Membrane-stabilizing effect of antiviral agent

This is an automatic translation.
Click here to read the publication in the original language.

Introduction

Due to the development of an unfavorable situation associated with the spread of the 2019-nCoV coronavirus, it is currently relevant to develop an effective drug to prevent the development of viral pneumonia, which is the main complication of influenza and other viral infections (ARVI, including coronaviruses).

The Research Center "Park of Active Molecules" in Obninsk develops and offers an original anti-influenza agent with a new chemical structure and an excellent mechanism of action. As a new anti-influenza agent, a chemical compound, an indole derivative, is proposed. The developed tool can be used to prevent the development of pneumonia caused by influenza viruses and other viral infections (ARVI, including coronaviruses).

Previous studies have shown that for the chemical series of 5-hydroxyindoles, to which the developed agent belongs, a high efficiency has been established in cell cultures in relation to various serotypes of influenza viruses, including coronaviruses (including the pathogen SARS) [1]. In addition to virus-specific activity, they showed a general membrane-stabilizing effect, which also helps to maintain the integrity and functionality of the lung alveoli in viral and bacterial infections. The membrane stabilizing effect is characteristic of the entire class of compounds and is also manifested under the action of radiation exposure and chemotherapy in experimental animals [2]. At the same time, there are still questions of the mechanism of membrane stabilizing action for the entire class of 5-hydroxyindole compounds, including the agent being developed.

It is known that the cell membrane is a dynamic structure that continuously changes its shape. The most important cellular processes, such as exo - and endocytosis, intracellular vesicular transport, are associated with topological rearrangements of membranes, fusion and division, requiring local disruption of the bilayer structure. These processes are associated with the formation of strongly curved membrane surfaces, a certain configuration and lipid composition of which is maintained due to specific lipid-protein interactions. Similar processes of local changes in the morphology of membranes are also realized at various stages of cell infection with the influenza virus, whether it is the absorption of the virion in the process of cellular endocytosis, the fusion of the viral and endosomal membrane, preceding the release of the viral genetic material into the cytoplasm of the infected cell, or the budding of daughter viral particles from the surface of its plasma membranes. It is clear that the energy expenditures required for such topological rearrangements of cell membrane structures will be determined by the mechanical properties of the membrane as a continuous medium. The necessary formalism for such a description was developed in the works of Helfrick [3] and supplemented by Hamm and Kozlov [4]. In these works, it was proposed to consider the lipid bilayer of a cell as a two-dimensional liquid crystal and to use the theory of elasticity of liquid crystals to describe the mechanical properties of membranes. Within this approach, the average orientation of lipid molecules is described by the vector field of unit vectors, called the director, n. The field of directors is set on a certain surface passing inside the lipid monolayer. The shape of a surface is characterized by the vector field of unit normals to it, N. In membrane mechanics, as a rule, three main independent deformations are considered: transverse bending, inclination, and lateral tension / compression. The deformation of the transverse bend corresponds to the appearance of an angle between the directors at close points of the dividing surface, and is quantitatively characterized by the divergence of the director, div (n). The tilt deformation corresponds to the deviation of the director from the normal and is quantitatively characterized by the tilt vector t = n - N. Lateral tension / compression is characterized by a relative change in the area of the dividing surface of the monolayer, (a - a ₀ ) / a ₀ , where a is the current area per lipid per dividing surface, a ₀ is the original area. Part of the free energy associated with deformations is represented by an expansion in a Taylor series in these deformation modes with respect to the spontaneous state under the assumption of small deformations. The curvature of a monolayer in the spontaneous state can be nonzero. This curvature is called spontaneous. Spontaneous curvature is determined by the lipid composition of the membrane. It is believed that the spontaneous curvature of a multicomponent monolayer is equal to the concentration-weighted average of the spontaneous curvature of individual components. Thus, the deformation part of the free energy turns out to be related to the “chemical” contribution due to the dependence of the spontaneous curvature on the concentrations of lipid components. Each of the deformations is characterized by its modulus, i.e. energy that must be spent to implement it. Experiments in various model lipid systems have shown that the greatest modulus is characteristic of the deformation of lateral stretching / compression of the membrane, and therefore such deformation is practically not realized in the processes of cellular morphogenesis, and the deformation of the transverse bending has the greatest effect on the processes of topological rearrangements of the membrane [5].

