How to Distinguish between siRNA, shRNA, and miRNA

In 2000, "Science" magazine recognized "Revisiting RNA" as one of the top ten breakthroughs in technology, highlighting siRNA, shRNA, and miRNA as stars in the RNA field. These RNAs are pivotal in RNA interference (RNAi), a phenomenon where specific short double-stranded RNAs induce gene silencing. Andrew Z. Fire and colleagues demonstrated that double-stranded RNA's inhibitory effect surpasses that of antisense RNA, coining the term RNAi.

RNAi garnered significant scientific acclaim, named a top ten breakthrough in 2001 and 2002 by "Science" and "Nature," respectively. In 2003, small RNA research continued its prominence, once again making the "Science" top ten list. Andrew Z. Fire and Craig C. Mello were awarded the Nobel Prize in Physiology or Medicine in 2006 for their research on RNAi mechanisms.

In 2018, gene silencing drugs based on RNAi principles successfully entered the market, reaffirming their status as a top ten breakthrough. This underscores the profound scientific significance of RNAi research, where siRNA, shRNA, and miRNA play crucial roles.

siRNA

siRNA, short for small interfering RNA, refers to a class of double-stranded RNA molecules approximately 21-23 nucleotides in length (literature reports suggest a range from 19 to 25 nucleotides). These molecules typically possess two unpaired nucleotides at the 3' end and function through the RNA interference mechanism to induce gene silencing of target genes.

Figure 1. Structure of siRNA

shRNA

Short hairpin RNA (shRNA) is a type of short double-stranded RNA structure with a specific stem-loop sequence, typically ranging in length from 19 to 25 nucleotides. This structure can be introduced into cells via vectors and undergo enzymatic cleavage to form siRNA within the cell. Subsequently, these siRNAs achieve precise modulation of target genes through the RNA interference mechanism. This technology provides a robust tool in modern molecular biology research, aiding in the deeper understanding of the intricate mechanisms governing gene expression regulation.

Figure 2. Structure of shRNA

miRNA

MicroRNAs (miRNA) are small RNA molecules ubiquitous in eukaryotes, typically measuring approximately 22 nucleotides in length (ranging from 21 to 24 nucleotides). Mature miRNAs exist as single-stranded RNA molecules and primarily regulate gene expression of target genes through mRNA degradation or inhibition of protein translation. They are widely distributed across eukaryotic organisms and play crucial roles in various biological processes.

Figure 3. Structure of miRNA

From a structural and definitional perspective, an in-depth examination reveals three primary similarities and distinctions among siRNA, shRNA, and miRNA:

Initially, these molecules share a characteristic in their mature form, typically spanning 20 to 25 nucleotides (nt). Specifically, siRNA manifests as double-stranded RNA (dsRNA) featuring two unpaired nucleotides at the 3' terminus and a phosphate group at the 5' terminus. shRNA, post-processing, transforms into siRNA, whereas miRNA ultimately adopts a single-stranded RNA structure.

Secondly, concerning their origins and structural attributes, siRNA predominantly comprises artificially synthesized linear dsRNA, whereas both shRNA and miRNA form local dsRNA structures through hairpin formation. Notably, siRNA and shRNA exhibit fully complementary double-stranded regions, while miRNA often displays imperfect complementarity.

Lastly, although shRNA and miRNA exhibit some structural resemblances, functionally they align more closely with siRNA. Specifically, post-Dicer enzyme cleavage within cells, shRNA can generate siRNA and exert interference via the siRNA pathway. In contrast, miRNA regulates target genes through a distinct pathway.

Despite their shared traits and potential for substitutability under specific contexts, each plays a distinct role in RNA interference mechanisms. Subsequent sections will delve into a detailed analysis of their specific associations and divergences.

Differences and Connections among siRNA, shRNA, and miRNA

shRNA, constructed via vectors to achieve siRNA in vivo, shares structural similarities with siRNA but demonstrates functional distinctions. Meanwhile, miRNA, another type of small RNA, exhibits close associations with siRNA and RNAi, yet distinct differences from siRNA are evident across multiple dimensions. Subsequent sections will delve into detailed discussions on the relationships and disparities between shRNA and siRNA, as well as miRNA and siRNA.

shRNA and siRNA

Figure 4. Schematic diagram of gene regulation by shRNA and siRNA

The figure above illustrates the fundamental mechanisms by which shRNA and siRNA regulate gene expression within cells, despite their different pathways, they converge on a common objective. siRNA can be synthesized exogenously or derived from longer double-stranded RNA precursors. These dsRNAs enter cells through transfection and undergo a series of enzymatic processes in the cytoplasm to mature into functional siRNAs. Mature siRNAs subsequently associate with the RNA-induced silencing complex (RISC), leading to mRNA degradation through complementary base pairing.

