In vitro analyses of protein transport into thylakoids

MISSING IMAGE

Material Information

Title:
In vitro analyses of protein transport into thylakoids a subset of chloroplast proteins are transported into thylakoids by a chloroplast seca-dependent pathway
Physical Description:
vii, 108 leaves : ill., photos ; 29 cm.
Language:
English
Creator:
Yuan, Jianguo, 1962-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 94-107).
Statement of Responsibility:
by Jianguo Yuan.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002028119
notis - AKL5730
oclc - 33032688
System ID:
AA00003646:00001

Full Text








IN VITRO ANALYSES OF PROTEIN TRANSPORT INTO THYLAKOIDS:
A SUBSET OF CHLOROPLAST PROTEINS ARE TRANSPORTED INTO
THYLAKOIDS BY A CHLOROPLAST SECA-DEPENDENT PATHWAY












By

JIANGUO YUAN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1994











ACKNOWLEDGEMENTS


The author thanks the members of his committee Drs. Gloria Moore, Ken

Cline, Al Lewin, Don McCarty, and Eduardo Vallejos for their care and advice

during the work and preparation of this dissertation. The author is especially

indebted to his adviser Dr. Ken Cline for financial and psychological support

throughout the entire studies; without such support the author's life in Gainesville

would have been miserable and this dissertation would have not been possible. The

author also thanks Mike McCaffery and Changjiang Li for their excellent technical

assistance, and Ralph Henry for many fruitful discussions. Thanks are also extended

to others who have associated with the author. This dissertation is dedicated to the

author's family for their understanding and tolerance of long-time absence of care.














TABLE OF CONTENTS


ACKNOWLEDGMENTS ....................


KEY TO ABBREVIATIONS ............

ABSTRACT .........................


.............. .......... v

. . . vi


CHAPTERS


1 INTRODUCTION ........................
Goals of the Research ......................
Structure and Function of Chloroplasts ..........
Biogenesis of Chloroplasts ...................
Protein Import into Chloroplasts ..............
Protein transport into Thylakoids ..............
Protein Export from Bacteria as a Model
for Studies of Protein Transport into Thylakoids .
Is Protein Transport into Thylakoids Homologous to
Protein Export from Bacteria? ................


. . 1
............... 2
. . 3
............... 4
............... 6
.............. 11

. . 15

.............. 18


2 CRYOPRESERVATION OF CHLOROPLASTS AND THYLAKOIDS
FOR STUDIES OF PROTEIN IMPORT AND INTEGRATION ....
Introduction ...........................................
M materials and M ethods ...................................
Results and Discussion ...................................
Conclusions ...........................................

3 STROMAL FACTOR PLAYS AN ESSENTIAL ROLE IN
PROTEIN INTEGRATION INTO THYLAKOIDS
THAT CANNOT BE REPLACED BY UNFOLDING
OR BY HEAT SHOCK PROTEIN HSP70 .....................
Introduction ...........................................
M materials and M ethods ...................................
R results ...............................................
D discussion ............................................






4 PLASTOCYANIN AND THE 33K OXYGEN-EVOLVING
PROTEIN ARE TRANSPORTED INTO THYLAKOIDS
WITH SIMILAR REQUIREMENTS AS PREDICTED
FROM PATHWAY SPECIFICITY .......................... 57
Introduction ........................................... 57
M materials and M ethods ................................... 58
R results ............................................... 60
D discussion ............................................ 73

5 ONE OF THREE PATHWAYS FOR PROTEIN TRANSPORT
INTO PEA THYLAKOIDS USES A SECA-DEPENDENT
TRANSLOCATION MECHANISM ......................... 77
Introduction ........................................... 77
Materials and Methods ................................... 79
Results and Discussion ................................... 83
Conclusions ........................................... 91

6 SUM M ARY ........................................... 92

REFERENCES ............................................. 94

BIOGRAPHICAL SKETCH ................................... 108












KEY TO ABBREVIATIONS


LHCP
PC
PCara
PCpea
OE33
OE23
OE17
Rubisco
SS
pLHCP
pPC
pOE33
pOE23
pOE17
pSS
(p)LHCP
LHCP-DP
mPC
mOE33
mOE23
mOE17
SE
BSA
Hsp70
DTT
PMF
Alp
E
CF1/CFo


Light-harvesting chlorophyll a/b protein
plastocyanin
plastocyanin from Arabidopsis thalliana
plastocyanin from pea
the 33-kDa subunit of the oxygen-evolving complex
the 23-kDa subunit of the oxygen-evolving complex
the 17-kDa subunit of the oxygen-evolving complex
ribulose- 1,5-bisphosphate carboxylase/oxygenase
small subunit of Rubisco
precursor form of LHCP
precursor form of PC
precursor form of OE33
precursor form of OE23
precursor form of OE17
precursor form of SS
pLHCP or LHCP
a characteristic protease degradation product of LHCP
mature form of PC
mature form of OE33
mature form of OE23
mature form of OE17
stromal extract
bovine serum albumin
70-kDa heat shock protein
dithiothreitol
proton motive force
transmembrane electric potential
einstein
coupling factor











Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IN VITRO ANALYSES OF PROTEIN TRANSPORT INTO THYLAKOIDS:
A SUBSET OF CHLOROPLAST PROTEINS ARE TRANSPORTED INTO
THYLAKOIDS BY A CHLOROPLAST SECA-DEPENDENT PATHWAY

By

JIANGUO YUAN

August, 1994

Chair: Gloria A. Moore
Cochair: Kenneth C. Cline
Major Program: Plant Molecular and Cellular Biology


Thylakoids are membranes within chloroplasts that carry out the light-driven

reactions of photosynthesis. Biogenesis of thylakoids requires the cooperative

expression of both the chloroplast genome and the nuclear genome. Nuclear

encoded thylakoid proteins are synthesized in the cytosol, imported into chloroplasts,

and then transported into thylakoids. To understand the mechanisms for protein

targeting to the thylakoids, in vitro assays with isolated chloroplasts and isolated

thylakoids were developed and used for the studies of thylakoid protein transport.

In the first part of this dissertation, a method for preservation of isolated

chloroplasts and isolated thylakoids is presented to facilitate investigations of

thylakoid protein biogenesis.






In the second part of this dissertation, a purified thylakoid membrane

precursor protein was used in in vitro assays to investigate the importance of soluble

factors in protein import into chloroplasts and integration into thylakoids. The

results show that cytosolic factors are not required for import into the chloroplast

stroma, but integration into the thylakoid membrane absolutely depends on a stromal

protein factor that appears to play an active role in the integration process. Purified

thylakoid precursor proteins were also used in substrate competition studies that

revealed three distinct targeting pathways for transport of proteins into or across the

thylakoid membrane.

In the third part of this dissertation, in vitro assays with isolated thylakoids

were developed for detailed studies on the requirements for transport of proteins on

one of the pathways. The results show that similar requirements are used for

transport of proteins on the same translocation pathway. Further studies on the

properties of this thylakoid translocation pathway show that it exhibits characteristics

of SecA-dependent protein export from bacteria, suggesting a SecA-homologous

pathway for protein transport into plant thylakoids. With an antibody to a conserved

peptide from an algal SecA, a SecA homologue of pea chloroplasts (CPSecA) was

identified and purified from stromal extract with a combination of conventional

purification methods. Purified CPSecA was shown to be able to support protein

transport into isolated thylakoids. Interestingly, only one of the three pathways for

protein transport into thylakoids uses the chloroplast SecA-dependent translocation

mechanism.












CHAPTER 1
INTRODUCTION



Chloroplasts make this planet green. They are the chlorophyll-containing

organelles in plant cells that are responsible for the conversion of light energy into

chemical energy in photosynthesis. Besides being the organelle of photosynthesis,

chloroplasts are also the site of many other biosynthetic functions such as synthesis

of amino acids, lipids, and pigments (Kirk and Tilney-Bassett, 1978). All these

functions are catalyzed by enzymes made up of proteins. Although chloroplasts have

their own machinery for making proteins, less than 15% of their proteins are made

in the chloroplast. The majority of chloroplast proteins are encoded in the nucleus,

synthesized in the cytosol, and posttranslationally imported into chloroplasts. The

mechanisms that govern protein targeting and transport are thus of exceptional

importance to the biogenesis of chloroplasts. However, despite intensive studies over

the last two decades, still very little is known about the mechanisms by which

nuclear-encoded chloroplast proteins are specifically targeted to the six chloroplast

subcompartments [see de Boer and Weisbeek (1991) and Theg and Scott (1993) for

a recent reviews].








Goals of the Research

The overall goal of our research presented here is to understand the strategies

and mechanisms by which newly made chloroplast proteins are specifically localized

to the various subcompartments. As an experimental system we have chosen as a

model the targeting and transport of cytosolically-synthesized chloroplast proteins to

the thylakoids because they have the most complex pathway for localization. They

have to cross up to three membranes and three aqueous compartments en route from

their site of synthesis in the cytosol to their functional locations in the thylakoid

membrane or the lumen. Biochemical studies of this process have made use of two

in vitro assays for the assessment of import into chloroplasts and transport into

thylakoids. The first is called the import assay that utilizes isolated intact

chloroplasts and radiolabeled precursors. The second is called the integration or

transport assay that is similar to the import assay with the exception that it employs

lysed chloroplasts and isolated thylakoids supplemented with stromal extract. To

facilitate such studies, the first specific goal of this dissertation is to develop a

method for preservation of active chloroplasts and thylakoids for studies of protein

import and integration. A second specific goal of this dissertation is to produce

chemical quantities of pure precursor proteins that can be used to address questions

about the importance of soluble factors in chloroplast protein transport and to carry

out substrate competition studies that could reveal pathway specificity for import into

chloroplasts or transport into thylakoids. The third specific goal of this dissertation

is to purify a stromal component required for transport of a subset of proteins into






3

the thylakoid lumen. The results from these studies are presented in four chapters

(Chapters 2-5). The purpose of this chapter is to give some background information

and to outline the questions related to the studies described in this dissertation. The

results and significance are summarized and discussed in the last chapter.



Structure and Function of Chloroplasts

Chloroplasts are the photosynthetic member of a family of organelles called

plastids that distinguish plants from other biological kingdoms. Chloroplasts are

found in leaves and other green tissues of plants. Chloroplasts are structurally

complex organelles. They are enclosed by a double membrane envelope and are

filled with an internal membrane system called the thylakoids. The envelope

membranes and the thylakoid membrane divide chloroplasts into several

subcompartments. The intermembrane space is the aqueous compartment between

the two envelope membranes. The thylakoid lumen is the aqueous compartment

enclosed by the thylakoid membrane. The stroma is the aqueous compartment

within the envelope and outside the thylakoids.

The chloroplast envelope separates the contents of the chloroplast from the

cytoplasm of the cell. The major role of the envelope is to serve as a selective

barrier and to allow regulated exchange of metabolites and ions between chloroplasts

and the cell cytosol (Douce and Joyard, 1991). The envelope also contains an import

machinery for transport of proteins into the chloroplast (Douce and Joyard, 1990).

In addition to its many transport functions, the envelope is also the site of many






4

biosynthetic reactions such as synthesis of lipids and pigments (Douce et al., 1984;

Douce and Joyard, 1991). The stroma is the site of dark reactions of photosynthesis,

the site of protein synthesis from chloroplast-encoded genes, and the site of many

other reactions. The thylakoids are the site of the light reactions of photosynthesis.

In the light reactions of photosynthesis, light energy is captured by the two

photosystems and used to drive electron transfer that results in the production of

NADPH and the build-up of a proton gradient across the thylakoid membrane. The

proton gradient is subsequently used by the thylakoidal ATP synthase to make ATP.

The thylakoid membrane contains the lipids, the pigments and most of the proteins

involved in the light reactions of photosynthesis. The only components of the light

reactions that do not permanently associate with the membrane are the electron

carriers plastocyanin and ferredoxin which, respectively, reside in the thylakoid lumen

and the chloroplast stroma (Bogorad, 1991).



Biogenesis of Chloroplasts

Different forms of plastids, each with distinct functions, are found in plants.

These different forms are all referred to as plastidss' because they are all, to some

extent, interconvertible (Kirk and Tilney-Bassett, 1978). The best-known plastid is

the chloroplast. Other plastids include proplastids, etioplasts, amyloplasts, and

chromoplasts. Proplastids are undifferentiated progenitor plastids present in

meristematic cells. All other plastids ultimately develop from proplastids by division

and differentiation (Mullet, 1988). Proplastids are small and contain little thylakoid






5

membrane system, but during development they increase in size, accumulate

numerous macromolecular components, and progressively deposit membrane

lamellae in their stroma. In leaves, the end result of proplastid development is

chloroplast with its highly intricate internal membrane system. The development of

chloroplasts from their tiny precursor proplastids involves synthesis of a great deal

of lipids, pigments, nucleic acids, and proteins.

Synthesis of the fatty acids occurs in the chloroplast stroma (Slabas and

Fawcett, 1992) along with the initial acylation of glycerol 3-phosphate (Alban et al.,

1988; Browse and Sommerville, 1991). The later steps of lipid synthesis are carried

out by enzymes in/on the envelope membranes (Joyard et al., 1991). The lipid of the

thylakoid membrane is thought to be transferred from the envelope as membrane

vesicles (Joyard et al., 1986). Because protein is the major component of

chloroplasts, chloroplast development involves a massive synthesis of proteins.

Chloroplasts contain DNA and have a machinery for protein synthesis. However, the

capacity of the chloroplast genome is too small to account for all the proteins present

in the chloroplast (Palmer, 1991). Indeed, the majority of the chloroplast proteins

are encoded in the nucleus, synthesized in the cytosol, and transported into

chloroplasts. Nuclear-encoded chloroplast proteins are first synthesized as larger

precursors with amino terminal extensions called transit peptides that are both

necessary and sufficient for directing import into the chloroplast (Berry-Lowe and

Schmidt, 1991). Some multimeric enzymes and protein complexes of chloroplasts are

composed of both chloroplast- and nuclear-encoded components. An interesting but






6

also very important question about chloroplast biogenesis is how the cells coordinate

the expression of chloroplast- and nuclear-encoded proteins destined for the same

functional complex and how proteins are specifically targeted to their correct location

in the chloroplast. The targeting and transport of proteins to the chloroplast and

their subsequent localization within the chloroplast are the focuses of the research

described in this dissertation.



Protein Import into Chloroplasts

A key development that allowed significant advances in our understanding of

protein import into chloroplasts was the reconstitution of protein transport with

isolated chloroplasts (Chua and Schmidt, 1978, 1979; Highfield and Ellis, 1978).

Studies using reconstituted system have shown that protein import into chloroplasts

is a multi-step process. The first step is the binding of precursor proteins to the

chloroplast surface (Cline et al., 1985; Su et al., 1992). The second step is

translocation of the bound precursors across the two envelope membranes into the

stroma (Schnell and Blobel, 1993). A third step, which occurs during or shortly after

translocation, is proteolytic processing of the precursor to remove the transit peptide

(Berry-Lowe and Schmidt, 1991). In vitro assays for import studies depend on the

availability of isolated intact chloroplasts. A method for long-term preservation of

intact active chloroplasts would be helpful for such studies.








Structure of Chloroplast Transit Peptides

Most chloroplast precursor proteins possess monopartite transit peptides for

targeting to the chloroplast stroma. These transit peptides from different precursor

proteins are functionally interchangeable (Berry-Lowe and Schmidt, 1991; de Boer

and Weisbeek, 1991). However, there is no apparent sequence homology between

transit peptides of different precursor proteins. Transit peptides are characterized

by large numbers of serine and threonine and the lack of tyrosine and negatively

charged amino acids (Berry-Lowe and Schmidt, 1991). Transit peptides have been

predicted to form random coils (von Heijne and Nishikawa, 1991). Theg and Geske

(1992) have investigated the biophysical properties of the transit peptide to the

gamma subunit of the chloroplast ATP synthase from Chlamydomonas reinhardtii.

They used circular dichroism measurements to show that the gamma subunit transit

peptide is largely a random coil, both in the presence and in the absence of sodium

laurylsulfate micelles. However, the secondary structure of the ferredoxin transit

peptide was reported to be modulated by its interaction with negatively charged

lipids (Horniak et al., 1993). In buffer, the transit peptide of ferredoxin was shown

to be in a random coil conformation. In the presence of micelles formed by anionic

lipids, a large increase in a-helix was observed for the ferredoxin transit peptide.

Transit peptides are removed in the stroma by a soluble processing protease (de

Boer and Weisbeek, 1991; Berry-Lowe and Schmidt, 1991).









