Abstract
The mRNAs of inflammatory factors often contain unstable AU-rich elements (AREs) within their 3’ untranslated regions (3’UTRs), and HuR is a key ARE-binding protein to recognize such elements. However, the roles of other Hu family member (s) remain unclear. In this study, we show an inflammatory stimulation-induced short-form and cytoplasm-localized HuB (Hel-N2). In cytoplasm, HuB interacts with HuR as well as HuR-associated pro-inflammatory cytokine/chemokine mRNAs and facilitates the recruitment of these mRNAs into the translation machinery. Furthermore, HuB recruits RNA helicase DHX9 to unwind the secondary structures presented in the 5’ untranslated regions (5’UTRs) of these mRNAs, thereby promoting their translation. The work has been carried out in human and mouse cell lines and female mice. Our finding not only broadens our understanding about how the Hu family members govern pro-inflammatory gene expression, but also suggests that targeting HuB is a viable strategy to treat inflammation-related diseases.
Similar content being viewed by others
Introduction
Eukaryotic cells determine the final protein output of their genetic programs not only through transcriptional control but also via post-transcriptional mechanisms, including the regulation of mRNA localization, translation, and turnover rate1,2. Typically, mRNAs that govern cell growth and the organism’s response to external stimuli, such as pathogens, inflammatory signals, and environmental factors, exhibit highly precise spatiotemporal expression patterns in cells3. For instance, mRNAs encoding inflammatory factors may remain stable over prolonged periods, so ceasing their synthesis does not immediately halt or redirect inflammation. Post-transcriptional mechanisms that modulate the stability and/or translation efficiency of inflammatory factor mRNAs offer more rapid and flexible regulatory controls, thereby particularly crucial in coordinating both the initiation and resolution of inflammation4,5.
Some of the most extensively studied short-lived mRNAs contain adenine/uridine-rich elements (AREs) within their 3′ untranslated regions (3′-UTRs). mRNAs with AREs exhibit instability but are subject to differential regulation in response to inflammatory stimuli, facilitated by the coordinated actions of ARE-binding proteins (AUBPs)6,7,8,9. While most AUBPs function as negative regulators of gene expression at post-transcriptional levels by decreasing mRNA stability, the embryonic lethal aberrant visual-like/human antigen (ELAVL/Hu) proteins generally serve as positive post-transcriptional regulators10,11,12. The Hu family includes four members: HuR, HuB, HuC, and HuD13,14,15. Among them, HuR is constitutively expressed across all tissues, while the other three have been considered primarily expressed in the nervous system and are collectively termed neuron-specific ELAVs (nELAVs)16,17,18,19. HuR is predominantly localized in the nucleus but shuttles into the cytoplasm by exploiting its nucleoplasmic shuttle sequence (HNS) under various stress conditions20,21,22. By binding to AREs, HuR functions as an adaptor protein for mRNA nuclear export and/or a stabilizer for its mRNA targets20. HuR’s involvement in various physiological and pathological processes is well-documented, especially in the post-transcriptional regulation of pro-inflammatory cytokine/chemokine expression and the pathogenesis of inflammatory diseases21,22,23. However, the roles of nELAVs in non-neural cells under physiopathological conditions remain largely unexplored.
HuB, to the best of our knowledge, is the first identified member of the ELAV family, and its functions in non-neural cells has garnered increasing attention to date. Its role in adipocyte differentiation has been documented; in adipocytes, co-sedimentation of HuB with polysome fractions during sucrose gradient centrifugation was observed, and electron microscopy revealed a co-localization of HuB with ribosomal structures, indicating its involvement in protein synthesis24. Recently, HuB’s function in follicular maturation has also come to light. The translation of DDX6, dependent on HuB, is essential for the assembly of P body-like granules and the formation of primordial follicles25. Additionally, HuB is highly expressed in certain tumor cells, and differs from HuR in being predominantly localized within the cytosol. Ectopic expression of HuB in cancer cells can initiate nuclear export of HuR, leading to the formation of a complex between HuB and HuR within the cytosolic fraction. This phenomenon is associated with the upregulated expression of ARE-containing mRNAs, including those encoding HIF-1α, c-Fos, c-MYC, and Est226.
The emerging roles of HuB in non-neural cells led us to investigate its involvement in the pathogenesis of inflammatory diseases. In this study, we demonstrated a significant induction of HuB following inflammatory stimulation, as well as its role in facilitating the effective translation of pro-inflammatory cytokines and chemokines. Located in the cytoplasm, HuB forms complexes with HuR and interacts with RNA helicase. This interaction enables the efficient translation of mRNAs associated with HuR that encode cytokines/chemokines.
Results
Inflammatory stimulation promotes the expression of cytoplasmic short-form HuB (Hel-N2)
To investigate the expression patterns of ELAVs upon inflammatory stimulation, human HEK293 and HeLa cells, mice RAW264.7 and U251 cells, as well as mice lungs were exposed to TNFα or LPS, followed by western blotting analyses. The results indicated that neuronal cells (U251) expressed all ELAV family members, which barely reacted upon TNFα induction. Whereas, non-neuronal cells (HEK293, HeLa and RAW264.7) and mice lungs tissues expressed a measurable level of HuB which can be markedly upregulated in response to TNFα or LPS, contrasting to that of HuC and HuD which was nearly undetectable with or without inflammatory stimulation (Fig. 1A, B). Given the homology among members of the Hu family proteins, we assessed the specificity of the antibodies. The results indicated that the HuR antibody does not recognize GFP-HuB. Conversely, the HuB antibody did not recognize GFP-HuR (Fig. S1A). Moreover, according to online predictions (https://www.proteinatlas.org), HuB is expressed in various other tissues (Fig. S1B), including approximately 60% of small cell lung cancer cases27,28.
The expression of HuR, HuB, HuC and HuD in various cell lines and mice lungs. HEK293, HeLa, RAW264.7 and U251 cells were treated with TNFα or LPS for 1 h. Subsequently, Western blotting was performed to detect the protein levels of HuR, HuB, HuC and HuD (A). Mice lungs were challenged with LPS (1 mg/kg) for 1 h. Then lung homogenates were prepared, Western blotting was performed to detect the protein levels of HuR, HuB, HuC and HuD (B). Quantification of the relative levels of HuR and HuB from the western blotting was achieved by normalization against that of β-actin. C Immunofluorescence analysis of the localization and expression of HuB in mice lungs tissue under inflammatory conditions. Mice lungs were challenged with LPS (1 mg/kg) for 1 h, Then frozen sections of mice lungs were prepared. Expression of HuB was detected by HuB antibody. Nuclei were counterstained with DAPI. Scale bar, 50 μm. D The Hel-N1 isoform of HuB is mainly expressed by brain tissues. Total RNA was extracted from various tissues and cells as indicated, and the purified RNA from each sample was transcribed to cDNA, and the HNS domain of HuB was amplified, The PCR products were sent for sequencing, and the information was compared with the NCBI database to identify the isoforms of HuB (Hel-N1 or Nel-N2). Shown is the diagram of the sequences of HuB mRNAs from the different tissues and cells. E Inflammation stimulation enhances the association of HuB with HuR. HEK293 cells were transfected with GFP-HuB plasmid, and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h. The cytosolic lysates were collected for co-IP assay by using GFP antibody, and then subjected to western blotting to detect the interaction of GFP-HuB with HuR. F ITC binding measurements of GST-HuR with GST-HuB. The purified recombinant proteins (GST and GST-HuR) were injected into calorimetric cells before GST-HuB injection. Theoretical titration curves as well as associated KD values were displayed. G, H Time dynamics of HuR shuttles to the cytoplasm in response to inflammatory stimulation. HEK293 cells were challenged with TNFα (10 ng/mL) for different time intervals, the cells were lysed as cytosolic extracts (CE) and nuclear extracts (NE), and the cytoplasmic location of HuR was analyzed by western blotting (G) and immunefluorescence staining (H) with HuR antibody. The profiles in the right panel (H) show the fluorescence intensity of HuR from line scans in the merged images in the left panel, analyzed by Image Pro Plus software (red line: HuR; blue line: DAPI). Scale bar, 20 μm. I Knockdown of HuB reduces the TNFα-induced cytoplasmic shuttling of HuR. HEK293 cells were transfected with siRNA targeting HuB or a control. After being mock-treated or exposed to TNFα (10 ng/mL) for 1 h, the CE and NE of cells were prepared, the efficacy of HuB knockdown was showed by western blotting with HuB antibody, and the shuttling of HuR was detected by western blotting with HuR antibody. J The ectopic over-expression of HuB but not HuR increases the cytoplasmic location of HuR. Flag-HuR or Flag-HuB together with GFP-HuR were overexpressed in HEK293 cells. The location of GFP-HuR was visualized with a *60 objective on a confocal microscope. Scale bar, 20 μm. The quantification from randomly three independent experiments was shown in the right panel. Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way ANOVA (A, B, E, J). n = 3 animals (B, C), n = 3 independent experiments (A, D–J), *p < 0.05, **p < 0.01, ***p < 0.001, ns: no significance. Source data are available in supplementary data 1 for this figure.
Next, the sub-cellular distribution of HuB was examined. Mice lungs were challenged with LPS, and HEK293 cells were incubated with TNFα. Immunofluorescence labeling assays revealed that proinflammatory stimuli induced a significant increase in HuB expression in cytoplasm (Fig. 1C; Fig. S1C). Unlike non-neuronal cells, HuB was mainly localized in the nucleus while it was also detectable in the cytoplasm in U87 neuronal cells (Fig. S1E). Human HuB exists in two variants: the long form (Hel-N1) which contains 13 additional amino acids compared to the short form (Hel-N2) located in HNS domain29. Hel-N2 is highly expressed in human brain tumor cells and oocytes, whereas Hel-N1 is dominant in normal adult brains30,31. Then we conducted an extensive investigation into which variant of HuB is expressed in HEK293, HeLa, U251 and Raw264.7 cell lines, as well as mice lung and brain tissues and human lung samples. The results indicated that all tested samples—except for mouse brain—expressed the short-form HuB (Hel-N2) (Fig. 1D and Fig. S1D). These findings suggested that HuB (Hel-N2) is induced by external signals and specifically localized within the cytoplasm of non-neuronal cells. In later description, all HuBs mentioned refer to the short-form (Hel-N2) ones.
