royalsocietypublishing.org/journal/rsif Research Cite this article: Liang S-P, Levenson R, Malady B, Gordon MJ, Morse DE, Sepunaru L. 2020 Electrochemistry as a surrogate for protein phosphorylation: voltage-controlled assembly of reflectin A1. J. R. Soc. Interface 17: 20200774. http://dx.doi.org/10.1098/rsif.2020.0774 Received: 22 September 2020 Accepted: 4 November 2020 Subject Category: Life Sciences–Chemistry interface Subject Areas: biomaterials, biophysics, biotechnology Keywords: electro-assembly, reflectin, electrochemistry, charge neutralization Authors for correspondence: Michael J. Gordon e-mail: gordon@ucsb.edu Daniel E. Morse e-mail: d_morse@lifesci.ucsb.edu Lior Sepunaru e-mail: sepunaru@ucsb.edu Electronic supplementary material is available online at https://doi.org/10.6084/m9.figshare. c.5221981. Electrochemistry as a surrogate for protein phosphorylation: voltage-controlled assembly of reflectin A1 Sheng-Ping Liang1, Robert Levenson2,3, Brandon Malady2, Michael J. Gordon4,5, Daniel E. Morse2,5 and Lior Sepunaru1 1 Department of Chemistry and Biochemistry, University of California Santa Barbara, Building 232, Santa Barbara, CA 93106-9510, USA 2 Department of Molecular, Cellular, and Developmental Biology, University of California, Santa Barbara, CA 93106-9625, USA 3 Soka University of America, Aliso Viejo, CA 92656, USA 4 Department of Chemical Engineering, University of California, Santa Barbara, CA 93106-5080, USA 5 Institute for Collaborative Biotechnologies, University of California, Santa Barbara, CA 93106-5100, USA LS, 0000-0002-4716-5035 Phosphorylation is among the most widely distributed mechanisms regulating the tunable structure and function of proteins in response to neuronal, hormonal and environmental signals. We demonstrate here that the lowvoltage electrochemical reduction of histidine residues in reflectin A1, a protein that mediates the neuronal fine-tuning of colour reflected from skin cells for camouflage and communication in squids, acts as an in vitro surrogate for phosphorylation in vivo, driving the assembly previously shown to regulate its function. Using micro-drop voltammetry and a newly designed electrochemical cell integrated with an instrument measuring dynamic light scattering, we demonstrate selective reduction of the imidazolium side chains of histidine in monomers, oligopeptides and this complex protein in solution. The formal reduction potential of imidazolium proves readily distinguishable from those of hydronium and primary amines, allowing unequivocal confirmation of the direct and energetically selective deprotonation of histidine in the protein. The resulting ‘electro-assembly’ provides a new approach to probe, understand, and control the mechanisms that dynamically tune protein structure and function in normal physiology and disease. With its abilities to serve as a surrogate for phosphorylation and other mechanisms of charge neutralization, and to potentially isolate early intermediates in protein assembly, this method may be useful for analysing never-beforeseen early intermediates in the phosphorylation-driven assembly of other proteins in normal physiology and disease. 1. Introduction Enzymatically catalysed phosphorylation is one of the most ancient and universally distributed mechanisms of biological signal transduction, regulating the structure, hierarchical assembly and function of proteins in response to a wide range of neuronal, hormonal and environmental triggers [1,2]. While several molecular mechanisms may govern this regulation, the phosphorylation of cationic protein domains most commonly acts via charge neutralization. This effect can be mimicked by pH titration or genetic engineering, as seen in the case of the complex protein reflectin A1 [3,4]. This cationic, intrinsically disordered block-copolymer-like protein mediates the neuronally controlled fine-tuning of colour reflected from the intracellular Bragg lamellae in squid skin for camouflage and communication [4–6]. In vivo, progressive charge neutralization by neuronally activated phosphorylation overcomes Coulombic repulsion of reflectin A1’s cationic domains, driving condensation, folding, and hierarchical assembly, resulting in osmotic dehydration of the reflectin-containing © 2020 The Author(s) Published by the Royal Society. All rights reserved. (a) 2 (b) current (µA) histidine O– 10 mM histidine 10 mM HClO4 H3N+ H+ of hydronium H+ of imidazolium H+ of N-term 10 mM histidine O H N N+ H glycine (Gly) O– O H3N+ H 10 mM glycine –1.0 –0.8 –0.6 –0.4 –0.2 E(V) versus SCE 0 0.2 Figure 1. Site-specific electrochemical