Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

While monocrotaline (MCT) intoxication has been used as a model of pulmonary hypertension since 1961 (1), the mechanism by which MCT causes pulmonary toxicity is still not completely determined (2). MCT is an 11-membered macrocyclic pyrrolizidine alkaloid (PA) that causes a proliferative pulmonary vasculitis, pulmonary hypertension, and cor pulmonale in Sprague-Dawley rats (3). The MCT rat model facilitates studies of pulmonary hypertension because it requires only a single injection that leads to progressive pulmonary vascular hypertension with prominent vascular lesions (2). Most macrocyclic PAs cause acute periacinar hepatic necrosis, and chronically cause random individual hepatocellular necrosis, megalocytosis, mitotic inhibition, hepatic fibrosis, and veno-occlusive disease leading ultimately to hepatic failure and portal hypertension. Monocrotaline administered to Sprague-Dawley rats causes acute periacinar hepatic necrosis only at high doses, but a single lower dose causes pulmonary hypertension with no hepatic lesions. Monocrotaline requires biotransformation by the liver to the putative reactive metabolite monocrotaline pyrrole (MCTP) (4, 5). The pulmonary vascular endothelium is thought to be a likely target of MCT intoxication based on circulatory proximity to the liver and the early changes observed in morphology, extravascular leakage of large molecular weight tracers, decreased 5-hydroxytryptamine clearance as well as increased thymidine uptake by endothelial cells in both pulmonary arteries and veins (2, 6, 7).

Proliferative inhibition of cells exposed to PAs has been shown previously in vivo and in vitro (8). A characteristic alteration of cells exposed to PAs, known as megalocytosis, is defined by nuclear and cytoplasmic gigantism, and this change occurs both in vivo and in vitro in numerous cell types (8, 11). A marked inability of transformed bovine kidney cells to proliferate after exposure to metabolized MCT was shown by Hincks and colleagues (12). Bovine pulmonary artery endothelial cells (BPAEC) exposed to MCTP became megalocytic and failed to proliferate through 15 d (13). Macrocyclic PAs cause cross-linking of macromolecules through bifunctional reactive sites that might be related to cell cycle disturbances leading to megalocytosis (2, 9). Cross-linking of DNA induced by MCTP in porcine pulmonary endothelial cells was associated with megalocytosis and loss of proliferative ability, but with a retention of the ability to synthesize DNA, RNA, and proteins (7, 14). This inability of pulmonary endothelial cells to proliferate may contribute to the pathogenesis of MCT-induced disease.

Previous studies in our laboratory utilizing BPAEC exposed to MCTP showed covalent binding to DNA associated with a cell-cycle arrest (15). The stage of the cell-cycle arrest was concentration dependent, with a G2 + M phase arrest in cells treated with a low concentration (5 µg/ml) of MCTP whereas an S phase arrest occurred in cells treated with a high concentration (34.5 µg/ml) of MCTP. The cell-cycle arrest was present as soon as 24-48 h in treated cells and persisted at least through 96 h. Treated cells, whether arrested in S phase or G2 + M phases, were unable to complete a monolayer, unlike control cells. This is similar to findings of an inability of MCT-treated renal epithelial (12) and MCTP-treated pulmonary artery endothelial (14) cells to proliferate. When these in vitro findings are correlated with the in vivo model, the early vascular leakage in the pulmonary parenchyma apparent between 48 and 168 h after injection with MCT (16, 17) could be related to an inability of the injured pulmonary vascular endothelium to undergo proliferative repair.

The passage of cells through the cell cycle depends on, among other things, tight regulation of cyclin gene transcription, timely degradation of cyclin proteins, and modification of the kinase subunits by phosphorylation (for review see the work by Pines [18]). In the G2/M phase transition, cdc2 kinase (also known as cyclin-dependent kinase 1) activation is the pivotal step. To be active as a kinase, cdc2 must be phosphorylated on threonine 161 (Thr 161; by cyclin-dependent kinase-activating kinase) and dephosphorylated at threonine residue 14 (Thr 14) and tyrosine residue 15 (Tyr 15) (by cdc25C phosphatase). Once the cdc2 kinase activity is maximal, the cell progresses through M phase.

Three different phosphorylation states of cdc2 can be detected by conventional sodium dodecyl sulfate (SDS) gel electrophoresis and Western blotting. The most electrophoretically retarded form is the hyperphosphorylated form of cdc2 with phosphorylations of Thr 14, Tyr 15, and Thr 161, representing the inactive form of cdc2 kinase present in G2 phase. The fastest migrating form represents the active kinase, present in M phase, in which only the Thr 161 residue is phosphorylated (19). The middle band represents an intermediate amount of phosphorylation on either Thr 14 or Tyr 15 residues (20).

DNA damage caused by numerous alkylating agents or ionizing radiation in various cell types has been associated with a G2 phase delay or arrest (21, 22). This G2 phase delay or arrest can occur by a variety of mechanisms. Nitrogen mustard-induced alkylation of CA46 cells is associated with the accumulation of cyclin B protein (to levels seen at mitosis), but there is no accumulation of activated cdc2 kinase during the G2 arrest due to the inability of cdc25C to become activated as a phosphatase (20, 23). High doses of ionizing radiation of mammalian cells in S phase are associated with G2 arrest and a decreased amount of cyclin B mRNA and protein (24). Both nitrogen mustard and ionizing radiation result in the accumulation of inactive cdc2 kinase and consequently an inability of the arrested cells to progress into M phase without a delay to repair the DNA damage.

