Synthesis and Characterization of Thiophosphoramidate Morpholino Oligonucleotides and Chimeras
Introduction
In the current era of personalized medicine, oligonucleotide therapies are increasingly recognized as a dependable platform for drug development. Due to their molecular size, which typically ranges between 6 and 10 kilodaltons, and their inherent electrical charge, oligonucleotides (ONs) exhibit significantly different pharmacodynamic and pharmacokinetic properties when compared to conventional small molecule drugs. While small molecule drugs exert their effects by directly interacting with disease-related proteins, ON drugs function by targeting the messenger RNAs (mRNAs) that serve as the blueprints for these proteins. This mechanism allows ON drugs to intervene in the disease process at a much earlier stage, effectively preventing pathogenesis.
Synthetic ON drugs possess the remarkable ability to interact with their specific mRNA targets through the well-established Watson-Crick base pairing rules. This interaction can lead to two primary outcomes: either it can physically block the RNA’s function, as seen in antisense-mediated exon-skipping and microRNA inhibition, or it can trigger the irreversible degradation of the target RNA, such as through RNase H1 or Argonaute 2 (Ago2) mediated cleavage. This degradation effectively halts the transfer of genetic information from the mutated gene to the production of the defective protein. Their unique mechanisms of action enable ON drugs to address therapeutic targets considered “undruggable” by traditional approaches. This category includes several rare genetic disorders for which no effective therapeutic modalities currently exist.
Depending on their structural design and the specific way in which they modulate protein expression, ON drugs are classified into various categories, including antisense oligonucleotides, small interfering RNAs (siRNAs), antagomirs, splice-switching oligonucleotides, and aptamers. Over the past four decades, extensive research has led to the development of numerous chemical modifications aimed at fine-tuning various biochemical properties of ON drugs. These modifications are designed to enhance binding affinity to their targets, improve resistance to nuclease enzymes that degrade nucleic acids, optimize in vitro and in vivo delivery to the site of action, and minimize potential immune stimulation and associated toxicities.
Among these modified oligonucleotides, a distinct class of synthetic ONs known as phosphorodiamidate morpholinos (PMOs) has garnered considerable attention. Two PMO drugs that can effectively target the mRNA of the human dystrophin gene, a challenging target, have received conditional approval for the treatment of Duchenne Muscular Dystrophy. Furthermore, two additional PMO-based therapies are currently undergoing Phase III clinical trials. Several oligonucleotide drugs incorporating the phosphorothioate (pS) linkage have also received approval from the Food and Drug Administration (FDA), including examples such as Vitravene®, Kynamro®, Spinraza®, Tegsedi™, and Givlaari™. Throughout their development process, the clinically relevant pharmacokinetic, biodistribution, metabolism, and toxicity profiles of both pS and PMO based ONs have been rigorously investigated and validated through extensive years of detailed testing and clinical trials. These studies have revealed that PMOs are characterized by rapid clearance from the bloodstream and exhibit minimal drug-related toxicity at clinically relevant doses. However, they also demonstrate limitations in cellular uptake and overall pharmacokinetics. In contrast, ONs containing pS internucleotide linkages can interact with a wide array of proteins, including plasma proteins, which can lead to off-target effects and associated toxicity. While pS-modified ONs possess superior enzymatic stability, their “stickiness” can result in reduced renal clearance, longer circulation half-lives, and good tissue distribution.
Despite the individual successes of PMO and pS-modified ONs, the chemical synthesis of oligonucleotide hybrids that incorporate both morpholino and DNA-pS subunits has been challenging due to inherent drawbacks associated with the phosphorus(V)-based synthetic strategy currently employed for generating PMOs. Furthermore, it is reasonable to hypothesize that second- and third-generation sugar phosphoramidates containing a P=S (thiophosphate) moiety could potentially improve their hydrolytic stability under acidic conditions and also enhance their resistance to degradation by intracellular nuclease enzymes. This rationale has led the researchers to focus entirely on oligonucleotides containing thiophosphoramidate morpholinos (TMOs) and TMO chimeras that incorporate DNA-pS linkages. In this article, the researchers evaluate the stability of TMO analogues when exposed to aqueous acid, their hybridization affinity towards complementary DNA and RNA sequences, as well as their ability to recruit the enzyme RNase H1. Additionally, they assess the biological potency of these TMO analogues as inhibitors of microRNA-15b using a dual-luciferase reporter assay.
Results and Discussion
The figure mentioned, Figure 1, provides chemical structures of several oligonucleotide analogues: phosphorodiamidate morpholino (PMO) (labeled 1a), phosphoramidate morpholino (labeled 1b), thiophosphoramidate morpholino (TMO) (labeled 1c), and TMO-DNA-pS chimeras (labeled 1d).
The text then discusses that modifications like 2′-O-methyl (2′-OMe), 2′-fluoro (2′-F), 2′-O-(2-methoxyethyl) (2′-O-MOE), and locked nucleic acid (LNA) are only available as phosphoramidite synthons, which are building blocks used in oligonucleotide synthesis. While various PMO hybrids have been synthesized using conventional phosphorus(V) chemistry, a reliable and adaptable synthetic method for creating chimeras that combine morpholino nucleotides with therapeutically relevant modifications has not yet been reported. The development of such chimeric analogues could be advantageous for creating novel therapeutic drugs with biological properties distinct from those of their original components.
Previous research towards this goal has focused on synthesizing morpholino-DNA chimeras incorporating different chemical linkages, including boranophosphoramidate, phosphoramidate, and alkyl phosphoramidate linkages. Other examples in the scientific literature include morpholino methylphosphonamidates and triazole-linked morpholino analogs. Zhang et al. reported the synthesis and characterization of phosphoramidate morpholino-DNA chimeras (10-12 nucleotides in length) containing up to three morpholino thymidine (mT) phosphoramidate linkages (as depicted in Figure 1b) and morpholino-RNA chimeras (21 nucleotides in length) with one or two morpholino uridine (mU) phosphoramidate linkages. Their findings indicated that the introduction of one or more morpholinophosphoramidate linkages resulted in a reduction in hybridization affinity compared to an unmodified DNA/RNA duplex. Notably, when siRNA duplexes containing one or two terminal morpholinophosphoramidate linkages (at the 3′- and 5′- ends) were treated with serum, their stability improved by 4 to 7-fold relative to unmodified oligonucleotides. Improved serum stability has also been observed with other phosphoramidate analogues. Of particular interest was the observation that siRNA duplexes containing morpholino phosphoramidate subunits exhibited activity at nanomolar concentrations in a dual luciferase reporter assay, a common method for studying gene expression.
