.4):
....,
.
. s

, . 2m.
LE
. an}:

 

, . . 4‘
.3231.
2 I

. I. a. _. ..
tawny}...

“15::-

i. v

:......L

1.2.? A )A

 

5.2.} . .c.

3.7,". , , mmamun . . a w: .

.._. 1.

 

 

This is to certify that the

dissertation entitled

The Biosynthesis of
Dimethylsulfoniopropionate
(DMSP) in Marine Algae

presented by

Wayne Anthony Hicks

has been accepted towards fulfillment
of the requirements for

 

 

Ph .D . degree in Biochemistry

flaw/{41 66155:
aMajo/r p/rofessor
J A

pm //1 /¢
Date@(_fi%z-g 20a;

MS U is an Affirmative Action/Equal Opportunity Institution 042771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIFIC/DateDuo.p65-p.15

 

THE BIOSYNTHESIS OF DIMETHYLSULFONIOPROPIONATE (DMSP) IN
MARINE DINOFLAGELLATES

By

Wayne Anthony Hicks

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Biochemistry

2002

ABSTRACT

THE BIOSYNTHESIS OF DIMETHYLSULFONIPROPIONATE (DMSP) IN
MARINE DINOFLAGELLATES.

By

Wayne Anthony Hicks

Dimethylsulsoniopropionate (DMSP) is a member of a class of
compounds known as “compatible osmolytes”. These compounds accumulate to
high cytoplasmic concentration in response to water stress without incurring any
damaging effects to cellular structures, such as proteins. DMSP is produced by
a few land plants in diverse plant families. It is also produced and accumulated
by seven different classes of marine algae. DMSP has a number of biological
functions. It helps to regulate osmotic pressure, it serves as a cryoprotectant, and
it also functions as a defense compound as a precursor to acrylic acid.

DMSP is a precursor to acrylic acid through the action of the enzyme
DMSP Iyase, produced by marine bacteria and some marine algae. The Iyase
cleaves DMSP into DM8 and acrylic acid. DMS produced from DMSP is a major
fraction of the total amount of sulfur in the global sulfur cycle. Marine sources
contribute up to 40 metric tons/ year of DM8 to the environment according to one
source. DMS in the atmosphere forms cloud condensation nuclei, and thereby
contribute to cloud albedo and the cooling of the earth’s surface. DMS is
oxidized in the atmosphere to sulfuric acid and precipitates as acid rain.

Some land plants also synthesize DMSP. Previous research has shown

that they synthesize DMSP from methionine via the enzymatic sequence of

- methylation, deamination/decarboxylation, and oxidation. Research is
described in this thesis that describes the synthesis of DMSP in marine algae.
The pathway in marine algae has no steps in common with that of the higher
plant pathway. Algae also synthesize DMSP from methionine, however they do
so through the enzymatic sequence of transamination, reduction, methylation
and oxidative decarboxylation. This research has led to the identification of a
novel biological compound, 4-dimethylsulfoniohydroxybutyrate (DMSHB) which
may also function as a compatible osmolyte. Dinoflagellates are the largest
producers of DMSP among the algal classes that synthesize DMSP. Previous
research on DMSP synthesis in the dinoflagellate Crypthecodinium cohnii
suggests that decarboxylation is the first step. This pathway was tested and
results indicate that the DMSP biosynthetic pathway in dinoflagellates is
consistent with other marine algae that proceed with transamination as the first
step of the pathway. The effect of nitrogen on increasing DMSP production by

dinoflagellates is also described.

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................... vi
LIST OF FIGURES .................................................................................... vii
LISTOF ABBREVIATIONS ......................................................................... ix
CHAPTER I: LITERATURE REVIEW
Introduction .............................................................................. 1
Biological Distribution of osmolytes ......................................... 2
Biological Function of Osmolytes ............................................. 7
DMSP Biosynthesis in Higher Plants ..................................... 16
Enzymes of the DMSP Biosynthetic Pathway
in Higher Plants ...................................................................... 24
Early Work in Algal Biosynthesis of DMSP ............................. 27
Research Objectives ............................................................... 32

CHAPTER II: A NEW ROUTE FOR THE SYNTHESIS OF

CHAPTERIII:

DIMETHYLSULFONIOPROPIONATE IN

MARINE ALGAE.

Introduction ............................................................................. 33
Materials and Methods ........................................................... 36
Results .................................................................................... 39
Fate of Methionine Nitrogen in DMSP Biosynthesis ............... 48
Discussion ............................................................................... 50
Future Directions ..................................................................... 55
BIOSYNTHESIS OF DMSP IN THE MARINE
DINOFLAGELLATE HETEROCAPSA TRIOUETRA
Introduction .............................................................................. 57
Materials and Methods ............................................................. 61
Results ..................................................................................... 68
Feeding Experiments with [35S]Methionine .............................. 69
MTOB, MTHB, and MTP .......................................................... 7O
DMSHB and DMSP .................................................................. 76
Cold Trapping with MTHB and MTP ........................................ 80
Discussion ................................................................................ 82

CHAPTER IV: THE BIOSYNTHESIS OF DIMETHYLSULFONIO-

PROPIONATE IN TWO MARINE DINOFLAGELLATES:

CRYPTHECOHDINIUM COHNII AND LINGLODINIUM
POL YEDRA

Introduction ............................................................................... 86
Materials and Methods ............................................................. 88
Results ...................................................................................... 94
Methionine Uptake Experiments ............................................... 94
Glycine Betaine and DMSP in C. cohnii .................................. 100
[35$]Met feeding Experiments in C. cohnii ............................... 102

Culture Conditions for L. polyedra ........................................... 106
[358]Met Feeding Experiments in L. polyedra .......................... 109
Discussion ................................................................................ 1 1 1

Table I.

Table II.

Table III.

Table IV.

Table V.

List of Tables

Detection limit for DMSHB .......................................................... 44
[‘5N1Met incorporation into amino acids ..................................... 49
Efficacy of MTP and MTHB as cold traps ................................... 82

DMSP production by Crypthecodinium cohnii
grown in L1 medium .................................................................... 96

DMSP production by Crypthecodinium cohnii
grown in reduced nitrogen medium ............................................. 96

vi

Figure 1.

Figure 2.

Figure 3.

Figure 4.
Figure 5.
Figure 6.
Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

List of Figures

Marine Phytoplankton and the Sulfur Cycle ................................ 4
Comparison of DMSP Biosynthesis in Higher Plants and Marine
Algae ........................................................................................... 5
Pathway proposed by Uchida for DMSP biosynthesis in
Crypthecohdinium cohnii ............................................................. 6
Hofmeister Series ....................................................................... 10
Different possible routes for biosynthesis of DMSP ................... 18
SMM cycle .................................................................................. 21

Link between DMSP synthetic pathway and synthesis
of gonyol, and gonyauline ........................................................... 30

Method for derivatization of sulfonium ions for analysis

by GC-MS ................................................................................... 43
Mass spectra of DMSHB and DMSP from stable isotope

labeling studies in Enteromorpha intestine/is .............................. 46
FABMS spectrum of [U-‘3C]Met .................................................. 47

Fragmentation Pattern of lsobutylchlorofonnate isobutyl
ester of Gln .................................................................................. 50

GC-MS spectrum of glutamine from [‘5N1Met fed
Enteromorpha intestine/is ............................................................ 52

GCMS spectrum of glutamine from 15NH40H fed
Enteromorpha intestine/is ............................................................ 53

Linglodinium polyedra .................................................................. 61

Methionine uptake by Heterocapsa triquetra in
L1 media and reduced nitrogen L1 media ................................... 72

TLE autoradiogram of organic acids extracted from
Heterocapsa triquetra .................................................................. 73

vii

Figure 17. TLE autoradiogram of organic acids extracted from
Heterocapsa triquetra in ether ..................................................... 75

Figure 18. Graph showing 358 uptake into selected metabolites
of Heterocapsa triquetra .............................................................. 76

Figure 19. Graph showing the major fates of [35S]Met in
Heterocapsa triquetra ................................................................. 79

Figure 20. TLE autoradiogram showing bands of 358 labeled

DMSHB from Heterocapsa triquetra ........................................... 80
Figure 21. Growth curves of Crypthecohdinium cohnii in various

media .......................................................................................... 97
Figure 22. Graph of methionine uptake by Crypthecohdinium

cohnii in various media ............................................................... 99
Figure 23. FABMS spectra of Zooxanthellae zwitterions ............................ 101

Figure 24. FABMS spectra of Crypthecohdinium cohnii
zwitterions .................................................................................. 102

Figure 25. TLE autoradiogram of organic acids extracted from
Crypthecohdinium cohnii ............................................................ 104

Figure 26. 2-D TLC/T LE autoradiogram of zwitterions extracted
from Crypthecohdinium cohnii ................................................... 107

Figure 27. TLE autoradiogram of organic acids extracted from
Linglodinium polyedra ................................................................ 108

Figure 28. Graph showing 358 incorporation into sulfonium
compounds extracted from Linglodinium polyedra .................... 110

viii

AdoHcy
Ala

Asn
Asp
BADH
DDH
DMS
DMSHB
DMSKB
DMSP
DMSP-aid

DMSP-amine

FABMS
GCMS
gfw

Gln

Glu
HMT
Km
MCW
MDH
Met
MMT
MTA
MTHB
MTOB
MTP
MTP-amine
NAD
NADP
NMR
PLP
QAC
SAM
SIM
SMM
TBDMS
TLC
TLE
TMSCN
TSC

Vmax

List of Abbreviations

Adenosyl homocysteine

alanine

asparagine

aspartate

Betaine aldehyde dehydrogenase
3—dimethylsulfoniopropionaldehyde dehydrogenase
dimethyl sulfide
4-dimethylsulfonio-2-hydroxybutyrate
4-dimethylsulfonio-Z-ketobutyrate
3-dimethylsulfoniopropionate
3-dimethylsulfonioprcpionaldehyde
3-dimethylsufoniopropylamine

Fast Atom Bombardment Mass Spectrometry
Gas Chromatography-Mass Spectrometry
grams fresh weight

glutamine

glutamate

Homocysteine methyltransferase
Michaelis-Menten constant
methanolzchloroformzwater

Malate dehydrogenase

methionine

S-methyl methionine S-methyltransferase
methylthio adenosine
4-methylthio-2-hydroxybutyrate
4-methylthio-2-oxobutyrate
methylthiopropionate
methylthiopropylamine

nicotinamide adenine dinucleotide
nicotinamide adenine dinucleotide phosphate
Nuclear Magnetic Resonance

pyridoxal phosphate

quatenary ammonium compound
S-adenosyl methionine (S-AdoMet)
selected ion monitoring
S-methylmethionine

t-butlydimethylsilyl

thin layer chromatography

thin layer electrophoresis
trimethylsilylcyanide

tertiary sulfonium compound

maximum velocity of a reaction

CHAPTER 1
LITERATURE REVIEW

Introduction

Dimethylsulfoniopropionate (DMSP) is a member of a class of compounds
known as biologically compatible osmolytes that are accumulated by halophytic
or drought tolerant plants and some microorganisms. This class of compounds
includes zwitterionic tertiary sulfonium compounds (TSCs), and quaternary
ammonium compounds (QACs), as well as polyols such as glycerol, and
mannitol. Amino acids and amino acid derivatives are also known to provide
osmoprotective effects, particularly proline, and proline betaine, a QAC, which
have been shown to accumulate in some osmotically stressed plants (37). Some
members of this class of compounds are known to accumulate to cytoplasmic
concentrations of up to 1.4 M (36) without disrupting cellular structures. Similar
cytoplasmic concentrations of chaotropic salts would be deleterious to protein,
and membrane structure. High internal concentrations of compatible solutes
reduce osmotic stress in dry or saline conditions. DMSP and other biologically
compatible osmolytes therefore serve a very important role in plant adaptation to
the environment. Their accumulation to such high levels regulate osmotic
pressure in the presence of high external salt concentrations, and may also

protect some organisms from freezing damage at low temperatures (14, 53, 92).

The role DMSP plays in biological systems is important in understanding
how salt tolerant plants regulate their internal environment; however DMSP also
has important environmental implications as a major source of atmospheric
dimethyl sulfide (DMS). DMSP is a precursor of dimethylsulfide, which is
released to the atmosphere from ocean and coastal waters (1, 2, 8,11). DMS is
produced from catabolism of DMSP through the action of a DMSP Iyase found in
some marine bacteria and marine algae (89). The Iyase produces DMS and
acrylate from DMSP (11, 13).

(A) (CthS’CH2CH2COOH —-) (CH3)2S + CH2=CHCOOH

DMSP DMS Acrylate

The estimate of biogenic sulfur emission is thought to be in the range of 32 to
103Tg/yr, which is equivalent to about half of the atmospheric sulfur from
anthropogenic sources (1, 78, 43). The percentage of biogenic sulfur attributable
to DMS is about 45% (76). DMS has been implicated as a nucleation center for
aerosol and cloud formation and also in acid precipitation by its oxidation to
sulfuric acid (2,). Its role as a cloud condensation nucleus directly impacts
climate regulation patterns (Fig.1) (2,11). The ecological role of DMSP as a
major precursor of DMS has provided an impetus to uncover the biosynthetic

mechanism by which it is produced.

Biological Distribution of Osmolytes
Betaines, small nitrogen containing compounds in which the nitrogen atom

is fully methylated, occur in a wide range of higher plants, and even in

mammalian kidney (50). The most widely studied and distributed member of this
class of QACs is glycine betaine (4, 5, 7, 15, 37). Glycine betaine has also been
found to occur in trace amounts in some plants, which do not accumulate this
compound to osmotically significant quantities. Glycine betaine is thought to
have evolved early in plant evolution and the class angiosperrneae seem to
retain glycine betaine production as a primitive evolutionary characteristic (21 ). In
some members of the genus Limonium of the Plumbaginacae (35), and
Wollastonia biflora (Compositae)(68), the osmoregulatory functions of glycine
betaine have been replaced or supplemented by other QACs or TSCs.
Wollastonia is among the few flowering plants that accumulate
dimethylsulfoniopropionate.

DMSP occurs in seven classes of marine algae, but is found in only a few
unrelated land plant families. It has been isolated in Spartina altemiflora
(Poaceae), an intertidal grass species (51, 78), sugarcane (Gramineae) (65), and

as mentioned above, Wollastonia biflora (Asteraceae) (36, 79).

 

 

Marine Sulfur Cycle

    
     

nuclei CCN

 
 

sto,
sog-

 
 
  
 

 

 

5 Water
5 droplet
5 ems-cu, CH,S
_., cn,so-
_ cn,so,-
Climate
feedback Marine Atmosphere

 

 

 

 
 
  
  

 

 

3" 5::
\
\
\
\
\
\

DMS
Phytoplankton

zooplankton DMSO

 

Cysteine
methionine

 

Figure 1) The role of Marine phytoplankton in the Sulfur Cycle

 

Comparison of DMSP Biosynthetic Pathways
in Higher Plants and Marine Algae

 

Flowering Plants NHZ Marine Algae
MCOO'
MethylationH/SMQ \rransamination
Spartina Wollastonia MTOB A1000-
>SM SMM 2000-
D rbo I t' .
eca xy a Ion Oxidation Tl Reduction
T OH
>S/\/\NH2 Deamination + \S M C 00'
DMSP-amine Decarboxylatton MTHB
Methylation
Deamination ¢
OH
>DMSP-ald /S
DMSHB
' ' Oxidation +
Oxrdatk A/Decarboxylation
\S+/\/COO-
/
DMSP

Figure 2)Side by side view of known DMSP biosynthetic pathways
in higher plants and marine algae

DMSP Biosynthetic Pathway in
Crypthecohdinium cohnii

NH2
\sMcoo- Met

1 Decarboxylation
\SMNHZ MTP-amine

l Deamination
\SN COO‘ MTP

l Methylation

Figure 3) DMSP biosynthetic pathway proposed by Uchida et al.
for the dinoflagellate C. cohnii (86)

Biological Function of Osmolytes

Compatible osmolytes serve a variety of biological functions. Their role in
adaptation to water stress has been extensively demonstrated (6, 9,14, 15,16,
32, 35, 37). This includes changes in temperature, salinity, water availability, and
nitrogen levels. The term “compatible osmolytes” differentiates between those
molecules which lower water potential and maintain turgor pressure without
perturbing effects on other biological molecules, and those, such as inorganic
salts, which effect a perturbation of the system.

There are two ideas, not mutually exclusive, that are commonly invoked to
explain the mechanism by which compatible osmolytes function. Osmolytes may
stabilize protein structure by direct interaction with proteins or ligands, or through
solvent-solute interactions (6, 92). Timasheff offered that glycerol was able to
stabilize protein structure by re-ordering water structure in such a way that
enhances the structure of the solvent layer of the protein (24, 25). The other
theory is that they function in much the same way as the “stabilizing ions” of the
Hofmeister series. It is worth noting that the highly methylated nature of
betaines, such as glycine betaine, and trimethylamine N-oxide, closely resemble
the non-organic salt ammonium sulfate (Fig. 4).

Ammonium occupies a position as a stabilizing cation in the Hoffmeister
series, in which ions are ranked according to their affinity for protein
denaturation, or their ability to promote a “salting out” effect. The stnicture of

methylamines is very similar to that of quaternary ammonium ions. Yancey and

Somero suggest it is this similarity that is responsible for the stabilizing effect of
betaines on protein structure (92). Indeed, they showed that the effects of
compatible and non-compatible osmolytes were algebraically additive. This
phenomenon has been taken advantage of in nature as some species of urea
accumulating vertebrates (e.g. Xenopus Iaevis) maintain a 2:1 ratio of
methylamines: urea. This strategy allows them to adapt to osmotic stress while
offsetting the protein denaturation effects through the inclusion of methylamines

(7, 50,92)

c.2026 o_c9.oo_o .65 new 65826 5993 :0 Soto .65 5

8:3 LoEoELoI 9.: Co 2.2 65 2 5.25m 2m 55 8.38.08 .8826» oEow

Am ...So 95.8.. 6.262%. £295 oflfimficv 8 ...£ 9.53.. 6.3026 590.5
o~=3mfi 2 use. >9: 85cc; 2 mm 896.6 mctcm 3.2083: 9: Co ace. 23 p.59“.

