Theoretical, biological, and electrochemical explanation of the complexity of Peramivir’s drug–DNA interaction Authors: Monika Wypych, Anna Ignaczak, Paulina Tokarz, Jagoda Seroka, Kamila Koszelska, Sylwia Smarzewska Published: 2025 ABSTRACT AbstractIn this study, an in-depth investigation was carried out on the interaction between the anti-influenza drug peramivir (PMV) and various DNA, including double-stranded DNA (dsDNA), single-stranded DNA (ssDNA), polyadenylic acid (PolyA), polyguanylic acid (PolyG) and plasmid pUC19, employing square-wave voltammetry (SWV), spectrophotometry, plasmid relaxation assays, and density functional theory (DFT). Voltammetric measurements demonstrated that peramivir significantly influences the unwinding of the DNA double helix. The data also suggest that PMV binds within the DNA groove and interacts with nucleobases, showing a stronger affinity toward guanine. These findings were confirmed by the results of DFT calculations, which additionally demonstrated the importance of PMV-nucleobase hydrogen bond formation and interactions of this drug with the sugar-phosphate backbones of DNA for the stabilization of DNA:PMV complexes. Complementarily plasmid relaxation assays indicate that PMV does not cause significant DNA damage such as single- or double-strand breaks, but also does not provide any protection against oxidative DNA damage. These findings imply that the drug’s mechanism of action is centred more on molecular interaction with DNA rather than on compromising its structural integrity.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-28369-9.Keywords: Peramivir, Influenza, DNA interactions, Voltammetry, DFTSubject terms: Chemistry, DNA Introduction The DNA double helix, found in all living organisms, plays a crucial role in essential cellular processes such as replication, transcription, and cell division1,2. However, its structure is vulnerable to the damage from endogenous and exogenous factors, including pathogens like influenza viruses3,4. There are four types of influenza viruses: A, B, C, and D. The dominant types, A and B, are responsible for the majority of influenza cases within the population5,6. These viral infections primarily target the immune and respiratory systems, commonly causing symptoms such as rhinitis, sore throat, cough, and fever. Additional symptoms may include muscle pain, headaches, and general fatigue. In severe cases, hospitalization may be required, and untreated influenza can be fatal4,5,7–9. Consequently, advancing research on anti-influenza drug development and their therapeutic effects remains a priority in the fight against influenza. Peramivir (PMV; (1 S,2 S,3R,4R)-3-[(1 S)-1-(Acetylamino)-2-ethylbutyl]-4-(carbamimidoylamino)-2-hydroxycyclopentanecarboxylic acid), is an anti-influenza drug, which belongs to the class of neuraminidase inhibitors (NAIs)10,11. Influenza viruses spread through the action of glycoproteins, including neuraminidase (NA) and hemagglutinin. The hemagglutinin protein enables the virus to enter the cells of a host organism, while NA facilitates the release of the virus from infected cells, thereby propagating the infection11. In the literature there are only few studies about PMV interactions, focused mainly on its binding to enzymes12–16. According to recent tests conducted on mice, the anti-influenza drugs oseltamivir and amantadine exert a pathogenic effect on the DNA content of hepatocytes in treated animals17. The interactions of oseltamivir with DNA have been earlier investigated using spectroscopic and computation methods18. It has been shown that the complexation of the drug by DNA occurs via Van der Waals and hydrogen bonds and the drug is bound via intercalation between the base pairs. However, there is a lack of literature addressing the impact of anti-influenza drugs on single- and double-stranded DNA and whether any differences exist between them. Given the fact of constant mutations and changes occurring in the structure of viruses, sustained research into interactions and monitoring of the obtained results are essential to enhance the efficacy of appropriate pharmacological agents. Identification of the binding mechanism between drug molecules and various forms of DNA could, in the future, contribute not only the development of novel and the refinement of existing anti-influenza drugs, but also to the advancement of therapeutics aimed at augmenting immune responses or targeting other virus-induced diseases. Over the last few decades, our understanding of DNA interactions with small molecules has significantly increased. Research has focused on elucidating the mechanisms by which chemical compounds bind to nitrogenous bases. Many of these studies have led to the development of various drugs, including those with anti-cancer properties. The literature suggests two main modes of a drug molecule binding to DNA: covalent and non-covalent. Furthermore, the non-covalent mode can occur in three ways: intercalation, groove-binding, and external binding16,19–21. Various techniques are available to study DNA interactions, including chromatography22–24, spectroscopy25–28, spectrometry29–32, fluorometry33–36, and electrochemistry37–39. Among the latter, voltammetry is the most frequently used electrochemical technique. Voltammetric measurements are an excellent tool for studying interactions due to their speed, simplicity, and sensitivity. Electrochemical methods offer several ways to investigate interactions, with the most common being incubation40,41, which allows for the examination of the effect of time on the interaction between DNA and substance molecules. Simple electrochemical DNA biosensors41,42 can also be designed to detect drug-DNA binding processes. Moreover, during analysis, it is possible to monitor changes in the oxidation signals of individual nitrogenous bases present in DNA, enabling the determination of interactions with compounds43,44. Electrochemical methods have been regularly employed to study interactions, demonstrating high sensitivity and accuracy in detecting even minor structural disturbances and oxidative damage to DNA45–47. All of the described achievements in the field of DNA interaction research constitute an extremely valuable resource. Based on documented discoveries and established analytical methods, it is possible to effectively design experimental approaches, depending on the specific objectives. Despite the extensive research efforts on the interactions between individual drug and various forms of DNA or single nitrogenous bases, it remains essential to pursue further investigations involving novel and structurally different compounds. In the context of drugs, such