Human Platelets Damage Aspergillus fumigatus Hyphae and May Supplement Killing by Neutrophils

Abstract

Neutropenia is considered a significant risk factor for invasive aspergillosis but is almost always associated with concurrent thrombocytopenia. Studies determined that platelets, like neutrophils, attached to cell walls of the invasive hyphal form of Aspergillus fumigatus. Organisms were damaged as shown by loss of cell wall integrity in scanning laser confocal microscopy and release of defined hyphal surface glycoproteins. Rapid expression appearance of surface antigen CD63 and release of markers of platelet degranulation confirmed activation during attachment to hyphae. Optimal platelet activation required opsonization of hyphae with fresh or heat-inactivated whole plasma. These effects of opsonization with whole plasma could not be duplicated by pooled human serum, immunoglobulin G, or fibrinogen, whether used separately or combined. Thus, platelets in the presence of whole plasma have the potential to play an important role in normal host defenses against invasive aspergillosis.

The crucial role of phagocytic cells, particularly polymorphonuclear neutrophils (PMN), in host defenses against invasive aspergillosis is supported by the occurrence of the disease under well-characterized clinical conditions where these cells are either dysfunctional (7) or scarce (37). A large body of in vitro evidence and animal data from our own (14, 15, 20) and other (30) laboratories corroborates and helps to explain the frequent association between neutropenia and disseminated aspergillosis.

However, chemotherapy-induced neutropenia rarely occurs without concomitant thrombocytopenia. Several clinical and experimental observations suggest that platelets, besides preventing hemorrhage, may contribute to host defenses against a broad range of infections (6, 5). For example, thrombocytopenic rabbits tend to have larger vegetations and higher density of streptococci in an experimental model of bacterial endocarditis (32). Platelets stimulated with α-thrombin release a cationic protein (platelet microbicidal protein [PMP]) which is microbicidal to some strains of Staphylococcus aureus (35) and Candida albicans (29). Recent data have shown that resistance of C. albicans to PMP increases the severity of experimental Candida endocarditis (36). Platelets also are cytotoxic in vitro against larvae of Schistosoma sp. (10, 18, 25) and inhibit growth of Plasmodium malariae in vitro (26).

Aspergillus hyphae typically invade blood vessels, causing local thrombosis with consequent distal hemorrhagic infarction. This characteristic propensity for intravascular invasion preceding and during dissemination, the association of dissemination with thrombocytopenia, and the known microbicidal capabilities of platelets led us to examine the potential role of platelets in host defenses against infection by Aspergillus fumigatus, the most common cause of invasive aspergillosis.

Our results established that A. fumigatus hyphae strongly activated human platelets in vitro and that platelets damaged A. fumigatus hyphae in vitro.

(This work was presented in part at the 34th Annual Meeting of the Infectious Diseases Society of America, New Orleans, La., 18–20 September 1996 [4a].)

MATERIALS AND METHODS

Reagents.

All reagents were purchased from Sigma Chemical, St. Louis, Mo., unless otherwise stated. Experiments involving PMN were done in phosphate-buffered saline (PBS; BioWhittaker, Walkersville, Md.) with 5.5 mM glucose, 3.4 mM CaCl₂, and 5.25 mM MgCl₂ (supplemented PBS). Platelets were suspended in buffer containing 3.8 mM HEPES with 140 mM NaCl, 3.75 mM NaH₂PO₄, 21 mM KCl, 1 mM CaCl₂, and 5.5 mM glucose (HEPES-Ca²⁺).

Organisms.

As in past experiments (22), A. fumigatus conidia from a clinical isolate were harvested from culture on Sabouraud dextrose agar slants, then suspended in Sabouraud dextrose broth at 10⁶/ml, and left overnight at room temperature on a gyrotatory shaker. The swollen conidia were then germinated at 37°C for 2.5 h. Under these conditions, ≥90% formed hyphae.

Opsonization of hyphae.

