You are on page 1of 158

1

Studies on Extraction of Cyanobacterial


Polysaccharides (sacran) from
Aphanothece sacrum
and its Structures and Properties





A dissertation for the degree

DOCTOR OF ENGINEERING

by


Maiko Okajima-Kaneko

Department of Organic and Polymeric Materials
Tokyo Institute of Technology

2009
2

CONTENTS
CHAPTER-1- 4
General Introduction
1-1: Introduction to cyanobacteria
1-2: Introduction to polysaccharides
1-3: Introduction to polysaccharides derived from cyanobacteria
1-4: Conclusion

CHAPTER-2- 20
Extraction of polysaccharides from Aphanothece sacrum
biomaterials
2-1: Introduction
2-2: History and overviews of Aphanothece sacrum
2-3: Observation of Aphanothece sacrum biomaterials
2-4: Optimization of polysaccharides extraction process from Aphanothece sacrum
2-5: Conclusion

CHAPTER-3- 43
Structural analyses of polysaccharides extracted from Aphanothece
sacrum
3-1: Introduction
3-2: Basic analyses of primary structures
3-3: Multi-step acid hydrolysis and determination of constituent sugars.
3-4: Higher-order structures of extracted polysaccharides
3-5: Conclusion
3

CHAPTER-4- 77
Solution properties and liquid crystalline natures of polysaccharides
extracted from Aphanothece sacrum
4-1: Introduction
4-2: Solution properties of extracted polysaccharides
4-3: Liquid crystalline nature of extracted polysaccharides
4-4: Conclusion

CHAPTER-5- 118
Metal ion sorption of extracted polysaccharides and its hydrogels
5-1: Introduction
5-2: Gel bead formation of extracted polysaccharides with various metal ions
5-3: Metal ion sorption of chemically cross-linked sacran gels
5-4: Metal ion sorption of semi-IPN PVA gels containing polysaccharides
5-5: Conclusion

CHAPTER-6- 150
Conclusive Remarks

Acknowledgements 153
Published Papers 156
4

-CHAPTER 1-

GENERAL INTRODUCTION


1-1: Introduction to cyanobacteria
Cyanobacteria are photoautotrophic prokaryotes which include a large variety of species
of widespread occurrence and with diverse morphological, physiological and biochemical
properties. Cyanobacteria, which are similar to bacteria such as Escherichia coli., have
fixed CO
2
for 3.5 billion years longer than plants on the earth. Cyanobacteria, which are
known as blue-green algae, blue-green bacteria or cyanophylla, are a phylum of bacteria
that obtain their energy through photosynthesis. The name "cyanobacteria" comes from the
color of the bacteria. They are a significant component of the marine nitrogen cycle and an
important primary producer in many areas of the ocean, but are also found in habitats other
than the marine environment; in particular cyanobacteria are known to occur in both
freshwater [1,2] and hypersaline inland lakes [3]. Stromatolites of fossilized
oxygen-producing cyanobacteria have been found from 2.8 billion years ago [4]. The
ability of cyanobacteria to perform oxygenic photosynthesis is thought to have converted
the early reducing atmosphere into an oxidizing one, which dramatically changed the
composition of life forms on Earth by provoking an explosion of biodiversity and leading
to the near-extinction of oxygen-intolerant organisms. Chloroplasts in plants and eukaryotic
algae have evolved from cyanobacteria via endosymbiosis.
Cyanobacteria are found in almost every conceivable environment, from oceans to fresh
water to bare rock to soil. Most are found in fresh water, while others are in marine, occur
in damp soil, or even temporarily moistened rocks in deserts. A few are endosymbionts in
lichens, plants, various protists, or sponges and provide energy for the host. Cyanobacteria
include unicellular and colonial species. Colonies may form filaments, sheets or even
hollow balls. Some filamentous colonies show the ability to differentiate into several
different cell types: vegetative cells, the normal, photosynthetic cells that are formed under
favorable growing conditions; akinetes, the climate-resistant spores that may form when
5

environmental conditions become harsh; and thick-walled heterocysts, which contain the
enzyme nitrogenase, vital for nitrogen fixation. Heterocyst may also form under the
appropriate environmental conditions (anoxic) wherever nitrogen is necessary.
Heterocyst-forming species are specialized for nitrogen fixation and are able to fix nitrogen
gas, which cannot be used by plants, into ammonia (NH
3
), nitrites (NO
2
-
) or nitrates (NO
3
-
),
which can be absorbed by plants and converted to protein and nucleic acids. Many
cyanobacteria also form motile filaments, called hormogonia, that travel away from the
main biomass to bud and form new colonies elsewhere. The cells in a hormogonium are
often thinner than in the vegetative state, and the cells on either end of the motile chain may
be tapered. In order to break away from the parent colony, a hormogonium often must tear
apart a weaker cell in a filament, called a necridium. Each individual cell of a
cyanobacterium typically has a thick, gelatinous cell wall. They differ from other
gram-negative bacteria in that the quorum sensing molecules autoinducer-2 [5] and
acyl-homoserine lactones [6] are absent. They lack flagella, but hormogonia and some
unicellular species may move about by gliding along surfaces. In water columns some
cyanobacteria float by forming gas vesicles, like in archaea. Some of these organisms
contribute significantly to global ecology and the oxygen cycle. The tiny marine
cyanobacterium Prochlorococcus was discovered in 1986 and accounts for more than half
of the photosynthesis of the open ocean [7].
Finally I introduce some representative cyanobacteria with their pictures. Photo 1 shows
a figure of subspecific Aphanothece stagnina in a rice field of Vietnam countryside. The
species were observed widely in tropical area. Aphanothece sacrum (Suizenzinori; main
subject of this thesis) was misclassified into the species before Suringers finding. Photo 2
shows a figure for a species of J apan-indigenous cyanobacterium living in Hokuriku area,
which is called Ashitsuki. The name of Ashitsuki is used in Manyoushu long ago. However
this is very rare similarly with Suizenjinori. Photo 3 shows a figure of a cyanobacteria
living in a hot spring of Venezia countryside (Italy). The diatomite hybrids with their
metabolites such as glycolipids have been traditionally used for repairing the joint pain in
the area and now widely used in Europe. Photo 4 shows a figure of most popular
cyanobacteria living on the ground worldwide. This cyanobacteria is famous for tolerate in
dry environments.


6


























Photo 1. Aphanothece stagnina in a rice field Photo 2. Nostoc verrucosum in a river
(Tien kien, Phu Tho, Veitnam) (Ashitsuki) (Toga river, Nanto, J apan)

Photo 3. Phormidium sp. in a hot spring Photo 4. Nostoc commune (Ishikurage) on the
(Presidentterme, Padova, Italy) ground (J AIST campus, Nomi, J apan)

7

1-2: Introduction to polysaccharides
1-2-1: Polysaccharides
Polysaccharides are relatively complex carbohydrates. They are polymers made up of
many monosaccharides joined together by glycosidic covalent bonds to be very large, and
are often branched. Most of them are amorphous and insoluble in water, and have no sweet
taste [8]. When all the monosaccharides in a polysaccharide are the same type the
polysaccharide is called a homopolysaccharide, but when more than one type of
monosaccharide is present they are called heteropolysaccharides. Examples include
storage polysaccharides such as starch and glycogen and structural polysaccharides such as
cellulose and chitin. Polysaccharides have a general formula of C
x
(H
2
O)
y
where x is usually
a large number more than 200. Considering that the repeating units in the polymer
backbone are often six-carbon monosaccharides, the general formula can also be
represented as (C
6
H
10
O
5
)
n
where n=40-100000. Starches are glucose polymers in which
glucopyranose units are bonded by alpha-linkages. It is made up of a mixture of Amylose
(15-20%) and Amylopectin (80-85%). Amylose consists of a linear chain of several
hundred glucose molecules while Amylopectin is a branched molecule made of several
thousand glucose units (every chain 24-30 glucose unit). Starches are insoluble in water but
can absorb water. They can be digested by hydrolysis, catalyzed by enzymes called
amylases, which can break the alpha-linkages (glycosidic bonds). Humans and other
animals have amylases, and then they can digest starches. Potato, rice, wheat, and maize are
major sources of starch in the human diet. The formation of starches is the way that plants
store glucose. Glycogen is a polysaccharide that is found in animals and is composed of a
branched chain of glucose residues. It is stored in liver and muscles. The structural
components of plants are mainly cellulose. Woods are largely composed of cellulose and
lignin, while paper and cotton are nearly pure cellulose. Cellulose is a polymer made with
repeated glucose units bonded together by beta-linkages. Humans and many other animals
lack an enzyme to break the beta-linkages, and then they do not digest cellulose. Certain
animals can digest cellulose, because bacteria possessing the enzyme to hydrolyze the
beta-linkages are present in their gut.

8


Structure of amylase


Structure of amylopectin


Structure of cellulose



1-2-2
Ba
pepti
comp
impo
bacte
lung)
is ma
polym
bene
new
appli

2: Bacteria
acterial pol
idoglycan a
pounds wh
ortant virule
erium to su
). Polysacch
ade under th
mer modific
fits and the
antibiotics
ications [9-
l polysacch
lysaccharide
and lipopo
hose functi
ence factor
urvive in ha
haride biosy
he subtle int
cation and s
eir biosynth
s and vac
10].
Xanthan g
harides
es represen
lysaccharid
ons range
rs (e.g. poly
arsh environ
ynthesis is
terplay betw
synthesis. B
eses should
ccines) and
gum
9
nt a diverse
des as caps
from stru
y(N-acetylg
nments (e.g
a tightly re
ween energy
Bacterial pol
d enable to d
d the com

Pullua
e range of
sules and
uctural cell
glucosamine
g. Pseudomo
gulated and
y regulation
lysaccharide
develop nov
mmercial ex

n
f macromol
exopolysac
l-wall com
e) in S. au
onas aerug
d energy int
n and conser
es show the
vel antibact
xploitation
Dextran
lecules that
charides. T
mponents, a
ureus) to pe
ginosa in th
tensive proc
rvation and
e enormous
terial strateg
to develo
n

t include
They are
and have
ermit the
he human
cess, and
d between
potential
gies (e.g.
op novel

10

Pathogenic bacteria commonly produce a thick, mucous-like, layer of polysaccharide.
This "capsule" cloaks antigenic proteins on the bacterial surface that would otherwise
provoke an immune response and thereby lead to the destruction of the bacteria. Capsular
polysaccharides are water soluble, commonly acidic, and have molecular weights on the
order of 100-1000 kDa. They are linear and consist of regularly repeating subunits of
one-six monosaccharides. There is enormous structural diversity; nearly two hundred
different polysaccharides are produced by E. coli alone. Mixtures of capsular
polysaccharides, either conjugated or native are used as vaccines. Bacteria and many
other microbes including fungi and algae often secrete polysaccharides as an evolutionary
adaptation to help them adhere to surfaces and to prevent them from drying out. Humans
have developed some of these polysaccharides into useful products, including xanthan gum,
dextran, gellan gum, and pullulan.
Gellan gum
Cell-surface polysaccharides play diverse roles in bacterial ecology and physiology.
They serve as a barrier between the cell wall and the environment, mediate host-pathogen
interactions, and form structural components of biofilms. These polysaccharides are
synthesized from nucleotide-activated precursors (called nucleotide sugars) and, in most
cases, all the enzymes necessary for biosynthesis, assembly and transport of the completed
polymer are encoded by genes organized in dedicated clusters within the genome of the
organism. Lipopolysaccharide is one of the most important cell-surface polysaccharides, as
it plays a key structural role in outer membrane integrity, as well as being an important
mediator of host-pathogen interactions. The enzymes that make the A-band
(homopolymeric) and B-band (heteropolymeric) O-antigens have been identified and the
metabolic pathways defined [11]. The exopolysaccharide alginate is a linear copolymer of
11

-1,4-linked D-mannuronic acid and L-guluronic acid residues, and is responsible for the
mucoid phenotype of late-stage cystic fibrosis disease. The pel and psl loci are two recently
discovered gene clusters that also encode exopolysaccharides found to be important for
biofilm formation. Rhamnolipid is a biosurfactant whose production is tightly regulated at
the transcriptional level, but the precise role that it plays in disease is not well understood at
present. Protein glycosylation, particularly of pilin and flagellin, is a recent focus of
research by several groups and it has been shown to be important for adhesion and invasion
during bacterial infection [12].



1-3: Introduction to polysaccharides derived from cyanobacteria
1-3-1: Cyanobacterial polysaccharides
Many cyanobacteria are known to be able to synthesize outermost slimy investments and
to release polysaccharide material into the culture medium during cell growth. These
released polysaccharides (RPSs), being easily recoverable from the culture medium, are
attracting much interest in view of their possibilities in several industrial applications. An
overview of the current knowledge on both RPS-producing cyanobacterial strains
(including the possible roles of the exopolysaccharides) and chemical characteristics of the
cyanobacterial RPSs is given, with particular emphasis on RPS properties and possible
industrial applications. On the whole, cyanobacterial RPSs are characterized by a great
variety in both number (from two to 10) and type of constitutive monosaccharides (various
arrangements of acidic and neutral sugars). Most polymers show an anionic nature due to
the presence of uronic acids and/or other charged groups such as pyruvyl or sulfate [13].
Polypeptide moieties as well as acetyl substituents have also sometimes been found,
causing additional structural complexity. All the cyanobacterial RPSs so far tested showed
a pseudoplastic behavior, but with marked differences in both viscosity values and shear
thinning. In terms of RPS production, the responses of cyanobacteria to changes of culture
conditions appear strain-dependent. RPS productivities shown by some cyanobacteria are
well comparable with those reported for other photosynthetic microorganisms proposed for
polysaccharide production, but very low in comparison with those of heterotrophic
microorganisms. Nevertheless, cyanobacteria may be regarded as a very abundant source of
structurally diverse polysaccharides, some of which may possess unique properties for
special applications, not fulfilled by the polymers currently available. However, much work
12

has still to be done to bridge the wide gap existing between data on the biology of the
RPS-producer strains and information concerning technological and other useful properties
of the cyanobacterial RPS.
Since the early 1950s, more than one hundred cyanobacterial strains, belonging to twenty
different genera, have been investigated with regard to the production and the released
exocellular RPS into the culture medium. The chemical and rheological properties show
that such polysaccharides are complex anionic heteropolymers, in about 80% cases
containing six to ten different monosaccharides and in about 90% cases containing one or
more uronic acids; almost all have non-saccharide components such as peptidic moieties,
acetyl, pyruvyl and/or sulfate groups. Based on such ingredients, cyanobacterial RPSs show
promise as thickening or suspending agents, emulsifying or cation-chelating compounds
and the residual capsulated cyanobacterial biomass, following RPS extraction, could be an
effective cation-chelating material. Indeed, when eleven unicellular and filamentous
RPS-producing cyanobacteria, selected on the basis of the anion density of their RPSs and
on the abundance of their outermost investments, were screened for their ability to remove
Cu
2+
from aqueous solutions, a quick and most effective heavy metal adsorption was
observed for the unicellular Cyanothece CE 4 and the filamentous Cyanospira capsulata.
These results suggest the possibility to accomplish, through the exploitation of
RPS-producing cyanobacteria, a multiproduct strategy to procure a wide range of
biopolymers suited to various industrial applications, in addition to the residual biomass
effective in the recovery of heavy metals from polluted waters [14,15].

1-3-2: Polysaccharides produced from cyanobacteria wearing solid jelly matrix
Most of cyanobacteria wearing jellylike matrix belong to Nostoc species which is very
popular from researchers to general people. The cyanobacterium Nostoc commune is
representative cyanobacteria which producing and secreting polysaccharide to solid
jellylike matrix, is have been studying widely by many researchers and adapted to the
terrestrial environment and has a cosmopolitan distribution. The role of extracellular
polysaccharides (EPS) in the desiccation tolerance of photosynthesis in N. commune was
examined. Although photosynthetic O
2
evolution was not detected in desiccated colonies,
the ability of the cells to evolve O
2
rapidly recovered after rehydration. The air-dried
colonies contained approximately 10% (wt/wt) water, and field-isolated, natural colonies
with EPS were highly water absorbent and were rapidly hydrated by atmospheric moisture.
13

The cells embedded in EPS in Nostoc colonies were highly desiccation tolerant, and O
2

evolution was not damaged by air drying. Although N. commune was determined to be a
mesophilic cyanobacterium, the cells with EPS were heat tolerant in a desiccated state. EPS
could be removed from cells by homogenizing colonies with a blender and filtering with
coarse filter paper. This treatment to remove EPS did not damage Nostoc cells or their
ability to evolve O
2
, but O
2

evolution was significantly damaged by desiccation treatment
of the EPS-depleted cells. Similar to the EPS-depleted cells, the laboratory culture strain
KU002 had only small amount of EPS and was highly sensitive to desiccation. In the
EPS-depleted cells, O
2

evolution was also sensitive to freeze-thaw treatment. These results
strongly suggest that EPS of N. commune is crucial for the stress tolerance of
photosynthesis during desiccation and during freezing and thawing. N. commune can
survive in a dry, metabolically inactive state, sometimes for long periods [16,17]. Such
aged and desiccated cells rapidly recover their physiological capacities following the
addition of water and then resume active growth. There is some understanding of the
mechanisms that permit cells to withstand extreme fluctuations in water availability,
including knowledge of factors that lead to the damaging of cell components, as well as a
knowledge of repair processes [18-23]. Despite much study of the different strategies used
to overcome acute water deficit, it is still unclear how physiological responses to drying
and rehydration are controlled at the whole-cell level or how a complement of gene
products could interact synergistically, through four dimensions, to provide desiccation
tolerance. The mechanism is that colonies of N. commune accumulate in environments
where water availability is intermittent. For example, many regions of the tropics and
subtropics have distinct dry and wet seasons that last for several months each and
sometimes longer. Typically, wetting events are initially erratic at the onset of the wet
season, and drying events are initially erratic at the end of the wet season [15]. The finding
that the rate of increase in the weight of colonies becomes higher with subsequent cycles of
wetting and drying may be of ecological relevance. Slow water uptake at the onset of the
wet season (after a prolonged drought) would prevent cells from becoming overcommitted,
in a metabolic sense, until there was a more regular availability of water. After such time,
very rapid water uptake would ensure that cells were able to utilize even transient rainfall
most efficiently. The role of gene expression in the process of physical loss of water from
colonies is unknown at this time. The data obtained by using inhibitors of transcription and
translation suggest that water flux in the extracellular milieu of N. commune colonies is
14

coupled to, and possibly regulated at, the level of gene expression. These findings
emphasize the complexities of desiccation tolerance even in simple organisms. A purified
fraction (>12 kDa) of an aqueous extract of the glycan from desiccated field material
contained glucose, Nacetylglucosamine, glucosamine, mannose, and galactosamine with
ratios of 3.1 : 1.4 : 1 : 0.1 : 0.06, respectively. Lipid soluble extracts of N. commune
contained trehalose and sucrose and the levels of both became undetectable following cell
rehydration. Intracellular cyanobacterial trehalase was identified using immunoblotting and
its synthesis was detected upon rehydration of desiccated field cultures. Elemental analysis
of glycan extracts showed a flux in the concentrations of salts in the glycan matrix
following rehydration of desiccated colonies. The role of the glycan in the desiccation
tolerance of N. commune is discussed with respect to structure/function relationships [24].
The extracellular glycan of N. commune is abundant, it is both structurally- and
chemically-complex, it effectively isolates the cells within the colony from their immediate
environment, and it undergoes physical and biochemical changes some of which are
marked and others subtle, in response to the principal environmental variable of water
availability. Sucrose and trehalose were both present in the non-polar extracts of desiccated
field materials of N. commune, and were present, albeit at much reduced levels, in
desiccated. The time of disappearance of trehalose following rehydration of desiccated
colonies matched the time at which a putative trehalase was detected in cell extracts using
Western blotting. No trace of the protein trehalase or the sugar trehalose was found in any
of the aqueous extracts. This is consistent with the fact that the enzyme trehalase is an
intracellular enzyme [25]. It is clear that there is a correlation between the disappearance of
the sugar trehalose and the appearance of the enzyme trehalase upon rehydration of
desiccated field material of N. commune, however, the significance of this to desiccation
tolerance in cyanobacteria is unknown at this time. In this discussion, we explain of
functions of the glycan. One principal function of the glycan is that it provides a repository
for water. The glycan represents a mixed system, where water and the polysaccharide tend
to mix as thoroughly as they can for thermodynamic reasons. Although microorganisms are
certainly present at the surfaces of the N. commune colonies, the outer silicon-rich layer
represents an impenetrable barrier for them. The silicon-rich layer must be made through
physico-chemical precipitation as it is hard to account for a concerted synthesis of this layer
on behalf of the cells. More likely the layer is the product of some oxygen/drying
dependent effect on the peripheral sheath, although in liquid cultures of N.commune strain
15

DRH 1 a discrete pellicular structure is also seen. The glycan represents the bulk of the
colony and constitutes a considerable diversion of the carbon and nitrogen budget. The
extracellular glycan appears to represent a buffer zone between the atmosphere and the
cells. The prodigious investments made in sheath synthesis and those components found
within the sheath, and our interpretation of structure and composition reported here suggest
a principal role for the glycan. It is a central component of the mechanisms used by N.
commune to tolerate desiccation. These mechanisms will be uncovered through
understanding the mode of synthesis of the glycan and of those components, such as
water-absorption and UV-absorbing pigments, present within it.[26-34].
Thus Nostoc commune is a representative cyanobacteria wearing jellylike matrix and
well studied. However the mass-cultivation method of the cyanobacteria has not been
developed yet, which prevent developing industrially. On the other hand, Aphanothece
sacrum is a representative cyanobacteria wearing jellylike matrix which has been already
mass-cultured in a natural river (Kogane river in Asakura, Fukuoka). However the
polysaccharide of A. sacrum is not studied widely for the following reasons; because A.
sacrum is not available worldwide for its ease to collapse during transportation, and
because the control of cultivation is too difficult to study it by the biological researchers.
In other viewpoint of science, I speculated that PS of A. sacrum, which has been
mass-cultivated in J apan for a long time, may have potential as the novel functional
biomaterials. I also detected that the usefulness and high potential of A. sacrum as a
biomass-producing microorganism. A. sacrum is still quite important for the J apanese
researchers, especially biomaterial scientists. The finding motivated me to study
Aphanothece sacrum-derived biomass and I first focused its polysaccharides in terms of
huge jellylike matrixes. Finally I exaggerate that this is a first challenge to extract
Aphanothece sacrum polysaccharides and to materialize them in the world.








