Embed Size (px)
Citation preview
Organic Geochemistry 38 (2007) 582–608
www.elsevier.com/locate/orggeochem
OrganicGeochemistry
Hydrogen isotope fractionation in freshwater algae:I. Variations among lipids and species
Zhaohui Zhang *, Julian P. Sachs 1
Department of Earth, Atmospheric and Planetary Sciences, Building E34-166, Massachusetts Institute of Technology,
77 Massachusetts Avenue, Cambridge, MA 02139, USA
Received 16 May 2006; received in revised form 19 November 2006; accepted 3 December 2006Available online 21 December 2006
Abstract
Five species of freshwater green algae, including three strains of Botryococcus braunii (two A Race, one B Race), Eudo-
rina unicocca and Volvox aureus, were cultured under controlled conditions in media containing different concentrations ofdeuterium. The hydrogen isotopic ratios of lipids in the algae, including alkadienes, botryococcenes, heptadecenes, fattyacids, and phytadiene, were measured by gas chromatography–isotope ratio-mass spectrometry (GC–IRMS) and foundto closely track water dD values. While correlation coefficients (R2) in excess of 0.99 for all lipids in all species suggest thatlipid dD values can be used to determine water dD values, hydrogen isotope fractionation was found to vary systematicallybetween lipids and lipid homologues within a single alga, as well as for the same lipid between species of algae. Undersimilar growth conditions, two species of Chlorophyceae (Eudorina unicocca and Volvox aureus) and three species of Tre-bouxiophyceae (Botryococcus braunii) produced palmitic acid (C16 fatty acid) that differed by 90–100& relative to water.Ubiquitous lipids such as palmitic acid, with a multitude of aquatic and terrestrial sources, are therefore not good targetsfor D/H-based paleohydrologic reconstructions. In addition to the use of source-specific biomarkers that derive unambig-uously from a single family or species, paleohydrologic applications of lipid D/H ratios will need to consider the as yetunstudied potential influence that environmental parameters such as nutrients, light and temperature, etc., may have onD/H fractionation during lipid synthesis.� 2006 Elsevier Ltd. All rights reserved.
1. Introduction
Hydrologic variations are difficult to reconstructfrom the geologic record and are poorly reproduced
0146-6380/$ - see front matter � 2006 Elsevier Ltd. All rights reserveddoi:10.1016/j.orggeochem.2006.12.004
* Corresponding author. Present address: Department of Geo-sciences, University of Massachusetts, 611 North Pleasant Street,Amherst, MA 01003, USA. Tel.: +1 413 545 1755; fax: +1 413545 1200.
E-mail address: [email protected] (Z. Zhang).1 Present address: University of Washington, School of Ocean-
ography, Box 355351, Seattle, WA 98195, USA.
by climate models, yet they are essential for under-standing the natural variation of climate, especiallyin the tropics. Hydrogen and oxygen isotope ratiosof lake and ocean water reveal hydrologic variationscaused by the higher vapor pressure of HHO andH2
16O relative to HDO and H218O. However, these
ratios are often not preserved in the geologic recorddue to isotopic exchange or diagenesis or, as withmineral phases, are influenced by non-hydrologicprocesses. The ‘holy grail’ of paleohydrologic recon-structions is therefore a robust recorder of water
.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 583
isotopic variations that does not undergo alterationon the timescale of study.
The hydrogen isotopic composition of algal lipidbiomarkers in lake sediments holds great potentialfor water isotopic reconstruction. All hydrogen inalgae derives from water and algae do not transpire,a process that overprints dD values in land plants.Whereas hydrogen atoms covalently bound toorganic nitrogen, sulfur, and oxygen are labile andexchange quickly and reversibly with hydrogenatoms in water (Werstiuk and Ju, 1989), hydrogenatoms in lipids are bound to carbon (Estep andHoering, 1980) and are nonexchangable (Epsteinet al., 1976; Sauer et al., 2001). Compared to bulkorganic matter, individual lipid biomarkers are notsubject to isotopic changes caused by preferentialdegradation of less stable compounds.
Despite the strong potential of algal lipids inD/H reconstructions technical challenges in contin-uous flow-isotope ratio mass spectrometry havelimited their application. With these technical prob-lems all but overcome in recent years (cf. Burgoyneand Hayes, 1998; Hilkert et al., 1999; Sessions et al.,1999) the next step in developing compound-specifichydrogen isotopic reconstructions of water D/Hratios is demonstrating that lipid biomarker dD
values closely track water dD values. Several recentstudies have addressed this issue (Sessions et al.,1999; Sessions, 2006; Sauer et al., 2001; Huanget al., 2002, 2004; Sachse et al., 2004; Chikaraishiet al., 2004a,b,c, 2005; Chikaraishi, 2006; Englebr-echt and Sachs, 2005; Schouten et al., 2006). How-ever, this is the first to systematically evaluate thefidelity with which dD values in a variety of lipidbiomarkers from cultured freshwater microalgaerecord water dD values.
Estep and Hoering (1980, 1981) pioneered hydro-gen isotopic biogeochemistry by measuring the dD
values of total organic matter in axenic algal cul-tures that were �90& to �110& relative to thewater. They concluded that (1) photosynthesis wasthe primary process causing hydrogen isotopic frac-tionation in plants, (2) some additional fraction-ation occurred in the dark reactions during thesynthesis of malic or pyruvic acids and lipids, and(3) little hydrogen isotope fractionation occurredduring respiration.
Englebrecht and Sachs (2005) cultured the marinemicroalga Emiliania huxleyi and demonstrated thatalkenones closely tracked water dD values(R2 > 0.999). Huang et al. (2004) showed that dD val-ues of palmitic acid (C16:0 fatty acid) in core-top sed-
iments from lakes in eastern North America werewell-correlated with both water dD values of the lakesand air temperature (R2 = 0.89), and Sachse et al.(2004) showed that n-alkanes in core top sedimentsfrom lakes in a meridional transect through centralEurope co-varied with dD values of precipitation.
To better constrain the extent to which algal lip-ids record water dD values and to address the spe-cies-dependence of hydrogen isotope fractionationin lipids we cultured five species of freshwater greenalgae – Eudorina unicocca, Volvox aureus (bothbelong to Chlorophyceae) and three strains of Bot-ryococcus braunii (Trebouxiophyceae) – under con-trolled conditions in the laboratory, with eachstrain grown in five media containing differentenrichments of deuterium. We measured dD valuesof several lipids from each culture, including alkadi-enes, botryococcenes, heptadecene, fatty acidmethyl esters (FAMEs), phytadiene and free fattyacids. Reported here are the results of hydrogen iso-topic fractionation in those lipids from the fivefreshwater algal species.
2. Methods
2.1. Algal cultures
2.1.1. Algal speciesFive species of freshwater green algae were cul-
tured: Eudorina unicocca, Volvox aureus and threestrains of Botryococcus braunii. Relative to redand brown algae, green algae are closer in originto land plants (Kirk, 1998).
Botryococcus braunii is a colonial member of Tre-bouxiophyceae (Senousy et al., 2004), characterizedby high production of lipids and widely distributedin freshwater lakes and ponds. Interest in Botryo-
coccus braunii arose in the 1960s when the algaewere discovered to synthesize large amounts ofhydrocarbons (cf. Maxwell et al., 1968; Douglaset al., 1969). Axenic inoculants of three strains ofBotryococcus braunii were provided by Dr. PierreMetzger of the Laboratoire de Chimie Bioorganiqueet Organique Physique in France. Two were fromthe A race, one from Morocco (a small pool inOukaidem, Atlas) and the other from Lake Titicaca(Bolivia). The third was from the B race, from Mar-tinique (French West Indies).
Eudorina unicocca and Volvox aureus both belongto Chlorophyta (green algae), chlorophyceae, volvo-cales and they are the most common single-celledgreen algae in natural water bodies. Each possesses
584 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
biflagellate cells held together in a coenobium ofdefined shape, usually spherical (Kirk, 1998). Theinoculants of Eudorina unicocca and Volvox aureus
were supplied by the Culture Collection of Algaeand Protozoa (CCAP) in Cumbria, United King-dom (now located in Dunstaffnage Marine Labora-tory, Oban, Scotland). The two cultured strainswere Eudorina unicocca G.M. Smith 1930 (CCAP24/1C, originated from freshwater near Blooming-ton, Indiana, USA) and Volvox aureus Ehrenberg1838 (CCAP 88/6, originated from freshwater inMalham Tarn, Yorkshire, England).
We conducted the culture experiments in the lab-oratory of Dr. Daniel Repeta at the Woods HoleOceanographic Institution from July to November,2003.
2.1.2. Preparation of deuterium-enriched water and
the determination of water dD values
The reference standard for D/H ratios in thiswork is Vienna Standard Mean Ocean Water(VSMOW) with D/H = 155.76 ± 0.05 · 10�6. dD isdefined as
dD¼ ½ðD=HÞsample�ðD=HÞVSMOW�=ðD=HÞVSMOW�1000‰
In order to investigate the relationship between lipiddD and water dD over a wide range of values, we setup five cultures for each species in waters with differ-ent deuterium concentrations. We mixed varyingamounts of Aldrich deuterium oxide (99.9% D) withdistilled water (dD = �65&) to produce mediumwith five different dD values, ranging from �65&
to +450&.Water dD values were measured on an H-device-
Thermo Finnigan Deltaplus XL mass spectrometerat Dartmouth College. The H-Device consists of aquartz reactor filled with chromium powder and aseries of valves. The quartz reactor was sealed atone end by a septum and at the other end by a pneu-matic valve. The reactor was held at 850 �C and theair evacuated by the vacuum system in the massspectrometer. 1 ll of water was injected throughthe septum and flash evaporated. The water wasreacted for one minute and the resulting gasesadmitted to the inlet system of the mass spectrome-ter. Precision of the water dD analyses was 0.5&.
2.1.3. Culture media
The medium for the culture experiments wasautoclaved before algal inoculants were introduced.Glassware was acid-leached overnight and rinsed
with deionized water, then autoclaved. Algal trans-fers were conducted in a laminar flow bench withsterilized labware.
B. braunii was grown on a modified CHU 13medium (composition in mg/l): KNO3, 200;K2HPO4Æ3 H2O, 54; MgSO4 Æ 7H2O, 100; CaCl2 Æ2H2O, 52; FeNaEDTA, 10; plus 5 ml of a solutionof micronutrients (mg/l): H3BO3, 286; MnSO4 ÆH2O, 154; ZnSO4 Æ 7H2O, 22; CuSO4 Æ 5H2O, 8;Na2MoO4 Æ 2H2O, 6; CoSO4 Æ 7H2O, 9. The pHwas approximately 7.4.
Cultures of E. unicocca and V. aureus were grownin Jaworski’s Medium (JM). The stock compositionin g/200 ml was: (1) Ca(NO3)2 Æ 4H2O, 4.0; (2)K2HPO4, 2.48; (3) MgSO4 Æ 7H2O, 10.0; (4)NaHCO3, 3.18; (5) FeNaEDTA, 0.45; Na2EDTA,0.45; (6) H3BO3, 0.496; MnCl2 Æ 4H2O, 0.278 g;(NH4)4Mo7O24 Æ 4H2O, 0.20 g; (7) Cyanocobala-min, 0.008 g; Thiamine HCl, 0.008 g; Biotin,0.008 g; (8) NaNO3, 16.0 g; (9) Na2HPO4 Æ 12H2O,7.2 g. Stock solutions 1–9 were diluted to 1 l withdeionized water.
2.1.4. Light intensity and temperature control
Continuous light was provided by two banks oftwo 25 W cool-white fluorescent tubes. Light inten-sity was measured with a Biospherical InstrumentsQSL-100 probe equipped with a QSP-170 sensor.The center of the chamber was �20 · 1015 quanta/scm2, or 320 lmol/m2 s. Illumination was on a12 h light:12 h dark photo cycle.
The 15 cultures of B. braunii (A1W1-5, A2W1-5and BBW1-5; A1: A race, Morocco strain; A2, Arace, Titicaca strain; BB: B race, Martinique strain)were grown at summer room temperature (July–September, 2003), which averaged 28 �C but variedfrom 26.5 �C to 30 �C.
E. unicocca and V. aureus were grown at 15 �C in aclosed chamber built from foam insulation andcooled with an air-conditioner (Haier, 5000 BTU).The temperature inside the chamber varied between14.5–16 �C during the course of the experiment(EUW1-5, VVW1-5; EU: E. unicocca; VV: V. aureus).
2.1.5. Aeration
To expedite algal growth the culture medium wasbubbled with 1% CO2 in purified air that had beenpassed through two 0.2 lm membrane filters. Metz-ger et al. (1985a) reported that hydrocarbon levelsof Botryococcus brauni were greatly affected by cul-ture conditions; air augmented with 1% CO2
increased the hydrocarbon contents from 5% in
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 585
unaerated cultures to 20–61% in aerated cultures,depending on the strain.
All plastic tubing and filters were autoclavedbefore use. Air was introduced into the mediumnear the bottom of the flask so that bubbling keptthe algae in suspension. The flow rate was�250 ml/min. 1% CO2 air was made by mixing N2
gas and gas from tanks of 95% O2 and 5% CO2 ina cell filled with glass beads.
Evaporation during the course of the experi-ments caused deuterium enrichment in the mediumthat we evaluated with a control experiment. A cul-ture flask with 1.5 l water (dD = �67.7%) was aer-ated with the algal cultures and sampled every fivedays. After 45 days the water was enriched in deute-rium by 45&. Aliquots of medium were thereforecollected from every culture flask throughout thecourse of the experiments for D/H analysis. dDwater
values at the start and end of each culture experi-ment are shown in Tables 1–3.
2.1.6. Cell density monitoring and harvest
Cell density was monitored daily by measuring asmall aliquot of each culture on a HP8452A DiodeArray Spectrophotometer. Because the aliquotswere not returned to the culture flasks (in order toprevent contamination) they typically constitutedthe largest source of water loss during the experi-ments. Growth curves for each strain of B. braunii,
and for E. unicocca and V. aureus were obtained bymeasuring the absorption of light at 400, 500, 600,and 700 nm. Absorption curves at different wave-lengths are nearly identical. Representative absorp-tion curves for 600 nm light are shown in SFig. 1(supplementary material). The continuous increasein light absorption (and by extension, cell density)with time was partly caused by evaporative loss ofwater resulting from aeration. B. braunii, and E.
unicocca/V. aureus cultures were harvested after 45days and 31 days, respectively, during the exponen-tial-phase on the growth curves.
Cultures were harvested by helium-pressurizedfiltration through 293 mm diameter Whatman GF/F filters (pore size 0.7 lm), and subsequently keptat �20 �C until analysis. A small aliquot of filtratewas filtered through a 0.2 lm membrane filter toobtain water for dD analysis.
2.2. Lipid extraction and isolation
Harvested algae on GF/F filters were freeze-dried,cut into 0.5 · 0.5 cm pieces, and extracted on a Dio-
nex ASE-200 pressurized fluid extractor with dichlo-romethane (DCM) and methanol (MeOH) (9:1) at1500 psi and 150 �C. The total lipid extract was frac-tionated on an aminopropyl cartridge-style SPE col-umn (Burdick & Jackson, size 500 mg/4 ml) withDCM/isopropyl alcohol (IPA) (3:1). Retained fattyacids were recovered with 4% acetic acid in diethylether, methylated with 10% BF3 in MeOH and puri-fied by urea adduction. The DCM/IPA fraction wasfractionated by column chromatography using 5%water-deactivated silica gel in a 29 cm · 1.2 cm glasscolumn. Hydrocarbons were eluted with hexane,FAMEs and phytadienes with 10% ethyl acetate(EtOAc) in hexane, and alcohols (including phytoland sterols) with MeOH. FAMES and phytadieneswere purified further by urea adduction.
