mg!

m ":53

f
“9.13::
wunzx

,
1z3i=

..u.
hrun
JL».«.- .. -
Jn ,..
h.«‘lu“n

“‘3‘

..-..-
«~ zazzca ~

«5.

5'9"
'%§§E§::di:i+
53%; ’ 'z i:
'13;

Ear:&.g.
“war“.w.:;,.
... . .«=.£

-<:z,

W... .m—
.-",. x
..r.
:'~ ..4

H...» .
um.

.n.

t‘?¥1§i§f€

mun "-4

$51.“:

 

ducal” ‘

.. f. I”

A
- .

arri§a};'ii§’

{::?%&a

5’

,.{EE#§:F:
, _ .

Z?‘

1::

wfia'

,g

:f;
1!.
.g.
..

«~‘v

if
,
fzfieqvfl
w...—
a
"A
‘ ‘Iit
.-

..—’
«a
‘f
,;H. .

in;
q,..‘
’3

,

.. A
y. 1
.3.»
w vw
auav

"'I
. v~
:n . 7’

- ,,..
“um-
1.2V.

m

\

f‘ir

”’21
.1»

‘.
1&1
:9’5
‘u
151:
fiQ.~
_....

I

%
zfi7P L
$512523?
“ 'ng§
‘ .

i
0!

in..."

'55?

. g'_
0‘ .
.

”‘41

w
{a :-

’3

-.~..
..--

. :71
..~.
~23

,.

P
J

. ,3.
,.
3?
3»

.,fl..»
.' 3m
.
‘€-=-
,. 5,_

'27
..e15‘

as“..-

t

4v. -... m.
”M
.4

w
u-

-.n»
4' ',“..a.~
‘ .3"
V. ‘ o
2" ‘13"
v .
é."

.....
.
v‘a‘A
..
.-
«s-

V¥IX

{ESS
:9
?

23 3i
‘§!1 1‘1111‘ if

u
I

«.3.

.,..‘~.
‘ -O¢
.....
\.. -

, .
.
..«
°~:2
’5“

1
~
'3

u

3" 1
4' \,
infig
v
u
G
.

\
.

..
..

inn

..:s?

I
I

.
‘fly

 

i?

fhréii‘fi

This is to certify that the

dissertation entitled

HOW PLANTS MANAGE THEIR FATTY ASSETS:
A STUDY INTO THE ORGANIZATION AND REGULATION
OF THE PLANT FATTY ACID BIOSYNTHETIC PATHWAY

presented by

David Ken Shintani

has been accepted towards fulfillment
of the requirements for

Ph.D. Botany

degree in

Q%

Major professor

 

 

4/4/(74

 

 

LIBRARY
Mlchlgan State }
University

 

 

 

PLACE IN RETURN BOX to remove thle checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU to An Afflrmettve AottorVEquel Opportunity lnetltwon
Wm:

 

 

HOW PLANTS MANAGE THEIR FATTY ASSETS:
A STUDY INTO THE ORGANIZATION AND REGULATION OF
THE PLANT FATTY ACID BIOSYNTHETIC PATHWAY

By

David Ken Shintani

A DISSERTATION

Submitted to
Michigan State University
in partial ful ment of the reqmrements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Botany and Plant Pathology
1996

ABSTRACT
HOW PLANTS MANAGE THEIR FATTY ASSETS:
A STUDY INTO THE ORGANIZATION AND REGULATION OT
THE PLANT FATTY ACID BIOSYNTHETIC PATHWAY.

By
David Ken Shintani

This dissertation represents the cumulative results from several individual
projects that pertained to different aspects of the organization and regulation of
plant fatty acid biosynthesis. The role of multiple isoforms of acyl carrier protein
(ACP) in plant fatty acid synthesis was addressed in studies where the
Arabidopsis ACPs were characterized. Through these studies, the predominant
leaf ACP isoform (LMI ACP) was purified and was used to identify the
corresponding cDNA clone. The tissue specific pattern of LMI ACP expression
was determined and the possible role of specific ACP isoforms in directing
mamrefattyacidstowards difi‘erentmetabolicfateswasdiscussed These studies
also lead to identification and characterization of an Arabidopsis mitochondrial
ACP. Immunolocalization and protein import studies were used to confirm the
identity the mitochondrial ACP. Furthermore, in vitro fatty acid synthase assays
indicated that the mitochondrial ACP was able to function as a co—factor for fatty
acid synthesis and that this ACP isoform appeared to facilitate the synthesis of
medium chain fatty acids. Studies were also conducted to look at the effect of
exogenously applied lipid on tobacco cell fatty acid synthesis. These studies
revealed that plant fatty acid synthesis is regulated by a product feedback
mechanism Analysis of acyl-ACP intermediates of fatty acid synthesis implied
that acetyl-CoA carboxylase (ACCase) was an important regulatory enzyme
involved in the feedback inhibition of plant fatty acid synthesis. Additionally,

it was determined that the feedback inhibition was most likely occurring through
a biochemical mechanism. Because these and other in viva findings suggested
that the plastid ACCase was a key enzyme involved in controlling the flux
through fatty acid biosynthesis, experiments were performed to alter the level of
ACCase in tobacco. The expression of the tobacco biotin carboxylase subunit.
of the plastid ACCase was altered using over- and antisense-expression
approaches in transgenic plants. The effect of altered biotin carboxylase on the
fatty acid synthesis pathway flux and on ACCase complex formation was
addressed in these studies. The combined results presented in this dissertation
have advancedomrmderstandingofplantlattyacidmetabolismand opened new
avenues of research in the plant lipid field.

To my parents, Kazumi and Doris Shintani,
whose sacrifices allowed me the luxury

of pursuing a career in science.

iv

ACKNOWLEDGMENTS

Without the support of my best-friend and wife, Christie Howard, I would
not have finished my PhD. It is through her strength and encouragement that I
was able to make it though the many low points of graduate school and fully
enjoy the high points. Our many late night scientific discussion before going to
sleep contributed significantly to the work presented in this dissertation. I want
to thank Christie for putting up with my poor moods and for picking up the slack
at home when I wasn’t able to (or was to lazy to) help out. I would also like to
thank Christie for reminding me that there is life outside of the laboratory and for
encouraging me to pursue that life.

I would like to thank John Ohlrogge for his willingness to take a chance
on me and accepting me into his laboratory. I have grown both intellectually and
personally through John’s guidance. He has been the ideal mentor who, though
his contagious enthusiasm, encouraged me to work hard and excel. I am grateful
to John for allowing me to pursue my own “hair-brained” ideas and his
willingness to discuss issues that went against the grain of the current dogma. I
would also like to thank Carol Ohlrogge for her kindness and generosity in
inviting Chrisitie and me into her home on several occasions. We have
considered Carol and John as part of our surrogate family since we have been
in Michigan and are grateful for their support.

My time in the Ohlrogge lab has been very rewarding and this is primarily
due to the many friendships that I have established with the past and present

V

members of this lab. I would like to thank Edgar Cahoon for teaching me how
to be a graduate student and for his generosity with his expertise in lipid
chemistry. Keith Roessler has been a good friend and the one person that I love
to argue with about anything. I also have enjoyed working with Xiaoming Bao,
who is also a good friend and good source of cigarettes. I would like to thank
Linda Savage for listening to me bitch about everything and nothing at all, Nicki
Engeseth for her sympathetic ear, Kathy Schmid for encouraging me to do things
that I didn’t feel like doing (like writing or preparing for talks), Sergei
Mekhedov for his bad Russian jokes, Basil Shorrosh for many intellectual
discussions, Sarah Hunter for fixing the computers after I screwed them up,
Dusty Post-Beittenmiller for her teaching me to think on my feet, Alenka
Housek-Radoj cic for her wonderful cakes, Peter Dormann for his biochemical
expertise, Paul Roesler for his promising to hire me to work in his hardware
store, Ray Pacovsky for his incredible energy, Haj ime Wada for continuing the
mtACP work, Vicki Eccleston for her good mood, and everybody else (i.e. Kuni
Nakahira, Adrian Butt, Oscar de llarduya, Becky Peck, James and Amy, Grattan
Roughan, Abi, Arif, and anyone I may have forgotten). I would also like to thank
Steve Schwartz, Susan Fuj imoto, Beth Rosen, Todd and Laura Black, John
Shanklin and Jim Dombrowski for their friendship during our stay at MSU. I
would also like to thank Jan J aworski for several helpful discussions during his
many visits to East Lansing.

I would also like to thank Pam Green for her support as a committee
member and, for her generosity in allowing me to use her facilities and
equipment. I would also like to thank Natasha Raikhel, Ray Hammerschmidt,

and Lee Macintosh for their guidance as members on my graduate committee.

TABLE OF CONTENTS

List of Figures .................................................................................................... x
Chapter 1: Introduction ...................................................................................... 1
The role of fatty acids in cellular metabolism ............................................ 1
Biochemical reactions of fatty acid synthesis ............................................ 2
Acetyl-CoA Carboxylase ......................................................................... 4
Fatty Acid Synthase ................................................................................. 5
1. Organization of the Fatty Acid Synthase Complex ...................... 5
Type I F AS Complexes ........................................................ 5
Type H FAS Complexes ....................................................... 7
H. Subcellular Localization of the Fatty Acid Synthase Complex. . 10
Regulation of Plant Fatty Acid Synthesis ................................................ 12
Biochemical Regulation of Fatty Acid
Synthesis in Plants .............................................................. 13
I. Light/Dark Regulation of Plant Fatty Acid
Synthesis .................................................................. 1 5
11. Product Feedback Inhibition of Plant Fatty Acid
Synthesis .................................................................. 16
III. Role of ACCase in Controlling the Flux Through
Fatty Acid Synthesis in Maize Leaves ...................... 17
IV. Molecular Approaches to Understanding the
Biochemical Regulation of Fatty Acid Synthesis ....... 18
Genetic Regulation of Fatty Acid Synthesis in Plants .................... 19
Literature Cited ...................................................................................... 21

vii

Chapter 2: The Protein Pmification and cDNA Cloning of the

Predominant Acyl Carrier Protein Isoform in Arabidopsis Leaves

Abstract ............................................................................................... 27
Introduction ............................................................................................ 28
Materials and Methods ........................................................................... 31
Results and Discussion ........................................................................... 35
Conclusion ............................................................................................. 47
Literature Cited ...................................................................................... 51

Chapter 3: The Characterization of a Mitochondrial Acyl Carrier

Protein Isoform Isolated From Arabidopsis thaliana

Abstract ................................................................................................. 53
Introduction ............................................................................................ 55
Materials and Methods ........................................................................... 56
Results and Discussion ........................................................................... 63
Conclusion ............................................................................................. 83
Literature Cited ...................................................................................... 85

Chapter 4: Feedback Inhibition of Fatty Acid Synthesis in

Tobacco Suspension Cells

Abstract ................................................................................................. 89
Introduction ............................................................................................ 90
Materials and Methods ........................................................................... 91
Results and Discussion ........................................................................... 96
Conclusion ........................................................................................... 1 17

viii

Chapter 4 (cont)
Literature Cited .................................................................................... 119

Chapter 5: Antisense- and over-expression of biotin carboxylase

in tobacco leaves.

Abstract ............................................................................................... 122
Introduction .......................................................................................... 123
Materials and Methods ......................................................................... 125
Results and Discussion ......................................................................... 128
Conclusions .......................................................................................... 142
Literature Cited .................................................................................... 144

Chapter 6: Conclusions and Future Research Perspectives

Understanding the role of multiple ACP isoforrns in plants ................... 147
Mitochondrial ACP and fatty acid synthesis ......................................... 149
Product feedback regulation of plant fatty acid synthesis ...................... 154
Antisense- and over-expression of the biotin carboxylase

gene in tobacco leaves ................................................................ 159
Literature Cited .................................................................................... 163

ix

LIST OF FIGURES

Figure 2-1: Purification of LMI ACP from Arabidopsis leaf ............................. 37

Figure 2-2: Nucleotide and deduced amino-acid sequence of the
cDNA clone corresponding to the LMI ACP ................................ 39

Figure 2-3: Immuno-blot analysis of E. coli expressed LMI ACP constructs
and ACP extracts from different Arabidopsis tissues ................... 42

Figure 2-4: Accumulation of LMI-ACP mRNA in different
Arabidopsis tissues ...................................................................... 45

Figure 2-5: Alignment of the LMI ACP protein sequence with
other Arabidopsis plastid ACPs and comparisons
of the transit peptide cleavage sites of other
Arabidopsis plastid ACPs ............................................................ 46

Figure 2-6: Hydrophobicity analysis of the Arabidopsis
ACP mature protein sequences ..................................................... 49

Figure 3-1: Nucleotide and deduced amino-acid sequence of pACP25 .............. 65

Figure 3-2: Alignment of mtACP-l protein sequence with
known mitochondrial ACPs and with the
Arabidopsis chloroplast ACPs ...................................................... 67

Figure 3-3: In vivo protein import of mtACP-l

precursor into pea leaf mitochondria ............................................. 72

Figure 3-4: In vitro protein import studies of the mtACP-l

precursor into spinach chloroplast ................................................ 74

Figure 3-5: Westem—blot analysis of Arabidopsis leaf
sub—cellular fiactions .................................................................... 76

Figure 3-6: Western-blot analysis of mitochondrial

membrane and matrix fi'actions ..................................................... 78

Figure 3-7: Utilization of mtACP-l and spinach
cpACP-l by chloroplast FAS ....................................................... 79

Figure 3-8: Product analysis of fatty acid synthesis reactions
supplemented with mtACP-l or spinach cpACP-l ........................ 82

Figure 4-1: Effect of Tween-18:0 and Tween-18:1 on
tobacco cell growth and acetate incorporation

into total lipid fractions ................................................................ 99

xi

Figure 4-2: The effect of Tween-18:1 on the rate of
[1-“C]acetate incorporation into fatty acids

of tobacco cells .......................................................................... 101

Figure 4-3: Effect of Tween-18:1 on [1-“C]acetate incorporation into
long chain acyl-ACPs of tobacco cells ....................................... 106

Figure 4-4: Effect of Tween-18:1 on [1-”C]acetate incorporation into
medium chain acyl-ACP of tobacco cells .................................... 109

Figure 4-5: Analysis of the free-, acetyl-, and malonyl-ACP pools

from control and Tween-18:1 treated tobacco cells ..................... 1 12

Figure 4-6: Irnmuno-blot analysis of ACCase in extracts isolated from
control and Tween-18:1 treated cells .......................................... 116

Figure 5-1: Biotin carboxylase levels in transgenic tobacco leaves .................. 129
Figure 5-2: Comparison of leaf BCCP and biotin carboxylase

levels in transgenic tobacco leaves containing

altered levels of biotin carboxylase ............................................. 133
Figure 5-3: Growth of wild type plants and plants

expressing increased and decreased levels of

biotin carboxylase under low light conditions ............................. 135

xii

Figure 5-4: Leaf fatty acid content of wild type, A5, and A11 plants
grown under low light and high light conditions .......................... 139

Figure 5-5: Plants that are homozygous for the A11
transgene locus show abnormal floral development .................... 141

xiii

Chapter 1
Introduction

The role of fatty acids in cellular metabolism

Fatty acids are essential compounds needed by almost all biological
systems and are involved in several cellular processes. The majority of the fatty
acid present in the cell is associated with membrane lipids. Fatty acids make up
the hydrophobic component of membrane lipids, whose physical properties, such
as melting temperature and packing volume, are greatly influenced by the chain
length and the degree of desaturation of the attached fatty acid groups.

Although fatty acids are primarily recognized as structural components of
membrane lipids, they are also important in many other cellular processes. Fatty
acids are precursors for the synthesis of storage compomrds such as fats, oils and
waxes (Ohlrogge, 1994; Battey et al., 1989). In plants, fatty acid precursors are
necessary for the production of epidermal waxes and cutin, substances that help
prevent desiccation of plant tissues and protect against pathogen attack
(Kolattukudy, 1987).

Fatty acids and their derivatives have also been implicated as signal
molecules involved in several signal transduction pathways. In animal systems,
fatty acids precursors are needed for the synthesis of prostaglandins, compounds
which have been implicated in a range of biological activities including the
perception of pain, tissue inflammation, and smooth muscle contraction (Smith
and Borgeat, 1985). J asmonic acid, a plant signaling molecule associated with
the induction of plant defense mechanisms, is synthesized from the fatty acid
linolenic acid (18:3) (Vick and Zimmerman, 1984).

Fatty acids are also involved in the post-translational modification of

2
specific proteins. Protein acylation has been shown to be involved in the
targeting and anchoring of proteins to specific membranes (Towler et al., 1988;
McIlhinney, 1990; Milligan et al., 1995). This process may also play a role in
facilitating the association of different proteins through the interaction of the
fatty acid chain of the acylated protein and a hydrophobic domain of a second
protein (Towler et al., 1988; McIlhinney, 1990).

Biochemical Reactions of Fatty Acid Synthesis

The biochemical reactions leading to the production of fatty acids have
been well studied and are known to occur by the same basic mechanism in all
organisms. The elongation of a growing fatty acid chain proceeds through the
step-wise addition of 2 carbons to a growing acyl chain. The source of the 2
carbon donor for the elongation of the growing fatty chain is derived from acetyl-
CoA. Acetyl-CoA is committed to fatty acid synthesis in the reaction catalyzed
by acetyl-CoA carboxylase (ACCase) which facilitates the ATP dependent
carboxylation of acetyl-CoA to yield malonyl-CoA. In a reaction catalyzed by
malonyl-CoAzACP transacylase (MCT) the malonyl-group from malonyl-CoA
is then transferred to the protein co-factor acyl carrier protein (ACP). ACP
provides the scaffolding from which the fatty acid molecule is built and acts as
a vehicle to transport the immature fatty acid chain between sequential
enzymatic steps in this biosynthetic pathway.

The elongation of the growing fatty acid chain involves the condensation
of an acyl-thioester primer with malonyl-ACP. This reaction is catalyzed by a
class of enzymes called B—ketoacyl-ACP synthases (KAS). The acyl-thioester
primer is transferred to a cysteine residue in the KAS active site which then
interacts with the 2 carbon donor malonyl-ACP. Although most of the

3
condensation reactions use saturated acyl-ACPs as primers for elongation,
acetyl-CoA serves as the primer in the first condensation reaction. In yeast and
animal FAS, the transfer of the acyl-thioester primer to KAS is catalyzed by the
enzyme acetyl-transacylase (ATA). KAS catalyzes the condensation reaction
which results in the release of CO2 and the formation of a B—ketoacyl-ACP
species that is 2 carbons longer than the reactant primer.

The [i-ketoacyl-ACP intermediate is reduced in a NADPH requiring
reaction catalde by B—ketoacyl-ACP reductase (KR) to form B—hydroxybutyrl—
ACP. Through the loss of H20, the hydroxyacyl-ACP intermediate is converted
to an enoyl-ACP by B-hydroxyacyl-ACP dehydratase. The enzyme, enoyl-ACP
reductase, then catalyzes the NADPH dependent reduction of the enoyl-ACP
intermediate which results in a saturated butyryl-ACP species. The plant—enoyl-
ACP reductase (ER) is able to use either NADPH or NADH in this reaction.

Depending on the organism, the FAS complex can catalyze seven or eight
rormds of elongation resulting in the formation 16 or 18 carbon acyl-ACPs. The
final reactions of fatty acid synthesis involve the removal of the mature fatty acid
chains fiom the ACP co-factor. Two different types of reactions are involved in
the termination of fatty acid synthesis. First, the acyl-ACP could be cleaved by
a acyl-ACP thioesterases to yield a free fatty acid molecule. Alternatively, the
acyl group could be transferred from the ACP to an acceptor molecule (such as
free CoA yielding an acyl-CoA molecule, or glycerol-3-phosphate (G-3-P)) to
form lyso-phosphatidic acid (the first intermediate of glycerolipid synthesis).
Animals terminate fatty acid synthesis via a thioesterase, while the termination
reaction in fungal fatty acid synthesis involves an acyl-transferase that transfers
the mature fatty acid to CoA. In bacteria, the mature fatty acid is transferred
from ACP to either the sn-1 or sn-2 position of glycerol. Plants terminate fatty

4
acid synthesis in reactions catalyzed by acyl-ACP thioesterases and

acyltransferases similar to those occurring in bacteria.

Acetyl-CoA Carboxylase

Acetyl-CoA carboxylase (ACCase) has been identified as an important
regulatory enzyme in bacterial (Magnuson et al., 1993), plant (Post-Beittenmiller
et al., 1991 & 1992; Page et al., 1994; Shintani & Ohlrogge, 1995) yeast
(Karniryo and Numa, 1973) and animal (V agelos, 1971) fatty acid synthesis. The
reaction catalyzed by ACCase commits carbon in the form of acetyl-CoA
towards fatty acid synthesis. ACCase is composed of three subunits: a biotin
carboxylase, a biotin carboxyl carrier protein (BCCP), and a transcarboxylase.
The formation of malonyl-CoA catalyzed by this enzyme occurs through a two
step mechanism. First, the biotin carboxylase subunit catalyzes the activation of
HCO3- by ATP which is then used to carboxylate a biotin prosthetic group of the
BCCP subunit. The transcarboxylase subunit then catalyzes the transfer of the
carboxyl group from the BCCP subunit to an acetyl-CoA molecule.

The animal, fungal, and bacterial ACCases are localized in the cytosol.
Plant ACCases are localized in both the plastid and the cytosol (Alban et al.,
1994; Sasaki et al., 1995). The malonyl-CoA produced by the plastid ACCase
is utilized primarily for de nova fatty acid synthesis, while the plant cytosolic
ACCase is utilized in the synthesis of phenylpropanoids (Ebel and Hahlbrook,
1977) and very long chain fatty acid synthesis (Ohlrogge et al., 1978). Plant
cytosolic derived malonyl-CoA has also been shown to be needed for the
inactivation of l-aminocyclopropane 1-carboxylate (ACC) (Kionka and
Amrhein, 1984).

The organization of the ACCase subunits differs between different

5

organisms. In animals (Lopez-Casillas et al., 1988), yeast (Al-Feel et al., 1992),
and diatoms (Livne and Sukenik, 1990; Roessler, 1990) ACCases are organized
as homomeric complexes of single multifirnctional polypeptides. The bacterial
ACCase is organized in a heteromeric complex composed of four soluble
polypeptides: the BCCP subunit, the biotin carboxylase subunit, and two
transcarboxylase subunits (Magnuson et al., 1993 and references within). The
organization of the plant ACCase has been an issue of great debate for several
years. Recent studies have shown that dicots and non-gramineaceous monocots
contain a plastid heteromeric ACCase and a cytosolic homomeric ACCase
(Alban et al., 1994; Shorrosh et al., 1995; Choi et al., 1995; Sasaki et al., 1995).
Gramineaceous monocots, on the other hand, contain homomeric ACCases in
both their plastids and cytosol (Egli et a1, 1995).

Fatty acid Synthase
1. Organization of the Fatty Acid Synthase Complex

The organization of the individual enzymatic activities that collectively
make up the fatty acid synthase (FAS) complex differs significantly among
different organisms (Bloch and Vance, 1977). The yeast and animal FAS
enzymes are organized as type I enzyme complexes where the individual
enzymatic activities are collectively organized on multifunctional polypeptides.
The plant and bacteria FAS enzymes are organized as type II complexes where
each enzymatic activity is associated with an individual soluble enzyme.

Type I FAS Complexes
The animal and fungal FAS enzymatic activities are organized on large
mutifimctional polypeptides and are referred to as type I enzyme complexes. The

6

animal FAS complex is organized as a homodimer with a native molecular mass
of greater than 500,000 (W akil, 1989). Each complex is composed of two
identical subunits with molecular weights of 260,000. Results from limited
proteolysis studies, showed that each animal FAS subunit was composed of
three major domains; designated as FAS domains I, II, and III; with respective
molecular weights of 127,000, 107,000, and 33,000 (Mattick et al., 1983a). The
largest domain, domain 1, contained the following enzymatic activities: [3-
ketoacyl synthase (KAS); malonyl transacylase (MCT); and acetyl transferase
(ATA) (Mattick et al., 1983a). Domain H contained the ACP protein co-factor,
and the D-ketoacyl reductase, B hydroxyacyl dehydratase, and enoyl reductase
activities (Mattick et al., 1983a). The 33 kD domain III, contained the
thioesterase activity (Mattick et al., 1983b).

The firngal FAS complex is composed of two different multifunctional
subunits, the 212,000 D rat-subunit and the 203,000 D [i-subunit (Stoops et a1,
1978). The tar-subunit contains the KAS, [i-ketoacyl reductase activities and the
ACP moiety (Stoops and Wakil, 197 8). The fi-subtmit contains the B-
hydroxyacyl dehydratase, enoyl reductase, MCT, ATA, and acyl-transferase
activities (Stoops and Wakil, 1978). The yeast native FAS complex has a «,0,
configuration with a molecular weight of 2,400,000 (Stoops and Wakil, 1980).

The animal and frmgal FAS complexes are capable of only 7 rounds of
elongation to produce the 16 carbon saturated acyl-ACP, palmitoyl-ACP. The
terminal reaction of the animal FAS involves the hydrolysis of the acyl-ACP
thioester linkage resulting in free palmitate. This fatty acid is rapidly converted
to an acyl-CoA through the action of acyl-CoA synthetase. The terminal reaction
of the fungal F AS involves an acyl-ACP transacylase which catalyzes the
transfer of the palmitoyl-group from ACP to CoA. In both animal and fungal

7
cells, the palmitoyl-CoA can then be directed towards glycerolipid synthesis or
be further elongated to stearoyl-CoA by a fatty acyl-CoA elongase that is
localized on the cytosolic face of the endoplasmic reticulum. The stearoyl-CoA
can then serve as a substrate for glycerolipid synthesis or undergo desaturation

by stearoyl-CoA desaturase to yield oleyl-CoA.

Type II FAS Complexes

The plant and bacterial FAS complexes are comprised of individual
soluble enzymes and are referred to as type H enzyme complexes (Vagelos,
1974; Shimakata and Stumpf, 1982a). Although type H FAS complexes may
appear to be less sophisticated then the animal or yeast type I FAS complexes,
their organization is certainly more complex. A functional type II FAS complex
requires the association of at least nine individual soluble enzyme activities.
These enzymes must be coordinated to catalyze the correct reaction in the
correct sequence. Additionally, plants and bacteria possess multiple isoforrns of
certain FAS components, some of which are responsible for specialized
fimctions.

It is not known how the individual FAS subunits are associated with one
another in a type H complex. Isolation of an intact FAS complex from bacterial
or plant cell extracts has proven difficult because FAS components are so easily
disassociated (Shimakata and Stumpf, 1982b). It is possible that the soluble
enzymes comprising the type II FAS complex are arranged as loosely associated
ordered complex. The components of other multienzyme pathways, such as the
citric acid cycle (Srere et al., 1987) and the Calvin cycle (Gontero et al., 1988;
Suss et al., 1993), have been reported to be organized in a sequential fashion to
form fimctional complexes. Although little evidence has been obtained to support

8

the hypothesis that the type H FAS enzymes are organized in an ordered
complex, results from in vitra plant FAS assays suggest that FAS components
in cell free extracts need to reassociate in order to function. It has been shown
that the addition of exogenous ACP is necessary to obtain high rates of FAS
from cell free extracts. Rutkoski and J aworski (1978) observed that when
exogenous ACP was added to a spinach chloroplast extract, a 15 minute lag
period was observed before 14C-malonyl-CoA incorporated into fatty acids. This
lag period was eliminated if the extract was pre-incubated with the exogenous
ACP for 15 minutes prior to the addition of label, suggesting that the lag period
was necessary for FAS complex formation.

Like the animal and yeast type I FAS, the bacterial and plant type H FAS
are able to synthesize palmitic acid. However, because type H FAS complexes
are much more fluid than the type I FAS, additional enzymes with specialized
functions are able to integrate into the type H complex and allow for the
production of a variety of different fatty acid products. For example, plants
possess a specific KAS isoform, KAS H that is responsible for the elongation of
palmitoyl-ACP (16:0-ACP) to stearoyl-ACP (18:0-ACP) (Shimakata and
Stumpf, 1982c). The plant 18:0-ACP can then be modified by a soluble acyl-
ACP desaturase that can introduce a single double bond at the A9 position to
produce oleoyl-ACP (18:1A9-ACP) (McKeon and Stumpf, 1982). Each of these
long chain acyl-ACPs can be cleaved by a thioesterase to produce palmitic acid
(16:0), stearic acid (18:0), and oleic acid (18:1A9) products (McKeon and
Stumpf, 1982). These products are then esterified to CoA and subsequently
serve as substrates for glycerolipid biosynthesis (Roughan and Slack, 197 7). E.
coli cells contain two KAS isoforms, KAS I and H, that are involved in the
synthesis of long chain saturated and mono-unsaturated fatty acids (Garwin et

9

al., 1980). Both E. coli KAS isoforms can catalyze the condensation reactions
leading to the synthesis of 16:0-ACP, however, KAS I is also capable of
elongating the mono-unsaturated 10 carbon fatty acyl-ACP, cis-3-decenoyl-
ACP, in a series of condensation reactions to produce the mono-unsaturated 16
carbon fatty acyl-ACP, palmitoleoyl-ACP (16:1A9-ACP). 16:1A9-ACP can then
be elongated by KAS H to form cis-vaccenoyl-ACP (18:1A11-ACP). The E. coli
long chain acyl-ACPs are then rapidly converted to acyl-CoAs by
acyltransferases and these acyl-CoAs subsequently enter gylcerolipid synthetic
pathways.

The type H configuration also allows plants and bacteria to alter the
production of certain fatty acid species in order to respond to environmental
changes. E. coli is known to alter the fluidity of their membranes in response to
fluctuations in temperature by modulating the synthesis of the two mono-
unsaturated fatty acids cis-vaccinic acid (18:1A11) and palmitoleic acid
(16:1A9) such that at elevated temperatures, the ratio of 18:1A11 to 16:1A9
levels decreases, while at low temperatures this ratio increases (Garwin and
Cronan, 1980). This change in E. coli membrane lipid fatty acid composition
was attributed to the thermoregulation of KAS H, the enzyme responsible for the
elongation of 16:1A9—ACP to 18: 1A1 l-ACP. It was shown that the increased
KAS II activity at lower temperatures was due to the enzymes unusually low
temperature optimum and not due to changes in the levels of KAS H mRNA or
protein. Plant cell suspension cultures have also been reported to alter the
synthesis of 18 carbon acid species in response to changes in temperature and
the addition of exogenous fatty acids (MacCarthy and Stumpf, 1980a). In plants,
18 carbon fatty acids make up the majority of the rmsaturated fatty acid species
while the saturated fatty acid, 16:0, is the most abundant 16 carbon fatty acid

10
species. It observed that plant suspension cultures altered the fluidity of their
membranes in response to increased temperature by decreasing the degree of
membrane lipid fatty acid desaturation. It was shown that KAS H was labile at
elevated temperatures which let to an increase in the ratio of 16 carbon to 18
carbon fatty acids (MacCarthy and Stumpf, 1980b).

The type H FAS configuration also allows for the introduction of FAS
enzymes with specialized functions into the normal FAS complex to produce
unusual fatty acid products. One such example can be seen in plants that produce
oils containing high levels of medium chain fatty acids. The production of lamic
acid (12:0) in the fruits of the California Bay Laurel tree has been attributed to
an rmusual acyl-ACP thioesterase that specifically cleaves fatty acids from acyl-
ACPs of 12 carbons in length. cDNA clones corresponding to the lauryl-ACP
thioesterase have been expressed in both E. coli (Ohlrogge et al., 1995) and
Brassica napus (Voelker et al., 1992), two species which do not normally
produce medium chain fatty acids. In both cases, it appeared that the medium
chain acyl-ACP thioesterase was able to integrate into the host FAS complex
and produce medium chain fatty acids. Additionally, Cahoon et al. (1992), were
able to express a coriander cDNA clone corresponding to a A4-palmitoyl-ACP
desaturase which was involved in the synthesis of an unusual fatty acid known
as petroselinic acid in tobacco cells and observe the synthesis of this unusual
fatty acid product.

11. Subcellular Localization of the Fatty Acid Synthase Complex

While the animal, fungal and bacterial FAS activities are predominantly
localized with in their respective cytosols, the plant FAS is predominantly
associated with the plastid compartment (Ohlrogge, 1979). The plastid

1 1

localization of a prokaryotic type FAS I complex in plant cells is thought to have
originated through the endosymbiotic relationship between an eukaryotic single
cell organism and a photosynthetic bacteria. For this reason, Anabaena, a
cyanobacterium which is known to contain a type H FAS, has been used by
some researchers as a model system to study plant fatty acid synthesis. Euglena,
a single cell green algae that contains both a cytoplasmic type I FAS and a type
H plastid localized FAS (Hendren and Bloch, 1979), may represent an organism
that is closely related to the hypothetical photosynthetic endosymbiotic
organism.

Evidence has been emerging for the presence of a type H FAS in the
mitochondria of animals, frmgi, and plants. The work of Brady and Mikolaj czk
(l 98 8) showed that Neuraspara crassa mitochondria contained a soluble low
molecular weight ACP. Antibodies made to the Neuraspara ACP, were reported
to cross- react with protein extracts isolated from plant mitochondria (Chuman
and Brody, 1989). Mitochondrial ACPs have been purified to homogeneity from
Neumspara (Sackmann et al., 1991) and bovine heart muscle (Runswick et al.,
1991) mitochondria and cDNA clones corresponding to these (Runswick et al.,
1991; Sackmann et al., 1991) and the Arabidopsis (Shintani and Ohlrogge,
1994) mitochondrial ACP have recently been cloned. Radiolabeling experiments
done with isolated Neurospora mitochondria suggest that these mitochondria
may be capable of de novo fatty acid synthesis (Mikolaj czyk and Brady, 1990;
Zenson et al., 1992). In further support of a mitochondrial fatty acid synthase,
Harrington et al. (1993), reported the identifitcation of a [i-ketoacyl synthase that
is localized within yeast mitochondria.

