A Legionella pneumophila Effector Protein Encoded in a
Region of Genomic Plasticity Binds to Dot/Icm-Modified
Shira Ninio1, Jean Celli2, Craig R. Roy1*
1Section of Microbial Pathogenesis, Yale University School of Medicine, Boyer Center for Molecular Medicine, New Haven, Connecticut, United States of America,
2Laboratory of Intracellular Parasites, Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Hamilton, Montana,
United States of America
Legionella pneumophila is an opportunistic pathogen that can cause a severe pneumonia called Legionnaires’ disease. In the
environment, L. pneumophila is found in fresh water reservoirs in a large spectrum of environmental conditions, where the
bacteria are able to replicate within a variety of protozoan hosts. To survive within eukaryotic cells, L. pneumophila require a
type IV secretion system, designated Dot/Icm, that delivers bacterial effector proteins into the host cell cytoplasm. In recent
years, a number of Dot/Icm substrate proteins have been identified; however, the function of most of these proteins
remains unknown, and it is unclear why the bacterium maintains such a large repertoire of effectors to promote its survival.
Here we investigate a region of the L. pneumophila chromosome that displays a high degree of plasticity among four
sequenced L. pneumophila strains. Analysis of GC content suggests that several genes encoded in this region were acquired
through horizontal gene transfer. Protein translocation studies establish that this region of genomic plasticity encodes for
multiple Dot/Icm effectors. Ectopic expression studies in mammalian cells indicate that one of these substrates, a protein
called PieA, has unique effector activities. PieA is an effector that can alter lysosome morphology and associates specifically
with vacuoles that support L. pneumophila replication. It was determined that the association of PieA with vacuoles
containing L. pneumophila requires modifications to the vacuole mediated by other Dot/Icm effectors. Thus, the localization
properties of PieA reveal that the Dot/Icm system has the ability to spatially and temporally control the association of an
effector with vacuoles containing L. pneumophila through activities mediated by other effector proteins.
Citation: Ninio S, Celli J, Roy CR (2009) A Legionella pneumophila Effector Protein Encoded in a Region of Genomic Plasticity Binds to Dot/Icm-Modified
Vacuoles. PLoS Pathog 5(1): e1000278. doi:10.1371/journal.ppat.1000278
Editor: Ralph R. Isberg, Tufts University School of Medicine, United States of America
Received May 16, 2008; Accepted December 22, 2008; Published January 23, 2009
This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public
domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Funding: This work was supported by NIH Grant 2R01AI041699-12 (C.R.R.), by The Human Frontiers Science Program (S.N.), and by the Intramural Research
Program of the National Institutes of Health, National Institute of Allergy and Infectious Diseases (J.C.).
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: Craig.email@example.com
L. pneumophila is the causative agent of a severe pneumonia
called Legionnaires’ disease [1,2]. In the environment it can be
found in fresh water reservoirs , in a very large spectrum of
environmental conditions . In these environments L. pneumophila
resides within protozoan hosts, where it is able to survive and
replicate . A large number of protozoa species can provide a
habitat for L. pneumophila, among them Acanthamoeba castellanii,
Hartmanella sp. and Naeglaria sp. . When humans come in
contact with aerosolized contaminated water sources, L. pneumo-
phila can access human alveolar macrophages. The bacterium is
engulfed by these cells, where it is able to proliferate, and may
cause severe disease . L. pneumophila is not transmitted between
individuals  and is therefore thought to have evolved to survive
within its Protozoan environmental hosts, and only infect humans
as an accidental pathogen.
To survive within eukaryotic cells, L. pneumophila requires a type
IV secretion system designated the Dot/Icm system [7,8] that
delivers bacterial effector proteins into the host cell cytoplasm
[9,10]. The Dot/Icm system is crucial for the ability of the
bacterium to remodel the vacuole in which it resides by preventing
delivery of the vacuole to lysosomes , and promoting
recruitment of endoplasmic reticulum (ER)-derived vesicles to this
vacuole to create a unique organelle in which the bacterium
survives and replicates [12–16]. To date, four L. pneumophila
serogroup1 isolates have been fully sequenced. These are the
Philadelphia1 strain , which was derived from the original
isolate obtained from the eponymous outbreak at an American
Legion convention in 1976 , the Lens and Paris strains , an
epidemic and endemic strain, respectively, isolated in France, and
CP000675). Sequence comparison revealed a high degree of
genomic plasticity, with a large number of strain-specific genes
found in each genome .
Using genomic data in conjugation with genetic and biochem-
ical methods, many Dot/Icm substrate proteins have been
identified [18–29]. The function of most of these substrates
remains unknown, however, for some effectors biochemical and
genetic studies demonstrate activities important for the biogenesis
of an organelle that is permissive for L. pneumophila replication
(reviewed in ). The number of substrate proteins identified to
date is higher than was initially predicted, and it is not yet clear
why so many effectors are required for the survival of the bacteria.
PLoS Pathogens | www.plospathogens.org1 January 2009 | Volume 5 | Issue 1 | e1000278
Genomic plasticity and effector abundance could be related to
the versatile lifestyle of L. pneumophila. These bacteria can survive
within a variety of protozoan hosts found in different environ-
ments. Because natural environments probably support a defined
subset of protozoan hosts, it can be predicted that L. pneumophila
strains that have evolved in different environments would possess
slightly different sets of effector proteins that best facilitate the
survival within their environmental hosts. As a first step in
addressing this hypothesis, we have focused our investigation on a
chromosomal region that displays a high degree of plasticity
among the sequenced L. pneumophila genomes. We show that
effectors of the Dot/Icm system are abundant in this region and
demonstrate that one of the effectors encoded in this region is
recruited to vacuoles containing L. pneumophila by a process
requiring Dot/Icm-dependent modifications to the vacuole
Identification of lpg1965 as a substrate of the Dot/Icm
In previous work aimed at identifying novel L. pneumophila
effectors, a screen was conducted using the Dot/Icm component
IcmW as bait in a yeast-two-hybrid system. The screen was
successful at identifying several effectors . Further analysis of
data generated in that screen has led to the identification of an
additional protein fragment capable of interacting with the IcmW
protein. This fragment consists of amino acids 715 to 988 of the
protein encoded by open reading frame (ORF) lpg1965. A
calmodulin-dependent adenylate cyclase (Cya) gene fusion ap-
proach was used to test whether the lpg1965 gene encodes a Dot/
Icm-translocated substrate protein [31–33]. Because Cya enzymatic
activity is very low in the absence of calmodulin, this enzyme is
inactive in bacterial cells, and is activated when delivered into
eukaryotic cells. Fusion of Cya with a translocated effector results in
delivery of the hybrid protein into host cells, resulting in a dramatic
elevation in cAMP levels. Infection of CHO cells with wild-type L.
pneumophila expressing the Cya-lpg1965 fusion protein resulted in an
increase in cAMP that was three logs above the background levels
found in uninfected control cells. When a dotA mutant devoid of a
functional Dot/Icm system was used, cAMP levels were similar to
background levels, indicating that translocation of lpg1965 is
mediated by the Dot/Icmsystem. Because lpg1965 wasidentifiedas
an IcmW-interacting protein, dependency of the IcmS-IcmW
complex for efficient translocation of lpg1965 was tested. As
demonstrated forotherIcmW-interactingsubstrates oftheDot/Icm
system , translocation of lpg1965 was highly dependent on the
IcmS-IcmW protein complex (Figure S1).
The lpg1965 gene is found in a chromosomal region of
high genomic plasticity
Analysis of other sequenced L. pneumophila strains revealed that
lpg1965 is absent in the Lens and Paris genomes. Genomic
plasticity in the chromosomal region encoding lpg1965 was
apparent upon local sequence alignment between the four
available genome sequences (Figure 1). We decided to investigate
the genomic region delineated by the housekeeping genes encoded
by ORFs lpg1962 (peptidyl-prolyl cis-trans isomerase, (ismr)) and
lpg1977 (ThiJ protease, (thiJ)) to determine whether other effector
The survival of intracellular pathogens often involves the
modification of the host vacuole in which the pathogen
resides. This can be achieved through the function of
effector proteins that are delivered into the host cell
cytoplasm using specialized transport machinery. In the
case of Legionella pneumophila, the bacterium that causes
a severe pneumonia known as Legionnaires’ disease, a
type IV secretion system, termed Dot/Icm, delivers a
number of proteins into host cells, resulting in altered
trafficking of the L. pneumophila–containing vacuole. The
mechanisms by which effector proteins are spatially and
temporally regulated in the host cell to facilitate the
survival of the pathogen are not well understood. In this
work, we report the identification of several L. pneumo-
phila effectors encoded in a genomic region of high
plasticity, among them the protein PieA. We demonstrate
the Dot/Icm dependent recruitment of PieA to the L.
pneumophila vacuole and show that the protein binds to
the cytoplasmic face of the vacuole as a result of L.
pneumophila–induced modifications to this vacuole. Our
findings demonstrate that the association of an effector
with host vacuoles can be spatially controlled through
activities mediated by other effector proteins.