Samples of fresh blood were either anticoagulated with EDTA or left to clot at 37°C and then centrifuged for preparation of autologous plasma or serum, respectively. Platelet-poor plasma was obtained by centrifugation (2,000 × g for 10 min). Germinated hyphae were resuspended in glass tubes at 50 × 10⁶/ml with the following opsonic solutions: pooled human plasma (BioWhittaker), fresh or heat-inactivated (30 min at 56°C) autologous plasma or serum, fibrinogen (3 mg/ml), human immunoglobulin G (IgG; 1 mg/ml), and various combinations of these opsonins. Following 20 min of incubation at 37°C, hyphae were washed twice, resuspended in the working buffer, and kept at room temperature until used.

Fluorescent labeling of A. fumigatus hyphae.

Biotinylation of hyphal cell wall glycoproteins was performed by a previously published method (4), with the following modifications. Freshly germinated hyphae were suspended at 3 × 10⁷/ml in 3 ml of 100 mM phosphate buffer (pH 8.0) containing N-hydroxysuccinimidobiotin (0.1 mg/ml). Following 15 min of incubation at 37°C with gentle shaking, the hyphae were washed sequentially with 50 mM (pH 6.0) and 10 mM (pH 7.4) phosphate buffers and were resuspended (3 × 10⁷/ml) in HEPES-Ca²⁺. The biotinylated hyphae were then incubated with 5 μg of 5-([4, 6-dichlorotriazin-2-yl]amino)fluorescein (DTAF)-conjugated streptavidin (Jackson ImmunoResearch Laboratories, West Grove, Pa.) per ml for 10 min at room temperature, hand mixed, and protected from light. Following two washes with supplemented PBS to remove any excess of DTAF-conjugated streptavidin, hyphae were resuspended in working buffer and wrapped in aluminum foil until used. Biotinylation with or without labeling with fluorescent conjugate did not alter hyphal viability, ability to activate platelets, or susceptibility to damage by either platelets or neutrophils.

Microscopy.

Transmitted light imaging and fluorescent light imaging were performed with a Nikon (Garden City, N.Y.) microscope. Samples obtained at intervals were photographed with a Nikon 2000 camera and Ektachrome film (Eastman Kodak, Rochester, N.Y.). For selected experiments, hyphal damage was visualized with a Leica laser scanning confocal microscope (Leica, Lasertechnik, GmbH, Heidelberg, Germany) equipped with an argon ion laser (2 to 50 mW). The smallest apertures were applied to the detection pinhole for optimal signal. Z-series optical sections were recorded at 0.5 to 1.0 μm, using a 63× (numerical aperture = 1.4) oil immersion lens and stored on optical disks (27, 34). Computer images were generated and prints were made with Adobe Photoshop software (San Jose, Ca.).

Preparation of PMN and platelets.

As in past experiments, PMN (20) and platelets (11) were freshly prepared for each experiment from blood of healthy volunteer donors. Blood was anticoagulated with either heparin or citrate for PMN or platelet preparation, respectively. Blood for PMN preparation was diluted 1:1 with isotonic 3% dextran and incubated for 20 min at 37°C to accelerate erythrocyte sedimentation. Supernatants were then centrifuged on Ficoll-metrizoate gradients. Residual erythrocytes were lysed twice with hypotonic NaCl. Yield of functional PMN, assessed by nitroblue tetrazolium reduction, was >98%. Immediately before use, PMN were resuspended in supplemented PBS to final concentrations of 3 × 10⁶ to 6 × 10⁶/ml and kept on ice until used. Platelet-rich plasma was obtained by centrifugation of citrate anticoagulated blood. Platelets were separated on Sepharose 2B columns (11), then diluted to final concentrations of 6 × 10⁷ to 20 × 10⁷/ml in HEPES-Ca²⁺, and kept at 37°C with continuous gentle rocking until used. Platelets were prepared in this manner to minimize the potential for spontaneous activation as shown previously (11). Spontaneous activation of platelets during the preparation procedure was identified by change in platelet shape, aggregation, spontaneous degranulation (judged by β-glucuronidase release), and lack of response to thrombin stimulation in controls. Platelets evidencing activation during preparation were discarded.

