JOURNAL OF FERMENTATION AND BIOENGMEERING

Vol. 83, No. 1, 17-20. 1997

Fermentative Metabolism to Produce Hydrogen Gas and Organic

Compounds in a Cyanobacterium, Spirulina platensis
KATSUHIRO AOYAMA,1.2 IEAKI UEMURA,2 JUN MIYAKE,3s4 AND YASUO ASADA3*
Research Institute of Innovative Technology for the Earth, I-1 Higashi, Tsukuba, Ibaraki 305, Frontier Technology Research
Institute, Tokyo Gas Co. Ltd., l-7-7 Suehiro-cho, Tsurumi-ku, Yokohama 230,2 National Institute of Bioscience and
Human-Technology, AIST/MITI, I-l Higashi, Tsukuba, Ibaraki 305,3 and National Institute of Advanced
Interdisciplinary Research, AIST/MITI, l-l-4 Higashi, Tsukuba, Ibaraki 305,4 Japan

Received 24 June 1996/Accepted 26 October 1996

The non nitrogen-fixing and 6lamentous cyanobacterium Spirulina platensis NIES-46 produced hydrogen
gas, ethanol, and low molecular organic acids auto-fermentatively under dark and anaerobic conditions. The
fermentative productivity was enhanced by incubating the cyanobacterium under nitrogen-starved conditions.
Cell-free extracts of the cyanobacterium catalyzed hydrogen production by the addition of acetyl-coenzyme A
and pyruvate. Pyruvate-degrading and acetaldehyde dehydrogenase activities were observed in the cell-free
extracts. These results suggest that the fermentation was dependent on the anaerobic degradation of en-
dogenous glycogen via pyruvate.

[Key words: Spirulina, hydrogen, organic acid, ethanol]

Since Gaffron and Rubin (1) discovered auto-fermenta- with air for mixing.
tive and light-dependent hydrogen production by a green Fermentation and analysis of products The cyano-
alga, Scenedesmus sp., hydrogen production has been bacterial cells were harvested in the late logarithmic
extensively studied not only with green algae but also growth phase by filtration using a filter paper (no. 2
with cyanobacteria (for a review, see 24). One of the filter paper, Advantech-Toyo, Tokyo), washed, and resus-
enzymes responsible, hydrogenase, has also been studied pended in sodium phosphate buffer (20mM, pH 7.0,
in these photoautotrophic organisms (for a review, see 5, unless otherwise stated). Twenty ml of the cell suspen-
6). Cyanobacterial hydrogenases, were first investigated sion was sealed in an Erlenmeyer flask (60ml total vol-
in the context of the recovery of hydrogen gas produced ume) with a rubber stopper and the gas phase was re-
by a nitrogenase-dependent mechanism (7, 8). Subse- placed with nitrogen gas. The flasks were shaken for a
quent studies, however, demonstrated that these hy- period ranging from overnight to one day in darkness
drogenases can be active in the evolution of molecular at 30C. A small portion of the gas phase was occasion-
hydrogen, especially in dark and fermentative metab- ally withdrawn by a pressure-lock syringe and injected
olism, by various cyanobacteria (9-14). into a GCRlA (Shimadzu, Kyoto) gas chromatograph
The cyanobacterium Spirulina maxima was demonstrat- equipped with a molecular sieve 5A column (18) to deter-
ed to contain hydrogenase nearly 2 decades ago (15), but mine the hydrogen concentration. Organic compounds
its physiological role remained obscure. We were the first excreted by S. platensis were analyzed by gas chromatog-
to demonstrate fermentative hydrogen production by raphy using a Shimadzu GCRlA gas chromatograph
Spirulina species (11). Heyer and Krumbein (16) deter- equipped with a Tenax GC (Waters, Milford, MA, USA)
mined the excretion products in fermentation by column, or by a Shimadzu LC8A liquid chromato-
cyanobacteria, including Spirulina, in which hydrogen graph equipped with a Shimadzu Shim-pack SCR-1OlH
production was supposed to be involved. However, little column.
or no information exists on the characteristics of in vitro Determination of glycogen The glycogen content in
fermentative activities, especially the electron donating S. platensis cells was determined according to Ernst et al.
reaction for hydrogen production. Here, we describe (19).
hydrogen production and the related fermentative Preparation of cell-free extracts S. platensis cells
metabolism by intact cells and cell-free extracts of harvested in the late logarithmic growth phase were
Spirulina platensis NIES-46. washed and suspended in 50 mM Tris-HCl, pH 7.5, and
then passed through a French press (Ohtake, Tokyo) at
800 kg/cm2. After centrifugation at 2,400 x g for 10 min,
MATERIALS AND METHODS
the supernatants were used as cell-free extracts.
Strain and cultivation An axenic culture of the Determination of protein concentration The pro-
cyanobacterium S. platensis NIES-46 was obtained tein concentrations of the cell-free extracts were deter-
from the National Institute of Environmental Studies, mined according to Peterson (20).
the Environment Agency, Japan. The cyanobacterium Analysis of dissolved hydrogen gas Production of
was cultivated at 30C in SOT medium (17) under con- hydrogen gas by cell-free extracts were measured by
tinuous illumination of fluorescent lamps (about 5 continuously monitoring dissolved hydrogen gas with a
klx). Nitrogen-free in the text indicates the omission hydrogen (Uebayasi, M. et al., Abstr. Ann. Meet. Sot.
of sodium nitrate from SOT. Cultures were bubbled Plant Physiol., Japan, p. 165, 1981)-oxygen electrode
(model 5300, Yellow Springs Instruments, USA) system
* Corresponding author. (10). The experiments were done under anaerobic condi-

