Kinetic Study of Ligand Binding and Conformational Changes in

Jul 2, 2018 - *Telephone: +49-(0)721-608 43401; Fax: +49-(0)721-608 48480; E-mail: [email protected]. .... Figure 1 shows a close-up of the active site...
1 downloads 0 Views 1MB Size
Subscriber access provided by NEW MEXICO STATE UNIV

Article

A Kinetic Study of Ligand Binding and Conformational Changes in Inducible Nitric Oxide Synthase Michael Horn, Karin Nienhaus, and Gerd Ulrich Nienhaus J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b05137 • Publication Date (Web): 02 Jul 2018 Downloaded from http://pubs.acs.org on July 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

A Kinetic Study of Ligand Binding and Conformational Changes in Inducible Nitric Oxide Synthase

Michael Horn1, Karin Nienhaus1, G. Ulrich Nienhaus1,2,3*

1

Institute of Applied Physics Karlsruhe Institute of Technology (KIT) Wolfgang-Gaede-Str. 1 D-76131 Karlsruhe Germany 2

Institute of Nanotechnology (INT) and Institute of Toxicology and Genetics (ITG) Karlsruhe Institute of Technology (KIT) D-76344 Eggenstein-Leopoldshafen Germany 3

Department of Physics University of Illinois at Urbana-Champaign 1110 West Green Street Urbana, IL 61801 USA Running Title: Conformational changes in iNOS *To whom correspondence should be addressed. Telephone: +49-(0)721-608 43401 Telefax: +49-(0)721-608 48480 E-mail: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Nitric oxide synthases (NOSs) are heme enzymes that generate highly reactive nitric oxide from L-arginine (L-Arg) in a complex mechanism that is still only partially understood. We have studied carbon monoxide (CO) binding to the oxygenase domain of murine inducible NOS (iNOS) by using flash photolysis. The P420 and P450 forms of the enzyme, assigned to a protonated and unprotonated proximal cysteine, through which the heme is anchored to the protein, show markedly different CO rebinding properties. The data suggest that P420 has a widely open distal pocket that admits water. CO rebinding to the P450 form strongly depends on the presence of the substrate L-Arg, the intermediate Nω-hydroxy-L-arginine and the cofactor tetrahydrobiopterin. After adding these small molecules to the enzyme solution, the CO kinetics change slowly over hours. This process can be described as a relaxation from a fast rebinding, metastable species to a slowly rebinding, thermodynamically stable species, which we associate with the enzymatically active form. Our results allow us to determine kinetic parameters of L-Arg binding to ferrous deoxy iNOS protein for the first time, and also provide clues regarding the nature of structural differences between the two interconverting species.

2 ACS Paragon Plus Environment

Page 2 of 37

Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

INTRODUCTION Proteins are linear polymers of amino acids that typically fold into tightly packed threedimensional structures. These intricate architectures are only weakly stabilized and perform large-scale structural fluctuations under physiological conditions that oftentimes play an important role in their biological function. Laser flash photolysis studies of the rebinding of small ligands (e.g., carbon monoxide, CO, or dioxygen, O2) to myoglobin (Mb) and other heme proteins over wide temperature ranges, pioneered by Frauenfelder and coworkers,1 played a key role in establishing this dynamic view of proteins. Indeed, a great deal of insight into protein dynamics has been gained by merely following the time course of a simple protein reaction.1-3 The protein energy landscape concept, in which conformational changes are described as trajectories on an energy surface comprising a huge number of microscopic states (“conformational substates”), has originated from such studies.4-6 The catalytic activity of enzymes relies on large-scale protein motions that modulate substrate binding, turnover and product release. The energy landscape will be reshaped by interactions with substrate molecules, but how the resulting changes of structure and dynamics affect catalysis has largely remained elusive. In-depth experimental studies of enzyme function are in general challenging because transient species may be short-lived and not build up to a detectable population, and chemical reactions occurring during catalysis might mask important dynamic events. In recent years, we have studied protein conformational changes during enzymatic action of heme containing oxygenases, an important class of enzymes that transfer oxygen from O2 to a substrate molecule.7-11 The reaction starts with the formation of a ternary complex consisting of 3 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the enzyme, an O2 ligand bound to the heme iron and a substrate molecule residing close to the heme-bound ligand.11 To investigate the assembly of the ternary complex, we block the enzymatic turnover by replacing O2 by CO, which resembles O2 in size and heme binding properties. By using flash photolysis, we monitor the intense Soret absorption bands in the blue region of the visible spectrum, which report on the heme iron oxidation and ligation states, with nanosecond time resolution. In addition, we frequently employ infrared spectroscopy of the CO stretching absorption as a superb complementary marker for probing the active site.8, 10, 12

