Am. J. Respir. Cell Mol. Biol. Vol. 19, pp. 867–874, 1998Internet address: www.atsjournals.org

Neutrophil Apoptosis during the Development and Resolution of Oleic Acid-Induced Acute Lung Injury in the Rat

Naveed Hussain, Fengying Wu, Li Zhu, Roger S. Thrall, and Mitchell J. Kresch

Departments of Pediatrics and Medicine, University of Connecticut School of Medicine, Farmington, Connecticut

The oleic acid (OA) model of acute lung injury in rats is characterized by a massive and rapid influx ofpolymorphonuclear neutrophils (PMN) within 1 h, with a peak inflammatory response at 4 h and resolu-tion by 72 h. We hypothesized that PMN apoptosis is involved in the resolution of OA-induced acute lung

injury. To test this hypothesis, healthy adult Fischer 344 rats were given 30

m

l OA in 0.1% bovine serumalbumin (BSA) intravenously; controls were given BSA alone and killed at 1, 4, 24, and 72 h after OA toobtain bronchoalveolar lavage fluid (BALF) and lung tissue. Cell pellets from BALF and formalin-fixed,paraffin-embedded tissue section samples were processed for terminal deoxyribonucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) to identify apoptotic cells. Propidium iodide was usedto counterstain nuclei. Percentage of nuclei undergoing apoptosis was counted under a fluorescent micro-scope. Control rats showed only resident alveolar macrophages (AM) in the BALF with no apoptosis. Atthe peak of injury, 1 h and 4 h after OA injection, we observed a massive PMN response without any evi-dence of apoptosis. At 24 h, when the OA injury is clinically and histologically in early resolution, we ob-served intense apoptosis of PMN nuclei along with evidence of apoptotic bodies in the cytoplasm of AM.Some of the AM also showed apoptotic nuclei at 72 h. Similar observations were made in the lung tissuesections. The results of the TUNEL assay were confirmed by DNA ladders and electron microscopy. Weconclude that apoptosis of PMN and clearance by AM is an important mechanism in resolution of OA-

induced acute lung injury.

Hussain, N., F. Wu, L. Zhu, R. S. Thrall, and M. J. Kresch. 1998. Neutro-phil apoptosis during the development and resolution of oleic acid-induced acute lung injury in the

rat. Am. J. Respir. Cell Mol. Biol. 19:867–874.

The progression from injury to acute inflammatory re-sponse is characterized by an influx of polymorphonuclearneutrophilic leukocytes (PMN). The resolution of acuteinflammation depends on the timely clearance of these in-flammatory cells. Emigration of PMN back into circula-tion does not occur after acute lung injury (1). The acutePMN response during inflammation in the lungs clears bytwo possible mechanisms: (

1

) cellular lysis and break-down, which results in liberation of proteases and otherenzymes; or (

2

) apoptosis and phagocytosis by macro-phages and other cells.

Apoptosis (2) (a form of programmed cell death) is anactive process involving gene transcription and proteinsynthesis that leads to an orderly sequence of events, re-sulting in cytoplasmic shrinking, budding, nuclear frag-mentation (180- to 200-bp fragments), and formation ofapoptotic bodies that are ingested by the neighboring cells(3, 4). Because the cell membrane remains intact duringthis process, there is no release of proinflammatory stimuliand this has been considered to be important in the resolu-tion of inflammation (5). Grigg and colleagues (6) haveshown that PMN apoptosis occurs in cells obtained fromtracheal aspirates in infants recovering from respiratorydistress syndrome. Cox and coworkers (7) have demon-strated that PMN apoptosis and engulfment by alveolarmacrophages contributes to the resolution of bacterial li-popolysaccharide (LPS)-induced acute lung inflammation.It remains to be determined whether this process is spe-cific to the LPS injury or common to other models of re-solving lung inflammation.

Oleic acid (OA) induces lung injury that mimics theacute inflammation of the adult respiratory distress syn-drome (ARDS) (8), which has been well characterized as amodel for studying acute lung inflammation (8–10), butthe role of apoptosis in this process has not been studied.

(

Received in original form August 4, 1997 and in revised form March 25,1998

)

Address correspondence to:

Naveed Hussain, M.B.B.S.; D.C.H., Depart-ment of Pediatrics, University of Connecticut School of Medicine, Farm-ington, CT 06030-2948. E-mail: hussain@nsol.uchc.edu

Abbreviations:

alveolar macrophage(s), AM; bronchoalveolar lavagefluid, BALF; bovine serum albumin, BSA; deoxyuridine triphosphate,dUTP; lipopolysaccharide(s), LPS; phosphate-buffered saline, PBS; poly-morphonuclear neutrophil, PMN; terminal deoxyribonucleotidyl trans-ferase-mediated dUTP-biotin nick end labeling, TUNEL.

