Electrostatics affects formation of Watson-Crick complex between DNA bases in monolayers of nucleolipids deposited at a gold electrode surface

1 Electrochimica Acta, 2021, 390,138816 Electrostatics affects formation of Watson-Crick complex between DNA bases in monolayers of nucleolipids deposited at a gold electrode surface Francisco Prieto Dapenaa*, ZhangFei Sub, Julia Alvarez Malmagroa,b, Manuela Ruedaa*, Jacek Lipkowskib* a Department of Physical Chemistry, University of Seville, C/Professor García González n◦ 2, 41012 Seville, Spain. b Department of Chemistry, University of Guelph, Guelph, Ontario, Canada N1G 2W1. Corresponding authors: Francisco Prieto: dapen[email protected] Manuela Rueda: m[email protected] Jacek Lipkowski:[email protected] In honor of Professor Sergio Trasatti on the occasion of his retirement as editor of Electrochimica Acta and in recognition of his outstanding contributions to electrochemistry Abstract Chronocoulometry was applied to determine charge in a monolayer of nucleolipid (1,2dipalmitoyl-sn-glycero-3-(cytidine diphosphate)) deposited at a gold electrode surface. The immersion method was used to measure the potential of zero charge of the interface (Epzci), which is a sum of charge on the monolayer of the nucleolipid and charge on the gold surface. Photon polarization infrared reflection absorption spectroscopy (PM IRRAS) was used to determine formation of the Watson-Crick complex between terminal cytidine moiety of the nucleolipid and guanine (its complementary base) added to the solution. The combination of electrochemical and spectroscopic studies allowed one to demonstrate that the Watson-Crick complex is formed when the interface is positively charged. The potential applied to the electrode affects not only the complex formation but also orientation of the cytosine moiety. The complex is formed when the cytosine moiety is oriented assuming a small angle with respect to the electrode surface. 2 1.- Introduction Lipid monolayers and bilayers supported on solid electrodes constitute systems that mimic, both, the chemical and the electrostatic environments of biological membranes. Therefore, the study of the behavior and the interactions of molecules biologically relevant located on these modified electrodes by electrochemical and spectroelectrochemical methods can provide phenomenological and structural information about their behavior in nature. Additionally, in the case of the DNA bases located on electrode surfaces, their specific Watson-Crick interactions between complementary bases, [1], can be used to develop new electrochemical biosensors. We have made several efforts to describe molecular recognition reaction between complementary DNA bases at an electrode surface. Our first attempts to study nucleotides directly adsorbed at a gold electrode surface [2–8] were not successful because metal-base interactions were stronger than interactions between co-adsorbed complementary bases. To overcome these limitations, we have employed monolayers of nucleolipids. Nucleolipid monolayers have shown specific molecular recognition capabilities to the complementary DNA bases [9–16]. Therefore, a monolayer of 1,2-dipalmitoyl-sn-glycero-3-(cytidine diphosphate) nucleolipid (16:0 DG-CDP) (Figure 1) was deposited at a gold electrode surface [17]. Since cytidine moiety was separated from the metal surface by a long dipalmitoyl chains, the formation of Watson-Crick with guanine could be observed.[18] The coupling between the C=O stretching vibrations of cytosine and guanine in the Watson-Crick complex was similar to that observed in the complex of polynucleotide helices (poly-cytosine:poly-guanine) [19]. However, the cross section of the polar head group of the nucleolipid is much larger than the cross section of the acyl chains (see Figure 1) and overcrowding of the polar heads could affect the Watson-Crick complex formation. To address this concern a mixed monolayer of DG-CDP and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) was assembled [20]. The DG-CDP and DPPC with (7:3) molar ratio displayed property of an ideal solution in which DG-CDP molecules were diluted by DPPC. Indeed, in the mixed 3 monolayer