Solvation Dynamics of Trimethylamine N-Oxide in Aqueous Solution

Jul 27, 2015 - ... Trimethylamine Oxide in Muscles of Deep-Sea Gadiform Teleosts: A ...... spectroscopy system using 10 GHz asynchronous optical sampl...
0 downloads 0 Views 16MB Size
Page 1 of 83

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

Solvation Dynamics of Trimethylamine N-Oxide in Aqueous Solution Probed by Terahertz Spectroscopy Lukas Knake, Gerhard Schwaab,∗ Konstantin Kartaschew, and Martina Havenith∗ Department of Physical Chemistry II, Ruhr-University Bochum, 44780 Bochum E-mail: [email protected]; [email protected]

keywords: Hydration, Terahertz Spectroscopy, TMAO



To whom correspondence should be addressed

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 We have studied the hydration dynamics of Trimethylamine N-oxide (TMAO) in aqueous solution using a combination of concentration dependent THz/far-infrared and Raman spectroscopic techniques. Terahertz/FIR absorption was measured using narrow band (76 - 93 cm-1 ) p-Ge laser and broad band (30 - 400 cm-1 ) Fourier Transform spectroscopy. We used principal component analysis in combination with a semi-ideal chemical equilibrium model to dissect the spectra into linear and non-linear contributions of the solvated solute extinction. We attribute the linear part to the average extinction/Raman scattering of TMAO-water aggregates with approximately 3–4 water strongly hydrogen-bonded to TMAO. An additional non-linear concentration dependence indicates a decrease of the number of attached water molecules with increasing TMAO concentrations due to a shift in association equilibria. The Raman spectra reveal a frequency shift of the (narrowband) intramolecular vibrations with decreasing dilution. Based on the results of a detailed analysis and isotopic substitution the experimentally absorption bands at 0 cm-1 , 176 cm-1 , and 388 cm-1 could be assigned to water relaxation modes, an intermolecular TMAO–H2 O stretch and the C-N-C bending mode, respectively. Our results provide evidence for a local modification of the water structure.

2 ACS Paragon Plus Environment

Page 2 of 83

Page 3 of 83

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 Trimethylamine N-oxide (TMAO) (CH3 )3 N+ O− is a natural osmolyte, which is found in the tissue of various marine animals living under extreme pressure conditions. 1–4 As protecting osmolyte, TMAO is accumulated in the cytoplasm of many organisms in order to raise the osmotic pressure when the cell is under stress. TMAO is known to stabilize the conformational structure of proteins against hydrostatic pressure 5,6 and to counteract the influence of denaturing osmolytes, such as urea. 7,8 Although the biological functions of TMAO are well studied, the underlying molecular mechanism is still under debate. Several theories have been proposed in an attempt to explain how osmolytes promote protein stability. The exclusion theory proposes that protecting osmolytes are preferentially excluded from the peptide surface because of unfavorable interactions between the osmolyte and the peptide backbone. 9 In a second model the stabilization is attributed to an indirect effect: TMAO is proposed to influence the water network structure, causing a strengthening of the hydrogen-bond network. 7,10–14 Alternatively, recent MD simulations suggest that TMAO preferentially binds to the folded state (of a hydrophobic polymer) due to the favorable free energy of insertion of a single osmolyte near the folded state compared to the unfolded state. 15 Neutron-scattering experiments 7 have revealed that the oxygen of TMAO forms strong hydrogen bonds with two or three water molecules. On average, the water oxygen–water oxygen distance indicates a more compact water hydrogen-bond network with reduced long range order compared to bulk water. Based upon broadband dielectric spectroscopic and femtosecond midinfrared spectroscopic measurements Rezus and Bakker 16,17 proposed that the librations of the water hydroxyl groups close to the hydrophobic part of the molecule (i.e. the three methyl groups) are slowed down by a factor of four from 2.5 to >10 ps. They concluded that each methyl group immobilizes four water molecules. In contrast, the reorientation time of the mobile water fraction was found to be decreased for TMAO compared to bulk water and other osmolytes. 17 This increased flexibility was attributed to a higher density of network defects in the hydrogen-bond network of water. 2DIR experiments of TMAO solution by Bakulin et al. 18 show also a significant retardation of the 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

