# Fileset

[A neutralizable dimeric anti-thrombin aptamer with potent anticoagulant activity in mice.pdf](https://mdr.nims.go.jp/filesets/0fea8462-8147-4b41-82d0-e2400663ae6d/download)

## Creator

Masanobu Nagano, Kazuki Kubota, Asuka Sakata, Rei Nakamura, [Toru Yoshitomi](https://orcid.org/0000-0003-3847-1812), Koji Wakui, Keitaro Yoshimoto

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[A neutralizable dimeric anti-thrombin aptamer with potent anticoagulant activity in mice](https://mdr.nims.go.jp/datasets/0b662967-6470-4c18-96ae-9ef6126bb8dc)

## Fulltext

A neutralizable dimeric anti-thrombin aptamer with potent anticoagulant activity in miceOriginal ArticleA neutralizable dimeric anti-thrombin aptamerwith potent anticoagulant activity in miceMasanobu Nagano,1 Kazuki Kubota,1 Asuka Sakata,2 Rei Nakamura,1 Toru Yoshitomi,1,3 Koji Wakui,1and Keitaro Yoshimoto11Department of Life Sciences, Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro, Tokyo 153-8902, Japan; 2Medicinal Biology ofThrombosis and Hemostasis, Nara Medical University, 840 Shijo-cho, Kashihara, Nara 634-8521, JapanReceived 8 April 2023; accepted 31 July 2023;https://doi.org/10.1016/j.omtn.2023.07.038.3Present address: Research Center for Functional Materials, National Institute forMaterials Science, 1-1 Namiki, Tsukuba Ibaraki 305-0044, JapanCorrespondence: Keitaro Yoshimoto, Department of Life Sciences, GraduateSchool of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro,Tokyo 153-8902, Japan.E-mail: ckeitaro@mail.ecc.u-tokyo.ac.jpHeparin-induced thrombocytopenia (HIT) is a complicationcausedby administration of the anticoagulantheparin.Althoughthe number of patients with HIT has drastically increasedbecause of coronavirus disease 2019 (COVID-19), the currentlyused thrombin inhibitors for HIT therapy do not have antidotesto arrest the severe bleeding that occurs as a side effect; therefore,establishment of safer treatments for HIT patients is imperative.Here, we devised a potent thrombin inhibitor based on bivalentaptamers with a higher safety profile via combination with theantidote. Using an anti-thrombin DNA aptamer M08s-1 as apromising anticoagulant, its homodimer and heterodimer withTBA29 linked by a conformationally flexible linker or a rigidduplex linker were designed. The dimerized M08s-1-based ap-tamers had about 100-fold increased binding affinity to humanandmouse thrombin comparedwith themonomer counterparts.Administration of these bivalent aptamers into mice revealedthat the anticoagulant activity of the dimers significantly sur-passed that of an approved drug for HIT treatment, argatroban.Moreover, adding protamine sulfate as an antidote against themost potent bivalent aptamer completely suppressed the antico-agulant activity of the dimer. Emerging potent and neutralizableanticoagulant aptamers will be promising candidates for HITtreatment with a higher safety profile.INTRODUCTIONHeparin is a naturally occurring heparan sulfate and the first antico-agulant agent in history.1,2 Because heparin is in widespread clinicaluse not only for treatment of serious thromboembolisms, such asheart attack and disseminated intravascular coagulation, but alsofor prevention of thrombosis as a result of kidney dialysis, extracor-poreal membrane oxygenation (ECMO), or cardiopulmonary bypassmachines, it is included in the World Health Organization (WHO)Model List of Essential Medicines as an effective medicine neededin a healthcare system.3Any heparin therapy can induce heparin-induced thrombocytopenia(HIT) as a severe complication in up to 0.2%–3% of patients.4–7 HITis caused by the immune response to a complex of heparin with plateletfactor 4 (PF4), a neoantigen,which then drives abnormal production ofactivated coagulant factor II, referred to as thrombin. The hypercoag-ulable state induced by thrombin results in life-threatening arterial or762 Molecular Therapy: Nucleic Acids Vol. 33 September 2023 ª 2023This is an open access article under the CC BY license (http://creativenous thrombosis with development of stroke, myocardial infarction,and deep vein thrombosis (DVT). It is recommended that patientsseverely affected by coronavirus disease 2019 (COVID-19) receive hep-arin for prophylaxis of thrombosis in an ECMO machine or for treat-ment of thrombosis incurred as a complication, implying that thepotential number of patients with HIT is increasing globally.8–10HIT treatment is performed via an intravenous drip infusion ofthrombin inhibitors, such as a small molecule and a peptide, argatro-ban and bivalirudin (Figure S1).5,11 However, when severe life-threat-ening bleeding occurs during HIT treatment with conventionalthrombin inhibitors, the only way to reverse the anticoagulant effectis to cease the infusion; persistent bleeding for longer than the half-lives (around 30 min to 1 h) of the inhibitors is a considerable burdenfor patients.10 Therefore, developing an anticoagulant that can imme-diately neutralize the effects when necessary is critical for achievingsafe treatment of HIT.Nucleic acid aptamers are ligands composed of a single-strandedoligonucleotide (single-stranded DNA [ssDNA] or RNA) possessinga high binding affinity and specificity to a target of interest.12,13Because the complement sequence against such an aptamer is a highlyspecific, low-cost, and rapidly eliminated antidote, development of aneutralizable anticoagulant, a combination of the bioactive aptamerand its antidote, has long been a research target.13–16 TBA15 (HD1)is the first reported anti-thrombin DNA aptamer, and many anti-thrombin aptamers have been reported since (Figure S2A).17–19TBA15 exerts its anticoagulant activity by preventing fibrin formationvia binding to exosite I of thrombin and blocking fibrinogen binding(Figure 1A).20,21 Currently, the only anti-thrombin aptamer in a clin-ical trial is NU172, an updated version of TBA15, investigated fortreating off-pump coronary artery grafting bypass, with no reportsfor other diseases (Figure S2C).22 One of the reasons for this narrowThe Authors.vecommons.org/licenses/by/4.0/).https://doi.org/10.1016/j.omtn.2023.07.038mailto:ckeitaro@mail.ecc.u-tokyo.ac.jphttp://crossmark.crossref.org/dialog/?doi=10.1016/j.omtn.2023.07.038&domain=pdfhttp://creativecommons.org/licenses/by/4.0/Figure 1. Optimization of duplex linker length with the M08s-1 variant and TBA29(A) The ternary complex of human thrombin with TBA15 and TBA29 (PDB: 5EW1). (B) Schematic representation of the duplex linker (X is the number of base pairs). (C) Clottingtime assay to see the thrombin inhibition of the aptamer. (D) Length dependency of the duplex linker composed of A-T base pairs. (E) Length dependency of the duplex linkercomposed of A-T/G-C mixed base pairs. (F–I) Secondary structures of M08s-1 containing bivalent anti-thrombin aptamers in this study: Lin08-08, Lin08-29, Pse08-08, andPse08-29. The secondary structure of M08s-1 is an estimated structure based on previous studies and the Quadruplex forming G-Rich Sequences (QGRS) mapper.23www.moleculartherapy.orgdisease scope is that the in vivo anticoagulant activity of NU172 anddifferent, relatively new anti-thrombin aptamers has not been fullyinvestigated. Therefore, ones with higher anticoagulant activity in vivothan NU172 will be suitable candidates for HIT therapy and treat-ment for other diseases.Recently, we discovered an anti-thrombin DNA aptamer, M08s-1, us-ing systematic evolution of ligands by exponential enrichment (SE-LEX) with microbead-assisted capillary electrophoresis (MACE)and revealed that M08s-1 possesses higher anticoagulant activitythan NU172 (Figure S2E).24–26 Despite the promising anticoagulantactivity in vitro, the anticoagulant activity in an animal model hasnot been assessed. Furthermore, M08s-1 has sufficient structuralscope to enhance anticoagulant activity via dimerization.27,28 Forthese reasons, we are motivated to investigate the in vivo anticoagu-lant activity and reversal of the activity of M08s-1 and its dimers todevelop a neutralizable drug candidate for HIT.Here, we designed four M08s-1-based bivalent aptamers, where thelinker between the monomeric aptamer was constructed with classicflexible poly-deoxythymidine (dT) or a rigid duplex. Both linker