Platelet activation by Aspergillus was detected by fluorescent labeling of CD42b (GPIb), an antigen present on plasma membranes of both resting and activated platelets (31), and of CD63, which is present on plasma membranes of activated platelets only (24). Platelets were mixed with Aspergillus hyphae (ratio 100:1) or activated with α-thrombin (9 nM) and incubated for 1 h at 37°C. DTAF-conjugated mouse anti-human CD42b or DTAF-conjugated mouse anti-human CD63 (Becton Dickinson, San Jose, Calif.) was added at 5 μg/ml (final concentration). After incubation for 15 min at 37°C, reactions were stopped with 3.7% buffered formalin. Samples were compared by fluorescence microscopy.

Platelet degranulation.

Following germination, hyphae were washed once with HEPES-Ca²⁺ and resuspended in this buffer at 4 × 10⁷/ml. Resting platelets were mixed with germinated hyphae in a 40/1 ratio and incubated for 30 min at 37°C with gentle mixing. Supernatants were obtained by rapid centrifugation (twice at 10⁴ × g for 4 min) through 80:20 (vol/vol) Dow Corning Contour oil (Nye Lubricants, New Bedford, Mass.). Samples were kept frozen at −70°C. Samples were diluted 1:10 in HEPES-Ca²⁺ buffer for assays of released platelet granule constituents. Markers used were platelet factor 4 (PF4) for α-granule release, β-glucuronidase for lysosomal granule release, and serotonin for δ (dense)-granule release.

To determine β-glucuronidase release, 100 μl of sample was mixed with 200 μl of 6 mM 4-methylumbelliferyl-β-d-glucuronide in 100 mM acetate buffer (pH 5.0) plus 200 μl of acetate buffer (500 μl, final volume). Samples were incubated for 30 min at 37°C shielded from light, 500 μl of 200 mM glycine (pH 10.5) was added to each sample, and fluorescence was read immediately (excitation, 360 nm; emission, 448 nm; Perkin-Elmer [Weston, Mass.] 650-10S fluorescence spectrophotometer). β-Glucuronidase release was determined as a fraction of total β-glucuronidase content obtained from 0.1% Triton X-100-lysed platelets corrected for background supernatant fluorescence prior to stimulation.

Serotonin release was measured as previously reported (12). Concentrated gel-filtered platelets (4 × 10⁹/ml) were loaded with [³H]serotonin (4 × 10⁻⁵ mCi/ml) for 20 min at 37°C. To prevent serotonin reuptake, the serotonin analog 13.3 nM imipramine was added within 30 s prior to activation. After stimulation, platelets were centrifuged through contour oil as noted above. After determination of ³H in each supernatant, serotonin content was expressed as a fraction of total serotonin content of 0.1% Triton X-100-lysed platelets.

PF4 release was determined by enzyme-linked immunosorbent assay (ELISA). After centrifugation of platelet supernatants through contour oil, 10 μl of each supernatant was diluted with 90 μl buffer in 96-well ELISA microtiter plates (Dynatech Laboratories, Inc., Chantilly, Va.) and kept at 4°C for 14 to 16 h. Following serial washes to remove unbound contents, plates were blocked with 3% bovine serum albumin in PBS and labeled with 0.17 μg of horseradish peroxidase-bound mouse anti-PF4 antibody (a generous gift from Repligen, Inc., Needham, Mass.) per ml, incubated, and washed again. TBM Microwell substrate (Kirkegaard & Perry, Inc., Gaithersburg, Md.) was used for development, and reactions were stopped with 0.36 N sulfuric acid. Absorbance was read at 450 nm. Data were normalized to PF4 release by platelets maximally stimulated with 9 nM α-thrombin.

Comparison of hyphal damage by platelets and activated platelet supernatants.

Resting platelets were split into two batches, one kept resting and the other maximally stimulated with 9 nM α-thrombin for 5 min. Supernatants were obtained by centrifugation (2,000 × g for 10 min). Following germination and opsonization with pooled human serum, hyphae were resuspended in HEPES-Ca²⁺.

