Progressive loss of the macrophage respiratory burst in oxygen toxicity

Jounud Printed

of Free Radicals

in the USA.

iu Biology

All rights

& Medicine.

Vol.

2. pp. 129-134.

0748-5514/86 0 1986 Pergamon

1986

reserved.

PROGRESS IVE LOSS OF THE MACROPHAGE OXYGEN TOXICITY

RESPIRATORY

GERALD HARRISON and HENRY JAY

S3.00+ Journals

.OO Ltd.

BI JRST IN

FORMAN

Research Division, Department of Medicine, Graduate Hospital, Philadelphia, PA 19146 Department of Physiology, University of Pennsylvania. Philadelphia, PA 19104 and Developmental Lung Biology Research Center, Laboratory of Biochemical Toxicology, Division of Neonatology & Pediatric Pulmonology, Department of Pediatrics and Department of Pathology and Institute for Toxicology, University of Southern California, Childrens Hospital of Los Angeles, Los Angeles, CA 90027 (Recciwd

6 Augrtsr

1986)

Abstract-The respiratory burst of rat alveolar macrophages stimulated by a variety of agents declines as a function of time of exposure to hyperoxia. Previous studies have evaluated this effect in terms of the stimulated 02: production of a population of cells. The present study was designed to determine whether this decline is due to a “turning off” of the respiratory burst activity of some cells within the alveolar macrophage population or a general suppression of the activity of all cells. The phorbol myristate acetate (PMA) initiated respiratory burst of individual rat alveolar macrophages was monitored using the reaction of nitroblue tetrazolium (NBT), which results in the formation of a precipitate on active cells. The formazan staining was evaluated using black and white photographs of the cells and comparison to a scale constructed from photographed cells of four differing intensities of staining. Frequency distributions indicated that when the respiratory burst capability in the population of alveolar macrophages is impaired approximately 50% by oxygen exposure and/or culture in plastic vessels with artificial media, there is a gradual shift in NBT reduction rather than an “all or nothing” mechanism, in which the distribution would have reflected a shift from darkly stained cells to the very lessened or negligible staining observed at the end stage of oxygen toxicity.

Keywords-Macrophage,

Respiratory burst, Superoxide, Tetrazolium, Oxygen toxicity the effect of hyperoxia on the alveolar macrophage respiratory burst both in vivo and in vitro.

INTRODUCTION

In the past several years there have been several reports demonstrating alteration of alveolar macrophage functions by hyperoxic exposure including phagocytosis, chemotaxis, and respiratory burst activity.‘-* The alveolar macrophage functions in the lung as both the first line of defense against infection by foreign organisms and the producer of a variety of agents that affect other cells, such as neutrophil chemotactic agents, interleukin-1, and fibroblast growth factor. Thus, while this cell type is not a structural component of the lung, alteration of these functions may participate in the pathology of hyperoxic lung damage. In studies of the loss of the ability to stimulate 02; production, the effect was evaluated in terms of the 02; production by a population of cells. The present study was designed to determine whether this decline is due to a suppression of the respiratory burst activity of a subpopulation of alveolar macrophages or a general suppression of the activity of all cells. We evaluated

MATERIALS

AND

METHODS

Reagents

Cytochrome c (type VI), nitroblue tetrazolium, phorbol myristate acetate (PMA) and superoxide dismutase were obtained from Sigma Chemical Co. (St. Louis, MO). Nembutal (sodium pentobarbital) and sterile saline for injection were obtained from Abbott Laboratories (Nbrth Chicago, iL). Baralyme@ was obtained from Allied Healthcare Products, Inc. (St. Louis, MO). RPM1 1640 culture medium was obtained from GIBCO Laboratories (Grand Island, NY). Exposure

of rats to hyperoxia

Male Sprague-Dawley specific pathogen free rats (Crl: CD(SD)BR, Charles River Breeding Laboratories, Wilmington, MA) weighing 250 to 300 g were exposed to a 100% O2 normobaric environment for up to 60 h. The O2chamber consistedof a 270 1Plexiglas@ box containing 6 individual cages. Baralyme@ to ab-

Correspondence to: Dr. Henry Jay Forman, Division of Neonatology & Pediatric Pulmonblogy. Childrens Hospital of Los Angeles, Box 54700, Los Angeles, CA 90054-0700.

