点击显示 收起
University of Oklahoma Health Sciences Center and Veterans Affairs Medical Center, Oklahoma City
Veterans Affairs Puget Sound Health Care System and University of Washington, Seattle
Patients with disseminated cryptococcosis typically have measurable levels of cryptococcal polysaccharide in serum samples but minimal leukocyte infiltration into infected tissues. In vitro data have shown that cryptococcal polysaccharide induces L-selectin (CD62L) shedding from leukocytes. To assess shedding in vivo, we compared leukocyte L-selectin levels in human immunodeficiency virus (HIV) type 1negative and -positive subjects with and without circulating cryptococcal polysaccharide. Results showed that subjects with cryptococcal polysaccharide in serum samples have significantly lower percentages of neutrophils, monocytes, and CD3+ T cells with L-selectin on their surfaces than do healthy subjects, regardless of HIV status. There was significantly more soluble L-selectin in serum samples from subjects with cryptococcosis than in those from uninfected subjects. Reduced L-selectin levels on leukocytes in subjects with circulating cryptococcal polysaccharide and increased serum levels of soluble L-selectin indicates that surface L-selectin shedding is a mechanism that likely explains reduced leukocyte extravasation into infected tissues of patients with disseminated cryptococcosis.
Cryptococcus neoformans is an encapsulated, yeastlike organism that causes severe disseminated infections in individuals with compromised cell-mediated immune (CMI) responses, particularly in those infected with HIV-1. Subjects with disseminated cryptococcosis typically have high levels of cryptococcal capsular polysaccharide in their serum samples and cerebrospinal fluid (CSF) [1, 2] but little to no leukocyte infiltration into infected tissues [3, 4]. Lack of cellular recruitment, especially of neutrophils, into CSF is not due to a chemoattractant deficiencyChaka et al. [5] found increased levels of a neutrophil chemoattractant, interleukin-8, but few neutrophils in the CSF of HIV-infected patients with cryptococcal meningitis. In vivo data have suggested that cryptococcal polysaccharide or its main component, glucuronoxylomannan (GXM), has a key role in suppressing the inflammatory recruitment of leukocytes [510]. HIV-infected patients with cryptococcal meningitis have ratios of GXM in serum to GXM in CSF that inversely correlate with the numbers of leukocytes in CSF [6]. Intravascular GXM has been shown to inhibit the recruitment of neutrophils, monocytes, and lymphocytes into peripheral gelatin sponges that contain an acute inflammatory stimulus in the mouse model [7] and into the CSF of rabbits with bacterial meningitis [8]; to inhibit the progression of group B streptococcal arthritis [9]; and to suppress collagen-induced arthritis in rats [10]. Because of such findings, there is growing interest in discovering the mechanisms by which circulating capsular polysaccharide inhibits leukocyte migration from the bloodstream into infected tissues.
Dong and Murphy [11] and Dong et al. [12] have shown that cryptococcal polysaccharide induces the shedding of L-selectin (CD62L) from surfaces of human neutrophils and T lymphocytes in vitro. It has been well established that the first step in leukocyte extravasation is the binding of L-selectin on leukocyte surfaces to its ligand on inflamed endothelial cells at the infection site [13]. The binding and subsequent shedding of L-selectin from leukocyte surfaces causes leukocytes to slow and roll along the endothelium and stimulates firm leukocyte attachment to endothelial cells via interactions between CD11a/CD18 (leukocyte function antigen1) and endothelial intercellular adhesion molecule1 [13]. L-selectin and leukocyte rolling have critical roles in neutrophil entry into CSF in meningitis [14] and other tissues [13]. Therefore, GXM-induced shedding of L-selectin from circulating neutrophils could explain their absence from tissues in cryptococcosis. GXM could have adverse effects on lymphocyte migration into lymph nodes, because naive T cells enter lymph nodes by L-selectin binding to CD34 and glyCAM-1 on high endothelial venules [15]. Thus, the early loss of L-selectin from a variety of blood leukocytes caused by circulating GXM has great potential to interfere with innate resistance and the induction and propagation of adaptive immune responses during systemic cryptococcal disease.
