Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Phosphatidylcholine (PC) is the major component of pulmonary surfactant, which is produced by alveolar epithelial type II cells (1). The increased production of pulmonary surfactant during the latter part of gestation is of paramount importance for the ability of the newborn to establish regular air-breathing. Surfactant deficiency due to lung immaturity is a main factor responsible for respiratory distress syndrome (RDS) in premature neonates. Although the production of surfactant is set into gear during the latter part of gestation, the underlying molecular mechanism for the developmental increase in surfactant PC synthesis remains unknown. The major pathway for PC biosynthesis in eukaryotic cells is the CDP-choline or Kennedy pathway. Cytidine triphosphate (CTP):phosphocholine cytidylyltransferase (CCT) catalyzes a rate limiting step in the CDPcholine pathway in many cells, including type II cells (1, 2). Recent studies have shown that CCT also plays a rate-regulatory role in de novo PC synthesis by maturing type II cells (3). Specifically, PC synthesis increased in type II cells during late fetal gestation and this increase in PC production was accompanied by an increase in CCT activity and a shift of enzyme from cytosol to endoplasmic reticulum (3). The molecular mechanism responsible for this subcellular translocation of CCT from cytosol to endoplasmic reticulum (ER) is unknown. Three CCT isoforms have been identified, CCT, CCT1, and CCT2 (4, 5). The CCTs are splice variants of the same gene, differing at their C-termini (4). CCT is encoded by a separate gene and differs from CCT at the amino- and carboxy-terminus (4). CCT is a phosphoenzyme that is phosphorylated at up to 16 serine sites located within the C-terminal region (6). The soluble, relatively inactive form of CCT is much more highly phosphorylated than the membrane-bound, activated form (2, 7). The level of phosphorylation of CCT has also been correlated with the stage of the cell cycle. Using relative electrophoretic mobility as an estimate of phosphorylation, Jackowski (8) tracked the changes of phosphorylation state of CCT during the cell cycle of Bac 1.2F5 cells. The enzyme was more phosphorylated through late G₁, S, and G₂/M phase and declined precipitously as cells exit mitosis and reenter G₀/G₁. Enzyme activity varied inversely with the phosphorylation state, suggesting a role for phosphorylation regulation of CCT activity through the cell cycle. Changes of subcellular localization of CCT during the cell cycle were not studied. Recent studies with IIC9 fibroblasts showed that the wave of PC synthesis during G₀ exit is accompanied by CCT activation, CCT translocation to membranes, and redistribution from nucleus to ER (7). In contrast, DeLong and coworkers (9) reported that the nuclear localization of active CCT was independent of cell cycle position in many mammalian cells. To add to the nuclear localization controversy, we have recently reported continuous nuclear exclusion of CCT in unsynchronized pulmonary epithelial cells (10). A recent study showed an enhanced expression of CCT during S phase, which precedes increases in CCT activity and PC synthesis (11), suggesting cell-cycle-regulated transcription of CCT. Other regulatory mechanisms, such as cell-cycle-dependent CCT protein degradation, have not been investigated.

Because fetal type II cells arrest in the G₀/G₁ stage with advancing gestation (12), we investigated whether the arrest at this stage in the cell cycle may be responsible for the increase in CCT activity and surfactant PC synthesis in maturing type II cells. We used a mouse lung epithelial cell line (MLE-15), which displays some characteristics of type II cells such as expression of surfactant proteins A, B, and C, but lacks others, including lamellar body secretion (13). We observed that elevated PC synthesis in G₀/G₁ stage is associated with CCT activation, membrane affinity, and phosphorylation, but not with CCT protein synthesis, degradation, or intracellular distribution. In contrast to other cells (9), CCT was not localized to the nucleus during any stage of the cell cycle. These data suggest that CCT association with membrane lipids in G₀/G₁ arrested MLE-15 cells may lead to increased PC synthesis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and Transfection

The MLE-15 cell line was maintained in Ham's F12, insulin, transferin,  estradiol, and sodium selenite (HITES) medium (13) supplemented with 2% (vol/vol) charcoal-treated fetal bovine serum (FBS). For expression of Flag-tagged CCT protein in MLE-15 cells, the Flag (Asp-Tyr-Lys-Asp-Asp-Asp-Lys) sequence was connected to the C-terminus of a cDNA encoding full-length CCT and subcloned in the expression vector pMP6a (generous gift of Dr. M. Philip, Applied Immune Sciences, Santa Clara, CA). The resultant plasmid, CCT-Flag, was purified and cotransfected with pcDNA3 (ratio 5:1) into MLE-15 cells, using cationic liposomes and plasmid DNA at a ratio of 10:1. The neomycin-resistant clones were selected with 0.5 µg/ml G418 in HITE medium with 2% (vol:vol) charcoal-treated FBS. The colonies were assayed by Western analysis to confirm expression of CCT-Flag mutant protein.

