Phosphorylation-dependent remodeling of the XIAP IRES by hnRNPA1

Phosphorylation-dependent remodeling of the XIAP IRES by hnRNPA1 | Research Square Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Article Phosphorylation-dependent remodeling of the XIAP IRES by hnRNPA1 Sayan Das*, Louise Dunnett*, Hayden Fisher, Vincenzo Venditti, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8543209/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract The heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) is a ubiquitously expressed RNA-binding protein with essential roles in splicing, mRNA stability, and translation. Its activity as an internal ribosome entry site (IRES) trans-acting factor (ITAF) is particularly relevant in stress adaptation and cancer, where dysregulated IRES-mediated translation promotes cell survival and therapy resistance. In small-cell lung cancer (SCLC), FGF-2 signalling activates S6K2-dependent phosphorylation of hnRNPA1 at serines 4 and 6, selectively enhancing expression of the anti-apoptotic XIAP and Bcl-xL protein. Here, we combine quantitative binding assays, X-ray crystallography, NMR spectroscopy, and multi-microsecond molecular dynamics (MD) simulations to define how phosphorylation modulates hnRNPA1–XIAP IRES interactions. We show that phosphorylation confers RNA- and sequence-specific recognition, with binding resolving into two cooperative interactions of distinct affinities at the RRM1 and RRM2 domains. This behaviour is consistent with phosphorylation-enhanced RNA melting activity that exposes otherwise inaccessible motifs. Structural and spectroscopic analyses reveal that phosphorylation does not induce structural rearrangements but perturbs the conformational ensemble of the intrinsically disordered (IDR) N-terminal tail, reshaping transient intramolecular contacts with the RRM domains. Our findings reveal that fine-tuning of IDR conformational dynamics is a key component of RRM-mediated RNA recognition, coupling post-translational regulation of RNA-binding proteins to translational control. *Sayan Das & Louise Dunnett contributed equally. Biological sciences/Biochemistry/Proteins/RNA-binding proteins Biological sciences/Structural biology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) is one of the most abundant and ubiquitously expressed proteins of the hnRNP family, a diverse group of DNA- and RNA-binding proteins involved in multiple stages of nucleic acid metabolism 1 . HnRNPA1 plays a critical role in post-transcriptional gene regulation, including the processing of microRNA precursors, constitutive and alternative splicing, mRNA stability, and internal ribosome entry site (IRES)-mediated translation 2 , 3 . Although predominantly localized in the nucleus, hnRNPA1 shuttles between the nucleus and cytoplasm in association with its mRNA cargo in response to specific cellular signals 4 . HnRNPA1 binds single-stranded RNA (ssRNA) and has been shown to also interact with telomeric ssDNA and unwind G-quadruplex structures of telomere, crucial for maintenance of telomere length 5 . HnRNPA1 exerts these diverse functions through a modular domain architecture composed of an N-terminal RNA-binding domain (RBD) and a C-terminal auxiliary domain. The RBD, called Unwinding Protein 1 (UP1), comprises two RNA recognition motifs (RRM1 and RRM2) arranged in an antiparallel orientation 6 , 7 . Although the two RRMs have a highly conserved fold, they are non-redundant and functionally non-equivalent 8 . Structural studies of UP1 bound to RNA/DNA have revealed a preferential binding polarity, with the 5’ end of the nucleic acid oriented toward the C-terminus and the 3’ end toward the N-terminus of the RRM domains 9 , 10 , 11 , 12 . The C-terminal low-complexity domain (LCD), also referred to as a prion-like domain (PrLD), is a glycine-rich (Gly-rich) intrinsically disordered region (IDR) 13 . It facilitates both protein interactions and non-sequence-specific nucleic acid binding via an RGG box. The LCD also contains the M9 motif, a non-classical nuclear localization signal, essential for hnRNPA1 nucleocytoplasmic shuttling 2 . Under physiological and pathological cellular stresses (i.e., hypoxia, nutrient deprivation, mitosis, DNA damage, and cellular differentiation) that impair canonical cap-dependent translation initiation, hnRNPA1 functions as an IRES trans-acting factor (ITAF) by binding to IRES elements in mRNAs and regulating their translation 14 . Cellular IRESs are cis-acting translational regulatory elements located in the 5′ untranslated regions (UTRs) of a subset of eukaryotic mRNAs 15 . Interestingly, it’s been shown that these mRNAs are capped and ITAFs play a key role in mediating the switch between canonical cap-dependent translation to IRES-dependent (cap-independent) translation initiation upon cellular stress 16 . These IRES-containing mRNAs often encode proteins involved in critical processes such as cell survival, proliferation, apoptosis, and angiogenesis, all of which are closely associated with tumorigenesis and cancer progression 17 . Consistently, dysregulation of hnRNPA1 ITAF activity has been directly linked to cancer progression 18 and resistance to therapy 19 , 20 , 21 , 22 . However, the precise molecular mechanisms underlying hnRNPA1’s function as an ITAF remain incompletely understood. Emerging evidence indicates that different post-translational modifications (PTMs) can modulate its IRES-binding activity 22 , 23 , 24 . Specifically, it has been shown that in Small Cell Lung Cancer (SCLC)—an aggressive and rapidly progressing subtype of lung cancer 25 —hnRNPA1 phosphorylation downstream of fibroblast growth factor 2 (FGF-2) signalling modulate its ITAF activity. FGF-2 trigger activation of the p70 ribosomal S6 kinase 2 (S6K2) 26 , which in turn phosphorylates hnRNPA1 on Serine 4 (S4) and S6 22 . These phosphorylations enhance hnRNPA1's ability to selectively bind, within the nucleus, to the IRES elements of B-cell lymphoma-extra-large (BCL-xL) and X-linked inhibitor of apoptosis (XIAP) mRNAs 22 . Phosphorylated hnRNPA1 mediates the nuclear export of these mRNAs, leading to increased expression of XIAP and Bcl-xL proteins. This cascade suppresses apoptosis, thereby recapitulating the pro-survival effects of FGF-2 signalling in SCLC 22 . Several attempts to reconcile the idiosyncratic mechanisms of hnRNPA1 have elegantly shown that its two RRMs are thermodynamically coupled, and that this interdomain communication is essential for allosteric and cooperative RNA binding 11 , 12 , 27 . However, our understanding of how phosphorylation affects RNA/DNA-binding proteins within ribonucleoprotein complexes remains limited 28 . Phosphorylation events in flexible regions have been shown to trigger global conformational changes or disorder-to-order transitions, thereby altering affinity for oligonucleotides 29 , 30 . Alternatively, the charged and bulky phosphate groups can directly engage in interaction networks or modify RNA/DNA association 31 . Phosphorylation can also influence intracellular protein localization, affecting the pool of accessible oligonucleotides 32 . To investigate the role of hnRNPA1 phosphorylation in IRES binding activity, specifically focusing on the well-characterized XIAP IRES, we performed a series of binding affinity measurements. We found that phosphorylation at S4 and S6 confers RNA, and not DNA, sequence-specific binding and identified a region near the 5’ end of the XIAP IRES that is preferentially recognized by the phosphorylated protein. Nuclear magnetic resonance (NMR) and X-ray crystallography revealed that phosphorylations at S4 and S6 do not induce major conformational changes nor alter interdomain communication. Extended molecular dynamics (MD) simulations of hnRNPA1 in phosphorylated and non-phosphorylated states, both free and bound to XIAP IRES minimal binding sequence, showed that phosphorylation at S4 and S6 modulates the conformational space sampled by the IDR N-terminal tail (N-tail). These phosphorylation-induced changes may be linked to the ability of UP1 to destabilize RNA secondary structures, thereby positioning the IDR flanking the RRM as a central component of the RNA recognition. Methods Protein Expression and purification UP1 wild type (wt) (1-196) were subcloned from pET9d-hnRNP-A1 (AddGene #23026) into pETM-14 and subsequently overexpressed as an N-terminal (His)6-tagged fusion protein in BL21(DE3)OneShot E. coli cells (Invitrogen). Mutations (S4DS6D, S4E/S6E, F17A/F59A, F108A/F150A) were introduced by site directed mutagenesis. Cells were grown in LB media, induced with 0.5 mM IPTG at 0.6 OD and grown overnight at 20°C. After harvesting, cells were resuspended in 50 mM Tris-HCl pH 8.00, 300 mM NaCl, 0.01%(v/v) Triton X-100, 20 mM imidazole, 5%(v/v) glycerol supplemented with DNase and protease inhibitor (Roche) and lysed using a high pressure homogeniser (Avestin Emulsiflex C3). Proteins were purified via nickel affinity chromatography on His-Trap columns (Cytiva). Eluted proteins were incubated overnight with prescission protease for His-tag removal. Proteins were further purified by cation exchange chromatography using 20 mM MES-NaOH pH 6.00, 5 mM β-mercaptoethanol, 1 mM EDTA, 5%(v/v) glycerol and eluted with an NaCl gradient. Samples were then loaded onto a Superdex 75 16/60 gel filtration column (Cytiva) and eluted in 20 mM MES-NaOH pH 6.00, 150 mM NaCl, 5 mM β-mercaptoethanol, 1 mM EDTA, 5%(v/v) glycerol for crystallisation samples or in 20 mM sodium phosphate buffer pH 6.5, 100 mM NaCl, 5 mM β-mercaptoethanol for NMR and affinity measurements samples. Protein purity was checked by SDS–PAGE. Dianthus spectral shift binding assay DNA and RNA oligos were purchased labelled with Cyanine5 (Cy5) at the 5’ (Merk) and were resuspended in water to a final concentration of 100 µM. To measure affinity between UP1 (wt, S4DS6D, S4ES6E) and oligos, 12.5 nM Cy5-labeled oligos was mixed with proteins, using a 12 points two-fold serial dilution with a maximum protein concentration of 3.5 µM, in a 20 µL final volume in 384-well microplates (Nanotemper Technologies). Spectral shift measurements were carried out in triplicate on a Dianthus instrument at 25°C (Nanotemper Technologies). Data were analysed with GraphPad Prism (version 10.5.0) using a nonlinear regression sigmoidal function. Secondary structures in Cy5-labeled oligos was analysed via thermal unfolding using an Andromeda X (Nanotemper Technologies). Approximately, 3 µL of 50 nM oligo was loaded into high sensitivity capillaries and measurement were carried out in duplicate (Nanotemper Technologies). Data were analysed using Andromeda X analysis software (Nanotemper Technologies). Expression of N isotopically-labelled UP1 Proteins were expressed in a 15 N labelled minimal media (M9: 6.5 g/L Na 2 HPO 4 , 3 g/L KH 2 PO 4 , 0.5 g/L NaCl, 4 g/L Glucose, 1g/L 15 NH 4 Cl, 120 mg/L MgSO 4 , 11 mg/L CaCl 2 , and 10 mL/L of 100x MEM Vitamin mix (Gibco)), according to the previous published protocol 33 . Briefly, the M9 media was pre-warmed at 37°C prior to inoculation with 1%(v/v) pre-culture. Cells were induced with 0.5 mM IPTG at 0.6 OD and grown overnight at 20°C. Proteins were purified as described above. Crystallisation UP1 wt, UP1 S4ES6E and S4DS6D were crystallised as previously reported for UP1 34 and UP1 bound to telomeric DNA 9 . Briefly, the purified proteins were concentrated to 20 mg/mL and were crystallised in 24 well hanging drop plates. 3 µl UP1 20 mg/ml was mixed with 3 µl of precipitant solution (100 mM Tris-HCl pH 7.50, 25%(v/v) PEG-4000, 20%(v/v) MPD). For the co-crystals, UP1 20 mg/ml was incubated on ice for 30 min in presence of and equimolar amount of telomeric DNA (TTAGGGTTAGGG), before mixing 1 µl of protein-DNA with 1 µl of precipitant solution (100 mM Tris-HCl pH 7.50, 1.8 M dibasic ammonium phosphate, 15%(v/v) glycerol). The crystals were grown at 16°C by the hanging drop vapor diffusion method within 1 day. Data collection and structure determination For the 100 K data collection, the crystals were flash-cooled in liquid nitrogen. X-ray diffraction data were collected at the Diamond Light Source beamlines I02 and I24. The diffraction images were processed using xia 3dii or xia2 dials 35 , followed by AIMLESS 36 . The structures were determined by molecular replacement using PDB ID: 1L3K and 2UP1. and refined using PHENIX 1.20.1 37 , followed by model building using Coot 38 . Structure have been deposited on the PDB: 9F1S and 8RZV for UP1 S4ES6E alone and in complex with telomeric DNA, respectively, and 9GEA and 9GPJ for UP1 S4DS6D alone and in complex with telomeric DNA, respectively. Structure prediction XIAP IRES mRNA sequence was retrieved from the IRESite database: XIAP (-174) 5'-uauuuagaauuagaauguuucuuagcggucguguaguuaguuauuuuuaugucauaaguggauaauuuguuagcuccuauaacaaaagucuguugcu uguguuucacauucuggauuuccuaauauaauguucucuuuuuagaaaagguggacaaguccuauuuucaagagaag aug (+ 3) IRES mRNA secondary structure was predicted using MXfold2 web server 39 and visualised with R2R software 40 . The structure of UP1 in complex with the 16-mer (5’-UUAGCGGUCGUGUAGU-3’) was predicted using AlphaFold 3 41 . NMR spectroscopy All NMR spectra were acquired on Bruker 700 or Bruker 800 spectrometer equipped with z-shielded gradient triple resonance 5-mm TCI cryoprobe. NMR samples containing 0.6-1mM UP1 wt, UP1 