Studies in this paper were designed to: (1) determine whether the MCTP-induced cell-cycle arrest persists through 4 wk after a 48-h exposure, (2) correlate cell-cycle arrest with megalocytosis over this same time period, (3) determine whether MCTP-induced cell-cycle arrest is reversible after a short-term exposure of BPAEC to MCTP, (4) determine by Western blotting whether MCTP-induced cell-cycle arrest is associated with an altered phosphorylation state of cdc2 protein, and (5) determine by cdc2 kinase assay whether MCTP-induced cell-cycle arrest in BPAEC is associated with inactive cdc2 kinase.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents

MCT was obtained from coarsely ground seeds of Crotalaria spectabilis (kindly provided by Dan Stankey, U.S. Department of Agriculture, Plant Materials Center, Brooksville, FL). The extraction and purification of MCT was as previously described (25), except that 0.25 N HCl was used instead of 0.5 N HCl. Monocrotaline was converted to MCTP by the method of Mattocks and associates (26), with the pyrrole being recrystallized from hexane:ethyl ether (1:1). Chemicals for the radioimmunoprecipitation assay (RIPA) lysis buffer were obtained from Sigma (Sigma Chemical Company, St. Louis, MO). Chemicals for the polyacrylamide gels were obtained from Bio-Rad (Bio-Rad Laboratories, Hercules, CA). All other chemicals and supplies, unless otherwise noted, were obtained from Fisher Scientific (Springfield, NJ).

Cell Culture

Endothelial cells used in these studies were isolated and purified from a single bovine pulmonary artery obtained from an abattoir. These endothelial cells were initially identified by the characteristic cobblestone morphology of endothelial cell monolayers, positive staining for factor VIII-related antigen, and uptake of low-density lipoprotein. Cells used in these studies were between passages 16 to 19 and the cdc2 kinase assay was also repeated on passage 27 cells. The cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 1.89 mM L-glutamine and 7% fetal bovine serum (FBS) (Sigma). Cells were maintained in an incubator with 5% CO₂ at 37.4°C with three exchanges of medium per week unless otherwise noted. Additionally, the cdc2 kinase experiment was also repeated using passage 16 cow pulmonary artery endothelial cells purchased from ATCC (CCL-209; Waldorf, MD). The ATCC cells were grown similarly to the above except, as recommended by ATCC, the medium was supplemented with 20% FBS. In the following experiments, all concentrations of MCTP were added in 5 µl dimethylformamide (DMF) unless otherwise noted.

Cell Cycle

Persistence of cell-cycle arrest after 48-h MCTP exposure. To determine whether BPAEC exposed to MCTP remained in cell-cycle arrest for longer than 96 h, a single confluent 75-cm² flask was passaged into 25-cm² flasks (n = 27) for the long-term post-MCTP exposure experiment. The 25-cm² flasks were allowed to become confluent before initiating the experiment. The flasks were treated with 5 µg/ml MCTP at time 0 and allowed to incubate for 48 h, at which time the medium containing MCTP was removed and the cells were washed twice (2 min) with 5 ml of nonsupplemented DMEM, before adding regular supplemented DMEM. Cells were collected (as described in FLOW CYTOMETRY section, below) at 1, 2, and 4 wk. There were three replicate flasks for each control and treatment group; due to the decrease in cell numbers over time, the 4-wk MCTP-treated group was prepared by pooling three flasks for each sample (n = 2). Control and treated flasks were examined by phase contrast microscopy 3 times a week. Cell images were captured as described (in Cell Morphometry section, below).

Confluent monolayer: 1-h MCTP exposure. To determine the effect on the cell cycle of a brief MCTP exposure in confluent BPAEC, a single confluent 75-cm² flask was passaged into 25-cm² flasks (n = 27). Flasks were allowed to reach a confluent monolayer before beginning the experiment. Cells (n = 3 flasks per time point) were treated with 5 µg/ml MCTP (low concentration) or 34.5 µg/ml MCTP (high concentration) for 1 h, followed by washing twice (2 min) with nonsupplemented DMEM. The medium was changed thereafter every 2 d. Start control cells were collected immediately before MCTP treatment was initiated. Treated cells were collected at 24, 48, 96, and 168 h after initiation of the experiment. All time points were compared with the start control cells because previous experiments indicated that there are insignificant changes in the cell-cycle distribution of confluent BPAEC over the time course of this experiment (see Figure 1; Thomas and coworkers [15], and unpublished observations). Cells were examined daily by phase microscopy.

View larger version (13K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 1.   Long-term effects on cell cycle in BPAEC after a 48-h exposure to a low concentration of MCTP. Confluent BPAEC were treated with 5 µg/ml of MCTP, delivered in 5 µl of DMF vehicle, at time 0 and incubated for 48 h, at which time the cells were washed and supplemented DMEM was returned and changed every 3 d for the remainder of the experiment. Control cells were treated with supplemented DMEM and then treated as above. Control and treated cells were collected for flow cytometric examination at time 0 (control only) and 1, 2, and 4 wk after initiation of the experiment. Each data point represents the mean ± SE of three separate flasks of cells, with the exception of the 4-wk treated data point, which represents six flasks pooled into two samples. (A) Percentage of cells in G1 phase; (B) percentage of cells in S phase or G2 + M phase. *Significant difference from control cells (P  0.05).

Log phase: 1-h MCTP exposure. To determine the effect on the cell cycle of a short-duration MCTP exposure in rapidly proliferating BPAEC, two confluent 75-cm² flasks were passaged into 25-cm² flasks (n = 48). Cells were treated beginning in log phase (~ 50% confluent) by applying a low or high dose of MCTP to the cells for 1 h, and then washing and changing the medium as above. Start control cells were collected immediately before MCTP treatment was initiated. Both control and treated cells were collected at 8, 24, 48, 96, and 168 h after initiation of the experiment. Three replicate flasks were examined for each time point except the 168-h high-concentration experiment, where cell numbers were insufficient for analysis by flow cytometry. Cells were analyzed daily by phase microscopy.

Flow Cytometry

After treatment, attached cells were collected from flasks by replacing the medium with 1.5 ml of 0.05% trypsin, 0.53 mM ethylene diaminetetraacetic acid (Life Technologies GIBCO BRL, Gaithersburg, MD). Before transfer and centrifugation (250 × g, 10 min, 4°C) 3 ml of DMEM supplemented with 25% FBS was added. Cells were next washed with 4 ml of ice-cold phosphate-buffered saline (PBS), pelleted, and resuspended with vortexing in 1.4 ml of ice-cold 70% ethanol. The suspension was stored at 4°C until use. On the day of examination, the cells were recentrifuged, washed with 1 ml of ice-cold PBS, recentrifuged, and decanted. Washed cells were incubated (37°C, 30 min) with RNAase A (Sigma; 250 µl; 500 units/ml in 1.12% Na citrate buffer), followed by the addition of propidium iodide (Molecular Probes Inc., Eugene, OR; 250 µl; 0.05% solution in 1.12% Na citrate buffer) (15). Cells were allowed to incubate for an additional 20 min in the absence of light. The cells were then examined on a FACScan flow cytometer (Becton Dickinson, Mountain View, CA) and analyzed using MULTICYCLE software (Phoenix Flow Systems, San Diego, CA) written and developed by P. S. Rabinovitch of the University of Washington (Seattle, WA).