16

References
[1] A. Oren, A proposal for further integration of the cyanobacteria under the
Bacteriological Code. Int. J. Syst. Evol. Microbiol. 54, 18951902 (2004).
[2] B. Betsey, D. Dyer, A Field Guide to Bacteria, Cornell University Press (2003).
[3] C. M. Hogan, Makgadikgadi, The Megalithic Portal, ed. A. Burnham (2008)
[4] J . M. Olson, Photosynthesis in the Archean era. Photosyn. Res. 88, 109-117 (2006).
[5] J . Sun, R. Daniel, I. W. Dbler, A. -P. Zeng Is autoinducer-2 a universal signal for
interspecies communication? A comparative genomic and phylogenetic analysis of the
synthesis and signal transduction pathways, BMC Evol. Biol. 4, 36 (2004).
[6] E. Dittmann, M. Erhard, M. Kaebernick, C. Scheler, B. A. Neilan, H. von Dohren, T.
Borner, Altered expression of two light-dependent genes in a microcystin-lacking mutant of
Microcystis aeruginosa PCC7806, Microbiology 147, 31133119 (2001).
[7] S. Nadis, The Cells That Rule the Seas, Sci. Amer., 289, 52-53 (2003).
[8] N. A. Campbell, B. Williamson, R. J . Heyden Biology: Exploring Life. Boston,
Massachusetts: Pearson Prentice Hall (2006).
[9] M. Ullrich (ed.) Bacterial Polysaccharides: Current Innovations and Future Trends.
Caister Academic Press (2009).
[10] Rehm BHA (ed.). Microbial Production of Biopolymers and Polymer Precursors:
Applications and Perspectives. Caister Academic Press (2009).
[11] H. Guo, W. Yi, J . K.Song, P. G. Wang Current understanding on biosynthesis of
microbial polysaccharides, Curr Top Med Chem 8, 141151 (2008).
[12] P. Cornelis (ed.). Pseudomonas: Genomics and Molecular Biology (1st ed.). Caister
Academic Press (2008)..
[13] Y. Tamaru, Y. Takani, T. Yoshida, T. Sakamoto, Crucial Role of Extracellular
17

Polysaccharides in Desiccation and Freezing Tolerance in the Terrestrial Cyanobacterium
Nostoc commune. Appl. Environ. Microbiol. 73277333 (2005).
[14] D. R. Hill, T. W. Keenan, R. F. Helm, M. Potts, L. M. Crowe, J . H. Crowe.
Extracellular polysaccharide of Nostoc commune (Cyanobacteria) inhibits fusion of
membrane vesicles during desiccation. J. Appl. Phycol. 9, 237-248 (1997).
[15] M. Potts, In B. A. Whitton, M. Potts (ed.), The ecology of cyanobacteria: their
diversity in time and space. Kluwer Academic. Publishers, Dordrecht, The Netherlands.
p.465504 (2000).
[16] R. D. Philippis, M. Vincenzini, Exocellular polysaccharides from cyanobacteria and
their possible applications, FEMS Microbiol. Rev. 22, 151-175 (1998).
[17] R. D. Philippis, C. Sili, R. Paperi, M. Vincenzini, Exopolysaccharide-producing
cyanobacteria and their possible exploitation: A review, J. Appl. Phycol. 13, 293-299
(2001).
[18] J . R. Battista, A. M. Earl, M. J . Park, Why is Deinococcus radiodurans so resistant to
ionizing radiation? Trends Microbiol. 7, 362365 (1999).
[19] D. Billi, D. J . Wright, R. F. Helm, T. Prickett, M. Potts, J . H. Crowe, Engineering
desiccation tolerance in Escherichia coli. Appl. Environ. Microbiol. 66, 16801684. (2000)
[20] F. Bloom, P. Price, G. Lao, J . L. Xia, J . H. Crowe, J . R. Battista, R. F. Helm, S.
Slaughter, M. Potts, Engineering mammalian cells for solidstate sensor applications.
Biosens. Bioelectron. 16, 603608 (2001).
[21] J . H. Crowe, F. A. Hoekstra, L. M. Crowe, Anhydrobiosis, Annu. Rev. Physiol. 54,
579599 (1992).
[22] R. F. Helm, Z. Huang, D. Edwards, H. Leeson, W. Peery, M. Potts, Structural
characterization of the released polysaccharide of desiccationtolerant Nostoc commune
18

DRH-1, J. Bacteriol. 182, 974982 (2000).
[23] B. Shirkey, D. P. Kovarcik, D. J . Wright, G. Wilmoth, T. F. Prickett, R. F. Helm, E. M.
Gregory, M. Potts, Active Fe-containing superoxide dismutase and abundant sodF mRNA
in Nostoc commune (cyanobacteria) after years of desiccation, J. Bacteriol. 182, 189-197
(2000).
[24] R. H. Donna, A. Peat, M. Potts, Biochemistry and structure of the glycan secreted by
desiccation-tolerant Nostoc commune (Cyanobacteria), Protoplasma 182, 126-148 (1994).
[25] W. Boos, U. Ehmann, E. Bremer, A. Middendorf, P. Postma, Trehalase of Escherichia
eoli, J. Biol. Chem. 262, 13212-13218 (1986).
[26] M. J . Burke, The glassy state and survival of anhydrous biological systems, A.C.
Leopold (ed.) Membranes, metabolism and dry organisms, Cornell University Press, Ithaca,
p.358-363 (1987).
[27] H. Barbosa, R. Alterthum, The role of extracellular polysaccharide in cell viability and
nitrogenase activity of Beijerinckia derxii. Can J Microbiol A 38, 986-988 (1992).
[28] G. BazzicheIIi, N. Abdelahad, F. Ventola, Structural modifications in the extracellular
investment of Nostoc commune Vauch. during the life cycle. I. Motile and nonmotile
hoimogonium, biseriate stage, J Uitrastruct Res 91, 174-181 (1985).
[29] G. M. Caiola, R. Ocampo-Friedmann, E. I. Friedmann, Cytology of long-term
desiccation in the desert cyanobacterium Chroococeidiopsis (ChroococcMes), Phycologia
32, 315-322 (1993).
[30] W. K. Dodds, R. W. Castenholz, Effects of grazing and light on the growth of Nostoc
pruniforme (Cyanobacteria), Br. Phycol. J. 23, 219-227 (1987).
[31] F. Garcia-Pichel, R. W. Castenholz, Characterization and biological implications of
scytonemin; a cyanobacterial sheath pigment, J. Phycol. 27, 395-409 (1991).
19

[32] F. Garcia-Pichel, R. W. Castenholz, Occurrence of UV-absorbing, microsporine-tike
compounds among cyanobacterial isolates and an estimate of their screening capacity, Appl.
Environ. Microbiol. 59, 163-169 (1993).
[33] C. E. Wingard, R. W. Castenholz, Evidence regarding the UV sunscreen role of a
mycosporine-like compound in the cyanobacterium Gloeocapsa sp, Appl Environ
Microbiol. 59, 70-176 (1993).
[34] D. R. Hill, S. L. Hladun, S. Scherer, M. Potts, Water stress proteins of Nostoc
commune (Cyanobacteria) are secreted with UV-A/Babsorbing pigments and associate with
1,4-13-D-xylanxylanohydrolase activity, J. Biol. Chem. 269, 7726-7734 (1994).

20

-CHAPTER 2-
Extraction of polysaccharides
from Aphanothece sacrum biomaterials

2-1: Introduction
Aphanothece sacrum (Sur.) Okada produces their own cell colony as jellylike matrix
which may contain the polysaccharides (PS) in a high content. However the extraction
of PS from the capsular matrix is difficult comparing to the non-gelating
exopolysaccharides secreted out of cells. Then, first of all, we simply observed A.
sacrum biomaterials and analyzed them by staining method using cationic pigments,
and tried to extract PS [1]. In this chapter, I introduce A. sacrum briefly and then
describe the extraction studies.

2-2: History and overviews of Aphanothece sacrum
Aphanothece sacrum (Sur.) Okada was biologically classified in the 19th century
by Suringar [2], is a freshwater unicellular cyanobacterium indigenous to J apan. A.
sacrum grew naturally more than 30 years ago in the river of spring water in north and
middle area (Kumamoto and Fukuoka) of Kyushu island in J apan, but recently became
an incredibly rare because of environment deterioration, similarly with other wild
species going to extinct. Currently A. sacrum is artificially kept in existential methods
by only three cultivation companies in Kumamoto and Fukuoka prefectures. These
companies commercially produce processed foods of A. sacrum biomaterials as
SUIZENJ INORI by 30-130 ton / year. It is said that there are many cyanobacteria of
21

about 2500 species in the worlds, however, the edible cyanobacteria are only 5-6 ones
all over the world. In this point of view, A. sacrum is a valuable procaryote not only in
J apan but also in the world. Nevertheless the research on A. sacrum is not made so
widely. Though A. sacrum has been known for more than 100 years, only 36 papers
except for patents are hit in results of a literature search using the keyword
Aphanothece sacrum (by Scifinder). If 6 papers of our studies and 4 J apanese papers
were out of survey, 26 English papers were remained. Out of them, the number of
papers focusing mainly on Aphanothece sacrum is only 22. Such a lack of research
activity is due to the low availability of experimental biomaterials and difficulty of an
efficient cultivation. Here I classified 23 researches of A. sacrum; only 3 papers were
published after 2000 while 19 papers were before 2000. In terms of the main keywords,
the 23 papers were classified as follows; 1 paper on vitamin B12 [3] (publication in
2006), 1 paper on a pure culture [4] (publication in 2004), and 21 papers on proteins
such as ferredoxins [5-24] (published in 1983-2003). (cf. Ferredoxins are small
iron-sulfur proteins that mediate electron transfer in a range of metabolic reactions.
Ferredoxins contain iron and sulfur atoms organized as iron-sulfur clusters. These
biological "capacitors" can accept or discharge electrons, the effect being change in the
oxidation states (+2 or +3) of the iron atoms.) Almost all of the papers on ferredoxins
are related with the structural analyses of the proteins and only one paper focused on
nitrogen fixation.
Thus most researches of A. sacrum metabolites were related with vitamin B12 and
ferredoxins. Prof. K. Kabata have researched metabolites of A. sacrum and found useful
biochemicals such as dimethyl fucose [1, 25], zeaxanthin, fucoxanthin, lycopene,
chlorophyll a, -carotene, phycobilliprotein, pheophytin, medium chain fatty acids [26].
22

By reviewing these reports, one can recognize again that A. sacrum is an excellent
microorganism capable of producing useful many biochemicals by fixing carbon
dioxide under the sunlight. However no reports except ours have addressed PS from A.
sacrum.
As related in Chapter 1, I have selected Aphanothece sacrum (Sur.) Okada, which
has a huge jellylike matrix with a cluster size from several centimeters to several
decimeters (Fig.2-1), out of cyanobacteria wearing the gels. The jellylike matrix is
formed by following mechanism, cells produce and secret PS to outside and
instantaneously bound with calcium and ferric ions in the river to form solid matrix
mainly composed of PS. However, A. sacrum has higher water content than other
popular jelly organisms such as Nostoc commune, according to the literature [27]. From
this information, I can analogize that ECM of A. sacrum has a high water absorption
capacity. The fact that the agars of seaweeds [28] and N. commune [29] are comprised
of PS easily leads to the assumption that a main construct of A. sacrum agar is PS. On
the other hand, the polysaccharide produced by eukaryotic, such as plants (cellulose,
alginate), animals (hyaluronic acid, chondroitin sulfate), fungi (Xanthan gum) and so on,
are wildly utilized as daily products and industrial materials. Whereas cyanobacterial PS
have never been developed except for spirulina [30-32] in spite of having possibility
and promising for new materials [33, 34]. Then I extracted exopolysaccharides from A.
sacrum which is edible to imagine high safety under the speculation that they have
available a lot to study basic physical property for materialization.



23

2-3: Observation of Aphanothece sacrum (Sur.) Okada biomaterials
2-3-1: Methods
Microscopy: Microscopic observation was made by Olympus microscope BX51
equipped with a digital camera DP70.
Alcian blue staining: A. sacrum biomaterials were washed with water and ethanol to
give decolored samples, which were then boiled in xylene at 120
o
C to deparaffinize
them. The samples were stained by being immersed in an alcian blue solution of pH 2.5
or pH 0.5 for 3 minutes. The stained samples were thoroughly washed by distilled water
and dehydrated by ethanol to give blue-stained samples. The samples were sliced into
thin films, observed by the Olympus microscopy, and their pictures taken by an
ultra-high performance CCD camera offering 12.8 million pixels. The thin films were
sandwiched between two glass plates in their wet state just before observation [35].
UV-vis spectroscopy: Ultraviolet-visible absorption spectra of PS aqueous solution were
measured at 25
o
C on a Perkin Elmer Lambda 25 UV/VS spectrometer.
2-3-2: Results and discussion:
We can see an abundance of a jelly-like extracellular matrix (ECM) shown by optical
micrograph (Fig.2-2). One can see that many cell bodies with a diameter of 3-4 m are
dispersed over the jelly matrix, forming a group with a size range less than 80 m. In
addition, about half of the area of the matrix contains no cells. From the results of the
microscopic observation, we could roughly estimate the volume percent of the ECM as
70-85 vol%. The swollen weight of A. sacrum compared to its dry weight was 62-64.
In order to investigate structural characteristics of main constituents for the ECM, we
performed the alcian blue staining test for the A. sacrum raw biomaterials. This method
has been widely used to investigate the mucopolysaccharide distribution in the
24

connective tissues of animals, because alcian blue binds to both carboxylate and sulfate
anions at pH 2.5, but binds to only sulfate at below pH 1.0 [36]. The A. sacrum
biomaterials stained at pH 2.5 were harder and showed a strong blue color, whereas
those stained at pH 0.5 were softer and showed a lighter blue (Fig.2-3). Fig.2-4 shows
microscopic images of the stained samples, demonstrating that alcian blue bound to the
A. sacrum biomaterials under both pH conditions. A blue color of the biomaterials at pH
0.5 indicates the presence of sulfate groups. The color in the biomaterials stained at
pH 2.5 was deeper (Fig.2-4b) than at pH 0.5 (Fig.2-4a), the same as the macroscopic
observations, which indicates the presence of carboxylate groups. Fig.2-4a shows that in
the focused cell bodies are seen a blue ring whose staining intensity was greater than in
the ECM region, indicating that the inside of the cells was not stained; thus, the sulfate
groups were concentrated just around the cell bodies. From these results, it was
inferred that PS is present in all A. sacrum materials except for the inside of the cell
bodies. In addition, the sulfate concentration around the cells suggested that PS may
be secreted by the cells.

2-4: Optimization of polysaccharides extraction process from
Aphanothece sacrum
Most of PS can be easily extracted from sea algae, plant seeds, and a cyanobacterium
Nostoc commune just using pure hot water. However cellulose, chitin/chitosan, and
mannan which have solid structure or special functional groups for example -bond and
amide, N-acetyl, carboxyl, and sulfate groups, are extracted under an alkali or acidic
condition, respectively. In the case of A. sacrum, because we did not have ideas how to
extract PS, initially we tried to use hot water with strong agitation, but homogeneous
25

and pure PS could not be obtained.
Then we extracted PS using alkaline solution to ionize the carboxylic acid, which
may induce the increase in PS solubility in the water as follows (Fig.2-5). The A.
sacrum biomaterials were freeze-thawed and washed in pure water, to remove the
red-purple pigments such as phycobiliproteins (Fig2-6) which was eluted from inside of
cells (fluorescence microscope: inset of Fig.2-6). The samples were washed three times
using a large amount of ethanol with shaking (120 rpm) overnight, and then collected by
filtration using gauze. The ethanol-washed samples were put into 0.1 M NaOH aq at
100
o
C, and agitated at constant temperature for 4 h to yield the transparent solution.
The solution was neutralized with HCl until the pH value decreased to 8.0-9.0, and then
filtrated. The solution containing PS was slowly poured into 100% acetone to
precipitate white fibrous material (Fig. 2-7). The fibers were dissolved in hot water
again, concentrated, and reprecipitated, and these operations were repeated three times
in total. The aqueous solutions of the fibers showed a pH value of about 6.0-7.0. Further
purification was achieved by gel filtration chromatography (Sephadex LH-20, GE
Healthcare) to produce a transparent solution containing PS solute (yield; 60-70%).
The picture of as-extracted PS is shown in Fig.2-8. PS is the white-light brown
fibers harder than widely-known cottons, and is soluble in hot water and
dimethylsulfoxide (DMSO) but insoluble in any other widely-used organic solvents.
The aqueous solution was highly viscous as shown in Fig.2-9 and showed no specific
absorption in the wavelength range of 220-600 nm in ultraviolet-visible (UV-vis)
spectroscopy (Fig.2-10), which confirmed these were not contaminated by proteins,
nucleic acid, chromophores, and/or other chemicals with aromatic rings. The polarized
microscopic observation showed that the precipitates were composed of oriented fibers
26

with a length of less than 3 cm and a diameter of less than 50 m. (Fig.2-11). Fig.
2-11a shows a microscopic photo of the fibers taken under the cross-nicol and Fig.
2-11b shows a corresponding photo taken using a first-order retardation plate (=530
nm) inserted in the light path. The fiber birefringence was negative, as evidenced by
subtractive birefringence (blue color) in the fiber lying from upper left to lower right.
The negative birefringence strongly suggests the PS main chains along the fiber axis.

2-5: Conclusion
Aphanothece sacrum (Sur.) Okada was biologically classified in the 19th century by
Suringar, and now I found its importance for the following reasons; 1) because it is
edible to be safe for the living organisms, 2) because it has a high performance for
producing biochemicals, e.g. vitamin B12, ferredoxins, lipids, and chromophores, 3)
because it has jellylike matrix with a high water content, which may contain PS with a
high-performance hydration behavior. Alcian blue staining test for the A. sacrum
biomaterials suggested the presence of sulfate groups and carboxylate groups in ECM.
The PS was successfully extracted from A. sacrum using alkaline solution. The
extracted PS was soluble in hot water and dimethylsulfoxide but insoluble in any other
widely-used organic solvents. Further it was found that the obtained fiber of PS had a
high purity by UV spectrum and a high orientation behavior by polarized microscopy.





27

References
[1] M. Okajima, M. Ono, K. Kabata, T. Kaneko, Extraction of Novel Sulfated
Polysaccharide from Aphanothece sacrum (Sur.) Okada, and its Spectroscopic
Characterization, Pure Appl.Chem., 79, 2039-204 (2007).
[2] W.F.R. Suringer, Illustration des Algues du Japon, Leide, E. J . Brill (1872).
[3] F. Watanabe, E. Miyamoto, T. Fujita, Y. Tanioka, Y. Nakano, Characterization of a
Corrinoid compound In The edible (Blue Green) Alga, SUIZENJ I-NORI Biosci.
Biotech. Biochem. 70, 3066-3068 (2006).
[4] T. Fujishiro, T. Ogawa, M. Matsuoka, K. Nagahama, Y. Takeshima, H.
Hagiwara, Establishment of a Pure Culture of the Hitherto Uncultured Unicellular
Cyanobacterium Aphanothece sacrum, and Phylogenetic Position of the Organism,
Appl. Environ. Microbiol. 70, 33383345 (2004)
[5] T. Hase, K. Wada, H. Matsubara. Amino acid sequence of the major component of
Aphanothece sacrum ferredoxin. J. Biochem. 79, 329343 (1976).
[6] T. Hase, S. Wakabayashi, K. Wada, H. Matsubara. Amino acid sequence of
Aphanothece sacrum ferredoxin II (minor component) structural characteristics and
evolutionary implications. J. Biochem. 83, 761770 (1978).
[7] H. L. Liu, Y. -M. Lin, Stability and unfolding mechanism of the N-terminal
-hairpin from [2Fe-2S] ferredoxin I by molecular dynamics simulations. J. Chin.
Chem. Soc. 50, 799-808 (2003).
[8] M. S. Searle, R. Zerella, D. H. Williams, L. C. Packman, Native-like -hairpin
structure in an isolated fragment from ferredoxin: NMR and CD studies of solvent
effects on the N-terminal 20 residues. Protein Eng. 9, 559-565 (1996).
[9] S. Ikemizu, M. Bando, T. Sato, Y. Morimoto, T. Tsukihara, K. Fukuyama, Structure
28

of [Fe-2S] ferredoxin I from Equisetum arvense at 1.8 .ANG. resolution. Acta Cryst.
Sec. D: Biol. Cryst. D50, 167-174 (1994).
[10] T. Miyaji, G. Tamura, Isolation and partial characterization of homogeneous nitrite
reductase from a cyanobacterium, Aphanothece sacrum. Biosci. Biotech. Biochem.
56, 1333-1334 (1992).
[11] H. Matsubara, K.Fukuyama, L. J . Rogers, Common structural features of a red
algal flavodoxin and cyanobacterial ferredoxins. Flavins Flavoproteins Proc. Int.
Symp., 10th 441-444 (1991)
[12] T. Tsukihara, K. Fukuyama, H. Matsubara, Y. Katsube, Refined structure of an
iron-sulfur [2Fe-2S] ferredoxin I from Aphanothece sacrum: intramolecular
interaction and structural organization. Protein Struct. Anal., Folding Des. 93-104
(1990).
[13] T. Tsukihara, K. Fukuyama, M. Mizushima, T. Harioka, M. Kusunoki, Y. Katsube,
T. Hase, H. Matsubara, Structure of the iron-sulfur [2Fe-2S] ferredoxin I from the
blue-green alga Aphanothece sacrum at 2.2 .ANG. resolution. J. Mol. Biol. 216,
399-410 (1990).
[14] Y. Minami, Y. Sugimura, S. Wakabayashi, K. Wada, Y. Takahashi, H. Matsubara,
Isolation, properties, and amino acid sequence of a ferredoxin from a multinuclear,
unicellular green alga, Bryopsis maxima. Physiol Vegetale, 23, 669-678 (1985).
[15] T. Tsutsui, T. Tsukihara, K. Fukuyama, Y. Katsube, T. Hase, H. Matsubara, Y.
Nishikawa, N. Tanaka. Main chain fold of a [2Fe-2S]ferredoxin I from
Aphanothece sacrum at 2.5 .ANG. resolution. J. Biochem. 94, 299-302 (1983).
[16] K. Wada, H. Matsubara, R. K. Chain, D. I. Arnon, A comparative study of the
biological activities of two molecular species of chloroplast-type ferredoxins. Plant
29

Cell Physiol. 22, 275-81 (1981).
[17] H. Matsubara, T. Hase, S. Wakabayashi, K. Wada, Gene duplications during
evolution of chloroplast-type ferredoxins. Evol. Protein Mol., Proc. Symp. 209-219
(1978).
[18] A. Kunita, M. Koshibe, Y. Nishikawa, K. Fukuyama, T. Tsukihara, Y. Katsube, Y.
Matsuura, N. Tanaka, M. Kakudo, et al. Crystallization and a 5 .ANG. x-ray
diffraction study of Aphanothece sacrum ferredoxin. J. Biochem. 84, 989-92
(1978).
[19] T. Hase, K. Inoue, H. Matsubara, M. M. Williams, L. J . Rogers, Amino acid
sequence of Synechocystis 6714 ferredoxin: a unique structural feature of
unicellular blue-green algal ferredoxin, J. Biochem. 92, 1357-1362 (1982).
[20] K. Wada, T. Hase, H. Matsubara, Evolutionary information involved in primary
structures of chloroplast-type ferredoxins. J. Biochem. 78, 637-639 (1975).
[21] T. Hase, K. Wada, H. Matsubara, Minor component of ferredoxin from
Aphanothece sacrum cells. J. Biochem. 78, 605-610 (1975).
[22] K. Wada, T. Hase, H. Tokunaga, H. Matsubara, Amino acid sequence of Spirulina
platensis ferredoxin. Far divergency of blue-green algal ferredoxins. FEBS Lett. 55,
102-104 (1975).
[23] K. Wada, H. Kagamiyama, M. Shin, H. Matsubara, Ferredoxin from a blue-green
alga, Aphanothece sacrum. J. Biochem. 76, 1217-1225 (1974).
[24] T. Hase, K. Inoue, N. Hagihara, H. Matsubara, M. M. Williams, L. J . Rogers,
Ferredoxin from Aphanothece halophitica, a unicellular blue-green alga: close
relationship to ferredoxins from filamentous blue-green algae and phylogenetic
implications, J. Biochem. 94, 1457-1464 (1983).
30

[25] K. Kabata, C. Okamoto, N. Sasada, M. Ono, K. Igoshi, H. Kobayashi, C. Masuoka,
Y. Ito, Kyushu Tokai Daigaku Nogakubu Kiyo 24, 37-43 (2005).
[26] K. Kabata C. Okamoto, T. Odahara, T. Nakazono, M. Kikuchi, Studies on the
general analyses and cultivation methods of freshwater cyanobacterium
Suizenzinori (Aphanothece sacrum (Sur.) Okada), Kyushu Tokai Daigaku Kiyo, 16,
47-53 (1997).
[27] K. Hori, T. U. Mohri, T. Oikita, Absorption of color additives and settling volume
in water of blue-green alga, ishikurage (Nostoc commune) Plant Foods Hum. Nutr.
42, 31-36 (1992).
[28] D. Renn. Biotechnology and the red seaweed polysaccharide industry: status, needs
and prospects, Trends Biotechnol. 15, 9-14 (1997).
[29] Y. Tamaru, Y. Takani, T. Yoshida, T. Sakamoto, Crucial Role of Extracellular
Polysaccharides in Desiccation and Freezing Tolerance in the Terrestrial
Cyanobacterium Nostoc commune, Appl. Environ. Microbiol. 71, 73277333
(2005).
[30] Z. Cohen, M. Reungjitchachawali, W. Siangdung & M. Tanticharoen, Production
and partial purification of gamma-Linolenic acid and some pigments from Spirulina
platensis, Journal of Applied Phycology 5, 109-115 (1993).
[31] G. Andrich, A. Zinnai, U. Nesti, F. Venturi, R. Fiorentini, Supercritical Fluid
Extraction of Oil From Microalga Spirulina Platensis, Acta Alimentaria, 35,
195-203 (2006).
[32] M. Muhling, A. Belay, B. A. Whitton, Variation in fatty acid composition of
Arthrospira (Spirulina) strains, J. Appl. Phycol. 17, 137146 (2005).
[33] A. Oren, A proposal for further integration of the cyanobacteria under the
31

Bacteriological Code. Int. J. Syst. Evol. Microbiol. 54, 18951902 (2004).
[34] B. Betsey, D. Dyer, A Field Guide to Bacteria, Cornell University Press (2003).
[35] M. K. Okajima, T. Bamba, Y. Kaneso, K. Hirata, E. Fukusaki, S. Kajiyama, T.
Kaneko, Macromolecules 41, 4061-4064 (2008).
[36] J . R.Harkema, J . A. Hotchkiss, Ozone- and endotoxin-induced mucous cell
metaplasias in rat airway epithelium: novel animal models to study toxicant-induced
epithelial transformation in airways. Toxicology lett. 68, 251-263 (1993).