On the other hand, shRNA relies on vectors, viruses, or Agrobacterium as carriers to enter cells via transfection or infection. Within the nucleus, shRNA transcripts form hairpin double-stranded structures and undergo processing by enzymes like Drosha to generate mature shRNAs. Afterward, mature shRNAs are transported to the cytoplasm where they are processed into siRNA by ribonucleases, entering the RNAi pathway.

In summary, while shRNA and siRNA employ distinct mechanisms for gene regulation within cells, both ultimately achieve mRNA targeting, demonstrating the diversity and flexibility of gene regulation mechanisms. Despite their functional similarities in final outcomes, shRNA and siRNA are not interchangeable.

This distinction primarily stems from several key aspects:

Transfection Pathways Variability

Initially, direct transfection of siRNA or long double-stranded RNA in animal cells shows suboptimal efficiency. Similarly, direct transfection is unsuitable for plant cells due to challenges in protoplast regeneration. In contrast, viral vectors such as lentiviruses or adenoviruses carrying shRNA exhibit markedly superior infection efficiency. Moreover, when shRNA vectors are transformed into Agrobacterium, they readily infect plant callus tissues.

Divergent Interference Mechanisms

Figure 5. siRNA interference pathway

According to the findings of Jarve et al., upon transfection into cells, long dsRNA rapidly translocates into the cell nucleus within 15 minutes. Subsequently, over the next four hours, it extensively distributes in both intact and dissociated forms throughout the cytoplasm. Within the cytoplasm, dsRNA forms complexes with Dicer, TRBP (known variably as R2D2 in Drosophila), and PACT, the latter being an activator protein. Dicer, leveraging its double-stranded RNA-specific ribonuclease III activity, efficiently processes the dsRNA into siRNA molecules typically ranging from 21 to 23 nucleotides in length, bearing two unpaired nucleotides.

These siRNA fragments subsequently associate with the RNA-induced silencing complex (RISC). Importantly, siRNAs that are meticulously designed and processed prior to cell transfection can bypass the processing by Dicer, TRBP, and PACT, directly binding to RISC. RISC primarily comprises proteins such as Argonaute-2, Dicer, and TRBP.

Following association with RISC, the two strands of siRNA separate, with one strand dissociating from the complex. The RISC complex carrying the antisense strand is responsible for identifying and guiding the complex to target mRNA, ultimately leading to the cleavage and degradation of the mRNA.

Figure 6. shRNA interference pathway

Compared to siRNA, the synthesis of shRNA occurs within the cell nucleus. Upon entering the cell via transient or stable transfection methods, the shRNA vector is initially transported to the nucleus. Inside the nucleus, shRNA is processed by the Drosha/DGCR8 complex, converting it into pre-shRNA. Subsequently, the Exportin-5 protein transports pre-shRNA into the cytoplasm. Within the cytoplasm, the Dicer complex removes the hairpin structure of pre-shRNA, thereby generating siRNA. Following this, siRNA binds to the RNA-induced silencing complex (RISC) and releases one RNA strand, enabling the complex to identify and degrade target mRNA.

Thus, it is evident that the processing pathway of shRNA post-nuclear export is broadly similar to that of dsRNA, ultimately forming a siRNA structure. However, shRNA exhibits superior characteristics in terms of stability and sustainability.

Duration of Action

Due to its transient conversion nature, siRNA cannot achieve prolonged maintenance and detection, nor does it possess the ability to be inherited by subsequent generations. In contrast, shRNA can integrate into the cellular genome, ensuring its intended effects persist in descendant cells.

Expression Levels

Within cells, siRNA levels are directly influenced by RNA quantity and transfection efficiency and decrease over time. However, shRNA can precisely control expression levels through inducible promoters and maintains sustained expression.

Off-Target Probability

Despite the high specificity of RNAi as a gene silencing mechanism, it still exhibits specific and non-specific off-target effects. Specific off-target effects may arise from partial complementarity between the sense or antisense siRNA strands and non-target mRNA sequences, posing potential issues in both siRNA and shRNA applications. According to literature, off-target inhibition can occur with as few as seven nucleotides of complementarity, influenced by factors such as sequence composition surrounding the complementary region, the specific location of the sequence on mRNA, and the number of copies of the corresponding sequence in mRNA.