868

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 19 1998

In tissues, apoptosis and phagocytosis proceed rapidly andare completed in a matter of hours (3, 11). Visualization ofsmall percentages of

in situ

stained apoptotic cells yieldsbiologically significant data. A rate of tissue regression asrapid as 25% per day can result from apoptosis in 2% to3% of the cells at any one time (11). The terminal deoxyri-bonucleotidyl transferase-mediated deoxyuridine triphos-phate (dUTP)-biotin nick end labeling (TUNEL) assayhas been shown to be a sensitive and specific method fordetecting apoptosis in histologic sections (12, 13).

Based on the hypothesis that inflammatory cell apopto-sis is necessary for the resolution of acute pulmonary in-flammation, the purpose of this study was to determine therole of PMN apoptosis during the course of oleic acid-in-duced lung injury and resolution in rats.

Materials and Methods

Administration of Oleic Acid

Adult male, pathogen-free, Fisher 344 inbred rats weigh-ing 220 to 260 g (Charles River Laboratories, Worthing-ton, MA) were given 30

m

l of OA (Sigma Chemical Co.,St. Louis, MO) suspended in 270

m

l 0.1% bovine serumalbumin (BSA) (Sigma Chemical Co.) intravenously via afemoral vein cutdown under ketamine (15 mg/kg) (BristolLaboratories, Syracuse, NY) anesthesia to induce acutelung injury, as detailed in our previous reports (8, 14, 15).Control animals were given an equal volume of 0.1% BSAin a similar manner.

Bronchoalveolar Lavage

Animals were killed at 1, 4, 24, and 72 h after administra-tion of OA. Bronchoalveolar lavage (BAL) was per-formed

in situ

by instilling 50 ml (five 10-ml aliquots) ofsterile saline via a cannula ligated in the trachea (16). Ap-proximately 45 ml of fluid was retrieved by BAL in eachanimal. Total leukocyte counts were made using a standardAmerican Optical hemocytometer (Buffalo, NY). BALcells were obtained after centrifugation at 600

3

g

for 6 min,resuspended at a concentration of 1

3

10

6

cells/ml, andfixed in 4% neutral buffered formalin for 10 min at roomtemperature. One hundred fifty microliters of cell suspen-sion was dried on a microscope slide for the TUNEL assay.

Lung Tissue

After lung lavage, the rats’ partially inflated lungs werefixed in 4% neutral buffered formalin for 24 h, paraffinembedded, and processed routinely. Slides were deparaf-finized in xylene and rehydrated in graded ethanol, andthe protein was digested with 20

m

g/ml proteinase K(Sigma Chemical Co.) for 15 min at room temperature forthe TUNEL assay.

TUNEL Assay

The slides were washed with two changes of phosphate-buffered saline (PBS) and the TUNEL assay (17) was per-formed using the Fluorescein Apoptag

in situ

apoptosisdetection kit (Oncor, Gaithersburg, MD) according to themanufacturer’s instructions. Briefly, after equilibrationwith the buffer provided, terminal deoxyribonucleotidyltransferase (TdT) enzyme was applied and incubated with

digoxigenin-labeled dUTP for 1 h at 37

8

C. The reactionwas then stopped using the Stop/Wash buffer providedand antidigoxigenin-fluorescein was applied for 30 min atroom temperature. After being washed with PBS, the slidewas counterstained with propidium iodide/Antifade (On-cor) and used for fluorescent microscopy. Appropriatecontrols with or without TdT enzyme were used to detectfalse positives.

Fluorescent Microscopy

After TUNEL staining, the slides were viewed by epifluo-rescence using standard fluorescein excitation at 494 nmand emission at 523 nm wavelength. An intense green sig-nal in the nucleus indicated TUNEL-positive apoptoticcells. The absence of green nuclear emission indicatedTUNEL-negative nonapoptotic cells (18). Propidium io-dide was viewed with excitation at 596 nm, and a brightred emission was seen in all nuclei present, regardless ofthe condition of the cell. This served as the counterstainand also helped in identification of cell type (PMN or mac-rophage) based on the nuclear size and morphology. Per-cent apoptosis was calculated by number of TUNEL-posi-tive cells

3

100/total number of cells counted with propidiumiodide stain. A minimum of 500 cells was counted fromeach sample of BAL fluid (BALF) using a Zeiss Axiovert100 TV fluorescence microscope (Zeiss, Thornwood, NY)under oil immersion at

3

630 magnification; alternatively,a minimum of 10 fields was analyzed from each sample oflung tissue after image capture using the PhotometricsCCD imaging camera (Tucson, AZ).