water dynamics that correlates with slow spectral diffusion. Molecular dynamics (MD) simulations by Laage et al. 19 predicted only a moderate slowdown of the water molecules around the methyl groups of TMAO. They observed an inhomogeneous retardation dynamics, i.e. that the retardation of water molecules is more pronounced in the vicinity of the hydrophilic N-O group compared to water close to the hydrophobic groups. In agreement with this result, Hunger et al. 20 reported a significant interaction of water molecules with the hydrophilic site of the TMAO molecule. They proposed the formation of stable TMAO·2H2 O and/or TMAO·3H2 O complexes. Mazur et al. 21 have investigated aqueous TMAO solutions using ultrafast Optical Kerr effect spectroscopy in the far infrared region. They reported three resonances centered at 45 cm-1 , 175 cm-1 and 380 cm-1 . The band at 380 cm-1 was assigned to an intramolecular mode. 21 The first two bands were also found in the Raman spectrum of bulk water. 21,22 The authors concluded that TMAO does not disturb the water structure significantly. Infrared and Raman spectroscopic measurements 23–26 yielded no evidence for either TMAO self-aggregation for TMAO. Recently, by combining concentration dependent Raman spectroscopy and electronic structure calculations of small TMAO water aggregates (up to eight water) Munroe et al. 27 supported the view that TMAOwater aggregates involve on average at least three strong hydrogen bonds located at the hydrophilic site. We report a combination of concentration dependent THz/far-infrared(FIR) absorption measurements in the frequency range between 30 and 400 cm-1 (1.1-12 THz) and Raman measurements to investigate the hydration of TMAO. THz/FIR spectroscopy provides a sensitive tool to probe changes in the collective hydrogen bond network around solutes such as proteins 28–30 small osmolytes 30–33 and ions. 34–37

4 ACS Paragon Plus Environment

Page 4 of 83

Page 5 of 83

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

Experimental Sample Preparation Trimethylamine N-oxide dihydrate (purity 99.0%) was purchased from Fulka. The weighted amount was dissolved in HPLC grade ultra pure water for each concentration. Density measurements were performed with a DMA 58 density meter from Anton Paar.

THz Fourier Transform Spectroscopy Concentration dependent THz spectra of TMAO in water were recorded in the frequency range between 30 and 400 cm-1 using a Bruker Vertex 80v Fourier Transform Infrared spectrometer with an Infared Laboratories liquid-helium-cooled silicon bolometer as detector and a silicon carbide lamp as light source. We used a Bruker liquid cell with chemical vapor deposition-grown diamond windows (Diamond Materials, GmbH). The sample layer thickness was fixed by a Kapton spacer to 25 ± 1 µm. The precise layer thickness was determined by recording the interference pattern of the empty cell. The sample compartment was purged with nitrogen gas to minimize humidity. The temperature of the sample cell was kept constant at 20.0 ± 0.2 ◦ C. Each spectrum was measured with a resolution of 2 cm-1 and averaged three times over 128 scans. The absorption coefficient was determined using Lambert-Beer’s law:   I0 (ν) 1 α(ν) = ln d I(ν)

(1)

with α(ν), d, I0 (ν) and I(ν) as frequency dependent absorption coefficient, layer thickness, initial intensity and transmitted intensity, respectively. The effective solute absorption was determined by subtracting the scaled water contribution cw εbulk from the absorption coefficient of the sample solution, assuming in the first instance that all water in the sample behaves like bulk water:

α eff = αsolution − cw εbulk

5 ACS Paragon Plus Environment

(2)

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 6 of 83

with cw as the water concentration in the sample solution. α eff probes the solute induced changes of the solvation dynamics compared to bulk water. The water concentration within the sample cw was derived from density measurements. The effective molar extinction ε eff was determined by dividing α eff by the solute concentration csolute : ε eff = α eff /csolute

(3)

