Mol Biol Rep (2014) 41:7115–7120 DOI 10.1007/s11033-014-3586-7
A novel measurement of allele discrimination for assessment of allele-specific silencing by RNA interference Masaki Takahashi • Hirohiko Hohjoh
Received: 8 July 2013 / Accepted: 5 July 2014 / Published online: 19 July 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Allele-specific silencing by RNA interference (ASP-RNAi) is an atypical RNAi that is capable of discriminating target alleles from non-target alleles, and may be therapeutically useful for specific inhibition of diseasecausing alleles without affecting their corresponding normal alleles. However, it is difficult to design and select small interfering RNA (siRNAs) that confer ASP-RNAi. A major problem is that there are few appropriate measures in determining optimal allele-specific siRNAs. Here we show two novel formulas for calculating a new measure of allelediscrimination, named ‘‘ASP-score’’. The formulas and ASP-score allow for an unbiased determination of optimal siRNAs, and may contribute to characterizing such allelespecific siRNAs. Keywords RNA interference Small interfering RNA Allele-specific silencing Allele discrimination Novel formula Abbreviations RNAi RNA interference siRNA Small interfering RNA ASP-RNAi Allele-specific RNAi DR Discrimination ratio
Electronic supplementary material The online version of this article (doi:10.1007/s11033-014-3586-7) contains supplementary material, which is available to authorized users. M. Takahashi H. Hohjoh (&) Department of Molecular Pharmacology, National Institute of Neuroscience, NCNP, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187-8502, Japan e-mail: [email protected]
Introduction RNA interference (RNAi) is the gene silencing machinery triggered by 21–25-nucleotide-long small interfering RNA (siRNA) duplexes homologous to silenced genes, and has become a powerful tool for suppressing the expression of a gene of interest [1, 2]. The application of RNAi has been expanding to various fields including drug-discovery and medical science [3, 4]. Allele-specific RNAi, or ASP-RNAi, is an advanced application of RNAi, and it can specifically inhibit the expression of target alleles of interest, but not suppress the expression of their corresponding non-target alleles [4, 5]. Therefore, ASP-RNAi may become an applicable tool for suppression of disease-causing alleles with a minimal inhibition against their corresponding wild-type (normal) alleles; and its application as a therapeutic use for dominantly inherited diseases such as familial Alzheimer’s disease and Huntington’s disease may be particularly promising [4, 6–15]. To induce ASP-RNAi, the design and selection of siRNAs that confer ASP-RNAi is vital, but quite difficult. Generally, an in vitro assessment of designed siRNAs is performed using reporter genes carrying target (mutant) and non-target (normal) allelic sequences. In the conventional ways of assessment, the effects of designed siRNAs against target and non-target reporter alleles are separately (independently) examined, and subsequently a careful normalization for a precise comparison among such independent data needs to be carried out. Our previous studies established an in vitro heterozygous assay system for assessment of designed siRNAs for ASP-RNAi, and the system substantially mitigated the difficulty [5, 12]. Briefly, the Photinus and Renilla luciferase reporter genes carrying mutant (target) and normal (non-target) allelic sequences in their 30 untranslated regions were constructed. The effects of designed siRNAs against
123
7116
the target mutant reporter allele in allele-specific silencing, and off-target silencing against the normal reporter allele, were simultaneously examined under a heterozygous condition generated by cotransfection of the reporter alleles and siRNAs into cultured mammalian cells. Introduction of an artificial nucleotide mismatch (es) into siRNAs often improves allele discrimination, or allelespecific silencing [4, 10, 11, 14, 16–19]. However, the followings are still unpredictable: which nucleotide position in introducing a mismatch into siRNAs is effective for improving ASP-RNAi? and which nucleotide of three mismatched nucleotides can work most effectively? Currently, designed siRNAs including such mismatched (modified) siRNAs must be tested one by one, and a precise assessment needs to be carried out. A problem over the course of assessment of designed siRNAs is that there are few appropriate measures in determining optimal allele-specific siRNAs, regardless of the data obtained from either conventional methods or our in vitro assay system. In this report, we describe two novel formulas for calculating a new measure of allele discrimination, named ‘‘ASP-score’’. The formulas and ASP-score may provide us with a precise determination of optimal allele-specific siRNAs and may contribute to understanding the characteristics of such allele-specific siRNAs.
Mol Biol Rep (2014) 41:7115–7120
Photinus and Renilla luciferase reporter genes carrying mutant and wild-type allelic sequences in their 30 untranslated regions were constructed as a mutant allele and a wild-type allele, respectively; and vice versa. The day before transfection, HeLa cells were treated with trypsin, suspended in fresh medium without antibiotics, and seeded onto 96-well culture plates at a cell density of 1 9 105 cells/cm2. The pGL3-TK-backbone reporter plasmid (60 ng), phRL-TK-backbone reporter plasmid (10 ng), pSV-b-Galactosidase control vector (20 ng) (Promega, Fitchburg, WI, USA), and 20 nM (final concentration) of siRNA duplex were added to each well; a lipofectamine 2000 transfection reagent (Invitrogen) was used for transfection according to the manufacturer’s instructions. Cell lysate was prepared 24 h after transfection, and the expression level of luciferase and b-galactosidase was examined using a Dual-Luciferase Reporter Assay System (Promega) and a Beta-Glo Assay System (Promega), respectively. The luminescent signals were measured using a Fusion Universal Microplate Analyzer (PerkinElmer, Waltham, MA, USA). The expression level of mutant and wild-type reporter alleles was normalized to the level of bgalactosidase studied as a control, and the ratios of the mutant and wild-type reporter expression in the presence of test siRNAs were further normalized against the control ratios obtained with a non-silencing control siRNA (siControl).
