II: oil
.1315"!

. Silil}
shat-(5.1.. n

L V
.. .ihzwumlfi 4"
. .t . .1

at

Effidtsiaflllciig z.
.6"). r: . ....!..as..\azNu

.513!!! ii: . I).

c
It... in
397:

\s!.h.

2"
N3 n."
‘ d I
z.....f%z..v...-
.331}. 1..
.. .5332
52);). ,1
it?! .2. 9....
‘ with... . ,1 . z
3 2; . .s
1315i...

 

is... tlZ-aP

[,k

flf‘

IO
1
I.
I

 

LIBRARY
Micruyan State
University

 

 

 

This is to certify that the
dissertation entitled

SECRETION SIGNALS AND CHAPERONE FUNCTION IN
ERWINIA AMYLOVORA

presented by

LINDSAY R. TRIPLETT

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Plant Pathology

 

 

45%

 

Major Professor’s Signature
/ 2 - ’7“ — 0 <7

Date

 

MSU is an Affinnative Action/Equal Opportunity Employer

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

p

 

 

 

 

 

5108 KzlProleocaProleIRCIDateDmhdd

SECRETION SIGNALS AND CHAPERONE FUNCTION IN ER WINIA AMYLOVORA
By

Lindsay R. Triplett

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

2009

ABSTRACT
SECRETION SIGNALS AND CHAPERONE FUNCTION IN ER WINIA AMYLOVORA
By
Lindsay R. Triplett

Like many gram-negative bacterial pathogens, the plant pathogen Erwinia
amylovora requires a type III secretion system (TBSS) to establish infection in the host.
This work investigates the factors required for type III secretion of the E. amylovora
pathogenicity factor DspE and for its interaction with the specific chaperone DspF. DspE
N-terminal fragments fused to the CyaA translocation reporter were used to identify a
minimal secretion and translocation signal within the first 51 amino acids, and an optimal
translocation signal within 150 N-terminal amino acids. Mutagenesis of portions of the
translocation signal revealed that residues 2-10 were required for protein translocation,
and that mutagenesis of a segment containing a highly conserved stretch of 5 serines led
to diminished translocation levels. Yeast two-hybrid assays, in-vitro pull-down assays,
and site-directed mutagenesis were used to map the chaperone binding domain of DspE
to residues 51-100, with a crucial role for residues 71-85, although the DspE-specific
chaperone DspF and the N-terminal chaperone binding domain were not required for
translocation of an N-terminal DspE-CyaA fusion.

Type III secretion chaperones have a highly conserved tertiary structure, but few
analyses have been completed to determine which residues are critical for which 1
function. One site postulated, but never tested, to be important in chaperone-effector
interaction is the helix-binding groove. A homology-based model of DspF was used to

identify 11 amino acids with putative localization in the helix-binding groove. Site-

directed mutagenesis of DspF identified four highly-conserved residues required for
virulence in DspF, and three of these residues were required for interaction with the N-
terrninus of DspE. This study supports the significance of polar residues of the helix-
binding groove in chaperone-effector interactions.

T3SS chaperones may interact with multiple substrates in the bacterial cell,
including regulatory proteins, multiple effectors, and components of the secretion
apparatus. A directed yeast two-hybrid screen identified E. amylovora secreted proteins
Eopl and Eop3 and homologs to the type HI secretion apparatus proteins Hch and Hch
as candidate interactors of DspF. Annotation of the unpublished Erwinia genome
confirmed or revealed the presence of putative chaperone genes next to eopI and eop3,
but not other effectors such as eop4. Deletion of dspF positively affected CAMP
accumulation mediated by translocation of Eopl- and Eop3-CyaA, but not Eop4-CyaA.
Deletion of the putative Eopl chaperone, escI , negatively affected translocation of Eopl-
CyaA but not of DspE-CyaA, indicating that Escl chaperone action is highly specific.
Deletion of Eopl, Eop3, and the corresponding putative chaperone genes had no
significant effect on virulence, although Eal l89AdspFAesc1 showed slightly increased

levels of early growth on pear compared with Ea1189AdspF.

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. George Sundin, for giving me the
opportunity to study in this field, for his continual support and positive attitude, and for
being an all-around great mentor. I would also like to thank the members of my
committee, Drs. Frances Trail, Cindy Arvidson, and Sheng Yang He, for their helpful
advice and comments.

Many thanks to my many former and current co-workers in the Sundin lab, for
giving advice and always making the job a little more fun; to my parents, for always
believing in me, and to my wonderfiil husband Preston, whose love and friendship helped

me all the way.

iv,

TABLE OF CONTENTS

List of Tables ............................................................................................................. vii
List of Figures ............................................................................................................ viii
Chapter 1: Literature Review .................................................................................. 1
I. Signals targeting proteins for bacterial secretion
A. Type I, H, IV, and V secretion signals and chaperones .................................. 1
B. Type III secretion system ................................................................................ 2
C. Type III secretion signals
i. The N-terminus ............................................................................ 3
ii. Downstream signals that influence secretion and translocation. 4
iii. Characteristics of the chaperone binding domain ........................ 5
11. Class I type HI secretion chaperones
A. Definition and overview ................................................................................. 7
B. General structural features of the chaperone-effector interaction ................... 7
C. Substrate specificity of T3 SS chaperones ....................................................... 8
D. The roles of TBSS chaperones
i. Chaperones dock effector proteins to the secretion apparatus ...... 9
ii. Chaperones confer secretion specificity ........................................ 10
iii. Chaperones stabilize or prevent aggregation of cognate effectors 12
iv. Chaperones may regulate the expression of cognate effectors ..... 13
v. Chaperones may impose a temporal hierarchy of secretion .......... 14
III. Chaperone-effector interaction in Erwinia amylovora
A. Erwinia amylovora biology and economic significance ................................. 16
B. The secreted effector DspE ............................................................................ 17
C. The T3SS chaperone DspF ............................................................................. 18
IV. Rationale and project goals ................................................................................... 19
Works Cited .......................................................................................................... 24

Chapter 2: Functional analysis of the N-terminus of the Erwinia amylovora
secreted effector DspA/E reveals features required for secretion,
translocation, and interaction with the chaperone DspB/F.

Abstract ....................................................................................................................... 33
Introduction ................................................................................................................. 34
Materials and Methods ................................................................................................ 37
Results ......................................................................................................................... 41
Discussion ................................................................................................................... 58
Tables .......................................................................................................................... 63
Works Cited ................................................................................................................ 67

Chapter 3: Homology-based modeling of the Erwinia amylovora type III
secretion chaperone DspF used to identify amino acids required for virulence
and interaction with the effector DspE

Abstract .......................................................................................................................
Introduction .................................................................................................................
Materials and Methods ................................................................................................
Results .........................................................................................................................
Discussion ...................................................................................................................
Tables ..........................................................................................................................
Works Cited ................................................................................................................

Chapter 4: A Study of the Roles of DspF and other Class I Secretion
Chaperones in Erwinia amylovora in Effector Translocation and Virulence
Abstract .......................................................................................................................
Introduction .................................................................................................................
Materials and Methods ................................................................................................
Results .........................................................................................................................
Discussion ...................................................................................................................
Tables ..........................................................................................................................
Works Cited ................................................................................................................

Chapter 5: Conclusions and Future Directions
Summary of Work .......................................................................................................
Questions for the Future .............................................................................................

Appendix 1: Genetic Differences Among Blight-Causing Erwinia Species with
Differing Host Specificities Identified by Suppression Subtractive
Hybridization .............................................................................................................

vi

72
73
75
78
90
97
100

104
105
108
110
128
133
136

140
145

147

List of Tables

Table 2-1. Strains and plasmids used in Chapter 2 ..................................................... 43
Table 2-2. Oligonucleotides used in Chapter 2 ............................................................ 45
Table 3-1. Bacterial strains and plasmids used in Chapter 3 ....................................... 97
Table 3-2. Oligonucleotide primers used for mutagenesis of dspF ............................. 99
Table 4-1. Bacterial strains and plasmids used in Chapter 4 ....................................... 133
Table 4-2. Primers for cloning and mutagenesis used in Chapter 4 ............................ 134

Table 4-3. Yeast two-hybrid identification of putative DspF-interacting proteins ...... 135

Table A-1. Erwinia spp. strains and their indigenous plasmid content ....................... 165
Table A-2. Libraries generated by suppressive subtractive hybridization .................. 166
Table A-3. Sequence analysis of SSH inserts specific to E. amylovora Eal 10 ........... 167

Table A-4. Sequence analysis of SSH inserts specific to Ejp556 and E. pyrifoliae.... 169

Table A-5. Analysis of the distribution of gene sequence among blight-causing
Erwinia strains .............................................................................................................. 170

vii

List of Figures

Figures in this dissertation are presented in color.
Figure 1-1. Variations in the binding pattern and roles of T3SS chaperones ...........

Figure 2-1. Pathogenicity and cell death phenotypes of Erwim'a amylovora
Eal 189, Ea1189AdspF, and Ea1189AdspF (pLRT198) ............................................

Figure 2-2. Kinetics of CAMP accumulation after infiltration into tobacco leaves

of Eal 189MspE + DspE (1-203)-CyaA .....................................................................

Figure 2—3. Mapping of the secretion and translocation domain of DspE .................

Figure 2-4. Analysis of alanine replacement mutants of amino acids 2-40 of DspE.

Figure 2-5. Mapping of the DspF-interacting regions in the N—terminal half of

DspE ............................................................................................................................

Figure 2-6. Alanine-replacement mutagenesis of the N-terminal chaperone-
binding domain of DspE .............................................................................................

Figure 2-7. Functional analysis of amino acids 51-100 of DspE ...............................

Figure 3-1. Structural modeling and site-directed mutagenesis of dspF ...................
Figure 3-2. Virulence phenotype of site-directed mutants ........................................

Figure 3-3. Alanine replacement of predicted helix-binding groove residues of
DspF disrupts the interaction with the N-terminus of DspE in yeast .........................

Figure 3-4. Site-directed mutants cause moderate reductions in virulence and
DspE (1-737)-CyaA translocation when expressed in wild-type E. amylovora .........

Figure 4-1. DspE-6xHis is detected in Ea1189AdspF and partially complements
the dspF pathogenicity phenotype ..............................................................................

Figure 4-2. Gene organization and interactions of putative chaperone-effector
pairs in E. amylovora ..................................................................................................

Figure 4-3. CAMP accumulation in tobacco leaves mediated by E. amylovora
expressing CyaA fusion proteins ................................................................................

Figure 4-4. CAMP accumulation in tobacco leaves mediated by E. amylovora
deletion mutants expressing Eopl and DspE-CyaA fusion proteins ..........................

Figure 4-5. Bacterial population counts on immature pear fruit inoculatied with
Ea1189 and mutant derivatives ...................................................................................

viii

23

43

45
47

50

53

55
57
81

84

86

90

114

120

124

125

127

Figure A-l. Symptom expression and bacterial growth in immature pear fruit
following inoculation with Erwinia amylovora strains ............................................... 150

ix

Chapter 1: Literature Review

Bacteria secrete proteins and peptides in order to break down substrates, acquire
nutrients, antagonize competitors, or to modify the physiology of hosts or symbionts.
Protein transport across two cell membranes and a peptidoglycan layer presents a special
challenge, and gram-negative bacteria have evolved at least eight unique protein export
systems. At least three of these systems, including the type HI secretion system (T3SS),
transfer proteins directly into eukaryotic cells. This chapter discusses the nature of the
type HI secretion signal, the role of chaperones in type IH secretion, and the T3SS in
Erwim'a amylovora.

I. Signals targeting proteins for bacterial secretion.

IA. Type I, H, IV, and V secretion signals and chaperones. Signals targeting bacterial
proteins for export vary greatly in their nature and complexity. For example, ABC
transporter (Type I) secreted proteins are targeted for secretion by non-cleaved, highly
variable C-terminal secretion signals rich in oz-helices (25). For secretion through the Sec-
dependent pathway known as type II secretion (as well as the type IV pilus system), a
protein is first targeted to the general secretory pathway by an N-terminal signal peptide
which is cleaved upon entry into the periplasm, then targeted for extracellular secretion.
Attempts to define the protein elements required for TZSS targeting have been met with
very limited success, and have suggested that the secretion signal is highly complex and
may be bipartite or dependent on folding (72). Proteinaceous substrates of the type IV
secretion system generally have no conserved signal sequence, although the C-terminus is
required for type IV secretion in several secreted substrates of A. tumefaciens (27).

Specific chaperones are required for secretion of at least some type IV secretion

substrates, and these bind close to the signal sequence (66). Proteins secreted through the
autotransporter (Type V) secretion pathway require an N-terminal signal peptide for Sec-
dependent secretion into the periplasm before exiting the cell via the B—barrel membrane
pore in the C-terminus (86).

1B. The Type III Secretion System. The T3SS is a conserved syringe-like apparatus
spanning the inner and outer membrane. Roughly 25 proteins are required to build the

T3 SS, including nine core apparatus proteins universally conserved among plant
pathogens (22). Three general components of the apparatus are required for protein
translocation into host cells: 1). The base, which spans the inner and outer membrane, is
composed of a set of inner membrane rings facing the cytosol, a central rod traversing the
periplasm, and a set of outer membrane rings. A number of peripheral inner membrane
proteins associate with the inner membrane ring. 2). The needle or pilus is a filamentous
macromolecular structure of secreted proteins extending from the bacterial outer
membrane rings to the eukaryotic host cell. 3). The translocation apparatus, or translocon,
is a structure that may bridge the tip of the pilus to the host cell membrane and create a
pore allowing the entry of bacterial proteins. Substrates are most likely unfolded by the
inner membrane ATPase as they enter the secretion system before traversing the pilus
and refolding in the host cytoplasm (37).

The T3SS is critical for pathogenesis in a wide variety of animal and plant
pathogenic bacterial species (22). Phylogenetic analyses have divided the secretion
systems into seven families. Plant pathogens in the genera Erwinia and Pseudomonas
encode secretion systems of the hrpl family; the regulatory mechanisms controlling hrpl

systems differ substantially from the hrp2 family secretion systems found in

Xanthomonas and Ralstonia (80). Although many components of the core secretion
system and secretion mechanism are conserved among all T3 SS, the apparatus in plant
pathogens share a few key adaptations for expression and infection within the plant
system. For example, hrpl systems feature a needle of extended relative length,
presumably required to traverse the thick plant cell wall. Plant pathogen systems also lack
orthologs to many known translocon proteins in Salmonella and Yersinia, suggesting a
divergent method of cell entry (19). Despite major advances in the past 15 years,
relatively little is known about how proteins are targeted, loaded, and injected fiom the
T3SS into the host cytosol, especially when the host is a plant. While many exciting
recent findings have been made regarding the regulation, assembly, and mechanism of
the bacterial T3SS, the discussion here will focus on substrate targeting and the role of
effector-specific chaperones in secretion.

IC. Type III secretion signals.

i. The N-terminus. There is no conserved signal sequence targeting proteins for type III
secretion. The N-terminal secretion domain of type III-secreted proteins do share some
common features, such as a hydrophobic third amino acid and an amphipathic, serine rich
N-terminus (8, 43). However, these characteristics are not definite indicators that an
expressed protein will be secreted by the T3SS (36). The N-terminal 15 amino acids are
usually sufficient for secretion of an effector protein into culture media (6, 46), and that
the first seven amino acids are especially important in this process (69, 73). While some
authors have recorded a very small amount of secretion in the absence of the N-terminal
signal (21), most reports concur that the N-terminal secretion domain is absolutely

required (17, 49).

Extensive mutational analyses of the N-termini of secreted proteins in Yersim'a
have failed to identify a single sequence feature absolutely required for secretion (52, 69).
The DNA and protein sequences of the N-terminal 15 amino acids are highly tolerant of
synonymous mutations and frameshift mutations (6, 46, 68), and even a series of
synthetic sequences encoding amphipathic N-termini were sufficient to confer secretion
to a reporter fusion (53). The ability of the N-terminus to tolerate frameshifi mutations
caused some debate about whether the N-terminal secretion signal is an RNA signal or
proteinaceous (52, 53, 68), but biochemical and microscopic studies on secretion signal-
reporter gene fusions have confirmed the secretion-competence of pre-synthesized
effector proteins (29, 52, 91). It is possible that an amphipathic, serine-rich, disordered N-
terminus is sufficient for the type IH secretion of any protein, although it is not intuitive
that such a general signal would be sufficient for selection of a few proteins among
thousands. Whatever the nature of the secretion signal, it appears to be universal, as the
Erwinia chrysanthemi and Xanthomonas campestris T3 SSS can secrete Yersinia effectors,
and the Yersinia secretion apparatus secretes P. syringae effectors AvrB and Ath0 (7,
70).
ii. Downstream signals that influence secretion and translocation.

The studies of the N-terminal secretion signal discussed above relied heavily on
analyses of the supernatants of bacterial cultures, but the observations gained from these
studies were not always consistent with the requirements for translocation of the native
effector protein into host cells. For example, while the first 15 amino acids of YOpE can
be secreted into the culture medium, 50 amino acids were required for translocation into

host cells, but only in the presence of a YopE binding chaperone protein, Sch (l 7, 77).

This and similar observations led to the proposal of a bipartite translocation signal- an N-
terrninal signal necessary for secretion, followed by a downstream sequence necessary for
translocation (48).

Why would a secretion signal be sufficient for secretion into the medium, but not into
host cells? How can the two processes be separated? One possible explanation is that the
secretion conferred by the N-terrninal signal sequence is mediated by an alternate
secretion system. In other words, the N-terminal sequence could be an ancient secretion
signal sufficient for secretion by the flagellar secretion system, but not by the more
recently evolved T3 SS (48). A second possibility is that requirements for secretion are
different in the artificial medium environment than in the host. Although Cheng and
coworkers reported that the N-terminal 15 amino acids of YopE was secreted from a
Yersinia enterocolitica strain (21), another study found no trace of YopE in the culture
media during HeLa cell infections (49). It is also possible that some unknown signal is
transmitted upon contact with host cells, changing the requirements for entering the T3SS
and preventing translocation. In summary, a secretion signal longer than the N-terminal
minimal secretion signal is often required for effector translocation into host cells.
Although the reason for this is not yet clear, the longer translocation signal often
coincides with the domain which interacts with a type III secretion chaperone (48, 77).
iii. Characteristics of the chaperone binding domain.

Chaperone-binding T3SS-secreted effectors characterized thus far all have a
single chaperone-binding domain (CBD), although at least one effector binds two
chaperone dimers in this region (2). The CBD is almost always found close to the N-

terminus of the effector between amino acid residues 50-200 (51). Notable exceptions are

members of the Pseudomonas syringae effector HopOl-l family, which has a CBD in the
middle third of the protein, and HpaB, which binds to the secreted effector HpaA at its C-
terminus (42, 56).

The CBD is typically required for protein translocation into host cells (28, 32, 50, 92).
However, the chaperone binding domain can exert an inhibitory effect on secretion in the
absence of the chaperone. Yersinia YopE and SicP from Salmonella were both secreted
without the chaperone present, but only after fragments of the chaperone binding domain
had been deleted (17, 48). The CBD was not required for some functions of YopE. The
CBD may confer instability to the effector protein when not stabilized by a chaperone; in
fact, several secreted substrates were found to aggregate and destabilize in the absence of
the chaperone (17, 32). The CBD of the Yersinia enterocolitica effector YopO doubles as
' a membrane localization signal, providing further evidence that chaperones may bind
here to prevent aggregation (50). However, in a study of three P. syringae effectors, the
domain conferring chaperone-dependent stability was found to be separate from the
chaperone binding domain itself, suggesting that the function of the CBD is varied among
effectors (5 8). A second proposed explanation for the apparent secretion-inhibitory role
of the chaperone-binding domain is that the chaperone binding domain may form an
inhibitory structure, or bind to an unknown inhibitor. This theory is based on the
observation that the deletion of the chaperone binding domain of YopE led to
translocation in the absence of the chaperone (73). Although the chaperone —effector
interaction may have evolved as a means to promote effector stability, stability is not the
only functional role of the CBD. The other diverse functions of the CBD are synonymous

with those of the chaperone itself, which will be discussed below.

11. Class I T3SS Chaperones
IIA. Chaperones are small, dimerized proteins that bind to one or more secreted
effector protein.

Effector-specific (Class 1) type IH secretion chaperones were discovered in
Yersinia as small, acidic proteins required for the accumulation of their cognate effectors
and for bacterial virulence (34). Early studies in Yersim'a confirmed that the Syc
chaperones bound directly to Yop effector proteins (35, 88), and this direct binding
relationship has been consistently confirmed with other chaperone-effector pairs.
Although T3SS chaperones are generally required for secretion of partner effector
proteins, several studies have shown evidence of residual levels of chaperone-
independent secretion in the absence of the CBD or the chaperone (82, 87, 88).
Chaperones also bind to regulatory or secretion system proteins (discussed below). It has
been generally observed that chaperones are not themselves secreted; however, evidence
for Spa15 secretion into media and host cells was recently reported (31). T3SS
chaperones were first discovered in plant pathogens in 1998 (38, 84) and have been
characterized or predicted in a wide range of gram-negative phytopathogenic bacteria.

IIB. General Structural features of the chaperone-effector interaction.

Several crystallization studies have revealed that despite a complete lack of sequence
conservation among the chaperones, they have a highly conserved structure consisting of
three a-helices and a five-stranded B-sheet (54, 59, 67, 85). Crystal structures of
chaperones bound to effectors Show that the unfolded chaperone-binding domain of the
effector protein wraps around the exterior chaperone dimer (12, 51, 76). Cross-

complementation studies have suggested that the T3SS targeting mechanism is universal

and conserved between animal and plant pathogens (7, 70); Birtalan et al. proposed that
the conserved three-dimensional structure of the effector-chaperone complex might be
part of this universal signal (13).

IIC. Chaperone-effector specificity.

Most characterized chaperones bind to a single partner effector with a high degree
of specificity. Some exceptions are known; Page et al. (2002) reported that the Shigella
dysentariae chaperone Spa15 binds multiple effectors, leading the authors to propose that
Spa15 belongs to a subclass of chaperones (class IIB) with broad specificity. Although it
has been speculated that a novel structural variant of Spa15 might be a feature associated
with broad specificity (85), the molecular basis of chaperone-effector specificity has not
been determined. CesT and Ian are other examples of chaperones that interact with and
facilitate secretion of multiple effectors (28).

Although specific chaperone-effector pairs have been characterized, T3SS
chaperones in at least two well-characterized plant pathogen systems interact with
multiple secreted effectors. The P. syringae chaperones Sth, Sthl, and Sth2 can
each bind, stabilize, and facilitate secretion of three cognate effectors with sequence
similarity (42), and the X. campestris chaperone HpaB is a “global chaperone” required
for secretion of five different effectors and for pathogenicity (18). It is possible that,
because P.3yringae and X. campestris express a large number of effectors, it would be
beneficial for T3 SS chaperones in these species to bind more than one effector protein.
Chaperone-effector binding range in plant pathogens with fewer effectors, such as

Dickeya chrysanthemi and Erwinia amylovora, has not been reported, but it could

provide clues as to whether broad effector-binding specificity is an effect of high effector
number.
IID. The roles of T3SS chaperones.

i. Chaperones dock effector proteins to the secretion apparatus.

The requirement of chaperones for secretion of some effectors has long led to
speculation that chaperones may serve as a docking device for effector interaction with
the T388. Indeed, since 2003, in-vivo co-immunoprecipitation and co-localization studies
have confirmed that several class I chaperones interact directly with the ATPase
component of the TBSS. Akeda and Galan (2005) found that the Salmonella enterica sv.
Typhimurium ATPase InvC not only binds to the T3SS chaperone SicP, but also
catalyzes the dissociation of the chaperone fiom its cognate effector in an ATP-dependent
manner. The chaperone-effector complex of CesT and Tir from enteropathogenic
Escherichia coli also interacts with the ATPase in a chaperone-dependent manner (40,
82). However, membrane localization of CesT is not ATPase-dependent.

Recently, evidence has emerged that some T3SS chaperones interact with
secretion system components other than the ATPase. For example, HpaB in
Xanthomonas campestris pv. vesicatoria interacts with the conserved inner membrane
proteins Hch and HrcV in vitro (57). Hch and its homologs (i.e., F lhB and Ych)
interact with T3 SS substrate specificity switches and may be involved in T3SS substrate
recognition (5, 33, 55). Another potential interactor with a T3SS chaperone is CdsQ in
Chlamydia trachomatis, orthologous to the putative C-ring component of T388 and
flagellar secretion apparatuses (75). Members of the Ych and HrpQ families are known

interactors with the HrcN central, ATPase (5 7). The recently-discovered interactions

between chaperones and multiple components of the apparatus base suggest a more
complex T3SS-targeting role for secretion chaperones than previously assumed.

The structural basis of the interaction between T3 88 chaperones and secretion
apparatus components is not known. Evidence suggests that the C-terminal end of the
ATPase is required for interacting with chaperones (2, 94). The recent solution of the first
crystal structure of a T388 ATPase, InvC, led to the identification of a putative chaperone
docking interface on the ATPase (94). Mutation of a conserved hydrophobic residue at
this site (V394) abolished the ability of the ATPase to bind a T3 SS chaperone. Since this
structure is not present on ATPases of the flagellar secretion system, it was proposed that
pathogenicity-associated T3SS ATPases are specially adapted to bind class I chaperones
(94). Less is known about what chaperone factors may be required for interaction with
the ATPase. Hydrostatic amino acids of two chaperones, SigE from S. enterica and CesT
fi'om E. coli, have been identified that are required for effector secretion but not for
effector binding (47, 82). The authors speculated that these residues may be involved in
interaction with the T3 SS ATPase, but no specific amino acids have been confirmed to be
involved in the chaperone-T3SS interaction.

ii. Chaperones confer secretion system specificity to effectors.

Consistent with the effector-ATPase docking model, several T3 SS chaperones are
thought to confer secretion system specificity upon various effectors. Lee and Galan
reported that although S. enterica effectors SopE and SptP are secreted through the SPI-l
T3SS in the presence of the chaperone binding domain (CBD), a mutant SptP or SopE
that does not bind the chaperone is only secreted through a second T3SS normally

dedicated to secretion of flagellar components (48). Another group reported that SopE

10

chaperone binding-negative mutants are secreted through both the flagellar T3SS and the
SPI—l system, but are not translocated into host cells (28). It was unclear what led to the
difference in their findings, and why the SopE mutants were able to be secreted but not
translocated by the same secretion system. However, the collected findings of these
studies point to the conclusion that the chaperone and CBD are important in trafficking
Salmonella effectors to the correct secretion system. The authors speculated that the N-
terminal 15-20 amino acids could represent an “ancestral” secretion signal recognized by
the flagellar T3SS, and this recognition may be somehow overcome by chaperone
binding downstream (28).

Because chaperones bind directly to the T3SS apparatus and play a role in
secretion system specificity, the chaperone-T3 SS interaction is one possible point at
which effector-TBSS specificity might be conferred. This role would be especially
important in the presence of two pathogenicity-associated T3 SS, as is the case in
Salmonella. Whether chaperones can interact with a broad or narrow spectrum of T3 SS
components has not been determined, although both chaperones and the secretion
apparatus are highly structurally conserved. It has been reported that the Y. pestis effector
YopE can be secreted by Salmonella enterica and the plant pathogen Xanthomonas
campestris, but only in the presence of its cognate chaperone Sch (70). This might
suggest that Sch could interact with a wide variety of T3SS ATPases, but it also could
be possible that the stabilization of YopE by Sch is sufficient to allow low levels of
secretion by multiple T3SSs. Interestingly, another study demonstrated that YopE can be
secreted by the Erwinia chrysanthemi T3SS without the Sch chaperone present (7), but

this study was conducted in an E. coli expression system that resulted in extremely low

11

levels of secretion. In any case, it is clear that the activity of chaperones is conserved in
- multiple systems. Whether this includes direct binding to T3SS components is a question
yet to be addressed in the literature.

iii. Chaperones stabilize or prevent aggregation of cognate effectors.

The observation that the chaperone binding domain may inhibit secretion in the absence
of a chaperone led some to question whether the chaperone may function in masking a
structure inhibitory to secretion. Ehrbar et al. proposed that the chaperone could
competetively prevent the CBD from binding to an unknown secretion inhibitor (28), but
other studies have demonstrated that the chaperone binding domain is itself secretion-
inhibitory, conferring destabilization to the effector protein (50). Effectors in several
different systems have reduced half-lives in the absence of the chaperone (39, 58). It was
originally proposed, based on the crystal structures of several secreted effectors, that
effectors tend to have a disordered, proteolysis-labile N-terminus. The chaperone might
mask this disorder and prevent proteolysis (13, 76). However, not all chaperone-aided
effectors have a disordered N-terminus (67). Letzelter et al. (2006) found that the
chaperone binding domains of several effectors in Y. pestis are prone to autoaggregation,
conferring insolubility to the protein in the absence of the chaperone (50). In one of these
effectors, YopO, the CBD coincided with a membrane localization domain responsible
for targeting the effector to the host cell membrane. So, in some cases the chaperone
directly binds to destabilizing domain necessary for function in the host cell.

In other cases, there is evidence that the chaperone may confer stability by a
slightly different mechanism. In the plant pathogen Pseudomonas syringae, stability

assays showed that the chaperones Sthl, Sth, and SthI protect effector proteins

12

from degradation by the Lon protease, but that the target for proteolysis lies outside the
chaperone binding domain (58). The stabilization function of chaperones is not universal,
as some effectors are not destabilized in the absence of their cognate chaperone (18, 84,
88). It should be noted however, that the possibility of stabilization by other undiscovered
chaperones was not ruled out.

iv. Chaperones may regulate the expression of cognate effectors.

Some of the earliest reports of chaperones proposed a fimction as regulators of
effector gene expression (34). Although this initial characterization was due to effector
destabilization in the absence of the chaperone, it was later discovered that some T3SS
chaperones also regulate transcription or translation of specific effector proteins. In
Salmonella typhimurium, the chaperone SicA binds directly to the transcription factor
InvF to cause it to activate transcription from effector gene promoters (23). The
chaperone-like protein Est in P. aeruginosa binds to and inhibits the anti-activator
Est, relieving Est-mediated transcriptional repression (24). Among plant pathogens,
a protein similar in structure'to classic T3SS chaperones, HrpG, is an indirect
transcriptional de-repressor of the P. syringae T3SS (89).

The Yersinia pestis YopH chaperone SycH works at the posttranscriptional level
to activate expression of yop genes; by binding to translational repressors Ysch and
YscM2, SycH relieves the repression to initiate Yop synthesis (20). Recently, it was
shown that the chaperone Sch also binds to Ysch (Dittman et al. 2007), suggesting
that posttranscriptional regulation by chaperones could be more widespread than

originally suspected.

l3

v. Chaperones may impose a temporal hierarchy of secretion.

Upon contact with the host cell, the T 3SS of S. enterica begins translocating a
pre-synthesized pool of effectors into the host cell within seconds (74). While the
phenomenon of substrate-specificity switching between the secretion apparatus and
effectors has long been recognized, microscopic and biochemical translocation assays
have only recently demonstrated a hierarchical order of translocation among the effectors
themselves (93). This fine—tuning of effector secretion could be important to deliver
proteins critical early in the interaction first (for example, cell invasion and immune
response avoidance factors), or to temporally separate effectors with opposing functions
(for example, actin polymerization and depolymerization factors).

The observation that some effectors have chaperones while others do not led to
the proposal that chaperones may function in the preferential secretion of “early burst”
effectors (13). There is now evidence that chaperones are involved in imposing this
hierarchy of secretion in both Yersinia and Salmonella. For example, the N-terminus of
the Y. pestis effector YopE is translocated in the absence of its chaperone if all other
effector genes are deleted from the cell. Addition of the other effectors reduces YopE
translocation levels unless the chaperone Sch binds, suggesting that Sch may confer a
competetive advantage to YopE (17), which in turn may limit the translocation levels of
other effectors (1). In Salmonella enterica, the effector SipA is translocated earlier than
the effector SptP, even when both effectors are pre-formed in the bacterial cytosol (91).
Computer modeling experiments demonstrated that this hierarchy of secretion might
result from the SipA and SptP cognate chaperones binding to limited T3SS sites at

slightly different affinities. In enterohemmorhagic E. coli, a system in which the

14

chaperone CesT binds many effectors, differential effector-chaperone binding affinity
was proposed to impose the secretion hierarchy (62). Although it is logical to assume that
differential chaperone-T3 SS or chaperone-effector binding affinity could confer a
competetive advantage to some effectors over others, this theory has not been tested
biochemically.

While Sch and SptP may help YopE compete with other effectors for access to
the secretion system, the Yersinia chaperone SycH works to impose a hierarchy using a
more advanced mechanism. SycH binds to the negative regulator Lch to form what may
be an inhibitory complex at the secretion apparatus (93). This complex decreases the
secretion levels of all Yops except YopH, allowing YopH to be secreted earlier than the
other Yops.

It is not known whether a temporal hierarchy of effector secretion is common in
plant pathogens, and even substrate specificity switching mechanisms of the T3SS in
plant pathogens are poorly understood. Pseudomonas and Erwinia effector secretion into
culture media can occur during active growth of the Hrp needle (45), which could suggest
that substrate specificity switching is not regulated in the same manner as in Yersinia.
However, a Hrp needle is required for secretion of effectors into the media, even during
full levels of effector expression (90). It is possible that secretion could be differently
regulated in the plant environment than in vitro, and some plant pathogens are known to
secrete T3SS substrates in a specific order. Ralstonia solanacearum T3SS secretion is
host cell contact-dependent, for example (4), and substrate specificity of the
Xanthomonas campestris T3SS is controlled by the regulatory protein HpaC and the

T388 chaperone HpaB (55). HpaB is sequestered by the secretion control protein HpaA,

15

preventing other secreted effectors from premature Hpa-dependent secretion (5 6). The
secretion of HpaA prior to other effectors is the first known example of a temporal
secretion hierarchy among effectors in plant pathogenic bacteria. Of the many complex
roles of chaperones discussed here, mediation of secretion dynamics among effectors is
perhaps the least-understood.
III. Erwinia amylovora as a system for studying chaperone-effector
interaction

IIIA. Erwinia amylovora biology and economic significance

Erwinia amylovora, a gram-negative bacterium of the family Enterobacteriaceae,
causes a devastating fire blight disease of apple trees and other plants in the family
Rosaceae. The bacterium enters the tree through blossoms or through wounds in actively
growing shoots, replicating rapidly and causing tissue necrosis (83). Infected shoots have
a crooked shape and burnt appearance characteristic of fire blight. E. amylovora spreads
throughout the tree via the xylem, and produces copious amounts of a saccharide-rich
exopolysaccharide (EPS) which aids the bacteria in movement throughout the plant and
in dispersal by wind and insects (83). E. amylovora causes significant yield losses every
year in apple and pear growing regions, and can lead to the death of young and highly
susceptible trees. Unfortunately, options for chemical control of fire blight are very
limited; the few commercially available antibiotic sprays are becoming reduced in
efficacy due to bacterial antibiotic resistance (60). Research into the molecular
mechanisms of E. amylovora pathogenesis will be important in revealing new targets for

control.