129

130

G. HARRISON and H. J. FORMAN

sorb CO2 and boric acid to absorb NH3 were placed in the bottom of the chamber. Humidity was unregulated; however, the flow rate of 5 I O?/min prevented precipitation within the chamber. This flow rate also maintained a constant 100% O2 atmosphere, .which was monitored with a flow-through sensor (Teledyne Analytical Instruments, City of Industry, CA). Rats were given free access to water without food; contrgl groups were maintained in room air and also denied food. Ceil prepnration

Alveolar macrophageswere isolated from rats exposedto either air or I atmosphereabsolute OZ. Cells were isolated from 200 to 300 g male rats (Crl: CD(SD)BR, Charles River Breeding Laboratories, Wilmington, MA) by’ lavage with phosphate-buffered saline (PBS) (140 mM NaCI, 5 mM KCI, 8.6 mM Na2HOPq,pH 7.4) in a modification of the method of Myrvik, et al. 9.‘oCells were suspendedin Buffer A at a concentration of I x IO6 cells/ml. Buffer A contained the ingredients of PBS plus I .OmM MgSO.,, I .3 mM CaC12,5 mM glucose, and IO mM HEPES, pH 7.4. Cell suspensionswere counted using a hemocytometer (American Optical Corp., Buffalo, NY) and dye exclusion measuredwith erythrosin B. Typically, ~95% of the cells excluded dye and 98% of these “viable” cells were alveolar macrophages. Macropkage

culture

In vitro cell populations were prepared by obtaining cells from unexposedrats and culturing them in RPMI1640 with medium equilibrated and maintained under either air or 100% O2 at I atmospherefor 24 h. Cells were removed from the surface using a gentle force of Buffer A delivered through a Pasteur pipet. The cells were centrifuged in an Eppendorf microfuge for 30 s and resuspendedin Buffer A. Assay of stimulated

02: production

Alveolar macrophageswere incubated in Buffer A in the presence of 2 X lo-’ M ferricytochrome c for I5 min at 37”. Superoxide production was then stimulated by the addition of IO nmollml PMA. OZ; production was measured as the superoxide dismutaseinhibitable reduction of ferricytochrome c followed at 550-540 nm in a dual wavelength spectrophotometer. The small baseline rate of non-O?’ dependent cytochrome c reduction was subtracted. All rates were expressed as the amount of 02: produced in 4 min following the lag phase.”

Respiratory

burst measurement

by formazart

staining

Approximately I x IO’ cells were suspendedin I ml of Buffer A containing NBT (0.5 mg/ml) as above in a I .5 ml Eppendorf microcentrifuge tube. PMA was added to a final concentration of IO nmol/ml and incubated I5 min at 37°C then allowed to stand at room temperature for an additional I5 min. The reaction was stopped by 5 min centrifugation and decanting of the supernatant. The cell pellet was washedwith phosphate buffered saline (PBS) (Buffer A minus CaCl?, MgSOJ and glucose) recentrifuged for 5 s, supernate decanted and pellet resuspendedin 0.5 ml PBS. Photographs were taken of freshly prepared slides of these suspensions(I 2 to I5 p,I of cell suspension under coverslip). All photographs of cell suspensions were taken under identical conditions of magnification (200 X), illumination and exposure time with an Olympus BH-2 light microscope adapted to use Polaroid Type 55 Pos/Neg, black and white film ASA 50. This film type generatesa negative from which prints were made using identical paper type, exposure and development times. The original photographs were not suitable due to a wipe-on coating which caused additional variations in background tint not encountered when using the negatives and Kodak Polycontrast paper. Evaluation of the degree of formazan staining was made by comparison to a scale constructed from photographed cells. Cells showing no noticeable formazan precipitate were graded “I”; those with slight staining were graded “2”; with some additional darker areas, “3”; and with most of the area being darkly stained, “4.” The scale was constructed by one of the investigators (H.J .F.) and evaluation was made by an observer, who was unaware of the technique or purpose of the investigation. The only instruction given to the observer was to compare the individual cells to the scale and mark each as it was recorded. Since neutrophils appeared in the later stages of the in vivo exposures, the contribution of thesecells were eliminated by the investigator. A count of the cells in each category was also made by the investigator and was in very close agreement with that by the observer; however, to avoid any prejudice, only the observer’s counts were used.

