Gene Expression Regulation in Eukaryotes (PDF)

7.2 RNA Interference ASOs-based therapies may be approved in the next years. Nevertheless, the failure of some clinical trials using ASOs or the toxicity effects observed in some studies also recommend proceeding with caution, in order to build solid preclinical and clinical data ensuring the safet...

7.2 RNA Interference ASOs-based therapies may be approved in the next years. Nevertheless, the failure of some clinical trials using ASOs or the toxicity effects observed in some studies also recommend proceeding with caution, in order to build solid preclinical and clinical data ensuring the safety and efficacy of therapies. 7.2 RNA Interference The discovery of the RNA interference (RNAi) pathway [21] revolutionized the molecular biology research field, providing important tools for studying genes and importantly constituting a new opportunity for gene therapy to treat genetic dominant disorders, where silencing of the mutant gene is essential to counteract the disease phenotype. Several observations made in plant experiments, based on the introduction of antisense, exogenous genes, led researchers to postulate the existence of a posttranscriptional gene silencing mechanism [22–25]. In those studies, researchers observed a reduction in the expression of endogenous and exogenous genes, in a phenomenon named co-suppression. However, it was only in 1998 that Andrew Fire and Craig C. Mello demonstrated an efficient interference of gene expression by double-stranded RNA molecules in an animal species - C. elegans [21]. In this breakthrough study, the Nobel-awarded researchers (2006 Nobel Prize in Medicine or Physiology) observed that the introduction of exogenous double-­stranded RNAs (dsRNAs) led to the silencing of a particular target gene. Now, it is known that RNAi is a cellular pathway conserved among several organisms, including in eukaryotes, being crucial in the regulation of gene expression and in the innate defense against invading viruses. RNAi-based tools proved very successful in biomedical research, allowing researchers to apply gene silencing strategies for target identification and validation. Their use as therapeutics is also quite promising, and several clinical trials have been performed or are ongoing using small regulatory RNAi molecules. Importantly, their use is widespread in different biotechnology areas, including crop improvement [26]. 135 7.2.1  ene Expression Regulation G in Eukaryotes The central dogma of molecular biology establishes a unique direction of information flow from the DNA to RNA and finally to protein. It is now known that there are exceptions to this dogma, although it remains actual and accurate for most of the eukaryotic protein-coding genes. The human genome is estimated to have around 20,000 genes that codify for proteins (Table 7.3), whereas the number of genes that codify for different RNA species (e.g., noncoding RNAs, small regulatory RNA) is much higher. However, in a given cell and at a particular timepoint, only approximatively half of these genes are expressed, i.e., transcribed into RNA [27, 28]. This happens because gene expression in each cell is tightly, and differently, regulated. Differences in gene expression provide different features and properties to each cell type of a particular organism, despite all of them having the same DNA in their nuclei. Some important genes are expressed in all cells (constitutive genes), whereas the expression of others is dependent on internal cell factors and external factors and stimuli. In order for cells to maintain their functions and homeostasis, gene expression is regulated at different points (Fig. 7.5). The first point of regulation is the chromatin control. Inside the cell nucleus, DNA binds to histone proteins, ­constituting the chromatin. When the DNA is tightly wrapped around histones, excess packaging prevents gene transcription, by preventing the coupling of transcription factors to the genes. Moreover, epigenetic events and the involvement of small regulatory RNAs contribute to gene silencing at the chromatin level, thus regulating gene expression in cells [29]. The next control step concerns the regulation of transcription. Even if the DNA is unwound and the genes are exposed, for transcription to occur several proteins need to be recruited. The binding of transcription factors can promote or repress transcription of a given gene, thus regulating its expression. Next, there is the regulation of gene expression at the RNA processing level, in which the transcribed RNA undergoes different modifica- 136 7 Gene Therapy Strategies: Gene Silencing Fig. 7.4 Antisense oligonucleotides-based therapy for Duchenne muscular dystrophy (DMD). DMD is caused by different mutations in the DMD gene, leading to abnormal mRNA splicing and resulting in the production of a prematurely truncated, unstable, and nonfunctional dys- trophin protein. A gene therapy for DMD has been approved, based on ASOs that target the splicing process. It leads to the skipping of exon 51, generating an internally deleted dystrophin