Trafficking Pathways and Characterization of CD4 and CD8 Cells Recruited to the Skin of Humans Experimentally Infected with Haemophilus ducr
http://www.100md.com
感染与免疫杂志 2005年第7期
Departments of Medicine Pathology and Laboratory Medicine Dermatology Microbiology and Immunology, Indiana University, Indianapolis, Indiana
Department of Pathology, Harvard Medical School
Joint Program in Transfusion Medicine, Children's Hospital, Boston, Massachusetts
ABSTRACT
T-cell homing to infected skin is not well studied in humans. We examined sites experimentally infected with Haemophilus ducreyi by immunohistochemistry and flow cytometry for expression of receptors and ligands involved in cutaneous T-cell homing and determined the phenotypes of the T cells that trafficked to skin. Endothelial cells expressed E-selectin in infected but not uninfected skin, while peripheral node addressin (PNAd) was minimally expressed in all samples. CC chemokine ligand 27 (CCL27) was expressed in the epidermis and endothelium of both infected and uninfected skin. Interestingly, CCL21, a chemokine thought to be associated principally with T-cell trafficking in the lymphatic compartment, was highly expressed on the endothelium of infected skin. Few naive cells were present in experimental lesions, emphasizing the combined role of PNAd and CCL21 in trafficking of this subset. Memory cells (CD45RA–) dominated both CD4 and CD8 T-cell populations at the site of infection. Effector memory (CD45RA– CD27–) CD4+ and CD8+ T cells were enriched in lesions. Although the CC chemokine receptor 7-positive (CCR7+) population of both central memory (CD45RA– CD27+) and effector memory cells was not enriched in the skin compared to peripheral blood, CCR7+ cells were not precluded from entering infected skin. Taken together with our previous work (D. Soler, T. L. Humphreys, S. M. Spinola, and J. J. Campbell, Blood 101:1677-1683, 2003), these studies led us to propose a model of memory T-cell trafficking to skin in response to experimental H. ducreyi infection.
INTRODUCTION
T cells move out of peripheral blood into tissues by a multistep process: an initial loose tethering mediated by selectins, followed by a tighter interaction mediated by chemokines, and then a firm adhesion mediated by integrins, finally resulting in diapedesis (21). Differential expression of ligands and their receptors allows for tissue-specific migration of T cells (30). For homing to skin, cutaneous lymphocyte antigen (CLA) and the CC chemokine receptor 4 (CCR4) and CCR10 expressed by T cells bind to their ligands E-selectin, CC chemokine ligand 17 (CCL17) (thymus- and activation-regulated chemokine), and CCL27 (cutaneous T-cell-attracting chemokine), respectively, on endothelial cells (6, 11, 18, 32). In contrast, T cells that enter secondary lymphoid tissues express L-selectin and CCR7 in order to bind to their respective ligands peripheral node addressin (PNAd) and CCL21 (6-cysteine chemokine, 6Ckine) on endothelial cells (29). CCL21 is also expressed outside of secondary lymphoid organs in humans, including in the small intestine, thyroid, pancreas, trachea, and salivary glands (16). Recently, CCL21 was found to be expressed by human dermal endothelial cells during T-cell autoimmune diseases such as atopic dermatitis, lichen planus, and graft-versus-host diseases. Lymphocytes infiltrating these lesions expressed CCR7 (8), suggesting that the CCR7-CCL21 interaction is also operative in T-cell homing to skin. Whether expression of this receptor-ligand pair occurs in response to a bacterial skin infection has not been studied.
In addition to expression of receptors that allow for tissue-specific homing, migratory T cells are also defined as naive, memory, or effector cells by combinations of surface markers, including but not limited to the CD45 isotypes RA and RO, the costimulatory molecule CD27, L-selectin, and chemokine receptors such as CCR7, CXCR3, CXCR4, or CCR5 (9, 17, 29, 30). For example, CD4 populations may be classified as naive (CD45RA+ CD27+ CCR7+), central memory (CD45RA– CD27+ CCR7+), and effector memory (CD45RA– CD27– CCR7–) (9, 29). Similarly, CD8 cells may be defined as naive (CD45RA+ CD27+ CCR7+), central memory (CD45RA– CD27+ CCR7+), effector memory (CD45RA– CD27– CCR7–),and effector (CD45RA+ CD27– CCR7–) (12, 13, 29). Functionally, naive and central memory CD4 cells are similar, capable of producing only interleukin-2 (IL-2) upon in vitro stimulation, whereas effector memory populations produce IL-4, IL-5, IFN-, and some IL-2. Among CD8 T cells, IFN- and perforin are produced only by the effector memory and effector populations.
Haemophilus ducreyi is an obligate human pathogen and the causative agent of the genital ulcer disease chancroid, which facilitates transmission of human immunodeficiency virus type 1 (4). To study the human immune response to H. ducreyi, we developed an experimental inoculation model with human volunteers (2, 33, 37). In the model, subjects are inoculated at multiple sites on the upper arm with an allergy testing device, which delivers the bacteria to both the epidermis and the dermis. Papules develop at inoculated sites, which evolve into pustules, simulating natural chancroid. The histopathology of experimental infection closely parallels that of naturally occurring disease (22, 25, 26). The lesion consists of a polymorphonuclear leukocyte infiltrate that forms an epidermal abscess and a deep dermal perivascular infiltrate of T cells and macrophages that resembles a poorly formed granuloma that extends interstitially to just below the epidermis (26, 36). By 48 h of infection, T cells make up about 50% of the leukocyte infiltrate and remain abundant throughout the course of experimental infection (20).
Most cutaneous homing studies are performed in the context of noninfectious inflammatory skin diseases (8, 18) or by using CLA+ subsets in peripheral blood as a surrogate system to study cells with skin-homing potential (19, 29, 38). There is a limited amount of information about T cells that home to skin in response to infection (23, 28, 32). Human inoculation experiments with H. ducreyi provide a unique opportunity to study the trafficking and phenotypes of T cells recruited chiefly to the dermis in response to this pathogen (20, 32). Previous work in our laboratory showed that CD4 cells isolated from H. ducreyi-infected pustules are memory cells that are enriched for the skin-homing markers CLA and CCR4. A smaller proportion express the cutaneous marker CCR10, but the CCR10+ cells are not significantly enriched compared to peripheral blood (32). In addition, about half the CD4 T cells in H. ducreyi-infected skin specimens are CD45RA– CD27+ CCR7+ central memory cells, but it was not clear whether these cells were enriched relative to peripheral blood (32).
In this study, we further defined the subsets of CD4 T cells present in lesions and for the first time examined the homing and memory/effector phenotypes of CD8 T cells that traffic to infected sites. We also examined longitudinally obtained biopsy specimens for expression of the skin-homing ligands E-selectin and CCL27 and compared them to expression of the lymph node-homing ligands PNAd and CCL21. Taken together with our previous work, these studies allowed us to define the trafficking pathways and phenotypes of T cells recruited in response to experimental H. ducreyi infection in the skin.
MATERIALS AND METHODS
Human subjects. Twenty-five healthy adult volunteers contributed 43 tissue specimens for this study (Table 1). Biopsy samples designated CS or Control were obtained from seven subjects who participated in previous human challenge trials and one subject who donated uninfected skin (2, 26). Tissue specimens for flow cytometry analysis were obtained from six volunteers who were infected for the purposes of this study, six subjects who participated in a mutant-parent trial (35), and five subjects who participated in a reinfection trial (34) (Table 1). Informed consent was obtained from the subjects in accordance with the guidelines for human experimentation of the U.S. Department of Health and Human Services and the Institutional Review Board of Indiana University-Purdue University at Indianapolis.
The enrollment procedures, exclusion criteria, preparation of the bacteria, and inoculation procedures for the human challenge experiments are described in detail elsewhere (2, 36, 37). Each volunteer was inoculated at two or three sites on the upper arm with live H. ducreyi 35000HP (a human-passaged isolate of strain 35000) and at one site with heat-killed 35000HP. Each subject donated up to three biopsy samples (Table 1). For immunohistochemical analysis, several subjects were assigned to biopsy 1 to 4 days after inoculation, while others were biopsied at the clinical endpoint, defined as the development of a painful pustule or a maximum of 14 days of infection. For flow cytometry analysis, the subjects donated tissue and peripheral blood at the endpoint. After biopsy, the subjects were treated with two doses of oral ciprofloxacin as previously described (26).
Immunohistochemical analysis. Banked sections of paraffin-embedded tissue were stained by immunohistochemistry exactly as described previously (20). Primary antibodies included anti-human E-selectin/CD62E, anti-mouse CCL27, anti-human CLA, anti-human CCL21 (R&D Systems, Minneapolis, MN), and anti-mouse PNAd (MECA-79, BD Pharmingen, San Diego, CA). Biotinylated secondary antibodies (donkey anti-goat immunoglobulin G; Jackson Immunoresearch Laboratories, West Grove, PA) were detected with streptavidin-horseradish peroxidase, followed by 3,3'-diaminobenzidine. Controls included uninfected skin; normal tonsil; omission of the primary antibody for E-selectin, CCL27, and CCL21; and an isotype control for PNAd and CLA. A dermatopathologist analyzed the samples and scored them as positive if at least three separate sections exhibited staining.
