|Home | About | Journals | Submit | Contact Us | Français|
Anaplasma phagocytophilum, the causative agent of human granulocytic anaplasmosis (HGA), shares the same enzootic life cycle as Borrelia burgdorferi, the causative agent of Lyme disease. Although La Crosse, WI, is a well-recognized Lyme disease focus with an abundance of Ixodes scapularis vector ticks and the first documentation of HGA occurred in patients from northwestern Wisconsin, local transmission of A. phagocytophilum has not to date been documented. In this study, we evaluated DNA extracted from 201 ticks captured locally by a real-time PCR that targeted a unique region within msp2, and 24 samples (12%) yielded positive results. The PCR also detected A. phagocytophilum DNA in blood samples obtained from 53 patients with clinical abnormalities consistent with HGA, and sequencing confirmed that the DNA was recovered from the Ap-ha variant of A. phagocytophilum, associated exclusively with human infection. The findings therefore confirmed that the upper Midwestern focus for HGA endemicity now includes the regions immediately surrounding La Crosse, WI. The results also validated the utility of the real-time msp2 PCR test for confirming acute HGA in the clinical setting.
Anaplasma phagocytophilum, previously known as Ehrlichia phagocytophila, Ehrlichia equi, or human granulocytic ehrlichiosis agent, is an obligate intracellular bacterium that infects neutrophilic and rarely eosinophilic granulocytes (11). The infection and associated illness, now termed human granulocytic anaplasmosis (HGA), were first identified in 1994, when a patient from Wisconsin developed a severe febrile illness after a tick bite and subsequently died (4). Since then, the number of HGA cases has increased annually (17), especially in the northeastern and upper Midwestern United States, where the tick vector Ixodes scapularis is prevalent.
Patients with HGA frequently present with a febrile illness shortly after a tick bite. In addition, leukopenia, thrombocytopenia, and a mild to moderate elevation of hepatic enzymes are often present (7, 16). However, the results from routine hematological and chemistry blood tests are not pathognomonic, which makes accurate diagnosis challenging. Moreover, other factors such as age, neutrophilia, lymphopenia, anemia, and immunosuppression can increase the severity of the illness and risk for hospitalization (7, 11, 19).
Borrelia burgdorferi, the causative agent of Lyme disease, and A. phagocytophilum share the same enzootic life cycle, where ticks of the Ixodes persulcatus complex become infected as they feed on small mammals (27). Therefore, persons who reside in areas where Lyme disease is endemic may also be at risk of contracting HGA. However, a previous investigation (21) failed to detect A. phagocytophilum in mice or ticks captured near La Crosse, WI, even though the region is a well-recognized Lyme disease focus (10, 21). Despite this, area clinicians continued to report increasing numbers of patients with clinical signs and symptoms that suggested HGA. In addition, HGA was first described in Wisconsin patients (4), and the northwestern section of the state is a well-recognized (6) focus of endemicity. We therefore continued our efforts to document expansion of the region of endemicity to areas surrounding La Crosse, WI, by reexamining I. scapularis ticks for the presence of A. phagocytophilum DNA. We also evaluated the DNA recovered from the blood of patients with tick bites or recent exposure to ticks and acute symptoms that could be associated with HGA using a highly specific and sensitive PCR test (13, 23).
Adult female I. scapularis ticks were collected from 2 sites in regions directly north and south of La Crosse, WI. Site SC was approximately 5 miles south of Blair, WI, in Trempealeau County. Site CB was approximately 1 mile southeast of Coon Valley, WI, in Vernon County. Adult I. scapularis ticks were collected by flagging the underbrush from September to November 2008. The ticks were transported immediately to the laboratory and stored in groups of 5 to 10 at 8°C and 100% humidity in mesh-covered vials containing about 1/2 inch of plaster of paris. The vials were moistened periodically with distilled water.
