A molecular dynamics simulation of two neuraminidase-sialic acid (NA-SA) complexes show a difference of the level of stability between sialic acid and neuraminidases that originated from viruses A/Tokyo/3/67 (Structure A) dan A/Pennsylvania/10218/84 (Structure B). Analyses of sialic acid RMSD and the change of torsional angles suggest that the sialic acid in Structure A is much more twisted and able to be influenced more by the binding of the neuraminidase functional residues than Structure B. Moreover, analyses upon hydrogen bond occupancy and binding free energy of both complexes showed that Structure A had more stable hydrogen bonds and each complex’s binding free energy were calculated to be –1.37 kcal/mol and 17.97 kcal/mol for Structure A and Structure B, respectively, further suggesting stability more dominant in Structure A than Structure B. Overall, Structure A has a more stable enzyme-substrate than Structure B.
There are two pathogenicity levels of the avian influenza virus: the High Pathogenic Avian Influenza (HPAI) virus and the Low Pathogenic Avian Influenza (LPAI) virus. These levels of virus pathogenicity are classified based on the Intravenous Pathogenicity Index (IVPI) of a six-week-old chicken. An influenza virus is said to be highly pathogenic when it is able to kill >75% of chickens ages 4 - 6 weeks in a ten-day window post-inoculation and features an IVPI larger than 1.2. The viruses that do not meet the HPAI criteria are called LPAI viruses [1].
Based on a few in vitro and in vivo studies [2-8], the enzymatic activity of neuraminidase (NA) participates in defining the pathogenicity level of avian influenza virus. Neuraminidase plays a major role in viral replication. During virus budding, neuraminidase cleaves the newly mature virion from its host cell. After the virion is detached from its host cell, it is then able to infect other host cells. In the previous studies, it was found that the neuraminidase of the HPAI viruses have the ability to cleave the sialic acid (SA) more effectively than the LPAI viruses.
Research efforts in neuraminidase inhibition throughout the years have resulted in three commercially available drugs (Oseltamivir, Laninamivir and Peramivir). However, at the rate in which the neuraminidase enzyme is mutating, these drugs could lose their potency over time. In other words, those three inhibitors are not stable in the confines of the neuraminidase binding pockets and may lead to the virus becoming resistant to the inhibitors.
Intricate studies in this area (NA-SA interaction) are very important, keeping in mind that the main principle of inhibition is to have a certain residue bind with the neuraminidase more than with the natural substrate, sialic acid. This study aims to shed light on neuraminidase so that structure-based drug design could solve the observed problem of neuraminidase resistance to drug molecules such as Oseltamivir, Zanamivir, Laninamivir and Peramivir in the future.
The computational approach of defining the correlation between pathogenicity and the neuraminidase’s ability to bind to a substrate is still on the rise, and because of neuraminidase’s rate of mutation, the knowledge in this field will complement experimental approaches to gain a more specific design to neutralize the neuraminidase activity more completely. While the standard intravenous pathogenicity index test can produce results in 10 days [9], a faster identification method through NASBA only focuses on hemagglutinin [10]. There needs to be an alternative to these experimental tests that integrates all of the different methods and approaches in order to determine the viral pathogenicity in an accurate and timely manner. Computational approaches are valuable resources in this process, and this study serves as a starting point on how pathogenicity could be viewed on a molecular scale.
In this study, the structure of the A/Tokyo/3/67 avian influenza virus that was isolated in 1967 [11], during a time of prevalent infection, will be compared to the structure of the A/Pennsylvania/10218/84 avian influenza virus that is a non-pathogenic avian influenza virus [12]. These avian influenza viruses were chosen to delegated two different level of pathogenicity. The relationship between the pathogenicity level of an avian influenza virus and the stability of the neuraminidase functional residues upon binding to sialic acid will be examined. Additionally, the natural substrate sialic acid will be examined in detail to observe its response to neuraminidase during molecular dynamics simulations.
2. Materials and Methods2.1. Structure Preparation2.1.1. Sequence Alignment and Template Searching
The amino acid sequences of both neuraminidases were obtained from the influenza database in NCBI [13] with accession code AAB05621 for the A/Tokyo/3/67 virus and BAF48360 for A/Pennsylvania/10218/84, which will be denoted as Structures A and B, respectively. A sequence alignment was executed using the fast pairwise alignment method that uses the BLOSUM 30 scoring matrix.