16

IIIB. The E. amylovora effector DspE

Biochemical analysis of culture supematants has indicated that E. amylovora
secretes at least four putative effector proteins into media via the T3SS: EOpl, Eop3,
Aerpt2Ea/Eop4, and DspE (63). eopI and aerptZEa deletion mutants have been
characterized and were shown to affect host range and virulence, respectively, in the
laboratory (9, 95). In contrast, the effector DspE is required for E. amylovora
pathogenicity and population growth in planta. The dspE gene was identified as the result
of several transposon mutagenesis screens for mutants that abolished pathogenicity but
not the hypersensitive response on tobacco (10, 78). DspE is 1,838 amino acids long,
among the largest secreted effectors characterized, with homologs in several important
plant pathogens including Pseudomonas syringae, Ralstonia solanacearum, and Pantoea
stewartii. Deletion of these DspE homologs typically leads to a reduction in virulence or
complete loss of pathogenicity (16). aer, a dspE homolog in PSeudomonas syringae,
can partially complement a dspE mutant, suggesting functional conservation within the
DspE class of effectors (14). The sequence of dspE is highly conserved among strains of
E. amylovora and E. pyrzfoliae (41), and dspE from E. pyrifoliae and from Rubus-
infecting species of E. amylovora complement the virulence of a dspE knockout mutant
of E. amylovora strain Ea1189 (Triplett, unpublished). As the above studies indicate,
DspE is an ancient, essential component of many pathogen-host interactions.

Although its specific function is unknown, DspE suppresses salicylic acid-
mediated cell wall defenses when transiently expressed in plant cells, and induces
necrotic cell death in host and nonhost cells (16, 26). The DspE homolog Aer is

functionally redundant to the P. syringae effector Hole, which mediates the

17

degradation of vesicle trafficking pathway components (64), so it is possible that Aer
and DspE work through a similar mechanism to suppress host immunity. The firnctional
targets of DspE are unknown, but several leucine rich-repeat (LRR) kinases have been
identified in apple which interact with DspE in yeast and in vitro (61). RNAi silencing of
these DspE-interacting proteins of Malus, or DIPMs, led to a slight increase in resistance
to E. amylovora (15). Although the localization of DspE in the host cell has not been
reported, the sequences of DspE and its homologs include a WXXXE motif required for
fimction and another conserved motif similar to an endoplasmic reticulum targeting
signal (44). Taken together, these studies suggest that DspE could fimction in multiple
and novel roles in the plant infection process.

IIIC. The E. amylovora chaperone DspF.

The first sequence analyses of the dsp region of E. amylovora showed that dspE
shared a transcriptional unit with a small orf, named dspB or dspF (14, 38). For this
discussion, the gene will heretofore be referred to as dspF. The predicted gene product of
the dspF orf had the defining characteristics of a class I T3SS chaperone: a size of 16
kDa, an acidic pl, and a helical structure. Disruption of the dspF gene caused a drastic
reduction in virulence and led to an apparent disapperance of DspE protein fi'om the
supernatant, strongly indicating that dspF encoded a DspE chaperone (3 8). Biochemical
studies showed that the protein DspF binds directly to DspE in vitro, as well as another
unknown interactor of about 80 kDa (39). DspF was required for immunodetection of
DspE in E. amylovora cells, although the chaperone did not significantly affect DspE
transcriptional levels, indicating that DspF must affect either stability or translation of

DspE. Despite the apparent lack of DspE in the culture supematants in the dspF mutant

18

background, the fact that the dspF mutant retains a reduced degree of T3 SS-dependent
pathogenicity suggests that DspE is still secreted at some level in the absence of its
chaperone (39).

Many of the sequenced bacterial pathogens containing a dspE homolog also carry
a gene homologous to dspF. The role of a dspF homolog has only been characterized in
Pantoea stewartii (44). The P. stewartii gene wtsF has a positive effect on the stability of
the effector WtsE, but is not required for virulence, possibly due to the presence of a
redundant chaperone. In fact, among characterized chaperones in plant pathogens, only
DspF and HpaB fiom Xanthomonas campestris have a significant effect on virulence. (As
mentioned previously, HpaB is the only known T3SS chaperone in X. campestris). In
other effector systems studied, redundancy of effectors and chaperones precludes a
virulence phenotype for mutants of a single chaperone. The requirement of DspF in
virulence makes E. amylovora a useful system for the study of chaperone function in
plant pathogens.

IV. Rationale and project goals.

There are few studies illustrating the role of type HI secretion chaperones in plant
pathogen systems. However, these studies have expanded the known diversity of
chaperone functional roles and challenged some previous paradigms of secretion
signalling. Figure 1-1 compares three variations of chaperone-effector binding between
P. syringae, X. campestris, and enterobacterial pathogens such as Salmonella enterica sv.
Typhimurium. In the Enterobacteriaceae, chaperones each bind to the N-terminus of a
single secreted effector in a highly specific manner. Little or no effector secretion occurs

in the absence of each cognate chaperone (Figure 1-1A). However, in P. syringae, which

19

expresses a large number of chaperones and secreted effectors during infection, some
chaperones may interchangeably bind to and stabilize several paralogous effector proteins
(Figure l-lB). While Xanthomonas campestris pv. vesicatoria also harbors numerous
putative secreted effectors, HpaB is the only known chaperone. HpaB is required for the
translocation, but not the stabilization, of multiple effector proteins, but it also aids in the
development of a secretion hierarchy via a novel mechanism (Figure 1-1C). Thus, the
roles and binding patterns of chaperone-effector pairs may vary substantially between
systems. How great is the diversity of chaperone roles among other plant pathogens and
symbionts, and what selective pressures among different organisms contributed to this
divergence? Study of chaperone-effector interaction in a wider variety of pathogens and
symbionts is needed to understand the spectrum of T3SS regulation and targeting
mechanisms.

The purpose of this research is to identify translocation signals of the effector DspE
and further characterize DspF chaperone fimction in E. amylovora, a system that is both
enterobacterial and plant-associated. As previously mentioned, the strong virulence
phenotype of dspE and dspF deletion strains makes this effector-chaperone pair a
conducive system for study. First, the project aims to characterize the general mechanism
of interaction between the critical E. amylovora effector, DspE, and its chaperone, DspF.
Toward this objective, I performed an extensive series of targeted mutagenesis
experiments in both DspE and DspF. Secretion assays, adenylate-cyclase reporter-based
translocation assays, and protein-protein interaction studies were used to analyze these
mutants. In Chapter 2, I demonstrated that between 31 and 51 N-terminal amino acids are

required for both DspE secretion and for translocation into tobacco cells, while longer

20

fragments lead to higher translocation levels. I identified a conserved regions between
amino acids 2-10 and 31-40 required for optimal translocation, and showed that an N-
terrninal chaperone binding domain of DspE lies between amino acids 51-100, although
neither this domain nor the presence of DspF was required for translocation. In Chapter 3,
I used a computer- generated, homology-based model of DspF to identify amino acid
residues required for full virulence of E. amylovora. In mutagenesis studies, four highly-
conserved residues with predicted localization in a helix-binding groove were identified
as required for DspF fimction; three of these residues were required for interaction with
DspE in yeast, supporting the putative role of electronegative residues of the helix-
binding groove in chaperone function.

The second aim of this project has been to identify potential new roles of DspF by
identifying putative DspF interactors in E. amylovora, and assessing the function of some
of these predicted interactions (Chapter 4). In a directed yeast two-hybrid screen, four
novel putative DspF-interacting proteins were identified, including two putative
components of the T388 machinery (Hch and Hch) and two effector proteins with
little previous characterization (Eopl and Eop3). Eopl and Eop3 have predicted
chaperones, while Eop4, an effector that does not interact with DspF in yeast, has no
predicted chaperone. Although DspF interacted with Eopl and Eop3, translocation assays
indicated that deletion of dspF did not have a negative impact on effectors other than
DspE. Interestingly, translocation levels of Eopl and Eop3 were increased in the absence
of DspF, leading to the hypothesis that DspF could have a direct or indirect negative role
on translocation of some effectors by facilitating competition by DspE or through a

regulatory mechanism. The phenotype of a dspF deletion mutant could be complemented

21

by ectopic DspE expression, suggesting that DspF has a role in maintaining adequate
DspE concentrations within the cell. Thus, this study provides support for formerly
hypothesized roles of DspF, and points to new interactors and potential DspF functions.
A fourth, unrelated study I completed on genomic differences between Erwinia
amylovora and close taxonomic relatives is included in the appendix. In combination
with the mapping and characterization of the DspE-DspF interaction, the work in this
dissertation contributes significantly to the understanding of Erwinia amylovora

molecular pathogenesis.

22

Y. pestis P. syringae C X. campestris

A

a.)
0::
: J

lllllllllllllllllllllllll

 

<— Sthl "1

. y , HopO 1 -1 p aB
Y0 E Sthl !
p \ A HopSl HpaA
SYCO YOpO HOpOl -m I I

 

 

 

YopM
(chaperone- .$ chaperone- HpaB-dependent
independent) independent effectors

effectors

Figure 1-1. Variations in the binding pattern and roles of T3SS chaperones. A. Specific
chaperone-effector interactions in Yersinia pestis. The N-terminus of secreted effectors
including YopE and YopO each bind a single chaperone, Sch (in green) and Sch (in
orange), respectively, which may stabilize and dock the effector to peripheral inner
membrane components of the secretion system. A small number of effector proteins, such
as the effector YopM, have no known chaperone. B. While some chaperone-effector
interactions in the plant pathogen Pseudomonas syringae are thought to be highly
specific, some chaperones interact with multiple related effectors. The chaperones Sthl
(orange) and SthI (blue) can interchangeably facilitate secretion of HopOl-l (light
orange) and HopSl (light blue). Chaperone binding sites have been mapped to the N-
terrninus or to the central portion of various effectors. C. The Xanthomonas campestris
chaperone HpaB (in black) interacts with the C—terminus of the secretion control protein
HpaA (light blue). Upon HpaA translocation into host cells, HpaB binds the N-terminus
of multiple other HpaB-dependent effectors and renders them competent for secretion.

23

10.

11.

WORKS CITED

. Aili, M., E.L. Isaksson, S.E. Carlsson, H. Wolf-Watz, R. Rosqvist and M.S.

Francis. 2008. Regulation of Yersinia Yop-effector delivery by translocated YopE.
Int. J. Med. Microbiol. 298: 183-192.

Akeda, Y., and J.E. Galan. 2004. Genetic analysis of the Salmonella enterica type
III secretion-associated ATPase InvC defines discrete functional domains. J.
Bacteriol. 186: 2402-2412.

Akeda, Y. and J.E. Galan. 2005. Chaperone release and unfolding of substrates in
type IH secretion. Nature 437: 911-915.

Aldon, D., B. Brito, C. Boucher, and S. Genin. 2000. A bacterial sensor of plant
cell contact controls the transcriptional induction of Ralstonia solanacearum
pathogenicity genes. EMBO J 19: 2304—23 14.

Allaoui, A., S. Woestyn, C. Sluiters, and GR. Cornelis. 1994. Ych, a Yersinia
enterocolitica inner membrane protein involved in Yop secretion. J. Bacteriol. 176:
4534-4542.

Anderson, D.M., O. Schneewind. 1997. An mRNA signal for the secretion of Yop
proteins by Yersinia enterocolitica. Science 278: 1140-1143.

Anderson, D.M., D.E. Fonts, A. Collmer and O. Schneewind. 1999. Reciprocal
secretion of proteins by the bacterial type IH machines of plant and animal pathogens

suggests universal recognition of mRNA targeting signals. Proc. Natl. Acad. Sci. USA
96: 12839-12843.

Arnold, R., S. Brandmaier, F. Kleine, P. Tischler, E. Heinz, S. Behrens, A.
Niinikoski, H.W. Mewes, M. Horn, T. Rattei. 2009. Sequence-based prediction of
type III secreted proteins. PLoS Pathog 5: e10003 76.

Asselin, J.E., K.N. Yip, and S.V. Beer. 2008. Eopl differs in strains of Erwinia
amylovora that differ in host specificity. Proceedings of the XI International
Workshop on Fire Blight. Ed. K.B. Johnston and V0. Stockwell. Acta Hort 793:
213-214.

Barny, M.A., M.H. Guoinebretiere, B. Marcais, E. Coissac, J.P. Paulin and J.
Laurent. 1990. Cloning of a large gene cluster involved in Erwinia amylovora CFBP
1430 virulence. Mol. Microbiol. 4: 777-786.

Belleman, P., and K. Geider. 1992. Localization of transposon insertions in
pathogenicity mutants of Erwinia amylovora and their biochemical characterization.
J. Gen. Microbiol. 138: 931-940.

24

12. Birtalan, S., and P. Ghosh. 2001. Structure of the Yersinia type III secretory system
chaperone Sch. Nat Struct Biol. 8: 974-8.

13. Birtalan, S.C., R.M. Phillips, and P. Ghosh. 2002. Three-dimensional secretion
signals in chaperone-effector complexes of bacterial pathogens. Mol. Cell 9: 971-980.

14. Bogdanove, A.J., J.F. Kim, Z. Wei, P. Kolchinsky, A.0. Charkowski, A.K.
Conlin, A. Collmer, and S.V. Beer. 1998. Homology and functional similarity of an
hrp-linked pathogenicity locus, dspEF, of Erwinia amylovora and the avirulence
locus aer of Pseudomonas syringae pathovar tomato. Proc. Natl. Acad. Sci. USA 95:
1 325-1330.

15. Borejsza-Wysocka, E.E., M. Malnoy, X. Ment, J.M. Bonasera, R.M. Nissinen,
J.F. Kim, S.V. Beer, and HS. Aldwinckle. 2006. The fire blight resistance of apple
clones in which DspE-interacting proteins are silenced. Acta Hort. 704: 509-513.

16. Boureau, T., H. El Maarouf-Boteau, A. Garnier, M.-N. Brisset, C. Perino, I.
Pucheu, and M.-A. Barny. 2005. DspA/E, a type III effector essential for Erwinia
amylovora pathogenicity and growth in planta, induces cell death in host apple and
nonhost tobacco plants. Mol. Plant Microbe Interact. 19: 16-24.

17. Boyd, A.P., I. Lambermont, and GR. Cornelis. 2000. Competition between the
Yops of Yersinia enterocolitica for delivery into eukaryotic cells: role of the Sch
chaperone binding domain of YopE. J. Bacteriol. 182: 4811-4821.

18. Biittner, D., D. Gurlebeck, L.D. Noel, and U. Bonas. 2004. HpaB from
Xanthomonas campestris pv. vesicatoria acts as an exit control protein in type III-
dependent protein secretion. Mol Microbiol 54: 755—768.

19. Bilttner, D. and S.-Y. He. 2009. Type 111 protein secretion in plant pathogenic
bacteria. Plant Physiol. 150: 1656-1664.

20. Cambronne, E.D., J.A. Sorg and O. Schneewind. 2004. Binding of SycH
chaperone to Ysch and YscM2 Activates effector yop expression in Yersinia
enterocolitica. J. Bacteriol. 186: 829-841.

21. Cheng, L.W., D.M. Anderson and O. Schneewind. 1997. Two independent type III
secretion mechanisms for YopE in Yersinia enterocolitica. Mol. Microbiol. 24: 757-
765.

22. Cornelis, G. 2006. The Type III secretion injectosome. Nature Rev. Microbiol. 4:
81 1-825.

23. Darwin, KB. and V.L. Miller. 2001. Type HI secretion chaperone-dependent

regulation: activation of virulence genes by SicA and InvF in Salmonella
typhimurium. EMBO J. 20: 1850-1862.

25

24. Dasgupta, N., G.L. Lykken, M.C. Wolfgang, and T.L. Yahr. A novel anti-anti-
activator mechanism regulates expression of the Pseudomonas aeruginosa type III
secretion system. Mol. Microbiol. 53: 297-308.

25. P. Delepelaire. 2004. Type I secretion in gram-negative bacteria. Biochim. Biophys.
Acta 1694: 149-161.

26. Debroy, S., R. Thilmony, Y.-B. Kwack, K. Nomura, and S.-Y. He. 2004. A family
of conserved bacterial effectors inhibits salicylic acid-mediated basal immunity and
promotes disease necrosis in plants. Proc. Natl. Acad. Sci. USA 101: 9927-9932.

27. Ding, 2., K. Atmakuri and P. J. Christie. 2003. The outs and ins of bacterial type
IV secretion substrates. Trends in Microbiol. 11: 527-535.

28. Ehrbar, K., B. Winnen and W. Hardt. 2005. The chaperone binding domain of
SopE inhibits transport via flagellar and SPI-l TTSS in the absence of Ian. Mol.
Microbiol. 59: 248-264.

29. Enninga, J., J. Mounier, P. Sansonetti and GT. Nhieu. 2005. Secretion of type III
effectors into host cells in real time. Nat Methods 2: 959—965.

30. Evans, L.D.B., G.P. Stafford, S. Ahmed, G.M. Fraser and C. Hughes. 2006. An
escort mechanism for cycling of export chaperones during flagellum assembly. Proc.
Natl. Acad. Sci. USA 103: 17474-17479.

31. Faherty, C.S., and A.T. Morelli. 2009. Spal 5 of Shigellaflexneri is secreted
through the type-III secretion system and prevents staurosporine-induced apoptosis.
Infection and Immunity online ahead of print. doi:10.1128/IAI.00800-09.

32. Feldman, M.A., Muller, S., Wuest, E., and Cornelis, GR. 2002. Sch allows
secretion of YopE-DHFR hybrids by the Yersinia enterocolitica type HI Ysc system.
Mol. Microbiol 46: 1 183-1 197.

33. Ferris, H.U., Y. Furukawa, T. Minamino, M.B. Kroetz, M. Kihara, K. Namba,
and RM. McNab. 2005. FlhB regulates ordered export of flagellar components via
autocleavage mechanism. J. Biol. Chem. 280: 41236- 41242.

34. Forsberg, A. and H. Wolf-Watz. 1990. Genetic analysis of the YopE region of
Yersinia spp.: Identification of a novel conserved locus, yerA, regulating Yop gene
expression. J. Bacteriol. 172: 1547-1555.

35. Frithz-Lindsten, E., R. Rosqvist, L. Johansson and A. Forsberg. 1995. The
chaperone-like protein YerA of Yersinia pseudotuberculosis stabilizes YopE in the

cytoplasm but is dispensable for targeting to the secretion loci. Mol. Microbiol. 16:
635-647.

26

36. Furutani, A., M. Takaoka, H. Sanada, Y. Noguchi, T. Oku, K. Tsuno, H. Ochiai,

and S. Tsuge. 2009. Identification of novel type III secretion effectors in
Xanthomonas oryzae pv. oryzae. Mol. Plant Microbe Interact. 22: 96-106.

37. Galan, J., and H. Wolf-Watz. 2006. Protein delivery into eukaryotic cells by Type

38.

39.

40.

41.

42.

43.

45.

ID secretion machines. Nature 444: 567-5 73.

Gaudriault, S., L. Malandrin, J.-P. Paulin and M.-A. Barny. 1997. DspA, an
essential pathogenicity factor of Erwinia amylovora showing homology with Aer of
Pseudomonas syringae, is secreted via the Hrp secretion pathway in a DspB-
dependent way. Mol. Microbiol. 26: 1057-1069.

Gaudriault, S., J.-P. Paulin and M.A. Barny. 2002. The DspB/F protein of Erwinia
amylovora is a type IH secretion chaperone ensuring efficient intrabacterial
production of the Hrp-secreted DspA/E pathogenicity factor. Mol. Plant Pathology 3:
3 13-320.

Gauthier, A. and BB. Finlay. 2003. Translocated intimin receptor and its chaperone
interact with ATPase of the type HI secretion apparatus of enteropathogenic
Escherichia coli. J. Bacteriol. 185: 6747-6755.

Giorgi, S. and M. Scortichini. 2005. Molecular characterization of Erwinia
amylovora strains from different host plants through RFLP analysis and sequencing
of hrpN and dspA/E genes. Plant Pathology 2542 789-798.

Guo, M., S.T. Chancey, F. Tian, Z. Ge, Y. Jamir, and J.R. Alfano. 2005.
Pseudomonas syringae type IH chaperones SthI, SthI, and SthZ facilitate
translocation of their cognate effectors and can substitute for each other in the
secretion of HopOl-l. J. Bacteriol. 187: 4257-4269.

Guttman DS, Vinatzer BA, Sarkar SF, Ranall MV, Kettler G, Greenberg JT.
2002. A fimctional screen for the type III (Hrp) secretome of the plant pathogen
Pseudomonas syringae. Science 295: 1722—1726.

. Ham, J.H., D.R. Majerczak, A.S. Arroyo-Rodriguez, D.M. Mackey and D.L.

Coplin. 2006. WtsE, an Aer-family effector protein fiom Pantoea stewartii subsp.
stewartii, causes disease-associated cell death in corn and requires a chaperone
protein for stability. Mol. Plant Microbe Interact. 19: 1092-1102.

Jin, Q., W. Hu, 1. Brown, G. McGhee, P. Hart, A. Jones and S.-Y. He. 2001.
Visualization of secreted Hrp and Avr proteins along the Hrp pilus during type HI
secretion in Erwinia amylovora and Pseudomonas syringae. Mol. Microbiol. 40:

1129-1139.

27

46. Karavolos, M.H., M. Wilson, J. Henderson, J.J. Lee and C.M.A. Khan. 2004.
Type HI secretion of the secreted effector SopE is mediated by an N-terminal amino
acid signal and not an mRNA sequence. J. Bacteriol. 187 : 1559- 1567.

47. Knodler, L.A., M. Bertero, C. Yip, N.C.J. Strynadka and O. Steele-Mortimer.
2006. Structure-based mutagenesis of SigE verifies the importance of hydrophobic
and electrostatic residues in type IH chaperone function. Mol. Microbiol. 62: 928-940.

48. Lee, 8.11. and J.E. Galan. 2004. Salmonella type III secretion-associated chaperones
confer secretion pathway specificity. Mol. Microbiol. 51: 483-495.

49. Lee, V.T., D.M. Anderson and O. Schneewind. 1998. Targeting of Yersinia Yop
proteins into the cytosol of HeLa cells: one-step translocation of YopE across

bacterial and eukaryotic membranes is dependent on Sch chaperone. Mol.
Microbiol. 28: 593-601.

50. Letzelter, M. I. Sorg, L. J. Mota, S. Meyer, J. Stalder, M. Feldman, M. Kuhn, I.
Callebaut and GR. Cornelis. 2006. The discovery of Sch highlights a new
function for type III secretion effector chaperones. EMBO J. 25: 3223-3233.

51. Lilic, M., M. Vujanac and E. Stebbins. 2006. A common structural motif in the
binding of virulence factors to bacterial secretion chaperones. Mol. Cell 21: 653-664.

52. Lloyd, S.A., M. Norman, R. Rosqvist and H. Wolf-Watz. 2001. Yersinia YopE is
targeted for type III secretion by N-terrninal, not mRNA, signals. Mol. Microbiol. 39:
520-532.

53. Lloyd, S.A., M. Sjostrom, S. Andersson and H. Wolf-Watz. 2002. Molecular
characterization of type III secretion signals via analysis of synthetic N-terminal
amino acid sequences. Mol. Microbiol. 43: 51-59.

54. Locher, M., B. Lehnert, K. Krauss, J. Heesemann, M. Groll, and G. Wilharm.
2005. Crystal structure of the Yersinia enterocolitica type IH chaperone SycT. J. Biol.
Chem. 280: 31149-55.

55. Lorenz, C., S. Schulz, T. Wolsch, O. Rossier, U. Bonas, and D. Buttner. 2008.
HpaC controls substrate specificity of the Xanthomonas type III secretion system.
Plos Pathogens 4: doi e1000094.

56. Lorenz, C., O. Kirchner, M. Egler, J. Stuttmann, U. Bonas, and D. Buttner.
2008. HpaA from Xanthomonas is a regulator of type III secretion. Mol. Microbiol.
69: 344-360.

57. Lorenz, C. and D. Buttner. 2009. Functional characterization of the type III
secretion ATPase HrcN from the plant pathogen Xanthomonas campestris pv.
vesicatoria. J. Bacterial. 191: 1414-1428.

28

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

Losada, LC. and SW. Hutcheson. 2005. Type HI secretion chaperones of
Pseudomonas syringae protect effectors from Lon-associated degradation. Mol.
Microbiol. 55: 941-953.

Luo, Y., M.G. Bertero, E.A. Frey, R.A. Pfuetzner, M.R. Wenk, L. Creagh, S.L.
Marcus, D. Lim, F. Sicheri, C. Kay, C. Haynes, B.B. Finlay, and N.C.J.
Strynadka. 2001. Structural and biochemical characterization of the type 1H
secretion chaperones CesT and SigE. Nature Struct. Biol. 8: 1031-1036.

McManus, P.S., V.O. Stockwell, G.W. Sundin, and A.L. Jones. Antibiotic use in
plant agriculture. Annu. Rev. Phytapathal. 40: 443-465.

Meng, X, J .M. Bonasera, J .F. Kimm R. M. Nissinen and S.V. Beer. 2006. Apple
proteins that interact with DspA/E, a pathogenicity factor of Erwinia amylovora, the
fire blight pathogen. Mol. Plant Microbe Interactions 19:53-61.

Mills, E., K. Baruch, X. Charpentier, S. Kobi, and I. Rosenshine. 2008. Real-time
analysis of effector translocation by the type IH secretion system of enteropathogenic
Escherichia coli. Cell Host and Microbe 3: 104-113.

Nissinen, R.M., A.J. Ytterburg, AJ. Bogdanove, K.J. Van Wijk and S.V. Beer.
2006. Analyses of the secretomes of Erwinia amylovora and selected hrp mutants

reveal novel type HI secreted proteins and an effect of Her on extracellular harpin
levels. Mol. Plant Pathol. 8: 55-67.

Nomura, K., S. Debroy, Y. H. Lee, N. Pumplin, J. Jones, and S.-Y. He. 2006. A
bacterial virulence protein suppresses host innate immunity to cause plant disease.
Science 313: 220-223.

Page, A.L., P. Sansonetti, and C. Parsot. 2002. Spa15 of Shigellaflexneri, a third
type of chaperone in the type III secretion pathway. Mal. Microbial. 43: 1533-1542.

Pattis, 1., E. Weiss, R. Laugks, R. Haas and W. Fischer. 2007. The Helicabacter
pylori CagF protein is a type IV secretion chaperone-like molecule that binds close to

the C-tenninal secretion signal of the CagA effector protein. Microbiology 153: 2896-
2909.

Phan, J., J.E. Tropea, and D.S. Waugh. 2004. Structure of the Yersinia pestis type
HI secretion chaperone SycH in complex with a stable fragment of YscM2. Acta
Crystallographica 60: 1591-1599.

Ramamurthi, K.S., and O. Schneewind. 2002. Yersinia enterocolitica type III
secretion: mutational analysis of the yapQ secretion signal. J. Bacteriol. 184: 3321-
3328.

29

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

Ramamurthi, K, and O. Schneewind. 2005. A synonymous mutation in Yersinia
enterocolitica yopE affects the function of the YopE type III secretion signal. J.
Bacterial. 187: 707-715.

Rossier, 0., K. Wengelnik, K. Hahn, and U. Bonas. 1999. The Xanthomonas hrp
type 111 system secretes proteins from plant and mammalian bacterial pathogens.
Proc. Natl. Acad. Sci. USA 96: 9368-73.

Rosqvist, R., S. Hakansson, A. Forsberg and H. Wolf-Watz. 1995. Functional
conservation of the secretion and translocation machinery for virulence proteins of
Yersiniae, Salmonellae, and Shigellae. EMBO J. 14: 4187-4195.

Sandkvist, M. 2004. The biology of type H secretion. Mal. Microbial. 40: 271-283.

Schesser, K., E. Frithz-Lindsten and H. Wolf-Watz. 1996. Deliniation and
mutational analysis of the Yersinia pseudotuberculosis YopE domains which mediate

translocation across bacterial and eukaryotic cellular membranes. J. Bacterial. 178:
7227-7233.

Schlumberger, M.C., A.J. Muller, K. Ehrbar, B. Winnen, I. Duss, B. Stecher,
and W.D. Hardt. 2005. Real-time imaging of type III secretion: Salmonella SipA
injection into host cells. Proc. Natl. Acad. Sci. USA 102: 12548—12553.

Spaeth, KE., Y.-S. Chen, and RH. Valdivia. 2009. The Chlamydia type III
secretion C-ring engages a chaperone-effector protein complex. PloS Pathog.:
e1 0005 79.

Stebbins, GE. and J.E. Galan. 2001. Maintenance of an unfolded peptide by a
cognate chaperone in type III secretion. Nature 414: 77-81.

Sory, M.P., A. Boland, I. Lambermont, and GR. Cornelis. 1995. Identification of
the YopE and YopH domains required for secretion and internalization into the

cytosol of macrophages, using the cyaA gene fusion approach. Proc. Natl. Acad. Sci.
USA 92: 11998-12002.

Steinberger, E.M. and S.V. Beer. 1988. Creation and complementation of
pathogenicity mutants of Erwinia amylovora. Mol. Plant Microbe Interactions 1:
1 35-144.

Tamano, K., Katayama, E., Toyotome, T. & Sasakawa, C. 2002. Shigella Spa32 is
an essential secretory protein for functional type III secretion machinery and

uniformity of its needle length. J. Bacteriol. 184: 1244—1252.

Tang, X., Y. Xiao, and J .M. Zhao. 2006. Regulation of the type III secretion system
in phytopathogenic bacteria. Mol. Plant Microbe Interact. 19: 1159-1 166.

30

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

Thomas, J., G.P. Stafford and C. Hughes. 2004. Docking of cytosolic chaperone-
substrate complexes at the membrane ATPase during flagellar type HI protein export.
Proc. Natl. Acad. Sci. USA 101: 3945-3950.

Thomas, N., W. Deng, J. Puente, E.A. Frey, C.K. Yip, N.C.J. Strynadka and BB.
Finlay. 2005. CesT is a multi-effector chaperone and recruitment factor required for
the efficient type III secretion of both LEE and non-LEE-encoded effectors of
enteropathogenic Escherichia coli. Mol. Microbiol. 57: 1762-1779.

Vanneste, J.L. 2000. Fire Blight: The disease and its causative agent, E. amylovora.
Oxfordshire, UK: CABI publishing.

van Dijk, K., V.C. Tam, A.R. Records, T. Petnicki-chieja, and J.R. Alfano.
2002. The ShcA protein is a molecular chaperone that assists in the secretion of the
HopPsyA effector from the type III (Hrp) protein secretion system of Pseudomonas
syringae. Mal. Microbial. 44: 1469-1481.

Van Eerde, A., C. Hamiaux, J. Perez, C. Parsot, B.W. Dijkstra. 2004. Structure of
Spa15, a type IH secretion chaperone from Shigellaflexneri with broad specificity.
EMBO Rep. 5: 477-483.

Veiga, E., Sugawara, E., Nikaido, H., de Lorenzo, V., and Fernandez, L. A.
2002. EMBOJ. 21: 2122—2131.

Wattiau, P., and GR. Cornelis. 1993. Sch, a chaperone-like protein of Yersinia
enterocolitica involved in the secretion of YopE. Mol. Microbiol. 8: 123-131.

Wattiau, P., Bernier, B., P. Deslee, T. Michiels, and GR. Cornelis. 1994.
Individual chaperones required for Yop secretion by Yersinia. Proc. Natl. Acad. Sci.
USA 91:10493-10497.

Wei, C.-F., W.-L. Deng and H.-C. Huang. 2005. A chaperone-like HrpG protein
acts as a suppressor of HrpV in regulation of the Pseudomonas syringae pv. syringae
type IH secretion system. Mol. Microbiol. 57: 520-536.

Wei, W., PIovanich-Jones, A., W.-L. Deng, Q.-L. J in, A. Collmer, H.-C. Huang
and S.-Y. He. 2000. The gene coding for the Hrp pilus structural protein is required

for type HI secretion of Hrp and Avr proteins in Pseudomonas syringae pv. tomato.
Proc. Natl. Acad. Sci. 97: 2247-2252.

Winnen, B., M.C. Schlumberger, A. Sturm, K. Schupbach, S. Siebenmann, P.

Jenny and W. Hardt. 2008. Hierarchical effector protein transport by the Salmonella
typhimurium SPI-l type III secretion system. PLOS One 3: e2178 1-8.

31

92. Woestyn, S., M.-P. Sory, A. Boland, 0. Lequenne, and GR. Cornelis. 1996. The
cytosolic Sch and SycH chaperones of Yersinia protect the region of YopE and

YopH involved in translocation across eukaryotic cell membranes. Mal. Microbial.
20:1261-1271.

93. Wulf'f-Strobel, C.R., A.W. Williams and SC. Straley. 2002. Lch and SycH
function together at the Ysc type III secretion system in Yersinia pestis to impose a
hierarchy of secretion. Mol. Microbiol. 43: 411-423.

94. Zarivach, R., Vuckovic, M., Deng, W., Finlay, B.B., and Strynadka, N.C.J. 2007.
Structural analysis of a prototypical ATPase from the type III secretion system.
Nature Struct. and Mal. Biol. 14: 131-137.

95. Zhao, Y., S.-Y. He and G. Sundin. 2006. The Erwinia amylovora aerpt2EA gene

contributes to virulence on pear and Aerpt2EA is recognized by Arabidopsis RPS2
when expressed in P. syringae. Mol. Plant Microbe Interact. 19: 644-654.

32

Chapter 2: Functional analysis of the N-terminus of the Erwinia amylovora secreted
effector DspA/E reveals features required for secretion, translocation, and binding
to the chaperone DspB/F.

This chapter has been modified from a publication of the same title in Molecular Plant
Microbe Interactions (2009, vol. 22: 1282-1292), co-authored by Maeli Melotto and
George Sundin.

ABSTRACT

DspA/E is a type III secreted effector protein required for pathogenicity in the
apple and pear pathogen Erwinia amylovora, and DspB/F is a small chaperone protein
involved in DspA/E secretion. While the secretion and translocation signals of many type
IH secretion effector proteins in human enteric pathogens have been characterized
extensively, relatively little is known about the translocation requirements of many
effectors in plant pathogens, including large DspE-like proteins. In this study, we report a
functional analysis of the N-terminus of DspE. The minimal requirements for secretion,
translocation, and chaperone binding were characterized. Translocation assays using an
adenylate cyclase (CyaA) reporter indicated that the first 51 amino acids of DspE were
sufficient for translocation, and that 150 amino acids were required for optimal
translocation levels. The minimal translocation signal corresponded with the
requirements for secretion into culture media. Mutations of conserved regions in amino
acids 2-10 and 31-40 were found to influence translocation levels of an N-terminal DspE-
CyaA fusion. Yeast two-hybrid and in-vitro pull-down assays revealed a chaperone
binding site within amino acids 51-100 of DspE and binding to DspF in this region was
disrupted by specific mutations. However, neither disruption of the chaperone binding

domain nor deletion of the dspF gene had a significant impact on translocation levels of

N-terminal DspE-CyaA fusions.

33

INTRODUCTION

Type III secretion systems (T3SS) are syringe-like structures central to the
infection strategy of many bacterial pathogens and symbionts. Through these systems,
effector proteins are injected into host cells and then function to suppress innate defenses
and/or induce infection. Although the N-terminal 15 amino acids is sufficient for the
secretion of many effectors into the external medium, this minimal portion is often
insufficient to elicit measurable effector translocation into host cells, and a longer
downstream translocation signal is required (45, 47). The translocation signal often
coincides with a domain within the N-terminal 200 amino acids that binds to a type HI
secretion chaperone (17, 22, 32, 53).

Type HI secretion chaperones are small proteins that facilitate secretion of
effector, translocator, and secretion apparatus proteins through the T3SS. These
chaperones have been divided into several classes; of these, class I chaperones are small
structurally conserved dimers that interact specifically with T3SS effector proteins, while
class II and III chaperones interact with the translocon and secretion apparatus,
respectively (43). Chaperones typically bind to a single site within residues 30 through
150 of the cognate effector protein (35). Co-crystallization studies have provided a model
of the chaperone-effector interaction in which the chaperone binding domain of the
effector wraps around the exterior of the chaperone dimer, interacting with a helix-
binding groove and hydrophobic surfaces on the chaperone (3, 48). In many cases, the
chaperone binding domain coincides with or is adjacent to a destabilizing, aggregation-
prone, or otherwise secretion inhibitory domain masked by chaperone binding (6, 19, 34).

