Z O O E C O . O R G

---

Genetic Variants of Ehrlichia phagocytophila in the United States

Robert F. Massung(1)*, Michael J. Mauel(3), Jessica H. Owens(1), Nancy Allan(1), Joshua W. Courtney(1), Kirby C. Stafford III(2), and Thomas N. Mather(3)

1. Division of Viral and Rickettsial Diseases, CDC, Atlanta, Georgia
2. Connecticut Agricultural Experiment Station, New Haven, CT
3. Center for Vector-Borne Disease, University of Rhode Island, Kingston, RI

* Corresponding author. Mailing address: Centers for Disease Control and Prevention, MS G-13, 1600 Clifton Rd., N.E., Atlanta, GA, USA. Phone: (404) 639-1082. Fax: (404) 639-4436. e-mail: rfm2@cdc.gov.
 

ABSTRACT

Primers were used to amplify a 561-bp region of the 16S rRNA gene of Ehrlichia phagocytophila from Ixodes scapularis ticks and small mammals collected in Rhode Island and Connecticut. DNA sequences determined for all 50 positive samples collected from 1996 through 1998 in southwestern Connecticut (Bridgeport area) were identical to the sequence previously reported for Ehrlichia phagocytophila DNA extracted from confirmed human infections. In contrast, the sequences from 92 of 123 positive Rhode Island samples collected from 1996 through 1999 included several variants which differed from that found in the agent infecting humans by 1-2 nucleotides, including a variant previously described in ticks from Rhode Island and white-tailed deer from Maryland and Wisconsin. Whereas 11.9% of 67 E. phagocytophila-positive ticks collected during 1997 in Rhode Island harbored ehrlichiae with sequences identical to the sequence of the human agent (E. phagocytophila-human agent), the remainder (88.1%) had variant sequences including a predominant variant ehrlichia (79.1%) not previously described. Rhode Island chipmunks and white-footed mice also harbored the human agent and 3 variant sequences. The number of questing ticks collected in Rhode Island that were infected with E. phagocytophila was dramatically higher in 1997 (46.4 and 38.3% in nymphs and adults, respectively) compared to 1996 (16.1%), 1998 (5.5%), and 1999 (4.9 and 12.3% in nymphs and adults, respectively). The frequency distribution of genetic variants of E. phagocytophila also differed between these years. Examination of 79 nymphal ticks collected from Bluff Point in southeastern Connecticut, near the Rhode Island border, showed a distribution of E. phagocytophila variants similar to that noted in Rhode Island. A genetic variant of E. phagocytophila was first detected in ticks collected in Bridgeport, Connecticut in 1999 suggesting that the geographic range of this variant may be expanding westward. Although the function and biological significance of these genetic types is unknown, the low incidence of human ehrlichiosis in Rhode Island may be due, in part, to these variant ehrlichiae interfering with maintenance and transmission of the known agent of human disease.
 

Introduction

Members of the genus Ehrlichia are obligate, intracellular bacteria within the order Rickettsiales. Although ehrlichial infections of veterinary importance were first described in 1935, the first case of human ehrlichiosis in the United States was reported in 1987 (1). The human pathogen was subsequently identified as Ehrlichia chaffeensis (2) and the number of human cases now exceed 740 (3). In 1994, a second ehrlichial infection in humans was reported, and was referred to as human granulocytic ehrlichiosis (HGE) due to the proclivity of the agent to infect neutrophils (4). The majority of HGE cases have been diagnosed in northeastern and upper midwestern areas of the Unites States, although a limited number of cases have been reported in Europe and in northern California (5-12).