Cell Morphometry

Cell images were captured and digitized at 48 h and at 1, 2, and 4 wk, during the persistence of the cell-cycle arrest experiment, using a Sony video camera mounted on a Leitz Fluovert inverted phase contrast microscope and a video frame grabber card (Data Translations, Inc., Marlboro, MA) mounted in a Macintosh IIvx microcomputer. For control samples, three microscopic fields (×100) were captured per flask (n = 3) per time point. In control samples, 15 randomly selected cells per field were then measured (a total of 135 cells measured per treatment) for all time points, with the exception of the 48-h sample where a total of 45 cells were measured. In MCTP-treated samples, more fields were captured due to the large size of the cells and a total of 24, 108, 100, or 94 cells were measured from the 48-h, 1-wk, 2-wk, or 4-wk samples, respectively. Surface area occupied by individual cells was measured by planimetry performed with the public domain program NIH Image version 1.54 (NTIS, 5285 Port Royal Rd., Springfield, VA 22161; part number PB93-504868).

Western Blotting for cdc2 Protein

A time course was performed to examine the alteration of the relative amounts of the multiple phosphorylated forms of cdc2 in control and MCTP-treated cells. For each experiment a single confluent 75-cm² flask of BPAEC was split into 75-cm² flasks (n = 6) and upon reaching confluency, the six flasks were split into 75-cm² flasks (n = 30). BPAEC were grown to 90% confluency, at which time they were lysed (control cells) or treated with 5 µg/ml MCTP for 24, 48, or 96 h and then lysed. A sample consisted of three similarly treated flasks of cells that were pooled by serially adding the lysate buffer from the first to the second and then the third flask. Each treatment had an n = 3 (nine flasks total). A replicate flask for each treatment was collected by trypsin dissociation for cell counts. Cells were lysed in 400 µl of modified RIPA buffer (50 mM Tris-HCl, pH 7.4; 1% NP-40; 0.25% Na deoxycholate; 150 mM NaCl; 1 mM ethyleneglycol-bis-(-aminoethyl ether)- N,N'-tetraacetic acid [EGTA]; 1 mM phenylmethylsulfonyl fluoride; 1 µg/ml aprotinin; 1 µg/ml leupeptin; 1 µg/ml pepstatin; 1 mM Na₃VO₄; 1 mM NaF) (Upstate Biotechnology, Inc., Lake Placid, NY; 1995 catalogue protocol). The lysate was centrifuged (16,000 × g, 20 min, 4°C) and the supernatant collected. The supernatant was affinity-purified by adding 50 µl of p13^suc1-agarose beads to concentrate cyclin-dependent kinases (Upstate Biotechnology), and gently rocked overnight at 4°C. The samples were centrifuged (1,300 × g, 6 min, 4°C) to pellet the agarose beads. The supernatant was removed and saved for later protein analysis. The beads were then washed twice with ~ 200 vol modified RIPA buffer followed by centrifugation. Next, 50 µl of 2× treatment buffer (0.125 M Tris, 4% SDS, 20% vol/vol glycerol, 0.2 M dithiothreitol [DTT], 0.02% bromophenol blue, pH 6.8) was added; the samples were heated to 100°C (5 min) and centrifuged, and the supernatant (affinity-purified sample) was collected and frozen at 80°C until use. The affinity-purified samples as well as the corresponding supernatants were thawed and separated by SDS polyacrylamide gel electrophoresis (SDS-PAGE) according to the protocol of Laemmli (27) on a polyacrylamide (4% T, 2.7% C stacking, and 11% T, 2.7% C running) gel of 0.75-mm thickness and 12-cm length. Each gel was loaded with control samples collected from the same experiment as were the treated samples (25-µl sample per lane). The gels were run at 4°C in a Hoefer SE 600 model vertical tank chamber (Hoefer Scientific Instruments, San Francisco, CA) at 10 mA per gel for an average of 6 h. Gels were transferred at 95 V in a Hoefer vertical tank system for 90 min at 4°C in transfer buffer (0.025 M Tris, 0.192 M glycine, 0.0375% wt/vol SDS, 20% vol/vol MeOH) onto 0.2-µm polyvinylidene difluoride (PVDF) membranes (Bio-Rad).

The PVDF membrane was then blocked with 3% non-fat milk (NFM) (Bio-Rad) in Tris-buffered saline (TBS) buffer (50 mM Tris, 150 mM NaCl, pH 7.4) for 30 min at room temperature by gentle rocking. The membrane was covered by TBS containing 3% NFM and 2 µg/ml of rabbit antihuman C-terminal cdc2 antibody (Upstate Biotechnology) overnight at 4°C with gentle rocking. The membrane was washed twice with TBS buffer (5 min, 50 ml) and then exposed to 3% NFM, TBS solution containing a 1/2,000 dilution of alkaline phosphatase conjugated antirabbit secondary antibody (Bio-Rad) for 1 h at room temperature with gentle rocking. The membrane was then washed twice with Tween TBS (0.05% Tween 20, 50 mM Tris, 150 mM NaCl, pH 7.4) (5 min, 50 ml), and then three times with TBS (5 min, 50 ml). Identification of cdc2 was accomplished by incubating the membrane for 90 min at room temperature with a premixed substrate reagent kit (Bio-Rad) containing 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium. The reaction was terminated by washing twice with ddH₂O (10 min).

