- Email: [email protected]
P-selectin is upregulated early in the course of hyperoxic lung injury in mice
Free RadicalBiology& Medicine,Vol. 21, No. 4, pp. 567-574, 1996 Copyright© 1996 ElsevierScienceInc. Printed in the USA.All rightsreserved 0891-5849/96 $15,00 + .00 ELSEVIER
PII S0891-5849(96)00121-9
Brief Communication P-SELECTIN
IS
UPREGULATED
HYPEROXIC
LUNG
EARLY
IN
THE
INJURY
IN
MICE
COURSE
OF
T. ZEB,* B. PIEDBOEUF, t M. GAMACHE, ~: C. LANGSTON, t and S. E. WELTY* *Department of Pediatrics, Baylor College of Medicine, Houston, TX 77030, USA; tUnite de Pediatrie, Centre de Recherche du CHUL Universite, Ste Foy, Qurbec, Canada; and *Department of Pathology, Baylor College of Medicine, Houston, TX 77030, USA (Received 7 April 1995; Revised 10 October 1995; Re-Revised 24 January 1996; Accepted 22 February 1996)
Abstract--While treatment with supplemental oxygen is often essential in patients with lung disease, prolonged therapy may cause lung injury by itself. Although the mechanisms responsible for initiating hyperoxic lung damage almost certainly involve primary oxidative transformations, the possible contributions of inflammation to the tissue injury have been attracting increasing research activity. Increases in intercellular adhesion molecule-1 (ICAM-1) coincide with the inflammation, but in other models of inflammation transient adhesion mediated by members of the Selectin gene family was found to be essential before ICAM-1/#2 interactions could occur. We, therefore, wondered whether a similar sequence of initial transient adhesion followed by subsequent responses would be observed in hyperoxic lung inflammation. We, therefore, determined the effects of hyperoxia exposure on lung mRNA for Pand E-Selectin in mouse lungs. We found that there was no detectable mRNA for E-Selectin through 72 h of hyperoxia exposure by Northern blotting, but that mRNA for P-Selectin was detectable as early as 48 h after initiation of hyperoxia. To determine the location of P-Selectin upregulation we examined hyperoxia-exposed mouse lungs by in situ hybridization and found that the upregulation of P-Selectin at 48 h was localized to large muscularized vessels, at 72 h expression was detected in some medium size muscularized vessels, and at 96 h abundant expression was observed also on nonmuscularized small vessels. In conclusion, increases in mRNA for P-Selectin early in the course of hyperoxia exposure suggest that P-Selectin expression in hyperoxic lungs increases in parallel with upregulation of ICAM-1, leading to the accumulation of neutrophils in hyperoxic lungs, and that interventions targeting these two adhesion molecules may lead to a diminution in hyperoxic lung inflammation and lung injury. Keywords--Oxygen toxicity, Inflammation, Lung injury, Gene expression, P-Selectin, Free radicals
INTRODUCTION
sequence of adhesive interactions between neutrophils and e n d o t h e l i a l c e l l s are r e q u i r e d b e f o r e neut r o p h i l s can a c c u m u l a t e at a site o f i n f l a m m a t i o n , z'3 I n i t i a l i n t e r a c t i o n o f n e u t r o p h i l s with e n d o t h e l i a l c e l l s is c h a r a c t e r i z e d b y t r a n s i e n t a d h e s i o n and rolling o f the n e u t r o p h i l on e n d o t h e l i a l cells. 5-7 This initial t r a n s i e n t a d h e s i o n is m e d i a t e d b y m e m b e r s o f the s e l e c t i n g e n e f a m i l y i n c l u d i n g E- and P - S e l e c t i n on e n d o t h e l i a l c e l l s and L - S e l e c t i n on n e u t r o p h i l s . 3'8'9 T h e i n i t i a l t r a n s i e n t a d h e s i o n o f n e u t r o p h i l s on end o t h e l i a l cells is i n s u f f i c i e n t for n e u t r o p h i l a c c u m u l a t i o n in tissue u n l e s s a d d i t i o n a l i n t e r a c t i o n s occur, t° A f t e r the initial a d h e s i o n in an i n f l a m m a t o r y focus, t h e r e m a y be a d d i t i o n a l s t i m u l i that attract and activate n e u t r o p h i l s . 1~-~3 T h e final steps in the p r o c e s s
D e s p i t e a d v a n c e s in clinical m e d i c i n e and b i o t e c h n o l ogy, the m a i n s t a y o f therapy for patients with respiratory failure is the administration o f s u p p l e m e n t a l o x y g e n . Unfortunately, p r o l o n g e d e x p o s u r e to supplemental o x y g e n m a y contribute to respiratory failure b y causing lung d a m a g e b y itself. Therefore, investigations o f the p a t h o g e n e s i s o f h y p e r o x i c lung injury continue, and recently there has been increasing interest in the potential contribution o f acute inflammation to the course o f h y p e r o x i c lung d i s e a s e J In s e v e r a l e x p e r i m e n t a l m o d e l s o f i n f l a m m a t i o n a Address correspondence to: Stephen E. Welty, M.D., One Baylor Plaza, Room 337D, Houston TX, 77030. 567
T. ZEB et al.
568
leading to neutrophil accumulation in a tissue are firm adhesion and transendothelial migration, and are often mediated by neutrophil f12 integrins with I C A M - 1 on endothelial cells. 14-~6 In previous studies we determined that the expression of ICAM-I in mouse lungs was increased in response to hyperoxia-exposure, messenger RNA being increased at 48 h and protein being increased at 72 h, and that the increased expression was localized to cells in the distal airspaces, which is also the area in which neutrophil accumulation o c c u r s . 17 Other studies using monoclonal antibodies indicate that ICAM-1 upregulation and interactions with neutrophil ]~2 integrins contribute significantly to hyperoxic lung inflammation and injury, ors In order to determine whether hyperoxic lung inflammation might also conform to the multiple adhesive interaction model for the development of inflammation, we examined hyperoxic mouse lungs for the expression of E- and P-Selectin mRNA. Finding increases in P-Selectin m R N A in mouse lungs well before the development of inflammation, we used in situ hybridization to determine the cell specific localization of increased P-Selectin expression. Because the upregulation of P-Selectin is one of the earliest changes we have identified in relationship to potential mechanisms of hyperoxic lung inflammation, it may be the step at which hyperoxic inflammatory responses may be controlled more effectively, in the context of other studies, suggesting that controlling the inflammatory response may be helpful. MATERIALS AND METHODS
Experimental protocol We studied male FVB mice that were approximately 6 weeks old, weighing 20 to 25 g. We placed the mice in a Plexiglas chamber that allowed them free access to food and water, and exposed them to greater than 95% oxygen continuously by administration of pure oxygen at 2 liters per minute while removing excess carbon dioxide using soda lime (Sodasorb, Dewey & Almy Chemical Division, Grace & Co., Lexington, MA.). We observed the mice and measured the oxygen concentration twice daily during hyperoxia exposure. In order to obtain mouse lungs for the analyses, we sacrificed our mice with 200 mg/kg of pentobarbital, IP. We sacrificed mice after 12, 24, 48, 72, or 96 h of hyperoxia exposure for RNA studies. We removed the lungs via a midline thoracotomy, placed the lungs in liquid nitrogen, and stored them at - 8 0 ° prior to isolation of total RNA. Air-breathing mice served as controis in all experiments and are designated as 0 h in > 9 5 % oxygen.
In studies for in situ hybridizations we cannulated the tracheas of mice with PE 50 tubing after administering pentobarbital. We then perfused the airways with zinc formalin (ANATECH, Battle Creek, MI) until the lungs were fully distended, and removed the heart and lung in block for subsequent paraffin embedding.
Isolation and analysis of RNA We isolated total lung RNA using a modification of the procedure of Chomcynzski and Sacchi j9 described by Hamburg and co-workers, z° After isolating RNA, we resuspended the RNA in 10 mM Tris pH 7.6, 1 mM EDTA, and quantified the amount of RNA by measuring the absorption at 260 nm. We loaded 20 #g of total RNA per sample per lane on a 1% agarose/formaldehyde denaturing gel, separated the RNA by electrophoresis and transferred the separated RNA from the denaturing gel to a nitrocellulose filter. We then allowed the nitrocellulose filter to air dry at 68 ° for 3 h.
