|Home | About | Journals | Submit | Contact Us | Français|
Many species of Rickettsia are well-known mammalian pathogens transmitted by blood-feeding arthropods. However, molecular surveys are continually uncovering novel Rickettsia species, often in unexpected hosts, including many arthropods that do not feed on blood. This study reports a systematic molecular characterization of a Rickettsia infecting the psocid Liposcelis bostrychophila (Psocoptera: Liposcelidae), a common and cosmopolitan household pest. Surprisingly, the psocid Rickettsia is shown to be Rickettsia felis, a human pathogen transmitted by fleas that causes serious morbidity and occasional mortality. The plasmid from the psocid R. felis was sequenced and was found to be virtually identical to the one in R. felis from fleas. As Liposcelis insects are often intimately associated with humans and other vertebrates, it is speculated that they acquired R. felis from fleas. Whether the R. felis in psocids causes disease in vertebrates is not known and warrants further study.
Many species of Rickettsia are well-known mammalian pathogens that are transmitted by blood-feeding arthropods via bites or feces and can cause mild to fatal diseases in humans (33). Some species are also considered potential bioterrorism agents (4). Most Rickettsia research has focused on pathogens that are found in two closely related species groups, the typhus and spotted fever groups, such as Rickettsia prowazekii, Rickettsia rickettsii, and Rickettsia typhi, the causal agents of epidemic typhus, Rocky Mountain spotted fever, and murine typhus, respectively (3, 4, 33). However, recent surveys suggest that Rickettsia bacteria are much more widespread than previously suspected and that they are being detected in novel hosts, the vast majority of which are arthropods, including many that do not feed on blood (29, 45).
The number of new rickettsial species that cause diseases in humans is rapidly increasing (33). One such species that has been generating much interest in recent years is Rickettsia felis, the causative agent of a murine typhus-like disease (1, 2, 13, 16, 17, 28, 44). The disease is often unrecognized, and even though it is considered clinically mild, it can cause severe illness and death in older patients and in cases of delayed diagnosis (2). R. felis was identified only in 1990 (1) and has since been found worldwide in fleas, where it is maintained transovarially and can reach high infection rates (e.g., 86% to 94% in cat fleas) (2, 3, 44), as well as in ticks and mites (34). While experimental infections have confirmed that R. felis is transmitted to vertebrate hosts via blood feeding and that R. felis occurs in an infectious extracellular state (39), it is not known whether transmission can also occur through contamination of broken skin by infected vector feces, as in R. typhi (3, 34).
A number of features distinguish R. felis from species in both the typhus and spotted fever groups. Lately, it has been proposed that R. felis be in its own group, allied with Rickettsia akari and Rickettsia australis, the causal agents of rickettsial pox and Queensland tick typhus, respectively, and a number of recently discovered strains infecting insects that do not feed on blood (16, 17, 29, 45). Moreover, R. felis was the first Rickettsia species shown to have a plasmid (28). While plasmids now appear to be quite widespread in the genus, the R. felis plasmid stands out with respect to its relatively large size and distinctive gene content (5, 6, 9, 14, 17).
This study reports that a common and cosmopolitan insect, the psocid Liposcelis bostrychophila (Psocoptera: Liposcelidae) harbors R. felis. Liposcelids are the closest free-living relatives of parasitic lice (19) and are well-known for their close proximity to humans, particularly as pests in houses and grain storage facilities (8, 41). Through 16S rRNA gene sequencing, L. bostrychophila was recently shown to harbor a strain of Rickettsia (29, 30, 42). A systematic molecular characterization of this Rickettsia was conducted, demonstrating that it is authentic R. felis. Furthermore, the psocid symbiont plasmid was sequenced and was shown to be virtually identical to the plasmid from R. felis that infects cat fleas.