Synchronization of Cells

MLE-15 cells were grown to semiconfluence and then synchronized using serum starvation or pharmacologic blockade with aphidicolin and nocodazole. To generate semiquiescent (G₀/G₁ phase) cells, cultures were washed twice with serum-free Hanks' balanced salt solution without calcium and magnesium (HBSS) and incubated for 24 h in serum-free HITES medium. For reversible arrest at G₁/S boundary, MLE-15 cells were washed twice with serum-free HBSS and then incubated for 24 h in HITES containing 2% (vol/vol) charcoal-treated FBS and 5 µg/ml aphidicolin. MLE-15 cells blocked with aphidicolin (G₁/S) were released into S phase following washout of the drug with HBSS and reincubation for 3 h with HITES containing 10% (vol/vol) charcoal-treated fetal calf serum (FCS). To arrest MLE-15 cells in G₂/M-phase, cultures were washed with HBSS, exposed for 24 h to HITES supplemented with 2% (vol/vol) charcoal-treated FBS and 50 ng/ml nocodazole, and mitotic cells harvested by mitotic shake-off technique. All cell-cycle positions were verified by fluorescent-activated cell sorter (FACS) analysis as previously described (12).

[³H]Choline Incorporation into PC

Synchronized MLE-15 cells in six-well cell culture plates were pulsed with 5 µCi/ml [methyl-³H]choline. After a 3-h incubation, medium was removed and cells were washed with serum-free HITE. Following trypsinization to remove the cells from the plate, cellular lipids were extracted by the method of Bligh and Dyer (14). PC was isolated from the lipid extract by thin layer chromatography on silica H plates using CHCl₃/MeOH/H₂O (65:25:4, vol/vol) as developing solution. The PC was visualized with a bromothymol blue solution, scraped into scintillation vials, and radioactivity quantified by liquid scintillation counting.

Cytidylyltransferase Assay

MLE-15 cells were grown to semiconfluence in 75-cm² tissue culture flasks. After synchronization in either G₀/G₁, G₁/S, S, or G₂/M phase cells were collected by scraping in homogenization buffer of 145 mM NaCl, 50 mM Tris-HCl (pH 7.4), 50 mM NaF, and 2.5 mM EDTA and CCT activity was assayed in the forward direction by measuring the rate of incorporation of [methyl- ¹⁴C]phosphocholine into CDPcholine (15).

CCT Promoter Analysis

MLE-15 cells were seeded at a density of 5 × 10⁵ cell/well in 12-well culture plates in HITES + 2% (vol/vol) charcoal-treated FBS and allowed to adhere overnight. The nonadherent cells were removed by aspiration and adherent cells were washed with serum-free HITES and then transfected with CCT promoter-luciferase construct (pGL3 basic [promoterless]; LUC.C5 [2068/ +38], LUC.C8 [210/+38]; generous gifts of Dr. Bakovic, University of Alberta, Edmonton [16]) and -galactosidase plasmid (ratio 10:1), using cationic liposomes and plasmid DNA at a ratio of 10:1. After 5 h, the serum-free medium was replaced with HITES containing 2% (vol/vol) charcoal-treated FBS and the cells were cultured overnight. The cells were then synchronized in G₀/G₁, G₁/S, S or G₂/M phase as described above. Preparation of cell extracts and luciferase and -galactosidase assays were performed according to Promega kits (Madison, WI).

CCT mRNA Content

Total RNA was isolated from synchronized MLE-15 cells, reverse transcribed, and amplified by 15 cycles of polymerase chain reaction (PCR) using primers against the central conserved region of all CCT isoforms (17). The reverse transcriptase (RT)- PCR product (225 bp) was further analyzed by Southern blotting using CCT cDNA labeled with ³²P (17). The RT-PCR reaction was linear up to 30 cycles of amplification and no signals were detected without the initial addition of reverse transcriptase. In additional experiments, primers were designed to specifically amplify CCT1 and 2 mRNA (4). Using agarose gel electrophoresis and ethidium bromide staining, transcripts for both CCT isoforms (256 and 586 bp, respectively) were detected in MLE-15 cells after 40 cycles of PCR. Employing the primers against the conserved region, CCT mRNA was already detectable after 20 cycles of PCR. This suggests that CCT is the most abundant isoform and that mRNA changes observed with 15 cycles of PCR can be attributed to CCT.