From a chemical synthesis perspective, the solid-phase synthesis of morpholinophosphoramidate oligonucleotides resulted in low overall yields. Furthermore, the crude reaction mixtures could not be purified using the conventional dimethoxytrityl (DMT)-On/Off procedure, a standard purification technique in oligonucleotide synthesis, due to their inherent instability in the presence of aqueous acid. The researchers then hypothesized that replacing a non-bridging phosphorus-oxygen (P=O) linkage in morpholino…
Chemical Synthesis
The morpholino nucleosides of N6-benzoyl adenosine (mABz), N4-benzoyl cytidine (mCBz), N2-isobutyryl guanosine (mGiBu), and thymidine (mT) (designated as compounds 3a-d) were synthesized according to previously published literature protocols (as outlined in Scheme 1). In brief, 7.42 mmol of a 5′-dimethoxytrityl (DMT) protected ribonucleoside (specifically, 5.0 g of compound 2a, 4.87 g of 2b, 4.82 g of 2c, or 4.16 g of 2d) was dissolved in 500 mL of anhydrous methanol. Subsequently, 1.2 equivalents of sodium periodate (1.9 g, 8.9 mmol) and 1.2 equivalents of ammonium biborate tetrahydrate (2.35 g, 8.9 mmol) were sequentially added to the solution. The resulting mixture was stirred at 25 degrees Celsius for 6 hours or until no unreacted ribonucleoside could be detected using thin-layer chromatography (TLC). The reaction mixture was then filtered through a pad of Celite®, and to the filtrate, 2 equivalents of sodium cyanoborohydride (0.93 g, 14.8 mmol) and 2 equivalents of glacial acetic acid (0.85 mL, 14.8 mmol) were added. This mixture was stirred for 16 hours, filtered again, and the solvent was removed using a rotary evaporator. The crude reaction mixture was purified by column chromatography on a silica gel column that was pre-equilibrated with ethyl acetate (EtOAc) containing 3% triethylamine. The column was first eluted with 100% EtOAc, followed by a mixture of EtOAc and Methanol (70:30) to yield the 6′-DMT protected morpholino nucleosides (3a-d): mABz (2.75 g, 56% yield), mGiBu (2.21 g, 46% yield), mCBz (2.40 g, 51% yield), and mT (2.33 g, 58% yield).
Zhang et al. have previously described the chemical synthesis of morpholino phosphorodiamidites of mU and mT. Following slight modifications to these established protocols, the phosphorodiamidites of mABz, mGiBu, mCBz, and mT (designated as 4a-d, as shown in Scheme 1) were synthesized. Briefly, one equivalent of each morpholino nucleoside (ranging from 2.6 to 3.1 g, 4.8 mmol) and 1.2 equivalents of 2-cyanoethyl N,N,N’,N’-tetraisopropyl phosphorodiamidite (1.83 mL) were reacted with 0.5 equivalents of 0.25M 5-(Ethylthio)-1H-tetrazole (ETT, 9.6 mL) in dichloromethane for 30 minutes. After evaporating the solvent to dryness, the crude product was quickly purified using column chromatography. The pure phosphorodiamidites were isolated in moderate yields: 4a (2.75 g, 67% yield), 4b (2.37 g, 59% yield), 4c (2.63 g, 66% yield), and 4d (2.93 g, 82% yield). The detailed synthetic protocol and comprehensive characterization data, including 1H NMR, 31P NMR, 13C NMR, and ESI-mass spectrometry (ESI-MS) data for all synthesized morpholino phosphorodiamidites, are provided in the Supporting Information (SI, Section 2 and SI pages S45-S48).
Optimization of TMO Synthesis
Conventional nucleoside phosphoramidites, characterized by the general formula (R1O)(R2O)P[(N,N’-(iPr)2], typically undergo activation at a single P(III)-N linkage during the coupling step. In contrast, phosphorodiamidite morpholino synthons (4a-d) possess two P(III)-N linkages, both of which can be activated to varying extents depending on the specific reaction conditions employed.
Barone and colleagues demonstrated that when a phosphorodiamidite containing two amino substituents, specifically P-Methoxy-(N,N’-diisopropylamino,N-morpholino) phosphorodiamidite, was reacted with 1H-tetrazole, the primary activation pathway involved the N,N’-diisopropylamino group. This major pathway led to the formation of a morpholino-substituted P-Methoxy phosphoramidite as the predominant product, accounting for 95% of the reaction outcome, with only 5% resulting in the morpholino-activated product.
Since the initial step in phosphoramidite activation involves protonation of the amine leaving group and the weakening of the P-N bond, the more basic diisopropylamino group, with a reported pKa of 11.07 in water, exhibits a greater tendency for protonation compared to the morpholino ring nitrogen, which has a pKa of 8.49 in water. This difference in basicity suggests that the diisopropylamino group is a superior leaving group in this context. Consistent with this, prior research comparing the activation of 5’-DMT-2’-deoxythymidine phosphoramidites containing either an N-morpholino or an N,N’-diisopropylamino amine leaving group revealed a 95% activation of the diisopropylamino group within one minute under tetrazole-mediated activation, while only 5% activation of the morpholino group was observed under identical conditions. This observation further implies that the morpholino ring nitrogen might be less susceptible to activation compared to the N,N’-diisopropylamino group.
Previous investigations have indicated that the selectivity of activator interaction, specifically the preference for activating the morpholino ring nitrogen versus the N,N’-diisopropylamino group, is influenced by the choice of activator. When 0.5 equivalents of ammonium tetrazolide salt were used to activate P-Methoxy-(N,N’-diisopropylamino,N-morpholino) phosphorodiamidite, a 10% formation of the morpholino-activated product was observed, contrasting with the 5% observed with 1H-tetrazole. This suggests a potential for competing activation pathways depending on the activator. Additionally, the use of Activator 42, identified as 5-[3,5-bis(trifluoromethyl)phenyl]-1H-tetrazole with a pKa of 3.4, resulted in the lowest yields of the desired product. This outcome could be attributed to competing morpholino activation, potentially leading to premature self-capping of the growing TMO chain as a 5’-N,N’-diisopropyl phosphoramidate. Furthermore, the high acidity of Activator 42 might have contributed to stepwise double activation of the incoming activated phosphorodiamidite or degradation of the TMO backbone during the extended coupling time of 600 seconds.
