American Journal of Pathology, Vol. 138, No. 3, March 1991 Copyright ©D Amencan Association of Pathologists
The Effects of Monocrotaline Pyrrole on Cultured Bovine Pulmonary Artery Endothelial and Smooth Muscle Cells J. F. Reindel* and R. A. Rotht From the Departments of Pathology* and of Pharmacology and Toxicology,t Michigan State University, East Lansing, Michigan
Monocrotaline pyrrole (MCTP), a reactive electrophile, induces delayed and progressive pulmonary edema; vascular remodeling, and pulmonary hypertension after a single intravenous administration to rats. The effects of a single exposure of cultured bovine pulmonary artery endothelial cells (BEC) and bovine pulmonary artery smooth muscle cells (BSMC) to MCTP were examined Monocrotaline pyrrole caused a dose-dependent, delayed4 and progressive cell detachment and release of lactate debydrogenase activity from monolayers of BECs but not BSMCs. Monolayers of BECs also released increased concentrations of 6-keto-prostaglandin F,,, the stable metabolite ofprostacyclin as the post-treatment interval increased Progressive and marked endothelial cell hypertrophy occurred after exposure to a nominal concentration of5 or50 Ag/ml ofMCIP but not after 0.5 Lg/ml. Morphologic changes in monolayers of BSMCs were minimal, even up to 2 weeks after exposure. Ultrastructurally the hypertrophic, MCTP-treated BECs had enlarged cell profiles with enlarged nuclei. The nucleoli were prominent, occasionally multiple, and had separation of granular and fibrillar components. Cytoplasmic microtubules and perinuclear intermediate filaments were prominent in some cells as were the golgi apparatus and
endoplasmic reticulumr Degenerative changes were not prominent in cells that remained in the monolayer. Monocrotaline pyrrole inhibited proliferation of both cell types at concentrations (0.5 pLg/ml) that were not cytotoxic. These findings indicate that MCTP induces direct, dose-dependent injury to cells in culture that is delayed and progressive, and the expression of this injury depends in part, on the cell type. (Am JPathol 1990, 138:707-719)
Monocrotaline (MCT) is a pyrrolizidine alkaloid (PA) found in the foliage of Crotalaria spectabilis.' Ingestion of MCT has been associated with hepatic and pulmonary disease in humans and many animal species. In rats, a single administration of MCT results in pneumotoxic changes that are delayed in onset and progressive.23 These include pulmonary vascular leak, pulmonary arterial vascular remodeling, and pulmonary hypertension. Pulmonary damage is not caused directly by MCT but by its metabolic products. Monocrotaline and other pneumotoxic PAs are converted to reactive pyrroles by the mixed function oxidase system of the liver, but not of the lungs.4 6 It is presumed that small quantities of these pyrrolic metabolites escape binding in the liver and travel via the blood stream to the pulmonary vascular bed, where they bind covalently to cellular macromolecules.7 The responses of pulmonary vascular cells to this binding are largely unknown, and how they relate to the delayed and progressive pneumotoxicity and vascular remodeling is not well understood. For example, it is unknown whether pyrrolic derivatives of MCT produce their effects by direct, cytotoxic interactions with cells of the lung or whether cell injury is indirect, requiring secondary factors such as cellular release of inflammatory mediators or activation of the immune system. Monocrotaline pyrrole (MCTP) is an unstable, electrophilic, putative metabolite of MCT.5 When a single injection of a low dose of chemically synthesized MCTP is injected into tail veins of rats, delayed and progressive pulmonary injury, pulmonary hypertension, and right heart hypertrophy results that is similar to that induced by MCT itself.8 Associated with these changes is remodeling of the pulmonary arteries and microvasculature.>12 Circulating blood cells and pulmonary vascular endothelial cells are presumably among the cells first exposed Supported by NIH grant ES02581. Dr. Reindel was supported by NIH Training Grant ES07146 and the Center for Environmental Toxicology, Michigan State University. Accepted for publication November 2, 1990. Address reprint requests to Dr. Robert A. Roth, Department of Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824.
707
708
Reindel and Roth
AJP March 1991, Vol. 138, No. 3
to MCTP after intravenous injection. Indeed endothelial cells of the pulmonary arterial vasculature and capillary bed probably are exposed to higher concentrations of the short-lived pyrrole than are other lung cells. In morphologic studies of MCT- or MCTP-induced lung disease, structural changes in endothelial cells occur as the pulmonary disease progresses.12-16 These changes are relatively subtle early in the course of the disease but become more apparent with time. Smooth muscle cells of blood vessels and cells of the alveolar septal interstitium might be exposed to lower concentrations that escape binding in the vascular compartment and survive passage into the interstitial fluid. These cells also undergo delayed and progressive alterations in vivo. The response of endothelial cells and vascular smooth muscle cells in vitro may aid in understanding the delayed and progressive pathologic changes that occur in vivo in the pulmonary parenchyma and microvasculature after exposure to MCTP. For example, such responses may help define which of the effects observed in vivo are direct consequences of MCTP-cell interactions and which ones may be responses to endogenously synthesized mediators, the release of which might be triggered by MCTP or by the pulmonary hypertensive response. Accordingly we tested the hypothesis that cells of the pulmonary vasculature are directly injured after a single exposure to MCTP and that this injury is delayed and progressive. To this end, we examined the cytotoxic and proliferative responses to MCTP of bovine pulmonary artery endothelial cells (BEC) and bovine pulmonary artery smooth muscle cells (BSMC) grown in culture. Our finding of delayed and progressive alterations in these cells is reminiscent of the response to MCTP in vivo.