1cm
Figure21.PictureofAphanothecesacrum.
(livinginthefreshwaterriver)
32
10m
Cell
Jellylikematrix
Figure22.Opticalmicrographof
Aphanothecesacrum.
33
Figure23.Aphanothecesacrumbiomaterialsstained
byalcian blueunderpH0.5andpH2.5.
34
a:pH=0.5 b:pH=2.5
Figure24.Opticalmicrographof Aphanothece
sacrumbiomaterialsstainedbyalcian blue
underpH0.5andpH2.5.Scalebars:10m
35
1.Watersolublepigmentswereremovedbyfreeze/thawing.
2.Lipophilic moleculeswereremovedbyethanolwashing.
3.Thesamplewassolubilization ofpolysaccharidesin0.1N
NaOHsolution.
4.Thealkalinesolutionwas neutraized againstdistilledwater
for72huntilpHvaluebecamedownto8.0-9.0,andthen
filtrated.
5 The filtrate was evaporated 5.Thefiltratewasevaporated.
6.Thesolutionwaspouredintopureisopropanoltoform
fibrousprecipitation.
Figure25.Extractionflowsofpolysaccharides
fromAphanothecesacrumbiomaterials
36
50m
Fluorescentmicroscope
ofthebiomaterials
(excitationwavelength:
520550nm)
Phycobiliproteins
50m
biomateirals
Figure26.Aphanothecesacrumbiomaterialsfreeze
th d t l t l l ti t i i thawedtoelutepurpleaqueoussolutioncontaining
phycobiliproteins fromcells.
37
Figure 27 Polysaccharides extracted from Figure27.Polysaccharidesextractedfrom
Aphanothecesacrum inaceton.
38
Figure28.Polysaccharidesextractedfrom
Aphanothecesacrumindrystate.
39
Figure 29 Viscose aqueous solution of Figure29.Viscoseaqueoussolutionof
polysaccharidesextractedfromAphanothecesacrum.
40
0 35
0.4
0.45
0.5
0.2
0.25
0.3
0.35
o
r
b
a
n
c
e
0.05
0.1
0.15
a
b
s
o
0
200 300 400 500 600
Wavelength(nm)
Figure210.UVvisabsorbancespectrumofaqueous
solutionofpolysaccharidesextractedfrom
Aphanothecesacrum
41
a
b
Figure211.Polarizedmicroscopicphotosof g p p
polysaccharidesextractedfromAphanothecesacrum.
Scalebars:2mm
42
43

-CHAPTER 3-
Structural analyses of polysaccharides extracted from
Aphanothece sacrum

3-1: Introduction
In the chapter 2, we successfully extracted PS from A. sacrum using alkaline
solution. In the course of research development of PS functions, it is tremendously
important to analyze the structure. I have already suggested the possibility of the
presence of PS containing carboxylic acid and sulfate group in ECM of A. sacrum. In
this chapter, I investigate how many these functional groups exist in PS [1].
Furthermore, I report a detail on the primary structure of PS and the result of
higher-order structure. From these results, we may obtain the basic knowledge on
properties and functions. Here I found that the extracted PS was novel because a novel
sugar, sulfated muramic acid, was detected as a constituent sugar of the PS. Then I
named the extracted novel PS sacran on the basis of the species name sacrum and the
meaning of polysaccharides. Muramic acid in bacteria exist as a constituent of
peptideglycan in cell wall but there is reported a capsular polysaccharide having a
muramic acid constituent. Then the muramic acid constituent in polysaccharides is not
so strange.

3-2: Basic analyses of primary structures
3-2-1: Methods
Spectroscopy: Fourier transform infrared spectroscopy (FT-IR) spectra of sacran were
44

recorded at 25
o
C on a Perkin Elmer Spectrum One spectrometer between 4000-600
cm
-1
using a diamond-attenuated total reflection (ATR) accessory.
1
H nuclear magnetic
resonance (NMR) of sacran was measured at 25
o
C in DMSO-d
6
solution on a Varian
FT-NMR spectrometer (UNITY 500plus, 500MHz), using the residual proton resonance
of water as an internal standard (4.79 ppm).
1
H and
13
C NMR spectra of
monosaccharides were measured by J EOL -500 spectrometer in chloroform-d or
pyridine-d
5
at room temperature using tetramethylsilane as an internal standard (0.00
ppm). Electron spectroscopy for chemical analysis (ESCA) was performed with a
Physical Electronics PHI 5600 ESCA system employing Mg K radiation (1253.6 eV)
and a pass energy of 31.5 eV.
Molecular weight: The relative molecular weight of sacran was determined by GPC
(J ASCO HPLC system with a Shodex OHpak SB-806M HQ column, 80 mmID x 300
mmL), calibrated with pullulan standards (eluent: 0.01M (NH
4
)
2
SO
4
aq). Charged
aerosol detector was used for solute detection. As a result of calibration, the following
equation for Mw determination was given:
Log Mw =13.9515 +-0.4768 tR (R
2
=0.999)
where tR is retention time (min). Multi-angle laser light scattering (MALLS) was
performed in static mode at 30
o
C with a fully computerized DLS-7000 system including
a compact goniometer system (Super Dynamic Light scattering spectrometer, Otsuka
Electronics). The angles ranged between 15, 17, 19, 22, 25, 30, and 40
o
. A He-Ne laser
(
0
=632.8 nm) was used as the light source, and the scattering of toluene was used as
the primary standard. The refractive index increment, dn/dc, was chosen at 0.1435
mLg
-1
. Sacran solutions with concentrations of 0.27, 0.18, 0.135, and 0.09 mgml
-1

were used. J ust before each measurement, the sample solutions were slowly filtered at
45

least 3 times using a syringe filter (pore size 5.00 m, Ministart, sartorius, Germany)
and dispensed directly into the cylindrical light scattering cuvette (20 mm, Pyrex tubes,
Iwaki, Tokyo).
Other analyses: Elemental analysis was made by Yanako CHN coder MT-6 (at Center
for Organic Elemental Microanalysis at Kyoto University).
3-2-2: Results and discussion
Spectroscopic study: FT-IR spectra of sacran showed several distinct peaks at
wavenumbers of 3354 (hydroxyls), 2924 (aliphatic chains), 1613 (carbonyls), 1415
(aliphatic chains), 1363 (sulfates), 1223 (sulfates), and 1022 cm
-1
(hydroxyls) (Fig.3-1).
The spectra indicated that sacran has carbonyl and sulfate groups, as well as the typical
groups of sugars. Since the carbonyl type was unclear, the FT-IR spectrum of
acid-treated sacran was studied. The peak appeared at 1736 cm
-1
, indicating that
sacran from A. sacrum has carboxyls presumably from uronic acid. Uronic acid content
was estimated as 22 % by the carbazole-sulfuric acid method (525 nm). Fig. 3-2 shows
ESCA diagram of sacran, showing a peak at a binding energy of 169 eV, with a shoulder
around 170 eV, which is characteristic of sulfate peak in ESCA. Next We tried to
measure
1
H NMR spectra of sacran in DMSO-d
6
(Figure 3-3). This figure shows a
distinct signal, whose maximum lies at a chemical shift of 1.24 ppm, and multiple
broad signals at 2.8-3.6 ppm. All the signals were difficultly shown because the
peak at 1.24 ppm was much smaller than multiple broad signals at 2.8-3.6 ppm.
While a signal at 1.24 ppm may imply the presence of 6-deoxy sugar unit, the broad
signals at 2.8-3.6 ppm may imply the presence of methoxyls, methylenes, and
methines. Since all other cyanobacterial PS contain at least 6 units of different
monosaccharides [2-6], one can analogize that sacran from A. sacrum may also contain
46

several different monosaccharides and the multiple signals at 2.8-3.6 ppm may be
overlapping of too many signals. Then it is impossible to make an in-depth
investigation of the multiple signals in the
1
H NMR spectra of not-hydrolyzed sacran.
The elemental analysis of desalinated sacran by a cation exchange resin (DOWEX
50W-X8, 50-100 mesh, H form) showed results as follows; C; 36.04 %, H: 5.91 %,
N:0.30 %, S:2.07 %. If average molecular weight for a sugar unit is assumed to be 162,
the molar composition of S to the total sacran can be estimated as 10 mol%. The
spectroscopy and elemental analyses indicated the extracted sacran contains carboxyls
and sulfate groups. The degree of sulfation was estimated as about 10 mol%.
Molecular weight measurement: Gel permeation chromatography (GPC) of the sacran
solution with a concentration of 0.01% in 0.01M (NH
4
)
2
SO
4
aq indicated that sacran had
a high weight-average Mw of 1.7 x 10
7
g/mol with a narrow polydispersity of 1.7
(Fig.3-4a). This value is so high that I can speculate the molecular weight may be from
sacran aggregates. Then, I examined changes in the peak size and the retention time by
diluting the solution. By dilution, the peak size decreased but the retention time did not
change even if very dilute solutions such as 0.001% and 0.0001% were measured. On
the other hands, no distinct peak was detected in the concentration of 0.00001% solution,
presumably beyond the limitation of detector. It is most important that a new peak did
not appear at all in lower-molecular-weight areas in any solutions. On the other hand,
the generation of too small molecules to detect cannot be denied. Then the change of
peak intensity as a function of sacran concentration was quantitatively investigated and
showed the plotts in Fig.3-4b. Peak area and peak height were both proportionate to the
sacran concentration. A least-square approach made clear that the connection of these
plots showed a straight line under sufficient accuracy; coefficient of correlation,
47

R=0.999. The peak intensity reduced by just that deceased amount of sacran solutes.
This result indicated that sacran chains in the specimens showed only the peak of 17
million in solutions with any concentration. Then the molecular weight of an elemental
sacran chain is 17 million. Since the elemental chain is not dissociated by dilution, I
can consider that it is a single chain of sacran, however, another possibility of the
elemental chains is chain associates bound by an unbelievably-strong force. Nonetheless,
the molecular weight value is merely relative, and it possesses lower reliability from the
fact that a maximum molecular weight of pullulan standard is 700 000. Then we
measured the absolute molecular weight using multi-angle (from 15
o
to 40
o
) laser light
scattering (MALLS). We successfully obtained a typical Zimm-Berry plot with a very
small error of 1.4% (Fig.3-5) by repeating the challenges under careful filtration (pore
size: 5 m) under a pressure. When the absolute Mw of the sacran in pure water was
measured as a first try, the error value was high due to a severe dependence on sacran
concentration, and we did not obtain a valid lattice shape of the Zimm-Berry plot. The
addition of NaNO
3
may be useful for adopting the not-electrostatically-restrained
conformation of sacran. From an analysis of the Zimm-Berry plot, the absolute Mw
and radius of gyration, <s
2
>
1/2
, for sacran were estimated at 1.6 x 10
7
and 402 nm,
respectively. Second virial coefficient is 1.51 x 10
-9
.

Since the value is plus, sacran
chains in the concentration range between 0.027 and 0.009 wt% are repulsive,
indicating sacran chains may be difficult to associate in a dilute aqueous solution. The
result support my speculation that the elemental chain of sacran is a single chain. The
absolute Mw was comparable with the relative Mw obtained by gel permeation. To the
best of our knowledge, this is the first confirmation of an absolute Mw over 10
7
Da for a
water-soluble bio-derived polymer (e.g. hyaluronic acid and xanthan gum: ~10
6
Da)
48

[7,8] although it has been reported that some associative PS had extremely high
apparent Mw values (~10
9
Da)

[9].

3-3: Multi-step acid hydrolysis and determination of constituent sugars.
3-3-1: Methods:
Multi-step acid hydrolysis: The sacran solution with a concentration of 0.5 wt% (sacran:
1g) were mixed with cation exchange resins and were slowly stirred to exchange Na
canterions of sacran into H ions to obtain an acidic fragment of sacran. When the
mixture was filtrated to remove the cation exchange resins using vacuum filtration
implement, a fraction was obtained as a gel covering the resins (fraction A). The
fraction A may be the acidic fraction presumably rich in uronic acid. The filtrate was
separated from the acidic fragment and, and then was concentrated by evaporation and
poured into methanol. Here a fibrous fraction appeared in methanol (fraction B) and the
soluble parts in methanol was obtained (fraction C). The components of fraction C were
analyzed by TLC to confirm oligomers. In order to determine constituent sugars of
fraction C, it was methanolyzed after evaporation and dry, and then was
trimethylsilylated to analyze gas chromatography (GC) and gas chromatography-mass
spectrometry (GC-MS). The methanolysis and trimethylsilylation procedure are showed
as follows. 1 N HCl-MeOH which was prepared by adding dry MeOH (50 ml) into
CH
3
COCl (3.6 ml) was added to the fraction C (1 mg) and heated to 70
o
C and kept for
16 hours with agitation. The methanolyzed sample was evaporated with washing
t-BuOH in three times. Then the sample was trimethylsilylated with TMSI-C (200 ul) at
room temperature for 1 hour, to analyze constituent sugars by GC and GC-MS. On the
other hand, the insoluble parts in methanol (fraction B) were acid-hydrolyzed using 4 M
49

TFA at 110
o
C for 4 hours and then methanolyzed for 24 hours by the procedure similar
with fraction C. Further, standard monosaccharides such as Glc, Gal, Man, Fuc, Rha,
Xyl, Ara, GalA, GlcA, GlcN, GalN, GlcNAc, GalNAc, Rib, Mur, and MurNAc were
also methanolyzed and trimethylsilylated, and used as external standards. However
guluronic acid (GulA) and mannuronic acid (ManA) are not available commercially.
Then alginate was acid-hydrolyzed using 4M TFA at 110
o
C for 48 hours and
methanolyzed for 44 hours in order to obtain external standards of GulA and ManA.
The uronic acid rich fraction (fraction A) was methanolyzed for 21 hours without
hydrolyzed and was analyzed after trimethylsilylation by GC and GC-MS. Further, the
same sacran sample which was used multi-step acid-hydrolysis was also
acid-hydrolyzed (4M TFA for 4 hours and methanolyzed for 16 hours (parent sample of
sacran for multi-step method; sample D).
GC and GC-MS analysis: Monosaccharides constituting the jelly matrix were analyzed
by the following procedure. After the poly- and oligosaccharides were acid-hydrolyzed,
methylated, and trimethylsilylated, the constituent sugars were analyzed by GC-MS
apparatus (Trace DSQ, Termo Fisher Scientific Inc., Waltharm, MA, USA) and GC
(GC-18A, Simadzu, Kyoto, J apan) equipped with a 30 m 0.25 mm i.d. fused-silica
capillary column coated with 0.25-m TC-1 (GL Sciences, Tokyo, J apan). The
operational conditions were as follows: GC-MS; the injector temperature was 200 C;
the flow rate of helium gas was 1 ml/min; the column temperature program was started
at 80 C, increasing to 260 C at 6 C/min; the transfer line and the ion source
temperatures were 250 C. GC; the injector temperature was 200 C; the detection
(flame ionization detection) temperature was 250 C; the column temperature was held
at 140 C for 2 min, then raised by 8 C/min to 260 C, which was kept for 13 min; the
50

carrier gas was nitrogen, 100 kPa.
FT-ICR-MS analysis: For fourier-transformation ion cyclotron resonance
mass-spectroscopy (FT-ICR-MS), the dried sacran was acid-hydrolyzed with 4M TFA
at 110
o
C for 78 hours. FT-ICR-MS was performed by electrosplay ionization (ESI) of
degraded sugar chains. ESI FT-ICR mass spectra were recorded on a Bruker
BioAPEXII 70e FT-ICR mass spectrometer (Bruker Daltonics Inc. Billerica, MA)
equipped with a 7 T superconducting magnet and an external ESI source (Analytica of
Branford Inc.). The sample solution was infused into the ESI source in a negative ion
mode with a syringe pump at 60 L/h and was desolvated by a countercurrent nitrogen
gas heated to 250
o
C. The nitrogen needle-gas was passed through the grounded coaxial
needle to the metal-capped glass capillary (-3.5 kV). The operation of the FT-ICR mass
spectrometer was described elsewhere [10]. The possible molecular formulae (all
possible combination of atomic masses with the least deviation from the measured
mass) were obtained using the mass analysis module (Bruker XMASS package). All
FT-ICR mass spectra were identically calibrated using a mixture of caffeine, adenocine
monophosphate, tetrapeptide MRFA, and ultramark 1620.
3-3-2: Results and discussion
Biologically important carbohydrate polymers occurring extracellularly and on the
surfaces of cells often contain uronic acid residues which are ionized at physiological
pH's. However determination of primary sequences of polysaccharides containing
uronic acids is often complicated by the unusual stability to acid hydrolysis of
glycosidic bonds formed by these residues. The most direct and obvious method to
facilitate quantitative depolymerization of polymeric uronides under nondestructive
hydrolysis conditions is to convert them to esters which in turn can be reduced to the
51

more readily hydrolyzable neutral polymers [11-14]. Since sacran is a polyanion with
carboxylate and sulfate groups, it may be very difficult to hydrolyze for analyzing by
one-step acid-hydrolysis. Then I established multi-step acid-hydrolysis method to
clarify the constituent sugars and their compositions of sacran.
GC and GC/MS analyses of trimethylsilylated samples of sacran (derived from
Aphanothece sacrum cultivated in summer season) hydrolyzed by one-step method
(with 4M TFA at 110
o
C for 78h) indicated that the main monosaccharides were Glc,
Gal, Man, Xyl, Rha, Fuc, GalA, and GlcA, with a composition of 25.9: 11.0: 10.0: 16.2:
10.2: 6.9: 4.0: and 4.2. We also confirmed the presence of trace amounts (ca. 1.0 %) of
Ara, GalN, and Mur. However the monosaccharide eliminating at the early stage of the
acid-hydrolysis may be degraded by further acid-treatment not to detect by GC and
uronic acid is also easy to degrade. The monosaccharide degradation changes the result
of sugar residue compositions. Then I established multi-step acid-hydrolysis method
and the results are shown follows. In the analytic method, the muramic acid was
additionally detected with sugar residues detected by the normal method from the acidic
fraction (fraction A). GlcA was not detected while a large peak was detected at a
retention time between polar monosaccharides such as uronic acids and less polar ones
such as pentoses. I speculate that the large peak may be assigned GlcA derivative with
hydrocarbons such as a methyl group. Next constituent sugars and their compositions
of a methanol-soluble fraction (fraction C) were analyzed; Glc, Gal, Man, Xyl, Rha, Fuc,
GalA, GlcA, GalN, and HexA with a composition of 35.1: 3.0: 4.9: 14.9: 19.9: 0.0: 0.0:
0.0: 0.0: 0.0 (Fig.3-6). This result indicates that constituents of Glc, Gal, Man, Xyl, Rha,
were easily eliminated as a result of mild acid treatment using cation-exchange resin
while Fuc, GalA, GlcA, GalN, and HexA were remained sacran backbone or sacran
52

oligomer backbone. Then Glc, Gal, Man, Xyl, and Rha may constitute the sugar
chains around the chain-end and/or side chains of sacran. Whereas, constituent sugars
and their compositions of a fibrous fraction (fraction B) were analyzed; Glc, Gal, Man,
Xyl, Rha, Fuc, GalA, GlcA, GalN, and HexA with a composition of 26.9: 8.2: 4.8: 10.3:
10.0: 4.7: 1.2: 1.4: 4.4: and 11.4 (Fig.3-7). On the other hand, sample D (parent sample
of sacran for multi-step method) were analyzed; Glc, Gal, Man, Xyl, Rha, Fuc, GalA,
GlcA, GalN, and HexA with a composition of 31.1: 15.3: 16.5: 12.5: 13.8: 2.0: 1.0: 1.6:
1.8: and 4.4 (Fig.3-8). Comparing with the results of sample D, composition increases
of Fuc, GalA, GalN, HexA were found to support the results in fraction C. In two
parent sacrans, the compositions of hexoses such as Glc, Gal, Man and neutral sugars
such as Xyl and Rha were larger than those of Fuc and uronic acids such as GalA and
GlcA were smaller. However, sample D (derived from Aphanothece sacrum cultivated
in winter season) was found to contain a small amount of GalN and unknown uronic
acid presumably due to the milder hydrolyses condition than the former. One can say
that the natural materials produced by organisms are dependent on cultivation seasons.
Thus, constituent sugars and their ratio of sacran depend on the picked-up season from
cultivation river, but It is noticeable that the sort of the main sugar resides were kept in
parent sacrans regardless on cultivation month or other environments. Moreover the
analyses of GC and GC-MS confirmed the absence of ManA, GulA, or Rib as sugar
residues of sacran, indicating the sacran is a unique PS quite different from
seaweed-derived PS or contaminated with nucleic acids.
One can presume that the N element detected by elemental analysis was derived
from GalN and Mur. Fourier Transform Ion Cyclotron Resonance Mass Spectroscopy
(FT-ICR-MS) of methanolyzed sacran (Fig.3-9) showed a milli-mass unit value of
53

358.0806, corresponding to a [M-H]
-
=358.0808 from sulfated dimethylmuramic acid
(structure; Fig.3-9) whose two methyl groups were considered to be attached to
reducing terminal and carboxylic acid group in the process of the methanolysis.
Furthermore, the MS/MS measurement showed 278.1244 to [M-H]
-
=278.1240 from
dimethylmuramic acid, indicating the elimination of a sulfate group. Then Mur detected
by GC method was sulfated in the state of sacran constituent. Sulfated muramic acid is a
newly-found sugar in nature, and thus one can state that the extracted sacran was a
novel one. We also found sulfated muramic acid in the sacran extracted using the
following method: the ECM of A. sacrum was eluted by hot pure water using an
autoclave, and ultracentrifuged (at 50000 rpm for 30 min) to obtain a clear supernatant
which was poured into iso-propanol to precipitate the sacran fibers. FT-ICR-MS
indicated that the presence of di- and trisaccharides composed of hexoses and muramic
acid derivatives. The results of connection of muramic acid with hexoses
demonstrated that the muramic acid derivative found as a sacran sugar unit was not
merely a contaminant from cell wall constituents but rather a constituent of the capsular
sacran of A. sacrum because muramic acid generally exists with N-acetyl glucosamine
in the cell walls of bacteria [15]. It was demonstrated that sacran was an ampholytic SC
with an imbalanced charge ratio resulting in a high content of anionic sugars with
sulfate and carboxylic acid groups, and a low content of cationic amino sugars
(anions/cations; ca. 30/1).
In the next step, sacran was partially acid-hydrolyzed by more mild condition than
described above to give mixtures of oligomers, and partial structures were analyzed by a
milli-mass estimation of the obtained oligomer mixture by FT-ICR-MS methods.
Initially, the dry sample of sacran (0.5 g) was hydrolyzed with 4M trifluoroacetic acid
54

(50 ml) for 10 h at 121
o
C, and during hydrolysis treatment we repeatedly checked the
decomposition state using TLC analyses to confirm the presence of oligomers. The
hydrolysis was further made for 6 h, and it was confirmed that trisaccharides,
tetrasaccharides and pentasaccharides were contained in acid-degraded products. After
additional hydrolyses for 5 h (total hydrolysis was 21 h), the hydrolysate was found to
degrade into trisaccharides, disaccharides, and monosaccharides, and it was considered
that these molecular weights were low (below 700) enough to analyze by FT-ICR-MS.
Next, reducing terminals of obtained products were methylated with methanol (50
ml)/acetyl chloride (3.6ml) to measure mass by FT-ICR-MS. Compositions of
oligosaccharides were analyzed based on m/z values of main peaks. For example in
entry 5 of Table 3-1, the measured value of 301.0239 corresponds with the calculated
one for a [M-H]
-
(301.023) from sulfated dimethyl uronic acid on the level of milli-mass
units, and other corresponding compositions were not found even if various
combination of sugars, substituents, and ionization were used in calculations. Thus, the
peak was assigned to sulfated dimethyl uronic acid. Eight other sugar constituents found
by an analogous method are shown in Table 3-1, where we extracted two trisaccharide
sequences whose structures were shown in Fig.3-10. The sequence common in entries
of 1, 2, 3, 7, and 8 contains an uronate diad while that common in entries of 6 and 9
contains muramic acid. It is presumable that these sequences are build into the
compartment of polysaccharides because intensity of each peak was high. Here, it was
cleared that uronic acid corresponds to glucuronic acid or galacturonic acid as clarified
above. Other structural characters are shown below; 1) some sulfonic acids are
proximal to carboxylic acid. 2) sacran is composed of a structure that has many
hydroxyl groups located around carboxylic acids. On account of facts, one can say that
55

sacran has diverse partial structures around carboxylic acids. The determination of
in-depth primary structure containing the diverseness of constituent sugars similarly
with other cyanobacterial polysaccharides [16] are currently proceeded, using enzyme
degradation treatments according to literature on Nostoc commune polysaccharides [17].