The close genetic and antigenic relationship of the HGE agent to two previously characterized species, E. phagocytophila, noted for infections of ruminants in Europe, and E. equi, the agent of equine granulocytic ehrlichiosis, has led to the suggestion that these three be reclassified as a single species, with E. phagocytophila as the precedent name. The 16S rRNA gene has been amplified and sequenced from confirmed human infections in both North America and Europe, and all sequences have been identical to the original published sequence for the HGE agent (4, 8), with the exception of two cases recently reported from Northern California that were the same as the E. equi 16S rRNA gene sequence (12). A variant that showed a 2-bp difference compared to the sequence of the HGE agent was reported in white-tailed deer in Maryland and Wisconsin, and Ixodes scapularis ticks collected in Rhode Island (13,14). Likewise, the 16S rRNA sequences determined from documented infections of horses and ruminants by various E. phagocytophila strains have differed from the sequences noted for the HGE agent by several bases. None of these latter-mentioned variant forms have been shown to cause human disease. Another ehrlichia that has been found in nature and is closely related to E. phagocytophila, but apparently does not cause human disease, is the white-tailed deer agent (15). Ehrlichiae closely related to E. phagocytophila recently have been demonstrated in Colorado, an area of the US where human ehrlichiosis is not endemic (16). These data suggest that only a subset of the E. phagocytophila strains that exist in nature may cause human disease. Although the HGE agent is now considered to be a member of the species E. phagocytophila, hereafter we will designate the isolates responsible for human disease as E. phagocytophila-human agent (EP-ha) to differentiate them from the 16S rRNA sequence variants described in the current study.

Rhode Island and Connecticut are adjacent northeastern states. Previous studies have shown these two states to have a similar frequency and distribution of vector Ixodes scapularis ticks, and primary reservoir hosts, white-footed mice and chipmunks (17-19). However, the number of human infections with E. phagocytophila reported to date in CT is dramatically higher than in RI. The current study was undertaken to examine the frequency and distribution of E. phagocytophila, including EP-ha and E. phagocytophila-related variants, in potential reservoir and vector populations in Rhode Island and Connecticut.

Materials and Methods

Tick and Mammal Collections

Questing nymphal and adult black-legged ticks (I. scapularis) were collected from 4 sites in the region surrounding South Kingston, Rhode Island and Bridgeport, Connecticut each year from 1996 to 1999 (Figure 1). Questing nymphal ticks were also collected from Bluff Point in southeastern CT in 1997. The Rhode Island sites are all located in the zone of highest I. scapularis density within the state (17). Small rodents, including white-footed mice (Peromyscus leucopus) and chipmunks (Tamias striatus) were live-trapped at the same locations. Samples of blood were obtained following procedures approved by the institutional animal care and use committee. Blood was stored in EDTA at –80oC until being tested for the presence of ehrliciae by polymerase chain reaction (PCR) techniques and DNA sequencing.
 

Figure 1. Map of northeastern United States with geographic location of tick and rodent sampling sites.

Sample preparation

DNA was extracted directly from blood samples using a QIAamp Blood Extraction Kit (Qiagen, Chatsworth, CA). The protocol followed was as suggested by the manufacturer. Briefly, detergent lysis was in the presence of proteinase K for 10 min at 70oC. The lysed material was applied to a spin column containing a silica gel-based membrane and washed twice. Purified DNA was eluted from the columns in 200 μl Tris-HCl (10 mM, pH 8.0) and stored at 4oC until used as template for PCR amplification.

DNA was extracted from I. scapularis ticks using a modification of the manufacturer’s protocol for the QIAamp Tissue Kit (Qiagen) as previously described (13).
 

PCR analysis

A nested PCR that amplified the 5’ region of the 16S rRNA gene was used to identify granulocytic ehrlichiae in tick and wildlife samples (13). Briefly, primary amplifications consisted of 40 cycles in a Perkin Elmer 9600 thermal cycler with each cycle consisting of a 30-sec denaturation at 94oC, 30-sec annealing at 55oC, and a 1-min extension at 72oC. The 40 cycles were preceded by a 2-minute denaturation at 95oC, and followed by a 5-minute extension at 72oC. Primary amplifications used primers ge3a and ge10r and reagents from the GeneAmp PCR Kit with AmpliTaq DNA Polymerase (Perkin Elmer, Applied Biosystems Division, Foster City, Calif.). Each reaction contained 5 µl of purified DNA as template in a total volume of 50 µl, and 200 µM each dNTP (dATP, dCTP, dGTP and dTTP), 1.25 units Taq polymerase, and 0.5 µM each primer. Reaction products were subsequently maintained at 4oC until analyzed by agarose gel electrophoresis or used as template for nested reactions.

