Rickettsiae and Rickettsial Infections: The Current State of Knowledge
+ Author Affiliations
- Reprints or correspondence: Dr. David H. Walker, Dept. of Pathology, Center for Biodefense and Emerging Infectious Diseases, 301 University Blvd., Galveston, TX 77555-0609 (dwalker@utmb.edu).
Abstract
New human rickettsial pathogens have been
discovered, and long-known rickettsiae of undetermined pathogenicity
have been demonstrated
to cause illness. Disease associated with Rickettsia slovaca
has unique clinical manifestations, including prominent lymphadenopathy
without fever and rash. Rickettsial genomes are highly
conserved, with reductive evolution leading to a
small genome that relies on the host cell for many biosynthetic
functions.
Advances in the evaluation of the pathogenesis of
rickettsial disease include identification of rickettsial adhesins, a
host
cell receptor, signaling elements associated with
entry of rickettsiae by induced phagocytosis, rickettsial enzymes
mediating
phagosomal escape, and host actin-based rickettsial
cell-to-cell spread. Disruption of adherens junctions of infected
endothelial
cells likely plays a role in the critical
pathophysiologic mechanism: increased microvascular permeability.
Production of
reactive oxygen species by infected endothelium
injures these cells. However, disseminated intravascular coagulation
rarely
occurs. Immunity is mediated by reactive
cytokine-activated rickettsicidal nitrogen and oxygen species and by
clearance of
rickettsiae by cytotoxic CD8 T cells.
Taxonomic Conundrum
During the past 2 decades, many novel Rickettsia isolates have been characterized by recently developed methods for genetic analysis. A large number of these Rickettsia species are agents of human diseases in areas of the world where rickettsioses had not previously been investigated in depth
(e.g., R. japonica in Japan and Korea; R. honei in Australia and Southeast Asia; R. africae throughout sub-Saharan Africa and in the French West Indies; R. felis globally; the R. sibirica mongolotimonae strain in Asia, Europe, and Africa; R. parkeri in North and South America; "R. heilongjiangensis" in northeastern Asia; and R. aeschlimannii in Africa) [1,2,3,4,5,6–7]. A novel isolate (Astrakhan strain) of Rickettsia conorii has been recovered from patients in Astrakhan, Russia, and in Chad and from ticks in Kosovo. R. conorii strain Israeli has been isolated not only from patients in Israel but also from those in Portugal.
Interestingly, this proliferation of named species has generated controversy among rickettsiologists regarding the appropriate
taxonomy of Rickettsia species. One proposed set of criteria for establishing rickettsial species is based on the premise that all previously named
species are substantially different from one another [8]. However, in a recent publication in Nature Reviews Microbiology
discussing reevaluation of prokaryotic species, a group of taxonomic
experts rejected this approach, stating that “Defining
species limits by using levels of sequence
similarity typically found within existing named species is clearly
inappropriate”
[9, p. 735]. For rickettsiae, a proposed criterion for the establishment of a new species (i.e., a divergence of the rrs gene of 0.2%) differs markedly from the threshold of genetic divergence (i.e., a divergence of the rrs
gene of 0.5%) of variants of other bacteria with similar evolutionary
selective pressure in an obligately intracellular lifestyle
involving an arthropod host (i.e., Orientia tsutsugamushi, Coxiella burnetii, and Ehrlichia chaffeensis). Of note, such organisms as R. conorii, R. sibirica, R. africae, and R. parkeri
could be considered to be strains or subspecies if they were newly
identified in other genera. The inertia of history and
habit would justify preserving the current taxonomy
and establishing a moratorium against using the proposed criteria to
establish
scores of new species.
The motivation driving the “genomic
splitters” appears to be based on the gratifying experience of baptizing
more organisms
with creative names rather than using taxonomic
criteria consistent with other similar organisms toward the aim of
scientific
utility. Any criteria for determining prokaryotic
species are arbitrary, and there is near universal agreement that a
taxonomic
scheme is justified only if it serves a useful
scientific purpose. Some tentative criteria relate to putative, unique
clinical
manifestations and epidemiologic factors. Actually,
the clinical manifestations of most rickettsioses constitute a
continuous
spectrum (e.g., case-fatality rates and the
percentage of patients with eschars). Indeed, even when the proportion
of patients
with an eschar is considered to be a clinical
characteristic, the incidence varies for patients infected with the same
strain
in different geographic areas (e.g., R. conorii
strain Israeli in Portugal, where eschars are detected frequently, and
that in Israel, where eschars are rarely identified).
