Inquiry Basket
Synthesis of siRNA and ASOs can be challenging because they require sequence specificity to target a gene of interest for knockdown. Also, chemical modifications are needed for both siRNA and ASOs in order to be resistant to metabolism. High purity is also required for further development of these oligonucleotides for therapeutic use. Selecting the sequence of a siRNA or ASO, while also considering the various backbone and sugar modifications necessary for activity, potency, selectivity and pharmacokinetics as well as meeting elevated purity requirements poses numerous challenges that lead to a much more complex manufacture process than regular DNA oligos.
The functional mechanisms of four small nucleic acids.1,5
siRNA and ASOs share many properties. siRNA and ASOs are similar in that they are chemically synthesized nucleotide analogues which induce sequence specific silencing of their target gene. They differ because siRNA leads to degradation of target mRNA while antisense oligonucleotides lead to the sterical blocking of translation or degradation of target mRNA. Both siRNAs and ASOs lead to degradation of target mRNA by binding complementary to target mRNA and initiating RNase H catalyzed degradation. For this reason siRNA and antisense oligonucleotides can both be used to silence targets that are deemed "undruggable." Therapeutically siRNA and antisense oligonucleotides can silence non-coding RNAs and dominant negative mutations with high sequence specificity that small molecules cannot achieve. Because of their ability to silence these targets oligonucleotides can be used to treat rare disease and cancer. There are however many challenges associated with delivery.
| Feature | siRNA | ASOs |
| Structure | Double-stranded RNA (21-23 nt) | Single-stranded DNA or RNA (15-25 nt) |
| Mechanism | RISC-mediated mRNA cleavage | Ribonuclease H recruitment or steric blocking |
| Chemical modifications | Extensive (2'-O-methyl, LNA, PS) | Extensive (PS backbone, 2'-MOE, LNA) |
| Cellular delivery | Challenging; requires formulation | Moderate; gapmer designs improve uptake |
| Off-target effects | Sequence-dependent; immunogenicity | Sequence-dependent; protein binding effects |
Table 1 Comparison of siRNA and Antisense Oligonucleotide Characteristics
Synthetic siRNAs can also induce RNAi. The duplex becomes part of RISC, the passenger strand is cleaved and the guide strand pairs with target mRNA. Ago proteins cleave the target mRNA in a sequence-specific manner resulting in RNAi. The process acts enzymatically: one round of RNAi can destroy multiple target mRNA molecules. ASOs are generally single stranded and complementary to stretches of targeted mRNA. Hybridization of the ASO to its target creates a RNA/DNA or RNA/RNA hybrid. Recruitment of RNase H degrades the RNA of the hybrid, resulting in downregulation of the targeted protein.
siRNAs and ASOs are more complicated to synthesize than typical oligonucleotides because they require chemical modifications for stability, and have more complex structures. siRNAs are double stranded, so two separate strands need to be chemically synthesized and annealed together with accurate base pairing. The product needs to be purified to achieve homogenous dsRNA with defined overhangs. This purification step essentially doubles the synthetic workload for siRNAs over typical oligonucleotides. Both siRNA and ASOs require chemical backbone modifications (such as phosphorothioates) to increase stability within the cell. Backbone modifications further complicate synthesis because they often exist as mixtures of diastereomers with differing biological activity. Finally, siRNAs and ASOs require extensive purification to ensure that lack of failure products and batch-to-batch consistency for medical use. The types of impurities found in therapeutic oligonucleotides necessitate characterization with analytical methods.
One hurdle to siRNA and ASO drug synthesis difficulty and efficiency is their design. The ideal design for siRNA and ASOs would show potency to their target with minimal off-target effects. Sequence- and modification-based parameters that determine siRNA design can include length, symmetry, and type and placement of modification. Additionally, siRNA and ASOs should be stable in vivo and capable of entering the cell, whether by itself or through some sort of delivery via a carrier vehicle. The siRNA and ASOs used as therapeutics are structurally different from conventional laboratory-made oligonucleotides. Therapeutic siRNA and antisense oligonucleotides require certain structural qualities that allow them to maintain structural integrity so that they do not degrade as easily and can enter the target cell. Chemical stability and delivery method are key aspects to consider when designing siRNA and antisense oligonucleotides to be highly efficient.