the polar head group of the nucleolipid was more parallel to the electrode surface than in the monolayer of pure nucleolipid. The objective of this work is to explore the molecular recognition capabilities to guanine of mixed monolayers of 16:0 DG-CDP and DPPC transferred to gold electrode surfaces, with higher separation between nucleolipid polar heads than in pure 16:0 DGCDP monolayers. With the help of chronocoulometry we will determine charge of the polar head of the nucleolipid. We will also determine the potential of zero charge of the interface (Epzci) which corresponds to the sum of charge on monolayer of the nucleolipid and on the surface of the metal, and show that the Watson-Crick complex is formed when the interface is positively charged. N NH2 ON O OHOH HH HH OPO OO P O O O O O O O NH N N H N O NH2 c) Guanine (G) b) 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) a) 1,2-dipalmitoyl-sn-glycero-3-(cytidine diphosphate) nucelolipid - 16:0 DG-CDP OP O O O O O O O N NN O N H H H H N N N N N O H H H H H d) Cytosine:Guanine (C:G) complex Figure. 1 Molecular structure of a) 1,2-dipalmitoyl-sn-glycero-3-(cytidine diphosphate), b) 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) ,c) guanine (G) and d) Cytosine:Guanine (C:G) Watson-Crick complex. 4 With the help of PM IRRAS we will show that the complex is oriented nearly parallel to the electrode surface. In this way we will provide new information how electrostatics influence the complex formation and orientation of the polar head of the nucleolipid. The data for the mixed monolayer with and without guanine will be compared and the effect of dilution of the nucleolipid with the phospholipid will be discussed. 2.- Experimental. 2.1. Reagents, solutions, and electrodes. The stock solutions of lipids with the desired mole fraction of 16:0 DG-CDP were prepared by mixing the required volumes of 1 mg mL-1 solutions of 1,2-dipalmitoyl-snglycero-3-cytidine diphosphate (16:0 CDP DG) and 1,2-dipalmitoyl-sn-glycero-3phosphocholine (Avanti Polar Lipid) in chloroform. The solutions were stored at -20 C. The NaF 0.1 M (Sigma-Aldrich - BioXtra, 99%) supporting electrolyte was prepared from NaF powder cleaned in an UV ozone chamber for 15 min. Ultrapure water, freshly purified with a Milli-Q system was used to make solutions, for electrochemical measurements, and D2O (Cambridge Isotope Laboratories) was employed to make solutions for PM-IRRAS experiments. The saturated solution of guanine (Sigma-Aldhrich) (c.a. 0.025 mM) was prepered in 0.1 M NaF electrolyte. The electrochemical and PM-IRRAS measurements were performed in a threeelectrode glass cells. Single crystal gold (111) electrodes, home prepared according to the Clavilier method[21], and supplied by Mateck were used as working electrodes for the electrochemical and PM-IRRAS measurements, respectively. Ag/AgCl (saturated KCl, Pine Research Instrumentation, -0.045 V vs SCE) was employed as a reference electrode. The auxiliary electrode was a platinum foil. All the measurements were performed at room temperature (22 ± 2 °C). All potentials are reported versus the saturated calomel electrode (SCE). The glassware was cleaned overnight in an acidic solution of potassium permanganate and rinsed first with diluted piranha mixture (3:1 H2SO4/H2O2 v/v) and next 5 meticulously rinsed with Milli-Q water. The PTFE components were cleaned by immersion in piranha mixture followed by exhaustive rinsing with Milli-Q water. 2.2. Preparation of the monolayers, measurements of the Langmuir isotherms and transfer of the monolayers to the gold single crystal (111) electrode. Langmuir isotherms were registered with a Nima 611D Langmuir trough (270 cm2 of area and c.a. 150 cm3 of subphase volume), equipped with two PTFE moving barriers and a PS4 pressure sensor and located in a Perspex© cabinet to avoid contamination and air stream interferences. The Wilhelmy plate (a strip of 1 cm-wide Whatman© chromatographic paper) was freshly made for every measurement. The PTFE trough and the moving barriers were cleaned by repetitive rinsing with methanol and Milli Q water. The mixed monolayers of DPPC and 16:0 DG-CDP were prepared by spreading a volume of 