Materials and Methods Preparation of Endothelial Cells Bovine endothelial cells were isolated from segments of pulmonary artery by modifications of the methods of Jaffe et al17 and Booyse et al.18 Segments of the mainstem pulmonary artery from young calves were rinsed with Hank's balanced salt solution (HBSS) containing 1% penicillin (100 units/ml), streptomycin (100 ,ug/ml), and fungizone (0.25 ,ug/ml) (PSF; Gibco, Grand Island, NY) and the adventitia and external elastic laminae were removed. The 5- to 6-cm segments of vessels were opened and placed luminal surface down into 60-mm petri dishes containing 3 to 5 ml of collagenase solution (type IV, 0.1 %) (Sigma Chemical Co., St. Louis, MO). The abluminal surface of the vessels was exposed briefly to ultraviolet light to kill remaining adventitial fibroblasts and the surface layer of smooth muscle cells. The tissue was incubated for 3 to 5 minutes at 370C and sheets of endothelial cells were brushed gently from the luminal surface
with a rubber policeman. These brushings were dispersed into 100-mm petri dishes containing HBSS. Individual small sheets or clusters of the endothelial cells were transferred to wells of 12-well tissue culture cluster plates (Costar, Cambridge, MA) containing culture medium (see below). Cells were allowed to attach to the well surface for 3 to 4 hours. Wells then were rinsed vigorously with HBSS to remove unattached cells and fresh Dulbecco's Modified Eagles Medium (DMEM, Gibco) was added into the wells. Plates were incubated until individual colonies developed (4 to 7 days), and selected wells free of fibroblasts and smooth muscle cells (spindle cells) were identified for expansion of cell lines. Endothelial cells were identified by the characteristic cobblestone morphology of monolayers, ultrastructural features, positive staining for factor VIII antigen, and negative staining for desmin antigen.
Preparation of Smooth Muscle Cells Bovine smooth muscle cells were isolated from small sections of pulmonary artery from which the external elastic laminae was removed. The tissue was rinsed vigorously and the luminal and adventitial surfaces of the vessel were exposed to ultraviolet light to kill surface smooth muscle and endothelial cells. The vessel intima and internal elastic lamina were removed from the vessel, and fragments of the wall, measuring approximately 1 mm,3 were excised from the muscular layer and placed in tissue culture wells containing culture medium. Within 4 to 7 days, outgrowths of bipolar spindle cells spread from the tissue sections on the plastic surface of the plate. Tissue fragments then were removed from the plate and the attached cells continued to proliferate to form monolayers. Bovine pulmonary smooth muscle cells isolated in this manner stained positively for desmin antigen, negatively for factor VIII antigen, and were spindle shaped in appearance. These cells grew in a hill-and-valley, whorling pattern. They did not form a uniform monolayer but grew in overlapping layers. Cell culture medium consisted of DMEM containing 10% fetal bovine serum (Gibco), 2 mmol/l (millimolar) glutamine and 1% PSF solution (Gibco). Both BECs and BSMCs were passed using trypsin-ethylenediamine tetraacetic acid (EDTA) solution (0.025% trypsin, 0.27 mmolA EDTA in Ca' '-free HBSS) and were used at passages less than 10.
Preparation of MCTP Monocrotaline pyrrole was prepared from MCT (Transworld Chemicals, Washington, DC) via an N-oxide intermediate by the method of Mattocks.19 Monocrotaline pyrrole isolated by this synthesis procedure has Ehrlich
Monocrotaline Pyrrole-induced Cell Injury
709
AJP March 1991, Vol 138, No. 3
activity6 and a structure consistent with MCTP as determined by mass spectrometry and nuclear magnetic resonance.20 Monocrotaline pyrrole was maintained in N,N-dimethylformamide (DMF) vehicle as a stock solution of 20 mg/ml under nitrogen at - 20°C. All dilutions of MCTP were with DMF solvent immediately before use. A 2.5-,ul volume of MCTP solutions or DMF (control) per milliliter of medium was used in all studies to achieve the nominal concentrations of MCTP (0.005 to 50 ,ug/ml) used in the study.
Effects of MCTP on Cell and Monolayer Morphology Bovine endothelial cells and BSMCs were plated into 12or 24-well tissue culture plates and allowed to proliferate to confluency. Five microliters of MCTP solution was added to test tubes containing 1 ml of media, the tubes were rapidly shaken, and the mixture was poured immediately into wells already containing 1 ml of culture medium. Monolayer morphology was examined at daily intervals using phase-contrast microscopy. The medium was replaced with fresh medium containing no MCTP 1 day after treatment and three times per week thereafter for morphologic studies. At selected intervals, monolayers in replicate plates were fixed with 2% glutaraldehyde in 0.1 mol/l (molar) phosphate-buffered saline (pH 7.4) or 10% neutral-buffered formalin for morphometric studies. Cells were stained with hematoxylin and eosin or with crystal violet (0.5%) stains. Morphometric assessment of cell size (two-dimensional surface area) was determined with the aid of a Joyce Loebl image analysis system attached to an inverted Nikon TMS microscope (Nikon Instruments Division, Garden City, NY). The two-dimensional cell-surface area of approximately 200 cells per dose were measured 1, 5, and 15 days after exposure to MCTP.
Transmission Electron Microscopy Confluent monolayers of BECs were grown on Thermanox tissue culture coverslips (Lux, Miles Laboratories, Elkhart, IN) and exposed to 0, 5, or 50 ,ug/ml MCTP. Five to seven days after exposure, monolayers were washed twice with HBSS and fixed for at least 4 hours with 2% glutaraldehyde in 0.1 moVI phosphate buffer (pH 7.4). Monolayers subsequently were washed twice with 0.1 mol/l phosphate buffer. Fixed cells were scraped from coverslips and embedded in Epon Araldite resin (Electron Microscopy Services, Fort Washington, PA). Ultrathin gold sections were cut from blocks with an LKB ultramicrotome (Cambridge Instruments, Deerfield, IL), stained with uranyl acetate and lead citrate and viewed with a Phillips 301 transmission electron microscope (Phillips Electronic Instruments, Mahwah, NJ).
Cell Proliferation Assays Five hundred BECs or BSMCs were plated onto 1 00-mm tissue culture dishes containing 10 ml of medium. Triplicate or quadruplicate plates were used for each treatment group. Cells were allowed to attach for 4 to 6 hours before exposure to a single administration of the test compound and were given fresh medium 7 days after treatment. At 14 days, cells were fixed with 10% neutralbuffered formalin (pH 7.4) and stained with crystal violet (0.5%). Colonies of cells containing approximately 50 or more cells per colony were enumerated for each plate and results were expressed as a percentage of control.
