523
Normal and Deficient Neutrophils Can Cooperate to Damage Aspergillus fumigatus Hyphae John H. Rex, John E. Bennett, John I. Gallin, Harry L. Malech, and David A. Melnick
From the Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland
Although relatively uncommon, invasive infections with the ubiquitous mold Aspergillus are often fatal to the individuals who develop them. Hosts deficient in either number or function of their leukocytes are at risk. This includes patients who have prolonged neutropenia from cytotoxic agents used in cancer chemotherapy or after organ transplants and patients with a major functional deficiency of their neutrophils as seen with chronic granulomatous disease (eGD) [1, 2]. Oxidative mechanisms are critical at several junctures in the host's interaction with Aspergillus. While the pulmonary alveolar macrophage that first encounters an inhaled conidia uses nonoxidative mechanisms to control or prevent the germination of the conidia [3-5], both neutrophils and monocytes have been shown to damage the hyphal form via an oxidative mechanism [6-8]. The neutrophil is clearly the key cell in host defense once conidia have germinated into hyphae. Acute infection with Aspergillus usually produces either necrosis or suppuration, and in either case neutrophils are the prominent cell in the nonneutropenic patient [9]. The ability of normal phagocytes to damage hyphae has been studied in three ways. Diamond et ale ·[10] showed electron microscopic evidence for loss ofhyphal cytoplasmic architecture in the area of phagocyte attachment. Schaffner et ale [7] demonstrated that a large number of phagocytes can sterilize a much smaller number of hyphae. Finally, decreased hyphal
Received 26 October 1989; revised 1 February 1990. Reprints and correspondence: Dr. John H. Rex, Bldg. 10, Room llNI07, National Institutes of Health, Bethesda, MD 20892. The Journal of Infectious Diseases 1990;162:523-528 This article is in the public domain. 0022-1899/90/6202-0036
metabolism following exposure to phagocytes has been demonstrated. Diamond et ale [6, 10] have shown reduction in hyphal uptake of a radiolabeled substrate, whereas Levitz and Diamond [11] showed reduced hyphal conversion of tetrazolium salts to their reduced form. To study the interactions of normal and deficient neutrophils with hyphae and with each other, we modified the test of Levitz and Diamond so that it could be used to study the effects of the neutrophil over a broad range of cell-to-hyphae ratios. This assay takes advantage of the fact that metabolically active hyphae will convert the yellow tetrazolium salt 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) to a blue formazan derivative with maximum absorbance at 560-570 nm.
Materials and Methods Fungus. A strain of Aspergillusjumigatus, originally isolated from a patient with invasive aspergillosis, was used throughout. After growth on Czapek solution agar (Difco) for 3-5 days at 30°C, conidia were harvested in phosphate-buffered saline (PBS), filtered through sterile gauze, washed twice in PBS and stored at rv1Q7/ml at 4°C. At use, a suspension of 5 x 1()4 conidia/ml was prepared by making an appropriate dilution of the concentrated conidia into yeast-nitrogen base that had been supplemented with 2 % glucose. Aliquots of 1 ml were dispensed into the 24-mm-diameter wells of a Costar plate (Costar, Cambridge, MA) and incubated at 30°C for 14-16 h until the conidia had germinated to produce 80-120 I-tm hyphae. Microscopic examination of such plates detected only occasional ungerminated conidia. Hyphal growth was stopped at this stage, if necessary, by storing the plates at 4°C. Plates were never stored for more than a few hours before use. Hyphae germinated in this manner were found to be very adherent to the bottom of the well and were not removed even by multiple washings.
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
Using a metabolic test of hyphal viability, the interaction between neutrophils and Aspergillus hyphae was investigated over a broad range of hyphae-to-neutrophil ratios. Normal neutrophils were found to damage hyphae whereas neutrophils from patients with both chronic granulomatous disease (CGD) and myeloperoxidase (MPO) deficiency did not. Further, both azide and catalase + superoxide dismutase inhibited the ability of normal neutrophils to damage hyphae, suggesting that this damage is mediated by products of the respiratory burst and by the MPO-halide system. Also, mixtures of small numbers of normal neutrophils with larger numbers of CGD neutrophils (range, 1:5 to 1:15) damaged hyphae more efficiently than either population of cells alone. Further, mixtures of CGD and MPO-deficient neutrophils, neither of which alone could efficiently damage hyphae, were able to damage the hyphae almost as well as a comparable number of normal neutrophils. These data demonstrate that intact neutrophils can cooperate to synergistically damage Aspergillus hyphae, possibly by extracellular mixing of hydrogen peroxide and MPO.
524
1ID 1990;162 (August)
Rex et al.
100 C
.. 0
'; CD
80
>
C
0
0
60
1=
:IE
-..
'0
40
C
0
0
0
20
0~
o+--.....--,---.----r--........--r-----.----.---.-----, o
2
345
Number of Normal PMNL x105 per well Figure 1. Damage to Aspergillus hyphae by normal neutrophils (PMNL). 0 to 5 X 105 normal neutrophils were added to wells in which 5 x 1()4 conidia had been germinated. Incubation time was 2 h. All points are mean ± standard error of data from three or four experiments.