In this regard, the aim of this work was to study the effect of the indoles derivative on the bending rigidity of the membrane depending on the presence of key components of cell membranes in it, namely: cholesterol, charged lipids, as well as at different degrees of saturation of the hydrocarbon chains of lipid molecules.

1 Lipid nanotube model and its application for the analysis of the mechanical properties of membranes

To carry out these studies, a lipid nanotube (NT) model was chosen, first proposed by the staff of the laboratory of bioelectrochemistry of the Institute of Physical Chemistry, Russian Academy of Sciences, and well-proven in studies of the process of lipid membrane morphogenesis in the process of cellular endocytosis [6, 7].

To study the effect of the indoles derivative on the mechanical properties of membranes, we used an experimental setup consisting of a PAR-175 universal generator (Princeton Applied Research, USA), an EPC-8 patch-clamp amplifier (HEKA Elektronik, Germany), an F-900 four-pole filter (Frequency Devices , USA) and an OS-1420 oscilloscope (GOULD, England). The position of the pipette relative to the BLM was varied using Newport Motion Controller (Model 860-C2) and a pre-calibrated Model ESA-CSA motion microcontroller (Newport, USA) (a piezo controller that allows you to change the vertical position of the pipette with an accuracy of 0.1 μm.).

The formation of a model lipid bilayer membrane (BLM) was carried out by the Mueller-Rudin method [8] on holes in a copper lattice (EMS, USA) with a diameter of ~ 100 µm, located in the volume of a Petri dish. The grid holes were treated with a phospholipid solution (Avanti Polar Lipids Inc., USA) in a 1/1 octane / decane mixture (Sigma, USA) at a concentration of 10 mg / ml and dried in an atmosphere. After that, the Petri dish was filled with an electrolyte solution (10 mM KCl, 1 mM Hepes, 0.1 mM EDTA, pH = 7.0; 50 mM KCl, 5 mM Hepes, 0.5 mMEDTA, pH = 7.0 or 100 mM KCl , 10 mM HEPES, pH 7.0), and a drop of a solution of phospholipids in squalane (Sigma, USA) at a concentration of 10 mg / ml was applied to the holes in the lattice with a brush, which spontaneously formed BLM within several minutes. In this case, the solvent and excess lipid went into the bulk phase surrounding the BLM, that is, into the meniscus. The following lipid solutions were used in the experiments: 1,2-dioleoyl-sn-glycero-3-phosphatidylcholine (DOPC), 10 mg / ml 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC), 1,2 -dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE), 1,2-dioleoyl-sn-glycero-3-phosphatidylserine (DOPS) and cholesterol (Cholesterol) (Avanti Polar Lipids Inc., USA) in chloroform, concentration 10 mg / ml, in the ratios indicated in the text.

After the formation of the BLM, a sawtooth signal from the generator was applied to the silver chloride electrodes by means of a patch-clamp amplifier, and the capacitive current of the membrane was recorded in the potential clamping mode. Using a micromanipulator, a patch pipette was brought to the membrane, so that a tight contact was formed between the tip of the micropipette and the membrane (contact resistance 1-10 GΩ), which was evident from a sharp decrease in the capacitive current. To destroy the membrane under the pipette, the hydrostatic pressure was abruptly changed, which led to the appearance of a conduction current. After that, removing the pipette from the flat membrane, the membrane tube was pulled out. The electrical conductivity of the tube was measured in the mode of fixing the potential difference between the measuring electrode inside the patch pipette and the ground electrode in the external volume of the electrolyte using a current amplifier. The values of the potentials used are indicated in the text.