20-30 L at the air/water interphase in the Langmuir trough with open barriers. The solvent of the lipid solution was allowed to evaporate for 20 min. Then, the barriers compressed the monolayer at a rate of 25 cm2 min 1, while the surface pressure was registered. The monolayers were transferred to the gold single crystal (111) electrode surfaces at the equilibrium spreading pressure (c.a. 30 mN m-1) by the horizontal touch (LangmuirShaefer method) technique. The electrode was previously flame annealed and allowed to cool down in the Langmuir trough cabinet for 30 min. The electrode was horizontally touched to the monolayer using a computer-controlled step motor. It was then immediately lifted. Afterwards, it was dried in Ar atmosphere for 60 min. 2.3. Electrochemical instrumentation and measurements. Electrochemical measurements were carried out using the HEKA PG590 potentiostat. Before the working electrode was brought into the hanging meniscus configuration, oxygen in the electrolyte solution was removed by purging argon during 30 min. The procedure to measure surface charge densities, was described in the previous paper [17]. The charge density versus potential plots were measured by integrating the current transients registered when potential was stepped from a given potential E to E = - 6 0.85 V vs SCE [22,23]. The monolayer is completely detached from the gold surface at potential -0.85 V vs SCE. At this potential charge density for the electrode initially covered by the monolayer is equal to charge density at the bare electrode surface. The potential of zero charge for the electrode free from the monolayer was determined independently from the position of the minimum of diffuse double layer on the differential capacity curve. It was used to covert the charge differences between a given potential E and E = -0.85 V vs SCE into the absolute charge densities. The potential of zero free charge of the interface (Epzci) was determined in an independent experiment by the immersion method [24–26]. Figure SI 1 of the supporting information, plots charge densities determined by the immersion method. The intersection of this linear relation with zero charge provided the value of Epzci = 0.08 V vs SCE. 2.4. PM-IRRAS measurements. Nicolet Nexus 870 (Thermo Fisher) was used to collect PM IRRAS spectra. It was equipped with an external tabletop optical mount (TOM) box which incorporated a MCTA detector cooled with liquid nitrogen, the sample compartment and the optical head (II/ZS50 ZnSe 50 kHz) of the photoelastic modulator (PEM, PM-90 from Hinds Instruments) and a sampling demodulator (GWC Instruments Synchronous Sampling Demodulator). A CaF2 equilateral prism (Boxin Photoelectronic Co.) was the IR window as described in [27,28]. The EG&G PAR362 potentiostat was used to control potential of the working electrode. The Omnic software was used for spectra acquisition by means of a macro that triggered the potential steps and the spectra collection. The gold (111) electrode modified with the nucleolipid monolayer incubated in 0.025 mM guanine solution was transferred to the spectroelectrochemical cell containing 0.1 M NaF in D2O. The supporting electrolyte was then de-aerated by purging with Ar during 20 min. The electrode was pressed against the rectangular face of the CaF2 prism, maintaining a thin layer (1-10 m) of the electrolyte between the electrode and the prism, determined by the procedure described elsewhere [27]. The spectrometer and the TOM box were purged with dry and CO2 free air from purge gas generator (Parker Blaston). Electrode potential was changed in steps of -0.1 V from 0.36 to -0.85 V vs SCE. At each potential, 4000 interferograms were recorded with an instrument resolution of 4 cm-1and averaged. 