The regulation of miR-155 strand selection by CELF2, FUBP1 and KSRP proteins

Abstract

The microRNA-155 exists in two forms, miR-155-5p and miR-155-3p, produced from either strand of the double-stranded precursor of miR-155 in a mutually exclusive manner. The more abundant and better-studied miR-155-5p has been implicated in numerous biological processes, with dysregulated expression observed in various human diseases. miR-155-5p plays an essential role in supporting inflammatory responses in macrophages. Activating macrophages with lipopolysaccharide (LPS) elevates miR-155-5p, while the anti-inflammatory cytokine interleukin-10 (IL10) reduces miR-155-5p levels. Recently, researchers have suggested that miR-155-3p also plays a role in macrophage function, although its specific function in this context remains unclear. We found that LPS stimulation of macrophages results first in the elevation of miR-155-3p levels, followed by an increase in miR-155-5p levels. In this paper, we investigate the mechanisms underlying the maturation of pre-miR-155 into either miR-155-5p or miR-155-3p. We describe the contribution of three RNA-binding proteins, CELF2, FUBP1 and KSRP, to pre-miR-155 processing. Our data suggest that CELF2 regulates the selection of miR-155-5p and miR-155-3p strands. FUBP1 may support the expression of miR-155-3p for specific subcellular functions, while KSRP appears to inhibit both miR-155-5p and miR-155-3p maturation without altering the relative expression of each strand.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-15004-w.

Introduction

MicroRNAs (miRNAs) are short, noncoding regulatory RNAs that post-transcriptionally regulate gene expression through sequence-specific targeting of mRNA. MiR-155 is among the most extensively studied miRNAs due to its involvement in various biological and pathological processes, including immunity, inflammation and cancer^1–7. The MIR155HG locus encodes miR-155, and the primary transcript of miR-155 (pri-miR-155) undergoes Drosha-mediated cleavage to the precursor of miR-155 (pre-miR-155)⁸. Pre-miR-155 can undergo processing into two alternate forms, the dominant strand miR-155-5p and the non-dominant strand miR-155-3p⁹ depending on which strand from the pre-miR-155 is selected. The selected strand is loaded onto the RNA-induced silencing complex (RISC) and becomes the guide strand, while the other strand undergoes degradation¹⁰. The 5p and 3p strands have different seed sequences and thus target different mRNAs¹¹. Although miR-155-5p is the most abundant and studied strand, miR-155-3p was recently shown to have biological functions^12–14. Furthermore, the dysregulated expression of miR-155-3p alone has been observed in certain cancers and diseases^15–20.

Activating macrophage cells is essential in the human host’s defense against pathogens^21,22. We²³ and other^2,3,24 showed that activation of macrophage cells with lipopolysaccharide (LPS), a cell wall component of gram-negative bacteria that binds to the toll-like receptor 4 (TLR4) on host cells, induced expression of miR-155-5p. miR-155-5p was the only strand studied and required for macrophage production of inflammatory cytokines such as tumour necrosis factor-alpha (TNFα). However, persistent macrophage activation can become harmful, leading to inflammatory diseases, cardiovascular diseases and cancers^25–27. Consequently, activated macrophages also release an autocrine, anti-inflammatory cytokine called interleukin-10 (IL10) to attenuate the pro-inflammatory effects of LPS²⁸. IL10 is a pleiotropic cytokine characterized initially as a cytokine synthesis inhibitory factor produced by murine Th₂ T-cells to prevent cytokine production by murine Th₁ T-cells²⁸. However, deactivating the activated macrophages is its major in vivo function²⁹. IL10 receptor (IL10R) signalling uses both the Signal Transducer and Activator of Transcription 3 (STAT3)^30–32 and Src Homology 2 domain-containing Inositol-5 Phosphatase 1 (SHIP1)-dependent^23,30,33 pathways to deactivate macrophages^{23,31,33–35}. We previously reported that SHIP1 could form a complex with STAT3 (SHIP1:STAT3) in response to IL10R signalling or through the exposure of cells to small molecules that bind to induce a conformational change in SHIP1³⁰. The formation of the SHIP1:STAT3 complex is sufficient to inhibit macrophage production of TNFα. We also showed that IL10 inhibition of miR-155-5p required SHIP1 and STAT3²³.

We examined how IL10 interferes with the LPS-induced expression of miR-155-5p and found that IL10 inhibits the maturation of pre-miR-155 into mature miR-155-5p²³. We then used mass spectrometry to identify proteins that bind to pre-miR-155 to characterize the mechanism by which IL10R signalling inhibits pre-miR-155 maturation³⁶. CUGBP Elav-Like Family member 2 (CELF2) protein is an RNA-binding protein known for its post-transcriptional regulation of mRNAs through alternative splicing of mRNAs^37,38 or polyadenylation³⁹. CELF2 and its family member CELF1 also participate in microRNA regulation⁴⁰. Treiber et al.. showed that CELF2 and CELF1 proteins bind to the stem of pre-miR-140 and inhibit its maturation to miR-140⁴⁰. In our previous study, we found that CELF2 protein associates with pre-miR-155 in an IL10-dependent manner, and deletion of CELF2 enhanced LPS-induced expression of miR-155-5p, suggesting CELF2’s role in inhibiting miR-155-5p expression³⁶.

Recently, Simmonds et al. showed that LPS stimulation also leads to the early expression of miR-155-3p, followed by miR-155-5p at later time points⁴¹. Thus, in the current study, we examined whether CELF2 and two other proteins we identified to bind pre-miR-155, Far Upstream Element Binding Protein 1 (FUBP1) and KH-type Splicing Regulatory Protein (KSRP)^36,42might be involved in regulating miR-155-3p and miR-155-5p levels, especially the switch from expressing the miR-155-3p to the miR-155-5p strand. FUBP1 belongs to a family of RNA-binding proteins (FUBP family), including KSRP⁴². Previously, studies implicated KSRP in controlling miR-155-5p expression in macrophages^24,43 and dendritic cells⁴⁴. Ruggiero et al. reported that pre-miR-155 co-immunoprecipitated with KSRP and depletion of KSRP impaired the expression of mature miR-155-5p while simultaneously causing the accumulation of pri-miR-155 and pre-miR-155²⁴. KSRP, also known as Far Upstream Element Binding Protein 2 (FUBP2), binds to the terminal loop of pre-miR-155 via four distinct hnRNPK-Homology (KH) domains and promotes its maturation^24,45,46.

We now describe the roles of CELF2, FUBP1 and KSRP in controlling the expression of miR-155-5p and 3p. We used CRISPR-Cas9 mediated silencing to construct cells deficient in CELF2, FUBP1, or KSRP. Our data suggest that CELF2 is involved in the timing of miR-155-3p expression and affects the overall strand ratio of miR-155-3p and miR-155-5p. FUBP1 specifically inhibits miR-155-3p, while KSRP inhibits the expression of both strands. Finally, we show that FUBP1 and KSRP proteins interact with pre-miR-155 via its KH3 domain. These data provide insights into mechanisms that might underlie dysregulated miR-155-5p and miR-155-3p levels observed in certain diseases.

