INTRODUCTION

Protein kinase C (PKC)¹ is a family of serine/threonine kinases which play a central role in signal transduction and have been widely implicated in control of cell growth, differentiation, and transformation (1-4). The primary physiological activator of PKC is diacylglycerol (DAG), which is transiently generated by agonist-induced hydrolysis of phosphoinositides and other membrane phospholipids (1, 5). PKC activation, often accompanied by increased association of the enzyme with cellular membranes and/or cytoskeletal elements (6-9), initiates a signaling cascade leading to alterations in gene expression and modulation of a variety of cellular functions. Initial interest in PKC stemmed from its identification as the major cellular receptor for tumor-promoting phorbol esters (10), suggesting a role in stimulation of cell growth and transformation. While studies in a number of tissues have implicated PKC in positive control of cell growth (11-15) and transformation (16, 17), accumulating evidence also points to its involvement in cell growth inhibition and differentiation (14, 15, 18-21). A key to understanding these diverse responses is the observation that PKC represents a multigene family of 11 closely related enzymes with varying structures and enzymological characteristics (1, 2): the conventional PKC isozymes , I, II, and , which require Ca²⁺ for activity and respond to phorbol esters; the novel isozymes , , , , and µ, which do not require Ca²⁺, and the atypical isoforms  and , which neither require Ca²⁺ nor respond to phorbol esters. Individual PKC isozymes also exhibit varying substrate specificity, tissue distribution, and subcellular localization (1, 2, 22, 23); these differences, together with the varied consequences of PKC activation in the same cell (e.g. Refs. 24-26), the expression of more than one isozyme in most cell types (1, 2, 21, 27), and conservation of the isozymes in higher organisms (2), argue that individual isozymes play specific, specialized roles in cell signaling. However, understanding of the biological functions of individual isozymes and of the molecular regulatory pathways in which they participate remains limited.

The growth-regulatory consequences of PKC activation suggest a link between PKC signaling and control of the cell cycle machinery. Activation of PKC has been shown to result in alterations in cell cycle progression in either stimulatory or inhibitory directions in several systems (14, 19, 20, 28-30). Moreover, limited evidence supports a role for individual isozymes in modulation of major cell cycle transitions. For example, PKC / (29) and  (31) have been implicated in control of the G₁  S transition in vascular smooth muscle cells and NIH 3T3 fibroblasts, respectively, while PKC II has been shown to play a requisite role in progression from G₂ into M phase in HL-60 cells (30), and PKC  has been associated with control of M phase in Chinese hamster ovary cells (32). Orderly progression through the cell cycle is now known to be dependent on the coordinated interaction between key cell cycle regulatory molecules including cyclins, cyclin-dependent kinases (cdks), and cdk inhibitory proteins such as p21^waf1/cip1 and p27^kip1 (33-35). Together, these molecules control the activity of a number of important substrates including the retinoblastoma gene product (Rb), a critical regulator of the G₁  S transition (36, 37). The interplay between specific PKC isozyme signaling and control of the cell cycle machinery is an important question that remains to be addressed at the molecular level.

Previous studies in this laboratory have identified the mammalian intestinal epithelium as a useful model system in which to define the physiological role(s) of individual PKC isozyme(s) in control of cell growth and cell cycle progression (27). This dynamic and complex tissue system undergoes continuous and rapid renewal (38); its polarized architecture, with well-defined regions of cell proliferation, growth arrest, and differentiation, allows correlation of the expression and activation of PKC isozymes with specific stages of cell development (27). Using a combined biochemical and morphological approach to detect changes in PKC isozyme expression and activation status at the individual cell level, our previous studies revealed that multiple PKC isozymes are expressed in the intestinal epithelium and that they are differentially regulated with respect to cell growth and differentiation (27). Of particular note was the finding that four members of the PKC family, PKC , II, , and , undergo marked changes in expression and subcellular distribution indicative of activation precisely at the point in the mid-crypt at which cells cease dividing, suggesting that one or more of these molecules are involved in negative growth-regulatory signaling pathways in this tissue.

In the present study, we have extended the characterization of PKC isozyme expression and activation in the rat intestinal epithelium in situ and, based on this analysis, have explored the role of PKC isozyme(s) in control of cell growth and cell cycle progression using the non-transformed IEC-18 immature crypt cell line as a complementary in vitro model system. In keeping with the finding that activation of specific isozymes coincides with cell growth arrest in the intestinal crypt, PKC agonists were found to block IEC-18 cell cycle progression in G₁ and delay transit through S and/or G₂/M phases. Analysis of the mechanism(s) underlying PKC-mediated regulation of the G₁  S transition revealed that PKC agonist-induced G₁ arrest is accompanied by Rb hypophosphorylation and rapid induction of the Cip/Kip family cdk inhibitors p21^waf1/cip1 and p27^kip1. Taking advantage of differential down-modulation of individual PKC isozymes under different PKC agonist treatment conditions, PKC  was identified as sufficient to mediate these negative growth-regulatory effects. Taken together, the data presented provide evidence for control of the cell cycle machinery in the intestinal epithelium by PKC-mediated pathways and demonstrate a mechanism for the integration of environmental anti-mitogenic stimuli with regulation of cell division in this tissue. Furthermore, they indicate that PKC , in particular, is linked to a pathway which negatively modulates cell growth and cell cycle progression in intestinal epithelial cells both in vitro and in situ.

EXPERIMENTAL PROCEDURES

Monoclonal antibody specific for the catalytic domain of PKC  was obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Polyclonal rabbit anti-PKC , II, , , and  were purchased from Life Technologies, Inc. and Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies specific for PKC , , , and  were purchased from Transduction Laboratories, Inc. (Lexington, KY) and Santa Cruz Biotechnology. The antibodies used in this study have been extensively characterized for the absence of cross-reactivity with other PKC isozymes (27). To ensure reliability of the data obtained, at least two antibodies were used for each isozyme studied. Polyclonal antibody specificity was confirmed by competition assays with the appropriate antigenic peptide as described previously (27). Polyclonal anti-Rb and anti-p21^waf1/cip1 antibodies were obtained from Santa Cruz Biotechnology. Monoclonal antibody specific for p27^kip1 was obtained from Transduction Laboratories. Horseradish peroxidase-conjugated rat anti-mouse and goat anti-rabbit secondary antibodies were purchased from Boehringer Mannheim.