HuB interacts with HuR and facilitates the cytoplasmic retention of the latter
The members of the Hu family are known for their interactions with one another, favoring the post-transcriptional regulation of ARE-containing mRNAs32,33,34. To investigate whether inflammatory stimulation promotes the interaction between HuB and HuR, HEK293 cells overexpressing GFP-HuB were incubated with or without TNFα for 1 h, followed by co-immunoprecipitation (co-IP) assay by using an anti-GFP antibody. Western blotting revealed a notable interaction between HuB and HuR in HEK293 cells, which was further enhanced by inflammatory stimulation (Fig. 1E). Moreover, HEK293 cells were treated with or without TNFα, followed by the performance of proximity ligation assay (PLA). Notably, the interaction between HuR and HuB was observed predominantly located in cytoplasm (Fig. S1F). Next, the physical interaction between HuR and HuB was investigated. Pull-down assays revealed that GST-HuB interacts with His-HuR and the heterogeneous interaction between HuR and HuB is stronger than the homogeneous dimerization of HuB (Fig. S1G). Domain mutants of GST-HuR were constructed, and the result indicated that the HNS domain of HuR is essential for its interaction with HuB (Fig. S1G). Furthermore, we conducted an isothermal titration calorimetry (ITC) analysis in vitro, which indicated the binding affinity (Kd) of two proteins was 8.89e-6 ± 1.76e-6M, confirming a substantial interaction of HuR/HuB (Fig. 1F).
In response to TNFα stimulation, the cytoplasmic HuR was increased, but there was no significant reduction in the abundance of nuclear HuR (Fig. 1G); correspondingly, the HuR level in whole cell was slightly upregulated (Fig. S1H). As revealed from the kinetics of cytoplasmic HuR distribution, HuR exists in the cytoplasm with a basal level even absent of TNFα stimulation, which may reflect a need of HuR in regulating physiological targets at posttranscriptional level. To clarify whether the increase of cytoplasmic HuR is due to its shuttling to the cytoplasm or its upregulated expression, protein synthesis inhibitor CHX was utilized, and the abundance of HuR at the whole-cell, cytoplasmic and nuclear levels was examined. The results showed that HuR was still significantly mounted in the cytoplasm upon inflammatory stimulation, indicating that HuR indeed shuttles into the cytoplasm (Fig. S1I). Then we speculate that the induced HuB/HuR interaction may facilitate the cytoplasmic redistribution of HuR. GFP-tagged HuB was overexpressed, and we found that expression of GFP-HuB increased the level of HuR in the cytoplasm without affecting the expression of HuR at the whole-cell level (Fig. S1J). Moreover, when cells were transfected with siRNA targeting HuB, the abundance of HuR at the whole-cell level remained almost unchanged (Fig. S1K), but the inhibition of HuR’s cytoplasmic shuttling was observed (Fig. 1I). Furthermore, GFP-HuR was co-expressed with either Flag-HuR or Flag-HuB. As illustrated in Fig. 1J, expression of Flag-HuB significantly elevated GFP-HuR signals in the cytosol even in the absence of inflammatory stimulation. Likewise, we also found that GFP-HuB promotes the endogenous HuR shuttling to the cytoplasm (Fig. S1L). Taken together, these results demonstrated that inflammatory stimulation enhances the expression of HuB and its interaction with HuR, which is essential for the cytoplasmic retention of the latter.
HuB interacts with HuR-associated targets and enhances the expression of proinflammatory genes
To gain further insight into the cooperative role of HuB with HuR in inflammatory responses, we investigated the impact of HuB on the binding of HuR to mRNAs. TNFα is one of the most investigated pro-inflammatory cytokine35, and CXCL1 and CXCL2 are well known C-X-C chemokine family members potent on neutrophil chemotaxis36. Thus, our studies have been focusing on the expression of these proinflammatory genes22,37. In the present study, HEK293 cells were transfected with siRNA targeting HuB and incubated with or without TNFα, followed by RNA-IP assays. The mRNA levels of HuR-associated targets (TNFa, CXCL1, and CXCL2) in TNFα-treated cells were significantly elevated compared to those in mock-treated cells; however, transient knockdown of HuB resulted in an impaired binding of HuR to these mRNAs (Fig. 2A). Conversely, overexpressed of HuB enhanced the association of HuR with its targets (Fig. S2A). These results demonstrated that HuB can effectively enhance the binding of HuR to mRNA targets.
A HuB enhances HuR binding with inflammatory genes’ mRNAs. HEK293 cells were transfered with siRNA targeting HuB or with control siRNA, and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h. The cytosolic lysates were collected and RNA-IP was conducted using HuR antibody. Half of the bead-antibody-protein/mRNA complexes were utilized for western blotting to assess equal loading/input of HuR, and the remaining half was subjected to real-time PCR to detect the levels of HuR antibody-precipitated TNFα, CXCL1, and CXCL2 mRNAs via normalization to the mRNA levels detected from the whole-cell lysates as the input. The levels of mRNAs precipitated from the group transfected with control siRNA and absent of TNFα treatment were set as 1, to normalize the mRNA levels of other samples. B Biotin-labeled tandem ARE repeat-containing RNA oligonucleotides. C Both HuB and HuR bind with ARE-containing RNA. GST, GST-HuB, and GST-HuR were incubated with biotin-labeled tandem ARE repeat-containing RNA oligos as indicated. The gel retardation assay was performed to detect the binding of the recombinant proteins to the probes. D Biotin-labeled TNFα mRNA’s 3’UTR ARE repeat-containing RNA oligonucleotides. E Both HuB and HuR bind with TNFα mRNA’s ARE. GST, GST-HuB and GST-HuR were purified and eluted, and then incubated with TNFα ARE. The gel retardation assay was performed to detect the binding of the recombinant proteins to probes. F HuR and HuB bind to the same RNA sequence as a single complex. GST-HuB and GST-HuR were purified and eluted, and then incubated with TNFα ARE. The gel retardation assay was performed to detect the binding of the recombinant proteins to probes. G, H HuB knockdown mitigates the expression of inflammatory genes. HEK293 cells were subjected to siRNAs targeting HuB and then mock treated or exposed to TNFα (10 ng/mL) for 1 h; Western blotting shows the efficacy of HuB knockdown; Real-time PCR was performed to detect the mRNA expression of TNFα, CXCL1 and CXCL2 (G). The level of TNFα protein was detected by western blotting. Quantitative assessment of the relative levels of TNFα and HuB was achieved via western blotting and normalized to that of β-actin (H). I HuB knockdown decreases proinflammatory mediators’ mRNA half-lives in TNFα exposed cells. HEK293 cells were subjected to siRNA targeting HuB followed by being mock-treated or exposed to TNFα (10 ng/mL) for 1 h to boost the transcription of proinflammatory mediators, and then subjected to transcriptional inhibition with or without TNFα maintenance for various times as indicated; Western blotting shows the efficacy of HuB knockdown. Quantitative assessment of the relative levels of HuB was achieved via normalization to that of β-actin. Real-time PCR was performed to assess the remaining mRNA levels of TNFα. Data were normalized to β-actin mRNA, and the relative amount of mRNA without Act D treatment was taken as 100%. Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way (A, G, and H) or two way (I) ANOVA test. n = 3 independent experiments (A–I), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Source data are available in supplementary data 1 for this figure.
Next, we performed RNA electrophoretic mobility shift assays (RNA-EMSA) utilizing recombinant proteins. GST-HuB or GST was incubated with biotin-labeled tandem ARE repeat-containing RNA oligonucleotides as previously described (Fig. 2B)22. The results showed that GST-HuB, but not GST, formed a complex with the probe (Fig. S2B). The introduction of titrated cold probes resulted in a dose-dependent competition (Fig. S2C), while the addition of an antibody led to a supershift of the complex (Fig. S2D), confirming the specificity of HuB’s binding. Moreover, the binding of HuR and HuB was compared. The result indicated that both HuB and HuR bound to the same ARE-containing probe, but HuB had a significantly increased affinity (Fig. 2C). To specify the function of HuB in regulation of mRNAs of pro-inflammatory genes, isometric RNA oligonucleotides containing three AREs present in the native 3′UTRs of TNFα mRNA were designed (Fig. 2D). We also compared the binding of HuB and HuR to TNFα 3’UTR ARE, it is worth noting that HuB formed more protein-RNA complex compared with HuR (Fig. 2E). Moreover, we performed RNA-EMSA by using equal molar amount of GST-HuR and GST-HuB, as well as GST-HuR/GST-HuB mixture with half molar amount of each. The results indicated that, HuB could form more complex with TNFα 3’UTR ARE, consistent with the data shown in Fig. 2C, E; however, HuB/HuR mixture yielded more protein/RNA complex than HuB alone (Fig. 2F). This result not only indicated HuR and HuB bind to the same RNA sequence as a single complex, but also suggested that the HuR/HuB/RNA complex is more stable, maybe due to that HuR and HuB are more potent to form heterodimer than homodimer of their own (Fig. S1G). Collectively, these data, combined with those findings indicating the induced heterodimerization between HuB and HuR as well as enhanced retention of HuR in the cytoplasm, suggested once the HuR-mRNA complex is transferred to the cytoplasm, HuB may efficiently engage with these mRNA targets.
To date, the role of HuB in proinflammatory gene expression has not been thoroughly investigated. To this end, HEK293 cells were transfected with siRNA targeting HuB and then stimulated with or without TNFα. Real-time PCR analysis revealed that the TNFα-induced expression of proinflammatory genes (TNFα, CXCL1, and CXCL2) was significantly reduced following HuB knockdown (Fig. 2G). Western blotting analysis using antibodies against TNFα demonstrated corresponding changes at protein levels (Fig. 2H). In contrast, overexpression of GFP-HuB significantly promoted the expression of these proinflammatory genes (Fig. S2E). In our previous study22, a number of inflammatory factors (e. g. Ccr4 and Ccr6) whose expression is not regulated by HuR. Interestingly, the present study revealed that CCR4 and CCR6 was not significantly reduced following HuB knockdown (Fig. S2F). Next, we measured the stability of TNFα mRNAs following HuB knockdown or overexpression. HEK293 cells were exposed to TNFα for 1 h, thereafter, a transcription inhibitor Act D was introduced with or without continued exposure to TNFα. The levels of remaining mRNAs were tested. In mock-treated cells, TNFα mRNA levels rapidly declined upon the addition of Act D. Whereas, HuB manifested the potential to improve the stability of TNFα mRNA (Fig. 2I and Fig. S2G). Collectively, these data suggest that HuB plays a crucial role in regulation of inflammatory gene expression at posttranscriptional level.