Results

LPS induces the association of FUBP1 to pre-miR-155

We^23,36 and others³⁴ have shown that IL10 inhibits LPS-induced miR-155-5p expression in macrophage cells. We further demonstrated that IL10 inhibits the maturation of pre-miR-155 to miR-155-5p²³ and that the RNA-binding protein CELF2 contributes to the process³⁶. However, our mass spectrometry-based examination of pre-miR-155-associated proteins also identified FUBP1 and KSRP as other proteins that might interact with pre-miR-155 in a stimulus (LPS ± IL10) dependent manner³⁶.

We analyzed pre-miR-155 pull-down samples for CELF2, FUBP1 and KSRP protein by immunoblotting, to follow up on these mass spectrometry-identified protein candidates. As described in our previous study³⁶the amount of CELF2 detected in the pre-miR-155 pull-down samples is the greatest in the LPS + IL10 samples (Fig. 1). LPS stimulation of the cells also increases the amount of FUBP1 protein in the pre-miR-155 pull-down samples; the addition of IL10 increases this further. KSRP levels in the pre-miR-155 pull-down samples appear to rise with LPS and further with IL10 addition, but these changes are statistically insignificant (Fig. 1). The levels of all three proteins in the total cell lysate remained constant in all stimulation conditions (Supplementary Fig. 1), suggesting that the association of these proteins with pre-miR-155 might represent post-transcriptional modifications of those proteins, which alters their ability to bind pre-miR-155.

Fig. 1.

Increased interaction of FUBP1 protein with pre-miR-155 in response to LPS stimulation. RAW264.7 cells were transfected with biotinylated pre-miR-155 oligonucleotide and stimulated with LPS ± IL10 for 1 h before collecting the samples. (A) Expression levels of CELF2, FUBP1 and KSRP protein interacting with pre-miR-155 oligonucleotide were determined by immunoblotting. (B)The graphs show the CELF2, FUBP1, and KSRP band intensities in the pull-down sample, normalized to the same protein in the total cell lysate. One-way ANOVA with Tukey’s correction calculated the significance of the difference between the stimulations. ** p < 0.01, * p < 0.05, ns = not significant. The data represent the results of 5 experiments.

CELF2, FUBP1 and KSRP deficiency differentially affects miR-155-5p and miR-155-3p expression

To investigate the role of CELF2, FUBP1 and KSRP in LPS and IL10 regulating miR-155-5p and miR-155-3p expression, we used CRISPR-Cas9-mediated targeting to knock down the expression of each of the three proteins in RAW264.7 cells. We transduced CRISPR-Cas9 expressing RAW264.7 cells with lentiviruses harbouring a blasticidin resistance gene and a guide RNA (gRNA) sequence targeting either FUBP1 or KSRP, as we described previously for CELF2 KD cells³⁶. Cells were selected with 10 µg/mL Blasticidin, and the bulk drug-resistant population was used for experiments. We call CRISPR-Cas9/gRNA transduced cells “KD” for knockdown rather than KO (for knockout) because we use a bulk, uncloned population, which may contain cells with different degrees of target gene deletion. We use a bulk population to reduce the impact of insertional mutagenesis effects, which the use of cell clones would magnify. As illustrated in Fig. 2, FUBP1 KD and KSRP KD cells express significantly less FUBP1 or KSRP protein than the control RAW264.7 cell.

Fig. 2.

Generation of FUBP1 and KSRP knockdown cells by CRISRP-Cas9 mediated gene silencing. RAW264.7/Cas9 cells transduced with FUBP1 sgRNA or KSRP sgRNA were treated with 2 µg/mL of doxycycline for 48 h to induce knockdown of FUBP1/KSRP proteins. FUBP1 and KSRP protein expressions were determined by immunoblotting. Data plotted represents FUBP1 and KSRP protein band intensity normalized to GAPDH. One-way ANOVA with Tukey’s correction determined the comparison indicated with braces. **** p < 0.0001, *** p < 0.001, ns = not significant. The data are representative of 3–5 experiments.

After confirming the reduction of FUBP1 and KSRP proteins in the KD cells, we stimulated the cells with LPS ± IL10 for 0, 2 and 4 h, isolated RNA, and quantified the levels of miR-155-5p and miR-155-3p by qPCR. Figure 3A shows every cell’s data normalized to the LPS-stimulated sample (2 h) of the control RAW264.7 cells, allowing comparison among all the cells. Figure 3B graphs the same data in a time course format for each cell type. Figure 3B shows that in RAW264.7 cells, the LPS-stimulated miR-155-3p levels are detectable at 2 h and plateaus by 4 h. In contrast, the LPS-stimulated miR-155-5p level significantly increases only by 4 h and continues to rise past 4 h (data not shown). As we described earlier³⁶. IL10 inhibited the expression of miR-155-5p at 4 h but not at 2 h. IL10 inhibited the expression of miR-155-3p at 4 h (Fig. 3 A and B), but the inhibition at 2 h was statistically significant when the comparison was within just the RAW264.7 cells (Fig. 3B).

Fig. 3.

CELF2, FUBP1, and KSRP deficiency alters the expression of miR-155-5p and miR-155-3p in response to LPS ± IL10 stimulation. FUBP1 KD, CELF2 KD, KSRP KD or the control RAW264.7 cells were stimulated with 1 ng/mL LPS ± 1 ng/mL of IL10 for (A) 2 and 4 h before total RNA extraction. The expression levels of miR-155-5p and miR-155-3p were determined by qPCR and normalized to snoRNA202 levels. The data plotted represents the expression levels of miR-155-5p and miR-155-3p levels normalized to the 2-hour LPS-stimulated control RAW264.7 cells. (B) The data from panel A were replotted as time courses for each cell type. miR-155-5p and miR-155-3p levels were normalized to the unstimulated sample of control RAW264.7 cells. Two-way ANOVA with Tukey’s correction determined the comparisons indicated with braces between different time points or cell types as * and the significance between stimulations as †. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, †††† p < 0.0001, ††† p < 0.001, †† p < 0.01, † p < 0.05, ns = not significant. The data are representative of 4 experiments.

Plotting the ratio of miR-155-3p to miR-155-5p helps illustrate the change in the selection of miR-155-3p and miR-155-5p strands for expression. Similar to Simmond et al.’s observation in human macrophages stimulated with LPS⁴¹we found that LPS treatment increases the ratio of miR-155-3p to miR-155-5p in control cells (Fig. 3A, rightmost graph). The miR-155-3p to miR-155-5p ratio is unchanged by IL10, suggesting IL10 controls miR-155 expression by inhibiting the general miR-155 maturation without altering the miR-155 strand selection. The LPS effect on miR-155-3p to miR-155-5p ratio disappears by 4 h.