Cellular Release of Lactate Dehydrogenase Cytotoxicity for BECs and BSMCs was assessed as the appearance of lactate dehydrogenase (LDH) activity in the medium above monolayers of cells exposed to a single administration of MCTP. Lactate dehydrogenase activity was determined by the method of Bergmeyer and Bernt21 using pyruvate solution (Sigma Chemical) and reduced nicotinamide adenine dinucleotide (NADH; Sigma Chemical). All monolayers were treated on day 0 and replicate plates were used for analysis of LDH activity daily for 5 days after MCTP treatment. Culture medium from triplicate wells for each treatment and posttreatment interval was analyzed for LDH activity. After removal of medium, monolayers in these wells were rinsed with HBSS and the cells were lysed with 15 ,ul of 10% Triton x 100 (Sigma Chemical) in 2 ml of fresh medium to determine cell-associated LDH activity. Percentage release of LDH for each well was defined by the following formula: % release = LDH in medium above monolayer LDH from lysed cells + LDH in medium above monolayer
10
The mean of triplicate wells for each study was considered a replication for statistical analysis (n = 5 replications for BECs; n = 4 replications for BSMCs).
Determination of 6-Ketoprostaglandin
Fla
Bovine endothelial cell monolayers in replicate 24-well tissue culture plates were exposed on day 0 to a single administration of 50, 25, 5, or 0.5 ,ug/ml MCTP or to vehicle in a total volume of 2 ml of medium. At daily intervals, medium from above monolayers of one of the replicate plates was collected in a polystyrene tube containing 15 ,ug of indomethacin. Tubes were centrifuged at -20°C
710
Reindel and Roth
AJP March 1991, Vol. 138, No. 3
and supernatant fluids were collected and stored at -200C until analyzed. Unextracted samples of medium were analyzed for 6-ketoprostaglandin Fla (6-keto-PGF1,), the stable breakdown product of prostacyclin (PGI2), by radioimmunoassay. Standard curves were constructed using DMEM for dilution. Specific antibodies and antigen were purchased from Seragen (Boston, MA). The crossreactivity at 50% maximal binding of antigen (B/Bo) for 6-keto-PGF1,, antibody was 7.8% for PGFia; 6.8% for 6keto-PGE1; 2.2% for PGF2a; 0.7% for PGE1; 0.6% for PGE2; and less than 0.1% for PGD2, PGA2, PGA1, PGB1, PGB2, TxB2, 15-keto-PGF2, 15-keto-PGF2a, DHKE2, or DHKF2a (Seragen). Quadruplicate wells were used at each dose on each day and 6-keto-PGFia accumulation in the medium above each of the exposed monolayers was determined for 4 or 5 days after a single exposure to MCTP.
Enumeration of Attached Cells After removal of medium for 6-keto-PGFia determination, monolayers were washed twice with 1 ml of HBSS to remove nonadherent cells. Adherent cells were detached by treatment with 0.025% trypsin, 0.27 mmol/l EDTA solution (Gibco) in Ca' '-free HBSS and were enumerated from quadruplicate wells using a hemacytometer.
Analysis of Data Results were expressed as mean ± standard error. Percentage data was transformed by arcsin-1 transformation. Data were analyzed by a completely random analysis of variance. Nonhomogeneous data were log transformed to attain homogeneity before analysis. Tukey's w-test was used for individual comparisons, except morphometric data, for which Bonferroni's correction for multiple comparisons was used. The criterion for significance was set at P < 0.05.
was evident two to three days after treatment and became progressively more pronounced as the post-
treatment interval increased. By 3 to 4 days, small gaps in the monolayers were apparent between the enlarged endothelial cells of the 50 ,ug/ml dose group but generally not at the lower doses. Cytoplasmic vacuolation was evident in some endothelial cells after 5 days post-treatment and this became more widespread and severe with time. Nuclei of cells in the 5 and 50 ,ug/ml wells enlarged as the post-treatment interval increased, and nucleoli were prominent. As cells enlarged, filamentous strands or stress fibers radiating from the perivascular region often became evident in the cytoplasm. Figure 1 shows the dramatic change that occurs in morphology of endothelial monolayers by 14 days after a single exposure to 0, 0.5, 5, and 50 ,ug MCTP/ml of medium. Morphology of cells and cell monolayers exposed to the 0.5 ,ug/ml dose did not differ appreciably from control. The morphology of monolayers of BSMCs was not markedly altered after MCTP administration. There was no distinct evidence of enhanced cell detachment from monolayers at any dose. In monolayers exposed to 50 ,ug/ml there appeared to be a slight decrease in cellularity of BSMC monolayers compared to control, some cells appeared more spindle-shaped, and a mild enlargement of nuclei occurred in some scattered cells after the first 5 to 7 days after exposure. These changes, however, were subtle (Figure 2) compared to those occurring in BECs. By day 14, monolayers of BSMCs remained intact and cell degeneration (ie, cytoplasmic vacuolation, nuclear pyknosis) was not enhanced compared to controls. Morphometric assessment of BEC size revealed a dose-dependent increase in cell-surface area 5 days after exposure but not at day 1 (Figure 3). At the 5 and 50 ,ug/ml doses, the cells continued to enlarge as the posttreatment interval increased. By day 15, the mean surface area of BECs exposed to 5 and 50 ,ug/ml MCTP were 5.9 and 7.9 times that of control, respectively.