.2326 Son o:_E<
azwlzwlwwx -ooolxolm
o o o ”3.2
_ _ _ +

I I I

+~um +30 +35. +3 +ao +oz
-mzo -_ Lm

E - Base
gazzsooo

+¥

Egg—2:25 no.5

01".. zf
" \
“£0 0"
u + / m
9.5 2 1
9.3.5::—
oESom 3.56.2 o:.En>5¢E_.F
m m
-o :o :o
/ N _ m o _lh
\\ol zonal :6 d4 :0
o £0 :6
9.3.32» .8 9.35:3:02
m
+412 +~zz£mzuv +2435: 25:8
.0896 -mow -moc L 235‘

:8 - Base
23.35

10

DMSP is a sulfonium compound, the structure of which closely resembles
that of the QACs. It has been shown to function in the same capacity as that of
the QACs (4, 5,14). In some plants, such as Spartina altemiflcra, and
Wollastonia biflora, DMSP and glycine betaine are both present. Some research
suggests that DMSP serves as a constitutive osmoticum is such systems, while
glycine betaine functions as a quick response compound whose concentration is
adjusted in response to changing conditions of salinity (9).

Colmer and Higashi authored a paper (9) in which they looked at the
effects of salinity, and nitrogen availability on the concentration of sugars, glycine
betaine, proline, and DMSP in Spartina altemr‘flora. They found elevated levels
of glycine betaine when plants were exposed to high concentrations of NaCI, and
nitrogen was provided at a concentration greater than 0.5 M. DMSP showed
decreased levels in response to the high NaCl, high nitrogen growth regimen.
However, when nitrogen levels were reduced below 0.5 M DMSP accumulated to
higher levels vs. control plants. The accumulation of glycine betaine was sharply
reduced compared to that seen for high salt and nitrogen > 0.5 M. In marine
environments where nitrogen levels are more limiting, DMSP plays more of a
primary role in osmotic adjustment.

Grone and Kirst conducted several studies in which DMSP levels were
monitored in response to a variety of environmental stressors (31, 32). In a 1992
paper they studied the effects of nitrogen availability, and the intracellular
methionine pool size on the DMSP content of Tetraselmis subcordifonnis (32).

(As will be discussed in detail below, earlier studies by Greene in 1962

11

established that DMSP was produced from methionine) (30). Cultures of
Tetraselmis grown in nitrogen deficient media showed a two-phase increase in
DMSP production. In the first 24 h. DMSP content/cell was increased by ~30°/o.
After 14 days DMSP content increased almost three-fold. This increase is DMSP
was accompanied by a decrease in total protein where cultures grown in nitrogen
deficient media showed a decrease from 52 pg/cell to 20 pg/cell. The onset of
protein reduction occurred within the first 48 h of culture in nitrogen deficient
media. When exogenous methionine was provided to cultures growing in a
complete medium a 3-fold increase in DMSP/cell was observed. Cultures were
also subjected to hyperosmotic shock, protease inhibitors were added, and
DMSP content was assayed. Control cultures, which were subjected to
hyperosmotic shock without the addition of protease inhibitors, underwent an
initial period in which DMSP content /cell was significantly reduced. This was
followed by a period of synthesis in which intracellular DMSP recovered to its
initial values. The cultures with added proteases did not recover to starting
values after the initial decline. This indicated that cells were able to synthesize
DMSP as a result of the increase of the free methionine pool resulting from
protein degradation.

DMSP’s suggested role as a cryoprotectant is naturally demonstrated as
polar macroalgae have a significantly higher DMSP content than their
counterparts from temperate or tropical regions. Chlorophyceaen species

collected from polar regions have an average DMSP content of

12

40 - 70 mmol kg'1 fresh weight, whereas species from temperate waters contain
25 — 30 mmol kg’1 fresh weight (43, 44). Several species of green macroalgae
collected from an Antarctic habitat were cultivated for one year at temperatures
of 0°C and 10°C. Those cultivated at 0°C showed a 2 — 5.5-fold increase In
intracellular concentrations of DMSP. The increase in DMSP production seems
to be linked to its ability to stabilize enzyme structure at reduced temperatures.
Malate dehydrogenase (MDH) was assayed at 30°C and at —2°C in the
presence of increasing amounts of DMSP. Enzyme activity measured at 0°C
was stimulated to a level of 165% of the control in the presence of DMSP
concentrations up to 300 mM; at 30°C just the opposite was found, increasing
DMSP concentration up to 300 mM lowered enzyme activity by up to 35%. This
cryoprotective response was not seen with the addition of proline or sucrose (45).
The ability of DMSP to promote enzyme activity in the presence of high
salt concentrations was tested using four enzymes isolated from Tetraselmis
subcordifomiis (31 ), glucose-6-phosphate dehydrogenase, glutamate
dehydrogenase, MDH, and a proteolytic enzyme. Of the four enzymes tested
only MDH showed a pronounced sensitivity to salt concentration. When MDH
was assayed in the presence of saturating substrate concentration,100 mM NaCI
produced a 35% decrease in activity, whereas the addition of the chloride salt of
DMSP (DMSP-CL) at a concentration of 100 mM did not have such an effect
(31). 200 mM NaCI, and 200 mM DMSP-Cl produced similar levels of inhibition.
Under substrate limiting conditions near the K.“ value for MDH, DMSP proved

much less inhibitory than NaCI. MDH activity in the presence of NaCI or

13

DMSP-Cl showed a decrease in Vmax and an increase in Km as is characteristic of
mixed inhibition. DMSP-CI, alone, lowered Vmax by an additional 7%.
There is a 2.5 fold increase in the rate of malate synthesis to oxaloacetate
synthesis under hypersaline conditions when DMSP is not present (31). The Km
of MDH for oxaloacetate in 100 mM Na Cl was four-fold higher versus a two-fold
increase in the presence of DMSP-Cl (31). Furthermore under hypersaline
conditions when the oxaloacetate hydrating activity of MDH is assayed with no
DMSP present; malate acts as an uncompetetive inhibitor by lowering both Vmax
and Km. At low oxaloacetate concentration malate even acts as an activator
rather than an inhibitor (31). When DMSP is present Km and Vmax are both
decreased, but the enzyme is still responsive to inhibition by malate. Increased
substrate levels can overcome inhibition of enzyme activity by NaCL. Overall, the
most significant effect of NaCl within a biochemical pathway, maybe its effect on
turnover balance. In this case, turnover balance is referred to as the rate of
malate formation to oxaloacetate formation (31). In the presence of high NaCI,
when oxaloacetate levels are low, malate becomes an activator rather than an
inhibitor. In the presence of DMSP, malate continues to act as an inhibitor and
the MDH turnover balance is maintained closer to that of control values. The
effect of DMSP in this system would be to maintain regulation of the pathway in
the presence of hypersaline conditions by preventing malate buildup (31).

The correlation of DMSP production with photosynthetic organisms
suggests a link (32). The effect of daylength, and light intensity was tested for its

effect on DMSP production (43). Five separate benthic species of Antarctic

14

Chlorophyceae were grown under differing lighting regimens of low, moderate
and high intensity, as well as long and short day lengths (18:6 light/dark cycle, or
6:18 light/dark cycle). The concentration of DMSP increased linearly with
increased light intensity, and with longer daylength. Three of the five species
tested also showed significant reduction of DMSP concentration when cultured in
darkness, although one showed such a reduction only after 170 days. This may
have been due to senescence of the culture. The link between DMSP
biosynthesis and photosynthetic processes may be due to the requirements of
DMSP biosynthetic enzymes for the reductants NADPH and ATP (37).

Acrylate produced from DMSP via DMSP Iyase has been shown to be an
effective chemical defense agent against herbivore grazing (88, 89).
Microzooplankton, which are the primary herbivores of marine phytoplankton,
have been shown to avoid DMSP accumulating species when offered a choice of
prey. DMSP-Iyase is produced by DMSP accumulating marine algae, but is kept
physically separate from internal DMSP. This is demonstrated in that
exponentially growing cultures produce little DMS, however stationary cultures
containing senescent cells produce significant amount of DMS (88, 89). When
DMSP containing phytoplankton are ingested by herbivores DMS emission is
increased as a result of the mixing of DMSP-Iyase and DMSP in the predator gut

(88, 89).

15

DMSP Biosynthesis in Higher Plants

The biosynthesis of DMSP in plants is a complex story. There are at least
two distinct pathways for synthesizing DMSP in higher plants. (Fig.2). In marine
algae one biosynthetic pathway has been established for the classes
Bacillariaphyceae, Chlorophyceae, Prasionophyceae, and Prymnesiophyceae
(Fig. 2) and a second has been postulated for Crypthecohdinium cohnii, a
heterotrophic dinoflagellate (Fig. 3). This means the pathway has evolved
independently at least three times. There are no other examples of such
biosynthetic diversity for a single naturally occurring compound. The biological
relevance of DMSP must be very significant, possibly because, unlike QACs it is
an effective compatible osmolyte that does not tie up significant amounts of a
limiting nutrient, nitrogen. It’s role as a defense compound subsequent to
catabolism by DMSP-Iyase may add to its biological value.

Prior to the work done in Wollastonia biflora, nothing was known of the
intermediates between the putative precursor methionine and DMSP. The
necessary steps to convert methionine to DMSP involve deamination,
methylation, decarboxylation and oxidation, however the sequence of these
steps was not known (Fig. 5). Hanson et. al. (36)used a strategy of radioisotope
labeling with [U-“C]Met in feeding studies. He also conducted feeding studies
with 1“C labeled putative precursors (36) S-methylmethionine (SMM) and
methylthiopropionate (MTP) to identify SMM as an intermediate in DMSP

biosynthesis. This work was complemented with cold trapping studies, stable

16

isotope labeling, and mass spectrometry to produce several lines of evidence,
which unequivocally demonstrated that methylation of methionine to produce
SMM was the first step in the DMSP biosynthetic pathway. While SMM is
ubiquitous in higher plants as part of the SMM cycle (62), in vivo double labeling
of the two methyl groups showed that SMM was directly in the DMSP pathway
and not first converted back to methionine.

In pulse chase experiments with [U-“C]Met ,SMM acquired and lost label
in a manner that is kinetically consistent with its role as an intermediate. When

[“CJSMM was provided; DMSP efficiently acquired label.

17

.EoEc 9:93:02 5 5365535: .335.
on. 28:35 Zeta oocmmo on... .EoSanm 3?QO c_ ooccoo 6353
9.: 9365 Soto >25... 05 mmzo .6 $3553 5. 852 62:88 22650 3 2:9“.

 

 

 

 

 

 

 

 

ogeodmso
\
0
$520 \ A 8.20 \
\ :0
9655 /
omsa m\ 000 w\ o 2.5 m\
-ooo\/\+ / A - J\/\+ / A litmuo ~/.\/\+ /
/ O llllllllllll :2
A ttttttttttt oc_Emdm2n_ \ g
co-
_ omsam\ A N:2 m\
oxo\/\+ / /\/\+ / cos.
~12

acct. Lose: 5 cmsa Lo £85585

18

When unlabeled SMM was used as a cold trapping pool in combination
with [u-“ciiuet flux of 14c label from [U-“ClMet into DMSP was diminished.
The evidence from these experiments was consistent with SMM as an
intermediate, but interpretation was complicated by the SMM cycle (Fig.6).

The SMM cycle, as described by Mudd and Datko (61, 62) showed that
Met and SMM are interconnected in the species Lemna paucicostata through
what was described as a futile cycling between the two compounds (Fig.6). In
this system of two interdependent, and opposing reactions; SMM is synthesized
from methionine and S-adenosylmethionine (S-AdoMet) to yield SMM and
S-adenosylhomocysteine (AdoHcy). This irreversible reaction is catalyzed by S-
methylmethioninc-methyltransferase. The opposing reaction, catalyzed by
homocysteine S-methyltransferase, utilizes SMM as a methyl donor to convert
homocysteine to Met.

The presence of the SMM cycle provides for the possibility that SMM
acquires label from [U-14C]Met not because it is an intermediate in DMSP
biosynthesis, but because it is connected to Met through the SMM cycle. To test
for this possibility a dose of racemic SMM, in which the methyl groups were
differentially labeled with 13CH3 and CzHa, was provided to leaf discs of W. biflora
(36). Analyses by FABMS showed that 93 - 96% of the DMSP came from the
doubly labeled SMM as evidenced by the DMSP ion at m/z 139, which indicates
a four mass unit increase above that of unlabeled DMSP MH+ ion. If SMM did
give rise to DMSP indirectly through the SMM cycle, the methyl carbons would

have been randomized. Met synthesized from homocysteine and SMM would

19

have a mixture of methyl groups labeled with ”CH3, ”CH3 -, C2H3, C2H3 -, 13CH3.
CzHa in a 1:2:1 ratio. This would have produced mass shifts of M+2, M+4 and

M+6 respectively.

20

ling,+

\S/\/kCOO'

Met
AdoMet

HMT MMT

HCY NH; AdoHcy

COO'

SMM

Figure 6) The SMM Cycle: MMT (methionine methyltransferase) uses Met and
AdoMet to form SMM and homocysteine. HMT (homocysteine
methyltransferase) catlyzes the opposing reaction to form Met from
homcysteine and SMM.

21

Hanson and co-workers used similar methods to demonstrate that 3-
dimethylsulfoniopropionaldehyde (DMSP-aid) was the next intermediate in the
DMSP pathway (42). [3°S]SMM was given to Wollastonia leaf discs and DMSP-
ald incorporated and lost label in parallel with the decline in specific activity of the
SMM pool. In addition unlabeled DMSP-aid was supplied as a trapping pool and
shown to be effective in decreasing flux of label from [3°S]SMM into DMSP. The
methylation of Met to form SMM, and oxidation of DMSP-aid to DMSP can be
explained by ”well established” biochemical reactions

The conversion of SMM to DMSP-aid requires two steps, transamination,
or oxidative deamination, and decarboxylation. DMSP-amine is a stable species
and would be a likely intermediate resulting from decarboxylation of SMM.
Hanson et al. never isolated a free intermediate. The possible transamination
product 4-Dimethylsulfonio-2-ketobutyrate (DMSKB) is an unstable species,
which spontaneously undergoes B, y—elimination to yield DMS, and
vinylglyoxalate (10). The absence of DMSP-amine in the labeling studies and
the inherent instability of DMSKB led Hanson and co-workers to propose that a
unique multi-enzyme complex catalyzes the SMM —+ DMSP-aid conversion in
which there are no free intermediates (68).

The studies in Wollastonia were extended to Spartina altemiflora (42).
Both plants were known to be DMSP accumulators, however they are
phylogenetically distant. In Spartina, SMM and DMSP also acquired label when
a pulse of [35$]Met was given. Unlike Wollastonia labeled DMSP-amine, was

detected. As discussed above DMSP-amine is the stable decarboxylated

22

product of SMM. When [35S]SMM was supplied DMSP-amine, acquired and lost
label in response to the depletion of radiolabeled SMM from the SMM pool. In
studies which used [3°S]DMSP-amine as a precursor the DMSP-aid became
labeled in a manner consistent with its role as a downstream intermediate in the
DMSP pathway. The labeling experiment with [35S]DMSP-amine provided one
additional surprise. Regardless of whether a large or small dose of precursor
was supplied; about 20% was metabolized to DMSP. This indicated a high
capacity for the metabolism of DMSP-amine. A metabolic model developed to fit
the data (42) suggests there are large storage pools of DMSP-amine and SMM,
and small metabolically active pools with a low flux rate between them compared
to rate of DMSP synthesis. Exogenously supplied DMSP-amine, therefore first
enters the storage pool prior to being metabolized to DMSP. The evidence
garnered from this experiment indicated that the DMSP synthetic pathway in
Spartina was distinct from that seen in Wollastonia sharing several steps, but

with a different enzyme yielding DMSP amine as a free intermediate.

23

Enzymes of the DMSP Biosynthetic pathway in Higher Plants

The S-adenosyl methionine (SAM) dependent methionine-S-
methyltransferase (MMT) was purified and characterized from W. bifiora (51 ).
This enzyme catalyzes the formation of SMM, which is the first committed
intermediate in DMSP biosynthesis in higher plants. MMT appears to be a novel
enzyme. It is a homotetramer of 110 KDa subunits. Most O and N-
methyltransferases are typically monomers or dimers with subunits in the range
of 20 — 45 KDa in size (21). Mammalian small molecule methyltransferases
involved in DNA modification are typically ~250 -300 residues in length (21).
The AdoMet binding domain for MMT is contained in its first 100 residues which
suggests it may contain a novel methyltransferase domain. The 750 residue C-
tenninus of the protein shows some similarity to starch and glycogen
phosphorylases. The highest region of similarity of MMT is to the pyridoxal
5’phosphate (PLP) domain of glucan phosphorylases. No other known methyl
transferase is PLP-dependent. In phosphorylase this domain includes the lysine
to which PLP is bound as a Schiff base; however, in MMT this lysine is absent. lf
PLP is present in MMT, it must be bound in a different position. Spectroscopic
studies reveal a small absorption peak at 330 nm and fluorescence at 550 nm,
similar to that of PLP-bound phosphorylase. It is important to emphasize that the
meanings of the unusual features of MMT have yet to be determined. Kinetic

analyses of MMT indicate that it has an ordered Bi Bi mechanism in which

24

AdoMet is the first substrate to bind and AdoHomocysteine (AdoHCy) is the last
product released.

The second step of the DMSP anabolic pathway in Wollastonia converts
SMM to DMSP-aid. This is a two step process requiring
deamination/transamination, and decarboxylation. The unstable nature of the 2-
keto acid intermediate DMSKB suggests two possibilities (10). One possibility is
that DMSKB only exists as a tightly bound intermediate for an unusual enzyme
capable of canying out both required steps. The second possibility is that the
two enzymes operate as a multifunctional complex in which the unstable 2-keto
acid intermediate is immediately channeled from one active site to the other.
Both possibilities represent unusual biochemical reactions. Stable isotope
labeling and GCMS experiments were carried out using 15N labeled Met to help
determine the enzyme mechanism (69). Amino acid pools were analyzed by
GCMS to ascertain the nature of the amino acceptor. MS analyses showed the
glu pool to be more highly labeled than gln or asn. This indicated a
transamination mechanism rather than oxidative-deamination, which would
generate free 15NH3. It would be expected that free ‘5NH3 would subsequently be
assimilated into the amide nitrogen of glutamine pools via the GS-GOGAT
(glutamine synthase—glutamate dehydrogenase, glutamate aminotransferase)
pathway.

