (adj.P = 0.0075, P < 0.0001) and sPre-B cells (adj.P = 0.37, P = 0.021) from cKO/Igh B1-8hi mice (Fig. 6e, g).We also found a comparable number of UPR genes affected by UPF1 depletion in Igh WT early LPre-, Igh B1-8hi early LPre-, and Igh B1-8hi sPre-B cells (Fig. 6i).These data suggest that, unlike IFN-related genes, the omission of Igh recombination through B1-8 hi knock-in did not mitigate the abnormally high expression of UPR genes in Upf1-deficient B cells.Some of the UPR genes were downregulated during the differentiation from early LPreto sPre-B stage in Upf1-Ctrl/Igh B1-8hi B cells (GSEA: adj.P = 0.060) (Fig. 6j and Supplementary Fig. 6a), suggesting a delayed downregulation of UPR genes during the transition from early LPre-to sPre-B stage in Upf1-deficient B cells.Collectively, these data demonstrate that UPF1 is critical for ensuring the transcriptome shift from the early LPre-to sPre-B stage by preventing abnormal expression of genes related to cell cycle, IFN response, and the UPR.

UPF1 downregulates the expression of its target RNAs in early LPre-and sPre-B stages
We next searched for RNAs directly targeted by UPF1 in B cells.It was reported that immunoprecipitation with anti-p-UPF1 (S1111) successfully isolated NMD-target mRNAs, while steady-state UPF1 largely nonspecifically associated with mRNAs 22 .We conducted RNAimmunoprecipitation sequencing (RIP-seq) analysis using anti-p-UPF1 antibody in activated splenic B cells since it was difficult to isolate a sufficient amount of early LPre-B cells for this experiment (Fig. 7a and Supplementary Fig. 7a).We identified 812 p-UPF1-bound RNAs, including prominent NMD targets such as Smg5, Gadd45b, and Ddit3 (Fig. 7b and Supplementary Data 3).Gene ontology analysis revealed various functions of UPF1-target mRNAs, such as NF-κB activation and the UPR (Fig. 7c).
Interestingly, we found that the RNAs directly targeted by p-UPF1 were enriched in Upf1-cKO early LPre-B, cKO/Igh B1-8hi early LPre-B cells, and cKO/Igh B1-8hi sPre-B cells (Fig. 7d-f).These data suggest that UPF1 directly downregulates the expression of the target mRNAs in both early LPre-and sPre-B stages.On the other hand, the RNAs identified as UPF1-targets were not enriched in Cnot3-deficient or Zfp36l1/l2-deficient progenitor B-cells, indicating that not all UPF1-binding RNAs are concurrently regulated by ZFP36L1/L2 or CNOT3 despite the transcriptome similarity between Upf1-deficient early LPre-and Zfp36l1/l2deficient progenitor B cells (Supplementary Fig. 7b, c).These results suggest that although the initial dysregulations of RNA metabolism differ, Upf1-cKO and Zfp36l1/l2-DCKO B cells eventually converge toward similar RNA expression profiles.
Since our RIP-seq analysis did not cover the entire transcripts in early LPre-and sPre-B cells, identifying UPF1-target RNAs among all the highly expressed genes in Upf1-deficient progenitor B cells posed a challenge (Supplementary Fig. 7d).However, we observed that UPR genes targeted by UPF1 were highly expressed in Upf1-deficient early LPre-and sPre-B cells (Fig. 7g).In addition, UPF1-targeted mRNAs in IFN response-or cell cycle-related genes were also highly expressed in Upf1-deficient early LPre-B cells (Fig. 7h, i).These findings suggest that UPF1-mediated direct mRNA suppression contributes to the appropriate gene expression status at the early LPre-and sPre-B stages of B cell development.