In line with these findings, initial attempts to synthesize TMOs using 0.45 M 1H-tetrazole or 0.25 M ETT also resulted in low yields and a significant amount of failure products, indicating poor overall efficiency and the formation of substantial side products. Similarly, low yields were observed when a 0.12 M concentration of 1H-tetrazole (TET, pKa = 4.9) was employed. The presence of a considerable quantity of short failure products suggests a lower stepwise coupling efficiency associated with TET activation.
Given that 5-Ethylthio-1H-Tetrazole (ETT, pKa = 4.3) is more acidic than 1H-tetrazole, subsequent experiments explored the use of 0.12 M ETT buffered with 0.01 M DMAP and unbuffered 0.12 M ETT. Surprisingly, the highest yield of the full-length product was achieved with the unbuffered 0.12 M ETT, yielding 85.9%. The less acidic and more nucleophilic activator 4,5-dicyanoimidazole (DCI, pKa = 5.2) also produced TMO1 in relatively high yields (82.3% at 0.12 M concentration). Attempts to enhance nucleophilicity by using DMAP-buffered ETT resulted in lower yields of the expected product compared to unbuffered ETT. These observations suggest that factors beyond activator pKa, such as nucleophilicity, coupling time, and activator concentration, play a crucial role in determining stepwise yields during TMO synthesis. Based on these results, 0.12 M ETT was selected as the preferred activator for solid-phase synthesis.
To further investigate the impact of activator concentration on product selectivity, the solution-phase reactivity of 6’-DMT-morpholinothymidine phosphorodiamidite (4d) with 3’-O-acetyl thymidine was monitored using 31P NMR spectroscopy. While lower ETT concentrations (0.25 equivalents) led to slower reactions and a higher degree of morpholino activation, the use of 1.0 equivalent or more of ETT resulted in stepwise activation of both amino substituents, yielding significant amounts of symmetrical dithymidine phosphotriester.
In summary, the solid-phase synthesis of TMOs described herein closely parallels conventional phosphorothioate DNA synthesis with two key modifications: the ETT concentration was reduced to 0.12 M, deviating from the manufacturer’s recommended 0.25 M, and a 600-second coupling time was implemented. For the synthesis of TMO-DNA-pS chimeras, a 30-second coupling time was employed for commercially available 2’-deoxyribonucleotide 3’-phosphoramidites.
Optimization of DMT-On/Off Purification
Prior research into the kinetics and mechanisms of 3’,5’-thiophosphoramidate hydrolysis across a wide pH range, specifically from pH 1 to 9, has indicated that protonation of the 3’-nitrogen atom results in the formation of an N(H)-P(V) linkage. This specific linkage has been shown to be susceptible to degradation under acidic conditions, particularly below pH 5. Given that TMO subunits contain a P(V)-(N-morpholino) linkage, we conducted an evaluation of the hydrolytic stability of TMO1 under acidic pH conditions.
In this evaluation, TMO1 was subjected to treatment with varying concentrations of aqueous acetic acid, and the rate at which it degraded over time was monitored using liquid chromatography-mass spectrometry, or LCMS. Under the conditions typically employed for conventional DMT-On/Off purification, which involves 80% aqueous acetic acid, TMO1 underwent complete degradation within a 10-minute timeframe at a temperature of 25 degrees Celsius. Consequently, to mitigate this rapid degradation, the concentration of the acetic acid was reduced to 50%. The rate of TMO1 degradation over time under this modified condition was subsequently monitored. Our findings indicated that a 5-minute treatment with 50% aqueous acetic acid, followed by an immediate quenching step using triethylamine, resulted in greater than 99% detritylation of TMO1 while minimizing degradation. This observation led us to adopt this specific protocol as our preferred method for DMT deprotection.
Employing this optimized deprotection protocol, the final DMT-Off oligonucleotides achieved an average purity exceeding 85%, as determined by LCMS analysis. Following this established procedure, several TMO mixmer oligonucleotides and chimeric TMO-DNA-pS oligonucleotides were synthesized, and their characterization data obtained via LCMS is compiled. Where applicable, the percentage yields of the full-length DMT-On product relative to any observed failure products were determined by integrating the ultraviolet profiles obtained from the LCMS data.
To further validate the robustness of the solid-phase synthesis procedure for TMOs, as outlined in the relevant scheme, the spectra of crude, unpurified reaction mixtures for TMOs 14-16 were analyzed. Integration of the ultraviolet profiles of the mixmer oligonucleotides revealed that the 6’-DMT On product was obtained in yields ranging from 30 to 45%. We observed that in addition to the percentage of guanine-cytosine content and the overall length of the oligonucleotide, the purity and moisture content of the morpholino phosphorodiamidites played a critical role in achieving satisfactory synthesis yields. Consequently, the average stepwise yields during the coupling reactions ranged from 95 to 97%, as monitored by trityl analysis on an ABI-394 synthesizer. Following DMT-On/Off purification, the TMO oligonucleotides were obtained in moderate yields, ranging from 47 to 276 nanomoles. Comparable yields were also achieved for the TMO-DNA-pS chimeras.
For the purpose of yield comparison, a 14-mer DNA-pS oligonucleotide (ON12) and an unmodified DNA oligonucleotide (ON13) produced yields exceeding 800 nanomoles of pure product when synthesized under the newly optimized TMO synthesis conditions. TMOs 8, 10, and 11, which contain a 3’-morpholino moiety, were synthesized using Universal Support III. In these specific cases, the initial morpholino phosphorodiamidite coupling step was followed by an oxidation reaction instead of sulfurization. All subsequent synthesis steps remained consistent with those used for other solid supports. The cleavage of oligonucleotides from the Universal Support III was performed using a 2M solution of ammonia in methanol for a duration of 1 hour, followed by a deprotection step using 28% aqueous ammonia at 55 degrees Celsius for 16 hours.
Enzymatic Stability
While phosphorothioate oligonucleotides exhibit enhanced resistance to enzymatic degradation compared to unmodified DNA, it is established that non-ionic phosphorodiamidate morpholino oligomers, or PMOs, are not susceptible to degradation by any known enzyme. Given that TMOs are hybrid molecules incorporating both morpholino nucleosides and phosphorothioate linkages, we proposed that they might demonstrate superior enzymatic stability against nucleases when compared to unmodified DNA.