titration of amino acids histidine (His) and glycine (Gly). (a) Cyclic voltammograms of His and Gly in 100 mM KCl as a supporting electrolyte, with and without 10 mM perchloric acid (curves offset for clarity). Negative current peaks seen in the voltammograms between −300 and −900 mV versus SCE correspond to reduction of hydronium (yellow), the ring imidazolium (NH+, red), and the terminal amine (NH+3 , blue), as shown in (b). (b) Molecular structures of His+ ( protonated His under acidic conditions) and Gly, with imidazolium and terminal amine protonation sites coloured in red and blue, respectively. Bragg lamellae. This reversible dehydration, in turn, modifies the lamellar spacing and effective refractive indices, leading to a change in reflected colour [7,8]. In vitro, reduction of the protein’s histidine residues by pH-titration acts as a surrogate for the charge neutralization by phosphorylation in vivo, driving reflectin A1 assembly [3,4]. Unique among the cationic amino acids in proteins, the pKa of histidine’s imidazolium side chain lies in the physiological range of pH, enabling its tunable ionization to control the structures, interactions and activities of many proteins [9]. Examples include the catalytic triads of hydrolases [10], other enzymatic proton shuttles [11] and conformational changes and/or assembly (e.g. in firefly luciferase [12], viral haemagglutinin [13] and reflectin A1 [4]). We demonstrate here that the imidazolium of histidine in the monomeric state, in oligopeptides and in proteins in solution can be directly deprotonated under mild acidic conditions on a platinum (Pt) electrode at a formal reduction potential that is clearly distinguishable from those of hydronium and the terminal primary amine. In the case of the histidine-rich reflectin A1 protein, transitory contact of freely diffusing reflectin with an appropriately biased Pt electrode is shown to reduce the protonated imidazolium side chains of histidine residues in the protein, driving its assembly as an in vitro surrogate for the protein’s charge neutralization by phosphorylation in vivo. 2. Electrochemical results To distinguish between the electrochemical reduction potentials of the imidazolium side chain of histidine (His), the terminal primary amine and the hydronium ion, cyclic voltammetry (CV) with a Pt electrode was performed on freely diffusing His and glycine (Gly) in 40 mM KCl supporting electrolyte, in the presence and absence of 10 mM perchloric acid (figure 1). When no perchloric acid was added, one redox wave appeared between −300 and −900 mV versus SCE (saturated calomel electrode) in the CVs of both Gly (black) and His (blue) in figure 1a. Under such conditions, the terminal amine [14] ( pKa ∼ 9–10, also see electronic supplementary material, figure S1 is the only possible protonated chemical moiety in both His and Gly. Thus, the redox waves between −750 and −900 mV (designated in the light blue region) were assigned to the direct electrochemical deprotonation of the terminal primary amine (NH+3 ). When perchloric acid was mixed with His at a 1 : 1 molar ratio, two redox waves appeared between −300 and −900 mV (figure 1a, red). The dissociated proton from perchloric acid populates the imidazole ring of the His side chain and forms imidazolium (NH+) due to the approximately 6.5 pKa of His. As such, the redox waves between −500 and −750 mV (light red region) were assigned to direct electrochemical deprotonation of the imidazolium side chain of histidine. Further addition of perchloric acid (2 : 1 ratio of HClO4 : His) produced three redox waves (figure 1a, yellow). Under this condition, both weakly acidic molecular moieties—the terminal amines and the imidazolium of His—are fully protonated. Thus, the redox waves between −300 and −500 mV (yellow region) can be assigned to direct ‘classical’ electrochemical reduction of hydronium (H3O+) on the Pt surface [15]. Overall, these voltammograms demonstrate the selective reduction of protons on different chemical moieties in amino acids via electrochemistry. Moreover, the voltammetric approach offers an analytic route to probe and distinguish chemical moieties both with and without titratable side chains. Given the ability to electrochemically reduce specific protonated moieties in freely diffusing amino acids, we sought to demonstrate that the same effect could be observed for His within a polymer, as seen in the tripeptide Glycine– Glycine–Histidine (Gly–Gly–His), in comparison to Glycine–Glycine–Glycine (Gly–Gly–Gly). Histidine was chosen as the primary target for selective electroreduction