Graphical header for the Optical Technology Division

[skip navigation] Physics Laboratory Home Optical Technology Division Home NIST Home Optical Technology Division Home Division Research Areas Division Products and Services Meetings of Interest

CW THz Spectroscopy of Peptide Nanotubes and Hydrated Crystals

Description:

High-resolution continuous-wave terahertz (THz) methods are being developed for spectroscopic investigations of condensed phase biomolecular systems. THz radiation interrogates the lowest frequency vibrational modes of a biomolecule. These modes characterize the incipient motions for the large scale conformational changes along the torsional degrees of freedom responsible for the flexibility of protein, polynucleotide and polysaccharide backbones. THz vibrational features, therefore, provide a sensitive probe of the force constants influencing the collective nuclear motions that extend over a large portion of the framework and may be used to validate and improve current force field and biomolecular dynamics models. Furthermore, the energy available at THz frequencies is comparable to that available thermally at room temperature making possible detailed investigations of the mechanical anharmonicity important in these large amplitude motions [1-2].

Many of the studies described here are focused on crystalline solids. The discrete THz absorption features of crystals characterize highly degenerate vibrational modes and reflect, to varying degrees, lattice translations, intermolecular librations and intramolecular vibrations. A common property of biomolecular crystals is the extensive hydrogen bonding network that exists throughout the lattice. For peptide crystals, these networks extend along the peptide backbone as well as between the zwitterionic end groups (NH3+ & CO2-). Subtle modifications to such networks via hydration are shown here to impact the phonon frequencies of the lattice to varying degrees depending on the hydrophobic or hydrophilic nature of the interactions.

A brief review of the biomolecular systems under investigation is given following a short description of the cw THz spectrometer.

High Resolution THz Spectrometer

A schematic diagram of the continuous-wave THz spectrometer is given in Fig. 1. A full description of its performance for high resolution THz laser studies is given elsewhere [1-2]. Briefly, the system consists of a low-temperature-grown GaAs photomixer [3] driven at the difference frequency of two near-infrared lasers. The two lasers include a fixed frequency diode laser operating near 850 nm and (ΔνFWHM ∼approximate 0.0001 cm-1) and a newly constructed standing-wave Ti:Sapphire (Ti:Sapp) laser having a resolution of (ΔνFWHM approximate 0.04 cm-1) [4]. The Ti:Sapp laser is broadly tunable and gives more than an order-of-magnitude improvement in resolution over that of the unstabilized ring laser discussed previously [1-2]. The laser is seeded by feedback from an external grating-tuned cavity through a 4% output coupler. This configuration improves the scan-to-scan repeatablility of the spectrometer to better than ±0.03 cm-1, as determined from scans of water vapor in the THz region. One such scan of the water vapor spectrum is illustrated in Fig. 2. The absolute frequency stability is important to minimize non-statistical noise induced by small shifts in the standing wave interference patterns (with modulation depths up to 50% of the transmitted power) when background ratios are performed.

Figure 1a
Figure 1b
Figure 1. A schematic diagram of the cw THz laser spectrometer. The home-built grating-tuned Ti:Sapp laser shown in the lower panel is tunable over a 100 cm-1 interval at a resolution of 0.04 cm-1 [4] and used in combination with a second fixed-frequency laser to generate broadly tunable THz radiation for studies of condensed phase biomolecular systems.

Schematic diagram of the photomixer-based THz spectrometer.
Figure 2. Water vapor absorption spectrum obtained using the THz spectrometer compared to that calculated using the HITRAN database. An expanded portion in the lower panel containing pressure broadened water lines (ΔνFWHM approximately˜ 0.2 cm-1) illustrates the resolving power.
The Ti:Sapp beam is combined with the output from a fixed frequency diode laser and focused by an aspherical lens onto the photomixer. The Ti:Sapp laser frequency is scanned at 0.04 cm-1 resolution to a difference frequency of 100 cm-1. The diode and Ti:Sapp beams are chopped at 400 Hz prior to entering the vacuum chamber containing the photomixer assembly. The maximum THz power generated by the photomixers is approximately1 µW from 0.06 THz to 3 THz, with output power decreasing as ω-4 beyond the peak value at 0.6 THz. The focused THz beam passes through the cryogenically cooled sample at 4.2 K and is detected by a liquid-helium-cooled silicon-composite bolometer. Power detection sensitivity of the bolometer is <1 nW up to 3 THz in a 400 Hz bandpass (NEP of the bolometer is 1 pW/Hz1/2). Both the amplified bolometer signal and a voltage proportional to the photomixer current are sampled using lock-in amplifiers. These signals are oversampled at an interval 5 times smaller than the 30 ms time constant of the bolometer to eliminate lineshape distortion. The Ti:Sapp frequency is measured with a wavemeter every half a second to an accuracy of 0.02 cm-1.