We hypothesized that there would be reduced densities of L-selectin on the surfaces of leukocytes and/or reduced numbers of L-selectinpositive leukocytes in patients with disseminated C. neoformans infection who have GXM in their circulation. We tested this hypothesis by analyzing L-selectin densities on the surfaces of circulating leukocytes, determining the percentage of L-selectinpositive cells, and measuring soluble L-selectin (sL-selectin) in serum samples from patients with systemic cryptococcosis in the absence or presence of coinfection with HIV-1. We compared these findings with those from equivalent leukocyte populations or serum samples from HIV-1negative or positive subjects without cryptococcosis. The presence of intravascular cryptococcal polysaccharide significantly reduced the surface density of L-selectin and/or the percentage of L-selectinpositive cells from each cell lineage, and sL-selectin levels in serum samples from HIV-negative patients with cryptococcosis were significantly higher than those in serum samples from control subjects. The loss of L-selectin likely contributes substantially to diminished inflammatory responses in innate and adaptive CMI immune responses to C. neoformans infection.
SUBJECTS, MATERIALS, AND METHODS
Study design and subjects.
The cohort was made up of 4 groups of subjects: (1) healthy individuals without HIV-1 infection (HIV-Cn-); (2) patients with disseminated C. neoformans infection but without HIV-1 infection (HIV-Cn+); (3) clinically healthy individuals with HIV-1 infection (HIV+Cn-); and (4) patients with disseminated C. neoformans infection and concurrent HIV-1 infection (HIV+Cn+). Blood was drawn from a healthy, HIV-negative control subject each time blood was collected from a C. neoformansinfected subject or an HIV-infected subject. All HIV-Cn- subjects tested negative for HIV-1 p24 antigen by ELISA, and both HIV-Cn- and HIV+Cn- subjects tested negative for cryptococcal polysaccharide by a cryptococcal latex agglutination test. The diagnosis of disseminated cryptococcosis was established by positive blood culture or positive cryptococcal polysaccharide titer in serum or CSF. Informed consent was obtained from each patient or their legal guardian. The study was conducted in agreement with the guidelines for human experimentation from the US Department of Health and Human Services and was approved by the institutional review board of the University of Oklahoma Health Sciences Center.
Maintenance of endotoxin-free conditions.
All experiments were performed by use of endotoxin-free plasticware or glassware heated for 3 h at 180°C. Reagents contained <8 pg/mL endotoxin (minimal detectable level) according to the Limulus assay (Whitaker Bioproducts).
Peripheral blood leukocyte isolation.
Blood was drawn into EDTA Vacutainer tubes (BD Biosciences) and immediately put on ice. Neutrophils were enriched by use of a modified Ficoll-Hypaque method [16]. Enriched neutrophil and peripheral blood mononuclear cell (PBMC) layers were collected, and residual red blood cells (RBC) were lysed in 0.85% NH4Cl. Neutrophils and PBMCs were washed twice in PBS, resuspended in RPMI 1640 (Gibco), and processed for immunofluorescent staining within 1 h after collection. Both enriched leukocyte populations were >95% pure, as determined by differential counts by use of Diff-Quick (Baxter). Leukocyte viability was >95% by trypan blue dye exclusion. Blood from each subject was collected into tubes without anticoagulant. Serum was used for the determination of cryptococcal polysaccharide and sL-selectin levels and in the viral assay.
Cryptococcal antigen titer, viral load, and sL-selectin levels.
Cryptococcal capsular polysaccharide was detected with an IMMY kit, in accordance with the manufacture's instructions (IMMY). The lower limit of detection for the assay was 3.212.5 ng/mL cryptococcal antigen. HIV RNA was quantified by the Roche Amplicor HIV-1 Monitor test (Roche). sL-selectin levels in serum samples were assessed by ELISA, in accordance with the manufacturer's instructions (Bender MedSystems). The lower limit of detection for this assay was 0.3 ng/mL.
Monoclonal antibodies (MAbs).
MAbs purchased from BD Biosciences were phycoerythrin-labeled antiL-selectin (Leu-8; mouse IgG2A), peridinin-chlorophyll-protein complexlabeled anti-CD3 (Leu-4; mouse IgG1), fluorescein isothiocyanate (FITC)labeled anti-CD14 (Leu-M3; mouse IgG2b), Cy-Chromelabeled anti-CD4 (mouse IgG1,k), and FITC-labeled anti-CD45RA (mouse IgG2b,k). Appropriately labeled isotype-matched MAb controls were included for each MAb.