Each fraction was analyzed by gas chromatogra-phy–mass spectrometry (GC–MS) to positivelyidentify each lipid and determine its purity, thenby gas chromatography with flame-ionization detec-tion (GC-FID) to determine individual lipid concen-trations. Agilent 6890 gas chromatographs wereoperated with programmable temperature vaporiza-tion (PTV) inlets, 60 m Varian Chrompac CP-Sil 5capillary columns with 0.32 mm i.d. and 0.25 lmfilm thickness, and helium carrier gas. Oven temper-ature programs differed for each lipid class.
Alkadienes in the Titicaca strain and botryococc-enes in the Martinique strain in B. braunii cultureswere hydrogenated with Raney Nickel in order toelucidate their exact structures.
2.3. Determination of molecular D/H ratios
The hydrogen isotopic composition of each lipidfraction was measured by gas chromatography–iso-tope ratio-mass spectrometry (GC–IRMS) with aFinnigan Deltaplus XP mass spectrometer, equippedwith a Trace GC and a Combustion III interface.
The Trace GC was equipped with a PTV inletoperated in splitless mode, a 30 m DB-5 capillarycolumn (J&W Scientific) with 0.25 mm i.d. and0.25 lm film, and was operated at a constant heliumflow rate of 1 ml/min. The oven temperature pro-gram differed for each lipid class. For hydrocarbons,the starting temperature was 90 �C, rising to 230 �Cat 13 �C/min, then to 325 �C at 5.5 �C/min, followedby 13 min at 325 �C. Effluent from the GC enteredthe GC-C III interface, a graphite-lined ceramic tubeat 1400 �C, where quantitative pyrolysis to graphite,hydrogen gas and carbon monoxide occurred (Bur-goyne and Hayes, 1998). The hydrogen gas stream
Tab
le1
Hyd
roge
nis
oto
pe
rati
os
and
D/H
frac
tio
nat
ion
inh
ydro
carb
on
sfr
om
thre
est
rain
so
fB
.bra
unii
Cu
ltu
res
H2O
dDH
2O
dDC
27
alk
adie
ne
C29
alk
adie
ne
C29
alk
atri
ene
C31
alk
adie
ne
star
th
arve
stdD
ra
edD
ra
edD
ra
edD
ra
e
Botr
yoco
ccus
bra
uni,
Ara
ce,
Moro
cco
stra
in
A1W
1�
64.5
�29
.5�
184.
93.
70.
840
�16
0.1
�23
5.7
1.0
0.78
8�
212.
4�
232.
63.
20.
791
�20
9.2
�22
5.5
1.9
0.79
8�
201.
9
A1W
288
.411
7.1
�84
.30.
90.
820
�18
0.2
�12
2.3
1.4
0.78
6�
214.
3�
123.
50.
30.
785
�21
5.4
�11
6.2
2.8
0.79
1�
208.
9
A1W
318
7.7
220.
8�
35.8
2.1
0.79
0�
210.
2�
58.1
3.4
0.77
2�
228.
5�
64.1
3.1
0.76
7�
233.
3�
51.5
1.3
0.77
7�
223.
1
A1W
428
5.6
316.
932
.62.
40.
784
�21
5.9
18.6
2.2
0.77
3�
226.
510
.22.
90.
767
�23
2.9
23.4
3.1
0.77
7�
222.
8
A1W
547
2.2
497.
216
8.6
2.9
0.78
0�
219.
516
2.5
1.1
0.77
6�
223.
515
1.9
1.9
0.76
9�
230.
616
9.4
3.2
0.78
1�
219.
0
Ave
rage
0.80
3�
197.
20.
779
�22
1.0
0.77
6�
224.
30.
785
�21
5.1
Sta
nd
ard
dev
iati
on
0.02
625
.90.
007
7.3
0.01
111
.20.
009
9.4
Cu
ltu
res
H2O
dDH
2O
dDC
27
alk
adie
ne
C29
alk
adie
ne
C31
alk
adie
ne
star
th
arve
stdD
ra
edD
ra
edD
ra
e
Botr
yoco
ccus
bra
uni,
Ara
ce,
Tit
icaca
stra
in
A2W
1�
64.0
�26
.0�
224.
01.
40.
797
�20
3.3
�24
4.9
2.2
0.77
5�
224.
8�
243.
71.
90.
776
�22
3.5
A2W
287
.612
8.6
�12
2.9
0.9
0.77
7�
222.
9�
120.
41.
00.
779
�22
0.6
�12
2.6
0.8
0.77
7�
222.
6
A2W
318
2.1
221.
3�
55.0
2.2
0.77
4�
226.
2�
57.6
2.8
0.77
2�
228.
4�
56.2
2.9
0.77
3�
227.
2
A2W
428
0.7
321.
120
.44.
90.
772
�22
7.6
7.0
3.3
0.76
2�
237.
77.
33.
40.
762
�23
7.5
A2W
545
8.0
499.
713
5.7
4.9
0.75
7�
242.
713
0.8
5.2
0.75
4�
246.
013
1.8
6.8
0.75
5�
245.
3
Ave
rage
0.77
5�
224.
50.
769
�23
1.5
0.76
9�
231.
2
Sta
nd
ard
dev
iati
on
0.01
414
.10.
010
10.3
0.01
09.
8
Cu
ltu
res
H2O
dDH
2O
dDC
30
bo
tryo
cocc
ene
C31
bo
tryo
cocc
ene
C34
bo
tryo
cocc
ene(
4)C
34
bo
tryo
cocc
ene
iso
(5)+
?aC
32
+C
33
bo
tryo
cocc
enea
C34
bo
tryo
cocc
ene
iso
?b
star
th
arve
stdD
ra
edD
ra
edD
ra
edD
ra
edD
ra
edD
ra
e
Botr
yoco
ccus
bra
unii
,B
race
,M
art
iniq
ue
BB
W1
�64
.0�
18.9
�28
8.7
1.4
0.72
5�
275.
0�
310.
33.
90.
703
�29
7.0
�33
4.4
4.1
0.67
8�
321.
5�
325.
54.
60.
687
�31
2.5
�33
1.1
3.5
0.68
2�
318.
2�
373.
41.
50.
639
�36
1.3
BB
W2
102.
216
1.1
�17
2.9
2.4
0.71
2�
287.
6�
184.
81.
20.
702
�29
7.9
�22
4.3
0.9
0.66
8�
331.
9�
206.
01.
10.
684
�31
6.2
�21
3.3
1.6
0.67
8�
322.
4�
258.
73.
10.
638
�36
1.6
BB
W3
184.
123
0.2
�10
2.1
2.0
0.73
0�
270.
1�
125.
82.
10.
711
�28
9.4
�16
2.7
1.2
0.68
1�
319.
4�
148.
72.
80.
692
�30
8.0
�15
1.3
2.9
0.69
0�
310.
1�
205.
82.
20.
646
�35
4.4
BB
W4
273.
732
7.6
�16
.33.
80.
741
�25
9.0
�56
.75.
70.
711
�28
9.5
�91
.43.
20.
684
�31
5.6
�81
.46.
60.
692
�30
8.0
�85
.93.
70.
689
�31
1.5
�14
7.0
4.1
0.64
3�
357.
4
BB
W5
455.
151
3.6
101.
42.
80.
728
�27
2.3
76.8
6.1
0.71
1�
288.
648
.11.
70.
692
�30
7.5
58.4
2.1
0.69
9�
300.
737
.84.
20.
686
�31
4.3
�29
.62.
20.
641
�35
8.8
Ave
rage
0.72
7�
272.
80.
708
�29
2.5
0.68
1�
319.
20.
691
�30
9.1
0.68
5�
315.
30.
641
�35
8.7
Sta
nd
ard
dev
iati
on
0.01
010
.30.
005
4.6
0.00
98.
90.
006
5.8
0.00
55.
10.
003
3.0
Eac
hst
rain
was
gro
wn
infi
vem
edia
con
tain
ing
diff
eren
tco
nce
ntr
ati
on
so
fd
eute
riu
m.
Th
edD
valu
eo
fea
chli
pid
was
mea
sure
din
trip
lica
tean
dth
eav
erag
ean
dst
and
ard
dev
iati
on
are
rep
ort
ed.
Th
edD
valu
eo
fth
ecu
ltu
rem
ediu
mat
the
star
tan
dh
arve
star
eal
sosh
ow
n.
iso
,is
om
er;r
,st
and
ard
dev
iati
on
.
Fra
tio
nat
ion
fact
or,
a,is
defi
ned
as(D
/H) l
ipid
/(D
/H) w
ate
r=
(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
).
Fra
ctio
na
tio
n,e,
isd
efin
edas
(a�
1)*1
000
=[(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
)�
1]*
1000
.a
Th
ep
eak
so
fC
34
bo
tryo
cocc
ene
(co
mp
ou
nd
5),
C32
bo
tryo
cocc
ene
(co
mp
ou
nd
6)
and
C33
bo
tryo
cocc
ene
(co
mp
ou
nd
7)
coel
ute
into
two
pea
ks
and
itis
har
dto
dis
tin
guis
ham
on
gth
em.
Th
ep
eak
of
C34
bo
tryo
cocc
ene
iso
mer
(5)
+?
isli
kel
yth
eco
elu
tio
no
fC
34
iso
mer
(co
mp
ou
nd
5)
and
the
un
iden
tifi
edp
eak
bef
ore
it.
Th
eas
sign
edC
32
+C
33
bo
tryo
cocc
ene
cou
ldev
enin
clu
de
the
smal
lp
eak
of
C34
bo
tryo
cocc
ene
iso
mer
wh
ich
elu
tes
afte
rth
em.
bT
his
pea
kco
uld
be
ano
ther
C34
bo
tryo
cocc
ene
or
ano
ther
ho
mo
logu
ew
ith
mo
reca
rbo
nat
om
s.
586 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
Tab
le2
Hyd
roge
nis
oto
pe
rati
os
and
D/H
frac
tio
nat
ion
infr
eefa
tty
acid
s,n
atu
ral
FA
ME
san
dp
hyt
adie
nes
inth
ree
stra
ins
of
B.b
raunii
Cu
ltu
res
H2O
dD
atst
art
H2O
dD
ath
arve
st
C16
fatt
yac
idC
18:1
fatt
yac
idC
20:1
fatt
yac
idC
26:1
fatt
yac
idC
28:1
fatt
yac
idC
30:1
fatt
yac
id
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
Botr
yoco
ccus
bra
unii
,B
race
,M
art
iniq
ue
stra
in
BB
W1
�64
.0�
18.9
�19
2.1
3.4
0.82
3�
176.
6�
191.
63.
10.
824
�17
6.0
�17
7.9
0.3
0.83
8�
162.
1�
187.
91.
70.
828
�17
2.3
�18
9.4
3.5
0.82
6�
173.
8�
191.
00.
90.
825
�17
5.4
BB
W2
102.
216
1.1
�37
.80.
50.
829
�17
1.3
�31
.10.
20.
835
�16
5.5
�11
.04.
00.
852
�14
8.2
�33
.04.
00.
833
�16
7.1
�30
.91.
50.
835
�16
5.4
�27
.12.
40.
838
�16
2.0
BB
W3
184.
123
0.2
15.9
2.0
0.82
6�
174.
219
.43.
50.
829
�17
1.3
33.2
5.8
0.84
0�
160.
212
.66.
30.
823
�17
6.9
20.4
2.2
0.82
9�
170.
5N
DN
DN
DN
D
BB
W4
273.
732
7.6
95.4
3.6
0.82
5�
174.
986
.84.
90.
819
�18
1.3
115.
31.
30.
840
�15
9.9
ND
ND
ND
ND
96.8
1.2
0.82
6�
173.
8N
DN
DN
DN
D
BB
W5
455.
151
3.6
228.
80.
70.
812
�18
8.1
233.
73.
50.
815
�18
4.9
256.
21.
10.
830
�17
0.0
ND
ND
ND
ND
242.
73.
00.
821
�17
9.0
ND
ND
ND
ND
Ave
rag
e0.
823
�17
7.0
0.82
4�
175.
80.
840
�16
0.1
0.82
8�
172.
10.
828
�17
2.5
0.83
1�
168.
7
Sta
nd
ard
dev
iati
on
0.00
76.
50.
008
7.8
0.00
87.
80.
005
4.9
0.00
55.
00.
009
9.5
Botr
yoco
ccus
bra
unii
,A
race
,T
itic
aca
stra
in
A2W
1�
64.0
�26
.0�
206.
51.
30.
815
�18
5.4
�20
5.6
1.9
0.81
6�
184.
4�
175.
94.
40.
846
�15
3.9
�16
7.2
5.8
0.85
5�
145.
0�
184.
24.
90.
838
�16
2.4
�17
5.6
0.2
0.84
6�
153.
6
A2W
287
.612
8.6
�85
.10.
50.
811
�18
9.4
�79
.21.
00.
819
�18
1.4
�61
.92.
00.
831
�16
8.8
�57
.34.
30.
835
�16
4.8
�62
.71.
10.
831
�16
9.5
�59
.11.
20.
834
�16
6.3
A2W
318
2.1
221.
3�
25.3
1.5
0.79
8�
201.
9�
6.6
2.5
0.81
3�
186.
6N
DN
DN
DN
D�
10.2
6.0
0.81
0�
189.
6�
2.2
1.1
0.81
7�
183.
06.
03.
30.
824
�17
6.3
A2W
428
0.7
321.
185
.44.
70.
822
�17
8.4
72.5
2.1
0.81
2�
188.
197
395.
50.
831
�16
8.9
ND
ND
ND
ND
95.7
2.1
0.89
�17
0.6
109.
15.
00.
840
�16
0.5
A2W
548
5.0
499.
718
9.4
5.5
0.79
3�
206.
920
6.8
5.9
0.80
5�
195.
322
8.4
3.0
0.81
9�
180.
9N
DN
DN
DN
D21
8.7
4.0
0.81
3�
187.
422
9.0
2.8
0.82
0�
180.
5
Ave
rag
e0.
808
�19
2.4
0.81
3�
187.
20.
832
�16
8.1
0.83
4�
166.
40.
825
�17
4.6
0.83
3�
167.
4
Sta
nd
ard
dev
iati
on
0.01
211
.80.
005
5.2
0.01
111
.10.
022
22.3
0.01
010
.30.
011
11.1
Botr
yoco
ccus
bra
unii
,A
race
,M
oro
cco
stra
in
A1W
1�
64.5
�29
.5�
219.
93.
10.
804
�19
6.2
�21
0.0
3.4
0.81
4�
185.
9�
185.
53.
40.
839
�16
0.7
�19
7.5
0.5
0.82
7�
173.
0�
199.
81.
40.
825
�17
5.4
�20
6.4
4.3
0.81
8�
182.
0
A1W
288
.411
7.1
�10
3.0
1.6
0.80
3�
197.
1�
91.0
2.8
0.81
4�
186.
3�
67.2
1.1
0.83
5�
165.
0�
77.5
3.6
0.82
6�
195.
95.
41.
10.
824
�17
6.5
0.2
4.4
0.81
9�
180.
7
A1W
318
7.7
220.
8�
21.4
1.8
0.80
2�
198.
4�
8.6
0.4
0.81
2�
187.
912
.45.
30.
829
�17
0.8
�18
.37.
40.
804
�19
5.9
5.4
1.1
0.82
4�
176.
50.
24.
40.
819
�18
0.7
A1W
428
5.6
316.
969
.21.
20.
812
�18
8.1
81.0
4.4
0.82
1�
179.
1N
DN
DN
DN
DN
DN
DN
DN
D81
.32.
20.
821
�17
8.9
ND
ND
ND
ND
A1W
547
2.2
497.
219
8.0
4.3
0.80
0�
199.
920
5.7
3.8
0.80
5�
194.
724
0.2
7.3
0.82
8�
171.
722
9.9
1.8
0.82
1�
178.
522
0.3
3.2
0.81
5�
185.
0N
DN
DN
DN
D
Ave
rag
e0.
804
�19
5.9
0.81
3�
186.
80.
833
�16
7.0
0.82
0�
180.
40.
823
�17
7.5
0.82
1�
178.
9
Sta
nd
ard
dev
iati
on
0.00
54.
60.
006
5.6
0.00
55.
20.
011
10.6
0.00
54.
90.