Immunofluorescence analysis of PKC isozyme expression and subcellular distribution in rat intestinal epithelial tissue was performed on 4-6-µm cryosections as described previously (27).

Sequential release of epithelial cell populations from rat small intestine was performed as described previously (27, 39). Briefly, populations of cells at different developmental stages were isolated using timed incubations in a calcium-chelating buffer. Cells were washed in PBS, and fractions were pooled into crypt (proliferating), lower villus (differentiating), and upper villus (functional) populations. Membranes were prepared from isolated crypt or villus cells as described previously (27).

The IEC-18 cell line (ATCC CRL-1589) is an immature, non-transformed cell line derived from rat ileal epithelium which maintains many characteristics of proliferating crypt cells (27, 40, 41). IEC-18 cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM) with 4 mM glutamine, 10 µg/ml insulin, and 5% fetal calf serum (FCS).

PKC isozymes were activated in IEC-18 cells by treatment with either 100 nM phorbol 12-myristate 13-acetate (PMA; Sigma), 100 nM phorbol 12,13-dibutyrate (PDBu; Sigma), or 20 µg/ml 1,2-dioctanoyl-sn-glycerol (DiC₈; Sigma) for various times. PMA and PDBu were dissolved in ethanol, with a final vehicle concentration in the medium of <0.1%; DiC₈ was dissolved in acetone, with a final vehicle concentration of <0.2%. Control cells were treated with the appropriate vehicle alone. PMA was administered either as a 15-min pulse, followed by two washes in PBS and a return to fresh medium for various times, or in continuous exposure. DiC₈ treatments required addition of fresh drug every 6 h, as DiC₈ is rapidly metabolized by the cell (42, 43). Depletion of PKC , , and  from IEC-18 cells was accomplished by treatment with 1 µM PDBu for 24 h.

IEC-18 cells were synchronized in G₀/G₁ by serum deprivation. Briefly, subconfluent cells growing in complete medium were washed twice in DMEM with no FCS and incubated with DMEM containing 0.5% FCS and 4 mM glutamine (no insulin) for 72 h. More than 90% of cells were arrested in G₀/G₁ by this method, as determined by flow cytometric analysis. Cells were released from G₀/G₁ arrest by addition of complete growth medium containing 5% FCS and 10 µg/ml insulin.

IEC-18 cells were synchronized in G₁/S phase by incubation with 1 µg/ml aphidicolin for 24 h. Cells were released from G₁/S arrest by removal of aphidicolin and incubation in complete growth medium.

Subconfluent IEC-18 cells were briefly washed in PBS, harvested by trypsinization, fixed in 70% ethanol, and treated with 0.04 mg/ml RNase A (Sigma) in 20 mM Tris, pH 7.5, 250 mM sucrose, 5 mM MgCl₂, and 0.37% Nonidet P-40 (Sigma). Cellular DNA was stained with 25 µg/ml propidium iodide (Sigma) in 0.05% sodium citrate and quantified by flow cytometry. Cell cycle analysis was performed using the Winlist and Modfit programs (Verity Software House, Topsham, ME).

IEC-18 cells were partitioned into soluble (cytosolic) and particulate fractions essentially as described previously (27). Briefly, cells were washed twice in cold PBS and scraped in an extraction buffer containing 20 mM Tris, pH 7.5, 2 mM EGTA, 2 mM EDTA, 0.5 mg/ml digitonin, 10 mM NaF, 4 mM phenylmethylsulfonyl fluoride, 2 mM benzamidine, 10 µg/ml leupeptin, and 10 µg/ml aprotinin (digitonin buffer). Digitonin-soluble (cytosolic) and -insoluble (particulate) fractions were separated by ultracentrifugation at 100,000 × g for 40 min at 4 °C. The cytosolic protein in the supernatant was precipitated with 10% trichloroacetic acid for 10 min on ice, pelleted, washed in acetone, solubilized in 100 mM NaOH, and neutralized by the addition of 100 mM HCl. The particulate pellet was incubated on ice for 30 min in digitonin buffer containing 1% Triton X-100 (Triton buffer). The membrane sample was then cleared by centrifugation (10,000 × g) for 30 min at 4 °C. Cytosolic and membrane fractions were boiled in Laemmli sample buffer (44) for 5 min before being subjected to SDS-PAGE and Western blot analysis.

Subconfluent IEC-18 cells were rapidly washed twice in PBS at 4 °C and incubated for 5-15 min on ice in 20 mM Tris, pH 7.6, 120 mM NaCl, 100 mM NaF, 200 µM Na₃V0₄, 4 mM phenylmethylsulfonyl fluoride, 2 mM benzamidine, 10 µg/ml leupeptin, 10 µg/ml aprotinin, and 0.5% Nonidet P-40 (Sigma). Cell extracts were cleared by a 30-min centrifugation (10,000 × g) at 4 °C and boiled in Laemmli sample buffer (44) for 5 min before being subjected to SDS-PAGE and Western blot analysis.

Cell lysates (30 µg) were subjected to SDS-PAGE (44), using 7.5% (Rb), 10% (PKC isozymes), or 15% (p21^waf1/cip1, p27^kip1) polyacrylamide gels. Protein was electrophoretically transferred to nitrocellulose membrane, and membranes were blocked in Tris-buffered saline (TBS; 20 mM Tris-HCl, 137 mM NaCl, pH 7.6) containing 5% non-fat dried milk (TBS/milk) for 30 min at 37 °C. Membranes were incubated for 2 h at room temperature or overnight at 4 °C with primary antibody in TBS/milk with 0.1% Tween 20 (TBSt/milk; Sigma), followed by six 5-min washes in TBSt/milk. Blots were then incubated for 1 h at room temperature in secondary HRP-conjugated antibody in TBSt/milk, followed by six 5-min washes in TBSt/milk and three 5-min washes in TBS. Bound horseradish peroxidase was then detected using the SuperSignal CL system (Pierce). Specificity of the antibodies was determined using the relevant antigenic peptide in competition experiments as recommended by the manufacturer. All data presented are representative of at least three independent experiments.