RNA hybridization We used the partial complementary DNA for murine E-Selectin that was generously provided by C. M. Ballantyne, which had been cloned into pBluescript S K + (Stratagene, La Jolla, CA), and we generated a partial complementary DNA for murine P-Selectin using reverse transcriptase polymerase chain reaction (PCR) of total lung RNA isolated from the lung of a mouse exposed to 2 mg/kg of lipopolysacharide (Salmonella abortus equi, Sigma Chemical Comp., St. Louis, MO). In the reverse transcriptase PCR, we designed downstream and upstream primers based on the published sequence of the murine P-Selectin, and which also contained restriction sites on their 5' ends to facilitate ligation of the PCR product into pBluescript S K + (Strategene, La Jolla, CA). We made radioactive antisense transcripts to E- and P-Selectin m R N A with T7 RNA polymerase and nucleotide triphosphates with 32p-labeled uridine triphosphate (3000 Ci/mmol, Amersham Life Sciences, Arlington Heights, IL). After making the antisense transcripts, we prehybridized the nitrocellulose filters containing separated RNA in 50% formamide, 50 rnM Na2PO4, 0.8 M NaC1, 1.0 mM EDTA, 0.5× Denhardt's reagent, and 250 #g/ml of heat-denatured herring sperm DNA for approximately 3 h at 55 °. After prehybridization, we discarded the solutions and hybridized the nitrocellulose filters at 55 ° for 16 h using the same solution used for prehybridization and adding the radioactive antisense transcripts. We then washed the hybridized filters three times for 20 rain with 0.1 × SSC and 0.1% SDS at 68 ° to decrease nonspecific binding of radioactive transcripts to the filter. Following
P-selectin in oxygen toxicity these washes we exposed x-ray film to the hybridized filters to generate autoradiograms. In every Northern blot hybridization in which we analyzed R N A from hyperoxic lung tissue we included an analysis of R N A from lungs o f mice treated with 2 mg/kg o f IP lipopolysaccharide (Salmonella abortus equi, Sigma Chemical Company, St. Louis, MO). Because endotoxin is a potent inducer o f E- and P-selectin in mouse lungs, these R N A samples served as our positive controls.
569
=888888888888888888
28S
.
.
.
.
18S
28S In situ hybridization We made complementary R N A (cRNA) probe labeled with [33p]-UTP (DuPont Canada, Markham, Ont., Canada) as described previously. 2~ After making the c R N A probe, we digested the D N A template with R N A s e free D N A s e I (Promega Corp., Madison, WI), and extracted the resulting solution containing c R N A probe with an equal volume of phenol/chloroform (1:1), followed by ethanol precipitation and resuspension in diethylpyrocarbonate-treated water. Before hybridization, we performed limited alkaline hydrolysis o f the c R N A probe to reduce transcript length to between 100 and 300 bp. 22 We determined the size o f the
888881888888888888._
- '
18S
Fig, 2. Autoradiogram from a Northern blot hybridization for PSelectin messenger RNA in hyperoxia-exposed mouse lungs. Mice were exposed to 0, 24, 48, 72, and 96 h of >95% oxygen continuously, or 2 mg/kg of endotoxin IP and lungs obtained (2 h after endotoxin dosing) for isolation of total RNA. Twenty micrograms of RNA was separated electrophoretically, transferred to a nitrocellulose membrane, and hybridized with -~2P-labeled antisense P-Selectin transcripts. Signal for P-Selectin is clearly evident just below the 28S ribosomal RNA signal in the lung RNA from mice exposed to endotoxin, and there is detectable signal in lung RNA from mice exposed to 48 through 96 b of hyperoxia. The ethidium bromidestained filter in the lower part of the figure coincides to this autoradiogram and documents equivalent quantity and quality of RNA.
~
28S
-,,~--18S
28S 18S Fig. 1. Autoradiogram from a Northern blot hybridization for ESelectin messenger RNA in hyperoxia-exposed mouse lungs. Mice were exposed to 0, 12, 24, 48, and 72 h of > 9 5 % oxygen continuously, or 2 mg/kg of endotoxin IP and lungs obtained (2 h after endotoxin dosing) for isolation of total RNA. Twenty micrograms of RNA was separated electrophoretically, transferred to a nitrocellulose membrane, and hybridized with 32P-labeled antisense E-Selectin transcripts. Signal for E-Selectin is clearly evident in the lung RNA from mice exposed to endotoxin just below the 28S ribosomal RNA signal, but there is no detectable signal in lung RNA from mice exposed to up to 72 h of hyperoxia. The ethidium bromidestained filter in the lower part of the figure coincides with the autoradiogram and documents equivalent quantity and quality of RNA.
partially hydrolyzed transcripts by agarose gel electrophoresis. For hybridization, we used methods described previously. 2L23 W e carried out hybridization on lung sections overnight at 54 ° in 50% formamide, 0.3 M NaC1, 10 m M Tris-C1 (pH 8.0), 1 m M E D T A , 1× Denhardt's solution, 10% dextran sulfate, 0.5 mg/ml yeast transfer RNA, and 17 ng/ml o f c R N A probe in a moist chamber containing 0.3 M NaC1, 50% formamide. After hybridization, we treated the sections with RNAse, 2~ followed by a stringent wash at 68 ° in 0.1× SSC for 30 min. W e exposed the sections to N B T - 2 photographic emulsion (Kodak Company, Rochester, NY) for 10 d for autoradiography and then counterstained the sections with hematoxylin and eosin before photomicrography. In all experiments we evaluated nonspecific binding using sense c R N A probe.
RESULTS
General appearance No mice died in hyperoxia exposure before the times chosen to obtain the lungs for analysis. Furthermore, the mice had no visual evidence of illness through 72 h o f hyperoxia exposure. At 96 h of hyperoxia exposure
/
~t:,:k
0 Fig. 3. In situ hybridizations for P-Selectin mRNA in hyperoxia-exposed mouse lungs. Mice were exposed to >95% oxygen for up to 96 h, the lungs inflated with and fixed in zinc formalin, and embedded in paraffin. Sections were cut and hybridized with antisense P-Selectin labeled with 3sp. The sections were then developed by autoradiography. In the sections from A to F there are two figures of identical sections, one examined with standard light microscopy, and one examined utilizing dark-field microscopy. The purpose of the dark-field examination is to illustrate a signal for P-Selectin mRNA, which is identified by bright staining (white arrows). The purpose of the light microscopy examination is to identify the regions in the lung where the PSelectin mRNA signal is observed in the corresponding dark-field section (at low power, signal for P-Selectin mRNA is not detectable in H&E sections). The section labeled 3A is a representative lung section (10×) from a mouse exposed to room air only. There is no detectable P-Selectin mRNA observed by dark-field examination. The section labeled 3B is a representative lung section (10×) from a mouse exposed to 48 h of byperoxia. In the section P-Selectin mRNA is clearly evident in endothelial cells in large muscularized vessels (white arrows on dark field) that are identifiable as arteries by light microscopy because of the vessel's proximity to the large airway (black arrow). Note also that P-Selectin is not upregulated throughout the artery, suggesting that the upregulation is focal even within the artery examined. The section labeled 3C is a higher power view of the same section to further illustrate the endothelial cell localization of the signal (white arrow). The figure labeled 3D is a representative section (20× for dark-field view and 10× for light microscopy) from a mouse exposed to 72 h of hyperoxia, and circumferential signal is observable in the medium sized vessel (white arrow on dark-field view), which is identifiable as an artery by its muscular wall (arrow on light microscopy). The sections labeled 3E (10×) and 3F (20×) are from a mouse exposed to 96 h of hyperoxia, In addition to the findings observed in mice exposed to shorter durations of hyperoxia, these lungs also have P-Selectin evident in even smaller lung vessels (white arrows on dark-field, black arrows on light microscopy). The smaller vessels are identifiable as arterial by the fact that they are relatively small vessels in an intralobular locale, whereas veins of this size would be interlobular in locale. While P-Selectin expression is detectable in these small vessels there is no detectable signal in alveolar capillaries.