protein that is nonetheless functional. tions, for example splicing, capping and addition of a poly-A tail, in order to yield functional mature RNA molecules to be exported from the nucleus. The control of these types of modifications also constitutes a level of gene expression regulation [30]. At this point, alternative splicing of primary transcripts is yet another way for cells to regulate gene expression, by altering the mRNA products of the same gene. The following regulation point concerns RNA stability. The lifetime of a mRNA molecule is an important factor for their translation into proteins. Binding of mRNA to RNA-binding proteins or to small regulatory RNAs such as RNAi molecules can prevent translation and target mRNA for degradation or, on the contrary, increase their stability promoting translation. Thus, RNA stability is now recognized as an important player in the regulation of gene expression [31]. The translation process is yet another important point, being divided into three phases: initiation, elongation and termination. Initiation is probably the most important step to be regulated, and players involved in the, process other than the mRNA molecules, such as initiation factors or ribosomes, can also be involved in the regulation of mRNA translation and, ultimately, of gene expression [32]. Finally, the post-translational modifications are the ultimate point of regulation of gene expression. Translated proteins can undergo a variety of modifications: some are permanent (e.g., proteolytic cleavage), whereas others are reversible (e.g., phosphorylation), but both types can affect protein activity, function and turnover, ultimately constituting a form of gene expression regulation. The RNAi pathway interferes with gene expression through the action of particular small RNA 7.2 RNA Interference 137 Table 7.3 Overview of the human genome genes and their function (according to GENCODE, Human Release 30). Gene class Messenger RNA Long noncoding RNA Small noncoding RNA MicroRNA Small nuclear RNA Small nucleolar RNA Antisense RNA Function Protein coding Estimated number 19,986 Gene regulation 16,193 Gene regulation 7,576 Translational inhibition and mRNA degradation Processing of pre-mRNA Processing of rRNA, tRNA, and snRNA 1,881 Gene regulation 1,901 942 5,611 species that bind to mRNA targets and lead to their cleavage or degradation, or block their translation. The following sections describe these molecules and their functions, in the context of the two subpathways into which RNAi can be divided. 7.2.2  he Small Interfering RNA T Pathway Small or short interfering RNAs (siRNA) are ~22-nucleotide-long, double-stranded molecules with two-nucleotide 3’overhangs. The siRNA pathway starts with the cleavage of double-­ stranded RNAs (dsRNAs) by Dicer to form the siRNAs. The dsRNAs’ can originate from viruses or transposons. The resulting siRNAs are then assembled into the minimal RNA-induced silencing complex (RISC) with an Argonaute protein. In this complex, the siRNA double strand is separated, with the passenger strand being discarded and the guide strand remaining linked to Argonaute. Then, RISC binds to a complementary mRNA sequence and silences it via the cleavage activity of Argonaute (Fig. 7.6). In humans, there are four Argonaute proteins (1–4), although only Ago2 presents slicer activity [33]. Nevertheless, other human Ago proteins can still participate in gene regulation by binding the target mRNA molecules and inhibiting their translation. 7.2.3 The MicroRNA Pathway MicroRNAs (miRNAs) are ~22-nucleotide-long, double-stranded molecules encoded by specific genes in the nucleus that were first discovered in 1993, independently from the RNAi pathway [34]. The miRNA pathway starts in the nucleus, where miRNA genes are transcribed by RNA polymerase II, forming the primary-miRNA (pri-miRNA), which is at least 1000-nucleotide long, with single or clustered double-stranded hairpins (Fig. 7.7). The pri-miRNA is then cleaved by Drosha in the nucleus, resulting in the formation of a precursor miRNA (pre-miRNA) with around 70 nucleotides. The pre-miRNA associates with Exportin-5 to be exported to the cytoplasm. There, the miRNA pathway converges with the siRNA pathway, and the pre-­ miRNA is also processed by Dicer and assembled into RISC (Fig. 7.6). 7.2.4  mall Interfering RNAs Versus S MicroRNAs Despite converging in the final steps of the RNAi pathway, siRNAs and miRNAs share important differences that provide some specific features to each pathway, including their origin, number of mRNA targets or the molecular mechanism of gene regulation (Table 7.4). One important difference between siRNAs and miRNAs is that siRNAs have a complete complementarity to the target mRNA (in the coding region), whereas miRNAs are only partially complementary, typically in the 3’UTR region of target mRNA (Fig. 7.8) (nevertheless, siRNAs designed with partially complementary binding sites have been implemented to study translational repression) [35]. From this difference emerge other important dissimilarities between the two pathways. First, if