were poststained with lead citrate. The grids were viewed in a Philips EM-300 at 60 kYo
Results Characterization of the assay. As Levitz and Diamond reported [11], we too found that the amount of MTT-formazan produced by Aspergillus hyphae is proportional to the initial inoculum of conidia (data not shown). Addition of up to 5 X 1Q5 neutrophils to wells containing Aspergillus hyphae gave the dose-response curve shown in figure 1. Each well was inoculated with 5 X 1()4 conidia, so it can be calculated that a readily detectable effect of neutrophils on the hyphae occurred when there were at least 5 neutrophils for each germinated conidia. As 5 X lQ5 neutrophils (i.e., 10 neutrophils for each germinated conidia) reliably produced substantial damage to the hyphae, this number of neutrophils was chosen for further studies. A time course study using 5 X lQ5 neutrophils per well is shown in figure 2 and demonstrates that the neutrophils act quickly, reducing the percentage of control MTT conversion to 47.3 % ± 4.9 % at 30 min. Ifthe hyphae were incubated with the neutrophils at 4°C, then no damage to the hyphae resulted (data not shown). In preliminary studies it was noted that normal neutrophils would almost always loosen some of the otherwise adherent hyphae from the bottom of the Costar well, resulting in their loss during the wash steps. To demonstrate that the reduced amount of MTT conversion was not due simply to loss oflive, metabolically active hyphae, the fluid from each successive wash was collected in a 15-mm conical tube, the eluted hyphae were washed twice with sterile water (to remove the deoxy-
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
Phagocytes. Neutrophils were isolated using described techniques [12], suspended in Hank's balanced salt solution (HBSS) plus 20 mMHEPES, pH 7.0, adjusted to 107/ml by counting with a Coulter counter (Coulter, Hialeah, FL), and held on ice until used. Diffquick (Harleco, Gibsontown, NJ)-stained cytospin preparations of the neutrophil suspensions typically showed >95 % neutrophils. Cell viability as assessed by trypan blue exclusion was always >98 %. Serum. All assays were done using serum from a single donor. The serum was either used fresh or stored frozen at -70°C until needed. Damage to hyphae. To Costar wells containing germinated hyphae were added 0.1 ml of serum, neutrophils, and HBSS with 20 mM HEPES, pH 7.0, to give a final volume of 1.0 ml. Wells were usually set up in triplicate. The neutrophils were gently sedimented to the bottom of the well by centrifuging at 20 g for 1 min. After incubation at 37°C, the fluid in the wells was aspirated through a 19-9auge needle and the neutrophils were lysed by adding 0.3 ml of 0.5 % sodium deoxycholate. The deoxycholate was aspirated, and the wells were then washed four times by adding 0.5 ml of sterile water, swirling gently, and aspirating. RPMI 1640 with L-glutamine, 1.0 ml, containing 0.5 mg/ml MTT was then added to each well, and the plates were incubated at 37°C for 3 h. The wells were then aspirated dry and 0.2 ml of isopropanol with 0.05 NHCl was added to each well. The acidified isopropanol was swirled gently until all of the blue precipitate had solubilized (2-3 min), and 0.15 ml of the isopropanol was transferred into an Immulon I plate (Dynatech Laboratories, Chantilly, VA). The optical density (aD) of the wells was determined on an MR600 microplate spectrophotometer (Dynatech) using the difference between a test wavelength of 570 nm and a reference wavelength of 630 nm. A well containing only 0.15 ml acidified isopropanol was used as a blank. A group of wells containing only hyphae was treated identically to the wells containing neutrophils and was always used on each plate. All data from a given plate were divided by the mean aD of that plate's hyphae-only group of wells. As a background control, a series of wells duplicating the additives in the test wells was routinely set up in a parallel Costar plate that contained no hyphae. This reagent-blank plate was then processed identically to the plates containing hyphae. The ODs of these control wells were averaged and subtracted from those of the test wells. Results were expressed as percentage of control MTT conversion as calculated by the formula: % control MTT conversion = 100 x [(mean aD in the test wells - mean aD in blank wells)/(mean aD in wells containing only hyphae - mean aD in blank wells)]. Electron microscopy. To examine the Aspergillus hyphae for evidence of ultrastructural damage, wells were taken at the end of the 2-h incubation period, aspirated dry (without lysing the neutrophils), fixed in 2 % glutaraldehyde in PBS, washed twice in 0.1 M cacodylate buffer, pH 7.4, and then fixed for 1 h in 1% osmium tetroxide/ 0.1 M cacodylate buffer, pH 7.4. The wells were then washed twice in water, incubated overnightin 7.7% uranyl acetate, and dehydrated in ethanol. The base of the well was then cut free (or, alternatively, the hyphae were grown on a coverslip in the bottom of the well, which was simply removed), and it was overlaid with Epon/Araldite and polymerized at 60°C for 2 days. The plastic was then removed from the substrate and glued with a cyanoacrylate (Krazy Glue) to an empty block of Epon/Araldite. Thin sections were picked up on formvar and carbon-coated 160-mesh hexagonal grids. Some grids
JIO 1990;162 (August)
Aspergillus and Neutrophils
525
100 C
..
100
c
.. 0
0
·iii
·iii CD
> c
> C
0
0
u
(J
1=
60
-..
40
"0
20
-
-..
40
C
0
0
(J
0
Normal Neutrophils X-linked eGO neutrophils
:E
C
-
60
1=
:E
"0
80
CD
80
(J
20
0
";;;.!!.
";;;.!!.
0
0
-..----.--......--.----.-........---~--.-------,
30
60
90
120
Time (minutes)
Table 1. Effect of inhibitors on damage by normal neutrophils. % control of MTT conversion*
None Catalase 3000 units/ml + superoxide dismutase, 150 units/ml Sodium azide, 1 rnM
15.1
± 3.3 (7)
86.3 ± 5.8 (3)t
± 6.7 (4)t * Results are given as mean ± standard error (number of experiments). t Difference from no inhibitor by Tukey test
(P
o
10
20
Number of PMNL x105 per well
Figure 2. Time course of damage to Aspergillus hyphae by normal neutrophils. 5 X 105 nprmal neutrophils were incubated with hyphae for indicated time. All points are mean ± standard error of three or four experiments.
Inhibitor, final concentration
.......-....,.---.,...........- _ o 30 40 50
O+----.......---r-~-.,....
82.0
< .01).
cholate and neutrophil debris), and then the hyphae were incubated with 1.0 ml of MTT in RPMI 1640 in a fashion analogous to the hyphae in the Costar wells. The detached hyphae were consistently found to produce at most 3 % of the MTT-formazan produced by the hyphae in the control wells (data not shown). Microscopic examination demonstrated that these detached hyphae remained largely free of blue MTT granules, indicating a major loss of metabolic activity. In sharp contrast, "normal" hyphae could always be seen to contain a substantial amount of blue granules, consistent with the active reduction of MTT. Similar results were obtained when the wash fluids were filtered across a Nuc1eopore filter (8-J.'m pore diameter), followed by incubation of the filter in MTT. Thus, the loss of hyphae following incubation with neutrophils represented loss of damaged, metabolically inactive hyphae, and was not a significant source of error in the MTT readings. Addition ofinhibitors. As shown in table 1, the addition of inhibitors to 5 X 105 normal neutrophils produced a pattern suggestive of an oxidative process. Sodium azide, a direct inhibitor of myeloperoxidase, produced a marked reduction in damage to the hyphae. Addition of superoxide dismutase
Figure 3. Lack of effect of X-linked COD neutrophils (PMNL). Results are mean ± standard error of 15 experiments (COD neutrophils) or 4 experiments (normal neutrophils).
plus catalase, enzymes that inactivate toxic 'oxygen metabolites, also produced a sharp decrease in damage to the hyphae. lilck ofdamage to hyphae by CGD neutrophils. To further evaluate the r C
...
