Anaerobic Coculture of Microalgae with Thermosipho globiformans

and Methanocaldococcus jannaschii at 68°C Enhances Generation of

n-Alkane-Rich Biofuels after Pyrolysis
Kunio Yamane, Shigeru Matsuyama, Kensuke Igarashi, Motoo Utsumi, Yoshihiro Shiraiwa, Tomohiko Kuwabara
Faculty of Life and Environmental Sciences, University of Tsukuba, Tsukuba-Shi, Ibaraki, Japan

We tested different alga-bacterium-archaeon consortia to investigate the production of oil-like mixtures, expecting that n-al-
kane-rich biofuels might be synthesized after pyrolysis. Thermosipho globiformans and Methanocaldococcus jannaschii were
cocultured at 68°C with microalgae for 9 days under two anaerobic conditions, followed by pyrolysis at 300°C for 4 days. Arthro-
spira platensis (Cyanobacteria), Dunaliella tertiolecta (Chlorophyta), Emiliania huxleyi (Haptophyta), and Euglena gracilis
(Euglenophyta) served as microalgal raw materials. D. tertiolecta, E. huxleyi, and E. gracilis cocultured with the bacterium and
archaeon inhibited their growth and CH4 production. E. huxleyi had the strongest inhibitory effect. Biofuel generation was en-
hanced by reducing impurities containing alkanenitriles during pyrolysis. The composition and amounts of n-alkanes produced
by pyrolysis were closely related to the lipid contents and composition of the microalgae. Pyrolysis of A. platensis and D. tertio-
lecta containing mainly phospholipids and glycolipids generated short-carbon-chain n-alkanes (n-tridecane to n-nonadecane)
and considerable amounts of isoprenoids. E. gracilis also produced mainly short n-alkanes. In contrast, E. huxleyi containing
long-chain (31 and 33 carbon atoms) alkenes and very long-chain (37 to 39 carbon atoms) alkenones, in addition to phospholip-
ids and glycolipids, generated a high yield of n-alkanes of various lengths (n-tridecane to n-pentatriacontane). The gas chroma-
tography-mass spectrometry (GC-MS) profiles of these n-alkanes were similar to those of native petroleum crude oils despite
containing a considerable amount of n-hentriacontane. The ratio of phytane to n-octadecane was also similar to that of native
crude oils.

C rude petroleum oil reservoirs are generally located in deep

geological formations with environments characterized by
anaerobic, high-temperature, and high-pressure conditions that
harbor various microorganisms (1, 2). Petroleum fluids are pro-
duced by biological modification and thermal cracking of fossil
organic materials in source rocks during geologically long-term
subsurface storage, and expelled fluids migrate toward traps that
are eventually formed by reservoir and cap rocks (3). The major
detectable hydrocarbons in crude oils are n-alkanes and related
compounds (4). Understanding the activity of such microorgan-
isms is essential to search for potential associations with crude oil
generation.
We previously sampled mixtures of crude oils and production
water directly from subsurface oil deposits to search for subsur-
face microorganism activity during oil generation. Analysis of 16S
rRNA gene sequences led to the identification of a Thermacetoge-
nium-like bacterium (92% identity in 16S rRNA gene sequence
with Thermacetogenium phaeum) and a Methanothermobacter-
like archaeon (96% identity with Methanothermobacter thermau-
totrophicus) as the major indigenous populations in one oil deposit at
74°C (5). Cocultured T. phaeum and M. thermautotrophicus syn-
trophically degraded acetate to CH4 at 55°C under anaerobic condi-
tions in a methanogenic reactor for wastewater (6). Based on these
findings, we hypothesized that syntrophic pairs of bacteria and ar-
chaea would be the major indigenous microorganisms in oil reser-
voirs and that thermophilic pairs might have some involvement in
the subsurface generation of petroleum from organic materials. We
then supposed that the oil generation process could be analyzed in
subsurface microalgae during a short period under anaerobic, high-
temperature, and atmospheric conditions using syntrophic pairs that
have rapid-growth ability. However, we could not isolate either the
bacterium or the archaeon from either production water or oils from
our samples.
On the other hand, Kuwabara et al. (7) isolated a thermostable,
anaerobic bacterium (Thermosipho globiformans) from a hydro-
thermal vent at Suiyou Seamount in the Western Pacific Ocean.
The growth of this bacterium at 68°C under anaerobic conditions
was enhanced 5- to 7-fold in coculture with the thermostable ar-
chaeon Methanocaldococcus jannaschii (8, 9) compared with mon-
oculture. Furthermore, cocultures of the two microorganisms
rapidly proliferated and reached the stationary phase of growth
within 16 h (K. Igarashi and T. Kuwabara, unpublished data).
This growth was similar to that of cocultures of Thermotoga
maritima and M. jannaschii at 80°C (10, 11), whereas it was 30- to
50-fold faster than that of pairs of T. phaeum and M. thermau-
totrophicus at 55°C (6) and of Pelotomaculum thermopropionicum
and M. thermautotrophicus at 55°C (12). Therefore, we cocultured
T. globiformans and M. jannaschii because they comprised a bac-
terial and archaeal pair similar to those found in the subsurface oil
deposit. We then tested various alga-bacterium-archaeon consor-
tia to determine the generation of oil-like mixtures within a short
period under anaerobic, thermophilic, and atmospheric condi-
Received 29 August 2012 Accepted 20 November 2012
Published ahead of print 26 November 2012
Address correspondence to Kunio Yamane, yamane.kunio.fb@u.tsukuba.ac.jp.
Supplemental material for this article may be found at http://dx.doi.org/10.1128
/AEM.01685-12.
Copyright © 2013, American Society for Microbiology. All Rights Reserved.
doi:10.1128/AEM.01685-12