To investigate this hypothesis, the susceptibility of TMO8, which is exclusively modified with thiophosphoramidate linkages, and a DNA-phosphorothioate control oligonucleotide, ON12, to 3’-exonuclease degradation was evaluated using Snake Venom Phosphodiesterase I, commonly known as SVPDE. In these enzymatic hydrolysis experiments, TMO8 and ON12, both at a concentration of 13.3 micromolar, were incubated at 37 degrees Celsius in a reaction mixture containing 100 millimolar Tris-HCl buffer at pH 8.5, 14 millimolar magnesium chloride, 72 millimolar sodium chloride, and SVPDE enzyme at a concentration of 0.1 units per milliliter. Aliquots of 45 microliters were withdrawn from the reaction mixture at various time intervals, subjected to heat inactivation to halt enzymatic activity, and then stored on dry ice until analysis by analytical reverse-phase high-performance liquid chromatography, or RP-HPLC.
Based on the analytical RP-HPLC data, no discernible degradation or peak broadening was observed for TMO8 over a 23-hour period. The undegraded TMO8, which eluted with a retention time between 13 and 16 minutes, constituted the major product. While two minor products were detected at retention times of 11 and 11.8 minutes, these were tentatively identified as potential degradation products; however, the remaining minor peaks were attributed to the enzyme itself. In contrast, when ON12 was subjected to the same assay conditions, significant peak broadening was evident within 7 hours. Furthermore, a peak eluting rapidly at a retention time of 1.5 minutes, presumably corresponding to 5’-phosphorothioate-thymidine mononucleotide, appeared during this time frame.
It is possible that undegraded TMO8 and its potential enzymatic degradation products, such as those resulting from the removal of one or two terminal nucleosides, might co-elute on a reverse-phase C18 column during analytical RP-HPLC. To address this possibility, the fractions corresponding to the peaks observed at each time point during the SVPDE assay were collected, pooled, and subsequently reanalyzed using liquid chromatography-mass spectrometry, or LCMS. In the case of TMO8, the analytical HPLC fractions corresponding to the ultraviolet signal at a retention time of 13 to 16 minutes at various time points during the SVPDE assay, specifically at 1, 4, 8, and 23 hours, were reanalyzed by LCMS. By comparing these reanalysis profiles with that of untreated TMO8 and analyzing their corresponding mass-to-charge ratio data at time zero and 23 hours, we definitively confirmed that TMO8 did not undergo degradation within 23 hours under our SVPDE assay conditions. Consequently, we concluded that in comparison to DNA-phosphorothioate oligonucleotides, TMOs exhibit a high degree of stability towards Snake Venom Phosphodiesterase I.
The SVPDE assays for DNA-phosphorothioate oligonucleotide ON12 and unmodified phosphate DNA oligonucleotide ON13 were performed using the same protocol as described for TMO8. The fractions corresponding to the broad peak eluting between 14 and 16 minutes in the analytical RP-HPLC profiles were reanalyzed by LCMS. Although precise quantification of each degradation product was challenging due to partially overlapping signals, the LCMS analysis of ON12 at 7 hours post-SVPDE treatment indicated that essentially complete degradation had occurred. In comparison, unmodified DNA oligonucleotide ON13 was completely degraded within 15 minutes under the same experimental conditions.
We routinely utilize 1.0 micromolar succinyl controlled pore glass, or CPG, loaded with 2’-deoxynucleosides or 2’-O-methyl nucleosides for the synthesis of TMOs, as exemplified by all oligonucleotides listed in Table 1 with the exception of 8, 10, and 11. Therefore, the 3’-exonuclease susceptibility of TMO7, which contains a 2’-deoxythymidine moiety at its 3’-terminus, was also evaluated using the aforementioned method with the following conditions: 100 millimolar Tris buffer at pH 9, 14 millimolar magnesium chloride, 72 millimolar sodium chloride, 13.3 micromolar TMO7, and SVPDE enzyme at a concentration of 2 times 10 to the power of minus one units per milliliter at 37 degrees Celsius. Consistent with our previous findings, no degradation of TMO7 was observed over a 23-hour period, a result that was confirmed by LCMS reanalysis.
The binding affinity of an oligonucleotide is a critical factor influencing its therapeutic potential. Since the introduction of locked nucleic acids, or LNAs, several sugar modifications that conformationally restrict the nucleoside have been developed to enhance the binding affinity of oligonucleotides. Among unnatural backbone structures, PMOs and peptide nucleic acids, or PNAs, improve hybridization affinity towards RNA, whereas stereorandom oligonucleotides containing phosphorothioates and boranophosphates decrease binding affinity. While phosphoramidate oligonucleotides and chimeric morpholino phosphoramidate-DNA have shown reduced RNA binding affinity, N3’-P5’ phosphoramidates and their thiophosphoramidate variants, as well as triazole-linked morpholino oligonucleotides, have been demonstrated to form highly stable duplexes with RNA.
Encouraged by the enhanced RNA binding affinity observed with various morpholino and thiophosphoramidate analogs, we investigated the hybridization affinity of TMO and TMO-DNA-phosphorothioate chimeras 3 through 6 towards complementary DNA and RNA. Control duplexes with the same sequence, consisting of PMO, DNA-phosphorothioate, or canonical DNA, were also evaluated under identical conditions. Briefly, oligonucleotides 3 through 6 and their DNA or RNA complements, each at a concentration of 1.0 micromolar, were dissolved in a suitable buffer containing 50 millimolar Tris-HCl, 50 millimolar potassium chloride, and 1 millimolar magnesium chloride at pH 8.3. DNA and RNA duplexes of PMO, DNA-phosphorothioate, and canonical DNA were also prepared under the same conditions and used as controls. Each duplex, with a total volume of 1.0 milliliter, was placed in a 10-millimeter path length cuvette and positioned in a 6×6 Peltier thermostatted multi-cell holder. The samples were heated from 15 to 90 degrees Celsius at a rate of 3 degrees Celsius per minute, maintained at 90 degrees Celsius for 5 minutes, and then cooled to 10 degrees Celsius at a rate of 1 degree Celsius per minute before initiating thermal denaturation runs at a rate of 0.5 degrees Celsius per minute. Each experiment was repeated at least twice per duplex, resulting in four melting curves. Representative thermal denaturation runs for all duplexes are provided.