Although chemical modifications of siRNA, such as methylation of the 2'-ribose of the second base, have been reported to effectively reduce off-target effects, caution is warranted in applying such modifications as they may decrease overall inhibitory capacity, thereby affecting the practical application of RNAi technology.

It is noteworthy that evidence suggests shRNA exhibits superior off-target effects compared to siRNA. Although the specific mechanisms behind this phenomenon require further exploration, one plausible explanation is the higher stability of shRNA transcribed in the cell nucleus, whereas siRNA may undergo degradation in the cytoplasm, leading to off-target effects.

miRNA and siRNA

According to early definitions of RNAi, miRNA was not initially included in its scope. Despite being discovered in 1993, the detailed functional studies of miRNA lagged behind those of siRNA with RNAi definitions primarily focusing on siRNA. However, with advancements in scientific research, three consecutive articles published in Science in 2001 elucidated the characteristics and functional roles of miRNA in Caenorhabditis elegans, revealing its crucial regulatory roles in biological growth and development.

As research progressed, scientists discovered numerous similarities between miRNA and siRNA, with both sometimes utilizing the same gene silencing pathways. Therefore, while there is no definitive literature concluding whether miRNA-mediated gene regulation falls under RNAi, many studies consider miRNA as another implementation pathway of RNAi. This article supports this viewpoint, namely that siRNA, shRNA, and miRNA each represent distinct implementations of RNAi, complementing each other to form a comprehensive framework of RNA interference.

Figure 7. RNAi pathways of miRNA and siRNA

The figure above provides a detailed overview of the specific pathways through which miRNA and siRNA execute RNAi within cells. Previously, we extensively discussed the pathway of siRNA. By comparison, it is evident that the miRNA pathway exhibits significant similarities to the shRNA pathway. Specifically, following transcription within the cell nucleus, the miRNA vector undergoes processing by the enzyme Drosha to form pre-miRNA. Subsequently, pre-miRNA is exported from the nucleus to the cytoplasm through the action of proteins such as Exportin-5. Within the cytoplasm, pre-miRNA undergoes further processing by enzymes like Dicer, which removes its hairpin structure. Afterwards, mature miRNA is loaded onto the RNA-induced silencing complex (RISC), simultaneously dissociating the other RNA strand. Subsequently, miRNA-RISC pairs with target mRNA in a manner that is not fully complementary. Ultimately, this process may lead to mRNA degradation or inhibition of its translation, thereby interfering with the expression of target genes.

It is crucial to clarify here that the prevailing view suggests that when miRNA achieves complete complementarity with mRNA, it triggers mRNA degradation. In cases of partial complementarity, miRNA binds to mRNA, thereby inhibiting protein translation. Indeed, when complementarity reaches a high level, it results in mRNA cleavage and degradation within the RNAi pathway, a process executed by Ago proteins possessing specific capabilities to cleave double-stranded RNA. However, specific literature defining the threshold for "high complementarity," such as 99% or 89%, remains absent.

In animal cell systems, the relationship between miRNA and mRNA is predominantly partial, typically anchoring within the mRNA's 3'UTR region. In contrast, in plant cells, complete complementarity between miRNA and mRNA is more common, often leading to mRNA degradation. Importantly, miRNA's complementarity involves interactions not only with mRNA but also within its own stem-loop double-stranded structure. In most cases, these two complementarity relationships do not simultaneously reach a complete state; if they do, miRNA's properties will approximate those of shRNA.

Additionally, Table 1 provides a detailed comparison of the general properties of siRNA and miRNA , serving as a crucial reference for related research.

Table 1 Characteristics of SiRNA and miRNA

miRNA and siRNA exhibit numerous similarities but also notable differences, often leading to confusion in their distinction. To clarify these differences comprehensively, this article elaborates on three dimensions: their origins, processing pathways, and interference mechanisms.

Origins

siRNA typically originates from external sources such as viral infections or artificially introduced dsRNA that undergoes cleavage to produce exogenous genes within cells. These genes exhibit transient characteristics and lack heritability. In contrast, miRNA is an endogenous substance derived from non-coding pri-mRNA that undergoes processing via cleavage. miRNA constitutes a regulatory system naturally evolved over long periods in organisms, characterized by temporal, conservative, and tissue-specific expression patterns.