Crystalline Trialanine: Extreme Sensitivity to β-sheet Structure and Co-crystallized Water

High resolution THz absorption spectra (0.06 THz to 3 THz) have been obtained at 4.2 K for three crystalline forms of trialanine [5]. The crystal structures differ in their β-sheet forms (parallel vs. anti-parallel) and in their water composition (hydrated vs. dehydrated anti-parallel β-sheet). The crystal structures of p-Ala3 and ap-Ala3-H2O reported from X-ray studies [6] are shown in Fig. 3.

Figure 3
Figure 3. X-ray crystal structures of the parallel (top two panels) and anti-parallel (bottom two panels) β-sheet forms of trialanine.

The mid-infrared FTIR spectra of p-Ala3 and ap-Ala3-H2O [7] obtained at room temperature are shown as the upper and middle traces of Fig. 4, respectively. While overall the FTIR spectra are quite similar, subtle but distinguishing features sensitive to different β-sheet forms are seen in the bend and stretch regions. However, the FTIR spectrum of the dehydrated structure, ap-Ala3, shown in the lower part of Fig. 4 is indistinguishable from that obtained for the hydrated form, ap-Ala3-H2O.

Figure 4.

Figure 4. FTIR spectra of three crystalline forms of Ala3. The spectra illustrate this region is mildly sensitive to the β-sheet form and very insensitive to the water in the structure.		       Figure 5

Figure 5. THz spectra of three crystalline forms of Ala3. All three spectra are unique and illustrate the extreme sensitivity in this region to changes in the β-sheet form and the co-crystallized water weakly bound in the lattice.


The corresponding THz spectra of the three crystalline structures are shown in Fig. 5. All spectra are nearly vibrationally resolved with little absorption below 1 THz. Compared to the mid-IR region, the spectral patterns of p-Ala3 and ap-Ala3-H2O are qualitatively different illustrating the extreme sensitivity to changes in the intermolecular hydrogen bonding networks that characterize the different β-sheet forms. Moreover, in sharp contrast to the mid-infrared region, the THz spectra of the hydrated and dehydrated forms of the anti-parallel β-sheet again share nothing in common. This single result illustrates the enormous impact that a hydrogen bonded solvent like water has on the global nuclear motions probed in the THz region.

The three different forms of trialanine serve as a benchmark system to investigate the lowest frequency vibrational modes of a peptide crystal and to examine the impact the hydrogen bonding network has on these vibrational modes. Furthermore, these systems are just simple enough to treat at the full quantum level of theory using density functional theory (DFT). Predictions obtained from a classical force field (CHARMM) and DFT (PW91) models for periodic solids are shown together with the experimental THz spectrum of p-Ala3 in Fig. 6.       Figure 6.
Figure 6. THz spectra observed (bottom) and calculated using the CHARMM force field (top) and density functional theory (middle) for periodic crystal structures of p-Ala3. Notice that both levels of theory adequately predict the lowest energy region (below 1.5 THz). However, the nuclear motions associated with the predicted THz modes near 1.4 THz are very different as illustrated in Figure 7.


The vibrational modes calculated at 1.4 THz using CHARMM and DFT are shown in Fig. 7. Notice that the nuclear motions at these two levels of theory are qualitatively different and that the similarities apparent in the predicted spectra are therefore only coincidental. Similar calculations (not shown) have been performed on the ap-Ala3 and ap-Ala3-H2O forms. In general, the results for the parallel β-sheet are in better agreement with experiment than those of the anti-parallel β-sheet forms. For all three structures, however, the hydrogen bond distances are under-estimated at both levels of theory and the predicted absorption features are significantly red-shifted for ap-Ala3 and ap-Ala3-H2O. These results indicate the PW91 functional is not sufficient to treat the weak inter-sheet hydrogen bonding present in the anti-parallel β-sheet forms and strongly suggest the need for improved force field models that include three-atom hydrogen bonding terms for periodic solids. It should also be stressed that meeting these objectives represents an extreme theoretical challenge.

Figure 7b. MOVIE of Normal Mode of p-Ala3 at 1.4 THz from DFT.
Normal mode vibration predicted at approximately˜1.4 THz for the p-Ala3 β-sheet using CHARMM force field (left) and DFT/PW91 (right) functional.