When compared to the canonical DNA/DNA duplex, the TMO3/DNA duplex exhibited a melting temperature depression of -12.6 degrees Celsius, which corresponds to -0.6 degrees Celsius per TMO modification. This loss of binding affinity is more pronounced than that observed for the DNA-phosphorothioate/DNA duplex, which showed a melting temperature depression of 8.2 degrees Celsius or -0.4 degrees Celsius per modification. In contrast, fluorescently labeled PMO/DNA displayed a melting temperature close to that of the native DNA/DNA duplex, with a loss of only 1.5 degrees Celsius, a reduction that could be attributed to the presence of a 3’-fluorescein label on the PMO. The TMO-DNA-phosphorothioate chimera 4, which is 32% TMO modified with a gapmer configuration consisting of 7 TMO linkages, exhibited a melting temperature depression of -6.6 degrees Celsius or -0.9 degrees Celsius per TMO modification. Surprisingly, TMOs 5 and 6, composed of alternating TMO and DNA-phosphorothioate linkages, showed an increase in melting temperature of approximately +6.5 degrees Celsius or +0.6 degrees Celsius per TMO modification when duplexed with complementary DNA. To summarize, the melting temperature values for oligonucleotide/DNA duplexes of oligonucleotides 3 through 6 followed the trend: TMO5/DNA ~ TMO6/DNA > DNA/DNA ~ Fluorescein-PMO/DNA > TMO4/DNA > DNA-phosphorothioate/DNA > TMO3/DNA.
In contrast to its DNA duplex, the TMO3/RNA duplex showed an increase in melting temperature of +10 degrees Celsius or +0.48 degrees Celsius per TMO linkage when compared to the unmodified DNA/RNA duplex. Similarly, the Fluorescein-PMO/RNA duplex also exhibited a melting temperature increase of +0.4 degrees Celsius per modification. TMO5 and TMO6, which contain alternating morpholino thiophosphoramidate and DNA-phosphorothioate linkages, showed a melting temperature increase greater than or equal to +9.6 degrees Celsius or +0.88 to 1 degree Celsius per TMO linkage, representing the highest melting temperature increase observed for any duplex discussed in this study.
The DNA-phosphorothioate/RNA control duplex showed a -0.4 degrees Celsius depression per DNA-phosphorothioate linkage, as anticipated, and the gapmer TMO4/RNA duplex showed a depression of -0.5 degrees Celsius per TMO linkage. This latter observation may be attributed to the design of the gapmer, which features short TMO ‘wings’ and a 14-mer DNA-phosphorothioate ‘gap’. Thus, the RNA duplexes of TMOs 3 through 6 followed the trend: TMO5/RNA ~ TMO3/RNA ~ TMO6/RNA > Fluorescein-PMO/RNA > DNA/RNA > TMO4/RNA > DNA-phosphorothioate/RNA.
To summarize the thermal denaturation studies, alternately modified TMO5 and TMO6 demonstrated an increase in binding affinity, ranging from +6 to +10 degrees Celsius, towards both complementary DNA and RNA. We hypothesize that the insertion of alternating 2’-deoxynucleotides after each TMO subunit might provide increased flexibility to an otherwise rigid construct, thereby facilitating more efficient binding with the complementary strand. A significant loss in melting temperature, greater than 12 degrees Celsius, was observed for the TMO3/DNA duplex. Similarly, Fluorescein-PMO also did not exhibit increased binding affinity towards complementary DNA, suggesting that the exclusive presence of rigid morpholino moieties results in a backbone conformation that is not conducive to efficient DNA binding. The DNA-phosphorothioate/DNA duplex also showed a reduced melting temperature of -8.2 degrees Celsius, as expected. Thus, it can be hypothesized that the TMO3 construct, being a complete hybrid of both rigid morpholino sugars and phosphorothioate internucleotide linkages, experiences a cumulative loss of DNA binding ability. In contrast, both TMO3 and Fluorescein-PMO showed superior RNA binding affinity, ranging from +8 to +10 degrees Celsius.
Apart from the overall percentage of modification, the specific position of TMO linkages within the oligonucleotide sequence appears to play a crucial role in determining its binding affinity. This becomes evident when comparing the melting temperature values of TMO5 and TMO6, both with 50% TMO modification, and the gapmer TMO4, which has 32% TMO modification. When complexed with complementary RNA or DNA, the binding affinity of the latter was approximately 14 degrees Celsius lower.
The conformation of canonical DNA/RNA heteroduplexes typically lies between the A and B form helical geometries. While both A and B form duplexes exhibit maxima and minima at similar wavelengths in their circular dichroism spectra, a greater ellipticity between 270 and 280 nanometers is characteristic of the A-form, whereas a B-form duplex shows approximately equal positive and negative bands above 220 nanometers with a crossover point at 261 nanometers. To gain further insights into the overall helical structure of our TMO-containing duplexes, RNA heteroduplexes of TMOs 3, 5, and 14 were analyzed using circular dichroism spectroscopy and compared with control duplexes of DNA-phosphorothioate/RNA and canonical DNA/RNA. Given that the relative position and intensities of characteristic negative bands at 210 nanometers and the positive band around 280 nanometers for DNA/RNA hybrids are highly sequence-dependent, TMOs 3 through 6 and a DNA-phosphorothioate control with an identical sequence were used in this experiment, with the exception of TMO14.
The circular dichroism spectra of RNA duplexes of TMOs 3, 5, and 14 display the characteristic features of an A-type helical environment, exhibiting negative bands of varying intensity around 210 nanometers and a positive band of strong intensity around 270 nanometers. The positive Cotton band and the negative band at 210 nanometers, which are indicative of an A-form helical structure, are most prominent in the TMO5/RNA heteroduplex and closely resemble a typical A-form RNA/RNA duplex. This observation further supports the hypothesis that the increased conformational flexibility achieved by inserting 2’-deoxynucleotide linkages between TMO subunits promotes a more pronounced A-form conformation in its duplexes, resulting in superior binding affinity for this specific design.