Electron Microscopy

Lung tissues and BALF cell pellets from control and OA-injured rats were fixed for 1 to 2 h in 2.5% glutaraldehydein 0.1 M sodium cacodylate buffer, pH 7.4, postfixed in 1%OsO

4

0.8% potassium ferricyanide in cacodylate buffer, andstained in block with 0.5% aqueous uranyl acetate. Afterdehydration in ethanol, the samples were embedded in Po-lybed resin (Polysciences, Warrington, PA). Thin sectionswere cut with a diamond knife, collected on 200 mesh coppergrids, stained with uranyl acetate and lead citrate, and ex-amined in a Philips CM10 transmission electron microscope(Philips, N.V., Hillsboro, OR). Samples were examined forthe presence of characteristic features of apoptosis, such ascondensation of nuclear chromatin, fragmentation of nuclei,condensation of cytoplasm, and/or presence of condensednuclear fragments in the cytoplasm of neighboring cells.

DNA Ladders

Lung tissue homogenate was suspended in lysis buffer andincubated overnight at 37

8

C (19). The DNA was extractedwith an equal volume of 25:24:1 phenol, chloroform, andisoamyl alcohol mixture and then ethanol-precipitated for20 h at

2

80

8

C. The DNA was washed twice (70% EtOH),air-dried, and then resuspended in RNase buffer for 1 h at37

8

C. After RNase activity was quenched (56

8

C for 5 min),10-

m

g samples of DNA (determined spectrophotometri-cally) were electrophoresed on a 1% agarose gel in tris-buffered ethylenediamine tetraacetic acid (50 V, 2.5 h).Electrophoresed DNA was stained with ethidium bromide(1

m

g/ml) and visualized with ultraviolet irradiation.

Hussain, Wu, Zhu,

et al.

: Apoptosis and Lung Inflammation 869

In tissues with low apoptotic rates, a modification of theprevious procedure using end labeling with

32

P was used(20). Briefly, 1

m

g of genomic DNA was end labeled in 30

m

l of reaction buffer with 0.5

m

Ci of

32

P-deoxycytidinetriphosphate (dCTP) at room temperature for 30 min.DNA was precipitated with ethanol and resuspended in100

m

l of TBE. About 10

m

l of DNA from each samplewere loaded on a 1% agarose gel for electrophoresis fol-lowed by autoradiography (X-OMAT-AR; Kodak, Roch-ester, NY). The size of DNA fragments (180 to 200 bp)was confirmed by ladders of 123-bp molecular weightmarkers.

Statistical Analysis

Statistical comparisons of single parameters between twogroups was performed with unpaired Student’s

t

tests (21).

P

,

0.05 was considered significant.

Results

Inflammatory Cell Numbers and Apoptosis in BALF

As reported previously (15, 22), administration of OA in-duced a rapid influx of PMN in BALF (Table 1). The ab-solute numbers of PMN increased sharply within 4 h,peaking around 24 h, and decreasing by 72 h after OA ad-ministration. PMN apoptosis was observed to be maximal(62

6

3%,

P

,

0.0001) at 24 h after OA. AM apoptosiswas seen in a small but significant percentage of cells (1

6

0.1%,

P

,

0.05) about 72 h after OA.

Relative Proportion of Cells in BALF

The cells in BALF were identified as PMNs on the basis oftheir characteristic nucleus and smaller size comparedwith alveolar macrophages. The alveolar macrophageswere identified on the basis of their relative larger size anddistinct nuclear morphology with propidium iodide stainand fluorescent microscopy. The proportions of inflamma-tory cells as a percentage of BALF cells were studied at 1,4, 24, and 72 h after OA administration and the results areshown in Figure 1. In control rats at all times, the BALFcells were predominantly (99%) alveolar macrophages

(AM) with

,

1% PMN (Figures 1A and 1B). The propor-tion of BALF cells that were PMN increased rapidly 1 and4 h after OA administration. At 24 h after OA injury, amajority of PMN stained TUNEL-positive with the typicalmorphology of apoptosis (

P

,

0.05). At the same time, therelative proportion of PMN fell, and therefore there was acorresponding increase in the proportion of AM. By 72 hafter OA injury, the relative proportion of BALF cells hadreverted close to control values with some PMN and occa-sional AM showing apoptotic changes.

