THEORETICAL & APPLIED MECHANICS LETTERS 4, 041009 (2014)

Wavelet method applied to specific adhesion of elastic solids via

molecular bonds
Lei Zhang, Jizeng Wang,a) Youhe Zhou
Key Laboratory of Mechanics on Disaster and Environment in Western China,
Ministry of Education, College of Civil Engineering and Mechanics, Lanzhou University,
Lanzhou 730000, China
(Received 25 April 2014; revised 12 May 2014; accepted 21 May 2014)

Abstract We propose a wavelet method to analyze the stochastic-elastic problem

of specific adhesion between two elastic solids via ligand-receptor bond clus-
ters, which is governed by a nonlinear integro-differential equation with a sin-
gular Cauchy kernel to describe the mean-field coupling between deformation of
elastic materials and stochastic behavior of the molecular bonds. To solve this
problem, Galerkin method based on a wavelet approximation scheme is adopted,
and special treatment which transforms the singular Cauchy kernel into a smooth
one has been proposed to avoid the cumbersome calculation of singular integrals.
Numerical results demonstrate that the method is fully capable of solving the spe-
cific adhesion problems with complex nonlinear and singular equations. Based on
the proposed method, investigations are performed to reveal the relation between
steady-state pulling force and mean surface separation under different stress con-
centration indexes, which is crucial for assembling the overall constitutive rela-
tions for multicellular tumor spheroids and polymer-matrix microcomposites.

c 2014 The Chinese Society of Theoretical and Applied Mechanics. [doi:10.1063/2.1404109]

Keywords wavelet, cell adhesion, nonlinearity, singularity, force-separation relation

Biological cells must adhere to an extracellular matrix (ECM) or other cells to perform their
physiological functions like differentiation, spreading, and healing.1 For instance, tumor cells
usually live in an environment formed by other host cells, ECM and extracellular liquid,2,3 whose
duplication, reorganization, and deformation are deeply affected by the adhesion molecules which
bind them together by exerting force of measurable strength;3 ECM elasticity significantly affects
the integrin activation in bone marrow mesenchymal stem cells, as experimentally confirmed by
Du et al.,4 and theoretically explained by Xu et al.5 Lifetime for individual molecular bonds is
usually very short.6–9 When many single bonds form into a molecular bond cluster, the lifetime
becomes much longer due to stochastic group effects. Qian et al.10,11 have investigated how the
lifetime of such a cluster is affected by the elastic property of the cell-substrate system through
Monte Carlo simulations. This finding is consistent with the fact that cells may strongly respond
to changes in cellular and environmental stiffness. To gain further insight into this issue, for a
specific adhesion system with two elastic solids contacted via a molecular bond cluster, Wang
and Gao12 and Wang et al.13 have proposed an analytical model by considering the mean-field

a) Corresponding author. Email: jzwang@lzu.edu.cn.

041009-2 L. Zhang, J. Z. Wang, Y. H. Zhou Theor. Appl. Mech. Lett. 4, 041009 (2014)