A smooth change in the length of the tube was carried out using a piezo controller. The obtained data - the current through the tube, converted by the patch clamp by the amplifier into voltage (gains from 1mV / pA to 30mV / pA), the values of the electric voltage applied to the ends of the tube and the data from the indicator of the piezo controller - were recorded on the computer hard disk after preliminary digitization using ADC L-305 / L-1210 (L-card, Russia). The sampling rate is 1 kHz. The signals were passed through a low-pass filter (F-900) before being entered into a computer; cutoff frequency 0.5 kHz. The piezoelectric controller readings were converted to the values of the vertical displacement of the pipette using a calibration curve. At the transition from MT to NT, the conduction current fell sharply to zero (Figure 1, curve 1), after which the movement of the pipette stopped. Further, by amplifying the signal from the conduction current by a factor of 10-30 and bringing the pipette closer to the membrane with a piezo controller, the changes in the conduction current through the NT were recorded, and if an increase in the current was observed, then it was possible to say about the presence of NT between the pipette and the membrane. Then the dependence of the conduction current on the position of the micropipette was recorded.

The peculiarities of NT formation allow us to monitor the change in the conductivity of the drawn tube depending on its length (Figure 2, C). The conductivity measured during the experiment includes the conductivity of the NT and the conductivity of the leak at the point of contact of the membrane with the patch pipette. Assuming that the shape of the NT differs slightly from the cylindrical one, and assuming that the leakage conductivity is constant (when the pipette moves, only the length L of the NT changes), the electrical resistance RHT should linearly depend on the change in the length ΔLNT. Indeed, the linear function gives a good approximation of the R (L) dependence (Figure 2, D). Thus, since:

,

where ρ _sp is the resistivity of the electrolyte, then both the length L NT and its radius r _NT can be calculated.

If the NT contains charged lipids, it is necessary to take into account the fact that the concentration of electrolyte ions inside the NT is higher than outside [10].

,

where ρ _NT / ρ _ulk is the specific conductivity of the electrolyte inside the NT / outside the NT, φ® is the potential distribution inside the NT.

The equilibrium radius of such NT is defined as:

The method [9] was used to measure the mechanical parameters. The essence of the method is to analyze the dependence of the measured radius of NT on the value applied to its ends of the potential difference. The fact is that the equilibrium radius of NT is determined by the root of the ratio of the modulus of bending of the NT membrane to the lateral tension of the BLM, and according to the Lippmann electrocapillarity equation, the lateral tension decreases with the appearance of transmembrane stress. Thus, as a result of a drop in the electric potential along the NT, the transmembrane potential on the NT wall changes from the applied value at one end to 0 at the opposite end. This leads to the fact that the shape of the NT deviates from the cylindrical one. However, as shown by our theoretical calculations, the electrical resistance of the NT retains a linear dependence on its length. Thus, the measured radius of the NT corresponds to the radius of some effective cylinder, which has the same conductivity and length as the NT. Moreover, the value of the effective radius of the NT is related to the mechanical parameters and the value of the potential difference applied to the ends of the NT by the following expression:

where K is the bending modulus of the membrane, σ ₀ is the lateral tension of the BLM, C _sp is the specific electrical capacity of the BLM, U is the voltage applied to the ends of the NT.

Consequently, from the linearization of the dependence of the inverse square of the NT radius on the square of the applied stress, it is possible to find both the membrane bending modulus and the lateral tension of the BLM (Figure 3).

2 Results of studies of the interaction of the indoles derivative with lipid nanotubes

To study the effect of the test compound on the mechanics of the lipid matrix of cell membranes using a lipid nanotube model, a lipid composition was selected similar to that described for the Maidin-Derby canine kidney cells (MDCK) (the main components were phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and cholesterol), on which a detailed study of cellular and viral lipidome, their common features and differences [10]. In addition, lipids were used with both saturated and unsaturated lipid tails, which is characteristic of cell membranes. In a number of experiments, when assessing the effect of charged lipids on the interaction of the test compound with lipid membranes, the charged membrane components were removed and replaced with neutral dioleoylphosphatidylcholine (DOPC). Similarly, in the case of experiments with different amounts of cholesterol, the cellular composition was also deviated.