17
18 AOYAMA ET AL. J. FERMENT.BIOENG.,

tions by previously sparging the reaction mixture with

argon gas. The oxygen electrode was used to ascertain
anaerobic conditions.
Assay of pyruvate-degrading and related enzyme activi-
ties Alcohol dehydrogenase, pyruvate decarboxylase,
and acetaldehyde dehydrogenase in the cell-free extracts
were measured by the change in the absorbance at 340
nm due to oxidation of NADH. Reaction mixtures con-
sisted of the S. platensis extract, 50mM Tris-HCl buffer
(pH 7.9, and 0.5 mM NADH with the following addi-
tions: for alcohol dehydrogenase, 1 mM acetaldehyde
as a substrate; for pyruvate decarboxylase, 20mM pyru-
vate as a substrate and an excess amount of alcohol
dehydrogenase from yeast (Boehringer-Mannheim-
Yamanouchi); for acetaldehyde dehydrogenase (acylat- 0 6 12 18 24
ing), 5 mM acetyl-coenzyme A as a substrate and an ex- Time (h)
cess amount of yeast alcohol dehydrogenase.
FIG. 2. Effects of cell concentration and nitrogen starvation on
fermentative hydrogen production by S. platensis. Nitrogen-starved
RESULTS AND DISCUSSION cells of S. platensis prepared by incubation for 72 h as described in Fig.
1 were suspended in 20 mM sodium phosphate, pH 7.0, at different cell
Fermentative production of hydrogen gas and organic concentrations. The following open symbols indicate nitrogen-starved
compounds by intact cells S. platensis NIES-46 ac- cells: 0, 1.624; n, 4.060; 0, 8.700 (mg dry wt/ml of suspension).
cumulated glycogen up to about 15-50X of the cell dry Filled circles (0) indicate non-nitrogen-starved cells at a concentra-
wt. when photoautotrophically incubated for 2-3 d in a tion of 1.624 mg dry wt/ml of suspension.
nitrogen-free culture medium (Fig. 1). Cells incubated
for 3 d in the nitrogen-free medium had higher fermenta- hydrogenases are acid-labile (15, 22), a decrease in in-
tive hydrogen production activity than those grown in tracellular pH could also affect hydrogen production
the ordinary culture medium (Fig. 2). No substrates for (refer also to Fig. 4).
hydrogen production were supplied artificially. Glycogen Time courses of the production of organic compounds
could thus be a candidate for the source of hydrogen by the thin suspension are shown in Fig. 3. The organic
gas, in which case the intracellular content of glycogen products-mainly acetate, formate, lactate, and ethanol
would supposedly be associated with the hydrogen pro- -all increased with the passage of time. The product
ducing activity. profile of S. platensis NIES-46 is rather different from
The rate and sustainability of hydrogen production those of the three strains of Spin&a reported by Heyer
were also investigated in suspensions with different cell and Krumbein (16).
concentrations. Hydrogen production was found to be The effects of the pH and concentration the of buffer
inversely associated with the cell concentration (Fig. 2). solution on the production of hydrogen and organic
Likely explanations for this could be product inhibition acids were also investigated (Fig. 4). The hydrogen pro-
or a change of pH; fermentation products or intracellu- ductivity was less at lower pH and in the lower buffer
lar pH might reach an inhibitory level earlier in a dense concentration, which is in accordance with our supposi-
suspension than in a thin one. Since cyanobacterial tion from the results of Fig. 2, that hydrogen production
could be inhibited by acidic products. In the case of the
1.2, 160