siRNAs and ASOs can be anywhere from 15nt up to greater than 25nt in length. These variations can impact not only the ease of chemical synthesis, but also cellular function. The majority of siRNAs are chemically synthesized as 21–23nt duplexes with two nt 3'overhangs. This translates to two complementary strands that are separately synthesized and annealed to make a duplex. Both strands of the duplex must be fully complementary to each other. If either strand is in excess or if they do not anneal properly there will be fewer siRNAs capable of assembling into functional silencing complexes. ASOs can be anywhere from 15-25nt's in length. Shorter molecules have less off target effects, however they also have lower binding affinities to their target and often require chemical modifications that increase their binding affinity. Longer molecules can repress more effectively, but can sometimes form secondary structures that inhibit target binding. Additionally, siRNA duplexes require symmetric design. Each strand of the duplex has opposing thermodynamic stability at the 5' end. The strand with the less stable 5' end will become the guide strand of the siRNA and directs silencing to the target transcript. If the passenger strand has lower stability, silencing can be directed to off target transcripts.
One drawback to oligonucleotide drugs is off-target effects, which can occur via hybridization-dependent and -independent pathways. Hybridization-dependent off-target effects can occur when the therapeutic oligonucleotide sequence is partially complementary to another transcript. Examples include siRNAs triggering microRNA-like functions when the three-prime untranslated region (UTR) of a mRNA is partially complementary to the seed region of an siRNA. Seed-dependent off-target gene regulation has been shown to impact hundreds of transcripts with only partial sequence homology to the intended target. Both microRNA-like activity and ASOs can trigger off-target effects through partial binding to other transcripts. Interestingly, ASOs with longer lengths have been shown to have higher affinity for their intended target as well as higher tolerance for mismatches that could lead to off-target activity. Selection against sequences with high homology to other genes in the transcriptome through bioinformatic approaches can help reduce these effects. Site-specific chemical modification can also be utilized to induce higher specificity for mismatches. Ultimately, determining the extent of off-target activity through transcriptome-wide expression analysis is required before in vivo testing.
Synthetic chemistry of drug siRNAs and antisense oligos is more complex compared to regular DNA oligos. Drug siRNAs and antisense oligos require ribonucleotide based chemistries, which have additional challenges associated with protecting the two prime hydroxyl that will react without modification. Backbone modifications, sugar modifications, and conjugations to therapeutic oligos ends add additional synthetic complexity as well as stereochemical complexity. Modifications must be installed at the correct location to ensure pharmacokinetic (PK) and binding properties are optimized without affecting biological activity (knockdown of target mRNA via RNAi or RNase H activation, for example). Due to this complexity crude reaction mixtures can contain diastereomers of the product along with modified byproducts that may be difficult to purify and analyze via traditional methods.
Chemically, RNA synthesis differs from DNA synthesis in several ways. The additional two-prime hydroxyl group is reactive and must be protected during the phosphitylation cycle in order to prevent unwanted migrations of phosphates along the growing strand (resulting in strand degradation). It is most commonly protected using sterically-demanding silyl groups (such as TBDMS or TIPS) requiring fluoride ion for removal, unlike DNA nucleotides which are unprotected at this position. Thus, protecting RNA during synthesis requires additional steps and potential side reactions compared to DNA synthesis. In addition, the protected ribonucleoside phosphoramidites behave differently (primarily in kinetics of coupling and solubility) from their deoxy equivalents and optimized reaction conditions are required. Coupling times may also need to be extended. The phosphodiester linkage in RNA is also more prone to alkaline hydrolysis during synthesis and workup procedures. All these factors lead to lower overall synthetic efficiencies and yield than DNA oligomers of similar size, as well as more complex mixtures of side-products during synthesis that must be separated during purification.
Therapeutic siRNAs and ASOs contain substantial chemical modifications to allow for proper pharmacokinetics in vivo. Backbone modifications like phosphorothioate linkages, in which the non-bridging oxygen is substituted with sulfur, can confer nuclease resistance while also increasing protein binding affinity; however, they also create stereochemical variation as every modified phosphate becomes a mixture of diastereomers. Sugar modifications at the 2'-position, such as methoxy, fluoro, or methoxyethyl groups, can increase binding affinity to complementary RNA while decreasing immune system recognition. The placement of these modifications matters; heavy modification of the siRNA seed region at the two-prime position can disrupt target recognition, while selective modification of the passenger strand can inhibit off-target loading of the siRNA into RISC. Modification strategies for antisense oligonucleotides commonly use gapmer design: an unmodified region of DNA 'sandwiched' between modified 'wings.' These heterogeneous molecules pose issues with synthetic replication and QC.