Nested amplifications used primers ge9f and ge2 and 1 µl of the primary PCR product as template in a total volume of 50 µl. Each nested amplification contained 200 µM each dNTP (dATP, dCTP, dGTP, and dTTP), 1.25 units Taq polymerase, and 0.2 µM each primer. Nested cycling conditions were as described for the primary amplification, except 30 cycles were used. Reactions were subsequently maintained at 4oC until analyzed by agarose gel electrophoresis or purified for DNA sequencing.

DNA sequencing and data analysis

All samples producing positive PCR products were subjected to DNA sequencing reactions using fluorescent-labeled dideoxynucleotide technology (Dye Terminator Cycle Sequencing Ready Reaction Kit; Perkin-Elmer, Applied Biosystems Division). Sequencing reaction products were separated, and data were collected using an ABI 377 automated DNA sequencer (Perkin-Elmer, Applied Biosystems Division). The full sequence was determined for both strands of each DNA template to ensure maximum accuracy of the data. Sequences were edited and assembled using the Staden software programs (20), and analyzed using the Wisconsin Sequence Analysis Package (Genetics Computer Group, Madison, Wis.)(21).
 

Results

A total of 50 of 375 (13.3%) I. scapularis ticks from Bridgeport, Connecticut were PCR positive for ehrlichiae (Table 1). The percentage positive within each of the 4 years ranged from a low of 6.1% positive in nymphs in 1998 to 23.3% in adults in 1996. Less year-to-year variation was noted among adult ticks where infection prevalence ranged from 11.7% (1997) to 23.3% (1996). PCR analysis of EDTA blood samples from white-footed mice collected in Connecticut during the summer and fall of 1997 and spring of 1998 showed that 17 of 47 (36.2%) were positive in 1997, and 3 of 5 (60%) in 1998 (22). The amplification products were sequenced for each ehrlichia PCR positive mouse and tick. All products from samples collected in the Bridgeport, CT area from 1996 through 1998 had sequence identical to the 16S rRNA gene (EP-ha) previously amplified and sequenced from documented human infections in the Northeast and Upper Midwest US, and in Europe (4). The 16S rRNA sequence determined from adult ticks collected from Bridgeport in 1999 showed that all 12 positives also contained the human agent (EP-ha), although one of the ticks produced a mixed sequence suggesting the presence of more than one agent. The PCR products from this tick were cloned, and individual clones purified and sequenced. These data confirmed the presence of a mixed population of ehrlichiae containing 16S rRNA sequences that matched EP-ha, and that differed from EP-ha by 2 nucleotides. The latter sequence was identical to a variant (hereafter, variant 1) previously described in Rhode Island ticks, and in deer in Maryland and Wisconsin (13,14) (Table 2). In contrast to ticks and rodents examined from the Bridgeport area, nymphal ticks collected in 1997 from Bluff Point in southeastern Connecticut contained a nearly equal distribution of EP-ha (5 of 9 positives; 55.6%) and variant 1 (4 of 9 positives; 44.4%) ehrlichia.
 

Table 1. Spatial and temporal variation in the occurrence and distribution of EP-ha and E. phagocytophila variants.

* Includes one tick that was positive for both the HGE agent and variant 1.

Table 2. Variation within nucleotide region 74 to 446 among 16s rRNA gene sequences obtained for EP-ha, the Rhode Island variants, and E. equi. The number designations for the EP-ha 16S rDNA sequence correspond to those reported by Chen et al. (4). Gene Bank Accession No. U02521.