Materials and methods Cell culture
Results and discussion
HeLa cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Wako) supplemented with 10 % fetal bovine serum (Invitrogen), 100 units/ml penicillin, and 100 lg/ml streptomycin (Wako) at 37 °C in a 5 % CO2 humidified chamber.
Conventional discrimination ratio and novel ASP-score
RNA and DNA oligonucleotides DNA oligonucleotides and siRNAs used in this study were synthesized by and purchased from Sigma-Aldrich (St Louis, MO, USA). Construction of reporter alleles, transfection and heterozygous reporter assay Construction of mutant and wild-type reporter alleles, transfections, and the in vitro assay using the reporter alleles were carried out as described previously [5, 11, 12, 15, 20]. The DNA oligonucleotide sequences of the target and non-target alleles used in the construction of the reporter alleles and the sequences of siRNAs examined are indicated in Supplementary Table S1 and S2. Briefly, the
123
The assessment of siRNAs designed for ASP-RNAi depends upon an in vitro reporter assay, and the data obtained from target and non-target reporter alleles are compared each other after normalization with the data obtained from a control reporter. The most important part in the assay is about how to evaluate the allele discrimination between target and non-target alleles precisely, and discrimination ratio (target-allele expression/non-targetallele expression) is usually used as a measure of allele discrimination. However, the discrimination ratio is not a perfect measure and may be sometimes not suitable for such an assessment; for example, when siRNAs confer a strong inhibition to target alleles together with a moderate suppression to non-target alleles (detailed in Fig. 1), it may be difficult to select competent siRNAs for ASP-RNAi only by discrimination ratio. Therefore it is of importance and necessary to develop a new measure for allele discrimination as an alternative to the discrimination ratio, for a precise and objective assessment of siRNAs designed for ASP-RNAi.
Mol Biol Rep (2014) 41:7115–7120
7117
Fig. 1 Comparison between ASP-score and discrimination ratio (DR). a In vitro heterozygous reporter assay. The effects of representative siRNAs (indicated) that were studied previously [15, 20] on target mutant reporter alleles (open bars) and non-target wildtype reporter alleles (solid bars) were exhibited. Data are average of four independent determinations. Error bars represent standard deviations. b Discrimination ratios. Discrimination ratios (DR) (target-allele expression/non-target-allele expression) of indicated siRNAs were calculated based on the data in A. c, d ASP-scores. ASP-scores of indicated siRNAs were calculated by Formula I (c) and Formula II (d) with the same data in A. e Histogram of siRNAs in DR and ASP-score. Histogram was created based on DRs and ASP-scores
of 119 siRNAs which we designed and examined in the previous studies [15, 20]. X-axis is a relative value of DR or ASP-score. Difference in the distribution of the siRNAs between DR and ASPscore graphs was analyzed by a v2 test, and a significant difference between them was detected (p = 0.0004). f Abundance ratio of siRNAs capable of inhibiting wild-type alleles as well as mutant alleles. Of 119 siRNAs, above-average siRNAs in DR ([Ave-DR) or ASP-score ([Ave-ASP) were grouped, and the percentage of siRNAs that retained an ability to inhibit 50 % or more of the expression of wild-type alleles was examined. In addition, whole siRNAs (all siRNAs) as a control group were also examined
We developed two novel formulas for calculating a new measure of allele discrimination between target and nontarget alleles, and named the new measure ‘‘ASP-score’’: Formula I: [non-target (normal) allele Exp.R.—target (mutant) allele Exp.R.] 9 non-target (normal) allele Exp.R. Formula II: [non-target (normal) allele Exp.R.—target (mutant) allele Exp.R.] 9 [1—target (mutant) allele Exp.R.] Non-target (normal) allele Exp.R.: relative expression ratio of a normalized expression level of a non-target allele treated with a test siRNA to a normalized expression level of the non-target allele treated with a control non-silencing siRNA (siControl) as 1. Target (mutant) allele Exp.R.: relative expression ratio of a normalized expression level of a target allele treated
with the test siRNA to a normalized expression level of the target allele treated with the siControl as 1. Formula I is suitable for selection of siRNAs conferring a marked allele-discrimination without affecting non-target alleles, and Formula II is capable of selecting siRNAs that confer allele-discrimination with a strong inhibition against target alleles even though the siRNAs retain some ability to cleave non-target alleles (Supplementary Fig. S1); in other words, if a severe adverse effect by suppressing normal alleles is a concern, and if a strong inhibition against disease-causing alleles in spite of some suppression of normal alleles is required, Formula I and II, respectively, may be useful for selecting optimal allelespecific siRNAs against target disease-causing alleles.
