JOURNAL
OF BIOSCIENCE
AND BIOENGINEERING
Vol. 87, No. 2, 189-193. 1999
Treatment of Aquarium Water by Denitrifying Photosynthetic Bacteria
Using Immobilized Polyvinyl Alcohol Beads
HISASHI NAGADOMI,’
TAKAKO HIROMITSU,2 KENJI TAKEN0,2
AND KEN SASAK12*
MASANORI
WATANABE,3
Nagao Co. Ltd., 256 M&aura, Okayama 72912121,l Materials and ScienceEngineering, Hiroshima-Denki Institute of
Technology, 6-20-I Nakano, Aki-ku, Hiroshima 739-0321 ,2 and Department of Molecular Biotechnology, Graduate
School of Engineering, Hiroshima University, 1-4-I Kagamiyama, Higashi-Hiroshima 739-0046,3 Japan
Received5 June 19981Accepted6 November 1998
During the purification of an aquarium for carp breeding, a relatively high level of chemical oxygen demand
(COD) was removed by filtration systemspacked with both alginate- and polyvinyl alcohol i.PVA)-immobilized
gel beads of Rhodobacter sphaeroides S. Low nitrate accumulation was observed in the alginate gel beads
packed system due to denitrihcation, but high levels of nitrate and nitrite accumulation were observed in the
PVA gel beads packed system. This phenomenon was caused by the inhibitory effect of PVA on nitrite
reductase. The boric acid used for hardening gel beads of PVA slightly inhibited nitrate reductase. On the other
hand, during the denitrifying growth experiments for this strain, boric acid inhibited cell growth, but PVA only
partially inhibited cell growth. Based on electron equivalent (Y,.), growth yields using various kinds of concentrations of PVA were almost identical. It was suggested that PVA might only limit the growth rate of this
strain by the inhibition of nitrite reductase.
[Key words: denitrification, polyvinyl alcohol, alginate, immobilization, Rhodobacter sphaeroides]
In closed aqua-environments such as fish breeding
ponds and ponds on golf courses, eutrophication has
been frequently observed due to the accumulation of
phosphorous and nitrogenous compounds. This polluted
environment sometimes causes to sudden death of fish
or foul odors. Therefore, circulating filtration systems that
remove organic materials, phosphorous and nitrogenous
compounds have been developed and applied (1). These
systems, however, are usually based on the activities of
microorganisms that grow naturally in a given environment. Sometimes, the capacity of the systems to remove
phosphorous and nitrogenous compounds is not sufficient to prevent eutrophication. In addition, more effective and compact equipment is required to minimize the
space needed and maintenance costs (1).
The removal of nitrogenous compounds such as ammonia, nitrate and nitrite has been performed in bioreactors using denitrifying bacteria (2-5) and polyvinyl alcohol (PVA)-immobilized
photosynthetic bacteria with
denitrifying activity (6). The authors noted in a previous
study that the purification of an aquarium for breeding
fish (carp) could be accomplished using a simple circulating filtration system that was packed with alginate-gelimmobilized photosynthetic bacteria with denitrifying
ability (7). The authors then succeeded in effectively removing chemical oxygen demand (COD) and nitrogenous
compounds. Alginate gel beads, however, decomposed
gradually and the removal effect decreased after 2 weeks.
However, the PVA gels were quite stable over a one
month period (8, 9).
Photosynthetic bacteria can consume various types of
organic matter with a relatively high growth rate and can
consume phosphorous and nitrogenous compounds
simultaneously (10). Therefore, photosynthetic bacteria
have been used for water purification of fish breeding
ponds in Japan. In addition, photosynthetic bacteria
have also been used as a feed supplement to intensify the
color of carp (Nishiki-goi) and to prevent fish disease
during breeding (10).
This study investigates the characteristics of water
purification and the denitrification of PVA-immobilized
photosynthetic bacteria by focusing on the long-term
purification of fish breeding ponds and the maintenance
of the denitrifying ability of the photosynthetic bacterium, Rhodobacter sphaeroides. In addition, the accumulation of nitrite in the water by PVA-immobilized photosynthetic bacteria was also analyzed in order to elucidate the denitrifying activities in PVA gel beads.