The latter example suggests that neither geography
nor inconsistent clinical manifestations define the taxonomy of the
microbial
agent. The R. sibirica mongolotimonae
strain, which has been found on 3 continents, is another example. The
name “lymphangitis-associated rickettsiosis”
has been proposed for the disease caused by this
strain, despite the fact that lymphangitis is observed in only 40% of
cases
and has been described in patients infected with
other Rickettsia species [1].
In general, the spotted fever rickettsiosis syndromes are similar. Those with distinctive features include infections caused
by R. slovaca and R. africae. Tickborne lymphadenopathy (TIBOLA) and Dermacentor-borne necrosis-eschar-lymphadenopathy (DEBONEL) are descriptive disease names for R. slovaca infection [3], a disease that occurs in Europe during the winter. Dermacentor
ticks most frequently attach to the occipital scalp, where, 7–9 days
later, an eschar appears in association with painful
cervical lymphadenopathy. Fever and rash are seldom
present. Alopecia at the eschar site in 24% of cases and persistent
asthenia
may occur even after antirickettsial treatment.
This syndrome is novel among rickettsial diseases.
Similarly, African tick-bite fever is a highly prevalent and, apparently, non–life-threatening rickettsiosis [2, 7].
It differs from most other rickettsioses of mild-to-moderate severity
in that it produces painful regional lymphadenopathy,
multiple eschars, nuchal myalgia, and, on occasion,
a sparse and sometimes vesicular rash. Predictably, the illness
associated
with the closely related agent R. parkeri is similar [4]. R. parkeri and R. slovaca are examples of rickettsiae that, for decades, were presumed to be nonpathogenic or of undetermined pathogenicity and are
now known to cause illness in humans.
The possibility that 1 or more other strains of Rickettsia
identified only in an arthropod host might be pathogenic in humans
remains to be proven. In the United States, few infectious
diseases specialists have focused investigative
attention on rickettsial diseases. Isolation of rickettsiae from
patients
and prospective clinical studies are rarely
undertaken. Diseases caused by such pathogens as the highly prevalent R. felis, which was shown to cause illness in humans more than a decade ago, and R. parkeri, which was discovered >60 years ago, remain virtually uninvestigated.
New developments in the epidemiologic profile of rickettsioses include the emergence of spotted fever rickettsiosis due to
R. rickettsii in Argentina; the reemergence of spotted fever rickettsiosis due to R. rickettsii in Brazil, Colombia, and the United States; identification of other Rickettsia species in ticks in South America and Mexico; an outbreak of Rocky Mountain spotted fever in Arizona, transmitted for the
first time in the United States by Rhipicephalus sanguineus ticks; outbreaks of louseborne typhus in Burundi, Russia, Peru, and Algeria; and the reemergence of scrub typhus [10,11,12,13–14]. A National Institutes of Health map of emerging infectious diseases that lacked any mentions of rickettsioses led D.H.W.
to create a map containing only rickettsiae (figure 1). Reports of some rickettsiae (e.g., R. helvetica) as human pathogens are based on inconclusive data.
Contributions Of Genome Sequencing To Understanding Rickettsiae
Rickettsial genomes are highly conserved, with similar gene synteny and content [15].
Their small genomes have resulted from gene decay, with there being
many pseudogenes and a high proportion of noncoding
DNA. Their cytosolic niche, which is rich in
nutrients, amino acids, and nucleotides, has allowed rickettsiae to
jettison
the genes encoding enzymes for sugar metabolism and
for lipid, nucleotide, and amino acid synthesis, a characteristic most
likely responsible for our inability to cultivate
them in cell-free medium. Rickettsia species also contain as
many as 5 autotransporters, proteins with 3 domains, a leader sequence
that mediates transport across
the cell membrane, a passenger sequence, and a
transporter sequence that is inserted as a β-barrel into the outer
envelope
to transport the passenger sequence to the outer
surface of the cell wall. Among the autotransporters, outer membrane
protein
(Omp) A is present only in rickettsiae in the
spotted fever group, and OmpB is present in all Rickettsia species. Sca 1, Sca 2, and Sca 3 exist as split genes (interrupted into 2–4 open reading frames) in at least 1 Rickettsia species. Sca 4, which shares sequence similarity, is not an autotransporter, because it lacks the transporter domain.
Rickettsia-Host Cell Interactions
Obligately intracellular rickettsiae attach to the host cell receptor Ku70 by means of the most abundant surface protein,
OmpB (figure 2) [16].