004
4.3
Cu
ltu
res
H2O
dD
atst
art
H2O
dD
ath
arve
st
Ph
ytad
len
eN
atu
rall
yo
ccu
rin
gF
AM
EdD
dDS
tdev
ae
C16
Std
eva
eC
18:1
Std
eva
eC
20:1
Std
eva
eC
28:1
Std
eva
eC
30:1
Std
eva
e
Botr
yoco
ccus
bra
unii
,B
race
,M
arti
niq
ue
stra
in
BB
W1
�64
.0�
18.9
�29
1.2
7.8
0.72
2�
277.
6�
175.
51.
60.
840
�15
9.6
�18
2.3
0.1
0.83
3�
166.
6�
162.
45.
70.
854
�14
69.3
�17
9.7
3.5
0.83
6�
163.
9�
182.
83.
20.
833
�16
7.0
BB
W2
102.
216
1.1
�19
7.5
1.6
0.69
1�
308.
8�
49.1
1.0
0.81
9�
181.
0�
46.1
3.3
0.82
2�
178.
4�
22.5
3.9
0.84
2�
158.
1�
45.2
6.3
0.82
2�
177.
7�
49.4
2.4
0.81
9�
181.
2
BB
W3
184.
123
0.2
�15
0.1
1.4
0.69
1�
309.
27.
02.
70.
819
�18
1.5
1.3
2.7
0.81
4�
186.
025
.05.
40.
833
�16
6.8
4.7
0.5
0.81
7�
183.
3N
DN
DN
DN
D
BB
W4
273.
732
7.6
ND
ND
ND
ND
69.3
6.9
0.80
5�
194.
579
.26.
10.
813
�18
7.1
106.
24.
60.
833
�16
6.7
78.8
2.6
0.81
3�
187.
4N
DN
DN
DN
D
BB
W5
455.
151
3.6
ND
ND
ND
ND
205.
88.
50.
797
�20
3.3
208.
66.
50.
799
�20
1.5
ND
ND
ND
ND
217.
70.
70.
804
�19
5.5
ND
ND
ND
ND
Ave
rag
e0.
702
�29
8.5
0.81
6�
184.
00.
816
�18
3.9
0.84
1�
159.
50.
818
�18
1.6
0.82
6�
174.
1
Sta
nd
ard
dev
iati
on
0.01
818
.10.
017
16.6
0.01
312
.80.
007
9.7
0.01
211
.80.
010
10.1
Botr
yoco
ccus
bra
unii
,A
race
,T
itic
aca
stra
in
A2W
1�
64.0
�26
.0N
DN
DN
DN
D�
202.
14.
20.
819
�18
0.8
�21
4.3
1.2
0.80
7�
193.
3�
192.
71.
90.
829
�17
1.1
�20
8.0
3.3
0.81
3�
186.
8�
204.
75.
50.
817
�18
3.4
A2W
287
.612
8.6
ND
ND
ND
ND
�78
.90.
90.
816
�18
3.8
�83
.82.
30.
812
�18
8.2
�69
.15.
00.
825
�17
5.1
�72
.73.
40.
822
�17
8.3
ND
ND
ND
ND
A2W
318
2.1
221.
3N
DN
DN
DN
D8.
54.
60.
826
�17
4.3
�0.
50.
90.
818
�18
1.6
27.8
4.7
0.84
2�
158.
51.
82.
80.
820
�17
9.7
ND
ND
ND
ND
A2W
428
0.7
321.
1N
DN
DN
DN
D69
.91.
50.
810
�19
0.1
72.7
1.6
0.81
2�
188.
0N
.DN
.DN
DN
D82
.37.
20.
819
�18
0.8
ND
ND
ND
ND
A2W
548
5.0
499.
7N
DN
DN
DN
D20
8.0
2.1
0.80
5�
194.
520
6.2
1.0
0.80
4�
195.
724
5.2
0.1
0.83
0�
169.
7N
DN
DN
DN
DN
DN
DN
DN
D
Ave
rag
e0.
815
�18
4.7
0.81
1�
189.
40.
831
�16
8.6
0.81
9�
181.
40.
817
�18
3.4
Sta
nd
ard
dev
iati
on
0.00
87.
90.
005
5.5
0.00
77.
20.
004
3.7
ND
ND
Eac
hst
rain
was
gro
wn
infi
vem
edia
con
tain
ing
diff
eren
tco
nce
ntr
atio
ns
of
deu
teri
um
.T
he
dDva
lue
of
each
lip
idw
asm
easu
red
intr
ipli
cate
and
the
aver
age
and
stan
dar
dd
evia
tio
nar
ere
po
rted
.dD
valu
eso
ffr
eefa
tty
acid
sw
ere
mea
sure
do
nm
eth
yles
ter
der
ivat
ives
and
corr
ecte
dfo
ris
oto
pic
com
po
siti
on
of
the
add
edm
eth
ylgr
ou
p.
Th
edD
valu
eo
fth
ecu
ltu
rem
ediu
mat
the
star
tan
dh
arve
star
eal
sosh
ow
n.
std
ev,
stan
dar
dd
evia
tio
n;
ND
,n
ot
det
erm
ined
du
eto
insu
ffici
ent
qu
anti
tyo
fm
ater
ial.
ais
defi
ned
as(D
/H) l
ipid
/(D
/H) w
ate
r=
(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
);e
isd
efin
edas
(a�
1)*1
000
=[(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
)�
1]*1
000.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 587
Tab
le3
Hyd
roge
nis
oto
pe
rati
os
and
D/H
frac
tio
nat
ion
infr
eefa
tty
acid
s,F
AM
Es
and
ph
ytad
ien
ein
E.
unic
occ
aan
dV
.aure
us
Cu
ltu
res
H2O
dDat
star
tH
2O
dDat
har
vest
C14
fatt
yac
idC
16
fatt
yac
idC
18u
nsa
tfa
tty
acid
aC
18sa
tfa
tty
acid
a
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
dDS
tdev
ae
Eudori
na
unic
occ
a
EU
W1
�62
.4�
50.5
�13
5.7
0.9
0.91
0�
89.8
�10
4.2
1.9
0.94
3�
56.6
�97
.31.
60.
951
�49
.3�
116.
54.
80.
931
�69
.5
EU
W2
78.1
84.1
�32
.24.
10.
893
�10
7.3
35.3
2.1
0.95
5�
45.0
41.5
4.4
0.96
1�
39.3
9.4
5.4
0.93
1�
68.9
EU
W3
218.
622
0.7
110.
40.
40.
910
�90
.416
7.2
2.4
0.95
6�
43.9
162.
44.
10.
952
�47
.815
3.7
3.4
0.94
5�
54.9
EU
W4
340.
734
0.8
201.
25.
00.
896
�10
4.1
273.
40.
20.
950
�50
.2N
DN
DN
DN
DN
DN
DN
DN
D
EU
W5
458.
846
3.7
298.
53.
80.
887
�11
2.8
386.
03.
70.
947
�53
.041
1.8
8.7
0.96
5�
35.5
379.
511
.60.
942
�57
.5
Ave
rag
e0.
899
�10
0.9
0.95
0�
49.8
0.95
7�
42.9
0.93
7�
62.7
Sta
nd
ard
dev
iati
on
0.01
010
.30.
005
5.3
0.00
76.
70.
008
7.6
Volv
ox
aure
us
C18:3
/1b
VV
W1
�62
.5�
51.2
ND
ND
ND
ND
�87
.33.
30.
962
�38
.1�
77.0
0.9
0.97
3�
27.2
�90
.02.
40.
959
�41
.0
VV
W2
90.8
106.
5N
DN
DN
DN
D23
.24.
20.
925
�75
.340
.13.
60.
940
�60
.1N
DN
DN
DN
D
VV
W3
226.
123
2.6
ND
ND
ND
ND
149.
91.
90.
933
�67
.115
8.5
2.1
0.94
0�
60.1
161.
42.
10.
942
�57
.7
VV
W4
334.
833
8.5
ND
ND
ND
ND
265.
54.
80.
945
�54
.5N
DN
DN
DN
DN
DN
DN
DN
D
VV
W5
452.
145
2.8
ND
ND
ND
ND
372.
23.
20.
945
�55
.539
7.2
4.0
0.96
2�
38.3
405.
42.
80.
967
�32
.7
Ave
rag
e0.
942
�58
.10.
954
�46
.40.
956
�43
.8
Sta
nd
ard
dev
iati
on
0.01
414
.10.
016
16.4
0.01
312
.8
Cu
ltu
reH
2O
dDat
star
tH
2O
dDat
har
vest
Ph
ytad
ien
eH
epta
dec
ene
Nat
ura
lly
occ
urr
ing
FA
ME
dD
dDst
dev
ae
dDst
dev
ae
C16
std
eva
eC
18
un
satc
std
eva
eC
18
satd
std
eva
e
Eudori
na
unic
occ
a
EU
W1
�62
.4�
50.5
ND
ND
ND
ND
�14
3.5
2.3
0.90
2�
98.0
�11
0.3
1.8
0.93
7�
63.0
�10
6.0
1.7
0.94
2�
58.5
�12
0.9
2.4
0.92
6�
74.2
EU
W2
78.1
84.1
ND
ND
ND
ND
�22
.10.
80.
902
�98
.022
.74.
30.
943
�56
.723
.54.
20.
944
�55
.9�
5.2
4.2
0.91
8�
82.4
EU
W3
218.
622
0.7
ND
ND
ND
ND
60.0
2.2
0.86
8�
131.
712
1.7
2.4
0.91
9�
81.1
135.
57.
40.
930
�69
.8N
DN
DN
DN
D
EU
W4
340.
734
0.8
ND
ND
ND
ND
179.
81.
70.
880
�12
0.0
243.
72.
90.
928
�72
.425
4.5
0.2
0.93
6�
64.4
ND
ND
ND
ND
EU
W5
458.
846
3.7
ND
ND
ND
ND
302.
22.
30.
890
�11
0.3
354.
90.
50.
926
�74
.335
5.7
3.3
0.92
6�
73.8
ND
ND
ND
ND
Ave
rag
e0.
888
�11
1.6
0.93
0�
69.5
0.93
6�
64.5
0.92
2�
78.3
Sta
nd
ard
dev
iati
on
0.01
514
.60.
010
9.7
0.00
87.
50.
006
5.8
Volv
ox
aure
us
C18:3
/1e
Std
evC
18
e Std
ev
VV
W1
�62
.5�
51.2
�28
5.1
6.5
0.75
3�
246.
6�
132.
82.
10.
914
�86
.1�
106.
01.
80.
942
�57
.8�
81.6
2.1
0.96
8�
32.1
�98
.61.
60.
950
�50
.0
VV
W2
90.8
106.
5�
178.
17.
90.
743
�25
7.2
�11
.52.
90.
893
�10
6.6
22.9
0.7
0.92
4�
75.6
26.9
0.9
0.92
8�
71.9
24.3
3.9
0.92
6�
74.3
VV
W3
226.
123
2.6
ND
ND
ND
ND
76.7
N/A
h0.
874
�12
6.4
90.9
N/A
f0.
885
�11
4.9
103.
8N
/Af
0.89
6�
104.
5N
DN
DN
DN
D
VV
W4
334.
833
8.5
ND
ND
ND
ND
ND
ND
ND
ND
224.
01.
00.
915
�85
.524
5.2
2.7
0.93
0�
69.7
259.
17.
20.
941
�59
.2
VV
W5
452.
145
2.8
ND
ND
ND
ND
ND
ND
ND
ND
317.
26.
80.
907
�93
.433
3.6
3.5
0.91
8�
82.1
358.
86.
80.
935
�64
.7
Ave
rag
e0.
748
�25
1.9
0.89
4�
106.
40.
922
�78
.10.
936
�63
.90.
938
�62
.0
Sta
nd
ard
dev
iati
on
0.00
87.
50.
020
20.2
0.01
515
.30.
022
21.9
0.01
010
.2
Eac
hst
rain
was
gro
wn
infi
vem
edia
con
tain
ing
diff
eren
tco
nce
ntr
atio
ns
of
deu
teri
um
.T
he
dDva
lue
of
each
lip
idw
asm
easu
red
intr
ipli
cate
and
the
aver
age
and
stan
dar
dd
evia
tio
nar
ere
po
rted
.dD
valu
eso
ffr
eefa
tty
acid
sw
ere
mea
sure
do
nm
eth
yles
ter
der
ivat
ives
and
corr
ecte
dfo
ris
oto
pic
com
po
siti
on
of
the
add
edm
eth
ylgr
ou
p.
Th
edD
valu
eo
fth
ecu
ltu
rem
ediu
mat
the
star
tan
dh
arve
stis
also
sho
wn
.
ND
,n
ot
det
erm
ined
du
eto
insu
ffici
ent
amo
un
t;S
tdev
,st
and
ard
dev
iati
on
,b
ased
on
trip
lem
easu
rem
ents
un
less
no
ted
oth
erw
ise.
a,is
defi
ned
as(D
/H) l
ipid
/(D
/H) w
ate
r=
(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
);e,
isd
efin
edas
(a�
1)*1
000
=[(dD
lip
id+
1000
)/(d
Dw
ate
r+
1000
)�
1]*1
000.
aA
llfa
tty
acid
sw
ith
C18
skel
eto
nar
em
erge
din
totw
op
eak
so
nly
,u
nsa
tura
ted
and
satu
rate
d.
Un
satu
rate
dar
em
ain
lyC
18:1
and
C18:3
,b
ut
po
ssib
lyin
clu
des
smal
lp
eak
of
C18:2
.
bA
llfa
tty
acid
sw
ith
C18
skel
eto
nar
em
erge
din
totw
op
eak
so
nly
,u
nsa
tura
ted
and
satu
rate
d.
Un
satu
rate
dis
pre
do
min
ated
by
C18:3
,fo
llo
wed
by
C18:1
and
,to
am
uch
less
exte
nt,
C18:2
.In
VV
W5,
un
satu
rate
dis
on
eh
um
p.
cF
or
EU
W1,
2an
d3,
all
fatt
yac
ids
wit
hC
18
skel
eto
nar
em
erge
din
totw
op
eak
so
nly
,u
nsa
tura
ted
and
satu
rate
d.
Un
satu
rate
dar
em
ain
lyC
18:1
and
C18:3
,b
ut
incl
ud
essm
all
pea
ko
fC
18:2
;In
EU
W4
and
5,u
nst
aru
ated
iso
ne
hu
mp
.
dF
or
EU
W3,
4an
d5,
satu
rate
dF
AM
Ed
oes
no
th
ave
suffi
cien
tam
ou
nt
for
dDm
easu
rem
ent.
eA
llF
AM
Es
wit
hC
18
skel
eto
nar
em
erge
din
totw
op
eak
so
nly
,u
nsa
tura
ted
and
satu
rate
d.
Un
satu
rate
dis
pre
do
min
ated
by
C18:3
,fo
llo
wed
by
C18:1
and
,to
am
uch
less
exte
nt,
C18:2
.In
VV
W4
and
5,u
nst
aru
ated
iso
ne
hu
mp
.
fV
VW
3n
atu
ral
FA
ME
and
hep
tad
ecen
efr
acti
on
has
on
lyo
ne
mea
sure
men
td
ue
toli
mit
edam
ou
nt
afte
rp
roce
ssin
gan
dn
ot
incl
ud
edin
aver
age
calc
ula
tio
n.
588 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 589
was introduced to the mass spectrometer via an opensplit, where a flow of helium carried the H2 gas intothe mass spectrometer.
Sensitivity of the instrument was monitored withsix pulses of commercial H2 gas (ultra high puritygrade) via a second open split, four at the beginningand two at the end of each run (SFig. 2). Interfer-ence from Hþ3 , which is formed in the ion sourcevia ion-molecule reactions between Hþ2 and neutralH2, impedes the accurate determination of HD+
(Sessions et al., 1999). A calibration curve, or an‘‘Hþ3 factor’’, was determined daily to correct forthe Hþ3 interference. Hþ3 was determined by measur-ing the (m/z 3)/(m/z 2) response of 10 injections ofH2 reference gas. A low and stable value of less thanseven was typically achieved.