The absorbance (540 nm) of the cdc2 bands was determined by densitometry on a Shimadzu CS-9301PC dual wavelength flying spot scanning densitometer (Shimadzu, Columbia, MD) in reflectance mode. The densitometric scans were performed from ~ 2 cm above (toward the origin) to ~ 1 cm below the cdc2 bands. Two cdc2 positive bands (occasionally three) were detected by the densitometer (Figure 2C). The ratio of the area of the slower migrating band (or bands) to the area of the faster migrating band was calculated and control and MCTP-treated samples were compared. The total area of the cdc2 bands per lane was compared in control and MCTP-treated samples. The total area of cdc2 data was then normalized for number of cells that were affinity-purified. A separate gel was run simultaneously with the above gel and was loaded with the supernatants (25 µl per lane) from the corresponding p^13suc1 affinity-purified sample. The gel was fixed and stained with Coomassie blue to detect protein. The gel lanes were analyzed on the densitometer (540 nm) in transmission mode, from the origin to the end of the gel. As an estimate of total protein, the total area for each lane (corresponding to approximately the total protein) was determined and control and MCTP-treated samples were compared with statistical significance determined as above. Alternatively, the total area per lane was normalized for the number of cells that were affinity-purified and similar calculations performed.

View larger version (15K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 2.   MCTP-induced alterations in phosphorylation state of cdc2 protein. BPAEC, 90% confluent, were treated with 5 µg/ml (low concentration) MCTP for 0 (control), 24, 48, or 96 h and lysed, followed by affinity purification with p13suc1 and separation by SDS-PAGE, then Western blotting and immunodetection by C-terminal cdc2 antibody. (A) Membrane 1: lanes 1 and 10, cdc2 standard (Santa Cruz); lanes 2, 4, and 8, control BPAEC; lanes 3, 7, and 9, 96-h MCTP-treated BPAEC; lanes 5 and 6, molecular weight standards (Bio-Rad). The upper band represents inactive cdc2, while the lower band represents active cdc2. (B) Membrane 2: lane 1, cdc2 standard (Santa Cruz); lanes 2, 5, and 8, control BPAEC; lanes 3, 6, and 9, 24-h MCTP-treated BPAEC; lanes 4, 7, and 10, 48-h MCTP-treated BPAEC. (C ) Representative densitometric scans of control and 96-h MCTP-treated BPAEC lanes from membrane 1, above. The peak closest to the y-axis represents the most electrophoretically retarded cdc2 band (inactive), while the furthest peak represents the active cdc2 band.

cdc2 Kinase Assay

To determine the cdc2 kinase activity in BPAEC the Promega SignaTECT^TM cdc2 protein kinase assay (Promega, Madison, WI) was used. Cells were 85 to 90% confluent before initiating treatment. Experiments consisted of three groups as follows. Vehicle control (asynchronous) BPAEC were exposed to 10 µl of DMF for 48 h prior to harvest; MCTP was administered for 48 h at 5 µg/ml in a total of 10 µl of DMF; Colcemid was given at 0.4 µl/ml for 20 h prior to harvest. Three flasks (75 cm²) were assigned per group. Cells were washed with PBS and subsequently lysed with 100 µl of RIPA while retained in the flasks. Samples were centrifuged at 100,000 × g for 1 h and supernatant removed from the pellet. Protein concentrations were determined on 10-µl samples of supernatant using the bicinchoninic acid assay according to manufacturer's specifications (Pierce, Rockford, IL). Standard curves were prepared with bovine albumin (fraction V; Sigma), each standard and blank containing an aliquot of RIPA buffer equivalent to that of the sample to be assayed. Protein kinase cdc2 was assayed according to Technical Bulletin 227 (Promega). Briefly, 5 µg of protein was used per assay and each flask was assayed in triplicate. Total volume of the assay was 25 µl and contained 1,250 pmol of [³³P]-adenosine triphosphate (2.57 µCi); 50 mM Tris-HCl, pH 7.5; 10 mM MgCl₂; 1 mM EGTA; 2 mM DTT; 40 mM -glycerolphosphate; 20 mM p-nitrophenylphosphate; and 0.1 mM sodium vanadate. The biotinylated substrate (PKTPKKAKKL), derived from histone H1, was present at a concentration of 25 µM. Total activity was calculated by subtracting the amount of phosphorylation occurring in the absence of substrate from reactions containing substrate. Reaction mixtures were preincubated for 5 min at 37°C prior to addition of the protein samples; incubations were extended for an additional 5 min, and reactions were terminated with the addition of 12.5 µl of 7.5 M guanidine hydrochloride. An aliquot of 15 µl was spotted on the kit-provided streptavidin-coated membrane. Membranes were washed once for 30 s with 2 M NaCl, 3 times for 2 min each with 2 M NaCl, 4 times for 2 min with 2 M NaCl in 1% H₃PO₄, and twice for 30 s with distilled water. Membranes were then assayed for radioactivity by scintillation counting. Data was expressed as picomoles of [³³P] incorporated into substrate/ min/µg of protein.

Statistics

For the cell-cycle experiments, statistical significance was determined by a paired two-tailed Student's t test performed using Microsoft Excel Analysis Toolpak software (Microsoft Corp., Redmond, WA). Significant differences between groups were determined using a P value  0.05. For cell size analysis and Western blots, statistically significant differences were determined by analysis of variance (ANOVA) followed by post hoc Fisher's least significant difference test (P  0.05) using Data Desk version 4 software (Data Description, Inc., Ithaca, NY). For the cdc2 kinase assay the statistically significant differences were determined by a pooled t test (P  0.05) using Data Desk version 4 software.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Persistence of Cell-cycle Arrest after 48-h MCTP Exposure

One week after confluent monolayer BPAEC were exposed to MCTP (5 µg/ml for 48 h), there was a marked reduction in the percentage of cells in G1 phase (84.0 ± 0.4 to 5.8 ± 0.2%, control and MCTP, respectively; Figure 1). Alteration of S phase proportion was also apparent with a decrease in treated cells. This was associated with a large increase in the percentage of G2 + M phase cells in the treated samples (11.74 ± 0.2 to 93.1 ± 0.3%, control and MCTP, respectively; Figure 3). Similarly, at 2 wk, the percentage of cells in G1 was reduced even further and S phase-treated cells were still decreased compared with 2-wk controls. In the 2 wk after treatment with MCTP, the percentage of G2 + M phase cells continued to increase (12.7 ± 0.2 to 96.2 ± 0.1%, control and MCTP, respectively). Similar trends were present through 4 wk, as the proportion of G1 phase cells were still markedly reduced, and there was still a downward trend in treated S phase cells, although not statistically significant. The arrest in G2 + M phase was still present at 4 wk (15.9 ± 0.3 to 94.8 ± 2.2%, control and MCTP, respectively).

