Physics Letters A 381 (2017) 276–279

Contents lists available at ScienceDirect

Physics Letters A
www.elsevier.com/locate/pla

Electronic transport on the spatial structure of the protein:

Three-dimensional lattice model
R.G. Sarmento a , N.F. Frazão b , A. Macedo-Filho c,∗
a
Departamento de Ciências Biológicas, Universidade Federal do Piauí, 64800-000 Floriano, PI, Brazil
b
Centro de Educação e Saúde, Universidade Federal de Campina Grande, 581750-000 Cuité, PB, Brazil
c
Campus Prof. Antonio Geovanne Alves de Sousa, Universidade Estadual do Piauí, 64260-000 Piripiri, PI, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: We report a numerical analysis of the electronic transport in protein chain consisting of thirty-
Received 17 November 2016 six standard amino acids. The protein chains studied have three-dimensional structure, which can
Accepted 21 November 2016 present itself in three distinct conformations and the difference consist in the presence or absence of
Available online 25 November 2016
thirteen hydrogen-bondings. Our theoretical method uses an electronic tight-binding Hamiltonian model,
Communicated by Z. Siwy
appropriate to describe the protein segments modeled by the amino acid chain. We note that the
Keywords: presence and the permutations between weak bonds in the structure of proteins are directly related
Electronic transport to the signing of the current–voltage. Furthermore, the electronic transport depends on the effect of
Lattice model temperature. In addition, we have found a semiconductor behave in the models investigated and it
Protein suggest a potential application in the development of novel biosensors for molecular diagnostics.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction Table 1
The twenty standard amino acids.

Alanine (A) Cysteine (C) Aspartic Acid (D) Glutamic Acid (E)
Proteins are the main functional elements in all living organ- Phenylalanine (F) Glycine (G) Histidine (H) Isoleucine (I)
isms. In biological systems, the genetic information stored in DNA Lysine (K) Leucine (L) Methionine (M) Asparagine (N)
is transcribed by the mRNA. The combinations of three of the four Proline (P) Glutamine (Q) Arginine (R) Serine (S)
Threonine (T) Valine (V) Tryptophan (W) Tyrosine (Y)
nucleotides of the mRNA identify an amino acid, known as tRNA,
which assists in the translation of the genetic code contained in
DNA and mRNA, thereby producing a specific protein [1,2]. There- ganization of the protein depends on the amino acid comprising
fore, proteins are macromolecules consisting of a chain of amino and as they are available in relation to others in the chain. [2,3].
acid residues linked by peptide bonds, thus forming a primary Nowadays, biomolecules are strong candidates to be part of
structure. sensors for diagnosis and prognosis in biomedical area. Generally,
The biological function of a protein is related to its spatial enzymes (proteins) are often used in sensors due to their speci-
structure, in other words the primary structure has to bend to ficity and the reaction products may be electrochemical transducer
form a secondary structure and then packaged in a tertiary struc- for measurement. When enzymes are immobilized by an electrode
ture. In some cases, the tertiary structures of several proteins or enzymatic reactions is activated and can be translated by electron
subunits have joined to form the quaternary structure. That last transport [4,5].
structure is kept stable by several weak bonds among the subunits The idea of using biomolecules as electronic components is
of the same kind, which maintain the tertiary structure. Some not new. In 1974, Aviram and Ratner were the ones who first
agents may cause changing in the protein conformation, such as: suggested the construction of a simple organic electronic device,
temperature, changing in the medium acidity, and substitution of a called a rectifier, which is composed of a donor and an accep-
single amino acid [2,3]. Following the twenty standard amino acids tor system connected by a bridge of carbon single bonds [6].
which form all proteins are represented in Table 1. The spatial or- Since then, the charge transport through an electrode–molecule–
electrode junction is extensively studied and it is of key impor-
tance in molecular electronics applications [7–10].
* Corresponding author. The aim of this study is to investigate the electronic transport
E-mail address: amfilho@gmail.com (A. Macedo-Filho). in a three dimensional lattice model, which is represented in three

http://dx.doi.org/10.1016/j.physleta.2016.11.026
0375-9601/© 2016 Elsevier B.V. All rights reserved.
 R.G. Sarmento et al. / Physics Letters A 381 (2017) 276–279 277