Energetics of Protein Import into Chloroplasts

Import of proteins into chloroplasts is an energy-dependent process

(Grossman et al., 1980; Pain and Blobel, 1987). Unlike the import of proteins into

mitochondria which requires the presence of both ATP and an electric field across

the inner mitochondrial membrane (Chen and Douglas, 1987; Eiler et al., 1987),

import of proteins into chloroplasts needs only ATP (Theg et al., 1989). ATP is

required both for efficient binding of precursor proteins to the chloroplast envelope

and for translocation of bound precursors across the two envelope membranes into

the stroma. Binding requires only micromolar concentrations of ATP (Olsen et al.,

1989). The site of ATP utilization for binding is the intermembrane space between

the two envelope membranes (Olsen and Keegstra, 1992). The ATP requirement for

protein transport across the envelope is about tenfold higher than that for precursor

binding to the chloroplast surface (Olsen et al., 1989; Theg et al., 1989). The site of

ATP hydrolysis for import is in the stroma (Pain and Blobel, 1987; Theg et al., 1989).



Cytosolic Factors for Protein Import into Chloroplasts

Over the past several years, a number of studies have used purified precursor

proteins to investigate the requirement for cytosolic factors for protein import into

chloroplasts. Chemical quantities of purified precursor proteins were obtained by

overexpression in E. coli and then purification from cell extract by conventional

chromatography or by isolation of inclusion bodies. Purified precursor proteins were

dissolved in 8 M urea and assayed for import in the presence or absence of cytosolic






9

proteins immediately after dilution from urea. When the precursor protein to the

light-harvesting chlorophyll a/b protein (pLHCP) was used in such studies, no import

was obtained after urea-denatured pLHCP was rapidly diluted into the import

mixture (Waegemann et al., 1990). Import of pLHCP into isolated chloroplasts was

only observed after the denatured pLHCP was dialyzed in the presence of leaf

extract. The effect of the leaf extract on the import of pLHCP into chloroplasts was

only partly compensated by Hsp70. Based on these experiments, the authors

concluded that import of proteins into chloroplasts requires at least two cytosolic

factors. On the other hand, efficient import was obtained for purified precursors of

plastocyanin (de Boer et al., 1991) and ferredoxin (Pilon et al., 1990) in the absence

of cytosolic factors, indicating that cytosolic factors are not necessary for protein

import into chloroplasts. Apparently, there is a difference between the import

requirement for pLHCP and that for pre-ferredoxin and pre-plastocyanin. Does this

difference in import requirements reflect the difference between precursor proteins

or between the apparatus for import of pLHCP and that for import of pre-ferredoxin

or pre-plastocyanin? If the latter were true, it would mean that different

mechanisms were used for the import of specific proteins into the chloroplast. As

will be seen in chapter 3, our studies challenge the conclusions of Waegemann et al.

(1990).









Membrane Proteins Involved in Chloroplast Protein Import

Before import into chloroplasts, precursors must bind to the envelope of

chloroplast. Binding has been shown to be specific and saturable (Cline et al., 1985;

Friedman and Keegstra, 1989). Binding was greatly reduced when chloroplasts were

treated with protease (Cline et al., 1985). These results suggest that binding of

precursor proteins to the chloroplast envelope is mediated by proteinaceous

receptors, consistent with the fact that high affinity binding requires low levels of

ATP. In an effort to identify the import receptor(s), Cornwell and Keegstra (1987)

used a noncleavable heterobifunctionalcrosslinker and crosslinked a putative 66-kDa

receptor protein to a radiolabeled small subunit of ribulose-1,5-bisphosphate

carboxylase precursor (pSS). However, they were unable to identify a 66-kDa

protein on stained gels of purified envelope proteins.

Kaderbhai et al. (1988) performed cross-linking to isolated inner and outer

envelope membranes using a synthetic peptide corresponding to the transit peptide

of pSS and identified a 30-kDa and a 52-kDa protein. Pain et al. (1988) used

antiidiotypic antibodies and also identified a 30-kDa and a 52-kDa protein. The 52-

kDa protein was shown to be the large subunit of ribulose bisphosphate carboxylase.

The 30-kDa protein was found by sequencing of a cDNA clone to be the same

protein identified previously as the phosphate translocator (Schnell et al., 1990;

Flugge et al., 1991; Wiley et al., 1991).

A 51-kDa protein of possible involvement in chloroplast protein import was

identified by yet another approach that used 32P-ATP to examine differential







11

phosphorylation of envelope proteins upon interaction with precursor proteins (Hinz

and Flugge, 1988). More recently, an 86-kDa protein, a 34-kDa protein, and a 70-

kDa heat shock cognate protein were identified and shown to be in a complex that

could interact with precursor proteins in an ATP-dependent manner (Waegemann

and Soll, 1991; Soll and Waegemann, 1992; Soll and Alesen, 1993).

Most recently, Perry and Keegstra (1994) performed label transfer crosslinking

experiments and labeled two proteins of 75-kDa and 86-kDa of potential involvement

in chloroplast protein import. Labeling of both proteins required pSS with its transit

peptide. Labeling of the 75-kDa protein occurred only when low levels of ATP were

present. Both proteins were identified as proteins of the outer envelope membrane.

However, the labeled form of the 75-kDa protein could only be detected in fractions

containing mixed envelope membranes. These results suggest that pSS first binds to

the 86-kDa protein and that the ATP-requiring step for binding is associated with the

75-kDa protein.



Protein Transport into Thylakoids

Nuclear-encoded proteins destined for the internal membrane system of

chloroplasts must be further transported into or across the thylakoid membrane in

order to reach their functional locations. Studies with a number of different

thylakoid proteins (Smeekens et al., 1986; Ko and Cashmore, 1989; Cline et al., 1993;

Howe and Merchant, 1993) have indicated that thylakoid precursor proteins are first

translocated across both envelope membranes into the chloroplast stroma, and are







12

then transported into the thylakoid membrane or across the membrane into the

lumen.



Signals for Localization to the Thylakoids

In addition to the signals for envelope transfer, precursors of thylakoid

lumenal proteins also contain a sequence for thylakoid transfer on their amino

terminal extensions. The amino terminal extensions of lumenal proteins are thus

bipartite in structure with a stroma-targeting transit peptide at the N-terminus

proximal and a lumen-targeting sequence for thylakoid transfer at the C-terminus

proximal (de Boer and Weisbeek, 1991). Lumen-targeting sequences are structurally

similar to signal peptides of bacterial and endoplasmic reticulum (ER) secretary

proteins. They are all characterized by having a hydrophobic core region of ~ 15

residues, a positively charged N-flanking region and an A-X-A cleavage site (von

Heijne et al., 1989). Lumen-targeting sequences are removed in the lumen by a

thylakoidal peptidase (Halpin et al., 1989).

There is no cleavable signal sequence for the integration of proteins into the

thylakoid membrane. Integration of membrane proteins such as the light-harvesting

chlorophyll a/b-binding protein of photosystem 2 (LHCP) is directed by targeting

signals within the mature protein sequence (Lamppa, 1988; Viitanen et al, 1988;

Hand et al., 1989). The exact location of these signals in LHCP has not been

determined, but appears to lie within one or more of the three hydrophobic

transmembrane segments (Auchincloss et aL, 1992; Huang et al., 1992).








Pathways for Protein Transport into Thylakoids

Several lines of evidence suggest that a common translocation mechanism is

used for the transport of proteins across the envelope membranes. The most

convincing evidence comes from competition studies. Purified precursor proteins and

synthetic transit peptides compete for the import of other precursor proteins (Perry

et al., 1991; Schnell et al., 1991; Oblong and Lamppa, 1992; Cline et al., 1993).

Import of both stromal and thylakoid precursor proteins was severely inhibited by

import saturating concentrations of pOE23 (Cline et al., 1993).

In contrast to protein transport across the envelope into the stroma, transport

of proteins into thylakoids appears to be mediated by several distinct translocation

mechanisms. Evidence for the operation of multiple pathways for protein transport

into thylakoids comes from recent substrate competition studies that revealed three

distinct precursor specificity groups (Cline et al., 1993). Lumen-resident proteins

OE23 and OE17 constitute one group, lumenal proteins OE33 and PC a second, and

the membrane protein LHCP a third. Two additional proteins PS1-N and PS2-T

were subsequently shown to be on the OE23 and OE17 pathway (Henry et al., 1994;

Nielsen et al., 1994). The specificity group determined from competition studies

correlate exactly with the precursor protein-specific requirements for transport of

proteins into thylakoids (Cline et al., 1992a). However, it remains to be seen

whether the various pathways represent distinct translocation systems.








Energetics of Protein Transport into Thylakoids

In contrast to the singular ATP requirement for import of proteins into

chloroplasts, transport of proteins into thylakoids requires both nucleotide

triphosphates and a proton motive force (PMF) (Mould and Robinson, 1991; Cline

et al., 1989, 1992; Reed et al., 1990). The ApH of the PMF is the active component

for thylakoid protein transport. The energy requirements for thylakoid protein

transport are further complicated by the fact that the energy requirements for

thylakoid transport are precursor protein-specific (Cline et al., 1992). Transport of

the 33-kDa subunit of the oxygen-evolving complex (OE33) and integration of LHCP

requires ATP and is stimulated by the trans-thylakoidal proton gradient (Kirwin et

al., 1989; Bauerle and Keegstra, 1991; Mould and Robinson, 1991; Cline et al., 1992).

Transport of plastocyanin (PC) requires only ATP (Theg et al., 1989; Cline et al.,

1992). Transport of OE23 and OE17 (the 23-kDa and 17-kDa subunit of the oxygen-

evolving complex) require only a proton gradient (Cline et al., 1992). Why are there

different energy requirements for protein transport into thylakoids? Does this reflect

different mechanisms for thylakoid transport?



Soluble Factors Required for Thylakoid Transport

The stromal factor required for integration of LHCP into the thylakoid

membrane is the first soluble factor ever reported for chloroplast protein transport

(Fulsom and Cline, 1988). This factor was originally discovered during in vitro

integration studies with separated chloroplast subfractions (Cline, 1986). It was







15

found that membrane integration of pLHCP occurred only in the presence of both

thylakoids and stromal extract. The stromal factor has been shown to be a protein

factor with an active sulfhydryl group (Fulsom and Cline, 1988). Payan and Cline

(1991) have shown that one function of the stromal factor is to maintain the

solubility and insertion competence of LHCP by converting LHCP into a large and

soluble species. Efforts to identify and isolate this factor have already been

undertaken in our laboratory. Purification was performed with conventional protein

fractionation methods based on integration activity. Unfortunately, integration

activity was lost when the resolution of separation methods was increased. Stromal

factor was also reported for efficient transport of OE33 into the thylakoid lumen

(Mould et al., 1991). It was suggested that the stromal processing activity was the

stromal stimulating activity for OE33 transport (Kirwin et al., 1989; Mould et al.

1991). However, their experiments could not rule out the possibility that a stromal

factor other than the processing protease is the stimulating activity for OE33

transport.



Protein Export from Bacteria as a Model
for Studies of Protein Transport into Thylakoids

The evolutionary ancestor of chloroplasts is thought to be a cyanobacterium

that formed an endosymbiotic existence with the host primordia eukaryote. Similarly,

the evolutionary ancestor of the mitochondrion is thought to be an endosymbiontic

purple bacterium. In this context, there is good reason to believe that the protein

sorting and/or transport system that occurs within chloroplasts and mitochondria is







16

derived from the machinery present in the endosymbiotic prokaryotes at the time of

endosymbiosis. This concept has been referred to as "Conservative Sorting" in the

mitochondrial literature (Hartl and Neupert, 1990). The "Conservative Sorting"

hypothesis suggests that intraorganellar protein transport systems resemble those of

modern-day prokaryotes. The mechanisms for thylakoid protein transport might also

be very similar to the transport of proteins across the bacterial membrane (Smeekens

et al., 1990). Indeed, there is much similarity between thylakoid protein transport

and protein export in bacteria. First, signal sequences for thylakoid transport are

like bacterial signal peptides. Second, the thylakoid transfer sequences have been

shown to function in E. coli as signal peptides and are able to export the passenger

protein from the cytoplasm. Third, the thylakoidal peptidase that removes the

thylakoid transfer domain has identical substrate specificity as that of the bacterial

signal peptidase that removes signal peptides from bacterial secretary proteins (de

Boer and Weisbeek, 1991; Berry-Lowe and Schmidt, 1991; and references within).



The General Secretory Pathway in Bacteria

Currently, the best understood system for protein export from bacteria is that

of E. coli. In E. coli, signal sequence-bearing proteins are exported by a

translocation system that depends on SecA and SecY/E proteins and is powered by

the combined action of ATP hydrolysis and a proton motive force (Pugsley, 1993).

Some preproteins also require SecB for efficient export (Kumamoto and Bechwith,

1983, 1985). Two additional proteins SecD and SecF are essential for protein export






17
in vivo (Gardel et al., 1990), but have no effect in vitro (Matsuyama et al., 1993).

SecA and SecB are found in the cytosol, although SecA is also found associated with

the cytoplasmic membrane (Oliver, 1993). SecY, SecE, SecD and SecF are

membrane proteins. SecY seems to span the membrane 10 times with its amino and

carboxyl termini in the cytosol (Ito, 1992). SecE from E. coli is thought to span the

membrane three times, although the Bacillus subtilis SecE only spans the membrane

one time (Schatz et al., 1989; Ito, 1992). SecD and SecF both appear to span the

membrane six times (Gardel et al., 1990).

SecB is a chaperone protein that binds to precursor proteins and maintains

them in a export competent conformation. SecB also helps targeting precursor

proteins to the translocase (an enzyme made up of SecA and SecY/E proteins) via

its affinity for SecA (Hartl et al., 1990). SecA is the peripheral domain of the

protein translocase. SecA hydrolyzes ATP and initiates translocation by getting the

first 20 amino acid residues across the E. coli membrane (Schiebel et al., 1991). In

the absence of a PMF, SecA and ATP can complete translocation by promoting

transport of ~ 20 amino acid residues each cycle across the membrane (Schiebel et

al., 1991). Azide inhibits SecA-dependent transport (Oliver et al., 1990) and removal

of membrane-associated SecA by urea extraction prevents transport (Cunningham

et al., 1989). Addition of either purified SecA or SecA-containing fraction restores

transport activity to urea-washed membranes. SecY and SecE may form a pore for

the actual passage of preproteins across the membrane (Joly and Wickner, 1993).







18

The functions of SecD and SecF are not yet clear, but recent studies suggest that

they may be involved in the PMF-driven transport (Arkowitz and Wickner, 1994).



Is Protein Transport into Thylakoids Homologous to Protein Export from Bacteria?

The endosymbiotic origin of chloroplasts from an ancestor cyanobacterium,

the similarity between the thylakoid targeting sequences and the bacterial signal

sequences, the identical cleavage specificity of the thylakoidal and bacterial

peptidases, the discovery of secA homologous genes in the chloroplast genomes of

several algae all point to the existence of a SecA type of protein translocation system

in plant chloroplasts. However, direct evidence for the operation of a SecA type of

translocation mechanism for protein transport into plant thylakoids must come from

the identification of a chloroplast homologue of SecA and the demonstration of its

involvement in protein transport into thylakoids.












CHAPTER 2
CRYOPRESERVATION OF CHLOROPLASTS AND THYLAKOIDS
FOR STUDIES OF PROTEIN IMPORT AND INTEGRATION



Introduction

Recent advances in our understanding of the mechanisms of chloroplast

protein biogenesis derive primarily from the availability of in vitro biochemical assays

for chloroplast protein import and in organello protein synthesis. Active chloroplasts

are prepared for these assays by homogenization of fresh tissue followed by a

combination of differential and density gradient centrifugation (Walker et al., 1987).

Chloroplasts are then incubated under conditions that either promote synthesis of

proteins within the organelle (Bhaya and Jagendorf, 1984; Mullet et al., 1986) or

allow the import of cytoplasmically synthesized plastid proteins into the organelle

(Mishkind et al., 1987). Isolation of chloroplasts from amenable plant species, such

as spinach and peas, is not difficult but requires about two intensive hours from start

to finish. This preparation time has become a limiting factor as investigations of

transport mechanisms have become more sophisticated and the duration of

experiments has increased. A ready source of preserved but active chloroplasts

would greatly facilitate such studies. Furthermore, the ability to preserve active

chloroplasts for extended periods will obviate the problem of seasonal variation in







20
import competence of chloroplasts and permit investigators to comparatively analyze

plastids isolated from different plant tissue at different times.