123
7118
Assessment of siRNAs based on ASP-scores We calculated both ASP-scores and discrimination ratios of siRNAs that we previously examined by our in vitro assay system [15, 20]. Figure 1 shows representative examples. The siG356D_A10(G18), siRs331_T9 and siRs331_T11 siRNAs, which conferred a strong inhibition and also a moderate inhibition to their target and non-target alleles, respectively (Fig. 1a), exhibited a high value in discrimination ratio (Fig. 1b). The siG356D_A9_3- and siRs331_T10 siRNAs that appear to confer an ideal ASP-RNAi (Fig. 1a), similarly indicated a high value in discrimination ratio (Fig. 1b). When their ASP-scores were calculated by Formula I, the ASP-scores of siG356D_A10(G18), siRs331_T9 and siRs331_T11 were
Fig. 2 Characteristics of competent siRNAs conferring allele-specific silencing. a Top 20 siRNAs conferring a strong ASP-RNAi. Based on the ASP-scores of 119 siRNAs we designed previously [15, 20], the top 20 siRNAs were arranged in descending order of the ASC-score, and aligned with reference to the position of target nucleotide variations (target mutations) which were highlighted in blue. Introduced mismatch-nucleotides are indicated by small letters and highlighted in green. The position of 50 end, 30 end and introduced mismatches (MM), relative to the target mutation point given as 0, is indicated (right panel). The expression levels of target and non-target reporter alleles treated by the top 20 siRNAs are indicated in Supplementary Fig. S2. b Graphic representation of the relation between ASP-scores and the position of the 50 or 30 ends of
123
Mol Biol Rep (2014) 41:7115–7120
markedly lower than those of siG356D_A9_3- and siRs331_T10, and also appeared to be as low as those of the siRs273_A8 and siG356D_A10(G16) siRNAs that were incapable of inducing RNAi (Fig. 1c). Accordingly, siG356D_A10(G18), siRs331_T9 and siRs331_T11 were rated as unsuitable siRNAs for ASP-RNAi by their ASP-scores. When Formula II was used instead of Formula I, the ASP-score of siG356D_A9_3- and siRs331_T10 remained a high value; however, siG356D_A10(G18), siRs331_T9 and siRs331_T11, which exhibited a similar discrimination ratio, indicated a different profile in ASP-score (Fig. 1d): this difference might be attributable to the level of silencing of non-target alleles. There was a significant difference in distribution of siRNAs between ASP-score and discrimination ratio
the top 20 siRNAs. The position of either the 50 or 30 ends of the siRNAs, relative to the mutation point given as 0, was exhibited. c Mismatched position. The relation between ASP-scores and the position of introduced mismatch-nucleotides relative to the mutation point was indicated. All designed siRNAs that carried mismatched nucleotides were examined and the siRNA belonging to the top 20 siRNAs were indicated by green circles. d Schematic drawing of a sense-strand siRNA. Open and solid circles represent nucleotides, and the solid circle indicates the position corresponding to a target mutation point. Possibly favorable 50 and 30 ends and mismatch position of siRNAs conferring a strong allele-specific silencing are indicated by arrows. (Color figure online)
Mol Biol Rep (2014) 41:7115–7120
(Fig. 1e). When above-average siRNAs in ASP-score or discrimination ratio were investigated, approximately 30 % of siRNAs in the group of discrimination ratio appeared to retain an ability to inhibit 50 % or more of the wild-type allele expression, whereas there was no such siRNAs in the corresponding group of ASP-score (Fig. 1f). These findings suggested that ASP-score might be an appropriate criterion which is different from the conventional discrimination ratio. When taken together, the data we presented here suggest that Formula I or Formula II may give siRNA a matching ASP-score for its allele discrimination ability, or allele-specific silencing, and may allow for a more objective assessment of ASP-RNAi. Potential characteristics of optimal allele-specific siRNAs Of 119 siRNAs we designed and examined previously [15, 20], we aligned the siRNAs with reference to the position of target nucleotide variations (target mutations), and arranged them in descending order of the ASP-score. Figures 2a, b show the top 20 siRNAs conferring a strong ASP-RNAi. Based on the alignment, possible characteristics of optimal allele-specific siRNAs might be that: (i) a favorable 50 end may be at nucleotide position -9 to -8 relative to the mutation point that is given as 0 (Fig. 2b), or (ii) a favorable 30 end may be at nucleotide position 11–12 relative to the mutation point (Fig. 2b), and (iii) introduction of nucleotide mismatches at positions 3–7 relative to the mutation point may be effective for improving ASPRNAi (Fig. 2c, d). In addition, it looked like target mutations were located at a central position of their siRNAs. Because active RNA-induced silencing complexes (RISCs) can cleave target RNAs at the position corresponding to the center of guide siRNAs (nucleotide position between 10 and 11 from the 50 end of guide siRNA) which are incorporated into the RISCs [21, 22], it is conceivable that base pairing (matching) and un-pairing (mismatching) at the cleavage site of guide siRNAs greatly influence discrimination between target (mutant) and non-target (normal) RNAs. Accordingly, the position of nucleotide mismatch(es) (or un-pairing) of guide siRNAs against nontarget RNAs as well as target RNAs may be a key element for influencing allele discrimination, or allele-specific silencing, and also for designing siRNAs that confer ASPRNAi. To further confirm and understand the structural characteristics of optimal allele-specific siRNAs, more extensive studies need to be carried out. The formulas and ASP-score we presented here may help performance of such studies and contribute to clarifying optimal allelespecific siRNAs.