3-4: Higher-order structures of extracted polysaccharides
3-4-1: Methods
All Atomic force microscopy (AFM) experiments were performed using a
commercial AFM unit (SPA-400, Seiko Instruments, J apan) equipped with a calibrated
20 m xy-scan and 10 m z-scan range PZT-scanner. For the imaging of sugar chains
by AFM, a stiff cantilever (SI-DF20, Seiko Instruments, force constant is 13 N/m in
typical value, typical resonant frequency is 130 kHz, pyramidal tip shape, tip curvature
radius is 10 nm) were used and imaging was taken in the dynamic force modulation
(DFM) mode at optimal force. The scan speed was 2 m/sec. TEM images were
obtained with a Hitachi HF-2000 Field Emission Transmission Electron Microscope
operated at an acceleration voltage of 100kV at a magnification of 35,000x. The
specimens were prepared by slow evaporation of a drop of a diluted MeOH/Water (20/1
v/v) solution of sacran (ca. 10 ppm) on a carbon-coated copper mesh grid.
3-4-2: Results and discussion
Atomic force microscopy (AFM; Fig.3-11a) of specimens dried from a diluted
aqueous solution (0.5 ppm) on a mica sheet showed the formation of network structures
containing nanosized loops whose diameter ranged from 100 nm to 2 m. On the
other hand, we took AFM images of sacran dried in the presence of NaCl and found that
sacran chains drew almost straight lines (Fig.3-11b) indicating that cross-linking points
56

of sacran networks were broken. The lines were 0.4-0.7 nm in thickness (Fig.3-12) and
then were composed of the single or double sacran chain to suggest the adoption of the
extended conformation. Chain extension enhancing the hydration power countervailed
the effects of added salts. The formation of the physical hydrogels and the variability of
the MALLS measurement without NaNO
3
may be due to the dispersion of the
electrostatically-associating points. Since sacran may be semi-rigid-rod as other
natural polysaccharides [7], one can discuss that they recovered extended conformation
by the breakage of electrostatic association. The chain extension gave an important
discussion. Generally it is quite difficult to observe the AFM image of extended
polymer chains but sacran easily gave a chance. Fortunately I caught a photo of sacran
chains extended to the length of 13 m which corresponds to the thickness of human
thin hair (Fig.3-13). The length of fully-extended sacran was calculated to be about 50
m from molecular weight. The observed length was a bit shorted than the calculated
one but the length is a surprising value indicating the sacran is longest of all the
polymers whose length was measured.
Transmission electron microscopy (TEM; Fig.3-14) of specimens dried from a diluted
MeOH/Water (20/1 v/v) solution of sacran (ca. 10 ppm) on a carbon-coated Cu grid
showed that almost all of the sacran chains appearing as black threads which spreads out
all over the copper grid. Energy dispersive X-ray spectroscopy of the threads was
simultaneously performed. Since the Na and O elements, as well as Cu, were distinctly
confirmed (Fig.3-15), it was concluded that these strings were sacran chains. Sacran
chains formed nanoloops supporting the AFM results and its size is about 400 nm which
is in good agreement with the result of MALLS. This form was often seen in
microscopic images of RNA and specific proteins as a lariat structure, indicating that
57

the associating points dispersed in the polymer chains. In sacran, the association can
occur by electrostatic attraction of a small amount of amino cations with carboxylate
and sulfate ones.

3-5: Conclusion
The structure of Aphanothece sacrum-derived PS, sacran, was analyzed. Sacran was
composed of more than 11 sugar residues containing hexoses, pentoses, deoxyhexoses,
uronic acid, and a novel sugar of sulfated muramic acid. The content of carboxylic
groups was 22 % and that of sulfate groups was 11%. GPC characterization of sacran in
very thin solution demonstrated that weight-average molecular weight of sacran
elemental chains was 1.7 x 10
7
g/mol. I found the surprising character of sacran
structure, i.e. absolute molecular weight was 1.6 x 10
7
g/mol which was the highest
value of all the extracted biomolecules reported thus far. AFM and TEM microscopy
showed that sacran adopted an extended conformation under salt and was a
megamolecule with a length of 13 m. Sacran was difficult to hydrolyze completely and
then I established multi-step acid-hydrolysis methods. It was found that the main
backbone of sacran was composed of Fuc, GalA, GlcA, GalN, and HexA. Ssacran did
not contain ManA, GulA, or Rib, meaning sacran was quite different from
seaweed-derived PS. Further analyses of oligomeric fractions implied that sacran had
diverse partial structures around carboxylic acids.

References
[1] M.Okajima, M.Ono, K.Kabata, T.Kaneko, Extraction of Novel Sulfated
Polysaccharide from Aphanothece sacrum (Sur.) Okada, and its Spectroscopic
58

Characterization, Pure Appl.Chem., 79(11), 2039-204 (2007).
[2] Y. Tamaru, Y. Takani, T. Yoshida, T. Sakamoto, Crucial Role of Extracellular
Polysaccharides in Desiccation and Freezing Tolerance in the Terrestrial
Cyanobacterium Nostoc commune, Appl. Environ. Microbiol. 71, 73277333
(2005).
[3] J . B. Lee, T. Hayashi, K. Hayshi, U. Sankawa, M. Maeda, T. Nemoto, H.
Nakanishi, Further Purification and Structural Analysis of Calcium Spirulan from
Spirulina platensis J. Nat. Prod. 61, 1101-1104 (1998).
[4] D. Garozzo, G. Impallomeni, E. Spina, L. Stuuriale, The structure of the exocellular
polysaccharide from the cyanobacterium Cyanospira capsulate, Carbohydr. Res.
307, 113-124 (1998).
[5] P. Li, Z. Liu, R. Xu, Chemical characterisation of the released polysaccharide from the
cyanobacterium Aphanothece halophytica GR02, J. Appl. Phycol. 13, 71-77 (2001).
[6] J . M. Panoff, B. Priem, H. Morvan, F. J oset, Sulphated exopolysaccharides produced
by two unicellular strains of cyanobacteria, Synechocystis PCC 6803 and 6714 Arch.
Microbiol. 150, 558-563 (1988).
[7] J . R. E. Fraser, T. C. Laurent, U. B. G. Laurent, Hyaluronan: its nature, distribution,
functions and turnover, J. Intern. Med. 242, 27-33 (1997).
[8] J . -H. Kim, S. -J . Yoo, D. -K. Oh, Y. -G. Kweon, D. -W. Park, C. -H. Lee, G. -H. Gil,
Selection of a Streptococcus equi mutant and optimization of culture conditions for
the production of high molecular weight hyaluronic acid, Enzyme Microbial Technol.
19, 440-445 (1996).
[9] B. E. Christensen, O. Smidsroed, A. Elgsaeter, B. T. Stokke, Depolymerization of
double-stranded xanthan by acid hydrolysis: characterization of partially degraded
59

double strands and single-stranded oligomers released from the ordered structures
Macromolecules 26, 6111-6120 (1993).
[10] K. Miyabayashi, Y. Naito, M. Miyake, K. Tsujimoto, Structure characterization of
polyaromatic hydrocarbons in Arabian mix vacuum residue by electrospray
ionization Fourier transform ion cyclotron resonance mass spectrometry Eur. Mass
Spectrom., 6, 51-59 (2000).
[11] R. L. Taylor, H. E. Conrad, Stoichiometric Depolymerization of Polyuronides
andGlycosaminoglycuronans to Monosaccharides following Reduction of Their
Carbodiimide-Activated Carboxyl Groups. Biochemistry 11, 1383-1388 (1972).
[12] W. S. York, A. G. Darvill, M. Mcneil, T. T. Stevenson, P. Albersheim, Isolatiion and
Characterization of Plant Cell Walls and Cell Wall Components, Methods Enzymol.
118, 3-40 (1985).
[13] B. Quemener, M. Lahaye, C. Bobin-Dubigeon, Sugar determination in ulvans by a
chemical-enzymatic method coupled tohigh performance anion exchange
chromatography, Journal of Applied Phycology 9, 179188 (1997).
[14] A. Cesaro, G. Liut, C. Bertocchi, L. Navarini, R. Urbani, Physicochemical
properties of the exocellular polysaccharide from Cyanospira capsulate, Int. J. Biol.
Macromol., 12. 79-84 (1990).
[15] H. Frank, M. Lefort, H. H. Martin, Chemical analysis of a mucopolymer
component in cell walls of the blue-green alga Phormidium uncinatum. Biochem.
Biophys. Res. Commun. 7, 322-325 (1962).
[16] T. Uragami, T. Wada, M. Sugihara, Studies on syntheses and permeabilities of
special polymer membranes. 62. Active transport of alkali metal ions through
poly(vinyl alcohol) membranes crosslinked with glutaraldehyde, Angew. Makromol.
60

Chem. 138, 173-183 (1986).
[17] R. D. Philippis, C. Sili, R. Paperi, M. Vincenzini, Exopolysaccharide-producing
cyanobacteria and their possible exploitation: A review J. Appl. Phycol., 13,
293-299 (2001).
pH8.5
a
n
s
m
i
t
t
a
n
c
e
R-COOH
pH2.5
T
r
a
R-SO
4
-
4000 3500 3000 2500 2000 1500 1000 500
Wavenumber (cm
-1
)
Figure31.IRspectraofsacranprecipitated
fromaqueoussolutionofpH2.5andpH8.5.
61
S
155 160 165 170 175 180
Binding energy (eV)
Figure32.ESCAspectrumofsacran.
62
In DMSO-d
6
DMSO
Chemical Shift, (ppm)
Figure33.
1
HNMRspectraofsacran.
63
Mn:16000000
Mw:20600000
Mn/Mw:1.2
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Retention time /min
Figure34.GPCchromatogramofsacransolutionin0.1MNaNO
3
.
64
1.00E-03
1.20E-03
6.00E-04
8.00E-04
1.00E 03
}
1
/
2
/
(
m
o
l
/
g
)
1
/
2
0 00E+00
2.00E-04
4.00E-04
{
K
c
/
R
(

)
}
0.00E+00
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
sin
2
(/2) +0.05c
Figure 3 5 Zimm Berry plot Figure35. ZimmBerryplot
ofsacransolutionin0.1MNaNO
3
.
65
7.0
8.0
9.0
10.0
uV(x1,000)
3.0
4.0
5.0
6.0
5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 min
-1.0
0.0
1.0
2.0
Figure36.GCMSspectrumofsacranfractionC
aftercationexchangeresintreatment.
Thefractionissolubleinmethanol.
66
1 75
2.00
uV(x100,000)
1.00
1.25
1.50
1.75
0.25
0.50
0.75
5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 min
0.00
Figure37.GCMSspectrumofsacranfractionB
aftercationexchangeresintreatment.
Thefractionisinsolubleinmethanol.
67
1.4
1.5
uV(x100,000)
0 7
0.8
0.9
1.0
1.1
1.2
1.3
0.1
0.2
0.3
0.4
0.5
0.6
0.7
5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 min
0.0
Figure38.GCMSspectrumofsacransampleD
beforecationexchangeresintreatment.
68
S/ S
O
H
H
H
-
O
3
SO
O
H
H
H
HO
MS/MS
O
H
O
H
NH
2
OMe
HO
MeOOC
O
H
O
H
NH
2
OMe
MeO
HOOC
280 275 360
[MH]

358.0808
[MH]

278.1240
280 275 360
Figure 3 9 FT ICR MS spectrum of methanolyzed Figure 39. FTICRMS spectrum of methanolyzed
sacran showing the peak of m/z=358.0806
assigned to sulfated dimethyl muramic acid (right
structure). MS/MS spectrum showed the peak of
m/z=278.1224 assigned to dimethyl muramic acid
(left structure), to illustrate the elimination of
sulfate group.
69
Figure310.Partialstructuresofsacrandetermined
fromTable31.Upperstructure:fromentriesof
1,2,3,7,and8.Lowerstructure:fromentriesof6and
9intable31.
70
a)
b)
Figure311. Atomicforcemicroscopicimagesof
specimensdriedfromadilutesolutionofsacran(a)
anddriedfromadilutesalinesolutionofsacran (b).
71
Figure312.CrosssectionaldiagramofAFMimageof
sacran, indicatingthelinethicknessis0.40.7nm
correspondingwiththicknessofsacran chains.
72
13 m
Figure313.AFMimageofsacranchains.The
specimenwaspreparedbydryingthe verydilute
solutionofsacranwithaconcentrationofppborder
onamicaplate.
73
Figure314.transmissionelectronmicroscopic
i f i imagesofspecimens
driedfromadilutesolutionofsacran.
74
Fig.315EDXspectrumofsacran recordedwhileTEM
l A h d O d N l t i i l analyses.ArrowsshowedOandNaelementsoriginal
tosacran chains.
75
Table3-1 m/z values of main peaks in milli-MS measurement and correspondingsugars Table 3-1 m/z values of main peaks in milli-MS measurement and corresponding sugars
entry m/z calculated m/z Sugar constituents Mw measurement
modes
1 355.0869 356.095 [M-H]
-
uronic acid +hexose 356.095 negative
2 369.0667 370.075 [M-H]
-
uronic acid +uronic acid 370.075 negative
3 545.1362 546.144 [M-H]
-
methylated (uronic acid +uronic
acid +hexose)
546.143 negative
4 693.2089 694.217 [M-H]
-
methylated (uronic acid +hexose +
hexose +hexose)
694.217 negative
5 301.0239 301.023 [M-H]
-
sulfated dimethyl uronic acid 302.031 negative
6 358.0819 359.090 [M-H] sulfated dimethyl muramic acid 359.089 negative
7 435.1127 435.112
[M+Na]
+
dimethylated (uronic acid +uronic
acid)
412.122 positive
8 463.0773 464.076 [M-H]
-
sulfated dimethylated (hexose +
uronic acid)
464.084 negative
uronic acid)
9 630.2163 630.224 [M-H]
-
methylated (hexose + hexose+
N-acethylmuramic acid)
631.232 negative

76
77

-CHAPTER 4-
Solution properties and liquid crystalline natures
of polysaccharides extracted from Aphanothece sacrum

4-1: Introduction
In the previous chapter, we demonstrated that sacran contains carboxyl, sulfate and a
trace of amide functional groups. Furthermore, it was found that specific
monosaccharide of dimethylated fucose and sulfated muramic acid were contained, and
sacran was composed of more than 11 monosaccharides as constituent sugars by
spectroscopy and GC, GC-MS and FT-ICR-MS analyzing. The molecular weight and
radius of gyration of sacran were estimated at 1.6 x 10
7
Da and 402 nm, respectively, by
GPC and MALLS measurements. That is to say, it is revealed that sacran is a
megamolecule largest in the extracted natural molecule. Sacran contains many hydroxyl
groups locating around carboxylic acids and diverse partial structures relating with
carboxylic acids. In this chapter, we tried to investigate solution properties such as
viscosity, gelation, and the solution behavior under various salts or ions. Furthermore,
dielectronics, liquid crystallinity, and rheooptics of sacran are discussed.

4-2: Solution properties of extracted polysaccharides
4-2-1: Methods
Titration: Hydrogen chloride solution used for titration was prepared by 10 fold dilution
of concentrated HCl (12M, Kanto Chemical Co. Inc.). The titration of the sacran
solution (0.33 wt%, 150ml) was performed as follows. A very small volume (50 l) of
78

HCl aqueous solution (pH 0.23) was dropped stepwise into the sacran solution with a
moderate agitation under a nitrogen atmosphere at 25
o
C. The pH, conductivity, and
turbidity of the sample at each step was measured (initial conditions: pH 7.52, turbidity
0.197, and conductivity 24.3 Scm
-1
). The pH was measured by a Tix-90i pH meter
(Toko Chemical Laboratories Co. Ltd., Japan). The turbidity was measured using a
UV/VIS Spectrometer (Lambda 25, Perkin Elmer) at a wavelength of 600 nm. Each
sample was set in a glass cell (light path: 1 cm). The conductivities were measured by a
conductivity meter (SevenEasy Conductivity Mettler-Toledo GmbH, Switzerland).
Viscosity measurement: Viscoelastic behavior of polysaccharide solutions were
measured as follows: A cone plate (25 mm ) was used as a probe for the static rotation
viscosity, which was measured at room temperature (approximately 25.0 0.5 C) by a
rotation viscometer (HAAKE MARS2, Ger). The solution thickness was 1 mm
(thickness at the center position: 0.053 mm). The probe was pre-rotated at an angular
velocity of 0.01 rpm for 60 seconds before the measurement was started. The rotation
viscosity was recorded with changes in angular velocity from 10
-3
to 2 x 10
4
rpm.
Zero-shear viscosity was estimated by extrapolation of linear plots to zero velocity.
4-2-2: Results and discussion
Titration: The content of sulfate groups in sacran was estimated at 10 % by elemental
CHNS analysis in Chapter 3 [1]. Unfortunately, the concentration of carboxylate
groups cannot be estimated so simply since they are present in various sugar residues
such as in uronic acid, muramic acid, and others. Therefore, we performed a pH titration
of sacran sodium salt via the addition of HCl solution, and the pH, turbidity, and
conductivity were measured step by step. Fig. 4-1a shows the pH change of the sacran
solution as a function of the amount of added protons. The titration curve showed only a
79

simple line. In the course of the pH changes, the turbidity of the sacran solution
decreased initially, and showed a minimum at the inflection point of Fig.4-1b, but then
increased again to a stable level over pH 2.79. Although the turbidity of the sacran
solution initially decreased upon adding the transparent HCl solution, the carboxylate
groups may be simultaneously protonated. The protonated sacran chains may show
some partial aggregation, increasing the turbidity from pH 3.72 with increasing pH, but
became stable at pH 2.79. The content of carboxylate groups were determined by
conductivity titration as shown in Fig. 4-2. Fig.4-2a show the conductivity change of
sacran solution (150ml, 0.033 %) by adding HCl aq solution (pH=0.23). Conductivity, ,
increased with an increase in HCl concentration but the value is lower than of HCl
aq solution without sacran. The difference may be due to H
+
adsorption to carboxylate
ions of sacran. is plotted against concentration of HCl (Fig.4-2b) and is determined
to be 112 S/cm in the saturated adsorption of H
+
to carboxylates. Since molar
conductivity of H
+
is 349.8 S m
2
/mol, the value was converted into carboxyl
composition 17.6% to total sugar residue of sacran. The content was comparable with
the total content of uronic acids determined by the carbazole-sulfuric acid method
(22%).
Gelation under ions: In general, high-molecular-weight polymers are difficult to
dissolve in water because of their own strong segregation effects [2]. On the other
hand, sacran had both water-solubility and an ultra-high molecular weight, which can
lead to some unique properties. Although physical hydrogels of sacran were stable in
the still-standing state, they transited into the sol state by vigorous agitation, which was
confirmed by viscosity dependence on shear velocity (Fig.4-3a). On the other hand,
one can confirm from this figure that hyaluronic acid did not show a large velocity
80

dependence. The zero-shear viscosity was obtained by extrapolation of the plots in
fig.4-3a. The rotation viscosity measurement of the sacran sol (1 wt%) showed a very
high zero-share viscosity value (83 000 cps) whereas hyaluronic acid (Mw~150-180
kDa) showed a value of 8 900 cps. Such a large difference in the viscosity can be
attributed to the difference in the Mw. Although the addition of NaCl (0.9%) reduced
the viscosity of the hyaluronic acid solution to 4 400 cps as a result of charge screening
effects of salts [3], it raised the viscosity of the sacran sol to 153 000 cps (Fig.4-3).
This result indicated that the addition of NaCl had some crucial effects on the chain
state of sacran. As shown in Chapter 3, sacran chains adopt the extended
conformation under the salt and then sacran may be semi-rigid-rod as other natural
polysaccharides [4]. One can discuss that sacran chains recovered extended
conformation by the breakage of electrostatic association. Some reports pointed out the
existence of supergiant sugar chains with a Mw over 10
7
Da in nature [5] but their
properties have never been confirmed due to the low extraction efficiency. For
example, DNA is the highest Mw biopolymer [6], but the amount of DNA that can be
extracted with a Mw over 10
7
Da is too small to investigate its physical properties.
Sugar sequence and bonding pattern of sacran have not been determined yet since
cyanobacterial sugar chains generally had very complex structures composed of a great
variety of constitutive monosaccharides without repeating unit [7].
The sacran physical gel absorbed very large quantities of pure water (6100 mlg
-1
)
to its dry weight, as confirmed by the tea-bag method using a membrane with
submicroscaled pores 0.8 m smaller than most widely-used filters with microscaled
pores (Fig.4-4) [8]. The water absorption of hyaluronic acid is 1200 mlg
-1
to the dry
weight using the submicro-pored filter, and -carrageenan showed only 700 mlg
-1
.
81

We showed that the water absorption efficiency of sacran was quite high as compared to
these typical water-absorbable polysaccharides, which may be due to the
extremely-high Mw of sacran. Nanoloopes can hold water molecules efficiently as the
chemically cross-linked hydrogels inside of their compartments. Subsequently, we
showed that the value for saline absorption was as high as 2700 mlg
-1
to dry weight.
Although the saline absorption efficiency of sacran decreased to about one half of the
value for fresh water, the value was still very high. The saline absorption efficiency of
hyaluronic acid was 240 mlg
-1
, about one sixth lower than its pure water absorption,
which is a normal phenomenon due to the charge screening of polyelectrolytes by Na
and Cl ions to reduce their hydration performance. One hypothesizes that the small
decrease in sacran may be related to the salt-induced nanostructure transformation to
extend chains to micrometer-scale in length. Although the addition of NaCl could
break the nanoloop structures to attenuating the water absorbability, it contributes to
increasing the hydration performance by extending the chains. Furthermore, sacran
still showed high absorption values for other salines (0.9%) containing multivalent
metal ions such as Ca and Mg at 2000 and 2200 mlg
-1
, respectively. The result that a
lower saline-absorption value was observed for the larger valence ions may be due to
the stronger electrostatic forces with the polyanions to cross-link the sugar chains. In
the case of an artificial urine containing CaCl
2
(0.02%), MgSO
4
(0.04%), NaCl (0.8%),
and urea (2.0%), the saline absorption was also a high value at 2600 mlg
-1
(Fig.4-4).
The sacran physical gels was successfully wrapped in widely-used nonwoven cloth even
if no chemical cross-linker was used, and one can expect that sacran can be utilized as a
high-performance urine absorber. All the results shown above suggest that sacran
physical gels can be expected to have other various applications such as high
82

moisturizing agents, good thickening agents, high-performance cell scaffolds,
drug-release carriers in blood, and tree-planting materials for deserts. Here I discuss
relationship between water retention capacity and polymer chain forms under the
condition of water-retaining at maximum which is schematically illustrated at the
bottom of Fig. 4-4. I speculate that the condition of water-retaining at maximum can
be regarded as polymer form in c*. c* is the critical concentration at which chain
overlap occurs. As c* is closely correlated with molecular weight and discussed in
detail in the section of 4-3-3, the value of c* is calculated as 0.012 wt %. Whereas the
value of concentration, c
w
, where polymer chains retain water at maximum is calculated
as 0.016 wt % from the measured value of water-retention-capacity (6100 ml/g). These
values show a good agreement and the same result is obtained in hyarulonic acid
(Mw:2.0x10
6
g/mol; c
w
=0.083 wt %, c*=0.063 wt %), suggesting that my speculation
may be believable. As a result of my discussion, the high value of water retention
capacity in sacran is due to its ultra-high molecular weight leading to the high c*.