Knockdown of CELF2 protein enhanced miR-155-5p expression at 2 and 4 h (Fig. 3A) compared to control RAW264.7 cells. miR-155-3p expression is slightly enhanced in CELF2 KD compared to RAW264.7 cells at 2 h but reduced at 4 h of LPS stimulation (Fig. 3A). Of note, CELF2 KD cells show a dramatic acceleration in the kinetics of miR-155-5p and miR-155-3p expression. Compared to the control and other cells, miR-155-5p expression in CELF2 KD cells appears to have already reached a plateau by 4 h (Fig. 3B). miR-155-3p expression peaked earlier at 2 h and decreased to basal expression level by 4 h. Interestingly, the miR-155-3p/miR155-5p ratio in CEL2 KD at 2 and 4 h (Fig. 3 A, right panel) significantly decreased compared to the control RAW264.7 cells. IL10 failed to inhibit the expression of either miR-155-5p or miR-155-3p in the CELF2 KD cells. The LPS-stimulated miR-155-3p/miR-155-5p ratio stays the same regardless of the addition of IL10.

In FUBP1 KD cells, LPS-stimulated miR-155-5p expression is unchanged compared to its action in control RAW264.7 cells (Fig. 3A). In contrast, the level of miR-155-3p is enhanced, suggesting a selective role of FUBP1 in regulating miR-155-3p expression, but the overall cellular ratio of miR-155-3p/miR-155-5p is unchanged. Like that observed in CELF2 KD cells, IL10 could not inhibit either miR-155-5p or miR-155-3p in FUBP1 KD cells. In the KSRP KD cells, miR-155-5p and miR-155-3p expression is significantly enhanced without a change in the relative expression. This elevation occurs without significant changes in pri-miR-155 or pre-miR-155 levels (Supplementary Fig. 2). Unlike that observed in FUBP1 and CELF2 KD cells, IL10 could inhibit miR-155-5p and miR-155-3p in cells lacking KSRP.

Expression of miR-155 target genes in CELF2, FUBP1 and KSRP KD cells

To assess the biological relevance of the changes in miR-155-5p and miR-155-3p, we examined the levels of Arg2 (a validated miR-155-5p target)^47,48; Apbb2 (a predicted target of miR-155-3p)^49,50; and Myd88 (a target of both miR-155 strands)^51,52. The expression of the miR-155 targets should decrease in cells where miR-155 levels are increased compared to RAW264.7 cells, and conversely increase in cells where miR-155 levels are reduced. As expected, Arg2 levels are lower in the CELF2 and KSRP KD cells (Fig. 4), where miR-155-5p is higher (Figs. 3A and 4-hour data). Apbb2 levels are higher (Fig. 4) in CELF2 KD cells, where miR-155-3p levels are lower (Figs. 3A and 4-hour data) than in RAW264.7 cells. As expected, myd88 levels are lower in all three KD cell types (Fig. 4) since miR-155-5p or miR-155-3p strands are elevated in all three protein KD cells (Fig. 3 A). Experiments are underway to distinguish between the effect of CELF2 and KSRP knockdown on the target genes being mediated by: (i) CELF2 and KSRP KD-dependent upregulation of miR-155-5p and miR-155-3p (Fig. 3), or (ii) the potential direct effects of CELF2 and KSRP1 on the target gene mRNA.

Fig. 4.

Expression of miR-155 target genes in CELF2 KD, FUBP1 KD, KSRP KD, or control RAW264.7 cells. Cells were stimulated with 1 ng/mL of LPS for 4 hours before extraction of total cellular RNA. The expression levels of Arg2, Apbb2 and Myd88 were determined by qPCR and normalized to GAPDH levels. Two-way ANOVA with Tukey’s correction determined the comparisons indicated with braces. **** p< 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant. The data represents three experiments.

IL10 regulation of inflammatory cytokine production in CELF2, FUBP1, and KSRP KD cells

Previously, we^24,36 and others³⁴ reported that miR-155 (miR-155-5p was assessed) deficiency increases LPS-induced inflammatory cytokine production^24,36. Therefore, we examined IL10’s action on TNFα production in FUBP1 and KSRP KD cells. Cells were stimulated with LPS ± 0.5 ng/mL or 10 ng/mL IL10 for 1 h, and TNFα levels in the culture supernatants were quantified by ELISA. Figure 5 shows that in control RAW264.7 cells, IL10 inhibits LPS-induced TNFα expression in a dose-dependent manner. However, in FUBP1 KD cells, IL10 could not decrease LPS-induced TNFα levels at all the concentrations of IL10 tested. This impairment of IL10 action suggests FUBP1 is essential for IL10 inhibition of LPS-induced TNFα. We previously showed that CELF2 deficiency partially impaired IL10-dependent TNFα inhibition³⁶. In contrast, the absence of KSRP protein resulted in higher LPS-stimulated TNFα production than in RAW264.7 cells, but IL10 could inhibit TNFα levels.

Fig. 5.

FUBP1 and KSRP deficiency alters the expression of TNFα in response to LPS ± IL10. FUBP1 and KSRP KD cells were stimulated with 1 ng/mL LPS ± indicated concentration of IL10 for 1 h before collecting the cell culture supernatant. The level of TNFα in the supernatants was determined by ELISA. Two-way ANOVA with Tukey’s correction determined the comparisons indicated with braces. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant. The data are representative of 3–5 experiments.

SHIP1 and STAT3 dependence of miR-155-5p and miR-155-3p expression

We previously showed that IL10 receptor signalling requires SHIP1 and STAT3 to inhibit miR-155-5p production in LPS-stimulated macrophages^23,33. However, we had not examined the expression of the alternate miR-155 strand, miR-155-3p. Thus, we extracted peritoneal macrophages (perimacs) from SHIP1 or STAT3 wild type (WT) or knockout (KO) mice, rested the cells for 2 h, stimulated the cells with LPS ± IL10 for 4 h, isolated RNA, and quantified the levels of miR-155-5p and miR-155-3p.

Figure 6 A shows the miR-155-5p and miR-155-3p expression in perimacs from STAT3 WT/KO (both C57Bl6 background) and SHIP1 WT/KO (both Balb/C background) mice 4 h after LPS ± IL10 stimulation. The qPCR data were normalized to the LPS-stimulated WT sample to compare the expression in WT and KO cells. In STAT3 WT and KO perimacs, the basal and LPS-stimulated levels of miR-155-3p are similar in the corresponding sample of each cell type (Fig. 6A). However, LPS-stimulated levels of miR-155-5p are reduced in the STAT3 KO, and IL10 could not inhibit the expression of either miR-155-5p or miR-155-3p in STAT3 KO cells and instead increased their levels in STAT3 KO cells. Normalizing the qPCR gene expression data to each cell type’s LPS-stimulated sample also reveals the effects of STAT3 deficiency on LPS and IL10 regulation of miR-155-5p and miR-155-3p. (Fig. 6B).

Fig. 6.

STAT3 and SHIP1 dependence of miR-155-5p and miR-155-3p expression at 4 h. miR-155-5p or miR-155-3p expression levels in STAT3 WT and KO, or SHIP1 WT and KO cells were determined by qPCR and normalized to snoRNA202 levels. (A) Data were plotted to represent the RNA expression normalized to the LPS-stimulated sample of the STAT3 WT. (B) The data in panel A was replotted with RNA expression normalized to each cell’s own LPS-stimulated sample. Two-way ANOVA with Tukey’s correction determined the comparison between the stimulations indicated with braces. **** p < 0.0001, *** p < 0.001, ** p < 0.01, ns = not significant. The data are representative of 3 experiments.