P-selectin in oxygen toxicity
571
Fig. 3. Continued.
most mice had evidence of illness exhibited by an apparent increase in their work of breathing. Northern blot hybridizations and in situ hybridizations An autoradiogram of a Northern blot hybridization for E-Selectin is depicted in Fig. 1. E-Selectin m R N A signal is clearly evident visually just below the 28S
signal in the lung RNA from animals exposed to endotoxin, whereas there is no detectable E-Selectin in the lungs of mice exposed to hyperoxia for up to 72 h. The ethidium bromide-stained membrane at the bottom of Fig. 1 documents even loading. An autoradiogram of a Northern blot hybridization for P-Selectin is depicted in Fig. 2. In both lung RNA samples from endotoxin-treated mice there is signal for P-Selectin m R N A clearly evident. In addition, lung RNA samples
572
T. ZEB et
from mice exposed to 48 h or more of hyperoxia have signals for P-Selectin evident. To study the cell-specific regulation of P-Selectin mRNA, in situ hybridizations were performed. In lungs from air-breathing mice (Fig. 3A), P-Selectin transcripts were not detected. After 48 h of hyperoxia exposure, P-Selectin mRNA was detected in endothelial cells of the lungs. However, the intensity of the expression as well as the type of vessel where the transcripts were observed changed with the length of exposure in hyperoxia. After 48 h, P-Selectin transcripts were observed in large muscularized vessels (Fig. 3B), which are identifiable as arteries because of the proximity to the airways and the thickness of the muscular walls. After 72 h of exposure, P-Selectin transcripts were detected in medium size vessels, which are identifiable as arteries by their proximity again to airways (Fig. 3C), and after 96 h of hyperoxia exposure, abundant expression of P-Selectin transcripts was observed (3D) along with expression in nonmuscularized smaller vessels (Fig. 3E). P-Selectin transcripts were not detected in capillaries or in sections hybridized with radioactive sense transcripts, although we cannot rule out the possibility that P-Selectin was upregulated on the capillary interstitium, but was below the limits of detection. DISCUSSION Recent evidence suggests that lung inflammation is an important contributor to the development of hyperoxic lung injury, and that specific therapy directed against neutrophil adhesion may be useful. To date, the evidence targeting neutrophil adhesion in hyperoxic lung injury has been directed against neutrophil/32 integrins and one of their ligands, ICAM-I. The interactions of f12 integrins with ICAM- 1 mediates firm adhesion, ~4-16 and these interactions are dependent on prior transient adhesion of neutrophils endothelial cells. Lawrence and co-workers in studies in vitro in which neutrophil adhesion was assessed under conditions of flow, found that ICAM-1 -mediated neutrophil adhesion could not take place until P-Selectin-mediated neutrophil adhesion occurred. From the studies by Lawrence et al. and other investigators, a concept has emerged that multiple adhesive interactions between neutropbils and endothelial cells are required before neutrophil accumulation and neutrophil-dependent cytotoxicity c a n o c c u r . 3'24 If neutrophil accumulation in the hyperoxia-exposed lungs is also dependent on sequential neutrophil interactions with adhesion molecules, then the studies supporting a role for/32 integrins and ICAM-1 are addressing a more limited picture of
al.
mechanisms leading to neutrophil accumulation in hyperoxic lungs. Our finding of increased P-Selectin mRNA early in the course of hyperoxia-exposed lungs suggests a role for P-Selectin in hyperoxia-induced lung inflammation, similar to that indicated by Lawrence et al. in their model of inflammation. The lack of E-Selectin upregulation by northern blotting suggests that E-Selectin is not involved in transient adhesion in hyperoxia-exposed lungs. The lack of demonstrable upregulation in E-Selecting is in contrast to our findings in hyperoxia-exposed rats, in which ESelectin is upregulated early in the hyperoxia-exposure. 25 The most likely explanation for differences in rat and mouse E-Selectin expression in hyperoxia is that the regulation of E-Selectin gene is dependent on the species studied. Of particular interest is the localization of the PSelectin during hyperoxia exposure. The fact that PSelectin upregulation occurs in the large, medium, and small arterial vessels of the lung with no detectable mRNA expression in capillaries suggests that P-Selectin gene transcription in endothelial cells of the larger vessels is more sensitive to increases in oxidative stress, although we cannot absolutely rule out that capillary expression is present but below the level of detection using in situ hybridization. An apparent inconsistency with the hypothesis that P-Selectin gene transcription is sensitive to oxidative stress is the finding that the earliest upregulation of P-Selectin occurs in pulmonary artery endothelium (prior to exposure to the high alveolar PO2) rather than in the pulmonary capillaries or veins (directly exposed to high PO2). In the in situ hybridizations the arteries are distinguished from the veins by determining whether the vessel involved is in proximity to airways, because of the fact that arteries branch with the airways into the lung and the veins return blood from the capillaries between lung lobes. Surprisingly, using these criteria for the identification of pulmonary arteries and veins, the upregulation is most marked on pulmonary arteries (Fig. 3B, note the large airway adjacent to the vessel with P-Selectin mRNA evident in the endothelium). The next issue is the functional significance of an increase in P-Selectin mRNA in lung endothelial cells. P-Selectin protein has been shown be present in platelets and endothelial cells in preformed granules that may be translocated rapidly to the cell surface in response to endogenous stimuli, such as histamine and thrombin) 6'27 In addition, some investigators have observed increases in P-Selectin mRNA in tissues of animals exposed to lipopolysaccharide 2. and in endotheliomas exposed to tumor necrosis factor-o<29 In the study of endotheliomas the increase in P-Selectin mRNA was associated with increased P-Selectin pro-
P-selectin in oxygen toxicity
tein, and the investigators speculated that P-Selectin protein was upregulated by increased transcription. It is likely that the observed increase in P-Selectin mRNA leads to an increase in P-Selectin protein, though we have not tested this hypothesis. The role of the upregulation of P-Selectin in hyperoxic lung inflammation is still undetermined. There are basically two different views in regard to the role that P-Selectin might play in arterial side of the pulmonary circuit. The first view is simply that upregulation of P-Selectin in the arterial side of the pulmonary circuit plays no role in hyperoxic lung inflammation. The second view is that the upregulation does play a role in hyperoxic lung inflammation by mediating transient adhesion and rolling of neutrophils in the pulmonary circulation. P-Selectin upregulation on the arterial side of the pulmonary circulation would place P-Selectin in a favorable position to mediate transient adhesion and rolling prior to the pulmonary capillaries, at which point firm adhesion could occur via different mechanisms at the level of the alveolar capillary, which is the site in which neutrophils have been observed to enter lung tissue. 3° If this second view is correct, it would be in contrast to inflammatory responses in the systemic circulation in which neutrophil adhesion and transendothelial migration are well established as occurring in the venules. Support for this contrast in localization for neutrophil adhesion and transendothelial migration in the lung is supported by studies in which delays were observed in lung neutrophil transit proximal to pulmonary veins. 31 The altered cell specific expression of P-Selectin as a function of hyperoxia exposure time is interesting demands further study. Studies of the 5' flanking sequences of P-Selectin to determine sequences crucial to the cell specific expression is likely to be fruitful in determining the mechanisms for the alterations in cellular expression as a function of hyperoxia exposure. The P-Selectin promoter studies could be carried out in in vitro transfection studies in which cells are transfected with serial deletions of the 5' flanking region of P-selectin ligated to a reporter gene, after which the cells would be exposed to oxidants, or in in vivo studies in which mice made transgenic with a similar series of promoter-reporter constructs are exposed to hyperoxia. Determining the role of individual adhesion molecules in hyperoxic lung inflammation and injury is an extremely complex issue. To illustrate the potential confusion, studies in which anti-ICAM-1 and antiCD 1 1a antibodies were administered to hyperoxia-exposed mice the treatment was protective, while in studies from our own lab knockout mice deficient in ICAM-1 and with reduced CD18 expression are not protected (unpublished results). There are a number of potential reasons for this apparent contradiction in ex-
573
perimental results, such as that temporal and spatial interactions of adhesion molecules with their ligands is more complex than anticipated. For example, in hyperoxia-exposed animals there may be specific durations of exposure or specific regions wherein adhesion molecule interactions may mediate "appropriate" inflammatory responses, while at other durations or in other regions adhesion molecule interactions may mediate "inappropriate" inflammatory responses. The complexity of the inflammatory response and the redundancy of the adhesion molecule system make the issue of determining the role of inflammation in hyperoxic lung injury problematic. However, molecular approaches in which adhesion molecule expression can be tightly regulated and more than one adhesion molecule affected will be helpful in better understanding the role of inflammation in hyperoxic lung injury and in designing more specific interdictions when appropriate. Furthermore, expression studies such as the present study are essential as a foundation on which to determine which molecules are likely to be involved in an inflammatory response. For example, the present studies suggest that exploring E-Selectin in hyperoxic mice in more detailed studies is unwarranted, while studying the role of P-Selectin expression by itself and in concert with other adhesion molecules is appropriate. In summary, P-Selectin mRNA is upregulated early in the course of hyperoxic lung injury and is localized primarily to the arterial side of the pulmonary circuit. These findings support the idea that P-Selectin participates in hyperoxic inflammation and that future studies inhibiting P-Selectin function or upregulation are necessary to determine the extent of the role of P-Selectin in hyperoxic lung injury and inflammation. Our finding that E-Selectin mRNA is not upregulated supports the idea that E-Selectin does not play a role in hyperoxic lung inflammation. At this time we cannot rule out the possibility that E-Selectin is upregulated in hyperoxic lungs, but that Northern blot hybridization is simply not sensitive enough to detect the upregulation. Acknowledgements- This work was supported by NIH Grant AI19031, NIH Grant 5 P30 HD27823-04, and an A H A grant, Texas Affiliate.