CCT Protein Content

MLE-15 cells stably transfected with CCT-Flag were seeded in six-well culture plates at a density of 106 cells/well in HITES medium plus 2% (vol/vol) charcoal-treated FBS and incubated overnight. Following cell-cycle treatment as decribed above, cells were twice washed with phosphate-buffered saline (PBS), pelleted by centrifugation, and lysed in 50 µl sodium dodecylsulfate (SDS)-sample buffer [10% (vol/vol) glycerol, 2% (vol/wt) SDS, 5% (vol/vol) -mercaptoethanol, 0.0025% (wt/vol) bromophenol blue, 0.06 M Tris, pH 8.0] by sonication. The protein lysates were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) and then transferred to nitrocellulose membrane. CCT-Flag was detected using mouse anti-Flag M2 monoclonal antibody (1:1,500; Sigma) and horseradish peroxidase-conjugated anti-mouse immunoglobulin G (IgG) (1:20,000) followed by enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech, Piscataway, NJ).

CCT Protein Turnover

MLE-15 cells stably transfected with CCT-Flag were grown to semiconfluence in 75-mm culture flasks and synchronized in either G₀/G₁, G₁/S, S, or G₂/M phase as described above. The cells were then washed with HBSS and cysteine- and methionine-free Eagle's minimum essential medium (MEM) and incubated for 30 min in same medium containing 200 µCi/ml [³⁵S]Translabel (³⁵S-methionine and ³⁵S-cysteine; Amersham Pharmacia Biotech). Radioactive medium was aspirated, cells were washed twice with serum-free MEM (containing 0.16 mM cysteine and 0.1 mM methionine) and the incubation in serum-free MEM (containing 0.16 mM cysteine and 0.1 mM methionine) continued. At various times (0, 1, 2, 4, 8, and 24 h), the medium was aspirated and the cells were scraped in 1 ml of 1% (vol/vol) NP-40 in PBS and sonicated. Insoluble material was removed by centrifugation at 14,000 × g for 10 min, and the supernatant was incubated with 2 µl/ml of anti-Flag M2 monoclonal antibody overnight at 4°C. Sepharose G beads (100 µl; Amersham Pharmacia Biotech) were added and the samples were incubated for 1 h at 4°C. The protein G beads were washed three times with 0.5M NaCl in PBS, 50 µl of SDS sample buffer was added to the beads and the immune complexes were dissociated by boiling for 5 min. The samples were subjected to 10% (wt/vol) SDS-PAGE and radiolabeled CCT-Flag was revealed by autoradiography and quantitated using a phosphoimager (410A and Image Quant software; Molecular Dynamics, Sunnyvale, CA).

Digitonin Permeabilization

MLE-15 cells stably transfected with CCT-Flag were grown to semiconfluence in 75-cm² culture flasks and synchronized in either G₀/G₁, G₁/S, or S phase as described above. The cells were then rinsed with ice-cold PBS, placed on a rocking platform at 4°C, and permeabilized in 10 mM Hepes, pH 7.4, 0.1 M KCl, 2 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 0.2 mg/ml digitonin. After 30 min, medium was collected and the cell ghosts were scraped in permeabilization buffer. Aliquots of both fractions were subjected to 10% SDS-PAGE and Western blotting using mouse anti-Flag M2 monoclonal antibody (1:1,500; Sigma) and horseradish peroxidase-conjugated anti-mouse IgG (1:20,000) followed by ECL (Amersham Pharmacia Biotech). In separate experiments, CCT activity was measured in both fractions.

³²P Labeling of CCT

MLE-15 cells stably transfected with CCT-Flag grown to semiconfluence in 75-cm² culture flasks were synchronized in either G₀/G₁, G₁/S, S, or G₂/M phase as described above. The cells were then washed three times in phosphate-free MEM and incubated in the same medium containing [³²P]orthophosphate (0.2 mCi/ml). After a 3-h incubation, the medium was removed and the cells were washed twice with ice-cold PBS and lysed as described above in PBS containing 1 mM EGTA, 10% (vol/vol) glycerol, 1% (vol/ vol) Triton X-100, 100 mM sodium fluoride, 10 mM pyrophosphate, 200 µM Na-orthovanadate, 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Lysates were immunoprecipitated with monoclonal Flag M2 antibody (mAb) and resolved as descibed above. ³²P-labeled CCT-Flag was detected by autoradiography.

