Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

Contents lists available at ScienceDirect

Journal of the Mechanical Behavior of

Biomedical Materials
journal homepage: www.elsevier.com/locate/jmbbm

Engineering hydrogel viscoelasticity T

a,b a a,b b,c,d
Ludovica Cacopardo , Nicole Guazzelli , Roberta Nossa , Giorgio Mattei ,
Arti Ahluwaliaa,b,
⁎

a
Research Centre "E. Piaggio", University of Pisa, Largo Lucio Lazzarino 1, 56122 Pisa, Italy
b
Department of Information Engineering, University of Pisa, Via Girolamo Caruso 16, 56122 Pisa, Italy
c
Optics11 B.V., De Boelelaan 1081, 1081 HV Amsterdam, the Netherlands
d
Biophotonics & Medical Imaging and LaserLaB, VU University Amsterdam, De Boelelaan 1105, 1081 HV Amsterdam, the Netherlands

ARTICLE INFO ABSTRACT

Keywords: The aim of this study was to identify a method for modifying the time-dependent viscoelastic properties of gels
Viscoelasticity without altering the elastic component. To this end, two hydrogels commonly used in biomedical applications,
Damping component modulation agarose and acrylamide, were prepared in aqueous solutions of dextran with increasing concentrations (0%, 2%
Agarose and 5% w/v) and hence increasing viscosities. Commercial polyurethane sponges soaked in the same solutions
Polyacrylamide
were used as controls, since, unlike in hydrogels, the liquid in these sponge systems is poorly bound to the
Mechanobiology
polymer network. Sample viscoelastic properties were characterised using the epsilon-dot method, based on
compression tests at different constant strain-rates. Experimental data were fitted to a standard linear solid
model. While increasing the liquid viscosity in the controls resulted in a significant increase of the characteristic
relaxation time ( ), both the instantaneous (Einst ) and the equilibrium (Eeq ) elastic moduli remained almost
constant.
However, in the hydrogels a significant reduction of both Einst and was observed. On the other hand, as
expected, Eeq – an indicator of the equilibrium elastic behaviour after the occurrence of viscoelastic relaxation
dynamics – was found to be independent of the liquid phase viscosity.
Therefore, although the elastic and viscous components of hydrogels cannot be completely decoupled due to
the interaction of the liquid and solid phases, we show that their viscoelastic behaviour can be modulated by
varying the viscosity of the aqueous phase. This simple-yet-effective strategy could be useful in the field of
mechanobiology, particularly for studying cell response to substrate viscoelasticity while keeping the elastic cue
(i.e. equilibrium modulus, or quasi-static stiffness) constant.

1. Introduction (Awad et al., 2004; Engler et al., 2006a, 2006b; Jay et al., 2009; Kumar,
2014; Nemir and West, 2009; Tse and Engler, 2010; Yang et al., 2011).
Hydrogels are characterised by a biphasic structure: a solid network, In a seminal work, Engler and co-workers demonstrated that stem cells
which confers elastic properties, surrounded by an aqueous solution, alter their commitment depending on substrate stiffness using PAAm
which is associated with their viscous behaviour (Hoffman, 2012). hydrogels prepared with different concentrations of bis-acrylamide
Since hydrogel composition and mechanical properties are similar to (Engler et al., 2006a, 2006b). A few years later, Pek et al. reported on
those of biological tissues, they are of great interest as biomaterials. the differentiation of 3D mesenchymal stem cell (MSC) cultures in
Numerous reports on engineering hydrogels to better recapitulate the thixotropic polyethylene glycol silica gels (Pek et al., 2010). In parti-
in-vivo milieu focus on characterising and tuning their elastic beha- cular, they observed different levels of neural, myogenic and osteogenic
viour, typically through the use of crosslinking strategies (e.g. physical, factor expression as the gel liquefaction stress (i.e. minimum stress re-
chemical, enzymatic) (Hoffman, 2012) and/or adding reinforcements quired to liquefy the hydrogel) increases. Moreover, the expression of
(e.g. fibres, fillers) (Bhattacharya et al., 2006). This is likely due to the these factors was higher in the 3D gels in comparison to the 2D culture,
wealth of literature on cell response to stiffness (Humphrey et al., 2014; indicating that the modulation of stiffness of a three-dimensional en-
Ingber, 1997; Wells, 2008). In particular, polyacrylamide (PAAm) is the vironment is more effective on cell differentiation. Stiffness is also
workhorse of mechanobiology studies and has been used extensively fundamental for cellular migration within 3D matrices. Zaman et al.