0.0; m
20 40 60
J
so0 .

0 6 12 18 24
Tie(h)
Time (h)
FIG. 1. Accumulation of glycogen in S. platensis during incuba-
tion under nitrogen-starved conditions. S. platensis cells were asepti- FIG. 3. Production of hydrogen and organic compounds by
cally harvested in the late logarithmic growth phase, washed with a nitrogen-starved cells of S. platensis. Nitrogen-starved cells of S.
nitrogen-free culture medium, transferred into the nitrogen-free cul- platensis prepared as described in Fig. 2 were suspended at a concen-
ture medium and then incubated under the same light and temperature tration of 1.624 mg dry wt/ml, and incubated under the same condi-
conditions as those used for cultivation. Symbols: 0, cell concen- tions as in Fig. 2. Symbols: 0, hydrogen gas; 0, acetate; 0, for-
tration; A, glycogen content. mate; A, lactate; 0, ethanol.
VOL. 83, 1997 FERMENTATIVE METABOLISM OF SPIRULINA 19

TABLE 2. Enzyme activities related to pyruvate degradation in the

cell-free extracts of S. platensis
Activity
(nmol/min/mg protein)
Alcohol dehydrogenase 0.76
Pyruvate decarboxylase 19.07
Acetaldehyde dehydrogenase (acylating) 1.83
The reaction mixture contained 1.5mg/ml of protein.