To every sample was added a set of co-injectionstandards with known dD values that were chosento bracket the peaks of interest in the chromato-grams. All isotopic standards were obtained fromDr. Arndt Schimmelmann at the BiogeochemicalLaboratory, Indiana University. For hydrocarbons,a mixture of C14 n-alkane, C16 n-alkane, 5a-andro-stane, and C40 n-alkane was used. C14 and C16 n-alk-anes were used as ‘‘throw-away’’ peaks to avoid anypossible hydrogen isotope memory effect, while 5a-androstane and C40 n-alkane were used as isotopicstandards for the computation of lipid dD valueswith IsoDat 2.0 software (Thermo Finnigan)(SFig. 2). A different mixture of standards was pre-pared for each lipid fraction.
Each sample was run in triplicate and the standarddeviation was usually less than 5& (Tables 1–3), sim-ilar to values reported by Sessions et al. (1999) andEnglebrecht and Sachs (2005). A set of 15 n-alkaneswith known dD values (Mixture A or B) acquiredfrom Indiana University were injected every 6–9 runsto ensure the accuracy of our data. dD values arereported with reference to the VSMOW standard.
2.4. Correction of dD contribution from –CH3 added
during methylation
Fatty acid dD values were measured on methylester derivatives. The dD value of the three H atomsadded during the methylation reaction with BF3 inMeOH was determined by experimentation withphthalic acid (Dr. Arndt Schimmelmann, IndianaUniversity) having a known dD value. Three exper-iments performed over the course of this study, eachmeasured in triplicate, indicated a dD value for thethree derivative H atoms of �115.7 ± 0.5&. This
value was used to correct the measured dD valuesof fatty acid methyl esters for the added hydrogenatoms by mass balance.
3. Results
3.1. Algal growth
Growth curves for all cultures indicated exponen-tial growth and, by extension, healthy cultures(SFig. 1). With a relatively small (but uncertain) con-tribution to light absorption caused by evaporation,most of the progressive increase in light absorptioncan be attributed to an increasing biomass. Growthrates were determined using the endpoints of theexponential portion of the growth curves (by curvefitting, SFig. 1) as k (divisions/d) = log2(N1/N0)/(T1 � T0), where N1 and N0 are the absorption valuesat the end and beginning, respectively, and T1 and T0
are the time (days) at the end and beginning, respec-tively, of the exponential growth period (Adolf et al.,2003). Calculated growth rates for the Morocco,Titi–caca and Martinique strains of B. braunii were0.128–0.138, 0.120–0.133 and 0.124–0.135 divi-sions/day, respectively. E. unicocca and V. aureus
had growth rates of 0.156–0.166 and 0.159–0.170divisions/day, respectively.
3.2. Hydrocarbon distributions within B. braunii,
E. unicocca and V. aureus
3.2.1. B. braunii, A race (Titicaca and Morocco)
A race cultures contained substantial quantitiesof odd carbon-numbered hydrocarbons in the rangeof C23–C33, each with a terminal double bond(Metzger et al., 1985a, 1986; Metzger and Largeau,1999). Hydrocarbons in the Titicaca strain consistedalmost exclusively of C25–C31 odd n-alkadienes:C27H52 (1, 18; E/Z), C29H56 (1, 20; E/Z), C31H60
(1, 22; E/Z) (SFig. 3a) (structure 1 in AppendixA), in accord with previous reports (Metzgeret al., 1986). Minor alkenes included C23H44 (1,14; E/Z), and C25H48 (1, 16; E/Z), and C29 alkatri-enes (SFig. 3a). On GC-IRMS, the Z/E isomers co-eluted, resulting in a single isotopic value.
Hydrocarbons in the Morocco strain differedfrom those in the Titicaca strain in that only oneisomer of each odd C25–C31 n-alkadiene wasobserved and an appreciable concentration of theC29H54 alkatriene eluted between the C29 and C31
alkadienes (SFig. 4a), as previously reported byMetzger et al. (1986).
590 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
3.2.2. B. braunii, B race (Martinique)
3.2.2.1. Botryococcenes. Polymethylated triterpenesof the generalized formula CnH2n�10, where 306 n 6 37, termed botryococcenes, are produced bythe B race (Metzger et al., 1985a,b, 1988; Metzgerand Largeau, 1999). Botryococcene mixtures exhibita large range of molecular mass and isomerismrelated to genetic and physicochemical factors(Metzger et al., 1985b). Seven botryococcenes wereidentified by mass spectrometry and coinjectionwith authentic standards provided by Dr. PierreMetzger (SFig. 5a and Appendix A), the most abun-dant of which were two isomers of the C34 botryo-coccene (structures 4 and 5 in Appendix A), withlesser quantities of C30, C31, C32 and C33 botryo-coccenes (structures 2, 3, 6 and 7 in Appendix A).
Positive identification of all botryococcenes dur-ing GC–IRMS was usually not possible due to inad-equate peak separation caused by the largeinjections required for dD determinations. Forexample, in GC–IRMS chromatograms, com-pounds 5–7, which were baseline-separated on theGC-FID (SFig. 5a), were not well-resolved, co-elut-ing in as few as two peaks. The last peak in GC–IRMS chromatograms may be another isomer ofthe C34 botryococcene or it may be a differenthomologue altogether. In either case, pending posi-tive structural identification by GC–MS, we refer tothat compound as C34 iso in Table 1 and SFig. 5a.
The C30 botryococcene is the precursor for theC31–C34 botryococcenes which are synthesized bymethylation on positions 3, 7, 16 and/or 20 of theC30 backbone (Metzger et al., 1987; Okada et al.,2004). The relative abundance of C30 botryococcenein a population results from a balance between itsproduction and its loss via methylation to formC31–C34 botryococcenes (Okada et al., 2004). As aculture ages and botryococcenes accumulate, the syn-thesis of botryococcenes is shifted toward the longerhomologues, resulting in minor amounts of C30 bot-ryococcene late in a culture cycle (Okada et al.,2004). Growing for 45 days, C34 botryococcenes weremost abundant in our cultures and C30 abundanceswere low (SFig. 5a).
3.2.2.2. Phytadiene. Quantities of phytadiene suffi-cient for D/H analysis existed in only three culturesowing to their loss during urea adduction ofFAMEs (the two co-eluted during column chroma-tography). Four isomers were observed, each with aparent ion at m/z 278. Structure determination wasbased on the comparison of our electron impact
mass spectra and gas chromatographic retentiontimes with published data (Fukushima et al., 1992).
Phytadienes are degradation products of phytol,the ester-linked side-chain of chlorophyll-a. Theycan be produced (i) at elevated temperatures fromchlorins, (ii) during GC analysis of underivatizedphytol, and (iii) by acidic dehydration of phytol(Grossi et al., 1996). However, a controlled experi-ment in which we injected underivatized phytol(Sigma-Aldrich) onto the GC demonstrated that lessthan 1% was converted to phytadienes, and thoseproduced had a different distribution of isomers.Furthermore, chlorophyll (Pfaltz & Bauer) subjectedto identical ASE extraction conditions resulted in noproduction of phytadienes. We therefore concludethat phytadienes occurred naturally in the B Raceof B. braunii, probably by dehydration of phytol(Volkman and Maxwell, 1986; Fukushima et al.,1992; Grossi et al., 1996). Here we report only thedD values of the most abundant isomer, neophytadi-ene (structure 8 in the Appendix A) (Table 2).Phytols were likely still present, along with sterols,in the alcohol fractions which were not analyzed.
3.2.3. Hydrocarbon distributions in Eudorina
unicocca and Volvox aureus
Both E. unicocca and V. aureus had similar hydro-carbon distributions characterized almost exclusivelyby 8-heptadecene (C17:1) (structure 9 in the AppendixA) with minor amounts of heptadecane (C17) and 10-nonadecene (C19:1) (SFig. 6a). This distribution ofhydrocarbons differs from the short-to-mid chainlength alkane distributions found in many algae(Gelpi et al., 1970). Because the hydrocarbon fractioncontained a single abundant compound it was com-bined with the FAME fraction for dD analysis(SFigs. 6c and 7b). Though no published reports oflipid distributions in E. unicocca or V. aureus exist,and few reports of 8-heptadecene are in the literature,it is likely derived from glycerol monoolein in the cellmembrane (Peterson, 1980).
Phytadienes occurred in E. unicocca and V. aur-
eus (SFigs. 6c and 7b), but losses during sampleworkup prevented D/H analysis of all but the mostabundant isomer (compound 8 in the Appendix A)in two cultures of V. aureus (Table 3).
3.3. Fatty acid distributions
The B race (Martinique) contained even num-bered monocarboxylic acids ranging from n-C14 ton-C30, with C16, C18:1 and C28:1 predominating
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 591
(SFig. 5b). Except for the C16 fatty acid (palmitic),all other fatty acids were unsaturated. Linoleic acid(C18:2x6) and linolenic acid (C18:3x3) eluted closelywith monounsaturated C18:1. Of the longer homo-logues, C28:1 was most abundant, followed bysubstantial quantities of C20:1, C26:1 and C30:1, andtrace amounts of C22:1 and C24:1. Such distributionsare in agreement with previous reports (Douglaset al., 1969).
Unlike the hydrocarbons, fatty acid distributionswere similar between A and B races. A race fatty acidswere primarily even numbered, varying in lengthfrom C16 to C32, with C16, C18:1 and C28:1 predomi-nating (SFig. 3b). The only major differences betweenthe A and B race fatty acids were a higher abundanceof C16 relative to C18:1 and a more pronounced even-over-odd predominance in the A race.
Compared to fatty acids in B. braunii, E. unicocca
and V. aureus produced fatty acids with substan-tially shorter chain lengths (SFig. 6b). The mostabundant fatty acid in both species was palmiticacid, C16:0. Other saturated fatty acids includedC18, C14, and trace amounts of C17 (SFig. 6b). Theprimary unsaturated fatty acids in the two specieswere C18:2, C18:3 and C18:1, with lesser amounts ofC20:1 (SFig. 7a). Differences between the fatty aciddistributions in the two species included relativelylarger amounts of C18 in V. aureus, and relativelylarger quantities of C20:1 in E. unicocca.
3.4. Naturally occurring fatty acid methyl esters
(FAMES)
Though not previously reported, our cultured B.
braunii contained large quantities of naturally occur-ring FAMEs. Both A (Titicaca) and B (Martinique)races had FAME distributions that closely followedtheir fatty acid (FA) distributions, with C16, C18:1
and C28:1 predominating and a pronounced even-over-odd predominance (SFigs. 3c and 5c).
A control experiment was conducted to ensurethat our lipid extraction procedure (DCM/MeOH(9:1 v/v), 150 �C, 1,500 psi) did not cause methyla-tion of fatty acids. C16 and C18 fatty acids (Sigma-Aldrich) were subjected to ASE extraction usingthree solvent systems: 100% DCM, 10% MeOH inDCM (1:9 v/v) and 50% MeOH in DCM (1:1 v/v).Extractions with DCM and 10% MeOH in DCMproduced no measurable fatty acid methyl esters.Extraction with 50% MeOH in DCM converted lessthan 0.1% of both C16 and C18 FAs to FAMEs, sub-stantially less than observed in our cultures.
While we recognize that the control experimentswith pure fatty acids lacked the potential catalyticand matrix effects other compounds and compo-nents of cultured algae (on a filter) may provide,we have no evidence to indicate that the FAMEs inthe cultures were produced during sample workup.Furthermore, when the Martinique and Titicacacultures were extracted at room temperature byultrasonication in 100% DCM, a gentle extractionprocedure, significant quantities of FAMEs werepresented.
Fatty acid methyl/ethyl esters have been reportedin the extracts of algae (Weete, 1976), fungus (Las-eter et al., 1968; Laseter and Weete, 1971), mamma-lian tissues (Saladin and Napier, 1967), pollen(Fathipour et al., 1967), and protozoans (Chuet al., 1972). Hydroxy fatty acid methyl esters werefound in marine algae (Sinninghe Damste et al.,2003).
Though we cannot exclude the possibility that asmall fraction of the FAMEs in our B. braunii wereproduced during sample preparation, the bulk ofthem are likely natural, begging the question whythey have not previously been reported in algal cul-tures. We hypothesize that the common procedureof methylation of fatty acids or hydrolysis of totallipid extracts before column chromatography andGC–MS analysis destroyed or masked any naturalFAMEs. The biosynthesis of FAMEs was investi-gated in the bacterium Mycobacterium phlei:S-adenosylmethionine as the most effective methyldonor and fatty acids as acyl acceptors (Akamatsuand Law, 1970).
E. unicocca and V. aureus also produced appre-ciable amounts of FAMEs. As with B. braunii,FAMEs in both species had very similar distribu-tions to the free fatty acids, with C16 FAME mostabundant, followed C18 and C18:1/3/2 FAMEs, anda pronounced even-over-odd predominance (SFigs.6c and 7b).
3.5. Hydrogen isotope fractionation in hydrocarbons
In quantifying D/H fractionation during lipidsynthesis we adopt the traditional definitionof ‘‘fractionation factor’’ for reactions under ther-modynamic equilibrium, i.e., alipid–water = (D/H)lipid
/(D/H)water = (dDlipid + 1000)/(dDwater + 1000).Because the natural variability of a is small anapproximation of ‘‘fractionation,’’ termed an‘‘enrichment factor’’, is often reported, with geo-chemists tending to use eA–B � 103 · lnaA–B (Hoefs,
592 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
2004), and ecologists, e = (aA–B � 1) · 1000 (Lajthaand Marshall, 1994). Both quantities were calcu-lated for each lipid in each culture and in the ensu-ing discussion we use the ‘‘ecologist’s’’ definition,elipid–water = (alipid–water � 1) · 1000 = [(dDlipid +1000)/(dDwater + 1000)� 1] · 1000 (primarily becausethe e values in five cultures calculated this way aremore consistent). Because lipids are depleted in deu-terium relative to water the value of fractionation isalways negative. A ‘‘larger’’ isotope fractionationmeans the absolute value of e is larger.
For continuity with recent literature we reportthe linear regression equations for five cultures ofeach species, y = slope · dDwater + intercept. Theobservation that the intercept 5(slope � 1) · 1000,a point discussed by Sessions and Hayes (2005), willbe discussed in detail in a subsequent paper. Fornow we simply note that the y-intercept of the linearregression of dDlipid versus dDwater is the D/H frac-tionation when dDwater = 0.
dD values of C27, C29 and C31 n-alkadienes in B.
braunii, A Race (Titicaca and Morocco) closelytracked water dD values, with R2 > 0.99 (Table 1;Fig. 1a and d). In spite of the large water dD range,a for C29 n-alkadiene in the Titicaca strain was nearlyconstant in five cultures, varying from 0.754 to 0.779,and averaging 0.769 with a standard deviation (r) of0.01. Similarly, a for the C29 n-alkadiene in the Mor-occo strain averaged 0.779 ± 0.007. Therefore, asimple approximation of a = (D/H)lipid/(D/H)water =(dDlipid + 1000)/(dDwater + 1000) is suitable for theexperiments reported here.
The e values for the C29 n-alkadienes varied from�220.6& to �246.0& in the Titicaca strain, averag-ing �231.5 ± 10.3&, and varied between �212.4&
to �228.5&, averaging �221.0 ± 7.3& in the Mor-occo strain (Table 1). Average a values were 0.769and 0.779, respectively. The dD values of C27 alkadi-enes in two of the Morocco strain cultures (A1W1and A1W2) differed substantially from the C29 alk-adiene dD values (Table 1), an observation forwhich we have no explanation.
dD values of C30–C34 botryococcenes in theB race, Martinique strain, of B. braunii closelytracked water dD values, with R2 > 0.99 (Fig. 2a).The e values of C30, C31 and C34 averaged�273&, �293&, and �310& to �359&, respec-tively (Table 1).
The dD values of 8-heptadecene in E. unicocca
and V. aureus closely tracked water dD values, withR2 > 0.99 (Fig. 3b and d). The e values averaged�111.6& and �106.4&, respectively (Table 3).