stochastic-elastic coupling to investigate the dependence of adhesion strength on the cluster size
and interfacial shape.
However, the above progress is still far from enough towards the fully understanding of
mechanisms on cell adhesion. One subject connected to this study is the force-separation rela-
tion between adhered cells in a multi-cellular biological medium, such as multi-cellular tumor
spheroids,14,15 and other tissues.16 Studies have demonstrated that adhesive forces between cell-
substratum and cell-cell critically determines the organization, ultimate morphology, and cytoar-
chitecture formed by liver tissue cells.16 Quantitative study on constitutive relations of multi-
cellular biological systems will help us to understand the mechanism of the tumor growth,14,15
in which cell adhesion plays a vital role. However, the governing equations of the specific ad-
hesion model of cellular interactions proposed by Wang and Gao12 and Wang et al.13 consist of
one nonlinear integro-differential equation with singular Cauchy kernel and one strong nonlinear
differential equation. Therefore, how to efficiently and precisely solve these equations becomes
crucially important for the understanding of the adhesion behavior of the system.
For linear integro-differential equations with singular Cauchy kernel, approximate solutions
can be obtained by using numerical methods based on the so-called airfoil polynomials17 and
Legendre polynomials.18 For coupled nonlinear singular integro-differential equations, the power
series based iterative algorithm has been adopted by Wang and Gao12 and Wang et al.13 Appli-
cations of this method involves the calculation of a number of Cauchy type singular integrals,
which usually needs sophisticated mathematical skills. In order to avoid this cumbersome situa-
tion and establish a generally efficient mathematical tool to solve a broad range of strong nonlinear
mathematical problems arising from molecular and cellular biomechanics studies, we adopt a re-
cently developed wavelet based numerical method for the solution of the adhesion problems in
the present study. For the differential equations, such a method has been demonstrated very ef-
ficient and accurate no matter the nonlinearity of the equations are weak or strong, in contrast to
conventional approaches for nonlinear problems, which may be valid only for specific individual
problems. This outstanding characteristic of the wavelet method can be ascribed to the fact that
the method is established based on a modified wavelet approximation scheme of square-integrable
functions, which allows each expansion coefficient being explicitly expressed by a single-point
sampling of the function being approximated.19,20 For the singular Cauchy kernel in the equations,
a mathematical technique is proposed to change it step-by-step into a smooth one.
In the present study, we adopt the stochastic-elastic model of Wang and Gao12 for the mean-
field description of specific adhesion of soft elastic media via clusters of molecular bonds. Fig-
ure 1 shows the adhesion of two elastic half-spaces contacting through a cluster of ligand-receptor
bonds, which forms an adhesion patch with size 2a and stretched by a remotely applied force P
(per unit out-of-plane thickness). We neglect such non-specific interactions as the van der Waals
force. Cartesian coordinates (x, z) are located at the middle of the adhesion region of the lower
elastic half-space. We assume that the initial bond density distribution ρ0 is a constant, while the
density of the closed bonds is denoted as ρb (x,t).
The Young’s modulus, Poisson’s ratio, and vertical displacements of the lower and upper
elastic half-spaces are denoted as E1 , E2 , ν1 , ν2 , and u1 , u2 , respectively. Continuum contact
mechanics is applied to model the deformation of elastic media. Then the relation between normal
041009-3 Wavelet method applied to specific adhesion of elastic solids via molecular bonds

E2, v2
ξ, ρb, h, 2a p(s)
E1, v1

Fig. 1. Schematic plot of two elastic half-spaces connected via a cluster of ligand-receptor bonds.

traction along the interface and the surface separation u = u2 − u1 + h can be written as12,21
 a
∂u 2 p(s) dh
= ds + , (1)
∂ x πE ∗ −a x−s dx

where 1/E ∗ = (1 − ν12 )/E1 + (1 − ν22 )/E2 .

a
Define the mean surface separation as ua = [1/(2a)] −a u dx. In this study, we consider the
closed bonds as Hookean springs of stiffness ξ . Within the adhesion region, we have normal
interfacial traction as p(x) = −ξ ρb u, and the total adhesion force as
 a  a
P= p(x) dx = ξ ρb u dx. (2)
−a −a

Following Bell,9 the association and dissociation kinetics of the bonds can be described by

dρb / dτ = γ (ρ0 (x) − ρb (x, τ )) − eξ u/Fb ρb (x, τ ), (3)

in which τ = k0t represents the normalized time, γ the dimensionless rebinding rate of each pair
of open bonds, k0 the dissociation rate of a closed bond under no applied force. The parameter Fb
is a reference force scale about 4 pN. When the steady state is considered, Eq. (3) can be reduced
to
 
ρb = γρ0 γ + eξ u/Fb . (4)

We define dimensionless variables as follows

uξ x ρb h ξ aρ0 P E∗
û = , r = , ρ̂ = , ĥ = , α = , P̂ = , β = , (5)
Fb a ρ0 a E∗ aρ0 Fb ρ0 Fb

where the dimensionless parameter α has been recognized as the so-called stress concentration
index. It can be seen from Eq. (5) that α is proportional to the bond density, bond stiffness,
the adhesion size, and the inverse of the effective elastic modulus. Numerical solutions12 have
shown that, for α < 0.1, the stress along the interface appears very uniform, while it becomes
concentrated to the adhesion edge for α > 1.
041009-4 L. Zhang, J. Z. Wang, Y. H. Zhou Theor. Appl. Mech. Lett. 4, 041009 (2014)

Based on Eq. (5), Eqs. (1), (2), and (4) can be reduced to
 1
∂ û 2α ρ̂ û dĥ
=− ds + β α , (6)
∂r π −1 r − s dr
 1
P̂ = ρ̂ û dr, (7)
−1
ρ̂ = γ /(γ + eû ). (8)

Considering the cases for small stress concentration index α , then we have ∂ û/∂ r ≈ 0, or
û ≈ const. This will further lead to ρ̂ ≈ const. and