RESULTS

IEC-18 cells are non-transformed, immature intestinal epithelial cells derived from rat ileum, which retain many phenotypic characteristics of proliferating intestinal crypt cells (40, 41). To evaluate this cell line as an in vitro model system in which to examine directly the involvement of PKC isozymes in intestinal epithelial cell growth regulation, we compared PKC isozyme profiles in proliferating crypt cells in situ and cultured IEC-18 cells. Previous morphological and biochemical analysis of the expression and subcellular distribution of individual PKC isozymes in the rat intestinal epithelium revealed that PKC , II, , , and , but not PKC  or I, are present in cells of the crypt-villus unit and are differentially regulated with respect to cell growth and differentiation (27). In situ immunofluorescence analysis demonstrated that marked changes in the expression and subcellular distribution of several PKC isozymes coincides precisely with cell growth arrest in the mid-crypt region. Representative data are shown for PKC  in Fig. 1A: proliferating lower crypt cells exhibit diffuse cytosolic immunostaining for this isozyme (arrowhead), presumably reflecting its presence in an inactive conformation. At cell position 14-18 from the crypt base (i.e. the mid-crypt), the region in which cells cease division and commit to differentiation (38), levels of PKC  expression markedly increase, and PKC  immunostaining becomes clearly detectable in the lateral membrane domains and in the developing brush-border microvilli. Similar changes, confirmed by biochemical analysis (27), were observed for PKC II, , and  (27), suggesting that one or more of these PKC isozymes play a role in signaling pathway(s) related to negative control of cell growth in the intestinal epithelium in situ.

Further analysis has revealed the presence of three additional isozymes, PKC , , and , in this tissue (Fig. 1B). Based on the premise that association of PKC isozymes with the particulate fraction may be indicative of activation (6), the pattern of regulation of these PKC isozymes along the crypt-to-villus axis was examined by comparing the level of membrane-associated expression in isolated proliferating (crypt), differentiating (lower villus), and functional (upper villus) intestinal epithelial cells. As shown in Fig. 1B, crypt cell membranes express low levels of PKC , and only trace amounts of PKC  and . Since PKC  and  were also essentially undetectable in whole cell extracts of intestinal crypt cells (data not shown), these isozymes appear to be absent from proliferating cells. Thus, taken together with our previous findings (27), these data demonstrate that in situ proliferating cells of the intestinal crypts express PKC , II, , , , and . Interestingly, in contrast to cells of the proliferation zone, differentiating and functional cells of the villus express readily detectable levels of membrane-associated PKC , , and  (Fig. 1B), indicating that increased expression/activation of these isozymes is associated with post-mitotic events in intestinal epithelial cells in situ. Immunofluorescence localization of PKC  revealed that this isozyme is expressed mainly in cells of the mid- to upper villus, in association with the endoplasmic reticulum (data not shown), suggesting a role in enterocyte mature function (45). Due to a lack of suitable reagents, we were unable to perform morphological analysis of the distribution of PKC  and  in tissue enterocytes.

As shown in Fig. 1C, Western blot analysis revealed that IEC-18 cells express the same panel of PKC isozymes as tissue crypt cells. While PKC , II, , , , and  were readily detectable in IEC-18 whole cell extracts, PKC  and , which are expressed only in villus cell populations in situ, were found to be absent from these cells. In the present study, the IEC-18 cell line was used as an in vitro model system in which to address the hypothesis that one or more PKC isozymes play a role in negative regulation of cell growth and cell cycle progression in the intestinal epithelium in situ.

To determine the effects of PKC activation on IEC-18 cell cycle progression, asynchronously growing IEC-18 cell populations were treated with a panel of PKC agonists to activate PKC isozyme(s) directly, and perturbations in cell cycle distribution were determined by flow cytometric analysis after various times over a 24-h period. Three PKC agonists were used for this analysis: the phorbol esters PMA (100 nM) and PDBu (100 nM) and the DAG analogue, DiC₈ (20 µg/ml). Phorbol esters are potent activators of most members of the PKC family, with the exception of PKC  and  (1, 10, 22), and can be used to mimic the effects of DAG and to bypass normal agonist-mediated control of the enzyme. The use of two phorbol esters in this study was based on evidence that different members of this class of agents can induce distinct cellular responses (22). The membrane-permeant DAG analogue DiC₈, a less potent but more physiological PKC agonist, was used to confirm the involvement of a PKC-mediated pathway in observed phorbol ester cell cycle-specific effects. PMA was administered either as a 15-min pulse or in continuous exposure, while PDBu and DiC₈ were added only in continuous exposure.

As shown in Fig. 2A.i, treatment of IEC-18 cells with PMA, added as a 15-min pulse or in continuous exposure, resulted in a marked reduction (70-75%, n = 6) in the relative number of cells in S phase by 6 h. This decrease was accompanied by a 20-30% (n = 6) increase in the relative number of cells in G₁, indicating that PKC activation negatively regulates transit through this phase of the cell cycle. Although quantitative analysis revealed no significant change in cell cycle distribution at 3 h following addition of PMA, the number of cells with DNA content corresponding to early S phase was markedly reduced by this time (arrows), indicating that entry into S phase had been restricted; thus, the G₁  S block was evident by 3 h. Pulse or continuous treatment of IEC-18 cells with PMA for 6 h also resulted in a 70-80% (n = 6) increase in the proportion of cells with apparent 4n DNA content, consistent with an accumulation of cells in G₂/M and/or possibly late S phase of the cell cycle. This increase could reflect a PMA-mediated delay in transit of IEC-18 cells through G₂/M and/or delayed exit from S phase. Inhibition of cell cycle progression was transient; reversal of the G₂/M and/or S phase effects occurred prior to release of the G₁/S block, resulting in accumulation of the majority (>80%) of the population in G₁ by 12 h. The G₁/S block in cell cycle progression was consistently released 12-14 h following addition of PMA. By 24 h, cells could be seen in all phases of the cell cycle. Treatment of IEC-18 cells with the phorbol ester PDBu produced similar results (Fig. 2A.ii); an accumulation of cells in G₁ phase as well as in G₂/M and/or late S phase of the cell cycle was seen by 6 h of continuous exposure to this PKC agonist. However, release of the G₁ block was evident by 9-10 h, indicating that the kinetics of the effect were slightly accelerated in PDBu-treated cells.