924 aem.asm.org Applied and Environmental Microbiology p. 924 –930 February 2013 Volume 79 Number 3

tions. The processes involved in oil generation, such as n-alkane
production, can be duplicated by laboratory pyrolysis. Higher
temperatures are needed to drive these reactions within a few
hours or days rather than the millions of years in nature. Pyrolysis
is an established and accepted research technology that has been
applied to generate liquid fuels from higher plants, including
woody materials (13). The production of biofuels from marine
microalgae has also been studied in detail because microalgae are
primarily renewable producers in oceans and they constitute the
largest biomass in nature. Furthermore, they do not compete with
the food industry for starch stocks and arable land (14–17).
The present study aimed to compress the time required to gen-
erate oil from microalgae and to mimic the subsurface reaction
using rapidly proliferating model cocultures of T. globiformans
and M. jannaschii with pyrolysis at 300°C for 4 days. This strategy
generated high yields of n-alkane-rich oils from microalgae. We
also identified a close relationship between the n-alkane composition
in hexane subfractions after pyrolysis and the lipid composition of
Arthrospira platensis (Cyanobacteria), Dunaliella tertiolecta (Chloro-
phyta), Emiliania huxleyi (Haptophyta), and Euglena gracilis (Eugle-
nophyta) microalgae as biofuel sources. A considerable amount of
n-alkanes with compositions similar to those of native petroleum
crude oils was generated from E. huxleyi, which contains more long-
chain alkenes and very long-chain alkenones (18).
E. huxleyi is a unicellular, calcifying coccolithophorid alga of
the phylum Haptophyta that is widely distributed in oceans, where
it frequently forms massive blooms that can cover ⬎100,000 km2
of the ocean surface. During such blooms, most of the algae tend
to sink and sediment on the ocean floor (19, 20). Alkenones are
less labile and have been identified in sea sediments from the Eo-
cene age and in Cretaceous black shale (21).
MATERIALS AND METHODS
Bacterial and archaeal strains. The bacterial strain Thermosipho globifor-
mans MN14T was isolated from a hydrothermal vent at Suiyo Seamount,
Izu-Bonin Arc, western Pacific Ocean (28°34=N, 140°38=E), at a depth of
1,384 m and a temperature of 230°C and was identified in our laboratory
(7). The archaeal strain Methanocaldococcus jannaschii DSM 2661 was
isolated from black sediments sampled at the East Pacific Rise hydrother-
mal vents at 21oN (8, 9). The strain was purchased from Deutsche Sam-
mlung von Mikroorganismen und Zell Kulturen (DSMZ) GmbH, Braun-
schweig, Germany. Monocultures of T. globiformans and M. jannaschii
grow poorly in Tc medium without sulfur powder (Tc-So [see below]) at
68°C under N2-H2-CO2 (80:10:10) and H2-CO2 (80:20) gas phases, re-
spectively. The cell densities were 7 ⫻ 107 and 5 ⫻ 106 cells/ml, respec-
tively, after 16 h of cultivation. In contrast, cocultures of the two strains at
68°C for 16 h in Tc-So medium under a gas phase of N2-H2-CO2 (80:10:
10) stimulated their growth 5- to 7-fold (T. globiformans, 4 ⫻ 108 cells/ml;
M. jannaschii, 3 ⫻ 107 cells/ml).
The Tc-So medium contained (liter⫺1) 25 g NaCl, 0.33 g KCl, 2.8 g
MgCl2 · 6H2O, 1.66 g MgSO4, 0.3 g K2HPO4, 0.25 g NH4Cl, 10 mg NaBr,
25 mg Fe2SO4 · 7H2O, 10 ml each of trace mineral and vitamin solutions
(22), 3 g of yeast extract (Difco), 3 g of tryptone (Difco), and 1 mg of
resazurin. Sulfur powder was omitted from the Tc medium (7). After
adjusting the pH to 6.5, the medium was sterilized by autoclaving and
stored at room temperature.
Before anaerobic cultivation, the medium was anoxidized by adding
1/20 volume of 10% (wt/wt) Na2S · 9H2O and 10% (wt/wt) L-cysteine HCl
in an anaerobic workstation (Ruskinn Technology Ltd., Leeds, United
Kingdom) under an N2-H2-CO2 (80:10:10) gas phase. Separate T. globi-
formans and M. jannaschii monocultures were simultaneously inoculated
into serum bottles (final concentrations, 1 ⫻ 106 and 1 ⫻ 105 cells/ml,
February 2013 Volume 79 Number 3
n-Alkane-Rich Biofuels from Microalgae
respectively) with an initial N2-H2-CO2 (80:10:10) gas phase in the work-
station.
Microalgae. Dry, powdered freshwater algae, A. platensis (Cyanobac-
teria) and E. gracilis (Euglenophyta), were provided by DIC Life Tec Co.
Ltd., Tokyo, Japan, and by Euglena H1 Co. Ltd., Tokyo, Japan, respec-
tively. Saltwater algae, D. tertiolecta NIES 2258 (Chlorophyta) and E.
huxleyi NIES 837 (Haptophyta), were cultured in artificial seawater (Marine
Art SF-1; Osaka Yakken, Osaka, Japan) enriched with Erm-Schreiber me-
dium (ESM) and 10 nM sodium selenite. The ESM, containing (liter⫺1)
12 g NaNO3, 0.5g K2HPO4, 25.9 mg FeEDTA, 33.2 mg MnEDTA, 10 mg
thiamine-HCl, 0.1 mg biotin, 0.1 mg cyanocobalamin, and 100 g Tris at
pH 8.0, was mixed (10 ml) with 990 ml of artificial seawater immediately
before use (23). The cells were constantly illuminated at 100 ␮mol of
photons m⫺2 s⫺1 by fluorescent lights in a 2-liter flat, oblong glass vessel
aerated by air bubbling at 20°C for 12 days. The cells were harvested by
centrifugation at 5,000 ⫻ g for 10 min, lyophilized, and stored at 5°C.
Microscopy. Bacterial and archaeal monocultures and cocultures
were routinely sampled for cell counting using a Nikon model Eclipse
E600 microscope with phase-contrast and epifluorescence capabilities.
Pyrolysis. Lyophilized powdered algae (100 mg each) were placed in
Pyrex glass tubes (12 by 120 mm) to make ampoules. The air was com-
pletely replaced with nitrogen gas, and then the ampoules (12 by 70 to 75
mm) were evacuated using an oil vacuum pump, sealed, and heated in a
muffle furnace at 300°C for 4 days. Thereafter, the ampoules were cooled
and connected to a silicon tube, one side of which was sealed using a
syringe sampler with a silicon septum. The air in the capped silicon tubes
was evacuated, the tip of the ampoule was broken, 0.5 ml of the gas phase
of the pyrolysis product was withdrawn, and hydrocarbons were analyzed
by gas chromatography (GC). The remaining organic materials in the
ampoules were Vortex mixed and then extracted twice with 0.5 ml of
chloroform to analyze n-alkanes and other hydrocarbons.
Lipid analysis. Microbial cells were sedimented by centrifugation, and
then total lipids were extracted using chloroform-methanol as described
by Ames (24). The total lipids in the chloroform fraction were dried by an
evaporator and nitrogen blow down, and then they were dissolved in 1.0 ml of
5% hydrogen chloride-methanol (Wako Pure Chemicals Industries Ltd.,
Osaka, Japan) and heated at 100°C for 60 min for methanolysis. The samples
were dried by an evaporator and nitrogen blow down, dissolved in 1 ml of
chloroform, and fractionated by silica gel column chromatography.
Silica gel column chromatography. A 150- by 5-mm column was
plugged at the bottom with glass wool, rinsed with methanol, left to dry,
and then dry packed with 0.5 g of activated Wako Gel C200 silica gel
(Wako Pure Chemical Industries Ltd.). The gel was conditioned with 5 ml
of hexane, which was discarded. Dried samples were suspended in hexane,
applied to the column, and washed with an additional 0.5 ml of hexane.
Samples were then eluted stepwise with 5 ml of hexane, 5 ml of 5% (vol/
vol) diethyl ether in hexane (5% EH), 5 ml of 30% (vol/vol) diethyl ether
in hexane (30% EH), and 5 ml of diethyl ether. The eluates were collected
in calibrated glass centrifuge tubes, concentrated by evaporation, and
dried under nitrogen gas. Dried samples were weighed, dissolved in 0.5 ml
of chloroform, and stored at ⫺20°C.
GC-mass spectrometry (MS). The lipid and oil composition was an-
alyzed using a 6890N gas chromatograph (Agilent Technology Inc., Santa
Clara, CA) equipped with an MS-600H mass spectrometer (JEOL, Tokyo,
Japan) and a 30-m by 0.25-mm (inside diameter [i.d.]) (0.25-␮m film)
DB-5MS fused-silica capillary column (Agilent Technology). The carrier
gas was helium (1 ml/min). The injection and detector temperatures were
set at 300°C and 320°C, respectively. The temperature program comprised
1 min at 50°C, ramping to 320°C in 10°C/min increments, and holding for
12 min.
Coculture of T. globiformans and M. jannaschii with sonicated mi-
croalgae. The microalgae A. platensis, D. tertiolecta, E. huxleyi, and E.
gracilis (0.5 g each) were suspended in 5 ml of Tc-So medium, frozen at
⫺20°C, thawed, and sonically disrupted for 5 min at 5 kc with glass beads
(except for E. huxleyi). The pH of the suspensions was adjusted to 6.5, and
aem.asm.org 925
Yamane et al.

