microRNA strand selection: Unwinding the rules

In the paper

“miR‐574* (5p) is oncogenic in colorectal cancer, NSCLC, and PTC (Cui, Tang, Chen, & Wang, 2014; Ji et al., 2012; Ma et al., 2016; X. Wang, Lu, Geng, Yang, & Shi, 2017; Zhou et al., 2015; Zhou et al., 2016), while miR‐574 (3p) is tumor suppressive in breast cancer and leukemia (Chiyomaru et al., 2013; Ujihira et al., 2015; Xu et al., 2020; Yao, Wu, Lindner, & Fox, 2017).”

What the sources say

• Disputed

  Cui et al.’s Results report that reducing miR-574-5p increased proliferation and invasion in colorectal cancer cells, while increasing it had the opposite effect. The Discussion concludes that miR-574-5p plays a suppressive role in colorectal cancer liver metastasis, contrary to the oncogenic role attributed to this paper. The Discussion also mentions an earlier study reporting reduced tumor growth after miR-574-5p inhibition, but that is not Cui et al.’s finding. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  cites Cui, Tang, Chen, & Wang, 2014 sourceExperimental — accuracy not yet measured
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ji et al., 2012
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ma et al., 2016

In the paper

“Consistent with this idea, the less abundant miR* strands were shown repress mRNAs containing predicted miR* target sites in flies and mammals, suggesting that miR* strands are functional (Okamura et al., 2008; Yang et al., 2010).”

What the sources say

• ?Not checked — no full text of the sourcecites Okamura et al., 2008 source
• Supported

  The Abstract reports transcriptome-wide repression of seed-matched targets by star strands and direct repression of five miR-19* targets through star-seed sites. The Introduction also describes detectable miRNA* effects on target networks in Drosophila, while identifying the paper’s own tests as vertebrate work. (LLM, conf=0.91)

  cites Yang et al., 2010 source

In the paper

“Interestingly, these ALG‐1 mutations do not reverse strand selection for all miRNAs, suggesting that additional intrinsic or extrinsic factors may affect C. elegans miRNA strand selection differently for distinct miRNAs (Zinovyeva et al., 2015).”

What the sources say

• Supported

  The Discussion reports that ALG-1 mutations affect strand selection differently depending on the particular microRNA. It also raises the possibility that microRNA features or additional extrinsic mechanisms influence which strand is selected, matching the citing sentence's qualified interpretation. (LLM, conf=0.93)

  cites Zinovyeva et al., 2015 source

In the paper

“microRNAs (miRNAs) play a central role in the regulation of gene expression by targeting specific mRNAs for degradation or translational repression.”

What the sources say

• Supported

  The Introduction describes miRNAs as regulatory RNAs with conserved interactions with human mRNAs. The miRNA Regulatory Mechanisms section explains that repression of target mRNAs occurs through mRNA decay and reduced translation, supporting both mechanisms in the claim. (LLM, conf=0.94)

  cites Bartel, 2018 source
• Supported

  The Abstract says miRNAs recognize and downregulate target mRNAs. The CANONICAL SITE TYPES section identifies mRNA destabilization and translational repression as mechanisms of gene-expression downregulation, while noting that destabilization is the main mechanism. (LLM, conf=0.93)

  cites Kim, Chang, & Baek, 2017 source

In the paper

“In addition, for majority of miRNAs, both strands have precisely defined 5′ termini, compared to their 3′ ends, suggesting evolutionary pressure to maintain the target‐determining sequence of both strands (Okamura et al., 2008; Ruby, Jan, et al., 2007; Ruby, Stark, et al., 2007; Seitz, Ghildiyal, & Zamore, 2008).”

What the sources say

• ?Not checked — no full text of the sourcecites Okamura et al., 2008 source
• Partly supported

  The mir-1003 sequencing passage (Abstract; page not provided) says that each of two complementary small-RNA sets had a more consistent 5′ than 3′ terminus. The paper also identifies the 5′ seed as important for targeting and notes conservation of two mirtronic miRNA seeds (Abstract; page not provided). It does not establish this pattern for a majority of miRNAs or attribute it to evolutionary pressure to preserve the target-determining sequences of both strands. (LLM, conf=0.84)

  cites Ruby, Jan, et al., 2007 sourceExperimental — accuracy not yet measured
• Partly supported

  The Results and Discussion report that, in flies, the 5′ ends of both miRNA and miRNA∗ strands are typically more precisely defined than their 3′ ends. The paper presents evolutionary pressure to preserve seed sequences as one possible explanation, alongside constraints on Argonaute loading. It does not establish the unqualified claim for miRNAs generally. (LLM, conf=0.83)

  cites Seitz, Ghildiyal, & Zamore, 2008 sourceExperimental — accuracy not yet measured

In the paper

“Additional pairing between nucleotides in the 3′ half of the miRNA and the target mRNA also contribute to miRNA‐mediated gene targeting (reviewed in Chipman & Pasquinelli, 2019).”