Recent studies of plastid protein biogenesis have also focused on stromal

factors necessary for assembly of thylakoid proteins (Chitnis et al., 1987; Cline, 1986;

Fulsom and Cline, 1988). Purification schemes for the stromal factors require

reconstitution with active thylakoids for activity assays. Such assays are frequently

conducted one or several days after the original isolation of stroma and require the

isolation of fresh chloroplasts for thylakoid preparation. Accordingly, it would be

very helpful if thylakoids from the original chloroplast preparation could be frozen

for later use.

Methods for preservation of photosynthetically competent thylakoids have

been available for several years (Farkas and Malkin, 1979; Santarius, 1986; 1990), but

until now, methods for preserving intact chloroplasts have not been described. The

present communication describes protocols for the preservation of intact chloroplasts

and isolated thylakoids that are, respectively, active for subsequent studies of protein

import and integration. Approximately 65-70% of the chloroplasts stored in liquid

nitrogen in the presence of dimethyl sulfoxide remained intact upon thawing and

were fully functional for the import of precursor proteins. Preserved thylakoids were

nearly as active for protein integration studies as freshly prepared thylakoids. The

ability to store chloroplasts and subfractions for extended periods will facilitate

investigations of plastid protein biogenesis.








Materials and Methods

Materials

Tritium-labeled leucine was purchased from New England Nuclear. RNasin

and SP6 polymerase were from Promega Biotech. Miracloth was from Behring

Diagnostics. Mg-ATP, thermolysin, and Percoll were from Sigma. DMSO was from

Aldrich Chemical Company. Ethylene glycol and glycerol were from Fisher

Chemical Company. All other chemicals were reagent grade. The in vitro expression

plasmid for pLHCP, psAB80XD/4, is an SP6 derivative of psAB80 (Cashmore, 1984)

and has been described elsewhere (Cline et al., 1989). The expression plasmid for

LHCP, P2HPLC, is a pUC18 plasmid that harbors the coding sequence for the

mature form of LHCP from petunia (Viitanen et al., 1988), and was generously

provided by Dr. Paul Viitanen. The expression plasmid for the precursor to the

Rubisco small subunit from pea, pSMS64 (Anderson and Smith, 1986), was the

generous gift of Dr. Steven Smith. The expression plasmid for the precursor to

plastocyanin, pSPPC74 (Smeekens et al., 1985), was kindly provided by Drs. Thomas

Lubben and Kenneth Keegstra.



Preparation of Chloroplasts, Lysates, Thylakoids, and Stromal Extract

Intact chloroplasts were isolated from 10- to 12-day-old pea (Laxton's Progress

9) seedlings as described (Cline, 1986). Chloroplast lysates were obtained from

intact chloroplasts by resuspending chloroplast pellets in 10 mM Hepes/KOH (pH

8), 10 mM MgC12, and after 5 min adjusting to import buffer, 10 mM MgCI2.







22

Thylakoids were prepared from chloroplast lysates by centrifugation at 3200 x g for

8 min at 4 C. Stromal extract was prepared from the resulting supernatant by

further centrifugation at 42,000 x g for 30 min at 4C. Lysates prepared at a

concentration of 0.5 mg Chlorophyll/mL are arbitrarily referred to as 1X lysate and

the stromal extract resulting from such lysates as 1X stroma. Chlorophyll (Chl)

concentrations were determined according to Arnon (1949). For cryopreservation,

intact chloroplasts were resuspended in import buffer containing varying amounts of

cryoprotectant (0, 10, 20, or 30% [v/v] DMSO, ethylene glycol, glycerol; or 0.1 M, 0.2

M, or 0.3 M glycine) at final concentrations of 3.0 5.0 mg Chl/mL. Thylakoids to

be cryopreserved were resuspended in storage buffer (20 mM Hepes/KOH, pH 8, 20

mM sorbitol, and 125 mM KC1) containing either 30% [v/v] of DMSO or 30% [v/v]

of ethylene glycol to 3.0 5.0 mg Chl/mL. Aliquots (1.0 to 2.0 mL) of chloroplasts

or thylakoids were placed either in microcentrifuge tubes or thin walled screw-cap

plastic vials and plunged into liquid N2. Cryopreserved chloroplasts were thawed at

room temperature, diluted with import buffer to about 1.0 mg Chl/mL, and

repurified on 35% Percoll cushions (Cline, 1986). The percentage of intact

chloroplasts was determined from the Chl content of chloroplasts that sedimented

through the Percoll relative to the total amount of Chl applied to the cushion.

Chloroplasts that sedimented through Percoll were verified to be intact by the

ferricyanide reduction assay (Walker et al., 1987). Repurified chloroplasts were

washed twice with import buffer before use. Cryopreserved thylakoids were thawed

at room temperature, diluted to about 1.0 mg Chl/mL with storage buffer and







23

recovered by centrifugation at 3200 x g for 8 min. Thylakoid pellets were

subsequently washed twice with import buffer, 10 mM MgCl2 and resuspended in the

same buffer to 1.0 mg Chl/mL.



Microscopic Examination of Chloroplast Integrity

Small aliquots of chloroplasts in import buffer were fixed by adding an equal

volume of 3% glutaraldehyde in 50 mM potassium phosphate (pH 7.5) and 2 mM

MgCl2. After 45 min on ice, chloroplasts were pelleted and sequentially washed with

50 mM Hepes buffer (pH 8), containing decreasing sorbitol concentrations (0.33 M,

0.16 M, 0.00 M). Chloroplasts were post-fixed with 2% Os04 in phosphate buffer

(50 mM potassium phosphate, pH 7.5) at room temperature for 2 hours and

dehydrated in a graded ethanol series, from 12.5, 25, 40, 60, 75, 85, 95, to 100%,

followed by 3 cycles in 100% acetone. Dehydrated chloroplasts were embedded in

Spurr's resin (Spurr, 1969) in a graded series of 15%, 30%, 50%, 75%, 90%, and

100% plastic. Thin sections were prepared on an LKB ultramicrotome and were

post-stained with uranyl acetate for 20 min followed by lead citrate for 10 min.

Sections were examined with a Hitachi HU-11E electron microscope.



Preparation of Radiolabeled Precursors

RNA for pLHCP, pSS, and pPC was prepared by SP6 polymerase

transcription of EcoRI-linearized plasmids (Cline, 1988) and translated in the

presence of [3H] leucine in a wheat germ system (Cline et al., 1989). Translations







24
were diluted approximately six-fold and adjusted to import buffer containing 30 mM

leucine before use.



Assays for Import and Integration of Precursor Proteins

Import assays were carried out in vitro essentially as described (Cline, 1986).

Each assay contained 200 IpL of chloroplast preparation (0.5 mg Chl/mL), 50 IL of

adjusted translation product, and 50 pL of 60 mM Mg-ATP in import buffer or 50

pL of import buffer alone (light-driven assays). Assays were initiated with addition

of precursor and incubated at 25 C either in the presence of white light (~ 150

pE.m2s-') or in darkness. Assays were for 10 min with manual shaking at 5 min

intervals. After the reaction, chloroplasts were treated with thermolysin to remove

surface-bound precursors, repurified on Percoll cushions, and washed with import

buffer containing 5 mM EDTA as described (Cline et al., 1985).

Integration assays with chloroplast lysates were performed basically as

described (Cline, 1988). Each assay contained 200 pL of lysates (0.5 mg Chl/mL),

50 pL of 60 mM Mg-ATP in import buffer, 20 mM MgClz or 50 pL of import buffer,

20 mM MgCl2 alone (light-driven assays), and 50 pL of adjusted translation product.

When assays were conducted with reconstituted lysates, reaction mixtures contained

100 pL of thylakoids (1.0 mg Chl/mL) and 100 tL of 2X stromal extract in place of

the lysates. Reaction mixtures were incubated at 25C for 30 min either in darkness

or in white light ( ~ 150 pE.m2s1) with manual shaking every 5 min. Thylakoids were






25

recovered by centrifugation and were further treated with thermolysin to remove

non-integrated molecules as described (Cline, 1986).



Analysis of Precursors, Recovered Chloroplasts and Subfractions

Samples of precursors, recovered chloroplasts, and recovered chloroplast

subfractions were subjected to SDS-PAGE (Laemmli, 1970). For analysis of the

assembly of imported SS into Rubisco, stromal extracts of the recovered chloroplasts

were subjected to 6% nondenaturing PAGE (Payan and Cline, 1991). After

electrophoresis, gels were prepared for fluorography (Bonner and Laskey, 1974) and

placed on X-ray film. Radioactive proteins were extracted from the dried gels

(Walter et al., 1981) and quantified as described (Cline, 1986).



Results and Discussion

Preservation of Isolated Intact Chloroplasts

Various compounds have been reported to act as cryoprotectants and to

impart protective qualities to biological tissues during low temperature storage (Chen

and Li, 1989; Finkle et al., 1985). Among them, DMSO has long been known as one

of the most effective cryoprotectants for both plant and animal tissues (Finkle et al.,

1985); ethylene glycol, glycerol, and glycine have also been documented as good

cryoprotective agents in some cases (Farkas and Malkin, 1979; Finkle et al., 1985).

Here, these cryoprotectants at different concentrations were examined for their

ability to preserve intact pea chloroplasts. Immediately after isolation, chloroplasts








Table 2-1. Average percentage of intact chloroplasts preserved.

Protective Substances [v/v] Percentage

Freshly Prepared 85.0
No Cryoprotectants 00.0
10% DMSO 45.4
20% DMSO 67.6
30% DMSO 34.5
10% Ethylene Glycol 62.7
20% Ethylene Glycol 58.3
30% Ethylene Glycol 25.3
10% Glycerol 20.0
20% Glycerol 21.0
30% Glycerol 20.0

Note: Values in this table were derived from the average of three experiments
conducted under the same conditions. Frozen samples were stored overnight in
liquid nitrogen (-196 oC).


were resuspended in import buffer containing a cryoprotectant and placed in liquid

nitrogen for preservation. Samples were thawed at room temperature and the

percentage of intact chloroplasts was determined with a Percoll cushion assay

(Methods). The average percentage of intact chloroplasts preserved under different

conditions is shown in Table 2-1. Typically, the average percentage of intactness was

85% for freshly prepared chloroplasts. With preserved materials, best results were

found to be approximately 65% intact for samples stored in 10% ethylene glycol or

in 20% DMSO. For those stored in glycerol, less than 30% intact chloroplasts were

preserved at several concentrations, and nearly all chloroplasts were broken when

stored in glycine at 0.1 to 0.3 M. Because of the poor preservation of chloroplasts

stored in glycerol and glycine, only DMSO- (specifically, 20% [v/v]) and ethylene

glycol- (10% [v/v]) preserved chloroplasts were used in subsequent experiments.







































Figure 2-1. Ultrastructural features of fresh and cryopreserved pea
chloroplasts. (a) Freshly prepared chloroplasts; (b) chloroplasts
preserved in 20% (v/v) DMSO; (c) chloroplasts preserved in 10% (v/v)
ethylene glycol. Cryopreserved samples were stored in liquid N2 for
2 weeks. Bar, 2 pm.








Microscopic Examination of Chloroplast Integrity

Cryopreserved chloroplasts were examined by thin section transmission

electron microscopy. Low magnification micrographs gave a good impression of the

intactness of chloroplasts (Fig. 2-1). Examination of high magnification photographs

revealed no significant differences between freshly prepared chloroplasts and those

stored at low temperature in the presence of cryoprotectants (not shown). Outer

and inner envelope membranes were present and seemingly unchanged in

cryopreserved chloroplasts. Thylakoid structure appeared to be the

same in both newly prepared and preserved chloroplasts, with a nearly identical

shape and degree of stacking. Percoll cushion repurified samples contained only

intact chloroplasts (Fig. 2-1). Examination of unpurified samples revealed that

approximately 70% of preserved and 85% of freshly prepared chloroplasts were

intact (not shown). These results were fairly consistent with those obtained from

Percoll cushion assays (refer to Table 2-1).



Import of pLHCP by Preserved Intact Chloroplasts

Cryopreserved intact chloroplasts were assayed for their ability to import

proteins. Both DMSO- and ethylene glycol-preserved intact chloroplasts imported

pLHCP into an external protease resistant state (Fig. 2-2). Import could be driven

either by light or by exogenous ATP, indicating that preserved chloroplasts had

retained both a functional photosynthetic apparatus and a functional adenylate

translocator. Almost no pLHCP was imported when assays were carried out in












c 100
0
S80

0

S20-
0

0
C DMSO EG
120
imported b is 323 (light-driven assays) and 227 (ATP-TP
100 .
280
060




chloroplasts were incubated either with 20% [v/v] DMSO or 10%








ethylene glycol (EG). Cryoprese for 10 min. Incubated chloroplasts were thawed, repurified, and
washed twice with import buffer and assayed for their ability to import pLHCP (Methods). Each assay
pLHCP (Methods). Each assay received 202450 molecules of pLHCP per chloroplast. The control ()
thwas freshly prepared chloroplasts. e absolute number of molecules imported
imported per chloroplast is 32398 (light driven assays) and 22302 (ATP-driven assays)
driven assays) for the control. () Effects of preincuaion with
chCoroplasts were incubated either with 20% [vlv] DMSO or 10%
ethylene glycol (EG) for 10 mn. Incubated chloroplasts were then
washed twice with import buffer and assayed for their ability to import
pLHCP (Methods). Each assay received 2450 molecules of pLHCP

the 10 mrai incubation. The absolute number of molecules imported
per chloroplast is 398 (light driven assays) and 302 (ATP-driven assays)
for the control.







30
darkness in the absence of ATP (not shown). Furthermore, import ability was

virtually abolished by pretreatment of chloroplasts with thermolysin (not shown).

These latter observations indicate that import occurred by the physiological

mechanism. As import activity is concerned, cryopreserved chloroplasts were up to

85% as active in import assay as freshly prepared chloroplasts. Of the two kinds of

cryoprotectants used, DMSO preservation resulted in chloroplasts that were

moderately more active. DMSO and ethylene glycol on their own did not

significantly affect freshly prepared chloroplasts to import pLHCP.



Correct Assembly of Imported Proteins

Chloroplasts, cryopreserved for 6 months, were used to import several

different precursor proteins, including pLHCP, the precursor to the small subunit

(pSS) of Rubisco, and the precursor to plastocyanin (pPC). Following import,

chloroplasts were recovered and subfractionated into thylakoids and stroma to assess

the localization and assembly status of the imported proteins (Fig. 2-3). LHCP was

correctly processed to mature size, and properly assembled into the thylakoids as

determined by characteristic partial resistance to protease (Cline, 1986). Only a

trace amount of LHCP was detected in the stromal fraction. Similarly, mature-sized

plastocyanin was recovered with the thylakoid fraction in a protease-protected state.

The intermediate-sized plastocyanin precursor (trace amount), which has previously

been described (Smeekens et al., 1986), was present in the stromal fraction. Mature

-sized SS was recovered with the stromal fraction. Analysis of the stroma by 6%



















pLHCP pSS pPC
LY TH TT SE LY TH TT SE LY TH TT SE
pLHCP-
SLHCP-OP: &W4I N1
LHCP-DPP
Cr sr PC-
SS* CLL

pLHCP-
LHCP-: Z
SLHCP-DP* W pPC-
N pss- N iPc.
o ss -









Figure 2-3. Targeting and assembly of proteins imported into
cryopreserved chloroplasts. Chloroplasts were cryopreserved with 20%
DMSO for 6 months. Chloroplasts were thawed, repurified, and used
for import assays (600 jiL) with pLHCP, pSS, and pPC. Following
import, chloroplasts were treated with thermolysin to remove surface-
bound molecules, repurified on Percoll cushions (Methods), and lysed
with 75 IL of 10 mM Hepes/KOH (pH 8). An aliquot (25 il) of each
lysate (LY) was removed and the remainder centrifuged at 10,000 x g
for 10 min to separate stroma from thylakoids. Stromal extract (SE)
was obtained from the supernatant by further centrifugation
(Methods). The pelleted thylakoids were resuspended in 800 pLL of
import buffer. Aliquots (400 tiL) of thylakoids were treated with
thermolysin at 50 Ig/mL. Thermolysin-treated thylakoids (TT) as well
as untreated thylakoids (TH) were then resuspended in 25 pL of 20
mM EDTA. All samples were analyzed by SDS-PAGE/fluorography.
A photograph of the fluorogram is shown.







32
nondenaturing PAGE demonstrated that SS was assembled into the Rubisco

holoenzyme (not shown). These results are virtually identical to those obtained with

fresh chloroplasts and demonstrate that the assembly apparatus within chloroplasts

has remained functional in the cryopreserved chloroplasts.