P̂ = 2γ ûa /(γ + eûa ), (9)

which gives the relation between steady-state pulling force and the mean surface separation at
small α .
When the pulling force is small, Taylor-series expansion of the right side of Eq. (9) gives a
linear relation between the normalized pulling force and the averaged separation between the two
opposing surface as P̂ = 2γ ûa /(γ + 1).
In order to solve Eqs. (6)–(8), we define x̄ = 2r − 1. By applying integration in terms of
variable x̄ to Eq. (6), we can transform the Cauchy kernel into a weakly singular logarithm kernel
as follows
  1
4α 1 4α
û(x̄) − û(0) + ρ̂ (û)û(ȳ) ln |x̄ − ȳ| dȳ − ρ̂ (û)û(ȳ) ln ȳ dȳ =
π 0 π 0
β α ĥ(x̄) − β α ĥ(0). (10)

Define a new function as s1 (x̄, ȳ) = ln |x̄ − ȳ| dȳ. After performing integration by parts to the
logarithm kernel, then Eq. (10) can be changed into
 1
4α 4α d
û(x̄) − ρ̂ (û(0))û(0)I(x̄, 0) − I(x̄, ȳ) (ρ̂ (û)û) dȳ + c1 = β α ĥ(x̄), (11)
π π 0 dȳ

where the kernel I(x̄, ȳ) = s1 (x̄, ȳ) − s1 (x̄, 1) is a smooth function, c1 is a constant. Equations (7)
and (11) form a well-posed problem.
Based on the work of Wang,22 for a function f (x) ∈ L2 [0, 1], we have

2n  
k
f (x) ≈ Pnx f (x) = ∑f 2n
ϕn,k (x), x ∈ [0, 1], (12)
k=0

in which ϕn,k (x) = ∑−1 i=2−3N+M1 T0,k (i/2 , β 0 )φ (2 x − i + M1 ) + φ (2 x − k + M1 ) for k ∈ [0, 3],
n n n
n
ϕn,k (x) = φ (2n x − k + M1 ) for k ∈ [4, 2n − 4], ϕn,k (x) = ∑2i=2−1+M n +1 T1,2n −k (i/2 , β 1 )φ (2 x −
1 n n

i + M1 ) + φ (2 x − i + M1 ) for k ∈ [2 − 3, 2 ] and φ (x) is the generalized Coiflet-type or-

n n n

thogonal scaling function developed by Wang,22 Supp(φ (x)) = [0, 3N − 1], and T0,k (x, β 0 ) =
∑3i=0 β0,i (ζ0,i,k /i!)xi , T1,k (x, β 1 ) = ∑3i=0 β1,i (ζ1,i,k /i!)(x − 1)i , β 0 = {β0,i = 1}, β 1 = {β1,i = 1},
041009-5 Wavelet method applied to specific adhesion of elastic solids via molecular bonds

i = 0, 1, 2, 3, where ζ 0 = 2−in ζ0,i,k and ζ 1 = 2−in ζ1,i,k , i, k = 0,1,2,3, are given by Wang22 as
⎡ ⎤ ⎡ ⎤
1 0 0 0 1 0 0 0
⎢ ⎥ ⎢ ⎥
⎢ −11/6 3 −3/2 1/3 ⎥ ⎢ 11/6 −3 3/2 −1/3 ⎥
⎢ ⎥ ⎢ ⎥
ζ0 = ⎢ ⎥, ζ 1 = ⎢ ⎥. (13)
⎢ 2 −5 4 −1 ⎥ ⎢ 2 −5 4 −1 ⎥
⎣ ⎦ ⎣ ⎦
−1 3 −3 1 1 −3 3 −1

In order to employ the wavelet Galerkin method, we need to apply Eq. (12) to approximate
n n
various functions in Eqs. (7) and (11) as û(x̄) ≈ ∑2k=0 ûk ϕn,k (x̄), ĥ(x̄) ≈ ∑2k=0 ĥk ϕn,k (x̄), I(x̄, 0) ≈
n n n n  (ȳ).
∑2k=0 I(x̄k , 0)ϕn,k (x̄), I(x̄, ȳ) ≈ ∑2k=0 ∑2j=0 I(x̄k , ȳ j )ϕn,k (x̄)ϕn, j (ȳ), dρ̂ (û)û/ dȳ ≈ ∑2i=0 ρ̂ (ûi )ûi ϕn,i
Substituting the above expansions into Eqs. (7) and (11), and applying Galerkin method,
yields