TABLE 1

BALF inflammatory cell apoptosis in OA lung injury

Time after OA Injury

PMN Number

†

(

3

10

6

cells

)PMN % Apoptosis

(

mean

6

SEM

)AM % Apoptosis

(

mean

6

SEM

)

Controls 0.03

6

0.01 0

6

0 0

6

04 h 5.6

6

1.2* 1

6

0.2 0

6

024 h 7.6

6

0.5* 62

6

3.0** 0

6

072 h 2.4

6

0.3* 11

6

2.2* 1

6

0.1*

Data are presented as means

6

SEM from four animals. Data from controlanimals at each time point were the same and are therefore represented by asingle set of numbers. Percent apoptosis was calculated by number of TUNEL-positive nuclei

3

100/total number of nuclei stained with propidium iodidefrom a total of 1,000 cells counted using a fluorescent microscope with

3

630magnification. AM and PMN were distinguished on the basis of their morpho-logical features. There was a significant induction of apoptosis at 24 h after OAinjury (**

P

,

0.0001)

compared with controls. By 72 h after injury, PMN apop-tosis was decreasing and occasional AM apoptosis (*

P

,

0.05) was observedcompared with controls. (

†

Data from Shiue and Thrall [15]; *

P

,

0.05.)

Figure 1. Time course of changes in the relative proportion of ap-optotic (TUNEL-positive) and nonapoptotic (TUNEL-negative)inflammatory cells in the BALF obtained after administration of30 ml OA intravenously to rats. (a) Neutrophils as a percentage ofcells in BALF. (b) AM as a percentage of cells in BALF. (c) AMcontaining neutrophil apoptotic bodies identified by TUNEL-positive bodies in their cytoplasm as a percentage of cells inBALF. Data are shown as means 6 SEM, n 5 4–5; **P , 0.0001;*P , 0.05.

870

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 19 1998

Figure 2. Maximum apoptosis is seen 24 h after OA-induced acute lung injury. BALF cells and lung tissue sections were processed asdescribed in MATERIALS AND METHODS using the Fluorescein Apoptag kit (ONCOR, Gaithersburg, MD). The top panel shows resultsfrom the BAL, and the bottom panel shows results from lung tissue. Identical fields were observed for TUNEL-positive green fluores-cence (a, b, c) and propidium iodide (red) counterstain for all nuclei (d, e, f). No TUNEL-positive cells are seen at 4 h after OA (a and bin both panels), and these results were identical to those of the controls and 1-h samples (not shown). Maximum number of TUNEL-positive apoptotic cells were seen at 24 h after OA (b and e in both panels). Decreased TUNEL-positive cells were seen at 72 h after OA(c and f in both panels). Note the presence of apoptotic PMN in the macrophage in the top panel, b. Bar 5 120 m for top panel; 60 m forbottom panel.

Hussain, Wu, Zhu, et al.: Apoptosis and Lung Inflammation 871

TUNEL Assay of BAL and Lung Tissue

To determine whether the observations made on BALFinflammatory cells reflected the tissue response of thelung, we examined histologic sections of the lung at 1, 4,24, and 72 h after OA-induced injury, using the TUNELassay. Representative fields from each set of slides afterfluorescent microscopy are shown in Figure 2.

The top panel in Figure 2 shows results from the BAL,and the bottom panel shows results from lung tissue. Iden-

tical fields from each sample were observed for TUNEL-positive bright green fluorescence (a, b, c) of apoptosisand propidium iodide (red fluorescence) nuclear counter-stain (d, e, f) to assess the percentage of nuclei undergoingapoptosis. No TUNEL-positive cells were seen at 4 h afterOA (a in both panels of Figure 2) and these results wereidentical to those of the controls and 1-h samples (notshown). Maximum number of TUNEL-positive apoptoticcells were seen at 24 h after OA (b in both panels of Figure2). Decreased numbers of TUNEL-positive cells wereseen at 72 h after OA (c in both panels of Figure 2). Lungsections from control rats did not show TUNEL-positivecells (results not shown).

DNA Ladders from Lung Tissue

The fragmentation of DNA due to apoptosis was alsodemonstrated by the typical ladder pattern on agarose gelelectrophoresis of DNA, extracted from lung tissue ofOA-treated rats and untreated controls, as shown in Fig-ure 3. Control rat lungs did not show a DNA ladder pat-tern even with 32P labeling, whereas typical ladder patternswere observed in specimens from OA-treated rat lungs at24 and 72 h.

Lung Electron Microscopy

Because apoptosis has been defined in morphologic terms(12), the confirmation of this process was done by electronmicroscopic examination of the lung in rats treated witholeic acid at various time points. The peak of apoptosis at24 h after OA was confirmed by finding a number of apop-totic PMN in BALF as shown in Figure 4a. The findingof apoptotic bodies ingested by macrophages in BALF,shown in Figure 4b, demonstrates the occurrence of pha-gocytosis that was suggested by observations with TUNELand hematoxylin and eosin (H&E) staining.

Lung Histopathology

Figure 5 shows the histopathologic changes using the H&Estains of BALF and lung tissue. Increased cellularity withacute inflammatory cells (PMN) was seen at 4 h after OA.Cellularity showed a decrease by 24 h, and was revertingback to control levels by 72 h after OA. AM with apop-totic bodies in their cytoplasm (arrows, Figure 5) couldstill be seen under high-power light microscopy (magnifi-cation: 3400) in these lavaged lung specimens.