RESULTSAND DISCUSSION Rats were exposed to I atmosphere absolute O2 for either 36 h or 60 h and alveolar macrophages isolated. Alveolar macrophageswere also exposed in vitro to I atmosphereabsolute O2for 24 h. Theseexposure protocols produce reproducible inhibition of the

Lung

macrophage

respiratory burst stimulated by either PMA or other agents.“-” Although there are variations of approximately 25% in PMA-stimulated 02; production between control groups of rats, the standard error within a control group is less than 10%. The mean O1; production observed after 36 h of exposure was 35% and was significantly different from controls by Student’s t test with P < 0.05.” In the present studies, our goal was to compare the change in the whole population with the pattern of formazan staining that could be observed on individual cells. Because this procedure can only be done once to an individual cell, we had to design our experiment as a comparison between the distribution pattern of staining in populations of different respiratory burst capacity. For the effect of in vivo exposure, the cells from two rats were pooled for each exposure period or for controls. For in vitro exposure, the cells from four rats were pooled and then divided between hyperoxic and normoxic incubation. The PMA stimulated respiratory burst of each preparation was measured using the ferricytochrome c assay and each was stained with NBT as described in Methods and Materials. Although NBT reduction may not be a stoichiometric measurement of 02: released by the plasma membrane NADPH oxidase in the respiratory burst,ls production of the formazan stain has been shown to correlate directly with this activity. 16Thus, while quantitation of O?; production by formazan staining has not been attempted here, the percentage loss of activity in the ferricytochrome c assay has been used as an index for evaluation of the contribution of individual cells to the total respiratory burst. Figure I shows the formazan staining pattern of a portion of the cells. Cells incubated with either NBT or PMA, rather than a combination of these reagents, were unstained (all cells wereclassified as grade “I”). Figure 2 shows histograms for the distribution of staining from counts from several photographs of each preparation. 100 cells from each preparation were counted except for the 60 h 02-exposed cells (78 alveolar macrophages). The histograms indicate that there was a loss of formazan staining of alveolar macrophages from rats exposed to 100% O2 either in vivo or in vitro. In addition, the results also indicate that in vitro culture itself, produced a decrease in the amount of formazan staining observed with PMA stimulation. Using the ferricytochrome c reduction assay, it has not been possible to determine whether the gradual loss of respiratory burst activity occurred because some cells became desensitized to PMA stimulation while others remained active or because all cells were inhibited gradually in their ability to produce a respiratory

function

131

in 0:

burst. Formazan staining of individual cells, however, did make it possible to decide which of these two proposed mechanisms was a better fit. To aide in determination of the possible pathways for loss of the respiratory burst due to hyperoxia and/ or in vitro exposure, we made two theoretical calculations. We assumed that the 60 h exposed cells represented 100% inhibition since the respiratory burst measured in this population by the ferricytochrome c reduction method was nil. Table I shows the percentage inhibition from the ferricytochrome c assay, the distribution of staining for each preparation, and two theoretical distributions. Theory I assumes that for a given percentage inhibition (X) of the exposed cells, the distribution in each category will change by X percent between the control and the 60 h exposure (100% inhibition.) The change in each category would thus be linear, as would be expected for an “on-off” mechanism. Theory 2 assumes that a gradual inhibition occurs. Because staining of the cells is evaluated by assigning cells into ranges rather than by absolute quantitation, theory 2 can only approximate a gradual change. In designing this model, the capacity for staining of each cell is assumed to shift X percent toward a lower category rather than changing from an “on” to an “off” state. The mathematical relationship that describes this gradual loss of activity is a series expansion where the distribution in each category can be calculated as: Yl, = Yl j + X( Yzi - Yzo + S( Yji - YJ, +

x(y4i

- y.lJ))