The test compound was added to the system in three different ways: from a solution in deionized water with a concentration of 0.1 mg / ml, from a solution in 96% ethyl alcohol with a concentration of 1 mg / ml, or by adding a solution of the test compound in chloroform with a concentration of 1 mg / ml directly to the lipid mixture at the stage of BLM formation.

In the first case (an aqueous solution with a concentration of 0.1 mg / ml), the test compound did not dissolve completely: a suspension remained in the water, which did not disappear for a long time (two weeks). Therefore, for further work, this solution was kept in an ultrasonic bath and filtered through pores with a diameter of 100 nm. The resulting solution was studied by atomic force microscopy (AFM) on a mica substrate. For this, it was applied to mica (the diameter of the substrate was 1.5 cm) in a volume of about 200 μl and incubated for 15 minutes at room temperature, then a drop of the solution was dried with an argon flow. The resulting sample was scanned in the resonance mode of the instrument operation (tapping mode).

From the resulting image (Figure 4), it can be seen that spherical particles of matter are presented on the surface of mica, the sizes of which vary greatly: the particle heights are from 0.6 to 1.4 nm, the particle diameter is in the range of 20-80 nm. The data obtained indicate that in water the test compound is presented mainly in the form of agglomerates.

Both in chloroform and in 96% ethanol solution, the test compound at a concentration of 1 mg / ml was completely dissolved. The AFM images showed that in these cases the test compound is represented by particles with heights ranging from 0.2 to 0.9 nm and a diameter ranging from 12 to 20 nm (Figure 5).

However, a solution of the test compound in ethanol after two weeks acquired a yellowish color, the intensity of which depended on its concentration. This may indicate the transition of the compound under study from the hydrochloride form to the base form.

We studied the interaction of an alcohol solution of the test compound with a membrane of the following lipid composition: DOPC / DOPE / POPC / Cholesterol = 49/13/13/25 mol%. (DOPC - dioleoylphosphatidylcholine; DOPE - dioleoylphosphatidylethanolamine; POPC - palmitoyleoylphosphatidyloline)

It was shown that the addition of an alcohol solution of the test compound to the system to a final concentration of 1 μg / ml (buffer 10 mM KCl, 1 mM Hepes, 0.1 mM EDTA, pH = 7.0) statistically significantly reduces the bending modulus of the membrane by half. ... A further increase in the concentration of the test compound in solution to 2 μg / ml did not lead to a statistically significant change in the flexural modulus. It was found that the lateral tension of the membrane did not depend on the presence of the test compound in the solution. In control experiments, it was shown that when ethanol was added to the membrane (the maximum value of the volume fraction of alcohol in the solution reached 0.2%, which corresponded to a similar addition of an alcohol solution of the test compound), the membrane flexural modulus did not change. However, the fact that in an alcoholic solution of the test compound, possibly, passes from the hydrochloride form to the base form, does not allow us to draw a definite conclusion about the effect of the initial test compound on the membrane when it is added in an alcoholic solution.

A study was carried out of the properties of the membrane in the presence of the test compound in it. For this, the test compound, also dissolved in chloroform (concentration 1 mg / ml), was added to a solution of DOPC / DOPE / POPC / Cholesterol lipids dissolved in chloroform. The experiments were carried out for the following ratio of components: DOPC / DOPE / POPC / Cholesterol / FS = 49-X / 13/13/25 / X mol%, where X takes values: 0.02; 0.1; 0.5 mol%. The obtained values of the membrane bending modulus did not depend on the initial concentration of the test compound in the mixture, and amounted to (0.3 ± 0.1) × 10 ^-19 J. The value of the flexural modulus in the absence was (0.9 ± 0.1) × 10 ^{- 19} , that is, the PS initially added to the membrane reduces its bending modulus by a factor of 3. Moreover, the fact that there is no dependence of the bending modulus on the concentration of the test compound in the membrane, most likely, indicates that already at a concentration of 0.02 mol%, its practically maximum possible concentration in the bilayer lipid membrane is reached.