What the sources say

• Supported

  In “Target recognition and regulation by miRNAs,” the review says that, after seed binding, nucleotides in the miRNA’s 3′ half can bind the target site and that 3′-end pairing can affect targeting specificity. “Same seed, different targets” also describes additional 3′-portion pairing as contributing to target interactions. (LLM, conf=0.94)

  cites Chipman & Pasquinelli, 2019 source

In the paper

“Selectively blocking the phosphorylation of the 5′ end of the guide strand in vitro leads to the loading of the passenger strand, consistent with either strand having the capacity to be loaded into miRISC (Kawamata et al., 2011).”

What the sources say

• Supported

  In “RISC loading requires 5′ phosphate of the guide” (Fig. 1C–D; page not provided), the authors report that blocking the original guide strand’s 5′ end while phosphorylating the passenger strand led to Ago1-RISC formation with the passenger strand acting as guide. This supports the claim that either strand can be loaded as the guide under the tested conditions, provided it has a 5′ phosphate. (LLM, conf=0.92)

  cites Kawamata et al., 2011 source

In the paper

“Consistently, the most commonly observed 5′ nucleotide of miRNA guide strands is a Uracil, whereas the passenger strand often begins with a Cytosine at the 5′ end (Ghildiyal et al., 2009; Mi et al., 2008; Takeda, Iwasaki, Watanabe, Utsumi, & Watanabe, 2008; Warf, Johnson, & Bass, 2011; Figure 2c).”

What the sources say

• Supported

  In the section on miRNA/miRNA* duplex structure and Argonaute loading, the paper reports that 72% of miRNAs detected in fly heads began with U. It also reports that C was the most frequent first nucleotide of miRNA* (passenger) strands, at 32.1%. (LLM, conf=0.91)

  cites Ghildiyal et al., 2009 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Mi et al., 2008
• ?Not checked — no full text of the sourcecites Takeda et al., 2008 source

In the paper

“The first four nucleotides of each end of the duplex are believed to establish thermodynamic asymmetry, and the difference of a single extra hydrogen bond is sufficient to drive preferential loading of the less stable strand in vitro (Khvorova, Reynolds, & Jayasena, 2003; Krol et al., 2004; Schwarz et al., 2003).”

What the sources say

• Not checked yet

  Checked before — its earlier verdict is held back until a new check runs.

  cites Khvorova, Reynolds, & Jayasena, 2003 source
• Partly supported

  The “Thermodynamic Features of miRNA Precursors” section compares stability at the two duplex ends, including their terminal four nucleotides, but finds that the terminal two nucleotides distinguish miRNA from miRNA* most effectively. It does not report that one extra hydrogen bond suffices to drive preferential loading in vitro. Page numbers are unavailable in our copy. (LLM, conf=0.86)

  cites Krol et al., 2004 sourceExperimental — accuracy not yet measured
• Supported

  The Supplementary data (Figure S1 caption; page unavailable) identifies the relative stability of the first four base pairs at each strand’s 5′ end as explaining functional asymmetry. The Results and Discussion (page unavailable) report that a single-hydrogen-bond difference measurably affects RISC assembly and that the less stably paired strand is favored; the supplementary experiments describe in vitro RNAi reactions. (LLM, conf=0.91)

  cites Schwarz et al., 2003 source

In the paper

“Interestingly, in some closely‐related species, the preferred miRNA arm for homologous miRNA genes is not the same (de Wit et al., 2009; Griffiths‐Jones et al., 2011; Marco et al., 2010), consistent with the idea that, in at least some cases, miRNA arm switching could be evolutionarily favored.”

What the sources say

• Supported

  The miRNA evolution mechanisms section reports that mir-246 has a dominant 3′ arm in C. elegans and C. briggsae but a dominant 5′ arm in C. remanei. The Figure 3 caption identifies this difference between related species as an example of arm switching. The suggestion that switching could be evolutionarily favored is the citing authors’ interpretation of that observation. (LLM, conf=0.91)

  cites de Wit et al., 2009 source
• Supported

  In Results and Discussion, under “Arm choice has evolved in the miR-100/10 family” (page unavailable), the paper reports opposite dominant-arm usage for orthologous miR-10 in D. melanogaster and T. castaneum. It also describes arm switching as a way to change miRNA targets and function during evolution. The sentence’s suggestion that switching could be evolutionarily favored is a cautious interpretation of that finding, not a claim that this paper demonstrated selection for it. (LLM, conf=0.91)

  cites Griffiths-Jones et al., 2011 source
• Partly supported

  The Results (Functional Shifts in Insect miRNAs; page unavailable) report different dominant arms for homologous miRNAs in Drosophila and Tribolium, and the Discussion calls arm switching a source of evolutionary novelty. But the Discussion says these species’ conserved miRNAs span approximately 350 million years, so this example does not establish the claim about closely related species. The paper also does not show that switching was evolutionarily favored rather than simply observed. (LLM, conf=0.85)

  cites Marco et al., 2010 sourceExperimental — accuracy not yet measured

In the paper

“However, miRNA guide‐strand selection appears to occur independently of R2D2 as miRNA guide strand loading is not perturbed by loss of R2D2 function (Nishida et al., 2013; Okamura, Robine, Liu, Liu, & Lai, 2010).”