L. bostrychophila insects were collected by Manoj Nayak (Queensland Department of Primary Industries) in the Darling Downs region of Queensland, Australia, in 1997. The culture was maintained in confined vessels on a culture medium consisting of whole wheat and rice at 27 ± 1°C and 70% ± 2% relative humidity (RH). Sequences from an L. bostrychophila culture collected by George Opit (Entomology and Plant Pathology, Oklahoma State University) from a grain elevator in Manhattan, KS, in 2006 and from an L. bostrychophila culture collected on organic rice by Alex Wilson from Tucson, AZ, in 2004 were also obtained in order to confirm the presence of R. felis in insects from different locales.
Insects were dipped in a soap solution, quickly surface disinfected in 70% ethanol, and washed in sterile phosphate-buffered saline (PBS) and processed whole. Total DNA was extracted from pooled adult L. bostrychophila insects (n = 20) using a DNeasy Purification Kit (Qiagen, Hilden, Germany), with modifications. Samples were first incubated with enzymatic lysis buffer for 2 h at 37°C. Samples were then frozen in liquid nitrogen for 5 min and heated to 65°C for 10 min, and this freeze-thaw cycle was performed three times. The remaining steps were as described in the manufacturer's protocol for Gram-positive bacteria. Samples were processed independently to remove the possibility of contamination.
16S rRNA, gltA, ompA, ompB, and cell surface antigen sca4 (gene D) were amplified and sequenced using previously described primers and methods (see Table S1 in the supplemental material). All PCR amplifications were performed on a Mastercycler Gradient (Eppendorff, Hamburg, Germany) with 1 μM each primer in 3 mM MgCl2, 20 μM each deoxyribonucleoside triphosphate, and 1.25 units of Taq polymerase in a total volume of 25 μl of 1× reaction buffer (Invitrogen, Carlsbad, CA). PCR products were separated in a 1% (wt/vol) agarose gel in TBE buffer (45 mM Tris-borate, 1 mM EDTA) and stained with ethidium bromide (0.1 mg/ml). Negative controls (no DNA added) were always performed in parallel. No products were obtained from these controls. All PCR products were sequenced in both directions at the Macrogen, Inc., Sequencing Center (Seoul, South Korea). Sequence chromatograms were visually inspected, verified, aligned, and annotated using Geneious Pro (Biomatters).
In order to assess the phylogenetic relationships of the psocid bacterium, the 16S rRNA, gltA, ompA, ompB, and cell surface antigen sca4 (gene D) sequences obtained from our analyses were compared with the appropriate sequences from all validated Rickettsia spp. (12, 18, 22, 35-37). Phylogenetic analysis was performed using maximum likelihood, implemented in PhyML (11, 20). Modeltest, version 3.7 (32), was used to determine the most appropriate model of nucleotide substitution. Each gene was analyzed separately, and analyses were also performed on a concatenated data set of the slowly evolving genes gltA and 16S rRNA and a concatenated data set of ompA, ompB, and gene D. To assess confidence in nodes, 100 bootstrap replicates were performed. Maximum-parsimony analyses implemented in PAUP*, version 4.0 (40), yielded similar topology.
To determine the possible presence of a plasmid, the diagnostic primer sets pRFc-pRFd and pRFh-pRFi were initially used (28). To rule out the possibility that the fragments amplified by these primers were chromosomal and to confirm the presence of a plasmid, plasmid sequences from the completed R. felis URRWXCal2 genome were used to design primers (described in Table S2 in the supplemental material) that would amplify large overlapping fragments that would close the circle of the plasmid. Large fragments were amplified using an Expand Long Template PCR system (Roche, Mannheim, Germany) according to the manufacturer's protocol. Once the presence of a plasmid via long PCR was confirmed, primers were designed in order to sequence the entire plasmid genome (also described in Table S2). All new primers in Table S2 were designed using Primer3 (http://biotools.umassmed.edu/bioapps/primer3_www.cgi.). The plasmid from R. felis infecting fleas has been previously reported to occur in up to two forms that may differ in stability (13, 28). The present study focused on the large form.