Discussion
In this study, we have elucidated the vital role of UPF1 in B cell development.The disruption in B cell development observed in the absence of UPF1 is directly attributed to the impediment of Igh V H -D H J H recombination during the early LPre-B stage.The association established between UPF1 and V H -D H J H recombination elucidates a fundamental mechanism, underscoring the indispensable role of UPF1 in orchestrating B cell differentiation from early LPre-to late LPre-B stage through V H -D H J H recombination in the mouse BM.In addition, UPF1 is essential for sPre-immature B cell differentiation.UPF1 suppresses a Remarkably, we observed a significant upregulation in the protein expression and phosphorylation of UPF1 during the early LPre-B cell stage, coinciding with the occurrence of V H -D H J H recombination. Several explanations for this phenomenon are plausible.Firstly, it is conceivable that the translation efficiency of Upf1 mRNA is augmented in early LPre-B cells.Alternatively, the activity of SMG1 kinase, a key player in UPF1 phosphorylation, might be modulated during B cell differentiation.Subsequent investigations are warranted to elucidate whether specific signaling pathways exist to directly control the translation and phosphorylation of UPF1, thereby exerting precise control over B cell differentiation.) sPre (Igh B1-8hi )  ).The upregulated genes and the downregulated genes in sPre-B cells from early LPre-B cells are colored purple and green, respectively.The statistical analyses were performed using edgeR.d Enrichment of the upregulated genes in sPre-B cells (purple) and the downregulated genes in sPre-B cells (green) in Upf1-cKO/Igh B1-8hi sPre-B cells.e, g Normalized enrichment score of GSEA (Upf1-cKO/Igh B1-8hi vs Ctrl/Igh B1-8hi ) in sPre-B cells (e) and early LPre-B cells (g).
The gene sets positively or negatively enriched in Upf1-cKO/Igh WT early LPre-B cells (Fig. 2b) are pointed with pink or green arrows, respectively.f The ratio of differentially expressed transcripts derived from Igh v region (Upf1-cKO vs Ctrl in early LPre-B) and Igk/Igl v region (Upf1-cKO/Igh B1-8hi vs Ctrl/Igh B1-8hi in sPre).The statistical analyses were performed using limma (Igh WT ), and edgeR (Igh B1-8hi ).h, i The ratio of highly expressed IFNA/IFNG RESPONSE genes (h) and UNFOLDED PROTEIN RESPONSE genes (i) in Upf1-deficient B cells with indicated genotypes and cell types.Hallmark gene sets from GSEA are used (IFN ALPHA RESPONSE and IFN GAMMA RESPONSE are combined without gene overlap).The statistical analyses were performed using limma (Igh WT ), and edgeR (Igh B1-8hi ).j Comparison of UPRrelated gene expression between early LPre-and sPre-B cells (Ctrl/Igh B1-8hi ). Green dots indicate significantly upregulated genes in sPre-B cells compared to early LPre-B cells.Only the genes highly expressed in Upf1-cKO/Igh B1-8hi sPre-B cells are shown.Source data are provided as a Source Data file.a Scheme used to identify UPF1 target RNAs in B cells with RIP-seq.Isolated splenic B cells were stimulated for 3 days.After stimulation, the B cells were lysed in RNA IP lysis buffer and incubated with anti-phospho-UPF1 Ab bound to protein G magnetic beads.The precipitated RNAs were eluted from the beads and analyzed by RNAseq.b RNAs precipitated with p-UPF1 antibody in splenic B cells.The x-axis indicates Log2FC normalized with RNA expression in splenic B cells (input RNAs) and the y-axis indicates Log2FC normalized with RNAs precipitated with control IgG antibody.Yellow dots indicate highly enriched RNAs; we defined them as UPF1binding targets (see "Methods").Known NMD targets are highlighted in light green.c Gene ontology analysis of UPF1-target RNAs.The statistical analyses were performed using Metascape.d-f Enrichment score plot of UPF1-binding targets (yellow) and all genes detected in splenic B cells (gray) in Upf1-cKO vs Ctrl (d); early LPre-B, Igh WT , (e); early LPre-B, Igh B1-8hi , (f); sPre-B, Igh B1-8hi ).g Comparison of "response to unfolded protein expression (GO:0006986)" gene expression between Upf1-cKO vs Ctrl in indicated genotypes and cell types.The number adjacent to each bar indicates adj.P (red; adj.P < 0.05).Only the genes identified as UPF1-target are shown.The statistical analyses were performed using limma (Igh WT ), and edgeR (Igh B1-8hi ).h, i Volcano plot comparing Upf1-cKO and Ctrl (early LPre-B) of the genes included in INTERFERON ALPHA RESPONSE (h) and CELL CYCLE CHECKPOINTS (i).Yellow dots indicate the genes identified as UPF1-targets.The statistical analyses were performed using limma.

