Repositorio Institucional de Documentos

Silvia Caballero Mancebo – Final Master Project i Acknowledgements A number of people deserve thanks for their support and help during this Master thesis. It is therefore my greatest pleasure to express my gratitude to them all. Foremost, I would like to express my sincere gratitude to my advisor Dr. Anabel Gracia Lostao for her support, her patience, motivation and her insight that lead to the original proposal of studying for the first time the nanomechanical properties of an enzyme-substrate complex. I wish to thank Dr. Marta Martínez Júlvez (at the Department of Biochemistry and Cellular and Molecular Biology) for the laboratory materials she provided me with. My appreciation also goes out to José Luis Díaz for his time, assistance and good advices dealing with the AFM. I would also like to extend my gratitude to the whole Faculty of this Master, especially to the Master Coordinator, Dr. Pilar Cea whose dedication and hard work has certainly helped us along the way. She always took time out from her busy schedule to give a helping hand both academically and personally. I would like to thank Cátedra SAMCA for granting the SAMCA Nanotechnology Scholarship to join this program. Thanks to my fellow lab mates. Carlos, thank you so much for introducing me to the exciting world of force spectroscopy and for guiding me during my first steps with the AFM. Without your unparalleled support, this work would not have gone ahead. Mari Carmen, thank you for your optimism and contagious joy, for your uncontrollable desire to help and for those hours of company (you can’t imagine how sorry I am for missing that day, congratulations!). And finally Miren, thank you for your desire to learn, for your hard questions, and above all, for your great help this summer. Good luck in Barcelona! My special "thanks" (you know how!, no pun intended) to my classmates. I won’t forget the endless lab demos, last minute reviews and our after lunch talks. Thank you for listening to me cheering me up when I needed it. I expect visits!! A special mentioned is due for my friends. You have forgiven my absences, born my burdens and supported me in everything. Paula, it has been great meeting you and living with you. Without you, this year could have been unbearable. Silvia Caballero Mancebo – Final Master Project ii Last but not least, I owe more than thanks to my parents and sister. There is so much for me to thank them that I do not know where to start. They have always supported me, no matter what and have been pivotal in all my successes. Elena, although distance is going to be a hard thing to overcome, I know I will always count on you. Best of luck and reach for the stars. You deserve this and more! Silvia Caballero Mancebo – Final Master Project iii Table of contents Acknowledgements.............................................................................................................................. i Abbreviations ..................................................................................................................................... v Abstract ............................................................................................................................................ viii 1. INTRODUCTION ...................................................................................................................... 1 1.1. Atomic Force Microscopy ..................................................................................................... 1 1.1.1. Basic principles of AFM .............................................................................................. 2 1.1.2. AFM operating modes .................................................................................................. 3 1.2. Atomic Force Spectroscopy (AFS) ....................................................................................... 5 1.2.1. Force-distance curves ................................................................................................. 6 1.2.2. Theoretical models to analyse AFS data .................................................................... 7 1.3. Immobilization strategies ...................................................................................................... 9 1.4. Ferrodoxin NADP+ reductase .............................................................................................. 10 2. OBJECTIVES .......................................................................................................................... 13 3. MATERIALS AND METHODS ............................................................................................ 14 3.1. Protein labelling .................................................................................................................. 14 3.2. FNR-PDP spectrophotometric quantification ..................................................................... 15 3.3. Enzymatic activity measurements ....................................................................................... 15 3.4. FNR-PDP immobilization on mica ..................................................................................... 17 3.5. AFM topography and scratching ........................................................................................ 18 3.6. AFM tip functionalization ................................................................................................... 19 3.7. AFS experiments ................................................................................................................. 21 4. RESULTS ................................................................................................................................. 