There is not much information on the SMM -> DMSP-amine conversion or

the DMSP-amine -) DMSP-aid potential conversion. The final step in which

DMSP-aid is converted to DMSP is common to both pathways and some initial

25

characterization has been done. The oxidation of DMSP-aid to produce DMSP
has been shown to occur via 3-dimethylsulfoniopropionaldehyde dehydrogenase
(DDH) (83). This enzyme prefers NAD to NADP and is localized to chloroplast
stroma. It has a low Km for DMSP-ald of 1.5 pM, which is consistent with the
calculated small pool sizes of available substrate. DDH follows Michaelian
kinetics, and is subject to substrate inhibition. Other small aldehydes are also
capable of acting as inhibitors, including betaine aldehyde at concentrations
below 0.4 mM. In fact, this dehydrogenase is likely related to the
dehydrogenase involve in glycine betaine biosynthesis. Antiserum to betaine
aldehyde dehydrogenase (BADH) neutralizes DDH activity (83). DMSP-aid is
also a superior substrate for BADH than is betaine aldehyde. This strongly

suggests structural similarity between the two enzymes (21).

26

Early work on Algal Biosynthesis of DMSP

In 1962 Ronald Greene published the first paper on the synthesis of
DMSP in marine algae (30). His study used Ulva lactuca, a macro-algae of the
class Chlorophyceae, to establish that DMSP was synthesized from methionine.
In several rounds of labeling Greene provided cultures of algae with [3°S]Met,
[‘4c11Met, and [“C2]Met and analyzed the alkaline hydrolyzed products of
DMSP (DMS and acrylate) for their incorporation of label from methionine. This
work established that the carbon skeleton of DMSP originated from methionine,
which undenlvent methylation, deamination, decarboxylation, and oxidation to
form the penultimate product. Greene also used [“CHalMet to show that the
methyl donor for the formation of DMSP was also methionine. In an attempt to
Identify some of the intermediates in the pathway he used [35$]SMM in the same
type of feeding experiments and found no labeling of DMSP resulting from SMM.

Uchida et al, has done previous research on DMSP biosynthesis in
dinoflagellates. The work of Uchida et al. was done using (86) Crypthecodinium
cohnii, a heterotrophic dinoflagellate. (86) This is an interesting case, because
biosynthesis of DMSP has typically been linked to photosynthesis (17, 37, 43,
48). Uchida et al. also used [35S]Met in feeding studies to determine
intermediates in the pathway; however, they did very long term labeling (1 week).
The feeding studies were done in the dark, whereas others have always been
done under illumination. The authors suggest a different pathway for synthesis

of DMSP in C. cohnii, in which the first step is decarboxylation (Fig. 3).

27

Their scheme is based on research which investigated the ability of
putative intermediates to effect a diminution in flux of the 358 label from [35$]Met
into DMSP. The results showed that 4-methyl-2-oxobutyrate (MTOB), which
would be the product if the first step in the pathway were deamination of Met, and
SMM were both ineffective as trapping pools. In contrast, methylthiopropionate
(MTP) was a very effective as a trapping pool. This compound could be formed
by several routes, but they extrapolated that decarboxylation was the first step in
the DMSP pathway based on the purification of a methionine decarboxylase from
C. cohnii. Methionine decarboxylase has previously been shown to be present in
algae and plants that do not accumulate DMSP (66). Without direct evidence
Uchida et al. proposed that methylthiopropylamine (MTP-amine) is the first
intermediate in the DMSP pathway, because it is the product of methionine
decarboxylation. The research by Uchida and co-workers while provocative
doesn’t make a convincing case for MTP and MTP-amine as intermediates.
There were no direct feeding studies reported that showed these compounds
exhibited the type of labeling kinetics that would be expected of an intermediate.
The presence of a decarboxylase that uses methionine as a substrate doesn’t
argue convincingly either, as amino acid decarboxylases are widespread (72, I
86).

Nakamura and colleagues have done some interesting work with the
autotrophic dinoflagellate Linglodinium polyedra. They sought to establish a link

between the biosynthesis of DMSP, and two other sulfonium compounds, gonyol,

28

and gonyauline (Fig. 7). Nakamura used a combination of stable isotope labeling
and NMR in his investigation. He fed [methyl-13C11Met and [methyl-‘3C1]SMM

and determined that the methyl groups in gonyauline come from Met (63).

29

.638 new 65.3963 8 £95556 2 cozooccoo
gamma @5265 coma cctmE 5 5353 255585 5.65. C. 2:9".

amsfi
w\
.000\/\... / I
\\\. I],
.963 R cc=am>com
m\ \
-OOO\Un\V/\+ / -OOO/Dv\+m/
4, t!
// mrmsa \\\
IO
mI.:2 m0.:2 5.2

.000 m IV boo m IV .08 m/
/

IO 12

Eboéoq .4 5 2355605 ems—Q

30

Using [1-‘°C]Met they found no 13C in gonyauline. He also fed [U-‘3CIDMSP and
found that the C1-C3 positions and the methyl groups were incorporated into
gonyauline. ln gonyol however only the methyl groups and the CS-C5 positions
labeled. This suggested gonyauline was synthesized from Met, but the carbon
atom, which occupies the C1 position, was different in both gonyauline and
gonyol when labeled Met was used as a precursor. Nakamura et al. proposed
that the C1 and CZ atoms of gonyol come from acetate. This was supported by
feeding studies with [1 ,2-13C]AcOH, which showed that 13C was incorporated
into the C1 and CZ positions of gonyol. They concluded that gonyol and
gonyauline were synthesized from DMSP, however the C1 of gonyauline
originates from 002 and the C1 and C2 positions of gonyol are derived from
acetate. Their statement that the C1 position of gonyauline comes from 002 is
based on the increase of 13C incorporation at the C1 position when NaH1°C03
was added to the medium. This increase was only seen when the organisms
were cultured in the light, not in the dark. This led them to the conclusion that the
incorporation of C02 was photosynthesis dependent. The conclusions drawn
from the work of Nakamura et al. require additional confirmation as the long term
nature (7days) of his experiments allows for the possibility of scrambling of

carbon atoms from the supplied precursors into other intermediates.

31

Research Objectives

The work described in the following chapters is directed toward elucidating
the DMSP biosynthetic pathway in marine algae. This will be done using a
strategy of isotope labeling with radioactive and stable isotopes, thin layer
chromatography, Thin layer electrophoresis, mass spectrometry, and
autoradiography In a similar means to that used in defining DMSP biosynthesis in
higher plants. This work was first done in species of four diverse classes of
marine algae, Enteromorpha intestinalis (Chorophyceae) Tetraselmis
subcordifcnnis (Prasinophyceae) Melosira nummuloides (Bacillariaphyceae) and
Emiliana huxleyi (Prymnesiophyceae) and then extended to dinoflagellates
because they accumulate more DMSP and produce more DMS than marine
algae of other classes. Heterocapsa triquetra is used as a model organism for
dinoflagellates, but DMSP biosynthesis in Crypthecodinium cohnii, and
Linglodinium polyedra is also investigated because of the unusual features

highlighted by previous research done in these organisms.

32

CHAPTER II

A NEW ROUTE FOR THE SYNTHESIS OF

DIMETHYLSULFONIOPROPIONATE IN MARINE ALGAE

Acknowledgement
The work contained in this chapter was previously published in the journal Nature ‘
as Gage, D. A., Rhodes, 0., Nolte, K. D., Hicks, W. A. Leustek, T., Cooper, A. J.
L., and Hanson, A. D., (1997) A new route for synthesis of

dimethylsulfoniopropionate in marine algae. Nature, 387, 891-894

Introduction

As discussed in Chapter 1 the biosynthesis of DMSP has been
established in higher plants. Flowering plants synthesize DMSP through the
sequential methylation, decarboxylation, transamination, and oxidation of
methionine. There are two variants of the pathway found in Spartl'na and
Wollastonia. Algae also synthesize DMSP from methionine, but until recently
none of the intermediate steps were known, only that the carbon skeleton and
sulfur of DMSP were incorporated intact into DMSP (31).

DMSP synthesis in marine algae has now been shown to share no
common steps with the pathway in higher plants. There were some early studies
on the synthesis of DMSP in the dinoflagellate Linglodinium polyedra by

Nakamura et al, (63). They used radiolabeling methods to show that the sulfur,

33

CZ, and C3 atoms of methionine are incorporated into DMSP. This was
consistent with Greene’s result in Ulva Iactuca that DMSP is derived from
methionine.

Uchida et al also conducted some early experiments in the dinoflagellates.
They used the heterotrophic species Crypthecodinium cohnii. The evidence they ,
obtained suggesting the distinct pathway Met —> MTP-amine —> MTP -> DMSP
was based on indirect evidence. The conversion of Met to MTP-amine, which is
the first step of their pathway, was based on the purification of a methionine
decarboxylase from C. cohnii. Methionine decarboxylases are however present
in many other species which do not produce DMSP (72, 77, 87).

At the time I joined the lab experiments were underway to define the route
for biosynthesis of DMSP in Enteromorpha intestinalis, a macroalgae of the class
Chlorophyceae, Melosira nummulcides, a diatom, Tetraselmis, a Prasinophyte,
and Emiliana Huxleyi, a prymnesiophyte. Methods of in vivo isotope labeling
were combined with chromatography, electrophoresis, and mass spectrometry to
elucidate the biosynthetic pathway of DMSP in these diverse marine algae.

From this work the pathway has now been shown to synthesize DMSP from the
sequential transamination, reduction, methylation, and oxidative -
decarboxylation of methionine to form DMSP. This pathway includes the
biosynthesis of the novel sulfonium compound 4-dimethylsulfonio-2-
hydroxybutyrate (DMSHB). The intermediate DMSHB has also been
demonstrated to serve many of the same biological functions as DMSP. This

project was a collaborative effort between the laboratories of Andrew Hanson,

Univ. of Florida, Tom Leustek, Rutgers Univ. Arthur Cooper, Cornell Univ. David
Rhodes, Purdue University, and Doug Gage. A collaborative paper was
published in the journal, Nature (20) that described the pathway elucidation and
included my analytical studies presented below.

My specific role, described herein, was quantitative and qualitative
analysis of metabolites isolated from stable isotope feeding studies. The
objective of my work was to define the enzymatic mechanisms involved in two of
the biosynthetic steps: the deamination of methionine and the
decarboxylation/oxidation of DMSHB. New analytical methods were developed

as a part of this project.

35

Materials and Methods

Materials. Andrew Hanson, Florida State University, provided DMSHB
that was used as a standard for mass spectrometric measurements, zwitterionic
fractions isolated from Enteromorpha that contained DMSHB and DMSP, and
DMSHB and DMSP that was labeled with stable isotopes. David Rhodes,
Purdue University, provided amino acid samples that were labeled with 15N and
isolated from Enteromorpha. All other chemicals and reagents were purchased
from commercial sources.

Derivatization of sulfonium compounds. The sulfonium compounds
isolated from the zwitterionic fractions of Enteromorpha intestinalis were
derivatized to their tert-butyl dimethylsilyl (TBDMS) esters for analysis by GC-
MS. The zwitterionic fractions of algal lysates were lyophilized to complete
dryness. The dried sample was then further dried overnight in a vacuum
dessicator in the presence of P205 in order to remove all traces of water. A
derivatization cocktail was prepared by mixing 20 uL N~(tert-butyldimethylsilyl)—
N-methyltrifluoroacetamide with 2 pl of 10% (TMSCN) in chloroform, 4ul 10%
trimethylpyridine In chloroform, and 2 uL 10% dimethoxypropane in chloroform.
The above reactants were mixed in the above order. 25 pL of the above
derivatization cocktail was added to the dried sample. The sample and cocktail
were incubated for 2 h at 60°C.

Gas Chromatography-Mass Spectrometry of TBDMS derivatives of
sulfonium compounds. Mass spectrometric data was obtained using a JEOL

AX-505H double focusing mass spectrometer coupled to a Hewlett-Packard

36

5890J gas chromatograph. Gas chromatography was acquired using a 30 m DB1
column (0.25mm ID and 0.25pm film coating). The column was procured from
J&W Scientific. Direct (splitless) injection was used. Helium gas flow was
approximately 20 ml/min. GC conditions were as follows: Injector temperature
260°C, initial temperature 100°C, initial time 2 min, rate 20°C/min, final
temperature 310 °C, final time 0 min MS conditions were as follows: separator
temperature 285 °C, ion source temperature 200°C, electron energy 30 eV, scan
rate 45-500 a.m.u. at 0.8 scans/sec. DMSHB was derivatized as its
t-butyldimethylsilyl ester/ether and analyzed by GC-MS with selected ion
monitoring after on-column nucleophile assisted S-demethylation. Authentic
DMSHB was used to calibrate the selected ion monitoring parameters. The
diagnostic fragment ion clusters at mass/charge ratio (mlz) 321, due to loss of a
t-butyl radical, was monitored in the SIM mode with peak centering at 8’26”,
which corresponds to the appropriate retention time. The SIM parameters for 13C
labeled DMSHB were programmed to monitor ions from 321.2 mlz - 327.2 mlz in
increments of 1, with peak centering at 8’25”.

Derivatization of ”W labeled amino acids. 100 mg batches of
Enteromorpha intestine/is were lncubated for 1, 2, and 4h with 1 umole of
[1°N]ammonium chloride of with 5 pmole [‘leMet in 2.5 ml of seawater. These
were extracted using a standard methanol-chloroform-water extraction
procedure. I analyzed samples of the neutral amino acid fraction from [‘5N]Met
feeding studies that were provided by David Rhodes. The samples were

derivatized for analyses by mass spectrometry using 60 ul of a 60:30:10 mixture

37

of water-isobutanol-pyridine, 10 pl of isobutylchlorofonnate and 60 pl of
chloroform.

Gas chromatography-mass spectrometric analyses of Glutamine and
Asparagine. GC-MS data was acquired as described above except that the GC
conditions were as follows: 280°C, initial temperature 100°C, initial time 2 min,
rate 10°C/min, final temperature 300°C. MS conditions were as follows:
separator temperature 300°C, ion source temperature 200°C, electron energy 30

eV, scan rate 0.8 scans/sec over a mass range of 45-500 a.m.u.

38

Results

In order to determine the biosynthetic intermediates in the marine algae
Enteromorpha intestinalis, large scale labeling studies were carried out in the
laboratory of Andrew Hanson at the University of Florida with the assistance of
Doug Gage, Michigan State University, David Rhodes, Purdue University,
Thomas Leustek, Rutgers University, and Arthur Cooper, Cornell University.
Andrew Hanson coordinated the project, and oversaw the labeling studies with all
of the algae except those done with Tetraselmis, which were conducted by David
Rhodes. David Rhodes also developed a computer model to analyze the kinetics
of metabolite interconversion in the DMSP pathway. Isolation and mass
spectrometric analyses of acidic, and basic amino acid fractions from [‘5N]Met
and 15NH40H fed Enteromorpha was also done by David Rhodes. I performed
GC-MS analyses of DMSHB samples and analyzed FAB-MS spectra of DMSP
isolated from [1°C]Met fed Enteromorpha. I also performed GC-MS analyses of
neutral amino acid fractions isolated from Enteromorpha that had been fed
[‘5N1Met or 1°NH4OH.

100 mg fronds of algae were incubated with a 500 pmol dose of [3°S]Met.
The metabolic fates of the Met were followed using thin layer chromatography
(TLC) and thin layer electrophoresis (TLE). The major fates of Met were
incorporation into protein and conversion to DMSP. Two compounds labeled in
a manner that is consistent with the expected kinetics for an intermediate. These

two compounds were 4-methylthio-2—hydroxybutyrate (MTHB) and its

39

S-methylated product dimethylsulfonio-Z-hydroxybutyrate (DMSHB). The most
likely route from Met to MTHB is via the relatively unstable ot-keto acid , 4-
methylthio-Z-oxobutyrate (MTOB), which can undergo spontaneous
decarboxylation to form methylthiopropionate (MTP). An alternative gentle
extraction method was employed to quantify MTOB. MTOB was also converted
to a stable derivative, MTHB, with sodium borohydride. Comparisons of reduced
and unreduced samples were used to estimate MTOB pool size. Both methods
showed that MTOB acquired and lost 358 label in parallel with MTHB. MTP that
was produced as a result of the spontaneous decarboxylation of MTOB could
also be estimated based on the difference between reduced and unreduced
samples of MTOB. Analysis of MTP levels by this procedure shows that a
significant amount, but not all of the MTP which labels in the [°°S]Met labeling
experiment is derived from MTOB (personal communication from Andrew
Hanson).

Compounds other than MTOB, MTHB, and DMSHB which were
considered likely intermediates based on other studies of DMSP biosynthesis, or
because of known biological reactions were also monitored for uptake of label
from [35S]Met. Two of the known intermediates in higher plant DMSP
biosynthesis, S-methylmethionine (SMM) and 3—dimethylsulfoniopropionaldehyde
(DMSP-aid) (36, 42) showed little or no uptake of label. MTP-amine, which was
proposed to be an intermediate of DMSP biosynthesis in dinoflagellates by
Uchida et al (86), slowly accumulated label as a minor end product. MTP, .

another compound from the pathway suggested by Uchida et al., labeled in a

40

variable manner, from undetectable to a level commensurate with that of MTOB.
This was shown in repeated experiments with different lots of E. intestinalis.
The data from MTP can be explained based on a catabolic route involving
oxidative decarboxylation of MTOB to MTP, which occurs in other plants and
algae that do not accumulate DMSP (12, 58, 66). This same pathway is also
present in animals and varies with nutritional status.