efforts bring significant benefits, contributing to the improvement of pharmaceutical quality. Ongoing studies and systematic documentation in this area drive scientific progress toward the development of solutions that ensure the effective performance of therapeutic agents. Furthermore, explaining the mechanisms underlying drug interactions and specific DNA forms may facilitate the design of future drugs with improved efficacy and reduced adverse effects. In the long term, these advancements not only support people’s health but also benefit the entire pharmaceutical industry, as research progress may reduce the likelihood of drugs being suddenly withdrawn or having their usage restricted. To the best of our knowledge, voltammetric studies on the interaction between PMV and DNA have not been performed to date. In the current study, the interactions between the anti-influenza drug PMV and dsDNA, ssDNA, Polyadenylic acid (PolyA), and Polyguanylic acid (PolyG) were examined using square-wave voltammetry (SWV) with a glassy carbon electrode (GCE) as the working electrode. In voltammetric studies, interactions were examined in two ways. The first method involved determining the effect of increasing PMV concentration on ds/ssDNA. Subsequent studies involved incubating DNA and PMV molecules to observe the course of interaction over time. The second method focused on studying interactions between single nitrogenous bases, using PolyA and PolyG. The measurements were carried out using the developed biosensors, which were incubated in the presence of PMV. To obtain a comprehensive understanding of the phenomena in the electrochemical system, the strength and nature of interactions between PMV and DNA were investigated at the molecular level using computational chemistry methods. The complexation and interaction energies for several different configurations of the dsDNA:PMV complexes in water were evaluated using density functional theory (DFT). In addition, to estimate and analyze the strength of interactions of PMV with DNA single strands and with individual nucleobases in dsDNA:PMV, the DFT calculations were performed for such systems extracted from the complexes. Plasmid relaxation assays were also conducted to determine whether PMV causes single-strand breaks (SSBs) and double-strand breaks (DSBs) in DNA or provides any protection against oxidative damage. Experimental Apparatus An µAutolab type III (Metrohm-EcoChemie, Netherlands) potentiostat combined with an M164 electrode stand (mtm-anko, Poland) controlled by a computer with GPES software (version 4.9) and µStat 300 (Metrohom-DropSens, The Netherlands) potentiostat with M164 electrode stand (mtm-anko, Poland) controlled by DropView 8400 Software (version 3.78) was used for voltammetric studies. The three-electrode system consisted of glassy carbon electrode (GCE) as the working electrode (BASi, USA; diameter: 3 mm), reference electrode Ag/AgCl (3 mol L− 1 KCl), and Pt wire as the auxiliary electrode. All measurements were performed in a standard 10 mL voltammetric cell. For all voltammetric measurements, a frequency of 25 Hz, an amplitude of 0.025 V, and a step potential of 0.005 V were used. The surface of the working electrode was polished on an alumina slurry each time before the next measurement. After that, the surface of the GCE was carefully rinsed with distilled and deionized water. Additionally, between measurement series, the electrode was cleaned using ultrasounds for 5 min. A Cary 100 Bio UV-Vis spectrophotometr (Aglient, USA) was used for spectrometric studies. The wavelength range of all spectrophotometric measurements was 190–800 nm. A pH/mV/ion meter (Elmetron, Poland) was used to examine the pH of the prepared buffer solution. Room temperature (20–22 °C) was upheld during all experiments. All measurements were repeated three times. Chemicals and reagents Peramivir, double-stranded salmon sperm DNA, Polyadenylic acid, Polyguanylic acid, 2,2-Diphenyl-1-picrylhydrazyl (DPPH), and phosphate-buffered saline tablets (PBS) were obtained from Sigma-Aldrich (Germany). The standard solution of PMV was prepared by dissolving an appropriate amount of the analyte in distilled water. In present work, PBS of pH = 7.4 was used as the supporting electrolyte for ds- and ssDNA interaction studies. For PolyA and PolyG interaction studies, acetate buffer, pH = 4.5, was used. This buffer was prepared by mixing specific volumes of pre-prepared solutions containing the appropriate concentrations of sodium acetate (Chempur, Poland) and acetic acid (POCH, Poland). The DPPH standard solution was prepared by weighing the appropriate amount of the substance and then dissolving it in the DMSO (Avantor, Poland). For the dsDNA standard solution preparation, the required mass of the compound was dissolved in PBS. ssDNA was prepared by performing a denaturation process. The previously prepared dsDNA standard solution was heated in a water bath at a temperature of approximately 90–95 °C for 10 min, and then rapidly cooled in an ice bath47. PolyA and PolyG stock solutions were prepared by dissolving the appropriate amount of compounds in distilled water. Solutions were kept refrigerated at 4 °C when not in use. For the purpose of the spectrophotometric studies, solutions of: ds/ssDNA (c(ds/ssDNA) = 20 mg mL− 1), DPPH (c(DPPH) = 1.0 · 10− 6 mol L− 1) and a mixture containing ds/ssDNA (c(ds/ssDNA) = 20 mg mL− 1), peramivir (c(PMV) = 1.0 · 10− 6 mol L− 1) and DPPH (c(DPPH) = 1.0 · 10− 6 mol L− 1) were made. Incubation procedure Procedure 1 – ds/ssDNA interactions with peramivir SWV measurements of interactions were performed in the presence of ds/ssDNA (c(ds/ssDNA) = 20 mg mL− 1) and peramivir (c(PMV) = 1.0 · 10− 3 mol L− 1) in PBS at pH = 7.4 as a supporting electrolyte. PBS was placed in the 10 mL voltammetric cell, and then the blanks were recorded. Interactions were examined in two ways. First, the effect of time on the interaction between dsDNA and PMV was examined. For this purpose, an addition of appropriate volume of the dsDNA solution was placed in a measuring cell using a micropipette. Then, after recording the voltammograms for dsDNA, a constant volume of peramivir was added. Measurements were taken at intervals of 30 min, and a total time of incubation was 2 h. The next step was to examine the effect of increasing PMV concentration on dsDNA. In order to do this, appropriate volumes of anti-influenza drug solution were added to the cell containing the DNA, and then voltammograms were recorded. All procedures were repeated for ssDNA. Procedure 2 – PolyA/G interactions with peramivir SWV was used in the case of PolyA/PolyG-PMV interactions. The supporting electrolyte, which was an acetate buffer (pH = 4.5), was placed in a 10 mL voltammetric cell, and then the blanks were measured. The voltammograms for PolyA and PolyG were recorded after addition of appropriate volumes of solutions of mentioned polymerized nitrogen bases (c(PolyA/PolyG) = 5 mg mL− 1). PolyA/G-PMV interactions were studied using electrochemical biosensors. Before starting the measurements, the electrode was thoroughly cleaned. 