⁎
Corresponding author at: Research Centre "E. Piaggio", University of Pisa, Largo Lucio Lazzarino 1, 56122 Pisa, Italy.
E-mail address: [email protected] (A. Ahluwalia).

https://doi.org/10.1016/j.jmbbm.2018.09.031
Received 10 May 2018; Received in revised form 14 September 2018; Accepted 20 September 2018
Available online 21 September 2018
1751-6161/ © 2018 Elsevier Ltd. All rights reserved.
L. Cacopardo et al. Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

showed that keeping biochemical parameters (such as matrix ligands) dependent on the interaction with water and temperature. Its com-
constant, the migration speed of human prostate carcinoma cells was pressive stiffness varies as a function of its concentration and physical
higher within soft scaffolds (Zaman et al., 2006). Finally, neurite ex- gels, stable at an ambient temperature of around 20–25 °C, are formed
tension in agarose gels with different concentration was found to be without external crosslinkers. Water molecules bonded to this system
inversely proportional to the stiffness of the gels (Balgude et al., 2001). contribute to the stability of the agarose double helix and to the stiff-
However, hydrogels as well as biological tissues are intrinsically ness of the gel (Arnott et al., 1974; Nijenhuis, 1997). On the other hand,
viscoelastic. Recognising this, some attention has recently been directed as reviewed above, the stiffness of PAAm depends not only on the initial
to tuning material viscoelasticity thanks to studies demonstrating that quantity of monomer used, but also on the quantity of crosslinker and
cells also respond to the time-dependent properties of their substrates. initiators (Keplinger et al., 2013). Porous polyurethane (PU) sponges
PAAm gels with different viscous properties can be realized by adding soaked in dextran solutions were used as controls. Since the liquid
high molecular weight linear polymers, which are sterically entrapped phase is not bound to the polymer network in these sponges, they are
in the gels (Charrier et al., 2018), or simply by varying the proportion not biphasic gel systems. Thus, unlike agarose and PAAm hydrogels, the
of acrylamide monomer and bis-acrylamide (Cameron et al., 2011). interaction between the solid sponge and liquid filling its pores is weak,
Using the former method, Charrier and colleagues generated gels with a therefore their viscoelastic properties should be directly related to the
loss modulus (G″) of 200 and 500 Pa for 1.8% and 2.75% w/v of viscosity of the liquid phase. The compressive viscoelastic properties of
polymer. The authors reported that fibroblast spreading decreases with both PU sponges and agarose and PAAm hydrogels were measured at
increasing gel viscosity and that this effect is different in case of col- various constant strain rates, using the epsilon dot method (Tirella
lagen or fibronectin coatings, suggesting that different biochemical et al., 2013). Stress-time data collected at different strain rates were
pathways influence cell sensitivity to viscosity. Moreover, they showed then globally fitted to a Maxwell standard linear solid (SLS) model to
that for a constant stiffness (storage modulus G′), the differentiation of determine the effect of increasing dextran solution viscosity on the
primary rat hepatic stellate cells to fibrogenic myofibroblasts on stiff viscoelastic parameters, i.e. the instantaneous and equilibrium elastic
substrates decreases as the viscosity increases. The authors hypothesise moduli, and the characteristic relaxation time.
that cell traction forces decrease in response to relaxation phenomena
(Charrier et al., 2018). Cameron and coworkers prepared PAAm hy- 2. Material and methods
drogels with different ratios of acrylamide and bis-acrylamide. The
values of G″ at 0.005 rad s−1 were 130, 10 and 1 Pa for 15/0.0125%, 2.1. Sample preparation
12/0.0358% and 8/0.1% w/v acrylamide/crosslinker weight ratio, re-
spectively. They observed increased spreading area and proliferation of Agarose solutions were prepared at 1% w/v polymer concentration
MSCs and a higher tendency to differentiate into different lineages as by dissolving agarose powder (Sigma A9539) in aqueous media with
the viscosity increased. Thus, they conclude that “a reduction in both different viscosities using 0% and 5% w/v dextran (Sigma D1037, MW
passive and actively generated isometric cytoskeletal tension is caused 425–570000). In the range of dextran molecular weights and at the
