Microsecond-Resolved Infrared Spectroscopy on Nonrepetitive Protein Reactions by Applying Caged Compounds and Quantum Cascade Laser Frequency Combs

Cascade Laser Frequency Combs

Article

Microsecond-Resolved Infrared Spectroscopy on Nonrepetitive

Protein Reactions by Applying Caged Compounds and Quantum

Mohamad Javad Norahan, Raphael Horvath, Nathalie Woitzik, Pierre Jouy, Florian Eigenmann,

Klaus Gerwert, and Carsten K ö tting*

Cite This: Anal. Chem. 2021, 93, 6779 −6783

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: Infrared spectroscopy is ideally suited for the

investigation of protein reactions at the atomic level. Many systems

were investigated successfully by applying Fourier transform

infrared (FTIR) spectroscopy. While rapid-scan FTIR spectrosco-

−

py is limited by time resolution (about 10 ms with 16 cm

resolution), step-scan FTIR spectroscopy reaches a time resolution

of about 10 ns but is limited to cyclic reactions that can be repeated

hundreds of times under identical conditions. Consequently, FTIR

with high time resolution was only possible with photoactivable

proteins that undergo a photocycle. The huge number of

nonrepetitive reactions, e.g., induced by caged compounds, were

limited to the millisecond time domain. The advent of dual-comb

quantum cascade laser now allows for a rapid reaction monitoring

in the microsecond time domain. Here, we investigate the potential to apply such an instrument to the huge class of G-proteins. We

compare caged-compound-induced reactions monitored by FTIR and dual-comb spectroscopy by applying the new technique to the

α subunit of the inhibiting G protein and to the larger protein−protein complex of Gα with its cognate regulator of G-protein

signaling (RGS). We observe good data quality with a 4 μs time resolution with a wavelength resolution comparable to FTIR. This is

more than three orders of magnitude faster than any FTIR measurement on G-proteins in the literature. This study paves the way for

infrared spectroscopic studies in the so far unresolvable microsecond time regime for nonrepetitive biological systems including all

GTPases and ATPases.

INTRODUCTION

but with a conventional globar as the light source, signal-to-

■

noise ratios (S/N) are not suﬃcient for single-shot experi-

Fourier transform infrared (FTIR) spectrometers revolution-

ized infrared spectroscopy in the 1970s. Likely, the advent of

stable mid-IR quantum cascade lasers (QCLs) will impact

infrared spectroscopy to the same extent. Conventional QCLs

are superior with regard to brilliance, but they lack the

multiplex advantage of FTIR that is especially helpful in time-

resolved measurements of proteins. For systems that can be

excited repetitively, step-scan FTIR can provide time-resolved

ments on biological systems. Synchrotron irradiation in

combination with a dispersive setup and an array detector

permits single-shot experiments with microsecond resolution.

However, even here S/N was not suﬃcient for single-shot

experiments of a protein reaction.

The recent development of dual-comb QCL-based spec-

trometers allows for microsecond time resolution as well, but

11,12

spectra with nanosecond resolution. For these repetitive

with a much smaller footprint.

In these instruments, two

systems, tunable QCLs provide time-resolved spectra with 10

broad band lasers that emit at many discrete wavelengths are

used as the light source (Figure 1). The wavelength spacing of

^t^h^e^ﬁ^r^s^t^l^a^s^e^r⁽^frep,1) is close but not identical to the other laser

ns resolution. Pump-probe experiments (visible pump and

IR probe) even allow for femtosecond time-resolved IR

spectroscopy. However, samples that allow only single

excitations can either be measured only at a single wavelength

or with rapid-scan FTIR. The time resolution of rapid-scan

FTIR depends on the scanning velocity of the Michelson

interferometer and is limited to about 10 ms at a wavelength

resolution of 16 cm⁻within high-end research FTIR

instruments. The implementation of a faster Michelson

interferometer allows, in principle, for a higher time resolution,

Received: February 11, 2021

Accepted: April 8, 2021

Published: April 21, 2021

2021 American Chemical Society

779

Anal. Chem. 2021, 93, 6779−6783

Analytical Chemistry

Article

coupling GDP with pHP-caged P . pHP-P was obtained in ﬁve

steps from para-hydroxy-acetophenone and dibenzylphos-

phate.