Processing and Maturation

Figure 8. RNAi model in Drosophila

siRNA primarily originates from long double-stranded RNA, which is cleaved into double-stranded siRNA by Dicer enzyme. Each long double-stranded RNA can yield varying numbers of siRNA fragments through this process. In contrast, miRNA undergoes transcriptional cleavage within the nucleus before being released into the cytoplasm, where it undergoes Dicer-mediated cleavage to achieve maturity. It is noteworthy that both siRNA and miRNA undergo maturation processing by the same Dicer enzyme in humans and Caenorhabditis elegans. However, in Drosophila melanogaster and Arabidopsis thaliana, there exist two and four types of Dicer enzymes, respectively, resulting in siRNA processing by different Dicer enzymes in these organisms. Specifically, in Drosophila, pre-miRNA is processed by Dicer-1 in association with its partner protein Loqs, whereas dsRNA is processed by Dicer-2 and R2D2.

This clarification provides a detailed comparison of the mechanisms involved in siRNA and miRNA processing across different organisms, underscoring the nuanced differences in enzyme utilization for their maturation pathways.

Interference Mechanisms

Figure 9. Interference mechanism of miRNA and siRNA

Mature siRNA and double-stranded miRNA are loaded onto the RNA-induced silencing complex (RISC), after which one strand departs. Subsequently, RISC carrying antisense RNA binds to mRNA, effectively inhibiting target gene expression through mRNA degradation and translation repression pathways. Due to the diversity of small RNAs, there are differences between these two types of RISCs. The core enzyme of RISC is the Argonaute enzyme from the AGO protein family, which has multiple members in most species. In Drosophila and human cells, miRNA binds with AGO-1 protein to form non-slicing RISC, interfering with target gene expression by inhibiting mRNA translation. In contrast, siRNA binds with AGO-2 protein to form slicing RISC, disrupting target gene expression through mRNA cleavage and degradation.

In Arabidopsis thaliana, mRNA degradation primarily involves AGO-1, and this process includes siRNA or fully complementary miRNA. The assembly of RISC and its mechanism targeting mRNA are highly complex and vary significantly across different species.

Furthermore, there may be questions regarding the mechanism of action of miRNA. Typically, miRNA targets the 3'-UTR region of the target gene and does not directly cleave mRNA. However, considering that mRNA translation initiates from the 5' end, how does miRNA inhibit translation? When miRISC binds to the 3'-UTR region of mRNA, although it does not directly act on the 5' end, it can still influence the function of the 5' region through spatial folding. For instance, miRISC can compete with translation initiation factors for binding to mRNA, induce mRNA deadenylation, thereby inhibiting mRNA circularization, and induce premature ribosomal drop-off, thereby inhibiting the translation process.

Selection Between miRNA, siRNA, or shRNA

The preceding sections have extensively elucidated the differences among miRNA, siRNA, and shRNA. Researchers are encouraged to make informed decisions based on their experimental objectives. The summarized recommendations are as follows:

siRNA: Recommended when cells are amenable to transfection or electroporation, when researchers prioritize experimental turnaround time and rapid results, or when fluorescence or other molecular labeling of small RNAs is desired. siRNA can be synthesized directly by specialized RNA synthesis companies, and purification methods of PAGE or higher are recommended to ensure quality.

shRNA: Preferable when aiming for comprehensive gene silencing, specificity in gene silencing is desired, induction of interference effects at specific time periods is intended, proteins with longer half-lives need to be silenced, or when stable genetic traits in cells are required. In plant gene silencing studies, shRNA is also favored.

Plant Research: shRNA is considered the primary choice for gene silencing studies in plants.

Biological Trait Studies: shRNA is the preferred choice due to its unique advantages in RNAi-based biological trait studies.

miRNA: Highly favored for studies concerning endogenous non-coding small RNA regulation due to its specific functional characteristics.

Comprehensive Regulation Pathways: miRNA is unequivocally the best choice when investigating the comprehensive regulatory impact of a small RNA on all regulatory pathways associated with a trait.

Moderate Interference Effects: miRNA is preferred when target genes cannot be completely silenced to avoid potential cell death due to its milder interference effects.

In conclusion, the selection of RNA interference pathways should comprehensively consider experimental objectives and research requirements to ensure smooth experimentation and obtain accurate and reliable research outcomes.