Crossover bands in the circular dichroism spectra were observed between 205 and 215 nanometers and between 227 and 233 nanometers for all duplexes examined, with the exception of the DNA-phosphorothioate/RNA duplex, which did not exhibit a crossover in these specific regions. The positive absorption band at 273 nanometers for the canonical DNA/RNA duplex, which had a crossover at 257 nanometers, was blue shifted to approximately 265 nanometers for all TMO duplexes as well as the DNA-phosphorothioate/RNA duplex, which showed crossover wavelengths between 250 and 255 nanometers. Apart from the TMO5/RNA duplex, the gapmer TMO4/RNA duplex and the DNA-phosphorothioate/RNA duplex also displayed an intense negative peak in the 200 to 213 nanometer range, a characteristic feature of an A-form duplex. In the case of TMO14, this negative peak closely resembled that of the canonical DNA/RNA control duplex, with both duplexes showing crossovers at 203 and 211 nanometers. However, a less intense negative peak was observed for the TMO3/RNA duplex, highlighting the influence of sequence on the circular dichroism spectra. In summary, the circular dichroism spectra of the newly synthesized TMO heteroduplexes reveal structural similarities to the DNA-phosphorothioate/RNA duplex. TMOs 3, 4, and 14 adopt helical conformations that are more akin to an A-form duplex rather than a B-form duplex. Interestingly, the uniformly TMO modified constructs, TMO3 and 14, formed RNA duplexes with notably different biophysical characteristics compared to the chimeric TMO5. Specifically, the TMO5/RNA duplex closely resembles an A-form RNA/RNA duplex, providing a plausible explanation for its higher binding affinity towards complementary DNA and RNA. Further experiments designed to evaluate the helical conformations adopted by various TMO/DNA duplexes are currently in progress.
Ribonuclease H1 enzymes constitute a large family of endonucleases that specifically catalyze the degradation of the RNA strand within an RNA/DNA hybrid duplex. These evolutionarily conserved enzymes are involved in numerous essential biological processes, including the removal of RNA primers from Okazaki fragments during DNA replication, homologous recombination, DNA repair, and RNA interference. Given that retroviral reverse transcriptases contain an RNase H domain crucial for the reverse transcription of viral complementary DNA, the RNase H domain of HIV-RT is being investigated as a potential therapeutic target for drug-resistant HIV-1 strains.
Antisense-based RNase H therapeutics garnered significant interest several decades ago when Zamecnik and colleagues demonstrated that the addition of short synthetic oligonucleotides, specifically 13-mers targeting the viral repeat sequence of Rous sarcoma virus, inhibited viral replication and oncogenic transformation of chicken fibroblast cells. They successfully demonstrated the ability of oligonucleotides to inhibit gene expression by binding to and subsequently promoting the degradation of the target RNA through the action of RNase H1. Since this groundbreaking work, a variety of chemically synthesized oligonucleotide analogs have been evaluated for their potential in RNase H1-mediated antisense therapies. This approach is attractive because the reaction proceeds catalytically, allowing the oligonucleotide drug to be released for multiple successive rounds of RNA-targeted inhibition. Examples of chemical modifications that confer RNase H1 activity include phosphorothioates, boranophosphates, mesylphosphoramidates, and sugar modifications such as arabinonucleic acid, cyclohexene nucleic acid, and 2’-fluoro arabinonucleic acid.
TMOs exhibit higher RNA binding affinity compared to canonical control duplexes, potentially enabling them to efficiently invade RNA secondary structures. They also possess superior enzymatic stability, making them ideal candidates for RNase H1-based therapeutics. To evaluate this aspect, TMO3 was preannealed with a 5’-fluorescein-labeled complementary RNA and subjected to digestion with Escherichia coli RNase H1. The results were analyzed by polyacrylamide gel electrophoresis, with a detailed experimental protocol provided, and visualized using a transilluminator to detect the 5’-fluorescein tag attached to the RNA. No degraded 5’-fluorescein-RNA fragments were observed in the case of the TMO/RNA duplex after RNase H1 treatment. Therefore, we concluded that the TMO3/RNA duplex is not RNase H1 active.
It is well established that oligonucleotides lacking inherent RNase H1 activity can be rendered active by adopting a gapmer oligonucleotide design, where such modifications flank an RNase H1-active DNA or DNA-phosphorothioate segment. Consequently, we explored the RNase H1 activity of gapmer TMO4, which contains TMO wings and a 14-mer DNA-phosphorothioate gap. The experiment was conducted using the same protocol as described for TMO3. The results illustrate that TMO4 does indeed stimulate RNase H1 activity.
Further insights into the RNase H1 recruiting ability of certain chemical modifications versus those that are inactive can be derived from the crystal structure analyses of human RNase H1 complexed with various DNA/RNA duplexes. While the RNA-binding domain of the enzyme dictates positional preference for the cleavage site, its catalytic core primarily interacts with the minor groove of an 11-base pair RNA/DNA duplex segment. The RNA strand precisely fits into one of the two grooves present in the active site, where it adopts a regular A-form structure and is largely recognized through interactions with four 2’-hydroxyl residues located around the scissile phosphorus within the catalytic core. DNA modifications exhibiting high flexibility have been shown to abolish RNase H1 activity. Nevertheless, the specificity for the DNA strand is imposed by a requirement for conformational flexibility that allows its accommodation into the phosphate-binding pocket. Given that significant distortions to the torsion angles of the oligonucleotide backbone are necessary to position the DNA strand within this pocket, only a flexible B-form oligonucleotide can meet this requirement. Additionally, this strand must be able to precisely fit within the DNA-binding channel formed by the “basic protrusion” site and must undergo numerous physical distortions within this channel, resulting in a heteroduplex that adopts a conformation intermediate between the A and B forms. The specificity of DNA over RNA is further dictated by a steric requirement, specifically the absence of 2’-hydroxyl groups, which in turn permits hydrogen bonding and van der Waals interactions within the basic protrusion channel. Thus, RNase H1 possesses highly complex structural requirements such that even a single chemical modification, when appropriately positioned within a DNA strand, can abolish its activity.
RNase H1 activity at a previously characterized cleavage site.65-66
Based on these findings, we hypothesize that the absence of RNase H1 activity in the TMO3/RNA duplex could be attributed to its limited conformational flexibility. This lack of flexibility arises from the consistent presence of rigid, six-membered morpholino rings throughout its backbone structure, which may prevent it from properly accessing the phosphate-binding pocket or interacting effectively with the DNA-binding channel within the enzyme’s catalytic site. This hypothesis is further supported by the work of Østergaard and colleagues, who recently demonstrated that a fluorinated nucleoside modification containing a rigid six-membered ring, when systematically incorporated across the gap region of a 3-9-3 gapmer, led to a significant reduction in RNase H1 activity at most positions compared to its 2’-fluoro-2’-deoxynucleoside analog.