Fig. 1. Schematic models for the electron transport in three types of proteins represented in three dimensional structure, where (a) P A , (b) P B , and (c) P C connected between
two electrodes. The circles denote sites (electrode or amino acid); the solid lines represent the peptide bonds (t c = 0.3 eV is the hopping term between electrode and chain
and t = 0.3 eV is a peptide bonds energy); the dashed lines denote several hopping terms (h = 0.03 eV, hydrogen bonds).

different conformations labeled as P A , P B , and P C . These mod- Table 2

els have a specific amino acid sequence, differing each other by Connections among amino acid residues. The indexes i and j identify the amino
acids at sequence previously cited and numbered from 1 to 36.
the hydrogen bondings distribution (see Fig. 1), where thirteen hy-
drogen bonds try to mimic a mutation (disease) in a real protein Protein P B Protein P C
as suggested by [11]. For the three proteins under consideration, i j i j i j i j
we take into account the chains of proteins made up of thirty-six 2 31 17 32 3 6 15 24
amino acids obeying the following sequence NKTVVGEPWHCLLF- 15 22 18 21 4 29 16 27
PRRDKNQMSYLTGIAGEDSAAI. The protein P A has a simple struc- 15 24 18 23 5 23 17 24
16 21 19 32 6 27 17 26
tural arrangement built of amino acids linked by covalent bonds
16 31 20 31 7 10 19 32
(peptide bonds). The proteins P A and P B present a structural ar- 17 24 27 30 14 25 20 31
rangement created of amino acids joined through peptide bonds 17 26 14 27
and non-covalent bonds (hydrogen bonds). We carried out the
electronic transport analysis of the protein by fixing the protein
Table 3
between two electrodes. When a potential difference is applied
The average values of the ionization energy of each specific amino acid A , B , ..., Y .
across it, we can obtain an electronic signature of the protein on
the effects of low temperature and its protein conformation, serv- ε A = 8.84 eV εC = 8.68 eV ε D = 8.34 eV ε E = 7.99 eV
ε F = 9.24 eV εG = 8.00 eV ε H = 8.72 eV ε I = 8.77 eV
ing as a parameter for possible disease investigations or for future ε K = 8.15 eV εL = 8.88 eV εM = 8.41 eV εN = 8.51 eV
applications in nanotechnology. ε P = 8.75 eV ε Q = 8.35 eV ε R = 8.17 eV ε S = 8.34 eV
εT = 8.54 eV ε V = 8.51 eV εW = 8.32 eV εY = 8.87 eV
2. Model and implementation

the peptide bond energy and h i , j represents hydrogen bond en-

To address the influence of the previously mentioned factors
ergy, that are weak intermolecular bonds in comparison to pep-
on the charge transport efficiency, we used the time indepen-
tide bonds, depicted in Figs. 1(b) and 1(c) by t (solid line) and
dent Schrodinger equation to calculate the current–voltage for spa-
tial structure of the protein. The tight-binding model Hamiltonian h (dashed line). Despite the hydrogen bonds h i , j to be weak, the
that describes the electronic transport, through electrode–protein– large quantities and the form how the bonds are arranged among
electrode system, is given by (see Fig. 1): amino acids, in the chain, can stabilize and to determine the spatial
structure of the proteins. Here, we consider 13 connections useful
H = H p + He + Hc. (1) to stabilize these conformational arrangements as it is possible to
be seen in Table 2.
The first term on the right side of Eq. (1) describes the hamiltonian In Table 3, the variable εi represents the energy on-site matrix
through the protein chain and it can be written as elements, which is given by the average values of the energy of
ionization of the nucleic acids of RNA, i.e., εG = 7.77 eV (Guanine),

36 
35
 
Hp = εi |i i | + t i ,i ±1 cos(θ)|i i ± 1| εC = 8.68 eV (Cytosine), ε A = 8.26 eV (Adenine) and εU = 9.32 eV
(Uracil); that was estimated by V. M. Orlov and collaborators, using
i =1 i =1
photoionization mass-spectrometry in gas phase [13–15].