Electron Microscopy
Results Cell Morphology Morphologic evaluation indicated no evidence of distinct injury to monolayers of BECs until 20 to 48 hours after a single exposure to a dose of MCTP of 5 or 50 ,ug/ml. The injury first was evident as a subtle increase in cell detachment that became progressively more pronounced in the subsequent 2 to 3 days. The cells remaining attached to the plate enlarged, so that monolayer confluency was maintained despite loss of cells. Cellular enlargement first
Endothelial cells 5 to 7 days after exposure to 5 and 50 ,ug/ml had larger cross-sectional profiles than control cells (Figure 4). The cell enlargement was not due solely to a thinning and spreading of cytoplasmic processes but consisted of an increase in nuclear and cytoplasmic area. Monocrotaline pyrrole-treated cells had large, oval, elongate nuclei with abnormally prominent, often multiple, nucleoli. These nucleoli often had segregation of granular and fibrillar components into distinct zones. Cell cytoplasm contained moderate amounts of rough endoplasmic reticulum, many polysomes, prominent golgi, and
Monocrotaline Pyrrole-induced Cell Injury
711
AJP March 1991, Vol. 138, No. 3
V.P.-
Figure 1. Changes in cell morphology of bovline pulmonarv arterial endothelial cells 14 days after a single exposure to MCTP. A: 0 pg/ml (DMF vehicle). B: 0.5 pg/ml. C: 5 i±g/ml. D: 50 ig/ml. Cells exposed to 5 and 50 iAg/ml MCTP were markedly enlarged compared to dhose exosed to 0 or 0. 5 p.g/ml. Photographs u'ere taken at the same magnification using a phase-contrast microscope (160x magnification, bar = 150 1Lm).
many microtubules and pinocytotic vesicles. Caveoli were common at the cell surface. Some MCTP-treated cells had increased perinuclear intermediate filaments and cytoplasmic microtubules. Intercellular tight junctions were evident between cells, and lamellarlike processes of plasma membrane overlapped at the cell boundaries. Some MCTP-treated and control cells had variably sized, perinuclear autophagic vacuoles. Monocrotaline pyrroletreated cells in the process of detaching from the monolayer had small vesicular dilations of endoplasmic reticulum and increased numbers of autophagic vacuoles.
Lactate Dehydrogenase Release Lactate dehydrogenase activity in the media in which BEC monolayers were incubated was not elevated significantly at any dose at day 1, but thereafter it increased in a dose-dependent manner (Figure 5). At the two higher MCTP doses, a tendency toward increased LDH activity in the medium occurred at day 2, and this increase became statistically significant and progressively greater
with time. The magnitude of the LDH release from the BECs was dose dependent. The lowest MCTP concentration tested (0.5 jig/ml) did not result in increased LDH release at any time during the 6-day experiment. In contrast to the BEC, LDH activity in medium from BSMC did not increase at any time or dose as a result of MCTP exposure.
Prostacyclin Production At day 1 after treatment, the concentration of 6-ketoPGFia in media from BEC monolayers exposed to MCTP was not different from controls at any dose (Figure 6). Thereafter it increased in wells exposed to MCTP at concentrations of 5 ,ug/ml and more. The magnitude of the increase was largely dose-dependent, although monolayers exposed to 50 ,ug/ml MCTP consistently produced slightly less 6-keto-PGF1,, than did monolayers exposed to 25 ,ug/ml. The four replications in this study yielded similar findings, although the magnitude of the increase in 6-keto-PGF1, varied considerably among studies.
712
Reindel and Roth
AJP March 1991, Vol. 138, No. 3
Figure 2. Changes in cell morphology of bovine smooth muscle cells 7 days after a single exposure to MCTP. A: Cells exposed to 0 Ag/ml (DMF vehicle); B: 0.5 pg/ml; C: 5 pg/l; D: 50 pg/ml. Photographs were taken at the same magnification using phase-contrast microscopy of hematoxylin-stained cell monolayers (160X magnification, bar = 150 pm).
Monolayer Cellularity Enumeration of BECs in monolayers from which PGI2 release was determined showed that monolayer cellularity was not altered by MCTP at day 1 after treatment (Figure
7). Thereafter the number of cells remaining in the BEC monolayers exposed to an MCTP concentration of 5 ,ug/ml or more decreased with time. The cellularity of monolayers exposed to 25 ,ug/ml MCTP was not significantly different from that of monolayers exposed to 50
-, 6000 C
0
o
*
5000
E 4' 4000 cr
%w 3000 *
< 2000 Li 1000
*
0 DAY 1
DAY 15 DAY 5 MCTP CONCENTRATION (,.g/ml)
Figure 3. Effect of MCTP on surface area of bovine pulmonary arterial endothelial cells 1, 5, and 15 days after a single treatment with MCTP. Results are expressed as mean + SEfor 150 to 200 cells per treatment group. Asterisks indicate significant differences from cells exposed to 0 p.g/ml MCTP, P < 0.05.
Monocrotaline Pyrrole-induced Cell Injury 713 AJP March 1991, Vol. 138, No. 3
A....
B.
Figure 4. Transmission electron micrographs ofbovine pulmonary arterial endothelial cells 5 davs after exposure to 0 (A) or 50 pig/ml (B) MCTP. The endothelial cell exposed to MCTP has an enlarged cell profile and a nucleus with prominent, multiple nucleoli. The nucleolus is separated into distinct zones of granular andfibrillar components (3500 x magnification).
,ug/ml. The cellularity of BEC monolayers receiving 0.5 or O ,ug/ml increased slightly after day 1, with the number of cells in the 0 ,ug/ml group consistently greater than 0.5 ,ug/ml group.
Colony-forming Ability Despite the differences in morphologic and cytotoxic responses between BECs and BSMCs, the effect of MCTP on colony-forming efficiency was similar (Figure 8). Both cell types exposed to 0.5 ,ug/ml MCTP exhibited significantly reduced ability to proliferate. No colonies devel-
oped at concentrations of 5 and 50 ,ug/ml, although scattered, extremely large individual cells were attached to the plate surface. These cells were viable as judged by Trypan blue exclusion. Many of them had vacuolated cytoplasm and large nuclei with prominent nucleoli that became increasingly evident as the post-treatment interval increased.
Discussion A single intravenous administration of a low dose of MCTP produces pulmonary injury, pulmonary vascular
Reindel and Roth
714
AJP March 1991, Vol. 138, No. 3
BOVINE ENDOTHELIAL CELLS 7^_'
50-
LUJ
40
T
c-j -J :
*@ T
60-
LJ VI
Figure 5. Lactate debydrogenase activity (% maximal release) above monolayers of bovine pulmonary arterial endothelial cells (A) and bovine smooth muscle cells (B) after a single exposure to MCTP on day 0. Results are
30
v
T
T
20
10 0 0
1
3
2
A
I
50.0 ,Ltg/m
T
5.0 ug/ml
expressed as apercentage ofLDH releasefrom the monolayers. Asterisks indicate difference from 0 pg/ml MCTP dose groups, P < 0.05 (n =5 for BECs; n = 4 for BSMCs).