Why a mitochondrial localized FAS exists is unclear since it has been

shown that cytosolic FAS, in the cases of yeast and animals, and the plastid

12
FAS, in the case of plants, are sufficient to meet the cellular demand for fatty
acid synthesis. It has been proposed that a mitochondrial FAS may be involved
in the synthesis of mitochondrial specific lipids such as cardiolipin (Brody et al.,
1990). It is also possible that the mitochondrial ACP may be involved in cellular
processes other than fatty acid synthesis. Recently, researchers studying the
mitochondrial electron transport have reported that the mitochondrial ACP may
be a component of the NADqubiquinone oxidoredutase (complex I) (Runswick
et al., 1991; Sackmann et al., 1991). It has been shown in that the disruption of
the gene coding for the mitochondrial ACP lead to the inability of the
NADHqubiquinone oxidoreductase (complex I) to form (Schulte et al., 1994)
suggesting that the mitochondrial ACP may be a structural component of

complex I of the mitochondrial electron transport chain.

Regulation of Plant Fatty Acid Synthesis

Although the biochemistry of plant fatty acid synthesis has been
thoroughly studied, relatively little is known about the regulation of this
pathway. Plants appear to synthesize only enough fatty acids to satisfy the
demand of glycerolipid synthetic pathways. This is clearly exemplified by the
observation that the accumulation of rmesterified fatty acids is never observed
in plants. The high degree of FAS regulation acts to prevent the over production
of fatty acids which would be energetically wasteful and could lead to an
accumulation of fatty acids or acyl-CoAs whose detergent nature would be
detrimental to the cell. It appears that high rates of fatty acid synthesis are
primarily observed in tissues where high rates of glycerolipid synthesis are
occurring, such as developing oilseeds and rapidly dividing and expanding cells.
Although the rates of fatty acid synthesis are significantly lower in mature

l3

difl'erentiated cells, these rates appear to be sufficient for the maintenance and

repair membrane lipids.

Modes of fatty acid synthesis regulation in plants.

Plants fatty acid synthesis is regulated through both biochemical and
genetic mechanisms. The biochemical regulation of fatty acid synthesis allow
plants to respond immediately to changes in the demand for fatty acids caused
by short term fluctuations in environmental conditions or metabolite levels. The
genetic regulation of fatty acid synthesis allows the plant to respond to long term
changes in cellular demand for fatty acids by increasing the amount of the
biosynthetic machinery. Depending on the specific species or tissue, one or a .
combination of the two types of regulation may be involved in controlling fatty
acid synthesis.

Biochemical Regulation of Fatty Acid Synthesis

The rates of fatty acid synthesis in plants are strongly influenced by
factors efl‘ecting both the metabolic state of the cell and the fluidity of cellular
membranes. Fatty acid synthesis is an energetically costly process, such that the
synthesis of one mole of palmitic acid (16:0) requires 8 moles of acetyl-CoA, 7
moles of ATP, 7 moles of NADH, and 7 moles of NADPH.

Biochemical regulation typically occurs through covalent or allosteric
modifications which either enhance or attenuate the activity of specific
regulatory enzymes. These modifications efiecting enzymatic activity are usually
fully reversible. In contrast to the regulation of synthetic pathway by the
induction of gene expression, which occurs within the range of tens of minutes,

biochemical modes of regulation occur much more rapidly, within the range of

14
fractions of minutes.

Much effort has been expended to identify enzymes that regulate the flux
of carbon through fatty acid synthesis. Researchers have tried to identify the rate
limiting enzymes in fatty acid synthesis by comparing the in vitra activities of
FAS enzymes measured from plant extracts. Of the FAS enzymes surveyed, the
in vitro velocity of the reaction catalyzed by acetyl-CoAzACP transacylase
(ATA), was by far the slowest and was therefore putatively identified as the rate
limiting enzyme controlling fatty acid biosynthesis in plants (Shimakata and
Stumpf, 1983). However, later studies suggested that ATA may play only a
minor role in plant fatty acid synthesis and may actually catalyze a side reaction
of the condensing enzyme KAS IH (Jaworski et al., 1993). These findings
indicate that although in vitra measurements of enzymatic rates can sometimes
help identify regulatory enzymes within a pathway, in many cases the
conclusions based on these measurements may not apply to what is occurring in
viva.

Recently, in viva methods have been used to monitor changes in the levels
of FAS intermediates that occurred in response to changes in the flux of carbon
through the fatty acid synthetic pathway. These methods take advantage of the
fact that all FAS intermediates are covalently liked to the protein co-factor ACP.
Because acyl-ACP intermediates are extremely labile, with half lives in the range
of seconds, purification protocols had to be developed that would ensure the
stability of these intermediates (Post-Beittenmiller et al., 1991). The purified
acyl-ACP intermediates could be fiactionated on non-denaturing polyacrylamide
gels containing varying concentrations of urea. The relative mobility of
individual acyl-ACP intermediates in this gel system was shown to be highly
dependent on the nature of the attached acyl-group. The identity of these acyl-

15
ACP intermediates could then be determined by their co-migration with in vitro
synthesized acyl-ACP standards. By analyzing the accumulation or depletion of
specific acyl-ACP intermediate pools, it is possible to identify specific enzymatic
steps which were involved in the regulation of this pathway.

1. Light / Dark Regulation of Plant Fatty Acid Synthesis

One example of the biochemical regulation of fatty acid biosynthesis is the
light/dark regulation of fatty acid synthesis in leaves. It has been shown that the
rate of fatty acid synthesis is induced 5-7 fold when plants are transferred from
the dark into the light (Browse et al., 1981). The light activation of FAS occurs
within seconds after the transfer to the light and is rapidly reversed when plants
are transferred back to the dark.

When acyl-ACP and acyl-CoA pools were analyzed from light and dark
treated spinach leaves and chloroplasts, acetyl-CoA carboxylase (ACCase) was
identified as an important enzyme involved in the dark/light regulation of plant
FAS (Post-Beittenmiller et al., 1991 and 1992). These studies showed that the
levels of malonyl-CoA (the product of ACCase) and malonyl-ACP (which is
thought to be in equilibrium with malonyl-CoA) increased when leaves or
chloroplast that were incubated in the light (Post-Beittenmiller et al., 1991; Post-
Beittenmiller et al., 1992), indicating that ACCase activity was stimulated by the
light treatment. It has also been shown that ACCase activity is significantly
higher in extracts isolated fi'om When ACCase activity was measured rapidly

Although the mechanism by which ACCase is regulated by light is not
well tmderstood, Nakahira and Ohlrogge (manuscript in prep) showed that the
chloroplast electron transport system was somehow involved in the light

activation of ACCase in isolated spinach chloroplast. Using various

16

combinations of artificial electron donors and acceptors they showed that
electron flow between the cytochrome b/f-plastocyanin region and photosystem
I was required for the light activation of ACCase. Furthermore, they showed that
a membrane potential, but not a pH gradient across the thylakoid membrane was
required for the light activation of ACCase. In experiments where the thylakoid
membrane potential was dissipated by the addition of the specific K+ ionophore
valinomycin, the light activation of ACCase was not observed. However, when
the pH gradient across the thylakoid membrane was disrupted by the addition of
NH4C1, the light activation of ACCase was not effected. The relationship
between electron transport, the establishment of a membrane potential, and the
light activation of ACCase are not well understood.

Interestingly, the electron transport to photosystem I has also been shown
to be required for the activation of ribulose bisphosphate carboxylase
(RUBISCO) by RUBISCO activase (Campbell and Ogren, 1990). However, in
contrast to the light activation of ACCase, Campbell and Ogren (1990) showed
that the activation of RUBISCO required a pH gradient but does not require the
establishment of a thylakoid electron potential. It is possible that the light
activation of ACCase and RUBISCO may be occurring through similar
mechanism. However, as of yet, how electron transfer to photosystem I is
involved in the activation of either ACCase or RUBISCO is unknown.

II. Product Feedback Inhibition of Plant Fatty Acid Synthesis

A second example of biochemical regulation of plant fatty acid synthesis
can be seen during product feedback inhibition. It has been shown that the
addition of exogenous fatty acids in the form of Tween-fatty acid esters can

cause a significant decrease in the rates of fatty acid synthesis in soybean and

17

tobacco suspension cultures (T erzaghi et al., 1985; Terzaghi, 1986; Shintani and
Ohlrogge, 1995). Rates of fatty acid synthesis decreased by 3 - 5 fold within 15
minutes of the addition of oleoyl-Tween (Shintani and Ohlrogge, 1995). By
analyzing the pools of acyl-ACPs in Tween treated and control cells, it was
shown that ACCase played an important role in regulating fatty acid synthesis
during product feedback inhibition (Shintani and Ohlrogge, 1995). Due to rapid
onset of the inhibition and the observation that protein and mRNA levels of the
subunits of the plastid ACCase and FAS enzymes did not change during the
feedback inhibition, it appeared that the regulation of ACCase during product
feedback inhibition was probably occurring through a biochemical mechanism
rather than a genetic mechanism.

The identity of the feedback molecule was not determined, nor was the
mechanism by which ACCase was inhibited. What was clear from the analysis
of the acyl-ACP pools was that long chain acyl-ACPs were probably not acting
as the feedback molecules effecting ACCase. This assumption was based on the
observation that upon the onset of the feedback inhibition, the pools of long
chain acyl-ACPs become rapidly depleted. One might predict that if the long
chain acyl-ACPs were acting as feedback molecules, that the pools of these acyl-
ACPs would increase instead of decreasing as was observed. Furthermore, no
inhibition was detected in studies where long chain acyl-ACPs were added to
partially purified plastid ACCase (Roesler et al., 1995).

III. Role of ACCase in Controlling the Flux Through Fatty Acid Synthesis
in Maize Leaves.

Specific enzyme inhibitors have also been used to study potential
regulatory enzymatic steps in fatty acid synthesis. Page et al. (1994) studied the

1 8
effect of two highly specific inhibitors of the maize plastid ACCase, sethoxydim
and fluazifop, on rates of fatty acid synthesis. By correlating changes in rates of
ACCase activity due to the addition of either of the two inhibitors with changes
in the rate of fatty acid synthesis, they were able to show that ACCase exerted
a high degree of control fatty acid biosynthesis.

Molecular Approaches to Understanding the Biochemical Regulation of
Fatty acid Synthesis

Recently, many of the genes coding for components of the plant fatty acid
biosynthetic machinery have been cloned. This deve10pment should allow for
rapid advancements leading to a better understanding of the biochemical
mechanisms involved in the regulation of this central biosynthetic pathway.
Because it is now possible to alter levels of specific enzymes within the fatty
acid synthetic pathway in transgenic plants and determine what effect these
changes would have on pathway flux, we should now be able to use the
principles of metabolic control theory to test our assumptions about possible
regulatory points within the plant fatty acid synthetic pathway. The major idea
behind metabolic control theory is based on the assumption that the flux through
a given synthetic pathway is controlled through the cumulative control of the
sum of the pathway enzymes. Therefore, if the level of a particular enzyme is
changed, the degree of change in the over-all flux through the pathway caused
by this change in enzyme level reflects the degree of control that the enzyme has
over the entire pathway. So if a small change in the levels of a particular
enzymes causes large changes in the flux through the pathway, this enzyme has
large degree of control over the pathway. Alternatively, if it is necessary to make
large changes in enzyme levels to make large changes in pathway flux, then the

19
enzyme will have a low degree of control over flux through the pathway.

Genetic Regulation of Fatty Acid Synthesis in Plants

Several examples of the genetic regulation of fatty acid synthesis has
been shown to occur in plants. These include the increased gene expression of
FAS related proteins during the period of early leaf development and prior to the
period of oil deposition in oilseeds (Hannapel and Ohlrogge, 1988). In each of
these cases the change in the rate of fatty acid synthesis has been attributed to
an increase in the levels of FAS related enzymes.

Fatty acid synthesis has been reported to occur at significantly higher rates
in young, expanding spinach leaves than in older mature spinach leaves
(Kannagara et al., 1973). This increased rates of fatty acid synthesis
corresponded to increase in the level of F AS machinery, which was evident by
an increase in the mRNA and protein levels of ACP (Hannapel and Ohlrogge,
1988). Additionally, the increased rate of fatty acid synthesis observed in
developing seeds during the period of oil deposition has been shown to be due
to an increase in FAS related gene expression. The levels of ACP mRNA
(Hannapel and Ohlrogge, 1988) and protein (Ohlrogge and Kuo, 1984) were
shown to increase to high levels just prior to the onset of oil synthesis in
developing soybean seeds. Furthennore, in developing Brassica napus seed, the
profile of enoyl-ACP reductase and stearoyl-ACP desaturase mRN A
accumulation coincides with the period of increased FAS activity (F awcett et al.,
1994). These observations suggest that the genetic regulation of FAS during seed
development may involve the coordinate increase in the synthesis of the entire
FAS machinery.

So far very little is known about cis- or trans-acting factors involved in the

20

transcriptional induction of genes coding for components of fatty acid synthetic
machinery. This may be due to the fact that until recently, few of the genes
encoding proteins involved in fatty biosynthesis have been cloned from plants.
However, fairly extensive studies have been done to characterize the promoter
region of Arabidopsis ACP genes (Baerson and Lamppa, 1993; Baerson et al.,
1994) Through this work, cis elements have been identified in the promoter
region of a constitutively expressed Arabidopsis ACP gene that are necessary
for leaf, root, and seed expression (Baerson et al., 1994). Although
advancements in this area have been slow in coming as of late, the recent cloning
of many interesting cDNAs and genes involved in fatty acid synthesis (Topfer
and Martini, 1994) should lead to the rapid growth of this particular area.

Literature Cited

Al-Feel W, Chirala SS, Wakil SJ 1992) Cloninfilpf the yeast FAS3 one and
Emma-1315 3 4 5siz’tgucture of yeast acetyl-Co carboxylase. oc Natl Acad Sci SA 89:

Alban C Baldet P, Douce R (1994) Localization and characterization of two
structurally different forms of acetyl-CoA carboxylase 1n youn pea leaves of

\5N5h7ic151605ne is sensitive to aryloxyphenoxypropronate herbicides. rochem J 300:

Baerson SR, Lamp a GK (1993) Developmental regulation of an la/qyl carrier
riotgrsnsgzegp promo er in vegatrtrve and reproductrve trssues. Plant olec BlOI

Baerson .SR, Vanderheiden MG, Lamppa GK (1994) Identification of
domains in an Arabidopsis acyl camer rotern gene 6promoter required for
maxrmal organ-specrfic expressron. Plant olec B1012 : 1947-1959.

Batte JF Schmid KM Ohlro e JB 1989 Genetic e ' eerin for lant
oils: phtenfial and limitations. TIHTECthl 2-126. ngm g p

Bloch K, Vance D (197 7) Control mechanisms in the synthesis of saturated fatty
acrds. Ann Rev Brochem 46: 263-298.

Brody S, Mikolaj k S (198% Neuraspara mitochondria contain an acyl
carrier protem. Eur J rochem l : 353-3 9.

Brady S, Mikolajczyk S, Chuman L (1993)) Studies on de nova fatty acid
flfnthesrs. 1n mitochondria. In. PJ Quinn, Harwood, eds, Plant L1 1d
1igchemrstry, Structure, and Utrhzatron. Portland Press, Ltd, London, pp 1 7-

Browse J, Rougan PG, Slack CR (1981) Light control of fatty acid 5 thesis
Sgt; diurnal fluc trans of fatty acrd composrtron in leaves. Brochem J 1 6: 347-

Cam bell WJ, Ogren WL 1990 Electron transport through photosystem I
strm ates h t actrvatron o .rrb ose brsthosphate carborgylase/oxygenase
(RUBISCO) y RUBISCO activase. Plant hysrol 94: 479-4 4.

Cahoon EB, Shanklin J, Ohlrogge JB (1992) Egression. of a coriander
desaturase results in etroselmmc acrd pro uctron 1n ansgemc tobacco. Proc
Natl Acad 801 USA 9: 11184-11188.

Choi JK, Yu F, Wurtele ES, Nikolau BJ (1995) Molecular cloning and
charactenzatron of the cDNA coding for the brotm-contarmng subumt of the
chloroplastrc acetyl-CoA carboxylase. Plant Physrol 109: 619-625.

Chuman L, Brody S (1989) Acyl carrier protein is present in the mitochondria
of plants and eukaryotrc nucro-orgamsms. Eur J Brochem 184: 643-649.

21

22

Fawcett T Simon WJ, Swinhoe Shanklin J, Nishida 1, Christie WW,
Slabas AR (1994) Expressronof A and steady-state levels of Bprotein
1

rlssosfopgés of enoyl-ACP reductase from Brassica napus. Plant Molec 01 26:

Ebel J Hahlbrock K (1977? Enzymes of flavone and flavonol coside
brosynthesrs. Coordmateand seectrve mductron 1n cell—suspensron c es of
Petraselinum hartense. Eur J Biochem 75: 201-209.

Egli M Lutz S, Gengenbach B (1995) cDNA sequence of a multifunctional

acetyl- oA carbo lase from maize. Absract m the proceedings from the l995

IS3mchemrstry and olecular Brology of Plant Fatty Ac1ds and Glycerolrprds
ymposrum.

Garwin JL, Klages Cronan JE 1980 Structural, enzymatic and enetic
studies of beta-ketoagll-acyl carrier irroteiir synthases I and H ofEschgfichia
coli. J B101 Chem 255: 11949-11956.

Garwin JL, Cronan JE (1980} Thermal modulation of fatty acid 5 thesis in
Ezglgrefichia coli does not mvo ve de novo synthesrs. J Bacterrol 41: 1457-

Gontero Cardenas ML, Ricard J 1988 A functional five-enzyme com lex
of chlororlril’astS involved in the Calvin(cycle). Eur J Biochem 173: 437-443.p

Hannapel DJ, Ohlro¥e JB (1988) Re ation of acyl carrier protein
rlnlelszep 78 RNA levels unng seed and le development. Plant Physrol 86:

Harrington A, Hebert CJ, Tung lB Getz GS, Slonimski PP (1993)
Identification of a new nuclear gene ( Ell/fl) encodmg a rotem homologous to
a beta-keto-acyl synthase which 18 essential for nntoc ondrral respiratlon 1n
Saccharomyces cerevisiae. Molecular Microbiology 9: 545-555.

Hendren RW, Bloch K (197.9).Fa acid 5 thetases from Euglena gracilis:
separatron of component actrvrters o the AC -dependent fa acrd s thetase
2111;?) fairgrérilg punficatron of the beta-ketoacyl-ACP reductase. 3101 C cm 255:

J aworski JG, Post-Beittenmiller D, Ohlroggde JB (1993) Acetyl-acyilacrrier
cr .

Brotern rs not a major intermediate in fatty a biosynthesis 1n spinac Eur J
iochem 273: 981-987.

Kamiryo T, Numa S (1973) Reduction of acetyl-CoA carboxylase content of
Saccharamyces cerevisiae by exogenous fatty acrds. FEBS Letters 38: 29-32.

Kannan%a CG, Jacobson BS Stumpf PK (1973) Fat metabolism in higher
plants. L : A comparison of fatty acrd-synthesrzmg enzymes 1n chloro ast

11s6ollated from mature and immature leaves of spmach. Plant Physiol. 52: T56-

23

Kojnka C, Amrhein N (1984 The enzymatic malonylation of 1-

ammoc clo ro ane-l-carbo 11c am in homo enates of mun bean h 1 ,
Planta 62:p 2 6-235. xy g g YPOCOW S

Kolattukudy PE (1987) Lipid-derived defensive polymers and waxes and their
role in .plant-nucrobe mteractron. .In PK. Stumpf EE Conn, eds, The
Brochemrs of Plants: A comprehensrve treatrse, V0 9, Acadenuc Press, New

York, pp 2 1-314

Livne A, Sukenik (1990) Acetgl-COA carboxylase from the marine
prymnesrophyte Isochrysis galbana. lant Cell Physrol 31: 851-858.

Lo ez-Casillas F, Bai DH, Luo X, Kong IS, Hermodson MA, Kim KH
(1 8) Structure of the codm se uence and 811m ammo-acrd sequence of
acetyl-CoA carboxylase. Proc at Acad 801 SA 8 : 5784-57 88.

MacCarthy. JJ, Stumpf PK (19.80).The efl‘ect of different temperatures on fatgy
231319 synthesrs and polymsaturatron in cell suspensron culture. Planta 147: 38 -

MacCarthg' JJ, Stumpf PK (1980 Incorporation of [2-“C]malonyl-CoA into
fatty acrds y a cell free extract of atharanthus roseus suspensron culture cells.
Planta 150: 412-418.

Ma uson K, Jackowski S, Rock CO Cronan JE (1993).Regulation of fatty
acr biosynthesrs in Escherichia coli. Microbiological Revrews 57: 522-542.

Mattick JS Tsukamoto Y, Nickless J, Wakil SJ (1983a) The architecture of
the animal fir acrd s thetase: I. Proteolytrc dissection and peptide mappmg.
J Brol Chem 58: 15 91-15299.

Mattick JS, Nickless J, Mizugaki M, Yarzg CY, Uchi ama S, Wakil SJ
(1983b) The archrtecture of the annual fatty acr . synthetase: . Separatlon of the

core and thioesterase functions and determmatron of the N-C onentation of the
subunit. J Brol Chem 258: 15300-15304.

McIlhinne/y RAJ (19909 The fats of life: the importance and function of protein
acylatron. IBS 1 : 38 -391.

McKeon TA, Stumpf PK (1982) Purification and characterization of the
stearoyl-acyl earner protem desaturase and acyl-ac 1 carrier protem tluoesterase
from maturing seeds of safflower. J 3101 Chem 2 7: 12141-12147.

Mikolajczk S, Brody S (1990) De nova fatty acid sgrthesis mediated by acyl
232nm protern 1n Neuraspora crassa mitochondria. ur J Biochem 187: 43 -

° ' nG ParentiM Ma eeAI 1995 Thed 'c role of almj lation
insigng'altrairsductionTlBSfil:181-886. ) W p ty

24

Nakahira K, Ohlrogge JB (manuscript in preparation). Photosystem I
mhyolveinetnt in regulation of acetyl-Co carboxylase actrvrty in spinach
0 crap as.

Ohlro Savage L, Jaworski J, Voelker T, Post-Beittenmiller D (1995)
Altera .on .0 ac l-ac l earner protem pools amd ace l-CoA carboxylase

expressron m Esc ric ia coli by a plant medium chain acy -acyl earner protein
tluoesterase. Arch Brochem Biophys 317: 185-190.

Ohlro e JB 1994 Desi of new lant roducts: E ' eerin of fa acid
metabcfirgsm. Pl(ant Pliysiol 04: 821-836. p ngm g tty

Ohlrogge JB, Kuo TM 1984) Control of lipid synthesis during soybean seed

develo ment: c an immunochemical assay of acyl carrier protein. Plant
Physro 74: 622- 25.

Ohlrogge JB, Kuhn DN, Stum f PK (1979) Subcellular localization of acyl
egrnlerl' lt’otelngrén leaf protoplast o Spinacia oleracea. Proc Natl Acad Scr USA

Ohlr e JB, Pollard. MR, Stumpf PK £1 97 8) Studies on the biosynthesis of
waxes y developmg j0joba seed tissue. 1p1ds 13: 203-210.

Page RA, Okada S, Harwqod JL (1994) Ace l-CoA carbo lase exerts strong
flux control hprd synthesrs 1n plants. Brochrm rophys Acta 210: 369-372.

Post-Beittenmiller D, Jaworski JG, Ohlroggie JB (1991) In viva pools of free
and acylated acyl carnerjpfiotelrréshm 3 ma h vrdence for srtes of regulatron of
10 e

c :
fatty acrd biosynthesis. m 66: 1858-1865.

Post-Beittenmiller D, Rou an PG, Ohlrogge JB (1992 Regérlation of plant
fatty acrd biosynthesrs: ysrs of acyl-Co and acyl-A P su strate pools in
spinach and pea chloroplast. Plant Physrol 100: 923-930.

RoeslerKR, Savage L, Shintani .DK, Shorrosh BS, Ohlrogge JB (1996? C0-
punficatron, co-rmmuno recrprtatron, and coordinate. expressron during or seed
development of acetyl- oA carboxylase actrvrty, blOtln carboxylase roetrn and
blOtln carboxyl camer protem 1n hrgher plants. Planta 198:(m pressg

Roessler PG (1990) Purification and characterization of ace l-CoA carboxylase
from the dratom Cyclotella crypitca. Plant Physrol 92: 73- 8.

Ron an PG Slack CR 1977 Lon chain ac l-coenzyme A s thestase
eggvztg/lof spinach chloroplgst is ioncegtrated in the envelope. Bich3enm J 158:

Runswick MJ, Fearnle IM Skehel JM, Walker JE (1991) Presence of an
acyl carrier protem 1n ADH2ub1 mnone oxrdoreductase from bovrne heart
mitochondria. FEBS Lett 286: 121- 24.

25

Rutkoski A Jaworski JG (1978) Fa acid synthetase. from chloro last of
soybean cot}! edons: ACP activation and 0A mh1 1t10n. Brochem Bioplgys Res
Comm 8 :428-430.

Sackmann U, Zensen R, Rohlen D, Jahnke U Weiss H (1991) The acyl
carrier totem 1n Neurospora crassa mitochondria rs a subunit of
NADqu 1qu1none oxrdoreductase (complex I). Eur J Brochem 200: 463-469.

Sasaki Y, Konishi T, Nagano Y.(1995 The corrépartmentation of acetyl-CoA
carboxylase m plants. Plant Physml 10 : 445-44 .

SchulteU, Fecke W Krull _C, Nehls U, Schmiede A, Schneider R, Ohnishi
1T1} Weiss . H. (1994’) In vrvo drssectron of the mitochondrial respiratory

ADqubr urnone oxidoreductase (complex 1). Biochim et Biophysrca Acta
1187: 121- 24.

Shimakata T, Stumpf PK (1983) The purification and function of acetyl-
CoAzacyl earner proetem transferase. J B101 Chem 258: 3592-3598.

Shimakata T, Stumpf PK (1982b) The prokaryotic nature of the fatty acid
3 . thetase .of develqpmg Carthamus t1nctor1us L. (saffflower) seed. Arch
iochem Brophys 21 : 1 4-154.

Shimakata T Stumpf PK 1982b) Fatty acid synthase of Spinacia oleracea
leaves. Plant Phys1ol 69: 12 7-126 .

Shimakata T, Stumpf PK (1982c) Isolation and fimction of spinach leaf beta-
lscrsetlozacyl-[acyl earner protem] synthases. Proc Natl Acad 801 USA 79: 5808-

Shintani DK, Ohlrogge JB (1994) The characterization of a mitochondrial ac l
egrzrieri 53mm Isoform rsolated from Arabidopsis thaliana. Plant Physrol 10 :

Shintani DK, Ohlroggge JB 1995 Feedback inhibition of fa acid 8 thesis
in tobacco suspension cells. The Pfant Journal 7: 577-587. tty yn

Shorrosh BS Roesler KR, Shintani D, van de loo FJ, Ohlrogge JB (1995)
Structural analysrs, lastrd local1zatron, and expressmn of the brotm carboxylase
subumt of acetyl- 0A carboxylase fiom tobacco. Plant Physml 108: 805-812.

Smith WL, Borgeat P (1985) The eicosanoids: Prosta andins, thromboxanes,
leukotnenes, an hydroxy-ercosaenorc acrds. In DE ance, JE Vance, eds,

Biochemis of Li ids and Membranes The Ben'amin/Cummm' gs Publrshin
Company, tIrnyc, Meiilo Park CA, pp 325-360. J g

Srere PA, Sum ' She D 1987 Organlzatr' 'onal as ects of the citric acid
cycle. Biochem ocBSymp 4: 193-132. p

26

Stoops JK, Awad ES, Arslanian MJ Gunsherg S Wakil SJ, Oliver RM
(197 ) Studies on the east fatty acrd synthetase: Subumt com osition and

structural or tion 0 a lar e multiftmctronal e e com lex. Biol Chem
253: 44641375. 8 3 p

Stoops JK, Wakil SJ (1978) The isolation of the two subunits of yeast fatty
acrd synthetase. Brochem Biophys Res Commun 84: 225-231.

Stoops JK, Wakil JS (1980) Yeast fatty acid synthetase: Structure-function
relatlonsh1p and nature 0 the beta-ketoacyl synthetase s1te. Proc Natl Acad Sci
USA 177: 4544-4548.

Suss KH, Arkona C Manteuffel lfijAdler K (1993) Calvin cycle multienzyme
complexex are bormd’ to chloroplast ylakord membranes of h1gher plant in situ.
Proc Natl Acad Sc1 USA 90: 514-5 18.

Terza ° W 1986? Metabolism of Tween-fatty acid esters by cultured soybean
cells. lant P ysro 82: 7 80-7 86.

Terzaghi W, Roth j, Keim P Honeycutt RJ, Lark KG (1985 A‘somatic
enetrc s stem for so bean. In Freelmg, ed, Plant Genetics, Alan Lrss, Inc.,
ew Yogic, pp 771-7 9.

Topefer R, Martini N (1994 Molecular cloning. of cDNAs or genes encodin
prlo6 4111255mvolved 1n de nova atty acrd syntheSIS 1n plants. J Plant Physml 14 :

Towler DA, Gordon J1, Adams SP, Glaser L (19.88 The biolo and
enzymology of eukaryotrc protern acylatron. Ann Rev Bloc em 57: 69- .

Va os RP 1971 Re ation of fa acid bios thesis. Curr To Cell Re
4: ffl9-166.( ) gul tty W p gul

 

W08 RP (1974 Biosynthesis of saturated fatty acids. In TWGoodwin1 ed,
Intematron Revrew of Scrence. Brochemrstry of Lrprds, V0 4,
Butterworth, London, pp 100-140.

Vick BA, Zimmerman DC (1984 Biosynthesis of jasmonic acid by several
plant specles. Plant Physrol 7 :45 -461.

Voelker T Worrell AC Anderson L, .Bleinbaum J, Fan C, Hawkins DJ,
Radke. SE, avies HM (1992) Fatty acrd biosynthe31s red1rected to medlum
chams 1n transgemc allseed plants. Scrence 257: 72-74.

Wakil S 1989 Fa acid 5 thase, a roficient multifunctional enzyme.
BiochemisiryZ :452 -4530. yn p

Zenson R,Husmann .H, Schneider R, Peine T Weiss H $1992) De nova
synthes1s and desaturation of fatty acrds at the mrtochondnal acy carner proteug

a subunit of NADqubiquinone oxidoreductase in Neurospora crassa. FEB
Lett 310: 179-181.

Chapter 2
The protein purification and cDNA cloning of the predominant acyl

carrier protein isoform in Arabidopsis leaves

Abstract

The predominant ACP isoform formd in Arabidopsis leaf tissues (the leaf
major isoform [LMI] ACP) was purified to homogeneity from Arabidopsis leaf
tissue. The sequence of the first 25 amino acid residues of the amino-terminus
of the mature LMI ACP were determined and used to identify a cDNA clone
corresponding to this ACP isoform from the Arabidopsis EST cDNA collection.
The identity of the LMI ACP cDNA was confirmed by the co-migration of the
Escherichia coli (E. call) expressed LMI cDNA sequence with the endogenous
LMI ACP on native polyacrylamide gels. Although the LMI ACP is the most
abundant ACP isoform in leaves, LMI ACP is also found in other green tissues
and mature seeds. As was determined by immuno-blot and Northern blot
analysis, the LMI ACP was present in extracts made from several green tissues
and mature seeds, but was not detectable in roots. Although the deduced amino
acid sequence of the mature LMI ACP shared a high degree of similarity with
the previously cloned Arabidopsis ACPs, the LMI contained several structural
features that distinguished it from other Arabidopsis ACP isoforms.

27

Introduction

Acyl carrier protein (ACP) is an essential protein cofactor involved in
fatty acid synthesis (Ohlrogge, 1987). This protein acts as a scaffold which
shuttles the growing acyl-chain between the various enzymes in the fatty acid
synthase complex. All ACPs pocess a highly conserved amino-acid sequence
surrounding a serine residue to which the phosphopantetheine prosthetic group
is attached. The growing acyl-chain is covalently attached to this prosthetic
group through a thioester linkage. The animal and yeast ACPs are present as
domains with in type I FAS multifunctional polypeptide complexes (W akil et al.,
1983), where as plant and bacterial ACPs exist as soluble proteins with
molecular weights of approximately 10,000 and pI's of around 4 (Ohlrogge,
1987)

In plants, the plastid localized ACP has been the most well characterized
component of the FAS complex. Plastid ACP proteins and genes have been
isolated from a number of plant species (von Wettstein-Knowles et al., 1994 and
references within). Of the known plant ACPs, all show a high degree of amino
acid similarity, especially in the region surrounding the prosthetic attachment site
(Ohlrogge, 1987).

Plant species have been shown to express multiple isoforms of the plastic
localized ACP (Ohlrogge & Kuo, 1985). The number of plastid ACP isoforms
present differ not only between plant species, but also between different tissues
within a given plant species. Spinach contains two different plastid ACP
isoforms in leaves, but only one ACP isoform in roots and seeds (Ohlrogge &
Kuo, 1985). The multiple isoforms of plastid ACP have been shown to be coded
for by small nuclear gene families in many plant species (von Wettstein, 1994).
While the different plastid ACP genes clearly correspond to proteins with

28

29
different peptide sequences, evidence from the sequencing of multiple ACP
cDNAs from Brassica napus showed that some individual ACP isoforms are
coded for by multiple genes (Safford et al., 1988).