7119 Acknowledgements This work was supported by research grants from the Ministry of Health, Labour and Welfare of Japan, and also by Grants-in-Aid for Scientific Research and Yong Scientists from Japan Society for the Promotion of Science.
References 1. Dykxhoorn DM, Novina CD, Sharp PA (2003) Killing the messenger: short RNAs that silence gene expression. Nat Rev Mol Cell Biol 4:457–467 2. Meister G, Tuschl T (2004) Mechanisms of gene silencing by double-stranded RNA. Nature 431:343–349 3. Seyhan AA (2011) RNAi: a potential new class of therapeutic for human genetic disease. Hum Genet 130:583–605 4. Hohjoh H (2013) Disease-causing allele-specific silencing by RNA interference. Pharmaceuticals 6:522–535 5. Hohjoh H (2010) Allele-specific silencing by RNA interference. Methods Mol Biol 623:67–79 6. Sah DW, Aronin N (2011) Oligonucleotide therapeutic approaches for Huntington disease. J Clin Invest 121:500–507 7. Rodriguez-Lebron E, Gouvion CM, Moore SA, Davidson BL, Paulson HL (2009) Allele-specific RNAi mitigates phenotypic progression in a transgenic model of Alzheimer’s disease. Mol Ther 17:1563–1573 8. Rodriguez-Lebron E, Paulson HL (2006) Allele-specific RNA interference for neurological disease. Gene Ther 13:576–581 9. Feng X, Zhao P, He Y, Zuo Z (2006) Allele-specific silencing of Alzheimer’s disease genes: the amyloid precursor protein genes with Swedish or London mutations. Gene 371:68–74 10. Miller VM, Gouvion CM, Davidson BL, Paulson HL (2004) Targeting Alzheimer’s disease genes with RNA interference: an efficient strategy for silencing mutant alleles. Nucleic Acids Res 32:661–668 11. Ohnishi Y, Tamura Y, Yoshida M, Tokunaga K, Hohjoh H (2008) Enhancement of allele discrimination by introduction of nucleotide mismatches into siRNA in allele-specific gene silencing by RNAi. PLoS ONE 3:e2248 12. Ohnishi Y, Tokunaga K, Kaneko K, Hohjoh H (2006) Assessment of allele-specific gene silencing by RNA interference with mutant and wild-type reporter alleles. J RNAi Gene Silencing 2:154–160 13. Sierant M, Paduszynska A, Kazmierczak-Baranska J, Nacmias B, Sorbi S, Bagnoli S, Sochacka E, Nawrot B (2011) Specific Silencing of L392V PSEN1 Mutant Allele by RNA Interference. J Alzheimers Dis, Int, p 809218 14. Pfister EL et al (2009) Five siRNAs targeting three SNPs may provide therapy for three-quarters of Huntington’s disease patients. Curr Biol 19:774–778 15. Takahashi M, Watanabe S, Murata M, Furuya H, Kanazawa I, Wada K, Hohjoh H (2010) Tailor-made RNAi knockdown against triplet repeat disease-causing alleles. Proc Natl Acad Sci USA 107:21731–21736 16. Miller VM, Xia H, Marrs GL, Gouvion CM, Lee G, Davidson BL, Paulson HL (2003) Allele-specific silencing of dominant disease genes. Proc Natl Acad Sci USA 100:7195–7200 17. Huang H et al (2009) Profiling of mismatch discrimination in RNAi enabled rational design of allele-specific siRNAs. Nucleic Acids Res 37:7560–7569 18. Schwarz DS et al (2006) Designing siRNA that distinguish between genes that differ by a single nucleotide. PLoS Genet 2:e140
123
7120 19. Sibley CR, Wood MJ (2011) Identification of allele-specific RNAi effectors targeting genetic forms of Parkinson’s disease. PLoS ONE 6:e26194 20. Takahashi M, Katagiri T, Furuya H, Hohjoh H (2011) Diseasecausing allele-specific silencing against the ALK2 mutants, R206H and G356D, in fibrodysplasia ossificans progressiva. Gene Ther 19(7):781–785
123
Mol Biol Rep (2014) 41:7115–7120 21. Liu J et al (2004) Argonaute2 is the catalytic engine of mammalian RNAi. Science 305:1437–1441 22. Meister G, Landthaler M, Patkaniowska A, Dorsett Y, Teng G, Tuschl T (2004) Human Argonaute2 mediates RNA cleavage targeted by miRNAs and siRNAs. Mol Cell 15:185–197
A novel measurement of allele discrimination for assessment of allele-specific silencing by RNA interference Masaki Takahashi • Hirohiko Hohjoh
Received: 8 July 2013 / Accepted: 5 July 2014 / Published online: 19 July 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Allele-specific silencing by RNA interference (ASP-RNAi) is an atypical RNAi that is capable of discriminating target alleles from non-target alleles, and may be therapeutically useful for specific inhibition of diseasecausing alleles without affecting their corresponding normal alleles. However, it is difficult to design and select small interfering RNA (siRNAs) that confer ASP-RNAi. A major problem is that there are few appropriate measures in determining optimal allele-specific siRNAs. Here we show two novel formulas for calculating a new measure of allelediscrimination, named ‘‘ASP-score’’. The formulas and ASP-score allow for an unbiased determination of optimal siRNAs, and may contribute to characterizing such allelespecific siRNAs. Keywords RNA interference Small interfering RNA Allele-specific silencing Allele discrimination Novel formula Abbreviations RNAi RNA interference siRNA Small interfering RNA ASP-RNAi Allele-specific RNAi DR Discrimination ratio