View larger version (21K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   View larger version (31K): [in this window] [in a new window]   Figure 3.   Long-term effects on cell cycle in BPAEC after a 48-h exposure to a low concentration of MCTP, representative flow cytometric DNA distribution histograms. See Figure 1 caption for experimental details. The x-axis represents propidium iodide DNA fluorescence, while the y-axis represents cell numbers. (a) Start control cells; (b) 1-wk control; (c) 1-wk treated; (d) 2-wk treated; (e) 4-wk control; (f ) 4-wk treated cells.

In the control samples at all time points 25,000 cells per flask were examined by flow cytometry. In the 1-wk treated samples 25,000 cells were examined per flask, whereas in the 2-wk treated samples > 20,000 cells were examined per flask; however, in the 4-wk treated samples 6,675 or 10,080 cells were examined per sample. The average %BAD* value for these DNA histograms was 0.35%. It is recommended that %BAD be less than 20% to properly analyze cell-cycle phase distributions. The average chi-square value for these analyses was 3.3. The MULTICYCLE analysis package provides a chi-square goodness-of-fit test to judge the fit of the analysis model to the actual DNA histogram. A goodness-of-fit value of 1.0 is considered excellent, while values > 10.0 are suspect.

Cell Morphometry

The persistent G2 + M phase arrest of MCTP-treated confluent monolayer BPAEC (5 µg/ml, 48 h) was associated with a significant and continued increase in cell size, beginning by 48 h, through 28 d (Figures 4 and 5). The control cells also increased slightly in size through the experiment.

View larger version (29K): [in this window] [in a new window]   Figure 4.   Long-term effects on cell size in BPAEC after a 48-h exposure to a low concentration of MCTP. Cell images were captured and digitized at 48 h, 1 wk, 2 wk, and 4 wk on the cells described in Figure 1 legend. See MATERIALS AND METHODS section on cell morphometry for number of cells measured per time point and treatment. Surface area occupied by individual cells was determined by planimetry of the captured images. Each data point represents the mean surface area per cell ± SE per treatment per time point. *Significant difference from all other data except those marked with * (P  0.05). **Significant difference from all other data (P  0.05). #Significant difference from all other data except those marked with # (P  0.05). ##Significant difference from all other data except those marked with ## (P  0.05). ###Significant difference from all other data (P  0.05).

View larger version (111K): [in this window] [in a new window]   View larger version (93K): [in this window] [in a new window]   Figure 5.   Morphologic changes induced by MCTP in BPAEC. Confluent BPAEC were treated with (A) 0.0 or (B) 5.0 µg MCTP/ml medium for 48 h, and photomicrographs were taken at 168 h after initiating treatment. Photomicrographs were taken at an identical magnification, and the bars represent 25 µm.

Confluent Monolayer: 1-h MCTP Exposure

Low concentration (5 µg/ml). At 24 h after the initiation of a 1-h treatment of confluent BPAEC with 5 µg/ml of MCTP, there was a slight decrease in the proportion of G1 phase cells (Figure 6). In treated cells there was a 1.8× increase in the percentage of S phase cells compared with start control cells, while the proportion of treated G2 + M phase cells at 24 h was the same as control. By 48 h, a 1.4× decrease in the percentage of treated G1 phase cells was present. The proportion of treated S phase cells was similar to control, and the percentage of treated G2 + M phase cells was 2.8× higher than control. After 96 h, the decrease in the percentage of treated G1 phase cells was leveling off, while the proportion of treated S phase cells was 3× less than control cells. The percentage of treated G2 + M phase cells was still increasing to 3.7× control. The percentage of treated G1 phase cells at 168 h was 1.5× lower than controls, while the proportion of treated S phase cells was only 1.8× lower than controls. After 168 h, the percentage of treated G2 + M phase cells was 3.6× higher than controls.

View larger version (17K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 6.   Changes in cell cycle in confluent BPAEC treated with a low or high concentration of MCTP for 1 h. Confluent BPAEC were treated with 5 µg/ml (low concentration) or 34.5 µg/ml (high concentration) MCTP, delivered in 5 µl of DMF vehicle, at time 0 h and incubated for 1 h, at which time the cells were washed and supplemented DMEM was returned and changed every 2 d for the remainder of the experiment. Start control cells were collected at time 0. Treated cells were collected for flow cytometric examination at 24, 48, 96, and 168 h after initiation of the experiment. Each data point represents the mean ± SE of three separate flasks of cells. (A) Percentage of cells in G1 phase; (B) percentage of cells in S phase; and (C) percentage of cells in G2 + M phase. *Significant difference from start control cells (P  0.05)

High concentration (34.5 µg/ml). At 24 h after the initiation of a 1-h treatment of confluent BPAEC with 34.5 µg/ ml of MCTP, there was a slight decrease in the percentage of G1 phase cells as compared with start control cells (Figure 6). In treated cells there was a 2× increase in the proportion of S phase cells compared with start control cells, while the percentage of treated G2 + M phase cells declined slightly. After 48 h, the percentage of treated G1 phase cells continued to decline while the proportion of treated S phase cells continued to increase by 3.7× over control. The proportion of treated G2 + M phase cells returned to start control levels by 48 h. At 96 h, the percentage of treated G1 phase cells continued to decrease to less than one half of control while the percentage of treated S phase cells continued to increase to 5× higher than control and the proportion of treated G2 + M phase cells remained at start control levels. The percentage of treated G1 phase cells at 168 h continued to decline while the proportion of treated S phase cells continued to increase to over 6× higher than control. At 168 h, however, the percentage of treated G2 + M phase cells began to increase to over 2.5× start control cells.