23 4.1. Protein labelling and functionalization................................................................................ 23 4.2. AFM imaging ...................................................................................................................... 24 4.3. Force Spectroscopy ............................................................................................................. 26 Silvia Caballero Mancebo – Final Master Project iv 4.4. Dissociation kinetics for the FNR:NADP+ complexes ........................................................ 28 5. DISCUSSION ........................................................................................................................... 31 6. CONCLUSIONS ...................................................................................................................... 35 7. REFERENCES ......................................................................................................................... 37 Silvia Caballero Mancebo – Final Master Project v Abbreviations Abs Absorbance AFM Atomic Force Microscopy APTES (3-Aminopropyl)triethoxysilane C Concentration cyt c Cytochrom c Da Dalton DFS Dynamic Force Spectroscopy DNA Desoxyrribonucleic acid DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid ET Electron transfer F Applied force F* Most probable unbinding force FAD Flavin adenine dinucleotide (oxidised) FADH Flavin adenine dinucleotide (reduced) Fd Ferredoxin Fld Flavodoxin FMN Flavin mononucleotide FNR Ferredoxin-NAPD+ mnReductase Fz Force-distance curves kB Boltzmann’s constant Silvia Caballero Mancebo – Final Master Project vi kcat see TON koff Dissociation rate constant at zero applied force l Path length NADP+ Nicotinamide adenine dinucleotide phosphate (oxidised) NADPH Nicotinamide adenine dinucleotide phosphate (reduced) NHS N-hydroxysuccinimide PDB Protein Data Bank PBS Phosphate Saline Buffer PEG Poly(ethylene glycol) R Loading rate rpm Revolutions per minute SDS Sodium dodecyl sulfate SMFS Single Molecule Force Spectroscopy SNOM Scanning Near Field Optical Microscopy SPM Scanning Probe Microscopy STM Scanning Tunnelling Microscopy Sulfo-LC-SPDP sulfosuccinimidyl 6-[3’(2-pyridyldithio)-propionamido] hexanoate T Temperature τ Lifetime TON Turnover number wt wild type ε Attenuation coefficient Silvia Caballero Mancebo – Final Master Project vii Abstract The aim of this thesis is to study for the first time the interaction forces between an enzyme and its enzymatic substrate using a single molecule approach, Atomic Force Spectroscopy (AFS). The enzyme chosen is the Ferredoxin-NADP+ reductase (FNR) involved in the photosynthetic chain accepting electrons from flavoproteins ferredoxin (Fd)/ flavodoxin (Fld) and transfering them to its substrate, NADP+ which is then reduced to NADPH. To that matter, the enzyme FNR was immobilized on the surface of flat mica pieces and its substrate, NADP+, was immobilize on the surface of prefunctionalized AFM tips carrying a maleimide-PEG linker. The effect that the immobilization process had on the structure and activity of FNR was tested by AFM imaging and enzymatic assays, respectively, showing only a slight decrease on the catalytic activity of the enzyme. Both the functionalized tip and sample were subsequently used for the AFS measurements. These experiments gave data consisting of hundreds of Fz curves at different loading rates that were used to construct histograms that provided the values for the most probable unbinding force at each different loading rate. 136 pN was obtained as the most probable unbinding force at a loading rate of 10 nN/s. The unbinding data was analysed in the framework of the Bell-Evans model in order to obtain mechanostability parameters associated with the transition state of the dissociation process. The results showed a single regime, associated to a one-step dissociation kinetic for the FNR:NADP+ complex. The dissociation rate constant, koff, and the xβ (width of the energy barrier) gave 0.0198 s-1 and 0.0205 nm, respectively. These values were indicative of the mechanical stability of the FNR:NADP+ complex, higher than that of FNR with either one of its protein partners. This, together with the lifetime value of the complex, 50.6 s, corroborated our expectations of a strong specificity of FNR for NADP+. Silvia Caballero Mancebo – Final Master Project Introduction 7 Fig. 5. Diagram of a typical force-distance curve showing a specific unbinding event (Bizzarri and Cannistraro, 2010). 1.2.2. Theoretical models to analyse AFS data The most common approach used to describe the behaviour of force-induced transformations is based on the seminal works of Bell. Bell’s model was proposed for the analysis of cell-cell adhesion bonds and provides a phenomenological description of the effect that mechanical forces have on molecular interactions (Bell, 1978). The main idea is that a forced dissociation can be described as a thermally activated escape over a transition-state barrier and treated within the frame work of the reaction rate theory. At equilibrium, a pair of interacting molecules changes from the bound (AB) to the unbound (A+B) state proceeding over a single transition state (A:B) with a certain activation energy barrier ( Δ G*) (Fig. 5). When an external force is applied, the energy profile of the unbinding process changes by lowering the activation energy of the transition state by F·xβ (where F is the applied force and xβ the width of the energy barrier). Now, the dissociation rate constant depends on the applied force, F, as follows (Bizzarri and Cannistraro, 2009): ( ) ( ⁄) (Eq.4) where koff is