...
'iii 100
> C 0
Gl
0
80
0
0
~ ~
~
-
'0 ...
40
0
0
20
0
'0
~
';;.!!.
0
Normal alone Normal + eGO
1
2
20
';;.!!. 0
O+------,..----r-------.----,.--~---.
o
40
c
0
0
60
~
60
- '0 ... c
80
3
4
5
Number of Normal PMNL x105 per well
0 0
10
20
30
40
50
Number of eGO PMNL X105 per well
Flgure S. Effect of mixing a large, fixed nwnber of CGD neutrophils (PMNL) with an increasing nwnber of normal neutrophils. Either normal neutrophils alone or normal neutrophils plus 25 X lOS CGD neutrophils were added to wells containing Aspergillus hyphae. Results are mean ± standard error of three experiments. Curves are different by an omnibus sign test, P < .05, where a sign test was performed for each experiment and the resulting P values were aggregated using Fisher's omnibus technique [13].
Figure 6. Effect of mixing an increasing number of CGD neutrophils (PMNL) with a fixed number of normal neutrophils. To each well was added 2 x lOS normal neutrophils in addition to indicated nwnber of CGD neutrophils. All points are mean ± standard error of two or three experiments. * different from wells with no CGD neutrophils at P < .05 by Dunnett test.
there is a decline in the ability of the mixture to damage the hyphae. This effect is probably due to displacement of the nonnal neutrophils from the hyphae by the dysfunctional can neutrophils.
of their cells to damage hyphae was studied (figure 7), While neither the can neutrophils nor the myeloperoxidase-deficient neutrophils could, in isolation, efficiently damage hyphae, the mixture of these deficient cell types gave a dose-response curve similar to that produced by normal neutrophils. In contrast with these data, we found that a mixture of two types of can neutrophils, each with a genetically distinct defect, could not cooperate to produce damage to the hyphae. In this experiment, neutrophils from a patient with autosomal recessive
Effect ofmixing myeloperoxidase-deficient neutrophils and CGD neutrophils. A patient with well-documented complete myeloperoxidase-deficiency [14] was studied simultaneously with two different X-linked can patients. In addition to studying their cells alone, the ability of mixtures of equal numbers
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
F1gure 4. Transmission electronmicrographs of neutrophils adherent to hyphae. Both parts (X16,800) show a neutrophil (top) and a hypha (bottom). A, CGD neutrophil associated with a normal-appearing hypha. Note well-preserved internal organelles of hypha. B, Normal neutrophil associated with a severely damaged hypha that has lost essentially all of its internal structure. Bar = I ",m,
Aspergillus and Neutrophils
JID 1990;162 (August)
140
c:
.. o
'iii CD
> c:
120 100
o
o
~
80
e-c:
60
:E
o
o '5 '#-
40
eGO
20
MPO-deficient 1:1 CGO:Mpo-deficient Normal
o
10
20
30
40
50
Number of PMNL x105 per well Figure 7. Effect of mixing CGD and myeloperoxidase (MPO)deficient neutrophils (PMNL). Indicated amounts of individual cell types or a 1:1 mixture of MPO-deficient and CGD neutrophils were added to wells. Results are mean ± standard error of four experiments.
eGO (deficient in a cytosolic 47-kDa factor) and neutrophils from one of the X-linked eGO patients (deficient in membrane-bound cytochrome b558 ) were combined (data not shown). The results provide evidence that, even in the close proximity of attachment to hyphae, neutrophils from patients with different forms of eGO cannot exchange cytosolic or membrane-bound components to reconstitute their oxidative pathways. As a further control for the mixing experiments, the eGO and myeloperoxidase-deficient neutrophils were mixed indirectly by sequential application. For these experiments, 2.5 x 105 , 5 x lOS, or 7.5 X 105 neutrophils of one type were incubated for 1 h with the hyphae. These neutrophils were then removed by lysis with sterile distilled water (three washes of 5 min each; complete lysis verified visually), and then an equivalent number of neutrophils of the other type were added for 1 h. This second group of neutrophils was removed by lysis with the sodium deoxycholate, and the plates were then handled in the standard fashion. No significant reduction in conversion ofMTT was seen with either sequence of application (i.e., myeloperoxidase-deficient first, eGO second, or vice versa). Discussion The neutrophil plays a central role in host defense against infection by Aspergillus and other fungi. Unlike bacteria and yeast, which can be ingested and killed by a single neutrophil, Aspergillus hyphae present a large, nonphagocytosable surface to which multiple neutrophils can adhere. The mechanism by which neutrophils kill such extracellular pathogens
remains uncertain. Further, the level of interaction between distinct cells adherent to the same target is not known. We have studied the ability of normal and genetically deficient neutrophils to damage Aspergillus hyphae. Our data suggest that distinct populations of neutrophils can act synergistically to damage an extracellular target. The MTT reduction assay, first described by Levitz and Diamond [11], proved to be a simple and reproducible means of quantitating the effect of neutrophils on hyphae. While all metabolic assays have the common problem that they measure damage rather than killing, several lines of evidence suggest that this assay reflects a serious, possibly lethal, injury to hyphae. First, using other techniques, others have shown that neutrophils can both alter the ultrastructure of the hyphae [10] and completely sterilize a small inoculum of hyphae [7]. Second, incubation of the hyphae with the neutrophils at 4°e resulted in no damage to the hyphae. Third, neutrophils with genetic defects affecting microbicidal activity showed defects in their ability to affect the MTT assay. Finally, electron micro.graphs confirmed that incubation of hyphae with normal neutrophils, but not eGO neutrophils, resulted In loss of internal structure of the hyphae. The MTT reduction assay thus appears to be a rapid, simple, and reproducible means of quantitating fungal viability. In agreement with others [6], we found that superoxide dismutase plus catalase, and the myeloperoxidase inhibitor sodium azide, had a protective effect on Aspergillus hyphae during incubation with normal human neutrophils. Similarly, neutrophils with a genetic deficiency of cytochrome b558 (Xlinked chronic granulomatous disease), which cannot generate a respiratory burst, did not affect fungal viability as measured by MTT reduction. Also, myeloperoxidase-deficient neutrophils were unable to efficiently damage hyphae. Taken together, these data strongly suggest that the respiratory burst and the myeloperoxidase-halide system playa key role in neutrophil-mediated damage to these nonphagocytosable pathogens. The current data demonstrate that neutrophils with different defects in their microbicidal mechanisms can interact to damage Aspergillus hyphae. The addition of eGO neutrophils enhanced the antifungal effect of normal neutrophils when submaximal numbers of normal cells were used. Maximum enhancement was seen with ratios of eGO-to-normal PMNL between 5: 1 and 15: 1. Higher ratios resulted in decreased enhancement, perhaps due to crowding out of normal cells during competition for binding to hyphae. This synergistic interaction was strikingly demonstrated by the combination of myeloperoxidase-deficient and eGO neutrophils. While neither cell type alone could effectively damage hyphae, a 1:1 mixture of the deficient cells produced effects that closely paralleled those of normal neutrophils. This effect was not reproduced when the deficient cell types were allowed to interact sequentially with the hyphae, indicating that both
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
o-l---~~::;:==:::;:==~=~
527
528
Rex et al.