In addition to stability, chaperone binding can confer secretion system specificity to a

34

cognate effector or modulate the order in which effectors are translocated into host cells
(10, 11, 17). Interaction with a chaperone is necessary for substantial levels of protein
translocation of many T3 SS effectors (19), although chaperone-independent translocation
of some effectors can be detected (13, 18, 49).

In the past decade, it has become increasingly clear that class I T3SS chaperones
have an important role in plant pathogenesis. T3SS chaperones in plant pathogenic
bacteria, like their counterparts in animal pathogens, exhibit a conserved three-helix
architecture and bind directly to cognate effector proteins to confer stability and efficient
secretion (11). Despite these similarities, studies focusing on the role of the effector-
chaperone interaction in plant pathogenesis have yielded several novel finds. For
example, while class I T3SS chaperones were previously thought to universally bind to
the N-terminus of their cognate effector, the Pseudomonas syringae effector HopOl -1
was reported to have a chaperone binding site in the middle third of the protein (24).
Second, while the chaperone binding domain itself is a destabilization domain for several
Yersinia-secreted effectors, the chaperones of P. syringae effectors were shown to
prevent proteolytic cleavage targeted to a site outside the chaperone binding domain (36).
Third, studies of secretion signalling in plant pathogens have provided several new
examples of broad specificity chaperones that facilitate the secretion of multiple partners
(10, 24). Thus, additional examinations of chaperone-effector interactions in plant
pathogens could greatly further our understanding of chaperone-effector specificity and
secretion system targeting.

Erwinia amylovora, the causal agent of fire blight disease of apple and pear,

secretes at least four effector-like proteins in a T3 SS-dependent manner (41). Of these,

35

only the effector called DspA or DspE (henceforth DspE) is known to be required for
pathogenicity (1, 5, 20, 55). At 189 kDa, DspE is an unusually large secreted effector
protein that interacts with at least four apple proteins (38). The functional role of DspE in
apple pathogenesis is unknown, although it has been shown to suppress salicylic acid-
mediated host defenses and cause necrotic cell death in host and nonhost plants (7, 16).
DspE binds directly to a Class I T3SS chaperone protein, DspF (20), which is also
apparently required for efficient intrabacterial DspE production or stability of DspE,
although it is not thought to affect transcription levels (21). Likewise, the related
chaperone protein WtsF from Pantoea stewartii subsp. stewartii is involved in the
temperature-dependent stability and secretion of WtsE, a DspE homolog (25). Thus,
initial analyses of DspF-like chaperones have been performed; however, to our
knowledge, a detailed analysis of the chaperone-effector interaction and of sequences in
the effector important to chaperone binding have not been conducted.

We hypothesized that sequence features in the N-tenninus of DspE would affect
the binding of DspF, with the caveat that additional sequences independent of DspF
binding could also affect the translocation of DspE. In this study, we used a combination
of secretion and translocation assays, virulence studies, and protein-protein interaction
studies to characterize the minimal and optimal requirements for secretion, translocation,
and chaperone binding in the N-terminus of DspE. We demonstrate that the N-terrrrinus
of DspE interacts with DspF and identify specific sequences that contribute to this
interaction. Finally, we show that site-directed and deletion mutations of DspE that
abolish DspF binding to the N-terminus do not abolish the translocation of N-terrninal

DspE reporter firsions into host cells.

36

MATERIALS and METHODS
Genetic manipulations and analyses. The bacterial strains and plasmids utilized in this
study are listed in Table 1. Primers are listed in supplementary table 81. Genetic
manipulations were performed using standard techniques and protocols (2). The inserts of
all plasmid constructs were sequenced at the Michigan State University genomics facility.
Sequence alignments were performed using T-Coffee (42,
http://www.ch.embnet.org/software/TCoffeehtml). Secondary structural prediction was
performed by Psipred (29, http://bioinf.cs.ucl.ac.uk/psipred/) and supported by other
secondary structure prediction programs.
Deletion mutagenesis of dspE and dspF. E. amylovora nonpolar chromosomal deletion
mutants were constructed according to the Red recombinase method (15). Briefly, the
primers DspE_F1 and DspE_R1, and DspF_F 1 and DspF_Rl were used to generate
recombination fi'agrnents consisting of 50-nucleotide homology arms of the dspE or dspF
genes flanking the neomycin phosphotransferase cassette from the plasmid pKD4. PCR
products were purified, concentrated, and electroporated into Ea1189 expressing
recombinase genes from the helper plasmid pKD46. Mutants were selected on LB agar
media amended with 50 pg mL-l kanamycin. Single gene disruption was confirmed by
PCR, sequencing, and functional complementation. Virulence assays were conducted
using immature pears as previously described (54).
Cloning of dspE and derivative deletion mutants. To create pLRTS, a 6.5-kb stretch
encompassing the dspE and dspF orfs was amplified using primers DFl and DspF-R and
the Expand Long Template PCR system (Roche Diagnostics; Indianapolis, IN), according

to the manufacturer’s instructions. The product was cloned into pGem T-Easy, and

37

complementation ofEa1189AdspE and AdspF was confirmed. To create site-directed and
deletion mutants of the full-length DspE, fragments were amplified fiom pLRT113,
pLRT114, and pLRT116 using primers DFl -Apa and DR7 and cloned into the ApaI and
BsmI sites of pLRT5.

Site-directed alanine replacement and deletion mutagenesis. Alanine replacement
constructs of DspE (pLRT 113, 114, 116, 149, 150, 151, and 152) and deletions of the
DspE N-terminus (pLRT109) were created using the SLIM mutagenesis technique (14).
A set of two short primers were designed flanking each mutation site. In addition, two
tailed primers were designed consisting of the short primer sequences with a 5’ tail
encoding a stretch of 10 alanines. Mutagenesis primer sequences are shown in
supplementary figure S1; FT and RT denote tailed mutagenesis primers and FS and RS
denote short flanking primers for each mutagenesis site. The four primers were combined
in a PCR reaction to amplify the entire plasmid pLRT14 with complementary overhang
ends. The PCR product was digested with DpnI to remove the template, the
complementary ends were hybridized together, and the products were transformed into
Escherichia coli strain DHSOL Plasmids were extracted from the transforrnants and
screened by digestion to test for the addition of a Sad] site incorporated into the alanine
linker. Inserts were sequenced using the primers DFl and CyaAR. The sequenced alanine
replacement or deletion fragments were amplified, digested, and ligated into pMJ H20 to
create pLRT113, 114, and 116, into pGilda to create pLRT149, 150, 151, and 152, and
into pLRT5 to create pLRT137, 138, and 140.

The CyaA translocation assay. N-terrninal DspE fragments of increasing length,

preceded by the native promoter, were amplified with primers DFl and DR] through

38

DR8 and cloned into the XbaI and SmaI sites of the plasmid pMJH20 in fusion with
residues 2-402 of the adenylate cyclase CyaA. DspE-CyaA constructs were introduced
into Ea1189AdspE and Ea1189AdspF. Stable fusion protein expression was confirmed
via western blot with anti-CyaA antibody (Santa Cruz Biotechnology; Santa Cruz, CA),
and in-vitro calmodulin-dependent adenylate cyclase activity of DspE(1-203)-CyaA was
confirmed using a published assay (46). Overnight cultures were suspended to 6 x 10"8
cfu/mL in phosphate-buffered saline and infiltrated into the youngest three fully
expanded leaves of 8-wk old Nicotiana tabacum cv. Samsun plants. Each treatment was
infiltrated into three replicate leaves on different plants. Two l-cm leaf disks were
collected and flash-frozen at 20 h post-inoculation. CAMP was extracted according to the
method of Casper-Lindley et al. (12); leaf disks were ground in liquid nitrogen and
resuspended in 325 uL HClO4 (1.1 M), and 280 uL of the supematants was neutralized
with 37.5 uL K2CO3 (6M). cAMP levels in the resulting supematants were measured
using the cyclic AMP EIA kit (Cayman Chemical Co.; Ann Arbor, MI) according to the
manufacturer’s instructions. Protein levels in leaf pellets were measured using the Bio-
Rad protein assay. Statistical analyses were done using a one-way analysis of variance
and a mean separation was accomplished using Fisher’s protected least significant
difference test.

Yeast two-hybrid assays. dspE fragments were cloned in fusion with the LexA DNA
binding domain in the BamHI and XhoI sites of the bait vector pGilda (Clontech;
Mountain View, CA). dspF was cloned into the BamHI and EcoRI sites of the prey vector
pB42AD. Bait and prey constructs were co-transfonned into the Saccharomyces

cerevisiae strain EGY48 p80(pLacZ) using the Yeast Transformation Kit (Zymo

39

Research Corporation; Orange, CA). Transforrnants were selected on minimal SD agar
amended with -Ura/-His/-Trp dropout supplement (Clontech). Reporter expression was
visualized on minimal SD-galactose agar amended with —Ura/-His/-Trp/-Leu dropout
supplement and 80 pg mL-l 5-bromo-4-chloro-3-indolyl-beta-D-galactopyranoside (X-
gal). Stable expression of prey constructs was confirmed by Western Blot using anti-
LexA antibody (Millipore).

Secretion assays. 50 mL overnight cultures of Eal l89AdspE strains expressing DspE-
CyaA fusions were grown overnight in LB broth, pelleted by centrifugation, washed with
40 mL water, and resuspended in 20 mL Hrp-inducing minimal medium, pH 5.7 (27).
Cultures were induced for 9 h at room temperature. Induced cultures were pelleted for 20
min at 2600xg. The uppermost 35 mL of supernatant was removed, amended with PMSF
to a concentration of 0.5 mM, centrifuged again, and the uppermost 30 mL of supernatant
was removed and centrifuged a third time. The resulting supernatant was concentrated to
75 pL (400x) using the Amicon 15 mL centrifugal filter unit (10 kDa MWCO) and the
Ultrafree 0.5 mL microcentrifugal filter unit (Millipore; Billerica, MA). Protein
concentrations were measured by Bradford assay and adjusted to 200 pg/mL; 3 pg of
total supernatant protein was analyzed by western blot using anti-CyaA antibody (Santa
Cruz Biotechnology, Santa Cruz, CA). Anti-NPTH antibody (US Biological,
Swarnpscott, MA) was used as a lysis control.

In-vitro protein interaction assay. dspF was cloned into the BamHI and XhoI sites of
the vector pGex 4T-l (GE Healthcare; Piscataway, NJ) in fusion with an N-terminal

glutathione S-transferase tag. E. coli DHSor cells carrying the firsions were induced in 1

mM isopropyl B-D-l-thiogalactopyranoside for 4 h at 37° C. To induce 15 mL overnight

40

cultures of E. amylovora, Eal 1 89AdspF cells harboring DspE-CyaA fusion vectors were
washed twice with water and resuspended in 30 mL modified Hrp-inducing minimal
medium, and incubated at 25° C for 6 h. Induced E. coli and E. amylovora cells were
pelleted and lysed in 1 mL B-Per reagent (Therrno-Fisher; Rockford, IL). Lysates were
diluted 1:1 in tris buffered saline (TBS; 20 mM tris, 150 mM NaCl, pH 7.4), cleared by
centrifugation, and amended with phenylrnethanesulphonylfluoride to a final
concentration of 0.5 mM. 0.5 mL bait (DspF-GST) lysate was incubated with 0.5 mL
prey (DspE-CyaA) lysate and 50 pL TBS-equilibrated glutathione agarose (GE Life
Sciences) overnight at 4 °C. The glutathione agarose fiaction was washed eight times in
ten bed volumes of TBS and boiled for 10 minutes in 100 uL SDS loading buffer (2%
sodium dodecyl sulfate, 4 mM EDTA, 5% B-mercaptoethanol, 0.02% bromophenol blue,
and 10% glycerol). 20 pL was resolved on a 12% polyacrylamide gel, transferred to
nitrocellulose, and probed with anti-CyaA or anti-GST antibody (Precision Antibody;
Columbia, MD).

RESULTS
Creation and assessment of a dspF deletion mutant in E. amylovora Ea1189.
Gaudriault et al. reported that a dspF mutant of E. amylovora was weakly pathogenic on
pear seedlings (21). To confirm these results and to determine whether a dspF mutant
also caused symptoms on immature pear fruits, E. amylovora knockout strains
Ea1189AdspE and Ea1189AdspF were constructed as described in Materials and
Methods. Ea1189AdspF was complemented with the plasmid pLRTl98, a construct
containing the dspF gene and its promoter sequence. In a virulence assay on immature

pear, Ea1189AdspE caused no symptoms, while Ea1189AdspF caused symptoms that

41

were dramatically reduced compared with that of the wild type strain (Figure 2-1A). The
virulence phenotype of Eal 189AdspF was complemented by pLRTl98. When infiltrated
into leaves of Nicotiana tabacum cv. Samsun, wild type Ea1189 caused complete HR-like
cell collapse within 24 h. dspE and Ea1189AdspF caused little to no necrosis, while

Eal l89AdspF(pLRT198) caused near wild-type levels of cell collapse (data not shown).
Within 48 h, Ea1189AdspE and Eal 189AdspF caused a cell collapse, but this did not
appear as severe as that caused by wild-type Ea1189 (Figure 2-1B). Taken together, these
results indicate that while DspE is required for E. amylovora pathogenicity on pear, its
chaperone DspF is not. However, Eal l89AdspF causes greatly diminished symptoms on

pear and elicits a response on tobacco visually identical to that of Eal l89AdspE.

42

    

Eal 189 AdspF AdspF
(pLRT] 98)

 

Figure 2-1. Pathogenicity and cell death phenotypes of Erwinia amylovora Ea1189,

Eal 189MspF, and Ea1189AdspF (pLRTl 98). A. Symptom development on immature
pears 5 days afler inoculation with 3 x 103 cfir of three E. amylovora strains. B. Tobacco
leaf cell death 48 hours afier infiltration with suspensions ofEa1189 (l), Ea1189AdspE
(2), Ea1189AdspF (3), and Ea1189AdspF (pLRTl98) (4).

43

Kinetics of DspE(1-203) -CyaA translocation into tobacco leaves. Bocsanczy et al.
recently demonstrated that the N-terrninal 737 amino acids of DspE confer translocation
of the reporter CyaA into tobacco leaf cells from wild-type E. amylovora, and that
translocation levels increase steadily between two and 10 h post-infiltration (4).
Adenylate cyclase levels cannot be measured from tobacco infiltrated with wild-type E.
amylovora afier 12 h due to rapid cell collapse. To determine the time required for
maximum levels of DspE translocation into tobacco cells, and to determine the kinetics of
effector translocation into tobacco over a more extended time period, we examined
translocation levels of a DspE-CyaA firsion from Ea1189AdspE. Eal 189AdspE causes a
delayed and greatly reduced cell collapse, allowing an extended period of time in which
translocation assays can be performed. A fusion of DspE residues 1-203 to the CyaA
reporter fragment was constructed and introduced into Ea1189AdspE. The resulting strain
was infiltrated into tobacco leaves, and CAMP accumulation in the leaves was measured
after various timepoints. To control for the possibility that bacterial infiltration itself
causes increase in CAMP levels, we also measured CAMP accumulation elicited by

Eal l89AdspE carrying the parent vector pMJH20. CAMP accumulation of DspE(l-203)-
CyaA expressed by Eal l 89AT3 SS was measured to confirm that DspE translocation is
dependent on the T3 SS. Timepoints were optimized by preliminary studies indicating that
a significant increase in CAMP levels was first detectable between 2 and 8 h post-
infiltration with Eal 189AdspE expressing DspE(l-203)-CyaA (data not shown). CAMP
level increased significantly between 0 and 8 h, increased to a greater degree between 7

and 11 h, and remained steady between 11 and 24 (Figure 2-2). CAMP levels did not

44

increase in leaves infiltrated with the pMJH20 or type HI secretion-deficient controls.
These results show that, in the absence of cell collapse, E. amylovora requires up to 11 h
to reach maximum levels of effector translocation. CAMP levels can fluctuate transiently
in plant cells (28, 31), so the persistence of high CAMP levels after 24 h suggests that E.
amylovora maintains maximum effector expression and translocation levels for at least a
day in a non-host system. Based on these results, leaf samples in all subsequent studies

were collected at 16 h post-infiltration.

25 - +TBSS
+pMJH20
2° +1-203-CyaA

    

 

 

 

 

 

 

 

.5

o

E

fl-

2215-

E

510

3.5

:. fl

ya—

0 ' t

Figure 2-2. Kinetics of CAMP accumulation after infiltration into tobacco leaves of
Ea1189AdspE +DspE(1-203)-CyaA ( A), Eal189AT3SS+DspE(1-203)-CyaA (6), and
Ea1189AdspE + pMJH20 (I). Samples were collected at 0, 8, 11, and 24 hours. Data
points represent the mean and standard error of three replicate samples.

Mapping of the secretion and translocation signals of DspE. The minimal secretion
signal of most model type III effector proteins is shorter than the minimal sequence
required for translocation into host cells, and this translocation signal often coincides
with the Chaperone binding domain (13). The N-terminal 15, 33, 51, 77, 101, 150, 203,

and 737 amino acids of DspE were fused to the CyaA reporter protein to map the

minimal secretion and translocation signal of DspE. Fusion constructs were introduced

45

into Ea1189AdspE, and stable expression in Hrp-inducing minimal medium was
confirmed by western blot (Figure 3A).

To determine the minimal signal for secretion of DspE, secretion of the shortest
three firsion proteins from Ea1189AdspE was assayed by western blot of supematants as
described in Materials and Methods. While DspE(1-51)-CyaA was strongly detected in
the E. amylovora supernatant fi'action, secretion ostpE(1-15)-CyaA and DspE(1-33)-
CyaA could only be detected very faintly (Figure 3B). These results suggest that while
the N-terrninal 15 amino acids of DspE are sufficient to allow residual levels of secretion,
51 amino acids are required for substantial levels of secretion into culture medium.
Fusions of longer DspE fragments were secreted at levels similar to that ostpE(1-51)-
CyaA (data not shown).

Translocation levels of the CyaA firsions were determined by assaying for CAMP
accumulation in leaf disks. DspE(l-15)-CyaA and DspE(l-33)-CyaA did not cause an
increase in CAMP levels in tobacco compared with the vector control, but fiagments of
51 amino acids or longer caused an increase in CAMP levels that were significantly
different from the pMJ H20 control (Figure 3C). These results indicate that the minimal
signal for translocation is within the first 51 amino acids, but not the first 33 amino acids,
of DspE. Fragments of 150 and 737 amino acids caused significantly greater levels of
CAMP accumulation in plant cells than the first 51 amino acids, indicating that 51 amino
acids are not sufficient for optimal translocation. Instead, optimal levels of translocation
occurred with fragments of 151 amino acids or greater, even though, for unknown
reasons, the DspE(1-101)-CyaA fragment was translocated at lower levels than DspE(1-

51)-CyaA. The experiments were repeated three additional times with consistent results.

46

A 15 33 51 77 101 150 203

g.-
-..‘-

B 15 33 51 15 33 51
we» II! -awa
anti-NntII m -29 kDa

P

    

#-

O
m
j 2

 

 

 

 

E. *t*

3335-

E30—

00251 **

5201

““151

5...

~51 I

g. . l<-7_‘_._-_
€5$§slis§§§§m
veeoiiiizw
GGCCSESRQGZ
TTF‘TTTC‘BW 3°
53:33:35 4.33
mmmmmtgmm 5'3
%%%%%m%% all-J
QQQQQCQQ a

Figure 2-3. Mapping of the secretion and translocation domain of DspE. A. The first 15,
33, 51, 77, 101, 150, and 203 amino acids of DspE were expressed in firsion with amino
acids 2-402 of CyaA, and stable fusion expression in Eal l89AdspE was confirmed by
western blot. B. Western blot of CyaA fusion proteins in supematants (SN) and pellets
(P) of Eal 189MspE expressing the first 15, 33, and 51 amino acids of DspE fused to
CyaA. Samples were probed with anti-NptII antibody as a cell lysis control. C.
Translocation levels of N-terrninal fragments of DspE as measured by CAMP
accumulation in tobacco leaves 16 h after infiltration with 10A8 Cfu/mL Ea1189AdspE
cells expressing DspE-CyaA fusions. Bars represent mean CAMP levels of three leaves
on separate plants, plus or minus standard error. Asterisks above bars denote values
significantly different fi'om that of the pMJH20 control (P<0.05). Asterisks above
brackets denote values significantly different from one another (P<0.05).

47

Mutational analysis of conserved segments of the N-terminal 50 amino acids of
DspE. To identify features that might be important for translocation within the first 50
residues of DspE, the N-terminal fiagment of DspE was aligned with that of seven
homologs from related Erwinia species as well as Pantoea agglomerans pv. gypsophilae,
Pantoea stewartii, and Pseudomonas syringae B728A (Figure 2-4A). While the N-
terminal 50 amino acids of these proteins did not share a high degree of conservation,
there were two highly conserved stretches within residues 1-11 and 28-40. To determine
whether specific residues within the translocation signal of E. amylovora might play an
important role as part of the translocation signal, we used site-directed ligase independent
mutagenesis (SLIM) to replace amino acids 2-10, 11-20, and 31-40 with stretches of
consecutive alanine residues on the plasmid pLRT14, yielding plasmids pLRT113, 114,
and 116. Alanine stretch replacement is a common strategy to assess the functional role
of a sequence without imposing extreme steric constraints (33). Due to the nature of the
SLIM technique, the number of amino acids varied by one or two; residues 2-10 and 11-
20 were replaced with ten alarrines, while 31-40 were replaced with eight alanine
residues. Alanine-replacement mutants were subcloned into pLRT5, a full length, hi gh-
copy Clone of dspE, to create plasmids pLRT137, 138, and 140. These constructs were
tested for complementation of the virulence phenotype of Ea1189AdspE on immature
pear. While dspE (2-10A) did not complement Ea1189AdspE on immature pear after four
days, dspE (l 1-20A) and dspE (31-40A) complemented the mutant to levels similar to
that of the parent vector (Figure 2-4B), indicating that only residues 2-10 were required
for translocation. To measure translocation levels of the site-directed mutants, DspE (l-

203; 2-10=10A)-CyaA, DspE (1-203; 1 1-20=10A)-CyaA, and DspE (1-203; 31-40=8A)-

48

CyaA were amplified and re-Cloned into the CyaA fusion vector pMJ H20. Constructs
were introduced into Ea1189AdspE, expression was confirmed, and levels of
translocation into tobacco were assayed. Translocation levels of DspE (1-203; 11-
20=10A)-CyaA were not significantly different than those of DspE(l -203)-CyaA, but
DspE (1-203; 31-40=8A)-CyaA was consistently translocated at levels significantly
lower than the wild-type. DspE (1-203; 2-10= 10A)-CyaA was translocated at greatly
reduced levels, although CAMP levels were significantly higher than the vector control
(Figure 2-4C). We did not obtain an alanine-replacement mutant of the conserved
sequences between residues 41 and 50, but deletion of residues 35-125 from DspE(1-
203)-CyaA did not have a significant impact on translocation levels (data not shown),
indicating that these residues are not important for protein translocation. Together, these
results suggest that the first 10 amino acid residues of DspE are required for translocation
into host cells and pathogenicity, and that other portions of the translocation signal, while
tolerant to mutagenesis, could have a small but additive effect on translocation levels.
The fact that decreases in translocation levels did not correspond to a decrease in
virulence suggests that small reductions in DspE translocation do not adversely affect

virulence levels.

49

>

   
 
  

Ea1189 "iELKSLGu VHTAAHNPVGHGV j“
E-pyr- Mimosa; AVQIAAHNPVGQGV :
E- tasm fivr‘mLoe'rrl~ IQTATHGPAGPG ‘ _‘
D. dad. ”RENLHAfi TTVQNVENPNNSTIPP"
P. gyp. r
Pss. “GKRLLKW FQKSRAPRQSAARPP

lesion diamet (m
H H
c N A ex ea c N
h?
_ v
A)
:

AdspE wt 2-10 11-20 31-40

h

C bd

AO\®°.,
ccoac

pg cAMP/ug protein
N
O

 

O

AdspE Wt 2-10 11-20 31-40

Figure 2-4. Analysis of alanine replacement mutants of amino acids 2-40 of DspE. A.
Sequence alignment of amino acids 1-50 of DspE and homologs from other
phytopathogenic bacteria. Bacterial species and Genbank accessions are as follows: E.
amylovora (AAC04850) E. pyrifaliae (AY530755), E. tasmaniensis (YP_001906490),
Pantoea agglomerans (AAF 76343), P. stewartii (AAG01467), and Pseudomonas
syringae B728A (YP_234280). B. Lesion diameters of immature pears infected with
Eal l89AdspE and Eal l89AdspE carrying pLRT5, a full-length Clone expressing DspE,
or pLRT137, 138, and 140, pLRT5 variants expressing DspE with amino acids 2-10, 11-
20, or 31-40 replaced with stretches of alanine. C. CAMP levels of tobacco leaves
infiltrated with Eal l 89AdspE carrying the pMJH20 parent vector, pLRT l4 (DspE (l-
203)-CyaA) or alanine replacement variants pLRT113, 114, and 116. Bars represent
means plus or minus standard error. Letters above bars denote groups with significant
differences (P<0.05).

50

DspF binds to DspE between amino acids 51-100. Affinity blot experiments
have previously detected a direct interaction between DspE and DspF (21). To date,
almost all reported T3 SS chaperones have interacted with cognate effector proteins via a
single Chaperone binding domain near the N-terminus (43). However, recent studies
showed that some Chaperones of Pseudomonas effectors may bind to the middle third of
an effector protein (24). Here, a yeast two-hybrid (Y2H) approach was used to identify
the portion of the N-terminus of DspE that interacts with the secretion Chaperone DspF.
N-terrninal DspE fragments of varying lengths were fused to the LexA DNA binding
domain and tested for interaction with a B42-HA-DspF fusion in yeast as described in
Materials and Methods. The N-tenninal 800, 600, 400, 200, 150, and 100 amino acids of
DspE all interacted strongly with B42-HA-DspF, but the N-tenninal 50 amino acids did
not (Figure 2-5A). Further analyses showed that residues 51-100 of DspE were sufficient
for strong interaction with DspF, while deletion of residues 51-100 from LexA-DspE(1-
150) abolished the interaction with DspF (Figure 2-5A). Constructs expressing LexA -
DspE(25-75) and LexA -DspE(75-125) interacted very weakly with B42-HA-DspF (data
not shown), suggesting that the Chaperone binding domain is located squarely within
residues 51-100 of DspE. Deletion of residues 51-100 from LexA -DspE(l-800) reduced
the strength of the interaction with DspF, but did not abolish the interaction. Further
analysis mapped additional DspF yeast two-hybrid interaction sites to residues 151-200,
401-450, 651-700, and 1000-1200 of DspE (data not shown). Because these secondary
candidate chaperone binding sites interacted weakly with DspF, and because the N-
terrninal binding site in residues 51-100 corresponds to the Chaperone binding sites of the

majority of Characterized type III effector proteins, this study is focused on

51

Characterization of DspE residues 51 to 100 as a probable requirement for DspF
chaperone firnction.

To confirm the DspE-DspF interaction in vitro, a glutathione S-transferase pull-
down assay was performed. A GST-DspF fusion immobilized on gluthione agarose was
tested for the ability to pull down DspE(1-l 50)-CyaA and DspE(1-150A51-100)-CyaA.
DspE(1-150)-CyaA was pulled down by GST-DspF in vitro, but DspE(l-150A51-100)-
CyaA was not (Figure 2-5B). This supports the yeast two-hybrid evidence that residues

51 to 100 are required for interaction of the DspE N-terminus with DspF.

52

A

 

 

 

 

 

 

 

 

 

 

 

 

Interaction
with B42-
LexA - DspE HA-DspF
1-800 ._ "‘ _: :.f::;:“:’: +
1-600 +
1-400 4-
1-200- +
1-150 4.
1-100 +
1-50 .
51-100 4.
101-150 .
1-150(A51-100) l:-:l _ _ - .. .. _ -. _ p _ -
1-800(A51-100) . ‘* ..".'t:-"";;...'i:":_::.; ‘_'. .'.'-_::*‘. r; ’ "__ :iffg 5 4.].
B.
DspF-GST + + -
DspE 1-150-CyaA + - +

 

DspE 1-150(A51-100)-CyaA . + -

 

anti-CyaA __
Bound fraction _ ,
anti-GST B - I
Input anti-CyaA I H! ”4

Figure 2-5. Mapping of the DspF- interacting regions in the N-terrninal half of DspE. A.
Yeast two-hybrid analysis of the N-terrninal 800 amino acids of DspE. LexA-DspE
fusions were co-expressed with fusions of DspF with B42-HA. B. Pull-down analysis of
DspE (1-150)-CyaA and DspE (1-150 A51-100)-CyaA with DspF-GST. Lysates of

Eal 189AdspE carrying plasmids pLRT13 (DspE(1-150)-CyaA) or pLRT109 (DspE (l-
150A51-100)-CyaA) were incubated overnight with lysates of E. coli overexpressing
GST-DspF and glutathione beads, or with glutathione beads alone.

53

DspE residues 51-60 and 71-85 affect the yeast two-hybrid interaction between
DspE (1-150) and DspF. Crystallization studies and analyses of amino acid sequences
have identified secondary structures and specific residues thought to be involved in
Chaperone-effector binding. Lilic and Stebbins showed that deletions of a single 3-residue
motif, the B—motif, in the Chaperone binding domains of several effector proteins led to a
loss of Chaperone-effector binding (35). Co-crystallization studies have indicated that
hydrophobic grooves in Class I Chaperones interact with one or more a-helices on their
cognate effector protein, but few functional analyses have been performed of specific
Chaperone binding features on the effector (l 7, 35). To identify DspE sequences within
residues 51-100 that were important for DspF binding, and to determine whether single or
multiple sequences were involved, an alanine replacement strategy was employed. SLIM
mutagenesis was used to replace DspE residues 51-60, 61-70, 71-85, and 86-100 with
stretches of 8 to 10 alanines as described in Materials and Methods. Residues 71-85 were
mutated at once because they correspond to a putative a-helix-rich region predicted by
the secondary structure prediction program PsiPred. In addition, we identified a potential
B—motif consisting of residues L79, L83, and K86 which would be functionally disrupted
by the 71-85 mutant (Figure 2-6A). Alanine replacement constructs are described in
Table 2-1. DspE residues 1-150 containing site-directed mutants in the 51-100 region
were fused to the LexA protein and tested for interaction with B42-HA-DspF in a yeast
two-hybrid assay. Replacement of residues 61-70 and 86-100 had no apparent effect on
the interaction between LexA-DspE (1-150) and B42-HA-DspF (Figure 2-6B). However,
replacement of residues 51-60 appeared to cause a slight decrease in yeast growth and

color development, while replacement of residues 71-85 caused a drastic reduction in

54

growth and color development (Figure 2-6B). Faint color development was still visible
when the 71-85 alanine replacement strain was plated in high concentrations on a less-
selective medium amended with leucine, although this color development was not seen in
a double mutant in which residues 51-60 and 73-82 were replaced with 7 and 10 alanines,
respectively (data not shown). These results demonstrate that residues 71 -85, which
contain a putative a-helix and chaperone binding motif, contribute strongly to the DspE-

DspF interaction, and that residues 51-60 may also contribute to this interaction.

A

   

 

 

NRGKMPRIHQPSTAADGI SAAHQQKKS FSERGCEGTEKFSRSAPQGQPGTT
50 60 70 so 90 100

B LexA-DspE

 

:3 1-150
1-150 [551-100
. 1-150; 51-60=7A
0 1-150; 61-70=10A
1-150; 71-85= 8A
,0 1-150; 86-100=10A
, pGilda + B42-HA-DspF
‘ pLexA -53 + pB42AD-T

 

Figure 2-6. Alanine-replacement mutagenesis of the N-tenninal Chaperone-binding
domain of DspE. A. Amino acid sequence of residues 51-100. Cylinders represent
predicted or-helices. Underlined residues denote residues of a potential B-motif. B. Yeast
two-hybrid interaction between alanine-replacement variants of LexA-DspE (1-150) and
B42-HA-DspF. The negative control strain carries the empty pGilda bait vector and B42-
HA-DspF; the positive control strain canies pLexA-53 and pB42AD-T.

55

The DspF-binding domain is not required for translocation of the N-terminus of
DspE. The N-tenninal chaperone binding domain is required for translocation of many
type IH-secreted effector proteins (26, 32). To determine whether the Chaperone binding
domain in residues 51-100 of DspE affects protein translocation levels, a plasmid
expressing DspE(l -150)-CyaA lacking residues 51-100 was constructed. CyaA
expression levels in Ea1189AdspE were confirmed by western blot to be equivalent to
Eal l89AdspF, and translocation levels into tobacco cells were tested. Deletion of amino
acids 51-100 from residues 1-150 did not cause a change in translocation levels (Figure
2-7A). To control for the possibility that residual Chaperone binding sites outside the 51-
100 range were not affecting translocation levels, we also tested constructs with residues
35 to 125 deleted, and these were not reduced in translocation (data not shown). We also
tested constructs of DspE(l -73 7)-CyaA with residues 51-100 deleted, and found no effect
of residues 51-100 on the longer fragment (data not shown). Although we hypothesized
that translocation of the N-terminus of DspE would be DspF-dependent, DspE(1-150)-
CyaA was translocated from Ea1189AdspE and Ea1189AdspF at similar levels (Figure 2-
7B). PCR and pear virulence assays were used to confirm that the bacterial samples were
not contaminated with any DspF-expressing cells. These results suggest that the
accumulation and translocation of plasmid-expressed N-terrninal fragments of DspE is
not dependent on DspF binding, and that the Chaperone binding domain itself does not

inhibit or destabilize the N-terrninus of DspE in the absence of DspF.

56

45 A 353
a 40‘
“g 35- 3“
Eso- 25‘
DD
E1251 20.
z 20- 15.
315- 10
“'10-

5 5‘

o 0

 

 

DspE(l-150)-CyaA
DspE(1-150A51-100)-
CyaA

pMJH20

DspE(1-150)-CyaA
Ea1189AdspE

DspE(l-150)-CyaA
Ea1189AdspF
pMJ H20

Figure 2-7. Functional analysis of amino acids 51-100 of DspE. A. CAMP accumulation
in tobacco leaves infiltrated with Eal l89AdspE expressing DspE(1-150)-CyaA and DspE
(1-150A 51-100)-CyaA. B. CAMP accumulation in leaves infiltrated with Ea1189AdspE
and Ea1189AdspF expressing DspE(l-150)-CyaA. Results are the means and standard
errors from three replicate leaves on separate plants.

57

DISCUSSION

This work reports a multi-faceted analysis of the N-terrninal 200 amino acids of
the E. amylovora secreted effector protein DspE. Three areas were examined with regards
to functionality: requirements for secretion, requirements for translocation, and
requirements for binding to the Chaperone DspF. Our results demonstrated that, while
DspF is required for virulence and binds to a specific site on the N-terminus of DspE, the
N-terrrrinal minimal Chaperone binding domain is not required for translocation of the
DspE N-terminus .

