RESEARCH ARTICLE
Proton conductivity of glycosaminoglycans
John Selberg
☯
, Manping Jia
☯
, Marco Rolandi
ID
*
Department of Electrical Engineering, University of California, Santa Cruz, CA, United States of America
☯ These authors contributed equally to this work.
* mrolandi@ucsc.edu
Abstract
Proton conductivity is important in many natural phenomena including oxidative phosphory-
lation in mitochondria and archaea, uncoupling membrane potentials by the antibiotic Gram-
icidin, and proton actuated bioluminescence in dinoflagellate. In all of these phenomena, the
conduction of protons occurs along chains of hydrogen bonds between water and hydro-
philic residues. These chains of hydrogen bonds are also present in many hydrated biopoly-
mers and macromolecule including collagen, keratin, chitosan, and various proteins such as
reflectin. All of these materials are also proton conductors. Recently, our group has discov-
ered that the jelly found in the Ampullae of Lorenzini- shark’s electro-sensing organs- is the
highest naturally occurring proton conducting substance. The jelly has a complex composi-
tion, but we proposed that the conductivity is due to the glycosaminoglycan keratan sulfate
(KS). Here we measure the proton conductivity of hydrated keratan sulfate purified from
Bovine Cornea. PdH
x
contacts at 0.50 ± 0.11 mS cm
-1
, which is consistent to that of Ampul-
lae of Lorenzini jelly at 2 ± 1 mS cm
-1
. Proton conductivity, albeit with lower values, is also
shared by other glycosaminoglycans with similar chemical structures including dermatan
sulfate, chondroitin sulfate A, heparan sulfate, and hyaluronic acid. This observation sup-
ports the relationship between proton conductivity and the chemical structure of
biopolymers.
Introduction
Proton (H
+
) conductivity is important in many natural phenomena[1] including oxidative
phosphorylation in mitochondria and archaea[2–4], uncoupling membrane potentials by the
antibiotic Gramicidin[5], and proton actuated bioluminescence in dinoflagellate[6]. In all of
these phenomena, the conduction of H
+
occurs along chains of hydrogen bonds between
water and hydrophilic residues. These chains are often referred to as proton wires[3]. This
conduction follows the Grotthus mechanism in which a hydrogen bond is exchanged with a
covalent bond contributing to the effective transfer of an H
+
from a molecule to its next-door
neighbor[7]. Following this mechanism, proton conductivity in hydrated biopolymers and
macromolecules is widespread including collagen[8], keratin[9], chitosan[10], melanin[11],
peptides[12], and various proteins such as bovine serum albumin[13] and reflectin[14, 15]. In
addition to the ability to support proton wires, typically these materials include an acid or a
base group that serve as H
+
or OH
-
dopants and provide charge carriers for proton
PLOS ONE | https://doi.org/10.1371/journal.pone.0202713 March 8, 2019 1 / 8
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Selberg J, Jia M, Rolandi M (2019)
Proton conductivity of glycosaminoglycans. PLoS
ONE 14(3): e0202713. https://doi.org/10.1371/
journal.pone.0202713
Editor: Nikolai Lebedev, US Naval Research
Laboratory, UNITED STATES
Received: August 1, 2018
Accepted: February 15, 2019
Published: March 8, 2019
Copyright: © 2019 Selberg et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the manuscript and its Supporting
Information files.
Funding: This work was supported by the National
Science Foundation DMR- 1648815.
Competing interests: The authors have declared
that no competing interests exist.
conductivity [16–18]. Following this trend, for example, the synthetic polymer Nafion, with a
high proton conductivity of 78 mS cm
-1
, contains very strong acid groups that donate H
+
to
the water of hydration for proton conduction [19]. Our group has recently demonstrated that
the jelly contained in the ampullae of Lorenzini, the electrosensing organ of sharks and skates,
is the highest naturally occurring proton conductor[20]. We proposed that keratan sulfate
(KS), a glycosaminoglycan (GAG), was the material responsible for proton conductivity due to
its similar chemical structure to other known proton conductors such as chitosan, and the abil-
ity to form many hydrogen bonds with water when hydrated (Fig 1A)[21, 22]. Given that it is
difficult to purify KS from the shark jelly due to small amounts of sample per organism, we set
to explore KS from different sources that were available to perform these measurements. Here,
we have measured the proton conductivity of KS derived from bovine cornea [23, 24] and
other GAGs using Pd based proton conducting devices [10].
GAGs are long, linear, hydrophilic biopolymers composed of repeating of disaccharide
units with many acidic groups that may support the presence of proton wires (Fig 1B) that
transport protons through the Grotthuss mechanism [25]. Among these are hyaluronic acid
(HA), heparan sulfate (HS), chondroitin sulfate A (CSA), dermatan sulfate (DS), and KS[26,
27]. Additionally, GAGs have important biological functions in regulating hydration and
water homeostasis of tissues, which is derived from their ability to absorb very large amounts
of water at high humidity[28]. They are also implicated in many fundamental operations such
as cell patterning [29], cell signaling, and regulation[30].