Immunofluorescence Microscopy

MLE-15 cells stably transfected with CCT-Flag seeded on glass coverslips were synchronized in G₀/G₁, G₁/S, or S as described above. After synchronization cells were fixed in 4% (wt/vol) paraformaldehyde in PBS and permeabilized with 0.2% (vol/vol) Triton-X100 in PBS containing 0.1% (wt/vol) bovine serum albumin (BSA). Coverslips were then washed with PBS and treated successively with 5% (wt/vol) normal goat serum (NGS) and 0.1% (wt/vol) BSA in PBS (NGS/BSA) for 1 h, first antibody solution in NGS/BSA overnight at 4°C and fluoroisothiocyanate (FITC)- or rhodamine-labeled second antibody solution in NGS/ BSA for an additional hour, with excessive washing with PBS in each interval. Mouse anti-Flag M2 mAb and anti-calnexin mAb were diluted 1:100 in NGS/BSA (1:100), and used as primary antibodies. The FITC and rhodamine-labeled secondary antibodies to mouse (goat IgG) were diluted 1:100. Coverslips were mounted on microscope slides with mounting medium containing 4',6-diamidino-2-phenylindole (DAPI) and stained cells were then observed using fluorescence microscopy. No signal was observed when primary antibodies were omitted (control).

Immunogold Electron Microscopy

Synchronized MLE-15 cells transfected with CCT-Flag and grown in 75-cm² culture flasks were washed three times with PBS, collected by gently scraping, transferred to 1-ml centrifuge tubes, and pelleted by centrifugation for 5 min at 300 × g. The cells were resuspended and fixed for 1 h in 4% (vol/vol) paraformaldehyde containing 0.1% (vol/vol) glutaraldehyde. The fixed cells were then washed in PBS, transferred to beam capsules, dehydrated through an ascending ethanol series, incubated twice for 4 h in 1:1 diluted lowicryl KM4 in ethanol followed by 100% lowicryl KM4 (Canemco Inc, Montreal, PQ, Canada) and then exposed overnight to UV light at 20°C. After verification of the presence of cells by 1% (vol/vol) toluidine blue staining of 0.5-µm thick sections, 70-80 nm ultrathin sections were cut using a diamond knife with a Reichert Ultracut (Reichert Instruments, Wetzlar, Germany) and placed onto formvar-coated nickel grids. The grids were then washed with PBS and treated successively with PBS containing 0.5% (wt/vol) glycine and 1% (wt/vol) BSA followed by three treatments of 10 min each with 1% (wt/vol) BSA in PBS (BSA/PBS), first antibody solution in BSA/PBS for 1 h at room temperature and 10-nm gold-labeled second antibody solution in BSA/PBS for an additional hour, with excessive washing with BSA/PBS in each interval. Grids were washed three times with PBS and three times with distilled water. Finally, the samples were stained with uranyl acetate and lead citrate. After washing with distilled water, cells were examined using a Philips 430 electron microscope. Rabbit anti-CCT-N antiserum and mouse Flag M2 mAb were diluted 1:100 or 1:150, respectively, and used as primary antibodies. The gold-labeled secondary antibodies to rabbit (goat IgG) and mouse (goat IgG) were both diluted 1:20. Controls included no primary antibody and preimmune serum.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Synchronization of MLE-15 Cells

FACS analysis confirmed that serum starvation or pharmacologic treatment of nontransfected and stably transfected MLE-15 cells synchronized the cells in the appropriate phases of the cell cycle. Serum starvation for 24 h arrested the MLE-15 cells in G₀/G₁. Inclusion of aphidicolin in the medium resulted in a block at the G₁/S boundary. However, the aphidicolin block was reversible as cells were released in S phase after removal of drug. Nocodazole blocked the cells in G₂/M phase (Figure 1).

View larger version (81K): [in this window] [in a new window]   Figure 1.   Cell-cycle positions of MLE-15 cells. Cells were grown to semiconfluence and then synchronized using serum starvation (G0/G1) or pharmacologic blockade with 5 µg/ml aphidicolin (G1/S), 5 µg/ml aphidicolin followed by washout and 3-h pulse with 10% serum (S), and 50 ng/ml nocodazole (G2/M). Cell cycle positions were verified by cell sorter analysis.

Phosphatidylcholine Biosynthesis and CCT Activity Peak at G₀/G₁

The incorporation of [³H]choline into PC peaked in G₀/G₁, declined at the G₁/S boundary, and remained low at S and G₂/M (Figure 2A). To determine whether this pattern is accompanied by changes in CCT activity, we measured CCT activity in synchronized MLE-15 cells. As can be seen in Figure 2B, the cell cycle pattern of CCT activity paralleled that of PC. In separate experiments we immunoprecipitated CCT-Flag and determined its activity. In agreement with the activity pattern of total (endogenous and CCT-Flag) CCT activity, CCT-Flag activity was highest at G₀/G₁and declined with cell cycle progression (data not shown), suggesting that the Flag epitope does not influence CCT activity.

