$
-.I ~
.1: zz
1» .l?.-.‘v
. v u“..’& f.

 

 

 

.1. . x
..;mo;.h.
4 ,

. . 1..
x I. 2‘!» . IT..9 u.‘.:\
Hr... 92..-... . 1
. .. .

a»

I.

I1:-
.f

-. 5‘
7.7-. l
.u .1.-
..vl. .....~. 7:...1.
‘1‘ »....v,...-.
.u. .7:

. 4:71 .7. .I.

clflhv

r?!“

.19.)
ti .iA

(ii. I .

1“
:VIS...’
. . , ‘ f;!vi5 $2 :7..-
v y : ’.\.|o..l-a
€57.15: 9,3...)

..,..€...... .
.5

o
r|wmdt§ N

1.. $7,334.11.

 

31......
5...»... 3va
. . 21.... ten.
, . . ‘
sax. Ila-{5.2. wow: «\L l‘a
. 5. .. ! I:

MEN!!!

'..l
MICHIGAN STATE

llliflllilllll .llllll‘

RSITY Ll RARIE

 

Till .mmll‘

3 1293 00891 0394

 

 

 

 

 

 

l!

 

 

 

This is to certify that the

dissertation entitled

Molecular and Cellular Characterization of Non-
specific Lipid Transfer Proteins From Spinach
and Arabidopsis

presented by

Sharon Leah Thoma

has been accepted towards fulfillment
of the requirements for

PhD Botany and Plant Pathology

degree in

 

 

    

Major professor

Date i2 / i l (I. Z’"/

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE m nerunu BOX to remove {his checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE‘ DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Adlai/Equal Opportunity Institution
omens-9.1

 

MOLECULAR AND CELLULAR CHARACTERIZATION OF NON-SPECIFIC LIPID
TRANSFER PROTEINS FROM SPINACH AND ARABIDOPSIS

By

Sharon Leah Thoma

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Botany and Plant Pathology

1992

ABSTRACT

Molecular and cellular characterization of non-specific
lipid transfer proteins from spinach and Arabidopsis

By

Sharon Leah Thoma

The mechanism by which lipids move from their main sites of synthesis, the
endoplasmic reticulum (ER) and chloroplast, to other cellular organelles, is
unknown. Lipid transfer proteins (LTPs) are a class of proteins which are capable
of mediating lipid transfer between natural and artificial membranes in vitro.
Despite a lack of experimental evidence, it has been assumed that LTPs carry out
a similar role in vivo. This dissertation describes the investigation into the in viva
function of plant LTPs. Using in vitro transcription and translation in the presence
of microsomal membranes, it was shown that a spinach LTP is cotranslationally
inserted into the ER. To determine the exact location of LTP in plants, antibodies
were raised against an Arabidopsis LTPzprotein A fusion protein which had been
produced in Escherichia coli. Using immunoelectron microscopy, the Arabidopsis
LTP was localized to the cell wall, and was observed predominantly in epidermal
cells. To investigate temporal and spatial expression patterns of UPS, transgenic
plants containing an Arabidopsis LTP promoter-B glucuronidase fusion were

produced. The LTP promoter was active in very specific cell and tissue types, and

in young seedlings was regulated in a developmental manner. The localization of
LTP to the cell wall, and its expression in a specific subset of cell and tissue types
precludes any role that this protein may have in lipid transfer. To elucidate the role
of LTP within a plant, transgenic Arabidopsis containing a greatly reduced level of
LTP due to the presence of an Arabidopsis cDNA in reverse orientation behind a
constitutive promoter, were produced. These plants exhibit no marked phenotype
to indicate-the function of LTP. The predominance of LTP in epidermal cell walls
led to the hypothesis that it may be involved in cuticle formation. However,
analysis of wax and cutin composition in antisense plants showed no alteration in
these compounds. Analysis of the temporal and spatial activity of the Arabidopsis
LTP promoter indicate that the protein may be involved in some aspect of
phenylpropanoid metabolism. The antisense plants provide a system with which

to test this and other hypotheses concerning the function of UPS.

ACKNOWLEDGEMENTS

I thank Chris Somerville for the opportunity to work in his laboratory, for his
insights into my project and science in general, and for teaching me to become
an independent researcher. I also thank the members of my committee, Pam
Green, Karen Klomparens, John Ohlrogge, and Mike Thomashow for their
guidance and helpful suggestions during the course of this project. Also, thanks

to Natasha Raikhel for the support and encouragement she has given me.

I am grateful to all the past and present members of the Somerville laboratory for
the helpful advice and for providing me with countless amusing stories to recount.
I especially thank John Shanklin and Clint Chapple for their technical help and for

all the discussions they endured about LTP.

Special thanks goes to Sarah Zill, John Shanklin, Clint Chapple, Antje Heese, Scott
Peck, and Theresa and Sebastian Bednarek for their friendship and for many
memorable times. Also, I thank my parents, who did not always understand what

I did, or why I did it, but were always supportive.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................. vii
LIST OF FIGURES .............................................................................................. viii
CHAPTER 1: Introduction .................................................................................. 1

CHAPTER 2: Isolation of a cDNA clone for a spinach
lipid transfer protein and evidence that the protein

is synthesized by the secretory pathway .......................................................... 24
ABSTRACT ........................................................................................................... 25
INTRODUCTION .................................................................................................. 26
EXPERIMENTAL PROCEDURES ....................................................................... 27
RESULTS .............................................................................................................. 32
DISCUSSION ........................................................................................................ 43
REFERENCES ...................................................................................................... 46
CHAPTER 3: An Arabidopsis non-specific lipid transfer

protein is a cell wall protein ................................................................................ 49
ABSTRACT ........................................................................................................... 50
INTRODUCTION .................................................................................................. 51
EXPERIMENTAL PROCEDURES ....................................................................... 53
RESULTS .............................................................................................................. 58
DISCUSSION ....................................................................................................... 71
REFERENCES ..................................................................................................... 77

V

CHAPTER 4: An Arabidopsis lipid transfer protein
promoter specifies complex expression In transgenic

plants ..................................................................................................................... 80
ABSTRACT ............................................................................................................ 81

INTRODUCTION ................................................................................................... 82
EXPERIMENTAL PROCEDURES ........................................................................ 85
RESULTS ............................................................................................................... 90
DISCUSSION ....................................................................................................... 1 17
REFERENCES ..................................................................................................... 126

CHAPTER 5: Analysis of transgenic plants Arabidopsis
containing a reduced level of a non-specific lipid

transfer protein ................................................................................................... 131

ABSTRACT .......................................................................................................... 132
INTRODUCTION ................................................................................................. 133
EXPERIMENTAL PROCEDURES ...................................................................... 137
RESULTS ............................................................................................................. 142
DISCUSSION ....................................................................................................... 1 63
REFERENCES ..................................................................................................... 168
CHAPTER 6: Summary and conclusion .......................................................... 172

vi

LIST OF TABLES

Table 1-1. Biochemical properties of plant non-specific

lipid transfer proteins ............................................................................... 8
Table 4—1. Primer sequences used to amplify promoter

fragments from ALTP1 ........................................................................... 89
Table 4-2. Transgenic lines used for histochemical

analysis of ALTP1 activity ...................................................................... 100
Table 5-1. Antisense lines tested by Western analysis ................................ 147

vii

LIST OF FIGURES

Figure 1-1. Comparison of the amino acid sequences of plant
non-specific lipid transfer proteins ......................................................... 13

Figure 2-1. Composite nucleotide sequence of the cDNA for
spinach lipid transfer protein and the deduced amino
acid sequence .......................................................................................... 34

Figure 2-2. Nonhem blot of total RNA from spinach probed
with the cDNA of spinach lipid transfer protein .................................... 36

Figure 2-3. Southern blot of total spinach DNA probed with
the cDNA of spinach lipid transfer protein ............................................ 38

Figure 2-4. Processing of in vitro translation products by
canine microsomal membranes ............................................................. 40

Figure 2-5. Processing of in vitro translation products by
microsomal membranes from maize endosperm
cultures ...................................................................................................... 41

Figure 2-6. Protease sensitivity of translation products ................................. 42

Figure 3-1. Structure of plasmids used to produce LTP
fusion proteins .......................................................................................... 59

Figure 3-2. SDS-PAGE of purified LTP-protein A fusion
from E. coli. ............................................................................................... 61

Figure 3-3. Western blot analysis of an Arabidopsis leaf
protein extract produced with anti-LTPzprotein A
antibodies ................................................................................................. 62

Figure 3-4. Tissue distribution of Arabidopsis LTP ......................................... 63

Figure 3-5. Immunocytochemical localization of LTP in the
cell wall of Arabidopsis leaf tip, stem, and petiole
using immunogold labeling ..................................................................... 66
viii

Figure 3-6. Immunocytochemical localization of LTP in
Arabidopsis guard cells, stem cortical cells, and
xylem vessels using immunogold labeling ............................................ 68

Figure 3-7. Immunocytochemical localization of LTP in

Arabidopsis root and pistil using immunogold labeling ...................... 70
Figure 4-1. Southern blot of Arabidopsis genomic DNA

probed with an Arabidopsis LTP cDNA ............................................... 92
Figure 4-2. Sequence of ALTP1 promoter and coding region .................... 94

Figure 43- Comparison of putative regualtory
sequences in ALTP1 and in genes of the general

phenylpropanoid pathway ...................................................................... 96
Figure 4-4. ALTP1-B-glucuronidase constructs used to

produce transgenic plants ..................................................................... 98
Figure 4-5. Transgenic Arabidopsis seedlings after X-Gluc

staining at different developmental stages .......................................... 102
Figure 4-6. Longitudinal section through shoot meristem

of 14 day old seedling ........................................................................... 103
Figure 4-7. ALTP1 activity in root tissue ....................................................... 105
Figure 4-8. ALTP1 activity in leaf and stem tissue ........................................ 108
Figure 4-9. ALTP1 activity in flowers ............................................................... 110
Figure 4-10. ALTP1 activity in siliques ............................................................ 111

Figure 4-11. Activity of CaMV 35$ promoter in transgenic
Arabidopsis .............................................................................................. 114

Figure 4-12. ALTP1 activity in lines containing 1 Kb,
700 bp, and 280 bp promoter fragments ............................................ 116

Figure 5-1. Sequence analysis of the Arabidopsis
cDNA clone ............................................................................................. 143

Figure 5-2. Comparison of the Arabidopsis LTP
sequence to LTP consensus sequence ............................................... 144

ix

Figure 5-3. Structure of constructs used to produce
transgenic lines ....................................................................................... 146

Figure 54. Western analysis comparing levels of LTP
in wild type and transgenic plants ........................................................ 149

Figure 5-5. Growth of wild type and antisense plants .................................. 151

Figure 56. Control and antisense plants at 15 and 25
days after planting .................................................................................. 154

Figure 5-7. Time of flowering in control and antisense
plants ........................................................................................................ 155

Figure 58. Percent germination of antisense and
control plants ........................................................................................... 156

Figure 59. Gas chromatography analysis of wax
composition ............................................................................................. 157

Figure 5-10. Gas chromatography analysis of cutin
compostion .............................................................................................. 158

Figure 5—11. Scanning electron microscopy analysis
of epicuticular wax of wild type and antisense
plants ........................................................................................................ 160

Figure 5-12. Transmission electron microscopy
analysis of the wax and cutin of wild type
and antisense stems .............................................................................. 162

CHAPTER 1

INTRODUCTION

2

Phospholipids are the major lipid component of most eukaryotic cell
membranes. It has been estimated that eukaryotes may contain as many as 1000
chemically distinct phospholipids (Raetz, 1982) and the metabolic regulation and
biological significance of the large number of lipid species is not well understood.
In eukaryotes, most enzymes responsible for lipid synthesis are on the cytoplasmic
face of the. endoplasmic reticulum (ER) (Alberts, et at, 1989) and all plants also
synthesize lipids in the chloroplast. Although lipid biosynthesis is restricted to a
few organelles within a cell, lipids are found in all membrane systems of the cell,
with different organelles containing specific and unique lipid compositions as well
as an asymmetric distribution of lipid species across the bilayer.

The maintenance of a complex intracellular distribution of lipids must involve
specific lipid transport mechanisms. However, the identification of such a transport
mechanism in living cells has been difficult. There are three likely general
mechanisms for lipid movement: (1) transport of molecules from one organelle to
another by vesicle budding and fusion, (2) transport of molecules by lateral
diffusion between organelles connected by membrane bridges, and (3)
spontaneous or protein-mediated transport of lipid monomers through the cytosol.
There has been evidence that vesicular transport is important in delivering
phospholipids from their site of synthesis to the plasma membrane (Mills, of al.,
1984). However, this process has not been demonstrated to be important for
delivering lipids to other membrane systems within a cell, so it is possible that

other mechanisms are operational.

other mechanisms are operational.

Phospholipids rapidly migrate in the plane of the bilayer of biological
membranes and in model membranes in the liquid crystalline state. By this
process, phospholipids could move between organelles via transient or permanent
interconnections between membranes. Associations between the ER and the
chloroplasts have been observed in many plants including Equisetum telmateia
(McLean, et al., 1988), Acer pseudoplantanus and Pinus pinea (Wooding and
Northcote, 1965), Phaseolus vulgaris (Whatley, at al., 1991 ). Continuities have also
been observed between the ER and other organelles, including the nucleus, the
Golgi apparatus, the plasma membrane, and the outer mitochondrial membrane
in the fern gametophyte of Pteris vittata L. (Crotty and Ledbetter, 1973).

Spontaneous transfer of native phospholipids does occur within a cell, but
it occurs at rates too slow to account for the rate of membrane biogenesis within
a cell (McLean and Phillips, 1982; ertz, 1982). Also spontaneous transfer is non-
selective and would lead to a randomization of lipid components in the
membranes. Protein mediated lipid transfer would require that the rate of
membrane biogenesis within a cell is equal to the rate of diffusion of the lipid-
protein complex through the cytosol. This, in fact, has been shown to be the case
for the formation of certain organellar membranes. The specific and unique lipid
composition of different cellular membranes could be maintained by this
mechanism, as there could be a specificity as to which lipids particular proteins

could bind and transport.

4

Protein mediated transport of phospholipids between membranes was first
observed in 1968 by ert2 and Zilversmit. The low level of exchange seen when
rat liver microsomes and mitochondria were incubated together was significantly
increased when a 105,000 9 liver supernatant was added to the preparation.
Subsequently, liver and beef heart preparations were shown to contain a factor
which stimulated phospholipid exchange. This transport factor (LTP) was first
partially purified from beef heart (ert2 and Zilversmit, 1970). LTPs have since
been purified from other mammalian tissues, yeast, bacteria, and plant tissue.
Based on their ability to catalyze lipid transport in vitra, it has been suggested,

despite the lack of experimental evidence, that LTPs carry out a similar role in viva.

Mammalian tissue contains at least three different classes of UPS. One
class, purified initially from rat and bovine liver (Kamp, at al., 1973; Lumb, et al.,
1976), is highly specific for phosphatidylcholine (PC) and are called PC—transfer
proteins (PC-TPs). The second class of proteins, the phosphatidyllnositol (Pl)
transfer proteins (Pl-TPs), isolated originally from bovine brain and heart (DiCorleto,
at al., 1979; Helmkamp, at al., 1974), has a preference to bind Pl, but can also
transfer PC. An LTP purified from yeast (Szolderits, at al., 1989) has a dual
specificity closely resembling that of mammalian Pl-TPs. However, the mammalian
and yeast Pl-TPs share no sequence similarity with respect to primary structure
(Dickeson, at al., 1989; Bankaitis, et al., 1990). A phosphatidylserine (PS) transfer

protein (PS-TP) has also been isolated from yeast (Paltauf and Daum, 1992). The

5
PS-TP can also transfer PE, cardiolipin, phosphatidic acid and ergosterol, but not

PC or Pl. No mammalian counterpart to the PS-TP has been identified. The third
protein class, first purified from bovine and rat liver (Bloj and Zilversmit, 1977; Grain
and Zilversmit, 1980), transfers a large variety of phospholipids, as well as
cholesterol and neutral glycosphingolipids. These proteins are designated non-
specific Iipid transfer proteins (nsLTPs). The nsLTP from mammalian tissue has
been shown to be identical to a protein which was previously designated steral
carrier protein 2 (Scallen, at al., 1985a). This protein appears to function in several
steps of cholesterol metabolism (Scallen, stat, 1985b). A nsLTP has been purified
from the bacteria, Fihadapseudamanas sphaeraides (T ai and Kaplan, 1985), from
the filamentous fungus Candida trapicalis (Tan, at al., 1990), and nsLTPs have

been purified to homogeneity from several plant species.

Assay for lipid transfer activity

Lipid transfer activity is determined by measuring the stimulation of
phospholipid transfer between natural membranes, such as microsomes,
chloroplasts, or mitochondria, and artificially prepared, membranes, such as
liposomes. One membrane, the donor membrane, is prepared to contain a
radiolabeled lipid (such as [am-phosphatidylcholine), and is then incubated with
the non-labeled, acceptor membrane. After incubation of the membrane mixture
with LTP, the donor and acceptor membranes are separated, and the appearance

and increase of label in the acceptor membrane is analyzed. As it is rarely

6

possible to quantitatively separate donor and acceptor membranes, donor
membranes are additionally labeled with a nontransferable tracer molecule (such
as [“Cl-cholesteryl oleate). Such a tracer can monitor for cross contamination or
incomplete recovery of membranes, and determination of the 3H/“C ratio in
acceptor membranes then reflects the rate of lipid transfer.

Other, less common assays, involve the use of uni- and multilamellar
vesicles which are separable by centrifugation, or the incubation of LTP with
liposomes containing spin labeled lipids, whose movement can be followed by

ESR spectroscopy (Nishida and Yamada, 1985).

Exchange versus not transport

The observation that phospholipid transport in vitra can be catalyzed by a
protein raises the question of whether the movement takes place in both directions
(exchange) or in one direction (net transport). There have been several reports
in which an attempt has been made to elucidate the nature of this protein
mediated transport, and the results have been somewhat disparate. In the
presence of bovine PC-TP, net transfer of PC was observed to acceptor vesicles
prepared from PE and phosphatidate (Wirtz, at al, 1980). However, the
concentration of LTP used in this experiment was much higher than physiological
levels and when lower concentrations were used, only PC exchange was observed
(Helmkamp, 1980). Grain and Zilversmit (1980) demonstrated that a bovine liver

nsltp catalyzed net transfer of PC and PI from PC/Pl multilamellar vesicles to intact

7

or delipidated human high density lipoproteins. Under identical conditions,
however, it was observed that a bovine PC-TP and a bovine heart Pl/PC-TP

catalyzed exchange of PC and PI.

Purification of non-specific llpld W prom from plants

It was shown by Kader (1975) that soluble extracts from potato tubers
contained a protein component which enhanced the movement of lipids between
microsomal fractions and mitochondria isolated from cauliflower and potato tuber.
Since that time, LTPs have been purified to homogeneity from maize seedlings
(Douady, at al, 1982), spinach leaves (Kader, at al., 1984), and castor bean
seedlings (T akishima, at al., 1986; Watanbe, at al, 1986), and are quite well
characterized at the biochemical level. A barley protein, first classified as a
probable amylase/proteinase inhibitor (Mundy and Rogers, 1986), was
subsequently identified as an LTP (Bernhard and Somerville, 1989; Breu, et al,

1989).

Biochemical properties

Spedfidty afpbntlpid Iransferpratehe As previously suggested, all UPS
which have been isolated and extensively characterized from plants are non-
specific; i.e., they transfer a large variety of phospholipids in vitra assays (Table
1-1). The maize and barley LTPs have been shown to transfer PC, PI, and

phosphatidylethanolamine (PE) (Douday, et al., 1985; Kader, 1985), and the

Nmmw

3:89.. 395209.. 5.8a 29. same a 288m 9:.
ESSEsEmossBocoe docs. "agnosiafimofi .E ”esozcfiafimofi on.

 

 

 

 

 

 

 

 

 

assassin am 822,

. «8. 42m
, .coEasoWooth ad mad ofiEo.r

82 .s. s
.92": ”32 8.8 25.08: as ca 52%

.a. a gene. .oe.mn_._a.oa

$2

an... .688 maids. a.» 3 mans.
52 80.2.0“.
out 68¢ Mazda. 2: am .658
82 :89. maid“. - 2 Sam
I as. - - sew- ,..-.,..,>,__ol___lowmml I line“. .1 l I liar“: - I I 8.8»
22.83 88208. 538.22

w2_w._.0mn. mmuwZé DE... 0.n__0waw.zoz ._.Z<._n_ “.0 memmmOmm .._<O__zwIOO.m

655982555885 .me ”8.225558% .8 Baa.§£<

9

proteins isolated from spinach and caster have been shown to transfer PC, Pl, PE,
phoshphatidylglycerol (PG), and monogalactosyldiacylglycerol (MGDG) (Kader, et
al., 1984; Tchang, at al., 1988; Tsuboi, at al., 1991). Four isoforms of LTP, A, B,
C, and D have been isolated from castor bean seedlings, and the substrate
specificities are similar among isoforms (T suboi, at al., 1991).

Besides transferring lipid molecules, it has been demonstrated that lipid transfer
proteins can bind fatty acids. Rickers, at al. (1984) isolated a fatty acid binding
protein from onion leaves and it was subsequently shown that this protein had the
ability to transfer PC in an in vitra assay (Arondel, at al., 1989). It was later
demonstrated that the spinach LTP has the ability to bind long chain fatty acids
(oleic and linoleic) and oleoyl-CoA (Rickers, et al., 1985). The castor bean isoform
nsltp-A also has a binding capacity for oleic acid and oleoyl CoA (Yamada, et al.,

1990)

Molecuhr mass The apparent molecular mass of plant lipid transfer
proteins, as determined by SDS-polyacrylamide gel electrophoresis and gel
filtration is approximately 9 kDa (Table 1-1). These values correspond well with the
molecular masses determined from protein sequences (Bouillon, at al., 1987) and
are close to those established for nsltps from animal sources, which range from
11.2 to 14.5 kDa (Helmkamp, 1990). A dimeric form of maize LTP (M, 18,000 1
1000) has been observed when electrophoresis is carried out under non-reducing

conditions (Douady, at al., 1985). Specific LTPs from mammalian and yeast tissue

10
are larger, ranging from 25-35 Kda (Helmkamp, 1990).

lsaelectric poht Determination of Pi by chromatofocusing or isoelectric
focusing shows that plant nsLTPs are basic (Table 1-1). Maize, spinach, and
castor proteins have pls of 8.8, 9.0, and 10.5, respectively. A 10:3 proportion of
basic residues versus acidic residues gives a tomato LTP a calculated isoelectric
point value of 8.85 (T ones-Schumann, at al., 1992). Acidic proteins (pl of about
4.5) have also been detected in spinach leaf and castor bean seedlings (T anaka

and Yamada, 1979), but these proteins have not been well characterized.

awry Plant nsLTPs are very stable proteins. The castor protein isolated
by Watanbe and Yamada (1986) retained transfer activity after storage for one
month at 4°C, or 30 minutes at 90°C. Similarly, the maize nsLTP is stable for one
month at 4°C and it retains 50% of its transfer activity after 30 minutes at 95°C

(Arondel and Kader, 1990).

Abmdanaa Although specific LTPs from mammalian sources are of
relatively low abundance (0.05% - 0.25%) (Helmkamp, 1990), plant nsLTPs have
been reported to be highly abundant. Using immunological techniques, Grosbois
at al. (1987) determined that maize LTP constituted 4.1% of the total soluble
protein in crude extracts of maize seedlings and it has been reported that LTP

accounts for 10% of the total soluble protein at certain stages of maize seedling

11

development (Grosbois, at al., 1989). The Arabidopsis LTP, however, is not so
abundant as it constitutes no greater than 0.1% of the total soluble protein in crude

leaf extracts (Thoma, S., unpublished data).

Protein structure

Amino acid sequence for the nsLTPs from maize seedlings (Dubois, et al.,
1987; Tchang, et al., 1988), spinach leaves (Bouillon, at al., 1 987; Bernhard, at al.,
1991), castor bean seedlings (T akishima, at al., 1986), barley caryopses
(Svensson, at al., 1986), sunflower (Arondel, at al., 1990), carrot (Sterk, at al.,
1991) and wheat (Dieryck, at al., 1992) have been determined by Edman
degradation of the purified protein or by analysis of the cDNA corresponding to
the protein. These plant LTPs, while lacking any obvious sequence homology to
LTPs from yeast, bacteria, or mammalian cells, share a high amino acid similarity
among themselves (Fig. 1-1). This is especially noticeable in the central portion
of the protein, where hydropathy profiles indicate the presence of a deep
hydrophilic region surrounded by two hydrophobic regions (Arondel and Kader,
1990; Yamada, 1992). It has been hypothesized that the hydrophobic regions are
involved in binding to the acyl chains of a lipid molecule, and the hydrophilic
region may associate with the polar region of the lipid molecule (T akishima, et al.,
1988). These proteins also exhibit a hydrophobic N-termlnal end and it has been
suggested that this region of the protein may be important for lipid-membrane

interactions (Arondel and Kader, 1990). Secondary structure analysis of a barley

12

Figure 1-1. Comparison of the amino acid sequences of plant non-specific lipid
transfer proteins. Amino acids which are identical to the consensus sequence are
indicated by a shaded gray area 6%). Conserved substitutions are represented by
the single letter amino acid designation which is highlighted with gray.

13

AAAA.AAAVA
"333635333
SARAAAASAA
TNTYRCRTR.
RPNPPPPAPP
STAAAAAAVA
RTLAVAAAAA
GAC.VVTLA.
CLFVVVLLLL
DVLVHHHVHV
RHRICCLLLL
LYSCLL.A.e
GHLLLVVVVV
LVLLVVLVLL
AHLVP!.A.¢
LHLACCATAA
HAVCAAAAVA
VSAIIAVAA

PFLRRHVH.

SVRSGLALQ

SVIVALLHL

RNVHIVQVV

LRABRQQQQ

VHSHHATAe

G 8 RR R

H A AA A

I “I H

Anmmmuo
ammt
finno
Tobacco
many!
Emu
mum
(human

swmwh

aware

A  8SA af_.viTP

c
aig.;:flf],lflfic
.TSTAA...BGSO.
ALALQA...GAAG.
PPJfiiTGRGPGTKS

“CVLWCKGGGBRQQP

IPSTRAGLLIGAGa
VVATSTLPP.8AP9

.GNVDDflPRTPGTTG
“FVRPK

hi..m AR
u;VA TR 0 G

09, n 9 De
Lurrrxxnooan..
fla-§fiiifi.PASVL

RRR<LTTTT._Y
GGPPGRGAPTSABQ
ta I!V$§aik£§iL
.. :Cu.:- C

8 G.TTRTGRA

.X}.IxajTau££VL

UABARSRRGRAAL.

~76 RH;I,Af$f:S

"TUNDASTTDASDe
,,..Fifi71515.v
spiprflngfifffiao

wMGT7SinU_TiL,G

;G,TR$EUT
LATfiEVRRI
.G.Lifif.A

EwDRP.
gfigggw
f¥2.A

Mmmwmua
Cumn
aumrh
aware
aware
awmrb
Spinach
Tobacco
TuMm
mmuyr
mm"
mm.
Hm“
(”mama

Q.‘<T.TI.QERGR
it Ye 33.3L

. s:
..z _ .JC
. G BRQ.T T UN
. . -,,..p fC

...,.Afl g T fl§%.fA

P
§¥RY§w igficl
AAARSGRRAGGR.¢
RNaARNNGRRASRe
.;€Ts,fi..a.t»a
.TA 3 .....
GQ TRSSSSGNSGQ

QTRIATTITTAAS.
QR RRQRQRQRRQ.
me RR Rngb 7R

PfiiXD PE G .SA
, VV .4§m_OT
.,TP . LWBM $.T

RR  PRRARLQRRRe
R 3:. i¢.y[A

suxkosancuuhln
, RA nVLn .3"
ta :TfJAL

NGRBQGARNDSSR.
CRQHQR L:IRR R
.w, £GgwfifiTle
TV?G.ITTR£GRTG
SNARTNGGGNSSDQ

.uunmmu.
flunk
uwmmh
muuma
auumc
auumb
Spinach
Tobacco
hmno
away:
hue
my.
mum
immuuu

AR

C
NDNNRNNDDDDDD.
. T:IIT
xgfifiifflfn RS
T.RRT:L. T.TALP
m3 .flqf.l..8
. . yii....a I
RRPPPRARRTTAT.

IverRfi..ZKY
S NRGMLFQ-.GTP
.-XR¥~ y fig!
Dvx 3 SR.
Iffim V,

N

,: ..,, V
”maRR ._G
afANHTV..RyR
RARYAGGSSSSGP.
; .pflmGflfll,fl.P
§§§7§§§G.flnf1
GGSSRSGGGSSSSQ
588! 8RA
_._§§  u:fl,F.A
RAAAPLRRRNNRR.
GNDDQRGGRRGGD.
ALQRDPYLLLAARe
xxxfifanugfign
..§$¥a§¥$T¢Y:L
.3TNflgcfiffjus
8.PP.PR8RN..Re
GT.......BSSI.

nmummna
flunk
mwmrh
aware
auumc
gamma
Spinach
Tobacco
Rmmo
Emmy!
mun.
hwi
mum
Cmuuuml

R
V
T
A
L
B
D N
SVJLRRRRQTxS
Tfifgfivi W

1.ABT8AKT;
RNQBBQK

SRVN

uuummua
Carrot
unmrh
mumra
unwrc
uner
summh
Tobacco
Rmno
mmuyr
mums
Hm“
lhmuuml

Rmi

14
LTP (Madrid and von Wettstein, 1991), indicates that the protein has three

amphiphilic regions which could form a pocket for the binding of
phosphoglycerides, and it is proposed that the hydrophilic surface of the protein
allows for the solubility of the lipid-protein complex in the cytosol. All plant nsLTPs
characterized to date contain 8 cysteine residues, and the positions of these
residues are completely conserved. It has been demonstrated that the cysteine

residues in the castor bean protein form four disulphide bridges (T akishima, et al.,

1988).