by the inherent creep of substrates with a high loss modulus” (Cameron concentrations used, the aqueous phases are Newtonian (Guo et al.,
et al., 2011). Finally, thanks to a computational model and cell ex- 2013). The solutions were stirred until boiling and then cast into
periments on alginate hydrogels, Chaudhuri and coworkers showed that custom moulds, obtaining 13 mm diameter - 8 mm height cylindrical
cell spreading is higher on ionically crosslinked viscoelastic soft gels samples. PAAm gels were prepared by dissolving 15.6% w/v of Acry-
with respect to covalently cross-linked stiff elastic ones (Chaudhuri lamide monomer (Biorad) in 0%, 2% and 5% dextran in water with the
et al., 2015). addition of 0.02% w/v Bis-acrylamide (N,N0-methylene bis-acryla-
From this short overview, it is clear that to date hydrogel viscoe- mide, Sigma M7256), 0.084% w/v of ammonium persulfate (Biorad)
lastic properties have been modulated by acting on the solid network, and 0.056% v/v Tetramethylethylenediamine (TEMED) (Sigma T9281).
e.g. by varying the polymer and/or crosslinker concentration. This is All the components were mixed and the final solution was first degassed
likely to alter not only the hydrogel viscoelastic dynamics, described by for 2 min and then cast in the custom moulds. The polymerization was
one (or more) characteristic relaxation time(s), but also the residual completed in 90 min at room temperature. Commercial PU sponges
equilibrium elastic properties after the occurrence of viscoelastic phe- (UR303810, water adsorption 0.1%, Goodfellow, Huntingdon – UK)
nomena, described by the equilibrium elastic modulus. Indeed, we re- were cut into cubes (sides≃10 mm) to obtain control samples which
cently reported that glutaraldehyde-crosslinked gelatin hydrogels were submerged in the different aqueous media for about 30 min at
showed increased instantaneous and equilibrium elastic moduli as well room temperature to allow fluid absorption. They were also tested in
as characteristic relaxation time with increasing crosslinker con- dry conditions to obtain a baseline viscoelastic behaviour in the absence
centration, meaning that – for the same polymer concentration – more of liquids.
crosslinked hydrogels are not only stiffer but there is also a concomitant
change in their viscoelastic behaviour towards a more elastic one 2.2. Viscosity measurements
(Mattei et al., 2017).
Given the emerging importance of hydrogel viscoelastic properties The dynamic viscosity (µ) of the different aqueous phases (0%, 2%
in regulating cell behaviour and lineage commitment, it could be of and 5% dextran) was characterised with an AMVN Automated micro-
interest to decouple viscoelastic characteristics from elastic ones, by viscometer (Anton Paar GmbH, Germany), which measures the rolling
modulating viscoelastic relaxation dynamics while keeping the equili- time of a ball through transparent and opaque liquids according to
brium elastic modulus constant. To achieve this, we chose to act on the Höppler falling ball principle (Kulicke and Clasen, 2013). For each so-
hydrogel liquid phase rather than the solid polymer network. In parti- lution, both density and viscosity are measured within the range of
cular, we investigated the effect of changing the viscosity of the aqu- room temperature (16 °C, 21 °C and 26 °C) using a capillary with
eous phase in agarose and polyacrylamide (PAAm) hydrogels through 1.6 mm inner diameter and a 1.5 mm diameter ball. Six repeats were
the addition of dextran. Dextran is a homo-polysaccharide of glucose performed for each measurement.
containing consecutive α(1,6)-linkages, which contribute to the high
solubility of the molecule in water (Naessens et al., 2005). The molecule 2.3. Mechanical testing and statistical analysis
is often used to increase the viscosity of aqueous solutions, for example,
to mimic blood viscosity (Campolo et al., 2014; Liepsch et al., 1991). The viscoelastic properties of the sponges and gels were tested using
Agarose is a linear polysaccharide that is normally insoluble in organic the epsilon-dot method (Tirella et al., 2013). Unconfined bulk com-
solvents, but highly soluble in water. Its gelling capacity is strongly pression tests were performed at different constant strain rates (0.0005,

163
L. Cacopardo et al. Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

decreases with temperature.