Acknowledgment We thank Ellen DeCarlo, Robert Karp, Philip Karp, Linda Herzman, and the nurses of 11 East for their contributions, John Albert
for electron microscopy, and David Alling and Steve Banks for statistical assistance.
References 1. Rippon JWR. Medical mycology, 3rd ed. Philadelphia: W. B. Saunders, 1988:618-650 2. Gallin JI, Buescher ES, Seligmann BE, Nath J, Gaither T, Katz P. Recent advances in chronic granulomatous disease. Ann Intern Med 1983;99:657-674 3. Schaffner A, Douglas H, Braude AI, Davis CEo Killing of Aspergillus spores depends on the anatomical source of the macrophage. Infect Immun 1983;42:1109-1115 4. Levitz SM, Selsted ME, Ganz T, Lehrer RI, Diamond RD. In vitro killing of spores and hyphae of Aspergillusfumigatus and Rhizopus oryw.e by rabbit neutrophil cationic peptides and bronchoalveolar macrophages. J Infect Dis 1986;154:483-489 5. Levitz SM, DiBenedetto DJ. Paradoxical role ofcapsule in murine bronchoalveolar macrophage-mediated killing of Cryptococcus neoformans. J Immunol 1989;142:659-665 6. Diamond RD, Clark RA. Damage to Aspergillus fumigatus and Rhizopus oryzae hyphae by oxidative and nonoxidative microbicidal products of human neutrophils in vitro. Infect Immun 1982;38:487-495 7. Schaffner A, Douglas H, Braude A. Selective protection against conidia by mononuclear and against mycelia by polymorphonuclear phagocytes in resistance to Aspergillus. J Clin Invest 1982;69:617-631 8. Washburn RG, Gallin JI, Bennett JE. Oxidative killing of Aspergillus fumigatus proceeds by parallel myeloperoxidase-dependent and -independent pathways. Infect Immun 1987;55:2088-2092 9. Pena CEo Aspergillosis. In: Baker RD, ed. Human infection with fungi, actinomycetes and algae. New York: Springer-Verlag, 1971:762-831 10. Diamond RD, Krzesicki R, Epstein B, Jao W. Damage to hyphal forms of fungi by human leukocytes in vitro. AmJ PathoI1978;91:313-328 11. Levitz SM, Diamond RD. A rapid colorimetric assay of fungal viability with the tetrazolium salt MTT. J Infect Dis 1985;152:938-945 12. B~yum A. Isolation of lymphocytes, granulocytes, and macrophages. Scand J Immunol Suppl 1976;5:9-15 13. Fisher RA. Statistical methods for research workers. 11th ed. London: Oliver and Boyd, 1950:99-101 14. Nauseff WM, Root RK, Malech HL. Biochemical and immunologic analysis of hereditary myeloperoxidase deficiency. J Clin Invest 1983; 71:1297-1307 15. Hamers MN, de Boer M, MeerhofLJ, Weening RS, Roos D. Complementation in monocyte hybrids revealing genetic heterogeneity in chronic granulomatous disease. Nature 1984;307:553-555 16. Buescher ES, Gallin JI. Leukocyte transfusion in chronic granulomatous disease. N Engl J Med 1982;13:800-803 17. Quie P. The white cells: use of granulocyte transfusions. Rev Infect Dis 1987;9:189-193
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
cell types must be present simultaneously. A plausible mechanism for this cooperative antifungal effect is that the myeloperoxidase-deficient cells are stimulated to generate hydrogen peroxide, which is then used by the eGO neutrophils to damage hyphae via the myeloperoxidase-halide system. This might occur via extracellular transfer of hydrogen peroxide from myeloperoxidase-deficient to eGO neutrophils. Alternatively, extracellular mixing of hydrogen peroxide and myeloperoxidase might occur after degranulation of the eGO neutrophils. The failure of autosomal recessive eGO cells (lacking a 47kDa cytosolic phosphoprotein) to complement the function of X-linked eGO cells (lacking cytochrome b558 ) suggests that this interaction does not occur at the level of formation of a syncytium allowing exchange of cytosolic components. Such cellular complementation has been described in somatic cell hybrids between cultured cell lines derived from autosomal recessive and X-linked eGO donors [15]. This work provides theoretical support for the use of white blood cell transfusions in the treatment of serious, refractory Aspergillus infections in patients with circulating dysfunctional neutrophils. Although the transfused cells can persist up to 42 h in the sputum [16], no study has documented conclusive evidence that these transfusions are of benefit [17]. It is also interesting that the mothers of patients with X-linked eGO may be healthy with as few as 5 % circulating normal neutrophils (Malech and Gallin, unpublished data). In conclusion, this work further describes and validates a simple reproducible assay of hyphal viability that can be used to quantitate the effect of neutrophils on Aspergillus hyphae. Using this assay, we demonstrated that mixtures of normal and defective (X-linked eGO) neutrophils and mixtures of different types of defective neutrophils (myeloperoxidasedeficient and X-linked eGO) can cooperate to damage Aspergillus hyphae, and that this may proceed via extracellular mixing of toxic oxidative products and myeloperoxidase. Of particular interest is the observation that a small number of normal neutrophils can cooperate with a larger number of dysfunctional neutrophils to produce effective damage of Aspergillus hyphae in our assay system. This is the first in vitro observation to support the use of white blood cell transfusions in eGO patients with refractory Aspergillus infections.