typesof bivalent aptamers showed approximately 100-fold higher affinityto human and mouse thrombin than the monomeric counterpartM08s-1. Intravenous injection of these aptamers into mice showedsignificantly stronger anticoagulant activity than argatroban andNU172. Moreover, the anticoagulant activity of the discovered biva-lent dimers could be partially but strongly reversed by a short comple-mentary strand and even readily neutralized by protamine sulfate,indicating that they are potential alternatives to the drugs currentlyused for HIT therapy.Molecular Therapy: Nucleic Acids Vol. 33 September 2023 763http://www.moleculartherapy.orgTable 1. Evaluation of affinity of anti-thrombin aptamers to thrombin by SPRAptamerHuman thrombin Mouse thrombinka (1/Ms) kd (1/s)KD(nM) ka (1/Ms) kd (1/s)KD(nM)M08s 7.04 � 105 3.33 � 10�2 47.2 3.56 � 105 1.76 � 10�1 495TBA29 1.34 � 105 4.94 � 10�3 36.9 2.76 � 105 6.07 � 10�3 22.0Lin(08-08) 1.63 � 106 6.94 � 10�4 0.4 6.41 � 105 4.30 � 10�3 6.7Pse(08-08) 1.19 � 106 5.10 � 10�4 0.4 7.35 � 105 3.06 � 10�3 4.2Pse(08–29) 1.33 � 106 3.47 � 10�3 2.6 6.72 � 105 4.26 � 10�3 6.3Human or mouse thrombin was immobilized to the CM5 chip with a low RU of 500.The KD of Pse(08–29) was calculated as 1:1 binding by approximating that M08s andTBA29 possess the same KD as the homodimer.Molecular Therapy: Nucleic AcidsRESULTSDesign of M08s-1-based bivalent anti-thrombin aptamersTo increase the potency of the thrombin inhibitor aptamer M08s-1in vivo, harnessing avidity by dimerization is a simple and robustway to achieve it. So far, a heterodimeric aptamer, HD1-22, in whichTBA29 (HD22) as an anti-thrombin aptamer targeting another exosite(exosite II) of thrombin29,30 is linked to TBA15 via a conformationallyflexible single-stranded linker, has been well studied (Figure 1A).31Later, the optimal single-stranded linker between TBA15 and TBA29wasoptimizedby in vitro selection, resulting in that single-strand linkerfolding into a conformationally rigid duplex by taking a pseudo-circu-lar structure.32 BecauseM08s-1binds to exosite I of thrombin,33 similarto TBA15, we designed a heterodimer of M08s-1 and TBA29 with theduplex linker and optimized the length and composition of the linker(Figures 1B and S3A). Screening of the duplex linker was performedbased on thrombin clotting time (TCT)with purifiedfibrinogen, wherefibrin cleaved from fibrinogen by thrombin spontaneously aggregatesto increase the turbidity of the solution as a result of “clotting”; inhibi-tion of thrombin by the aptamer prolongs the clotting time (Fig-ure 1C).34 When the duplex linker was composed of base pairs withhomo-dT andhomo-dAchains, where the basepair lengthwas variable(X = 0, 5, 9, 13, 17, and 21 bp in Figure 1C), the clotting was nearlycompletely inhibited by a linker length with 5 and 17 bp, reflectinghigh anticoagulant activity via thrombin inhibition (Figure 1D).Next, after changing the composition of the duplex linker fromhomo-dT homo-dA chains to the mixed base pairs of A:T and G:Cbased on a previously reported duplex linker,35 a linker length of 5and 17 bp was also critical for anticoagulant activity, among others(Figure 1E). These results suggest that inhibition of thrombin by theheterodimer ofM08s-1 andTBA29with a duplex linker is independentof the linker sequence but strictly dependent on the specific linkerlength. Because the 17-bp length showedmore potent inhibition activ-ity than the 5-bp linker, we selected the 17-bp linker. Furthermore, thelinker compositions were determined to be mixed A:T and G:C basepairs because of high thermodynamic stability compared with thehomo-dA:homo-dT duplex. Therefore, we designed M08s-ds17-TBA29 (Pse08-29), having our optimized duplex linker as a hetero-dimer of M08s-1 and TBA29, aiming for a potent pseudo-circularbivalent aptamer in vivo (Figure 1I).764 Molecular Therapy: Nucleic Acids Vol. 33 September 2023Because the inhibition activity is not derived from TBA29 butM08s-1, examination of a homodimer of M08s-1 is also needed inaddition to pse08-29 to investigate the best dimer with high anticoag-ulant activity. Furthermore, in vivo, the structural properties of thelinker-connecting aptamers would also affect behavior; therefore, aclassic flexible linker should also be investigated. Finally, we designedthree M08s-1-based bivalent aptamers with M08s-dT17-M08s andM08s-dT17-TBA2936 (Lin08-08 and Lin08-29) with the classic linearof 17 mer dT linker (Figures 1F and 1G), and M08s-ds17-M08s(Pse08-08) with a rigid linker with 17-bp duplex for further experi-ments (Figure 1H).Affinity evaluation of dimeric anti-thrombin aptamersNext, we evaluated the affinity of the bivalent aptamers to human andmouse thrombin using surface plasmon resonance (SPR) using acco-ciation rate constant (ka), dissociation rate constant (kd), and dissoci-ation constant (KD). Upon immobilization of thrombin onto a car-boxymethylated matrix, the response unit (RU) was set to a lowvalue (500) to bias the 1:1 binding of the dimeric aptamer to thrombinand avoid multiple binding modes (Figures S4C–S4D). All dimericaptamers were readily prepared from the corresponding ssDNAs us-ing an annealing procedure. With human thrombin, although M08sand TBA29 bound with a similar dissociation constant (KD) of 47.2and 36.9 nM, respectively (Table 1; Figure S5A), whereas M08s-1 ho-modimers, Lin08-08, and Pse08-08 increased their affinity to almost100-fold that of the monomer counterpart independent of the linkertype (KD = 0.4 nM). Pse08-29, the heterodimer with a rigid linker,showed 15-fold higher affinity than the monomer counterparts butweaker affinity than M08s-1 homodimers, presumably because ofthe distant binding site between exosite I and II (Figures S4Cand S4D).Understanding the binding affinity of aptamers to mouse thrombinis important for performing mouse dosing experiments. The bind-ing affinity of monomeric M08s-1 to mouse thrombin had a10-fold decreased KD value compared with that of human thrombin(47.2 vs. 495 nM) (Table 1; Figure S5B). However, TBA29 showedsimilar affinity regardless of the species difference (36.9 nM vs.22 nM). Although the affinity of the M08s-1 homodimers decreased10-fold compared with that against human thrombin, this stillmaintained a 100-fold higher binding affinity to the monomer coun-terparts (KD = 495 nM vs. 6.7 and 4.2 nM, respectively), indicatingthat the considerable anticoagulant activity of dimeric aptamerswould be observed in mice. The pseudo-circular heterodimerPse08-29 maintained a similarly high binding affinity to humanthrombin (KD = 2.6 nM vs. 6.3 nM). Although determination ofthe KD of Lin08-29 with the flexible poly(dT) linker was not feasiblebecause of the complex binding modes (Figure S4E), an avidity anal-ysis based on the koff value was performed to compare all dimers inparallel, which determined that the degree of avidity of Lin08-29 wassimilar to that of other dimers (Figure S6).37 Collectively, thedimeric aptamers possessed high binding affinity to one of themonomer counterparts. The loss of the binding affinity to mousethrombin for M08s-1 could be overcome by increased bindingFigure 2. Analysis of anticoagulant activity of aptamers and drugs using aPTT in spiked plasma(A) Schematic of aPTT. (B) The aPTT in human plasma. (C) The aPTT in mouse plasma. (D) Concentration dependency of aptamers on the anticoagulant activity in mouseplasma. The clotting time of the aptamer (�) as a control in human plasma andmouse plasmawas 27.1 and 25.1 s, and the relative clotting time at the y axis was calculated bysubtraction of the averaged value of the control gained in triplicate. Error bars indicate SD (N = 3). The statistical significance was tested using a t test. ***p < 0.001,****p < 0.0001.www.moleculartherapy.orgaffinity via dimerization, suggesting the value of testing the antico-agulant activity in mice.Evaluation of the anticoagulant activity of M08s-1-basedbivalent aptamers by activated partial thromboplastin time(aPTT) in spiked plasmaWith the confirmation of high binding affinity of dimeric aptamers tothrombin, we next assessed the anticoagulant activity in plasma bymeasuring aPTT, a clinically used monitoring method for parenteralanticoagulant drugs such as heparin and argatroban.38,39 