Immunophenotyping. Pustules from 17 subjects were biopsied after 5 to 9 days of infection. When multiple sites from one subject were biopsied, the cells were pooled. Biopsy samples were processed as described previously (7, 24). Briefly, 4- to 6-mm punch biopsy specimens were minced in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco-BRL, Rockville, MD) with 5 mM EDTA (Sigma, St. Louis, MO) and incubated with vigorous stirring for 4 to 6 h at 4°C. Single cells were collected using a 40-μm sieve. Peripheral blood mononuclear cells were isolated from peripheral blood by Ficoll-Hypaque density gradient centrifugation.
Immunostaining was performed with a five-step method as described previously (32). Antibodies used included CD45RA-phycoerythrin (PE)-Texas Red (clone 2H4; Beckman-Coulter, Fullerton, CA), CD27-PE (M-T271; Pharmingen), CD4-allophycocyanin (APC) (SK3; Pharmingen), CD4-APC-cyanine dye 7 (Cy7) (S3.5; CalTag Laboratories, Burlingame, CA), CD8-APC-Cy7 (3B5; CalTag), and CCR7 (150503; R&D Systems). CCR7 was detected with biotinylated goat anti-mouse immunoglobulin G (heavy and light chains; Jackson Immunoresearch Laboratories), followed by streptavidin-PE-Cy7 (Pharmingen). Data were acquired on a dual-laser MoFlo cytometer (Cytomation, Ft. Collins, CO) and analyzed using Summit 3.0 software (Cytomation). Approximately 100,000 peripheral blood CD4 cells, 40,000 peripheral blood CD8 cells, 7,634 ± 5,640 (mean ± standard deviation) lesional CD4 cells, and 3,971 ± 2,636 lesional CD8 cells were analyzed from each subject.
Statistical analysis. Statistical analysis was performed with pairwise comparisons. Using the Bonferroni adjustment to account for multiple comparisons, P values of <0.017 or 0.013 were considered significant when three or four groups were analyzed, respectively.
RESULTS
Expression of ligands involved in T-cell homing in H. ducreyi-infected skin. Using biopsy samples obtained 1 to 9 days after inoculation (Table 1), we examined H. ducreyi-infected skin for expression of ligands reported to be important in T-cell trafficking to the skin compartment (E-selectin) versus the lymph node compartment (PNAd). Immunostaining with anti-E-selectin antibody demonstrated marked expression of E-selectin in endothelial cells of all infected samples but not of uninfected skin (Fig. 1A and B and Table 2). Probing of samples with anti-PNAd antibody demonstrated occasional focal staining of endothelial cells in uninfected skin and samples infected with live bacteria (Fig. 1C and D and Table 2). PNAd staining generally occurred in only one to three vessels in the superficial vascular plexus per sample, as opposed to a diffuse expression of E-selectin in numerous vessels throughout the tissue. Endothelial cells in tonsil demonstrated strong PNAd staining (data not shown). Thus, E-selectin was upregulated within 24 h of infection, while PNAd expression was not altered by infection.
The CLA epitope is expressed by lymphocytes trafficking through skin and is associated with E-selectin binding. By immunohistochemistry, control samples contained few lymphocytes, most of which were CLA+. Infected samples contained many lymphocytes, most of which expressed CLA (Fig. 2 and Table 2). CLA+ lymphocytes were occasionally present in the epidermis. In the dermis, CLA+ lymphocytes were present near the epidermis, deeper in the dermis, and in perivascular areas. A few granulocytes in pustules were also CLA+, but CLA staining was less prominent on granulocytes than lymphocytes. Therefore, CLA was expressed by most lymphocytes and a few granulocytes that migrated to the dermis and epidermis in response to H. ducreyi infection. These data are consistent with our previous observations made by flow cytometry that approximately 75% of the CD4 T cells in lesions express CLA (32).
CCL27, a chemokine made by keratinocytes and endothelial cells of the superficial dermal plexus, recruits CCR10+ T cells into inflamed skin (18). We have shown previously that 10% of CD4 cells recruited to infected sites express CCR10 (32). Samples from uninfected skin demonstrated expression of CCL27 in the keratinocytes, endothelial cells, and lymphocytes (Fig. 1E). Examination of skin infected with H. ducreyi also showed positive staining in keratinocytes, endothelial cells, and the mononuclear cell infiltrate (Fig. 1F). CCL27 staining was uniformly positive throughout infected and uninfected sections and appeared unchanged in response to H. ducreyi infection in the skin.
CCL21, a chemokine expressed by endothelial cells in the lymph nodes, small intestine, and spleen, allows for trafficking of CCR7+ T cells to lymphoid tissues. Because it is a potent T-cell chemoattractant and we had previously found CCR7+ cells in infected skin (32), we examined H. ducreyi-infected skin for expression of CCL21. Endothelium in uninfected skin did not react with anti-CCL21 antibody (Fig. 1G). In all but one sample infected with live H. ducreyi, CCL21 was expressed by endothelial cells in the dermis (Fig. 1H and Table 2). In addition to endothelial cell staining, occasional mononuclear cells in the perivascular infiltrate expressed CCL21. Thus, CCL21 was upregulated during infection but not expressed in uninfected skin.
Expression of memory/effector markers by CD4 cells in infected skin. Previously, we examined the skin-homing phenotypes of the CD4 T-cell infiltrate in three pustules using the markers CD45RA, CLA, CCR7, CCR4, and CCR10 (32). Those experiments showed that the CD4+ T cells in infected sites are predominantly CD45RA– CLA+ CCR4+. A subset of the CCR4+ cells are also CCR10+. To better define the T-cell populations recruited to the skin, we analyzed pustules from 15 subjects. We initially examined expression of CD45RA and CD27 on CD4 cells. Lesional cells were compared to each subject's peripheral blood cells at the time of biopsy. Approximately 55% of the skin CD4 cells were CD45RA– CD27+ cells and 40% were CD45RA– CD27– cells, while <5% were CD45RA+ CD27+ (Fig. 3A). The CD45RA– CD27– population was significantly enriched relative to donor-matched peripheral blood cells (P < 0.0001).
Because CCL21 was upregulated in infected skin, we examined T cells isolated from lesions for expression of the CCL21 receptor, CCR7. We analyzed expression of CCR7 and CD27 on the CD45RA– subset in three subjects (Fig. 3B). Although CD45RA– CD27+ cells in both skin and blood were predominantly CCR7+, there were significantly fewer cells expressing CCR7 in the biopsy sample than in the blood (P = 0.003). The CD45RA– CD27– populations in both skin and blood were also predominantly CCR7+. However, in the biopsy samples, there were significantly more CCR7– cells than in blood (P = 0.013). Thus, CCR7+ cells are not precluded from accumulating in the skin, but there was no evidence that the CCL21-CCR7 pathway leads to enrichment of CCR7+ CD4 cells in the skin relative to peripheral blood.
Expression of memory/effector and skin-homing markers by CD8 cells in infected skin. We isolated CD8 T-cell populations from infected skin from 11 subjects and compared them to CD8 cells in donor-matched peripheral blood. We initially examined expression of CD45RA and CD27 and found that there were significantly fewer CD45RA+ CD27+ CD8 cells in the skin relative to peripheral blood (20% versus 50%, P < 0.0001) (Fig. 4A). The proportion of CD45RA– CD27+ and CD45RA+ CD27– cells in the skin was not significantly different than in peripheral blood, but the cutaneous CD8 cells were significantly enriched in CD45RA– CD27– cells relative to blood (40% versus 10%, P = 0.0006) (Fig. 4A).
We examined CCR7 expression on the CD45RA– subset of CD8 cells isolated from skin and compared them to peripheral blood CD8 cells from three donors. As expected, the majority of naive CD8 cells in peripheral blood expressed CCR7 (data not shown). There were few naive (CD45RA+ CD27+) CD8 cells in the skin, but those present were predominantly CCR7– (data not shown). Less than 1% of the CD45RA+ CD27– cells in both blood and skin expressed CCR7 (data not shown). The CD45RA– CD27+ and CD45RA– CD27– populations of CD8cells were not enriched in CCR7 expression in skin versus peripheral blood (P > 0.013, Fig. 4B). Thus, CCR7+ CD8 T cells are also not prohibited from entering the skin compartment, but there was no evidence that the CCL21-CCR7 pathway leads to enrichment of CCR7+ CD8 cells in the skin.
DISCUSSION
The CD4 and CD8 T cells recruited to pustules of experimental H. ducreyi infection consisted mostly of central memory and effector memory cells, with very few naive cells and effector (cytotoxic) CD8 cells relative to peripheral blood. Endothelial cells in infected sites expressed high levels of E-selectin and CCL21 relative to normal skin, and T cells recruited to infected sites expressed CLA. In contrast, CCL27 was expressed at similar levels in both infected and uninfected skin. Although CCL21 was expressed by the endothelium, there was no evidence that CCR7-expressing cells were enriched at infected sites compared to peripheral blood.
The use of multiple markers such as CD45RA, CD27, and CCR7 in flow cytometry is predictive of the functional subsets of T cells (10). Because the relationship between these markers and function is well established and due to limited cell numbers obtained from infected sites, we did not confirm function by intracellular cytokine staining. Using expression of CD45RA and CD27, we determined that the CD4 population in skin was primarily composed of memory and effector cells, and the latter cells were significantly enriched relative to peripheral blood. The majority of CD4+ CD27+ memory cells also expressed CCR7, a marker of central memory cells (28). Our findings are similar to those on erythema migrans, in which CD4 cells that migrate to infected skin consist predominantly of central memory cells and effector memory cells. The CD8 cells isolated from H. ducreyi-infected skin were significantly enriched in memory effector cells and depleted in naive cells relative to peripheral blood. These findings are also similar to those on erythema migrans, except that central memory CD8 cells are enriched in erythema migrans relative to peripheral blood (28).