Ax − (4α /π)By Dc1 = β α Q , D Ty = P̂/2,

By +D (14)

where matrices, A = {al,k = 01 ϕn,l (x̄)ϕn,k (x̄)dx̄}, B = AIM AIM, except for its first vector B l,1 =
1  
AI k,1 , I = {Ik, j = I(x̄k , ȳ j )}, M = {m j,i = 0 ϕn,i (ȳ)ϕn, j (ȳ) dȳ}, D = {dl = 01 ϕm,l (x̄) dx̄},
AIM j,1 +AI
vectors x = {ûk = û(k/2n )}, y = {ρ̂ (ûi )ûi }, Q = hA hA, h = {hk = ĥ(x̄k )}, l, k, j, i = 0, 1, · · · , 2n . The
generalized connection coefficient m j,i , al,k , and dl can be obtained exactly by using the procedure
suggested by Wang.22 Equation (14) is a set of nonlinear algebraic equations with 2n + 2 unknown
variables.
In order to justify the proposed wavelet Galerkin algorithm, we consider the solution of the
following integro-differential equation with the same singular kernel as in Eq. (6)
 1  1
d f (r) f (s) 1+r 4
−λ ds = 2r(1 + λ ) + λ (1 − r2 ) ln , f (r) dr = − . (15)
dr −1 r−s 1−r −1 3

It can be easily verified that the exact solution of Eq. (15) is f (r) = r2 − 1. Figure 2 shows the
absolute error of the numerical solution obtained by using the wavelet method for different values
of λ ; with an increase of λ , the singularity of the equation can be enhanced accordingly. It can
be seen from Fig. 2 that the maximum error of the proposed wavelet Galerkin method can keep
below 0.1% when λ becomes large.
Organic tissues are usually made up of different types of cells and the components that con-
stitute the extracellular matrix. When the mechanical properties of these tissues including tumors
are studied, quantitative descriptions on the individual adhesive interactions between adjacent
cells or between cells and the ECM become crucial. As a mean-field approximation, we consider
the force-separation relation between two elastic solids contacted via a cluster of ligand-receptor
bonds as shown in Fig. 1. Figure 3 plots the relation between the steady-state pulling force and
the mean surface separation for the adhesion system. It can be seen from Fig. 3 that the force-
separation relation is obviously nonlinear. When the stress concentration index α is less than 1,
this relation is in good agreement with the limiting analytical relation of Eq. (9). The relation starts
to deviate significantly from the analytical relation when the stress concentration index α reaches
2.5. In our previous studies,12 we have demonstrated that α can be viewed as an index to indicate
041009-6 L. Zhang, J. Z. Wang, Y. H. Zhou Theor. Appl. Mech. Lett. 4, 041009 (2014)

how the interfacial traction is distributed. As shown in Fig. 4, increasing α can lead to stress
concentration near the edge of the contact region. If α is much smaller than 1, the distribution of
interfacial traction becomes nearly uniform. Therefore, more uniform distribution of the interfa-
cial traction corresponds to better agreement between the numerically obtained force-separation
relation and the analytical formula of Eq. (9).

103
λ = 1/p2 λ = 1.0 0.8
λ = 1/p λ=p γ/2

Normalized pulling force

4
10 ^
h/0.4
0.6 Analytical
β/1
Absolute error

105 Linear
0.4 α = 1.0Τ104
α = 1.0
106 α = 1.5
0.2 α = 2.0
α = 2.5
107 0
0 0.2 0.4 0.6 0.8 1.0 1.2
Normalized mean surface separation
108
0 0.2 0.4 0.6 0.8 1.0
x Fig. 3. Relation between the mean normalized
surface separation and normalized pulling force
Fig. 2. Distribution of the absolute errors of the for different values of the stress concentration in-
numerical solution along x. dex α .

α = 0.1
Normalized interfacial traction

0.30
α = 1.0
α = 1.5
α = 2.0
0.20

γ/2
^
0.10 h/0.4
β/1
^
p/0.3
0
1.0 0.5 0 0.5 1.0
r

Fig. 4. The normalized interfacial traction distribution under different values of the stress concentration
index α .