Hyphae were incubated for 30 min at 37°C with gentle mixing together with platelets (platelet/hypha ratio of 100:1 or 400:1) or with supernatants (originally obtained from maximally activated or resting platelets). Reactions were stopped with 0.5 ml of ice-cold double-distilled H₂O (ddH₂O). Hyphal pellets were washed twice with cold ddH₂O to lyse platelets and remove supernatants. For measurement of hyphal metabolic activity, 2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]2H-tetrazolium-5-carboxanilide inner salt (XTT) was dissolved by warming in a water bath at 60°C for 30 min, cooled to room temperature, and mixed with 8 μg of 2,3-dimethoxy-5-methyl-1,4-benzoquinone (Co-Q₀) per ml in PBS. Hyphal pellets were resuspended and mixed with 0.5 ml of the XTT–Co-Q₀ mixture in PBS (22). Samples were incubated at 37°C for 30 min with gentle tumbling, and 100 μl of centrifuged (10 times at 10³ × g for 2 min at 4°C) supernatants were transferred to microtiter plate wells. Absorption was measured at 570 nm. Optical density at 650 (OD₆₅₀) nm was also read to control for nonspecific absorption. Hyphal damage was calculated by using the equation 1 − ([OD₅₇₀ hyphae + PMN or platelets or PMN + platelets]/OD₅₇₀ hyphae − [OD₅₇₀ PMN or platelets or PMN + platelets/OD₅₇₀ hyphae]) × 100.

Estimate of hyphal damage by PMN and platelets.

Triplicate experiments were performed in 1.5-ml polypropylene microcentrifuge tubes. Tubes containing hyphae were filled with 60 μl (6 × 10⁵ hyphae) of organisms in HEPES-Ca²⁺. PMN and platelets, separately or in combination, were incubated with hyphae. Incubations were performed as noted above. Incubations thus included hyphae alone, PMN alone, platelets alone, and PMN plus platelets combined in each concentration tested (660-μl final volumes in HEPES-Ca²⁺).

Following 1 h of incubation at 37°C with gentle tumbling, 0.5 ml of ice-cold ddH₂O was added to stop reactions. Samples were centrifuged at 10⁴ × g for 2 min at 4°C. Pellets were resuspended with 1 ml of sterile ddH₂O, briskly mixed, and allowed to stand at room temperature for 10 min to achieve hypotonic lysis of PMN. Platelets and hyphal pellets were washed again with ddH₂O; hyphae were resuspended in 1 ml of fresh Sabouraud dextrose broth and incubated for 0 to 4 h at 37°C with gentle tumbling. Specimens were centrifuged (10 times at 10³ × g for 2 min at 4°C), and pellets were resuspended with 0.5 ml of XTT–Co-Q. Following 30 min of incubation at 37°C with mixing, damage was calculated by using the formula stated above.

Data analysis.

Data were analyzed with Sigmastat 1.0 software (Jandel Scientific, San Rafael, Calif.). The Student t test was used to assess differences of means for normally distributed variables. Results are presented as mean ± standard error of the mean (SEM) of percentage of controls unless otherwise specified.

RESULTS

Platelet adherence and associated loss of hyphal surface proteins.

Platelets adhered to surfaces of opsonized A. fumigatus hyphae (Fig. 1 and 2A). In the early period after attachment, platelets retained rounded oval configurations (Fig. 2A). With activation, attached platelets tended to contract and flatten as they spread over hyphal surfaces, imparting a markedly irregular appearance to hyphal surfaces (Fig. 1). In contrast, hyphal surfaces without attached platelets retained their normally smooth, regular appearances (Fig. 2A to C). Over a 1-h incubation period, platelets induced loss of fluorescent hyphal cell wall constituents, as observed by laser scanning confocal microscopy (Fig. 2). Fluorescence of fluorescein isothiocyanate-labeled cell walls in direct contact with attached platelets decreased progressively, whereas fluorescence intensity did not decrease on the surfaces of hyphal walls without attached platelets (Fig. 2B to J). Eventually, hyphae within masses of surrounding, attached platelets lost nearly all fluorescence, while the masses of surrounding platelets accumulated fluorescence in homogeneous patterns (Fig. 2D to J).

FIG. 1.

