1 3
Top Curr Chem (Z) (2018) 376:24
configuration with interacting aromatic side chain in a stacked conformation (Fig. 9c), an
analogous trend is found. In this case, a sizeable electronic quartic coupling is expected,
and it is found equally at CAS(14,13)
δ
and RAS(2,3|8,8|2,2)
δ
levels, while the energy
shift (Δ) is significantly overestimated at the RAS(4,7|0,0|4,6)
δ
level.
Here, a successful benchmark study aimed at identifying efficient multi-configurational/multi-reference approaches in terms of the computational cost/accuracy
ratio has been demonstrated in detail for benzene and phenol chromophores involved
in protein systems. Analogous studies have been performed for indole [82] and adenine [57] chromophores, while the remaining nucleobases are currently under investigation. Overall, we believe that such theoretical benchmark studies will provide a
full set of parameters and computational recipes that will enable the construction of
excitonic model Hamiltonians for application in large systems (such as full proteins
or large DNA/RNA sequences) and simulating 2DUV spectra of dimeric and small
oligomeric systems under realistic conditions with unprecedented accuracy.
In the next sections, we will illustrate how accurate simulation of 2DUV spectroscopy could be used as a powerful tool to investigate physicochemical properties
of biological systems.
4.2 2DES for Tracking GS Conformational Dynamics
In Sect. 3.1, we illustrated how 2DUV spectroscopy holds the potential to resolve
electronic couplings associated with the interaction between UV chromophores in a
model protein system containing two aromatic side chains. In particular, the energy
shift (Δ) of mixed doubly excited states due to quartic coupling clearly affects the
2DUV–UV spectrum of the solvated CFYC tetrapeptide, enabling discrimination
between peptide conformations with interacting and non-interacting chromophores.
The CFYC peptide was chosen for modeling protein folding/unfolding dynamics
because its terminal cysteine residues can form a disulfide bond that holds the tetrapeptide in a cyclic closed conformation while, if this bond breaks, CFYC will naturally unfold in an open conformation (Fig. 10). QM/MM geometry optimization of
the closed CFYC, solvated in water solution, yields a structure in which the two
aromatic side chains are rather close to other, in a so-called T-stacked conformation
[52]. On the other hand, the open CFYC conformation is associated with unstacked
(i.e. non-interacting) chromophoric units (Fig. 10a).
As shown in Fig. 10b, the excited-state manifolds of both closed T-stacked and
open unstacked QM/MM optimized structures have been fully characterized at the
CAS(14,13)
δ
level [37, 52] showing how they are strongly differentiated by the
chromophoric electronic couplings, which have a twofold effect: (i) red-shift of the
mixed states by quartic coupling, and (ii) strong red-shift of charge-transfer (CT)
states (from Phe to Tyr and vice versa) with a concomitant increase in the TDMs
associated with the corresponding S 1-2 → CT transitions. The former would cause
off-diagonal (negative) signals to appear in the 2DUV–UV maps (see Fig. 9c), while
the latter would show up exclusively at lower energies, i.e. in 2DUV–Vis spectra. Figure 10c shows the SOS//QM/MM computed 2DUV–Vis spectra for closed
T-stacked and open unstacked conformations. In these spectra, few (positive) ESA
87
Reprinted from the journal
Top Curr Chem (Z) (2018) 376:24
configuration with interacting aromatic side chain in a stacked conformation (Fig. 9c), an
analogous trend is found. In this case, a sizeable electronic quartic coupling is expected,
and it is found equally at CAS(14,13)
δ
and RAS(2,3|8,8|2,2)
δ
levels, while the energy
shift (Δ) is significantly overestimated at the RAS(4,7|0,0|4,6)
δ
level.
Here, a successful benchmark study aimed at identifying efficient multi-configurational/multi-reference approaches in terms of the computational cost/accuracy
ratio has been demonstrated in detail for benzene and phenol chromophores involved
in protein systems. Analogous studies have been performed for indole [82] and adenine [57] chromophores, while the remaining nucleobases are currently under investigation. Overall, we believe that such theoretical benchmark studies will provide a
full set of parameters and computational recipes that will enable the construction of
excitonic model Hamiltonians for application in large systems (such as full proteins
or large DNA/RNA sequences) and simulating 2DUV spectra of dimeric and small
oligomeric systems under realistic conditions with unprecedented accuracy.
In the next sections, we will illustrate how accurate simulation of 2DUV spectroscopy could be used as a powerful tool to investigate physicochemical properties
of biological systems.
4.2 2DES for Tracking GS Conformational Dynamics
In Sect. 3.1, we illustrated how 2DUV spectroscopy holds the potential to resolve
electronic couplings associated with the interaction between UV chromophores in a
model protein system containing two aromatic side chains. In particular, the energy
shift (Δ) of mixed doubly excited states due to quartic coupling clearly affects the
2DUV–UV spectrum of the solvated CFYC tetrapeptide, enabling discrimination
between peptide conformations with interacting and non-interacting chromophores.
The CFYC peptide was chosen for modeling protein folding/unfolding dynamics
because its terminal cysteine residues can form a disulfide bond that holds the tetrapeptide in a cyclic closed conformation while, if this bond breaks, CFYC will naturally unfold in an open conformation (Fig. 10). QM/MM geometry optimization of
the closed CFYC, solvated in water solution, yields a structure in which the two
aromatic side chains are rather close to other, in a so-called T-stacked conformation
[52]. On the other hand, the open CFYC conformation is associated with unstacked
(i.e. non-interacting) chromophoric units (Fig. 10a).
As shown in Fig. 10b, the excited-state manifolds of both closed T-stacked and
open unstacked QM/MM optimized structures have been fully characterized at the
CAS(14,13)
δ
level [37, 52] showing how they are strongly differentiated by the
chromophoric electronic couplings, which have a twofold effect: (i) red-shift of the
mixed states by quartic coupling, and (ii) strong red-shift of charge-transfer (CT)
states (from Phe to Tyr and vice versa) with a concomitant increase in the TDMs
associated with the corresponding S 1-2 → CT transitions. The former would cause
off-diagonal (negative) signals to appear in the 2DUV–UV maps (see Fig. 9c), while
the latter would show up exclusively at lower energies, i.e. in 2DUV–Vis spectra. Figure 10c shows the SOS//QM/MM computed 2DUV–Vis spectra for closed
T-stacked and open unstacked conformations. In these spectra, few (positive) ESA
87
Reprinted from the journal