Overcoming challenges in the development of RNA therapeutics.2,5
Stereochemistry presents another problem when using phosphorothioates. Each phosphorothioate modification center is a new stereocenter. Diastereomeric phosphorus centers may give rise to biologically active stereoisomers. All PS modification sites are racemic during conventional syntheses, leading to mixtures that have varying properties (e.g., chromatographic behavior) and may have different levels of biological activity, binding affinity, nuclease resistance, and toxicity. Evaluation of the ratios of diastereomers may be necessary and may require ion-pairing reversed-phase chromatography or chiral column chromatography to determine the stereochemical purity of a PS oligonucleotide sample. Evaluation of biological activity should also consider whether there are differences between the diastereomers. To properly characterize these products as quality attributes becomes more demanding as sequence length and degree of modification increase.
| Modification Type | Chemical Structure | Primary Function |
| Phosphorothioate | Phosphate with sulfur substitution | Nuclease resistance, protein binding |
| 2'-O-Methyl | Methyl group at 2' position | Enhanced binding, reduced immunogenicity |
| 2'-Fluoro | Fluorine at 2' position | Increased stability, stronger binding |
| 2'-MOE | Methoxyethyl group at 2' position | Extended half-life, improved affinity |
| Locked nucleic acid | Methylene bridge between 2'-O and 4'-C | Exceptional binding affinity |
Table 2 Chemical Modifications in siRNA and ASO Synthesis
Large scale solid phase synthesis of therapeutic siRNA strands and antisense oligonucleotides is often bottlenecked due to the sequential nature of phosphoramidite synthesis. In order to achieve strand elongation, each monomer residue is added sequentially to a bound-growing chain on an insoluble surface. With each synthetic cycle, there is inherent process loss that limits the overall achievable yield, complicates purification due to increasing sequence-based impurity complexity, and creates challenges in assembling ds siRNA therapeutics. Additionally, introduction of chemically modified nucleotides such as phosphorothioates, 2'-substituted nucleotides and terminal modifications introduce steric and stereochemical constraints that slow coupling kinetics relative to unmodified DNA.
Solid-phase oligonucleotide synthesis reactions are cumulative in that each step decreases the concentration of growing full-length product. Therefore, the percent yield of full-length product synthesized is highly dependent on coupling efficiency. Standard nucleoside phosphoramidites used in solid-phase synthesis have coupling efficiencies greater than 99%. Despite this high coupling efficiency, many truncated products are formed as the oligonucleotide chain length increases beyond ~20nt. Synthetic therapeutic siRNAs and ASOs frequently use chemically modified nucleotides which can decrease coupling efficiency due to steric bulk from protecting groups or the modified-amidite itself, faster decomposition of the activated phosphoramidite and decreased solubility. Chemically speaking, this means that if you have an oligonucleotide synthesis that is 98% efficient per coupling step, your percent yield of full-length product will only be 60% when the synthesis is complete. The other 40% is considered failed sequences and has to be separated during the purification process. Yield is dependent on reaction condition, quality of reagents and scale of synthesis so these parameters must be optimized for cost-effective production of therapeutically relevant oligonucleotides.
Chemical synthesis of both strands separately and subsequent annealing to form a siRNA duplex adds complexity on top of the chemical synthesis considerations for a single-stranded oligonucleotide. Equal amounts of perfectly matched guide and passenger strands with correct sequence and modification pattern must be combined. Any mismatch or excess of strand will reduce the amount of functional siRNA. Modified nucleotides, especially at the ends of the strand can significantly affect hybridization rates and stability of the strands. Reaction conditions for annealing like temperature profiles, salt concentrations, and strand ratios may need to be optimized for each siRNA. Incomplete annealing can lead to formation of inactive heteroduplexes that have different base pairing. Excess of starting materials can lead to activation of the immune system or off-target effects. Modified RNA strands can have different properties from unmodified DNA or RNA strands. Some modified strands can be more hydrophobic, such as when using two prime modifications. If strands are too hydrophobic, they can aggregate during annealing or product storage.
One of the significant hurdles to overcome in developing therapeutic siRNAs and antisense oligonucleotides is purification. They are synthesized chemically through solid-phase synthesis. This results in a crude product mixture that contains truncated failures, incomplete deprotectants, and stereoisomers that need to be separated out. Additional purification is necessary to meet clinical-grade purity. Another difference between regular oligos and therapeutic RNAs or ASOs is chemical modifications. Therapeutic RNAs/ASOs usually have modified backbones (often phosphorothioate backbones which create diastereomers) and modified sugars. In addition to having to remove unwanted byproducts of synthesis, these chemical modifications change how the therapeutic oligos will behave during purification. Lastly, siRNAs need to be annealed after synthesis. The purification process needs to not only remove all undesired byproducts but also leave behind a product that is sufficiently active and homogenous from batch to batch.