Spotted fever group rickettsiae also use OmpA as an adhesin. Adhesion
of OmpB to the membrane-spanning protein Ku70 results
in recruitment of additional Ku70 molecules to the
cell membrane, where further OmpB binding occurs. Ubiquitin ligase is
also
recruited to the future rickettsial entry site
where Ku70 is ubiquitinated, and signal transduction events lead to
recruitment
of Arp2/3 complex. Cdc42 (a small guanidine
triphosphatase), protein tyrosine kinase, phosphoinositide 3-kinase, and
Src-family
kinases activate Arp2/3, resulting in phagocytosis
of the attached rickettsia occurring as a result of a zipper mechanism
involving alteration of cytoskeletal actin at the
entry site [17]. Another rickettsial protein—RickA, expressed on the rickettsial surface—activates Arp2/3, which initiates polymerization
of host cell actin [18, 19].
The filaments of actin push the rickettsia to the surface of the host
cell, where the host cell membrane is deformed outward
and invaginates into the adjacent cell. Disruption
of both cell membranes enables the rickettsia to enter the adjoining
cell
without being exposed to the extracellular
environment. Some rickettsiae exit via the luminal surface of blood
vessels into
the bloodstream. Typhus rickettsiae do not
stimulate actin-based mobility, and they accumulate to massive
quantities intracellularly
until the endothelial cell bursts, releasing
rickettsiae into the blood.
To enter the cytosol of the host cell where nutrients, adenosine triphosphate, amino acids, and nucleotides are available
for growth and to avoid phagolysosomal fusion and death, rickettsiae must escape from the phagosome [20]. Rickettsiae secrete phospholipase D and hemolysin C, which disrupt the phagosomal membrane and permit the rapid escape
of the rickettsiae.
Rickettsial Pathogenic Mechanisms
The major pathophysiologic effect of
rickettsial infections is increased microvascular permeability due to
the disruption
of adherens junctions between infected endothelial
cells, development of interendothelial gaps, formation of stress fibers,
and conversion of the shape of endothelial cells
from polygons to large spindles [21].
The current hypothesis regarding the mechanism of injury of
rickettsia-infected endothelial cells concerns oxidative stress.
Endothelial cells infected with spotted fever group
rickettsiae in vitro produce reactive oxygen species that cause lipid
peroxidative damage to the host cell membranes.
There is evidence that rickettsial infection causes oxidative stress in
experimentally
infected animals [22],
but the extent to which this mechanism explains the pathologic findings
of spotted fever group rickettsioses and the possibility
of other pathogenic mechanisms, such as the effects
of cytokines, other mediators, or cytotoxic T cells, remains
undetermined.
Rickettsial infection of endothelial cells activates nuclear factor κB, which inhibits apoptosis and mediates the production
of proinflammatory cytokines [23].
Rickettsia-infected endothelium produces IL-6, IL-8, and monocyte
chemoattractant protein 1. Delay of endothelial cell
death allows further intracellular rickettsial
growth. A promising avenue for discovering novel general principles of
pathobiology
and immunobiology is in vitro investigation of the
effects of rickettsial infection on primary endothelial cells from
relevant
organs in the presence of cellular components of
blood and physiologic shear forces of flow and with appropriate in vivo
models.
Reports of cases of rickettsioses often
claim, without supporting evidence other than thrombocytopenia, that the
patient had
disseminated intravascular coagulation. Case series
based on such reports submitted to public health agencies and other
reports
based on retrospective reviews of medical records
lacking sufficient coagulation studies have perpetuated the myth that
disseminated
intravascular coagulation occurs in a substantial
portion of patients with severe rickettsiosis. Careful analysis of such
data reveals that disseminated intravascular
coagulation occurs only rarely in persons with rickettsial infections [24].
Patients with Rocky Mountain spotted fever and Mediterranean spotted
fever develop a procoagulant state resulting from
rickettsia-induced disseminated endothelial injury,
release of procoagulant factors, and activation of the coagulation
cascade
with generation of thrombin, platelet activation,
increased fibrinolytic factors, and consumption of natural
anticoagulants.
The result is a remarkably homeostatic state,
considering the extent of the endothelial injury: life-threatening
hemorrhages
and vaso-occlusive thrombi leading to infarcts do
not occur often, even in fatal rickettsioses. Even though hemorrhagic
and
thrombotic complications are not common, the
presence of thrombin might play a role in increased vascular
permeability, as
has been demonstrated by in vitro experiments using
endothelial cell monolayers.
Serial investigations of the coagulation
system in a mouse model of lethal spotted fever rickettsiosis revealed
severe multifocal
endothelial injury, increased generation of
thrombin, decreased levels of factor VII, increased factor V
procoagulant activity,
decreased prekallikrein and tissue plasminogen
activator activity, and increased activity of plasminogen activator
inhibitor
without disseminated intravascular coagulation [25].