because its imidazolium side chain produces a clear voltammetric response that is well separated from water electrolysis, primary (terminal) amine reduction and hydronium reduction (figure 1a). To clearly show the feasibility of J. R. Soc. Interface 17: 20200774 –1.2 5 mM histidine 10 mM HClO4 (His+) royalsocietypublishing.org/journal/rsif 100 µA (a) 0 H+ of N-term H+ of imidazole (b) H+ O– O histidine H3N+ 0 NH O NH –2 O –4 NH Gly–Gly–Gly N+ H 0 –2 Gly–Gly–Gly –4 O Gly–Gly–His H3N+ 0 –4 reflection –0.8 –0.6 –0.4 –0.2 E(V) versus Ag/AgCI NH NH O O H 0 Figure 2. (a) Differential pulse voltammograms (DPVs in 40 mM NaCl at pH 3, measured versus Ag/AgCl) for 1 mM histidine, 2 mM tripeptide H–glycine–glycine– glycine–OH (Gly–Gly–Gly), 2 mM tripeptide H–glycine–glycine–histidine–OH (Gly–Gly–His), and 25 µM reflectin A1 protein. The reflectin A1 DPV was baseline subtracted for peak deconvolution (red, imidazolium; grey, hydronium). Arrows show the evolution of DPV peak potential for the designated electrochemically active imidazolium NH+ (red) and terminal amine (NH+3 , blue), as represented in (b). (b) Molecular structures of Gly–Gly–His and Gly–Gly–Gly tripeptides, with imidazolium and terminal amine protonation sites coloured in red and blue, respectively. electrochemical reduction of specific chemical moieties within diffusing peptides, these experiments were carried out using differential pulse voltammetry (DPV). Increased sensitivity is provided by DPV, as only the Faradaic contribution to the reduction reaction is monitored (unlike CV), thus allowing analysis of lower concentrations of redox active species. Histidine and the tripeptide samples were adjusted to pH 3 in 40 mM NaCl to ensure a dominant population of imidazolium, while maintaining electrolyte support (we verified that NaCl and KCl can be used interchangeably, see electronic supplementary material, figure S2). The resulting DPV of His shows three clear reduction waves (figure 2a). The reduction wave peaking at −450 mV corresponds to the aqueous solvent’s hydronium proton reduction, as shown in electronic supplementary material, figure S3. Because we confirmed that the reduction potentials of the protonated amino acid moieties are correlated with their pKa values (figure 1), we can assign the reduction wave peaking at −600 mV to the deprotonation of imidazolium, and the shoulder near −770 mV to deprotonation of the terminal amine. In Gly–Gly–Gly, investigated as a control tripeptide with non-titratable side chains (cf. figure 1b), the reduction wave peaking at −450 mV reflects hydronium reduction, while the reduction wave peaking at −720 mV corresponds to deprotonation of the terminal amine, as illustrated in figure 2. Similar to Gly–Gly–Gly, Gly–Gly–His also shows two waves: the hydronium reduction wave peak at −450 mV and a broad wave between −600 and −800 mV. Comparison of the molecular structures of the tripeptides (figure 2b) indicates that Gly–Gly–His at pH 3 will have an extra protonated imidazolium moiety in addition to the protonated terminal amine. We thus attribute the broad peak of Gly–Gly–His to the overlap of two redox waves, namely the terminal amine and the imidazolium. Shifts in the reduction potentials of protons on the Gly and His residues in the two tripeptides are associated with shifts in local pKa values due to the chemical environment in the tripeptide, an effect that has been documented [16–20] and seen in other pH-titration experiments (electronic supplementary material, figure S3). The aforementioned experiments demonstrate that the imidazolium of histidine, in freely diffusing form or in a peptide, can be electrochemically deprotonated; as such, we next sought to see if this phenomenon could be extended to histidines contained within a macromolecular protein. We successfully demonstrated this effect with reflectin A1, the histidine-rich (31 His/350 AAs) protein that functions as a sensitive transducer of neuronal signals governing the dynamic changes in skin colour for camouflage and communication in squids [6]. Reflectin A1 was chosen for this test because recently it was shown that its biological activity is mediated by charge neutralization-driven assembly to form large multimers, and that pH-titration of its histidine residues can be used as an in vitro surrogate for the neurotransmittermediated phosphorylation that triggers this assembly in vivo [3,4]. This assembly effect thus provides a sensitive measure of histidine deprotonation within the protein. Additionally, reflectin A1 monomers at low pH are known to be intrinsically disordered [4], indicating