We report a time-course analysis of DspE(1-203 )-CyaA translocation into tobacco
leaves. CAMP accumulation in tobacco cells was detected at moderate levels after 8 h, but
had reached maximum levels within 11 h. This supports the recent finding of Boczanscy
et al. that, after a short initial lag period, DspE translocation levels increase continuously
up to 10 h post-inoculation (4). hrp gene promoters require about 6 h for expression
during induction in tobacco (52), which could explain the delay between leaf disk
infiltration and CAMP accumulation. In repeated studies, DspE-CyaA induced CAMP
levels remained at maximum levels until at least 24 h post-inoculation. Work by Schecter
et al. demonstrated that Xanthomonas protein translocation (CAMP accumulation) levels
into nonhost tobacco cells drops precipitously after about 8 h post-infiltration (2004), and
Wei et al. showed a decrease in hrp gene transcription in E. amylovora during the same
timefrarne (52). Our findings indicate that in the absence of DspE-induced cell death,
maximum translocation levels can be maintained for extended periods of time.

Studies on a variety of Type III- secreted effectors in animal and plant pathogens

have led to a model of two functional T3SS signals- a short N-terminal fi'agment

58

sufficient for secretion into culture media, and a longer fragment required for
translocation into host cells (40, 47). Here, we found that the N-terminal 15 and 33 amino
acids of DspE were sufficient to allow very low levels of secretion of the CyaA reporter
gene into Hrp-inducing minimal medium, while DspE(l-51)-CyaA was strongly secreted.
These results suggest that DspE does have a minimal secretion signal separate from the
translocation signal, but that optimal levels of secretion require a functional translocation
domain. Previous work in plant pathogens indicated that while the X. campestris effector
AvrBsZ has separate secretion and translocation signals (12, 40), the secretion and
translocation requirements of the P. syringae effector Ath0 are not separate (44).

This study found that translocation of the CyaA reporter protein required greater
than 33 N-terrninal amino acids of DspE, and that translocation levels increased with
greater than 51 amino acids. This finding echoes previous work which reported minimal
and optimal translocation signals of the secreted effectors AvrPto, AvrBsZ, SspH2, and
YopE, among others (12, 39, 44, 45). These findings suggest that the translocation signal
itself could have multiple components. The basis for increasing secretion competence of
longer fi'agments is not known. In some cases, the structural conformation of effector or
effector-reporter hybrids may be more secretion compatible in longer fragments. In some
cases, longer gene fragments may allow greater levels of chaperone binding.

Mutational studies on the minimal translocation domain demonstrated that the
sequence in residues 2-10 is required for all but the smallest detectable levels of
translocation, a finding supported by previous studies on type III secretion signalling (6,
45). Alanine replacement of residues 11-20 caused a minor, but not significant, decrease

in translocation levels. Mutation of residues 31-40, a stretch rich in serine, caused a 40%

59

to 60% reduction in translocation levels of DspE(1-203)-CyaA in multiple studies.
Residues 31-40 contain a stretch of five highly conserved serines. A high serine content
(>10%) is a defining feature of the N-terminus of T3 SS effectors in plant pathogens,
although the N-terminus of DspE does not contain any serines outside of residues 31-40.
While at least one serine residue in the N-terminus of the Pseudomonas syringae effector
protein Aerpt2 was involved in protein stability, a fi'agment containing the serine-rich
motif SSASS increased secretion or translocation efficiency without affecting levels of
protein accumulation (51). Mutagenesis of residues 31-40 did not affect accumulation of
DspE(l-203)-CyaA in western blot experiments. Further studies are needed to determine
whether the serine-rich stretch of DspE itself contributes to translocation efficiency.
Yeast-two hybrid and GST pulldown analyses in this study identified a DspF
binding domain within the first 150 residues of DspE, specifically within residues 51 to
100. Fifty amino acid-domain is comparable to the sizes of other characterized Chaperone
binding domains (3). Disruption of residues 51-60 and 71-86, respectively, strongly
interfered with this interaction. Yeast two-hybrid studies also pointed to the possibility of
several downstream sites of Chaperone interaction, but these interactions have not been
confirmed in vitro. To our knowledge, there are no previous reports of multiple distinct
Chaperone binding sites on an effector. Bogdanove et al. reported that a deletion in the C-
terminal half of DspE is complemented with a cosmid expressing the overlapping C-
terminal fragment, suggesting that DspE forms two independently folding domains (5).
The mechanism of secretion of the C-terrninal fragment is not known, but given the large
size of DspE and the possibility of multiple secreted domains, multiple DspF interaction

sites could be important for stabilization and secretion targeting of downstream portions

60

of the protein. Although this study focused on the N-terminus of DspE, it will be
interesting in future studies to determine whether DspF binds to multiple downstream
sites of DspE in a functionally significant manner.

Deletion of amino acids 51-100 of DspE or deletion of dspF did not result in a
decrease in translocation levels ostpE(1-150)-CyaA. Furthermore, DspE (1-150)-CyaA
showed no signs of destabilization in the absence of the Chaperone binding domain or
DspF. These results would support a model in which the chaperone binding domain does
not have an effect on translocation of a stably expressed N-terrninus of DspE. This was
surprising, since the N-terrninal Chaperone binding domain is required for the T3SS-
dependent translocation of the N-terminus of many effectors (43). One possibility for this
difference is that a primary role of the chaperone binding domain is to protect DspE fiom
destabilization targeted to the C-terminal portions of the protein, as does the Chaperone
binding domain of HothoM (36). If this were the case, deletion of the Chaperone binding
domain would have little effect on translocation of N-terrninal DspE fragments.
Alternatively, expression of the N-terrninal CyaA fusion of DspE from the medium-copy
vector pMJH20 could overcome the need for the Chaperone binding domain, as multiple
copies of the gene could potentially saturate proteolytic degradation or help DspE
compete against other effectors in a hypothetical secretion hierarchy. However, it would
be very unlikely for ectopic expression of the CyaA fusion to abolish the importance of
Chaperone binding, as effectors such as SspA and YSCP require the Chaperone for
translocation even when expressed as a CyaA firsion from pMJH20-derived vectors (8,
37). This would suggest that the role of the N-terminal Chaperone binding domain of

DspE differs from that of YscP and SspA, and our current work is directed at determining

61

whether alternate chaperones or Chaperone binding sites contribute to DspE secretion.
Together, our results indicate that the N-terminal 150 amino acids of DspE are secretion-
and translocation-competent even in the absence of DspF binding, and that the Chaperone
binding domain within residues 51-100 of DspE is not inhibitory to translocation in the
absence of DspF. While Chaperone-independent translocation of some effectors has been
documented (50), the virulence phenotype and secretion profile of E. amylovora dspF
mutants demonstrates a strong role for the Chaperone in translocation of wild-type DspE
(21). Studies involving full-length DspE will be helpful in determining the true functional
significance of the N-terminal Chaperone binding domain and its interaction with DspF.
In summary, our results provide evidence that the DspE Chaperone binding
domain and the DspE-DspF interaction have the same basic configuration as other
Characterized Chaperone-effector pairs, including a minimal and optimal translocation
signal and an N-terrninal Chaperone binding domain. However, we have shown that the
N-terrrrinus of DspE has some Characteristics unlike that of many smaller secreted
effectors, in that the N-terminus of DspE required a functional translocation signal for
substantial levels of secretion in vitro, and did not require interaction with DspF for

translocation competence.

62

Table 2-1. Strains and plasmids used in Chapter 2.

 

 

 

 

 

 

 

Strain or plasmid Characteristics Source
Escherichia coli strain

F- 80d1acZ AM15 A(lacZYA-argF)UI 69 endAI
DH50t recAI hst17(rK-mK+) deaR thi-I supE44

gyrA96 relAI h— Invitrogen

Erwinia amylovora strains
Eal 189 Wild type 9
Ea1189AdspE dspE deletion mutant, KmR This study
Ea1189AdspF dspF deletion mutant, KmR This study
Eal 189AT3SS T3SS deletion mutant, KmR 56
pBRRl-MCSS broad host-range cloning vector, GmR 30
Plasmids

pLRT198 pBRRl-MCSS expressing dspF This study
pMJ H20 pWSK29 containing codons 2 to 406 of CyaA, AmpR 39
pLRT8 pMJH20 expressing DspE(l-15)-CyaA This study
pLRT9 pMJ H20 expressing DspE(1—33)-CyaA This study
pLRTl 0 pMJ I120 expressing DspE(l-51)-CyaA This study
pLRTll pMJH20 expressing DspE(1-77)-CyaA This study
pLRT12 pMJH20 expressing DspE(1-lOl)-CyaA This study
pLRT 13 pMJH20 expressing DspE(1-150)-CyaA This study
pLRT14 pMJH20 expressing DspE(l-203)-CyaA This study
pLRT201 pMJH20 expressing DspE(1-737)-CyaA This study
pLRT109 pMJH20 expressing DspE(1-150d. 51-100)-CyaA This study
pLRTl l3 pMJH20 expressing DspE (1-203; 2-10=10A)-CyaA This study
pLRTl 14 pMJH20 expressing DspE (1-203; 1 1-20=10A)-CyaA This study
pLRT116 pMJ H20 expressing DspE (1-203 31-40=8A)-CyaA This study
pGilda HIS3 LexA BD bait vector, AmpR Clontech
pB42AD TRPI B42 AD prey vector, AmpR Clontech
pLexA-53 positive control bait vector, AmpR Clontech
pB42-HA-T positive control prey vector, AmpR Clontech
pLRTl92 LexA-DspE(l-800) This study
pLRT216 LexA-DspE(l-400) This study
pLRT217 LexA-DspE(l-ZOO) This study
pLRT3l LexA-DspE(l-ISO) This study
pLRT30 LexA-DspE(l -1 00) This study
pLRT29 LexA-DspE(l-SO) This study
pLRT32 LexA-DspE(Sl-IOO) This study
pLRT35 LexA-DspE(101-150) This study
pLRT37 LexA-DspE(151-200) This study
pLRT155 LexA-DspE(l-ISO A51-100) This study
pLRT l 49 LexA-DspE(1-203; 51-60=7A) This study
pLRTISO LexA-DspE (1-203; 61-70=10A) This study

 

 

 

63

 

Table 2-1 (cont’d).

 

 

pLRT151 LexA-DspE (1-203; 71-86= 8A) This study

pLRT152 LexA-DspE (1-203; 86-100= 10A) This study

pLRT215 B42-HA-DspF This study

pGEM T-Easy high-copy TA Cloning vector Promega

pLRT5 5.6 kb dspEF locus in pGem T-Easy This study
pLRT5 derivative expressing DspE (1-203; 2-

pLRTl37 10=10A) This study
pLRT5 derivative expressing DspE (1-203; 11-

pLRTl38 20=10A) This study
pLRT5 derivative expressing DspE (1-203; 31-

pLRT140 40=8A) This study

pGEX 4T-1 Bacterial expression vector expressing a GST tag GE Healthcare

pLRT184 pGEX 4T-1 derivative expressing GST-DspF This study

pKD4 AmpR, KmR, mutagenesis cassette template 15

pKD46 AmpR, expresses 1» red recombinase 15

 

 

64

TABLE 2-2. Oligonucleotides used in Chapter 2.

 

Name

sequence

 

DF 1
DR1-15
DR2-33
DR3-51
DR4-77
DRS- l 01
DR6-150
DR7-203
DR8-737
DspF-R
DFl -Apa
CyaA-R
dspE_F1
dspE_R1
dspF_Fl
dspF_Rl
2-1 OFT
2-10FS
2-10RT
2-10RS

1 1-20FT
l 1-20FS
1 l-20RT
1 l-20RS
31—40FT
31-40FS
31-40 RT
31-40 RS
51-100FT
51-100FS
51-100RT
51-100RS
51-60FT
51-60FS
51-60RT
51-60RS
61-70FT
61 -7OFS
61 -7ORT
61-70RS
71 -85FT
71 -85FS

gttctagacctggtgggaaccggttgcag

gtccataatattgtgtactgccgccttgtgtt
tggtttaatattctgctgtaaggcaacaccat

agttcaatattac gatttttgccttctgcc g
Catggaatattgaaggatttcttttgctggtgag

tagccaatatttttgctgtgggtggtacctg
gatctCaatatttttcaccattggccgcccgg

gactgaatattcgggatgtgcggcggtgtaa
catgaatattgaaggagcgaatgtcccctttttc
lattatgccgcgctactctcgtcta

gtgggccccctggtgggaaccggttgcag

tcggcaatcaggctggtggaatg
atggaattaaaatcactgggaactgaacacaaggcggcagtacacacagcgcgattgtgtaggctggagct

ttagctcttcatttccagCCcttccttcttcaaatcagtaagcgcagatgattccggggatccgtcgacc
atgacatCgtcacageagCgggttgaaaggtttttacagtatttctccgcgcgattgtgtaggctggagct
ttatgccgcgctactctcgtctaattgcgctatatactcacgcacttctgattccggggatccgtcgacc
gCggcagccgctgcagctgccgcggcagctaaggcggcagtacacacagcgg
aaggcggcagtacacacagcgg
agctgccgcggcagctgcagcggctgccgcgacccgttgcccccaccctctg
gacccgttgcccccaccctctg
gCggcagccgctgcagctgcCgcggcagctcctgtggggcatggtgttgcct
cctgtggggcatggtgttgcct
agctgccgcggcagctgcagcggctgccgcgtgttcagttcccagtgatttt
gtgttcagttcccagtgatttt

 

gcggCagCCgctgcagctgccgcggcagcttcattggcggcagaaggcaaaa
tcattggcggcagaaggcaaaa
(agctgccgcggcagctgcagcggctgccgcctgctgtaaggcaacaccatgc
ctgctgtaaggcaacaccatgc
tggcggcagaaggcaaaaatcacagcaaaggggcaacattgc
leacagcaaaggggcaacattgc
atttttgccttctgccgccaatgatgcagcggcattttgc
atgatgcagcggcattttgc
gCggcagccgctgcagcthCgcggcagcttctactgcggctgatggtatca
tctactgcggctgatggtatca
agctgccgcggcagctgcagcggctgccgcatttttgccttctgccgccaat
atttttgccttctgccgccaat
gcggcagccgctgcagctgccgcggcagctcaccagcaaaagaaatccttcag
caccagcaaaagaaatccttcag
agctgccgcggcagctgcagcggctgccgctggctggtgaattctcggcat
tggctggtgaattctcggcat
gcggcagccgctgcagctgccgCggcagctaaaaaattttccagatcggcaccgcagggc
aaaaaattttccagatcggcaccgcagggc

 

 

65

 

Table 2-2 (cont’d).

 

71-85RT agctgccgcggcagctgcagcggctgccgcgctgataccatcagccgcagtag
71-85RS gctgataccatcagccgcagtag
86-100FT gcggcagCCgctgcagctgccgcggcagctcacagcaaaggggcaacattgc
86-100FS cacagcaaaggggcaacattgc
86-1 OORT agctgccgcggcagctgcagcggcthCgccgtccccaaacagcccctgaga
86-100RS C gtccccaaacagcccctgaga

 

 

66

WORKS CITED

1. Asselin, J.E., KN. Yip, and S.V. Beer. 2008. Eopl differs in strains of Erwinia
amylovora that differ in host specificity. Acta Hort. 793: 213-214.

2. Ausubel, F.M., R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith,
and K. Struhl. eds. 1997. Current Protocols in Molecular Biology. Wiley, New York.

3. Birtalan, S.C., R.M. Phillips, and P. Ghosh. 2002. Three-dirnensional secretion
signals in Chaperone-effector complexes of bacterial pathogens. Mol. Cell 9: 971-980.

4. Bocsanczy, A.M., R.M. Nissinen, C.-S. Oh, and S.V. Beer. 2008. HrpN of E.
amylovora functions in the translocation of DspA/E into plant cells. Mol. Plant Pathol. 9:
425-434.

5. Bogdanove, A.J., J.F. Kim, Z. Wei, P. Kolchinksy, A.0. Charkowski, A.K. Conlin,
A. Collmer, and S.V. Beer. 1998. Homology and functional similarity of an hrp-linked
pathogenicity locus, dspEF of Erwinia amylovora and the avirulence locus aer of
Pseudomonas syringae pathovar tomato. Proc. Natl. Acad. Sci. USA 95: 1325-1330.

6. Boyd, A.P., I. Lambermont, and G.R. Cornelis. 2000. Competition between the
Yops of Yersinia enterocolitica for delivery into eukaryotic cells : role of the Sch
Chaperone binding domain of YopE. J. Bacteriol. 182: 4811-4821.

7. Boureau, T., H. El Maarouf-Boteau, A. Garnier, M.-N. Brisset, C. Perino, I.
Pucheu, and M.A. Barny. 2005. DspA/E, a type IH effector essential for Erwinia
amylovora pathogenicity and growth in planta, induces cell death in host apple and
nonhost tobacco plants. Mal. Plant-Microbe Interact. 19: 16-24.

8. Bronstein, P.A., E.A. Miao, and 8.]. Miller. 2000. Ian is a type HI secretion
chaperone specific for SspA. J. Bacteriol. 182: 6638-6644.

9. A. Burse, H. Weingart, and M.S. Ullrich. 2004. The phytoalexin-inducible multidrug
efflux pump AcrAB contributes to virulence in the fire blight pathogen, Erwinia
amylovora. Mal. Plant-Microbe Interact. 17: 43-54.

10. Buttner, D., D. Gurlebeck, L.D. Noel, and U. Bonas. 2004. HpaB from
Xanthomonas campestris pv. vesicatoria acts as an exit control protein in type III-
dependent protein secretion. Mol. Microbiol. 54: 755-768.

11. Buttner, D. and U. Bonas. 2006. Who comes first? How plant pathogenic bacteria
orchestrate type III secretion. Curr. Opin. Microbiol. 9: 193-200.

12. Casper-Lindley, C., D. Dahlbeck, E.T. Clark, and B.J. Staskawicz. 2002. Direct

biochemical evidence for type IH-dependent translocation of the AvrBsZ effector into
plant cells. Proc. Natl. Acad. Sci. USA 99: 8336-8341.

67

13. Cheng, L.W., D.M. Anderson, and O. Schneewind. 1997. Two independent type HI
secretion mechanisms for YopE in Yersinia enterocolitica. Mol Microbiol 24: 757—765.

14. Chin, J., P.E. March, R. Lee and D. Tillett. 2004. Site-directed, ligase-independent
mutagenesis (SLIM): a single-tube methodology approaching 100% efficiency in 4 h.
Nucleic Acids Res. 32: e174.

15. Datsenko, KA. and Wanner, B.L. 2000. One-step inactivation of Chromosomal
genes in Escherichia coli using PCR products. Proc. Natl. Acad. Sci. USA 97 : 6640-6645.

16. Debroy, S., R. Thilmony, Y.-B. Kwack, K. Nomura, and S.-Y. He. 2004. A family
of conserved bacterial effectors inhibits salicylic acid-mediated basal immunity and
promotes disease necrosis in plants. Proc. Natl. Acad. Sci. USA 101: 9927-9932.

17. Ehrbar, K., B. Winnen, and W. Hardt. 2005. The Chaperone binding domain of
SopE inhibits transport via flagellar and SPI-l TTSS in the absence of Ian. Mal.
Microbial. 59: 248-264.

18. Elliott, S.J., S.W. Hutcheson, M.S. Dubois, J.L. Mellies, L.A. Wainwright, M.
Batchelor, G. Frankel, S. Knutton, and J.B. Kaper. 1999. Identification of CesT, a
Chaperone for the type IH secretion of Tir in enteropathogenic Escherichia coli. Mal
Microbial 33: 1 176—1 189.

19. Feldman, M.A., S. Muller, E. Wuest, and G.R. Cornelis. SyCE allows secretion of
YopE-DHFR hybrids by the Yersinia enterocolitica type IH Ysc system. Mal. Microbial
46: 1183-1197.

20. Gaudriault, S., L. Malandrin, J.-P. Paulin, and M.-A. Barny. 1997. DspA, an
essential pathogenicity factor of Erwinia amylovora showing homology with Aer of
Pseudomonas syringae, is secreted via the Hrp secretion pathway in a DspB-dependent
way. Mol. Microbiol. 26: 1057-1069.

21 .Gaudriault, S., J.-P. Paulin, and M.-A. Barny. 2002. The DspB/F protein of
Erwinia amylovora is a type III secretion Chaperone ensuring efficient intrabacterial
production of the Hrp-secreted DspA/E pathogenicity factor. Mol. Plant Pathol. 3: 313-
320.

22. Ghosh, P. 2004. Process of protein transport by the type III secretion system.
Microbial. Mal. Biol. Rev. 68: 771-795.

23. Gouet, P., E. Courcelle, D.I. Stuart, and F. Metoz. 1999. ESPript: multiple
sequence alignments in PostScript. Bioinformatics 15: 305-308.

24. Gun, M., S.T. Chancey, F. Tian, Z. Ge, Y. Jamir, and J.R. Alfano. 2005.
Pseudomonas syringae type III Chaperones Sthl, Sth1, and Sth2 facilitate

68

translocation of their cognate effectors and can substitute for each other in the secretion
of HopOl-l. J. Bacteriol. 187: 4257-4269.

25. Ham, J.H., D.R. Majerczak, A.S. Arroyo-Rodriguez, D.M. Mackey, and D.L.
Coplin. 2006. WtsE, an Aer-family effector protein fiom Pantoea stewartii subsp.
stewartii, caused disease-associated cell death in corn and requires a Chaperone protein
for stability. Mal. Plant-Microbe Interact. 19: 1092-1102.

26. Higashide, W. and D. Zhou. 2006. The first 45 amino acids of SopA are necessary
for Ian binding and SPI-I secretion. J. Bacteriol. 188: 2411-2420.

27. Huynh, T.V., D. Dahlbeck, and B.J. Staskawicz. 1989. Bacterial blight of soybean:
regulation of a pathogen gene determining host cultivar specificity. Science 245: 1374-
1 3 77.

28. Jiang, J., L.W. Fan, and W.H. Wu. 2005. Evidences for the involvement of
endogenous CAMP in Arabidopsis defense responses to verticillium toxins. Cell Res. 15:
585-592.

29. Jones, D.T. 1999. Protein secondary structure prediction based on position-specific
scoring matrices. J. Mol. Biol. 292: 195-202.

30. Kovach, M.E., P.H. Elzer, D.S. Hill, G.T. Robertson, M.A. Farris, R.M. Roop,
and KM. Peterson. 1995. Four new derivatives of the broad-host-range Cloning vector
pBBRlMCS, carrying different antibiotic-resistance cassettes. Gene 166: 175-176.

31. Kurosaki, F., and A. Nishi. 1993. Stimulation of calcium influx and calcium cascade
by cyclic AMP in cultured carrot cells. Archiv. Biochem. Biophys. 302: 144-51.

32. Lee, SH. and J.E. Galan. 2004. Salmonella type IH secretion-associated Chaperones
confer secretion-pathway specificity. Mol. Microbiol. 51: 483-495.

33. Lefevre, F., M.H. Remy, and J.M. Masson. 1997. Alanine-stretch scanning

mutagenesis: a simple and efficient method to probe protein structure and function.
Nucleic Acids Res. 25: 447-448.

34. Letzelter, M., I. Sorg, L.J. Mota, S. Meyer, J. Stalder, M. Feldman, M. Kuhn, I.
Callebaut, and G.R. Cornelis. 2006. The discovery of Sch highlights a new firnction
for type HI secretion effector Chaperones. EMBO J. 25: 3223-3233.

35. Lilic, M., M. Vujanac, and CE. Stebbins. 2006. A common structural motif in the
binding of virulence effectors to bacterial secretion chaperones. Mol. Cell 21: 653-664.

36. Losada, LC. and SW. Hutcheson. 2005. Type III secretion Chaperones of

Pseudomonas syringae protect effectors from Lon-associated degradation. Mal.
Microbial. 55: 941-953.

69

37. Matsumoto, H., and GM. Young. 2009. Essential role of the SycP chaperone in
Type IH secretion of the YspP effector. J. Bacteriol. 191: 1703-1715.

38. Meng, X., J.M. Bonasera, J.F. Kim, R.M. Nissinen, and S.V. Beer. 2006. Apple
proteins that interact with DspA/E, a pathogenicity factor of Erwinia amylovora, the fire
blight pathogen. Mal. Plant-Microbe Interact. 19: 53-61.

39. Miao, E.A., and 8.]. Miller. 2000. A conserved amino acid sequence directing
intracellular type III secretion by Salmonella typhimurium. Proc. Natl. Acad. Sci. USA
97: 7539-7544.

40. Mudgett, M.B., 0. Chesnokova, D. Dahlbeck, E.T. Clark, 0. Rossier, U. Bonas,
and B.J. Staskawicz. 2000. Molecular signals required for type IH secretion and

translocation of the Xanthomonas campestris AvrBsZ protein to pepper plants. Proc.
Natl. Acad. Sci. USA 97: 13324-13329.

41. Nissinen, R. M., A.J. Ytterberg, A.J., Bogdanove, K van Wijk, and S.V. Beer.
2007. Analyses of the secretomes of Erwinia amylovora and selected hrp mutants reveal

novel type III secreted proteins and an effect of Her on extracellular harpin levels. Mol.
Plant Pathol. 8:55-67.

42. Notredame, C., D. Higgins, and J. Heringa. 2000. T-Coffee: a novel method for
multiple sequence alignments. J. Mol. Biol. 302: 205-207.

43. Parsot, C., C. Hamiaux, and A.-L. Page. 2003. The various and varying roles of
specific chaperones in type III secretion systems. Curr. Opin. Microbiol. 6: 7-14.

44. Schechter, L.M., KA. Roberts, Y. Jamir, J.R. Alfano, and A. Collmer. 2004.
Pseudomonas syringae type III secretion system targeting signals and novel effectors
studied with a CyaA translocation reporter. J. Bacteriol. 186: 543—555.

45. Schesser, K, E. Frithz-Lindsten, and H. Wolf-Watz. 1996. Delineation and
mutational analysis of the Yersinia pseudotuberculosis YopE domains which mediate
translocation across bacterial and eukaryotic cellular membranes. J. Bacteriol 178: 7227-
7233.

46. Sory, M.-P., and G.R. Cornelis. 1994. Translocation of a hybrid YopE-adenylate
cyclase from Yersinia enterocolitica into HeLa cells. Mol. Microbiol. 14: 583-594.

47. Sory, M.-P., A. Boland, I. Lambermont, and G.R. Cornelis. 1995. Identification of
the YopE and YopH domains required for secretion and internalization into the cytosol of

macrophages, using the cyaA gene fusion approach. Proc. Natl. Acad. Sci. USA 92:
1 1998-12002.

48. Stebbins, CE. and J.E. Galan. 2001. Maintenance of an unfolded peptide by a
cognate Chaperone in type III secretion. Nature 414: 77-81.

70

49. Thomas, N.A., W. Deng, J.L. Puente, E.A. Frey, C.K Yip, N.C.J. Strynadka, and
B.B. Finlay. 2005. CesT is a multi-effector Chaperone and recruitment factor required for
the efficient type 1H secretion of both LEE and non-LEE encoded effectors of
enteropathogenic Escherichia coli. Mol. Microbiol. 57: 1762-1779.

50. Trulzsch, K, A. Roggenkamp, M. Aepfelbacher, G. Wilharm, K Ruckdeschel,
and J. Heesemann. 2003. Analysis of Chaperone-dependent Yop secretion/translocation

and effector function using a mini-virulence plasmid of Yersinia enterocolitica. Intern. J.
Med. Microbiol. 293: 167-177.

51. Vinatzer, B., Jelenska, J., and Greenberg, J.T. 2005. Bioinformatics correctly
identifies many type III secretion substrates in the plant pathogen Pseudomonas syringae
and the biocontrol isolate P. fluorescens SBW25. Mol. Plant Pathol. 18: 877-888.

52. Wei, Z. M., B.J. Sneath, and S.V. Beer. 1992. Expression of Erwinia amylovora hrp
genes in response to environmental stimuli. J. Bacteriol. 174: 1875-1882.

53. Woestyn, S., M.-P. Sory, A. Boland, 0. Lequenne, and G.R. Cornelis. 1996. The
cytosolic Sch and SycH Chaperones of Yersinia protect the region of YopE and YopH
involved in translocation across eukaryotic cell membranes. Mol. Microbiol. 20:1261-71.

54. Zhao, Y., S.E. Blumer, and Sundin, G.W. 2005. Identification of Erwinia
amylovora genes induced during infection of immature pear tissue. J. Bacteriol. 187 :
8088-8103.

55. Zhao, Y., S.-Y. He, and G. Sundin. 2006. The Erwinia amylovora aerptZEa gene
contributes to virulence on pear and Aerpt2EA is recognized by Arabidopsis RPSZ
when expressed in Pseudomonas syringae. Mal. Plant-Microbe Interact. 19: 644-654.

56. Zhao, Y., G.W. Sundin, and D. Wang. 2009. Construction and analysis of
pathogenicity island deletion mutants of Erwinia amylovora. Can. J. Microbial. 55: 45 7-
464.

71

Chapter 3: Homology—based modeling of the Erwinia amylovora type III secretion
Chaperone DspF used to identify amino acids required for virulence and interaction
with the effector DspE.

ABSTRACT

Chaperones involved in bacterial type III secretion are critical to effector
translocation into host cells and to virulence of many pathogens. In the past decade,
crystallization studies have elucidated the highly conserved structure of numerous type
III secretion Chaperones and identified a general pattern of surface sites predicted to be
essential to interaction with the effector. However, few mutational analyses have been
performed to determine the relevance of these sites. Here, the structure of DspF, a T3SS
chaperone required for virulence of the fi'uit tree pathogen Erwinia amylovora, was
modeled based on predicted structural homology to characterized T3SS chaperones. This
model, along with observed sequence conservation, was used to hypothesize which
residues could be important for interaction with the secreted effector DspE. Eleven site-
directed alanine replacement mutants were constructed and assessed for effect on
virulence complementation, dimerization, and interaction with the N-terminal chaperone-
binding site of DspE. Four amino-acid residues were identified that did not complement
the virulence defect of a dspF knockout mutant, and three of these residues were required
for interaction with the N-terminus of DspE. In addition, these four site-directed mutants
of DspF caused a moderate reduction in virulence when introduced into wild-type
Ea1189, and two significantly lowered translocation levels of a DspE-CyaA fusion
protein. This study supports the significance of the predicted B-sheet helix-binding

groove in DspF chaperone function, and suggests that structural homology modeling may

be a useful tool in the functional characterization of many T3SS Chaperones.

72

INTRODUCTION

A wide variety of bacterial pathogens and symbionts require a type IH secretion
system (T3 SS) to inject host cells with effector proteins important for suppressing host
defenses and establishing infection. Many of these effector proteins interact with specific
type HI secretion Chaperones prior to secretion. T3SS effector Chaperones may perform
numerous functions; some Chaperones have been shown to bind directly to a specific
T3SS apparatus or to hold the N-terminus of the effector in an unfolded position,
suggesting a direct loading function (1, 12, 24, 33). T 3 SS chaperones may also cover an
inhibitory or destabilizing domain on the effector, organize a temporal secretion
hierarchy among effectors of varying importance, or regulate the transcription of cognate
effectors (6, 10, 25, 38).

Despite a lack of sequence conservation, T3 SS effector-specific Chaperones
(heretofore T3SS Chaperones) are highly structurally conserved; they are small, acidic
dimers with a Characteristic shape of three or-helices and five B-sheets (28). The
publication of at least 10 T3SS Chaperone crystal structures in the last decade has
revealed that most Chaperones possess two surface- exposed hydrophobic patches, one of
which lies in a twisted B-sheet groove often referred to as the “helix-binding groove” (3,
8, 11, 26, 32, 33, 37 ). Co-crystal structures of effector-bound chaperones have indicated
that the N-terminus of the effector protein wraps around the exterior of the Chaperone,
burying the helix-binding groove on each monomer as well as a second hydrophobic site
(4, 26, 30, 33). Mutagenesis studies of the Salmonella enterica sv. T yphimurium
chaperone gene SigE and the Escherichia coli gene cesT determined that selected external

hydrophobic residues play an important role in the Chaperone-effector interaction (20,

73

27). However, there are many additional residues in each chaperone structure predicted to
interface with secreted effectors, including hydrophobic and electrostatic residues
exposed in the B-sheet helix-binding groove. Very few mutational analyses of Chaperones
have been performed, and the relative contribution of structural features to different
chaperone functions is still poorly understood.

Erwinia amylovora is a gram-negative bacterium that causes a devastating blight
disease of apple and pear trees. Unlike the model bacterial phytopathogen Pseudomonas
syringae, members of the genus Erwinia are enterobacterial plant pathogens, with strong
taxonomic and genomic similarity to Salmonellae and Escherichiae (3 5). Pathogenesis of
E. amylovora is dependent on the secretion of DspE, a large T3SS-secreted effector of
unknown function (14). DspF, the specific Chaperone of DspE, is a Chaperone involved in
DspE stability and secretion into the medium (14, 15). DspF is a member of a large
family of T3 SS chaperones in plant pathogens, with putative or confirmed sequence
homologs in Pseudomonas syringae, Pantoea agglomerans, Pantoea stewartii, and
Dickeya dadantii (5, 13, 18, 31). Deletion of dspF from E. amylovora causes a strong
reduction in virulence (36), making E. amylovora a useful system for genetic studies of
Chaperone firnction.

Recently, we used site-directed mutation and protein-protein interaction studies to
identify N-terrninal features of DspE required for secretion and for binding DspF (36).
Here, a mutagenesis approach was used to characterize sequence and structural
requirements of DspF Chaperone function. The tertiary structure of DspF was modeled
based on sequence and secondary structural similarities between DspF and structurally

Characterized type III secretion Chaperones fi'om human pathogens. The model guided an

74

alanine-scanning mutagenesis study of the conserved residues of the predicted helix-
binding groove of DspF. We identified multiple amino acid residues required for E.
amylovora virulence and/or interaction with DspE, suggesting that electronegative groove
residues are important to the Chaperone-effector interaction.

MATERIALS and METHODS
Bacterial strains and plasmids and genetic techniques. The bacterial strains and
plasmids used in this study are listed in Table 1. PCR, sequencing, and cloning was
performed using standard techniques and protocols (2). All sequencing reactions were
conducted at the Michigan State University Research Tedhnology Support Facility.
Site-directed alanine replacement mutagenesis. Alanine-replacement mutants were
constructed using the Quikchange mutagenesis system following the manufacturer’s
recommendations (Agilent Technologies, Stratagene Division; La Jolla, CA). The entire
plasmid pLRT7, including the dspF gene and its native promoter, was amplified using
complementary mutagenic primers and the following PCR reaction conditions: 94°C for
30s, 57°C for 305, and 68°C for 7 m. Primers for mutagenesis of dspF are shown in Table
2. One microliter of DpnI was added to 50 pL reaction mixture and incubated at 37°C for
l h. PCR products were then transformed into E. coli DHSa. Plasmid inserts were
sequenced to ensure that a single-site nucleotide alteration was the only difference
between mutant and wild-type vector inserts. pLRT7 and derivatives were introduced into
E. amylvora Eal l89AdspF by electroporation. The presence of the dspF plasmids in each
complemented strain was confirmed by PCR.
Structural prediction methods. 3D-Jury, a consensus fold-recognition server combining

nine structural prediction programs on Metaserver (16;

75

http://meta.bioinfo.pl/submit_wizard.pl), was used to search for structural homologs of
DspF. Although 3D-Jury hits included most known T3SS Chaperone structures, the
structure of the chaperone SicP (1 jyo_A) from Salmonella enterica sv. T yphimurium was
ranked the most similar to DspF by the prediction servers. A Phyre analysis also
identified SicP as the best match to DspF, with an equivalent overall alignment (19;
http://www.sbg.bio.ic.ac.uk/phyre/). Using the sequence alignment ranked most highly by
3D-Jury, a backbone homology model of DspF was constructed from the SicP structure
using in-house software. Side chain atom positions were also taken from ljyo for the
homology model, but only if the aligned residues in DspF and SicP were identical. The
remaining side Chain conformations were then constructed using SCWRL (22;
http://dunbrack.fccc.edu/scwrl4/SCWRL4.php). The DspF structural model is available
upon request from the authors. The predicted molecular surface of the structure was
visualized using Deepview/swiss PDB viewer (17). To identify conserved residues of
DspF, an alignment of DspF, WtsF, and Aer protein sequences was constructed using T-
Coffee (28).