Aggregated platelets completely covering A. fumigatus hyphae. Arrows (length of wide arrow = 5 μm) denote the irregular pattern formed by the complete coating of hyphal surfaces with spread, retracted platelets, which have lost the typical round or oval appearance of unactivated platelets (arrows), consistent with platelet activation. A broader mass of platelets was present at the hyphal tip (wide arrow), reflecting the focal variations in the sizes of platelet aggregates attached to hyphae in these experiments.

FIG. 2.

Light and laser confocal scanning microscopy of fluorescein-labeled, biotinylated A. fumigatus hypha during the course of 0 to 1 h of incubation with platelets. Color images depict computer-reconstructed stereoscopic fluorescence images formed by summation of fluorescence of all optical sections of 1-μm thickness. The intensity of fluorescence is represented by a range of colors, with brown, orange, yellow, and blue-white reflecting progression from low to intermediate to high fluorescence intensities. These color differences can be compared within each of the separate color images. However, apparent color differences between each of the separate photographs do not necessarily represent relative changes or differences in fluorescence intensities. The system did not allow sufficiently reliable calibration of hyphal and background fluorescence to permit accurate comparisons of relative differences in fluorescence between independent color images. (A) A transmitted light image after 5 min of incubation shows early attachment of platelets to a hyphal surface. Note the oval shapes suggesting unactivated platelets (straight arrows), as well as the preservation of the normally smooth hyphal surface (curved arrows). (B) A color confocal image (left) shows a regular pattern of orange to yellow fluorescence of the hyphal surface (straight arrow = 10 μm) surrounding the less fluorescent brown hyphal interior after 1 h of incubation (curved arrow). (C) Corresponding to panel B, the transmitted light image (right) defines a single oval platelet (curved arrow) attached to a small portion of the surface of one hypha after 1 h of incubation. Note the distinct, regular outline of the hyphal surface in the absence of attached platelets (straight arrows). (D) In this confocal image showing opsonized hyphae completely coated by an aggregated mass of platelets after 1 h of incubation, note the almost complete loss of discernible hyphal outline and some focal areas of increased fluorescence within platelet aggregates. Arrows delineate a small area corresponding to a remnant of a hyphal cell wall within the platelet mass. (E) A transmitted light image corresponding to panel D shows a discernible hyphal cell wall remnant within the mass of aggregated platelets (arrows). (F) A confocal image shows diffuse redistribution of increased fluorescence throughout a mass of aggregated platelets (arrows) after 1 h of incubation. (G) Arrows indicate a definable remnant of a hyphal cell wall in this transmitted light image corresponding to panel F. Note the location of the hyphal remnant compared to the irregular distribution of increased fluorescence shown in the confocal image (F). (H) For most of the hyphae that were completely surrounded by platelets, cell walls became indistinct and fluorescence was dispersed throughout the masses of platelets (arrows; top arrow = 22.5 μm) within 2 h or less, as in this confocal image taken after 1 h of incubation. (I) In this barely visible negative image of nonfluorescent hyphal remnant marked by arrows, some residual fluorescence is scattered within the mass of adherent platelets. (J) Shown here in a mixed confocal/transmitted light image, hyphae free of platelets retained relatively uniform surface fluorescence during 3 h of incubation, even though some labeled hyphal proteins were released spontaneously over time (Fig. 3), as expected from previously published work by others (33). An unattached, oval, apparently unactivated platelet is shown near the labeled hypha (arrow).

Consistent with this apparent loss of hyphal wall integrity observed by microscopy, platelets induced a time-dependent increase in release of distinctive fluorescent biotinylated cell wall proteins into supernatants (Fig. 3A). Data previously published by others established that live hyphae release certain surface proteins during normal growth (33). Evidence from Wosten and colleagues (33) suggests that most such proteins are released by secretion at hyphal tips. Beyond this spontaneous release of hyphal proteins, separation of A. fumigatus proteins from supernatants established that platelets induced release of additional proteins, in levels well above those occurring spontaneously over time (Fig. 3B). Controls verified that biotinylation and fluorescent labeling did not alter hyphal viability, ability to activate platelets, or susceptibility to damage by host cells.