Notably, theaPTT assay differs from the simple TCTwith purified fibrinogen (Fig-ure 1C) because the aPTT assay allows assessment of the outcome ofthe coagulation cascade from all coagulant factors in an intrinsicpathway (Figures 2A and S7). For instance, exosite I is also exposedon prothrombin, where binding of the aptamers to exosite I canreduce the rate of prothrombin activation and could prolong aPTTwithout having a direct effect on thrombin activity in plasma;40,41therefore, aPTT reflects inhibition of thrombin and prothrombin.The “relative aPTT” was calculated by subtracting the aPTT of PBSspiked in human plasma (27 s) from the aPTT of a tested samplespiked in human plasma. We first investigated the anticoagulant ac-tivity of the monomeric M08s-1 and compared this with the knownanti-thrombin DNA aptamers TBA13, TBA29, RE31,42 and NU172(Figures S2A–S2D); in addition, the argatroban and bivalirudinanti-thrombin drugs indicative of HIT were tested (Figures S1Aand S1B).Measurement of the drugs andmonomeric aptamers at mo-lecular equivalence (0.33 mM) in human plasma using the aPTT assaydemonstrated a relative aPTT ofM08s-1 at 20 s, which was the longestof those tested and was greater than that of previously reported ap-tamers and even of drugs (Figure 2B). This tendency was alsoobserved in mouse plasma (Figure 2C), suggesting that M08s-1 at amonomer level was the most potent thrombin inhibitor among thetested aptamers and drugs.We next compared the M08s-1-based dimers with the known dimericaptamers RA-36,43 TBA15-dA15-TBA29,44 and 0/0A2/0A4,45 which area homodimers of TBA15 linkedwith a single dT linker, a classic hetero-dimer with TBA15 and TBA29 attached by a flexible poly(dA) linker,and a rationally designed heterodimer with RE31 and TBA29 with arigid duplex linker, respectively (Figures S2F–S2H). The aPTT assayin human plasma showed that the two heterodimers, Lin08-29 andPse08-29, exhibited a similar degree of prolongation of the relativeaPTT than the previously reported heterodimer 0/0A2/0A4 (relativeaPTT = 49 s). However, the two M08s-1-homodimers, Lin08-08 andMolecular Therapy: Nucleic Acids Vol. 33 September 2023 765http://www.moleculartherapy.orgFigure 3. Anticoagulant activity of the aptamers after their systemic administration to mice(A) Schematic of the experiment. (B) The maximum anticoagulant effect at the 3-min time point using a uniform dose of 1.3 mmol/kg. The statistical significance was testedusing a t test. *p < 0.1, **p < 0.01, ****p < 0.0001. (C) Time course of clotting time plotted against collected serum after the injection at several time points. Relative clotting timewas obtained by subtracting the treated data by 22.6 s, an averaged value of aptamer (�) as control aPTT in a triplicate. (D) Pharmacokinetics (PK) study of the administeredaptamers, where the concentration inmouse plasmawas calculated based on the standard curve from anticoagulant activity in spikedmouse plasma (Figure 2D). Argatrobanwas not calculated because of lack of aPTT time. (E) The half-lives (t-half) of the aptamers are based on PK analysis.Molecular Therapy: Nucleic AcidsPse08-08, showed70sand85sof relativeaPTT, respectively (Figure2B).When the therapeuticwindowof unfractionatedheparin,which is in the2- to 3-fold range against naive aPTT as a control (27 s),46 is used forevaluation of the aptamers, the therapeutic window in relative aPTTcan bedefined to be a range of 27–54 s, suggesting that bothM08s-1-ho-modimers showing aPTTs of �85 s bear enough effectiveness. Formouse plasma, compared with the aPTT of dimers, the M08s-1-based-heterodimers Lin08-29 and Pse08-29 showed a higher anticoag-ulant effect (40 s and 50 s of relative aPTT) than M08s-1-homodimersand other previously reported dimers (Figure 2C).We next evaluated the concentration dependency of the aptamers andargatroban for the anticoagulant activity in mouse plasma (Fig-ure 2D). Although argatroban required approximately 12 mM to reachthe maximum therapeutic window of 50 s of relative aPTT, M08s-1could reach the time at a 3-fold lower concentration (4 mM). Further-more, four M08s-1-based bivalent aptamers nearly reached 50 s ofrelative aPTT at 0.3–0.45 mM, suggesting that these aptamers wereeffective at 30-fold and 10-fold lower concentrations comparedwith argatroban and monomeric M08s-1, respectively. BecauseM08s-1-scm, whose nucleobase sequence were designed by scram-bling that ofM08s-1, did not reach themaximum therapeutic windowof more than 30 mM, the working range of the dimeric aptamers at alow 0.3–0.45 mMwas likely the outcome of structured M08s-1 unit tothrombin and prothrombin in plasma. Overall, the homodimers766 Molecular Therapy: Nucleic Acids Vol. 33 September 2023Lin08-08 and Pse08-08 and heterodimers Lin08-29 and Pse08-29showed the strongest anticoagulant activity in human plasma ormouse plasma, among others, suggesting that dimerization basedon M08s-1 was a reasonable approach to produce potent anticoagu-lant activity.Assessment of anticoagulant activity of the bivalent aptamersin vivo using aPTTBecause the M08s-1-based dimeric aptamers showed high anticoagu-lant activity using the aPTT assay with spiked plasma, we investigatedtheir activity in mice. This experiment followed the way of clinicalmonitoring, where plasma from collected blood samples after injec-tion of the aptamers was assessed using the aPTT assay (Figure 3A).Although argatroban and bivalirudin are administered by continuousinfusion in the clinic, in this study, the aptamers were injected at a sin-gle dose level to investigate the efficacy and pharmacokinetics inadvance. To do this, argatroban, NU172, M08s-1, and four M08s-1-bivalent aptamers (Lin08-08, Lin08-29, Pse08-08, and Pse08-29)were administered separately to mice via intravenous bolus injectionat molar equivalents (0.13 mmol/kg dose). Then, the aPTT 3 min afterinjection was defined as the maximum aPTT, where all bivalent ap-tamers showed a more prolonged relative aPTT than argatroban orthe monomeric aptamers NU172 and M08s-1 (Figure 3B). The heter-odimer with a rigid linker, Pse08-29, showed the highest anticoagu-lant activity among M08s-1-based dimers. Next, mice were dosedwww.moleculartherapy.orgwith monomeric aptamers and argatroban at 0.13 mmol/kg dose andfour M08s-1-dimeric aptamers at 0.04 mmol/kg, and blood wascollected until 1 h after administration. Despite their lower injectiondoses than the monomers and argatroban, all dimeric aptamersshowed high anticoagulant activity based on the aPTT (Figure 3C).Besides aPTT, prothrombin time (PT) is also measured for approveddirect thrombin inhibitors (DTIs), including argatroban.47 Therefore,we performed the PT assay in the same manner via mouse injections(Figures S7, S8A, and S8B). Among the aptamers, M08s-based Lin08-08 and Pse08-29 prolonged the PT slightly longer than others, whichis the same trend as with aPTT. In all cases, the effects of aPPT weresignificant compared with PT, which was the same trend as for DTIsseen with the therapeutic dose.48Conversion of relative aPTT to molar concentration based on thestandard curves shown in Figure 2D can estimate the aptamer phar-macodynamics (Figure 3D). Two monomeric aptamers, NU172 andM08s-1, were rapidly diminished in the distribution phase after3min and behaved in a biphasic manner. In contrast, the dimers actedmore in a monophasic manner. The b-phase elimination half-lives inblood were less than 10 min, although the homodimers, Lin08-08 andPse08-08, showed slightly longer half-lives up to 31 min (Figure 3E).Collectively, intravenous administration of the aptamers and argatro-ban to mice followed by aPTT testing revealed that the heterodimerpse08-29 with the rigid duplex linker had the most potent anticoag-ulant activity in mice.Serum stability of bivalent anti-thrombin aptamersChemically unmodified aptamers are known to be labile against nu-cleases in blood. To better understand the mechanism of fast clear-ance in vivo, the stability of the monomeric M08s-1 and four dimericaptamers (Lin08-08, Lin08-29, Pse08-08, and Pse08-29) in 50% hu-man or mouse serum was investigated (Figures S10 and S11). Thehalf-lives of all the aptamers in human and mouse serum were greaterthan 30 min, implying that aptamer