13
  To investigate the electron transport behavior in a spatial
+ h i , j cos(θ)|i  j | , (2)
structure of the protein on the effects of temperature, which
i = j
is known to be one of the principal parameters in experiments
where |i  represents a localized state of an electron at the i-th with biomolecules, since variation of the temperature induces
amino acid. The quantities t i ,i +1 and h i , j are the nearest-neighbor structural disorder and fluctuations of the system. Thus, we con-
electronic overlaps (hopping terms) between the amino acid in sider the variation of the temperature to the hopping integrals
protein chain, with their values respectively given by 0.3 eV and in terms of twist-angle fluctuations. In this context, we assume
0.03 eV. Each one values are within the range of values obtained θ being the twist-angle fluctuations between neighboring amino
by Chemical Quantum Calculation [12]. The term t i ,i +1 represents acid, which means the angle of torsion applied along a chemi-
278 R.G. Sarmento et al. / Physics Letters A 381 (2017) 276–279

cal bond. Here we consider that θ follows a Gaussian distribution

such that θ = 0 and θ 2  = k B T / I 2 from the equipartition the-
orem with I 2 = 250 K, where T is temperature in kelvin, I is
the reduced moment of inertia for the relative rotation of the two
adjacent amino acid and  is the oscillator frequency of the vibra-
tion mode. Using these expressions, the temperature-dependent
hopping integrals can be obtained t i ,i +1 cos(θ) ≈ t i ,i +1 (1 − θ 2 /2)
and h i , j cos(θ) ≈ h i , j (1 − θ 2 /2) [16].
The second term, related to the two semi-infinite metallic elec-
trodes, reads

0
 
He = εss |i i | + t ss |i i ± 1| + t ss |i ± 1i |
i =−∞
∞
  
+ εss |i i | + t ss |i i ± 1| + t ss |i ± 1i | . (3)
i =37

Here εss and t ss are the ionization energy and hopping term of the
electrode (platinum), whose values are given by εss = 5.36 eV and
t ss = 12 eV [8,17].
Finally, the third term describes the coupling between the pro-
tein chain and semi-infinite metallic electrodes yielding
   
H c = t C |01| + |10| + t C |3637| + |3736| , (4)
where t C is the hopping term between the electrode and the pro-
tein chain. Also, we assume t C = 0.3 eV (t C ∼ t protein ) to allow a
better characterization of intrinsic conduction properties [16].
The current I is defined as the rate of charge transport, and
it is of direct interest since it corresponds to an experimentally
observable quantity. The current–voltage characteristic is obtained
from the Landauer–Büttiker [8,18] formula and expressed as
Fig. 2. Current–voltage curves for protein P A (top box), protein P B (middle box),
∞ and protein P C (bottom box) at low temperature (0 K: solid line, 16 K: dashed
2e   line, and 40 K: dotted line). The nonlinear I –V curves exhibit a current gap at low
I (V ) = T ( E ) f L ( E ) − f R ( E ) dE . (5)
h applied bias, which is an indication of semiconductor behavior.
−∞