S4DS6D or UP1 S4ES6E were prepared in 20mM Sodium Phosphate (pH 6.5), 100mM NaCl, 2mM DTT and 90% H 2 O/10% D 2 O buffer. Spectra were processed and analysed using NMRPipe 42 and POKY 43 , respectively. Backbone 1 H N , 15 N H , 13 C α , 13 C β and 13 C’ chemical shifts of UP1 S4DS6D were assigned using triple resonance correlation experiments (HNCO, HNCA, HN(CO)CA, HNCB, HN(CO)CACB). Out of the 196 residues (which includes 6 prolines), 180 residues were assigned (95% excluding prolines). The program MARS 44 was used to unravel the UP1 S4DS6D and UP1 S4ES6E TROSY peak assignment out of the experimental chemical shifts. The assignments were deposited on the BioMagResBank (accession code 53410 and 53413). The weighted combined 1 H/ 15 N chemical shift perturbations (CSP) resulting on the 1 H- 15 N TROSY spectra of UP1 wt from introduction of the phopshomimetic mutations were calculated using the following equation: $$\:CSP=\sqrt{{\left({\varDelta\:\delta\:}_{H}{W}_{H}\right)}^{2}+{\left({\varDelta\:\delta\:}_{N}{W}_{N}\right)}^{2}}$$ 1 , where W H (= 1) and W N (= 0.154) are the weighting factors for the 1 H and 15 N amide chemical shifts and ∆δ H and ∆δ N symbolize the 1 H and 15 N chemical shift differences in ppm between wild type and mutant UP1. 15 N longitudinal ( 15 N-R 1 ) and transverse ( 15 N-R 2 ) relaxation rates were obtained at 800 MHz and 25°C using 15 N-R 1 and R 1ρ experiments. The spin-lock field for the R 1ρ experiment was set to 1 kHz. Decay durations were 0, 80, 240, 400, 560, 720, 880 and 1080 ms for R 1 and 2, 45, 89, 178, 267, 356, 445, 534, and 800 ms for R 1ρ . Backbone amide 15 N Carr-Purcell-Meiboom-Gill (CPMG) relaxation dispersion experiments were carried out at 800 MHz and 25 o C using a pulse sequence that measures the exchange contribution for the TROSY component of the 15 N magnetization. Off-resonance effects and pulse imperfections were minimized using a four-pulse phase scheme. Experiments were performed with a fixed relaxation delay (20 ms) but a changing number of refocusing pulses to achieve different effective CPMG fields (50, 100, 150, 200, 250, 300, 350, 400, 450, 500, 600, 700, 800, 900 and 1000 Hz). Experimental errors on the transverse relaxation rates were estimated from the noise level estimated with the POKY software. Residue specific 15 N-Rex values were calculated by taking the difference between the R 2 values measured at CPMG fields of 50 and 1000 Hz, respectively. Molecular Dynamics (MD) Simulations 2µs simulations was performed by using Gromacs 2023 package 45 . The NMR solution structure (PDB code: 2LYV) 46 was used as a starting conformation. S4D and S6D mutations, and S4 and S6 phoshorylations were introduced using PyMOL. The protein topology was prepared with Amber03 force field 47 . The system was placed in a TIP3P water box with the distance from the surface of the water box to all the atoms of the solute is set to 10 Å. Counterions Na + and Cl − were added to neutralize the charge. Energy minimization was carried out using the steepest descent method with a step size of 0.01nm. Minimization was run until the maximum force fell below 1000 kJ/mol/nm for 50000 steps. The system equilibrated at 300 K for 100 ps with a timestep of 2 fs for 50000 steps and equilibrated at a constant pressure (1 atm) for 100 ps with a timestep of 2 fs for 50000 steps, both using the leap frog integrator. LINCS was used to constrain the covalent hydrogen bonds with a Verlet cutoff scheme with a 1.0 cutoff radius for neighbour list. The particle mesh Ewald scheme was used to treat long range interaction with a Fourier grid spacing of 0.16 nm. The modified Berendsen thermostat (velocity-rescale thermostat) and the Parrinello-Rahman barostat were used to maintain the temperatures at 300 K and 1 atm. 2 µs production MD simulation was carried out with a 2 fs timestep with 1000000000 steps. The leap frog integrator was used for integrating Newton’s equations of motion. LINCS was used to constrain the covalent hydrogen bonds with a Verlet cutoff scheme with a 1.0 cutoff radius for neighbour list. The particle mesh Ewald scheme was used to treat long range interaction with a Fourier grid spacing of 0.16nm. The modified Berendsen thermostat (velocity-rescale thermostat) and the Parrinello-Rahman barostat were used to maintain the temperatures at 300K and 1atm. Distance measurements were performed using the MDAnalysis 48 distance analysis tool. Analysis of the UP1-RNA contacts was performed using GetContacts ( https://getcontacts.github.io/))gith ub package. MD trajectory were visualised using PyMOL 49 . Results XIAP IRES structure Prediction Initially identified in viruses, IRESs are characterised by complex RNA structures that mediate ribosome recruitment to initiate translation in a cap-independent manner. Unlike viral IRESs, cellular IRESs lack conserved sequences or structural motifs and are not classified into distinct groups 50 . This has limited our ability to propose general mechanisms of IRES-mediated regulation, and cellular IRESs have thus far been characterised individually. Currently, no full-length 3D structures exist for the XIAP IRES (or any other cellular IRES), and available information has been derived from structural prediction tools 51 , 52 . To identify ssRNA regions likely to be recognised by hnRNPA1, we predicted the XIAP IRES secondary structure using MXfold2, a deep learning-based prediction tool that has been shown to achieve high accuracy for non-coding RNA structure predictions 39 . Since mRNAs fold co-transcriptionally in vivo 53 , we assumed that the XIAP IRES adopts its structure independently of the downstream coding region. Accordingly, and in contrast to previously published XIAP IRES models 51 , 52 , we predicted the secondary structure using the full-length XIAP mRNA sequence from positions − 174 to + 3 (start codon). In the resulting model (Fig. 1 ) it is possible to identify a multibranch loop (m1), three hairpin loops (h1–h3), five internal loops (i1–i5), and three bulge loops (b1–b3). To map hnRNPA1 binding sites on the XIAP IRES, we designed short oligonucleotides covering all identified loops and including all hnRNPA1 minimal binding motif, the 5’-AG-3’ dinucleotide (Fig. 1 ). UP1 binding site mapping on the XIAP IRES To map hnRNPA1 binding site on the XIAP IRES and identify difference in binding affinities between the phosphorylated and unphosphorylated protein, we performed spectral shift experiments (Fig. 2 A-D). We focused our studies on UP1, as hnRNPA1 LCD is unsuitable for structural studies and does not participate in the RNA-sequence-specific recognition of the XIAP IRES. Due to the challenges of producing large