Electronic supplementary material The online version of this article (doi:10.1007/s11033-014-3586-7) contains supplementary material, which is available to authorized users. M. Takahashi H. Hohjoh (&) Department of Molecular Pharmacology, National Institute of Neuroscience, NCNP, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187-8502, Japan e-mail: [email protected]
Introduction RNA interference (RNAi) is the gene silencing machinery triggered by 21–25-nucleotide-long small interfering RNA (siRNA) duplexes homologous to silenced genes, and has become a powerful tool for suppressing the expression of a gene of interest [1, 2]. The application of RNAi has been expanding to various fields including drug-discovery and medical science [3, 4]. Allele-specific RNAi, or ASP-RNAi, is an advanced application of RNAi, and it can specifically inhibit the expression of target alleles of interest, but not suppress the expression of their corresponding non-target alleles [4, 5]. Therefore, ASP-RNAi may become an applicable tool for suppression of disease-causing alleles with a minimal inhibition against their corresponding wild-type (normal) alleles; and its application as a therapeutic use for dominantly inherited diseases such as familial Alzheimer’s disease and Huntington’s disease may be particularly promising [4, 6–15]. To induce ASP-RNAi, the design and selection of siRNAs that confer ASP-RNAi is vital, but quite difficult. Generally, an in vitro assessment of designed siRNAs is performed using reporter genes carrying target (mutant) and non-target (normal) allelic sequences. In the conventional ways of assessment, the effects of designed siRNAs against target and non-target reporter alleles are separately (independently) examined, and subsequently a careful normalization for a precise comparison among such independent data needs to be carried out. Our previous studies established an in vitro heterozygous assay system for assessment of designed siRNAs for ASP-RNAi, and the system substantially mitigated the difficulty [5, 12]. Briefly, the Photinus and Renilla luciferase reporter genes carrying mutant (target) and normal (non-target) allelic sequences in their 30 untranslated regions were constructed. The effects of designed siRNAs against
123
7116
the target mutant reporter allele in allele-specific silencing, and off-target silencing against the normal reporter allele, were simultaneously examined under a heterozygous condition generated by cotransfection of the reporter alleles and siRNAs into cultured mammalian cells. Introduction of an artificial nucleotide mismatch (es) into siRNAs often improves allele discrimination, or allelespecific silencing [4, 10, 11, 14, 16–19]. However, the followings are still unpredictable: which nucleotide position in introducing a mismatch into siRNAs is effective for improving ASP-RNAi? and which nucleotide of three mismatched nucleotides can work most effectively? Currently, designed siRNAs including such mismatched (modified) siRNAs must be tested one by one, and a precise assessment needs to be carried out. A problem over the course of assessment of designed siRNAs is that there are few appropriate measures in determining optimal allele-specific siRNAs, regardless of the data obtained from either conventional methods or our in vitro assay system. In this report, we describe two novel formulas for calculating a new measure of allele discrimination, named ‘‘ASP-score’’. The formulas and ASP-score may provide us with a precise determination of optimal allele-specific siRNAs and may contribute to understanding the characteristics of such allele-specific siRNAs.
Mol Biol Rep (2014) 41:7115–7120
Photinus and Renilla luciferase reporter genes carrying mutant and wild-type allelic sequences in their 30 untranslated regions were constructed as a mutant allele and a wild-type allele, respectively; and vice versa. The day before transfection, HeLa cells were treated with trypsin, suspended in fresh medium without antibiotics, and seeded onto 96-well culture plates at a cell density of 1 9 105 cells/cm2. The pGL3-TK-backbone reporter plasmid (60 ng), phRL-TK-backbone reporter plasmid (10 ng), pSV-b-Galactosidase control vector (20 ng) (Promega, Fitchburg, WI, USA), and 20 nM (final concentration) of siRNA duplex were added to each well; a lipofectamine 2000 transfection reagent (Invitrogen) was used for transfection according to the manufacturer’s instructions. Cell lysate was prepared 24 h after transfection, and the expression level of luciferase and b-galactosidase was examined using a Dual-Luciferase Reporter Assay System (Promega) and a Beta-Glo Assay System (Promega), respectively. The luminescent signals were measured using a Fusion Universal Microplate Analyzer (PerkinElmer, Waltham, MA, USA). The expression level of mutant and wild-type reporter alleles was normalized to the level of bgalactosidase studied as a control, and the ratios of the mutant and wild-type reporter expression in the presence of test siRNAs were further normalized against the control ratios obtained with a non-silencing control siRNA (siControl).