Membrane Insertion of (p)LHCP by Chloroplast Lysates

Many of our studies focus on the transport of proteins into and across the

thylakoid membranes. These studies utilize an assay in which thylakoid protein

transport is reconstituted in chloroplast lysates. Thus, it was important to determine

if lysates from preserved chloroplasts were functional in these assays. Cryopreserved

intact chloroplasts were lysed and assayed for their ability to integrate (p)LHCP.

The results showed that lysates from both DMSO- and ethylene glycol-preserved

chloroplasts were capable of integrating (p)LHCP into their thylakoid membranes

(Fig. 2-4). Relative integration of (p)LHCP by lysates from preserved chloroplasts

was approximately 85% of that by lysates from freshly prepared chloroplasts for

ATP-driven assays, and approximately 70% for light-driven assays. As has previously

been shown for lysates from fresh chloroplasts (Cline, 1986), virtually no integration

occurred when assays were carried out in darkness in the absence of ATP or when

stroma was omitted from the assay mixture (not shown). DMSO preservation

resulted in lysates that were slightly more active in integration assays than ethylene

glycol preservation (Fig. 2-4).
















120
m Ught E ATP
-E 1 0 0 .................................. ..................



60 )


0
aL-

C DMSO EG




Figure 2-4. Relative integration of (p)LHCP by lysates from
cryopreserved chloroplasts. Chloroplasts were cryopreserved overnight
in the presence of either 20% [v/v] DMSO or 10% ethylene glycol
(EG). Chloroplasts were thawed, repurified, and lysed (Methods).
Lysates were assayed for the ability to integrate (p)LHCP into their
thylakoid membranes (Methods). Each assay received 5380 molecules
of pLHCP per chloroplast equivalent. Control experiments (C) were
conducted with lysates prepared from fresh chloroplasts. The absolute
number of molecules integrated per chloroplast equivalent is 249
(light-driven assays) and 296 (ATP-driven assays) for the control.








Integration of (p)LHCP by Cryopreserved Thylakoids

Previous studies have identified optimal conditions for the cryopreservation

of the photosynthetic competence of isolated thylakoids (Farkas and Malkin, 1979).

Similar conditions were examined for their ability to preserve the integration

competence of isolated pea thylakoids. Preserved thylakoids were supplemented with

fresh stroma and assayed for their ability to integrate (p)LHCP. The results showed

that both DMSO- and ethylene glycol-preserved thylakoids were able to integrate

(p)LHCP into their membranes (Fig. 2-5). Relative integration by cryopreserved

thylakoids was from 95 to 98% of that by freshly prepared thylakoids (ATP-driven

assays). For light-driven assays, ethylene glycol-preserved thylakoids were slightly

more active than DMSO-preserved thylakoids (Fig. 2-5), suggesting that ethylene

glycol may be marginally better than DMSO for stabilizing the photosynthetic

apparatus. This result is consistent with previous observation by Farkas and Malkin

(Farkas and Malkin, 1979).



Conclusions

In this study, we have described for the first time a simple and very

convenient protocol for cryopreservation of isolated intact chloroplasts. By following

this protocol, we were able to preserve chloroplasts fully functional for protein

import and assembly for six months (Fig. 2-3). We found that DMSO is the superior

preserving media for intact chloroplasts and recommend the following protocol.

Immediately after isolation of the chloroplasts, the final pellet should be resuspended















120
o0 N Ught B ATP
t100-

0 80E-

O 60 -
40-

20
20-

C N DMSO EG





Figure 2-5. Relative integration of (p)LHCP by cryopreserved
thylakoids. Isolated thylakoids were cryopreserved for four weeks in
storage buffer containing either 30% [v/v] DMSO or 30% ethylene
glycol (EG). Preserved thylakoids were thawed, washed, and
resuspended in import buffer plus 10 mM MgCl2 to a Chi
concentration of 1.0 mg/mL. Purified thylakoids were furnished with
fresh stroma in integration assays. Each assay received 4800 molecules
of pLHCP per chloroplast equivalent. The positive control (C) was
freshly prepared thylakoids. The negative control (N) was frozen
thylakoids with no protective substances. The absolute number of
molecules integrated per chloroplast equivalent is 226 (light-driven
assays) and 274 (ATP-driven assays) for the positive control.







36
in import buffer containing 20% DMSO (v/v) at a Chl concentration of 3 to 5

mg/mL. Chloroplasts should then be allowed to equilibrate for about 10 min on ice

before freezing. Freezing should be achieved rapidly by submerging small vials in

liquid N2 and thawing be allowed to occur at room temperature. For best results,

chloroplasts should be repurified on Percoll cushions before use.

Chloroplasts stored under the above conditions retained their ability to import

a variety of different precursor proteins and to correctly localize and assemble them

within the chloroplast. These latter processes frequently require transport into or

across the thylakoid membrane. The proper localization of plastocyanin and LHCP

by cryopreserved plastids, as well as the ability of lysates from these plastids to

assemble LHCP demonstrate that the thylakoid transport apparatus has been

preserved during low temperature storage. Preserved chloroplasts were also active

for in organello protein synthesis and produced a similar pattern of labeled

polypeptides although in a somewhat reduced amount (not shown). For preservation

of isolated thylakoids, a modified method of that reported by Farkas and Malkin

(1979) was an effective means of preserving thylakoids for use in protein integration

analysis.

We hope that this method will facilitate future studies of chloroplast protein

biogenesis. We recommend that investigators test the efficacy of these preservation

methods when different precursors or a different source of plastids is to be

employed.











CHAPTER 3
STROMAL FACTOR PLAYS AN ESSENTIAL ROLE IN PROTEIN
INTEGRATION INTO THYLAKOIDS THAT CANNOT BE REPLACED
BY UNFOLDING OR BY HEAT SHOCK PROTEIN HSP70



Introduction

Soluble factors necessary for efficient transport of proteins across or into

membranes have been discovered in nearly every translocation system studied in

depth. Many soluble factors such as the cytosolic Hsp70, the mitochondrial Hsp60,

and the bacterial SecB, DnaK and GroEL are molecular chaperones that function

to prevent nonproductive reactions of preproteins such as premature folding (Chirico

et al., 1988; Murakami et al., 1988; Bochkareva et al., 1988; Kumamoto, 1989;

Phillips and Silhavy, 1990; Koll et al., 1992), aggregation (Lecker et al., 1989), or

nonspecific membrane association (Hartl et al., 1990). Certain soluble factors, such

as SecB of bacteria and the signal recognition particle (SRP) of mammalian cells,

also provide a targeting function. SecB facilitates preprotein targeting to the

translocase via its affinity for SecA (Hartl et al., 1990). SRP binds to signal peptide

of nascent preproteins (Walter and Blobel, 1982) and targets them to the

endoplasmic reticulum (ER) by subsequent binding to the SRP receptor (Gilmore

et al., 1982). Some soluble factors, e.g. the bacterial SecA (Schiebel et al., 1991),

also participate in the mechanisms of membrane translocation.






38
The light-harvesting chlorophyll a/b protein (LHCP) is nuclear-encoded and

must cross the chloroplast envelope before being integrated into thylakoids.

Integration of LHCP into thylakoids has been reconstituted in vitro and shown to

require thylakoids, ATP, and an as yet unidentified stromal protein factor (Cline,

1986; Fulsom and Cline, 1988). Both the intact precursor (pLHCP) and the mature-

sized protein (LHCP) can serve as substrates for the reconstituted reaction (Cline,

1986; Viitanen et al., 1988). Payan and Cline (1991) showed that one function of the

stromal factor is to maintain the solubility and integration competence of (p)LHCP

(LHCP or pLHCP). The stromal factor accomplishes this by converting (p)LHCP

into a large and soluble species, most likely a complex between (p)LHCP and part

of the stromal factor. This putative complex has an estimated molecular weight of

about 120 kDa. The ability of the stromal factor to convert pLHCP into the 120-

kDa complex correlates with the ability of stromal protein to promote pLHCP

integration, supporting the idea that complex formation is a step leading to thylakoid

integration.

Recently, the precursor form of LHCP has been over-expressed in E. coli and

the purified pLHCP used in vitro to study the importance of soluble factors in

protein import into chloroplasts or integration into thylakoids. First, Waegemann

et al. (1990) reported that urea-solubilized pLHCP was not competent for import;

dialysis of pLHCP with soluble proteins was essential to obtain import competence.

Second, Yalovsky et al. (1992) reported that when urea-denatured pLHCP was

directly diluted into the integration reaction, it was inserted into thylakoids in the






39

absence of stromal extract. Further, they reported that stromal extract was required

only if urea-denatured pLHCP was dialyzed prior to the integration reaction and that

the plastid Hsp70 could replace stromal extract in this reaction, i.e. Hsp70 is the

stromal factor.

We have reexamined the soluble factor requirements) for pLHCP import and

integration using purified E. coli-made pLHCP as well as in vitro-translated pLHCP.

Our studies have shown that whereas unfolded pLHCP is efficiently imported into

chloroplasts without cytosolic factors, its integration into thylakoids absolutely

requires stromal extract. Hsp70 could not replace stromal extract in the integration

reaction and depleting stromal extract of Hsp70 did not impair its ability to support

integration, i.e. Hsp70 is not the stromal factor. When considered with the essential

nature of stromal extract, the finding that pLHCP is surprisingly stable in aqueous

solution indicates that the stromal factor most likely has an active role in the

integration reaction.



Materials and Methods

Materials

[3H]-leucine was from Du Pont-New England Nuclear. RNasin and SP6 RNA

polymerase were from Promega. Mg-ATP, thermolysin, fast flow DEAE-Sepharose,

C-8 linked ATP-agarose, protein A-Sepharose 4B, and Percoll were from Sigma. All

other chemicals were reagent grade. Plasmid psAB80XD/4 is the in vitro expression

plasmid for pLHCP (Cline, 1988). Plasmid pETHPLHCP is the E. coli over-







40
expression plasmid for pLHCP (Cline et al., 1993). Several antibodies were used in

this study. The antibody to LHCP has been described (Payan and Cline, 1991).

Antibody to the E. coli DnaK was kindly provided by John McCarty and Caroline

Donnelly (Massachusetts Institute of Technology, Cambridge). Antibody to tomato

cytosolic Hsp70 was a gift of Dr. Nover (Neumann et al., 1987).



Preparation of Radiolabeled Precursor, Chloroplasts, Lysates. Thylakoids, and SE

[3H]-pLHCP was prepared either by in vitro transcription (Fulsom and Cline,

1988) and translation (Cline, 1986, 1988) or by over-expression in E. coli in the

presence of ['H]-leucine (Cline et al., 1993). Chloroplasts were isolated from 9- to

10-day-old pea (Laxton's Progress 9) seedlings as described (Cline, 1986; Yuan et al.,

1991) and were resuspended in import buffer (50 mM Hepes/KOH, pH 8, 0.33 M

sorbitol). Lysates, thylakoids, and stromal extract (SE) were prepared from isolated

chloroplasts (Fulsom and Cline, 1988; Yuan et al., 1991). Lysates prepared at 0.5

mg/ml chlorophyll (Chl) were arbitrarily referred to as IX lysate and the SE resulting

from such lysate as 1X SE (~ 3 mg/ml protein).



Purification of Hsp70 from SE and Preparation of Anti-Hsp70 Antibody

Stromal Hsp70 was purified according to Welch and Feramisco (1985) as

follows. Stromal protein (100 mg) in buffer A (20 mM Hepes/KOH, pH 8, 20 mM

KC1, 5 mM MgCl2, 1 mM DTT) was loaded at 1 ml/min onto a 10 ml fast flow

DEAE-Sepharose column. After the column was washed at 2 ml/min with 30 ml of






41
buffer A, the bound proteins were eluted at 1 ml/min with a 50 ml, 20 to 500 mM

KCl gradient. The fractions containing Hsp70 at ~250 mM KCI were pooled,

diluted 5 fold with buffer A, and applied at 10 ml/hour to a 5 ml ATP-agarose

column that had been equilibrated with buffer A. The column was sequentially

washed with 10 ml buffer A, 20 ml 0.5 M KCI in buffer A, and 20 ml buffer A, and

was then eluted with 30 ml buffer A containing 3 mM ATP. The resulting

preparation was about 70% Hsp70 and 25% Hsp60 as determined by densitometry

of Coomassie blue-stained gels. The identity of Hsp70 was confirmed by specific

reaction with antibody against E. coli Hsp70/DnaK (Results). Hsp70 was further

purified by SDS gel electrophoresis, electroeluted from gel bands, and used for

antibody production in rabbits.



Depletion of Hsp70 from SE

Protein A-Sepharose 4B (1 ml aliquot) was washed (Payan and Cline, 1991),

suspended in 3 ml of 10 mM Hepes/KOH (pH 8), and then mixed with 3 ml of either

2% bovine serum albumin (BSA), preimmune serum, or anti-Hsp70 serum. The

mixture was incubated overnight at 4C and the supernatant removed by

centrifugation at 500 x g for 2 min. After 3 washes (5 ml each) with column buffer

(25 mM Hepes/KOH, pH 8, 50 mM KCI, and 10 mM MgCI2), protein A-Sepharose

matrices were transferred into syringe columns with glass fiber filter supports.

Stromal extract (2 ml 1X SE) was applied to each column and allowed to pass

through after 15 min incubation. The flow-through was reapplied to the column two







42
more times in a similar fashion and was finally recovered from the column by

centrifugation at 500 x g for 3 min.



Assays for Import, Integration, and Soluble Complex Formation

Import assays were conducted for 10 min as described (Cline, 1986; Yuan et

al., 1991) except that E. coli-made pLHCP was used instead of in vitro-translated

pLHCP. Assays were started by adding urea-denatured pLHCP to the assay mixture

as described in the figure legends. Unless otherwise specified, import reactions (300

il) contained chloroplasts equivalent to 0.33 mg/ml Chl, 50 mM Hepes/KOH, pH 8,

0.33 M sorbitol, 10 mM Mg-ATP, 2 mM DTT, 0.2 M or less urea, and approximately

0.2 pM pLHCP. Integration assays were performed both with E. coli-made pLHCP

and with in vitro-translated pLHCP for 30 min essentially as described (Cline, 1986,

1988). Integration reactions (300 il) with E. coli-made pLHCP received thylakoids

equivalent to 0.33 mg/ml Chl, 9 mg/ml stromal protein, 17 mM Hepes/KOH, pH 8,

55 mM sorbitol, 10 mM Mg-ATP, 1 mM DTT, 0.2 M or less urea, and approximately

0.2 tM pLHCP. To minimize microcentrifuge tube-adsorbed pLHCP, washed

thylakoids recovered from integration assays were transferred to new tubes before

analysis. Assays for soluble complex formation were performed according to Payan

and Cline (1991).








Sample Analyses

Samples recovered from the above assays were subjected to electrophoresis

on 12.5% SDS polyacrylamide gels (Laemmli, 1970) and fluorography (Bonner and

Laskey, 1974). About 10% of the chloroplasts or thylakoids recovered from each

assay were loaded per gel lane. Quantification of import or integration was

accomplished by scintillation counting of radiolabeled proteins extracted from excised

gel bands (Cline, 1986; Walter et al., 1981).



Miscellaneous Methods

Chl concentrations were determined according to Arnon (1949). Protein

assays were performed by the BCA method (Pierce) for samples without DTT or by

the Bradford method (Bradford, 1976) for samples with DTT using BSA as a

standard. Immunoprecipitation and immunoblotting were carried out as described

(Payan and Cline, 1991).