quantities of homogeneously phosphorylated protein in vitro , we used two UP1 phosphomimetic mutants, S4DS6D and S4ES6E, which have been shown to behave like the phosphorylated protein in cell 22 . Assuming the phosphorylations on S4 and S6 only affect hnRNPA1 ITAF activity, we used the human telomeric ssDNA repeats (TR2) as positive control (Fig. 2 A). In perfect agreement with Burd et al 54 , we measured low nM affinities for UP1 wild type (wt), which were identical to the dissociation constants (K D s) of S4DS6D and S4ES6E (Table 1), as expected based on our crystallographic data (here below). Previous studies reported that UP1 binds with similar modes and affinities both ssDNA and ssRNA 2 . As such, we conducted this initial mapping study using DNA oligos with an equivalent sequence to the XIAP IRES. To map UP1 binding site on XIAP IRES, we initially focused our search around the 5’ of the IRES which is predicted to have extended single stranded regions. Indeed, we found that both UP1 wt and phosphomimetic mutants recognise with similar affinities a region between nucleotides − 155 and − 112 (Fig. 2 B-D) (Table 1). This is in contrast with earlier binding data for hnRNPA1 wt and the XIAP IRES, which mapped the binding region towards the 3’ of the IRES, between nucleotides − 62 to − 34. However, in this study a substantially shorter IRES sequence (-100 to + 3) was used, which did not include the binding region we identified 55 . As a control, we also tested oligonucleotides encompassing all predicted shorter single-stranded regions of the XIAP IRES, confirming absence of additional recognised sites (Fig. S1 A-E) We also noticed our measured K D s were ~ 10-fold higher than the positive control and other previously reported cognate RNA sequences 12 , 56 , 57 . We therefore reasoned that phosphorylation at S4 and S6 may confer both RNA-specific and sequence-specific recognition, as RNA and DNA adopt distinct conformational landscapes 58 . We repeated the titration, this time using RNA oligos. We noticed no difference in affinity between the wt and the phosphomimetic mutants for the R4 oligo spanning region − 131 to -112 (Fig. S1 A), with K D s similar to those measured for the D4 DNA oligo (Table 2). On the contrary, a more complex binding scenario was observed for the oligo R3, which spans nucleotides − 154 to − 134 of the XIAP IRES. For UP1 wt, we detected one single predominant binding event with lower affinity (about 1.5-fold) compared to the D3 DNA oligo (Fig. 2 A). Differently, using the same protein concentration range as for the wt, we observed a more pronounced biphasic-binding curves for UP1 S4DS6D and S4ES6E (Fig. 2 A). Biphasic-binding curves are usually the result of two binding events on two binding sites with different binding affinities 59 . Since UP1 has two well characterised ribonucleoprotein consensus sequences (RNPs) on each RRM domain 3 , it is reasonable to assume that R3 binds with different affinities to the RRM1 and RRM2 domains of the phosphomimetic mutants. Interestingly, the K D s of the low-affinity R3 binding events (K D low ) in both UP1 S4DS6D and S4ES6E were comparable to that of the wt (Fig. 2 B). To characterise the high-affinity states more confidently, we repeated the titration experiments at lower protein concentrations (Fig. 2 C). The estimated K D high for UP1 S4DS6D and S4ES6E were 36.05 nM and 20.25 nM, respectively (Table 2), while, consistently with our previous titration, the signal to noise ratio for UP1 wt was too low for the data to be analysed. Although these K D s may not be measured accurately due to the inability to fully decouple the two binding events, both UP1 S4DS6D and S4ES6E bind the R3 oligo with ~ 10-fold higher affinity than the wt, and ~ 5-fold higher affinity than the equivalent D3 DNA oligo (Table 1). Taken together, these data demonstrate that phosphomimetic mutations not only enable hnRNPA1 to recognize a specific region of the XIAP IRES but also allow it to discriminate between RNA and DNA with equivalent sequences. Furthermore, our results indicate that the enhanced specificity for XIAP IRES is linked to a change in binding mode induced by the phosphomimetic mutations. UP1 identification of the minimal binding sequence A recent study by Jain et al. 56 showed that the presence of two closely spaced 5’-UAG-3’ motifs enhances the binding affinity of hnRNPA1. The R3 oligonucleotide (20-mer) contains three UAG motifs spaced by more than two nucleotides and corresponds to the only region within the XIAP IRES that harbours YAG motifs (where Y is a pyrimidine) in close proximity within single-stranded regions or loops. Two additional 5’-YAG-3’ motifs, separated by two or more nucleotides, are located between positions − 172 and − 159, a region predicted to be double-stranded (Fig. 1 ). This oligonucleotide binds the wt and phosphomimetic mutants with ~ 3 and ~ 30-fold lower affinity, respectively, than the 20-mer (Table 2) (Fig. S2B), suggesting that while the presence of two YAG motifs may be a minimal requirement for binding, additional factors or structural elements likely contribute to sequence-specific recognition. To identify the minimal UP1-binding sequence within the 20-mer and determine which UAG site is involved in binding, we generated a shorter 16-mer oligonucleotide that lacked the third UAG and contained only two UAG motifs separated by eight bases. Interestingly, wt and phosphomimetic UP1 mutants bind with similar low nM affinities (Fig. 2 D), comparable to the K D high estimated for the 20-mer (Table 2). We hypothesized that this difference may stem from distinct secondary structures between the 16-mer and 20-mer oligos. Indeed, the 20-mer spans a loop region within the XIAP IRES and is predicted to adopt a loop conformation when isolated (Fig. S3A). In contrast, the 16-mer is not predicted to form such a structure (Fig. S3B). Thus, we extrapolated presence of secondary structures in all oligos tested by measuring their melting temperatures (Tm) and confirmed that the 20-mer is the only oligo we tested with a secondary structure (Fig. S4). Together with the binding affinity data, these observations suggest that one UAG site on the 20-mer (the lower-affinity site) is accessible to both wt and phosphomimetic UP1. However, the phosphomimetic mutant can access the second UAG motif more easily, indicating that phosphorylation may enhance the RNA-unwinding activity of UP1. Previous studies on the loop region of pri-mir-18a suggested that hnRNPA1 has RNA melting properties. To further validate this model, we divided the 20-mer into two 10-mer oligos: R3-1, which contains the first UAG motif, and R3-2, which contains