In the human challenge model, papules develop at virtually all sites inoculated with estimated delivered doses in the range of 1 to 120 CFU (33). These papules either spontaneously resolve or evolve into pustules. There are host and gender effects on the outcomes for the sites (5, 34). Men are 2.8-fold (95% confidence interval, 1.6- to 5.1-fold) more likely than women to develop pustules (5) (data not shown). Experimental infection with H. ducreyi to the pustular stage does not confer protection against subsequent challenge (1). Subjects who initially are classified as resolvers or pustule formers and who are rechallenged tend to repetitively resolve infection at all sites or repetitively form pustules (34). In pustules, H. ducreyi is surrounded by polymorphonuclear leukocytes and macrophages, which fail to ingest the organism (3). However, there is no correlation in the ability of phagocytes isolated from resolvers and pustule formers to ingest the organism in vitro and clinical outcome (34). These data suggest that the overall composition of the lesion modulates the ability of phagocytes to clear the bacteria. The CD4 cells recovered from pustules in this study had memory or effector phenotypes, consistent with our ability to isolate H. ducreyi-specific CD4 T-cell lines from pustules (15). Interestingly, some of these lines have characteristics of regulatory T cells, which could promote phagocytic failure (34). The significance of the recruitment of memory effector CD8 cells to a site of an extracellular bacterial infection is unclear.
In H. ducreyi-infected skin, the majority of infiltrating T cells were CLA+ and there was upregulation of E-selectin on dermal endothelial cells. These data are consistent with earlier studies on CLA expression by skin-homing T cells (27) and by T cells recruited to sites of inflammation such as atopic dermatitis, contact dermatitis, psoriasis, and delayed-type hypersensitivity reactions (31, 32). Similarly, Koelle's group recently showed that antigen-specific CD8+ T cells specific for skin-tropic viruses (herpes simplex virus type 2 [HSV-2]) are CLA+ while CD8+ cells specific for non-skin-tropic viruses (Epstein-Barr virus and cytomegalovirus) are CLA– (23). A greater proportion of HSV-specific CD8 cells recovered from lesions express CLA relative to HSV-specific CD8 cells recovered from peripheral blood, and E-selectin is upregulated in HSV-2-infected tissue (23). E-selectin was initially thought to bind to the CLA epitope on T cells, but recent evidence emphasizes the role of fucosyltransferase VII and E-selectin binding ligands other than CLA as initial steps in the transmigration of T cells from blood vessels into skin (38-40). Nevertheless, CLA is a useful marker for expression of these E-selectin binding ligands.
By immunohistochemistry, CCL27 was expressed at similar levels in both uninfected and infected skin. In a previous study, we showed that about 10% of CD4 T cells in chancroidal lesions and in a delayed-type hypersensitivity response to dermal injection of Candida antigens express CCR10, the receptor for CCL27 (32). These data contrast with findings on psoriasis and atopic or allergic contact dermatitis, in which CCL27 expression is increased relative to normal skin and the majority of skin-infiltrating lymphocytes express CCR10 (18). The small percentage of cells expressing CCR10 and the lack of upregulation of CCL27 due to infection indicate that this ligand-receptor pair is not a major factor in T-cell homing in response to H. ducreyi. In experimental H. ducreyi infection and in delayed-type hypersensitivity responses, the bacteria and antigens are delivered primarily to the dermis and most of the T cells infiltrate the deep region of the dermis. In contrast, in psoriasis and contact dermatitis, most of the T cells infiltrate the upper dermis and epidermis. We speculate that CCL27 and CCR10 play a major role in trafficking of T cells to sites of epidermal inflammation rather than deep dermal inflammation.
The receptor-ligand pair CCR7-CCL21 is crucial for entry of T cells into lymphoid tissue but until recently has not been associated with trafficking to skin. CCR7 and CCL21 are also important in the emigration of Langerhans cells from skin to lymph nodes (14). We found that CCL21 was expressed in the dermal endothelium of infected skin but not in healthy skin. Similar results were found by Hromas and colleagues for atopic dermatitis, lichen planus, and graft-versus-host disease (8). In that study, CD45RO and CCR7 were expressed on T cells that infiltrated diseased skin, as demonstrated by immunohistopathology (8). While these data suggest that the CCL21-CCR7 pathway is operative in cutaneous T-cell homing, the number of CCR7-expressing cells in the skin was not compared to that in peripheral blood. Our data show quantitatively that there is no enrichment of CCR7+ CD4 or CD8 central memory or effector memory cells in infected skin relative to peripheral blood. Thus, CCR7+ cells are not precluded from entering the skin in response to H. ducreyi, but it is unlikely that the CCR7-CCL21 pathway plays a major role in selective enrichment of memory T cells. CCL21 may play a role in the migration of Langerhans cells out of skin and into draining lymph nodes.
Taking our present results together with those of our previous studies (32), we can propose a model outlining the receptor and ligand pairs that likely are most involved in homing of T cells to the dermis of H. ducreyi-infected skin. An initial loose association occurs between E-selectin on the endothelial cells and CLA+ T cells. The predominance of CCR4+ cells in lesions (32) suggests that CCR4-CCL17 mediates a subsequent interaction that allows migration of CCR4+ cells into the tissue. Unfortunately, our attempts to stain for CCL17 in tissue were technically unsuccessful. Although CCL21 is upregulated at infected sites, there is no enrichment of lesional CCR7+ cells relative to populations present in peripheral blood, and CCR7-CCL21 likely does not play a major role in T-cell recruitment. Similarly, the presence of CCL27 in both uninfected skin and infected skin, along with a minority population of CCR10+ cells in lesions, indicates that CCR10-CCL27 has a lesser role in trafficking of T cells in this model.
The combinations of PNAd-L-selectin and CCR7-CCL21 are known to be critical for trafficking of naive T cells to lymphoid tissue (29). Although CCR7- and CCL21-expressing cells were found in the skin, there was minimal expression of PNAd in normal and diseased skin. The fact that few naive T cells accumulated in the lesions emphasizes the importance of the combined role of PNAd-L-selectin and CCR7-CCL21 in homing of naive T cells to lymphoid tissue.
In summary, we determined that T cells that traffic to the dermis of H. ducreyi-infected skin likely utilize the CLA-E-selectin and CCR4-CCL17 pathways, while the CCR7-CCL21 and CCR10-CCL27 pathways play minor roles in this process. Our data indicate that the T cells that home to skin infected with H. ducreyi are predominantly central and effector memory cells, with an enrichment of effector memory cells relative to T cells present in peripheral blood. Studies to determine the effect that these T cells have on outcome in the model are ongoing in our laboratory.
ACKNOWLEDGMENTS
This study was funded by National Institutes of Health grants AI27863,AI31494, and AI46784. T.L.H. was supported by T32AI07367 from the National Institute of Allergy and Infectious Diseases. The human challenge trials were also supported by grant M01RR00750 to the General Clinical Research Center at Indiana University.
We thank Barry Katz for statistical analyses; Margaret Bauer for thoughtful criticism of the manuscript; Kate Fortney, Stacy Bennett, Marti Greenwald, and Beth Zwickl for assistance with the human challenge trials; and the volunteers who participated in the study.
REFERENCES
1. Al-Tawfiq, J. A., K. L. Palmer, C.-Y. Chen, J. C. Haley, B. P. Katz, A. F. Hood, and S. M. Spinola. 1999. Experimental infection of human volunteers with Haemophilus ducreyi does not confer protection against subsequent challenge. J. Infect. Dis. 179:1283-1287.
2. Al-Tawfiq, J. A., A. C. Thornton, B. P. Katz, K. R. Fortney, K. D. Todd, A. F. Hood, and S. M. Spinola. 1998. Standardization of the experimental model of Haemophilus ducreyi infection in human subjects. J. Infect. Dis. 178:1684-1687.
3. Bauer, M. E., M. P. Goheen, C. A. Townsend, and S. M. Spinola. 2001. Haemophilus ducreyi associates with phagocytes, collagen, and fibrin and remains extracellular throughout infection of human volunteers. Infect. Immun. 69:2549-2557.
4. Bong, C. T. H., M. E. Bauer, and S. M. Spinola. 2002. Haemophilus ducreyi: clinical features, epidemiology, and prospects for disease control. Microbes Infect. 4:1141-1148.
5. Bong, C. T. H., J. Harezlak, B. P. Katz, and S. M. Spinola. 2002. Men are more susceptible to pustule formation than women in the experimental model of Haemophilus ducreyi infection. Sex. Transm. Dis. 29:114-118.
6. Campbell, J. J., G. Haraldsen, J. Pan, J. Rottman, S. Qin, P. Ponath, D. P. Andrew, R. Warnke, N. Ruffing, N. Kassam, L. Wu, and E. C. Butcher. 1999. The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells. Nature 400:776-780.
7. Campbell, J. J., K. E. Murphy, E. J. Kunkel, C. E. Brightling, D. Soler, Z. Shen, J. Boisvert, H. B. Greenberg, M. A. Vierra, S. B. Goodman, M. C. Genovese, A. J. Wardlaw, E. C. Butcher, and L. Wu. 2001. CCR7 expression and memory T cell diversity in humans. J. Immunol. 166:877-884.