4-3: Liquid crystalline nature of extracted polysaccharides
4-3-1: Methods
Transmission electron microscopy (TEM): TEM images were obtained with a Hitachi
HF-2000 Field Emission Transmission Electron Microscope operated at an acceleration
voltage of 100kV at a magnification of 35,000x. The specimens were prepared by slow
evaporation of a drop of a diluted MeOH/Water (20/1 v/v) solution of sacran (ca. 10
ppm) on a carbon-coated copper mesh grid.
X-ray diffraction Wide angle X-ray diffraction (WAXD) patterns were recorded using
an X-ray diffractometer (RINT PowerX18) equipped with a scintillation counter using
83

Ni-filtered CuK radiation (40 kV, 100 mA; wavelength = 1.5418 ) in reflection
geometry. A 2 (: diffraction angle) scanning speed of 2
o
min
-1
with a sampling
interval of 0.02
o
was used. Specimens such as dried sacran fibers, the polysaccharide
solutions and In
3+
-bound wet gels inserted into a glass capillary (diameter: 1.5
mm, thickness: 0.01 mm, Markrhrchen aus Glas Nr.10, Germany) were held on the
non-reflective surface of a Si board.
Atomic force microscopy (AFM): All AFM experiments were performed using a
commercial AFM unit (SPA-400, Seiko Instruments, Japan) equipped with a calibrated
20 m xy-scan and 10 m z-scan range PZT-scanner. For the imaging of sugar chains
by AFM, a stiff cantilever (SI-DF20, Seiko Instruments, force constant is 13 N/m in
typical value, typical resonant frequency is 130 kHz, pyramidal tip shape, tip curvature
radius is 10 nm) were used and imaging was taken in the dynamic force modulation
(DFM) mode at optimal force. The scan speed was 2 m/sec.
Dielectric relaxation: The complex dielectric constant * of sacran aqueous solutions
was measured by an ac two-terminal method using an LCZ meter (HIOKI 3532-50).
The frequency ranged from 42 Hz to 5 MHz, and the applied voltage was 0.1 V. The
sample cell used in the present study was a coaxial type of cylindrical condenser with
stainless-steel electrodes. The dielectric measurement was carried out at room
temperature (r.t.) at approximately 25.0 0.5 C.
Optorheometry: Optometry was equipped with the abovementioned rheometry
apparatus. The relative birefringence change of the polysaccharide solution as a function
of shear velocity was recorded by measuring the intensity of white light transmitted
through the solution (gap: 0.3 mm) under cross-nicol polarimetry. The measurement
system included a rheometer with optics, as schematically illustrated in Fig.4-5.
84

Dynamic viscosity measurements were made by changing the frequency at room
temperature using the same systems under static mode. The strain amplitude was set to
10 %.
4-3-2: Results in measurement of liquid crystallinity
Fig.4-6a shows a microscopic photo of sacran fibers in the dry state, taken under the
cross-nicol using a first-order retardation plate (=530 nm) inserted into the light path.
The polarized microscopic observation shows oriented fibers with a thickness of less
than 10 m. The fiber birefringence is negative, as evidenced by both subtractive
birefringence (blue color) in the fiber lying from the upper left to the lower right and
additive birefringence (orange color) in the fiber lying from the upper right to the lower
left. The negative birefringence strongly suggests that sacran backbones lie along the
fiber axis. X-ray diffraction diagrams of the fibers in dry states showed only a broad
halo without crystalline peaks, indicating sacran fibers were non-crystalline (Fig.4-7),
presumably due to many kinds of sugar residues as demonstrated in the Chapter 3.
The dry fibers easily dissolved in hot water to create very viscous and translucent
solution. Fig.4-6b shows crossed-polarizing microscopic photos of sacran solutions
with a concentration of 0.5 wt %. The bright region contains many sacran rods lining
up but darkened by a 45
o
(or 135
o
) rotation for lines to be parallel with a polarizer (or
analyzer) axis. The aforementioned phenomenon is characteristic of self-orientation.
X-ray diffraction diagrams of microfibers in aqueous solutions still showed only a broad
halo without crystalline peaks, indicating microfibers were still non-crystalline (Fig.4-7).
The birefringence in the solution is generally caused by crystalline particle dispersions
or the molecular orientation of solutes. In the present case, X-ray studies did not
confirm crystalline particles. Fig. 4-6c shows the image of the sacran aqueous solution
85

with a concentration of 0.5 wt % (3 ml) put on a glass dish with a round bottom. The
solution was irradiated (brightly) by the white backlight under crossed-polarizers, to
show a birefringence and continuous brushes around point defects with a disclination of
s = 1/2 (arrows). These textures including brushes, so called Schlieren textures,
indicate that sacran chains form the nematic LC phase [9]. We can speculate that the
LC phase may be attributed to a special form of sacran, as a microbial polysaccharide,
schizophyllan, which forms a rigid triple helix to show an LC phase [10].

It was
noticeable that the domain surrounded by brushes was very big (millimeter to several
centimeter scale) even without any treatment by external forces, i.e. sacran chains
automatically aligned in the macroscopic range.
In order to analyze the structure of sacrans in more detail, AFM images of sacran
chains dried from solutions with concentrations of 1 ppm on mica substrates were
observed. Fig. 4-8a shows sacran chains as whitish lines, indicating a few sacran
chains formed bundles with twisted morphologies (a representative bundle is indicated
by the arrow; ca. 3 m length). Close-up images taken by TEM revealed a double
helix-like form of the sacran bundle (Fig. 4-8b), where helixes were loosely wound as
shown in the right illustration of Fig. 4-8b, and its thickness and helical pitch are about
20 nm and 120 nm, respectively. The thickness is too large comparing to conventional
double helix such as DNA duplex, then I can guess that the helix may be regarded as
coiled-coil structure formed by sacran chain bundles. The driving force for the bundle
formation is not clear but we can speculate that interchain interactions of sacran may be
related with functional groups such as uronic acid, sulfated sugars, and amino sugars,
which were all detected in the Chapter 3.

Helical chains generally behave like rigid rods
seen as almost-straight lines and possibly helical chains contributed to exhibit LC
86

phases in sacran solutions. We attempted to catch bundle orientations as AFM images
using specimens dried from sacran solution at c = 0.01 wt % and obtained Fig. 4-8c
where many twisted ropes of sacran chains lined up, suggesting a close relationship
between sacran helixes and LC.
4-3-3: Results in dielectronic measurement
When sacran transform from single chains to coiled-coil structures, the ionic
environment around sacran chains also changes. In order to investigate the ionic
environment of sacran chains under various concentrations, we investigated the
dielectric behavior of sacran solutions with changing concentrations.
The relative complex dielectric constant * is defined by the following equation,
" ' *
i + =
(1).
Here, and stands for the real and imaginary part of the relative complex dielectric
constant, respectively. The frequency dependence of the dielectric constant of sacran
aqueous solution with 0.02 wt.% (as-measured data) is shown in Fig. 4-9. The dielectric
constant in lower frequencies took a huge value of ~10
3
and it largely decreased
satisfying with a power law as '~
-n
(1.4<n<1.6 in the present experiment). It is well
known that ion-blocking electrodes give rise to a large frequency-dependent
polarization at low frequencies, which is called the electrode polarization effect [11].
The large dielectric constant of the sacran aqueous solution observed here is clearly
interpreted as the electrode polarization effect, not the intrinsic value of the dielectric
constant. As described in the introduction, the dielectric response of the sacran on which
we are focusing is buried under the large electrode polarization. Accordingly, we
decomposed the as-measured dielectric constant into the dielectric constant of sacran
aqueous solutions '
s
and that of the electrode polarization '
EP
;
87

'
EP
' '
+ =
s (2)
In order to extract the dielectric constant of sacran '
s
, we used the dielectric constant of
KCl aqueous solutions '
KCl
. This is the method assuming that the component of the
electrode polarization is equal to the dielectric constant of saline solutions with nearly
the same electric conductivity as the sample, '
EP
'
KCl
.
Typical dielectric spectrum for the KCl aqueous solution is also shown in Fig. 4-9.
Similar electrode polarization effect was observed in the dielectric spectrum of HCl
aqueous solution as well as sacran. However, note that the dielectric constant of KCl
aqueous solution was clearly lower than that of sacran in the frequency range of 10
2
-10
5

Hz. This strongly suggests that the sacran aqueous solution shows a low-frequency
dielectric relaxation in the frequency range. The dielectric constant of the sacran at 10
6

Hz was 85, which was equal to the dielectric constant of the KCl aqueous solution (=85)
or pure water (=84). This means no dielectric relaxation due to the counterions which
were dissociated from sacran occurs above 10
6
Hz. The inset in Fig. 4-9 shows the
dielectric spectrum of sacran aqueous solutions '
s
after subtracting the dielectric
constant of KCl aqueous solutions '
KCl
from the as-measured data '. Large
low-frequency relaxation and high-frequency one were observed around at 100 Hz and
100 kHz, respectively. Based on the analysis method, I could get the dielectronic
spectra of sacran with various concentrations (Fig.4-10a). Both relaxations were fitted
by the following equation consisting of two Debye relaxations [12]:

+
+

+
+ =
) 2 / cos( cosh
sinh
1
2

) 2 / cos( cosh
sinh
1
2

H
H H
L
L L '
w
'
s


x
x
x
x

Where x
L,H
=ln
L,H
; is the angular frequency of the electric field,
L,H
is the mean
relaxation time of low or high frequency relaxations. '
w
is the dielectric constant of
88

pure water ('
w
= 82). '
L,H
is the dielectric increment of low or high relaxations,
respectively. It is widely accepted that polyelectrolyte solutions exhibit two dielectric
relaxations; one is a low-frequency relaxation in the kHz range and another is a
high-frequency relaxation in the MHz range. The low-frequency relaxation arises
from fluctuations in bound counterions along the polymer chain. The high-frequency
relaxation is due to fluctuations in loosely bound counterions perpendicular to the
polymer chain [13, 14]. According to the fitted equation, the low-frequency relaxation
time of very dilute solutions of c= 0.002 wt % were determined to be =8.1 ms. The
aforementioned result indicates that the fluctuation length
L
of bound counterions
along the sacran chain is 4.5 m using a relation of
L L
D 2 = ; D is the diffusion
constant of Na
+
ion in a free medium (1.2210
-9
m
2
/s). According to the theory of
Mandel [15]

, the contour length
L
12

corresponding to the length of the sacran chain
was calculated as 16 m which is in the same order of the length obtained from AFM
(Fig.3-13), that is, the low-frequency relaxation is due to the counterion fluctuation
along sacran chains (Fig.4-11a and 4-11b). Such long

have been observed in another
giant macromolecule, Na-DNA ( = 68 ms: degree of polymerization 12000) [16] or a
polyelectrolyte gel (0.7 < < 2.1 ms) [17], suggesting that sacran has a long and
electrostatically continuous chain.
Fig. 4-10b shows the dielectric constant at 100 kHz for sacran aqueous solutions
under various concentrations. The dielectric constant gradually increased with the
sacran concentration but the increasing rate clearly rose at c = 0.09 wt %. In addition,
the relaxation time of bound counterions along sacran chains suddenly shortened
between 0.08-0.09 wt % (Fig.4-12a). Actually, the relaxation time of bound
89

counterions at c = 0.09 wt % was determined to be 0.95 ms, which corresponds to a
fluctuation length of 1.5 m. The helical structure of sacran chains might restrict the
long fluctuation of bound counterions. We can speculate that some association points
of anions and cations might make terminals on fluctuations pass as schematically shown
in Fig. 4-11c. Otherwise, the short-time relaxation may be caused by the enhanced
mobility of bound counterions because the electrostatic potential valley along the chain
is weakened by electrostatic associations of sacran chains.
Fig. 4-10c shows the dielectric constant at 2 MHz for sacran aqueous solutions with
various concentrations. The dielectric constant remained constant below 0.02 wt %
but suddenly increased at c = 0.02 wt %. Since the Mw of sacran is very high, 1.6 x
10
7
, c* the critical concentration at which chain overlap occurs is very low 0.012 wt %
as calculated using the following equation [18]:
c* a
-3
N
-4/5
,
Where a is the length of the monosaccharide (0.65 nm), and N is the number of
monosaccharide residues in sacran (8.9 x 10
4
). The calculated c* agrees with the
critical value of 0.02 wt %. We estimated the high-frequency relaxation time using the
same procedure for estimating the low-frequency relaxation. High-frequency
relaxation times at c = 0.002 and 0.09 wt % were estimated to be 1.1 s and 54 ns,
which correspond to fluctuation lengths of 52 and 11 nm, respectively. Loosely bound
counterions have short relaxation times and fluctuation lengths at c > 0.02 wt %, which
originate from the decrease in the correlation length of sacran chains and loosely bound
ions fluctuate among chains are shown in Fig. 4-12b.
4-3-4: Results in rheooptical measurements
We next investigated viscosities and polarized optical behaviors of sacran solutions
90

under various concentrations in order to determine the critical concentration of liquid
crystallization. Fig. 4-13a shows the shear rate dependence of rotation viscosities of
sacran solutions with concentrations ranging between 0.06 - 1.5 wt %. Sacran solution
viscosities increased upon increasing concentrations. Fig. 4-14a shows zero shear
viscosities, which were estimated by extrapolating viscosity data at a shear rate range of
10
-3
-10
-2
1/s plotted against concentration. Zero shear viscosity also showed an increase
with concentration but with an inflexion point around a concentration of 0.25-0.5 wt %.
Since the inflexion point may relate with the liquid crystallization of sacran solutions,
their polarized optical behaviors were investigated. In order to evaluate the orientation
state of sacran chains, the intensity of white light transmitted through sacran solutions
under crossed polarizers was measured by a photometry system equipped with a
rheometer as illustrated in Fig.4-5. Fig. 5b shows the light intensity change as a
function of shear rate. Light intensities of sacran solutions with concentrations of 1.5,
1.0 and 0.5 wt % were higher than those of 0.25, 0.13, and 0.06 wt % when shear rates
were very low. Then we estimated the zero shear intensity by extrapolating intensity
data at a shear rate range of 10
-3
-10
-2
1/s and plotted it against concentrations (Fig.
4-14b). Zero shear intensities of sacran solutions jumped up at concentrations of
0.25-0.5 wt %, indicating that the orientation of sacran chains increased drastically in
this concentration range. The light intensity under the condition of mono-domain
orientation at a shear rate of 10 1/s also showed jump-up at the same concentration
range with zero-shearing. These results indicate the critical concentration of liquid
crystallization of sacran solution lay in 0.25-0.5 wt % where the inflexion point
appeared in the zero shear viscosity plot (Fig. 4-14a). In Fig.4-15, volume fractions at
liquid crystalline transition in various polymers are summarized. One can see that
91

sacran showed much lower volume fraction than conventional liquid crystalline
polymers.
The sacran solution showed a thixotropic effect; the viscosity of sacran solutions with
concentrations of 0.06, 0.13, 0.25, 0.5, 1.0, and 1.5 wt % showed 3.3, 3.8, 4.1, 4.5, 5.5,
and 6.4 times decrease, respectively, with a 10 times increase in shear rate (thixotropy
index; Fig.4-16). The thixotropic effect may be attributed to the interchain sliding of
sacran single chains and sacran helixes, which can orient along the shear flow. The
high thixotropic value in concentrations higher than 0.5 wt % may be related to the LC
state where sacran chains orient very smoothly. An increase in the orientation degree
of sacran solutions by shear rate increase was confirmed by cross-nicol polarimetry and
images of Fig. 5c show the intensity increase of light transmitted through sacran
solutions with a shear rate increase. The quantitative change in light intensity in
Fig.4-13b shows two-step viscosity increases in sacran solutions in LC states with
concentrations of 0.5, 1.0, and 1.5 wt % and an increase in shear rate. The two-step
increase was not seen in other polysaccharides such as xantham gum or hyaluronic acid
(concentration: 0.5 wt %, Fig.4-17). The first increase in light intensity occurred in a
shear rate range of 10
-2
10
0
1/s while the second increase occurred in a shear rate
range of 10
2
10
4
1/s. From the equation of Zimm relaxation time [19],
Zimm
:

Zimm

s
R
g
3
/k
B
T,
Where
s
is viscosity, R
g
is the mean square radius of polymer chains, k
B
is the
Boltzmann constant, and T is the absolute temperature, we calculated the R
g
scale of
sacran chains showing light intensity increases. The first increase occurred in the R
g

scale range between 100-300 nm which corresponds with the mean square radius of
random-coiled sacran chains (200 nm) [1],

while the second increase occurred in the R
g

92

scale between 30-50 nm which was in the same order of helical associates thickness.
As a consequence, we can summarize that the first increase in light intensity is
attributed to the macroscopic orientation of sacran chains in the LC state while the
second light intensity increase is attributed to the extension of sacran chains by strong
shear stress.
We measured dynamic moduli of sacran solutions under various concentrations in
order to investigate gelation behaviors (Fig.4-18). Storage moduli, G, of sacran
solutions were higher than loss moduli, G in concentrations of 0.5, 1.0, and 1.5 wt %
while G is almost equal to G in concentration of 0.25 wt % and is lower than G in
concentrations of 0.13 and 0.06 wt %. These aforementioned results indicate that
sacran solutions adopted a gel state in concentration ranges more than 0.25 wt%.
Although the formation mechanism of the cross-linking junction is not clear, we can
assume that the bridge formation of coiled-coil aggregates, which is faintly seen in Fig.
2c (arrows), may be a driving force for gel formation. Gels were very soft and easily
deformed without breaking, suggesting that sacran chains can form coiled-coil
structures dynamically and interchain interactions may not be very strong.
4-3-5: Discussion about solution behavior
Based on illustrations from Fig.4-11, we can discuss structural changes in sacran
solutions as a function of concentration. Dielectric relaxation studies indicated c* is
0.02 wt %. In concentration ranges between 0.02-0.09 wt %, sacran chains have a
number of opportunities to overlap one another but do not show association behaviors
(Fig.4-11b). Over 0.09 wt %, sacran chains form rigid coiled-coil structures
(Fig.4-11c) and this coiled-coil fraction may increase with the concentration. At c =
0.25 wt %, the sacran solution becomes a critical state of sol-gel phase presumably due
93

to an increased fraction of chain entanglement (Fig.4-11d) and Fig.5b shows a slight
increase of the birefringence with shear rate due to the presence of physical
cross-linking junctions. The nematic LC phase is exhibited when the concentration
exceeds 0.5 wt % (Fig.4-11e). The concentration showing the LC phase is extremely
low because xanthan gum and shizophyllan which are well-known high-performance
LC polysaccharides showed an LC phase above 6 wt % [20] and 13 wt% [21],
respectively. Even crystallite rods of charged celluloses with an average length of 115
nm showed a critical LC concentration of 5 wt % [22]. According to the literature [22,
23], electrostatic interactions of mesogenic chains are important to show the LC phase
in dilute solutions. The efficient LC formation of sacran chains may be attributed to
electrostatic interactions with other polysaccharides and to ultra-high Mw and lengths of
sacran chains, so we calculated the aspect ratio, X, of sacran mesogenic rods from
Florys lattice theory [24]:
= 8/X(1-2/X) 8/X,
where is the volume fraction. When of the sacran solution is 0.005, X is estimated to
be 1600 (Fig.4-15) which is an anomalous value since other LC polysaccharides such as
Xanthan gum and shizophyllan show X=517 and X=95, respectively [18]. The
thickness of the sacran coiled-coil structure is estimated to be about 20 nm from Fig. 2,
and the persistence length of sacran mesogenic chains in the LC state is calculated as 32
m, which is longer than the calculated and observed length of sacran rods. Sacran
mesogenic chains may be derived from the bridged sacran rope and show a LC effect.
Thus huge LC domains in centimeter scale are constructed in sacran solution under no
external treatments. Fig.4-15 shows volume fraction for liquid crystal transition in
several rigid-rod polymers with various aspect ratios [25]. According to these data,
94

liquid crystalline polymers with single strand helix and shizophyllan with a triple strand
helix all showed a disagreement with a Florys lattice theory curve (dotted line) under
the condition of too high aspect ratio. The literature [25] discussed the phenomenon
shown below; the entropical effects on decreasing the persistence length of polymer
chain mesogens which may deform at some portions in very long chains. Then I can
speculate that sacran coiled-coil structure may be so rigid that very low volume fraction
below 0.005 came true for the first time.
Finally we confirmed that the white backlight transmitted through the discolored
region during cross-nicol polarimetry (Fig. 4-19), indicated that A. sacrum biomaterials
show a birefringence. Thus we inferred that the main component of A. sacrum, sacran,
might be in the LC state in vivo and giant rods may reinforce ECMs since the ECM of A.
sacrum has a water content of 97.5 - 98.3 wt % (c = 1.7 - 2.5 %). Moreover A. sacrum
biomaterials show strong laser-light scattering (Fig.4-20) presumably from sacran rods
in the LC state, like LC gels, which are effective upon the dispersion of irradiated light
in floating colonies of phototrophic A. sacrum cells.