5 µL of PolyA solution (c(PolyA) = 50 µg mL− 1) was applied to the electrode surface, waiting for the electrode to dry completely. The electrochemical biosensor prepared in this way was incubated in a cell containing acetate buffer (pH = 4.5) and PMV (c(PMV) = 1.0 · 10− 3 mol L− 1). Incubation was carried out for different times: 60, 30 and 15 min. After the appropriate time, the electrode was removed and carefully washed with distilled water to remove unbound drug molecules. Then, the biosensor was immersed in a cell containing only the supporting electrolyte and voltammograms were recorded. The preparation of the biosensor and measurements for the PolyG were performed in an identical manner. Plasmid relaxation assay The pUC19 plasmid was isolated from the DH5α E. coli cells with AxyPrep Plasmid Miniprep Kit (Axygen Scientific, Inc., Union City, CA, USA) according to the manufacturer’s instructions. The native plasmid, mainly in covalently closed circular (CCC) form, was digested with the restrictase PstI according to the manufacturer’s protocol (New England Biolabs, Ipswich, MA, USA) to induce linear (L) form as a positive control. The plasmid (150 ng) was incubated with peramivir (PMV) for 24 h at 37 °C to assess the potential of the compound to induce DNA breaks. The induction of oxidative DNA damage in the plasmid was conducted by the Fenton reaction with minor modification (Lloyd et al., 1999). The plasmid (150 ng) was pre-incubated with copper(II) sulfate (50 · 10− 6 mol L-1) or ammonium iron(II) sulfate (75 · 10− 6 mol L-1 for 15 min at 37 °C followed by the treatment with hydrogen peroxide (H2O2) at 75 mM for 1 h at 37 °C. The treatment induced a DNA single-strand break resulting in the relaxation of supercoiled (CCC) plasmid to open circular (OC) form. Structural differences between CCC, OC and L plasmid forms account for their different electrophoretic mobility. Protection against DNA strand breaks induced by the Fenton reaction was employed to assess the antioxidative properties of PMV. Briefly, PMV was dissolved in H2O to a stock concentration of 15 mM. Supercoiled plasmid (150 ng) was pretreated with PMV for 15 min at 37 °C, followed by the Fenton reaction. Immediately after incubation, bromophenol blue dye was added, and the samples were subjected to 1% agarose gel electrophoresis followed by ethidium bromide staining. The visualization was conducted under UV light (302 nm), scanning by a CCD camera. Densitometric analysis was performed with the GeneTools by Syngene (Cambridge, UK) software. Signal intensity was calculated as area under a peak. Computational studies The theoretical approach used in the present work was similar to this applied in our previous study on the interactions of the anticancer drug ponatinib with dsDNA48. First, a conformational search was performed to find the lowest energy conformer of PMV in water. This was achieved by applying the three-stage procedure presented in the flow chart Fig. S5. Briefly, in Stage 1, several thousand structures were crated in the Hyperchem program49 by randomly changing thirteen selected torsion angles (Fig. S6). All these structures were fully optimized with the molecular mechanics force field Amber99. In Stage 2, all structures were re-optimized without any constraints in vacuo using the semiempirical method PM7 combined with the MOZYME procedure available in the MOPAC program50. Additionally, a short semiempirical molecular dynamics simulation was performed in the Gabedit program51 to verify whether any lower energy structure can be found. From all structures obtained in Stage 2, 1000 conformers having the lowest PM7 heat of formation were selected. In Stage 3, they were fully re-optimized using the density functional theory (DFT) method M062X-GD352,53 with the basis set 6-31G(d,p) and the Polarizable Continuum Model (PCM) of water54. The DFT calculations were performed using the Gaussian 16 package55. At this stage, the PMV structure with the lowest DFT energy was selected to be used in further studies. More detailed description of this method can be found in the Supporting Information to ref48. (section Procedure S1). In the calculations performed for the dsDNA:PMV complexes the same two dsDNA models were used as in the work on the dsDNA:ponatinib complexes. [Fig. 8 in ref48]. They were based on two double helix B-DNA dodecamers having different sequences, taken from the RCSB Protein Data Bank, namely 1BNA (5’-D(*CP*GP*CP*GP*AP*AP*TP*TP*CP*GP*CP*G)-3 3’))56 and 119D ((5’-D(*CP*GP*TP*AP*GP*AP*TP*CP*TP*AP*CP*G)-3 3’))57. Their structures were obtained by terminating these dodecamers with hydrogens and optimizing them with the PM7/MOZYME method in water represented by the implicit COSMO model58. For each dsDNA model, four sites for binding of PMV were considered: intercalation (InC), external binding (ExB), major groove (MaG) and minor groove (MiG) (Fig. 1a). As in the previous work, the procedure used in the research on the interaction of PMV with dsDNA included three steps and its general concept is presented in Fig. 1b, while technical details can be found in the ESI to ref48. (section Procedure S2). The final structures were obtained from the partial optimizations performed with the M062X-GD3/6-31G(d,p) method in water (PCM) for the complexes selected after the semiempirical PM7 calculations. In the optimization, the PMV molecule was fully relaxed, while the dsDNA was frozen. (a) Schematic picture of the four binding sites in dsDNA considered in the calculations. (b) The flow chart of the procedure used to explore stability of the dsDNA:PMV complexes. From the DFT calculations, the complexation and interaction energies ( and , respectively) were calculated according to the formulas: where is the energy of the partially optimized complex dsDNA: PMV, - the energy obtained from the single point (SP) DFT calculations performed for the initial dsDNA structures, - the DFT energy of the most stable conformer of PMV obtained in water (PCM), and - the energy from the single point calculations performed for the PMV structure with its geometry in a given dsDNA:PMV complex. Additionally, the strength of PMV interactions with each of two DNA strands in the MaG and MiG complexes was evaluated. From the final structure of each