AM Phagocytosis of Apoptotic PMNs

About 7% to 9% of AM in BALF showed evidence ofTUNEL-positive staining in their cytoplasm by 24 to 72 hafter OA injury (Figure 1c). This staining represents theapoptotic PMNs that had been ingested by AM as shownin Figure 2 (bottom panel, b) and Figure 5 (arrow). Theconfirmation of this process was done by electron micro-graphic demonstration of phagosomes with apoptotic bod-ies in AM (Figure 4b).

Observations of Other Cell Types in Lung and BALF

Capillary endothelium. OA injury is primarily due toendothelial damage, and this was seen in the OA-injuredlungs as cellular swelling with electron microscopy, but no

Figure 3. DNA ladders. DNA was extracted from lung tissue ofcontrol rats and those treated with OA for 4, 24, and 72 h, as de-scribed in MATERIALS AND METHODS. Equal amounts of DNAwere labeled with 32P-dCTP and subjected to gel electrophoresisfollowed by autoradiography. DNA samples shown are lane 1:OA-injured at 72 h, DNA ladders faintly visualized; lane 2: OA-injured at 24 h, maximal intensity of ladders; lane 3: OA-injuredat 4 h; lane 4: Control at 4 h, no ladders are seen; and lane 5: 123-bp molecular weight marker.

872 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 19 1998

endothelial apoptosis was noted. The TUNEL staining didnot identify positive staining in the endothelium.

Alveolar epithelium. Evidence of injury was seen withcellular and interstitial swelling and presence of fibrinousstrands 24 h after OA treatment, but no morphologicchanges of apoptosis were noted with electron microscopy.In the BALF, the two major cell types noted with electronmicroscopy were PMN and AM. The AM were twice theirnormal size after 24 h of injury with OA and demonstratedphagosomes with both lamellar bodies (not shown) andapoptotic bodies (Figure 4b).

DiscussionWe used the OA-induced lung injury model of acute in-flammation that resolves spontaneously to demonstratethat (1) PMN influx is followed by PMN apoptosis be-tween 1 and 24 h, (2) onset of PMN apoptosis correlateswith histologic signs of resolution between 24 and 72 h,and (3) phagocytosis of apoptotic PMNs by macrophageshelps in their clearance.

In this report of apoptosis of lung cells in the OA-induced model of acute lung inflammation, we have suc-

cessfully demonstrated use of the TUNEL assay for evalu-ation of apoptosis in BALF and tissues with acute lung in-jury. Previous studies using purely light microscopicfeatures to quantitate apoptosis (7) suffer from the draw-backs of false positives and decreased specificity. TheTUNEL assay is more specific and sensitive than light mi-croscopy for detection of apoptosis (18). The TUNEL as-say of lung tissue that we have done demonstrates thatPMN apoptosis also occurs in lung parenchyma and is notrestricted to PMNs in the alveolar fluid. The results fromDNA ladders and electron microscopy confirm the find-ings from the TUNEL assay.

Apoptosis as a form of programmed cell death involvedin clearing up cells and tissues has been demonstrated dur-ing embryogenesis (23), during physiologic involution oforgans such as the thymus (24) and ovary (25), and inadaptive response to noxious stimuli (26). We have re-cently demonstrated the ontogeny of apoptosis duringlung development in the rat (27). The role of apoptosis inthe clearance of inflammatory cells has been suggested bySavill (5) and Haslett (28) on the basis of findings in cellculture studies. Its role in resolution of granulation tissuein dermal healing was elegantly demonstrated by Desmou-

Figure 4. Lung electron microscopy. Representative electron microscopic images from the BAL sample of an OA-injured lung at 24 hafter the injury. Panel a (magnification: 318,000) shows a typical apoptotic PMN with the typical condensation of the chromatin in ashrunken vacuolated cell, along with a condensed isolated apoptotic body. Panel b (magnification: 310,500) shows a normal macro-phage with an ingested apoptotic PMN (arrow); a normal PMN is also shown alongside for comparison.

Hussain, Wu, Zhu, et al.: Apoptosis and Lung Inflammation 873

liere and colleagues (29), who showed that myofibroblastnumber in granulation tissue is reduced by apoptosis of su-perfluous cells during wound healing. Apoptotic bodies inhealing muscle tissue suggest the role of apoptosis in clear-ance of inflammatory cells (30).

There have been reports about the role of apoptosis inlung injury caused by bleomycin (31, 32) and hyperoxia(33). Its role in inflammation and repair was suggested byPolunovsky and coworkers (34), who showed that apopto-sis is involved during the resolution of fibrosis, but in thisreport, the cell types involved were not identified. Griggand colleagues (6) showed that neutrophil apoptosis oc-curs in BALF cells during the resolution of neonatal respi-ratory distress syndrome.