The functional significance of multiple plastid ACP isoforms is not
known. However, in leaf tissue, one possible role for multiple ACP isoforms
may be to direct fatty acids towards use in either microsomal or chloroplast
glycerolipid synthetic pathways (Roughan and Slack, 1982). The majority of the
leaf membrane lipids are synthesized through a microsomal localized
glycerolipid synthetic pathway. Fatty acids are directed towards this pathway
through the action of long chain acyl-ACP thioesterases which cleave the mature
fatty acids from the ACP backbone. The free fatty acids then move across the
plastic membrane where they are rapidly converted to acyl-CoAs which serve
as substrates for microsomal localized acyltransferases. In addition to the
microsomal glycerolipid pathway, the leaves of certain plant species such as
Arabidopsis and spinach, synthesize a significant amount of their glycerolipids
through a chloroplast localized pathway. In these plants, long chain acyl-ACPs
serve as substrates for chloroplast localized acyltransferases which transfer fatty
acids to the sn-1 and sn-2 position of glycerol backbones (Browse and
Sommerville, 1991). It has been proposed that specific ACP isoforms may serve
as preferred acyl-ACP substrates either for acy-ACP thioesterases which would
direct fatty acids towards microsomal glycerolipid pathways or for plastic
localized acyl-ACP acyltransferases. The results from in vitro studies suggested
that in spinach leaves, the ACP I isoform was the preferred substrate for acyl-
ACP thioesterases, while the ACP H isoform served as a better substrate for the
chloroplast local ized glycerol-3 -phosphate acyltransferase (GPAT) (Guerra et
al., 1986). However, although in viva measurements of acyl-ACP pools in

30
spinach leaves showed that both ACP I and ACP H participated equally in the
initial reactions of fatty acid synthesis, in viva evidence does not exist to indicate
whether either plastic ACP isoform is preferred over the other by either GPAT
or any of the thioesterases (Post-Beittenmiller et al., 1991).

Arabidopsis thaliana provides a particularly attractive system to study the
role of the multiple ACP isoforms in plants. Previous work has shown that
Arabidopsis contains at least four and as many as six different plastid ACP
isoforms, some of which are expressed constitutively in all tissues while others
appear to be expressed in a tissue specific manner (Hlousek-Radojcic et al.,
1991). The genes coding for three different Arabidopsis ACP isoforms have
previously been cloned (Post-Beittenmiller et al., 1989; Lamppa et al., 1991;
Hlousek-Radojcic et al., 1991) and appear to be expressed constitutively in all
tissues (Hlousek-Radojcic et al., 1991).

Arabidopsis leaves have been shown to contain at least 5 different
isoforms of ACP (Hlousek-Radojcic et al., 1991). One ACP isoform, designated
as the leaf major isoform (LMI) ACP, was present at significantly higher levels
than the other ACP expressed in leaves and appeared to be expressed in a leaf
specific manner.

The work reported here will describe the protein purification of the leaf
major ACP isoform and the subsequent identification of the corresponding
cDNA clone. Furthermore, patterns of expression will be shown by immunoblot
and Northern blot analysis.

Materials and Methods
(Unless otherwise indicated, all procedures were done at 4°C.)
Purification the leaf major ACP isoform (LMI)

Two hundred grns of Arabidopsis thalina leaf material, which was
previously frozen in liquid nitrogen and stored at -80°C, was homogenized in
500 ml of 50 mM sodium phosphate pH 6.8, 1 mM DTT using a Brinkrnann
polytron. The homogenate was centrifuged at 6,500 RPM in a Sorval GS-3 rotor
for 20 minutes and the supematant was filtered through miracloth and collected.
Ammonia sulfate was added to the supematant to 60% saturation and stirred for
one hour. The (NH4),SO.1 precipitate was pelleted by centrifugation at 10,000 g
for 20 minutes. The supernatant was collected and adjusted to 5% trichloroacetic
acid (TCA). After stining for 1 hour, the TCA precipitate was collected by
centrifugation at 10,000 g for 20 minutes. The TCA pellet was then resuspended
in 50 ml of 100 mM Tris pH 9.0, 100 mM dithiolthrietol (DTT). The
resuspended pellet solution was diluted with H20 until the conductivity was
equal to that of a 100 mM NaCl solution. DTT was added to a final
concentration of 10 mM and the solution was stirred for 1 hour. After
centrifugation at 10,000 g for 10 minutes to remove any insoluble material the
solution was applied at 0.833 ml/minute to a 25 ml DEAE cellulose column
(DE52, Whatrnann) equilibrated with 10 column volumes of 10 mM 2-(N--
morpholino) ethanesulfonic acid (MES) pH 6.1, 10 mM DTT at 0.83 ml/minute.
After washing with 2 column volumes of 10 mM MES pH 6.1, 10 mM DTT.
protein was eluted from the colmnn using a linear gradient of 0 to 500 mM NaCl
in 10 mM MES pH 6.1, 10 mM DTT. Fractions containing ACP were identified
using the E. cold acyl-ACP synthetase assay as described by Kuo and Ohlrogge
(1984). Pooled fractions containing the LMI ACP were further purified on a

31

32
Mono-Q HRS/5 FPLC column (Pharmacia) equilibrated with 20 mM Tris pH
7.5. Protein was eluted from the column with a O to 300 mM MgCl2 gradient in
10 mM Tris pH 7.5 at a flow rate of 1.5 ml/minute. Protein elusion was
monitored by absorbance at 225 nm and individual protein elution peaks were
collected by hand. Fractions were analyzed for ACP as described above.

E. coli Expression of LMI ACP

Polymerase chain reaction (PCR) was used to facilitate the cloning of the
mature LMI ACP sequence into the pET 3D E. coli expression vector (Studier
et al., 1990). For the LMI expression construct containing two alanine residues
at the amino terminus, a forward primer (J 03 5) with the following sequence was
synthesized: CATGCCATGGCCGCGAAAGCAGAGACGGT. For the LMI
expression construct that contained a single alanine residue the amino terminus,
the forward primer; CATGCCATGCGAAAGCAGAGACGGTGCA was used.
Both primers contained NcaI sites at the 5' end of each primer. The reverse
primer in both cases was designed from the SP6 promoter contained in the
cDNA vector. The mature LMI sequences were amplified with the Pfu DNA
polymerase (Stratagene, La Jolla, CA) as described by Mullis and Falloona
(1987) using thirty cycles as follows: 1 min at 94C, 2 min 55°C, 2 min 72°C.
The PCR products were first treated as described by Crowe et al. (1990) and
were then digested with BamHI and N001 and subcloned into pET3D. The single
and double alanine LMI expression constructs were designated as pl -ala and p2-
ala respectively. Both DNA constructs were sequenced as described by Sanger
et al. (1977) and no PCR generated errors were detected in either construct.
These constructs were transformed into the E. coli strain, BL21(DE3) (Studier
et al., 1990).

33

Twenty-five ml cultures of BL21(DE3) cells transformed with the LMI
pl-ala and p2-ala constructs, which were started from overnight cultures, were
grown at 37°C in Lmia broth containing 50 mg/ml ampicillin to a final optical
density of 0.6 measured at 600 nm. The expression of the LMI protein
expression was then induced by the addition of 0.4 mM isopropyl-B-D-
thiogalactopyranoside (IPTG), after which each culture was grown for an
additional 4 hours at 37°C. After centrifugation at 6000 g for 10 minutes, the cell
pellets were resuspended and homogenized in 5% TCA. The TCA precipitate
was centrifuged at 10,000 g for 20 minutes. After resuspending the TCA pellet
in 200 mM Tris pH 9.5, 100 mM DTT the ACP was deacylated by incubating
the resulting solution at 65 °C for one hour. The solution was then adjusted to a
final concentration of 5% TCA and incubated on ice for 20 minutes. The TCA
pellet was collected as described above. The pellet was washed with 1% TCA
and resuspended in 50 mM N-2hydroxyethylpiperazine-N‘-2-ethanesulfonic acid
(HEPES)-KOH pH 7.8, 5 mM DTT. Insoluble material was then removed by
centrifugation at 10,000 g at 4°C for 20 minutes.

Native PAGE and Immnuoblot analysis

For immunoblot analysis, ACPs fi'om Arabidopsis tissues were extracted
as described for the extraction of the E. coli expressed ACPs except that
approximately 1 gm of each tissue was ground to a fine powder in liquid nitrogen
prior to the homogenization in 5% TCA. All ACP samples were fractionated by
native 15% PAGE as described by Post-Beittenmiller et al. (1991). Prior to
electrophoresis, samples were treated with either 5 mM DTT or 5 mM N-
ethylmaleiamide (NEM). The protein was blotted to nitrocellulose filters and
ACPs were detected using antibodies to spinach ACP I as described by Post-

34
Beittenmiller et al. (1991). Native gels were silver stained as described by Merril
(1990)

RNA Extraction and Northern Blot Analysis

All aqueous solutions used in RNA preparations were treated with 0.1%
diethylpyrocarbonate as described by Sambrook et al. (1989). Fresh tissue was
harvested and immediately frozen in liquid nitrogen and stored at -80°C until
proccessed. The frozen tissue was ground in liquid nitrogen to a fine powder.
Three and one halfml of homogenization buffer at 80°C (0.2 M sodium borate,
30 mM [ethylenebis(oxyethylenenitrilo)]-tetraacetic acid (EGTA), 1% SD S, 1%
deoxycholate, 2% polyvinylpyrrolidone 40,000, 10 mM DTT) was added to one
gm of frozen ground tissue. One volume of 1:1 (v/v) phenol: chloroform was
added immediately to each sample and the mixture was vortexed for several
seconds. Phases were seperated by centrifugation at 10,000 g for 20 minutes at
room temperature and the aqueous phase was extracted twice with One volume
of chloroform. The final aqueous fi'action was adjusted to 2M LiCl and
incubated on ice overnight. The LiCl precipitate was then collected by
centrifuging at 10,000 g for 20 minutes and the pellet was washed one time with
2 M LiCl. The washed LiCl pellet was resuspended in 500 n] of H20 and
reprecipitated by adding 1/10 volume of 2M potassium acetate pH 5.5 and two
volumes of ethanol. The ethanol precipitate was pelleted at 10,000 g for 20
minutes and resuspended in 500 pl of H20 and stored at -20°C. Twenty pg of
total RNA was then fractionated on formaldehyde gels and blotted to Zeta-Probe
nylon membranes (BioRad, Richmond CA) as described by Sambrook et al.
(1989). The membranes were prehybridized overnight at 42°C in SXSSC, 10X
Denhardt's solution, 0.1% SDS, 0.1 M potassium phosphate pH 6.8, 100 ng/ml

35
salmon sperm DNA. The membranes were then hybridized overnight in SXSSC,
10X Denhardt's solution, 0.1 M potassium phosphate pH 6.8, 100 ug/ml salmon
sperm DNA, 10% dextran sulfate, 30% fonnamide with l x 10° CPM / m1
random primer 32P-dCTP labelled probe. The blots were washed 3 times at 65°C
in 0. IXSSC, 0.1% SDS.

Results and Discussion
Purification of the Arabidopsis LMI ACP

The pattern of ACP isoform expression was extensively characterized in
leaf, seed, and root tissue fiom Arabidopsis (Hlousek-Radojcic et al., 1991).
These studies showed that Arabidopsis leaves contained at least five different
isoforms of ACP. One leaf ACP, the leaf major ACP isoform (LMI), was shown
to be the predominant ACP in leaves but undetectable in root and seed tissues.
Due to this apparant leaf specific pattern of expression of the LMI ACP
(Hlousek-Radojcic et al., 1991) and the hypothesis that different leaf ACPs may
be involved in directing fatty acids towards different metabolic pathways
(Guerra et al.,1986), we became interested in characterizing this particular
Arabidopsis ACP isoform.

The Arabidopsis LMI ACP isoform was purified to homogeneity after two
rounds of anion-exchange chromatography. The crude 5% trichloroacetic acid
(TCA) pellet, containing the total complement of leaf ACPs, was fractionated
into three peaks of ACP after elusion with a salt gradient from a conventional
DEAE-cellulose column (Fig. 2-1a). Western blot analysis of these peak
fractions showed that the majority of the LMI was present in the third peak (Fig.
2-1b). The DEAE-cellulose eluted fractions containing the LMI ACP were then
applied to a mono-Q anion-exchange FPLC column from which the LMI ACP

36

Figure 2-1. Purification of LMI ACP from Arabidopsis leaf. A) The elution
profile of ACPs from a DEAE-cellulose column. ACPs were eluted from the
DEAE column using a 0 to 500 mM NaCl gradient. Fractions containing ACP
were identified using the E. coli acyl-ACP synthtease assay. B) Irnmunblot
analysis of DEAE fractions containing ACP. Aliquots from DEAE fractions
were run on 15% native PAGE gels, blotted to nitrocellulose and ACPs were
then detected with antibodies to spinach ACP I. C) Silver stained native PAGE
gel containing an aliquot of the LMI ACP eluted from the mono-Q FPLC

column.

 

4.j_k. Wu _.—._. -.

 

37

.>

20000 -
16000 —
12000 —
8000 —

4000 —

o __

nanograms ACP

 

 

fraction number

6 m Omar...

27 29 34 35 36 38 39 4o 41 Pellet peak fraction

 

 

38
eluted in a single fraction. The mono-Q eluted LMI ACP was purified to

homogeneity as demonstrated by Western blot analysis and silver staining of
native PAGE gels (Fig. 2-1c).

Protein Sequencing and cDNA Identification

The purified LMI ACP was subjected to amino-terminal protein sequence
analysis. Approximately 100 pmoles of LMI ACP was applied to 3 Applied
Biosystems 477A Protein Sequencer. The first 25 amino acids of the mature LMI
ACP amino-terminus were identified through this analysis (the underlined amino
acid sequence, Fig. 2-2). Initial attempts to clone a cDNA corresponding to the
LMI ACP using PCR approaches were unsuccessful. However, a corresponding
ACP cDNA clone containing this 25 amino acid sequence was identified from
the Arabidopsis EST cDNA collection at Michigan State University (N ewmann
et al., 1994). Four EST cDNA clones (EST H)s: T45818, T43755, T44265,
T43443) shared 100% identity with the first 25 amino acids of the mature LMI
ACP. Clone T4581 8 was obtained from the Arabidopsis Biological Resource
Center at Ohio State University and the entire cDNA sequence was determined
( Fig. 2-2).

Confirmation of the identity of the Putative LMI ACP cDNA

Although the amino-terminal sequence was identical to the deduced
sequence fi'om the ACP cDNA identified from the EST collection, further
confirmation was needed to prove that the cDNA corresponded to the LMI ACP.
Therefore, the cDNA sequence corresponding to the deduced mature ACP was
expressed in E. coli and its mobility was compared to the mobility of the

endogenous LMI ACP from a leaf extract on native poly acrylamide gels. Native

11
121

31
181

51
241

71
301

91
361

111
421

131
481

541

601

661

Figure 2-2. Nucleotide and deduced amino acid sequence of the cDNA clone
corresponding to the LMI ACP. The underlined amino acid sequence
represents the amino-terminal sequence determined from protein sequencing of
the LMI ACP purified from Arabidopsis leaf. The closed triangle denotes the
position of the transit peptide cleavage site. Amino acid sequence in bold type

represents the deduced mature LMI ACP protein sequence. The amino acid

39

NCNTGNCCGNTCCGGAATTCCNTGGTCGACCCACGCGTCCGTTTTTTTTTTGGGTCTCTG

M A S L S T T S L S
AGTTGTGTATTGAGCTTCATCTCCTTCAAATGGCTTCCTTATCAACCACTTCCCTAAGCT

F K A P S T T I S Q V L R K A S S S Q S
TCAAAGCTCCGTCCACCACAATTTCACAGGTTTTAAGGAAAGCCTCAAGTTCTCAGTCTG
V

V T F G R F T S S T K S L R L Q I S C A
TNACCTTTGGTCGTTTCACGTCTTCAACAAAGAGCCTTCGTCTTCAAATCAGTTGTGCCG

A I A I ,T 414,0 K V SLLD I V .KL,I Q L A L A
CGAAAGCAGAGACGGTGCAGAAAGTGAGTGACATTGTAAAAGAACAATTGGCTTTAGCTG

A Q I P L T A B S K F 8 A L G A D 8 L D
CTGATGTTCCGCTCACTGCTGAATCCAAGTTCTCTGCTCTTGGTGCCGATTCTCTCGACA

T V 8 I V M A L B B X P N I 8 V B B 8 D
CCGTGGAGATAGTGATGGCGTTGGAGGAAAAGTTTAACATAAGTGTGGAGGAATCTGATG

A Q N I T T I L B A A D L I I D L V Q I
CTCAAAACATTACGACCATCCTAGAGGCGGCTGATTTGATAGAGGACCTTGTTCAAAAGA

K P A A B T 8
AACCCGCGGCGGAAACTTCCTGAACCGGCTGTCTTTTCTTTCCTGACCTAATCTTGCCNT

CAGAGCGTTTTATTACGGTATATTAGNCCGTTTGTTANTTTTATTTATTTCGGAAGCTAA
GATTGGGTGTTGACAATGAGACNACTTTGGTACGTTGAGATTGATTCTTTGTGGTGATTG

GTCTTATGACGAATTCATAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

sequence in normal type represents the LMI ACP transit peptide.

10
120

30
180

50
240

70
300

90
360

110
420

130
480

137
540

600

660

718

40
PAGE has been shown to be very sensitive to small changes in ACP
conformation such that a single amino acid change can significantly alter its
mobility (Hlousek-Radojcic et al., 1991).

However, because the results from the amino-terminal sequencing was
ambiguous as to whether the amino-terminus of the mature LMI ACP contained
one or two alanine residues, two PCR derived DNA sequences encoding mature
LMI ACPs containing either one or two alanine residues at their amino termini
were expressed in E. coli. These constructs were designated as pl-ala and p2—ala
respecitively.

The mobility of each of the expressed proteins was compared to that of
the endogenous LMI ACP from an Arabidopsis leaf extract on native poly
acrylamide gels. Whereas the E. coli expressed p2-ala protein (Fig. 2-3a) co-
migrated exactly with the LMI ACP when from anArabidapsz‘s leaf extract (Fig.
2—3a), the mobility of the pl-ala protein differed substantially from the LMI ACP
(Fig. 2-3a). The co-migration of the p2-ala E. coli expressed LMI protein with
the LMI ACP in leaf extracts confirms that the putative LMI cDNA codes for
the LMI ACP. Furthermore, these results also confirm that the cleavage of the
LMI ACP transit peptide results in the occurance of two alanines at the amino
terminus of the mature LMI ACP.

Tissue Specific Patterns of LMI Expression

The tissue specific expression of the LMI ACP was determined by
immuno-blot analysis of protein extracts isolated from various Arabidopsis
tissues. The LMI ACP was detected only in green tissues; including leaves (Fig.
2-3a), above-ground tissues of young seedlings (Fig. 2-3b), siliques (Fig. 2-3b),
and whole flower pods (Fig. 2-3b). The level of LMI protein expression was

41

Figure 2-3. Immunoblot analysis of E. coli expressed LMI ACP constructs
and ACP extracts from different Arabidopsis tissues. A) Extracts from
Arabidopsis leaf and two E. coli expressed LMI ACP constructs containing
either one or two alanine residues at their respective amino-terminus were
fractionated on native 15% PAGE gels. B) The E. coli expressed LMI ACP (p2-
ala) and TCA extracts made from different Arabidopsis tissues were fractionated
on native 15% PAGE gels. In both cases (A) and (B), the gels were then blotted
to nitrocellulose and ACPs were identified with antibodies to spinach ACP I.

 

 

 

42

E
A' a .2 3
9 z -
E. o. 3
- —E.collACP
_ACP-1
—ACP-2
..... 4... —Agg-3
“fut-Ah?
B.
i .
C
a E
“a
é‘a‘ “ i5 .. s .
a; i 3§ 8 ..=‘f 5
III II. < M or .
ACP-Z—
v ... --~«—- -ACP-1

ACP-2— I. on...“ _ 83$

ACP-Y— _ ._ ' —A
LMI Acp—m'w

 

 

 

 

 

 

 

43

highest in leaf extracts where it represented approximately 70% of the total leaf
ACP (Fig. 2-3a). In aboveground tissues of young seedlings, which consist
primarily of stems and leaves, the LMI ACP represented approximately 30% of
the total ACP content, (Fig. 2-3b). The LMI ACP appeared to be minor ACP
isoform in siliques, mature seeds and flower pods where it represented less 10%
of the ACP complement in both tissues (Fig. 2-3b). The LMI ACP was not
detected in roots (Fig. 2-3b).

The immune-blot analysis also showed that Arabidopsis contained at least
five different ACP isoforms. As was reported by I-Housek-Radovcic et al.
(1992), three ACP isoforms, which have been identified as ACP-1, ACP-2, and
ACP-3, were expressed in all tissues. A yet to be purified ACP isoform
(designated ACP-X) also appeared to be expressed in all tissues, however, at a
much lower level than ACP-1, ACP-2 and ACP-3 (Fig. 2-3b). Two additional
ACPs designated as ACP-Y and ACP-Z appeared to be expressed exclusively
in flowers and mature seeds respectively (Fig. 2-3b). The expression level of
these two proteins appeared to be fairly low relative to other ACP isoforms
detected in their rspective tissues.

The levels of LMI ACP mRNA in difierent Arabidopsis tissues were
determined by Northem blot analysis. While LMI ACP mRNA was present in
leaves, siliques, and flower buds, it was not detectable in roots (Fig. 2-4). This
pattern of expression was consistent with the tissue specific pattern of LMI

protein accumulation.

Structural comparisons of LMI ACP with other Arabidopsis ACPs
The LMI ACP shared between 56% and 60% amino acid sequence
identity with the three known Arabidopsis ACP protein sequences and contained

44

h
0
3
£2
ll.

Leaf
Root

   

Figure 2-4. Accumulation of LMI-ACP mRNA in different Arabidopsis
tissue. Approximately 20 ug of total RNA isolated from Arabidopsis whole
flower, leaf, root and silique tissues was fractionated under denaturing
conditions on a 1.2% agarose gel. The RNA was blotted to nylon filters and
hybridized with a random primed probe made from the LMI-ACP cDNA.

45

all of the features associated with a chloroplast localized ACP (i.e. an amino-
terminal extension corresponding to a plastic transit peptide, and a
phosphopantetheine attachment site). Although the LMI ACP shared significant
amino-acid sequence similarity to the three other cloned Arabdiopsis ACPs,
several features of the LMI ACP distinguished it from these and other plant
ACPs. First, the LMI ACP contained two deviations in the concensus sequence
corresponding to the phosphopantetheine prosthetic group attachment site (i.e.
GADSLTVEIVMGLEEEF) which has been shown to be conserved in all other
Arabidopsis ACPs and almost all plant ACPs. The first difference occurred at
position 48 of the mature protein sequence (Fig. 2-5A) where a glycine has been
subsituted for by an alanine. The second change, occurred at position 52 (Fig 2-
5A) where a glutamate residue was replaced by a lysine residue. While the
exchange at position 48 of a glycine to an a: anine would not be predicted to
significantly alter protein structure, the exchange at position 52 of a positively
charged amino-acid for a negatively charged amino acid may alter the character
of the prosthetic attachment site. These changes were unique to the LMI ACP
and have not been observed in any other plant ACPs where the prosthetic group
attachment site concensus sequence is strictly conserved.

The LMI ACP also differs from the previously characterized Arabidopsis
ACP isoforms at the location of the transit peptide cleavage site. Each of the
Arabidopsis ACPs, including the LMI ACP, shared a significant degree of amino
acid sequence similarity in the region surrounding the transit peptide cleavage
site (Fig. 2-5B). However, as was determined by the amino termial sequencing
of the endogenous LMI ACP and confirmed by the E. coli expression of the
mature sequence determined from the LMI ACP cDNA, the cleavage site of the
LMI ACP transit peptide was shift relative to the other three Arabidopsis ACPs

46

 

IX. 1 10 20 30 40 so
I | | | | |
LMI -3. H onvsnrnxaenanaanyznransxasanGADSLDTVErVMALa
ACP-1 AKQETIEKVSAIVKKQLSLTPDKKVVAETKFADLGADSLDTVEIVMGLE
ACP-2 AKPETVDKVCAVVRKQLSLKEADEITAATKEAALGADSLDTVEIVMGLE
ACP-3 AKPETVDKVCAVVRKQLSLKEADEITQATKEAALGADSLDTVEIVMGLE
51 60 70 so
I | | l
LMI BKFNIMESDKQHIITIWLIBDEVQKKM
ACP-1 EBFNIQMAEEKAQKIATVEQAAELIEELINERK
ACP-2 BEFGIEMAEEKAQSIAEVEQAAALIEELLFEKAK
ACP-3 EEFGIEMAEEKAQSIATVEQAAALIEEBLLEKAK
I3.
V
LMI Raorse aaanz

ACP - 1 RLSVSCA AKQET
ACP - 2 RLTVSCA AKPET
ACP - 3 RLTVTCA AKPE‘T

Figure 2-5. A) Alignment of the LMI ACP protein sequence with other
Arabidopsis plastid ACPs and B) comparisons of the transit peptide cleave
sites of other Arabidopsis plastid ACPs. Sequences in bold type represent
conserved amino acids. Amino acids that are identical in all ACPs are denoted
by shaded characters. Underline amino-acids represent residues that are unique
to the LMI ACP. Closed triangle represents the transit peptide cleavage site.

47
by one amino acid residue towards the amino terminus of the pro-protein.

Additionally, hydrophobicity analysis of the four different Arabidopsis
ACPs showed that the LMI ACP contained a unique stretch of hydrophobic
amino acids in the region between amino acid residues 10 and 22 of the mature
protein (Fig. 2-6). This region corresponded to a particularly hydrophillic region
of the ACP] , ACP2 and ACP3 isoforms. The calculated pI of the LMI ACP also
distinguished it fi'om the other Arabidopsis ACP isoforms. The p13 of the
Arabidopsis ACP isoform differed significantly enough to cause the four
Arabidopsis leaf ACPs to eluted from the DEAE cellulose column as would have
been predicted from their calculate pl‘s. ACPl, which had a the highest calulated
pI (i.e. 4.4) was the first ACP isoform to elute from the DEAE column. ACP2
and ACP3, which both have calculated pI's of 4.2, co-elute in the second ACP
peak, while the LMI ACP, which had a calculated pI of 3.94, eluted in the final
ACP peak.

It is not clear whether these unique features of the LMI ACP have any
physiological significance. However, if the hypothesis that specific ACP
isoforms act as preferred substrates for specific enzymatic reactions is correct,
one would expect that these ACPs would differ structurally from one another.
The unique hydrophobic nature of the amino terminal end of the LMI ACP and
its more acidic pI may facilitate preferential protein-protein interactions between

the LMI ACP and certain fatty acid synthetic enzymes.

Conclusion
The cloning of the LMI ACP is a particularly important achievement
because it the first "tissue" specific ACP isoform gene cloned from Arabidopsis.
We now have the opportunity to not only alter the expression of the LMI in

48

Figure 2-6. Hydrophobicity analysis of the Arabidopsis ACP mature
protein sequences. The matue protein sequences were subjected to hydophocity
analysis using Kyte-Doolittle amino acid classification parameters. The
horizontal axis designate amino acid residue position. The vertical axis denotes
the degree of hydrophobic / hydrophillic character. Increasing positive values
coorelate with increasing hydrophobicity and decreasing negative values
coorelate with increasing hydrophillicity. (A) = LMI ACP, (B) = ACP-1, (C) =
ACP-2, (D) = ACP-3.

49

 

I'll'l‘ll“

iilJiii

l.T

 

lr-.....iuilrii

111.1,..-an

 

ruiiiii..il,,”“I

 

 

[IT'T|' ................... .

 

 

 

 

.Il

 

Tlllflniwltri”

 

l

""I'Imnpll'

 

“lllllllewl

 

[m]

 

.—
—

 

 

 

88

1“

 

.,,Illlluil.l,.1|l

1

[Mil iiiil. .-
1 l

 

- uTurITIII‘IHI

l

 

 

n
—

 

 

 

8t

 

 

180

 

.,.,ulllluil.l,.1|i

 

 

 

l

-iaITItlTITIIITI-

 

 

11111.1”

 

illui.I

 

 

 

50
tissues where it is normally found, but it is also possible to express this "tissue"
specific ACP isoform in tissues where it is not normally expressed. Such
strategies can then be used to determine what, if any, specific role the LMI ACP
plays in plant fatty acid synthesis.

Literature Cited

Browse J, Sommerville C 1991 Glyceroligid synthesis: Biochemistry and
Regulat10n.Ann Rev Plant P ysro 42: 467-5 .

Crowe JS Cooper HJ, Smith MA, Sims MJ, Parker D Gewert D (1991)
Improved clomng eflicrency of . polymerase cham reactron CR) products after
protemase K dlgestron. Nuclerc A01d Res 19: 184.

Guerra DJ, Ohlrogge JB, Frentzen M (1986 Activity of acyl carrier rotein

isoforms 1n reactrons of plant fatty acid metabo 'sm. Plant Physrol 82: 8-453

Hlousek-Radojcic A, Post-Beittenmiller D, Ohlrogge JB (1991) E ression
of constrtutrve and trssue-spec1fic acyl carner protem1sof0rms 1n Ar idopsis.
Plant Physrol 98: 206-214.

Lammpa G, Jacks C (1991) Analysis of two linked genes coding/for the ac l
2351;? 7pgrotem (ACP) from Arabzdopszs thalima (calumbia). Plant 01 B101 1 :

~llvgenil CR (1990) Gel staining techniques. Methods in Enzymology 182: 47 7-

Ohlrogge KuoTM (1985) Plants have isoforms for acyl carrier 32 otein that
are expresse drfferently 1n drfferent trssues. J B101 Chem 260: 80 -8037

Ohlrogge JB (g8? Biochemistry of plant ac 1 carrier roteins. In PK Stumpf,
EE 1Cefimi,5e7ds, e lochemrstry of Plants, V0 10. Aca ermc Press, New York,
PP ' ~

Post-Beittenmiller MA, Hlousek-Radojcic A Ohlrogge JB (1989) DNA
sequence ofa enomrc clone encoding an Arabidopsis acyl carner protern.
Nucleic Ac1d cs. 15: 7197.

Post-Beittenmiller D, Jaworski JG, Ohlro%e JB (1991) In vivo pools of free
and acylated acyl carrier rotems in spinach: vrdense for s1tes 0f regulatron of
fatty acrd brosynthesrs. . B101 Chem 266: 1858-1865.

Rou an PG, Slack CR (19.82)3 Cellular organization of glycerolipid
meta olrsm. Ann Rev Plant Physrol 3: 97-132.

Sambrook J, Fritseh EF, Maniatis T (1989 Molecular Biology: A laboratory
manual, second ed1t101§C Nolan, ed, C01 Spring Harbor La oratory Press,
Cold Sprmg Harbor, N .

Sanger _F, .Nicklen S Coulson AR (l977) DNA sequencing with chain
termmatrng mhrbrtors. oc Natl Acad Sc1 USA 74: 5463-5467.

Safford R, Windust JHC, Lucas C De SilvaJ, James CM, Hellyer A,
Smith CG Slabas AR, Hughes SG (l988)Plast1d-locahsed seed acyl. carrier

rotein of rassica nagus rs encoded by a distinct nuclear multigene fanuly. Eur
Brochem 174: 287- 95

51

52

Studier FW, Rosenburg AH, Dunn JJ, Dubendorff JW 1990) Use of T7
golyérgerase to direct expressron of cloned genes. Methods of nzymology 185:

von Wettstein-Knowlea P Knauf V, Ohlrogge JB, Lamppa GA, Safford R,
Soucret G (1994 Nomenclature for enes encodmg acyl carner protern (ACP).
Plant Molecular rology Reporter : S31-S34.

Wakil SJ Stoo Joshi VC 1983 Fa acid s thesis and its re ation.
Annu Rex3 Bioc em 2: 537-579( ) tty yn gul

Chapter 3
The Characterization of a Mitochondrial Acyl Carrier Protein Isoform
Isolated From Arabidopsis tlraliana.1

Abstract

A cDNA clone was isolated from an Arabidopsis leaf cDNA library that
shared a high degree of protein sequence identity with mitochondrial acyl carrier
proteins (mtACPs) isolated from Neurospara crassa and bovine heart muscle.
The cDNA encoded a 88 amino acid mature protein that was preceded by a
putative 35 amino acid presequence. In vitro protein import studies have
confirmed that the presequence specifically targets this protein into pea
mitochondria, but not into chloroplasts. These studies indicated that pea
mitochondria were not only able to import and process the precursor protein, but
also possessed the ability to acylate the mature protein. The mitochondrial
localization of this protein, mtACP-l, was confirmed by western blot analysis.
Arabidopsis mitochondrial protein extracts contained two cross reacting bands
that co-migrated with the mature mtACP-l and acylated mtACP-l proteins. The
acylated form of mtACP-l was approximately four times more abundant than the
rmacylated form and appeared to be localized predominantly in the mitochondrial
membrane where as the unacylated mtACP-l was present mostly in the matrix
fraction. Using a chloroplast fatty acid synthase (FAS) system, mtACP-l was
able to function as a co-factor for fatty acid synthesis. However, predominantly
short and medium chain fatty acids were produced in FAS reactions

supplemented with mtACP-l , suggesting that mtACP-l may be causing

 

1The material in this chapter was published previously [Shintani DK,
Ohlrogge JB (1994) Plant Physiol 104: 1221-1229]

53

54
premature fatty acid chain termination.

 

Introduction

Acyl carrier proteins (ACPs) are protein co-factors essential in de novo
fatty acid synthesis (Ohlrogge, 1987). ACPs fimction to shuttle the elongating
fatty acid chain through the reactions of fatty acid synthesis. These proteins are
also involved in the initial desaturation and acyl-transfer of 16 and 18 carbon
fatty acids (Guerra et al., 1986; Ohlrogge, 1987). In animals and fungi, ACP
exists as a sub-domain of the mulitfunctional fatty acid synthase polypeptide
(W aldl et al.,1983), whereas in plants and most bacteria, ACPs are soluble low
molecular weight acidic proteins (Ohlrogge, 1987).

