INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

ALVEOLAR MACROPHAGES (AM) constitute an important first line of host defense against inhaled viable and nonviable particles. AM are regarded to be very robust phagocytic cells. However, recent studies have shown that their phagocytic activity can be markedly impaired by small amounts of ingested aggregates of ultrafine carbon particles (0.2 µg/10⁶ AM) and also by long-term (22-44 h) incubation with low concentrations of interferon- (IFN-; 12.5 U/ml) (17, 18). These results are of particular interest as epidemiological studies have shown a correlation between moderate levels of particles in the ambient air and acute effects such as mortality in heart and lung diseases and chronic lung morbidity (1, 4, 6, 21). Ingested ambient air particles in AM may impair the phagocytic capacity of these cells after an episode of high particle concentration, and this impairment should be of particular importance for persons with infections, which increase the production of IFN- (2, 5).

Results showing that the phagocytic process by AM is vulnerable have increased the interest in systematic studies of this process. Hed (13) described a method for investigation of phagocytosis by using fluorescein-labeled test particles. The numbers of particles attached to the cell surface (blue stained and fluorescence quenched by trypan blue) and particles ingested by the cell (fluorescent) can then be discriminated and quantified. In recent years, the data on attached and ingested particles per cell have been interpreted in a new way (9, 11, 12, 14, 19, 20, 23). The sum of attached and ingested particles have been defined as the accumulated attachment (AA). This parameter describes the attachment process, as all ingested particles must first have been attached to the cell surface. The number of ingested particles per AM divided by AA have been defined as the ingested fraction (IF). As the number of ingested particles is related to the cumulative number of attached particles, IF should reflect the ingestion process. The first aim of the present study was to deduce from the experimental parameters the time of ingestion of particles and to use this calculated parameter and the experimental data to ascertain whether there is any relationship between attachment and ingestion processes.

The second aim of this study was to compare how the experimental and calculated parameters vary among particles for which uptake occurs via different surface receptors. We used particles of amorphous silica to represent particles of nonbiological material. Without opsonization, amorphous silica particles probably are attached to the AM surface by scavenger receptors (16). We also used heat-killed Candida albicans, which binds to the mannose receptor (15). Because of its capsule, Cryptococcus neoformans is hardly at all phagocytized by AM during the first hours (3, 12). However, opsonization with specific IgG results in a considerable phagocytosis of the yeast by AM, and the uptake takes place via attachment to the Fc receptor (9). We therefore studied the uptake of heat-killed C. neoformans opsonized with specific IgG.

IFN- has a fundamental role in activating white blood cells, including macrophages (2, 5). It was therefore surprising that long-term (22-44 h) preincubation of AM with IFN- even at a low concentration (12.5 U/ml) markedly impaired subsequent phagocytosis of silica particles (17, 18). However, most studies of IFN- effects on phagocytic cells have dealt with microbial particles, e.g., fungi and bacteria. A third aim of the study, therefore, was to investigate whether phagocytosis of the yeasts that bind to different AM receptors than silica particles would also be impaired by long-term treatment with IFN-.

	MATERIAL AND METHODS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

For a schematic description of the different experiments, see Table 1.

Phagocytosis of test particles was measured in control AM and IFN--treated AM obtained from the same rat. Test particles were silica particles, heat-killed C. albicans, or heat-killed C. neoformans opsonized with specific IgG. The test of phagocytosis started after the AM had been preincubated with or without IFN- for 22 h. For AM preincubated with IFN-, which was also present during the test of phagocytosis, 12.5 U IFN-/ml was used in all experiments. In experiments 3 and 6 (see Table 1 for descriptions), 50 U/ml was also used.

Silica particles and yeast strains. Particles of inert silica with a diameter of 3.2 ± 0.4 µm (mean ± SD), coated with aminopropyl groups (Spherisorb S 3 NH; Phase Separations, Queensferry, Clwyd, UK) were used as test particles.

Preparation of FITC-labeled silica particles and yeast. Heat-killed C. neoformans or C. albicans suspended in 10 mM sodium phosphate buffer (pH 7.6) were pelleted by centrifugation for 10 min at room temperature and resuspended with Ringer acetate solution (pH 6; Kabi Pharmacia AB, Sweden). Fluorescein isothiocyanate (FITC; Sigma Chemical, St. Louis, MO), 10 mg diluted in 20 ml of sodium carbonate buffer (pH 10.2), was added to 1 ml of the yeast suspensions or 5 mg of silica particles. Final suspensions were incubated at 37°C for 1 h. Suspensions were then washed by centrifugation three to four times in Ringer acetate solution. Yeast cells and particles were counted in a Bürker hemocytometer and diluted in Ringer acetate solution to a concentration of 10⁸/ml. Finally, suspensions were divided into 0.5-ml aliquots and stored at 20°C.