there is complete complemen- 138 Fig. 7.5 Possible points of gene expression regulation in eukaryotes. The starting point correspondes to chromatin remodeling (1), as condensed chromatin prevents gene transcription. The second point of regulation involves the control of the transcription process (2), through the binding of transcription factors. Next, RNA processing in the nucleus (3) also constitutes an opportunity for regulation of gene expression. In the cytoplasm, the transport of 7 Gene Therapy Strategies: Gene Silencing mRNA and, importantly, control of its stability are other points of gene expression regulation (4). The next level of control is protein translation (5), since different mechanisms and players can promote or repress the translation of a particular mRNA. Finally, several post-translational modifications (6) of the proteins can alter their functionality and turnover, thus constituting a last level of control of gene expression. 7.2 RNA Interference tarity, which is the case of the siRNA pathway, then Argonaute cleaves the mRNA through its catalytic activity. On the other hand, in the miRNA pathway, while the incomplete complementarity may also lead to the degradation of the target mRNA, this does not occur through the catalytic activity of Argonaute, and the pathway may also lead to translational repression of the mRNA instead. Second, due to its complete complementarity, siRNAs target one single mRNA, whereas miRNAs can target multiple mRNAs. This particularity of miRNAs may be an important advantage in the treatment of complex multigenic diseases using gene therapy, which require the targeting of multiple genes. Another important difference between the two RNAi pathways concerns the biogenesis of each regulatory molecule. MiRNAs are constitutively expressed by specific genes, whereas siRNAs have their origin in transposons and viruses. In the context of gene therapy applications, both siRNAs and miRNAs have advantages and disadvantages, although the former was far more tested in clinical studies. Nevertheless, they are both quite attractive as therapeutic tools, as they can target virtually any gene, which is difficult or almost impossible to achieve using small molecules or protein-based products. 7.2.5  mall Interfering RNAs Versus S Short Hairpin RNAs The use of siRNAs as a molecular tool for gene knockdown experiments or as a therapeutic agent has some important disadvantages. For example, siRNAs do not allow a long-term expression, limiting their use in cells to a 72 hours maximum and demanding continuous applications in their use as therapy. The transient effect also requires that siRNAs be administered in high doses, to allow a therapeutic effect. This important fact contributes to innate immune response and toxicity events, detected upon synthetic siRNAs administration, which strongly limits their clinical application as therapeutic agents. Moreover, in systemic delivery, siRNAs are rapidly degraded by blood nucleases, and their uptake is low in most organs and cells (except for the liver). 139 Trying to overcome these limitations, an alternative to siRNAs can be the use of short hairpin RNAs (shRNAs), which are artificial noncoding RNAs with a stem-loop structure, which can be expressed in the nucleus and use part of the miRNA pathway. They mimic a pri-miRNA and are processed by Drosha and exported to the cytoplasm by Exportin-5. The shRNAs are usually delivered through viral vectors and thus can achieve a permanent expression when using integrative virus. Like siRNAs, they are fully complementary to the target mRNA, although, in the last years, several studies showed less off-target effects of shRNAs comparing to siRNAs (Table 7.4) [36]. 7.2.6  ene Therapy Applications of G RNAi The discovery of the RNAi mechanism led to a huge development in gene therapy preclinical trial studies, which ultimately resulted in several clinical studies involving siRNAs, shRNAs and miRNAs. Besides the efficacy issue, clinical studies involving RNAi molecules also focus on toxicity and delivery aspects. To date, more than 30 clinical trials were carried out using RNAi molecules (mainly siRNAs) for different therapeutic indications, especially cancer and ophthalmic conditions [37]. Cancer is a preferential target for RNA silencing molecules, aiming to inhibit genes related to uncontrolled cell growth, angiogenesis, metastasis and drug resistance. On the other hand, ocular conditions are also a good target as eyes are immuneprivileged sites where local delivery can be achieved. Nevertheless, siRNA, shRNA or miRNA molecules were also studied as therapeutic effectors for several other conditions, including cardiovascular or infectious diseases. Recently the first RNAi-based gene therapy product was approved in Europe and the USA for hereditary transthyretin-­ mediated amyloidosis. Onpattro™ (patisiran) is a siRNA encapsulated within a liposome nanoparticle, administrated intravenously once every 3 weeks, which was demonstrated to elicit an average knockdown of 87% (maximum of 96%) of the protein causing the disease (transthyretin - TTR) [20]. 