that many of its amino acids are likely exposed to the solvent, and therefore accessible for direct electrochemical reduction with a Pt electrode. Of the numerous isoforms of reflectin known, A1 is also the one that has been most extensively characterized biophysically [3–5]. The DPV of reflectin A1 (figure 2) exhibits a large reduction wave at ca −450 mV (versus Ag/AgCl) corresponding to hydronium reduction (for the solvent at pH 3), with a pronounced shoulder at −550 to −650 mV that becomes more apparent upon deconvolution (red curve). This shoulder and its resolved peak clearly occur in the range of imidazolium reduction (−500 to −700 mV) as seen in the previous data in figures 1 and 2, with the observed variation in reduction wave width for reflectin A1 being attributable to differences in pKa for multiple histidine residues in the protein that depend on the local sequence environment. The presence of J. R. Soc. Interface 17: 20200774 –2 O– royalsocietypublishing.org/journal/rsif –2 –4 current (µA) 3 Gly–Gly–His reflection monomers (a) potentiostat (b) electrochemical reduction Ref Iscattering (t) la s er phosphorylation WE (c) (d) iv 30 000 v vi size distribution (volume %) iii 20 000 15 000 10 000 i 5000 15 ii OCP i ii iii iv v vi –475 mV i ii iii iv v vi –700 mV i ii iii iv v vi 10 5 15 10 5 15 10 0 5 0 5 10 time (min) 15 20 102 101 hydrodynamic diameter, DH (nm) 103 Figure 3. Electrochemical assembly of reflectin A1 protein. (a) In the squid, a neurotransmitter triggers enzymatic phosphorylation, neutralizing reflectin and driving its condensation, folding and assembly [5]. In vitro, electrochemical reduction of histidine imidazolium postulated to act analogously to pH titration, neutralizing the protein and driving assembly. (b) Experimental set-up to electrochemically trigger reflectin assembly with in situ DLS. (c) Reflectin A1 DLS intensity (count rate) for OCP (red), −475 mV (blue) and −700 mV (green) conditions with respect to Ag/AgCl. (d) Reflectin A1 particle size distributions (volume %) measured by DLS at times (i)–(vi) indicated in (c). Reflectin A1 monomer DH = 8–12 nm. this reduction wave for reflectin A1, and its corresponding thermodynamic equivalency with the freely diffusing histidine, indicates that direct charge exchange occurs between the imidazolium side chains in reflectin A1 and the Pt electrode. This observation of selective electrochemical reduction of cationic histidine residues in the protein is mechanistically distinct from earlier reflectin thin film work demonstrating that anionic amino acids in the native sequence of wild-type reflectin A1 in the solid state can act as a conducting proton ‘shuttle’ under high humidity (greater than 70% relative humidity) conditions [21]. We further confirmed our hypothesis of this direct charge exchange mechanism involving imidazolium side chains using dynamic light scattering (DLS, figure 3) and TEM (figure 4). By combining electrochemistry with DLS to measure the apparent size distributions of reflectin A1 within the domain of a Pt coil electrode (figure 3b), we observed no change in the size of reflectin A1 monomers under open circuit potential (OCP), as well as after 20 min application of −475 mV (versus Ag/AgCl), a potential sufficient to reduce only the hydronium ion (figure 3c,d). At this low potential, only free hydronium would be reduced, gradually changing the solution pH near the electrode surface; apparently, the limitation of this effect by diffusion makes it ineffective in driving assembly. In marked contrast, dramatic, time-dependent assembly of reflectin A1 is triggered by application of −700 mV, which was shown above (cf. figures 1 and 2) to be sufficient to reduce the imidazolium of histidine in various states. Since the intensity of Rayleigh scattering scales with particle radius [22], the back-scattered photon count rate from the DLS photodetector was used to directly measure reflectin A1 assembly ( particle formation) as time progressed (figure 3c). Decay of the autocorrelation function of the scattering allows determination of particle diffusivity [23], and from this, via the Stokes–Einstein relation, the size distribution of the particles’ effective hydrodynamic diameters (figure 3d). To gain deeper insight into the electro-assembly process, the reflectin A1 particle size distributions were analysed at six time points (figure 3c,d). At time (i), analyses at all potentials show similar count rates, with size distributions of particles in all samples showing the approximate diameter of the reflectin A1 monomer (ca 10 nm). At time (ii), the −700 mV (green curve) sample exhibited a bimodal size distribution, while all other samples remained as monomers. With progressively longer times, the size distributions of particles in the −700 mV sample increased steadily until reaching a plateau (approx. 