Rhode Island samples from I. scapularis ticks, white-footed mice, and chipmunks contained E. phagocytophila variants as well as EP-ha. A total of 123 of the 538 ticks (22.9%) examined were positive for E. phagocytophila by PCR, including 61 of 232 adults (26.3%) and 62 of 306 nymphs (20.3%). DNA sequencing was performed on 92 of these PCR products and overall, only 24 (26.1%) showed sequences identical to those of EP-ha. Fifteen (16.3%) ticks showed sequences corresponding to variant 1. The remainder of the ticks (53 or 57.6%) showed a novel sequence differing from EP-ha by 2 nucleotides, and from variant 1 by 4 nucleotides (hereafter, variant 2)(Table 2). PCR testing of blood samples from Rhode Island chipmunks in 1996 (n=19) detected 11 positives (57.9%). DNA sequencing of these PCR products showed that 9 were identical to the sequence of EP-ha, and the remaining 2 represented novel variant sequences each differing from EP-ha by just 1 nucleotide (variants 3 and 4; Table 2). The host and vector association of EP-ha and the 4 E. phagocytophila variants found in Rhode Island are shown in Table 3.

Table 3. Host and vector associations of E. phagocytophila and the four 16S rDNA sequence variants detected in Rhode Island from 1996 through 1999.

1. Based on samples from 35 confirmed human infections from various states including Rhode Island and Connecticut.
2. Variant 1 detected only in ticks in RI and CT; positive deer samples collected in Maryland and Wisconsin (13, 14).

Discussion

Strains of Ehrlichia phagocytophila exist in nature that are capable of causing disease in sheep, cattle, horses, dogs, cats, and humans. These strains, including the species previously known as the HGE agent and E. equi, are grouped as a single species based on the close relationship of these agents at the genetic and antigenic levels. However, there are clearly biological and ecological differences between strains of E. phagocytophila including varying host pathogenicity, vectors, DNA sequence, and geographic distribution. Small ribosomal subunit (16S) DNA sequences are very highly conserved in bacteria and are often used to identify and differentiate bacterial species. The 16S rRNA gene sequences that have been amplified from every confirmed human case, with the exception of two isolated cases in Northern California, have been identical to the E. phagocytophila-human agent (EP-ha) sequence determined by Chen et al. (4). Recently, an ehrlichia with a 16S rRNA gene sequence differing from EP-ha by a single nucleotide was identified in white-tailed deer from Maryland and Wisconsin, and in I. scapularis from Rhode Island (13,14).

Examination of every PCR-positive white-footed mice (n=20) and I. scapularis ticks (n=38) collected in Bridgeport, Connecticut from 1996 through 1998 showed that they harbored E. phagocytophila identical in sequence to EP-ha for a 546-bp region of 16S rRNA gene (4). Sequence analysis of PCR products from two Connecticut deer blood samples had DNA identical to the EP-ha p44 gene sequence (23). However, examination of mice and ticks from Rhode Island showed a much lower percentage of isolates like EP-ha, and several E. phagocytophila variants with novel 16S rRNA gene sequences. These data indicate that variant forms of E. phagocytophila, not yet associated with human or veterinary disease, frequently occur in Rhode Island. The same or additional E. phagocytophila variants may also occur in other regions of the United States. Most PCR assays will amplify products from these agents of similar size to the human agent PCR product, so the variants are indistinguishable when analyzing products only by agarose gel electrophoresis. Our findings also suggest that results from other human-infection prevalence surveillance studies in ticks and rodents that have not included PCR product sequencing may be misleading. For example, had we not sequenced our PCR products for the year 1997, we would have concluded that 46.4% and 38.3% of nymphal and adult I. scapularis ticks, respectively, collected in southern Rhode Island were positive for EP-ha. In truth, only 11.9% of the positives that were sequenced, and an estimated 5.0% of the total tested ticks, corresponded to EP-ha positives, with the remainder of the 1997 RI-positives consisting of genetic variants not yet associated with human disease.