FIG 1 Flowchart of the study. A to D represent experiments.

the volume was brought up to 100 ml using Tc-So medium. Media in
500-ml bottles were anoxidized using Na2S · 9H2O and L-cysteine HCl in
the anaerobic workstation and then inoculated with T. globiformans and
M. jannaschii cells (which had been separately cultured). We monitored T.
globiformans and M. jannaschii populations, as well as CH4 production in
the gas phases of the media. We then established practically optimal cul-
ture conditions as follows: microalgae were cocultured during the first
stage at 68°C for 48 h under an N2-H2-CO2 (80:10:10) gas phase and in the
second phase under H2-CO2 (80:20) at 68°C for 7 days. The gas phases
were replaced at room temperature.
Insoluble materials in the media were collected after the second stage
by centrifugation at 5,000 ⫻ g for 20 min, lyophilized, and stored at 5°C
before being sealed in Pyrex test tubes for pyrolysis.
Flowchart of the study. Figure 1 summarizes possible relationships
between lipid composition and the generation of n-alkane-rich biofuels
from cocultured microalgae in a flowchart.
Gas chromatography. The methods for gas chromatography are de-
scribed in the supplemental material.
while that of the archaeal cells increased to 1.1 ⫻ 108 cells/ml. In
addition, the gas phase comprised almost 60% CH4 after 7 days of
second-stage cultivation under an H2-CO2 (80:20) gas phase. At
the third stage, when the gas phase was replaced with H2-CO2
(80:20), M. jannaschii maintained CH4 production. Both T. globi-
formans and M. jannaschii rapidly grew in a syntrophic fashion
during the first stage (Fig. 2) by utilizing H2 produced via the
bacterial metabolism of organic materials. However, only M. jan-
naschii proliferated during the second stage under a high H2 con-
centration and produced more CH4. The second stage of the co-
culture in basal Tc-So medium resulted in 4 ⫻ 108 cells/ml of T.
globiformans, 3 ⫻ 107 cells/ml of M. jannaschii, and high (35%)