8. Christopherson, K. W., II, A. F. Hood, J. B. Travers, H. Ramsey, and R. A. Hromas. 2003. Endothelial induction of the T-cell chemokine CCL21 in T-cell autoimmune diseases. Blood 101:801-806.
9. DeJong, R., M. Brouwer, B. Hooibrink, T. Van der Pouw-Kraan, F. Miedema, and R. A. W. Van Lier. 1992. The CD27– subset of peripheral blood memory CD4+ lymphocytes contains functionally differentiated T lymphocytes that develop by persistent antigenic stimulation in vivo. Eur. J. Immunol. 22:993-999.
10. De Rosa, S. C., L. A. Herzenberg, L. A. Herzenberg, and M. Roederer. 2001. 11-color, 13-parameter flow cytometry: identification of human naive T cells by phenotype, function, and T-cell receptor diversity. Nat. Med. 7:245-248.
11. Duijvestijn, A. M., E. Horst, S. T. Pals, B. N. Rouse, A. C. Steere, L. J. Picker, C. J. L. M. Meijer, and E. C. Butcher. 1988. High endothelial differentiation in human lymphoid and inflammatory tissues defined by monoclonal antibody HECA-452. Am. J. Pathol. 130:147-155.
12. Faint, J. M., N. E. Annels, S. J. Curnow, P. Shields, D. Pilling, A. D. Hislop, L. Wu, A. N. Akbar, C. D. Buckley, P. A. H. Moss, D. H. Adams, A. B. Rickinson, and M. Salmon. 2001. Memory T cells constitute a subset of the human CD8+CD45RA+ pool with distinct phenotypic and migratory characteristics. J. Immunol. 167:212-220.
13. Fukada, K., Y. Sobao, H. Tomiyama, S. Oka, and T. Masafumi. 2002. Functional expression of the chemokine receptor CCR5 on virus epitope-specific memory and effector CD8+ T cells. J. Immunol. 168:2225-2232.
14. Geissmann, F., M. C. Dieu-Nosjean, C. Dezutter, J. Valladeau, S. Kayal, M. Leborgne, N. Brousse, S. Saeland, and J. Davoust. 2002. Accumulation of immature Langerhans cells in human lymph nodes draining chronically inflamed skin. J. Exp. Med. 196:417-430.
15. Gelfanova, V., T. L. Humphreys, and S. M. Spinola. 2001. Characterization of Haemophilus ducreyi-specific T-cell lines from lesions of experimentally infected human subjects. Infect. Immun. 69:4224-4231.
16. Gunn, M. D., K. Tangemann, C. Tam, J. G. Cyster, S. D. Rosen, and L. T. Williams. 1998. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc. Natl. Acad. Sci. USA 95:258-263.
17. Hamann, D., P. A. Baars, M. H. G. Rep, B. Hooibrink, S. R. Kerkhof-Garde, M. R. Klein, and R. A. W. Van Lier. 1997. Phenotypic and functional separation of memory and effector human CD8+ T cells. J. Exp. Med. 186(9):1407-1418.
18. Homey, B., H. Alenius, A. Muller, H. Soto, E. P. Bowman, W. Yuan, L. McEvoy, A. I. Lauerma, T. Assmann, E. Bunemann, M. Lehto, H. Wolff, D. Yen, H. Marxhausen, W. To, J. Sedgwick, T. Ruzicka, P. Lehmann, and A. Zlotnik. 2002. CCL27-CCR10 interactions regulate T cell-mediated skin inflammation. Nat. Med. 8:157-165.
19. Hudak, S., M. Hagen, Y. Liu, D. Catron, E. Oldham, L. M. McEvoy, and E. P. Bowman. 2002. Immune surveillance and effector functions of CCR10+ skin homing T cells. J. Immunol. 169:1189-1196.
20. Humphreys, T. L., C. T. Schnizlein-Bick, B. P. Katz, L. A. Baldridge, A. F. Hood, R. A. Hromas, and S. M. Spinola. 2002. Evolution of the cutaneous immune response to experimental Haemophilus ducreyi infection and its relevance to HIV-1 acquisition. J. Immunol. 169:6316-6323.
21. Kelly, K. A. 1997. Modulation of leukocyte-endothelial cell interactions by infectious agents. Bull. Inst. Pasteur 95:147-159.
22. King, R., S. H. Choudhri, J. Nasio, J. Gough, N. J. D. Nagelkerke, F. A. Plummer, J. O. Ndinya-Achola, and A. R. Ronald. 1998. Clinical and in situ cellular responses to Haemophilus ducreyi in the presence or absence of HIV infection. Int. J. STD AIDS 9:531-536.
23. Koelle, D. M., Z. Liu, C. M. McClurkan, M. S. Topp, S. R. Riddell, E. G. Pamer, A. S. Johnson, A. Wald, and L. Corey. 2002. Expression of cutaneous lymphocyte-associated antigen by CD8+ T cells specific for a skin-tropic virus. J. Clin. Investig. 110:537-548.
24. Kunkel, E. J., J. J. Campbell, G. Haraldsen, J. Pan, J. Boisvert, A. I. Roberts, E. C. Elbert, M. A. Vierra, S. B. Goodman, M. C. Genovese, A. J. Wardlaw, H. B. Greenberg, C. M. Parker, E. C. Butcher, D. P. Andrew, and W. W. Agace. 2000. Lymphocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine (TECK) expression distinguish the small intestinal immune compartment: epithelial expression of tissue-specific chemokines as an organizing principle in regional immunity. J. Exp. Med. 192:761-767.
25. Magro, C. M., A. N. Crowson, M. Alfa, A. Nath, A. Ronald, J. O. Ndinya-Achola, and J. Nasio. 1996. A morphological study of penile chancroid lesions in human immunodeficiency virus (HIV)-positive and -negative African men with a hypothesis concerning the role of chancroid in HIV transmission. Hum. Pathol. 27:1066-1070.
26. Palmer, K. L., C. T. Schnizlein-Bick, A. Orazi, K. John, C.-Y. Chen, A. F. Hood, and S. M. Spinola. 1998. The immune response to Haemophilus ducreyi resembles a delayed-type hypersensitivity reaction throughout experimental infection of human subjects. J. Infect. Dis. 178:1688-1697.
27. Picker, L. J., S. A. Michie, L. S. Rott, and E. C. Butcher. 1990. A unique phenotype of skin-associated lymphocytes in humans. J. Pathol. 136:1053-1068.
28. Salazar, J. C., C. D. Pope, T. J. Sellati, H. M. Feder, Jr., T. G. Kiely, K. R. Dardick, R. L. Buckman, M. W. Moore, M. J. Caimano, J. G. Pope, P. J. Krause, and J. D. Radolf. 2003. Coevolution of markers of innate and adaptive immunity in skin and peripheral blood of patients with erythema migrans. J. Immunol. 171:2660-2670.
29. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708-712.
30. Sallusto, F., D. Lenig, C. R. Mackay, and A. Lanzavecchia. 1998. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J. Exp. Med. 187:875-883.
31. Santamaria-Babi, L. F. 2004. CLA+ T cells in cutaneous diseases. Eur. J. Dermatol. 14:13-18.
32. Soler, D., T. L. Humphreys, S. M. Spinola, and J. J. Campbell. 2003. CCR4 versus CCR10 in human cutaneous Th lymphocyte trafficking. Blood 101:1677-1683.
33. Spinola, S. M., M. E. Bauer, and R. S. Munson, Jr. 2002. Immunopathogenesis of Haemophilus ducreyi infection (chancroid). Infect. Immun. 70:1667-1676.
34. Spinola, S. M., C. T. H. Bong, A. L. Faber, K. R. Fortney, S. L. Bennett, C. A. Townsend, B. E. Zwickl, S. D. Billings, T. L. Humphreys, M. E. Bauer, and B. P. Katz. 2003. Differences in host susceptibility to disease progression in the human challenge model of Haemophilus ducreyi infection. Infect. Immun. 71:6658-6663.
35. Spinola, S. M., K. R. Fortney, B. P. Katz, J. L. Latimer, J. R. Mock, M. Vakevainen, and E. J. Hansen. 2003. Haemophilus ducreyi requires an intact flp gene cluster for virulence in humans. Infect. Immun. 71:7178-7182.
36. Spinola, S. M., A. Orazi, J. N. Arno, K. Fortney, P. Kotylo, C.-Y. Chen, A. A. Campagnari, and A. F. Hood. 1996. Haemophilus ducreyi elicits a cutaneous infiltrate of CD4 cells during experimental human infection. J. Infect. Dis. 173:394-402.
37. Spinola, S. M., L. M. Wild, M. A. Apicella, A. A. Gaspari, and A. A. Campagnari. 1994. Experimental human infection with Haemophilus ducreyi. J. Infect. Dis. 169:1146-1150.
38. Takahashi, R., Y. Mizukawa, Y. Yamazaki, K. Hayakawa, J. Hayakawa, A. Kudo, and T. Shiohara. 2003. In vitro differentiation from naive to mature E-selectin binding CD4 T cells: acquisition of skin-homing properties occurs independently of cutaneous lymphocyte antigen expression. J. Immunol. 171:5769-5777.
39. Wagers, A. J., J. B. Lowe, and G. S. Kansas. 1996. An important role for the a1,3 fucosyltransferase, FucT-VII, in leukocyte adhesion to E-selectin. Blood 88:2125-2132.