The labeling data acquired from [3°S]Met labeling is consistent with the
pathway Met —> MTOB —+ MTHB —) DMSHB -—> DMSP. This pathway was
further supported when 353 labeled versions of the putative intermediates, SMM,
MTHB and DMSHB were supplied. Very little SMM was metabolized. Taken
together with the data from the [35S]Met labeling studies SMM does not appear to
be an intermediate in the algal pathway. MTHB was mainly converted to Met,
which was then converted to protein, and a small amount was also found in
DMSP. These data are consistent with the conversion of Met —-) MT HB through
a freely reversible transamination reaction. DMSHB was metabolized to DMSP.
The identification of intermediates in the 35S labeling studies was based on co-
migration with standards during TLC and TLE. In order to further support the
data; experiments using stable isotope labeling were conducted. E. intestinalis
was supplied [U-‘°C]Met in atmospheres containing 1"02 or 1802. The latter was
used to Investigate the enzymatic reactions involved, since incorporation of 1"0
into DMSP would provide evidence if an oxygenase was involved in the formation
of DMSP from DMSHB (20). The zwitterionic fractions were isolated and these

samples were submitted to me for analyses by GC-MS, and fast atom

41

bombardment mass spectrometry (FABMS). These analyses were done as a
means to verify the presence of DMSHB, which is a novel biological compound,
and to test if an oxygenase reaction is involved in its conversion to DMSP.
Analysis of the sulfonium compound, DMSHB, required the development
of a new technique, because this metabolite is found in low levels and is difficult
to completely separate from other compounds. DMSHB was converted to its
TBDMS-derivative and analyzed by GCMS in order to take advantage of the
separation capability and sensitivity of the method. Trimethylsllylcyanide
(TMSCN) was included in the derivatization mixture as a source of CN' ions,
which promoted the S-demethylation of DMSHB (Fig. 8). As salts, sulfonium
compounds would not normally be amenable to analysis by GCMS because of
their fixed charge and polar nature. This novel approach for analysis of
sulfonium compounds was developed from a similar approach used to analyze
acylcamitines (39). The derivatization and demethylation procedure greatly
enhances the volatility of sulfonium compounds, and makes their analyses by
GCMS possible (19, 39). I used single ion monitoring (SIM) for the GCMS
analyses. This was shown to be very sensitive with a 10 pmol detection limit
(Table I). The S-demethylated DMSHB-TBDMS derivative had a retention time of
8 min 26 s. The M-57 peak at mlz 321 (loss of a t-butyl group) was diagnostic for

the presence of DMSHB.

42

CH3

\
/S

CH3

GHQ

 

i

 

OH

I
-cn2— CH2 — CH — coo-

 

 

1. Dry over P205

2. N-(tert-butyldimethylsilyl)-N-methyl

trifluoroacetamide
+ 2,2-dimethoxypropane
+ trimethylsilylcyanide

 

_l,_

-5;—

I
O

+ I I.
/S-CH2 — CH2 - CH - COO-SI
CH3. l

 

 

Inject reaction mixture directly into GC

-CN ions from TMSCN act as nucleohiles
in the heated injection port resulting in
the S-demthylation of the TBDMS derivative.

 

.+

——Si—

I
O

+ I I,
CH3- s‘CHz '— CH2 "' CH — COO—Sit

Volatile S-demethylated product

Figure 8) Method for the dervatization of DMSHB for analysis by GCMS

43

 

 

 

Detection Limit for DMSHB

 

 

 

 

 

 

 

 

 

 

SampLe Retention time Peak area
800 pmol DMSHB + 1 - Z 8 min. 27 s 20.142
pmol DMSP

200 pmol DMSHB + 1 - 2 8 min. 27 5 1.0691
pmol DMSP

40 pmol DMSHB + 1 — 2 8 min. 27 s 0.2252
pmol DMSP

20 pmol DMSHB 8 min. 26 5 7.8909

10 pmol DMSHB 8 min. 25 3. 0.7460

10 pmol DMSHB * 8 min. 25 3 6.2817

 

Table I) GCMS data acquired in the SIM mode for the S-demethylated TBDMS
esters of DMSHB.

(Note: The first three entries in the table are samples from ion-exchange purified
extracts of Enteromorpha that was spiked with the 20, 5, and 1 nmols of DMSHB
respectively. The amount of DMSHB entered in the table is the amount injected
onto the GC column assuming 100% recovery from the ion exchange columns.
The last entry in the table (*) was taken with the collection slit opened to 250 this
increased the signal by a factor of 8.37.)

 

The E. intestinalis samples that had been supplied with [U-13C]Met in chambers
with atmospheres containing 1602 or 1802, showed a significant peaks at mlz 326,
which correspond to the M+5 peak of the M-57 peak of the S-demethylated
DMSHB-TBDMS ion (Fig.9). Because the analyte is S-demethylated, the sixth
carbon is not directly detected. Nevertheless, we can infer that the entire carbon
skeleton of DMSHB ls derived from Met, which is consistent with Greene’s earlier
result in Ulva Iactua (30). Met is the likely methyl donor in the methylation of
MTHB to form DMSHB (via S-AdoMet). There was no indication of 18O
incorporation into DMSHB, which suggests that oxygens in the hydroxyl and
carboxyl groups are not derived from molecular oxygen

Because DMSP is an abundant metabolite, a less sensitive MS technique,
fast atom bombardment mass spectrometry could be used for its analysis.
Derivatization is not required for this technique. FABMS of DMSP samples from
the [U-13CjMet feeding study done in an 1602 atmosphere showed significant
M+4 and M+5 peaks (Fig. 9). The M+4 peak is likely the result of the
[1°C4]homocysteine moiety. Homocysteine is recycled back to Met after it has
been used in methylation reactions as AdoMet, with the resulting formation and
release of AdoHcy. GCMS and FABMS (Fig.10) confirmed the presence of the
recycled Met. Modeling studies indicated the [‘3C4JDMSP/[130510MSP ratio
agreed with expected values for the recycling of [‘3C4]homocysteine into DMSP.
GCMS data of DMSHB formed from [U-‘°C]Met in an 180 atmosphere show no
significant increase in the M+2 peak indicating the a-I’leI’OXYI group does not

originate from 02. The FABMS spectrum of DMSP resulting from E. intestinalis

45

treated with [U-‘aClMet I180 shows strong signals at m/z 141 (13C4,‘°O) and 142

(1305, 180) indicating that 30-40% of the DMSP formed became labeled with 18o.

46

A B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DMSHB DMSP
2.0 1 CONTROL 30 I CONTROL
20 r
1.0 «
l 10 I
0.0 u I E r 1 I 1 Lu JUL J! A I ; I’ fit
‘73
2.o - [‘3C]Metli°Oz g 30« [‘3C1Met/1502
El 2
a ' a
g E,”
Q. - | l E 10-I
°-° In . . . . . Jr g ALLA-H4 : A
2.0 - [‘3ClMet11302 2° ‘ [13CIM9U 1”02
1.0 - 10 .
321 322 323 324325 326 327 328 135136 137138139140 141 142
mlz ' mlz
Figure 9) Mass Spectral evidence for DMSHB and for its conversion

to DMSP by an oxygenase reaction. E. intestinalis fronds
(100 mg fresh weight) were incubated for 24 h with
L-[U-‘3C]Met (5 pmol) In a flask Initially containing 21%

1602 or 1802 in N2; controls received no Met and no 1802.

A) SIM analysis of DMSHB. The peak areas in endogenous
DMSHB was replaced by 13C4-and 13C5-labelled species

the S-demethylatedderivatives of [‘3-041- originate from 02.
B) FABMS analysis of DMSP. In the [1"C]Met/“‘O2 treatment,
the peaks at mlz 141 and 142 ([1304, 18O]-and [1305, 1aO]DMSP)
show that 30-40% of the 1‘ic-labelled molecules contain an
18O atom.

47

Photosynthetically generated 1602 likely lowered this percentage. The
information gathered from the 13C and 18O labeling data indicate that an
oxygenase reaction is involved in the oxidative decarboxylation of DMSHB to

DMSP.

Fate of Methionine Nitrogen in DMSP Biosynthesis

Oxidative deamination or transamination are both plausible mechanisms
for the deamination of Met to form MTOB. Stable isotope labeling was used to
differentiate between the two mechanisms. If MTOB were formed via oxidative
deamination, free‘sNHg from the a—‘SNHa of [‘5N]Met would have been
generated. Incorporation of free 15NH3 into the amide nitrogen of Gln through the
action of glutamine synthase would then be expected. In the case of
transamination the a-‘5NH3 of methionine would have been transferred directly to

a-ketoglutarate, oxaloacetate, or pyruvate. This would result in the formation of

15N labeled Glu, Asp and Ala respectively (69). 5 pmol of [‘SN]Met was
exogenously supplied to 100 mg of E. intestinalis and GCMS was used to
examine the pattern of labeling in the amino acid pool. Glu (7.0% abundance at
2 hrs), Asp, and Ala acquired ‘SN rapidly (Table II). This is consistent with a
transamination mechanism.

The neutral amino acid fraction was isolated, and I converted them to their

N-butylcarbonyl isobutyl esters prior to analyses by GCMS (Fig. 11)(40, 88).

48

 

nmol 15N amino acid/gfw (rounded to the nearest nmol)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

amio T = 0 h Met 1 Met 2 h Met 4 h NH4+ NH4+ NH4+ 4
acid h 1 h 2 h h

Ala O 13 11 20 137 150 84

Met n.d. 4,374 5,815 8,965 n.d. n.d. n.d.

Asn 0 1 2 7 29 61 68

Gln 0 15 13 25 702 812 612

Asp 0 6 11 12 26 41 49

Glu 0 51 77 119 283 383 374

 

Table II) 80 mg fresh weight samples of E. intestinalis were fed 5 pmol
[‘SNJMet or 1 pmol 15NH... The amino acids were isolated and
derivatized to their N(O,S)—heptafluorobutyryl isobutyl esters for
analyses by GCMS. Quantitation was based on 15N abundance
and predetermined pool sizes.
(Personal communication from David Rhodes, Purdue University)

49

 

56 he .Smo_b:nom_ oumgofiegzebanom. ho Egan 5295808“. A: 2:9“.

 

 

 

 

wand—c A\
ammoozz )V. o
w Ms...» A .I
ammoomrz- .
Sm- NE. a N
3
ammoozz ~12 m 8sz 002 I

 

 

A‘DmNOO .1 SmNoo
zMC . r. _

+

 

50

The abundance of the M+1 peak of the ion containing the amide nitrogen at m/z
84, for Gln, and m/z 69, for Asn, was compared to the value for the naturally
occurring isotopes in order to determine isotopic enrichment. The isotopic
enrichment was less than 1% at 2 h. The low level of the isotopic enrichment in
the Gln amide nitrogen is also consistent with transamination of Met, but not with
oxidative deamination (Fig.12). A mechanism involving oxidative deamination
would have resulted in the labeling of the amide nitrogen of Gln via the operation
of glutamine synthetase. The presence of glutamine synthetase in E. intestinalis
was verified by supplying the algae with a 1 pmole/100 mg dose of “"NH4+
(Fig.13). The amide nitrogen of Gln showed a 42% increase in abundance at 2h.

Several lines of evidence therefore indicated that the first step of DMSP
biosynthesis in E. intestinalis is transamination, as opposed to oxidative
deamination. The above stable isotope labeling data are consistent with
transamination, but not oxidative deamination. The conversion of [35$]MTHB to
Met also indicates the reversibility of the Met to MTOB step, which is also

consistent with transamination.

Discussion
Taken collectively the data above establish a pathway for the
synthesis of DMSP in marine algae for the first time. The findings were extended
to other species of marine algae, which are known to be large'producers of

DMSP.

51

assesses .m .8. 8.2.2.: Ea... sass...
So; N .50 so :2 E0893 955350 2 oEEm .o 5.50ch 9200 Am? 2:9".

 

 

 

 

 

 

 

mod Udfi-Dfll 13.03513 0500

52

mam§mm§ .m 8:33 4:2.“P Ea: 0383
So: N .50 co co_ EoEmg 9:58:00 2 mEEm cc 8288.» £200 A? 239....

 

$—
3....
V
l"‘-q '
g3...
l
g. .
’0

 

 

 

 

 

 

mar-r 0+5-30 CDSCUUCUN

53

Emiliana huxleyi (Prymnesiophyceae) and Melosira nummulcides
(Bacillariophyceae) both contained small pools of DMSHB, which acquired and
lost label in parallel to a dose of supplied [35$]Met. Both also metabolized a dose
of [35S]DMSHB to [35S]DMSP. It is very likely that they share the same pathway.

This newly established pathway for the biosynthesis of DMSP shares no
common steps with the pathway in higher plants. This indicates that the DMSP
pathway has evolved independently at least twice. The pathway proposed by
Uchida et al. for the heterotrophic dinoflagellate, C. cohnii, presents the intriguing
possibility for yet another biological means for the synthesis of the same
molecule.

The transamination step at the beginning of the algal pathway for the first
time provides a possible explanation for the observation of increased DMSP
synthesis in response to nitrogen limitation. Mobilization of the nitrogen in a-
amino groups via amino acids provides a supply for the synthesis of other
nitrogenous metabolites. This may explain the 50% reduction in protein content
of Tetraselmis cultured in nitrogen reduced medium seen by Grone and Kirst
(32). Methionine transport and its subsequent deamination would provide a
ready supply of reduced nitrogen. The resulting a-keto acid , MTOB, can then be
shunted through the DMSP synthetic pathway.

DMSHB was also identified as a novel sulfonium compound, and a
possible substrate for DMSP Iyase. Some preliminary evidence in feeding
studies indicates significant catabolism of DMSHB supplied to marine algae. It is

therefore another possible contributor to DMS emissions. Understanding of the

synthesis of DMSP in marine algae and its regulatory mechanism should
contribute to an improved insight to biogenic sulfur emissions and its impact on
our environment.

Some preliminary isolation and characterization of the first three enzymes
in the DMSP biosynthetic pathway has been done (80). The methionine
aminotransferase at the head of the DMSP pathway has been shown to be non-
enantioselective with respect to substrate methionine. It has a preference for
MTOB or 2-oxoglutarate as the amino acceptor and for L-Glu as the amino donor
for the reverse reaction. All algae would be expected to show some methionine
aminotransferase activity as a constituent of the Met salvage pathway (12). In E.
intestinalis this activity was 30-100 times higher than in non-DMSP accumulating
species (80).

The reductase that catalyzes the MTOB -> MTHB conversion is NADH
linked, and appears to prefer NADH to NADPH. It also appears to be
enantiospecific in its conversion of MTOB to MTHB, producing exclusively the D-
form of MTHB. The MTOB reductase is also a speculative point of connection in
linking DMSP synthesis to photosynthesis, because of its need for NADH, whose
levels are photosynthesis dependent.

The formation of the sulfonium compound DMSHB in E. intestinalis is an
AdoMet dependent methyltransferase. Studies conducted in the Hanson
laboratory show it has a preference for D-MTHB (80). It has not been shown to
have any MTP methylating activity. This enzyme also appears to be a novel

methyltransferase whose presence is exclusive to DMSP accumulating algae.

55

The enzyme data are intriguing and could be diagnostic for characterizing
some of the DMSP producing classes of marine algae whose pathways have not
yet been rigorously characterized.

The unique nature of the sulfonium compound DMSHB, and its formative
enzyme, MTHB methyltransferase could be used as a single point of data to
define the DMSP pathway in other species. The heterotrophic dinoflagellate,
Crypthecodinium cohnii, could provide a good point of comparison. The
presence, or lack of DMSHB could indicate whether the pathway defined for E.
intestinalis is operative, or as a heterotrophic species whether it has evolved

another means for DMSP synthesis.

Future Directions

The newly identified pathway presented in this chapter represents a
significant step in our understanding of the biosynthetic origin of DMSP. The
identification of the novel sulfonium compound DMSHB provides for the
possibility that it may have osmoregulatory properties. Comparison of its
osmoregulatory properties, if any, with DMSP will provide another variable, which
must be considered when investigating osmotic adjustment in DMSP
accumulating organisms. Furthermore, what are the phylogenetic ramifications
of its discovery? If DMSHB is an effective osmolyte, the question must be asked
if there are organisms that have evolved a mechanism to produce DMSHB and
not DMSP. We must also consider DMSHB as a new possible source of DMS

emissions to the atmosphere.

56

The methyltransferase that catalyzes the formation of SMM from Met in
the DMSP biosynthetic pathway of higher plants has been shown to have some
unique characteristics among methyltransferases (41 ). The MTHB
methyltransferase in the marine algal pathway offers an interesting target for
study, as it catalyzes the formation of a similar sulfonium product. The data
herein establishes the pathway in four of the seven classes of marine algae that
produce DMSP (Chlorophyceae, Prasinophyceae, Bacillariaphyceae, and
Prymnesiophyceae). The dinoflagellates however, as a class accumulate more
DMSP and produce more DMS than any of the other classes (48). The
possibility of another distinct pathway for the synthesis of DMSP in C. cohnii is
presented in the paper by Uchida et al. These are some of the compelling
reasons to extend the research tools developed in the elucidation of DMSP
synthesis in E. intestinalis, to DMSP biosynthesis in dinoflagellates. This will
resolve what is now an outstanding biochemical question. Do dinoflagellates

have their own unique pathway for the biosynthesis of DMSP?

57

Chapter III

BIOSYNTHESIS OF DMSP IN THE MARINE DINOFLAGELLATE.
HE TEROCAPSA TRIQUETRA.

Introduction

Dinoflagellates have long generated scientific interest as the culprit in
producing “red tides”. These massive algal blooms, which kill large numbers of
fish every year, are produced by a variety of species. Some of the neurotoxins
produced by dinoflagellates. such as the neurotoxic shellfish poison, saxitoxin,
are far more potent than curare (76). Dinoflagellates produce the well-known
paralytic shellfish poisoning (PSP). PSP is accumulated by mollusk through
filtering, and accumulates in mollusk tissue. The accumulation of these toxins
can render many mollusks used as food sources inedible at particular times of
the year when dinoflagellate populations are high. In areas where dinoflagellate
populations are consistently high shellfish can never be eaten.

Dinoflagellates are the largest producers of DMSP among the seven
classes of marine algae that accumulate DMSP. DMSP concentrations among
these organisms show interspecies variation. Some of these species which
produce large amounts of DMSP are also red tide organisms. Linglodinium
polyedra (Fig. 14) (formerly known as Gonyaulax polyedra) is a red tide organism
that produces ciguatoxins. Exposure to ciguatoxin is seldom fatal, but their

effects can be long term and severe (76). There are even anecdotes of

58

fishermen becoming ill simply by breathing the air in the vicinity of large algal
blooms. Understanding the physiology of dinoflagellates is therefore very
important both from an economic and public health standpoint.

The marine algal synthetic pathway for DMSP is completely different from
that of higher plants (Chapter II). This points to independent evolution of the
pathway in the two different phyla.

DMSP biosynthesis is usually linked to photosynthesis as its production
has been shown to be light dependent (43). As discussed in Chapter 1 the
marine dinoflagellate Crypthecodinium cohnii is a heterotrophic dinoflagellate that
produces DMSP in a non-light dependent manner. In a paper by Uchida et al. a
synthetic route for DMSP was proposed, which is distinct from the pathway in
higher plants or other marine algae. The evidence for this pathway however,
was somewhat equivocal. Dinoflagellates therefore present the intriguing
possibility of a third or even fourth route for the synthesis of DMSP.