Although ACPs are most commonly associated with fatty acid
metabolism, ACPs are also known to frmction in the synthesis of membrane
derived oligosaccharides (Therisod et al.,1986) and polyketides (Shen et
al.,1992). ACP has also been implicated as an acyl-donor involved in the
activation of haemolysin, an E. coli membrane targeted toxin (Issartel et al.,
1991)

Recently, ACPs have been reported to occur in mitochondria of
Neuraspora and bovine heart muscle (Brody and Mikolaj czyk, 1988; Sackmann
et al., 1991; Runswick et al., 1991). The role of mitochondrial ACP is not well
understood. The mtACP has been identified as a component of the
NADqubiquinone oxidoreductase complex (complex I) of Neu rospora
(Sackmann et al., 1991) and bovine heart muscle mitochondria (Runswick et al.,
1991). The acylated form of the mtACP is found to be associated with a matrix
domain of the membrane bound NADqubiquinone oxidoreductase complex.
The function of the mtACP in complex I has not yet been determined. It has also

wworted that Neuraspora mitochondria are capable of de nova fatty acid

55

56

synthesis (Mikolajczk & Brody, 1990; Zenson et al.,l992), implying that ACP
may act as a co-factor of mitochondria FAS. However, because the animal or
fungal cytoplasmic FAS is thought to be sufficient to maintain fatty acid
metabolism in the cell, it is unclear why this redundant pathway exists. It has
been suggested that mitochondrial FAS may be necessary for the synthesis of
mitochondria specific membrane lipids such as cardiolipin (Mikolajczyk &
Brody, 1990; Zenson et al., 1992). In this regard, Brody et al. (1990) reported
that in vitro labeling of Neuraspora mitochondria with [2-“C] malonic acid
resulted in the radiolabel accumulating as myristoyl moeities in cardiolipin and
phospholipid fractions.

In this report, we describe the isolation and characterization of a putative
Arabidopsis mtACP cDNA clone. This acyl carrier protein, mtACP-1, shares
a high degree of sequence similarity to the Neuraspora and bovine mtACPs and
contains an N-terrninal extension that resembled a mitochondrial presequence.
The subcellular localization of this ACP and its relationship to fatty acid

synthesis are the focus of this research.

Materials and Methods

Screening of the cDNA Library

The mtACP clone, pACP25, was isolated from a A. ZAP (Stratagene, La
J 0113, CA) cDNA library constructed from mRNA purified from cold acclimated
Arabidopsis plants (Lin C & Thomashow MF, 1992). Approximately 60,000
plaques transferred to duplicate nitrocellulose filters (Sambrook et al. 1989)
were screened using a 5' end 32P-labelled (Sambrook et al., 1989) 66 nucleotide
oligomer corresponding to a region containing the phosphopantetheine binding

57

site (Hansen, 1987). The filters were prehybridized for 2 hours and hybridized
overnight in 5 X SSPE, 1% (w/v) SDS, 0.1% (w/v) blotto at 42°C. The filters
were then washed three times for 15 minutes in 5 X SSPE, 1% (w/v) SDS at
42°C and then used to expose x-ray film overnight at -80°C with intensifying
screens. pBluescript SK- (Stratagene) plasmids containing desired cDNA
sequences were excised fiom the A ZAP vector (Stratagene) as described by
Short et. a1 (1988). All further manipulations were performed in pBluescript
KS+ as described by Sambrook et a1 (1989). The identities of positive cDNA
clones were confirmed by double stranded dideoxy sequencing with Sequenase
2.0 (U .8. Biochemical, Cleveland, OH) (Sanger et al., 1977).

in vitro Transcription/Translation of Arabidopsis pre-mtACP-I and Spinach
pre-cpACP-I precursors.

The DNA sequence corresponding to the mtACP-1 precursor peptide was
amplified away from the non-coding sequences of the cDNA clone pACP25 by
polymerase chain reaction (PCR) (Mullis and Faloona, 1987). The forward
primer (J0107 = 5'-CGGGATCCATGGCACTGAGAAATGC-3') and reverse
primer (J 094 = 5'-CGGGATCCGCTAGACATTGGATGATT-3') had BamHl
sites engineered into their respective 5' ends to facilitate subcloning into
pBluescript H KS+. The PCR reaction contained 2.0 uM each primer, 10 ng of
pACP25 template DNA, 2.5 imits Taq DNA polymerase (Perkin-Elmer Cetus,
Emeryville, CA), 200 nM each dNTP, 50 mM KCl, 10 mM Tris-HCl, pH 8.3,
1.5 mM MgCl2 0.01% (w/v) gelatin in a 100 nL volume. The reaction mix was
first denatured for 2 minutes at 94°C. The sequence was amplified using 30
cycles of the following thermocyle profile: 1 minute at 94 °C, 2 minutes at 55 °C,
2 minutes at 72°C. The PCR product was digested with BamHI, subcloned into

58
pBluescript H KS+, and was designated p0527. p0527 was linearized with
EcoRV and used as a template to generate pre-mtACP-I transcript using T7
RNA polymerase (Boehringer Mannheim, Indianapolis, H1). The spinach
cpACP-l RNA template construct (Schmid & Ohlrogge, 1990) was linearized
with BamHI and the pre-cpACP-l transcript was synthesized using T3 RNA
polymerase (Boehringer Mannheim). 10 ug of each precursor transcript was

translated in the presence of 35S-methionine using a rabbit reticulocyte lysate

system (Promega, Madison, WI).

Mitochondrial and Chloroplast Protein Import Studies

Mitochondria were purified from 400 g of fourteen day old green pea
seedlings (Little Marvel) (Burpee Seeds) grown at 20°C on a 12 hour light/dark
photoperiod as described by Fang et a1 (1987). The intactness of the
mitochondria was determined by the latency of cytochrome C oxidase activity
after Triton-X-100 treatment (N eurburger et al., 1982).

Mitochondrial protein import reactions were performed as described by
Watts et al. (1992) with the following modifications. Mitochondria
corresponding to approximately 150 p g of protein were incubated for 1 hour at
25°C with 20 uL of in vitro translation product in a final volume of 100 uL 0.3
M sorbitol, 20 mM TBS, 1 mM DTT, 0.5 mM EDTA, 2 mM K,HPO,, 20 mM
KCl, 9 mM creatine phosphate, 5.0 mM methionine, 1 mM NADH, 120 ng/mL
creatine kinase, 2 mM ATP, pH 7 .2. The import reactions were then incubated
on ice for 15 minutes followed by an additional 30 minutes with or without 20
ug/mL proteinase K. PMSF was added to a final concentration of 10 mM and
the reactions were incubated on ice for 15 minutes. Intact mitochondria were
reisolated by centrifugation at 5000 g for 5 minutes through a 1 mL 26% (v/v)

59

Percoll cushion containing 0.25 M sucrose, 0.1% (w/v) BSA, 10 mM TES pH
7.2. The mitochondrial pellets were washed with 200 11L of 0.25 M sucrose,
0.1% (w/v) BSA, 10 mM TES pH 7.2 and repelleted at 5000 g for 5 minutes.
The mitochondria were resuspended in 50 uL of IX SDS-PAGE sample buffer
and boiled for 5 minutes. Where noted, the mitochondria were resuspended in
100 mM Tris pH 9.0, 100 mM DTT, incubated at 65°C for 30 minutes and then
precipitated with 5% (w/v) TCA. The resulting TCA pellet was then
resuspended in 50 uL of 1X SDS-PAGE sample buffer. The mitochondrial
extracts were electrophoresed on 15% SDS-polyacrylamide gels and blotted
onto nitrocellulose filters (Battey and Ohlrogge, 1990). Radioactivity on the
blots was detected and analyzed by phosphorimaging (Molecular Dynamics,
Stmnyvale, CA).

Spinach chloroplasts were isolated from spinach leaves as described by
Roughan (1987). A 200 uL import reaction mixture containing 20 nL of in vitro
translation product, 8 mM ATP, 0.33 M sorbitol, 50 mM HEPES/KOH pH 8.0,
and 100 uL of chloroplasts was incubated for 45 minutes at 25°C. The reactions
were incubated on ice for 30 minutes with or without the addition of 20 ug/mL
proteinase K (Boehringer Mannheim). PMSF was added to a final concentration
of 10 mM and the reactions were incubated for an additional 15 minutes on ice.
Intact chloroplasts were re-isolated by centrifugation at 1,500 g for 6 minutes
through a 1 mL cushion of 40% (v/v) Percoll in 50 mM HEPES/KOH, pH 8.0,
0.33 M sorbitol. The chloroplast pellet was washed in 50 mM HEPES/KOH pH
8.0, 0.33 M sorbitol and repelleted by centrifuging at 1,500 g for 6 minutes. The
washed chloroplasts were resuspended in 50 nL 25 mM I-IEPES, pH 8.0 and
incubated on ice for 15 minutes to cause lysis. The lysed chloroplasts were then
centrifuged at 14,000 g for 10 minutes. The resulting supernatant was analyzed

60

on a 15% SDS-polyacrylamide gel and blotted onto nitrocellulose as described
above. The blot was then used to expose X-ray film.

Expression of pACP25 in E. coli and polyclonal antibody production
PCR was used to facilitate the cloning of the mature mtACP-l sequence
into the E. coli protein expression vector, pET3a (Studier et al., 1990). The
forward primer, J095 (J095 = 5'-
CGGGATCCATATGTCGCACGATGATCATCTT-3') contained a Nde I site
designed into the 5' end. The M13 reverse sequencing primer was used as the
reverse primer. The mature peptide sequence was amplified by PCR as
described above. The PCR product was cloned into the EcoRV site of
pBluescript H KS+ and then subcloned as a NdeI BamHI fragment into pET3a.
The resulting plasmid was designated p0914A-10. The induced
expression of mtACP-1 from E. coli BL21 (DE3) cells carrying the plasmid
p0914A-10 by IPTG resulted in the production of an abundant protein with a
molecular mass of approximately 18 kD which was easily detected in Coomassie
blue stained SDS-PAGE gels. To produce antibodies to this protein, a three
liter culture of E. coli strain BL21 (DE3) containing p01914A-10 was induced
by 0.4 mM IPTG when the culture density reached an optical density of 0.7 at
600 nm. The culture was incubated for an additional 6 hours at 37 °C after
which the cells were harvested by centrifugation at 6000 g for 20 minutes. The
cell pellets were washed in homogenization buffer (50 mM MOPS pH 7.5, 10
mM DTT) and resuspended in 30 mLs of the same solution containing 1 mg of
DNase I (Sigma). The cells were lysed in a French pressure cell at 20,000 psi.
The lysate was diluted to 100 mL with homogenization buffer and centrifuged
at 10,000 g for 20 minutes. (NI~I,),SO4 was added to the supernatant to 60%

61

saturation and was stirred for 30 minutes. The insoluble material was removed
by centrifugation at 10,000 g for 30 minutes. TCA was added to the supernatant
to a final concentration of 5% (w/v), and was then mixed slowly for 30 minutes.
The TCA insoluble material was pelleted at 10,000 g for 30 minutes. The TCA
pellet was washed with 1% (w/v) TCA and resuspended in 40 mL of 100 mM
Tris pH 9.5, 100 mM DTT. The resuspended TCA pellet was incubated at 65 °C
for 30 minutes to deacylate ACPs. The deacylated ACPs were then precipitated
in 5% (w/v) TCA as described above. The washed TCA pellet was resuspended
in 40 mL of 50 mM MES pH 6.1, 10 mM DTT. The mitochondrial ACP was
separated from E. coli ACP on a DEAE cellulose (Whatrnan) column using a
linear gradient of0 - 0.5 M NaCl in 10 mM MES pH 6.1,10 mM DTT. Column
fractions were analyzed for ACP using an acyl-ACP synthetase assay (Kuo and
Ohlrogge, 1984). Fractions containing mitochondrial ACP were pooled, TCA
preciptated, and resuspended in 10 mM Tris/HCl pH 7.5. The mtACP was
further purified on a Mono-Q HRS/5 column (Pharmicia) using a gradient of 0 -
0.3 M MgCl2 in 10 mM Tris/HCl pH 7.5.

Antibodies were generated by injecting the E. coli expressed mtACP-1
that was emulsified with Hunter's TiterMax Research Adjuvant R1 (Cthx
Corp., Norcross, GA) into New Zealand white rabbits.

Extraction of Arabidopsis ACPs

Arabidopsis leaf ACP extracts were prepared from 10 g of Arabidopsis
leaves. The leaves were ground in liquid nitrogen and homogenized in 100 ml
of 10 mM HEPES pH 7.8, 10 mM DTT using a polytron. The homogenate was
centrifuged and the supernatant was TCA precipitated and deacylated as
described above. The deacylated Arabidopsis leaf ACP extract was

62
resuspended in 2.5 mL of 10 mM HEPES pH 7 .8, 10 mM DTT and stored at -
20°C in 500 pL aliquots.

Arabidopsis mitochondria were isolated from 18 day old plants grown at
23°C in a 12 hour photoperiod. Plants were placed in the dark for 12 hours
prior to mitochondria isolation. Mitochondria were extensively purified on
continuous PVP-40/Percoll gradients as described by Day et al. (1985). The
isolated mitochondria contained 266 ug/mL protein. The specific activity of
cytochrome C oxidase in this preparation was 77 nmoles/min/mg protein. The
mitochondria were 86. 5% intact as determined by the latency of cytochrome C
oxidase activity before Triton X 100 treatment (N eurburger et al., 1982). The
mitochondria were free of chloroplast contamination as determined by

chlorophyll content. Mitochondria extracts were deacylated as described above.

FAS Assay

Spinach chloroplast fatty acid synthase (FAS) was purified as described
by Guerra et a1 (1987). Spinach cpACP-l and E. coli ACP were purified as
described by Ohlrogge and Kuo (1985) and Rock and Cronan (1980)
respectively. FAS assays contained 1 mM NADH, 2 mM NADPH, 1 mM
acetyl-CoA, 5 mM [i-mercaptoethanol, 100 ug/mL spinach leaf protein, 100 mM
Tris/HCl pH 8.0 , 400,000 DPM/mL of 54 nCi/umole [2-"'C] malonyl-CoA.
FAS reactions were supplemented with mtACP-1 or cpACP-l at designated
concentrations. The reactions were incubated at 30°C. The reactions were then
saponified with 0.1 volumes of 10 M KOH and incubated at 65°C for 15
minutes. 0.2 volumes of 5 M H2SO4 and 1 volume of a solution of 1 mM
palmitic acid in 1M acetic acid in isopropanol were added to each sample. The
saponified fatty acids were then extracted three times with 2 volumes of hexane

63
and analyzed by scintillation counting. F AS reaction products were dried under
N2 and derivatized in 1 mL of boron trichloride at 90°C for 10 minutes. 1 mL
of H20 was added and the fatty acid methyl esters (FAME) were extracted 3
times with 2 mL of hexane. The FAME were then separated on KC18 reverse
phase TLC plates (Whatrnan, Maidstone, England) developed in
acetonitrilezmethanolzwater (65:35 20.5 vol/vol). The amount of 1"C label

incorporated into each FAME species was determined by phosphorimaging
analysis of the TLC plates.

Results and Discussion

Isolation of a cDNA encoding a mitochondrial form of acyl carrier protein

An oligonucleotide probe derived from a highly conserved amino acid
sequence surrounding the ACP phosphopantetheine binding site (Hansen, 1987)
was used to screen an Arabidopsis leaf cDNA library. Forty positive cDNA
clones were isolated in this screening, twelve of which were purified and
sequenced. Seven of the twelve cDNAs corresponded to two previously cloned
Arabidopsis ACP isoforms, ACP-2 (fHousek-Radoj cic et al., 1992) and ACP-3
(Lamppa & Jacks, 1991). The identity of four of the sequenced cDNAs was not
conclusively determined. The remaining cDNA clone, pACP25, appeared to
encode a novel Arabidopsis ACP isoform.

pACP25 contained a 2400 base pair cDNA insert. Restriction mapping
and Southern blot analysis using the 66mer probe indicated that the ACP
sequence was located in a 950 base pair BamHi HindIII fragment located in the
3' end of pACP25. This cDNA fragment was subsequently sequenced and two
tandem open reading frames were detected. One open reading frame encoded

64

a protein that showed a high degree of identity to the Pseudomonas enzyme, N-
carbamyl-L-amino acid amidohydrolase (W atanabe et al., 1992). The second
open reading flame, encoded a protein showing a high degree of identity to ACP.
The two reading flames were separated by an EcoRI site suggesting that two
independent cDNA fragments were could have been inadvertently subcloned
into the phage vector during the library construction. Because the two open
reading flames appeared to be related to two proteins that have very different
functions, it is unlikely that the two open reading flames originated flom the
same mRNA species.

The protein encoded by the second open reading flame of pACP25,
designated mtACP-1, contained the phosphopantetheine binding domain which
is highly conserved among all ACPs (Ohlrogge, 1987). A putative translation
initiation sequence (Heidecker & Messing, 1986) (underlined in Fig. 3-1)
surrounded the AUG initiation codon. No polyadenylation site consensus
sequence or poly (A) tail was found in the 3' untranslated sequence.

The mature mtACP-1 protein sequence was compared to the Arabidopsis
cpACP sequences and the known mtACP sequences (Fig 3-2). Although the
amino acid sequence corresponding to the region surrounding the prosthetic
group attachment site (underlined in Fig. 3-2) was highly conserved among
mtACP-1 and all other ACPs, two amino acid changes common to each of the
mtACPs were detected in this region. First, at position 44 (denoted by * in Fig.
3-2), a highly conserved alanine seen in all cpACPs was replaced by a leucine
in the mtACPs. A second alteration in the conserved sequence of the cpACPs
was seen at position 54 (denoted by " in Fig. 3-2) where a highly conserved
glycine was replaced by an alanine in the mtACP sequences. The region where

these changes occur is totally conserved among all known cpACPs sequences

Figure 3-1 Nucleotide and deduced amino acid sequence of pACPZS. The

65

m a 1 r n a i 1 r h 1 r v p v q
QAAAGAGACGAAAATGGCACTGAGAAATGCAATTCTTCGTCACCTGAGGGTTCCGGTGCA
t 1 g 1 n q s k i g f 1 g t i r a f S S
AACCCTAGGATTGAATCAGTCTAAAATTGGGTTCCTTGGTACGATCCGGTCATTTTCTTC
H D D H L S R E A. V 'V D R ‘V L D V ‘V K S
GCACGATGATCATCTTAGCAGAGAAGCTGTCGTCGATAGAGTACTCGATGTTGTCAAGAG
F P K V’ D P S K V‘ T P E V H P Q N D L G
CTTCCCCAAAGTCGATCCCTCTAAGGTGACTCCTGAGGTTCATTTCCAAAACGATTTGGG
LDILDTVIIVMAIIIII’KLE
ATTAGATAGTTTGGACACAGTGGAGATAGTGATGGCTATTGAAGAGGAATTCAAGCTGGA
I P D K E .A D K I D S C S L .A I E Y V' Y
AATTCCAGACAAAGAAGCTGACAAGATCGATTCTTGCTCTCTCGCCATTGAAIACGTTTA
N H P M S S
CAATCAICCAATGTCTAGCTAATCGCTGTGTGCTCTCTCATGGTTTCTCCTTCCTTTTTC
TTGTATGAGGTTCTTCTCTCTTTGAIAATGCCATAATGGGATTTGGCAAAAAAATCACTT
TTATTTGGAGAGCATTGTTATTGACTCTTTTAACGGTTCCAGTTTCTTAAAAAGAAAACC
TTTTTGGAATTCC

putative mitochondrial presequence sequence is denoted by the lower case single

letter amino acid code. The underlined nucleotide sequence represents the

translational initiation consensus sequence (Heidecker and Messing, 1986). The

bold upper case single amino acid sequence represents the conserved sequence

surrounding the phosphopantetheine attachment site.

66

.38: sea as amend $5.:
:28an cases £30.sz n 29.9 a... a: as s 03 2a 3:83 20... Hanson... seas ”Estes: u
2030 2 fies: a a 5:583:80 20... 038520 sense £8381: u 23.8 2 53: a a
8:383: :0... FEBS? 8.58 u :03: 0: ”ES: .3 a sausage :0... 36:28? See Eoeesz
n mugs: 0 Z 50¢. 35553:: 02333 humoEBEV n 7:038 2 .86 E08583 05058:.mqonnmosm 05
wqfigotsm 08033 “0080800 05 3:823: 8:038 355:5 05. .mm0< 3:30:83: 00:5 05 E 30>u00amob
8:23: 05:0? 0:: 0:650— »: “000030: :00: 023 35 £0»: 0:838:20 :0 E :08 3:200: 053% E: 05:2:
0020800 bawfi 80:38: 33:00:00: < :5: ._. $200 05:8 3200.50 08:0: 3 afloat. «nines—:0 fiumommeeev.

05 .53 :5 we: 08¢: itceoaooza :32:— 5? A3 00:250.». 502:: VHDAJE we Bonn—52¢. Tm 0.5mm...—

67

MdfiUUAAflfiHA<¢u0fl>H4umouxmfidxuH0»u”aqua“nuuan”u«(J"((fixaddfiHmndmxnmqoxmuu¢Uu8uuBfim¥< MImU(QO

 

X(XMhHAmmHA(<uOW>HflumouxmmdzuHunuu”awn”uuu"u“u“ta«(dfifihédfiHflQdflKflmAOMmu"flung””Bumx<. Nlmudflo
KKNZHANMHAu<uOfl>B<u«Ouxmfi(ZGquuu"A0"uuuuuu“u«(JH"¢E¥BH(5>¥uQuBAmAO&u"HdmuKmHBHOIC HIQUdQU
mmZNmzw>MHH¢AmUmQondmxamquxhummH¢2>HH>BQAmQflOQQZObm>NmB>¥mmn>¥mhmx>>nn>mn>>Kfl¢mAonmmm HImU¢UE

_ _ _ < _ e _ _ _ _

om ch ow on o¢ om ON OH

mw>nxxn¢Hun>HmOmoznumu"Huuuumwu"aux"«Huuuonuuuuuaux!"umz>mauwuuHunwnuAufluuuuwaunaummmddu mudufi
z¢du0mAHuu>uxn>umuouuonuu«uHmuu«nuuuu>uuuuuuuuuuuu¢a"(Hmquxuuz>¥ah0mn>ddemh>OQXXAmUQ<d mQ<£E
mm2mzzw>¥deflmUmQHxnimxnmHwnthmuH¢Z>HN>BOAmQQUAQZOhm>mmB>¥mmn>¥mhmx>>DH>MQ>>KMmmflmnnmmm Hlmudufi
_ _ _ <_ 0.... _ _ _
cm on cm on ov on em OH

 

68

but is difi‘erent in mtACP-l and the bovine and Neuraspora mtACP sequences.
In addition to these specific changes in the highly conserved phosphopantetheine
binding domain, when compared to known ACP protein sequences, mtACP-1
shared the highest degree of identity with the mtACPs. The Neuraspora and
bovine heart muscle mtACP protein sequences are 68% and 67% identical to
mtACP-1 respectively, whereas the Arabidopsis cpACPs share less then 60%
identity with mtACP-1. The Arabidopsis cpACP sequences share at least 69%
identity among themselves (Hlousek-Radoj cic etal.,1992); thus the sequence of
the mtACP-1 protein appeared to be more conserved with widely divergent
organisms of three eukaryotic phylogenetic kingdoms than with cpACPs isolated
from the same species, Arabiabpsis. The alignment of mtACP-1 amino acid
sequence with the Neuraspora and bovine mtACPs sequences suggested that
mtACP-1 encoded an 88 amino acid mature protein preceded by a 35 amino acid
N-terminal extension. Both the calculated pI of 4.26 and molecular mass of
9,914 daltons were consistent with other ACPs of both chloroplast and
mitochondrial origin.

The 35 amino acid N-terminal extension (small case letters in F ig.3-1)
shared many of the structural features common to mitochondrial presequences
as defined by von Heijne et a1 (1989). These features included a high degree of
alpha helical nature at the N-terminus, hydrophobic amino acids at the C-
terminus of the presequence between amino acids 24 and 34, and a highly
conserved arginine at position 32. Using these criteria it was hypothesized that
the first 35 amino acids of the protein encoded by pACP25 represented a

mitochondrial presequence.

69

E. coli Expression of pACPZS

The cDNA sequence encoding the mature mtACP-1 protein was
subcloned into the E. coli protein expression vector pET3a. mtACP-1 could be
clearly separated fi'om the E. coli ACP by DEAE-cellulose anion exchange
chromatography and was purified to near homogeneity as determined by silver
staining of SDS-PAGE gels (data not shown). The E. coli mtACP-l migrated
as a 18.0 kD doublet when analyzed by SDS-PAGE. Pure acyl carrier proteins
have been reported to sometimes migrate as doublets during PAGE (J ackowski
and Rock, 1987). The purified mtACP-1 protein was used to generate
polyclonal antibodies in rabbits. The resulting antibodies were highly specific
for mtACP-1. Although the antibodies did cross react with other ACP isoforms
due to the high degree of similarity shared among all ACP isoforms, these
antibodies showed a marked preference for mtACP-l. At a 1:500 dilution, the
crude antiserum was sufficient to detect as little as 10 nanograms of mtACP-1
on a Western blot (data not shown).

In vitro protein import studies.

To confirm the targeting specificity of the mtACP-1 presequence, in vi tro
mitochondrial and chloroplastic protein import studies were conducted. The
competency of mitochondria and chloroplasts for the import of the mtACP-1 and
spinach cpACP-1 precursors was compared. Pea mitochondria and spinach
chloroplasts were incubated in the presence of radiolabelled apo-mtACP-l or
spinach apo-cpACP-l precursor proteins with and without the addition of
proteinase K. Proteins imported into the organelles were protected from
digestion by exogenously added protease and were detected in mitochondrial or
chloroplast extracts by autoradiography. The identities of processed proteins

70
were determined by mobilities on SDS-PAGE gels relative to mobilities of ACP
standards.

The mtACP-1 presequence proved to be sufficient for targeting protein
import into pea mitochondria. Although several radiolabelled proteins appeared
to be associated with the reisolated mitochondria after incubation with in vitro
translated apo-mtACP precursor (Fig. 3-3, lane 1), only two proteins, with
apparent molecular masses of 18 kD and 12 kD, were protected from protease
treatment (Fig. 3-3, lane 2).

The 18 kD radiolabelled protein co-migrated with the mature—holo
mtACP-1 standard (Fig. 3-3, lanes 2 and 5) and the 12 kD band appeared to
migrate slightly higher than the mature palmitoyl-mtACP-l standard (Fig. 3-3,
lanes 2 and 4). Because the 12 kD band did not migrate exactly with the
palmitoyl-mtACP-l standard, it is possible that the mtACP acylated during the
import reaction was composed of a fatty acid of a different chain length since
acyl-ACPs are known to migrate on native PAGE differently depending on the
fatty acids chain length (Post-Beittenmiller et al., 1990). To confirm that the 12
kD protein did correspond to an acylated form of mtACP-1, post-import
mitochondrial extracts were deacylated by treatment with 100 mM DTT at pH
9.0 (Lakin-Thomas and Brody, 1985). Upon deacylation, the 12 kD
radiolabelled band disappeared while the presence of the 18 kD band was not
affected (Fig. 3-3, lane 3). Thus it appears that the pea mitochondria were not
only able to import and process the mtACP-1 precursor, but were also able to
acylate the mature mtACP-1 protein. These results contrast with chloroplast
protein import experiments where acylated forms of ACP were not detected
(Fernandez and Lamppa, 1990a and 1990b). Although Neuraspora
mitochondria have been reported to be capable of de novo fatty acid synthesis

71

Figure 3-3 in vitro protein import of mtACP-l precursor into pea leaf
mitochondria. Mitochondrial protein import studies were performed as
described in the Materials and Methods section. The radiolabelled proteins were
detected on protein blots using a Molecular Dynamics phosphorimaging system.
Lanes 1-4 were derived from images obtained when the background to signal
range was set at 10 to 1000 counts. Lanes 6 and 7 were derived from images
obtained when the background to signal range was set at 10 to 50 counts. Lane
1, import mixture containing pea mitochondria and radiolabelled mtACP-1
precursor. Lane 2, same as lane 1 but the mitochondria were incubated with
proteinase K afier import. Lane 3, same as lane 1, but the mitochondria were
treated with proteinase K and then deacylated (Lakin-Thomas and Brody, 1985).
Lane 4, acyl-mtACP-l standard that was synthesized in vitro from mtACP-1 and
14C-palmitic acid using E. coli acyl ACP synthetase (Kuo and Ohlrogge, 1984).
Lane 5, mtACP-1 protein standard derived from E. coli expressed pACP25 and
detected by Western-analysis. Lane 6, radiolabelled mtACP-1 precursor
standard synthesized from rabbit reticulocyte lysates and transcripts derived
from p0527. Lane 7, radiolabelled apo-mtACP-l synthesized fi'om rabbit
reticulocyte lysate and transcripts derived from p0914A-10.

protease

deacylated

mature holo-
mtACP-1

acyl mtACP-1

 

-+.+
--+
a

—>"""" “has

1‘ 3"
- . ‘0. '2‘ .r

72

acyl mtACP-1

 

standards
5
95’ '7
e 8
a E
:2 g
‘3 a
E a.

mature apo-mtACP-t

I b ‘ v' .
.0 filth-‘3' ‘
.. ,

°. '27.:-
, "-1.

'.- 2‘.

o ‘ «\a ..9
. a.
'fl-z: 1.
‘30. Ol-

‘ l

7

pre-apo-

. <— mtACP-1

mature apo-
mtACP-1

 

73
(Mikolajczk and Brody, 1990; Zensen et al., 1992), it is not known if the acyl
chain esterified to the mature mtACP-1 is derived from mitochondrial de novo
fatty acid synthesis or from an acyl-transfer reaction from a preformed acyl
group.

Interestingly, no radiolabelled protein was detected in the post-import
mitochondrial extract that co-migrated with the mature apo-mtACP-l standard
(Fig. 3-3, lane 7). This suggests that the pea mitochondria are able to very
efficiently add the phosphopantetheine prosthetic group to imported mtACP-1
protein. These results contrast with chloroplast protein import studies in which
the complete conversion of the apo-cpACP I to the holo form was not seen
(Fernandez and Lamppa, 1990a and 1990b). Our results further indicate that the
mitochondria may contain holo ACP synthetase (HAS), the enzyme responsible
for the addition of the prosthetic group. Elhussein et al. (198 8) have reported
that HAS activity was associated with both cytoplasmic and plastid cell fractions
with relatively little HAS activity being detected in the mitochondria. Although
the results shown in figure 3 strongly suggest that the pea mitochondria possess
HAS, it is possible that the mtACP precursor was modified prior to import by
contaminating cytoplasmic HAS present in the mitochondria preparation.
However, it is not likely that the prosthetic group was added during the in vitro
translation reaction since it is known that the rabbit reticulocyte lysates contain
no HAS activity (Linda Savage & Dusty Post-Beittenmiller, unpublished data).
In additional control experiments it was shown that the spinach cpACP-1
precursor was not able to direct protein import into the mitochondria (data not
shown).

When the mtACP-1 precursor was incubated with spinach chloroplasts,

it was able to bind to the outer chloroplast envelope (Fig. 3-4, lane 1) however,

74

protease - +

Dre-apo-mtACP-r
standard

.. pre-apo-
‘” mtACP-1

Figure 3-4 in vitro protein import studies of the mtACP-1 precursor into
spinach chloroplasts. Chloroplast protein import studies were performed as
described in the Materials and Methods section. The radiolabelled proteins were
detected by autoradiography. Lane 1, import mixture containing spinach
chloroplasts and radiolabelled mtACP-1 precursor. Lane 2, same as lane 1 but
the chloroplast were incubated with proteinase K afier import. Lane 3,
radiolabelled mtACP-1 precursor standard synthesized fiom rabbit reticulocyte
lysates and transcripts derived from p0527 .

75
after protease treatment, no chloroplast import had occurred (Fig. 3-4, lane 2).
However, as expected, the spinach cpACP-1 precursor was imported and

correctly processing by spinach chloroplasts (data not shown).

Sub-cellular localization

To confirm the mitochondrial localization of mtACP-1 , Western analysis
of Arabidopsis leaf extracts was performed. Blots of Arabidopsis protein
extracts run on 15% SDS-PAGE gels were probed with antibodies specific for
mtACP-1. To simplify the number of potential forms of mtACP-1, protein
extracts were first deacylated (Lakin-Thomas and Brody, 1985). A 18 kD
doublet that cross reacted with the mtACP-1 antibodies was detected in
deacylated whole leaf and deacylated purified mitochondria extracts (Fig. 3-5,
lanes 1 and 2). These proteins co-migrated with the E. coli expressed mtACP-1.
The anti-mtACP-l antibodies weakly cross reacted with proteins that migrated
to the known positions of the Arabidopsis cpACP isoforms (Battey and
Ohlrogge, 1990) in whole leaf extracts (Fig. 3-5, lane 1). Based on the total
ACP content of Arabidopsis leaf, as estimated from acyl- ACP synthetase, and
the level of mACP-l , as approximated from Western blots, we determined that
the mtACP-l may constitute as much as 10% of the total Arabidopsis leaf ACP.

In addition to the rmacylated form of mtACP-1, intact Arabidopsis
mitochondria contained a cross-reactive protein that co-migrated with the acyl-
mtACP-l synthesized in vitro (Fig. 3-5, lane 3). As estimated from the staining
intensity of Western blots, the acylated form of mtACP-1 appeared to be present
at approximately 4 times the level of free mtACP-1 in purified mitochondria.
This result differs from spinach cpACPs where the free ACP accounted for
approximately 60% of the cpACP pool (Post-Beittenmiller et a1, 1990). Further

76

.2 .2
3 3’2 3'2
2: 22 32
:72 38: :E
8 8T: '5
18-4KD— 6. ——mtACP-1
14.3kD —
..... ~ —acyI-mtACP-1
6.2kD —
- —cpACPs
3.0kD --
1 2 3

Figure 3—5 Western-blot analysis of Arabidopsis leaf sub-cellular fractions.
Arabidopsis whole leaf and mitochondrial protein extracts were purified as
described in the Materials and Methods section. Where noted, the extracts were
deacylated by treatment with 100 mM DTT at pH 9.0 for 30 minutes at 65 °C.
Deacylated leaf protein extract (lane 1), deacylated mitochondrial protein extract
(lane 2), untreated mitochondrial protein extract (lane 3).