TMO4, in contrast, contains a substantial 14-nucleotide DNA-phosphorothioate gap and exhibits RNase H1 activity. This observation aligns with previous research indicating that a gap length of 7 to 10 nucleotides is sufficient to support efficient RNase H1 catalysis, while a complete loss of activity occurs when the gap length is less than 5 deoxynucleotides. Furthermore, comprehensive sequencing-based experiments characterizing the sequence preferences of various RNase H1 enzymes have shown that RNA cleavage sites predicted to have the highest probability for RNase H1 cleavage are located within this optimal gap length.
In summary, oligonucleotides modified exclusively with TMO units, such as TMO3, do not elicit RNase H1 activity, likely due to their inherent conformational rigidity. However, a gapmer design like TMO4 can be rendered RNase H1 active. This suggests that oligonucleotides with uniform TMO modification are potential candidates for exon skipping experiments, where they function as steric blockers, whereas gapmers incorporating TMO wings can be utilized for antisense experiments that aim to degrade messenger RNA through activation of the RNase H1 pathway.
MicroRNAs are short non-coding RNA molecules that play critical roles in post-transcriptional gene regulation by precisely controlling both the stability and the rate of translation of messenger RNAs. They suppress translation by guiding the miRNA-induced silencing complex, or miRISC, to partially complementary sites located in the 3’-untranslated region of their messenger RNA targets, leading to deadenylation, translational repression, or messenger RNA degradation. Their widespread involvement in the pathogenesis of various human diseases makes them attractive targets for oligotherapeutic interventions.
Antisense microRNA inhibitors, also known as anti-miRs, are chemically modified short oligonucleotide analogs that exhibit complementarity to an endogenous mature microRNA of interest. They function by competitively binding to and sequestering the microRNA target within the cytoplasm, resulting in microRNA inhibition and the subsequent derepression of its messenger RNA target. Although the precise mechanism of action is still under investigation, it was initially hypothesized that anti-miRs stimulated microRNA degradation through an RNase H1-dependent pathway. However, more recent studies have demonstrated that anti-miRs inhibit microRNA function primarily through a steric blocking mechanism, physically associating with Argonaute-bound microRNAs. The consistently poor activity of potent gapmer designs in microRNA assays has led to the hypothesis that the Argonaute-associated microRNA/anti-miR complex may not be accessible to RNase H1. Chemical modifications such as 2′-O-methyl, 2′-O-methoxyethyl, 2′-fluoro, and locked nucleic acid have demonstrated effective microRNA targeting in both in vitro and in vivo experiments.
Luciferase reporter gene assays are commonly used to evaluate the potency of anti-miRs in microRNA-mediated gene regulation. To investigate whether TMOs can function as microRNA inhibitors in vitro, we utilized a HeLa cell line with two stably integrated luciferase reporters. The Firefly luciferase gene, under the transcriptional control of a weak upstream constitutive promoter and a downstream polyadenylation signal, served as an internal control. A perfectly complementary binding site for hsa-miR-15b-5p was inserted between the Renilla luciferase coding sequence and the polyadenylation signal in a specific plasmid to transcribe a chimeric Renilla messenger RNA containing both the Renilla-coding sequence and the miR-15b-5p target site. When this reporter plasmid is stably expressed in HeLa cells, Renilla luciferase expression is suppressed by endogenous miR-15b-5p, presumably through perfect complementary binding to its 3’UTR and subsequent degradation of the Renilla messenger RNA. Effective inhibition of miR-15b-5p using anti-miRs leads to the derepression of Renilla gene expression. This increased expression of Renilla luciferase can be monitored by measuring the luciferase activity of both Renilla and Firefly luciferase using a dual luciferase reporter assay. Thus, an increase in Renilla luciferase production correlates with the effectiveness of the anti-miR in binding to endogenous miR-15b, while Firefly luciferase provides an internal control for normalization of experimental values. The dynamic range of reporter assays is proportional to the copy number of the chosen microRNA target in cells. Inhibiting more abundant microRNAs can result in a larger fold change in Renilla luciferase expression compared to miR-15b-5p, which has moderate abundance in HeLa cells.
As an initial screen for biological activity, TMOs 3, 4, and 5, designed with varying numbers and placements of TMO linkages, were tested in the 0.1 to 1 micromolar concentration range in the HeLa-15b system described above. A non-targeting control and appropriate mock controls, involving only transfection reagent to account for its effects on cells, were included. An anti-miR with an identical sequence but exclusively 2’-O-methyl modified with alternating phosphate and phosphorothioate linkages was used as a positive control. A DNA-phosphorothioate oligonucleotide with an identical sequence served as the negative control. All oligonucleotides used in this study, except the non-targeting control, were full-length reverse complements of the endogenous mature miRNA-15b-5p.
To conduct the experiment, each oligonucleotide was transfected into 60% confluent HeLa-15b cells at concentrations ranging from 0.1 to 1 micromolar per well in a 96-well plate using lipofection. Following transfection, the cells were harvested, lysed, and analyzed after 24 hours. Biological activity was assessed using the Dual-Luciferase Reporter Assay System according to the manufacturer’s instructions. The ratio of Renilla luminescence to Firefly luminescence for oligonucleotide-treated cells was quantified using a luminometer, and this data was normalized to the ratio of mock-transfected cells, reported as fold change in Renilla/Firefly luminescence relative to the lipid-only transfection control, which was set to 1. Thus, a taller bar on a histogram indicates higher Renilla luciferase expression due to more efficient miR-15b-5p inhibition. The results of this experiment are presented.
The DNA-phosphorothioate control exhibited poor activity throughout this assay, whereas the chimeric 2’-O-methyl control and TMO5 showed a 6 to 7-fold improvement in activity within 24 hours of transfection. TMO3, which was exclusively TMO modified, displayed moderate levels of activity, resulting in a 3 to 4-fold change in Renilla luciferase expression. The gapmer TMO4 did not show appreciable levels of activity. The non-targeting control TMO9 showed an approximately 2-fold increase in activity in the 0.25 to 1 micromolar concentration range, suggesting the presence of off-target effects at higher concentrations. The maximum fold-change in Renilla luciferase expression was observed for the 2’-O-methyl control, followed by TMO5, with no further increase in activity observed when concentrations exceeded 100 nanomolar. This plateau in activity could be due to the low to moderate copy numbers of miR-15b-5p in HeLa cells, implying that maximum inhibition might have been achieved at 100 nanomolar. Although a clear dose-dependent inhibition could not be observed at higher concentrations, this initial experiment demonstrated that TMO5 is approximately 6-fold more active compared to the DNA-phosphorothioate negative control.