Removal of full-length product from shorter chain failure sequences is a major purification issue since failure in solid-phase synthesis steps is cumulative. Separation of n-minus-one species from full-length product can be especially difficult because chromatographic properties are very similar. Ion-pair reversed-phase chromatography using lipophilic ion-pairing reagents resolves oligonucleotides on the basis of hydrophobicity with longer chains being more retained. Anion-exchange chromatography is another useful technique which resolves species on the basis of charge differences. The full-length product will have one more negatively charged phosphate group than any shorter failure sequence allowing separation. Separation becomes more difficult if phosphorothioate linkages are included because PS oligonucleotides are more hydrophobic and peak shapes are broader due to the diastereomeric character of PS linkages. Purities greater than what can be achieved with single column chromatography may be required for therapeutic oligonucleotides. To reach the low level of impurities required by regulatory agencies, combinations of chromatography using different separation principles may be necessary.
siRNAs are typically purified from their single-stranded precursors. Thus, purification must address both single-stranded and double-stranded species. Purification of the two strands separately prior to annealing ensures stoichiometric mixing and full annealing; however, two chromatography runs are needed and incomplete annealing may result if there is a significant difference in purity between strands. Alternatively, crude single strands can be annealed and purified on-column as a duplex. Double-stranded RNAs behave differently than single strands during chromatography. By purifying siRNAs in duplex form, better resolution of full-length siRNA from truncated species and excess strand can be achieved. Chromatography must be performed under non-denaturing conditions (low temperature, proper salt concentration) to keep the strands annealed. Performing chromatography under denaturing conditions will allow determination of strand purity because the siRNA will be dissociated into two separate oligonucleotides. The purified therapeutic should contain minimal amounts of single-stranded siRNA because excess strand has the potential to activate the innate immune response or cause off target gene silencing.
Therapeutic siRNAs and antisense oligonucleotides must undergo extensive quality control testing. Unlike materials made for research purposes which can be verified using basic analytical techniques like gel chromatography or electrophoresis, molecules destined for therapeutic use require full characterization and qualification using orthogonal methods. Determining identity involves confirming that the produced product matches the intended design. Purity is determined by measuring the amount of full-length product versus impurities created during manufacturing. Strand integrity of siRNA molecules means that the two strands are correctly hybridized to each other and haven't degraded.
Quality control of therapeutic oligonucleotides involves confirming sequence identity, determining chemical purity, and validating structural integrity (duplex integrity for duplex therapeutic oligonucleotides). Sequence identity is established by mass spectrometry to confirm the measured mass agrees with the predicted mass of the designed sequence, though sequencing can also be performed where appropriate. Chemical purity is most commonly assessed by high-performance liquid chromatography (HPLC) to determine the amount of full-length product compared to truncated products, deletion products, and synthetic impurities. Final product chemical purity is typically >90% for therapeutic oligonucleotides. Strand integrity of siRNA duplexes is checked to confirm complete annealing between the sense and antisense oligonucleotide strands and to ensure no excess single-stranded siRNA is present, which can be done by native gel electrophoresis or by non-denaturing chromatography.
In addition to physicochemical characterization data, functional validation data demonstrate that the purified oligonucleotide behaves in cell-based assays and biochemical assays as expected. Evaluation of potency demonstrates that the siRNA or ASO reduces the levels of its intended target in a dose-dependent manner in a relevant cell type and helps confirm mechanism of action by use of appropriate controls. Evaluations of specificity generally focus on off-target effects and are designed to detect unintended modulation of genes other than the intended target, which could impact the clinical safety profile. Stability studies demonstrate that the product maintains potency during storage under recommended conditions for its proposed shelf life and may include testing under conditions beyond those recommended to accelerate stability-related degradation. Stability, potency, and mechanism of action studies are typically conducted with analytical methods that have been qualified for use in regulatory submissions and have documented precision and accuracy. Functional validation assures that the manufactured drug substance or drug product has the biological activity expected of the product.