Log Phase: 1-h MCTP Exposure

Low concentration (5 µg/ml). After a 1-h treatment of log phase BPAEC with 5 µg/ml MCTP, there was a slight decrease in the percentage of treated G1 phase cells at 8 h compared with 8-h control cells (Figure 7). The percentage of treated S phase cells had more than tripled, while there was a slight decrease in the proportion of treated G2 + M phase cells. By 24 h, there was a continued decrease in the percentage of treated G1 cells. Treated S phase cells at 24 h were slightly elevated. The percentage of treated G2 + M phase cells was similar to control. The decrease in the percentage of treated G1 phase cells was beginning to slow by 48 h, while the percentage of treated cells in S phase was below control levels. The proportion of treated G2 + M phase cells was 1.4× higher than control by 48 h. By 96 h, the percentage of treated G1 phase cells was increasing while treated S phase cells were similar to control. The percentage of treated G2 + M phase cells was declining similar to control cells. By 168 h, the percentage of treated G1 phase cells continued to increase and was similar to control levels. The proportion of treated S phase cells had increased to control levels, while the percentage of treated G2 + M phase cells was still decreasing toward control levels. The percentages of cells in G1, S, or G2 + M phase varied over time in the treated group as determined by an ANOVA test (P < 0.01). By 120 h after MCTP treatment, many of the cells had returned to normal size and mitotic figures could be observed again by phase microscopy. However, occasional megalocytes were still present at 120 h (data not shown).

View larger version (18K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Figure 7.   Changes in cell cycle in log phase BPAEC treated with a low or high concentration of MCTP for 1 h. Exponentially growing BPAEC were treated with 5 µg/ml (low concentration) or 34.5 µg/ml (high concentration) of MCTP, delivered in 5 µl of DMF vehicle, at time 0 and incubated for 1 h, at which time the cells were washed and supplemented DMEM was returned and changed every 2 d for the remainder of the experiment. Start control cells were collected at time 0. Control and treated cells were collected for flow cytometric examination at 8, 24, 48, 96, and 168 h after initiation of the experiment. Each data point represents the mean ± SE of three separate flasks of cells. (A) Percentage of cells in G1 phase; (B) percentage of cells in S phase; (C) percentage of cells in G2 + M phase. *Significant difference from control cells (P  0.05)

High concentration (34.5 µg/ml). After a 1-h treatment of log phase BPAEC with 34.5 µg/ml of MCTP, there was no difference at 8 h in the percentage of cells in G1 phase compared with the 8-h control cells (Figure 7). However, the percentage of treated S phase cells had more than tripled by 8 h. There was a slight decrease in the percentage of G2 + M phase cells at 8 h. By 24 h there was a slight decrease in the percentage of treated G1 phase cells. The proportion of treated S phase cells was almost 3× higher than the 24-h control cells, while the percentage of 24-h G2 + M phase cells had declined by 1.5× compared with control. The percentage of G1 phase cells (48 h) continued to decline after the MCTP treatment with a 1.6× decrease. By 48 h, there was a 4.4× increase in the percentage of treated S phase cells, while the percentage of treated G2 + M phase cells was similar to control. A continued decline in the percentage of treated G1 phase cells was present at 96 h. By 96 h, the percentage of treated S phase cells was decreasing while the percentage of treated G2 + M phase cells was beginning to increase.

Western Blotting for cdc2 Protein

The majority of cdc2 protein in control asynchronous BPAEC was composed of the faster-migrating form, while the majority of cdc2 in MCTP-treated BPAEC was composed of the slower-migrating form at all time points examined (Figure 2). On blot 1, the average ratio of the slower-migrating form to the faster-migrating form of cdc2 was 0.7 ± 0.14 for the control samples, whereas the ratio was 2.5 ± 0.46 for the 96-h MCTP-treated samples (Figure 8A). On blot 2, the average ratio was 0.3 ± 0.05 for control samples, whereas the ratio was 1.44 ± 0.27 for the 24-h MCTP treatment and 1.29 ± 0.20 for the 48-h MCTP treatment (Figure 8A).

View larger version (23K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   Figure 8.   MCTP-induced alterations in phosphorylation status and amount of affinity-purified cdc2 protein: see Figure 7 caption for experimental details. (A) In treated and control cells, ratios of the area of inactive bands (upper) to area of active cdc2 bands (lower) are reported in relative absorbance units. (B) Total area of the cdc2 bands was calculated from the densitometric scans and normalized for number of cells affinity-purified. For both graphs, means ± SE of three lanes are shown per data point. *Significant difference from control cells in blot 2 (P  0.05). **Significant difference from control cells in blot 1 (P  0.05).

Cell counts for the first membrane were 2.32 × 10⁶ cells in the replicate control flask and 606,000 cells in the replicate 96-h MCTP-treated flask. If multiplied by three to simulate the pooling of samples for affinity purification, then ~ 7 × 10⁶ control and 1.8 × 10⁶ treated cells were affinity-purified per sample run for blot 1. Similarly, ~ 4 × 10⁶ control, 3.6 × 10⁶ 24-h MCTP, and 1.5 × 10⁶ 48-h MCTP cells, respectively, were affinity-purified per sample run for blot 2. 

The total area of cdc2 bands normalized for number of cells affinity-purified was determined by densitometry. There was significantly more cdc2 protein per treated cell at the 48- and 96-h time points relative to their respective controls (Figure 8B). When non-normalized data were examined, there was no difference in the average total area of cdc2 bands between treated and control lanes (data not shown).

The total area of the Coomassie blue-stained bands (as measured by absorbance at 540 nm) within each lane (corresponding to approximately total protein) was determined by densitometry on a gel loaded with supernatants corresponding to the affinity-purified samples run in gel 1. The average area of a control lane was 474.4 ± 38.4, while the average 96-h MCTP lane was 470 ± 65.8. If, however, the total area was normalized for numbers of cells, the average relative area per cell of the control lanes was 6.8 × 10⁵ ± 5.5 × 10⁶ and 26.1 × 10⁵ ± 3.7 × 10⁵ for the 96-h MCTP lanes (data not shown). The 96-h MCTP-treated samples contained significantly more total protein (3.8×) on a per-cell basis than did controls (P < 0.01) by paired two-tailed Student's t test.

cdc2 Kinase Assay

Kinase activity for cdc2 was assayed in asynchronous control BPAEC, BPAEC treated with 5 µg/ml MCTP for 48 h, or BPAEC treated with 0.4 µg/ml Colcemid for 20 h. Maximum cdc2 kinase activity was present in Colcemid-treated BPAEC, as expected (Figure 9). MCTP-treated cells' cdc2 kinase activity was slightly higher than control asynchronous BPAEC. Serum-starved BPAEC had the lowest cdc2 kinase activity (at background levels), as would be expected (data not shown). Serum-starved cells are in G₀ phase where cdc2 kinase activity is negligible. The cdc2 kinase assay experiment was repeated using BPAEC obtained from ATCC, and similar results to those above were obtained (data not shown).