The x-ray crystallography structure of the neuraminidase-sialic acid (NA-SA) complex for Structure A [14] was obtained from the RCSB protein data bank [15] with accession code 2BAT. Furthermore, the crystal structure for Structure B was obtained by homology modeling as explained in the next section.
2.1.2. Refinement, Homology Modeling and Explicit Solvation Process
The generation of Structure B was done by homology modeling. Homology modeling was initiated by alignment of the amino acid sequence of the target structure sequence with the template. For the template itself, the 2BAT structure was chosen. In the 2BAT model, there is a sialic acid ligand in the structure. During modeling, the sialic acid was considered to be a rigid structure.
The template was refined by removing unwanted water and calcium molecules, and the remaining NA-SA structure was then used to generate Structure B. The modeling of Structure B was completed by characterizing the structure with the charmm forcefield, which in turn added the missing hydrogen atoms to the structure. Following structure generation, both Structure A and Structure B were solvated in a TIP3P water box.
2.2. Minimization and Molecular Modeling
Both complexes were subjected to two steps of energy minimization. The first step was executed by the Steepest Descent algorithm with a targeted energy gradient of 0.5 kcal/mol and a 1,000,000-step maximum. The second step was executed by the conjugate gradient with a targeted energy gradient of 0.1 kcal/mol and a 1,000,000- step maximum. For both complexes, a nonbond list radius of 14 Å was used, and a switching function was applied between 10 - 12 Å for computational efficiency. To gain a long-range electrostatic energy contribution, it was visualized in a spherical cutoff mode.
The molecular dynamics simulation was executed with heating during the first 20 ps of the simulation; the temperature rose from 0 - 300 K. The parameters used were 20,000 steps, 0.001 time step, 0 K initial temperature, 300 K target temperature, nonbond list radius identical to that in energy minimization, and the trajectory data were stored every 0.1 ps.
2.3. End-Point Energy Calculation
Calculation of the ligand-receptor complex was based on the equation below:
is the average Gibbs energy, is the electrostatic and nonpolar free energy from implicit solvation. In this particular study, Generalized Born with Molecular Volume (GBMV) was applied for implicit solvation. The last term (TS) is the temperature and entropy contribution, while the first term of the right hand side () is the energy term produced by the applied forcefield, which is the potential energy of the system [16]:
(2)
The relationship between the ligand, receptor and complex energy is given in the next equation:
Equation (1) is average Gibbs energy which constructed each component energy in Equation (3). Furthermore, all phases of the study described in this section from structure preparation to molecular dynamics simulation were conducted with Discovery Studio 2.1 (Accelrys).
3. Results
A sequence alignment indicated that the two neuraminidases had 91% amino acid similarity. This similarity allowed us to generate a homology model for Structure B based on the 2BAT template structure, since it was higher than the required minimum of 50% similarity [17]. In addition, energy minimization of the solvated structures resulted in a decrease in energy to –711, 107.21 and –599,227.56 kcal/mol for Structure A and Structure B, respectively.
Heating simulation was executed for the first 20 ps (20,000 steps) of the molecular dynamics simulation (from 0 - 300 K for each complex) to raise the system temperature to room temperature. The heating process is illustrated in Figure 1(a) for both the high pathogenic and low pathogenic complexes. It can be seen that the increase in temperature from 0 - 300 K occurs between 0 and 2.5 ps, while the rest of the molecular dynamics simulation continued in equilibrium at a 300 K average temperature. The increment of system temperature in 2.5 ps designed to computational efficiency. For both systems, the increase in temperature and the subsequent thermal stability of the system appeared to be similar. This can be seen by the overlapping curvatures in Figure 1(a). This could suggest that the interactions of both complexes are along similar energetic pathways.
To evaluate the stability of the systems during simulation, the root mean square deviation (RMSD) of the backbone and all atoms of the complexes were calculated for all conformations throughout the simulation. As a whole, the structural stability of the systems was wellmaintained and is illustrated in Figure 1(b). In that diagram, the RMSD of all atoms was below 1 Å and had very little fluctuation. The RMSDs for both systems were 0.89 Å and 0.93 Å for Structure A and Structure B, respectively. Furthermore, the movement and change in stability of the neuraminidase molecules were not that significant. This is indicated by the backbone RMSDs that were well below 0.6 Å (0.52 Å for Structure A and 0.51 Å for structure B). The RMSDs of all neuraminidase atoms suggest that both systems behaved in a similar manner.