Isolate B cells (CD43
An intriguing question that remains to be addressed is the precise mechanism by which UPF1 orchestrates the regulation of V H -D H J H recombination in early LPre-B cells.We observed that the loss of cell cycle quiescence in Upf1-deficient early LPre-B cells is one potential factor contributing to the inhibition of V H -D H J H recombination.Moreover, previous studies have provided evidence that germline gene transcripts (GLTs) of Igh play a supportive role in DNA recombination 31,32 .Notably, the expression of transcripts originating from the V region of Igh was significantly reduced in Upf1-deficient early LPre-B cells.In contrast, D H -J H recombination, a process preceding V H -D H J H recombination, proceeded even in Upf1-deficient B cells, with the levels of transcripts arising from the D or J region of Igh remaining unaltered in Upf1-deficient early LPre-B cells.Therefore, this suggests that UPF1 might exert control over the expression of GLTs of the Igh V region, potentially initiating V H -D H J H recombination. Further studies are required to shed light on the intricate mechanisms governing how UPF1 regulates V H -D H J H recombination.
Through transcriptome analysis, we have identified a spectrum of genes that exhibit differential expression in the Upf1-deficient early LPre-B cells.Among these changes, one of the most striking alterations was the marked increase in the expression of ISGs.The abnormally heightened expression of ISGs was attenuated upon the knockout of the IFN receptor, implying the activation of the IFN signaling pathway in Upf1-deficient early LPre-B cells.Interestingly, pre-rearrangement of Igh, conducted by crossing Igh B1-8hi knock-in mice, also inhibited the abnormally high expression of IFN-response genes in Upf1-deficient early LPre-B cells.On the other hand, Upf1-cKO/Igh B1-8hi sPre-B cells, in which V L -J L recombination occurs, showed higher expression of IFNresponse genes.Given that both Upf1-cKO early LPre-B cells and Upf1-cKO/Igh B1-8hi sPre-B cells undergo DNA recombination at D H -J H and V L -J L regions, respectively, it is tempting to hypothesize that DNA recombination triggered the aberrant activation of IFN pathway in Upf1deficient B cells.Despite this IFN signaling activation, the B cell developmental blockade was not rectified in Upf1/Ifnar1 double knockout mice.These results suggest that the accumulation of ISGs due to Upf1 deficiency does not significantly influence the transition from early to late LPre-B cells.It is well documented that spontaneous expression of ISGs can be induced by the abnormal accumulation of double-stranded (ds) RNA, often stemming from disruptions caused by the failure of RNA metabolisms, such as the loss of METTL3, an N 6methyladenosine (m 6 A) methyltransferase 33 , or the absence of an RNA editing enzyme ADAR1 34 .Given that impaired NMD due to Upf1 deficiency may lead to the accumulation of aberrant mRNAs, it is also possible that UPF1 plays a critical role in suppressing IFN responses by preventing aberrant buildup of dsRNA.
Previous studies have reported that the deficiency of certain other RBPs leads to a halt in B cell differentiation, typically occurring just before the V H -D H J H recombination 4,7,8,[10][11][12]35 . Amog these RBPs, CNOT3 is a component of the CCR4-NOT deadenylase complex, responsible for promoting poly(A) tail shortening in various mRNA decay machinery, including NMD and ZFP36L1/2-mediated mRNA decay.However, our findings indicate that the RNA expression profile of early LPre-B cells in Upf1-cKO mice is closer to that of Zfp36l1/l2-DCKO B cells than Cnot3-cKO B cells.We must mention that the slight disparities in data acquisition methods among these three data sets, such as variations in the gating strategy for early developing B cells, might impact the analysis results.Furthermore, our investigation revealed that the expression of p-UPF1-binding RNAs was not enriched in either Zfp36L1/L2-DCKO or Cnot3-cKO B cells.In the context of the NMD pathway, RNA degradation is mediated by the NMD-specific endonuclease SMG6, alongside the SMG5 and SMG7 proteins, which recruit the CCR4-NOT deadenylase complex 15,18 .SMG6, under the guidance of UPF1, triggers endonucleolytic cleavage of mRNAs, independent of deadenylation 16,17 .Thus, the SMG6-mediated RNA degradation pathway may exhibit a preference for early LPre-B cells.
An unresolved question pertains to how UPF1 regulates the differentiation from sPre-B cells to immature B cells, a process involving Ig light chain recombination and subsequent expression of BCR (IgM) on the cell surface.While Upf1-deficient sPre-B cells were capable of undergoing Ig light chain recombination, they failed to express BCR on the cell surface.Our investigation revealed elevated expression of UPR genes, some of which are directly targeted by UPF1, in Upf1-deficient sPre-B cells.The increased expression of UPR genes in Upf1-deficient cells might exacerbate unnecessary UPR, potentially impeding the expression of BCR and its proper transport to the cell surface [28][29][30]36 . Thedisruption could hinder the differentiation process from sPre-B cells to immature B cells.
We found that UPF1 directly suppresses a specific set of RNAs associated with cell cycle regulation, the UPR, and IFN responses during B cell differentiation.However, it remains unclear whether UPF1 targets 40 to 60% of highly expressed genes in Upf1-deficient progenitor B cells, as the RNAs identified in RIP-seq using splenic B cells did not encompass all the genes expressed in early LPre-or sPre-B cells.Therefore, there is a possibility of additional target mRNAs specifically expressed in early LPre-and sPre-B cells that are regulated by UPF1.Furthermore, considering the dynamic changes in the abundance of phosphorylated UPF1 during B cell differentiation, it is intriguing to investigate whether the binding specificity of UPF1 varies depending on the stages of B cell differentiation.
In summary, this study collectively underscores the intrinsic role of UPF1 as a pivotal factor in the transition from early to late LPre-B and sPre-B to Immature-B differentiation.UPF1 is prerequisite for V H -D H J H recombination without affecting D H -J H recombination. Furthermore, UPF1 ensures the transcriptome dynamics from early LPre-to sPre-B cells.Our findings contribute to a deeper understanding of the intricate molecular mechanisms governing B cell differentiation.

