Lymph node hemophagocytosis in rickettsial diseases: a pathogenetic role for CD8 T lymphocytes in human monocytic ehrlichiosis (HME)?
© Dierberg and Dumler; licensee BioMed Central Ltd. 2006
Received: 23 December 2005
Accepted: 21 July 2006
Published: 21 July 2006
Human monocytic ehrlichiosis (HME) and Rocky Mountain spotted fever (RMSF) are caused by Ehrlichia chaffeensis and Rickettsia rickettsii, respectively. The pathogenesis of RMSF relates to rickettsia-mediated vascular injury, but it is unclear in HME.
To study histopathologic responses in the lymphatic system for correlates of immune injury, lymph nodes from patients with HME (n = 6) and RMSF (n = 5) were examined. H&E-stained lymph node tissues were examined for five histopathologic features, including hemophagocytosis, cellularity, necrosis, and vascular congestion and edema. The relative proportions of CD68 macrophages, CD8 and CD4 T lymphocytes, and CD20 B lymphocytes were evaluated by immunohistochemical staining.
Hemophagocytosis was similar in HME and RMSF, and was greater than in control cases (p = .015). Cellularity in HME was not different from controls, whereas RMSF lymph nodes were markedly less cellular (p < 0.002). E. chaffeensis-infected mononuclear phagocytes were infrequent compared to R. rickettsii-infected endothelial cells. More CD8 cells in lymph nodes were observed with HME (p < .001), but no quantitative differences in CD4 lymphocytes, macrophages, or B lymphocytes were identified.
Hemophagocytosis, CD8 T cell expansion, and the paucity of infected cells in HME, suggest that E. chaffeensis infection leads to macrophage activation and immune-mediated injury.
The tick-borne obligate intracellular bacteria Ehrlichia chaffeensis and Rickettsia rickettsii are the causes of human monocytic ehrlichiosis (HME) and Rocky Mountain spotted fever (RMSF) diseases, respectively . E. chaffeensis is transmitted by the Lone Star tick, Amblyomma americanum, and human infection typically presents with fever, myalgias, pancytopenia, and mild to moderate elevation of serum transaminases . This clinical presentation is easily confused with that of RMSF, which typically presents with fever, rash, and headache. R. rickettsii is known to infect endothelial cells, leading to direct rickettsia-mediated vascular injury accompanied by a vigorous but protective Th1 immune response . E. chaffeensis is known to infect cells in the monocyte/macrophage lineage, but the pathogenesis of HME is less clear . Lymphohistiocytic-rich infiltrates, foamy macrophage infiltrates, and non-caseating granulomas suggest diffuse activation of the mononuclear phagocyte system and a role for host inflammation and immunity in its pathogenesis [5, 6]. In this study, we examined documented cases of HME and RMSF in order to quantitate and further delineate the pathology and underlying pathogenesis of E. chaffeensis infection.
Patients and tissue samples
H&E-stained and unstained slides or formalin-fixed paraffin-embedded lymph node tissue blocks were obtained for 6 cases of HME (5 autopsies and 1 biopsy) and 5 representative cases of RMSF autopsies. All cases were documented by serologic studies (HME 3 cases; RMSF 1 case), PCR amplification from peripheral blood (HME 2 cases; RMSF 0 cases), or immunohistochemical demonstration of either R. rickettsii or E. chaffeensis (HME 6 cases; RMSF 5 cases). The patients ranged in age from 6 to 80 years. In addition, lymph node tissue from 8 control cases were examined, including 4 infection controls (1 case each of murine typhus [autopsy], cat scratch disease [biopsy], tuberculosis [biopsy], and gram-positive coccus lymphadenitis [biopsy]) chosen to reflect a spectrum of potential Th1 and Th2 immune induction, and 4 consecutive, random autopsy cases from the Johns Hopkins Hospital for which lymph node tissues were available at the time of study inception (random controls, including 1 patient with pulmonary embolism/cardiomyopathy, 1 patient with congestive heart failure and chronic obstructive pulmonary disease, 1 patient with a subarachnoid hemorrhage and 1 patient with myasthenia gravis and bronchopneumonia). Control cases were selected to represent a broad, but not comprehensive spectrum of potential changes observed with other infections and non-infectious processes. The lymph node specimens taken during autopsy represent hilar and/or mediastinal lymph nodes, and for biopsies, cervical or inguinal lymph nodes were sampled. Generally one or two lymph nodes per patient were evaluated and an entire 5 μm tissue section for each lymph node was reviewed for relevant histopathologic features. No lymph nodes draining skin lesions or other obvious inflammatory foci were used. Approval for study of the tissues was obtained from The Johns Hopkins Medicine Institutional Review Board.