REFERENCES 1. Wegner, C. D.; Wolyniec, W. W.; LaPlante, A. M.; Marschman, K.; Lubbe, K.; Haynes, N.; Rothlein, R.; Letts, G. L. Intercellular adhesion molecule-1 contributes to pulmonary oxygen toxicity in mice: Role of leukocytes revisited. Lung 170:267-279; 1992. 2. Lawrence, M. B.; Smith, C. W.; Eskin, S. G.; Mclntire, L. V. Effect of venous shear stress on CD18-mediated neutrophil adhesion to cultured endothelium. Blood 75:227-237; 1990. 3. Lawrence, M. B.; Springer, T. A. Leukocytes roll on a selectin
574
4. 5, 6. 7. 8.
9.
10.
11.
12.
13. 14.
15.
16.
17.
18.
T. ZEB et al. at physiologic flow rates: Distinction from and prerequisite for adhesion through integrins. Cell 65:859-873; 1991. Butcher, E. C. Leukocyte-endothelial cell recognition: Three (or more) steps to specificity and diversity. Cell 67:1933-1936; 1991. Anderson, B. O.; Brown, J. M.; Shanley, P. F.; Bensard, D. D.; Harken, A. H. Marginatiing neutrophils are reversibly adherent to normal lung endothelium. Surgery 109:51-61; 1991. Springer, T. A.; Lasky, L. A. Sticky sugars for selectins. Nature 349:196-197; 1991. Patarroyo, M.; Makgoba, M. W. Leukocyte adhesion to cells in immune and inflammatory responses. Lancet 2:1139-1142; 1989. Ley, K.; Gaehtgens, P.; Fennie, C.; Singer, M. S.; Lasky, L. A.; Rosen, S. D. Lectin-like cell adhesion molecule 1 mediates leukocyte rolling in mesenteric venules in vivo. Blood 77:25532555; 1991. Hammer, D. A.; Apte, S. M. Stimulation of cell rolling and adhesion on surfaces in shear flow: General results and analysis of selectin-mediated neutrophil adhesion. Biophysiol. J. 63:35-57; 1992. Arfors, K. E.; Lundberg, C.; Lindbom, L.; Lundberg, K.; Beatty, P. G.; Harlan, J. M. A monoclonal antibody to the membrane glycoprotein complex CD18 inhibits polymorphonuclear leukocyte accumulation and plasma leakage in vivo. Blood 69:338340; 1987. Carden, D. L.; Smith, J. K.; Korthuis, R. J. Oxidant-mediated, CD 18-dependent microvascular dysfunction induced by complement-activated granulocytes. Am. J. Physiol. 260:H1144-H1152; 1991. Mulligan, M. S.; Varani, J.; Warren, J. S.; Till, G. O.; Smith, C. W.; Anderson, D. C.; Todd, R. F.; Ward, P. A. Roles of /32 integrins of rat neutrophils in complement and oxygen radical-mediated acute inflammatory injury. J. lmmunol. 148:1847-1857; 1992. Westwick, J.; Li, S. W.; Camp, R. D. Novel neutrophil-stimulating peptides, lmmunol. Today 10:146-147; 1989. Smith, C. W.; Marlin, S. D.; Rothlein, R.; Toman, C.; Anderson, D. C. Cooperative interactions of LFA- 1 and Mac- 1 with intercellular adhesion molecule-I in facilitating adherence and transendothelial migration of human neutrophils in vitro. J. Clin. Invest. 83:2008-2017; 1989. Smith, C. W.; Rothlein, R.; Hughes, B. J.; Mariscalco, M. M.; Rudloff, H. E.; Schmalsteig, F. C.; Anderson, D. C. Recognition of an endothelial determinant for CD18-dependent human neutrophil adherence and transendothelial migration. J. Clin. Invest. 82:1746-1756; 1988. Furie, M. B.; Tancinco, M. C. A.; Smith, C. W. Monoclonal antibodies to leukocyte integrins CDIIa/CD18 and CDllb/ CD 18 or intercellular adhesion molecule- 1 inhibit chemoattracrant-stimulated neutrophil transendothelial migration in vitro. Blood 78:2089-2097; 1991. Welty, S. E.; Rivera, J. L.; Geske, R. S.; Montgomery, C. A.; Hansen, T. N. Increases in intercellular adhesion molecule-1 (ICAM-1) in hyperoxic mouse lungs are localized to cells in the alveolar interstitium. Pediat. Res. 32:160A; 1993. Ramsay, P. L.; Welty, S. E.; Bree, M. P.; Hansen, T. N. Lung
19.
20. 21.
22.
23.
24. 25. 26.
27. 28. 29.
30. 31.
injury and inflammation are reduced in hyperoxic mice by inhibition of neutrophil /32 integrin CDlla/CDI8. Pediatr. Res. 35:396A; 1994. Chomczynski, P.; Sacchi, N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem. 162:156-159; 1987. Hamburg, D. C.; Tonoki, H.; Welty, S. E.; Montgomery, C. A.; Hansen, T. N. Endotoxin induces glutathione reductase activity in lungs of mice. Pediatr. Res. 35:311-315; 1994. Piedboeuf, B.; Johnston, C. J.; Watkins, R. H.; Hudak, B. B.; Lazo, J. S.; Cherian, M. G.; Horowitz, S. Increased expression of tissue inhibitor of metalloproteinases (TIMP-1) and metallothionein in murine lungs after hyperoxic exposure. Am. J. Respir. Cell Mol. Biol. 10:123-132; 1994. Cox, K. H.; DeLeon, D. V.; Angerer, L. M.; Angerer, R. C. Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes. Dev. Biol. 101:485-502; 1984. Veness-Meehan, K. A.; Cheng, E. R.; Mercier, C. E.; Blizt, S. L.; Johnston, C. J.; Watkins, R. H.; Horowitz, S. Cellspecific alterations in expression of hyperoxia-induced mRNAs of lung. Am. J. Respir. Cell Mol. Biol. 5:516-521; 1991. Albelda, S. M.; Smith, C. W.; Ward, P. A. Adhesion molecules and inflammatory injury. FASEB J. 8:504--512; 1994. Ramsay, P. L.; Welty, S. E. E-selectin increases in hyperoxic lungs and is associated with inflammation and injury. Am. J. Respir. Crit. Care Med. 151:A343; 1995. Johnston, G. I.; Kurosky, A.; McEver, R. P. Structural and biosynthetic studies of the granule membrane protein GMP-140, from human platelets and endothelial cells. J. Biol. Chem. 264:1816-1823; 1989. Celi, A.; Furie, B.; Furie, B. C. PADGEM: An adhesion receptor for leukocytes on stimulated platelets and endothelial cells. Proc. Soc. Exp. Biol. Med. 198:703-709; 1991. Auchampach, J. A.; Oliver, M. G.; Anderson, D. C.; Manning, A. M. Cloning and sequence comparison and in vivo expression of the gene encoding rat P-Selectin. Gene 145:251-255; 1994. Weller, A.; Isenmann, S.; Vestweber, D. Cloning of the mouse endothelial selectins. Expression of both E- and P-selectin is inducible by tumor necrosis factor alpha. J. Biol. Chem. 267:15176-15183; 1992. Worthen, G. S.; Avdi, N.; Vukajiovich, S.; Tobias, P. S. Neutrophil adherence induced by lipopolysaccharide in vitro../. Clin. Invest. 90:2526-2535; 1992. Downey, G. P.; Doherty, D. E.; Schwab, B.; Elson, E. L.; Henson, P. M.; Worthen, G. S. Retention of leukocytes in capillaries: Role of cell size and deformability. J. Appl. Physiol. 69:17671778; 1990.