Here, we have analyzed the effects of cofactor and substrate on CO rebinding to the heme iron in a mammalian inducible nitric oxide synthase (iNOS). The enzyme is expressed in macrophages and up-regulated upon inflammatory and immunologic stimulation so as to generate highly reactive nitric oxide (NO) to protect against microbial pathogens.13 Overexpression of iNOS leads to excessive amounts of free NO radicals and destruction of healthy tissue. Therefore, it is not surprising that iNOS is implicated in the development of various diseases including arthritis, multiple sclerosis, septic shock, diabetes, and transplant rejection,13-17 and efforts are ongoing to identify potent and highly selective inhibitors to regulate its activity.18-20 The iNOS enzyme and its endothelial NOS (eNOS) and neuronal NOS (nNOS) isoforms21-23 catalyze the production of NO in a two-step monooxygenation of L-arginine (L-Arg) to NO and Lcitrulline via the intermediate Nω-hydroxy-L-arginine (NOHA).24-26 iNOS is only catalytically active as a homodimer,24, 27, 28 with its protomers consisting of two domains. The N-terminal oxygenase domain (iNOSoxy) contains the heme prosthetic group and binds the 4 ACS Paragon Plus Environment

Page 4 of 37

Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

tetrahydrobiopterin (H4B) cofactor as well as the substrate L-Arg. The C-terminal reductase domain (iNOSred) contains NADPH, FAD and FMN as redox-active cofactors.29 The two domains are linked by an α-helical calmodulin binding sequence. Calmodulin binding mediates interactions between iNOSred on one protomer and iNOSoxy on the other one, which is essential for proper electron injection into the active site during catalysis.30 To simplify structure-function studies, the polypeptide chain of 130 kDa can be cleaved by trypsin into iNOSoxy (56 kDa) and iNOSred (74 kDa) to study them separately. Both domains maintain their structural integrity, but only iNOSoxy can still form an active dimer via a head-to-head alignment.31 Furthermore, deletion of the first 65 residues of iNOSoxy results in a variant, Δ65 iNOSoxy, having essentially the same activity as wildtype iNOSoxy.32 Residues 66 ‒ 114, however, consisting of a β-hairpin hook, a zinc binding motif, and an H4B-binding segment, are essential for dimerization.33 Figure 1 shows a close-up of the active site of murine iNOSoxy. The heme group is anchored to the protein via Cys194 at the fifth coordination of the heme iron, which is a characteristic of P450 type enzymes.34 Absorption spectra of iNOSoxy-CO often display two Soret bands in the blue region of the visible range, denoted P420 and P450. They are assigned to protein species having a protonated and deprotonated Cys194 side chain, respectively.35-37 In the P450 form, Cys194 is stabilized by a hydrogen-bond interaction with Trp188, and the H4B cofactor is hydrogen-bonded to one of the two propionate groups of the heme. The L-Arg substrate points into the heme pocket, positioning its guanidino moiety in close proximity to the oxygen molecule at the sixth coordination of the heme iron. By using steady state and nanosecond time-resolved optical spectroscopy at room temperature, we have investigated CO binding to the iNOSoxy domain. The P450 and P420 species rebind CO markedly differently, and CO 5 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