The results of the mechanical tests for the sponges are reported in
Fig. 3: the relaxation time increases significantly with increasing dex-
tran concentration; conversely, both Einst and Eeq are fairly constant for
the wet sponges. This indicates that only the SLS dashpot parameter
changes, while both SLS springs remain unaltered as the viscosity in-
creases. The dry sponges have significantly higher Einst (74.47 ± 4.22
Fig. 1. The SLS model and symbols used to describe the lumped parameters kPa) and Eeq (25.39 ± 0.49 kPa) as well as a significantly lower τ
obtained from fitting experimental data to the epsilon dot equations. Note that (1.60 ± 0.21) than the wet ones.
a lower relaxation time τ denotes a more viscous-like behaviour (i.e. when τ → Fig. 4 illustrates the results obtained for agarose gels: the increase in
0 the material is ideally viscous and vice versa when τ → ∞ we have an ideally viscosity of the liquid phase resulted in a significant reduction of both
elastic material). instantaneous elastic modulus and relaxation time, indicating that the
viscoelastic components of the system (i.e. E2 and η, which are placed in
0.001, 0.005 and 0.01 s−1) with a Zwick/Roell ProLine Z005 uniaxial series in the SLS parallel model, Fig. 1) are altered. The Eeq (equivalent
testing device (Germany). All samples except the dry PU sponges were to E1 in Fig. 1) does not vary significantly between agarose gels pre-
tested while being partially submerged in their respective aqueous pared in 0% and 2% w/v dextran solutions, while it was significantly
phase to maintain hydration. Measurements at each strain rate were reduced for hydrogels prepared in 5% w/v dextran.
performed on n = 3 different samples to avoid any effect due to re- In the case of PAAm gels (Fig. 5), Einst and τ follow the same de-
peated testing cycles, leading to a total of 4 (strain rates) x 3 (replicates) creasing trend observed for agarose, while the equilibrium elastic
= 12 samples tested per material investigated. For each sample en- modulus is almost constant with increasing dextran concentration. The
gineering stress-time curves recorded at the same strain-rate were ANOVA analysis shows a significant reduction of both instantaneous
averaged. Then, the mean curves obtained at the different strain rates elastic modulus and characteristic relaxation time with increasing
investigated were globally fitted to a standard linear solid (SLS) model dextran concentration.
as described in (Tirella et al., 2013), obtaining E1, E2 and η (Fig. 1),
which were used to compute the instantaneous (Einst=E1 +E2) and 4. Discussion
equilibrium (Eeq=E1) elastic moduli and the characteristic relaxation
time (τ = η/ E2). In particular, Einst refers to the initial elastic response, In the control PU sponges, there is little or no interaction between
which is given by the sum of all the springs in the lumped model since the spongiosum structure and the liquid phase. Water diffuses inside the
the viscous components are “shorted” out and do not deform while small pores formed during the manufacture of foams, which are pro-
sustaining the load. The equilibrium modulus refers instead to the duced by adding inert gases to the monomers precursors to induce
“static” response after the viscoelastic dynamics have occurred, and it is bubble formation (Sabbahi and Vergnaud, 1993). Thus, the only factor