140 Fig. 7.6 The RNA interference (RNAi) pathway is a cellular pathway conserved among several organisms, including eukaryotes, being crucial in the regulation of gene expression and in the innate defense against invading viruses. RNAi has two main subpathways: the small interfering RNA (siRNA) and the microRNA (miRNA) pathways. The siRNA pathway starts with the introduction of a dsRNA molecule into the cell (1A), while the 7 Gene Therapy Strategies: Gene Silencing miRNA pathway is initiated with the transcription of pri-­ miRNA and its processing in the nucleus (1B). In the cytoplasm of the cell, the two pathways converge, as both molecules are processed by Dicer (2) and assembled into the RISC complex (3). This is followed by the elimination of the passenger strand (4) and the target mRNA silencing (5) through mRNA cleavage or by translational repression. 7.2 RNA Interference 141 Fig. 7.7 MicroRNA secondary structure, highlighting the processing sites for Drosha and Dicer enzymes, as well as the seed region, which is the key binding location responsible for translation silencing. In the nucleus, the pri-miRNA is cleaved by Drosha into pre-miRNA, which is 60–70-nucleotide long. In the cytoplasm, the pre-­ miRNA, with its characteristic loop, is cleaved by Dicer, yielding the mature miRNA molecule. Table 7.4 Comparison between the main characteristics of the different molecules of regulatory RNAs from the RNAi pathway that can be used in gene therapy. Properties Origin Structure (prior to Dicer processing) Structure Complementarity siRNA Transposons, viruses, exogenous (synthetic) Double-stranded RNA that contains 30 to over 100 nucleotides 21–23 nucleotide RNA duplex Fully complementary to target mRNA (coding region) miRNA Encoded by their own genes, exogenous (synthetic) Pre-miRNA contains 70–100 nucleotides with mismatches and hairpin structure 19–25 nucleotide RNA duplex mRNA target Mechanism of gene expression regulation Delivery to the cell One Cleavage of mRNA Non-viral systems Partially complementary to mRNA, typically targeting the 3’UTR region of the target mRNA Multiple Cleavage of mRNA / translational repression / degradation of mRNA Viral and non-viral systems Persistence Degraded after 48 hours Expressed for up to 3 years Dosage required Likelihood of “off-target” effects High Low Low High 7.2.7 RNAi Terms Glossary dsRNA – a long double-stranded noncoding RNA with more than 100 nucleotides, from different sources (e.g., transposons, viruses or artificially introduced in cells), which is processed by Dicer generating siRNAs. shRNA Exogenous but encoded by genes (synthetic) Double-stranded with loop 21–23 nucleotide RNA duplex Fully complementary to target mRNA (coding region) One Cleavage of mRNA Viral and non-viral systems Expressed for up to 3 years Low Low siRNA – a noncoding regulatory small double-­ stranded RNA molecule with 21–23 nucleotides, with two-nucleotide 3’ overhangs, originated from Dicer cleavage of dsRNAs. microRNA – a noncoding regulatory double-­ stranded RNA molecule (~22 nucleotides) encoded by specific genes in the nucleus and processed by Drosha and Dicer. 7 142 Gene Therapy Strategies: Gene Silencing Fig. 7.8 Complementarity of siRNA and miRNA to target mRNA. siRNAs have a complete complementarity to target mRNA (in the coding region), whereas miRNAs are only partially complementary, through the seed region, typically to the 3’UTR region of the target mRNA. shRNA – a noncoding small regulatory double-­ stranded RNA molecule that contains a loop structure, which is artificially introduced in cells, expressed in the nucleus, and processed to siRNA by Dicer in the cytoplasm. Drosha – a ribonuclease (RNase) III enzyme that processes pri-miRNAs and shRNAs in the nucleus. Dicer – an RNase III enzyme vital for the siRNA and miRNA pathways, generating small double-­stranded RNA molecules suitable to be loaded into Argonaute protein. RISC – a ribonucleoprotein complex, consisting of an Argonaute protein bound to a guide strand from a siRNA or miRNA molecule, that fragments the target mRNA. Argonaute – an essential protein for RISC assembly, functioning in the recognition of the guide strand, target cleavage and recruitment of other important proteins involved in the silencing process. dsRBP – a protein important for the Dicer processing of dsRNAs and subsequent passage to the RISC (e.g., TAR RNA-binding protein, TRBP). Guide strand – a single-stranded RNA molecule resulting from the separation of the siRNA or miRNA, that has complementarity the target mRNA, providing specificity to the RNAi process. Seed sequence – a region of 2–8 nucleotides in the guide strand whose complementarity to the target mRNA is critical to the silencing success. 7.2.8  ene Silencing as Therapy for G Machado-Joseph Disease/ Spinocerebellar Ataxia Type 3 Machado-Joseph disease is a fatal dominant inherited neurodegenerative disease, caused by an abnormal CAG expansion in the coding region of the ATXN3 gene. Along with the other eight polyglutamine diseases, it constitutes a larger group of monogenic inherited neurodegenerative