In SHIP1 knockout cells, LPS-stimulated expression of both miR-155-5p and miR-155-3p decreases compared to SHIP1 WT cells. (Fig. 6A). To see if IL10 could inhibit the LPS-induced expression of miR-155 in the SHIP1 KO, we also normalized the RNA expression data to each cell’s own LPS-stimulated sample. Figure 6B shows that IL10 inhibits both miR-155-5p and miR-155-3p levels in SHIP1 KO cells, but the inhibition is less effective compared to SHIP1 WT cells. These observations suggest that IL10 requires SHIP1 to inhibit miR-155-5p and that SHIP1 enables the LPS-induced expression of both miR-155-5p and miR-155-3p.

We then examined the kinetics of miR-155-5p and miR-155-3p expression at 0, 1, 2 and 4 h (Fig.7). We normalized the qPCR data to the 0-hour time point of the appropriate WT cell control. In STAT3 WT cells stimulated with LPS, miR-155-3p rises rapidly and reaches a plateau at 2 h, while miR-155-5p rises more slowly but continues to rise towards 4 h. The presence of IL10 (LPS + IL10) reduced the levels of miR-155-3p but not miR-155-5p in STAT3 WT cells. The different miR-155-5p and miR-155-3p kinetics resemble those observed by Simmonds et al. in human peripheral blood-derived macrophages⁴¹. In STAT3 KO cells, miR-155-5p and miR-155-3p levels stimulated by LPS show similar kinetics to the STAT3 WT. Remarkably, the presence of IL10 (LPS + IL10) enhanced rather than inhibited the expression of both miR-155-5p and miR-155-3p. We did not look at longer time points because the effect of autocrine cytokines will contribute to gene expression at longer stimulation times.

Fig. 7.

Kinetics of miR-155-5p and miR-155-3p expression in STAT3/SHIP1 WT and KO cells. The expression level in SHIP1 WT and KO cells of miR-155-5p or miR-155-3p was determined by qPCR and normalized to snoRNA202 levels. (A) Kinetics of miR-155-5p and miR-155-3p expression with RNA normalized to the WT 0-hour sample. (B) The data plotted represents the comparison of LPS-stimulated miR-155 levels at 4 h between different cell types. (C) The data plotted represents the ratio of miR-155-3p and miR-155-5p. Two-way ANOVA with Tukey’s correction determined the significance of the difference in values (if any) in the LPS ± IL10 stimulated sample simultaneously. **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant. The data are representative of 3 experiments.

In SHIP1 WT cells, LPS-stimulated miR-155-3p can be detected by 1 h and continues to rise towards 4 h. LPS induced high levels of miR-155-5p at 4 h. In SHIP1 WT cells, IL10 inhibited the LPS-stimulated expression of miR-155-3p and miR-155-5p at all time points. In SHIP1 KO cells, both miR-155 strands show elevated basal levels and LPS-induced miR-155-5p is significantly reduced compared to WT. LPS induces miR-155-3p levels to a similar extent as in SHIP1 WT cells; however, miR-155-3p becomes undetectable at 4 h. IL10 was impaired in reducing miR-155-5p levels and completely failed to inhibit miR-155-3p levels at all time points.

As Fig. 7A shows, the kinetics of both miR-155-5p and miR-155-3p in the SHIP1 WT cells are slightly delayed compared to those in STAT3 WT cells. SHIP1 WT cells also have considerably lower miR-155-5p levels (Fig. 7B) than STAT3 WT, and miR-155-3p/miR-155-5p ratio is about two-fold (Fig. 7 C) higher than the STAT3 WT cells. Furthermore, STAT3 WT shows no difference in the ratio of miR-155 strands between LPS and IL10 stimulation, but SHIP1 WT shows a significant increase in the ratio in IL10 stimulation. This difference might reflect the different genetic backgrounds of the SHIP1 WT/KO (Balb/C) and STAT3 WT/KO (C57BL/6) mice.

The KH3 domains of both FUBP1 and KSRP interact with pre-miR-155

The KH domains in the FUBP family of RNA-binding proteins, which include FUBP1 and KSRP, have been described as binding RNA⁵³. Ruggiero et al. showed that the RNA-binding protein KSRP interacts with pri-miR-155 and pre-miR-155 to regulate miR-155 expression^24,44. However, since previous research has not described FUBP1 as participating in miRNA processing, we examined its ability to bind to pre-miR-155. KSRP primarily interacts with RNA through its third KH domain (KH3), and substituting the key GXXG residues of the KH domain with GDDG significantly reduces the interaction of KSRP with RNA⁴⁶. To test whether FUBP1 interacts with pre-miR-155 through its KH3 domain, we generated recombinant WT and KH3 domain GDDG mutants for FUBP1 and KSRP (Fig. 8 A). We then measured the interaction of these recombinant proteins to pre-miR-155 using biolayer interferometry (BLI). Figure 8B and C show that KSRP WT and FUBP1 WT interact with pre-miR-155, but the KH3 GDDG mutants of KSRP and FUBP1 do not.

Fig. 8.

The KH3 domain of KSRP and FUBP1 are essential for interacting with pre-miR-155. We expressed and purified KSRP/FUBP1 wildtype (WT) or KSRP/FUBP1 KH3 GDDG mutant proteins as described in the Materials and Methods section. (A) Coomassie Blue gel staining was used to assess the purity of the purified protein. (B) BLI biosensors loaded with biotinylated pre-miR-155 were dipped in wells containing KSRP WT, KSRP KH3 GDDG mutant, FUBP1 WT, FUBP1 KH3 GDDG mutant proteins for 10 min, followed by a dissociation step in the assay buffer for 20 min. (C) Data plotted to represent the BLI response of indicated proteins with biotinylated pre-miR-155. Unpaired Student’s t-test determined the comparison between the WT and KH3 GDDG mutant. **** p < 0.0001. The data represent the results of 3–5 experiments.

Discussion

Most investigators have focused on studying the more abundant miR-155-5p strand, which is elevated in inflammatory disease and cancer^2,5,6,34while miR-155-3p was thought to be a by-product or passenger strand¹⁸. However, Simmonds et al. were the first to report that miR-155-3p is incorporated in RISC in LPS-stimulated human macrophages, indicating the non-dominant miR-155-3p is regulated and biologically functional⁴¹. In LPS-activated macrophages, miR-155-3p expression peaks at earlier LPS stimulation (4 h) and disappears by 24 hours^41,44. In contrast, miR-155-5p levels do not rise until about 2–4 h and remain detectable in higher numbers at 24 hours^41,44. These kinetics suggest that miR-155-3p and miR-155-5p, which target distinct RNAs, play different roles in regulating cell function. Moreover, overexpression of miR-155-3p has also been reported in various cancer and disease cells¹⁸. The dysregulated increase of miR-155-3p may be due to enhanced expression of the MIR155HG primary transcript⁵⁴elevating both miR-155-5p and miR-155-3p strands^55–58. However, in some cases, miR-155-3p is upregulated preferentially over miR-155-5p^12,14,59,60so the ratio of the strands expressed and target availability may be important factors determining the overall effects of miR-155 dysregulation on cancer growth⁶¹ and disease progression⁶². For these reasons, we characterized the proteins involved in controlling the maturation of pre-miR-155 to miR-155-5p and miR-155-3p.