In conclusion, we have developed a wavelet Galerkin method for the analysis of the spe-
cific adhesion between two soft elastic media via clusters of ligand-receptor bonds. This method
can efficiently and accurately deal with the nonlinear Cauchy type singular integrals, providing
a promising numerical tool to study a broad range of biomechanics problems with features of
strong nonlinearity and singularity. By using the proposed wavelet method, we have numerically
obtained the nonlinear force-separation relations for the specific adhesion systems. In the limiting
cases of the stress concentration index being much smaller than 1, this relation has been derived
analytically. This study is crucial for assembling the overall constitutive relations from biological
tissues including tumors to polymer-matrix microcomposites, etc.
041009-7 Wavelet method applied to specific adhesion of elastic solids via molecular bonds

This work was supported by the National Natural Science Foundation of China (11032006 and
11121202), National Key Project of Magneto-Constrained Fusion Energy Development Program
(2013GB110002), and the Fundamental Research Funds for the Central Universities (lzujbky-2013-1).

1. X. Chen, B. M. Gumbiner. Crosstalk between different adhesion molecules. Current Opinion in Cell Biology 18,
572–578 (2006).
2. W. M. Klieser. Multicellular spheroids. Journal of Cancer Research and Clinical Oncology 113, 101–122 (1987).
3. D. Ambrosi, L. Preziosi. Cell adhesion mechanisms and stress relaxation in the mechanics of tumours. Biomechanics
and Modeling in Mechanobiology 8, 397–413 (2009).
4. J. Du, X. Chen, X. Liang, et al. Integrin activation and internalization on soft ECM as a mechanism of induction of
stem cell differentiation by ECM elasticity. Proceedings of the National Academy of Sciences 108, 9466–9471 (2011).
5. G. K. Xu, C. Yang, J. Du, et al. Integrin activation and internalization mediated by extracellular matrix elasticity: A
biomechanical model. Journal of Biomechanics 47, 1479–1484 (2014).
6. F. Li, S. D. Redick, H. P. Erickson, et al. Force measurements of the α5 β1 integrin–fibronectin Interaction. Biophysical
Journal 84, 1252–1262 (2003).
7. A. D. Bershadsky, N. Q. Balaban, B. Geiger. Adhesion-dependent cell mechanosensitivity. Annual Review of Cell and
Developmental Biology 19, 677–695 (2003).
8. R. Merkel, P. Nassoy, A. Leung, et al. Energy landscapes of receptor–ligand bonds explored with dynamic force
spectroscopy. Nature 397, 50–53 (1999).
9. G. I. Bell. Models for the specific adhesion of cells to cells. Science 200, 618–627 (1978).
10. J. Qian, J. Wang, H. Gao. Lifetime and strength of adhesive molecular bond clusters between elastic media. Langmuir
24, 1262–1270 (2008).
11. J. Qian, J. Wang, Y. Lin, et al. Lifetime and strength of periodic bond clusters between elastic media under inclined
loading. Biophysical Journal 97, 2438–2445 (2009).
12. J. Wang, H. Gao. Size and shape dependent steady-state pull-off force in molecular adhesion between soft elastic
materials. International Journal of Fracture 166, 13–19 (2010).
13. J. Wang, J. Yao, H. Gao. Specific adhesion of a soft elastic body on a wavy surface. Theoretical and Applied Mechanics
Letters 2, 014002 (2012).
14. D. Ambrosi, F. Mollica. On the mechanics of a growing tumor. International Journal of Engineering Science 40,
1297–1316 (2002).
15. V. Cristini, J. Lowengrub, Q. Nie. Nonlinear simulation of tumor growth. Journal of Mathematical Biology 46, 191–
224 (2003).
16. M. J. Powers, L. G. Griffith. Adhesion-guided in vitro morphogenesis in pure and mixed cell cultures. Microscopy
Research and Technique 43, 379–384 (1998).
17. A. Badr. Integro-differential equation with Cauchy kernel. Journal of Computational and Applied Mathematics 134,
191–199 (2001).
18. A. Mennouni, S. Guedjiba. A note on solving integro-differential equation with Cauchy kernel. Mathematical and
Computer Modelling 52, 1634–1638 (2010).
19. X. J. Liu, J. Z. Wang, Y. H. Zhou. Wavelet solution of a class of two-dimensional nonlinear boundary value problems.
Computer Modelling in Engineering & Science 92, 493–505 (2013).
20. Y. H. Zhou, X. M. Wang, J. Z. Wang, et al. A wavelet numerical method for solving nonlinear fractional vibration,
diffusion, and wave equations. Computer Modelling in Engineering & Science 1981, 1–24 (2011).
21. K. L. Johnson. Contact Mechanics. Cambridge University Press, Cambridge (1987).
22. J. Wang. Generalized theory and arithmetic of orthogonal wavelets and applications to researches of mechanics in-
cluding piezoelectric smart structures, [PhD Thesis]. Lanzhou Univeristy, Lanzhou (2001) (in Chinese).