RESULTS
Characterization of T. globiformans and M. jannaschii cocul-
tures to determine optimal culture conditions for media con-
taining microalgae. We assessed the growth of T. globiformans
and M. jannaschii cells and CH4 production in various culture
media to determine the optimal coculture conditions for these
organisms in modified Tc-So medium containing sonically oscil-
lated microalgae. Basal Tc-So medium comprised 0.3% tryptone
and 0.3% yeast extract. We prepared three additional media that
included 0.5, 0.75, and 1.0% (each) tryptone and yeast extract.
Both microorganisms were cultured under an N2-H2-CO2 (80:10:
10) gas phase at 68°C under atmospheric conditions for 48 h dur-
ing the first stage of coculture. The gas phase was then replaced
with H2-CO2 (80:20), and the second stage of culture continued at
68°C for 7 days under atmospheric conditions. Figure 2 indicates FIG 2 Production of CH4 in three gas phases by cocultured T. globiformans
the amounts of CH4 produced in the gas phases. The populations and M. jannaschii in Tc-So medium containing various concentrations of tryp-
of T. globiformans and M. jannaschii reached 2 ⫻ 109 and 7.2 ⫻ 107 tone and yeast extract at 68°C under an N2-H2-CO2 (80:10:10) gas phase in
cells/ml, respectively, and the CH4 content in the N2-H2-CO2 (80: stage I and H2-CO2 (80:20) in stages II and III. 䊐, Basal Tc-So medium con-
taining 0.3% tryptone and 0.3% yeast extract;
, Tc-So medium containing
10:10) gas phase reached 12% after 48 h of the first stage of culture 0.5% tryptone and 0.5% yeast extract; Œ, Tc-So medium containing 0.75%
in medium enriched with 1% tryptone and 1% yeast extract. In tryptone and 0.75% yeast extract; Œ, Tc-So medium containing 1% tryptone
contrast, the bacterial cell density was reduced to 1 ⫻ 109 cells/ml, and 1% yeast extract.

926 aem.asm.org Applied and Environmental Microbiology

 n-Alkane-Rich Biofuels from Microalgae

levels of CH4 release into the gas phase. We also added organic TABLE 1 Main components of fatty alcohols, fatty acids, wax esters,
acids, carbohydrates, and amino acids to the Tc-So medium and alkenes, and alkenones in the total lipids extracted from microalgae
measured CH4 production and the growth of the two microorgan- cocultured with T. globiformans and M. jannaschii
isms. However, none of these compounds exerted any significant Contenta
effect. Lipid component A. platensis D. tertiolecta E. huxleyi E. gracilis
Analysis of cell growth and CH4 production in the coculture
Fatty alcohols
of T. globiformans and M. jannaschii in the presence of microal-
C12 — — — 0.9
gae. The cell populations and CH4 production in medium con- C13 — — — 1.4
taining A. platensis were almost identical to those in Tc-So basal C14 — — — 13.9
medium during the first and second stages. In contrast, CH4 pro- C15 — — — 1.1
duction and cell growth were reduced to 70% of those in Tc-So C16 — — — 3.2
basal medium when media contained D. tertiolecta and E. gracilis
after 7 days of cultivation at the second stage. Growth and CH4 Isoprenoid hydrocarbons 4.3 4.5 2.2 3
production were obviously repressed during the first and second (5 components)
stages in medium containing E. huxleyi. The CH4 concentration in
Fatty acid methyl esters
the gas phase was ⬍0.5%, and the densities of T. globiformans and C12:0 — — — 1.9
M. jannaschii were ⬍5 ⫻ 107 and ⬍1 ⫻ 106 cells/ml, respectively, C13:0 — — — 4.1
at both stages. The growth of T. globiformans and M. jannaschii C14:0 — 0.7 7.7 18.8
reached the stationary phase at 24 h of cultivation in the first stage C16:0 68.3 24.5 5.1 14.9
of Tc-So and in enriched medium and needed 48 h for growth to C16:1 3.8 0.8 0.8 —
the stationary phase in cocultivation with microalgae. These re- C16:2 — 1.6 — —
sults indicated that the microalgae interacted with T. globiformans C16:4 — 11.2 — —
and M. jannaschii to repress their growth and CH4 formation. C18:0 1.3 0.4 0.5 11.4
C18:1 — 4.8 1 10.5
Regardless, such repression enhanced the production of n-al-
C18:2 11.1 2.4 0.2 0.2
kane-rich oils because of a decrease in the number of impurities
C18:3 11.2 46.3 8.5 —
during pyrolysis. Based on these findings, we established opti- C18:4 — 2.8 2.8 —
mal coculture conditions in basal Tc-So basal medium contain- C18:5 — — 6.7 —
ing microalgae. C20:2 — — — 1.5
Lipid composition of microalgae cocultured with T. globifor- C20:3 — — — 2.9
mans and M. jannaschii and n-alkanes produced by pyrolysis. C20:4 — — — 3.9
We analyzed biofuel production from the four microalgae by C22:0 — — 1.6 —
comparing the lipid composition of the cocultured algae with the C22:4 — — — 0.2
content and composition of n-alkanes produced by pyrolysis. Lip- C22:6 — — 7.7 —
ids extracted with chloroform-methanol from the algae were dried
Wax esters
and heated with 5% HCl-methanol at 100°C for 60 min for meth- C24 — — — 0.6
anolysis, and then the lipid composition was analyzed by GC-MS C26 — — — 1.1
(Table 1). The major lipid components of A. platensis and D. ter- C28 — — — 2.4
tiolecta were methyl esters of 16- and 18-carbon-chain fatty acids C30 — — — 0.7
that were derived from glycolipids and phospholipids and small C32 — — — 1.2
amounts of isoprenoids. The lipid contents of the two microalgae C34 — — — 0.2
accounted for 15 to 20% (wt/wt). In addition to these methyl
Long-chain alkenes
esters, considerable amounts of docosahexaenoic acid (DHA)
C31:1 — — 2.5 —
methyl ester, long-chain (31 and 33 carbon atoms) alkenes, and
C31:2 — — 6.9 —
very long-chain (37, 38, and 39 carbon atoms) hydrocarbon alk- C33:3 — — 5.6 —
enones (18) were detected in lipid fractions from the cocultured E.
huxleyi sample. The lipid content in E. huxleyi was 40 to 45% Very long-chain
(wt/wt), and the total alkene and alkenone content reached over alkenones
55% of the total lipid. C37 — — 23.2 —
Many fatty alcohols, wax esters, and fatty acid methyl esters C38 — — 16.1 —
were detected in the lipid fraction of E. gracilis (Table 1), as de- C39 — — 0.9 —
a
scribed by Koritala (25). The lipid content in E. gracilis was 20 to Components were calculated as ratios (%) of each total lipid sample solubilized in
chloroform from peak areas of GC-MS profiles after extracted samples were
25% of the dry weight.
methanolized and analyzed. —, Undetectable or very small peaks. The values are
C16:0 fatty acid methyl ester accounted for 80% of the lipid averages of triplicate experiments.
composition of cocultured T. globiformans and M. jannaschii cells,
and GC-MS showed that the methanolysis products of lipids ex-
tracted from microalgae with and without cocultures of the two
microorganisms essentially did not differ. four cocultured algae. Samples of A. platensis and D. tertiolecta, in
Table 2 shows the main hydrocarbons and their relative which the major lipids comprised fatty acids, both contained
amounts in hexane subfractions separated by Wako Gel C200 col- short-chain n-alkanes (nC13 to nC19). In addition, we detected
umn chromatography after pyrolysis and GC-MS analysis of the large amounts of phytane (iC20) in the A. platensis sample, as well