MATERlALS
AND METHODS
Microorganism and culture medium
A denitrifying
photosynthetic bacterium, R. sphaeroides S (7, 8) was
used in this study. Stock culture was maintained as
described previously (11).
Glutamate-malate (GM) medium (11) which contained
5 g/l of KN03 was used for denitrifying growth of this
strain. GM medium without KN03 was used for precultivation medium.
Cultivations
Precultivation was performed in a
300-ml conical flask (with a silicon stopper, 200ml medium) for approximately 3 d at 30°C under static-light conditions using two tungsten bulbs (5 klx, 20 W/m2, on the
surface of a culture vessel) as described previously (11).
The main cultivation (denitrifying culture) was performed at 30°C for 5 d under dark anaerobic conditions
in a 1.5-l Roux bottle (with a silicon stopper) containing
GM medium with 5 g/f of KN03. After inoculation with
the preculture broth (2%, v/v), the medium was sparged
with sterile nitrogen gas (99.9%) in order to make it
anaerobic. Culture bottles were mixed vigorously by
hand at 12-h intervals in order to maintain their homogeneity.
After cultivation, cells were harvested by centrifuga-
* Correspondingauthor.
189
190
NAGADOMI ET AL.
tion (10,000 xg, 20min) and washed twice with deionized water, then used immediately for cell immobilization.
Cell immobilization
Cell immobilization was performed according to the conventional immobilization
procedure using Na-alginate (Wako Chemical Co. Ltd.,
Osaka) and polyvinyl alcohol (PVA, Kuraray PVA-HC,
Kuraray Co, Ltd., Osaka) (9). The cell suspension in
deionized water (ODGa=5.0, approximately 2.5 g/c) was
mixed completely with 4% (w/w) autoclaved alginate or
PVA solution and titrated into CaC12 (2%, w/w) or saturated boric acid solution, respectively to form the gel
beads. The gel beads with immobilized cells (3-4mm in
diameter) were then soaked in respective solution for
20 h at 4-7°C in order to harden the gel beads. These gel
beads with immobilized cells were used to purify aquarium water and in denitrifying experiments. The gel beads
were washed several times with large amounts of
dechlorinated tap water (approximately 25-30°C) before
use in order to remove the extra alginate or PVA on the
surface of the gel beads.
Purification of fish breeding water
Fish (carp) breeding was performed as described previously (7) using
25 1 of dechlorinated tap water in an acryl aquarium
equipped with an airlift circulated filter vessel with a
diameter of 10 cm, a height of 7 cm and an air supply of
1 Urnin. In the filter tank, 30 or 60 g (wet base) of gel
beads as packed to the glass wool in the filter vessel.
Five carp of approximately 5 cm in length (Nishiki-goi,
one year old, approximately 10 g) were bred in this
aquarium and were fed with 1.0 g/25 1 of pellet feed
(Swimy mini, Japan Pet Feed Co. Ltd., Tokyo) per day in
each aquarium for 15 to 20d. Three aquariums of the
same type were prepared: one was as a control (no gel
beads) and the other 2 were packed with 30 and 60g
respective gel beads.
Denitrification by gel beads
In order to analyze the
denitrifying characteristics of the gel beads, 0-30g of gel
beads as put into a 1.5-l Roux bottle with 1 1 of
denitrifying medium (10) and denitrification was carried
out for 20 d at 30°C under dark anaerobic conditions
(the medium was sparged with nitogen gas for several
minuts during each sampling period). The medium with
gel beads was gently agitated with a magnetic stirrer (approximately 100rpm) in order to maintain its homogeneity. The composition of the denitrifying medium was as
follows (g/l): glucose 1.0, KN03 1.08, KH2P04 0.11,
(NH&SO4 1.O, yeast extract 0.1, thiamine-HCl 1 x 10 3,
nicotinic acid 1 x lop3 and biotin 1 x low5 (10).
Denitrification
by resting cells
Resting cells were
established in a 300-ml conical flask (with 200 ml liquid
medium with a silicon stopper) using freshly prepared
suspended cells in denitrifying medium so as to separately study the effects of PVA and boric acid on denitrification. The cell concentration was set at ODm=2.0 (1 .Og
cells/l denitrifying medium) and 0, 5 and 10% PVA was
added to each of three conical flasks (300m1, liquid
volume 200 ml). These 3 flasks were incubated for 20d
at 30°C under dark anaerobic conditions. Flasks were
shaken vigorously by hand at 12-h intervals in order to
maintain homogeneity.