Occurrence of leofonns

There have been reports of multiple isoforms of UPS in a variety of plant
species. Four isoforms of LTP have been isolated from castor bean seedlings,
and they have been shown to be differentially expressed in a tissue specific
manner (T suboi, et al., 1991). Two peaks of activity observed during purification
of maize LTP suggest the presence of at least two isoforms in this plant (Arondel,
et ‘31.. 1990). This result was confirmed by Southern analysis of maize genomic
DNA which indicates the presence of several maize LTP genes (T chang. et al.,
1989). Purification of wheat LTP led to the isolation of three proteins which had
PC transfer activity in vitro, two with a molecular mass of 9 k0 and the other 7 k0
(Dieryck, et al., 1992). The presence of isoforms in spinach tissue was suggested
by the observation that the protein separated into two fractions upon purification

(Kader, er al., 1984). However, high stringency Southern analysis indicates the

15

presence of only one gene (Bernhard, et al., 1991; chapter 2 of this dissertation)
suggesting that the isoforms are not closely related at the amino acid level.
Southern analysis of tomato DNA under high stringency conditions was consistent
with the presence of only one gene, but low stringency hybridization indicates the
presence of several LTP-like genes (T ones-Schumann, et al., 1992). Southern
analysis shows the presence of a single gene in the carrot genome (Sterk, et al.,
1 991). Two different LTPs have been found in tobacco. The first was isolated as
an anther specific protein is found to be expressed predominantly in tapetal cells
(Koltunow, et al., 1990) and the other is expressed in all aerial tissues of the plant,
but was not expressed in tapetal cells (Fleming, et al., 1992). Barley also encodes
at least two LTPs. The first was originally identified as an amylase/proteinase
inhibitor based on sequence homology (Mundy and Rogers, 1986) and the other

has more recently been identified (R. Kalla, personal communication).

Localization of plant nsLTPs

Because of their proposed function of intracellular lipid transfer, it has long
been assumed that LTPs were located in the cytosol (Arondel and Kader, 1990:
Helmkamp, 1990). However, recent evidence indicates that nsLTPs are not
located in the cytoplasm. it has been reported that the maize LTP has an amino
terminal extension and is synthesized on membrane-bound polysomes (T chang.
et al., 1988; Vergnolle, et al., 1988) and that spinach and barley LTPs contain

signal peptides and are cotranslationally inserted into microsomal membranes

16
(Bernhard, at al., 1991; chapter 2 of this dissertation; Madrid, 1991). Since these

proteins lack the endoplasmic retention signal, KDEL (Munro and Pelham, 1987),
they would be expected to be secreted or targeted to a specific organelle. In
support of these expectations, Mundy and Rogers (1986) reported that the barley
protein which was subsequently identified as an LTP (Breu, et al., 1989) was found
in aleurone cell culture medium and Stark, et al. (1991) demonstrated that a carrot
LTP was secreted by embryogenic cell cultures. Using immunogold labeling and
light microscopy, Sossountzov, et al. (1991), have shown that a maize LTP is
localized to the periphery of the epidermal cells in maize coleoptile and suggest
that the labeling is associated with the plasma membrane. It has more recently
been demonstrated (Thoma, et al., 1993; chapter 3 of this dissertation), using
immunocytochemical labeling at the ultrastructural level, that an Arabidopsis LTP

is localized to the cell wall, and that it is mainly seen in epidermal cell walls.

Scope of dissertation

Although LTPs have a demonstrated ability to catalyze lipid transfer between
membranes in vitro, it has been difficult to assign an in viva function to these
proteins. A yeast Pl-TP is the only LTP to which the in vitro activity correlates to
a function in viva (Bankaitis, et al., 1990). It appears that this protein is involved
in elevating Pl/PC ratios of yeast Golgi membranes by an unknown mechanism
(Cleves, et al., 1991). The nsLTPs from plants, however, lack sequence homology

to the yeast Pl-TP and are likely to have a different biological role.

17

The original objective of the research described in this dissertation was to
examine the possible role of nsLTPs in membrane biogenesis. As a first step, a
cDNA corresponding to a spinach nsLTP was cloned. Analysis of this clone
revealed the presence of a signal peptide. and was subsequently shown to be
cotranslationally inserted into microsomal membranes (Chapter 2). These
observations raised questions about the localization of LTP and thus, about the
role of LTP within the cell. To address these questions, antibodies raised against
a protein AzArabidopsis LTP fusion protein were used to localize the protein by
immunological techniques at the ultrastructural level (Chapter 3) and transgenic
plants containing an Arabidopsis LTP promoter - B—glucuronidase fusion were
analyzed (Chapter 4). Finally, in an attempt to determine the role of this protein
within a plant, Arabidopsis plants containing a very reduced level of LTP due to the
presence of the Arabidopsis LTP cDNA in reverse orientation behind the
Cauliflower Mosaic Virus (CaMV) 35$ promoter were produced and analyzed
(Chapter 5). A summary of the conclusions of this work and suggestions for future

research are discussed in chapter 6.

18
REFERENCES

Alberta, 3., Bray, D., Lewis. J., Raff, M., Roberts. K. and Watson. J. 1990. Molecular
biology of the cell. Garland Publishing, Inc., New York.

Arondel, V., Tchang, P., Vergnofle, C., Jollot, A., Grosbois, M., Guerbette, P.,
Moroh, M.-D., Pemollet, J.-C., Deleony, M., Pulgdomemeoh, P. and Kader, J.-C.
1989. Plant lipid-transfer proteins: structure, function and molecular biology. In:
Biological role of plant lipids, P.A. Biacs, K Gruiz, and T. Kremmers, eds. Plenum
Publishing, New York, pp. 341-350.

Arondel, V. and Kader, J.-C. 1990. Lipid transfer in plants. Experientia 46:579-585.

Bankaitis, V.A., Altlten, J.R., Cloves, A.E. and Dowhan, W. 1990. An essential role
for a phospholipid transfer protein in yeast Golgi function. Nature 347:561-562.

Bernhard, W3. and Somervllle, CR. 1989. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from
amino acid sequence homology. Arch. Biochem. Biophys. 269:695-697.

Bemhard, W.R., 'l'l'loma. 3., Hotels. J. Md Somervle, C.Fl. 1991. Isolation of a
cDNA clone for spinach lipid transfer protein and evidence that the protein is
synthesized by the secretory pathway. Plant Physiol. 95:164—170.

Bio], B. and Zlversmlt. 0.8. 1977. Flat liver proteins capable of transferring
phosphatidyethanolamine. Purification and transfer activity of other phospholipids
and cholesterol. J. Biol. Chem. 252:1613-1619.

Boulllon, P., Drleohel, C., Vergnolle, C., Duranton, H. and Kader, J.-C. 1987. The
primary structure of spinach-leaf phospholipid-transfer protein. Eur. J. Biochem.
1 66:387-391 .

Breu, V., Guerbelte, P., Kader, J.-C., Gamlnl Kernangara, C., Svenuon, B. and
von Wettsteln Knowles, P. 1989. A 10 kD barley basic protein transfers

phosphatidylcholine from liposomes to mitochondria. Carlsberg Res. Commun.
54:81 -84.

Cloves, A.E., McGee, T.P., Whitters, EA, Chanplon, KM, Altken, J.Fl., Dowhan,
W-. Goebel, M. and Bankaltls, VA. 1991. Mutations in the CDP-choline pathway for
phospholipid biosynthesis bypass the requirement for an essential phospholipid
transfer protein. Cell 64:789-800.

19

Crab, RC. and Zlveramlt. DB. 1980. Two non specific phospholipid exchange
proteins from beef liver. 1. Purification and characterization. Biochemistry 1921433-
1439.

Crotly, WJ. md Wetter, MC. 1973. Membrane continuities involving
chloroplasts and other organelles in plant cells. Science 182:839-841.

Dickson, S.K., le, C.N., Schuyler, G.T.. Dalton, T.P., Helmkamp, GM and
Yarbrough, LR. 1989. Isolation and sequence of cDNA clones encoding rat
phosphatidyllnositol transfer protein. J. Biol. Chem. 264:16557-16564.

DlCorleto, ' P.E., Wench, J.B. and 211mm, 0.8. 1979. Purification and
characterization of two phospholipid exchange proteins from bovine heart. J. Biol.

Chem. 254:7795-7802.

Dleryok, W.. Gautier, M.-F., Lulllon, V. ltd Joucl'ler, P. 1992. Nucleotide sequence
of a cDNA encoding a lipid transfer protein from wheat (Triu‘cum durum Desf.)
Plant Mol. Biol. 19:707-709.

Dandy, D., Groobole, M., W, F. and Kader, J.~C. 1982. Purification of a
basic phospholipid transfer protein from maize seedlings. Biochim. Biophys. Acta
710:143-153.

Douady, D., Guerbette, F., Grosbois, M. and Kader, J.-C. 1985. Purification of
phospholipid transfer protein from maize seeds using a two-step chromatographic
procedure. Physiol. veg. 23:373«380.

Flenhg, A.J., Mandel, T. Holman, 8., sum. P., de Vrlec, S.C. and Kuhlemeler, C.
1 992. Expression pattern of a tobacco lipid transfer protein gene within the shoot
apex. Plant J., in press.

Grosbois, M., Guerbotte, F., Dandy, D. and Kader, J.-C. 1987. Enzyme
immunoassay of a plant phospholipid transfer protein. Biochim. Biophys. Acta
917:162-168.

Grosbois, M., Guerbatto, F. and Kader, J.-C. 1989. Changes in the level and
activity of phospholipid transfer protein during maturation and germination of maize
seeds. Plant Physiol. 90:1560-1564.

Helmkfl'np, GM. 1980. Concerning the mechanism of action of bovine liver

phospholipid exchange protein: exchange or net transfer. Biochem. Biophys. Res.
Commun. 97:1091-1096.

20

Helmkamp, GM. 1990. Transport and metabolism of phosphatidylinositol in
eukaryotic cells. In: Subcellular biochemistry, H.J. Hilderson, ed., Plenum
Publishing Corp., pp.129—174.

Helmkamp, G.M., Harvey. M., erlz, KWA and van Deenen, LLM. 1974.
Phospholipid exchange between membranes. Purification of bovine brain proteins
that preferentially catalyze the transfer of phosphatidylinositol. J. Biol. Chem.
249:6382-6389.

Kader, J.-C. 1975. Proteins and the intracellular exchange of lipids. l. Stimulation
of phospholipid exchange between mitochondria and microsomal fractions by
proteins isolated from potato tuber. Biochim. Biophys. Acta 380:31-44.

Kader, J.-C. 1985. Lipid binding proteins in plants. Chem. Phys. Lipids 38:51-62.

Kader, J.-C., Julienne, M. and Vergnolle, C. 1984. Purification a nd characterization
of a spinach-leaf protein capable of transferring phospholipids from liposomes to

mitochondria or chloroplasts. Eur. J. Biochem. 139:411-416.

Kemp, H.H., erlz, KWA and Van Deenen, LLM. 1973. Some properties of
phosphatidyl exchange protein purified from beef liver. Biochim. Biophys. Acta
318:313-325.

Koltunow, AM, Truettner, J., Cox, KH., Wallroth, M. and Goldberg, RB. 1990.
Different temporal and spatial gene expression patterns occur during anther
development. Plant Cell 2:1201-1224.

Lumb, H.H., Kloosten'nen, A.D., ertz, KWA and Van Deenen, LLM. 1976. Some
properties of phospholipid exchange proteins from rat liver. Eur. J. Biochem.
69:15-22.

Madrid, S. 1991. The barley lipid transfer protein is targeted into the lumen of the
endoplasmic reticulum. Plant Physiol. Biochem. 29:695-703.

Madrid, SM. and von Weltsteln, D. 1991. Reconciling contradictory notions on
lipid transfer proteins inhigher plants. Plant Physiol. Biochem. 29:705-711.

Molaan, B., W, J.M. and Junlper, BE. 1988. Continuity of chloroplast and
endoplasmic reticulum membranes in Chara and Equisetum. New Phytol. 109:59-
65.

McLean, LR. and Phflllpo, MC. 1982. Cholesterol desorption from clusters of
phosphatidylcholine and cholesterol in unilamellar vesicle bilayers during lipid
transfer or exchange. Biochemistry 21 :4053-4059.

21

“Is, J.T., Furlong, ST. and Dalildowlcz, EA. 1984. Plasma membrane biogenesis
in eukaryotic cells: translocation of newly synthesized lipid. Proc. Natl. Acad. Sci.

USA 81:1385-1388.

Mundy, J. and Rogers, J.C. 1986. Selective expression of a probable
amylase/protease inhibitor in barley aleurone cells: comparison to the barley
amylase/subtilisin inhibitor. Planta 196:51-63.

Mmro, S. and Pelham, H.R.B. 1987. A C-terminal signal prevents secretion of
luminal ER proteins. Cell 48:899-907.

lehlda, l. and Yamada, M. 1985. Semisynthesis of a spin-labeled
monogalactosyldiacylglycerol and its application to the assay for galactolipid-
transfer activity in spinach leaves. Biochim. Biophys. Acta 813:298-306.

Pdtauf, F. and Daum, G. 1992. Phospholipid transfer proteins from yeast. Methods
Enzymolon 209:514-522.

Raetz, C.R.l-l. 1982. Genetic control of phospholipid bilayer assembly. In:
Phospholipids, J.N. Hawthorne and GB. Ansell, eds., Elsevier, Amsterdam, pp.
435-477.

Halters, J., Tobor, I. and Kader, J.-C. 1984. Purification and binding characteristics
of a basic fatty acid binding protein from Avena safiva seedlings. Biochim. Biophys.
Acta 784:313-319.

Halters, J., Spenor. F. and Kader, J.-C. 1985. A phospholipid transfer protein that
binds long-chain fatty acids. FEBS 180:29-32.

swan, T.J., Nolmd, B.J., Gavey, K.L, Base, NM. Ookner, R.K.. Chanderbhan, R.
std Vahouny, G.V. 1985a. Sterol carrier protein 2 and fatty acid binding protein.
separate and distinct functions. J. Biol. Chem. 260:197-200.

Scallen, T.J., Noland, B.J., Gavey, KL. Base, NM. Ookner, R.K., Chanderbhan, R.
and Vahouny, G.V. 1985b. Sterol carrier protein 2 and fatty acid binding protein.
J. Biol. Chem. 260:4733-4739.

Soseountzov, L, Rilz-Avlla, L, Vlgnols, F., Jolllot. A., Arondel, V., Tcahng, F.,
Grosbois, M., Guerbette, F., Mlgnlac, E., Deleony, M., Pulgdomeneoh, P. and
Kader, J.-C. 1991. Spatial and temporal expression of a maize lipid transfer protein
gene. Plant Cell 3:923-933.

22

Stark. P., Boolf, H.. Schellekene, GA, Van Kamrnen, A. and De Vrlee. S.C. 1991.
Cell-specific expression of the carrot EP2 lipid transfer protein. Plant Cell 3:907-
921.

Svenooon, B., Aeono, K., Janoeoon, l.B., Pouleen, F.M., Mundy, J. and Svendsen,
LB. 1986. A 10 kd barley seed protein homologous with an B-amylase inhibitor
from Indian finger millet. Carlsberg Res. Commun. 51:493-500.

Szolderlts, e... W. A., Paltauf, F. and Daum, e. 1989. Membrane properties
modulate the activity of a phosphatidylinositol transfer protein from the yeast,
Saccharomyces cerevisiae. Biochim. Biophys. Acta 986:301-309.

Tel, S.-P. and Kaplan, S. 1985. Intracellular localization of phospholipid transfer
activity in Rhodopseudomonas sphaeroides and a possible role in membrane
biogenesis. J. Bacteriolon 164:181-186.

Taklshlma, K, Watanbe, S., Yamada, M. and Mamlya, G. 1986. The amino-acid
sequence of nonspecific lipid transfer protein from germinated castor bean.
Biochim. Biophys. Acta 870:248-255.

Taklchlma, K, Watanbe, 5., Yamada, M. and Mamlya, G. 1988. Amino acid
sequences of two nonspecific lipid transfer proteins from germinated castor bean.
Eur. J. Biochem. 177:241-249.

Tan, H.. Okazakl. K., Kubota, l., Kamlryo, T. and Utlyema. H. 1990. A novel
peroxisomal nonspecific lipid-transfer protein from Candida Vopicalis. Gene
structure, purification and possible role in B-oxidation. Eur. J. Biochem. 190:107-
112.

Tanaka. T. and Yamada, M. 1979. A phosphatidylcholine exchange protein
isolated from germinated castor bean endosperms. Plant Cell Physiol. 20:533-542.

Tchang, F., Thls, P., Stlefel, V., Arondel, V., Moroh, M.-D., Pages, M.,
Pulgdonmenoh, P., Grellet, F., Delaney, M., Boulflon, P., Huet, J.-C., Guerbette, F.,
Beauvals-Cante, F., Duranton, H. and Kader, J.-C. 1988. Phospholipid transfer
protein: full length cDNA and amino acid sequence in maize. J. Biol. Chem.
263:16849-16855.

Thoma. S.L, Kaneko, Y. and Somewllle, C. 1993. An Arabidopsis lipid transfer
protein is a cell wall protein. Plant Journal, in press.

Torres-Schumann, S., Godoy, JA. and Plntor-Toro, J.A. 1992. A probable lipid
transfer protein gene is induced by NaCl in stems of tomato plants. Plant Mol. Biol.
18:749-757.

23

Tsubol, S., Suga, T., Taklshlma, K., Mamlya, G., Matsul, K., Ozekl, Y. and Yamada,
M. 1991. Organ-specific occurrence and expression of the isoforms of nonspecific
lipid transfer protein in castor bean seedlings, and molecular cloning of a full-
Iength cDNA for a cotyledon-specific isoform. J. Biochem. 110:823-831.

Vergnolle, C., Arondel, V., Tchang, F., Grosbois, M., Guerbctte, F., Jolllot. A. and
Kader. J.-C. 1988. Synthesis of phospholipid transfer proteins from maize
seedlings. Biochem. Biophys. Res. Commun. 157:37-41.

Watanbe, S. and Yamada, M. 1986. Purification and characterization of a non-
specific Iipid transfer protein from germinated castor bean endosperms which
transfers phospholipids and galactolipids. Biochim. Biophys. Acta 876:116-123.

Whatley, J.M., McLean, B. and Juniper, BE. 1991. Continuity of chloroplast and
endoplasmic reticulum membranes in Phaseolus vulgaris. New Phytol. 117:209-
217.

erlz, KMIA 1982. Phospholipid transfer proteins. In: Lipid-protein interactions,
P.C. Jost and OH. Griffith, eds., Elsevier, Amsterdam, pp. 151-223.

Wlltz, KWA and Zilversmit, 0.8. 1968. Exchange of phospholipids between liver
mitochondria and microsomes in vitro. J. Biol. Chem. 243:3596-3602.

erlz, K.W.A. and Zilversmit, DB. 1970. Partial purification of phospholipid
exchange protein from beef heart. FEBS Lett. 7:44-46.

ertz, KWA, Dovaux, P.F. and Blenvenue, A. 1980. Phosphatidylcholine exchange
protein catalyzes the net transfer of phosphatidylcholine to model membranes.
Biochem. 19:3395-3399.

Woodlng, F.B.P. and Northcote, DH. 1965. Association of the endoplasmic
reticulum and the plastids in Acer and Pinus. Amer. J. Bot. 52:526-531.

Yamada, M. 1992. Lipid transfer proteins in plants and microorganisms. Plant Cell
Physiol. 33:1 -6.

Yamada, M., Tsuboi, S., Oeafme, T., Suga, T. and Taklehlma, K. 1990.
Multifunctional properties of non-specific lipid transfer proteins from higher plant.
In: Plant lipid biochemistry, structure, function, and utilization. P.J. Quinn and J.L
Harwood, eds. Portland Press Ltd., London, pp. 278-280.

CHAPTER 2
ISOLATION OF A CDNA CLONE FOR A SPINACH LIPID TRANSFER PROTEIN

AND EVIDENCE THAT THE PROTEIN IS SYNTHESIZED BY THE SECREI'ORY
PATHWAY

This work was originally published in Plant Physiology
Reference: Bernhard, W.R., Thoma, S., Botella, J. and Somerville, CR. 1991.

Isolation of a cDNA clone for spinach lipid transfer protein and evidence that the
protein is synthesized by the secretory pathway. Plant Physiol. 95:164-170.

24

25
ABSTRACT

A cDNA clone encoding a non-specific lipid transfer protein from spinach
was isolated by probing a library with synthetic oligonucleotides based on the
amino acid sequence of the protein. Determination of the DNA sequence indicated
a 354-nucleotide open reading frame which encodes a 118-amino acid residue
polypeptide. The first 26 amino acids of the open reading frame, which are not
present in the mature protein, have all the characteristics of a signal sequence
which is normally associated with the synthesis of membrane proteins or secreted
proteins. In vitro transcription of the cDNA and translation in the presence of
canine pancreatic microsomes or microsomes from cultured maize endosperm
cells indicated that proteolytic processing of the preprotein to the mature form was
associated with cotranslational insertion into the microsomal membranes. Since
there is no known mechanism by which the polypeptide could be transferred from
the microsomal membranes to the cytoplasm, the proposed role of this protein in
catalyzing lipid transfer between intracellular membranes is in doubt.

Although the lipid transfer protein is one of the most abundant proteins in
leaf cells, the results of genomic Southern analysis were consistent with the
presence of only one gene. Analysis of the level of mRNA by Northern blotting
indicated that the transcript was several-fold more abundant than an actin
transcript in leaf and petiole tissue, but was present in roots at less than 1% of the

level in petioles.

26
INTRODUCTION

Lipid transfer proteins, which have been isolated from animals, yeast, plants
and bacteria, are characterized by their ability to catalyze exchange of lipids
between natural or artificial membranes in who (Kader, 1985; Helmkamp, 1986;
WII'tZ and Gadella, 1990). Under conditions in which the donor membrane is
actively engaged in lipid synthesis, plant lipid transfer proteins have also been
shown to catalyze net transfer from the donor to acceptor membrane (Miguel, et
al., 1987). On this basis, it has been proposed that they participate in the transfer
of lipids between membranes in intact cells (Miguel, et al., 1987). However, this
family of proteins is relatively poorly characterized in several respects, and a role
in intracellular lipid transport is not established (Bishop and Bell, 1988).

In higher plants, both specific and non-specific lipid transfer proteins have
been characterized (Kader, 1985; Watanabe and Yamada, 1986). The basic lipid
transfer proteins from maize seedlings (Douady, et al., 1982), spinach leaves
(Kader, et al., 1984) and castor bean seedlings (Watanabe and Yamada, 1986)
are non-specific. All non-specific plant lipid transfer proteins are soluble proteins
which may account for as much as 4% of the total soluble protein (Kader, 1985).
The intracellular location of the lipid transfer proteins has not been established but
they are located outside the chloroplast (Schwitzguebel and Siegenthaler, 1985)
and have been thought to be cytosolic. The most thoroughly characterized
proteins have a pl of about 9 and a molecular mass of about 9000 Da (Kader, et

al., 1984; Watanabe and Yamada, 1986). The amino acid sequences have been

27

determined for the spinach leaf (Bouillon, et al., 1987) and castor seedling proteins

(T akishima, et al., 1986). In addition, on the basis of amino acid sequence identity

to the known lipid transfer proteins, several polypeptides from barley and finger

millet, which were originally described as probable amylase inhibitors (Mundy and

Rogers, 1986), were identified as lipid transfer proteins (Bernhard and Somerville,

1 989; Breu, et al., 1989). Recently, a cDNA of the maize lipid transfer protein has

been characterized (T chang, et al., 1986). Comparison of the deduced amino
acid sequences of cDNA clones encoding the maize and barley lipid transfer
proteins with the directly determined amino terminal sequences of the mature
proteins indicated that these proteins are synthesized as precursors containing 27
or 25 additional N-terminal amino acids, respectively.

In order to establish conditions for the analysis of the role of lipid transfer
proteins by genetic methods, a full length cDNA clone for the non-specific lipid
transfer protein from spinach leaves has been isolated and characterized. l have
characterized the level of mRNA abundance in various tissues and the number of
closely related genes in the spinach genome was examined. In addition, I
describe here the results of experiments indicating that synthesis of the mature

protein involves processing of a pre-protein by the secretory pathway.

EXPERIMENTAL PROCEDURES
Materials - A Agt11 cDNA library derived from mRNA of spinach leaves

(Spinacia oleracea L American Hybrid 424), was obtained from W. L. Ogren

28
(United States Department of Agriculture, Urbana, IL). Plasmid pSP65AB30, which

carries a cDNA encoding the light harvesting chlorophyll a/b binding protein from
Lemna gibba under transcriptional control of the SP6 promoter (Kohorn, et al.,
1986) was obtained from E. Tobin (UCLA, Los Angeles, CA). Plasmid pSPBP1
encoding a cDNA for bovine prolactin under transcriptional control of the SP6
promoter was obtained from V.R. Lingappa (UCSF, San Francisco, CA). Plasmid
pSAc3 carrying an actin gene from soybean (Shah, et al., 1982) was obtained from
R. Meagher (University of Georgia, Athens, GA). The plasmid pBluescript (B81)
was purchased from Stratagene (San Diego, CA). Oligonucleotides were
synthesized by the phosphoramidite method on an Applied Biosystems 380A
instrument.

A preparation of microsomal membranes suitable for in vitro translocation
and processing of precursor proteins was generously provided by J. Miernyk (US
Department of Agriculture, Peoria, IL). The microsomes were prepared from maize
endosperm cultures (Miernyk, 1987) by a modification (R.G. Shatters Jr. and J.A.

Miernyk, manuscript in preparation) of the method of Burr and Burr (1981).

Plaque Screening - The cDNA library was plated on Escherichia coliY1090
and nitrocellulose plaques lifts were screened with the oligonucleotide mixtures
SLTP-1 (ATGTG(C/T) GGNGTNCA(C/T)AT) and SLTP-2
(AA(A/G)GGNATU/C)AA('I’/C)TA-(T/C)GG) which were labeled to an average

specific activity of 10" cpm pg" with [y-‘FP1ATP (3000 Ci mmol") and T4

29
polynucleotide kinase (Maniatis, et al., 1982). The filters were prehybridized 3 to

5 h at 42°C for SLTP-1 and at 39°C for SLTP-2, In 6 x 880*, 50 mM NaPO4 (pH
6.8), 5 x Denhardt’s solution and 0.5 mg ml'1 sonicated herring DNA (Sigma). The
hybridization was carried out at the same temperatures for 36 h with the addition
of 2 pmol of the appropriate oligonucleotide mixture. The filters were washed in
4 x SSC at the same temperature as the hybridization, with 4 changes of the

washing buffer at 30 min intervals.

Nucleic Acid Manipulations - RNA and DNA was extracted from 6 week old
spinach by grinding 5 g of tissue in liquid N2, then stirring for 15 min at 60°C in
a mixture of 50 ml of 100 mM Tris-HCI pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5%
(w/v) SDS, 2% (v/v) 2-mercaptoethanol and 50 ml of phenol. The mixture was then
extracted once with 100 ml of chloroform/Isoamyl alcohol (24:1), once with
phenol/chloroform/Isoamyl alcohol (24:24:1) and the nucleic acids recovered by
ethanol precipitation. The pellet was resuspended in 20 ml water, 1 volume of 4
M LiCl was added and after incubation for 16 h at -20°C the RNA was pelleted by
centrifugation at 12,000 x g for 60 min. The RNA pellet was washed with 70%
ethanol. DNA was isolated from the supernatant of the UCI precipitation by adding
2 volumes of ethanol and centrifuging at 15,000 x g for 60 min, then purified by
CsCl centrifugation.

Fragments to be sequenced were subcloned into the appropriate restriction

site of pBluescript and sequenced by the chain termination method using

30
Sequenase (US Biochemicals).

For Southern analysis DNA was restricted and resolved by electrophoresis
on a 0.8 % agarose gel (10 pg per lane). Following depurination in 0.25 M HCI
and denaturation in 0.5 M NaOH, 1.5 M NaCI the DNA was transferred in 0.25 M
NaOH, 1.5 M NaCI to nylon membrane (Hybond N, Amersham). The membrane
was prehybridized 3 to 5 h at 50°C in 6 x SSC, 5 x Denhardt’s solution and 1 mg
ml" sonicated herring DNA. The hybridization was carried out at 50°C for 10 h
with the addition of 0.2 pg of the purified insert from the plasmid pWB2 labelled
with [a-“PJdCTP by random priming (Feinberg and Vogelstein, 1983) to an
average specific activity of 5 x 10‘ cpm pg". The membranes were washed once
with 1 x SSC, 0.2% SDS at 50°C and 3 times with 0.1 x SSC, 0.2% SDS at the
same temperature.

For Northern analysis 20 pg RNA was resolved by electrophoresis in an
agarose gel containing formaldehyde (Maniatis, et al., 1982). After the transfer (in
10 x SSC) to nitrocellulose, the blot was prehybridized at 42°C in 5 x SSC, 5 x
Denhardt’s solution, 50 mM NaHPO, (pH 7.0), 0.01% SDS, 50% deionized
formamide and 1 mg ml" sonicated herring DNA. The hybridization was carried
out at the same temperatures for 10 h with the addition of 0.4 pg of the purified
insert of the plasmid pWB2 labelled with [a-32P]dCTP by random priming. The
membranes were washed twice with 2 x SSC, 0.1% SDS at 23°C, twice with 2 x

SSC, 0.1% SDS at 60°C and once with 0.2 x SSC, 0.1% SDS at 60°C.

31

In vitro Transcription and Translafion - Transcripts were prepared from
linearized plasmids with T3 RNA polymerase in the presence of 500 pM
m7G(5’)ppp(5’)G (Pharmacia) and 2.5 mM of each ribonucleotide triphosphate at
38°C for 1 h. The reaction was terminated by addition of 1 U of DNAse per pg
of template DNA. After 10 min the reaction was extracted with chloroform/phenol
(1:1), then with chloroform, adjusted to 0.3 M sodium acetate and RNA recovered
by ethanol precipitation.

Translation assays containing 0.5 pg RNA, 15 pl of wheat germ extract
(Promega), 15 pCi aE’s-methionine (1149 Ci/mmol; Amersham) in a final volume of
30 pl were incubated 1 h at 30°C. No additional KOAc or MgOAc was required
for maximal rates of translation of transcripts from pWB2 or pSP65AB30. Maximal
rates of translation of transcripts from pSPBP1 were obtained following addition
of 33 mM KOAc. For some experiments 3 pl of canine pancreatic microsomes
(Promega), or 8 pl (21.4 pg of protein) of microsomes from cultured cells of maize
endosperm tissue, were added to the translation reactions. The translation
products were resolved electrophoretically by loading equal amounts of
incorporated radioactivity on 10 to 20% gradient SDS polyacrylamide gels. The
resolved proteins were electrophoretically transferred to nitrocellulose then the

filters were autoradiographed.

Protease Protection - Translation reactions were terminated by addition of

1 pl of a solution containing 3 mM cycloheximide and 0.12 M methionine, and

32

divided into three identical samples. One sample was left untreated as a control.
CaCl2 (10 mM) and proteinase K (1 mg ml") were added the second sample. The
third aliquot was adjusted to contain 1% Triton X-100 before addition of the
protease. The reactions were incubated at 24°C for 1 h and the treatments were

terminated by addition of 0.1 volume of 100 mM phenethylmethyl sulfonyl fluoride

in isopropanol.