the dissociation rate at the equilibrium, at zero applied force, with koff = koff (0). Evans and Ritchie, starting from this assumption, derived a description of the unbinding process in terms of a crossing over a single, sharp barrier through the application of a timedependent force, F(t). In consequence, this model provides the dependence of the unbinding force on the loading rate, R = dF/dt, at which the external force is applied. R is usually given by the Silvia Caballero Mancebo – Final Master Project Introduction 8 product between the tip retraction speed and the spring constant of the cantilever (Bizzarri and Cannistraro, 2010). The Bell-Evans model is based on the following assumptions: - The loading rate during a measurement is constant. - The unbinding process of a single ligand-receptor pair occurs. - The rupture time is longer than the diffusional relaxation time, and any rebinding process is neglected. - The pulling coordinate coincides with the reaction coordinate. - The width of the energy barrier, xβ, is independent of the applied force. On this basis, the most probable unbinding force, F*, can be obtained by calculating the maximum of a probability distribution of the unbinding force as follows (Evans and Ritchie, 1997): ( ) ( ) (Eq.5) F* being the most probable unbinding force, kB is the Boltzmann constant, T the temperature at which the experiment has been conducted, R the loading rate and koff and xβ the parameters that characterize the complex dissociation, in the case of this project, the FNR:NADP+ enzymatic system (see section 1.2). The former expression predicts a linear relationship between the most probable unbinding force, F*, and the natural logarithm of the loading rate, R. By plotting F* as a function of lnR, the equilibrium parameters koff and xβ can be extracted from the intercept and the slope of the fitting linear curve, respectively (Fig. 6A). In some cases this function exhibits two or more distinct linear regimes (Fig. 6B). Then, two different sets of koff and xβ values can be extracted by two independent linear fits. Such behaviour may be traced back to the presence of two intermediate states in the unbinding process (Bizzarri and Cannistraro, 2009). Fig. 6. (Left) Scheme of the energy profile for a dissociation process of a biomolecular complex. (Right) Corresponding trend for the most probable unbinding force as a function of the loading rate. Silvia Caballero Mancebo – Final Master Project Introduction 9 1.3. Immobilization strategies The detection of unconstrained ligand-receptor recognition requires a specific linker design (Hinterdorfer et al., 1998). This is one of the most time-consuming steps in an AFM experiment in which the receptor has to be immobilized on a nanometrically flat surface and its ligand has to be attached to the AFM tip. Generally speaking, there are two main approaches for immobilization of molecules to both the substrate and the tip: either electrostatically or covalently. The main difference between them is the type of established bond. Covalently coupling of both receptor and ligand guarantees that they are sufficiently tight attached as covalent bonds are around 10 times stronger than typical receptorligand bonds (Hinterdorfer and Dufrêne, 2006). Loose receptor or ligand fixation may lead to their pull-off from the surface or tip respectively upon the recognition event, resulting in the blocking of the receptor-ligand recognition. Fig. 7. Different immobilization strategies on AFM tips. Moreover, the ligand has to be provided with enough motional freedom around the tip so that the recognition process is not influenced by steric restrictions (Bizzarri and Cannistraro, 2012). This is done by using a spacer between the tip and the ligand. Poly(ethylene glycol) (PEG) (Fig. 7C) is used most of the times as crosslinker. It is a water-soluble, nontoxic and non-adhesive flexible polymer. On the one hand, PEG spacers facilitate the encounters thank to an increase of the motional freedom of the ligand. On the other hand, the use of PEG as spacer also solves one of the main drawbacks of force spectroscopy experiments: providing a fingerprint for identifying the specific events corresponding to the rupture of events. When the receptor on the surface of the sample binds to a ligand on the tip, an attractive force develops upon withdrawing of the tip from the sample B C A Silvia Caballero Mancebo – Final Master Project Introduction 10 (retrace line in Fig. 5) and increases with tip-sample separation. The same happens when a nonspecific interaction between the cantilever and the sample occurs. It is the shape of the curve that determines whether that peak corresponds to a specific interaction or not: the shape of curve in a specific event is determined by the elastic properties of the flexible PEG linker that shows a nonlinear, parabolic-like profile that reflects the increment of the spring constant of the linker during extension. Therefore, specific receptor-ligand recognition events are easily distinguishable from linearly shaped non-specific tip-sample adhesion events (Fig. 8) (Hinterdorfer et al., 2000). 