Linglodinium polyedra is a photosynthetic dinoflagellate that produces two
sulfonium compounds, whose structures are very similar to DMSP. Nakamura et
al. (63) showed that the synthesis of these two compounds, gonyol and
gonyauline, were linked to DMSP.

Heterocapsa triquetra was selected as a model organism for the study of
DMSP biosynthesis in dinoflagellates. because it is known to accumulate DMSP
and it grows well in laboratory conditions. There is the added benefit that H.
triquetra, unlike L. polyedra, is not known to produce any sulfonium compounds

other than DMSP.

59

This study utilizes a strategy of labeling with radioactive and stable isotopes to
establish the pathway of DMSP biosynthesis in Heterocapsa triquetra, a

photosynthetic dinoflagellate.

60

 

Figure 14) Linglodinium polyedra is an armored dinoflagellate.

This particular example has lost its flagella. Normally there would

be one flagella within the transverse funow, which is the deep groove
encircling the cell, and one flagella in the longitudinal furrow, which

is the groove dividing the cell into right and left halves. The armored
appearance is composed of a polysaccharide and microfibrils. It is actually
interior to the cell and is covered by a cell membrane.

61

Materials and Methods

Chemicals. Racemic MTHB was purchased from Sigma and was
recrystallized from aqueous ethanol. SMM-Cl (SMM chloride salt), MTP, DMSP-
CI (DMSP chloride salt) and MTP- amine were purchased from TCI America.
MTOB was purchased from Sigma.

Radiochemicals. L-[35S]SMM was synthesized by treating L-[3°S]Met with
250 mM methanol in 6 M HCI at 110°C for 4 hrs. L-[358]DMSHB was synthesized
using a modification of the method of Zappia 8 Schenk) (93) by treating 7 nmol
L-[35$]SMM with 9 nmol H2304 in 60 pl H20 and 5 pl glacial acetic acid. 10 pl of
NaNOz was added dropwise to the reaction vial kept on ice. The reaction was
incubated at 22°C for 3 hrs. The product was isolated by ion exchange
chromatography using 2 columns arranged in tandem. The first column was 2 ml
of mixed bed resin (2:1 AG1 "OH, and BR70 H‘). The second was an AG50-X8
(H‘) column. The sample was applied to the mixed bed column and washed with
2 ml of H20. The product was eluted from the second column with 5 ml of 2.5 N
HCI. The column eluate was lyophilized to dryness, and resuspended in double
distilled (dd) H20 to give the desired final concentration. Product purity was
ascertained by co-migration with standard during TLE according to the protocol
for TLE of sulfonium compounds in the following section. [35$]Met 43.5 TBq/nmol
(NEG-009A) was purchased from NEN Life Science products. It was treated with
4% thioglycollic acid at 95°C for 4 h to reduce Met sulfoxide. Basic impurities

were removed by treating with AG50 (NHI) resin.

62

Acidic impurities were removed with AG1 (COO‘) resin. Racemic MTHB was
purchased from Sigma and was recrystallized from aqueous ethanol.

Growth conditions for Heterocapsa triquetra. H. triquetra was obtained
from The Center for Culture of Marine Phytoplankton (CCMP449) and cultured
axenically in L1 medium at 18°C and a 14:10 light :dark regime with a photon
fluence of 105 uE m'2 s". A daughter culture was made by diluting an aliquot of
the stock culture into reduced nitrogen L1 medium in a 1:3 ratio. This was
passaged on the fourth day of culturing by removing 1/3 of the total volume and
replacing with an equal volume of the reduced nitrogen medium. The final
nitrogen concentration after the first passage was 0.26 mM. Cell populations
were calculated using a hemacytometer with a 0.1 mm cell depth.

Labeling with [35$]Met. Two labeling experiments were done. Labeling of
cultures grown in reduced nitrogen medium were used on the seventh day. Nine
12 ml subcultures were prepared from the reduced nitrogen culture. The nine
subcultures corresponded to harvest time points of 0, 5, 10, 30, 60, 120, 240 and
1200 min. Cells were used in exponential growth phase at a population density of
3.85 x 104 cells/ ml. Cultures were inoculated with 19 uCi (0.4 nmol [35$]Met)
and incubated on a ZD rocker at a low to moderate speed. Labeling experiments
with cells grown in a complete L1 medium were done with four 12 ml subcultures
made from a stock in exponential growth at a cell density of 4.86 x 104 cells /ml.
Cultures were inoculated with 8 uCi (0.4 nmol [35$]Met). Lighting and
temperature conditions were identical to those used in culturing. One culture

was used as a 0 time point (control). We centrifuged the culture in order to pellet

63

the cells. The supernatant was removed and a dose of [358]Met was added that
was equivalent to that added to the other cultures. An aliquot was removed and
assayed by scintillation counting. This was done to provide a baseline that could
be used to monitor uptake of radioactivity by the other cultures. The amount of
radioactivity remaining in the supernatant was assayed, and the percent uptake
was calculated using the value obtained from the control culture as a baseline
value.

Extraction of metabolites. Algal cultures were gently centrifuged after
labeling at 4300 rpm using an S4180 rotor in a Beckman GS15R centrifuge. The
pellets were frozen in liquid nitrogen. 1.0 pmol of MTP, MTOB, MTHB, SMM,
MTP-amine, DMSP, and DMSHB were added as carriers. l.0 ml of 0.1 N HCI
was added. The pellet was vortexed’ and centrifuged at 4300 rpm. The lysate
was fractionated into acidic, basic, and zwitterionic components using a series of
three columns AG1 (-OH), Bio-Rex 70 (H+), and AGSO-XB arranged in tandem.
Cell lysates were washed through the three columns in series. Columns were
eluted individually to isolate acidic, basic, and zwitterionic fractions.

Column eluates were lyophilized, resuspended in 300 pl of 60% methanol and
lyophilized in a vacuum concentrator. The residues were suspended in a
minimal volume of water for application to TLC plates.

Isolation of metabolites. MTP, MTOB, and MTHB and Met were isolated
using a 1.5 ml column packed with AG1 - X8 ('OH) 200 - 400 mesh resin. Cell
lysates were applied in a 1 ml volume. The column was washed with 15 ml dd

H20, and eluted with 5.0 ml 2.5 N HCI. SMM and MTP-amine were isolated by

passing the entire wash fraction of the AG1 column through a 1 ml column
packed with Bio-Rex 70 (*H) 100 - 200 mesh resin. The column was washed
with 15 ml of dd H20 and eluted with 15 ml of 1.0 N HCI.

DMSHB and DMSP were isolated by passing the entire wash fraction from the
Bio-Rex 70 column through a 1.0ml column packed with AG50W-X8 (H‘)

200 - 400 mesh resin. The column was washed with 5.0 ml of dd H20 and
eluted with 5.0 ml of 2.5 N HCI.

TLE of cellular metabolites. 35S labeled metabolites were identified by
co-migration with authentic standards during thin layer electrophoresis.

TLE was done using a Hunter thin layer peptide mapping system. Organic acids,
amino acids, and bases were separated by TLE in a 1:1 :38 mixture of
pyridine/acetic acid/H20 for 20 min. at 1.8 kV. Sulfonium compounds were
separated in 1.5 N formic acid at 1.8 W for 20 min.

Identification of metabolites. Metabolites isolated from column fractions
were spotted onto TLC plates in individual lanes. Authentic standards of
expected metabolites within the column fractions were spotted in adjacent lanes.
After TLE an autoradiogram was prepared. The radiolabeled bands were
localized, excised from the plate, and counted by scintillation. The lane
containing the authentic standards was then stained with a reagent specific to the
type of metabolites isolated from the column fractions. Dragendorff’s reagent
was used to react with and positively identify sulfonium compounds, and QACs
(DMSHB and DMSP) (20, 36) isolated from the AG50-X8 column. Ninhydrin was

used to identify basic compounds (SMM, MTP amine) from the Bio-Rex 70

65

column (20, 36). lodoplatinate was used for organic acids from the AG1 column
(KTMB, MTP, MTHB) (20, 36). The mobilities of the standards could then be
compared to the bands from the autoradiogram, and the individual metabolites
subsequently identified.

Identification of DMSHB. TLE of the sulfonium compounds from AG50-X8
fractions produced an intense band corresponding to DMSP, that obscured the
zone of DMSHB migration. The zone corresponding to DMSHB was scraped
from the plate and eluted using 3 x 300 ml of 80% methanol. The samples were
lyophilized and reapplied to TLC plates for separation by TLE.

Base destruction of DMSP. Half of the 3"’8 labeled sulfonium fractions
were reserved. To these fractions 0.5 pmol and 0.5 pmol of DMSHB were
added. The samples were then treated with 17% NaOH. The tubes were left
open to air in a hood for evolution of [3°S]DMS. The samples were neutralized
using 1.7 vols (relative to NaOH) of 2.5 N HCI. The samples were diluted to 0.5
ml and applied to a two column series consisting of a 1 ml mixed bed (2:1 AG1 ’
OH and Bio-Rex 70 H”) column, and a 0.5 ml AGSO-XB (H+) column respectively.
Both columns were washed through with 2 ml of H20, and the A650 column was
washed with an additional 5 ml of H20. The A650 column was eluted with 5 ml
of 2.5 N HCI. The eluate was lyophilized, suspended in 300 ml of 60% methanol,
and lyophilized again in a vacuum concentrator. The residues were resuspended
in 5 pl of H20 and applied to a 0.1 mm glass backed cellulose plate for TLE.

Quantification of radioactivity. The amount of radioactivity recovered from

columns was assayed by adding a small aliquot to 5 ml of Ready Gel scintillation

66

fluid and assaying by scintillation counting. Radioactivity recovered from plates
was calculated by scraping radioactive bands, and adding the scrapings to 2 ml
of H20. This was mixed with 2 ml of Ready Gel scintillation fluid and assayed by
scintillation counting.

Autoradiography. Subsequent to separation by TLE, the
electrophoretograms were enclosed in an X-ray film cassette with one sheet of X-
ray film. Exposure times varied according to the amount of radioactivity present
on the TLE plates.

Quantitation of radioactivity in protein and lipids. 35$ incorporation into
protein and lipid fractions was assayed by a modification of the method described
by Mudd and Datko (1986). Extracted pellets were washed with 1 ml of 0.1 N
HCI, vortexed and centrifuged. The wash was discarded. The pellets were then
extracted with 2 x 1 ml of methanol:chloroform:water (12:5:1). The pellet was
vortexed and centrifuged at 4300 rpm using an S 4180 rotor in a Beckman
GS15R centrifuge. The supematants were combined and phases broken by the
addition of 0.5 ml CHCI3 and 0.75 ml H20. The pellet (containing the lipid free
protein fraction) was suspended in 1 ml of H20 and vortexed. A 100 ml aliquot
was removed and added to a 2 ml of H20. 2 ml of Ready gel scintillation fluid
was added and the mixture was vortexed. The mixture was assayed by
scintillation. Efficiency correction was carried out by adding a known amount of
[358 Net and calculating an efficiency factor for each sample. 353 incorporation

into lipids was assayed by taking 200 pl of the organic phase and adding it to 4

67

ml of Ready gel scintillation fluid. The mixture was counted using a scintillation
counter.

Cold trapping studies. A six-day-old culture of H. triquetra was subdivided
into seven 12 ml subcultures. Three were used as controls. One of these three
was used as a 0 time point. This culture was centrifuged to pellet the cells. 8
pCi of [35$]Met was added to the supernatant. A 0.5 ml aliquot was added to 5
ml of Ready Gel scintillation fluid and assayed. The other six were used
painrvise to provide a 30 min. time point and a 4 h time point. The control

cultures had only 8 pCi [35$]Met added. The other four cultures had either MTP

or MTHB added 1 min. prior to the addition of an 8 pCi dose of [35$]Met.

68

Results

A small dose of [3°S]Met was supplied to 12 ml cultures of axenic,
exponentially growing cultures of H. triquetra. The fate of the 35S label was
followed into downstream metabolites, and into end products. The major fate of
[35$]Met was incorporation into protein. DMSP also accumulated significant
amounts of 358. Several compounds, which have been identified or suggested
as intermediates in other organisms that accumulate DMSP, also acquired label.

In agreement with studies done on DMSP biosynthesis in the chlorophyte
alga Enteromorpha intestinalis, 4-methylthio-2-oxobutyrate (MTOB), 4-
methylthio-Z-hydroxybutyrate (MTHB), and 4-dimethylsulfonio-2-hydroxybutyrate
all acquired label. MTOB and MTHB labeled heavily at early time points, and
levels diminished as cellular pools of free [35$]Met were depleted. This pattern
of labeling is consistent with what would be expected of an early intermediate in
the pathway. DMSHB acquired label at a constant rate in parallel with DMSP.
This may indicate that there are storage pools of DMSHB, which accumulated
label slowly in conjunction with the increase of intracellular DMSP levels.
MTP also labeled quite heavily, and paralleled the pattern seen in MTOB.

S-methylmethionine (SMM), the first intermediate in the plant DMSP
biosynthetic pathway, was not detected. We also did not detect the presence of
3-methylthiopropylamine (MTP-amine). Uchida et al. suggested MTP-amine was
a metabolite of the DMSP pathway in the heterotrophic dinoflagellate
Crypthecodinium cohnii, although no direct evidence for its presence was

provided. (86).

69

The data from the [35$]Met labeling studies favor the pathway Met —>
MTOB —-> MTHB —> DMSHB —> DMSP previously established in E. intestinalis
(20). The pathway proposed by Uchida et al. is not operative in this
dinoflagellate, because we failed to find any MTP-amine. However, the labeling
pattern of MTP makes it difficult to rule this out as a potential intermediate. The
data could be interpreted to support the pathway Met —9 MTOB -> MTP -+
DMSP. In this scenario either MTP or MTHB would be a substrate for the
methyltransferase, which catalyzes the formation of the sulfonium product
DMSP, in the case of the former, and DMSHB in the case of the latter. I
therefore complemented the [35$]Met labeling experiment with a cold trapping
experiment. Either MTP or MTHB was provided as a cold trap in conjunction with
[35$]Met to test for diminution of 35S flux from [35$]Met into the sulfonium products

of the two possible pathways.

Feeding Experiments with [”SlMethionine

A culture of Heterocapsa triquetra was grown in reduced nitrogen L1
medium for a week. This was done to maximize uptake of tracer Met, and
production of DMSP. Met uptake was increased two-fold in the reduced nitrogen
medium, as compared to similar feeding experiments in unmodified L1 medium
(Fig. 15). Experiments with another dinoflagellate, C. cohnii, indicated that
DMSP production was increased two-fold on a per molecule methionine basis.
The combined effect was a five-fold increase in DMSP production at 2h (See

Chapter 4.) This is consistent with the frequently observed inverse correlation

70

between nitrogen availability and DMSP production (9, 32, 49). The mother
culture was subdivided into eight 12 ml cultures. Each culture was inoculated
with 19 pCi of [358]Met. Each of the eight cultures was harvested at eight
different time points of O, 5, 10, 30, 60, 120, 240, and 1200 minutes. Cells from
the harvested cultures were disrupted and the soluble components fractionated
using a series of 3 ion exchange columns. The column eluate was separated by
TLE, and an autoradiogram was made to evaluate the components that

incorporated 35$.

MTOB, MTHB and MTP
The autoradiogram of the TLE plate from the organic acid containing AG1
fraction showed bands corresponding to MTOB, MTHB, MTP and Met. There
was an unknown band that appears at 120, and 240 minutes and migrates
intermediate to MTHB and MTOB (Fig. 16). When the bands were excised from
the TLE plate, carryover from this band skewed the quantitation for MTOB, and

MTHB upwards.

71

 

 

nmol Met uptake by H. triquetra in L1 media

 

0.45 i

0.3

 

O O
(.0 «b

025 - 11.—"”3‘aanigil

C
N

0.15 ~

0
_L

0.05 -

nmol Met remaining In
medium

 

 

 

0 20 40 60 80 100 120
rnfiunes

 

 

 

 

nmol Met uptake by H. triquetra in low N L1

 

0.45

O
Dias:
0001-5

1A

025* - "'_._"'f"""i;mo.ii.r;

 

O
N

0.1-5 4

0.05 a

o
_s

 

 

nmol Met remaining In
medium

 

0 20 40 60 80 100 120
minutes

 

 

Figure 15) Graphs depict the nmol of Met remaining in the
supernatant after cells were provided with exogenous methionine
and centrifuged. A) Uptake of met by cells cultured in L1 media.

B) Uptake of met by cells grown in L1 where nitrogen concentration

was reduced ten-fold.

72

5.2

8:625 5.2

at).

mI._.S_

mot).

.61 2 rd 5 0909on 88356380331 E0: upon 23sz 8? 8:9“.

m o _. on 00 our ovm comp

 

l_j'l‘V

 

 

moSEE

73

I, therefore carried out another [35S]Met feeding experiment and this time the
organic acids were extracted from the acidified cell lysate using ether. The ether
extracted fraction, and the acidic fraction were analyzed by TLE. The ether
extract contained MTOB, but not the unknown band (Fig.17). This showed that
radioactivity incorporated into MTOB was drastically reduced after 60 min.

MTOB is somewhat unstable and can undergo spontaneous
decarboxylation to form MTP. This was thought to be the source for some of the
MTP formed in the experiments. The ether extraction was also done to improve
recovery of MTOB and provide a more accurate assessment of MTOB and MTP
levels. The autoradiogram of the ether extracted fraction still showed a band
corresponding to MTP, however its labeling pattern was more variable,
increasing then decreasing intermittently.

MTOB is the most likely intermediate between Met and MTHB, which
increases its likelihood as a pathway intermediate (12, 20, 28). The increased
flux through the DMSP pathway under nitrogen limiting conditions is also
consistent with a transaminase reaction at the head of the pathway.

MTHB is a stable compound and its labeling kinetics are quite consistent
with that of a pathway intermediate (Fig.18). It is also a natural precursor to
DMSHB via methylation. It should be noted that in representative samples of
DMSP accumulating algae no MTP methylating activity has been found
previously (80). A methyltransferase, which catalyzes the formation of DMSHB
from D-MTHB has been identified in E. intestinalis (80).

74

 

75

10

30

240 120 60

1 200

Figure 17) Organic acids extracted from Heterocapsa triquetra in ether. This

shows disappearance of MTOB after 30 minutes.

