NOTE! This site uses cookies and similar technologies.

If you not change browser settings, you agree to it. Learn more

I understand

Bio-molecular systems

Models development

Minimalist models for biopolymers

MusCaDe implements the minimalist model for biopolymers[1] [2] [3] representing polypeptides and nucleic acids as un-branched chains, with the interacting center ("bead") placed on C-alpha or on P, as shown in Fig 1, and with implicit solvent. These models are the simplest which can explicitly represent secondary strucutures [4] (hence "minimalist), and can therefore be considered paradigmatic among the CG ones [5], besides providing very favorable cost to benefit ratio.
Our efforts are currently devoted to optimizing these models parameterization, according to two criteria: (i) using analytical physics-chemistry based FF terms, (ii) including information at different levels (bottom-up and top-down) and from different sources (experimental or other levels simulations). (i) helps rationalizing the parameterization and reaching a possible standard for the FF; (ii) helps combining accuracy and predictive power into the model. MusCaDe provides tools for the statistics based parameterization[6] (SecStAnT), including different algorithms for the optimisation, such as the Iterative Boltzmann Inversion and the MonteCarlo parameters space exploration (AsParaGs). This approach has already produced minimalist models for helical polypeptides [7], DNA and RNA [8]; previous version of the model including a partial structural bias were applied to a number of biomolecular systems of different sizes, ranging from single proteins to ribosomes (see here for further information).
Fig 1: Minimalist models for polypeptides (left) and nucleic acids (right).

Meso-Scale model for the cytoplasm

Fig 2: (a) Meso scale model for the cytoplasm (b) Sphere-of-bead parameterization of the crowders FF (c) minimalist model for the Green Fluorescent Proteins and (d) its "parallel" multi-scale combination with the cytoplasm model.
We recently developed a meso-scale model for the cytoplasm representing crowder molecules as soft spheres of different sizes and different diffusion coefficients [9] (Fig 2,a). The FF is developed according to similar criteria as to the minimalist model: bottom-up and top-down information are included. In particular, the inter-crowders potential is developed by means of a sphere-of-beads approach (Fig 2,b), so that the meso-scale model is automatically compatible with the mimalist models for biopolymers, which allows using it both in serial and in parallel multi-scaling (Fig 2,b,c) [10]. A user-interface for the parameterization and implementation of the cytoplasm model is under development.

Multi-scale applications

Fluorescent proteins and Rhodopsins

The above mentioned mentioned cytoplasm model was used with the minimalist model of the Green Fluorescent Protein (GFP, Fig 2,c) in parallel combination, as previously done to study the effect of crowding on the dynamics of HIV-1 protease (see here for further information). However GFP, due to its relative simplicity and to the large amount of experimental data, was considered as a test case for the whole range of multi-scale methods in the last decade (see here and here for further information). In particular, the tuning of GFP optical properties is a challenging problem for hybrid QM/MM methods, and we analyzed the performance of TD-DFT, CASPT2 and SAC-CI in combination with non-polarizable Amber force field [11]. The role of the electrostatic field in GFP tuning was investigated in a mixed experimental-computational study (in preparation), where the average electrostatic potential felt by the chromophore was calculated by all-atom MD simulations. More recent multi-scale applications are to rhodopsins, for which we developed a hybrid atomistic-CG model to represent the whole photocycle (in the case of bacterial rhodopsin) or the main steps of it (for bovine rhodopsin) without using massive computing resources [12].

Membrane-active Peptides

Short amino-acid sequences interacting with the lipid bilayer are being actively studied for application such as new antimicrobial agents and drug-delivery vectors. We employ a top-down approach encompassing various scales and computational methods to design new membrane-active sequences and study peptide-bilayer interactions. At the coarser scale, we developed statistical-based models describing and predicting peptide antimicrobial activity, and successfully designed novel highly effective antimicrobial peptides (AMPs) also including non-natural amino acids [13]. Prediction of peptide structure at the atomic level is performed by extensive all-atom MD simulations, comparing different force fields in order to assess the robustness of the results, and with enhanced configuration sampling by using the Hamiltonian Replica Exchange (HREX) method. Interaction of peptides and other organic molecules with the lipid bilayer is investigated by both all-atom and coarse-grained (Martini force field) MD simulations. Metadynamics and Umbrella Sampling techniques provide free energy scans of coordinates describing the position/orientation of the molecule with respect to the bilayer.
Fig 3:Distribution of solvent molecules surrounding an amphipatic alpha-helix AMP, GMG_01. The figure displays the isodensity surface of water (red) and TFE molecules (green) around the AMP during the molecular dynamics simulation, clearly revealing the amphipatic nature of the peptide.