70 nm) after ca 10–15 min. The results shown in figure 3c,d are especially interesting because they appear to differ from those obtained when reflectin A1 assembly is driven by a pH jump, change in ionic strength or genetic engineering [3,4,24]. All methods trigger assembly by charge neutralization or screening, but unlike the bulk process seen conventionally, electrochemical reduction inherently requires mass transport of reflectin A1 to and from the electrode, as well as multiple His+ → His turnovers, resulting in more gradual and controllable change in assembly size. The latter may therefore offer a J. R. Soc. Interface 17: 20200774 OCP –475 mV –700 mV 25 000 count rate (kcps) photodetector 4 royalsocietypublishing.org/journal/rsif CE size distribution (%) (a) (b) OCP 5 –700 mV royalsocietypublishing.org/journal/rsif 200 nm Figure 4. Transmission electron microscopy (TEM) images of aliquots of reflectin A1 ( pH 3, 40 mM NaCl): (a) before any bias was applied (OCP); (b) after −700 mV versus Ag/AgCl had been applied for 30 min. Samples were collected from the centre of the DLS/Pt coil as shown in figure 3b. unique possibility of using electrochemistry to isolate early and potentially never-before-analysed intermediates in the assembly of proteins. Results of the DPV and DLS analyses of reflectin A1 were further confirmed by TEM (figure 4). In the sample analysed under OCP conditions, only reflectin A1 monomers (D < 15 nm) were seen, whereas, the sample exposed to −700 mV for 30 min contained significantly larger particles (D = 40–70 nm), further supporting our electrochemical spectroscopy results. 3. Conclusion Bentley et al. [25] reported that the protons of acidic amines can be directly reduced at Pt electrodes with reduction potentials correlated with their pKa values. This observation suggests that electrochemistry can be used to selectively target the deprotonation of specific amino acid residues within proteins. Histidine residues are ideal targets for this because their side chain has the lowest pKa of all positively charged amino acids. Accordingly, we have shown that the imidazolium of histidine in the monomeric state, in oligopeptides and in proteins in solution can be directly deprotonated under acidic conditions on a Pt electrode, at a formal reduction potential that is clearly distinguishable from those of the hydronium cation and terminal primary amine. Cyclic and differential pulse voltammetry clearly showed that the formal reduction potentials of these protonated moieties correlated well with their respective pKa values. The relatively low electrochemical reduction potentials observed make this method especially useful for the biophysical analysis of charged proteins. Specific electroreduction of imidazolium moieties in the His-rich protein, reflectin A1, revealed variations in pKa sensitive to local environments in the protein, and led to the unprecedented discovery that assembly could be manipulated electrochemically by charge neutralization of amino acid side chains. As noted above, (i) the histidine residues within reflectin A1 that have undergone electrochemical reduction exhibit a range of pKa values apparently dependent on their local sequence environments; (ii) this electroreduction is limited by the brief diffusional contact with the electrode; and (iii) the structures of the resulting reflectin A1 assemblies differ from those driven by bulk changes in pH, ionic screening or genetic engineering [3,4], thus suggesting they may be kinetically trapped early intermediates in the assembly process. We conclude that only a specific subset of histidines in the protein has been electrochemically reduced, giving rise to the electrochemical signature and assemblies observed. Charge neutralization of reflectin A1 previously was shown to progressively overcome Coulombic repulsion of the cationic, initially disordered protein, driving condensation and secondary folding with potential emergence of previously cryptic hydrophobic domains that can facilitate hierarchical assembly [4]. The resulting assemblies are then quickly stabilized by a complex process of dynamic arrest, apparently consisting of both non-covalent interactions, such as π–π, cation–π and sulfur–π interactions, and electrostatic interactions characteristic governing colloidal stability [4]. Together, these findings help