Negative regulation of miRNAs sorting in EVs: the RNA-binding protein PCBP2 impairs SYNCRIP-mediated miRNAs EVs loading

Reviewer #1 (Public review):

In this study, Marocco and colleages perform a deep characterization of the complex molecular mechanism guiding the recognition of a particular CELLmotif previously identified in hepatocytes in another publication. Having miR-155-3p with or without this CELLmotif as initial focus, authors identify 21 proteins differentially binding to these two miRNA versions. From these, they decided to focus on PCBP2. They elegantly demonstrate PCBP2 binding to miR-155-3p WT version but not to CELLmotif-mutated version. miR-155-3p contains a hEXOmotif identified in a different report, whose recognition is largely mediated by another RNA-binding protein called SYNCRIP. Interestingly, mutation of the hEXOmotif contained in miR-155-3p did not only blunt SYNCRIP binding, but also PCBP2 binding despite the maintenance of the CELLmotif. This indicates that somehow SYNCRIP binding is a pre-requisite for PCBP2 binding. EMSA assay confirms that SYNCRIP is necessary for PCBP2 binding to miR-155-3p, while PCBP2 is not needed for SYNCRIP binding. Then authors aim to extend these finding to other miRNAs containing both motifs. For that, they perform a small-RNA-Seq of EVs released from cells knockdown for PCBP2 versus control cells, identifying a subset of miRNAs whose expression either increases or decreases. The assumption is that those miRNAs containing PCBP2-binding CELLmotif should now be less retained in the cell and go more to extracellular vesicles, thus reflecting a higher EV expression. The specific subset of miRNAs having both the CELLmotif and hEXOmotif (9 miRNAs) whose expressions increase in EVs due to PCBP2 reduction is also affected by knocking-down SYNCRIP in the sense that reduction of SYNCRIP leads to lower EV sorting. Further experiments confirm that PCBP2 and SYNCRIP bind to these 9 miRNAs and that knocking down SYNCRIP impairs their EV sorting.

In the revised manuscript, the authors have addressed most of my concerns and questions. I believe the new experiments provide stronger support for their claims. My only remaining concern is the lack of clarity in the replicates for the EMSA experiment. The one shown in the manuscript is clear; however, the other three replicates hardly show that knocking down SYNCRIP has an effect on PCBP2 binding. Even worse is the fact that these replicates do not support at all that PCBP2 silencing has no effect on SYNCRIP binding, as the bands for those types of samples are, in most of the cases, not visible. I think the authors should work on repeating a couple of times EMSA experiment.

Reviewer #2 (Public review):

Summary:

The author of this manuscript aimed to uncover the mechanisms behind miRNA retention within cells. They identified PCBP2 as a crucial factor in this process, revealing a novel role for RNA-binding proteins. Additionally, the study discovered that SYNCRIP is essential for PCBP2's function, demonstrating the cooperative interaction between these two proteins. This research not only sheds light on the intricate dynamics of miRNA retention but also emphasizes the importance of protein interactions in regulating miRNA behavior within cells.

Strengths:

This paper makes important progress in understanding how miRNAs are kept inside cells. It identifies PCBP2 as a key player in this process, showing a new role for proteins that bind RNA. The study also finds that SYNCRIP is needed for PCBP2 to work, highlighting how these proteins work together. These discoveries not only improve our knowledge of miRNA behavior but also suggest new ways to develop treatments by controlling miRNA locations to influence cell communication in diseases. The use of liver cell models and thorough experiments ensures the results are reliable and show their potential for RNA-based therapies

Weaknesses:

The manuscript is well-structured and presents compelling data, but I noticed a few minor corrections that could further enhance its clarity:

Figure References: In the response to Reviewer 1, the comment states, "It's not Panel C, it's Panel A of Figure 1"-this should be cross-checked for consistency.
Supplementary Figure 2 is labeled as "Panel A"-please verify if additional panels (B, C, etc.) are intended.

Western Blot Quality: The Alix WB shows some background noise. A repeat with optimized conditions (or inclusion of a cleaner replicate) would strengthen the data. Adding statistical analysis for all WBs would also reinforce robustness.