What the sources say

• Partly supported

  The Results report that miR-bantam’s distribution between Ago1 and Ago2 was hardly affected by R2D2 loss, and the Discussion says Ago1-associated miRNA levels were hardly changed. These findings support largely unchanged miRNA loading, but the paper’s explicit guide-versus-passenger strand-selection result concerns siRNAs, not miRNAs; it does not directly establish R2D2-independent miRNA guide-strand selection. (LLM, conf=0.85)

  cites Nishida et al., 2013 sourceExperimental — accuracy not yet measured
• Supported

  The Results report that mature miRNA abundance was largely unchanged in AGO1 complexes from RNAi-defective mutants. The Discussion adds that AGO1 continued to preferentially incorporate mature miRNA strands when the RNAi pathway was disrupted, supporting the claim about guide-strand selection and loading despite loss of R2D2. (LLM, conf=0.90)

  cites Okamura, Robine, Liu, Liu, & Lai, 2010 source

In the paper

“It has been suggested that the ratio of miR::miR* strands may be influenced by target availability, as the presence of suitable mRNA targets can stabilize cognate miRNAs (Chatterjee, Fasler, Büssing, & Großhans, 2011; Tsai et al., 2016).”

What the sources say

• Supported

  The Introduction states that expressing miRNA targets increases cognate miRNA levels in C. elegans and that target availability may alter miR:miR∗ ratios. The Summary also reports that cognate target mRNAs protect miRNAs from degradation and can promote accumulation of certain miR∗ strands. (LLM, conf=0.94)

  cites Chatterjee, Fasler, Büssing, & Großhans, 2011 source
• Supported

  The Results section reports that adding binding sites or target 3′UTRs increased the levels of the corresponding miR-193a strands, and attributes differences in their ratio to target-gene expression. The Discussion states that target mRNA abundance can stabilise miRNA* strands. Page numbers are not provided in our copy. (LLM, conf=0.94)

  cites Tsai et al., 2016 source

In the paper

“miR‐155* (3p) targets the Hsp40‐family genes Dnaja2 and Dnajb1, whose down‐regulation appears to directly contribute to the disease pathogenesis (Mycko et al., 2015), (Table 1).”

What the sources say

• Supported

  The targeting results (Fig. 4C) show that miR-155-3p directly targets Dnaja2 and Dnajb1 and reduces their expression. The Discussion reports that these genes suppress EAE and identifies their transcripts as critical to autoimmune demyelination, supporting the hedged claim that their down-regulation contributes to pathogenesis. Page numbers are not available in the supplied copy. (LLM, conf=0.93)

  cites Mycko et al., 2015 source

In the paper

“While often supported by in vivo observations, such as the observation that most miRNAs begin with a 5′ Uracil (Ghildiyal et al., 2009; Mi et al., 2008; Takeda et al., 2008; Warf et al., 2011), how well these rules hold up in a living organism throughout its lifespan is yet to be established.”

What the sources say

• Supported

  The section on miRNA/miRNA* duplex structure and Argonaute loading reports that 72% of miRNAs expressed in fly heads began with U, supporting the observation for which this paper is cited. The sentence’s question about whether strand-selection rules hold throughout an organism’s lifespan is the citing authors’ framing, not a finding attributed to this source. (LLM, conf=0.93)

  cites Ghildiyal et al., 2009 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Mi et al., 2008
• ?Not checked — no full text of the sourcecites Takeda et al., 2008 source

In the paper

“Complicating the matters of miRNA strand stability being dictated by cognate target presence, numerous groups have reported target‐directed miRNA degradation (TDMD), a mechanism for target dependent miRNA decay (reviewed in Wightman, Giono, Fededa, & de la Mata, 2018).”

What the sources say

• Supported

  The Introduction describes this paper as a review of findings on TDMD. The General Features of TDMD section defines TDMD as target RNAs triggering degradation of mature miRNAs and cites multiple studies reporting it, supporting both the mechanism and the claim that multiple groups have reported it. (LLM, conf=0.93)

  cites Wightman, Giono, Fededa, & de la Mata, 2018 source

In the paper

“Further, modifications to the 3′ overhangs can also directly affect TS and can influence strand selection (Bramsen et al., 2009; Bramsen & Kjems, 2013; Ui‐Tei et al., 2012).”