1.4. Ferrodoxin NADP+ reductase In this project, the system under study is the enzyme Ferredoxin NADP+ reductase (FNR). FNR is a 36 kDa protein (PBD code: 1QUE) that can be found in plants, bacteria and some algae. The FNR used in this project belongs to the cyanobacteria Anabaena PCC 7119 and plays a crucial role during the electron transfer (ET) in photosynthesis as the last component of the Photosystem I (PSI) (Fig. 9). Fig. 8. Real AFS force curves. (A) Specific unbinding event due to the parabolic profile of the peak. (B) Non-specific event since the linear slope of the retrace extends beyond the contact point. B A Fig. 9. Scheme of the light reactions during photosynthesis. Silvia Caballero Mancebo – Final Master Project Introduction 11 It catalyses the sequential transfer of two electrons from two different Ferredoxins (Fd) molecules to a molecule of NADP+ (nicotinamide adenine dinucleotide phosphate) in order to reduce it to NADPH (Fig. 10A) (Jelesarov and Bosshard, 1994). NADPH will then contribute to the Calvin Cycle as a major source of reducing power. Fig. 10. (A) Diagram of the redox process in which FNR is involved during photosynthesis. (B) Overall reaction catalysed by FNR. The main structural feature of FNR is that it shows two domains: a flavin-binding domain (more precisely a flavin adenine dinucleotide, FAD) in which the cofactor is accommodated and the NADP+-binding domain which is responsible for the binding of the nucleotide substrate (Fig. 11) (Medina and Gómez-Moreno, 2004). Fig. 11. (A) Tridimensional structure of FNR (PBD: 1QUE). FAD is shown in green sticks. (B) FAD and NADP+ binding sites. The protein surface is shown in yellow and FAD (orange) and NADP+ (shadowed white) are shown in sticks. A B Silvia Caballero Mancebo – Final Master Project Introduction 12 The interactions between FNR and its protein partners (Fd and its substitutive flavoprotein Flavodoxin (Fld) in iron-deficient environments) have been extensively studied (Hurley et al., 2006; Medina and Gómez-Moreno, 2004). It is known that both Fd and Fld interact electrostatically with FNR at the same interaction surface, and that the interaction between the former is stronger mainly due to a more specific interaction (Martínez-Júlvez et al., 2009). In order to get a further insight into the physico-chemical properties of these interactions, our group used Dynamic Force Spectroscopy (DFS) to analyse the nanomechanical properties of the protein complexes formed by FNR and found that the mechanical stability of the FNR complex with its natural electron donor, Fd, is almost 3-fold higher and its lifetime longer than in the case of the complex with Fld (GómezMoreno and Lostao, 2013; Marcuello et al., 2012; Gómez-Moreno et al., 2011; Marcuello et al., in preparation). Regarding the interaction of FNR and its enzymatic substrate NADP+, which is the aim of this project, it is known that certain aminoacid residues in the NADP+ binding pocket of FNR interact directly with the 2’-P group of NADP+. In addition, some structural rearrangements in the protein that assure the coenzyme specificity are also needed (Medina and Gómez-Moreno, 2004). We proposed the following working hypothesis: due to the high specificity that the enzyme FNR shows for its enzymatic substrate, NADP+, their interaction in terms of unbinding force for the FNR: NADP+ complex may be stronger than that seen between FNR and its electron donor Fd. With this study, we intend to fully characterize the interactions between FNR and its substrate and protein partners, broadening the extensive knowledge that our group has gather of this enzyme along the years. On the other hand, it should be highlighted that this is the first nanomechanical study on an enzyme-substrate system. Silvia Caballero Mancebo – Final Master Project Objectives 13 2. OBJECTIVES These were the objectives proposed at the beginning of the project: 1. Immobilize the protein Ferredoxin-NADP+ Reductase on the surface of flat mica pieces in order to be used for molecular recognition studies preserving its functionality. 2. Design a reproducible protocol for the immobilization of nicotine adenine-based nucleotides (such as NADP+) on AFM tips for their use as functionalized probes during the force measurements. 3. Study for the first time the molecular recognition process as well as the interaction forces established between the enzyme FNR and its enzymatic substrate NADP+. To that matter, the most probable unbinding force has to be obtained at different loading. 4. To determine the specificity of the binding events obtained in the AFS experiments by means of blocking experiments. 5. To obtain the mechanostability parameters of the FNR:NADP+ complex using the Ritchie-Evans model in order to be able to elucidate the energy landscape of the complex. Silvia Caballero Mancebo – Final Master Project Materials and Methods 14 3. MATERIALS AND METHODS 3.1. Protein labelling The protein FNR from Anabaena PCC 7119 was purified from Escherichia coli BL21 cultures in the Biochemistry and Cellular and Molecular Biology Department of the University of Zaragoza by Dr. Marta Martínez Júlvez. Then it was labelled using the heterobifunctional crosslinker sulfoLC-SPDP (sulfosuccinimidyl 6-[3’(2-pyridyldithio)-propionamido] hexanoate; Pierce) (Fig. 12). This crosslinker contains an amine-reactive end, an N-hydroxysuccinimide (NHS) ester group, that binds to primary amines (-NH2) from the lateral chain of lysine residues at the surface of FNR. The other terminal group (a 2-pyridyldithiol group) is a sulfhydryl-reactive portion which reacts optimally with sulfhydryl groups resulting in displacement of a pyridine-2-thione molecule and the formation of a disulfide bond. Fig. 12. Sulfo-LC-SPDP crosslinker. These are the steps followed for the FNR labelling: 1) SPDP binding to FNR: A fresh 20mM SPDP solution in 50 mM Tris-HCl, pH 8 was prepared. Then 15 µl of this solution were added per milligram of protein to be labelled. This solution was incubated during 50 min at room temperature under mild stirring. The reaction produces tagged-species of free FNR carrying a C9-long arm ending in a 2pyridyldithiol group (FNR-PDP) (Fig. 13). Fig. 13. Reaction of the amine groups on the surface of FNR with the NHS group from SPDP to yield labelled FNR (FNR-PDP). Silvia Caballero Mancebo – Final Master Project Materials and Methods 15 2) FNR-PDP purification: FNR-PDP molecules were isolated from unreacted SPDP and other molecules and concentrated using a Microcon-10kDa Centrifugal Filter Unit. Centrifugation was performed at 4 ºC and 4000 rpm during 15 minutes. 