Figure 1. High genomic plasticity in the region encoding lpg1965. The Philadelphia1 genomic region flanking lpg1965 (pieC) was analyzed to
obtain the specific genomic context of pieC. ORFs of novel translocated effectors are shown in black. Below is a low-resolution sequence alignment of
this region, to scale, of all four sequenced genomes. Regions absent from the genomes of Lens, Paris and Corby are highlighted with grey bars.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org2January 2009 | Volume 5 | Issue 1 | e1000278
proteins are present. Several genes that reside within this
chromosomal region are found in all four strains, where they
share extremely high sequence identity, and then there are
multiple genes that are absent from one or more of the genomes.
One mechanism that could account for genomic plasticity within
this region is the acquisition of genetic material by horizontal gene
transfer, followed by incorporation of the foreign DNA into the
genome . Genetic material incorporated by horizontal gene
transfer typically has a different GC content compared with the
average GC content of the receiving genome . When
compared to the average genomic GC content of 38.3%,
lpg1965 and its neighboring genes that are not present in all four
strains have a significantly lower GC content of 30.4% (lpg1963),
27.3% (lpg1964), 33.3% (lpg1965) and 33.2% (lpg1966). Although
this analysis supports the hypothesis that these genes were
acquired through a process of horizontal gene transfer, validation
of this hypothesis requires further analysis. Regardless of the
mechanism, these data indicate that lpg1965 is located in a region
where genomic rearrangements have occurred.
Multiple proteins translocated by the Dot/Icm system are
encoded in the lpg1965 region
The observation that the Dot/Icm substrate encoded by
lpg1965 was located in a region of genomic plasticity suggested
a location where other potential substrates of the Dot/Icm system
might reside. To directly test whether additional proteins in the
lpg1965 genomic region are Dot/Icm translocated substrates we
fused Cya to the amino terminus of nine predicted proteins
encoded in this region that were either novel or contained
eukaryotic-like domains, and to three proteins encoded elsewhere
on the chromosome that were predicted paralogues of proteins
encoded in the plasticity region. This analysis revealed ten
additional substrates of the Dot/Icm system (Figure 2A). Thus,
these genes encode Pie (Plasticity Island of Effectors) proteins that
are translocated substrates of the Dot/Icm system. Proteins within
the region of genomic plasticity were designated PieA to PieG.
Proteins outside of the Pie region were designated PpeA and PpeB,
for the two translocated PieE paralogues, and PpgA for the
translocated PieG paralogue. Similar to lpg1965 (PieC), translo-
cation of the other Pie proteins was reduced greatly in a mutant
strain of L. pneumophila deficient in the IcmSW protein complex
(Figure S3). Even with the observation that these pie genes encode
proteins with a functional C-terminal secretion signal recognized
by the Dot/Icm system, expression of the pie genes was analyzed
by reverse-transcription-PCR (RT-PCR) to ensure that these were
not pseudogenes. These data show all the pie genes are expressed
by L. pneumophila (Figure S2).
In Table 1 the Pie proteins and paralogues are organized into
families based on amino acid identity. The degree of homology
between the different family members was calculated using the
multiple sequence alignment software ClustalW . Proteins
PieC and PieD share 14.7% sequence identity and 22.6%
similarity. PieE shares 17.3% and 20.7% identity with PpeA and
PpeB respectively, and 29.4% and 33.1% similarity with these
proteins, respectively. PieG shares 15.7% and 16% identity with
lpg1975 and PpgA respectively, and 23.2% and 25.4% similarity,
The Pie proteins are not essential for the intracellular
multiplication of L. pneumophila
L. pneumophila strain SN178 is derived from parental strain Lp01
and is deficient in nine of the translocated Pie proteins and related
paralogues. SN178 has in-frame chromosomal deletions removing
the genes pieA, pieB pieC, pieD, pieE, pieG, ppeA, and ppeB. Insertional
inactivation of the ppgA gene in SN178 resulted in the strain
SN179, which is a mutant deficient in ten of the Pie proteins and
related paralogues. A. castellanii was infected with Lp01, SN178
and SN179, and intracellular growth of these strains was
compared to an isogenic Dot/Icm-deficient strain having a
mutation in dotA. These data indicate that both SN178 and
SN179 replicate as well as the parental strain Lp01 in A. castellanii
(Figure 2D). The fold increase in colony-forming units (cfu)
recovered from cells infected with Pie-deficient L. pneumophila was
similar to the number recovered from cells infected with the
parental strain Lp01. As expected, the dotA mutant did not
replicate in these cells. Similar results were obtained when
replication was measured in bone marrow-derived macrophages
from an A/J mouse (Figure 2C). Thus, a strain deficient in the
repertoire of Pie proteins and related paralogues has no
measurable intracellular growth defect in macrophages or
protozoan host cells, indicating that these proteins do not play
an essential role in establishment and maintenance of a vacuole
that supports replication of L. pneumophila in cell culture conditions.
Differential localization of Pie proteins in eukaryotic cells
Several of the Pie proteins contain predicted eukaryotic
homology domains (Table 1). Putative coiled coil regions are
found in PieA, PieC, PieD, PieF, and in the PieE family. This
domain is predominantly found in eukaryotic proteins where it
participates in the establishment of protein-protein interactions
involved in a wide range of cellular processes including membrane
tethering and vesicle transport . Another eukaryotic homology
domain identified is the RCC1 motif found in PieG and the
related protein PpgA. RCC1 is a guanine nucleotide exchange
factor for the Ran-GTPase, which is involved in cell cycle control
and other cellular processes . The presence of these putative
domains in the Pie proteins suggests that once within the host cells,
Pie proteins might function to mimic and manipulate cellular
processes to facilitate the intracellular survival of L. pneumophila.
Because subcellular localization of effectors can provide important
insight into their biochemical functions, CHO cells were
transfected with plasmids encoding GFP fusions of different Pie
proteins to examine the distribution of these protein in
mammalian cells. As shown in Figure 2B, GFP-Pie fusion proteins
had different subcellular localization properties. There were
several Pie proteins that appeared to localize to intracellular
membranes. GFP-PieA was concentrated on vesicular structures in
the perinuclear region of the cell. GFP-PieE displayed an ER-like
reticulate pattern, and GFP-PieG localized to small vesicular-like
structures throughout the cell (Figure 2B). None of the Pie proteins
disrupted the structure of the Golgi apparatus when overproduced
(data not shown), which is a phenotype observed for a number of
other Dot/Icm effectors [24,25,39]. Thus, Pie proteins have
unique subcellular distribution phenotypes that could relate to
their ability to target different host proteins and possibly vesicular
PieA is recruited to the L. pneumophila vacuole
The localization of PieA during infection was investigated
further to independently address whether Pie proteins are
translocated into host cells during infection. A polyclonal antibody
specific for the PieA protein was used to determine whether PieA is
found on vacuoles containing L. pneumophila. Vacuoles were
isolated from U937 macrophage-like cells two hours after infection
with L. pneumophila. PieA staining was evident on vacuoles
containing wild-type bacteria (Figure 3A). No staining was
observed on vacuoles containing a pieA mutant (Figure 3A). PieA
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org3January 2009 | Volume 5 | Issue 1 | e1000278
staining was conducted in the absence of permeabilization, and
under conditions where the majority of the vacuoles remain intact.
Thus, the PieA associated with the vacuoles corresponded to
protein on the cytoplasmic face of the vacuole.
In cells ectopically producing GFP-PieA there was a clustering of
LAMP-1-positive late-endosomal/lysosomal vesicles (Figure 3B).
The GFP-PieA protein was found in association with these LAMP-
1-positive vesicles. These data suggest that PieA overproduction
leads to an alteration in the morphology of host endocytic
compartments. Cells producing GFP-PieA were infected with L.
pneumophila to see if PieA overproduction interfered with any cellular
processes important for L. pneumophila trafficking and growth.