February 2013 Volume 79 Number 3 aem.asm.org 927

Yamane et al.

TABLE 2 Main hydrocarbons in hexane subfractions from Wako Gel TABLE 3 Comparison of main hydrocarbon and nitrile contents in
C200 column chromatography after pyrolysis of the four cocultured chloroform-soluble fractions from pyrolysis products of cocultured and
algae noncocultured E. gracilis
Contenta Contenta
A. D. E. E. Hydrocarbon or nitrile Cocultured alga Noncocultured alga
Hydrocarbon Abbreviation platensis tertiolecta huxleyi gracilis nC12 8.7 6.4
n-Dodecane nC12 0.5 0.5 2.1 1.8 nC13 16.1 12.2
n-Tridecane nC13 2.2 1.4 4.5 12.5 nC14 24.1 19.7
n-Tetradecane nC14 4.8 3 4.7 22.6 nC15 12.6 9.1
n-Pentadecane nC15 16.5 9.5 5.2 15.3 nC16 14.6 12.4
n-Hexadecane nC16 5.1 3.7 3.1 21.5 Tetradecanenitrile (C14N) 0.3 17.6
n-Heptadecane nC17 31.1 3.1 5.1 12.6 nC17 14.1 3.9
Pristane iC19 3.1 44.1 0.2 0.6 iC19 0.1 0.4
n-Octadecane nC18 2.6 2.3 2.7 1.9 nC18 1.5 0.6
Phytane iC20 30.9 30.1 3.6 7.7 iC20 6.6 3.1
n-Nonadecane nC19 3.2 1.5 6.5 1.3 Hexadecanenitrile (C16N) 0 9.2
n-Eicosane nC20 — 0.8 2.7 0.5 nC19 1.3 0.3
n-Heneicosane nC21 — — 3.9 0.4 Octadecanenitrile (C18N) 0 5.1
n-Docosane nC22 — — 3.4 0.4 a
Components were calculated as ratios (%) of total amounts of chloroform- soluble
n-Tricosane nC23 — — 3.1 0.7 fractions based on peak areas of GC-MS profiles. All values are averages of two
n-Tetracosane nC24 — — 2.1 0.1 experiments.
n-Pentacosane nC25 — — 1.1 0.1
n-Hexacosane nC26 — — 0.8 —
n-Heptacosane nC27 — — 1.4 — tions and amounts of n-alkanes and nitriles detected in the
n-Octacosane nC28 — — 1.9 — GC-MS profiles of chloroform-soluble pyrolysis fractions of E.
n-Nonacosane nC29 — — 0.9 — gracilis samples with and without coculture. All peaks detected by
n-Triacontane nC30 — — 5.5 —
total ion scan chromatography in the cocultured samples coin-
n-Hentriacontane nC31 — — 17.4 —
cided with those of the extracted ion profile of the m/z 71 frag-
n-Dotriacontane nC32 — — 1.8 —
n-Tritriacontane nC33 — — 5.3 — ment, which is one characteristic fragment of n-alkanes. In con-
n-Tetratriacontane nC34 — — 3.9 — trast, many noncoincident peaks were detected in samples
n-Pentatriacontane nC35 — — 7.1 — without coculture. The major noncoincident peaks were detected
a
Hydrocarbon contents were calculated as percentages (%) in total amounts of each as tetradecanenitrile (C14N), hexadecanenitrile (C16N), and octa-
subfraction from peak areas of GC-MS profiles. —, undetectable. The values are decanenitrile (C18N), eluted in the 5% EH subfraction on Wako
averages of at least two experiments. Gel C200. Those alkanenitriles were also detected in the other
three microalgal samples without coculture. Therefore, we specu-
lated that reducing the amounts of impurities, such as nitriles,
as pristane (iC19) and phytane in the D. tertiolecta sample. Short- during pyrolysis enhanced the production of n-alkane-rich oil.
chain n-alkanes (nC13 to nC19) and iC20 were also major compo- Table 4 summarizes the total dry weight (mg) of crude oil
nents of the E. gracilis sample, which contained nC20 to nC25 production in the hexane, 5% EH, 30% EH, and diethyl ether
alkanes as minor components. In contrast, the E. huxleyi sample of subfractions (from Wako Gel C200) of the chloroform-soluble
major lipids, comprising long-chain alkenes and very long-chain fractions and the oil recovered (mg) in the hexane subfraction of
alkenones, contained n-alkanes of various chain lengths (nC13 to the four microalgae (100 mg each) with and without coculture
nC35) and small amounts of pristane and phytane. These n-alkane after pyrolysis. Table 4 also summarizes the ratios (%) of iso-
and isoprenoid profiles were similar to those of native crude oils prenoids, fatty acid methyl esters, fatty alcohols, wax esters, alk-
(4, 26) except for the considerable amount of nC31 alkane. Two enes, and alkenones that were calculated from each GC-MS peak
peaks indicating biomarkers (steranes) were detected in the pro- area of lipid samples. More crude oil was produced by pyrolysis,
files of an extracted ion (m/z 217). However, the ion (m/z 191) for and almost twice as much of the n-alkane-rich biofuels was recov-
hopanes was very small. These results indicated that the n-alkane ered from microalgae that had been cocultured with the two
profiles determined by GC-MS in hexane subfractions from the microorganisms. About 30% of the dry weight of cocultured E.
pyrolysis products of microalgae samples are closely related to the huxleyi was recovered as crude oil, and 8 to 10% was recovered as
lipid composition of the microalgae. Long-chain lipid compo- n-alkane-rich oil in the hexane subfraction after pyrolysis.
nents, in addition to basic fatty acid methyl esters, increased the Characterization of hydrocarbon gases resulting from algal
amounts of generated oils and the appearance of long-chain n-al- pyrolysis. We used gas chromatography to analyze the composi-
kanes in each alga. The final concentrated samples of hexane sub- tion of hydrocarbons in 0.5 ml of the 6 ml of gases resulting from
fractions from A. platensis, D. tertiolecta, and E. gracilis were liq- the pyrolysis of microalgal samples (100 mg each). Figure S1 in the
uid, whereas that from E. huxleyi was solid. supplemental material shows that at least 20 peaks were detected
Effect of coculture with T. globiformans and M. jannaschii in E. huxleyi-cocultured samples. The same peaks were detected in
on biofuel production. To demonstrate the possible effects of all four microalgal samples with and without coculture. Among
cocultures on the production of n-alkane-rich oils, we analyzed these peaks, the contents of methane, CO2, ethylene, ethane, pro-
the n-alkane compositions of pyrolysis products from the four pylene, propane, i-butane, n-butane, and n-pentane were mea-
algae that were not cocultured. Table 3 compares the composi- sured using a calibrated chromatograph (see Table S1 in the sup-