Our examination of the CELF2, FUBP1, and KSRP KD cells revealed different miR-155-5p and miR-155-3p expression patterns compared to the control RAW264.7 cells. (1) Without CELF2 protein, LPS-induced expression of miR-155-5p increases significantly compared to control RAW264.7 cells at 2 and 4 h. In contrast, miR-155-3p is modestly increased by 2 h and decreased compared to control RAW264.7 cells at 4 h. The resulting decrease in the miR-155-3p/miR-155-5p ratio suggests that the CELF2 protein plays a role in strand switching to miR-155-3p from miR-155-5p (Fig. 3). (2) Knocking down FUBP1 protein increased miR-155-3p expression, but the overall cellular ratio of miR-155-3p/miR-155-5p remains unchanged. This observation suggests that the FUBP1-mediated miR-155-3p molecules may have subcellular compartment/location-specific functions, as reported for other miRNAs⁶³. (3) The absence of KSRP protein resulted in a substantial increase in the levels of both strands of miR-155. KSRP may thus be a general inhibitor of the miR-155 maturation process.

Our previous studies showed that the CELF2 protein inhibits miR-155-5p maturation³⁶similar to the CELF2 function on miR-140 maturation⁴⁰. However, our current studies showed that the CELF2 protein differentially regulates the miR-155 strands (Figs. 3 and 4). Without CELF2 protein, levels of miR-155-5p are enhanced, but miR-155-3p is reduced (Figs. 3 and 4). We have also looked at the levels of the primary transcript of miR-155 (pri-miR-155) and the precursor of miR-155 (pre-miR-155) in these cells (Supplementary Fig. 2). The pri-miR-155 and pre-miR-155 levels in CELF2 KD cells were similar to that in control RAW264.7 cells, suggesting that the change in miR-155 levels occurs at the stage of maturation of miR-155. These observations are consistent with CELF2 protein favouring the selection of miR-155-3p at early time points (2 h) after LPS stimulation. After 2 h, miR-155-3p levels fall, and miR-155-5p levels rise. Studies are underway to determine how the CELF2-dependent selection of miR-155-3p transitions to allow miR-155-5p expression.

The CELF2 protein has been described as possessing anti-tumour functions, with low expression of this protein detected in various cancer cells^64,65. The overexpression of CELF2 results in a reduction of tumour size^64,66–68. Elevated expression of miR-155-5p^6,69–72 and miR-155-3p^18,56,57 have been observed in many cancers. Decreased CELF2 protein expression may underlie the high levels of miR-155-5p and miR-155-3p in certain cancers^52,55,57,61. Future studies investigating the correlation between miR-155-5p and miR-155-3p levels and CELF2 expression levels in various cancer types will provide further insight into the regulation of miR-155 and the role of the CELF2 protein in this process.

We concluded that KSRP inhibits the expression of miR-155-5p and miR-155-3p based on our data, which shows that LPS-stimulated levels of both are elevated in KSRP KD cells compared to control cells. Zhou et al.⁴⁴ and Ruggiero et al.²⁴ also examined the role of KSRP in regulating miR-155. Zhou et al. showed that the knockdown of KSRP via siRNA in plasmacytoid dendritic cells increased expression miR-155-3p in response to the TLR7 agonist R837, which is consistent with our data⁴⁴. For miR-155-5p, however, they observed a decrease in its expression. This disagreement with our observed increase in miR-155-5p in KSRP KD cells may be due to the differences in cell type (macrophages vs. dendritic cells) and agonist (LPS, a TLR4 ligand vs. R837, a TLR7 ligand) used. Ruggiero et al. studied LPS-stimulated miR-155-5p expression in KSRP siRNA knockdown macrophage cells²⁴. In their study, siRNA-mediated knockdown of KSRP protein reduced miR-155-5p levels while pri-miR-155 levels increased, suggesting that KSRP facilitates the maturation of miR-155-5p²⁴. Both observations differ from ours, where we found that KSRP KD cells result in increased miR-155-5p levels (Fig. 3) with no effect on pri-miR-155, consistent with KSRP inhibiting the maturation of miR-155-5p (Supplemental Fig. 1).

The discrepancy between our observations and Ruggiero’s might be due to differences in experimental approaches. Ruggiero et al. used super-physiological concentrations of LPS (100 ng/mL) compared to the physiological concentration of LPS (1 ng/mL) used in our work. Very high ligand concentrations can alter the kinetics of gene/protein expression, including the production of autocrine cytokines such as IL10^73,74, which can suppress miR-155 levels^23,34,36. Autocrine factors can lead to feedback effects not observed at lower ligand concentrations. Another experimental variation involves the nucleotide sequences used by Ruggiero et al. versus those used by us for silencing KSRP expression. We designed our KSRP targeting sequences based on the online Synthego CRISPR Design Tool, and a nucleotide BLAST (nBLAST) search confirms their specificity to the KSRP locus (Supplementary Fig. 3 and Supplementary Fig. 4). However, a nBLAST search of the sequence Ruggiero et al. used to silence KSRP mRNA showed complementarity to several other RNAs other than the intended KSRP mRNA (Supplementary Tables 1 and Supplementary Fig. 4). Among the off-target RNAs, Gm35853 is a long noncoding RNA (lncRNA) that is abnormally increased in hematopoietic stem and progenitor cells from a murine model of lupus compared to control mice⁷⁵. Overexpression of miR-155 has been implicated in patients with systemic lupus erythematosus⁷⁶. Thus, one could speculate that increased levels of GM35853 may lead to elevated miR-155. Then, reducing lncRNA Gm35853 by siRNA KSRP could decrease miR-155 levels, as observed by Ruggiero et al..

Our data show that FUBP1 suppresses miR-155-3p expression (Fig. 3 A) with no significant influence on miR-155-5p. Since the number of miR-155-3p molecules in cells is much less than that of miR-155-5p (Fig. 3, and Simmonds et al.)⁴¹the overall cellular ratio of miR-155-3p/miR-155-5p remains similar in FUBP1 KD and control RAW264.7 cells (Fig. 3 A, right-hand graph). These observations suggest that the subcellular localization of miR-155-3p may be necessary for its biological action. miRNAs primarily localize in the cytoplasm of cells, exerting their best-described function as post-transcriptional regulators by binding to target mRNAs to inhibit their translation or promote their degradation⁷⁷. However, recent research has revealed that miRNAs can also be found in the nucleus, mitochondria, RNA granules and endomembranes, suggesting potential roles beyond cytoplasmic regulation depending on their subcellular localization^78,79.