FIG. 3.

Platelet-triggered release of biotinylated hyphal cell wall surface proteins. (A) Time-dependent release of fluorescein-labeled cell wall surface glycoproteins into supernatants of hyphae incubated with (○) and without (•) platelets. (B) Western blot showing time-dependent release of biotinylated cell wall surface glycoproteins into supernatants of hyphae exposed to platelets. Lanes 1 to 3, sequential protein release into supernatants from hyphae incubated with platelets for 1, 2, and 3 h, respectively; lanes 4 and 5, spontaneous protein release into supernatants from hyphae incubated for 0 and 3 h without platelets. Consistent with previously published observations by other investigators (33), certain particular hyphal proteins were secreted spontaneously and were detectable in increasing amounts over time.

Platelet activation and degranulation.

Adherence to A. fumigatus hyphae resulted in platelet activation. As expected from published data by others (31), the CD42b surface antigen proved to be constitutively expressed on surfaces of both resting and activated platelets (Fig. 4A). Interaction of platelets with hyphae clearly induced expression of the CD63 antigen on platelet surfaces (Fig. 4B), an established antigenic marker for platelet activation (24). In contrast, surfaces of resting platelets with no attached hyphae did not bind antibodies to the CD63 activation antigen, as they remained nonfluorescent.

FIG. 4.

(A) Staining with fluorescein-conjugated monoclonal anti-CD42b (GPIb) antibody, which binds to an antigen that is constitutively expressed on surfaces of both resting (arrowhead) and stimulated platelets (arrows), identifying platelets adhering to large areas of hyphal cell wall (arrow = 5 μm). (B) Staining of platelets attached to hyphae with fluorescein-conjugated monoclonal anti-CD63 (GPIIIa) antibody, reactive with an antigen that is present exclusively on surfaces of activated platelets. Platelets coating hyphal cell wall were brightly fluorescent, consistent with activation. (C) Corresponding transmitted light image showing unattached platelets (arrows) which remained nonfluorescent (arrow = 5 μm).

Further supporting hypha-induced platelet activation, A. fumigatus hyphae that were opsonized with autologous human plasma also triggered release from all three platelet granule types: PF4 from α-granules, β-glucuronidase from lysosomal granules, and serotonin from δ-granules (Fig. 5A). The relative efficiency of different opsonins in supporting platelet activation was assessed by comparing hypha-induced β-glucuronidase release to release of that enzyme from platelets after maximal stimulation with α-thrombin (MAX). Optimal opsonization occurred with whole plasma (Fig. 5B). No significant differences occurred with fresh autologous plasma (82.4% ± 9.2% of MAX; mean ± SEM of 12 separate triplicate experiments), heat-inactivated (56°C, 30 min) autologous plasma (98.8% ± 11.7% of MAX; mean ± SEM of 3 separate triplicate experiments), or pooled plasma (60.0% ± 8.3% of MAX; mean ± SEM of 8 separate triplicate experiments). In contrast, only minimal release from platelet granules was triggered by unopsonized hyphae or by hyphae which had been opsonized with pooled human serum, IgG, or fibrinogen, whether used separately or in various combinations in three or more separate experiments.

FIG. 5.

Platelet degranulation following interaction with A. fumigatus hyphae. (A) Effects of A. fumigatus hyphae opsonized with fresh autologous plasma on release of markers for platelet α-granules (PF4), δ-granules (serotonin [5-OHT]), and lysosomal granules (β-glucuronidase [β-glu]). Results depict mean ± SEM of 3 to 12 separate experiments, each performed in triplicate. Open bars, responses to opsonized hyphae; solid bars, responses to unopsonized hyphae. (B) Effects of various opsonins on platelet degranulation triggered by A. fumigatus hyphae. β-Glucuronidase release following hyphal opsonization with fresh autologous plasma (AP), heat-inactivated fresh autologous plasma (HI-AP), pooled plasma (PP), pooled human serum (PHS), heat-inactivated pooled human serum (HI-PHS), fibrinogen (F), pooled human IgG, and various combinations compared to that of unopsonized hyphae (UOH). Results represent mean ± SEM of 3 to 12 separate experiments, each performed in triplicate.