928 aem.asm.org Applied and Environmental Microbiology

 n-Alkane-Rich Biofuels from Microalgae

TABLE 4 Lipid compositions of cocultured algae and comparison of oil production from pyrolyzed microalgae with and without coculture of
T. globiformans and M. jannaschii
Value in microalgaed
A. platensis D. tertiolecta E. huxleyi E. gracilis
Parameter (Cyanobacteria) (Chlorophyta) (Haptophyta) (Euglenophyta)
Lipid composition (%)a
Isoprenoids 4.3 4.5 2.2 3.0
Fatty acid methyl esters 95.7 95.5 42.6 70.3
Fatty alcohols ND ND ND 20.5
Wax esters ND ND ND 6.2
Alkenes ND ND 15.0 ND
Alkenones ND ND 40.2 ND

Crude oil productionb (%) in chloroform fraction 17.5 13.2 29.4 20.4
14.6 8.7 23.2 17.2
Oil recoveryc (%) in hexane subfractions 4.2 2.8 8.9 5.8
1.2 1.5 5.7 3.1
a
Contents of lipid compositions were calculated as ratios (%) of total lipid composition in samples treated with 5% HCl-methanol determined from peak GC-MS areas.
b
Total dry weight of Wako Gel C200 subfractions (hexane, 5% EH, 30% EH, and diethyl ether subfractions) of chloroform fractions obtained from pyrolysis products of algae (100
mg) with and without (boldface) coculture.
c
Dry weight of Wako Gel C200 hexane subfractions of chloroform fractions obtained from pyrolysis products of 100 mg of algal samples with and without (boldface) coculture.
d
All values are averages of at least three experiments, and the standard deviation of each is ⫾10 to 20%. ND, not detected.