RESULTS

Isolation and Sequence Analysis of a cDNA Clone - On the basis of the
amino acid sequence of the lipid transfer protein of spinach (Bouillon, et al., 1987)
two oligonucleotide mixtures composed of 64 (SLTP-1) and 96 (SLTP-2) different
17-mers were prepared, which contained all possible DNA sequences which could
encode the corresponding regions of the protein sequence. Screening of a Agt11
cDNA library derived from spinach leaf poly(A+) RNA resulted in the recovery of
two positive plaques from among approximately 75,000 screened. Both were
identical and contained two EcoRl fragments of 580 and 730 bp, respectively.
These inserts were subcloned into the EcoRI site of pBluescript to produce the
plasmids pWB2 and pWB3 and the DNA sequence of the inserts in both plasmids
was determined. The insert in plasmid pWB2 was found to encode the spinach
lipid transfer protein. The insert in pWB3, which also had a poly-(At) sequence at

one end, appeared to be a fragment of an unrelated cDNA which could not be

33

identified by sequence homology to any previously determined DNA or protein
sequence and was not characterized further.

The composite nucleotide and deduced amino acid sequences of pWB2 are
shown in Fig 2-1. The entire cDNA was 564 nucleotides long and contained one
open reading frame of 354 bp which began 13 bp from the 5’ end and terminated
197 bp from the 3' poly(A) tail. The open reading frame encoded a polypeptide
of 118 amino acids. The carboxy terminal 92 residues of the open reading frame
were almost identical to the directly determined amino acid sequence of the
mature protein (Bouillon, et al., 1987). The only differences were that the deduced
amino acid sequence had cysteines rather than serines at residues 27 and 28.
Cysteines are also found at these positions in all other homologous proteins
(Bernhard and Somerville, 1989). The deduced amino acid sequence also
contained 26 amino terminal residues which were not found on the mature
polypeptide. This observation is consistent with evidence from the analysis of the
maize (T chang, et al., 1988) and barley (Mundy and Rogers, 1986; Bernhard and
Somerville, 1989) genes indicating that these polypeptides are produced by

cleavage of a preprotein.

RNA Analysis - In order to assess the size of the transcript encoding the
spinach lipid transfer protein, and to evaluate the tissue specificity of gene
expression, the cDNA clone was used to probe a Northern blot containing spinach

RNA from various tissues (Fig. 2-2A). The results of this experiment indicated that

TTGCATTATA TTT ATC CCT AGC TCC CCT GTT ATC AAC TTA CCT TGT 46
H A S S A V I K L A C '16

GCA CTC CTG TTG TGC ATC GIG GTC CCT GCA CCA TAC CCT GAA CCA 91
A V L L C I V V A A P Y A E A '1

661 ATA ACT TGT GGG ATG GIT TCA AGC AAA CTT GCT CCT TGC ATT 136
G I T C G H V S S K L A P C I 15

SOS TAC CTT AAA GGA GGC CCC TTG CCC GGT GGT TGC TGT GGT GGA 181
G Y L K G G P L G G G C C G G 30

ATT AAC GCC CTC AAC GCG CCA CCT CCC ACC ACT CCT GAC AGG AAA 226
I K A L N A A A A T T P D R K 45

ACT GCA TGC AAT TGC CTC AAA ACT CCT GCT AAT CCC ATT AAG GGA 271

T A C N C L K S A A N A I K G 60

ATC AAC TAC GGA AAC CCT CCT GGT CTC CCT GGT ATG TGT GGC CTC 316
I N Y C K A A G L P G H C G V 75

CAT ATT CCC TAC GCC ATT ACC CCC ACC ACC AAC TGC AAC CCC GTC 361
H I P Y A I S P S T N C N A V 90

CAC TAA ACCGCAAATGTTATAACAAAAATCGAAGATGGAGCTACATAGGAGTGGCC 417
H *

CAGTTACTAAGCTCCTAGAGTGTATGATAATAAAAGAAGAGAGATCATCTTTGCCAAGT 476
CGCTAGCTTCTATTTCTTGTTTCATGTATTATTGCAACTTTTCTATTACTTTTCGGGTT 535

ACAAATATCCTAATATTACCAAAAAAAAA 56‘

Figure 2-1. Composite nucleotide sequence of the cDNA for spinach lipid transfer

protein and the deduced amino acid sequence. The amino acids are numbered

from the glycine residue found at the amino terminus of the mature protein

(Bouillon, et al., 1987). The regions of sequence homologous to the oligonucleotide

probes used to isolate the gene are underlined. The cDNA was isolated by

\gemer Bernhard. Sequencing was carried out by Werner Berhard and Jose
otella.

35
the probe hybridized to an mRNA of approximately 700 bp in RNA preparations

from leaves and petioles. Since this is similar to the size of the cDNA clone in
pWB2, it appears that the insert in pWB2 represents a nearly full-length cDNA.
There was a striking difference in expression of the mRNA in green tissues versus
root tissue, where accumulation of the lipid transfer protein mRNA was not readily
apparent. .On prolonged exposure of the filter a faint signal could be observed in
the lane of root RNA (results not presented). This signal was estimated to be
approximately 100-fold less abundant than the leaf signal. In order to verify that the
preparation of RNA from the various tissues were of comparable quality, the filter
was stripped and reprobed with an actin gene which was previously shown to
hybridize to a similar extent to RNA from roots, shoots and hypocotyls of soybean
(Hightower and Meagher, 1985). The presence of hybridization signals of similar
intensity in the root and petiole lanes when probed with the actin gene (Fig 2-23)

verified that the two preparations of RNA were of similar quality and quantity.

Southem Analysis - Two protein fractions with lipid transfer activity have
been reported for spinach (Kader, et al., 1984), suggesting the presence of
isoforms of lipid transfer protein. To investigate the number of homologous genes
present in the spinach genome, Southern analysis was performed with genomic
DNA cut with several restriction enzymes which do not have recognition sites in the
cDNA sequence. For 5 of the 6 restriction enzymes used, the cDNA probe

hybridized to only one fragment, and in 4 of these cases the fragment was less

36

 

LLI LLI
5'1— e'r-
‘érs 35.8
BEE 40.0:
||| ||l
4.40-A B

2.37 -

1.35 - I ..

0.24 -

 

 

 

 

 

LTP ACTIN

Figure 2-2. Northern blot of total RNA from spinach probed with the cDNA of
spinach lipid transfer protein labelled to a specific activa of 2 x 109 dpm/ug DNA
(A) then stripped and reprobed with an actin cDNA labeled to a specific activity of
3 x 109 dpm/ug DNA (B). Twenty pg of RNA was loaded in each lane. The filter in
A was exposed 2 days, the filter in B was exposed 5 days. The size (in kb) and
location of molecular weight markers is indicated at the left of the figure.

37
than 4 kb in size (Fig. 2-3). Thus, it appears likely that only one gene encoding the

lipid transfer protein is present in the spinach genome. The presence of two
bands in the lane containing DNA restricted with Dral suggests the presence of at
least one intron in the genomic sequence. In view of the very high abundance of
the protein in leaf tissues, the mechanism by which a single gene appears to
regulate accumulation of the protein may merit further investigation with respect

to protein and mRNA turnover and rate of transcription.

In Vitro Translation and Processing - The sequence of the 26 amino-terminal
residues of the deduced amino acid sequence of the spinach lipid transfer protein
contained many of the features associated with signal sequences (Watson, 1984;
von Heijne, 1985). The general characteristics of signal peptides are: (1) they are
20 to 40 amino acids long, (2) there is a charged residue within the first 5 amino
acids in the amino-terminal direction from the cleavage site and this is followed by
a core of at least nine hydrophobic residues, (3) A helix-breaking residue (glycine
or proline) frequently occurs 4 to 8 residues before the cleavage site. The most
stringent requirement is that an alanine residue occurs at the minus-1 and minus-3
positions. The prepeptide of the spinach lipid transfer protein satisfies all of these
criteria.

The role of the amino-terminal sequence was tested by transcribing and
translating the cDNA in vitro under conditions which lead to processing of signal

Peptides by the secretory pathway (Walter and Lingappa, 1986). When transcripts

38

 

E E
2 a E 8 Er: a
X CD I Lu 0 m
I I l I I I
9.4-

6.6- ~
4.3- . p.
I '-

U
2.0- t
0.6-

 

 

 

Figure 2-3. Southern blot of total spinach DNA probed with the cDNA of spinach
lipid transfer protein. The size (in kb) and location of molecular weight markers is
indicated at the right of the figure. Southern analysis was carried out by Werner
Bernhard.

39

from the cDNA were translated without addition of microsomal membranes (Fig.
2-40), or when the dog microsomal membranes were added after translation was
complete (Fig. 2-4F), the translation product was of the size expected from the
open reading frame. By contrast, when dog microsomal membranes were present
during translation, a translation product of the size expected for the mature
polypeptide accumulated (Fig. 2-4E). Identical results were obtained with the
positive control, prolactin (Fig. 2-4A-C), whereas the presence of dog microsomes
had no effect on the size of the chlorophyll a/b binding protein (Fig. 2-4G,H) which
is inserted into chloroplast membranes by a different mechanism. Replacement of
dog microsomes with maize microsomal membranes resulted in processing of the
primary translation product to a processed form of the same size as that obtained
with the dog microsomes (Fig. 2-5). Thus, the amino termius of the preprotein
functions as a signal peptide in both plants and animals.

In order to determine if the lipid transfer protein was cotranslationally
inserted into the microsomal membranes, the protease sensitivity of the translation
products was assessed by treating the translation reactions with proteinase K, then
inactivating the proteinase with phenethylmethylsulfonyl fluoride prior to SDS-
PAGE. When translated in the absence of microsomes, all of the translation
products from the lipid transfer gene and the control gene (prolactin) were
protease sensitive (Fig. 2-6B,G). By contrast, when transcripts from the lipid
transfer protein and prolactin genes were translated in the presence of dog

microsomes, the low molecular weight product was protected from proteolysis but

4O

 

 

BP LTP .LHCP
A B C D E F G H
42.1 -
30.4 ‘ ..
m
18.2 -
13.7 -
81 _ -t.
2.7 -

 

 

 

Figure 2-4. Processing of in vitro translation products by canine microsomal
membranes. Transcripts from the bovine prolactin gene (BP), the spinach lipid
transfer protein gene (LTP) and the chlorOphyll a/b binding protein (LHCP) were
translated without addition of microsomes (A,D,G), with microsomes (B,E,H) and
with microsomes added after translation had been terminated (C,F). The numbers
at the left indicate the apparent molecular weight of standards (in kDa).

 

41

 

43.0 -

25.7 -

 

—L—L
our-co
meat:

I

3.0 -

 

 

 

Figure 2-5. Processing of in vitro translation products by microsomal membranes
from maize endosperm cell cultures. Transcripts of the lipid transfer protein were
translated without the addition of microsomes (A), with maize microsomes (B), and
wrth canine microsomes (C). The numbers at the left indicate the apparent
molecular weight of standards (in kDa).

42

 

 

30.4 -

18.2 -
13.7 -

 

 

2.7 -

 

 

Figure 2-6. Protease sensitivity of translation products. Transcripts from the bovme
prolactin gene (BP) and the spinach lipid transfer protein gene (LTP) were
translated without addition of microsomes (A,B,F.G) or with dog microsomes
(C.D.E,H,I,J). After translation, the reaction products were incubated WIth
proteinase K for 1 h (B,D,E,G,I,J). In lanes E and J, Triton X-100 was added along
with the proteinase K.

43
the high molecular weight product was not (Fig. 2-6D,I). The presence of

microsomes also resulted in the appearance of a translation product from the lipid
transfer gene of higher molecular weight. This does not appear to be due to N-
Iinked glycosylation since the polypeptide does not contain the glycosylation sites
Asn-X-Ser or Asn-X-Thr. Addition of low amounts of Triton-X100 to these
translation mixtures concomitantly with addition of the proteinase K rendered the
low molecular weight products protease sensitive (Fig. 2-6E,J). Thus, the evidence
indicates that the mature lipid transfer protein is sequestered in the endoplasmic
reticulum. This result is consistent with the results of recent experiments indicating
that maize lipid transfer protein is synthesized on membrane-bound polysomes

(Vergnolle, et al., 1988).

DISCUSSION

The mechanisms by which lipids are transferred between intracellular
membranes are not established (Bishop and Bell, 1988). However, the
demonstrations that lipid transfer proteins could mediate rapid not transfer of lipid
between a donor and acceptor membrane in vitro (Miguel, et al., 1987) suggests
that this mechanism could be important in vivo. According to this hypothesis, the
rate of transfer of lipid between membranes would be proportional to the rate of
diffusion of the protein-lipid complex between donor and acceptor membranes.
In this respect, the high abundance of the spinach leaf protein (i.e., up to 4% of

total soluble protein) in rapidly expanding tissue is consistent with such a

44

proposed role. By contrast, the very low abundance of the transcript in root tissue
suggests that the protein is not involved in lipid transport in this tissue. This could,
in principle, be due to the involvement of different isozymes in this tissue. Lipid
transfer in leaves and other green tissues involves substantial flux between
endoplasmic reticulum and chloroplast membranes (Browse, et al., 1986). By
contrast, in roots, relatively little flux of lipid is thought to take place between
plastids and endoplasmic reticulum. Thus, lipid transfer proteins with different
characteristics may be utilized in green and non-green tissues. The possibility of
additional isozymes of related proteins is not excluded by the Southern analysis
reported here since the conditions used would not have revealed homology to any
DNA sequence of less than about 80% sequence identity. Because of the
relatively large size of the spinach genome, and the fact that it has not yet been
possible to genetically transform spinach, a detailed analysis of a weakly
homologous gene family in this organism is not considered worthwhile. However,
we have used the spinach cDNA clone described here to isolate the homologous
cDNA and genomic clones from Arabidopsis maliana (results not presented). The
greatly reduced genome size of this organism, and the ease with which it can be
genetically transformed, should facilitate a critical analysis of the question of tissue-
specific isoforms.

The evidence, presented here, indicating that synthesis of the spinach lipid
transfer protein involves cotranslational insertion into the endoplasmic reticulum,

suggests that either this protein is synthesized by a novel mechanism, or it does

45

not participate in intracellular lipid transfer. Since it lacks the carboxy-terminal
KDEL sequence associated with proteins retained in the lumen of the endoplasmic
reticulum (Munro and Pelham, 1987) it is presumably transferred to another
compartment of the cell by additional post-translational processing. Because the
mature protein is normally recovered as a soluble extrachloroplastic protein, it has
been assumed to be located in the cytosol (Kader, 1985). However, if the mature
protein is primarily located in the cytosol, it would appear to be the first instance
in which a cytosolic protein is synthesized via the secretory pathway with cleavage
of the signal peptide. The only reported instance in which synthesis of a cytosolic
protein involves cleavage of a signal sequence by signal peptidase is in the case
of the hepatitis B virus pro-core protein (Garcia, et al., 1987). However, in this
case, translocation of the protein into the endoplasmic reticulum during synthesis
is aborted at an early stage so that the protein is not inserted into the membrane.
By contrast, the protease protection experiments indicate that the lipid transfer
protein is taken up by the endoplasmic reticulum. There is no known mechanism
by which it could subsequently be transfered to the cytoplasm.

If the spinach lipid transfer protein is not located in the cytosol, it cannot
carry out the proposed role in transferring lipid between the endoplasmic reticulum
and the other organelles. Unfortunately, the precise intracellular location of the
protein has not been critically addressed by previous studies. Thus, a resolution
of the alternate possibilities must await the results of an immunoelectron

microscopic analysis of the cellular location of the protein.

46
REFERENCES

Bernhard, W.R. and Somervflle, CR. 1988. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from
amino acid sequence homology. Arch. Biochem. Biophys. 269:695-697.

Blshop, W.R. and Bell, RM. 1988. Assembly of phospholipids into cellular
membranes: Biosynthesis, transmembrane movement and intracellular
translocation. Annu. Rev. Cell Biol. 4:579-610.

Boulllon, P., Drlschel, C., Vergnolle, C., Duranton, H. and Kader, J.-C. 1987. The
primary structure of spinach-leaf phospholipid-transfer protein. Eur. J. Biochem.
166:387-391.

Breu, V., Guerbette, F., Kader, J.-C., Kamangara, C., Svenseon. B. and Von
Wettateln-Knowlea, P. 1989. A 10 kd barley basic protein transfers
phosphatidylcholine from liposomes to mitochondria. Carlsberg Res. Commun.
54:81-84.

Browse,J.,Warwlck, N.,Somervllle, C.R. andSlack, CR 1986. Fluxes through the
prokaryotic and eukaryotic pathways of lipid synthesis in the 16:3 plant

Arabidopsis thaliana. Biochem. J. 235:25-31.

Burr, FA. and Burt, B. 1981. In vitro uptake and processing or prozein and other
maize preproteins by maize membranes. J. Cell Biol. 90:427-434.

Douady, D., Grosbois, M., Guerbette, F. and Kader, J.-C. 1982. Purification of a
basic phospholipid transfer protein from maize seedlings. Biochim. Biophys. Acta
710:143-153.

Felnberg, AP. and Vogelsteln, B. 1983. A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-
13.

Garcla, P.D., Ou, J.H., m, W.J. and Walter, WD. 1987. Targeting of the
hepatitis B virus precore protein to the endoplasmic reticulum membrane: After
signal peptide cleavage translocation can be aborted and the product released
into the cytoplasm. J. Cell Biol. 104:705-712.

Helmkarnp, G.M., Jr. 1986. Phospholipid transfer proteins: Mechanism of action.
J. Bioenerg. Biomem. 18:71-93.

HIghtower. RC. and Meagher, RB. 1985. Divergence and differential expression
of soybean actin genes. EMBO J. 4:1-8.

47
Kader, J.-C.1985. Lipid binding proteins in plants. Chem. Physics Lipids 38:51 -62.

Kader, J.-C., Julienne, M. and Vergnolle, C. 1984. Purification and characterization
of a spinach-leaf protein capable of transferring phospholipids from liposomes to
mitochondria or chloroplasts. Eur. J. Biochem. 139:411-416.

Kohorn, B.D., Harel, E., Chllnls, P.R., Thomber, J.P. and Tobln, EM. 1986.
Functional and mutational analysis of the light-harvesting chlorophyll a/b protein
of thylakoid membranes. J. Cell Biol. 102:972-981.

Manlatla, T., Frlbcl'l, EF. and Sambrook, J. 1982. Molecular Cloning. A Laboratory
Manual. Cold Spring Harbor, New York.

Miernyk, J.A. 1987. Extracellular secretion of acid hydrolases by maize endosperm
cells grown in liquid medium. J. Plant Physiol. 129:19-32.

Mlquel, M., Block, MA. Joyard, J., Dome, A.J., Dubaoq, J.P., Kader, J.-C. and
Bones, R. 1987. Protein-mediated transfer of phosphatidylcholine from liposomes
to spinach chloroplast envelope membranes. Biochim. Biophys. Acta 937:219-228.

Mundy, J. and Rogers, J.C. 1986. Selective expression of a probable
amylase/protease inhibitor in barley aleurone cells: Comparison to the barley
amylase/subtilisin inhibitor. Planta 196:51-63.

Munroe, S. and Pelham, H.R.B. 1987. A C-terminal sequence prevents secretion
of luminal ER proteins. Cell 48:899-907.

Schwltzguebel, J.P. and Slegenthaler. PA. 1985. Evidence for a lack of
phospholipid transfer protein in the stroma of spinach chloroplasts. Plant Sci.
40:167-171.

Shah, D.M., Hightower, RC. and Meagher, RB. 1982. Complete nucleotide
sequence of a soybean actin gene. Proc. Natl. Acad. Sci. USA 79:1022-1026.

Taklshlma, K., Watanabe, S., Yamada. M. and Mamlya. G. 1986. The amino-acid

sequence of nonspecific lipid transfer protein from germinated castor bean.
Biochim. Biophys. Acta 870:248-255.

Tchano. F.. This. P., Stlefel, V., Arondel, v., Morch, M.D., Pages, M., Pulgdomeoh,
P., Grellet, F., Debeny, M., Boulllon, P., Huet, J.C., Guerbette, F., Beauvais-Cante,
F., Duranton, H.. Pemollet, J.C. and Kader, J.-C. 1988. Phospholipid transfer

protein: Full-length cDNA and amino acid sequence in maize. J. Biol. Chem.
263:16849-16855.

48

Vergnolle, C., Arondel, V., Tchang, F., Grosbols, M., Guerbette, F., Jolllot. A. and
Kader, J.-C. 1989. Synthesis of lipid transfer proteins from maize seedlings.
Biochim. Biophys. Res. Comm. 157:37-41.

von Heine, G. 1985. A new method for predicting signal sequence cleavage sites.
Nucleic Acid Res. 14:4683-4690.

Walter, P. and Ungappa. V. 1986. Mechanism of protein translocation across
endoplasmic reticulum membranes. Annu. Rev. Cell Biol. 2:499-516.

Wau'Iabe, S. and Yamada, M. (1986) Purification and characterization of a non-
specific lipid transfer protein from germinated castor bean endosperms which
transfers phospholipids and galactolipids. Biochim. Biophys. Acta 876:116-123.

Watson, M.E.E. 1984. Compilation of published signal sequences. Nucleic Acid
Res. 12:5145-5159.

erlz, K.WA. and Gm T.W.J., Jr. 1990. Properties and modes of action of
specific and non-specific phospholipid transfer proteins. Experientia 46:592-599.

CHAPTER 3

AN ARABIDOPSIS NON-SPECIFIC LIPID TRANSFER PROTEIN
IS A CELL WALL PROTEIN

This paper has been submitted to and accepted by 7716 Plant Journal

Reference: Thoma, S., Kaneko, Y. and Somerville, C. 1993. The non-specific lipid
transfer protein from Arabidopsis is a cell wall protein. Plant J., in press.

49

ABSTRACT

Lipid transfer proteins (LTPs), mediate the transfer of phospholipids
between membranes in vitro. However, the in vivo function of LTPs is not known.
To determine the precise location of the non-specific LTP from Arabidopsis, a
cDNA clone was used to produce an Arabidopsis LTP:protein A fusion. Antibodies
raised against the fusion protein were used to localize the Arabidopsis LTP by
immunoelectron microscopy. LTP was found to be located in the cell wall, mainly
in epidermal cells. This location appears to be inconsistent with the proposed role
of the protein in intracellular lipid transfer, but is consistent with a possible function

in cuticle formation.

51
INTRODUCTION

The mechanism by which lipids move from their main sites of synthesis, the
endoplasmic reticulum (ER) and the chloroplast, to other cellular organelles is
unknown. The search for proteins capable of intracellular lipid transfer resulted in
the isolation of several small proteins which had the ability to mediate the in vitro
transfer of radiolabeled phospholipids from liposomal donor membranes to
mitochondrial and chloroplast acceptor membranes (Douday et al., 1982, Kader
et al., 1984). It was, therefore, suggested that these proteins might catalyze lipid
transfer between membranes within a cell (Arondel and Kader, 1990). However,
the turnover number for the in vitro activity is very low and there is no evidence
that these proteins carry out this role in vivo.

Lipid transfer proteins (LTPs) have been isolated from fungi, bacteria, plants,
and mammals. Both specific and nonspecific LTPs (nsLTPs) have been isolated
from mammalian tissue (ert2 and Gadella, 1990), but only nsLTPs have been
found in plant tissue (Arondel and Kader, 1990). Non specific LTPs have been
isolated from maize (Douday et al., 1982), spinach (Kader et al., 1984), barley
(Mundy and Rogers, 1986), and castor (T akishima et al., 1986). Plant nsLTPs.
While lacking any obvious sequence homology to animal LTPs, share high amino
acid sequence similarity to the nsLTPs from other plant species. All known plant
LTPs have several common characteristics; they are soluble, basic proteins with
a molecular mass of approximately 9 kDa (Arondel, et al., 1989) and, except for

one cysteine to glutamine conversion in the ragi protein, all plant LTPs have 8

52

conserved cysteine residues (Bernhard and Somerville, 1989).

Based on the proposed function of intracellular lipid transfer, it has been
assumed that LTPs were cytosolic (Arondel and Kader, 1990). However, recent
evidence suggests that LTPs are not located in the cytosol. It has been reported
that the maize LTP has an amino terminal extension and is synthesized on
membrane bound polysomes (T chang, et al., 1988, Vergnolle, et al., 1988) and that
spinach LTP contains a signal peptide and is cotranslationally inserted into
microsomal membranes (Bernhard, et al., 1991). Since these proteins lack the
endoplasmic reticulum retention signal, KDEL, they would be expected to be
secreted or targeted to a specific organelle. In support of these expectations,
Mundy and Rogers (1986) reported that a protein, which was subsequently
identified as a barley LTP, was found in aleurone cell culture media. Similarly,
Sterk, etal. (1991), receme demonstrated that a carrot LTP was secreted by carrot
embryogenic cell cultures.

The expression pattern of LTP has been studied in cultured embryogenic
carrot cells and in developing maize (Sossountzov, et al., 1991; Sterk, et al., 1991).
In situ hybridization studies show that the carrot LTP mRNA is localized to the
protoderm cells of somatic and zygotic embryos and in suspensor cells and the
integument of zygotic embryos. The maize LTP RNA is expressed in the epidermis
of the coleoptile and the scutellum. The localization of maize LTP has been
investigated using immunogold labeling and light microscopy (Sossountzov, et al..

1991). The highest concentrations of LTP were found in the outer epidermis of

53

coleoptiles and in the vascular tissue of the leaf.

To date, the localization of plant nsLTP in the aerial parts of mature plants
has not been determined and the precise localization of LTPs has not been
determined using electron microscopy. Therefore, I have used antibodies against
an Arabidopsis LTP:protein A fusion for immunocytochemical labeling of various
Arabidopsis tissues at the ultrastructural level. I show here that Arabidopsis LTP
is located mainly in epidermal cell walls. The location of LTP outside of the plasma
membrane and the preferential expression in epidermal cells are consistent with

a previous proposal that LTP may participate in cuticle formation (Sterk, et al.,

1991).

EXPERIMENTAL PROCEDURES

Plant growm conditions - Columbia wild type of Arabidopsis thaliana (L.)
Heynh. was used in all experiments. Plants were grown on a
perlitezvermiculitezsphagnum mixture (1 :1 :1 ), at 22°C, under continuous illumination

(100-150 pE m'zs").

Plasmid construction - The synthetic oligonucleotides.
5’TAAGGATCCAATACATCGAACGCT3’and5’AACTGCAGTACTCTGAAA‘ITI'CA3’.
were used as primers to generate a 510 bp PCR product from a previously
isolated Arabidopsis LTP cDNA (Genbank accession M80566). This product was

digested with EcoRV and BamHI and was ligated into the corresponding restriction

54
sites of pBluescript (Stratagene, San Diego, CA), to yield the plasmid pSLT1. The

Pstl and BamHI fragment from pSLT1 was ligated into pRIT2T (Pharmacia,
Piscataway, NJ), a vector which contains the affinity tail of staphylococcal protein
A, to yield the plasmid pSLT2, which was subsequently used to produce a
LTP:protein A fusion protein. The protein A-LTP junction was sequenced to assure
that these fragments were in the correct reading frame. pSLT3, which was used
to produce an LTszaItose binding protein fusion, was prepared by ligating the
BamHI and Pstl fragment of pSLT1 into the corresponding site of le821 (New
England Biolabs, Beverly, MA), a plasmid which contains malE, the E. coli maltose

binding protein. Restriction digests and ligation reactions were carried out as

described by Maniatis et al. (1982).

Protein expression and purification - For production of the LTP:protein A
fusion protein, E. coli strain N4830-1 (Pharmacia, Piscataway, NJ) containing
pSLT2 was grown in 500 ml L-broth containing 50 mg/L ampicillin at 37°C until late
log phase. At this point, the incubation temperature was rapidly increased to 42°C
for 2 hours to induce expression of the fusion protein. Bacteria were collected by
I centrifugation, resuspended in 30 ml TST (50 mM Tris-Cl, pH 7.6, 150 mM NaCl,
0.05% Tween 20) and lysed using a French press. The lysed cells were
centrifuged at 10,000 x g for 10 minutes and the supernatant was applied to an
lgG affinity column (4 X 1.5 cm, Pharmacia, Piscataway, NJ). The fusion protein

was eluted using 0.5 M HOAc, adjusted to pH 3.4 with NH.OAc. The eluent was

55
lyophilized and resuspended in 25 mM sodium phosphate buffer, pH 75, The

sample was then applied to a MonoQ column (5.5 X 0.7 cm) and eluted with a 50
ml linear gradient of 0 to 1 M NaCl in 40 mM Tris-Cl, pH 8.0. The purity of the
protein was monitored by SDS-PAGE (Laemmli, 1970).

For production of the LTPzMaIE fusion protein, E. coli strain DH5a
containing. pSLT3 was grown in 1 liter L-broth containing 50 mg/L ampicillin to mid
log phase. At this point, IPTG was added to a final concentration of 0.3 mM. After
2 hours, cells were collected by centrifugation and resuspended in 40 ml 10 mM
sodium phosphate, 30 mM NaCl, 0.25% Tween 20, 10 mM fi-mercaptoethanol, 10
mM EDTA, pH 7. Cells were lysed using a French press, centrifuged at 9000 x g
for 30 min and the supernatant applied to an amylose column (13 X 2.5 cm). The
amylose column was prepared adding 60 ml 5 N NaOH and 30 ml epichlorhydrin
to 40 ml 25% (w/v) amylose. The resulting gel was fragmented in a Waring
blender for approximately 5 seconds. The protein was eluted from the amylose

column with 3 ml 10 mM NaPO,, 0.5 M NaCl, 10 mM fi-mercaptoethanol, 10 mM

maltose (pH 7.0).

Antibody production - Approximately 50 ug of the purified fusion protein was
suspended in Freund’s complete adjuvant and subcutaneously injected into a New
Zealand white rabbit. Three and five weeks after the initial injection, the rabbit was
reinjected with 25 ug of the protein suspended in Freund’s incomplete adjuvant.

The rabbit was bled at two week intervals after the final injection for a total of 10

56

weeks.

Approximately 30 mg of purified LTP-MalE fusion protein was to bound to
3 ml Affigel 15 (Bio Rad, Richmond, CA) according to the manufacturer’s
instructions. Fifteen ml crude antiserum was added to the affinity column, the
column was washed with several volumes of 25 mM NaHzPO4 pH 7.4, and the anti
LTP antibody was eluted with 3 ml 200 mM glycine pH 2.8. The specificity of the

affinity purified antibody was checked by Western analysis.

Protein isolation and protein gel blot analysis - An Arabidopsis protein
extract was prepared by homogenizing tissue in 20 mM glycine pH 8.4, 5 mM
MgCl2, 2.5 mM EDTA. Extracts were centrifuged twice at 10,000 X g for 10 min.
Protein concentration was determined by the BCA protein assay (Pierce Biochem.
Rockford, IL). Forty ug of protein was electrophoretically resolved on a 10—20%
gradient SDS-polyacrylamide gel. Proteins were transferred to nitrocellulose and
Western analysis was carried out essentially as described by Towbin, et al. (1979).
Goat anti-rabbit lgG-alkaline phosphatase conjugate was used as the second
antibody. Alkaline phosphatase activity was revealed using nitroblue tetrazolium

and 5-bromo-4-chloro-3 indolyl phosphate as substrates.