Anti-C. neoformans polyclonal IgG production. Male New Zealand White rabbits were immunized intravenously with 10⁸/ml heat-killed C. neoformans in sterile 0.9% saline. Animals received three injections weekly for 12 wk. Animals were bled, and serum was collected 7 days after conclusion of the immunization schedule (9).

Opsonization. Anti-C. neoformans polyclonal IgG, subagglutinin concentrations of a 1:64 dilution, was used as the source of opsonin. Opsonization of the yeast was carried out for 1 h at 37°C.

Animals and lung lavage. AM were obtained by lavage from healthy male Sprague-Dawley rats (Charles River, Uppsala, Sweden) that weighed 250-300 g. The study was approved by the local animal ethics committee. Rats were killed by an overdose of pentobarbital sodium. Lungs were excised and lavaged with Hanks' balanced salt solution (without Ca²⁺ and Mg²⁺, pH 7.4, 37°C) by using brief massage. Approximately 30 ml of lavage fluid, containing 10-15 × 10⁶ cells, were obtained. More than 90% of the cells was estimated to be AM, as determined from typical AM morphology by light microscopy. Cells were washed once by a 10-min, 300-g centrifugation at room temperature; the resulting cell pellet was resuspended in HEPES-buffered medium 199, pH 7.4 (GIBCO, Paisley, Scotland), with 10% inactivated calf serum, 100 U/ml penicillin, and 100 µg/ml streptomycin (complete medium). Cells were counted in a Bürker hemocytometer.

Phagocytic assay. Phagocytic activity was studied by using a modification of the method described by Hed (13). One million AM in 2 ml of complete medium per culture dish were incubated with 12.5 or 50 U IFN-/ml (IFN--treated AM) or without IFN- (control AM). Culture dishes were placed for 1 h in an incubator at 37°C with 5% CO₂ in air and 80% relative humidity to allow cells to attach to the glass. Recombinant rat IFN- was used (Nova Kemi, Stockholm, Sweden). After an additional 20 h in the incubator, the medium was again exchanged, and 1 ml of complete medium with 10 × 10⁶ FITC-labeled silica or yeast particles was added to each culture dish. Whenever the culture medium was exchanged, new IFN- was added. After the end of the phagocytosis period, ice-cold Ringer acetate solution was added to the dishes to interrupt phagocytosis. Both at 1 and 20 h, the density of AM cell layers was carefully inspected and appeared to be fairly similar, essentially a monolayer. In an earlier study (11), the loss of AM during a 24-h period was measured to be <10%. Extracellular free-test particles were rinsed off, and the AM were stained with trypan blue (2 mg/ml) for 30 s. As trypan blue does not enter viable cells, ingested yeast particles were distinguished by their yellow fluorescence and the surface-attached ones by their staining with trypan blue (Fig. 1). The numbers of ingested and attached yeast particles per macrophage were determined with a Zeiss fluorescence microscope. In each sample, 100 consecutive AM were scored.

View larger version (226K): [in this window] [in a new window]   Fig. 1.   Silica particles attached to and ingested by alveolar macrophages (AM) after 15 min. Fluorescence from the attached particles is extinct by trypan blue, and these particles are stained blue by the dye. Ingested particles are still fluorescent (yellow). Photo by Anders Magnusson.

Mathematical model of phagocytosis. A mathematical model for the attachment and ingestion processes was developed. In the model, it is assumed that the rate of attachment, defined as dAA/dt, is proportional to the number of free particles per AM

(1)

where N is the number of particles available for phagocytosis per AM at the beginning of the experiment (in our experiments, N = 10), k is a rate constant specific to the different types of particles used, and t is time. The solution to this differential equation (Eq. 1) is

(2)

From Eq. 2, k can be solved as

(3)

For a certain type of particle, IT is defined as the time it takes from the moment of attachment of a particle to the cell surface until the time it is ingested. For a certain duration of the phagocytosis period of these particles, it is assumed that IT is the same for all particles. Under this assumption, all particles that become attached between time 0 and t  IT will have been ingested at time t, but none of the particles that become attached after time t  IT. IP at time t is thus, according to Eq. 2