The protocols were reviewed and approved by the Institutional Review Board (IRB) of the Gundersen Lutheran Medical Center. To facilitate clinical care, patient blood samples were tested for the presence of HGA-related DNA at the request of the attending clinicians. Blood samples were collected during 2008 and 2009 from patients who presented with clinical abnormalities that could be caused by infection with A. phagocytophilum. Samples were transported to the laboratory within 24 h of collection. In addition, the medical records of the PCR-positive patients were examined to determine whether the epidemiological histories and clinical symptoms and signs were consistent with HGA (3, 5, 7, 19, 20, 29).
The DNA from the blood samples was immediately recovered using the QIAamp DNA blood minikit (Qiagen, Valencia, CA). Briefly, 1 ml of whole blood was centrifuged at 2,500 × g for 10 min. Following centrifugation, 200 μl of buffy coat was combined with 20 μl of protease, 200 μl of buffer AL (Qiagen), 1 μl of carrier DNA (Sigma-Aldrich, St. Louis, MO), and 5 μl of exogenous DNA processing control. The DNA processing control consisted of lambda phage that contained 5 kb of mouse hepatitis virus DNA (EraGen, Madison, WI). The exogenous DNA was then detected in the msp2 PCR by including forward primer 5′-CCTGTGCGGGCAAGAAAG-3′, reverse primer 5′-CGCATCCAGTGCGAAGGT-3′, and probe 5′-hexachloro-6-carboxyfluorescein (HEX)-CGAGTTTAACGACAAGCCCCAAAGTCA-black hole quencher 1a (BHQ1a)-5HEX-3′. After the suspension was mixed thoroughly and incubated at 56°C for 10 min, 200 μl of 100% ethanol was added, the mixture was transferred to a column and washed, and the DNA was eluted in buffer AE as recommended by the manufacturer (Qiagen).
Individual ticks were also processed using the QIAamp DNA minikit (Qiagen) with minor modifications. Each tick was placed in a sterile 1.5-ml microcentrifuge tube (Sarstedt, Inc., Newton, NC) that contained 180 μl of buffer ATL (Qiagen) and then cut in half using a scalpel fitted with a disposable sterile blade. After the tick was cut, 20 μl of proteinase K was added and the suspension was mixed by vortexing and then incubated at 56°C for 2 h. Following incubation, 200 μl of buffer AL, 5 μl of exogenous DNA processing control, and 1 μl of carrier DNA were added, and the suspensions were mixed and incubated at 70°C for 10 min. Following incubation, the samples were centrifuged briefly to sediment the remaining exoskeleton; the supernatant was then transferred to a sterile 1.5-ml microcentrifuge tube (Sarstedt), and 200 μl of 100% ethanol was added. The mixture was then transferred to a column and washed, and the DNA was eluted with buffer AE (Qiagen) as recommended by the manufacturer. To prevent DNA carryover, individual ticks were cut only in individual sterile tubes and separate disposable sterile blades were used for each tick. The extracted DNA samples were then stored at −20°C until tested.
The extracted DNA was amplified by using a real-time PCR (13) that targeted A. phagocytophilum msp2 with primers ApMSP2f (5′-ATGGAAGGTAGTGTTGGTTATGGTATT-3′) and ApMSP2r (5′-TTGGTCTTGAAGCGCTCGTA-3′) and probe ApMSP2p-FAM (5′-6-carboxyfluorescein [FAM]-TGGTGCCAGGGTTGAGCTTGAGATTG-BHQ1a-FAM-3′). Five microliters of extracted DNA was combined with 20 μl of master mix that contained 12.5 μl of AP buffer mix (10× AmpliTaq Gold buffer, 2.5 mM MgCl2, deoxynucleoside triphosphates [dNTPs]), 4.5 μl of AP primer/probe mix (20 μM), 2.5 μl of exogenous control primers and probe, and 0.5 μl of AmpliTaq Gold DNA polymerase (1.5 U). The DNA was then amplified using a Mx3000P real-time thermal cycler (Stratagene, Cedar Creek, TX) under the following conditions: 1 cycle at 95°C for 10 min, 40 cycles at 95°C for 15 s, 60°C for 1 min, and 72°C for 30 s, and 1 cycle at 25°C for 5 s.