Surprisingly, there was a redistribution of GFP-PieA observed in
cells infected with L. pneumophila. The GFP-PieA protein was found
circumferentially localized to vacuoles containing L. pneumophila
(Figure 3C). The observed redistribution of the protein upon
infection is unique to PieA, and was not observed for any of the
other GFP-Pie fusion protein (data not shown). The GFP-PieA
staining on vacuoles containing replicating L. pneumophila delineated
the membrane surrounding the bacteria. Anti-KDEL staining was
used to visualize ER proteins with this retention motif, and showed
that vacuolescontainingL. pneumophila that stained positivefor GFP-
PieA also stained positive with anti-KDEL (Figure 4A). Thus, the
GFP-PieA-positive organelles containing L. pneumophila have the
expected properties of the specialized ER-derived vacuoles that
support L. pneumophila replication.
GFP-PieA co-localization with ER markers was observed only
with the L. pneumophila-containing vacuoles in infected cells,
Figure 2. Pie proteins are translocated substrates of the Dot/Icm system and show distinct localization patterns when produced in
eukaryotic cells. (A) CHO FccRII cells were infected, at an MOI of 30, with L. pneumophila strains wild-type (black bars) or dotA (grey bars) harboring
plasmids expressing Cya fusion with the indicated proteins. One hour after infection cells were lysed and cAMP was extracted and quantified as
described under Materials and Methods. Levels of cAMP were also determined for uninfected cells (uninfected), or cells expressing Cya alone (pCya).
Each bar represents the mean cAMP value obtained from triplicate wells6standard deviation. (B) Epifluorescence micrographs of CHO FccRII cells
expressing the indicated Pie proteins N-terminally fused to GFP, demonstrating unique subcellular distribution phenotypes. (C,D) L. pneumophila
growth rates were determined in mouse bone marrow-derived macrophages (C), and in the protozoan host A. castellanii (D). Intracellular growth of
strain SN179 (open squares) was compared to that of strain SN178 (closed squares), wild-type L. pneumophila strain Lp01 (triangles), and DdotA
mutant strain CR58 (circles). Each time point represents the fold increase in the mean number of viable bacteria recovered from triplicate wells. The
Pie proteins were not essential for the intracellular multiplication of L. pneumophila in these cell types.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org4 January 2009 | Volume 5 | Issue 1 | e1000278
suggesting that protein recruitment occurs in response to a
pathogen-mediated alteration in the vacuole. To investigate
whether pathogen subversion of the ER to create a vacuole that
permits replication was sufficient to induce relocalization of PieA
to an ER-derived vacuole, GFP-PieA producing cells were infected
with Brucella abortus, which similarly to L. pneumophila requires a
type IV secretion system to create an ER-derived vacuole that
supports intracellular replication . GFP-PieA showed partial
co-localization with LAMP-1-positive compartments in B. abortus-
infected cells, but no co-localization of GFP-PieA with the ER
marker calreticulin was detected in these cells, and no co-
localization of GFP-PieA was observed with the B. abortus-
containing vacuole (Figure 4B). These data suggest that intracel-
lular L. pneumophila induce a specific modification to the vacuole in
which they reside, and that this change mediates GFP-PieA
recruitment to the vacuole.
Deletion analysis reveals PieA domains important for
recruitment to mature vacuoles containing L.
Deletion derivatives were constructed to identify amino acid
regions within PieA that are important for interaction of the
protein with vacuoles containing L. pneumophila. All of the
eukaryotic expression plasmids encoding the GFP-PieA deletion
constructs described in Figure 5 produced similar levels of protein
after transfection (data not shown). The recruitment of each PieA
deletion derivative to vacuoles containing L. pneumophila was
measured by fluorescence microscopy after infection (Figure 5).
These data show that a GFP fusion protein containing C-terminal
residues 513–699 of PieA was recruited to vacuoles containing L.
pneumophila as efficiently as the full-length GFP-PieA protein. This
region of PieA was designated the Vacuole Recruitment Domain
(VRD). The GFP-PieA(1–512) protein, having the C-terminal
VRD deleted, did not co-localize with vacuoles containing L.
pneumophila, which indicates that the VRD is both sufficient and
important for vacuole recruitment of PieA. A central region of
PieA was found to have homology to the C-terminal region
containing the VRD (Figure 5, grey bars). Although the internal
region with similarity to the VRD region could not mediate
recruitment of GFP-PieA(1–512) and GFP-PieA(1–614) to the
vacuole, production of GFP-PieA(1–320) resulted in localization of
the protein to the vacuole at low efficiency. Thus, there are
discrete regions in PieA that can target this effector protein to
vacuoles containing L. pneumophila.
In vitro binding of PieA to isolated vacuoles containing L.
To better characterize the binding of PieA to vacuoles, a cell-
free system was established. Vacuoles containing an L. pneumophila
pieA mutant were isolated from infected U937 macrophage-like
cells and immobilized on glass coverslips. Purified PieA(513–699)
protein was incubated with vacuoles. PieA association was
determined by immunofluorescence microscopy following staining
of vacuoles with an aPieA antibody (Figure 6). Fluorescence
microscopy clearly revealed PieA(513–699) protein surrounding
vacuoles containing L. pneumophila. These data show that the C-
Table 1. Pie protein families and their homologues.
PieA (lpg19636) lpc1442 Coiled coil
PieC (lpg1965)lpc1443 Coiled coil
lpp1947lpc1446 Coiled coil
PieE (lpg1969) lpp1952lpl1941 lpc1452 Coiled coil
PpeA (lpg1701)lpp1666 lpl1660lpc1130Coiled coil
PpeB (lpg1702)lpp1667 lpl1661 Coiled coil
PieF (lpg1972)lpp1955lpl1950 lpc1459 Coiled coil
PieG (lpg1976)lpp1959lpl1953 lpc1462 RCC1
PpgA (lpg2224) RCC1
1GeneBank accession number AE017354.
2GeneBank accession number CR628336.
3GeneBank accession number CR628337.
4GeneBank accession number CP000675.
5Homology domains were identified using the SMART  and CD-Search 
6The original ORF name is indicated in parenthesis.
7Each family of parologous proteins is placed in a single cell.
Figure 3. PieA associates with vacuoles containing L. pneumo-
phila. (A) Representative epifluorescence micrographs of L. pneumo-
phila vacuoles isolated from U937 macrophage-like cells infected with
bacteria expressing the fluorescent protein DsRed. Endogenous PieA
was detected on vacuoles isolated from cells infected with wild-type L.
pneumophila using a PieA-specific antibody and FITC-labeled secondary
antibodies. No PieA staining was detected on vacuoles from cells
infected with a pieA mutant strain. CHO FccRII cells transiently
expressing the fusion protein GFP-PieA were (B) fixed and stained for
the lysosomal marker LAMP-1, revealing GFP-PieA staining in regions of
clustered lysosomes, or (C) infected with wild-type L. pneumophila for
seven hours, and stained with DAPI to identify bacterial DNA and host
cell nuclei. Vacuoles containing L. pneumophila are magnified in the
inset of each image. GFP-PieA was observed in association with
vacuoles containing L. pneumophila.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org5January 2009 | Volume 5 | Issue 1 | e1000278
terminal VRD region in PieA defined in vivo mediates protein
binding to isolated vacuoles in vitro. Vacuoles containing DdotA
mutant bacteria were used to determine whether in vitro binding of
PieA to vacuoles containing L. pneumophila was dependent on Dot/
Icm-mediated alterations to the organelle (Figure 6A and 6B).
There was no detectable binding of PieA(513–699) to vacuoles
containing DdotA bacteria. Thus, Dot/Icm-dependent modifica-
tions to vacuoles containing L. pneumophila are required for PieA
binding both in vivo and in vitro. The efficiency of PieA binding to
vacuoles varied depending on the time the vacuoles were isolated
after infection. Vacuoles isolated later in infection achieved a
higher level of PieA binding. Optimal PieA binding was obtained
for vacuoles isolated from cells that were infected for at least two
hours (Figure 6D). PieA binding to vacuoles containing the icmS
icmW double mutant was also tested because the translocation of
multiple effector proteins is abrogated in this mutant [26,41]. PieA
binding to vacuoles containing the icmS, icmW mutant was
impaired, with only 50% binding activity relative to wild-type-
containing vacuoles (Figure S4). Taken together, these data
indicate that Dot/Icm-dependent maturation events mediated by
effectors, requiring IcmSW function for translocation, enable the
efficient binding of PieA to vacuoles containing L. pneumophila.
