**1. Introduction**

In biological cells, transmembrane proteins are embedded in the plasma membrane, composed of a lipid bilayer [1–3]. Certain transmembrane proteins act as receptors for external stimuli [4,5] such as light, mechanical stress, or the presence of specific molecular compounds [5–9]. Transmembrane receptor proteins transform the stimulus into signals that can be processed further downstream by the molecular machinery of the cell [2,4,10]. The stimulus produces structural or conformational changes in the transmembrane protein that will, for instance, result in an activation or deactivation of a specific ion channel, the release of a G protein (G protein-coupled receptors) or the activation of certain enzymes (enzyme-coupled receptors) [1,2,6,9,11–13].

Archaerhodopsin proteins (Arch) are among a class of transmembrane receptor proteins called photoreceptors that react to light [14,15]. Upon illumination with green light, Arch will undergo deprotonation of a Schiff base, resulting in proton pumping [16–19]. In this way, Arch actively transports protons through the membrane and out of the cell [14,18,19]. In vivo, the resulting proton gradient enables ATP synthases to produce ATP [18,20]. Photoreceptors typically consist of seven transmembrane helices and the chromophore retinal [16,18,21]. The structure of Arch corresponds to G protein-coupled

**Citation:** Khangholi, N.; Finkler, M.; Seemann, R.; Ott, A.; Fleury, J.-B. Photoactivation of Cell-Free Expressed Archaerhodopsin-3 in a Model Cell Membrane. *Int. J. Mol. Sci.* **2021**, *22*, 11981. https://doi.org/ 10.3390/ijms222111981

Academic Editor: Masoud Jelokhani-Niaraki

Received: 14 October 2021 Accepted: 3 November 2021 Published: 5 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

receptors that include rhodopsin. Because bacteriorhodopsin and Arch share similar structures and functions, their activation is assumed to be the same (see Figure 1c). Rhodopsin acts as a photoreceptor as well [1,15,22]. However, Arch acts as an ion channel, not as a G protein-coupled receptor. There are different kinds of Arch, which have slightly different properties [21]. Besides being photo active, Arch-3 fluorescence is also voltage sensitive [16,23], and it is often used in optogenetics as a voltage sensor [16,17]. To our knowledge, an electrophysiological characterization of Arch-3 has not ye<sup>t</sup> been reported in the literature [18].

In this work, we present a new and simple microfluidic approach [24] to study the conducting properties of ion channels. We incorporated recombinantly produced Arch-3-EGFP into a free-standing lipid bilayer that mimics a biological membrane. A cell-free expression system was used for the production and reconstitution of Arch-3-EGFP. Such in vitro systems enable better control of various biochemical parameters and processes than in vivo systems [25–28]. They are regularly used to study gene circuits or reaction cascades [29–31]. In vitro, proteins, in the form of wild-type proteins or with modifications, can easily be produced recombinantly by the simple addition of the corresponding coding DNA [32–34]. Moreover, in vitro systems can mimic in vivo systems, allowing the transfer of results to the in vivo situation [35–37]. Using this approach, we report the first electrophysiological characterization of single Arch-3 channels in a model cell membrane. The method presented here (sketched in Figure 1) can be understood as a first step towards the further investigation of different signaling cascades.

**Figure 1.** (**a**) A schematic of the formation of a lipid bilayer at the intersection of two microfluidic channels. First, the channel is filled with oil–lipid solution. Then, a cell-free expression reaction solution containing synthesized Arch-3-EGFP fused to the membrane of vesicles is injected into both channels. A remaining oil–filled lipid bilayer separates the two channels (left). Lipids from the oil phase and lipid vesicles with the inserted Arch-3-EGFP from the aqueous phase form a monolayer at both oil–water interfaces. As oil drains into the PDMS, the two interfaces gently meet to form a lipid bilayer containing Arch-3-EGFP (middle). The visualization of the inserted Arch-3-EGFP is achieved by exciting EGFP with blue light. AgCl electrodes inserted into the microfluidic device are used for electrical measurements (right). (**b**) The image shows the fluorescent signal of the EGFP tag from Arch-3 at the site of the suspended bilayer; PDMS elements remain dark. (**c**) The image shows the conformational changes of Arch-3 upon light exposure mediated by the deprotonation of retinal. This results in proton pumping which can be detected using electrophysiological measurements (scheme based on channel rhodopsin, adapted from [22]).

#### **2. Results and Discussion**

A DOPC/DPhPC bilayer containing Arch-3-EGFP was formed in a microfluidic device as described in the method section and as sketched in Figure 1. To verify the formation of a lipid bilayer, electrophysiological measurements were performed by applying a potential difference of 20 mV between both channels containing the reaction solution. The lipid bilayer separated the two ion-conducting water reservoirs, whereas the water–oil–water sandwich acted as a capacitor. Measuring the capacitance of such a sandwich in real time enabled the detection of bilayer formation; the graph in Figure 2a shows the related data from our experiments. The initial signal fluctuating around 10 pF corresponds to the situation with two monolayers separated by a macroscopic oil layer. The jump in the capacitance signal corresponds to the formation of a bilayer, a so-called zipping process. The following gradual increase in the capacitance demonstrates the growth of the bilayer area. This is due to the drainage of the oil to the PDMS at the plateau border.