View larger version (26K): [in this window] [in a new window]   Figure 9.   MCTP- and Colcemid-induced alterations in cdc2 kinase activity: BPAEC (85 to 90% confluent) were treated with DMF (10 µl) or MCTP (5 µg/ml) for 48 h, or 0.4 µg/ml Colcemid for 20 h before harvest. cdc2 Kinase activity was determined by Promega SignaTect cdc2 protein kinase assay as described in MATERIALS AND METHODS. Three flasks (75 cm2) were assigned per treatment group and each flask was assayed in triplicate. Data was expressed as mean picomoles (33P) incorporated into substrate (PKTPKKAKKL)/min/µg protein. *Significantly different from control (P = 0.05) by pooled t test.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

PAs are known to cause proliferative inhibition both in vivo and in vitro. We have previously shown that MCTP exposure of BPAEC resulted in a concentration-dependent cell-cycle arrest over a relatively short time course, with a low concentration (5 µg/ml) inducing an S phase delay followed by a G2 + M phase arrest, whereas a high concentration (34.5 µg/ml) resulted in an S phase arrest. In this study, we showed by flow cytometry that the G2 + M phase arrest induced by MCTP (low concentration, 48 h) was irreversible through the 4 wk examined. Flow cytometric DNA analysis of propidium iodide-stained cells does not distinguish between a G2 and an M phase cell because G2 and M phase cells contain similar amounts of DNA. The G2 phase of the cell cycle generally lasts between 1 and 6 h and begins at the end of DNA replication, lasting to the beginning of the condensation of chromosomes (28). The hallmark of the G2/M phase transition is the activation of cdc2 kinase by cdc25C mediated dephosphorylation of the inhibitory phosphate groups on cdc2 (29). Therefore, the difference between G2 phase and M phase cells can be distinguished when the phosphorylation status of cdc2 is examined. In this study, MCTP induced a G2 arrest in low-concentration-exposed BPAEC as evidenced by the alteration of cdc2 phosphorylation status. The most electrophoretically retarded cdc2 band represented the inactive state of cdc2 present in G2 phase, with phosphate groups present on Tyr 15 and Thr 14 residues along with Thr 161 (19). In this study, the relative amount of the electrophoretically retarded bands increased by 24 h and remained elevated through at least 96 h compared with control asynchronous cells. The alteration in cdc2 phosphorylation state was present by as early as 12 h of MCTP exposure (data not shown).

Confirmation of the MCTP-induced cdc2 changes observed by Western blotting was performed with a biochemical cdc2 kinase activity assay. The results of this assay were similar to the Western results in that MCTP-treated cells arrested in G2 phase had only slightly higher cdc2 kinase activity than asynchronous control cells despite having almost twice the percentage of cells in G2 + M phases, as previously determined by flow cytometry. Asynchronous BPAEC treated under similar conditions have about 18% of the cells in the G2 + M phases, while MCTP treatment for 48 h results in about 30% of the cells being in the G2 + M phases (15). A 20-h treatment of BPAEC with 0.4 µg/ml Colcemid results in about 70% of the cells being in G2 + M phases as determined by flow cytometry (unpublished observations). However, Colcemid arrests cells in metaphase where cdc2 kinase activity is at its peak; therefore, it is likely that about 70% of the cells are arrested in metaphase and not in G2. As expected, there was a massive cdc2 kinase activity increase in Colcemid-treated BPAEC consistent with this metaphase arrest. If it was assumed that the MCTP-arrested cells (30%) were also blocked in metaphase instead of G2 phase, then one would expect that cdc2 kinase activity would be only about 2.3-fold less than in Colcemid-arrested cells. This is not the case; there was an approximately 15-fold increase in cdc2 kinase activity in the Colcemid-treated cells compared with the MCTP-arrested cells.

When cells are exposed to DNA damaging agents, a G2 phase delay or arrest is a common occurrence (30). The G2 checkpoint is activated when there is DNA damage or when DNA replication is not completed. This checkpoint prior to cell division is enforced by inhibiting activation of cdc2 kinase, which then allows time for repair of the defects and/or completion of DNA replication; therefore, the fidelity of the genetic information passed to the daughter cells is ensured (29, 31). We have previously shown that MCTP covalently binds to pulmonary endothelial cell DNA (15) and others have shown DNA-DNA cross-links (14). It is likely that this DNA damage causes activation of the G2 checkpoint, resulting in a G2 arrest in MCTP- exposed BPAEC. In nitrogen mustard alkylation of DNA, G2 arrest is evoked due to an inability to dephosphorylate the inhibitory sites on cdc2. Also, at low doses of ionizing radiation, a similar phenomenon occurs; however, at high doses there are reductions of cyclin B mRNA and protein which result in G2 arrest (32). The inability of MCTP- exposed BPAEC to progress into mitosis and proceed in the cell cycle is not due to a lack of cdc2 protein, because cdc2 protein actually increases with exposure to MCTP. Total cdc2 protein normalized for cell number was increased significantly by 48 h and continued to be increased through 96 h. Instead, the present study demonstrates that, much like nitrogen mustard-induced G2 arrest, the G2 arrest in MCTP-exposed BPAEC may be due to an inability to dephosphorylate the inhibitory phosphorylations on cdc2. Further studies into the mechanism regulating this cdc2 alteration are in progress.

In most experimental models of G2 arrest (including alkylating agents or ionizing radiation), the arrest generally lasts for less than 24 h (20, 24). In nitrogen mustard- associated cross-linking, the cross-links are removed to the level of control samples within 24 h after treatment (30). Wagner and colleagues showed that MCTP-induced crosslinking in porcine pulmonary endothelial cells remained above control cells for 10 d (14). Part of the reason that MCTP-induced G2 arrest persists for so long may be that the DNA damage cannot be effectively repaired.