Histologic and immunohistologic staining
H&E-stained lymph node tissues were examined for five histopathologic features, including hemophagocytosis (macrophage activation), cellularity (inflammatory cell infiltration), necrosis (tissue damage), and vascular congestion (non-cellular tissue inflammation) and edema (vascular permeability). The relative proportions of CD68 macrophages (clone KP-1), CD8 (clone 1A5) and CD4 T lymphocytes (clone 1F6), and CD20 B lymphocytes (clone L26) were determined by immunohistochemical staining of paraffin-embedded tissues from 5 of 6 HME cases, 1 case of each of murine typhus and cat scratch disease, and 2 autopsy controls for which additional slides or tissue were available. No unstained slides or formalin-fixed paraffin-embedded lymph node tissue blocks were available for 4 of the 5 RMSF cases, prompting exclusion of RMSF from evaluation for immunophenotypic markers. All immunophenotyping antibodies were obtained from Ventana Medical Systems, Inc., (Tucson, AZ) and a modified avidin-biotin complex method was used after pretreatment of slides in citrate buffer. Specificity of each antibody reaction was confirmed prior to use consistent with the Quality Control recommendations of the College of American Pathologists Laboratory Accreditation Program, Anatomic Pathology Checklist, or the NCCLS Approved Guideline, and were confirmed with each run by using normal lymph node or tonsil tissues with defined distributions of the cell subsets studied.
Owing to the potential that semiquantitative assessment may not produce normally distributed results, we elected to use non-parametric studies to compare HME, RMSF, and control groups. This involved blinding the slides as to diagnosis and ranking the 19 cases for each histologic feature and the 9 cases for each immunophenotypic feature, respectively [7–10]. Cases were evaluated for quantity, distribution, and severity of each feature by slide-to-slide comparisons, until each case was assigned a rank number that corresponded to rank position for each feature. Ranking was conducted simultaneously by 2 microscopists to assure consensus. These ranks were evaluated for each feature using one-tailed Mann-Whitney U-statistic nonparametric tests where p values <.05 were considered significant. For each comparison of two groups (HME vs. controls, HME vs. RMSF, RMSF vs. controls), all features were re-ranked for appropriate application of Mann-Whitney U-statistical tests.
Immunohistologic demonstration of E. chaffeensis and R. rickettsii was performed using biotinylated human anti-E. chaffeensis for direct detection , E. chaffeensis monoclonal antibody 1A9 , or polyclonal rabbit anti-R. rickettsii  and appropriate biotinylated antibodies, followed by detection using either streptavidin-alkaline phosphatase (Dako) and naphthol phosphate/fast red/levamisole substrate solution or streptavidin-horseradish peroxidase and diaminobenzidine substrate. Extensive washing with phosphate-buffered saline was performed after each incubation. All specimens were counterstained with Mayer's hematoxylin, mounted with Crystal Mount (Biomeda, Hayward CA) and dried before applying coverslips. All immunohistologic preparations were examined at a magnification of 400× for identification of infected cells. E. chaffeensis morulae and individual R. rickettsii were identified by substrate reaction product, and assessed for bacterial load by a semiquantitative method ranging from none detected, graded as 0, and 1, 2, or 3 if infected cells were found in <10%, 10 to 25%, and >25% of microscopic fields, respectively.
Median (range) of histopathologic and immunohistologic rankings of features among patients with HME, RMSF, or controls. High ranks correspond to a greater degree for each histologic or immunohistologic feature.