HuB serves as a bridge between HuR and the translation machinery
Given the proposed role of HuB in protein synthesis24, we prepared the HuB-associated immunoprecipitates from HEK293 cells and conducted mass spectrometry analysis38. We noticed that the HuB interactome contained the majority of small subunit (40S) ribosomal proteins, key translation initiation factors such as eIF4H and translation regulators such as RNA helicases DHX9 (Supplementary data 2: HuB interactome analysis by mass spectrometry-iBAQ values and Supplementary data 3: HuB interactome analysis by mass spectrometry-PeptideGroups), implying the involvement of HuB in translation and/or translation regulation. By immunoprecipitation-coupled western blotting analysis, we confirmed the binding of cytoplasmic HuB to the 40S ribosomal subunits (RPS) RPS3, DHX9 as well as eIF4H. RNase treatment was used to validate the presence of these direct protein interactions (Fig. 3A). To confirm the capability of DHX9 binding with HuB in vitro, recombinant GST and GST-HuB was purified and incubated with cell lysates derived from HEK293 cells for pull-down assays. The results supported a physical interaction between HuB and DHX9 (Fig. S3A). Then we investigated the distribution of HuB within polysome fractions in response to inflammatory stimulation. Firstly, using a ribosome protein (RPL7) from the large subunit to define where is located the 80S and the polysome, we showed the global effect of the ribosome profile fractionation. The sample treated with 50 mM EDTA was taken as a control39, which leads to the disassembly of 80S ribosomes and polysomes (Fig. S3B). Then siRNA targeting HuB was introduced into HEK293 cells followed by a treatment with TNFα for 1 h. Ribosome fraction separation was conducted, and western blotting showed that HuB, HuR and DHX9 were mainly presented in the translation initiating fractions but also distributed in polysomes, the knockdown of HuB hardly affects the assembly of polysomes, but resulted in a reduce of HuR and DHX9 from the translation initiating and polysome fractions in response to inflammatory stimuli (Fig. 3B). Moreover, real-time PCR was conducted, and the change in relative distribution of TNFα mRNA was analyzed and compared under different conditions. The result showed HuB depletion selectively decreased the translation of TNFα mRNA but not the translation of the control TUBB mRNA (Fig. 3C and Fig. S3C). This polysome fractionation analysis provided strong support for a role for HuB as a ribosome-associated protein, thereby modulating the translation efficiency of pro-inflammatory genes’ mRNAs.
A Confirmation of the enhanced association of HuB with the proteins revealed by proteomic mass spectrometry in response to inflammation stimulation. HEK293 cells were transfected with plasmid expressing GFP or GFP-HuB, and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h. The cytosolic lysates were collected for protein immunoprecipitation assay with GFP antibody and then subjected to western blotting to detect the interaction of GFP-HuB with DHX9, eIF4H as well as RPS3. B, C HuB facilitates the recruitment of DHX9, HuR and the associated mRNAs into polysome fractions. HEK293 cells were transfected with siRNA targeting HuB or control, and incubated with TNFα (10 ng/mL) for 1 h. Then the isolation of ribosomal fractions was undertaken, and the location of HuB, DHX9 and HuR in polysome fraction was analyzed by western blotting (B). Real-time PCR was performed to detect the mRNA levels of TNFα (C). Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way ANOVA test (A, C). n = 3 independent experiments, **p < 0.01, *p < 0.05. Source data are available in supplementary data 1 for this figure.
HuB is crucial for DHX9-mediated enhancement of the expression of pro-inflammatory genes
DHX9 has been recognized as a translation regulator through unwinding the secondary structure of 5’UTR40,41. However, its role in regulation of pro-inflammatory cytokines/chemokines’ expression has not been documented. To this end, firstly, we investigated the endogenous expression of DHX9 in HEK293 cells under inflammatory conditions; the results indicated that the level of DHX9 remained unchanged upon stimulation (Fig. 4A). DHX9 was predominantly localized within the nucleus, whereas also detectable in the cytoplasm (Fig. 4B). Subsequently, DHX9-knockdown was carried out, and HEK293 cells were treated with or without TNFα. The result of real-time PCR analysis revealed that TNFα-induced expression of CXCL2 was significantly reduced due to DHX9-knockdown (Fig. 4C); However, TNFα-induced upregulation of CCR4 and CCR6, did not significantly decrease by DHX9 silencing (Fig. S4A), in line with the effect of HuB knockdown (Fig. S2B). Given that DHX9 exists in the nucleus in a large quantity and the upregulation of inflammatory factors may be achieved through both transcriptional and posttranscriptional levels, dual-report assays were set up to detect the roles of DHX9 in activating promoter and resolving 5′UTR blockage. The results showed that, indeed, DHX9 improve TNFα-induced CXCL2 expression both through upregulating promoter activity and overcoming the potential 5′UTR-posed impeding (Fig. 4D, E, Fig. S4B, C), indicating that DHX9 implicates in expression of pro-inflammatory cytokine/chemokine through both transcriptional and posttranscriptional regulations.
A The expression of DHX9 in HEK293 cells. HEK293 cells were mock-treated or exposed to TNFα (10 ng/mL) for 1 h. Cell lysates were collected for western blotting with DHX9 antibody. B The localization of DHX9 in cells. HEK293 cells were mock-treated or exposed to TNFα (10 ng/mL) for 1 h. Then the cells were lysed by CE (left) or NE (right) buffer, and analyzed by western blotting with DHX9 antibody. C DHX9 knockdown down-regulates the expression of inflammatory genes. HEK293 cells were subjected to siRNAs targeting DHX9 and then mock treated or exposed to TNFα (10 ng/mL) for 1 h; Western blotting shows the efficacy of DHX9 knockdown and the protein level of TNFα; Real-time PCR was performed to detect the mRNA level of CXCL2. D Structural illustration of pGL3-Control and CXCL2-5′UTR reporters. E DHX9 knockdown suppresses the expression of the reporter gene with CXCL2-5′UTR. HEK293 cells were interfered with siRNA targeting DHX9 for 12 h, and then co-transfected with reporter plasmid expressing firefly luciferase driven by SV40 promoter and carrying CXCL2-5′UTR, as well as the control reporter plasmid Renilla, for another 12 h. After being mock-treated or exposed to TNFα (10 ng/mL) for 1 h, Luciferase activity was analyzed, the firefly luciferase level was normalized to Renilla (right). Western blotting shows the efficacy of DHX9 knockdown (left). F, G HuB assists in the binding of DHX9 to inflammatory gene mRNAs. HEK293 cells were transfered with GFP-HuB (F) or siRNA targeting HuB (G), and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h. The cytosolic lysates were collected, and then RNA-IP was conducted using DHX9 antibody. Half of the bead-antibody-protein/mRNA complexes was utilized for western blotting to assess equal loading/input of DHX9 and the remaining half was subjected to real-time PCR to detect the levels of the precipitated TNFα, CXCL1 and CXCL2 mRNAs via normalization to the mRNA levels detected from whole-cell lysates as the input. The levels of mRNA precipitated from the group transfected GFP (F) or siCTR (G) and absent of TNFα treatment were set as 1, to normalize the mRNA levels of other samples. Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way ANOVA test (A–C, E–G). *p < 0.05, **p < 0.01, ***p < 0.001. Source data are available in supplementary data 1 for this figure.
Since knockdown of HuB resulted in a reduce of DHX9 abundance in the translation initiating fractions and polysomes induced by inflammatory stimuli (Fig. 3); and HuB was mainly located in the cytoplasm (Fig. 1), we further investigated a function of HuB-DHX9 interaction in the cytoplasm in upregulating the expression of pro-inflammatory genes. The beneficial role of HuB in the binding of DHX9 to target mRNAs was examined. HEK293 cells overexpressing GFP-HuB were incubated with or without TNFα, followed by RNA-IP assays using DHX9 antibody. The levels of DHX9-associated mRNAs in TNFα-treated cells were significantly elevated compared to those in mock-treated cells, with a more pronounced effect observed in HuB-overexpressing cells (Fig. 4F). Conversely, transient knockdown of HuB resulted in a declined binding of DHX9 to the target mRNAs. (Fig. 4G). Collectively, data suggested that in the cytoplasm, DHX9 enhances the expression of pro-inflammatory genes through its interaction with HuB.
HuB promotes the translation from the mRNAs with stem-loop structure in 5′-UTR and ARE in 3′-UTR via cooperation with DHX9
Proinflammatory cytokine/chemokine mRNAs usually contain complex secondary structures such as stem-loop in their 5ʹ-UTR. Thus, to investigate the molecular mechanism by which HuB/DHX9 association promotes the translation of proinflammatory genes, firstly, a set of firefly (FF) luciferase reporters were utilized (with Renilla luciferase taken as an internal control). ATP5O-5’UTR-FF contains a short 5ʹ-UTR region of ATP5O (8 nt: GGGAGAGG) located upstream of the initiation codon of FF. A stable stem-loop structure was introduced in front of the initiation codon of FF of the plasmid ATP5O-5ʹUTR-FF, and the generated construct was named as ATP5O-5’UTR-SL-FF41. Thereafter, multiple copies of AUUUA elements were inserted into these two plasmids as the 3ʹ-UTR regions of the reporter gene FF (Fig. 5A). Through dual reporter assay, we confirmed that the secondary structure in the 5ʹ-UTR significantly inhibited the translation process. Nevertheless, the translation efficiency was increased upon TNFα stimulation (Fig. 5B). Next, HuB-deficient cells were incubated with or without TNFα. Dual reporter assay showed that HuB knock-down strongly repressed TNFα-induced increase in luciferase activity, in line with a significant decrease in the mRNA level of reporter gene (Fig. 5C); Likewise, DHX9 knock-down resulted in a similar effect as that of HuB (Fig. S5A). In contrast, overexpression of GFP-HuB significantly promoted the expression of the reporter gene with or without TNFα stimulation (Fig. S5B).