Gα_i₁and RGS proteins were expressed and puriﬁed as

described by Mann et al. In the puriﬁed proteins, the

nucleotide GDP was exchanged for pHP-GTP. The exchange

rate was >95% as checked by reversed-phase high-performance

liquid chromatography (HPLC; LC-2010; Shimadzu) [mobile

phase: 50 mM Pi (pH 6.5), 5 mM tetrabutylammonium-

bromide, 7.5% (vol/vol) acetonitrile; stationary phase: ODS-

Hypersil C18 column]. For the intrinsic Gα measurements,

Gα ·pHP-GTP was lyophilized. For the samples, lyophilized

protein was resuspended in buﬀer to reach the following

concentrations: 5 mM Gα , 200 mM N-(2-hydroxyethyl)-

piperazine-N′-ethanesulfonic acid (Hepes; pH 7.5), 150 mM

NaCl, 5 mM MgCl , 200 mM dithiothreitol (DTT), and 0.1%

(

vol/vol) ethylene glycol. For the RGS-catalyzed reactions, a

:1 molar ratio of Gα_i₁with RGS was lyophilized and

resuspended to reach 5 mM Gα ·RGS, 100 mM Hepes (pH

Figure 1. (A) Principle of dual-comb spectroscopy. (B) Schematic of

the setup of the dual-comb spectrometer used here.

7.5), 100 mM Tris (pH 7.5), 150 mM NaCl, 5 mM MgCl , 20

mM DTT, and 0.1% (vol/vol) ethylene glycol. The samples

(

f_r_e_p_,₂). Overlaying the two lasers produces a set of beatings

were packed between two CaF windows, separated by a spacer

spaced by Δf = f_r_e_p_,₁− f in the radiofrequency domain

ring yielding a pathlength of about 60 μm, that were sealed

with silicon grease and mounted either in a Bruker Vertex 80v

spectrometer or an IRsweep IRis-F1 dual-comb spectrometer.

All measurements were done at room temperature (293 K).

The reactions were initiated by ﬂashes of a XeCl-excimer laser

(308 nm, 150 mJ, Coherent LPX Pro 240). FTIR measure-

rep

rep,2

measured by a high-bandwidth mercury−cadmium−telluride

(

MCT) detector. From this measurement, the entire

heterodyne beating pattern can be recovered. In our setup,

Δf leads to a time resolution of 4 μs. The spectral window of

rep

−

the lasers is from 1207.0 to 1276.8 cm . By guiding the QCLs

through a sample that is irradiated by a pulsed UV laser, UV

light-induced spectral changes of the sample can be obtained

−

ments were recorded at a 4 cm spectral resolution,

manipulated by zero ﬁlling by a factor of 2, and Fourier-

transformed using Mertz phase correction and with a

Blackman−Harris three-term apodization function.

with a time resolution of 4 μs.

Caged compounds are a tool for the investigation of

reactions of biological systems that can otherwise not be

initiated by light. Caged compounds release the reactive

compound upon irradiation and cannot be excited repetitively.

Rapid-scan FTIR spectrometers usually have a time resolution

of about 10 ms at a 16 cm⁻spectral resolution. Due to limited

S/N, the actual time resolution for measuring protein reactions

is often even slower and often artiﬁcial low temperatures are

necessary to slow down the reaction. Most prominently,

The operating principles of the IRis-F1 spectrometer have

11,12,20

been described in detail.

Brieﬂy, two continuous-wave

QCL frequency combs with Δf = 3.2 MHz were used to

rep

obtain the heterodyne beating signal. Under the operating

conditions used here, f_r_e_p_,₁and f_r_e_p_,₂were about 9.82 GHz, as

determined from the interline beating measured with a

spectrum analyzer (FSW26, Rohde & Schwarz). This gives a

−

native spectral resolution of 0.3275 cm . The spectral output

of the individual lasers is presented in Figure S1. The power at

the sample was adjusted to be about 3 mW. We obtained data

in two measurement modes, as described further in Figure S2.