4-4: Conclusion
Sacran solution showed a high zero-shear viscosity (83 000 cps, 1 wt%) which
increased by addition of salts. The water retention capacity was also high 6100 mlg
-1
to
the dry weight. Saline retention capacity of sacran was 2700 mlg
-1
and 10 times than
that of hyaluronic acid. Surprisingly sacran showed still a high capacity for urine
retation (2600 mlg
-1
) even if in the presence of divalent ions such as calcium and
magnecium ions. Moreover I found sacran aqueous solution showed liquid crystalline
phase. Sacrans were observed as self-orienting micro rods longer than 3 m in dilute
95

solution at c = 0.01 wt % by optical microscopes. Sacran chains form coiled-coil
structures at c > 0.09 wt % and form huge domains (centimeter scale) of liquid crystals
at c > 0.5 wt % which was quite low when compared to conventional liquid crystalline
polysaccharides. Mesogenic helical chains of sacrans had extremely high aspect ratios
of 1600 for highly persistent lengths of 32 m. The sacran mesogen was longest in all
the liquid crystalline molecules reported thus far.
References
[1] M. K. Okajima, T. Bamba, Y. Kaneso, K. Hirata, E. Fukusaki, S. Kajiyama, T.
Kaneko, Supergiant Ampholytic Sugar Chains with Imbalanced Charge Ratio Form
Saline Ultra-absorbent Hydrogels, Macromolecules, 41, 4061-4064 (2008).
[2] P. -G. De Gennes, Scaling Concepts in Polymer Physics Cornel Univ. Press. Ithaca
and London (1979).
[3] Y. -Q. Zhang, T. Tanaka, M. Shibayama, Super-absorbency and phase transition of
gels in physiological salt solutions Nature 360, 142-144 (1992).
[4] J. R. E. Fraser, T. C. Laurent, U. B. G. Laurent, Hyaluronan: its nature, distribution,
functions and turnover, J. Intern. Med. 242, 27-33 (1997).
[5] For example: K. L. Smiley, J. A. Boundy, D. E. Hensley Action patterns of
immobilized dextranase, Carbohydr. Res. 104, 319-324 (1982).
[6] J. Richter, R. Seidel, R. Kirsch, M. Mertig, W. Pompe, J. Plaschke, H. K. Schackert,
Nanoscale Palladium Metallization of DNA, Adv. Mater. 12, 507-510 (2000).
[7] R. De Philippis, M. Vincenzini, Exocellular polysaccharides from cyanobacteria and
their possible applications FEMS Microbiol. Rev. 22, 151-175 (1998).
[8] Y. Nohata, J. Azuma, R. Kurane, Structural studies of a neutral polysaccharide
produced by Alcaligenes latus, Carbohydrate Res. 293, 213-222 (1996)
96

[9] I. Dierking, Textures of Liquid Crystals Wiley-VCH, Verlag, (2003).
[10] T. Norisuye, Triple-stranded helical structure of schizophyllan and its antitumor
activity in aqueous solution, Makromol. Chem. Suppl. 14, 105-118 (1985).
[11] For example, W. Scheider, Theory of the frequency dispersion of electrode
polarization. Topology of networks with fractional power frequency dependence, J.
Phys. Chem., 79, 127-136 (1975).
[12] K. S. Cole, R. H. Cole, Dispersion and absorption in dielectrics, J. Chem. Phys. 9,
341-351 (1941).
[13] Y. Nagamine, K. Ito, R. Hayakawa, Low- and high-frequency electric birefringence
relaxations in linear polyelectrolyte solutions, Langmuir, 15, 4135-4138 (1999).
[14] Y. Nagamine, K. Ito, R. Hayakawa, Low- and high-frequency relaxations in linear
polyelectrolyte solutions with different counter-ion species, Coll. Surf. A
Physicochem. Eng. Aspects 148, 149-153 (1999).
[15] M. Mandel, The electric polarization of rod-like charged macromolecules, Mol.
Phys. 4, 489-496 (1961).
[16] M. Sakamoto, H. Kanda, R. Hayakawa, Y. Wada, Dielectirc relaxation of DNA in
aqueous solutions Biopolymers 15, 879-892 (1976).
[17] T. Mitsumata, K. Ikeda, J. P. Gong, Y. Osada, Low-frequency dielectric relaxation
of polyelectrolyte gels, J. Phys. Chem. 102, 5246-5251 (1998).
[18] P. G. De Gennes, Physics of Liquid Crystals. Oxford Univ. Press, Oxford (1995).
[19] M. Doi, S. F. Edwards, The Theory of Polymer Dynamics Oxford, Clarendon,
(1986).
[20] K. Van, T. Norisuye, A. Teramoto, Liquid Crystal Formation in Aqueous Solutions
of a Polysaccharide Schizophyllan, Mol. Cryst. Liq. Cryst. 78, 123-124 (1981).
97

[21] X. M. Dong, D. G. Gray, Effect of Counterions on Ordered Phase Formation in
Suspensions of Charged Rodlike Cellulose Crystallites, Langmuir 13, 2404-2409
(1997).
[22] R. Oertel, W. M. Kulicke, Viscoelastic properties of liquid crystals of aqueous
biopolymer solutions, Rheol. Acta. 30, 140-150 (1991).
[23] A. Teramoto, T. Sato, Liquid crystal formation in semiflexible polymer solutions:
effects of chain stiffness, electrostatic interaction, and polydispersity, Lecture notes
in physics 415, 399-410 (1993).
[24] P. J. Flory, Phase Equilibria in Solutions of Rod-Like Particles, Proc. Roy. Soc. A
234, 73-89 (1956). b) P. J. Flory, Molecular theory of liquid crystals,.Adv. Polym.
Sci. 59, 1-36 (1984).
[25] A. Teramoto, M. Kobayashi, T. Norisuye, (Ed.) Ordering in Macromolecular
Systems pp156, Springer-Verlag Berlin and Heidelberg GmbH & Co. K. (1994).
6
6.5
a
3
3.5
4
4.5
5
5.5
6
p
H
a
1
1.5
2
2.5
3
0.00001 0.0001 0.001 0.01 0.1
Amount of added protons (mol)
b
pH =2.79
p ( )
0 18
0.2
0.22
T
u
r
b
i
d
i
t
y
pH =3.73
0.12
0.14
0.16
0.18
Amountofaddedprotons(mol)
0.1
0 0.001 0.002 0.003 0.004 0.005
Figure 41 pH titration curves of sacran solution Figure41.pHtitrationcurvesofsacransolution
(a)anditsturbiditychangeasafunctionofpH.
98
1000
1200
c
m
)
a
400
600
800
c
t
i
v
i
t
y
,

S
/
c

0
200
0 0.5 1 1.5 2 2.5
C t ti f HCl ( M)
C
o
n
d
u
c
ConcentrationofHCl (mM)
100
120
140
m
)
b
40
60
80

S
/
c
Fi 4 2 C d ti it tit ti
0
20
0 0.5 1 1.5 2
ConcentrationofHCl (mM)
Figure42.Conductivitytitrationcurve
ofsacransolution.
99
10
5
c
p
s
)
Sacran
pure water
saline
a)
10
4
S
h
e
a
r

v
i
s
c
o
s
i
t
y

(
c
Hyaluronan
purewater
li
purewater
Shear velocity(1/s)
0.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0
10
3
S
saline
Shear velocity (1/s)
Sacran
s
i
t
y

(
c
p
s
)
b)
Z
e
r
o

s
h
e
a
r

v
i
s
c
o
s
Figure43.Shearvelocitydependenceonshear
Hyaluronan
NaCl concentration (wt%)
Z
g y p
viscositiesofsacranandhyaluronan solutionunder
varioussaltconcentrations.
100
7000
3000
4000
5000
6000
n

c
a
p
a
c
i
t
y

(
m
l
/
g
)
PureNaClCaCl
2
MgCl
2
MgSO
4
Artificial PureNaCl
water urineWater
0
1000
2000
R
e
t
e
n
t
i
o
n
sacran hyarulonan
Polymerchains
water
Figure44.Retentioncapacityofvarioussaltsolutions
f d h l C diti th t
101
forsacranandhyaluronan.Conditionthatsacran
chainsretainwateratmaximum(bottom)
White Light
Photometry
30 25
Polarizer
Glass Plate (parallel)
Sample
0.3
5
Glass substrate
Analyzer
Rheometry
CCD camera
Figure45.Schematicillustrationofphotometry
systemequippingwithrheometry formeasuring
relativebirefringencechangesofpolysaccharide
solutionasafunctionofshearstress.
102
500 m
a
b
c
45
o
disclination
1 cm
cross-nicol
Figure46.a)Crossedpolarizingmicroscopic(PLM)
photosofdriedfiberswithafirstorderretardation
plate(=530nm).b)PLMimagesofsacranaqueous
solutionwith0.5wt%.Left:brightregion.Right:
bright region darkened upon a 45
o
rotation. c) brightregiondarkenedupona45 rotation.c)
Schlierentexture(arrows)ofsacranaqueoussolution
with0.5wt%takenundercrossnicolpolarimetry.
103
sacranfibers
In/sacrangel
sacran sol
/ g
Diffractionangle2/deg
Figure47.Xraydiffractiondiagramsofsacranfibers,
sacran solution, sacransolution,
andgelsofsacranwithInions(Chapter5).
104
a b c
Figure 48. Microscopic images of sacran chains. a)
Atomic force microscopic (AFM) images of sacran Atomic force microscopic (AFM) images of sacran
chains dried from very dilute solutions on mica
substrates (Arrow: a representative bundle of sacran
chains). b) Transmission electron microscopic (TEM)
f d d b d images of specimens dried on a carboncoated Cu
grid. Illustration: helical form of sacran chains
speculated as the basis of TEM images. c) AFM
images of sacran chains dried from more g
concentrated solutions than the solution in fig.a.
Arrows show sacran chains bridging sacran ropes.
105

10
4
200
250
300
10
3

'
50
100
150
200
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
2
10
1
10
2
10
3
10
4
10
5
10
6
10
7
f (Hz)
Figure49.Frequencydependencesofthereal
partofdielectricconstantforsacranandKCl
aqueoussolutions.Inset:Dielectricspectrum
of sacran aqueous solution after subtracting ofsacranaqueoussolutionaftersubtracting
theelectrodepolarizationeffect.
106
a
140
100
110
120
130
0.002wt%
0.004wt%
0.01wt%
0.02wt%
0.04wt%
0.06wt%
0.08wt%
0.09wt%
0.1wt%
0.2wt%
0.4wt%
0.6wt%

'
r
80
90
100
10
3
10
4
10
5
10
6
10
7
f (Hz)
b
c
110
120
130
140
100 kHz

'
r
90
95

'
r
2 MHz
80
90
100
10
3
10
2
10
1
10
0
c (wt%)

80
85
10
3
10
2
10
1
10
0
c (wt%)
c(wt%)
c(wt%)
Figure 410. Dielectric constants for sacran aqueous
solutions with various concentrations. (a) Dielectric
spectra, (b) concentration dependence of dielectric spectra, (b) concentration dependence of dielectric
constant at 100 kHz, and (c) concentration
dependence of dielectric constant at 2 MHz.
107

sacran

a b
+

counter ions

sacran
chain
+

t i

0.02 wt%

c*
counter ions
counter ions

0 09 wt%

c
helical (rigid-rod)
0.09 wt%

0.5 wt% 0.25 wt%


d
e
LC gels
gels
Figure411.Schematicillustrationofsacranself g
organizationuponaconcentrationincrease.Black
points:aminosugarresidues.
108
10
-2
a:100kHz
10
-3
L

(
s
)
10
-4

L
10
-5
b:2MHz
10
-7
10
-6

H

(
s
)
10
-8
10
7
10
-3
10
-2
10
-1
10
0
( t%)

c (wt%)
Figure412.Concentrationdependenceof
l ti ti f th l ti ) relaxationtime, ,ofthesacransolutions.a)
Frequency:100kHz,b)Frequency:2MHz
109
10
2
10
3
Steady State Measurement
1.5 wt.%
1.0wt.%
05wt %
a
(
P
a

s
)
120
140
Birefrincence Intensity
1.5wt.%
1.0wt.%
0.5wt.% [
a
.
u
.
]
b
a
.
u
.
)





Shear rate (1/s)
10
-1
10
0
10
1
0.5wt.%
0.25wt.%
0.13wt.%
0.06wt.%


[
P
a
s
]
o
n

v
i
s
c
o
s
i
t
y

(
40
60
80
100
0.5wt.%
0.25wt.%
0.13wt.%
0.06wt.%
r
i
n
g
e
n
c
e

I
n
t
e
n
s
i
t
y

g
h
t

i
n
t
e
n
s
i
t
y

(
a
c
10
-3
10
-2
10
-4
10
-2
10
0
10
2
10
4
Shear Rate [1/s]
R
o
t
a
t
i
0
20
10
-4
10
-2
10
0
10
2
10
4
B
i
r
e
f
r
Shear Rate [1/s]
Shear rate (1/s)
L
i
g
c
0 10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
1.510
4
Shear rate (1/s)
crossed
polarizer
shear
Figure 413 a) Shear rate dependence of rotation
viscosity for sacran solutions with various
concentrations. b) Intensity of white light
i d h h l i i h i transmitted through sacran solutions with various
concentrations, measured under a crossed polarizer.
c) Crossedpolarizing images of white light
transmitted through sacran solutions with a g
concentration of 0.5 wt %. Dotted lines show critical
concentrations in a lightintensity increase.
110
10
2
10
3
Zero-shear Viscosity
i
t
y

[
P
a
s
]
a
s
i
t
y

(
P
a

s
)
10
1
10
r
o
_
s
h
e
a
r

V
i
s
c
o
s
i
s
h
e
a
r

v
i
s
c
o
s
10
0
10
-2
10
-1
10
0
10
1
z
e
r
Concentration [wt.%]
30
Zero-shear Birefringence Intensity [a.u.]
b
Concentration (wt%)
Z
e
r
o

)
15
20
25
r

I
n
t
e
n
s
i
t
y

[
a
.
u
.
]
b
n
t
e
n
s
i
t
y

(
a
.
u
.
)
0
5
10
10
-2
10
-1
10
0
10
1
Z
e
r
o
_
s
h
e
a
r
Z
e
r
o

s
h
e
a
r

i
n
10 10 10 10
Concentration [wt.%]
Concentration (wt%)
Figure 414. a) Change in zero shear viscosity for
sacran solutions as a function of concentrations. b)
Change in intensity of white light transmitted through g y g g
sacran solutions under a crossed polarizer as a
function of concentrations.
111
0.1
1
n

OtherLCpolymers
(singlechains)
0.01
V
o
l
u
m
e

f
r
a
c
t
i
o
n
Frolys
latticetheory
Shizophyllan
(triplehelixes)
0.001
1 10 100 1000 10000 100000
Aspectratio X
V
sacran(=0.005)
p
Figure415. Volumefractionforliquid
crystaltransitioninrigidrodpolymers
ith i t ti
112
withvariousaspectratios.
i
n
d
e
x
6
6.5
7
T
h
i
x
o
t
r
o
p
y
4
4.5
5
5.5
Concentration(%)
3
3.5
0 0.5 1 1.5 2
Figure416.Plotts ofthixotropy index
againstsacranconcentration
i l ti inaqueoussolution.
113
0.5wt.%
150
200
Montnant
Xanthan Gum
Hyaluronan
a
.
u
.
]
Sacran
100
Hyaluronan
f
r
i
n
g
e
n
c
e

[
a
n
s
i
t
y

(
a
.
u
.
)

0
50
B
i
r
e
f
L
i
g
h
t

i
n
t
e
0
10
-3
10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
10
5
Shear Rate [1/s]
Shearrate(1/s)
Figure417. Intensityofwhitelighttransmitting
throughpolysaccharidesolutionswitha
concentrationof0.5wt%,measuredunder
crossedpolarizer.
114
10
2
10
1
1.5 wt.%
G' tan
10
2
10
1
1.0wt.%
G' tan
10
2
10
1
0.5wt.%
G' tan


(
P
a
)
a: 1.5 wt% b: 1.0 wt% c: 0.5 wt%
G
G
10
-2
10
-1
10
0
10
1
10
-1
10
0
10
-3
10
-2
10
-1
10
0
10
1
10
2
G
G"
tan
G
'
,
G
"
[
P
a
]
t
a
n

f [Hz]
10
-2
10
-1
10
0
10
1
10
-1
10
0
10
-3
10
-2
10
-1
10
0
10
1
10
2
G
G"
tan
G
'
,
G
"
[
P
a
]
1
0
w
p
T
a
n

d
e
l
f [Hz]
10
-2
10
-1
10
0
10
1
10
-1
10
0
10
-3
10
-2
10
-1
10
0
10
1
10
2
G"
G
'
,
G
"
[
P
a
]
t
a
n

f [Hz] L
o
s
s

M
o
d
u
l
i

G

T


a

G
G
G
G
G
G
Tan
Tan
Tan
f [Hz] f [Hz] f [Hz]
10
-1
10
0
10
1
10
2
10
0
10
1
0.13wt.%
G'
G"
tan
G
'
,
G
"
[
P
a
]
t
a
n

10
-1
10
0
10
1
10
2
10
0
10
1
0.25wt.%
G'
G"
tan
G
'
,
G
"
[
P
a
]
t
a
n

10
-1
10
0
10
1
10
2
10
0
10
1
0.06wt.%
G'
G"
tan
G
'
,
G
"
[
P
a
]
t
a
n

a
g
e

M
o
d
u
l
i

G

,

L
n



















d: 0.25 wt% e: 0.13 wt% f: 0.06 wt%
G
G
G
G
Tan Tan
Tan
10
-2
10
-1
10
-3
10
-2
10
-1
10
0
10
1
10
2
f [Hz]
10
-2
10
-1
10
-3
10
-2
10
-1
10
0
10
1
10
2
f [Hz]
10
-2
10
-1
10
-3
10
-2
10
-1
10
0
10
1
10
2
f [Hz] S
t
o
r
a










F r e q u e n c y (Hz)
G
G
G
Figure 418. Dynamic moduli for sacran solutions with
various concentrations measured at r.t. In all plots, 1
st
and 2
nd
vertical axes show dynamic moduli and tan and 2 vertical axes show dynamic moduli and tan ,
respectively, and the horizontal axis shows frequency.
115
a
PP (-)
PP (+)
1cm
PP ()
bb
cross-nicol
Figure 419.AppearanceofAphanothecesacrum
biomaterialpartiallylostphycobiliproteins PP
duringcultivation.(a)Normalimage.(b)Crossed
polarizingimage.Thebirefringencewaswell
confirmedinPP()regioninA.sacrum .
116
laser
1 cm
Scatteredlight
Figure 420.AppearanceofAphanothece
sacrum biomaterialwhosetopregionwas
irradiatedbyagreenlaserpointer(=532
) G li ht tt d id l nm).Greenlightwasscatteredwidelyover
thebiomaterials.
117
118

-CHAPTER 5-
Metal ion sorption of extracted polysaccharides and its hydrogels

5-1: Introduction
In previous chapters, it was found that sacran is a polyanion which contains
carboxyl acid and sulfate and a megamolecule constituted of 11 sorts of
monosaccharides. Because sacran is polyanions, we speculate that sacran may
efficiently adsorb metal ions. On the other hand, sacran forms specific structures such as
liquid crystals. Moreover, since sacran shows high viscosity sensitive to the salt, it is
considered that sorbability of cation closely related with sacran solution structures. Then
in this chapter, the adsorption function of sacran with metal ions is clarified by studying
of relationship between metal adsorption property and the sacran structure.

5-2: Gel bead formation of extracted polysaccharides with various metal ions
5-2-1: Methods
Materials: Heavy metal salts used for adsorption tests were listed below. Sc
tifluoromethanesulfonic acid, Cr(III) chloride hexahydrate, Fe(III) chloride
hexahydrate, Mo(IV) acid disodium dehydrate, Pr chloride, Sm chloride hexahydrate,
Dy chloride hexahydrate, Er chloride hexahydrate, Tm chloride, Lu acetate tetrahydrate
were purchased from Wako Pure Chemical Industries. Mn chloride tetrahydrate, Fe(II)
chloride tetrahydrate, Co(II) chloride hexahydrate, Ni(II) chloride hexahydrate, Cu(II)
chloride dyhydrate, Zn chloride, Sr chloride hexahydrate, In chloride, Ba chloride
dyhydrate, Eu chloride, Gd sulfate hexahydrate, Tb nitrate hexahydrate, Ho chloride
119

hexahydrate were purchased from Kanto Chemical Co. Inc. Al(III) chloride and Ca
chloride were purchased from NACALAI Tesque, Inc. Mg chloride hexahydrate were
purchased from Sigma-Aldrich Co. Y chloride, La chloride, Ce chloride, Nd chloride,
Yb chloride were purchased from Nippon Yttrium Co. Ltd. Ethylenediamine
N,N,N,N-tetraacetic acid tetrasodium salts (EDTA) were purchased from Kanto
Chemical Co. Inc. and used as received.
Evaluation of metal ion adsorption: Aqueous solutions of individual metal ions at
concentrations of 10
-1
M, 10
-2
M, 10
-3
M, 10
-4
M, and 10
-5
M were prepared just before the
adsorption tests. Viscous solutions of the polysaccharides such as sacran and sodium
alginate (0.5wt%) were dropped stepwise into the individual metal ion solutions and
shaken softly, and after 10 minutes of standing still, we checked whether gel beads had
formed in the solution. In more diluted sacran solutions, the adsorption tests were
performed using the same procedure as above, but we checked the gelation behavior
after 10 minutes and 2 days. If gel beads were formed, then an EDTA aqueous solution
at a concentration of 0.5M was added into the solution containing the gel beads. Next,
we checked whether the gel beads were broken by the metal ion chelation effects of
EDTA. The amount of metal ions adsorbed to the polysaccharides chains was estimated
from the concentration of residual metal ions in the supernatant of the test samples
containing the gel beads, as measured by ICP (inductively coupled plasma) analysis
using a Shimadzu ICPS-8100.
5-2-2: Results and discussion
Gelation with metal ions: Viscous solutions of anionic polysaccharides such as sacran
and sodium alginate at a concentration of 0.5 wt% were dropped into the solution of
divalent metal ions. The gel bead formation behaviors were summarized in Table 5-1.
120

Sacran formed gel beads in solutions (10
-1
M) of heavier alkaline earth metal ions such
as Sr
2+
and Ba
2+
, but did not form gel beads in other divalent ions such as Mg
2+
, Mn
2+
,
Co
2+
, Ni
2+
, Cu
2+
, Zn
2+
, Fe
2+
, and Ca
2+
.

According to the literature, alginate gel beads
were formed as a result of the physical cross-linkage of the polysaccharide chains by
electrostatic binding and coordination of the divalent metal ions to the chain backbones
[1]. Based on these alginate studies,

we inferred that gel beads of sacran could be
formed by sacran chain cross-linkage following metal ion binding. This inference was
supported by the finding that the sacran gel beads were broken by the addition of EDTA,
which efficiently captured the metal ions. For Fe
2+
, thin precipitates instead of gel beads
were formed, and the precipitates were much more difficult to collect than the gel beads,
which were easily picked up by a pair of tweezers. The precipitate formation may be
due to poor binding of the metal ions.
On the other hand, alginate gel beads were formed in solutions (10
-1
M) of Co
2+
, Ni
2+
,
Cu
2+
, Zn
2+
, Sr
2+
, Ba
2+
and Ca
2
. Consequently, sodium alginate had a higher efficiency
for adsorption with divalent metal ions than sacran. No beads were formed for either
polysaccharide in divalent Mg
2+
, Mn
2+
, and Fe
2+
ionic solutions (Table 5-1). In thinner
solutions of Sr
2+
and Ba
2+
, the sacran formed gel beads at a concentration of 10
-2
M, but
no gel beads were found at 10
-3
M. Therefore, the critical concentrations of sacran/Sr
2+

and sacran/Ba
2+
gel formation were between 10
-2
to 10
-3
M. Sodium alginate showed
critical concentrations of gelation with Sr
2+
and Ba
2+
in the same range as sacran. If only
the ionic interactions of the polysaccharide anions with the metal cations were
correlated with gel bead formation, then metal ions with smaller ionic radii among
identically valent ions should show more effective gelation. In second congeners
containing alkaline earth metals, the ionic radii (VI coordination number) of Be
2+
, Mg
2+
,
121

Ca
2+
, Sr
2+
, Ba
2+
, and Ra
2+
are 35, 72, 100, 113, 136, 143 pm, which increased in order of
increasing atomic number. The larger two ions, Sr
2+
and Ba
2+
, should show weaker
interactions with the anions. However, these two ions showed higher gelation activity.
In metal ions with atomic numbers of 25-30, Mn
2+
, Fe
2+
, Co
2+
, Ni
2+
, Cu
2+
, and Zn
2+

showed a simple decrease in ionic radii but only the two smallest metal ions, Cu
2+
and
Zn
2+
, formed gel beads with alginates. In contrast, Ca
2+
is the largest divalent ions in the
fourth row, and efficiently formed gel beads. These results showed that the effects of the
ionic radii are not very important for gel bead formation; instead, the suitability of the
polysaccharide structures with the metal ions may be important. On the other hand,
sacran formed gel beads only in solutions of Sr
2+
and Ba
2+
in these tests of divalent
metal ions, which suggests that sacran should have good suitability with large metal
ions located at the fifth and sixth rows, correlating with the electrons in the N electron
orbital.
With respect to trivalent metal ions, sacran formed gel beads in solutions of Al
3+
, Sc
3+
,
Cr
3+
, Fe
3+
, Y
3+
, In
3+
and lanthanoid ions over a concentration range of 10
-1
to 10
-3
M,
but no gel beads formed at 10
-4
to 10
-5
M. Therefore, the critical concentrations of
formation for sacran gel beads with these trivalent ions were between 10
-3
and 10
-4
M,
which was lower than for sacran/divalent-metal-ion gels. In particular, Fe
3+
formed gel
beads more efficiently than Fe
2+
. These results indicate that trivalent ions have a
higher efficiency for cross-linking sacran chains than divalent ions, presumably due to
stronger electrostatic interactions. Similarly, this difference between the two Fe ions
for sodium alginate indicates an advantage of the trivalent metal ions, but the result that
no gel beads were formed at a concentration of 10
-3
M suggests that the effect of the
charge increase from +2 to +3 is not very strong. In addition, we were aware that
122

sacran formed fiber aggregates in very dilute solutions (10
-4
M) of rare earth metal ions
such as Sc
3+
, Y
3+
, and lanthanoid ions. The fibers were easily removed from the
solutions by a pair of tweezers like gel beads, which shows an advantage in metal ion
recovery. Sacran had a higher ability to form gel beads adsorbing trivalent metal ions,
as shown in Table 5-2, than sodium alginate. The diffusion coefficient of sacran, whose
radius of gyration is ca. 400 nm as reported in the Chapter 3, could be estimated at
~10
-12
m
2
s
-1
. This is roughly 10
3
times lower than the metal ions (~10
-9
m
2
s
-1
),
indicating that the spreading speed of the droplets corresponding to the diffusion speed
of the polysaccharides was much higher than the rate of metal ion cross-linking. Since
the Mw of sodium alginate is as high as ~10
5
, the diffusion coefficient should be lower
than ~10
-11
m
2
s
-1
. The lower diffusion coefficient of sacran than for sodium alginate
may be more effective for gel bead formation.
Fig. 5-1 is a periodic table summarizing whether gel beads are formed in order to
compare sacran and sodium alginate. Although sodium alginate had an advantage to
form gel beads in divalent ionic solution, sacran has the advantage in trivalent solutions.
Moreover, sacran has an advantage for heavier metal ions from the fifth and sixth rows.
We attempted a gelation test for Pb
2+
, which is divalent but heavier than lanthanoid
(Table 5-2). Although sodium alginate did not form gel beads in a solution of Pb
2+
at a
concentration of 10
-3
M, sacran successfully formed gel beads with Pb
2+
at 10
-3
M.
These results also support the suitability of sacran for heavy metal ions from the sixth
row. This finding is meaningful in terms of toxic metal recovery from industrial waste.
Heavy metal ions with N electron orbitals show various coordination styles, which may
be suitable for sacran chains containing very complex structures composed of more than
123