Dipeptide Nanotubes

We have recently obtained THz absorption spectra [8] for a series of dipeptides known from x-ray studies to form nanotube structures [9]. Depending on the residue combination, either hydrophobic or hydrophilic core regions are formed. Examples of the crystal structures of each type are shown in Fig. 8. THz absorption spectra obtained for a series of hydrophobic structures are shown in Fig. 9. They are given in order of decreasing core size (from 5.2 Å for AV to 3.3 Å for IV) and share in most cases the strong lowest energy feature. Even more interesting is that these nanotubes may support different solvents. A number of solvated crystal structures have been reported including a few containing water. The THz absorption spectra of the hydrated and dehydrated forms of AV and its retro-analog, VA, are shown in Fig. 10.

Figure 8

Figure 8. Top view of the crystal structures of the hydrophobic dipeptide nanotube, Valine-Alanine (VA - left) and the hydrophilic nanotube, Isoleucine-Leucine (IL - right).

Figure 9

Figure 9. THz spectra obtained for a series of hydrophobic dipeptide nanotubes (A - alanine, V - valine, I - isoleucine). The spectra are ordered according to decreasing van der Waals diameters of the core regions from 5.2 Å to 3.3 Å

Figure 10

Figure 10. THz spectra of AV and its retro-analog, VA. The two spectra in each panel were obtained for dehydrated samples (blue) and for samples where water is present in the hydrophobic core region (red).

So, in contrast to the drastic spectral changes found with hydration for the anti-parallel sheet of Ala3, the THz spectra of these hydrophobic structures change very little. Only a slight red shift of 1 cm-1 occurs with hydration. It is clear that the absence of hydrogen bonds between the solvent and the lattice structure results in a mild perturbation of the THz absorption features. The calculated CHARMM spectrum of VA (without water) is shown in Fig. 11. The spectral predictions are in remarkably good agreement with the observed features (as are the predictions for a number of other nanotube structures). Notice also the nuclear motion associated with the strong low-energy feature (Fig. 12) involves significant breathing motion of core region. Theoretical studies using CHARMM and DFT are currently underway to investigate how these THz vibrational frequencies change with hydration.

Figure 11
Figure 11. THz spectra of valine-alanine (VA) calculated using the CHARMM force field (top) compared to the observed spectra of VA with (middle) and without water (bottom).


Normal Mode of VA at 1.1 THz from CHARMM

Figure 12. The predicted normal mode from the CHARMM force field associated with the lowest-energy feature. The mode involves significant breathing motion of core region.

References
  1. T.M. Korter and D.F. Plusquellic, Chem. Phy. Lett. 385 45-51 (2004).
  2. D.F. Plusquellic, T.M. Korter, G.T. Fraser, R.J. Lavrich, E.C. Benck, C.R. Bucher, J. Domench, and A.R. Hight Walker, "Continuous-Wave Terahertz Spectroscopy of Biomolecules and Plasmas," in Terahertz Sensing Technology. Volume 2: Emerging Scientific Applications and Novel Device Concepts (World Scientific, 2003, ed. by D.L. Wollard, W.R. Loerop, and M.S. Shur), 13(4), 385-404 (2003).
  3. a) K.A. McIntosh, E.R. Brown, K.B. Nichols, O.B. McMahon, W.F. DiNatale, T.M. Lyszczarz, Appl. Phys. Lett. 67, 3844 (1995);
    b) S. Verghese, K.A. McIntosh, E.R. Brown, Appl. Phys. Lett. 71, 2743, (1997);
    c) E.R. Brown, Appl. Phys. Lett. 75, 769-771 (1999);
    d) S.M. Duffy, S. Verghese, K.A. McIntosh, A. Jackson, A.C. Gossard, S. Matsuura, IEEE Trans. Microwave Theory and Tech. 49, 1032 (2001).
  4. K. Siegrist and D.F. Plusquellic, Rev. Sci. Instr. internal review (WERB).
  5. K. Siegrist, C.R. Bucher, I. Mandelbaum, A.R. Hight Walker, R. Balu, S.K. Gregurick, and D.F. Plusquellic, J. Am. Chem. Soc. (submitted)
  6. a) A. Hempel, N. Camerman, and A. Camerman, Biopolymers 31, 187 (1991).
    b) J.K. Fawcett, N. Camerman, and A. Camerman, Acta. Cryst. B 31 658-665 (1975).
  7. W. Qian, J. Bandekar, and S. Krimm, Biopolymers 31, 193 (1991).
  8. K. Siegrist and D.F. Plusquellic, in preparation.
  9. C. H. Gorbitz, Acta, Cryst. B, 58, 849 (2002) and references therein.

A Media Player is required. This software may be downloaded without charge.

For technical information or questions, call:
David F. Plusquellic
Phone: (301) 975-3896
FAX: (301) 975-2950
Email: david.plusquellic@nist.gov

OTD Home Page   |   Technical Inquiries   |   Site Comments
Online: March 2000   -   Last updated: March 2006