Study approval
All animal experiments were conducted in compliance with the regulations approved by the Committee for Animal Experiments of the Graduate School of Medicine and Institute for Frontier Life and Medical Sciences, Kyoto University.

Mice
The targeting vector was constructed by inserting a 2.0 kb fragment containing exons 4 to 6 of the Upf1 gene flanked by loxP sites, 8.0 kb of a 5′ sequence, 1.0 kb of a 3′ sequence, and a neomycin (neo)-resistant gene flanked by FRT sites into a PGKneoF2L2DTA vector.The targeting vector was linearized and transfected into EGR-101 embryonic stem (ES) cells derived from C57BL/6 mice by electroporation, and G418resistant clones were screened for homologous recombination by PCR and Southern blot analysis.The primers used for PCR are listed in Supplementary Table 1.The successfully recombined clones were microinjected into blastocytes derived from ICR mice and transferred to pseudopregnant females.Mating of chimeric male mice to C57BL/6 female mice resulted in the transmission of the floxed allele to the germline.The Upf1 Flox/+ mice were crossed with FLP-Cre mice (C57BL/6 background) to eliminate the neo gene and then crossed with the indicated Cre expressing mice (C57BL/6 background).
Mb1 cre/+ mice (C57BL/6 background) 37 were kindly provided by Dr. Michael Reth (University of Freiburg).V H B1-8 high knock-in mice (termed in this study as B1-8 hi ) (C57BL/6 background) 25 were kindly provided by Dr. Michel C. Nussenzweig (The Rockefeller University).Mice were used at 8 to 16 weeks old, male and female, and were age-matched.Mice were euthanized by cervical dislocation.Mice were co-housed with the control mice and maintained under specific pathogen-free (SPF) conditions at Kyoto University animal facility with a 12 h light/ dark cycle and access to food and water ad libitum.Room temperature was maintained at 23 ± 3 °C, with a relative humidity of 50 ± 20% and all 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 licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence 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 licence, visit http://creativecommons.org/ licenses/by/4.0/.

Fig. 2 |
Fig. 2 | Transcriptome analysis of Upf1-depleted early LPre-B cell.a Volcano plot of the genes identified in the RNA-seq analysis of Upf1-cKO and Ctrl early LPre-B cells.Pink dots indicate highly expressed genes in cKO (Log 2 FC > = 2, adj.P < 0.05), and green dots indicate low expressed genes in cKO (Log 2 FC < = − 2, adj.P < 0.05).Known NMD targets (Gas5, Gadd45b, and Ddit3) are highlighted in yellow.The statistical analyses were performed using limma.Source data, including exact adjusted p-values, are provided as a Source Data file.b Normalized enrichment score of GSEA using hallmark gene sets.A positive number indicates the gene sets enriched in highly expressed genes in Upf1-cKO early LPre-B cells, and a negative number indicates those enriched in low expressed genes in Upf1-cKO.Significantly enriched gene sets (adj.P < 0.01) are colored in light blue.The statistical analyses were performed using GSEA.c-h Enrichment score plots of "INTERFERON ALPHA RESPONSE" (c), "E2F TARGES" (e), and "IMMUNOGLOBULIN COMPLEX" (g) and bar graph showing Log2FC (Upf1-cKO vs Ctrl) of the representative genes in "INTER-FERON ALPHA RESPONSE" (d), "E2F TARGETS" (f), and "IMMUNOGLOBULIN COMPLEX" (H).The top 10 genes showing the highest (d) or lowest (f, h) Log 2 FC are picked up, and the number adjacent to each bar indicates adj.P.The statistical analyses were performed using limma.i The result of GSEA using all gene sets.Each node indicates identified gene sets and overlapping gene sets are grouped.The size of each dot indicates a normalized enrichment score.Pink and green nodes indicate the gene sets enriched in highly expressed genes and low expressed genes in cKO, respectively.