rebinding to the P450 form is strongly affected by the presence of the substrate L-Arg, the intermediate NOHA and the cofactor H4B. Upon addition of these small molecules to the enzyme solution, the CO kinetics change slowly over hours. The process can be described as a relaxation from a fast rebinding, metastable species to a slowly rebinding, thermodynamically stable species, which we associate with the enzymatically active form. From the L-Arg concentration dependence of the CO rebinding kinetics, we were able to extract the association and dissociation rate coefficients of substrate L-Arg binding to the ferrous deoxy iNOS protein for the first time.

EXPERIMENTAL SECTION Sample Preparation. The murine iNOSoxy domain, with its first 65 residues deleted (Δ65 iNOSoxy, referred to as iNOSoxy in the following), was expressed and purified as previously described.32 To prepare CO-ligated iNOSoxy samples, 1200 µl buffer solution (100 mM potassium phosphate, pH 7.1, unless otherwise noted, with L-Arg, NOHA or H4B (all Sigma Aldrich, St. Louis, MO, USA) added in the desired concentrations) were equilibrated with an atmosphere containing either 0.05 or 1 atm CO partial pressure for 15 min in an airtight 1 × 1 × 3 cm3 glass cuvette (Hellma, Müllheim, Germany). 25 µl of a freshly prepared, concentrated ferric iNOSoxy stock solution was mixed with the buffer to obtain a final protein concentration of ~10 µM. A two-fold molar excess of an anaerobically prepared sodium dithionite solution (Sigma Aldrich) was added to the cuvette with a gas-tight Hamilton syringe to produce ferrous iNOSoxy-CO.

6 ACS Paragon Plus Environment

Page 6 of 37

Page 7 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

UV/Vis Absorption Spectroscopy. Absorption spectra were measured on a Cary 100 (Agilent, Waldbronn, Germany) dual beam UV/Vis absorption spectrometer from 200 to 700 nm with a resolution of 1 nm. The home-built sample holder was kept at constant temperature with the help of a temperature-controlled water bath (Neslab RTE 7, Thermo Scientific, Waltham, MA, USA). Flash Photolysis. In our home-built flash photolysis apparatus,7 samples were photodissociated by a 6-ns (full width at half maximum) pulse from a frequency doubled Nd:YAG laser (Surelite II10, Continuum, Santa Clara, CA). Ligand rebinding was monitored in the Soret region with light from a tungsten source that was passed through a monochromator set to the desired wavelength. The light transmitted through the sample was detected by a photomultiplier tube (R5600U, Hamamatsu Corp., Middlesex, NJ). The amplified photocurrent was converted to a voltage by a transimpedance amplifier; its two voltage outputs were fed into two analog-todigital (A/D) converter PCI cards (NI-PCI 5114, NI-PCI 6221, National Instruments, Newbury, Berkshire) with differing sampling times (4 ns time steps up to 2 ms and 4 µs steps for 2 µs – 100 s) to capture absorbance changes over ~10 orders of magnitude in time. The time traces from the two A/D converters were time-averaged and combined to yield equally spaced data points on a logarithmic timescale from 10 ns – 100 s. Up to 100 individual transients were averaged for a single final trace. For better visualization, we present only a limited time range in the figures below. Analysis of Flash Photolysis Data with Kinetic Heterogeneity. Frequently, the time dependence of ligand recombination to heme proteins at ambient temperature is fitted either with an

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 37

exponential decay function or a sum of exponentials to model multiple, discrete processes. Such an approach assumes implicitly that any intrinsic kinetic heterogeneity within these processes is averaged on time scales faster than the characteristic time of the processes themselves. In our experiments with iNOSoxy, we have observed that CO rebinding occurs in multiple nonexponential processes. Such a kinetic trace can generally be described by a continuous superposition of exponential processes with apparent rate coefficients, λ,  =   log λ λ exp−λ.