(4)

According to Eqs. 2 and 4, the measured IF is related to k, AA, and IT as follows

(5)

From Eqs. 3 and 5, IT can be explicitly expressed as a function of k and IF

(6)

IT calculated in this way is an estimate of the average ingestion time for a certain type of particle and for a certain duration of phagocytosis.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Studies with silica particles. Variances in AM samples within and between rats were compared for the experimentally determined parameters AA and IF as well as for the calculated parameter IT (experiment 1). For each of six animals, phagocytosis of silica particles by AM during 30 min with and without IFN- was studied in triplicate. Table 2 shows large variances for AA that are mainly due to differences among rats (P < 0.01, ANOVA). In contrast, variances for IF and IT, in relation to the mean value, are relatively small both within and among rats. Means of AA, IF, and IT all differed significantly between control AM and AM treated with IFN- (P < 0.001, ANOVA).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Calculated accumulated attachment (AA) for silica particles as functions of time from measured AA at 15, 30, and 45 min, respectively. Calculations were made both for control () and IFN--treated AM (). Dotted lines, curves calculated from the 15-min observations; dashed lines, curves calculated from the 30-min observations; solid lines, curves calculated from the 45-min observations.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Relationships between the experimentally determined ingested fraction (IF), the calculated ingestion time (IT), and the experimentally determined values of the AA for silica particles. A: relationship between IF and AA for the 30-min observations. B: relationship between IF and AA for the 45-min observations. The calculated relationship between IF and AA (from Eqs. 2 and 3) using the relationship between AA and IT in C is also shown. C: relationship between IT and AA for the 15 ()-, 30 ()-, and 45-min () observations. D: relationship between AA and the difference between the experimentally determined AA in controls and IFN--treated AM.

Studies with C. albicans. Phagocytosis of C. albicans was studied during minutes 15, 30, and 45 by control and INF--treated AM from 4 rats (experiment 5). Figure 4 shows the measured AA at the various observation times for the control and INF--treated AM together with the corresponding curves calculated from each of the measured AA values by using Eq. 2. For both the control AM and for the INF--treated AM, the three calculated AA curves agreed well (Fig. 4). Table 4 shows the AA, IF, and IT at the three time points. At all time points, AA differed significantly between controls and IFN--treated AM (P < 0.01, two-tailed, paired t-test). IF and IT differed significantly between controls and IFN--treated AM only at the 45-min observation (P < 0.05).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Calculated AA for Candida albicans as a function of time from measured AA at minutes 15, 30, and 45, respectively. Calculations were made both for control () and IFN--treated AM (). Dotted lines, curves calculated from the 15-min observations; dashed lines, curves calculated from the 30-min observations; solid lines, curves calculated from the 45-min observations.

View larger version (8K): [in this window] [in a new window]   Fig. 5.   Relationship (linear regression) between the experimentally determined IF, the experimentally determined values of AA for C. albicans, and the 30-min observations.

Studies with C. neoformans. Phagocytosis of C. neoformans opsonized with specific IgG antibodies during 45 min was measured in six rats by using AM not treated and treated with IFN- (experiment 8; Table 5). Phagocytosis of C. neoformans was compared with that of silica particles. There was a significant decrease in AA by the IFN- treatment for both C. neoformans and silica particles (P < 0.01, two-tailed paired t-test). IF was significantly higher for C. neoformans than for silica particles both for control AM and AM treated with IFN- (P < 0.01, two-tailed paired t-test).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

For silica particles, there was a large variance (mean square) in AA among samples from different rats but not within samples from the same rat. The large variance among animals is thus mainly caused by differences in the attachment process among animals and not by uncertainties in the method. Although AA values (means of observations of 100 AM) did not vary much in samples from the same rat, there was a considerable variation among individual AM from the same rat (Fig. 1). Variances of IF and IT were small among both rats and samples from the same rat.

Both for controls and IFN--treated AM, which had phagocytized silica particles or C. albicans during three different time periods, experimentally determined AA corresponded well with the single theoretical rate constant. For each of these particles and for control AM, both experimental IF and calculated IT are independent of AA at small and moderate values of AA (Figs. 3, A-C, and 5). For large AA values, i.e., when AA approaches the total number of particles/AM (10 particles/AM in our study), both IF and IT increase slightly with AM. According to Eq. 5, IF should increase rapidly when AA approaches 10 particles/AM (Fig. 3B). However, for silica particles, such an increase was not seen for the experimentally determined IF. A probable explanation for this is that the number of particles per AM during phagocytosis was >10 because of loss of AM during the 20-h preincubation period.