PKC agonists inhibit cell cycle progression in IEC-18 cells. A, subconfluent cultures of asynchronously growing IEC-18 cells were treated with PKC agonists for various times over a 24-h period, and cell cycle distribution was determined by flow cytometric analysis of DNA content in propidium iodide-stained cells as described under "Experimental Procedures." i, IEC-18 cells were treated with vehicle alone (Untreated) or 100 nM PMA (either as a 15-min pulse or in continuous exposure), and relative DNA content was determined after 3, 6, 12, 14, and 24 h. ii, IEC-18 cells were treated continuously with 100 nM PDBu, and cell cycle distribution was analyzed at 6, 9, 10, and 24 h. iii, IEC-18 cells were treated with 20 µg/ml DiC₈, and cell cycle distribution was determined at 6 and 24 h. Numbers indicate percentage of cells in G₀/G₁, S, and G₂/M phases. Arrows indicate initial restriction of entry into S phase at 3 h. Data are representative of three or more independent experiments. B, subconfluent cultures of IEC-18 cells synchronized in G₀/G₁ phase by serum withdrawal for 72 h were released from growth arrest by addition of 5% FCS (time 0) and treated with 100 nM PMA or 20 µg/ml DiC₈ 8 h later; cell cycle distribution was determined at 14, 18, and 22 h after serum stimulation (i.e. 6, 10, and 14 h of PKC agonist treatment). Data are representative of three independent experiments. C, subconfluent cultures of IEC-18 cells synchronized in G₁/S phase with aphidicolin were released from cell growth arrest by removal of the drug and treated with PKC agonists 4 h later; cell cycle distribution was determined at 6 and 8 h after removal of aphidicolin (i.e. 2 and 4 h of PKC agonist treatment). Data are representative of three independent experiments.

In contrast to the transient cell cycle arrest resulting from phorbol ester treatment, DiC₈ produced a sustained inhibition of cell cycle progression in IEC-18 cells over the 24-h experimental period (Fig. 2A.iii). As with phorbol ester treatment, a significant reduction (~60%, n = 4) in the relative number of cells in S phase was consistently evident by 6 h following administration of the DAG analogue; this reduction was accompanied by a 20-30% (n = 4) increase in the proportion of cells in G₁ and a 30-40% (n = 4) increase in the proportion of cells with apparent 4n DNA content. The G₁ effects were maintained over the 24-h period, as indicated by a sustained >30% (n = 3) increase in the proportion of G₁ cells relative to control cell populations. The persistence of a significant population of cells with apparent 4n DNA content at the 24-h time point despite a sustained G₁/S block suggests that a delay in G₂/M progression and/or exit from S phase is also sustained over the 24-h experimental period. Taken together, these data demonstrate that activation of PKC by three different PKC agonists results in inhibition of IEC-18 cell cycle progression at two (or more) points in the cell cycle: one in G₁ and the other in G₂/M and/or possibly S phase. The duration of these effects was found to differ with the PKC agonist used; while phorbol esters produced transient cell cycle arrest in IEC-cells, DiC₈ treatment resulted in a sustained inhibition of cell cycle progression over the 24-h experimental period.

To determine if PKC-mediated cell cycle arrest is associated with differentiation of IEC-18 cells into absorptive enterocytes, we examined PKC agonist-treated cells for expression of the differentiation marker, alkaline phosphatase (39), using an in situ enzyme cytochemical approach. Neither control nor PKC agonist-treated cells were found to express the differentiation marker (data not shown). Furthermore, PKC agonists did not induce the expression of PKC  or , isozymes associated with differentiated cells of the villus in situ (data not shown). Taken together, the data indicate that PKC-mediated withdrawal from the cell cycle is not associated with differentiation in this system.

To confirm and further investigate PKC-mediated regulation of cell cycle progression in IEC-18 cells, experiments were conducted using synchronized cell populations. IEC-18 cells were synchronized in G₀/G₁ by serum deprivation for 72 h and released from growth arrest by addition of 5% FCS (time 0). Entry of cells into S phase was evident by 12-14 h following serum stimulation, and by 22 h, cells had progressed through G₂/M and begun to re-enter G₀/G₁ (Fig. 2B). Treatment of IEC-18 cells with either PMA or DiC₈ in mid-G₁ (6-8 h following serum stimulation) resulted in inhibition of cell cycle progression into S phase. PMA treatment resulted in transient G₁ arrest, with release kinetics similar to those seen in unsynchronized cells; i.e. arrest was maintained for ~12 h following addition of phorbol ester. DiC₈ treatment, on the other hand, resulted in sustained inhibition of cell cycle progression into S phase. At 22 h after serum stimulation (14 h after addition of DiC₈), DiC₈-treated cells were still accumulated in G₁, in sharp contrast to both PMA-treated and control cells.

To examine PKC-mediated effects on IEC-18 cell cycle progression through S and G₂/M phases, cells were synchronized in G₁/S phase by treatment with 1 µg/ml aphidicolin for 24 h. Following removal of aphidicolin, cells progressed through S and G₂/M phases, returning to G₁ by 6 h (Fig. 2C). Consistent with the increase in the proportion of cells with apparent 4n DNA content observed following treatment of asynchronously growing IEC-18 cells with PKC agonists (Fig. 2, A.i, A.ii, and A.iii), the addition of PMA or DiC₈ 4 h following removal of aphidicolin markedly retarded progression of IEC-18 cells into G₁ phase, likely reflecting delayed transit through G₂/M and/or S phase of the cell cycle. Progression of cells into G₁ phase was delayed for at least 6 h following addition of either PKC agonist (data not shown).