(1)

Extracting the rate spectrum, or rate distribution function, f(λ), by an inverse Laplace transform is an ill-conditioned mathematical operation. We use the Maximum Entropy Method (MEM) to find the best-fit model distribution, f(λ), by maximizing the Shannon-Jaynes entropy.38 Multiple processes are typically visible in f(λ) as discrete subdistributions, fi(λ). Usually, f(λ) can be fitted with a sum of Gaussians, λ = ∑   λ = ∑





exp −



 ! λ"〈 ! λ 〉 

%

(2)

to extract log-averaged rate coefficients, 〈log λi〉, and full widths at half maximum, wi = 2 (2 ln 2)1/2 σi, as a measure of the kinetic heterogeneity within these processes. The width of a subdistribution in the MEM-derived f(λ) is determined by the statistical uncertainty of the data, which usually depends on the timescale of the subprocess, and by kinetic heterogeneity within the process, and great care has to be taken when attempting to disentangle the effects. The fractional populations of the subdistributions are given by the weight factors, ai.

8 ACS Paragon Plus Environment

Page 9 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

RESULTS AND DISCUSSION

P450 and P420 species of iNOSoxy-CO show very different CO recombination. To study CO rebinding to the P420 and P450 forms, we freshly prepared an iNOSoxy-CO sample from the oxidized (met) form (equilibrated with 1 bar CO). Initially, only the P450 species was present; subsequently, the fraction of P420 species slowly increased. We started the kinetic experiments 2 h after preparation, at which point the fractions of P420 and P450 were comparable. Thus, the two Soret bands of iNOSoxy-CO centered on 421 (P420) and 444 nm (P450) are of similar height (Figure 2A, inset); the Soret band of the deoxy species (obtained by adding a two-fold molar excess of an anaerobically prepared sodium dithionite solution to a met sample equilibrated with N2) is broad and peaks at 414 nm. Figure 2A displays kinetic traces after photodissociation at selected wavelengths. The absorbance changes can be positive or negative, depending on the monitoring wavelength, as indicated by the crossing of the steady-state absorption spectra (Figure 2A, inset). We collected kinetic traces between 400 and 450 nm every 2 nm to obtain time-dependent optical difference spectra (Figure 2B, C), which represent spectral changes associated with bimolecular CO rebinding from the solvent. Between 1 µs and 1 ms after photodissociation, CO recombination causes the Soret band of the CO-bound P450 species at 444 nm to recover at the expense of the deoxy Soret band at shorter wavelengths (Figure 2B). The isosbestic point at 432 nm confirms that these changes represent a two-state interconversion. Between 1 ms and 1 s after photodissociation, we observe the recovery of the P420 species (Figure 2C), although with an ~20-fold smaller amplitude. Figure 2D shows a two-dimensional plot (wavelength vs. log (time/s)) of the absorbance change between successive time points of the kinetic traces, 9 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

representing the numerical time derivative. Of note, the contour spacing is decreased by a factor of ten above 1 ms to enhance the weak signal from the CO-bound P420 species. The two red peaks are markedly shifted in time, indicating that CO recombination in the P420 species is ~20-fold slower than in the P450 species. The P420 and P450 species are present in roughly equal proportions (Figure 2A, inset) and it is well established that CO can be photolyzed from the heme iron with ~100% quantum yield.39, 40 So why are the photolysis-induced absorbance differences of the two species on the microsecond timescale so different and, moreover, why are they kinetically so different? We can address these questions with the help of a three-well model, a somewhat simplified kinetic description capturing ligand recombination in heme proteins. The model includes a ligandbound state, a geminate state with the ligand in the heme interior, and a (deoxy) state without a ligand in the protein. The three states are separated by enthalpy barriers governing the transition probabilities. Notably, the validity of the three-well model at physiological temperatures was experimentally proven for Mb in a pioneering study by Eaton and coworkers.41 After photodissociation, CO may either rebind without leaving the protein (geminately, within