We found that the KH3 domain of FUBP1 participates in binding to pre-miR-155, which is analogous to the requirement of the KSRP KH3 domain for binding to miR-155-5p²⁴. The KH3 domain of FUBP1 and KSRP share 79% sequence homology (Fig. 9 A). Thus, KSRP and FUBP1 may recognize the same or similar sequences in pre-miR-155. The KSRP KH3 domain has been shown to interact with the terminal loop of the pre-miR-155^24,44,80 and the known consensus sequence for the FUBP1 KH3 domain (UUGUG)⁸¹ seems to interact with the stem of the pre-miR-155 (Fig. 9B). Interestingly, the sequence recognition motif for CELF2 protein (UGU-N_x-UGU)⁸² exactly overlaps two of three expected FUBP1 recognition motifs in the pre-miR-155 sequence, suggesting the possibility of FUBP1 to compete with CELF2 for binding to pre-miR-155.

Fig. 9.

Schematic representation of FUBP1/KSRP/CELF2 regulation, homology, and interaction. (A) The sequence alignment of the KH3 domain of FUBP1 and KSRP protein. Sequences in black and red indicate matching and mismatching sequences, respectively. (B) Schematic diagram showing the predicted interaction site of FUBP1, CELF2 and KSRP to pre-miR-155. The bases of pre-miR-155 are gray; blue bases show the miR-155-5p sequence, and black bases indicate the miR-155-3p sequence.

As a next step in investigating the control of miR-155-5p and miR-155-3p expression in macrophage functions, we examined the contribution of IL10 receptor signalling. IL10 receptor signalling involves both STAT3 and SHIP1-STAT3 pathways³⁰. Our studies showed that in the absence of STAT3 protein, IL10 no longer inhibits either miR-155-5p or miR-155-3p production but instead enhances expression of both by 4 h and 2 h, respectively (Fig. 7 A). In the absence of SHIP1, the LPS-induced level of both miR-155 strands is significantly reduced, IL10-dependent inhibition of miR-155-5p is partly impaired and IL10-dependent inhibition of miR-155-3p is abolished (Fig. 6B). Our data suggest that STAT3 and SHIP1 underlie the regulation of miR-155 by the IL10R pathway, as summarized in Fig. 10. We found that in WT mouse macrophages, LPS-induced miR-155-3p expression increases more rapidly than miR-155-5p (Fig. 7). In the C57BL/6 mouse (STAT3 WT), miR-155-3p reaches a plateau at around 2 h. In the Balb/C mouse (SHIP1 WT), the miR-155-3p level continuously rises to 4 h. The difference in miR-155-3p kinetics between STAT3 WT and SHIP1 WT may be attributed to the variation in mouse strain (C57BL/6 vs. Balb/C), which contributes to the described differences in immune response between these strains^83,84.

Fig. 10.

Schematic representation of miR-155 regulation. Schematic diagram representing the regulation of miR-155-5p and miR-155-3p expression in response to LPS and IL10. The proteins in blue font represent the protein required for the shown action, and the proteins in orange font represent proteins that suppress the shown action.

We had previously examined the kinetics of IL10 inhibition of LPS-induced TNFα expression³⁰. We demonstrated that the early (< 2 h) response required a SHIP1:STAT3 complex, whereas the late phase (> 2 h) only required STAT3. In the current study, we support the previous findings⁴¹ that LPS-induced miR-155-3p was detected earlier (2–4 h) than miR-155-5p (> 4 h). Furthermore, in the absence of STAT3 or SHIP1 protein, IL10-dependent inhibition of miR-155-3p is abolished, whereas, for miR-155-5p, SHIP1 KO only resulted in partial impairment of IL10-dependent inhibition of miR-155-3p (Fig. 7 A). These observations suggest that IL10 inhibition of miR-155-3p may be necessary for IL10 inhibiting the early (SHIP1:STAT3 dependent) phase of TNFα production. In contrast, miR-155-5p inhibits the late (STAT3-dependent) phase of TNFα expression. However, the mechanism by which STAT3 or SHIP1:STAT3 complexes control miR-155-5p and miR-155-3p remains to be determined. One possibility is that both lead to protein expression that participates in the control of pre-miR-155 processing. In fact, since the absence of STAT3 results in IL10 enhancing rather than reducing LPS-induced miR-155-5p and miR-155-3p levels, we predict these STAT3-induced proteins may suppress pre-miR-155 processing (Fig. 11).

Fig. 11.

Schematic representation of IL10R downstream signaling process and potential mechanism on a different phase of pro-inflammatory response. Schematic diagram of IL10R downstream signaling pathway affecting various stages of pro-inflammatory response.

Work is underway to determine how CELF2, SHIP1, and STAT3-dependent pathways affect miR-155 strand selection. Two strands of miRNA can be produced from a single precursor RNA molecule, but only one strand is selected to load into the RNA-induced silencing complex⁸⁵. Strand selection of miRNAs may occur due to a change in the cleavage site of Dicer⁸⁶ or through actions of RNA editing enzymes on the pri/pre-miR-155 molecule⁸⁵ that affects the thermodynamic stability of the strands or its interaction with RNA binding proteins. Altered thermodynamic stability or Dicer cut site may result in selective processing of one or the other strand into the one that becomes the guide strand⁸⁷. In the case of miR-155, the CELF2 protein exhibits strand selection characteristics, favouring the expression of miR-155-3p, while FUBP1 favours the expression of miR-155-5p. We will also investigate whether RNA editing contributes to the expression of miR-155-5p versus miR-155-3p and whether these differences are involved in the aberrant expression of miR-155-5p versus miR-155-3p observed in certain cancers^52,55,57,61. These investigations will also provide insight into regulating the expression of other miRNAs, including those controlled by CELF2 and KSRP^{40,43,80,88,89}.

Methods

Mouse colonies

SHIP1 WT or SHIP1 KO (in the BALB/c background) mice were provided by Dr. Gerald Krystal (BC Cancer Research Centre, Vancouver, BC). STAT3 WT and KO mice (in the C57BL/6 background) were generated as described³⁰. All mice were maintained following the animal care protocols approved by the University of British Columbia Animal Care Committee. Mice were euthanized according to the University of British Columbia’s standard operating procedure #ACC-03-2012 Euthanasia of Adult Rodents using Inhalant Anesthetic, followed by Carbon Dioxide.

Cell lines

The murine cell line RAW264.7 (ATCC TIB-71) was maintained in Roswell Park Memorial Institute 1640 medium (RPMI-1640, SH30027, HyClone, Logan, UT) supplemented with 9% fetal bovine serum (FBS, SH30396, HyClone, Logan, UT). As described below, FUBP1 KD RAW264.7 cells were generated by transduction of RAW 264.7 expressing iCas9 (called RAW264.7 control cells). The generation of CELF2 KD RAW264.7 cells was described in Yoon et al.³⁶.