View larger version (38K): [in this window] [in a new window]   Figure 2.   PC synthesis and CCT activity. (A) Synchronized MLE-15 cells were pulsed for 3 h with [3H]choline and incorporation into PC was measured. (B) CCT activity was measured in homogenates of synchronized MLE-15 cells. The data represent the average ± SEM of three separate experiments performed in triplicate.

CCT Synthesis and Degradation Is Not Influenced by Cell Cycle

To test whether the increase in CCT activity in G₀/G₁ was due to an increase in CCT synthesis, we first assessed CCT transcriptional activity. CCT promoter activities were deduced from luciferase assays in transiently transfected and synchronized MLE-15 cells. To synchronize cells in S-phase, MLE-15 cells were incubated in serum-containing medium for 5 h instead of 3 h after aphidicolin block removal. A promoterless-luciferase control construct was used as negative, and the SV40 luciferase construct was employed as positive control. Both CCT constructs (2068/ +38 and 201/+38) possessed significant promoter activity (Figure 3A). The full-length promoter (2068/+38) showed a 2-fold higher luciferase expression compared with the 5'-deleted 201/+38 CCT promoter construct. The promoter activity of both constructs was not influenced by the cell cycle positions of the cells (Figure 3A). The message levels of CCT also remained constant during the cell cycle, indicating that CCT mRNA degradation was not altered during the cell cycle (Figure 3B). These findings suggest that the cell-cycle-dependent pattern of CCT activity is not due to altered transcription rate or mRNA stability of CCT. We then investigated whether the cell cycle changes in CCT activity were due to alterations in CCT protein degradation. Synchronized cells stably transfected with CCT-Flag cDNA were pulsed-labeled with ³⁵S-labeled methionine/cystine, extensively washed, chased with cold methionine/cysteine (original concentration in MEM), and CCT-Flag immunoprecipitated using monoclonal Flag antibodies at various times after start of the chase. Immunoprecipitates were resolved by SDS-PAGE and quantified by PhosphorImager (Figures 4A and 4B). Independent of cell cycle position, the rate of disappearance of CCT was ~ 2-3 h. As can be seen, the incorporation of [³⁵S]Translabel into CCT-Flag during the 30-min pulse was the same for each cell cycle position. Because CCT-Flag protein pool size remained constant during the cell cycle (Figure 4C), this observation suggests that CCT-Flag mRNA translation is not affected by the cell cycle. Thus, the pattern of change in CCT activity during the cell cycle is likely neither caused by changes in CCT expression or degradation.

View larger version (39K): [in this window] [in a new window]   Figure 3.   Promoter activity and mRNA levels of CCT. (A) MLE-15 cells were transfected with either minimal (201/+38 bp) or long-form (2,068/+38 bp) CCT-promoter coupled to a luciferase reporter gene. Cells were then synchronized and luciferase was measured. The data represents the average ± SEM of three separate experiments performed in duplicate. (B) Total RNA was isolated from synchronized cells and message levels were measured by low-cycle RT-PCR followed by Southern blotting of the PCR products with a 32P-labeled CCT probe. RNA integrity and equal loading was measured by performing low-cycle RT-PCR followed by Southern blotting of the PCR products with a 32P-labeled -actin probe The experiment was repeated three times with similar results.

View larger version (20K): [in this window] [in a new window]   Figure 4.   Degradation of CCT protein. (A) Synchronized MLE-15 cells stably transfected with CCT-Flag were pulsed for 3 h with [35S]Trans label and then chased with cold methionine/ cysteine. After lysing, CCT-Flag was immunoprecipitated with anti-Flag mAb and resolved by SDS-PAGE. (B) 35S-labeled CCT-Flag was quantified by PhosphorImage analysis. The data represents the average ± SEM of four separate experiments. (C) Equal amounts of lysates from synchronized MLE-15 cells stably transfected with CCT-Flag were subjected to SDS-PAGE and immunoblotted with anti-Flag mAb. The experiment was repeated twice with similar results.

CCT Localization Remains Unaltered during Cell Cycle

Because recent data suggest that CCT activation during Go G1 transition of IIC9 fibroblasts is accompanied by a translocation from nucleus to cytosolic ER (7), we examined the intracellular distribution CCT-Flag in MLE-15 cells during the cell cycle by laser confocal microscopy and immunogold electron microscopy. The Flag epitope was attached to the C-terminus of the CCT isoform, and to confirm that epitope tagging did not interfere with intracellular localization of CCT we compared CCT-Flag localization with that of N-terminal HA (haemagglutinin)-tagged CCT (HA-CCT) as well as endogenous CCT using CCT-specific antibodies (18).