7 In order to optimize the mean square electric field strength (MEFS), the incidence of the IR beam and the half-wave retardation of the photoelastic modulator were optimized for different spectral regions (CH stretching region around 3000 cm-1 and CO, CC and CN stretching region around 1600 cm-1). The procedure of PM-IRRAS measurements and the signal processing, including the correction of the PEM response functions are described in [27,28]. After correction for PEM response functions, PM-IRRA spectra measure ∆S: ∆𝑆=𝑅𝑝−𝑅𝑠 𝑅𝑝+𝑅𝑠=2.3 𝜀 Γ (1) where, Rp and Rs are the reflectivities with p and s polarized radiation, Γ is the surface concentration of the absorbing species and  is the molar absorption coefficient of the adsorbed species. Theoretical PM-IRRA spectra for mixed monolayers of DPPC and 16:0 DG-CDP, with a random orientation of the molecules, were simulated by solving the Fresnel equations for the system of four parallel homogeneous phases (CaF2|D2O|DPPC|Au) or (CaF2|D2O|16:0 DG-CDP|Au) , as described in [20]. 3.- Results and discussion. 3.1. Langmuir isotherms of mixed DG-CDP|DPPC monolayers. The molecular recognition reaction between guanine and DG-CDP|DPPC monolayer was investigated in 0.1M NaF solution. The compression isotherm of the mixed DG-CDP|DPPC monolayers were studied in [20] using pure water as the subphase. The substitution of water by 0.1M NaF solution was serendipitous. However, it gave very interesting results. Unexpectedly, significant differences between properties of the mixed monolayers spread on 0.1M NaF and on pure water were observed. Figure 2 compares isotherms of DG-CDP|DPPC monolayers measured at the two subphases at different molar fractions of the nucleolipid. The monolayers spread at 0.1M NaF are much more expanded 8 Amolec /Å2 40 80 120 160 200  / mN·m-1 0 20 40 60 80 Amolec / Å2 40 80 120 CS-1 / mN·m-1 100 200 a) Amolec /Å2 40 80 120  / mN·m-1 0 20 40 60 80 Amolec / Å2 40 80 120 CS-1 / mN·m-1 0 100 200 b) Figure 2. Langmuir isotherms measured at 22oC for the mixed DG-CDP:DPPC monolayers with molar fractions of DG-CDP (xDG-CDP): xDG-CDP =0 (red solid line), xDG-CDP =1 (blue dashed line), xDG-CDP =0.2 (green dashed line), xDG-CDP =0.4 (black dot-dash line) and xDG-CDP =0.7 (pink dashed line), measured at aqueous 0.1 M NaF subphase. a), and at pure water subphase, b). Insets plot compression modulus, Cs-1, defined as 𝐶𝑠−1 = −𝐴𝑚𝑜𝑙𝑒𝑐(𝑑𝜋 𝑑𝐴𝑚𝑜𝑙𝑒𝑐 ⁄) versus the area per molecule. Data for pure water subphase taken from [20]. than that measured at pure H2O. This is reflected in the onset of compression that takes place at much higher mean molecular areas and in much lower values of the compression modulus than for films compressed at pure water (see insets to Figure 2 a and b). For monolayers spread on 0.1M NaF, the collapse pressure does not change with the mixture composition. In contrast, it changes appreciably for the monolayer spread on H2O. Such changes are consistent with significant differences in formation of DG-CDP vesicles in pure water and in NaCl solutions observed in [29]. 9 Aexc / Å2 -10 0 10 20 30 xDG-CDP 0.0 0.2 0.4 0.6 0.8 1.0 Gexc / J mol-1 -300 0 300 Aexc / Å2 -20 -10 0 10 xDG-CDP 0.0 0.2 0.4 0.6 0.8 1.0 Gexc / J mol-1 -300 0 300 a) b) c) d) Figure 3. Mean molecular excess area, a) and b), and excess Gibbs free energy, c) and d), plotted as a function of the mole fraction of DG-CDP at surface pressures of 5 mN m-1 (red circles), 15 mN m-1 (green triangles) and 30 mN m-1 (blue squares). Left panel data for 0.1M NaF subphase, right panel data for pure H2O as the subphase. The differences between the monolayers spread at H2O and 0.1M NaF can be quantified by calculating the excess molecular area, ∆𝐴𝑒𝑥𝑐: ∆𝐴𝑒𝑥𝑐 =𝐴𝑚𝑜𝑙𝑒𝑐 − (𝑥𝐷𝐺−𝐶𝐷𝑃 ·𝐴(𝑥𝐷𝐺−𝐶𝐷𝑃=0)+(1−𝑥𝐷𝐺−𝐶𝐷𝑃)·𝐴(𝑥𝐷𝐺−𝐶𝐷𝑃=1)) (2) and the excess Gibbs energy, ∆𝐺𝑒𝑥𝑐 defined by: ∆𝐺𝑒𝑥𝑐 =∫∆𝐴𝑒𝑥𝑐 𝜋 0𝑑𝜋 (3) They are plotted in Figures 3 a and b. The differences are significant. For monolayers spread at 0.1M NaF both ∆𝐴𝑒𝑥𝑐 and ∆𝐺𝑒𝑥𝑐 are predominantly positive, indicating that the interactions between molecules in the monolayer are repulsive. In contrast, for monolayers spread at water both ∆𝐴𝑒𝑥𝑐 and ∆𝐺𝑒𝑥𝑐 are negative when the mole fraction of DG-CDP is higher than 0.4, pointing out that the interactions between the molecules in the film are 16 transmission spectrum of guanosine diphosphate and hence may be assigned to non-bonded guanine. The bands at ~1700 and ~1690 cm-1 are signatures of Watson-Crick complex and originate from the up-shift of the guanine C=O band caused by intermolecular coupling in the complex [19,37,38]. The presence of two bands suggests presence of complexes