3.6. Hydrogen isotope fractionation in free fatty acids
3.6.1. Fatty acid dD values in B. braunii
Fatty acid dD values in all three A and B race B.braunii strains closely tracked water dD values, withR2 > 0.99 (Figs. 1b, e and 2b). In the A race Titicacastrain, e values of C16, C18:1 and C28:1 fatty acidsaveraged �192.4&, �187.2& and �174.6&,respectively, while e values of C20:1 fatty acids aver-aged �168.1& (Table 2). In the A race Moroccostrain, e values of C16 and C18:1 fatty acids averaged�195.9& and �186.8&, respectively (Table 2),while e values of C20:1 and C28:1 fatty acids averaged�167.0& and �177.5&, respectively. In the B race(Martinique), e values of C16, C18:1 and C28:1 fattyacids averaged �177.0&, �175.8& and �172.5&,respectively, compared to �160.1& for the C20:1
fatty acid. Taken together, the C16, C18:1 and C28:1
fatty acids of A and B race B. braunii had e valuesof �174& to �196&, while the C20:1 fattyacid was consistently enriched in deuterium by�15&.
3.6.2. Fatty acid dD values in E. unicocca and
V. aureus
Fatty acid dD values in both E. unicocca and V.
aureus closely tracked water dD values, with R2 >0.99 (Fig. 3a and c). The e values of the C16 fatty acidwere�49.8& in E. unicocca and�58.1& in V. aureus,significantly less than in B. braunii and, for E. unic-
occa, about 50& enriched in deuterium relative tothe C14 fatty acid (Table 3).
Some uncertainty in the absolute dD values ofC18 fatty acids exists as a result of partial co-elutionof C18:2, C18:3, C18:1, etc. (see notes in Table 3), attimes requiring the integration of all unsaturatedC18 acids as a single peak. Nevertheless, excellentlinear relationships between C18 fatty acid (satu-rated and unsaturated) and water dD values, andsimilar e values for the C18 and C16 fatty acids sug-gests that the uncertainty is relatively small.
3.7. Hydrogen isotope fractionation in FAMEs and
phytadienes
3.7.1. FAME and phytadiene dD values in B. braunii
dD values of naturally occurring C16, C18:1 andC28:1 FAMEs in B. braunii A (Titicaca) and B (Mar-tinique) races closely tracked water dD values, withR2 > 0.99 (Figs. 1c and 2c). The e values of the A race(Titicaca) FAMEs were �181.4& to �189.4& forC16, C18:1 and C28:1 FAMEs (Table 2). The C20:1
-300
-200
-100
0
100
200δD
of a
lkan
dien
es (
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.708x - 219.016
y = 0.691x - 207.429
(A race, Titicaca strain)
C31- diene
C29-diene
C27- diene
(C27:2)
(C29:2 )
-300
-200
-100
0
100
200
δD o
f alk
adie
nes
and
alka
trie
nes
(‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.750x - 215.365
y = 0.661x - 169.218
(A race, Morocco strain)
C31-diene
C29-triene
C29-diene
C27- diene (C27:2)
(C29:2)
-200
-100
0
100
200
300
δD o
f fre
e fa
tty a
cids
(co
rrec
ted,
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.776x - 157.036
y = 0.773x - 163.884
y = 0.782x - 180.955
y = 0.768x - 184.316
(A race, Titicaca strain)
C20:1
C28:1
C18:1
C16
(C20:1 )
(C16 )
(C18:1 )
(C28:1 )
-300
-200
-100
0
100
200
300
δD o
f fre
e fa
tty a
cids
(‰
)
0 100 200 300 400 500
δD of water at harvest (‰ )
y = 0.808x - 162.687
y = 0.795x - 171.904
y = 0.801x - 195.320
(A race, Morocco strain)
C20:1
C28:1
C18:1
C16
(C20:1)
(C28:1)
(C16)
-300
-200
-100
0
100
200
300
δD o
f nat
ural
occ
urin
g F
AM
E (
‰)
0 100 200 300 400 500D of water at harvest (‰)
y = 0.837x - 169.571
y = 0.836x - 183.940
y = 0.800x - 187.147
y = 0.778x - 177.098
(A race, Titicaca strain)
C20:1
C28:1
C18:1
C16(C20:1 )
(C16 )
(C18:1 )
(C28:1 )
a
d
e
b
c
Fig. 1. Relationships between lipid and water dD values in cultures of B. braunii, A race, Titicaca (a–c) and Morocco (d–e) strains. Allalkadiene, alkatriene, fatty acid and FAME dD values were highly correlated with water dD values (R2 > 0.99). (a) C27, C29 and C31
alkadienes in the Titicaca strain. C27 alkadienes were deuterium-enriched compared to other homologues. (b) C16, C18:1, C20:1 and C28:1
fatty acids in the Titicaca strain. C20:1 was most deuterium-enriched, followed by C28:1 and C16. (c) C16, C18:1, C20:1 and C28:1 FAMES inthe Titicaca strain. Chain-length-dependant dD variations were identical to the fatty acids. (d) C27, C29 and C31 alkadienes and C29
alkatriene in the Morocco strain. C27 was most deuterium-enriched. (e) C16, C18:1, C20:1 and C28:1 fatty acids in the Morocco strain. C20:1
was most deuterium-enriched, with C28:1 slightly enriched compared to C16.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 593
-300
-200
-100
0
100
δD o
f ph
ytad
iene
s (
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.557x - 282.096
-200
-100
0
100
200
300
δD o
f nat
ural
ly o
ccur
ing
FA
ME
(‰
)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.770x - 148.222
y = 0.746x - 165.893
y = 0.715x - 162.069
C20:1
C28:1
C18:1
C16
(C20:1 )
(C28:1 )
(C16 )
-200
-100
0
100
200
300
δD o
f fre
e fa
tty a
cids
(co
rrec
ted,
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.810x - 153.424
y = 0.807x - 167.963
y = 0.791x - 169.872
C20:1
C28:1
C18:1
C16
(C20:1 )
(C28:1 )
(C16 )
-400
-300
-200
-100
0
100
200δD
of b
otry
ococ
cene
s (
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.647x - 359.821
y = 0.726x - 329.067
y = 0.730x - 297.445
y = 0.751x - 277.948(Martinque)
C34bot iso
C34bot(5)+?
C34bot(4)
C32+C33 bot
C31bot
C30bot(C30 bot)
(C34 bot [4])
(C31 bot)
(C34 bot iso ?)
a
c d
b
Fig. 2. Relationships between lipid and water dD values in cultures of B. braunii, B race, Martinique strain. All botryococcene, fatty acid,FAME and phytadiene dD values were highly correlated with water dD values (R2 > 0.99). (a) C30–C34 botryococcenes (see structures inAppendix A). Deuterium-depletion increased with increasing carbon number. The compound labeled ‘‘C34 bot iso?’’ (i.e., the last peak inSFig. 5a) may be an isomer of the C34 botryococcene, or possibly a longer homologue. (b) C16, C18:1, C20:1 and C28:1 fatty acids. C20:1 wasconsistently the most deuterium-enriched homologue. (c) FAMEs. (d) Phytadiene.
594 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
FAME was �20& enriched in deuterium relative toother homologues, as observed for the free fattyacids. The e values of B race (Martinique) FAMEswere from �181.6& to �184.0& for C16, C18:1 andC28:1 (Table 2). And, similar to the A race, the C20:1
FAME was enriched in deuterium relative to otherFAMEs by �20&, with an e value of �159.5&.
Phytadiene dD values were measured in three Brace (Martinique) cultures that had sufficient mate-rial. dD values of phytadiene were linearly corre-lated with water dD values, with R2 > 0.99(Fig. 2d). The e values averaged �298.5 ± 18.1&
(Table 2), substantially greater than for fatty acids
and FAMEs. We also note that the slope of theregression was low, 0.557, compared to 0.73–0.83for fatty acids (Fig. 2d; Table 4). With only threecultures to define the slope, however, we are unableto evaluate its significance.
3.7.2. FAME and phytadiene dD values in
E. unicocca and V. aureus
dD values of C16 and C18:1 FAMEs from E. unic-
occa closely tracked water dD values, with R2 > 0.99(Fig. 3b). The e values for two FAMEs in E. unic-
occa were �69.5& and �78.3&, somewhat morenegative than the corresponding FAs (Table 3).
-100
0
100
200
300
400
-50 50 150 250 350 450
δD of water at harvest (‰ )
y = 0.895x - 63.069
y = 0.849x - 104.543
(Eudorina unicocca)
C18 FAME-unsat
C16 FAME
C17:1-ene
( C17:1-ene )
(C18-unsat )
-100
0
100
200
300
400
δD o
f fre
e fa
tty a
cids
(co
rrec
ted,
‰)
-50 50 150 250 350 450
δD of water at harvest (‰ )
y = 0.970x - 67.527
y = 0.858x - 93.322
y = 0.949x - 49.435
(Eudorina unicocca)
C18 FA
C14 FA
C16 FA
(C18 )
(C16 )
(C14 )
-150
-50
50
150
250
350
450
δD o
f fre
e fa
tty a
cids
(co
rrec
ted,
‰)
-50 50 150 250 350 450
δD of water at harvest (‰)
y = 0.948x - 45.885
y = 0.931x - 56.259
(Volvox aureus)
C18FAunsat
C18 FA
C16 FA
(C16 )
(C18 unsat)
-300
-200
-100
0
100
200
300
400
δD o
f FA
ME
s, C
17:1
-ene
and
phy
tadi
ene
(‰)
δD o
f FA
ME
s, C
17:1
-ene
and
phy
tadi
ene
(‰)
-50 50 150 250 350 450
δD of water at harvest (‰ )
y = 0.925x - 59.797
y = 0.740x - 93.548
y = 0.840x - 71.505
(Volvox aureus)
phytadiene
C18 FAME
C17:1-ene
C16 FAME(C18 FAME )
(C16 FAME )
(C17:1-ene )
a b
dc
Fig. 3. Relationships between lipid and water dD values in cultures of Eudorina unicocca and Volovx aureus. All fatty acid, FAME andn-alkene dD values were highly correlated with water dD values (R2 > 0.99). (a) C14, C16 and C18 fatty E. unicocca cultures. C14 wasdeuterium-depleted compared to both C16 and C18. (b) Heptadecenes and C16 and C18 FAMEs in E. unicocca. Heptadecene is substantiallydepleted in deuterium relative to fatty acids and FAMEs. (c) C16 and C18 fatty acids in V. aureus. (d) FAMEs, heptadecene andphytadienes in V. aureus. dD values of heptadecene (in three cultures for which we have data) were depleted in deuterium relative to fattyacids and FAMEs. Although only two analyses of phytadienes were performed both were significantly depleted in deuterium relative to allother lipids.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 595
dD values of the C16 FAME in V. aureus also closelytracked water dD values, with R2 = 0.99 (Fig. 3d).The e value for the C16 FAME in V. aureus,�78&, was more negative than in E. unicocca.The slope of the water–lipid dD regression for theC18:1 FAME was larger than for the C16 FAME(Fig. 3d; Table 5).
Two V. aureus cultures had sufficient phytadienefor dD determination. The e values of �246.6& and�257.2& indicate substantial deuterium depletionin phytadiene relative to FAs and FAMEs. A simi-
lar isotopic depletion in phytadiene relative to FAsand FAMEs was observed in B. braunii (Marti-nique) (Fig. 2d; Table 2).
4. Discussion
4.1. Hydrogen isotopes in lipids from B. braunii
Lipids within a single class generally fell within anarrow dD range of <50&. Between lipid classeshydrogen isotopic differences of 50–200& were
Table 4Summary of D/H fractionation parameters in lipids from B. braunii
Strain Analyte R2 Slopea Intercepta ab Stdevc ed
Hydrocarbon
Titicaca Strain C27-alkadiene 0.999 0.691 �207.4 0.7755 0.0141 �224.5C29-alkadiene 0.999 0.708 �219.0 0.7685 0.0103 �231.5C31-alkadiene 0.999 0.709 �219.0 0.7688 0.0098 �231.2
Morocco strain C27-alkadiene 0.995 0.661 �169.2 0.8028 0.0259 �197.2C29-alkadiene 0.998 0.750 �215.4 0.7790 0.0073 �221.0C29-alkatriene 0.998 0.723 �213.9 0.7757 0.0112 �224.3C31-alkadiene 0.998 0.744 �207.1 0.7849 0.0094 �215.1
Martinique strain C30 botryococcene 0.994 0.751 �277.9 0.7272 0.0103 �272.8C31 botryococcene 1.000 0.730 �297.4 0.7075 0.0046 �292.5C34 botryococcene 0.997 0.726 �329.1 0.6808 0.0089 �319.2C34 bot isomer + ?e 0.999 0.724 �316.3 0.6909 0.0058 �309.1C32 + C33 bot (?)e 0.999 0.699 �318.3 0.6847 0.0051 �315.3C34 bot (?)e 0.999 0.647 �359.8 0.6413 0.0030 �358.7
Fatty acids
Titicaca strain C16 fatty acid 0.992 0.768 �184.3 0.8076 0.0118 �192.44C18:1 fatty acid 0.999 0.782 �181.0 0.8128 0.0052 �187.2C20:1 fatty acidf 0.999 0.776 �157.0 0.8319 0.0111 �168.1C28:1 fatty acid 0.997 0.773 �163.9 0.8254 0.0103 �174.6
Morocco strain C16 fatty acid 0.999 0.801 �195.3 0.8041 0.0046 �195.9C18:1 fatty acid 0.998 0.797 �183.5 0.8132 0.0056 �186.8C20:1 fatty acidf 1.000 0.808 �162.7 0.8330 0.0052 �167.0C28:1 fatty acid 0.999 0.795 �171.9 0.8225 0.0049 �177.5
Martinique strain C16 fatty acid 0.998 0.791 �169.9 0.8230 0.0065 �177.0C18:1 fatty acid 0.998 0.790 �168.2 0.8242 0.0078 �175.8C20:1 fatty acid 0.997 0.810 �153.4 0.8399 0.0078 �160.1C28:1 fatty acid 0.999 0.807 �168.0 0.8275 0.0050 �172.5
Titicaca strain C16 FAME 0.998 0.778 �177.1 0.8153 0.0079 �184.7C18:1 FAME 0.998 0.800 �187.1 0.8106 0.0055 �189.4C20:1 FAMEf 0.998 0.837 �169.6 0.8314 0.0072 �168.6C28:1 FAMEf 0.999 0.836 �183.9 0.8186 0.0037 �181.4C16 FAME 1.000 0.715 �162.1 0.8160 0.0166 �184.0C18:1 FAME 1.000 0.736 �166.4 0.8161 0.0128 �183.9C20:1 FAMEf 0.999 0.770 �148.2 0.8405 0.0097 �159.5C28:1 FAME 1.000 0.746 �165.9 0.8184 0.0118 �181.6
Phytadiene
Martinique strain Phytadieneg 0.996 0.557 �282.1 0.7015 0.0181 �298.5
a The values of slopes and intercepts are derived from the linear regressions of dD values of biomrakers and dD values of waters atharvest; n is typically 5, unless noted otherwise.
b Fractionation factor, a, was calculated for each culture from the measured dD values of biomarkers and water at harvest according tothe equation: alipid–water = (D/H)lipid/(D/H)water = (dDlipid + 1000)/(dDwater + 1000). Reported is average value of five cultures withstandard deviation given n = 5 unless otherwise noted.
c Stdev, standard deviation of measured fractionation factors n = 5 unless otherwise noted.d Fractionation, e, is defined as 1000 · (a � 1) = [(dDlipid + 1000)/(dDwater + 1000) � 1] · 1000, where a is the fractionation factor.e Peak identification with certain uncertainty. See text for details.f Where n = 4.g Where n = 3.
596 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
common (Tables 1–3). In B. braunii A (Titicaca) andB (Martinique) races fatty acids and FAMEs withthe same carbon skeleton had similar dD values(Fig. 4a and c), suggesting that the hydrogen atoms
on the FAME were derived from the same pool ofhydrogen from which fatty acids were producedand that the esterification reaction imparted littleor no hydrogen isotope fractionation.