Materials and methods Cell culture
Results and discussion
HeLa cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Wako) supplemented with 10 % fetal bovine serum (Invitrogen), 100 units/ml penicillin, and 100 lg/ml streptomycin (Wako) at 37 °C in a 5 % CO2 humidified chamber.
Conventional discrimination ratio and novel ASP-score
RNA and DNA oligonucleotides DNA oligonucleotides and siRNAs used in this study were synthesized by and purchased from Sigma-Aldrich (St Louis, MO, USA). Construction of reporter alleles, transfection and heterozygous reporter assay Construction of mutant and wild-type reporter alleles, transfections, and the in vitro assay using the reporter alleles were carried out as described previously [5, 11, 12, 15, 20]. The DNA oligonucleotide sequences of the target and non-target alleles used in the construction of the reporter alleles and the sequences of siRNAs examined are indicated in Supplementary Table S1 and S2. Briefly, the
123
The assessment of siRNAs designed for ASP-RNAi depends upon an in vitro reporter assay, and the data obtained from target and non-target reporter alleles are compared each other after normalization with the data obtained from a control reporter. The most important part in the assay is about how to evaluate the allele discrimination between target and non-target alleles precisely, and discrimination ratio (target-allele expression/non-targetallele expression) is usually used as a measure of allele discrimination. However, the discrimination ratio is not a perfect measure and may be sometimes not suitable for such an assessment; for example, when siRNAs confer a strong inhibition to target alleles together with a moderate suppression to non-target alleles (detailed in Fig. 1), it may be difficult to select competent siRNAs for ASP-RNAi only by discrimination ratio. Therefore it is of importance and necessary to develop a new measure for allele discrimination as an alternative to the discrimination ratio, for a precise and objective assessment of siRNAs designed for ASP-RNAi.
Mol Biol Rep (2014) 41:7115–7120
7117
Fig. 1 Comparison between ASP-score and discrimination ratio (DR). a In vitro heterozygous reporter assay. The effects of representative siRNAs (indicated) that were studied previously [15, 20] on target mutant reporter alleles (open bars) and non-target wildtype reporter alleles (solid bars) were exhibited. Data are average of four independent determinations. Error bars represent standard deviations. b Discrimination ratios. Discrimination ratios (DR) (target-allele expression/non-target-allele expression) of indicated siRNAs were calculated based on the data in A. c, d ASP-scores. ASP-scores of indicated siRNAs were calculated by Formula I (c) and Formula II (d) with the same data in A. e Histogram of siRNAs in DR and ASP-score. Histogram was created based on DRs and ASP-scores
of 119 siRNAs which we designed and examined in the previous studies [15, 20]. X-axis is a relative value of DR or ASP-score. Difference in the distribution of the siRNAs between DR and ASPscore graphs was analyzed by a v2 test, and a significant difference between them was detected (p = 0.0004). f Abundance ratio of siRNAs capable of inhibiting wild-type alleles as well as mutant alleles. Of 119 siRNAs, above-average siRNAs in DR ([Ave-DR) or ASP-score ([Ave-ASP) were grouped, and the percentage of siRNAs that retained an ability to inhibit 50 % or more of the expression of wild-type alleles was examined. In addition, whole siRNAs (all siRNAs) as a control group were also examined
We developed two novel formulas for calculating a new measure of allele discrimination between target and nontarget alleles, and named the new measure ‘‘ASP-score’’: Formula I: [non-target (normal) allele Exp.R.—target (mutant) allele Exp.R.] 9 non-target (normal) allele Exp.R. Formula II: [non-target (normal) allele Exp.R.—target (mutant) allele Exp.R.] 9 [1—target (mutant) allele Exp.R.] Non-target (normal) allele Exp.R.: relative expression ratio of a normalized expression level of a non-target allele treated with a test siRNA to a normalized expression level of the non-target allele treated with a control non-silencing siRNA (siControl) as 1. Target (mutant) allele Exp.R.: relative expression ratio of a normalized expression level of a target allele treated
with the test siRNA to a normalized expression level of the target allele treated with the siControl as 1. Formula I is suitable for selection of siRNAs conferring a marked allele-discrimination without affecting non-target alleles, and Formula II is capable of selecting siRNAs that confer allele-discrimination with a strong inhibition against target alleles even though the siRNAs retain some ability to cleave non-target alleles (Supplementary Fig. S1); in other words, if a severe adverse effect by suppressing normal alleles is a concern, and if a strong inhibition against disease-causing alleles in spite of some suppression of normal alleles is required, Formula I and II, respectively, may be useful for selecting optimal allelespecific siRNAs against target disease-causing alleles.