Therefore, the electron donation for hydrogenase is

thought to involve ferredoxin. However, soluble fer-
redoxin purified from Microcystis aeruginosa NIES-44, a
al- similar cyanobacterial fermenter to Spirulina (9, 10) did
6 7 8
not work as an electron mediator for the hydrogenase
partially purified from M. aeruginosa (22). The hypothe-
PB sis by Krogmann (23) that cytochrome csso may mediate
FIG. 4. Effects of pH and buffer concentration on production of the donation of electrons from ferredoxin to hydrog-
hydrogen and organic acids. Nitrogen-starved cells of S. platensis enase in cyanobacteria needs to be proved with purified
prepared by incubation for 72 h as described in Fig. 1 were suspended proteins.
at a concentration of 4.06Omg dry wt/ml of suspension in 20 or Pyruvate-degrading and related enzyme activities
100 mM sodium phosphate buffer, at pH 6.0, 7.0, or 8.0. Symbols: The pyruvate-degrading enzyme activities in cell-free
0 , n , hydrogen; 0, 0, acetate; A, A, formate (open and filled sym-
bols indicate 20 and 100 mM sodium phosphate buffer, respectively).
extracts were detected and are shown in Table 2. It is
suggested that pyruvate might be degraded to acetyl-
CoA with ferredoxin reduction by ferredoxin: pyruvate
20-mM buffer at pH 7.0, the pH was reduced to 6.6 after oxidoreductase and simultaneously to acetaldehyde by
the fermentation. Production of the organic acids them- pyruvate decarboxylase, and that acetyl-CoA can be
selves, however, was also lower at acidic pH and at the further metabolized in two ways; to acetate by hydrolytic
lower buffer concentration. reactions and acetaldehyde by acetaldehyde dehydrog-
Hydrogen production by cell-free extracts The enase (acylating).
hydrogen producing activity of cell-free extracts in the Conclusion Although further studies are required
presence of various electron donors was assayed (Table to determine the quantitative metabolic fluxes and direct
1). The maximal rate of hydrogen production was ob- electron mediator for the hydrogenase, the metabolic
served in the presence of an artificial electron mediator, pathways for the production of hydrogen gas and organ-
methyl viologen reduced by an excess amount of ic compounds are hypothesized to be roughly as shown
hydrosulfite. Among the physiological electron donors, in Fig. 5.
pyruvate plus CoA was significantly effective, giving a Besides the studies on fermentative metabolism report-
rate near that in the presence of hydrosulfite alone, ed here, we are also investigating the utilization of the
which was supposed to fully reduce an electron donor in fermented organic compounds by photosynthetic bacte-
the extracts. The rate of hydrogen production in the ria for material production, the results of which will be
presence of NADH and NADPH with and without ATP published elsewhere.
was much less than that with pyruvate plus CoA.
The results in Table 1 suggest that the supply of reduc-
ACKNOWLEDGMENTS
ing power for hydrogenase may involve CoA-dependent
cleavage of pyruvate by the reaction of pyruvate-fer- This work was performed under the management of Research
redoxin oxidoreductase, which has been shown to be the
case in Anabaena (21): pyruvate alcohol
decarboxylase dehydrogenase
pyruvate + Fdox + CoA-tacetyl-CoA + CO1 + Fdred glycogen+** pyruvate U acetaldehyde -> ethanol

TABLE 1. Hydrogen production by cell-free extracts of S. platensis

Additive HZ production rate

(nmol/min/mg protein)
None
Pyruvate 20 mM
0.0110
0.0000
1 I A NADH
acetaldehyde
dehydrogenase
Pyruvate 20 mM + Coenzyme A 0.4 mM 0.0856 (acylating)
(cyt csso?) CH3CO-COA
a-Ketoglutarate 20 mM + Coenzyme A 0.4 mM 0.0007
Formate 10 mM 0.0090
NADH 5 mM 0.0000
NADH 5 mM + ATP 5 mM 0.0007 1
hydrogenase I
NADPH 5 mM 0.0009
NADPH 5 mM + ATP 5 mM 0.0009 CH,CCOH + CoA
Hydrosulfite 5 mM 0.102 1
Methyl viologen 1 mM + hydrosulfite 5 mM 0.731 Hz

The reaction mixture contained 6.2mg/ml of protein and the FIG. 5. Hypothetical pathway of fermentative production of
additive. hydrogen gas and organic compounds by S. platensis.
20 AOYAMA ET AL. J. FERMENT.BIOENG.,