In time-resolved mode, each acquisition is 33.6 ms long,

sampled at 2 GS/s, and fast Fourier transforms are taken at

intervals of 4 μs, which determines the time resolution of the

ATPases and GTPases were investigated in detail by means

of caged nucleotides. As caged compounds, the P -[1-(2-

nitrophenyl)ethyl] ester (NPE) and the P -[para-hydrox-

yphenacyl] ester (pHP) of the nucleotides were used.

Here, we used, for the ﬁrst time, a dual-comb QCL-based

spectrometer for the investigation of GTPases in the

microsecond time domain. First, we measured the photolysis

reaction of two caged GTPs, NPE-GTP and pHP-GTP, in

solution without a protein present and compare the results

with conventional rapid-scan FTIR. In the next step, we

applied the technique to the Gα subunit of the heterotrimeric

^Δ^frep

11,21

measurement. Time resolutions up to

are possible,

but

given the rates of the processes investigated here, and a drop in

signal-to-noise ratio at lower integration times, 4 μs was

chosen. In this mode, each additional acquisition takes a

relatively long time to complete and serves only to improve the

signal-to-noise ratio by synchronizing with repeated excita-

tions. In the long-term acquisition mode, each acquisition is 1

ms long and repeats with a frequency of approximately 100 Hz.

The time-resolved dual-comb IR data were obtained from

combining measurements from both modes, which were each

logarithmically averaged. Before merging, the data were scaled

according to the diﬀerence spectra at peaks wavenumber of

each dataset. Furthermore, to avoid discontinuities, we allowed

for an ∼15 ms overlap between both datasets. The originally

recorded high spectral resolution of the IRsweep instrument at

G protein and monitor its GTPase reaction. Finally, the

regulator of G-protein signaling (RGS)-catalyzed reaction of

Gα was measured for the ﬁrst time at ambient temperatures by

time-resolved infrared spectroscopy. At room temperature, the

reaction is completed before the ﬁrst datapoint of a rapid-scan

FTIR measurement can be recorded. With dual-comb IR, it is

well resolved, including a so far unknown intermediate.

EXPERIMENTAL SECTION

The P -[1-(2-nitrophenyl)ethyl] ester (NPE) of GTP was

obtained from Jena Bioscience (Jena, Germany). The P -[para-

■

−

−1

hydroxyphenacyl] ester (pHP) of GTP was synthesized by

0. 3275 cm was also averaged to 4.2 cm for better S/N.

780

Anal. Chem. 2021, 93, 6779−6783

Analytical Chemistry

spectrometer (Figure 2 ).

Article

The data were further analyzed by a global ﬁt (eq 1). The

time-resolved absorbance change ΔA(ν,t) is described by the

The photolysis diﬀerence spectrum of the FTIR experiment

can be compared with the same diﬀerence from the dual-comb

experiment. In both cases, the spectra are shown in a way that

the newly formed absorptions are facing upward and vanishing

absorptions are facing downward. Figure 2B shows that the

spectra agree nicely. The spectral window of the dual-comb

experiment shows nicely the vanishing band of the combined

asymmetric stretching vibrations of the phosphate groups of

absorbance change induced by photolysis a (ν) followed by a

number n of exponential functions ﬁtting the amplitudes a for

each wavenumber ν. The amplitude spectra have negative

peaks for absorptions that disappear (or lose intensity) and

positive peaks for absorptions that emerge (or gain intensity).

a_l(ν)(1 − e⁻

^kl^t

)

the aci-nitro-GTP intermediate.