 

3"’S Incorporation into Stable Intermediates

 

i + ' ”was“ 7;
.+ DMSHB;

 

 

 

 

 

 

500
450

 

350 -
300 f" ' W "k fr“.
250 - i ' MTHB?
200 3:939:91]
150
100
50

o -

nCiIgfw

 

 

 

 

 

Figure 18) Graph showing incorporation of 358 into selected metabolites.
A) Quantitation of MTHB and DMSHB from acid extracted algae. This
shows the skewing of MTHB quantities from 60 min onward due to
presence of a contaminating band B) MTHB from ether fraction of acid
extracted algae. There is still an increase at 120 minutes (plotted as a
single point) but later values show a general downward trend.

76

 

DMSHB and DMSP

DMSHB accumulated label over the entire time period, as did DMSP
(Fig.19). This was a somewhat surprising result as intracellular free Met levels
showed a steady decrease in radioactivity. A large source of [35S]Met between
the 240 min and 1200 min was likely protein bound met, which continued to
increase up to the 240 min. time point, but showed a sharp decrease at 1200 min
(Fig. 19). Grone and Kirst have previously shown that an increase in DMSP
synthesis is accompanied by as much as a 50% decrease in protein content/cell
for Tetraselmis grown in nitrogen deficient media.

DMSP and DMSHB migrate very closely together during TLE.
Autoradiography of one half of sulfonium containing fraction showed a strong
tailing from the DMSP band that extended through the zone of migration for
DMSHB. The zone corresponding to DMSHB was scraped, eluted and a new
TLE, and autoradiogram was made, which showed a small band corresponding
to DMSHB (Fig.20). The other half of the sulfonium fraction was treated with
NaOH as another means to confirm the identity of DMSHB. DMSP is base labile,
and undergoes a’,B—elimination to yield DMS and acrylate, but DMSHB is stable
to base treatment (12). Once DMSP was eliminated from the sample, TLE
showed a strong band corresponding to DMSHB. Subsequent to scraping and
quantification of the radioactive bands from the TLE plate the plate was stained
with Dragendorff’s reagent. There were no Dragendorff's positive bands in the
sample lanes with electrophoretic mobility identical to DMSP. If DMSP had

survived the base treatment there would have been a positive staining band just

77

above the zone of migration for DMSHB. A single band was found in the sample
lanes that had an electrophoretic mobility identical to that of authentic DMSHB

added to adjacent lanes.

Heterocapsa did not take up [35S]DMSHB. We were therefore unable to

confirm its role as the direct precursor through direct feeding studies.

78

 

 

Major fates of [35S]Met in H.triquetra

 

{Lorin—SP7" " ';
ii—r—free Met 3

100000 1

0 c — a . '

0 5 10 30 80 120 240 1200
minutes

 

 

 

 

Figure19) Graph showing the major fates of a 0.4 nmol dose of
P5S]Met. The disparity between disappearance of intracellular
free Met and products is likely due to production of

DMS.

79

 

 

 

 

 

 

Figure 20) Autoradiogram of TLE showing the appearance of DMSHB
bands after scraping lower zone of DMSP band from previous TLE,
and eluting. DMSHB is the very faint band beneath the heavier DMSP spot.

80

Cold trapping studies with MTHB and MTP

The data from the [35S]Met labeling experiments in which MTOB, MTHB
and DMSHB all acquired label seems to confirm the pathway previously reported
for other classes of marine algae (20). In order to further conflrrn this conclusion.
I carried out cold trapping experiments to determine the ability of a pool of
exogenously supplied MTHB, and MTP to diminish flux of label from [35$]Met into
the downstream sulfonium products DMSHB and DMSP (Table III). The
experiment was done for two time points 0.5 h and 4 h. Six day old cultures were
presented with 0.8 nmol of MTHB, or MTP 1 min. prior to the addition of a 0.4
nmol dose (0.8 pCi) of [”SlMet. We monitored cellular uptake of radioactivity;
and activity recovered in DMSP bands from TLE plates

Uptake of methionine was inhibited by 4.0% at 0.5 hrs. and 6.0% at 4 hrs
for the culture with added MTHB In the culture with added MTP uptake was only
slightly decreased with 5.0% and 7.0% respectively. The amount of radioactivity
recovered from the sulfonium fraction in the MTHB treated sample was 62% of
the control at 0.5 h and 59% at 4h. MTP was slightly less effective containing
61% and 75% respectively.

The effectiveness of MTP as a cold trap would seem to suggest that it

both MTP and MTHB may be intermediates.

81

 

 

 

 

 

 

 

 

Sample Time (min) [ SjMet uptake (dpm) F8] in sulfonium
* cmpds.

Control 30 218,867 2,525

Control 240 272,1 13 28,799

MTHB 30 189,905 2,169

MTHB 240 234,112 17,597

MTP 30 182,647 2,002

MTP 240 237,897 20,651

 

 

 

 

Tablelll) A single study showing the effect of a 1.0 nmol dose of MTHB or MTHB
on the incorporation of [358] from Met —> DMSP and DMSHB.

82

 

Discussion

The data from the [35$]Met labeling experiments are consistent with the
pathway, Met —> MTOB —> MTHB —+ DMSHB -+ DMSP. They are not consistent
with the pathway proposed by lshida of Met -> MTP-amine ——> MTP -> DMSP.
This is due primarily to the lack of [3°S]MTP-amine in the radiotracer studies.

The labeling of MTP can be explained by oxidative decarboxylation of MTOB. It
is likely that MTHB ‘s effectiveness as a cold trap is the result of its conversion to
methionine. This is to be predicted based on the sequence of enzymatic steps in
the known algal pathway. In this pathway both the transamination of Met to form
MTOB, and the reduction of MTOB to form MTHB are reversible. It was also
found that [35$]MTHB was mostly converted to Met in E. intestinalis, although
some was converted to DMSP (20). It is worth noting that the reductase
responsible for the MTOB —> MTHB conversion exclusively synthesizes D-MTHB.
This means that the methyltransferase that converts DMSHB to DMSP is may
also prefer the D-enantiomer. This was also shown in E. intestinalis. Most of
the L-MTHB provided was likely converted to Met with only a small amount
continuing through the pathway to DMSP. This small amount of conversion to
DMSP may explain why MTHB is more effective as a cold trap than MTP.

The labeling pattern of MTP and its effectiveness as a cold trap are
consistent with a role as an intermediate. MTP could be formed via the oxidation
of MTP or MTHB and methylated to form DMSP in a parallel pathway. MTP is
most likely not an intermediate. It is more likely that the efficacy of MTP as a

cold trap is due to recycling back to Met via methylthioadenosine (MTA), or a

83

reduction in Met transamination via an inhibition mechanism. No MTP
methylating activity has been demonstrated in algal extracts of DMSP producers.
If a parallel pathway is operative it may be unique to dinoflagellates.

The labeling pattern of DMSHB is a somewhat surprising result. It would
be expected that its position as the immediate precursor of DMSP would produce
a pattern opposite to that seen in the radiotracer experiments. One would expect
higher levels at early time points with a progressive diminution of label at the later
stages of the time course. One possible explanation of this pattern is a storage
pool of DMSHB exists in slow equilibrium with a small metabolically active pool,
which carries a high flux. This would produce the observed labeling pattern as
the storage pool slowly becomes labeled over the time course of the experiment,
while the bulk of the 358 flux passes through the metabolically active pool into
DMSP (20). This would be expected to produce a rapid accumulation of label in
DMSP while DMSHB incorporates label at a much slower rate. A computer
model developed to fit the labeling data for the DMSP synthetic pathway may
support this explanation (20).

The data in this chapter establishes a pathway for DMSP biosynthesis in
dinoflagellates for the first time. This suggests that the pathway is strongly
conserved among marine algae. Dinoflagellates represent another class of
marine algae for which the DMSP pathway has been defined. The novel
compound DMSHB has been found in four other DMSP accumulating algae,
Tetraselmis subcordiforrnis (Prasinophyceae) Melosira nummulcides

(Bacillariophyceae) Enteromorpha intestinalis (Chlorophyceae) and Emiliana

huxleyi (Prymnesiophyceae) and may itself be diagnostic for this pathway, as its
presence has not been reported in any non-DMSP producers. Neither has it
been reported in any DMSP producing organisms, which do not share this
pathway.

It is important to point out that our data do not support the DMSP pathway
proposed by Uchida et. al. Their model was developed for a heterotrophic
species, while ours is for a photosynthetic species. Linglodinium polyedra is
another dinoflagellate species in which some preliminary work has been done on
DMSP biosynthesis. Nakamura confirmed Greene’s result that DMSP was
synthesized using methionine as a substrate. He also suggested that two other
sulfonium compounds, gonyol and gonyauline, were synthesized from DMSP. It
is likely that L. polyedra and H. triquetra share a common pathway for DMSP
biosynthesis. The compound DMSHB provides an alternative substrate to DMSP
however for the synthesis of gonyol, and gonyauline.

The DMSP story in dinoflagellates should be completed by extending the
methods used in this study to C. cohnii and L. polyedra. Experiments to address

this question are presented in Chapter IV.

85

Chapter IV

THE BIOSYNTHESIS OF DIMETHYLSUFONIOPROPIONATE IN TWO MARINE
DINOFLAGELLATES: CRYPTHECOHDINIUM COHNII AND LINGLODINIUM
POL YEDRA.

Introduction
Crypthecohdinium cohnii and Linglodinium polyedra (formerly
known as Gonyaulax polyedra) are two unusual DMSP producing organisms. C.
cohnii is the only known heterotrophic organism that accumulates DMSP. It
produces DMSP in the dark as well as in the light (48, 86). For the experiments
described herein C. cohnii cultures were grown in the dark. The biosynthesis of
two sulfonium compounds in addition to DMSP, whose structures are very similar
to DMSP is the unique feature of L. polyedra. Gonyauline
(cis-2-(dimethylsulfonio) cyclopropane carboxylate) is a novel compound whose
function is associated with period shortening of in the bioluminescent circadian
rhythm of L. polyedra (12, 63). Gonyol (3S-5—dimethylsulfonio-3-
hydroxypentanoate) is also found in some other dinoflagellates such as
Amphidinium carterae (63). There is no defined physiological function for gonyol.
DMSP biosynthetic studies have been conducted previously in both of
these organisms. Uchida et al. authored a paper on the synthesis of DMSP in C.
cohnii in which they proposed the first step of the pathway is decarboxylation of

Met. They described the complete pathway as the following. Met —-)

86

methylthiopropylamine (MTP-amine) —> methylthiopropionate (MTP) —> DMSP.
Their evidence was largely based on cold trapping experiments with MTP and
the purification of a methionine decarboxylase, an activity that is known in many
non-DMSP accumulating organisms. ( 12, 54, 58, 77). There were no direct
labeling studies done, however, no other investigators have done follow-up
studies to support or disprove the pathway.

Nakamura et al. confirmed Greene’s result (30) that DMSP was
synthesized from Met in L. polyedra. They also suggested that the carbon
skeleton of gonyol was derived from DMSP and acetate, and that gonyauline was
also synthesized from DMSP, with it’s C1 carbon originating from 002 (63).

In this chapter [35$]Met feeding studies are described which are consistent

with the pathway: Met —> MTOB -) MTHB —+ DMSHB —> DMSP in both species.

87

Materials and Methods

Growth conditions for Linglodinium polyedra. Stocks were obtained from
the Provasoli-Guillard National Center for the Culture of Marine Phytoplankton.
(Stock number CCMP 407) Cultures of L. polyedra were grown in L1 media at
18°C on a 14:10 light dark cycle with a photon fluence of 390 pE. Cells were
used in exponential growth phase or early stationary phase.

Growth conditions for Crypthecodinium cohnii. Stock cultures were
obtained from Martek Biosciences. Cultures of C. cohnii were grown In three
types of media. 50 ml cultures were grown in a yeast-glucose medium prepared
from 50 glL glucose and 6 g/L yeast extract in filtered seawater. The cultures
were maintained in 250 ml Erlenmeyer flasks in ambient light at room
temperature. Cultures in L1 medium were maintained using the same conditions
as those used for the yeastglucose cultures. In addition a defined medium
prepared according to the method described by Guillard and Keller (33) was
used. The protocol was modified by reducing the concentration of (NH4)2$O4
tenfold to a concentration of 5.0 mg/L. Na2804 and NazEDTA were also
eliminated from the stock solution. Na2$O4 was added on the fourth day of
culture, or was replaced with 2.5 mM cysteine. NazEDTA was added on the
fourth day of culture. Vitamin levels were augmented to the levels suggested by
Tuttle and Loeblich (85) 50 ml cultures were grown in 250 ml Erlenmeyer flasks
that were maintained in the dark. Subcultures were started by inoculating 50 ml

of the defined medium with 1 ml of cells from a culture not more than 10 days old

88

that was grown in 50 g/L glucose and 6 g/L yeast extract medium. Cell
populations were calculated using a hemacytometer with a 0.1 mM cell depth

Media preparations for uptake experiments. The defined medium from
above was used as the basis for all experiments. A low nitrogen variant was
prepared from the basic media protocol as described above. Complete medium
was supplemented with 2.25 mg (NH4)2$O4 to raise the nitrogen level to that
prescribed in the protocol. High salt medium was made by adding 0.995 g NaCI
and 2.25 mg (NH4)2SO4. Another variant of the medium was made by replacing
$04 with 2.5 mM cysteine. A high salt, low nitrogen medium was prepared by
adding 0.995 9 NaCl. A high salt medium was also prepared in which 804 was
replaced with 2.5 mM cysteine. This medium was supplemented with 2.25mg
(NH4)2SO4. A high salt, low nitrogen, 2.5 mM cysteine medium was made
exactly as the preceding, except no (NH4)2$O4 was added.

Radiochemicals. [35$]Met 43.5 TBq/nmol (NEG-009A) was purchased
from NEN Life Science products. It was treated with 4% thioglycollic acid at 95°C
for 4 h to reduce Met sulfoxide and with A650 (NH4+) to remove basic impurities
and AG1(COO') to remove acidic impurities (51). Racemic MT HB was
purchased from Sigma and was recrystallized from aqueous ethanol.

Labeling with [35S]Met. C. cohnii cultures were used on the seventh day
following inoculation. Seven 12 ml subcultures were prepared from the reduced
nitrogen culture. Cells were used in exponential growth phase at a population

density of 7.5 x 105 cells/ ml. Cultures were inoculated with 19 pCi

89

(0.4 nmol Met) and incubated on a 2-D rocker at a low to moderate speed. L.
polyedra cultures were used in late exponential or early stationary phase. A 52
pCi, 0.8 nmol dose of [35$]Met was added to each of four cultures corresponding
to harvest time points of 0, 20, 60, and 120 min. Lighting and temperature
conditions were identical to those used in culturing. One culture was used as a 0
time point (control). Cultures were centrifuged to pellet the cells. The
supernatant was removed and a dose of [35$]Met was added that was equivalent
to that added to the other cultures. An aliquot was removed and assayed by
scintillation counting to determine the amount of radioactivity remaining outside
the cells. This was done to provide a baseline that could be used to monitor
uptake of radioactivity by the other cultures. The amount of radioactivity
remaining in the supernatant was assayed, and the percent uptake was
calculated using the value obtained from the control culture as a baseline value.
Extraction of metabolites. L. polyedra cultures were gently centrifuged
after labeling at 4300 rpm using an S4180 rotor in a Beckman GS15R centrifuge.
Freezing in liquid nitrogen and adding l.0 ml of 0.1 N HCI lysed the pellets. One
pmol of MTP, MTOB, MTHB, SMM, MTP-amine, DMSP, and DMSHB were
added as carriers to enhance recovery. The pellet was vortexed and centrifuged
at 4300 rpm. The lysate was fractionated into acidic, basic, and zwitterionic
components using a series of three columns A61 (-OH), Bio-Rex 70 (H+), and
AG50-X8 arranged in tandem. The cell lysate was washed through the series of
columns, and columns were eluted individually. C. cohnii cultures were also

extracted with 2 ml of methanolzchloroformzwater (MCW) in a 12:5:1 mixture. The

90

residue was re-extracted with an additional 2 ml of MCW. The extracts were
pooled, and an additional 1ml of chloroform and 1.5 ml of H20 were added in that
order. The aqueous phase was kept and acidified with 25 pmol of HCI. The
sample was lyophilized, redissolved in 1 ml of H20 and added to the ion-
exchange column series. Column eluates from both algal species were
lyophilized, resuspended in 300 pl of 60% methanol and lyophilized in a vacuum
concentrator. The residues were suspended in a minimal volume of water for
application to TLC plates.

Isolation of metabolites. MTP, MTOB, and MTHB and Met were isolated
using a 1.5 ml column packed with A61 - X8 ('OH) 200 — 400 mesh resin. Cell
lysates were applied in a 1 ml volume. The column was washed with 15 ml dd
H20, and eluted with 5.0 ml 2.5 N HCI. SMM and MT P-amine were isolated by
passing the entire wash fraction of the A61 column through a 1 ml column
packed with Bio-Rex 70 (*H) 100 — 200 mesh resin. The column was washed
with 15 ml of dd H20 and eluted with 15 ml of 1.0 N HCI. DMSHB and DMSP
were isolated by passing the entire wash fraction from the Bio-Rex 70 column
through a 1.0ml column packed with A650W-X8 (H’) 200 - 400 mesh resin. The
column was washed with 5.0 ml of ddHZO and eluted with 5.0 ml of 2.5 N HCI.

TLE of cellular metabolites. 358 labeled metabolites were identified by
co-migration with authentic standards during thin layer electrophoresis.

TLE was done using a Hunter thin layer peptide mapping system and glass
backed 0.1 mm cellulose TLC plates from Merck. Organic acids, amino acids,

and bases were separated by TLE in a 1:1:38 mixture of pyridine/acetic acid/H20

91

for 20 min at 1.8 kV. Sulfonium compounds were separated in 1.5 N formic acid
at 1.8 W for 20 min.

2-Dimensional TLC/T LE. Zwitterionic fractions isolated from C. cohnii
were separated using a 2-dimensional TLC/T LE sytem. TLC plates were glass
backed 0.1 mm cellulose plates from Merck. TLC was done first using a 6:2:2
butanolzacetic acidzHZO solvent system. When TLC was finished the plates were
allowed to dry completely overnight. TLE was then done in an orthogonal
direction in 1.5N formic acid at 1.8kV for 20 min.