What the sources say

• Partly supported

  The Results section ('Modified overhangs alter strand selection by RISC'; page unavailable) reports that 3′-overhang modifications can favour incorporation of the opposing strand. It does not establish that those modifications directly affect TS; the Discussion instead says the strand-selection effect occurs irrespective of the duplex’s thermodynamic asymmetry. (LLM, conf=0.84)

  cites Bramsen et al., 2009 sourceExperimental — accuracy not yet measured
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Bramsen & Kjems, 2013
• Partly supported

  The section on thermodynamic control by chemical modifications (page unavailable) says that modifications at the 3′ ends of the guide and passenger strands alter duplex stability and can improve selective guide-strand loading. The paper identifies siRNAs as having 3′ overhangs, but it does not establish that the reported stability changes come specifically from modifying those overhangs rather than other 3′-end nucleotides. (LLM, conf=0.84)

  cites Ui-Tei et al., 2012 sourceExperimental — accuracy not yet measured

In the paper

“Here, miR‐24‐2* (5p) directly targets PKCα (protein kinase C), a factor that is required for the survival of MCF‐7 breast cancer cells (Martin et al., 2012; Weldon et al., 2005).”

What the sources say

• Supported

  The Results section reports that miR-24-2* represses PKCα mRNA and its 3′ UTR in MCF-7 cells. The Discussion describes this as direct inhibition and states that PKCα is critical for MCF-7 cell survival. Page numbers are not provided in our copy. (LLM, conf=0.93)

  cites Martin et al., 2012 source
• ?Not checked — no full text of the sourcecites Weldon et al., 2005 source

In the paper

“suggesting that regulation of TUT4/TUT7 activity can dictate strand selection of miR‐324 in humans, resulting in arm switching across tissues (Kim et al., 2020).”

What the sources say

• Partly supported

  The section ‘miR-324 arm switching is controlled by TUT4 and TUT7’ reports that TUT4/7 knockdown changes the miR-324 strand ratio in human HEK293T cells. It also links TUT4/7 levels to strand ratios across mouse tissues, but the cited passages do not establish arm switching across human tissues specifically. (LLM, conf=0.84)

  cites Kim et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“Similar observations have been reported in flies, where several miRNA duplexes were found to exhibit tissue‐specific differences in miRNA strand selection (Ghildiyal et al., 2009).”

What the sources say

• Supported

  The Discussion reports different miRNA/miRNA* ratios for several duplexes in fly heads and ovaries, including a reversal of which strand is more abundant for miR-92a. This supports the cited observation of tissue-specific differences in strand selection; the paper presents the cause of those differences as a possibility, not a finding. (LLM, conf=0.91)

  cites Ghildiyal et al., 2009 source

In the paper

“The abundance of miR‐193a‐5p, but not miR‐193‐3p is significantly decreased in breast cancer and is believed to be due to reduced availability of miR‐193a‐5p targets (Tsai et al., 2016).”

What the sources say

• Partly supported

  The Results section (‘Different arm expression preferences of miR-193a in breast cancer’; page unavailable) reports significantly lower miR-193a-5p in breast cancer tissue, with no difference in miR-193a-3p. The Discussion proposes that target-gene abundance may explain the differing arm levels, but does not establish that miR-193a-5p targets become less available in breast cancer; it specifically suggests that greater stabilization of miR-193a-3p may explain why that arm did not decline. (LLM, conf=0.84)

  cites Tsai et al., 2016 sourceExperimental — accuracy not yet measured

In the paper

“Interestingly, miRNA arm switching has been proposed to influence miRNA evolution, as altered selection of a miRNA strand by Argonaute would be expected to result in a different mRNA target repertoire (de Wit, Linsen, Cuppen, & Berezikov, 2009; Griffiths‐Jones, Hui, Marco, & Ronshaugen, 2011; Marco, Hui, Ronshaugen, & Griffiths‐Jones, 2010).”

What the sources say

• Supported

  The Discussion proposes arm switching as a mechanism of miRNA evolution. The paper also explains that either hairpin arm can supply the miRNA loaded into RISC and that the mature miRNA’s seed determines target recognition. Together, these passages support the proposed change in target repertoire when the selected arm changes; the paper uses “RISC” rather than naming Argonaute in these passages. No page numbers are available in our copy. (LLM, conf=0.88)

  cites de Wit, Linsen, Cuppen, & Berezikov, 2009 source
• Supported

  The Introduction says that miRNAs from opposite arms can be sorted into different Argonaute complexes and that changes in arm usage may alter miRNA function. In Results and Discussion (‘Arm switching impacts mRNA target regulation’), the paper reports that targets predicted for the two arms do not overlap more than expected by chance. The supplied passages do not give page numbers. (LLM, conf=0.90)

  cites Griffiths-Jones, Hui, Marco, & Ronshaugen, 2011 source
• Supported

  The Results section reports that shifts in use of the 5′ and 3′ arms are predicted to change a miRNA’s target profile because the two arms produce different mature sequences. The Discussion identifies arm switching as a source of evolutionary novelty. The paper describes strand association with the RNA-induced silencing complex, though it does not name Argonaute in these passages. (LLM, conf=0.88)

  cites Marco, Hui, Ronshaugen, & Griffiths-Jones, 2010 source

In the paper

“As ethanol exposure promotes changes in the 5p/3p ratios of certain miRNAs, it is possible that de‐regulation of miRNA strand selection could play some role in alcohol abuse disorders (Ghezzi et al., 2016).”