Yeast two-hybrid assays. Site-directed mutant sequences of dspF were amplified using
the primers DspFECF (5’- catgaattcatgacatcgtcacagcagcgg-3’) and DspFXer (5’-
gatCCTCGAGttatgccgcgctactctcgtc-B’), digested with EcoRI and XhoI, and ligated into
the bait vector pB42AD. To construct the prey vectors pLRT93 and pLRT219, dspF was
amplified with primers DspFBamlF (gatcggatcctcatgacatcgtcacagcagcgg) and DspFXhR,
and nucleotides l-l800 of dspE were amplified with primers DspEBamF (5’-
gcggatcctcatggaattaaaatcactggg-3’) and DspEXhoR800 (5’-

CgctcgagttagaaggagcgaatgtCCCC-3’). PCR products were digested with BamHI and X110]

76

and ligated in fusion with the LexA DNA binding domain in the prey vector pGilda
(Clontech; Mountain View, CA). Bait and prey constructs were co-transformed into the
Saccharomyces cerevisiae strain EGY48 p80(lacZ) using the Yeast Transformation Kit
(Zyrno; Orange, CA). Transformation reaction mixtures were plated on SD medium
supplemented with -ura-his-trp dropout solution (Clontech). Transforrnants were re-
plated on SD-gal medium (Clontech) supplemented with 5-bromo-4-chloro-3-indoly1-
beta-D-galactopyranoside (X-gal; 80 pg/mL), and color development was observed after
incubation at 30° C for three days. Expression of DspF-B42-HA was confirmed by
western blot of lysed yeast cells using anti-HA-Hrp antibody (Roche; Indianapolis, IN,
USA)

Virulence assays. Bacterial virulence was assayed using an immature pear fi'uit assay as
previously described (39). Briefly, surface-sterilized immature pears were wounded, and
bacterial strains (1.5 x 10"6 Cfu/mL) were pipetted on the wound. Pears were incubated
in a humid Chamber at 28° C for 9 d. After three days, pears developed a single, Circular
black lesion spreading out from the inoculation site. Lesion diameters were measured in
two perpendicular directions using digital calipers, and the average of these two
measurements was recorded as the average lesion diameter for each pear. If the lesions
spread to the sides or back of the pears, a piece of twine was used to measure the lateral
length of the lesion around the Circumference of the pear.

Translocation assays. Alanine-replacement mutants of dspF and the native dspF
sequence were amplified with primers FHF (5’ gcataagcttcgcatctgcgcttactgatttga 3’) and
FxbR (5’ gatctctagattatgccgc gctactctcgtcta 3 ’), digested with HindIII and XbaI, and

ligated into the broad host-range vector pBRRl -MC82 to create plasmids pLRT98 and

77

pLRT 158-161. pLRT201, expressing DspE(l -73 7)-CyaA, was introduced into wild-type
E. amylovora E31189. CyaA fusion expression was confirmed via western blot with anti-
CyaA antibodies (Santa Cruz Biotechnology; Santa Cruz, CA). pBRRl-MCSZ
derivatives were introduced into Ea1189/pLRT201 by electroporation. CyaA-expressing
overnight cultures were adjusted to 4.5 x 108 Cfu/mL (CD. = 0.3) and infiltrated into the
leaves of 8-wk old Nicotiana tabacum cv. Samsun plants. Leaf disks were collected after
11 h. CAMP was extracted and quantified as previously described (3 6).

RESULTS
Structural modeling of DspF. A combination of secondary structural alignment and
computational modeling was used to identify candidate residues of importance for DspF
Chaperone firnction. The tertiary structure of DspF was modeled based on predicted
structural homology to S. enterica subsp. T yphimurium SicP and other type IH secretion
Chaperones as described in Materials and Methods. The predicted tertiary structure of
DspF is shown in Figure 3-1A; structural alignment with SicP and more Closely-related
Chaperones from other plant pathogens is shown in Figure 3-1B. Although DspF and SicP
share only 15% sequence identity, residues 31-98 of DspF, which are predicted to form a
dimerization domain and helix-binding groove, share a much higher degree of similarity
with SicP. The amino acids predicted to form the B—sheets of the helix-binding groove
were roughly 30% identical and 60% similar between SicP and DspF (Figure 3-1B). In
addition, the secondary structure of SicP was the most Closely aligned to DspF using
multiple secondary alignment algorithms in the 3D-Jury suite. Thus, the sequence and
secondary structure of the central portion of DspF could be Closely aligned to that of

SicP. Based on these similarities, the DspF structure is predicted to include the

78

Characteristic features common to Class I type III secretion Chaperones, including the
conserved 3 a-helical structure and an external groove similar to that implicated in helix-
binding.

Identification of candidate DspF residues involved in virulence and effector
interaction. Sequence conservation and surface-exposed side Chains are potential
indicators that an amino acid could be involved in protein-protein interactions. To
identify the most highly conserved amino acids of DspF, psiBLAST search was used to
identify unique sequence homologs of DspF including DspF sequences from the related
enterobacterial plant pathogens Erwinia pyrifoliae, Pantoea agglomerans pv.
gypsaphilae, Pantoea stewartii, and Pectobacterium carotavara subsp. atrosepticum. The
sequence of the related type III secretion Chaperone Aer fi'om Pseudomonas syringae
was also included. When these protein sequences were aligned, 29 firlly conserved
residues were identified in DspF (Figure 3-1B). The DspF structural homology model
was used to screen conserved residues for putative substrate-exposed side chains. We
identified seven amino acids with predicted surface exposure and absolute conservation
among the Closest homologs of DspF: E39, E44, P46, H55, F78, D91, and L98 (Figure 3-
1A). These conserved residues had a predicted localization in a groove similar to the
helix-binding groove identified in crystallization studies of type HI Chaperones (3, 27, 32,
37). To allow for a complete analysis of the predicted helix-binding groove of DspF, we
identified four additional arrrino acids, L53, R57, E60, and H94, predicted to face the
helix-binding groove, although these were not completely conserved (Figure 3-1A and 3-

113).

79

Figure 3-1. Structural modeling and site-directed mutagenesis of DspF. A. Homology-
based model of DspF. Two monomers are depicted in green and grey. Residues selected
for alanine-replacement mutagenesis are labeled; residues completely conserved among
DspF homologs are labled in a red font. Alanine replacement of side Chains depicted in
red prevented complementation of virulence in E. amylovora, while replacement of
residues colored blue did not affect complementation. Structure was rendered using
MolMol (21). B. Alignment of the amino acid sequence of DspF with that of 5 Closely-
related Chaperones from plant pathogenic bacteria and with Salmonella typhimurium
SicP. Residues completely conserved among DspF family Chaperones are indicated in
black; conservatively replaced residues are indicated in grey. Red asterisks specify
highly-conserved DspF residues targeted for site-directed mutagenesis. Blue asterisks
indicate the SicP residues buried by interaction with the effector SptP, previously
calculated by Singer et al. (32) based on the work of Stebbins and Galan (33). Genbank
accessions for aligned sequences are: E. amylovora, AAC04851; E. pyrifoliae,
AAS45451; Pantoea stewartii, AAG01468, Pantoea agglomerans, AASZO350;
Pseudomonas syringae pv. phasealicola, YP273526; Pectobacterium carotavarum,
ZP_03833469; and S. typhimurium, AAC38655.

80

mm5<
NUVH
mun:
mnwn
wwmm
wuwn
mwnw

mam—:5 wL.

 

*fi...

  

‘.

 

81

To determine whether these residues aligned with helix-binding groove residues
in other Chaperones, the SicP-based structural model of DspF was compared to three
other Closely related structures: P. syringae AverhF Orfl (PDB accession 1528),
Yersinia pestis YscB (PDB accession lxkp), and the T110839 Chaperone protein from
Synechococcus elongatus (PDB accession 2plg). Despite little sequence conservation, a
majority of the 11 amino acids predicted to face the helix-binding groove of DspF aligned
with residues previously determined to be exposed in the corresponding groove of
AverhF orfl (11 aligned /11 groove residues), YscB (9/11) and T110838 (8/11) in 3D-
Jury secondary structure alignments (data not shown). In addition, the 11 candidate helix-
binding residues of DspF aligned closely in both sequence and secondary structural
alignments with previously identified effector-buried residues of SicP (Figure 3-1A).
Thus, the conserved residues in DspF predicted to face the helix-binding groove are
strong candidates for interaction with DspE.

Identification of four conserved residues of DspF required for virulence function. To
determine the functional significance of residues in the putative helix-binding groove, 11
site-directed alanine-replacement mutants of dspF were constructed from the parent
vector pLRT7, carrying the dspF orf and its immediate upstream promoter. Alanine
replacement is a common strategy for determining the role of single amino acid surface
residues without altering the structural integrity of the protein (9). The alanine-
replacement pLRT7 derivatives were introduced into Ea1189AdspF, yielding 11
complemented strains. In a virulence assay, immature pear fruits were inoculated with
each complemented strain as described in Materials and Methods. Lesion diameters were

measured at eight days post-inoculation. Four site-directed mutants, E39A, E44A, H55A,

82

and F78A, did not complement the virulence phenotype of Eal l89AdspF . Strains
carrying these plasmids caused lesions that were significantly smaller than the strain
complemented with wild-type dspF (Figure 3-2). These four amino acids were considered
to be required for virulence. The remaining seven of the plasmids expressing dspF site-
directed mutants fully complemented the virulence phenotype of the dspF deletion
mutant, with no significant difference in virulence between strains complemented with
the wild-type and mutant derivative vectors (Figure 3-2). While Eal 189AdspF
complemented with dspF(E39A) caused lesions not significantly larger than that
complemented with wild-type DspF, Ea1189AdspF complemented with dspF(E44A),
dspF(HSSA), or dspF(F78A) caused levels of virulence intermediate between

Ea1189AdspF and Eal l89AdspF/pLRT7 (Figure 3-2).

83

\l
O
h?
9
E

Lesion diameter (mm)

 

 

AdspF wt E39 E44 P46 L53 H55 R57 E60 F78 L91 H94 L98

 

1189 1189 1189 1189 1189 1189
AdspF AdspF AdspF AdspF AdspF AdspF
(pLRT7) (E39A) (E44A) (H55A) (F78A)

Figure 3-2. Virulence phenotype of site-directed mutants. A. Average lesion diameters
on pears inoculated with Ea1189AdspF carrying a plasmid with wild-type dspF, an empty
vector, or one of 11 site-directed dspF mutants. Error bars depict lesion diameters iSE.
Letters a, b, and C denote treatments equivalent to wild-type, significantly different from
both the wild-type and the vector control (p<0.05), and Significantly different from the
wild-type only (p<0.001), respectively. B. Immature pears after infection with selected
strains as described in A. Photographs and measurements were taken 8 days post-
inoculation.

84

Three DspF residues required for virulence are also required for interaction with
the N-terminus of DspE. Previously, we mapped a DspF-binding site to within amino
acids 1-150 of DspE (36). Here, a LexA-based yeast two-hybrid (Y2H) assay was used to
determine whether site-directed mutants of DspF affected in virulence were affected in
their interaction with the N-terminal 150 amino acids of DspE. As previously reported, a
fusion of wild type DspF with the B42 activation domain showed a strong interaction
with LexA-DspE(l-l 50) (36). Replacement of DspF residues E39, E44, or H55 with
alanine abolished this interaction (Figure 3-3A). Only one virulence-deficient mutant,
B42AD-DspF (F 78A), showed a strong interaction with the DspE fusion protein. As a
control, B42AD-DspF constructs with alanine replacements in residues E60, D91, or
H94, residues not required for virulence, interacted strongly with DspE (1-150) in the
Y2H assay (Figure 3-3A), suggesting a correlation between effector-binding ability and
complementation of virulence phenotype. In a western blot analysis of yeast lysates, B42-
HA-DspF and variants were all similarly expressed and migrated at the same size,
indicating that a lack of binding was not due to loss of expression (data not shown). All
site-directed mutants showed a positive interaction with a longer fi'agment of DspE,
LexA-DspE(l-600), although color development appeared weaker in the interaction
between LexA-DspE (1-600) and B42-HA-DspF (E39A), B42-HA-DspF (E44A), and
B42-HA-DspF (H55A) (Figure 3-3B). All B42-HA-DspF fusions interacted with a LexA-
DspF fusion in yeast (Figure 3-3C), indicating that the site-directed mutations did not

affect DspF homodimerization.

85

- wt E39 E44 H55 F78 E60 D91 H94

{Lil

   

 

.3. I --‘".“":~._ . .
I .Q ~ ‘ .
C t. - fl 4.; .25 f ‘3 C

Figure 3-3. Alanine replacement of predicted helix-binding groove residues of DspF
disrupts the interaction with the N-terminus of DspE in yeast. A. Yeast two-hybrid
interaction ofLexA-DspE(1-150) with pB42-HA-T (-), B42-HA-DspF (wt), or site-
directed alanine replacement mutants. Dark coloration indicates a positive interaction. B.
Interaction of LexA-DspE(1-600) with B42-HA-DspF and site-directed alanine
replacement derivatives. C. Interaction of LexA-DspF with B42-HA-DspF and site-
directed alanine replacement derivatives. All yeast strains were incubated on the same
plate and photographed after 72 h at 30° C.

86

Site-directed mutants reduce virulence of wild-type Ea1189 'and moderately reduce
translocation levels of the N-terminus of DspE. Having observed that site-directed
mutants of DspF interact with wild-type DspF in yeast, we hypothesized that ectopic
expression of DspF site-directed mutants would have a dominant-negative effect on wild-
type DspF function; that is, the alanine-replacement mutants would negatively affect
secretion of DspE and virulence of E. amylovora. Negative dominance would indicate
that site-directed DspF mutants were interacting with wild-type DspF and abrogating its
function. On the other hand, a lack of negative dominance would suggest that the DspF
dimer may be able to fimction normally with only one copy of the amino acid in question.
To maximize levels of dspF expression and any potential effect of negative dominance,
we cloned dspF and dspF site-directed mutants into a pGem high-copy cloning vector to
create constructs pLRT215-219. pLRT215, expressing wild-type dspF, complemented
Ea1189AdspF more strongly than pLRT7 did (data not shown). In an immature pear
assay, all strains of Eal 189 expressing site-directed dspF mutants showed decreased
virulence after five days compared with wild-type Ea1189 or Ea1189 carrying wild-type
dspF (Figure 3-4A, 3—4B). To determine whether this could be the result of reduced
translocation levels of DspE, dspF derivatives DspF (E44A), DspF (H55A) and DspF
(F78A) were subcloned into the broad host-range vector pBBRl-MCSZ and introduced
into Ea1189 expressing DspE (l-737)-CyaA. Translocation levels were assessed as
described in Materials and Methods. Expression of each site-directed mutant caused a
decrease in translocation levels, although this reduction was not significant for co-
expression with DspF (H55A) (Figure 3-4C). To determine whether the reduction in

DspE fusion translocation was due to a reduction in expression levels, strains oo-

87

expressing DspE(l-737)-CyaA and DspF variants were grown in hrp-inducing minimal
medium, and levels of DspE(1-737)-CyaA were analyzed by western blot. DspE-CyaA
fusion proteins accumulated to similar approximate levels across treatments (data not
shown). Together, these results indicate that overexpression of site-directed mutants of
DspF has a moderate dominant-negative effect on virulence, and negatively impacts

DspE (1-737)-CyaA translocation to varying degrees.

88

Figure 3-4. Site-directed mutants cause moderate reductions in virulence and DspE(1-
737)-CyaA translocation when expressed in wild-type E. amylovora. A. Lesions on
immature pears five days after inoculation with wild-type E31189 (Wt) and Ea1189
carrying site-directed dspF mutants. B. Average lesion diameters on pears inoculated
with wild-type Ea1189 (wt), Ea1189 carrying wild-type dspF on a plasmid (Wt/dspF)
and Ea1189 carrying site-directed dspF mutants. Bars represent the average of ten
replicate pears +/- SE. C. Effect of dspF and dspF variant expression on translocation
levels of DspE(1-73 7)-CyaA. dspF and dspF variants were co-expressed with DspE (1-
737)-CyaA in wild type E. amylovora. Bars represent the average of three replicate
leaves +/ SE. Asterisks denote means which are significantly different from the wild-
type (p<0.05). D. In-vitro expression levels of DspE(1-737)-CyaA when co-expressed
with dspF, dspF (E44A), dspF (H55A), and dspF (F78A), assayed by western blot of
strains induced in hrp minimal medium.

89

 

wt/ wt/ wt/ wt/ wt/
dspF dspF dspF dspF dspF
(E39A) (E44A) (H55A) (F78A)

B

30

 

 

average lesion diameter (mm)

wt wt/ wt/ wt/ wt/ wt/
dspF dspF dspF dspF dspF
(E39A) (E44A) (HSSA) (F78A)

{h
i: 160
-:: 140
“5 20
2 1
N 100
a 80
z 60
:5 40
g 20
0

 

wt/ wt/ wt/ wt/
dspF dspF dspF dspF
(E44A) (H55A) (F7 8A)
Figure 3-4.

90

DISCUSSION

Type III secretion chaperones possess a remarkably conserved structure despite a lack
of sequence homology. This conserved shape is suggestive of a divergent co-evolution of
chaperones with cognate effector secretion signals, preserving T3 SS targeting features
while conferring specificity to the chaperone-effector interaction. In this study, we took
advantage of structural similarities between T3 SS chaperones to predict the residues
likely to be exposed in the effector-binding groove of DspF, a member of a conserved
family of chaperones important in plant pathogenesis. Of seven highly-conserved amino
acid residues with predicted localization in this groove, mutagenesis of four residues led
to reduced virulence; none of the four non-conserved residues tested influenced
virulence. All but one of the 11 site-directed mutants tested was able to complement the
Ea1189AdspF mutant to some degree, indicating that even the mutants impaired in
virulence retained some function. These observations, combined with the high degree of
secondary structural similarity among chaperones and the frequency of conserved
sequence features between SicP and DspF, support the hypothesis that the residues
investigated here face the chaperone-effector interaction site.

Conserved surface hydrophobic patches have been the focus of most discussions
regarding the mechanism of T3SS chaperone-effector interactions; the presence and
arrangement of these sites have been proposed to contribute to the stability, and perhaps
even specificity, of the interaction (3, 26, 27). One of the conserved hydrophobic sites is
typically located in the helix-binding groove (3, 8). However, the structure of the
AverhF chaperone from Pseudomonas syringae pv. phasealicola demonstrated a

surprising lack of hydrophobic residues exposed in the corresponding groove, contrasting

91

sharply with the structure of Sch and suggesting a divergent mechanism of chaperone-
effector binding (32). The structural model presented here suggests that DspF also lacks
an extended hydrophobic surface in the predicted effector binding groove; only four of
the 11 side chains predicted to project into the groove are nonpolar. It will be interesting
to learn whether a missing or diminished hydrophobic patch at this site is a common
feature of chaperones in plant pathogens.

Alanine replacement of three polar amino acids diminished the virulence
phenotype of DspF. Electrostatic interactions mediated by external polar residues were
proposed to be important contributors to chaperone-effector interactions (4), and we
propose that polar DspF residues required for virulence could contribute to electrostatic
interactions with the cognate effector DspE. DspF residue H55 aligns with an asparagine
at position 49 in SicP, which is predicted to be involved in hydrogen bonding with the
effector (33). The other two polar amino acids found to impact virulence are glutamate
residues; glutamates have also been identified as predicted contributors to chaperone-
effector hydrogen bonding (4, 33). The mechanism by which these three polar residues
influence DspE binding and translocation remains unsolved, but the intermediate
virulence phenotype of the DspF E44A and H55A mutants suggests that these variants
may still interact with DspE at a reduced level. While the possibility of mutation-induced
localized structural destabilization was not ruled out experimentally, a previous
mutagenesis study of external T3 SS chaperone residues identified no alanine replacement
which caused chaperone misfolding (20).

The only nonpolar DspF residue found to be required for chaperone function in

this study was F78, which is predicted to extend into the groove from the dimerization

92

helix of the opposite DspF monomer. The alanine-replacement strategy used here does
not disrupt a hydrophobic surface and may be limited in determining the significance of
many hydrophobic residues; however, alanine replacement is still useful to determine the
significance of bulky hydrophobic residues or those with a specialized function. Many
crystallized chaperones feature a large hydrophobic or ringed side chain in this relative
position; while this feature is absent from SicP, secondary structure alignments suggest
that the role of F78A of DspF is likely analogous to that of F69 in Sch, L73 of YscB,
F 78 of AverhF Orfl, W100 in chaperone T110839, and L74 in CesT (PDB accessions
1n5b, lxkp, 1s28, 2plg, and 1k3e, respectively). In addition, phenylalanine stacking has
been identified as a probable mechanism of chaperone-effector binding (33).
Replacement of DspF F78 with alanine caused a significant decrease in virulence and
translocation of an N-terminal filSiOl‘l of DspE, although dimerization and binding to the
N-terminus of DspE in yeast were not apparently diminished. Its possible that the
interaction of this mutant with the effector is stable enough to yield a positive result in the
yeast two-hybrid assay, but destabilized enough to prevent a virulence phenotype.
However, if mutation of F78 truly uncouples the virulence of DspF from interaction with
DspE, F78 might be involved some other aspect of type HI secretion. To date, at least
three other site-directed chaperone mutants have been identified which appear to
uncouple chaperone binding from secretion (20, 34), but these were all mutations in
surface-exposed electronegative residues. While F 78A and similarly-positioned
hydrophobic residues in other chaperones are predicted to face the edge of the helix-
binding groove, the possibility that conserved external hydrophobic residues could play a

role in translocation efficiency cannot be ruled out. Although the mechanism is still

93

unknown, F 78 represents a common conserved feature of DspF family type III
chaperones with a role in effector translocation.

Previously, we identified a strong chaperone binding site within residues 51-100
of DspE, although several downstream sites on DspE also interacted with DspF in yeast
(3 6). Here, we show that alanine replacement of certain DspF residues disrupted the yeast
two-hybrid interaction between DspF and DspE 1-150, but not between DspF and longer
fragments of DspE. This could indicate that DspF could bind to alternate binding sites by
slightly different mechanisms, or that the yeast two-hybrid interaction between DspF and
downstream segments of DspE is due to weak, nonspecific binding. The requirement of
virulence-associated DspF residues for binding the N-terminal 150 amino acids of DspE
provides further support for the role of the N-terminus of DspE in chaperone binding and
virulence.

This study included analysis of the negative-dominance phenotype of site-directed
dspF mutants. Although our previous work found no difference in translocation levels of
DspE(1-737)-CyaA when compared between Ea1189AdspE and Ea1189AdspF (36),
more recent unpublished work indicates that DspE(1-73 7)-CyaA translocation levels
from Ea1189AdspF and AdspE are lower than translocation levels from the wild-type
Ea1189. Therefore, although DspF is not required for translocation of the N-terminus of
DspE, DspF may be necessary for full levels of N—terminal DspE translocation. Here, we
showed that expression of some dspF site-directed mutants strongly reduced translocation
of DspE(1-73 7)-CyaA. This finding supports a function of DspF in translocation of the
N-terminus of DspE, and shows that co-expression of wild-type DspF with DspF mutants

abrogates this function. Introduction of dspF (H55A) into 1189 did not significantly

94

reduce N-terminal DspE translocation levels; it is unknown whether this is because DspF
has a weaker requirement for the H55 residue or whether the H55A variant is otherwise
limited in function. Site-directed DspF mutants caused a very small reduction in
virulence only when expressed from a hi gh-copy vector; this may be because moderate
reductions in DspE translocation do not typically lead to a reduction in virulence (36). To
our knowledge, this is the first analysis of negative dominance in T3 SS chaperones.

As noted in the introduction, the helix-binding groove is not thought to be the sole
effector-interaction site of type HI secretion chaperones. In addition, several studies have
focused on a second hydrophobic binding crevice partially composed of three broadly
conserved hydrophobic residues in the BI sheet, facing away from the helix-binding
groove (26, 32, 33). In SicP, this pocket was predicted to be composed of residues C27,
L29, and L31 (26, 33), which align in DspF with C30, L32, and N34. The second
hydrophobic patch could play an important role in DspF interaction with DspE, in
addition to the relevance of putative helix-binding groove residues investigated here.
While these hydrophobic surface patches have been identified as important contributors
to chaperone-effector interactions, hydrogen bonds were also proposed to mediate the
specificity of the interaction (4), and external electrostatic residues have a demonstrated
role in chaperone function (20). Our results also lend further support for the significance
of individual hydrophilic residues in the helix-binding groove.

This study presents a homology-based prediction of a conserved DspF helix-
binding groove and evidence for the involvement of several residues in this putative
groove structure in virulence and chaperone binding. The high degree of conservation of

residues required for virulence among chaperones of the DspF family suggests that,

95

despite significant divergence, DspF-like chaperones bind to cognate effectors using a
conserved mechanism. Our results support the importance of exposed polar and
hydrophobic residues in T3 SS chaperone function, and identify a novel conserved
hydrophobic residue required for chaperone function but not for dimerization or effector
binding. Through comparative modeling studies and firrther functional analysis, we
submit that known chaperone structures may serve as a rich resource in revealing the

mechanisms of T3SS targeting.

96

Table 3-1. Bacterial strains and plasmids used in Chapter 3.

 

 

 

 

 

 

 

 

 

Strain or plasmid Characteristics Source
Escherichia coli strain

F-80dlacZ AMIS A(lacZYA-argF)U169 endAl Invitrogen,
DHSa recAl hst17 (rk-mk+)deoR thi-l supE44 gyrA96 Carlsbad, CA,

relAl )\- USA

Erwinia amylovora strains
Eal 189 Wild type 7
Ea1189AdspF dspF deletion mutant, KmR 36
Plasmids

pAlter-l Cloning vector for mutagenesis, TetR Promega
pLRT7 pAlter-l derivative expressing dspF This study
pLRT8 DspE(1-15)-CyaA 36
pLRT18 pLRT7 derivative expressing dspF (E39A) This study
pLRT19 pLRT7 derivative expressing dspF(E44A) This study
pLRT20 pLRT7 derivative expressing dspF (P46A) This study
pLRT21 pLRT7 derivative expressing dspF (L53A) This study
pLRT22 pLRT7 derivative expressing dspF (H55A) This study
pLRT23 pLRT7 derivative expressing dspF (R57A) This study
pLRT24 pLRT7 derivative expressing dspF (E6OA) This study
pLRT25 pLRT7 derivative expressing dspF (F78A) This study
pLRT26 pLRT7 derivative expressing dspF (D91A) This study
pLRT27 pLRT7 derivative expressing dspF (H94A) This study
pLRT28 pLRT7 derivative expressing dspF (L98A) This study
pGilda HIS3 LexA BD bait vector, AmpR Clontech
pLRT93 LexA-DspF This study
pLRT219 LexA-DspE(1-600) This study
pB42AD TRPl B42 AD prey vector, AmpR Clontech
pLRT215 B42-HA-DspF 36
pLRT61 B42-HA-DspF E39A This study
pLRT57 B42-HA-DspF E44A This study
pLRT58 B42-HA-DspF H55A This study
pLRT59 B42-HA-DspF F78A This study
pLRT60 B42-HA-DspF E6OA This study
pLRT62 B42-HA-DspF D91A This study
pLRT63 B42-HA-DspF H94A This study
pBBRl-MCSZ broad host-range cloning vector, KmR 21
pLRT98 pBBRl-MCSZ derivative expressing dspF This study
pLRT158 pBBRl-MCS2 derivative expressing dspF (E39A) This study
pLRT159 pBBRl-MCSZ derivative expressing dspF (E44A) This study
pLRTl6O pBBRl-MCSZ derivative expressing dspF (H55A) This study
pLRTl61 BBRl-MCS2 derivative expressing dspF (F78A) This study

 

 

 

97

 

Table 3-1 (cont’d).

 

 

 

pGem T-Easy TA cloning vector, ApR Promega
pLRT201 DspE(1-73 7)-CyaA 36
pLRT219 pGem T-Easy expressing dspF This study
pLRT220 pGem T-Easy expressing dspF (E39A) This study
pLRT221 pGem T-Easy expressing dspF (E44A) This study
pLRT222 pGem T-Easy expressing dspF (H55A) This study
pLRT223 Gem T-Easy expressing dspF (F 78A) This study

 

98

Table 3-2. Oligonucleotide primers used for

mutagenesis of dspF

 

primer

sequence (5’-> 3’)

 

E39F
E39R
E44F
E44R
P46F
P46R
L53F
L53R
H55F
HSSR
R5 7F
RS7R
E6OF
E6OR
F78F
F 78R
D91F
D91R
H94F
H94R
L98F
L98R

taacgaacaagatgagggggcggcggtgctgga
ccagcaccgccgcgggctcatcttgttcgtta
ggcggcggtgctggggtaccgcaacacag
ctgtgttgcggtactggcagcaccgccgcc
cggtgctggaagtagggcaacacagcgaca
tgtcgctgtgttggggtacttccagcacc g
cacagcgacagcctggcirctacactgccg
cggcagtgtag’gggcaggctgtcgctgtg
cagcgacagcctgttactagggtgccgaatcattgagg
cctcaatgattcggcaggctagtaacaggctgtcgctg
ctgttactacactgggcaatcattgaggctg

cagcctcaatgattggggagtgtagtaacag
cactgcgaatcattggggctgacccacaaac

gtttgtgggtcagcgggaatgattcggcagtg
ctauacagctgaatgggaaatggcggcc
ggccgccatttcaggattcagctgtaatag
gttggctggcgctgggtgaactgcacaacg
cgttgtgcagttcaggcagcgccagccaac
gctggatgaactggggaacgtgcgtttatg
cataaacgcacgttgggcagttcatccagc
ctgcacaacgtgcgtggtgttttcagcagtc
gactgctgaaaacatggcgcacgttgtgcag

a. Underlined residues denote the altered codon.

99

10.

11.

WORKS CITED

. Akeda, Y. and J.E. Galan. 2005. Chaperone release and unfolding of substrates in

type III secretion. Nature 437 : 911-915.

Ausubel, F.M., R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith,
and K. Struhl, eds. 1997. Current Protocols in Molecular Biology. Wiley, New
York.

Birtalan, S., and P. Ghosh. 2001. Structure of the Yersinia type III secretory system
chaperone Sch. Nat. Struct. Biol. 8: 974-978.

Birtalan, S.C., R.M. Phillips, and P. Ghosh. 2002. Three-dimensional secretion
signals in chaperone-effector complexes of bacterial pathogens. Mol. Cell 9: 971-980.

. Bogdanove, A.J., J.F. Kim, Z. Wei, P. Kolchinsky, A.O. Cahrkowski, A.K.

Conlin, A. Collmer, and S.V. Beer. 1998. Homology and functional similarity of an
hrp-linked pathogenicity locus, dspEF, of Erwinia amylovora and the avirulence
locus aer of Pseudomonas syringae pathovar tomato. Proc. Natl. Acad. Sci. USA 95:
1 325-1330.

Boyd, A.P, I. Lambermont, and G.R. Cornelis. 2000. Competition between Yops
of Yersinia enterocolitica for delivery into eukaryotic cells: role of the Sch
chaperone binding domain of YopE. J. Bacteriol. 182: 4811-4821.

Burse, A., H. Weingart, and M.S. Ullrich. 2004. NorM, an Erwinia amylovora
multidrug efflux pump involved in in vitro competition with other epiphytic bacteria.
Appl. Environ. Microbiol. 70: 693-703.

Buttner, C.R., G.R. Cornelis, B.W. Heinz, and H.H. Niemann. 2005. Crystal
structure of Yersinia enterocolitica type III secretion chaperone SycT. J. Mol. Biol.
375: 997-1012.

Cunningham, B.C., and J.A. Wells. 1989. High-resolution epitope mapping of
hGH-receptor interactions by alanine-scanning mutagenesis. Science 244: 1081-1085.

Darwin, KH., and V.L. Miller. 2001. Type III secretion chaperone-dependent
regulation: activation of virulence genes by SicA and InvF in Salmonella
typhimurium. EMBO J. 20: 1850-1862.

Edokimov, A.G., J. Tropea, K. Routzahn, and D.S. Waugh. 2002. Three-

dimensional structure of the type IH secretion chaperone Sch from Yersinia pestis.
Acta Cryst. 58: 398-406.

100

12. Ehrbar, K., B. Winnen, and W. Hardt. 2006. The chaperone binding domain of
SopE inhibits transport via flagellar and SPI-l TTSS in the absence of Ian. Mal.
Microbial. 59: 248-264.

13. Frederick, R.D., M. Ahmad, D.R. Majerczak, A.S. Arroyo-Rodriguez, S.
Manulis, and D.L. Coplin. 2001. Genetic organization of the Pantoea stewartii
subsp. stewartii hrp gene cluster and sequence analysis of the hrpA, hrpC, hrpN, and
wtsE operons. Mol Plant Microbe Interact. 14: 1213-1222.

14. Gaudriault, S., Malandrin, L., Paulin, J.-P., and Barny, M.-A. 1997. DspA, an
essential pathogenicity factor of Erwinia amylovora showing homology with Aer of
Pseudomonas syringae, is secreted via the Hrp secretion pathway in a DspB-
dependent way. Mol. Microbiol. 26: 1057-1069.

15. Gaudriault, S., Paulin, J.-P., and Barny, M.A. 2002. The DspB/F protein of
Erwinia amylovora is a type III secretion chaperone ensuring efficient intrabacterial
production of the Hrp-secreted DspA/E pathogenicity factor. Mol. Plant Pathol. 3:
313-320.

16. Ginalski K, A. Elofsson, D. Fischer, and L. Rychlewski. 2003. 3D-Jury: a simple
approach to improve protein structure predictions. Bioinformatics. 19: 1015- 1018.

17 . Guex, N. and M.C. Peitsch. 1997. Swiss-Model and the Swiss-de viewer: an
environment for comparative protein modeling. Electrophoresis 18: 2714-2723.

18. Ham, J.H., D.R. Majerczak, A.S. Arroyo-Rodriguez, D.M. Mackey, and D.L.
Coplin. 2006. WtsE, an Aer-family effector protein fiom Pantoea stewartii subsp.
stewartii, causes disease-associated cell death in corn and requires a chaperone
protein for stability. Mal. Plant Microbe Interact. 19: 1092-1 102.

19. Kelley, L.A., and MJ.E. Sternberg. 2009. Protein structure prediction on the web: a
case study using the Phyre server. Nature Protocols 4: 363—371.

20. Knodler, L.A., M. Bertero, N.C.J. Strynadka, and 0. Steele-Mortimer. 2006.
Structure-based mutagenesis of SigE verifies the importance of hydrophobic and
electrostatic residues in type HI chaperone function. Mol. Microbiol. 62: 928-940.

21. Koradi, R., M. Billeter, and K. Wuthrich. 1996. MOLMOL: A program for display
and analysis of macromolecular structures. J. Mol. Graphics 14: 51-55.

22. Kovach, M.E., P.H. Elzer, D. Steven Hill, G.T. Robertson, M.A. Farris, R.M.
Roop II, and KM. Peterson. 1995. Four new derivatives of the broad-host-range
cloning vector pBBRlMCS, carrying different antibiotic-resistance cassettes. Gene
166: 175-176.