Nucleic acid aptamers for intracellular delivery

Thanks to their ability to recognize biomolecular targets with high affinity and specificity, nucleic acid aptamers are increasingly investigated as diagnostic and therapeutic tools, particularly when their targets are cell-surface receptors. We investigate the relationship between the folding of an anti-mouse transferrin receptor DNA aptamer and its interaction with the transferrin receptor both in vitro and in living cells. We identified and purified two aptamer conformers by means of chromatographic techniques [14]. Fluorescence-anisotropy measurements showed that only one fold is able to bind mouse transferrin receptor. Besides displaying enhanced endocytosis in living mouse fibroblasts, the purified active fold is internalized also in human pancreatic cancer cells. Starting from these observations, we rationally designed variations of the parent sequence aimed at stabilizing the active fold, and consequently increase aptamer activity.

References

  1. Valentina Tozzini - Minimalist models for proteins: a comparative analysis Quarterly Reviews of Biophysics 03:333-371 (2010)
  2. Fabio Trovato, Valentina Tozzini - Supercoiling and Local Denaturation of Plasmids with a Minimalist DNA Model The Journal of Physical Chemistry B 42:13197-13200 (2008)
  3. Fabio Trovato, Valentina Tozzini - Minimalist models for biopolymers: Open problems, latest advances and perspectives AIP Conference Proceedings :187-200 (2012)
  4. Valentina Tozzini, Walter Rocchia, J. Andrew McCammon - Mapping All-Atom Models onto One-Bead Coarse-Grained Models:  General Properties and Applications to a Minimal Polypeptide Model Journal of Chemical Theory and Computation 3:667-673 (2006)
  5. Valentina Tozzini - Coarse-grained models for proteins Current Opinion in Structural Biology 2:144-150 (2005)
  6. Giuseppe Maccari, Giulia L. B. Spampinato, Valentina Tozzini - SecStAnT: secondary structure analysis tool for data selection, statistics and models building Bioinformatics 5:668-674 (2014)
  7. Giulia L. B. Spampinato, Giuseppe Maccari, Valentina Tozzini - Minimalist Model for the Dynamics of Helical Polypeptides: A Statistic-Based Parametrization Journal of Chemical Theory and Computation 9:3885-3895 (2014)
  8. Filip Leonarski, Fabio Trovato, Valentina Tozzini, Andrzej Leś, Joanna Trylska - Evolutionary Algorithm in the Optimization of a Coarse-Grained Force Field Journal of Chemical Theory and Computation 11:4874-4889 (2013)
  9. Fabio Trovato, Valentina Tozzini - Diffusion within the Cytoplasm: A Mesoscale Model of Interacting Macromolecules Biophysical Journal 11:2579-2591 (2014)
  10. Fabio Trovato, Riccardo Nifosì, Armida Di Fenza, Valentina Tozzini - A Minimalist Model of Protein Diffusion and Interactions: The Green Fluorescent Protein within the Cytoplasm Macromolecules 20:8311-8322 (2013)
  11. Pietro Amat, Riccardo Nifosì - Spectral “Fine” Tuning in Fluorescent Proteins: The Case of the GFP-Like Chromophore in the Anionic Protonation State Journal of Chemical Theory and Computation 1:497-508 (2013)
  12. Francesco Tavanti, Valentina Tozzini - A Multi-Scale–Multi-Stable Model for the Rhodopsin Photocycle Molecules 9:14961-14978 (2014)
  13. Giuseppe Maccari, Mariagrazia Di Luca, Riccardo Nifosí, Francesco Cardarelli, Giovanni Signore, Claudia Boccardi, Angelo Bifone - Antimicrobial Peptides Design by Evolutionary Multiobjective Optimization PLoS Computational Biology 9:e1003212- (2013)
  14. David Porciani, Giovanni Signore, Laura Marchetti, Paolo Mereghetti, Riccardo Nifosì, Fabio Beltram - Two Interconvertible Folds Modulate the Activity of a DNA Aptamer Against Transferrin Receptor Molecular Therapy—Nucleic Acids 1:e144- (2014)