degradation in vivo was not themain reason for the rapid activity loss, likely because of distribution totissues or clearance to the kidneys.49,50Neutralization of the anticoagulant activity of the bivalentaptamersEven if aptamers have a short half-life, developing an antidote torapidly reverse their anticoagulant effects in severe bleeding in HITpatients is crucial. To achieve our goal of developing a neutralizablethrombin-inhibitory aptamer, we screened optimal antidote se-quences from 9 short complementary strands against the M08s-1site of Pse08-29, which showed the most potent anticoagulant activityin mice (Figures S12A and S12B). Mixing of antidotes with a 16-foldconcentration of 0.33 mM of Pse08-29 in human serum followed byaPTT, the assessment identified [M08s-1 G2]c (22-mer, entry 3) ashaving the most efficient reversal of relative aPTT from 43 s to 19s. Then the dose dependency of [M08s-1 G2]c was tested in aPTTand revealed that it plateaued at a 4-fold concentration in the aPTTassay (Figures 4A and 4B). This suggested that the complementarysequence of theM08s-1 site on Pse08-29 could not completely reversethe anticoagulation effect of the aptamer.It is reported that some anticoagulant aptamers can be neutralizedwith protamine sulfate, the antidote to heparin.51,52 Therefore, wewanted to investigate whether polycationic protamine sulfate couldbe used as a reversing agent for the bivalent aptamer Pse08-29 bearinga negatively charged phosphate backbone. Protamine sulfate success-fully reversed the anticoagulant effect of Pse08-29 in a dose-depen-dent manner. It almost completely reversed at an aptamer: antidoteratio of 1:2 (w/w), restoring the aPTT to baseline (Figure 4C).Thus, not only the complementary strand but also protamine sulfatewere found to be valuable antidotes to control the activity of Pse08-29,which is vital for mitigating the bleeding risks associated with argatro-ban and bivalirudin used in HIT therapy.DISCUSSIONDeveloping a thrombin inhibitor with high anticoagulant activity andits antidote is an important issue for HIT therapy. The recentCOVID-19 pandemic raised the number of patients with HIT becauseof the increased use of heparin; unfortunately, currently used drugsfor HIT have no antidotes to arrest bleeding as a side effect. In thisstudy, we devised M08s-1-based bivalent aptamers with higher anti-coagulant activity than other aptamers and even current drugs usedfor HIT. We demonstrated that an antidote could reverse the activityand is useful for developing safer drugs for HIT treatment.Although many dimeric anti-thrombin aptamers have been reported,most are heterodimers, where TBA15 targeting exosite I and TBA2953targeting exosite II of thrombin are linked with conformationally flex-ible poly(dA) or poly(dT) linkers as a hinge for bivalent tight binding(Figures 1A and S2B).31,44,54,55 Ahmad et al.32 used in vitro selectionto optimize the linker composition between TBA15 and TBA29, re-sulting in a conformationally rigid duplex-type linker in which theoverall structure becomes a pseudo-circular (dumbbell-like) bivalentaptamer with a nick. Based on this knowledge, screening of the lengthof the duplex linker betweenM08s-1 and TBA29 resulted in high anti-coagulant activity with a linker with 5 or 17 bp. This trend is presum-ably due to the 12-bp difference in length between them. Because ahelical pitch of the DNA duplex is approximately 10 bp,56 these 5-and 17-bp lengths may match the topology of the two aptamers toinduce tight binding to thrombin by forming a 2:2 or 3:3 complexother than standard 1:1 or 2:1 formation (Figure S3B).45 Finally, wedesigned four M08s-1-based bivalent aptamers with a 17-bp rigidduplex (Pse08-08 and Pse08-29) or flexible poly(dT) linker (Lin08-08 and Lin08-29) to identify the optimal linker for M08s-1 dimers.In the experimental results obtained from SPR analysis, there was adifference in species specificity for the binding affinity, in which theone of monomeric M08s-1 to human thrombin reduced 10-foldcompared with the mouse thrombin. The reason why the affinity ofM08s-1 differed between human and mouse thrombin may bebecause the exosite I of human thrombin targeted by M08s-1 couldbe structurally different from that of mouse thrombin. ThisMolecular Therapy: Nucleic Acids Vol. 33 September 2023 767http://www.moleculartherapy.orgFigure 4. Neutralization of the anticoagulant effect by antidotes(A) Secondary structure of Pse08-29 and its partial complement sequence, [M08s G2]c, as an antidote. The solid black line inside gray is deoxyguanines predictedG-quadruplex structures suggested by QGRS mapper as G-sore, 36. (B) Neutralization assay with a short complementary strand assessed by aPTT. Pse08-29; 0.33 mM,complementary strand; 0, 0.66, 1.32, 2.64, or 5.28 mM. (C) Neutralization assay with protamine sulfate assessed by aPTT. Pse08-29; 5.6 mg/mL, protamine; 2.8, 5.6, or11.2 mg/mL. Error bars indicate SD (N = 3).Molecular Therapy: Nucleic Acidshypothesis was supported by alignment of the amino acid sequencesof both thrombins, where the mouse thrombin sequence containedmany differences in the exosite I region (Figure S9), and a similarresult between human and mouse thrombin was also seen in the in-hibition assay for TBA15 targeting exosite I.57 Overall, the bivalentaptamers yielded a 100-fold higher affinity than themonomeric coun-terpart to human and mouse thrombins, which benefit from retainingthe activity in vivo.The in vitro aPTT assay in spiked plasma revealed that monomericM08s-1 had higher anticoagulant activity than currently used drugsfor HIT treatment and other identified aptamers. The anticoagulantactivity of the designed dimers was approximately 10-fold higherthan that of M08s-1. The performance of the M08s-1-homodimersin human plasma using the aPTT assay was superior to that of theM08s-1-based-heterodimers. Intriguingly, the result in mouse plasmawas the opposite, likely because of the influence of decreased affinityof M08s-1 from human to mouse thrombin identified in the SPRexperiment (Table 1).The aPTT assay with the plasma of mice after administration of thebivalent aptamers suggested that the anticoagulant activity signifi-cantly surpassed that of the approved drugs for HIT and the clinicalcandidate aptamers argatroban and NU172, respectively. Two mono-meric aptamers, NU172 and M08s-1, were rapidly diminished in thedistribution phase after 3 min and behaved in a biphasic manner,whereas the dimers behaved in a monophasic manner; this differencewas probably due to the effect of the different molecular weights (8–13 kDa vs. 35–38 kDa) on their filtration by the kidneys (Figure 3D).58This residence trend of the dimeric aptamers within 3 min afteradministration would also be the reason for having high anticoagu-lant activity in mice compared with the monomeric aptamers. Theb-elimination half-life of the aptamers was less than 10 min, exceptfor M08s-1-based homodimers, which were similar to that of the re-ported unmodified aptamer (Figure 3E).50 A long half-life in mice wasseen in M08s-homodimeric aptamers, especially Lin08-08, up to31 min; however, the ones of the other dimers and monomers re-mained up to 8.5 min, which is typical behavior for aptamers (Fig-768 Molecular Therapy: Nucleic Acids Vol. 33 September 2023ure 3E). Although a chemically modified therapeutic aptamer, pegap-tanib, which targets vascular endothelial growth factor (VEGF) forocular vascular disease, prolonged the aPTT as a side effect becauseof the nonspecific interaction with plasma proteins at a 5 mg/kgdose,59,60 the M08s-1-based bivalent aptamers were chemically un-modified and showed sufficient efficacy at the lower dose of0.04 mg/kg, suggesting less risk for nonspecific prolongation ofaPTT. Although the reasons for the potent anticoagulant activityin vivo of pse08-29 compared with the other three bivalent dimersare unknown, our study suggests that the keys to the high potencyof the discovered bivalent aptamer, Pse08-29, would be at least (1)the selection of M08s-1, (2) the avidity harnessed by dimerization,(3) the high binding affinity of TBA29 to mouse thrombin, and (4)the high residence in blood at the distribution phase.Argatroban and bivalirudin, used to treat HIT, have no