DNA sequencing was performed on a region of the 16S rRNA gene as described previously (11). Briefly, primers GE9f and GE10r were combined with the DNA extracted from the patient blood samples, and products were amplified under the following conditions: 1 cycle at 94°C for 5 min, 40 cycles at 94°C for 30 s, 52°C for 30 s, and 72°C for 1 min, and 1 cycle at 72°C for 7 min. The resulting 919-bp DNA fragment was then ligated into the pGEM-T vector (Promega, Madison, WI), and plasmid preps were forwarded for sequencing (Laragen, Inc., Los Angeles, CA).
An appropriate positive control was confirmed prior to beginning the study. A. phagocytophilum DNA was extracted from the blood of a patient who presented to Gundersen Lutheran Medical Center in 1997 with fever, chills, headaches, and myalgia. In addition, the patient had thrombocytopenia and elevated alanine and aspartate aminotransferase levels, and morulae were observed in granulocytes. The extracted DNA was amplified using primers GE9f and GE10r and then ligated into the pGEM-T vector (Promega, Madison, WI) for sequencing (Laragen). After confirmation that the amplified DNA sequence had >99% (918/919 bp) identity with A. phagocytophilum variant Ap-ha (GenBank accession number U02521 ) by a BLAST search, the extracted patient DNA was amplified using the msp2 primers, and the amplification product was ligated into a pGEM-T vector. Limiting dilution in whole blood confirmed that the real-time PCR detected as few as 8 copies of the plasmid, and approximately 200 copies of the plasmid were then used as a positive control for each PCR run. In addition, negative findings were considered valid only when the 94-bp region of the exogenous control (mouse hepatitis virus) DNA was detected.
Since B. burgdorferi-infected I. scapularis ticks are highly endemic in the region surrounding La Crosse (10, 21), the clinicians also routinely requested a screening test for Lyme disease, and the results were noted when the medical charts of the PCR-positive patients were reviewed. The sera were tested by a commercially available whole-cell enzyme immunoassay (EIA) that detects but does not discriminate between IgM and IgG antibodies (Vidas IgG/IgM Lyme screen) according to the manufacturer's instructions. Samples that yielded a test value ≥1.0 were considered positive.
We collected 171 and 30 female adult I. scapularis ticks from sites SC and CB, respectively. Nineteen (11%) of the ticks from site SC and 5 (17%) of the ticks from site CB yielded positive PCR results. The findings therefore confirmed that significant numbers of ticks native to adjacent areas located directly north and south of La Crosse were harboring A. phagocytophilum.
A. phagocytophilum DNA was also detected in the blood samples from 53 (7%) of 711 individuals with epidemiologic findings and symptoms that the ordering physician thought were possibly consistent with HGA. In addition, false-negative reactions were not problematic, since the unrelated extraction control DNA was detected in each of the PCR-negative samples. The PCR-positive patients resided in La Crosse or communities located in areas surrounding the city, and none had traveled outside the region within 30 days prior to developing the illness. The average age was 55 years, and males (68%) predominated (Table 1). The illness also persisted for an average of 5 days (range, 1 to 25 days) before the patient sought treatment. In addition, 16 (30%) patients recalled a tick bite, and fever (81%), myalgia (68%), headache (58%), and fatigue (53%) were common. Additionally, 5 (9%) PCR-positive patients also had single lesions on the trunk characterized by the clinician as erythema migrans (EM) lesions characteristic of Lyme disease. Moreover, 6 (11%) additional patients required hospitalization because of complications that included respiratory distress and previous diagnosis of myasthenia gravis (n = 1), history of pneumonia within the previous 2 months (n = 1), concurrent pneumonia confirmed by abnormal chest X ray (n = 1), syncope (n = 1), high fever (n = 1), or mental confusion (n = 1).