Fixation of tissue - Plant tissue was fixed for 1.5 h at room temperature in
2% (v/v) glutaraldehyde in 50 mM potassium phosphate, pH 7.0. The samples

were washed three times with phosphate buffer and post-fixed for 1 h in 1% 030.

57

in 50 mM potassium phosphate buffer, pH 7.0 at room temperature. The tissue
was washed three times in phosphate buffer, dehydrated in a graded series of
acetone (10-95%), and infiltrated with LR White resin (Polysciences, Warrington,
PA). Infiltration was carried out over 36 h with several changes of resin.
Dehydration and infiltration were carried out at 4°C. The resin was polymerized for

24 h at 55‘?C, in 1 ml gelatin capsules.

Immunogold labeling for electron microscopy - Ultrathin sections (gold,
approximately 80-100 nm, on average 40 sections of each of 5 sample prepartions
of each tissue type) of various tissues were mounted on 300 mesh nickel grids.
The sections on grids were treated with 5% NalO, for 30 min and 0.1 N HCI for 10
min. After a 5 min wash in distilled water, the grids were incubated in TBST (10
mM Tris, 150 mM NaCl, 0.05% Tween 20, 0.02%NaN3, adjusted to pH 7.5 with HCI)
containing 1% BSA for 20 min. The sections were labeled overnight with a 1:20
dilution of the primary antibody in TBST-BSA. After labeling, the grids were
washed 3 times with TBST and then treated with a 1:50 dilution of 15 nm colloidal
gold-conjugated protein A (EY Labs, San Mateo, CA) for 45 min. The grids were
I washed three times in TBST, treated with 1% glutaraldehyde in 0.05 M potassium
phosphate buffer for 5 minutes and rinsed in distilled water. The sections were
stained for 30 min with a saturated uranyl acetate solution and 5 min with
Reynold’s lead citrate. The sections were viewed and micrographs were taken

with a Philips 201 electron microscope (Philips Netherlands, Eindhoven).

58
Immunocytochemical controls - The specificity of labeling was tested by (1)

labeling with preimmune serum in place of the anti-LTP primary antibody; (2)
incubating sections with secondary antibody without prior exposure to primary
antibody; (3) incubating antibody with maltose binding proteinzLTP fusion protein
before labeling; (4) treating sections made from transgenic plants in which the
presence of LTP was reduced by the expression of an antisense construct of the

LTP cDNA.

RESULTS

Protein expression and antibody producfi'on - In order to produce anti-LTP
antibodies of suitable quality for immunoelectron microscopy, two different LTP-
fusion proteins were constructed which permitted facile purification of the LTP by
affinity chromatography. A PCR product corresponding to the complete open
reading frame of the mature Arabidopsis LTP was subcloned into the plasmid
pRIT2T, to yield the plasmid pSLT2 (Fig. 3-1A). This plasmid, which contains the
LTP open reading frame fused to the protein A affinity tail, permits purification of
the fusion protein on an lgG column. The same PCR product was subcloned into
the plasmid le821, to yield pSLT3 (Fig. 3-1 B). This plasmid, which contains the
LTP open reading frame fused to the E. coli maltose binding protein, permits
affinity purification of the fusion protein on an amylose column.

Expression of an Arabidopsis LTP:protein A fusion in E. coli facilitated

purification of a fusion protein which was uncontaminated by other plant proteins

59

 

BamI-ll

Smal

chl g /

EcoRl
§ /
'Zgzgij; R
Protein A Amp //

A

Sac1.Kpn1.Eagl.Baml-l1

 

Figure 3-1. Structure of plasmids used to produce LTP fusion proteins.

A) The LTP open reading frame is fused to the affinity tail of staphylococcal
protein A. Abbreviations: Tail, 3’end of protein A gene; AmpR, ampicillin
resistance; ori, origin of replication.

B) The LTP open reading frame is fused to malE, the E. coli maltose binding
protein. Abbreviations: laqlq, hyper-expressed lac repressor; P tac, tac promoter;

malE, maltose binding protein; lacZ, fi-galactosidase; AmpR, ampicillin resistance;
ori, origin of replication.

60
(Fig. 3-2). Western analysis using antibodies raised against the purified

LTP:protein A fusion showed that the antibody recognized a protein in crude
Arabidopsis leaf extracts with a molecular weight of about 9 kDa, which is the size
expected for the mature LTP. However, there was weak cross reaction with
several other proteins (data not shown). To ensure that the anti-LTP antibodies
would cross react only with LTP during immunolocalization, the antibodies were
purified by affinity chromatography on an LTszaltose binding protein fusion
coupled to Affigel 15. The resulting antibodies recognized a single 9 kDa protein
in Western blots of total protein extracts from Arabidopsis leaves (Fig. 3-3). No

immunostaining was observed with the preimmune control serum (Fig. 3-3).

Arabidopsis L TP is found in many tissues - Protein extracts were made from
a variety of tissue types, including root, leaf, petiole, stem, silique, and flower.
Western analysis showed that LTP was present in all tissue types examined (Fig.
3-4). Expressed as a percentage of total protein, the LTP level in the stem was the
highest of all tissue types examined. The LTP level in flowers was higher than that
in leaves, petioles, siliques, and roots. LTP was detected in root tissue, although

at much lower levels than in other plant tissue (Fig. 3-4).

Arabidopsis LTP has an extracellular location - Immunogold labeling at the
ultrastructural level was used to determine the specific cellular location of LTP.

Examination of sections from leaf, petiole, stem, root, and flower showed that the

61

 

97 -

 

43 -
w'-

31 -

21 -

 

 

 

Figure 3-2. SDS-PAGE of purified LTP-protein A fusion from E. coli.

Proteins were expressed in E. coli, purified by lgG affinity chromatography, and
resolved by SDS-PAGE. Lane A) Protein A; Lane B) LTP-protein A fusion; Lane
C) LTP-protein A fusion after further purification by anion exchange chromatography.

62

 

46 -

31 -1

 

 

 

Figure 3-3. Western blot analysis of an Arabidopsis leaf protein extract produced
with anti-LTP:protein A antibodies.

An Arabidopsis leaf protein extract (40 pg per lane) was resolved by SDS-PAGE,
transferred to nitrocellulose, and immunodetected with affinity purified anti-LTP
antibodies (Lane I) or preimmune serum (P).

63

ABCDEFG

 

 

 

 

Figure 3-4. Tissue distribution of LTP.

Protein extracts (40 pg per lane) were prepared from various tissues of
Arabidopsis plants, resolved by SDS-PAGE, transferred to nitrocellulose and
immunodetected with anti-LTP antibodies. A) root; B) petiole; C) newly emerging
leaf; D) fully expanded leaf; E) stem; F) silique; G) flower.

54
LTP was located outside of the plasma membrane in the cell wall (Figs. 3-5, 3-6,

3-7). The highest concentration of LTP was seen in the epidermal cell walls in the
aerial parts of the plant. Figure 4 shows the immunolabeling of epidermal cell
walls from sections through the tip of a mature leaf (Fig. 3-5A), a petiole (Fig. 3-
50), and a stem (Fig. 3-5E). The labeling in the leaf, petiole, and stem sections
was confined to the cell wall and cuticle. There was essentially no labeling when
preimmune serum is used in place of anti LTP antibodies (Fig. 3-58, D, F), or when
anti-LTP antibodies were quenched with an LTP:maltose binding protein fusion
protein before labeling (data not shown). Thus, the labeling was specific for LTP.
Examination of sections of leaf tissue indicated that the protein was found
in equal amounts in both the upper and lower epidermis, and revealed that the
walls of guard cells were much more heavily labeled than the walls of other leaf
epidermal cells (Fig. 3-5A, 3-6A). Labeling was observed around the entire
epidermal cell; the outer periclinal wall, the anticlinal wall, and the inner periclinal
wall (data not shown). Labeling was also observed in cell walls of stem cortical
cells (Fig. 3-68), and petiole cortical cells (data not shown). Light labeling was
also observed in the walls of leaf mesophyll cells. In stem cortical cells, labeling
was more intense in the cell walls adjacent to intercellular space than in walls
between two adjacent cortical cells (Fig. 3-7B). An unusual type of cellular
material, possibly a secretory product, was seen in certain intercellular spaces of
stem cortical cells. The exact nature of this material is unknown, but it was heavily

labeled (data not shown). Labeling was observed in the secondary walls of

65

Figure 3-5. Immunocytochemical localization of LTP in the cell wall of Arabidopsis
leaf tip, stem and petiole using immunogold labeling.

Anti-LTP antibodies label the cell wall of leaf, petiole and stem epidermal cells. On
average, 40 sections of each of 5 sample preparations of each tissue type were
examined for each treatment (using immune serum and preimmune serum). There
is essentially no labeling when preimmune serum is used in place of anti-LTP
serum.

A) Leaf tip, longitudinal section, treated with anti-LTP serum.

8) Leaf tip, longitudinal section, treated with preimmune serum.
C) Petiole epidermal cell wall, treated with anti-LTP serum.

D) Petiole epidermal cell wall, treated with preimmune serum.
E) Stem epidermal cell wall, treated with anti-LTP serum.

F) Stern epidermal cell wall, treated with preimmune serum.

A-F, Bar = 0.25 pm
cu, cuticle; cw, cell wall; cyt, cytoplasm; v, vacuole

 

67

Figure 3-6. Immunocytochemical localization of LTP in Arabidopsis guard cells,
stem cortical cells, and xylem vessels using immunogold labeling.

Anti-LTP antibodies label the cell wall surrounding guard cells and stem cortical
cells. On average, 40 sections of each of 5 sample preparations of each tissue
type were examined for each treatment (using immune serum and preimmune
serum).

A) Guard cells of leaf epidermis, longitudinal section, treated with anti-LTP serum.
Gold particles are present in the cell wall surrounding these cells. Bar = 0.5 pm
B) Stem cortical cells, treated with anti-LTP serum. Gold particles are present in
the cell walls, most notably of cells adjacent to intercellular space (arrowheads).
Bar = 0.25 pm

C) Xylem vessel from stem, treated with anti-LTP serum. The secondary cell wall
is labeled. Bar = 0.5 pm

D) Xylem vessel from stem, treated with preimmune serum. The secondary cell
. wall is labeled. Bar = 0.5 pm

cu, cuticle; cw, cell wall; cyt, cytoplasm; is, intercellular space; sw, secondary
cell wall; v, vacuole; xp, xylem parenchyma

68

w. _,.. e. fee ..

_ ,. flee: .

 

69

Figure 3-7. Immunocytochemical localization of LTP in Arabidopsis root and pistil
using immunogold labeling.

Anti-LTP antibodies label the root epidermal cell walls, the stigma papillae and cell
wall at the stigma surface, and the intercellular space between the parenchymous
cells directly below the stigma surface. On average, 40 sections of each of 5
sample preparations of each tissue type were examined for each treatment (using
immune serum and preimmune serum).

A) Root epidermal cell wall, treated with anti-LTP serum. Gold particles are seen
in the wall.

8) Two root endodermal cells, treated with anti-LTP serum. Note the absence of
gold particles associated with these cells.

C) Stigma surface, longitudinal section, treated with anti-LTP serum. Gold
particles are seen associated with the cell wall and stigma papillae.

D) Stigma surface, longitudinal section, treated with preimmune serum. Note the
absense of gold particles.

E) lntercellular space of stigma, treated with anti-LTP serum. Notice the heavy
labeling over this area.

F) lntercellular space of stigma, treated with preimmune serum. Note the absence

of labeling in this area.

A-F, Bar = 0.25 pm
cu, cuticle; cw, cell wall; cyt, cytoplasm; ml, middle lamella; v, vacuole

70

 

71

vascular tissue. However, this labeling occurred when either anti-LTP antiserum
or preimmune serum was used (Fig. 3-6C, D), indicating that the labeling was not
specific for LTP. There was light labeling of root epidermal walls (Fig. 3-7A) but
no labeling associated with endodermal cells (Fig. 3-7B) or with any other cell type
in the root. LTP was present in the walls of the cells on the surface of the stigma
(Fig. 3-7C). and in the cell walls of the stigma parenchyma. Fig. 3-7E shows the
junction between four parenchymous cells just below the surface of the stigma in
which labeling was associated with the middle lamella. Essentially no labeling was
associated with the stigma surface or intercellular space when preimmune serum

is used in place of anti-LTP antibodies (Fig. 3-7D, F).

DISCUSSION

Lipid transfer proteins were originally isolated based on their ability to
catalyze lipid transfer between membranes in viva. However, it has not been
possible to assign an in viva function to these proteins. A yeast
phosphatidylinositol transfer protein (Pl-TP) is the only LTP to which the in vitro
activity correlates to a function in viva (Bankaitis, et al., 1990). This protein
A appears to be involved in elevating the phosphatidylinositol/phosphatidylcholine
ratios of yeast Golgi membranes by an unknown mechanism (Cleves, et al., 1991).
However, the nsLTPs from plants lack sequence homology to the yeast Pl-TP and
thus, likely, have a different biological role.

Although it has long been assumed that LTPs are cytosolic (Helmkamp,

72

1986), there is no evidence to support such an assumption, and recent evidence
indicates that plant LTPs are not located in the cytosol, but are secreted
(Bernhard, et al., 1990; Mundy and Rogers, 1986; Sterk, et al., 1991). Using
immunocytochemical techniques and light microscopy, Sossountzov, et al. (1991),
demonstrated that a maize LTP is found at the periphery of leaf and coleoptile
epidermal .cells and suggested that the staining was associated with the
cytoplasm. However, since light microscopic techniques lack the capability to
resolve the plasma membrane, it is generally necessary to use electron
microscopic imaging techniques to determine on which side of the plasma
membrane a protein is located. In order to resolve this issue I have examined an
extensive series of ultrathin sections from all major tissues of mature Arabidopsis
plants by immunoelectron microscopy using highly purified monospecific
antibodies. My results indicate that the LTP is primarily or solely located in the cell
wall and cuticle of all cell types where it is represented. Although this technique
cannot exclude the possibility of relatively low levels of accumulation in other
cellular compartments, the weight of evidence appears to be inconsistent with a
cytosolic location. In particular, since the Arabidopsis LTP contains a signal
peptide and the similar proteins in spinach and barley enter the secretory pathway
(Bernhard, et al., 1991; Madrid, 1991), it was not unexpected to find this protein
in the cell wall. This localization, however, would seem to preclude any direct
involvement of this LTP in lipid transfer between organellar and subcellular

membranes.

73
The Arabidopsis LTP was detected in the epidermal cell walls of all tissue

types examined; leaves petioles, stems, flowers, and, to a minor extent, roots. The
amount of labeling seen in the stem is higher than that seen in other green tissues
examined. Guard cells were more heavily labeled than the surrounding epidermal
cells. Although it has been reported that LTP is strongly expressed in vascular
strands (Sossountzov, et al.,1991), I did not observe specific labeling of vascular
tissue. Western analysis shows that the Arabidopsis LTP is present in low
concentrations in roots, and immunocytochemical labeling shows that this protein
can be detected in the cell walls of epidermal cells. LTP is present in the walls of
cells on the stigma surface and is also associated with the walls of the
parenchymous cells of the stigma. The epidermal and subepidermal layers of a
stigma can produce a secretion that eventually form a film over the epidermal walls
(Esau, 1977). The secretion contains mainly lipid and phenolic compounds, chiefly
anthocyanins, flavonoids, and cinnamic acids (Martin and Brewbaker, 1971 ). LTP
could, in principle, play a role in deposition of the lipid components of the
secretion.

The preferential accumulation of LTP in the cell walls and cutin layer of
epidermal cells is consistent with the results reported by Sterk, at al. (1991) and
Sossountzov, et al. (1991), and is consistent with speculation that the protein might
play a role in cuticle formation (Sterk, etal, 1991). The cuticle is a continuous layer
of predominantly lipophilic material found on the outermost surface of the aerial

parts of plants (Holloway, 1982). The major component of the cuticle is cutin, an

74

insoluble polymer composed mainly of C., and C18 hydroxylated fatty acids
(Holloway, 1982). The cuticle also contains wax, which is composed of a diverse
group of long chain hydrocarbons, including long chain fatty acids esterified with
alcohols, and free aldehydes, ketones, and alcohols (Kolattukudy, 1980). The
aliphatic chains present in plant waxes and cutin are generated within the
epidermal cell and are delivered to the outside of the cell where it is thought that
esterification takes place (Croteau and Kolattukudy, 1974). It is possible that LTPs,
which have a demonstrated ability to bind fatty acids (Rickers, et al., 1985), could
bind, and therefore solubilize, wax or cutin monomers upon release from the
plasma membrane and deliver them to the developing cuticle where they would
be available for polymerization. However, since it is not generally accepted that
roots contain cutin, the presence of LTP in this tissue is inconsistent with a role for
this protein in cuticle formation.

Another possible function of plant LTPs could be a general role in plant
defense. A protein, which is homologous to nsLTPs, has recently been isolated
from radish seeds and shows antifungal activity in vitro (W. Broekaert, personal
communication). Also, thionins, which have toxic effects on bacteria, fungi, and
I insects (Bohlman and Apel, 1991), have several features in common with plant
LTPs; they are low molecular weight, basic proteins with several conserved
cysteine residues. A barley leaf thionin is found in the cell wall, with the highest
concentration in the outer wall of the epidermal cell layer (Bohlman, et al., 1988:

Reimann-Philip, et al.. 1989). This localization is similar to that of Arabidopsis LTP

75

in all green tissue examined.

A complication concerning the possible role of LTP is that in some plant
spades there are isoforms which might have different functions. Four isoforms of
LTP isolated from castor bean seedlings are differentially expressed in a tissue-
specific manner (T suboi, et al., 1991). In a preliminary report, Yamada and
coworkers.(1990) have reported that in castor bean cotyledons, the LTP-A isoform
is present in glyoxosomes and in the secondary wall of xylem vessels. This
observation is very surprising since different mechanisms are normally required to
target proteins to these cellular compartments. Two different LTPs have been
found in tobacco, one of which is present predominantly in tapetal cells (Koltunow,
et al., 1990). A second tobacco LTP gene is expressed in all aerial portions of the
plant (Fleming, et al., 1992). A second gene was detected in Arabidopsis by
probing Southern blots at high stringency with the LTP cDNA (chapter 4 of this
dissertation). A second LTP from Arabidopsis has recently been isolated (F.
Garcia-Olmeda, personal communication), and it is possible that this isoform has
a differing localization than that described in this chapter. The anti-LTP antibodies
which are described in this chapter will be tested for cross reaction with the
recently isolated isoform.

In conclusion, due to its extracytoplasmic location, the nsLTP from
Arabidopsis is unlikely to be directly involved in intracellular lipid transfer. Thus,
assuming that the lipid transfer activity is not incidental to the function of the

protein, it is possible that this protein is involved in cuticle formation or in some

76

aspect of general plant defense. Analysis of transgenic plants in which the amount
of nsLTP has been greatly reduced by expression of an antisense construct of the
LTP cDNA (S. Thoma, U. Hecht, and C. Somerville, unpublished results) may

facilitate the elucidation of an in viva function for plant nsLTPs.

REFERENCES

Arondel, V., Tchang, F., Vergnolle, Ch., Jolllot. A., Grosbois, M., Guerbette, F.,
March, M.-D.. Pemollet, J.-C., Delseny, M., Pulgdomenech, P., and Kader, J.-C.
1989. Plant lipid-transfer proteins: structure, function, and molecular biology. In:
Biological role of plant lipids. PA. Biacs, K Gruiz, and T. Kremmer, eds. Akadémiai

Kiadé, Budapest, pp. 341 -350.
Arondel V and Kader, J.-C. 1990. Lipid transfer in plants. Experentia 46.579-585

Balltallls, VA, Allken, J.R., Cloves. AE. and Dowhan, W. 1990. An essential role
fora phospholipid transfer protein in yeast Golgi function. Nature 347: 561 -562.

Bernhard, W.R. and Somervllle, C.R. 1989. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from

amino acid sequence homology. Arch. Biochem. Biophys. 269 695-697

Bernhard, W.R., Thoma. S., Botella, J. and Somervllle. C.R. 1991. Isolation of a
cDNA alone for spinach lipid transfer protein and evidence that the protein if

synthesized by the secretory pathway. Plant Physrol. 95 164-170

Bohlman H Clauserl S Behnke, 8., Glass, H.. HIIIer, C., Relnmann-Phlllpp, U.,

Salvador, 6., Burkholt, V. and Apel, K. 1988. Leaf-speCific thionins of barley a
novel class of cell wall proteins toxic to plant pathogenic fungi and possibly

involved in the defense mechanism of plants. EMBO J. 7. 1559-1565
Bohlman, H. andApel, K. 1991. Thionins. Annu. Rev. Plant Physiol. Plant Mol. Biol.

42: 227-240.

Cloves, A.E., McGee, T. and Bankaltls, V. 1991. Phospholipid transfer proteins a
biological debut. Trends Cell Biol. 1: 30-34.

Croteau. R. and Kolattukudy, PE. 1974. Biosynthe3is of hydroxyfatty aCld
POIymers. Enzymatic synthesis of cutin from monomer acids by cell-free
Preparations from the epidermis of Vicia faba leaves. Biochemistry 13: 3193-3202.

Douady, D., Grosbola, M., Guerbette, F. andKader. J..-C 1982. Purification ofa

basic phospholipid transfer protein from maize seedlings. Biochim. Biophys. Acta,

71 0:1 43-1 53.
Esau. K 1 977. Anatomy of seed plants, second edition. John Wiley and Sons, New

York.

78
Hamlin. A.J., Mandel, T. Hofmen, S.. Sterk, P., de Vrles, S.C. and Kuhlemeler, C.

1992. Expression pattern of a tobacco lipid transfer protein gene within the shoot

apex. Plant J., in press.

Halouay. P.J. 1982. Structure and biochemistry of plant cuticular membranes: an
overview. In: The plant cuticle. D.F Cutler, K.L Alvin, and CE. Price, eds.

Academic Press, London, pp. 1-32.

Kader, J.-C.. Jullenne, M. and Vergnolle, C. 1984. Purification and characterization
of a spinach-leaf protein capable of transferring phospholipids from liposomes to

mitochondria or chloroplasts. Eur. J. Biochem. 139: 411-416.

Kolatumdy, PE. 1980. Cutin, suberin, and waxes. In: The Biochemistry of Plants,
Vol.4, Lipids: Structure and Function. P.K. Stumpf and BE. Conn, eds. Academic

Press, NY. pp. 571 -645.
Kollunow, A.M., Truotlner, J., Cox, K.H., Wallroth, M. and Goldberg, RB. 1990.

Different temporal and spatial gene expression patterns occur during anther
development. Plant Cell 2:1201-1224.

Laemrnll. UK 1970. Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature 277: 680-685.

Madrld, SM. 1991. The barley lipid transfer protein is targeted into the lumen of
the endoplasmic reticulum. Plant Physiol. Biochem. 29: 695-703.

Maniatis, T., Frltsch, EF. and Sambrook, J. 1982. Molecular cloning. A laboratory
manual. Cold Spring Harbor, New York.

Martin, F.W. and Brewbaker, J.L 1971. The nature of stigmatic exudate and its role
in pollen germination. In: Pollen development and physiology. J. Heslop-Harrison,

ed. Butterworth, London, pp.262-272.

Mundv. J. and Rogers, J.C. 1986. Selective expression of a probable
amylase/protease inhibitor in barley aleurone cells: comparison to the barley

amylase/subtilisin inhibitor. Planta, 196: 51-63.

More, J-. Spener, F. and Kader, J.-C. 1985. A phospholipid transfer protein that
binds long-chain fatty acids. FEBS Lett. 180, 29-32.

3W. L, Ruiz-Avila, L, Vlgnols, F., Jolllot. A., Arondel, v., Tchang, F.,
Gm. M., Guerbette, F., Mlgniac, E., Delseny, M., Pulgdomenech, P. and

Kader, J.-C. 1991. Spatial and temporal expression of a maize lipid transfer protein
gene. Plant Cell 3: 923-933.

79

Slant, P., Bool, H.. Schellekena, GA, Van Kammen, A. and De Vrles, S.C. 1991.
Coll-specific expression of the carrot EP2 lipid transfer protein gene. Plant Cell 3:

907-921 .

Taldshlma. K., Watanabe, 8., Yamada, M. and Mamlya. G. 1986. The amino-acid
sequence of nonspecific lipid transfer protein from germinated castor bean.

Biochim. Biophys. Acta 870: 248-255.

Tchang, F., This, P., Stlefel, V., Arondel, V., March, M.-D., Pages, M.,

Pilgdomenech, P., Grellet, F., Deleeny, M., Boulllon, P., Huet, J.-C., Guerbette, F.,
Beams-Canto, F., Duranton, H.. Pemollet, J.-C. and Kader, J.-C. 1988.

Phospholipid transfer protein: full length cDNA and amino acid sequence in maize.
J. Biol. Chem. 263: 16849-16855.

Towbln, H.. Staehelln. T., and Gordon, J. 1979. Electrophoretic transfer of proteins
from polyacrylamide gels to nitrocellulose sheets: procedure and some practical

applications. Proc. Natl. Acad. Sci. USA 76: 4350-4354.
Vergnolle, C., Arondel, V., Tchmg, F., Grosbois, M., Guerbette, F., Jolllot. A., and
Kader, J.-C. 1989. Synthesis of lipid transfer proteins from maize seedlings.
Biochim. Biophys. Res. Commun. 157:37-41.

Witz. K.WA. and Gadella, Jr., T.W.J. 1990. Properties and modes of action of
specific and nonspecific phospholipid transfer proteins. Experentia 46: 592-599.

Yamada, M., Tsuboi, S., Osafune, T., Suga, T. and Taklshlma, K. 1990.
Multifunctional properties of non-specific lipid transfer protein from higher plants.
In: Biological role of plant lipids. P.A. Biacs, K Gruiz, and T. Kremmer, eds.

Akadémiai IGado, Budapest, pp. 278-280.

 

CHAPTER 4

AN ARABIDOPSIS UPID TRANSFER PROTEIN PROMOTER SPECIFIES COMPLEX
EXPRESSION IN TRANSGENIC PLANTS

80

81

ABSTRACT
Lipid transfer proteins (LTPs) are a class of proteins which have the ability

to stimulate phospholipid transfer between membranes in vitro. Based on their
extracellular location, however, it is unlikely that these proteins have a similar role
in viva. To better understand a possible role for these proteins, an LTP gene from
Arabidopsis has been cloned. The regulation of the LTP gene was examined by

analysis of GUS activity in transgenic plants containing LTP promoter-B-

glucuronidase fusions. The LTP promoter was active early in seedling

development and was strongly expressed in the tips of cotyledons, in the shoot

meristem, in stipules, and at the base of lateral roots. In adult plants, the gene

fusion was expressed in leaf guard cells, stem epidermal cells, and the stigma,

pollen grains and floral nectaries of mature flowers. Analysis of the promoter
region shows the presence of sequences homologous to putative regulatory
elements of several genes encoding proteins of the phenylpropanoid biosynthetic
pathway. These sequence data, along with the GUS expression pattern are
consistent with a role for the Arabidopsis nsLTP in phenylpropanoid metabolism

or regulation of expression by the same or similar mechanisms.

82
INTRODUCTION

The mechanisms by which lipids move from their sites of synthesis, the

endoplasmic reticulum (ER) and the chloroplast, to other cellular organelles, is
unknown. A class of proteins, called lipid transfer proteins (LTPs) was first isolated
based on the ability to transfer phospholipids between natural and artificial
membranes in vitro (Wirtz and Zilversmit, 1968). It has been difficult to assign an

in viva function to these proteins, but it has long been assumed that they are

 

involved in intracellular lipid transport (Helmkamp, 1990).

LTPs have been isolated from mammalian tissue (Kamp, et al., 1973;
Helmkamp, et al., 1974; for review see Wirtz, 1982), bacteria (T ai and Kaplan,
1985), yeast (Szolderits, et al., 1989; Paltauf and Daum, 1992), filamentous fungi
(Tan, et al., 1990), and several plant species, including spinach (Kader, et al., 1984,
Bouillon, et al., 1987), maize (Douady, et al., 1982), barley (Mundy and Rogers,
1986), castor (T akishima, et al., 1986), and sunflower (Arondel, et al., 1990). The

plant LTPs are all small, soluble, basic proteins. Unlike some of the mammalian

 

and yeast proteins, which transfer only a specific class of phospholipids, plant
LTPs have a broad specificity for the type of lipid they will transfer, and the spinach

and castor LTPs have a demonstrated ability to bind fatty acids (Rickers, et al.,
1985; Yamada, et al., 1990).

Based on their proposed function of intracellular lipid transfer, it has been
presumed that LTPs were cytosolic proteins (Arondel and Kader, 1990; Helmkamp,

1990). However, recent evidence demonstrates that plant LTPs are not located in

83

the cytoplasm. It has been reported that the maize LTP has an amino terminal
extension and is synthesized on membrane bound polysomes (T chang, et al.,
1988) and the spinach and barley LTPs contain signal peptides and are
cotranslationally inserted into the ER (Bernhard, et al., 1991; chapter 2 of this
dissertation; Madrid, 1991). Since these proteins lack the ER retention signal,
HIKDEL (Munro and Pelham, 1987), they would be expected to be secreted or
targeted to a specific organelle. In support of these expectations, Mundy and
Rogers (1986) reported that a barley protein, which was subsequently identified
as an LTP (Bernhard and Somerville, 1989; Breu, et al., 1989) was found in the
aleurone cell culture medium, and Sterk, et al., (1991) demonstrated that a carrot
LTP was secreted by embryogenic cell cultures. Sossountzov, et al. (1991) have
shown that a maize LTP is localized to the periphery of the epidermal cells in
maize coleoptiles, and suggest that the labeling is associated with the cytoplasmic
side of the plasma membrane. It has more recently been demonstrated, using
immunocytochemical labeling at the ultrastructural level, that an Arabidopsis LTP
is localized to the cell wall, and that it is mainly seen in epidermal cell walls
(Thoma, et al., 1993; chapter 3 of this dissertation).
With the mounting evidence that at least some plant nsLTPs are
extracellular, the proposed function of these proteins in intracellular lipid transfer
comes into question. Based on the observation that a carrot LTP is secreted and

the localization of an Arabidopsis LTP to the walls of epidermal cells, it has been

proposed that the protein may play a role in cuticle formation or in general plant

 

34
defense (Sterk, et al., 1991; Thoma, et al., 1993; chapter 3 of this dissertation).