Cox and coworkers (7) showed that PMN apoptosis inthe LPS model of acute lung injury occurs between 6 and72 h after injury, and they demonstrated the ingestion ofapoptotic PMNs by macrophages. The question is thenraised whether the PMN apoptosis and macrophage pha-

gocytosis occur as specific responses to the LPS that wasintroduced intratracheally, or whether they are physiologi-cally regulated mechanisms that occur in response to otherforms of injury as well.

The OA model of pulmonary injury is well suited to an-swer this question because the route of administration isnot intratracheal but intravenous. Moreover, it is a well-characterized model of acute lung injury and inflammationthat bears similarities with ARDS, a significant clinicalproblem in humans (8, 9).

Interestingly, our findings of the timing of PMN influxand apoptosis are similar to those of Cox and coworkers(7). We found peak apoptotic response at 24 h after the in-jury. By 72 h, neutrophil numbers and percent apoptosishad decreased significantly. It is therefore reasonable tosuggest that the timing of apoptosis and the resolution re-sponse are not related to a specific injurious agent or routeof delivery but are more general responses during resolu-tion of an acute inflammatory process.

It has been shown that dexamethasone inhibits neutro-phil apoptosis in vitro in a dose-dependent manner (35).Our previous reports have shown that dexamethasone at-tenuates OA injury (15, 22) but delays the resolution of in-jury (15), which may well be due to its inhibition of PMNapoptosis. Conversely, it has also been shown that inter-leukin-10 given intratracheally during LPS-induced lunginjury in rats increases PMN apoptosis and facilitates reso-lution of inflammation (36). These findings suggest that lo-cal or paracrine factors could modulate PMN apoptosisand thus affect resolution of lung inflammation. Identifica-tion of these factors may lead to an understanding of thetriggers that determine whether an acute inflammatory re-sponse would lead to repair or progress to chronic inflam-mation and fibrosis.

Macrophages are known to function as scavengers ofcellular material. Scavenging of debris from lytic PMN hasbeen demonstrated in AM but has been shown to causeAM activation and release of proinflammatory cytokines(37). Phagocytosis of apoptotic PMN by AM does not acti-vate the macrophages and is probably the optimal methodof clearance of PMN (38). This process has been demon-strated to occur in vitro (39) and also in vivo in acute ar-thritis (40). We have shown in this study that apoptoticbodies are found in AM during the resolution of OA-induced inflammation, and this may be a common mecha-nism by which the superfluous and effete inflammatorycells are removed.

Kazzaz and coworkers have shown that 48 h of expo-sure to hyperoxia (100% oxygen) in mice resulted in in-creased lung cell apoptosis (33), but lung cells in culturedied by necrosis with similar injury. In their study, timepoints other than 48 h were not reported and the specificcell types undergoing apoptosis in the lungs were not iden-tified. As the authors speculated, it is possible that the ap-optosis seen in mice may have been a “downstream phe-nomenon,” after the initial injury. This may be correlatedwith our findings of increased apoptosis 24 h after OA in-jury, which appears to be temporally related to the onsetof the repair mechanisms. However, the factor(s) involvedin this hypothetical switch from injury and inflammationto apoptosis and resolution remains to be identified.

Figure 5. Lung sections stained with H&E, obtained 4 to 72 h af-ter OA administration. (a) Control rat 24 h after intravenoussaline. (b) Acute inflammatory response at 4 h after OA. ( c)Decreasing inflammation at 24 h after OA. (d) Resolving inflam-matory response at 72 h after OA. Note the apoptotic bodies inmacrophages (arrows). Bar 5 60 m.

874 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 19 1998

The probable course of events that occurs during acutelung inflammation and resolution is that an initial PMN in-flux occurs immediately in response to the insult or injury.Upon termination of the injury, PMN influx wanes andPMN apoptosis is initiated by mechanisms hitherto un-known. This is associated with macrophage ingestion ofapoptotic PMNs and effective clearance without liberationof proteases or inflammatory mediators (as suggested byHaslett and colleagues [41]), thus circumventing the in-flammatory response. The AM themselves do not undergostimulation or respiratory burst (41), and therefore thewhole process clears effectively. Some defunct AM maythemselves undergo apoptosis and clearance by neighbor-ing cells.

It is possible that acute inflammation does not resolveeither if (1) the injury persists, (2) apoptosis of neutrophilsis not initiated, (3) the macrophages are not able to clearthe apoptotic neutrophils, or (4) a combination of thesefactors occurs. Further investigations along these linesmay yield vital clues to optimizing the mechanisms of heal-ing and repair.