To investigate the mechanism(s) underlying phorbol ester- and DiC₈-mediated inhibition of cell cycle progression in IEC-18 cells and to examine the basis for the differential responses to these agents, the effects of each agonist on PKC isozyme expression and subcellular distribution were examined at various times during the 24-h treatment period. In the absence of known specific in vivo substrates for the individual PKC isozymes examined in this system, PKC isozyme translocation (i.e. association with the particulate fraction) and down-regulation were used as measures of agonist-induced isozyme-specific effects and possible indicators of PKC isozyme activation (6, 22). IEC-18 cells were treated with 100 nM PMA (administered as a 15-min pulse or in continuous exposure), 100 nM PDBu (in continuous exposure), or 20 µg/ml DiC₈ (in continuous exposure), harvested at various times (i.e. 15 min, 2 h, 6 h, 12 h, or 24 h), and partitioned into soluble and particulate fractions; PKC isozyme levels in each fraction were determined by Western blot analysis. Treatment with PMA or PDBu resulted in essentially complete translocation of PKC , , and  to the particulate subcellular fraction within 15 min (Fig. 3). In cells exposed continuously to either PMA or PDBu (Fig. 3, A and B), down-regulation of PKC , , and  was evident by 6 h, and these isozymes were essentially depleted from the cells by 12 h. PKC  was more resistant to PMA-mediated down-regulation than PKC  or , with significant levels persisting in the particulate fraction for longer than 6 h. PMA pulse treatment produced similar effects on PKC isozyme subcellular distribution and expression, with a notable difference in the regulation of PKC . As shown in Fig. 3C, PMA pulse treatment also resulted in marked down-regulation of PKC  and  over the 24-h period. PKC  and cytosolic PKC  were depleted from the cells by 12 h, and membrane-associated PKC  was significantly down-regulated under these treatment conditions. Low levels of down-regulation-resistant PKC  were frequently observed in the particulate fraction following either pulse or continuous treatment with PMA for 24 h (see Fig. 3, A and C). PKC , on the other hand, showed a unique response to pulse treatment with PMA. While down-regulation of PKC  was evident in the particulate subcellular fraction by 6 h, this effect was accompanied by reappearance of the isozyme in the soluble fraction. By 12 h, control patterns of PKC  expression and subcellular distribution had been re-established in PMA pulse-treated cells (Fig. 3C). Phorbol ester treatment did not affect the expression or subcellular distribution of the atypical PKC isozymes  and  (data shown only for PKC , Fig. 3D). Interestingly, these PKC agonists also had no effect on the expression or subcellular distribution of PKC II.

Phorbol ester treatment results in translocation and subsequent down-regulation of PKC , , and , but does not affect PKC II or , in IEC-18 cells.

DiC₈, in contrast to phorbol esters, induced translocation but did not significantly down-regulate PKC , , and  over the 24-h experimental period (Fig. 4). Interestingly, while DiC₈ treatment resulted in complete translocation of PKC  and  to the particulate fraction, PKC  was only partially translocated by this DAG analogue. PKC II, , and  were neither translocated to the particulate fraction nor down-regulated by this agent (data not shown). Taken together, the data demonstrate that induction and maintenance of PKC-mediated inhibition of cell cycle progression in IEC-18 cells is coincident with the presence of PKC , , and/or  in the particulate subcellular fraction. Thus, phorbol ester treatment results in transient growth arrest and transient compartmentalization of these PKC isozymes in the particulate fraction. In contrast, DiC₈ produces sustained inhibition of cell cycle progression and sustained presence of PKC , , and  in this subcellular compartment. Therefore, the data support a role for PKC , , and/or  in negative growth regulatory signaling in IEC-18 intestinal epithelial cells. This notion is further supported by the observation that release of the PMA-induced G₁/S block in cell cycle progression in synchronized IEC-18 cells is also coincident with down-regulation of PKC , , and  (data not shown).

DiC₈ treatment results in sustained translocation of PKC , , and  to the particulate fraction in IEC-18 cells.

To explore the molecular pathways by which PKC interacts with the intestinal epithelial cell cycle and induces G₁ arrest, the effects of phorbol esters or DiC₈ on the expression and activity of key G₁ control molecules was investigated. The retinoblastoma gene product (Rb) is a major regulator of the G₁/S transition (36, 37). In its underphosphorylated, active form, assumed during late mitosis, Rb binds the transcription factor E2F and prevents it from participating in expression of genes required for DNA synthesis. During the course of G₁, Rb is functionally inactivated by sequential phosphorylation by several cyclin·cdk complexes, resulting in release of E2F and allowing transcription of S phase genes.

To examine the effects of PKC activation on the functional state of Rb, asynchronously growing IEC-18 cells were exposed to PKC agonists for various times, and Rb phosphorylation state was examined by Western blot analysis. This analysis revealed that PMA (administered as a 15-min pulse or in continuous exposure) or PDBu produced a marked increase in the levels of hypophosphorylated, active Rb in IEC-18 cells (Fig. 5, A and B). Rb hypophosphorylation was evident at both 2 h and 6 h following addition of phorbol ester. However, increased levels of hypophosphorylated Rb were no longer detectable at 24 h, indicating that the effect was transient. More detailed analysis of the time course of the effect revealed a significant loss of hyperphosphorylated, inactive Rb, with concurrent appearance of the hypophosphorylated, active form of the protein, by 60-90 min following addition of these agents (Fig. 5C). Hypophosphorylation of Rb was maintained for 8-10 h and was undetectable by 12 h (Fig. 5D). Thus, PKC agonist-induced translocation of PKC , , and/or  in these cells was associated with a shift in Rb phosphorylation favoring the hypophosphorylated, transcription factor-binding form which retards cell cycle progression. Down-regulation of PKC , , and , on the other hand, was accompanied by a shift favoring the hyperphosphorylated, growth-permissive form of Rb and coincided with release of phorbol ester-mediated cell cycle arrest in IEC-18 cells.

DiC₈ treatment also produced an accumulation of the hypophosphorylated, growth-suppressive form of Rb by 2 h, confirming the involvement of a PKC-mediated signal transduction pathway in modulation of Rb phosphorylation state in IEC-18 cells (Fig. 5E). In contrast to the effects seen with PMA and PDBu, hypophosphorylation of Rb was maintained over the 24-h experimental period. The sustained change in Rb phosphorylation state is consistent with the sustained presence of PKC , , and  in the particulate subcellular fraction and sustained cell growth arrest, supporting a role for one or more of these isozymes in negative growth-regulatory pathways involving modulation of the Rb growth suppressor protein in IEC-18 cells.