Construction of the FUBP1 or KSRP-pLX-sgRNA targeting vectors

The FUBP1 sgRNA targeting sequence (5’ GCTAAATCCGACCATCCCATC) or KSRP sgRNA targeting sequence (5’ TCCGGTGTTGTAGTCCGACA) was designed using the CRISPR Gold online tool⁹⁰ or Synthego CRISPR Design Tool (https://design.synthego.com/#/). Oligonucleotides corresponding to this sequence were cloned into the pLX-sgRNA vector using overlap-extension PCR as described³⁶. The pLX-sgRNA vector with target-specific insert was transformed into chemically competent Stbl3 E. coli cells, and colonies were selected using ampicillin. The resulting FUBP1-pLX-sgRNA construct was confirmed by sequencing.

Generation of RAW 264.7 cells expressing FUBP1 or KSRP pLX-sgRNA

The FUBP1-pLX-sgRNA vector was transduced into RAW264.7 cells expressing iCas9 (called RAW264.7 control cells)³⁶ using lentiviruses. Lentiviruses harbouring FUBP1-pLX-sgRNA were prepared by co-transfecting FUBP1-pLX-sgRNA vector into HEK293T (ATCC CRL-3216) cells with the packaging plasmid R8.9 and VSVG. 24 h after the transfection, the supernatant was collected and incubated with RAW264.7 control cells in the presence of 8 µg/mL protamine sulfate. Cells transduced with FUBP1-pLX-sgRNA viruses were selected using 10 µg/mL blasticidin2 µg/mL doxycycline was added to the culture media for up to 48 h to induce the expression of Cas9 and knockdown of FUBP1.

RNA-oligonucleotide transfection and RNA pull-down assay

RAW264.7 cells (ATCC TIB-71) were seeded at 8.4 × 10⁶ cells per 10 cm dish a day before the transfection. Biotinylated pre-miR-155 oligonucleotides (Biotin-UAAUUGUGAUAGGGGUUU UGGCCUCUGACUGACUCCUACCUGUUA) were obtained from Invitrogen Life Technologies (ThermoFisher Scientific, Nepean, ON). RNA oligonucleotides and Lipofectamine-3000 (L3000-015, ThermoFisher Scientific, Nepean, ON) were prepared in Opti-MEM (31985, ThermoFisher Scientific, Nepean, ON) separately. They were mixed at a 1:1 (v/v) ratio and incubated at room temperature for 20 min to allow the formation of Lipofectamine-oligonucleotide complexes. 1 µg of RNA-oligonucleotide to 1.125 µL of Lipofectamine-3000 was used. After incubation, the RNA-Lipofectamine solution was diluted ten-fold with 9% FBS/RPMI-1640 and added to cells. The cells were incubated with the transfection solution in a chamber at 37 °C supplemented with 5% CO₂. After 6 hours, the solution was replaced with 9% FBS/RPMI-1640, and cells were allowed to recover overnight.

Following the 9% FBS/RPMI-1640 overnight incubation, the transfected RAW264.7 cells were stimulated with 1 ng/mL LPS (Escherichia coli serotype 0111:B4; Millipore-Sigma, Oakville, ON) ± 50 ng/mL IL10 for the indicated length of time in a chamber at 37 °C supplemented with 5% CO₂. Following stimulation, the media was removed, and cells were chilled with cold (4 °C) phosphate-buffered saline (PBS, SH30256, ThermoFisher Scientific, Nepean, ON) for 2 min. Next, the PBS was removed, and the cells lysed by the addition of protein solubilization buffer (PSB, 50 mM HEPES, 100 mM NaF, 10 mM NaPPi, 2 mM NaVO₄, 2 mM NaMoO₄, 4 mM EDTA, 0.125% Triton X-100, protease inhibitor cocktail (11836145001, Millipore-Sigma, Oakville, ON), and 0.5 mM Tris(2-carboxyethyl)phosphine (TCEP, M115, Soltec Ventures, Beverly, MA)). Cell lysates were collected with a cell scraper and gently agitated at 4 °C for 30 min. Insoluble material was removed by centrifugation at 12,000 RPM for 20 min at 4 °C.

Clarified cell lysates were added to streptavidin magnetic beads (1164786001, Millipore-Sigma, Oakville, ON) in 1.5 mL microfuge tubes and incubated for 90 min at 4 °C on a nutator. The tubes were then briefly centrifuged at 5,000 RPM, and magnetic beads were immobilized using a magnetic tube stand (12321D, ThermoFisher Scientific, Nepean, ON). Lysates were removed, and the beads were resuspended in the wash buffer (0.1% Tween-20 containing PSB) and rocked for 5 min at 4 °C on a nutator. The washing was repeated 3 times. The proteins were eluted by boiling 2x SDS-PAGE sample buffer (0.125 M Tris, pH 6.8, 5% 2-mercaptoethanol, bromophenol blue, 13.5% glycerol, 4.5% SDS) for immunoblot analysis.

Immunoblot analysis

Proteins were separated by 10% SDS-PAGE, followed by electroblotting onto polyvinylidene fluoride (PVDF) membrane (IPFL00010, Millipore-Sigma, Oakville, ON). The membranes were blocked in 3% bovine serum albumin (BSA), then probed with the following primary antibodies overnight: 1:1000 KSRP (ab140648, Abcam, Toronto, ON), 0.1 µg/mL GAPDH (G9545, Millipore-Sigma, Oakville, ON), 1 µg/mL CELF2 (Sc-47731, Santa Cruz, Dallas, TX) and 1 µg/mL FUBP1 or KSRP (Sc-136137, Santa Cruz, Dallas, TX). The membranes were washed three times in Tris-buffered saline containing 0.05% Tween-20 (TBST), incubated with either Alexa Fluor 660 anti-mouse IgG (A21055) or Alexa Fluor 680 anti-rabbit IgG (A21109, Invitrogen, Burlington, ON), and imaged using a LI-COR Odyssey Imager.

Isolation of and stimulation of mouse peritoneal macrophages

Primary peritoneal macrophages (perimacs) were isolated from mice by peritoneal lavage with 3 mL of sterile PBS. Perimacs were seeded at 2.0 × 10⁶ cells per well in a 6-well tissue culture plate or 1.18 × 10⁶ cells per well in a 24-well tissue culture plate in Iscove’s Modified Dulbecco’s Medium (IMDM, SH30228, HyClone, Logan, UT) supplemented with 10% FCS. Cells were allowed to adhere for 2 h, rinsed with room temperature PBS to remove non-adhered cells, and fresh media was added. The media was changed after 1 h, and the cells were stimulated with 1 ng/mL LPS ± 1 ng/mL IL10 for 1, 2, or 4 h. Triplicate wells were used for each stimulation condition.

RNA extraction and qPCR

Total RNA was extracted using Tri-Reagent (T9424, Millipore-Sigma, Oakville, ON) according to the manufacturer’s instructions. 1–3 µg of RNA was treated with RNase-free DNase I (04716728001, Millipore-Sigma, Oakville, ON) for 20 min at 37 °C, followed by the addition of 0.1 M EDTA to a final concentration of 8 mM to inactivate DNase I.

