[論文レビュー] Computational Modeling of Exciton-bath Hamiltonians for LH2 and LH3 Complexes of Purple Photosynthetic Bacteria at Room Temperature

Daniel Villanueva Montemayor, Eva Rivera+1 more|arXiv (Cornell University)|Jan 2, 2026

Photosynthetic Processes and Mechanisms被引用数 0

ひとこと要約

The paper combines all-atom MD simulations with TD-DFT and exciton-bath modeling to explain spectral differences between LH2 and LH3, and to construct a common exciton-bath Hamiltonian for these complexes.

ABSTRACT

Light harvesting 2 (LH2) complex is a primary component of the photosynthetic unit of purple bacteria that is responsible for harvesting and relaying excitons. The electronic absorption line shape of LH2 contains two major bands at 800 nm and 850 nm wavelength regions. Under low light condition, some species of purple bacteria replace LH2 with LH3, a variant form with almost the same structure as the former but with distinctively different spectral features. The major difference between the absorption line shapes of LH2 and LH3 is the shift of the 850 nm band of the former to a new 820 nm region. The microscopic origin of this difference has been subject to some theoretical/computational investigations. However, the genuine molecular level source of such difference is not clearly understood yet. This work reports a comprehensive computational study of LH2 and LH3 complexes so as to clarify different molecular level features of LH2 and LH3 complexes and to construct simple exciton-bath models with a common form. All-atomistic molecular dynamics (MD) simulations of both LH2 and LH3 complexes provide detailed molecular level structural differences of BChls in the two complexes, in particular, in their patterns of hydrogen bonding (HB) and torsional angles of the acetyl group. Time-dependent density functional theory calculation of the excitation energies of BChls for structures sampled from the MD simulations, suggests that the observed differences in HB and torsional angles cannot fully account for the experimentally observed spectral shift of LH3. Potential sources that can explain the actual spectral shift of LH3 are discussed, and their magnitudes are assessed through fitting of experimental line shapes.

研究の動機と目的

• Clarify molecular-level factors differentiating LH2 and LH3 spectroscopic properties.
• Develop simple, common exciton-bath models usable for LH2 and LH3.
• Assess how hydrogen bonding and acetyl group orientation affect Qy excitation energies.
• Provide MD-derived geometric and coupling parameters to inform exciton dynamics simulations.

提案手法

• All-atom MD simulations of LH2 and LH3 in membrane environments to obtain structural parameters and dynamics.
• TD-DFT calculations of Qy excitation energies for BChls with proximate protein residues from MD snapshots.
• Transition charge approximation (TrEsp) to compute electronic couplings J_ss'(n-m) from MD data with a dielectric screening factor f=0.55.
• Coarse-grained exciton-bath Hamiltonian H_T including H_e^0, disorder H_e, H_b, and H_eb.
• Classical energy-gap correlation function C_E,sn^cl(t) from MD to obtain bath spectral densities J_s(n)(ω) via Fourier transform.
• Discussion of sources beyond HB effects (e.g., acetyl group rotation) that could explain LH3 spectral shifts.

実験結果

リサーチクエスチョン

• RQ1 What molecular features differentiate LH2 and LH3 and how do they affect their absorption spectra?
• RQ2 Can a common exciton-bath Hamiltonian adequately describe LH2 and LH3 dynamics using MD-informed parameters?
• RQ3 To what extent do hydrogen bonding, acetyl group orientation, and local environment contribute to the B800/B850 (B820) spectral shifts?
• RQ4 How do MD-derived geometric parameters and electronic couplings influence the ensemble line shapes of LH2 and LH3?

主な発見

Parameter	LH2 (cm^-1)	LH3 (cm^-1)	rel. diff. %
J_{αβ}(0)	244.74 ± 4.97	238.74 ± 5.11	2.45
J_{αβ}(1)	139.75 ± 10.22	74.13 ± 10.69	46.96
J_{αα}(1)	-59.07 ± 1.01	-61.21 ± 1.01	3.62
J_{ββ}(1)	-29.27 ± 1.61	-21.54 ± 1.70	26.41
J_{βα}(1)	13.83 ± 0.17	13.52 ± 0.18	2.24
J_{αβ}(2)	12.66 ± 0.35	10.83 ± 0.40	14.45
J_{γγ}(1)	-24.46 ± 0.77	-23.45 ± 0.75	4.13
J_{αγ}(1)	28.29 ± 0.59	26.58 ± 0.52	6.04
J_{βγ}(1)	2.63 ± 0.61	3.31 ± 0.54	25.86
J_{γα}(0)	-12.31 ± 0.42	-12.17 ± 0.37	1.14
J_{γβ}(0)	-2.68 ± 0.83	0.88 ± 0.77	67.16
λ_{α}	23.9	27.1	13.39
λ_{β}	28.4	26.4	7.04
λ_{γ}	73.3	72.4	1.23

• MD sampling shows LH2 and LH3 have nearly identical tertiary structures and pigment arrangements, with small differences in azimuthal orientation and some angles.
• Nearest-neighbor couplings between α- and β-BChls from adjacent protomers (J_αβ(1)) drop by about half in LH3 relative to LH2.
• Electronic couplings averaged over MD trajectories are generally smaller than fixed-crystal-structure estimates, consistent with recent higher-level calculations.
• Spectral densities J_s(n)(ω) show γ-BChl (B800) has larger bath coupling and more low-frequency content than α/β-BChls (B850/B820), due to more hydrophilic surroundings.
• A common exciton-bath model structure is feasible for LH2 and LH3, with disorder in site energies modeled as Gaussian, and off-diagonal couplings treated via the TrEsp method.
• Disorder in site energies is necessary to reproduce experimental ensemble line shapes; the standard deviations are suggested to be on the order of ~200 cm^-1 for B850/B820 units.