given by E1, as the viscous components are on “open circuit”, so they which conditions the movement of water through the pores during
cannot sustain any load or stress. Finally, a 1-way ANOVA analysis was unconfined compression is the viscosity of the aqueous solution. This
performed to investigate the effect of varying the dextran solution means that water (which has a lower viscosity than dextran) needs less
concentration on the resultant viscoelastic properties. It should be time to re-distribute through the pores, resulting in a lower relaxation
noted that all the mechanical parameters illustrated in Fig. 1 and de- time τ. Increasing dextran concentration and hence µ, results in a higher
rived here in the strain-rate domain have been demonstrated to be τ because the fluid resists motion.
equivalent to those obtained via frequency domain testing (e.g. dy- The water swollen hydrogels networks cannot be considered in the
namic mechanical analysis, DMA). In particular, frequency-domain same light as the PU. In particular, the aqueous phase is formed of both
storage and loss moduli can be estimated from strain-rate measure- ‘bound water’ and ‘free water’. The latter is defined as the amount of
ments, and – dually – strain rate domain viscoelastic constants (Einst, water that does not interact with the polymeric chains as all the hy-
Eeq, and τs) can be obtained from frequency tests (Bartolini et al., 2018; drophilic and hydrophobic groups have been saturated with the ‘pri-
Zeltmann et al., 2016). mary and secondary bound water’ respectively. As reported by Hoffman
et al. the ‘free water’ is assumed to fill the space between network
chains (Hoffman, 2012), thus it is able to redistribute within the net-
3. Results
work when the gel is deformed. Increasing the dextran concentration in
the liquid phase interferes with hydrogen bonding between agarose
The dynamic viscosity, µ, of the solutions as a function of dextran
chains and water, allowing the latter to move more easily through the
concentration and temperature (16, 21 and 26 °C) is shown in Fig. 2. As
hydrogel molecular pores and redistribute within its network
expected, µ increases with increasing dextran concentration and
(Nijenhuis, 1997). This results in a reduction in compressive stiffness as
well as a decrease in relaxation time because water moves faster as it is
less bound to the network.
PAAm gels are highly hydrophilic but present a low hydrolytic
stability (Dursun et al., 2004). Their gelification occurs through the
chemical crosslinking between acrylamide and bis-acrylamide (N,N0-
methylene bis-acrylamide, in our case), which results in the formation
of a porous network with covalent bonds between polymer chains
(Dursun et al., 2004). In the samples used here, despite the fact that the
instantaneous elastic moduli of the agarose and PAAm gels in the ab-
sence of dextran were fairly similar, the latter gels have a lower τ and
exhibit a higher difference between instantaneous and equilibrium
elastic modulus than the former. It is likely therefore that, even in the
absence of dextran, water molecules are less bound to the PAAm net-
work than they are to agarose, and this effect is amplified with in-
Fig. 2. Dynamic viscosity of the dextran solutions as a function of concentra- creasing amounts of the polysaccharide. As in the case of agarose, our
tion. results suggest that as dextran concentration increases water moves