Here f L ( R ) = {exp[β( E − μ L ( R ) )] + 1}−1 is the Fermi–Dirac distri-
bution, β = 1/k B T and μ L ( R ) stands for electrochemical potentials
of the metallic electrodes. We choose μ L = E f + (1 − η)eV and
μ R = E f − ηeV , where E f is the equilibrium Fermi energy, V is
the applied voltage and η is a parameter describing the possible
asymmetry of contact to leads, we assume here as E f = 7.5 eV
and η = 1/2 respectively. It is important to mention that there is
a direct influence of the transmittance T(E) on the current–voltage,
since the transmittance behavior depends on the effect of temper-
ature.
3. Numerical results and discussion
The results for the current–voltage (I × V ) for three proteins at
different temperatures are plotted in Fig. 2. The results show the
behavior of the current in the positive and negative biased protein.
For both the positive and negative polarities, there is a character-
istic insulator region for −2 V ≤ V bias ≤ 2 V, and nonlinear regions
indicating transition toward saturation currents for V bias < −2 V
and V bias > 2 V in all frames. All of them have semiconductors
characteristics, as in the case of dry proteins [19,20].
Each protein has a standard current characteristic, which was
assumed to depend mainly on the structure, keeping fixed the
thirty-six amino acids arranged in ordered sequence. The result of
these calculations is that simple proteins P A , P B and P C can be
distinguished by means of charge transport measurements. Note
that, for a given fixing temperature, the result of the current–
voltage of the protein P A is smaller than the other proteins stud-
ied. This result is due to the fact that proteins P B and P C present
a more number of charge transport channels, due to the hydrogen
bond, so the current I ( P A ) < I ( P B ) < I ( P C ).
The conformation of a protein is stabilized largely by weak in-
teractions. Although the proteins P B and P C present the same
amounts of hydrogen bonds (13 links), protein P C has a saturation
current greater than the protein P B , that is justified by hydrogen
bonds strategically distributed along the protein, preventing the
structural arrangement is flexible, see Fig. 1. The conformational
of a protein is more thermodynamically stable when it tends to
lower Gibbs free energy, when it reaches this stage the protein
tends to have a folded conformation. The unfolded state of a pro-
tein is characterized by a high degree of conformational entropy
[11]. Therefore, for a given set temperature, the protein has a sta-
ble conformation has a higher saturation current, lower Gibbs free
energy and the lowest degree of entropy.
Our numerical results indicate that variations of temperature
(0 K, 16 K and 40 K) in the system significantly alter the current–
voltage curves of each protein. At low temperature, the twist-angle
fluctuation causes the averaged hopping matrix element to de-
crease and to fluctuate, leading to a significant thermal enhance-
ment of charge transport. On the other hand, with increasing tem-
perature, the system hopping term decreases, it can be seen easily
by Eq. (2), the reduction hopping term produces a decrease of the
resonance peaks of the transmission coefficients (not shown here,
for similar behavior, see W. Ren et al. [21]), consequently causes
a decrease in current–voltage. All this demonstrates that current–
voltage in proteins are strongly sensitive to the conformational
structure and temperature of the chain.
4. Conclusions
To summarize, we have analyzed the electronic transport on
three types of protein chains composed of thirty-six standard
 R.G. Sarmento et al. / Physics Letters A 381 (2017) 276–279