Yz, = Y?j + X(Yj( - YJ” + X( Y4i - YJ,)) - X( Yli -

Yl, + -tj Yji - Y30 + X( Ydj - Yb)))

Y,,, = Yj, + x( YJi - Y,) - -et y.3; - Y,,) +

dY4,

-

Y,))

Y,,. = Yq; - x( YJi - Y,) where x = X/ 100, Y,, is the number of cells expected in a category I, YZi is the number of cells found for category 2 in the control; and Y,, is the number of cells found for category 3 in the 60 h exposed cell population. The experimental and predicted distributions were compared by a chi-square goodness-of-fit test. I7 A chisquare value of less than 7.8 1 indicated that the theory fit the data with a = 0.05. The chi-square values shown in Table I indicate that Theory I did not fit in any comparison while Theory 2 fit in three of the four comparisons. Although Theory 2 did not significantly

132

G. HARRISON

and

H. J. FORMAN

(B)

1

2

3

4

CD)

03 Fig. I. Formazan staining of PMA-stimulated rat alveolar macrophagcs. untreated rats; (B) from 36 h O,-exposed rats; (C) from 60 h O:-exposed culture in 100% O>. The small darkly stained cells in C are neutrophils.

Representative rats; (D) after

photographs 24 h culture

are shown in air; and

for: (A) (E) after

from 24 h

Lung Table

1. Observed

and Theoretical

Distributions

macrophage

function

for Formazan Staining In Vitro Culture

133

in Oz of Alveolar

Distribution Alveolar macrophage preparation

unexposed rats 60 h Oz in vivo 36 h O2 in vivo C (Theory 1 from A B: 40.2% inhibition) C (Theory 2 from A B: 40.2% inhibition) (D) 24 h air cultured (E) 24 h O2 cultured E (Theory 1 from D B: 32.5% inhibition) E (Theory 2 from D B: 32.5% inhibition) E (Theory 1 from A B: 73.2% inhibition) E (Theory 2 from A B: 73.2% inhibition) D (Theory I from A B: 60.2% inhibition) D (Theory 2 from A B: 60.2% inhibition)

Respiratory (nmol 01:14

burst min)

Macrophages

(%

Chi-square vs. observed

Predicted II

III

IV

and

32

10

18

40

100.7

and

3

22

35

40

28.3

and

36

19

28

17

15.0

and

21

29

33

17

1.1

and

58

14

10

18

86.9

and

30

29

2.3

18

4.8

and

48

12

13

27

80. I

and

16

28

29

27

7.3

3.13 0 1.87

1.25 0.84

16 23

III 29 3 53

19 32

40 30

fit when compared with the data for in vivo exposure to 36 h of 02, the expected values from Theory 2 came much closer than did those for Theory I. Considering how unlikely it is that a precise mathematical relationship would fit the nonparametric data obtained by formazan staining, Theory 2 appeared to approximate the pathway of inhibition. Practically, it appears that some pattern of gradual inhibition would best describe the observed data.

t;

IV 67 0 43

25 15

In contrast, Theory 1 obviously did not come close to predicting the pattern of inhibition by either hyperoxia and/or culture conditions. Since Theory 1 described a situation in which X% of the alveolar macrophages were “turned off” while Theory 2 described a situation in which all the cells were gradually inhibited, it appears that the latter better describes the situation. It is likely that such a pattern of inhibition represents toxic or pathologic changes in the cells

25

t f L 0

Rir

36hr02

60hr

O2

Air

In uiuo Fig. 2. Histogram

and/or

I

(A) (B) (C)

II 4 18 4

to Hyperoxia

cells)

Observed I 0 79 0

Exposed

indicating

the PMA-stimulated

formazan

24hr

O2

In vitro staining

pattern

of air and t&-exposed

alveolar

macrophages.

G. HARRISON

134

rather than a “down regulation” in which subpopulations of Aveolar macrophages would have been selectively “turned off.” Ack,to,c’/edgme,trs-Supported by NJH grants HL3 I83 I and HL37556. The authors thank June Nelson for technical assistance, Drs. James Glasgow and Victor Damiano for instruction in photography of the cells, Dr. Linda Chan for statistics consultation, and Tina-Karen Alexia Forman for evaluation of cell staining.