IT values calculated from the experiments with control AM at various observation times and with different particles are presented in Table 6. The values agree fairly well between experiments in which the same type of particles and same observation time was used. One uncertainty in the calculated IT is, as mentioned above, that the number of particles per AM during the time of phagocytosis probably is >10. This means that the real IT should be smaller than the calculated one. The lower limit of IT (IT_lim) is reached when AA increases linearly with time, and this limit can be calculated from the following equation

(7)

IT_lim values are also given in Table 6. As expected, the deviation between IT and IT_lim is largest for experiments with large AA values but is, in general, rather small. The similar appearances of the AM densities after 1 and 20 h and our earlier measured AM loss of ~10% during 24 h (11) indicate that the uncertainty in the average IT values due to AM losses is small also for experiments with large AA values, as in experiment 2. With a loss of 20% of the AM, IT is calculated, from Eqs. 3 and 6, to be 25.0 ± 4.7 min compared with 26.4 ± 5.0 min in Table 6. However, overestimation of IT due to losses of AM increases with increasing AA and might well explain the positive correlation between IT and AA in the 45-min experiments with silica particles (Fig. 3C). This is illustrated by the fact that if a loss of 20% of the AM is assumed in all 45-min samples in Fig. 3C, there will no longer be any significant correlation between IT and AA (R² = 0.21).

IT increased with the time of observation (between minutes 15 and 30 and between minutes 30 and 45 for the silica particles and between minutes 15 and 30 for C. albicans). A probable reason for this is that IT is an estimate of the average ingestion time (IT) for a certain particle type up to a certain time of observation and that there may be a large unknown distribution of IT. According to Eq. 7, for t > IT and IF = 0 for t  IT, the calculated IT becomes smaller than the average real IT if the observation time does not include the whole distribution of IT. This can be illustrated by dividing an arbitrary IT distribution in two fractions, A and B, with respective magnitudes of a and b (a + b = 1). Average IT for these fractions are assumed to be IT_A and IT_B, and all IT values in the distribution are assumed to be <t. From Eq. 7 follows

(8)

From this equation, average IT for the whole distribution is expressed as

(9)

If it is assumed that all IT values in fraction B are t and all IT values in fraction A are <t, the following IT is calculated from Eq. 7 for t > IT and IF = 0 for t  IT

(10)

As IT_B is > t, this value is less than the real average.

This explanation is compatible with the results that silica particles, for which IT is long, showed an increase in IT in the whole range from 15 to 45 min, whereas C. albicans, for which IT is shorter, showed an increase in IT only between 15 and 30 min.

For control AM, there was no correlation between AA and IF or IT (except for large AA values) for either silica particles or C. albicans despite large variations in AA. This result is in agreement with our laboratory's earlier study with C. neoformans opsonized with specific IgG antibodies (9). In that study, AA varied in a dose-related way with IgG concentration. The highest IgG concentration, 64 times the lowest one, gave 4-5 times larger AA values. However, IF did not vary with IgG concentration. This lack of correlation in the present and earlier study between AA and IF or IT strongly indicates that, for a given particle type, attachment and ingestion processes are independent.

When comparing phagocytosis of test particles by AM, it is important to consider receptors on the AM cell membrane for these particles and the number of receptor ligands on the particles. Our earlier study with C. neoformans has shown very little attachment and ingestion of the yeast incubated up to 3 h with AM with complete medium, including heat-inactivated serum (12). Therefore, it is reasonable to assume that C. neoformans opsonized with specific IgG binds to the Fc receptor. Silica particles are likely to bind to one or several scavenger receptors, a class of receptors that is present on the surface of macrophages that bind chemically altered or oxidized proteins and lipoproteins (22). Some members of this receptor class can probably bind particles of nonbiological material such as titanium dioxide, iron oxide, quartz, and latex particles (16). It is reasonable to assume that C. albicans mainly binds to the mannose receptor (15), but it may also bind to scavenger receptors. This yeast should have a different binding pattern to receptors from that of silica particles because silica particles are unlikely to bind to the mannose receptor. It is thus reasonable to assume that the three types of particles bind to different AM receptors. Even if surface densities of receptor ligands on the three types of particles are not known, average AA values were fairly similar for the three types of particles; average values were 3.5 silica particles/AM after 30 min for all 20 rats studied compared with 3.8 C. albicans/AM for all 10 rats studied. Average AA values after 45 min for the same six rats were 3.2 silica particles/AM and 3.5 C. neoformans/AM. Moreover, IF and IT were rather independent of AA for silica particles and C. albicans, and earlier experiments with C. neoformans opsonized with IgG antibodies (9) demonstrated that the IgG concentration used in opsonization markedly affected AA, but not IF. It is, therefore, possible to rank the duration of the ingestion mediated by the three groups of receptors. Particles bound to the mannose receptor were ingested in the shortest time (Table 6), and particles bound to the Fc receptor were ingested nearly as rapidly, whereas particles bound to scavenger receptors were ingested markedly slower.