What the sources say

• Partly supported

  The Discussion reports ethanol-related changes in the abundance of four miRNA passenger strands in Drosophila and suggests that those strands may have regulatory potential. It does not establish changes in 5p/3p ratios or propose that disrupted strand selection plays a role in alcohol abuse disorders. (LLM, conf=0.81)

  cites Ghezzi et al., 2016 sourceExperimental — accuracy not yet measured

In the paper

“Under TDMD, the presence of a target mRNA promotes cognate miRNA degradation, often with specific target mRNAs responsible for controlling miRNA decay (Ameres et al., 2010; Cazalla, Yario, Steitz, & Steitz, 2010; de la Mata et al., 2015; Ghini et al., 2018; Kleaveland, Shi, Stefano, & Bartel, 2018).”

What the sources say

• Partly supported

  The “miRNA sensors affect steady-state miRNA levels” section (page not provided) reports that engineered egfp mRNAs with complementary sites reduced the abundance of their corresponding miRNAs. This supports target-dependent, cognate miRNA loss, but does not establish that specific target mRNAs often control miRNA decay generally; the paper’s “Sequence requirements for miRNA remodeling” section says canonical animal miRNA-binding sites did not alter miRNA stability. (LLM, conf=0.84)

  cites Ameres et al., 2010 sourceExperimental — accuracy not yet measured
• Partly supported

  The supplied Abstract passage reports that the viral noncoding RNA HSUR 1 directs degradation of mature miR-27 through sequence-specific binding. This supports target-RNA-directed miRNA decay, but the identified RNA is not an mRNA, and the paper does not establish that specific target mRNAs often control miRNA decay. No page numbers are supplied for the passages. (LLM, conf=0.87)

  cites Cazalla, Yario, Steitz, & Steitz, 2010 sourceExperimental — accuracy not yet measured
• Supported

  The Introduction defines TDMD as decay of miRNAs triggered by highly complementary target RNAs, including mRNAs. The Discussion reports that expression of highly complementary target mRNAs induces potent, specific miRNA degradation in neurons. (LLM, conf=0.90)

  cites de la Mata et al., 2015 source

In the paper

“miR‐574* (5p) targets QKI6 (Quaking), while miR‐574‐3p targets ACVR1B (activin receptor type‐1), and depleting one miRNA strand's target phenocopies the loss of the other miRNA strand in gastric cancer cells (Zhang et al., 2019).”

What the sources say

• Partly supported

  The Results section, “miR-574-5p/3p target endogenous QKI6 and ACVR1B, respectively” (page unavailable), identifies QKI6 and ACVR1B as direct targets of miR-574-5p and miR-574-3p, respectively. It reports that restoring target expression partly rescues the effects of the corresponding miRNA, but does not show that depleting one strand’s target phenocopies loss of the other strand. (LLM, conf=0.87)

  cites Zhang et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“alters the sequence of miR‐146a* (3p) strand, and likely leads to changes in target repression, contributing to PTC (Jazdzewski et al., 2009; Labbaye & Testa, 2012), (Table 1).”

What the sources say

• ?Not checked — no full text of the sourcecites Jazdzewski et al., 2009 source
• Supported

  The “miR-146a and cancer” section (page not provided) says the G/C polymorphism produces two passenger-strand isoforms, miR-146a*G and miR-146a*C, with distinct target genes. It also links the resulting changes in gene expression in heterozygotes to predisposition to papillary thyroid cancer. (LLM, conf=0.90)

  cites Labbaye & Testa, 2012 source

In the paper

“Interestingly, over one‐third of the genes regulated by miR‐574‐3p and miR‐574‐5p overlap, highlighting that although each strand contains a unique seed sequence, there can be substantial shared targets of the two strands (Zhang et al., 2019).”

What the sources say

• Partly supported

  The Results section on QKI6 and ACVR1B (Fig. 3j; page unavailable) reports that approximately one-third of the genes regulated by miR-574-5p and miR-574-3p overlap. It does not establish that the proportion is over one-third, and it describes the common targets as arising through direct or indirect pathways. (LLM, conf=0.87)

  cites Zhang et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“In animal models of multiple sclerosis, miR‐155* (3p) is highly up‐regulated in T‐helper (Th) cells, and promotes Th17 cell differentiation (Mycko et al., 2015).”