A comparison of the structures of both backbones of neuraminidase and its substrate (sialic acid) at the end of the simulation is shown in Figure 2. The superimposed
structures of the complexes, including the neuraminidase (shown as a ribbon) and the sialic acid (shown as a tubular shape), suggest that they do not differ from each other significantly at the end of the molecular dynamics simulation. This does not mean that the non-bonding interactions in the complexes are the same. Therefore, the observation of the interactions between the substrate and the functional residues are necessary to determine the cause of the pathogenicity of avian influenza virus neuraminidase.
Calculation of the binding free energy at the end of the simulation with added implicit solvation using the Generalized Born with Molecular Volume (GBMV) method, resulting in the values of –1.37 kcal/mol for Structure A and 17.97 kcal/mol for Structure B. The resulting values indicate that ligand binding is more favorable in Structure A than in Structure B [18,19].
3.1. Overall Substrate Stability
Figure 3 depicts the overall movement of the sialic acid in response to the hydrogen bonds that form between the sialic acid and neuraminidase and the long range interactions in the 14 Å spherical cutoff range. As shown in Figure 3, the substrate RMSD drastically increased from the starting point to the 6000th step until it reached the 0.65 Å mark. This could be caused by the increase in kinetic energy of the atoms with the rise in temperature. After the heating phase, both substrates were observed to be relatively stable until the 9000th step. Between the 9000th and 16,000th step, the RMSD of Structure A’s sialic acid decreased to 0.45 Å. This suggests that there is higher electrostatic interaction towards the initial position of sialic acid in Structure A than Structure B. The following steps of the curvature showed that the RMSD
ReferencesWorld Organization for Animal Health, “Avian Influenza: 2.7.12, Terrestrial Animal Health Code,” World Organi- zation for Animal Health, France, 2007.L. J. Mitnaul, M. N. Matrosovich, M. R. Castrucci, A. B. Tuzikov, N. V. Bovin, D. Kobasa and Y. Kawaoka, “Ba- lanced Hemagglutinin and Neuraminidase Activities Are Critical for Efficient Replication of Influenza A Virus,” Journal of Virology, Vol. 74, No. 13, 2000, pp. 6015- 6020. doi:10.1128/JVI.74.13.6015-6020.2000S. Tarigan, R. Indriani and Darminto, “Karakterisasi Aktivitas Enzimatik Neuraminidase Virus Influenza H5N1,” JITV, Vol. 12, No. 2, 2007, pp. 153-159.C. W. Lee, D. L. Suarez, T. M. Tum-pey, H. W. Sung, Y. K. Kwon, Y. J. Lee, S. J. Choi-Joh, M. C. Kim, E. K. Lee, J. M. Park, X. Lu, J. M. Katz, E. Spackman, D. E. Swayne and J. H. Kim, “Characterization of Highly Pa- tho-genic H5N1 Avian Influenza A Viruses Isolated from South Korea,” Journal of Virology, Vol. 79, No. 6, 2005, pp. 3692-3702. doi:10.1128/JVI.79.6.3692-3702.2005V. T. Peltola, K. G. Murti and J. A. McCullers, “The Influenza Virus Neuraminidase Contributes to Secondary Bacterial Pneumo-nia,” Journal of Infectious Diseases, Vol. 192, No. 2, 2005, pp. 249-257. doi:10.1086/430954T. Suzuki, T. Takahashi, C. T. Guo, K. L. P. J. Hidari, D. Miyamoto, H. Goto, Y. Kawaoka and Y. Suzuki, “Sialidase Activity of Influenza A Virus in an Endocytic Pathway Enhances Viral Replication,” Journal of Virology, Vol. 9, No. 18, 2005, pp. 11705-11715.