Table 5Summary of D/H fractionation parameters in lipids from E. unicocca and V. aureus
Strain Analyte na R2 Slopeb Interceptb ac Stdevd ee
Hydrocarbon
Eudorina unicocca 8-heptadecene 5 0.993 0.849 �104.5 0.8884 0.0146 �111.6
Fatty acids
Eudorina unicocca C14 fatty acid 4 0.997 0.858 �93.3 0.8991 0.0103 �100.9C16 fatty acid 5 0.999 0.949 �49.4 0.9502 0.0053 �49.8C18 FA-unsaturated 4 0.999 0.984 �47.0 0.9571 0.0067 �42.9C18 fatty acid 4 0.999 0.970 �67.5 0.9373 0.0076 �62.7
Volvox aureus C16 fatty acid 5 0.994 0.931 �56.3 0.9419 0.0141 �58.1C18 FA-unsat 4 0.992 0.948 �45.9 0.9536 0.0164 �46.4C18 fatty acid 3 0.996 0.978 �47.9 0.9562 0.0128 �43.8
Natural occurring fatty acid methyl esters (FAME)
Eudorina unicocca C16 FAME 5 0.998 0.895 �63.1 0.9305 0.0097 �69.5C18 FAME-unsat 5 0.999 0.898 �57.6 0.9355 0.0075 �64.5
Volvox aureus C16 FAME 5 0.988 0.840 �71.5 0.9146 0.0212 �78.1C18 FAME-unsat 5 0.982 0.838 �55.4 0.9280 0.0262 �63.9C18 FAME 4 0.998 0.925 �59.8 0.9380 0.0102 �62.0
Notes:a n indicates the number of cultures in which dD values of a biomarker were measured.b The values of slopes and intercepts are derived from the linear regressions of dD values of a biomarkers and dD values of waters at
harvest.c Fractionation factor, a, was calculated for each culture from the measured dD values of biomarkers and water at harvest according to
the equation: alipid–water = (D/H)lipid/(D/H)water = (dDlipid + 1000)/(dDwater + 1000). Reported is average value of five cultures withstandard deviation given n = 5 unless noted otherwise.
d Stdev, standard deviation of measured fractionation factors n = 5 unless noted otherwise.e Fractionation, e, is defined as 1000 · (a � 1) = [(dDlipid + 1000)/(dDwater + 1000) � 1] · 1000, where a is the fractionation factor.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 597
Hydrocarbons were depleted in deuterium rela-tive to fatty acids in all cultures: Alkadienes in theA race (Titicaca and Morocco) by about 40&,and botryococcenes in the B race (Martinique) by100–180& (Fig. 4a–c; Tables 1 and 2). While inagreement with Estep and Hoering (1980), whoreported deuterium enrichments in fatty acids rela-tive to hydrocarbons in higher plants, these findingsare at odds with Sessions et al. (1999) who reporteddeuterium depletions of �50–100& in fatty acidsrelative to hydrocarbons in both a higher plant Dau-
cus carota (carrot) and a microalga Isochrysis gall-
ana (haptophyte).Carbon chain length had a minor and variable
influence on dD values of fatty acids, FAMEs andalkadienes in B. braunii, but a significant influenceon dD values of botrycoccenes. In the Martiniquestrain, dD values of C16 and C18:1 fatty acids werevery similar while in both the Titicaca and Moroccostrains dD values of C18:1 fatty acids were slightlyelevated compared to those of C16 fatty acids. How-ever, in all three strains, dD values of C28:1 wereabout 10& more positive than those of C16 andC18:1. The C20:1 fatty acid, a minor component,
was always the most deuterium-enriched fatty acid,typically by �20& relative to C16, C18:1 and C28:1
FAs. Similar isotopic relationships were observedfor FAMEs (Table 2).
The trend toward deuterium enrichment withincreasing fatty acid chain length reported by Ses-sions et al. (1999) is not borne out by our data.Nor are the large deuterium enrichments of 112–163& in C18 relative to C16 FA in marine red andbrown macroalgae reported by Chikaraishi et al.(2004c). Chikaraishi et al. (2004c) also reported aprogressive deuterium depletion of �117& to�181& with increasing degree of unsaturation inC18:0 to C18:4 FAs. One possible explanation forthe discordant findings might be isotopic fraction-ation during lipid purification when argentationchromatography was used. Chikaraishi et al.(2004c) converted FAs into FAMEs and then usedAgNO3-impregnated silica gel to separate unsatu-rated from saturated FAs, a procedure likely tocause D/H fractionation in carbon double bonds,particularly when multiple unsaturations exist (deLigny, 1976). Alternatively, there might be real dif-ferences in isotopic fractionations during fatty acid
-300
-200
-100
0
100
200
300δD
of a
lgal
lipi
ds (
‰)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.782x - 180.955
y = 0.715x - 220.492
(Titicaca strain)
C18:1 FAME
C18:1 FA
C29-diene
(C18:1 FA )
(C29:2 diene)
-300
-200
-100
0
100
200
300
δD o
f alg
al li
pids
(‰
)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.797x - 183.499
y = 0.752x - 215.536
(Morocco strain)
C29:1 triene
C18:1 FA
C29-diene
(C18:1 FA )
(C29:2 diene)
-400
-300
-200
-100
0
100
200
300
δD o
f alg
al li
pids
(‰
)
0 100 200 300 400 500
δD of water at harvest (‰ )
y = 0.730x - 297.445
y = 0.726x - 329.067
y = 0.791x - 169.872
(Martinique strain)
phytadiene
C30bot
C34 bot
C16 FAME
C16 FA
(C30 bot)
(C34 bot)
(C16 FA)
a b
c
Fig. 4. D/H fractionation in different lipid classes within three strains of B. braunii. (a) A race, Titicaca strain. Alkadienes were �40&
depleted in deuterium relative to fatty acids. (b) A race, Morocco strain. Alkadienes and alkatrienes were �40& depleted in deuteriumrelative to fatty acids. (c) B Race, Martinique strain. Botryococcenes and phytadienes were >110& depleted in deuterium relative to fattyacids and FAMEs.
598 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
synthesis between the green algae we cultured andthe red/brown algae Chikaraishi et al. (2004c) inves-tigated. Additional culture experiments and stan-dardized lipid purification procedures are requiredto address this discrepancy.
Hydrogen isotope fractionation in alkadienesfrom the A race may have a slight carbon-chain-length dependence. In the Titicaca strain, dD valuesof C29 and C31 alkadienes were almost identical,while the C27 alkadiene was enriched in deuteriumby �7& relative to C29 and C31 (Fig. 1a; Table 1).In the Morocco strain dD values of C27 alkadieneswere enriched by �24& relative to C29 (Table 1).The one alkatriene analyzed, C29:3, had a dD value
very close to the C29:2-diene (Table 1), suggesting fur-ther desaturation did not cause D/H fractionation.
Botryococcene dD values had a discernible car-bon-chain-length dependence, with longer chain-lengths generally associated with more negative dD
values (Fig. 2a; Table 1). As such, the C30 botryo-coccene—the precursor to all other botryococc-enes—had the least negative dD value (�272.8&),followed by the C31(�292.5&), and C34 botryococc-enes (�319.2&) (Table 1). We hypothesize that themethyl donor in botryococcene synthesis is depletedin deuterium relative to the C30 botryococcene,resulting in increasingly negative dD values withincreasing carbon number. Nevertheless, our con-
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 599
clusion must be considered tentative due to co-elu-tion of the C32 and C33 botryococcenes on theGC–IRMS (see Section 3.2.2.1).
Phytadienes in the B race had e values that aver-aged �298.5&, significantly deuterium-depleted rel-ative to fatty acids (Table 2), consistent with priorstudies of D/H fractionation in phytol from marinealgae (Estep and Hoering, 1980; Sessions et al.,1999) and vascular plants (Chikaraishi et al.,2004a). Hydrogen isotope fractionation betweenphytadiene and water was similar to that betweenC34 botryococcenes and water (Fig. 4c; Tables 1and 2).
4.2. Hydrogen isotopes in lipids from E. unicocca and
V. aureus
FAMEs in both E. unicocca and V. aureus cul-tures were 20–27& depleted in deuterium relativeto fatty acids (Table 3). The only dominant hydro-carbon in both species, 8-heptadecene, was 48–62& more negative than the C16 FAs.
A chain length effect may exist for fatty acids in E.
unicocca cultures in which the C14 FA was 50& morenegative than the C16 and C18 fatty acids (Table 3).
Two analyses of phytadiene dD values were madein V. aureus, yielding deuterium depletions of�246&
and�257& relative to water (Table 3), similar to thedeuterium depletions in B. braunii (Martinique).
4.3. Hydrogen isotope fractionation during lipid
synthesis
During photosynthesis water is oxidized to O2
and NADP+ is reduced to NADPH (Raven et al.,1999), with the latter associated with a large D/Hfractionation (Estep and Hoering, 1980; Luo et al.,1991). 3-Phosphoglyceric acid (PGA) (also calledglycerate 3-phosphate) is produced when ribulose-1,5-bisphosphate and CO2 are catalyzed by theRubisco enzyme. The PGA is then converted to gly-ceraldehydes-3-phosphate (G3P) using ATP asenergy and NADPH as reductant. During this con-version, NADPH yields hydrogen that is incorpo-rated into organic matter.
Hydrogen isotopic variations within and betweenlipids in a single cell, and between lipids in differentspecies, come from three sources: (1) the isotopiccomposition of biosynthetic precursors, (2) isotopeeffects during biosynthetic reactions, and (3) the iso-topic composition of added hydrogen, primarilyfrom NADPH and NADH (Smith and Epstein,
1970). The third source of isotopic variation is notrelated to the flow of substrates used in the assemblyof carbon skeletons, providing a distinct differencefrom carbon isotope studies (Sessions et al., 1999).
Lipids in plants derive from three primary biosyn-thetic pathways (Chikaraishi et al., 2004a). Straight-chain (n-alkyl) lipids are produced via the acetogenicpathway using acetyl coenzyme-A (acetyl-CoA). Iso-prenoid (i.e., branched) lipids are synthesized via themevalonic acid (MVA) or non-mevalonic-acid path-way (DOXP/MEP, or 1-deoxy-D-xylulose 5-phos-phate/2-C-methylerythritol 4-phosphate) usingisopentenyl pyrophosphate (IPP).
4.3.1. D/H fractionation in acetogenic lipids: fatty
acids
Fatty acids are the precursor to all other aceto-genic lipids and acetyl-CoA is the direct precursorto fatty acids (Harwood, 1988). The principle pho-tosynthate in the plant cell, sucrose, is convertedto glucose 1-phosphate and further degraded inthe cytoplasm to either malic or pyruvic acid(Stumpf, 1980). These two respiratory substratesenter the mitochondria where pyruvate is oxida-tively decarboxylated to form acetyl-CoA and CO2
by the pyruvate dehydrogenase complex (Fig. 5a).Palmitic acid (C16) is synthesized from acetyl-
CoA and malonyl-CoA via several enzymatic reac-tions involving acyl carrier protein (ACP), NADPHand NADH (Fig. 5a) (Stumpf, 1980; Ohlrogee,1987). Its immediate precursor is palmitoyl-ACP,which can have three fates: elongation to stearoyl-ACP, use in glycerolipid synthesis, or hydrolysis topalmitic acid. Hydrogen isotope fractionation inC16 fatty acid in algae is the sum of these processes.
Elongation of palmitoyl-ACP to stearoyl-ACP,and the subsequent desaturation of stearoyl-ACPto form oleoyl-ACP do not, in all likelihood, causemuch hydrogen isotope fractionation, as evidencedby similar dD values in C16 and C18:1 fatty acids inthree strains of B. braunii (Table 2). This is consistentwith the findings of Behrouzian et al. (2001) whoreported that desaturation by stearoyl-ACP D9
desaturase proceeded without a measurable kineticisotope effect. Stearoyl-ACP desaturase has such ahigh activity that stearoyl-ACP rarely accumulates,with near complete conversion to oleate (Harwood,1997). Supporting that notion is our observation ofmuch higher concentrations of C18:1 than C18:0 fattyacids in B. braunii (SFigs. 3b, 4b, and 5b).
Longer FA homologues are synthesized fromoleic acid by two different elongases, one in the
Fig. 5. Lipid biosynthetic pathways in B. braunii with estimated D/H fractionation. (a) Fatty acids and alkadienes in the A race (Titicaca)are produced via the acetogenic pathway using acetyl coenzyme-A as a precursor. (b) Botryococcenes and phytadiene in the B race(Martinique) are isoprenoid (i.e., branched) lipids synthesized via the DOXP/MEP pathway using isopentenyl pyrophosphate as aprecursor (IPP). D values shown represent D/H fractionation when dDwater = 0. Acronyms are as follows, Co-A: coenzyme A; ACP: acylcarrier protein; DMAPP: dimethylallyl diphosphate; DOXP: 1-deoxy-D-xylulose 5-phosphate; FPP: farnesyl diphosphate; GPP: geranyldiphosphate; GGPP: geranylgeranyl diphosphate; IPP: isopentenyl diphosphate; MEP: 2-C-methyl-D-erythritol 4-phosphate; PSPP:presqualene diphosphate. (Referred from Pollard et al., 1979; Stumpf, 1980, 1987; Templier et al., 1984, 1991; Agrawal and Stumpf, 1985;Harwood, 1988; Schwender et al., 1997, 2001; Lichtenthaler, 1999; Rohmer, 1999; Charon et al., 1999, 2000; Szkopinska, 2000; Wankeet al., 2001; Okada et al., 2004; Sato et al., 2004).
600 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
endoplasmic reticulum (C18-CoA elongase, catalyz-ing C18:1! C20:1), the other in the Golgi apparatus(C20-CoA elongase, catalyzing C20:1! C22:1!C30:1) (Agrawal and Stumpf, 1985a,b; Lessireet al., 1985). Agrawal and Stumpf (1985b) furtherdemonstrated that both the C18:1! C20:1 andC20:1! C22:1 elongations can proceed withNADPH as a reductant, but that only NADH canbe used for the C18:1! C20:1 elongation (Fig. 5a).
We hypothesize that consistently higher dD val-ues of C20:1 relative to all other fatty acids (Table2; Figs. 1b, e and 2b) resulted from either or bothits synthesis from C18:1 in the endoplasmic reticulum
or NADH serving as the sole reductant. Likewise,similar dD values of both C28:1 and C30:1 fatty acids(Table 1) may result from their synthesis in theGolgi apparatus. The hydrogen isotope data thussupport the notion that the C20:1 fatty acid is synthe-sized from C18:1 in a different location than that inwhich subsequent elongations occur.
4.3.2. D/H fractionation in acetogenic lipids:
alkadienes
The observed �40& deuterium depletion of alk-adienes relative to fatty acids (Fig. 4a, b) can beattributed to D/H fractionation during elongation
Fig. 5 (continued)
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 601
and decarboxylation of C18:1 fatty acids during alk-adiene (and alkatriene) biosynthesis (Templier et al.,1984, 1991). Considering that D/H fractionationduring elongation of fatty acids is significantly lessthan 40& it is possible that the decarboxylationprocess is characterized by a large kinetic isotopeeffect (Fig. 5a).
Furthermore, we note that the C29 and C31 alkadi-enes had similar dD values in both (Titicaca andMorocco) A race strains, whereas the C27 alkadienewas enriched in deuterium (Fig. 1a and d; Table 1),suggesting a common biosynthetic route for C29
and C31 that differed from that for C27. The C29
alkatriene dD values were similar to the C29 alkadiene
602 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
values suggesting that further desaturation resultedin little or no D/H fractionation.
4.3.3. D/H fractionation in isoprenoid lipidsdD values of isoprenoid (i.e., branched) lipids,
such as botryococcenes and phytadienes, were muchlower than dD values of acetogenic lipids (Tables 1–3). This pattern has been observed previously byseveral researchers and attributed to the differentbiosynthetic pathways for those two classes of lipids(c.f., Sessions et al., 1999; Hayes, 2001).