What the sources say

• Supported

  The Discussion reports that miR-155-3p is highly upregulated in T-helper cells during EAE, an animal model of multiple sclerosis. The Results section, “miR-155-3p presence in CD4+ T cells promotes Th17 development,” reports that its upregulation promotes Th17 differentiation after T-cell receptor activation. (LLM, conf=0.94)

  cites Mycko et al., 2015 source

In the paper

“Both strands appear to target Cyclin‐D (CCND1), Ajuba LIM protein (AJUBA) and heart developmental protein with EGF like domains (HEG1) and knocking down those targets phenocopies restored miR‐193b levels (Choi et al., 2019).”

What the sources say

• Supported

  The Results section on direct binding reports that both miR-193b strands inhibit CCND1, AJUBA, and HEG1 through interactions with their mRNAs’ 3′UTRs (Figure 4). The Results section on target-gene down-regulation reports that knocking down each gene reduced invasion, migration, and colony formation in lung cancer cells, consistent with the effects of miR-193b (Figure 5). (LLM, conf=0.93)

  cites Choi et al., 2019 source

In the paper

“TUT4/TUT7 uridylate the 3′ end of miR‐324‐3p, which changes the site of Dicer processing to generate a shorter miR‐324 duplex that undergoes reversed strand selection, stabilizing the miR‐324‐3p strand instead of the typically more abundant miR‐324‐5p strand (Figure 4c; Kim et al., 2020).”

What the sources say

• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites Kim et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“In fly ovaries, the miR‐92a‐3p, miR‐988‐3p, and miR‐284‐5p strands are expressed at higher levels than their corresponding passenger strands, while the miR‐92a‐5p, miR‐988‐5p, and miR‐284‐3p strands are expressed in excess within fly heads (Ghildiyal et al., 2009).”

What the sources say

• Partly supported

  The Discussion (page unavailable) reports the claimed ovary-versus-head abundance patterns for miR-92a, miR-988, and miR-284, comparing each annotated miRNA with its miRNA* strand. Our copy does not identify those strands by 3p or 5p designation, so it does not establish that the specifically named strands correspond to the reported comparisons. (LLM, conf=0.82)

  cites Ghildiyal et al., 2009 sourceExperimental — accuracy not yet measured

In the paper

“Consistently, restoration of miR‐223 (3p) in BRCA1, BAP1‐deficient cancer cells is synthetic lethal, probably due to repression of NHEJ components (Srinivasan et al., 2019).”

What the sources say

• Supported

  The Discussion states that a single miR223-3p transfection induced synthetic lethality in BRCA1- and BAP1-deficient malignancies, likely by destroying the required alternative NHEJ components CtIP and Pso4. The Results also report cytotoxicity in the tested BRCA1- and BAP1-deficient cancer cells. Page numbers are not provided in our copy. (LLM, conf=0.94)

  cites Srinivasan et al., 2019 source

In the paper

“In another example, both miR‐193b (3p) and miR‐193b* (5p) are down regulated in lung cancer and restoring the levels of either strand suppresses malignant phenotypes including metastatic potential and cell proliferation (Choi et al., 2019), (Table 1).”

What the sources say

• Supported

  The Results section reports that both strands are down-regulated in metastatic lung cancer cells. It also reports that restoring either strand reduces migration and invasion; Figure 2 describes colony-forming assays showing reduced proliferative potential. The paper’s evidence is particularly focused on metastatic lung cancer cells. (LLM, conf=0.91)

  cites Choi et al., 2019 source

In the paper

“In another example, a single point mutation (U–C) that affects the relative TS of the human miR‐133a duplex results in altered 5p/3p ratios of miR‐133a and is associated with atrial fibrillation presumably through effects on mRNA targeting (Ohanian et al., 2013; Figure 6d).”

What the sources say

• Partly supported

  The Abstract Results and Discussion report that the 79T>C MIR133A2 variant weakens predicted duplex thermostability and changes the miR-133a-5p/3p ratio. The Results identify the variant in one patient with atrial fibrillation, but the Discussion calls its relationship to AF “questionable”; effects on mRNA targets are predicted, not demonstrated in patient tissue. Thus the claimed association through mRNA targeting overstates the evidence. (LLM, conf=0.90)

  cites Ohanian et al., 2013 sourceExperimental — accuracy not yet measured

In the paper

“Intriguingly, TUT‐4/7 are up‐regulated in glioblastoma and miR‐324 arm switching appears to correlate with a poor prognosis (Kim et al., 2020).”