A study was carried out of the interaction of an aqueous solution of the test compound, prepared as described above, with the membrane, depending on the concentration of cholesterol in it. Experiments were carried out with a buffer solution: 50 mM KCl, 5 mM Hepes, 0.5 mMEDTA, pH = 7.0 lipid composition DOPC / DOPE / POPC / Cholesterol = 49/13/13/25 mol% and DOPC / DOPE / POPC / Cholesterol = 34/13/13/40 mol%. At 25 mol% cholesterol in the absence of the test compound, the flexural modulus is s, after adding the test compound, the modulus decreased to a value of (0.4 ± 0.1) × 10 ^-19 J, which, within the measurement error, corresponds to the results obtained when it was introduced. directly into the composition of the membrane from a solution in chloroform (the bending modulus drops by a factor of 2-3). In the case of 40 mol% cholesterol, which corresponds to its level in the composition of the lipid envelope of the influenza virus [11], in the absence of the test compound, the flexural modulus value coincides with the value for the membrane at 25 mol% cholesterol (0.9 ± 0.1) × 10 ^{- 19} J. After adding an aqueous solution of the test compound, the value of this modulus was (0.6 ± 0.1) × 10 ^-19 J, that is, its effect on the bending stiffness of the membrane decreased in proportion to the concentration of cholesterol in the membrane.

When charged lipids (phosphatidylserine) are introduced into the system to a level corresponding to the average content of these lipids in cell membranes (composition DOPC / DOPE / POPC / Cholesterol / DOPS = 34/13/13/25/15 mol%) (DOPS - dioleoylphosphatidylserine), the effect of the test compound on reducing the bending stiffness of the membrane became most pronounced: a decrease in the flexural modulus was observed from (0.88 ± 0.14) × 10 ^-19 J to (0.22 ± 0.10) × 10 ^-19 J.

Conclusion

Thus, we studied the effect of the compound under study on the mechanical properties of lipid membranes using a lipid nanotube model. Three possible options for adding the test compound to the system were considered: in the form of a filtered aqueous solution, in the form of a solution in alcohol, and by direct introduction into the membrane from a solution in chloroform. In all these cases, it was observed that when the test compound was added to the lipid membrane, the flexural rigidity of the latter decreased. However, in the case of an alcohol solution, the studied compound was transferred from the hydrochloride form to the base form; therefore, these solutions were not used for further studies. The fact that the results for the case of adding the test compound in the form of an aqueous solution to the buffer washing the membrane with the results for its direct incorporation into the BLM show that in all cases the incorporation of the test compound into the membrane is observed, leading to a 3-fold decrease in the flexural rigidity of the latter. Moreover, the limiting concentration of the test compound in the lipid bilayer does not exceed 0.02 mole percent. Since the low flexural rigidity of membranes, as mentioned above, simplifies local topological rearrangements of cell membranes [9, 10], the compound under study can potentially accelerate the processes of cell life, synaptic transmission, phagocytosis, etc. It should be noted that the maximum effect of a decrease in flexural rigidity was observed in the presence of charged lipids in the membrane, i.e. precisely in the case when the lipid composition of nanotubes mirrors the composition of cell membranes as closely as possible.

Experiments with different cholesterol content in the lipid membrane showed that this component of cell membranes, most likely, prevents the incorporation of the test compound into the lipid bilayer, i.e. its introduction into the lipid membrane of the virion enriched with cholesterol seems to be more difficult than into the cell membrane. However, this statement requires additional verification, which will be carried out at the next stage of research.

Thus, the indoles derivative was found not only to have a pronounced virus-specific effect, but also to change the flexibility and increase the resistance of the epithelial cell membranes of the bronchopulmonary system, and contribute to the creation of specific immune protection. The results of the studies performed allow considering the compound under study for use before and during a viral epidemic, including for preventing the development of viral pneumonia, which is the main complication of influenza and other viral infections (ARVI, including coronaviruses).