While siRNA and antisense oligonucleotide synthesis for research purposes is relatively simple to perform on a small scale, producing these molecules for therapeutic use requires large-scale synthesis and delivery, while meeting good manufacturing practice guidelines, so that every batch is safe and of high quality suitable for use in humans. Manufacturing technologies must also be validated, equipment qualified and quality management systems established. Facilities must be able to not only produce consistent siRNAs and antisense oligonucleotides at larger scales to meet clinical demand but also pass inspections by regulatory agencies who review facility submissions. Successfully manufacturing chemically modified oligonucleotides that maintain potency can be challenging. Manufacturing facilities need specialized clean room space, analytical equipment, and trained personnel to perform well-documented standardized operating procedures.
Batch-to-batch variability is tightly controlled during production of oligonucleotide drugs. Identity, purity, and potency of each batch should meet established specifications. Processes used to synthesize therapeutic oligonucleotides are well characterized and validated. Reaction coupling efficiencies, reagent concentrations and reaction times are monitored and controlled during oligonucleotide synthesis. Installation qualification, Operational qualification, and Performance qualification (IQ/OQ/PQ) of synthesis platforms, purification systems, and analytical equipment guarantees these tools perform consistently for multiple batches. Analytical assays are in place to ensure batch integrity at critical steps during oligonucleotide synthesis. If there are any deviations from the expected results these can be caught before product quality is affected. All manufacturing parameters, involved personnel, and processing conditions are recorded in batch records allowing for full traceability of each manufactured lot. Analytical methods used to determine identity, purity and potency are validated to ensure they are reliable over the expected range of sample attributes. Specifications including full-length content, impurities, and potency are established with appropriate acceptance criteria. Finally, demonstration of batch-to-batch reproducibility using multiple consecutive batches ensures consistent drug dosing for patients.
Production processes for therapeutic oligonucleotides are subject to regulation by authorities such as the FDA or EMA. Guidance documents provided by regulatory agencies recommend that a complete chemistry, manufacturing and controls (CMC) package be prepared, which demonstrates the quality and reproducibility of the drug substance. Information provided on synthetic methods, downstream processing, analytical methods and acceptance criteria should be thoroughly described with suitable justification (development data, risk assessment, etc.). Overall compliance with current good manufacturing practices (GMP) is needed to ensure the product is manufactured in suitably designed and controlled facilities by qualified personnel using properly maintained equipment and following written procedures. Quality systems are required to ensure control of changes, evaluation of deviations and investigating root causes as well as implementing corrective and preventive actions (CAPAs). Documentation should allow for traceability of all steps in the production process. This includes receipt of raw materials, packaging materials, and finished product shipping as well as detailed batch records describing the entire process. These batch records include all relevant process parameters, in-process controls, and environment monitoring. Stability indicates expiration dating and should be supported by accelerated and real-time studies storing the product under different conditions. Stability should be verified annually for the commercial life of the product. Regulatory inspections ensure that these practices are being adhered to. Any observations noted during an inspection are required to be addressed and corrected.
| GMP Element | Key Requirements | Implementation Strategy |
| Process validation | Demonstrated consistency across batches | Protocol-driven process performance qualification |
| Analytical validation | Qualified methods for all quality attributes | Systematic method development and validation studies |
| Quality systems | Document control, change control, deviation management | Electronic quality management systems |
| Supply chain | Qualified suppliers, raw material testing | Supplier audits, incoming material control |
| Manufacturing flexibility | Multi-product capabilities, capacity management | Campaign planning, validated cleaning procedures |
Table 3 GMP Considerations for Therapeutic Oligonucleotide Manufacturing
siRNA and antisense oligonucleotides (ASOs) present significantly greater synthetic complexity than standard DNA primers. RNA chemistry sensitivity, dense backbone modifications, duplex integrity, and stringent purity requirements all demand optimized manufacturing control. Our capabilities in siRNA and antisense oligonucleotide synthesis include:
By integrating controlled synthesis chemistry, purification expertise, and analytical validation, siRNA and antisense oligonucleotides can be produced to meet demanding research and therapeutic development requirements.
Reduced gene silencing efficiency, unexpected off-target activity, or inconsistent experimental outcomes may be linked to sequence design, modification density, or oligonucleotide quality limitations. You may benefit from a technical review of your synthesis strategy if your program involves:
Aligning molecular design with synthesis feasibility and purification requirements early in development can help reduce downstream risk and timeline delays. Contact our scientific team to discuss your siRNA or antisense oligonucleotide requirements and identify the most appropriate synthesis and quality strategy for your application.
References