The same reaction was investigated before by a step-scan

ΔA(ν, t) = a (ν) +

∑

l=1

(1)

FTIR experiment with a 10 μs time resolution at a spectral

For easier comparison, we give the half-life (t₁_/₂= ln(2)/k) in

the manuscript. Due to the limited intensity of the excitation

beam and the insuﬃcient overlap of the excitation laser with

the absorption spectrum of the caged compounds, not the

complete sample is photolyzed by one laser ﬂash. Note that a

more intense excitation beam could also lead to an increased

heat artifact. To estimate the amount of photolysis, the signal

−1 24

resolution of 15 cm . The data were obtained within 5 h of

measurement time, using 200 × 200 μm segments from ﬁve

samples of 1 cm area. In comparison, the dual-comb

experiment was done with a single sample. Compared to the 5

h of the step-scan FTIR experiment, the dual-comb measure-

ment only took a few seconds. Further, much smaller amounts

of sample are needed. The saving in sample consumption is

about a factor of 10 in our experiment, but could be improved

by another factor of 10 using a cuvette optimized for the QCL

proﬁle. The diameter of the laser beam is about 3 mm and

much smaller than the conventional IR cuvettes, as shown in

Figure S5 . The quantum yield of the NPE-GTP hydrolysis with

our 308 nm XeCl-excimer laser is limited. For this reason, we

repeated the experiment 20 times with the same sample in the

same position and co-added the observed changes. The sample

response after each shot was monitored (Figure S3 ). The

kinetics in Figure 2B were obtained by combining the time-

resolved and long-term measurement modes of the IRis-F1

(details of the data treatment are given in the Experimental

Section). We measured a half-life of 46 ms at 293 K for the

intermediate.

−

strength at the most intense analyte peak (1253 cm ) was

integrated over the ﬁrst 2 ms after excitation as a function of

sample excitation number. This was repeated for the negative

times (−2 to 0 ms) as a control, as no spectral features are

expected there. Figure S3 indicates that for NPE-GTP after 20

excitations, no further photolysis occurs. For pHP-GTP after

0 −15 excitations, no further photolysis is observed (Figure

S4 ), and the data shown are the averages of the ﬁrst 10

excitations. In the ﬁgures of the main text, we always show the

results of measurements obtained by a single sample.

RESULTS AND DISCUSSION

■

Photolysis of NPE-GTP does not produce GTP directly; ﬁrst,

an intermediate aci-nitro anion is formed (Figure 2A).

Depending on the reaction conditions, GTP is formed from

this intermediate, usually in the millisecond time regime. We

measured NPE-GTP photolysis with both a Bruker Vertex 80v

FTIR spectrometer and an IRsweep IRis-F1 dual-comb

pHP-GTP is the superior caged compound for the

investigation of fast reactions because photolysis is fast and

without an intermediate. Indeed, we observe very fast

production of GTP from pHP-GTP with only one minor

and very fast kinetic rate that might be a heat artifact (Figure

S4 ). For this reason, we use pHP-GTP for the protein

reactions shown below.

The protein with its GTPase domain (orange) and all-α

domain (yellow) is shown in Figure 3A. The central nucleotide

is shown in a ball-and-stick representation. Figure 3B shows

the reaction scheme of Gα . After irradiation with a 308 nm

excimer laser, we expect the photolysis reaction and

subsequently the hydrolysis. The hydrolysis reaction is slow

and can be observed by rapid-scan FTIR as a control for our

ﬁrst dual-comb experiments with a protein. In both cases, a

global ﬁt of the time-resolved data by applying eq 1 with n = 1

describes the data well. This can be recognized for the dual

comb experiment by the nice agreement of the time-resolved

−

data at 1215 cm with the curve from the ﬁt. Comparing dual

comb and FTIR spectra, the amplitude spectra of photolysis

a_p_hand hydrolysis spectra a both agree nicely (Figure 3C,D).

The absorption peaks were assigned in the literature by

19,26

isotopic labeling.