doi:10.1128/JVI.79.18.11705-11715.2005T. M. Tumpey, D. L. Suarez, L. E. L. Perkins, D. A. Senne, J. Lee, Y. J. Lee, I. P. Mo, H.W. Sung and D. E. Swayne, “Characterization of a Highly Pathogenic H5N1 Avian Influenza A Virus Isolated from Duck Meat,” Journal of Virology, Vol. 76, No. 12, 2002, pp. 6344- 6355. doi:10.1128/JVI.76.12.6344-6355.2002E. Spackman, D. E. Swayne, D. L. Suarez, D. A. Senne, J. C. Pedersen, M. L. Killian, J. Pasick, K. Handel, S. P. S. Pillai, C. W. Lee, D. Stallknecht, R. Slemons, H. S. Ip and T. Deliberto, “Characterization of Low-Pathogenicity H5N1 Avian Influenza Viruses from North America,” Journal of Virology, Vol. 81, No. 21, 2007, pp. 11612- 11619. doi:10.1128/JVI.01368-07Anon, “Highly Pathogenic Avian Influenza (Fowl plague),” In: OIE manual of standards for diagnostic tests and vaccines,” 2nd ed. OIE, Paris, France, A15, 1992, pp. 123-129.R. A. Collins, L. S. Ko, K. L. So, T. L. Ellis, T. Lau, Y. A. C. Hoi, “A NASBA Method to Detect High- and Low- Pathogenicity H5 Avian Influenza Viruses,” JSTOR- Avian Diseases, Vol. 47, pp. 1069-1074.M. S. Pereira, P. Chakraverty and A. R. Pane, “The Influence of An-tigenic Variation on Influenza A2 Epidemics,” Journal of Hygiene Epidemiology Microbiology and Immunology, Vol. 67, 1969, pp. 551-557.N. V. Kaverin, I. A. Rudneva, N. A. Ilyushina, N. L. Varich, A. S. Lipatov, Y. A. Smirnov, E. A. Govorkova, A. K. Gitelman, D. K. Lvov and R. G. Webster, “Structure of Antigenic Sites on the Haemagglutinin Molecule of H5 Avian Influenza Virus and Phenotypic Variation of Es-cape Mutants,” Journal of General Virology, Vol. 83, 2002, pp. 2497-2505.Influenza Virus Sequence Database.
http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/nph-select.cgi?go=databaseJ. N. Varghese, J. L. Mc-Kimm-Breschkin, J. B. Caldwell, A. A. Kortt and P. M. Colman, “The Structure of the Complex between Influenza Virus Neuraminidase and Sialic Acid, the Viral Receptor,” Proteins, Vol. 14, No. 3, 1992, pp. 327-332. doi:10.1002/prot.340140302A Resource for Studying Biological Macromolecules.
http://www.rcsb.org/pdb/home/home.doA. R. Leach, “Molecular Modelling: Principles and Applications,” 2nd Edi-tion, Prentice Hall, Upper Saddle River, 2001, pp. 165-181.E. Martz, “Homology Modeling for Beginners with Free Software,” June 2001.
http://www.umass.edu/molvis/workshop/homolmod.htmX. L. Guo, D. Q. Wei, Y. S. Zhul and K. C. Chou, “Cleavage Mechanism of the H5N1 Hemagglutinin by Trypsin and Furin,” Amino Acids, Vol. 35, No. 2, 2008, pp. 375-382.P. Decha, T. Rungrotmongkol, P. Intharathep, M. Malaisree, O. Aruksa-kunwong, C. Laohpongspaisan, V. Para- suk, P. Sompornpisut, S. Pianwanit, S. Kokpol and S. Hannongbua, “Source of High Pathogenicity of an Avian Influenza Virus H5N1: Why H5 Is Better Cleaved by Furin,” Biophysical Journal, Vol. 95, No. 1, 2008, pp. 128-134. doi:10.1529/biophysj.107.127456V. Stoll, K. D. Stewart, C. J. Maring, S. Muchmore, V. Giranda, Y.-G. Y. Gu, G. Wang, Y. Chen, M. Sun, C. Zhao, A.L. Ken-nedy, D.L. Madigan, Y. Xu, A. Saldivar, W. Kati, G. Laver, T. Sowin, H.L. Sham, J. Greer, D. Kempf, “Influenza Neuramini-dase Inhibitors: Structure- Based Design of a Novel Inhibitor Series,” Biochemistry, Vol. 42, No. 3, 2003, pp. 718-727.
doi:10.1021/bi0205449O. Aruksakunwong, M. Malisree, P. Decha, P. Som- pornpisut, V. Parasuk, S. Pianwanit and S. Hannongbua, “On the Lower Susceptibility of Oseltamivir to Influenza Neuraminidase Subtype than Those in N2 and N9,” Biophysical Journal, Vol. 92, No. 3, 2007, pp. 798-807.