40. Wagers, A. J., L. M. Stoolman, R. Kannagi, R. Craig, and G. S. Kansas. 1997. Expression of leukocyte fucosyltransferases regulates binding to E-selectin. J. Immunol. 159:1917-1929.(Tricia L. Humphreys, Lee )
Department of Pathology, Harvard Medical School
Joint Program in Transfusion Medicine, Children's Hospital, Boston, Massachusetts
ABSTRACT
T-cell homing to infected skin is not well studied in humans. We examined sites experimentally infected with Haemophilus ducreyi by immunohistochemistry and flow cytometry for expression of receptors and ligands involved in cutaneous T-cell homing and determined the phenotypes of the T cells that trafficked to skin. Endothelial cells expressed E-selectin in infected but not uninfected skin, while peripheral node addressin (PNAd) was minimally expressed in all samples. CC chemokine ligand 27 (CCL27) was expressed in the epidermis and endothelium of both infected and uninfected skin. Interestingly, CCL21, a chemokine thought to be associated principally with T-cell trafficking in the lymphatic compartment, was highly expressed on the endothelium of infected skin. Few naive cells were present in experimental lesions, emphasizing the combined role of PNAd and CCL21 in trafficking of this subset. Memory cells (CD45RA–) dominated both CD4 and CD8 T-cell populations at the site of infection. Effector memory (CD45RA– CD27–) CD4+ and CD8+ T cells were enriched in lesions. Although the CC chemokine receptor 7-positive (CCR7+) population of both central memory (CD45RA– CD27+) and effector memory cells was not enriched in the skin compared to peripheral blood, CCR7+ cells were not precluded from entering infected skin. Taken together with our previous work (D. Soler, T. L. Humphreys, S. M. Spinola, and J. J. Campbell, Blood 101:1677-1683, 2003), these studies led us to propose a model of memory T-cell trafficking to skin in response to experimental H. ducreyi infection.
INTRODUCTION
T cells move out of peripheral blood into tissues by a multistep process: an initial loose tethering mediated by selectins, followed by a tighter interaction mediated by chemokines, and then a firm adhesion mediated by integrins, finally resulting in diapedesis (21). Differential expression of ligands and their receptors allows for tissue-specific migration of T cells (30). For homing to skin, cutaneous lymphocyte antigen (CLA) and the CC chemokine receptor 4 (CCR4) and CCR10 expressed by T cells bind to their ligands E-selectin, CC chemokine ligand 17 (CCL17) (thymus- and activation-regulated chemokine), and CCL27 (cutaneous T-cell-attracting chemokine), respectively, on endothelial cells (6, 11, 18, 32). In contrast, T cells that enter secondary lymphoid tissues express L-selectin and CCR7 in order to bind to their respective ligands peripheral node addressin (PNAd) and CCL21 (6-cysteine chemokine, 6Ckine) on endothelial cells (29). CCL21 is also expressed outside of secondary lymphoid organs in humans, including in the small intestine, thyroid, pancreas, trachea, and salivary glands (16). Recently, CCL21 was found to be expressed by human dermal endothelial cells during T-cell autoimmune diseases such as atopic dermatitis, lichen planus, and graft-versus-host diseases. Lymphocytes infiltrating these lesions expressed CCR7 (8), suggesting that the CCR7-CCL21 interaction is also operative in T-cell homing to skin. Whether expression of this receptor-ligand pair occurs in response to a bacterial skin infection has not been studied.
In addition to expression of receptors that allow for tissue-specific homing, migratory T cells are also defined as naive, memory, or effector cells by combinations of surface markers, including but not limited to the CD45 isotypes RA and RO, the costimulatory molecule CD27, L-selectin, and chemokine receptors such as CCR7, CXCR3, CXCR4, or CCR5 (9, 17, 29, 30). For example, CD4 populations may be classified as naive (CD45RA+ CD27+ CCR7+), central memory (CD45RA– CD27+ CCR7+), and effector memory (CD45RA– CD27– CCR7–) (9, 29). Similarly, CD8 cells may be defined as naive (CD45RA+ CD27+ CCR7+), central memory (CD45RA– CD27+ CCR7+), effector memory (CD45RA– CD27– CCR7–),and effector (CD45RA+ CD27– CCR7–) (12, 13, 29). Functionally, naive and central memory CD4 cells are similar, capable of producing only interleukin-2 (IL-2) upon in vitro stimulation, whereas effector memory populations produce IL-4, IL-5, IFN-, and some IL-2. Among CD8 T cells, IFN- and perforin are produced only by the effector memory and effector populations.
Haemophilus ducreyi is an obligate human pathogen and the causative agent of the genital ulcer disease chancroid, which facilitates transmission of human immunodeficiency virus type 1 (4). To study the human immune response to H. ducreyi, we developed an experimental inoculation model with human volunteers (2, 33, 37). In the model, subjects are inoculated at multiple sites on the upper arm with an allergy testing device, which delivers the bacteria to both the epidermis and the dermis. Papules develop at inoculated sites, which evolve into pustules, simulating natural chancroid. The histopathology of experimental infection closely parallels that of naturally occurring disease (22, 25, 26). The lesion consists of a polymorphonuclear leukocyte infiltrate that forms an epidermal abscess and a deep dermal perivascular infiltrate of T cells and macrophages that resembles a poorly formed granuloma that extends interstitially to just below the epidermis (26, 36). By 48 h of infection, T cells make up about 50% of the leukocyte infiltrate and remain abundant throughout the course of experimental infection (20).
Most cutaneous homing studies are performed in the context of noninfectious inflammatory skin diseases (8, 18) or by using CLA+ subsets in peripheral blood as a surrogate system to study cells with skin-homing potential (19, 29, 38). There is a limited amount of information about T cells that home to skin in response to infection (23, 28, 32). Human inoculation experiments with H. ducreyi provide a unique opportunity to study the trafficking and phenotypes of T cells recruited chiefly to the dermis in response to this pathogen (20, 32). Previous work in our laboratory showed that CD4 cells isolated from H. ducreyi-infected pustules are memory cells that are enriched for the skin-homing markers CLA and CCR4. A smaller proportion express the cutaneous marker CCR10, but the CCR10+ cells are not significantly enriched compared to peripheral blood (32). In addition, about half the CD4 T cells in H. ducreyi-infected skin specimens are CD45RA– CD27+ CCR7+ central memory cells, but it was not clear whether these cells were enriched relative to peripheral blood (32).
In this study, we further defined the subsets of CD4 T cells present in lesions and for the first time examined the homing and memory/effector phenotypes of CD8 T cells that traffic to infected sites. We also examined longitudinally obtained biopsy specimens for expression of the skin-homing ligands E-selectin and CCL27 and compared them to expression of the lymph node-homing ligands PNAd and CCL21. Taken together with our previous work, these studies allowed us to define the trafficking pathways and phenotypes of T cells recruited in response to experimental H. ducreyi infection in the skin.
MATERIALS AND METHODS
Human subjects. Twenty-five healthy adult volunteers contributed 43 tissue specimens for this study (Table 1). Biopsy samples designated CS or Control were obtained from seven subjects who participated in previous human challenge trials and one subject who donated uninfected skin (2, 26). Tissue specimens for flow cytometry analysis were obtained from six volunteers who were infected for the purposes of this study, six subjects who participated in a mutant-parent trial (35), and five subjects who participated in a reinfection trial (34) (Table 1). Informed consent was obtained from the subjects in accordance with the guidelines for human experimentation of the U.S. Department of Health and Human Services and the Institutional Review Board of Indiana University-Purdue University at Indianapolis.
The enrollment procedures, exclusion criteria, preparation of the bacteria, and inoculation procedures for the human challenge experiments are described in detail elsewhere (2, 36, 37). Each volunteer was inoculated at two or three sites on the upper arm with live H. ducreyi 35000HP (a human-passaged isolate of strain 35000) and at one site with heat-killed 35000HP. Each subject donated up to three biopsy samples (Table 1). For immunohistochemical analysis, several subjects were assigned to biopsy 1 to 4 days after inoculation, while others were biopsied at the clinical endpoint, defined as the development of a painful pustule or a maximum of 14 days of infection. For flow cytometry analysis, the subjects donated tissue and peripheral blood at the endpoint. After biopsy, the subjects were treated with two doses of oral ciprofloxacin as previously described (26).
Immunohistochemical analysis. Banked sections of paraffin-embedded tissue were stained by immunohistochemistry exactly as described previously (20). Primary antibodies included anti-human E-selectin/CD62E, anti-mouse CCL27, anti-human CLA, anti-human CCL21 (R&D Systems, Minneapolis, MN), and anti-mouse PNAd (MECA-79, BD Pharmingen, San Diego, CA). Biotinylated secondary antibodies (donkey anti-goat immunoglobulin G; Jackson Immunoresearch Laboratories, West Grove, PA) were detected with streptavidin-horseradish peroxidase, followed by 3,3'-diaminobenzidine. Controls included uninfected skin; normal tonsil; omission of the primary antibody for E-selectin, CCL27, and CCL21; and an isotype control for PNAd and CLA. A dermatopathologist analyzed the samples and scored them as positive if at least three separate sections exhibited staining.
Immunophenotyping. Pustules from 17 subjects were biopsied after 5 to 9 days of infection. When multiple sites from one subject were biopsied, the cells were pooled. Biopsy samples were processed as described previously (7, 24). Briefly, 4- to 6-mm punch biopsy specimens were minced in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco-BRL, Rockville, MD) with 5 mM EDTA (Sigma, St. Louis, MO) and incubated with vigorous stirring for 4 to 6 h at 4°C. Single cells were collected using a 40-μm sieve. Peripheral blood mononuclear cells were isolated from peripheral blood by Ficoll-Hypaque density gradient centrifugation.