PCR products were cloned using a StrataClone PCR cloning kit (Stratagene, San Diego, CA) according to manufacturer's instructions. Clones from selected colonies were purified with a QIAprep Spin Miniprep Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Clones were then sequenced in both directions (Macrogen, Seoul, South Korea). The plasmid was annotated using NCBI ORF Finder (open reading frame finder) (http://www.ncbi.nlm.nih.gov/projects/gorf/) and the CGView server (http://stothard.afns.ualberta.ca/cgview_server/index.html).
The nucleotide sequences of the 16S rRNA, gltA, ompA, ompB, and gene D genes were deposited in the GenBank database under accession numbers GQ329872 to GQ329880 and GQ385243. The nucleotide sequences of the plasmid diagnostic genes were deposited in the GenBank database under accession numbers GQ385244 to GQ385247. The pRF plasmid sequence was deposited in the GenBank database under accession number GQ329881.
Sequencing and phylogenetic analysis of five genes from L. bostrychophila originally collected in Australia revealed that the psocid Rickettsia is authentic R. felis (Fig. (Fig.1).1). Citrate synthase (gltA) (382 bp) and cell surface antigen sca4 (gene D) (2,013 bp) sequences were identical to those of R. felis URRWXCal2. 16S rRNA (1,205 bp), outer membrane protein rOmpA (ompA) (1,705 bp), and outer membrane protein rOmpB (ompB) (4,893 bp) sequences exhibited three, one, and five nucleotide differences, respectively, compared to R. felis URRWXCal2. 16S rRNA (1,205 bp) and rOmpB (ompB) (750 bp) sequences obtained from L. bostrychophila from the Kansas culture and 16S rRNA (1,202 bp) and rOmpB (ompB) (5,180 bp) sequences obtained from L. bostrychophila from the Arizona culture were identical to those obtained from the Australia culture, thus confirming that R. felis is present in L. bostrychophila from different origins.
Diagnostic PCR, long PCR, and, finally, sequencing of the complete plasmid show that the psocid Rickettsia harbors a plasmid that is virtually identical to the R. felis URRWXCal2 pRF plasmid. The plasmid sequenced from the psocid Rickettsia (Queensland) was 62,882 nucleotides in length, had a GC content of 33.63%, and was 98% identical in the DNA level to the pRF plasmid from R. felis URRWXCal2. There were no differences in gene order between the plasmids. The few differences in gene content are summarized in Table Table11 and Fig. Fig.2.2. These can be largely attributed to accumulation of nonsense mutations that have caused shortening of 14 predicted genes. Two R. felis URRWXCal2 plasmid genes (pRF09 and pRF39) have been split into two separate genes in the psocid Rickettsia plasmid as a result of a new stop codon. Two R. felis URRWXCal2 plasmid genes (pRF54 and pRF55) have been merged into one in the psocid Rickettsia plasmid, and five genes (pRF22, pRF38, pRF57, pRF58, and pRF60) are absent. Finally, diagnostic PCR and sequencing also identified the plasmid in psocid cultures collected in Kansas and Arizona.
This study identifies a new host for R. felis: L. bostrychophila, a cosmopolitan insect pest. These highly polyphagous insects are infamous for their close associations with human dwellings and food storage facilities (8, 41). Furthermore, there are reports of intimate contact between liposcelids and humans, including feeding on dead skin of human toe nails (23) and infesting scalps (41). In addition, these insects are often found in close associations with vertebrates as they were recovered from rat nests, bird nests, and even on puppy fur (7, 26, 41). They also eat dead insects and insect eggs (41). Thus, it is feasible that during the course of evolution, while sharing the same microhabitat and in close association with humans or other vertebrates, R. felis was horizontally transmitted from fleas to L. bostrychophila.