Hydrostatic control of the flows enabled us to keep the bilayer area fairly constant for 1 h. The fluorescent image of EGFP tagged to Arch-3 under blue illumination, as shown in Figure 1b, confirmed the presence of Arch-3-EGFP in the vicinity of the lipid bilayer. After switching the laser illumination from blue to green, Arch-3 was activated, and an ion current was detected across the suspended lipid bilayer in real time upon light stimulation. The graph in Figure 2b shows the current intensity as a function of time, measured in the absence of light and after a short (~10 s) exposure to green laser light (532 nm) at ~40 s. Before light exposure, the current stayed almost constant. During the light pulse, the electrical current peaked, which means that ions were passing through the protein. After a few minutes, the signal decayed to its initial value. We interpret this observation as the signature of light-induced activation and subsequent deactivation of Arch-3. According to the literature, in any type of a rhodopsin photocycle, the deprotonation of the Schiff base after light excitation occurs within a range of picoseconds [38,39]. The deprotonation opens the channel; however, the subsequent recovery of the Schiff base requires milliseconds to several seconds [38–41]. Until recovery, the channel remains open for protons to pass through. Fitting an exponential decay to the curve from the time it begins to fall gives a recovery time τ of about 84 s.

**Figure 2.** (**a**) Electrical capacitance as measured between the two microfluidic channels, separated by an oil phase (see Figure 1a) as a function of time. The signature of bilayer formation is the sudden increase in capacitance. (**b**) A real-time current recording of the bilayer containing Arch-3-EGFP. In the absence of light, we observe a current signal fluctuating around a constant value. The sudden jump in the current signal demonstrates the activation of Arch-3 caused by a green laser pulse applied to the bilayer at ~40 s ≤ t ≤ 50 s. The signal decays, with a time constant of ~84 s, to its initial dark value. The recovery time corresponds to the re-protonation of the Schiff base as reported in [38–40].

To characterize Arch-3 activation more deeply, we continuously exposed an Arch-3-containing bilayer to monochromatic light with a wavelength of 532 nm. During light exposure, we observed an overall increase in the conductance, which was composed of a

stepwise increase and decrease in the current signal, as shown in Figure 3a. This signal is caused by a combination of opening and closing channels and the simultaneous fusion of Arch-3-EGFP containing vesicles from the aqueous phase with the bilayer. If Arch-3- EGFP is added, the current recording will expose a positive jump. The individual current steps Δ*I*, as indicated by the red line in Figure 3a, were obtained using the "change point analysis" algorithm as part of the software Origin (Origin 2021b; OriginLab Corporation, Northampton, MA, USA). In total, over 300 steps were analyzed. The histogram shown in Figure 3b displays the frequency of certain current steps Δ*I* grouped within intervals of 2.5 pA. From the histogram, it can be seen that most of the jumps were in the range of (16.25 ± 6.25) pA, where the full width at half maximum was used to determine the experimental uncertainty. The obtained value is in excellent agreemen<sup>t</sup> with the value found in the literature for bacteriorhodopsin (BR) [42]. The large current steps Δ*I* causing the asymmetric distribution towards larger current steps in the histogram could be due to multiple channels in the bilayer simultaneously being activated or deactivated.

Although Arch-3 is a proton pump, protons can passively pass while the channel is open. Considering the current step with the highest probability, i.e., the maximum peak of the histogram at Δ*I* = 16.25 pA, and assuming that Arch-3 has a passive opening in the bilayer, we can determine the radius of a single Arch-3 channel from the relation [43]

$$r = \sqrt{\frac{l\Delta I}{\pi GCV}}\tag{1}$$

where *l* ≈ 3 nm is the length of the protein assumed to be close to the thickness of the bilayer, Δ*I* = 16.25 pA is the amplitude of the typical current jump for one channel opening or closing, and *V* = 20 mV is the applied voltage. The molarity *C* of the reaction solution was converted from the osmolarity, which was measured as 1.02 osmol/kg. Because of the complexity of the reaction solution, the number of dissociable particles per molecule (*n*) was assumed to be two, as is the case for NaCl. Thus, the molarity was obtained as *C* = 0.5 M. Using this value, the molar conductivity results as *G* =18.4 S·M−1·m<sup>−</sup>1. The corresponding prediction of the pore radius of an Arch-3 channel, *r* ≈ (0.31 ± 0.02) nm, is in excellent agreemen<sup>t</sup> with the radius of the pore of BR (0.4–1.1) nm [42] and of rhodopsin (0.45–0.7) nm [41], which also consist of seven transmembrane helices forming the same structure.

**Figure 3.** (**a**) Current signal as a function of time for an Arch-3-containing bilayer under continuous 532 nm laser illumination (black line). The red line was obtained by ''change point analysis" from the software Origin (Origin 2021b; OriginLab Corporation, Northampton, Massachusetts, USA). It shows the mean value of each current step. (**b**) The histogram shows the frequency of certain current steps grouped in intervals of 2.5 pA.