doi:10.1529/biophysj.106.092528M. Shu, Z. Lin, Y. Zhang, Y. Wu, H. Mei and Y. Jiang, “Molecular Dynamics Simulation of Oseltamivir Resistance in Neuraminidase of Avian Influennza H5N1 Virus,” J Mol Model, 2010.N. R. Taylor, M. von Itzstein. “Molecular Modeling Studies on Ligand Binding to Sialidase from Influenza Virus and the Mechanism of Catalysis,” Journal of Medicinal Chemistry, Vol. 7, No. 5, 2004, pp. 616-624.J. Gong, W. Xu and J. Zhang, “Structure and Functions of Influenza Virus Neuraminidase,” Current Medicinal Che- mistry, Vol. 14, No. 1, 2007, pp. 113-122.
doi:10.2174/092986707779313444R. J. Russell, L. F. Haire, D. J. Stevens, P. J. Collins, Y. P. Lin, G. M. Blackburn, A. J. Hay, S. J. Gamblin and J. J. Skehel, “The Structure of H5N1 Avian Influenza Neuraminidase Suggests New Opportu-nities for Drug Design,” Nature, Vol. 443, No. 2006, pp. 45-49.
doi:10.1038/nature05114R. Chachra, R.C. Rizzo, “Ori-gins of Resistance Conferred by the R292K Neuraminidase Mutation via Molecular Dynamics and Free Energy Calcula-tions,” Journal of Chemical Theory and Computation, Vol. 4, No. 9, 2008, pp. 1526-1540. doi:10.1021/ct800068vJ. L. McKimm-Breschkin, A. Sahasrabudhe, T. J. Blick, M. McDonald, P. M. Colman, G. J. Hart, R. C. Bethell and J. N. Varghese, “Mutations in a Conserved Residue in the Influenza Virus Neuraminidase Active Site Decreases Sensitivity to Neu5Ac2en Derivatives,” The Journal of Virology, Vol. 72, No. 3, 1998, pp. 2456-2462.L. Le, E. Lee, K. Schulten and T. N. Truong, “Molecular Modeling of Swine Influenza A/H1N1, Spanish H1N1 and Avian H5N1 Flu N1 Neuraminidase Bound to Tamiflu and Relenza,” PloS Current Beta, Vol. 1, 2009, pp. 1015-1028. V. P. Mishin, F. G. Hayden and L. V. Gubareva, “Susceptibilities of Antiviral-Resistant Influenza Viruses to Novel Neuraminidase Inhibitors,” Antimicrobial Agents and Chemotherapy, Vol. 49, No. 11, 2005, pp. 4515-4520. doi:10.1128/AAC.49.11.4515-4520.2005T. G. Sheu, V. M. Deyde, M. Okomo-Adhiambo, R. J. Garten, X. Xu, R. A. Bright, E. N. Butler, T. R. Wallis, A. I. Klimov and L. V. Gubareva, “Surveillance for Neuraminidase Inhibitor Resis-tance among Human Influenza A and B Viruses Circulating Worldwide from 2004 to 2008,” Antimicrobial Agents and Chemotherapy, Vol. 52, No. 9, 2008, pp. 3284-3292.N. T. Wetherall, T. Trivedi, J. Zeller, C. Hodges-Savola, J. L. McKimm-Breschkin, M. Zambon and F. G. Hayden, “Evalua-tion of Neuraminidase Enzyme Assays Using Different Sub-strates to Measure Susceptibility of Influenza Clinical Isolates to Neuraminidase Inhibitors: Report of the Neuraminidase In-hibitor Susceptibility Network,” Journal of Clinical Microbi-ology, Vol. 41, No. 2, 2003, pp. 742-750. doi:10.1128/JCM.41.2.742-750.2003J. L. McKimm-Breschkin, T. Trivedi, A. Hampson, A. Hay, A. Klimov, M. Tashiro, F. G. Hayden and M. Zambon, “Neura-minidase Sequence Analysis and Susceptibilities of Influenza Virus Clinical Isolates to Zanamivir and Oseltamivir,” Antim-icrobial Agents and Chemotherapy, Vol. 47, No. 7, 2003, pp. 2264-2272.