JID 1990;162 (August)
Normal and Deficient Neutrophils Can Cooperate to Damage Aspergillus fumigatus Hyphae John H. Rex, John E. Bennett, John I. Gallin, Harry L. Malech, and David A. Melnick
From the Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland
Although relatively uncommon, invasive infections with the ubiquitous mold Aspergillus are often fatal to the individuals who develop them. Hosts deficient in either number or function of their leukocytes are at risk. This includes patients who have prolonged neutropenia from cytotoxic agents used in cancer chemotherapy or after organ transplants and patients with a major functional deficiency of their neutrophils as seen with chronic granulomatous disease (eGD) [1, 2]. Oxidative mechanisms are critical at several junctures in the host's interaction with Aspergillus. While the pulmonary alveolar macrophage that first encounters an inhaled conidia uses nonoxidative mechanisms to control or prevent the germination of the conidia [3-5], both neutrophils and monocytes have been shown to damage the hyphal form via an oxidative mechanism [6-8]. The neutrophil is clearly the key cell in host defense once conidia have germinated into hyphae. Acute infection with Aspergillus usually produces either necrosis or suppuration, and in either case neutrophils are the prominent cell in the nonneutropenic patient [9]. The ability of normal phagocytes to damage hyphae has been studied in three ways. Diamond et ale ·[10] showed electron microscopic evidence for loss ofhyphal cytoplasmic architecture in the area of phagocyte attachment. Schaffner et ale [7] demonstrated that a large number of phagocytes can sterilize a much smaller number of hyphae. Finally, decreased hyphal
Received 26 October 1989; revised 1 February 1990. Reprints and correspondence: Dr. John H. Rex, Bldg. 10, Room llNI07, National Institutes of Health, Bethesda, MD 20892. The Journal of Infectious Diseases 1990;162:523-528 This article is in the public domain. 0022-1899/90/6202-0036
metabolism following exposure to phagocytes has been demonstrated. Diamond et ale [6, 10] have shown reduction in hyphal uptake of a radiolabeled substrate, whereas Levitz and Diamond [11] showed reduced hyphal conversion of tetrazolium salts to their reduced form. To study the interactions of normal and deficient neutrophils with hyphae and with each other, we modified the test of Levitz and Diamond so that it could be used to study the effects of the neutrophil over a broad range of cell-to-hyphae ratios. This assay takes advantage of the fact that metabolically active hyphae will convert the yellow tetrazolium salt 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) to a blue formazan derivative with maximum absorbance at 560-570 nm.
Materials and Methods Fungus. A strain of Aspergillusjumigatus, originally isolated from a patient with invasive aspergillosis, was used throughout. After growth on Czapek solution agar (Difco) for 3-5 days at 30°C, conidia were harvested in phosphate-buffered saline (PBS), filtered through sterile gauze, washed twice in PBS and stored at rv1Q7/ml at 4°C. At use, a suspension of 5 x 1()4 conidia/ml was prepared by making an appropriate dilution of the concentrated conidia into yeast-nitrogen base that had been supplemented with 2 % glucose. Aliquots of 1 ml were dispensed into the 24-mm-diameter wells of a Costar plate (Costar, Cambridge, MA) and incubated at 30°C for 14-16 h until the conidia had germinated to produce 80-120 I-tm hyphae. Microscopic examination of such plates detected only occasional ungerminated conidia. Hyphal growth was stopped at this stage, if necessary, by storing the plates at 4°C. Plates were never stored for more than a few hours before use. Hyphae germinated in this manner were found to be very adherent to the bottom of the well and were not removed even by multiple washings.
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
Using a metabolic test of hyphal viability, the interaction between neutrophils and Aspergillus hyphae was investigated over a broad range of hyphae-to-neutrophil ratios. Normal neutrophils were found to damage hyphae whereas neutrophils from patients with both chronic granulomatous disease (CGD) and myeloperoxidase (MPO) deficiency did not. Further, both azide and catalase + superoxide dismutase inhibited the ability of normal neutrophils to damage hyphae, suggesting that this damage is mediated by products of the respiratory burst and by the MPO-halide system. Also, mixtures of small numbers of normal neutrophils with larger numbers of CGD neutrophils (range, 1:5 to 1:15) damaged hyphae more efficiently than either population of cells alone. Further, mixtures of CGD and MPO-deficient neutrophils, neither of which alone could efficiently damage hyphae, were able to damage the hyphae almost as well as a comparable number of normal neutrophils. These data demonstrate that intact neutrophils can cooperate to synergistically damage Aspergillus hyphae, possibly by extracellular mixing of hydrogen peroxide and MPO.
524
1ID 1990;162 (August)
Rex et al.
100 C
.. 0
'; CD
80
>
C
0
0
60
1=
:IE
-..
'0
40
C
0
0
0
20
0~
o+--.....--,---.----r--........--r-----.----.---.-----, o
2
345
Number of Normal PMNL x105 per well Figure 1. Damage to Aspergillus hyphae by normal neutrophils (PMNL). 0 to 5 X 105 normal neutrophils were added to wells in which 5 x 1()4 conidia had been germinated. Incubation time was 2 h. All points are mean ± standard error of data from three or four experiments.