Immunostaining was performed with a five-step method as described previously (32). Antibodies used included CD45RA-phycoerythrin (PE)-Texas Red (clone 2H4; Beckman-Coulter, Fullerton, CA), CD27-PE (M-T271; Pharmingen), CD4-allophycocyanin (APC) (SK3; Pharmingen), CD4-APC-cyanine dye 7 (Cy7) (S3.5; CalTag Laboratories, Burlingame, CA), CD8-APC-Cy7 (3B5; CalTag), and CCR7 (150503; R&D Systems). CCR7 was detected with biotinylated goat anti-mouse immunoglobulin G (heavy and light chains; Jackson Immunoresearch Laboratories), followed by streptavidin-PE-Cy7 (Pharmingen). Data were acquired on a dual-laser MoFlo cytometer (Cytomation, Ft. Collins, CO) and analyzed using Summit 3.0 software (Cytomation). Approximately 100,000 peripheral blood CD4 cells, 40,000 peripheral blood CD8 cells, 7,634 ± 5,640 (mean ± standard deviation) lesional CD4 cells, and 3,971 ± 2,636 lesional CD8 cells were analyzed from each subject.
Statistical analysis. Statistical analysis was performed with pairwise comparisons. Using the Bonferroni adjustment to account for multiple comparisons, P values of <0.017 or 0.013 were considered significant when three or four groups were analyzed, respectively.
RESULTS
Expression of ligands involved in T-cell homing in H. ducreyi-infected skin. Using biopsy samples obtained 1 to 9 days after inoculation (Table 1), we examined H. ducreyi-infected skin for expression of ligands reported to be important in T-cell trafficking to the skin compartment (E-selectin) versus the lymph node compartment (PNAd). Immunostaining with anti-E-selectin antibody demonstrated marked expression of E-selectin in endothelial cells of all infected samples but not of uninfected skin (Fig. 1A and B and Table 2). Probing of samples with anti-PNAd antibody demonstrated occasional focal staining of endothelial cells in uninfected skin and samples infected with live bacteria (Fig. 1C and D and Table 2). PNAd staining generally occurred in only one to three vessels in the superficial vascular plexus per sample, as opposed to a diffuse expression of E-selectin in numerous vessels throughout the tissue. Endothelial cells in tonsil demonstrated strong PNAd staining (data not shown). Thus, E-selectin was upregulated within 24 h of infection, while PNAd expression was not altered by infection.
The CLA epitope is expressed by lymphocytes trafficking through skin and is associated with E-selectin binding. By immunohistochemistry, control samples contained few lymphocytes, most of which were CLA+. Infected samples contained many lymphocytes, most of which expressed CLA (Fig. 2 and Table 2). CLA+ lymphocytes were occasionally present in the epidermis. In the dermis, CLA+ lymphocytes were present near the epidermis, deeper in the dermis, and in perivascular areas. A few granulocytes in pustules were also CLA+, but CLA staining was less prominent on granulocytes than lymphocytes. Therefore, CLA was expressed by most lymphocytes and a few granulocytes that migrated to the dermis and epidermis in response to H. ducreyi infection. These data are consistent with our previous observations made by flow cytometry that approximately 75% of the CD4 T cells in lesions express CLA (32).
CCL27, a chemokine made by keratinocytes and endothelial cells of the superficial dermal plexus, recruits CCR10+ T cells into inflamed skin (18). We have shown previously that 10% of CD4 cells recruited to infected sites express CCR10 (32). Samples from uninfected skin demonstrated expression of CCL27 in the keratinocytes, endothelial cells, and lymphocytes (Fig. 1E). Examination of skin infected with H. ducreyi also showed positive staining in keratinocytes, endothelial cells, and the mononuclear cell infiltrate (Fig. 1F). CCL27 staining was uniformly positive throughout infected and uninfected sections and appeared unchanged in response to H. ducreyi infection in the skin.
CCL21, a chemokine expressed by endothelial cells in the lymph nodes, small intestine, and spleen, allows for trafficking of CCR7+ T cells to lymphoid tissues. Because it is a potent T-cell chemoattractant and we had previously found CCR7+ cells in infected skin (32), we examined H. ducreyi-infected skin for expression of CCL21. Endothelium in uninfected skin did not react with anti-CCL21 antibody (Fig. 1G). In all but one sample infected with live H. ducreyi, CCL21 was expressed by endothelial cells in the dermis (Fig. 1H and Table 2). In addition to endothelial cell staining, occasional mononuclear cells in the perivascular infiltrate expressed CCL21. Thus, CCL21 was upregulated during infection but not expressed in uninfected skin.
Expression of memory/effector markers by CD4 cells in infected skin. Previously, we examined the skin-homing phenotypes of the CD4 T-cell infiltrate in three pustules using the markers CD45RA, CLA, CCR7, CCR4, and CCR10 (32). Those experiments showed that the CD4+ T cells in infected sites are predominantly CD45RA– CLA+ CCR4+. A subset of the CCR4+ cells are also CCR10+. To better define the T-cell populations recruited to the skin, we analyzed pustules from 15 subjects. We initially examined expression of CD45RA and CD27 on CD4 cells. Lesional cells were compared to each subject's peripheral blood cells at the time of biopsy. Approximately 55% of the skin CD4 cells were CD45RA– CD27+ cells and 40% were CD45RA– CD27– cells, while <5% were CD45RA+ CD27+ (Fig. 3A). The CD45RA– CD27– population was significantly enriched relative to donor-matched peripheral blood cells (P < 0.0001).
Because CCL21 was upregulated in infected skin, we examined T cells isolated from lesions for expression of the CCL21 receptor, CCR7. We analyzed expression of CCR7 and CD27 on the CD45RA– subset in three subjects (Fig. 3B). Although CD45RA– CD27+ cells in both skin and blood were predominantly CCR7+, there were significantly fewer cells expressing CCR7 in the biopsy sample than in the blood (P = 0.003). The CD45RA– CD27– populations in both skin and blood were also predominantly CCR7+. However, in the biopsy samples, there were significantly more CCR7– cells than in blood (P = 0.013). Thus, CCR7+ cells are not precluded from accumulating in the skin, but there was no evidence that the CCL21-CCR7 pathway leads to enrichment of CCR7+ CD4 cells in the skin relative to peripheral blood.
Expression of memory/effector and skin-homing markers by CD8 cells in infected skin. We isolated CD8 T-cell populations from infected skin from 11 subjects and compared them to CD8 cells in donor-matched peripheral blood. We initially examined expression of CD45RA and CD27 and found that there were significantly fewer CD45RA+ CD27+ CD8 cells in the skin relative to peripheral blood (20% versus 50%, P < 0.0001) (Fig. 4A). The proportion of CD45RA– CD27+ and CD45RA+ CD27– cells in the skin was not significantly different than in peripheral blood, but the cutaneous CD8 cells were significantly enriched in CD45RA– CD27– cells relative to blood (40% versus 10%, P = 0.0006) (Fig. 4A).
We examined CCR7 expression on the CD45RA– subset of CD8 cells isolated from skin and compared them to peripheral blood CD8 cells from three donors. As expected, the majority of naive CD8 cells in peripheral blood expressed CCR7 (data not shown). There were few naive (CD45RA+ CD27+) CD8 cells in the skin, but those present were predominantly CCR7– (data not shown). Less than 1% of the CD45RA+ CD27– cells in both blood and skin expressed CCR7 (data not shown). The CD45RA– CD27+ and CD45RA– CD27– populations of CD8cells were not enriched in CCR7 expression in skin versus peripheral blood (P > 0.013, Fig. 4B). Thus, CCR7+ CD8 T cells are also not prohibited from entering the skin compartment, but there was no evidence that the CCL21-CCR7 pathway leads to enrichment of CCR7+ CD8 cells in the skin.
DISCUSSION
The CD4 and CD8 T cells recruited to pustules of experimental H. ducreyi infection consisted mostly of central memory and effector memory cells, with very few naive cells and effector (cytotoxic) CD8 cells relative to peripheral blood. Endothelial cells in infected sites expressed high levels of E-selectin and CCL21 relative to normal skin, and T cells recruited to infected sites expressed CLA. In contrast, CCL27 was expressed at similar levels in both infected and uninfected skin. Although CCL21 was expressed by the endothelium, there was no evidence that CCR7-expressing cells were enriched at infected sites compared to peripheral blood.
The use of multiple markers such as CD45RA, CD27, and CCR7 in flow cytometry is predictive of the functional subsets of T cells (10). Because the relationship between these markers and function is well established and due to limited cell numbers obtained from infected sites, we did not confirm function by intracellular cytokine staining. Using expression of CD45RA and CD27, we determined that the CD4 population in skin was primarily composed of memory and effector cells, and the latter cells were significantly enriched relative to peripheral blood. The majority of CD4+ CD27+ memory cells also expressed CCR7, a marker of central memory cells (28). Our findings are similar to those on erythema migrans, in which CD4 cells that migrate to infected skin consist predominantly of central memory cells and effector memory cells. The CD8 cells isolated from H. ducreyi-infected skin were significantly enriched in memory effector cells and depleted in naive cells relative to peripheral blood. These findings are also similar to those on erythema migrans, except that central memory CD8 cells are enriched in erythema migrans relative to peripheral blood (28).