3) Finally, the concentrated protein was resuspended in 50 mM Tris-HCl, pH 8, quantified and stored at -20 ºC. 3.2. FNR-PDP spectrophotometric quantification The protein concentration was quantified using a Cary 100Bio spectrophotometer from Varian and quartz cuvettes. Spectra acquisition of free and labelled FNR (FNRwt and FNR-PDP respectively) as well as the base line correction were performed in 50mM Tris-HCl, pH8 using a scan rate of 50 nm/min and a wavelength range between 200 and 800 nm. According to the Lambert-Beer law, the absorbance of a substance at a certain wavelength can be related to the concentration of that substance in the cuvette as long as the absorbance is kept as a linear function of the concentration. It is known that FNR has a characteristic absorbance peak at 458 nm (Medina and Gómez-Moreno, 2004) and therefore, knowing its molar absorption at this wavelength, FNR concentration can be calculated: (Eq. 6) ε being the molar absorption coefficient of FNR at 458 nm (9.4 mM-1cm-1), C the concentration of FNR in the cuvette and l the path length (1 cm). Once the protein was labelled and quantified, it was again filtered using a Microcon-10kDa Centrifugal Filter Unit. Centrifugation was performed at 4 ºC and 4000 rpm during 15 minutes. Then the protein was stored in phosphate buffer saline (PBS) at -20 ºC. 3.3. Enzymatic activity measurements In order to determine if the labelling and functionalization process affects functionality and enzymatic activity of FNR, kinetic measurements were performed using both free and labelled protein in solution. Silvia Caballero Mancebo – Final Master Project Materials and Methods 16 A great variety of enzymatic activities have been described for FNR. Among them, the cytochrome с reductase (cyt c) activity is one of the most used in order to describe the catalytic activity of FNR. Taking advantage of the fact that cytochrome с of mammals reduces Fd quite fast, this activity can be measured by monitoring the absorbance increment at 550 nm, being the molar absorption coefficient of cytochrome с at this wavelength around 20 mM-1cm-1. Fig. 14. Electronic transference steps described for the cytochrome с reductase. The activity measurements were performed in a Cary 100Bio spectrophotometer from Varian. The reaction volumes used for these assays and the starting concentrations are shown in table 1. The measurement has to be taken half a minute after all reactants have been mixed. Although FNR may directly reduce some types of cytochrome such as cytochrome f (Zanetti and Forti, 1966), it is very inefficient reducing cytochrome с from mammals when Fd is absent. This is the reason why Fld or Fd have to be added latest into the cuvette. Reactives Reference FNRwt FNR-PDP 50 mM Tris-HCl, pH8 900 µl 648 µl 648 µl Cyt с (7.5 mg/ml) 100 µl 94 µl 94 µl NADPH (2 mM) - 121 µl 121 µl Fld (0.47 mM) - 137 µl 137 µl FNR (4 µM) - 1 µl 1 µl Table 1. Reaction volumes used for the enzymatic activity measurements in solution. In order to be able to calculate the turnover number (TON), also known as kcat, the slope of the graphs obtained at 60 s has to be extrapolated. Silvia Caballero Mancebo – Final Master Project Results 23 4. RESULTS 4.1. Protein labelling and functionalization The objective for tagging the protein with a crosslinker is to immobilize it onto a flat mica surface in order to perform force measurements using an AFM. Protein immobilization in a functional active form is fundamental for its activity. Exceptionally, a methodology for a controlled and oriented FNR immobilization towards its protein partners was pursued, as previously developed by our group in which the binding site of FNR protein partners was protected during the labelling process. This allowed to largely increase the effectiveness in AFS (Marcuello et al., 2012). However, in this work, as usual in AFS, a random labelling was performed due to the fact that in this case the interaction we were interested in studying was that between FNR and its substrate NADP+ whose binding site defers from that of FNR protein partners. The results, in terms of percentage of tip approaches that produced specific events, obtained using a random labelling and attachment ranged between 4 and 18%. These data are similar to those found in the literature for this kind of experiments using random procedures. With the purpose of assuring the integrity of the tagged species of FNR, a steady-state enzymatic assay was performed as explained in section 3.3 in which the cytochrome c reductase activity in solution of labelled FNR (FNR-PDP) was compared to that of the free enzyme (FNRwt) (Fig. 20). Fig. 20. Cytochrome c reductase activity assay results. As seen, the enzymatic activity of FNRwt reached higher values in terms of absorbance at 550nm after one and a half minutes. Silvia Caballero Mancebo – Final Master Project Results 24 The catalytic activity of an enzyme is measured in terms of its turnover number (TON), also known as kcat, and it is defined as the number of molecules that the enzyme converts into product per catalytic site per unit of time. FNRwt exhibited a TON value of 7.56 s-1 while FNR-PDP showed a rate of 5.08 s-1 (Table 2). All measurements were performed at a FNR concentration of 4 nM that is the usual concentration for measuring FNR activity in solution. Although there is a 1.5-fold decrease of the turnover number in labelled FNR with respect to free FNR, the steady-state enzymatic assay gave positive data showing, on the one hand, that the functionality of the samples was only slightly affected by the tagging process and, on the other hand, that the enzyme is still able to recognize and bind efficiently to its substrate. This concurs with previous results in our group which showed that random labelling processes always affects enzyme activity somehow (Marcuello et al., 2012). Sample Δ Abs (mUA) [FNR] (nM) TON, kcat (s-1) FNR wt 0.03629 4 7.56 FNR-PDP 0.02439 4 5.08 Table 2. Kinetic parameters of FNR-PDP for cytochrome c reductase activity. 