Histologic Feature n = 19
HME n = 6
RMSF n = 5
P HME vs. RMSF
Controls n = 8
P HME vs. controls
P RMSF vs. controls
P HME vs. controls
CD68 [n = 9]
7  (1–10)
3.5  (2–8)
CD4 [n = 8**]
5 [4**] (3–9)
6  (2–8)
CD8 [n = 9]
8  (6–10)
2.5  (1–5)
CD20 [n = 7**]
6 [3**] (3–8)
3.5  (1–7)
Immune-mediated control of facultative and obligate intracellular bacterial infection is particularly problematic due to the sequestered location and the lack of immune effector access to the pathogens. Thus, although many of these infections are thought to cause tissue damage by direct bacteria-mediated cell injury, an open question is to what degree host inflammation and immunity contribute to overall disease pathogenesis. Although much study of the histopathology of both HME and RMSF has been conducted, little correlation of these findings with potential pathogenetic mechanisms has been attempted; for example, the occurrence of hemophagocytosis in both entities is well described [5, 13–15], but the potential importance of this feature in immunopathologic reactions is not well addressed. Tissue injury seen in RMSF is most likely a combination of direct rickettsia-mediated endothelial cell injury, manifested by vasculitis and increased vascular permeability [16, 17], as well as by immune-mediated injury, demonstrated here by hemophagocytosis. Animal studies have shown that in balance, the potentially detrimental effects of Th1 immunity, local macrophage activation, and generation of cytotoxicity associated with CD8 or NK cells in rickettsial infection are beneficial to the host, owing to effective pathogen control [16, 18]. IFNγ, TNFα, and other pro-inflammatory mediators produced by these cells when responding to obligate intracellular infections are known to activate macrophages, which inhibit growth of both R. rickettsii and E. chaffeensis. However, this process could also cause bystander tissue injury as suggested by the detection of hemophagocytosis [17, 19]. Given the differences in bacterial load between RMSF and HME, tissue damage in HME is more likely a result of poorly controlled macrophage activation and release of effector molecules, including nitric oxide and reactive oxygen species [3, 17, 20].
The excessive cytokine production induced with E. chaffeensis infection is believed to contribute to the toxic and septic shock-like presentation seen in many cases of HME . As demonstrated in this study, hemophagocytosis and evidence of inflammation-mediated tissue injury is observed in HME and RMSF to an extent greater than that in the selected infectious and non-infectious control cases. In addition, HME cases demonstrated significantly more infiltration by CD8 T lymphocytes than the control cases. This further indicates that CD8 cells and possibly cellular cytotoxicity likely play a significant role in HME pathogenesis. Whether a similar role for CD8 cells exists for RMSF, as suggested by studies in animal models, could not be adequately assessed here and needs further investigation.
Unlike the situation for the vasculotropic rickettsioses, evidence for direct ehrlichia-mediated cell injury in vivo has not been found, but the frequent presence of hemophagocytosis implicates inflammatory- and immune-mediated injury as possible major contributors. E. chaffeensis can cause cytolytic death of infected cells in vitro, but this is unlikely to be responsible for the degree of leukopenia, neutropenia and/or thrombocytopenia observed in patients with HME given the low frequency of infected cells in vivo as well as the lack of competent infection in platelets, neutrophils, and megakaryocytes. The relative lack of infected cells in HME compared to RMSF, despite similar degrees of hemophagocytosis, and the marked CD8 T cell expansion that could drive such responses, suggests that Ehrlichia chaffeensis infection often leads to macrophage activation, generation of inflammatory effectors, mediators, and cytokines, non-specific phagocytic activity, and consequent immune-mediated injury . Hematologic data were not available for these cases, which in general reflect the more severe end of the disease spectrum. However, it is well recognized that patients with HME and RMSF, especially when severe, demonstrate leukopenia, thrombocytopenia, and anemia, although the disease process and degree of cytopenias differ from those observed in severe acquired or hereditary hemophagocytic syndromes .
The precise mechanism by which E. chaffeensis modulates the host innate or adaptive immunity to generate such immunologic injury is unclear. Several recent studies have begun to shed light on ehrlichial immunopathogenesis, with infection of human macrophage cell lines resulting in transcriptional alterations associated with down-regulated macrophage function , with production of dysfunctional cytotoxic TNFα by CD8 T lymphocytes , or by direct stimulation of NKT cells to produce excessive quantities of IFNγ .
Potential pitfalls of this study include the biased selection of patients that represent the most severe spectrum of HME or RMSF – fatal cases – and the relatively small number of cases examined. Both of these situations result from the relatively rare access to pathologically-examined cases of both HME and RMSF, limiting the power of this study. Similarly, the small number of infectious controls could equally under-represent immunopathologic responses among these entities. Despite these limitations, the small number of cases allowed at least partial study, and provides evidence of rickettsial immunopathogenicity in humans that should provide impetus for further study in both HME and RMSF, even among mild to moderately severe infections.