plemental material). The CO2 content was very high, and the gases
were rich in methane and ethane. The compositions and ratios of
the hydrocarbon gases produced by pyrolysis were similar among
the algae and with and without cocultivation. The n-hexane con-
tent was very low, and a peak for n-hexane could not be quantified.
Cocultured T. globiformans and M. jannaschii cells pro-
duced essentially no n-alkane-rich biofuel. We prepared two
batches of lyophilized cocultured bacterial and archaeal cells to
determine whether they can produce biofuels. One batch was
cocultured for 2 days in Tc-So medium under a N2-H2-CO2 (80:
10:10) gas phase, followed by 7 days under an H2-CO2 (80:20) gas
phase at 68°C for comparison with cocultured microalgae (9-day
cocultures). The other was cocultured in Tc-So medium enriched
with 1% yeast extract and 1% tryptone at 68°C for 24 h under an
N2-H2-CO2 (80:10:10) gas phase. After pyrolysis at 300°C for 4
days (100 mg per sample), hydrocarbons in the chloroform-solu-
ble fractions and the hexane, 5% EH, 30% EH, and diethyl ether
subfractions from the two samples separated by column chroma-
tography on Wako Gel C200 were analyzed by GC-MS. Figure S2A
to E in the supplemental material shows the results for cells that
were cocultured for 9 days. The major components of the two
chloroform-soluble fractions were tetradecanenitrile (C14N),
pentadecanenitrile (C15N), and hexadecanenitrile (C16N) and a
minimal amount of n-alkanes (nC14, nC15, and nC16) (see Fig. S2A
in the supplemental material). The n-alkanes eluted in the hexane
subfractions (see Fig. S2B in the supplemental material), and ni-
triles eluted in the 5% EH subfraction (see Fig. S2C in the supple-
mental material). Branched alkanes were detected only in the
hexane subfraction (see Fig. S2B in the supplemental material)
from the cells cocultured for 9 days. The 30% EH subfractions (see
Fig. SD in the supplemental material) contained phenol and
naphthalene, as well as their derivatives, which were detected by
analyzing the ion profiles of m/z 94, 108, 122, 128, 142, and 156
extracted from the GC-MS values (see Fig. S3 in the supplemental
material). The dry weight of each subfraction was 0.1, 2.5, 2.5, and
3.4 mg of the 9-day cocultures and 0.2, 6.1, 5.4, and 4.8 mg of the
cells cocultured in enriched medium. The total weight (mg) of the
four subfractions was assumed to be the dry weight of the chloro-
form-soluble fraction, and the total values corresponded to 8.5
and 16.5% of the lyophilized starter cells, respectively. Since the
ratios of the hexane subfractions to the starter cells were only 0.1%
and 0.2% (wt/wt), we assumed that the cocultured cells essentially
did not generate n-alkane-rich biofuels. All values are averages of
triplicate experiments.
DISCUSSION
We tested different alga-bacterium-archaeon consortia to investi-
gate the production of oil-like mixtures, suggesting that they
could potentially be used for the synthesis of n-alkane-rich biofu-
els for a very short period. A model coculture of T. globiformans
and M. jannaschii under anaerobic and atmospheric conditions at
68°C and subsequent pyrolysis at 300°C for 4 days were intro-
duced in the process. The yield of crude and n-alkane-rich oil was
highest when the marine microalga E. huxleyi served as the raw
material. The GC-MS profiles of n-alkanes and isoprenoids in the
pyrolysis products were similar to those of natural crude oils, ex-
cept for the higher nC31 alkane levels. Petroleum crude oils con-
tain many compounds that are undetectable by GC-MS. There-
fore, analysis of hydrocarbons in pyrolysis products is limited.
However, n-alkanes and related hydrocarbons can be quantita-
tively recovered using gas and silica gel column chromatography
(27).
Adding tryptone and yeast extract to cocultures in the enriched
media enhanced the growth of the two microorganisms and CH4
production (Fig. 2), whereas adding sonicated microalgae (in par-
ticular E. huxleyi) repressed or did not affect either cell growth or
CH4 production. Introducing the coculture step increased the
production of crude and n-alkane-rich oils generated from the
microalgae by pyrolysis. We suppose that one reason for the en-
hanced production of biofuels from microalgae is a reduction in
the formation of impurities containing alkanenitriles during
pyrolysis. Alkanenitriles comprised the major pyrolysis products
from cocultured T. globiformans and M. jannaschii cells (see Fig.
S2 in the supplemental material) and from algal samples without