In the photolysis, a combined asym-

metric stretching vibration of the phosphate groups is assigned,

and in the hydrolysis, the asymmetric PO stretching mode of

the α-phosphate of GTP is shown. Further, the single-

Figure 2. (A) Photolysis of the NPE-GTP reaction scheme. (B)

Photolysis spectra obtained by FTIR (black) and dual-comb (red)

obtained by global ﬁts of the time-resolved data, respectively. (C)

Kinetics of the hydrolysis reaction obtained by dual-comb experi-

ments. The points correspond to the time-resolved data, and the line

corresponds to the global ﬁt according to eq 1.

exponential kinetics (Figure 3E) are in line with the

literature. After having demonstrated the basic functioning

of the dual-comb technique for GTPase reactions of proteins,

we want to investigate a very fast reaction, which cannot be

observed at room temperature by FTIR. The RGS-catalyzed

781

Anal. Chem. 2021, 93, 6779−6783

Analytical Chemistry

Article

Figure 4. (A) Structural model of Gαi·RGS from PDB ID 1AGR. (B)

Kinetics of the hydrolysis reaction of Gαi·RGS. (C) Additional

prehydrolysis rates obtained by dual-comb IR. In (B) and (C), the

points correspond to the time-resolved data and the line corresponds

to the global ﬁt according to eq 1. (D) Three-dimensional (3D) plot

of the changes obtained from the global ﬁt (eq 1) of the dual-comb

data as detailed in the Experimental Section. After photolysis, three

processes (n = 3) with half-lives of 38 μs, 86 μs, and 90 ms were

obtained.

Figure 3. (A) Structural model of Gai from PDB ID 2G83. (B)

Reaction scheme for the GTPase reaction of Gα . (C) Photolysis and

(D) hydrolysis spectra obtained by FTIR (black) and dual-comb

(red) obtained by global ﬁts of the time-resolved data, respectively.

(E) Kinetics of the hydrolysis reaction. The points correspond to the

time-resolved data, and the lines correspond to the global ﬁts

according to eq 1. The absorption of the dual-comb experiment was

normalized to the FTIR experiment, where a thinner sample was used.

reaction of Gα is much faster than the reaction of Gα alone

and cannot be resolved by rapid-scan FTIR at ambient

temperature. The ﬁrst datapoint in the FTIR measurements of

these larger protein−protein complexes (Figure 4A) is usually

above 100 ms (see, e.g., Figure 2B in ref 19). With the dual-

comb experiment, our ﬁrst datapoint is at 4 μs. The kinetics at

experiments including the measurements of further Gα

mutants will be necessary for a clearcut assignment and is

not within the scope of this work. The complete reaction with

all of the information obtained in the dual-comb experiment is

shown in Figure 4D.

−

240 cm show nicely the decay of the α-GTP band due to

CONCLUSIONS

the GTPase reaction. Clearly, the reaction is almost complete

at 100 ms, indicating that the reaction could not be observed

by rapid-scan FTIR at all. A half-life of 90 ms was obtained.

Interestingly, there are even two additional very fast rates

■

Overall, we were able to show that the dual-comb setup is very

suitable for the investigation of proteins with caged

compounds. The only drawback of the new technique is the

relatively small spectral window of each dual-comb setup.

However, laser modules can be changed and modules for all

interesting wavelengths between 2200 and 900 cm⁻are

available, and with the about 100 times stronger source power,

single-shot analyses of weak absorbers are possible. Another

approach could be the measurement of full spectra by FTIR

(Figure 4C half-lives of 38 and 86 μs) resolved, preceding

hydrolysis. We can speculate that one rate corresponds to a

heating artifact (as observed for pHP-GTP alone, Figure S3)

and the other rate corresponds to a fast rearrangement within

the catalytic site. Such a rearrangement was also observed on a

slower timescale in a Gα mutant. However, further

782

Anal. Chem. 2021, 93, 6779−6783

Analytical Chemistry

The Supporting Information is available free of charge at

Article

with low time resolution and subsequent measurement of

interesting regions with a dual-comb setup.