What the sources say

• Supported

  In “Alteration of the miR-324 strand ratio affects cell cycle progression” (Figure 6B–C and Figure S6C), the paper reports that both TUT4 and TUT7 are significantly upregulated in glioblastoma and that patients with low miR-324-5p/3p ratios had poor prognosis. This supports the citing sentence’s cautious description of an association between altered arm usage and prognosis. (LLM, conf=0.94)

  cites Kim et al., 2020 source

In the paper

“miR‐21* (3p) downregulation of PPACTR4 appears to decrease protein phosphatase 1 (PP1) activity leading to increased phosphorylation and inactivation of the tumor suppressor Rb/E2F (retinoblastoma/E2 transcription factor), ultimately resulting in tumorigenesis (Ge et al., 2015; Ohanian et al., 2013).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ge et al., 2015
• ?Could not verify

  The Results section on the MIR133A2 variant reports altered miR-133a-5p abundance, and the Discussion addresses predicted miR-133a targets. Searches for miR-21*, PPACTR4, PP1, Rb/E2F, and tumorigenesis found no passage in our copy addressing the claimed pathway. (LLM, conf=0.76)

  cites Ohanian et al., 2013 sourceExperimental — accuracy not yet measured

In the paper

“Each miRNA is post‐transcriptionally processed into a duplex comprising two strands.”

In the paper

“The Argonaute‐loaded miRNA programs the gene target specificity of the miRISC, as nucleotides 2‐7, referred to as the miRNA “seed” sequence, are mainly responsible for determining the target mRNA identity through base pairing (Grün, Wang, Langenberger, Gunsalus, & Rajewsky, 2005; Haley & Zamore, 2004; Jackson et al., 2003; Krek et al., 2005; Lewis, Burge, & Bartel, 2005; Lewis, Shih, Jones‐Rhoades”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Grün et al., 2005
• ?Not checked — no full text of the sourcecites Haley & Zamore, 2004 source
• ?Not checked — no full text of the sourcecites Jackson et al., 2003 source

In the paper

“The target gene specificity of miRISC is determined by sequence complementarity between the Argonaute‐loaded miRNA strand and target mRNA.”

In the paper

“Residues within the Argonaute PIWI domain contribute to a phosphate binding tract that is formed along the interface of the MID and PIWI domains that is proposed to sense TS while a nucleotide selection loop within the MID domain senses the identity of the 5′ nucleotide (Suzuki et al., 2015; Figure 2a).”

What the sources say

• ?Not checked — no full text of the sourcecites Suzuki et al., 2015 source

In the paper

“Therefore, miRNA strand selection plays a defining role in dictating the specificity of miRISC toward its targets and provides a mechanism to alter gene expression in a switch‐like fashion.”

In the paper

“Consistently, the 5′ ends of guide strands have been found to contain an excess of purines, whereas the 5′ ends of passenger strands are pyrimidine rich in humans and flies, which likely contributes to thermodynamic asymmetry between the two duplex ends (Hu et al., 2009).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Hu et al., 2009

In the paper

“One strand of the miRNA duplex is selectively anchored into the Argonaute protein and determines the specificity of the miRISC based on sequence complementarity between the miRNA and the 3′UTR of the target mRNA (Figure 1, Step 5).”

In the paper

“Each strand of the miRNA duplex has the capacity to be loaded into miRISC and possesses a unique seed sequence.”

In the paper

“The miRNA duplex is then loaded into an Argonaute protein to form the miRNA‐Induced Silencing Complex (miRISC; Figure 1, Step 5).”

In the paper

“miRNAs that share a common seed sequence, referred to as miRNA families, can target the same group of mRNA substrates, at least in a partially redundant manner.”

In the paper

“While either strand can be selected for miRISC loading, frequently, a single miRNA strand predominates.”

In the paper

“One of the two miRNA strands is selectively loaded into an Argonaute protein to form the miRNA‐Induced Silencing Complex (miRISC) in a process referred to as miRNA strand selection. The other strand is ejected from the complex and is subject to degradation.”

In the paper

“For most miRNAs, one strand is preferentially loaded into Argonaute and becomes part of the miRISC, originally earning the designation of the “miR” or “guide” strand. The other strand of the miRNA duplex is ejected from the Argonaute protein and degraded (Figure 1, Step 5′).”

In the paper

“Perhaps not surprisingly, Argonaute protein appears to play a direct role in miRNA strand selection, as recombinant Argonaute MID domain is sufficient to recognize the preferred 5′ nucleotide and to favor thermodynamically unstable ends (Suzuki et al., 2015).”

What the sources say

• ?Not checked — no full text of the sourcecites Suzuki et al., 2015 source

In the paper

“Aberrant strand selection can lead to altered gene regulation by miRISC and is observed in several human diseases including cancer.”

In the paper

“While sequestering miRNA arms might not affect the 5p/3p ratio of a given miRNA, it could alter the balance of target repression by each miRNA arm.”

In the paper

“The ability of cells to switch the miRNA arm preference is also evident from cases of physiological or pathogenic states, further supporting the idea that miRNA strand selection is a regulated process.”