were poststained with lead citrate. The grids were viewed in a Philips EM-300 at 60 kYo
Results Characterization of the assay. As Levitz and Diamond reported [11], we too found that the amount of MTT-formazan produced by Aspergillus hyphae is proportional to the initial inoculum of conidia (data not shown). Addition of up to 5 X 1Q5 neutrophils to wells containing Aspergillus hyphae gave the dose-response curve shown in figure 1. Each well was inoculated with 5 X 1()4 conidia, so it can be calculated that a readily detectable effect of neutrophils on the hyphae occurred when there were at least 5 neutrophils for each germinated conidia. As 5 X lQ5 neutrophils (i.e., 10 neutrophils for each germinated conidia) reliably produced substantial damage to the hyphae, this number of neutrophils was chosen for further studies. A time course study using 5 X lQ5 neutrophils per well is shown in figure 2 and demonstrates that the neutrophils act quickly, reducing the percentage of control MTT conversion to 47.3 % ± 4.9 % at 30 min. Ifthe hyphae were incubated with the neutrophils at 4°C, then no damage to the hyphae resulted (data not shown). In preliminary studies it was noted that normal neutrophils would almost always loosen some of the otherwise adherent hyphae from the bottom of the Costar well, resulting in their loss during the wash steps. To demonstrate that the reduced amount of MTT conversion was not due simply to loss oflive, metabolically active hyphae, the fluid from each successive wash was collected in a 15-mm conical tube, the eluted hyphae were washed twice with sterile water (to remove the deoxy-
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
Phagocytes. Neutrophils were isolated using described techniques [12], suspended in Hank's balanced salt solution (HBSS) plus 20 mMHEPES, pH 7.0, adjusted to 107/ml by counting with a Coulter counter (Coulter, Hialeah, FL), and held on ice until used. Diffquick (Harleco, Gibsontown, NJ)-stained cytospin preparations of the neutrophil suspensions typically showed >95 % neutrophils. Cell viability as assessed by trypan blue exclusion was always >98 %. Serum. All assays were done using serum from a single donor. The serum was either used fresh or stored frozen at -70°C until needed. Damage to hyphae. To Costar wells containing germinated hyphae were added 0.1 ml of serum, neutrophils, and HBSS with 20 mM HEPES, pH 7.0, to give a final volume of 1.0 ml. Wells were usually set up in triplicate. The neutrophils were gently sedimented to the bottom of the well by centrifuging at 20 g for 1 min. After incubation at 37°C, the fluid in the wells was aspirated through a 19-9auge needle and the neutrophils were lysed by adding 0.3 ml of 0.5 % sodium deoxycholate. The deoxycholate was aspirated, and the wells were then washed four times by adding 0.5 ml of sterile water, swirling gently, and aspirating. RPMI 1640 with L-glutamine, 1.0 ml, containing 0.5 mg/ml MTT was then added to each well, and the plates were incubated at 37°C for 3 h. The wells were then aspirated dry and 0.2 ml of isopropanol with 0.05 NHCl was added to each well. The acidified isopropanol was swirled gently until all of the blue precipitate had solubilized (2-3 min), and 0.15 ml of the isopropanol was transferred into an Immulon I plate (Dynatech Laboratories, Chantilly, VA). The optical density (aD) of the wells was determined on an MR600 microplate spectrophotometer (Dynatech) using the difference between a test wavelength of 570 nm and a reference wavelength of 630 nm. A well containing only 0.15 ml acidified isopropanol was used as a blank. A group of wells containing only hyphae was treated identically to the wells containing neutrophils and was always used on each plate. All data from a given plate were divided by the mean aD of that plate's hyphae-only group of wells. As a background control, a series of wells duplicating the additives in the test wells was routinely set up in a parallel Costar plate that contained no hyphae. This reagent-blank plate was then processed identically to the plates containing hyphae. The ODs of these control wells were averaged and subtracted from those of the test wells. Results were expressed as percentage of control MTT conversion as calculated by the formula: % control MTT conversion = 100 x [(mean aD in the test wells - mean aD in blank wells)/(mean aD in wells containing only hyphae - mean aD in blank wells)]. Electron microscopy. To examine the Aspergillus hyphae for evidence of ultrastructural damage, wells were taken at the end of the 2-h incubation period, aspirated dry (without lysing the neutrophils), fixed in 2 % glutaraldehyde in PBS, washed twice in 0.1 M cacodylate buffer, pH 7.4, and then fixed for 1 h in 1% osmium tetroxide/ 0.1 M cacodylate buffer, pH 7.4. The wells were then washed twice in water, incubated overnightin 7.7% uranyl acetate, and dehydrated in ethanol. The base of the well was then cut free (or, alternatively, the hyphae were grown on a coverslip in the bottom of the well, which was simply removed), and it was overlaid with Epon/Araldite and polymerized at 60°C for 2 days. The plastic was then removed from the substrate and glued with a cyanoacrylate (Krazy Glue) to an empty block of Epon/Araldite. Thin sections were picked up on formvar and carbon-coated 160-mesh hexagonal grids. Some grids
JIO 1990;162 (August)
Aspergillus and Neutrophils
525
100 C
..
100
c
.. 0
0
·iii
·iii CD
> c
> C
0
0
u
(J
1=
60
-..
40
"0
20
-
-..
40
C
0
0
(J
0
Normal Neutrophils X-linked eGO neutrophils
:E
C
-
60
1=
:E
"0
80
CD
80
(J
20
0
";;;.!!.
";;;.!!.
0
0
-..----.--......--.----.-........---~--.-------,
30
60
90
120
Time (minutes)
Table 1. Effect of inhibitors on damage by normal neutrophils. % control of MTT conversion*
None Catalase 3000 units/ml + superoxide dismutase, 150 units/ml Sodium azide, 1 rnM
15.1
± 3.3 (7)
86.3 ± 5.8 (3)t
± 6.7 (4)t * Results are given as mean ± standard error (number of experiments). t Difference from no inhibitor by Tukey test
(P
o
10
20
Number of PMNL x105 per well
Figure 2. Time course of damage to Aspergillus hyphae by normal neutrophils. 5 X 105 nprmal neutrophils were incubated with hyphae for indicated time. All points are mean ± standard error of three or four experiments.
Inhibitor, final concentration
.......-....,.---.,...........- _ o 30 40 50
O+----.......---r-~-.,....
82.0
< .01).
cholate and neutrophil debris), and then the hyphae were incubated with 1.0 ml of MTT in RPMI 1640 in a fashion analogous to the hyphae in the Costar wells. The detached hyphae were consistently found to produce at most 3 % of the MTT-formazan produced by the hyphae in the control wells (data not shown). Microscopic examination demonstrated that these detached hyphae remained largely free of blue MTT granules, indicating a major loss of metabolic activity. In sharp contrast, "normal" hyphae could always be seen to contain a substantial amount of blue granules, consistent with the active reduction of MTT. Similar results were obtained when the wash fluids were filtered across a Nuc1eopore filter (8-J.'m pore diameter), followed by incubation of the filter in MTT. Thus, the loss of hyphae following incubation with neutrophils represented loss of damaged, metabolically inactive hyphae, and was not a significant source of error in the MTT readings. Addition ofinhibitors. As shown in table 1, the addition of inhibitors to 5 X 105 normal neutrophils produced a pattern suggestive of an oxidative process. Sodium azide, a direct inhibitor of myeloperoxidase, produced a marked reduction in damage to the hyphae. Addition of superoxide dismutase
Figure 3. Lack of effect of X-linked COD neutrophils (PMNL). Results are mean ± standard error of 15 experiments (COD neutrophils) or 4 experiments (normal neutrophils).
plus catalase, enzymes that inactivate toxic 'oxygen metabolites, also produced a sharp decrease in damage to the hyphae. lilck ofdamage to hyphae by CGD neutrophils. To further evaluate the r C
...