4.2. AFM imaging AFM imaging was used as a tool to monitor the results of the immobilization process of FNR on the surface of flat mica pieces at a molecular level. Topography images showed that the enzyme forms a homogeneous monolayer when at least 4 µg of protein were added to each mica piece of approximately 1 cm2. (Fig. 21). The rest of the functionalizations were performed adding 4 µg of protein to each mica piece. Topography images of the functionalized mica surfaces show a height between 8 and 12 nm (Fig. 21 and 22). This height is consistent with the expected length of the reaction products used during the functionalization and coincides with that obtained in the scratching profile (Fig. 22B). Another aspect of the functionalized mica worth mentioned is that the layer, as imaged by AFM, is not regular but it rather displays mounts and valleys randomly distributed over the surface. Silvia Caballero Mancebo – Final Master Project Results 25 Fig. 21. AFM topography images of functionalized mica with different amounts of FNR. (A) 1 µg, (B) 1.5 µg, (C) 2 µg, (D) 4 µg and (E) 6 µg. A homogenous monolayer appears when at least 4 µg of FNR are added. Fig. 22. (A) Distribution of FNR-PDP covalently bond on mica and the corresponding height profile showing a maximum height of 11.43 nm. (B) Scratching on the functionalized mica and its height profile showing a maximum height between 11 and 12 nm. Silvia Caballero Mancebo – Final Master Project Results 26 4.3. Force Spectroscopy Force data from the analysis of the force-distance curves were obtained as explained in section 3.7. A typical mica-FNR:NADP+-tip force scan is depicted in Figure 23. Starting from the zeroforce point, the tip is moved closer to the sample with an increment in the force (blue line) until tip and sample come into contact. Pushing the tip further towards the surface requires higher forces causing the bending of the cantilever. Retraction of the tip (red line) produces a sharp jump (the peak seen in the curve) indicating that a sudden release has occurred between the tip and the sample. This is due to the rupture of the bond between the two interacting molecules (fu). The specific length of the peak (lu), coincident with that of the stretched maleimide-PEG linker (approximately 20 nm), indicates that the force, fu, can be attributed to a specific event. Fig. 23. Force curve showing a single specific event for a FNR:NAPD+ complex. Molecules are brought into contact when the tip is moved towards the sample (blue line). In the retraction line (red line) the force unbinding force required is shown (fu). The total length at which the unbinding occurs (lu) is defined as the unbinding length. Figure 24 shows the histograms built at the four different loading rates used in the measurements. Each histogram represents the relative frequency (in terms of percentage) of specific unbinding events versus the unbinding force. As it can be seen, data in all histograms can be fit under two peaks. The first peak indicates the relative frequency of events that occur at the displayed force and accounts for those interactions in which a single FNR molecule interacts with a single NADP+. Those rupture events observed at higher force values are the ones in which two proteins on Silvia Caballero Mancebo – Final Master Project Results 27 the mica are binding to two different NADP+ molecules on the AFM tip. This is the reason why the second peak appears at force values that are double of those for the single event. Fig. 24. Force distributions obtained at loading rates of 6 (A), 10 (B), 20 (C) and 78 nN/s (D) for the FNR:NADP+ complex. The most probable unbinding force is indicated for each distribution. The results show that the bond probability increases with the loading rate for the FNR:NADP+ complex, as more multiple events are seen at higher velocities. This suggests that the faster the tip approaches to the surface the more it favours the formation of bonds between the enzyme and its substrate, thus the number of multiple events increases. It is worth mentioned that at a loading rate of 10 nN/s (Fig. 24B), the second peak is too small to be statistically significant, however, it follows the tendency of multiple events appearing at every loading rate. The fact that multiple events are so uncommon at this loading rate may indicate that this is the optimal loading rate for this complex under the conditions it was tested. Silvia Caballero Mancebo – Final Master Project Results 28 Regarding the force data, the intermolecular forces fall into the expected trend considering the fact that the most probable unbinding force increases with the loading rate. The corresponding mean rupture forces, obtained by fitting the histograms to a Gaussian function, were 102±28 pN (R = 6 nN/s), 136±34 pN (R = 10 nN/s), 363±94 pN (R = 20 nN/s) and 587±90 pN (R = 78 nN/s), Both the specificity of the interaction probed in this experiment and the quality of the