Laboratory findings confirmed thrombocytopenia (mean, 75 × 103 platelets/μl; range, 25 × 103 to 133 × 103 platelets/μl) and leukopenia (mean, 2.5 × 103 leukocytes/μl; range, 1.4 × 103 to 3.5 × 103 leukocytes/μl) in 29 (66%) and 18 (41%) of the 44 PCR-positive patients who were tested, respectively (Table 2). Elevated levels of aspartate transaminase (80%) and alanine transaminase (52%) were also common. In addition, serum samples from 50 PCR-positive patients were tested for antibodies against B. burgdorferi by a whole-cell enzyme-linked immunosorbent assay (ELISA), and 6 (12%) had significant reactivity, including the 5 patients with concurrent EM skin lesions. The cases were treated with doxycycline (100 mg twice daily [b.i.d.]) for at least 10 days (29), and the antibiotic treatment resulted in the complete resolution of the clinical abnormalities in each patient.
As final confirmation of accuracy, we reamplified the DNA from 10 msp2 PCR-positive patients by using a PCR that targeted a 919-bp region of the 16S rRNA gene. The DNA sequences from 5 patients had 100% identity with the prototype A. phagocytophilum variant Ap-ha (GenBank accession number U02521), and the DNA from the remaining five patients had only 1 or 2 base pair differences (Table 3).
Anaplasmosis is a tick-borne disease first described in 1994 among residents of Wisconsin and Minnesota (4). The causative agent, A. phagocytophilum, has the same tick vector (I. scapularis) as B. burgdorferi, and the geographic area surrounding and including La Crosse, WI, is a well-recognized (10, 21) Lyme disease focus. However, previous efforts to document A. phagocytophilum in ticks collected from the region yielded negative findings, despite the presence of a well-recognized focus of HGA endemicity in the upper Midwest (4) that apparently originated in northwestern Wisconsin (6). For example, Jackson et al. (21) failed to detect the organisms by PCR in 713 ticks captured in the late 1990s from multiple sites, several of which overlapped the areas sampled in this investigation.
In this study, however, our findings confirmed that the region surrounding La Crosse is now also an area in which A. phagocytophilum is endemic. In support, we detected A. phagocytophilum DNA in ticks collected from sites located directly north and south of La Crosse and also in blood samples from 53 patients with no recent histories of travel outside the region. In addition, sequencing the DNA from a segment of the 16S rRNA gene from 10 msp2 PCR-positive patients confirmed that they were infected with the Ap-ha strain of A. phagocytophilum, associated exclusively with human illness (24).
A possible explanation for the failure to detect A. phagocytophilum in the previous investigation (21) was the decreased sensitivity of the PCR test, especially since the test targeted the 16S rRNA gene, which is present in lower copy numbers than msp2 targeted by the PCR in the present study. However, Massung and Slater (23) showed that PCR tests that targeted either the 16S rRNA or msp2 gene yielded similar sensitivities when A. phagocytophilum-spiked HL60 cells were tested.
Another possible explanation, therefore, may be that the emergence of A. phagocytophilum-infected ticks in the region reflects continued expansion of the upper Midwestern focus of HGA endemicity, which is a smaller focus contained within the well-characterized (10, 21) Lyme disease region. In support, we examined >300 archived DNA samples extracted from ticks collected from the region for the previous study by Jackson et al. (21) using the msp2 PCR test and failed to detect A. phagocytophilum (data not shown). In addition, monitoring exposure of canines to B. burgdorferi and A. phagocytophilum by serologic testing is useful for determining the severity and extent of a focus of endemicity (8), and recent polling confirmed that area veterinarians have detected antibodies that recognize A. phagocytophilum for only the past several years.