Platelet and PMN effects on hyphal metabolic responses.

Activated platelets decreased the ability of hyphae to reduce XTT, a previously established indicator of fungal cell damage (22). As shown in Fig. 6A, increasing ratios of platelets to hyphae from 50:1 to 400:1 in incubations induced progressive declines in the ability of hyphae to reduce XTT. A 50:1 ratio of platelets to hyphae resulted in a 12.5% ± 4.9% decrease in hyphal reduction of XTT (P <0.01 compared to corresponding controls incubated for 1 h in buffer with no platelets or to zero-time controls, i.e., hyphae mixed with platelets but not incubated). Significantly greater effects occurred with a 200:1 ratio of platelets to hyphae (P < 0.001 compared to results obtained with a 50:1 ratio). Prolongation of hyphal incubations with platelets beyond 1 h did not further alter effects of platelets on XTT reduction. However, it remained possible that platelet-mediated effects on hyphal metabolism were reversible, causing only temporary suppression of the ability of hyphae to reduce XTT. To determine whether platelets induced sustained suppression of hyphal metabolic activity, after 1 h of exposure to platelets, incubations of hyphae with XTT were prolonged for up to 2 h. Platelet/hypha ratios of 50:1 and 167:1 were used in eight separate experiments, each performed with two to four replicates. With both ratios, platelet-induced decrements in hyphal metabolic activity were sustained, as there was no recovery of the ability to reduce XTT during more prolonged incubation with the dye (Fig. 6B). In contrast to intact platelets, supernatants of α-thrombin-activated platelets induced no detectable alterations in the capacity of hyphae to reduce XTT.

FIG. 6.

Effects of platelets on XTT reduction by hyphae. (A) Effects of various ratios of platelets to hyphae on reduction of XTT by A. fumigatus hyphae. Platelets were incubated with hyphae for 1 h in ratios ranging from 50 platelets/1 hypha to 400 platelets/1 hypha. Effects of platelets on reduction of XTT by hyphae were then determined by incubations of hyphae with XTT for the standard 1-h time period. Results represent mean ± SEM percentage decrease in XTT reduction determined in three to eight experiments, each performed with two to four replicates, except for the two experiments performed with a 400:1 ratio of platelets to hyphae. (B) Time course of XTT reduction by hyphae. After incubation with platelets for 1 h, XTT reduction by hyphae was measured after the standard 1-h period of exposure to XTT and after a second hour in XTT (•, results of incubations using a 50:1 ratio of platelets to hyphae; ▪, results of incubations using a 167:1 ratio of platelets to hyphae).

Optimum numbers of intact platelets were less effective than PMN in damaging hyphae, as 10:1 to 30:1 ratios of PMN to hyphae resulted in >65% decrements in hyphal XTT reduction. The high efficiency of PMN fungicidal effects precluded detection of any potential effects of platelets on augmenting PMN responses at optimum ratios of PMN to hyphae. Experiments performed with platelets combined with suboptimum numbers of PMN did not reveal any significant increase in damage compared to effects of PMN alone. However, incubation of organisms together with both PMN and platelets resulted in more pronounced aggregation (data not shown), so that irregular loss of hyphae prior to incubations with XTT might then have contributed to an observed high variability in results of these assays (data not shown).

DISCUSSION

These data show that platelets adhered to A. fumigatus hyphae, the invasive form of the fungus, and tended to spread over hyphal surfaces. This interaction induced platelet activation. Opsonization of hyphae with whole plasma was required for optimal platelet activation. Plasma would be readily available for any interactions in vivo, since platelets encounter hyphae only in the circulation. Invading hyphae would then be coated with plasma opsonins under such normal conditions. Plasma contains several factors that might potentially serve as opsonins and potentiate platelet adhesion to hyphae and/or activation. Despite the presence of Fc receptors on platelet surfaces (1) which could interact with antibody-coated hyphae, our data indicated that opsonization of hyphae with normal human IgG alone resulted in only weak platelet activation. Although gamma globulins from healthy individuals include anti-Aspergillus antibodies (28), this does not preclude the possibility that higher antibody titers of patients with aspergillosis might improve platelet responses. Nevertheless, our data do not support a major role for immunoglobulins in platelet-mediated host defenses in nonimmunosuppressed individuals.