0.5 ug/ml 0 'ag/mI
5
4
DAYS AFTER MCTP TREATMENT
BOVINE SMOOTH MUSCLE CELLS 7(
i
LUJ L-J
* 50
A
A
A-A
5(
o-o
u)
LUI
*
6(
Ag/ml
5 Ag/ml 0.5 Aug/mI O ug/ml
40 -
c-j I
30 -
-J
20 :2
10-
I1
===4-I 100000060z
00
2
1
B
3
4
I
5
DAYS AFTER MCTP TREATMENT 6-KETO
PGF1a
PRODUCTION
*
Ann -T IVUU
I A-
- -
,
I /I .// ~
0
0/
--iiz
400 -
,'
U-ZS
0 0-
.'
/
1
200
*
-
E,
(0 y
..
/, ..." I---
tDJ
n. u
i-
1
2
3
.
..
T 25 ,Ag/ml I
-A
...........''''
* 1
50 ,ug/ml
5 Lug/ ml
Figure 6.
6-Keto-PGF,. concentration in me-
0.5 Ag3/ml dium above monolayers of bovine pulmonary arterial endotbelial cells exposed to 0-A--A O ,tg/ ml MCTP on day 0. Results represent mean of 4 replicate wells. Asterisks indicate significant q differences from concentrations of 6-keto-PGFia in medium from above mono5 4 layers exposed to 0 Ag/ml (DMF vebicle), P < 0.05. Standard error for individual compar-
DAYS AFTER MCTP TREATMENT
isons was 10.24 by Tukey's w-test.
Monocrotaline Pyrrole-induced Cell Injury 715 AJP March 1991, Vol. 138, No. 3
CEI
ENUMERATION
o n rl
1g/ml
0 n Ct)
Li]
-j
C-) -lJ
Figure 7. Effects of MCTP on numbers of bovine pulmonary arterial endotbelial cells in cell monolayers after a single exposure to MCTP on day 0. Results represent mean of four replicate wells (n = 4). Asterisk-s indicate significant differences from monolayers exposed to 0 pig/ml, P < 0.05. Standard error for individual comparisons was 10.24 bI
Lu]
m
Lg/ml 50 ,ug/ml
z
Tukey's w-test.
remodeling, and pulmonary hypertension in rats.8912 Pronounced lung injury, however, is slow to develop and is not prominent until several days after MCTP administration.8'1012 This injury initially is apparent as a slight increase in vascular leak and mild pulmonary interstitial edema. Thereafter lung edema becomes more widespread and severe and is accompanied by progressive interstitial inflammation and arterial vascular remodeling. The delay in development of major lung injury in MCTP-induced disease is enigmatic because MCTP is an unstable, reactive electrophile that binds covalently to tissue macromolecules or is rapidly inactivated in aqueous environments.5 '22 The delayed and progressive nature of lung injury has led some investigators to suggest that MCTP or other pyrrolic metabolites of MCT may not cause cytotoxic injury directly, but rather that binding of metabolite(s) in the lung trigger indirect mechanisms of cell injury that lead ultimately to overt lung injury.8,9,2325 Our findings in vitro indicate that endothelial cells and cell monolayers are damaged directly by a single exposure to MCTP and, as in vivo, the damage is delayed in onset and progressive. Monocrotaline pyrrole-induced injury to BECs was dose dependent. Higher doses caused enhanced BEC detachment and lysis and resulted in progressive deterioration of the endothelial cell monolayers over a period of several days. The lowest concentration (0.5 ,ug/ml) suppressed BEC proliferation but was not associated with overt cell injury or monolayer disruption. Even in monolayers exposed to high concentrations of pyrrole, many cells did not die soon after exposure but survived for 2 weeks or longer, maintaining a substantial degree of monolayer integrity. Spreading and hypertrophy of endothelial cells that survived the early toxic effects of MCTP tended to compensate for the gradual cell detachment
DAYS AFTER MCTP TREATMENT
that occurred. Bovine endothelial cell monolayers exposed to 5 ,ug/ml MCTP appeared intact throughout the 2-week period of these studies. The capacity of endothelial cells to spread and undergo hypertrophy after MCTP exposure was striking. The surface area of endothelial cells increased to approximately eight times the size of control cells by day 15 after treatment. The rate of enlargement seemed to depend, in part, on cell density, ie, room available for cell expansion. Cells exposed when monolayers were subconfluent enlarged more rapidly than cells exposed in stable, confluent monolayers. In addition, cells in confluent monolayers exposed to 5 ,ug/ml MCTP attained the size of endothelial cells exposed to higher MCTP concentrations if portions of the monolayer were removed mechanically to allow room for cell expansion (unpublished observations). Thus it appeared that expression of the hypertrophic response was limited by contact with other cells and room available for cell expansion. Higher doses of MCTP caused greater cell detachment and lysis, which consequently provided more room for cell enlargement. The hypertrophic effects observed in BECs in culture are indeed seen in endothelial cells of MCTP- and MCT-exposed rats in vivo and are reminiscent of PAinduced megalocytotic responses in liver parenchymal cells.91426 1306 Some investigators described thickened profiles of endothelial cells in the pulmonary microvasculature.13 Others have reported that small blood vessels become occluded by extremely enlarged endothelial cells, particularly in late stages of the disease.9 1426 The findings in this study suggest that the hypertrophic change in endothelium in vivo is a direct effect of MCTP. The presence of hypertrophied endothelial cells in vivo in part may reflect a loss of adjacent endothelial cells, which would have allowed some room for cell enlargement.
Reindel and Roth
716
AJP Mmach 1991, Vol. 138, No. 3
BOVINE ENDOTHELIAL CELLS
A 1
C) z
LlJ C
92 (DZ 0 z-00
*
-C-) L-
z 0
*
0 C)
0o
0
0
B C) Z
50
0.5 5 0.005 0.05 MCTP CONCENTRATION (,utg/ml)
BOVINE SMOOTH MUSCLE CELLS
1004 1-i/
z L)
Ll-0" 80H--
Z°
60-
U..