two UAG motifs separated only by a single base, so unable to bind both RRMs simultaneously. These 10-mers lack the ability to form stable secondary structures and separate the UAG motifs. Binding assays showed that R3-1 and R3-2 bind to S4DS6D and S4ES6E with affinities comparable to K D low and K D high estimated for the 20-mer, respectively (Table 1). As expected, in the absence of secondary structure constraints, all UAG sites are accessible, and the affinities measured for the 10-mers with UP1 wt match those observed with the phosphomimetic mutants (Fig. 2 E-F). This indicates that we successfully separated the two binding events observed during the 20-mer titration with UP1 phosphomimetic mutants, confirming the reliability of the K D estimates. Taken together, these data indicate that phosphorylation at S4 and S6 enhances UP1’s ability to remodel RNA secondary structure, thereby exposing otherwise inaccessible binding sites. To investigate the topology of RNA binding to the UP1 RRMs, we mutated conserved RNP residues in RRM1 (F17A/F59A; hereafter rrm1RRM2) and RRM2 (F108A/F150A; hereafter RRM1rrm2). These mutations have previously been shown to reduce, but not abolish, RNA binding 11 . This strategy was preferred over separating the two RRMs, as domain splitting has been shown to reduce RRM2 stability 27 . The mutants were then titrated with R3-1, R3-2, and the 20-mer RNA. The affinity for R3-1 decreased (over 2-fold) in the rrm1RRM2 mutant but was unchanged in the RRM1rrm2 mutant (Table 3). These results indicate that R3-1 primarily binds RRM1 and that RRM1 constitutes the lower-affinity binding site and, as a consequence, RRM2 represents the higher-affinity site, in agreement with previous studies 11 . We observed a more complex binding behaviour for R3-2. No binding was detected for the UP1 rrm1RRM2 mutant, whereas the phosphomimetic rrm1RRM2 mutants exhibited a significant reduction in affinity (~ 4–10-fold) compared with UP1 containing only the phosphomimetic substitutions. This behaviour is consistent with previously reported cooperative binding 12 , whereby initial binding to RRM1 enhances subsequent binding to RRM2 (Fig. 3 A). For all RRM1rrm2 mutants, two binding events were detected. However, the low affinity, poor signal-to-noise ratio, and substantial overlap of the two binding curves precluded reliable estimation of K D values (Fig. S5). It is therefore reasonable to assume that R3-2, which contains two UAG motifs, may also interact with RRM1, albeit with very low affinity. This is supported by the trend observed for the 20-mer RNA (Table 3). The phosphomimetic rrm1RRM2 mutants showed a reduction in affinity (~ 3-fold) relative to the phosphomimetic-only mutants, while the phosphomimetic RRM1rrm2 mutant exhibited two overlapping low-affinity binding events (Fig. S5). Although similar to the biphasic curve we described above for the UP1 S4DS6D/S4ES6E, the affinities were further reduced, and the low signal-to-noise ratio and curve overlap prevented quantitative determination of binding constants. By contrast, no binding was detected for UP1 rrm1RRM2, whereas two extremely weak binding events were observed for UP1 RRM1rrm2, which were weaker than those observed for the phosphomimetic RRM1rrm2 mutants and likewise insufficient for K D estimation. Taken together, these data suggest that the 20-mer is initially recognized by RRM1 of UP1, and only in the phosphorylated protein the RNA loop structure is destabilized, allowing the RNA to additionally engage RRM2 (Fig. 3 B) Phosphomimetic mutations do not induce a disorder-to-order transition. To unravel the mechanism through which the phosphorylation at S4 and S6 enables hnRNPA1 to specifically bind to IRES sequences in XIAP mRNAs, we investigated whether these modifications induce local or global conformational changes. We determined crystal structures of both the apo and DNA-bound (5’-TTAGGGTTAGGG-3’) forms of UP1 S4ES6E and S4DS6D (diffraction data and refinement statistics are provided in Table S.1). In all structures, the flexible N-tail is not visible, suggesting that protein activation does not involve a disorder-to-order transition in this region. The two RRM domains are overall identical to the UP1 wt structure, both in the apo and DNA-bound states. Each RRM adopts the canonical β1α1β2β3α2β4 topology, with the inter-domain linker becoming ordered upon DNA binding (Fig. 4 A–B). We further validated these findings using NMR spectroscopy, to discard presence of crystallographic biases. Previous NMR studies on UP1 have shown that the apo solution structure more closely resembles the DNA-bound crystal structure than the apo one, which has a rotation of 15° of one RRM in respect to the other 46 , also present in our phosphomimetic mutants crystal structures. Similarly, the DNA-bound crystallographic structure displays a crystallographic 2:2 stoichiometry that is not observed in solution and is not representative of the UP1–RNA complex in solution 12 . To investigate whether phosphomimetic mutations induce structural rearrangements, we acquired TROSY spectra of UP1 wt, S4DS6D and S4ES6E mutants. The spectra are overall very similar (Fig. 5 A), with minor differences in chemical shifts only for residues in direct contact with the inserted mutations at positions 4 and 6 (Fig. 5 B), which, in agreement with our crystallographic data, indicates lack of major conformational changes induced by the phosphomimetic mutation. However, to further investigate potential subtle structural changes, we analysed the NMR secondary chemical shift (SCS) propensities of both wt and mutants. Resonance assignments for 1 H N , 15 N H , 13 C α , 13 C β and 13 C’ in the UP1 mutants was performed using triple-resonance NMR experiments 60 . Assignments for the wt protein were reported previously 46 . Random coil chemical shifts were generated for both proteins using GGXGG-based nearest-neighbour correction for glycines 61 . The differences between experimental and generated random coil values for C α and C β for both the wt and the mutated UP1 were plotted against each other (Fig. 5 E). Analysis of these plots reveals only minor deviations from the expected linear correlation, confirming that the phosphomimetic mutations do not perturb the secondary structure of UP1. To characterize the impact of phosphorylation on protein dynamics, we performed solution NMR relaxation experiments. Addition of the 16- or 20-mer RNA oligonucleotides to UP1 resulted in extensive line broadening in the 1 H– 15 N correlation spectra, preventing analysis of the RNA-bound complexes (Fig. S6). Therefore, NMR characterization was restricted to the free protein. Fast (ps–ns) motions