For the boric acid tests, 0, 3, 6 g/l of boric acid was
added to cells suspended in a denitrifying medium (OD6m
=2.0) in a 300-ml conical flask (with 200ml liquid
volume) and the cultures were maintained at 4-7’C for
about 20 h in a refrigerator. The suspended cells were
then collected by centrifugation (lO,OOOxg, 20min) and
.I. BI~SCI.
BIOENC~. ,
resuspended in fresh denitrifying medium in a conical
flask (300ml , liquid 200ml) without boric acid so that
the cell concentration was OD660=2.0. The same procedure was conducted for the PVA test with the flasks
incubated for 20 d.
In denitrification by resting cells, KN03 (1.08 g/l) or
KNOz (1.0 g/l) was used as the terminal electron acceptor for denitrification (see Results and Discussion).
Denitrifying growth experiment
A 300-ml conical
flask (200 ml of denitrifying medium) with a silicon stopper was used in the experiment. After the addition of 0
(control), 2 or 5% PVA and 3 or 5 g/l of boric acid to
the medium, the flasks were autoclaved (lOlb, lOmin),
then the preculture broth was added (lo%, v/v). The
flasks were incubated for 3 d at 30°C under dark anaerobic conditions after sparging with pure nitrogen gas.
Flasks were shaken by hand at 12-h intervals to maintain
homogeneity.
Analyses
Cell concentration, COD, residual nitrate
(as mg/l of NOi--N), nitrite (as mg/l of NOT-N), and glucose concentrations were analyzed as described previously (11). Pack-test (Kyouritsu Rikagaku Co. Ltd., Tokyo)
was used for semiquantitative nitrite analysis for carp
breeding experiments.
RESULTS AND DISCUSSION
Carp breeding and water purification by gel beads
The water quality profiles from aquariums for carp
breeding using different gel beads in the water circulation filter were compared. As shown in Fig. la, the COD
increase in the aquarium with alginate gel beads packed
in the filter vessel was slightly suppressed during the 2week carp breeding period compared with that of the
water quality of the control experiment (no gel beads).
?
30
2 20
g
10
0
0
30
7, 20
E
2
I '0
90
Time(d)
FIG.
1. Purification
of an aquarium
for fish (carp) cultivation
using (a) alginateand 0~) aolvvinvl
alcohol
(PVA)-immobilized
gel
beads bf R.-sphaeroide~
s backed in a circ&ting
kdter vessel. Syk
bols: 0, control
(no gel beads);
0, 30 g (wet base) of gel beads;
q ,
60 g (wet base) of gel beads.
VOL.
TREATMENT
87, 1999
OF AQUARIUM
WATER BY PHOTOSYNTHETIC
BACTERIA
191
experiments
In addition,nitratewasnot accumulated
in the water ed in the waterof PVA gel bead-packed
when the alginate gel beads were packed, whereas the
opposite was observed in the control experiment. This
indicates that denitrification is active in the alginate gel
beads as has been noted previously (7). Around the gel
beads, organic matter such as feces of carp and microorganisms in the water was accumulated like sediment
mud. This organic matter was decomposed to volatile
fatty acids or other organic acids by digestion under
microaerobic conditions. Strain S probably utilized these
organic acids as substrates (electron donor). Gel beads,
however, began to decompose after 2 weeks and COD
removal activity began to decrease.
For the PVA gel beads (Fig. lb), COD reduction
was active as similarly observed in the aquarium with
alginate gel beads, however, nitrate accumulation was
observed in all cases. It was suggested that denitrification
was not so active in PVA gel beads compared with in
alginate gel beads. In addition, 2 carp died in the aquarium with 60 g of PVA gel beads which did not occure in
aquariums with alginate gel beads. In that aquarium,
approximately 1.Omg/l of NOT-N accumulation was detect(a)
2
0
5
10
15
20
Time(d)
(b)
(the analysis was semi-quantitative using a test kit).
Generally, nitrite was relatively toxic for the carp (7).