A The structural illustration of firefly (FF) luciferase reporters. ATP5O-5′UTR-FF-3 with a short 5′′UTR region of ATP5O in front of the initiation codon of FF, and multiple copies of AUUUA elements were constructed in 3′-UTR. A stable stem-loop structure was introduced in front of the initiation codon of FF of the plasmid ATP5O-5′UTR-FF, multiple copies of AUUUA elements were constructed in 3′-UTR, the generated construct was named ATP5O-5′UTR-SL-FF-3. B TNFα induces the firefly (FF) luciferase reporter gene expression. HEK293 cells were co-transfected with a Renilla (internal control) and ATP5O-5′UTR-FF-3 or ATP5O-5′UTR-SL-FF-3, and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h, Luciferase activity was analyzed by Luciferase assay, and the firefly luciferase level was normalized to that of Renilla. C HuB knockdown suppresses the expression of firefly (FF) luciferase with 5′-UTR stem-loop and 3’UTR ARE. HEK293 cells were interfered with siRNA targeting HuB, and then co-transfected with reporter plasmids ATP5O-5’UTR-SL-FF-3 and Renilla (internal control). The level of HuB protein was detected by western blotting. After being mock-treated or exposed to TNFα (10 ng/mL) for 1 h, luciferase activity was analyzed, and its mRNA levels were examined by real-time PCR. The levels of firefly luciferase activity and mRNA were normalized to that of Renilla. D Structural illustration of TNFα-5’UTR-Luc, TNFα-3′UTR-Luc, and TNFα-Luc reporters. E ARE enhances the expression of firefly (FF) luciferase. HEK293 cells were respectively transfected with plasmid ATP5O-5’UTR-FF-3, TNFα-5′UTR-Luc, TNFα-3′UTR-Luc, or TNFα-Luc (along with Renilla as internal control), and then exposed to TNFα (10 ng/mL) for 1 h. Luciferase activity was analyzed, the firefly luciferase level was normalized to that of Renilla. F HuB knockdown down-regulates the expression of TNFα-Luc. HEK293 cells were co-transfected with a Renilla (internal control) and TNFα-Luc, then transfered with siRNA targeting HuB, and then mock-treated or exposed to TNFα (10 ng/mL) for 1 h. The level of HuB protein was detected by western blotting. Luciferase activity was analyzed; the firefly luciferase level was normalized to that Renilla. G In vitro translation assay verifies the role of HuB in protein synthesis. In vitro translation assays were performed using the Flexi® Rabbit Reticulocyte Lysate System (Promega) according to the manufacturer’s instructions. TNFα-Luc luciferase mRNA was used as template as described in Materials and Methods. Different amounts of GST and GST-HuB were added to the assay mixture. The input of GST, GST-HuB used in in vitro translation assays was shown by coomassie staining (upper). The fluorescence signal was measured by Victor 3 microplate reader (lower). Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way ANOVA test (B, C, E–G). n = 3 independent experiments (B, C, E–G), *p < 0.05, **p < 0.01, ***p < 0.001. Source data are available in supplementary data 1 for this figure.
Additionally, we predicted the potential secondary structure of the 5′-UTR of pro-inflammatory cytokine/chemokine mRNAs via RNAfold web server (http://rna.tbi.univie.ac.at/RNAfold/bLLumHdDvY). The results indicated the formation of the putative stem-loop structures (Fig. S5C). Then, a set of reporter plasmids containing native elements from TNFα mRNA were constructed. TNFα-5’UTR-Luc represents that the 5’-UTR of TNFα was introduced in front of the initiation codon of FF; TNFα-3’UTR-Luc was constructed by inserting the 3′-UTR of TNFα following the stop codon of FF. TNFα-Luc contains both of these two regions from TNFα mRNA (Fig. 5D). Dual reporter assays demonstrated that in response to TNFα stimulation, TNFα 5′UTR region indeed constituted a translation barrier, whereas the efficiency of the translation from TNFα-Luc and TNFα-3’UTR-Luc mRNAs was significantly higher than that of ATP5O-5′UTR-FF-3, indicating the binding site of HuB (TNFα 3’UTR region) is essential for the translation (Fig. 5E). To investigate the impact of HuB on the translation process, HuB-deficient cells were incubated with or without TNFα. The depletion of HuB significantly inhibited the TNFα-induced increase in TNFα-Luc’s luciferase activity (Fig. 5F). On the other hand, overexpression of GFP-HuB markedly enhanced TNFα-Luc’s expression, regardless of TNFα stimulation (Fig. S5D).
To further elucidate the role of HuB and DHX9 in the protein synthesis process, the rabbit reticulocyte lysate system was utilized to perform the in vitro translation analysis. Recombinant proteins GST-HuB and His-DHX9A were respectively introduced into the rabbit reticulocyte lysate system. The addition of GST-HuB or His-DHX9A increased the translation efficiency from the TNFα-Luc’s mRNA that contained both 5’-UTR and 3’UTR regions of TNFα mRNA in a dose-dependent manner (Fig. 5G and Fig. S5E).
Cell-penetrating peptide TAT-HuR-HNS3 disrupt HuR/HuB interaction and mitigates LPS-induced pulmonary inflammation
HuR exploits its HNS domain to interact with HuB (Fig. S1G). Our previous studies demonstrated that HuR engages with itself or PARP1 through the C-terminal end of its HNS domain. The peptide TAT-HuR-HNS3 (YGRKKRRQRRRSPMGVDHMSGLSGVNVPGNASSG), derived from this C-terminal segment, disrupts the interactions between HuR and its partners42,43,44. Thus, we investigated the effect of this interfering peptide on the interaction between HuR and HuB. Recombinant GST or GST-HuB was purified and incubated with cell lysates obtained from HEK293 cells in the presence or absence of TAT-HuR-HNS3 (10 μM). Pull-down assays revealed that TAT-HuR-HNS3 effectively disrupted the interaction between HuR and HuB (Fig. 6A). Additionally, GFP-tagged HuR along with Flag-tagged HuB were overexpressed in HEK293 cells, which were subsequently treated with or without TAT-HuR-HNS3. The increase in cytosolic GFP-HuR induced by overexpression of HuB was abolished following treatment with the peptide (Fig. 6B). Moreover, we explored the impact of TAT-HuR-HNS3 on the associations of DHX9 with inflammatory genes’ mRNAs. HEK293 cells were pretreated with the peptide before stimulation with TNFα. Cell lysates were immunoprecipitated using antibodies against DHX9, followed by RNA-IP assays. Results indicated that challenge with TNFα significantly increased the engagement of DHX9 with TNFα, CXCL1 and CXCL2 mRNAs; however, this increase was suppressed by pretreatment with TAT-HuR-HNS3 (Fig. 6C). Thereafter, a mouse lung inflammation model was employed to assess the subepithelial accumulation of leukocytes within lung tissues. LPS exposure markedly induced neutrophil accumulation, which was mitigated by prior administration of TAT-HuR-HNS3 (Fig. 6D). These data with the findings from our previous studies substantiates the therapeutic potential of this cell-penetrating peptide to treat inflammation-related diseases.
A TAT-HuR-HNS3 inhibits HuB-HuR interaction. HEK293 cells were mock-treated or exposed to TNFα (10 ng/mL) for 1 h, GST and GST-HuB were incubated with equal amounts of whole-cell extracts from HEK293 cells in the presence of peptides TAT-HuR-HNS3 or not. The levels of pulled-down HuR were detected by western blotting. Quantitative assessment of the relative levels of HuR was achieved via normalization to that of GST. B TAT-HuR-HNS3 abolishes the increase in GFP-HuR in the cytosol induced by overexpression of HuB. HEK293 cells were transfected with a plasmid expressing Flag-HuB along with GFP-HuR, in the presence or absence of peptides HNS3. The cytoplasm location of HuB and HuR was analyzed by an anti-FLAG antibody or the GFP fluorescence. Nuclei were counterstained with DAPI. Scale bar, 20 μm. C TAT-HuR-HNS3 suppresses the association of RNA helicase with target mRNAs. HEK293 cells pretreated with TAT-HuR-HNS3 were challenged with TNFα (10 ng/mL), and 1 h later, the cytosolic lysates were collected, and then RNA-IP was conducted using DHX9 antibody. Half of the bead-antibody-protein/mRNA complexes was utilized for western blotting to assess the equal loading/input of DHX9 and the remaining half was subjected to real-time PCR to detect the levels of pulled-down TNFα, CXCL1 and CXCL2 mRNAs via normalization to the mRNA levels detected from whole-cell lysates as the input. The levels of mRNA precipitated from the group absent of TNFα treatment were set as 1, to normalize the mRNA levels of other samples. D TAT-HuR-HNS3 attenuates LPS-induced lung inflammation. Mice were pretreated (i.n.) with saline or TAT-HuR-HNS3 (100 ng per mice) for 1 h and then challenged (i.n.) with LPS (1 mg/kg) for another 12 h. Lung tissue sections were processed for staining with H&E (hematoxylin-eosin staining) to examine the subepithelial accumulation of leukocytes. Scale bar, 100 μm (above), 20 μm (below). Quantification shows mean ± SD based on three independent experiments, with significance of the difference determined by one-way ANOVA test (A, C). n = 3 independent experiments (A–C), n = 3 animals (D), **p < 0.01. Source data are available in supplementary data 1 for this figure.
Discussion
Numerous studies have elucidated the functions of HuR in various post-transcriptional processes, including polyadenylation, alternative splicing, nucleocytoplasmic shuttling, mRNA stability, and translation20,45,46; however, the roles and mechanisms of three other members of the ELAVL/Hu family expressed outside the nervous system remain significantly less understood17,18. The present study reveals that cytoplasm-localized HuB is induced in response to inflammatory stimuli to enhance translation from pro-inflammatory cytokine/chemokine mRNAs. Once redistributed to the cytoplasm, HuR associates with HuB, facilitating the delivery of pro-inflammatory gene mRNAs into the translation machinery; concurrently, RNA helicase DHX9 is recruited by HuB to unwind secondary structures within the 5'UTRs of these target mRNAs. Consequently, a prompt translation of cytokines and chemokines is achieved.
Among the Hu family of proteins, to the best of our knowledge, HuB was first identified as a trans-factor that binds to AU-rich elements (ARE) in mRNA 3’-UTRs. However, research on HuB has historically been limited compared to that on HuR. This study demonstrated that HuB is expressed across various non-neuronal cell types and its expression is upregulated upon stimulation (Fig. 1). Notably, two variant forms of HuB exist: the long form (Hel-N1), which contains 13 additional amino acids compared to the short form (Hel-N2) between the second and third RNA recognition motifs (RRMs). This suggests potential functional differences between Hel-N1 and Hel-N229,31. Previous studies have documented that Hel-N2 is highly expressed in human brain tumor cells while Hel-N1 predominates in normal adult brains30. Our data indicate that both neuronal and non-neuronal cell lines express short-form HuB (Hel-N2), whereas brain tissue predominantly expresses long-form HuB (Hel-N1). In neuronal cells, constitutively expressed HuB primarily localizes within the nucleus (Fig. S1). We propose that, unlike neural cells, most other tissue cells may preferentially express the short form of HuB (Hel-N2)—which could function as an induced protein responsive to processes such as cell differentiation, inflammatory responses, and tumor development. The current study elucidates how inflammation-related gene expression is significantly enhanced by HuB (Hel-N2). In comparison to HuR, HuB (Hel-N2) demonstrates a stronger association with target mRNAs (Fig. 2). Furthermore, HuB (Hel-N2) interacts with components of the translation machinery and RNA helicase DHX9 (Fig. 3). These findings suggest a distinct role for HuB (Hel-N2) in regulating pro-inflammatory cytokine/chemokine expression at the translational level.