For measurement of miR-155-5p, miR-155-3p, and small nucleolar RNA MBII-202 (snoRNA202) levels, 20 ng of DNase I treated RNA was used to generate cDNAs using the miRNA TaqMan Reverse Transcription Kit (4366597, ThermoFisher Scientific, Nepean, ON), Multiscribe™ reverse transcriptase (4319983, ThermoFisher Scientific, Nepean, ON), and miR-155-5p (002571), miR-155-3p (464539_mat) or snoRNA202 (001232, ThermoFisher Scientific, Nepean, ON) primers according to the manufacturer’s instructions. qPCR quantification of miR-155-5p and miR-155-3p and snoRNA202 cDNA was performed using the TaqMan fast advanced master mix (4444557, ThermoFisher Scientific, Nepean, ON) and the appropriate TaqMan probes on a StepOne Plus™ instrument (4376582, Invitrogen, Burlington, ON). miRNA levels were analyzed using the comparative CT method with snoRNA202 as the normalization control.

For measurement of pri-miR-155, pre-miR-155, and GAPDH, 200 ng of DNase I treated RNA were reverse transcribed using SuperScript™ IV reverse transcriptase (18090200, ThermoFisher Scientific, Nepean, ON), oligodT primer (18418020, ThermoFisher Scientific, Nepean, ON) and random hexamers. qPCR quantification of pri-miR-155, pre-miR-155, and GAPDH were achieved with primers for pri-miR-155, pre-miR-155, and GAPDH in conjunction with the SYBR Green master mix (100029284, ThermoFisher Scientific, Nepean, ON). RNA levels were analyzed using the comparative CT method with GAPDH as the normalization control.

Molecular cloning of FUBP1 and KSRP

The open reading frame (ORF) of mouse FUBP1 and KSRP was obtained by PCR on cDNA generated from RNA isolated from SHIP1 WT perimacs. The FUBP1/KSRP ORF was inserted into the Gateway entry vector (pENTR1A, Invitrogen, Burlington, ON) via restriction digest and ligation. The products were transformed into DH5α E. coli chemically competent cells, and colonies were selected using 50 µg/mL of kanamycin. The sequences of the FUBP1/KSRP pENTR1A vectors were confirmed by sequencing. Using site-directed mutagenesis, the GXXG sequences in the third KH domain of FUBP1 and KSRP were mutated to the GDDG sequence. PCR with non-overlapping primers (5΄ phosphorylated forward primer with desired mutation and non-overlapping reverse primer) was used to generate FUBP1/KSRP daughter plasmids containing the desired mutation. The resulting PCR product was extracted with phenol-chloroform, treated with Dpn I to remove the parental vector, daughter plasmids ligated with T4 ligase, and transformed into DH5α chemically competent cells. The FUBP1/KSRP WT and KH3 GDDG mutant pENTR1A vectors were transferred to lentiviral vector FUGWBW using Gateway LR reactions as described³⁰.

Recombinant FUBP1 and KSRP protein expression

HEK293T cells were transfected with FUBWBW vectors encoding FUBP1/KSRP WT or KH3 GDDG mutant proteins. After transfection (48 h), the cells were lysed with the lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM imidazole, 1% NP-40, protease inhibitor cocktail, and 10 mM TCEP). The lysates were incubated for 45 min at 4 °C on a nutator, then centrifuged at 12,000 rpm for 20 min, and the supernatants were transferred to tubes containing cobalt affinity beads (TALON metal affinity resin, #635504, Takara Bio, Mountain View, CA). The cell lysates were incubated with the beads for 2 h, and the beads were washed three times with wash buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM Imidazole, 0.1% NP-40, 0.5 mM TCEP) before eluting with the elution buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 150 mM imidazole, 0.5 mM TCEP, and 0.02% Tween-20).

Measurement of TNFα production

Cells were seeded at 2.0 × 10⁴ cells per well in a 96-well tissue culture plate and allowed to adhere overnight. The media was changed the next day, 1 h before stimulation. Cells were stimulated with 1 ng/mL of LPS ± indicated concentration of IL10 for 1 h. Triplicate wells were used for each stimulation condition. The supernatant was collected, and secreted TNFα protein levels were measured using a BD OptEIA Mouse TNFα Enzyme-Linked Immunosorbent Assay (ELISA) kit (558534, BD Biosciences, Mississauga, ON).

Biolayer interferometry

The binding affinity between the FUBP1 or KSRP proteins to pre-miR-155 was examined using biolayer interferometry with super-streptavidin (SSA) biosensor tips (18-5057, ForteBio, Fremont, CA). SSA biosensor tips were hydrated in BLI assay buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5 mM TCEP, 0.2% Tween-20) before coating with biotinylated pre-miR-155 and blocking with 0.1% BSA. The kinetic measurements were done at 30 °C with an orbital flow of 1000 rpm. A 60-second baseline was established using BLI assay buffer. The pre-miR-155 coated biosensors were then dipped into wells containing KSRP or FUBP1 protein, and association was monitored for 600 s. The sensors were then transferred to wells containing only BLI assay buffer, and protein dissociation was monitored for 600 s. The raw data were analyzed using the Octet Red Data Analysis software (ver. 8.2).

Statistical analysis

Band intensities were quantified in immunoblots using LI-COR Odyssey imaging system and Image Studio™ Lite software (LI-COR Biosciences, Lincoln, NE). Graph-Pad Prism 6 (GraphPad Software Inc., La Jolla, CA) performed all statistical analyses. Statistical details can be found in figure legends. Values are presented as means ± standard deviations. Unpaired Student’s t-tests were used where appropriate to generate two-tailed P values. Two-way ANOVA was performed where required with appropriate multiple comparison tests. The differences were considered significant when p ≤ 0.05. The full uncropped blots and SDS-gel images for Figs. 1, 2 and 3 appear in Supplemental Figs. 5, 6 and 7 respectively.

Ethical statement

All experimental protocols were approved by the relevant University of British Columbia (UBC) research safety and ethics committee. Protocols followed relevant UBC guidelines and regulations. The perimacs were derived from mice in accordance with the animal care protocol (A21-0203) approved by the University of British Columbia Animal Care Committee. The cell culture experiments are done in accordance with UBC Biosafety requirements (B16-0206). All methods are reported in accordance with ARRIVE guidelines.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, J.Y., and A.M.; methodology, J.Y.; validation, J.Y., T.C., N.L. and A.M.; formal analysis, J.Y.; investigation, J.Y., N.C. and A.M.; Resources, J.Y., and T.C.; data curation, J.Y., and A.M.; writing—original draft preparation, J.Y., and A.M.; writing—review and editing, J.Y., T.C.,NL and A.M.; visualization, J.Y.; supervision, A.M.; project administration, T.C., and A.M.; funding acquisition, A.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by operating grants from the Natural Science of Engineering Research Council (RGPIN-2014-05662) and the Canadian Institutes of Health Research (MOP-133415).

Data availability

All data not provided in the manuscript and the supplementary information are available from alice.mui@ubc.ca upon request.

Declarations

Competing interests

The authors declare no competing interests.