To examine the mechanism(s) underlying PKC-mediated modulation of Rb phosphorylation state, the effects of phorbol esters and DiC₈ on the expression of the Cip/Kip family cdk inhibitors p21^waf1/cip1 and p27^kip1 were examined by Western blot analysis. p21^waf1/cip1 and p27^kip1 represent a family of molecules which block the activity of the cyclin·cdk complexes responsible for phosphorylation of Rb (46). Treatment of IEC-18 cells with 100 nM PMA (administered as a 15-min pulse or in continuous exposure) resulted in a marked increase in levels of p21^waf1/cip1 by 2 h; levels of this cdk inhibitor remained elevated at 6 h but were indistinguishable from those in control cells by 24 h (Fig. 6A). Strong induction of p21^waf1/cip1 expression was seen within 60 min of PMA addition and reached a maximum at 2 h (Fig. 6B). The level of p21^waf1/cip1 induction by PMA was comparable with that observed 2 h following UV irradiation, a strong inducer of the protein (47). As with other effects of phorbol ester treatment, the increase in p21^waf1/cip1 expression was transient. Elevated levels were maintained for at least 6 h, but were found to diminish significantly thereafter (Fig. 6, A and C). DiC₈ (20 µg/ml) treatment also produced an increase in p21^waf1/cip1 expression; in this case, however, increased levels were sustained over the 24-h treatment period (Fig. 6D). Taken together, the data indicate that a PKC-mediated signaling cascade modulates p21^waf1/cip1 expression in IEC-18 cells; reversal of the effect in phorbol ester-treated cells correlates with down-regulation of PKC , , and , reappearance of the hyperphosphorylated form of Rb, and release of the block in cell cycle progression. Treatment with DiC₈, which resulted in sustained translocation of PKC ,  and , produced a sustained elevation of p21^waf1/cip1 levels in IEC-18 cells. Analysis of the levels of the related cdk inhibitor p27^kip1 consistently revealed similar, but more modest, changes in response to continuous PMA treatment for 2 or 4 h (Fig. 7). A reproducible ~1.5-fold increase in p27^kip1 levels was seen at these times following PKC agonist treatment.

To determine if PKC agonists induce changes in cell cycle control molecules when applied specifically in the G₁ phase, the effects of PKC agonists on Rb phosphorylation state and on the expression of p21^waf1/cip1 and p27^kip1 were examined in G₁-synchronized cells. IEC-18 cells synchronized by serum deprivation were released from growth arrest by serum stimulation and treated with PMA 8 h later; control and treated cells were collected at 10 and 16 h following release from growth arrest, and lysates were subjected to Western blot analysis. PMA treatment in mid-G₁ resulted in hypophosphorylation of Rb and induction of p21^waf1/cip1 and p27^kip1 within 2 h of PKC agonist treatment (Fig. 8). Hypophosphorylation of Rb was maintained in PMA-treated cells at 16 h after serum stimulation, a point at which control cells express fully hyperphosphorylated Rb.

PKC , , and/or  Are Required for PKC Agonist-mediated Cell Cycle-specific Effects in IEC-18 Cells

To confirm the requirement for PKC , , and/or  in mediating growth-inhibitory effects in IEC-18 cells, we examined the ability of PKC agonists to induce cell cycle arrest, Rb hypophosphorylation, and p21^waf1/cip1 expression in IEC-18 cells depleted of these isozymes. Treatment of IEC-18 cells with 1 µM PDBu for 24 h resulted in complete down-regulation of PKC , , and  (Fig. 9A), without affecting levels of PKC II, , or  (data not shown). These results are consistent with reported evidence that long term treatment with high concentrations of phorbol ester promotes proteolytic degradation and thus depletion of PKC isozymes in many biological systems (48), allowing assessment of PKC agonist effects in the absence of specific members of this enzyme family. Following PDBu treatment, the cells were washed extensively to remove the agonist. Subsequent retreatment of these PKC -, -, and -depleted cells with 100 nM PMA failed to produce the cell cycle arrest, Rb hypophosphorylation, or induction of p21^waf1/cip1 which were evident in cells in which PKC isozymes had not been depleted by previous exposure to PDBu (Fig. 9, B-D). These data provide further support for the requirement for one or more of these isozymes in mediating PKC agonist-induced growth-inhibitory effects. They also indicate that PKC II, , and  are not involved in PKC agonist-induced negative growth regulation in this system.

Depletion of PKC , , and  from IEC-18 cells prevents cell cycle arrest, modulation of Rb phosphorylation, and p21

PKC  Is Sufficient to Mediate PKC Agonist-induced Inhibition of Cell Cycle Progression, Rb Hypophosphorylation, and p21^waf1/cip1 Induction in IEC-18 Cells

As shown in Fig. 3C, pulse treatment (15 min) of IEC-18 cells with 100 nM PMA results in translocation followed by eventual depletion of cytosolic PKC  and  by 24 h. In contrast, following initial translocation, PKC  disappears from the particulate fraction and begins to re-emerge in the cytosolic compartment by 6 h. Thus, restimulation of IEC-18 cells with PKC agonists 24 h after a PMA pulse treatment addresses the response of these cells under conditions in which the major PKC agonist-responsive isozyme is PKC . The small amount of membrane-associated, down-regulation-resistant PKC  expressed by these cells is unlikely to play a role in negative growth-regulatory signaling events in this system, since the presence of comparable levels of this isozyme was unable to sustain growth arrest in cells exposed continuously to PMA over a 24-h period. Retreatment of these PKC /-deficient cells with 100 nM PMA resulted in complete translocation of PKC  to the particulate fraction (Fig. 10A), inhibition of IEC-18 cell cycle progression (Fig. 10B), hypophosphorylation of Rb (Fig. 10C), and accumulation of p21^waf1/cip1 (Fig. 10D). Taken together, these data support the ability of PKC  alone to mediate negative growth-regulatory signaling in IEC-18 cells, involving pathways controlling Rb phosphorylation and cdk inhibitor expression. Moreover, they indicate that PKC  and  do not play an obligate role in mediating growth arrest in intestinal epithelial cells.

PKC  is sufficient to mediate cell cycle arrest, modulation of Rb phosphorylation, and p21

DISCUSSION

Despite extensive evidence for a role of the PKC family of signal transduction molecules in positive or negative control of cell growth (3), understanding of the underlying mechanisms involved and of the specific growth-regulatory function(s) of individual isozymes remains limited. We have previously shown a strong correlation between activation of PKC family members and control of cell growth in intestinal crypts in situ (27). In the present study, the role of PKC isozymes in growth-inhibitory signaling was established directly in the IEC-18 crypt cell line, using a panel of PKC activators to bypass normal agonist-mediated control of PKC isozyme expression and activation state. The data presented provide strong support for the involvement of one or more specific PKC isozymes in negative regulation of cell cycle progression in G₁ phase and demonstrate that PKC  alone is sufficient to mediate this growth-regulatory effect. PKC-mediated G₁ arrest was found to involve induction of Cip/Kip family cdk inhibitors and activation of the growth-suppressive function of the Rb protein, thus linking PKC  to control of cdk inhibitor expression and Rb phosphorylation state in this system. The data also demonstrate a secondary cell cycle effect of PKC agonists in G₂/M and/or possibly S phase, indicating that PKC isozyme(s) can act at multiple stages to inhibit cell cycle progression in IEC-18 cells.