It should also be noted that the clinical signs and symptoms of the HGA patients were typical (5, 7, 11, 15, 17), with fever, headache, and myalgia predominating, and leukopenia, thrombocytopenia, and elevated transaminase levels common laboratory abnormalities. Moreover, coinfection with B. burgdorferi was also relatively common since 5 (9%) msp2 PCR-positive patients had positive results in a Lyme disease screening test and each also presented with a characteristic EM lesion. As further support, we recently found that 12 (7%) of 171 locally captured I. scapularis ticks submitted to the laboratory for analyses during 2010 were infected with A. phagocytophilum and B. burgdorferi. It is also interesting that each patient with a concurrent EM lesion was also seropositive for B. burgdorferi. A likely explanation, however, is that the high sensitivity was inflated because the whole-cell EIA, which detects both IgM and IgG antibodies, also detected cross-reactive antibodies produced in response to A. phagocytophilum (28) or to past exposure to other unrelated organisms (9). Another possibility is that the individuals with coinfections were exposed frequently to B. burgdorferi-infected ticks and the screening test was confounded by the past exposure. In support, 3 (60%) of the 5 HGA patients with concurrent EM lesions had relatively low (1.0 to 1.5) positive Lyme disease ELISA values.
The findings are also important for another reason. The nonspecific nature of HGA can make diagnosis difficult, and, in fact, there were no significant clinical differences between the patients with positive PCR findings and those that yielded negative results. However, accurate detection of the illness can be essential, especially since complicating factors such as increased age and coinfection with B. burgdorferi can significantly increase morbidity and lengthen the clinical manifestations (5, 7, 18, 26). The most common laboratory methods for confirming the illness include seroconversion or a 4-fold increase in antibody titer (1, 5, 6), but the accuracy is often dependent on comparing the antibody responses in paired acute- and convalescent-phase sera (5), which are rarely obtained after successful treatment. In addition, most cases present within the first week of infection, when antibodies are rarely present. More confounding, IgG antibodies may persist for months to years after infection (5, 14) and there can be significant seropositivity in healthy individuals who reside in foci of endemicity (2, 6). An alternative, therefore, is to identify characteristic intragranulocytic inclusions in Wright- or Giemsa-stained peripheral blood smears (3). However, the procedure is more labor-intensive, sensitivity decreases significantly if morulae are scarce, and other inclusions or overlying platelets can confound detection.
In contrast, PCR is a sensitive test (67 to 90%) for confirming acute-phase HGA (3, 22), and previous (13, 23) laboratory studies confirmed that real-time PCR that targets a unique region within msp2 is also specific. More significantly, our results extended these findings by confirming that the msp2 PCR test can also be used for detecting HGA in the clinical setting. In support, false-positive results were unlikely since each PCR-positive patient had appropriate epidemiologic histories and clinical signs and symptoms, and, while we did not specifically evaluate sensitivity, the A. phagocytophilum genome contains multiple (>80) copies of msp2 (25, 30) and the PCR detected as few as 8 copies. In addition, inhibitory factors were not problematic since an appropriate (12) unrelated extraction control DNA was detected in each PCR-negative sample.
In summary, B. burgdorferi and A. phagocytophilum share the same enzootic life cycle, so it is especially important to monitor Lyme disease foci for the additional possibility of HGA. La Crosse, WI, is centrally located in a well-documented Lyme disease focus, and the findings confirm that the region should now also be considered part of the upper Midwestern focus of endemicity for HGA. The results also validate the utility of the msp2 PCR test for confirming infection with A. phagocytophilum during the acute phase of the illness.
We gratefully acknowledge Lauren Leopold for assistance with the collection and processing of ticks. We also thank Mohandas Pillai for valuable assistance with DNA analyses.
Financial support was provided by the Gundersen Lutheran Medical Foundation, La Crosse, WI.
Published ahead of print on 14 September 2011.