123
7118
Assessment of siRNAs based on ASP-scores We calculated both ASP-scores and discrimination ratios of siRNAs that we previously examined by our in vitro assay system [15, 20]. Figure 1 shows representative examples. The siG356D_A10(G18), siRs331_T9 and siRs331_T11 siRNAs, which conferred a strong inhibition and also a moderate inhibition to their target and non-target alleles, respectively (Fig. 1a), exhibited a high value in discrimination ratio (Fig. 1b). The siG356D_A9_3- and siRs331_T10 siRNAs that appear to confer an ideal ASP-RNAi (Fig. 1a), similarly indicated a high value in discrimination ratio (Fig. 1b). When their ASP-scores were calculated by Formula I, the ASP-scores of siG356D_A10(G18), siRs331_T9 and siRs331_T11 were
Fig. 2 Characteristics of competent siRNAs conferring allele-specific silencing. a Top 20 siRNAs conferring a strong ASP-RNAi. Based on the ASP-scores of 119 siRNAs we designed previously [15, 20], the top 20 siRNAs were arranged in descending order of the ASC-score, and aligned with reference to the position of target nucleotide variations (target mutations) which were highlighted in blue. Introduced mismatch-nucleotides are indicated by small letters and highlighted in green. The position of 50 end, 30 end and introduced mismatches (MM), relative to the target mutation point given as 0, is indicated (right panel). The expression levels of target and non-target reporter alleles treated by the top 20 siRNAs are indicated in Supplementary Fig. S2. b Graphic representation of the relation between ASP-scores and the position of the 50 or 30 ends of
123
Mol Biol Rep (2014) 41:7115–7120
markedly lower than those of siG356D_A9_3- and siRs331_T10, and also appeared to be as low as those of the siRs273_A8 and siG356D_A10(G16) siRNAs that were incapable of inducing RNAi (Fig. 1c). Accordingly, siG356D_A10(G18), siRs331_T9 and siRs331_T11 were rated as unsuitable siRNAs for ASP-RNAi by their ASP-scores. When Formula II was used instead of Formula I, the ASP-score of siG356D_A9_3- and siRs331_T10 remained a high value; however, siG356D_A10(G18), siRs331_T9 and siRs331_T11, which exhibited a similar discrimination ratio, indicated a different profile in ASP-score (Fig. 1d): this difference might be attributable to the level of silencing of non-target alleles. There was a significant difference in distribution of siRNAs between ASP-score and discrimination ratio
the top 20 siRNAs. The position of either the 50 or 30 ends of the siRNAs, relative to the mutation point given as 0, was exhibited. c Mismatched position. The relation between ASP-scores and the position of introduced mismatch-nucleotides relative to the mutation point was indicated. All designed siRNAs that carried mismatched nucleotides were examined and the siRNA belonging to the top 20 siRNAs were indicated by green circles. d Schematic drawing of a sense-strand siRNA. Open and solid circles represent nucleotides, and the solid circle indicates the position corresponding to a target mutation point. Possibly favorable 50 and 30 ends and mismatch position of siRNAs conferring a strong allele-specific silencing are indicated by arrows. (Color figure online)
Mol Biol Rep (2014) 41:7115–7120
(Fig. 1e). When above-average siRNAs in ASP-score or discrimination ratio were investigated, approximately 30 % of siRNAs in the group of discrimination ratio appeared to retain an ability to inhibit 50 % or more of the wild-type allele expression, whereas there was no such siRNAs in the corresponding group of ASP-score (Fig. 1f). These findings suggested that ASP-score might be an appropriate criterion which is different from the conventional discrimination ratio. When taken together, the data we presented here suggest that Formula I or Formula II may give siRNA a matching ASP-score for its allele discrimination ability, or allele-specific silencing, and may allow for a more objective assessment of ASP-RNAi. Potential characteristics of optimal allele-specific siRNAs Of 119 siRNAs we designed and examined previously [15, 20], we aligned the siRNAs with reference to the position of target nucleotide variations (target mutations), and arranged them in descending order of the ASP-score. Figures 2a, b show the top 20 siRNAs conferring a strong ASP-RNAi. Based on the alignment, possible characteristics of optimal allele-specific siRNAs might be that: (i) a favorable 50 end may be at nucleotide position -9 to -8 relative to the mutation point that is given as 0 (Fig. 2b), or (ii) a favorable 30 end may be at nucleotide position 11–12 relative to the mutation point (Fig. 2b), and (iii) introduction of nucleotide mismatches at positions 3–7 relative to the mutation point may be effective for improving ASPRNAi (Fig. 2c, d). In addition, it looked like target mutations were located at a central position of their siRNAs. Because active RNA-induced silencing complexes (RISCs) can cleave target RNAs at the position corresponding to the center of guide siRNAs (nucleotide position between 10 and 11 from the 50 end of guide siRNA) which are incorporated into the RISCs [21, 22], it is conceivable that base pairing (matching) and un-pairing (mismatching) at the cleavage site of guide siRNAs greatly influence discrimination between target (mutant) and non-target (normal) RNAs. Accordingly, the position of nucleotide mismatch(es) (or un-pairing) of guide siRNAs against nontarget RNAs as well as target RNAs may be a key element for influencing allele discrimination, or allele-specific silencing, and also for designing siRNAs that confer ASPRNAi. To further confirm and understand the structural characteristics of optimal allele-specific siRNAs, more extensive studies need to be carried out. The formulas and ASP-score we presented here may help performance of such studies and contribute to clarifying optimal allelespecific siRNAs.