doi:10.1128/AAC.47.7.2264-2272.2003H. Yen, N. A. Ilyushina, R. Salomon, E. Hoffmann, R. G. Webster and E. A. Govorkova, “Neuraminidase Inhibitor-Resistant Recombinant A/Vietnam/1203/04 (H5N1) Influenza Viruses Retain Their Replication Efficiency and Pathogenicity in Vitro and in Vivo,” Journal of Virology, Vol. 81, No. 22, 2007, pp. 12418-12426.
doi:10.1128/JVI.01067-07A. Meijer, A. Lackenby, O. Hungnes, B. Lina, S. van der Werf, B. Schweiger, M. Opp, J. Paget, J. van de Kassteele, J. Hay and M. Zambon, “Osel-tamivir-Resistant Influenza Virus A (H1N1), Europe, 2007-08 Season,” Emerging Infectious Diseases, Vol. 15, No. 11, 2009, pp. 552-560. doi:10.3201/eid1504.081280A. S. Monto, J. L. McKimm-Breschkin, C. Macken, A. W. Hampson, A. Hay, A. Klimov, M. Tashiro, R. G. Webster, M. Aymard, F. G. Hayden and M. Zambon, “Detection of Influenza Viruses Re-sistant to Neuraminidase Inhibitors in Global Surveillance dur-ing the First 3 Years of Their Use,” Antimicrobial Agents and Chemotherapy, Vol. 50, No. 7, 2006, pp. 2395-2402.
doi:10.1128/AAC.01339-05D. Tamura, K. Mitamura, M. Yamazaki, M. Fujino, M. Nirasawa, K. Kimura, M. Kiso, H. Shimizu, C. Kawakami, S. Hiroi, S. Takahashi, M. Hata, H. Minagawa, Y. Kimura, S. Kaneda, S. Sugita, T. Horimoto, N. Sugaya and Y. Kawaoka, “Oseltamivir-Resistant Influenza A Viruses Circulating in Japan,” Journal of Clinical Microbiol-ogy, Vol. 47, No. 5, 2009, pp. 1424-1427.
doi:10.1128/JCM.02396-08N. A. Roberts, “Treatment of Influenza with Neuraminidase Inhibitors: Virological Implica-tions,” The Royal Society, Vol. 356, No. 1416, 2001, pp. 1895-1897.
doi:10.1021/ja8085643 R. E. Amaro, X. Cheng, I. Ivanov, D. Xu and J. A. McCammon, “Characterizing Loop Dynamics and Li- gand Recognition in Human- and Avian-Type Influ-enza Neuraminidases via Generalized Born Molecular Dynam-ics and End-Point Free Energy Calculations,” Journal of the American Chemical Society, Vol. 13, No. 13, 2009, pp. 4702-4709. K. M. Masukawa, P. A. Kollman and I. D. Kuntz, “Investigation of Neuraminidase-Substrate Recognition Using Molecular Dynamics and Free Energy Calculations,” Journal of Medicinal Chemistry, Vol. 46, No. 26, 2003, pp. 5628-5637. doi:10.1021/jm030060qB. J. Smith, P. M. Colman, M. von Itzstein, B. Danylec and J. N. Varghese, “Analysis of Inhibitor Binding in Influenza Virus Neuramini-dase,” Protein Science, Vol. 10, No. 4, 2000, pp. 689-696. doi:10.1110/ps.41801M. von Itzstein, W. Y. Wu, G. B. Kok, M. S. Pegg, J. C. Dyason, B. Jin, T. V. Phan, M. L. Smythe, H. F. White, S. W. Oliver, P. M. Colman, J. N. Varghese, D. M. Ryan, J. M. Woods, R. C. Bethell, V. J. Hotham, J. M. Cameron and C. R. Penn, “Rational Design of Potent Sialidase-Based Inhibitors of Influenza Virus Replica-tion,” Nature, Vol. 363, 1993, pp. 418-423. doi:10.1038/363418a0
is the average Gibbs energy, 
is the electrostatic and nonpolar free energy from implicit solvation. In this particular study, Generalized Born with Molecular Volume (GBMV) was applied for implicit solvation. The last term (TS) is the temperature and entropy contribution, while the first term of the right hand side (
) is the energy term produced by the applied forcefield, which is the potential energy of the system [
(2)