HuR relies on its nucleoplasmic shuttling to execute its functions47. To date, serine phosphorylation within the HNS has been identified as a key mechanism governing the nucleocytoplasmic shuttling of HuR48,49. Additionally, methylation and ubiquitination have also been shown to influence the cellular localization of HuR50,51. Our previous studies demonstrated that PARylation of HuR by PARP1 enhances its cytoplasmic localization, ultimately stabilizing mRNAs encoding pro-inflammatory cytokines and chemokines22. In this study, we further clarify that beyond post-translational modifications, the interaction between HuB (Hel-N2) and HuR plays a critical role in determining the cytoplasmic distribution of HuR (Fig. 1). Members of the Hu family are known to interact with one another; this may facilitate their functional roles. While it has been observed that HuD can form dimers and trimers, no complex formation was detected for HuC in mammalian neuronal cells32,33. Both our studies and those conducted by others have confirmed that HuR forms homodimers or heterodimers with HuB26,52. This study provides further validation that the functions of members within the Hu family are coordinated and synergistic during the post-transcriptional regulation of pro-inflammatory cytokine/chemokine expression.
DHX9, also known as nuclear DNA helicase II (NDH II) or RNA helicase A (RHA), is an NTP-dependent helicase capable of unwinding both RNA and DNA, as well as aberrant polynucleotide structures53. In the nucleus, DHX9 interacts with numerous proteins involved in DNA replication and responses to DNA damage, thereby promoting DNA replication and maintaining genomic stability54. Additionally, DHX9 plays a role in nuclear factor-κB (NF-κB)-mediated transcriptional activation and enhances transcription in response to various inductive stimuli55. In contrast, the functional research of DHX9 in the cytoplasm is relatively scarce but is suggested to participate in translation regulation in a target- or structure-specific manner56,57. The translation efficiency of human mRNAs is affected by the folding/unfolding dynamics of rG4s within their 5’-UTRs, which depend on both DHX36 and DHX958. Our current study reveals that DHX9 is implicated in the expression of pro-inflammatory cytokines/chemokines through both transcriptional and post-transcriptional regulation. While the role of DHX9 in the transcriptional regulation of pro-inflammatory genes remains a conundrum requiring further investigation, our present study elucidates the mechanism by which DHX9 promotes translation of pro-inflammatory cytokines/chemokines. In response to inflammatory stimulation, HuB directs DHX9 to unwind the stem-loop structure in 5’-UTR of pro-inflammatory cytokine/chemokine mRNAs (Figs. 4, 5). Recently, a study unveiled a function of DHX9 in resolving RNA damage by forming stress granules (DHX9-SGs), which induce dsRNA-related immune responses and translation shutdown to promote the survival of newborn daughter cells. Typically localized in the nucleus, DHX9 rapidly aggregates into droplets within the cytoplasm following cell division when two daughter cells are formed59. Our immunofluorescence experiments further demonstrated that DHX9 did not aggregate into droplets upon stimulation with TNFα (Fig. S4D). This observation may be attributed to the less damaging conditions present in this study, which elicit a different functional response from DHX9.
Interfering peptides (IPs) have emerged as valuable agents specifically targeting protein interactions and are garnering increasing attention60. Our previous research established that HuR-HNS3, derived from the C-terminus (the final one-third) of the HuR-HNS domain, effectively blocks the interaction between PARP1 and HuR while inhibiting PARylation of HuR. This action subsequently reduces cytoplasmic redistribution of HuR and destabilizes pro-inflammatory gene mRNA44. Furthermore, this peptide interferes with HuR dimerization/oligomerization, thereby facilitating Argonaute 2-based miRNA-induced silencing complex binding to target mRNAs and decreasing stability of pro-inflammatory cytokine/chemokine mRNAs52,61. In this study, we further elucidated that HNS is utilized by HuR to interact with HuB. The peptide HuR-HNS3 obstructed the formation of the HuB/HuR heterodimer, impairing cytoplasmic localization of HuR and reducing binding affinity between RNA helicase DHX9 and HuR-associated mRNAs (Fig. 6). Consequently, TAT-HuR-HNS3 holds potential for suppressing expression levels of pro-inflammatory cytokines/chemokines associated with HuR through multiple mechanisms. This suggests a promising avenue for employing this peptide in clinical treatment.
In summary, our current study has revealed a role of the short-form HuB (Hel-N2) in non-neuronal cells. HuB (Hel-N2) is induced by stress stimuli, such as inflammatory agents, and is localized in the cytosol. In this context, it serves as a hub that connects HuR-associated pro-inflammatory cytokine and chemokine mRNAs with RNA helicase DHX9, thereby facilitating an efficient translation. Our findings also suggest a potential therapeutic strategy for diseases closely linked to increased mRNA stability and translation—such as inflammation-related disorders—through the inhibition of HuB/HuR dimerization via interfering peptides.
Materials and methods
Antibodies and reagent
Polyclonal antibody against HuB (14008-1-AP) (1:3000 for WB, 1:1000 for IP and 1:300 for IF), monoclonal antibody against LaminB (66095) (1:2000 for WB), monoclonal antibody against TNFα (60291-1) (1:3000 for WB), polyclonal antibody against HuC (55047-1-AP) (1:3000 for WB), monoclonal antibody against HuD (67835-1-Ig) (1:3000 for WB), polyclonal antibody against DHX9 (17721-1-AP) (1:3000 for WB, 1:300 for IF and 1:1000 for IP), polyclonal antibody against HuR (11910-1-AP) (1:3000 for WB, 1:100 for IP and 1:200 for IF), polyclonal antibody against RPS3 (11990-1-AP) (1:3000 for WB), polyclonal antibody against EIF4H (11012-1-AP) (1:3000 for WB) were from Proteintech (Wuhan, China). Monoclonal antibodies against Ribosomal Protein S6 (RPS6) (sc-74459) (1:1000 for WB) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-β-Tubulin (HC101) (1:8000 for WB), anti-GFP (HT801) (1:8000 for WB and 1:1000 for IP), anti-GST (HT601) (1:8000 for WB), anti-His (HT501) (1:8000 for WB) and anti-β-Actin (HC201) (1:8000 for WB) mouse monoclonal antibodies were purchased from TRANS (Beijing, China). In inflammatory stimulation experiments, the working concentration of recombinant human TNFα (Peprotech, 300-01 A) was 10 ng/mL, and that of Lipopolysaccharide (LPS, L2630, Sigma) was 500 ng/mL. Cell-penetrating peptides (CPPs) were purchased from DG peptides Co., Ltd (Hangzhou, China) and purified to >95%, and their identity was confirmed by mass spectrometry, Lipofectamine 2000 (Invitrogen) was used for transfection of plasmids, and Lipofectamine 3000 (Invitrogen) was used for transfection of siRNA.
Constructs
Plasmid GST-HuB was kindly provided by Dr. Hua Lou (Department of Genetics, Case Comprehensive Cancer Center, and Center for RNA Molecular Biology, School of Medicine, Case Western Reserve University). To construct GFP-HuB and His-HuB plasmids, the full-length HuB was cloned into the vector pGFP-C1 and PET30a. Plasmids GST and GST-HuR were kindly provided by Dr. Myriam Gorospe (Laboratory of Cellular and Molecular Biology; National Institute on Aging, National Institutes of Health, USA). To construct Flag-HuR and His-HuR plasmids, the full-length HuR was cloned into the vector pCMV-N-FLAG and PET30a. To construct reporter plasmid CXCL2-luc, the promoter of CXCL2 (NCBI Reference Sequence: NC_000004.12) was amplified by PCR using human genome as the template and was cloned into the vector pGL4.2. To construct reporter plasmid CXCL2-5′UTR, the 5′UTR of CXCL2 (NCBI Reference Sequence: NC_000004.12) was amplified by PCR using human genome as the template and was cloned into the vector pGL3-Control. His tag DHX9A contains the catalytic domain of DHX9 was a gift from Nicola Burgess-Brown (Addgene plasmid # 42456; http://n2t.net/addgene:42456 ; RRID:Addgene_42456), pCDNA3 ATP5O (5′UTR)-FF was a gift from Ivan Topisirovic (Addgene plasmid # 85486; http://n2t.net/addgene:85486; RRID:Addgene_85486), pCDNA3 ATP5O (5’UTR)-SL-FF was a gift from Ivan Topisirovic (Addgene plasmid # 85491 ; http://n2t.net/addgene:85491; RRID:Addgene_85491)62, ATP5O-5′UTR-FF-3 and ATP5O-5′UTR-SL-FF-3 were mutants various copies of AUUUA elements in 3’UTR developed from ATP5O-5’UTR-FF and ATP5O-5′UTR-SL-FF. TNFα-luc was insert TNFα-5′UTR (NCBI Reference Sequence: NC_000006.12) before ATP5O-5’UTR-FF’s Open Reading Frame (ORF) and TNFα-3’UTR (NCBI Reference Sequence: NC_000006.12) after ATP5O-5’UTR-FF’s ORF, TNFα-3’UTR-luc was insert TNFα-3′UTR after ATP5O-5′UTR-FF’s ORF, TNFα-5’UTR-luc was insert TNFα-5’UTR before ATP5O-5’UTR-FF’s ORF.
Small interfering RNAs
Small interfering RNAs (siRNAs) targeting human-HuR (5’-UGCCGUCACCAAUGUGAAAGU-3’), human HuB (#1: 5’-UUAUUGUUUUGGUUUGAAGUC-3’, #2: 5’-GACAGAGUACUGCAGGUCU-3’) and human DHX9 (#1: 5’-GGACUAGUAGCAACAUUGA-3’, #2: 5’-AUUCUAGCAUCAUCCCAGGTT)-3’ were commercially synthesized from Shanghai GenePharma were used at 100 nM. Cells were seeded in plates, incubated in a growth medium without antibiotics overnight, and then transiently transfected with RNA oligos using lipofectamine 3000 (Invitrogen) following the manufacturer’s instructions. At 4–6 h post-transfection, the medium was replaced with the fresh complete medium to promote recovery, after 48 h, cells were lysed as cytosolic extracts (CE), nuclear extracts (NE) and whole cell extracts (WE), for western blotting or immunoprecipitation.
Cell culture
HEK293 (Shanghai Institutes for Biological Sciences, CSTR:19375.09.3101HUMSCSP5209), RAW264.7 (Shanghai Institutes for Biological Sciences, CSTR:19375.09.3101MOUSCSP5036), HeLa (Shanghai Institutes for Biological Sciences, CSTR:19375.09.3101HUMSCSP5209), U251 (Shanghai Institutes for Biological Sciences, CSTR:19375.09.3101HUMTCHu58) and U87 (Shanghai Institutes for Biological Sciences, CSTR:19375.09.3101HUMSCSP5432) cells were grown in DMEM supplemented with 10% (v/v) FBS (HyClone), 10 mM glutamine, 100 IU/mL penicillin, and 100 mg/mL streptomycin. Cell lines STR (Short Tandem Repeat) identification has been conducted. Cells were regularly tested for mycoplasma contamination by mycoplasma detection kit.