40-
z
*
0
20
*
C-)
n- .I.7
I
-
0
0.005
0.05 0.5 MCTP CONCENTRATION
5
(Ag9 /ml)
Ultrastructural studies of rat lungs have indicated that endothelial cell destruction occurs but is not widespread after administration of MCTP.14'1531 Early after MCTP treatment in vivo, leakage of intravenously administered particles from the pulmonary vasculature occurred in vessels with relatively normal-appearing endothelial cell profiles.15,31 Similarly electron microscopic examination of cultured endothelial cells 5 days or more after exposure to MCTP did not reveal marked degenerative changes in the vast majority of enlarged cells remaining attached to the plate. Indeed, other than thickened endothelial cell profiles and changes in the nucleus and nucleolus, consistent differences from control cells were not identified. If the responses of BECs in vitro are similar to the re-
_lift 50
Figure 8. Effect of MCTP on colony-forming efficiency of bovine pulmonary arterial endothelial cells (A) and bovine smooth muscle cells (B) determined 14 days after a single exposure to MCTP. Results are expressed as a percentage of the mean number of colonies observed in plates receiving 0 .g/ml (DMF vehicle). Each value represents the mean + SE (n = 4). Asterisks indicate significant differencesffrom controls (0 p.g/ml), P < 0.05.
sponses that occur in vivo, then slow death of some cells with spreading and hypertrophy of surviving endothelial cells could allow for substantial, albeit temporary, maintenance of an intact vascular endothelium, thereby retarding development and progression of interstitial edema. This may explain, in part, why major lung leak and overt lung injury appear to be delayed after MCT or MCTP is given in vivo.3'8 Initially endothelial cell spreading may compensate for limited cell loss and may temporarily maintain integrity of the endothelial barrier. However, when the limit of this compensatory response is reached in the presence of a block in cell replication and with continued cell attrition, the endothelium presumably would become leakier, and exposure of the subendothelial matrix would occur. Exposure of the subendothelium
Monocrotaline Pyrrole-induced Cell Injury
717
AJP March 1991, Vol. 138, No. 3
would result in increased adherence of platelets. Indeed platelets sequester in the lungs of animals treated with MCTP32 but not until several days after treatment. Enhanced retention of platelets in the lung occurs coincident with the development of major pulmonary vascular leak and lung injury. These findings in vivo are consistent with our observation in vitro that gaps in the endothelium do not appear until several days after endothelial cell injury is initiated. Thus the compromised ability of endothelium to repair via the process of cell proliferation may be critical to the development of delayed and progressive vascular injury in this model. Based on studies in vivo in which rats were given Crotalaria spectabilis, several investigators suggested that endothelial cell hyperplasia in pulmonary arteries occurs in end-stage lung disease.26'27 This inference was based on an increase in nuclear profiles in blood vessels26 and on a slight increase in tritiated thymidine incorporation by endothelial cells that occurs in the pulmonary vasculature at days 7 and 21.27 If endothelial cells of rat pulmonary arteries in vivo respond similarly to BECs in vitro, then MCTP probably does not induce this hyperplastic response directly. Monocrotaline pyrrole induces a block in cell proliferation in each of the several cell types we have tested to date33 (also, unpublished observation), thus lending support to this premise. Rather any hyperplastic response that occurs in vivo may represent a reparative response of vascular cells that were not exposed to the toxicant. It should be mentioned that increased thymidine incorporation may represent unscheduled DNA synthesis and repair in cells with DNA damaged by MCT metabolites rather than an indication of cell proliferation. Endothelial cells are known to produce a variety of factors that can influence smooth muscle cell growth or vasomotor activity. PGI2 is one such factor. The accumulation of 6-keto-PGFia, the stable metabolite of PGI2, was enhanced in media above monolayers of BECs treated with MCTP. The enhanced production of PGI2 was dependent on the dose of MCTP. There does, however, appear to be an optimal dose for maximal release of this product because 50 ,ug/ml MCTP consistently resulted in less accumulation of 6-keto-PGF1, in the media than did 25 ,ug/ml. A similar maximal release of PGI2 by endotoxintreated endothelial cells in culture was reported.34 As with the other markers of endothelial injury, the enhanced release of PGI2 was delayed, gradually increased over time, and paralleled the enhanced detachment of cells from monolayers. The results in Figure 6 were expressed as the concentration of immunoreactive 6-keto-PGF1,,, in the medium. Because the number of surviving cells was reduced by the higher MCTP doses (Figure 7), the amount of PGI2 produced per surviving cell was probably substantial at these higher doses. If data were expressed relative to the number of viable cells remaining in the
monolayer, the response would be amplified. However this latter approach to normalizing data also has its bias because cells in the process of detaching and dying may produce substantial amounts of prostacyclin, and their contribution to the overall concentration would be underestimated. On close inspection, there is an apparent, minor disparity between the results of the cell enumeration assay and the morphometric results. The cell enumeration data indicate that there is more extensive loss of cells from the monolayer than that indicated by the morphometric studies. For example, by day 5 the number of cells remaining in the monolayers exposed to 50 ,ug MCTP/ml were fewer than 20% of that at day 1. At this time, the BEC monolayer appeared largely confluent, despite only a three- to fourfold increase in cell size. The increase in cell size would not seem sufficient to maintain monolayer confluency. This discrepancy appeared to be due to the lysis of some of the MCTP-treated cells during enzymic and mechanical dissociation of cells for enumeration. The lysis of fragile cells during preparation may have led to slight underestimation of MCTP-treated cell numbers presented in Figure 7. The responses of BSMCs to MCTP indicate that cell types can differ in sensitivity to certain effects of MCTP. Concentrations that enhanced cell detachment and caused release of LDH from monolayers of BECs produced little change in monolayer morphology or LDH release from BSMCs. The resistance of BSMCs to the cytotoxic effects of MCTP may be due