with stronger and weaker intermolecular coupling. The bands at 1656, 1645 and 1612 cm-1 can be assigned to cytosine moiety The band at 1612 cm-1 is assigned to cytosine ring vibrations and bands at 1656, 1645 cm-1 are predominantly carbonyl band vibrations coupled to the skeletal C=C vibrations. The coupling gives symmetric and antisymmetric stretches which result in splitting of the cytosine C=O band [37]. Figure SI 7 of SI shows that they are quite sensitive to the molecular environment. Both bands are present in the mixed monolayers. In the presence of guanine, they are somewhat blue shifted. However, in the monolayer of pure nucleolipid with guanine they are significantly blue shifted, and they overlap strongly with the guanine bands. For comparison, in the IR spectrum of dG5C5 the double stranded complex shows two carbonyl cytosine bands while the spectrum of d(GC)8 complex has only one carbonyl band. The structure of d(G5C5) has A-form and is similar to poly(dG)- poly(dC) while d(GC)8 has B-form structure with alternating G and C bases in the stack. These differences illustrate that the coupling of the carbonyl stretch and C=C stretch vibrations of cytosine is affected by the intermolecular interactions [37]. Consistent with the literature [19,37], the up shift of the C=O guanine band is strong when the Watson-Crick complex is formed and the down shift of the cytosine band is weak. The shape of guanine bands changes significantly with potential while the changes of cytosine bands are less dramatic. This point is illustrated by Figure 7 which plots intensities of guanine and cytosine bands as a function of the electrode potential. The C=O bands of guanine forming complex essentially disappear while the band corresponding to the nonbonded guanine increases at negative potentials. 17 Integrated intensities / cm-1 0.00 0.05 0.10 -0.8 -0.4 0.0 E-Epzfc /V E /V vs SCE -0.8 -0.4 0.0 0.4 0.00 0.05 0.10 a) guanine b) cytosine Figure 7. Integrated intensities vs potential plots corresponding to the bands a) at 1678 cm1 (blue triangles-down), 1690 cm-1 (purple diamonds) and 1700 cm-1 (black triangles-up); b) 1645 (green squares) and 1656 cm-1 (red circles), obtained from deconvolution of the PM-IRRA spectra of a mixed monolayer of DP-CDP:DPPC 7:3 incubated in the presence of 25 M guanine and transferred to a gold (111) electrode at the equilibrium pressure. This behavior indicates that the complex is stable when the total charge at the interface (charge on the metal plus the charge of the lipid monolayer) is positive, in agreement with the studies of guanine bonding to the monolayer of the pure nucleolipid [18]. In contrast to the behavior of guanine in the complex, the bands of cytosine are increasing by moving from positive to negative potentials. They follow the trend for the band of non-bonded guanine suggesting that complex formation has not strong effect on the cytosine bands. We will show below that the change of the cytosine bands are caused chiefly by the change of the orientation of the cytosine moiety rather than a change of the complex formation/disappearance. 18 Additional information about the properties of the C=O stretching region is provided by the generalized 2D COS analysis of the spectra. The electrode potential constituted the external perturbation for this analysis (see ref [39]). The 2D COS spectra are shown in Figures 8 a and b in which red color indicates positive bands and blue color negative bands. The synchronous spectra are plotted in Figure 8a. The bands located at the diagonal are auto-correlation bands, the bands located off diagonal are cross-correlation bands. All C=O stretching bands of guanine and cytosine have strong auto-correlation bands. No auto-correlation band is seen for cytosine ring vibrations at ~1612 cm-1. The synchronous spectrum is dominated by strong red square at frequencies between 1675 and 1645 cm-1. It indicates that the non-bonded guanine band at ~ 1678 cm-1 and cytosine bands at 1656 and 1645 cm-1 are strongly correlated and change in the same direction. The crosscorrelation bands of 1675, 1656 and 1645 cm-1 with the complexed guanine bands at 1700 and 1690 cm-1 indicate that these changes are also correlated but their negative sign indicates that they change in the opposite direction. This result is consistent with plot of intensities in Figure 7. 