We express our deep gratitude and gratitude to the head of the laboratory of bioelectrochemistry and his staff of the Institute of Physical Chemistry and Electrochemistry named after V.I. A.N. Frumkin of the Russian Academy of Sciences, Professor, Ph.D., Associate Professor O.V. Batishchev for his help in researching the membrane-stabilizing action of the indoles derivative.

Authors (LLC Research Center Park of Active Molecules, Obninsk):

• Erimbetov Kenes Tagaevich, Doctor of Biological Sciences
• Goncharova Anna Yakovlevna, Ph.D.
• Bondarenko Ekaterina Valerievna, Ph.D.
• Rakhimdzhan Akhmetdzhanovich Roziev, Ph.D.
• Zemlyanoy Ruslan Alexandrovich, postgraduate student

List of sources used

1. Dondurei E.A., Osidak L.V., Golovacheva E.G., Golovanova A.K., Gladchenko L.N., Kostitsina E.K. Arbidol in the complex therapy of combined lesions of the respiratory and gastrointestinal tracts in children // Children's infections. - 2007. - No. 2. - P.39-44.
2. Miroshnichenko A.G. The effectiveness of arbidol in the correction of the nephrotoxic effect of antineoplastic chemotherapeutic agents (experimental study) // Abstract of the thesis for the degree of candidate of medical sciences. - 2010 .-- 21 p.
3. Helfrich W. Elastic properties of lipid bilayers: theory and possible experiments // Z. Naturforsch - 1973 - 28 - P. 693-703.
4. Hamm M. Elastic energy of tilt and bending of fluid membranes / Hamm M., Kozlov MM // Eur. Phys. J. - 2000 - P. 323-335.
5. Galimzyanov T.R. Linear tension and structure of the raft boundary, calculated taking into account the deformations of bending, inclination and tension / compression / T.R. Galimzyanov, R. Yu. Molotkovsky, S.A. Akimov // Biol. Membranes - 2011- 28 (5) - P.415-422.
6. Bashkirov PV GTPase cycle of dynamin is coupled to membrane squeeze and release, leading to spontaneous fission / PV Bashkirov, SA Akimov, AI Evseev, SL Schmid, J. Zimmerberg, VA Frolov // Cell - 2008 - 135 (7) - P. 1276-1286.
7. Shnyrova AV Geometric catalysis of membrane fission driven by flexible dynamin rings. / PV Bashkirov, SA Akimov, TJ Pucadyil, J. Zimmerberg, SL Schmid, VA Frolov // Science - 2013- 339 (6126) - P. 1433-1436.
8. Mueller P. Methods for the formation of single bimolecular lipid membranes in aqueous solution / P. Mueller, DO Rudin, HT Tien, WC Wescott // J. Phys. Chem. - 1963 - 67 - P. 534-535.
9. Bashkirov P.V. Membrane nanotubes in an electric field as a system for measuring the mechanical parameters of the lipid bilayer // Biol. Membranes - 2007 - 24 (2) - P. 183-192.
10. Gerl MJ, Quantitative analysis of the lipidomes of the influenza virus envelope and MDCK cell apical membrane / MJ Gerl, JL Sampaio, S. Urban, L. Kalvodova, J.-M. Verbavatz, B. Binnington, D. Lindemann, CA Lingwood, A. Shevchenko, C. Schroeder, K. Simons // J. Cell Biol. - 2012 - 196 (2) P. 213-221.
11. Campelo F. Helfrich model of membrane bending: from Gibbs theory of liquid interfaces to membranes as thick anisotropic elastic layers / F. Campelo, C. Arnarez, SJ Marrink, MM Kozlov // Adv. Colloid Interface Sci. - 2014 -208 P. 25-33.

Download the article in PDF (394 KB)

This is an automatic translation.
Click here to read the publication in the original language.