Results

Stromal Components Are Absolutely Essential for Integration of pLHCP into
Thylakoids

Precursor LHCP was produced by over-expression in E. coli. The over-

expressed pLHCP was sequestered in inclusion bodies. Isolation of inclusion bodies

yielded pLHCP that was 90-95% pure as determined by densitometry of Coomassie

blue-stained gels. The specific radioactivity of E. coli-made pLHCP was 400,000 to

800,000 dpm per jig protein. When purified pLHCP was denatured in 8 M urea and







44





Import Integration
P 1 2 3 4 5 P 1 2 3 4 5

A O--


LHCP-DP

ATP + + + ATP + + +
Ught + SE + + -
SF + + RT + -




Figure 3-1. Soluble factors are not required for import of purified
pLHCP into chloroplasts whereas unfolded pLHCP still requires
stromal extract for integration into thylakoids. E. coli made-pLHCP
was dissolved in 8 M urea, 8 mM DTT at room temperature for 4 hrs
and then directly diluted into assays for import into chloroplasts or
integration into thylakoids. Import assays were carried out either in
dark with 10 mM ATP (lanes 1, 4, and 5), without ATP (lane 2); or in
light (~ 70 pmol.m-.s-) with no added ATP (lane 3). Some assays
also received soluble factors (SF) i.e. wheat germ protein (lane 4, 1
mg) or rabbit reticulocyte protein (lane 5, 3 mg). Integration assays
were carried out in dark with 10 mM ATP (lanes 1, 3, and 4) or
without ATP (lanes 2 and 5). Some assays also received 3 mg protein
of stromal extract (SE, lanes 1 and 2) or rabbit reticulocyte lysate (RT,
lane 3). Recovered chloroplasts and thylakoids were either extracted
with 0.1 M NaOH (A) or treated with thermolysin (B). Gels were
loaded on an equivalent Chl basis (4.5 g/lane). Lane P, purified
pLHCP. LHCP-DP, a diagnostic protease degradation product of
correctly assembled LHCP or pLHCP (Cline, 1988; Andersson et al.,
1982; Mullet, 1983). Marker (*) points at a protease degradation
product previously characterized as inserted but resulting from
incompletely assembled pLHCP (Reed et al., 1990).






45

then directly diluted into reaction mixtures, it immediately adopted a form competent

either for import into chloroplasts or for integration into thylakoids (Fig. 3-1).

Purified pLHCP was able to import into isolated chloroplasts in the absence

of any soluble factors as long as energy (ATP or light) was provided (Fig. 3-1B). The

E. coli-made substrate was imported as efficiently as in vitro-translated pLHCP; up

to 15% of the added pLHCP was imported during the 10 min incubation. Addition

of cytosolic components, i.e. wheat germ extract or reticulocyte lysate, did not

stimulate the import of purified pLHCP. On the contrary, elevated quantities of

wheat germ extract inhibited import of E. coli-made pLHCP (Fig. 3-1, lane 4) as well

as in vitro-translated pLHCP (unpublished results). Waegemann et al. (1990)

previously reported that urea-denatured pLHCP was not competent for import unless

it was first dialyzed with cytosolic factors (leaf extract). It is apparent from

Waegemann et al. (1990) that cytosolic factors can be helpful to pLHCP import

under certain conditions. However, the results in Figure 3-1 demonstrate that they

are not essential.

In contrast, integration of purified pLHCP into isolated thylakoids absolutely

required the presence of stromal extract in addition to ATP (Fig. 3-1B). Titration

of the stromal dependence showed that at least 2.5 mg/ml stromal protein was

necessary to obtain appreciable integration (data not shown). In the presence of 9

mg/ml stromal protein, as much as 15% of the added pLHCP was integrated into

thylakoids during a 30 minute incubation. Neither wheat germ extract (not shown)















) N-Q.
(o CO X

A -t pLHCP

W -- LHCP-DP
B

1 2 3




Figure 3-2. Hsp70 is not able to support integration of purified
pLHCP into isolated thylakoids. E. coli-made pLHCP was solubilized
in 8 M urea, 8 mM DTT for 4 hrs at room temperature and then
dialyzed to remove the urea before assay for integration. Dialysis was
initiated by mixing 10 pl ( 6 tLg) of urea-solubilized pLHCP with 400
pl of buffer A containing either 5 mg stromal protein (lane 1), no
addition (lane 2), or 100 pig ATP-agarose affinity-purified stromal
Hsp70 (lane 3). Dialysis was conducted at 4C for 30 min against
buffer A on number 3 (molecular weight cut off, 3500 Da) Spectra/por
dialysis membrane (Spectrum Medical Industries, Inc., Los Angeles,
California). At the end of dialysis, 200 pl of the dialysis mixture was
mixed with ATP, washed thylakoids, and assayed for integration.
Recovered thylakoids were treated either by alkali extraction (A) or
with thermolysin (B). LHCP-DP and marker (*), see legend to Figure
3-1.






47
nor rabbit reticulocyte lysate (Fig. 3-1B, lane 3) could replace stromal extract for

pLHCP integration.

The above conclusions were made based on protease-resistance of integrated

(p)LHCP. It has been shown that membrane-integrated (p)LHCP is largely resistant

to protease digestion, yielding a characteristic degradation product (LHCP-DP) upon

treatment with proteases (Cline, 1988; Andersson et al., 1982; Mullet, 1983). Using

resistance to alkali extraction as a criterion for integration, Yalovsky et al. (1992)

concluded that stromal extract was not required for thylakoid integration of

denatured/unfolded pLHCP. However, comparison of these two treatments showed

that alkali extraction was not sufficiently rigorous to differentiate integrated from

surface-bound pLHCP (compare Fig. 3-1, A and B). NaOH-resistant pLHCP was

obtained with or without stromal extract, with or without ATP (Fig. 3-1A). Alkali

extraction was also not effective in removing pLHCP bound to the surface of intact

chloroplasts (Fig. 3-1A). Similar conclusions have recently been reached by two

other research groups even with in vitro-synthesized pLHCP (Auchincloss et al., 1992;

Huang et al., 1992).



Hsp70 Is Not the Stromal Factor Required for pLHCP Integration

Yalovsky et al. (1992) reported that urea-denatured pLHCP lost integration

competence upon dialysis unless stromal extract or purified Hsp70 was present

during dialysis. We also observed that inclusion of Hsp70 during dialysis of urea-

denatured pLHCP led to increased alkali-resistant pLHCP (Fig. 3-2A). However,



















Anf- Anti-
LHCP HSD70
PI IM PI IM
A --- Stromal Hsp70
Sample Well

B I ( (Soluble Complex


-- Dye Front
1234




Figure 3-3. Hsp70 is not part of the pLHCP soluble complex. [3H]-
pLHCP (translated in vitro) was mixed with stromal extract and
incubated for 15 min at 25C in the presence of 5 mM ATP. The
mixture was then subjected to immunoprecipitation analyses either
with preimmune (PI) sera (lanes 1 and 3) or with immune (IM) sera
(lanes 2 and 4). After removal of the pellet by centrifugation, the
supernatant was analyzed for the presence of Hsp70 by
immunoblotting (A) or for the presence of the soluble pLHCP
complex by native gel electrophoresis (B).






49

integration of pLHCP as judged by protease-resistance only occurred when stromal

extract was present during dialysis (Fig. 3-2B). We suggest that dialysis with Hsp70

increases association of pLHCP with the thylakoids because Hsp70 maintains pLHCP

in solution throughout dialysis. In our experiment, most of the buffer-dialyzed

pLHCP was adsorbed to the dialysis apparatus.

Payan and Cline (1991) have previously correlated the integration-promoting

activity of stromal extract with its ability to convert pLHCP into a 120-kDa soluble

complex. They showed that the plastid Hsp60 is not a component of the complex,

but could not determine whether the plastid Hsp70 was a component. To assess

whether Hsp70 was part of the 120-kDa pLHCP soluble complex, we prepared an

antibody to the plastid Hsp70 and used this antibody in immunoprecipitation

analyses. Such analyses showed that the pLHCP soluble complex was not removed

by treatment with anti-Hsp70 antibody (Fig. 3-3, lane 4). In contrast, all of the

complex was removed by treatment with anti-LHCP antibody (Fig. 3-3, lane 2).

The possible involvement of Hsp70 in integration was further assessed by

assaying pLHCP integration with Hsp70-depleted stromal extract obtained by

immunoaffinity chromatography on an anti-Hsp70 IgG protein A-Sepharose column.

Control stromal extract was treated identically either with a preimmune or a BSA

mock-treated column. Depletion of Hsp70 was verified by Coomassie blue staining

(Fig. 3-4A) and by immunoblotting with anti plastid Hsp70 or anti E. coli DnaK

antibody (Fig. 3-4B). By using a dilution calibration curve for immunoblotting with

the anti plastid Hsp70, we determined that the depletion treatment removed more

















*Wx 1 2 3 4 1 2 3 4 5
S74

A kCs *0




P 123456
-And- ." ,4 -P
A W a'eW *6* 8Deceorn^
1 23 1 23





Figure 3-4. Hsp70-depleted stromal extract supports pLHCP
integration equally well as does Hsp70-containing stromal extract.
Hsp70-depleted stromal extract (SE) was prepared as described in
"Materials and Methods". A. Coomassie blue staining of stromal
protein passed through a BSA mock column (lane 1), passed through
a preimmune IgG column (lane 2), passed through an anti-Hsp70 IgG
column (lane 3); lane 4, ATP-agarose affinity-purified stromal Hsp70
(shows the enrichment of Hsp70 as well as the presence of Hsp60).
B. Immunoblotting of stromal protein untreated (lane 1), passed
through a BSA mock column (lane 2), passed through an anti-Hsp70
IgG column (lane 3). C. Complex formation with SE (lane 1); without
SE (lane 2); with SE passed through a BSA mock column (lane 3),
passed through a preimmune IgG column (lane 4), passed through an
anti-Hsp70 IgG column (lane 5). D. Integration with 1X SE (lane 1);
without SE (lane 2); with purified stromal Hsp70 (lane 3); with 1.5X
SE passed through a BSA mock column (lane 4), passed through a
preimmune IgG column (lane 5), passed through an anti-Hsp70 IgG
column (lane 6). Lane P, in vitro-translated pLHCP. LHCP-DP and
marker (*), see legend to Figure 3-1.















-r160-
CO 140 -- BSAMock
C 120- -- PI IgG
S100 "..- IMIgG
g) 80-

*) 40- 40
S20

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
[SE] (X)





Figure 3-5. Stromal protein concentration-dependent integration of
pLHCP is independent of the presence of Hsp70. Quantification of
pLHCP integration (see legend to Figure 3-4 for details) with different
concentrations (based on protein assays) of stromal extract (SE) that
had been depleted of Hsp70 with an anti-Hsp70 IgG column (IM IgG)
or treated identically with a preimmune IgG column (PI IgG) or with
a BSA mock column (BSA Mock). The protein concentration of 1X
stromal extract was 2.45 mg/ml (0.82 mg/ml final concentration in
assay). Data are expressed as a percentage of the integration obtained
with 1X untreated SE (1000 dpm per ig Chl).







52

than 95% of the stromal Hsp70. Yet, the Hsp70-depleted stromal extract was as

competent as control stromal extract for integrating pLHCP into thylakoids (Fig. 3-

4D) or converting pLHCP into the 120-kDa soluble complex (Fig. 3-4C). Marshall

et al. (1990) reported that stromal extract contains a second Hsp70. This second

Hsp70 is present in very low amounts in stromal extract and, although difficult to

visualize by Coomassie staining, can be monitored by immunoblotting with an

antibody against tomato cytosolic Hsp70 (Neumann et al., 1987). With a similar

dilution calibration curve, we determined that 50-60% of the minor Hsp70 was

removed by our depletion treatment (data not shown). In the experiments with

depleted stromal extract, the stromal component was the limiting factor for

integration (Fig. 3-5). The identical response of pLHCP integration to different

amounts of control or Hsp70-depleted stromal extract demonstrates that Hsp70 is

not the stromal factor previously described.



pLHCP, Diluted out of Urea, Is Stable as a Substrate for Import and Integration

The notion that unfolded pLHCP is sufficient for integration and that Hsp70

is the stromal factor presupposes that pLHCP rapidly folds into an integration

incompetent conformation. We examined this by diluting urea-denatured pLHCP

to a low urea concentration (-0.2 M) and then assaying aliquots for import and

integration at various times after dilution. The amount of import or integration for

each time point is plotted in Figure 3-6 as a percentage of the no-preincubation

control. The amount of import or integration was about 70% of the no-preincubation

















S* ...... ....
E40
0 25'C
0 4 8 12 16 20 24 28
Time (min)
120.
**o B



4 ~ a W

0 4 8 12 16 20 24 28
Time (min)





Figure 3-6. Urea-denatured pLHCP remains competent for import
and integration during incubation under renaturing conditions. E. coli-
made pLHCP was denatured in 8 M urea, 8 mM DTI' at room
temperature for 4 hrs and then diluted to -~ 0.2 M urea into import
buffer plus 2 mM DTT. At various times after the dilution, pLHCP
was assayed (at a calculated concentration of ~0.2 jIM) either for
import into chloroplasts (A) or for integration into thylakoids in
chloroplast lysates (B). Preincubation of pLHCP in buffer was
performed at 0C or at 25C. Assays for both import and integration
were carried out in light ( ~ 70 jimol.m-.s-1) in the presence of 10 mM
added ATP. The absolute number of pLHCP imported or integrated
was, respectively, ~ 5.5 x 10' or ~ 5.7 x 104 molecules per chloroplast
or chloroplast equivalent for the no-preincubation controls. About
14% of the added pLHCP was imported or integrated. Inset in A,
fluorogram of import with pLHCP preincubated at 0C. Inset in B,
fluorogram of integration with pLHCP preincubated at 0C.






54

control when diluted pLHCP was preincubated for 30 min at 0C, and around 50%

of the no-preincubation control when preincubated for 30 min at 250C. Thus,

although a time- and temperature-dependent decline in import and integration

competence occurred, it was not rapid.

We also measured the amount of pLHCP present in each aliquot and found

similar declines with time (data not shown). Presumably, pLHCP in 0.2 M urea is

prone to adsorption to surfaces such as the plastic tubes and pipet tips used in this

experiment. This further supports our earlier assertion (Payan and Cline, 1991) that

measurable loss of pLHCP integration competence results from non-productive

reactions that lead to insolubility, e.g. aggregation and adsorption, rather than from

folding.



Discussion

Our data clearly show that whereas urea-denatured pLHCP is sufficient for

import into intact chloroplasts (Fig. 3-1), integration into thylakoids requires a

stromal factor that can not be bypassed by unfolding (Fig. 3-1). The apparent

integration reported by Yalovsky et al. (1992) in the absence of stromal extract can

be entirely explained as unproductively associated pLHCP that was not removed by

alkali extraction. Alkali extraction was originally designed to differentiate integral

from peripheral membrane proteins in their native forms (Steck and Yu, 1973).

Resistance to alkali extraction has traditionally been interpreted as evidence for

protein-membrane interaction. However, caution must be taken when such a






55
criterion is used to assess the behavior of denatured or non-native proteins such as

those used in transport studies. In the case of pLHCP integration, the current

evidence indicates that the association of pLHCP with thylakoids in the absence of

stromal extract is not physiologically relevant

Our experiments have also demonstrated that Hsp70 is not the stromal factor

(Figs. 3-2, 3-4 and 3-5) We cannot rule out a role for Hsp70 in pLHCP integration

because of the technical difficulty of removing all Hsp70 proteins from assay

mixtures. In the experiment shown in Figures 3-4 and 3-5, Hsp70 proteins (verified

by immunoblotting) were associated with the washed thylakoids used in the assays

and were present in the wheat germ extract used for preparing the pLHCP. It was

necessary to use in vitro-translated pLHCP in the depletion experiments because it

is not technically feasible to prepare Hsp70-depleted stromal extract sufficiently

concentrated for detectable integration of E. col-produced pLHCP, owing to its

much lower specific radioactivity. Nevertheless, our results show that plastid Hsp70

is not the stromal factor previously described. First, no integration occurs in the

absence of stromal extract even though Hsp70s are present on thylakoids and in the

translation mixture (Fig. 3-4). Second, since the stromal factor is limiting in the

integration reaction, its removal in total or in part would be reflected in a decrease

in integration. As shown in Figure 3-5, there is no difference in the integration-

promoting activity between control and Hsp70-depleted stromal extract.

It has been shown that premature folding can prevent membrane transport of

a variety of authentic and chimeric preproteins (Liu et al., 1989; Weiss and Bassford,







56

1990; della-Cioppa and Kishore, 1988; Eilers and Schatz, 1986). However, folding

does not appear to be a problem for pLHCP import and integration. The apparent

decrease in import and integration competence (Fig. 3-6) can be explained by the

time-dependent loss of pLHCP from solution. This is not surprising as a stably

folded pLHCP is expected only upon its interaction with the lipid bilayer. A

molecular chaperone would help to prevent loss of pLHCP from solution. Indeed,

our previous analyses demonstrated a chaperone function for the stromal factor

(Payan and Cline, 1991). However, such a role is not essential as demonstrated by

the experiment in Figure 3-6, where pLHCP lost only ~50% of its initial

competence during an incubation that mimicked the assay conditions for integration.

The fact that the stromal factor is absolutely essential (Figs. 3-1 and 3-2) supports

the proposal that the stromal factor has an additional function, most probably in

targeting pLHCP to the membrane or possibly even in the mechanism by which

pLHCP folds into the lipid bilayer.