Protein on the cytoplasmic surface of the vacuole
containing L. pneumophila is required for PieA binding
The in vitro assay was used to determine whether there is a
protein determinant on the vacuole containing L. pneumophila that
is important for PieA binding. Proteins on the surface of vacuole
containing L. pneumophila were digested with Proteinase-K (PK)
prior to incubation with purified PieA(513–699). Treatment of
vacuoles with PK greatly reduced PieA(513–699) binding
(Figure 7A, left panel). PieA(513–699) was associated with 7961
percent of vacuoles in the untreated control reactions compared to
less than 3 percent of the vacuoles that were digested with PK
(Figure 7B). Because PK treatment might disrupt the vacuole
membrane surrounding L. pneumophila, membrane integrity was
assessed after PK digestion by staining isolated vacuoles with an
antibody that binds to LPS on the bacterial surface (aLP). The
percentage of untreated vacuoles that stained positive using the
aLP antibody (2460.6) did not increase after PK treatment
(1663). All of the vacuoles stained positive with the aLP antibody
when the surrounding membrane was permeabilized with
methanol before antibody incubation (Figure 7A and 7B). Thus,
PK treatment did not affect the integrity of the membrane
surrounding isolated vacuoles containing L. pneumophila, indicating
that the inability of PieA to bind to these vacuoles is caused by
digestion of a protein exposed on the cytoplasmic surface of the
Genomic plasticity has been demonstrated for the L. pneumophila
genome in work comparing the genomes of two serogroup1
isolates Lens and Paris . More than ten percent of the genes in
each genome were found to be strain specific . How genomic
plasticity might impact the repertoire of translocated effectors and
bacterial phenotypes has not been investigated. It is possible that L.
pneumophila strains have evolved in association with different
protozoan hosts and that the predominant host species encoun-
tered by an individual strain in nature will impact the acquisition
and maintenance of genes encoding translocated effectors. In
support of this hypothesis, we demonstrate in this study a region of
genomic plasticity that encodes multiple translocated effectors of
the Dot/Icm system, which we called the Pie region.
The Pie region within the strain Philadelphia1 was found to
contain genes encoding seven different Dot/Icm substrates.
Analysis of the Pie region in other sequenced strains of L.
pneumophila revealed that the pieB gene is found only in the
Philadelphia1 genome, pieE, pieF and pieG are present in all four
genomes, and pieA, pieC and pieD are present in two or more of the
four genomes (Figure 1). pieA, pieB, pieC and pieD have a
significantly lower GC content than the genomic average. This
finding suggests that these pie genes were acquired through
horizontal gene transfer, from foreign DNA. Because some of the
pie genes are not present in all L. pneumophila genomes it is possible
that they were acquired after the strain sub-speciation took place.
Another possibility is that the genes were lost from the Lens and
Paris genomes due to lower selection pressure for their existence in
This work clearly illustrates that regions of genomic plasticity in
L. pneumophila can contain genes encoding Dot/Icm substrates. L.
pneumophila effectors have been identified using several genetic and
biochemical methods [21–29], including a systematic search for
eukaryotic-like L. pneumophila proteins [18–20]. Our data suggests
that further examination of other regions of genomic plasticity will
likely reveal additional L. pneumophila effectors that have no obvious
homology to eukaryotic proteins, which would include Dot/Icm
Figure 4. GFP-PieA and KDEL co-localize on the membrane of
vacuoles containing L. pneumophila, but not on those contain-
ing the ER-resident bacterium B. abortus. (A) CHO FccRII cells
transiently expressing the fusion protein GFP-PieA were Infected with
wild-type L. pneumophila for seven hours and stained for the ER marker
KDEL to identify the membrane of vacuoles containing L. pneumophila.
Vacuoles containing bacteria are magnified in the inset of each image.
GFP-PieA is found decorating the L. pneumophila containing vacuole
where it co-localizes with the ER marker KDEL. (B) HeLa cells were
transfected to express GFP-PieA and infected with B. abortus strain
DsRedm-2308 for 24 hours before processing for confocal microscopy
analysis. Bacteria (appear red) do not recruit GFP-PieA (appears green)
to their replicative vacuoles. GFP-PieA remains associated with LAMP1-
positive vesicles (appear blue, left panel) and does not co-localize with
the ER marker calreticulin (appears blue, right panel) around the B.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org6 January 2009 | Volume 5 | Issue 1 | e1000278
substrates that arose by convergent evolution of proteins unrelated
to the eukaryotic factors they mimic or perturb.
PieA was further investigated because of the unique ability of this
protein to bind to vacuoles containing L. pneumophila. Data obtained
from in vivo recruitment and in vitro binding of PieA to vacuoles
containing L. pneumophila indicate that a protein on the cytoplasmic
surface of vacuoles is required for PieA interaction. Other
requirements for PieA binding were that the bacteria within the
vacuoles must have a functional Dot/Icm system and vacuoles must
have matured in the host cell for at least one to two hours. These
data suggest that accumulation of an effector or effector complex on
the vacuole is necessary for PieA recruitment and that it takes
roughly two hours to achieve the required concentration of the PieA
recruitment-determinant on the vacuole. Alternatively, it is possible
that the Dot/Icm system mediates vacuole recruitment of a host
protein that is not found on other ER-derived vesicles in the cell,
and it is this host determinant that is critical for PieA binding.
Importantly, PieA was not recruited to the ER-derived vacuole
containing B. abortus. These data suggest that PieA recruitment to
vacuoles is not a general phenomenon that occurs after invasion of
the ER by pathogens with a functional type IV secretion system.
Thus, it is likely that PieA recruitment is either mediated directly by
a L. pneumophila effector or indirectly by an activity mediated by a L.
pneumophila effector that is not present in B. abortus.
The survival of L. pneumophila within host cells involves an
ordered series of events that are controlled by the Dot/Icm system.
Within minutes of infection, the Dot/Icm system stimulates
efficient uptake of L. pneumophila by the host phagocyte ,
rapidly prevents fusion of endocytic vesicles with the vacuole [43–
45], and stimulates transport and binding of ER-derived vesicles to
the limiting membrane of the vacuole [14,15]. During replication
there is evidence the Dot/Icm system stimulates ubiquitination of
protein on the vacuole surface , modulates NF-kB activation
, and interferes with protein translation . At late stages of
infection the Dot/Icm system assists in controlling bacterial egress
from the spent host cell .
There is evidence that L. pneumophila effectors involved in
controlling distinct cellular processes are spatially and temporally
regulated. Transcriptional control of effector protein expression is
one mechanism that could account for temporal regulation
[28,29]. Spatial regulation of DrrA on the vacuole has been
demonstrated , and is mediated by host determinants on the
plasma membrane that are presumably lost or modified as
vacuoles mature and acquire new membrane from early secretory
vesicles. Proteasome-mediated degradation of ubiquitinated effec-
tors , and phosphoinositide metabolism on the vacuole
membrane , are other mechanisms that have been proposed
to spatially control L. pneumophila effectors. Studies presented here
on PieA indicate that specific modifications to vacuoles controlled
by other L. pneumophila effectors provide spatial and temporal
information that is recognized by other effectors. Based on the
observation that PieA binding requires a protein determinant on
the cytoplasmic face of the vacuole membrane, we speculated that
some effectors act as scaffolding proteins that function at specific
stages of infection to recruit and retain a subset of effectors that
have biochemical functions important for stage specific maturation
events. Future studies will focus on determining the proteins on the
vacuole membrane that interact with PieA, functioning as
determinants important for spatial regulation of PieA. This may
lead to the identification of L. pneumophila effectors involved in
regulating different stages of vacuole maturation.
Materials and Methods
Strains and media
All bacterial strains, plasmids and oligonucleotide primers used in
this study are listed in Table S1. Unless otherwise noted, chemicals
were purchased from Difco. L. pneumophila strains used in this study
were grown on charcoal-yeast extract (CYE) plates as described
previously [50,51]. When needed, chloramphenicol was added to
the media at a concentration of 10 mg ml21.
Primary cells and cell lines were cultured at 37uC in 5% CO2.