Fig 1. The keratan sulfate. (A) Chemical structure of KS. (B) An illustration of a three-monomer segment of KS. Possible intra- and inter-molecular hydrogen bonds as
well as the hydrogen bonds between the water of hydration and the polar parts of the molecule form a continuous network comprised by hydrogen-bond chains. The
sulfate group interacts with the hydrogen-bond network and forms an H
3
O
+
(hydronium) ion.
https://doi.org/10.1371/journal.pone.0202713.g001
Proton conductivity of glycosaminoglycans
PLOS ONE | https://doi.org/10.1371/journal.pone.0202713 March 8, 2019 2 / 8
Materials and methods
Materials
Glycosaminoglycan samples were received from the Linhardt laboratory at Rensselaer Univer-
sity and stored dry at -15C. Including, 70% pure CSA isolated from bovine trachea (average
MW: 20kDa), HA sodium salt from streptococcus zooepidemicus (average MW: 100kDa), DS
from porcine intestinal mucosa (average MW: 30kDa), HS (porcine intestinal mucosa (average
MW: 14.8kDa), and KS isolated from the bovine cornea (average MW: 14.3kDa) a biochemical
description of the KS can be found at Weyers et al.[24].
Device fabrication
Two-terminal measurements were performed on Si substrates with a 0.1μm SiO
2
layer. Con-
ventional photolithography was used to pattern 0.1μm thick Au and Pd contacts. Pd contacts
were 500 μm wide and separated by different channel lengths, L
SD
= 5, 10, 20, 50, 100, 200,
500 μm. We performed both two terminal device measurements and transmission line mea-
surements (TLM) to reduce the influence of contact resistance on the conductivity [11].
Deposition of glycosaminoglycans
All lyophilized samples were rehydrated in DI water (pH 6.7) at a concentration of 0.15–0.2
mg μl
-1
and drop casted onto the devices. The samples were the dehydrated into a film with
dry nitrogen gas flow.
Proton conductivity measurements
Direct current–resistance measurements were performed using a Keithley 4200 source-meter
and a two-contact probe station arrangement on devices. The devices were enclosed in an
environmental chamber at room temperature in an atmosphere of nitrogen or hydrogen with
controlled relative humidity (RH). We controlled RH by bubbling gases through a bubbler
containing DI water at pH 6.7. Hydrated in sequence from dry to 75%RH in N
2
, 90%RH in
N
2
, 90% RH in a mixture of 95% N
2
with 5% hydrogen, and 90% RH in a mixture of 95% N
2
with 5% deuterium gas to form PdH
x
or PdD
x
contacts. A one-hour incubation period was
carried out after switching between humidity and gas compositions. During the measurement,
the Pd/PdH
x
electrodes were contacted with tungsten probes. When we applied a source-drain
potential difference, V
SD
, the PdH
x
source injected protons (H
+
) into drain through the sam-
ples, inducing measurable electrical current in the circuit.
Results and discussion
Proton conductivity measurements
Palladium (Pd) devices are useful for studying proton transport in materials due to the nature
of Pd to reversibly form palladium hydride (PdH
x
)[31–34]. Several mechanisms for the forma-
tion of PdH
x
are known (Eqs 1–4).
H
2
þ Pd ! 2PdH
ads
ð1Þ
Equation one describes the adsorption and splitting of H
2
molecules into two adsorbed H
on the Pd metal surface without electron transfer in a reaction described by Tafel kinetics.
H
2
þ Pd ! PdH
ads
þ H
þ
þ e
ð2Þ
Equation two is the Heyrovsky reaction in which a H
2
is split into an adsorbed H atom and
Proton conductivity of glycosaminoglycans
PLOS ONE | https://doi.org/10.1371/journal.pone.0202713 March 8, 2019 3 / 8
a H
+
, e
-
pair at the Pd surface, this e
-
is transferred into the metal.
H
þ
þ Pd þ e
$ PdH
ads
ð3Þ
The Volmer reaction in Eq 3 describes a third mechanism, which involves an electron
transfer to a H
+
near the Pd surface allowing it to adsorb as PdH
ads
. Once PdH
ads
is formed on
the metal surface, H can diffuse into the subsurface bulk forming PdH
x
(Eq 4). [10, 35, 36].
PdH
ads
$ PdH
x
ð4Þ
Pd devices were designed such that PdH
x
formation occurs spontaneously by Eq 1 in a 5%
H
2
atmosphere on two Pd contacts. These Pd/PdH
x
contacts are separated by a channel con-
sisting of a GAG film which completes the circuit (Fig 2A and 2B). A voltage V
SD
between the
Pd/PdH
x
contacts induces a current of H
+
to exit one Pd contact, travel through the film chan-
nel, and enter the second Pd contact according to Eq 3. In this manner, one e
-
travels through
the circuit and is recorded as I
d
for each H
+
that is conducted through the channel. Consider-
ing the conductivity of the GAG films is expected to be much less than the conductivity of elec-
trons in electrodes, the current that we measure during the experiments is indicative of the
conductivity of the channel.