Acknowledgment: The authors thank Susan Kreuger and Frank Morgan of theCenter for Biomedical Imaging Technology for their expert technological as-sistance. They are also grateful to Dr. Melinda Sanders for her help with theprocessing of tissue for paraffin sections, and Dr. Arthur Hand for help withelectron microscopy. This study was funded in part by the University of Con-necticut Health Center Faculty Research Grant and the American Lung Asso-ciation Connecticut Affiliate Research Awards (N.H.).

References

1. Hughes, J., R. Johnson, A. Mooney, C. Hugo, K. Gordon, and J. Savill.1997. Neutrophil fate in experimental glomerular capillary injury in therat. Am. J. Pathol. 150:223–234.

2. Kerr, J. F. R., A. H. Wyllie, and A. R. Currie. 1972. Apoptosis: basic biolog-ical phenomenon with wide ranging implications in tissue kinetics. Br. J.Cancer 26:239–257.

3. Wyllie, A. H., J. F. R. Kerr, and A. R. Currie. 1980. Cell death: the signifi-cance of apoptosis. Int. Rev. Cytol. 68:251–306.

4. Wyllie, A. H., R. G. Morris, A. L. Smith, and D. Dunlop. 1984. Chromatincleavage in apoptosis: association with condensed chromatin morphologyand dependence on macromolecular synthesis. J. Pathol. 142:67–77.

5. Savill, J. 1994. Apoptosis in disease. Eur. J. Clin. Invest. 24:715–723.6. Grigg, J. M., J. S. Savill, C. Sarraf, C. Haslett, and M. Silverman. 1991. Neu-

trophil apoptosis and clearance from neonatal lungs. Lancet 338:720–722.7. Cox, G., J. Crossley, and Z. Xing. 1995. Macrophage engulfment of apop-

totic neutrophils contributes to the resolution of acute pulmonary inflam-mation in vivo. Am. J. Respir. Cell Mol. Biol. 12:232–237.

8. Dickey, B. F., R. S. Thrall, J. R. McCormick, and P. A. Ward. 1981. Oleicacid-induced injury in the rat: failure of indomethacin treatment or com-plement depletion to ablate lung injury. Am. J. Pathol. 103:376–383.

9. Julien, M., J. J. M. Hopffel, and M. R. Flick. 1986. Oleic acid lung injury insheep. J. Appl. Physiol. 60:433–440.

10. Selig, W. M., C. E. Patterson, D. P. Henry, and R. A. Rhoades. 1986. Roleof histamine in acute oleic acid-induced lung injury. J. Appl. Physiol.61:233–239.

11. Brusch, W., S. Paffe, B. Putz, G. Barthel, and R. Schulte-Hermann. 1990.Determination of the length of the histological stages of apoptosis in nor-mal liver and in altered hepatic foci of rats. Carcinogenesis 11:847–853.

12. Gavrieli, Y., Y. Sherman, and S. A. Ben-Sasson. 1992. Identification of pro-grammed cell death in situ via specific labeling of nuclear DNA fragmen-tation. J. Cell Biol. 119:493–501.

13. Ben-Sasson, S. A., Y. Sherman, and Y. Gavrieli. 1995. Identification of dy-ing cells—in situ staining. Methods Cell Biol. 46:29–39.

14. Watts, F. L., B. L. Oliver, G. M. Johnson, and R. S. Thrall. 1990. Superoxideproduction by rat neutrophils in the oleic acid model of lung injury. Free

Radical Biol. Med. 9:327–332.15. Shiue, S. T., and R. S. Thrall. 1991. Effect of corticosteroid therapy on the

acute injury and recovery stage of oleic acid induced lung injury in the rat.Exp. Lung Res. 17:629–638.

16. Thrall, R. S., R. W. Barton, D. A. D’Amato, and S. B. Sulavik. 1982. Differ-ential cellular analysis of bronchoalveolar lavage fluid obtained at variousstages during the development of bleomycin-induced pulmonary fibrosisin the rat. Am. Rev. Respir. Dis. 126:488–492.

17. Schmitz, G. G., R. Seibl, T. Walter, and C. Kessler. 1991. Nonradioactive la-beling of oligonucleotides in vitro with the hapten digoxigenin by tailingwith terminal transferase. Anal. Biochem. 192:222–231.

18. Gorczyca, W., J. Gong, and Z. Darzynkiewicz. 1993. Detection of DNAstrand breaks in individual apoptotic cells by the in situ terminal deoxynu-cleotidyl transferase and nick translation assays. Cancer Res. 53:1945–1951.

19. Mizgerd, J. P., R. M. Molina, R. C. Stearns, J. D. Brain, and A. E. Warner.1996. Gadolinium induces macrophage apoptosis. J. Leuk. Biol. 59:189–195.