Institute of Innovative Technology for the Earth (RITE) as a part evolve molecular hydrogen under dark and anaerobic condi-
of the Project for Biological Production of Hydrogen supported by tions. J. Ferment. Technol., 64. 553-556 (1986).
the New Energy and Industrial Technology Development Organiza- 12. Howartb, D. C. and Codd,. G.A.: The uptake and production
tion (NEDO). of molecular hydrogen by unicellular cyanobacteria. J. Gen.
Microbial., 131, 1561-1569 (1985).
REFERENCES 13. Almoo, H. and Biiger, P.: Hydrogen metabolism of the unicel-
lular cyanobacterium Chroococcidiopsis thermalis ATCC29380.
1. Gaffron, H. and Rubin, J.: Fermentative and photochemical FEMS Microbial. Lett., 49, 445-449 (1988).
production of hydrogen in algae. J. Gen. Physiol., 26, 219-240 14. Oost, J. and Cox, R. P.: Hydrogenase activity in nitrate-grown
(1942). cells of the unicellular cyanobacterium Cyanothece PCC 7822.
* Smith, G. D.. Ewart. G. D., and Tucker, W.: Hydrogen
L. Arch. Microbial., 151, 40-43 (1989).
production by cyanobacteria. Int. J. Hydrogen Energy, 17, 15. ___
Llama. M. J.. Serra. J. L.. Rao. K. K.. and Hall. D. 0.: Isot=
695-698 (1992). tion and characterization of the hydrogenase activity from the
3. Miyamoto, K.: Hydrogen production by photosynthetic bacte- non-heterocystous cyanobacterium Spirulina maxima. FEBS
ria and microalgae, p. 771-786. In Murooka, Y. and Imanaka, Lett., 98, 342-346 (1979).
T. (ed.), Recombinant microbes for industrial and agricultural 16. Heyer, H. and Krumbein, W. E.: Excretion of fermentation
applications. Marcel Dekker, New York (1993). products in dark and anaerobically incubated cyanobacteria.
4. Markov, S. A., Bazin, M. J., and Hall, D. 0.: The potential of Arch. Microbial., 155, 284-287 (1991).
using cyanobacteria in photobioreactors for hydrogen produc- 17. Ogawa, T. and Ten& G.: Studies on the growth of Spiruiina
tion. Adv. Biochem. Eng. Biotech., 52, 60-86 (1995). platen&, I. On the pure culture of Spirulina platensis. J. Fer-
5. Houchins, J. P.: The physiology and biochemistry of hydrogen ment. Technol., 48, 361-367 (1970).
metabolism in cyanobacteria. Biochim. Biophys. Acta, 768, 18. Asada, Y., Tonomura, K., and Nakayama, 0.: Hydrogen evo-
227-255 (1984). lution by an isolated strain of Anabaena. J. Ferment. Tech-
6. Schulz, R.: Hydrogenase and hydrogen production in eukaryot- nol., 57, 280-286 (1979).
ic organisms and cyanobacteria. J. Marine Biotech., 4, 16-22 19. Ernst, A., Kirschenlohr, H., Diez, J., and Boger, P.: Glycogen
(1996). content and nitrogenase activity in Anabaena variabilis.
7. Eisbrenner, G., Roos, P., and Bothe, H.: The number of Arch.Microbiol., 140, 120-125 (1984).
hydrogenases in cyanobacteria. J. Gen. Microbial., 125, 383- 20. Peterson, G. L.: A simplification of the protein assay method
390 (1981). of Lowry et al. which is more generally applicable. Analyt.
8. Stewart, W. D. P.: Some aspects of structure and function in Biochem., 83, 346-356 (1977).
N2-tixing cyanobacteria. Ann. Rev. Microbial., 34, 497-536 21. Bothe, H., Falkenberg, B., and Nolteernsting, U.: Properties
(1980). and function of the pyruvate: ferredoxin oxidoreductase and
9. Asada, Y. and Kawamura, S.: Hydrogen evolution by from the blue-green alga Anabaena cylindrica. Arch.
Microcystis aeruginosa in darkness. Agric. Biol. Chem., 48, Microbial., 96, 291-304 (1974).
2595-2596 (1984). 22. Asada, Y., Kawamura, S., and Ho, K.-K.: Hydrogenase from
10. Asada, Y. and Kawamura, S.: Hydrogen evolving activity the unicellular cyauobacterium, Microeystis aeruginosa.
among the genus, Microcystis, under dark and anaerobic condi- Phytochemistry, 26, 637-640 (1987).
tions. Rept. Ferment. Res. Inst., 63, 39-54 (1985). 23. Krogmann, D. W.: The low-potential cytochromes in cyanobac-
11. Asada, Y. and Kawamura, S.: Screening for cyanobacteria that teria and algae. Biochim. Biophys. Acta, 1058, 35-37 (1991).