were evaluated using 15 N R₁ and R₂ measurements, while slower (µs–ms) dynamics were probed by 15 N CPMG relaxation dispersion experiments. Plots of the 15 N R₂/R₁ ratios and the chemical-exchange contributions to R₂ ( 15 N Rex) as a function of residue index (Fig. 5 B–C) indicate that the phosphomimetic mutations do not meaningfully alter these relaxation parameters. Thus, phosphorylation does not significantly affect the overall timescales of internal motions within UP1 and do not alter the reciprocal dynamics and communications between the RRM1 and RRM2. Phosphorylation perturbs the dynamics of the N-tail To fully investigate how phosphorylation affects the conformational dynamics of the N-terminal IDR of UP1, we performed MD simulations on four UP1 constructs: the wild-type protein (UP1 wt), the phosphomimetic mutants S4D/S6D and S4E/S6E, and the doubly phosphorylated form (phospho-UP1). For each variant, five independent 2 µs simulations were carried out, providing a total of 10 µs of trajectory data per system. In addition, five independent 2 µs simulations were performed for complexes formed by UP1 wt or phospho-UP1 in complex with RNA. These structures were generated using AlphaFold 41 with the 16-mer minimal binding sequence described above. The RNA adopts a horseshoe-shaped conformation on UP1 (Fig. 6 A), closely resembling models reported in previous studies 11 , 12 . In this structure, A 3 G 4 are positioned in direct contact with F17 and F59, and A 14 G 15 contact F108 and F150, in excellent agreement with our topology studies. Analysis of the MD trajectories revealed that phosphorylation significantly and differentially alters the conformational behaviour of the N-tail in the free and RNA-bound UP1. In the free protein, substitution of S4 and S6 with negatively charged groups introduces electrostatic interactions with residues R92 and K166 (Fig. 6 B). These interactions destabilize contacts between the N-tail and the RRM1 domain, increasing the flexibility of the N-tail and promoting transient proximity to RRM2 (Fig. 6 C). In the RNA-bound state, the simulations show that the N-tail of unphosphorylated UP1 forms several direct contacts with the RNA ( Fig. S7 ) that position it near RRM2 (Fig. 6 C), significantly reducing the sampled conformational space compared to the unbound protein. Phosphorylation at S4 and S6 weakens these interactions ( Fig. S7 ), thereby reshaping the conformational ensemble sampled by the N-tail during RNA engagement (Fig. 6 C), rotating it by ~ 90º compared to the non-phosphorylated protein. Taken together, the MD data demonstrate that phosphorylation at S4 and S6 primarily influences long-range intramolecular contacts rather than altering the intrinsic dynamics of the RRM domains. Specifically, phosphorylation remodels interactions involving the N-tail, thereby modifying the conformational space sampled by UP1, which in turn alters direct contacts between the N-tail and RNA and is likely linked to the differences observed in our affinity measurements. Discussions The RBP hnRNPA1 plays essential roles in fundamental cellular processes, including splicing, regulation of mRNA stability and nuclear export, and IRES-mediated translation 2 , 3 . Accordingly, hnRNPA1 has been implicated in numerous pathological contexts such as cancer, neurodegeneration, and viral infection 62 , 63 , 64 . Interest in hnRNPA1 has steadily increased, due to its multifunctional roles and idiosyncratic RNA binding mechanisms 2 . A central aspect of hnRNPA1 regulation is its dynamic nucleocytoplasmic shuttling, which allows it to selectively bind RNA sequences in the nucleous and promote cytosolic export 22 . HnRNPA1 has been involved in the reprogramming of metabolic and survival pathways in cancer cells, via phosphorylation on S4 and S6 by the kinase S6K2 22, 65 . Our study provides molecular insights into how phosphorylation of hnRNPA1 at S4 and S6 regulates its ITAF activity, with particular focus on the binding of the XIAP IRES. We demonstrated that phosphomimetic mutations, shown to behave like the phosphorylated protein in cell 22 , do not induce major structural rearrangements in the RRM domains or trigger a disorder-to-order transition in the N-tail. Instead, phosphorylation perturbs the conformational ensemble sampled by the N-tail, altering intramolecular contacts and enabling the protein to destabilise local RNA structures. The likely enhanced RNA melting activity of the phosphomimetic mutants facilitates binding to otherwise less accessible recognition motifs, thereby providing sequence- and RNA-specific binding. The enrichment of 5’-YAG-3’ motifs in recognized sequences is consistent with the previously characterized hnRNPA1 optimal motif 56 and with the preferential binding to oligos containing two closely spaced UAG motifs 12 . Our proposed mode of XIAP IRES binding (Fig. 6 ) closely resembles the horseshoe-like conformation observed for the single-stranded human intronic splicing silencer ISS-N1 11 , yet in an opposite orientation but similar polarity. In contrast to the single-stranded ISS-N1, our 20-mer adopts a weak loop structure and is the only oligo we tested with a detectable secondary structure, specifically recognised by the phosphomimetic mutants. Kooshapur et al. 12 demonstrated that hnRNPA1 can recognize RNA loops and that cooperative binding of both RRM domains facilitates melting of target stem–loop RNAs. Indeed, the authors noticed that the binding of UP1 to a 17-mer RNA helical stem was inconsistent with a single binding event. The concentration-dependent enthalpy changes followed a biphasic curve, which the author interpreted as reflecting sequential unwinding and binding events that can be captured by ITC 12 . Our titration closely resembles this behaviour and mode of binding; however, using spectral shifts approach, we were able to resolve our biphasic curve as two distinct binding events at the two RRM domains. In line with Kooshapur et al. 12 , we also detected cooperativity of binding. This was supported by mutating the RNPs in the two RRMs and showing that binding on the RRM2 is reduced in absence of binding or sequence specific binding on the RRM1. Mechanistically, we did not detect any phosphorylation-induced structural changes in UP1 based on our NMR and crystallographic data. However, our MD simulations revealed that phosphorylation reshapes the conformational ensemble by altering contact patterns between the N-tail and the RRMs. Similar phosphorylation-induced modulation of disordered-region dynamics has been reported for other RBPs, including HuR and Ets1 66, 67 . Baños-Jaime et al. 66 showed that phosphorylation at Y5 