experimental design were evaluated by means of blocking experiments at a loading rate of 10 nN/s, as it has been stated as the standard loading rate for biological complexes. These results are summarized in Figure 25. Very similar data was observed between the blocked and the non-blocked samples in terms of the most probable unbinding force, as shown in the Figure. This indicates that both interactions are within the same range of forces assuring the quality of the experiment. Moreover, as expected, the rupture events frequency decreases (from 18 to 4%) for the blocked sample indicating the specificity of the forces recorded. Fig. 25. Force distributions for the non-blocked (A) and the blocked (B) samples obtained at a loading rate of 10 nN/s. 4.4. Dissociation kinetics for the FNR:NADP+ complexes The energy landscape of a bond rupture explored by AFS defines the force-driven pathway along the pulling direction until the bond rupture. A typical energy landscape is a one-dimension plot representing the energy of the system versus the reaction coordinates (Kramers, 1940). The shape of this landscape is thus constituted by the height of the energy barrier and the energy barrier width between the valley and the summit of the peak. The height of the energy barrier is characterized by the koff value, whereas the energy barrier width is described by the xβ parameter. Silvia Caballero Mancebo – Final Master Project Results 29 In order to calculate both values, the most probable unbinding force is represented as a function of the loading rate (Fig. 26). This plot exhibits one linear regime. Such behaviour can be traced back to the presence of one intermediate state in the dissociation process of FNR and NADP+ (Strunz et al., 1999). This means that NADP+ dissociates from FNR through a single energy barrier between the initial and the transition state of the highest energy to which the system must be raised before dissociation can occur. The koff and xβ parameters can be then extracted from the linear fit. Fig. 26. Loading rate dependence on the most probable unbinding forces. This indicates that a single energy barrier is crossed during the unbinding process (Yuan et al., 2000). According to Evans theory (Evans and Ritchie, 1997), by applying a linear force on the bond, the energy landscape is tilted such that the energy barrier is reduced. Fitting the F* versus lnR plot with equation 5, gives both kinetic parameters at zero force conditions (xβ and koff) by relating to the slope and the intercept of the linear fit, respectively. Table 3 summarizes the kinetic parameters of the dissociation of the FNR:NADP+ complex obtained from the fitting of data in Figure 24. Unbinding Force for a Single Complex (pN) koff (s-1) τ (s) xβ (nm) FNR:NADP+ 136±34 0.0198 50.6 0.0205 Table 3. Mechanical parameters obtained for the dissociation of FNR:NADP+ complex from fitting data shown in Figure 26. The intermolecular force for a single complex was measured at R 10 nN/s. Silvia Caballero Mancebo – Final Master Project Results 30 The analysis gave a position the energy barrier along the reaction coordinate, xβ, of 0.0205 nm. The calculated koff value was of 0.0198 s-1. This value is related to the characteristic lifetime, τo, of the complex (τo = koff -1). The expected life for this complex is 50.6 s. This data provides information on the specificity of the reaction: it is thought that a greater half-time is associated with greater specificity in the biorecognition process (Robert et al., 2007). Silvia Caballero Mancebo – Final Master Project Discussion 31 5. DISCUSSION During this Master Project the interaction forces between an enzyme and its enzymatic substrate were studied at a single molecule level for the first time using AFS. The biological complex chosen was the FNR-NADP+ system that is involved in the last step of the photosynthesis light reactions in plants, algae and cyanobacteria. Our group has been involved in many studies of this enzyme as it belongs to a large family of proteins called flavoproteins, one of the main research areas of our group. In previous works, the interaction of FNR with its protein partners, Fd and Fld, and its substrate, NAPD+ has been extensively studied using classical biochemistry techniques. However, these methods provide average signals and behaviour of the total of molecules involved in the study. AFS was used in our group to define the energy landscape and the mechanostability parameters of the dissociation of FNR from both its protein partners so this project fulfils the characterization of all the interactions of FNR at the single molecule level. A typical AFS experiment requires the immobilization of the interacting partners, one on the surface of a substrate and the other one at the AFM tip. In this work the enzyme FNR was immobilized on the surface of flat mica pieces. It is of pivotal importance that both the labelling process and the immobilization process have no effect over the functionality and the catalytic activity of the enzyme. To monitor this, two different control experiments were carried out. On the one hand, the absorbance spectrum of the enzyme was checked after the labelling process. The characteristic peak of FNR appeared at 458 nm, assuring that the structure of the FMN center of the enzyme was not affected by this process. On the other hand, the catalytic activity of the enzyme after the tagging process was compared to the catalytic activity of the wild type one by monitoring the cytochrome c reductase activity. Although the activity of the labelled FNR exhibited a 1.5-fold decrease with respect to the free FNR (Table 2), it still had high catalytic activity. This