Identification of metabolites. Metabolites isolated from column fractions
were spotted onto TLC plates in individual lanes. Authentic standards of
expected metabolites within the column fractions were spotted in adjacent lanes.
After TLE was carried out an autoradiogram was made. The radiolabeled bands
were localized, excised from the plate, and counted by scintillation. The lane
containing the authentic standards was then stained with a reagent specific to the
type of metabolites isolated from the column fractions. Dragendorff’s reagent
was used to positively identify sulfonium compounds (DMSHB and DMSP)
isolated from the A650-X8 column (36). Ninhydrin was used to identify basic
compounds (SMM, MTP-amine) from the Bio-Rex 70 column. lodoplatinate was
used for organic acids from the A61 column (KT MB, MTP, MTHB).

The mobilities of the standards were then compared to the bands from the

autoradiogram, and the individual metabolites subsequently identified.

92

Autoradiography. Subsequent to separation by TLE, the
electrophoretograms were enclosed in an X-ray film cassette with one sheet of X-
ray film. Exposure times were dependent on the quantity of radioactivity in the

sample.

Quantitation of radioactivity. The amount of radioactivity recovered from
columns was assayed by adding a small aliquot to 5 ml of Ready Gel scintillation
fluid and assaying by scintillation counting. Radioactivity recovered from plates
was calculated by scraping radioactive bands, and adding the scrapings to 2 ml
of H20. This was mixed with 2 ml of Ready Gel scintillation fluid and assayed by

scintillation counting.

93

Results

Production of DMSP in C. cohnii was strongly influenced by culture
conditions. When cells were cultured in a defined medium with limited nitrogen;
70% of a 0.4 nmol dose of [35$]Met was taken up by the cells at the 30 min. time
point. This compares to little, or no uptake of Met by cells growing in the same
defined medium wherein nitrogen was not limited and a 3.2-fold increase in Met

uptake compared to cultures grown in L1 media (Tables IV & IV).

Methionine Uptake experiments

Culture conditions of C. cohnii were optimized to maximize Met
uptake, population density, and health of cultures as assessed by cell
morphology. A defined medium for cell culture was prepared as described by
Guillard and Keller (33). The medium was modified by augmenting vitamin
concentrations to that specified by Tuttle & Loeblich (85). Seven different
versions of the cell medium were made. In one version the nitrogen
concentration was reduced ten-fold. A high salt medium was made by increasing
the NaCL concentration from 0.35 M to 1 M. The other variable tested was sulfur
source. This was changed from sulfate to cysteine. The variables were tested
alone and in concert. All of the cultures containing cysteine in lieu of sulfate did
poorly. This was marked by a lack of swimming cells. The high salt cultures

were distinguished by large rounded cells, but with few swimming cells.

When high salt was combined with low nitrogen, a larger number of swimming
cells were observed, however growth was inhibited. All of the cultures in which at
least two of the three varied factors were combined did poorly. The low nitrogen
cultures were easily the healthiest cultures when division rates (Fig. 21) cell
morphology, and metabolic activity (this was associated with swimming) were
considered together.

Experiments designed to monitor Met uptake of cultures grown in the
various types of media showed the low nitrogen cultures to be far more efficient
in assimilating Met. This was evidenced by a 70% depletion from the media of

exogenously supplied methionine at 30 mins. and 84% at 60 mins. (Fig. 22)

95

 

TablelV. DMSP production by Crypthecodinium cohnii grown in L1 medium

 

 

 

 

 

 

 

 

time DMSP (nmols) DMSP (nmols/nmol
Met)

15 mins. 4.2 x 10" 7.9 x 10'3

30 mins 9.7 x 10‘ 1.1 x 10'2

60nmm. 1.6x10'3 1.8x10'2

120 mins 4.0 x 10':r 2.6 x 10‘2

240mins. 1.2 x 10'1 5.2 x 1o"I

 

 

 

TableV. DMSP production by Crypthecodinium cohnii grown in reduced N
medium

 

 

 

 

 

 

 

 

time DMSP (nmols) DMSP (nmols/nmol
Met)

5 mins. 4.7 x 10'3 4.3 x 10"

15 mins 1.0 x10"2 4.8 x10’T

30 mins. 1.2 x 10‘2 4.3 x 10'2

60 mins 2.0 x 10'2 5.4 x 10'2

120mins. 2.1 x 10‘2 5.4 x 10'2

 

 

Table IV) Quantitation of DMSP after a 0.4 nmol dose of exogenously
supplied [35$]Met. The amount of DMSP is based on the specific
activity of the supplied [3°S]Met in uCi/nmol. Cultures for this
experiment were grown in L1 media.

Table IV) Same as above, except that C. cohnii cultures were grown in a
reduced nitrogen medium.

96

 

 

30.8 5 ©9585 9m :o_fitomm_u\m_mo£ £5 E mommEs
.mfioE maotm> 5 85:8 EEoo Eacfiooocabo .2 $23 5265 SN 9:9“.

 

zflamwaiumq2lfl+gczenwlrz
2325252231.;
2.350.510.2271.
5:53:09. .271,
2325.512

92253 ,2

652.27l

Sanoosz, “

(svou) 'Iumnw

 

 

 

 

 

Enos m:o_._u> E ===oo.o be £320

97

.EzfimE =8 9.: c. @5595. 53:328. .0 anEm 9.: mm oozoa m_ 9.9a: .mczcaoo

cozm==£om E oo>mmmm mm; ESmanam 9: new cosmméEcoo >9 622.3 0.63

$60 of. .EE on cam .om .2. £an 6E: 695 .m 93.8 .8 we come ho 8395 _E F m

mEEEmm B oocogcoE mm; 9.9a: .EEcmE coccmo m so 325:? .893 :_ $505
$58 EEEoooochO 3 Dogma. “32.3345 232.0396 so omon SE: to m _o 9.8a: ANN 2:9...

98

 

 

 

 

 

 

 

 

 

 

 

 

 

II... .I F p P + _ b _ p o
2
<9: on + 6.2 .25 F + as .25 3 ll .il , 882
2 3.5.? m6 2.: md IT l -- rill/ix
. 808w.
235.282le ,, ,,
/, - ooooom
32532535318227... _/
o.
2 35 3 I... -. 888 m
m3 _>_Em.m , ooooom
.052? I.-- l ..
”all!” Ill. dill . . occcow
i a.
3996 lol 1
.. ooooou
o 65:00 l¢|
ooooom

 

 

 

9.9% 8.26%

99

Glycine Betaine and DMSP in C. cohnii

During an early experiment with C. cohnii a Dragendorff's positive band
was detected in the zwitterionic fraction of cell lysate. The spot had an
electrophoretic mobility similar to that of glycine betaine, a functionally equivalent
compatible osmolyte. We had earlier observed the presence of both DMSP and
glycine betaine in the FABMS spectrum of the zwitterionic fraction of
zooxanthellae isolated from Tridacnid clams. There appeared to be a reciprocal
relationship between the two peaks (Fig. 23). We therefore looked at the FABMS
spectrum of the zwitterionic compounds isolated from C. cohnii and we observed
a small peak at mlz 118, which could be attributable to the presence of glycine
betaine (Fig. 24). It has been reported in the literature that addition of betaine to
growth media helps stimulate growth of C. cohnii cultures (48). Tetraselmis
subcordiformis, a Prasinophyte algae and glycine betaine producer, is known to
accumulate up to two-fold more DMSP in nitrogen limited conditions. It has been
previously suggested in the literature that DMSP, and glycine betaine levels may
be inversely related with glycine betaine being produced when nitrogen is
available, and DMSP produced when nitrogen is limiting (48). This relationship
was another factor in deciding to use nitrogen limited medium for cell culture (32,

48).

100

SI947E

 

 

 

 

 

 

 

 

 

 

 

 

 

118.82
lees '
$8
.‘59'95 135.113
j .132.“ 144.04 176.5 l
r r T 7 F * 5 fl 1’ f I ' r fi 1
115 129 125 1:16 135 110 145 159 155 160 165 170 175 180 185 190 195 298
Ill:
155332 (.9 ea
Ima '
l
53‘ 135.81.
158.8
111.97 '13331
1 I l 175.35 135.113
9“ A- l A All 21 1 Al- 11 1A A - L121 AA 1‘ A
T T r 1— 1 fl —r—* r Y I ' T f V
110 115 120 125 130 135 148 145 150 155 160 165 170 175 100 195 139 135 aea
on
539159 “5.95
1 1
135.12
52— '
476.29
, 149.13 ,7
95.95 112.213 L 11 ' '1601 .
B-IAM I r4 ‘; 21L71 AAA A_‘Tl l 1 ll IL--A 129‘; LIAA‘ _ l
r " t f T V I V r " 1
95 1% 105.110 115 120 125 130 135 149 145 150 135 [SB 155 179 175 180
on

Figure 23) FABMS spectra of zwitterionic fraction isolated from Tridacnid clams.
The clams have a symbiotic relationship with zooxanthellae. The ions
at mlz 118 and 135 are due to the presence of glycine betaine and
DMSP respectively, produced by the zooxanthellae. Note that in the
top spectrum the major peak is at mlz 118 (glycine betaine). In the
following two spectra the increase in the intensity of the ion at mlz
135 (DMSP) is accompanied by a decrease in glycine betaine (118).

101

6583 osozo oo

oocomoa or: 9 can m_ w: NE om xmoa :95 on... .dmsfi 9 mocoawmtoo mm? NE om xmoa
5.55 cap .38 52558255 52. 85.8 cases. 285:? ca 52584. mzmfi i 2:9“.

 

 

 

 

N\E
so. mm. an. mm. am. mm. am. mv. 0v. mm. an. mm. am. m_. a._ ma.
_ 4 1r 0 .r . , 1 . F. . T p . . . p 1
1‘.— 14 1114 .1 4t 114441111-4144 4‘ 1411} 111—. I471 11 1 l 1 11d ‘4 didd 1111's
_ z _ 2
\ nu
wmumn_ mm.wm_ 4.
. mm.nsm
mm.me. 4 cm
nm.sn_
mm.vmm +.ss.
momma
.m.m u .acn cmmu u coo: co—
nm._o H occum c.e __.. a pm

o.mou n.s " each
n Eocuuuam «as: o

[35$]Met feeding experiments in C. cohnii

Three separate 358 trace labeling studies were carried out with C. cohnii.
The first experiment was done with C. cohnii cultures grown in L1 media. A 0.4
nmol dose of [35S]Met was provided and followed into metabolites. We were
able to identify MTOB, MTHB, and MTP as metabolites that assimilated tracer
Met (Fig. 25). There was no labeled MTP-amine, which would be expected if the
pathway suggested by Uchida et al. were operative. We also saw intensely
labeled DMSP. The dark bands seen in the TLE autoradiographs of the column
fraction containing DMSP most likely obscured any DMSHB. Because DMSHB
was the putative immediate precursor of DMSP, earlier time points with a higher
specific activity tracer dose may reveal its presence. The slow rate of Met uptake
was another consideration in this experiment. Only 18% of supplied Met was
taken up at 15 min with a gradual increase to 39% over the entire 4 h labeling
time course. Such a slow rate of uptake may provide insufficient flux to see

DMSHB at early time points, if pool sizes are small.

103

104

 

60

120

30 15

Minutes

organic acids extracted from C. COhnll

mg of

Figure 25) 35$ label

The second [35S]Met labeling experiment was done with a culture grown in
a low nitrogen variant of C. cohnii media. A seven day old, large single culture
was divided into six 12 ml subcultures, corresponding to five time points, and one
that was used as a control. A 0.4 nmol dose of [35$]Met was provided. The
decrease in nitrogen concentration not only improved the rate of Met uptake
significantly; it also increased flux through the DMSP pathway. The low nitrogen
media improved the number of moles of DMSP produced/ number of moles Met
taken up by 2-fold (Table V). Unfortunately, sampling at earlier time points failed
to provide a clear signal for the presence of DMSHB, or MTP-amine. We had
identified early precursors of both possible pathways, yet we needed to find
either DMSHB or MTP-amine to clearly differentiate between the two possible
routes for DMSP biosynthesis. One of the perceived problems was the smearing
of bands in the TLE autoradiographs of the zwitterionic fraction. We thought this
might be due to oils in the cellular extract. Oils are the major storage product for
C.cohnii, and low nitrogen conditions are known to trigger an increase in the size
of the storage pools (personal communication from Martek Biosciences). It was
decided to use an alternative extraction procedure in another labeling
experiment.

I supplied a 0.4 nmol dose of tracer Met to six 12 ml subcultures taken
from a single large culture of C. cohnii. In this experiment the cells were
extracted with MCW. We analyzed the zwitterionic fraction of the cell lysate
using a ZD-TLC/T LE system. This gave improved resolution between DMSP and
DMSHB. We observed a clear spot that corresponded to DMSHB (Fig. 26).

105

I also analyzed the basic fraction of the cellular lysate and as in previous
experiments there was no evidence of MTP-amine. This experiment finally
provided the piece of evidence needed to establish the pathway: Met —> MTOB
—> MTHB —> DMSHB -> DMSP as the operative pathway in C. cohnii.

Culture conditions for L. polyedra

The challenges in doing feeding studies in L. polyedra begin with the
difficult task of obtaining high density cultures. L. polyedra has a well deserved
reputation for being an organism that does not do well at high or low density. I
was able to obtain very robust cultures by starting with stock that was derived
from a robust culture. We also found that having a sufficient depth of medium in
the culture vessel was necessary to support the daily vertical migration of the
species. Cultures that we attempted to maintain in Erlenmeyer flask never
obtained very high density. l intuited this to be due to the lack of depth of the
culture. The superior surface/volume ratio of flask culturing seemed to be of no
advantage. When we began using 100 ml of volume in 75 cm2 tissue culture
flask, the cultures became very healthy, and were aftenrvards exceedingly easy to
maintain. Whereas, other dinoflagellate species need to be passaged every third
or fourth day, it was no problem for the cultures of L. polyedra to remain without

being passaged for a period of weeks with no perceived loss of culture vitality.

106

 

ions form Crypthecohdini m

ion of P5$]Met.

ium fract

Figure 26) 2-D TLC/T LE of sulfon

it

cohnii 1200 minutes after add

107

m
0
1..
E

MTP

 

108

60

120

20

Minutes

Figure 27) Organic acids extracted from L. polyedra

[35$]Met feeding experiments in L. polyedra

The outstanding problem with doing these labeling studies in L. polyedra
is with the separation of the sulfonium compounds. DMSP, DMSHB, gonyol, and
gonyauline are all very similar in structure and separate very poorly during TLE. I
also tried a to separate them by TLC using a wide variety of solvent systems.
We finally decided on a 20 TLC/T LE separation system. This provided the best
results among the alternatives considered. We were able to obtain a clean
separation for DMSHB and DMSP, although addition of the other sulfonium
products still made it difficult to effect a clean separation.

A single large culture of L. polyedra was subdivided into four 24 ml
subcultures. The four subcultures corresponded to 3 time points and a control.
We added a 0.8 nmol dose of Met. The results indicated labeling of MTOB,
MTHB, and MTP in the organic acid fraction (Fig. 27). These results are
consistent with those obtained for all of the other species of marine algae that
were tested. There was also strong incorporation of label into DMSHB, DMSP,
gonyol, and gonyauline. Labeling of DMSHB and DMSP preceded that of gonyol
and gonyauline. The quantity of 358 contained in each metabolite was estimated.
Because of poor separation quantitation was difficult for 60 min and 120 min as
gonyol and gonyauline began to incorporate label. The TLE was divided into 4
zones where standards were known to migrate and label incorporation into each

individual metabolite was quantitated by counting these zones (Fig. 28).

109

328 E 8582a 96 £85 £5 5 «some;

._o.€oo new 05.868 *0 coszLo. 9: 2 98585 8 mar—20 8m mesE 5?
E2228 2 m2... ._o>coo new 65.8.38 2 .25 amio new $555 5 «pE.:anom man
.88on 4 S 8:3an8 E:_:o=:m 25 8:98.85 8.25m»: 9:265 :85 SN 9:9“.

0E:
.mSE om .mEE om

_ . r Illqu‘llo

. oooom

 

 

 

 

 

 

 

 

 

r oooov

 

 

 

 

 

- ooooo

dazed 588

mImEQ n. D.
gigool ooooop .m

. 12988,. . 88m.

 

. coco:

. ooooop

 

 

 

f oooomp

 

 

 

ooooom

.230on . 4 *0 3:89:00 82:02.5 2:. 8:22:85 m3

110

Discussion

The labeling information garnered from these experiments, together with
the data from Heterocapsa triquetra establish the biosynthetic pathway in
dinoflagellates for the first time. It is clear that this pathway is strongly conserved
in marine algae. C. cohnii and L. polyedra represent two special cases, which
might reasonably be considered to diverge from other marine algae. The
synthesis of gonyol and gonyauline appear to have evolved from the DMSP
pathway. The kinetics of the labeling of sulfonium compounds in L. polyedra are
consistent with DMSHB and DMSP as precursors of gonyol and gonyauline. The
question as to whether DMSP or DMSHB is the immediate precursor of the other
two compounds remains unsolved. Nakamura’s work suggested that DMSP was
the immediate precursor to gonyol and gonyauline. Nothing in our studies
contradicts this conclusion. Nakamura harvested cells 24 hrs after feeding
labeled precursors and the possibility of recycling of compounds into other
compounds is a very real possibility as the labeling kinetics of our studies show
that flux through the pathway is very high. This means that DMSHB is still the
possible immediate precursor to gonyol or gonyauline, or possibly both.

The study conducted by Uchida et al. was done well in advance of our
work with E. intestinalis. We now know much more about the three compounds
they tested as potential intermediates by their ability to act as cold traps. SMM is
cleariy not an intermediate in marine algal DMSP biosynthesis. MTOB is a
relatively unstable compound and may have been subject to decomposition in

these experiments.

111

We have also found that MTP is an effective cold trap in H. triquetra, but this
activity is likely the result of recycling back to Met via the Met salvage pathway
and not due to it’s role as a DMSP precursor. This is further corroborated by the
lack of MTP methylating activity in cellular extracts of DMSP producing algae
(80). The methionine decarboxylase activity in C. cohnii is widespread in algae

and other plants which do not produce DMSP.

112

References

113

10.

11.

References

Aneja, VP. and W. J. Cooper, Biogenic sulfur emissions: A review:
1989. In E. S. Saltzman, W. J. Cooper, eds, Biogenic Sulfur in the
Environment. American Chemical Society, Washington DC. p. 2-13

Bates, T. S. Charlson, R. J. and R. H. Gammon, Evidence for the climatic
role of marine biogenic sulfur. Nature, 1987.329: p. 319-320.