THz Narrow Band Absorption The averaged change of the THz absorption in the frequency range from 76 to 93 cm−1 (2.12.8 THz) was measured using our p-type germanium laser difference spectrometer. The system is described in detail elsewhere. 38,39 We used a double-beam configuration with sample and reference cell to compensate for small drifts of the experimental conditions, like temperature, humidity, and laser power. The temperature was kept at 20.0 ± 0.2 ◦ C and the humidity was controlled to be less than 3%. A liquid helium cooled photoconductive germanium detector served as detector. Each data point was averaged over 30000 pulses. We measured the absorption difference between the sample and bulk water which can be described by Beer’s law:   Isample I0,reference 1 ∆α = αsample − αwater = − ln · d I0,sample Ireference

(4)

with Isample , Ireference being the intensity of the sample solution and the intensity of bulk water under identical conditions, respectively. I0,sample and I0,reference designate the THz absorption measurements of neat water in the sample and reference channel. We added the absorption of bulk water 36 to ∆α in order to obtain the absorption of the sample solution: α = ∆α + cw0 εbulk where cw0 is the water concentration in bulk water.

6 ACS Paragon Plus Environment

(5)

Page 7 of 83

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

Raman Spectroscopy Raman measurements were carried out using a confocal Raman microscope (alpha300 R/A/S) from WITec (Ulm, Germany). A frequency-doubled Nd:YAG laser with a wavelength of 532 nm was used for excitation. The laser beam was coupled into the microscope through a single mode optical fiber and then focused on the sample surface with a 20x objective (NA=0.4). The resulting Raman scattered light was collected with the same objective and registered by a back-illuminated electron multiplying charge-coupled device (EMCCD) detector (200*1600 pixels, cooled to -60 ◦ C)

after passing a multimode fiber (50 µm diameter) and a diffraction grating (600 grooves per

mm). All measurements were carried out in the spectral range from 300 to 1600 cm-1 with a spectral resolution of 10 ps. They concluded that each methyl group immobilizes four water molecules. In contrast, the reorientation time of the mobile water fraction was found to be decreased for TMAO compared to bulk water and other osmolytes. 17 This increased flexibility was attributed to a higher density of network defects in the hydrogen-bond network of water. 2DIR experiments of TMAO solution by Bakulin et al. 18 show also a significant retardation of the 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

water dynamics that correlates with slow spectral diffusion. Molecular dynamics (MD) simulations by Laage et al. 19 predicted only a moderate slowdown of the water molecules around the methyl groups of TMAO. They observed an inhomogeneous retardation dynamics, i.e. that the retardation of water molecules is more pronounced in the vicinity of the hydrophilic N-O group compared to water close to the hydrophobic groups. In agreement with this result, Hunger et al. 20 reported a significant interaction of water molecules with the hydrophilic site of the TMAO molecule. They proposed the formation of stable TMAO·2H2 O and/or TMAO·3H2 O complexes. Mazur et al. 21 have investigated aqueous TMAO solutions using ultrafast Optical Kerr effect spectroscopy in the far infrared region. They reported three resonances centered at 45 cm-1 , 175 cm-1 and 380 cm-1 . The band at 380 cm-1 was assigned to an intramolecular mode. 21 The first two bands were also found in the Raman spectrum of bulk water. 21,22 The authors concluded that TMAO does not disturb the water structure significantly. Infrared and Raman spectroscopic measurements 23–26 yielded no evidence for either TMAO self-aggregation for TMAO. Recently, by combining concentration dependent Raman spectroscopy and electronic structure calculations of small TMAO water aggregates (up to eight water) Munroe et al. 27 supported the view that TMAOwater aggregates involve on average at least three strong hydrogen bonds located at the hydrophilic site. We report a combination of concentration dependent THz/far-infrared(FIR) absorption measurements in the frequency range between 30 and 400 cm-1 (1.1-12 THz) and Raman measurements to investigate the hydration of TMAO. THz/FIR spectroscopy provides a sensitive tool to probe changes in the collective hydrogen bond network around solutes such as proteins 28–30 small osmolytes 30–33 and ions. 34–37

4 ACS Paragon Plus Environment

Page Page46 4 of 83 41

Page 47 5 ofof41 83

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

Experimental Sample Preparation Trimethylamine N-oxide dihydrate (purity 99.0%) was purchased from Fulka. The weighted amount was dissolved in HPLC grade ultra pure water for each concentration. Density measurements were performed with a DMA 58 density meter from Anton Paar.