1/ cm-1 1600165017001750 2 / cm-1 1600 1650 1700 1750 a) 1 / cm-1 1600165017001750 2 / cm-1 1600 1650 1700 1750 b) Figure 8. Synchronous ,a) and c), and Asynchronous 2Dcos analysis of the PM-IRRA spectra of the 1800-1600 cm-1 , a) and b), and the 1600-1500 cm-1, c) and d), sub-regions of the polar head spectra of a mixed monolayer of DP-CDP:DPPC 7:3 incubated in the 19 presence of 25 M guanine and transferred to a gold (111) electrode at the equilibrium pressure. The external variable is the decreasing potential and the reference is the averaged spectrum shown at the axis. Positive cross-correlations are represented in red and negative cross correlations are represented in blue. The diagonals of the surface plots correspond to the autocorrelations. The asynchronous spectrum Figure 8b provides information about sequential changes of IR bands with potential. It has only cross-correlation bands. The positive sign of the cross-correlation band denoted as (v1, v2), with v1 and v2 being wavenumbers on the x and y axes, indicates that changes at v1 take place before changes of v2. The negative sign indicates that v2 precedes v1. Table SI 1 of the SI lists signs of the cross-correlation bands. The cross correlation bands at (1645, 1700), (1656, 1700) and (1678, 1700) are negative showing that the potential induced changes of the complex band at 1700 cm-1 precedes changes of the non-bonded guanine and cytosine C=O bands. However, the (1690, 1700) band is positive indicating that changes of the “weakly bonded complex” band take place before changes of the “strongly bonded complex” band. Further, (1645, 1678) and (1658, 1678) bands are positive pointing to changes in the cytosine bands are driving the change of the “non bonded” guanine band. The negative sign of the (1645, 1658) band suggests that changes of the 1658 cm-1 band precede changes of the 1645 cm-1 and that the changes of the two bands are not fully synchronized. Interestingly, the asynchronous spectrum has a negative band at (1612, 1658). The 1612 cm-1 ring vibrations band was absent in the synchronous spectrum indicating that its changes are predominantly asynchronous and are to some extent following changes of the main C=O band of cytosine. In conclusion, the 2D COS analysis showed that the potential controlled changes of the spectra are initiated by the changes in the complex followed by the changes of the main cytosine band at 1658 cm1. The other bands follow the changes of the main cytosine band. Below we will show that this behavior correlates with changes of the orientation of the cytosine moiety with potential. 3.3.2.2 Cytosine and guanine ring vibrations region 20 For selected potentials between 0.36 and -0.24 V vs SCE, Figure 9 compares deconvoluted spectra in the 16001450 cm-1 region, which contains bands corresponding to ring vibrations of cytosine and guanine moiety. The strong bands at 1525 and 1505 cm1 correspond to cytosine moiety and weak bands at 1550, 1567 and 1578 cm-1 correspond to guanine. The bands below 1500 cm-1 correspond to sugar ring vibrations. The assignment of various bands in the PM IRRAS spectra is summarized in Table SI3 of SI. In addition, Table SI4 lists uncertainties of the band positions, their FWHM and uncertainties of FWHMs. 1550 1500 1450 0.36 V vs SCE 0.002 a.u. S 0.26 V vs SCE 0.16 V vs SCE 0.06 V vs SCE -0.04 V vs SCE -0.14 V vs SCE wavenumber / cm-1 -0.24 V vs SCE 21 Figure 9. Deconvolution of the PM-IRRA spectra in the potential range 0.36 to -0.24 V vs SCE obtained for a mixed monolayer of DP-CDP:DPPC 7:3 incubated in the presence of 25 M guanine and transferred to a gold (111) electrode at the equilibrium spreading pressure. Dashed lines are located at 1505, 1525, 1550, 1567 and 1578 cm-1. E / V vs SCE -0.8 -0.4 0.0 0.4 Integrated intensities / cm-1 0.00 0.01 0.02 0.04 0.06 0.08 E-Epzfc / V -0.8 -0.4 0.0 cytosine guanine Figure 10. Integrated intensities vs potential plots corresponding to the individual bands at 1505 cm-1 (back triangles-up), 1525 cm-1 (red circles), 1550 