unify our understanding of the initial drivers of reflectin A1 assembly (by electroreduction, phosphorylation, pH-titration, ionic screening or genetic engineering) with the pioneering observations of Guan et al. [26], who showed that a variety of small aromatic compounds can dramatically facilitate assembly, pointing to the importance of π–π and related interactions. Our findings also demonstrate that DPV allows selective deconvolution and probing of different energetic (redox) processes and/or charge neutralization events that are important in the thermodynamically controlled assembly processes, the latter being exemplified by electrically triggered assembly of reflectin A1 above a threshold equilibrium electrochemical (thermodynamic) potential. As such, electrochemistry offers a new approach to unravel some of the mechanisms governing protein assembly as well as to potentially control assembly processes. For example, the direct, site-specific electroreduction of reflectin A1 presented herein acts as a surrogate for its physiological charge neutralization by phosphorylation. J. R. Soc. Interface 17: 20200774 200 nm 4. Methods 4.1. Reflectin A1 expression and purification 4.2. Protein solubilization and sample preparation Lyophilized reflectin A1 was solubilized by the addition of 0.22 μm-filtered 25 mM sodium acetate, pH 4.5 buffer. Concentration was determined by measuring absorbance (A280) using a calculated extinction coefficient of 120 685 l mol−1 cm−1 [27]. For electrochemical measurements in buffered solution, the protein was dialysed 1000-fold three times against buffer solution (with pH and chemical composition as described) to remove unwanted counter ions from protein purification. For measurements in unbuffered solutions, samples were initially dialysed into 25 mM sodium acetate, pH 4.5 as above, and then further dialysed 1000-fold three times into an unbuffered solution, with final pH adjusted with perchloric acid. Final pH after dialysis was confirmed in all cases with a pH meter. Protein was stored at 4°C and filtered through a 0.1 μm syringe filter shortly before use. DLS and TEM were used to confirm the monomeric state of the protein at the start of all experiments [3,4]. All DPV measurements were done in a miniaturized threeelectrode configuration with the Pt electrode prepared via an identical polishing protocol as described above. In this series of experiments, the electrochemical ‘cell’ was a 10 μl droplet of analyte solution on the Pt disc electrode pointed upward. A Pt wire counter electrode was inserted between the working and reference electrode. DPV measurements were done with a potential step size of 5 mV, pulse height of 10 mV, pulse duration of 50 ms and interval time of 0.5 s. 4.5. Constant potential electrochemistry with in situ dynamic light scattering Constant potential electrochemical measurements were carried out in a three-electrode configuration (working electrode: Pt coil; counter electrode: Pt wire; reference electrode: Ag/AgCl) with an Autolab M204 electrochemical workstation (Metrohm). The three electrodes were assembled in a DLS compatible optical cuvette and sealed with parafilm to prevent dust from the surroundings. DLS was performed with a Malvern Zetasizer Nano ZS (Worcestershire). The samples were probed with a 632.8 nm HeNe gas laser with a beam diameter of 0.63 mm (1/e²) and detected by an Avalanche photodiode (Q.E. greater than 50% at 633 nm) in a backscattering configuration at 7° from normal. The measurements were done with 2 ml sample volumes at 25°C (see main text for the applied potentials). All samples were measured at OCP for at least 30 min to ensure signal stability over time. 4.6. Transmission electron microscopy Electrochemically driven reflectin A1 assemblies and their parallel controls were collected straight from the DLS cuvette (before and after applying −750 mV for 30 min) in the vicinity of the working electrode with micropipette and directly applied to 400-mesh carbon-coated grids (Electron Microscopy Services, Hatfield PA). Prior to sample application, grids were treated by glow discharge for 20 s. Five microlitres of freshly prepared sample was applied for 2 min before wicking away excess solution with filter paper. Samples were then negatively stained by application of 20 μl freshly filtered 1.5% uranyl acetate three times, with wicking in between each application. Samples were examined in a 200 kV ThermoFisher Talos G2 TEM in bright field mode. 4.7. Titrations Titration experiments were done by monitoring pH while gradually adding 2.5 M NaOH in 100 mM of pH 2 analyte solution. The pH of the analyte solution was adjusted with perchloric acid. pH was monitored with a pH meter (Fisher Scientific, Accumet Basic 150). Data accessibility. All data are reported in the manuscript and electronic supplementary material. 