antidote,increasing the risk of the treatment. Therefore, designing an antidoteto Pse08-29, which showed high anticoagulant activity in vivo, iscrucial. We designed nine complementary sequences of the M08s-1site on Pse08-29, but they could not completely reverse the anticoagu-lation effect. Likely, Pse08-29 linked by a rigid duplex linker maybenefit the high thermodynamic stability and high anticoagulant ac-tivity, however, it could lose structural flexibility to be hijacked bycomplementary strands. In turn, using clinically approved protaminesulfate as an antidote for heparin successfully suppressed the highanticoagulant activity of pse08-29, strengthening the utility of thebivalent aptamer with existing drugs for HIT without an antidote.Further in vivo evaluations with bleeding and prevention of thrombusformation are currently underway.MATERIALS AND METHODSAll oligonucleotides at high performance liquid chromatography(HPLC) grade were purchased from Eurofins Genomics (Japan)(Table S1). Before use, each 150 mM aptamer in PBS was annealedby heating at 95�C for 3 min using a dry-bath incubator (Major Sci-ence, Taiwan) and cooling immediately in a heat block previouslystored at 25�C. The annealed aptamer was diluted in PBS for thedesired concentrations. All solutions were prepared using ultrapurewww.moleculartherapy.orgwater from a Milli-Q water purification system (Merck Millipore,USA). Human fibrinogen, mouse serum, 2-amino-2-hydroxymethyl-1,3-propanediol; Tris, EDTA-2Na-2H2O, 40 w/v% acrylamide/bismixed solution 19:1, ammonium persulfate (APS), N,N,N,N0-tetra-methylethylene; TEMED, 10w/v% polyoxyethylene(20) sorbitanmonolaurate solution; 10% Tween 20, Dulbecco’s PBS (�), urea,and protamine sulfate from salmon were purchased fromFUJIFILMWako Pure Chemicals Industries (Japan). Normal humanserum and normal mouse plasma were purchased from Cosmo Bio(Japan). Human thrombin and bivalirudin trifluoroacetate salt werepurchased from Sigma-Aldrich (MO, USA). Mouse thrombin (Crea-tive Biomart, NY, USA), argatroban monohydrate (TCI, Japan),normal human plasma (George King Bio-Medical, KS, USA), and10 mM sodium acetate (pH 5.5; Cytiva, MA, USA) were used asreceived. C57BL/6 mice were purchased from Japan SLC (Shizuoka,Japan).Inhibitory activity analysis of the aptamers in thrombin andfibrinogen mixed solutionThe experiment was performed based on a previous report.36 Briefly,to a 0.6-mL tube, 100 mL of pre-annealed 12.5 nM aptamer in PBS and100 mL of 6.25 nM human thrombin in PBS were added and incu-bated at room temperature for 15 min. 80 mL of the mixture wasadded to each well of a 96-well Costar� assay plate (Corning, WA,USA). Immediately after adding 20 mL of 6 mM human fibrinogenin PBS to each well of assay plate using an 8-serial pipettor, the clot-ting curve was measured as an increase in absorbance at 350 nm asso-ciated with fibrin gel formation using a UV-visible microplate reader(Epoch, Agilent Technologies, CA, USA). The initiation point of thecoagulation was determined by Origin 7.0 software. The experimentswere performed in duplicate.Affinity analysis of anti-thrombin aptamers using SPRThe running buffer for SPR analysis (PBS [pH 7.4], 0.05 [v/v%] sur-factant Tween 20) was filtered using Sartolab� RF 1000 (Sartorius,Germany). Pre-annealed aptamers in SPR running buffer werefiltered through an Acrodisc� syringe filter (Paul, WA, USA). SPRspectroscopy-based binding analysis was performed using a BiacoreX100 (Cytiva, MA, USA). 1 mM human thrombin and mousethrombin in 10 mM sodium acetate (pH 5.5) were coupled on aCM5 sensor chip CM5 (Cytiva, UK) by 1-ethyl-3-(3-dimethylamino-propyl)carbodiimide hydrochloride (EDC)/N-hydroxysuccinimide(NHS) chemistry at a flow rate of 10 mL min�1 in SPR running bufferuntil the RU reached 500. Using single-cycle kinetics, the concentra-tion series of the aptamers was injected using 1MNaCl solution as theregeneration buffer. The obtained data were fitted with a 1:1 bindingmodel using the Biacore X100 evaluation software (Cytiva, UK), andthe dissociation constant (KD) was calculated.aPTT assayaPTT in spiked plasma1.6 mL of each pre-annealed 30 mM aptamers in PBS and 48.4 mL ofnormal human plasma or normal mouse plasma were incubated at37�C for 1 min using a fully automated blood coagulation instrument(CA-620; Sysmex, Japan). Then, 50 mL of aPTT reagent (Thrombo-check; Sysmex) was added and further incubated at 37�C for 2 min.Then, 50 mL of 0.025 M calcium chloride was added, and aPTT wasmeasured immediately by tracking the change in scattered light inten-sity over time at 660 nm. The experiments were performed in tripli-cate. The statical analysis was performed using Tukey’s test andordinary one-way ANOVA (GraphPad Prism v.9.4.1).aPTT in plasma from aptamer-injected mice50 mL of the collected plasma sample stored at �147�C was thawedand incubated at 37�C for 1 min using a fully automated blood coag-ulation instrument (CA-620, Sysmex). Then, 50 mL of aPTT reagent(Thrombocheck, Sysmex) was added and further incubated at 37�Cfor 2 min. Then, 50 mL of 0.025 M calcium chloride was added, andaPTT was measured immediately by tracking the change in scatteredlight intensity over time at 660 nm.Injection into mice and collection of plasma samplesC57BL/6 mice (22 g, 7 weeks old, n = 3 or 5) were anesthetized byinhalation of isoflurane (Viatris, Canonsburg, PA, USA) and injectedwith pre-annealed aptamer in PBS with 0.04 mmol/kg (16 mM, 50 mL)or 0.13 mmol/kg (52 mM, 50 mL) by bolus intravenous injection via ju-gular vein, and blood was collected from the inferior vena cava 3, 10,30, or 60 min after administration. Collected blood samples wereimmediately mixed with 3.8% sodium citrate (Muto Pure Chemicals,Tokyo, Japan) (blood:3.8% sodium citrate = 9:1 [v/v]) and centrifugedat 1,500� g for 15 min at 25�C using a centrifuge (Thermo Fisher Sci-entific, MA, USA), and then the plasma layer was collected and storedat �147�C. Mice were euthanized by cervical dislocation under iso-flurane anesthesia after collection of blood at each time point.PT assay in plasma from aptamer-injected mice50 mL of the collected plasma sample stored at �147�C was thawedand incubated at 37�C for 1 min using a fully automated blood coag-ulation instrument (CA-620, Sysmex). Then, 100 mL of PT reagent(Thrombocheck, Sysmex) was added and further incubated at 37�Cfor 2 min. Then, 50 mL of 0.025 M calcium chloride was added, andPT was measured by tracking the change in scattered light intensityover time at 660 nm.Pharmacodynamics analysisThe aPTT resulting from the administered aptamers at each timepoint were converted into concentrations based on standard curvesgenerated from the in vitro anticoagulant effect from spiked mouseplasma in Figure 2D. The concentration at time points of 3, (6), 10,30, and 60 min were subjected to one-phase decay using GraphPadPrism v.9.4.1. to analyze their in vivo half-lives.Serum stability assessment of anti-thrombin aptamersThe experiment was performed based on a previous study.61 In a 0.6-mL low-absorption tube, 40 mL of pre-annealed 4 mM aptamer in PBSwas mixed with 40 mL of normal human serum or 40 mL of normalmouse serum and then incubated at 37�C using a cool incubator(Ikuta Industry, Japan). The sample solution was quenched by addingMolecular Therapy: Nucleic Acids Vol. 33 September 2023 769http://www.moleculartherapy.orgMolecular Therapy: Nucleic Acids20 mL of 100 mM EDTA-2Na-2H2O solution and refrigerated at 4�Cuntil its analysis. Then, 100 mL of 2� loading buffer containing 8 Murea, 2 mM EDTA-2Na-2H2O, and 2 mM Tris was added and heatedat 95�C for 5 min using a dry-bath incubator (Major Science). Afterdenaturation, the samples were subjected to electrophoresis using12% acrylamide gel containing urea at 200 V for 30 min. Then, thegel was stained by Gel Star Nucleic Acid Gel Stain (Lonza), and theband derived from the aptamer was visualized by UV irradiation us-ing Fusion Solo S (Vilber Lourmat, France). The bands were quanti-fied by Fusion Solo 6S Edge software (Vilber Lourmat).Neutralization analysis of the anticoagulant activity using aPTTin plasma48.3 mL of normal human plasma (George King Bio-Medical), 0.83 mLof a 60 mM pre-annealed aptamer (final concentration: 0.333 mM,5.6 mg/mL), and 