within the disordered N-terminal tail of HuR increases its dynamics and transient contacts with the β-sheet surface of RRM1 compared to the non-phosphorylated protein. Likewise, phosphorylation of two residues in the disordered N-terminal region of Ets1 core domain drastically changes the conformational ensemble of the N-terminal region and increases the contact with the folded domain 67 . In both cases, phosphorylated tails competed with nucleic acid binding, thereby reducing RNA/DNA binding affinity. By contrast, phosphorylation of the UP1 N-tail alters the contact between the IDR and the RNA, which is likely linked to UP1 unwinding activity. Indeed, a similar behaviour was recently described for the closely related hnRNPD (also known as AUF1) 68 . hnRNPD contains two uncoupled RRM domains flanked by IDRs. Lee et al. 68 elegantly demonstrated that direct interactions between IDRs and RNA are required both to enhance protein–RNA interactions and to promote RNA remodeling activity. Beyond XIAP, phosphorylation of hnRNPA1 at S4 and S6 is required for binding to the Bcl-xL 22 and mouse mammary tumor virus (MMTV) IRESs. Barrera et al. 69 showed that these phosphorylations are necessary for efficient retroviral gene expression, suggesting that phosphorylation-driven structural remodeling of target RNAs may represent a generalizable mechanism for ITAF regulation. In future, it will be important to determine whether modulation of IDR dynamics constitutes a shared mechanism underlying RNA recognition processes. Declarations Acknowledgements We acknowledge Diamond Light Source for time on Beamline I02 and I24 under Proposal MX32787. We thank the Centre for Biomolecular Spectroscopy at King’s. This work was supported by BBSRC grant (BB/X018997/1) to F.P, and National Institute of General Medical Sciences with grant no. R35GM133488 to V.V.. Research reported in this publication was supported by NIH (award number S10-OD032235) for a 700 MHz NMR upgrade (BNMRF, Iowa State University). References Dreyfuss G, Matunis MJ, Pinol-Roma S, Burd CG (1993) hnRNP proteins and the biogenesis of mRNA. Annu Rev Biochem 62:289–321 Levengood JD, Tolbert BS (2019) Idiosyncrasies of hnRNP A1-RNA recognition: Can binding mode influence function. Semin Cell Dev Biol 86:150–161 Dunnett L, Das S, Venditti V, Prischi F (2025) Enhanced identification of small molecules binding to hnRNPA1 via cryptic pockets mapping coupled with X-ray fragment screening. J Biol Chem 301:108335 Lichtenstein M, Guo W, Tartakoff AM (2001) Control of nuclear export of hnRNP A1. Traffic 2, 261–267 Zhang QS, Manche L, Xu RM, Krainer AR (2006) hnRNP A1 associates with telomere ends and stimulates telomerase activity. RNA 12:1116–1128 Xu RM, Jokhan L, Cheng X, Mayeda A, Krainer AR (1997) Crystal structure of human UP1, the domain of hnRNP A1 that contains two RNA-recognition motifs. Structure 5:559–570 Shamoo Y, Krueger U, Rice LM, Williams KR, Steitz TA (1997) Crystal structure of the two RNA binding domains of human hnRNP A1 at 1.75 A resolution. Nat Struct Biol 4:215–222 Mayeda A, Munroe SH, Xu RM, Krainer AR (1998) Distinct functions of the closely related tandem RNA-recognition motifs of hnRNP A1. RNA 4, 1111–1123 Ding J, Hayashi MK, Zhang Y, Manche L, Krainer AR, Xu RM (1999) Crystal structure of the two-RRM domain of hnRNP A1 (UP1) complexed with single-stranded telomeric DNA. Genes Dev 13:1102–1115 Morgan CE et al (2015) The First Crystal Structure of the UP1 Domain of hnRNP A1 Bound to RNA Reveals a New Look for an Old RNA Binding Protein. J Mol Biol 427:3241–3257 Beusch I, Barraud P, Moursy A, Clery A, Allain FH (2017) Tandem hnRNP A1 RNA recognition motifs act in concert to repress the splicing of survival motor neuron exon 7. Elife 6 Kooshapur H et al (2018) Structural basis for terminal loop recognition and stimulation of pri-miRNA-18a processing by hnRNP A1. Nat Commun 9:2479 Sharma K et al (2023) Cryo-EM Structure of the Full-length hnRNPA1 Amyloid Fibril. J Mol Biol 435:168211 Bonnal S et al (2005) Heterogeneous nuclear ribonucleoprotein A1 is a novel internal ribosome entry site trans-acting factor that modulates alternative initiation of translation of the fibroblast growth factor 2 mRNA. J Biol Chem 280:4144–4153 Spriggs KA, Stoneley M, Bushell M, Willis AE (2008) Re-programming of translation following cell stress allows IRES-mediated translation to predominate. Biol Cell 100:27–38 Leppek K, Das R, Barna M (2018) Functional 5' UTR mRNA structures in eukaryotic translation regulation and how to find them. Nat Rev Mol Cell Biol 19:158–174 Mahe M, Rios-Fuller T, Katsara O, Schneider RJ (2024) Non-canonical mRNA translation initiation in cell stress and cancer. NAR Cancer 6:zcae026 Gupta A, Bansal M, Ding J, Pandit M, Sudarshan S, Ding H-F (2024) DKC1-mediated pseudouridylation of rRNA targets hnRNP A1 to sustain IRES-dependent translation and ATF4-driven metabolic adaptation. bioRxiv , 2024.2012.2012.628228 Benavides-Serrato A et al (2023) m(6)A-modification of cyclin D1 and c-myc IRESs in glioblastoma controls ITAF activity and resistance to mTOR inhibition. Cancer Lett 562:216178 Holmes B et al (2016) Mechanistic Target of Rapamycin (mTOR) Inhibition Synergizes with Reduced Internal Ribosome Entry Site (IRES)-mediated Translation of Cyclin D1 and c-MYC mRNAs to Treat Glioblastoma. J Biol Chem 291:14146–14159 Benavides-Serrato A, Saunders JT, Holmes B, Nishimura RN, Lichtenstein A, Gera J (2020) Repurposing Potential of Riluzole as an ITAF Inhibitor in mTOR Therapy Resistant Glioblastoma. Int J Mol Sci 21 Roy R et al (2014) hnRNPA1 couples nuclear export and translation of specific mRNAs downstream of FGF-2/S6K2 signalling. Nucleic Acids Res 42:12483–12497 Gao G, Dhar S, Bedford MT (2017) PRMT5 regulates IRES-dependent translation via methylation of hnRNP A1. Nucleic Acids Res 45:4359–4369 Martin J et al (2011) Phosphomimetic substitution of heterogeneous nuclear ribonucleoprotein A1 at serine 199 abolishes AKT-dependent internal ribosome entry site-transacting factor (ITAF) function via effects on strand annealing and results in mammalian target of rapamycin complex 1 (mTORC1) inhibitor sensitivity. J Biol Chem 286:16402–16413 Siegel RL, Miller KD, Jemal A (2018) Cancer statistics, 2018. CA Cancer J Clin 68:7–30 Pardo OE et al (2006) FGF-2 protects small cell lung cancer cells from apoptosis through a complex involving PKCepsilon, B-Raf and S6K2. EMBO J 25:3078–3088 Levengood JD et al (2024) Thermodynamic coupling of the tandem RRM domains of hnRNP A1 underlie its pleiotropic RNA binding functions. Sci Adv 10:eadk6580 Thapar R (2015) Structural basis for regulation of RNA-binding proteins by phosphorylation. ACS Chem Biol 10:652–666 