4.3. Cyclic voltammetry measurements Electrochemical measurements using Autolab M204 electrochemical workstation were done in a three-electrode configuration (working electrode: Pt disk; counter electrode: Pt wire; reference electrode: fritted Ag/AgCl in 1 M KCl(aq) or SCE. The potential difference between SCE and AgCl (in 1 M KCl(aq)) was experimentally determined to be +7.0 mV. The Pt disc working electrode (RPt = 1.5 mm) was polished three times (2 min each) using 1 μm, 0.25 μm and then 0.05 μm MetaDiTM polycrystalline diamond suspension (Buehler, Lake Bluff, IL, USA) on a microcloth polishing pad. The polished working electrode was then sonicated in water for 2 min. CV measurements were done at a scan rate of 100 mV s−1. Authors’ contributions. S.P.L. performed electrochemistry experiments; S.P.L. and R.L. carried out DLS, electrochemistry, and titration experiments. R.L. and B.M. expressed, prepared and purified reflectin protein for experiments. R.L. performed TEM measurements. S.P.L. and R.L. worked up and analysed data. S.P.L., R.L., M.J.G., D.E.M. and L.S. all interpreted data and contributed to writing the manuscript. Competing interests. We declare we have no competing interests. Funding. Research was sponsored by the U.S. Army Research Office and accomplished under cooperative agreement W911NF-19-2-0026 and contract W911NF-19-D-0001 for the Institute for Collaborative Biotechnologies. The content of the information herein does not necessarily reflect the position or the policy of the U.S. Government, and no official endorsement should be inferred. 6 J. R. Soc. Interface 17: 20200774 Recombinant Doryteuthis opalescens reflectin A1 was synthesized from a recombinant DNA construct and purified by methods described previously [3,4]. Briefly, Rosetta 2 (DE3) E. coli cells were grown in liquid cultures from freshly plated transformants in the presence of 50 mg ml−1 kanamycin to maintain selection of the recombinant DNA plasmid. Expression was induced in the logarithmic growth phase by addition of 1 mM isopropylthioglactoside. Expression proceeded for approximately 6 h, after which cells were centrifuged and frozen at −80°C. Reflectin A1 inclusion bodies then were purified from thawed cell pellets with BugBuster medium (Novagen, Inc., Madison, WI, USA), as directed by the manufacturer. Inclusion bodies were solubilized in 5% acetic acid/8 M urea. Reflectin was purified initially by ion exchange over a HiTrap XL (GE Healthcare) cation exchange column eluted with a gradient of 5% acetic acid/6 M guanidinium chloride. Fractions containing reflectin were collected and diluted in 5% acetic acid/8 M urea, concentrated by centrifugal filtration and loaded onto a HPLC reverse-phase C10 column equilibrated with 0.1% aqueous trifluoracetic acid and eluted with a gradient of 95% acetonitrile/ 0.1% TFA. The resulting purified protein was lyophilized and stored at −80 °C until solubilization for use. Purity was confirmed by SDS-PAGE on 10% tris-acetate SDS-PAGE gels (Life Sciences, Carlsbad, CA, USA). 4.4. Differential pulse voltammetry measurements royalsocietypublishing.org/journal/rsif Moreover, the possibility to isolate early intermediates in protein assembly using electrochemistry suggests that it may be useful in analysing never-before-seen early intermediates in charge-mediated regulation of protein structure and function in normal physiology and disease. References 2. 3. 5. 6. 7. 8. 9. 10. Dodson G. 1998 Catalytic triads and their relatives. Trends Biochem. Sci. 23, 347–352. (doi:10.1016/ S0968-0004(98)01254-7) 11. Bachovchin WW, Roberts JD. 1978 Nitrogen-15 nuclear magnetic resonance spectroscopy. The state of histidine in the catalytic triad of .alpha.-lytic protease. Implications for the charge-relay mechanism of peptide-bond cleavage by serine proteases. J. Am. Chem. Soc. 100, 8041–8047. (doi:10.1021/ja00494a001) 12. Branchini BR, Magyar RA, Murtiashaw MH, Anderson SM, Zimmer M. 1998 Site-directed mutagenesis of histidine 245 in firefly luciferase: a proposed model of the active site. Biochemistry 37, 15 311–15 319. (doi:10.1021/bi981150d) 13. Mair CM, Meyer T, Schneider K, Huang Q, Veit M, Herrmann A. 2014 A histidine residue of the influenza virus hemagglutinin controls the pH dependence of the conformational change mediating membrane fusion. J. Virol. 88, 13 189–13 200. (doi:10.1128/jvi.01704-14) 14. Chemical Rubber Company. 