In summary, preliminary screening at high oligonucleotide concentrations indicated that one or more TMO designs may have the potential for further investigation as anti-miRs. However, it is well known that cellular toxicity, anti-proliferative effects, and the manifestation of off-target effects due to non-specific binding increase with increasing oligonucleotide concentration. Therefore, we attempted to identify the lowest effective doses for the aforementioned TMO designs by performing this experiment at a lower concentration range, from 0.1 to 100 nanomolar. The transfection protocols were carried out as previously described.
The DNA-phosphorothioate anti-miR, serving as a negative control, showed poor activity at all tested concentrations. This can be explained by its reduced RNA binding affinity. The 2’-O-methyl positive control containing 50% phosphorothioate linkages effectively inhibited miR-15b-5p at a 100 nanomolar concentration, resulting in a greater than 7-fold change in Renilla luciferase expression. This is likely because the increased binding affinity of 2’-O-methyl sugars in the 22-mer oligonucleotide compensates for the decrease in melting temperature caused by the 11 phosphorothioate linkages, leading to a biologically active anti-miR with high target affinity and enhanced nuclease stability. In comparison, TMO5 and TMO6, consisting of alternating TMO and DNA-phosphorothioate linkages, also produced a 7-fold change in Renilla luciferase expression at a 100 nanomolar concentration. Despite its high RNA binding affinity, the exclusively TMO modified anti-miR construct, TMO3, was less active and showed only a three-fold inhibition relative to mock-transfected cells. Although the precise reasons remain unclear, this reduced activity could be attributed to its inability to invade and bind the miRISC complex, poor transfection efficiency, or altered intracellular distribution compared to TMO5 and TMO6. While the gapmer anti-miR TMO4 with TMO wings and a 14-mer DNA-phosphorothioate gap is capable of eliciting RNase H1 activity, it was not a potent inhibitor of miR-15b-5p, presumably due to its poor target binding ability and inherent susceptibility to endonucleases.
As shown, the highest miRNA inhibition levels were achieved at 100 nanomolar. A maximal fold-response of approximately 8-fold was detected for TMO5 at a 100 nanomolar concentration at the 48-hour time point. No further increase in activity was observed at higher TMO5 concentrations or at longer time points for either the TMO-based designs or the 2’-O-methyl positive control. This is likely because the dynamic range of the luciferase assay is limited by the moderate copy numbers of miR-15b-5p in HeLa cells.
The anti-miR activities observed at 48 hours were sustained at 72 hours, suggesting that TMOs 5 and 6 possess sufficient in vitro stability over this time period. Although the 2’-O-methyl positive control displayed the most potent inhibition in the 25 to 50 nanomolar range, its activity dropped below that of TMOs 5 and 6 at the lowest tested concentration of 10 nanomolar. The higher potency of TMOs 5 and 6 at 10 nanomolar might be due to their enhanced nuclease stability, resulting in a greater availability of undegraded anti-miRs over a longer duration. Additionally, a dose-dependent increase in Renilla luciferase expression was observed for alternately modified TMOs 5 and 6, as well as the 2’-O-methyl control, in the 10 to 100 nanomolar concentration range; a 2 to 5-fold increase in Renilla luciferase levels was observed at 10 nanomolar, but more robust Renilla luciferase derepression resulted in a 7 to 8-fold inhibition at 100 nanomolar. The DNA-phosphorothioate control and gapmer TMO4 did not effectively inhibit miR-15b-5p and showed negligible activity during this assay. The non-targeting control TMO9 maintained similar activity levels as the DNA-phosphorothioate negative control in the 10 to 100 nanomolar concentration range, suggesting that higher sequence-specific binding and fewer off-target binding events might occur under these conditions.
In summary, appropriately designed TMOs can effectively function as miR-15b-5p inhibitors in vitro. A dose-response study in the 10 to 100 nanomolar range demonstrated that higher levels of derepression of Renilla messenger RNA can be achieved with increasing concentrations of TMOs. This experiment further illustrated that the number and distribution pattern of TMO linkages play a critical role in determining biological activity. This finding aligns with several literature reports that have shown that the placement of high-affinity modifications such as 2’-fluoro, 2’-O-methyl, 2’-O-methoxyethyl, locked nucleic acid, or non-nucleotide modifiers wholly dictates the potency of the resulting oligonucleotide to modulate the activity of miR-21. Further experiments are underway to elucidate both the in vitro and in vivo potency of various TMO designs in relevant disease models.
In conclusion, we have successfully developed a chemical synthesis strategy for preparing a novel class of synthetic molecules termed thiophosphoramidate morpholinos using phosphoramidite chemistry. In contrast to P(V)-based PMO synthesis, the phosphoramidite-based approach facilitates the straightforward incorporation of several valuable antisense modifications, such as 2’-O-methyl, 2’-O-methoxyethyl, locked nucleic acid, and others that are commercially available, providing an accessible synthetic route to generate a range of previously unexplored drug candidates for oligotherapeutics. The combination of morpholino nucleobases and phosphorothioate linkages resulted in an oligonucleotide construct with exceptional nuclease stability. Based on a 3’-exonuclease degradation assay using Snake Venom Phosphodiesterase I, TMOs exhibited minimal degradation over 23 hours and were three times more stable than a DNA-phosphorothioate control. Thermal denaturation studies revealed that both the number and placement of TMO modifications are crucial in determining binding affinity; while a TMO-DNA-phosphorothioate gapmer performed poorly, alternately modified TMO-DNA-phosphorothioate chimeras displayed excellent hybridization affinity towards both DNA and RNA complements. Additionally, ART899 it appeared that the uniform placement of rigid morpholino sugars throughout the oligonucleotide backbone results in a relatively rigid construct that binds with high affinity to RNA but not DNA. This lack of flexibility is also supported by the observation that an exclusively TMO modified oligonucleotide was RNase H1 inactive. Circular dichroism spectra indicated that the overall helical structure of all TMO/RNA heteroduplexes lies between the A and B forms, although the alternately TMO modified chimera displayed prominent A-form bands, approaching that of a typical RNA/RNA duplex. Preliminary bioactivity screens assessing the potency of various TMO designs as anti-miRs showed that some TMO designs may have applications as microRNA inhibitors. In summary, we have demonstrated that TMOs hold significant promise as effective research tools in molecular biology and may possess considerable potential as novel drug candidates in oligotherapeutics. Current efforts are focused on further expanding the scope of TMO-based drug candidates in various therapeutic modalities and elucidating their mechanisms of action.