Thus, it was observed that in aquariums with PVA gel
beads denitrification was suppressed and nitrite sometimes accumulated. However, COD reduction with the
PVA gel beads was almost the same as with the alginate
gel beads. In addition, the mechanical stability of the
PVA gel beads was very high compared with the alginate
gel beads (9).
A detailed analysis of the characteristics of denitrification with PVA gel beads was performed.
Denitrifying ability of gel beads
To determine the
denitrification by gel beads in more detail, a denitrifying
medium was used for the denitrification by strain S
because the authors have already analyzed in detail the
characteristics of the denitrifying growth of this strain
using denitrifying medium (11).
As shown in Fig. 2a, COD and nitrate were reduced
when 30g alginate gel beads were added to the
denitrifying medium. In particular, the relatively rapid
reduction of nitrate was observed and approximately
50 mg/l of nitrite accumulation was observed after 10 d,
but this nitrite was reduced again after 20d. This
phenomenon was in good agreement with the denitrification of R. sphaeroides S in which nitrate was reduced as
follows: NOj+NOz+NO+N20+N2
(11). In general,
NO was not observed as an intermediate product of this
step in the denitrification of bacteria (12). In addition,
N20 could not also be detected in the denitrification by
the photosynthetic bacteria, strain S (7, 13). From the
results in Fig. 2a, denitrifying activity including cell
growth was still active in the alginate gel beads.
On the other hand, in the case of PVA gel beads,
COD and nitrate reductions were low as compared with
those in the alginate gel beads. In particular, a large
amount of nitrite accumulation (maximum 85 mg/l) was
observed but reconsumption was not noted when 30g of
PVA gel beads were used. It appears that PVA immobilization of cells seemed to suppress denitrification. In addition, the death of carp in the aquarium with 60g of
PVA gel beads may have been caused by the accumulation of nitrite (approximately 1.O mg/l of NOT-N) in the
water.
Therefore, in order to elucidate the suppression of
denitrification by PVA, the effects of PVA and boric
acid on denitrification were analyzed separately using
resting cells.
Effects of PVA
0
5
10
15
Time(d)
FIG. 2. Denitrification with (a) alginate- and (b) polyvynil alcohol
(PVA)-immobilized
gel beads of R. sphueroides S. Denitrification was
performed in a denitrifying medium (1 f) with 10 and 30 g (wet base)
gel beads under dark anaerobic conditions. Symbols: 0, control (no
gel beads); 0, 10 g gel beads was mixed into the broth; q , 30 g gel
beads was mixed into the broth.
and boric
acid on denitrification
The effects of boric acid and PVA on denitrification
with resting cells are shown in Figs. 3 and 4. In Fig. 3a,
nitrate reduction was suppressed with the increase in the
boric acid concentration. The resting cells were soaked
for ca. 20 h in a boric acid-containing denitrifying medium and then they were incubated for 15 d in a boric
acid-free denitrifying medium. A small amount of nitrite
accumulation was observed in the control experiment
only (Fig. 3a). When nitrite was used as an electron
acceptor in place of nitrate (Fig. 3b), the level of nitrite
consumption was almost the same as that in the control
experiment. These results suggest that boric acid seems
to act as an inhibitor for nitrate reductase during the
20-h soaking period to harden the PVA gel beads.
On the other hand, when PVA was added without
soaking in boric acid, the nitrate comsumption in the
PVA-added experiments was comparable to that in the
192
NAGADOMI
200
ET AL.
J.
-
BIOSCI. BIOENC;.,
(a)
cm
$
o
3
6
9
12
0
l!
4
Time(d)
a
12
16
Time(d)
(b,
200
t
0
0
3
6
9
12
15
Time(d)
FIG. 3. Effects of boric acid on denitrilication in the resting cell
system of R. sphaeroidesS. Cells were suspended in 0,3 or 6 g/l boric
acid in a denitrifying medium within a 300-ml conical flask (containing
200 ml of liquid) and stored at 4-7°C for 20 h. After that, cellswere
collected by centrifugation and washed twice using a boric acid-free
denitrifying medium. Cells were resuspended in fresh denitrifying
medium without boric acid. Cells were incubated for 15 d at 3O“C
under dark anaerobic conditions. Nitrate (a) and nitrite (b) were used
as the electron acceptor in denitrifying medium. Symbols: 0, control
(no boric acid was added); 0,3 g/l of boric acid was added; 0 ,6 g/l
of boric acid was added.
control experiment, however, nitrite was accumulated in
the PVA-added (5 and 10%) experiments (Fig. 4a). In
addition, when nitrite was used as an electron acceptor in
place of nitrate (Fig. 4b), the nitrite consumption in the
PVA-added experiment was reduced as compared with
the control experiment. These results indicate that PVA
mainly inhibits nitrite reductase.