Green algae use the DOXP/MEP pathway exclu-sively for isoprenoid synthesis (Lichtenthaler, 1999;Schwender et al., 2001; Sato et al., 2003). Hydrogenatoms in isoprenoids have two sources. The decar-boxylation of pyruvate and G3P yields DOXP(Schwender et al., 1997; Charon et al., 1999,2000). A rearrangement followed by a reductionleads to the formation of MEP, which possessesthe C5 isoprene backbone (Fig. 5b) (Charon et al.,1999, 2000). That step uses NADPH as a hydrogendonor, thus incorporating deuterium-depletedhydrogen to C1 of MEP (corresponding to C-4 ofIPP) (Fig. 5b). Conversion of MEP into IPP resultsin the elimination of three water molecules via tworeductions and one phosphorylation (Lichtenthaler,1999; Rohmer, 1999). Hydrogen atoms on C2 andC4 of IPP and/or DMAPP (dimethylallyl diphos-phate) comes from NADH or NADPH (Charonet al., 1999), while the remaining hydrogen atoms
-200
-100
0
100
200
300
400
δD o
f C1
6
fatty
aci
ds (
corr
ecte
d, ‰
)
0 100 200 300 400 500
δD of water at harvest (‰)
y = 0.768x - 184.3 r2 = 0.992
y = 0.791x - 169.9 r2 = 0.998
y = 0.869x - 74.6 r2 = 1.000
y = 0.931x - 56.3 r2 = 0.994EU
A2
BB
VVHT
VV
(VV at 15˚C)
(VVHT at 25˚C)
( BB at 28˚C )
(A2 at 28˚C )
a b
Fig. 6. Species and family dependence of D/H fractionation during synthdifferences in D/H fractionation in a single lipid are observed between famwhich has a small influence (see, for example, the lower slope for the high-tthe 15 �C experiment (VV)). Acronyms are as follows: EU: E. unicocca groaureus grown at 25 �C. BB: B. braunii, B race, Martinique strain grown a
comes from DOXP/MEP. It is likely that the IPPsynthesis is responsible for much of the deuterium-depletion in isoprenoids from green algae (Fig. 5b).
IPP and farnesyl diphosphate (FPP), the directprecursors C30 botryococcene, are synthesized viathe DOXP/MEP pathway in B. braunii (B Race)(Sato et al., 2003) via a 2-step reaction in the chlo-roplast (Okada et al., 2004) (Fig. 5b). In the firststep two molecules of FPP are condensed to formpresqualene diphosphate (PSPP). In the secondstep, the cycloprapane ring in PSPP is cleaved, fol-lowed by reduction with NADPH (Fig. 5b) (Okadaet al., 2004).
C30 botryococcene is rapidly converted to highermolecular weight (C31–C34) botryococcenes by suc-cessive methylation reactions (Metzger et al.,1987). The source of the additional methyl groupsis presumably methionine, the origin of whichremains unknown (Metzger et al., 1987). The factthat we observed progressively more deuterium-depletion with increasing carbon number (i.e., C30–C34) in botryococcenes suggests either that themethionine is depleted in deuterium relative to theC30 botryococcene and/or there is a kinetic isotopeeffect associated with the methylation reaction. Theposition of methylation on the C30 botryococcenemay also influence the dD values of different isomers.
Phytadiene (specifically, neophytadiene (Com-pound 8 in the Appendix A), the most abundant iso-mer) is produced by dehydration of phytol (Volkman
-200
-100
0
100
200
300
400
δD o
f C16
fat
ty a
cid
met
hyl e
ster
(‰
)
0 100 200 300 400 500
δD of water at harvest (‰ )
y = 0.778x - 177.1 r2 = 0.998
y = 0.772x - 70.6 r2 = 1.000
y = 0.715x - 162.1 r2 = 1.000
y = 0.840x - 71.5 r2 = 0.988
(VVHT at 25˚C)
(VV at 15˚C)
(BB at 28˚C)
(A2 at 28˚C)
esis of (a) palmitic acid (C16:0) and (b) C16 FAME. Large (�100&)ilies of green algae. The effect is much larger than for temperature,
emperature (VVHT, 25 �C) experiment with V. aureus compared town at 15 �C. VV: V. aureus grown at 15 �C. VVHT1 and VVHT2: V.
t 28 �C. A2: B. braunii, A race, Titicaca strain, grown at 28 �C.
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 603
and Maxwell, 1986; Grossi et al., 1996), which inturn is synthesized from GGPP (geranylgeranyldiphosphate) during three hydrogenation reactionsusing NADPH (Chikaraishi et al., 2004a). IPP andDMAPP serve as precursors to GGPP (Fig. 5b).Though we did not measure the dD value of phytolwe did observe e values of�251& to�298& in phyt-adienes, relative to water, in three species (B. braunii
(B race), E. unicocca and V. aureus), in accord withChikaraishi et al. (2004a).
4.3.4. Summary of D/H fractionation during lipidsynthesis
Though fatty acids, botryococcenes and phytadi-enes in the Martinque strain of B. braunii, and fattyacids and phytadiene in V. aureus are all synthesizedin the chloroplast (cf. Sato et al., 2003), presumablywith the same pool of NADPH, significant hydro-gen isotopic differences exist between botryococc-enes/phytadienes and fatty acids in B. braunii, andphytadiene and fatty acids in V. aureus. Thesehydrogen isotopic differences between acetogenicand isoprenoid lipids result from different biosyn-thetic pathways for the two types of lipids.
4.4. Species-dependence of hydrogen isotope
fractionation in green algae
A small but significant difference of �10–15&
in D/H fractionation during lipid synthesis wasobserved in different species, while a large differenceof�90–100& was observed between families of greenalgae (B. braunii, from the Trebouxiophyceae; E.
unicocca and V. aureus from the Chlorophyceae)(Fig. 6a and b).
The water–lipid isotopic difference for two com-pounds (C16 FA and FAME) from four species andtwo families was compared (Fig. 6a and b). Notwith-standing a small influence of temperature on D/Hfractionation during C16 FA and FAME synthesisof�–3&/�C (Zhang and Sachs, unpublished results),deuterium depletion in C16 FAs (Fig. 6a) and FAMEs(Fig. 6b) was much greater (i.e., >110&) in B. braunii
(Trebouxiophyceae) compared to either V. aureus orE. unicocca (Chlorophyceae), suggesting a large inter-family difference in D/H fractionation.
Between species the water-C16 FA isotopic differ-ence was much smaller, amounting to 15& betweenthe B (Martinique) (�177&) and A (Titicaca)(�192&) races of B. braunii (Table 2), and 8&
between E. unicocca (�50&) and V. aureus
(�58&) (Table 3). Similarly, the water-C16 FAME
isotopic difference was 1& between Martinique(�184&) and Titicaca (�185&) strains of B. brau-
nii, and 8& between E. unicocca (�70&) and V.
aureus (�78&) (Tables 2 and 3).Schouten et al. (2006) studied the effect of growth
rate on D/H fractionation in alkenones from marinecoccolithophorids. In a subsequent paper, we willexplore the role that growth rate may play in D/Hfractionation of lipids. Here, we note that a 5-folddifference in growth rate (at constant temperature)caused little change in e. So although the large dif-ference in D/H fractionation during C16 FA andFAME synthesis in B. braunii vs. E. unicocca andV. aureus may result in part from differing growthrates, there are likely to be large differences in thehydrogen isotopic fractionation during synthesis ofa single lipid by different families of algae.
4.5. Implications for paleohydrologic reconstructions
from lipid D/H
The near-perfect linear correlation (R2 > 0.99)between all lipids studied and water dD values infive species of freshwater microalgae, and in alke-nones from the marine phytoplankton Emiliania
huxleyi (Englebrecht and Sachs, 2005), provides asound basis for using sedimentary algal lipid dD val-ues to reconstruct water dD values through time. Animportant caveat to doing so, however, is that verylittle is known yet about the role environmentalparameters such as nutrients, light, temperatureand salinity may play in influencing D/H fraction-ation during lipid synthesis. Studies we and otherresearchers are conducting ought to shed light onthese potential influences in the near future.
Tables 4 and 5 list the linear regression equationsfor all lipids we have studied as well as a and e values.Though the empirical linear regression equations canusually be used to derive water dD values from lipiddD values, neither the slope nor the intercept areequivalent to D/H fractionation as expressed by aand e. Furthermore, simply adding e to lipid dD val-ues to reconstruct water dD values could be inappro-priate if water dD values span a large range. Wetherefore recommend reconstructing water dD valuesfrom lipid dD values using a, which accounts for theD/H discrimination between lipid and water, usingthe equation: dDwater = (dDlipid + 1000)/a � 1000.
To demonstrate the utility of this approach wecalculated water dD values from palmitic acid dD
values in five B. braunii (Martinique) cultures usingthe average a value (0.816) reported in Table 4, and
604 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
compared this value to the measured dDwater valuesin those cultures at the time of harvest. The differ-ences were 0.5&, 8.1&, 4.2&, 3.4&, and �20.4&
for BBW1 to BBW5, respectively. Except forBBW5, which had dDwater = 500&, the recon-structed dDwater values are satisfactory.
Furthermore, if non source-specific lipids, such asmost fatty acids and sterols, are used to reconstructwater dD values, the influence of changing organicmatter inputs to the particular lake or ocean sedi-ment must be considered. Compounds such as pal-mitic acid (C16) and oleic acid (C18:1) areubiquitous, being found in higher plants, algae andbacteria. Because large (>90&) differences in D/Hfractionation were observed in palmitic acid in differ-ent families of green algae (Fig. 6a) it is prudent toassume similarly large differences occur in othertypes of algae. Consequently, a change in the propor-tion of different algae in a lake would be expected toalter the dD value of palmitic acid deposited in sedi-ment, absent any change in the lake water D/H ratio.
Palmitic acid and other lipids that derive fromboth aquatic and terrestrial plants and algae furtherconfound the interpretation of down-core changesin dD measured on a non source-specific (i.e., ubiq-uitous) lipid because higher plant lipids are enrichedin deuterium by �30–60& relative to aquatic lipids(Sachse et al., 2004). Thus any change in the propor-tion of palmitic acid derived from aquatic algae andhigher plants would be expected to alter the dD
value of sedimentary palmitic acid in the absenceof any change in lake or meteoric water dD values.
Clearly the use of non-specific lipids in lake sed-iment, such as palmitic acid, for reconstructing lakewater dD values and paleohydrology is susceptibleto misinterpretation (cf Huang et al., 2002). A betterapproach for down-core lake water dD reconstruc-tions is the use of a lipid biomarker specific to afamily of plants or algae, if not a species (the inter-species differences being rather small at �5–15&),and for which the fractionation factor, a, or anempirical water–lipid dD calibration has been devel-oped. An example would be the C34 botryococcene,produced solely by the B race of B. braunii.
5. Conclusions
Here we have shown that lipids from five speciesof cultured green algae, including E. unicocca, V.
aureus and three strains of B. braunii, were near-per-fect recorders (R2 > 0.99) of water D/H ratios. Bar-ring any as yet unknown environmental influences
on D/H fractionation during lipid synthesis, algallipid dD values can therefore be used as surrogatesfor water dD values when the lipids derive from asingle species or family.
All algal lipids were highly depleted in deuteriumrelative to environmental water, the result of kineticisotope effects during enzymatic processes. In thegreen algae we studied, deuterium depletion in lipidsvaried with the biosynthetic pathways that pro-duced them. As observed previously in other plantsand algae, isoprenoid lipids (botryococcenes andphytadiene), which are synthesized via the DOXP/MEP pathway, were highly depleted in deuteriumrelative to acetogenic lipids (straight-chain fattyacids, alkadienes, etc.), which are synthesized viathe acetogenic pathway.
Systematic hydrogen isotopic differences wereassociated with carbon chain length. The C20:1 fattyacid was consistently deuterium-enriched relative tothe C28:1 and C30:1 fatty acids, both of which wereenriched in deuterium relative to the C16 and C18:1
fatty acids. We attribute these differences to the siteof elongation, which differs for C18:1–C20:1, and forC20:1–C28:1/C30:1 fatty acids. Similarly, the C30 bot-ryococcene is enriched in deuterium relative tolonger (C31–C34) homologues, implicating the meth-ylation reactions as an additional source of hydro-gen isotopic fractionation.
Alkadienes, the products of elongation anddecarboxylation of the C18:1 fatty acid, are depletedin deuterium by 40& relative to that precursor, pre-sumably the result of isotopic discrimination duringthe decarboxylation reaction.
D/H fractionation in lipids also varies betweenspecies of green algae. Deuterium depletion in a sin-gle compound, such as the C16 fatty acid (palmiticacid), was approximately 100& greater in B. braunii
than in both E. unicocca and V. aureus.Attempts at reconstructing water dD values
through time using non-source-specific lipids, suchas palmitic acid, therefore run the risk of misinter-preting changes in the source of sedimentary lipidsfor changes in water D/H ratios. Lipid biomarkersunique to a family (or genus) should be targetedfor down-core reconstructions of water D/H ratiosusing empirically derived water–lipid fractionationfactors (a) established from culture experiments.
Acknowledgements
We are indebted to Dr. Pierre Metzger in theLaboratoire de Chimie Bioorganique et Organique
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 605
Physique (France) for supplying Botryococcous
braunii inocculants and for sharing his wisdom onculturing and understanding this alga, and to Dr.Daniel Repeta at the Woods Hole OceanographicInstitution for assisting with the culture experi-ments. We thank Drs. Stefan Schouten, Alex Ses-sions and Mark Pagani for constructive reviewsthat greatly improved this manuscript, and AmyCash for useful editorial comments. We are gratefulto Carolyn Colonero and Roger Summons for assis-tance with GC–IRMS instrumentation at MIT, An-thony Faiia for water dD analyses at DartmouthCollege, and Yanek Hebting for assistance in hydro-genating botryococcenes and alkadienes. We thank
1
23
4
5
24
7
8
9
10
11
12
131
23 6
25
2627
28
CH3(CH2)7CH CH(CH2)xCH CH21.
2.
3.
4.
5.
6.
7.
8.
CH3(CH2)6CH CH(CH2)7CH39.
the Culture Collection of Algae and Protozoa inScotland, UK for providing inoculants of E. unic-
occa and V. aureus. Funding for this research wasprovided in part by the Gary Comer Science andEducation Foundation (J.S.), the Jeptha H. andEmily V. Wade Award for Research at MIT (J.S.),and a Henry L. and Grace Doherty Professorshipat MIT (J.S.).
Appendix A
Molecular structures of hydrocarbons discussedin this paper. Compounds 2–7 are botryococcenescontaining 30–34 carbon atoms.
4
15
1617
18
19
2021
2229
30
C30 bot
C31 bot
C32 bot
C34 bot
C34 bot-iso
C33 bot
Where x=15, 17 or 19
Phytadiene
606 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
Appendix B. Supplementary data
Supplementary data associated with this articlecan be found, in the online version, at doi:10.1016/j.orggeochem.2006.12.004.
Associate Editor—S. Schouten
References
Adolf, J.E., Stoecker, D.K., Harding, L.W., 2003. Autotrophicgrowth and photoacclimation in Karlodinium micrum (Dino-phyceae) and Storeatula major (Cryptophyceae). Journal ofPhycology 39, 1101–1108.
Agrawal, V.P., Stumpf, P.K., 1985a. Characterization andsolubilizatoin of an acyl chain elongation system in micro-somes of leek epidermal cells. Archives of Biochemistry andBiophysics 240, 154–165.
Agrawal, V.P., Stumpf, P.K., 1985b. Elongation system involvedin the biosynthesis of erucic acid from oleic acid in developingBrassica juncea seeds. Lipids 20, 361–366.
Akamatsu, Y., Law, J.H., 1970. The enzymatic synthesis of fattyacid methyl esters by carboxyl group alkylation. Journal ofBiological Chemistry 245, 709–713.