CHO cells were grown in minimal essential alpha medium (Gibco)
Figure 5. Deletion analysis of the PieA protein reveals a C-terminal domain sufficient for the recruitment of PieA to vacuoles
containing L. pneumophila. CHO FccRII cells transiently expressing the indicated PieA deletion constructs N-terminally fused to GFP were infected
with wild-type L. pneumophila. Cells were fixed seven hours post infection and scored for the percent of PieA-expressing infected cells where PieA
was detected decorating vacuoles containing L. pneumophila. Values are means from two independent experiments in which 30 vacuoles were
scored for each condition. A schematic illustration of the PieA protein sequence shows the different truncation constructs, as well as the predicted
coiled-coil domains (black bars) and the identified vacuole recruitment domain (VRD - grey bars). PieA C-terminal residues 513–699 were found to be
sufficient for mediating the recruitment of the protein to vacuoles containing L. pneumophila.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org7 January 2009 | Volume 5 | Issue 1 | e1000278
containing 10% heat-inactivated fetal bovine serum (FBS). U937
cells were grown in RPMI-1640+10% FBS. The cells were
activated with PMA for 48 h and replated in 6-well tissue culture
dishes at a concentration of 36106per well before infection with L.
pneumophila. Bone-marrow derived macrophages were cultured
from female A/J mice as described previously . HeLa cells
Figure 6. PieA binds in vitro to vacuoles containing L. pneumophila with an intact Dot/Icm system, but not to vacuoles containing a
dotA mutant. L. pneumophila vacuoles were isolated from U937 macrophage-like cells infected with either a pieA mutant or a dotA mutant strain
expressing the fluorescent protein DsRed. (A) Representative epifluorescent micrographs of L. pneumophila containing vacuoles that were incubated
with purified PieA(513–699) protein. Bound protein (PieA) was detected using a polyclonal antibody directed against PieA and was found on vacuoles
containing the pieA strain, but not on vacuoles containing the Dot/Icm deficient strain dotA. (B) L. pneumophila containing vacuoles were scored to
quantify the percent of vacuoles that bound to PieA. Values are means6standard error of mean for three independent experiments in which 300
vacuoles were scored for each condition. (C) An immunoblot probed with an aPieA antibody. A specific band corresponding to a protein with the
same molecular weight predicted for the pieA product was detected in whole-cell lysates from wild-type L. pneumophila. The aPieA–reactive product
was not detected in lysates isolated from pieA mutant L. pneumophila. The positions of molecular weight standards (kDa) are indicated to the left of
the immunoblot. (D) PieA binding to vacuoles isolated from U937 macrophage-like cells at different times post infection with wild-type L.
pneumophila. Values are means6standard error of mean for three independent experiments. PieA binding reaches a maximum in vacuoles isolated
two hours post infection.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org8 January 2009 | Volume 5 | Issue 1 | e1000278
(ATCC clone CCL-2) were cultured in Dulbecco’s-Modified Eagle
Medium (DMEM) supplemented with 10% fetal calf serum (FCS)
and 2 mM L-glutamine.
Construction of isogenic pie deletion strains
All L. pneumophila mutants used in this study were derived from
the wild-type strain Lp01. Gene deletions were introduced onto
the chromosome of L. pneumophila by allelic exchange as described
previously . Deletion alleles of the pie genes were constructed
using polymerase chain reaction (PCR) to generate DNA
fragments encoding regions of flanking homology that were
immediately 59 to the start codon and 39 to the termination codon
of each gene, or set of consecutive genes. The primers used were
SN76–SN77 and SN78–SN79 for the ppeA–ppeB deletion allele,
SN152–SN153 and SN154–SN155 for the pieE deletion allele,
SN129–SN149 and SN150–SN151 for the pieA–pieD deletion
allele, SN133–SN134 and SN135–SN136 for deletion allele
lpg1975–pieG. For each deletion allele, the 59 and 39 PCR
products were joined by recombinant PCR and the final product
was digested with the enzymes indicated in Table S1, and ligated
into the gene replacement vector pSR47s digested similarly,
creating the plasmids listed in Table S1. Deletion mutant strain
SN178 was created by progressive gene deletions using the above
replacement vectors. For the insertional-inactivation of pie genes,
primers SN72–SN73 were used for the ppgA allele, and primers
SN74–SN75 for the pieA allele. The resulting PCR products were
ligated into suicide vector pSR47 creating plasmids pSN44 (ppgA)
and pSN45 (pieA). Mutant strain SN179 was constructed by
integrating plasmid pSN44 into the genome of strain SN178.
Mutant strain SN122 was constructed by integrating plasmid
pSN45 into the genome of strain Lp01.
The following sets of primers were used for cloning potential
effector proteins: SN34 and SN35 for PieC, SN44 and SN45 for
SN51 for PieA, SN141 and SN142 for PieB, SN145 and SN146 for
lpg1975, SN166 and SN167 for PpeA, SN168 and SN169 for PpeB,
Figure 7. PieA binds to the cytoplasmic face of vacuoles containing L. pneumophila, in a mechanism dependent upon protein–
protein interaction. Representative epifluorescence micrographs of L. pneumophila vacuoles isolated from U937 macrophage-like cells infected
with a pieA mutant strain expressing the fluorescent protein DsRed. (A) L. pneumophila containing vacuoles that were treated with PK to eliminate
surface exposed protein epitopes (Proteinase-K), lost their ability to bind PieA (PieA), as compared to vacuoles treated under the same conditions
without PK (Control). PK treated vacuoles remained intact and their content remained inaccessible to an antibody directed against L. pneumophila
(aLP), unless the vacuoles were first permeabilized using cold methanol (Permeabilzed). (B) The isolated vacuoles were scored to quantify the percent
of vacuoles that bound to PieA(513–699) (PieA) or stained positive for L. pneumophila (aLP). Values are means6standard error of mean for three
independent experiments in which 300 vacuoles were scored for each condition.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org9 January 2009 | Volume 5 | Issue 1 | e1000278
SN170 and SN171 for PpgA, SN175 and SN176 for PieF, SN143
and SN144 for lpg1968, SN177 and SN178 for lpg1973. The
resulting PCR products were digested with the appropriate enzymes
or vector pEGFP-C2 digested BglII/PstI.The resulting pCya derived
vectors encode for fusion proteins consisting of an amino-terminal
M45 epitope tag, followed by amino acid residues 2–399 of the B.
pertussis CyaA enzyme followed by the indicated L. pneumophila
protein. Expression of the Cya fusion proteins is driven by the icmR
promoter located upstream. The resulting pEGFP derived vectors
encode for fusion proteins consisting of EGFP followed by the
indicated L. pneumophila protein. The names and content of the
generated plasmids are listed in Table S1. For the construction of
6His-tagged PieA primers SN91 and SN99 (PieA) or SN147 and
SN148 (PieA513–699) were used. The resulting PCR products were
digested with NdeI and BamHI, and ligated with the pET15b vector
digested similarly, resulting in plasmids pSN63 and pSN73
respectively. For the construction of a DsRed-Express L. pneumophila
expression vector, primers DsRed T1 Fwd and DsRed T1 Rev were
used to amplify the DsRed-Express coding sequence from plasmid
pDsRed-Express. The PCR product was digested with BamHI and
HindIII, and ligated with plasmid pMMB207 digested similarly,
resulting in plasmid pEMC22 encoding for the DsRed-Express
protein under the control of the tac and icmR promoters located
fused to GFP, plasmid pSN39 was digested with NheI and HindIII,
NheI and EcoRI, or NheI and SacI to obtain DNA fragments
encoding for GFP-PieA(1–323), GFP-PieA(1–512) and GFP-PieA(1–
614) respectively. The resulting fragments were ligated with plasmid
pEGFP-C2 digested similarly, creating plasmids pSN56, pSN55 and
pSN54 respectively. For the construction of PieA N-terminal deletion
constructs fused to GFP, plasmid pSN39 was digested with HindIII
and DraIII, EcoRI and DraIII, or SacI and DraIII to obtain DNA
fragments encoding PieA(324–699), PieA(513–699) and PieA(615–
699), respectively. The resulting fragments were ligated with plasmid
pEGFP-C1 (for PieA(324–699) and PieA(513–699)), or with plasmid
pEGFP-C3 (for PieA(615–699)) digested similarly, creating plasmids
pSN59, pSN60 and pSN61 respectively.
Translocation of potential substrates into host cells was assayed
using the Cya fusion approach described previously [26,33].