ABBREVIATIONS
I C A M - 1- - i n t e r c e l l u l a r adhesion molecule- 1 IP--intraperitoneal cRNA--
complementary RNA
PII S0891-5849(96)00121-9
Brief Communication P-SELECTIN
IS
UPREGULATED
HYPEROXIC
LUNG
EARLY
IN
THE
INJURY
IN
MICE
COURSE
OF
T. ZEB,* B. PIEDBOEUF, t M. GAMACHE, ~: C. LANGSTON, t and S. E. WELTY* *Department of Pediatrics, Baylor College of Medicine, Houston, TX 77030, USA; tUnite de Pediatrie, Centre de Recherche du CHUL Universite, Ste Foy, Qurbec, Canada; and *Department of Pathology, Baylor College of Medicine, Houston, TX 77030, USA (Received 7 April 1995; Revised 10 October 1995; Re-Revised 24 January 1996; Accepted 22 February 1996)
Abstract--While treatment with supplemental oxygen is often essential in patients with lung disease, prolonged therapy may cause lung injury by itself. Although the mechanisms responsible for initiating hyperoxic lung damage almost certainly involve primary oxidative transformations, the possible contributions of inflammation to the tissue injury have been attracting increasing research activity. Increases in intercellular adhesion molecule-1 (ICAM-1) coincide with the inflammation, but in other models of inflammation transient adhesion mediated by members of the Selectin gene family was found to be essential before ICAM-1/#2 interactions could occur. We, therefore, wondered whether a similar sequence of initial transient adhesion followed by subsequent responses would be observed in hyperoxic lung inflammation. We, therefore, determined the effects of hyperoxia exposure on lung mRNA for Pand E-Selectin in mouse lungs. We found that there was no detectable mRNA for E-Selectin through 72 h of hyperoxia exposure by Northern blotting, but that mRNA for P-Selectin was detectable as early as 48 h after initiation of hyperoxia. To determine the location of P-Selectin upregulation we examined hyperoxia-exposed mouse lungs by in situ hybridization and found that the upregulation of P-Selectin at 48 h was localized to large muscularized vessels, at 72 h expression was detected in some medium size muscularized vessels, and at 96 h abundant expression was observed also on nonmuscularized small vessels. In conclusion, increases in mRNA for P-Selectin early in the course of hyperoxia exposure suggest that P-Selectin expression in hyperoxic lungs increases in parallel with upregulation of ICAM-1, leading to the accumulation of neutrophils in hyperoxic lungs, and that interventions targeting these two adhesion molecules may lead to a diminution in hyperoxic lung inflammation and lung injury. Keywords--Oxygen toxicity, Inflammation, Lung injury, Gene expression, P-Selectin, Free radicals
INTRODUCTION
sequence of adhesive interactions between neutrophils and e n d o t h e l i a l c e l l s are r e q u i r e d b e f o r e neut r o p h i l s can a c c u m u l a t e at a site o f i n f l a m m a t i o n , z'3 I n i t i a l i n t e r a c t i o n o f n e u t r o p h i l s with e n d o t h e l i a l c e l l s is c h a r a c t e r i z e d b y t r a n s i e n t a d h e s i o n and rolling o f the n e u t r o p h i l on e n d o t h e l i a l cells. 5-7 This initial t r a n s i e n t a d h e s i o n is m e d i a t e d b y m e m b e r s o f the s e l e c t i n g e n e f a m i l y i n c l u d i n g E- and P - S e l e c t i n on e n d o t h e l i a l c e l l s and L - S e l e c t i n on n e u t r o p h i l s . 3'8'9 T h e i n i t i a l t r a n s i e n t a d h e s i o n o f n e u t r o p h i l s on end o t h e l i a l cells is i n s u f f i c i e n t for n e u t r o p h i l a c c u m u l a t i o n in tissue u n l e s s a d d i t i o n a l i n t e r a c t i o n s occur, t° A f t e r the initial a d h e s i o n in an i n f l a m m a t o r y focus, t h e r e m a y be a d d i t i o n a l s t i m u l i that attract and activate n e u t r o p h i l s . 1~-~3 T h e final steps in the p r o c e s s
D e s p i t e a d v a n c e s in clinical m e d i c i n e and b i o t e c h n o l ogy, the m a i n s t a y o f therapy for patients with respiratory failure is the administration o f s u p p l e m e n t a l o x y g e n . Unfortunately, p r o l o n g e d e x p o s u r e to supplemental o x y g e n m a y contribute to respiratory failure b y causing lung d a m a g e b y itself. Therefore, investigations o f the p a t h o g e n e s i s o f h y p e r o x i c lung injury continue, and recently there has been increasing interest in the potential contribution o f acute inflammation to the course o f h y p e r o x i c lung d i s e a s e J In s e v e r a l e x p e r i m e n t a l m o d e l s o f i n f l a m m a t i o n a Address correspondence to: Stephen E. Welty, M.D., One Baylor Plaza, Room 337D, Houston TX, 77030. 567
T. ZEB et al.
568
leading to neutrophil accumulation in a tissue are firm adhesion and transendothelial migration, and are often mediated by neutrophil f12 integrins with I C A M - 1 on endothelial cells. 14-~6 In previous studies we determined that the expression of ICAM-I in mouse lungs was increased in response to hyperoxia-exposure, messenger RNA being increased at 48 h and protein being increased at 72 h, and that the increased expression was localized to cells in the distal airspaces, which is also the area in which neutrophil accumulation o c c u r s . 17 Other studies using monoclonal antibodies indicate that ICAM-1 upregulation and interactions with neutrophil ]~2 integrins contribute significantly to hyperoxic lung inflammation and injury, ors In order to determine whether hyperoxic lung inflammation might also conform to the multiple adhesive interaction model for the development of inflammation, we examined hyperoxic mouse lungs for the expression of E- and P-Selectin mRNA. Finding increases in P-Selectin m R N A in mouse lungs well before the development of inflammation, we used in situ hybridization to determine the cell specific localization of increased P-Selectin expression. Because the upregulation of P-Selectin is one of the earliest changes we have identified in relationship to potential mechanisms of hyperoxic lung inflammation, it may be the step at which hyperoxic inflammatory responses may be controlled more effectively, in the context of other studies, suggesting that controlling the inflammatory response may be helpful. MATERIALS AND METHODS
Experimental protocol We studied male FVB mice that were approximately 6 weeks old, weighing 20 to 25 g. We placed the mice in a Plexiglas chamber that allowed them free access to food and water, and exposed them to greater than 95% oxygen continuously by administration of pure oxygen at 2 liters per minute while removing excess carbon dioxide using soda lime (Sodasorb, Dewey & Almy Chemical Division, Grace & Co., Lexington, MA.). We observed the mice and measured the oxygen concentration twice daily during hyperoxia exposure. In order to obtain mouse lungs for the analyses, we sacrificed our mice with 200 mg/kg of pentobarbital, IP. We sacrificed mice after 12, 24, 48, 72, or 96 h of hyperoxia exposure for RNA studies. We removed the lungs via a midline thoracotomy, placed the lungs in liquid nitrogen, and stored them at - 8 0 ° prior to isolation of total RNA. Air-breathing mice served as controis in all experiments and are designated as 0 h in > 9 5 % oxygen.
In studies for in situ hybridizations we cannulated the tracheas of mice with PE 50 tubing after administering pentobarbital. We then perfused the airways with zinc formalin (ANATECH, Battle Creek, MI) until the lungs were fully distended, and removed the heart and lung in block for subsequent paraffin embedding.
Isolation and analysis of RNA We isolated total lung RNA using a modification of the procedure of Chomcynzski and Sacchi j9 described by Hamburg and co-workers, z° After isolating RNA, we resuspended the RNA in 10 mM Tris pH 7.6, 1 mM EDTA, and quantified the amount of RNA by measuring the absorption at 260 nm. We loaded 20 #g of total RNA per sample per lane on a 1% agarose/formaldehyde denaturing gel, separated the RNA by electrophoresis and transferred the separated RNA from the denaturing gel to a nitrocellulose filter. We then allowed the nitrocellulose filter to air dry at 68 ° for 3 h.