...
'iii 100
> C 0
Gl
0
80
0
0
~ ~
~
-
'0 ...
40
0
0
20
0
'0
~
';;.!!.
0
Normal alone Normal + eGO
1
2
20
';;.!!. 0
O+------,..----r-------.----,.--~---.
o
40
c
0
0
60
~
60
- '0 ... c
80
3
4
5
Number of Normal PMNL x105 per well
0 0
10
20
30
40
50
Number of eGO PMNL X105 per well
Flgure S. Effect of mixing a large, fixed nwnber of CGD neutrophils (PMNL) with an increasing nwnber of normal neutrophils. Either normal neutrophils alone or normal neutrophils plus 25 X lOS CGD neutrophils were added to wells containing Aspergillus hyphae. Results are mean ± standard error of three experiments. Curves are different by an omnibus sign test, P < .05, where a sign test was performed for each experiment and the resulting P values were aggregated using Fisher's omnibus technique [13].
Figure 6. Effect of mixing an increasing number of CGD neutrophils (PMNL) with a fixed number of normal neutrophils. To each well was added 2 x lOS normal neutrophils in addition to indicated nwnber of CGD neutrophils. All points are mean ± standard error of two or three experiments. * different from wells with no CGD neutrophils at P < .05 by Dunnett test.
there is a decline in the ability of the mixture to damage the hyphae. This effect is probably due to displacement of the nonnal neutrophils from the hyphae by the dysfunctional can neutrophils.
of their cells to damage hyphae was studied (figure 7), While neither the can neutrophils nor the myeloperoxidase-deficient neutrophils could, in isolation, efficiently damage hyphae, the mixture of these deficient cell types gave a dose-response curve similar to that produced by normal neutrophils. In contrast with these data, we found that a mixture of two types of can neutrophils, each with a genetically distinct defect, could not cooperate to produce damage to the hyphae. In this experiment, neutrophils from a patient with autosomal recessive
Effect ofmixing myeloperoxidase-deficient neutrophils and CGD neutrophils. A patient with well-documented complete myeloperoxidase-deficiency [14] was studied simultaneously with two different X-linked can patients. In addition to studying their cells alone, the ability of mixtures of equal numbers
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
F1gure 4. Transmission electronmicrographs of neutrophils adherent to hyphae. Both parts (X16,800) show a neutrophil (top) and a hypha (bottom). A, CGD neutrophil associated with a normal-appearing hypha. Note well-preserved internal organelles of hypha. B, Normal neutrophil associated with a severely damaged hypha that has lost essentially all of its internal structure. Bar = I ",m,
Aspergillus and Neutrophils
JID 1990;162 (August)
140
c:
.. o
'iii CD
> c:
120 100
o
o
~
80
e-c:
60
:E
o
o '5 '#-
40
eGO
20
MPO-deficient 1:1 CGO:Mpo-deficient Normal
o
10
20
30
40
50
Number of PMNL x105 per well Figure 7. Effect of mixing CGD and myeloperoxidase (MPO)deficient neutrophils (PMNL). Indicated amounts of individual cell types or a 1:1 mixture of MPO-deficient and CGD neutrophils were added to wells. Results are mean ± standard error of four experiments.
eGO (deficient in a cytosolic 47-kDa factor) and neutrophils from one of the X-linked eGO patients (deficient in membrane-bound cytochrome b558 ) were combined (data not shown). The results provide evidence that, even in the close proximity of attachment to hyphae, neutrophils from patients with different forms of eGO cannot exchange cytosolic or membrane-bound components to reconstitute their oxidative pathways. As a further control for the mixing experiments, the eGO and myeloperoxidase-deficient neutrophils were mixed indirectly by sequential application. For these experiments, 2.5 x 105 , 5 x lOS, or 7.5 X 105 neutrophils of one type were incubated for 1 h with the hyphae. These neutrophils were then removed by lysis with sterile distilled water (three washes of 5 min each; complete lysis verified visually), and then an equivalent number of neutrophils of the other type were added for 1 h. This second group of neutrophils was removed by lysis with the sodium deoxycholate, and the plates were then handled in the standard fashion. No significant reduction in conversion ofMTT was seen with either sequence of application (i.e., myeloperoxidase-deficient first, eGO second, or vice versa). Discussion The neutrophil plays a central role in host defense against infection by Aspergillus and other fungi. Unlike bacteria and yeast, which can be ingested and killed by a single neutrophil, Aspergillus hyphae present a large, nonphagocytosable surface to which multiple neutrophils can adhere. The mechanism by which neutrophils kill such extracellular pathogens
remains uncertain. Further, the level of interaction between distinct cells adherent to the same target is not known. We have studied the ability of normal and genetically deficient neutrophils to damage Aspergillus hyphae. Our data suggest that distinct populations of neutrophils can act synergistically to damage an extracellular target. The MTT reduction assay, first described by Levitz and Diamond [11], proved to be a simple and reproducible means of quantitating the effect of neutrophils on hyphae. While all metabolic assays have the common problem that they measure damage rather than killing, several lines of evidence suggest that this assay reflects a serious, possibly lethal, injury to hyphae. First, using other techniques, others have shown that neutrophils can both alter the ultrastructure of the hyphae [10] and completely sterilize a small inoculum of hyphae [7]. Second, incubation of the hyphae with the neutrophils at 4°e resulted in no damage to the hyphae. Third, neutrophils with genetic defects affecting microbicidal activity showed defects in their ability to affect the MTT assay. Finally, electron micro.graphs confirmed that incubation of hyphae with normal neutrophils, but not eGO neutrophils, resulted In loss of internal structure of the hyphae. The MTT reduction assay thus appears to be a rapid, simple, and reproducible means of quantitating fungal viability. In agreement with others [6], we found that superoxide dismutase plus catalase, and the myeloperoxidase inhibitor sodium azide, had a protective effect on Aspergillus hyphae during incubation with normal human neutrophils. Similarly, neutrophils with a genetic deficiency of cytochrome b558 (Xlinked chronic granulomatous disease), which cannot generate a respiratory burst, did not affect fungal viability as measured by MTT reduction. Also, myeloperoxidase-deficient neutrophils were unable to efficiently damage hyphae. Taken together, these data strongly suggest that the respiratory burst and the myeloperoxidase-halide system playa key role in neutrophil-mediated damage to these nonphagocytosable pathogens. The current data demonstrate that neutrophils with different defects in their microbicidal mechanisms can interact to damage Aspergillus hyphae. The addition of eGO neutrophils enhanced the antifungal effect of normal neutrophils when submaximal numbers of normal cells were used. Maximum enhancement was seen with ratios of eGO-to-normal PMNL between 5: 1 and 15: 1. Higher ratios resulted in decreased enhancement, perhaps due to crowding out of normal cells during competition for binding to hyphae. This synergistic interaction was strikingly demonstrated by the combination of myeloperoxidase-deficient and eGO neutrophils. While neither cell type alone could effectively damage hyphae, a 1:1 mixture of the deficient cells produced effects that closely paralleled those of normal neutrophils. This effect was not reproduced when the deficient cell types were allowed to interact sequentially with the hyphae, indicating that both
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
o-l---~~::;:==:::;:==~=~
527
528
Rex et al.