Figure 5A shows a double indirect labeling of unsynchronized MLE-15 cells transfected with both CCT-Flag and HA-CCT. Carboxy-terminal-tagged CCT-Flag localized mainly to the cytoplasm. The cytoplasmic fluorescence had a punctate appearance and was concentrated in the perinuclear region. Some fluorescence localized to the cell membrane; however, no nuclear localization was noted. A similar cytoplasmic distribution was observed for N-terminal tagged HA-CCT using an anti-HA rhodamine-tagged antibody, demonstrating that the localization of epitope tag on CCT did not affect its distribution. We then examined the distribution of CCT in synchronized MLE-15 cells. To determine the cytoplasmic localization of CCT-Flag, we performed double labeling with anti-Flag monoclonal antibody for CCT-Flag and an antibody against calnexin, an integral membrane protein of ER (Figure 5B). The distribution of CCT-Flag was not influenced by the cell cycle position. It remained cytoplasmic and showed a distribution similar to that of calnexin, suggesting that cytoplasmic CCT colocalizes with the ER.

View larger version (49K): [in this window] [in a new window]   Figure 5.   Localization of CCT during the cell cycle. (A) Unsynchronized MLE-15 cells were cotransfected with cDNAs encoding for CCT-Flag and HA-CCT, respectively. Cells were costained with FITC-labeled anti-Flag mAb and rhodamine-labeled anti-HA antibodies. Cells were then mounted with medium containing DAPI. The images were acquired with ×100 objective and digitally zoomed to bring the single cell into the field of view. The overlay of the two images acquired with different filter sets was computer-mediated. Two different cells are shown. (B) Synchronized MLE-15 cells stably transfected with CCT-Flag were costained with FITC-labeled anti-Flag mAb and rhodamine-labeled anti-calnexin mAb. Cells were then mounted with medium containing DAPI. The images were digitally acquired with ×40 objective. The experiment was repeated twice with similar results.

To further identify the cytoplasmic structures with which CCT associates during the cell cycle, we performed immunogold transmission electron microscopy. Each of the three cell cycle positions (G₀/G₁, G₁/S, and S) was examined in three separate experiments. CCT-Flag was immunolocalized with monoclonal Flag antibody followed by goat anti-mouse IgG conjugated to 10-nm gold particles. The gold particles were seen extensively throughout the cytoplasm (Figure 6). Some of the cytoplasmic gold was in close proximity to the ER. There were no strong congregations of gold observed in these samples. The cytoplasmic-localized gold had a rather wide distribution. Gold particles were also associated with both the nuclear envelope and the plasma membrane, but were rarely localized at mitochondria or within the nucleus. There was no apparent difference in CCT distribution with cell cycle position. To confirm consistent localization between CCT-Flag and native CCT, cells were also immunostained with polyclonal CCT antibodies raised against the amino terminus (CCT-N), which differs significantly from CCT-1 and 2 (6) and therefore recognizes selectively CCT-. The CCT-N antibody, when tested with the synchronized MLE-15 cells, yielded results similar to those noted with anti-Flag mAb (Figure 6). Thus, there is no obvious distribution difference between CCT-Flag and native CCT. As controls, all experiments included omission of primary antibody, which consistently yielded extremely low occurrences of gold particles. Preimmune serum was also tested on each of the synchronized cell samples. Gold particles were infrequent and displayed no obvious localization similarity with the experimental samples (data not shown).

View larger version (114K): [in this window] [in a new window]   Figure 6.   Transmission electron micrographs of immunogold-localized CTT. MLE-15 cells stably transfected with CCT-Flag were synchronized in G0/G1, G1/S and S phase, fixed and processed for TEM after staining with either gold-labeled anti-Flag mAb (Flag) or anti-CCT-N (CCT-N). Circles identify gold particles. Left panel, the distributions of gold particles from anti-Flag treated samples showed a constant cytoplasmic distribution for all three synchronized cell cycle samples. Some nuclear envelope and plasma membrane labeling was noted. No gold particles were observed in the nucleus. Right panel, high magnification of cells stained with polyclonal anti- CCT-N feature similar CCT localizations to that seen with anti-Flag mAb. The experiment was repeated three times with similar results. Scale bar = 100 nm.