RNA hybridization We used the partial complementary DNA for murine E-Selectin that was generously provided by C. M. Ballantyne, which had been cloned into pBluescript S K + (Stratagene, La Jolla, CA), and we generated a partial complementary DNA for murine P-Selectin using reverse transcriptase polymerase chain reaction (PCR) of total lung RNA isolated from the lung of a mouse exposed to 2 mg/kg of lipopolysacharide (Salmonella abortus equi, Sigma Chemical Comp., St. Louis, MO). In the reverse transcriptase PCR, we designed downstream and upstream primers based on the published sequence of the murine P-Selectin, and which also contained restriction sites on their 5' ends to facilitate ligation of the PCR product into pBluescript S K + (Strategene, La Jolla, CA). We made radioactive antisense transcripts to E- and P-Selectin m R N A with T7 RNA polymerase and nucleotide triphosphates with 32p-labeled uridine triphosphate (3000 Ci/mmol, Amersham Life Sciences, Arlington Heights, IL). After making the antisense transcripts, we prehybridized the nitrocellulose filters containing separated RNA in 50% formamide, 50 rnM Na2PO4, 0.8 M NaC1, 1.0 mM EDTA, 0.5× Denhardt's reagent, and 250 #g/ml of heat-denatured herring sperm DNA for approximately 3 h at 55 °. After prehybridization, we discarded the solutions and hybridized the nitrocellulose filters at 55 ° for 16 h using the same solution used for prehybridization and adding the radioactive antisense transcripts. We then washed the hybridized filters three times for 20 rain with 0.1 × SSC and 0.1% SDS at 68 ° to decrease nonspecific binding of radioactive transcripts to the filter. Following
P-selectin in oxygen toxicity these washes we exposed x-ray film to the hybridized filters to generate autoradiograms. In every Northern blot hybridization in which we analyzed R N A from hyperoxic lung tissue we included an analysis of R N A from lungs o f mice treated with 2 mg/kg o f IP lipopolysaccharide (Salmonella abortus equi, Sigma Chemical Company, St. Louis, MO). Because endotoxin is a potent inducer o f E- and P-selectin in mouse lungs, these R N A samples served as our positive controls.
569
=888888888888888888
28S
.
.
.
.
18S
28S In situ hybridization We made complementary R N A (cRNA) probe labeled with [33p]-UTP (DuPont Canada, Markham, Ont., Canada) as described previously. 2~ After making the c R N A probe, we digested the D N A template with R N A s e free D N A s e I (Promega Corp., Madison, WI), and extracted the resulting solution containing c R N A probe with an equal volume of phenol/chloroform (1:1), followed by ethanol precipitation and resuspension in diethylpyrocarbonate-treated water. Before hybridization, we performed limited alkaline hydrolysis o f the c R N A probe to reduce transcript length to between 100 and 300 bp. 22 We determined the size o f the
888881888888888888._
- '
18S
Fig, 2. Autoradiogram from a Northern blot hybridization for PSelectin messenger RNA in hyperoxia-exposed mouse lungs. Mice were exposed to 0, 24, 48, 72, and 96 h of >95% oxygen continuously, or 2 mg/kg of endotoxin IP and lungs obtained (2 h after endotoxin dosing) for isolation of total RNA. Twenty micrograms of RNA was separated electrophoretically, transferred to a nitrocellulose membrane, and hybridized with -~2P-labeled antisense P-Selectin transcripts. Signal for P-Selectin is clearly evident just below the 28S ribosomal RNA signal in the lung RNA from mice exposed to endotoxin, and there is detectable signal in lung RNA from mice exposed to 48 through 96 b of hyperoxia. The ethidium bromidestained filter in the lower part of the figure coincides to this autoradiogram and documents equivalent quantity and quality of RNA.
~
28S
-,,~--18S
28S 18S Fig. 1. Autoradiogram from a Northern blot hybridization for ESelectin messenger RNA in hyperoxia-exposed mouse lungs. Mice were exposed to 0, 12, 24, 48, and 72 h of > 9 5 % oxygen continuously, or 2 mg/kg of endotoxin IP and lungs obtained (2 h after endotoxin dosing) for isolation of total RNA. Twenty micrograms of RNA was separated electrophoretically, transferred to a nitrocellulose membrane, and hybridized with 32P-labeled antisense E-Selectin transcripts. Signal for E-Selectin is clearly evident in the lung RNA from mice exposed to endotoxin just below the 28S ribosomal RNA signal, but there is no detectable signal in lung RNA from mice exposed to up to 72 h of hyperoxia. The ethidium bromidestained filter in the lower part of the figure coincides with the autoradiogram and documents equivalent quantity and quality of RNA.
partially hydrolyzed transcripts by agarose gel electrophoresis. For hybridization, we used methods described previously. 2L23 W e carried out hybridization on lung sections overnight at 54 ° in 50% formamide, 0.3 M NaC1, 10 m M Tris-C1 (pH 8.0), 1 m M E D T A , 1× Denhardt's solution, 10% dextran sulfate, 0.5 mg/ml yeast transfer RNA, and 17 ng/ml o f c R N A probe in a moist chamber containing 0.3 M NaC1, 50% formamide. After hybridization, we treated the sections with RNAse, 2~ followed by a stringent wash at 68 ° in 0.1× SSC for 30 min. W e exposed the sections to N B T - 2 photographic emulsion (Kodak Company, Rochester, NY) for 10 d for autoradiography and then counterstained the sections with hematoxylin and eosin before photomicrography. In all experiments we evaluated nonspecific binding using sense c R N A probe.
RESULTS
General appearance No mice died in hyperoxia exposure before the times chosen to obtain the lungs for analysis. Furthermore, the mice had no visual evidence of illness through 72 h o f hyperoxia exposure. At 96 h of hyperoxia exposure
/
~t:,:k
0 Fig. 3. In situ hybridizations for P-Selectin mRNA in hyperoxia-exposed mouse lungs. Mice were exposed to >95% oxygen for up to 96 h, the lungs inflated with and fixed in zinc formalin, and embedded in paraffin. Sections were cut and hybridized with antisense P-Selectin labeled with 3sp. The sections were then developed by autoradiography. In the sections from A to F there are two figures of identical sections, one examined with standard light microscopy, and one examined utilizing dark-field microscopy. The purpose of the dark-field examination is to illustrate a signal for P-Selectin mRNA, which is identified by bright staining (white arrows). The purpose of the light microscopy examination is to identify the regions in the lung where the PSelectin mRNA signal is observed in the corresponding dark-field section (at low power, signal for P-Selectin mRNA is not detectable in H&E sections). The section labeled 3A is a representative lung section (10×) from a mouse exposed to room air only. There is no detectable P-Selectin mRNA observed by dark-field examination. The section labeled 3B is a representative lung section (10×) from a mouse exposed to 48 h of byperoxia. In the section P-Selectin mRNA is clearly evident in endothelial cells in large muscularized vessels (white arrows on dark field) that are identifiable as arteries by light microscopy because of the vessel's proximity to the large airway (black arrow). Note also that P-Selectin is not upregulated throughout the artery, suggesting that the upregulation is focal even within the artery examined. The section labeled 3C is a higher power view of the same section to further illustrate the endothelial cell localization of the signal (white arrow). The figure labeled 3D is a representative section (20× for dark-field view and 10× for light microscopy) from a mouse exposed to 72 h of hyperoxia, and circumferential signal is observable in the medium sized vessel (white arrow on dark-field view), which is identifiable as an artery by its muscular wall (arrow on light microscopy). The sections labeled 3E (10×) and 3F (20×) are from a mouse exposed to 96 h of hyperoxia, In addition to the findings observed in mice exposed to shorter durations of hyperoxia, these lungs also have P-Selectin evident in even smaller lung vessels (white arrows on dark-field, black arrows on light microscopy). The smaller vessels are identifiable as arterial by the fact that they are relatively small vessels in an intralobular locale, whereas veins of this size would be interlobular in locale. While P-Selectin expression is detectable in these small vessels there is no detectable signal in alveolar capillaries.