steady-state enzymatic analysis showed that the labelled FNR was still able to bind to its substrate, proving the suitability of this labelling process for our final purpose. AFM imaging was used to control the immobilization of FNR on the mica surfaces. Images taken in tapping mode in fluid showed that a monolayer of FNR appeared when at least 4 µg of protein were added to each single mica piece. Topography images (Fig. 21) revealed the typical features of a protein monolayer. It is not a perfectly homogeneous monolayer but it displays Silvia Caballero Mancebo – Final Master Project Discussion 32 proteins molecules that appear as bigger and brighter circles in the images that are due to the globular shape of this protein and some holes. The height profiles seen in the functionalized mica pieces agreed with those found in other similar studies (Fig. 22). The scratching experiment confirmed that the total height of the functionalization product was around 12 nm. The height profile obtained in the scratching experiment is more accurate than the profile obtained in any other topography image because in the former, the surface of the mica piece is cleared and the height is measured exactly from the mica surface. One of the objectives of this project was to develop a reproducible procedure to functionalize NADP+ on surfaces, particularly on AFM tips. This is not a straightforward issue as too many factors have to be taken into account. The most important aspect to be considered is the reaction conditions. We are dealing with organic and biological molecules that are very sensitive to environmental changes, especially proteins, which can lose their structure and activity when pH changes just a few tenths. Another matter worth mentioned is the linker used to bind the ligand to the AFM tip. Although PEG is the most used spacer, there are other ones that can be used too. It is very important to know the stretching and flexibility characteristics of the spacer used because this is what will differentiate a specific event from an unspecific one in the force-distance curves. In our case, we used prefunctionalized tips with PEG ending in a maleimide group to which NADP+ was covalently bond. The presence of PEG allows the identification of the specific events quite easily: those peaks that displayed a non-linear, parabolic-like profile and whose length was around 20 nm. Typical force spectroscopy experiments were performed on a biological complex system involved in ET processes. The aim of these experiments was to gather further information on the mechanical properties of this specific enzyme-substrate complex and establish the basis for the study of other enzyme-substrate complexes. The experiments consisted in recording multiple forcedistance curves obtained by approaching the NADP+ functionalized AFM tip towards the FNR, which is immobilized on the chamber of the AFM instrument. DFS was applied and several hundreds of force-distance curves were recorded at four different loading rates. From the analysis of the curves, the most probable unbinding force, F*, was obtained for each loading rate (Fig. 24). At a loading rate of 10 nN/s, the most probable unbinding force was of 136±34 pN. When this value is compared to those previously obtained for the interaction between FNR and its protein partners Fd and Fld, 57±19 pN and 21±8 pN respectively (Marcuello et al., in preparation), it can be seen that the mechanical stability exhibited by the FNR:NADP+ complex is clearly higher: almost 3 times higher than the interaction between FNR and Fd and 7 Silvia Caballero Mancebo – Final Master Project References 39 Lee, G.U.; Chrisey, A.C. and Colton, R.J. 1994. Direct measurements of the forces between complementary strands of DNA. Science 266:771-773. Lee, G.U.; Kidwell, D.A. and Colton, R.J. 1994. Sensing discrete streptavidin-biotin interactions with atomic force microscopy. Langmuir 10:354-357. Marcuello, C.; de Miguel, R.; Gómez-Moreno, C. et al. 2012. An efficient method for enzyme immobilization evidenced by atomic force microscopy. Protein Eng Des Sel 25:715-723. Marcuello, C.; de Miguel, R.; Martínez-Júlvez, M. et al. Mechanostability of the electron-transfer complexes between Anabaena Ferredoxin-NADP+ Reductase and its redox protein partners studied by single molecule techniques. In preparation. Martínez-Júlvez, M.; Medina, M. and Velázquez-Campoy, A. 2009. Binding Thermodinamic of Ferredoxin NADP+ Reductase: Two different protein Substrates and One Energetics. Biophys J 96:4966-4975. Medina, M. and Gómez-Moreno, C. 2004. Interaction of ferredoxin-NAPD+ reductase with its substrates: optimal interaction for efficient electron transfer. Photosynthesis Research 79:113131. Noy, A. 2011. Force spectroscopy 101: how to design, perform, and analyse an AFM-based single molecule force spectroscopy experiment. Current Opinion in Chemical Biology 15:710-718. Robert, P.; Benoliel, A.M.; Pierres, A. et al. 2007. What is the biological relevance of the specific bond properties revealed by single-molecule studies? J Mol Recognit 20(6):432-447. Strunz, T.; Oroszlan, K.; Schäfer, R. et al. 1999. Dynamic force spectroscopy of single DNA molecules. Proc Natl Acad Sci USA 96:11277-11282. Yuan, C.; Chen, A.; Kolb, P. et al. 2000. Energy Landscape of Streptavidin Biotin Complexes Measured by Atomic Force Microscopy. Biochemistry 39:10219-10223. Zanetti, G. and Forti, G. 1966. Studies on Triphosphopyridine Nucleotide-Cytochrome f Reductase of Chloroplasts. J Biol Chem 241(2):279-285. Zhong, Q.; Inniss, D.; Kjoller, K. et al. 1993. Fractured polymer/silica fiber surface studied by tapping mode atomic force microscopy. Surface Sci Lett 290: 888-692.