to inherent differences in MCTP-cell interaction and/or in cellular protective mechanisms. In vivo, responses of pulmonary arterial smooth muscle cells observed after treatment of animals with MCTP or MCT include cell hypertrophy and hyperplasia leading to thickening of medial layers of pulmonary arteries and muscularization of nonmuscular arteries.11'12'25'27 Our findings in vitro indicate that even low concentrations of MCTP inhibit BSMC proliferation. In addition, MCTP did not induce marked changes in BSMC morphology when cells are exposed in a monolayer. This would suggest that the responses observed in smooth muscle in vivo are not caused by MCTP directly but may be indirectly induced through the release of growth factors, inflammatory mediators, or through other phenomena. The highly reactive nature of MCTP and the fact that smooth muscle cells in the vascular media are surrounded by several matrix barriers are consistent with this view because very little MCTP probably reaches smooth muscle cells in the vascular media. The MCTP-induced damage to BEC monolayers appears to develop much more slowly than damage caused by certain other endothelial cell toxicants. For example, bacterial endotoxin, homocysteine, and cyclos-
718
Reindel and Roth
AJP March 1991, Vol. 138, No. 3
porin A cause significant endothelial cell detachment at 6 hours or earlier and a pronounced release of LDH activity or 51Cr into the incubation medium by 24 hours.3>37 These toxicants caused marked disruption of endothelial cell monolayers within 24 hours of exposure as endothelial cells underwent contraction, pyknosis, and eventual detachment; however cell hypertrophy and spreading after exposure to these toxicants, as seen after MCTP exposure, has not been described. The responses of cultured endothelial cells matches in some respects the effects in vivo. Pulmonary edema induced by endotoxin or a-naphthylthiourea occurs much more rapidly than that induced by MCTP.'3 Furthermore lung edema can resolve relatively rapidly after a single exposure to doses of these toxicants that are not acutely lethal. This response is unlike the progressive nature of lung injury after a single exposure to MCTP or MCT. Thus the slowly developing nature of endothelial cell injury resulting from a single exposure to MCTP does not apply generally to other pneumotoxicants. Endothelial monolayer damage induced by hyperoxiae or gamma irradiation41 has several similarities to the damage induced by MCTP. For instance, hyperoxia can induce injury to endothelial monolayers that is delayed in development and progressive. In addition, hyperoxia or irradiation can induce endothelial cell hypertrophy and inhibit cell proliferation. Interestingly both hyperoxia and lung irradiation in vivo can cause chronic interstitial lung damage, vascular remodeling and chronic pulmonary hypertension similar in many aspects to that induced by MCTP.42A5 The fact that MCTP, hyperoxia, and irradiation can produce similar responses in cultured cells and also can induce persistent pulmonary hypertension in vivo suggests that inhibited cell proliferation with reduced capacity for repair of endothelium, gradual cell hypertrophy, and perhaps other comparable endothelial cellular changes may be important in the development of persistent pulmonary hypertension. In summary, MCTP causes direct, dose-dependent injury to BECs. The injury is slow to develop but results in progressive deterioration of endothelial cells. Cell proliferation is markedly inhibited by MCTP and limits effective repair of defects in endothelial cell monolayers. Similar responses in vivo might result in the delayed and progressive vascular leak and pulmonary edema that characterize MCT and MCTP pneumotoxicity.
Acknowledgments The authors thank D. Hummel for secretarial assistance in preparing this manuscript and K. Trosko, M. Boyce, J. Wagner, E. Shobe, K. Lovell, D. Craft, and A. Wenski-Roberts for technical assistance.
References 1. Adams R, Rodgers EF: The structure of monocrotaline, the alkaloid in Crotalaria spectabilis and Crotalaria retusa. I. J Am Chem Soc 1939, 61:2815-2819 2. Ghodsi F, Will JA: Changes in pulmonary structure and function induced by monocrotaline intoxication. Am J Physiol 1981, 240:H149-H155 3. Hilliker KS, Bell TG, Roth RA: Pneumotoxicity and thrombocytopenia after single injection of monocrotaline. Am J Physiol 1982, 242:H573-H579 4. Hilliker KS, Garcia CM, Roth RA: Effects of monocrotaline and monocrotaline pyrrole on 5-hydroxytryptamine and paraquat uptake by lung slices. Res Commun Chem Pathol Pharmacol 1983, 40:179-197 5. Mattocks AR: Toxicity of pyrrolizidine alkaloids. Nature 1968, 217:723-728 6. Mattocks AR, White INH: The conversion of pyrrolizidine alkaloids to N-oxides and to dihydropyrrolizidine derivatives by rat lung microsomes in vitro. Chem-Biol Interact 1971, 3:383-396 7. Barnes JM, Magee PN, Schoental R: Lesions in the lungs and livers of rats poisoned with the pyrrolizidine alkaloid fulvine and its N-oxide. J Pathol Bacteriol 1964, 88:521-531 8. Bruner LH, Hilliker KS, Roth RA: Pulmonary hypertension and ECG changes from monocrotaline pyrrole in the rat. Am J Physiol 1983, 245:H300-H306 9. Chesney CF, Allen JR, Hsu IC: Right ventricular hypertrophy in monocrotaline pyrrole treated rats. Exp Molec Pathol 1974, 20:257-268 10. Butler WH, Mattocks AR, Barnes JM: Lesions in the liver and lungs of rats given pyrrole derivatives of pyrrolizidine alkaloids. J Pathol 1970, 100:169-175 11. Lalich JJ, Johnson WD, Raczniak TJ, Schumaker RC: Fibrin thrombosis in monocrotaline pyrrole-induced corpulmonale in rats. Arch Pathol Lab Med 1977, 101:69-73 12. Reindel JF, Ganey PE, Wagner JG, Slocombe RF, Roth RA: Development of morphologic, hemodynamic, and biochemical changes in lungs of rats given monocrotaline pyrrole. Toxicol Appl Pharmacol 1990 (In press) 13. Valdivia E, Lalich JJ, Hayashi Y, Sonnad J: Alterations in pulmonary alveoli after a single injection of monocrotaline. Arch Pathol 1967, 84:64-76 14. Butler WH: An ultrastructural study of the pulmonary lesion induced by pyrrole derivatives of the pyrrolizidine alkaloids. J Pathol 1970,102:169-175 15. Hurley JV, Jago MV: Pulmonary oedema in rats given dehydromonocrotaline: A topographic and electronmicroscopic study. J Pathol 1975, 117:23-32 16. Rosenberg HC, Rabinovitch M: Endothelial injury and vascular reactivity in monocrotaline pulmonary hypertension. Am J Physiol 1988, 255:H1484-H1491 17. Jaffe EA, Nachman RL, Boecker CG, Minick CR: Culture of human endothelial cells derived from umbilical veins. J Clin Invest 1973, 52:2745-2756 18. Booyse FM, Sedlak BJ, Rafelson ME Jr: Culture of arterial
Monocrotaline Pyrrole-induced Cell Injury
719
AJP March 1991, Vol. 138, No. 3
19.