However, analysis of transgenic plants which contain a reduced level of LTP due
to the presence of an Arabidopsis LTP cDNA in reverse orientation behind the 35S
promoter (chapter 5 of this dissertation), shows no alteration in wax or cutin
composition under normal growth conditions. To better understand this protein
and its possible function, a genomic clone corresponding to an Arabidopsis LTP
has been isolated. I show here that LTP is encoded by at least two genes in
Arabidopsis and have designated this clone ALTP1. A series of transgenic plants
has been produced which contains the LTP promoter fused to the B-glucuronidase
reporter gene. l have analyzed the temporal and spatial expression of this
promoter and have found it to be expressed most strongly in the stigma surface,
epidermal cells of the leaf and stem, floral nectaries, and pollen grains of mature
plants. The promoter was also active in early seedling development and was
strongly expressed in the tips of cotyledons, in the shoot meristem, in stipules, and

at the base of lateral roots. The pattern of expression suggested a role for this
protein in phenylpropanoid metabolism. Computer aided analysis of the promoter

region shows that it contains several putative regulatory sequences found in

promoters of biosynthetic genes for the phenylpropanoid pathway (Cramer, et al.,

1989; Lois, et al., 1989; Ohl, et al., 1990).

EXPERIMENTAL PROCEDURES

Plant Material - Arabidopsis thaliana ecotype RLD was used to produce all
transgenic plants as it had the best transformation efficiency of all ecotypes tested
(J. Schiefelbein, unpublished results). Transgenic seed were surface sterilized with
a solution containing 5% hypochlorite and 0.02% Triton X-100, rinsed several times
with sterile. water, and plated on Murashige-Skoog (MS) medium (Murashige and
Skoog, 1962) containing kanamycin (50 pg/ml). Plants were grown at 22°C under
continuous illumination (100-150 pE M‘s"). With the exception of plants used for
GUS expression during seedling development, kanamycin resistant seedlings were
transplanted to pots containing a mixture of perlitezvermiculite:sphagnum (1:1 :1)

and grown under the same light and temperature conditions.

Isolafian of an Arabidopsis L T? genomic clone - A LGEM11 Arabidopsis
(ecotype Columbia) leaf genomic DNA library was plated on Escherichia coli KW51
cells and resulting plaques were lifted onto nitrocellulose filters. The filters were
prehybridized for 15 min at 65°C in a solution containing 200 mM NaCl, 20 mM
sodium phosphate, pH 7.7, 2 mM EDTA, 1% SDS, 0.5% dry milk powder, 10%
dextran sulfate. Hybridization was carried out for 8 h at 65°C in the same solution
containing an Arabidopsis LTP cDNA which was labeled with [a-FP] dCTP by
random priming (Feinberg and Vogelstein, 1983). The filters were washed three
times for 10 min at 65°C in a solution containing 2X SSC, 0.1% SDS and once in

a solution containing 0.5x SSC. 0.5% SDS under the same conditions. A 1.9 kb

86

EcoRI fragment containing the complete coding sequence for the Arabidopsis
LTP1 gene and over 1 kb of the 5’ upstream region was subcloned into the
corresponding site of pBluescript (Stratagene, La Jolla, CA). Single stranded DNA
was sequenced by the dideoxy chain-termination method (Sanger, et al., 1977).

All nucleic acid manipulations were carried out as described in Maniatis, et al.

(1982).

Nucleic acid manipulations - DNA was extracted from leaves of three week
old Arabidopsis by grinding 10 9 tissue in liquid N2 and adding to 10 ml of a 65°C
solution containing 2% (w/v) hexadiacyltrimethyl ammonium bromide (CTAB), 100
mM Tris-Cl. pH 8, 20 mM EDTA, 1.4 M NaCl, 1% polyvinylpyrrolidone. The mixture
was extracted once with chloroformzisoamyl alcohol (24:1). The aqueous phase
(10 ml) was added to 1 ml 10% CTAB, 0.7 M NaCl and this was extracted again
with chloroformzisoamyl alcohol. Ten ml of a solution containing 1% CTAB, 50 mM
Tris-Cl, pH 8, and 10mM EDTA was added to the aqueous phase and the mixture
was centrifuged at 10,000 x g for 10 min. The pellet was suspended in 2 ml high
salt TE (10 mM tris-Cl, pH 8, 1 mM EDTA, 1 M NaCI). The DNA was recovered by
1 ethanol precipitation and finally resuspended in TE.

For Southern analysis, DNA was digested and resolved by electrophoresis
through 0.8% agarose (1.7 pg/lane). The gel was then treated for 15 min with 0.25
N HCI, followed by 30 min in denaturation solution (0.5 M NaOH, 1.5 M NaCl) and

15 min in alkali transfer buffer (0.25 M NaOH, 1.5 M NaCI). The DNA was

87

transferred, in alkali transfer buffer, to a nylon membrane (Hybond N+,
Amersham). The membrane was prehybridized for 3 h at 65°C in a solution
containing 4X SET (3 M NaCI, 0.6 M Tris-HCI pH 7.4, 40 mM EDTA), 0.1 %
Na4P207, 0.2% SDS, 0.1% heparin. Hybridization was carried out at 65°C in a
solution containing 4X SET, 0.1% Na4P207, 0.2 % SDS, 10% dextran sulfate and an
Arabidopsis LTP gene fragment labeled with [a-“deCTP by random priming. The
membranes were washed for 15 min at 65°C in 2X SSC, 0.1% SDS, 15 min at 65°C

in 1X SSC, 0.1% SDS, and 15 min at 65°C in 0.1x SSC, 0.1%SDS.

Production of L TP promoter - B—glucuranidase constructs - Synthetic
oligonucleotides homologous to portions of the promoter and 5’ untranslated
region of the genomic clone were used as primers to amplify the promoter
fragment by PCR (Table 4-1). The upstream (5’) primers (UH20, UH21, UH23)
contained a recognition sequence for Hindlll and the downstream (3’) primer
(UH19) contained a BamHI recognition sequence. The amplified product was
digested with the appropriate restriction enzymes and subcloned into the
corresponding sites of pBl101. Three constructs were prepared in this manner,
one which contained about 1000 bp of the promoter fragment, one which

contained 700 bp of the promoter region, and another which contained 280 bp of

the promoter region.

 

 

 

88
Production of transgenic plants and screen for GUS activity The promoter-

GUS fusion constructs were transformed into Agrabacterium tumefaciens, strain
C58/pGV3850, by electroporation. Treated Agrabacterium were spread on LB

plates containing kanamycin (50 pg/ml), resistant colonies were picked, and a

 

TI

plasmid miniprep was performed to ascertain the presence of the plasmid. This

Agrabacterium was then used to transform Arabidopsis maliana var. RLD, by the

 

root transformation method described by Valvekens, et al. (1988). Few plants ‘I
produced roots, and thus were allowed to set seed in culture. The seed were
sown on MS plates containing kanamycin (50 pg/ml). Plants which were resistant
to kanamycin were transplanted to pots as soon as their primary leaves developed
and these plants were allowed to set seed. Leaves from transgenic plants were
homogenized in GUS lysis buffer (50 mM Na P0,, pH 7.0, 1 mM EDTA, 0.1% Triton
X-100, 0.7% B-mercaptoethanol) containing the chromogenic substrate 5-bromo-4-

chloro-3-indolyl B-D-glucuronide (X-gluc) and incubated at 37°C. The samples

were checked visually for appearance of blue color.

Histochemical localization of GUS activity - Plant tissue; leaves, stems,
petioles, siliques, flowers, and roots, were removed from the plant and placed in
100 mM phosphate buffer, pH 7.0, containing 1 mM spermidine and vacuum
infiltration was carried out for 15 minutes. Tissue was then incubated at 37°C with
a solution of 0.5 M KeFCN, 0.01% Triton-X100, 50 mM NaPO,, pH 7.0, 10 mM 8-

mercaptoethanol, 1 mM EDTA, 2 mM X-gluc, until a blue color appeared (usually

 

Table 4-1. Primer sequences used to amplify promoter

fragments from ALTP1.

 

UH21

ACCAAAAAGCTTAATGTAITTTGGTCGAAT

 

UH23

ACCAAAAGCI l | ICTTATTAGAGTCATGT

 

UH20
UH1 9

“ACCAAAAAGCTI’AATCTCAAAACCAAAGTC

 

ACCAAAGGATCCATA'ITGATCTCTTAGGTA

 

.‘;..fl£‘ -I' L

 

 

90
after 5 hours). Following incubation with the substrate, dehydration and

embedding were carried out essentially as described in De Block and Debrouwer
(1992), except that the activator concentration in the embedding step was
decreased from 0.8% to 0.6%. Eight micrometer sections were cut with a dry
glass knife and sections were flattened on a drop of water. Sections were viewed

using bright field microscopy on an Axiophot microscope (Zeiss).

 

Waunding and pathogen infection - Leaves of three week old plants were
mechanically wounded using a Dr. Scholl’s callous remover. Two hours after
wounding, both wounded and non-wounded leaves were removed from plants,
and leaves were stained with X-Gluc. Leaves of three week old plants were also
infected with the Arabidopsis fungal pathogen, Erisyphe cruciferae. Inoculation
was carried out by rubbing spores from an infected Arabidopsis leaf onto a test
leaf. Plants which were inoculated with fungus were marked with a drop of nail

polish on the tip of the leaf. Two days after infection, leaves were removed from

the plant and stained with X-Gluc.

 

RESULTS

Cloning and characterization of an Arabidopsis LTP (AL TP1 ) gene - An

Arabidopsis LTP genomic clone was isolated by screening a genomic library made

from Arabidopsis leaf tissue (J.C. Schneider, unpublished) with a labeled

91
Arabidopsis LTP cDNA (chapter 5 of this dissertation; Genbank accession

#M80566).

To determine the complexity of LTP genes in Arabidopsis, genomic DNA
was digested with a variety of restriction enzymes and hybridized with the labeled
Arabidopsis LTP cDNA. There are no cleavage sites for BamHl, Hindlll, Kpnl, and
Xbal within the DNA sequence (as determined from sequence analysis of the
genomic clone. Thus, one band would be expected on a Southern if there were
one gene. There are cleavage sites for EcoRV and Pvull near the middle of the
coding sequence. Two bands would be expected on a Southern if there were only
one gene. Under stringent hybridization and washing conditions, two to four
bands of different intensities were observed (Fig. 4-1), indicating that there are at

least two LTP genes in Arabidopsis.

A 1950 bp fragment, which included 1150 bp of the 5’ untranslated
region/promoter and 500 bp coding region was sequenced (Fig. 4-2). The
deduced amino acid sequence exactly matched that of the Arabidopsis LTP cDNA
(chapter 5 of this dissertation). Sequence comparison indicated that the coding
region contained a 130 bp intron, from position 396 to 526. The sequence exactly
matched that of the previously identified ALTP cDNA except for one Val to Gly

difference at position 528. This change is due to a one nucleotide change within
the coding sequence (the conversion of T-G in the second position of the codon).
The 5’ untranslated region of the LTP gene showed no overall sequence homology

to GenBank Release 70 listings, but several small regions were found which are

 

 

92

BEHKPX

23.1 -
. . ,.
6.6 - w
4.4 - .

 

K
2.3 -
2.0- , g
- -
-
0.56 -

 

 

 

Figure 4-1. Southern blot of Arabidopsis genomic DNA probed with an
Arabidopsis LTP cDNA.

Aliquots of Arabidopsis genomic DNA (1.7 ug) were digested with BamHI (B),
EcoRV (E), Hindlll (H), Kpnl (K), Pvull (P), and Xbal (X), separated on an agarose
gel, transferred to a nylon membrane, and hybridized with a labeled ALTP cDNA.
The blot was hybridized and washed under high stringency conditions. Size
markers on the left are in kilobases.

93

Figure 42. Sequence of ALTP1 promoter and coding region.

The putative transcription start site is at position +1 and a putative TATA box is
given in bold letters. The deduced amino acid sequence is shown below the
nucleotide sequence which contains a 130 bp intron from positions 396-526. The
signal peptide is underlined. Sequence elements corresponding to
phenylpropanoid promoters are indicated by shaded boxes. The regions used for
the various ALTP1 promoter-GUS fusion are marked with an arrow (r->). (This
sequence was determined by U. Hecht).

 

94

-1156 mmmmcmmmmmmmmcmcmca ~1097
nepotismoo
-1096 rmoccmcrcmcmccacmmmmcmunmcmmc -1037
-1on mnummmmmcmnmccmucmmccmcrm -977
-m tmmmummncmmcrmmmcr -917
-915 nmmnuccmmcnmcnmmmcrcrcmmxcmm -857
-356 tmmamacccmccmnmcmcmcmcmm -797
-796 mnmmmmmccmmwacmcmcnmmn -7 37
-736 mmmmmmmmmmmmm -677
cvpcusno
-676 mmmmcmmmmcmmccumcmmmcmu —617
-616 mummnccmunuucmmmmnmmcm -557
-556 mmummcmmnmmnmmmrmmmm 497
-496 cmmummmnmcmmmmm 437
«35 mmnmmcmmucmcmcwmmc ~377

-376 WMMGWWWMWTMMWCT -317

 

' ’ ' "epcuszeo
-256 macmmmmcmmmcrmcsammmmc -197

-196 nmmmmmmmcmmmmccrumnma -137
-136 mmcmmmcmmmcmemmmcmccmum -77

..............................
...............................

MH‘EACTCACTCCCCM -1 7

 

~76 CTATCGCATTCACACCACATAACAT
-16 mommmcn -1

 

123 mommmmmcmmmccmwccmcmccc 182
§EAAL8CGSVASSNLAACIG

1 83 rmmcammcmcccccmmceocmmmm 2 4 2
YVLQGGVIPPACCSGVKNLN

243 AGCATAGCCWGMCCCAGACCGTCAGCWWWWCGCT 302
SIARTTPDRQQACNCIQGAA

303 MMWWMCWWCCTWMGTGGA 362
RALGSGLNAGRAAGIPKACG

3 6 3 cremmccmmnmcmcmcmcmcmccmcrnmmcrm 4 2 2
VNIPYRISTSTNCKT

l 2 3 cmmmmnmcmcmmmcmn 4 8 2

4 8 3 ATTRACGGTGHMMWMMWAGCGTGAGGTGATGAGCTAGCA 54 2
G D E L A

54 3 ”WMMCWCMCCNMATMWGGWA 602
T V R "

6 O 3 mmmmmccmmmmmmcmcmcr 6 6 2
6 6 3 WMWGTMTMMGCMGCWWMMA 7 2 2
7 2 3 remcmmommuummmmcmm 7 8 2
7 8 3 1'36”!ch 7 9 5

 

 

95

conserved among promoters of phenylpropanoid biosynthetic genes (Figs. 4-2, 4-
3). A putative TATA box was found 117 bp from the translation initiation codon.
The putative TATA box had 75% homology to the consensus TATA sequence

TG/CTATAT/AA. A proposed transcription initiation site was determined by

comparison to transcription start sites of known plant genes (Joshi, 1987). An A
is at the transcription initiation site in 85% of genes examined, in the context

CTCATCA. This sequence closely resembles the sequence TCCATCA, which is

found in the ALTP1 gene.

Production of tansgenic Arabidopsis containing the LTP pramater:B-
glucuronidase fusion - To determine the expression pattern of LTP, various size
fragments of the LTP promoter were cloned into the binary vector pBl101 in front
of a promoterless B-glucuronidase (GUS) gene. The resulting constructs (Fig. 4-4)
were transformed into Arabidopsis by the root transformation method (Valvekens,
et al., 1988). Control plants containing GUS driven by the CaMV promoter were
also generated. Leaves from regenerated plants were homogenized and checked

visually for the appearance of a blue color in the presence of X-Gluc. Four lines
containing the 1 kb promoter fragment which had the highest levels of color
formation were chosen for further study. These lines were designated LTP-GUS
1A. 1 B, 10, and 10. Two strains containing the 700 bp fragment and one strain
Containing the 280 bp fragment also displayed color formation. These lines were

named LTP-GUS 700A, LTP-GUS 700B, and LTP-GUS 280A, respectively (See

in

Box 1 Arabi dopeis LTP1 -3 1 4 ACACMCT’ATAA
Arabi dopsis LTP1 -5 2 ACGTmaTAA
Parsley PAL -119 TCTCAcqw'ACCC
Bean PAL -8 9 ACCCACCIIMCA
Bean PAL +1 1 3 ACTCACCEACCC
Arabi dopsis PAL -435 ACACAL’CEI‘ACTC
Arabidopsis PAL -359 TCTCACCA'ACCG
Arabidopsis PAL -64 GCTtaccfthCA
consensus a c'rczicc'rAccg
BOX 2 Arabi dopsis LTP1 -l 1 0 ACMCCACCAAC
Parsley PAL —20 8 CCAACAAACCCC
Bean PAL -1 5 7 TCCACCAACCCC
Arabi dopsis PAL -1 6 3 1 CCAACACACCAC
Arabi dopsie PAL -1 0 9 TCAACCAACTCC
CONSENSUS §CAAC3AACC£C
Box 3 Arabi dopsis LTP1 +2 6 AACGAACAT
Bean CHS +23 MCCAACAA
Arabi dopsis PAL +23 MCCAACAA
CONSENSUS AACCAACAT

Figure 4-3. Comparison of putative regulatory sequences in ALTP1 and in genes
of the general phenylpropanoid pathway.

Putative regulatory sequences from bean, parsley and Arabidopsis phenylalanine
ammonia-lyase and bean chalcone synthase were aligned with the sequences
found in the ALTP1 clone. Matches of at least 5/8 (box 1), 3/5 (box 2), or 2/3 (box
3) are shown in the consensus sequence as full size letters. The positions of the
sequences are given relative to the proposed transcription start sites.

97

Figure 4—4. ALTP1- 8-glucuronidase constructs used to produce transgenic plants

1 Kb, 700 bp, and 280 bp fragments of the promoter and 5’ untranslated region
of the ALTP1 clone were placed in front of a promoterless B-glucuronidase gene
in pBl101. These constructs were transformed into Agrabacterium tumefaciens
and subsequently used to produce transgenic plants via the root transformation
method. (Constructs and transgenic plants were produced by U. Hecht).

98

 

a. eaeaeaeac - m

_EoEmE _ Lo. _

3 SN moz

Am EGG :EZ — mOZ—

 

36353 .c

 

 

 

 

 

 

Lo. Loo—ewe: .68an do.
8 88.6: - = “Z
ooh ESE .m
on: - , e2
mOZ 9928585 m _ Leos—mm: 8889:. no. _ —mOZ— Am :5: F52 —oa, mOZ

 

iii/its

(on

82-5.53 .<

99

Table 4-2 for a summary).

Expression during early seedling development - Seed from transgenic
plants were plated on MS medium containing kanamycin and seedlings were
stained with X-Gluc. GUS staining was strong in young seedlings which had
recently emerged from their seed coats (3 days after planting) in the hypocotyl
region adacent to the root (Fig. 4-5A). There was also light staining in the
cotyledon. At 5 days after planting, the entire cotyledons were weakly stained,
with more intense staining in the vascular tissue (Fig. 4-58). GUS staining was
also seen in the shoot meristem, in the primary leaves and in the stipules (Fig. 4-
5C). As the seedling developed, GUS staining in the cotyledon decreased and

eventually was confined to the tip of the cotyledon (Fig. 4-SD,F) and as the primary
leaves matured, GUS staining became localized to the tips of these structures as
well (Fig. 4-5E). After 10 days, staining was confined mainly to the stipules, with
weak staining in the shoot meristem (Fig. 4-11A). Sections through the shoot
meristem of 10 and 14 day old seedlings (Fig. 4-6) showed no staining in this
region, but intense staining was observed in the stipules. No staining was seen
in emerging roots, but as the roots developed, GUS staining was seen at the base

of lateral roots (Fig. 4-7A,B). Staining was occasionally observed at the tip of

lateral roots as they developed (Fig. 4-7C).

100

TABLE 4-2. Transgenic lines used for histochemical analysis of ALTP1 activity.

Relative

Promoter Staining
Length (bp) Intensity

I LTP-GUS 1A

LTP-GUS 1B
LTP-GUS 1C
LTP-GUS 1D
LTP-GUS 700A
LTP-GUS 7008

 

 

 

 

 

 

 

 

 

 

 

LTP-GUS 280A

 

101

Figure 4-5. Transgenic Arabidopsis seedlings after X-Gluc staining at different
developmental stages.
A. ALTP1 activity in 3 day old seedling.
B. ALTP1 activity in 5 day old seedling.
C. ALTP1 activity in the shoot meristem, stipules, and primary leaves of a
5 day old seedling.
D. ALTP1 activity in 6 day old seedling. The labeling in the cotyledon is
confined to the tip.
E. ALTP1 activity in the shoot meristem, stipules, and primary leaves of a
6 day old seedling.
F. ALTP1 activity near Iignified tissue at tip of cotyledon.

co, cotyledon; hy, hypocotyl; pl, primary leaf; r, root; sm, shoot meristem;
stip, stipule; t, trichome

 

102

 

103

Tantra

 

Figure 4-6. Longitudinal section through shoot meristem of 14 day old seedling.

pl, primary leaf; sl, secondary leaf; sm, shoot meristem; stip, stipule

104

Figure 4-7. ALTP1 activity in root tissue.
A. Lateral root forming on root of 5 day old seedling.
B. Lateral root forming on 5 day old seedling.
C. Lateral root of 14 day old seedling.

105

 

106
Expression of GUS in the adult plant - Tissue from kanamycin resistant

plants was collected, stained with X-Glu, and either observed directly or embedded
in Plastic and sectioned. Analysis of leaf tissue at 2, 3, and 4 weeks after planting
shows that GUS staining is most prominent in vascular tissue (Fig. 4-8A), guard
cells (Fig. 4-8B), hydathodes (Fig. 4-8C), and trichomes and their associated basal
cells (Fig. 4-BD). Weak staining was also observed in regular epidermal cells (Fig.
Fig. 4-8A). Stem tissue was taken and analyzed from three distinct regions of the
stem; at the base, from the middle, and from the top. Staining of epidermal cells,
cortical cells, and vascular tissue was observed in the top portions of the stem
(Fig. 4-8E). Staining was most prominent in the epidermal cells. Sections through
the lower portions of the stem revealed little or no visible staining (data not
shown).
Young flower buds (stage 6, Smyth et al., 1990) showed no GUS staining
(Fig. 4-9A), but as the flower developed, staining occurred. In the newly opened
flower (stages 11-12), weak staining was observed in the stigma (Fig. 4-98), and
in the mature flower (stage 15), intense staining was observed in the stigma (Fig.
4-9F). Anthers of stage 11-12 flowers did not exhibit staining (Fig. 4-90), but
Pollen grains of more mature flowers were intensely stained (fig. 4-9G). There was
little or no detectable staining in floral nectaries of stage 11 flowers (Fig. 4-90), but
as the flowers matured, staining developed in these structures (Fig. 4-9E). Whole
mounts of siliques shows strong staining is localized to the base and the tip of

these structures (Fig. 4-10A,B). There was weak staining observed in the ovary

107

Figure 4—8. ALTP1 activity in leaf and stem tissue.
A. Longitudinal section through 4 week, fully expanded leaf.
B. Longitudinal section through 4 week, fully expanded leaf.
C. Hydathode from 3 week old leaf.
D. Trichome from 3 week old leaf.

E. Cross section through stem of 6 week old plant. Section taken from top
of stem, near flowers.

co, cortical cells; epi, epidermis; gc, guard cell; v, vascular tissue

 

109

Figure 4-9. ALTP1 activity in flowers.

A. Longitudinal section through gynoecium of flower bud (stage 6, Smyth,
et al., 1990).

B. Longitudinal section through stigma and style of flower (stage 11).

C. Longitudinal section through anthers of stage 11 flower.

D. Longitudinal section through base of flower (stage 11).

E. Longitudinal section through base of mature flower (stage 15).

F. Longitudinal section through the stigma and style of mature flower (stage
15).

G. Longitudinal section through the anther of mature flower (stage 15).

n, nectary; ov, ovary; p, pollen; pe, petal; se, sepal; sti, stigma; sty, style

110

 

111

 

Figure 4-10. ALTP1 activity in siliques.
A. Tip of silique.
B. Base of silique. The arrowhead is pointing at zone of sepal and petal
abscission.

ov, ovary; ped, pedicel

1 12
wells (Fig. 4-10A).

Control plants transformed with pBl101. which contains a CaMV 35S-GUS
fusion, are shown in figure 4-11. The 35S promoter is active in most tissue and
cell types. In three day seedlings, staining is observed over the cotyledons,
hypocotyl and root (Fig. 4-11A). At five days (Fig. 4-11B) and six days (Fig. 4-
11C), staining is observed throughout the cotyledon, and is especially prominent
in the vascular tissue). Prominent staining is also observed in the root (Fig. 4-
11D). Light staining was observed in all cell types of the floral organs (Fig. 4-
11E,F) and of leaf tissue (Fig. 4-1 1 G). Staining was also observed throughout the
silique (Fig. 4-11H,I). No specific staining was observed in the pollen grains (Fig.
4-11J), and staining was present in the epidermal and cortical cells of stem tissue
(Fig. 4-1 1 K). Ught staining was also observed in stem pith cells (data not shown).
Preliminary analysis of GUS expression in plants containing the shortened
promoter fragments (Table 4-1), indicate that the activity of the promoter is similar
in all lines. Figure 4-12 represents analysis of 10 day old seedlings with truncated

promoter fragments.

Waunding and pathogen infecfian - Leaves of transgenic and control plants
were mechanically wounded and stained for GUS activity 2 hours after wounding.
There was no noticeable increase in staining after this treatment. Transgenic and
control plants were treated with the fungus, Erisyphe cruciferae, and leaves were

collected 48 h after infection. There was a marked increase in staining in the cells

113

Figure 4-11. Activity of CaMV 35$ promoter in transgenic Arabidopsis.

39"“;9 T" Fig.0.“)?

3 day old seedlings.
5 day old seedling.
6 day old seedling.
Primary root from 6 day old seedling.
Longitudinal section through floral organs. The staining is light, but is
present in all cells types at an equal intensity.
Longitudinal section through base of flower. Staining is light, but
present in all cell types at an equal intensity.
Longitudinal section through a leaf. Staining is present in all cell types.
Tip of silique.
Base of silique.
Longitudinal section through anther.
Cross section of a stem.

 

114

 

115

Figure 4-12. ALTP1 activity in lines containing 1 Kb, 700 bp, and 280 bp promoter
fragments. Seedlings were stained at 10 days.

Seedling containing 1 Kb fragment.

Seedling containing 700 bp fragment.
Seedling containing 280 bp fragment.

Roots of seedling containing 280 bp fragment.
Roots of seedling containing 700 bp fragment.

mane?

pl, primary leaf; sl, secondary leaf; stip, stipule

 

 

 

 

116

 

 

 

 

 

SUITOI

Thes

DISI

tiai

prc
de

0P

117

surrounding areas of damage and in the vascular tissue near wounded cells.

These results are still very preliminary, and thus, the data is not shown here.

DISCUSSION

Lipid transfer proteins were originally isolated based on their ability to
transfer lipids between membranes in vitro. It has long been assumed that the
proteins carry out a similar role in viva, carrying newly synthesized lipids to
developing membranes. A yeast phosphatidylinositol transfer protein (PI-TP) is the
only LTP to which the in vitro activity correlates to a function in viva (Bankaitis, et
al., 1990). It appears that this protein is involved in regulating
phosphatidylinositol/phosphatidylcholine ratios of yeast Golgi membranes (Cleves,
et al., 1991). However, plant nsLTPs lack sequence homology to the yeast Pl-TP
and, thus, are likely to have a different biological role. A role in intracellular lipid
transfer implies that LTPs are cytosolic and are present in all cell and tissue types
in which membranes are being synthesized. However, it has been demonstrated
that barley nsLTP is secreted into aleurone culture medium (Mundy and Rogers,
1986) and a carrot nsLTP is secreted by carrot embryogenic cell cultures (Sterk,
et al., 1991). More recently, an Arabidopsis LTP was shown to be localized to cell
walls and was found mainly in epidermal cells in the tissues examined (chapter 3
of this dissertation; Thoma, et al., 1993). A maize LTP was shown to be expressed
mainly in epidermal cells and vascular strands of maize coleoptiles (Sossountzov,

et al., 1991), and a tobacco LTP was highly expressed in leaf epidermal cells and

118
in the shoot apical meristem (Fleming, et al., 1992). In Arabidopsis (chapter 3 of

this dissertation; Thoma, et al., 1993), spinach (chapter 2 of this dissertation;
Bernhard, et al., 1991), maize (Sossountzov, et al., 1991), and tobacco (Fleming,
et al., 1992), LTP is found to be localized or expressed in the aerial portions of the
plant, with little or no localization or expression in root tissue. Clearly, the cell-type
and tissuespecific localization of plant nsLTPs is not consistent with a direct role
of these proteins in intracellular lipid transfer.

A genomic clone corresponding to an Arabidopsis nsLTP has been isolated.
Analysis of the complexity of the Arabidopsis genome reveals the presence of at
least two genes encoding for LTP, and this gene has been designated ALTP1.
The presence of multiple genes is not entirely surprising, as LTP isoforms have
been found in castor bean (Tsuboi, etal, 1991), barley (Mundy and Rogers, 1986;
R. Kalla, personal communication), tobacco (Koltunow, et al., 1 990; Fleming, et al..
1992), and wheat (Dieryck, et al., 1992), and Southern analysis indicates that there
may be several LTP genes in maize (T chang, et al., 1989) and tomato (T arres-
Schumann, et al., 1992). The deduced amino acid sequence of the ALTP1 clone
showed an exact match to that of a previously isolated cDNA clone (chapter 5 of
thisdissertation), except for one Val to Gly conversion at position 528. This could
indicate that we have cloned a gene other than the one corresponding to the
previously isolated cDNA. It is also possible that the difference observed is due
to an error in reading the DNA sequence, as the change is due to the conversion

of one base pair (T-rG) in the middle of the codon.

119

To better understand plant nsLTPs and to provide insight into their possible
function, transgenic plants containing ALTP1 promoter-B-glucuronidase fusions
were produced. The regulation of a reporter gene by a heterologous promoter in
transgenic plants has been shown to be an accurate representation of the intrinsic
regulatory properties of the introduced promoter (Benfey and Chua, 1989; Bevan,
et al., 1989). As anticipated in these experiments, there was variation in the relative
levels of expression in different transformed lines. Line LTP-GUS 1A had high
relative levels of GUS expression, LTP-GUS 1D had a low level of expression, and
LTP-GUS 1B and C had moderate expression levels. Transgenic lines containing
shorter promoter fragments also exhibited moderate expression levels. These
differences were probably due to positional effect from insertion of the gene
construct in various sites of the target genome. The problem was overcome by
looking at the expression pattern of several independently transformed lines. The
overall pattern of expression did not differ between lines containing the same
transgene.