0.83 mL of 120, 240, 480, or 960 mM complementsequence was incubated at 37�C for 1 min using a fully automatedblood coagulation instrument (CA-620, Sysmex). Then, 50 mL ofaPTT reagent (Actin FSL, Sysmex) was added and further incubatedat 37�C for 2 min. Then, 50 mL of 0.025 M calcium chloride wasadded, and aPTT was measured immediately by tracking the changein scattered light intensity over time at 660 nm. In the case of prot-amine sulfate as an antidote, 0.83 mL of 2.8, 5.6, or 11.2 mg/mL prot-amine sulfate in PBS was used instead of the complement sequence.The experiments were performed in triplicate.Ethical informationAll experiments using mice were conducted in accordance with theinstitutional guidelines approved by the Nara Medical UniversityInstitutional Animal Care and Use Committee (13239 and 13339).DATA AND CODE AVAILABILITYAll data generated during this study are included in this publishedarticle and its supplemental information or available upon request.SUPPLEMENTAL INFORMATIONSupplemental information can be found online at https://doi.org/10.1016/j.omtn.2023.07.038.ACKNOWLEDGMENTSK.Y. and A.S. received grants from AMED under grant number(JP22ak0101130), Japan. K.Y. received grants from the Japan Societyfor the Promotion of Science (JSPS) Grant-in-Aid for ScientificResearch (B) (18H02002), Japan; Grant-in-Aid for TransformativeResearch Areas (22H05049), Japan; and the SENSHIN MedicalResearch Foundation, Japan. M.N. received grants from JSPSGrant-in-Aid for Scientific Research (C) (21K06450), Japan; the Ike-tani Science and Technology Foundation, Japan; the Foundation forInteraction in Science & Technology, Japan; and the Research Foun-dation for Pharmaceutical Sciences, Japan.770 Molecular Therapy: Nucleic Acids Vol. 33 September 2023AUTHOR CONTRIBUTIONSK.Y. conceived this study. M.N. and A.S. designed the experiments.K.K., R.N., and A.S. performed the experiments. All authors wrotethe manuscript.DECLARATION OF INTERESTSM.N., K.K., A.S., and K.Y. are involved in a patent application relatedto this study (JP2022/212461). A.S. is a member of Medicinal Biologyof Thrombosis and Hemostasis established by Nara Medical Univer-sity and Chugai Pharmaceutical Co., Ltd. and the speaker’s bureaufrom CSL Behring. K.Y. is the CTO of LinkBIO Co., Ltd. K.Y. collab-oratively works under The University of Tokyo-LinkBIO Co., Ltd.;The University of Tokyo-Nara Medical University-Chugai Pharma-ceutical Co., Ltd.; or LinkBIO Co., Ltd.-Nara Medical University-Chugai Pharmaceutical Co., Ltd.REFERENCES1. Lim, G.B. (2017). Discovery and purification of heparin. Nat. Rev. Cardiol. 15, 69.https://doi.org/10.1038/nrcardio.2017.171.2. Oduah, E., Linhardt, R., and Sharfstein, S. (2016). Heparin: Past, present, and future.Pharmaceuticals 9, 38.3. World Health Organization (2021). WHO model list of essential medicines - 22ndlist, 2021. Tech. Doc. https://www.who.int/publications/i/item/WHO-MHP-HPS-EML-2021.02.4. Arepally, G.M. (2017). Heparin-induced thrombocytopenia. Blood 129, 2864–2872.5. Bakchoul, T., and Greinacher, A. (2012). Recent advances in the diagnosis and treat-ment of heparin-induced thrombocytopenia. Ther. Adv. Hematol. 3, 237–251.6. Watson, H., Davidson, S., and Keeling, D.; Haemostasis and Thrombosis Task Forceof the British Committee for Standards in Haematology (2012). Guidelines on thediagnosis and management of heparin-induced thrombocytopenia: second edition.Br. J. Haematol. 159, 528–540.7. Baroletti, S.A., and Goldhaber, S.Z. (2006). Heparin-Induced Thrombocytopenia.Circulation 114, 355–356.8. Warkentin, T.E., and Kaatz, S. (2020). COVID-19 versus HIT hypercoagulability.Thromb. Res. 196, 38–51.9. Spyropoulos, A.C., Levy, J.H., Ageno, W., Connors, J.M., Hunt, B.J., Iba, T., Levi, M.,Samama, C.M., Thachil, J., Giannis, D., et al. (2020). Scientific and StandardizationCommittee communication : Clinical guidance on the diagnosis , prevention , andtreatment of venous thromboembolism in hospitalized patients with. J. Thromb.Haemost. 18, 1859–1865.10. Schmidt, M., Hajage, D., Lebreton, G., Monsel, A., Voiriot, G., Levy, D., Baron, E.,Beurton, A., Chommeloux, J., Meng, P., et al. (2020). Extracorporeal membraneoxygenation for severe acute respiratory distress syndrome associated withCOVID-19 : a retrospective cohort study. Lancet Respir. Med. 8, 1121–1131.11. Vo, Q.A.T., Lin, J.K., and Tong, L.M. (2015). Efficacy and Safety of Argatroban andBivalirudine in Patients with Suspected Heparin-Induced Thrombocytopenia. Ann.Pharmacother. 49, 178–184.12. Dunn, M.R., Jimenez, R.M., and Chaput, J.C. (2017). Analysis of aptamer discoveryand technology. Nat. Rev. Chem 1, 0076.13. Zhou, J., and Rossi, J. (2017). Aptamers as targeted therapeutics: Current potentialand challenges. Nat. Rev. Drug Discov. 16, 181–202.14. Nimjee, S.M., Povsic, T.J., Sullenger, B.A., and Becker, R.C. (2016). Translation andClinical Development of Antithrombotic Aptamers. Nucleic Acid Therapeut. 26,147–155.15. Liu, M., Zaman, K., and Fortenberry, Y.M. (2021). Overview of the TherapeuticPotential of Aptamers Targeting Coagulation Factors. Int. J. Mol. Sci. 22, 3897–3917.16. Ponce, A.T., and Hong, K.L. (2019). A mini-review: Clinical development and po-tential of aptamers for thrombotic events treatment and monitoring. Biomedicines7, 55.https://doi.org/10.1016/j.omtn.2023.07.038https://doi.org/10.1016/j.omtn.2023.07.038https://doi.org/10.1038/nrcardio.2017.171http://refhub.elsevier.com/S2162-2531(23)00211-1/sref2http://refhub.elsevier.com/S2162-2531(23)00211-1/sref2https://www.who.int/publications/i/item/WHO-MHP-HPS-EML-2021.02https://www.who.int/publications/i/item/WHO-MHP-HPS-EML-2021.02http://refhub.elsevier.com/S2162-2531(23)00211-1/sref4http://refhub.elsevier.com/S2162-2531(23)00211-1/sref5http://refhub.elsevier.com/S2162-2531(23)00211-1/sref5http://refhub.elsevier.com/S2162-2531(23)00211-1/sref6http://refhub.elsevier.com/S2162-2531(23)00211-1/sref6http://refhub.elsevier.com/S2162-2531(23)00211-1/sref6http://refhub.elsevier.com/S2162-2531(23)00211-1/sref6http://refhub.elsevier.com/S2162-2531(23)00211-1/sref7http://refhub.elsevier.com/S2162-2531(23)00211-1/sref7http://refhub.elsevier.com/S2162-2531(23)00211-1/sref8http://refhub.elsevier.com/S2162-2531(23)00211-1/sref8http://refhub.elsevier.com/S2162-2531(23)00211-1/sref9http://refhub.elsevier.com/S2162-2531(23)00211-1/sref9http://refhub.elsevier.com/S2162-2531(23)00211-1/sref9http://refhub.elsevier.com/S2162-2531(23)00211-1/sref9http://refhub.elsevier.com/S2162-2531(23)00211-1/sref9http://refhub.elsevier.com/S2162-2531(23)00211-1/sref10http://refhub.elsevier.com/S2162-2531(23)00211-1/sref10http://refhub.elsevier.com/S2162-2531(23)00211-1/sref10http://refhub.elsevier.com/S2162-2531(23)00211-1/sref10http://refhub.elsevier.com/S2162-2531(23)00211-1/sref11http://refhub.elsevier.com/S2162-2531(23)00211-1/sref11http://refhub.elsevier.com/S2162-2531(23)00211-1/sref11http://refhub.elsevier.com/S2162-2531(23)00211-1/sref12http://refhub.elsevier.com/S2162-2531(23)00211-1/sref12http://refhub.elsevier.com/S2162-2531(23)00211-1/sref13http://refhub.elsevier.com/S2162-2531(23)00211-1/sref13http://refhub.elsevier.com/S2162-2531(23)00211-1/sref14http://refhub.elsevier.com/S2162-2531(23)00211-1/sref14http://refhub.elsevier.com/S2162-2531(23)00211-1/sref14http://refhub.elsevier.com/S2162-2531(23)00211-1/sref15http://refhub.elsevier.com/S2162-2531(23)00211-1/sref15http://refhub.elsevier.com/S2162-2531(23)00211-1/sref16http://refhub.elsevier.com/S2162-2531(23)00211-1/sref16http://refhub.elsevier.com/S2162-2531(23)00211-1/sref16www.moleculartherapy.org17. Bock, L.C., Griffin, L.C., Latham, J.A., Vermaas, E.H., and Toole, J.J. (1992). Selectionof single-stranded DNA molecules that bind and inhibit human thrombin. Nature355, 564–566.18. Macaya, R.F., Schultze, P., Smith, F.W., Roe, J.A., and Feigon, J. (1993). Thrombin-binding DNA aptamer forms a unimolecular quadruplex structure in solution.Proc. Natl. Acad. Sci. USA 90, 3745–3749.19. Riccardi, C., Napolitano, E., Platella, C., Musumeci, D., and Montesarchio, D. (2021).Pharmacology & Therapeutics G-quadruplex-based aptamers targeting humanthrombin: Discovery, chemical modifications and antithrombotic effects. Pharmacol.Ther. 217, 107649.20. Bock, P.E., Panizzi, P., and Verhamme, I.M.A. (2007). Exosites in the substrate spec-ificity of blood coagulation reactions. J. Thromb. Haemost. 