THz Fourier Transform Spectroscopy Concentration dependent THz spectra of TMAO in water were recorded in the frequency range between 30 and 400 cm-1 using a Bruker Vertex 80v Fourier Transform Infrared spectrometer with an Infared Laboratories liquid-helium-cooled silicon bolometer as detector and a silicon carbide lamp as light source. We used a Bruker liquid cell with chemical vapor deposition-grown diamond windows (Diamond Materials, GmbH). The sample layer thickness was fixed by a Kapton spacer to 25 ± 1 µm. The precise layer thickness was determined by recording the interference pattern of the empty cell. The sample compartment was purged with nitrogen gas to minimize humidity. The temperature of the sample cell was kept constant at 20.0 ± 0.2 ◦ C. Each spectrum was measured with a resolution of 2 cm-1 and averaged three times over 128 scans. The absorption coefficient was determined using Lambert-Beer’s law:   I0 (ν) 1 α(ν) = ln d I(ν)

(1)

with α(ν), d, I0 (ν) and I(ν) as frequency dependent absorption coefficient, layer thickness, initial intensity and transmitted intensity, respectively. The effective solute absorption was determined by subtracting the scaled water contribution cw εbulk from the absorption coefficient of the sample solution, assuming in the first instance that all water in the sample behaves like bulk water:

α eff = αsolution − cw εbulk

5 ACS Paragon Plus Environment

(2)

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 Page48 6 of 83 41

with cw as the water concentration in the sample solution. α eff probes the solute induced changes of the solvation dynamics compared to bulk water. The water concentration within the sample cw was derived from density measurements. The effective molar extinction ε eff was determined by dividing α eff by the solute concentration csolute : ε eff = α eff /csolute

(3)

THz Narrow Band Absorption The averaged change of the THz absorption in the frequency range from 76 to 93 cm−1 (2.12.8 THz) was measured using our p-type germanium laser difference spectrometer. The system is described in detail elsewhere. 38,39 We used a double-beam configuration with sample and reference cell to compensate for small drifts of the experimental conditions, like temperature, humidity, and laser power. The temperature was kept at 20.0 ± 0.2 ◦ C and the humidity was controlled to be less than 3%. A liquid helium cooled photoconductive germanium detector served as detector. Each data point was averaged over 30000 pulses. We measured the absorption difference between the sample and bulk water which can be described by Beer’s law:   Isample I0,reference 1 ∆α = αsample − αwater = − ln · d I0,sample Ireference

(4)

with Isample , Ireference being the intensity of the sample solution and the intensity of bulk water under identical conditions, respectively. I0,sample and I0,reference designate the THz absorption measurements of neat water in the sample and reference channel. We added the absorption of bulk water 36 to ∆α in order to obtain the absorption of the sample solution: α = ∆α + cw0 εbulk where cw0 is the water concentration in bulk water.

6 ACS Paragon Plus Environment

(5)

Page 49 7 ofof41 83

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

Raman Spectroscopy Raman measurements were carried out using a confocal Raman microscope (alpha300 R/A/S) from WITec (Ulm, Germany). A frequency-doubled Nd:YAG laser with a wavelength of 532 nm was used for excitation. The laser beam was coupled into the microscope through a single mode optical fiber and then focused on the sample surface with a 20x objective (NA=0.4). The resulting Raman scattered light was collected with the same objective and registered by a back-illuminated electron multiplying charge-coupled device (EMCCD) detector (200*1600 pixels, cooled to -60 ◦ C)

after passing a multimode fiber (50 µm diameter) and a diffraction grating (600 grooves per

mm). All measurements were carried out in the spectral range from 300 to 1600 cm-1 with a spectral resolution of