Mice with sublethal infection manifested an acute-phase reaction (e.g.,
increased plasma fibrinogen) and release of endothelial
cell contents (e.g., von Willebrand factor), but
without apparent activation of the coagulation system in the face of
marked
inhibition of fibrinolysis driven by endothelial
cells. Neither lethal nor sublethal infection resulted in uncontrolled
pathologic
thrombosis or consumption of clotting factors.
Mechanisms Of Immunity To Rickettsiae
Among the most interesting aspects of the
pathogenesis of rickettsial infections are the host defenses. Studies
of excellent
murine models of spotted fever and typhus group
rickettsioses have identified novel mechanisms of immunity, including
cytokine-mediated
activation of endothelial cell bactericidal control
of intracellular infection and the role of autophagy in rickettsial
killing.
Murine endothelial cells activated by IFN-γ and
TNF-α produce rickettsicidal nitric oxide via inducible nitric oxide
synthetase
[26].
Early in rickettsial infections, natural killer cells are activated and
inhibit growth of rickettsiae in association with
production of IFN-γ. Clearance of rickettsiae
requires cytotoxic CD8 T cells, which eliminate infected endothelial
cells by
inducing apoptosis that is dependent, at least in
part, on a perforin-mediated mechanism. Antibodies against rickettsial
OmpA
and OmpB, but not rickettsial lipopolysaccharide,
are protective against reinfection [27, 28].
However, antibodies to these proteins do not appear until after control
of the rickettsial infection and recovery from
the disease have occurred. Thus, antibodies do not
play an important role in immunity during the first exposure to a
pathogenic
rickettsia.
Both immune CD4 and CD8 T cells
contribute to protective immunity, and homing of lymphocytes and
macrophages to foci of infection
in the microcirculation is associated with
clearance of rickettsiae [26].
Expression of the chemokines CXCL9 (Mig), CXCL19 (IP-10), and CX3CL1
(fractalkine) in rickettsia-infected murine endothelial
cells and of CXCL9 and CXCL10 in cerebral
endothelial cells from patients with Rocky Mountain spotted fever has
not been mechanistically
linked to immune cell chemotaxis to
rickettsia-infected endothelium [29].
Expression of intercellular adhesion molecule–1 and vascular
endothelial cell adhesion molecule–1 and of rickettsial antigens
on infected endothelial cells could contribute to
perivascular migration of T cells. Perivascular CD4 and CD8 T cells,
macrophages,
and dendritic cells are presumed to be the sources
of the cytokines that activate endothelial rickettsicidal activities.
Human endothelial cells activated by
IFN-γ, TNF-α, IL-1β, and RANTES (regulated on activation, normally T
cell expressed and
secreted) kill intracellular rickettsiae through 2
bactericidal mechanisms, nitric oxide production and hydrogen peroxide
production [26].
Human macrophages, a minor target of rickettsial infections, kill
intracellular rickettsiae after activation by IFN-γ,
TNF-α, and IL-1β via production of hydrogen
peroxide and tryptophan starvation of rickettsiae associated with
degradation
of tryptophan by indoleamine-2,3-dioxygenase. The
relevance of these bactericidal mechanisms to human rickettsioses is
supported
by the expression of inducible nitric oxide
synthetase, IFN-γ, and indoleamine-2,3-dioxygenase in the skin lesions
of patients
with Mediterranean spotted fever.
The balance between the susceptibility of
the host to pathogenic mechanisms and resistance to rickettsial growth
is expressed
as host risk factors for severity of illness. For
humans, these factors include older age, glucose-6-phosphate
dehydrogenase
deficiency, treatment with a sulfonamide, diabetes
mellitus, and male sex. Experimental infections have confirmed that
older
male animals are killed by lower median doses of
rickettsiae. The undetermined importance of immune responses and of
oxidative
stress in human rickettsioses emphasizes the
unfortunate neglect of these life-threatening infections by infectious
diseases
physician-scientists.
The statements of Ted Woodward in his Maxwell Finland Lecture of 1972 regarding the need for scientists and physicians to
enter the field of rickettsiology played a role in my choice of rickettsial diseases as a career focus [30]. Dr. Woodward's.tift to me of the 1922 classic The Etiology and Pathology of Typhus, written by Wolbach, Todd, and Palfrey and containing the signature of S. Burt Wolbach, is the most cherished volume in my
library [31]. The studies of rickettsioses advanced by both Wolbach and Woodward remain unfinished.