Labudda for help with the sample preparation. They further

thank Dr. Jonas Schartner for the synthesis of pHP-GTP. This

work was supported by Deutsche Forschungsgemeinschaft

(DFG, German Research Foundation), individual Research

Grant GE 599/20-1 to K.G. and KO 3813/1-1 to C.K. and

Research Training Group 2341 “MiCon.” Further support was

provided by the Ministry for Culture and Science (MKW) of

North Rhine-Westphalia (Germany) through grant

111.08.03.05-133974 and the Protein Research Unit Ruhr

within Europe (PURE) funded by the Ministry of Innovation,

Science and Research (MIWF) of North Rhine-Westphalia

We demonstrate that with a single sample, a time resolution

in the microsecond regime can be obtained even for a larger

protein−protein complex. In our setups, we use a sample

thickness of 60 μm, a great advantage of the intense QCLs in

comparison to FTIR, where we use about 10 μm sample

thickness. This alone should lead to an about 6 times larger S/

N ratio. The larger pathlength also allows for a much easier

implementation of ﬂow through setups using conventional

microﬂuidics.

(

Germany).

ASSOCIATED CONTENT

sı Supporting Information

https://pubs.acs.org/doi/10.1021/acs.analchem.1c00666.

■

REFERENCES

■

(

1) Lorenz-Fonfria, V. A. Chem. Rev. 2020 , 120 , 3466 −3576.

2) Schultz, B.-J.; Mohrmann, H.; Lorenz-Fonfria, V. A.; Heberle, J.

Spectrochim. Acta, Part A 2018 , 188 , 666 −674.

Appl. Spectrosc. 1997 , 51 , 558 −562.

(4) Sanchez, M. L. K.; Sommer, C.; Reijerse, E.; Birrell, J. A.; Lubitz,

W.; Dyer, R. B. J. Am. Chem. Soc. 2019 , 141 , 16064 −16070.

(5) Neumann-Verhoefen, M.-K.; Neumann, K.; Bamann, C.; Radu,

I.; Heberle, J.; Bamberg, E.; Wachtveitl, J. J. Am. Chem. Soc. 2013 , 135 ,

Spectral coverage and intensities of the individual QCL

devices (Figure S1); illustration of the acquisition modes

of the dual comb spectrometer (Figure S2); photolysis

signal development during an experiment (Figure S3);

photolysis of pHP-GTP (Figure S4); and comparison of

the conventional IR cuvette with the beam diameter of

the QCL (Figure S5) (PDF)

3) Rammelsberg, R.; Hessling, B.; Chorongiewski, H.; Gerwert, K.

968 −6976.

6) Lin, J.; Gerwert, K.; Kötting, C. Appl. Spectrosc. 2014 , 68 , 531 −

35.

7) Souvignier, G.; Gerwert, K. Biophys. J. 1992 , 63 , 1393 −1405.

8) Suss, B.; Ringleb, F.; Heberle, J. Rev. Sci. Instrum. 2016 , 87 ,

No. 063113.

9) Schade, U.; Ritter, E.; Hegemann, P.; Aziz, E. F.; Hofmann, K. P.

Vib. Spectrosc. 2014 , 75 , 190 −195.

10) Ritter, E.; Puskar, L.; Kim, S. Y.; Park, J. H.; Hofmann, K. P.;

Bartl, F.; Hegemann, P.; Schade, U. J. Phys. Chem. Lett. 2019 , 10 ,

672 −7677.

11) Klocke, J. L.; Mangold, M.; Allmendinger, P.; Hugi, A.; Geiser,

M.; Jouy, P.; Faist, J.; Kottke, T. Anal. Chem. 2018 , 90 , 10494 −10500.

12) Lins, E.; Read, S.; Unni, B.; Rosendahl, S. M.; Burgess, I. J.

Anal. Chem. 2020 , 92 , 6241 −6244.

AUTHOR INFORMATION

Corresponding Author

■

Carsten Kötting − Competence Center for Biospectroscopy,

Center for Protein Diagnostics (ProDi), Ruhr University

Bochum, 44801 Bochum, Germany; Department of

Biophysics, Faculty of Biology and Biotechnology, Ruhr

University Bochum, 44801 Bochum, Germany;

Email: carsten.koetting@rub.de

Authors

Am. Chem. Soc. 1988 , 110 , 7170 −7177.