Balch, W. M. and K. Kilpatrick, Coccolith production and detachment by
Emiliana huxleyi (Prymnesiophyceae). Journal of Phycology, 1993. 29:
p. 566-575

Blunden, 6. and S. Gordon, Betaines and their sulphonio analogues in
marine algae. Progress in Phycological Research. 1986. 4: p. 40-80

Blunden, 6., et al., Betaines and tertiary sulphonium compounds from 62
species of marine algae. Biochemical Systematics and Ecology. 1992.
20(4): p. 373-388

Borowitzka, L. Osmoregulation in blue-green algae. Progress in
Phycological Research. 1986.4: p. 243-256

Burg, M., Molecular basis of osmotic regulation. 1995. Walter B. Cannon
Lecture. p. F983-F996

Challenger, F ., et al., Studies on Biological methylation. XVII. The natural
occurrence and chemical reactions of some thetins. 1957. Archives of
Biochemistry and Biophysics. 69: p. 514-523

Colmer, T. D., et al., Interactive effects of salinity, nitrogen and sulphur
on the inorganic solutes in Spartina altemiflora leaf blades. 1996. 296(47):
p. 369-375

Cooper, A. J. et al., Formation of highly reactive vinylglyoxalate
(2-oxo-3-butenoate) from amino acids with good leaving groups in the )-
position. Toxicological implications and therapeutic potential. 1989.
Biochemical Pharmacology. 38: p. 3895-3901

Cooper, A. J. L. and A. D. Hanson, Advances in enzymology of the

biogeochemical sulfur cycle. 1998. Chemtracts-Biochemistry and
Molecular Biology. 1 1: p. 729-747

114

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Cooper, A. J. L., Chemical and biochemical properties of sulfonium
compounds 1996. In R.Kiene, P. Visscher, G. Kirst, M. Keller eds.
Environmental and biological chemistry of DMSP and related
sulfonium compounds. Plenum, New York. p. 13-27

De Souza, M. P. and D. C. Yoch, Purification and characterization of
dimethylsulfoniopropionate Iyase from an Alca/igenes-like

dimethyl sulfide-producing marine isolate. 1995. Applied and
Environmental Microbiology. 61(1): p. 21-26

Dickson, D. M., et al., Steady state osmotic adaptation in Ulva lactua.
1980. Planta 150: p. 158-165

Dickson, D. M. J. and G. O. Kirst. The role of
fl-dimethylsulfoniopropionate, glycine betaine and homoarine in the
osmoacclimation of Platymonas subcordiformis. 1986. Planta 1 67:
p. 536-543

Edwards, D. M., et al., Organic solute accumulation in osmotically-
stressed Enteromorpha intestinalis. 1987. Marine Biology. 95: p. 583-592

Fall, R., et al., Laboratory studies of some environmental variables
controlling sulfur emissions from plants. 1988. Journal of Atmospheric
Chemistry. 6: p. 341-362

Fuhnnan, J. A. and C. A. Suttle, Vimses in marine planktonic systems.
1993. Oceanography 6(2): p. 51-63

Gage, D. A. and A. D. Hanson, Characterization of
3-dimethylsulfoniopropionate (DMSP) and its analogs with mass
spectrometry. 1996. In R.Kiene, P. Visscher, G. Kirst, M. Keller eds.
Environmental and biological chemistry of DMSP and related
sulfonium compounds. Plenum, New York. p 29-44

Gage, D. A., Rhodes, D., Nolte, K. D. Hicks, W. A., Leustek, T., Cooper,
a. J. L. and A. D. Hanson, A new route for the synthesis of
dimethylsulfoniopropionate in marine algae. 1997. Nature. 387: p. 891-894

Gage, D. A. and B. Rathinasabapathi, Role of glycine betaine and
dimethylsulfoniopropionate in water stress tolerance. 1999. In
Molecular Responses to Cold, Drought, Heat and Salt Stress in
Higher Plants. p. 125-152

Galleron, C. Synchronization of the marine dinoflagellate, Amphidinium
carterae in dense cultures. 1978. Journal of Phycology. 12: p. 69-73

115

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Galleron, C. and a. M. Durrand, Cell cycle and DNA synthesis in a marine
dinoflagellate Amphidinium carterae. 1979. Protoplasma. 100: p. 155-165

Gekko, K. and S. Timasheff, Mechanism of protein stabilization by
glycerol: Preferential hydration in glycerol-water mixtures. 1981.
Biochemistry 20: p. 4667-4676

Gekko, K. and S. Timasheff, Thermodynamic and kinetic examination
of protein stabilization by glycerol. 1981. Biochemistry 20: p4677-4686

Giovanelli, J. H., S. Mudd, and A. Datko, Recycling of methionine sulfur in
a higher plant by two pathways characterized by either loss or retention of
the 4-carbon moiety. 1981. Biochemical and Biophysical Research
Communications. 100(2): p. 831-839

Giovanelli, J. H., S. Mudd, and A. Datko, Quantitaive analysis of pathways
of methionine metabolism and their regulation n Lemna. 1985. Plant
Physiology. 78: p. 555-560

Givan, C. V. Aminotransferases in higher plants. 1980. In B. J. Mifflin ed.
The Biochemistry of Plants. 5: p. 329-357

Gold. K. and Barren, C. F. Growth requirements of Gyrodinium cohnii.
1966. Journal of Protozoology. 13: p. 255-257

Greene, R. C. Biosynthesis of dimethyl-fl-propiothetin. 1962. The Journal
of Biological Chemistry. 237(7): p. 2251-2254

Grone, T. and G. O. Kirst, Aspects of dimethylsulfoniopropionate effects
on enzymes isolated from the marine phytoplankter Tetraselmis
sucordifonnis (Stein) 1991. Journal of Plant Physiology. 1 38 p. 85-91

Grone, T. and 6. O. Kirst, The effect of nitrogen deficiency, methionine
and inhibitors of methionine metabolism on the DMSP contents of
Tetraselmis subcordiforrnis (Stein) 1992. Marine Biology. 497: p 497-503

Guillard, R. R. L. and M. P. Keller, Culturing Dinoflagellates. 1984. In
David L. Spector ed. Dinoflagellates. Academic Press, Orlando Fla. p
391-441

Hanson, A. D. and D. A. Gage, Identification and determination by fast
atom bombardment mass spectrometry of the compatible solute choline-
O-sulfate in Limonium species and other halophytes. 1991 . Australian
Journal of Plant Physiology. 1 8: p. 317-327

116

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

Hanson, A. D. et al, Osmoprotectant compounds in the Plumbaginaceae:
A natural experiment in metabolic engineering. 1994. Proceedings of the
National Academy of Sciences USA. 91: 306-310

Hanson, A. D., et al., Biosynthesis of 3-dimethylsulfoniopropionate in
Wollastonia biflora (L.) DC. 1994. Plant Physiology. 105: p. 103-110

Hare, P. D., Cress, W. A. and J. Van Staden, Dissecting the roles of
osmolyte accumulation during stress. 1998. Plant Cell and Environment.
21: p. 535-553

Harrison, P., J. Waters, R. E. and J. R. Taylor, A broad spectrum artificial
seawater medium for coastal and open ocean phytoplankton. 1980.
Journal of Phycology. 16: p. 28-33

Huang, 2. H., et al., Analysis of acylcamitines as their N-demethylated
ester derivatives by gas chromatography-chemical ionization mass
spectrometry. 1991. Analytical Biochemistry. 199: p. 98-105

Huang, 2. H. et al., Characterization of N-ethoxycarbonyl ethyl esters of
amino acids by mass spectrometry. 1993. Journal of Chromatography.
635: p. 271-281

James, F ., Nolte, K. D., and A. D. Hanson, Purification and properties of
S-adenosyI-L-methionine:Lmethionine S-methyltransferase from
Wollastonia biflora leaves. 1995. The Journal of Biological Chemistry.
270(28): p. 22344-22350

James, F., et al., Evidence implicating dimethylsulfoniopropionaldehyde as
an intermediate in dimethylsulfoniopropionate biosynthesis. 1995. Plant
Physiology. 108: p. 1439-1448

Karsten, U., Wiencke, C., and G. O. Kirst, The effect of light intensity and
daylength on the ,B-dimethylsulfoniopropionate (DMSP) content of marine
green macroalgae from Antarctica. 1990. Plant Cell and Environment. 13:
p. 989-993

Karsten, U., Wiencke, C., and G. O. Kirst, The
fi-dimethylsulfoniopropionate (DMSP) content of macroalgae from
Antarctica and southern Chile. 1990. Botanica Marina. 33: p. 143-146

Ka rsten, U., et al., Dimethylsulfoniopropionate production in phototrophic
organisms and its physiological function as a cryoprotectant. 1996. In
R.Kiene, P. Visscher, 6. Kirst, M. Keller eds.

Environmental and biological chemistry of DMSP and related

sulfonium compounds. Plenum, New York. p. 143-153

117

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Keller, S. E., Hutner, S. H., and D. E. Keller, Rearing the colorless
dinoflagellate Crypthecohdinium cohnii for use as a biochemical tool.
1968. Journal of Protozoology. 1 5(4): p. 792-795

Keller, D. M., et al., Media for the culture of oceanic ultraphytoplankton.
1987. Journal of Phycology. 23: p. 633-638

Keller, M. D., Bellows, W. K., and Guillard, R. L., Dimethyl sulfide
production in marine phytoplankton. 1989. In E. S. Saltzman, W. J.
Cooper, eds, Biogenic Sulfur in the Environment. American Chemical
Society, Washington 0.6 p. 167-182

Keller, M. D. and Bellows, W. K., Physiological aspects of the production
of dimethylsulfoniopropionate (DMSP) by marine phytoplankton. 1996. In
R.Kiene, P. Visscher, 6. Kirst, M. Keller eds.

Environmental and biological chemistry of DMSP and related

sulfonium compounds. Plenum, New York. p. 131-142

Kinne, R. H., The role of organic osmolytes in osmoregulation from
bacteria to mammals. 1993. The Journal of Experimental Zoology. 265:
p. 346-355

Koscis, M. 6., et al, Dimethylsulfoniopropionate biosynthesis in Spartina
altemiflora. 1998. Plant Physiology. 1 1 7: p. 273-281

Lavine, T. F., Floyd, N. F. and M. S. Cammaroti, The formation of
sulfonium salts from alcohols in sulfuric acid. 1953. p. 107-117

Le Rudulier, D., et al., Molecular basis of osmoregulation. 1984.
Science. 224: p. 1064-1068

Livesey, 6., Methionine degradation: anabolic and catabolic. 1984. trends
in Biochemical Sciences. January 1984. p. 27-29

Lund, J. W. 6., Kipling, C., LeCren, E. D. The inverted microscope method
of estimating algal numbers and the statistical basis of estimations by
counting. 1 958. Hydrobiology. 1 1 : p. 143-1 70

Malin, 6., The role of DMSP and DMS in the global sulfur cycle and
climate regulation. 1996. In R.Kiene, P. Visscher, G. Kirst, M. Keller eds.
Environmental and biological chemistry of DMSP and related

sulfonium compounds. Plenum, New York. p. 177-189

118

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

Mason, T. 6., and Blunden, G. Quatemary and tertiary sulfonium
compounds of algal origin as alleviators of osmotic stress. 1989. Botanica
Marina. 32: p. 313-316

Mazelis, M., and L. lngraham, The pyridoxal phosphate-depndent
oxidative decarboxylation of methionine by peroxidase. 1962. The Journal
of Biological Chemistry. 237: p. 109- 112

McNeil, S. D. M. L. Nuccio, and A. D. Hanson, Betaines and Related
Osmoprotectants. Targets for metabolic engineering of salt stress
Resistance. 1999. Plant Physiology. 120: p. 945-949

Morel, F. M. M., and R. R. L. Guillard, Aquil: A chemically defined
phytoplankton culture medium for trace metal studies. 1979. Journal of
Phycology. 15: p. 135-141

Mudd, H. S., and A. H. Datko, Methionine methyl group metabolism in
Lemna. 1986. Plant Physiology. 81: p. 103-1 14

Mudd, H. S., and A. H. Datko, The S-methylmethionine cycle in Lemna
paucicostata. 1990. Plant Physiology. 93: p. 623-630

Nakamura. H. et al., Biogenesis of sulfonium compounds in a
dinoflagellate; methionine cascade. 1997. Tetrahedron. 27: p. 9067-9074

O Brien, M C., and P. A. Wheeler, Short term uptake of nutrients by
Enteromorpha Prolifera (Chlorophyceae) 1987. Journal of Phycology.
23: p. 547-556

Paquet, L., et al., Accumulation of the compatible solute 3-
dimethylsulfoniopropionate in sugarcane and its relatives, but not other
Gramineous crops. 1994. Australian Joumal of Plant Physiology. 21: p.
37-48

Pokomy, M., E. Marcenko, and D. Keglevic, Comparative studies of L- and
D-methionine metabolism in lower and higher plants. 1970.
Phytochemistry. 9: p. 2175-2188

Ranocha, P., et al., The S-methylmethionine cycle in angiosperrns:
ubiquity, antiquity and activity. 2001. The Plant Journal. 25: .p. 575-584

Rhodes, D. and A. D. Hanson, Quatenary ammonium and tertiary

sulfonium compounds in higher plants. 1993. Annual review of Plant
Physiology. 44: p. 357-384

119

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

Rhodes, D. et al., S-Methylmethlonine conversion to
dimethylsulfoniopropionate: evidence for an unusual transamination
reaction. 1997. 115: p. 1541-1548

Russell, B. L., Rathinasabapathi, and A. D. Hanson, Osmotic stress
induces expression of choline monooxygenase in Sugar Beet and
Amaranth. 1998. Plant Physiology. 11 6: p. 859-865

Russell, D. W. and A. 6. Howard, The determination of DMSP in marine
algae and salt marsh plants. 1996. In R.Kiene, P. Visscher, G. Kirst, M.
Keller eds. Environmental and biological chemistry of DMSP and related
sulfonium compounds. Plenum, New York. p. 155-163

Scislowski, P. W. D., et al., Methionine metabolism by rat muscle and
other tissues. 1987. Biochemical Journal. 247: p. 35-40

Sciuto S., Piatelli, M., and Chillemi, R., (-)-(S)-4-Dimethylsulfonio-2-
methoxybutyrate from the red alga Rytiphloea tinctoria. 1982.
Phytochemistry. 21(1): p. 227-228

Shen. B., et al., Roles of sugar alcohols in osmotic stress adaptation.
Replacement of glycerol by mannitol and sorbitol in yeast. 1999. Plant
Physiology. 121: p. 45-52

Shrift, A., Nitrogen and sulfur changes associated with growth uncoupled
from cell division in Chlorella vulgaris. 1959. Plant Physiology. p. 505-512

Spector, D. L., Dinoflagellates. In Dinoflagellates. 1984. Academic Press.
Stevenson, D. E., Akhtar, M., and Gani, D., Streptomyces L-methionine
decarboxylase. Enzyme and stereochemical course of substrate
decarboxylation. 1990. Biochemistry. 29: p. 7660-7666

Steudtler, P. A., and B. J. Peterson, Contributions of gaseous sulfur from
salt marshes to the global sulphur cycle. 1984. Nature. 31 1: p. 455-457

Storey, R. et. al., Response of Melanthera biflora to salinity and water
stress. 1993. Journal of Experimental Botany. 44(267): p. 1551 -1560

Summers, P. S., et al., Identification and stereospecificity of the first three
enzymes of 3-dimethylsulfonipropionate biosynthesis in a Chlorophyte
algae. 1998. Plant Physiology. 1 16: p. 369-378

Sunda, W. G., and S. A. Huntsman, Interrelated influence of iron, light and
cell size on marine phytoplankton growth. 1997. Nature. 390: p. 389-392

120

82.

83.

85.

86.

87.

88.

89.

90.

91.

92.

Tarczynski, M. C. Jensen, R. 6., and H. J. Bohnert., Stress production of
transgenic tobacco production of the osmolyte mannitol. 1991. Science.
259: p. 508-510

Trossat, C., Nolte, K. D., and A. D. Hanson, Evidence that the pathway of
dimethylsulfoniopropionate begins in the cytosol and ends in the
chloroplast. 1996. Plant Physiology. 111: p. 1-9

Turner, S. M., Malin, G., and P. Liss, Dimethylsuflide and
(dimethylsulfonio) propionate in European coastal and shelf waters.
1989. In E. S. Saltzman, W. J. Cooper, eds, Biogenic Sulfur in the
Environment. American Chemical Society, Washington DC. p. 183-200

Tuttle, R. C., and A. R. Loeblich, An optimal growth media for the
dinoflagellate Crypthecohdinium cohnii. 1975. Phycologia. 14: p. 1-8

Uchida, A., et al., Biosynthesis of dimethylsulfoniopropionate in
Crypthecohdinium cohnii (Dinophyceae) 1996. In R.Kiene, P. Visscher, G.
Kirst, M. Keller eds.Environmental and biological chemistry of DMSP and
related sulfonium compounds. Plenum, New York. p. 97-107

Walters, D. S. and Steffens, J. C. Branched chain amino acid metabolism
in the biosynthesis of Lycopersicon penneli glucose esters. 1990. Plant
Physiology. 93: p. 1544-1551

Wang, J., et al., Analysis of amino acids by gas chromatography-flame
ionization detection and gas chromatography-mass spectrometry:
Simultaneous derivatization of functional groups by an aqueous-phase
chloroforrnate-mediated reaction. 1994. Journal of Chromatography. 663:
p. 71 -78

Wolfe, G. and M. Steinke, Grazing-activated production of dimethylsulfide
(DMS) by two clones of Emiliana huxleyi. 1996. Limnology and
Oceanography. 41(6): p. 1151-1160

Wolfe, G. M. Steinke, and G. O. Kirst, Grazing activated chemical
defence in a unicellular marine alga. 1997. Nature. 387: p. 894-897

White, R. H., Analysis of dimethyl sulfonium compounds in marine algae.
1982. Journal of Marine Research. 40(2): p. 529-535

Yancey, P., et al., Living with water stress: evolution of osmolyte systems.
1982. Science. 217: p. 1214-1222

121

93. Zappia, V., et al., The specificity of S-Adenosylmethionine derivatives in
methyltransfer reactions. 1969. The Journal of Biological Chemistry.
244(16): p. 4499-4509

122

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llll111111llllllllllllljlll