Previous studies on Rickettsia from Liposcelis suggest that its distribution and transmission may be similar to that in fleas. Rickettsiae are present in the ovaries and oocytes of the insects (30, 42), suggestive of maternal transmission. They also occur within host fat bodies and digestive tracts (10, 30), suggesting that like flea R. felis, the bacteria may be transmitted through host feces. Moreover, within the gut epithelium and gut lumen, the Rickettsia from Liposcelis was shown to be free in the cytosol and surrounded by a “halo” (10, 30), as was described for R. felis in various flea tissues (1, 34) and in tick-derived cell culture (31).
The role of R. felis in L. bostrychophila remains a great mystery. L. bostrychophila reproduces via parthenogenesis (although males were recently discovered in at least one population in Hawaii ); no other parthenogenetic Liposcelis species are known. No L. bostrychophila insects without Rickettsia infection have yet been found; cultures from Europe (United Kingdom and Czech Republic) (30, 42) and Asia (China) are infected (30), just like those from Australia and the United States presented in this study; R. felis has also not yet been found in any sexual Liposcelis species (A. Behar and S. J. Perlman, unpublished data). Furthermore, treating L. bostrychophila with antibiotics results in sterility (42). Thus, it is possible that R. felis is interfering with reproduction (38). In addition to the well-known insect symbionts Wolbachia (36) and Cardinium (47), one other Rickettsia species has been implicated in inducing parthenogenesis in its hosts (21). However, unlike L. bostrychophila, all of the insects thus far implicated in parthenogenesis induction have haplodiploid sex determination, and whether R. felis causes asexuality in L. bostrychophila remains unresolved. It should also be noted that, as yet, no symbionts other than Rickettsia have been conclusively demonstrated in L. bostrychophila. Recent studies reporting Wolbachia (24) and Cardinium (43) infection in L. bostrychophila do not provide sequences, only PCR results, and these results could not be repeated during the course of this study.
Alternatively, the bacterium could be providing its host with an essential function. A recent study proposed that the Rickettsia from Liposcelis is supplementing host nutrition (30) although this is hard to believe for a number of reasons. Rickettsia genomes are highly reduced and lacking most metabolic genes, leaving the bacterium totally dependent on host resources for growth and survival (15, 17). Although it is possible that the strain of R. felis infecting L. bostrychophila has acquired new genes, this seems unlikely, given that plasmid and outer membrane genes are virtually identical to those of the R. felis found in fleas. Moreover, closely related sexual species with the same ecology and nutritional requirements do not harbor Rickettsia (30). Finally, it is also possible that R. felis is maintained via addiction (25), for example, through a toxin-antitoxin system. The fact that L. bostrychophila is unable to reproduce when treated with antibiotics is especially intriguing since R. felis does not appear to induce such an effect in fleas (34).
The discovery of R. felis in the non-blood-feeding L. bostrychophila has a number of implications. This may be a much easier system potentially to study the molecular interactions between Rickettsia species (and R. felis specifically) and their arthropod hosts as the psocid can be cultured easily, grows rapidly, and does not require blood. Furthermore, comparative studies between R. felis from fleas and the psocid R. felis may provide researchers with valuable information on key processes regarding R. felis biology and epidemiology, including mechanisms of host recognition, maintenance, pathogenicity, and virulence.
Finally, future studies will need to determine the potential pathogenicity of the psocid R. felis for vertebrates. Thus far, there is no evidence that L. bostrychophila constitutes a threat to public health. Nevertheless, the results of the current study, combined with the intimate associations between Liposcelis and humans, suggest that this novel R. felis strain warrants further study.
We thank George Opit, Manoj Nayak, Paul Fields, and Alex Wilson for the Liposcelis cultures. Our sincere gratitude goes to John Hay for very helpful discussions and comments on an earlier version of the manuscript.
This work was supported by an NSERC Discovery Grant to S.J.P. and by the Canadian Institute For Advanced Research's Integrated Microbial Biodiversity Program.
Published ahead of print on 5 February 2010.
†Supplemental material for this article may be found at http://aem.asm.org/.