7119 Acknowledgements This work was supported by research grants from the Ministry of Health, Labour and Welfare of Japan, and also by Grants-in-Aid for Scientific Research and Yong Scientists from Japan Society for the Promotion of Science.
References 1. Dykxhoorn DM, Novina CD, Sharp PA (2003) Killing the messenger: short RNAs that silence gene expression. Nat Rev Mol Cell Biol 4:457–467 2. Meister G, Tuschl T (2004) Mechanisms of gene silencing by double-stranded RNA. Nature 431:343–349 3. Seyhan AA (2011) RNAi: a potential new class of therapeutic for human genetic disease. Hum Genet 130:583–605 4. Hohjoh H (2013) Disease-causing allele-specific silencing by RNA interference. Pharmaceuticals 6:522–535 5. Hohjoh H (2010) Allele-specific silencing by RNA interference. Methods Mol Biol 623:67–79 6. Sah DW, Aronin N (2011) Oligonucleotide therapeutic approaches for Huntington disease. J Clin Invest 121:500–507 7. Rodriguez-Lebron E, Gouvion CM, Moore SA, Davidson BL, Paulson HL (2009) Allele-specific RNAi mitigates phenotypic progression in a transgenic model of Alzheimer’s disease. Mol Ther 17:1563–1573 8. Rodriguez-Lebron E, Paulson HL (2006) Allele-specific RNA interference for neurological disease. Gene Ther 13:576–581 9. Feng X, Zhao P, He Y, Zuo Z (2006) Allele-specific silencing of Alzheimer’s disease genes: the amyloid precursor protein genes with Swedish or London mutations. Gene 371:68–74 10. Miller VM, Gouvion CM, Davidson BL, Paulson HL (2004) Targeting Alzheimer’s disease genes with RNA interference: an efficient strategy for silencing mutant alleles. Nucleic Acids Res 32:661–668 11. Ohnishi Y, Tamura Y, Yoshida M, Tokunaga K, Hohjoh H (2008) Enhancement of allele discrimination by introduction of nucleotide mismatches into siRNA in allele-specific gene silencing by RNAi. PLoS ONE 3:e2248 12. Ohnishi Y, Tokunaga K, Kaneko K, Hohjoh H (2006) Assessment of allele-specific gene silencing by RNA interference with mutant and wild-type reporter alleles. J RNAi Gene Silencing 2:154–160 13. Sierant M, Paduszynska A, Kazmierczak-Baranska J, Nacmias B, Sorbi S, Bagnoli S, Sochacka E, Nawrot B (2011) Specific Silencing of L392V PSEN1 Mutant Allele by RNA Interference. J Alzheimers Dis, Int, p 809218 14. Pfister EL et al (2009) Five siRNAs targeting three SNPs may provide therapy for three-quarters of Huntington’s disease patients. Curr Biol 19:774–778 15. Takahashi M, Watanabe S, Murata M, Furuya H, Kanazawa I, Wada K, Hohjoh H (2010) Tailor-made RNAi knockdown against triplet repeat disease-causing alleles. Proc Natl Acad Sci USA 107:21731–21736 16. Miller VM, Xia H, Marrs GL, Gouvion CM, Lee G, Davidson BL, Paulson HL (2003) Allele-specific silencing of dominant disease genes. Proc Natl Acad Sci USA 100:7195–7200 17. Huang H et al (2009) Profiling of mismatch discrimination in RNAi enabled rational design of allele-specific siRNAs. Nucleic Acids Res 37:7560–7569 18. Schwarz DS et al (2006) Designing siRNA that distinguish between genes that differ by a single nucleotide. PLoS Genet 2:e140
123
7120 19. Sibley CR, Wood MJ (2011) Identification of allele-specific RNAi effectors targeting genetic forms of Parkinson’s disease. PLoS ONE 6:e26194 20. Takahashi M, Katagiri T, Furuya H, Hohjoh H (2011) Diseasecausing allele-specific silencing against the ALK2 mutants, R206H and G356D, in fibrodysplasia ossificans progressiva. Gene Ther 19(7):781–785
123
Mol Biol Rep (2014) 41:7115–7120 21. Liu J et al (2004) Argonaute2 is the catalytic engine of mammalian RNAi. Science 305:1437–1441 22. Meister G, Landthaler M, Patkaniowska A, Dorsett Y, Teng G, Tuschl T (2004) Human Argonaute2 mediates RNA cleavage targeted by miRNAs and siRNAs. Mol Cell 15:185–197