Based on the above results, it was suggested that boric
acid slightly inhibits nitrate reductase and PVA inhibits
nitrite reductase relatively strongly. As a result, a large
amount of nitrite accumulation was observed during the
denitrification with PVA gel beads.
Shen and Hirayama established a bioreactor system
for denitrification, in which the photosynthetic bacterium, R. sphaeroides, was immobilized using the PVAboric acid method (6). However, they did not report on
nitrite accumulation levels. In addition, the purification
of wastewater by PVA-immobilized activated sludge has
also been noted, but there is no report about nitrite
accumulation to date (8, 9).
Effects of PVA on bacterial growth under the deniPVA inhibits nitrite reductase,
trifying condition
0
4
a
12
16
Time(d)
FIG. 4. Effects of polyvinyl alcohol (PVA) on denitrification in
the resting cells of R. sphaeroidesS. Cells were suspended in 2 or 5%
PVA in denitrifying medium and incubated for 16 d at 30°C under
dark anaerobic conditions. Nitrate (a) and nitrite @) were used as the
electron acceptor. Symbols: 0, control (no PVA was added); 0, 5%
PVA was added; q , 10% PVA was added.
therefore, growth experiments with PVA in the medium
were performed. A quantitative analysis was also carried
out to clarify the quantitative relationship between glucose and nitrate consumption in a resting cell system.
As shown in Table 1, cell growth, residual nitrate and
nitrite, and glucose are shown after 3 d culture when the
glucose was almost consumed in the control experiment
(no addition of PVA and boric acid). In the PVA added
experiments, a similar level of nitrite accumulation
(36.9-67.0mg/l)
was observed as in the resting cells,
however, no growth was observed in the boric acidadded experiments due to the long-term contact of cells
with boric acid.
In the denitrifying growth of R. sphaeroides S, the
evaluation of growth yield based on electron equivalent
(Y,,,) has already been analyzed in order to evaluate the
efficiency of energy conversion (11, 12).
Y,,. can be expressed as Eq. 1 (11, 12),
Ye,.= AX/ - (Epq03AN03 -E~ozAN0i )
(1)
where ENo3 and ENoz are the electron equivalents to the
reduction of nitrate and nitrite to NZ, respectively. ENo3
and EN02 are -5 and -3 equivalents/mol, respectively
(11).
On the other hand, the growth yield for glucose, Yx,s
can be expressed as,
VOL.
TREATMENT
81, 1999
TABLE
OF AQUARIUM
WATER BY PHOTOSYNTHETIC
BACTERIA
193
1. Quantitative analyses of glucose and NOT-N consumption and NOT-N accumulation after 3-d culture of R. sphueroides S under
denitrifying condition
Glucose
NOT-N
NOT-N
AGlucose
ANOF-N
AN02 -N
(g7$.)
(g/O
(w/O
(g/l)
(w/l)
@g/l)
Ow/O
3.23
0.05
13.0
35.3
0.95
16.9
2.54
0.25
None
0.98
248
0
0
0
0
no growth
Boric
3 g/l
0
no growth
acid
5 g/l
0.98
250
0
0
0
3.56
0.45
81.9
67.0
0.55
12.1
4.78
0.14
PVA
5%
0.54
9.78
2.64
0.14
3.42
10%
0.46
113
36.9
Initial concentration of elucose. NOT-N and NOT-N in the denitrifving medium were 1.OO,0.25, and 0 g/l, respectively. Indicates the amount
of change of each value after 3 d culture.
a Estimated by Eq. 3 using Yx,s=O.26 g cell/g glucose.