20. 21.
22. 23.
24.
25. 26. 27.
28. 29.
30.
31. 32.
endothelial cells. Thromb Diath Haemorrh 1975, 34:825839 Mattocks AR: Dihydropyrrolizine derivatives from unsaturated pyrrolizidine alkaloids. J Chem Soc 1969, C:1 155-1162 Roth RA: Effect of a mixed function oxidase inducer and inhibitor on monocrotaline pyrrole pneumotoxicity. Toxicol Appl Pharmacol 1986, 85:416-427 Bergmeyer HU, Bemt E: Lactate dehydrogenase UV assay with pyruvate and NADPH. In Bergmeyer HU, ed. Methods of Enzymatic Analysis, Vol. 2. New York, Academic Press, 1974, pp 574-579 White INH, Mattocks AR: Reaction of dihydropyrrolizines with deoxyribonucleic acids in vitro. Biochem J 1972, 128:291-297 Langleben D, Reid LM: Effect of methylprednisolone on monocrotaline-induced pulmonary vascular disease and right ventricular hypertrophy. Lab Invest 1985, 52:298-303 Kay JM, Heath D: Crotalaria spectabilis. The Pulmonary Hypertension Plant. Springfield, IL, Charles C. Thomas, 1969 Meyrick B, Gamble W, Reid L: Development of Crotalaria pulmonary hypertension: Hemodynamic and structural study. Am J Physiol 1980, 239:H692-H702 Tumer JH and Lalich JJ: Experimental cor pulmonale in the rat. Arch Pathol 1965, 79:409-418 Meyrick BO, Reid LM: Crotalaria-induced pulmonary hypertension: Uptake of 3H-thymidine by the cells of the pulmonary circulation and alveolar walls. Am J Pathol 1982, 106:84-94 Peterson JE: Effects of the pyrrolizidine alkaloid lasiocarpine-N-oxide on nuclear and cell division in the liver of rats. J Pathol Bacteriol 1965, 89:153-171 Downing DT, Peterson JE: Quantitative assessment of the persistent antimitotic effect of certain hepatotoxic pyrrolizidine alkaloids on rat liver. Aust J Exp Biol Med Sci 1968, 46:493-502 Jago MV: The development of hepatic megalocytosis of chronic pyrrolizidine alkaloid poisoning. Am J Pathol 1969, 56:405-422 Plestina R, Stoner HB: Pulmonary oedema in rats given monocrotaline pyrrole. J Pathol 1972, 106:235-249 White S, Roth RA: Pulmonary platelet sequestration is in-
33.
34. 35.
36.
37. 38.
39. 40.
41.
42.
43.
44.
45.
creased following monocrotaline pyrrole treatment of rats. Toxicol Appl Pharmacol 1988, 96:465-475 Reindel JF, Chang CC, Madhukar BV, Roth RA: Monocrotaline pyrrole induces antiproliferative and delayed cytotoxic responses in cultured cells. FASEB J 1988, 2:A1 181 Meyrick BO: Endotoxin-mediated pulmonary endothelial cell injury. Fed Proc 1986, 45:19-24 Harlan JM, Harker LA, Reidy MA, Gajdusek CM, Schwartz SM, Striker GE: Lipopolysaccharide-mediated bovine endothelial cell injury in vitro. Lab Invest 1983, 48:269-274 Zoja C, Furci L, Ghilardi F, Zilio P, Benigni A, Remuzzi G: Cyclosporin-induced endothelial cell injury. Lab Invest 1986, 55:455-462 Wall RT, Harlan JM, Harker LA, Striker GE: Homocysteineinduced endothelial cell injury in vitro: A model for the study of vascular injury. Thromb Res 1980, 18:113-121 Meyrick B, Brigham KL: Acute effects of Escherichia coli endotoxin on the pulmonary microcirculation of anesthetized sheep. Lab Invest 1983, 48:458-470 Fantone JC, Kunkel RG, Kinnes DA: Potentiation of a-naphthylthiourea-induced lung injury by prostaglandin El and platelet depletion. Lab Invest 1984, 50:703-710 Phillips PG, Higgins PJ, Malik AB, Tsan M: Effect of hyperoxia on the cytoarchitecture of cultured endothelial cells. Am J Pathol 1988,132:59-72 Rosen EM, Vinter DW, Goldberg ID: Hypertrophy of cultured bovine aortic endothelium following irradiation. Radiation Res 1989,117:395-408 Jones R, Zapol WM, Reid L: Pulmonary artery remodeling and pulmonary hypertension after exposure to hyperoxia for 7 days. A morphometric and hemodynamic study. Am J Pathol 1984, 117:273-285 Coflesky JT, Adler KB, Woodcook-Mitchell J, Mitchell J, Evans JN: Proliferative changes in the pulmonary arterial wall during short-term hyperoxic injury to the lung. Am J Pathol 1988, 132:563-573 Perkett EA, Brigham KL, Meyrick B: Chronic pulmonary hypertension following thoracic irradiation in sheep. J Appl Physiol 1986, 61:1875-1881 Meyrick B, Perkett EA, Brigham KL: Inflammation and models of chronic pulmonary hypertension. Am Rev Resp Dis 1987, 136:765-767