11 sugar residues, such as uronic acid, muramic acid, and mannose with vicinal
hydroxyl groups.
We investigated the amount of metal ions incorporated into the gel beads of sacran.
Polysaccharide solutions with a concentration of 0.5 % (1 ml) were dropped into InCl
3

solution with a concentration range from 10
-2
to 10
-3
M (10 ml). These results are
summarized in Table 5-3. The amount of metal ions absorbed into the gel beads
showed much smaller values than expected based on the carboxylate content of sacran,
and little dependence on the metal ion concentrations. On the other hand, gel beads
composed of sodium alginate absorbed more metal ions than sacran, and the amount of
absorbed ions decreased with a decrease in the metal ion concentration. This result
was unexpected, and then we picked up, broke, and observed the sacran/In
3+
gel beads.
We found that the sacran gel beads were composed of a gelatinous shell surrounding a
core of colloidal fluid like capsules, whereas the metal ions penetrated deep into the gel
beads composed of sodium alginates (Fig.5-2). This observation prompted us to
speculate about the formation process of the gel beads. We hypothesized that just after
the sacran droplet was surrounded by the metal ion solution, a stiff shell of sacran
chains cross-linked by metal ions was formed, and any subsequent metal ion
incorporation inside the gel beads was restricted by the dense shells. Sacran formed
capsular gel materials even in very dilute metal ion solutions. This highly efficient gel
capsule formation may be related to the sacran structure and metal ion size. Next, I
tried to measure the maximum adsorption ratio of neodymium ion to sacran chains from
the UV-vis absorption of supernatant. However, the sorption ratio increased with
increasing the concentration of neodymium ion solutions in any concentration range,
and then I failed to determine the maximal value by the method. This failure may be
124

due to the ion absorption into gel beads and the next try was done based on preventing
gel beads by reducing the concentration of sacran solution. However another difficulty
in evaluation is to distinguish adsorption of Nd ions to sacran chains precisely with
absorption. Then I give up using the UV-vis method and instead I tried to use the
conductivity measurement which may differentiate adsorption with absorption; if Nd
ions adsorb onto the sacran chains ionic interaction, they do not show conductivity. First
electrical conductance of NdCl
3
aq was measured and next a given amount of sacran
was added, and last the reduction of electrical conductance was measured. To confirm
the reliability of this method, I tried to examine an amount of adsorption of calcium ions
to alginates. I obtained a reasonable result that one Ca ion adsorbs onto two carboxylic
acids of alginates. This result agrees with well-known adsorption ratio and this method
is reliable. In the same way, adsorption ratio of Nd ions to alginate carboxylic acids was
1 : 3. Then alginates adsorbed trivalent and divalent ions without large difference.
Concerning sacran, the method showed that 1 : 2 adsorption of Nd ion to sacran
carboxylic acids but 1 : 3 adsorption to carboxylate and sulfate ions (Fig. 5-2). Thus, it
was demonstrated that these experimental systems were based on simple electrostatic
adsorption of all the ionic groups. Whereas, from the test of adsorption ratio of Ca ions
to sacran, it was found that 3 negative charges adsorbed to 0.1-0.2 Ca ions, which is so
small because adsorption did not attain to the equilibrated state in the concentration
range. Therefore, sacran can adsorb Nd ions selectively in the coexistence of divalent
ions if the condition is adjusted well. Finally I can claim the advantage of sacran in
Nd adsorption; sacran adsorbed Nd ions in lower concentration range i.e. higher
adsorption efficiency than alginates.

125

Structures
Sacran showed a lyotropic liquid crystalline phase in solutions with concentrations
over 0.25%, as reported in the Chapter 4 [2]. We examined the formation of the gel
beads by changing the concentration of the sacran solutions in order to investigate the
effects of the liquid crystalline state, where the sacran chains were self-organized and
self-oriented onto the gel beads. Fig. 5-4 shows the gel bead formation behavior when
sacran solutions of various concentrations of 1.00 %, 0.75 %, 0.5 %, 0.33 %, 0.1%,
0.075% and 0.05% were dropped into an In
3+
solution (10
-2
M). Ten minutes after the
sacran solution was dropped, we found that sacran solutions with concentrations of
1.00%, 0.75 %, 0.5 %, and 0.33% formed gel beads by binding the In
3+
ion, but nothing
was observed in the 0.075% and 0.05% solutions (Fig. 5-4, top lines). In the 0.1 %
solution, gel beads could be faintly seen, but they disappeared after one day. These
findings suggest a close relationship between the liquid crystalline state of the sacran
solution and gel bead formation. Moreover, birefringence of the sacran gel beads
formed in solutions over 0.25 % was confirmed by clear observation of the gel beads
under cross-nicol polarimetry, as shown in the second line of Fig. 5-4, indicating that
the orientation of the sacran chains was retained in the gel beads. The third line of Fig.
5-4 shows a crossed-polarized image of the sacran gel beads at two days after gel
formation. The size of the gel beads formed in solutions over 0.25 % showed little
change over two days, which indicated that the metal ion uptake into the gel beads was
almost finished within 10 minutes. On the other hand, in solutions below 0.25 %,
small gel beads appeared two days after mixing the sacran solutions with the In
3+

solutions. One can speculate that metal ion uptake into the sacran chains in the very
dilute solutions was very slow, but still faster than sacran diffusion, and thus the uptake
126

accumulated inside the sacran chain coils to increase their local concentration gradually.
When the sacran chain concentration exceeded 0.25 %, they could make a phase
transition to LC to accelerate the uptake, and therefore formed gel beads. In these cases,
the gel beads were smaller than the initial droplets, and were not capsular. In the
0.01% and 0.005% solutions, no gel beads were formed even 2 days after mixing the
sacran and In
3+
ion solutions, which might be correlated with the critical concentration
for sacran chain entanglement C* = 0.01 wt% calculated using an equation [3]:
c* a
-3
N
-4/5
, where a is the length of monosaccharide (0.65 nm), and N is the
number of monosaccharides in sacran (8.9 x 10
4
).
We investigated the structure of the gel beads by X-ray diffraction. Fig. 4-7 shows
WAXD diagrams of the sacran fibers in their dry state, sacran aqueous solution, and gel
beads of sacran with In
3+
in a water-swollen state. These diagrams show a very broad
diffraction of the amorphous halo seen in the non-crystalline structure, around
2 =19-23
o
(: diffraction angle) and 2=0.30-0.33
o
. These WAXD studies indicated
that the dry sacran fibers were non-crystalline as shown in the Chapter 4, presumably
due to heterostructures composed of more than 11 sugar residues, which allows sacran
to efficiently dissolve in water despite its extremely high Mw. In addition, the sacran
chains in the gel beads retained their structure in solution despite the metal ions binding.
From these results, we proposed the gel bead formation process as illustrated in Fig. 5-5.
Sacran solutions were initially in a nematic LC state, where the sacran chains
automatically oriented and packed more densely than in their amorphous state (left
illustration of Fig. 5-5). When the sacran solution was dropped into a trivalent metal ion
solution, the metal ions diffused into the droplet and bridged the sacran chains very
efficiently due to the sacran LC stacking (middle illustration). The metal ions bound
127

efficiently to the sacran chains on the droplet surface to maintain the LC structure,
which made it difficult for additional In
3+
to permeate into the gel beads (right
illustration). In the sacran gel beads, the efficient gelation of the droplet surface may
be attributed to strong electrostatic forces, the liquid crystalline state, an ultra-high
molecular weight, and the hetero-polysaccharide structure.
5-3: Metal ion sorption of chemically cross-linked sacran gels
5-3-1: Methods
Chemical cross-linking of extracted polysaccharides: Sacrans were cross-linked to yield
hydrogels and organogels by the following procedure. In the case of hydro-gel
formation, L-lysine was added into a sacran solution (1 wt.%, 100 ml) and then
1-ethyl-3-(3- dimethylaminopropyl) carbodiimide HCl salt (2 g) was added as an
condensation reagent. The solution was stirred strong, centrifuged to degas, and kept at
4
o
C for 72 hrs in the refrigerator, to yield the hydrogels. Organogels were prepared in
dimethylsulfoxide (DMSO) by the same procedure. The hydrogels and organogels were
purified by immersing in water and DMSO for 3 days at room temperature to remove
unreacted compounds, respectively. The solvent exchange from DMSO to water was
performed by immersion of organogels in distilled water for 1 week with replacing the
water.
Evaluation of metal ion adsorption: Aqueous solutions of individual metal ions with
concentrations of 10
-2
M, 5.0 10
-3
M, 2.5 10
-3
M, and 3.0 x 10
-3
M were prepared just
before adsorption tests. Viscous solutions of polysaccharides such as sacran, sodium
alginates (0.5 wt %) and carrageenan (0.5 wt %) were dropped into individual metal ion
solutions and shaken softly. Sacran ions to polysaccharides chains were estimated from
concentrations of residual metal ions in supernatants of adsorption test samples
128

containing gel beads. Gels were immersed in the same individual metal ion solutions.
The adsorption amount of metal measured by ICP (inductively coupled plasma)
analyses using a Shimadzu ICPS-8100.
5-3-2: Results and discussion
Chemical cross-linking: In general, it is reported that the macromolecules with high
molecular weight form gels efficiently [4]. Therefore we tried to prepare the gels of
Sacran by chemical cross-linking. The chemical cross-linking of polysaccharide is most
commonly made using the less efficient reaction of hydroxyl groups with cross-linkers
[4]. On the other hand, Sacran contained carboxyl groups and then we decided to use a
high efficient reaction such as dehydration reaction of carboxylic acids with diamines. It
has been reported that hyaluronic acid successfully reacted with ADH (structure;
Fig.5-6a) by using the abovementioned dehydration [4]. Then, we examined the gelation
of sacran using ADH by the procedure similar with the formation of hyaluronic acid gel
and obtained the aimed gels (Fig.5-7a). However ADH is not a natural product and then
it has some possibilities for militating against environments and bad for ecology. We
also investigated whether amino acid containing two amino groups L- lysine (structure;
Fig.5-6b) could be used as cross-linker, by following procedures. Sacran solution in
water and DMSO were prepared, and L-lysine was added, and then condensation agent
1-ethyl-3-(3-dimethylaminopropyl) carbodiimide HCl salt was added. Since the reaction
did not occur immediately, we had enough time to deaerate the solution by
centrifugation to gather all air bubbles on the top of solution. Thereby, homogeneous
solution without bubbles was obtained. The gelation after three days was confirmed by
turning a sample bottle containing reactants upside down. Then, the gels were took out
from a sample bottle and immersed in water for several days to remove impurities. The
129

swelling degree of hydrogels (amount of cross-linking agent added was 10 mol% to mol
of carboxylic acid in Sacran) was about 400 times. Although the value is very high
comparing with other hydrogels reported so far, the hydrogels were tough to process
into the rectangular shape by a normal cutter (Fig.5-7b). When the quantity of L- lysine
was decreased to 5 mol%, the gel was formed but it was more fragile than that prepared
at 10 mol%. Further, the gelation was not confirmed when 3 mol%. The swelling degree
of hydrogel when use of ADH as a cross-linker was about 240 times. Therefore it
seemed that reaction efficiency rate of L-lysine was somewhat lower than that of ADH,
presumably due to a lower nucleophilicity of -amine of L-lysine than that of
hydrorazide amine of ADH. However we discuss that L-lysin which is produced by
fermentation method has an advantage in the green-chemical aspect and can propose it a
cross-linking agent for sacran containing carboxylic acids. Moreover organogel of
Sacran was formed using L-lysine as a cross-linker in DMSO to create transparent gels.
Next, the organogel was immersed into pure water to replace DMSO into the water, and
then the gel shrunk and turned harder and opaque. One can considered that some sort of
self-assemblies were induced by solvent replacement, but the study in detail were
remained.
Metal ion sorption: Sacran is a sulfated polysaccharide containing sulfate groups in 10
mol % to total monosaccharides, -carrageenan which is also a sulfated polysaccharide
(150 mol %) was used as a control. Each polysaccharide solution was prepared at a
concentration of 0.5 wt % just like sacran solutions. Indium (In) ion and Gadolinium
(Gd) ions were used for the quantitative estimation of metal ion adsorption. First of all,
In ion and Gd ion solutions were adjusted to concentrations of 10
-2
M, 5.0 10
-3
M, 2.5
10
-3
M and 10
-3
M, respectively, and sacran and alginate solutions with 0.5 wt %
130

concentration were poured into Gd ion and In ion solutions at various concentrations
and agitated vigorously, and then left for a certain amount of time. Alginate finely
dispersed and sank to a nadir in all solutions, while sacran instantly formed stiff skin
layers with liquid crystalline structures, which were confirmed by X-ray diffraction
studies and observations from polarized microscopy. Furthermore, a sacran gel cube cut
out with 1 cm
3
sides was immersed in each metal ion solution and sorbed metal ions for
two days. The color of sacran gels varied from white to transparent as caused by the
sorption of metal ions (upper photo: Fig. 5-8), which might indicate that the metal ion
adsorbed to sacran chain networks as shown in lower figure of Fig.5-8. After two days,
alginate, carageenan, sacran and sacran gels were put off from each metal ion solution,
and the decrease in metal ion concentration in supernatants by polysaccharide additions
was measured by an ICP emission method. Accordingly, an amount of actual sorbed In
ions and Gd ions were largest in quantity in the case of pouring 0.5 wt % alginate into
ion solutions. Since 0.5 wt % alginate gave rise to slime under powerful agitation after
being poured into metal ion solutions, which increased the contact area with metal ions,
it was predictable that alginate actually sorbed metal ions the most. In contrast, when
0.5 wt % sacran solution was poured and agitated, sacran chains were instantly
cross-linked with metal ions and formed solid skin layers, so the skin layer might block
intruding metal ions inside capsular structures, showing a low sorbability. On the other
hand, sacran gel cubes with 1 cm
3
sides actually sorbed small amounts of metal ions.
Then we speculated such a small amount of sorbed metal ions to sacran gel may be
attributed to too low concentrations of sacran in the gel because of a high swelling
degree. Consequently, in the case of an amount of the metal ion sorption was
recalculated into the value per one mole sugar residue, the sorption value of sacran gel
131

was highest. Further we evaluated the sorbed amount of metal ions under electrostatic
adsorption of metal ions to polysaccharides. In other words, as both In and Gd ions are
trivalent, so the stoichiometric ratio to negative charges of polysaccharides is 0.33.
Actual evaluations were made in the following way: A / B, where A is the number of
moles of metal ion sorbed to polysaccharides, and B is the number of moles of negative
charge which is calculated from the number of moles of sugar residues and a negative
charge composition (alginate; 1, sacran; 0.33, and carageenan; 3).
We estimated the amount of sorbed metal ions to negative charges of 0.5 wt %
alginate, carageenan, sacran solutions and sacran gels from the above-described
calculation. Results show both In and Gd ions sorbed sacran gels exceeding electrostatic
stoichiometric values. Each polysaccharide solution with a 0.5 wt % concentration was
poured into metal ions solutions at various concentrations, and all values were below
0.33, especially alginate, which showed exceptionally low values of amounts of metal
ion sorption (Fig. 5-9). In these evaluations, it was demonstrated that sacran gels could
sorb metal ions more than ones own negative charge by cross-linking and gelation
effects. In previous investigations, we addressed that sacran formed instantaneously a
solid skin layer with liquid crystallinity when 0.5 wt % sacran was dropped into various
metal ion solutions. This phenomenon might indicate that the velocity of metal ion
adsorption to sacran rapidly forms sacran-metal ion complexes (skin layer) on surfaces,
which prevent the permeation of further metal ions into the inside through solid skin
layers, and inner sacran could not sorb metal ions. Next, we consider the mechanism of
excess metal ion sorption to only sacran gels. The sacran gel clouded in association with
metal ion adsorption (Fig. 5-8b), however, it could be observed that a core of the gel
also gradually clouded. Thus these phenomena reflected that metal ions intruded inside
132

of gels unlike the case of dropping 0.5 wt % sacran, and indicated most sacran chains
contributed to metal ion sorption, especially if one considers the gelation of sacran
could increase the ability of metal ion sorption. It is conceivable that chemical
cross-linking of sacran affects the disordering of sacran structures to prevent skin layer
formation. Through this effect, further metal ions smoothly intrude inside of gels. As a
result, the metal ion sorption ratio to negative charge of sacran by far exceeded the
stoichiometric ratio of 1:3. The phenomenon can be interpreted for follows. Sacran was
a supergiant macromolecule with a large number of negative charges from carboxyl and
sulfate groups, which strongly attracted metal ions by coulombic interaction. Since rare
earth metal ions formed water-insoluble salts with CO
3
2-
but they formed water-soluble
salts with SO
4
2-
, they might adsorb more efficiently to carboxylate groups of sacran
rather than sulfate ones. Even after the rare earth ions are trapped by carboxylate
groups, sulfate anions still attracts excess amount of metal ions by coulombic
interaction to absorb into sacran gels. Such synergic effects of carboxylate adsorption
and sulfate absorption may induce to the excess sorption phenomenon of sacran
hydrogels.

5-4: Metal ion sorption of semi-IPN PVA gels containing polysaccharides.
5-4-1: Methods
Preparation of semi-IPN PVA gels: Semi-IPN gels of PVA gels with polysaccharide
chains were shown as follows. 10 ml of 1 wt% poly(vinyl alcohol) (PVA) solution 10ml
of 0.8 wt% polysaccharides solution were mixed, and to make homogeneous by
agitation. After agitation, glutaraldehyde (2 mol% to PVA unit) was added to cross-link
PVA chains, and then degassed the solution by centrifugation to gather air bubbles on
133

the surface of solution and the solution was still-stood until gels were formed. For
comparison with semi IPN gels, the PVA gels cross-linked by glutaraldehyde were also
prepared. Then, gels were swelled in distilled water for 1 week with replacing the
water for purification.
Metal ion sorption: Swelling ratio of these gels were calculated as a weight ratio of
swollen gels in the equilibrated state to dried gels. On the other hands, aqueous
solutions of neodymium (Nd) metal ions with concentrations from 1.0 10
-7
to

1.0 10
0

M were prepared just before adsorption tests. Gels (semi IPN gels and PVA gels) were
cut into about 1 x 1 x 1 cm
3
and immersed into the individual Nd ion solutions. The
adsorption amount of Nd ions was measured by ICP (inductively coupled plasma)
analyses using a Shimadzu ICPS-8100.
5-4-2: Results and discussion
I speculated, in the case of chemically cross-linked gels of sacran, the amount
carboxylic acid related to metal ion sorption is decreased due to cross-linking by
reaction of amines of cross-linkers. In order to solve the problem, I fixed sacran
chains in PVA networks to prepare semi-IPN gels and I confirmed sacran chains did not
give the gels by glutaraldehyde. It is expected that sacran can scatter in the gel
networks of PVA without cross-linking. The swelling degree of hydrogels is
summarized as follows; PVA gels containing sacran was about 21 times, PVA gels
containing alginates was 19, and PVA gels was 6 times, respectively. The swelling
degree of PVA gels containing sacran was higher than that of PVA gels containing
alginates, which may be attributed to the difference in the water retention capacity.

Metal ion sorption: From the results of qualitative evaluation of metal ion sorption,
134

sacran efficiently adsorbed rare earth metal ions such as Nd ion. Nd ion solutions were
adjusted to concentrations of 10
1
-10
-7
M. Cubic gels (PVA gels, PVA gels containing
sacran, PVA gels containing alginates) with 1 cm
3
sides were immersed in metal ion
solutions and still-stood for a week. These gels were put off from each metal ion
solution, and measured their weights to determine the degree of shrinkage. The change
of swelling degree of gels did not occur in Nd ion solution with concentration range
below 10
-5
M. On the other hand, the swelling degree of gels containing polysaccharide
decreased from 10
-5
M to 10
-3
M, which suggest that Nd ions adsorb to polysaccharides
and cross-link the polysaccharide chains. In 10
-3
to10
-1
M, the swelling degree of gels
remained constant, indicating the constant physical cross-linking degree in the
concentration ranges. However, the swelling degree of gels increased again over the
metal ion concentration of 10
-1
M (Fig.5-10a).
In order to investigate the reason for the swelling degree change, the sorption ratio
was calculated from a decrease in metal ion concentration in supernatants measured by
an ICP emission method. Initially, I confirmed a sorption ratio of Nd ions to PVA gels
was negligibly low below Nd concentration of 10
-1
M, however, the sorption ratio
drastically increased over 10
-1
M. The re-swelling phenomenon of the gels over 10
-1
M
may be due to Nd ion adsorption to PVA networks without ionic neutralization to
increase osmotic pressure of gel inside. Furthermore, sorption ratio to carboxylic acid
of polysaccharides was much higher than those in polysaccharide solution system (not
gel system), and increased up to about 400 times. This reason can be explained by a
normal diffusion of Nd ions into the gels because Nd desorption from the Nd-sorbed
gels were easily made by just immersion into the pure water. Overall Nd ions sorption
ratio of PVA gels containing sacran was higher than that of PVA gels containing alginate,
135

similarly with the non-gel case. The high efficiency of Nd adsorption to sacran chains
was kept in PVA networks, which may lead to the use of Nd-recovery from Nd-adorbing
gels. In the future, I will make a metal recovery by direct reduction of Nd ions from the
inside of gels using electrodes.

5-5: Conclusion
Sacran contains carboxylate and sulfate groups. Then I investigated the gelation
properties of sacran binding with various heavy metal ions. The sacran chain adsorbed
heavier metal ions more efficiently, such as indium, rare earth metals, and lead ions to
form gel beads. In addition, trivalent metal ions adsorbed to the sacran chains more
efficiently than divalent ions. Gel bead formation may be closely correlated with the
liquid crystalline organization of sacran. The conductivity measurement revealed that
Nd-adsorption to sacran chains is 2:1 system. Next sacran was successfully cross-linked
with cross-linking agents such as l-lysine and adipoyl dihydrazide to form tough gels
with a high swelling degree, ca. 400-800 times to the weight of the dried gel. We used
highly swollen sacran gels (swelling degree: 700-800 times of dry weight) for metal
sorption. Sacran gels shrank and clouded in aqueous solutions of In and Gd ions and
showed more than 100 times adsorption ratios of these metal ions compared with non
cross-linked sacran presumably due to continuous uronic acids. Sacran have an
advantage in heavy metal adsorption to alginate system. Thus it would have advantages
in the adsorption of heavier metal ions. Although it is difficult to isolate heavy trivalent
metal ions, one can say that it was important to recover the group of various heavy
metal ions from metal ion mixtures containing monovalent or divalent metal ions. Many
cyanobacteria live in tropical areas and polysaccharides derived from cyanobacteria
136

generally have multiple structures such as many kinds of constituent monosaccharides.
This fact will lead to the technology of liquid waste disposal of heavy metal ions in
many tropical countries, which are beset with waste problems.