HME and RMSF are both caused by related obligate intracellular bacteria, and both induce hemophagocytosis in lymph nodes to a degree greater than a cohort of randomly selected normal and infectious disease controls. Animal models clearly show a direct role for pathogen-induced cytotoxic injury in vivo with R. rickettsii. However, the relatively low quantities of E. chaffeensis and the marked expansion of CD8 T cells in HME support a role for immunopathology in vivo in humans, perhaps related to the dysregulated proinflammatory cytokine generation as demonstrated in animal models. Much more study will be required to dissect the immunopathogenesis of Ehrlichia infections, and the studies here provide additional evidence that infection may involve cytotoxic T cell activation and proliferation that could drive excessive macrophage activation as a mechanism for generating tissue-damaging effectors in HME.
Human monocytic ehrlichiosis
Rocky Mountain spotted fever
cells Natural killer cells
Tumor necrosis factor
Supported in part by NIAID grant R01 AI 41213. The authors would like to thank the immunohistochemistry laboratory at The Johns Hopkins Hospital for assistance in preparation of the immunophenotyping stains.
- Dumler JS, Barbet AF, Bekker CP, Dasch GA, Palmer GH, Ray SC, Rikihisa Y, Rurangirwa FR: Reorganization of genera in the families Rickettsiaceae and Anaplasmataceae in the order Rickettsiales : unification of some species of Ehrlichia with Anaplasma, Cowdria with Ehrlichia and Ehrlichia with Neorickettsia, descriptions of six new species combinations and designation of Ehrlichia equi and 'HGE agent' as subjective synonyms of Ehrlichia phagocytophila. Int J Syst Evol Microbiol. 2001, 51: 2145-2165.View ArticlePubMedGoogle Scholar
- Paddock CD, Childs JE: Ehrlichia chaffeensis : a prototypical emerging pathogen. Clin Microbiol Rev. 2003, 16: 37-64. 10.1128/CMR.16.1.37-64.2003.View ArticlePubMedPubMed CentralGoogle Scholar
- Walker DH, Valbuena GA, Olano JP: Pathogenic mechanisms of diseases caused by Rickettsia. Ann N Y Acad Sci. 2003, 990: 1-11.View ArticlePubMedGoogle Scholar
- Olano JP, Walker DH: Human ehrlichioses. Med Clin North Am. 2002, 86: 375-392. 10.1016/S0025-7125(03)00093-2.View ArticlePubMedGoogle Scholar
- Walker DH, Dumler JS: Human monocytic and granulocytic ehrlichioses. Discovery and diagnosis of emerging tick-borne infections and the critical role of the pathologist. Arch Pathol Lab Med. 1997, 121: 785-791.PubMedGoogle Scholar
- Dumler JS, Dawson JE, Walker DH: Human ehrlichiosis: hematopathology and immunohistologic detection of Ehrlichia chaffeensis. Hum Pathol. 1993, 24: 391-6. 10.1016/0046-8177(93)90087-W.View ArticlePubMedGoogle Scholar
- Martin ME, Bunnell JE, Dumler JS: Pathology, immunohistology, and cytokine responses in early phases of human granulocytic ehrlichiosis in a murine model. J Infect Dis. 2000, 181: 374-378. 10.1086/315206.View ArticlePubMedGoogle Scholar
- Martin ME, Caspersen K, Dumler JS: Immunopathology and ehrlichial propagation are regulated by interferon-gamma and interleukin-10 in a murine model of human granulocytic ehrlichiosis. Am J Pathol. 2001, 158: 1881-1888.View ArticlePubMedPubMed CentralGoogle Scholar
- Scorpio DG, Akkoyunlu M, Fikrig E, Dumler JS: CXCR2 blockade influences Anaplasma phagocytophilum propagation but not histopathology in the mouse model of human granulocytic anaplasmosis. Clin Diagn Lab Immunol. 2004, 11: 963-968. 10.1128/CDLI.11.5.963-968.2004.PubMedPubMed CentralGoogle Scholar
- Browning MD, Garyu JW, Dumler JS, Scorpio DG: Role of reactive nitrogen species in development of hepatic injury in a C57bl/6 mouse model of human granulocytic anaplasmosis. Comp Med. 2006, 56: 55-62.PubMedGoogle Scholar