CD4+ cell counts.
CD4+ cell-surface markers from EDTA-anticoagulated blood were assessed by use of Simultest CD3/CD4 (BD Biosciences). Briefly, 10 L of antibody mixture was added to 50 L of blood and incubated for 15 min. RBCs were lysed with BD Lyse solution (BD Biosciences). Leukocytes were washed with PBS before fixation and were analyzed by flow cytometry within 4 h.
Expression of surface markers on leukocytes.
Leukocytes (106) were incubated with Fc-receptor blocker (1 g of purified myeloma protein [mouse IgG2a,k] UPC10; Organon Teknika) for 15 min, followed by the addition of staining buffer (PBS with 0.1% bovine serum albumin and 0.1% sodium azide) that contained 0.2 g of fluorescently labeled MAbs or isotype-control antibodies. After 30 min of incubation at 4°C, leukocytes were washed, fixed in 1% paraformaldehyde (wt/vol) in PBS, and analyzed with a FACSCaliber flow cytometer (BD Biosciences). Ten thousand events were collected in gated regions. Neutrophils, monocytes, and lymphocytes were gated by forward and side scatter. CD14+ cells in the monocyte gate and CD3+ cells in the lymphocyte gate were analyzed for L-selectin expression. Cells that stained positive for CD4 were defined as naive if they were in the lymphocyte gate and stained positive for CD45RA.
Statistical analysis.
The Mann-Whitney U test was used for analysis between 2 groups. The Kruskal-Wallis test (a nonparametric analysis of variation) was used for comparisons between multiple groups with Dunn's multiple comparison after testing. Correlations were performed by use of Pearson's rank test. P < .05 was considered to be significant.
RESULTS
Study subjects.
Twenty-six subjects were in the HIV-Cn- group, 6 were in the HIV-Cn+ group, 12 were in the HIV+Cn- group, and 13 were in the HIV+Cn+ group (table 1). Mean ages and age ranges for each of the subject groups were similar (P > .05). The 6 subjects in the HIV-Cn+ group had a median cryptococcal polysaccharide titer of 256 (range, 4512), which was significantly lower than the median cryptococcal antigen titer of 2048 (range, 1282,097,152) in HIV+Cn+ subjects (P < .01). Among HIV-positive subjects, the mean HIV RNA load was slightly, but not significantly, lower in HIV+Cn- subjects (n = 12) than in HIV+Cn+ subjects (table 1). However, the mean (±SE) CD4+ cell count in HIV+Cn+ subjects (35 ± 8.4 cells/L) was significantly lower than that in HIV+Cn- subjects (263 ± 89.8 cells/L) (P < .05). Thus, we selected a subset of HIV+Cn- subjects with CD4+ cell counts of <100 cells/L (n = 6; mean ± SE, 48 ± 17.7 cells/L) for analyses of L-selectin on neutrophils and monocytes as the most appropriate controls for HIV+Cn+ subjects. The subgroup of HIV+Cn- subjects had a mean ± SE HIV RNA log10 titer of 4.76 ± 0.48, compared with 4.69 ± 0.33 in HIV+Cn+ subjects. All subjects were included in the T lymphocyte analysis.
Analysis of L-selectin on neutrophils.
HIV-Cn+ and HIV+Cn+ subjects typically showed lower amounts of L-selectin on their neutrophils than did HIV-Cn- or HIV+Cn- subjects, respectively (figure 1A and 1B). The average mean fluorescence intensity (MFI) of L-selectin on neutrophils from HIV-Cn+ subjects was significantly lower than that of L-selectin on neutrophils from HIV-Cn- or HIV+Cn+ subjects (figure 1C and table 2). The MFI of L-selectin on neutrophils from HIV+Cn+ subjects was less, but not significantly so, than that on neutrophils from HIV+Cn- subjects (figure 1C and table 2). Patients with cryptococcosis, irrespective of HIV status, had significantly lower percentages of L-selectinpositive neutrophils than did their respective control groups (figure 1D and table 2).