In the paper

“Mechanisms influencing the production of isomiRs (highlighted in bold) have been experimentally demonstrated to influence 5p/3p ratios of miRNAs and can lead to miRNA arm switching, whereas other mechanisms remain largely hypothetical”

In the paper

“Overall, isomiRs may be a dominant mechanism driving miRNA strand selection, or arm switching, although it remains possible that unidentified regulators play substantial roles.”

In the paper

“It is also possible that circRNAs, which have been proposed to act as miRNA sponges (reviewed in Kristensen et al., 2019; Thomson & Dinger, 2016), could influence miRNA arm availability by sequestering miRNAs that have target sites within the circRNA.”

What the sources say

• ?Not checked — no full text of the sourcecites Kristensen et al., 2019 source
• ?Not checked — no full text of the sourcecites Thomson & Dinger, 2016 source

In the paper

“Thus, the presence of “decay targets” may in theory influence the 5p/3p ratios of certain miRNAs through TDMD.”

In the paper

“We propose that strand selection is likely the dominant mechanism leading to altered arm ratios, as miRNA turnover would be expected to influence miRNA arm ratios after a dominant miRNA strand is selected.”

In the paper

“While the 5′‐nucleotide preference and TS rules are considered to be the primary factors driving miRNA strand selection during Argonaute loading, strand selection of all miRNAs may not be explained by these two rules only.”

In the paper

“While the nature of these modifications in miRNA strand selection have not been fully explored, such alterations could presumably influence the TS of all small RNA duplexes and subsequently affect strand selection.”

In the paper

“however, deletion of the PAZ domain did not alter miRNA loading, suggesting PAZ domain is not required for sensing TS (Suzuki et al., 2015).”

What the sources say

• ?Not checked — no full text of the sourcecites Suzuki et al., 2015 source

In the paper

“These isomiRs can contain truncations or additions to the miRNA duplex and can alter the 5′ nucleotide identity of either, or both, strands.”

In the paper

“Additional accessory factors may also play a role in miRNA strand selection, most likely through isomiR formation.”

In the paper

“In human cells, DNA damage has been shown to promote changes in the expression of miRNAs (Dickey, Zemp, Martin, & Kovalchuk, 2011)”

What the sources say

• ?Not checked — no full text of the sourcecites Dickey, Zemp, Martin, & Kovalchuk, 2011 source

In the paper

“Overall, several mechanisms may contribute to altered miRNA 5p/3p ratios with regulation of miRNA strand selection likely playing a dominant role (Figure 5).”

In the paper

“In addition, miRBase‐annotated miRNA sequences or abundances may not be able to accurately reflect tissue‐dependent expression of miRNAs, making it difficult to define a single miRNA duplex from which any given miRNA strand originates.”

In the paper

“Notably, the miR‐194‐3p strand was nearly undetectable in brain and stomach tissues but was expressed at least as strongly as the miR‐194‐5p strand in other tissues including lungs, ovaries and uteri (Ro et al., 2007).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ro et al., 2007

In the paper

“While the miRNA guide strand is usually observed in extreme excess to its corresponding passenger strand, growing evidence supports that the relative levels of the two strands are highly regulated.”

In the paper

“Overexpression of miR‐574 *(5p) enhances cell proliferation while expression of miR‐574 (3p) suppresses proliferation.”

In the paper

“While data suggest that Argonaute proteins contribute to miRNA strand selection by recognizing both the 5′ nucleotide and relative TS of miRNA strands, it has become increasingly clear that these rules can be “circumvented” under different physiological or pathological conditions, primarily through duplex sequence alterations or isomiR production.”

In the paper

“Both miRNA strands can be functional, with arm switching playing a role during development through tissue‐specific or perhaps temporally‐regulated miRNA strand selection.”

In the paper

“Finally, multiple mechanisms may be involved in regulation of miRNA strand selection, with combinations of molecular inputs dictating which miRNA becomes loaded and to what extent.”

In the paper

“However, the rules governing miRNA strand selection were determined using a limited number of purified components; whether these rules are sufficient to drive strand selection in vivo has not been extensively explored.”

In the paper

“Similarly, altered miRNA strand selection is associated with disease, presumably due to targeting distinct sets of mRNAs (Figure 7).”

In the paper

“In squamous cell carcinoma, both miR‐21 (5p) and miR‐21* (3p) are up‐regulated and each is required for cell survival (Ge et al., 2015; Figure 6c, Table 1).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ge et al., 2015

In the paper

“Collectively, these data highlight that miRNA strand selection can be highly regulated and alterations to this normally robust process can have dramatic consequences and are prevalent in human disease.”

In the paper

“There are at least three known regulatory points for miRNA strand selection: (a) altered processing of the miRNAs by RNAse III endonucleases generating templated isomiRs, (b) nontemplated miRNA isomiRs generated through deamination (A‐I editing) or uridylation, which remodel the miRNA duplex, and (c) altered strand preference of Argonaute (Figure 7).”