Analysis of the temporal and spatial activity of the ALTP1 promoter
demonstrated that it was active in very specific cell and tissue types. Certain
aspects of the pattern observed were similar to the pattern of cells and tissues in
which phenylpropanoid biosynthesis occurs and in which the products of the
phenylpropanoid biosynthetic pathway reside. Phenylpropanoids are a class of
plant natural products derived from phenylalanine. These compounds play

important roles in plant development and in protection against environmental

120

stress. For instance, flavonoids are pigments and UV protectants in epidermal
cells, lignin is the ma’or structural component in xylem cell walls, and suberin is a
lipophilic substance commonly found in the casparian strip of the endodermis
(Esau, 1977). The induction of lignin in wheat (Moerschbacher, et al., 1990).
suberin deposition in potato (Roberts and Kolattukudy, 1989) and the
accumulation of phenylpropanoid derived phytoalexins help protect a plant against
mechanical damage and microbial attack (Dixon and Lamb, 1990).

The ALTP1 promoter was active in early seedling development. The pattern
of staining observed in the cotyledon and primary leaves follows the pattern of
xylem differentiation. Xylem development commonly occurs at the basal portion
of the leaf, then progresses in an acropetal manner. In leaves of adult plantS.
there was also staining associated with vascular tissue. More detailed analysis of
mature leaves needs to be carried out to determine if vascular expression
proceeds in a manner consistent with xylem differentiation.

In the developing seedling, staining was also prominent in emerging lateral
roots. Lateral roots arise from the pericycle and their emergence causes damage
to cortical cells. This results in an opening into the interior of the parent root, and
this "wound" may induce lignin and/or suberin deposition. Staining was also
observed in the stipules of young seedlings at least until two weeks of age. It is
difficult to assign a role for LTP in stipules, as this organ has no known function.
Staining was observed in the shoot apical meristem of developing seedlings. A

tobacco LTP has been shown to be highly expressed in the shoot meristem, and

121

in situ hybridization experiments show that the expression is located mainly in the
L1 layer. To see if the ALTP1 was active in a specific cell type, sections were
made through the shoot meristem of stained 10 d and 14 d seedlings. Although
there was weak staining observed in the shoot meristem in whole mounts of 10 d
seedling, expression of the ALTP1 promoter was not detected in sectioned tissue.
It is possible that the activity is at a low level which cannot be detected in thin
sections. Sections through younger tissue, where the meristematic acivity is more
prominent, need to be analyzed.

In the mature plant, the ALTP1 promoter was active in cells containing a
thick cuticle layer; leaf and stem epidermal cells, guard cells, and hydathodes.
Cutin is a polymer found associated with the cell wall of cells on the outer portion
of the aerial parts of plants, and phenolic acids such as p-coumarate and ferulic
acid are structural components (Riley and Kolattukudy, 1975). Staining observed
in stems was weak, but was more prominent near the top of the stem. This type
of developmental expression in which LTP is expressed at higher levels in the
younger part of the plant was also observed in tobacco (Fleming et al., 1992).
This differential expression was quite apparent in bulk stained stem tissue, but
more difficult to discern in sectioned tissue, and further studies to quantitate the
GUS activity by a fluorometric assay (Jefferson, et al., 1987) need to be carried
out.

The ALTP1 promoter was active during floral development. At the stage of

development where the sepals enclose the flower bud (stage 6, Smyth, et al.,

122
1990), no staining was observed. At stage 11-12, when the stigmatic papillae have

appeared and the petals are level with long stamens, staining is observed on the
stigma surface. By stage 15, when the stigma extends above the long anthers,
staining is observed in the stigma surface, the floral nectaries, and in developing
pollen grains. Floral nectaries are covered with a cuticular substance and stigma
cells can produce a secretion containing lipids and phenolic compounds such as
anthocyanins, flavonoids, and hydroxycinnamic acids (Martin and Brewbaker.
1971). Pollen grains are covered with a polymerized lipid material, spor0pollenin.
The chemical nature of sporopollenin is largely unknown, but there is evidence that
it may contain phenolic materials (Kolattukudy, 1980). Pigment synthesis also
occurs in developing pollen grains. Staining was observed at the base of the
ripening silique in the abscission zone of sepals and petals. This pattern is
consistent with that seen of an Arabidopsis phenylalanine ammonia-lyase (PAL)
promoter (Ohl, et al., 1990). Week staining was also observed in the silique walls.
Analysis of the promoter and 5’ untranslated region of the ALTP1 gene
revealed the presence of several sequence elements commonly found in
promoters of biosynthetic genes of the phenylpropanoid pathway (Cramer, et al.,
1989; Lois, et al., 1989; Ohl, et al., 1990). Two elements (boxes 1 and 2) are
homologous to sequences in bean, parsley, and Arabidopsis PAL promoters. In
parsley, these sequence elements have been shown to display elicitor-inducible
and light-inducible footprints in viva. A 9 bp AC rich element which is found in the

5’ untranslated regions of an Arabidopsis PAL gene and a bean chalcone synthase

123

(CHS) gene was also observed in this region of the ALTP1 gene. These sequence
data, along with the GUS expression pattern are consistent with a role for the
Arabidopsis nsLTP in phenylpropanoid metabolism or regulation of expression by
the same or similar mechanisms.

The role of possible regulatory sequences can be tested by examining the
staining pattern of plants containing truncated versions of the ALTP1 promoter.
Two of these shortened promoters, one containing a 700 bp fragment and the
other a 280 bp fragment of the promoter, have been transformed into plants. There
has not been extensive analysis of these plants, although preliminary data
suggests that expression patterns are similar to that observed with the 1 kb
promoter in 4 d and 10 d seedlings. To test the putative regulatory sequences
described in this paper, however, shorter promoter fragments and site specific
mutagenesis in specific regulatory sites must be carried out. Lignification and
suberization are induced in a plant upon wounding or pathogen infection.
Preliminary data suggest that ALTP1 may indeed be activated upon wounding.
Analysis of transgenic plants containing the ALTP1-GUS fusion show that there is
no GUS induction 2 hours after mechanical damage to the plant. However, 48 h
after infection with the fungal pathogen Erysiphe cruciferae, there is a significant
increase in staining around areas of cell damage in the infected leaves and in the
vascular tissue near the area of cell damage. The method of inoculation involves
rubbing an infected leaf over the test leaf, so it is not possible to tell if the cell

damage is due to fungal invasion, or to mechanical damage due to the inoculation

124

technique. It will be important to repeat these tests and to more fully assess the
effect of wounding and pathogen infection on the induction of ALTP1, using both
qualitative (histochemical) and quantitative (fluorometric) techniques.

A possible limitation to the GUS expression studies described here arises
from the possible role of nsLTPs in plant stress responses. If ALTP1 is induced
by stress,.the act of removing the tissue from the plant and adding it to the
staining solution may be stressful enough to trigger induction of the gene. Thus
I may be looking the stress induced pattern of expression. It may be important to
pre-fix the tissue in a fixative such as paraformaldehyde, to look at normal, non-
stress induced pattern of expression.

Activation of ALTP1 by biotic and abiotic stress requires further study. The
expression of a tomato nsLTP was shown to be induced by salt stress (T orres-
Schumann, etal, 1992), and proteins with regions of homology to LTPs have been
shown to be induced by drought stress. Analysis of transgenic Arabidopsis plants
which contain a reduced level of LTP due to the presence of an Arabidopsis LTP
cDNA in reverse orientation behind the CaMV 35S promoter, show no alteration
in their wax or cutin contents under normal growth conditions (chapter 5 of this
dissertation). Cutin composition from plants grown under drought conditions,
where the synthesis of cuticular components is a limiting factor in plant survival,
needs to be analyzed. Also it will be useful to examine the Iignin composition of
wounded and nonwounded antisense plants.

In conclusion, the ALTP1 promoter specifies a complex and specific

125

expression pattern in transgenic plants. The patterns observed, along with the
presence of promoter sequences homologous to putative regulatory elements of
phenylpropanoid biosynthetic genes, suggests that plant nsLTPs may be involved
in phenylpropanoid metabolism or the expression of LTP genes is regulated by the

same or similar mechanisms as phenylpropanoid genes.

1 26
REFERENCES

Arondel, V. and Kader, J.-C. 1990. Lipid transfer in plants. Experientia 46:579-585.

Arondel, V., Vergnole, C., Tchang, F. and Kader, J.-C. 1990. Bifunctional lipid-
transferzfatty acid-binding proteins in plants. Mol. Cell. Biochem. 98:49-56.

Bankaltle, V.A.. Allken, J.R., Cloves, AE. and Dowhm, W. 1990. An essential role
for a phospholipid transfer protein in yeast Golgi function. Nature 347:561-562.

Benley, RN. and Chua, N.-H. 1989. Regulated genes in transgenic plants. Science
244: 1 741 81 .

Bernhard, W.R. and Somervllle, C.R. 1989. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from
amino acid sequence homology. Arch. Biochem. Biophys. 269:695-697.

Bernhard, W.R., Thoma, S., Botella, J. and Somerville, C.R. 1991. Isolation of a
cDNA alone for spinach lipid transfer protein and evidence that the protein is
synthesized by the secretory pathway. Plant Physiol. 95:164-170.

Bevan, M., Shufflebottom, D., Edwards, K., Jefferson, R. and Schuch, W. 1989.
Tissue- and cell-specific activity of a phenylalanine ammonia-Iyase promoter in
transgenic plants. EMBO J. 8:1988-1906.

Boullon, P., Drbchel. C., Vergnofle, C., Duranton, H. and Kader, J.-C. 1987. The
primary structure of spinach-leaf phospholipid-transfer protein. Eur. J. Biochem.
166:387-391.

Breu, V., Guerbette, F., Kader, J.-C., Gamlnl Kannangara, C., Svensson, B. and
von Wetteteln -Knowlee, P. 1989. A 10 kD barley basic protein transfers
phosphatidylcholine from liposomes to mitochondria. Carlsberg Res. Commun.
54:81-84.

Cloves, A.E., McGee, T.P., Whltters, EA, Champlon, KM.. Altken, J.H., Dowhan,
W., Goebel, M. and Bankaitis, VA. 1991. Mutations in the CDP-choline pathway for
phospholipid biosynthesis bypass the requirement for an essential phospholipid
transfer protein. Cell 64:789-800.

Cramer, C.L, Edwards, K., Dron, M., Uang, X., Dlldlne, S.L, Bolwell, G.P., Dlxon,
RA, Lamb, OJ. and Schuch, W. 1989. Phenylalanine ammonia-lyase gene
organization and structure. Plant Mol. Biol. 12:367-383.

127

De Block, M. and Debrouwer, D. 1992. In situ enzyme histochemistry on plastic-
embedded plant material. The development of an artefact-free B-glucuronidase
assay. Plant J. 2:261 -266.

Dleryok. W., Gauller, M.-F., Lulllen, V. and Joudrler, P. 1992. Nucleotide sequence
of a cDNA encoding a lipid transfer protein from wheat (Triticum durum Desf.)
Plant Mol. Biol. 19:707-709.

Dlxon, RA. and Lamb, C.J. 1990. Molecular communication in interaction between
plants and microbial pathogens. Annu. Rev. Plant Physiol. Plant Mol. Biol. 41 :339-
367.

Douady, D., Grosbole. M., Guerbette, F. and Kader, J.-C. 1982. Purification of a
basic phospholipid transfer protein from maize seedlings. Biochim. Biophys. Acta
710:143-153.

Esau, K. 1977. Anatomy of seed plants, second edition. John Wiley and Sons, New
York.

Felnberg, AP. and Vogelsteln, B. 1983. A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-
13.

Flemlng,.,A.J Mandel,.,.,THofmanS Sterk.P., deVrIee,..SCandKuhlemeler,C.
1992. Expression pattern of a tobacco lipid transfer protein gene within the shoot
apex. Plant J. in press.

Helmkamp, GM. 1990. Transport and metabolism of phosphatidylinositol in
eukaryotic cells. In: Subcellular biochemistry, H.J. Hilderson, ed., Plenum
Publishing Corp., pp.129-174.

Helmkamp, G.M., Harvey. M., ertz. KWA and van Deenen, LLM. 1974.
Phospholipid exchange between membranes. Purification of bovine brain proteins
that preferentially catalyze the transfer of phosphatidylinositol. J. Biol. Chem.

249:6382—6389.

Jefferson, RA, Kavanagh, TA and Bevan, MW. 1987. GUS fusions: B-
glucuronidase as a sensitive and versatile gene fusion marker in higher plants.

EMBO J. 6:3901-3907.

Joehl, C.P. 1987. An inspection of the domain between putative TATA box and
translation start site in 79 plant genes. Nuc. Acid. Res. 15:6643-6653.

128

Kader, J.-C., Julienne, M. and Vergnolle, C. 1984. Purification a nd characterization
of a spinach-leaf protein capable of transferring phospholipids from liposomes to
mitochondria or chloroplasts. Eur. J. Biochem. 139:411-416.

Kemp, H.H., erlz, K.W.A. and Van Deenen, LLM. 1973. Some properties of
phosphatidyl exchange protein purified from beef liver. Biochim. Biophys. Acta
318:313-325.

Kolattukudy, PE. 1980. Cutin, suberin, and waxes. In: The Biochemistry of Plants,
Vol. 4, Lipids: Structure and Function. P.K. Stumpf and E.E. Conn, eds. Academic
Press, NY, . pp. 571 -645.

Koltunow, AM., Truettner, J., Cox, KH., Wallroth, M. and Goldberg, RB. 1990.
Different temporal and spatial gene expression patterns occur during anther
development. Plant Cell 2:1201-1224.

Lola, R., Dietrich, A., Hahlbrook, K. and Schulz, W. 1989. A phenylalanine
ammonia-lyase gene from parsley; structure regulation, and identification of elicitor
and light-responsive cis-acting elements. EMBO J. 8:1641-1648.

Medrld, S. 1991. The barley lipid transfer protein is targeted into the lumen of the
endoplasmic reticulum. Plant Physiol. Biochem. 29:695-703.

Manm, T., Frltech, E.F. and Sambrook, J. 1982. Molecular Cloning. A Laboratory
Manual. Cold Spring Harbor, New York.

Martln, F.W. and Brewbaker, J.L 1971. The nature of stigmatic exudate and its role
in pollen germination. In: Pollen development and physiology. J. Heslop-Harrison,
ed. Butterworth, London, pp.262-272.

Moerschbacher, B.M., Noll, U., Gorrlchon, L and Relsener, H.-J. 1990. Specific
inhibition of Iignification breaks hypersensitive resistance of wheat to stem rust.
Plant Physiol. 93:465-470.

Mundy, J. and Rogers, J.C. 1986. Selective expression of a probable
amylase/protease inhibitor in barley aleurone cells: comparison to the barley
amylase/subtilisin inhibitor. Planta 196:51-63.

Munro, S. and Pelham, H.H.B. 1987. A C-terminal signal prevents secretion of
luminal ER proteins. Cell 48:899-907.

Muraehlge, J. and Skoog, F. 1962. A revised medium for rapid growth and bio
assay with tobacco tissue cultures. Physiol. Plant. 115: 473-497.

129

Ohl, S., l-Iedrlck, 8., Chery, J. and Lamb, C.J. 1990. Functional properties of a
phenylalanine ammonia-lyase promoter from Arabidopsis. Plant Cell 2:837-848.

Paltauf, F. and Daum, G. 1992. Phospholipid transfer proteins from yeast. Methods
Enzymology 209:514—522.

Rlcltere, J., Spener, F. and Kader, J.-C. 1985. A phospholipid transfer protein that
binds long-chain fatty acids. FEBS 180:29-32.

Eloy, RG. and Kolattukudy, PE. 1975. Evidence for covalently attached p-
coumaric acid and ferulic aced in cutins and suberins. Plant Physiol. 56:650-654.

Roberta, E. and Kolatmkudy, PE. 1989. Molecular cloning, nucleotide sequence,
and abscisic acid induction of a suberization-associated highly anionic peroxidase.
Mol. Gen. Genet. 217:223-229.

Sanger, F., NIcklen, S. md Couleon, AR. 1977. DNA-sequencing with chain-
termination inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467.

Smyth, D.R., Bowman, J.L md Meyerowllz, EM. 1990. Early flower development
in Arabidopsis. Plant Cell 2:755-767.

Sossountzov, L, Rulz-Avfla, L, Vlgnola, F., Jolllot. A., Arondel, V., Tchang, F.,
Grosbois, M., Guerbette, F., Mignlac, E., Delseny, M., Pulgdomenech, P. and
Kader, J.-C. 1991. Spatial and temporal expression of a maize lipid transfer protein
gene. Plant Cell 3:923-933.

Sterk, P., Boolf, H.. Schellekens, GA, Van Kammen, A. and De Vrles, S.C. 1991.
Cell-specific expression of the carrot EP2 lipid transfer protein. Plant Cell 3:907-
921.

Szolderlte, G., Hermetter, A., Paltauf, F. and Daum, G. 1989. Membrane properties
modulate the activity of a phosphatidylinositol transfer protein from the yeast,
Saccharamyces cerevisiae. Biochim. Biophys. Acta 986:301-309.

Tal, S.-P. and Kaplan, S. 1985. Intracellular localization of phospholipid transfer
activity in Rhadapseudamanas sphaeraides and a possible role in membrane
biogenesis. J. Bacteriolon 164:1 81 -186.

Taklshlma, K, Watanbe, 5.. Yamada, M. and Mamlya, G. 1986. The amino-acid
sequence of nonSpecific lipid transfer protein from germinated castor bean.
Biochim. Biophys. Acta 870:248-255.

130

Tan, H.. Okazakl, K, Kubota, I., Kamlryo, T. and Utlyama, H. 1990. A novel
peroxisomal nonspecific lipid-transfer protein from Candida trapicalis. Gene
structure, purification and possible role in B-oxidation. Eur. J. Biochem. 190:107-
112.

Tchang, F., This, P., Stlefel, V., Arondel, V., March, M.-D., Pages, M.,
Pulgdonmenoh, P., Grellet, F., Delaney, M., Boufllon, P., Huet, J.-C., Guerbette, F.,
BeauvaIe-Cante, F., Duranton, H. and Kader, J.-C. 1988. Phospholipid transfer
protein: full length cDNA and amino acid sequence in maize. J. Biol. Chem.
263:16849-16855.

Thoma, S.L, Kaneko, Y. and Somervflle, C. 1993. An Arabidopsis lipid transfer
protein is a cell wall protein. Plant Journal, in press.

Torres-Schumann, S., Godoy. JA and PIntor-Toro, JA 1992. A probable lipid
transfer protein gene is induced by NaCl in stems of tomato plants. Plant Mol. Biol.

1 82749-757.

Tsuboi, S., Suga, T., Taklshlma. K, Mamlya. G., Matsul, K, Ozekl, Y. and Yamada,
M. 1991. Organ-specific occurrence and expression of the isoforms of nonspecific
lipid transfer protein in castor bean seedlings, and molecular cloning of a full-
length cDNA for a cotyledon-specific isoform. J. Biochem. 110:823-831.

Valvekens, D., Van Montagu, M. and Van ueebettens, M. 1988. Agrabacterium
tumefaciens-mediated transformation of Arabidopsis thaliana root explants by

using kanamycin selection. Proc. Natl. Acad. Sci. USA 85:5536-5540.

erlz, KWA 1982. Phospholipid transfer proteins. In: Lipid-protein interactions,
P.C. Jost and OH. Griffith, eds., Elsevier, Amsterdam. PP. 151-223.

ertz, KWA and leveremlt, 0.8. 1968. Exchange of phospholipids between liver
mitochondria and microsomes in vitro. J. Biol. Chem. 243:3596-3602.

Yamada, M., Tsuboi, S., Osafune, T., Suga, T. and Takishlma, K 1990.
Multifunctional properties of non-specific lipid transfer proteins from higher plant.
In: Plant lipid biochemistry, structure, function, and utilization. P.J. Quinn and J.L
Harwood, eds. Portland Press Ltd., London, pp. 278-280.

CHAPTER 5

ANALYSIS OF TRANSGENIC ARABIDOPSIS CONTAINING A REDUCED LEVEL
OF A NON-SPECIFIC UPID TRANSFER PROTEIN

131

1 32
ABSTRACT

A class of proteins, called lipid transfer proteins (LTPs), has been shown to
mediate transfer of phospholipids between membranes in vitro; however, there is
no evidence that these proteins catalyze lipid transfer in viva. I have recently
demonstrated by immunocytochemical labeling at the ultrastructural level that an
Arabidopsis LTP has an extracellular location, with its highest concentration in the
epidermal cells of aerial plant organs. The extracellular location and the cell-type
and tissue-type specific location is inconsistent with the proposed role of the
protein in intracellular lipid transfer. To determine the role of the protein, several
transgenic plants in which the level of LTP has been reduced by the expression
of an LTP antisense construct were generated. Western analysis indicated that
these plants contain less than 10% of the LTP of wild type plants. The transgenic
plants exhibited no visible phenotype to indicate a possible function of LTP within
the plant. The location of this protein and its ability to bind hydrophobic
molecules, leads to a speculation that the protein may be involved in cuticle
formation. The wax and cutin composition of the antisense plants has been
analyzed and there is no difference in comparison to wild type plants under normal

growth conditions.

1 33
INTRODUCTION

In eukaryotic cells, lipid synthesis occurs mainly on the lumenal face of the
endoplasmic reticulum (ER), and in all plant species, some lipid synthesis also
occurs in the chloroplast. Although lipid biosynthesis is restricted to a few
organelles within a cell, lipids are the ma'or component of all cellular membrane
systems of a cell, with different organelles containing unique and specific lipid
compositions. Thus, several membranes and organelles need to import their lipid
constituents. The mechanism(s) by which lipids move from their site of synthesis
to other cellular organelles is unknown.

In an attempt to elucidate the mechanism by which lipid movement occurs,
scientists began to search for proteins which were capable of transferring
phospholipids between membranes in vitro. Such proteins, called lipid transfer
proteins (LTPs), have subsequently been isolated from animal and plant tissue,
yeast, and bacteria. LTPs which transfer specific phospholipids, such as a
phosphatidylcholine transfer protein (PC-TP) and a phosphatidylinositol transfer
protein (Pl-TP), have been isolated from animal tissue and yeast (Wirtz, 1982;
Helmkamp, 1990). Non-specific LTPs (nsLTPs), which are capable of transferring
several classes of phospholipids, have been isolated from bacteria, plant and
animal tissue (Wirtz, 1982; Arondel and Kader, 1990). The plant nsLTPs have been
isolated from several species, including spinach (Kader, et al., 1984; Bouillon, et
al., 1987), maize (Douady, et al., 1982), barley (Mundy and Rogers, 1986).

sunflower (Arondel, et al., 1990), wheat (Dieryck, et al., 1992) and castor

134
(T akishima, et al., 1986). Although the plant LTPs share no obvious sequence

similarity to the nsLTPs from other organisms, they have considerable similarity
among themselves and they share several common features; they are soluble,
basic proteins with a molecular mass around 9 kDa.

The ability of LTPs to transfer phospholipids in vitro has led to the proposal
that LTPs act to shuttle lipids between organelles in viva. However, there has
been no direct evidence which supports this proposal. The only LTP to which the
in vitro activity correlates to a function in vivo is a yeast PI-TP which appears to be
involved in regulating the phosphatidylinositol/phosphatidylcholine ratios in yeast
Golgi membranes (Bankaitis, et al., 1990). It has been proposed that this protein
acts not by physically transferring lipids, but by sampling the Pl/PC ratio and
subsequently regulating PC synthesis (Cleves, et al., 1991). However, the plant
nsLTPs lack sequence homology with the yeast Pl-TP, and are thus likely to have
a different biological role.

Based on the proposed function of intracellular lipid transfer, it has been
assumed that LTPs were cytoplasmic (Arondel and Kader, 1990). However, there
is much evidence that contradicts a cytosolic location. It has been reported that
the maize LTP has an amino terminal extension and is synthesized on membrane
bound polysomes (T chang, et al., 1988, Vergnolle, et al., 1988) and that spinach
LTP contains a signal peptide and is cotranslationally inserted into microsomal
membranes (Bernhard, et al., 1991; chapter 2 of this thesis). As these proteins

lack the carboxy terminal ER retention signal, KDEL (Munro and Pelham, 1987),

135

they would be expected to. be secreted or targeted to a specific organelle. These
expectations were reinforced by reports that a secreted barley protein (Mundy and
Rogers, 1986) was homologous to nsLTPs (Bernhard and Somerville, 1989) and
was capable of in vitro lipid transfer (Breu, et al., 1989), and that a carrot nsLTP
was secreted by carrot embryogenic cell cultures (Sterk er al., 1991).
Immunocytochemical studies have shown that a maize LTP is localized to the
periphery of epidermal cells, and it has been suggested that the staining is
associated with the plasma membrane (Sossountzov, et al., 1991). An Arabidopsis
nsLTP has been localized to the cell wall and cutin layer in the aerial portions of
the plant, and is observed mainly in epidermal cells in leaves, petioles, stems, and
in the cells of the stigma surface (Thoma, et al., 1993; chapter 3 of this
dissertation).

The extracellular location of plant nsLTPs indicates that these proteins are
not directly involved in intracellular lipid transfer. This opens the possible in viva
role of these proteins to speculation. The preferential accumulation of the
Arabidopsis LTP to the epidermal cell walls is consistent with the prediction that
the protein may be involved in cuticle formation (Sterk, et al., 1991; Thoma, et al.,
1993; chapter 3 of this dissertation). The cuticle is a continuous layer of
predominantly lipophilic material found on the outermost surface of the aerial parts
of plants (Holloway, 1982). The cuticle is composed of cutin, an insoluble polymer
composed mainly of C1. and C1. hydroxylated and epoxygenated fatty acids, and

wax, which contains several classes of long chain hydrocarbons, including long

136
chain fatty acids (Kolattukudy, 1982). Wax and cutin monomers are synthesized

in epidermal cells and are delivered to the outside of the cell where it is thought
that esterification takes place (Croteau and Kolatukuddy, 1974). It is unknown how
the hydrophobic monomers pass through the aqueous environment of the cell
wall. Since LTPs have been shown to bind fatty acids (Rickers, at al., 1985), it has
been proposed that LTPs may play a role in the deposition of these aliphatic
chains to the outer cell surface.

Another proposed role for nsLTPs is involved with general plant defense.
A radish protein, originally isolated based on its in vitro antifungal activity, shares
sequence homology with plant nsLTPs (T erras, et al., 1992). Also, LTPs isolated
from barley and maize show strong antifungal activity in vitro assays (Garcia-
Olmeda, personal communication).

More recently, I have shown, using transgenic plants that contain an
Arabidopsis LTP promoterzfl-glucuronidase (GUS) fusion, that the promoter is
active in cells and tissues where phenylpropanoid metabolism is operative (chapter
4 of this dissertation). The GUS expression pattern, along with the presence of
sequence elements homologous to putative regulatory elements of
phenylpropanoid biosynthetic genes, has led to the proposal that LTP may play
a role in phenylpropanoid metabolism, or that expression of LTP genes is
regulated by similar mechanisms (chapter 4 of this dissertation).

To determine the role of nsLTPs in viva, transgenic plants which contain an

Arabidopsis nsLTP cDNA in reverse orientation behind the cauliflower mosaic virus

137

358 promoter were generated. Western analysis shows that these plants contain
a greatly reduced amount of protein when compared to control plants. Apart from
a delay in time of flowering, the plants containing the antisense construct, the
plants had no obvious phenotypic differences in comparison to control plants. To
test for the role of the protein in cuticle formation we analyzed the wax and cutin
composition of control and antisense plants using gas chromatography and
electron microscopy. To examine a possible role for LTP as an antifungal
compound, I also tested the ability of the antisense plants to withstand pathogen
attack. Experiments are also proposed to test the role of the protein in

phenylpropanoid metabolism. The results of these analyses are presented here.

EXPERIMENTAL PROCEDURES
Plant growth conditions - Unless othenrvise indicated, all plants were grown
at 22°C under continuous fluorescent illumination (100-150 pmol m2 s") on a

potting mixture of fine sphagnum:perlitezvermiculite (1 :1 :1).

Isolation of an Arabidopsis cDNA clone - A AGTf O Arabidopsis leaf cD NA
library was plated on Escherichia coli LE392 cells and nitrocellulose plaque lifts
were screened with a spinach LTP cDNA which was labeled with [a-nP] dCTP by
random priming (Feinberg and Vogelstein, 1983). Filters were prehybridized for
4 h at 42°C in a solution of 5X Denhardt’s, 5X SSC, 0.1% SDS, and 0.5 ug/ul

sonicated herring DNA. Hybridization was carried out for 16 h at 42°C in the same

138

solution containing the radiolabeled probe. The filters were washed twice at room
temperature, first in 4X SSC, followed by 2X SSC, and once at 42°C in 2X $80.
All washes were carried out for 15 min. The resulting cDNA was subcloned into
pBluescript and sequenced. All nucleic acid manipulations were carried out as

described in Maniatis, et al. (1982).

Production of antisense constructs - To produce plants in which the amount
of LTP has been reduced, three chimeric gene constructs were prepared. The
constructs were made by inserting fragments of an Arabidopsis LTP cDNA into the
binary vector, pBl121, in reverse orientation between the CaMV 35$ promoter and
the nos terminator. The B-glucuronidase gene was cut out of pBl121 with the
restriction enzymes Smal and Sacl. To prepare the constructs pANA1 and pANA2,
the Arabidopsis LTP cDNA was digested with EcoRI, which made a single out near
the center of the gene, and these two fragments were subcloned into the EcoRI
site of pBluescript. The LTP cDNA fragments were removed from pBluescript by
restriction digests with Sacl and Smal and the resulting fragments were ligated into
the corresponding sites of pBl121. To construct pANA3, which contained the full
length cDNA, the cDNA was amplified from the phage by PCR using primers
containing BamHl restriction sites. The PCR product was digested with BamHI
and subcloned into the corresponding site of pBluescript. The resulting plasmid
was then cut with Sacl and EcoRV and subsequently subcloned into the Sacl and

Smal sites of pBl121.

139
Production of transgenic plants - Each antisense construct and the vector,

pBl121, were transformed into Agrabacterium tumefaciens C58/pGV3850 by
electroporation. These Agrabacterium were used to transform Arabidopsis thaliana
(vars. RLD, NO-O, STD, and C24) by the root transformation method as described
by Valvekens, et al., 1988. Seed were collected from initial transformants, sterilized
and plated on Murashige-Skoog (MS) medium (Murashige and Skoog, 1962)
containing kanamycin (50 ug/ml). T2 plants which were resistant to kanamycin
were transplanted to pots and allowed to set seed. Segregation of kanamycin
resistance in the transgenic plants was followed in this manner until all lines were
homozygous for the transgene. Homozygous plants were used for all subsequent

analysis.