20. Narula, J., N. Haider, R. Virmani, T. G. DiSalvo, F. D. Kolodgie, R. J. Haj-jar, U. Schmidt, M. J. Semigran, G. W. Dec, and B. Khaw. 1996. Apoptosisin myocytes in end stage heart failure. N. Eng. J. Med. 335:1182–1189.

21. Colton, T. 1974. Inference of means. In Statistics in Medicine. T. Colton, ed-itor. Little, Brown & Co., Boston. 99–150.

22. Volpe, B. T., W. Lin, and R. S. Thrall. 1994. Effect of intratracheal dexa-methasone on oleic acid-induced injury in the rat. Chest 106:583–587.

23. Zakeria, Z., D. Quaglino, and H. S. Ahuja. 1994. Apoptotic cell death inthe mouse limb and its supression in the hammertoe mutant. Dev. Biol.165:294–297.

24. Compton, M. M., and J. A. Cidlowski. 1992. Thymocyte apoptosis—a modelof programmed cell death. Endo. 3:17–23.

25. Tilly, J. L. 1993. Ovarian follicular atresia: a model to study the mechanismsof physiological cell death. Endocrine J. 1:67.

26. Wyllie, A. H. 1981. Cell death: a new classification separating apoptosisfrom necrosis. In Cell Death in Biology and Pathology. I. D. Bowen andR. A. Lockshin, editors. Chapman & Hall, New York. 9–34.

27. Kresch, M. J., C. Christian, F. Wu, and N. Hussain. 1998. Ontogeny of apop-tosis during lung development. Pediatr. Res. 43:426–431.

28. Haslett, C. 1992. Resolution of acute inflammation and the role of apoptosisin the tissue fate of granulocytes [Editorial]. Clin. Sci. 83:639–648.

29. Desmouliere, A., M. Redard, I. Darby, and G. Gabbiani. 1995. Apoptosismediates the decrease in cellularity during the transition between granula-tion tissue and scar. Am. J. Pathol. 146:56–66.

30. Tidball, J. G., and B. A. St. Pierre. 1996. Apoptosis of macrophages duringthe resolution of muscle inflammation. J. Leuk. Biol. 59:380–388.

31. Hagimoto, N., K. Kuwano, H. Miyazali, R. Kunitake, M. Fujita, M. Ka-wasaki, Y. Kaneko, and N. Hara. 1997. Induction of apoptosis and pulmo-nary fibrosis in mice in response to ligation of Fas antigen. Am. J. Respir.Cell Mol. Biol. 17:272–278.

32. Hagimoto, N., K. Kuwano, Y. Nomoto, R. Kunitake, and N. Hara. 1997.Apoptosis and expression of Fas/Fas ligand mRNA in bleomycin-inducedpulmonary fibrosis in mice. Am. J. Respir. Cell Mol. Biol. 16:91–101.

33. Kazzaz, J. A., J. Xu, T. A. Palaia, L. Mantell, A. M. Fein, and S. Horowitz.1996. Cellular oxygen toxicity: oxidant injury without apoptosis. J. Biol.Chem. 271:15182–15186.

34. Polunovsky, V. A., B. Chen, C. Henke, D. Snover, C. Wendt, D. H. Ingbar,and P. B. Bitterman. 1993. Role of mesenchymal cell death in remodelingafter injury. J. Clin. Invest. 92:388–397.

35. Cox, G. 1995. Glucocorticoid treatment inhibits apoptosis in human neutro-phils: separation of survival and activation outcomes. J. Immunol. 154:4719–4725.

36. Cox, G. 1996. IL-10 enhances resolution of pulmonary inflammation in vivoby promoting apoptosis of neutrophils. Am. J. Physiol. 271:L566–L571.

37. Hurley, J. V. 1983. Termination of acute inflammation: I. Resolution. InAcute Inflammation. J. V. Hurley, editor. Churchill Livingstone, London.109–117.

38. Meagher, L. C., J. S. Savill, A. Baker, R. W. Fuller, and C. Haslett. 1993. Phago-cytosis of apoptotic neutrophils does not induce macrophage release ofthromboxane B2. J. Leuk. Biol. 52:269–273.

39. Kar, S., Y. Ren, C. Haslett, and J. S. Savill. 1993. Inhibition of macrophagephagocytosis in vitro of aged neutrophils undergoing apoptosis increasesrelease of neutrophil content. Clin. Sci. 85:27P. (Abstr.)

40. Savill, J. S., A. H. Wyllie, J. E. Henson, P. M. Henson, and C. Haslett. 1989.Macrophage phagocytosis of aging neutrophils in inflammation. J. Clin. In-vest. 83:865–875.

41. Haslett, C., J. S. Savill, M. K. Whyte, M. Stern, I. Dransfield, and L. C.Meagher. 1994. Granulocyte apoptosis and the control of inflammation[Review]. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 345:327–333.