164
L. Cacopardo et al. Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

Fig. 3. Instantaneous and equilibrium elastic moduli (A) and characteristic relaxation time (B) of sponges as a function of different dextran concentrations. A) * =
p < 0.05 with respect to other values of Eeq; B) * = p < 0.05.

Fig. 4. Instantaneous and equilibrium elastic moduli (A) and characteristic relaxation time (B) of Agarose samples as a function of different dextran concentrations.
* =p < 0.05 with respect to other values of Einst, Eeq and τ respectively.

Fig. 5. Instantaneous and equilibrium elastic moduli (A) and characteristic relaxation time (B) of PAAm samples as a function of different dextran concentrations.
* =p < 0.05.

165
L. Cacopardo et al. Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

Fig. 6. Schematic representation of the interaction between the solid network of PAAm and the liquid phase: without dextran hydrogen bonds are formed between
acrylamide and water. In the presence of dextran, water tends to form the bonds with dextran rather than with acrylamide. Thus, the rate of movement and
redistribution of the aqueous phase increases with increasing dextran concentration (µ0% < µ2% < µ5%). Free water molecules are shown as empty circles, while
bound water ones are represented as filled circles attached to the hydrogel polymer network (solid lines).

more easily through the network resulting in a lower relaxation time. 5. Conclusions
Our results are supported by the findings of Wu and Shanks who studied
how PAAm solutions behave with different glucose concentrations. In To date, most mechanobiology studies have focused on character-
particular, they found that when hydrogen bonds between glucose and izing biological tissue stiffness and investigating cell behaviour on
water overwhelm the bonds between PAAm amides and water, the substrates, typically hydrogels, on the bases of their elasticity.
acrylamide intramolecular interactions become higher than inter- However, both hydrogels and biological tissues are intrinsically vis-
molecular ones with the result that the amount of ‘free water’ increases. coelastic. In light of this, methods to modulate hydrogels viscoelastic
i.e. water is less bound to the chains and thus moves more easily properties have been recently proposed to study cell response to sub-
through the polymeric network (Wu and Shanks, 2003). This concept, strate viscoelasticity. In conclusion, as an alternative to the crosslinking
which is schematised in Fig. 6, can also be extended to dextran, as it is a of the polymer network, we demonstrate that modifying the viscosity of
polysaccharide composed of different glucose molecules (Hammond, the hydrogel liquid phase can also modulate its viscoelastic properties.
1969). The figure illustrates the formation of hydrogen bonds between Notably, this strategy allows changing the mechanical properties re-
the solid network and water in the absence of dextran and the tendency flecting viscoelastic phenomena (i.e. E2 and η, see Fig. 1) while keeping
of water to preferentially form bonds with the polysaccharide rather the equilibrium elastic response (Eeq=E1) constant. This decoupling is
than with PAAm when it is present in the hydrogel liquid phase. generally not achievable when using strategies that target the hydrogel
In unconfined compression experiments Eeq in principle represents polymer network, as they usually alter both the hydrogel viscoelastic
the elastic modulus of the polymer network at equilibrium, i.e. after the dynamics and the equilibrium elastic response. Future studies will be
viscoelastic dynamics have occurred and thus when the viscous com- focused on investigating the contribution of the liquid phase in different
ponent cannot support any load. We hypothesise that Eeq does not vary hydrogels with particular regard to the interaction between the
significantly in the hydrogels because the addition of dextran does not polymer network and the liquid phase, given its critical role in de-
modify the elastic network itself, but only the interaction between the termining the resultant bulk viscoelastic behaviour.
liquid and solid phase. On the other hand, Einst changes according to the
degree of water-polymer binding. Specifically, the instantaneous elastic References
modulus is higher for a network strongly bound to water and lower in
the case of weak water-polymer interactions. This suggests that varying Arnott, S., Fulmer, A., Scott, W.E., Dea, I.C.M., Moorhouse, R., Rees, D.A., 1974. The
the viscosity of the liquid phase in hydrogels by adding components to agarose double helix and its function in agarose gel structure. J. Mol. Biol. 90,
269–284. https://doi.org/10.1016/0022-2836(74)90372-6.
the aqueous solution may alter the interaction between the liquid phase Awad, H.A., Wickham, M.Q., Leddy, H.A., Gimble, J.M., Guilak, F., 2004. Chondrogenic
and the polymeric network and thereby modify the bulk viscoelastic differentiation of adipose-derived adult stem cells in agarose, alginate, and gelatin
response. In our case, dextran reduced this interaction in both agarose scaffolds. Biomaterials 25, 3211–3222. https://doi.org/10.1016/j.biomaterials.2003.
10.045.
and PAAm gels, facilitating liquid phase movement (and hence de- Balgude, A.P., Yu, X., Szymanski, A., Bellamkonda, R.V., 2001. Agarose gel stiffness de-
creasing the relaxation time). On the contrary, when these interactions termines rate of DRG neurite extension in 3D cultures. Biomaterials 22, 1077–1084.
are lower or absent, as in the case of the sponges, increasing the visc- https://doi.org/10.1016/S0142-9612(00)00350-1.
Bartolini, L., Iannuzzi, D., Mattei, G., 2018. Comparison of frequency and strain-rate
osity of the liquid phase results in an increase of τ, since the time re- domain mechanical characterization. Sci. Rep. 8, 13697. https://doi.org/10.1038/
quired for the liquid phase to move is higher. s41598-018-31737-3.
Bhattacharya, S., Srivastava, A., Pal, A., 2006. Modulation of viscoelastic properties of