We also examined the distribution of CCT using digitonin permeabilization. Digitonin permeabilization releases soluble proteins while organelle-trapped and membrane-bound proteins remain in the cell ghost fraction. Independent of the cell cycle position of the cells, digitonin released most of the CCT-Flag, whereas only a minor amount remained associated with the ghost (membrane) fraction (Figure 7). No significant changes in intracellular distribution of CCT-Flag mass were noted when cells were synchronized in G₀, G₁/S, or S phase of the cell cycle. CCT activity measurements confirmed that most of CCT is released from the cells (Table 1). Based upon the ratio of ghost/lysate total activity of CCT, the amount of organelle/membrane-associated CCT was greater in G₀/G₁ cells compared with cells synchronized in G₁/S and S phase of the cell cycle. Although opposing the Western blot analysis, the activity data suggest an increased organelle/membrane affinity of CCT in cells arrested in the G₀/G₁ phase. These cell cycle changes in organelle/membrane affinity are compatible with the observed changes in PC synthesis (Figure 2A).

View larger version (39K): [in this window] [in a new window]   Figure 7.   Distribution of CCT between particulate and cytosol. MLE-15 cells stably transfected with CCT-Flag were grown to semiconfluence in 75-cm2 cell culture plates. After synchronization in either G0/G1, G1/S or S phase, cells were digitonin permeabilized for 30 min at 4°C. Equal amount of protein from both ghost (particulate) and lysate fractions were subjected to SDS-PAGE and immunoblotted with anti-Flag mAb. The experiment was repeated twice with similar results.

CCT Phosphorylation in S Phase Correlates with Decreased Activity

Several studies have suggested that CCT interaction with membrane lipids is influenced by the phosphorylation status of the enzyme (2, 19). In Bac 1.2F5 cells, the phosphorylation status of CCT fluctuated during the cell cycle in concert with CCT activity (8). We labeled synchronized MLE-15 cells stably transfected with CCT-Flag with [³²P]orthophosphate for 3 h and immunoprecipitated CCT-Flag. The phosphorylation of immunoprecipitated CCT-Flag was minimal in G₀/G₁, increased at G₁/S, and peaked in S before declining in G₂/M (Figure 8). In a separate experiment we found a similar pattern of ³²P-labeled CCT when the cells were incubated for 10 h with ³²P-label, suggesting that the cell cycle-related changes in ³²P-labeled CCT are likely not due to differences in ³²P-labeling of ATP pools. As discussed above, CCT-Flag mass remains constant during the cell cycle (see Figure 4C) and, therefore, differences in CCT pool sizes are also not responsible for observed changes in ³²P-labeling of CCT. Thus, the cell cycle pattern of CCT activity is inversely related to CCT phosphorylation.

View larger version (46K): [in this window] [in a new window]   Figure 8.   Phosphorylation of CCT. Synchronized MLE-15 cells stably transfected with CCT-Flag were labeled for 3 h with [32P]orthophosphate, lysed, and CCT-Flag immunoprecipitated with Flag antibody. 32P-labeled CCT-Flag was detected by autoradiography. The experiment was repeated twice with similar results.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The current study clearly suggests that cell cycle-associated alterations in CCT activity are not due to changes in the synthesis and/or degradation of CCT. We showed that CTT promoter activity was not altered by cell cycle position, suggesting cell cycle-independent CCT gene transcription. The CTT mRNA levels remained constant during the cell cycle, suggesting no cell cycle-dependent changes in degradation rates of CTT mRNA. Pulse-labeling experiments and Western blotting suggested that CCT mRNA translation was cell cycle-independent. In addition, these experiments revealed that the rate of CCT protein turnover was cell cycle-independent. The localization of CCT in MLE-15 cells was predominantly cytoplasmic and did not change during the cell cycle. The cell cycle-associated alterations in CCT activity were, however, accompanied by changes in CCT phosphorylation and organelle/membrane association.

Similar to previous studies (7, 8), we showed that PC biosynthesis in MLE-15 cells is regulated in a cell cycle- dependent manner. The PC synthetic rate and CCT activity peaked at the G₀/G₁ position of the cell cycle and declined during G₀/G₁ transition to S phase. Expression of CCT was not altered during this transition, in agreement with studies using IIC9 fibroblasts (7). In contrast, based on promoter activity experiments and RNA blotting, Golfman and coworkers (20) reported recently enhanced expression of CCT during S phase in C3H10T1/2 fibroblasts. These data suggest that cell cycle-regulated transcription of CCT is cell type-specific. In line with our current observation, we have previously reported that the transcriptional rate for CCT, measured by nuclear run-on transcription assay, does not alter when fetal type II cells arrest in G₀/G₁ phase with advancing gestation (12, 17). In the present study, the absence of cell cycle associated changes in the synthetic and degradation pathways of CCT corroborated the finding of constant CCT protein levels. The rapid degradation of CCT (t_1/2 = 1.5 h) was somewhat unexpected. CCT activity is regulated by several mechanisms, including enzyme-membrane interactions and phosphorylation/dephosphorylation (2, 19). Thus, regulation of CCT activity by protein degradation seems to be redundant. On the other hand, an ubiquitous protein such as CCT, whose activity level varies with cell type, cell cycle, and development, may require multiple layers of regulation. Degradation of most short-lived regulatory cellular proteins is mediated by the ubiquitin-proteasome pathway (21) and a recent report suggests that CCT protein stability is controlled via this pathway (22). The cytoplasmic localization of CCT during all phases of the cell cycle contrasts with other published studies (7, 9). We showed that the CCT localization was not due to addition of the Flag epitope to the C-terminus because a similar cytoplasmic localization was observed with anti-CCT-N antibodies. Furthermore, a cytoplasmic localization was also noted with N-terminal HA-tagged CCT, implying that the localization of the epitope tag did not affect the subcellular distribution. We can only speculate that CCT nuclear localization is not universal to all cell types, as has been suggested (7). Similarly, serum-induced translocation of CCT from nucleus to cytoplasm during G₀ to S transition is specific to IIC9 fibroblasts (7). It is possible that the indirect detection of CCT-Flag in MLE-15 cells using Flag antibodies is not specific to CCT. However, no obvious signal was observed when we applied the Flag antibodies to nontransfected MLE-15 cells. In addition, we have recently reported nuclear exclusion of CCT in asynchronous pulmonary epithelial cells, including MLE-15 cells (10).