Two different mechanisms have been proposed to explain the ingestion process of particles by phagocytic cells, the "triggering" and the "zipper" mechanisms (7). For the triggering mechanism, it is assumed that attachment of a particle to a receptor on the cell membrane induces a signal that starts the ingestion process independent of other membrane receptors. The zipper mechanism instead involves step-by-step interaction between receptors of a certain type on the surface membrane and ligands on the particles, which results in a zipper-like inclusion of the particles in the cell membrane. Strong evidence for the zipper mechanism has been presented by Griffin and colleagues (7, 8). The authors inhibited the ingestion of erythrocytes or lymphocytes by mouse peritoneal macrophages in two ways. One way was to block the rest of the macrophage surface receptors once the respective test cell had been attached to the macrophages. The other way was to produce lymphocytes (used as particles) with ligands located only at a limited region of the lymphocyte surface. It is obvious that our recent results are well compatible with the trigger mechanism. The results may also be compatible with the zipper mechanism, but with this mechanism it would be expected that the attachment and ingestion processes for a certain type of particle would be positively correlated.

Long-term treatment of AM with IFN- markedly impaired the attachment process for all three types of particles. The result indicates that basic mechanisms for attachment to all types of receptors are affected. In experiments with silica particles, the effect of IFN- on AA correlated strongly with the AA level in the controls. Apparently, the most active AM, in regard to attachment of silica particles, were most affected by IFN- treatment. Such a correlation was not seen for C. albicans. This might be explained by the relatively small variance in AA in these experiments. In regard to the ingestion process, it seems to be a difference between the three groups of receptors. For the silica particles, IFN- clearly affected IF and IT, which indicates that the ingestion process via scavenger receptors is impaired. For C. albicans and C. neoformans opsonized with specific IgG, there might be a small impairment in the ingestion process by IFN- treatment.

In summary, both for controls and IFN--treated AM, which had phagocytized silica particles, or C. albicans, the experimental and calculated AA values agreed fairy well. AA varied largely among rats, but the variance was small within the rats. Variances in IF and IT were small, both among and within the rats.

For each type of particle separately, the ingestion process was, as estimated by IF and IT, independent of the attachment process.

One uncertainty in the calculated IT was that the number of particles per AM during phagocytosis was probably >10. This means that the real IT should be smaller than the calculated IT, but this uncertainty was estimated to be small.

IT increased with the time of observation both for the silica particles and for C. albicans. One reason for this is that IT is an estimate of the average IT for a certain particle and a certain observation time. At observation times shorter than the IT for some fraction of the particles, the calculated IT becomes shorter than the average real IT. Another reason (minor) for longer IT at long observation times is that the calculated IT is overestimated at large AA values because the number of particles per AM is >10.

It seems to be possible to rank the IT for three groups of receptors, scavenger receptors, the mannose receptor, and the Fc receptor, in the following way: scavenger receptors > Fe receptors  mannose receptors (Table 6).

The long-term treatment with IFN- markedly impaired the attachment process for all three types of particles, indicating that attachment via all three types of receptors is impaired. The most active AM concerned with attachment were most affected by IFN-. IFN- treatment clearly impaired the ingestion process only for particles bound to scavenger receptors, whereas the ingestion of particles bound to the mannose and Fc receptors were only marginally or not at all affected.

The two parameters IF (experimental) and IT (calculated) gave similar results in regard to the ranking of the ingestion process both for different types of particles and for the effect by IFN-. IT has, however, a clear advantage over IF in that it gives an estimate of an actual average IT.