CHAPTER 4
PLASTOCYANIN AND THE 33K OXYGEN-EVOLVING PROTEIN
ARE TRANSPORTED INTO THYLAKOIDS WITH SIMILAR
REQUIREMENTS AS PREDICTED FROM PATHWAY SPECIFICITY



Introduction

Plastocyanin (PC) and the 33K subunit of the oxygen-evolving complex

(OE33) are two of several thylakoid lumen-located proteins that are made in the

cytosol and transported into thylakoids. Recently, competition studies showed that

there are two pathways for protein transport into the thylakoid lumen and that PC

and OE33 are on the same pathway (Cline et al., 1993). As it has been shown in

other systems that transport requirements are intimately related to the mechanisms

of the translocation machinery (Neupert et al., 1990; Pugsley, 1993; Gilmore, 1993),

our expectation is that PC and OE33 share similar transport requirements.

It is generally agreed that OE33 and PC require ATP for transport (Kirwin

et al., 1989; Bauerle and Keegstra, 1991) and that OE33 transport is stimulated by

a proton gradient (Mould and Robinson, 1991; Cline et al., 1992a). However, the

stromal requirement for transport of PC and OE33 and the involvement of a proton

gradient in PC transport are still uncertain. For example, there are conflicting

reports on the role of a proton motive force (PMF) in PC transport (Theg et al.,

1989; Mould and Robinson, 1991; Cline et al., 1992a).






58

One problem with previous studies is that the experiments were conducted

with intact chloroplasts rather than with isolated thylakoids such that it was more

difficult to assess the contributions of different components of the energy

requirements and also the requirement for stromal factorss. The recent

development in our laboratory of an in vitro assay for PC and OE33 transport with

isolated thylakoids has allowed us to investigate the transport properties of these two

proteins in more detail. The results show that PC and OE33 are transported into

thylakoids with similar requirements; they both require ATP, stromal protein(s) and

the trans-thylakoidal proton gradient for maximum transport. These results add

support to the conclusions of Henry et al. (1994) and Robinson et al (1994) from

analyses of transport of chimeric precursor proteins that transport requirements are

intrinsic properties of translocation pathway and not of the translocated protein.



Materials and Methods

Preparation of Precursor Proteins

All reagents, enzymes, and standards were from commercial sources. In vitro

expression plasmids for pLHCP, pOE33, iOE33 and pPC from pea, and pPC from

Arabidopsis thalliana have been described (Cline et al., 1993). Radiolabeled

precursor proteins were prepared by in vitro transcription and subsequent in vitro

translation in a wheat germ system in the presence of [3H]-leucine (Cline et al.,

1993).








Preparation of Chloroplasts, Lysate, Thylakoids, and Stromal Extract

Intact chloroplasts were isolated from 9- to 10-day-old pea (Pisum sativum L.

cv. Laxton's Progress 9) seedlings (Cline, 1986). Chloroplast lysate, thylakoids and

stromal extract (SE) were prepared as described (Yuan et al., 1993). Thylakoids

were washed twice with import buffer (50 mM Hepes/KOH, pH 8, 0.33 M sorbitol)

plus 10 mM MgCl2 and finally resuspended in import buffer. When chloroplast lysate

was used in assays, chloroplasts were lysed immediately before the assay. Lysate

prepared at a chlorophyll (Chl) concentration of 0.5 mg/ml is arbitrarily referred to

as 1X lysate and the SE resulting from such lysate as 1X SE (approximately 2.5 mg

protein/ml).



Assays for Integration or Transport into Thylakoids

Assays for PC and OE33 transport into isolated thylakoids were conducted

for 30 min at 25C with chloroplast lysate or reconstituted lysate (thylakoids plus SE)

equivalent to 50 jig of Chl in a total volume of 150 pl containing 37 mM

Hepes/KOH (pH 8), 220 mM sorbitol, 4 mM MgCI2, and 4 mM Mg-ATP. Assays

were terminated by transfer to 0C, followed by recovery of thylakoids and

subsequent protease (thermolysin) treatment to remove surface-bound precursor

proteins. Assays for LHCP integration in this study were carried out exactly as for

PC and OE33 transport and were different from the original protocol (Cline, 1986)

in that the reaction mixture had lower concentrations of sorbitol, ATP, and MgCI2.









Use of lonophores and Coupling Factor Inhibitors

lonophores and the coupling factor (CFi/CFo) inhibitor venturicidin were

prepared in ethanolic stocks. The CFi inhibitor tentoxin was prepared in aqueous

stock. Upon addition of ionophores or tentoxin, reaction mixtures were routinely

incubated on ice for 10 min prior to assay. Venturicidin was added at 10 PM and

incubated with thylakoids at 25 C for 10 min before the addition of radiolabeled

precursor protein.



Analysis of Precursors, Recovered Chloroplasts and Thylakoids

Samples of precursors, recovered chloroplasts and thylakoids were subjected

to SDS-PAGE and fluorography and quantification was accomplished by scintillation

counting of radiolabeled proteins extracted from excised gel bands (Cline, 1986).



Miscellaneous

Protein assays were performed by the BCA method (Pierce) using bovine

serum albumin as a standard. Chl concentrations were determined according to

Arnon (1949).



Results

Transport of PC and OE33 into Isolated Thylakoids

Figure 4-1 shows the results of OE33 and PC transport assayed with

chloroplast lysate in light in the presence of added ATP. Transport of PC and OE33







61








P1234 P1234

pOE33 r, 1m i0E33 m_-

p^, P -"
pPCpea o pPCara
rm- -- -
Thermolysin + + + + + +
Sonication + +
Triton X-100 + +





Figure 4-1. Transport of PC and OE33 into thylakoids. Chloroplast
lysate (200 Ipg of Chl) was mixed with radiolabeled protein (pPCara,
pPCpea, pOE33, or iOE33) and incubated in the presence of 4 mM
Mg-ATP in a total volume of 600 pl of import buffer, 4 mM MgCI2 in
light (70 IE.m2s1) at 250C for 30 min. After the assay, the thylakoids
were recovered by centrifugation, washed one time with import buffer,
resuspended in 200 p of import buffer and divided into four equal
portions. One portion was mock-treated in import buffer with no
addition (lane 1), one portion was treated with 0.2 pg thermolysin per
Ipg Chl for 40 min at 4C (lane 2), a third portion was sonicated in a
bath sonicator during thermolysin treatment (lane 3), and the fourth
portion was treated with thermolysin in the presence of 1% Triton X-
100 (lane 4). Proteolysis was terminated by addition of EDTA to a
final concentration of 50 mM. An equal volume of 2 X SDS sample
buffer was added to samples immediately before boiling for 5 min, and
about 10% (-5 pg Chl) of each sample was loaded to gels and
analyzed by SDS-PAGE and fluorography. Lane P represents 2% of
the full-length or the intermediate-sized precursor added to each assay.
Positions are indicated for the precursor (p), intermediate (i), and
mature (m) forms of the proteins.







62

into the thylakoid lumen was evidenced by proteolytic processing by the lumenal

protease to the mature forms of the proteins (lane 1). Several lines of evidence

confirm that the appearance of mOE33 and mPC resulted from transport across

thethylakoid membrane. The accumulation of mPC and mOE33 was energy

dependent (see below), time dependent (increasing for up to 30 min) and

temperature dependent, i.e. accumulation of mature forms did not occur at 0C (data

not shown). mPC and mOE33 were found only with the thylakoid pellet upon

centrifugation (data not shown). mPC and mOE33 were resistant to added protease

(thermolysin) whereas associated precursors or intermediate forms were degraded

(lane 2). When the lumen was exposed to protease by sonication or treatment with

1% Triton X-100, mOE33 was degraded (lanes 3 and 4). Unexpectedly, mPC was

resistant to degradation under these conditions. A more comprehensive examination

with several proteases including trypsin, chymotrypsin, and proteinase K confirmed

that mPC is intrinsically resistant to proteolysis (data not shown). A combination of

trypsin and chymotrypsin that was reported capable of digesting mPC in an earlier

study (Bauerle and Keegstra, 1991) also failed to degrade mPC in our experiments.

To verify that mPC was indeed located in the lumen, thylakoids were treated with

0.05% Triton X-100, an amount sufficient to permeablize thylakoids without

solubilizing intrinsic membrane proteins (Ettinger and Theg, 1992; Yuan and Cline,

unpublished results). After treatment and centrifugation to pellet the thylakoids,

80% of the mPC was recovered in the supernatant.
























P12345
P-f
pOE33 m- -


pPCpea
m "m
Lvscte +


Thylakoids
[SE) (X)


P12345

i0E33 '-


pPCara


m-- --


246


246


Figure 4-2. Stromal extract stimulates PC and OE33 transport into
isolated thylakoids. Radiolabeled pPCara, pPCpea, pOE33, and iOE33
were assayed for transport either with chloroplast lysate (lane 1) or
with washed thylakoids supplemented with increasing amount of SE
(lanes 2-5) as designated below the panels. About 10% of each
sample was loaded to gels. Lane P represents 2% of the full-length or
the intermediate-sized precursor added to each assay.







64
We have observed that up to 28% of pPCpea, 16% of pPCara, or 10% of

pOE33 added to the reaction mixture was transported into the lumen. However, in

most assays the efficiency was about 5% for OE33 and 10% for PC transport. An

engineered iOE33 which lacked the stroma-targeting domain and was unable to

import into intact chloroplasts (data not shown) was as efficiently transported into

thylakoids as pOE33 (Figs. 4-1 and 4-2). This level of transport is significantly less

than that for two other lumenal proteins OE23 and OE17 (Cline et al., 1992a), but

approximately the same as is obtained for integration of LHCP (Cline, 1986; Yuan

et al., 1993). We examined transport of PCpea as well as PCara in all experiments,

but only showed the data for PCara in some experiments because of its higher level

of translation and overall transport (see below).



Stromal Components Are Necessary for Efficient Transport of PC and OE33 into

Thylakoids

It was necessary to include SE in assays for efficient transport of both the full-

length precursors (pPC and pOE33) and the intermediate-sized precursor (iOE33)

(Fig. 4-2). In general, transport increased with increasing concentrations of SE (Fig.

4-2). With 6X SE (stoichiometry to thylakoids in assay), transport was enhanced

approximately 3-fold for OE33 and 2-fold for PC from pea, respectively. The

response of PC fromArabidopsis was more complex. At relatively low concentrations

of SE (e.g. 2X SE), transport of PCara was usually stimulated about 2-fold, however,

further additions of SE reduced transport (Fig. 4-2, lower right panel).





















p-
m-
P1234567




Figure 4-3. SE stimulating activity is resistant to ribonuclease but
sensitive to heat and protease treatments. Radiolabeled pPCara was
mixed with thylakoids and assayed for transport with SE (lane 1),
without SE (lane 2), or with SE treated with heat (lane 3),
ribonuclease (lanes 4 and 5) or protease (lanes 6 and 7). Heat
treatment of SE (lane 3) was performed for 10 min at 60 C.
Treatment of SE (5 mg/ml) with ribonuclease A at 0.1 mg/ml (lane 4)
or at 1 mg/ml (lane 5) was carried out at 25 *C for 20 min. SE
became turbid after ribonuclease treatment. Treatment of SE (5
mg/ml) with active (lane 6) or PMSF-inactivated proteinase K (lane 7)
at 50 gg/ml was carried out at 4 C for 30 min. Active proteinase K
was deactivated after treatment. Proteinase K was deactivated at 4 C
for 30 min with the addition of freshly prepared PMSF to 10 mM.
After treatments, SE was centrifuged on a microcentrifuge for 1 min
at 15,000 rpm, 4 C. Thylakoids recovered from transport assays were
washed twice with import buffer without thermolysin treatment. Each
lane of the gel received approximately 10% of the recovered samples.
Lane P represents 2% of the precursor added to each assay.







66
To determine the nature of the stromal stimulating activity, SE was pretreated

with heat, protease and ribonuclease and then assayed for its ability to stimulate PC

and OE33 transport. Treatment of SE with ribonuclease (RNase A) did not affect

transport of PC into thylakoids, but treatment with heat or proteinase K abolished

its ability to stimulate PC transport (Fig. 4-3). Virtually the same results were

obtained with OE33 transport (data not shown). Thus, it appears that the stromal

stimulating component is a proteinaceous factor.



ATP Is Essential for PC and OE33 Transport into Thylakoids

Transport of pPC and pOE33 into thylakoids occurred in darkness with

exogenously added ATP or in the light without added ATP (Fig. 4-4, lanes 2 and 6).

As has been noted before (Cline et al., 1992a), both conditions can result in ATP

and a trans-thylakoidal PMF due to the reversible nature of the thylakoidal ATP

synthase. Previous work (Cline et al., 1992a) identified conditions for manipulating

ATP levels and the PMF in transport assays. Here, we used the same set of

conditions for PC and OE33 transport (Fig. 4-4). Parallel assays were also carried

out with LHCP as a control to verify the efficacy of the treatments. As can be seen

from Figure 4, transport of pPC and pOE33 in light required ATP as evidenced by

severe inhibition of transport in the presence of tentoxin (lane 8) or apyrase (lane

9). Apyrase depletes ATP (and GTP) via hydrolysis, whereas tentoxin prevents its

formation from photophosphorylation. In experiments designed to determine the

nucleotide specificity for PC transport, i.e. using translation products and stromal







67









Light + + + + + +
ATP + + +
Nig/Val + +
Apyrose +
Tentoxin + +
P1 2345678 9C
pLHCP--
LHCP-DP- O I 4

pOE33-
mOE33- 4

pPCara-4

mPCora- M -





Figure 4-4. Transport of PC and OE33 into thylakoids requires ATP
and is stimulated by the trans-thylakoidal PMF. PC and OE33
transport as well as LHCP integration assays were carried out under
conditions designed to provide both ATP and a PMF (light, dark +
ATP, light + ATP), ATP alone [dark + ATP + nigericin and
valinomycin (Nig/Val), dark + ATP + tentoxin], a PMF alone (light
+ tentoxin, light + apyrase), or neither energetic component (dark,
light + Nig/Val). Both integration and transport assays were
conducted with chloroplast lysate. Mg-ATP was added to 10 mM final
concentration. Apyrase was added to lysate at 2 units per 300 Pl assay
and the mixture incubated on ice for 10 min. Tentoxin was to 6 IM
final concentration. Nigericin and valinomycin were added to 0.5 tiM
and 1 uM, respectively, from an ethanolic stock. Control assays
received an equal amount of ethanol (lane C). The final ethanol
content was 0.33% in assays and was shown to have no effect on the
level of transport or integration. About 10% of the samples were
loaded to gels for analyses. Lane P represents 2% of the precursor
added to the assay reaction.







68
extract from which all nucleotides had been removed by desalting, we found that

ATP was the most effective nucleotide triphosphate but that GTP, CTP, and UTP

were also able to support transport (Yuan and Cline, unpublished results). These

results indicate that ATP is essential for transport of PC and OE33 into thylakoids.

Similar conclusions were also reached by Bauerle and Keegstra (1991) for PC and

Kirwin et al. (1989) and Hulford et al. (1994) for OE33 transport.



A PMF Contributes to Transport of PC and OE33 into Thylakoids

The results in Figure 4 also show that the trans-thylakoidal PMF was helpful

to transport of PC and OE33 into thylakoids. Inclusion of tentoxin (lane 4) or of a

combination of nigericin and valinomycin (lane 3) in assays conducted with ATP in

the dark reduced pOE33 transport up to 70% and reduced pPCara (Fig. 4-4) and

pPCpea (not shown) transport ~ 30-40%. The reduction in PC transport was

consistently observed whenever the trans-thylakoidal PMF was dissipated with

ionophores or the formation of a PMF was prevented with inhibitors of the

thylakoidal ATP synthase. This was true when assays were conducted with levels of

added ATP from 1 to 10 mM or with SE ranging from 0 to 6X (data not shown).

With higher levels of SE, up to 50% inhibition by ionophores was observed for PC

transport.

Inhibition of transport by ionophores appears to be a primary effect on the

PMF, rather than a secondary effect resulting from the possible ATP depletion by

the uncoupled ATPase activity of the thylakoidal ATP synthase. This was shown by




















pOE33
P1 2 3 4


100
S80
60
40
20
0
Nig +
Val +
Nig/Val +


pPCora
P1 2 3 4





-


Figure 4-5. ApH is the predominant component of PMF in stimulating
transport of PC and OE33 into thylakoids. Radiolabeled pPCara or
pOE33 was mixed with chloroplast lysate and assayed for transport in
darkness at 25C for 30 min in the presence of 4 mM added ATP.
Nigericin at 0.5 giM, valinomycin at 1 tiM, or a combination of
nigericin (0.5 jIM) and valinomycin (1 tiM) was added to assays as
indicated; transport control reactions received an equivalent amount
of ethanol. Approximately 10.6% of pPCara and 5.6% of pOE33
added to assays were transported into thylakoids in the control assays.
Lane P represents 2% of the precursor added to the transport
reaction.