Reverse transcription, electrophoresis PCR, and real-time PCR
Total RNA was isolated using the TRIzol reagent (15596-018, Invitrogen). Following this, cDNA was synthesized with the PrimeScript™RT reagent Kit plus gDNA Eraser (RR047A, Takara) according to manufacturer’s instruction. Real-time PCR was performed on the QuantStudio 3 Real-Time PCR Instrument (Applied Biosystems) with a TB Green® Premix Ex Taq™ (Tli RNaseH Plus) regent (RR420A, Takara). The mRNA expression levels of the target genes were normalized to the expression levels of the β-actin or Renilla gene. Data analysis was performed using the comparative cycling threshold method as previously described22. Quantitative PCR primers were as follows: Human β-Actin forward: 5’-CTCCATCCTGGCCTCGCTGT-3’, reverse: 5’-GCTGTCACCTTCACCGTTCC-3’, Human TNFα forward: 5’-TGCACTTTGGAGTGATCGG-3’, reverse: 5’-TCAGCTTGAGGGTTTGCTAC-3’, Human CXCL2 forward: 5’-CAAACCGAAGTCATAGCC-3’, reverse: 5’-GAACAGCCACCAATAAG-3’, Human CCR6 forward: 5’-GTTTTCAGCAATGCCACGTG-3’, reverse: 5’-CGGTAGTGTTCTGGATCGGA-3’, Human CCR4 forward: 5’-TGGTCCTCTTCCTTGGGTTC-3’, reverse: 5’-TGGGATTAAGGCAGCAGTGA-3’, Human TUBB forward: 5’-GGACCATGGACTCTGTTCGC-3’, reverse: 5’-AAGGAGAGTGCCCATTCCAGA-3’, and Firefly luciferase forward: 5’-GGTGGACATCACTTACGC-3’, reverse: 5’-CTCACGCAGGCAGTTCTA-3’, and Renilla luciferase forward: 5’-AGCCAGTAGCGCGGTGTATT-3’, reverse: 5’-TCAAGTAACCTATAAGAACCATTACCAGATT-3’ were purchased from Comate Bioscience Co.,Ltd (Jilin, China).
Quantitative real-time PCR was performed to detect mRNAs, The level of mRNA was expressed as a fold change using the △△Ct method. In addition to the real time qPCR, Regular PCR for amplification products was shown by agarose gel electrophoresis.
Identification of HuB variants
The human lung samples were provided by the Biobank of China-Japan Union Hospital of Jilin University (Changchun, China). The samples were the paracancerous tissues obtained from patients with small cell lung cancer during the surgery. The biological sample donors have signed an informed consent form and agreed to use the samples for relevant research. Experimental procedures, including RNA extraction and sequencing of the samples were approved by the Clinical Research Ethics Committee of China-Japan Union Hospital of Jilin University (Authorization number: 2024052303-1). All ethical regulations relevant to human research participants were followed.
HEK293, HeLa, U251, and Raw264.7 cell lines, as well as mice lung and brain tissues and human lung samples’ total RNA was extracted from various tissues and cells as indicated, and the purified RNA from each sample was transcribed to cDNA, and the HNS domain of HuB was amplified, Nucleotide sequence detection was conducted on the PCR products. (Comate Bioscience Co., Ltd (Jilin, China)). The nucleotide sequences of the samples were detected by Sanger sequencing, using the ABI 3700 automatic sequencer. The information was compared with the data provided by NCBI database to identify the isoforms of HuB (Hel-N1 or Nel-N2).
Measurement of mRNA stability
To measure the stability of the inflammatory mediator’s mRNA, a classical approach as previously described was applied22. Cells were exposed to TNFα (10 ng/mL) for 1 h for an immune boost, then the transcription inhibitor Act D (10 mg/mL) was added to the medium with or without the maintenance of TNFα for 0, 1, 2, 3, and 4 h. RNAs were isolated using TRIzol (15596-026, Invitrogen, Carlsbad, CA), and quantitative realtime PCR was performed using an Applied Biosystems thermocycler. The relative amount of mRNA without Act D treatment was taken as 100%.
Isothermal titration calorimetry (ITC) measurement
ITC assays were performed and carried out on a MicroCal PEAQ-ITC calorimeter (Malvern Paralytical) at 25 °C. PBS was used to dilute proteins. The concentration of recombinant GST-tagged HuR and GST were 20 μM, GST-HuB was diluted in PBS to 500 μM. The ITC experiments involved 19 injections of protein into protein to detect heat changes. A reference measurement (PBS) was carried out to compensate for the dilution of GST-HuB. Curve fitting to a single binding site model was performed using the ITC data analysis module of MicroCal Software provided by the manufacturer.
Cell fractionation
Whole cell lysis, as well as cytosolic and nuclear fractionation, were performed as described22. Briefly, Whole-cell lysates were prepared in Radioimmunoprecipitation assay (RIPA) buffer on ice for 30 min. Lysates were centrifuged and the supernatants were collected as the whole cell extract (WE). Cytoplasm and nuclear fractions were prepared by using the CelLytic NuCLEAR Extraction Kit (Sigma, NXTRACT, Saint Louis, MO). Cells were lysed with Cytosolic Lysis Buffer for 20 min, lysates were centrifuged (11000 × g, 1 min, 4 °C), and supernatants were collected as cytosolic extracts (CE). The pellets were washed twice with Cytosolic Lysis Buffer and lysed with Extraction Buffer. Nuclear lysates were clarified by centrifugation (21,000 × g, 5 min, 4 °C), and the supernatants were collected (nuclear extract, NE).
Western blotting and immunoprecipitation
Nonspecific rabbit or mouse IgG antibody was used as a negative control.
Cells were cultured and stimulated as described above and lysed. For western blotting, protein samples were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). The membranes were blocked with 5% nonfat milk for 1 h at room temperature before being incubated with primary antibodies overnight at 4 °C. Following this, Mouse- or rabbit-conjugated HRP second antibodies were added to the membranes for 50 min at room temperature, followed by washed 3 times with 1×Tris-buffered saline with Tween 20 (TBST) and visualized using Chemiluminescent HRP Substrate (Millipore) on Chemiluminescence Image System (Tanon Science & Technology Co., Ltd). The intensity of the bands was quantified using ImageJ software. To carry out immunoprecipitation, WE or CE extracts were incubated with GFP beads (Jilin Chenghe Trading Co., LTD, CH10001) for 3 h at 4 °C. The samples were further treated with 0.1 mg/mL RNaseA (MCE, USA) to remove RNA. The group transfected GFP and absent of TNFα treatment were used as a negative control. After washing by Radioimmunoprecipitation assay (RIPA) buffer three times, immunoprecipitated proteins were mixed with 1×loading buffer and boiled at 100 °C for 10 min followed by western blotting analysis. The antibody information is as shown above.
Mass spectrometry analysis of HuB interactome
HEK293 cells were cultured and stimulated as described above and lysed. Proteins complex (300 μg) immunoprecipitated with HuB antibody or IgG were subjected to SDS-PGAE and then in-gel protein digestion and peptides recovery conducted by Beijing Protein Inovation Co., Ltd. The solutions containing peptides released during in-gel digestion were measured using Nanoflow UPLC (Easy-nLC 1200 system, Thermo, Cambridge, MA, USA) coupled to Mass Spectrometer (Q Exactive™ Hybrid Quadrupole-Orbitrap™, Thermo). Briefly, the trypsinized peptides were firstly trapped and desalted on a 150 μm × 15 cm in-house made column packed with Acclaim PepMap RPLC C18 (1.9 μm, 100 Å, Dr. Maisch GmbH, Germany) with a pump system supplied moving phase A (0.1% formic acid in distilled water) and moving phase B (0.1% formic acid in acetonitrile). The profile of the gradient moving phase was as follows: 4–8% B for 2 min, from 8 to 40% B for 43 min, from 40 to 60% B for 10 min, from 60 to 95% B for 1 min and from 95 to 95% B for 10 min; and the flow rate was 600 nl/min. Different fractions of the eluate were injected into the Q-Exactive mass spectrometry set in a positive ion mode and the data dependent manner with a full MS scan from 300 to 1800 m/z. High collision energy dissociation was employed as the MS/MS acquisition method. The raw MS files were analyzed using MaxQuant (1.6.2.10) and deposited to protein database based (iProX, https://www.iprox.cn/) (Project ID: IPX0013585000, PXID: PXD068749). The mass spectrometry experiments were repeated two times.
Immunofluorescence microscopy
Cells were fixed with 10% (v/v) formaldehyde. After permeabilized with 0.5% (v/v) Triton X-100, cells were blocked with 2% (w/v) bovine serum albumin, and then incubated with primary antibodies recognizing HuB (1:300), HuR (1:200), DHX9 (1:200) or FLAG (1:200). Secondary antibodies (Goat Anti-Rabbit IgG (H&L) TRITC (CB417629683), Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd.) were used to detect primary antibody-antigen complexes with different color combinations as needed. The nuclei of the cells were stained with DAPI for 5 min. Images were acquired using a confocal microscope (LSM880, ZEISS, Germany).
Luciferase assay
Cells were seeded in the 24-well plates overnight in the absence of antibiotics. The cells were then respectively transfected with various firefly luciferase reporter plasmids together with renilla luciferase reporter plasmid (control) using Lipofectamine 2000 transfection reagent (Invitrogen). Cells were challenged with or without TNFα for 1 h, and then lysed in 100 μL lysis buffer, and the extracts (20 μL) were analyzed for luciferase activity using a Dual-Luciferase Reporter Assay kit (Promega). To further analyze the effects of the 3′-UTR on the expression of target genes, RNA was extracted and 1 U of DNase was used per 1 µg of RNA to eliminate plasmid DNA contamination. Firefly luciferase mRNA levels were measured by real-time PCR and calibrated to that of Renilla.