Phorbol ester- or DiC₈-mediated inhibition of cell cycle progression was shown to correlate directly with translocation of three members of the PKC family, PKC , , and , to the particulate subcellular fraction in IEC-18 cells. The involvement of these isozyme(s) in PKC agonist-induced cell cycle-specific effects was further indicated by the finding that the different abilities of phorbol esters and DAG analogues to sustain cell cycle-inhibitory effects in this system correlated with differential modulation of the expression of these molecules. Phorbol esters produced rapid down-regulation of these isozymes which coincided temporally with release of the block in cell cycle progression. In contrast, DiC₈ maintained significant levels of PKC , , and  in the particulate fraction over the 24-h experimental period, consistent with the sustained inhibition of cell cycle progression produced by this agent. Furthermore, depletion of PKC , , and  from IEC-18 cells abrogated PKC agonist-induced cell cycle effects in this system. Involvement of PKC II, , and  in these growth-regulatory effects was excluded by the findings that (a) IEC-18 cells expressing only PKC II, , and  failed to undergo growth arrest in response to PKC agonists (Fig. 9), and (b) neither the expression nor the subcellular distribution of these isozymes was affected by treatment with phorbol esters or DiC₈. Thus, of the six PKC isozymes expressed in the IEC-18 cell line, one classical (PKC ) and two novel (PKC  and ) isozymes were implicated in the observed growth-inhibitory response.

To gain insight into the individual contribution of PKC , , and/or  to negative growth regulation in IEC-18 cells, we used a pharmacological approach to manipulate specific PKC isozyme expression levels and subcellular distribution in this system. These studies implicated PKC , in particular, in mediating negative cell cycle regulatory events in IEC-18 cells. Conditions were established (i.e. a 15-min pulse treatment with 100 nM PMA followed by incubation in fresh medium for 24 h) which produced a population of IEC-18 cells in which the major PKC agonist-responsive isozyme present was PKC . Stimulation of these PKC - and -deficient cells with phorbol esters (Fig. 10) or DiC₈ (data not shown) resulted in translocation of PKC  to the particulate subcellular fraction and recapitulated the inhibition of cell cycle progression observed in cells expressing the full profile of PKC isozymes. As mentioned above, this effect was abrogated in IEC-18 cells also depleted of PKC  as a result of long term treatment with 1 µM PDBu (Fig. 9). Taken together with the finding that, in phorbol-ester treated cells, release of the block in cell cycle progression at 12 h correlated temporally with depletion of PKC , while PKC  and  exhibited different kinetics of down-regulation (Figs. 2 and 3), the data demonstrate that PKC  alone is sufficient to mediate PKC agonist-induced negative growth-regulatory signals in IEC-18 cells and that PKC  and  do not play a requisite role in these responses. Although the ability of PKC  and  to mediate growth-inhibitory signals has not been ruled out in this study, the notion that these isozymes do not participate in negative growth regulation in this system is consistent with the demonstration that the expression and subcellular distribution of PKC  do not change with cell growth arrest in the intestinal crypt in situ, and that the most pronounced changes in PKC  occur in association with enterocyte mature function on the villus (27). The ability of PKC  to mediate PKC agonist-induced growth-regulatory effects in IEC-18 cells correlated directly with its presence in the particulate subcellular fraction; i.e. disappearance of membrane-associated PKC  coincided with reversal of the cell cycle effects, even when accompanied by its reappearance in the cytosol. Based on evidence from a number of studies supporting a link between PKC  translocation and enzymic activation, the data point to a role for PKC  kinase activity in mediating negative growth-regulatory signaling in intestinal epithelial cells, a notion that will be addressed directly in future studies.

The data presented in this report implicating PKC isozyme(s) in negative growth-regulatory signaling pathways and inhibition of cell cycle progression are consistent with increasing evidence from studies in a number of cellular systems, including vascular smooth muscle cells (29), vascular endothelial cells (14, 49), melanoma cells (20), IMR-90 fibroblasts (28), and hematopoietic cells (50), demonstrating that PKC can mediate cell cycle arrest at the G₁/S boundary and/or in G₂/M phase. Thus, PKC isozyme-mediated inhibition of cell cycle progression is not unique to intestinal epithelial cells, but may reflect a widespread function of one or more PKC family members. Moreover, our data implicating PKC , in particular, in growth suppression in intestinal epithelial cells are in keeping with emerging information from a number of systems linking this isozyme with negative growth regulation. For example, specific activation of PKC  in K562 erythroleukemia cells results in cytostasis and megakaryocytic differentiation (24), whereas PKC II plays a requisite role in proliferation of these cells (30, 51). Similarly, overexpression of PKC  in R6 rat embryo fibroblasts results in marked growth inhibition (25), while overexpression of PKC I (25) or  (17) enhances the growth of these cells. Overexpression of PKC  in F9 teratocarcinoma cells (52), B16 melanoma cells (19), CHO cells (53), or 3Y1 fibroblasts (54) has also been shown to result in inhibition of cell growth/cell cycle progression. Taken together with our findings indicating that appropriately compartmentalized endogenous PKC  can negatively modulate cell cycle progression in intestinal epithelial cells, these data suggest that the biological role of PKC  is specifically associated with negative growth-regulatory signaling.

Our studies in the IEC-18 model indicate that PKC isozyme(s), and PKC  in particular, act by initiating a cascade of events resulting in increased levels of hypophosphorylated, transcriptionally repressive Rb. It is now well established that Rb is a major regulator of cell cycle progression; in its hypophosphorylated form, Rb represses transcription of genes essential for entry into S phase and thus induces cell cycle arrest. PKC agonists induced rapid (within 90 min) hypophosphorylation/activation of Rb in IEC-18 cells. Importantly, this change in Rb phosphorylation state occurred prior to cell growth arrest, supporting a cause-and-effect relationship between these events. Thus, PKC-mediated accumulation of the growth-suppressive form of Rb could account for the observed G₁ block in cell cycle progression. The involvement of Rb in PKC-mediated cell cycle-inhibitory effects is further supported by studies in human umbilical vein endothelial cells in which PDBu treatment was shown to induce G₁ arrest and Rb hypophosphorylation (14) and in NIH 3T3 cells, where ectopic expression of PKC  was shown to block normal growth factor-induced phosphorylation of the Rb protein (31). In addition, overexpression of PKC  in 3Y1 fibroblasts has been shown to decrease E2F transcriptional activity, further linking PKC to activation of the growth-suppressive function of Rb (54).