101

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

G. G. Krivov, M. V. Shapovalov, and R. L. Dunbrack, Jr. 2009. Improved
prediction of protein side-chain conformations with SCWRL4. Proteins, early view
DOI 10.1002/prot.22488.

Lee., S.H., and J.E. Galan. 2004. Salmonella type III secretion chaperones confer
secretion-pathway specificity. Mal. Microbial. 51:483-495.

Letzelter, M., I. Sorg, L.J. Mota, S. Meyer, J. Stalder, M. Feldman, M. Kuhn, I.
Callebaut. and G.R. Cornelis. 2006. The discovery of Sch highlights a new
function for type IH secretion effector chaperones. EMBO J. 25: 3223-3233.

Lilic, M., M. Vujanac, and CE. Stebbins. 2006. A common structural motif in the
binding of virulence factors to bacterial secretion chaperones. Mal. Cell 21: 653-664.

Luo, Y., M.G. Bertero, E.A. Frey, R.A. Pfuetzner, M.R. Wenk, L. Creagh, C.
Kay, C. Haynes, B.B. Finlay, and N.C.J. Strydnaka. 2001. Structural and
biochemical characterization of the type III secretion chaperones CesT and SigE. Nat.
Struct. Biol. 8: 1031-1036.

Notredame, C., D.G. Higgins, and J. Heringa. 2000. T-coffee: a novel method for
fast and accurate multiple sequence alignment. J. Mol. Biol. 302: 205-217.

Parsot, C., C. Hamiaux, and A.-L. Page. 2003. The various and varying roles of
specific chaperones in type III secretion systems. Curr. Opin. Microbiol. 6: 7-14.

Phan, J., J.E. Tropea, and D.S. Waugh. 2004. Crystal structure of the Yersinia
pestis type III secretion chaperone SycH in complex with a stable fragment of
YscM2. Acta Cryst. 60: 1591-1599.

Rojas, C.M., J. H. Ham, L.M. Schechter, J.F. Kim, S.V. Beer, and A. Colhner.
2004. The Erwinia chrysanthemi EC16 hrp/hrc gene cluster encodes an active Hrp

type III secretion system that is flanked by virulence genes fimctionally unrelated to
the Hrp system. Mal. Plant Microbe Interact. 17: 644-653.

Singer, A.U., D. Desveaux, L. Betts, J.H. Chang, Z. Nimchuk, S.R. Grant, J.L.
Dangl, and J. Sondek. 2004. Crystal structures of the type III effector protein

AverhF and its chaperone reveal residues required for plant pathogenesis. Structure
12: 1669-1681. ‘

Stebbins, C.E., and J.E. Galan. 2001. Maintenance of an unfolded polypeptide by a
cognate chaperone in bacterial type III secretion. Nature 414: 77-81.

Thomas, N.A., W. Deng, J.L. Puente, E.A. Frey, C.K. Yip, N.C.J. Strynadka, and
B.B. Finlay. 2005. CesT is a multi-effector chaperone and recruitment factor required
for the efficient type III secretion of both LEE and non-LEE encoded effectors of
enteropathogenic Escherichia coli. Mal. Microbial. 57: 1762-1779.

102

35.

36.

37.

38.

39.

Toth, I.K., L. Pritchard, and P.R.J. Birch. 2006. Comparative genomics reveals
what makes an enterobacterial plant pathogen. Annu. Rev. Phytopathol. 44: 305-336.

Triplett, L.R., M. Melotto, and G.W. Sundin. 2009. Analysis of the N-terminus of
the Erwinia amylovora secreted effector DspA/E reveals features required for

secretion, translocation, and interaction with the chaperone DspB/F. Mol. Plant
Microbe Interact. 22: 1282-1292.

van Eerde, A., C. Hamiaux, J. Perez, C. Parsot, and B.W. Dijkstra. 2004.
Structure of Spa15, a type III secretion chaperone fiom Shigellaflexneri with broad
specificity. EMBO Rep. 5: 477-483.

Wulff-Strobel, C.R., A.W. Williams, and S.C. Straley. 2002. Lch and SycH
function together at the Ysc type IH secretion system in Yersinia pestis to impose a
hierarchy of secretion. Mol. Microbiol. 43: 411-423.

Zhao, Y., S.E. Blumer, and G.W. Sundin. 2005. Identification of Erwinia

amylovora genes upregulated during infection of immature pear tissue. J. Bacterial.
187: 8088-8103.

103

Chapter 4: A Study of the Roles of Class I Secretion Chaperones in Erwinia
amylovora in Effector Translocation and Virulence.

ABSTRACT

Chaperones specific to type IH secreted effector proteins often interact with
multiple substrates in the bacterial cell, including regulatory proteins, effectors, and
components of the secretion apparatus. A directed yeast two-hybrid screen identified
Erwinia amylovora secreted proteins Eopl and Eop3 and the type III secretion apparatus
proteins Hch and Hch as putative interactors of the type III secretion system (T3SS)
chaperone DspF. Annotation of the unfinished E. amylovora genome confirmed or
revealed the presence of putative chaperone genes next to cap] and eop3, but not next to
the effector-encoding gene eop4, and it was hypothesized that DspF might influence
translocation levels of other effectors. Deletion of dspF affected translocation levels of
DspE, Eopl , and Eop3, but not Eop4, when examined using CyaA translational firsions.
Deletion of the putative Eopl chaperone, escI, negatively affected translocation of Eopl -
CyaA but not of DspE-CyaA. Deletion of Eopl, Eop3, and the corresponding putative
chaperone genes had no significant effect on virulence, although Ea1189AdspFAescI
showed very slightly increased levels of early growth on pear compared with

Eal 189AdspF.

104

INTRODUCTION

A central theme in the pathogenesis of gram-negative bacteria is the secretion of
effector proteins directly into host cells via the type III secretion system (T3SS). Three
classes of specific chaperone proteins bind to secreted proteins and components of the
T388 (10); Class I T3 SS chaperones (henceforth T3 SS chaperones) are chaperones that
specifically interact with effector proteins. Despite having little sequence similarity,
T3SS chaperones share a highly conserved tertiary structure and similar general mode of
substrate binding (28).

Nearly two decades of study of T3 SS chaperones in Escherichia, Salmonella, and
Yersinia have led to a general understanding of the functional role of chaperones in
secretion. Chaperones bind to the N-terminal portion of a cognate effector protein. This
interaction holds the N-terminus of the effector in a non-globular state, which may confer
secretion competence (16). The direct binding may also stabilize the effector by masking
an aggregative or secretion-inhibitory domain (15, 21). T3SS chaperones may interact
with multiple proteins other than effectors; for example, direct interaction between a
T3SS chaperone and transcriptional or translational regulators mediates the regulatory
roles of some T3SS chaperones (11, 12). In addition, T3SS chaperones have been
demonstrated to interact directly with the ATPase and other conserved components of the
secretion apparatus (1 , 9, 30). This interaction could function in docking the effector to
the secretion apparatus.

Despite the requirement of a chaperone for secretion of many T3 SS effectors,
some effectors do not bind or require a corresponding chaperone for secretion into host

cells, and secretion of some proteins without a chaperone was observed to be delayed

105

compared with secretion substrates utilizing a chaperone (7, 22). These results led to the
hypothesis that chaperones may establish patterns of intracellular effector dynamics,
establishing a temporal order of secretion (27 , 34). In Yersinia enterocolitica, chaperone-
independent secretion of YopE was detected only in the absence of all other effectors,
demonstrating a competetive secretion advantage conferred by the YopE chaperone (7,
22). Likewise, the Salmonella enterica effector SipA is translocated earlier than the
effector SptP despite simultaneous expression; the authors proposed that this hierarchy
might result from varying binding affinities between chaperones and the T3 SS (34). The
chaperone SycH also establishes a secretion hierarchy by forming a secretion-inhibitory
complex with the protein Lch (33).

Although little work has been done to examine the role of T3 SS chaperones in
plant pathogenic bacteria, recent studies have demonstrated a similar overall function in
these systems, with some novel variations. Like the chaperones previously characterized
in animal pathogens, chaperones in plant pathogens have a highly conserved tertiary
structure (29), and may function in effector stabilization or direct interaction with the
secretion apparatus (24, 25). Although the dynamics of effector secretion has not been
studied in depth in plant pathogens, there is recent evidence that a T3SS chaperone is
involved in establishing a secretion hierarchy in X. campestris. By sequestering the global
T3SS chaperone HpaB, the regulatory protein HpaA promotes the secretion of non-
effectors prior to effectors (23).

The plant pathogen Erwinia amylovora. is a member of the Enterobacteriaceae
with a close phylogenetic relationship to human enterobacterial pathogens. E. amylovora

secretes at least four effector proteins under hrp-inducing conditions in vitro: Eopl ,

106

Eop3, Aerpt2Ea/Eop4, and DspE (28). Deletion mutants of eopI, a predicted YopJ
family cysteine protease, and aerptZEa, a papain-like cysteine protease homolog, have
been characterized and were shown to mildly affect host range and virulence,
respectively (3, 36). eop3, a putative homolog to AverhE from Pseudomonas syringae,
has not been characterized in the literature. Only one effector, DspE, is required for
pathogenicity and population growth in planta (6, 28). Although its functional targets are
unknown, DspE suppresses salicylic acid-mediated cell wall defenses when transiently
expressed in plant cells, and induces necrotic cell death in host and nonhost cells (6, 14).
DspE is homologous to proteins critical for virulence in several important plant
pathogens including Pseudomonas syringae, Ralstonia solanacearum, and Pantoea
stewartii (6). DspE binds directly to the T3SS chaperone DspF, which has been
implicated in the stable expression and efficient secretion of DspE (18). A dspF deletion
mutant is pathogenic, but with a greatly reduced level of virulence, and DspE reporter
constructs containing the N-terminal 203 amino acids are translocated into plant cells in
the absence of DspF (32).

The strong virulence defect of the chaperone DspF has proven useful for studies
of chaperone-effector interactions in this organism, and we hypothesized that the
relatively small number of effectors produced could facilitate the study of intracellular
chaperone-effector dynamics. Because chaperones often interact with a variety of
proteins to complete various cellular roles, we decided to examine putative functional
interactions between DspE, DspF, and other proteins in the cell. Based on the roles of
other T3SS chaperones, we hypothesized that DspF may play a role in modulating

translocation levels of multiple effector proteins or in interaction with components of the

107

T3 SS apparatus. In a directed yeast two-hybrid screen, secreted effectors Eopl and Eop3
were identified as putative interactors of DspF, as were T3SS apparatus components
HrpQ and Hch. The eapI and eop3 genes are localized next to putative chaperone
genes, here named esc1 and esc3. To determine whether dspF, esc1, and esc3 influence
the translocation levels of multiple effector proteins, a series of chaperone and effector
mutants were constructed and assessed for virulence and population growth in planta and
translocation levels of effector-CyaA reporter constructs. Results indicated that deletion
of eopI and eop3 and corresponding specific chaperones had no effect on virulence or
translocation levels of a DspE-CyaA fusion. Deletion of the putative chaperones of Eopl
and Eop3 caused no effect on virulence, although chaperone gene deletions in the dspF
mutant background caused a small early spike in population growth on pears, suggesting
that other chaperones may slightly antagonize growth on pear. These findings point
toward a model of type 1H secretion in E. amylovora in which chaperones each promote
secretion of a single cognate effector.

MATERIALS and METHODS
Genetic techniques and bacterial strains, plasmids, and primers. Bacterial strains and
plasmids are listed in Table l. Oligonucleotides used for cloning are listed in Table 2.
PCR and gene cloning was performed using standard techniques (Ausubel). All
constructs were confirmed by sequencing, which was performed at the Research
Technology Support Facility at Michigan State University.
Mutant construction and complementation. All chromosomal mutants of E. amylovora
were constructed using the Red recombinase system using an adapted technique as

previously described (Zhao et al.2009, Triplett et al. 2009) The resistance cassette for

108

esc1 mutagenesis was amplified from pKD3 using primers esclmutF and esclmutR, and
the esc3 mutagenesis cassette was amplified using primers esc3mutF and R. Gene
knockout was verified by PCR amplification of the insertion cassette fi'om primers inside
the resistance cassette and flanking the inserted gene (Datsenko and Wanner, 2000), and
PCR products were sequenced to confirm the structure of the insert. To complement the
esc1 mutant, esc1 was amplified with primers esprH and escIRXh, digested with HindIII
and XhoI, and ligated into the vector pBBRl-MCSS.

Pear population counts and virulence assays. Bacterial population growth and lesion
size an immature pears was measured as previously described (Triplett et al. 2009, Zhao
et al. 2004). Briefly, bacterial strains were suspended in 0.5x phosphate buffered saline
(PBS) to a concentration of 1.5x10"8 cfu/ml, and 2 uL was inoculated onto surface-
sterilized immature pears. At the noted timepoints, a core of 0.5 cm was collected from
the point of inoculation on each pear, macerated in PBS, and serially diluted and plated
onto LB medium. Developing lesions on the remaining uncored pears were measured for
diameter using digital calipers as previously described (Triplett et al., 2009).

Yeast two-hybrid assays. esc1 and esc3 were amplified from genomic DNA using the
primers Escl F and R and Esc3 F and R, and PCR products were digested with BamHI
and EcoRI and ligated into the prey vector pB42AD. Bait vector inserts for eopI, eop3,
eop4, hch, hrcN, and hch were amplified with primers listed in Table 2, digested with
BamHI and XhoI, and ligated into pGilda in fi'ame with the LexA DNA binding domain.
Bait and prey vectors were co-transformed into Saccharomyces cerevisiae strain EGY48

p80 (pLacZ) using the Zymo yeast transformation kit (Zymo; Orange, CA). Yeast were

109

plated onto SD-galactose medium amended with X-gal, and color change was visualized
after incubation at 30° C for 72h.

Adenylate cyclase assays. Selected E. amylovora genes were amplified with primers
pEoplF and Eole, pEop3F and Eop3R, and pEop4F and Eop4R. Primers pEoplF,
pEop3F, and pEop4F incorporate a hip box promoter into the forward primer to confer
hrp-inducible gene expression. Eop3 was also cloned without the hrp promoter using
forward primer Eop3F. PCR products were digested with XbaI and 83p] and ligated into
pMJH20. Bacterial strains were transformed with CyaA filSiOIl constructs, and expression
was confirmed by western blot of cells grown in hrp-inducing conditions using anti-
CyaA antibody. Overnight cultures were resuspended in phosphate-buffered saline to an
OD. 600 of 0.3, and infiltrated into the leaves of 8-week-old tobacco plants. Two leaf
disks of 0.5 cm in diameter were collected between 8 and 9 hours post-infiltration, when
the first signs of cell collapse were becoming visible. cAMP levels were measured as
previously described (Triplett et al., 2009). Statistical analyses were done using a one-
way analysis of variance and mean separation was accomplished using Fisher’s protected

least significant difference test.

RESULTS
The requirement of DspF in virulence can be partially overcome by ectopic
expression of DspE. It was previously reported that the T388 chaperone DspF is
required for stable accumulation of the pathogenicity factor DspE in E. amylovora cells,
and for secretion into the medium (Gaudriault et al., 2002). However, we observed that a
fusion of the N-terminus of DspE with an adenylate cyclase reporter was stably expressed

and translocated into host cells from a dspF deletion mutant (Triplett et al., 2009). We

110

proposed that the requirement for dspF may either be conferred by the C-terminus of
DspE, or overcome by ectopic expression of DspE from a plasmid. To test whether the
requirement for DspF is overcome by plasmid expression, a full-length clone of dspE
with a C-terminal 6x-histidine tag was constructed in a high-copy cloning vector. This
clone (pLRT43) was introduced into 1189AdspE and Ea1189AdspF. Degradation or
repression of some effector proteins is mediated by the Lon protease (25). To determine
whether this is true of DspE, a deletion mutant of the Ian protease gene of E. amylovora
was generated in the Eal 189AdspF mutant background and transformed with pLRT43
(dspE-6xHis).

Western blot analysis of lysates fi'om cells grown eight hours in hrp-inducing
media revealed that DspE-6xHis accumulates to detectable levels in 1189AdspE and
Eal 189AdspF, with no difference in band intensity between the two strains. (Figure 4-
1A). DspE-6xHis was present in much greater levels in Ea1189AdspFAlon. These results
demonstrate a negative role for the Lon protease in DspE accumulation, and indicate that
DspE accumulates to equivalent levels in Ea1189AdspE and Eal l89AdspF.

Complemented strains were tested for virulence on immature pear fi'uit.
Ea1189AdspF carrying dspE-6xHis caused lesions significantly larger than Ea1189AdspF
alone, but significantly smaller than Ea1189 or Ea1189AdspE/dspE-6xHis (Figure 4-1B).
These results indicate that the dspF virulence phenotype is partially, but not fully,
compensated by ectopic expression of dspE.

Our previous work showed no difference between secretion levels of an N-
terrninal DspE-CyaA fusion from Ea1189AdspE or Eal 189AdspF. This study was

expanded to include secretion from wild-type Ea1189 and from Ea1189AdspF

111

 

complemented with dspF. To determine whether an N-terminal fragment of DspE is
secreted from the DspF mutant at levels lower than the wild-type, a CyaA assay was
performed as described in Materials and Methods. DspE(1-73 7)-CyaA was translocated
into host cells at lower levels from Eal 189AdspF and Ea1189AdspE than from Ea1189 or
Eal l89AdspF/dspF (Figure 4-1C). There was no clear difference in protein accumulation
between Ea1189, E31189AdspE, and Ea1189AdspF (data not shown). Together, these
results confirm that that full-length DspE can be secreted in the absence of DspF, and
indicate that the N-terminus of DspE is translocated at lower levels from Ea1189AdspF

and Ea1189AdspE than from wild-type Ea1189.

112

Figure 4-1. DspE-6xHis accumulates to reduced levels in Ea1189 AdspF and partially
complements the dspF pathogenicity phenotype. A. Western blots of lysates of
Ea1189AdspE, Ea1189AdspF, and Ea1189 AdspFAlan carrying pLRT43 (DspE-6xHis)
after 8 hr incubation hrp-inducing minmal medium. Ea1189AdspE and Eal l89AdspF
samples are from the same blot, whereas Ea1189AdspFAlon samples were visualized on
a separate blot and exposed for a shorter amount of time. B. Average lesion diameters on
immature pear fi'uits 5 days after inoculation with Ea1189, Ea1189AdspE, Eal l89AdspF,
Eal 189AdspE (pLRT43) and Ea1189AdspF (pLRT43). Error bars denote standard error.
Letters above bars indicate statistically separate groups (p<0.01). C. cAMP accumulation
in tobacco leaves infiltrated with Ea1189, Eal l89AdspE, Eal 189AdspF, and
Ea1189AdspF/dspF expressing DspE(l-737)-CyaA. Letters denote statistical groups
(p<0.05). D. Western blot of DspE(1-737)-CyaA expression in Ea1189, Ea1189AdspE,
and Ea1189AdspF.

113

A I _ _. -200kDa

v

AdspE AdspF AdspF

/Alon
B A 14 a
E 12-
E 10 b
a: .
E 8 c
as 6 ' d
S 4 ‘
E3 2 - e
0

 

'Ea1189 AdspE AdspF AdspE/ AdspF/
pLRT43 pLRT43

25 3

pg cAMP/ ug protein

 

1189 AdspE AdspF AdspF DspE
[dspF (1-15)-

CyaA
DspE (1-737)-CyaA

Figure 4-1.

114

A directed yeast two-hybrid screen identifies secreted proteins and T3SS apparatus
components as new putative interactors of DspF. Chaperone function is often
mediated by the formation of complexes between the chaperone, effector, and other
chaperone-interacting proteins. T3 SS chaperone interactors may include components of
the secretion apparatus, multiple effectors, or regulatory proteins. DspF interacts with at
least one protein other than DspE (17), and a recent report indicated that DspF interacted
with Eopl in a yeast two-hybrid assay (2). We used a yeast two-hybrid assay approach to
test whether DspF interacted with E. amylovora effector proteins or T3SS apparatus
components in yeast. In a study of hrp-dependent secreted proteins of E. amylovora,
Nissinen et al. identified three hrp-dependent secreted effector-like proteins in addition to
DspE, named Eopl, Eop3, and Aerpt2Ea (28). LexA firsions of these three effectors
were constructed and tested for interaction with DspF as described in Materials and
Methods. In a yeast two-hybrid assay, Eopl and Eop3 interacted with DspF, but
AerptZEa did not (Table 3). In addition to effector proteins, highly conserved
components of the secretion apparatus have been demonstrated to interact with various
T3SS chaperones: The EscN/HrcN ATPase (1, 19, 24), the predicted C-ring
(Hch/Ych/Spa33) (30), and a putative substrate specificity switch (I-Ich/Ych/Spa40)
(24). NCBI searches of previously annotated E. amylovora genes and Blast analysis of
the finished, unpublished E. amylovora genome (www.sanger.ac.uk) were used to
identify the homologous genes in E. amylovora to assay for interaction with DspF. LexA
fusions of E. amylovora HrcN and C-terminal derivatives did not interact with B42-HA-
DspF in yeast (Table 3). Attempts to co-purify HrcN and DspF in vitro were unsuccessful

due to difficulty solubilizing HrcN. B42-HA-DspF interacted strongly in yeast with a C-

115

terminal fragment of Hch corresponding to a predicted cleaved cytoplasmic tail. In
addition, a weaker interaction was detected between DspF and. the C-terminal 194 amino
acids of the putative C-ring Hch (Table 3). Neither Hch nor Hch demonstrated
transactivation activity when co-transformed into yeast with the empty pB42AD plasmid
(data not shown). Examination of a putative interaction between the full-length Hch and
DspF is still underway.

Annotation and yeast two-hybrid analysis of other putative chaperone genes
in E. amylovora. T3SS chaperone-encoding genes have often been identified as short
open reading frames immediately adjacent to genes encoding the cognate effector. There
are two putative T3SS chaperone genes adjacent to eapI ; these were designated orfA and
011C (2). 011C was annotated as a putative gene encoding a chaperone to HrpW, the
product of the other gene adjacent to aifC. To determine whether there are other T3SS
chaperones in E. amylovora, I searched the E. amylovora unpublished, finished genome
(www.sanger.ac.uk) for short, chaperone-like open reading frames encoded adjacent to
previously reported T3SS effector proteins. The secreted protein Eop3 is encoded
downstream of an open reading frame encoding a putative 137-aa protein predicted to
have the acidic pl and three at-helical structure characteristic of T3 SS chaperones (Figure
4-2A). However, annotation of the sequence surrounding the gene encoding the secreted
protein Aerpt2Ea, as well as that surrounding the putative effector gene HothoC, did
not yield any open reading frames with the characteristics of a type III chaperone.
Secondary structure prediction by 3D-Jury and Phyre Blast indicated that the product of
011C would not share the secondary structural characteristics of a Class I T3SS

chaperone. A search of the preliminary genome annotation of E. amylovora for genes

116

annotated as “secretion chaperone” yielded additional orfs predicted to be type III
secretion chaperones, but these were located in genomic islands associated with type III
secretion systems SSP-l and SSP-2, which are not expressed under the same conditions
as the hrp secretion system (Y. Zhao, pers. comm.) Blast searches of the E. amylovora
genome against the sequences of several characterized chaperones did not reveal any
other strong candidates for additional T3SS chaperones. In summary, three of the five
putative or confirmed effector-encoding genes previously reported in E. amylovora are
encoded adjacent to predicted chaperone sequences, and no additional putative
chaperones of the hip system were found. Because the effectors predicted to have
chaperones are called Eopl and Eop3, I propose that the putative chaperone genes
adjacent to these be called esc1 and esc3 for Erwinia secretion chaperones 1 and 3,
respectively.

Yeast two-hybrid assay for Escl and Esc3 interaction with E. amylovora effectors
and DspF. A previous study demonstrated Escl and Eopl interact directly in a yeast
two-hybrid assay (2). To determine whether Escl and Esc3 are capable of interaction
with multiple effector proteins in yeast, we examined the interaction of Escl and Esc3
with the three known E. amylovora secreted proteins. Both chaperones interacted with
Eopl , Eop3, and the N-terminal 600 residues of DspE, but did not interact with
Eop4/Aerpt2Ea or a RecA control fusion (Fig 2B). We previously reported that DspF
interacts strongly with residues 51-100 of DspE in yeast and in vitro (32). Escl and Esc3
fusions did not interact with the N-terminal DspF-binding domain of DspE (Figure 4-23).
The chaperone binding region of Eopl was mapped using further yeast studies. While

Escl and DspF both interacted with an N-terminal fragment of Eopl , the interaction

117

appeared much weaker for DspF than Escl. A putative C-terminal binding domain for
Escl, Esc3, and DspF was also identified, although this interaction was masked in longer
fiagments. Color development was stronger in the Esc3 and DspF interaction of the C-
terminus of Eopl than in the Escl interaction. To rule out the possibility of
heterodirnerization, DspF was also tested for interaction with Escl or Esc3, and no

interaction was found (data not shown).

118

Figure 4-2. Gene organization and interactions of putative chaperone effector pairs in E.
amylovora. A. eopl and eop3, but not eop4, are encoded next to putative chaperone
genes. Depiction of eopI region adapted from Vannesste et al., 2000; eap4 region
adapted from Zhao et al, 2006, and eop3 region based on analysis of finished E.
amylovora genome sequence (www.sanger.ac.uk). esc1 and esc3 are predicted to encode
effector-binding secretion chaperones, and orfC is predicted to encode a putative HrpW-
specific chaperone. Accessions for Eopl, Eop3, Eop4, Esc1(OrfA) and OrfC are
AAA98413.1, ABM65604.1, ABC70473, AAF63396.1, and AAC62316. The predicted
product of esc3 is 94% identical to accession YP_0026483 87 from the genome of
Erwinia pyrifoliae, a close relative of E. amylovora. yegS, efp, and blc are predicted to
encode proteins homologous to a diacyl glycerol kinase (nearly identical to Erwinia
pyrifoliae accession YP_0026483 85), elongation factor P (YP_002647518), and an outer
membrane lipoprotein (YP_002647520), respectively. B. Yeast two-hybrid interaction
between prey fusions of predicted chaperones Escl , Esc3, and DspF and bait fusions of
effectors Eopl, Eop3, DspF, and Eop4. C. Color development in Eopl-chaperone
interactions in yeast. Escl, Esc3, and DspF showed no color change when co-transformed
with the empty prey vector pGilda (not shown).

119

 

 

Eopl

Eop3
DspE

Eop4

Figure 4-2.

Parr
1-200 " fl ..
101-263 I

264-402 I :_', :3

120

 

u m
U) m
g a
+ +
+/. +
+ +/-
+ +
+ +
+ -

+ gosg

DspF does not facilitate translocation of Eopl, Eop3, or Eop4 in an adenylate
cyclase assay. Because Eopl and Eop3 were identified as putative interactors of DspF in
yeast, we wanted to determine whether DspF influences translocation levels of fusions of
Eopl, Eop3, or Eop4 in a CyaA translocation assay. To determine the effect of DspF on
translocation of heterologous effectors, plasmids expressing Eopl-CyaA, Eop3-CyaA,
and Eop4-CyaA from a hip box promoter were constructed. These clones were
introduced into Ea1189, Eal 189AdspF, and Ea1189AdspF/dspF. CyaA fusion expression
was confirmed by western blot of hip-induced cultures; levels of protein accumulation in
cultured cells appeared similar between strains (data not shown). cAMP levels in
infiltrated tobacco leaves was measured at between 8 and 9 hr post-inoculation, when the
wild-type and dspF complemented treatments were showing early signs of water-soaking.
Although E. amylovora expressing DspE(l-73 7)-CyaA accumulated cAMP to greater
levels when translocated fiom the wild-type Ea1189 than fi'om E31189AdspE or AdspF
(Figure 4-1C), CyaA fusions of Eopl and Eop3 elicited a greater accumulation of cAMP
levels from an Ea1189AdspF mutant than fi'om the wild-type (Figure 3A and 3B).
However, deletion of dspF did not have a significant influence on cAMP accumulation
mediated by Eop4-CyaA (Figure 3C). Preliminary results indicate that Eopl-CyaA and
Eop3-CyaA, but not Eop4-CyaA, also elicited increased cAMP levels when expressed
from Eal l89AdspE, indicating that the effect is not dspE-independent (data not shown).
DspE(1-15)-CyaA, a non-translocated fragment of DspE, was used as a negative control
to measure background levels of cAMP. These results demonstrate that DspF does not
promote the translocation of effectors other than DspE, and that DspF affects Eopl and

Eop3 translocation levels differently than Eop4.

121

Deletion of Escl causes a decrease in translocation levels of an Eopl-CyaA fusion,
but does not affect translocation levels of a DspE-CyaA fusion. Because Escl and
Esc3 both interacted with Eopl and Eop3 in yeast, I wanted to determine whether these
chaperones affect translocation of the effectors with which they interact. I focused these
studies on Escl due to time constraints. cAMP accumulation was measured in tobacco
leaves 8 to 9 hours after infiltration with Ea1189AdspF or Ea1189AescIAdspF
expressing Eopl-CyaA or DspE(1-737)-CyaA. cAMP accumulation elicited by Eopl-
CyaA was significantly lower in Ea1189Aesc1AdspF than in Ea1189AdspF, and this
phenotype was partially complemented by expression of esc1 from the plasmid
pLRT224. (Figure 4-4A). However, cAMP accumulation elicited by DspE(l-73 7)—CyaA
was not significantly affected by deletion of esc1 (Figure 4-4B). These results
demonstrate that esc1 is required for full levels of Eopl-CyaA translocation, but not for

translocation of a DspE-CyaA fusion.

122

Figure 4-3. cAMP accumulation in tobacco leaves mediated by E. amylovora
expressing CyaA fusion proteins. A-C. cAMP levels in tobacco after infiltration with
E31189, Eal l89AdspF, and Ea1189AdspF/dspF expressing Eopl-CyaA (A), Eop3-
CyaA (B), and Eop4-CyaA (C). DspE (1-15)-CyaA was expressed in Ea1189 as a
negative control. D. cAMP levels after infiltration with Eal l89AdspE or Ea1189AdspF
expressing a low-expression clone of Eop3-CyaA. Bars represent the means of three
samples collected from separate plants. Error bars denote standard error. Letters above
bars denote different statistical groups (p<0.05). Each experiment was performed at
least twice in different months with similar results

123

Figure 4-3.

pg cAMP/ug protein

A Eopl-CyaA

a
150 b
100 b
50
C

0

1189 AdspF AdspF DspE

/dspF (1-15)-
B CyaA
Eop3-CyaA

  

300
200

100

1 1 89 AdspF AdspF DspE
/dspF (1-15) -
CyaA

300 a Eop4-CyaA

 

1189 AdspF AdSpF DspE
/dspF (1-15)
-CyaA

124

 

Eopl-CyaA
180 ' a
140 r a
100 ‘
60 - b
20 ' C
.E 0 '
3 AdspF AescI- AescIAdspF DSPE
e AdspF lescI (1'15) '
a. CyaA
an
s B
A:
E 14 1-737-CyaA
° a
on 12
9' a
10
a
8 u
6
4
2 b
0

 

AdspF A9361 - AescI AdspF DspE
MSPF /escI (1-15) -
CyaA

Figure 4-4. cAMP accumulation in tobacco leaves mediated by E. amylovora strains
expressing effector-CyaA fusion proteins. A. cAMP levels in tobacco afier infiltration
with Eal l89AdspF, AescIAdspF, and AescIAdspF/escl expressing Eopl—CyaA. B.
cAMP levels in tobacco after infiltration with the same parent strains expressing DspE
(l-737)-CyaA. Bars represent the means of three samples from separate plants +/-
standard error. Letters above bars denote different statistical groups (p<0.05). Each
experiment was performed twice with similar results.

125

Deletion of esc1 from a dspF mutant background does not affect virulence, but
causes a slight increase in early population growth on pear. To further examine the
effects of the chaperone genes esc1, esc3, and dspF, I measured population growth and
virulence of chaperone and effector mutant strains in immature pears. Pears were
surveyed for population grth 0, 24, and 48 hr after inoculation with Ea1189, single or
double chaperone mutants, or effector mutant E. amylovora. After 24 hours, pears
inoculated with Ea1189AdspF accumulated roughly lO-fold fewer bacterial cells than did
the wild-type Ea1189-treated pears. Populations of Eal l89Aesc1AdspF consistently
accumulated to higher levels than Ea1189AdspF, accumulating to near wild-type levels
after 24 hours. However, there was no difference between Eal l89AdspF and
Ea1189Aesc1AdspF after 48 hr, and no significant difference in the lesion diameters
elicited by the two strains. Early population growth of Eal 189Aesc1 AdspF was reduced
by complementation with esc1 on pLRT224. Although the differences in early population
growth between Eal 189AdspF and Ea1189Aesc1AdspF are only five to seven-fold, they
are consistently repeatable and were statistically significant in two of four experiments.
We also noticed a consistent trend that Eal 189Aesc3AdspF accumulated higher
population grth than Ea1189AdspF after 24 hours (Figure 4-5B). Simultaneous
deletion of esc1 and esc3 caused a phenotype similar to that of deleting esc1 or esc3
alone, with a slight but not significant increase in population growth on day one, and no
effect on virulence (data not shown). Deletion of eopI, eop3, esc1, or esc3 from E31189
or the dspF mutant background had no significant effect on lesion development in
immature pear fruits. Together, these results indicate that the esc1, esc3, eapI , and eop3

genes do not contribute substantially to virulence in planta.

126

   
    

I'EHO' I Ea1189

I Aesc]
1.E+09- AdspF b
1.E+08‘ AescI/AdspF

I AescI/AdspF+escl

  
 
   
  

 

3" 1 E+07. I AeapI/AdspF
a .
O
1.E+06-
1.E+05-
1.E+04-
1.E+03‘ - . -
0h 24h 48h
a
1.E+10‘
I Ea1189
1.E+09‘ Aesc3
AdspF
1.E+08 - I Aesc3/AdspF
I Aeop3/AdspF
1.E+07'

cfu/g

1.E+06'
1.E+05 *

 

0h 24h 431.

Figure 4-5. Bacterial population counts on immature pear fruit inoculated with Ea1189
and mutant derivatives. Pears were inoculated with 1.5 x 10"3 cfu of Eal 189,

Eal l89AdspF, and orfA and eapI (A) or escB and eop3 (B) deletion derivatives. Bars
represent the mean of four replicate pears +/- SD. Letters represent groups of statistical
significance (p<0.05). Experiments were performed twice (B) or three times (A) on
separate days with similar results.

127

DISCUSSION

The events leading to type III secretion are complex and tightly regulated. While
pre- and post-transcriptional modulation of expression controls the timely abundance of
secretion apparatus components and substrates, mechanisms also exist to control
secretion levels of pre-formed proteins. The mechanism underlying this effect is still
under investigation. T3SS chaperones may function in both capacities, although the role
of these chaperones has not been tested in many plant pathogens. This study examined
the roles of DspF and (two other T3SS chaperones in interaction with potential chaperone
interactors and in effector translocation efficiency. We also tested the effect of these
effectors and chaperones on virulence and population growth in planta.

We also identified putative DspF interactors among known secreted proteins of E.
amylovora. Interestingly, two effectors encoded next to predicted chaperones interacted
with B42-HA-DspF, while an effector with no predicted chaperone did not. The capacity
of DspE for DspF-independent translocation and the putative interaction of DspF with
other secreted proteins led us to hypothesize that DspF might contribute to the
translocation of other effectors. Translocation studies demonstrated that deletion of esc1
had a negative effect on translocation levels of the predicted cognate effector Eopl but
not the non-cognate effector DspE. So, contrary to our initial hypothesis, DspF does not
facilitate translocation of multiple effectors.