70
conducting a time course analysis of transport in the absence or presence of

ionophores. If ATP were being depleted, the later stages of the transport reaction

would be affected to a much greater extent than the early stages. In fact, during a

40-min transport reaction in the presence of 10 mM Mg-ATP, a combination of 0.5

pM nigericin and 1.0 IM valinomycin was equally inhibitory throughout the course

of the reaction (data not shown).



ApH Is the Active Component of PMF in Stimulating PC and OE33 Transport

To determine the relative contribution of ApH and A4 to transport, individual

effects of nigericin and valinomycin were compared (Fig. 4-5). Nigericin acts as an

electroneutral H+/K+ antiporter and dissipates the proton gradient (ApH).

Valinomycin functions as an electrogenic K' uniporter and collapses the

electrochemical gradient (At). The results showed that nigericin by itself was as

effective in inhibiting PC and OE33 transport as the combination of nigericin and

valinomycin, whereas valinomycin alone exerted little if any inhibition, demonstrating

that ApH is the predominant component of the trans-thylakoidal PMF in stimulating

PC and OE33 transport.



PMF Does Not Change the ATP Requirement for PC Transport

Studies of protein export from Escherichia coli suggest that PMF alters the

response of the translocation machinery to ATP. Specifically, Shiozuka et al. (1990)

have shown that in the presence of a PMF, the Km value of the translocation























[ATP] (mM)


Figure 4-6. ATP concentration curves for PC transport in the
presence or absence of a PMF. Radiolabeled pPCara was mixed with
chloroplast lysate and assayed for transport in the presence of a PMF
(dark, light + venturicidin) or in the absence of a PMF (dark +
venturicidin, light + venturicidin + nigericin + valinomycin). ATP
was added to assays at 0.0, 0.1, 0.5, 2.0, and 5.0 mM. For the no-ATP
control, apyrase at 2 units/assay was included in assays (300 Il) to
deplete the residual amount of ATP ( 0.2 mM) present in translation
mixture and chloroplast lysate. The plotted ATP values were obtained
by combining the added ATP with the amount contributed by
translation mixture and lysate. Venturicidin (Vt) was added at 10 PM
to prevent formation of a PMF from ATP hydrolysis or to inhibit ATP
formation via photophosphorylation. Venturicidin was added to assays
from an ethanolic stock; control transport reactions received an equal
amount of ethanol. Venturicidin at 10 pM in the dark completely
blocked ATP-derived OE23 transport (data not shown) and was thus
effective in preventing ApH formation. A mixture of nigericin (N) at
0.5 p.M and valinomycin (V) at 1 IM was used to dissipate the trans-
thylakoidal PMF for light ATP-driven transport reactions. The
maximum transport of PCara for this experiment was 940 molecules
per chloroplast equivalent. Similar response was observed when this
experiment was performed with assay components depleted of ATP (by
gel filtration of SE and translation product), but the overall transport
activity was too low for quantitative analysis.






72
reaction for ATP is substantially lower than that in the absence of a PMF. We

examined this possibility for thylakoid protein transport by assaying PC transport

with varying levels of ATP in the presence or absence of the trans-thylakoidal PMF

(Fig. 4-6). For this experiment, the reversible activity of the thylakoidal ATP

synthase was inhibited by venturicidin, which blocks the proton pore of CFo (Wagner

et al., 1989).

The ATP response for PC transport in the presence of a PMF (venturicidin,

light) was virtually the same as the response in the absence of a PMF (venturicidin,

dark). One difference was that more PC was transported in the presence of a PMF.

In both cases transport was saturated by micromolar levels of ATP. If venturicidin

was omitted from the dark reactions such that the proton pumping ATPase was

active, the maximum amount of PC transported was greater than reactions that

contained venturicidin, but much higher levels of ATP were necessary to achieve the

maximum transport. These results demonstrate at least three points. First, a PMF

per se does not influence the quantity of ATP required for the transport reaction.

Second, conditions that permit or promote the ATPase function of the thylakoidal

ATP synthase lead to higher demands for ATP for the transport reaction, possibly

due to depletion of ATP. Finally, the effect of dissipating PMF on transport activity

is a primary effect and not due to secondary effect on changing the ATP requirement

for transport.








Discussion

Our analyses of PC and OE33 transport into isolated thylakoids show that PC

and OE33 transport is ATP-dependent (Figs. 4-4 and 4-6) and is enhanced by

stromal protein factors) (Figs. 4-2 and 4-3) and the trans-thylakoidal proton gradient

(Figs. 4-4, 4-5 and 4-6). The role of stromal factors in OE33 and PC transport has

not been entirely clear. Kirwin et al. (1989) reported that wheat OE33 was

efficiently transported into pea thylakoids in the absence of SE. Mould et al. (1991)

subsequently reported that efficient transport of OE33 into thylakoids required the

presence of SE, but suggested that the intermediate-sized OE33 was the substrate

for thylakoid transport and that it was the processing protease activity of SE that

stimulated OE33 transport. It can be seen from Figure 2 that transport of pea

iOE33 as well as pOE33 into pea thylakoids was greatly stimulated in the presence

of SE. The fact that iOE33 also depends on SE for efficient transport demonstrates

that stromal protein factors) perform some function other than processing. This

agrees with recently published studies of Hulford et al. (1994) that SE is required for

transport of wheat iOE33.

Bauerle and Keegstra (1991) reported that SE inhibited PC transport into

isolated thylakoids. They provided evidence that the full-length PCara precursor was

the active substrate for transport into thylakoids and that the intermediate-sized

forms of PCara were not active in thylakoid transport. They concluded that the

processing activity of SE was responsible for the inhibition of transport in the

presence of SE because the stromal processing enzyme converted some of the PC






74
precursor into intermediate-sized forms of the protein. Our results show that SE

stimulates transport of Arabidopsis PC as well as pea PC. However, there did appear

to be a stromal component that was inhibitory to Arabidopsis PC transport because

elevated levels of SE resulted in less transport than lower levels of SE (Fig. 4-2,

compare lane 2 with lanes 4 and 5). Other studies from our laboratory support the

notion that pPCara is the substrate for transport across pea thylakoid membrane.

When a time course for pPCara chloroplast import and assembly was examined using

a rapid stopping technique, the full-length precursor was the only species within

chloroplasts that exhibited the kinetics expected for a pathway intermediate (Cline

et al., 1992b). Furthermore, it was predominantly pPCara that accumulated in

chloroplasts when thylakoid transport was competed by saturating concentrations of

iOE33. These were not observed for pPCpea, where intermediate-sized species

displayed pathway kinetics (unpublished results of C. Li and K. Cline) and

accumulated during competition studies (Cline et al., 1993).

There has also been some confusion regarding the role of the trans-

thylakoidal proton gradient in PC and OE33 transport. Theg et al.( 1989) and Cline

et al. (1992a) failed to detect effects of ionophores on the localization of PC during

import into intact chloroplasts, whereas Mould and Robinson (1991) reported partial

inhibition of PC localization by ionophores. Earlier, Mould and Robinson (1991)

reported that a proton gradient is required for transport of OE33 into thylakoids.

Yet, recent studies from the same laboratory (Nielsen et al., 1994) failed to detect

stimulation of OE33 transport by the ApH. Previously, Cline et al.(1992a) and






75
Henry et al.(1994) reported that transport of OE33 into thylakoids is partially

inhibited by ionophores during import into intact chloroplasts. Here, we show that

ionophores cause severe reduction (60-70%) of OE33 transport and a small but

reproducible reduction (30-40%) of PC transport with isolated thylakoids (Figs. 4

and 5). We think it likely that due to the fact that ApH is stimulatory in these in

vitro assays, the effect of ApH may not be detected in certain experiments due to

specific methods of conducting the assays. For example, in the studies of Cline et

al. (1992a) and Henry et al. (1994), rapid stopping methods had to be used to avoid

subsequent transport of iOE33 during work-up procedures. Certainly, the relatively

small contribution of a ApH to PC transport together with the fact that it is more

difficult to assess the role of the PMF on thylakoid transport with intact chloroplasts

may explain previous discrepancies regarding the involvement of a PMF in PC

transport.

Even though ApH is only stimulatory to PC and OE33 transport in vitro, it

may play an essential role in their transport in vivo. In bacteria such as E. coli,

signal peptide-bearing proteins are exported by a transport system that depends on

SecA and SecY/E proteins and is powered by the combined action of ATP hydrolysis

and a PMF (Pugsley, 1993). SecA and ATP are essential and necessary for the

initial insertion of the precursor across the bilayer (Schiebel et al., 1991). Under

certain in vitro conditions, SecA and ATP can complete the transport in the absence

of a PMF. Thus, it is no wonder that a wide range of stimulatory effects of PMF has

been reported for bacterial protein export in vitro, whereas PMF appears to be








essential in vivo (Mizushima and Tokuda, 1990).

Because of the endosymbiotic origins of chloroplasts, protein transport into

thylakoids has frequently been assumed to occur by mechanisms similar to those for

protein export in bacteria (Smeekens et al., 1990). In this regard, we notice that

transport of PC and OE33 exhibits similar energy requirements to those of the SecA-

dependent protein translocation system in E. coli. Recent studies have also shown

that PC and OE33 transport is sensitive to azide (Cline et al., 1993; Henry et al.,

1994; Knott and Robinson, 1994; Yuan et al., In preparation), a characteristic

inhibitor of the SecA ATPase activity (Oliver et al., 1990). SecA homologous genes

have been found in the plastid genomes of two algal species (Scaramuzzi et al., 1992;

Valentin, 1992). Although secA homologues have not been found in plastid genomes

of land plants, we have recently detected a polypeptide of ~ 110-KDa in pea

chloroplasts that is immunoreactive with an antibody raised against a conserved

region of an algal chloroplast SecA (Yuan et al., in preparation). The results here,

demonstrating that OE33 and PC share similar requirements for transport across the

thylakoid membrane, are consistent with competition studies that indicate that OE33

and PC share at least one component of the translocation machinery (Cline et al.,

1993). It is possible, but remains to be directly demonstrated that OE33 and PC use

a common SecA-homologous apparatus for thylakoid transport.











CHAPTER 5
ONE OF THREE PATHWAYS FOR PROTEIN TRANSPORT INTO
PEA THYLAKOIDS USES A SECA-DEPENDENT
TRANSLOCATION MECHANISM



Introduction

Because of the endosymbiotic origin of chloroplasts from an ancestor

cyanobacterium, protein transport into plant thylakoids has been speculated to

resemble protein export across the cytoplasmic membrane in prokaryotes (Hartl and

Neupert, 1990; Smeekens et al., 1990). Indeed, evidence suggesting a thylakoid

system homologous to bacterial protein transport has been accumulating. First, the

lumen-targeting domains of thylakoid lumenal precursor proteins are similar to

prokaryotic signal sequences (Von Heijine et al., 1989) and have been shown to

function as such in E. coli (Seidler and Michel, 1990; Anderson and Gray, 1991).

Second, the cleavage specificity of the lumenal peptidase that removes thylakoid

targeting signals is identical to that of E. coli signal peptidase (Halpin et al., 1989).

Third, the energy and soluble factor requirements especially for PC and OE33

transport are very similar to those of the SecA-dependent translocation system (see

the results in chapter 6 of this dissertation).

Azide has been shown to be a specific inhibitor of bacterial protein export due

to its interference with the translocation ATPase of SecA (Oliver et al., 1990). We






78

first reported in our discussion that sodium azide specifically inhibits transport of PC

and OE33 but not of OE23 and OE17 or integration of LHCP into thylakoids (Cline

et al., 1993). Later, Knott and Robinson (1994) confirmed that azide reversibly

blocked PC and OE33 transport but was without effect on OE23 transport. Henry

et al. (1994) demonstrated that azide sensitivity is a pathway specific feature by

analyzing the transport of chimeric proteins. These results suggest that the PC and

OE33 pathway uses a translocation mechanism similar to that of the bacterial SecA-

dependent translocation system. This notion is further supported by the discovery

of secA and secY gene homologues in the plastid genome of red algae (Sccramuzi

et al., 1992a, 1992b; Valetine, 1993). However, previous efforts with antibodies to

E. coli SecA and by use of E. coli SecA to support transport were unsuccessful to

demonstrate a SecA involvement in the transport of proteins into thylakoids.

As an alternative approach to the identification of a SecA homologue in plant

chloroplasts, we overexpressed a highly conserved segment of an algal SecA and

produced antibody to this conserved peptide. This antibody reacts with SecA from

E. coli and also with a pea chloroplast protein of ~ 110 kDa on SDS-PAGE gels,

slightly larger than the E. coli SecA (102 kDa). The chloroplast SecA (CPSecA) is

present in low abundance in chloroplasts and mainly found in the soluble fraction.

Using immunoblotting as an assay, we have purified CPSecA from stromal extract

by a combination of conventional purification methods. Purified CPSecA was able

to support transport of PC and OE33 in an azide-sensitive fashion. In contrast,

purified CPSecA had no ability to support integration of LHCP or to stimulate






79
transport of OE23 and OE17 into thylakoids. Taken together, our studies have

defined one of the pathways for protein transport into thylakoids to be a chloroplast

SecA-dependent translocation pathway. Our studies also provide strong evidence for

conservative sorting of intra-chloroplast proteins. In addition, our strategy to the

identification of CPSecA should be of general use in the search of homologous

proteins for researchers from other field.



Materials and Methods

Chemicals and Expression Plasmids

Sodium azide, phenylmethyl sulfonylfluoride (PMSF), DEAE-sepharose (fast

flow), Sephacryl S-300 HR were from Sigma. Ammonium sulfate was from Fisher.

Hydroxylapatite was from BioRad. Mono Q HR 5/5 and Superose 6 HR 10/30 were

from Pharmacia. Hydropore-HIC hydrophobic interaction column was from Rainin.

Other reagents, enzymes, and standards were all from commercial sources. Plasmid

pMAQ805, which contains the complete coding sequence of the Pavlova lutherii secA

homologue, was kindly provided by Dr. Harold W. Stokes of Macquarie University,

Australia (Scaramuzzi et al., 1992). The in vitro expression plasmids for wheat

pOE33 and iOE33 were provided by Dr. Colin Robinson (Hulford et al., 1994). The

in vitro expression plasmids from pea and Arabidopsis were described earlier (Cline

et al., 1993; Yuan and Cline, 1994).






80
Cloning of A Conserved Segment of SecA into An E. coli Overexpression Plasmid

SecA protein sequences from P. lutherii (Pvlcpseca), Antithamnion spp

(Asplas), E. coli (Ecoseca) and Bacillus sutilis (Bacsubseca) were retrieved from the

GenBank database. The sequences were aligned using the GCG program PILEUP

(Devereux et al, 1984). A highly conserved region was identified by visual inspection.

A clone of the P. lutherii chloroplast secA was obtained (Scaramuzzi et al., 1992).

The DNA fragment containing the conserved region from bp 1043 to 1458 was

amplified by polymerase chain reaction (PCR) with a 5' primer 5'-GCTCCACCATA-

TGAAAATCGCCGAGATGAAGACAGG-3' containing an in-frame NdeI site and

a 3' primer 5'-GGAATGTTTCAAGCTITCGGGAGATTATTAGTGG-3' containing

a HindIII site to allow in-frame fusion of the 3' His-6 tag in pET24b. After digestion

with NdeI and HindIII, the PCR product was ligated with NdeI and HindIII digested

pET24b vector by standard procedures (Crowe et al., 1991). Three positive clones

designated pJYsecA were obtained and confirmed to be the same with no error in

their DNA sequence. For expression of JYsecA peptide (~ 16-kDa), pJYsecA was

introduced into E. coli BL21(ADE3).