Analysis of L-selectin on monocytes.
L-selectin density on monocytes was significantly lower in HIV-Cn+ subjects than in HIV-Cn- subjects and was significantly lower in HIV+Cn+ subjects than in HIV+Cn- subjects (table 2). Monocytes that shed L-selectin in patients with circulating cryptococcal antigen tended to lose L-selectin completely, because the percentage of L-selectinpositive monocytes, irrespective of whether the individual had HIV, was significantly lower than that in the respective control group (table 2). HIV+Cn- subjects actually displayed significantly higher L-selectin densities on monocyte surfaces than did HIV-Cn- subjects (P < .05).
Analysis of L-selectin on T cells.
The mean L-selectin MFI for HIV-Cn+ subjects' CD3+ T cells was significantly lower than that for similar cells in HIV-Cn- subjects (table 2). L-selectin densities were similar between HIV+Cn+ and HIV+Cn- subjects' T cells (table 2). Percentages of L-selectinpositive T cells were significantly lower in subjects with cryptococcosis than in subjects without cryptococcosis, irrespective of whether the individuals were HIV negative or positive (table 2).
Having found lower expression of L-selectin on T cells in subjects with cryptococcosis, we wanted to know whether the CD4 subset was similarly affected. This aspect of the study was limited to the HIV-Cn-, HIV+Cn-, and HIV+Cn+ groups because of difficulty in enrolling sufficient numbers of HIV-Cn+ subjects. When we compared cells from HIV-positive individuals with and without C. neoformans infection, it was necessary to determine the equivalency of viral loads in the 2 groups. There was no statistical difference in mean HIV RNA titers between HIV+Cn- and HIV+Cn+ groups (P > .05) (table 1). We found, as did Park et al. [17], that L-selectin density was reduced on CD4+ cells from healthy HIV-positive subjects, compared with that on CD4+ cells from control subjects (table 3). Although L-selectin density on CD4+ T cells from HIV+Cn- subjects was not reduced further when cryptococcosis was a concomitant disease (table 3), the L-selectin MFI on CD4+ cells and percentage of L-selectinpositive CD4+ cells was significantly lower than those in HIV-Cn- control subjects (table 3).
Analysis of L-selectin on naive T lymphocytes.
Naive CD4+CD45RA+ T cells have surface L-selectin, which facilitates their entry into lymph nodes, where they can become activated [15]. The average L-selectin MFI on CD4+CD45RA+ cells from the HIV+Cn- or HIV+Cn+ group was significantly lower than that on such cells from HIV-Cn- subjects (table 3). There was no difference in L-selectin density on CD4+CD45RA+ cells between HIV+Cn- and HIV+Cn+ subjects (table 3). The mean percentage of L-selectinpositive naive CD4+ T cells in HIV+Cn+ subjects was significantly lower than that in HIV-Cn- subjects (table 3).
sL-selectin levels in serum samples.
sL-selectin levels were significantly increased in serum samples from HIV-Cn+ subjects, compared with those in serum samples from HIV-Cn- subjects (figure 2). We found significantly increased levels of sL-selectin in the circulation of HIV+Cn- subjects, compared with those in the circulation of HIV-Cn- subjects (figure 2). In contrast, we did not find increased levels of sL-selectin in HIV+Cn+ subjects, compared with those in HIV-Cn- subjects (figure 2).
DISCUSSION
The results of a variety of in vivo and in vitro studies have suggested that cryptococcal polysaccharide triggers the shedding of L-selectin from leukocytes [512]. L-selectin on the surfaces of leukocytes is involved in the initial binding and rolling of leukocytes, a process that precedes extravasation into infected tissues and lymph nodes [13, 15, 18]. As a consequence, a reduction in L-selectin levels on circulating leukocytes that is of sufficient magnitude to interrupt normal trafficking patterns would also be expected to reduce the effectiveness of innate and adaptive host resistance mechanisms during infection. In the context of disseminated C. neoformans infection, this could have several manifestations, including decreased numbers of neutrophils and monocytes that enter infected tissues, such as the CSF. Monocytes are not only important in the innate resistance mechanisms (e.g., phagocytosis and killing of invading microbes) but also can develop into antigen-presenting cells in tissues and play a role in the induction of a protective CMI response. Similarly, naive T cells need surface L-selectin to enter lymph nodes, where the adaptive immune response to C. neoformans is induced [15, 19]. Because the anticryptococcal CMI response is the most effective host response against C. neoformans [20], impaired trafficking of these cells would have serious negative consequences for the host.