Fibrinogen also binds to fungal cell surfaces (9) and has been postulated to play a significant role in platelet attachment to C. albicans (21). Nevertheless, opsonization with fibrinogen, even in concentrations exceeding those present in normal plasma, did not trigger optimal platelet activation in our studies. This was not completely surprising, as components of the active platelet fibrinogen receptor are assembled not before, but only during the platelet activation process (17). Even so, fibrinogen opsonization of hyphae still might contribute to platelet activation in vivo. Coactivation processes might lead to fibrinogen receptor assembly which, in turn, might then contribute to further platelet activation after binding its ligand.

Aspergillus hyphae also activate complement (19), and platelets carry complement surface receptors (8). However, heat inactivation of complement did not impair the ability of plasma to support optimum platelet activation.

Supplementation of serum with pooled IgG, fibrinogen, alone or in combination, failed to restore the opsonic effect of whole fresh or heat-treated plasma (Fig. 5). Thus, further studies will be required to define specific plasma factors responsible for optimum opsonization. The nature of hyphal cell wall surface constituents binding these plasma factors also remains to be defined. We have initiated studies of mechanisms involved in these processes.

Taken together, results of confocal microscopic analysis, measurements of biotinylated surface wall glycoprotein release, and changes in ability to reduce the tetrazolium dye XTT indicated that platelets inflicted significant damage to hyphae. Confocal microscopy showed loss of cell wall constituents from hyphal surfaces that were covered by platelets (Fig. 2). Hyphae without surface-adherent platelets did not evidence loss of labeled surface proteins. Assays of biotinylated surface glycoprotein release showed increased shedding of hyphal cell wall components following exposure to platelets. Biotinylation with or without labeling with fluorescent conjugate did not alter hyphal viability, ability to activate platelets, or susceptibility to damage by either platelets or neutrophils. Of course, these data did not establish that shedding of these biotinylated wall constituents necessarily represented the sole sites of platelet-mediated damage. It is certainly possible that other unmeasured constituents of the cell wall or other sites might have sustained damage by platelets as well. Likewise, these data do not allow for quantification of the relative importance of biotinylated surface glycoprotein release compared to damage to or release of other cell wall constituents. Presumably, surface proteins are more likely to undergo biotinylation than deeper wall proteins. Release of these deeper proteins, if present, might go undetected because of absent labeling.

In addition, platelets impaired the capacity of the fungi to reduce XTT, a previously established indicator of the degree of damage to A. fumigatus hyphae (22). Detrimental effects of platelets on XTT reduction by hyphae were significant with ratios of platelets to hyphae between 50:1 and 400:1. Higher ratios of platelets to hyphae resulted in progressively greater declines in XTT reduction, though the increment between ratios of 200:1 and 400:1 were minimal. Hyphae did not show signs of recovering their ability to reduce XTT when incubations of hyphae with the dye were prolonged to 2 h, beyond the usual 1-h time period. Together with previously published data on XTT as an indicator of damage to A. fumigatus hyphae (22), these effects of platelets on XTT reduction are consistent with damage to the organisms by human platelets.

Invasive aspergillosis is a vasculotropic infection (37). Our previous studies showed that PMN protected endothelial cells against invasion by C. albicans hyphae (16). Interactions between damaged endothelial cells, platelets, and PMN have been extensively studied over the last decade, focusing essentially on vessel wall injury and degeneration. Since fungi have now been shown to readily interact with vascular endothelium, circulating platelets, and circulating phagocytes, and since all of these types of cell types have modulatory effects on one another (2, 13, 15, 23), it seems crucial to study fungal interactions with each one of them. The abundance of platelets in the circulation, the circumstantial association of thrombocytopenia with the occurrence of invasive aspergillosis, and our current data showing platelet-mediated damage to hyphae suggest that these latter cells may play a significant role in normally potent coordinated host defenses against this common opportunistic infection.