77
analysis showed that the acylated form of mtACP-1 was primarily associated
with the mitochondrial membrane fraction. The 100,000 g pellet of the lysed
mitochondria appeared to contain the majority of the acylated mtACP-1 along
with a small amount of free mtACP-1 (Fig. 3-6, lane 2). The free mtACP-1 was
present in relatively higher amounts than the acylated form in the 100,000 g
supernatant fraction (Fig. 3-6, lane 3). The membrane localization of the
acylated mtACP-1 was consistent with the report that the Neuraspora and
bovine mtACPs were only present in the acylated form when associated with the
NADqubiquinone oxidoreductase membrane complex (complex I) (Sackmann
eta1.,1991; erswick et al. 1991), further evidence is necessary to determine if

mtACP-1 is also associated with complex I in Arabidopsis mitochondria.

Fatty acid synthase (FAS) assays

Although Neuraspora mitochondria have been reported to be capable of
de novo FAS (Mikolajczyk & Brody, 1990; Zensen et al., 1992), no direct
evidence exists to show that the mtACP functions as a co-factor for fatty acid
synthesis. To determine whether or not mtACP-1 was able to act as a co-factor
in FAS, in vitro chloroplast FAS activities were compared using mtACP-1 and
spinach cpACP-1 concentrations ranging from 0.5 uM to 25 uM. The assays
were performed using concentrations of chloroplast F AS extracts that would
allow linear rates of FAS activity during a 15 minute time period. Although the
rates of fatty acid synthesis varied depending on the F AS preparation, mtACP-1
was in all cases able to act as a co-factor for spinach chloroplast FAS when
compared to spinach cpACP-1 (Fig. 3-7). These results suggest that despite its
structural divergence, mtACP-1 could function as a co-factor for chloroplast
F AS.

l .5
~ 8
H
O
a: 5
h H
0)
E 2 a.
o 0 3
.5 Q U)
0 or or
a a 8
,—, O. 0.
O O
‘5 o o
[I— I— F'
‘84“) -. I <— mtACP-1

14.3 kD

.. <— Acyl mtACP-1

6.2 kD- '

Figure 3-6 Westem—blot analysis of mitochondrial membrane and matrix
fractions. Arabidopsis mitochondria were purified as described in the Materials
and Methods section. The mitochondria were lysed by sonication and
fractionated by centrifugation at 100,000 g. Unfractionated mitochondria (lane
1), 100,000 g pellet from sonicated mitochondria (lane 2), and 100,000 g

supernatant from sonicated mitochondria.

79

 

50
""*— mtACP-1

   

40 + "O” cpACP-1

 

 

 

 

'0
a
S
O
2'
O
O
.5
g 30 "
9 l
E 20 -
a
E
3 10
O
s
D.

o

o 5 10 15 20 25

ACP(uM)

Figure 3-7 Utilization of mtACP-1 and spinach cpACP-l by chloroplast
FAS. 75 uL reactions were incubated for 15 minutes at 30 °C. Activity was
measured as incorporation of 14C malonyl-CoA into the lipid fraction. Lipid
extracted into hexane following saponification and neurtalization of the reaction
mixture were counted by liquid scintillation.

80

The profiles of the fatty acids synthesized by the chloroplast FAS in the
presence of mtACP-l and spinach cpACP-1 were compared. As expected, F AS
reactions incubated with spinach cpACP-1 yielded predominately long chain
fatty acids (Fig. 3-8A). However, when F AS reactions were supplemented with
mtACP-l, short, medium and long chain fatty acids were synthesized (Fig. 3-
8B). Caproic acid, lauric acid, myristic acid, and palmitic acid each made up
approximately 20% of the total fatty acid, while octanoic and stearic acids each
represented approximately 10%. A similar fatty acid profile was observed at
mtACP-1 concentrations ranging from 0. 5 uM to 5 .0 uM (data not shown). To
determine if the short and medium chain fatty acids derived from the FAS
reactions supplemented with mtACP-1 were present as acyl-ACPs, the reactions
were extracted prior to saponification to analyze only free fatty acids (Fig. 3-8A
and B). These analysis showed that the short and medium chain fatty acids
produced in reactions containing mtACP-1 were not esterified to mtACP-1. The
results suggest that the fatty acid products produced on mtACP-1 were
prematurely cleaved from the ACP. Although the accumulation of fatty acids of
intermediate-chain lengths were also observed when plant cpACPs were used
to supplement an E. coli FAS system (Simoni et al., 1967), the fatty acids were
apparently present as acyl-ACP intermediates and not as free fatty acids.

The observation of hydrolyzed short- and medium-chain fatty acids in the
FAS assays supplemented with mtACP-1 was particularly surprising because the
acyl-ACP thioesterases are known to be highly specific for long chain acyl-
ACPs (Ohlrogge et al., 1978). It may be possible that mtACP-1 facilitates
release of shorter chain fatty acids from ACP through its presentation of the
acyl-ACPs to acyl-ACP thioesterases. Brody and Mikolajczyk (1988) reported
that 3-hydroxytetradecanoyl-ACPS have been isolated from Neu rospora

81

Figure 3-8 Product analysis of fatty acid synthase reactions supplemented
with mtACP-1 or spinach cpACP-1. 1“C labelled fatty acids derived from
unsaponifed and saponified FAS reactions supplemented with either 22.5 uM
mtACP-1 or 22.5 uM spinach cpACP-1 were analyzed by reverse phase TLC
and quantitated by phosphorimaging. The FAME profiles derived from
unsaponified (solid bars) and saponified (open bars) FAS reactions
supplemented with spinach cpACP-1 are shown in figure 8A. The FAME
profiles derived from unsaponified (solid bars) and saponified (open bars) FAS
reactions supplemented with mtACP-1 are shown in figure 8B. In this
experiment, the total 14C incorporation into the F AMEs derived from the
unsaponified and saponified FAS reactions supplemented with mtACP-1 was
24,624 DPM and 28,543 DPM respectively. Unsaponified and saponified FAS
reactions supplemented with cpACP-1 had total 1“C incorporations of 48,906
DPM and 43,746 DPM. 1 mL FAS reactions were incubated for 1 hour. Similar

results were obtained in additional experiments.

 

 

 

82

 

100

80 _ - unaaponiflad

1:3 aaponlfiad

96 total fatty acid

 

 

 

 

 

 

08:0 010:0 012:0 014:0 018:0 018:0

 

100

- unaaponitiod

D aaponiflod
60 r

96 total tatty acid

 

'11;

08:0 010:0 012:0 014:0 018:0 018:0

 

83

mitochondria, suggesting that medium-chain acyl-ACP intermediates do occur
on mtACPs. Furthermore, in viva and in vitro labeling of Neuraspora
mitochondria with [2-“C]-malonic acid has resulted in the synthesis of
radiolabelled short- and medium-chain acyl-ACPs and free mryistic acid
(Mikolaj czyk and Brody, 1990). However, additional in vivo labeling
experiments in Neurospora (Zenson et al., 1992) indicated that long-chain acyl
intermediates were synthesizedon mtACP. Acyl-ACPs isolated from bovine
mitochondria were reported to have an acyl moiety with a molecular mass of 302
D, again suggesting that a long-chain acyl intermediate could be synthesized on
mtACPs (RUIISWiCk et al., 1991). Because of these divergent results, it is
difficult to ascertain the significance of the accumulation of short- and medium-
chain fatty acids in the chloroplast FAS assays supplemented with mtACP-1.

From the in vi tra chloroplast FAS experiments it was apparent that
mtACP-1 was structurally close enough to cpACP to at least initiate de nova
fatty acid synthesis, but sufiiciently dissimilar enough to participate differently
in the fatty acid termination reactions. Although the results of these experiments
could not predict what role if any mtACP-1 plays in mitochondrial de nava FAS,
they do suggest that mtACP-l is capable of acting as a co-factor in a FAS
system. Further experiments are necessary to determine if mtACP-1 does

participate in mitochondrial de nova FAS.

Conclusion
The cumulative results presented in this paper strongly support the
conclusion that mtACP-1 represents a mitochondrial localized ACP isoform.
First, mtACP-1 has been shown to be more closely related to the Neuraspora
and bovine mtACPs than to the Arabidopsis cpACPs. Secondly, the protein

84
import studies confirmed the specificity of the mtACP-1 presequence for import
into mitochondria. And finally, the detection of mtACP-l by western blot
analysis in mitochondrial extracts and the apparent membrane association of the
acylated form of mtACP-1 are both consistent with characteristics reported for
the Neuraspora and bovine mtACPs (Sackmann et al., 1991; Runswick et al.,
1991)

Although the function of mtACP-1 has not been determined, mtACP-1 is
clearly effective as a FAS co-factor in in vitro chloroplast FAS studies.
However, further studies using isolated mitochondria will be required to
determine if mtACP-1 plays a similar role in the mitochondria. Also, because
of the high degree of structural similarity that mtACP-1 shares with other
mtACPs, it is possrble that the plant mitochondria ACPs may also be associated
with NADH:ubiquinone oxidoreductase. The apparent membrane localization
of acyl mtACP-1 is consistent with the association of the Neuraspora and bovine
mtACPs with the NADH:ubiquinone oxidoreductase membrane complex
(Sackmann et al., 1991; Runswick et al., 1991). Regardless of the outcome,
further characterization of the fimction of mtACP-1 will expand present ideas
of the roles ACPs play in plant metabolism.

Acknowledgments
We wish to thank John Froehlich, Christie Howard, Dusty Post-Beittenmiller,
and Kathy Schmid for their helpfirl discussion of this manuscript. We would
also like to acknowledge Edgar Cahoon for help with fatty acid analysis and
Linda Savage for help with chloroplast protein import studies.

Literature. Cited .
BatteyJF, Ohlrogge JB (1990) Evolutronary and tissue-specific control of
:1: regssrgnggfbmultrp e acyl carner protein isoforms in plants and bacterra. Planta

Brod S Mikola' k S 1988 Neuras ra mitochondria contain an ac l
carrie¥ prbtein. Eui'fizBiochgm 1725: 353-3?90, y

Brody .S, Mikolajczyk S, Chuman L (1990‘)LStudies on de nova fatty acid
gnthesrs. in mitochondria. In. PJ . Harwoo eds, Plant 11 rd
11rgchemrstry, Structure and Utrlrzatron. Portland Press Lt London, pp 1 7 -

Chuman L, Brody S (1989)Acy1 carrier protein isopresent in the mitochondria
of plants and eucaryotrc mrcro-orgamsms. Eur J Biochem 184: 643-649.

Day DA, Neuburger M, Douce R (1985) Biochemical characterization of
glzrlsorophyll-free mitochondria from pea leaves. Aust J Plant Physrol, 12: 219-

Elhussein SA, Miern k JA Ohlrogge JB (1988) Plant halo-(acyl carrier
protern) synthase. Broc em J 252: 39 .

Fan TK, David NR, Miernyk JA, Randall DD 1987) Isolation and
purification of functronal pea leaf mitochondria ee of chlorophyll
contarrnnatron. Curr Top Plant Biochem Physrol 6: 17 5.

Fernandez MD, Lamppa GK (19903) Acyl carrier protein (ACP) import into
chloro last does not requrre the (phoszphogantetheme: evrdence for a chloroplast
holo- CP synthase The Plant ell : 1 5-206.

Fernandez MD Lamppa GK (1990b) Acyl carrier protein import into
chloroplast: Both the precursor and mature forms are substrates for
phos hopantetheme attachment b a soluble chloroplast holo-acyl carrier protem
syn se. J Brol Chem 246: 722 -7226.

Guerra DJ, Ohlrogge JB, Frentzen M (L986) Activity of acyl carrier rotein
1soforms in reactions of plant fatty acrd me bolrsm. Plant Physrol. 82: 8-453.

Hansen L (1987) Three cDNA clones for barley leaf acyl canier proteins I and
HI. Carlsberg Res Commun 52: 381-392.

Heidecker G, Messin J (1986 Structural analysis of plantgenes. In WR
nggs RL Jones, V W teds, ual Revrews of Plant hysrology, V0137.
Annual Revrews Inc, Palo Alto, CA pp 439-466.

Hlousek-RadojCic A, Post-Beittenmiller D, Ohlrogge JB (1992) E ression
of constrtutrve and tissue specific acyl carner protem isoforms in Ara idapsz's.
Plant Physrol 98: 206-214.

 

85

86

Issartel JP, Koranakis V, Hughes C (1991) Activation of Escherichia coli
prohaemol sm to the mature toxrn by acyl carrier protem-dependent fatty
acylation. ature 351: 759-761.

Jackowski S, Rock CO (1987) Altered molecular form of ac 1 carrier protein

assocrated vvrth -ketoacyl-acyl carrier protern synthase H abF) mutants. J
Bacterial 169: 1 69-1473.

Kuo TM, Ohlrogge JB (1984) Acylation of plant acyl canier proteins by ac 1
121101111 grotem synthetase from Escherichia coli. Arch Brochem Brophys 23 :

Lakin-Thomas PL, Brody. S (1985) A pantothenate derivative is covalently
llrggnd to mrtochondnal proteins in Neuraspora crassa. Eur J Brochem 146: 141-

 

Lamppa G, Jacks C (1991) Analysis of Mo linked genes coding/[for the aczl
2361513 7tlrtrotem (ACP) from Arabidopsis thalzana (Colurnbra). Plant 01 B101 1 :

Lin C Thomashow MF 1992 DNA sequence anal sis of a complementary

DNA for cold-re ated rabi . sis ene car15 an charactenzatron of the
COR15 polypeptr e. Plant Physro 99: 19-525.

Mikolajczyk S, Brody S (1990) De nova fatty acid 8 thesis. mediated by ac 1
carrier protern m Neuraspora crassa mrtochondna. ur J Brochem 187: 43 -

Mullis KB, Faloona FA (1987) Specific synthesis of DNA in vitro via a
polymerase catalyzed chain reaction. Meth Enzymol 155: 335-350.

Neurburger M, Journet EP, Bli y. R, Carde JP, Douce .R ((1982)
canon of plant mitochondria .y 150 cmc centrifugatron 1n ensrty
gradients of percoll. Arch Brochem Brophys 17: 312-323.

Ohlrogge JB (R83 Biochemistry of plant ac 1 carrier roteins. In PK Stumpf,
BE gammsgds, e rochennstry of Plants, Va 10. Aca emrc Press, New York,
PP ' -

Ohlrogge Kuo TM (1985) Plants have isoforms for acyl carrier rotein that
are expresse differently in different trssues. J Brol Chem 260: 80 -8037.

Ohlrogge JB,. Shine WE, Stumpf PK 1978) Fat metabolism in hi erplants.
Charactenzatron of plant acyl-ACP an acyl-CoA hydrolases. Arc Brochem
Brophys 189: 382-3 1.

Post-Beittenmiller D, Jaworski JG, thro e JB (1990) In viva pools of free
and acylated acyl earner rotems in s mach: vrdence for srtes of regulation of
fatty acrd brosynthesrs. Brol Chem 66: 1858-1865.

87

Post-Beittenmiller MA, Hlousek-Radojcic A, Ohlrogge JB (1989) DNA
se uence of a genomic clone encoding an Arabidopsis acyl carrier protein
(A P). Nuclelc Ac1ds Res 17: 1777.

Rock CO Cronan JR 81980) Improved purification of acyl carrier protein.
Anal Blochem 102: 362- 64.

Roughan G (19817)) Fatty acid thesis and. utilization by isolated chloroplasts.
In S Colowrclg 0 Kaplan, e , Methods 1n Enzymology, Vol 148. Acadenuc
Press, San D1ego, CA, pp 327-337.

Runswick MJ, Fearnle IM Skehel JM, Walker JE (1991) Presence of an
acyl carrier protem m ADI-l:ub1qu1none oxrdoreductase fiom bovme heart
mrtochondna. FEBS Lett. 286: 121-124.

Sackmann U, Zensen R, Rohlen D, Jahnke U Weiss H (1991) The acyl
earner totem 1n Neu rospora crassa mltochondna IS a subumt of
NADqu 1qumone reductase (complex I) Eur J Brochem 200: 463-469.

Sambrook J Fritsch EF, Maniatis T (1989) Molecular Cloning: A laboratory
Manual, Ed . Cold Sprmg Harbor Laboratory, Cold Spnng Harbor, NY.

Sanger .F, .Nieklen S Coulson AR (1977) DNA seguencing with chain
termmatmg InhlbltOI'S. Pro Natl Acad Sc1 USA 74: 5463- 467.

Schmid KM, Ohlrogge JB (1990) A root ac 1 carrier grotein-II from spinach
18 also expressed 1n leaves and seeds. Plant 01 3101 1 : 765-77 8.

Shen B, Summers RG, Gramajo H, Bibb .MJ, Hutchinson CR (1992)
Punficatron and characterlzatlon of the acyl earner rotem of the Stre tomyces
glaucescens tetracenomycm C polyketlde synthase. Bactenol 17 4: 3 18-3 821.

Short JM, Fernanda JM, Sorge JA, Huse WI) (1988) A, Zap; a bacterionhage
%5e8)§3r‘eg%18n vector w1th in viva excrsmn properties. Nuc elc Amds Researc 1 :

Simoni RD Criddle RS, Stumpf PK (1967) Fat metabolism in higher plants.
XXXI. Puniicatlon and ropertles of plant and bactenal acyl earner protems. J
Blol Chem 242: 573-5 1.

Studier FW, Rosenburhgr AH, Dunn JJ, Dubendorff JW (1990 Use of T7
RNA polymerase to ect expressron of cloned enes. In ethods of
Enzymology, Vol 185. Academlc Press, San D1ego, C pp 60-89.

Therisod H, Weissborn .AC, Kennedy EP (1986 An essential function for
acyl earner protem 1n the blOS thCSlS .of membrane- enved ohgosacchandes of
Escherichia coli. Proc Natl cad Sc1 USA 83: 7236-7240.

von Hei'ne G, Steppuhn J, Herrmann. RG 1989 .Domain structure of
rmtochon a1 and chloroplast targetmg peptldes. ur J lochem 180: 535-545.

88

Watanabe Ishikawa_T, Mukohara Y, Nakamura H 1992) Cloning; and
se uencmg o the genes involved 1n the conversmn of 5-su Sltuted hydantoms
to .e corres ndmg L-ammo ac1ds from the natlve plasmrd of Pseudomonas sp.
stram NS67 . J. Bactenol. 174: 962-969.

Wakil SJ Stoo Joshi VC 1983 Fa acid 5 thesis and its re ation.
AnnRev Biochgsms :537-579.( ) “y y“ SUI

Watts FZ, Walters AJ, Moore AL (1992) Characterization of PHSPl, a cDNA
encoding a mrtochondnal HSP70 from Pisum sativum. Plant Mol B101 18: 23-32.

Zensen R, Husmann. H, Schneider R, Peine T Weiss H (1992) De nova
syntheSIs and desaturation of fatty acids at the nutochondnal acy earner prote
igtltlglirllllt egalfiDqublqmnone ox1doreductase 1n Neu raspara crassa. FEB

 

Chapter 4
Feedback Inhibition of Fatty Acid Synthesis in Tobacco Suspension
Cells‘.

Abstract

The addition of exogenous oleic acid, in the form of oleoyl-Tween
(Tween-18:1) caused a 3 - 5 fold decrease in the rate of [1-"C] acetate
incorporation into tobacco suspension cell fatty acids. The decrease in acetate
incorporation occurred rapidly upon addition of Tween-18:1 and appeared to be
specific for fatty acid synthesis. In order to elucidate of possible regulatory
steps involved in the feedback regulation of fatty acid synthesis in plant cells,
tobacco cell acyl-ACP intermediates were analyzed using a combination of [l-
14C] acetate labelling and irnmuno-blot analysis. Within 30 minutes of
exogenous lipid addition, acetyl-ACP increased and long chain acyl-ACP
decreased, whereas medium chain acyl-ACP levels remained constant. These
acyl-ACP profiles observed during the feedback inhibition were those predicted
to occur under conditions where the flux through fatty acid synthesis is
decreased due to limiting levels of malonyl-CoA and therefore indicated that
acetyl-CoA carboxylase (ACCase) was centrally involved in the feedback
regulation of fatty acid synthesis. Immuno-blot analysis showed that ACCase
protein levels did not change during the feedback inhibition, suggesting that the
feedback inhibition of fatty acid synthesis in plant cells may occur through
biochemical or post-translational modification of ACCase and possibly other
fatty acid synthesis related enzymes.

 

1The material in this chapter was published previously [Shintani DK, Ohlrogge
JB (1995) The Plant Journal 7: 577-587.

89

Introduction

Although the biochemistry of plant fatty acid synthesis has been
extensively studied, comparatively little is known about the regulation of this
pathway (Ohlrogge et al.,l993). It appears that plant fatty acid synthesis is
regulated, at least in part, by the demand for fatty acids by microsomal and
plastidial enzymes involved in glycerolipid synthesis. This hypothesis is
supported by the observations that 1) non-esterified fatty acids do not
accumulate in plant cells; 2) the lipid to protein ratios of particular membranes
are stringently maintained; and 3) high rates of fatty acid synthesis are only
observed in tissues where high rates of glycerolipid production are occurring
such as in expanding leaves or developing oil seeds. The high degree of
regulation imposed on this pathway not only helps to economize cellular energy
expenditures, but also prevents the accumulation of free fatty acids and acyl-
CoAs, whose detergent nature would be detrimental to the cell.

Feedback regulation of fatty acid synthesis could provide one potential
mechanism that would coordinate the supply of fatty acids with this cellular
demand. Feedback regulation of fatty acid synthesis has been reported in
animal, bacterial, and fungal systems. The accumulation of fatty acid synthase
(FAS) products, in the form of long chain acyl-CoAs, was reported to cause the
inhibition of the animal (Goodridge, 1985) and yeast (Kamiryo & Numa, 1973)
acetyl-CoA carboxylase (ACCase), the key regulatory enzyme of fatty acid
synthesis in these organisms. Furthermore, the addition of exogenous fatty acids
to bacterial cultures of Lactobacillus caused a reduction in de novo fatty acid
synthesis (Henderson & McNeil, 1966).

The study of feedback inhibition in plants has been hindered by the

inability of plants to efficiently assimilate exogenous fatty acids. However,

90

91

Terzaghi has shown that fatty acids in the form of Tween-fatty acid esters can
be rapidly assimilated into the membranes of soybean suspension cells
(T erzaghi, 1986a and b). When Tween-fatty acid esters were added to soybean
culture medium, as much as 50% of the membrane lipids were derived from
exogenous fatty acids (Terzaghi, 1986a). Labeling studies indicated that the
Tween-fatty acid esters were imported into the cell where they were hydrolyzed
to yield free fatty acids. When Tween-fatty acid esters of oleic or stearic acid
were added to the culture medium of soybean suspension cultures, l4C acetate
incorporation into the fatty acids decreased relative to untreated cells (Terzaghi,
1986b). This observation provided preliminary evidence that fatty acid synthesis
could be regulated by feedback inhibition in plants.

In the present study, the Tween-fatty acid feeding system was adapted for
use in tobacco suspension cell cultures. The effect of exogenous fatty acids on
individual acyl-acyl carrier protein (ACP) intermediates of plant fatty acid
synthesis was investigated using a combination of l“C-acetate labelling and
immuno-blot analysis with antibodies to ACP. The information obtained from
these studies has been used to deduce possible enzymatic site(s) involved in the
feedback regulation of plant fatty acid synthesis. The results presented here
indicated that biochemical feedback regulation of FAS is responsible for the
decreased rate of acetate incorporation into fatty acids of Tween-18:1 treated
tobacco suspension cells, and that acetyl-CoA carboxylase is the most likely site
of action of the feedback regulation.

Materials and Methods
Cell culture and medium.

Tobacco (Nicotiana tabacum L. cv bright yellow 2) cells were grown

92
shaking at 28°C in 50 ml of liquid medium composed of Murashige and Skoog
basal salts (Gibco-BRL, Grand Island, NY), 3% sucrose, 2.5 mM Mes/KOH pH
5.7, 1 mg/ml thiamine, 1 mg/ml myo-inositol, 1 uM 2,4-D. Cultures were
subcultured weekly with a 5% (v/v) inoculum from a seven day old culture.

Determination of the effect of Tween-18:0 and Tween-18:1 on cell growth.

Tween-18:0 and Tween-18:1 solutions were prepared based on the
determination of Terzaghi (1986) that commercial Tween 80 is approximately
20% (w/w) oleic acid. Therefore a 90 mM Tween-18:1 stock was prepared by
diluting 6.4 gm of tissue culture grade Tween 80 (Sigma, St. Louis, MO) in 50
ml of water. A 90 mM Tween-18:0 stock solution was prepared by diluting 7 .1
gm of Tissue culture grade Tween 60 (Sigma, St. Louis, MO) in 50 ml of water.
50 ml tobacco cell cultures were grown. in Tween-18:0 and Tween-18:1
concentrations ranging from 225 uM to 900 uM and 450 11M to 900 uM
respectively for 11 days. The packed cell volume was determined daily afier
pelleting the cells at 1000 rpm for 5 minutes. The medium was aspirated away
and the cells were resuspended in fresh medium and fresh Tween-18:0 and 18:1

was added to maintain the initial concentration.

Acetate labelling.

Tobacco cells were used four days after subculturing. To maintain
homogeneity between treatments, two separate 50 ml cultures were combined,
mixed and then realiquoted. At various times afier the addition of Tween-18:1
to a final concentration of 900 uM to one of the two cultures, 2.5 ml of cells
were sampled and immediately added to a tube containing 25 uCi of [1-“C]
acetate (47 - 54 Ci/mole) (New England Nuclear, Boston, MA). The cells were

93
then incubated for 10 minutes shaking at 28°C after which the labelling was
quenched by the addition of TCA to a final concentration of 5%. The TCA
treated cells were then immediately frozen in liquid nitrogen and stored at -
80°C. The TCA treated cells were thawed on ice and homogenized for
approximately 30 seconds using a Brinkmann polytron at the highest setting.
The homogenates were then centrifuged at 10,000 g for 20 minutes at 4°C. The
resulting supernatant was aspirated away and the pellet was washed in 1% TCA.
The washed cell pellet was then resuspended in 2 m1 of 50 mM Hepes pH 7.8,
10 mM N-ethylmaleimide (NEM). Protein determination of this homogenate
was done as described by Lowry et al., (1951). The homogenate was then
centrifuged at 10,000 g for 20 minutes at 4°C and the resulting pellet and

supernatant were saved for lipid and acyl-ACP analysis respectively.

Lipid analysis.

Total lipid was extracted from the cell pellet remaining after resuspending
the TCA precipitate in 50 mM Hepes pH 7.8, 10 mM NEM by the method
described by Bligh and Dyer (1959). Methyl esters were derivatized from the
total lipid extract as described by Morrison and Smith (1964) using boron
trichloride. The total lipid extract was dried under N2 and resuspended in 300
pl of toluene. 1 ml of 10% (w/v) boron trichloride in methanol (Alltech) was
added and the mixture was incubated at 90°C for 1 hour. The mixture was
allowed to cool and 1 ml of water was added. Methyl esters were then extracted
three times with hexane.

To determine the amount of [l-l‘C] acetate incorporated into the fatty acid
methyl ester (FAME) fi'action, the total methyl ester mixture was fractionated on
silica 60 A TLC plates (Whatmann) developed in 60:40:1 (vzvzv) hexanezdiethyl

94
etherzacetic acid. The FAME fraction was identified by co-migration with 1“C
labelled FAME standards and then scraped from the TLC plate. The amount of
1“C-label incorporated into F AMEs was determined by liquid scintillation
counting.

The amount of label incorporated into individual fatty acid species was
determined by first fractionating the total methyl ester fraction by argentation
TLC. Argentation plates were prepared by first saturating silica 60 A TLC
plates (Whatmarm) with 15% AgNO3 in acetonitrile and then drying them
overnight. Saturated F AMEs and individual unsaturated FAMEs were
fractionated in three developments of the argentation TLC plates in toluene at -
20*C. Unsaturated FAME species were scraped from the TLC plates and
counted by liquid scintillation counting. Saturated FAMEs were extracted three
times from the argentation plate scrapings with 2:1 (vzv) hexanezdiethyl ether and
finther fractionated on KC18 reverse phase TLC plates (Whatmann) developed
in 65 23505 (vzvzv) acetonitrilezmethanolzwater. Individual saturated FAME
species were scraped from the TLC plates and counted by liquid scintillation

counting.

Analysis of l‘C-labelled acyl-ACPs.

The supernatant derived from the resupended TCA precipitate was
applied to a 0.5 ml DEAE cellulose column (Sigma) equilibrated with 10 mM
Mes pH 6.1. The column was then washed with 6 column volumes of 10 mM
Mes pH 6.1. The acyl-ACPs were eluted from the column with 3 column
volumes of 0.5 M LiCl in 10 mM Mes pH 6.1. The eluted acyl-ACPs were
concentrated by precipitation in 5% TCA. The acyl-ACPs were then
resuspended in 500 pl of 50 mM Hepes pH 7 .8, 10 mM NEM. The acyl-ACPs

95
were fractionated on a l M urea 15% native acrylamide gel and blotted onto
nitrocellulose filters as described by Post-Beittenmiller et al. (1991). The filters
were then wrapped in 1.5 pm mylar and used to expose phosphor-image screens
(Molecular Dynamics) for four days. 14C-labelled acyl-ACPs were identified by
comparing their migration to 1“C labeled acyl-ACP standards synthesized in vitro
using E. coli acyl-ACP synthetase (Kuo & Ohlrogge, 1984). The labeled acyl-
ACPs were then cut from the nitrocellulose filters. The acyl-ACPs were
derivatized to FAMEs by dissolving the nitrocellulose pieces in 1.5 ml of 0.5 M
sodium methoxide in methanol and incubated at room temperature for 1 hour.
The derivatization was stopped by adding 1.5 ml of 125 mM H,SO4 and the
FAMEs were then extracted three times with 3 ml of hexane. Individual F ANHE
species derived from the acyl-ACP fraction were fractionated and counted as

described above.

Analysis of ACP-SH, acetyl-ACP, and malonyl-ACP species.

Tobacco ACPs were isolated from 50 ml of seven day old tobacco
suspension cell cultures as described by Shintani and Ohlrogge (1994) up to the
DEAE chromatography step. Tobacco ACPs were deacylated by incubation at
65°C for 1 hour in 100 mM Tris-HCl, pH 9.0, 200 mM DTT. Malonyl-ACP
standards were prepared from the deacylated tobacco ACPs using malonyl
transferase isolated from E. coli as described by Stapleton and J aworski (1983).
Acetyl-ACP standards were prepared chemically by reacting the deacylated
tobacco ACPs with acetyl-irnidazole as described by Cronan and Klages (1981).
Tobacco acetyl-ACP, malonyl-ACP, and free ACP species were fractionated by
15% native PAGE as described by Post-Beittenmiller et al. (1991). The acetyl-
ACP, malonyl-ACP, and free ACP fractions from control and Tween-18:1

96
treated tobacco cells were isolated as described above for the radiolabelled acyl-
ACPs.

Immuno-blot Analysis of FAS Enzymes

5 ml aliquots were taken at various times after the addition of 900 uM
Tween-18:1 to one of two 4 day old tobacco cell cultures. The cells were
pelleted and frozen in liquid N2 and stored at -70°C. The cells were thawed on
ice in 50 mM Hepes pH 7 .5, 1 mM PMSF, 1 mM benzamidine, 1 ug/ml
leupeptin and homogenized using a Brinkrnann polytron at the highest setting for
1 minute. The homogenates were clarified by centrifuging at 10,000 g for 20
minutes at 4°C. 100 ug of protein fi'om the supernatant of each sample was run
on 10% SDS PAGE gels (Alternatively, the cytosolic ACCase which was run on
a 6% SDS PAGE gel). The gels were blotted onto nitrocellulose. Each filter
was blocked with 5% low fat powdered milk in 10 mM Tris pH 8.0, 150 mM
NaCl, 0.5% Tween 20 (TEST) for 1 hour at room temperature. The cytosolic
ACCase and the BCCP subunit of the plastid ACCase were identified using
1/500 dilution of commercial alkaline phosphatase conjugated antibody to biotin
(Sigma, St Louis, MO). The biotin carboxylase subunit of the plastid ACCase
was identified using a 1/500 dilution of an antibody against the caster bean

endosperm biotin carboxylase (Keith Roesler, personal communication).

Results and Discussion
Development of Tween feeding system for tobacco suspension cells
The Tween feeding system developed by Terzaghi (1986a) was adapted
for use with tobacco cell cultures. Initial studies were conducted to determine

if tobacco cells would respond as soybean cells did to the addition of oleoyl-

97

Tween (Tween-18:1) and stearoyl-Tween (Tween-18:0). Tobacco cells were
grown in the presence of various concentrations of either Tween-18:0 or Tween-
18:1 to determine the maximal concentration of each that could be tolerated
without effecting cell growth. In these experiments, tobacco cells were able to
grow at rates essentially the same as control cells in 900 pM Tween-18:1 (Fig.
4-lb) and were later found to tolerate concentrations as high as 1800 uM (data
not shown). This high tolerance to Tween-18:1 was similar to results reported
for soybean cells (Terzaghi, 1986a). However, while soybean cells were able
to tolerate relatively high concentrations of Tween-1 8:0 (i.e. 1000 uM), tobacco
cell growth was inhibited when Tween-18:0 was added at concentrations as low
as 225 uM (Fig. 4-1a).