Briefly, a stable CHO cell line producing FccRII  was used,
cells were plated at 16105cells per well in a 24-well tissue-culture-
treated dish, and infected on the next day with the desired L.
pneumophila strain carrying plasmids pSN24 (PieA), pSN65 (PieB),
pSN20 (PieC), pSN27 (PieD), pSN25 (PieE), pSN87 (PieF), pSN28
(PieG), pSN81 (PpeA), pSN82 (PpeB), pSN83 (PpgA), pSN66
(lpg1968), pSN88 (lpg1973) or pSN67 (lpg1975) expressing the
Cya fused to the gene of interest. The cells were infected at a
multiplicity of infection (MOI) of 30, and then spun five minutes at
1000 rpm to initiate contact and synchronize the infection.
Infected cells were incubated for one hour at 37uC with 5%
CO2. Cells were washed three times in ice-cold phosphate-
buffered saline (PBS) and lysed in cold buffer containing 50 mM
HCl and 0.1% triton x-100 for 30 minutes at 4uC. The lysates
were boiled for five minutes, and neutralized with 30 mM NaOH.
Levels of cAMP were determined using the cAMP Biotrak
enzymeimmunoassay (EIA) system (Amersham Biosciences).
Intracellular growth assays
Intracellular growth assays were conducted in A. castellanii
(ATCC strain 30234) or in bone marrow derived murine
macrophages, as described previously [26,55].
B. abortus infection
A derivative of B. abortus strain 2308 constitutively expressing
monomeric DsRed (DsRedm) was generated as described .
HeLa cells seeded on 12 mm glass coverslips in 24-well plates were
transfected using the FuGene 6TMtransfection reagent (Roche) to
express GFP-PieA 24 hours before infections. For infections,
bacteria grown to late log phase in Tryptic Soy broth were diluted
in complete medium and added to chilled cells at a theoretical
multiplicity of infection (MOI) of 500. Bacteria were centrifuged
onto cells at 4006g for 10 minutes at 4uC, and infected cells were
incubated for 30 minutes at 37uC under 7% CO2 atmosphere
following a rapid warm up in a 37uC water bath to synchronize
bacterial entry. Infected cells were then washed five times with
DMEM to remove extracellular bacteria, incubated for an
additional 60 minutes in complete medium before medium
containing 100 mg/ml gentamicin was added for 90 min to kill
extracellular bacteria. Thereafter, infected cells were maintained
in gentamicin-free medium. At 24 hours post infection infected
cells were washed three times with PBS, fixed with 3% PFA.
For localization of GFP-tagged effectors, CHO FccRII cells
were plated on 12-mm glass coverslips in 24-well tissue culture
plates at a density of 104cells per well. FuGene 6TM(Roche) was
used to transfect the cells with plasmids pSN39 (GFP-PieA),
pSN69 (GFP-PieB), pSN30 (GFP-PieC), pSN35 (GFP-PieD),
pSN37 (GFP-PieE) or pSN33 (GFP-PieG). After 18 hours of
expression cells were either directly fixed using 2% paraformal-
dehyde (PFA), or first infected with wild-type L. pneumophila strain
Lp01 at an MOI of 30, and fixed seven hours post infection. For
PieA deletion analysis, cells were seeded similarly, and transfected
with plasmids pSN39 (GFP-PieA), pSN54 (GFP-PieA(1–614)),
pSN55 (GFP-PieA(1–512)), pSN56 (GFP-PieA(1–323)), pSN59
(GFP-PieA(324–699)), pSN60 (GFP-PieA(513–699)) or pSN61
(GFP-PieA(615–699)). After 18 hours of expression cells were
infected with wild-type L. pneumophila strain Lp01 at an MOI of 20.
Seven hours post infection cells were washed with PBS
supplemented with 0.9 mM CaCl2and 1 mM MgCl2and pre-
permeabilized using 0.1% saponin in pipes buffer (80 mM pipes,
5 mM EGTA, 1 mM MgCl2, pH 6.8) for five minutes before
fixing with 2% PFA. For detection of host and bacterial DNA, cells
were stained with 4,6-diamidino-2-phenylindole (DAPI) for ten
minutes at 25uC. For Lamp1 detection cells were permeabilized
with 0.1% saponin (Sigma), and stained with UH1 mouse anti-
hamster Lamp1 monoclonal antibody (Developmental Studies
Hybridoma Bank), followed by secondary TexasRed anti–mouse
IgG (Invitrogen). For KDEL detection, cells were first permeabi-
lized using cold methanol, and stained with a mouse anti-KDEL
monoclonal antibody (Stressgene) followed by secondary Tex-
asRed anti–mouse IgG (Invitrogen). Polyclonal antibodies against
PieA were produced at Pocono Rabbit Farm and Laboratory
(Canadensis, Pennsylvania) using affinity purified histidine-tagged
protein as antigen to immunize rabbits. Digital images were
acquired with a Nikon TE300 microscope using a 10061.4 N.A
objective lens and a Hamamatsu ORCA-ER camera controlled by
IP Lab software. Images were exported as TIFF files and labeled in
For B. abortus infection experiments samples were blocked and
permeabilized in 10% horse serum, 0.1% saponin in PBS for
30 min at room temperature. Cells were labeled using mouse anti-
human LAMP-1 clone H4A3 antibody (developed by J. T. August
and obtained from the Developmental Studies Hybridoma Bank
developed under the auspices of the NICHD and maintained by
The University of Iowa, Department of Biological Sciences, Iowa
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org10January 2009 | Volume 5 | Issue 1 | e1000278
City, Iowa) or rabbit polyclonal anti-calreticulin antibodies
(Affinity BioReagents) for 45 min at room temperature. Bound
antibodies were detected using Cyanin 5-conjugated donkey anti-
mouse antibodies (Jackson ImmunoResearch Laboratories). Sam-
ples were observed and imaged on a Carl Zeiss LSM 510 confocal
laser-scanning microscope. Confocal images of 102461024 pixels
were acquired as projections of 3 consecutive slices with a 0.38 mm
step and assembled using Adobe Photoshop CS. For endogenous
PieA detection vacuoles were isolated from U937 macrophage-like
cells as described below and stained using rabbit polyclonal aPieA
antiserum at 1:100 dilution, followed by secondary FITC anti–
mouse IgG (Invitrogen).
In vitro PieA binding assay
U937 macrophage-like cells were seeded into 6-well tissue
culture plates at a density of 36106macrophages per well. The
next day, the cells were infected with the indicated L. pneumophila
strain harboring plasmid pEMC22, that were plate grown for two
days in the presence of 1 mM Isopropyl b-D-1-thiogalactopyrano-
side (IPTG) to induce the expression of the fluorescent protein
DsRed-Express. The bacteria were added to the cells at an MOI of
5 in the presence of 1 mM IPTG, and spun five minutes at
1000 rpm to initiate contact and synchronize the infection. One
hour post-infection extracellular bacteria were removed by
washing each well three times with warm PBS. Wells were
refreshed with tissue culture medium containing 1 mM IPTG and
incubated at 37uC for an additional hour. Next, the cells were
placed on ice, and the wells were washed with cold PBS, before
cells were lifted using a cell scraper into 1 ml cold homogenization
buffer (H.B) containing 250 mM sucrose, complete protease
inhibitor cocktail (Roche), and 20 mM Hepes pH 7.2. The cells
were homogenized using a ball-bearing homogenizer, and the cell
homogenate was spun for three minutes at 1500 rpm to sediment
cell nuclei and unbroken cells. The post-nuclear supernatant (PNS)
was diluted 1:5 in cold H.B. and spun onto poly-L-lysine coated
12-mm glass coverslips in 24-well tissue culture plates. The PNS
was fixed by the addition of PFA to 2% for 20 minutes at 25uC.
To test for PieA binding, coverslips were incubated for one hour at
4uC with a blocking solution containing 50 mM ammonium
sulfate and 2% goat serum, supplemented with 2 mg/ml of affinity
purified 6His-PieA(513–699) protein. To remove unbound
protein, the coverslips were washed three times in PBS. For
detection of bound protein coverslips were stained with aPieA
polyclonal antiserum at 1:500 dilution, followed by secondary
FITC anti–rabbit IgG (Invitrogen). Where indicated, vacuoles
were permeabilized for 10 s with ice-cold methanol. The integrity
of the vacuolar membranes was tested using polyclonal antiserum
specific for L. pneumophila serogroup1 (aLP), and FITC-conjugated
anti-rabbit IgG (Invitrogen). Where indicated, coverslips were first
treated with PK by incubating them for two hours at 37uC in PBS
to which PK was added to a final concentration of 10 mg/ml. The
reaction was stopped by washing the coverslips three times in PBS,
and then incubating them for 10 minutes in 1 mM PMSF in PBS,
and finally washing three times with PBS. Control samples were
treated similarly, but without the addition of PK.