Screening transgenic plants - Protein extracts were made from leaves of
individual transformed plants by homogenizing tissue in 20 mM glycine, pH 8.4, 5
mM MgCl2, 25 mM EDTA. Extracts were centrifuged twice at 10,000 X g for 10
min. Protein concentrations were determined by the BCA protein assay (Pierce
Biochem. Rockford, Ill) and equal amounts of protein from transgenic and wild
type plants were loaded onto 10-20% gradient SDS-polyacrylamide gels. The
resolved proteins were transferred to nitrocellulose and subjected to Western
analysis essentially as described by Towbin, et al. (1979) using anti- Arabidopsis
LTP antibodies (Thoma, et al., 1993; chapter 3 of this thesis). Goat anti-rabbit lgG-

alkaline phosphatase ooriugate was used as the second antibody. Alkaline

14o

phosphatase activity was shown using nitroblue tetrazolium and 5—bromo-4-chloro-
3-indolyl phosphate as substrates. Over 50 independently transformed lines were
screened in this manner. Plants which had low levels of protein in comparison to
wild type plants and control plants which were transformed with pBl121 containing
no antisense cDNA were rescreened by Western analysis. Four lines, LTP4, LTP5,
LTP12, and LTP18 had the greatest reduction in LTP, and were used for
subsequent analyses. T17, a transgenic line containing pBl121 without an
antisense cDNA, had similar levels of LTP as wild type and was used as a control

in subsequent analyses.

Growth curves - Seed from transgenic and wild type plants were sown on
the potting mixture described above. For growth measurements, the fresh weight
of the entire aerial portion of plants (n=10) was determined at 2 day intervals, from
day 4 to day 26. To measure time of flowering, the percentage of individual plants
that were flowering (n210) was measured each day from the time control plants
started bolting (day 14) until all plants were bolting. Percent germination was

determined by planting a known number of seed per pot and counting the number

of resulting plants.

Gas chromatographic analysis of wax and cutin composition - Epicuticular
wax was isolated by washing stems and leaves of wild type and transgenic plants

twice with chloroform. The chloroform extracts were concentrated by drying under

141
N2, and separated by gas chromatography on a 15 m X 0.53 mm ID Supelco SP8-

1 column, using flame ionization detection.

To isolate cutin, the residual plant material from the above preparation was
heated (80°C) in several changes of chloroformzmethanol (1:1) for 24 hours. The
tissue was then heated under reflux (75°C) in 3N methanolic HCI for 48 hours and
then extracted with NaCI (O.1M):hexane (1 :4). The hexane phase, which contained
the methyl esters of cutin monomers, was collected, concentrated, and was

separated by gas chromatography as described for wax esters.

Scanning electron microscopy- Stems from 5 week old transgenic and wild
type plants were removed and segments from similar portions of the stems were

viewed on an EMscope cryo stage of a JEOL 350F scanning electron microscope.

Preparation of samples for transmission electron microscopy - Stems from
5 week old transgenic and wild type plants were removed and similar portions of
the stems were fixed for 1.5 h, at room temperature in 2.5% (v/v) glutaraldehyde
in 0.1 M sodium phosphate, pH 7.0. The samples were washed 3 times with
phosphate buffer and post fixed for 1 h in 1% (v/v) osmium tetroxide in 0.1 M
sodium phosphate, pH 7.0 at room temperature. The tissue was washed 3 times
with phosphate buffer, dehydrated in a graded series of ethanol (ZS-100%) and
infiltrated with Spurr’s resin (Polysciences, Warrington, PA). Infiltration was carried

out over a 48 hour period with several resin changes. Dehydration and infiltration

142

were carried out at room temperature. The resin was polymerized for 24 h at
65°C. Ultrathin sections (90—100 nm) were cut with a diamond knife and mounted
on copper grids. On average, 40 sections of each of 5 sample preparations from
each antisense line were examined. Sections were stained with a saturated uranyl
acetate solution for 30 min and Reynolds lead citrate for 5 min. Sections were

viewed on a Philips 201 electron microscope.

Infection of plants with fungal pathogens - Transgenic plants containing an
antisense construct and plants containing the empty transformation vector were
grown in pots to 2.5 weeks of age and inoculated with Erysiphe cruciferae by
rubbing an infected leaf on the leaves of the test plant. Plants were placed in a

humidified chamber for 1 hour and left at 22°C for 7 d.

RESULTS

Isolation of an Arabidopsis nsL TP cDNA clone - A cDNA corresponding to
a spinach LTP was used as a probe to isolate an LTP cDNA clone from
Arabidopsis. The sequence of the cDNA was determined (Fig 5-1) and the
deduced amino acid sequence shows 49% identity to the consensus sequence
(chapter 1 of this thesis) of other known LTPs (Fig. 5-2). The presence of a 17
amino acid amino terminal extension is consistent with the reports that LTPs from

spinach, maize, and barley contain signal peptides.

143

1 AAAACTACGTCACTGGAATTCGGTGAAGTTGGCATGCTTGCTCTTGGCCTGCATGATTGT
H L A L G L H D C

61 ‘GGCCGGTCCAATACATCGAACGCTGCGCTAAGCTGTGGCTCAGTTAACAGCAACTTGGCA
G R S N T S N A A L 8 C G S V N S N L A

121 GCGTGCATTGGCTACGTGCTCCAAGGTTGTGTCATTCCCCCAGCGTGTTGCTCCGGCGTT
A C I G Y V L Q G G V I P P A C C S G V

181 AAAAACCTCAACAGCATAGCCAAGACGACCCCAGACCGTCAGCAAGCTTGCAATTGCATT
X N L N 8 I A. R T T P D R Q Q A C N C I

241 CAAGGTGCCGCTAGAGCCTTAGGCTCTGGTCTCAACGCTGGCCGTGCAGCTGGAATTCCT
Q G A A R A L G S G L N A. G R A A G I P

301 AAGGCATGTGGAGTCAATATTTCTTACAAAATCAGCACCAGCACCAACTGCAAAACCGTG
R A C G V N I S Y R I S T S T N C K T V

361 AGTGATGAGCTAGCAACGGTGAGATGATGCTACTACCGGAAGTTTCGAATCCTTATTATA
8 D E L A T V R *

421 TAATGGATGAGATTAATATTAAATAAGATGTTCGAATGTTTGTTTTTAGAGTTTTTAATT
481 TCTTGTCTTTTTCTATTGTGGTGTTCTTGTTATATGGGTTTGTCTGTACTATGTTCGCAG
541 GCAACAACGTTATATGAAATTTCAGAGTACTTGAAGTTTAAGTTAAAAAAAAAAAACCGA
601 ATTC

Figure 5-1. Sequence analysis of the Arabidopsis cDNA clone.

Nucleotide sequence of Arabidopsis cDNA and its deduced amino acid sequence.
The cDNA sequence was determined by W. Bernhard and J. Botella.

Arabidopsis

Consensus

Arabidopsis

Consensus

Arabidops 1s

CO!!! 01180..

Arabidopsis

Consensus

144

NLALGLRDCGRSITBIA

HAR'QVL**AAA*LV*LVL*AAP*A3A

ALSCGSVNSNLAACIGYVLQGGVIPPA-CCBGVKN

ooooooooooooooooo
....................

AIIE'c'QVtsivizxpcnaYL‘ecé'é'-*Pe**é'c'-cvxt

LNSIARTTPDRQQACNCIQGAARALGBGLNAGRAA

............
..............................

..................

................................

............
------------------
...................
............

u»rprxccvurpuurspsfi't‘éisni‘in

Figure 5-2. Comparison of the Arabidopsis LTP sequence to LTP consensus

sequence.

Deduced amino acid sequence of an Arabidopsis LTP determined from CDNA
sequence and comparison to the consensus sequence of other known LTPs (see

chapter 1 of this thesis).

Amino acids which are identical to the consensus

sequence are connected with (=55) and conserved substitutions are connected with

(z).

145

Production of antisense plants - To produce plants in which the amount of
LTP has been reduced, three chimeric gene constructs were prepared. As the
mechanism of antisense suppression of endogenous genes is still not well
understood, and there are varying reports as to the effectiveness of 3’ end vs. 5’
end vs. full length clone, we made constructs containing various portions of the
cDNA as well as the full length clone. The constructs were made by inserting the
full length Arabidopsis cDNA or portions of the cDNA into the binary plant
transformation vector, pBl121, in reverse orientation behind the CaMV 35$
promoter and the nos terminator (Fig. 5-3).

Antisense constructs and pBl121 were each transformed into Agrabacterium
tumefaciens strain 058 by electroporation. These Agrabacterium were used to
transform Arabidopsis plants by the root transformation method described by
Valvekens, et al. (1988). In an attempt to increase transformation efficiency.
several different ecotypes (RLD, NO-O, STD, and C24) of Arabidopsis were used
for the transformation. Plants which were kanamycin resistant were screened for
reduction in the level of LTP by Western analysis using an anti-Arabidopsis LTP
antibody (chapter 3 of this dissertation). Over 50 independently transformed lines
were screened by this method. Table 5-1 lists many of the transgenic lines and
the relative reduction in LTP as determined by Western analysis. The plants with
the greatest reduction appear, by qualitative measurement, to contain about 5—1 0%
of the level of LTP as wild type plants and of plants transformed with pBl121 (Fig.

5—4). These plants, designated LTP4, LTP5, LTP12, and LTP18 were selected to

146

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. pB1121

1:25 NPTII (Kan R) Eigmgs B-Glucmonidasc (GUS) 1:28
B. pANA1

:28 m1 (Kan R) 1385311335 Algfil'fafiDNA 1:38

C. pANA2

,2, mm ’32? 3.51.558 mm

D. pANA3

 

 

 

 

NOS NOS CaMV 35S ALTP cDNA NOS
pro NPTII (Kan R) tcr Ipromoter X full length tcr

 

 

 

 

 

Figure 5-3. Structure of constructs used to produce transgenic plants.

Three antisense constructs were prepared; pANA1 (B), containing the 5’ portion
of the Arabidopsis LTP cDNA; pANA2 (C), containing the 3’ portion of the cDNA;
and pANA3 (D), containing the full length cDNA. Some plants were also
transformed with pBl121 (A), as a negative control. The NPTll gene confers
kanamycin resistance. Constructs were prepared by J. Botella.

147

Table 5-1. Antisense lines tested by Western analysis. This table represents
plants which had a reduction in LTP in the primary screen and were later
rescreened. There are approximately 30 lines which showed no reduction in LTP
level in the primary screen, and these plants are not shown on this table. The
relative level of reduction of LTP is shown by minus signs. (--) indicates that the
level of protein was reduced to 5-10% of wild type levels, (-) indicates that the
level of protein was reduced to 25-30% of wild type levels, (-) indicates that the
level of protein was reduced to about 50% of wild type levels, and (0) indicates
that there was no noticeable reduction in LTP levels in comparison to wild type
plants.

Relative level

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Antisense line Antisense of LTP Parental
designation construct reduction ecotype
LTP1 pANA1 --- 8T0

I LTPZ pANA1 --- s'ro

H LTPB PANAB --— uo-o

fl LTP4 pANA2 --- RLD

I LTPS pANA3 --- RLD
LTP6 pANA3 -- RLD
LTP? pANA3 -- RLD
LTPB pANA3 - RLD
LTP9 pANA3 -- RLD
LTP10 pANA3 -- RLD

RLD
RLD
RLD
RLD
RLD
RLD
RLD
RLD .
RLD I

 

 

 

 

 

 

148

Table 5-1 continued

Relative level

 

       

 

 

 

 

 

 

 

 

 

 

 

 

Antisense line Antisense of LTP Parental
designation construct reduction V__ecc§yp¢_
0
0
LTP34 pANA1 ---
LTP35 pANA1 ---
LTP36 pANA2 -—-
LTP39 pANA2
LTP40 pANA2
LTP41 pANA1 0
LTP42 pANA2 ---
[L'rpu 4313111152 0

 

 

 

 

149

LTP18
LTP12

LTP5
LTP4
RLD

 

 

 

 

Figure 5-4. Western analysis comparing levels of LTP in wild type and transgenic
plants.

Equal amounts of total soluble leaf protein were resolved by SDS-PAGE,
transferred to nitrocellulose and immunodetected with anti-LTP antibodies. LTP4,
LTP5, LTP12 and LTP18 represent plants which have been transformed with
antisense LTP constructs. T17 represents a plant which has been transformed

with pBl121. RLD is the wild type parent.

150

be homozygous for kanamycin resistance and thus for the antisense LTP
construct. These lines were used for further analyses. All lines which were used
for further analysis were in the RLD background. There appears to be no
correlation with any particular antisense construct, and the level of reduction of
LTP (Table 5-1). LTP4 contains the construct pANA2, which harbors the 3’ half of
the ALTP cDNA, and LTP5, LTP12, and LTP 18 contain pANA3, which includes the
full length CDNA. Some plants containing pANA1, which harbors the 5’ half of the
cDNA, also have very low levels of LTP, but these plants were not in an RLD
background. T17, which contains pBl121, but no antisense cDNA, had similar

levels of LTP as wild type plants (Fig 5-4).

Phenotypic characten'zatian of transgenic plants - lf LTP played an essential
role in the plant, one would expect that plants which had a greatly reduced level
of the protein would exhibit some marked phenotypic changes with respect to
control plants. To quantitate the growth rate of the plants, a growth curve
measuring the fresh weight of the aerial portion of the plants was carried out.
There was essentially no differences between the growth rates of control and
antisense plants. LTP4, LTP5, and LTP18, were somewhat larger than the control
plants, throughout the period growth was measured, but LTP12 was smaller (Fig.
5-5). This minor variation does not correlate with the amount of residual LTP in the
antisense plants. Western analysis indicates that LTP4 has the greatest reduction

in protein levels, while LTP5, LTP18, and LTP12 appear to have slightly higher

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LTP4 LTP5
400 500
m-
A 300‘ + LTP4 a —o— LTP5
E -""' RLD é ‘ —o— RLD
E. '5.
'5 200 '6 5
1:" E
5' '5' =°°‘
8 8
§ 100- g
u‘ “- 100-
0 3 3 f o.
o 10 20 30 o 10 20 so
Time (Gav!) Time (days)
LTP12 LTP18
300
1
g E, —s— LTP18
E E —o— 121
3 '5
E a
g .9
3 m.
.C
3 5
“- u.
' ~ I . o :. ;_ . . , ,
Time (days) Time (days)

Figure 5-5. Growth of wild type and antisense plants.

Means of measurements of fresh weight of the entire aerial portion of the plant
(n=10) were determined. Each graph represents the growth of one antisense line
compared to wild type plants and control plants containing a CaMV-GUS fusion
(T17). Where no error bar (SE) is shown, the SE is less than the size of the

symbol.

152

levels. During the first two weeks of growth, antisense and control plants exhibited
no obvious morphological or developmental differences (Fig. 5-6A). At 14 days
after germination, control plants started bolting, but flowering was delayed in the
antisense plants (Fig. 5-6B, 5-7). LTP4, LTP5, and LTP18 started bolting 2-3 days
after control plants, and bolt initiation was delayed for 7 days in LTP12 (Fig. 5-7).
The germination rate of antisense and control plants was also tested by planting
a known number of seed in a pot and counting the number of resulting plants (Fig.
5-8). Germination rates for the control plants were approximately 80%. The
germination rates of antisense plants ranged from 60% to 90%. Again, there
appeared to be no correlation between the amount of residual LTP in the antisense
lines and germination rates.

To examine a possible role of LTP in cuticle formation, analysis of wax and
cutin composition in wild type and transgenic plants was performed using gas
chromatography. A comparison of chromatographs shows that there is no
quantitative or qualitative differences in either the wax or cutin of the low LTP
transgenic plants (Figs. 5-9, 5-10). The results of the analysis for LTP5 is
presented here as it is representative of all transgenic lines. To test if the
reduction in LTP had an effect on the structure of the epicuticular wax, the surface
of the stems of wild type and antisense plants was examined by scanning electron
microscopy (Fig. 5-11). The low levels of LTP appear to have no effect on the wax
structure. Also, transmission microscopic analysis of the stem cuticular layer

showed no obvious differences between wild type and antisense plants (Fig. 5-12).

153

Figure 56. Control and antisense plants at 15 and 25 days after planting.

A. Wild type, T17, and antisense plants at 15 days.
B. Wild type, T17, and antisense plants at 25 days.

 

155

LTP4 LTP5

 

 

Plants flowering (96)
8

Plants flowering ($6)
8

 

 

 

 

 

 

Time (days) Time (days)

LTP12 LTP18

 

 

tool "“— “912 1W“ —s— ltpla

m ‘ * M
—"- ‘2' —o— 121

Plants llowenng ($6)
8

Plants flowering (‘16)
8

 

 

 

 

 

 

AA A
o ' ———---' V o v -'

O 10 20 30 0 10 20 30

5 Time (days) Time (days)

Figure 5-7. Time of flowering in control and antisense plants.

Control and antisense plants were analyzed for the time of flowering. Plants
(n210) were observed from the time the first plant started flowering until all plants
were flowering. Each graph represents the percentage of plants flowering of one
antisense line as compared to the control plants. Where no error bar (SE) is
shown, the SE is less than the size of the symbol.

156

 

 

Percent Germlnatlon

 

 

 

Figure 58. Percent germination of antisense and control plants.

A known number (n=32) of seed were planted, and the number of plants which
resulted were counted.

157

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

... l
‘é’
m 2.0e5~
m .
a: l
I
o <
'- 4
8 109.5
I: ’ .
D
m .
a . l
a 1L 5 - .. t
c: 0 ‘ - , . .
o 20
A RETENTION TIME (min)
La)” .
§ ] RLD
a) 20e5~
LIJ .
II
I
o l
'—
8 to l
. 95
E l
O
I.“ «
E .
36 o ~~
0 2b
B RETENTION TIME (min)

Figure 59 Gas chromatography analysis of wax composition.

Epicuticular wax was extracted from leaves of wild type (RLD) and antisense
plants, and separated by gas chromatography on a 15 m X 0.53 mm ID Supelco
SPB-1 column. The elution profile of LTP5 is representative of all antisense lines
tested. This analysis was performed by U. Hecht.

158

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UJ 8.0e5 « LTP5
CD .
z
O
35 4
Lu 6.0e5 «
a 1
t:
O 1
I— .
8 4.0e5 «
l— .
Lu
0 l
g: l
i: 2.0e5?
”SJ .
O: 1 Mi}

0 .._l 4 ‘_¥

0 10 , . . 2;

A RETENTION TIME (min)
LU 8.0e5 RLD
a)
E .
fl .
m 6.0e5 <
a: .
O
5 .
UJ 4.0e5 =
[a .
D
g .
5: 2.085 =
j 4 “Metal L
g 4

o ._ l -

o 10 Y 2c?

8 RETENTION TIME (min)

Figure 5-10. Gas chromatography analysis of cutin composition.

Cutin was extracted from leaf tissue of wild type (RLD) and antisense plants and
separated by gas chromatography on a 15 m X 0.53 mm ID Supelco SPB-1
column. The elution profile of LTP5 is representative of all antisense lines tested.
This analysis was performed by U. Hecht.

159

Figure 5-11. Scanning electron microscopy analysis of epicuticular wax of wild
type and antisense stems.

Stem tissue of wild type (RLD) and antisense plants was frozen and viewed by
cryo scanning electron microscopy.

A. LTP 5

B. LTP12

C. LTP18

D. RLD

 

160

 

161

Figure 5-12. Transmission electron microscopy analysis of the wax and cutin of
wild type and antisense stems.

Stem tissue of wild type (RLD) and antisense plants was fixed, sectioned, and
viewed with a transmission electron microscope. On average, 40 sections of each
of 5 sample preparations were viewed.

A. LTP5

B. LTP12

C. RLD

162

 

163

To test the possible role of LTP in plant defense, control and antisense
plants were infected with the Arabidopsis fungal pathogen, Erysiphe cruciferae.
The results of this analysis were very ambiguous, as there was a variety of levels
of fungal growth on the infected plants, with no correlation as to the presence or

absence of an antisense construct.

DISCUSSION

Lipid transfer proteins were originally isolated based on their ability to
transfer phospholipids between natural and artificial membranes in vitro. It has,
however, been difficult to assign an in viva role to these proteins. Non-specific
LTPs have been isolated from several plant species, and recent evidence indicates
that these proteins do not play a part in intracellular lipid transfer. A maize LTP
was shown to be synthesized on membrane bound polysomes (T chang, et al.,
1988; Vergnolle, et al., 1988) and a spinach LTP was shown to be cotranslationally
inserted into the ER (Bernhard, et al., 1991; chapter 2 of this dissertation), thus
indicating that it enters the secretory pathway. A barley LTP has been found in
aleurone cell culture medium (Mundy and Rogers, 1986) and a carrot LTP has
been found in carrot embryogenic cell culture medium (Sterk, at al., 1991). More
recently, an Arabidopsis LTP has been localized to the cell wall (Thoma, et al.,
1993; chapter 3 of this dissertation).

In addition to its extracellular location, LTP is synthesized in very specific

164

tissue and cell types. LTPs from carrot, maize, tobacco are found exclusively in
the aerial portions of mature plants (Sterk, at al., 1991; Sossountzov, et al., 1991;
Fleming, at al., 1992), and an Arabidopsis LTP is found predominantly in the aerial
parts of mature plants (Thoma, at al., 1993; chapter 3 of this dissertation).
Immunocytochemical analysis showed that an Arabidopsis LTP is localized mainly
to epidermal cells (Thoma, et al., 1993; chapter 3 of this dissertation), and in situ
hybridization showed that expression of a tobacco LTP was confined mainly in
epidermal cells of the aerial portion of plants and the L1 layer of the shoot
meristem (Fleming, at al., 1992). A second tobacco LTP is expressed
predominantly in tapetal cells (Koltunow, et al., 1990). Analysis of transgenic
Arabidopsis containing an Arabidopsis LTP promoter-G US fusion, showed that
expression in adult plants is most prominent in stem and leaf epidermal cells, leaf
vascular tissue, the stigma, floral nectaries, pollen grains, and at the base of lateral
roots (chapter 4 of this dissertation).

Clearly, the extracellular location of the protein and its cell-type specific and
tissue specific expression preclude a role for this protein in intracellular lipid
transfer. In an effort to demonstrate an in viva function for plant nsLTPs transgenic
Arabidopsis plants that contain an Arabidopsis nsLTP cDNA in reverse orientation
behind the CaMV 35$ promoter have been produced. There have been numerous
reports in which gene expression of transgenic plants has been inhibited by
antisense RNA (Ecker and Davis, 1986; Smith, et al., 1988; Van der Krol, et al.,

1988). The reduction of target gene expression using antisense techniques has

165

been useful to generate mutants for the discovery of the biochemical and
biological function of target genes; for example, the role of the gene pTOM13 in
ethylene biosynthesis was determined in this manner (Hamilton, et al., 1990).
Arabidopsis LTP appears to be encoded by at least two genes as shown by high
stringency Southern analysis (chapter 4 of this dissertation). Assuming that these
genes are similar enough that both would be impaired by this technique, I should
be able to determine the function of the protein within the plant.

Antisense RNA techniques allowed us to produce plants which contained
reduced levels of LTP. There were varying levels of reduction with the most
dramatic containing no more than 10% of LTP as their wild type counterparts. If
LTP has an essential function in a plant under normal growth conditions, one
would expect to see an obvious alteration in phenotype of the antisense plants.
There is essentially no consistent difference in the growth rates (accumulation of
fresh weight) or germination rates between control and antisense plants. There
is a delay in flowering in the antisense plants, but at this time, this phenotype
cannot be reconciled with any possible role this protein may have. A more
detailed analysis of the late flowering phenotype is necessary.

Based on their extracellular location, cell- and tissue-type specific
localization and expression patterns (predominance in epidermal cells of aerial
portions of plants), and ability to bind fatty acids, it has been proposed that LTPs
may play a role in cuticle formation (Stark, et al., 1991; Thoma, et al., 1993:

chapter 3 of this thesis). To test this possible function for LTP, the wax and cutin

166

composition of antisense and wild type plants has been analyzed by gas
chromatography. Comparison of the results shows that there is essentially no
difference, quantitatively or qualitatively, in the wax or cutin of the antisense plants.
Low temperature scanning electron microscopy was used to observe the structure
of the wax on the stems of antisense and wild type plants and transmission
electron microscopy was used to compare the cuticle. Again, no difference was
observed between the transgenic and control plants. These data indicate that
either LTP is not involved in cuticle formation or that the protein is still present in
levels sufficient to carry out its role under normal growth conditions. As the cuticle
is the mai or barrier to water loss in plants (Holloway, 1982), wax and cutin samples
need to be extracted from plants which were grown under conditions, such as
drought or salt stress, in which the cuticle becomes a more limiting component in
plant survival.

The isolation of proteins which are homologous to LTPs and show
antifungal and antibacterial activity in vitro (T erras, et al., 1992; Garcia-Olmedo.
personal communication), has led to a hypothesis that LTPs carry out a defense
role in viva. To test this hypothesis, antisense and control plants were inoculated
with the Arabidopsis fungal pathogen, Erysiphe cruciferae. After one week, fungal
growth on all plants was apparent, but there was no correlation between amount
of fungal growth and the presence or absence of an antisense construct. Infection
with E. cruciferae, an obligate pathogen, involves rubbing an infected leaf onto the

leaf of a test plant. This method of inoculation is not quantitative, so if there is a

167

slight difference in susceptibility to this pathogen in the antisense plants, it will not
be apparent. This type of test needs to be repeated using fungal or bacterial
pathogens in which the amount of inoculum can be monitored more carefully.

The expression pattern of an Arabidopsis LTP - B-glucuronidase (GUS)
fusion in transgenic plants has led to the hypothesis that LTPs may be involved in
phenylpropanoid metabolism (chapter 4 of this dissertation). To test this
hypothesis, staining of stem sections could be carried out to determine if the lignin
content of antisense plants has been altered. Lignin and phenolic acid content of
antisense plants could also be determined by gas chromatographic analysis (C.
Chapple, personal communication). Results of a quantitative pathogenesis
experiment may be helpful here, as phenylpropanoid compounds are induced
upon pathogen infection (Dixon and Lamb, 1990). Again, a limiting factor to any
of these experiments is the presence of two LTP genes in Arabidopsis and the fact
that the low level of protein left in these plants may be enough to carry out the its
role.

In conclusion, the role of LTP within a plant is still unknown. Possibly,
further analysis of the antisense plants described in this paper will provide an

answer.

1 68
REFERENCES

Arondel, V. and Kader, J.-C. 1990. Lipid transfer in plants. Experientia 46:579-585.

Arondel. V., Vergnolle, C., Tchang, F. and Kader, J.-C. 1990. Bifunctional lipid-
transferzfatty acid-binding proteins in plants. Mol. Cell. Biochem. 98:49-56.

Bankaltls. V.A.. Altken, J.R., Cloves, AB. and Dowhan, W. 1990. An essential role
for a phospholipid transfer protein in yeast Golgi function. Nature 347:561-562.

Bernhard, W.R. and Somervllle, C.R. 1989. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from
amino acid sequence homology. Arch. Biochem. Biophys. 269:695-697.

Bernhard, W.R., Thoma. S., Botella, J. and Somervllle, C.R. 1991. Isolation of a
cDNA clone for spinach lipid transfer protein and evidence that the protein is
synthesized by the secretory pathway. Plant Physiol. 95:164-170.

Boulllon, P., Drlschel, C., Vergnolle, C., Duranton, H. and Kader, J.—C. 1987. The
primary structure of spinach-leaf phospholipid-transfer protein. Eur. J. Biochem.
166:387-391.

Breu, V., Guerbette, F., Kader, J.-C., Gamlnl Kannangara, C., Svensson, B. and
van Wettsteln -Knowlee, P. 1989. A 10 kD barley basic protein transfers
phosphatidylcholine from liposomes to mitochondria. Carlsberg Res. Commun.
54:81-84.

Clevee, A.E., McGee, T.P., Whitters, EA, Champlon, KM., Altken, J.R., Dowhan,
W., Goebel, M. and Bankallls, VA. 1991. Mutations in the CDP-choline pathway for
phospholipid biosynthesis bypass the requirement for an essential phospholipid
transfer protein. Cell 64:789-800.

Croteau, R. and Kolattukudy, RE. 1974. Biosynthesis of hydroxyfatty acid
polymers. Enzymatic synthesis of cutin from monomer acids by cell-free
preparations from the epidermis of Wcia faba leaves. Biochemistry 13: 3193-3202.

DleryclgW" Gautier, M..,-F Lulllen, V. andJoudrler, P. 1992. Nucleotide sequence
of a cDNA encoding a lipid transfer protein from wheat (Triticum durum Desf. )
Plant Mol. Biol. 19. 707-709.

Dixon, RA. and Lamb, C.J. 1990. Molecular communication in interaction between
plants and microbial pathogens. Annu. Rev. Plant Physiol. Plant Mol. Biol. 41 :339-
367.

169

Douady, D., Grosbob, M., Guerbette, F. and Kader, J.-C. 1982. Purification of a
basic phospholipid transfer protein from maize seedlings. Biochim. Biophys. Acta
710:143-153.

Ecker, J.R. and Davis, R.W. 1986. Inhibition of gene expression in plant cells by
expression of antisense RNA. Proc. Nat. Acad. Sci. USA 83:5372-5376.

Felnberg, AP. and Vogelsteln, B. 1983. A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-
13.

Fleming, M., Mandel, 'r. Hofman, s.. Sterk, P., de Vrles, so. and Kuhlemeler, c.
1992. Expression pattern of a tobacco lipid transfer protein gene within the shoot
apex. Plant J., in press.

Hamilton, A.J., Lycett, G.W. and Grierson, D. 1990. Antisense gene that inhibits
synthesis of the hormone ethylene in transgenic plants. Nature 346:284-287.

Helmkamp, GM. 1990. Transport and metabolism of phosphatidylinositol in
eukaryotic cells. In: Subcellular biochemistry, H.J. Hilderson, ed., Plenum
Publishing Corp., pp.129-174.

Holloway, P.J. 1982. Structure and biochemistry of plant cuticular membranes: an
overview. In: The plant cuticle. D.F Cutler, K.L. Alvin, and CE. Price, eds.
Academic Press, London, pp. 1-32.

Kader, J.-C., Jullenne, M. and Vergnolle, C. 1984. Purification and characterization
of a spinach-leaf protein capable of transferring phospholipids from liposomes to
mitochondria or chloroplasts. Eur. J. Biochem. 139:411-416.

Kolattukudy, PE. 1980. Cutin, suberin, and waxes. In: The Biochemistry of Plants,
Vol. 4, Lipids: Structure and Function. P.K. Stumpf and E.E. Conn, eds. Academic
Press, NY, pp. 571-645.

Koltunow, A.M., Tmettner, J., Cox, K.H., Wallrolh, M. and Goldberg, BB. 1990.
Different temporal and spatial gene expression patterns occur during anther
development. Plant Cell 2:1201-1224.

Maniatis, T., Frltech, E.F. and Sambrook. J. 1982. Molecular Cloning. A Laboratory
Manual. Cold Spring Harbor, New York.

Mundy, J. and Rogers. J.C. 1986. Selective expression of a probable
amylase/protease inhibitor in barley aleurone cells: comparison to the barley
amylase/subtilisin inhibitor. Planta 196:51-63.

170

Munro, S. and Pelham, H.H.B. 1987. A C-terminal signal prevents secretion of
luminal ER proteins. Cell 48:899—907.