Fig 2. Proton conduction measurement of KS. A) Palladium hydride(PdH
x
) electrode behavior. Under a V
SD
, PdH
x
source split into Pd, H
+
, and e
−
. Protons are
injected into the KS, whereas electrons travel through external circuitry and are measured. B) TLM geometry. Varying the distance between source and drain (L
SD
)
distinguishes between the fixed PdH
x
−KS interface contact resistance and the varying bulk resistance. C) Optical image of TLM geometry with hydrated KS on the
surface. Scale bar, 500μm. D) Transient response to a 1V bias in KS at 75%, 90%, 90% H
2
RH, in which the current under 90% with hydrogen is much higher than that
under 90% RH without hydrogen. E) Deuterium current (black) at 90% D
2
humidity is lower than proton current (red). F) The normalized resistance R
LN
as a function
of L
SD
, A linear fit gives a bulk material proton conductivity of 0.50 ± 0.11 mS cm
-1
.
https://doi.org/10.1371/journal.pone.0202713.g002
Proton conductivity of glycosaminoglycans
PLOS ONE | https://doi.org/10.1371/journal.pone.0202713 March 8, 2019 4 / 8
Materials characteristics
After deposited directly onto the transmission line measurement (TLM) (Fig 2C) device sur-
face without further processing, the KS film is thick, viscous, and optically transparent. After
one hour of incubating at 50%RH, the KS film dries to a non-homogenous film. The film rehy-
drates fully after incubating at 90%RH for one hour and appears as wet as when it was drop-
cast form solution (Fig 2C). This high water content of KS films is a result of sulfate groups
functionalizing either or both of the galactose and N-acetyl glucosamine sugars which make
up the repeating disaccharide unit of the GAGs. Considering the other members of GAGs fam-
ily, DS, HS, CSA, and HA also contain an abundance of repeating acidic groups which may
stabilize proton wires, as shown in Table in S1 File.
DC electrical measurements with PdH
x
proton-conducting contacts
With V
SD
= 1V on the Pd devices, we measured the drain current (I
D
) of KS, as shown in Fig
2D. First, at 75% RH in N
2
, I
D
(~ 0.5 nA) is small (black in Fig 2D). With the RH increased to
90% in N
2
, the increase in I
D
was negligible (red in Fig 2D). However, after we changed the gas
to 95%N
2
+ 5%H
2
, the I
D
increased more than 300 times to 155 nA (green in Fig 2D). The
same measurements were performed with DS, HS, CSA, and HA family and followed similar
trends (Figure A in S1 File). All GAGs displayed an increased current upon a 90%RH (5%H
2
)
atmosphere compared to a 90%RH N
2
atmosphere, indicating that protons predominately
contribute to the conductivity of GAGs materials at high relative humidity.
Kinetic isotope effect
To further test whether KS conductivity predominantly arises from protons, we investigated
the kinetic isotope effect. Measurements were repeated while hydrating the sample with deute-
rium oxide (D
2
O) instead of water and exposing the sample to deuterium gas rather than H
2
.
Like protons, deuterium ions (D
+
) can transport along proton wires and hydrated materials,
albeit with a lower mobility and an associated lower current due to the higher molecular
weight and higher binding energy during H-bonding[37]. The kinetic isotope effect in KS is
evident as a drop in the conductivity when deuterium replaces hydrogen as the atom being
transported (Fig 2E). Here, we observe a 15% drop in current when deuterium replaces hydro-
gen. The kinetic isotope effect observed with KS is relatively small. However, a similar small
kinetic isotope effect was observed for the proton conduction of bovine serum albumin[13].
The other members in GAGs family display a larger kinetic isotope effect, the current drop is
nearly 50% (Figure B in S1 File). The divergence of the KIE between the KS films and the other
GAGs may be due to regions different transport regimes for H
+
in KS films. Where the bind-
ing energy plays a big role in H-bond mediated transport by the Grotthuss mechanism it will
not be as noticeable by regions of bulk diffusion.
Transmission line measurement
TLM devices are designed with different lengths between the Pd source and the drain contacts
to eliminate the effect of contact resistance in the measurements of the proton conductivity
(Fig 2B) [20]. We applied V
SD
= 1 V on devices with L
SD
ranging from 5 to 500 um, measured
I
D
, and calculated the resistance of each device, R
L
. In this geometry, R
L
increases linearly with
L
SD
, but the contact resistance, R
C
, at the source–KS and drain–KS interface is constant. Con-
sidering that different devices contained KS with different thicknesses, we multiplied R
L
by the
sample thickness to get the normalized resistance, R
LN
. The slope of the plot of R
LN
as a func-
tion of L
SD
is proportional to the resistivity of KS, and the intercept on the R
LN
axis for L
SD
= 0
Proton conductivity of glycosaminoglycans
PLOS ONE | https://doi.org/10.1371/journal.pone.0202713 March 8, 2019 5 / 8