In the human challenge model, papules develop at virtually all sites inoculated with estimated delivered doses in the range of 1 to 120 CFU (33). These papules either spontaneously resolve or evolve into pustules. There are host and gender effects on the outcomes for the sites (5, 34). Men are 2.8-fold (95% confidence interval, 1.6- to 5.1-fold) more likely than women to develop pustules (5) (data not shown). Experimental infection with H. ducreyi to the pustular stage does not confer protection against subsequent challenge (1). Subjects who initially are classified as resolvers or pustule formers and who are rechallenged tend to repetitively resolve infection at all sites or repetitively form pustules (34). In pustules, H. ducreyi is surrounded by polymorphonuclear leukocytes and macrophages, which fail to ingest the organism (3). However, there is no correlation in the ability of phagocytes isolated from resolvers and pustule formers to ingest the organism in vitro and clinical outcome (34). These data suggest that the overall composition of the lesion modulates the ability of phagocytes to clear the bacteria. The CD4 cells recovered from pustules in this study had memory or effector phenotypes, consistent with our ability to isolate H. ducreyi-specific CD4 T-cell lines from pustules (15). Interestingly, some of these lines have characteristics of regulatory T cells, which could promote phagocytic failure (34). The significance of the recruitment of memory effector CD8 cells to a site of an extracellular bacterial infection is unclear.
In H. ducreyi-infected skin, the majority of infiltrating T cells were CLA+ and there was upregulation of E-selectin on dermal endothelial cells. These data are consistent with earlier studies on CLA expression by skin-homing T cells (27) and by T cells recruited to sites of inflammation such as atopic dermatitis, contact dermatitis, psoriasis, and delayed-type hypersensitivity reactions (31, 32). Similarly, Koelle's group recently showed that antigen-specific CD8+ T cells specific for skin-tropic viruses (herpes simplex virus type 2 [HSV-2]) are CLA+ while CD8+ cells specific for non-skin-tropic viruses (Epstein-Barr virus and cytomegalovirus) are CLA– (23). A greater proportion of HSV-specific CD8 cells recovered from lesions express CLA relative to HSV-specific CD8 cells recovered from peripheral blood, and E-selectin is upregulated in HSV-2-infected tissue (23). E-selectin was initially thought to bind to the CLA epitope on T cells, but recent evidence emphasizes the role of fucosyltransferase VII and E-selectin binding ligands other than CLA as initial steps in the transmigration of T cells from blood vessels into skin (38-40). Nevertheless, CLA is a useful marker for expression of these E-selectin binding ligands.
By immunohistochemistry, CCL27 was expressed at similar levels in both uninfected and infected skin. In a previous study, we showed that about 10% of CD4 T cells in chancroidal lesions and in a delayed-type hypersensitivity response to dermal injection of Candida antigens express CCR10, the receptor for CCL27 (32). These data contrast with findings on psoriasis and atopic or allergic contact dermatitis, in which CCL27 expression is increased relative to normal skin and the majority of skin-infiltrating lymphocytes express CCR10 (18). The small percentage of cells expressing CCR10 and the lack of upregulation of CCL27 due to infection indicate that this ligand-receptor pair is not a major factor in T-cell homing in response to H. ducreyi. In experimental H. ducreyi infection and in delayed-type hypersensitivity responses, the bacteria and antigens are delivered primarily to the dermis and most of the T cells infiltrate the deep region of the dermis. In contrast, in psoriasis and contact dermatitis, most of the T cells infiltrate the upper dermis and epidermis. We speculate that CCL27 and CCR10 play a major role in trafficking of T cells to sites of epidermal inflammation rather than deep dermal inflammation.
The receptor-ligand pair CCR7-CCL21 is crucial for entry of T cells into lymphoid tissue but until recently has not been associated with trafficking to skin. CCR7 and CCL21 are also important in the emigration of Langerhans cells from skin to lymph nodes (14). We found that CCL21 was expressed in the dermal endothelium of infected skin but not in healthy skin. Similar results were found by Hromas and colleagues for atopic dermatitis, lichen planus, and graft-versus-host disease (8). In that study, CD45RO and CCR7 were expressed on T cells that infiltrated diseased skin, as demonstrated by immunohistopathology (8). While these data suggest that the CCL21-CCR7 pathway is operative in cutaneous T-cell homing, the number of CCR7-expressing cells in the skin was not compared to that in peripheral blood. Our data show quantitatively that there is no enrichment of CCR7+ CD4 or CD8 central memory or effector memory cells in infected skin relative to peripheral blood. Thus, CCR7+ cells are not precluded from entering the skin in response to H. ducreyi, but it is unlikely that the CCR7-CCL21 pathway plays a major role in selective enrichment of memory T cells. CCL21 may play a role in the migration of Langerhans cells out of skin and into draining lymph nodes.
Taking our present results together with those of our previous studies (32), we can propose a model outlining the receptor and ligand pairs that likely are most involved in homing of T cells to the dermis of H. ducreyi-infected skin. An initial loose association occurs between E-selectin on the endothelial cells and CLA+ T cells. The predominance of CCR4+ cells in lesions (32) suggests that CCR4-CCL17 mediates a subsequent interaction that allows migration of CCR4+ cells into the tissue. Unfortunately, our attempts to stain for CCL17 in tissue were technically unsuccessful. Although CCL21 is upregulated at infected sites, there is no enrichment of lesional CCR7+ cells relative to populations present in peripheral blood, and CCR7-CCL21 likely does not play a major role in T-cell recruitment. Similarly, the presence of CCL27 in both uninfected skin and infected skin, along with a minority population of CCR10+ cells in lesions, indicates that CCR10-CCL27 has a lesser role in trafficking of T cells in this model.
The combinations of PNAd-L-selectin and CCR7-CCL21 are known to be critical for trafficking of naive T cells to lymphoid tissue (29). Although CCR7- and CCL21-expressing cells were found in the skin, there was minimal expression of PNAd in normal and diseased skin. The fact that few naive T cells accumulated in the lesions emphasizes the importance of the combined role of PNAd-L-selectin and CCR7-CCL21 in homing of naive T cells to lymphoid tissue.
In summary, we determined that T cells that traffic to the dermis of H. ducreyi-infected skin likely utilize the CLA-E-selectin and CCR4-CCL17 pathways, while the CCR7-CCL21 and CCR10-CCL27 pathways play minor roles in this process. Our data indicate that the T cells that home to skin infected with H. ducreyi are predominantly central and effector memory cells, with an enrichment of effector memory cells relative to T cells present in peripheral blood. Studies to determine the effect that these T cells have on outcome in the model are ongoing in our laboratory.
ACKNOWLEDGMENTS
This study was funded by National Institutes of Health grants AI27863,AI31494, and AI46784. T.L.H. was supported by T32AI07367 from the National Institute of Allergy and Infectious Diseases. The human challenge trials were also supported by grant M01RR00750 to the General Clinical Research Center at Indiana University.
We thank Barry Katz for statistical analyses; Margaret Bauer for thoughtful criticism of the manuscript; Kate Fortney, Stacy Bennett, Marti Greenwald, and Beth Zwickl for assistance with the human challenge trials; and the volunteers who participated in the study.
REFERENCES
1. Al-Tawfiq, J. A., K. L. Palmer, C.-Y. Chen, J. C. Haley, B. P. Katz, A. F. Hood, and S. M. Spinola. 1999. Experimental infection of human volunteers with Haemophilus ducreyi does not confer protection against subsequent challenge. J. Infect. Dis. 179:1283-1287.
2. Al-Tawfiq, J. A., A. C. Thornton, B. P. Katz, K. R. Fortney, K. D. Todd, A. F. Hood, and S. M. Spinola. 1998. Standardization of the experimental model of Haemophilus ducreyi infection in human subjects. J. Infect. Dis. 178:1684-1687.
3. Bauer, M. E., M. P. Goheen, C. A. Townsend, and S. M. Spinola. 2001. Haemophilus ducreyi associates with phagocytes, collagen, and fibrin and remains extracellular throughout infection of human volunteers. Infect. Immun. 69:2549-2557.
4. Bong, C. T. H., M. E. Bauer, and S. M. Spinola. 2002. Haemophilus ducreyi: clinical features, epidemiology, and prospects for disease control. Microbes Infect. 4:1141-1148.
5. Bong, C. T. H., J. Harezlak, B. P. Katz, and S. M. Spinola. 2002. Men are more susceptible to pustule formation than women in the experimental model of Haemophilus ducreyi infection. Sex. Transm. Dis. 29:114-118.
6. Campbell, J. J., G. Haraldsen, J. Pan, J. Rottman, S. Qin, P. Ponath, D. P. Andrew, R. Warnke, N. Ruffing, N. Kassam, L. Wu, and E. C. Butcher. 1999. The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells. Nature 400:776-780.
7. Campbell, J. J., K. E. Murphy, E. J. Kunkel, C. E. Brightling, D. Soler, Z. Shen, J. Boisvert, H. B. Greenberg, M. A. Vierra, S. B. Goodman, M. C. Genovese, A. J. Wardlaw, E. C. Butcher, and L. Wu. 2001. CCR7 expression and memory T cell diversity in humans. J. Immunol. 166:877-884.
8. Christopherson, K. W., II, A. F. Hood, J. B. Travers, H. Ramsey, and R. A. Hromas. 2003. Endothelial induction of the T-cell chemokine CCL21 in T-cell autoimmune diseases. Blood 101:801-806.