The experiments presented here tested the extent to which L-selectin was shed from circulating leukocytes in patients with disseminated cryptococcal infection in the presence or absence of concurrent HIV infection. The results showed that percentages of L-selectinpositive neutrophils, monocytes, and T cells were significantly lower in HIV-Cn+ and HIV+Cn+ subjects than in HIV-Cn- and HIV+Cn- subjects. In addition to reducing the numbers of cells bearing L-selectin, the density of L-selectin on the remaining L-selectinpositive neutrophils, monocytes, and T lymphocytes also was significantly lower in HIV-Cn+ subjects than in subjects without cryptococcosis. Widespread leukocyte shedding of surface L-selectin was corroborated by the finding of significantly higher levels of sL-selectin in serum samples from HIV-Cn+ and HIV+Cn- subjects than in serum samples from HIV-Cn- subjects. Although we also expected to see higher levels of sL-selectin in serum samples from HIV+Cn+ subjects, this was not the case. It is possible that the very high concentrations of cryptococcal polysaccharide in the subjects' serum samples interfered with the measurement of sL-selectin or that leukocytes from HIV+Cn+ subjects had shed their L-selectin at an earlier stage in the disease process and the leukocytes were not efficient at regenerating L-selectin, such that the continuous loss of L-selectin from the leukocytes' surfaces into the serum samples was not possible. Further studies must be conducted to explain the finding.
Cryptococcal polysaccharide causes L-selectin to shed from human neutrophils and T cells in vitro [11, 12]. Those data, along with the findings presented here, suggest that circulating cryptococcal polysaccharide in patients with cryptococcosis is responsible for reduced percentages of L-selectinpositive cells and reduced densities of L-selectin on L-selectinpositive leukocytes. The cryptococcal antigen titer in serum does not have to be excessively high to stimulate L-selectin shedding from neutrophil surfaces. This was demonstrated by a significant loss of L-selectin on neutrophils from HIV-Cn+ subjects, whose cryptococcal antigen titers ranged from 4 to 512. Moreover, there was a direct correlation between the level of cryptococcal antigen in serum samples from subjects with cryptococcosis and the level of L-selectin on neutrophils (r = 0.882; P < .05). The lowest densities of L-selectin on neutrophils occurred in subjects with median cryptococcal antigen titers of 256. This may, in part, be due to the polymerization of GXM molecules at higher concentrations but not at lower concentrations, reducing potential reactive sites. Ellerbroek et al. [21] showed that low-molecular-weight GXM was overrepresented at lower GXM concentrations, in contrast to a shift toward high-molecular-weight GXM at higher GXM concentrations. At the stage of cryptococcosis at which the trafficking of leukocytes to infected tissues or lymph nodes would be important in controlling the infection, cryptococcal antigen titers are most likely to be in the range effective to induce optimal L-selectin loss from neutrophils. Further studies with greater numbers of patients with cryptococcosis are needed to confirm our observation about the association of lower concentrations of cryptococcal antigen in serum with lower densities of L-selectin on neutrophils.
Our assessment of L-selectin on the surface of neutrophils and monocytes in HIV-positive subjects was limited to those who had a CD4+ cell count of 100 cells/L, so that the cryptococcosis-negative and-positive groups were more similar in this respect. The rationale for this was based on results from Elbim et al. [22], who showed that neutrophils from HIV-positive subjects with CD4+ cell counts of <200 cells/L did not shed L-selectin on stimulation with bacterial N-formyl peptides as readily as did neutrophils from HIV-negative control subjects or from a group of HIV-infected subjects with CD4+ cell counts of >200 cells/L. Those investigators also showed that subjects in early stages of an HIV infection (CD4+ cell count of >500 cells/L) displayed significant shedding of L-selectin from neutrophils [22]. These data [22] imply that, as HIV infection progresses, neutrophils become defective in their ability to shed L-selectin. The shedding of L-selectin from leukocytes is achieved by the leukocyte's metalloproteinase clipping off L-selectin [23]. The neutrophils from HIV-positive subjects in our study had levels of L-selectin similar to those of HIV-negative subjects. Our results are in agreement with the concept that, during late-stage HIV disease, neutrophils are resistant to L-selectin shedding, because, despite the fact that HIV+Cn+ subjects had high levels of cryptococcal polysaccharide, their neutrophils did not show a significant L-selectin loss, compared with neutrophils from HIV+Cn- subjects. Thus, the loss of L-selectin from neutrophils from HIV+Cn+ subjects was likely due to circulating cryptococcal polysaccharide; however, the amount of shedding of L-selectin seems to be affected by the stage of HIV infection.