The addition of either Tween-18:0 or Tween-18:1 caused significant
decreases in acetate incorporation into total lipid (Fig. 4-1c). While the addition
of Tween-18:0 caused a 2 fold decrease in acetate incorporation, the Tween-
l8:1 treatment proved to be a more potent inhibitor of lipid synthesis, causing
a 3 - 5 fold decrease in label accumulation (Fig. 4-lc). Since Tween-18:1 was
less toxic to cell growth and more effective in inhibiting acetate incorporation
into total lipid relative to Tween-18:0, all subsequent studies focused on the
effect of Tween-18:1 on tobacco cell fatty acid synthesis.

Tween-18:1 inhibits [1-“C] acetate incorporation into fatty acids.

The effect of Tween-18:1 on the rate of [l-“C] acetate incorporation into
fatty acids was both rapid and dramatic. Exponentially growing tobacco cell
cultures treated with 900 uM Tween-18:1 showed a 3 to 5 fold decrease in the
rate of [l-“C] acetate incorporation into fatty acids relative to control cells (Fig.

4-2a). This decrease in label accumulation was detected as early as 15 minutes

98

Fig. 4-1a,b,c Effect of Tween-18:0 and Tween-18:1 on tobacco cell growth
and acetate incorporation into total lipid fraction. a. Cell growth measured
in cultures treated with 0 uM (closed circle), 225 uM (open triangle), 450 uM
(open circle), or 900 uM (open square) Tween-18:0 b. Cell growth measured
in cultures treated with 0 uM (closed circle), 450 uM (open triangle), 675 uM
(open circle), and 800 uM (open square) Tween-18:1. c. [1-“C] acetate
incorporation into total lipid of untreated cells (closed circle), cells treated with
225 M Tween-18:0 (open triangle) and cells treated with 900 uM Tween-18:1

(open squares).

peeked cell volume (mL) packed cell volume (mL)

(1 bounced.)

DPM I mL packed cells

99

 

“ h

+ control
'0' 225 "I 10:0
'0- 450 all 10:0

    
 

'4 000 u“ 10:0

 

 

 

 

 

10

+control

f." 450 u“ 18:1

   
 

"0' 075 p“ 18:1
"'.” 000 u“ 10:1

 

 

 

 

150

120

days

 

+eontrot

    

’+ 225 pM1I:0
' O 000 uH1I:1

  

 

 

 

100

Fig. 4-2a,b The effect of Tween-18:1 on the rate of [l-“C] acetate
incorporation into fatty acids of tobacco cells. a. Effect of Tween-18:1 on the
rate of [l-l‘C] acetate incorporation into the total fatty acid fraction of rmtreated
cells (closed triangles) and cells treated with 900 uM Tween-18:1 (open circles).
2.5 ml of cells were incubated for 5 minutes in 25 uCi of [l-“C] acetate.
Labelling was stopped by adding TCA to 5% (v/v). The total lipid was extracted
from the TCA pellet and fatty acids were derivatized with boron trichloride.
Fatty acid methyl-esters (FAMEs) were purified by TLC after which [l-l‘C]
acetate incorporation was determined by scintillation counting. b. Ratio of de
novo synthesized 16 carbon fatty acids to 18 carbon fatty acids in untreated cells
(closed triangles) and Tween-18:1 treated cells (open circles). Individual
FAMEs species were fractionated first by argentation TLC and then by reverse
phase TLC. Each FAME species was scraped from the TLC plates and [1-“C]

acetate incorporation was determined by scintillation counting.

:etate
)n the
eated
cles).
state
acted
ride.
.Hq
lfdi

:ells

9’

101

 

 

 

 

 

 

 

30
+ control -9- Tween-18:1
.E
2
2
a. 20 ‘
a
E
‘~ \
.5 \
E 10 - \
.— °\
0 \
c P‘ _________ 0
0
0 1 2 3
b hours
to
I!
8 2.50
>, + control -9- Tween-18:1
"" o
3 2.00 - /
s x
.n /
3 1.50 - //
Q if”
'2 1.00 - [d/
=3 /
C 0.50 " {\\/‘
0
(D
.9 0.00 4
g 0 1 2 3

 

102

after Tween-18:1 addition (data not shown) and continued to decrease for one
hour (Fig. 4-2a), after which a minimal rate of fatty acid synthesis was
maintained for at least eight hours. In contrast, control cells showed an
increased rate of [1-”C] acetate incorporation into fatty acids after 1 hour and
then maintained this rate for several hours (F ig.4-2a). The increasing rate of
acetate incorporation into fatty acids in control cultures was consistent with the
observation that sycamore cells in log phase growth show an increase in both
fatty acid and phospholipid content (De Silva & Fowler, 1976). Although de
navo fatty acid synthesis was inhibited in Tween-18:1 treated cells, the amount
of oleate derived from the exogenously supplied Tween-18:1 was sufficient to
maintain membrane synthesis rates necessary for exponential growth.

Although the results of the [l-“C] acetate labelling experiments suggested
that the addition of an exogenous supply of lipid can lead to the feedback
inhibition of fatty acid synthesis, care was taken when interpreting these results
because the use of [1-“C] acetate in fatty acid labelling experiments could have
lead to erroneous conclusions due to the alteration of endogenous acetate pools
(Nurm et al., 1977). Several lines of evidence were examined to confirm that the
decrease in acetate incorporation into the fatty acids of Tween-18:1 treated cells
was due to the inhibition of the fatty acid synthesis and not due to alterations in
the endogenous acetate pool. First, the decrease in acetate incorporation into the
fatty acids of Tween-18:1 treated cells was observed over a 25 fold range of [1-
14C] acetate concentrations (i.e. 40 uM to 1 mM) (data not shown). Since the
Tween-18:1 induced inhibition of acetate incorporation was seen over a large
range of [1-“C] acetate concentrations, it was unlikely that the inhibition was
due to an alteration in endogenous acetate pool sizes. Secondly, the Tween-18:1
treatment did not inhibit [1-“C] acetate incorporation into sterols, indicating that

103

the inhibitory effect of Tween-1 8:1 was specific to fatty acid synthesis. In these
experiments, the rate of [l-I‘C] acetate incorporation into sterols was 0.52
uCi/gm fi'esh weight/minute for Tween-18:1 treated cells versus 0.43 uCi/gm
fresh weight/minute for control cells. If the Tween-18:1 treatment altered
endogenous acetate pools then the rate of radiolabel incorporated into sterols
would also be expected to have been reduced. Finally, as will be discussed later,
the Tween-18:1 induced decrease in radiolabel accumulation into fatty acids was
accompanied by an alteration in the profile of several acyl-ACP intermediates
that was consistent with decreased fatty acid synthesis. These cumulative
findings indicated that the addition of Tween-18:1 to tobacco cultures was
inhibiting acetate incorporation into fatty acids as the result of the feedback
regulation of fatty acid synthesis.

Tween-18:1 alters the products of fatty acid synthesis.

In addition to its inhibitory effect on acetate incorporation, Tween-18:1
also had an effect on fatty acid chain length and desaturation. Compared to
control cells, the Tween-18:1 treated cells accumulated significantly less label
into the total 18 carbon fatty acid fraction (i.e. stearic acid (C1820), oleic acid
(C18:1), linoleic acid (C182), and linolenic acid (C18:3)) relative to the amount
incorporated into palmitic acid (C16:0) (Fig. 2b). The addition of Tween-18:1
resulted in a four to five fold increase in the ratio of labelled 16 carbon fatty
acids relative to labelled 18 carbon fatty acids, whereas this ratio remained
constant in control cells (Fig. 4-2b). Because linolenic acid was the major fatty
acid found in the membranes of tobacco suspension cells (data not shown), the
decrease in 18 carbon fatty acid synthesis in Tween-18:1 treated cells also had
a major effect on the degree of fatty acid desaturation. Similar results were

104

reported for soybean cells where the ratio of saturated to unsaturated fatty acids
was found to increase as a result of Tween-18:1 feedings (Terzaghi, 1986b).

The results shown in Fig. 2b suggest that Tween-18:1 treated tobacco
cells were able to detect changes in the degree of membrane fatty acid
desaturation and then respond by modulating not only the quantity, but also the
quality of de novo synthesized fatty acids. Although the total rate of fatty acid
synthesis was decreased in Tween-18:1 treated cells, the remaining FAS activity
was redirected toward the production of predominantly palmitic acid. The
majority of the oleate, linoleate and linolenate of treated cells were derived from
the exogenously supplied Tween-1 8: 1, and therefore the cells were not required
to synthesize oleate via de novo fatty acid synthesis. Presumably the cells adapt
the products of de novo fatty acid synthesis to maintain some level of saturated
fatty acid synthesis while attenuating the level of unsaturated fatty acid synthesis
in Tween-18:1 treated cells. We considered that the Tween 18:1 treatment
caused a decrease in the activity of the enzyme that catalyzes the elongation of
16 carbon fatty acids to 18 carbon fatty acids, B-keto acyl-ACP synthase H
(KAS 11). However, as will be discussed later, the acyl-ACP profile of the

Tween-18:1 treated cells was inconsistent with this possibility.

Analysis of tobacco cell acyl-ACP intermediates.

The steady state pool sizes of the acyl-ACP intermediates of fatty acid
synthesis reflect the balance between their rate of synthesis and rate of turnover
(Soll & Roughan, 1988; Jaworski et al., 1993). Alterations in these pools in
response to changes in the rates of fatty acid synthesis have been used to identify
acetyl-CoA carboxylase (ACCase) as a regulatory enzyme involved in light/dark
control of fatty acid synthesis (Post-Beittenmiller et al., 1991, 1992b). In the

105
present study, the relative pool sizes of several acyl-ACP intermediates were
examined in control and Tween-18:1 treated cells in order to deduce possible

site(s) of regulation involved in the feedback inhibition of fatty acid synthesis.

Tween-18:1 alters [1-“C] acetate incorporation into long chain acyl-ACPs.

The addition of Tween-18:1 caused significant changes in the amount of
[l-“C] acetate incorporated into the C1620 and Cl 8:0 acyl-ACP pools. While
the amount of label associated with the C16:0 and C1810 acyl-ACP pools in
control cells increased over time, treated cells responded rapidly to the addition
of Tween-18:1 by incorporating less acetate into these acyl-ACP pools (Fig. 4-
3a and 4-3b). The decrease in acetate incorporation into the C1620 and C1820
acyl-ACP pools was detected as early as 30 minutes after the addition of Tween-
1821. The profile of [1-“C] acetate incorporation into the C1620 and C18:0 acyl-
ACP pools (Fig 4-3a and 4-3b) of both the control and treated cells closely
coincided with the profile of acetate incorporation into the fatty acid fractions of
these cells (Fig. 4-2a). No label was detected in the oleoyl-ACP fraction of
either the Tween-18:1 treated or control cells (data not shown). This result was
consistent with the observation that plant extracts contain very active long chain
acyl-ACP thioesterases that show a substrate preference for oleyl-ACP
(Ohlrogge et al., 1978).

Since the synthesis of the C1620 and C1820 acyl-ACPs was dependent on
condensation of a malonyl-ACP molecule with an acyl-ACP, the decreased
formation of these acyl-ACPs in Tween-18:1 treated cells was consistent with
either a decrease in available malonyl-ACP or the decrease in the rate of the
condensation reaction catalyzed by KAS H. Since an accumulation of 16:0-ACP
and medium chain acyl-ACPs was not observed in T ween-1 8:1 treated cells (Fig

0.40

1101] mg protelu

0.00 '

Fig. 4-3a,b. Effect of Tween-18:1 on [l-I‘C] acetate incorporation into long
chain acyl-ACPs of tobacco cells. a. [1-“C] acetate incorporated into 1620-
ACP fi’action of untreated cells (closed triangles) and cells treated with 900 uM
Tween-18:1 (open circles). - b. [l—l‘C] acetate incorporated into 18:0-ACP
fraction of untreated cells (closed triangles) and cells treated with 900 pM

Tween-18:1 (open circles). All values were derived from the mean of three

 

0.24 '

0.10 ’

 

 

 

hours

106

5"

nCl I ma proteln

0.17

0.14 '

0.10

0.07 ’

0.03 '

 

 

 

hours

individual experiments. Error bars represent standard error.

 

107
4-4a,b,c), it was unlikely that the decrease in Cl6:0 and C1820 acyl-ACPs pools
was the result of inhibition occurring at KAS II. It was more likely that the
decrease in long chain acyl-ACP synthesis in Tween-18:1 treated cells was due
to a decrease in available malonyl-ACP caused by the inhibition of ACCase.

The reduced [l-“C] acetate incorporation and the depletion of the long
chain acyl-ACP pools in Tween-18:1 treated plants indicated that although the
flux through fatty acid synthesis was reduced by the addition of Tween-1 8: 1, this
treatment did not substantially inhibit the turnover of these acyl-ACP pools. The
turnover of the C16:0 and C18:0 acyl-ACP pools is due to the presence of very
active long chain acyl-ACP thioesterases. These results suggest that the activity
of these thioesterases is not coupled to the feedback regulation of fatty acid
synthesis in tobacco cells.

The long chain acyl-ACP pools from developing seeds and light/dark
exposed chloroplasts responded similarly to changes in the rates of fatty acid
synthesis as did the feedback inhibited tobacco cells. Early in seed
development, when very active fatty acid synthesis is occurring, long chain acyl-
ACPs have been reported to accumulate to relatively high levels (Post-
Beittenmiller et al., 1992a). However, later as the rate of fatty acid synthesis
decreased, the long chain acyl-ACP pools became depleted (Post-Beittenmiller
et al., 1992a). Similarly, in spinach chloroplasts, long chain acyl-ACP levels
remained high in the light when rates of fatty acid synthesis were high and
decreased in the dark when fatty acid synthesis was inhibited ($011 and Roughan,
1982; Roughan and Nishida, 1990). These observations suggested that long
chain acyl-ACPs were probably not directly involved in the perception of the
feedback inhibition because these acyl-ACPs were reduced during the inhibition
of fatty acid synthesis. However, long chain acyl-ACPs have recently been

108

Fig. 4-4a,b,c. Effect of Tween-18:1 on [1-"C] acetate incorporation into
medium chain acyl-ACPs of tobacco cells. a. [l-“C] acetate incorporated
into-10:0-ACP fiaction of untreated cells (closed triangles) and cells treated with
900 uM Tween-18:1 (open circles). b. [1-“C] acetate incorporated into 12:0-
ACP fraction of untreated cells (closed triangles) and cells treated with 900 uM
Tween-18:1 (open circles). c. [l-”C] acetate incorporated into 14:0-ACP
fraction of untreated cells (closed triangles) and cells treated with 900 uM
Tween-18:1 (open circles). All values were derived from the mean of three

individual experiments. Error bars represent standard error.

109

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L 3 3
N . .
. z . A ~
f 2 8 ~
. _ m _ .
. 1 M A M a
. o a o
o . r p p 1 .
A‘- dl ” 5 . 1 ‘ 7 1 u l 1
o o. . 1w 2 9m .1 0 o
0 0 0 0 0 0. m M m «0. u M M
0 m 0. 0.
c
. .393 a... :0: .. £29:— uE :0: c a a
. .82 E \ .o:

a b C

ll Hill
\5

H

Li
\l

n .9
-1. 5‘

hours

1 10
suggested as a potential feedback inhibitor of bacterial fatty acid synthesis (J iang
and Cronan, 1994; Ohlrogge et al., manuscript submitted) and it is possible that
under other conditions this molecule may participate in the regulation of plant
fatty acid synthesis.

Medium chain acyl-ACP pools are not altered by Tween-18:1 treatment.

While the addition of Tween-1821 caused a decrease in acetate
incorporation into the long chain acyl-ACP pools, this treatment had no effect
on the medium chain acyl-ACPs. The amomt of [1-“C] acetate incorporated
into the C1020, C1220, and C14:0 acyl-ACP pools did not change for at least 3
hours in both Tween treated and control cells (Fig. 4-4a - 4-4c).

These results indicated that although the total flux through the fatty acid
synthetic pathway decreased upon the addition of Tween-18:1, the relationship
between the rate of synthesis and the rate of turnover of the medium chain acyl-
ACPs remained unchanged. Because both the synthesis and turnover of the
medium chain acyl-ACP intermediates is dependent on the condensation of the
elongating fatty acid chain with a malonyl-ACP molecule, the constant level of
medium chain acyl-ACP intermediates could be explained by either a decrease
in available malonyl-ACP or by inhibition occurring at the enzyme responsible
for the condensation reaction, B-keto acyl ACP synthase I (KAS I). If KAS I
were being inhibited in response to the addition of Tween-18:1, an increase in
the sizes of the short chain (i.e. 4 through 8 carbon) acyl-ACP and malonyl-ACP
pools should have been observed. Such an acyl-ACP profile was seen in spinach
chloroplast extracts treated with cemlinin, a potent inhibitor of KAS I (J aworski
et al., 1993). No such increase was detected in shorter chain acyl-ACP pools of

the Tween-18:1 treated tobacco cells. The simplest interpretation consistent

1 11
with the observation that the sizes of the medium chain acyl-ACP pools
remained constant in the presence and absence of Tween-18:1 is that the
decreased flux through the fatty acid synthesis pathway was due to a decrease
in malonyl-ACP corresponding to a decrease in malonyl-CoA as a result of the
feedback inhibition of ACCase.

The results in figure 4-4 also indicated that the decreased rate of fatty acid
synthesis was not due to inhibition occurring solely at the initial condensation
reaction catalyzed by [3 keto-acyl-ACP synthase HI (KAS III). If inhibition of
fatty acid synthesis was occurring at KAS 111, one would expect that the
malonyl-ACP that was no longer being used by KAS IH, would be utilized in
other condensation reactions such as those catalyzed by KAS I. If no newly
initiated acyl-ACP precursors were synthesized, then the medium chain acyl-
ACPs would be rapidly depleted as they were elongated. Because the medium
chain acyl-ACP pools were not depleted upon the addition of Tween-18:1, it was
not likely that the inhibition of KAS HI was solely responsible for the decreased
rate of fatty acid synthesis.

Tween-18:1 altered the sizes of the acetyl-ACP and free ACP pools.

The tobacco cell acetyl-ACP, malonyl-ACP and free ACP- SH pools were
analyzed by native PAGE immuno-blot analysis as described by Post-
Beittenmiller et al. (1991). Because tobacco cells contained two major ACP
isoforms, it was first necessary to identify individual free, acetyl- and malonyl-
ACP species in cell extracts by their co-migration with in vitro prepared
standards. In the case of both isoforms, the malonyl-ACP and acetyl-ACP
species were clearly resolved from the free ACP species (Fig. 4-5a).

Immuno-blot analysis indicated that the addition of Tween-18:1 caused

112

(a)

. l 2 3

. malonyl-ACPI .-

freeACPl____.—. 7 I IACPII
l ocet I-A Pl _ ’7 -___mo ony-

T y C - _freeACPll

l "'"’ —ocetyl-ACPII

(bl
can‘trols Tween-18:1
Hours 0 0.5 0 0.513 5

ACPI , . ...-. ACPI
C2:O-ACP 3877' '7' 7" 7" '9 1'.... 02:0-ACPI

ACP" ACP"
C2:0-ACPII 9.9“ .Q =3 .BtCZm-ACPII

 

Fig. 4-5a, b. Analysis of the free-, acetyl-, and malonyl-ACPs pools from
control and Tween-18:1 treated tobacco cells. a) ACP standards were
prepared in vitro from deacylated ACPs isolated from tobacco cells. Lane 1,
free ACP standards; lane 2, acetyl-ACP standards; and lane 3, malonyl-ACP
standards. b) Acyl-ACPs were extracted using 5% TCA from treated and
rmtreated cells at various times after the addition of Tween-18:1. TCA
precipitated proteins were resuspended in 50 mM HEPES pH 7. 8, 10 mM NEM.
Approximately 100 u g of TCA precipitated protein was loaded onto each gel.
All samples were run on 15% native PAGE gels and blotted to nitrocellulose.
ACP species were then identified using antibodies made to spinach ACP.

1 13
changes in the free and acetyl-ACP pools. As the rate of [1-‘4C] acetate
incorporation into fatty acids decreased in Tween-1821 treated cells (Fig. 4-2a),
a coincident decrease in fi'ee ACP level was observed (Fig. 4-5b). In control
cells where fatty acid synthesis was not inhibited (Fig. 4-2a), the free ACP levels
appeared to remain constant over the course of these experiments (Fig. 4-5b).

The acetyl-ACP levels had an inverse relationship to the rates of [1-“C]
acetate incorporation into the fatty acids of tobacco cells. The acetyl-ACP levels
increased after the addition of Tween-1 821 (Fig. 4-5b) and continued to increase
to a maximum level at 3 hours which coincided with the minimum rate of acetate
incorporation into the total fatty acid fraction (Fig. 4-2a). This increase of
acetyl-ACP levels was apparent within 30 minutes after the addition of Tween-
1821. Conversely, in control cells where label incorporation into fatty acids was
not inhibited (Fig. 4-2a), acetyl-ACP levels remained low (Fig. 4-5b). However,
as the rate of fatty acid synthesis began to plateau between 1 and 3 hours (Fig.
4-2a), a slight increase in acetyl-ACP levels in the control cells could be
detected (Fig. 4-5b).

Although the free ACPs and acetyl-ACPs were easily resolved by
immuno-blot analysis, malonyl-ACP levels were below the level of detection.
Several attempts to increase the sensitivity of the analysis, including specific
radiolabelling of the malonyl-ACPs with 14C bicarbonate, failed, suggesting that
either cellular levels of malonyl-ACP were extremely low or that the tobacco cell
malonyl-ACP was particularly labile. The former appeared to be more likely
since in vitro synthesized tobacco cell malonyl-ACP standard proved to be
stable during electrophoresis and could be stored at -20°C for several months
with no detectable breakdown.

The changes observed in the fiee and acetyl-ACP pools of Tween-18:1

114
treated tobacco cells are similar to free and acetyl-ACP pools of spinach
chloroplast during dark inhibition of FAS where it was concluded that ACCase
was being down regulated (Post-Beittenmiller et al., 1991, 1992b). In both
spinach leaf and tobacco suspension cultures, as the rate of F AS decreased, the
levels of acetyl-ACPs increased and the levels of free ACPs decreased.
J aworski et al. (1993) explained that the increase in acetyl-ACP was in response
to the inhibition of malonyl-CoA synthesis by ACCase and the uninhibited
acetyl-transferase activity of enzyme [3 keto-acyl-ACP synthase IH (KAS HI).

The feedback regulation of fatty acid synthesis was not caused by the
attenuation of ACCase levels.

Even though the acyl-ACP analysis implicated ACCase as an important
regulatory enzyme in fatty acid synthesis, the mechanism by which the regulation
occurred could not be deduced from those experiments. If the feedback control
mechanisms regulating fatty acid synthesis in plant cells were similar to yeast
and animals, then two different modes of regulation could have been responsible
for the feedback inhibition of fatty acid synthesis. First, in yeast and animals,
"short term" regulation of fatty acid synthesis has been shown to occur through
allosteric and covalent modifications of ACCase (Kim, 1983; Nikawa et al.,
1979). This type of regulation provides stringent control of fatty acid synthesis
and allows the organism to rapidly respond to momentary changes in cellular
demands for fatty acid synthesis. In addition to this " short term" control,
animals, yeast and bacteria response to chronic changes in the cellular demand
for fatty acid synthesis by altering the levels of ACCase protein and other FAS
related enzymes. When exogenous fatty acids were added to the culture medium
of yeast, rat hepatocytes, and human skin fibroblast, a significant decrease in the

1 15

levels of ACCase was observed (Kamiryo and Numa, 1973; Kitajirna et al.,
1975; Jacobs et al., 1973). Additionally, Li and Cronan (1993) have reported
that levels of E. coli ACCase directly correlated with the rate of cell growth.
The activity of other fatty acid synthetic enzymes have also been shown to be
regulated at the level of gene expression. For example, when rats were fed a diet
rich in polyunsaturated fats, the levels of FAS mRNA decreased (Clarke et al.,
1990). In Lactabacillus plantarum, the levels of ACP were reduced to 80% of
control levels when these bacteria were fed exogenous fatty acid (Sabaitis JE
and Powell GL, 197 6). Furthennore, the level of yeast stearoyl-CoA desaturase
mRNA was also reported to decrease when exogenous fatty acids were added
to the grth medium (Bossie and Martin, 1989).

Due to the immediacy of the response to the Tween-18:1 treatment, the
initial regulation of fatty acid synthesis in tobacco cells probably occurred at the
enzymatic level. However, because the inhibition of fatty acid synthesis was
observed over several hours it was also possible that the cells would adapt such
that the activities of ACCase and fatty acid synthetic enzymes were decreased
due to lower enzyme levels. In order to determine if the decreased ACCase was
due to the attenuation of enzyme levels, the levels of the biotin carboxylase and
biotin carboxyl carrier protein (BCCP) subunits of the plastid ACCase were
determined by immrmo-blot analysis. Antibodies made to the castor seed biotin
carboxylase, and biotin were used to detect the levels of the biotin carboxylase
and BCCP subunits of the plastidial ACCase in extracts prepared from control
and Tween-18:1 treated tobacco cells. Surprisingly, the levels of both subunits
did not change over a 32 hour period in either the control or Tween-18:1 treated
cells (Fig. 4-6). The levels of the 200 kD cytosolic form of ACCase were also
Lmafi'ected by the addition of Tween-18:1 (Fig. 4-6). Furthermore, the levels of

116

.ommoo< 0:885 08 Ea ommOU< 363a 05 no “.833 ARE 8: Son 88% 8 £85 8 888 8638mm 3
838 033888 885 88% 8 0800.4 cum—WE 05 mo H.853 own—>888 885 3% H88 05 8 838 8%qu
556 5; “Cocoa 825 88G 1:63:35 .88E 88338:: 8 @885 can flow m©<m-QO «8&2 so @8838
803 fix T503... 28 80 mo 85QO was 3585 8: E gem £3 0338 Set “888be £805 :38 we we

o3 i=8 68.3.5 fiat—83H was .9559 SP: @8838 8255 5 8.508.. we mam—an: 83-8585 6.9 .wmm

'rl-lrrlll'l'l I‘Ill‘ QwOUU<O=OmO+6

0809.. 88: .6 E38 .68 W

_

’0’! -4. 2'2! 00-101000!‘

9.0 U< DECK v.0 «7503
09.00 czoa

mam—onpmdo manpon_m.ooeaoz
.6 7:83.. 68:00

 

 

l 17

two other enzymes involved in fatty acid synthesis, specifically KAS III and
stearoyl-ACP desaturase, were also unaffected by the Tween-18:1 treatment
(data not shown). These results indicate that the decreased flux through the fatty
acid synthetic pathway was not due to a general decrease in FAS enzyme levels,
but was more likely due to the inhibition of enzyme activity arising through
allosteric or covalent post-translational modifications. However, it can not be
ruled out that expression of some specific FAS protein not measured in this
study was altered by the Tween-18:1 treatment. Individual filters were probed
with either antibodies made to the castor seed biotin carboxylase subunit of the
plastid ACCase to detect biotin carboxylase levels or antibodies made to biotin
to detect both the BCCP subunit of the plastid ACCase and the cytoplasmic
ACCase.

Conclusions

The analysis of the tobacco cell acyl-ACP pools allowed us to develop a
model depicting how fatty acid synthesis was being regulated by feedback
inhibition. The cumulative data obtained from these experiments indicated that
ACCase was being down regulated in response to the addition of exogenous
fatty acids. The observed acyl-ACP profiles of the Tween-18:1 treated cells
were consistent with what was observed and predicted to occur under conditions
where malonyl-CoA would be limiting. Although these studies could not rule
out the possibility that multiple components of fatty acid synthase were being
coordinately down regulated with ACCase, our data clearly showed that none
of the condensing enzymes were being individually inhibited by the feedback
regulation. Furthermore, the acyl-ACP analysis indicated that the long chain
acyl-ACPs thioesterases were not inhibited during the feedback inhibition of

 

 

1 18
fatty acid synthesis resulting in the rapid depletion of the long chain acyl-ACP
pools. This observation suggest that the long chain acyl-ACPs may not act be
the molecules involved in the perception of the feedback inhibition of fatty acid
synthesis. The results of the immuno-blot experiments suggested that the
feedback inhibition of fatty acid synthesis was not due to an attenuation of
ACCase levels, but was probably due to covalent or allosteric modifications of

this and possibly other fatty acid synthetic enzymes.

Acknowledgements

I would also like to acknowledge Edgar Cahoon for his help with the lipid
analysis and for helping to develop the long and medium chain acyl-ACP
analysis. I would like to thank Drs. Dusty Post-Beittenmiller, Keith Roesler, and
Mike Walsh for their critical discussion of this manuscript. We also thank Keith
Roesler for the biotin carboxylase antibodies, Jan J aworski and Hee Young Tai
for the KAS III antibodies and John Shanan for the stearoyl-ACP desaturase
antibodies. Acknowledgement is also made to the Michigan Agricultural
Experiment Station for its support of this work.

Literature Cited

Bli EG and er. WJ (1959) A rapid method of total lipid extraction and
p cation. Can Biochem Physrol 3 : 911-917

Bessie MA, Martin CE 1989 Nutritional re ation of east delta-9 fa acid
desaturase activity. J Baéterio 171: 6409-65”; y tty

Clarke SD Armstron MK, Jum DB 1990 Die 01 aturated fats
1115116111261 5 silgapress rat Inger fatty aci synth(ase arid S 11:42:21!le content. J Nutr

Cronan JE, Kla es AL (19812“?hemical synthesis of. 31%] thioesters of ac l
camer protein w1 natlve struc e. Proc Natl Acad SCI A 78: 5440-5 .

De Silva NS, Fowler MW (1976))h The lipid com osition of Acer
pseudoplatanus cells grown in culture. ytochennstry 15: 735-1740.

Goodrid e AG (1985) Fatty acidSynthase in Eukaryotes. 1]: DE Vance, JE
Vance, e Blochermstry of LlpldS and Membranes. Benjamin/Cummings
Pubhshmg Company Inc.

Henderson T0, McNeil ,JJ (1966) The control of fatty acid 5 thesis in
Lactobac111us plantarum. Biochem Biophys Res Comm 25: 662-66 .

Jacobs RA, Sly WS, Ma'erus PW 1973)1 The regulation of fatty acid
biosynthe51s mhuman skin 1 roblast. J 101 C em 248: 1268-1276.

Jaworski JG, Post-Beittenmiller D, Ohlrogge JB (1993) Acetyl-acyl carn'er
roteln 18 not a major mtermedlate 1n fatty ac1d blosynthems 1n spinach. Eur J
Biochem 213: 981-987.

Jiang P, Cronan JE (1 994) Inhibition of fatty acid synthesis in Escherichia coli
in the absence ofpho7sghol§p1d s thesrs and release of inhrbltlon by tlnoesterase
actlon. J Bacteriol 1 : 2 14- 821.

Kamiryo T Numa S (1973) Reduction of the acetyl coenzyme A carboxylase
Sgnggnt of Saccharomyces cereVISIae by exogenous fatty ac1ds. FEBS Lett 38:

gun1 Eli 5983) Regulation of acetyl-CoA carboxylase. Curr Top Cell Regul

Kita'ima K, Tashiro S, Numa S (1975) Acetyl-CoA carboxylase in cultured
hepa ocy_tes: Effects of exo enous_ fatty ac1ds on the content, synthes13 and
degradation of the enzyme. ur J Blochem 54: 373-3 83.

Kuo TM, Ohlrogge JB (19842}Acylation of plant acyl carrier protein by acgl

211%] carner protem synthetase om E. coli. Arch Biochem Biophys 230: 11

119

120

Lowry OH, Rosebrou NJ, Farr AL, Randall RJ (1951 Protein
measurements With the fo phenol reagent. J Biol Chem 193: 265- 75.

Morrison WR, Smith .LM 1964) Preparation of fatty acid methyl esters and
didnsiethylacetals from lipids w1 boron trifluoride methanol. J Lipid Res 5: 600-

Nikawa J, Tanabe T, Ogiwara H, Shiba T, Numa S 1979) Inhibitory effects
fifetltoil Zah23213 250 8l-CoA analogues on rat hver acetyl- oA carboxylase. F EBS

Nunn WD, Kelly DL, Stumfall MY (1977) Regulation of fatty acid synthesis
during the cessation of hospholipid biosyn em in Escherichia coli. J.
Bacteriology 132: 526-53 . '

thro eJB Jaworski JG, Post-Beittenmiller D (19989e nova fatty acid
biosyngigeSis. n TS Moore ed, Lipid Metabolism in Plants. C Press Inc., Ann
Arbor, MI, USA pp 3-32.

Ohlrogge JB,. Shine WE, Stumxf PK 197 8 Fat Metabolism in hi erplants.
haracterization of Blant acyl- CP an acy -CoA hydrolases. Arc Biochem
Biophys 189: 306-3 1.

Post-Beittenmjller MA, Jaworski JG, Ohlrogge JB (1992a? Regulation of

lipid syntheSis in castor seeds: AnalySis of the in vivo_ac¥l-ac carrier protein

001. n SL MacKen21e and DC Talor eds. Seed oils or e future. AOCS
ess, Champaign, IL, USA, pp 44- .

Pmt-Bdflenmfller D, Roughan. G, Ohlro e JB (1992b) Regulation of plant
fatty ac1d biosynthesis: Anaclinis of acyl- 0A and acyl-acy carrier rotem
substrate pools in spinach an pea chloroplast. Plant PhySiol. 100: 923- 30.
Post-Beittenmiller MA, Jaworski JG, Ohlro¥e JB (1991) In vivo pools of
free and acylated acyl earner rotems in s mach: v1dence for s1tes of regulation
of fatty ac1d biosynthCSis. J 101 Chem 66: 1858-1865.