These are relatively small refinements, and the manuscript is already in excellent shape. With these adjustments, it will be even stronger.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public review):

In this study, Marocco and colleagues perform a deep characterization of the complex molecular mechanism guiding the recognition of a particular CELLmotif previously identified in hepatocytes in another publication. Having miR-155-3p with or without this CELLmotif as the initial focus, the authors identify 21 proteins differentially binding to these two miRNA versions. From there, they decided to focus on PCBP2. They elegantly demonstrate PCBP2 binding to the miR-155-3p WT version but not to the CELLmotif-mutated version. miR-1553p contains a hEXOmotif identified in a different report, whose recognition is largely mediated by another RNA-binding protein called SYNCRIP. Interestingly, mutation of the hEXOmotif contained in miR-155-3p did not only blunt SYNCRIP binding but also PCBP2 binding despite the maintenance of the CELLmotif. This indicates that somehow SYNCRIP binding is a pre-requisite for PCBP2 binding. EMSA assay confirms that SYNCRIP is necessary for PCBP2 binding to miR-155-3p, while PCBP2 is not needed for SYNCRIP binding. The authors aim to extend these findings to other miRNAs containing both motifs. For that, they perform a small-RNA-Seq of EVs released from cells knockdown for PCBP2 versus control cells, identifying a subset of miRNAs whose expression either increases or decreases. The assumption is that those miRNAs containing PCBP2-binding CELLmotif should now be less retained in the cell and go more to extracellular vesicles, thus reflecting a higher EV expression. The specific subset of miRNAs having both the CELLmotif and hEXOmotif (9 miRNAs) whose expressions increase in EVs due to PCBP2 reduction is also affected by knocking-down SYNCRIP in the sense that reduction of SYNCRIP leads to lower EV sorting. Further experiments confirm that PCBP2 and SYNCRIP bind to these 9 miRNAs and that knocking down SYNCRIP impairs their EV sorting.

We thank this Reviewer for the time spent on our manuscript and for having appreciated our characterization of the present molecular mechanism controlling miRNA export/cellretention in hepatocytes.

While the process studied in this work is novel and interesting, there are several aspects of this manuscript that should be improved:

(1) First of all, the nature of the CELLmotif and the hEXOmotif they are studying is extremely confusing. For the CELLmotif, the authors seem to focus on the Core CELLmotif AUU A/G in some experiments and the extended 7-nucleotide version in others. The fact that these CELLmotif and hEXOmotif are not shown anywhere in the figures (I mean with the full nucleotide variability described in the original publications) but only referred to in the text further complicates the identification of the motifs and the understanding of the experiments. Moreover, I am not convinced that the sequences they highlight in grey correspond to the original CELLmotif in all cases. For instance, in the miR-155-3p sequence, GCAUU is highlighted in grey. However, the original CELLmotif is basically 7-nucleotide long: C, A/U, G/A/C, U, U/A, C/G/A, A/U/C or CAGUUCA in its more abundant version. I can only see clearly the presence of the Core CELLmotif AUUA in miR-155-3p; however, the last A is not highlighted in grey. It is true that there is some nucleotide variability in each position in the originally reported CELLmotif by the authors in ref. 5 and the hEXOmotifs in ref. 7; however, not all nucleotides are equally likely to be found in each position. This fact seems to be not to be taken into account by the authors as they took basically any sequence with any length and almost sequence combination as valid CELLmotif. This means that I cannot identify the CELLmotif in many cases among the ones they highlight in grey. Instead, they should really focus on the most predominant CELLmotif sequence or, instead, take a reduced subset of "more abundant" CELLmotif versions from the ones that could fit in the originally described CELLmotif. Altogether, the authors need to explain much better what they have considered as the CELLmotif, what is the Core CELLmotif and what is hEXOmotif in each case and restrict to the most likely versions of the CELLmotif and hEXOmotif.

We thank the Reviewer for having raised this concern and indeed we must agree with her/him and therefore, we modified the text and the figure accordingly.

In brief, as now stated, with respect to the CELL motif, miR 155-3p, miR-155-5p, miR-181d5p, miR-3084-5p, miR-122b-3p, miR-192-5p, miR-26b-3p, miR-31-3p, miR-195a-5p and miR-421-3p have the Core CELL motif (AUUA/G) described by Garcia- Martin and colleagues (ref. 5).

Other miRNAs (miR-345-3p, miR-23a-5p and miR-214-3p) share the described CELL motif (ref.5) with the most frequent nucleotides, considering also the reported variability. Also for the hEXO motif described by Santangelo and collaborators (ref.7), the most frequent nucleotides defining the motif sequence have been taken into consideration. The motifs have been better highlighted in the new version of Fig. 1 panel C.

(2) Validation of EV isolation method: first, a large part of Supplementary Figure 2 is not readable. EV markers seem to be enriched in EV isolates; however, more EV and cell markers should be assayed to fulfill MISEV guidelines.