P-selectin in oxygen toxicity
571
Fig. 3. Continued.
most mice had evidence of illness exhibited by an apparent increase in their work of breathing. Northern blot hybridizations and in situ hybridizations An autoradiogram of a Northern blot hybridization for E-Selectin is depicted in Fig. 1. E-Selectin m R N A signal is clearly evident visually just below the 28S
signal in the lung RNA from animals exposed to endotoxin, whereas there is no detectable E-Selectin in the lungs of mice exposed to hyperoxia for up to 72 h. The ethidium bromide-stained membrane at the bottom of Fig. 1 documents even loading. An autoradiogram of a Northern blot hybridization for P-Selectin is depicted in Fig. 2. In both lung RNA samples from endotoxin-treated mice there is signal for P-Selectin m R N A clearly evident. In addition, lung RNA samples
572
T. ZEB et
from mice exposed to 48 h or more of hyperoxia have signals for P-Selectin evident. To study the cell-specific regulation of P-Selectin mRNA, in situ hybridizations were performed. In lungs from air-breathing mice (Fig. 3A), P-Selectin transcripts were not detected. After 48 h of hyperoxia exposure, P-Selectin mRNA was detected in endothelial cells of the lungs. However, the intensity of the expression as well as the type of vessel where the transcripts were observed changed with the length of exposure in hyperoxia. After 48 h, P-Selectin transcripts were observed in large muscularized vessels (Fig. 3B), which are identifiable as arteries because of the proximity to the airways and the thickness of the muscular walls. After 72 h of exposure, P-Selectin transcripts were detected in medium size vessels, which are identifiable as arteries by their proximity again to airways (Fig. 3C), and after 96 h of hyperoxia exposure, abundant expression of P-Selectin transcripts was observed (3D) along with expression in nonmuscularized smaller vessels (Fig. 3E). P-Selectin transcripts were not detected in capillaries or in sections hybridized with radioactive sense transcripts, although we cannot rule out the possibility that P-Selectin was upregulated on the capillary interstitium, but was below the limits of detection. DISCUSSION Recent evidence suggests that lung inflammation is an important contributor to the development of hyperoxic lung injury, and that specific therapy directed against neutrophil adhesion may be useful. To date, the evidence targeting neutrophil adhesion in hyperoxic lung injury has been directed against neutrophil/32 integrins and one of their ligands, ICAM-I. The interactions of f12 integrins with ICAM- 1 mediates firm adhesion, ~4-16 and these interactions are dependent on prior transient adhesion of neutrophils endothelial cells. Lawrence and co-workers in studies in vitro in which neutrophil adhesion was assessed under conditions of flow, found that ICAM-1 -mediated neutrophil adhesion could not take place until P-Selectin-mediated neutrophil adhesion occurred. From the studies by Lawrence et al. and other investigators, a concept has emerged that multiple adhesive interactions between neutropbils and endothelial cells are required before neutrophil accumulation and neutrophil-dependent cytotoxicity c a n o c c u r . 3'24 If neutrophil accumulation in the hyperoxia-exposed lungs is also dependent on sequential neutrophil interactions with adhesion molecules, then the studies supporting a role for/32 integrins and ICAM-1 are addressing a more limited picture of
al.
mechanisms leading to neutrophil accumulation in hyperoxic lungs. Our finding of increased P-Selectin mRNA early in the course of hyperoxia-exposed lungs suggests a role for P-Selectin in hyperoxia-induced lung inflammation, similar to that indicated by Lawrence et al. in their model of inflammation. The lack of E-Selectin upregulation by northern blotting suggests that E-Selectin is not involved in transient adhesion in hyperoxia-exposed lungs. The lack of demonstrable upregulation in E-Selecting is in contrast to our findings in hyperoxia-exposed rats, in which ESelectin is upregulated early in the hyperoxia-exposure. 25 The most likely explanation for differences in rat and mouse E-Selectin expression in hyperoxia is that the regulation of E-Selectin gene is dependent on the species studied. Of particular interest is the localization of the PSelectin during hyperoxia exposure. The fact that PSelectin upregulation occurs in the large, medium, and small arterial vessels of the lung with no detectable mRNA expression in capillaries suggests that P-Selectin gene transcription in endothelial cells of the larger vessels is more sensitive to increases in oxidative stress, although we cannot absolutely rule out that capillary expression is present but below the level of detection using in situ hybridization. An apparent inconsistency with the hypothesis that P-Selectin gene transcription is sensitive to oxidative stress is the finding that the earliest upregulation of P-Selectin occurs in pulmonary artery endothelium (prior to exposure to the high alveolar PO2) rather than in the pulmonary capillaries or veins (directly exposed to high PO2). In the in situ hybridizations the arteries are distinguished from the veins by determining whether the vessel involved is in proximity to airways, because of the fact that arteries branch with the airways into the lung and the veins return blood from the capillaries between lung lobes. Surprisingly, using these criteria for the identification of pulmonary arteries and veins, the upregulation is most marked on pulmonary arteries (Fig. 3B, note the large airway adjacent to the vessel with P-Selectin mRNA evident in the endothelium). The next issue is the functional significance of an increase in P-Selectin mRNA in lung endothelial cells. P-Selectin protein has been shown be present in platelets and endothelial cells in preformed granules that may be translocated rapidly to the cell surface in response to endogenous stimuli, such as histamine and thrombin) 6'27 In addition, some investigators have observed increases in P-Selectin mRNA in tissues of animals exposed to lipopolysaccharide 2. and in endotheliomas exposed to tumor necrosis factor-o<29 In the study of endotheliomas the increase in P-Selectin mRNA was associated with increased P-Selectin pro-
P-selectin in oxygen toxicity
tein, and the investigators speculated that P-Selectin protein was upregulated by increased transcription. It is likely that the observed increase in P-Selectin mRNA leads to an increase in P-Selectin protein, though we have not tested this hypothesis. The role of the upregulation of P-Selectin in hyperoxic lung inflammation is still undetermined. There are basically two different views in regard to the role that P-Selectin might play in arterial side of the pulmonary circuit. The first view is simply that upregulation of P-Selectin in the arterial side of the pulmonary circuit plays no role in hyperoxic lung inflammation. The second view is that the upregulation does play a role in hyperoxic lung inflammation by mediating transient adhesion and rolling of neutrophils in the pulmonary circulation. P-Selectin upregulation on the arterial side of the pulmonary circulation would place P-Selectin in a favorable position to mediate transient adhesion and rolling prior to the pulmonary capillaries, at which point firm adhesion could occur via different mechanisms at the level of the alveolar capillary, which is the site in which neutrophils have been observed to enter lung tissue. 3° If this second view is correct, it would be in contrast to inflammatory responses in the systemic circulation in which neutrophil adhesion and transendothelial migration are well established as occurring in the venules. Support for this contrast in localization for neutrophil adhesion and transendothelial migration in the lung is supported by studies in which delays were observed in lung neutrophil transit proximal to pulmonary veins. 31 The altered cell specific expression of P-Selectin as a function of hyperoxia exposure time is interesting demands further study. Studies of the 5' flanking sequences of P-Selectin to determine sequences crucial to the cell specific expression is likely to be fruitful in determining the mechanisms for the alterations in cellular expression as a function of hyperoxia exposure. The P-Selectin promoter studies could be carried out in in vitro transfection studies in which cells are transfected with serial deletions of the 5' flanking region of P-selectin ligated to a reporter gene, after which the cells would be exposed to oxidants, or in in vivo studies in which mice made transgenic with a similar series of promoter-reporter constructs are exposed to hyperoxia. Determining the role of individual adhesion molecules in hyperoxic lung inflammation and injury is an extremely complex issue. To illustrate the potential confusion, studies in which anti-ICAM-1 and antiCD 1 1a antibodies were administered to hyperoxia-exposed mice the treatment was protective, while in studies from our own lab knockout mice deficient in ICAM-1 and with reduced CD18 expression are not protected (unpublished results). There are a number of potential reasons for this apparent contradiction in ex-
573
perimental results, such as that temporal and spatial interactions of adhesion molecules with their ligands is more complex than anticipated. For example, in hyperoxia-exposed animals there may be specific durations of exposure or specific regions wherein adhesion molecule interactions may mediate "appropriate" inflammatory responses, while at other durations or in other regions adhesion molecule interactions may mediate "inappropriate" inflammatory responses. The complexity of the inflammatory response and the redundancy of the adhesion molecule system make the issue of determining the role of inflammation in hyperoxic lung injury problematic. However, molecular approaches in which adhesion molecule expression can be tightly regulated and more than one adhesion molecule affected will be helpful in better understanding the role of inflammation in hyperoxic lung injury and in designing more specific interdictions when appropriate. Furthermore, expression studies such as the present study are essential as a foundation on which to determine which molecules are likely to be involved in an inflammatory response. For example, the present studies suggest that exploring E-Selectin in hyperoxic mice in more detailed studies is unwarranted, while studying the role of P-Selectin expression by itself and in concert with other adhesion molecules is appropriate. In summary, P-Selectin mRNA is upregulated early in the course of hyperoxic lung injury and is localized primarily to the arterial side of the pulmonary circuit. These findings support the idea that P-Selectin participates in hyperoxic inflammation and that future studies inhibiting P-Selectin function or upregulation are necessary to determine the extent of the role of P-Selectin in hyperoxic lung injury and inflammation. Our finding that E-Selectin mRNA is not upregulated supports the idea that E-Selectin does not play a role in hyperoxic lung inflammation. At this time we cannot rule out the possibility that E-Selectin is upregulated in hyperoxic lungs, but that Northern blot hybridization is simply not sensitive enough to detect the upregulation. Acknowledgements- This work was supported by NIH Grant AI19031, NIH Grant 5 P30 HD27823-04, and an A H A grant, Texas Affiliate.
