Critical analysis of the hypothesized SNHG1/miR-195-5p/YAP1 axis

How lncRNAs such as SNHG1 are integrated into cellular pathways to regulate, or be a regulatory component of, cellular processes is a question yet to be completely answered. A potential mechanism by which SNHG1 may exert its effect is through the so-called sponge effect. In this mechanism, SNHG1 (or other lncRNAs) is thought to bind miRNAs as a molecular sponge in a sequence-specific manner and inhibit their ability to bind and inhibit translation of a specific mRNA. The recent paper by Cheng et al. (Cheng et al. 2022) makes just these connections, in this case, between SNHG1 , miRNA-195-5p , and the Hippo pathway gene, YAP1 . In their proposed model, miRNA-195-5p binds to, and causes degra-dation of, YAP1 mRNA. MiRNA-195-5p is in turn modulated through the sponging activity of SNHG1. These interactions theoretically occur through a region of 7 nucleotide complementarity. Thus, YAP1 target protein expression can be modulated through a complex interplay involving the degree of SNHG1 and miR-195-5p expression. High SNHG1 should lead to greater sponging of miR-195-5p , and result in high YAP1 mRNA and protein levels, and conversely, low SNHG1 should lead to less sponging of miR-195-5p , and therefore low YAP1 mRNA and protein levels.

How lncRNAs such as SNHG1 are integrated into cellular pathways to regulate, or be a regulatory component of, cellular processes is a question yet to be completely answered. A potential mechanism by which SNHG1 may exert its effect is through the so-called sponge effect. In this mechanism, SNHG1 (or other lncRNAs) is thought to bind miRNAs as a molecular sponge in a sequence-specific manner and inhibit their ability to bind and inhibit translation of a specific mRNA. The recent paper by Cheng et al. (Cheng et al. 2022) makes just these connections, in this case, between SNHG1, , and the Hippo pathway gene, YAP1. In their proposed model, miRNA-195-5p binds to, and causes degradation of, YAP1 mRNA. MiRNA-195-5p is in turn modulated through the sponging activity of SNHG1. These interactions theoretically occur through a region of 7 nucleotide complementarity. Thus, YAP1 target protein expression can be modulated through a complex interplay involving the degree of SNHG1 and miR-195-5p expression. High SNHG1 should lead to greater sponging of miR-195-5p, and result in high YAP1 mRNA and protein levels, and conversely, low SNHG1 should lead to less sponging of miR-195-5p, and therefore low YAP1 mRNA and protein levels.
However, we would like to call attention to some aspects of the published study which are common in the field overall and to offer a counterpoint to the conclusions drawn. In the human genome, there are possibly 183,000 occurrences of the 7 nucleotide sequence being suggested to confer specificity in the binding of SNHG1 to miR-195-5p and, subsequently, YAP1 (calculated as the 3 billion bp human genome divided by the frequency of random appearance of the given 7 nt sequence -4 7 ). Within protein-coding regions, which are estimated to be 1.5% of the genome, there may be 2746 (4.5 × 10 7 /4 7 ) occurrences of this sequence. Indeed, miRNA target prediction websites (miRDB, miRDB.org and TargetScanHuman, targetscan.org) predict 1419-1508 transcript targets of miR-195-5p based on binding in the 3' UTR. Interestingly, and relevant to the Cheng et al. study, is not only predicted to interact with YAP1, but also LATS2 through the exact same sequence. LATS2 is a member of the Hippo pathway and a direct negative regulator of YAP1. This significantly increases the complexity of the potential effects of miR-195-5p on Hippo pathway signaling and profoundly affects the interpretation of the results of the Cheng study, which focuses exclusively on YAP1. Still, whether the biological effects observed after modulation of SNHG1 or miR-195-5p occur through the Hippo pathway, or one of the other predicted target genes, is unknown.
Another concern of the work of Cheng et al. is that their published sequence for SNHG1 does not appear in the SNHG1 processed transcript, but the entire 21 nt sequence is in an SNHG1 intron. This means that the SNHG1 sequence used in their reporter studies cannot act as a sponge for miR-195-5p. However, the correct SNHG1 sequence does have an 8 nt span of complementarity to miR-195-5p and has the potential to sponge miR-195-5p (Fig. 1). The resulting impact on their reported studies is that the luciferase reporter experiments using their SNHG1 sequence insert are not utilizing the correct sequences and are thus invalid.
The observations made after SNHG1 knockdown may be valid, but linking those phenotypes specifically to YAP1 through miR-195-5p is difficult. A connection between SNHG1 and YAP1 in squamous cell carcinoma was previously claimed in 2020 through miR-375. This study was subsequently retracted (Gao et al. 2020). Likewise, the first report of a connection between SNHG1 and miR-195 has also been retracted (Zhang et al. 2021). Others have reported the possible connection between SNHG1 and miR-195, but with many target genes, including CCND1, PDCD4, NEK2, and CDC42 (Li et al. 2019;Huang et al. 2019;Ji et al. 2020;Chen et al. 2021). Few of these reports show a direct effect on the target mRNA upon SNHG1 knockdown but rely on luciferase reporter constructs to show sequence-specific activity between SNHG1 and the reported miRNA, and then between the miRNA and the target mRNA. However, in the 2 Page 2 of 3 luciferase reporter constructs, it is only the identified complementary sequence that is in the reporter. This system is designed to work but does not reliably recapitulate the cell biology.
Likewise, the current study does not show a change in endogenous YAP1 expression upon SNHG1 knockdown and relies on the circumstantial evidence of a luciferase reporter assay. Indeed, our data in prostate cancer indicate a modest increase in YAP1 expression following SNHG1 knockdown (Fig. 2a). Furthermore, when we compare cellular proliferation after SNHG1 and YAP1 knockdown, we see a greater magnitude decrease in proliferation after SNHG1 knockdown than YAP1 knockdown, even though the knockdown of YAP1 (90%) was significantly better than SNHG1 (60%) (Fig. 2b). If YAP1 were more proximal in the pathway than SNHG1, then we would expect a greater effect with modulation of YAP1 level than SNHG1 level.
Overall, we point out that the proposed mechanism and findings of Cheng et al. lack specificity due to the short nucleotide region of complementarity between miR-195-5p and YAP1, and some of the data is questionable because of the use of an incorrect sequence for SNHG1. In addition, their findings rely on luciferase reporter assays rather than measurement of endogenous RNA. We have observed findings opposite to what their model predicts, although admittedly in cell lines from a different cancer. Finally, we show that miR-195-5p is also predicted to interact and potentially regulate LATS2, a negative regulator of YAP1. These issues with the work by Cheng et al. are consistent with problems observed in the SNHG1 literature overall, including reliance on luciferase reporter constructs and incorrect sequences. We also note a significant number of retractions in this field due to fraudulent data. Therefore, we urge caution in interpreting the results of Cheng and colleagues and the SNHG1 literature overall.
Author contribution S.Z. and F.C. had the idea for the article, S.Z. performed the literature search and data analysis, and S.Z. and F.C. drafted and/or critically revised the work.
Funding Partial funding support was received from the Michigan Prostate SPORE (P50CA186786) Career Enhancement Project award to SP Zielske and the Department of Defense Prostate Cancer Research Program Physician Research Award (W81XWH2010394) and Karmanos Cancer Institute (startup funds) to FC Cackowski.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethical approval Not applicable.

Competing interests
The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.