The 16S rRNA sequences obtained from tick and rodent samples collected from Bridgeport, CT from 1996 through 1998 were identical to EP-ha, but in 1999, one tick collected in Bridgeport was positive for both a variant (variant 1) previously found in RI, and EP-ha. A retrospective analysis of ticks collected in eastern CT (Bluff Point) demonstrated that variant 1 could be found as early as 1997 in a region of CT close to the RI border. The inability to detect variant 1 following extensive sampling in Bridgeport from 1996-1998, but its appearance in 1999 suggests that the geographic range of this variant may be expanding westward. Additionally, the identification of the co-infected tick is interesting in that it represents the first case of more than one strain of E. phagocytophila being detected in a single tick vector. This indicates that two strains of the agent are capable of co-existing in a single tick, at least transiently, and that they can survive the molting process since the co-infection was found in an unfed, host-seeking adult tick.

The results from RI samples collected in 1997 are unusual in several regards. First, there was a very high rate of E. phagocytophila-positive ticks (42.2% nymphs + adults) relative to all other tick populations sampled from 1996 through 1999, and many of the positives were shown to be variant 2 (79.1% of PCR positives sequenced). Second, the 1997 RI ticks represent the only population where the ehrlichia prevalence was higher in nymphs than adults (46.4% nymphs and 38.3% adults). Lastly, variant 2 sequences were also seen in samples collected in 1997 from white-footed mice and chipmunks, but were never noted again either before or after 1997. The fact that both nymphal and adult questing ticks were positive for variant 2 indicates that the variant was present in reservoir species during both larval and nymphal feedings, and likely spanned the fall of 1996 through the summer of 1997. The reason that this variant appeared only in 1997 in RI, was the prevalent strain in that year, and then completely disappeared is not known. It is possible that variant 2 may be more common in a reservoir species that we have not examined, and one that is a less-common target for questing ticks. It may be that during 1996-1997, host populations of preferred by immature I. scapularis (i.e. white-footed mice and chipmunks) were lower than normal resulting in a higher proportion of ticks feeding on atypical hosts that also harbored variant 2. These ticks could have then transmitted variant 2 to additional more common hosts after molting, resulting in the positive mice and chipmunks found in RI in 1997. Subsequent re-establishment of normal host populations may then have diluted the prevalence of variant 2 as immature tick feeding reverted to preferred hosts over the variant 2-bearing reservoir.

Although the function and biological significance of these genetic differences is unknown, we hypothesize that the variants may be interfering with maintenance and transmission of the true agent of human disease (EP-ha). Even accounting for an increased human ehrlichiosis case surveillance effort in Connecticut, the number of confirmed and suspected cases differs dramatically between the two neighboring states, with several hundred cases reported in Connecticut compared to <25 cases reported in Rhode Island. There were 178 confirmed or suspected cases between 1995 and 1997 (24) and case reports in Connecticut increased substantially in 1998 (228 provisional, 104 confirmed, CT Dept. of Public Health). These two states share a common border and many ecological factors known to support natural maintenance of both Borrelia burgdorferi and granulocytic ehrlichiae, such as populations of the tick vector I. scapularis and reservoir rodents, including white-footed mice (P. leucopus)(19,22). The incidence of Lyme disease in Connecticut and Rhode Island has been the highest in the nation for several years with Connecticut having a reported incidence only ca. 1.3-1.5 times higher than Rhode Island (25). By contrast, there appears to be a 50-fold difference in the incidence of reported ehrlichiosis between the two states. Therefore, it may be that the ehrlichia variants possess a competitive advantage over the EP-ha, possibly in infecting certain reservoir or vector populations. If true, a lower incidence of EP-ha and less human disease would be expected in areas where the variants predominate, since a lower percentage of ticks would harbor EP-ha.

Rickettsia rickettsii, the etiologic agent of Rocky Mountain spotted fever (RMSF), was first identified in the early 1900’s based on its association with human disease (26). Subsequent studies of veterinary infections and tick populations identified numerous additional Rickettsia species closely related to R. rickettsii, all clearly members of the spotted fever group but not associated with human disease. These species include R. montana, R. rhiphicephali, R. parkeri, R. bellii, and the “east side agent” R. peacockii (29, 30). It has been suggested that nonpathogenic rickettsiae interfere with the development of more virulent R. rickettsii in Dermacentor ticks, and that these nonpathogenic rickettsiae are often found with much greater frequency in ticks than are the more virulent species (29, 30). Our data suggest that a similar situation may exist among the granulocytic ehrlichiae, with both pathogenic and nonpathogenic genetic variants co-existing in nature. Isolation of the new variants will allow us to address the competitive advantage hypothesis experimentally in both ticks and mice through the use of mixed infections in the laboratory.