Reference:
[1] T. Gotoh, K. Matsushima, K. Kikuchi, Chemosphere 55, 135 (2004). b) G. T. Grant,
E. R. Morris, D. A. Rees, P. J. C. Smith, D. Thom, FEBS Lett. 32, 195 (1973).
[2] M. Okajima, D. Hayasaka, S. Miyazato, T. Kaneko, Macroscopic birefringence in
liquid crystals from novel cyanobacterial polysaccharide with an extremely high
molecular weight Proceeding of SPIE-International Congress on Optics and
Optoelectronics, 6587, 658711/1-658711/8 8 (2007).
[3] P.-G. de Gennes, Scaling Concepts in Polymer Physics, Section III-2-1 Cornell
University Press, Ithaca and London. (1979)
[4] Y. Osada. K. Kajiwara. Gels Handbook, Academic Press, Tokyo. (2001)
Metal ions
10
-1
M 10
-2
M 10
-3
M 10
-4
M 10
-5
M
Table 5-1. Gel beads formation behavior of sacran solutions
dropping into divalent metal ion solutions
a
Metal ion concentration
10 M 10 M 10 M 10 M 10 M
Mg
N (N) N (N) - - -
Mn
N (P) N (N) - - -
2+
2+
Fe
2+
P (P) N (N) N (N) N (N) N (N)
Ca
2+
N (G) N (G) N (N) N (N) N (N)
Co
N (G) N (P) - - -
Ni
N (G) N (P) - - -
Cu
N (G) N (G) - - -
Zn
N (G) N (G) - - -
2+
2+
2+
2+
Fe
( ) N (N) N (N) N (N) N (N)
( ) ( )
Sr
G (G) G (G) N (N) N (N) N (N)
Ba
G (G) G (G) N (N) N (N) N (N)
a) Marks of G, P, N, and - refer gel beads formation, precipitation, nothing,
2+
2+
Pb
G (G) G (G) G (P) N (N)
2+
-
and no test, respectively. Gelation behavior of sodiumalginate was shown in
parentheses
137
T bl 5 2 G l b d f ti b h i f l ti
Metal ion
10
1
M 10
2
M 10
3
M 10
4
M 10
5
M
Al
3
G (G) G (P) G (N) N (-) N (-)
Table 5-2. Gel beads formation behavior of sacran solutions
dropping into trivalent metal ion solutions
a
3
rd
row
Metal ion concentration
Al
G (G) G (P) G (N) N ( ) N ( )
Cr
3
G (G) G (G) G (P) N (-) N (-)
Fe
3
G (G) G (G) G (P) N (-) N (-)
Sc
3
G (G) G (G) G (P) F (-) N (-)
Y
3
G (G) G (G) G (P) F (-) N (-)
In
3
G (G) G (G) G (P) N (-) N (-)
3 row
4
th
row
5
th
row
La
3
G (G) G (G) G (P) F (-) N (-)
Ce
3
G (G) G (G) G (P) F (-) N (-)
Sm
3
G (G) G (G) G (P) F (-) N (-)
Nd
3
G (G) G (G) G (P) F (-) N (-)
Pr
3
G (G) G (G) G (P) F (N) - (-)
Eu
3
G (G) G (G)
G (P)
F (N) N ( ) Eu G (G) G (G)
G (P)
F (N) N (-)
Gd
3
G (G) G (G) G (P) F (N) N (-)
Tb
3
G (G) G (G) G (P) F (N) N (-)
Dy
3
G (G) G (G) G (P) F (N) - (-)
Ho
3
G (G) G (G) G (P) F (N) - (-)
Er
3
G (G) G (G) G (P) F (N) - (-)
T
3
G (G) G (G) G (P) F (N) ( )
Ln
Tm
3
G (G) G (G) G (P) F (N) - (-)
Lu
3
G (G) G (G) G (P) F (N) - (-)
a) Marks of G, P , N, F, and - refer to gel bead formation, precipitation,
fibrization, nothing, and no test, respectively. Gelation behavior of sodium
alginate was shown in parentheses
Yb
3
G (G) G (G) G (P) F (-) N (-)
138
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
H He
Li Be B C N O F Ne
M
2+
3+
Si P S Cl A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
H He
Li Be B C N O F Ne
M
2+
3+
Si P S Cl A
Na
Mg
2+
Al
3+
Si P S Cl Ar
K
Ca
2+
Sc
3+
Ti V
Cr
3+
Mn
2+
Fe
3+
Co
2+
Ni
2+
Cu
2+
Zn
2+
Ga Ge As Se Br Kr
Rb
Sr
2+
Y
3+
Zr Nb
Mo
6+
Tc Ru Rh Pd Ag Cd
In
3+
Sn
2+
Sb Te I Xe
Cs
Ba
2+
Ln
Hf Ta W Re Os Ir Pt Au Hg Tl
Pb
2+
Bi Po At Rn
Na
Mg
2+
Al
3+
Si P S Cl Ar
K
Ca
2+
Sc
3+
Ti V
Cr
3+
Mn
2+
Fe
3+
Co
2+
Ni
2+
Cu
2+
Zn
2+
Ga Ge As Se Br Kr
Rb
Sr
2+
Y
3+
Zr Nb
Mo
6+
Tc Ru Rh Pd Ag Cd
In
3+
Sn
2+
Sb Te I Xe
Cs
Ba
2+
Hf Ta W Re Os Ir Pt Au Hg Tl
Pb
2+
Bi Po At Rn
N N
N G
G G
G G
G N
G N
N G N G N G N G G N
G N
G N
G N N N N N
G N
N N
Fr Ra Ac Rf Db Sg Bh Hs Mt
La
3+
Ce
3+
Pr
3+
Nd
3+
Pm
Sm
3+
Eu
3+
Gd
3+
Tb
3+
Dy
3+
Ho
3+
Er
3+
Tm
3+
Yb
3+
Lu
3+
Ac Th Pa U Np Pu Am Cm Bk Cf Es Fm Md No Lr
Fr Ra Rf Db Sg Bh Hs Mt
La
3+
Ce
3+
Pr
3+
Nd
3+
Pm
Sm
3+
Eu
3+
Gd
3+
Tb
3+
Dy
3+
Ho
3+
Er
3+
Tm
3+
Yb
3+
Lu
3+
Ac Th Pa U Np Pu Am Cm Bk Cf Es Fm Md No Lr
G N G N G N G N G N G N G N G NG NG N G N G N G N G N
Ln
Ac
Figure 51. Periodic chart showing the gelation
behavior of sacran and sodium alginates when
polysaccharide solutions (0.5 wt%) were dropped p y ( ) pp
stepwise into individual metal ion solutions. G and
N refer to gelated and not gelated, respectively,
and the results for sacran and sodium alginates are
shown to the left and right of each individual element shown to the left and right of each individual element
mark, respectively.
139
Table5-3 Amount of indiumionadsorbedtopolysaccharidechains
a
Polysaccharides
In
3+
concentration
1.0 x10
-2
M 5.0 x10
-3
M 2.5 x10
-3
M 1.0 x10
-3
M
Sacran 29 39 51 45
Table 5 3. Amount of indium ion adsorbed to polysaccharide chains
Sodium alginate 210 200 195 55
-carrageenan 3.4 19 13 15
a) Amount of adsorbed In
3+
was evaluated by measuring In
3+
concentration of supernatant for
polysaccharides/InCl
3
mixture solution containing gels or precipitates after centrifugation.
140
In/sacrancomplex In/alginatecomplex
Figure52.Photoofgelsofsacranandsodium
alginateswhichwasformedbypolysaccharide
solutions(0.5wt%)droppingintoindiumion
solutions.
141
6 10
-4
7 10
-4
8 10
-4
Nd
3+
sacran with Nd
3+
(3mM)
y = 0.00034493x R= 0.99957
m
)

a
2 10
-4
3 10
-4
4 10
-4
5 10
-4

p

(
S
/
c
m
in the presence
of sacran

in the absence of
sacran
0 10
0
1 10
-4
0 0.5 1 1.5 2
c
s
(mM)
of sacran
210
-4
1 10
-4
1.5 10
-4
2 10
-4
[COO-]/[Nd3+]=3
/
c
m
)
0.957mM
alginate
b
0 10
0
5 10
-5
nd-sac(3mM)
nd-Alg(3mM)
[COO-]/[Nd3+]=1.5


(
S
/
0.278mM
Concentrationofadded
polysaccharides; 3mM, ca. 0.05%
sacran
Figure 53. a) Conductivity, , dependence on NdCl
3
concentration
in the presence or absence of sacran b) Change in conductivity
-5 10
-5
0 0.5 1 1.5 2 2.5
nd-Alg(3mM)
c
s
(mM)
polysaccharides;3mM,ca.0.05%
in the presence or absence of sacran. b) Change in conductivity
difference, , of NdCl
3
solution in the presence and absence of
polysaccharides, alginate and sacran.
142
0.05% 0.075% 0.10% 0.33% 0.50% 0.75% 1.00%
10min.
after
R
M
A
L
C
L
2 days
N
O
2days
after
C
L
Figure54.Photographsofgelbeadsformedby
droppingsacransolutionsofvarious
t ti i t i di t i hl id concentrationsintoanindiumtrichloride aqueous
solution(10
2
M).Normalrefersthedigitalimages
ofthegelbeadstakeninanormalmode,whereas
CLreferstheimagestakenundercrossnicol g
polarizingconditions.
143
sacran
drop
LCstate MetalIonuptake
Orientedgelstate
Metalion
Sacranchain
Figure55. Schematicillustrationofshell
formationcoveringthegelbeadswhensacran
solutionsintheirliquidcrystalline(LC)state
d d i t t l i l ti weredroppedintometalionsolution.
144
O
N
H
NH
2
a
N
H
O
N NH
2
N H
2
N H
2
COOH
b
2
COOH
Figure56. Chemicalstructureofcrosslinking
agentforsacran.a)adipoyl dihydrazide
(ADH).b)Llysin.
145
a b
Figure57.Imageofarepresentativegel
li k d b ADH ( ) d L l i (b) crosslinkedbyADH(a)andLlysin (b)
146
a b a b
1
1cm
1cm
metalions
Figure 58.a)Photo(upper)andnetworkstructure
(lower)ofsacrangel.b)Photo(upper)and
k (l ) h k b l i networkstructure(lower)shrunkbymetalion
adsorption.
147
3.5
2
2.5
3
sacran
alginate
sacrangel
sacran
a
t
i
o
Gd
0
1
1.5
2
alginate
sacrangel
carageenan
Crosslinking
effects
S
o
r
p
t
i
o
n

r
a
In
0
0.5
0 0.002 0.004 0.006 0.008 0.01 0.012
Metal ion concentration (M)
0.333
Figure 59 Concentration dependence on sorption Figure59.Concentrationdependenceonsorption
ratioofmetalionstoanionsofpolysaccharides.
148
20
25
(
g
/
g
)
PVAgelcontainingsacran
a
5
10
15
s
w
e
l
l
i
n
g

d
e
g
r
e
e
PVA gelcontainingalginates
PVA gel
0
1.E07 1.E06 1.E05 1.E04 1.E03 1.E02 1.E01 1.E+00
ConcentrationofNd
3+
(M)
450
e
s
6
b
200
250
300
350
400
N
d
3
+

t
o
c
a
r
b
o
x
y
l
a
t
o
l
/
m
o
l
)
0
1
2
3
4
5
PVAgel
containing
sacran
PVAgel
containing
alginates
0
50
100
150
200
S
o
r
o
t
i
o
n
r
a
t
i
o

o
f

(
m
o
1.E05 1.E04 1.E03 1.E02
g
1.E05 1.E04 1.E03 1.E02 1.E01 1.E+00
ConcentrationofNd
3+
(M)
Figure510.a)Swellingdegreechangeasafunctionof
concentrationofNd
3+
.b)SorptionratioofNd
3+
tocarboxylate
groupofpolysaccharidespenetratingtoPVAnetworks.
149
150

-CHAPTER 6-
Conclusive Remarks

I have focused cyanobacteria which are photoautotrophic prokaryotes having fixed
CO
2
for 3.5 billion years longer than plants on the earth. Especially cayanobacteria with
jellylike matrix containing a large amount of polysaccharides might be notice as an
important biomass in terms of the high performance of material productions. Of the
cyanobacteria with a jelly matrix, I focused Aphanothece sacrum (Sur.) Okada.
Aphanothece sacrum was biologically classified in the 19th century by Suringar, and
now I found its importance for the following reasons; 1) because it is edible to be safe
for the living organisms, 2) because it has a high performance for producing
biochemicals, e.g. vitamin B12, ferredoxins, lipids, and chromophores, 3) because it has
jellylike matrix with a high water content, which may contain polysaccharides with a
high-performance hydration behavior.
Alcian blue staining test for the A. sacrum biomaterials suggested the presence of
sulfate groups and carboxylate groups in extracellular matrix. The polysaccharide was
successfully extracted from A. sacrum using alkaline solution. The extracted
polysaccharide was soluble in hot water and dimethylsulfoxide but insoluble in any
other widely-used organic solvents. Further it was found that the obtained fiber of
polysaccharide had a high purity by UV spectrum and a high orientation behavior by
polarized microscopy.
The structure of Aphanothece sacrum-derived polysaccharide, sacran, was analyzed.
Sacran was composed of more than 11 sugar residues containing hexoses, pentoses,
151

deoxyhexoses, uronic acid, and a novel sugar of sulfated muramic acid. The content of
anionic groups is 33 %. I found the surprising character of sacran structure, i.e. absolute
molecular weight was 1.6 x 10
7
g/mol which was the highest value of all the extracted
biomolecules reported thus far. Atomic force microscopy and transmission electron
microscopy showed that sacran adopted an extended conformation under salt and was a
megamolecule with a length of 13 m. Sacran was difficult to hydrolyze completely and
then I established multi-step acid-hydrolysis methods. It was found that the main
backbone of sacran was composed of Fuc, GalA, GlcA, GalN, and HexA. Sacran did
not contain ManA, GulA, or Rib, meaning sacran is quite different from
seaweed-derived PS. Further analyses of oligomeric fractions implied that sacran has
diverse partial structures around carboxylic acids.
Sacran solution showed a high zero-shear viscosity (83 000 cps, 1 wt%) which
increased by addition of salts. The water retention capacity is also high 6100 ml g
-1
to
the dry weight. Saline retention capacity of sacran is 2700 ml g
-1
and 10 times than that
of hyaluronic acid. Surprisingly sacran showed still a high capacity for urine retention
(2600 ml g
-1
) even if in the presence of divalent ions such as calcium and magnesium
ions. Moreover I found sacran aqueous solution showed liquid crystalline phase.
Sacrans are observed as self-orienting micro rods longer than 3 m in dilute solution at
c = 0.01 wt % by optical microscopes. Sacran chains form double helixes at c > 0.09
wt % and form huge domains (centimeter scale) of liquid crystals at c > 0.5 wt % which
is quite low when compared to conventional liquid crystalline polysaccharides.
Mesogenic chains of sacrans have extremely high aspect ratios of 1600 for highly
persistent lengths of 32 m. The sacran mesogen is longest in all the liquid crystalline
molecules reported thus far.
152

Sacran contains carboxylate and sulfate groups. Then I investigated the gelation
properties of sacran binding with various heavy metal ions. The sacran chain adsorbed
heavier metal ions more efficiently, such as indium, rare earth metals, and lead ions to
form gel beads. In addition, trivalent metal ions adsorbed to the sacran chains more
efficiently than divalent ions. Gel bead formation may be closely correlated with the
liquid crystalline organization of sacran. Next sacran was successfully cross-linked with
cross-linking agents such as l-lysine and adipoyl dihydrazide to form tough gels with a
high swelling degree, ca. 400-800 times to the weight of the dried gel. We used highly
swollen sacran gels (swelling degree: 700-800 times of dry weight) for metal sorption.
Sacran gels shrank and clouded in aqueous solutions of In and Gd ions and showed
more than 100 times adsorption ratios of these metal ions compared with non
cross-linked sacran presumably due to continuous uronic acids. Poly(vinyl alcohol)
(PVA) networks successfully fixed sacran chains inside to create PVA hydrogels
containing sacran chains which also showed neodymium sorption more efficiently than
PVA hydrogels containing alginate chains. Thus it would have advantages in the
adsorption of heavier metal ions. Although it is difficult to isolate heavy trivalent metal
ions, one can say that it is important to recover the group of various heavy metal ions
from metal ion mixtures containing monovalent or divalent metal ions.
Many cyanobacteria live in tropical areas and polysaccharides derived from
cyanobacteria generally have multiple structures such as many kinds of constituent
monosaccharides. This fact will lead to the technology of liquid waste disposal of heavy
metal ions in many tropical countries, which are beset with waste problems. Through
these my researches, I hope the polysaccharides and other metabolic products from
cyanobacteria will be utilized as new biomass, medicine and industrial materials.
153

Acknowledgements:
I sincerely would like to thank Prof. Junji Watanabe for bestowing a chance of getting a
doctoral degree on me and being a precious advice about my unprofessional areas and
overall research composition. I could never have completed these works without
Professor Watanabes guidance and encouragement.

I would like to sincerely express my appreciation to Prof. Kazumasa Hirata and
Assistant Prof. Hiroyasu Nagase in Graduate School of Pharmaceutical Sciences,
University of Osaka, and Associate Prof. Takeshi Bamba in Graduate School of
Engineering, Osaka University. They taught about an importance of the application of
microbial metabolite for engineering and gave a chance of starting this research to me.
If it had not been for their guidance, I could not achieve my research.

I would also like to thank Associate Prof. Tatsuo Kaneko. In spite of lacking of my
enough chemical knowledge, he indefatigably educated and guided me at times, and
kindly encourage me at other times. If it were not for all of his supports, I could not
achieve my goal.

I would like to thank assistant Prof. Daisaku Kaneko. He supported me rheological
measurement of sacran solution and gave a novel finding about combined viscoelastic
and liquid crystalline properties.

I would also like to assistant Prof. Tetsu Mitsumata in Graduate School of Science and
Engineering, University of Yamagata. He made grate contributions to our study on
154

dielectric relaxation of sacran without the least regret. Owing to his supports, it was
clarified that the relationship between sacran molecule and metal ions. I believe that the
knowledge from his study will assist on the mechanism of metal ions sorption into
sacran.

I express thanks for the everyday support and help provided by all members of
KANEKO laboratory, especially Mr. Shinji Miyazato who launched the new studying
with me from the beginning and found the possibility of metal adsorption property of
sacran, and by Mr. Yashiro Kaneso who additionally evolve various solution properties
and clarification of higher-order structure of sacran by AFM observation often and often.
Owing to his effort, it was also confirmed the effects of salts to sacran.

I appreciate Kisendo company which provided us A. sacrum (suizenjinori) every time to
study, and advised us important information about mode of life of A. sacrum and natural
environment where A. sacrum is growing, which lead to a very significant ideas to
understand polysaccharide produced by A. sacrum.

I would like to thank Prof. Kiyotaka Kabata in Department of Animal Science, School
of Agriculture, Tokai university. He has been researching ecologies of A. sacrum for a
long time, and then he has given me various knowledge of A. sacrum and has addressed
germination experiments of plants using sacran as a water retention agent.

Finally, I am very happy to be fateful meeting with A. sacrum, and have been studying
the novel polysaccharide from A. sacrum. I will certainly try to develop various
155

products such as metal adsorbent and medicine. I hope that sacran is used as a standard
molecule with micro-scale length.

I thank for my families who encourage me every time, especially my husband, Tatsuo
and my parents.

156

Published Papers:
1) M. K. Okajima, S. Miyazato, T. Kaneko, Chemically Cross-Linking Effects on the
Sorption of Heavy Metal Ions to Hydrogels of Cyanobacterial Megamolecules, Sacran
Trans. MRS-J., in press.
2) M. K. Okajima, S. Miyazato, T. Kaneko The Cyanobacterial Megamolecule Sacran
Efficiently Forms LC Gels with Very Heavy Metal Ions Langmuir ASAP.

3) M. K. Okajima, D. Kaneko, T. Mitsumata, T. Kaneko, J. Watanabe, Cyanobacteria
Produce Megamolecules with Efficient Self-Orientations" Macromolecules 42(8),
2881-3218 (2009)(Cover of the issue)
4) M. K. Okajima, T. Bamba, Y. Kaneso, K. Hirata, S. Kajiyama, E. Fukusaki, T.
Kaneko, Supergiant Ampholytic Sugar Chains with Imbalanced Charge Ratio Form
Saline Ultra-absorbent Hydrogels Macromolecules,41(12), 4061-4064 (2008).

5) M. K. Okajima, S. Miyazato, T. Kaneko, Chemically Cross-Linked Gels Formed by
Novel Supergiant Polysaccharide, Sacran Trans. MRS-J., 33(2), 497-500 (2008).

6) T.Mitsumata, M. K. Okajima, T. Kaneko, Electric Properties of Ionic Polysaccharide
Sacran Aqueous Solutions Trans. MRS-J., 33(2), 431-434 (2008).
7) M. Okajima, M. Ono, K. Kabata, T. Kaneko, Extraction of Novel Sulfated
Polysaccharide from Aphanothece sacrum (Sur.) Okada, and its Spectroscopic
Characterization Pure Appl.Chem., 79(11), 2039-204 (2007).


157

Reference papers and reviews:
1)
309-311
2)
2 4 2009
3)

138-143
CMC 2009
4)
638 73 2008
6)M.Okajima, D.Hayasaka, S.Miyazato, T.Kaneko, Macroscopic birefringence in liquid
crystals from novel cyanobacterial polysaccharide with an extremely high molecular
weight Proceeding of SPIE-International Congress on Optics and Optoelectronics,
6587, 658711/1-658711/8 8 (2007).

Presentation in International Conferences:
1. Maiko Kaneko, Tatsuo Kaneko, Extraction of Novel Sulfated Polysaccharides from
Japan-Indigenous Cyanobacterium and Development of Environmentally-benign
Materials. IUPAC, 1st Green Sustainable chemistry international conference. In
Dresden (2006, September).
(Poster prize was awarded)
2. Maiko Okajima-Kaneko, Daisaku Hayasaka-Kaneko, Shinji Miyazato, Tatsuo
Kaneko, Macroscopic birefringence in liquid crystals from novel cyanobacterial
polysaccharide with an extremely high molecular weight.
SPIE Europe Optics and Optoelectronics. In Praha (2007, April).
3. Maiko Okajima-Kaneko, Shinji Miyazato, Tatsuo Kaneko. Extraction of Novel
Sulfated Polysaccharides from Japan-Indigenous Cyanobacterium and Development of
158

Environmentally-benign Materials. 7st GSC international symposium. In Tokyo (2007,
March).
4. Maiko Okajima-kaneko, Takeshi Bamba, Kazumasa Hirata, Tatsuo Kaneko,
Supergiant Sugar Chains from Jelly-like ECM of Aphanothece sacrum. 18st MRS
academic symposium. In Tokyo (2008).
5. Maiko Okajima-Kaneko, Takeshi Bamba, Kazumasa Hirata, Tatsuo Kaneko,
Supergiant Sugar Chains from Jelly-like ECM of Aphanothece sacrum. 10th Eurasia
Conference on Chemical Sciences. In manila (2008)
6. Miyazato Shinji, Maiko Okajima-Kaneko, Tatsuo kaneko, Efficient Metal-Recovery
by Novel Amphoteric Polysaccharides from Cyanobacteria Aphanothece sacrum
indigenous to Japan. 10th Eurasia Conference on Chemical Sciences. In manila (2008)
7. Maiko Okajima-Kaneko, Shinji Miyazato, Tatsuo Kaneko, Nobel Polysaccharide
Hydrogels Derived From Aphanothece Sacrum And Their Metal Recycling. 3th
International Biotechnology Symposium & Exhibition (IBS-2008). In Dalian (2008).
8. Yasuhiro Kaneso, Maiko Okajima-Kaneko, Tatsuo Kaneko, Novel Supergiant
Cyanobacterial Sugar Chains with Ultra-Absorbency. 3th International Biotechnology
Symposium & Exhibition (IBS-2008). In Dalian (2008)..
9. Takehiro Akashi, Maiko Okajima-Kaneko, Tatsuo Kaneko, Extraction of
chromoproteins from Aphanothece sacrum and their applications to optically-functional
materials. 3th International Biotechnology Symposium & Exhibition (IBS-2008). In
Dalian (2008)..
10. Tatsuo Kaneko, Maiko Okajima-Kaneko, Daisaku Kaneko, Tetsu Mitsumata, Junji
Watanabe,

Ultrahigh-Efficient Liquid Crystallization of Mega-Molecules, Sacran,
Polysaccharides from Aphanothece sacrum. The IUMRS International Conference in
Asia In Nagoya (2008).