We apologize for the low quality of the figure. In order to address this issue, we replaced the Supplementary Figure 2 panel A (now panel B) and we added further EV markers (TSG101, Alix, Flotillin) in Supplementary Figure 2 panel B (now panel A). Notably, in the same Western blot analysis we also addressed the expression of SYNCRIP and PCBP2 (that were found in the cellular end EV compartments or only in the intracellular compartment respectively).

(3) A key variable is missing in Supplementary Figure 2, which is whether PCBP2 or SYNCRIP knockdowns impair EV secretion rates. A quantification of the nr vesicles released per cell upon knocking down each of these factors would be essential to rule out that any of the effects seen throughout the paper are not due to reduced or enhanced EV production rather than miRNA sorting/retention.

We addressed this issue by quantifying the number of EVs per cell in shPCBP2 or shSYNCRIP with respect to the shCTR conditions. Data are shown in the new Supplementary Figure 2 panel C and indicate that there are not significant differences on EVs production rate upon PCBP2 or SYNCRIP knockdown.

(4) The EMSA experiment is important to support their claims. Given the weak bands that are shown, the authors need to show all their replicates to convince the readers that it is reproducible.

We are aware that the signals appear faint; the experimental replicates showing the robustness of the observation are reported below.

Author response image 1.

(5) Although the bindings of SYNCRIP and PCBP2 to miR-155-3p and other miRNAs having both hEXOmotif and CELLmotif seem clear, the need for SYNCRIP binding to allow for PCBP2-mediated cellular retention is counterintuitive. What happens to those miRNAs that only contain a CELLmotif in terms of cellular retention and SYNCRIP dependence for cellular retention? In this regard, a representative miRNA (miR-31-3p) is analyzed in several experiments, showing that PCBP2 does not bind to it unless a hEXOmotif is introduced (Figure 3). However, this type of experiment should definitely be extended to other miRNAs containing only CELLmotif without hEXOmotif.

Based on the Reviewer’s suggestion we confirmed previous findings by extending the observation to further two miRNAs embedding the sole CELL-motif (miR-195a-5p and 4213p) whose sequences are reported in Figure 4C. Data relative to qPCR amplification are reported in Figure 4D, Figure 5 panels A-B, Figure 6 and new Supplementary Figure 3. They confirm that miRNAs only containing CELL motif are not EV-exported in dependence of SYNCRIP and are cell-retained independently of PCBP2 silencing.

(6) Along the same line, I am missing another important experiment: the artificial incorporation of CELLmotif. For example, miR-365-2-5p lacks a CELLmotif but has a hEXOmotif. Does PCBP2 bind to this miRNA upon incorporation of CELLmotif? Does this lead now to enhanced cellular retention of this miRNA?

We are grateful for the Reviewer's concern. As suggested, we added RNA pull-down experiments for miR-365-2-5p in wild type form and in mutated form (with the inclusion of CELL motif). As reported in the new Figure 1 panel E, the addition of the CELL motif maintains SYNCRIP binding and allows PCBP2 interaction with this miRNA.

(7) What would be the net effect of knocking down both SYNCRIP and PCBP2 at the same time? Would this neutralize each other's effect or would the lack of one impose on the other? That could help in understanding the complex interplay between these two factors for mediating cellular retention and EV sorting.

SYNCRIP and PCBP2 play opposite roles in the dynamics of miRNA retention/export. SYNCRIP is involved in the loading of miRNAs into EVs through the recognition of hEXO motif. Instead, PCBP2 is involved in cellular retention of miRNAs, acting as a negative regulator of SYNCRIP activity. PCBP2 binding and function requires both CELL-motif and SYNCRIP binding in order to negatively regulate miRNAs export into EVs.

Being SYNCRIP silencing sufficient to cause miRNA retention (as shown in Supplementary Figure 3), we believe that the contemporary silencing of PCBP2 should not disclose any additional aspect on cellular retention and EV sorting dynamics.

(8) The authors have here a great opportunity to shed some light on an unclear aspect of miRNA EV sorting and cellular retention: whether the RBPs go together with the miRNA to the EVs or not. While the original paper describing hEXOmotif found SYNCRIP in EVs, another publication (Jeppesen et al, Cell 2019; PMID: 30951670) later found this RBP being very scarce in small EVs compared to cellular bodies or large EVs (Supplementary Tables 3 and 4 in that publication). Can the authors find SYNCRIP and PCBP2 in the EVs? Another important question would be the colocalization of these RBPs in the place where the miRNA selection is supposed to take place: in multivesicular bodies (MVB). Is there a colocalization of these RBPs with MVBs in the cell?

We are thankful for the Reviewer’s suggestions. As reported in Supplementary Figure 2A SYNCRIP is present in both the intracellular end EV compartment and PCBP2 is detectable only in the intracellular one.