dsDNA:PMV complex, two ssDNA: PMV systems were created by removing the first or second DNA strand. In their case, the interaction energy was calculated using the formula: where and are the energies of ssDNA: PMV and ssDNA, respectively, obtained from the single point DFT calculations performed for these systems in their geometries taken from the dsDNA:PMV complex. A similar approach was used to estimate the interaction energies of PMV with individual nucleobases in the dsDNA:PMV complexes. In this case, the selected nucleobase alone and PMV were extracted from the complex. The former connection of the nucleobase to DNA was replaced with the hydrogen atom with a default N-H distance 1.01 Å. For such pairs of molecules, the interaction energies were obtained from the single point DFT calculations as: where ‘nb’ stands for Ade, Gua or thymine (Thy). Results and discussion Voltammetric studies of the interaction between ds/ssDNA and peramivir Initial studies were aimed at preparing a comprehensive electroanalytical characterization of peramivir. For this purpose, a series of voltammetric measurements were carried out, in which the signals from the oxidation and reduction of the compound were analyzed in a wide pH range (measurements were performed at both, positive and negative potential sweep directions). Based on the results obtained, it was concluded that peramivir is an electrochemically inactive compound. The interaction studies between ds/ssDNA and peramivir were performed using square wave voltammetry. SWV technique was chosen due to its numerous possibilities in studying DNA interactions, as well as the speed of measurements, their accuracy and precision41–47. Studies were conducted using PBS buffer at a physiological pH of 7.4. The choice of this pH was based on the fact that the pH of healthy human body fluids ranges between 7.35 and 7.45, with an average of 7.40. This pH allows for the proper functioning of many biological processes59, and therefore, it is also suitable for conducting interaction studies. Considering that peramivir is an electrochemically inactive substance, the measurements were carried out in the potential range of 0 –1.5 V, where oxidation peaks from guanine (Gua) and adenine (Ade) are observed. First, the electrochemical behavior of ds/ssDNA in PBS buffer pH 7.4 was examined giving two signals visible in the SW voltammograms (data not shown). The small peak ca. E = + 0.9 V was attributed to the guanine, and the second one ca. E = + 1.2 V to the adenine oxidation41,42. In the next step the study of the influence of the incubation time on the ds/ssDNA:peramivir interaction was conducted. Appropriate volumes of DNA (c(ds/ssDNA) = 20 mg mL− 1) and peramivir (c(PMV) = 1 · 10− 3 mol L− 1) were added to the cell with the supporting electrolyte (PBS buffer pH 7.4). The total incubation time was 2 h. The electrode was cleaned after each measurement, and polished using a polishing kit and alumina slurry. First, the effect of time on the interaction of dsDNA and peramivir was examined (Fig. 2a). Over time, an increase in the current of signals coming from guanine was observed, however, after 1 h the peak heights start decreasing and the potential values tend to become slightly more positive. The opposite situation was observed in the case of signals coming from adenine. Initially, the Ade peak height decreases, but start increasing after 1 h. Figure 2b presents the influence of incubation time on the ssDNA:PMV interaction. In this case Gua peak current decreases, while Ade signals increases throughout the entire incubation time range. The obtained results suggested that additional time is needed to establish the proper interaction between dsDNA and PMV. Therefore, it was checked whether the signals would change after an additional 30 min (total time of incubation 2.5 h), but no signal shifts or changes were observed. The initial, gradual increase in the height of guanine peaks may be related to DNA conformational changes and the unwinding of the double helix under the influence of presence of peramivir in the solution. Based on this, it can be concluded that 1 h is sufficient time to unwind the DNA double helix. After this time, the molecules of peramivir and DNA can be bound, as evidenced by signals changes. In the case of ssDNA:PMV interaction, a decrease in peaks coming from Gua is visible. This happens because the double helix is already unwinded, so the binding of the molecules can occur immediately. For both dsDNA:PMV and ssDNA:PMV changes in the peak potential values are minimal indicating that a groove-binding interaction occurs between the molecules. (a,b) SW voltammograms for dsDNA (a)/ssDNA (b) (c(ds/ssDNA) = 1.0 · 10−2 mg mL− 1) and PMV (c(PMV) = 30.0 · 10− 6 mol L− 1) during the 2-hour incubation; 0 min (1), 30 min (2), 1 h (3), 1.5 h (4), 2 h (5). Supporting electrolyte: PBS pH 7.4. The experimental conditions: frequency of 25 Hz, amplitude of 0.025 V, step potential of 0.005 V (c,d) (c) SW voltammograms for dsDNA (1) (c(dsDNA) = 1.0 · 10−2 mg mL− 1) and various concentration of PMV (20.0 (2), 30.0 (3), 50.0 (4) · 10− 6 mol L− 1). (d) SW voltammograms for ssDNA (1) (c(ssDNA) = 1.0 · 10−2 mg mL− 1) and various concentration of peramivir (10.0 (2), 30.0 (3), 40.0 (4), 50.0 (5) · 10− 6 mol L− 1). Supporting electrolyte: PBS pH 7.4. The experimental conditions: frequency of 25 Hz, amplitude of 0.025 V, step potential of 0.005 V. In order to support the conclusions drawn, and to carefully investigate PMV interaction, the presence of increasing amount of peramivir on the height and potential of guanine and adenine signals of dsDNA and ssDNA was examined. For this purpose, gradually increasing volumes of peramivir solution were added to ss/dsDNA. In the case of dsDNA:PMV interaction (Fig. 2c), there is an increase in Gua peaks with increasing drug concentration in the cell. Unlike in the incubation-based studies, now the changes in the adenine peaks are clear and of a similar current magnitude. With increasing concentration of peramivir in the cell, a gradual decrease in Ade signal heights can be observed, without a shift in potentials. Based on the presented results, it can be concluded that in the initial stage of interaction, a double helix is unwinding under the influence of presence of PMV. By increasing the concentration of the drug in the cell, the unwinding of the double helix is accelerated, which results in an increase in the peak current for guanine. Moreover, the obtained voltammograms may also confirm the ongoing conformational changes in the dsDNA structure. The gradual increase in the signal from guanine is likely a consequence of the molecule’s arrangement relative to the