Recombinant protein purification
GST, GST target of HuB and HuR, His-HuB and His-DHX9A were expressed and purified in E. coli cells (Millipore) and induced with 0.6 mM IPTG at 16 °C overnight. Pelleted cells were re-suspended in lysis buffer (250 mM NaCl, 50 mM HEPES 7.5, 1 mM DTT, protease inhibitor). After sonication, lysates were centrifuged at 30,000 × g at 4 °C for 30 min. The supernatant (with 0.5% Nonidet P-40) was collected and incubated with balanced and reduced glutathione 4B (17075601, GE Healthcare Life Sciences, Uppsala, Sweden) or Ni-NAT agarose beads (30210, Qiagen, Hilden, Germany) on a rotator at 4 °C for 3 h or overnight. Beads were washed and eluted, and the puried recombinant proteins were confirmed by western blotting. For pull-down experiments, GST and GST-fused proteins immobilized on Glutathione Sepharose 4B were incubated with 1 mL of cell extract at 4 °C for 1–3 h. After three washes with Nonidet P-40 lysis buffer, the bound proteins were analysed by western blotting.
RNA-EMSA
To perform RNA EMSA and supershift analyses, a Chemiluminescent RNA EMSA Kit (20158, Thermo Fisher Scientific, Waltham, MA USA) was used. Briefly, GST or GST target of HuB and HuR proteins were purified and eluted in 100 μL buffer (50 mM Tris-HCl (pH > 8.0), 100 mM KCl and 40 mM glutathione). Recombined proteins were dissolved in the EMSA interaction buffer (3 mM MgCl2, 40 mM KCl, 5% glycerol, 2 mM DTT, 2 μg tRNA) and incubated with 20 μM of 5’-biotin-labeled RNA oligos for 30 min at room temperature. For supershift assays, 0.4 μg of specific antibodies or IgG were added to the mixture after 15 min of incubation at room temperature. The reaction mix was then loaded onto a 5% acrylamide native gel. RNA oligo probes utilized in the present study: AU-rich RNA oligo: 5′-AUUUAUUUAUUUAUUUAUUUAUUUA-3′ and TNFα ARE oligo: 5′-GAUUUUUAUUAUUUAUUUAUUAUUUAUUUAUUUAC-3′.
RNA immunoprecipitation
To isolate RNP complexes, a CE lysate from peptide-treated HEK293 cells, HuB knockdown HEK293 cells or GFP-HuB overexpressing HEK293 cells was precleared and then immunoprecipitated by using anti-HuB or anti-DHX9 antibodies coupled to Protein A/G Magnetic Beads (B23202, Selleck) overnight at 4 °C. Nonspecific rabbit IgG antibody was used as a negative control. After the beads were washed with NT2 buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1 mM MgCl2, and 0.05% NP-40), the mRNA was isolated by using an RNA sample kit (DP419, Tiangen, Beijing, China). The RNA levels in beads-Ab-protein/mRNA complex for each sample were assessed by reverse transcription coupled quantitative real-time PCR.
Polysome analysis
Polysome profiles were analyzed on sucrose gradients. Briefly, HEK293 cells with control or HuB knockdown were subjected to sucrose gradient fraction, cells were grown and treated as described above, and then dissolved in 1 mL extraction buffer (20 mM Tris pH 8.0, 140 mM KCl, 5 mM MgCl2, 0.5 mM DTT, 0.5% NP40 and 0.1 mg/mL cycloheximide) after washed twice with cold PBS containing 0.1 mg/mL cycloheximide. Extracts were clarified by centrifugation at 12,000 g for 10 min at 4 °C and then loaded onto 6 mL linear sucrose gradients (15–50%) prepared in extraction buffer. Polysomes were separated by centrifugation at 40,000 rpm/min for 2.5 h at 4 °C using a Beckman SW41 rotor. The gradients were collected into 500 μL fractions. For each sample, equal volumes of each fraction were pooled as follows: fractions 3 –6 for untranslated, fractions 7 − 10 for low translation and fractions 11 − 14 for high translation. By detecting OD260 absorption value, the absorption intensity of each scattered part is detected, and the absorption spectrum of the polysome is drawn. RNAs were isolated from polysome gradient fractions by using an RNA simple kit (DP419, Tiangen, China).
In vitro translation assay
In vitro translation assays were performed using the Flexi® Rabbit Reticulocyte Lysate System (Promega) according to the Manufacturer’s instructions. Briefly, indicated amounts of GST, GST-HuB or His-DHX9A (Catalytic domain containing DHX9) were added to the reaction mixture: Flexi® Rabbit Reticulocyte Lysate, 10 µM Amino Acid Mixture, RNasin® Ribonuclease Inhibitor, 7 mM KCl, 2 mM MgCl2, 1 mM DTT, 10 μg total cytoplasmic RNA (Cells were transfected with TNFα-Luc luciferase reporter plasmid using Lipofectamine 2000 (Invitrogen), total RNA was extracted using TRIzol reagent (Invitrogen)). The translation reaction proceeded at 30 °C for 90 min and the luciferase signal was recorded on a Victor 3 microplate reader.
Duolink (proximity ligation assay, PLA)
The proximity ligation assay (PLA) (Dolin In Situ Detection Reagents Red, DUO92008, sigma) was performed following the manufacturer’s instructions. Cells growing on slides were fixed with paraformaldehyde 4% and blocked for 1 h, then incubated with antibodies against HuB (Proteintech, 14008-1-AP) and HuR (Santa Cruz, 3A2, sc-5261) overnight at 4 °C. All samples were probed with anti-mouse MINUS and anti-rabbit PLUS Duolink secondaries. Duolink in situ red reagents were used according to the manufacturer’s protocol (Millipore Sigma) and mounted with Duolink DAPI mounting media, and generates a signal only when the two probes are in proximity. Images were acquired on a confocal microscope (ZEISS).
Mouse work
Six- to eight-week-old female C57BL/6 J mice (20–25 g) were purchased from Beijing HFK Bioscience Co., Ltd, China (ID: 91110000746555457 K). Mice were bred and cohoused 4–6 mice per cage in a specific pathogen-free facility at NENU (Changchun, Jilin, China) and allowed unlimited access to sterilized feed and water. They were maintained at 23 °C ± 1 °C and kept under a 12-h light/dark cycle. We have complied with all relevant ethical regulations for animal use. All mouse experiments were conducted in accordance with the protocols for animal use, treatment, and euthanasia approved by the Animal Care Committee of Northeast Normal University (Authorization number: 202302039). Following intraperitoneal administration of sodium pentobarbital (50 mg/kg) to anesthetize mice, LPS dissolved in normal saline (1 mg/mL) was administered to mice (1 mg/kg) via the intranasal route as described previously22,44. In some cases, TAT-HuR-HNS3 (1 mg/mL) was administered i.n. (100 ng per mice) 1 h before the challenge of LPS. To minimize potential confounders, administer the medication at fixed times of 9 a.m. each time, and each group should follow a fixed order of administration. After 12-h challenge of LPS, conscious mice were euthanized using the cervical dislocation method as per protocol. Subsequently, thoracic cavity was opened to extract lung tissue, which was either homogenized to prepare lysate or processed into frozen sections for further analysis. The experimental process adheres to the principle of blinding; individuals involved in group allocation are prohibited from participating in the conduction of the experiment, outcome assessment, and data analysis.
Statistics and reproducibility
All experiments were done at least three times independently confirming consistent results. The number used for each experiment is shown in figure legend. Animals and cells were randomized before treatments and no data was deliberately excluded from the analysis. Statistical analysis was performed using GraphPad Prism version 8 (GraphPad Software). Data are displayed as mean ± standard (SD). The p values were calculated from Student unpaired t test when comparing within two groups. One-way or two-way ANOVA were performed in the indicated figures whenmore than two groups were compared, n = 3 animals/independent experiments, The level of significance was accepted at * p < 0.05, ** p < 0.01, and *** p < 0.001, ns: no significance.
Data availability
Uncropped and unedited blots can be found in Supplementary Fig. 6 and source data for graphs can be found in Supplementary Data 1. Plasmids were uploaded in Addgene (ID:248280, 248281, 248279, 248282, 248283, 248284, 248285, 248286, 248287, 248288, 248289). The raw MS files were submitted to protein Database (iProX, https://www.iprox.cn/) (Project ID: IPX0013585000, PXID: PXD068749). All data generated or analyzed during this study are available from the lead contact on reasonable request.
References
Buccitelli, C. & Selbach, M. mRNAs, proteins and the emerging principles of gene expression control. Nat. Rev. Genet 21, 630–644 (2020).
Gehring, N. H., Wahle, E. & Fischer, U. Deciphering the mRNP Code: RNA-bound determinants of post-transcriptional gene regulation. Trends Biochem Sci. 42, 369–382 (2017).
Barreau, C., Paillard, L. & Osborne, H. B. AU-rich elements and associated factors: are there unifying principles?. Nucleic Acids Res. 33, 7138–7150 (2006).
Anderson, P. Post-transcriptional regulons coordinate the initiation and resolution of inflammation. Nat. Rev. Immunol. 10, 24–35 (2010).
Anderson, P. Post-transcriptional control of cytokine production. Nat. Immunol. 9, 353–359 (2008).
Du, J., Du, Y., Chen, L. & Liu, H. IL-17 promotes melanoma through TRAF2 as a scaffold protein recruiting PIAS2 and ELAVL1 to induce EPHA5. Biochim Biophys. Acta Mol. Cell Res. 1870, 119547 (2023).
von Roretz, C., Di Marco, S., Mazroui, R. & Gallouzi, I. E. Turnover of AU-rich-containing mRNAs during stress: A matter of survival. Wiley interdisciplinary reviews. RNA 2, 336–347 (2011).
Chen, C. Y. et al. AU binding proteins. recruit the exosome to degrade ARE-Containing mRNAs. Cell 107, 451–464 (2001).
Ghosh, S. & Jacobson, A. RNA decay modulates gene expression and controls its fidelity. Wiley Interdiscip. Rev. RNA 1, 351–361 (2010).
Brennan, C. M. & Steitz, J. A. HuR and mRNA stability. Cell Mol. Life Sci. 58, 266–277 (2001).
Hinman, M. N. & Lou, H. Diverse molecular functions of Hu proteins. Cell Mol. Life Sci. 65, 3168–3181 (2008).
Schoenberg, D. R. & Maquat, L. E. Regulation of cytoplasmic mRNA decay. Nat. Rev. Genet 13, 246–259 (2012).
Suresh Babu, S., Joladarashi, D., Jeyabal, P., Thandavarayan, R. A. & Krishnamurthy, P. RNA-stabilizing proteins as molecular targets in cardiovascular pathologies. Trends Cardiovasc Med 25, 676–683 (2015).
Simone, L. E. & Keene, J. D. Mechanisms coordinating ELAV/Hu mRNA regulons. Curr. Opin. Genet. Dev. 23, 35–43 (2013).
Schultz, C. W., Preet, R., Dhir, T., Dixon, D. A. & Brody, J. R. Understanding and targeting the disease-related RNA binding protein human antigen R (HuR). Wiley interdisciplinary reviews. RNA 11, e1581 (2020).