cm-1 (blue triangles-down), 1575 cm-1 (green squares)), obtained from deconvolution of the PM-IRRA spectra of a mixed monolayer of DP-CDP:DPPC 7:3 incubated in the presence of 25 M guanine and transferred to a gold (111) electrode at the equilibrium pressure. Figure 10 plots integrated intensities of cytosine and guanine bands as a function of the electrode potential. The cytosine bands increase while guanine bands decrease when potential becomes more negative. The most dramatic change takes place at the potential of zero charge at the interface. The integrated intensities of the cytosine band can be used to 22 determine orientation, and rotation of the cytosine moiety as a function of the applied potential using procedure developed in [5,17,18] and described in detail in the Supporting Information. Figure 11. a) Potential dependence of the tilt angles of the vibration tensors of the bands at 1505 (green triangles) and 1525 (blue diamonds) relative to the normal direction to the electrode surface, b) schematic representation of the directions of the transition dipoles of the vibrations at 1505 (green line) and 1525 (blue line) and c) Schematic representation of transition dipole direction of an arbitrary in-plane vibration (𝑂𝐵     ), normal direction to the electrode surface (𝑂𝐴     ), normal direction to the molecular plane (𝐴𝑃     ) and projected normal to the surface onto the molecular plane (𝑂𝑃     ). The integrated intensity is proportional to the cos2θ where θ is the angle between the direction normal to the surface and direction of the transition dipole of a given band. For bands at 1505 and 1525 cm-1 the angles θ calculated from the band intensities are plotted in Figure 11a. The directions of the transition dipoles of the 1505 and 1525 cm-1 bands of the cytosine moiety were determined with the help of DFT calculations in ref [8]. 3 N N H2N O 1 4 2 5 6 7 O P A B    a) b) c) E / V vs SCE -0.8 -0.4 0.0 0.4 / degrees 50 60 70 23 Their positions are shown in Figure 11b. Since the two transition dipoles are located in plane of the molecule one can calculate angle between surface normal and projection of the surface normal onto the plane of the molecule α and angle φ between direction of the transition dipole and projection of the surface normal onto the molecular plane. The definition of angles θ, α and φ is illustrated by Figure 11c and the procedure employed to calculate these angles is described in the Supporting Information. The angle α is a measure of the tilt of the cytosine moiety with respect to the surface normal while angle φ provides information about its rotation. -0.8 -0.4 0.0 E / V vs SCE -0.8 -0.4 0.0 0.4 cytosine / degrees 40 60 80 E - Epzfc / V Figure 12. Potential dependence of the tilt angles the cytosine moiety molecular plane for the mixed monolayer of DP-CDP:DPPC 7:3 incubated in the presence of 25 M guanine and transferred to a gold (111) electrode at the equilibrium pressure (filled red circles) and 24 in the absence of guanine (hollow red circles) and for a monolayer of pure nucleolipid incubated in the presence of guanine (black filled circles). Figure 12 compares changes of the tilt of the cytosine plane in the mixed DPCDP:DPPC 7:3 incubated in the presence of 25 M guanine, to the orientation in the pure DP-CDP with guanine and in the mixed monolayer without guanine. The plane of cytosine moiety assumes much smaller angle with respect to the normal in the mixed monolayer with guanine than in the mixed monolayer without guanine or pure nucleolipid monolayer with guanine. This trend correlates with the differences between charge of the diphosphate group shown in Figure 4b. The diphosphate group is fully dissociated in the mixed monolayer with guanine and the charge of the polar head group of the nucleolipid is -2, while it is partially dissociated in the two other monolayers where the charge of the diphosphate group is ~ -1.5. The polarity of the head group seems to have influence on the orientation of the cytosine moiety. The orientation of the cytosine moiety also depends on the presence of guanine. It seems that it opens the polar head region assisting in dissociation of the diphosphate group. In the absence