(9) In Figure 4C, the authors state in the text that CELLmotif and hEXOmotif are present in extra-seed region; however, for miR-181d-5p and miR-122-3p this is not true as their CELLmotifs fall within the seed sequence.

We apologize for our mistake. While for hEXO motif, it is confirmed that it is present in extraseed region on all analyzed miRNAs (as in ref. 7), the CELL motif on the cited miRNAs is overlapping with the seed sequence. We modified the text accordingly.

(10) The authors need to describe how they calculate the EV/cell ratio in gene expression in some experiments (for instance, Figures 1H, 4D, etc). Did they use any housekeeping gene for EV RNA content, the same RNA load, or some other alternative method to normalize EV vs cell RNA content?

We apologize for having not well clarified the calculation of EV/cell ratio in the cited figures. Data are shown as ratio of miRNAs expression in EVs with respect to the intracellular compartment. Expression of miRNAs in both compartments are normalized with respect to the spike-in sequence (cel-miR-39-3p), included in miRNAs sample (EVs and intracellular samples). This is also better clarified in the Materials and Methods section.

(11) I would suggest that the authors speculate a bit in the discussion section on how the interaction between PCBP2 and SYNCRIP takes place. Do they contain any potential interacting domain? The binding of one to the miRNA would impose a topological interference on the binding of the other?

We now speculate on the interaction between PCBP2 and SYNCRIP in the discussion section. Briefly, we described that PCBP2 interaction with several proteins have been reported (as in PMID 19881509 and 10772858), indicating the C-terminal domain including also the two KH1 and KH2 regions as the domains with the highest propensity interaction with proteins. Also in the case of SYNCRIP binding, the domains of interaction with proteins have been reported (as in PMID 10734137, 29483512 and 16765914) and we should hypothesize that these domains represent conserved regions responsible for its interaction also with PCBP2. Moreover, we also discussed that upon the interaction between SYNCRIP and the miRNAs a topological switch can occur, impacting the affinity of PCBP2 for the same miRNAs.

Reviewer #2 (Public review):

Summary:

The author of this manuscript aimed to uncover the mechanisms behind miRNA retention within cells. They identified PCBP2 as a crucial factor in this process, revealing a novel role for RNA-binding proteins. Additionally, the study discovered that SYNCRIP is essential for PCBP2's function, demonstrating the cooperative interaction between these two proteins. This research not only sheds light on the intricate dynamics of miRNA retention but also emphasizes the importance of protein interactions in regulating miRNA behavior within cells.

We thank this Reviewer for having appreciated our characterization of the molecular dynamics governing miRNA export/cell-retention in hepatocytes.

Strengths:

This paper makes important progress in understanding how miRNAs are kept inside cells. It identifies PCBP2 as a key player in this process, showing a new role for proteins that bind RNA. The study also finds that SYNCRIP is needed for PCBP2 to work, highlighting how these proteins work together. These discoveries not only improve our knowledge of miRNA behavior but also suggest new ways to develop treatments by controlling miRNA locations to influence cell communication in diseases. The use of liver cell models and thorough experiments ensures the results are reliable and show their potential for RNA-based therapies

Weaknesses:

Despite its strengths, the manuscript has several notable limitations. The study's exclusive focus on hepatocytes limits the applicability of the findings to other cell types and physiological contexts. While the interaction between PCBP2 and SYNCRIP is wellcharacterized, the manuscript lacks detailed insights into the structural basis of this interaction and the dynamic regulation of their binding. The generalization of the findings to a broader spectrum of miRNAs and RNA-binding proteins (RBPs) remains underexplored, leaving gaps in understanding the full scope of miRNA compartmentalization.

Furthermore, the therapeutic implications of these findings, though promising, are not directly connected to specific disease models or clinical scenarios, reducing their immediate translational impact. The manuscript would also benefit from a deeper discussion of potential upstream regulators of PCBP2 and SYNCRIP and the influence of cellular or environmental factors on their activity. Additionally, it is important to note that SYNCRIP has already been recognized as a major regulator of miRNA loading in extracellular vesicles (EVs). However, the purity of EVs is a concern, as the author only performed crude extraction methods without further purification using an iodixanol density gradient. The study also lacks in vivo evidence of PCBP2's role in exosomal miRNA export.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Improve figure quality in some cases (Figures 1A, 4B, Supplementary Figure 2).

Figures have been improved accordingly.

Reviewer #2 (Recommendations for the authors):

Questions for the Authors:

(1) Why was hepatocyte-specific data prioritized, and how generalizable are the findings to other cell types?