electrode, facilitating easier oxidation of Gua. In the case of adenine, the observed gradual reduction in peak current with increasing PMV concentration may indicate that PMV also interacts with this nitrogenous base. Significant differences in the structure of ds- and ssDNA affect their electrochemical behavior in the presence of peramivir. The dsDNA double helix must be unraveled before the actual interaction can occur, which is not the case for the already untied single-stranded DNA. For this reason, the behavior of ssDNA in the presence of increasing volumes of the drug was also examined. After adding PMV to the electrolyte containing ssDNA a notable decrease in the height of guanine peak can be observed (Fig. 2d). Subsequently, with increasing peramivir concentration, further decreases in the Gua signal height are visible. There are also changes in the adenine peak. Slight changes in the signal height were also visible in the previously performed study (incubation procedure) of the interaction of ssDNA and peramivir (Fig. 2b). The visible differences may indicate that PMV interacts not only with guanine, but also with adenine. However, in all the measurement series conducted thus far, no similar pattern in the changes of the Ade signal was observed, as seen with Gua. This might suggest that peramivir interacts with both nitrogen bases, but shows stronger bindings towards guanine. Voltammetric studies of the interaction between PolyA/PolyG and peramivir The earlier results revealed some ambiguity regarding the changes in signals from adenine. Specifically, while time-dependent studies did not show a clear correlation in the adenine peak, experiments examining the effect of peramivir concentration revealed distinct trends: a decrease in signal intensity with dsDNA and an increase with ssDNA interactions, prompting further investigation. Therefore, the direct interaction of peramivir with nitrogenous base polymers was also explored. Analyzing the drug’s effect on nitrogenous base polymers not only helps clarify the previously observed results but also provides a valuable approach for confirming these interactions and understanding the mechanisms by which the molecule binds to different DNA fragments41,60–62. For this purpose, SWV was applied, and an acetate buffer with pH 4.5 was used as the supporting electrolyte, as this pH significantly enhances the signals obtained from nitrogenous bases such as PolyA and PolyG62,63. First, voltammograms for PolyA and PolyG (c(PolyA/PolyG) = 5 mg mL− 1) were recorded revealing signals at EGua = + 1.12 V and EAde = + 1.38 V. The next step was to prepare PolyA– and PolyG–biosensors to study the interactions with peramivir (Experimental section: Procedure 2 – PolyA/G interactions with peramivir). Initially, voltammograms were recorded for the prepared biosensors without incubation in the peramivir solution to determine the height and potential of signals from polyadenine and polyguanine. In both cases, single signals were observed, with the PolyA peak at E = + 1.39 V and the PolyG peak at E = + 1.13 V. Subsequently, the biosensors were incubated for 15, 30, 60 min in a supporting electrolyte containing peramivir. Next, after careful rinsing with distilled water, SW measurements were performed after immersion of the biosensors in a cell containing bare acetate buffer. The recorded voltammograms for the PolyA/G:PMV interactions confirm previously obtained results. The signal from polyadenine increases significantly with increasing incubation time (Fig. 3a). The opposite situation is visible in PolyG peaks (Fig. 3b). The current of the signals decrease with longer incubation time. Moreover, after incubation, a second signal is visible (E = + 0.83 V). There is probability that the peak comes from Gua, and peramivir may affect the change of the polymer into the pure form of guanine. (a,b) SW voltammograms for PolyA (a) and PolyG (b) biosensors (c(PolyA/PolyG) = 5 µg mL− 1) and PMV (c(PMV) = 3.0 · 10− 5 mol L− 1) during the 1-hour incubation; 0 min (1), 15 min (2), 30 min (3), 1 h (4). Supporting electrolyte: acetate buffer pH 4.5. The experimental conditions: frequency of 25 Hz, amplitude of 0.025 V, step potential of 0.005 V. The results of these studies align with previous findings on the interaction between ssDNA and peramivir. As the double helix unwinds, the drug molecules gain direct access to the nitrogenous bases, similar to how the developed biosensor allows peramivir to bind with polyadenine or polyguanine deposited on the electrode surface. The observed changes in the signal intensities for adenine, guanine, and their respective polymers further indicate that peramivir interacts with both nitrogenous bases. Biological studies on the interaction between plasmid DNA and peramivir Voltammetric analysis demonstrated that PMV enhanced the unwinding of the DNA helix, followed by its interaction with purine bases Since ssDNA is more prone to damage than dsDNA63, we investigated whether peramivir could induce SSBs and DSBs in DNA. To evaluate this, a plasmid relaxation assay was conducted, where plasmid was incubated with increasing concentrations of peramivir for 24 h. The treated samples were then resolved on agarose gel (Fig. 4a). The native plasmid predominantly existed in the covalently closed circular (CCC) conformation (Fig. 4a, lane 1). As a positive control, the plasmid was digested with restriction enzyme PstI, which recognizes a specific site in pUC19 plasmid and generates a linear (L) form via a single DSB (Fig. 4a, lane PstI). Increasing concentrations of peramivir did not alter DNA plasmid conformation indicating that peramivir does not induce DNA breaks in the concentration range up to 1.0· 10− 3 mol L-1. Subsequently, the antioxidant potential of peramivir was assessed by examining its ability to protect DNA from damage induced by hydroxyl radicals generated through the Fenton reaction. Oxidative DNA damage results in the formation of SSBs and DSBs yielding open circular (OC) and L plasmid conformations, respectively. We observed that the native plasmid was predominantly in the CCC form (Fig. 4b, lane 1). Two positive controls were included: (i) plasmid treated with PstI generating L form (Fig. 4b, lane PstI), and (ii) plasmid treated with Fenton’s reagents (Fe2++H2O2 or Cu2++H2O2), leading to OC form, indicative of a single SSB (Fig. 4b, lane 2). To assess peramivir’s protective effects against DNA nicking, plasmid DNA was treated with Fenton’s reagents in the presence of increasing concentrations of peramivir (31.25–1000 · 10− 6 mol L-1). Plasmid relaxation assay. pUC19 plasmid DNA was resolved on a 1% agarose gel, stained with ethidium bromide and visualized