Designing siRNA: Guidelines and Considerations

Target Selection: siRNA offers broad flexibility in targeting mRNA at any position. To ensure experiment validity and reliability, it is advisable to select multiple targets judiciously. Specificity can be verified using the NCBI Primer-Blast tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). Targets showing non-specific matches of 15 or more nucleotides should generally be avoided.

Optimal Sequence Length: Based on extensive practice and validation, siRNA sequences ranging from 19-25 nucleotides typically exhibit optimal efficacy. Sequences exceeding 30 nucleotides may induce non-specific silencing, thus compromising experimental accuracy and reliability.

Sequence Design Criteria: Ideal sequences should conform to the AA(N)19-20TT or NA(N)19-20TT pattern, with a GC content between 35-55%. Typically, siRNA terminates with a free 3' end of UU nucleotides, although other dinucleotide structures can also be considered. Care should be taken to avoid GG structures, as single-stranded G residues are susceptible to ribonuclease cleavage.

Control Groups: To ensure result accuracy and comparability, appropriate control groups must be established. Typically, this includes at least two control groups: a scrambled control designed by reordering the sequence to disrupt specific targeting (Scrambled Control), and a non-targeting control composed of a sequence that does not anchor to any known gene of the target organism.

Thermodynamic Considerations: In advanced applications, siRNA's thermodynamic properties should be carefully considered. Optimal siRNA target sites that facilitate efficient binding with RISC during silencing are critical for enhancing silencing efficiency.

Utilizing Online Tools: For streamlined workflow and enhanced design efficiency, leveraging specialized online tools for siRNA design and optimization is recommended. These tools integrate robust algorithms and database support to facilitate rapid generation of high-quality siRNA sequences, thereby expediting experimental processes.

shRNA: Transcription and Design Considerations

shRNA typically relies on RNA polymerase III for transcription in vivo, commonly utilizing the human U6 and human H1 promoters. According to Petri et al., U6 promoter exhibits more pronounced interference efficacy with shRNA in the mouse brain compared to the H1 promoter. Therefore, prioritizing the U6 promoter is strongly recommended as the transcriptional driving mechanism.

Key considerations for shRNA design include:

Target selection should adhere to siRNA design principles.

Due to the polyT structure acting as a termination signal for RNA polymerase III, it is crucial to ensure that the shRNA sequence does not contain four or more consecutive T bases.

When using the U6 or 7SK promoters, the preferred start base of the target sequence should be G.

Avoid overlapping with single nucleotide polymorphism (SNP) regions when selecting target sites to mitigate potential interference.

In addition to scrambled and non-targeting controls, include a mock control to establish a baseline.

Utilize online tools for assistance in design. For instance, Thermo Fisher Scientific provides the RNAi Designer tool

miRNA: Design Principles and Considerations

In the field of biology, miRNA exhibits structural similarities to shRNA, thus design principles can be referenced from guidelines applicable to shRNA. However, significant differences between the two warrant specific attention:

Firstly, as an endogenous molecule, miRNA design does not involve target selection but rather entails screening suitable candidates from established miRNA libraries or directly retrieving from pertinent literature. For instance, when investigating the regulatory effects of known miRNAs on specific genes within a particular species, further exploration can determine whether these miRNAs possess homologous gene-regulatory capabilities across other species. While early studies primarily implicated miRNAs targeting mRNA's 3'-UTR region, subsequent research has identified miRNAs targeting CDS and 5'-UTR regions. Nevertheless, priority remains with the 3'-UTR region, particularly in studies involving animal cells; alternatively, CDS and 3'-UTR regions are viable options for plant cells.

Secondly, pre-miRNAs exhibit specific structural frameworks that may not be processed correctly into mature forms across different species. Therefore, for interspecies studies, it is advisable to utilize pre-miRNA frameworks native to the target gene's host.

Moreover, while instances exist of directly linking mature miRNA sequences downstream of the U6 promoter, such practices are uncommon and not broadly recommended.

In terms of miRNA expression regulation, U6 and H1 promoters are commonly employed constitutive promoters. Notably, a major advantage of both shRNA and miRNA pathways lies in their capacity to utilize inducible promoters, facilitating in-depth studies of RNAi phenomena within specific periods or tissues.

Regarding reference databases for miRNA design, several resources are available to researchers. miRBase serves as a comprehensive sequence database, offering published miRNA sequences, annotations, predicted gene targets, and more, thereby serving as a crucial public repository for miRNA information. PMRD focuses specifically on plant microRNA research, integrating information such as microRNA sequences, target genes, secondary structures, and expression profiles. Additionally, databases like Arabidopsis Small RNA Database and miRDB provide extensive microRNA target prediction data, supporting further advancements in related fields.