The identification and use of novel gene targets that are more variable than the 16S rRNA gene will allow us to better assess variability between strains of E. phagocytophila (31-33). Future studies should include E. phagocytophila from additional geographical areas where a significant number of human cases of granulocytic ehrlichiosis are reported (NY, WI, MN) in comparison to areas with similar vector densities but where little or no disease is present (NJ, PA, DE, MD, CA).

References

1. Maeda K, Markowitz N, Hawley RC, Ristic M, Cox D, McDade JE. 1987. Human infection with Ehrlichia canis, a leukocytic rickettsia. N Engl J Med 1987;316:853-6.

2. Anderson BE, Dawson JE, Jones DC, Wilson KH. Ehrlichia chaffeensis, a new species associated with human ehrlichiosis. J Clin Microbiol 1991;29:2838-42.

3. McQuiston JH, Paddock CD, Holman RC, Childs JE. The human ehrlichioses in the United States. Emerging Infect Dis 1999;5:635-642.
   

4. Chen SM, Dumler JS, Bakken JS, Walker DH. Identification of a granulocytotropic Ehrlichia species as the etiologic agent of human disease. J Clin Microbiol 1994;32:589-95.

5. Bakken JS, Dumler JS, Chen SM, Eckman MR, Van Etta LL, Walker DH. Human granulocytic ehrlichiosis in upper midwest United States. A new species emerging? JAMA 1994;272:212-8.

6. Gerwirtz, AS, Cornbleet PJ, Vugia DJ, Traver C, Niederhuber J, Kolbert CP, et al. Human granulocytic ehrlichiosis: report of a case in northern California. Clin Infect Dis 1996;23:653-4.

7. Hardalo CJ, Quagliarello V, Dumler JS. Human granulocytic ehrlichiosis in Connecticut: report of a fatal case. Clin Infect Dis 1995;21:910-4.

8. Petrovec M, Furlan SL, Zupanc TA, Strle F, Broqui P, Roux V, et al. Human disease in Europe caused by a granulocytic Ehrlichia species. J Clin Microbiol 1997;35:1556-9.

9. Broqui P, Dumler JS, Lienhard R, Brossard M, Raoult D. Human granulocytic ehrlichiosis in Europe. Lancet 1995;346:782-3.
   

10. Dumler JS, Bakken JS. Ehrlichial diseases of humans: emerging tick-borne infections. Clin Infect Dis 1995;20:1102-10.

11. Sumption KJ, Wright DJM, Cutler SJ, Dale BAS. Human ehrlichiosis in the UK. Lancet 1995;346:1487-8.

12. Foley, J.E., L. Crawford-Miksza, J.S. Dumler, C. Glaser, J.-S. Chae, E. Yeh, et al. Human granulocytic ehrlichiosis in Northern California: two case descriptions with genetic analysis of the ehrlichiae. Clin Infect Dis 1999;29:388-92.

13. Massung RF, Slater K, Owens JH, Nicholson WL, Mather TN, Solberg VB, et al. A nested PCR Assay for the detection of granulocytic ehrlichiae. J Clin Microbiol 1998;36:1090-5.

14. Belongia EA, Reed KD, Mitchell PD, Kolbert CP, Persing DH, Gill JS, et al. Prevalence of granulocytic Ehrlichia infection among white-tailed deer in Wisconsin. J Clin Microbiol 1997;35:1465-8.

15. Dawson JE, Warner CK, Baker V, Ewing SA, Stallknecht DE, Davidson WR, et al. Ehrlichia-like 16S rDNA sequence from wild white-tailed deer (Odocoileus virginianus). J Parasitol 1996;82:52-8.