166
L. Cacopardo et al. Journal of the Mechanical Behavior of Biomedical Materials 89 (2019) 162–167

physical gels by nanoparticle doping: Influence of the nanoparticle capping agent. Science & Business Media, pp. 1–120 doi: 3662107961.
Angew. Chem. - Int. Ed. 45, 2934–2937. https://doi.org/10.1002/anie.200504461. Kumar, S., 2014. Cellular mechanotransduction: stiffness does matter. Nat. Mater. 13,
Cameron, A.R., Frith, J.E., Cooper-White, J.J., 2011. The influence of substrate creep on 918–920. https://doi.org/10.1038/nmat4094.
mesenchymal stem cell behaviour and phenotype. Biomaterials 32, 5979–5993. Liepsch, D., Thurston, G., Lee, M., 1991. Studies of fluids simulating blood-like rheolo-
https://doi.org/10.1016/j.biomaterials.2011.04.003. gical properties and applications in models of arterial branches. Biorheology 28,
Campolo, J., Vozzi, F., Penco, S., Cozzi, L., Caruso, R., Domenici, C., Ahluwalia, A., Rial, 39–52.
M., Marraccini, P., Parodi, O., 2014. Vascular injury post stent implantation: different Mattei, G., Cacopardo, L., Ahluwalia, A., 2017. Micro-mechanical viscoelastic properties
gene expression modulation in human umbilical vein endothelial cells (HUVECs) viscoelastic properties viscoelastic properties of crosslinked hydrogels using the
model. PLoS One 9, e90213. https://doi.org/10.1371/journal.pone.0090213. nano-epsilon dot method (Basel). Materials 10, 889. https://doi.org/10.3390/
Charrier, E.E., Pogoda, K., Wells, R.G., Janmey, P.A., 2018. Control of cell morphology ma10080889.
and differentiation by substrates with independently tunable elasticity and viscous Naessens, M., Cerdobbel, A., Soetaert, W., Vandamme, E.J., 2005. Review Leuconostoc
dissipation. Nat. Commun. 9, 1–13. https://doi.org/10.1038/s41467-018-02906-9. dextransucrase and dextran: production, properties and applications. J. Chem.
Chaudhuri, O., Gu, L., Darnell, M., Klumpers, D., Sidi, A., Weaver, J.C., Huebsch, N., Technol. 80, 845–860. https://doi.org/10.1002/jctb.1322.
Mooney, D.J., Francisco, S., 2015. Substrate stress relaxation regulates cell spreading. Nemir, S., West, J.L., 2009. Synthetic materials in the study of cell response to substrate
Nat. Commun. 6. https://doi.org/10.1038/ncomms7365.Substrate. rigidity. Ann. Biomed. Eng. 38, 2–20. https://doi.org/10.1007/s10439-009-9811-1.
Dursun, S., Erdener, K., Nurettin, S., Olgun, G., 2004. The influence of preparation Nijenhuis, K. te, 1997. Agarose. In: Thermoreversible Networks: Viscoelastic Properties
methods on the swelling and network properties of acrylamide hydrogels with and Structure of Gels. Springer, Berlin, Heidelberg, pp. 194–202. https://doi.org/10.
crosslinkers. J. Macromol. Sci. Part A Pure Appl. Chem. A41, 419–431. https://doi. 1007/BFb0008710.
org/10.1081/MA-120028476. Pek, Y.S., Wan, A.C.A., Ying, J.Y., 2010. The effect of matrix stiffness on mesenchymal
Engler, A.J., Sen, S., Sweeney, H.L., Discher, D.E., 2006a. Matrix elasticity directs stem stem cell differentiation in a 3D thixotropic gel. Biomaterials 31, 385–391. https://
cell lineage specification. Cell 126, 677–689. https://doi.org/10.1016/j.cell.2006.06. doi.org/10.1016/j.biomaterials.2009.09.057.
044. Sabbahi, A., Vergnaud, J.M., 1993. Absorption of water by polyurethane foam. modelling