Roughan G, Nishida I. 1990) Concentrations of long chain acyl-acyl carrier
proteins during fa ac1 syntheSis by chloroplasts isolate from [pea (Pisum
sativum) safflower Carthamus linctons), and amaranthus (Amarant us lzvidus)

leaves. Arch Biochem Biophys 276: 38-46.

Sabitis JE Powell GL 1976) Acyl carrier rotein metabolism and re ation
agil‘aztty acid biosyntheSis y Lactobacillus p antarum. J Biol Chem 25 : 4706-

Shintani DK, Ohlrogge JR (199%) The characterization of a mitochondrial acyl
(1:31P? r10t162139180f0rm isolated om Arabidopsis thaliana. Plant PhySiology

Sol] J Ron an G 1982 Ac l-ac lcam'er rotein 00] sizes d ' -stead -
stgée lfgéty agild syntl(iesis y is)olate)cll spinachpchloroglast. FEBS Legers 14%:

121

Stapleton SR, Jaworski JG (1984) Characterization and purification of
malonyl-coenzyme A: [ac 1 carrier protein] transac lases from spinach and
Anabaena vanabilis. Bioc em Biophys Acta 794: 2 4-248.

Terzaghi WB (1986a) A system for manipulating the membrane fatty acid
compos1tion of soybean cell cultures b adding Tween-fatty acid esters to their
grth medium. lant PhySiol 82:7 -779.

Terzaghi WB 1(11986b) Metabolism of Tween-fatty acid esters by cultured
soybean cells. ant Physrol 82: 780-786.

Terzaghi WB, Roth J, Keim P Honeycutt RJ, Lark KG (1985 A somatic
genetic system for so an. In M Preelmg, ed, Plant Genetics. Alan Liss, Inc,
ew York, pp 771- 89.

Chapter 5

Antisense and over-expression of biotin carboxylase in tobacco leaves

Abstract

Several in vivo studies have implicated the plastid acetyl-CoA carboxylase
as an important regulatory enzyme in fatty acid synthesis. The plastid ACCase
is present as a heteromeric complex that is composed of four difl‘erent protein
subunits, specifically biotin carboxylase, biotin carboxyl carrier protein, and the
a- and B-subunits of the carboxyltransferase. In order to test our assumptions of
ACCase's role in regulating fatty acid synthesis and also to gain some insight into
the subunit organization of this heteromeric enzyme complex, biotin carboxylase
expression was altered in transgenic plants. Tobacco plants were transformed
with antisense— and over-expression tobacco biotin carboxylase constructs which
resulted in the generation of plants with biotin carboxylase levels ranging from
10% to 1000% of wild type levels. A comparison of leaf biotin carboxylase and
BCCP levels in plants showing elevated and decreased biotin carboxylase
expression, revealed that these two subunits of the plastid ACCase are not
maintained in a strict stoichiometric ratio. Tobacco plants containing elevated
and moderate decreases in leaf biotin carboxylase were indistinguishable from
wild type plants. However, one plant, with only 10% of wild type biotin
carboxylase levels, showed severely retarded growth when grown imder low
light conditions and significantly lower leaf fatty acid content than wild type
plants. Furthermore, the homozygous progeny of this plant exhibited aberrant

floral development.

122

 

Introduction

Fatty acid synthesis is central to the growth and development of most
organisms. Along with the central role that this pathway plays in the synthesis
of membrane lipid precursors, fatty acid synthesis is also necessary for the
synthesis of seed storage oils and waxes which are valuable agronomic
commodities.

Although much is known about the biochemistry involved in the synthesis
of fatty acids, little is miderstood about how this pathway is regulated. However
recently, several in vivo studies have implicated the plastid localized acetyl-CoA
carboxylase (ACCase) as an important regulatory enzyme in fatty acid synthesis.
First, careful measurements of acyl-acyl carrier protein (acyl-ACP) pools in light
and dark treated spinach leaves and chloroplasts indicated that the plastid
ACCase is involved in the light activation of fatty acid synthesis (Post-
Beittenmiller et al., 1991, 1992). It was also demonstrated through acyl-ACP
analysis that the plastid ACCase is involved in the product feedback inhibition
of fatty acid synthesis in tobacco suspension cultures (Shintani and Ohlrogge,
1995). Furthermore, using specific inhibitors of the maize plastid ACCase, Page
et al. (1994) demonstrated that ACCase exerts a high degree of flux control over
fatty acid synthesis in maize leaves. These findings agree with observations from
animal, fungal, and bacterial systems that implicate ACCase as an important
enzyme involved in the regulation of fatty acid synthesis in those organisms
(Goodridge, 1985; Kamiryo and Numa, 1973; Magnuson et al., 1993).

Due to its important role in regulating plant fatty acid synthesis, the plastid
ACCase has become the subject of intense interest. This interest has led to the
discovery that in dicot and non-gramineae monocot plants, the plastid ACCase

exists as a heteromeric enzyme complex composed of four independent

123

124
polypeptides corresponding to the biotin carboxyl carrier protein, the biotin
carboxylase, and the a- and [l-subunits of carboxyltransferase (Sasaki et
al.,l994; Alban et al., 1994; Shorrosh et al., 1995 and manuscript submitted;
Choi et al., 1995).

Recently, the genes coding for the biotin carboxylase, BCCP, and the a-
and B—subunits of carboxyltransferase have been cloned from plants (Sasaki et
al.,l994; Shorrosh et al., 1995 and in press; Choi et al., 1995). This achievement
has now made it possible to test assumptions about ACCase's regulatory role in
plant fatty acid synthesis. It is now possible to modify the level of the plastid
ACCase by altering expression of genes coding for one or more of the ACCase
subunits in transgenic plants. IfACCase has a strong regulatory influence on the
flux through fatty acid synthase, then changes in ACCase activity should
significantly alter rates of fatty acid synthesis.

We have previously reported the cloning of a tobacco cDNA clone
corresponding to the biotin carboxylase subunit of the plastid ACCase (Shorrosh
et al., 1995). In the present study, the biotin carboxylase cDNA was expressed
in the sense and antisense orientation behind a strong constitutive promoter in
transgenic tobacco plants. One goal of this work was to determine what effect
altering biotin carboxylase level would have on ACCase's ability to regulate the
flux through fatty acid synthesis. Additionally, we were interested in determining
how the expression of other subimits of the plastid ACCase would be effected
by altering levels of biotin carboxylase.

Materials and Methods

Construction of biotin carboxylase plant expression constructs and
transformation of these constructs into tobacco

The tobacco biotin carboxylase cDNA clone (Shorrosh et al., 1995) was
subcloned in the sense and antisense orientation as a PstI fragment in place of
the add intron I PstI fragment in the plant gene expression vector p1079. The
p1079 vector contained the promoter for the 355 transcript of the cauliflower
mosaic virus (CaMV) that was engineered to contain two copies of the CaMV
3 53 enhancer element and the nopoline synthase 3' terminator sequences. The
double CaMV 3SS promoter of p1079 was constructed by subcloning the HincH
EcoRV fragment corresponding to the CaMV 35S enhancer region of the CaMV
35$ promoter into the HincH site of the CaMV 35S promoter. The biotin
carboxylase sense and antisense expression constructs were subcloned as XbaI
fragments into the XbaI site of pBIN19 (Bevin, 1984). The resulting over- and
antisense-expression/Agrabacterium binary vector constructs were designated
as p0719S and p0719A respectively. These constructs were introduced into the
Agrabacterium tumefaciens strain LBA4404 (Hoekema et al. (1983) by direct
transformation as describe by An et al. (1988). Tobacco (Nicatiana tabacum)
leaf disc were transformed as described by Rogers et al. (1986).

Affinity purification of caster bean biotin carboxylase antibodies to
Escherichia coli (E. coli) expressed tobacco biotin carboxylase.

The DNA sequence from the tobacco biotin carboxylase cDNA clone
corresponding to the mature biotin carboxylase protein was amplified by
polymerase chain reaction (PCR) and subcloned into pETl 5b (N ovagen,
Madison, WI) as described by Shorrosh et al. (1995). The biotin carboxylase

125

126

was expressed in E. coli and purified as described by Shorrosh et al. (1995).

Approximately 4 mg of the purified E. coli-expressed biotin carboxylase
protein in 100 mM MES pH 5.8 was reacted with 1 ml of Afligel-IO (BioRad,
Richmond, CA) for 30 minutes at 4°C on a rocking platform, after which the gel
matrix was collected by centrifugation at 10,000 g for 10 minutes. The
supernatant was then reacted a second time with 1 ml of Affigel-IS (BioRad,
Richmond, CA) overnight at 4°C on a rocking platform. The reacted Afiigel-lO
and Affigel-l 5 were then combined and washed sequentially with 10 column
volumes of 10 mM Tris pH7.5, 10 column volumes of 100 mM citrate pH 2.5,
10 column volumes of 10 mM Tris pH 8.8, and 10 column volumes of TBST (10
mM Tris pH 8.0, 150 mM NaCl, 0.05% Tween 20, 0.02% sodium azide).

Ammonium sulfate was added to 5 0% saturation to 5 ml of the rabbit
antiserum against caster bean biotin carboxylase (Roesler et al., 1995). The
ammonium sulfate pellet was then resuspended in 2.5 ml of TBS (10 mM Tris
pH 8.0, 150 mM NaCl, 0.02% sodium azide). The solution was desalted on a
PD-10 desalting column (Pharmacia) equilibrated with TBS and bound to the
biotin carboxylase affinity column overnight at 4°C on a rocking platform. The
affinity column was washed with 10 column volumes of TBST and the antibody
was eluted with 10 column volumes of 100 mM citrate pH 2.5. The pH of eluted
fractions was neutralized by adding 1 M Tris pH 8.8.

Immuno-blot analysis of tobacco leaf extracts.

Tobacco leaves were immediately frozen in liquid nitrogen after being
harvested. Approximately 1 gm of leaf material was then groimd to a fine powder
in liquid nitrogen. The frozen ground tissue was then homogenized in 3 ml of 50
mM HEPES-KOH, 5 mM DTT, 1 mM phenylmethylsulfonyl fluoride (PMSF),

127

1 mM benzamidine using a Brinkmann polytron at the highest setting. SDS was
then added to the homogenate to a final concentration of 1%. This mixture was
then incubated in a boiling water bath for 5 minutes. The homogenate was
clarified by centrifuging at 10,000 g for 20 minutes at 15°C. The protein
concentration of each leaf extract was determine as described by Bradford
(1 97 6). One hundred u g protein aliquots of each leaf extract were fractionated
on 10% SDS-PAGE gels as described by Laemmli (1970). Proteins were then
transferred to nitrocellulose filters as described by Post-Beittemniller et al.
(1989) and the filters were probed with either a 1/250 dilution of the affinity
purified biotin carboxylase polyclonal antibodies or a 1/2000 dilution of biotin
antibodies as described by Roesler et al. (1995).

Fatty acid Analysis

Approximately 50 mg of fresh leaf tissue was homogenized in 1 ml of 50
mM HEPES, 5 mM DTT, 1 mM PMSF, 1 mM benzamidine plus 25 pg
heptadecanoic acid (17:0). Total lipid was extracted from the homogenate as
described by Bligh and Dyer (1959). The extracted lipids were concentrated
under nitrogen and fatty acid methyl esters (F AMEs) were prepared by
resuspending the dried lipid in 1 ml of 10% boron trichloride/methanol (Sigma).
This mixture was heated at 90*C for 30 minutes. The methylation reaction was
stopped by adding of 1 ml of water and FAMEs were extracted 2 times with 2
ml of hexane. Leaf FAMEs were analyzed by gas chromatography (GC) with a
Hewlett-Packard 5890 GC using a 30 m x 0.25 mm id. DB 23 column (J&W
Scientific, Rancho Cordova, CA) with a oven temperature programmed from
180*C (10 minute hold) to 230*C at 3*C/minute with a column head pressure
of 200 kPa of helium.

Results and Discussion

Over- and antisense-expression of biotin carboxylase resulted in the
generation of transgenic tobacco showing a wide range of leaf biotin
carboxylase expression levels.

The full length tobacco biotin carboxylase cDNA (Shorrosh et al., 1995)
was subcloned in the sense and antisense orientation behind the CaMV 35S
promoter. The tobacco biotin carboxylase over- and antisense-expression
constructs, designated p0719S and p0719A respectively, were used in the
Agrobacterium mediated transformation of tobacco leaf disc. These
transformations yielded 24 independent kanamycin resistant plants from each
construct.

The levels of biotin carboxylase protein expression in the leaves of
p0719S and p0719A transfonnants were determined by immuno—blot analysis
using antibodies affinity purified to the tobacco biotin carboxylase. A broad
range of biotin carboxylase expression levels was observed in the leaves of both
the sense (p0719S) and antisense (p0719A) transformants (Fig. 1A and B). Leaf
biotin carboxylase protein was significantly elevated over wild type levels in
several of the plants transformed with the p0719S biotin carboxylase over-
expression construct. These transformants contained biotin carboxylase levels
that ranged fi'om 2 to 5 fold over wild type levels (i.e. 82 and S6, Fig. 1A).
Northern blot analysis of RNA isolated from the leaves of these p0719S plants
indicated that biotin carboxylase mRNA levels also increased in correlation with
the observed increase in biotin carboxylase protein (data not shown). None of
the 24 independent p0719S transformants were found with more than a five fold

increase in leaf biotin carboxylase levels.

128

129

wt1wt2 $2 S3 $4 86

h... 1

B.
wt1 A5 A7 A8 A11 A14

Figure 5-1. Biotin carboxylase levels in transgenic tobacco leaves. One
hundred ug aliquots of leaf protein extract, isolated from plants transformed with
(A) the p0719S over-expression construct and (B) the p0719A antisense
construct, were fractionated on 10% SDS-PAGE gels and blotted to
nitrocellulose. Biotin carboxylase was detected using affinity purified antibodies
to the tobacco biotin carboxylase.

130

Decreased biotin carboxylase protein levels were observed in
approximately 30% of the plants transformed with the p0719A construct. Of
these plants, most showed decreases in leaf biotin carboxylase levels that ranged
from 30% to 50% of wild type levels (i.e. A14 and A15, Fig. 1B). However, of
the twenty-four independent p0719A transformants, only one plant was severely
deficient in leaf biotin carboxylase. This transfonnant, designated as All,
showed a 90% decrease in leaf biotin carboxylase protein relative to wild type
plants (Fig. 1B). Northern blot analysis of RNA isolated from the leaves of the
imder expressing p0719A plants using sense and antisense strand specific RNA
probes showed that levels of biotin carboxylase mRNA were almost
undetectable while a large accumulation of antisense RNA was observed (data
not shown). The observation that no plants were found containing biotin
carboxylase levels lower than 10% of wild type levels may suggest that biotin
carboxylase must be present at a certain minimal level to maintain the viability
of the plant. Furthermore, the low frequency of p0719A transformants containing
biotin carboxylase levels lower than 50% of wild type levels could also be
explained by the hypothesis that decreasing biotin carboxylase levels to very low
levels is detrimental to the plant.

While the expected increases and decreases in leaf biotin carboxylase
levels were observed in the majority of the plants transformed with the over- and
antisense-expression constructs, a small proportion of the p0719S and p0719A
transformants gave unexpected patterns of biotin carboxylase expression. For
example, of the 24 p0719S transfonnants, one plant (i.e S4, Fig 1A), contained
decreased levels of leaf biotin carboxylase protein. The decrease in biotin
carboxylase expression in these plants transformed with an over-expression

construct was most likely due to transgene inactivation. Similar results were first

131
observe during attempts to over-express the chalcone synthase gene using the
CaMV 35S promoter in petunia (Napoli et al., 1990; van der Krol et al., 1990).
Since then this phenomena has been observed by several groups and transgene
inactivation is now being used as an alternative to antisense approaches as a
means to decreasing gene expression in plants (Finnegan and McElroy, 1994 and
references within).

Surprisingly, several plants transformed with the antisense biotin
carboxylase construct contained elevated levels of leaf biotin carboxylase. Two
"antisense" transformants, A5 and A7 (Fig 1B), were found to contain biotin
carboxylase levels that were 5 to 10 fold higher than those measured in wild type
plants. These increased levels of biotin carboxylase expression rivaled and
surpassed those observed in the highest expressing p0719S plants (i.e. $2 and
S6 see Fig. 1A). Biotin carboxylase over-expression was also observed at a low
fiequency in tobacco suspension cells independently transformed with the
p0719A antisense construct (data not shown). At present, we do not have an
explanation for over-expression of biotin carboxylase occurring from plants
transformed with an antisense construct. However, because rearrangements of
the T-DNA have been reported to occur upon the integration during
Agrabacterium mediated transformation, one possible explanation could involve
the rearrangement of the antisense construct. Alternatively, the integration of the
p0719A T-DNA could have occurred such that the transgene was oriented in the
sense orientation behind a strong endogenous promoter. Extensive genomic
DNA analysis would be needed to distinguish between these and other
possibilities.

132
Increases and decreases in biotin carboxylase levels had no effect on the
expression of the BCCP subunit of ACCase

To determine if the levels of the different subunits of the plastid ACCase
were maintained in a strict stoichiometric ratio, the levels of BCCP were
determined in the leaves of plants with altered levels of biotin carboxylase.
Several reports indicate that the expression of individual subunits of various
chloroplast complexes are coordinately regulated with the levels of other
subunits or components of the same complex. For example, the expression of the
ribulose 1,5-bisphosphate oxygenase/carboxylase small subunit (S SU-
RUBISCO) was found to be dependent on the levels of the large subunit of
RUBISCO (LSU-RUBISCO) such that uncomplexed subunits were rapidly
degraded (Schmidt and Miskind, 1983). Furthennore, antisense-expression of the
SSU-RUBISCO in tobacco resulted in the coordinate decrease of the LSU-
RUBISCO (Rodennel et al., 1988). It has also been reported that in chlorophyll
biosynthetic mutants, light-harvesting chlorophyll a/b proteins (LHCPs) are
rapidly degraded due to their inability to assemble the correct pigment/protein
complex (Plumley and Schmidt, 1995). The stoichiometric ratios of RUBISCO
and LHC protein subunits are maintained by a post-translational mechanism
where excess subunits are proteolyticly degraded.

To determine if the subunits of the plastid ACCase are coordinately
regulated in a similar fashion to the RUBISCO or LHC subunits, BCCP levels
were measured in the leaves from wild type plants and the two transgenic plants
containing the highest and lowest levels of leaf biotin carboxylase, specifically
A5 and A11. The studies revealed that the BCCP levels were approximately
equivalent in A5 and A11 leaf extracts, but were slightly higher in wild type
leaves (Fig. 2). Although it is not clear why wild type plants expressed BCCP

133

wt1 A5 A11
Biotin
Carboxylase -

BCCP . Q .

Figure 5-2. Comparison of leaf BCCP and biotin carboxylase levels in
transgenic tobacco leaves containing altered levels of biotin carboxylase.
One hundred ug aliquots of leaf protein extract, isolated from wild type, A5, and
A11 plants were nm in duplicate on 10% SDS-PAGE gels and blotted to
nitrocellulose. Biotin carboxylase and BCCP were detected on separate blots
using affinity purified antibodies to the tobacco biotin carboxylase and antibodies
to biotin.

134

levels at slightly higher levels than either of the transgenic lines, these results
indicate that BCCP levels were not effected by either large increases or large
decreases in biotin carboxylase levels. Therefore, we conclude that biotin
carboxylase levels are not coordinately maintained in a strict stoichiometric ratio
with the BCCP subunit of the plastid ACCase. The presence of excess biotin
carboxylase subunits in over-expressing plants without an increase in BCCP
levels indicates that the plastid must be able to tolerate uncomplexed ACCase
subunits. It appears that the mechanism maintaining the strict coordination of
RUBISCO and LHC subunits does not apply to ACCase.

Plants containing extremely low levels of biotin carboxylase exhibit a
stunted phenotype when grown under low light conditions

While plants exhibiting high and moderately low levels of leaf biotin
carboxylase were phenotypically indistinguishable from wild type plants, the
grth and development of the All transformant, which contained very low
levels of leaf biotin carboxylase, was severely retarded when grown under low
light conditions (i.e. approximately 45 uE/m’) (Fig. 3). In addition to the stunted
growth, the leaves of the A11 plant were slightly chlorotic and less succulent
than the leaves of wild type plants that were grown under the same low light
conditions. However, when the A11 plant was transferred to higher light
conditions (i.e. >250 uE/mz), it grew vigorously and caught up developmentally
with wild type plants.

The stimted growth under low light conditions of the A11 plant was also
observed in the kanamycin resistant progeny of the self crossed primary
transformant suggesting that the antisense trait was segregating as a single
genetic locus (data not shown). The degree of stunting appeared to be more

 

-

.‘L—a- was

A.

_ --‘- ”Pi-~v‘
mm»... .- mifi,’ “ “ ‘1

135

 

     

wild type A5 A14 A11

 

 

Figure 5-3. Growth of wild type plants and plants expressing increased and
decreased levels of biotin carboxylase under low light conditions. A wild
type plant and primary p0719A transformants were grown in a growth chamber
at 25°C under constant illumination of 45 uE/mz.

136
severe in 1/4 of the second generation plants indicating that the antisense
phenotype was additive and possibly due to a gene dosage effect. When the
second generation A11 plants were grown under higher light conditions, the
stunted phenotype was again not observed (data not shown).

Decreased fatty acid content was associated with All plants grown under
both high and low light conditions.

Although a dwarf phenotype was associated with plants exhibiting large
decreases in leaf biotin carboxylase levels, it was not clear if this phenotype was
associated with decreased fatty acid synthesis due to lower ACCase activity in
these plants. Unfortunately, due to the instability of the ACCase and FAS
complexes, it was difficult to accin‘ately access rates of fatty acid synthesis and
chloroplast ACCase activity in isolated tobacco chlor0plast or leaf extracts. It
was however possible to measure leaf fatty acid content by gas chromatography
analysis. Such analysis allows for an indirect determination of the over-all effect
that changes in biotin carboxylase levels had on membrane lipid metabolism.
Fatty acid analysis was performed on leaves from plants expressing increased,
decreased,
and wild type levels of biotin carboxylase. Leaves that were of the same
approximate size were sampled from plants grown under low and high light
conditions.

These studies revealed that All leaves synthesized less fatty acid on a
fresh weight basis than the leaves of plants expressing wild type or elevated
levels of biotin carboxylase. This decrease in All leaf fatty acid content was
observed in plants grown under high and low light conditions. The stimted A11
plants grown under low light conditions had a 26% decrease in leaf fatty acid

137

content when compared to wild type plants grown under the same conditions
(Fig. 4A). Although the A1 1 plants were phenotypically indistinguishable from
wild type plants when grown under higher light conditions, a modest, but
statistically significant, 8% decrease in fatty acid content was measured in the
leaves of these plants (Fig. 43). Although the fatty acid content varied in
different leaves sampled fi'om the same plant, according to the Wilcoxon-Mann-
Whitney statistical test, the leaf fatty acid content of All plants, grown under
both light conditions, was significantly different (i.e. confidence level > 99%)
from that of wild type plants grown imder the same conditions. This same
statistical test also indicated that the leaf fatty acid content of AS, the transgenic
plant showing the highest levels of biotin carboxylase, was not significantly
different from wild type plants when grown imder either conditions. Additionally,
the leaf fatty acid content of plants showing only modest decreases in leaf biotin
carboxylase levels were also not significantly different from wild type plants
(data not shown).

The biotin carboxylase subunit of ACCase may be present in excess in wild
type tobacco leaves

The observation that biotin carboxylase levels could be decreased to as
low as 50% of wild type level without effecting the levels of BCCP, fatty acid
content or growth rate suggest that biotin carboxylase may be present in excess
in wild type plants. It was only after biotin carboxylase levels were decreased by
90% in All plants that changes in fatty acid content and growth rate were
observed. This result is consistent with those of antisense experiments using the
genes encoding pyrophosphate:fructose-6-phosphate phosphotransferase,
fructose-1,6-bisphosphatase, and citrate synthase. In these studies, a greater than

138

Figure 5-4. Leaf fatty acid content of wild type, A5 and A11 plants grown
under (A) low light and (B) high light conditions. Approximately 50 ug leaf
samples were taken from leaves of plants grown under 45 [Elm2 and 300 uE/m2
intensity light. Total lipid was extracted from leaves with 20 or 25 pg of
heptadecanoic acid (17:0) added as an internal standard. The total lipid was then
transesterified to fatty acid methylesters (FAMEs) using BCl3 in methanol.
FAMEs were then analyzed by gas chromatography. (A) Leaf fatty acid content
of plants grown under constant illumination by 45 irE/m2 light. (B) Leaf fatty
acid content of plants grown under constan illumination by 300 uE/m2 light. The
insert if (B) shows the same data expanded on a different scale. Horizontal bars
represent the mean value of the combined samples from each plant. A5, A11,
and wild type plants are represnted by closed circles, closed squares and close

triangles respectively.

139

222s 52. 9528 Be a:

 

 

 

 

 

 

 

 

 

 

A.A A A
e ¥
0.
A A A?!‘ W. LI. 8.
w ace—'-
1
I I I— II M I
04.0 M -
_ _ _ _ _ a _ _ _ T
4 2 O 8 6 4 2 M mm mm 8 6 4 2

Ego; swat mE \ Bow 3:8 a: £995 58.— 9: \ Bow >88 a:

B.

wildtype

A11

A5

140

80% decrease in enzyme levels was needed before changes in metabolism or
phenotype were detected (KoBmann et al., 1994; Hajirezaei et al., 1994;
Landschutze et al, 1995). According to metabolic control theory, in instances
where large changes in enzyme levels are needed to alter flux through
bio synthetic pathway, these enzymes are thought to exert little control over
pathway flux (Kacser and Porteous, 1987). However, the results from antisense
studies using genes coding for enzymes with previously determined regulatory
importance again showed that large decreases in enzyme levels were necessary
for altered metabolism. This apparent inconsistency has recently been addressed
by Stitt and Sonnewald (1995). They suggest that, under many circumstances,
the principles of metabolic control theory do not apply. Plants appear to have
multiple mechanisms to compensate for deficiencies with a metabolic pathway
such that control is usually shared between several enzymes. Furthermore,
several regulatory enzymes operate at well below their maximum activity which
allows these enzymes to rapidly respond to changes in metabolism.

Homozygous p0719A-11 plants had abnormal floral development

All plants grown under green house conditions were allowed to flower
and go to seed. Approximately 1/4 of the kanamycin resistant progeny from the
A11 self cross contained mature flowers that were male sterile due to aberrant
anther development (Fig. 5). In place of anthers, these plants contained "petal-
like " organs attached to the ends of their stamens. The filament portion of the
stamen appeared normal as did all of the other floral organs of these plants.
These plants were still fertile and did produce seed when crossed with pollen
from other plants.

Although it is not clear whether this abnormal flower morphology was due

2,111.
.ig.35})}3i§31-? \l’ 1.xli‘in") .

 

141

   

Figure 5-5 Plants that are homozygous for the A11 transgene locus show
abnormal floral development. In photographs (A) through (C), tissues
harvested from All plants are on the left and tissues harvested from wild type
plants are on the right. Panel (A) shows whole flowers with the front petals
removed. Panel (B) contains stamens dissected from All and wild type flowers.

Panel (C) contains flowers with all petals removed.

142
the decrease in biotin carboxylase levels or if it was due to the insertion of the
p0719A T-DNA into a gene involved in floral development, alterations in
different metabolic pathways have been implicated in altering floral
development. For example, mutations in lipid metabolism have resulted in
Arabidopsis male sterility. McConn and Browse (1993), reported that plants
lacking trienoic fatty acids were male sterile due to their inability to produce
jasmonic acid which is needed for pollen maturation. Furthermore, severe
decreases in citrate synthase in potato plants transformed witha citrate synthase
antisense construct was associated with inhibition of flower formation
(Landschutze et al., 1995). The ovaries of these plants failed to develop past a
early stages of flower development. It may be possible that the aberrant stamen
development in All plants is due to insufficient fatty acid synthesis associated

with the decrease in biotin carboxylase levels.

Conclusion

Several new insights into the regulation and complex structure of the
plastid ACCase have been revealed through these studies. First, it appears that
the levels of biotin carboxylase are not maintained in a strict stoichiometric ratio
with BCCP and presumably other subunits of the plastid ACCase. As observed
in plants expressing high levels of biotin carboxylase, the tobacco chloroplast
were able to tolerate the accmnulation of uncomplexed subunits of ACCase.
These results indicate that the formation and maintenance of the plastid ACCase
complex differs from other plastid localized protein complexes such as
RUBISCO and LHC.

Second, transgenic tobacco plants were able to tolerate 5 0% decreases in

biotin carboxylase levels without effecting growth or leaf fatty acid content.

143
These results suggest that the biotin carboxylase subunit of ACCase or the
ACCase complex may be present in excess in wild type leaves. However, under
low light conditions, severe retardation of grth was observed in the A11 plant
which expressed very low levels of leaf biotin carboxylase. Analysis of the leaf
fatty acid content of the A1 1 transformant suggested that the stunted phenotype
was associated with these plants inability to maintain sufficient rates of fatty acid
synthesis. In addition, aberrant stamen development was observed in the flowers
of homozygous A11 plants. Because only one transformant was generated that
exhibited these unusual phenotypic traits, it is not certain if these traits can be
directly associated with decreased biotin carboxylase levels. It is possible that
the atypical patterns of growth and development associated with the A11
transformant are due to the insertion of the p0719A T-DNA into region of the
tobacco nuclear genome which resulted in pleiotropic events leading to these
phenotypic changes.
Acknowledgements

We would like to thank Wan-ng Chu and Pam Green for the generous

gift of the p1079 plasmid.

Literature Cited

Alban C Baldet P, Douce R (1994) Localization and characterization of two
structuraflydrfi‘erent forms of acetyl-CoA carboxylase myoun pea leaves of

gtgh7ic§i6gne lS sensrtive to arloxyphenoxypropmate herbicides. iochem J 300:

An G, Ebert PR, Mitra A, .Ha SB (£2885) Binary vectors. In Plant Molecular
Biolo Manual SB Gelvrn and chilperoot eds, Kulwer Academic
Publis ers, pp A3: 1-19.

Bevin M 1984 B' A abacterium vectors for lant transformation. Nucl
Acids Res(12: 19711-8527? p

Bli h E_G, er WJ 1959 A. rapid method of total lipid extraction and
p cation. anJBioc em ySiol 7: 911-917.

Bradford MM (1976) A rapid and sensitive method for the ,quantitation of
micro .am quantities of rotem utihzmg the principle of protem-dye binding.
Anal iochem 72: 248-2 4.

Choi JK, Y_u F, Wurtele ES, Nikolau BJ (1995) Molecular cloning and
characterization of the cDNA coding for the biotin-contaimn subumt of the
chloroplastic acetyl-CoA carboxylase. Plant Physrol 109: 619- 24.

Finne an J, McElro D (1994) Transgene inactivation: Plants fight back.
Bio/tee ology12: 88 -888.

Goodridge AG 1985) Fattg acid synthase in eukaryotes. InBiochemistry of
lipids and Mom ranes, D Vance JE Vance, eds Benjamrn/Cummrngs
Publishing Company, Inc, Menlo Park, CA, pp 143-180.

Hajirezaei M, Sonnewald U, Viola R, Carlisle S Dennis D, Stitt M (1994
Trans enic potato plants wrth stron y decreased expresswn . o
pryopl osphate:fructose-6-phosphate. posp otransferase show no v1s1ble
Eggnog/p3?) and only minor changes in metabohc fluxes in their tubers. Planta

 

Hoekema A, Hirsch PR, Hooykaas PJJ,_SchilFeroort RA 1983) A binary
plant vector strate based . on separation o vir- and -region of the
Agrabacterium tume aciens Ti plasmid. Nature 303: 179-180.

Kacser H Porteous JW 1987 Control of metabolism: What do we have to
measure. Trends Biochem c1 1 : 14

Kamiryo T, Numa S (1973) Reduction of acetyl coenzyme A carbo lase
ggnfieznt of Saccharomyces cerevisae by exogenous fatty acids. FEBS Le . 38:

KoBmann J Sonnewald U, WillmitzerL (1994) Reduction of the chloroplastic
fiuctose-l ,6-bisphosfi1hatase in trans gemc potato plants impairs photosyntheSis
and potato growth. e Plant Journal. 6: 6 7-650.

 

144

145

Laemmli UK (1970) Cleavlage of structural proteins during the assembly of the
head of bacteriophage T4. ature 227: 680-685.

Landschutze V, WillmitzerL, Muller-Robor B (1995) Inhibition of flower
formation by antisense repressron of mrtochondnal citrate synthase in “21115151618118

tato lants leads to a s ific disinte tion of ov tissues of fl , E
914: 60-666. p“ 3” 3’? owers

Ma nson K, Jackowski SJ, Rock CO, Cronan JE (1993) Regulation of fatty
aci iosyntheSis in Escherichia coli. Microbiological Revrews: 57: 522-542.

McConn. MM, BrowseJA (1993) A mutant of Arabido sis thaliana deficient
in trienoic fatty. ac1ds 18 male. sterile. Proceeding of e 1993 Plant Lipid
Symposrum, National Plant Lipid Cooperative.

Na li C, Lernieux C, J orgensen R (1990) Introduction of a chimeric chalcone
e gene into petuma results in reversrble co-suppressron of homologoues
genes in trans. The Plant Cell 2: 279-289.

Page RA, Okada. S, Harwood JL (1994) Acetfigl-CoA carbo lase exerts stron
811619 gogtrol over hpid synthesrs in p ants. Bioc ca et Biop ysrca Acta 121 :

Plumley FG Schmidt GW 81995) Li t-harvesting chloro hyll A/B com lexes
304terdepend’ent prgrnent syn esrs an protein assembly. e Plant Cell : 689-

Post Beittenmiller MA, Schmid KM, Ohlrogge JB (119891) Expression of halo
and apo forms of the s mach ac 1 carrier protem- in caves of transgemc
tobacco plants. The Plan Cell 1: 8 9-899.