Mohamad Javad Norahan − Competence Center for

Biospectroscopy, Center for Protein Diagnostics (ProDi),

Ruhr University Bochum, 44801 Bochum, Germany;

Department of Biophysics, Faculty of Biology and

Biotechnology, Ruhr University Bochum, 44801 Bochum,

Germany

Raphael Horvath − IRsweep AG, 8712 Staefa, Switzerland

Nathalie Woitzik − Competence Center for Biospectroscopy,

Center for Protein Diagnostics (ProDi), Ruhr University

Bochum, 44801 Bochum, Germany; Department of

Biophysics, Faculty of Biology and Biotechnology, Ruhr

University Bochum, 44801 Bochum, Germany

Pierre Jouy − IRsweep AG, 8712 Staefa, Switzerland

Florian Eigenmann − IRsweep AG, 8712 Staefa, Switzerland

Klaus Gerwert − Competence Center for Biospectroscopy,

Center for Protein Diagnostics (ProDi), Ruhr University

Bochum, 44801 Bochum, Germany; Department of

Biophysics, Faculty of Biology and Biotechnology, Ruhr

University Bochum, 44801 Bochum, Germany

13) Klán, P.; Solomek, T.; Bochet, C. G.; Blanc, A.; Givens, R.;

(

Rubina, M.; Popik, V.; Kostikov, A.; Wirz, J. Chem. Rev. 2013 , 113 ,

119 −191.

(14) Barth, A.; Kreutz, W.; Mäntele, W. FEBS Lett. 1990 , 277 , 147 −

(

50.

15) Kötting, C.; Gerwert, K. Biol. Chem. 2015 , 396 , 131 −144.

16) Walker, J. W.; Reid, G. P.; McCray, J. A.; Trentham, D. R. J.

17) Park, C.-H.; Givens, R. S. J. Am. Chem. Soc. 1997 , 119 , 2453 −

463.

18) Givens, R. S.; Rubina, M.; Wirz, J. Photochem. Photobiol. Sci.

012 , 11 , 472.

(19) Mann, D.; Teuber, C.; Tennigkeit, S. A.; Schröter, G.; Gerwert,

K.; Kötting, C. Proc. Natl. Acad. Sci. U.S.A. 2016 , 113 , E8041 −E8050.

(20) Pinkowski, N. H.; Ding, Y.; Strand, C. L.; Hanson, R. K.;

Horvath, R.; Geiser, M. Meas. Sci. Technol. 2020 , 31 , No. 055501.

(

21) Coddington, I.; Newbury, N.; Swann, W. Optica 2016 , 3 , 414.

22) Barth, A.; Hauser, K.; Maentele, W.; Corrie, J. E. T.; Trentham,

D. R. J. Am. Chem. Soc. 1995 , 117 , 10311 −10316.

Complete contact information is available at:

23) Cepus, V.; Ulbrich, C.; Allin, C.; Troullier, A.; Gerwert, K.

https://pubs.acs.org/10.1021/acs.analchem.1c00666

Methods Enzymol. 1998 , 291 , 223 −245.

(24) Gerwert, K. Google Patents. WO1999040413A1.

Vib. Spectrosc. 1999 , 19 , 143 −149.

25) Rammelsberg, R.; Boulas, S.; Chorongiewski, H.; Gerwert, K.

26) Schröter, G.; Mann, D.; Kötting, C.; Gerwert, K. J. Biol. Chem.

015 , 290 , 17085 −17095.

27) Mann, D.; Höweler, U.; Kötting, C.; Gerwert, K. Biophys. J.

2017 , 112 , 66 −77.

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

The authors thank Dr. Daniel Mann for the preparation of G_α_i

and Dr. Christian Teuber, Adrian Höveler, and Kristin

■

783