in UV light. Structural differences between covalently closed circular (CCC), open circular (OC) and linear (L) forms of the plasmid accounted for their different electrophoretic mobility. (a) Lane 1 – negative control (pUC19 plasmid); lanes 2–8 pUC19 plasmid incubated with peramivir at indicated concentrations for 24 h at 37 °C; lane 9 – positive control of L form (the plasmid incubated with PstI restriction enzyme for 24 h for the induction of DNA double–strand break). (b) Lane 1 – negative control (pUC19 plasmid); lane 2 – positive control (the plasmid exposed to 50 · 10− 6 mol L− 1 CuSO4 or 75 · 10− 6 mol L− 1 (NH4)2Fe(SO4)2 for 15 min at 37 °C followed by the treatment with 75 · 10− 3 mol L− 1 H2O2 for 1 h at 37 °C, Fenton reaction); lanes 3–8 – pUC19 plasmid incubated with peramivir (PMV) at indicated concentrations for 15 min at 37 °C followed by Fenton reaction; lane 9 – positive control of L form (the plasmid incubated with PstI restriction enzyme for 24 h for the induction of DNA double-strand break). Original gels are presented in Fig. S1-S3 in the ESI. The results showed that peramivir did not prevent oxidative DNA damage, as evidenced by the lack of change in the plasmid’s CCC and OC forms when compared to plasmid treated solely with Fenton’s reagents (positive control). Densitometric analysis of resolved plasmid DNA (Fig. 5; Table 1) confirmed that peramivir exposure did not affect the distribution of CCC and OC plasmid forms. Therefore, we conclude that peramivir does not induce DNA breaks nor does it exhibit protective effects against oxidative DNA damage under the tested conditions, indicating its inert behaviour toward DNA integrity. Plasmid relaxation assay. Densitometric analysis of pUC19 plasmid resolved on a 1% agarose gel, stained with ethidium bromide and visualized in UV light. Plasmid conformations, covalently closed circular (CCC), open circular (OC) and linear (L) forms, are presented as peaks. Raw 1 – negative control (pUC19 plasmid); raw 2 – positive control (the plasmid exposed to 50 · 10− 6 mol L-1 CuSO4 or 75 · 10− 6 mol L-1 (NH4)2Fe(SO4)2 for 15 min at 37 °C followed by treatment with 75 mM H2O2 for 1 h at 37 °C, Fenton reaction); raws 3–8 – pUC19 plasmid incubated with peramivir (PMV) at indicated concentrations for 15 min at 37 °C followed by Fenton reaction; raw 9 – positive control (the plasmid incubated with PstI restriction enzyme for 24 h for the induction of DNA double-strand break). Table 1Densitometric analysis of pUC19 plasmid pre-incubated with peramivir (PMV) followed by Fenton reaction and resolved on agarose gel.CompoundOC form of DNA (%) ± SEMOC form of DNA (%) ± SEMCu2+ + H2O2Fe2+ + H2O2PMV (µM)021.13 ± 2.17032.50 ± 3.3131.2518.01 ± 2.0131.2530.60 ± 4.0962.519.24 ± 1.8662.534.64 ± 2.8712519.48 ± 2.0412534.49 ± 3.9225017.45 ± 1.7725034.12 ± 3.5350018.57 ± 1.9350029.40 ± 3.28100018.96 ± 2.08100030.32 ± 4.16The open circular (OC) form of DNA was quantified relative to total DNA in the lane. Densitometric analysis of pUC19 plasmid pre-incubated with peramivir (PMV) followed by Fenton reaction and resolved on agarose gel. The open circular (OC) form of DNA was quantified relative to total DNA in the lane. Spectrophotometric studies of the interaction between ds/ssDNA and peramivir To confirm the lack of antioxidant properties, additional, complementary spectrophotometric research were performed. For this purpose, 2,2-Diphenyl-1-picrylhydrazyl (DPPH), a reagent used to study antioxidant activity, was used. The spectrophotometric studies on this subject involved the absorption measurements of peramivir, DPPH (c(DPPH) = 1.0 · 10− 6 mol L− 1) and DNA. Initial studies allowed to establish that the antiviral drug does not give any signals, while in the case of DNA, for both dsDNA and ssDNA, a clear signal is visible at a wavelength of 260 nm. DPPH signals are visible at wavelengths 335 nm and 517 nm. Subsequently, it was examined how the peaks change when PMV and DPPH solutions are added to the solution of double- or single-stranded DNA (Fig. S4 in the ESI). The recorded spectra exhibit differences at wavelength range 350–550 nm, with significant increment of the signals originating from DNA. Such changes in the signals may indicate changes in the DNA structure, both structural and conformational, as a result of DPPH reduction, where peramivir plays the role of a mediator of this reaction. It is likely that PMV mediates electron transfer from DNA to DPPH, thus increasing the DNA signal by increasing its concentration in the active form. Additionally, the signals for ssDNA are significantly larger than for dsDNA, which may be due to the easier access of the molecules to a single helix. DFT studies of the interactions between DNA and Peramivir The research on the interaction of ssDNA and peramivir allowed to determine that the anti-influenza drug unwinds the double helix first, and only then the actual interaction of molecules occur. Nevertheless, the studies of the dsDNA:PMV interaction did not give unequivocal results, only suggesting the way in which interaction could have occurred. Therefore, it was decided to conduct additional studies, using for this purpose the computational approach. The results obtained for the four tested configurations of dsDNA:PMV clearly showed that the intercalation of PMV into dsDNA is very unlikely to occur. For both dsDNA models, no stable InC structure was obtained after the PM7 optimization, because PMV was always expulsed and the complex adopted the MaG configuration. This was to be expected, since the most stable conformer of PMV found in this work (Fig. 6) does not have any major planar fragment that could fit between the DNA base pairs. Its spatial structure is determined by various intramolecular interactions, including two hydrogen bonds, both formed by hydroxyl groups present in this molecule. PMV also has other hydrogen bond donors and acceptors, which, as will be shown below, play an important role in its binding by DNA. The most stable conformer of PMV indicated by the DFT results after the conformational search. Atom colors in (a): carbon – cyan, nitrogen – dark blue, oxygen – red, hydrogen –grey. The dashed lines mark the intramolecular hydrogen bonds. The complexation energies obtained for three other dsDNA:PMV configurations according to the formula (1) are presented as darker bars in Fig. 7a. Their values for the MaG and MiG sites obtained with the two dsDNA models oscillate around − 50 kcal mol− 1, which indicates similarly high stability of these configurations. The energy profit from binding PMV in the ExB site depends on the dsDNA model: in the 1BNA ExB complex it is much smaller than in the 119D ExB complex, but