Expression of JYsecA in E. coli and preparation of anti-JYsecA antibody

Cells harboring pJYsecA were grown in LB medium to an OD at 600 nm of

~ 1.0. Isopropyl thio-p-D-galactoside (IPTG) was added to 2 mM and the cells were

allowed to grow for another 4 hrs at 37C. Cells were harvested by centrifugation

and checked for expression of JYsecA. After induction, JYsecA accounted for about






81
5% of the E. coli protein and was found to be in inclusion bodies. Inclusion bodies

were isolated as previously described (Cline et al., 1993). Inclusion bodies were

solubilized in 8 M urea and JYsecA purified on a nickel column equilibrated with

6 M urea according to instructions provided by the manufacturer. For antibody

preparation, JYsecA was further purified by SDS-PAGE and electroeluted from

excised gel band. Antibody to SDS-denatured JYsecA was prepared by Cocalico

Biological Chemical Cooperation (PA).



Purification of CPSecA

CPSecA was purified from SE by a combination of ammonium sulfate

precipitation, DEAE-sepharose ion exchange, hydroxylapatite selective adsorption,

Sephacryl S-300 gel filtration, Mono Q anion exchange and Hydropore hydrophobic

interaction chromatographies. Fractions were monitored for CPSecA with antibody

to JYsecA. For ammonium sulfate precipitation, 30 grams of ammonium sulfate

solid was added to 260 mls of 2X SE (about 1.3 grams of protein) for the first cut

(20% saturation) to precipitate membrane vesicles and denatured proteins. After

removing the precipitated material by centrifugation, another 33 grams of ammonium

sulfate solid was added to the supernatant to reach a 40% saturation of ammonium

sulfate that is necessary to precipitate CPSecA protein. The precipitated proteins

were recovered by centrifugation and dissolved in Buffer A (25 mM Hepes/KOH, pH

8, 50 mM KCI, 5 mM MgCl2) plus 1 mM PMSF. The sample (120 mls) was then

applied to a 90 ml DEAE-sepharose column equilibrated with Buffer A. The column






82
was washed with 200 mls of Buffer A, followed by elution with 500 mls of a salt

gradient (50-350 mM KCI). Fractions (5 ml) were immunoblotted with anti-JYsecA

and those with CPSecA were pooled for hydroxylapatite chromatography. An equal

volume of Buffer A was mixed with the pooled sample (25 mls) and then applied to

a 45 ml column of hydroxylapatite equilibrated with 10 mM potassium phosphate

(pH 7). After wash with 100 mls of 10 mM potassium phosphate (pH 7) and 100 mls

of 1 M KCI (unbuffered), the column was eluted with 300 mls of a gradient of

potassium phosphate (10-300 mM). Fractions (3 ml) were immunoblotted for the

presence of CPSecA and those with CPSecA were pooled and concentrated to 1 ml

by centrifugation through Centriprep-10 (Amicon, Inc.). The concentrated sample

was then applied to a 80 ml Sephacryl S-300 gel filtration column equilibrated with

gel filtration buffer (20 mM Hepes/KOH, pH 8, 65 mM KCI, 1 mM DTT, 1%

ethylene glycol). Fractions (1.5 ml) were blotted for the presence of CPSecA and

those with CPSecA were pooled for anion exchange on a 1 ml Mono Q column

equilibrated with starting buffer (20 mM Hepes/KOH, pH 8, 50 mM KCI). After

wash with 10 mls of starting buffer, the column was eluted with 30 mls of a salt

gradient (50-300 mM KCI). Fractions (0.5 ml) were stained for the presence of

CPSecA and those with CPSecA were pooled and adjusted to PAS buffer (50 mM

potassium phosphate, pH 7, 1.5 M ammonium sulfate) for hydrophobic interaction

on a 7.85 ml polyethylene glycol matrix Rainin's Hydropore-HIC column equilibrated

with PAS. After wash with 20 mls of PAS, the column was eluted with 50 mls of a

descending salt gradient (1.5-0 M ammonium sulfate).








Use of Sodium Azide

Sodium azide was prepared in aqueous stock at 0.6 M in import buffer.

Chloroplasts or thylakoids were incubated with sodium azide for 10 min at 25C in

light (70 1iE.m2s-1) prior to the addition of precursor to assay mixture. Sodium azide

was used at the concentrations specified in the text and figure legends.



Miscellaneous Methods

Chloroplasts were isolated from 9- to 10-day-old pea (Laxton's Progress 9)

seedlings as described (Cline, 1986). Chl concentrations were determined according

to Arnon (1949). Chloroplast lysate, SE and thylakoids were prepared as described

(Cline et al., 1992; Yuan et al., 1991). Assays for import into chloroplasts or for

integration/transport into thylakoids were carried out as previously described (Cline

et al., 1993; Yuan and Cline, 1994). Sample analysis was carried out according to

Cline (1986). Immunoblotting was carried out as described (Payan and Cline, 1991).

Protein assays were performed by the Bradford method (Bradford, 1976) using

bovine serum albumin as a standard.



Results and Discussion

Effects of Azide on Protein Transport into Thylakoids

Azide is a specific inhibitor of protein export from bacteria (Oliver et al.,

1990). To examine whether protein transport into thylakoids is analogous to protein

export from bacteria, we investigated the effects of azide on transport or integration
















A B
P1234 120
pPCpI -
pOE3 ...a P
M Se. -m12 "8


pOE17 40
4Wlib d1 dew -0 20-
pOE23 :T -
ini 00 2 4 6 a 10
INNal(mM) 0 2 510 [Sodium AzMe] (mM)






Figure 5-1. Effect of azide on protein transport into thylakoids
assessed with intact chloroplasts during import (A) or with isolated
thylakoids during transport (B). Radiolabeled precursors (p) of PC,
OE33, OE23 and OE17 from pea were assayed for import and
subsequent localization with intact chloroplasts in the absence or
presence of increasing concentrations of sodium azide. Accumulation
of the mature forms (m) of the proteins indicates that the imported
proteins have been transported into thylakoids. Accumulation of
intermediate species (i) indicates that thylakoid transport of the
imported proteins was inhibited. Import assays were terminated with
HgC12 (Reed et al., 1990). Transport of PC (0), OE33 (A), OE23 (l)
and OE17 (A) from pea and PC (@) from A. thalliana and integration
of pLHCP (0) from pea into thylakoids were carried out with
chloroplast lysates in the absence or presence of increasing
concentrations of sodium azide.






85

of proteins into thylakoids. Figure 5-1A shows the results of an import experiment

carried out in the presence of increasing concentrations of sodium azide. Import of

proteins into chloroplasts was demonstrated by processing of the precursors to their

mature or stromal intermediate forms and by co-purification with intact chloroplasts.

Import of all four lumenal proteins (PC, OE33, OE23, and OE17) into chloroplasts

was not significantly affected by sodium azide at concentrations used in the

experiment. However, transport of the imported PC and OE33 into thylakoids was

severely inhibited in the presence of sodium azide, leading to the accumulation of

PC and OE33 stromal intermediates. In contrast, the localization of OE23 and

OE17 was unaffected.

The effect of azide on transport of PC, OE33, OE23, and OE17 into the

thylakoid lumen and on integration of LHCP into the thylakoid membrane was also

assessed in a reconstituted reaction with isolated thylakoids (Fig. 5-1B). The results

showed that while transport of PC and OE33 was inhibited in the presence of sodium

azide, integration of LHCP and transport of OE23 and OE17 were not affected by

sodium azide. The fact that azide does not interfere with LHCP integration or with

OE23 and OE17 transport is consistent with competition results that they utilize

different pathways for transport into thylakoids. In bacteria, azide has been shown

to target exclusively to the SecA protein and its inhibitory effect on protein transport

is attributed to its interference with the SecA ATPase activity (Oliver et al., 1990).

The fact that sodium azide specifically inhibits PC and OE33 transport suggests that














A (kMD) 1 2 3 4 5 6

204-
132-
65.0-

42.6-

B (k 12) 1234567

97.4- m
66.2-


31.0- .
21.5- -
S 12 34567





Figure 5-2. Detection and purification of a SecA homologue from pea
chloroplasts. (A) Immunoblotting of E. coli and chloroplast protein
with an antibody to a conserved peptide deduced from an algal secA
gene "Materials and Methods". Lane 1, total E. coli protein from a
wild type strain; lane 2, soluble extract of E. coli cells harboring the
SecA-overexpressing plasmid pT7-secA (Schmidt and Oliver, 1989);
lane 3, total chloroplast protein; lane 4, stromal protein; lane 5,
thylakoid protein; lane 6, total envelope protein. Total envelope
membranes were isolated according to Cline (1985). (B) Protein
profile of the Coomassie blue-stained gel of the CPSecA-containing
fraction from each purification step. Lane 1, total stromal protein
(12.5 gig); lane 2, ammonium sulfate precipitation (11.5 gig); lane 3,
DEAE-Sepharose ion exchange (3.5 jig); lane 4, hydroxylapatite
chromatography (5 ig); lane 5, Sephacryl S-300 gel filtration (1.75 jig);
lane 6, Mono-Q anion exchange (1 pg); lane 7, Hydropore
hydrophobic interaction (2 pg). (C) An immunoblot of the samples
shown in the Coomassie blue-stained gel.







87

they are transported by a similar mechanism to that of the bacterial SecA-dependent

translocation system.



Identification and Purification of a SecA Homologue from Pea Chloroplasts

A crucial step to demonstrate the involvement of a SecA type of system for

the transport of proteins into thylakoids is to identify a SecA homologue and to

demonstrate its involvement in protein transport into thylakoids. As antibody against

the entire SecA protein from E. coli failed to react with anything from plant

chloroplasts (Knott and Robinson, 1994), our strategy was to make an antibody to

a conserved segment of SecA and use that antibody to probe chloroplast proteins.

For this, protein sequences from all known SecAs (2 from bacteria and 2 from algal

chloroplasts) were aligned and compared for similarities. Eight conserved regions

were identified and 4 of them are clustered in a relatively small segment (128 amino

acid residues) of the SecA protein. This segment, from residue 95 to 222 of the P.

lutherii chloroplast secA gene, contains sites identified in E. coli for ATP binding and

azide sensitivity suppression (Oliver et al., 1990), and a site in B. subtilis SecA

necessary for functional complementation of E. coli SecA mutants (Klose et al.,

1993).

Antibody raised against this peptide reacted with SecA from E. coli (Fig. 5-

2A, lanes 1 and 2) and also with a protein of ~ 110 kDa (Fig. 5-2A, lanes 3-6) from

pea chloroplasts. The chloroplast SecA (CPSecA) was found mainly in the stromal

fraction of the chloroplast, but was present at low levels in the thylakoid and







88








P1 2 3 4
ppc- -
pPC
pOE33- ow
mOE. *poE
mOE3- mm
pLHCP- *
LHCP-DP- pLHoP

pOE17- -
mOE17- pOE17
IOE23- --
mOE23- *_M







Figure 5-3. Reconstitution of protein transport with purified CPSecA
and isolated thylakoids. Radiolabeled full-length or intermediate-sized
precursor proteins were mixed with thylakoids and assayed for
transport or integration with total stromal protein (315 jig/assay, lane
1), in the absence of soluble protein (lane 2), with purified CPSecA
(1.2 Ig/assay, lane 3), or with a stromal fraction from gel filtration that
lacked CPSecA (26 jig/assay, lane 4). Purified CPSecA was prepared
for assay by buffer exchange on a Superose 6 gel filtration column.
Assays were conducted as described (Yuan and Cline, 1994) except
that each received 4 mM ATP and 4 mM GTP. All assays were 75 pl
and contained the same buffer composition. Radiolabeled proteins
used in this experiment were pPC fromA. thalliana, pOE33 and iOE33
from wheat (Hulford et al., 1994), pLHCP, pOE17 and iOE23 from
pea.







89

envelope membrane fractions. By using immunoblotting as an assay, CPSecA was

purified from SE by a combination of conventional chromatographies (see 'Materials

and Methods' for purification details). Like the bacterial SecA, CPSecA exists as a

dimer in solution, migrating on gel filtration columns with a size of 200-25 kDa. The

purity of CPSecA from each step is shown in Figure 5-2B. Immunoblotting of the

pooled sample from each step is shown in Figure 5-2C. From a dilution series it is

estimated by immunoblotting that about 0.4% of the stromal protein is CPSecA.

Thus, CPSecA was enriched approximately 250 fold.



Purified CPSecA Is Able to Support Transport of PC and OE33 into Thylakoids

Purified CPSecA was tested for its ability to stimulate transport of PC and

OE33 into thylakoids. A gel filtration fraction that has much more stromal protein

but contains no CPSecA was used as a control to demonstrate specificity of CPSecA

in supporting transport. The results showed that purified CPSecA was able to

replace SE in supporting transport of PC and OE33 into isolated thylakoids (Fig. 5-

3). CPSecA, on the other hand, showed no activity to support LHCP integration into

the thylakoid membrane or to stimulate OE23 and OE17 transport across the

membrane into the lumen. Transport of PC and OE33 into thylakoids increases with

increasing concentrations of CPSecA protein and approaches saturation at about 150

nM (1.25 pg/75 iil assay) of CPSecA monomer (Fig. 5-4A). Sodium azide inhibits

CPSecA-dependent transport of PC and OE33, demonstrating that CPSecA is the

azide sensitive component (Fig. 5-4B).







90












A SE CPSecA
P 1 2 3 4 56 7 8




p/amay 125 0 01 0.S6 1 1.25 1.6
B SE cPSe
pi II r -
pPC-


pOE33- ,
azdMe(mM): 0 2 5 10 0 2 6 10







Figure 5-4. CPSecA-supported transport of PC and OE33 is CPSecA
concentration dependent (A) and sensitive to inhibition by sodium
azide (B). Radiolabeled pPC and pOE33 were assayed for transport
into isolated thylakoids in the presence of total stromal protein (SE,
125 pg/assay) or varying amounts of CPSecA (pig/assay). Sodium azide
at concentrations of 2, 5 and 10 mM was included in transport assays
(75 Il) supported by SE protein (315 pg/assay) or CPSecA (1.2
pg/assay) as described in "Materials and Methods". The precursors
were pPC from A. thalliana and pOE33 from wheat. All assays were
75 pl.








Conclusions

In this study, we have defined one of the pathways for protein transport into

thylakoids to be a chloroplast SecA-dependent translocation pathway. Proteins

transported on this pathway include PC and OE33, both of which are also present

in cyanobacteria, the progenitors of chloroplasts. Suggestive evidence for the

operation of a SecA type of translocation mechanism in plant chloroplasts for protein

transport into thylakoids was the observation that azide, which specifically inhibits

SecA-dependent protein export from bacteria, inhibits PC and OE33 transport but

not OE23 and OE17 transport or LHCP integration (Fig. 5-1 in this work; Cline et

al., 1993; Knott and Robinson, 1994; Henry et al., 1994). Direct evidence comes

from the identification of a SecA homologue from pea chloroplasts (Fig. 5-2) and the

demonstration that CPSecA alone is able to support transport of PC and OE33 into

isolated thylakoids (Fig. 5-3) in a concentration-dependent and azide-sensitive

manner (Fig. 5-4). Our studies are the first to identify and isolate an essential

component for chloroplast protein transport in vitro.











CHAPTER 6
SUMMARY



The research presented in this dissertation has gone a long way in advancing

our understanding of protein targeting and transport into thylakoids. The studies

described here made headway in revealing new information and principles about

thylakoid protein targeting and transport; they clarified confusions in the literature

regarding transport requirements and the role of soluble factors in protein import

into chloroplasts and integration into thylakoids; they resulted in the identification

and purification of an essential component of the thylakoid translocation machinery.

The key to our success was the use of chemical quantities of purified

precursor proteins. With purified precursor proteins, we were able to show that the

stromal factor required for LHCP integration plays an active role in the integration

reaction, most likely in targeting LHCP to the membrane or in folding LHCP into

the bilayer. The use of large amounts of purified precursor proteins additionally

allowed us to discover that import of proteins into the chloroplast stroma is much

faster than transport of proteins into thylakoids. This unexpected observation

allowed us to develop a novel in organello competition assay that lead to the

discovery of multiple pathways for protein transport into thylakoids.






93
One of the questions raised by the discovery of multiple pathways was whether

transport requirements are characteristic of translocation pathway or of translocated

precursor protein. The studies in chapter 5 show that transport requirements are

characteristic of translocation pathway and not of translocated precursor protein.

A long unanswered question regarding protein transport into thylakoids was

whether chloroplasts use a mechanism for protein transport into thylakoids that is

homologous to prokaryotic protein transport. This question was narrowed by the

discovery of multiple pathways down to the question of whether any of the three

pathways described for protein transport into thylakoids is analogous to the bacterial

SecA-dependent translocation system. Our studies were culminated with the

identification of a SecA homologue from pea chloroplasts and the demonstration

that purified chloroplast SecA is able to promote protein transport into isolated

thylakoids.