REFERENCES 1. Wegner, C. D.; Wolyniec, W. W.; LaPlante, A. M.; Marschman, K.; Lubbe, K.; Haynes, N.; Rothlein, R.; Letts, G. L. Intercellular adhesion molecule-1 contributes to pulmonary oxygen toxicity in mice: Role of leukocytes revisited. Lung 170:267-279; 1992. 2. Lawrence, M. B.; Smith, C. W.; Eskin, S. G.; Mclntire, L. V. Effect of venous shear stress on CD18-mediated neutrophil adhesion to cultured endothelium. Blood 75:227-237; 1990. 3. Lawrence, M. B.; Springer, T. A. Leukocytes roll on a selectin
574
4. 5, 6. 7. 8.
9.
10.
11.
12.
13. 14.
15.
16.
17.
18.
T. ZEB et al. at physiologic flow rates: Distinction from and prerequisite for adhesion through integrins. Cell 65:859-873; 1991. Butcher, E. C. Leukocyte-endothelial cell recognition: Three (or more) steps to specificity and diversity. Cell 67:1933-1936; 1991. Anderson, B. O.; Brown, J. M.; Shanley, P. F.; Bensard, D. D.; Harken, A. H. Marginatiing neutrophils are reversibly adherent to normal lung endothelium. Surgery 109:51-61; 1991. Springer, T. A.; Lasky, L. A. Sticky sugars for selectins. Nature 349:196-197; 1991. Patarroyo, M.; Makgoba, M. W. Leukocyte adhesion to cells in immune and inflammatory responses. Lancet 2:1139-1142; 1989. Ley, K.; Gaehtgens, P.; Fennie, C.; Singer, M. S.; Lasky, L. A.; Rosen, S. D. Lectin-like cell adhesion molecule 1 mediates leukocyte rolling in mesenteric venules in vivo. Blood 77:25532555; 1991. Hammer, D. A.; Apte, S. M. Stimulation of cell rolling and adhesion on surfaces in shear flow: General results and analysis of selectin-mediated neutrophil adhesion. Biophysiol. J. 63:35-57; 1992. Arfors, K. E.; Lundberg, C.; Lindbom, L.; Lundberg, K.; Beatty, P. G.; Harlan, J. M. A monoclonal antibody to the membrane glycoprotein complex CD18 inhibits polymorphonuclear leukocyte accumulation and plasma leakage in vivo. Blood 69:338340; 1987. Carden, D. L.; Smith, J. K.; Korthuis, R. J. Oxidant-mediated, CD 18-dependent microvascular dysfunction induced by complement-activated granulocytes. Am. J. Physiol. 260:H1144-H1152; 1991. Mulligan, M. S.; Varani, J.; Warren, J. S.; Till, G. O.; Smith, C. W.; Anderson, D. C.; Todd, R. F.; Ward, P. A. Roles of /32 integrins of rat neutrophils in complement and oxygen radical-mediated acute inflammatory injury. J. lmmunol. 148:1847-1857; 1992. Westwick, J.; Li, S. W.; Camp, R. D. Novel neutrophil-stimulating peptides, lmmunol. Today 10:146-147; 1989. Smith, C. W.; Marlin, S. D.; Rothlein, R.; Toman, C.; Anderson, D. C. Cooperative interactions of LFA- 1 and Mac- 1 with intercellular adhesion molecule-I in facilitating adherence and transendothelial migration of human neutrophils in vitro. J. Clin. Invest. 83:2008-2017; 1989. Smith, C. W.; Rothlein, R.; Hughes, B. J.; Mariscalco, M. M.; Rudloff, H. E.; Schmalsteig, F. C.; Anderson, D. C. Recognition of an endothelial determinant for CD18-dependent human neutrophil adherence and transendothelial migration. J. Clin. Invest. 82:1746-1756; 1988. Furie, M. B.; Tancinco, M. C. A.; Smith, C. W. Monoclonal antibodies to leukocyte integrins CDIIa/CD18 and CDllb/ CD 18 or intercellular adhesion molecule- 1 inhibit chemoattracrant-stimulated neutrophil transendothelial migration in vitro. Blood 78:2089-2097; 1991. Welty, S. E.; Rivera, J. L.; Geske, R. S.; Montgomery, C. A.; Hansen, T. N. Increases in intercellular adhesion molecule-1 (ICAM-1) in hyperoxic mouse lungs are localized to cells in the alveolar interstitium. Pediat. Res. 32:160A; 1993. Ramsay, P. L.; Welty, S. E.; Bree, M. P.; Hansen, T. N. Lung
19.
20. 21.
22.
23.
24. 25. 26.
27. 28. 29.
30. 31.
injury and inflammation are reduced in hyperoxic mice by inhibition of neutrophil /32 integrin CDlla/CDI8. Pediatr. Res. 35:396A; 1994. Chomczynski, P.; Sacchi, N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem. 162:156-159; 1987. Hamburg, D. C.; Tonoki, H.; Welty, S. E.; Montgomery, C. A.; Hansen, T. N. Endotoxin induces glutathione reductase activity in lungs of mice. Pediatr. Res. 35:311-315; 1994. Piedboeuf, B.; Johnston, C. J.; Watkins, R. H.; Hudak, B. B.; Lazo, J. S.; Cherian, M. G.; Horowitz, S. Increased expression of tissue inhibitor of metalloproteinases (TIMP-1) and metallothionein in murine lungs after hyperoxic exposure. Am. J. Respir. Cell Mol. Biol. 10:123-132; 1994. Cox, K. H.; DeLeon, D. V.; Angerer, L. M.; Angerer, R. C. Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes. Dev. Biol. 101:485-502; 1984. Veness-Meehan, K. A.; Cheng, E. R.; Mercier, C. E.; Blizt, S. L.; Johnston, C. J.; Watkins, R. H.; Horowitz, S. Cellspecific alterations in expression of hyperoxia-induced mRNAs of lung. Am. J. Respir. Cell Mol. Biol. 5:516-521; 1991. Albelda, S. M.; Smith, C. W.; Ward, P. A. Adhesion molecules and inflammatory injury. FASEB J. 8:504--512; 1994. Ramsay, P. L.; Welty, S. E. E-selectin increases in hyperoxic lungs and is associated with inflammation and injury. Am. J. Respir. Crit. Care Med. 151:A343; 1995. Johnston, G. I.; Kurosky, A.; McEver, R. P. Structural and biosynthetic studies of the granule membrane protein GMP-140, from human platelets and endothelial cells. J. Biol. Chem. 264:1816-1823; 1989. Celi, A.; Furie, B.; Furie, B. C. PADGEM: An adhesion receptor for leukocytes on stimulated platelets and endothelial cells. Proc. Soc. Exp. Biol. Med. 198:703-709; 1991. Auchampach, J. A.; Oliver, M. G.; Anderson, D. C.; Manning, A. M. Cloning and sequence comparison and in vivo expression of the gene encoding rat P-Selectin. Gene 145:251-255; 1994. Weller, A.; Isenmann, S.; Vestweber, D. Cloning of the mouse endothelial selectins. Expression of both E- and P-selectin is inducible by tumor necrosis factor alpha. J. Biol. Chem. 267:15176-15183; 1992. Worthen, G. S.; Avdi, N.; Vukajiovich, S.; Tobias, P. S. Neutrophil adherence induced by lipopolysaccharide in vitro../. Clin. Invest. 90:2526-2535; 1992. Downey, G. P.; Doherty, D. E.; Schwab, B.; Elson, E. L.; Henson, P. M.; Worthen, G. S. Retention of leukocytes in capillaries: Role of cell size and deformability. J. Appl. Physiol. 69:17671778; 1990.
ABBREVIATIONS
I C A M - 1- - i n t e r c e l l u l a r adhesion molecule- 1 IP--intraperitoneal cRNA--
complementary RNA