In contrast to the irreversible cell-cycle arrest induced by a 48-h exposure to MCTP, shorter duration (1 h) low-concentration exposure of log phase cells resulted in cell-cycle arrest by 48 h with a subsequent reversal evident at 96 h. The ability to reverse the cell-cycle block was dependent on the concentration of MCTP as well as the growth phase of the BPAEC, as confluent and high-concentration-treated cells had persistent arrest. In studies performed by others, higher concentrations of PA resulted in more interstrand DNA cross-links (12, 14). A known MCTP degradation product, dehydroretronecine, has been shown in vitro to bind DNA and cause cross-linking, with the amount of cross-linking steadily increasing with increased concentrations and reaction time from 0 to 100 h (33). It may be that low concentration (1 h MCTP exposure) log phase cells have less DNA cross-linking that can then be repaired, resulting in a reversal of the G2 arrest. This does not explain why there is a difference in response in confluent versus log phase low concentration-exposed cells. It may be that the log phase cells are less susceptible to MCTP toxicosis, or, alternatively, that log phase cells are more capable of repairing the DNA damage induced by MCTP. Mitogen-stimulated human lymphocytes have an enhanced ability to repair ionizing radiation- and alkylating chemical-induced DNA damage compared with nonproliferating lymphocytes (34). Another possibility for decreased susceptibility of log phase cells to MCTP might include an increased amount of glutathione or protein thiols in proliferating cells that react with the pyrrole first, thus minimizing DNA damage. Nonproliferating human microvascular endothelial cells have half the amount of glutathione present in proliferating human microvascular endothelial cells (35). Further examination of both of these possibilities in MCTP-induced toxicosis remains to be done.

The permanence and position of S phase arrest in MCTP (high concentration)-exposed BPAEC also depends on the length of exposure. In this study (1 h, high concentration), there was a prolonged S phase delay (from 8 to 96 h) with a gradual escape into G2 + M. Gradual movement of the S phase peak from early to middle and then late S phase occurred (representing increasing amounts of DNA replication) before an eventual escape into G2 + M phases was noted on the DNA histograms (data not shown). No evidence of an increase in G1 cells was detected through 168 h (in confluent cells), indicating that the cells remained arrested in G2 + M phases. This is quite different from the previously reported data (15) where cells exposed to high-concentration MCTP until harvested for flow cytometry initially had an early S phase peak at 8 h that moved to mid-S phase, where they remained arrested for the duration of the study. When BPAEC were exposed to ¹⁴C-MCTP (34.5 µg/ml; 1 to 96 h), covalent binding to DNA was observed to increase with length of exposure (15). This may be the reason why a long-term exposure resulted in an S phase arrest while short-term exposure resulted in an extremely prolonged S phase with an eventual escape into a G2 + M phase arrest. In the yeast Saccharomyces cerevisiae, the rate of ongoing S phase is retarded when the DNA is subjected to alkylation (36).

Megalocytosis is induced by a variety of PAs by an unknown mechanism in numerous cell types both in vivo and in vitro. Megalocytosis of pulmonary type II cells (37) and alveolar macrophages (38) has been demonstrated in vivo after exposure of rats to MCT, and a suggestion of megalocytosis induced by MCTP in pulmonary endothelial cells was shown by both Butler (39) and Reindel and associates (40). Megalocytosis has been previously shown in vitro in BPAEC exposed to MCTP (13, 41). In the present study, cell enlargement associated with G2 + M phase arrest began by 48 h, and the cells continued to enlarge through the entire 4 wk.

In rats exposed to PAs, megalocytosis of hepatocytes occurs more rapidly when a proliferative stimulus, such as hepatectomy, is applied (11). A combination of "mitotic inhibition" and a proliferative stimulus seems to be required to induce megalocytosis (8, 11). In this and previous studies (15), MCTP treatment appeared to induce a proliferative stimulus in exposed BPAEC, which was apparent by flow cytometric DNA analysis by 8 h in log phase cells. In fact, even in confluent cells, which are presumably contact-inhibited, a proliferative stimulus was apparent by 24 h. This stimulus does not appear to be a consequence of cell death, as cell loss did not appear to occur in appreciable numbers before 48 h (13, 15). However, recent evidence of apoptosis of pulmonary endothelial cells after in vivo treatment with MCT has been reported (42), and evidence of MCTP-induced apoptosis of BPAEC earlier than 48 h has been seen in our laboratory (manuscript submitted, Thomas and colleagues). MCTP-exposed (1 h, log phase) cells also became megalocytic associated with a G2 + M phase arrest through 96 h, but as the cell-cycle arrest reversed, the proliferating cells returned to normal size (data not shown). In this study, in addition to cell- cycle arrest, the cell enlargement was associated with an increased amount of cellular protein at 96 h compared with control. This is in agreement with the findings of others who showed increased DNA, RNA, and protein synthesis in pulmonary artery endothelial cells exposed to MCTP (7).

Pulmonary vascular endothelial cell dysfunction is thought to play a central role in the development of pulmonary hypertension both in people as well as in experimentally induced models (2, 43, 44). Monocrotaline and MCTP are known to cause delayed vascular leakage (16, 39) associated with endothelial cell dysfunction and cytotoxicity in vivo (40, 45), which corresponds in timing with the in vitro cell-cycle arrest and cytotoxicity (13, 15). The damage induced by MCTP in the pulmonary vascular endothelium, along with the associated proliferative inhibition, may result in an inability of the endothelium to repair the damage, consequently leading to persistent vascular remodeling. MCTP exposure of BPAEC resulted in megalocytosis associated with prolonged G2 phase cell-cycle arrest (up to 4 wk) which was similar to the prolonged mitotic inhibition and megalocytosis observed in the livers of rats exposed to MCTP (10). The G2 arrest was associated with an inactive cdc2 protein consistent with the inability to remove the inhibitory phosphate groups. The concentration and the length of exposure of the pulmonary vascular endothelium to MCTP in vivo are still not known, but could be important factors determining the nature and extent of potential cell-cycle arrest in vivo.