amino acids modeled in a three-dimensional lattice model. The
current–voltage curves, dependent on temperature, were obtained
by the use of tight-binding protein to an electrode–electrode-
system model. Aiming to further contribute to the present un-
derstanding of the electronic transport of the protein, we have
considered two lattice models three-dimensional of protein with
peptide links and hydrogen bonds, to compare them with one lat-
tice model three-dimensional without hydrogen bonds. Although
different circumstances may affect the links peptides and the links
hydrogens between the amino acids system in realistic conditions,
the knowledge gained from these three-dimensional lattice mod-
els and of the current–voltage results might serve as an insight to
help future experimental works on electronic transport properties
of protein molecule-based devices.
Based on our numerical results, all proteins studied showed the
saturation current curves with a characteristic semiconductor. The
effect of thermal fluctuation in the current–voltage curve for three
models of proteins were reported. In particular, the current–voltage
characteristics presented here show a promise a broad range of
nanocircuit developed to distinguish different protein profiles (nor-
mal or mutant). If electronic transport characteristics are antici-
pated to have some direct impact on the structure of protein, our
results open new directions of works in relation with biological
concerns.
Acknowledgements
This work was partially financed by the Brazilian Research
Agencie CNPq (Universal). A.M.F. received financial support from
CNPq Project 459961/2014-4.
279
References
[1] D. Voet, J.G. Voet, Biochemistry, vol. 1, 3rd ed., Wiley, Hoboken, NJ, 2004.
[2] D.L. Nelson, M.M. Cox, Lehninger Principles of Biochemistry, 5th ed., W.H. Free-
man, January 2008.
[3] J.M. Berg, J.L. Tymoczko, L. Stryes, Biochemistry: International Edition Hard-
cover, 2011.
[4] S. Viswanathan, T.N. Narayanan, K. Aran, K.D. Fink, J. Paredes, P.M. Ajayan, S.
Filipek, P. Miszta, H.C. Tekin, F. Inci, U. Demirci, P. Li, K.I. Bolotin, D. Liepmann,
V. Renugopalakrishanan, Mater. Today. 18 (2015) 513.
[5] H. Dong, K. Hao, Y. Tian, S. Jin, H. Lu, S. Zhou, X. Zhang, Biosens. Bioelectron.
53 (2014) 377.
[6] A. Aviram, M.A. Ratner, Chem. Phys. Lett. 29 (1974) 277.
[7] Y. Kitaguchi, S. Habuka, H. Okuyama, S. Hatta, T. Aruga, T. Frederiksen, M. Pauls-
son, H. Ueba, Sci. Rep. 5 (2015) 11796.
[8] R.G. Sarmento, U.L. Fulco, E.L. Albuquerque, E.W.S. Caetano, V.N. Freire, Phys.
Lett. A 375 (2011) 3993.
[9] J.I.N. Oliveira, E.L. Albuquerque, U.L. Fulco, P.W. Mauriz, R.G. Sarmento, Chem.
Phys. Lett. 612 (2014) 14.
[10] J.I.N. Oliveira, E.L. Albuquerque, U.L. Fulco, P.W. Mauriz, R.G. Sarmento, E.W.S.
Caetano, V.N. Freire, Europhys. Lett. 107 (2014) 68006.
[11] A.M. Gutin, V. Abkevich, E. Shakhnovich, Fold. Des. 3 (1998) 183.
[12] A. Voityuk, J. Jortner, M. Bixon, N. Roesch, J. Chem. Phys. 114 (2002) 5614.
[13] V.M. Orlov, A.N. Smirnov, Ya.M. Varshavsky, Tetrahedron Lett. 17 (1976) 4377.
[14] B.I. Verkin, L.F. Sukhodub, I.K. Yanson, Dokl. Biophys. 100 (1976) 226.
[15] S. Urano, X. Yang, P.R. LeBreton, J. Mol. Spectrosc. 214 (1989) 315.
[16] S. Roche, E. Maciá, Mod. Phys. Lett. B 17 (2004) 847.
[17] M.L. de Almeida, J.I.N. Oliveira, J.X. Lima Neto, C.E.M. Gomes, U.L. Fulco, E.L. Al-
buquerque, V.N. Freire, E.W.S. Caetano, F.A.B.F. de Moura, M.L. Lyra, Appl. Phys.
Lett. 107 (2015) 203701.
[18] R. Landauer, IBM J. Res. Dev. 1 (1957) 223.
[19] M. Büttiker, Phys. Rev. B 35 (1987) 4123.
[20] A.K. Bakhshi, Prog. Biophys. Mol. Biol. 61 (1994) 187.
[21] W. Ren, J. Wang, Z. Ma, H. Guo, Phys. Rev. B 72 (2005) 035456.