16. Zeidner NS, Burkot TR, Massung RF, Nicholson WL, Dolan MC, Rutherford JS, et al. Transmission of the agent of HGE by Ixodes spinipalpis ticks: Evidence of an enzootic cycle of co-infection with Borrelia burgdorferi in Northern Colorado. J Infect Dis 2000;182:616- 9.

17. Nicholson MC, Mather TN. 1996. Methods for evaluating Lyme disease risks using geographic information systems and geospatial analysis. J Med Entomol 1996;33:711-20.

18. Pancholi P, Kolbert CP, Mitchell PD, Reed KD, Dumler JS, Bakken JS, et al. Ixodes dammini as a potential vector of human granulocytic ehrlichiosis. J Infect Dis 1995;172:1007-12.

19. Telford SR III, Dawson JE, Katavolos P, Warner CK, Kolbert CP, Persing DH. Perpetuation of the agent of human granulocytic ehrlichiosis in a deer tick-rodent cycle. Proc Natl Acad Sci USA 1996;93:6209-14.

20. Dear S, Staden R. A sequence assembly and editing program for efficient management of large projects. Nucleic Acids Res 1991;19:3907-11.

21. Devereux J, Haeberli P, Smithies O. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 1984;12:387-95.

22. Stafford III KC, Massung RF, Magnarelli LA, Ijdo JW, Anderson JF. Infection with agents of human granulocytic ehrlichiosis, Lyme disease, and babesiosis in wild white-footed mice (Peromyscus leucopus) in Connecticut. J Clin Microbiol 1999;37:2887-92.

23. Magnarelli LA, Ijdo JW, Stafford III KC, Fikrig E. Infections of granulocytic ehrlichiae and Borrelia burgdorferi in white-tailed deer in Connecticut. J Wildl Dis 1999;35:266-74.

24. Centers for Disease Control and Prevention. Statewide surveillance for ehrlichiosis— Connecticut and New York. MMWR Morb Mortal Wkly Rep 1998;47:476-80.

25. Centers for Disease Control and Prevention. Lyme disease-United States. MMWR Morb Mortal Wkly Rep 1997;46:531-35.

26. Wolbach SB. Studies on Rocky Mountain spotted fever. J Med Res 1919;41:2-197.

27. Hackstadt T. The biology of rickettsiae. J Inf Dis 1996;5:127-43.

28. Niebylski ML, Schrumpf ME, Burgdorfer W, Fischer ER, Gage KL, Schwan TG. Rickettsia peacockii sp. Nov., a new species infecting wood ticks, Dermacentor andersoni, in western Montana. Int J Syst Bact 1997;47:446-52.

29. Burgdorfer W, Hayes SF, Mavros AJ. Nonpathogenic rickettsiae in Dermacentor andersoni: a limiting factor for the distribution of Rickettsia rickettsiae. In: Burgdorfer W, Anacker RL, editors. Rickettsiae and rickettsial diseases. New York: Academic Press; 1981. P. 585-94.

30. Burgdorfer W. Ecological and epidemiological considerations of Rocky Mountain spotted fever and scrub typhus. In: Walker DH, editor. Biology of Rickettsial Diseases. Boca Raton, FL: CRC Press; 1988. P. 33-50.

31. Massung RF, Owens JH, Ross D, Reed KD, Petrovec M, Bjoersdorff A, et al. Sequence analysis of the ank gene of granulocytic ehrlichiae. J Clin Microbiol 2000;38:2917-22.

32. Storey JR, Doros-Richert LA, Gingrich-Baker C, Munroe K, Mather TN, Coughlin RT, et al. Molecular cloning and sequencing of three granulocytic Ehrlichia genes encoding high-molecular-weight immunoreactive proteins. Infect Immun 1998;66:1356-63.

Zhi N, Ohashi N, Rikihisa Y. Multiple p44 genes encoding major outer membrane proteins are expressed in the human granulocytic ehrlichiosis agent. J Biol Chem 1999;274:17828-36.