The major site of PC synthesis is the ER (3, 23, 24). The colocalization of calnexin, an ER marker, and CCT suggests that CCT in MLE-15 cells localizes to this membrane-rich environment. In addition to ER, immunogold EM revealed that CCT also localizes to the plasma and nuclear membrane. The membranous localization of CCT is consistent with the idea of CCT activity regulation by reversible enzyme-membrane interactions (23, 25, 26). The enzyme exists in an inactive soluble form and an active membrane-bound form. When we assessed the distribution of CCT using digitonin permeabilization, only a minor amount of CCT-Flag remained associated with the membrane fraction. Thus, the majority of CCT is loosely associated with the membranous fraction and is mostly present in an inactive form. However, its presence in a membranous region permits fine control of PC synthesis when required. Only a few CCT molecules need to bind to the membrane to stimulate PC synthesis. Although we observed no significant cell cycle-dependent changes in intracellular distribution of CCT-Flag mass, it is possible that the digitonin release assay is not sensitive enough to detect small changes in organelle/membrane-associated CCT molecules. Indeed, our observation that a greater amount of the CCT activity was associated with the organelle/membrane fraction of G₀/G₁-arrested MLE-15 cells than cells synchronized in G₁/S or S phase corroborates this possibility. The CCT activity data from the digitonin permeabilization experiments are consistent with a redistribution of organelle/membrane-bound CCT to cytosol during cell cycle progression. Northwood and colleagues (7) reported also an intracellular redistribution of CCT during G₀ G₁ transition in IIC9 fibroblasts, but the redistribution occurred between nucleus and ER. As mentioned above, independent of cell cycle position CCT was a cytoplasmic protein in MLE-15 cells.

Reversible translocation of CCT between membrane and cytosol may be controlled by CCT phosphorylation/ dephosphorylation (2, 19). In line with previous studies (7), CCT activity and phosphorylation are inversely related during the cell cycle. Our data suggest that the decrease in CCT activity during G₀/G₁ S transition is due to release of membrane-bound CCT. Whether phosphorylation of CCT is primary or secondary to membrane release is unknown. In CCT activation experiments, dephosphorylation has been shown to be secondary of membrane binding (27, 28). Which of the numerous phosphorylation sites of CCT are getting phosphorylated during the cell cycle is unknown. Seven of the 16 C-terminal phosphorylation sites consist of Ser-Pro, suggesting that proline-directed kinases may be critical for the phosphorylation of CCT during the cell cycle. Indeed, proline-directed protein kinases, such as casein kinase II, glycogen synthetase kinase-3, cyclin-dependent kinase 2, protein kinase C, have been shown to phosphorylate pure CCT (29, 30). However, these in vitro phosphorylations did not change the activity of the enzyme. The precise determination of enzymes involved in CCT phosphorylation, as well as dephosphorylation and the functional importance of this process during the cell cycle, await more detailed genetic experiments.

Previously, we have reported that fetal type II cells arrest in the G₀/G₁ phase of the cell cycle with advancing gestation (12). This observation and the present finding that CCT activity peaks in this phase is consistent with the idea that the increase in surfactant PC synthesis at late gestation is at least in part due to dephosphorylation-induced membrane binding of CCT. A definitive determination of such mechanism will require transgenic animals that express CCT devoid of phosphorylation sites or that overexpress CCT phosphorylation domains, thereby interfering with endogenous CCT phosphorylation.