Reverse transcription–PCR (RT–PCR)
RNA was isolated from broth-grown L. pneumophila using the
TRIzol Max bacterial RNA isolation kit (Invitrogen). The RNA
was digested with DNase Using an On-Column DNase digestion
kit (Qiagen). RT-PCR was performed in two steps. First strand
synthesis was performed using superscript II reverse transcriptase
kit (Invitrogen) using 5 mg of total RNA and random primers
(Invitrogen), and included a negative control reaction without
reverse transcriptase. The PCR was preformed using taq-
polymerase (Invitrogen) with gene-specific primers, and a 1:50
dilution of the first strand mix as template.
The NCBI accession numbers for the proteins discussed in this
paper are L. pneumophila PieA (YP_095979), PieB (YP_095980),
PieC (YP_095981), PieD (YP_095982), PieE (YP_095985), PpeA
(YP_095992) and PpgA (YP_096236).
Found at: doi:10.1371/journal.ppat.1000278.s001 (0.10 MB PDF)
Strains, plasmids and primers
host cells in a Dot/Icm dependent manner. CHO cells were
infected, at a multiplicity of infection (MOI) of 30, with L.
pneumophila strains harboring plasmid pSN20 expressing the Cya-
lpg1965 fusion proteins, under the icmR promoter. One hour after
infection cells were lysed and cAMP was extracted and quantified
as described under Materials and Methods. To test for Dot/Icm
dependency, translocation was assayed both in cells infected with
the wild-type strain Lp01 (wt) as well as with the dotA mutant
CR58 (dotA). To test for IcmS and IcmW dependency,
translocation was assayed in cells infected with double mutant
CR503 lacking both icmS and icmW (icmW icmS). Levels of cAMP
were also determined in uninfected cells (uninfected). Each bar
represents the mean cAMP value obtained from triplicate
Found at: doi:10.1371/journal.ppat.1000278.s002 (9.78 MB TIF)
The protein encoded by lpg1965 is translocated into
expressed. RT-PCR analysis was performed using RNA isolated
from broth-grown L. pneumophila. Reactions were carried out using
primers specific for the genes pieA, pieB, pieC, pieD, pieE, pieF and
pieG. PCRs in which no reverse transcriptase was added during the
first-strand synthesis step were conducted to control for residual
DNA (no RT).
Found at: doi:10.1371/journal.ppat.1000278.s003 (2.37 MB TIF)
The pie genes encoded in the Pie genomic region are
the IcmS-IcmW complex. Translocation was assayed as described
earlier in cells infected with wild-type L. pneumophila or with a
double mutant CR503 lacking both icmS and icmW. Translocation
efficiencies were calculated by dividing cAMP levels measured for
the mutant strain by the cAMP levels measured in a parallel
infection using wild-type L. pneumophila producing the indicated
Cya fusion protein and multiplying by 100 to give percent
translocation (relative to wild-type). All infections were performed
in triplicate with a standard deviation of less than 10% of the
Found at: doi:10.1371/journal.ppat.1000278.s004 (9.32 MB TIF)
Translocation of all Pie proteins is dependent upon
impaired in the absence of the IcmS-IcmW complex. L. pneumophila
vacuoles were isolated from U937 macrophage-like cells infected
with either wild-type bacteria (wt), or a mutant strain lacking the
icmS and icmW genes (icmS icmW). Vacuoles were incubated with
purified PieA(513–699) protein and bound protein was detected
using a polyclonal antibody directed against PieA. L. pneumophila
containing vacuoles were scored to quantify the percent of
vacuoles that bound to PieA. Values are means6standard-error
of mean for three independent experiments.
Found at: doi:10.1371/journal.ppat.1000278.s005 (7.89 MB TIF)
PieA binding to vacuoles containing L. pneumophila is
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org11January 2009 | Volume 5 | Issue 1 | e1000278
We would like to thank Dr. Eva Campodonico for the construction of
plasmid pEMC22. Part of this work was first published at the Experimental
Workshop on Signal Transduction in Host-Bacterial Interactions,
Jerusalem, Israel, October 2007.
Conceived and designed the experiments: SN JC CRR. Performed the
experiments: SN JC. Analyzed the data: SN JC CRR. Contributed
reagents/materials/analysis tools: SN. Wrote the paper: SN CRR.
1. Fraser DW, Tsai TR, Orenstein W, Parkin WE, Beecham HJ, et al. (1977)
Legionnaires’ disease: description of an epidemic of pneumonia. N Engl J Med
2. McDade JE, Shepard CC, Fraser DW, Tsai TR, Redus MA, et al. (1977)
Legionnaires’ disease: isolation of a bacterium and demonstration of its role in
other respiratory disease. N Engl J Med 297: 1197–1203.
3. Fields BS (1996) The molecular ecology of legionellae. Trends Microbiol 4:
4. Fliermans CB, Cherry WB, Orrison LH, Smith SJ, Tison DL, et al. (1981)
Ecological distribution of Legionella pneumophila. Appl Environ Microbiol 41:
5. Atlas RM (1999) Legionella: from environmental habitats to disease pathology,
detection and control. Environmental Microbiology 1: 283–293.
6. Horwitz MA, Silverstein SC (1980) Legionnaires’ disease bacterium (Legionella
pneumophila) multiples intracellularly in human monocytes. J Clin Invest 66:
7. Berger KH, Isberg RR (1993) Two distinct defects in intracellular growth
complemented by a single genetic locus in Legionella pneumophila. Mol Microbiol 7:
8. Marra A, Blander SJ, Horwitz MA, Shuman HA (1992) Identification of a
Legionella pneumophila locus required for intracellular multiplication in human
macrophages. Proc Natl Acad Sci U S A 89: 9607–9611.
9. Segal G, Purcell M, Shuman HA (1998) Host cell killing and bacterial
conjugation require overlapping sets of genes within a 22-kb region of the
Legionella pneumophila genome. Proc Natl Acad Sci U S A 95: 1669–1674.
10. Vogel JP, Andrews HL, Wong SK, Isberg RR (1998) Conjugative transfer by the
virulence system of Legionella pneumophila. Science 279: 873–876.
11. Horwitz MA (1983) The Legionnaires’ disease bacterium (Legionella pneumophila)
inhibits phagosome-lysosome fusion in human monocytes. J Exp Med 158:
12. Horwitz MA (1983) Formation of a novel phagosome by the Legionnaires’
disease bacterium (Legionella pneumophila) in human monocytes. J Exp Med 158:
13. Swanson MS, Isberg RR (1995) Association of Legionella pneumophila with the
macrophage endoplasmic reticulum. Infect Immun 63: 3609–3620.
14. Tilney LG, Harb OS, Connelly PS, Robinson CG, Roy CR (2001) How the
parasitic bacterium Legionella pneumophila modifies its phagosome and transforms
it into rough ER: implications for conversion of plasma membrane to the ER
membrane. J Cell Sci 114: 4637–4650.
15. Kagan JC, Roy CR (2002) Legionella phagosomes intercept vesicular traffic from
endoplasmic reticulum exit sites. Nat Cell Biol 4: 945–954.
16. Kagan JC, Stein MP, Pypaert M, Roy CR (2004) Legionella subvert the functions
of rab1 and sec22b to create a replicative organelle. J Exp Med 199: 1201–1211.
17. Chien M, Morozova I, Shi S, Sheng H, Chen J, et al. (2004) The Genomic
Sequence of the Accidental Pathogen Legionella pneumophila. Science 305:
18. Cazalet C, Rusniok C, Bruggemann H, Zidane N, Magnier A, et al. (2004)
Evidence in the Legionella pneumophila genome for exploitation of host cell
functions and high genome plasticity. Nat Genet 36: 1165–1173.
19. Nagai H, Kagan JC, Zhu X, Kahn RA, Roy CR (2002) A bacterial guanine
nucleotide exchange factor activates ARF on Legionella phagosomes. Science 295:
20. de Felipe KS, Pampou S, Jovanovic OS, Pericone CD, Ye SF, et al. (2005)
Evidence for Acquisition of Legionella Type IV Secretion Substrates via
Interdomain Horizontal Gene Transfer. J Bacteriol 187: 7716–7726.