Analyzing Interference Effects: Methods and Techniques

Transcriptional Level:

RT-qPCR Analysis

RT-qPCR, or Real-Time Quantitative Polymerase Chain Reaction, represents an efficient and direct method for assessing mRNA expression levels. Widely utilized in research, this method excels in detecting mRNA degradation induced by siRNA, shRNA, and perfectly matched miRNA interference pathways. Its straightforward operational principle makes it highly favored among researchers.

RNA-Seq Analysis

RNA-Seq, or Transcriptome Sequencing, is a technique that deeply investigates changes in RNA expression at the cellular transcriptome level. Particularly suited for exploring complex regulatory relationships between miRNA and multiple mRNAs, RNA-Seq provides robust support for predicting miRNA function and screening potential regulatory genes.

Protein Level Analysis:

Dual-Luciferase Assay

The Dual-Luciferase assay is an effective method for directly detecting gene silencing effects caused by shRNA or miRNA at the protein level. During the experiment, the target gene is fused with Firefly Luciferase (FLuc) and co-expressed with shRNA or miRNA vectors. Interaction between shRNA or miRNA and the target gene leads to degradation of FLuc mRNA or translational inhibition, resulting in reduced FLuc expression levels. Consequently, the measured FLuc values decrease. Notably, the Dual-Luciferase assay is the sole method capable of quantitatively assessing changes in the translational levels of target genes among numerous techniques.

Western Blot Analysis (WB)

WB experiments similarly detect RNAi effects at the protein level, albeit achieving qualitative or semi-quantitative assessments of regulatory intensity. WB analysis offers advantages of simplicity and relatively short experimental durations; however, it relies on the use of specific antibodies. Therefore, careful selection of high-quality specific antibodies is crucial to ensure the accuracy and reliability of experimental results during WB analysis.

References:

1. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., et al. (1998). Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature, 391(6669), 806–811.
2. Jarve, A., Muller, J., Kim, I. H., et al. (2007). Surveillance of siRNA integrity by FRET imaging. Nucleic Acids Res, 35(18), e124.
3. Rao, D. D., Vorhies, J. S., Senzer, N., et al. (2009). siRNA vs. shRNA: similarities and differences. Adv Drug Deliv Rev, 61(9), 746-759.
4. Lin, X., Ruan, X., Anderson, M. G., et al. (2005). siRNA-mediated off-target gene silencing triggered by a 7 nt complementation. Nucleic Acids Res, 33(14), 4527-4535.
5. Hannus, M., Beitzinger, M., Engelmann, J. C., et al. (2014). siPools: highly complex but accurately defined siRNA pools eliminate off-target effects. Nucleic Acids Res, 42(12), 8049-8061.
6. Snove Jr., O., Rossi, J. J. (2006). Chemical modifications rescue off-target effects of RNAi. ACS Chem. Biol., 1, 274-276.
7. Lam, J. K., Chow, M. Y., Zhang, Y., et al. (2015). siRNA Versus miRNA as Therapeutics for Gene Silencing. Mol Ther Nucleic Acids, 4, e252.
8. Kim, V. N., Han, J., Siomi, M. C. (2009). Biogenesis of small RNAs in animals. Nat Rev Mol Cell Biol, 10(2), 126-139.
9. Tang, G. (2005). siRNA and miRNA: an insight into RISCs. Trends Biochem Sci, 30(2), 106-114.
10. Carthew, R. W., Sontheimer, E. J. (2009). Origins and Mechanisms of miRNAs and siRNAs. Cell, 136(4), 642-655.
11. Pushparaj, P. N., Aarthi, J. J., Manikandan, J., et al. (2008). siRNA, miRNA, and shRNA: in vivo applications. J Dent Res, 87(11), 992-1003.
12. Makinen, P. I., Koponen, J. K., Karkkainen, A. M., et al. (2006). Stable RNA interference: comparison of U6 and H1 promoters in endothelial cells and in mouse brain. J Gene Med, 8(4), 433-441.
13. Moore, C. B., Guthrie, E. H., Huang, M. T., et al. (2010). Short hairpin RNA (shRNA): design, delivery, and assessment of gene knockdown. Methods Mol Biol, 629, 141-158.