5, 81–94.21. Troisi, R., Balasco, N., Autiero, I., Vitagliano, L., and Sica, F. (2021). Exosite Bindingin Thrombin : A Global Structural/Dynamic Overview of Complexes with Aptamersand Other Ligands. Int. J. Mol. Sci. 22, 10803.22. Hutabarat, R.M., McCauley, T., Makim, A., Lewis, S., Olsen, K., Wagner-Whyte, J.,Diener, J., Matyugicheva, U., Scull, J., Levy, M.D., et al. (2007). Pharmakokinetic/pharmacodynamic profile of a novel aptamer direct thrombin inhibitor in cynomol-gus monkeys and yorkshire pigs following a single IV bolus administration.J. Thromb. Haemost. 5.23. Kikin, O., D’Antonio, L., and Bagga, P.S. (2006). QGRS Mapper: a web-based serverfor predicting G-quadruplexes in nucleotide sequences. Nucleic Acids Res. 34,W676–W682.24. Wakui, K., Yoshitomi, T., Yamaguchi, A., Tsuchida, M., Saito, S., Shibukawa, M.,Furusho, H., and Yoshimoto, K. (2019). Rapidly Neutralizable and HighlyAnticoagulant Thrombin-Binding DNA Aptamer Discovered by MACE SELEX.Mol. Ther. Nucleic Acids 16, 348–359.25. Nagano, M., Toda, T., Makino, K., Miki, H., Sugizaki, Y., Tomizawa, H., Isobayashi,A., and Yoshimoto, K. (2022). Discovery of a Highly Specific Anti-methotrexate(MTX) DNA Aptamer for Antibody-Independent MTX Detection. Anal. Chem.94, 17255–17262.26. Nagano, M., Oguro, T., Sawada, R., Yoshitomi, T., and Yoshimoto, K. (2021).Accelerated Discovery of Potent Bioactive anti-TNFa Aptamers by Microbead-Assisted Capillary Electrophoresis (MACE)-SELEX. Chembiochem 22, 3341–3347.27. Thomas, B.J., Porciani, D., and Burke, D.H. (2022). Cancer immunomodulation us-ing bispecific aptamers. Mol. Ther. Nucleic Acids 27, 894–915.28. Riccardi, C., Napolitano, E., Musumeci, D., and Montesarchio, D. (2020). Dimericand multimeric DNA aptamers for highly effective protein recognition. Molecules25, 5227.29. Krauss, I.R., Pica, A., Merlino, A., Mazzarella, L., and Sica, F. (2013). Duplex-quad-ruplex motifs in a peculiar structural organization cooperatively contribute tothrombin binding of a DNA aptamer. Acta Cryst. D69, 2403–2411.30. Pica, A., Krauss, I.R., Parente, V., Tateishi-Karimata, H., Nagatoishi, S., Tsumoto, K.,Sugimoto, N., and Sica, F. (2017). Through-bond effects in the ternary complexes ofthrombin sandwiched by two DNA aptamers. Nucleic Acids Res. 45, 1461–1469.31. Müller, J., Freitag, D., Mayer, G., and Pötzsch, B. (2008). Anticoagulant characteristicsof HD1-22, a bivalent aptamer that specifically inhibits thrombin and prothrombi-nase. J. Thromb. Haemost. 6, 2105–2112.32. Ahmad, K.M., Xiao, Y., and Soh, H.T. (2012). Selection is more intelligent than design: improving the affinity of a bivalent ligand through directed evolution. Nucleic AcidsRes. 40, 11777–11783.33. Troisi, R., Napolitano, V., Rossitto, E., Osman,W., Nagano, M., Wakui, K., Popowicz,G.M., Yoshimoto, K., and Sica, F. (2023). Steric hindrance and structural flexibilityshape the functional properties of a guanine-rich oligonucleotide. Nucleic AcidsRes. gkad634. https://doi.org/10.1093/nar/gkad634.34. De Cristofaro, R., and Di Cera, E. (1991). Phenomenological Analysis of the ClottingCurve. J. Protein Chem. 10, 455–468.35. Kuai, H., Zhao, Z., Mo, L., Liu, H., Hu, X., Fu, T., Zhang, X., and Tan, W. (2017).Circular Bivalent Aptamers Enable in Vivo Stability and Recognition. J. Am.Chem. Soc. 139, 9128–9131.36. Yoshitomi, T., Wakui, K., Miyakawa, M., and Yoshimoto, K. (2021). Design strategyof antidote sequence for bivalent aptamer: Rapid neutralization of high-anticoagulantthrombin-binding bivalent DNA aptamer-linked M08 with HD22. Res. Pract.Thromb. Haemost. 5, e12503.37. Säfsten, P., Klakamp, S.L., Drake, A.W., Karlsson, R., and Myszka, D.G. (2006).Screening antibody-antigen interactions in parallel using Biacore A100. Anal.Biochem. 353, 181–190.38. Owen, H., and Tsai, N. (2019). Pharmacological Review of Anticoagulants.39. Ikoma, H. (2002). Development of Argatroban as an Anticoagulant andAntithrombin Agent in Japan. Pathophysiol. Haemostasis Thrombosis 32, 23–28.40. Kretz, C.A., Stafford, A.R., Fredenburgh, J.C., and Weitz, J.I. (2006). HD1 , aThrombin-directed Aptamer , Binds Exosite 1 on Prothrombin with HighAffinity and Inhibits Its Activation by Prothrombinase. J. Biol. Chem. 281,37477–37485.41. Spiridonova, V.A., Novikova, T.M., Sizov, V.A., Shashkovskaya, V.S., Titaeva,E.V., Dobrovolsky, A.B., Zharikova, E.B., and Mazurov, A.V. (2019). DNAAptamers to Thrombin Exosite I . Structure Function Relationships andAntithrombotic Effects. Biochemistry (Mosc.) 84, 1521–1528. https://doi.org/10.1134/S0006297919120113.42. Ikebukuro, K., Okumura, Y., Sumikura, K., and Karube, I. (2005). A novel method ofscreening thrombin-inhibiting DNA aptamers using an evolution-mimicking algo-rithm. Nucleic Acids Res. 33, e108.43. Zavyalova, E., Golovin, A., Reshetnikov, R., Mudrik, N., Panteleyev, D., Pavlova, G.,and Kopylov, A. (2011). Novel Modular DNA Aptamer for Human Thrombin withHigh Anticoagulant Activity. Curr. Med. Chem. 18, 3343–3350.44. Müller, J., Wulffen, B., Pötzsch, B., and Mayer, G. (2007). Multidomain TargetingGenerates a Bivalent Aptamer. Chembiochem 8, 2223–2226.45. Hu, X., Tang, L., Zheng, M., Liu, J., Zhang, Z., Li, Z., Yang, Q., Xiang, S., Fang, L., Ren,Q., et al. (2022). Structure-Guided Designing Pre-Organization in Bivalent Aptamers.J. Am. Chem. Soc. 144, 4507–4514.46. Bjornsson, T.D., and Nash, P.V. (1986). Variability in Heparin Sensitivity of APTTReagents. Am. J. Clin. Pathol. 86, 199–204.47. Gosselin, R.C., Dager, W.E., King, J.H., Janatpour, K., Mahackian, K., Larkin, E.C.,and Owings, J.T. (2004). Effect of Direct Thrombin Inhibitors, Bivalirudin,Lepirudin, and Argatroban, on Prothrombin Time and INR Values. Am. J. Clin.Pathol. 121, 593–599.48. van Veen, J.J., Spahn, D.R., and Makris, M. (2011). Routine preoperative coagulationtests: an outdated practice? Br. J. Anaesth. 106, 1–3.49. Kovacevic, K.D., Gilbert, J.C., and Jilma, B. (2018). Pharmacokinetics , pharmacody-namics and safety of aptamers. Adv. Drug Deliv. Rev. 134, 36–50.50. Healy, J.M., Lewis, S.D., Kurz, M., Boomer, R.M., Thompson, K.M., Wilson, C., andMccauley, T.G. (2004). Pharmacokinetics and Biodistribution of Novel AptamerCompositions. Pharm. Res. (N. Y.) 21, 2234–2246.51. Oney, S., Lam, R.T.S., Bompiani, K.M., Blake, C.M., Quick, G., Heidel, J.D., Liu, J.Y.C.,Mack, B.C., Davis, M.E., Leong, K.W., and Sullenger, B.A. (2009). Development ofuniversal antidotes to control aptamer activity. Nat. Med. 15, 1224–1228.52. Soule, E.E., Yu, H., Olson, L., Naqvi, I., Kumar, S., Krishnaswamy, S., Sullenger, B.A.,Naqvi, I., Kumar, S., and Krishnaswamy, S. (2022). Article Generation of an antico-agulant aptamer that targets factor V/Va and disrupts the FVa-membrane interactionin normal and COVID-19 patient samples ll Article Generation of an anticoagulantaptamer that targets factor V/Va and disrupts the FVa-me. Cell Chem. Biol. 29, 215–225.e5.53. Tasset, D.M., Kubik, M.F., and Steiner, W. (1997). Oligonucleotide Inhibitors ofHuman Thrombin that Bind Distinct Epitopes. J. Mol. Biol. 272, 688–698.54. Hasegawa, H., Taira, K.I., Sode, K., and Ikebukuro, K. (2008). Improvement ofAptamer Affinity by Dimerization. Sensors 8, 1090–1098.55. Zhao, S., Tian, R., Wu, J., Liu, S., Wang, Y., Wen, M., Shang, Y., Liu, Q., Li, Y., Guo, Y.,et al. (2021). ADNA origami-based aptamer nanoarray for potent and reversible anti-coagulation in hemodialysis. Nat. Commun. 