Sellis D, Drosou V, Vlachakis D, Voukkalis N, Giannakouros T, Vlassi M (2012) Phosphorylation of the arginine/serine repeats of lamin B receptor by SRPK1-insights from molecular dynamics simulations. Biochim Biophys Acta 1820:44–55 Prischi F, Nowak PR, Carrara M, Ali MM (2014) Phosphoregulation of Ire1 RNase splicing activity. Nat Commun 5:3554 Thapar R, Marzluff WF, Redinbo MR (2004) Electrostatic contribution of serine phosphorylation to the Drosophila SLBP–histone mRNA complex. Biochemistry 43:9401–9412 Grammatikakis I, Abdelmohsen K, Gorospe M (2017) Posttranslational control of HuR function. Wiley Interdiscip Rev RNA 8 Cai M, Huang Y, Sakaguchi K, Clore GM, Gronenborn AM, Craigie R (1998) An efficient and cost-effective isotope labeling protocol for proteins expressed in Escherichia coli. J Biomol NMR 11:97–102 Jokan L, Dong AP, Mayeda A, Krainer AR, Xu RM (1997) Crystallization and preliminary X-ray diffraction studies of UP1, the two-RRM domain of hnRNP A1. Acta Crystallogr D Biol Crystallogr 53:615–618 Winter G, Lobley CM, Prince SM (2013) Decision making in xia2. Acta Crystallogr D Biol Crystallogr 69:1260–1273 Evans PR, Murshudov GN (2013) How good are my data and what is the resolution? Acta Crystallogr D Biol Crystallogr 69:1204–1214 Liebschner D et al (2019) Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol 75:861–877 Emsley P, Lohkamp B, Scott WG, Cowtan K (2010) Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66:486–501 Sato K, Akiyama M, Sakakibara Y (2021) RNA secondary structure prediction using deep learning with thermodynamic integration. Nat Commun 12:941 Sweeney BA et al (2021) R2DT is a framework for predicting and visualising RNA secondary structure using templates. Nat Commun 12:3494 Abramson J et al (2024) Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630:493–500 Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6:277–293 Lee W, Rahimi M, Lee Y, Chiu A (2021) POKY: a software suite for multidimensional NMR and 3D structure calculation of biomolecules. Bioinformatics 37:3041–3042 Narayanan RL, Durr UH, Bibow S, Biernat J, Mandelkow E, Zweckstetter M (2010) Automatic assignment of the intrinsically disordered protein Tau with 441-residues. J Am Chem Soc 132:11906–11907 Abraham MJ et al (2015) High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1–2 GROMACS:19–25 Barraud P, Allain FH (2013) Solution structure of the two RNA recognition motifs of hnRNP A1 using segmental isotope labeling: how the relative orientation between RRMs influences the nucleic acid binding topology. J Biomol NMR 55:119–138 Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA (2004) Development and testing of a general amber force field. J Comput Chem 25:1157–1174 Michaud-Agrawal N, Denning EJ, Woolf TB, Beckstein O (2011) MDAnalysis: a toolkit for the analysis of molecular dynamics simulations. J Comput Chem 32:2319–2327 Schrödinger L, DeLano W (2022) The PyMOL Molecular Graphics System, Version 2.4.0.) Yang Y, Wang Z (2019) IRES-mediated cap-independent translation, a path leading to hidden proteome. J Mol Cell Biol 11:911–919 Thakor N et al (2017) Cellular mRNA recruits the ribosome via eIF3-PABP bridge to initiate internal translation. RNA Biol 14:553–567 Baird SD, Lewis SM, Turcotte M, Holcik M (2007) A search for structurally similar cellular internal ribosome entry sites. Nucleic Acids Res 35:4664–4677 Filbin ME, Kieft JS (2009) Toward a structural understanding of IRES RNA function. Curr Opin Struct Biol 19:267–276 Burd CG, Dreyfuss G (1994) RNA binding specificity of hnRNP A1: significance of hnRNP A1 high-affinity binding sites in pre-mRNA splicing. EMBO J 13:1197–1204 Lewis SM, Veyrier A, Hosszu Ungureanu N, Bonnal S, Vagner S, Holcik M (2007) Subcellular relocalization of a trans-acting factor regulates XIAP IRES-dependent translation. Mol Biol Cell 18:1302–1311 Jain N, Lin HC, Morgan CE, Harris ME, Tolbert BS (2017) Rules of RNA specificity of hnRNP A1 revealed by global and quantitative analysis of its affinity distribution. Proc Natl Acad Sci U S A 114:2206–2211 Liu X et al (2017) Structure-Dependent Binding of hnRNPA1 to Telomere RNA. J Am Chem Soc 139:7533–7539 Butcher SE, Pyle AM (2011) The molecular interactions that stabilize RNA tertiary structure: RNA motifs, patterns, and networks. Acc Chem Res 44:1302–1311 Tang T et al (2022) Receptor-specific recognition of NPY peptides revealed by structures of NPY receptors. Sci Adv 8:eabm1232 Clore GM, Gronenborn AM (1998) Determining the structures of large proteins and protein complexes by NMR. Trends Biotechnol 16:22–34 Kjaergaard M, Poulsen FM (2011) Sequence correction of random coil chemical shifts: correlation between neighbor correction factors and changes in the Ramachandran distribution. J Biomol NMR 50:157–165 Bekenstein U, Soreq H (2013) Heterogeneous nuclear ribonucleoprotein A1 in health and neurodegenerative disease: from structural insights to post-transcriptional regulatory roles. Mol Cell Neurosci 56:436–446 Jean-Philippe J, Paz S, Caputi M (2013) hnRNP A1: the Swiss army knife of gene expression. Int J Mol Sci 14:18999–19024 Roy R, Huang Y, Seckl MJ, Pardo OE (2017) Emerging roles of hnRNPA1 in modulating malignant transformation. Wiley Interdiscip Rev RNA 8 Sun Y et al (2017) Phosphorylation of Ser6 in hnRNPA1 by S6K2 regulates glucose metabolism and cell growth in colorectal cancer. Oncol Lett 14:7323–7331 Banos-Jaime B et al (2024) Phosphorylation at the disordered N-end makes HuR accumulate and dimerize in the cytoplasm. Nucleic Acids Res 52:8552–8565 Kasahara K, Shiina M, Higo J, Ogata K, Nakamura H (2018) Phosphorylation of an intrinsically disordered region of Ets1 shifts a multi-modal interaction ensemble to an auto-inhibitory state. Nucleic Acids Res 46:2243–2251 Lee NC, Tilley HH, Acle GA, McGinnis PJ, Wilson GM (2025) Unstructured protein domains stabilize RNA binding and mediate RNA folding by AUF1. J Biol Chem 301:108442 Barrera A et al (2020) Post-translational modifications of hnRNP A1 differentially modulate retroviral IRES-mediated translation initiation. Nucleic Acids Res 48:10479–10499 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterials.docx Supplementary info Tables.docx Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information

Related documents

• LAT condensation gates PLCγ1 activation via bimodal LAT phosphorylation#241472
• Intracellular protein binders for imaging, control and future therapeutics.#671020
• Protein id is O95999#638781
• PPIscreenML is a method for structure-based screening of protein-protein interactions using AlphaFold.#122736
• A Peptide Photoaffinity Probe Specific for the Active Conformation of the Abl Tyrosine Kinase#619337