2020 CRC handbook of chemistry and physics. Boca Raton, FL: CRC Press. 15. Costentin C, Canales JC, Haddou B, Savéant J-M. 2013 Electrochemistry of acids on platinum. Application to the reduction of carbon dioxide in the presence of pyridinium ion in water. J. Am. Chem. Soc. 135, 17 671–17 674. (doi:10.1021/ ja407988w) 16. Kohda D, Sawada T, Inagaki F. 1991 Characterization of pH titration shifts for all the nonlabile proton resonances in a protein by two-dimensional NMR: the case of mouse epidermal growth factor. Biochemistry 30, 4896–4900. (doi:10.1021/ bi00234a009) 17. Kuramitsu S, Hamaguchi K. 1980 Analysis of the acid–base titration curve of hen lysozyme. J. Biochem. 87, 1215–1219. (doi:10.1093/ oxfordjournals.jbchem.a132855) 18. Forman-Kay JD, Clore GM, Gronenborn AM. 1992 Relationship between electrostatics and redox function in human thioredoxin: characterization of 19. 20. 21. 22. 23. 24. 25. 26. 27. pH titration shifts using two-dimensional homoand heteronuclear NMR. Biochemistry 31, 3442–3452. (doi:10.1021/bi00128a019) Gooley PR, Keniry MA, Dimitrov RA, Marsh DE, Keizer DW, Gayler KR, Grant BR. 1998 The NMR solution structure and characterization of pH dependent chemical shifts of the β-elicitin, cryptogein. J. Biomol. NMR 12, 523–534. (doi:10. 1023/A:1008395001008) Pahari S, Sun L, Alexov E. 2019 PKAD: a database of experimentally measured pKa values of ionizable groups in proteins. Database 2019, baz024. (doi:10. 1093/database/baz024) Ordinario DD, Phan L, Walkup IV WG, Jocson J-M, Karshalev E, Hüsken N, Gorodetsky AA. 2014 Bulk protonic conductivity in a cephalopod structural protein. Nat. Chem. 6, 596–602. (doi:10.1038/ nchem.1960) Seinfeld JH, Pandis SN. 2006 Atmospheric chemistry and physics, 2nd edn, vol. 5. Hoboken, NJ: John Wiley and Sons. Goldburg WI. 1999 Dynamic light scattering. Am. J. Phys. 67, 1152–1160. (doi:doi:10.1119/1.19101) Tao AR, DeMartini DG, Izumi M, Sweeney AM, Holt AL, Morse DE. 2010 The role of protein assembly in dynamically tunable bio-optical tissues. Biomaterials 31, 793–801. (doi:10.1119/1.19101) Bentley CL, Bond AM, Hollenkamp AF, Mahon PJ, Zhang J. 2015 Electrochemical proton reduction and equilibrium acidity (pKa) in aprotic ionic liquids: protonated amines and sulfonamide acids. J. Phys. Chem. C 119, 21 828–21 839. (doi:10.1021/acs.jpcc. 5b05723) Guan Z, Cai T, Liu Z, Luo S-J, Wang H-W, Xie C. 2017 Origin of the reflectin gene and hierarchical assembly of its protein. Curr.Biol.27, 833–2842. (doi:doi:10.1016/j.cub.2017.07.061) Wilkins MR, Gasteiger E, Bairoch A, Sanchez JC, Williams KL, Appel RD, Hochstrasser DF. 1999 Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol. 112, 531–552. (doi:10.1385/1-59259-584-7:531) J. R. Soc. Interface 17: 20200774 4. Theillet F-X et al. 2014 Physicochemical properties of cells and their effects on intrinsically disordered proteins (IDPs). Chem. Rev. 114, 6661–6714. (doi:10.1021/cr400695p) Nishi H, Shaytan A, Panchenko AR. 2014 Physicochemical mechanisms of protein regulation by phosphorylation. Front. Genet. 5, 270. (doi:10. 3389/fgene.2014.00270) Levenson R, Bracken C, Bush N, Morse DE. 2016 Cyclable condensation and hierarchical assembly of metastable reflectin proteins, the drivers of tunable biophotonics. J. Biol. Chem. 291, 4058–4068. (doi:10.1074/jbc.M115.686014) Levenson R, Bracken C, Sharma C, Santos J, Arata C, Malady B, Morse DE. 2019 Calibration between trigger and color: neutralization of a genetically encoded coulombic switch and dynamic arrest precisely tune reflectin assembly. J. Biol. Chem. 294, 16 804–16 815. (doi:10.1074/jbc.RA119.010339) Izumi M et al. 2010 Changes in reflectin protein phosphorylation are associated with dynamic iridescence in squid. J. R. Soc. Interface 7, 549–560. (doi:10.1098/rsif.2009.0299) Mäthger LM, Denton EJ, Marshall NJ, Hanlon RT. 2009 Mechanisms and behavioural functions of structural coloration in cephalopods. J. R. Soc. Interface 6, S149–S163. (doi:10.1098/rsif.2008.0366. focus) DeMartini DG, Krogstad DV, Morse DE. 2013 Membrane invaginations facilitate reversible water flux driving tunable iridescence in a dynamic biophotonic system. Proc. Natl. Acad. Sci. USA 110, 2552–2556. (doi:10.1073/pnas. 1217260110) Ghoshal A, DeMartini DG, Eck E, Morse DE. 2013 Optical parameters of the tunable Bragg reflectors in squid. J. R. Soc. Interface 10, 20130386. (doi:10. 1098/rsif.2013.0386) Liao SM, Du QS, Meng JZ, Pang ZW, Huang RB. 2013 The multiple roles of histidine in protein interactions. Chem. Cent. J. 7, 44. (doi:10.1186/ 1752-153X-7-44) royalsocietypublishing.org/journal/rsif 1. 7