Murashige, J. and Skoog, F. 1962. A revised medium for rapid growth and bio
assay with tobacco tissue cultures. Physiol. Plant. 115: 473-497.

Smlth, C.J.S., Watson, G.M., my, J., Blrd, C.R., Morrle, P.C., Schuch, W. and
Griereon, D. 1988. Antisense RNA inhibition of polygalacturonase gene expression
in transgenic tomatoes. Nature 334724726.

Sossountzov, L, Ruiz-Avila. L, Vlgnols, F., Jolllot. A., Arondel, V., Tcahng, F.,
Grosbois, M., Guerbette, F., Mignlac, E., Delseny, M., Pulgdomenech, P. and
Kader, J.-C. 1991. Spatial and temporal expression of a maize lipid transfer protein
gene. Plant Cell 3:923-933.

Sterk, P., Boolf, H.. smellekene, GA, Van Kammen, A. and De Vrlee, S.C. 1991.
Cell-specific expression of the carrot EP2 lipid transfer protein. Plant Cell 3:907-
921.

Taklehlma. K., Watanbe, 8., Yamada, M. and Mamlya, G. 1986. The amino-acid
sequence of nonspecific lipid transfer protein from germinated castor bean.
Biochim. Biophys. Acta 870:248-255.

Tchang, F., Thle, P., SIIefel, V., Arondel, V., March, M.-D., Pages, M.,
Pulgdonmenoh, P., Grellet, F., Delaney, M., Boulllon, P., Huet, J.—C., Guerbette, F.,
Beeuvale-Cante, F., Duranton, H. and Kader, J.-C. 1988. Phospholipid transfer
protein: full length cDNA and amino acid sequence in maize. J. Biol. Chem.
263:16849-16855.

Terrae, F.B.G., Goderls, l.J., Van Leuven, F., Vanderleyden, J., Cammue, EPA
and Broekart, W.F. 1992. In vitro antifungal activity of a radish (Raphanus sativus
L.) seed protein homologous to nonspecific lipid transfer proteins. Plant Physiol.

100:1055-1058.

Thoma, S.L, Kaneko, Y. and Somervllle, C. 1993. An Arabidopsis lipid transfer
protein is a cell wall protein. Plant Journal, in press.

Towbin, H.. Staehelin, T., and Gordon, J. 1979. Electrophoretic transfer of proteins
from polyacrylamide gels to nitrocellulose sheets: procedure and some practical
applications. Proc. Natl. Acad. Sci. USA 76: 4350-4354.

Valvekens, D., Van Montaau. M. and Van Lisebettens, M. 1988. Agrabacterium
tumefaciens-mediated transformation of Arabidopsis thaliana root explants by
using kanamycin selection. Proc. Natl. Acad. Sci. USA 85:5536-5540.

171

vm der Krol, A.R., Lentlng, P.E., Veenslra, J., van der Meer, I.M., Koes, R.E.,
Gerate, AG.M., Mal, J.N.M. and Stullle, AR 1988. An antisense chalcone synthase
gene in transgenic plants inhibits flower pigmentation. Nature 333:866-869.

Vergnolle, C., Arondel, V., Tchang, F., Grosbois, M., Guerbette, F., Jolllot. A. and
Kader, J.-C. 1988. Synthesis of phospholipid transfer proteins from maize
seedlings. Biochem. Biophys. Res. Commun. 157:37-41.

Mrlz. KWA. 1982. Phospholipid transfer proteins. In: Lipid-protein interactions,
P.C. Jost and OH. Griffith, eds., Elsevier, Amsterdam. 99. 151-223.

CHAPTER 6

SUMMARY AND CONCLUSION

172

173

Lipid transfer proteins were first isolated based on their ability to enhance
the rate of lipid transfer between natural and artificial membranes in vitro. It has
been a long held assumption that these proteins have a similar activity in viva, but
there has been no experimental evidence to support this supposition. My research
began as an attempt to examine the role of plant nsLTPs in membrane biogenesis
and to use existing technologies to demonstrate their role in this process in viva.
However, as my research progressed, it became clear that plant nsLTPs were not
directly involved in intracellular lipid transfer, and I began a search for an alternate
function. This chapter is a summary of the evidence which demonstrated that
nsLTPs are not actually involved in lipid transfer (chapters 2 and 3 of this
dissertation), the experimental observations which led to the formation of
hypotheses about the role of these proteins (chapters 3 and 4 of this dissertation),
the results of the experiments carried out to test these hypotheses (including
chapter 5 of this dissertation), and suggestions on future research.

The proposed mechanism of LTP mediated phospholipid transfer involves
the extraction of a lipid molecule from the cytoplasmic face of the ER (or other
biosynthetic membrane such as the chloroplast), movement of the lipid-protein
complex through the cytosol, and release of the lipid molecule to the outer leaflet
of the target membrane. The LTP would then be able to extract a lipid from the
target membrane (in the case of the exchange process) or be free to diffuse back
to the biosynthetic membrane to bind another lipid (in the case of net transfer).

Such a mechanism of action implies that LTPs are cytosolic, and there are

 

174
numerous reports which state that LTPs are cytoplasmic proteins (Arondel and

Kader, 1990; Helmkamp, 1990). These assertions are based on the fact that LTPs
are soluble proteins and that a direct role in intracellular lipid transfer obviates the
need for a cytoplasmic location. However, there is no experimental evidence to
support such statements.

There is much recent evidence which demonstrates that plant nsLTPs are,
in fact, extracytoplasmic. Comparison of the deduced amino acid sequence of a
spinach LTP cDNA (Bernhard, et al., 1991; chapter 2 of this dissertation) with a
directly determined amino acid sequence (Boulllon, at al., 1987) revealed the
presence of a 26 amino acid N-terminal extension. The extension had all the
properties of a signal peptide which directs proteins into the ER (van Heine,
1985): (1) an amino terminal location, (2) 20-40 amino acids in length, (3) Ala or
other small, uncharged residue in positions -1 and -3 from the cleavage site, (4)
a helix breaking residue such as Pro at the -5 position, and (5) a stretch of 810
small, uncharged residues N-terminal to the helix breaking residue. The presence
of a putative signal peptide was very surprising considering the supposed
cytosolic location and function of LTPs. Proteins which enter the secretory system
are either secreted or become localized within a compartment of the
endomembrane system, 6.9., the ER, the Golgi apparatus, or the vacuole
(Bednarek and Raikhel, 1992).

In vitro transcription and translation in the presence of microsomal

membranes and protease protection experiments demonstrated that the extension

 

 

175
was a functional signal peptide directing the spinach LTP into the ER (Bernhard,

et al., 1991; chapter 2 of this thesis). This supports the observation that a maize
LTP is synthesized on membrane bound polysomes (Vergnolle, et al., 1988). It
has subsequently been shown that a barley LTP (Barley I) also contains a signal
peptide (Madrid, 1991), and sequence analysis shows that carrot (Sterk, at al.,
1991), caster (T suboi, et al., 1991), tobacco (Fleming, at al., 1992), tomato (T arres-
Schumann, et al., 1992), wheat (Dieryck, at al., 1992), and Arabidopsis (chapter
5 of this dissertation) contain similar N-terminal sequences.

The entrance of plant nsLTPs into the secretory pathway is inconsistent with
the proposed role of the protein in intracellular lipid transfer. To determine the
precise location of the protein, and to resolve the possibility that the protein may
move from the ER to the cytoplasm by some novel mechanism, I produced
antibodies against an Arabidopsis LTP:protein A fusion and localized the protein
at the ultrastructural level using immunoelectron microscopy (chapter 3 of this
dissertation; Thoma, et al., 1993). The protein was found exclusively in the cell
wall, and it was most concentrated in epidermal cells. The cell wall localization
was consistent with other accounts reporting that plant nsLTPs were secreted. A
barley protein which was secreted in aleurone cell culture media (Mundy and
Rogers, 1986), was shown to contain considerable homology to nsLTPs (Bernhard
and Somerville, 1989) and to stimulate PC transfer from microsomes to potato
mitochondria in in vitro assays (Breu, at al., 1990). It has also been shown that a

carrot LTP is secreted by carrot embryogenic cell cultures (Stark, et al., 1991).

 

176

As all of this evidence pointing to an extracellular location accumulated, and
it became apparent that plant nsLTPs could not be directly involved in intracellular
lipid transfer or membrane biogenesis, it became imperative to try to determine the
actual function of these proteins. The next part of this chapter will discuss
suggested roles that LTP may have within a plant, and evidence which led to such
suggestions. Finally, recommendations on future research directions will be

discussed.

Glycerlde synthesis
Lipid metabolism in developing seeds consists mainly of triacylglycerol

(TAG) synthesis. TAGs serve primarily as a storage form of carbon in developing
seeds and rapid synthesis of lipids has been observed during the period of seed
weight gain (Murphy, at al., 1989). TAG synthesis and modification occurs in the
ER and these lipids are ultimately stored in oil bodies in the mature seed (Stymne
and Stobart, 1987).

The occurrence of nsLTP in developing maize and castor seeds (Grosbois,
et al.,1989; Yamada, et al., 1990) and its expression in aleurone tissue (Skriver, et
al., 1992) led to a hypothesis that the protein may act as an acyl carrier in TAG
synthesis (Yamada, et al., 1990). This hypothesis, however, is inconsistent with the

presence of LTP in non-lipid storing tissues. It is also inconsistent with the cell wall

localization of the protein.

 

177

Fatty acid degradatlon

Upon germination of many seedlings, breakdown of reserve triglycerides
begins and its products support seedling growth. Their degradation proceeds
mainly by oxidation in the B-position to the carboxyl group and sequential removal
of carbon units. The process of B-oxidation occurs in the glyoxosomes where it
is coupledto the glyoxolate cycle. This allows for the conversion of the carbon to
isocitrate which can eventually be shuttled to the cytoplasm and converted to
sucrose (Andrews and Ohlrogge, 1990).

lmmunoelectron microscopy has demonstrated that caster nsLTPA is
localized to the glyoxysomes and secondary cell wall of xylem vessels in castor
bean cotyledons (T suboi, at al., 1992). This partial localization to the glyoxysomes
plus the demonstration that nsLTPA could bind oleic acid and oleoyl CoA (T suboi,
et al., 1992) led to the hypothesis that this nsLTP could function as an acyl carrier
in B-oxidation. It has also been proposed that nsLTPA could function to enhance
the activity of acyl CoA oxidase, the rate limiting enzyme in B-oxidation, as it was
observed that the presence of nsLTPA increased enzyme activity in an in vitro
assay (T suboi, et al., 1992).

However, there are some uncertainties regarding this hypothesis. First,
subcellular fractionation showed that only 13% of the nsLTPA was found in the
glyoxysomal fraction, and the ma'or portion was found in the soluble fraction
(T suboi, et al., 1992). Proteins which are targeted to glyoxysomes or peroxisomes

are synthesized in the cytosol and are post-translationally transported into the

 

178
organelle. Thus, it is suggested that at least part of the protein associated with the

soluble fraction represents cytosolic protein. Secondly, signals which direct
proteins into glyoxysomes are generally a small sequence found on the C—terminal
portion of the protein (Gould, et al., 1989), and there is no such sequence found
in castor nsLTPA. Tsuboi and coworkers also show evidence that castor nsLTPA
is processed to a mature form in the presence of purified glyoxysomes. Import
of proteins into peroxisomes or glyoxysomes is not generally associated with the
removal of a presequence or with any other modification of the imported protein
(Gould, at al., 1989). There have been exceptions to this rule, for example a
watermelon glyoxysomal malate dehydrogenase is synthesized with an amino-
terminal transit peptide (Gietl, 1990). This transit peptide, however, bears no
resemblance the N—terminus of castor nsLTPA. The castor nsLTPA N-terminal
signal (T suboi, at al., 1991) has features representative of a signal peptide
directing proteins to the ER (von Heine, 1985). This is more consistent with the
co-localization of the protein to vessel cell walls, than with the localization of the
protein to the glyoxysomes. It is a possibility that the glyoxysome preparations
were contaminated with ER. In addition, the antibodies used for the localization
were not affinity purified, which could lead to artifacts. The discrepancy between
the localization of the protein to two locations which require different targeting
mechanisms, and the lack of a peroxisomal/glyoxysomal targeting signal on the

castor LTP, allow for reservations concerning the role of this protein in fatty acid

degradation.

 

 

179
Cuticleformatlon

The cuticle is a continuous layer of predominantly lipophilic material found
on the outermost surface of the aerial portions of plants (Holloway, 1982). A major
component of the cuticle is cutin, an insoluble polymer composed mainly of C“,
and 0,. hydroxylated and epoxygenated fatty acids (Kolattukudy, 1982). Another
component of the cuticle is wax. Plant epicuticular waxes are generally composed
of a diverse mixture of long chain hydrocarbons, including long chain (Cm-Ca.) fatty
acids (Kolattukudy, 1982). The aliphatic components of the cuticle are synthesized
in the epidermal cells and are delivered to the outside of the cell where it is
thought that esterification takes place (Croteau and Kolattukudy, 1974). The
mechanism by which the monomers are delivered to the outside of the cell is
unknown. It is possible that nsLTPs could bind the aliphatic monomers, carry
them through the aqueous environment of the cell wall, releasing them once they
have transversed the wall.

Tissue-type, cell-type, and subcellular localization of an Arabidopsis LTP
(chapter 3 of this dissertation; Thoma, at al., 1993) and expression patterns of the
carrot, maize, and tobacco LTP genes (Sterk, et al., 1991; Sossountzov, et al.,
1991; Fleming, at al., 1992), are all consistent with a hypothesized role for LTP in
cuticle formation.

lmmunoblot analysis demonstrated that Arabidopsis LTP is mainly present
in the aerial portions of the plant; flowers, siliques, stems, petioles, and leaves,

very little LTP present in root tissue. Immunocytochemical localization shows that

 

 

180
the Arabidopsis LTP is present mainly in the cell walls of epidermal cells, the cell

type in which the synthesis of cuticle monomers occurs. LTP is seen to be
particularly concentrated in guard cells. This is also consistent with a role in cuticle
formation. Normally, epidermal cells contain a cuticle layer only on their outer
surface. Guard cells, however, are covered with a thick cuticle that not only
extends over the outer surface, but also the surfaces facing the substomatal
chamber and the stomal pore (Esau, 1977). LTP is also heavily concentrated in
the cells walls surrounding cells at or near the stigma surface. The epidermal and
subepidermal layers of a stigma can produce a secretion containing lipid
components and phenolic compounds (Martin and Brewbaker, 1971) and it is
possible that LTP could play a role in the deposition of the lipid components of the
secretion. There is a small amount of LTP associated with the epidermal cells of
root tissue and LTP promoter-GUS fusions are active at the base of lateral roots
(chapter 4 of this dissertation). This is inconsistent with a role for LTP in cuticle
formation as roots are not generally thought to contain cutin.

RNA gel blot analysis shows that the carrot LTP is expressed in
embryogenic cell cultures, the shoot apex of seedlings, developing flowers, and
maturing seed (Sterk, et al., 1991). A tobacco LTP has been shown to be
expressed in the aerial parts of plants, primarily in epidermal cells and in the L1
layer of the shoot meristem (Fleming, et al., 1992). A maize LTP is highly
expressed in the epidermal cells of the aerial portions of maize seedlings

(Sossountzov, et al., 1991). In situ hybridization patterns show that the carrot LTP

 

181

mRNA is present in the protoderm (a cell layer which gives rise to the epidermis)
of somatic and zygotic embryos, in the tunica (top cell layer(s)) of the shoot apical
meristem, transiently in the epidermis of leaf primordia and floral organs, and in the
pericarp epidermis, seed coat, and outer epidermis of the integument in maturing
seeds (Stark, et al., 1991). As with the Arabidopsis LTP, the known locations of
cuticle formation correspond well with the locations in which the carrot LTP gene
is expressed. It has been suggested (Stark, et al., 1991) that the presence of LTP
in embryos may represent a mechanism to slow water uptake in a hypotonic
environment, thus restricting unlimited and potentially destructive cell expansion.

To test the role of LTP in cuticle formation, Arabidopsis plants containing a
greatly reduced level of LTP due to the introduction of an Arabidopsis LTP in
reverse orientation between the CaMV 35$ promoter and nos terminator were
produced (chapter 5 of this dissertation). Gas chromatographic analysis revealed
no qualitative or quantitative differences in the epicuticular wax or cutin
composition between antisense and wild type plants. Also electron microscopic
analysis revealed no differences in the wax or cutin structure (chapter 5 of this
dissertation). These data suggest that LTP is not involved in cuticle formation.
However, the negative results achieved by the antisense approach are not
conclusive. l was unable to obtain any plants that were completely devoid of LTP.
Thus, although the level of LTP in the plants used for analysis was greatly reduced,
it could be present at levels sufficient to carry out its function under normal growth

conditions. Also, it has been demonstrated that there are at least two closely

182
related LTP genes in Arabidopsis (chapter 4 of this dissertation). The antisense

approach may not have been successful in reducing the expression of both genes. '
Future experiments include analysis of wax and cutin composition of plants grown
under stress conditions where the cuticle is essential for the survival and/or growth

of the plant.

Protection again“ pathogen attack

Recently, a basic, 9 kD protein, which is homologous to nsLTPs, has been
isolated from radish seeds and shows antifungal activity in vitro (T erras, et al.,
1992). Also, proteins homologous to LTPs, which strongly inhibit bacterial and
fungal pathogens in in vitro assays, have been isolated from etiolated barley and
maize leaves. (F. Garcia-Olmedo, personal communication).

In addition, thionins, which have toxic effects on bacteria, fungi, and insects
(Bohlman and Apel, 1991), have several features in common with plant LTPs; they
are low molecular weight, basic proteins with several conserved cysteine residues.
A barley leaf thionin is found in the cell wall, with the highest concentration in the
outer wall of the epidermal cell layer (Bohlman, at al., 1988). This localization is
similar to that of Arabidopsis LTP in all green tissue examined.

In an attempt to determine whether LTPs had an antifungal activity in viva,
we infected wild type and antisense plants with the Arabidopsis fungal pathogen,
Erysiphe cruciferae. The manner of infection involves rubbing an infected leaf onto

a test leaf, which gave ambiguous, non quantitative results. This type of test bears

 

183

repeating using a system which will allow for more controlled and quantitative

testing.

Phenylpropanold metabohm

The analysis of transgenic plants containing an Arabidopsis LTP promoter-
GUS fusion showed that the promoter was active in developing Arabidopsis
seedlings and the pattern of expression observed in cotyledons and primary
leaves generally followed that of xylem differentiation. In adult plants, expression
was found mainly in vascular tissue or cell types covered with lipophilic substances
(chapter 4 of this dissertation). Lignin, the structural polymer of xylem, and
phenolic acids, a component of cuticular type substances, are both products of
the phenylpropanoid pathway. In addition, analysis of the promoter region of the
ALTP1 gene revealed the presence of several sequence elements commonly found
in the promoters of biosynthetic genes of the phenylpropanoid biosynthetic
pathway (Cramer, et al., 1989; Lois, ef al., 1989; Ohl, et al., 1990). These data
suggest that LTP may have a role in phenylpropanoid metabolism, or that LTP
genes are regulated in a manner similar to genes encoding enzymes of the
phenylpropanoid biosynthetic pathway.

Phenylpropanoids are a class of plant natural products derived from
phenylalanine. These compounds play important roles in plant development and
in protection against environmental stress. For instance, fiavonoids are pigments

and UV protectants in epidermal cells, Iignin is the major structural component in

184

xylem cell walls, and suberin is a lipophilic substance commonly found in the
casparian strip of the endodermis (Esau, 1977). The induction of lignin in wheat
(Moerschbacher, at al., 1990), suberin deposition in potato (Roberts and
Kolattukudy, 1989), and the accumulation of phenylpropanoid derived phytoalexins
help protect a plant against mechanical damage and microbial attack (Dixon and
Lamb, 1990).

When plants containing the ALTP1 promoter-GUS fusion were stained with
X-Gluc 48 hours after infection with E. cruciferae, several regions of staining
occurred. Examination of this tissue showed that each area was associated with
tissue damage (chapter 4 of this dissertation). At this point it is not known whether
this is due to hyphal growth into the tissue or mechanical damage due to the
wounding procedure. The experiment should be repeated using other fungal and
bacteria pathogens with different methods of inoculation.

In future research, the induction of LTP under different biotic and abiotic
stresses needs to be examined. A tomato nsLTP is greatly induced in stems by
NaCI, mannitol treatment, and ABA treatment (T orres-Schumann, et al., 1992).
Also, a small, basic barley protein, which has nearly 50% sequence identity with
maize LTP, was shown to be drought induced (Plant, at al., 1991). The production
of phenylalanine ammonia-lyase (PAL), an enzyme in the phenylpropanoid pathway
which contains similar sequence elements with the Arabidopsis LTP1 gene, is
induced by wounding, light, and HgClz treatment. Transgenic Arabidopsis

containing LTP promoter-GUS fusions are the ideal system to test for stress

 

185

induction of LTPs, and the transgenic Arabidopsis containing an LTP antisense
gene will be useful in testing the role of LTP in such stress responses in viva.

In conclusion, the role of non-specific lipid transfer proteins in plants is still
unknown. Their extracellular location and cell-type and tissue-type specific
expression, precludes any direct role for these proteins in lipid transfer, but has
provides us ‘with other hypotheses concerning the role- of this protein in viva.
Transgenic Arabidopsis containing reduced levels of LTP will provide us a tool with

which to test my hypotheses, and hopefully elucidate the role of plant non-specific

lipid transfer proteins.

 

1 86
REFERENCES

Andrews, J.E. and Ohlrogge, J. 1990. Fatty acid and lipid biosynthesis and
degradation. In: Plant physiology, biochemistry and molecular biology. D.T. Dennis
and DH. Turpin, eds., Longman Scientific, Essex, pp. 351 -362.

Arondel, V. and Kader, J.-C. 1990. Lipid transfer in plants. Experientia 46:579-585.

Bednarek, S.Y. and Raikhel, NM 1992. Intracellular trafficking of secretory proteins.
Plant Mol. Biol. 20:133-150.

Bernhard, W.R. and Somervlfle, C.R. 1989. Coidentity of putative amylase inhibitors
from barley and finger millet with phospholipid transfer proteins inferred from
amino acid sequence homology. Arch. Biochem. Biophys. 269: 695-697.

Bernhard, W.R., Theme, 8., Botella, J. and Somervllle, C.R. 1991. Isolation of a
cDNA clone for spinach lipid transfer protein and evidence that the protein if
synthesized by the secretory pathway. Plant Physiol. 95: 164-170.

Bohlman, H.. Clamen, S., Behnke, 8., Glass, H.. I-llller, C., Rehmann-Phlllpp, U.,
Schrader, G., Burkholt, V. and Apel, K. 1988. Leaf-specific thionins of barley - a
novel class of cell wall proteins toxic to plant pathogenic fungi and possibly
involved in the defense mechanism of plants. EMBO J. 7: 1559-1565.

Bohlman, H. andApsl, K. 1991. Thionins. Annu. Rev. Plant Physiol. Plant Mol. Biol.
42: 227-240.

Boulllon, P., Drlschel, C., Vergnolle, C., Duranton, H. and Kader, J.-C. 1987. The
primary structure of spinach-leaf phospholipid-transfer protein. Eur. J. Biochem.
166:387-391.

Breu, V., Guerbette, F., Kader, J.-C.. Gemini Kannangara, C., Svensson, B. and
van Wettsteln -Knowles, P. 1989. A 10 kD barley basic protein transfers
phosphatidylcholine from liposomes to mitochondria. Carlsberg Res. Commun.
54:81-84.

Cramer, C.L, Edwards, K., Dron, M., Uang, x, Dlldlne, S.L, Bolwell, G.P., Dlxon,
RA, Lamb, CJ. and Schuch, W. 1989. Phenylalanine ammonia-lyase gene
organization and structure. Plant Mol. Biol. 12:367-383.

Croteau, R. and Kolattulmdy, RE. 1974. Biosynthesis of hydroxy fatty acid
polymers. Enzymatic synthesis of cutin from monomer acids by cell-free
preparations from the epidermis of Vicia faba leaves. Biochemistry 13: 3193-3202.

187

Dleryck, W., Gautier, M.-F., Men, V. and Joudrler, P. 1992. Nucleotide sequence
of a cDNA encoding a lipid transfer protein from wheat (Tn‘ticum durum Desf.)
Plant Mol. Biol. 19:707-709.

Dixon, R.A. and Lamb, C.J. 1990. Molecular communication in interaction between
plants and microbial pathogens. Annu. Rev. Plant Physiol. Plant Mol. Biol. 41 :339-
367.

Esau, K. 1977. Anatomy of seed plants, second edition. John Wiley and Sons, New
York. '

Flsmhg, A.J., Mandel, T. Holman, S., Sterk, P., de Vrles, S.C. and Kuhlemeler, C.
1992. Expression pattern of a tobacco lipid transfer protein gene within the shoot
apex. Plant J., in press.

Gletl, C. 1990. Glyoxysomal malate dehydrogenase from watermelon is synthesized
with an amino-terminal transit peptide. Proc. Natl. Acad. Sci. USA 87:5773-5777.

Gould, S.J., Keller, G.-A., Hosksn, N., Wldnsm, J. and Subrarmnl, S. 1989. A
conserved tripeptide sorts proteins to peroxisomes. J. Cell Biol. 108:1657-1664.

Grosbols, M., Guerbette, F. and Kader, J.-C. 1989. Changes in the level and
activity of phospholipid transfer protein during maturation and germination of maize
seeds. Plant Physiol. 90:1560-1564.

I-Islmkamp, GM. 1990. Transport and metabolism of phosphatidylinositol in
eukaryotic cells. In: Subcellular biochemistry, H.J. Hilderson, ed., Plenum
Publishing Corp., pp.129-174.

Holloway, P.J. 1982. Structure and biochemistry of plant cuticular membranes: an
overview. In: The plant cuticle. D.F Cutler, K.L Alvin, and CE. Price, eds.
Academic Press, London, pp. 1-32.

Kolattukudy, PE. 1980. Cutin, suberin, and waxes. In: The Biochemistry of Plants,
Vol. 4, Lipids: Structure and Function. P.K. Stumpf and E.E. Conn, eds. Academic
Press, NY. PP. 571-645.

Lois, R., Dietrich, A., Hahlbrock, K. and Schulz. W. 1989. A phenylalanine
ammonia-lyase gene from parsley; structure regulation, and identification of elicitor
and light-responsive cis-acting elements. EMBO J. 8:1641-1648.

Madrld, S. 1991. The barley lipid transfer protein is targeted into the lumen of the
endoplasmic reticulum. Plant Physiol. Biochem. 29:695-703.

 

188

Marlin, F.W. and Brewbaker, J.L 1971. The nature of stigmatic exudate and its role
in pollen germination. In: Pollen development and physiology. J. Heslop-Harrison,
ed. Butterworth, London, pp.262-272.

mm, B.M., Noll, U., Gorrlchon, L and Relsener, H.-J. 1990. Specific
inhibition of lignification breaks hypersensitive resistance of wheat to stem rust.
Plant Physiol. 932465-470.

Murphy, D.J., Cumrnins, I. and Kang, AS. 1989. Synthesis of the ma’or oil-body
membrane protein in developing rapeseed (Brassica napus) embryos: integration
with storage lipid and storage protein synthesis and implication for the mechanism
of all body formation. Biochem. J. 258:285-295.

Ohl, 8., Hedrlck, 8., Chery, J. and Lamb, C.J. 1990. Functional properties of a
phenylalanine ammonia-Iyase promoter from Arabidopsis. Plant Cell 2:837-848.

Roberts, E. and Kolattdmdy, RE. 1989. Molecular cloning, nucleotide sequence,
and abscisic acid induction of a suberization-associated highly anionic peroxidase.
Mol. Gen. Genet. 217:223-229.

Skrlver, K., Leah, R., Muller-Uri, F., Olsen, F.-L and Mandy, J. 1992. Structure and
expression of the barley lipid transfer protein gene Lth. Plant Mol. Biol. 18. 585-
590.

Sterk, P., Baal, H.. Scheflekens, GA, Van Kammen, A. and De Vrles, so. 1991.
Cell-specific expression of the carrot EP2 lipid transfer protein gene. Plant Cell 3:
907-921.

Stymne, S. and Stobart, AK. 1987. Triacylglycerol biosynthesis. In: The
biochemistry of plants, vol. 9, PK Stumpf and E.E. Conn, eds., Academic Press,
New York, pp. 175-214.

Terras, F.B.G., Goderls, l.J., Van Leuvsn, F., Vanderleyden, J., Cammue, EPA
and Broekart, W.F. 1992. In vitro antifungal activity of a radish (Raphanus safivus
L) seed protein homologous to nonspecific lipid transfer proteins. Plant Physiol.
100:1055-1058.

Thoma, S.L, Kaneko, Y. and Somervllle, C. 1993. An Arabidopsis lipid transfer
protein is a cell wall protein. Plant Journal, in press.

Torres-Schumann, S., Godoy. J.A. and Pinter-Taro, J.A. 1992. A probable lipid
transfer protein gene is induced by NaCl in stems of tomato plants. Plant Mol. Biol.
18:749-757.

 

 

189

Tsubol, 8., Watanabe, 8., Ozekl, Y. and Yamada, M. 1989. Biosynthesis of
nonspecific lipid transfer proteins in germinating castor bean seeds. Plant Physiol.
90:841-845.

Tsuboi, S., Suga, T., Taklshlma, K., Mamlya, G., Matsul, K., Ozekl, Y. and Yamada,
M. 1991. Organ-specific occurrence and expression of the isoforms of nonspecific
lipid transfer protein in castor bean seedlings, and molecular cloning of a full-
length cDNA for a cotyledon-specific isoform. J. Biochem. 110:823-831.

Tsuboi, 8., Osafune, T., Tsugeld, R., Nishlmura, M. and Yamada, M. 1992.
Nonspecific lipid transfer protein in castor bean cotyledon cells: subcellular
localization and a possible role in lipid metabolism. J. Biochem. 111:500-508.

Vergnolle, C., Arondel, V., Tchang, F., Grosbois, M., Guerbette, F., Jolllot. A. and
Kader, J.-C. 1988. Synthesis of phospholipid transfer proteins from maize
seedlings. Biochem. Biophys. Res. Commun. 157:37-41.

van Heine, G. 1985. A new method for predicting signal sequence cleavage sites.
Nucleic Acid Res. 14:4683-4690.

Yamada, M., Tsuboi, S., Osatune, T., Suga, T. and Taklshlma, K. 1990.
Multifunctional properties of non-specific lipid transfer protein from higher plants.
In: Biological role of plant lipids. P.A. Biacs, K Gruiz, and T. Kremmer, eds.
Akadémiai I'Gada, Budapest, pp. 278-280.

I‘IICHIGRN STRTE UN

l llLlllllllllllll llllllll lllgllllllllllllllllll