Acknowledgment We thank Ellen DeCarlo, Robert Karp, Philip Karp, Linda Herzman, and the nurses of 11 East for their contributions, John Albert
for electron microscopy, and David Alling and Steve Banks for statistical assistance.
References 1. Rippon JWR. Medical mycology, 3rd ed. Philadelphia: W. B. Saunders, 1988:618-650 2. Gallin JI, Buescher ES, Seligmann BE, Nath J, Gaither T, Katz P. Recent advances in chronic granulomatous disease. Ann Intern Med 1983;99:657-674 3. Schaffner A, Douglas H, Braude AI, Davis CEo Killing of Aspergillus spores depends on the anatomical source of the macrophage. Infect Immun 1983;42:1109-1115 4. Levitz SM, Selsted ME, Ganz T, Lehrer RI, Diamond RD. In vitro killing of spores and hyphae of Aspergillusfumigatus and Rhizopus oryw.e by rabbit neutrophil cationic peptides and bronchoalveolar macrophages. J Infect Dis 1986;154:483-489 5. Levitz SM, DiBenedetto DJ. Paradoxical role ofcapsule in murine bronchoalveolar macrophage-mediated killing of Cryptococcus neoformans. J Immunol 1989;142:659-665 6. Diamond RD, Clark RA. Damage to Aspergillus fumigatus and Rhizopus oryzae hyphae by oxidative and nonoxidative microbicidal products of human neutrophils in vitro. Infect Immun 1982;38:487-495 7. Schaffner A, Douglas H, Braude A. Selective protection against conidia by mononuclear and against mycelia by polymorphonuclear phagocytes in resistance to Aspergillus. J Clin Invest 1982;69:617-631 8. Washburn RG, Gallin JI, Bennett JE. Oxidative killing of Aspergillus fumigatus proceeds by parallel myeloperoxidase-dependent and -independent pathways. Infect Immun 1987;55:2088-2092 9. Pena CEo Aspergillosis. In: Baker RD, ed. Human infection with fungi, actinomycetes and algae. New York: Springer-Verlag, 1971:762-831 10. Diamond RD, Krzesicki R, Epstein B, Jao W. Damage to hyphal forms of fungi by human leukocytes in vitro. AmJ PathoI1978;91:313-328 11. Levitz SM, Diamond RD. A rapid colorimetric assay of fungal viability with the tetrazolium salt MTT. J Infect Dis 1985;152:938-945 12. B~yum A. Isolation of lymphocytes, granulocytes, and macrophages. Scand J Immunol Suppl 1976;5:9-15 13. Fisher RA. Statistical methods for research workers. 11th ed. London: Oliver and Boyd, 1950:99-101 14. Nauseff WM, Root RK, Malech HL. Biochemical and immunologic analysis of hereditary myeloperoxidase deficiency. J Clin Invest 1983; 71:1297-1307 15. Hamers MN, de Boer M, MeerhofLJ, Weening RS, Roos D. Complementation in monocyte hybrids revealing genetic heterogeneity in chronic granulomatous disease. Nature 1984;307:553-555 16. Buescher ES, Gallin JI. Leukocyte transfusion in chronic granulomatous disease. N Engl J Med 1982;13:800-803 17. Quie P. The white cells: use of granulocyte transfusions. Rev Infect Dis 1987;9:189-193
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 21, 2015
cell types must be present simultaneously. A plausible mechanism for this cooperative antifungal effect is that the myeloperoxidase-deficient cells are stimulated to generate hydrogen peroxide, which is then used by the eGO neutrophils to damage hyphae via the myeloperoxidase-halide system. This might occur via extracellular transfer of hydrogen peroxide from myeloperoxidase-deficient to eGO neutrophils. Alternatively, extracellular mixing of hydrogen peroxide and myeloperoxidase might occur after degranulation of the eGO neutrophils. The failure of autosomal recessive eGO cells (lacking a 47kDa cytosolic phosphoprotein) to complement the function of X-linked eGO cells (lacking cytochrome b558 ) suggests that this interaction does not occur at the level of formation of a syncytium allowing exchange of cytosolic components. Such cellular complementation has been described in somatic cell hybrids between cultured cell lines derived from autosomal recessive and X-linked eGO donors [15]. This work provides theoretical support for the use of white blood cell transfusions in the treatment of serious, refractory Aspergillus infections in patients with circulating dysfunctional neutrophils. Although the transfused cells can persist up to 42 h in the sputum [16], no study has documented conclusive evidence that these transfusions are of benefit [17]. It is also interesting that the mothers of patients with X-linked eGO may be healthy with as few as 5 % circulating normal neutrophils (Malech and Gallin, unpublished data). In conclusion, this work further describes and validates a simple reproducible assay of hyphal viability that can be used to quantitate the effect of neutrophils on Aspergillus hyphae. Using this assay, we demonstrated that mixtures of normal and defective (X-linked eGO) neutrophils and mixtures of different types of defective neutrophils (myeloperoxidasedeficient and X-linked eGO) can cooperate to damage Aspergillus hyphae, and that this may proceed via extracellular mixing of toxic oxidative products and myeloperoxidase. Of particular interest is the observation that a small number of normal neutrophils can cooperate with a larger number of dysfunctional neutrophils to produce effective damage of Aspergillus hyphae in our assay system. This is the first in vitro observation to support the use of white blood cell transfusions in eGO patients with refractory Aspergillus infections.
JID 1990;162 (August)