Insight was also obtained into the mechanism(s) by which PKC isozyme(s), and PKC  in particular, modulate the activation state of the Rb protein. The phosphorylation state and activity of Rb are controlled by both positive and negative regulators (37). Cyclin·cdk complexes phosphorylate, and thus functionally inactivate, Rb during the course of G₁. The activity of cdks is regulated by cyclin binding, by positive and negative phosphorylation events, and by the accumulation or activation of cdk-inhibitory proteins such as p21^waf1/cip1 and p27^kip1 (46). Our data demonstrate that PKC isozyme(s), or PKC  alone, can mediate induction of Cip/Kip family cdk inhibitors in IEC-18 cells. The time course of p21^waf1/cip1 induction closely paralleled the appearance of hypophosphorylated Rb in response to PKC agonists, suggesting that PKC isozyme-mediated G₁ arrest occurs, at least in part, as a result of increased expression of cdk inhibitory molecules and subsequent inhibition of Rb phosphorylation. Further evidence for this pathway comes from the observation that p21^waf1/cip1 expression is induced in response to PMA treatment in leukemic cells (55, 56), that NIH 3T3 cells overexpressing PKC  show elevated levels of both p21^waf1/cip1 and p27^kip1 (31), and that PMA treatment of synchronized melanoma cells inhibits the down-regulation of p21^waf1/cip1 and p27^kip1 seen at the G₁/S border in these cells (57). Our demonstration that the DAG analogue DiC₈ also induces the expression of p21^waf1/cip1 confirms the involvement of PKC isozyme(s) in mediating this effect. While the molecular mechanisms underlying PKC isozyme-mediated control of cdk inhibitor levels remain to be determined, it is noteworthy in regard to p21^waf1/cip1 induction that the activity of the p53 transcription factor, a major inducer of this cdk inhibitory protein (reviewed in Ref. 46), has been shown to be positively modulated by PKC phosphorylation (58). Thus, treatment of IEC-18 cells with PKC agonists may result in phosphorylation and activation of p53 and subsequent induction of p21^waf1/cip1 gene expression. However, evidence also exists for p53-independent induction of p21^waf1/cip1 (59, 60); the involvement of these pathways in PKC-mediated p21^waf1/cip1 induction in IEC-18 cells is currently under investigation in our laboratory.

The physiological relevance of the findings presented in this report regarding PKC isozyme control of cell growth and cell cycle progression in the intestinal epithelium is supported by our previous demonstration that marked changes in the expression and subcellular distribution of four members of the PKC family, including PKC , coincide precisely with cell growth arrest in small intestinal epithelial crypts in situ (27). Similar regulation of PKC isozymes has recently been reported in rat (61) and human (62) colonic epithelium, where marked increases in expression of several PKC isozymes coincide with cell growth arrest in the upper region of colonic crypts. It should be noted that, although PKC agonists did not induce the expression of differentiation markers in IEC-18 cells, our findings do not exclude a possible role of PKC-mediated cell cycle control in regulating downstream differentiation events in the intestinal epithelium in situ. The data presented also offer a perspective on the role of disturbances in the PKC enzyme system in development of colonic neoplasms. Human and rat colonic tumors have been shown to exhibit reduced levels of PKC activity (63-65) and PKC isozyme expression (66-69) relative to paired adjacent normal colonic mucosa. Decreased levels of PKC activity have also been observed in colonic adenomas (63), indicating that changes in PKC isozyme regulation occur early in the multistage process of colon carcinogenesis. It is noteworthy in this regard that PKC  has been found to be markedly down-regulated in both rat (66, 69) and human colonic adenocarcinomas.² Furthermore, overexpression of PKC I in HT-29 human adenocarcinoma cells was demonstrated to result in marked inhibition of cell growth and reduction of tumorigenicity in nude mice (70), findings which led Weinstein and colleagues to suggest that, in some cell types, PKC acts as a tumor suppressor gene. Thus, reduced levels of PKC isozyme expression in colonocytes may impair normal growth-inhibitory signaling, leading to increased cell growth and contributing to the development of colonic neoplasia.

Taken together, the data presented indicate that PKC isozyme activation, and PKC  activation in particular, mediates inhibition of cell cycle progression in IEC-18 cells, preceded by rapid accumulation of cdk inhibitory proteins and appearance of the growth-suppressive form of Rb. The relevance of these findings to regulation of cell growth in the intestinal epithelium in situ is supported by evidence that activation of PKC isozymes, including PKC , coincides precisely with cell growth arrest (27) and with induction of p21^waf1/cip1 expression (71, 72) in the mid-crypt region. Levels of p21^waf1/cip1 and p27^kip1 are thought to play a critical role in the decision of a cell to proliferate or to withdraw from the cell cycle in response to environmental signals (73, 74); thus, activation of PKC  may play an important role in maintenance of balanced growth in intestinal epithelial cells in situ by integrating environmental anti-mitogenic signaling with the process of cell division via regulation of cell cycle inhibitory proteins. The importance of subsequent functional activation of the Rb protein to maintenance of cell cycle arrest in this system is emphasized by studies in transgenic mice, where sequestration of Rb by SV40 large T antigen in villus cells was shown to result in re-entry of these post-mitotic cells into the cell cycle (75, 76). Thus, in the intestinal epithelium, our data suggest that PKC functions as a physiological regulator of epithelial tissue homeostasis, acting to limit the expansion of actively proliferating populations by participating in the following program: activation of PKC  by growth-inhibitory factors initiates a signaling cascade, involving p53-dependent or -independent mechanisms, which leads to induction of cdk inhibitory molecules, suppression of phosphorylation of Rb in G₁ (and modulation of other key substrates in S and/or G₂/M phases), and subsequent cell cycle arrest.