9. DeJong, R., M. Brouwer, B. Hooibrink, T. Van der Pouw-Kraan, F. Miedema, and R. A. W. Van Lier. 1992. The CD27– subset of peripheral blood memory CD4+ lymphocytes contains functionally differentiated T lymphocytes that develop by persistent antigenic stimulation in vivo. Eur. J. Immunol. 22:993-999.
10. De Rosa, S. C., L. A. Herzenberg, L. A. Herzenberg, and M. Roederer. 2001. 11-color, 13-parameter flow cytometry: identification of human naive T cells by phenotype, function, and T-cell receptor diversity. Nat. Med. 7:245-248.
11. Duijvestijn, A. M., E. Horst, S. T. Pals, B. N. Rouse, A. C. Steere, L. J. Picker, C. J. L. M. Meijer, and E. C. Butcher. 1988. High endothelial differentiation in human lymphoid and inflammatory tissues defined by monoclonal antibody HECA-452. Am. J. Pathol. 130:147-155.
12. Faint, J. M., N. E. Annels, S. J. Curnow, P. Shields, D. Pilling, A. D. Hislop, L. Wu, A. N. Akbar, C. D. Buckley, P. A. H. Moss, D. H. Adams, A. B. Rickinson, and M. Salmon. 2001. Memory T cells constitute a subset of the human CD8+CD45RA+ pool with distinct phenotypic and migratory characteristics. J. Immunol. 167:212-220.
13. Fukada, K., Y. Sobao, H. Tomiyama, S. Oka, and T. Masafumi. 2002. Functional expression of the chemokine receptor CCR5 on virus epitope-specific memory and effector CD8+ T cells. J. Immunol. 168:2225-2232.
14. Geissmann, F., M. C. Dieu-Nosjean, C. Dezutter, J. Valladeau, S. Kayal, M. Leborgne, N. Brousse, S. Saeland, and J. Davoust. 2002. Accumulation of immature Langerhans cells in human lymph nodes draining chronically inflamed skin. J. Exp. Med. 196:417-430.
15. Gelfanova, V., T. L. Humphreys, and S. M. Spinola. 2001. Characterization of Haemophilus ducreyi-specific T-cell lines from lesions of experimentally infected human subjects. Infect. Immun. 69:4224-4231.
16. Gunn, M. D., K. Tangemann, C. Tam, J. G. Cyster, S. D. Rosen, and L. T. Williams. 1998. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc. Natl. Acad. Sci. USA 95:258-263.
17. Hamann, D., P. A. Baars, M. H. G. Rep, B. Hooibrink, S. R. Kerkhof-Garde, M. R. Klein, and R. A. W. Van Lier. 1997. Phenotypic and functional separation of memory and effector human CD8+ T cells. J. Exp. Med. 186(9):1407-1418.
18. Homey, B., H. Alenius, A. Muller, H. Soto, E. P. Bowman, W. Yuan, L. McEvoy, A. I. Lauerma, T. Assmann, E. Bunemann, M. Lehto, H. Wolff, D. Yen, H. Marxhausen, W. To, J. Sedgwick, T. Ruzicka, P. Lehmann, and A. Zlotnik. 2002. CCL27-CCR10 interactions regulate T cell-mediated skin inflammation. Nat. Med. 8:157-165.
19. Hudak, S., M. Hagen, Y. Liu, D. Catron, E. Oldham, L. M. McEvoy, and E. P. Bowman. 2002. Immune surveillance and effector functions of CCR10+ skin homing T cells. J. Immunol. 169:1189-1196.
20. Humphreys, T. L., C. T. Schnizlein-Bick, B. P. Katz, L. A. Baldridge, A. F. Hood, R. A. Hromas, and S. M. Spinola. 2002. Evolution of the cutaneous immune response to experimental Haemophilus ducreyi infection and its relevance to HIV-1 acquisition. J. Immunol. 169:6316-6323.
21. Kelly, K. A. 1997. Modulation of leukocyte-endothelial cell interactions by infectious agents. Bull. Inst. Pasteur 95:147-159.
22. King, R., S. H. Choudhri, J. Nasio, J. Gough, N. J. D. Nagelkerke, F. A. Plummer, J. O. Ndinya-Achola, and A. R. Ronald. 1998. Clinical and in situ cellular responses to Haemophilus ducreyi in the presence or absence of HIV infection. Int. J. STD AIDS 9:531-536.
23. Koelle, D. M., Z. Liu, C. M. McClurkan, M. S. Topp, S. R. Riddell, E. G. Pamer, A. S. Johnson, A. Wald, and L. Corey. 2002. Expression of cutaneous lymphocyte-associated antigen by CD8+ T cells specific for a skin-tropic virus. J. Clin. Investig. 110:537-548.
24. Kunkel, E. J., J. J. Campbell, G. Haraldsen, J. Pan, J. Boisvert, A. I. Roberts, E. C. Elbert, M. A. Vierra, S. B. Goodman, M. C. Genovese, A. J. Wardlaw, H. B. Greenberg, C. M. Parker, E. C. Butcher, D. P. Andrew, and W. W. Agace. 2000. Lymphocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine (TECK) expression distinguish the small intestinal immune compartment: epithelial expression of tissue-specific chemokines as an organizing principle in regional immunity. J. Exp. Med. 192:761-767.
25. Magro, C. M., A. N. Crowson, M. Alfa, A. Nath, A. Ronald, J. O. Ndinya-Achola, and J. Nasio. 1996. A morphological study of penile chancroid lesions in human immunodeficiency virus (HIV)-positive and -negative African men with a hypothesis concerning the role of chancroid in HIV transmission. Hum. Pathol. 27:1066-1070.
26. Palmer, K. L., C. T. Schnizlein-Bick, A. Orazi, K. John, C.-Y. Chen, A. F. Hood, and S. M. Spinola. 1998. The immune response to Haemophilus ducreyi resembles a delayed-type hypersensitivity reaction throughout experimental infection of human subjects. J. Infect. Dis. 178:1688-1697.
27. Picker, L. J., S. A. Michie, L. S. Rott, and E. C. Butcher. 1990. A unique phenotype of skin-associated lymphocytes in humans. J. Pathol. 136:1053-1068.
28. Salazar, J. C., C. D. Pope, T. J. Sellati, H. M. Feder, Jr., T. G. Kiely, K. R. Dardick, R. L. Buckman, M. W. Moore, M. J. Caimano, J. G. Pope, P. J. Krause, and J. D. Radolf. 2003. Coevolution of markers of innate and adaptive immunity in skin and peripheral blood of patients with erythema migrans. J. Immunol. 171:2660-2670.
29. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708-712.
30. Sallusto, F., D. Lenig, C. R. Mackay, and A. Lanzavecchia. 1998. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J. Exp. Med. 187:875-883.
31. Santamaria-Babi, L. F. 2004. CLA+ T cells in cutaneous diseases. Eur. J. Dermatol. 14:13-18.
32. Soler, D., T. L. Humphreys, S. M. Spinola, and J. J. Campbell. 2003. CCR4 versus CCR10 in human cutaneous Th lymphocyte trafficking. Blood 101:1677-1683.
33. Spinola, S. M., M. E. Bauer, and R. S. Munson, Jr. 2002. Immunopathogenesis of Haemophilus ducreyi infection (chancroid). Infect. Immun. 70:1667-1676.
34. Spinola, S. M., C. T. H. Bong, A. L. Faber, K. R. Fortney, S. L. Bennett, C. A. Townsend, B. E. Zwickl, S. D. Billings, T. L. Humphreys, M. E. Bauer, and B. P. Katz. 2003. Differences in host susceptibility to disease progression in the human challenge model of Haemophilus ducreyi infection. Infect. Immun. 71:6658-6663.
35. Spinola, S. M., K. R. Fortney, B. P. Katz, J. L. Latimer, J. R. Mock, M. Vakevainen, and E. J. Hansen. 2003. Haemophilus ducreyi requires an intact flp gene cluster for virulence in humans. Infect. Immun. 71:7178-7182.
36. Spinola, S. M., A. Orazi, J. N. Arno, K. Fortney, P. Kotylo, C.-Y. Chen, A. A. Campagnari, and A. F. Hood. 1996. Haemophilus ducreyi elicits a cutaneous infiltrate of CD4 cells during experimental human infection. J. Infect. Dis. 173:394-402.
37. Spinola, S. M., L. M. Wild, M. A. Apicella, A. A. Gaspari, and A. A. Campagnari. 1994. Experimental human infection with Haemophilus ducreyi. J. Infect. Dis. 169:1146-1150.
38. Takahashi, R., Y. Mizukawa, Y. Yamazaki, K. Hayakawa, J. Hayakawa, A. Kudo, and T. Shiohara. 2003. In vitro differentiation from naive to mature E-selectin binding CD4 T cells: acquisition of skin-homing properties occurs independently of cutaneous lymphocyte antigen expression. J. Immunol. 171:5769-5777.
39. Wagers, A. J., J. B. Lowe, and G. S. Kansas. 1996. An important role for the a1,3 fucosyltransferase, FucT-VII, in leukocyte adhesion to E-selectin. Blood 88:2125-2132.
40. Wagers, A. J., L. M. Stoolman, R. Kannagi, R. Craig, and G. S. Kansas. 1997. Expression of leukocyte fucosyltransferases regulates binding to E-selectin. J. Immunol. 159:1917-1929.(Tricia L. Humphreys, Lee )