Our finding that HIV-positive subjects' monocytes had significantly more L-selectin on their surfaces than did those of HIV-negative subjects (table 2) differ from those of Elbim et al. [24] and Trial et al. [25], who showed that subjects with late-stage HIV disease had reduced levels of L-selectin on their monocytes, compared with control subjects. The reasons for these divergent results are not clear. The MFI that we measured after staining for L-selectin was much higher (516 for HIV-negative subjects and 1180 for HIV-positive subjects) than was reported by the other 2 groups (90 and 140 for HIV-negative subjects and 6570 and 125 for HIV-positive subjects). It is possible that their procedures for the isolation of monocytes caused some shedding and that they were thus not observing all of the L-selectin on the monocytes as it occurs in the bloodstream.
Loss of L-selectin from naive T cells affects the induction of the CMI response by preventing the migration of naive T cells into lymph nodes. Thus, reduced numbers of L-selectinpositive naive T cells and reduced density on the remaining L-selectinpositive naive T cells, as we have shown in subjects with cryptococcosis, would result in diminished development of an anticryptococcal CMI response. The anticryptococcal CMI response is the most effective protective host response against C. neoformans [19], so any reduction in that response would be quite detrimental to the host.
In summary, we have demonstrated that L-selectin levels are reduced on neutrophils, monocytes, and lymphocytes in vivo in subjects with cryptococcosis who have circulating cryptococcal polysaccharide. These data, together with the finding of increased sL-selectin levels in serum samples from subjects with cryptococcosis, support the hypothesis that the loss of L-selectin is induced by cryptococcal polysaccharide in serum. Reduced levels of L-selectin on leukocytes during disseminated cryptococcosis is a likely mechanism whereby leukocytes are prevented from entering sites infected by C. neoformans as well as draining lymph nodes and, thus, can suppress both innate and adaptive host defense mechanisms.
References
1. Bloomfield N, Gordon MA, Elmendorf DF. Detection of Cryptococcus neoformans antigen in body fluids by latex particle agglutination. Proc Soc Exp Biol Med 1963; 114:647. First citation in article
2. Gordon MA, Vedder DK. Serologic tests in diagnosis and prognosis of cryptococcosis. JAMA 1966; 197:9617. First citation in article
3. Baker RD, Haugen RK. Tissue changes and tissue diagnosis in cryptococcosis: a study of twenty-six cases. Am J Clin Pathol 1955; 25:1424. First citation in article
4. Farmer SG. Histological response to capsule deficient Cryptococcus neoformans. Arch Pathol 1973; 96:3837. First citation in article
5. Chaka C, Heyderman R, Gangiadzo I, et al. Cytokine profiles in cerebrospinal fluid of human immunodeficiency virusinfected patients with cryptococcal meningitis: no leukocytosis despite high interleukin-8 levels. J Infect Dis 1997; 176:16336. First citation in article
6. Lipovsky M, Elden L, Walenkamp A, Dankert J, Hoepelman A. Does the capsule component of the Cryptococcus neoformans glucuronoxylomannan impair transendothelial migration of leukocytes in patients with cryptococcal meningitis J Infect Dis 1998; 178:12312. First citation in article
7. Dong ZM, Murphy JW. Intravascular cryptococcal culture filtrate (CneF) and its major component, glucuronoxylomannan, are potent inhibitors of leukocyte accumulation. Infect Immun 1995; 63:7708. First citation in article
8. Lipovsky M, Tsenova L, Coenjaerts FE, Kaplan G, Cherniak R, Hoepelman AI. Cryptococcal glucuronoxylomannan delays translocation of leukocytes across the blood brain barrier in an animal model of acute bacterial meningitis. J Neuroimmunol 2000; 111:104. First citation in article