12, 358.56. Crick, F., and Watson, J. (1953). Molecular structure of nucleic acids. Nature 25,737–738.Molecular Therapy: Nucleic Acids Vol. 33 September 2023 771http://refhub.elsevier.com/S2162-2531(23)00211-1/sref17http://refhub.elsevier.com/S2162-2531(23)00211-1/sref17http://refhub.elsevier.com/S2162-2531(23)00211-1/sref17http://refhub.elsevier.com/S2162-2531(23)00211-1/sref18http://refhub.elsevier.com/S2162-2531(23)00211-1/sref18http://refhub.elsevier.com/S2162-2531(23)00211-1/sref18http://refhub.elsevier.com/S2162-2531(23)00211-1/sref19http://refhub.elsevier.com/S2162-2531(23)00211-1/sref19http://refhub.elsevier.com/S2162-2531(23)00211-1/sref19http://refhub.elsevier.com/S2162-2531(23)00211-1/sref19http://refhub.elsevier.com/S2162-2531(23)00211-1/sref20http://refhub.elsevier.com/S2162-2531(23)00211-1/sref20http://refhub.elsevier.com/S2162-2531(23)00211-1/sref21http://refhub.elsevier.com/S2162-2531(23)00211-1/sref21http://refhub.elsevier.com/S2162-2531(23)00211-1/sref21http://refhub.elsevier.com/S2162-2531(23)00211-1/sref22http://refhub.elsevier.com/S2162-2531(23)00211-1/sref22http://refhub.elsevier.com/S2162-2531(23)00211-1/sref22http://refhub.elsevier.com/S2162-2531(23)00211-1/sref22http://refhub.elsevier.com/S2162-2531(23)00211-1/sref22http://refhub.elsevier.com/S2162-2531(23)00211-1/sref64http://refhub.elsevier.com/S2162-2531(23)00211-1/sref64http://refhub.elsevier.com/S2162-2531(23)00211-1/sref64http://refhub.elsevier.com/S2162-2531(23)00211-1/sref23http://refhub.elsevier.com/S2162-2531(23)00211-1/sref23http://refhub.elsevier.com/S2162-2531(23)00211-1/sref23http://refhub.elsevier.com/S2162-2531(23)00211-1/sref23http://refhub.elsevier.com/S2162-2531(23)00211-1/sref24http://refhub.elsevier.com/S2162-2531(23)00211-1/sref24http://refhub.elsevier.com/S2162-2531(23)00211-1/sref24http://refhub.elsevier.com/S2162-2531(23)00211-1/sref24http://refhub.elsevier.com/S2162-2531(23)00211-1/sref25http://refhub.elsevier.com/S2162-2531(23)00211-1/sref25http://refhub.elsevier.com/S2162-2531(23)00211-1/sref25http://refhub.elsevier.com/S2162-2531(23)00211-1/sref26http://refhub.elsevier.com/S2162-2531(23)00211-1/sref26http://refhub.elsevier.com/S2162-2531(23)00211-1/sref27http://refhub.elsevier.com/S2162-2531(23)00211-1/sref27http://refhub.elsevier.com/S2162-2531(23)00211-1/sref27http://refhub.elsevier.com/S2162-2531(23)00211-1/sref62http://refhub.elsevier.com/S2162-2531(23)00211-1/sref62http://refhub.elsevier.com/S2162-2531(23)00211-1/sref62http://refhub.elsevier.com/S2162-2531(23)00211-1/sref63http://refhub.elsevier.com/S2162-2531(23)00211-1/sref63http://refhub.elsevier.com/S2162-2531(23)00211-1/sref63http://refhub.elsevier.com/S2162-2531(23)00211-1/sref28http://refhub.elsevier.com/S2162-2531(23)00211-1/sref28http://refhub.elsevier.com/S2162-2531(23)00211-1/sref28http://refhub.elsevier.com/S2162-2531(23)00211-1/sref29http://refhub.elsevier.com/S2162-2531(23)00211-1/sref29http://refhub.elsevier.com/S2162-2531(23)00211-1/sref29https://doi.org/10.1093/nar/gkad634http://refhub.elsevier.com/S2162-2531(23)00211-1/sref30http://refhub.elsevier.com/S2162-2531(23)00211-1/sref30http://refhub.elsevier.com/S2162-2531(23)00211-1/sref31http://refhub.elsevier.com/S2162-2531(23)00211-1/sref31http://refhub.elsevier.com/S2162-2531(23)00211-1/sref31http://refhub.elsevier.com/S2162-2531(23)00211-1/sref32http://refhub.elsevier.com/S2162-2531(23)00211-1/sref32http://refhub.elsevier.com/S2162-2531(23)00211-1/sref32http://refhub.elsevier.com/S2162-2531(23)00211-1/sref32http://refhub.elsevier.com/S2162-2531(23)00211-1/sref33http://refhub.elsevier.com/S2162-2531(23)00211-1/sref33http://refhub.elsevier.com/S2162-2531(23)00211-1/sref33http://refhub.elsevier.com/S2162-2531(23)00211-1/sref34http://refhub.elsevier.com/S2162-2531(23)00211-1/sref35http://refhub.elsevier.com/S2162-2531(23)00211-1/sref35http://refhub.elsevier.com/S2162-2531(23)00211-1/sref36http://refhub.elsevier.com/S2162-2531(23)00211-1/sref36http://refhub.elsevier.com/S2162-2531(23)00211-1/sref36http://refhub.elsevier.com/S2162-2531(23)00211-1/sref36https://doi.org/10.1134/S0006297919120113https://doi.org/10.1134/S0006297919120113http://refhub.elsevier.com/S2162-2531(23)00211-1/sref38http://refhub.elsevier.com/S2162-2531(23)00211-1/sref38http://refhub.elsevier.com/S2162-2531(23)00211-1/sref38http://refhub.elsevier.com/S2162-2531(23)00211-1/sref39http://refhub.elsevier.com/S2162-2531(23)00211-1/sref39http://refhub.elsevier.com/S2162-2531(23)00211-1/sref39http://refhub.elsevier.com/S2162-2531(23)00211-1/sref40http://refhub.elsevier.com/S2162-2531(23)00211-1/sref40http://refhub.elsevier.com/S2162-2531(23)00211-1/sref41http://refhub.elsevier.com/S2162-2531(23)00211-1/sref41http://refhub.elsevier.com/S2162-2531(23)00211-1/sref41http://refhub.elsevier.com/S2162-2531(23)00211-1/sref42http://refhub.elsevier.com/S2162-2531(23)00211-1/sref42http://refhub.elsevier.com/S2162-2531(23)00211-1/sref43http://refhub.elsevier.com/S2162-2531(23)00211-1/sref43http://refhub.elsevier.com/S2162-2531(23)00211-1/sref43http://refhub.elsevier.com/S2162-2531(23)00211-1/sref43http://refhub.elsevier.com/S2162-2531(23)00211-1/sref44http://refhub.elsevier.com/S2162-2531(23)00211-1/sref44http://refhub.elsevier.com/S2162-2531(23)00211-1/sref45http://refhub.elsevier.com/S2162-2531(23)00211-1/sref45http://refhub.elsevier.com/S2162-2531(23)00211-1/sref46http://refhub.elsevier.com/S2162-2531(23)00211-1/sref46http://refhub.elsevier.com/S2162-2531(23)00211-1/sref46http://refhub.elsevier.com/S2162-2531(23)00211-1/sref47http://refhub.elsevier.com/S2162-2531(23)00211-1/sref47http://refhub.elsevier.com/S2162-2531(23)00211-1/sref47http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref48http://refhub.elsevier.com/S2162-2531(23)00211-1/sref49http://refhub.elsevier.com/S2162-2531(23)00211-1/sref49http://refhub.elsevier.com/S2162-2531(23)00211-1/sref50http://refhub.elsevier.com/S2162-2531(23)00211-1/sref50http://refhub.elsevier.com/S2162-2531(23)00211-1/sref51http://refhub.elsevier.com/S2162-2531(23)00211-1/sref51http://refhub.elsevier.com/S2162-2531(23)00211-1/sref51http://refhub.elsevier.com/S2162-2531(23)00211-1/sref52http://refhub.elsevier.com/S2162-2531(23)00211-1/sref52http://www.moleculartherapy.orgMolecular Therapy: Nucleic Acids57. Zavyalova, E., Golovin, A., Timoshenko, T., Babiy, A., Pavlova, G., and Kopylov, A.(2012). DNA Aptamers for Human Thrombin with High Anticoagulant ActivityDemonstrate Target- and Species-Specificity. Curr. Med. Chem. 19, 5232–5237.58. Zhou, J., and Rossi, J. (2017). Aptamers as targeted therapeutics: current potential andchallenges. Nat. Rev. Drug Discov. 16, 181–202.59. Pharmaceuticals and Medical Devices Agency (2008). Review Report for Macugen.772 Molecular Therapy: Nucleic Acids Vol. 33 September 202360. Ng, E.W.M., Shima, D.T., Calias, P., Cunningham, E.T., Guyer, D.R., and Adamis,A.P. (2006). Pegaptanib, a targeted anti-VEGF aptamer for ocular vascular disease.Nat. Rev. Drug Discov. 5, 123–132.61. Ueki, R., Uchida, S., Kanda, N., Yamada, N., Ueki, A., Akiyama, M., Toh, K., Cabral,H., and Sando, S. (2020). A chemically unmodified agonistic DNA with growth factorfunctionality for in vivo therapeutic application. Sci. Adv. 6, eaay2801.http://refhub.elsevier.com/S2162-2531(23)00211-1/sref53http://refhub.elsevier.com/S2162-2531(23)00211-1/sref53http://refhub.elsevier.com/S2162-2531(23)00211-1/sref53http://refhub.elsevier.com/S2162-2531(23)00211-1/sref54http://refhub.elsevier.com/S2162-2531(23)00211-1/sref54http://refhub.elsevier.com/S2162-2531(23)00211-1/sref55http://refhub.elsevier.com/S2162-2531(23)00211-1/sref56http://refhub.elsevier.com/S2162-2531(23)00211-1/sref56http://refhub.elsevier.com/S2162-2531(23)00211-1/sref56http://refhub.elsevier.com/S2162-2531(23)00211-1/sref57http://refhub.elsevier.com/S2162-2531(23)00211-1/sref57http://refhub.elsevier.com/S2162-2531(23)00211-1/sref57 A neutralizable dimeric anti-thrombin aptamer with potent anticoagulant activity in mice Introduction Results Design of M08s-1-based bivalent anti-thrombin aptamers Affinity evaluation of dimeric anti-thrombin aptamers Evaluation of the anticoagulant activity of M08s-1-based bivalent aptamers by activated partial thromboplastin time (aPTT)  ... Assessment of anticoagulant activity of the bivalent aptamers in vivo using aPTT Serum stability of bivalent anti-thrombin aptamers Neutralization of the anticoagulant activity of the bivalent aptamers Discussion Materials and methods Inhibitory activity analysis of the aptamers in thrombin and fibrinogen mixed solution Affinity analysis of anti-thrombin aptamers using SPR aPTT assay aPTT in spiked plasma aPTT in plasma from aptamer-injected mice Injection into mice and collection of plasma samples PT assay in plasma from aptamer-injected mice Pharmacodynamics analysis Serum stability assessment of anti-thrombin aptamers Neutralization analysis of the anticoagulant activity using aPTT in plasma Data and code availability Supplemental information Acknowledgments Author contributions Declaration of interests References