Behrouzian, B., Buist, P.H., Shanklin, J., 2001. Application ofKIE and thia approaches in the mechanistic study of a plantstearoyl-ACP D9 desaturase. Chemical Communications,401–402.
Burgoyne, T.W., Hayes, J.M., 1998. Quantitative production ofH2 by pyrolysis of gas chromatographic effluents. AnalyticalChemistry 70, 5136–5141.
Charon, L., Pale-Grosdemange, C., Rohmer, M., 1999. On thereduction steps in the mevalonate independent 2-C-meth-ylerythritol 4-phosphate (MEP) pathway for isoprenoidbiosynthesis in bacterium Zymomonas mobilis. TetrahedronLetters 40, 7231–7234.
Charon, L., Hoeffler, J.-F., Pale-Grosdemange, C., Lois, L.-M.,Campos, N., Boronat, A., Rohmer, M., 2000. Deuteriumlabeled isotopomers of 2-C-methylerythritol as tools for theelucidation of the 2-C-methylerythritol 4-phosphate for iso-prenoid biosynthesis. Biochemical Journal 40, 737–742.
Chikaraishi, Y., Naraoka, H., Poulson, S.R., 2004a. Carbon andhydrogen isotopic fractionation during lipid biosynthesis in ahigher plant (Cryptomeria japonica). Phytochemistry 65, 323–330.
Chikaraishi, Y., Naraoka, H., Poulson, S.R., 2004b. Hydrogenand carbon isotopic fractionations of lipid biosynthesisamong terrestrial (C3, C4 and CAM) and aquatic plants.Phytochemistry 65, 1369–1381.
Chikaraishi, Y., Suzuki, Y., Naraoka, H., 2004c. Hydrogenisotopic fractionation during desaturation and elongationassociated with polyunsaturated fatty acid biosynthesis inmarine macroalgae. Phytochemistry 65, 2293–2300.
Chikaraishi, Y., Matsumoto, K., Ogawa, N.O., Suga, H.,Kitazato, H., Ohkouchi, N., 2005. Hydrogen, carbon andnitrogen isotopic fractionations during chlorophyll biosyn-thesis in C3 plants. Phytochemistry 66, 911–920.
Chikaraishi, Y., 2006. Carbon and hydrogen isotopic composi-tion of sterols in natural marine brown and red macroalgaeand associated shellfish. Organic Geochemistry 37, 428–436.
Chu, I.M., Wheeler, M.A., Holmlund, C.E., 1972. Fatty acidmethyl esters in Tetrahymena pyriformis. Archives of Bio-chemistry and Biophysics 270, 18–22.
Douglas, A.G., Douraghi-Zadeh, K., Eglinton, G., 1969. Thefatty acids of the alga Botryococcus braunii. Phytochemistry 8,285–293.
Englebrecht, A.C., Sachs, J.P., 2005. Determination of sedimentprovenance at drift sites using hydrogen isotopes and unsat-uration ratios in alkenones. Geochimica et CosmochimicaActa 69, 4253–4265.
Epstein, S., Yapp, C.J., Hall, J.H., 1976. The determination ofthe D/H ratio of non-exchangeable hydrogen in celluloseextracted from aquatic and land plants. Earth and PlanetaryScience Letters 30, 241–254.
Estep, M.F., Hoering, T.C., 1980. Biogeochemistry of the stablehydrogen isotopes. Geochimica et Cosmochimica Acta 44,1197–1206.
Estep, M.F., Hoering, T.C., 1981. Stable hydrogen isotopesfractionation during autotrophic and mixotrophic growth ofmicroalgae. Plant Physiology 67, 474–477.
Fathipour, A., Schlender, K.K., Sell, H.M., 1967. The occurrenceof fatty acid methyl esters in the pollen of Zea mays.Biochimica et Biophysica Acta 144, 476–478.
Fukushima, K., Yasukawa, M., Muto, N., Uemura, T., Ishiwa-tari, R., 1992. Formation of C20 isoprenoid thiophenes inmodern sediments. Organic Geochemistry 18, 83–91.
Gelpi, E., Schneider, H., Mann, J., Oro, J., 1970. Hydrocarbonsof geochemical significance in microscopic algae. Phytochem-istry 9, 603–612.
Grossi, V., Baas, M., Schogt, N., Klein Breteler, W.C.M., DeLeeuw, J.W., Rontani, J.-F., 1996. Formation of phytadienesin the water column: myth or reality? Organic Geochemistry24, 833–839.
Harwood, J.L., 1988. Fatty acid metabolism. Annual Review ofPlant Physiology and Plant Molecular Biology 39, 101–138.
Harwood, J.L., 1997. Plant lipid metabolism. In: Mey, P.M.,Harbone, J.B. (Eds.), Plant Biochemistry. Academic Press,San Diego, pp. 237–272.
Hayes, J.M., 2001. Fractionation of the isotopes of carbon andhydrogen in biosynthetic processes. In: Valley, J.W., Cole,D.R. (Eds.), Stable Isotope Geochemistry. MineralogicalSociety of America, pp. 225–277.
Hilkert, A.W., Douthitt, C.B., Schluter, H.J., Brand, W.A., 1999.Isotope ratio monitoring gas chromatography/mass spec-trometry of D/H by high temperature conversion isotoperatio mass spectrometry. Rapid Communications in MassSpectrometry 13, 1226–1230.
Hoefs, J., 2004. Stable Isotope Geochemistry. Springer, Berlin,pp. 1–26.
Huang, Y., Shuman, B., Wang, Y., Webb, T., 2002. Hydrogenisotope ratios of palmitic acid in lacustrine sediments recordlate quaternary climate variations. Geology 30, 1103–1106.
Huang, Y.S., Shuman, B., Wang, Y., Webb, T., 2004. Hydrogenisotope ratios of individual lipids in lake sediments as noveltracers of climatic and environmental change: a surfacesediment test. Journal of Paleolimnology 31, 363–375.
Kirk, D.L., 1998. Volvox: Molecular-Genetic Origins of Multi-cellularity and Cellular Differentiation. Cambridge UniversityPress, Cambridge, pp. 16–44.
Lajtha, K., Marshall, J.D., 1994. Sources of variations in thestable isotopic composition of plants. In: Lajtha, K., Mich-ener, R.H. (Eds.), Stable Isotopes in Ecology and Environ-
Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608 607
mental Science. Blackwell Scientific Publications, Oxford, pp.1–21.
Laseter, J.L., Weete, J.D., 1971. Fatty acid ethyl esters ofRhizopus arrhizus. Science 172, 864–865.
Laseter, J.L., Weete, J.D., Weber, D.J., 1968. Alkanes, fatty acidmethyl esters, and free fatty acids in surface wax of Ustilago
maydis. Phytochemistry 7, 1177–1181.Lessire, R., Juguelin, H., Moreau, P., Cassagne, C., 1985.
Elongation of acyl-CoAs by microsomes from etiolated leekseedlings. Phytochemistry 24, 1187–1192.
Lichtenthaler, H.K., 1999. The 1-deoxy-D-xylulose-5-phosphatepathway of isoprenoid biosynthesis in plants. Annual Reviewof Plant Physiology and Plant Molecular Biology 50, 47–65.
de Ligny, C.L., 1976. The investigation of complex association bygas chromatography and related chromatographic and elec-trophilic methods. In: Giddings, J.C., Grushka, E., Cazes, J.,Brown, P.R. (Eds.), Advances in Chromatography, vol. 14.Marcel Dekker, New York, pp. 265–304.
Luo, Y-H., Sternberg, L.d.S.L., Suda, S., Kumazawa, S., Mitsui,A., 1991. Extremely low D/H ratios of photoproducedhydrogen by cyanobacteria. Plant and Cell Physiology 32,897–900.
Maxwell, J.R., Douglas, A.G., Eglinton, G., McCormick, A.,1968. The botryococcenes–hydrocarbons of novel structurefrom the alga Botryococccus braunii, Kutzing. Phytochemistry7, 2157–2171.
Metzger, P., Berkaloff, C., Casadevall, E., Coute, A., 1985a.Alkadiene- and botryococcenes-producing races of wild strainsof Botryococcus braunii. Phytochemistry 24, 2305–2312.
Metzger, P., Casadevall, E., Pouet, M.J., Pouet, Y., 1985b.Structures of some botryococcene: branched hydrocarbonsfrom the B-race of the green alga Botryococcus braunii.Phytochemistry 24, 2995–3002.
Metzger, P., Templier, J., Largeau, C., Casadevall, E., 1986. Ann-alkatriene and some n-alkadienes from the A race of thegreen alga Botryococcus braunii. Phytochemistry 25, 1869–1872.
Metzger, P., David, M., Casadevall, E., 1987. Biosynthesis oftriterpenoid hydrocarbons in the B-race of the green algaBotryococcus braunii. Sites of production and nature of themethylating agent. Phytochemistry 26, 129–134.
Metzger, P., Casadevall, E., Coute, A., 1988. Botryococcenedistribution in strains of the green alga Botryococcus braunii.Phytochemistry 27, 1383–1388.
Metzger, P., Largeau, C., 1999. Chemicals of Botryococcus
braunii. In: Cohen, Z. (Ed.), Chemicals from Microalgae.Taylor & Francis, London, pp. 205–260.
Ohlrogee, J.B., 1987. Biochemistry of plant acyl carrier proteins.In: Stumpf, P.K., Conn, E.E. (Eds.), The Biochemistry ofPlants, vol. 9. Academic Press, New York, pp. 137–158.
Okada, S., Devarenne, T.P., Murakami, M., Abe, H., Chappell,J., 2004. Characterization of botryococcene synthase enzymeactivity, a squalene synthase-like activity from the greenmicroalga Botryococcus braunii, Race B. Archives of Bio-chemistry and Biophysics 422, 110–118.
Peterson, D.C., 1980. Water permeation through the lipid bilayermembrane test of the liquid hydrocarbon model. Biochimicaet Biophysica Acta 600, 666–677.
Pollard, M.R., Mckeon, T., Gupta, L.M., Stumpf, P.K., 1979.Studies on biosynthesis of waxes by developing Jojoba seed.II. Demonstration of wax biosynthesis by cell-free homoge-nates. Lipids 14, 651–662.
Raven, P.H., Event, R.F., Eichhorn, S.E., 1999. Biology ofPlants. W.H. Freeman and Company Worth Publishers, NewYork, pp. 126–153.
Rohmer, M., 1999. The mevalonate-independent methylerythritol4-phosphate (MEP) pathway for isoprenoid biosynthesis,including carotenoids. Pure and Applied Chemistry 71, 2279–2284.
Sachse, D., Radke, J., Gleixner, G., 2004. Hydrogen isotoperatios of recent lacustrine sedimentary n-alkanes recordmodern climate variability. Geochimica et CosmochimicaActa 68, 4877–4889.
Saladin, T.A., Napier, E.A., 1967. Properties and metabolismof 2-alkylalkanoates. III. Absorption of methyl and ethyl2-methylpalmitate. Journal of Lipid Research 8, 342–349.
Sato, Y., Ito, Y., Okada, S., Murakami, M., Abe, H., 2003.Biosynthesis of the triterpenoids, botryococcenes and tetram-ethylsqualene in the B race of Botryococcus braunii via thenon-mevalonate pathway. Tetrahedron Letters 44, 7035–7037.
Sauer, P.E., Eglinon, T.I., Hayes, J.M., Schimmelmann, A.,Sessions, A., 2001. Compound-specific D/H ratios of lipidbiomarkers from sediment as a proxy for environmental andclimatic conditions. Geochimica et Cosmochimica Acta 65,213–222.
Schouten, S., Ossebaar, J., Schreiber, K., Kienhuis, M.V.M.,Langer, G., Benthien, A., Bijma, J., 2006. The effect oftemperature, salinity and growth rate on the stable hydrogenisotopic composition of long-chain alkenones produced byEmiliania huxleyi and Gephyrocapsa oceanica. Biogeosciences3, 113–119.
Schwender, J., Zeidler, J., Groner, R., Muller, C., Focke, M.,Braun, S., Lichtenthaler, F.W., Lichtenthaler, H.K., 1997.Incorporation of 1-deoxy-D-xylulose into isoprene and phytolby higher plants and algae. EFBS Letters 414, 129–144.
Schwender, J., Gemunden, C., Lichtenthaler, H.K., 2001. Chlo-rophyta exclusively use the 1-deoxyxylulose 5-phosphate/2-C-methylerythritol-4-phosphate pathway for the biosynthesis ofisoprenoids. Planta 212, 223–416.
Senousy, H.H., Beakes, G.W., Hack, E., 2004. Phylogeneticplacement of Botryococcus braunii (Trebouxiophyceae) andBotryococcus sudeticus isolate UTEX 2629 (Chlorophyceae).Journal of Phycology 40, 412–423.
Sessions, A.L., Burgoyne, T.W., Schimmelmann, A., Hayes,J.M., 1999. Fractionation of hydrogen isotopes in lipidbiosynthesis. Organic Geochemistry 30, 1193–1200.
Sessions, A.L., Hayes, J.M., 2005. Calculation of hydrogenisotopic fractionations in biogeochemical systems. Geochi-mica et Cosmochimica Acta 69, 593–597.
Sessions, A.L., 2006. Seasonal changes in D/H fractionationaccompanying lipid biosynthesis in Spartina alterniflora.Geochimica et Cosmochimica Acta 70, 2153–2162.
Sinninghe Damste, J.S., Rampen, S., Irene, W., Rupstra, C.,Abbas, B., Muyzer, G., Schouten, S., 2003. A diatomaceousorigin for long-chain diols and mid-chain hydroxy methylalkanoates widely occurring in Quaternary marine sediments:indicators for high nutrient conditions. Geochimica et Cos-mochimica Acta 67, 1339–1348.
Smith, B.N., Epstein, S., 1970. Biogeochemistry of the stableisotopes of hydrogen and carbon in salt marsh biota. PlantPhysiology 46, 738–742.
Stumpf, P.K., 1980. Biosynthesis of saturated and unsaturatedfatty acids. In: Stumpf, P.K., Conn, E.E. (Eds.), The
608 Z. Zhang, J.P. Sachs / Organic Geochemistry 38 (2007) 582–608
Biochemistry of Plants, vol. 4. Academic Press, New York,pp. 177–204.
Stumpf, P.K., 1987. Biosynthesis of saturated fatty acids. In:Stumpf, P.K., Conn, E.E. (Eds.), The Biochemistry of Plants,vol. 9. Academic Press, New York, pp. 121–136.
Szkopinska, A., 2000. Ubiquinone. Biosynthesis of quinone ringand its isoprenoid side chain. Intracellular localization. ActaBiochimica Polonica 47, 469–480.
Templier, J., Largeau, C., Casadevail, E., 1984. Mechanism ofnon-isoprenoid hydrocarbon biosynthesis in Botryococcus
braunii. Phytochemistry 23, 1017–1028.Templier, J., Largeau, C., Casadevail, E., 1991. Non-specific
elongation-decarboxylation in biosynthesis of cis and trans-alkadienes by Botryococcus braunii. Phytochemistry 30, 175–183.
Volkman, J.K., Maxwell, J.R., 1986. Acyclic isoprenoids asbiological markers. In: Johns, R.B. (Ed.), Biological Mark-ers in the Sedimentary Record. Elsevier, Amsterdam, pp. 1–42.
Wanke, M., Skorupinska-Tudek, K., Swiezewska, E., 2001.Isoprenoid biosynthesis via 1-deoxy-D-xylulose 5-phophate/2-C-methyl-D-erythritol 4-phosphate (DOXP/MEP) pathway.Acta Biochimica Polonica 48, 663–672.
Weete, J.D., 1976. Algal and fungal waxes. In: Kolattukudy, P.E.(Ed.), Chemistry and Biochemistry of Natural Waxes. Else-vier, Amsterdam, pp. 350–404.
Werstiuk, N.H., Ju, C., 1989. Protium–deuterium exchangeof benzo-substituted heterocycles in neutral D2O atelevated temperatures. Canadian Journal of Chemistry67, 812–815.