Post-Beittenmiller D, Jaworski JG, Ohlrog%e JB (1991) In vivo pools of free
and acylated acy carrjieerrotems in s mach: v1dense for srtes of regulation of
fatty acid biosyntheSis. Biol Chem 66: 1858—1865.

Post-Beittenmiller D, Rougihan G Ohlrrgge JB (1992) Regulation of plant
fatty .aCId biosyntheSis; Ana ySIS of acyl- oenzyme A and acyl-ac 1 carrier
Barbtem substrate pools in spinach and pea chloroplast. Plant Physrol 00: 923-

Rodermel SR, Abbott MS, .Bo orad L (1988) Nuclear-organelle interactions:
Nuclear antisense gene inhibits n ulose bisphosphate carboxylase enzyme levels
in transformed tobacco plants. Cell 55: 6 3-6 1.

Roosler KR, Savage LJ, Shintani DK, Shorrosh BS, Ohlrogge JB (1995)-Co-
purification, co-unmunoprecrptatio and coordinate exgressron of acetyl-CoA
carboxylase actrvrty, b10111). carbo ase, and biotin car oxyl carrier protein of
higher plants. Planta 198: (in pres?)

Rogers SG, Horsch RB, .Fraley RT (1986) Gene transfrer in plants: Production
ggéransformed plants usrng Ti plasnnd vectors. Methods Enzymol. 118: 627 -

146

Sasaki Y Hakamada K, Suama Y, Nagano Y Furusawa I, Matsuno R
(1994 Chloroplast-encoded rotem as a su umt of acetyl-CoA carboxylase in
pea p ant. J Biol Chem 268: 5118-25123.

Schmidt GW, Miskind ML (1983) Rapid degradation of unassembled ribulose
{ligfipgilos léiatze 2cgglgoxylase small subumts in chloroplast. Proc Natl Acad Scr

Shintani DK, Ohlro e JB 1995 Feedback inhibiton of fa acid 5 thesis
in tobacco suspensionggells. The Plgint Journal 7: 577-587. tty yn

Shorrosh BS, Roesler K, Shintani .DK, Van De Loo F, Savage L, Ohlrogge
JB (1995 Structural analysrs, lastid localization, and expressron of the biotin

)
carbo lase subunit of ace 1- 0A carbo lase from tobacco. Plant Ph siolo
108: 80812. ty xy y gy

Stitt M, Sonnewald U 1995) Re . ation of metabolism in transgenic plants.
Ann Rev Plant Phsyiol P ant M01 101 46: 341-368.

van der Krol AR, Mur LA, Belsd M, Mo] JN, Stuitje AR.(1990) Flavonoid
genes in. petuma: addition of a hmited mmber of gene copies may lead to a
suppressron of gene expreSSion. The Plant Cell 2: 2 1-299.

Chapter 6

Conclusions and Future Research Perspectives

Understanding the role of multiple ACP isoforms in plants

Acyl carrier protein (ACP) is an essential co-factor involved in fatty acid
synthesis. This protein fimctions to shuttle the growing fatty acid chain between
the enzymes of the fatty acid synthase complex. For an as yet unknown reason,
many plants plastids contain multiple isoforms of ACP (Ohlrogge and Kuo,
1985). Although results from in vitro studies suggest that in spinach chloroplast,
the different ACP isoforms may direct mature fatty acids towards different
metabolic fates (Guerra, et al., 1986), presently, no in viva data exist to support
this hypothesis.

However, because the expression of the different plastid ACP isoforms
have been extensively characterized in Arabidopsis (Hlousek-Radoj cic et al.,
1991; chapter 2) and DNA sequences corresponding to 4 different Arabidopsis
ACP isoforms have been cloned (Lamppa and Jacks, 1991; Hlousek-Radojcic,
1991; Post-Beittenmiller et al., 1989; Chapter 2), we now have a good system
to study the in viva role of multiple ACP isoforms in plants. Using transgene
technologies, it is now possible to alter the expression of individual ACP
isoforms in Arabidopsis and determine what, if any, efi‘ect these changes would
have on plant lipid metabolism.

The identification of a cDNA corresponding to the LMI ACP is
particularly helpful since, unlike the previously cloned Arabidopsis ACPs, the
LMI ACP is expressed almost exclusively in green tissues and is the
predominant ACP isoform in leaf tissue. Arabidopsis leaf fatty acids have been
shown to be partitioned between plastid and microsomal localized glycerolipid

147

148

biosynthetic pathways (Browse and Somerville, 1991). As was hypothesized by
Guerra et al. (1985), LMI ACP may be involved in directing fatty acid chains
towards one of these two metabolic fates. Because the plastid localized
glycerolipid pathway exists primarily in photosynthetic tissues, it is logical to
propose that the LMI ACP is involved in directing fatty acids towards this
pathway. If this were the case, lowering the expression of the LMI ACP in green
tissues might divert the flux of fatty acids away from the plastid localized
glycerolipid pathway and towards the microsomal pathway. This should result
in a change in the leaf fatty acid composition to resemble that of the Arabidopsis
actl mutant which is deficient in the ability to synthesize glycerolipids through
the plastid localized pathway (Kimst et al., 1988).

Over-expression studies of the LMI ACP in a tissue where it is not
normally found (i.e. the root), may also help determine what role it plays in fatty
acid synthesis. LMI ACP expression in root tissue may alter the fatty acid
composition of the root to more closely resemble the composition of green tissue.
Similar studies are now under way where the Arabidopsis ACP-1 isoform was
over-expressed in Arabidopsis leaves. ACP-l levels have been significantly
increased in leaf tissue and this isoform is now present at approximately the
same level as the LMI ACP. Preliminary results from these studies suggest that
the increased level of ACP-1 has increased the flux of fatty acids towards the
microsomal glycerolipid pathways. Although further experimentation is needed,
these results are very promising. Antisense and over-expression constructs of the
LMI ACP cDNA have been constructed and are now ready to transform into
Arabidopsis. These experiments are now under way and preliminary analysis of
LMI antisense and over-expressing plants will be available in the near future.

An alternative approach to imderstanding the role of multiple plastid ACP

149

isoforms in plants involves the measurements of acyl-ACP pools. Post-
Beittenmiller et al. (1991), reported that in spinach the two ACP isoforms
participated equally in the early enzymatic steps of fatty acid synthesis.
However, because they were unable to measure the long chain acyl-ACP pools
of both ACP isoforms, it was not possible to make any evaluations of how the
two ACPs participate in the terminal reactions of fatty acid synthesis. The
purification of the LMI ACP from Arabidopsis leaves has shown that deacylated
leaf ACP isoform can be fractionated by anion-exchange chromatography
(Chapter 2). The resolution of ACP isoforms is in part due to the differences in
the p1 and hydrophobic nature of the different Arabidopsis leaf ACPs (Chapter
2). It may be possible to exploit these physical difference to fractionate the
different ACPs such that we will be able to analyze acyl-ACPs of different chain
lengths in an isoform specific manner. Since we have cloned four of the five
different leaf ACP isoforms and expressed each in E. coli (Hlousek-Radojcic et
al., 1991; Chapter 2), it is possible to synthesize acyl-ACP standards for each
isoform (Kuo and Ohlrogge, 1985). These different acyl-ACP standards can then
be used to establish whether acyl-ACPs can be fractionated in an isoform
specific manner using different chromatographic methods. The ability to
fractionate the different acyl-ACPs in an isoform specific manner will allow us
to directly access their participation in vivo in the different terminal reactions of
fatty acid synthesis.

Mitochondrial ACP and fatty acid synthesis
Cumulative evidence from mitochondrial protein import studies, sequence
comparisons with known mitochondrial ACPs from animals and fungi, and

immunolocalization studies have proven that the protein encoded by the

1 50
Arabidopsis mtACP-l cDNA is targeted and localized within the plant
mitochondria (Chapter 3). Furthermore, these studies indicate that the majority
of mtACP-1 protein is present primarily in the acylated form and appears to be
membrane associated. When the mtACP-1 was used in an in vitro fatty acid
synthase assay, the fatty acid products were primarily medium chain fatty acids
(i.e. decanoic acid (C10:0), lauric acid (C12:0), myristic acid (C14:0)).

L Are plant mitochondria capable of de nova fatty acid synthesis and if so
why?

The observations described in Chapter 3 and summarized above are
consistent with those from studies of the Neuraspora and bovine heart muscle
mitochondrial ACPs. In all three cases, the acylated form of mtACP was found
to be membrane associated (Runswick et al., 1991; Sackmann et al., 1991;
Chapter 3). Furthermore, Neuraspora mitochondria have been reported to be
capable of de nova fatty acid synthesis (Brody et al., 1990; Zenson et a1, 1992)
which is consistent with our findings that the Arabidopsis mtACP-1 is able to
function as a F AS co-factor.

Although through the work described in Chapter 3, we were able to prove
that plants poccess a mitochondrial form of ACP, we were not able assign a
function to the plant mitochondrial ACP. Our data suggesting that the
Arabidopsis mtACP-1 participated in fatty acid synthesis was indirect and
needed to be followed up with careful labelling studies with isolated
mitochondria. Fortunately, Dr. Haj ime Wada from the University of Kyushu in
Japan, has continued the work on the plant mitochondrial ACP. Through
radiolabelling studies, he has shown that pea mitochondria are capable of de
novo fatty acid when provided with [l-“C]malonate. These labelling studies

1 5 1

revealed that the primary fatty acid products of pea mitochondrial FAS are 1-
hydroxymyristic acid and lauric acid. By analyzing the radiolabelled acyl-ACPs,
Dr. Wada identified l-hydroxymyristoyl-ACP and lauroyl-ACP as the primary
mitochondrial acyl-ACP intermediates and has thus proven directly that the plant
mtACP is a co-factor of mitochondrial FAS. These results also confirm the
findings that, in the presence of mitochondrial ACP, medium chain fatty acids are
synthesized in FAS assays (Chapter 3).

The necessity for a rediuidant fatty acid synthase localized in the
mitochondria is curious since cellular demand for fatty acid synthesis for
membrane biosynthesis appear to be met by the cytosolic and plastid localized
FAS in non-photosynthetic and photosynthetic eukaryotes. It is possible that
mitochondrial fatty acids are used in the post-translational acylation of specific
proteins. Acylation of mitochondrial proteins has been demonstrated by feeding
radiolabelled fatty acids to isolated rat liver mitochondria (Stucki et al., 1989).
The mitochondrial protein acylation reaction had a substrate preference for
medium chain fatty acids and radiolabelled proteins were resistant to hydrolysis
by hydroxylamine and base indicating that the acylation was probably occuning
through an arnide-linkage. Furthermore, Vassilev et al. (1995) have reported that
in Neuraspora, cytochrome c oxidase is myristylated via an amide linkage to a
specific lysine residue. These results are consistent with the findings from [1-“C]
malonate labelling of isolated pea mitochondria where several radiolabelled
proteins were shown to be resistant to base and hydroxylamine (W ada and

Ohlrogge, personal communication).

1 52
H. Is the plant mitochondrial ACP a subunit of the NADH:ubiquinone
oxidoreductase (complex 1)?

Groups worldng on NADH:ubiquinone oxidoreductase (complex I) in
bovine heart and Neu rospara mitochondria found that the acylated form of the
mtACP is an integral subunit of complex I (Runswick et al., 1991; Sackmann et
al., 1991). The Neuraspora complex I has been shown to be composed of two
tightly associated domains. A membrane domain, composed of proteins encoded
by the mitochondrial genome, constitutes the ubiquinone hydrogenase ftmction
of complex one (Weiss et al., 1991). The other domain, which is composed
primarily of proteins encode by nuclear genes, protrudes into the mitochondrial
matrix and contains the NADH dehydrogenase activity (Weiss et al., 1991). The
mitochondrial ACP has been shown to be associated with the matrix domain of
complex I (Sackmann et al., 1991). Although it is not clear what role the mtACP
plays in complex I, Schulte et al. (1994), reported that the formation of the
Neu raspara complex I is inhibited when the gene encoding the mtACP was
disrupted. This observation indicates that the presence of the mtACP is an
essential component of this protein complex.

The observations that only the acylated form of mtACP-1 is membrane
associated and that the Neuraspora and bovine mitochondrial ACPs are only
found in the acylated form when associated with complex I suggest that the
acylation may be key to mitochondrial ACPs role in complex I. It is known that
protein acylation is involved in the membrane anchoring of certain proteins
(McIlhinney, 1990; Towler et al., 1988). Furthermore, protein acylation has been
shown to facilitate protein-protein interactions (McIlhinney, 1990; Towler et al.,
1988). It is possible that the acylated form of the mitochondrial ACP is involved
in the anchoring of the matrix protruding domain of complex I to either the

1 53

mitochondrial irmer membrane or to the membrane domain of complex I. If this
is the case, then it would be interesting to investigate the role of the plant
mitochondrail ACP in complex I under conditions when electron flow is diverted
away from complex I and towards the internal rotenone insensitive NAD(P)H
dehydrogenase in plants. It is possible that the two domains of complex I could
become imcoupled through the deacylation of the mtACP-1. Such a mechanism
would inactivate complex I and allow NADH reduction to occur through the
rotenone insensitive NAD(P)H dehydrogenase.

So far there is little direct evidence for the involvement of the
mitochondrial ACP in the NADH:ubiquinone oxidoreductase complex in plants.
We have recently sent antibodies to the Arabidopsis mtACP to a group studying
the Arabidopsis NADH:ubiquinone oxidoreductase. Hopefully, they will be able
to determine if mtACP-1 is present in complex 1. Additionally, Dr. Wada is
attempting to alter the expression of the mtACP-1 in Arabidopsis using antisense
approaches and he is now in the process of analyzing plants transformed with the
mtACP-1 antisense construct. Although a total knock out of mtACP-1 expression
is likely to be lethal to the plant, it is possible that decreased expression of
mtACP-1 will lead to a decrease in mitochondrial respiration due to decreased

capacity for electron transfer from complex I.

III. Multiple isoforms of ACP may be present in Arabidopsis mitochondria.

Recently, we have identified a cDNA clone from the Arabidopsis EST
cDNA collection that encodes a protein that shows a high degree of sequence
similarity to the Arabidopsis mtACP-1 protein and other known mitochondrial
ACP sequences. Although similar to mtACP-1, the newly identified cDNA clone

appears to correspond to an tmique mitochondrial ACP isoform. We are now in

1 54
the process of characterizing this cDNA clone. This finding suggest, that like
plastids, mitochondria may also contain multiple isoforms of ACP.

Product feedback regulation of plant fatty acid synthesis

Plant fatty acid synthesis is tightly regulated such that fatty acids are
produce only at quantities needed to meet the demand of glycerolipid
biosynthetic enzymes. The results presented in Chapter 4 indicate that plants
employ a product feedback mechanism to control the flux through the fatty acid
biosynthetic pathway. In these studies, tobacco suspension cultures treated with
exogenous fatty acid in the form of Tween-18:1 incorporated [l-“C]acetate into
fatty acids at a rate 3 to 5 fold slower than untreated cells. Analysis of acyl-ACP
intermediate pools isolated from Tween treated and control cells revealed that
fatty acid synthesis was being regulated by feedback inhibition occurring at the
level of acetyl-CoA carboxylase (ACCase). These studies also showed that the
long chain acyl-ACPs were rapidly turned over upon the addition of Tween-18:1
indicating that the long chain acyl-ACP thioesterases were not afi‘ected by the
feedback regulation. The feedback inhibition appeared to be occurring through
a biochemical mechanism since ACCase subunits and fatty acid synthase enzyme
levels were not altered in response to the feedback inhibition. Preliminary studies
have shown that the feedback inhibition of fatty acid synthesis in tobacco
suspension cells is rapidly reversed by the removal of Tween-18:1. The rapid
onset and reversal of the inhibition are consistant with control occuring at the

biochemical level.

 

1 55
I. What is the feedback molecule involved in the feedback inhibition of plant
fatty acid synthesis?

Although ACCase was identified as an important regulatory enzyme
involved in the feedback inhibition of fatty acid synthesis, the biochemical
mechanism by which the inhibition occurs is unknown. The classical feedback
inhibition mechanism involves the allosteric modification of the regulatory
enzyme caused by the binding of a product or particular pathway intermediate.
It is possible that such a mechanism is employed in plants during feedback
inhibition of fatty acid synthesis, however at this point, the identity of the
feedback molecule is not known.

Several potential candidates exist for the feedback molecule involved in
the inhibition of plant fatty acid synthesis. Long chain acyl-ACPs have been
suggested as likely feedback molecules. In E. coli, the depletion of long chain
acyl-ACPs caused increases in the rate of ACCase and fatty acid synthesis
(Ohlrogge et al., 1995; Jiang and Cronan; 1994). These results suggest that long
chain acyl-ACPs act as feedback molecules involved in the product inhibition of
E. coli fatty acid synthesis. However, long chain acyl-ACPs do not seem to play
the same role in plants. The observations that the long chain acyl-ACP pools are
rapidly turned-over upon initiation of feedback inhibition of plant fatty acid
synthesis indicates that these F AS intermediates are probably not involved in in
the feedback mechanism. Furthermore, long chain acyl-ACPs had little effect on
ACCase activity when added to in vitro ACCase assays (Roesler et al., 1996).
It is also not likely that the medium chain acyl-ACPs are fimctioning as feedback
molecules since the sizes of these acyl-ACP pools remained constant under both
control and feedback conditions.

The results fi'om in vitro ACCase assays also cast doubt on the role of free

 

1 56
fatty acids as feedback molecules since the addition of palmitic (16:0) and oleic
(18:1) acids did not inhibit ACCase activity (Roesler et al., 1995).

Another possible inhibitor molecule that may interact directly with
ACCase is acetyl-ACP. Acetyl-ACP levels rapidly increase upon treatment with
Tween-18:1. The accumulation of acetyl-ACP closely correlated with the degree
of feedback inhibition observed in Tween treated cells. It is possible that the
initial inhibition of FAS is occurring at B-ketoacyl-ACP synthase I (KAS I), the
enzyme responsible for the tum-over of the acetyl-ACP pool. This inhibition
could result in a subsequent increase in acetyl-ACP which results in the
inhibition of ACCase. This hypothesis would require acetyl-ACP to be a highly
potent inhibitor of ACCase such that the Ki for acetyl-ACP would be very low.
However, in dispute of this hypothesis, Jaworski et al. (1993) suggest that the
accumulation of acetyl-ACP is a direct consequence of decreased ACCase
activity. They propose that acetyl-ACP levels increase because, due to decreased
availability of malonyl-CoA, acetyl-ACP can not be utilized in condensation
reactions catalyzed by B-ketoacyl-ACP synthase 1. To distinguish between these
alternatives, we could test the effect of acetyl-ACP on ACCase ativtiy in in vitro
assays. If ACCase proves to be inhibited by acetyl-ACP, studies should then be
done look at the effect of potential feedback molecules (i.e. free fatty acids and
acyl-ACPs) on KAS I. If this hypothesis is correct, then it would appear that the
inhibition of ACCase would be a secondary effect of the primary inhibition
occurring at a different point in plant fatty acid synthase.

H. Interaction between plastid fatty acid synthesis and microsomal
glycerolipid synthesis.
Rates of plastid fatty acid synthesis must be coordinately regulated with

1 57

cytosolic demands for fatty acids. Therefore, the feedback of fatty acid synthesis
must involve inter-organellar communication between these two pathways. The
link between these two pathway appears to involve the cytosolic acyl-CoA
pools. The cytosolic acyl-CoAs are synthesized in the chloroplast outer envelope
membrane. After being cleaved fiom ACP backbones by acyl-ACP thioesterases,
free fatty acids move across the chloroplast envelope where they are ligated to
coenzyme A (CoA) in a reaction catalyzed by acyl-CoA synthetase. These acyl-
CoAs serve as substrates for microsomal glycerolipid synthesis. The cytosolic
acyl-CoA pools are very small and difficult to measure. It is likely that plant cells
try to maintain a low concentration of long chain acyl-CoAs because due to their
detergent nature, accumulation of these compounds may be toxic. The addition
of exogenous fatty acids is likely to over-load the cytosolic acyl-CoA pools. The
plant cell may then sense this increase in cytosolic acyl-CoA and act to prevent
the de nova synthesis of fatty acids. Studies suggest that although feedback
inhibition of fatty acid synthesis occurred upon the addition of exogenous fatty
acids, there was little direct interaction between the exogenously applied fatty
acid and the plastid localized fatty acid biosynthetic enzymes. Terzaghi (1986a
& b) determined that rabiolabelled oleic acid (18:1) derived form Tween-18:1
incorporated almost exclusively into microsomal glycerolipid biosynthetic
pathways and little or no label was present in plastid derived glycerolipids.

One possible point of regulation is at the plastid outer envelope acyl-CoA
synthetase. Studies indicate that acyl-CoA synthetase is inhibited by the
accumulation of acyl-CoA. In preliminary expreriments, acyl-CoA synthetase
activity was inhibited by the addition of oleoyl-CoA (18:1-CoA) or palrnitoyl-
CoA (16:0-CoA) (data not shown). These results are consistant with the
hypothesis that a primary cytosolic factor, possibly long chain acyl-CoAs, may

1 5 8
be involved in sensing the need to decrease rates of fatty acid synthesis.

To test the hypothesis that cytosolic factors are involved in communicating
the demand of microsomal glycerolipid pathways for increases or decreases in
fatty acid synthesis, both biochemical and molecular approaches can be applied.
The major thrust of these studies should focus on the role of the cytosolic acyl-
CoA pools in communication between the plastid and microsomal pathways.
Initial studies could be done to see if the acyl-CoA or free CoA pools are altered
by the addition of exogenous fatty acids. Acyl-CoAs could be analyzed from
tobacco suspension cells treated with different levels of Tween-18:1 to determine
the capacity of the acyl-CoA pools to accept exogenous fatty acids. These
studies could be conducted in conjtmction with [1-“C]acetate labelling studies
to determine if there is a relationship between rates of fatty acid synthesis and
changes in the acyl-CoA pools.

Transgenic plants offer another approach to examine thet effect of
perturbing the cytosolic acyl-CoA pool on inter-organellar communication.
Several groups (i.e. Ernst Heinz, University of Hamburg; Matthew Hills, John
Irmes) have been started to characterize plant acyl-CoA synthetases and several
acyl-CoA synthetases have recently been identified from the Arabidopsis EST
cDNA collection. These efforts will result in the identification of a cDNA clone
corresponding to a plastid outer envelope localized acyl-CoA synthetase. Once
such a clone is in hand, it will be possible to determine if under- and over-
expression of this acyl-CoA synthetase in transgenic plants. Once plants with
altered acyl-CoA synthetase levels have been engineered, one can determine
what effect these changes will have on the plastid fatty acid synthesis and the
microsomal glycerolipid pathways' ability to respond to the addition of
exogenous fatty acids. These studies may lead to a an understanding of the

1 59
cytosolic acyl-CoA pool role in regulating interorganellar lipid metabolism.

Antisense- and over-expression of the biotin carboxylase gene in transgenic
tobacco leaves.

As was described in chapter 5, initial experiments were conducted to
evaluate the efiect of altered biotin carboxylase gene expression on the plastid
acetyl-CoA carboxylase (ACCase) complex formation and leaf fatty acid
metabolism in transgenic tobacco. Biotin carboxylase is a subunit of the
heteromeric plastid localized ACCase. Biotin carboxylase and three other protein
subunits, specifically biotin carboxyl carrier protein (BCCP), and the alpha and
beta subunit of carboxyltransferase, combine to form the plastid ACCase
complex. ACCase catalyzes the first committed step of fatty acid synthesis in
which acetyl-CoA is carboxylated to malonyl-CoA. The results from several in
viva studies indicate that the plastid ACCase plays a role in regulating rates of
plant fatty acid synthesis. Tobacco plants were transformed with constructs
containing the tobacco biotin carboxylase gene expressed in the sense and
antisense orientation behind the caulifower mosaic virus 35S transcript (CaMV)
promoter. Plants generated fiom these transformations contained leaf biotin
carboxylase expression levels ranging from 10% to 1000% of wild type leaf
levels. Altered levels of leaf biotin carboxylase had no efl‘ect on BCCP levels
indicating that the subunits of the plastid ACCase were not maintained in a strict
stoichiometric ratio. Although the phenotype and leaf fatty acid content of plants
showing high and moderately low levels of biotin carboxylase were
indistinguishable from wild type plants, stimted growth and decreased leaf lipid
content was observed in a single tranformant, A1 1, that was severely deficient

in leaf biotin carboxylase. The altered phenotype and leaf fatty acid content was

 

1 60
only observed when the A11 plant was grown under low light conditions.

I. Is the stunted phenotype and decreased leaf fatty acid content of A11
plants grown under low light conditions due to decreased biotin carboxylase
levels?

Under low light conditions the severe decrease in biotin carboxylase levels
in the leaves of All plants was associated with a decrease in leaf fatty acid
content and stimted phenotype. These effects could be attributed to decreased
ACCase levels, which subsequently caused the plant to be unable to synthesized
fatty acids at rates suflicient to maintain normal growth. Alternatively, it is
possible that the p0719A T-DNA inserted aberrantly into the tobaco genome
which resulted in the A1 1 phenotype. In order to distinguish between these two
possibilies, a series of diagnostic genetic crosses were made with homozygous
All plants.

The A11 phenotype may be due to an insertion of the p0719A T-DNA into
the tobacco genome which resulted the inactivation of an essential gene. Such
a mutation would most likely be recessive in nature. To determine if this were
the case, a homozygous All plant was backcrossed with wild type tobacco. The
progeny of this cross will be heterozygous for the T-DNA insertion and the wild
type locus. If the stunted A11 phenotype is the result of a recesive mutation
caused by the insertion of the T-DNA, one would expect the progeny to no
longer express the A11 phenotype due to the complementation by the wild type
locus. However, if the phenotype was due to the decrease in biotin carboxylase,
the progeny should appear stunted. The progeny of the back cross have been
germinated on kanamycin and were recently transferred to soil. So far, stimted

grth has not been detected in these plants.

161

One other possible explanation for the A11 phenotype is that the p0719A
T-DNA inserted into a regulatory region of some unknown gene causing a
dominant phenotype. To address this possiblity we crossed the homozygous A11
plant with transgenic tobacco expressing a cytosolic ACCase gene which has
been engineered to contain a chloroplast transit peptide. Unlike the heteromeric
plastid ACCase, the cytosolic ACCase is a homodimer of a single
multifunctional polypeptide that contains all of the ACCase subunits. When this
engineered cytosolic ACCase construct was transformed into Brassica napus,
it was correctly targeted to the plastid in an active form (Ohlrogge, personal
communication). If the A1 1 phenotype was due to the decrease in biotin
carboxylase which resulted in decreased ACCase activity, the cross with the
tobacco plant containing the plastid targeted homomeric ACCase should result
in progeny that no longer exhibit the A11 phenotype. We are now in the process
of analyzing the progeny from this cross. So far these plants are not exhibiting
the stunted phenotype that is associated with the A11 plant. Further analysis of
these plants will need to be done to confirm this preliminary result.

In further support for the hypothesis that the stunted phenotype was due
to the antisense expression of biotin carboxylase, the growth of some of the
progeny from the A14 self cross were also stimted when grown under low light
conditions. The original A14 transformant contained approximately 50% of wild
type biotin carboxylase levels in its leaves and grew normally under both low
light and greenhouse conditions. It is possible that the strmted phenotype
observed in the progeny of the A14 self cross is due to co-segragation of multiple
copies of the biotin carboxylase antisense loci. Again as in all of the diagnostic
crosses described above further analysis needs to be done on these plants to

confirm the conclusions made from these preliminary observations.

162
H. Why is stunting observed in plants expressing low levels of biotin
carboxylase only under low light conditions?

Ifthe A11 phenotype proves to be due to the observed decrease in biotin
carboxylase levels, then we need to understand why the phenotype is only
exhibited when plants are grown under low light conditions. Initial experiments
can be done to determine at what light intensity do biotin carboxylase levels
become limiting. We can dertermine this in two ways. First, All plants can be
grown imder different light intensities and patterns of growth can be recorded.
Second, [l-I4C]acetate labelling experiments can be done under different light
intensities on leaf disc isolated from All and wild type plants. The leaf disc will
be taken from All and wild type plants that are grown under high light
conditions to ensure that the tissues are developmentally similar. By comparing
rates of label incorporation into fatty acids, we can determine at what light
intensity biotin carboxylase becomes limiting The next set of experiments will
set out to answer how the limitation in biotin carboxylase effects ACCase
activity. We will attempt to isolate intact ACCase complexes from leaves of A11
plants grown under various light intensities and determine if the subunit
stoichiometry of the ACCase complex is altered in low light relative to high light
grown plants. Because antibodies to each of the ACCase subunits are available
we should be able to access levels of each subunit. However, we need to develop
methods to isolate active tobacco chloroplast ACCase complexes. This has so
far proven difficult and is possibly due the the high level of phenolics in tobacco
leaves. New methods of tobacco chloroplast isolation may be necessary to

address this question.

Literature Cited

Browse J, Somerville C 1991 Glyceroli%id s thesis: Biochemistry and
regulation. Ann Rev Plant P ysro Plant M01 101 2: 467-506.

Guerra DJ, Ohlrogge JB, Frentzen M (1986 Activity of acyl carrier rotein
isoforms in reactions of plant fatty acid metabo ism. Plant Physrol 82: 8-453.

Hlousek-Radojcic A, Post-Beittenmiller D, Ohlrogge JB (1991) E ression
of constitutive and tissue specrfic acyl carrier protein isoforms in Ara idopsis.
Plant PhySiol 98: 206-214.

Jaworski JG, Post-Beittenmiller D, Ohlro e JB (1993) Acetyl-ac 1 carrier
Breatezirlizts (not 39317211“ intermediate in fatty acr brosynthesrs in spinac . FEBS

Jian P, Cronan JE (1994) Inhibition of fatty acid synthesis in Escheriohia coli
in. e abscence of lphospholip7id s§ntheSis and release of mhrbition by
thioesterase action. J acteriol 6:2 14-2821.

Kuo TM, Ohlrogge JB (1984) Acylation of plant acyl carrier proteins by acyl
so 1 (121101613 grotem synthetase from Escherichia coli. Arch Biochem Biophys

Kunst L, Browse J, Somerville C (1988) Altered regulation of 1i id

biosynthesis in a mutant of Arabidopsis deficient in chlorosplast glycero -3-
phosphate acyltransferase actrvrty. Proc Natl Acad Scr USA 5: 41 -4147.

Lamppa G,.Jacks C 1991 Anagsis of two linked genes coding for the ac 1
carrier protein (ACP) om rabi psis thaliana. Plant Mol B10 6: 469-47 .

McIlinney RAJ 1990 The fats of life: the importance and fmetion of protein
acylation. TIBS 5: 3 7-390.

Ohlrogge JBf, Savage L, Jaworski J, Voelker T, Post-Beittenmiller D (1995)
Alteration _o ac l-aczl carrier protein pools and ace l-CoA carboxylase
expressron in Esc ric ia coli by a plant medium chain acy -acyl carrier protein
tlnoesterase. Arch Biochem Biophys 317: 185-190.

Ohlrogge Kuo TM (1984) Plants have isoforms for acyl carrier rotein that
are expresse differently in different tissues. J Biol Chem 260: 803 -8037.

Post-Beitenmiller MA, Hlousek-Radojcic A, Ohlrogge JB (1989) DNA
se uence of a enomic clone encoding an Arabidopszs acyl carrier protein
(A P). Nucleic crds Res. 17: 1777.

Post-Beittenmiller MA, Jaworski JG, Ohlro%e JB (1991)_In vivo pools of

free and acylated acyl carrier Etoteins in grinach: v1dence for srtes of regulation
of fatty ac1d brosynthesrs. J 101 Chem 66: 1858-1865.

163

164

Roesler .KR, Savage LJ, Shintani DK, Shorrosh BS, Ohlrogge JB (1995) Co-
punfication, co-nnmuno .l'eClpltathIl, and coordinate expressron of ace l-CoA
carboxylase actrvrty, bio carboxylase, biotin carboxyl carrier protein 0 higher
plants. Planta 198: (in press)

Runswick MJ, Fearnle IM Skehel JM, Walker JE (1991) Presence of an
acyl carnet protein in ADqubi uinone ox1doreductase from bovine heart
mitochondria. FEBS Lett 286: 121- 24.

SackmannU, Zenson R, Rohlen .D, Jahnke U Weiss H (1991) The acyl
carnerlprotem in Neuros ra crassa 1s a subumt of NADH:ubiquinone reductase
(comp ex I). Eur J Bioc em 200: 463-469.

SchultoU Fecke W Krull _C, Nehls U, Schmiede A, Schneider R, Ohnishi
T WEISS. H. (1994) . In vrvo dissection of the. mitochondrial respirato
gfigfblqmmm ox1doreductase (complex I). Biochim Biophys Acta 118 :

Stocki JW, Lehmann LH, Siege] E 1989 Acylation of groteins by myristic
acid in isolated mitochondria. J 101 em 64: 6376-638 .

Terzaghi WB (1986a) A system for mani ulating the membrane fatty acid
composrtion of sot/bean cell cultures by adding Tween fatty acid ester to their
growth medium. ant Physrol 82: 771-779

Terzaghi WB t1986b) Metabolism of Tween-fatty acid esters by cultured
soybean cells. P ant Physrol 82: 7 80-7 86.

Towler DA, Gordon J], Adams SP, Glaser L (1988 The biolo and
enzymology of eukaryotic protein acylation. Ann Rev Bioc em 57: 69- .

Vassilev AO, Plesofsky-Vig N, Brambl R (1995) C ochrome c oxidase in
Neuros ora crassa contains myristic ac1ds hnked to su unit 1. Proc Natl Acad
Sci US 92: 8680-8684.

Weiss H, Frieddrich T, Holhaus G, Preis D (1991!) The re irato -chain
NADH dehydrogenase (complex I) of mrtochondna. Eur Bichem 97: 5 3-57 6.