REFERENCES I.

2.

3.

4.

5.

6.

7.

L. M. Simon, S. G. Axline, and E. D. Robin. The effect of hyperoxia on phagocytosis and pinocytosis in isolated pulmonary macrophages. Lab. fwes/. 39: 541-546 (1978). N. Suttorp and L. M. Simon. Decreased bactericidal function and impaired respiratory burst in lung macrophages after sustained in vitro hvperoxia. Am. Rev. Rest. Dis. 128: 486-490 (1983). -’ T. A. Raffin. L. M. Simon. D. Braun, J. Theodore. and E. D. Robin. Impairment of phagocytosis by moderate hyperoxia (40% to 60% oxygen) in lung macrophages. Lab. Irwesr. 42: 622-626 ( 1980). A. L. Bowles, J. H. Dauber, and R. P. Daniele. The effect of hyperoxia on migration of alveolar macrophages in vitro. Am. Rev. Resp. Dis. 120: 541-545 (1979). C. Aerts and C. Voisin. In vitro toxicity of oxygen and oxygenparaquat association on alveolar macrophages surviving in gas phase. Bull. Ettrop. Physiopofh. Resp. 17: 145-15 I (suppl.) (1981). C. Voisin, C. Aerts, and A. B. Tonnel. Application a I’etude de la cytotocite de I’oxygene. Colloq. INSERM 84: 169-176 (1979). L. J. Wolff, L. A. Boxer. J. M. Allen, and R. L. Baehner. The selective effect of hyperoxia on the guinea pig alveolar mac-

and

H. J. FORMAN

8.

9.

IO.

II.

12.

13.

14.

15.

16.

17.

rophage membrane. J. Rericuloendorhclial Sot. 24: 377-382 (1978). M. Rister and R. L. Baehner. Effect of hyperoxia on superoxide anion and hydrogen peroxide production of polymorphonuclear leukocytes and alveolar macrophages. Br. J. Haemarol. 36: 241-248 (1977). Q. N. Myrvik, E. S. Leake, and B. Fariss. Studieson pulmonary alveolar macrophages from the normal rabbit: A technique to procure them in a high state of purity. J. Immunol. 86: l28132 (1961). H. J. Forman, J. Nelson, and A. B. Fisher. Rat alveolar macrophages require NADPH for superoxide production in the respiratory burst. J. Viol. Chem. 255: 9879-9883 (1980). H. J. Cohen and M. E. Chovaniec. Superoxide generation by digitonin-stimulated guinea pig granulocytes. A basis for a continuous assay for monitoring superoxide production and for the study of the activation of the generating system. J. C/in. Irwesr. 61: 1081-1087 (1981). H. J. Forman, J. J. Williams. J. Nelson, R. P. Daniele, and A. B. Fisher. Hyperoxia inhibits stimulated superoxide release by mt alveolar macrophages. J. Appl. Physiol. 53: 685-689 (1982). M. W. Sutherland, M. Glass, J. Nelson, Y. Lyen, and H. J. Forman. Oxygen toxicity: Loss of lung macrophage function without metabolite depletion. J. Free Rad. Biol. Med. I: 209214 (1985). H. J. Forman. J. Nelson, and G. Harrison. Hyperoxia alters the effect of Ca” on rat alveolar macrophage superoxide production. J. Appl. Physiol. 60: 13001305 (1986). J. K. Lace, J. S. Tan, and C. Watanakunakorn. An appraisal of the nitroblue tetrazolium reduction test. Am. J. Med. 58: 685-694 (1975). R. L. Baehncr, L. A. Boxer, and L. M. Ingraham. Reduced oxygen by-products and white blood cells. In: Free Radicals in Biology, Volume 5 (W. A. Pryor, ed.) pp. 91-l 12, Academic Press. New York (1982). L. Ott. Au inrrodrrcrion IO Smrisrical Methods, Duxbury Press, Boston, 775 pp. (1984).