of guanine, the cytosine moiety is nearly parallel to the monolayer surface preventing dissociation of the diphosphate group. Smaller tilt of cytosine facilitates sodium transfer into the bulk of the solution. Figure 12 provides second important information. The tilt angle of cytosine moiety is much higher at positive (E-Epzfc) where Watson-Crick complex is stable, indicating that complex assumes a small angle with respect to the plane of the monolayer. For the 1505 and 1525 cm-1 bands, changes of the rotation angles φ with potential are plotted in Figure SI 8a of the Supporting Information. For comparison, Figure SI8 b plots the rotation angles for the mixed monolayer without guanine and Figure SI8 c the rotation angles in pure monolayer of the nucleolipid with guanine. Relative to significant rotations of the plane of cytosine moiety observed in Figures SI 8b and c the rotation in the mixed monolayer with guanine is significantly smaller. Apparently, smaller tilt with respect to the surface normal prevents rotation of the cytosine moiety. The information concerning sequential changes of the ring stretching bands provides 2D COS analysis shown in Figures SI 9a and b. The synchronous spectra in Figure SI 9a show strong auto-correlation cytosine and weak guanine bands. However, the 25 guanine bands are present in the cross-correlation peaks that have negative sign indicating that changes of guanine and cytosine bands are in opposite direction. Asynchronous spectra are shown in Figure SI 9b. The (1505, 1525) cross correlation band is negative indicating that changes in 1505 band follow changes of the 1525 band. However, the (1525, 1550) and (1505, 1550) bands are negative suggesting that changes of the guanine band take place earlier than cytosine bands. This trend is consistent with the behavior of C=O stretch bands discussed earlier. 4.Summary and Conclusions To understand the mechanism of Watson-Crick complex formation between cytidine moiety of a monolayer of nucleolipid DP-CDP and guanine (its complementary base) several experimental techniques and several compositions of the monolayer were employed. The optimal condition for the transfer of the monolayer from the air-solution interface to the gold electrode surface were determined by recording compression isotherms. Monolayers of pure nucleolipid and mixed monolayers of nucleopipid and phospholipids (DPPC) without and with guanine were investigated. Mixed monolayer with composition (7:3) DP-CDP:DPPC molar ratio displayed ideal mixture properties and provided space between polar heads of the nucleolipid. The effect of the static electric field on the properties of the monolayer and the Watson-Crick complex formation was determined by measuring charge density at the electrode surface and the potential of zero charge at the interface. The electrosorption valencies and the charge at the diphosphate group of DP-CDP were calculated. These numbers showed that in the presence of guanine the diphosphate group is totally dissociated while in the monolayer of pure nucleolipid it is partially dissociated. Partial dissociation of the diphosphate group was also observed for the mixed monolayer without guanine. This behavior was explained with the help of PMIRRAS experiments which showed that cytosine moiety has large tilt angles when the diphosphate group is partially dissociated and smaller tilt angle when it is fully dissociated. The orientation of the cytosine moiety nearly parallel to the monolayer surface constitutes the barrier for exchange of sodium ions between the diphosphate group and the bulk of the solution. In the mixed monolayer with guanine, the guanine molecules promote smaller tilt angle of the cytosine moiety. Since no such effect was observed for a pure monolayer of 32 [40] X. Li, K. Feng, L. Li, L. Yang, X. Pan, H.S. Yazd, C. Cui, J. Li, L. Moroz, Y. Sun, B. Wang, X. Li, T. Huang, W. Tan, Lipid-oligonucleotide conjugates for bioapplications, Natl. Sci. Rev. 7 (2020) 1933–1953. doi:10.1093/nsr/nwaa161. [41] M. P. Thompson, M.-P. Chien, T.-H. Ku, A. M. Rush, N. C. Gianneschi, Smart Lipids for Programmable Nanomaterials, Nano Lett. 10 (2010) 2690–2693. doi:10.1021/nl101640k.