This work is based on our previous publication (Santangelo et al., Cell Reports), concerning the identification of the RBP SYNCRIP as an actor in the loading machinery of miRNAs in Extracellular Vesicles in hepatocytes. Since both SYNCRIP and PCBP2 are expressed in different cell types (Keerthikumar et al., 2016, PMID: 26434508; https://www.bgee.org/gene/ENSMUSG00000056851), is conceivable that our findings can be translated also in other cellular systems. To formally proof this hypothesis seems out of the scope of this manuscript.

(2) Can the authors elaborate on the functional impact of PCBP2-mediated miRNA retention? Which biological pathways are directly influenced by miRNAs retained by PCBP2?

We appreciate the suggestion; in line with this comment, we performed a Gene Ontology enrichment analysis of the targets of the retained miRNAs. In order to be the most exhaustive, we included both validated and predicted targets, respectively obtained from TarBase v9.0 database and DIANA-microT web server. As reported in the new figure 4, panel E, and in the new supplementary figure 4, the analysis highlighted several biological pathways collectively influenced by the PCBP2-dependent cell-retained miRNAs, including establishment of organelle localization, regulation of cell cycle and lymphocyte differentiation.

(3) What criteria were used to select the miRNAs (e.g., miR-155-3p) for this study?

miR-155-3p was selected as initial bait for RNA pulldown based on the reported presence of Core CELL motif in AML12 cell line (PMID: 34937935).

(4) How do the results using recombinant PCBP2 in RNA pull-down assays compare with those using native PCBP2 in cellular extracts?

The RNA-pull down with recombinant PCBP2 confirms the evidence obtained by RNA-pull downs with cellular extracts. Indeed, PCBP2 interacts with miR-155-3p in the wild type form and this interaction is lost upon the mutation of the CELL motif. Moreover, this experiment highlights a direct and sequence specific interaction.

(5) How much protein was loaded for Western blot analysis?

We’re sorry for not explaining the experimental procedure in depth. For protein expression analysis, as reported in supplementary figure 1 and 2, we loaded 30 µg of proteins. Half of the amount of the protein obtained upon either RNA pull-down or protein immunoprecipitation experiments performed using 2 mg of protein extract were analyzed. This information has been added to the methods section.

Suggested experiments to strengthen the manuscript:

(1) Purify EVs using an iodixanol density gradient to eliminate the possibility of soluble PCBP2 contamination.

We appreciate this suggestion. In order to avoid the effect of PCBP2 contamination that represents a source of variability in the experiments, we evaluated its presence in the purified extracellular vesicles protein extracts. As reported in the new figure Supplementary 2A, PCBP2 is completely absent in EV extracts as assessed by Western Blot; thus, accordingly with MISEV guideline we followed the differential ultracentrifugation method for EV purification.

(2) Perform gain- and loss-of-function assays by overexpressing or silencing PCBP2 in various models to observe downstream changes in miRNA-dependent pathways.

We chose to silence PCBP2 protein since its high expression in our cellular model. Overexpression of PCBP2 would probably have no other significant readout. We are aware that PCBP2 silencing would perturb miRNA biogenesis and in turn miRNA downstream pathway modulation. Indeed, its association with Dicer has been reported as propaedeutic to miRNA processing (Li, et al 2012, Cell metabolism). However, this aspect is out of the scope of the present manuscript, here we focus exclusively on PCBP2 role in the regulation of miRNA EV export. Moreover, to overcome the effects on miRNA processing we evaluated the expression level of each miRNA as ratio between the extracellular and intracellular compartment.

(3) Use a murine model with hepatocyte-specific PCBP2 knockout and track changes in EV miRNA content and their functional effects in target tissues.

We took advantage of our murine cells silenced for PCBP2 and evaluated miRNA content.

New functional assays (now included in the new figure 4F) with leukocytes obtained from C57BL/6J mice livers show a higher percentage of IFN-+ T cell, NK and myeloid cells upon shPCBP2 EVs treatment in comparison to the shCTR EVs; this suggests that PCBP2 silencing results into an EV-mediated modulation of the immune response.

(4) Conduct co-culture experiments to assess EV-mediated intercellular communication between donor and recipient cells.

We reasoned that co-culture experiments don’t limit the observed effect on EVs since the contribution of soluble factors can have a role on recipient cells. Conversely the treatments with purified EVs, here performed, allow the evaluation of the sole EV-mediated downstream effects.

These experiments would provide insights into the PCBP2-SYNCRIP axis, broaden the applicability of the findings, and enhance their translational relevance.