Addition
”
I_
Yx,s = AX/ AS
(2)
where AX is the amount of cell growth (g/l) and AS is
the amount of glucose consumed by cells (g/l).
From Eqs. 1 and 2
(3)
Y,. = Yx,sAS/-(ENO~ANO~ -&o~AN01)
The growth yield during denitrifying growth in glucose
minimal medium (denitrifying medium) was 0.26 g cell/g
glucose. This value was the same as the data in a previous study by us (11). Therefore, using Yx,s and Eq. 3,
Y,,. was calculated for each culture and is shown in
Table 1.
In the control experiment, the Y,,. value was in agreement with that of Pseudomonas denitrificans during
denitrifying growth (3.03 g cell/eq.) (12). In addition,
the Y,,, values in the PVA-added experiments were not
changed. This indicates that PVA did not affect the assimilatory reaction in cells such as the TCA cycle and
protein synthesis. PVA mainly affected the rate of nitrate respiration due to the inhibition of nitrite reductase.
ACKNOWLEDGMENT
This research was supported in part by a Grant-in-Aid
(No.
08650953) for Scientific Research from the Ministry of Education,
Science, Sports and Culture of Japan.
REFERENCES
1. Kikuta, T. and Fukada, A.: Water quality control of aquarium
water for admired fishes, p. 11-17. In Murakami, M. (ed.),
Examples of purification of the waters for environmental use.
Power-Sha, Tokyo (1996).
2. Nilson, I. and Ohlson, S.: Immobilized cells in microbial
nitrate reduction. Appl. Biochem. Biotechnol., 7, 39-41 (1982).
3. Wiffels, R. H., Shukking, G. C., and Tramper, J.: Characterization of a denitrifying bacterium immobilized in K-carrageenan.
Appl. Microbial. Biotechnol., 34, 399-403 (1990).
4. Nilson, I., Oldson, S., Haggstrom, L., MolIin, N., and Mosbath, K.: Denitrification of water using immobilized Pseudomonas denitrificans. Eur. J. Appl. Microiol. Biotechnol., 10, 261274 (1980).
5. Klapwik, A., van der Hoeven, J. C. M., and Lettinga, G.: Biological denitrification
in an upflow sludge blanket reactor.
Water Res., 15, l-6 (1981).
6. Shen, J. and Hirayama, 0.: Denitrification of PVA-immobilized denitrifying
photosynthetic
bacterium, Rhodobacter
sphaeroides. J. Ferment. Bioeng., 75, 43-47 (1993).
7. Sasaki, K., Hashimoto, G., Lin, T., Takeno, K., and Suzuki,
K.: Removal of nitrate ion by denitrification and the purification of fishing pond with immobilized cells. Mizushyori-Gijyutsu,
32, 29-35 (1991). (in Japanese)
8. Hashimoto, S. and Furukawa, K.: Immobilization of activated
sludge by the PVA-boric acid method. Biotechnol. Bioeng., 30,
52-59 (1987).
9. Hashimoto,
S.: Wastewater purification
by immobilized
microorganism. Yousui to Haisui, 29, 725-734 (1987).
10. Sasaki, K., Tanaka, T., and Nagai, S.: Use of photosynthetic
bacteria for the production of SCP and chemicals from organic
wastes, p. 247-290. In Martin, A.M. (ed.). Bioconversion of
waste materials to industrial products, 2nd ed. Blackie Academic & Professional (Chapman & Hall), London, New York,
Tokyo (1998).
11. Sasaki, K., Morii, H., Nishizawa, Y., and Nagai, S.:
Denitrifying and photoheterotrophic
growth of Rhodobucter
sphueroides S under anaerobic-dark and -light conditions. J.
Ferment. Technol., 66, 27-32 (1988).
12. Koike, I. and Hattori, A.: Growth yield of a denitrifying bacterium, Pseudomonas
denitrificnns,
under aerobic and
denitrifying conditions. J. Gen. Microiol., 88, l-10 (1975).
13. Satoh, T., Hoshino, Y., and Kitamura, H.: Rhodopseudomonas sphaeroides forma sp. denitrr$cans, a denitrifying strain as
a subspecies of Rhodopseudomonas sphaeroides. Arch. Microbiol., 108, 265-269 (1976).