9. Tissi L, Puliti M, Bistoni F, Mosci P, Kozel TR, Vecchiarelli A. Glucuronoxylomannan, the major capsular polysaccharide of Cryptococcus neoformans, inhibits the progression of group B streptococcal arthritis. Infect Immun 2004; 72:636772. First citation in article
10. Mirshafiey A, Ngoc Dzung DT, Murphy JW, et al. Tolerability and anti-inflammatory effects of glucuronoxylomannan in collagen-induced arthritis. Scand J Immunol 2004; 60:22632. First citation in article
11. Dong ZM, Murphy JW. Cryptococcal polysaccharide induces L-selectin shedding and tumor necrosis factor receptor loss from the surface of human neutrophils. J Clin Invest 1996; 97:68998. First citation in article
12. Dong ZM, Jackson L, Murphy JW. Mechanisms for induction of L-selectin loss from T-lymphocytes by a cryptococcal polysaccharide, glucuronoxylomannan. Infect Immun 1999; 67:2209. First citation in article
13. Ley K. Integration of inflammatory signals by rolling neutrophils. Immunol Rev 2002; 186:818. First citation in article
14. Granert C, Raud J, Lindquist L. The polysaccharide fucoidin inhibits the antibiotic-induced inflammatory cascade in experimental pneumococcal meningitis. Clin Diagn Lab Immunol 1998; 5:3224. First citation in article
15. Butcher EC. Lymphocyte homing and homeostasis. Science 1996; 272:606. First citation in article
16. Babcock GF. Neutrophil isolation procedure. In: Robinson JP, ed. Handbook of flow cytometry methods. New York: Wiley-Liss, 1993:22. First citation in article
17. Park SW, Royal W III, Semba RD, Wiegand GW, Griffin DE. Expression of adhesion molecules and CD28 on T lymphocytes during human immunodeficiency virus infection. Clin Diagn Lab Immunol 1998; 5:5837. First citation in article
18. Luscinskas FW, Ding H, Tan P, Cumming D, Tedder TF, Gerritsen ME. L- and P-selectins, but not CD49d (VLA-4) integrins, mediate monocyte initial attachment to TNF-alphaactivated vascular endothelium under flow in vitro. J Immunol 1996; 157:32635. First citation in article
19. Bauman SK, Nichols KL, Murphy JW. Dendritic cells in the induction of protective and nonprotective anticryptococcal cell-mediated immune responses. J Immunol 2000; 165:15867. First citation in article
20. Murphy JW, Schafer F, Casadevall A, Adesina A. Antigen-induced protective and nonprotective cellmediated immune components against Cryptococcus neoformans. Infect Immun 1998; 66:26329. First citation in article
21. Ellerbroek PM, Hoepelman A, Wolbers F, Zwaginga JJ, Coenjaerts F. Cryptococcal glucuronoxylomannan inhibits adhesion of neutrophils to stimulated endothelium in vitro by affecting both neutrophils and endothelial cells. Infect Immun 2002; 70:476271. First citation in article
22. Elbim C, Prevot MH, Bouscarat F, et al. Polymorphonuclear neutrophils from human immunodeficiency virusinfected patients show enhanced activation, diminished fMLP-induced L-selectin shedding, and an impaired oxidative burst after cytokine priming. Blood 1994; 84:275966. First citation in article
23. Chen A, Engel P, Tedder TF. Structural requirements regulate endoproteolytic release of the L-selectin (CD62L) adhesion receptor from the cell surface of leukocytes. J Exp Med 1995; 182:51930. First citation in article
24. Elbim C, Pillet S, Prevost MH, et al. Redox and activation status of monocytes from human immunodeficiency virusinfected patients: relationship with viral load. J Virol 1999; 73:45616. First citation in article
25. Trial J, Birdsall HH, Hallum JA, et al. Phenotypic and functional changes in peripheral blood monocytes during progression of human immunodeficiency virus infection. J Clin Invest 1995; 95:1690701. First citation in article