still significant. As can be seen in Fig. S7 in the ESI, in both these configurations the drug interacts only with the sugar-phosphate backbones of dsDNA. The large difference in for the two ExB models can be due to the fact that in 1BNA ExB the PMV molecule interacts with only one DNA strand, while in 119D ExB – with both strands. (a) The complexation energies (darker bars) and their corrected values (lighter bars) obtained for the dsDNA:PMV from the DFT calculations performed for two different models of dsDNA (1BNA, 119D) and three binding sites (ExB, MaG, MiG). (b) A comparison of the interaction energies of PMV with dsDNA and with the 1st and 2nd DNA strand (for their numbering see Fig. 8) in the dsDNA:PMV complexes. The numerical data is given in Tables S1 and S2 in the ESI. It should be remembered, however, that the initial dsDNA geometry was based on a crystallographic structure pre-optimized only with the semiempirical method. The values of should therefore be considered as rough estimates reflecting some trends. Indeed, single-point DFT calculations performed on dsDNA molecules extracted from the six dsDNA:PMV complexes tested showed that some of them had energies several kilocalories per mole lower than those of the initial dsDNA structures. Therefore, all dsDNA structures extracted from the complexes were re-optimized in water using the semiempirical PM7/COSMO method, and the DFT energies were calculated for the obtained geometries. Among them, the structures 1BNA and 119D with the lowest energies were selected (Fig. S8 in the ESI). Their energies (denoted as in Table S1 in the ESI) were used as the reference values in formula (1) to obtain the new, corrected complexation energies . They are shown as lighter bars in Fig. 7a. This correction does not change the overall trends, but decreases the absolute values of the complexation energy, which for the MaG and MiG complexes oscillate now around − 40 kcal mol− 1. However, these are still only approximate results, because for such a large system neither the dsDNA molecule nor the dsDNA:PMV complex could be fully optimized at the DFT level. The uncertainty about the final values described above does not occur in the case of interaction energies, the calculation of which is based on the structures present in a given complex. The most important are the MaG and MiG configurations, as in them PMV interacts with nucleobases (Fig. 8). Therefore, for these complexes the energies of interaction of PMV with dsDNA were analyzed (bars dsDNA in Fig. 7b). The very strong dsDNA:PMV interaction is found in the 1BNA MaG complex, followed by the 119D MiG complex. The difference in the trends observed in the and series is due to the deformation of PMV and dsDNA in the complex with respect to their structures shown in Fig. 6 and S8. The highest deformation energy for dsDNA, calculated as a difference between and listed in Table S1 in the ESI, is obtained for the 1BNA MaG complex (25.3 kcal mol− 1), while the smallest for the 1BNA MiG complex (1.2 kcal mol− 1). At the same time, the contribution coming from the deformation of PMV varies from 4.1 (119D MaG) to 9.5 kcal mol− 1 (119D MiG). When summed up, their contribution to the complexation energies ranges from 6.8 (119D MaG) to 30.9 kcal mol− 1 (1BNA MaG). Structures of the MaG and MiG configurations obtained from the DFT calculations for the two different models of the dsDNA:PMV complexes (a) and the corresponding schematic images of the dsDNA sequence, on which the nucleobases closest to PMV are marked (B). In (a) the PMV molecule is colored pink, while the DNA atom colors are: carbon – cyan, nitrogen – dark blue, phosphorus– yellow, oxygen – red, hydrogen –grey. Nucleobases symbols in (b): A – adenine, G – guanine, C – cytosine, T – thymine; 1st and 2nd indicate DNA strands numbering. Additionally, a strength of PMV interactions with a single strand of DNA was estimated by performing single point calculations for the structures of ssDNA: PMV complexes taken from the dsDNA:PMV complexes. For each such system, the interaction energy was calculated according to the formula (3). The values obtained for such ssDNA:PMV complexes containing the 1st or 2nd strand are compared to in Fig. 7b. Note the exceptionally strong interaction of -53.1 kcal mol− 1 between PMV and the 2nd strand of DNA in the 1BNA MaG complex. In this configuration, PMV interacts directly with both Gua and Ade. A similar situation occurs for the 2nd DNA strand in the 119D MaG complex, but here the binding is much weaker. Thus, in addition to interactions with the nucleobases, interactions with the DNA backbones contribute significantly to the overall interaction energy. This is confirmed by the large values of (Table S1 in the ESI) for the ExB configurations, in which PMV interacts only with the DNA backbones. Nevertheless, the results for ssDNA:PMV systems indicate that these complexes can also be stable in aqueous solution. Analysis of the nature and strength of interactions with the nucleobases alone can provide further valuable information. In all four complexes, PMV forms hydrogen bonds with several nucleobases (Fig. 9). Among them is always at least one Thy contributing to the overall binding, but below we focus on Ade and Gua, which are detected in the electrochemical measurements. In the MaG complexes, hydrogen bonds are formed with both Ade and Gua, while in the MiG configurations, with either only one or none of them. Analysis of the geometric parameters of hydrogen bonds with Ade and Gua in MaG complexes and their comparison with Jeffrey’s categorization64,65 suggests that the hydrogen bonds of PMV with Gua are moderately strong, while these formed with Ade are on the borderline between moderate and weak. As can be seen in Fig. 9, in the 1BNA MaG complex, there are two Gua…PMV hydrogen bonds: C = O…H-N with the O…H distance of 1.89 Å and (imidazole)N…H-N with the N…H distance of 1.93 Å. There are also two Ade…PMV hydrogen bonds: N…H (2.58 Å) and H…N (2.45 Å). Note that, in this complex the hydrogen atom from the carboxyl group of PMV was transferred to its amidine group, so the hydrogen bonds are formed by its two amine groups. In the 119D MaG complex, there is only one Gua…PMV bond C = O…H-N (1.91 Å) and also one Ade…PMV bond (imidazole)N…H-N (2.21 Å). In all complexes, the N…H hydrogen bonds are longer and therefore weaker than the O…H bonds. The nucleobases closest to PMV in the four complexes dsDNA:PMV and the hydrogen bonds (dashed black lines) formed between the PMV molecule and the nucleobases: thymine (T), adenine (A) and guanine (G). To evaluate the contribution of individual nucleobase-drug interactions to the total interaction, additional