- Yu X, Brouqui P, Dumler JS, Raoult D: Detection of Ehrlichia chaffeensis in human tissue by using a species-specific monoclonal antibody. J Clin Microbiol. 1993, 31: 3284-3288.PubMedPubMed CentralGoogle Scholar
- Dumler JS, Gage WR, Pettis GL, Azad AF, Kuhadja FP: Rapid immunoperoxidase demonstration of Rickettsia rickettsii in fixed cutaneous specimens from patients with Rocky Mountain spotted fever. Am J Clin Pathol. 1990, 93: 410-4.View ArticlePubMedGoogle Scholar
- Abbott KC, Vukelja SJ, Smith CE, McAllister CK, Konkol KA, O'Rourke TJ, Holland CJ, Ristic M: Hemophagocytic syndrome: a cause of pancytopenia in human ehrlichiosis. Am J Hematol. 1991, 38: 230-4.View ArticlePubMedGoogle Scholar
- Marty AM, Dumler JS, Imes G, Brusman HP, Smrkovski LL, Frisman DM: Ehrlichiosis mimicking thrombotic thrombocytopenic purpura. Case report and pathological correlation. Hum Pathol. 1995, 26: 920-925. 10.1016/0046-8177(95)90017-9.View ArticlePubMedGoogle Scholar
- Woodard BH, Farnham R, Bradford WD: Fatal Rocky Mountain spotted fever. Hematologic and lymphoreticular observations. Arch Pathol Lab Med. 1981, 105: 452-3.PubMedGoogle Scholar
- Feng HM, Walker DH: Mechanisms of intracellular killing of Rickettsia conorii in infected human endothelial cells, hepatocytes, and macrophages. Infect Immun. 2000, 68: 6729-36. 10.1128/IAI.68.12.6729-6736.2000.View ArticlePubMedPubMed CentralGoogle Scholar
- Valbuena G, Feng HM, Walker DH: Mechanisms of immunity against rickettsiae. New perspectives and opportunities offered by unusual intracellular parasites. Microbes Infect. 2002, 4: 625-633. 10.1016/S1286-4579(02)01581-2.View ArticlePubMedGoogle Scholar
- Walker DH, Olano JP, Feng HM: Critical role of cytotoxic T lymphocytes in immune clearance of rickettsial infection. Infect Immun. 2001, 69: 1841-1846. 10.1128/IAI.69.3.1841-1846.2001.View ArticlePubMedPubMed CentralGoogle Scholar
- Fisman DN: Hemophagocytic syndromes and infection. Emerg Infect Dis. 2000, 6: 601-8.View ArticlePubMedPubMed CentralGoogle Scholar
- Ismail N, Olano JP, Feng HM, Walker DH: Current status of immune mechanisms of killing of intracellular microorganisms. FEMS Microbiol Lett. 2002, 207: 111-20. 10.1111/j.1574-6968.2002.tb11038.x.View ArticlePubMedGoogle Scholar
- Ismail N, Soong L, McBride JW, Valbuena G, Olano JP, Feng HM, Walker DH: Overproduction of TNF-alpha by CD8+ type 1 cells and down-regulation of IFN-gamma production by CD4+ Th1 cells contribute to toxic shock-like syndrome in an animal model of fatal monocytotropic ehrlichiosis. J Immunol. 2004, 172: 1786-800.View ArticlePubMedGoogle Scholar
- Mosser DM: The many faces of macrophage activation. J Leukoc Biol. 2003, 73: 209-12. 10.1189/jlb.0602325.View ArticlePubMedGoogle Scholar
- Grom AA: Natural killer cell dysfunction: A common pathway in systemic-onset juvenile rheumatoid arthritis, macrophage activation syndrome, and hemophagocytic lymphohistiocytosis?. Arthritis Rheum. 2004, 50: 689-98. 10.1002/art.20198.View ArticlePubMedGoogle Scholar
- Zhang JZ, Sinha M, Luxon BA, Yu XJ: Survival strategy of obligately intracellular Ehrlichia chaffeensis : novel modulation of immune response and host cell cycles. Infect Immun. 2004, 72: 498-507. 10.1128/IAI.72.1.498-507.2004.View ArticlePubMedPubMed CentralGoogle Scholar
- Mattner J, Debord KL, Ismail N, Goff RD, Cantu C, Zhou D, Saint-Mezard P, Wang V, Gao Y, Yin N, Hoebe K, Schneewind O, Walker D, Beutler B, Teyton L, Savage PB, Bendelac A: Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections. Nature. 2005, 434: 525-529. 10.1038/nature03408.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2334/6/121/prepub