Fig. 3 | 6 NatureFig. 4 |
Fig. 3 | Blocking of IFN receptor signal does not rescue the B cell development in Upf1-cKO mice.a Appearance of mouse spleens from indicated genotypes.Results are representative of at least three independent experiments.b Flow cytometry plots of indicated populations in the splenocytes derived from indicated mice.Results are representative of at least two independent experiments.c Flow cytometry plots of indicated populations in the BM of indicated mice.d Flow cytometry plots of FVD − , CD19 + , B220 + , sIgM − cells derived from the BM of indicated mice.Results are representative of at least two independent experiments.e The mRNA expressions of indicated genes in early LPre-B cells were derived from the indicated mice (n = 3 per group).Each bar represents the mean ± SD from biological replicates.Statistical significance was calculated with a two-sided Student's t test.Source data are provided as a Source Data file.

Fig. 5 |
Fig. 5 | Phenotypic similarity between Upf1-cKO mice and Zfp36l1/l2-DCKO mice.a Enrichment score plot of the differentially expressed genes of Upf1deficient early LPre-B cells in Zfp36l1/l2-DCKO or Cnot3-cKO progenitor B cells.The red line indicates enrichment of the high expression genes in Upf1-cKO cells (Log2FC > = 2, adjP < 0.05), and the blue line indicates enrichment of the low expression genes in Upf1-cKO cells (Log2FC < = − 2, adjP <0.05).b, c The representative flow cytometry plots of BrdU assay in FVD − CD19 + , B220 mid , sIgM − , CD43 + , CD2 − cells (c) and the percentages of cells in S-phase (N = 5 for each genotype).Statistical significance was calculated with a two-sided Student's t test.d Highly expressed genes included in the "REACTOME CELL CYCLE CHECKPOINTS" gene set in Upf1-cKO early LPre-B cells.Bars indicate Log2FC (Upf1-cKO vs Ctrl), and the number adjacent to each bar indicates adj.P.The statistical analyses were performed using limma.Source data are provided as a Source Data file.

Fig. 6 |
Fig.6| UPF1 ensures transcriptome dynamics throughout the early LPre-to sPre-B cell differentiation.a PCR assay amplifying recombined Igk in sPre-B cells derived from the indicated mice.Genome DNA derived from the tail of WT was used as a negative control.Results are representative of two independent experiments.b Immunoblot showing protein expression of UPF1 in sPre-B cells from Ctrl/ Igh B1-8hi and cKO/Igh B1-8hi mice.Results are representative of at least three independent experiments.c Volcano plot of RNA expression comparison between early LPre-and sPre-B cells (Ctrl/Igh B1-8hi ).The upregulated genes and the downregulated genes in sPre-B cells from early LPre-B cells are colored purple and green, respectively.The statistical analyses were performed using edgeR.d Enrichment of the upregulated genes in sPre-B cells (purple) and the downregulated genes in sPre-B cells (green) in Upf1-cKO/Igh B1-8hi sPre-B cells.e, g Normalized enrichment score of GSEA (Upf1-cKO/Igh B1-8hi vs Ctrl/Igh B1-8hi ) in sPre-B cells (e) and early LPre-B cells (g).The gene sets positively or negatively enriched in Upf1-cKO/Igh WT early LPre-B cells + cells depletion) 3-day stimulation (IL-4, α-IgM Ab, α-CD40 Ab)

Fig. 7 |
Fig.7| RIP-Seq analysis to determine the direct binding targets of p-UPF1.a Scheme used to identify UPF1 target RNAs in B cells with RIP-seq.Isolated splenic B cells were stimulated for 3 days.After stimulation, the B cells were lysed in RNA IP lysis buffer and incubated with anti-phospho-UPF1 Ab bound to protein G magnetic beads.The precipitated RNAs were eluted from the beads and analyzed by RNAseq.b RNAs precipitated with p-UPF1 antibody in splenic B cells.The x-axis indicates Log2FC normalized with RNA expression in splenic B cells (input RNAs) and the y-axis indicates Log2FC normalized with RNAs precipitated with control IgG antibody.Yellow dots indicate highly enriched RNAs; we defined them as UPF1binding targets (see "Methods").Known NMD targets are highlighted in light green.c Gene ontology analysis of UPF1-target RNAs.The statistical analyses were