calculations were performed. From each complex, a pair of molecules containing one nucleobase and PMV was extracted and one hydrogen atom was added to a nitrogen atom in the nucleobase at the site of the missing link to DNA. For such pairs of molecules, without changing their original geometry in the complexes, the interaction energies (nb = Ade, Gua or Thy) were calculated and their values are given in Table S3. The strongest Gua-PMV interaction of -13.3 kcal mol− 1 occurs in 1BNA MaG, where two hydrogen bonds are formed between the nucleobase and the drug, and slightly weaker (-11.2 kcal mol− 1) in 119D MaG. In the case of Ade-PMV, it is the opposite: their interaction in 1BNA MaG is smaller than in 119D MaG (-5.7 and − 8.3 kcal mol− 1, respectively). As can be seen, in the complexes MaG the interaction with Gua is always stronger than with Ade by several kcal mol− 1, which confirms the conclusions drawn from the electrochemical results. However, it should be emphasized, that due to the time-consuming nature of quantum calculations, both the methodological approach and the DNA: PMV models are very simplified. It is obvious that the size of small DNA dodecamers is far from the size of real DNA, and the description of an aqueous solution using an implicit solvent model provides only an approximate estimate of its influence on the interactions between molecules. Additionally, due to technical limitations, full optimization of these systems is not possible. Therefore, the computational results presented in this paper are only approximate and specific numerical values should be treated with caution. It would be useful to verify these findings in the future by performing calculations on larger models, as well as using other computational approaches, for example simulation methods. The findings of this study have both broad scientific and practical significance, as they provide novel insights into antiviral drug–DNA interactions. This research not only contributes to the current understanding of antiviral drug–DNA interactions but also highlights the importance of multidisciplinary approaches in drug–DNA research. Only the synergy of various techniques allows for the study of multiple aspects of drug interactions with other molecules in the body, enabling the optimization of their therapeutic effects. It is crucial to determine whether a given drug binds to specific nitrogenous bases of the double helix and how it affects DNA. The results presented in this study provide a solid foundation for more effective antiviral drug design. A deeper understanding of the complex interactions between pharmaceutical compounds and DNA, as well as elucidation of the mechanisms of drug action, enables designing the appropriate experiment and more effective development of existing therapeutic approaches targeting viral diseases. Consequently, this can lead to increased drug efficacy and speed of action while minimizing the likelihood of adverse effects. It can also help reduce the risk of a drug having to be restricted for specific cases or withdrawn from the list of approved pharmaceutical products altogether. This fact directly impacts the financial aspect of pharmaceutical companies and industry, as it helps minimize expenses associated with the development of ineffective therapeutic agents or products with limited therapeutic range. Furthermore, popularizing available drug interaction research influences consumer awareness. Sharing knowledge about the therapeutic value of individual pharmaceutical compounds and their mechanisms of action not only increases awareness but also supports drug promotion, potentially leading to increased use and prescription for specific conditions. Conclusions In this work, a comprehensive study of the interaction between the anti-influenza drug peramivir (PMV) and different types of DNA was conducted using electrochemical, spectrophotometric, biological, and theoretical approaches. The interactions of PMV with individual nitrogenous bases, guanine (Gua) and adenine (Ade), were also investigated. The combined experimental and theoretical data provide a multifaceted understanding of the peramivir:DNA interaction, advancing the knowledge of its molecular mechanisms. The results indicate that PMV does not intercalate into dsDNA, and the drug noncovalently embeds itself in the dsDNA groove, preferentially at the MaG site. Furthermore, voltammetric measurements demonstrated a significant effect of PMV on the unwinding of the DNA double helix. It was found that PMV complexes with both dsDNA and ssDNA are stable due to strong PMV interactions (including N…H and O…H hydrogen bonds) with the sugar-phosphate backbones and nucleobases. In the case of the latter, stronger binding of the drug to Gua than to Ade was observed. This is due to the presence of an oxygen atom in Gua and its involvement in the formation of hydrogen bonds with the drug. Biological studies confirmed that PMV interacts with the DNA helix but does not induce significant DNA damage, such as strand breaks. Therefore, the drug’s action is based on binding to DNA, but does not disrupt its integrity. Moreover, no preventative action from of PMV against oxidative DNA damage was found. The uniqueness of the findings presented regarding dsDNA:PMV and ssDNA:PMV complexes is noteworthy because these interactions have not been previously studied from this perspective. Importantly, this work demonstrates significant differences between the interaction mechanisms of peramivir and compounds such as oseltamivir, highlighting the continued need for in-depth research in this field. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements DFT calculations have been carried out using resources provided by Wroclaw Centre for Networking and Supercomputing (http://wcss.pl), grant No. 443. This work was supported by the University of Lodz, Poland (Grant No. B2311113000177.07). Author contributions S.S., K.K., M.W., P.T., and J.S. conceived and planned the experiments. M.W., P.T., J.S. and K.K. carried out the experiments. A.I. planned and carried out the simulations. M.W., J.S. and P.T. contributed to sample preparation. S.S., K.K., M.W., A.I. P.T., and J.S. contributed to the interpretation of the results. S.S, A.I. and M.W. took the lead in writing the manuscript. All authors provided critical feedback and helped shape the research. All authors reviewed the manuscript. Data availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Monika Wypych, Email: monika.wypych@edu.uni.lodz.pl. Anna Ignaczak, Email: anna.ignaczak@chemia.uni.lodz.pl. Sylwia Smarzewska, Email: sylwia.smarzewska@chemia.uni.lodz.pl. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.