Interestingly, deletion of dspF had a moderate positive effect on cAMP
accumulation mediated by Eopl and Eop3-CyaA fusions, a negative effect on
translocation of a DspE-CyaA fusion. One possibility is that while effectors that interact

with chaperones compete for T3 SS access, effectors binding no chaperone are not

128

affected by chaperone levels. Alternatively, it is possible that DspF could have some
direct antagonistic effect, or that lowered DspE translocation levels could otherwise
increase translocation efficiency.

It should be noted that plant material was collected as the first visible signs of
hypersensitive cell death were beginning to appear, and we did not rule out the possibility
that early-stage cell death or other effects of DspE translocation might have caused
artificial reduction in cAMP levels in Ea1189 and Eal l89AdspF/dspF. Another
limitation of the cAMP assay strategy is that effector expression from a multicopy
plasmid may reduce the true effect of chaperones or competition between effectors;
studies with chromosomal reporter genes would be ideal to assess the role of DspF in its
true scope. Despite these limitations, patterns of cAMP accumulation induced by DspE
and Eop4 suggest that these effectors are affected by dspF in a different manner than the
chaperone-binding effectors Eopl or Eop3. Preliminary studies with a Eop3-CyaA fusion
expressed at low levels have been conducted, and indicate that Eop3 is translocated at
threefold greater levels from Eal l89AdspF than from Eal l89AdspE, suggesting that
DspF reduces Eop3 translocation in a manner that is neither DspE-dependent nor an
artefact of cell death. These results led us to hypothesize that some chaperones in E.
amylovora may negatively affect translocation mediated by other chaperone-effector
complexes. Further studies will be conducted to confirm a potential antagonistic effect of
DspF on Eopl and Eop3 translocation and determine its scope and mechanism.

Deletion of effector and chaperone genes in this study yielded no virulence
phenotype. In population growth studies, the only phenotype observed among the esc1

and esc3 chaperone deletions was a slight increase in early growth. This difference was

129

less than tenfold, and was only statistically significant in the esc1/dspF double mutant,
but was consistently repeatable and complemented by addition of esc1 on a plasmid.
Interestingly, Ea1189 strains expressing Eopl-CyaA or Eop3-CyaA displayed reduced
virulence in preliminary pear pathogenicity assays, and a slight negative effect was also
observed in strains overexpressing esc1 or esc3. Clearly, the role of these chaperones, if
any, in bacterial growth is minor. Further population studies, complementations, and
bacterial fitness assays will be performed to help confirm whether certain effectors and
chaperones confer a disadvantage in early growth on pear.

This study identifies two components of the T3 SS base as interactors of DspF in
yeast. Although the interaction between T3SS chaperones and the T3 SS ATPase have
been the focus of much study since 2003 (1, 19, 31), T3SS chaperones are now thought to
interact with other secretion system components (24, 30). The putative interactions
between DspF and secretion apparatus components have not been tested in vitro, but
because interactions between the orthologous secretion proteins and structurally
homologous chaperones have been identified in Xanthomonas campestris and Chlamydia
reinheartii (24, 30), we consider them strong candidate DspF interactors. Given the high
degree of conservation among structural components of the T3SS, it was proposed that
the conserved shape of T3 SS chaperones might act as a signal for recognition by the
secretion apparatus (5). The identification of interactions between diverse chaperones and
components of the T3SS base is very consistent with this theory. If the DspF- T3SS
interactions can be confirmed in vitro, compatability with the yeast two-hybrid strategy
could be helpful in determining the scope of these interactions and identifying the

requisite chaperone features. No functional role for interaction between T3 SS chaperones

130

and Hch or Hch homologs has been identified, and characterization of these
interactions would greatly expand our understanding of chaperone function.

This work includes an examination of the potential role of DspF and the Lon protease
on DspE accumulation. DspE-6xHis was detected at roughly equivalent levels in
Ea1189AdspE and Eal 189AdspF, although protein accumulation was greatly increased
by inactivation of the Lon protease in the dspF mutant background. This finding is
consistent with a previous report implicating T3 SS chaperones in protection from Lon-
mediated degradation in P. syringae (25). In addition, DspE-6xHis partially
complemented the virulence phenotype of Eal l89AdspF, confirming our previous
findings that dspF is not inherently required for DspE secretion. These results suggest
that the requirement for DspF in DspE translocation can be overcome in a concentration-
dependent manner. Although stabilization is a common fimction of chaperones, ectopic
expression is often not sufficient to overcome this function (26).

It is still unclear why DspE fusions accumulate in and are translocated from
Eal l89AdspE at reduced levels equivalent to Eal 189AdspF, rather than wild-type levels.
We considered the possibility that the mutation in Ea1189AdspE might be polar, but the
sequenced structure of the mutation is the same as the dozens of insertional mutants
confirmed to be non-polar by Datsenko and Wanner (13). Alternatively, DspE could
positively regulate its own secretion. Transcriptional and protein blot analysis will be
needed to confirm the putative self-regulation of DspE and determine whether it occurs at
the transcriptional, translational, or secretion level.

The aim of this study was to determine whether DspF functions in a variety of

roles common to other secretion chaperones. We determined that DspF interacts with

131

multiple chaperoned effectors and secretion system components in yeast, and affects
translocation of some effectors differently than others. In the process of characterizing
these roles, we also demonstrated that Eopl, Eop3, and Eop4 are translocated into
tobacco cells and that eapI and eop3 and the putative chaperones esc1 and esc3 are not
required for virulence. Further studies are being planned to determine whether the
protein-protein interactions identified here in yeast are replicated in vitro, and to
determine their functional significance. Many questions remain regarding the mechanism
of these chaperone roles and their significance to real-world pathogenesis, but these
findings contribute significantly to the understanding of Erwinia amylovora chaperone

function.

132

Table 4-1. Bacterial strains and plasmids used in Chapter 4.

 

 

 

 

strain relevant characteristics source
Escherichia coli DH5a F - 80d1acZ AM15 A(lacZYA-argF) U] 69 Invitrogen
endAI recAI hstI 7(rK-mK+) deaR
thi-I supE44 gyrA96 relAI ,1-
Erwinia amylovora
Ea1189 Wild type Burse et al. 2004
Ea1189AdspE dspE deletion mutant, KmR Triplett et al., 2009
Ea1189AdspF dspF deletion mutant, KmR Triplett et al., 2009
Ea1189Aesc1 escI deletion mutant, CmR This study
Eal 189Aesc3 esc3 deletion mutant, CmR This study
Ea1189AdspF/esc1 dspF/esc1 deletion mutant, CmR This study
Ea1189AdspF/esc3 dspF/esc3 deletion mutant, CmR This study
Eal l89AescI/esc3 esc1/esc3 deletion mutant, CmR This study
plasmids relevant characteristics source
pLRT43 DspE-6xHis This study
pLRT224 pBRRl-MCSS expressing esc1 This study
pLRT225 pBRRl—MCSS expressing esc3 This study
pLRT226 B42-AD-Escl This study
pLRT227 B42-AD-Esc3 This study
pLRT215 B42-AD-DspF Triplett et al., 2009
pLRT93 LexA-DspF This study
pLRT219 LexA-DspE (1-600) Triplett et al., 2009
pLRT131 LexA-DspE (1-150) Triplett et al., 2009
pLRT83 LexA-HrcN This study
pLRT131 LexA-HrcN(c200) This study
pLRT228 LexA-Hch(c105) This study
pLRT229 LexA-Hch This study
pLRT230 LexA-Hch (c128) This study
pLRTl 11 LexA-Eopl This study
pLRTl69 LexA-Eopl (1-200) This study
pLRT231 LexA-Eopl (135-263) This study
pLRT232 LexA-Eopl (264-402) This study
pLRT167 LexA-Eop3 This study
pLRT168 LexA-Eop4 This study
pLRT177 Eop 1 -CyaA This study
pLRTl78 Eop3—CyaA This study
pLRT209 Eop3—CyaA This study
pLRT2 1 0 Eop4-CyaA This study
pLRT201 DspE (l-737)-CyaA Triplett et al., 2009
pBRRl-MCSS broad host-range cloning vector, GmR Kovach et al. 1995
pGilda HIS3 LexA BD bait vector, AmpR Clontech
pB42-AD TRPl B42 AD bait vector, AmpR Clontech

 

 

 

133

Table 4-2. Primers for cloning and mutagenesis used in Chapter 4.

 

 

 

Primer Name Primer sequence 5’-> 3’

EoplFBam gatcggatccgcatgaaattatctggactgagtagc

EoleXh gatcctcgagctactttctttgttttcttgcggccag

Eop3FB gatcggatcctcatgagtaaaatgccttcaaggcat

E0p3RXh gatcctcgagttagaggtgcatcctcgcccaactctct

Eop4Fb gatcggatcctcgtgaaagtcagtcatctcacatcc

Eop4RXh gatcctc gagctaattttcactgtataacatggc g

HrcNFB gctggatcctcatggtgatgagcgcgctacagcag

HrcNRXh gatcctcgagtcaatcaggggcatgggcgctgac

HrcN 200F B gatggatcctcgccatcgtctttggcctgatcgg

HchFB acagaattcatgcaaagcttcaggcgctgat

Hch(c128)FB agaggatcctaatgaaaggccaccgtcatgc g

HchRXh atactcgagttattcggcttccagttcgatcacc

Hch(c105)FB gatcggatcctcgcgcagctggcgcactggctg

HchRXh gatcctcgagctatcatccaacctcctggcagg

pEop 1F gatctctagagggaaccggttgcagagaattgcaacataaaaatatctacaacgtttccagctgccatcag

Eole gcgaatatttcttgcggccagaaattcacctg

Eop3F gcgtctagactgttccagaaagtttttcatctg

pEop3F gcgtctagagggaaccggttgcagagaattgcaacataaaaatatctactgttccagaaagtttttcatctg

Eop3R gatcaatattgtgatggtcgttgaaaagataatca
gatctctagagggaaccggttgcagagaattgcaacataaaaatatctacgtgcagattggcgaagtgatt

pEop4F a

Eop4R gataatattattttcactgtataacatggcgtgtgg
gactctagagatctctagagggaaccggttgcagagaattgcaacataaaaatatctagctttagcatggc

EscIFH ggg

EscIRxh catcctcgagtcatgtgattttcccgg
atggcactatctcgtcagttcggtaaccgtgagagcccagcgcttgcccggggtgtaggctggagctgct

EoplmutF tc

EoplmutR ctactttctttgttttcttgcggccagaaattcacctgcccgctgtatttcatatgaatatcctcctta

Eop3mutF atgagtaaaatgccttcaaggcataatctgattttgatacataagactgagggtgtaggctggagctgcttc

Eop3mutR tcagtgatggtcgttgaaaagataatcaaaacttctgccataatactggccatatgaatatcctcctta

EsclmutF atgatcatacaagatttactgggtggattagccagacggctggaagccgggtgtaggctggagctgcttc

EsclmutR tcatgtgattttcccggcgagattcaacgttaatccctgctgcgccagcacatatgaatatcctcctta

Esc3mutF atgcgcaataaagagtttcactctctttgtggagccatcacaaaggtatagtgtaggctggagctgcttc

Esc3mutR ctcgagttacccgggttgcgtgacttccccacggtacatcccgtgcgccacatatgaatatcctcctta

LonmutF atgaatcctgagcgttctgaacgcattgaaatccctgtgttgccgttgcggtgtaggctggagctgcttc

LonmutR ctattttacgaggcaacctgcatgccataaggtgcattttgcagcgccagcatatgaatatcctcctta

 

134

Table 4-3. Yeast two-hybrid identification of putative DspF-interacting proteins.

 

Prey Interaction with DspF
Eopl +
Eop3 +
Eop4 -
HrcN -
HrcN (0200) -
Hch -
Hch (c128) +
Hch (c105) +

 

 

135

10.

ll.

WORKS CITED

. Akeda, Y. and J.E. Galau. 2005. Chaperone release and unfolding of substrates in

type III secretion. Nature 437: 91 1-915.

Asselin, K.E., C. Oh, RM Nissinen, and S.V. Beer. 2006. The secretion of EopB
from Erwinia amylovora. Proceedings of the X International Workshop on Fire
Blight. Ed. C. Bazzi and U. Mazzucchi. Acta Hort 704: 409—415.

Asselin, J.E., K.N. Yip, and S.V. Beer. 2008. Eopl differs in strains of Erwinia
amylovora that differ in host specificity. Proceedings of the XI International
Workshop on Fire Blight. Ed. K.B. Johnston and V.O. Stockwell. Acta Hort 793:
213-214.

Ausubel, F. M., R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A. Smith,
and K. Struhl, eds. 1997. Current Protocols in Molecular Biology. Wiley, New
York.

Birtalan, S.C., R.M. Phillips, and P. Ghosh. 2002. Three-dimensional secretion
signals in chaperone-effector complexes of bacterial pathogens. Mol. Cell 9: 971-980.

Boureau, T., H. El Maarouf-Boteau, A. Garnier, M.-N. Brisset, C. Perino, I.
Pucheu, and M.-A. Barny. 2005. DspA/E, a type III effector essential for Erwinia
amylovora pathogenicity and growth in planta, induces cell death in host apple and
nonhost tobacco plants. Mol. Plant Pathogen Interactions 19: 16-24.

Boyd, A.P., Lambermont, I. and Cornelis, G.R. 2000. Competition between the
Yops of Yersinia enterocolitica for delivery into eukaryotic cells : role of the Sch
chaperone binding domain of YopE. J. Bacteriol. 182: 4811-4821.

Burse, A., H. Weingart, and M.S. Ullrich. 2004. The phytoalexin-inducible
multidrug efflux pump Ach contributes to virulence in the fire blight pathogen,
Erwinia amylovora. Mal. Plant-Microbe Interact. 17: 43-54.

Buttner, D., and Bonas, 2006. Who comes first? How plant pathogenic bacteria
orchestrate type III secretion. Curr. Opin. Microbial. 9: 193-200.

Cornelis, G. 2006. The Type III secretion injectosome. Nature Rev. Microbiol. 4:
81 1-825.

Darwin, K.H. and V.L. Miller. 2001. Type III secretion chaperone-dependent
regulation: activation of virulence genes by SicA and InvF in Salmonella
typhimurium. EMBO J. 20: 1850-1862.

136

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Dasgupta, N., G.L. Lykken, M.C. Wolfgang, and T.L. Yahr. 2004. A novel anti-
anti-activator mechanism regulates expression of the Pseudomonas aeruginosa type
III secretion system. Mal. Microbial. 53: 297-308.

Datsenko, KA. and Wanner, B.L. 2000. One-step inactivation of chromosomal
genes in Escherichia coli using PCR products. Proc. Natl. Acad. Sci. USA 97 : 6640-
6645.

Debroy, S., R. Thilmony, Y.-B. Kwack, K. Nomura, and S.-Y. He. 2004. A family
of conserved bacterial effectors inhibits salicylic acid-mediated basal immunity and
promotes disease necrosis in plants. Proc. Natl. Acad. Sci. USA 101: 9927-9932.

Ehrbar, K., B. Winnen and W. Hardt. 2005. The chaperone binding domain of
SopE inhibits transport via flagellar and SPI-l TTSS in the absence of Ian. Mol.
Microbiol. 59: 248-264.

Galan, J., and H. Wolf-Watz. 2001. Protein delivery into eukaryotic cells by Type
III secretion machines. Nature 444: 567-573.

Gaudriault, S., L. Malandrin, J.-P. Paulin and M.-A. Barny. 1997. DspA, an
essential pathogenicity factor of Erwinia amylovora showing homology with Aer of

Pseudomonas syringae, is secreted via the Hrp secretion pathway in a DspB-
dependent way. Mal. Microbial. 26: 1057-1069.

Gaudriault, S., J.-P. Paulin and M.A. Barny. 2002. The DspB/F protein of Erwinia
amylovora is a type III secretion chaperone ensuring efficient intrabacterial
production of the Hrp-secreted DspA/E pathogenicity factor. Mol. Plant Pathology 3:
3 13-320.

Gauthier, A. and B.B. Finlay. 2003. Translocated intimin receptor and its chaperone
interact with ATPase of the type III secretion apparatus of enteropathogenic
Escherichia coli. J. Bacteriol. 185: 6747-6755.

Kovach, M.E., P.H. Elzer, D.S. Hill, G.T. Robertson, M.A. Farris, R.M. Roop,
and KM. Peterson. 1995. Four new derivatives of the broad host-range cloning

vector pBBRlMCS, carrying different antibiotic-resistance cassettes. Gene 166: 175-
1 76.

Letzelter, M. I. Sorg, L. J. Mota, S. Meyer, J. Stalder, M. Feldman, M. Kuhn, I.
Callebaut and G.R. Cornelis. 2006. The discovery of Sch highlights a new
function for type III secretion effector chaperones. EMBO J. 25: 3223-3233.

Lloyd, S.A., M. Norman, R. Rosqvist and H. Wolf-Watz. 2001. Yersinia YopE is

targeted for type III secretion by N-terminal, not mRNA, signals. Mol. Microbiol. 39:
520-532.

137

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Lorenz, C., O. Kirchner, M. Egler, J. Stuttmann, U. Bonas, and D. Buttner.
2008. HpaA from Xanthomonas is a regulator of type 1H secretion. Mol. Microbiol.
69: 344-360.

Lorenz, C. and D. Buttner. 2009. Functional characterization of the type III
secretion ATPase HrcN from the plant pathogen Xanthomonas campestris pv.
vesicatoria. J. Bacteriol. 191: 1414-1428.

Losada, L.C. and S.W. Hutcheson. 2005. Type III secretion chaperones of
Pseudomonas syringae protect effectors from Lon-associated degradation. Mal.
Microbial. 55: 941-953.

Matsumoto, H, and G.M. Young. 2009. Essential role of the SycP chaperone in type
III secretion of the YspP effector. J. Bacteriol. 191: 1703-1715.

Mills, E., K. Baruch, X. Charpentier, S. Kobi, and I. Rosenshine. 2008. Real-time
analysis of effector translocation by the type III secretion system of enteropathogenic
Escherichia coli. Cell Host and Microbe 3: 104-113.

Nissinen, R.M., A.J. Ytterburg, A.J. Bogdanove, K.J. Van Wijk and S.V. Beer.
2006. Analyses of the secretomes of Erwinia amylovora and selected hrp mutants

reveal novel type III secreted proteins and an effect of Her on extracellular harpin
levels. Mol. Plant Pathol. 8: 55-67.

Parsot, C., Hamiaux, C., and Page, A.L. 2003. The various and varying roles of
specific chaperones in type III secretion systems. Curr. Opin. Microbiol. 6: 7-14.

Spaeth, K.E., Y.-S. Chen, and RH. Valdivia. 2009. The Chlamydia type III
secretion C-ring engages a chaperone-effector protein complex. PloS Pathog.:
e1000579.

Thomas, N., W. Deng, J. Puente, E.A. Frey, C.K. Yip, N.C.J. Strynadka and B.B.
Finlay. 2005. CesT is a multi-effector chaperone and recruitment factor required for
the efficient type HI secretion of both LEE and non-LEE-encoded effectors of
enteropathogenic Escherichia coli. Mol. Microbiol. 57: 1762-1779.

Triplett, L.R., M. Melotto, and G.W. Sundin. 2009. Analysis of the N-terminus of
the Erwinia amylovora secreted effector DspA/E reveals features required for
secretion, translocation, and interaction with the chaperone DspB/F. Mol. Plant

Microbe Interact. 22: 1282-1292.
Wulff-Strobel, C.R., A.W. Williams and S.C. Straley. 2002. Lch and SycH

function together at the Ysc type III secretion system in Yersinia pestis to impose a
hierarchy of secretion. Mal. Microbial. 43: 411-423.

138

34. Winnen, B., M.C. Schlumberger, A. Sturm, K. Schupbach, S. Siebenmann, P.
Jenny and‘W. Hardt. 2008. Hierarchical effector protein transport by the Salmonella
typhimurium SPI-l type HI secretion system. PLOS One 3: e2178 1-8.

35. Zhao, Y., S.E. Blumer, and G.W. Sundin. 2005. Identification of Erwinia
amylovora genes upregulated during infection of immature pear tissue. .1. Bacterial.
187: 8088-8103.

36. Zhao, Y., S.-Y. He and G. Sundin. 2006. The Erwinia amylovora aerptZEA gene

contributes to virulence on pear and Aerpt2EA is recognized by Arabidopsis RPS2
when expressed in P. syringae. Mol. Plant Microbe Interact. 19: 644-654.

139

Chapter 5: Conclusions and Future Directions
Summary of Work

The type III secretion system of gram-negative bacteria is an elegant, ancient, and
powerful piece of equipment, enabling bacterial manipulation of eukaryotic cellular
pathways and establishment of countless pathogenic and symbiotic relationships that
shape our world. The research accomplishments of the last two decades in this area have
revealed the complexity and variety of these systems; they have also shown that an
understanding of the processes leading to secretion and the selective forces shaping it will
require contributions from a variety of fields and fi'om diverse organisms. Efforts to
elucidate the nature of the type III secretion signal and the role of secretion chaperones
have been rewarding, but much of this work so far has been limited to a few families of
T3SS in human pathogens.

My dissertation research has focused on the molecular signals required for the
type III translocation of effector proteins from Erwinia amylovora into plant cells. E.
amylovora is both a devastating pathogen and a system well-suited to translocation
research: it shares many genetic characteristics with other Enterobacterial species, it has
few chaperones and effectors in the genome, and it harbors a conserved chaperone-
effector pair required for pathogenesis. I began this research with the goal of determining
to what extent T3SS signaling and chaperone roles in Erwinia reflect the patterns
elucidated by research in mammalian pathogens. Based on work in these other systems, I
hypothesized that DspF and a DspF-binding domain would be critical for translocation of
DspE. Instead, my work demonstrates that the DspE secretion/ translocation signal is

distinct from the chaperone binding domain and that DspF facilitates but is not required

140

for DspE translocation. In addition, I provide an initial characterization of class I type III
secretion chaperones in the E. amylovora hrp system, candidate interactors, and putative
fimction.

Chapter 2 describes the N-terminal signals required for secretion of the
pathogenicity factor DspE. The most commonly proposed model of the type III secretion
signal is a bipartite signal, in which an N-terminal secretion signal is followed by a
domain necessary for translocation into host cells. In many ways, the N-terminal
characteristics targeting DspE for secretion are similar to those targeting proteins for
secretion from the well-studied T3S systems in Yersinia and Salmonella. DspE has an N-
terminal translocation signal and chaperone binding signal with common size and
predicted secondary structural characteristics. However, the secretion and translocation
signals coincided. Ifthe N-terminal secretion signal is actually an ancient secretion signal
for the flagellar T3 SS, it is possible that secretion of N-terminal DspE truncations is
suppressed by hrp-inducing, flagellar secretion system-repressing media conditions used
here. An experiment testing whether the N-terminal fragments of DspE are secreted
under flagellar-permissive conditions could lend support to the flagellar signal theory.

The secretion/translocation signal of DspE did not coincide with the chaperone-
binding domain, and neither this domain nor the chaperone itself were required for
translocation. Longer DspE fragments were translocated at greater levels, indicating the
presence of a minimal and optimal secretion signal, but the optimal secretion signal also
did not correspond with the chaperone binding domain. This finding demonstrates that
the DspF does not cover a secretion-blocking domain on DspE, nor is it an inherent

requirement for translocation. Although no single mutation completely abolished the

141

DspE translocation signal, I discovered that mutation of a segment including serine-
stretch reduced translocation levels significantly without reducing fusion expression.
Phylogenetic and functional studies would help determine whether the serine stretch is a
valuable part of a translocation signal in E. amylovora.

To determine the significance of dspF on translocation of the N-terminus of
DspE, DspE translocation was initially measured fi'om dspE and dspF deletion mutants to
allow avoidance of DspE-induced cell death and resulting effects on translocation levels
and cAMP accumulation. Unexpectedly, DspE (1-737)-CyaA was translocated at similar
levels from Eal 189AdspE and AdspF. Translocation studies later adapted to include
translocation from Ea1189 revealed that translocation levels from dspE and dspF mutant
strains were significantly lower than from the wild-type, and that the deletion of
Ea1189AdspE had an unforseen negative effect on DspE fusion protein translocation
levels. Translocation levels, as indicated by cAMP levels, were only reduced by about
50% in Eal 189AdspF compared with Ea1189. This finding contrasts with wild-type
DspE, secretion of which is barely detectable from Ea1189AdspF.

The mild phenotype of Ea1189AdspF in translocation assays and the
complementation of Ea1189AdspF by DspE-6xHis suggest that the phenotype of DspF is
partially overcome by ectopic expression of DspE. I did not rule out the possibility that
the C-terminal histidine tag may affect DspE secretion or accumulation levels, but my
findings suggest that the functional significance of dspF, whether in stabilization or
secretion system targeting or both, may be partially compensated for in a concentration-
dependent manner. As Eopl is also translocation-competent in the absence of its

chaperone, it appears that the functional role of E. amylovora chaperones is to make a

142

quantitative improvement in secretion, rather than impose a qualitative requirement. In
wild-type cells, effector levels may be maintained at low enough levels during infection
that chaperones are effectively required. However, given that previous observations
regarding chaperone fimction rely heavily on findings from plasmid-expressed fusion
proteins, my work underscores the importance of achieving wild-type effector expression
levels in future studies of chaperone function. Chaperones are thought to play a critical
role in controlling the temporal or concentration dynamics of effector secretion, and to
understand the scale and significance of these phenomena, future work should aim to
assess translocation levels from chromosomally tagged bacterial lines.

T3SS chaperones have a nearly universally conserved folded shape, confirmed by
numerous solved structures. Surface features proposed to mediate effector-binding
specificity include two conserved hydrophobic pockets, an general overall pattern of
hydrophobic surface sites, or hydrogen bonding mediated by electronegative surface
residues. Emotional studies of these features are few; although a conserved B-sheet
groove is often discussed as a probable site critical to the chaperone-effector interaction,
this site has never previously been functionally analyzed by systematic mutagenesis. I
reasoned that the structural conservation between chaperones could be used to identify
features required for function of diverse chaperones. Using a homology-based model of
DspF, eleven residues with probable placement in the helix-binding groove were
identified. Four hi ghly—conserved residues were identified to be required for DspF
virulence function, three of which are required for interaction with the N-terminus of
DspE, and two of which significantly reduced translocation of a DspE-CyaA fusion in a

dominant-negative manner. The DspF residues required for virulence were predicted to

143

cluster near the same point in the helix-binding groove, suggestive of an effector-binding
hotspot. My research provides evidence that polar residues in the effector-binding groove
are important contributors to the chaperone-effector interaction. Although it is likely that
this groove could interact with residues of the putative helices in residues 51-60 and 73-
82 of DspE, attempts to model such an interaction would be extremely speculative.
Chaperones serve diverse functions in secretion, and these firnctions are mediated
through interaction with diverse cellular proteins. The third goal of this research was to
identify interactors of DspF in yeast and to determine whether selected interactions have
functional significance. DspF interacted with E. amylovora secreted proteins Eopl and
Eop3 in yeast. Because DspF interacts with the same sites of Eopl and Eop3 as the
predicted cognate chaperones of these effectors, I predicted that DspF would influence
translocation levels of several secreted effector proteins, and that deletion of other
secreted effectors might influence virulence in a dspF mutant. In fact, chaperones had no
positive impact on translocation levels of non-cognate effectors, suggesting that
chaperone function is highly specific in E. amylovora. Interestingly, deletion of dspF had
a positive effect cAMP accumulation mediated by Eopl-CyaA and Eop3-CyaA, although
it is unknown whether this effect is indirect or related to the interaction with DspF.
Although deletion of eopI and eop3 and the presumed cognate chaperone genes esc1 and
esc3, respectively, had no effect on symptom development in pear, esc1 or esc3 deletion
led to a slight early increase in population growth. Based on the findings in this study, I
hypothesized that the three hip chaperones in E. amylovora may compete with or
antagonize translocation of other chaperoned effectors. This hypothesis will continue to

be tested in the Sundin lab.

144

DspF and other E. amylovora chaperones were also shown to interact with
putative components of the type HI secretion system in yeast. These interactions have not
been confirmed in vitro in E. amylovora, although similar interactions have been
confumed in other species. The increasingly apparent universality of chaperone-T3 SS
base interactions supports the theory that the three-dimensional chaperone structure
serves as a signal in this interaction.

Questions for the Future

Deletion of dspF could result in a reduction of accumulation of wild-type dspE
and this reduction may be dependent on Lon and other proteases. However, this reduced
accumulation phenotype was not observed in N-terminal DspE fragments or cloned full-
length DspE-6xHis. This suggests that, in contrast to previously characterized effectors
such as Sch and SopE, the chaperone binding domain of DspE is not inherently
destabilizing or secretion-inhibitory. Determining the mechanism by which dspF might
protect dspE from the effects of proteases will be an important step toward connecting the
significance of the chaperone-binding domain, protease targeting sites, and the
translocation signal.

Some of the most intriguing recent work on T3 SS targeting in plant pathogens has
been the first characterization of proteins with predicted involvement in secretion
substrate recognition, including a homolog of the Yersinia substrate-specificity switch
Ych. T3SS chaperones were shown to interact in-vitro with Ych or several other
putative cytosol-exposed components of the T3SS, demonstrating that T3S chaperones
could interact with the secretion apparatus in a manner much more complex than

previously suspected. Exploring these interactions in diverse T3SS and assessing their

145

firnctional significance will contribute substantially to our understanding of chaperone
function and type HI secretion. The materials developed through this work would provide
a good start for investigating these questions in Erwinia amylovora. E. amylovora
secretes a limited number of hip- associated chaperones, for which a series of deletion
mutants have been constructed in this work. Interactions between the three hrp
chaperones identified here and predicted T3SS apparatus components would be assessed
in-vitro, helping to address the question of whether all the chaperones of a T3SS bind to
the same secretion apparatus components with the same affinity. Because the E.
amylovora genome harbors five T3SS, these studies could be expanded to examine the
specificity of the chaperone-T3SS apparatus interaction. The structural model of DspF
could inform mutagenesis studies aimed at identifying surface features required for
interaction with the T3 SS, and translocation and secretion assays could help assess the
functional significance of these interactions. This line of research will add to the quest to
understand the murkiest areas of type III secretion- the substrate recognition machinery,
the translocation apparatus, and the uncharacterized peripheral proteins- and connect the

dots among them.

146

Appendix 1: Genetic Differences Among Blight-Causing Erwinia Species with
Differing Host Specificities Identified by Suppression Subtractive Hybridization

This appendix was published as an article with the above title, co-authored by Y. Zhao
and G.W. Sundin, in Applied and Environmental Biology (2006) vol. 72: 7359-7364.
ABSTRACT

The factors determining host specificity in Erwinia amylovora, the causal agent of
fire blight of apple and pear, are unknown. PCR-based subtractive hybridization was used
to isolate sequences from E. amylovora strain Ea110, pathogenic on apples and pears,
that were not present in three closely-related strains with differing host specificities: E.
amylovora MR1, pathogenic only on Rubus spp.; E. pyrzfoliae Ep1/96, the causal agent
of shoot blight of Asian pears; and Erwinia sp. Ejp556, the causal agent of bacterial shoot
blight of pear in Japan. Six subtractive libraries were constructed and analyzed.
Recovered sequences included type IH secretion components, hypothetical membrane
proteins, and ATP-binding proteins. In addition, we identified an Ea110-specific
sequence with homology to a type IH secretion apparatus component of the insect
endosymbiont Sodalis glassinidius, as well as an Ep1/96-specific sequence with

homology to the Yersinia pestis effector protein tyrosine phosphatase YopH.

147

INTRODUCTION

Erwinia amylovora is the causal agent of fire blight disease of apple, pear, and
many other rosaceous species. Interestingly, strains of E. amylovora that infect raspberry
and other brambles (Rubus spp.) are unable to infect apple and pear (5, 15). The Rubus-
infecting strains of E. amylovora are closely related to the apple and pear-infecting
strains, as evidenced by similar amplified fragment length polymorphism (AF LP) and
PCR fingerprints, species-level total DNA-DNA homology, and nearly identical
sequences in pathogenicity and virulence genes (13, 17, 24-26). Recently, other blight-
causing Erwinia spp. strains with restricted host ranges have been described; E. pyrifoliae
was identified as the cause of Asian pear blight (18), and Erwinia sp. strains isolated in
Japan cause bacterial shoot blight of pears (19). The Asian strains produced no symptoms
when inoculated on apple seedlings (19, 20). Comparisons of chromosomal and plasmid
sequences and AF LP profiles have indicated that E. pyrifoliae is genetically distinct from
but closely related to E. amylovora, and that the Erwinia sp. strains from Japan are more
closely related to E. pyrifoliae than to E. amylovora (18-20, 24, 25). Despite their
phenotypic and genetic similarities, the basis for differences in host specificity between
these Erwinia strains remains unknown. An understanding of the factors restricting host
range in these bacteria could provide important new insights into plant-pathogen
interactions and provide new targets for disease control.

Much of our understanding about the host specificity of bacterial phytopathogens
comes fiom the study of avirulence (avr) genes, the products of which elicit a
hypersensitive response and disease resistance in plants encoding corresponding

resistance (R) genes (22). However, gene-for-gene interactions have not been identified

148

among any of the blight-causing Erwinia species and their respective hosts. As more
bacteria] genome sequences become available, genomic comparison of closely related
strains with different host ranges can provide clues regarding the nature of host
specificity. This approach has led to the identification of several strain-specific genes
with possible host-limiting functions in pathogenic strains of Pseudomonas,
Xanthomonas, and Xylella, respectively (3, 12, 33).

Hybridization-based methods such as suppression subtractive hybridization (SSH)
are useful in the genetic comparison of organisms for which a genome sequence is
unavailable (1). SSH is a PCR-based method that has been used to identify differences
between prokaryotic genomes with differing phenotypes, including pathogenic and non-
pathogenic strains of the same species (27) and between different, closely-related species
(6, 38). SSH has also been used to analyze genetic differences between plant pathogenic
strains varying in host specificity (14, 40).

Our objective in this study was to identify genomic differences among plant
pathogenic Erwinia strains. We used SSH to generate six subtracted libraries to compare
the genomes of fruit tree-infecting E. amylovora strains with E. pyrifoliae, a Japanese
Erwinia sp. strain, and a Rubus-infecting strain of E. amylovora. We hypothesized that
this analysis could reveal novel virulence determinants required for disease on distinct
hosts. These experiments resulted in the identification of strain-specific sequences
including genes encoding a putative type III secretion system (T3SS) effector, a T388
apparatus component, and several putative membrane proteins.

Pathogenicity of Erwinia strains an immature pears. The bacterial strains used

in this study are listed in Table l. Immature pears were inoculated with E. amylovora

149

strains Eal 10, Ea273, and MR1, Erwinia Sp. Ejp556, and E. pyrifoliae Ep 1/96 as
previously described (41). After incubation for three days, pears infected with strains
Eal 10, Ea273, and Epl/96 displayed symptoms typical of fire blight infection including
extensive water-soaking and necrosis accompanied by the production of bacterial ooze
(Figure 1). Pears infected with Ejp556 displayed these same symptoms to a reduced
extent, and pears inoculated with MRI showed no signs of infection (Figure A-l). Cell
counts from inoculated pears revealed that populations of MR1 neither increased nor
decreased in number in immature pears over the four day period following inoculation

(data not shown).

150

 

 

E3110 Ea273 Epl/96 Ejp 556 MR1

10-
9_
3L
7.‘
6L
5."