21. Campodonico EM, Chesnel L, Roy CR (2005) A yeast genetic system for the
identification and characterization of substrate proteins transferred into host cells
by the Legionella pneumophila Dot/Icm system. Mol Microbiol 56: 918–933.
22. Conover GM, Derre I, Vogel JP, Isberg RR (2003) The Legionella pneumophila
LidA protein: a translocated substrate of the Dot/Icm system associated with
maintenance of bacterial integrity. Mol Microbiol 48: 305–321.
23. Luo ZQ, Isberg RR (2004) Multiple substrates of the Legionella pneumophila Dot/
Icm system identified by interbacterial protein transfer. Proc Natl Acad Sci U S A
24. Machner MP, Isberg RR (2006) Targeting of host Rab GTPase function by the
intravacuolar pathogen Legionella pneumophila. Dev Cell 11: 47–56.
25. Murata T, Delprato A, Ingmundson A, Toomre DK, Lambright DG, et al.
(2006) The Legionella pneumophila effector protein DrrA is a Rab1 guanine
nucleotide-exchange factor. Nat Cell Biol 8: 971–977.
26. Ninio S, Zuckman-Cholon DM, Cambronne ED, Roy CR (2005) The Legionella
IcmS-IcmW protein complex is important for Dot/Icm-mediated protein
translocation. Mol Microbiol 55: 912–926.
27. Shohdy N, Efe JA, Emr SD, Shuman HA (2005) Pathogen effector protein
screening in yeast identifies Legionella factors that interfere with membrane
trafficking. Proc Natl Acad Sci U S A 102: 4866–4871.
28. Altman E, Segal G (2008) The Response Regulator CpxR Directly Regulates
Expression of Several Legionella pneumophila icm/dot Components as Well as New
Translocated Substrates. J Bacteriol 190: 1985–1996.
29. Zusman T, Aloni G, Halperin E, Kotzer H, Degtyar E, et al. (2007) The
response regulator PmrA is a major regulator of the icm/dot type IV secretion
system in Legionella pneumophila and Coxiella burnetii. Molecular Microbiology 63:
30. Ninio S, Roy CR (2007) Effector proteins translocated by Legionella pneumophila:
strength in numbers. Trends in Microbiology 15: 372–380.
31. Sory MP, Cornelis GR (1994) Translocation of a hybrid YopE-adenylate cyclase
from Yersinia enterocolitica into HeLa cells. Mol Microbiol 14: 583–594.
32. Chen J, de Felipe KS, Clarke M, Lu H, Anderson OR, et al. (2004) Legionella
effectors that promote nonlytic release from protozoa. Science 303: 1358–1361.
33. Nagai H, Cambronne ED, Kagan JC, Amor JC, Kahn RA, et al. (2005) A C-
terminal translocation signal required for Dot/Icm-dependent delivery of the
Legionella RalF protein to host cells. Proc Natl Acad Sci U S A 102: 826–831.
34. Hacker J, Carniel E (2001) Ecological fitness, genomic islands and bacterial
pathogenicity. A Darwinian view of the evolution of microbes. EMBO Reports
35. Karlin S (2001) Detecting anomalous gene clusters and pathogenicity islands in
diverse bacterial genomes. Trends in Microbiology 9: 335–343.
36. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the
sensitivity of progressive multiple sequence alignment through sequence
weighting, position-specific gap penalties and weight matrix choice. Nucleic
Acids Res 22: 4673–4680.
37. Rose A, Schraegle S, Stahlberg E, Meier I (2005) Coiled-coil protein
composition of 22 proteomes - differences and common themes in subcellular
infrastructure and traffic control. BMC Evolutionary Biology 5: 66.
38. Moore JD (2001) The Ran-GTPase and cell-cycle control. Bioessays 23: 77–85.
39. Derre I, Isberg RR (2005) LidA, a translocated substrate of the Legionella
pneumophila type IV secretion system, interferes with the early secretory pathway.
Infect Immun 73: 4370–4380.
40. Celli J, de Chastellier C, Franchini DM, Pizarro-Cerda J, Moreno E, et al. (2003)
Brucella evades macrophage killing via VirB-dependent sustained interactions
with the endoplasmic reticulum. J Exp Med 198: 545–556.
41. Cambronne ED, Roy CR (2007) The Legionella pneumophila IcmSW complex
interacts with multiple Dot/Icm effectors to facilitate type IV translocation.
PLoS Pathog 3(12): e188. doi:10.1371/journal.ppat.0030188.
42. Hilbi H, Segal G, Shuman HA (2001) Icm/dot-dependent upregulation of
phagocytosis by Legionella pneumophila. Mol Microbiol 42: 603–617.
43. Berger KH, Merriam JJ, Isberg RR (1994) Altered intracellular targeting
properties associated with mutations in the Legionella pneumophila dotA gene. Mol
Microbiol 14: 809–822.
44. Roy CR, Berger KH, Isberg RR (1998) Legionella pneumophila DotA protein is
required for early phagosome trafficking decisions that occur within minutes of
bacterial uptake. Mol Microbiol 28: 663–674.
45. Swanson MS, Isberg RR (1996) Identification of Legionella pneumophila mutants
that have aberrant intracellular fates. Infect Immun 64: 2585–2594.
46. Dorer MS, Kirton D, Bader JS, Isberg RR (2006) RNA interference analysis of
Legionella in Drosophila cells: exploitation of early secretory apparatus dynamics.
PLoS Pathog 2(4): e34. doi:10.1371/journal.ppat.0020034.
47. Losick VP, Isberg RR (2006) NF-kappaB translocation prevents host cell death
after low-dose challenge by Legionella pneumophila. J Exp Med 203: 2177–2189.
48. Belyi Y, Niggeweg R, Opitz B, Vogelsgesang M, Hippenstiel S, et al. (2006)
Legionella pneumophila glucosyltransferase inhibits host elongation factor 1A.
Proceedings of the National Academy of Sciences 103: 16953–16958.
49. Weber SS, Ragaz C, Reus K, Nyfeler Y, Hilbi H (2006) Legionella pneumophila
exploits PI(4)P to anchor secreted effector proteins to the replicative vacuole.
PLoS Pathog 2(5): e46. doi:10.1371/journal.ppat.0020046.
50. Feeley JC, Gibson RJ, Gorman GW, Langford NC, Rasheed JK, et al. (1979)
Charcoal-yeast extract agar: primary isolation medium for Legionella pneumophila.
J Clin Microbiol 10: 437–441.
51. Roy CR, Isberg RR (1997) Topology of Legionella pneumophila DotA: an inner
membrane protein required for replication in macrophages. Infect Immun 65:
52. Celada A, Gray PW, Rinderknecht E, Schreiber RD (1984) Evidence for a
gamma-interferon receptor that regulates macrophage tumoricidal activity. J Exp
Med 160: 55–74.
53. Merriam JJ, Mathur R, Maxfield-Boumil R, Isberg RR (1997) Analysis of the
Legionella pneumophila fliI gene: intracellular growth of a defined mutant defective
for flagellum biosynthesis. Infect Immun 65: 2497–2501.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org 12January 2009 | Volume 5 | Issue 1 | e1000278
54. Joiner KA, Fuhrman SA, Miettinen HM, Kasper LH, Mellman I (1990) Download full-text
Toxoplasma gondii: fusion competence of parasitophorous vacuoles in Fc receptor-
transfected fibroblasts. Science 249: 641–646.
55. Coers J, Kagan JC, Matthews M, Nagai H, Zuckman DM, et al. (2000)
Identification of Icm protein complexes that play distinct roles in the biogenesis
of an organelle permissive for Legionella pneumophila intracellular growth. Mol
Microbiol. pp 719–736.
56. Starr T, Ng TW, Wehrly TD, Knodler LA, Celli J (2008) Brucella intracellular
replication requires trafficking through the late endosomal/lysosomal compart-
ment. Traffic 9: 678–694.
57. Letunic I, Copley RR, Pils B, Pinkert S, Schultz J, et al. (2006) SMART 5:
domains in the context of genomes and networks. Nucleic Acids Res 34:
58. Marchler-Bauer A, Bryant SH (2004) CD-Search: protein domain annotations
on the fly. Nucl Acids Res 32: W327–331.
PieA Recruitment to Vacuoles
PLoS Pathogens | www.plospathogens.org 13 January 2009 | Volume 5 | Issue 1 | e1000278