Berto, S., Usui, N., Konopka, G. & Fogel, B. L. ELAVL2-regulated transcriptional and splicing networks in human neurons link neurodevelopment and autism. Hum. Mol. Genet 25, 2451–2464 (2016).
Kraushar, M. L. et al. Temporally defined neocortical translation and polysome assembly are determined by the RNA-binding protein Hu antigen R. Proc. Natl. Acad. Sci. USA 111, E3815–E3824 (2014).
Ince-Dunn, G. et al. Neuronal Elav-like (Hu) proteins regulate RNA splicing and abundance to control glutamate levels and neuronal excitability. Neuron 75, 1067–1080 (2012).
Ohtsuka, T., Yano, M. & Okano, H. Acute reduction of neuronal RNA binding Elavl2 protein and Gap43 mRNA in mouse hippocampus after kainic acid treatment. Biochem Biophys. Res Commun. 466, 46–51 (2015).
Grammatikakis, I., Abdelmohsen, K. & Gorospe, M. Posttranslational control of HuR function. Wiley. Interdisciplinary Rev. RNA. 8, 1372 (2017).
Majumder, M., Chakraborty, P., Mohan, S., Mehrotra, S. & Palanisamy, V. HuR as a molecular target for cancer therapeutics and immune-related disorders. Adv. Drug Deliv. Rev. 188, 114442 (2022).
Ke, Y. Erratum: PARP1 promotes gene expression at the post-transcriptional level by modulating the RNA-binding protein HuR. Nat. Commun. 8, 15191 (2017).
Chen, J. et al. Circular RNA circRHOBTB3 represses metastasis by regulating the HuR-mediated mRNA stability of PTBP1 in colorectal cancer. Theranostics 11, 7507–7526 (2021).
Gantt, K. R., Jain, R. G., Dudek, R. W. & Pekala, P. H. HuB localizes to polysomes and alters C/EBP-beta expression in 3T3-L1 adipocytes. Biochem. Biophys. Res. Commun. 313, 619–622 (2004).
Kato, Y. et al. ELAVL2-directed RNA regulatory network drives the formation of quiescent primordial follicles. EMBO Rep. 20, e48251 (2019).
Hatanaka, T., Higashino, F., Tei, K. & Yasuda, M. The neural ELAVL protein HuB enhances endogenous proto-oncogene activation. Biochem. Biophys. Res. Commun. 517, 330–337 (2019).
Brennan, S. E. et al. The mRNA-destabilizing protein tristetraprolin is suppressed in many cancers, altering tumorigenic phenotypes and patient prognosis. Cancer Res 69, 5168–5176 (2009).
D’Alessandro, V. et al. Molecular detection of neuron-specific ELAV-like-positive cells in the peripheral blood of patients with small-cell lung cancer. Cell Oncol. 30, 291–297 (2008).
Gao, F. B., Carson, C. C., Levine, T. & Keene, J. D. Selection of a subset of mRNAs from combinatorial 3’ untranslated region libraries using neuronal RNA-binding protein Hel-N1. Proc. Natl. Acad. Sci. USA 91, 11207–11211 (1994).
Okano, H. J. & Darnell, R. B. A hierarchy of Hu RNA binding proteins in developing and adult neurons. J. Neurosci. 17, 3024–3037 (1997).
Chalupnikova, K. et al. An oocyte-specific ELAVL2 isoform is a translational repressor ablated from meiotically competent antral oocytes[J]. Cell Cycle 13, 1187–1200 (2014).
Kasashima, K., Sakashita, E., Saito, K. & Sakamoto, H. Complex formation of the neuron-specific ELAV-like Hu RNA-binding proteins. Nucleic Acids Res. 30, 4519–4526 (2002).
Filippova, N. et al. Hu antigen R (HuR) multimerization contributes to glioma disease progression. The. J. Biol. Chem. 292, 16999–17010 (2017).
Song, X., Shi, X., Li, W., Zhang, F. & Cai, Z. The RNA-binding protein HuR in digestive system tumors. BioMed. Res. Int. 2020, 9656051 (2020).
Kalliolias, G. D. & Ivashkiv, L. B. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat. Rev. Rheumatol. 12, 49–62 (2016).
De Filippo, K. et al. Mast cell and macrophage chemokines CXCL1/CXCL2 control the early stage of neutrophil recruitment during tissue inflammation [J]. Blood 121, 4930–4937 (2013).
Xue, Y. et al. Epigenetic control of type III interferon expression by 8-oxoguanine and its reader 8-oxoguanine DNA glycosylase1. Front Immunol. 14, 1161160 (2023).
Zhang J. & Ba, X. Mass spectrometry files—Mass spectrometry analysis of HuB/ELAVL2 [Dataset]. iProX https://www.iprox.cn/page/project.html?id=IPX0013585000 (2025).
Shetty, S., Hofstetter, J., Battaglioni, S., Ritz, D. & Hall, M. N. TORC1 phosphorylates and inhibits the ribosome preservation factor Stm1 to activate dormant ribosomes. EMBO J. 42, e112344 (2023).
Gulliver, C., Hoffmann, R. & Baillie, G. S. The enigmatic helicase DHX9 and its association with the hallmarks of cancer. Future Sci. OA 7, 650 (2020).
Petersen, C. P., Bordeleau, M. E., Pelletier, J. & Sharp, P. A. Short RNAs repress translation after initiation in mammalian cells. Mol. Cell 21, 533–542 (2006).
Gump, J. M. & Dowdy, S. F. TAT transduction: The molecular mechanism and therapeutic prospects. Trends Mol. Med. 13, 443448 (2007).
Wadia, J. S. & Dowdy, S. F. Transmembrane delivery of protein and peptide drugs by TAT-mediated transduction in the treatment of cancer. Adv. Drug Deliv. Rev. 57, 579596 (2005).
Wang, K. et al. Cell-penetrating peptide TAT-HuR-HNS3 suppresses proinflammatory gene expression via competitively blocking interaction of HuR with its partners. J. Immunol. 208, 2376–2389 (2022).
Sorge, R. E. et al. Inhibition of the RNA regulator HuR by SRI-42127 attenuates neuropathic pain after nerve injury through suppression of neuroinflammatory responses. Neurotherapeutics 19, 1649–1661 (2022).
Radhakrishnan, A. & Green, R. Connections underlying translation and mRNA stability. J. Mol. Biol. 428, 3558–3564 (2016).
Fan, X. C. & Steitz, J. A. HNS, a nuclear-cytoplasmic shuttling sequence in HuR. Proc. Natl. Acad. Sci. USA 95, 15293–15298 (1998).
Kim, H. H., Yang, X., Kuwano, Y. & Gorospe, M. Modification at HuR(S242) alters HuR localization and proliferative influence. Cell Cycle 7, 3371–3377 (2008).
Kim, H. H. et al. Nuclear HuR accumulation through phosphorylation by Cdk1. Genes Dev. 22, 1804–1815 (2008).
Vigouroux, C. et al. Methyl(R217)HuR and MCM6 are inversely correlated and are prognostic markers in non small cell lung carcinoma. Lung Cancer 89, 189–196 (2015).
Lucchesi, C., Sheikh, M. S. & Huang, Y. Negative regulation of RNA-binding protein HuR by tumor-suppressor ECRG2. Oncogene 35, 2565–2573 (2016).
Ke, Y. et al. Poly(ADP-ribosyl)ation enhances HuR oligomerization and contributes to pro-inflammatory gene mRNA stabilization. Cell Mol. Life Sci. 78, 1817–1835 (2021).
Lee, T. & Pelletier, J. The biology of DHX9 and its potential as a therapeutic target. Oncotarget 7, 42716–42739 (2016).
Powell, S. N. & Kachnic, L. A. Roles of BRCA1 and BRCA2 in homologous recombination, DNA replication fidelity and the cellular response to ionizing radiation. Oncogene 22, 5784–5791 (2003).
Tetsuka, T. et al. RNA helicase A interacts with nuclear factor kappaB p65 and functions as a transcriptional coactivator. Eur. J. Biochem 271, 3741–3751 (2004).
Shen, L. & Pelletier, J. General and target-specific DExD/H RNA helicases in eukaryotic translation initiation. Int. J. Mol. Sci. 21, 4402 (2020).
Hartman, T. R. et al. RNA helicase A is necessary for translation of selected messenger RNAs. Nat. Struct. Mol. Biol. 13, 509–516 (2006).
Murat, P. et al. Correction to: RNA G-quadruplexes at upstream open reading frames cause DHX36- and DHX9-dependent translation of human mRNAs. Genome Biol. 20, 124 (2019).
Zhou, Y. et al. RNA damage compartmentalization by DHX9 stress granules. Cell 187, 1701–1718.e28 (2024).
Bruzzoni-Giovanelli, H. et al. Interfering peptides targeting protein-protein interactions: the next generation of drugs?. Drug Discov. Today 23, 272285 (2018).
Bisogno, L. S. & Keene, J. D. RNA regulons in cancer and inflammation. Curr. Opin. Genet Dev. 48, 97–103 (2018).
Gandin V, et al. Nano CAGE reveals 5’ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs.nano CAGE reveals 5’ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs.nano CAGE reveals 5’ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs.nano CAGE reveals 5’ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs. Genome Res. 26(5):636-48 (2016).
Acknowledgements
We appreciate Dr. Hua Lou (Department of Genetics, Case Comprehensive Cancer Center, and Center for RNA Molecular Biology, School of Medicine, Case Western Reserve University) for kindly provided Plasmid GST-HuB. We appreciate Dr. Myriam Gorospe (Laboratory of Cellular and Molecular Biology; National Institute on Aging, National Institutes of Health, USA) for kindly provided Plasmid GST and GST-HuR. We also thank the following funding sources for their support: National Natural Science Foundation of China 32170724 and 31801182 to Y. K., and 32170591 to X. B., Provincial Science Foundation of Jilin, China 20210101356JC to J. W.
Author information
Authors and Affiliations
Contributions
J.Z., Y.T., X.Z., and X.H. performed biochemical and cellular experiments. X.H. and J.W. performed animal experiments. Y.K. and M.W. discussed the interpretation of the results. X.B. and J.Z. wrote the paper.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Communications Biology thanks Benjamin Weiss and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Tuan Anh Nguyen & Rosie Bunton-Stasyshyn.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.
About this article
Cite this article
Zhang, J., Huang, X., Tang, Y. et al. HuB serves as the central hub for connecting HuR-related mRNAs with translation machinery in inflammation. Commun Biol 9, 3 (2026). https://doi.org/10.1038/s42003-025-09228-9
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s42003-025-09228-9