4C

0 1 2 3 4

Log cfu / g

 

 

Days post-inoculation

Figure A-l. (A) Symptom expression in immature pear fruit following inoculation with
Erwinia amylovora strains. Each fi'uit was inoculated in four places. (B) Growth of E.
amylovora Eal 10 (I), Ea273 (Cl), and MR1 (A); E. pyrifoliae Ep1/96 (A); and Erwinia
sp. strain EJP556 (O) during infection of immature pears. The growth of bacterial strains
was monitored at 0, 1, 2, 3, and 4 days after inoculation. Data points represent the means
of three replicates 1 standard errors. Similar results were obtained in two additional

independent experiments.

151

SSH and isolation of E. amylovora-specific sequences. SSH is used to enrich for
PCR products unique to a tester strain, with amplification of sequences from a driver
strain suppressed. Six subtracted libraries were made: three were enriched for sequences
specific to E. amylovora Ea110- but not hybridizing to restricted host-range strains E.
pyrifoliae Epl/96, Erwinia sp. Ejp556, or E. amylovora MR1, and three were enriched
for sequences specific to the restricted host-range strains (Table 2). E. amylovora Ea273
was a strain of choice because a genome sequencing project of this strain is underway,
(http://www.sanger.ac.uk/Projects/E_arnylovora/) and so this strain was used as a driver
in subtractions enriching for restricted host-range strains. However, Ea273 also contains a
71.5 kb indigenous plasmid previously thought to be similar to a plasmid in E. amylovora
Ea322 (35). The plasmid from E. amylovora Ea322 has no known involvement in
virulence and no known counterpart in Ea110, Ep1/96, Ejp556, or MR1 (25). To avoid
the creation of subtractive libraries enriched with sequences fi'om this plasmid or the
ubiquitous plasmid pEA29, a plasmid-cured strain of E. amylovora Ea110 designated
Eal 10- was used as the tester in SSH experiments.

SSH was performed using the PCR-Select Bacterial Genome Subtraction Kit
(Clontech) following the manufacturer’s instructions except that the primary PCR
reaction was increased to 28 cycles. PCR products were ligated into the pGEM T-Easy
vector (Promega; Madison, WI) and transformed into chemically competent cells of
Escherichia coli DHSaL For the three subtractions that used E. amylovora Ea110- as the
tester, inserts from 96 randomly selected clones were amplified by PCR using primers
specific to the Oligonucleotide adapters, denatured for 10 minutes at 95° C, and spotted

onto duplicate Irnmobilon-P nylon membranes (Millipore; Bedford, MA). Tester and

152

driver genomic DNA was labeled with digoxigenin by random priming using the DIG
DNA labeling kit (Roche Applied Science; Indianapolis, IN) according to the
manufacturer’s instructions. Duplicate membranes were hybridized overnight to tester
and driver probes at 63° C as described in the DIG application manual (Roche). PCR
products hybridizing more strongly to tester DNA were purified, spotted onto new
membranes, and re-probed. Products hybridizing only to tester DNA were sequenced at
the Michigan State University Genomics Technology Support Facility, and sequences
were analyzed using BLASTx and BLASTn (http://www.ncbi.nlm.nih.gov/BLAST/).

The subtraction [Eal 10- - MR1] yielded 10 unique Ea110- -specific sequences,
including three with no significant BlastX match. Four sequences matched hypothetical
genes of unknown function, and two of these were similar to sets of predicted genes fi'om
the insect pathogen and nematode endosymbiont Photarhabdus luminescens (Table 3).
EM4 shared greater than 30% amino acid identity with several exopolysaccharide
acetyltransferases in Burkholderia, Xanthomonas, and Pseudomonas, including an ExoZ
homolog from Pseudomonas syringae pv. tomato. Although the firnctional role of EPS
acetylation has not been studied in plant pathogens, ExoZ is thought to reduce EPS
susceptibility to cleavage in Rhizobium melilati (39).

The subtraction [Ea110- - Epl/96] yielded nine non-redundant clones with
significant BlastX hits, including two with no predicted function (Table 3). A sequence
from clone EPl was similar to that of amsF from E. amylovora, a gene involved in
synthesis of the exopolysaccharide (EPS) arnylovoran (7). EPS, which comprises the

capsule, is an essential pathogenicity factor in E. amylovora. A sequence matching EPl

153

was obtained from the E. amylovora sequence database website
(http://www.sanger.ac.uk/Proj ects/E_amylovora/), revealing a 1.8 kb Orf

with 37% amino acid similarity to AmsF. However, the Orf does not appear to be flanked
by other ams gene homologs and thus is most likely in a different genomic location than
the ams operon. Another sequence, EP2, shared 31% amino acid identity with the type III
secretion T3SS gene ysaQ from the insect endosymbiont Sodalis glassinidius (Table 2).
This T3SS is required for host cell invasion and host transfer of the endosymbiont (10).
EP6 was 91% identical to 268 amino acids of the 506-aa protein MglA, a highly
conserved ABC transporter involved in galactose uptake in Salmonella typhimurium (34).
Although a role for sugar uptake in host specificity has not been determined, the ability to
utilize certain simple carbohydrates is critical for Erwinia virulence (2, 4). In addition, a
number of carbon source utilization and transporter genes were identified in a recent
analysis of genes expressed during E. amylovora pathogenesis (41).

The subtraction [Eal 10- - Ejp556] lead to the identification of seven unique
sequences with matches to known genes (Table 3). El 1 shared 24% amino acid similarity
to RfaK from E. coli, a protein thought to be involved in modification of the
lipopolysaccharide (LPS) core to facilitate O-antigen attachment (21). LPS O-antigen
biosynthetic genes have been found to differ significantly between closely related
Xanthomonas, Xylella, and Serratia marcescens genomes (33, 36, 40). LPS and EPS
form the interface at which Erwinia interacts with its host, so some basis for specificity
may lie in these molecules. More study is needed to determine whether BPS or LPS

structure has a role in Erwinia host range selection.

154

Sequence El 3 isolated in this library was identical to sequence EMS fi'om the
library [Ea110- - MR1], matching an E. coli ATPase; sequence EJ 6 was also similar to a
predicted ATPase. Other sequences isolated from this library matched OrfL and Orfs 106
and 81, predicted genes of unknown function previously identified in the E. amylovora
genome (l 1, 29).

Isolation of sequences specific to Asian Erwinia strains. For subtractions
enriching for sequences specific to Asian Erwinia strains, clone inserts were sequenced at
random prior to hybridization screening. Sequences were screened against the genome of
the driver, Ea273, to screen for tester-specificity. Results were then confirmed by dot-blot
hybridization as described above. Sequences with matches of greater than 80% in the
Sanger database always hybridized to DIG-labeled Ea273 total genomic DNA (data not
shown).

Of 48 clones sequenced from the subtraction [Ejp556 - Ea273], 14 clones were
tester-specific, with nine of these sharing significant homology with known sequences
(Table 4). J E1 contained eij 9, a putative membrane ABC transporter previously
identified in Ejp556 (24), and J E2, J E3, and J E4 also contained gene sequences with
similarity to hypothetical transmembrane proteins. Transporters and other membrane
proteins could influence the host range and lifestyle of a pathogen by determining the
range of substrates and niches available to it. Previous work has reported differences in
outer membrane proteins and an ABC transporter between PAGE profiles of Rubus
strains and other isolates of E. amylovora (5). Ejp556-specific clones J E6 and J E5 shared

69% identity with residues 40-75 of the 75 amino acid precursor to the bacteriocin

155

 

divergicin A, and to the serine protease-like domain of a hypothetical Burkholderia
chitin-binding protein, respectively.

In the subtraction [Epl/96 — Ea273], Ep1/96 plasmid DNA (300 ng) was added to the
primary and secondary hybridization of Ea273 genomic DNA and Epl/96 genomic DNA
to reduce preferential amplification of the small plasmids specific to this strain (25).
Nevertheless, over half of the tester-specific sequences resulting from this subtraction
matched known sequences from these small plasmids (data not shown). Of the remaining
eight tester-specific sequences, five unique sequences were identified with significant
BlastX matches (Table 3). Sequence analysis of clone PE3 showed a
predicted 48% amino acid similarity with the Yersinia pestis type IH effector YopH.
YopH is a protein tyrosine phosphatase that targets the actin cytoskeleton and a variety of
host signaling pathways involved in both adaptive and innate immunity (37). Comparison
of this sequence with the E. amylovora genomic database revealed a 1,587-nucleotide orf
with 57% nucleotide sequence similarity to the Ep1/96 putative YopH homolog, and 33%
similarity to YopH itself. The putative Erwinia YopH homologs, if functional, could
play a role in disrupting signaling pathways in host plants. Mutagenesis of these
sequences in E. amylovora and E. pyrifoliae is underway to determine whether the YopH
homologs have a role in virulence or host specificity.

Analysis of clone PEl revealed it to be nearly identical to clone J E2 from the
subtraction [Ejp556 - Ea273]. These sequences are predicted to share 44% amino acid
identity with predicted Thiamicraspira crunogena protein Tcr_0083, and over 30%
identity with several proteins with putative transmembrane domains. Sequence analysis

of Ep1/96-specific clone PES yielded a 479-nt Orf with homologs present in many gram-

156

negative pathogens of plants and animals. Downstream of this Orf is a sequence
containing a putative Duf7 96 domain, usually found in family 19 chitin-binding genes.
Family 19 chitin-binding proteins are common in plants, although they have been
recently found in several plant pathogenic bacteria, including Xanthomonas strains (33)
and a chitinolytic strain of Burkholderia gladiali (31).

Assay for sequences specific to MR1 but not Ea273. SSH libraries were also
constructed using Rubus-specific strain MR1 as the tester, but no MR1-specific
sequences were found after screening 36 sequences from two independent subtractions. A
previous SSH study of Xylella fastidiosa genomes failed to find tester-specific clones
when the tester genome was slightly smaller than that of the driver (14). Similarly, the
MRI genome could be smaller than that of Ea273. Comparison of Bordetella genomes
found evidence of substantial gene loss during speciation, and it was hypothesized that
genomic decay could have occurred subsequent to host range restriction (9, 28). It is
possible that, once restricted to bramble hosts, the MR1 genome may have lost genes
selected for in pathogens of apple and pear.

Distribution of isolated sequences among Erwinia strains. Ea110-specific
sequence EP2, which was similar to YsaQ from S. glassinidius, was blasted against the E.
amylovora Ea273 genomic sequence (http://www..sanger.ac.uk/Projects/E_amylovora/).
Further analysis of the shotgun database revealed two copies of a putative T3SS with
high similarity to S. glossinidius symbiosis island SSR-l, required for Sodalis invasion of
insect host cells (10). The first copy is approximately 23.5 kb in length and contains 20
predicted orfs. This is preceded by a putative phage integrase gene and has an overall

base composition (38.4% mol G+C) significantly lower than that predicted for the E.

157

amylovora genome (53.5% mol G+C). The second putative T3 SS is approximately 33.4
kb and contains 26 predicted orfs with an overall base composition of 43.4% mol G+C.
To determine the distribution of the new T3SS sequences among Erwinia strains,
DNA probes were generated from PCR products from four predicted Orfs including ysaQ
and yspB, a predicted apparatus protein and effector in the first putative T3 SS,
respectively; and ysa C2 and ysaHZ, both predicted apparatus components in the second
putative T3 SS. A fifth probe was created from ych, a hypothetical protein flanking the
first putative T3 SS. All of these probes hybridized to genomic DNA from strains of E.
amylovora, but not to DNA from other Erwinia strains (Table 5). Thus, if a Sodalis-like
T3SS does exist in the Asian pear pathogens, it is divergent from that of E. amylovora.
The Type IH (Hrp) secretion systems of E. pyrifoliae and E. amylovora associated with
plant infection, in contrast, are highly conserved, sharing 84-96% nucleotide sequence
similarity (32). E. amylovora is insect-disseminated and has been isolated fi'om orchard
populations of several orders of insects with diverse habitats (16), although the exact
interaction between the pathogen and its insect hosts remains poorly understood. The
presence of sequences highly similar to insect endosyrnbionts could indicate a common
ancestry and close phylogenetic relationship between Erwinia spp. and insect-related
enterics, raising the possibility that an insect host might be serving as a mixing vessel for
the exchange of genes between Erwinia strains and other enteric bacteria, facilitating
genetic variation and divergence. Another possibility is that E. amylovora could use these
type III secretion genes in its own symbiotic relationship with an insect host. Studies are
currently underway to determine whether these genomic islands have a role in the

virulence and spread of Erwinia spp.

158

DIG-labeled DNA probes were created from PCR products of several other
sequences isolated in this study to determine their distribution among Erwinia strains.
Results of dot-blot hybridizations are summarized in Table 5. Generally, Ea110- -
specific sequences hybridized to all four strains of E. amylovora tested. However, the
predicted AmsF homolog sequence EPl hybridized to DNA fiom strains Ejp556 and
Ejp557 as well as to E. amylovora, indicating that this sequence is not specific to apple
pathogens. Sequence J E5, a putative chitin-binding protein sequence specific to Ejp556,
hybridized to all four of the Asian Erwinia strains tested. However, both probes derived
from Ep1/96-specific sequences, including the putative YopH homolog PE5, hybridized
only to E. pyrifoliae strains (Table 5). A probe derived from the YopH homolog found in
Ea273 hybridized to all four E. amylovora strains.

Summary. In this study, we used SSH and dot-blot hybridization screening to
identify a number of genomic differences between Erwinia strains with differing host
ranges, including components of two novel type III secretion systems in E. amylovora, a
putative tyrosine phosphatase effector, and several sequences related to membrane
transport or polysaccharide biosynthesis. These putative genes could distinguish the
molecular arsenals of Erwinia strains from one another, and under further investigation
could help explain their differing abilities for invasion, survival, and virulence in their
respective hosts. Our findings also provide some clues about the nature and possible
origin of genetic variation between these Erwinia strains. Alternatively, host range
differences between Erwinia strains could stem from a number of complex factors not
detectable by SSH, including minute sequence differences in pathogenicity genes, post-

transcriptional protein modifications, or a combination of these.

159

10.

WORKS CITED

. Akopyants, N. S., A. Fradkov, L. Diatchenko, J. E. Hill, P. D. Siebert, S. A.

Lukyanov, E. D. Sverdlov, and D. E. Berg. 1998. PCR-based subtractive
hybridization and differences in gene content among strains of Helicobacter
pylori. Proc. Natl. Acad. Sci. USA 95:13108-131 13.

Aldridge, P., M. Metzger, and K. Geider. 1997. Genetics of sorbitol
metabolism in Erwinia amylovora and its influence on bacterial virulence. Mol.
Gen. Genet. 256: 611-619.

Bhattacharyya, A., S. Stilwagen, N. Ivanova, M. D'Souza, A. Berual, A.
Lykidis, V. Kapatral, I. Anderson, N. Larsen, T. Los, G. Reznik, E. Selkov,

J r., T. L. Walunas, H. Feil, W. S. Feil, A. Purcell, J.-L. Lassez, T. L.
Hawkins, R. Haselkorn, R. Overbeek, P. F. Predki, and N. C. Kyrpides. 2002.
Whole-genome comparative analysis of three phytopathogenic Xylellafastidiosa
strains. Proc. Natl. Acad. Sci. USA 99:12403-12408.

Bogs, J., and K. Geider. 2000. Molecular analysis of sucrose metabolism of

Erwinia amylovora and influence on bacterial virulence. J. Bacterial. 182:5351-
5358.

Brauu, P. G., and P. D. Hildebrand. 2005. Infection, carbohydrate utilization,
and protein profiles of apple, pear, and raspberry isolates of Erwinia amylovora.
Can. J. Plant Pathol. 27:338-346.

Brown, N. F., and I. R. Beacham. 2000. Cloning and analysis of genomic
differences unique to Burkholderia pseudomallei by comparison with B.
thailandensis. J. Med. Microbial. 492993-1001.

Bugert, P., and K. Geider. 1995. Molecular analysis of the ams operon required
for exopolysaccharide synthesis of Erwinia amylovora. Mal. Microbial. 15:917-
933.

Cheng, H.-P., and G. C. Walker. 1998. Succinoglycan is required for initiation
and elongation of infection threads during nodulation of alfalfa by Rhizobium
melilati. J. Bacterial. 180:5183-5191.

Cummings, C. A., M. M. Brinig, P. W. Lepp, S. v. (1. Pas, and D. A. Relman.
2004. Bordetella species are distinguished by patterns of substantial gene loss and
host adaptation. J. Bacteriol. 186:1484-1492.

Dale, C., S. A. Young, D. T. Haydon, and S. C. Welburn. 2001. The insect

endosymbiont Sodalis glossinidius utilizes a type III secretion system for cell
invasion. Proc. Natl. Acad. Sci. USA 98:1883-1888.

160

11.

12.

13.

14.

15.

16.

l7.

l8.

19.

20.

21.

Du, Z., V. J akovljevic, H. Salm, and K. Geider. 2004. Creation and genetic

restoration of Erwinia amylovora strains with low levan synthesis. Physio]. Mol.
Plant Pathol. 65:1 15-122.

Feil, H., W. S. Feil, P. Chain, F. Larimer, G. DiBartolo, A. Copeland, A.
Lykidis, S. Trong, M. Nolan, E. Goltsman, J. Theil, S. Malfatti, J. E. Loper,
A. Lapidus, J. C. Detter, M. Land, P. M. Richardson, N. C. Kyrpides, N.
Ivanova, and S. E. Lindow. 2005. Comparison of the complete genome

sequences of Pseudomonas syringae pv. syringae B728a and pv. tomato DC3000.
Proc. Natl. Acad. Sci. USA 102: 11064-11069.

Giorgi, S., and M. Scortichini. 2005. Molecular characterization of Erwinia
amylovora strains from different host plants through RF LP analysis and
sequencing of hrpN and dspA/E genes. Plant Pathol. 54:789-798.

Harakava, R., and D. W. Gabriel. 2003. Genetic differences between two

strains of Xylella fastidiosa revealed by suppression subtractive hybridization.
Appl. Environ. Microbiol. 69:1315-1319.

Heimann, M. F., and G. L. Worf. 1985. Fire blight of raspberry caused by
Erwinia amylovora in Wisconsin. Plant Dis. 69:360.

Hildebrand, M., E. Dickler, and K. Geider. 2000. Occurrence of Erwinia
amylovora on insects in a fire blight orchard. J. Phytopathol. 148:251-256.

Jock, S., and K. Geider. 2004. Molecular differentiation of Erwinia amylovora
strains from North America and of two Asian pear pathogens by analyses of
PF GE patterns and hrpN genes. Environ. Microbiol. 6:480-490.

Kim, W. S., L. Gardan, S.-L. Rhim, and K. Geider. 1999. Erwinia pyrifoliae
sp. nov., a novel pathogen that affects Asian pear trees (Pyrus pyrifolia Nakai).
Int. J. Syst. Bacterial. 49:899-906.

Kim, W. S., M. Hildebrand, S. Jock, and K. Geider. 2001. Molecular
comparison of pathogenic bacteria fi'om pear trees in Japan and the fire blight
pathogen Erwinia amylovora. Microbiology 147:2951-9.

Kim, W. S., S. Jock, J.-P. Paulin, S.-L. Rhim, and K. Geider. 2001. Molecular
detection and differentiation of Erwinia pyrifoliae and host range analysis of the
Asian pear pathogen. Plant Dis. 85:1 183-1 188.

Klena, J. D., R. S. d. Ashford, and C. A. Schnaitman. 1992. Role of
Escherichia coli K-12 tfa genes and the rfp gene of Shigella dysenteriae 1 in

generation of lipopolysaccharide core heterogeneity and attachment of 0 antigen.
J. Bacterial. 174:7297-7307.

161

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Leach, J.E., and RF. White. 1996. Bacterial avirulence genes. Annu. Rev.
Phytopathal. 34: 153-179.

Mavrodi, D. V., O. V. Mavrodi, B. B. McSpadden-Gardener, B. B Landa, D.
M. Weller, and L. S. Tomashow. 2002. Identification of differences in genome
content among pth-Positive Pseudomonasfluorescens strains by using PCR-
based subtractive hybridization. Appl. Environ. Microbial. 68:5170-5176.

Maxson-Stein, K., G. C. McGhee, J. J. Smith, A. L. Jones, and G. W. Sundin.
2003. Genetic analysis of a pathogenic Erwinia sp. isolated from pear in Japan.
Phytopathalogy 93: 1393-1399.

McGhee, G.C., E.L. Schnabel, K. Maxson-Stein, B. Jones, V.K. Stromberg,
G.H. Lacy, and A.L. Jones. 2002. Relatedness of chromosomal and plasmid

DNAs of Erwinia pyrifoliae and Erwinia amylovora. Appl. Environ. Microbial.
68: 6182-6192.

McManus, P. S., and A. L. Jones. 1995. Genetic fingerprinting of Erwinia
amylovora strains isolated from tree-fi'uit crops and Rubus. spp. Phytopathalogy
85: 1547-1553.

Miyazaki, J., W. Ba-Thein, T. Kumao, H. Azaka, and H. Hayashi. 2002.
Identification of a type III secretion system in uropathogenic Escherichia coli.
FEMS Microbial. Lett. 212:221-8.

Moore, R. A., S. Reckseidler—Zenteno, H. Kim, W. Nierman, Y. Yu, A.
Tuanyok, J. Warawa, D. DeShazer, and D. E. Woods. 2004. Contribution of
gene loss to the pathogenic evolution of Burkholderia pseudomallei and
Burkholderia mallei. Infect. Immun. 72:4172-4187.

Oh, C.-S., J. F. Kim, and S. V. Beer. 2005. The Hrp pathogenicity island of
Erwinia amylovora and identification of three novel genes required for systemic
infection. Molecular Plant Pathology 6: 125-138.

Ritchie, D. F., and E. J. Klos. 1977. Isolation of Erwinia amylovora
bacteriophage from the aerial parts of apple trees. Phytopathalagv 67:101-104.

Shimosaka, M., Y. Fukumori, T. Narita, X.-Y. Zhang, R. Kodaira, M.
Nogawa, and M. Okazaki. 2001. The bacterium Burkholderia gladiali strain
CHBlOl produces two different kind of chitinases belonging to families 18 and
19 of the glycosyl hydrolases. J. Biosci. Bioeng. 91:103-105.

Shrestha, R., K. Tsuchiya, S. J. Baek, H. N. Bae, I Hwang, J. H. Hur, and C.

K. Lim. 2005. Identification of dspEF, hrpW, and htpN loci and characterization
of the HrpNEp gene in Erwinia pyrifoliae. J. Gen. Plant Pathol. 71:211-220.

162

33.

34.

35.

36.

37.

38.

Silva, A.C.R., J. A. Ferro, F. C. Reinach, C. S. Farah, L. R. Furlan, R. B.
Quaggio, C. B. Monteiro-Vitorello, M. A. Van Sluys, N. F. Almeida, L. M. C.
Alves, A. M. do Amaral, M. C. Bertolini, L. E. A. Camargo, G. Camarotte, F.
Cannavan, J. Cardozo, F. Chambergo, L. P. Ciapina, R. M. B. Cicarelli, L. L.
Coutinho, J. R. Cursino-Santos, H. EI-Dorry, J. B. Faria, A. J. S. Ferreira, R.
C. C. Ferreira, M. I. T. Ferro, E. F. Formighieri, M. C. Franco, C. C.
Greggio, A. Gruber, A. M. Katsuyama, L. T. Kishi, R. P. Leite, E. G. M.
Lemos, M. V. F. Lemos, E. C. Locali, M. A. Machado, A. M. B. N. Madeira,
N. M. Martinez-Rossi, E. C. Martins, J. Meidanis, C. F. M. Menck, C. Y.
Miyaki, D. H. Moon, L. M. Moreira, M. T. M. Novo, V. K. Okura, M. C.
Oliveira, V. R. Oliveira, H. A. Pereira, A. Rossi, J. A. D. Sena, C. Silva, R. F.
de Souza, L. A. F. Spinola, M. A. Takita, R. E. Tamura, E. C. Teixeira, R. I.
D. Tezza, M. Trindade dos Santos, D. Truffi, S. M. Tsai, F. F. White, J. C.
Setubal, and J. P. Kitajima. 2002. Comparison of two Xanthomonas pathogens
with differing host specificities. Nature 417: 459-463.

Stamm, L. V., N. R. Young, and J. G. Frye. 1996. Sequences of the Salmonella
typhimurium mglA and mglC genes. Gene 171:131.

Steinberger, E.M., G.-Y. Cheng, and S.V. Beer. 1990. Characterization of a 56-
kb plasmid of Erwinia amylovora Ea322: Its noninvolvement in pathogenicity.
Plasmid 24:12-24.

Van Sluys, M. A., M. C. de Oliveira, C. B. Monteiro-Vitorello, C. Y. Miyaki,
L. R. Furlan, L. E. A. Camargo, A. C. R. da Silva, D. H. Moon, M. A. Takita,
E. G. M. Lemos, M. A. Machado, M. I. T. Ferro, F. R. da Silva, M. H. S.
Goldman, G. H. Goldman, M. V. F. Lemos, H. El-Dorry, S. M. Tsai, H.
Carrer, D. M. Carraro, R. C. de Oliveira, L. R. Nunes, W. J. Siqueira, L. L.
Coutinho, E. T. Kimura, E. S. Ferro, R. Harakava, E. E. Kuramae, C. L.
Marino, E. Giglioti, I. L. Abreu, L. M. C. Alves, A. M. do Amaral, G. S. Baia,
S. R. Blanco, M. S. Brito, F. S. Cannavan, A. V. Celestino, A. F. da Cunha, R.
C. Fenille, J. A. Ferro, E. F. Formighieri, L. T. Kishi, S. G. Leoni, A. R.
Oliveira, V. E. Rosa, Jr., F. T. Sassaki, J. A. D. Sena, A. A. de Souza, D.
Truffi, F. Tsukumo, G. M. Yanai, L. G. Zaros, E. L. Civerolo, A. J. G.
Simpson, N. F. Almeida, Jr., J. C. Setubal, and J. P. Kitajima. 2003.
Comparative analyses of the complete genome sequences of Pierce’s disease and
citrus variegated chlorosis strains of Xylella fastidiosa. J. Bacteriol. 185: 1018-
1026.

Viboud, G. I., and J. B. Bliska. 2005. Yersinia outer proteins: role in modulation
of host cell signaling responses and pathogenesis. Annu. Rev. Microbial. 59:69-
89.

Walker, J. C., and N. K. Verma. 2002. Identification of a putative pathogenicity
island in Shigellaflexneri using subtractive hybridisation of the S. flexneri and
Escherichia coli genomes. FEMS Microbiol. Lett. 213: 257-264.

163

 

 

39. York, G. M., and G. C. Walker. 1998. The succinyl and acetyl modifications of
succinoglycan influence susceptibility of succinoglycan to cleavage by the
Rhizobium melilati glycanases Eon and Est. J. Bacterial. 180:4184-4191.

40. Zhang, Q., U. Melcher, L. Zhou, F. Z. Najar, B. A. Roe, and J. Fletcher.
2005. Genomic comparison of plant pathogenic and nonpathogenic Serratia
marcescens strains by suppressive subtractive hybridization. Appl. Environ.
Microbiol. 71 :7716-7723.

41. Zhao, Y., S. E. Blumer, and G. W. Sundin. 2005. Identification of Erwinia

amylovora genes induced during infection of immature pear tissue. J. Bacteriol.
187:8088-8103.

164

 

Table A-1. Erwinia spp. strains and their indigenous plasmid content.

 

 

Species and Plasmids and Relevant Host Reference
strain(s) Characteristics
Erwinia amylovora
Eal 10 pEA29 Apple, Pear (30)
Ea110- cured of pEA29 Apple, Pear (41)
Ea273 pEA29, 71.5 Kb plasmid Apple, Pear CUC71;B02
Ea321 pEA29 Crataegus CFBP1367
MR1 pEA29 Rubus sp. (26)
Erwinia pyrifoliae
Ep1/96 pEP36 and 3 small plasmids Asian Pear (18, 25)
Ep4/97 pEP36 Asian Pear (18)
Erwinia sp.
Ejp556 pEJ30 Asian Pear (l9)
Ejp557 pEJ 30 Asian Pear Q9)

 

165

 

Table A-2. Libraries generated by suppressive subtractive hybridization.

 

 

Subtraction No. of unique sequences (No. of sequences
with significant BlastX match)

Ea110- (apple) - MR1 (Rubus spp.) 10 (7)

Ea110- (apple) - Ep1/96 (pear) 11 (9)

Ea110- (apple) - Ejp556 (pear) 10(7)

Ep1/96 (pear) - Ea273 (apple) 8 (5)

Ejp556 (pear) - Ea273 (apple) 14(9)
MR1 (Rubus spp.) - Ea273 (apple) 0

 

166

Table A-3. Sequence analysis of inserts specific to E. amylovora Ea110, but not E.
amylovora MR1, E. pyrifoliae Ep1/96, and Erwinia sp. Ejp556, respectively.

 

 

 

SSH BlastX E
clone Similar protein, organism Predicted fimction value accession no.
Phatorhabdus . .
EMl luminescens Plu3117 hypothetical protein 2 E-l7 CAE15491
a,b Phatorhabdus . .
luminescens Plu3116 hypothetical protein 9 E-17 CAE15490
Phatorhabdus . .
EM2 luminescens Plu 487 6 hypothetrcal protein 2 E-4l CAE17248
a Phatorhabdus . .
luminescens Plu 487 5 hypothetical protein 1 E-S CAE17247
Phatorhabdus DNA-damage
[Eal 10- - EM3 luminescens Plul 815 inducible protein 5 E7 CAE14108
MR1] a Shewanella denitrzficans . ZP_005 8343
S den_3233 phage mtegrase 2 E-36 0
EM 4 Burkholderia xenavarans exopolysaccharide 4 E _43 YP 553663
ExoZ brosynthesrs -
EMS Escherichia coli . ZP_0071476
c EcolB7_01003583 putat‘ve ATPase 9 E58 2
Oceanabacter sp. . .
EM6 Red 6 5-1 4712 hypothetical protein 4 E-28 ZP_0130638
EM7 Pseudomonas p uttda hypothetical protein 9 E-6l ZP_74489O
Pp2746
EPl E. amylovora Ams F ex°P°1ysa°°hflde 5 13-3 CAA54887
brosynthesrs
EP2b Sodalis glossmtdzus T3 SS invasron 2 E-23 AAS 66845
YsaQ protein
EP3b Pseudgzrzuggputzda putative arylesterase 7 E-28 NP_747354
EP4 Salmonélxénterzca ammonium transport 4 E-47 NP_458840
Pseudomonas syringae . .
[Ea110- _ EP5b PSYL 4778 hypothetical protein 8 E-l3 YP_237843
Epl/96] Salmonella typhimurium galactoside ABC
EP6 MglA transporter 7 E-127 AACM149
Klebsiella pneumoniae membrane
EP7 de dipeptidase 3 E-46 CAA41578
EP8 S“’m0"e”“ 8mm“ hypothetical protein 9 E-26 YP_215237
Sc0250
Escherichia coli C3433 hypothetical protein 2 E-6 2110272724
EP9a
Phatorhabdus ftsI precursor 2 E -5 CAE16033

luminescens FtsI

 

167

 

Table A-3 (cont’d).

 

lipopolysaccharide
biosynthesis
unlmown, chitin-
binding domain

EJ l Escherichia coli RfaK 9 E-6 AAA24523

EJ2 Erwinia amylovora Orfl. 1 E-87 AAX39449

EJ3c Efiiiinéii‘igia Predicted ATPase 9 E-58 ZP—Og71476
[E31 10' ' Escherg'chia coli ZP_0071476

Ej p5 56] El 4 EcolB7_01 003 5 86 hypothetrcal protein 1 E-96 5

EJ 5 Escherichia coli Orf 5 hypothetical protein 2 E-15 CAAl 1511
E J 6 Salmonella typhimurium putative cytoplasmrc 7 E-18 NP 462711

YrdR protern '-
E. amylovora Orf 106 hypothetical protein 1 E-37 CAH41996
E. amylovora Orf 81 hypothetical protein 7 E-33 CAH41995

EJ7a

 

a. Clones EMI, EM2, EM3, EP9, and EJ 6 contained sequences from two genes.

b. The sequences in clones EMl, EP2, EP3, and EP5 were each found in more than one
clone.

c. Clones EMS and EJ 3 contained identical insert sequences

168

 

Table A-4. Sequence analysis of SSH inserts specific to Erwinia sp. Ejp556 and E.
pyrifoliae, respectively, but not to E. amylovora Ea273.

 

 

 

SSH Similar protein, Predicted firnction BlastX accession no.
Clone organism E value
Erwinia sp. Ejp556 ATP binding ABC 6 E-40 NP_857628
JEl .
Ejp19 transporter
Thiamicrospira . . 7 E-18 AAB40679
JE2a crunagena Tcr_0083 hypothetrcal protein
JE3 Escherichia coli YbiO putative membrane 4 E-l9 AAN79366
transport protern
Yersinia mallareti . . 7 E-47 ZP_00823862
J E4 Ymola_01 003756 hypothetrcal protein
JES Burkholderiafungorum hypothetical chitin- l E-15 ZP_00281547
[556' Bcep02003426 binding protein
273] JE6 Carnabacterium diver .cin A comm 6 E-6 AAZZ9031
divergens DvnA g1 pr
Escherichia coli C2473 transposase 2 E-19 W809”
JE7 Erwinia carotavara in te se 4 E-ll YP_215278
Eca1054 gm
Caenarhabditis elegans . . l E-5 AAP82647
J E8 KO 6 A9. 1c hypothetrcal protein
Phatorhabdus . . 4 E-17 CAE15530
J E9 luminescens Plu3 l 5 6 hypothetrcal protein
Thiamicrospira . . 5 E-l 8 ABB40679
PEl crunogena Tcr_0083 hypothetrcal protein
PE2 Burkholderia sp. putative conserved 3 E-8 ABB06669
Bcepl 8194_c7625 lipoprotein
PE3b Aeromanisoifimamczda type III effector protein 6 E-8 ABD48950
[1/96 . I a 11
-273] PE4 Escherichia coli Tsr methyl-accepting 8 B16 83850
chemotaxrs protein
Bordetella h o the tical rotein 2 E-50 CAE40125
parapertussis BPPO716 yp p
PESb’C NOVOSP’i‘IgOb‘w” family 19 glycoside 3 13-24 YP_498016
aramatrczvarans h drolase
Sar002002220 y

 

169

 

Table A-5. Dot-Blot hybridization analysis of the distribution of gene sequence among
various blight-causing Erwinia strains.

 

 

SSH Ep 1/ Ep 4/
clone predicted function Eal 10 Ea273 Ea321 MR1 96 97 Ejp 556 Ejp557
EPS synthesis, similar
EPl to AmsF + + + + - - + +
E] l LPS biosynthesis + + + + - - - -
EJ 2 OrfL- unknown + + + + - - - -
EM4 P°iysa°°harlde + + + - - - - -
biosynthesrs
J E5 Chitin-binding protein - - - - + + + +
YopH-like PTPase
PE3 (Ep1/96) ' ’ ' ' + + - -
YopH-like PTPase
- (E 21273) + + + + - - - -
PE 5 Bordetella hypothetrcal _ _ _ _ + + _ _
protein
galactose transporter
EP6 Mgl A + + + + - - - -
EP7 OrD(- unlmown + + + + - - - -
EP2 YsaQ + + + + - - - -
- chZ + + + + — - - -
- YsaC2 + + + + - - - ..
- YsaHZ + + + + - - - -
- YspB + + + + - - - -
- DspE (positive control) + + + + + + + +

 

170