February 2013 Volume 79 Number 3 aem.asm.org 929

Yamane et al.
coculture. However, small amounts of alkanenitriles were de-
tected in the pyrolysis products of cocultured algal samples.
These results suggest that cocultured cells themselves and/or
the culture environment, such as anaerobic conditions, pre-
vented alkanenitrile production and enhanced n-alkane pro-
duction during pyrolysis. However, the actual reasons for the
elevated production remain unresolved.
Phospholipids and glycolipids are major lipids in the cell mem-
brane and photosynthetic apparatus of microalgae, and they are
generated from 14- to 22-carbon chain fatty acids. These lipids
were common and essential to the four microalgae. We consid-
ered that most of the methyl esters in the four microalgae detected
by GC-MS were derived from these 14- to 22-carbon chain fatty
acids by methanolysis. Analysis of A. platensis and D. tertiolecta
lipids by GC-MS indicated mainly methyl esters and small
amounts of isoprenoids. Short-chain (C13 to C19) n-alkanes and
more isoprenoids (iC19 and iC20) were produced from the mi-
croalgae by pyrolysis. The ratios of iC20 to nC18 in A. platensis and
D. tertiolecta were 11.9 and 13.1, respectively. In contrast, fatty
alcohols and wax esters, in addition to these methyl esters, were
detected in E. gracilis. The fatty alcohol and wax ester contents
were still low (20.5% and 6.2%, respectively). Therefore, short-
chain n-alkanes (C13 to C19) were also rich in pyrolyzed E. gracilis.
However, this sample contained some nC20 to nC25 alkanes that
reduced the iC20/nC18 ratio to 4.1. In contrast, more long-chain
(C31 and C33) alkenes (15%) and very long-chain (C37 to C39)
alkenones (40.2%) were detected in E. huxleyi lipids. Concomitant
with these findings, short-chain (nC13) to very long-chain (nC35)
n-alkanes were detected in pyrolysis products of the E. huxleyi
sample, and the iC20/nC18 ratio was also reduced to 1.3. The re-
ported ratio of iC20 to nC18 in natural crude oils is ⬍1.0 (28).
Thus, the GC-MS profile of the hexane subfraction of E. huxleyi
was similar to those of natural crude oils despite the presence of
nC31 alkane.
We found that oil generation from E. huxleyi can be accelerated
by introducing T. globiformans and M. jannaschii cocultures and
pyrolysis and that essentially no n-alkane-rich biofuel was pro-
duced from the cocultured bacterial and archaeal cells themselves
after pyrolysis. We suppose that alkenones of coccolithophorid
algae like E. huxleyi might serve as key raw materials for subsurface
crude oil generation.
ACKNOWLEDGMENTS
This study was supported in part by a grant-in-aid for Science Research
from the Ministry of Education, Culture, Sports, Science and Technology
of Japan.
We are grateful to N. Foster for critical reading of the manuscript.
We declare no conflict of interest.
REFERENCES
1. Hallmann C, Schwark L, Grice K. 2008. Community dynamics of anaer-
obic bacterium in deep petroleum reservoirs. Nat. Geosci. 1:588 –591.
2. Magot M. 2005. Indigenous microbial communities in oil fields, p 21–33.
In Ollivier B, Magot M (ed), Petroleum microbiology. ASM Press, Wash-
ington, DC.
3. Planckaert M. 2005. Oil reservoirs and oil production, p 3–19. In Ollivier
B, Magot M (ed), Petroleum microbiology. ASM Press, Washington, DC.
4. Wang Z, Fingas M, Li K. 1994. Fractionation of a light crude oil and
identification and quantitation of aliphatic, aromatic, and biomarker
compounds by GC-FID and GC-MS, part II. J. Chromatogr. Sci. 32:367–
382.
5. Yamane K, Hattori Y, Ohtagaki H, Fujiwara K. 2011. Microbial diversity
930 aem.asm.org
with dominance of 16S rRNA gene sequences with high GC content at 74
and 98oC subsurface crude oil deposits in Japan. FEMS Microbiol. Ecol.
76:220 –235.
6. Hattori S, Kamagata Y, Hanada S, Shoun H. 2000. Thermacetogenium
phaeum gen. nov., sp. nov., a strictly anaerobic thermophilic, syntrophic
acetate-oxidizing bacterium. Int. J. Syst. Evol. Microbiol. 50:1601–1609.
7. Kuwabara T, Kawasaki A, Uda I, Sugai A. 2011. Thermosipho globifor-
mans sp. nov., an anaerobic thermophilic bacterium that transforms into
multicellular spheroids with a defect in peptidoglycan formation. Int. J.
Syst. Evol. Microbiol. 61:1622–1627.
8. Jones WJ, Leigh JA, Mayer F, Woese CR, Wolfe RC. 1983. Methano-
coccus jannaschii sp. nov. an extremely thermophilic methanogen from a
submarine hydrothermal vent. Arch. Microbiol. 136:254 –261.
9. Whitman WB. 2001. Genus I. Methanocaldococcus gen. nov., p 243–245.
In Boone DR, Castenholz RW, Garrity GM (ed), Bergey’s manual of sys-
tematic bacteriology, 2nd ed, vol 1. Springer, New York, NY.
10. Johnson MR, Conners SB, Montero CI, Chou CT, Shockley KR, Kelly
RM. 2006. The Thermotoga maritima phenotype is impacted by syn-
trophic interaction with Methanococcus jannaschii in hyperthermophilic
coculture. Appl. Environ. Microbiol. 72:811– 818.
11. Muralidharan V, Rinker KD, Hirsh IS, Bouwer EJ, Kelly RM. 1997.
Hydrogen transfer between methanogens and fermentative heterotrophs
in hyper thermophilic cocultures. Biotechnol. Bioeng. 56:268 –278.
12. Ishii S, Kosaka T, Hori K, Hotta Y, Watanabe K. 2005. Coaggregation
facilitates interspecies hydrogen transfer between Pelotomaculum thermo-
propionicum and Methanothermobacter thermautotrophicus. Appl. Envi-
ron. Microbiol. 71:7838 –7845.
13. Ringer M, Putsche V, Scahill J. 2006. Large-scale pyrolysis oil production: a
technology assessment and economic analysis. Technical Report NREL/TP-510-
3779. National Renewable Energy Laboratory, Golden, CO.
14. Chisti Y. 2007. Biodiesel from microalgae. Biotechnol. Adv. 25:294 –306.
15. Gouveia L, Oliveira AC. 2009. Microalgae as a raw material for biofuels
production. J. Ind. Microbiol. Biotechnol. 36:269 –274.
16. Minowa T, Yokoyama S, Kishimoto M, Okakura T. 1995. Oil produc-
tion from algal cells of Dunaliella tertiolecta by direct thermochemical
liquefaction. Fuel 74:1735–1738.
17. Wu Q, Shiraiwa Y, Takeda H, Sheng G, Fu J. 1999. Liquid-saturated
hydrocarbons resulting from pyrolysis of marine coccolithophores Emili-
ania huxleyi and Gephyrocapsa oceanica. Mar. Biotechnol. 1:346 –352.
18. Volkmann JK, Eglinton G, Corner EDS, Forsberg TEV. 1980. Long-
chain alkenes and alkenones in the marine coccolithophorid Emiliania
huxleyi. Phytochemistry 19:2619 –2622.
19. Buitenhuis ET, van der Wal P, de Baar HJW. 2001. Blooms of Emiliania
huxleyi are sinks of atmospheric carbon dioxide: a field and mesocosm
study derived simulation. Global Biogeochem. Cycles 15:577–587.
20. Riebesell U, Zondervan I, Rost B, Tortell PD, Zeebe RE, Morel FM.
2000. Reduced calcification of marine plankton in response to increased
atmospheric CO2. Nature 407:364 –367.
21. Brassell SC, Eglinton G, Marlowe IT, Pflaumann U, Sarnthein M. 1986.
Molecular stratigraphy: a new tool for climatic assessment. Nature 320:
129 –133.
22. Balch WE, Fox GE, Mogrum LJ, Woese CR, Wolfe RS. 1979. Metha-
nogenes: reevaluation of a unique biological group. Microbiol. Rev. 43:
260 –269.
23. Danbara A, Shiraiwa Y. 1999. The requirement of selenium for the
growth of marine coccolithophorids, Emiliania huxleyi, Gephyrocapsa oce-
anica and Halladosphaera sp. (Prymnesiophyceae). Plant Cell Physiol. 40:
762–766.
24. Ames GF. 1968. Lipids of Salmonella typhimurium and Escherichia coli:
structure and metabolism. J. Bacteriol. 95:836 – 843.
25. Koritala S. 1989. Microbiological synthesis of wax esters by Euglena gra-
cilis. J. Am. Oil Chem. Soc. 66:133–134.
26. Yamane K, Maki H, Nakayama T, Nakajima T, Nomura N, Uchiyama
H, Kitaoka M. 2008. Diversity and similarity of microbial communities in
petroleum crude oils produced in Asia. Biosci. Biotechnol. Biochem. 72:
2831–2839.
27. Wang Z, Fingas M, Li K. 1994. Fractionation of a light crude oil and identi-
fication and quantitation of aliphatic, aromatic, and biomarker compounds
by GC-FID and GC-MS, part I. J. Chromatogr. Sci. 32:361–366.
28. Peters KE, Moldowan JM. 1993. The biomarker guide. Prentice Hall,
New York, NY.
Applied and Environmental Microbiology