Engler, A.J., Sen, S., Sweeney, H.L., Discher, D.E., 2006b. Matrix elasticity directs stem and experiments. Eur. Polym. J. 29, 1243–1246. https://doi.org/10.1016/0014-
cell lineage specification. Cell 126.4, 677–689. https://doi.org/10.1016/j.cell.2006. 3057(93)90155-9.
06.044. Tirella, A., Mattei, G., Ahluwalia, A., 2013. Strain rate viscoelastic analysis of soft and
Guo, C., Zhou, L., Lv, J., 2013. Effects of expandable graphite and modified ammonium highly hydrated biomaterials. J. Biomed. Mater. Res. - Part A 102A, 3352–3360.
polyphosphate on the flame-retardant and mechanical properties of wood flour- https://doi.org/10.1002/jbm.a.34914.
polypropylene composites. Polym. Polym. Compos. 21, 449–456. https://doi.org/10. Tse, J.R., Engler, A.J., 2010. Preparation of hydrogel substrates with Tunable mechanical
1002/app. properties. Curr. Protoc. Cell Biol. 1–16. https://doi.org/10.1002/0471143030.
Hammond, B.F., 1969. Dextran production by a human oral strain of Lactobacillus casei. cb1016s47.
Arch. Oral. Biol. 14, 879–890. https://doi.org/10.1016/0003-9969(69)90266-0. Wells, R.G., 2008. The role of matrix stiffness in regulating cell behavior. Hepatology 47,
Hoffman, A.S., 2012. Hydrogels for biomedical applications. Adv. Drug Deliv. Rev. 64, 1394–1400. https://doi.org/10.1002/hep.22193.
18–23. https://doi.org/10.1016/j.addr.2012.09.010. Wu, S., Shanks, R.A., 2003. Conformation of polyacrylamide in aqueous solution with
Humphrey, J.D., Dufresne, E.R., Schwartz, M.A., 2014. Mechanotransduction and extra- interactive additives and cosolvents. J. Appl. Polym. Sci. 89, 3122–3129. https://doi.
cellular matrix homeostasis. Nat. Rev. Mol. Cell Biol. 15, 802–812. https://doi.org/ org/10.1002/app.12528.
10.1038/nrm3896. Yang, M.T., Fu, J., Wang, Y.-K., Desai, R.A., Chen, C.S., 2011. Assaying stem cell me-
Ingber, D.E., 1997. Tensegrity: the architectural basis of cellular mechanotransduction. chanobiology on microfabricated elastomeric substrates with geometrically modu-
Annu. Rev. Physiol. 59, 575–599. https://doi.org/10.1146/annurev.physiol.59.1. lated rigidity. Nat. Protoc. 6. https://doi.org/10.1038/nprot.2010.189.
575. Zaman, M.H., Trapani, L.M., Sieminski, A.L., Mackellar, D., Gong, H., Kamm, R.D., Wells,
Jay, J.I., Shukair, S., Langheinrich, K., Hanson, M.C., Cianci, G.C., Johnson, T.J., Clark, A., Lauffenburger, D.A., Matsudaira, P., 2006. Migration of tumor cells in 3D matrices
M.R., Hope, T.J., Kiser, P.F., 2009. Modulation of viscoelasticity and HIV transport as is governed by matrix stiffness along with cell-matrix adhesion and proteolysis. Proc.
a function of pH in a reversibly crosslinked hydrogel. Adv. Funct. Mater. 19, Natl. Acad. Sci. USA, 103, 15–16. https://doi.org/Doi 10.1073/Pnas.0606087103.
2969–2977. https://doi.org/10.1002/adfm.200900757. Zeltmann, S.E., Bharath Kumar, B.R., Doddamani, M., Gupta, N., 2016. Prediction of
Keplinger, C., Yun Sun, J., Chiang Foo, C., Josef Michael Rothemund, P., Whitesides, strain rate sensitivity of high density polyethylene using integral transform of dy-
G.M., Keplinger, C., Sun, J.-Y., Suo, Z., 2013. Stretchable, transparent, ionic con- namic mechanical analysis data. Polym. (U. Kingd.) 101, 1–6. https://doi.org/10.
ductors. Sci. 341, 984–987. https://doi.org/10.1126/science.1240228. 1016/j.polymer.2016.08.053.
Kulicke, W.M., Clasen, C., 2013. Viscosimetry of polymers and polyelectrolytes. Springer

167