Inducible expression of catalytically active type 1 serine/threonine protein phosphatase in a human carcinoma cell line

Background One of the major cellular serine/threonine protein phosphatases is protein phosphatase type 1 (PP1). Studies employing many eukaryotic systems all point to a crucial role for PP1 activity in controlling cell cycle progression. One physiological substrate for PP1 appears to be the product of the retinoblastoma susceptibility gene (pRB), a demonstrated tumor suppressor. The growth suppressive activity of pRB is regulated by its phosphorylation state. Of critical importance is the question of the in vivo effect of PP1 activity on pRB and growth regulation. As a first step towards addressing this question, we developed an inducible PP1 expression system to investigate the regulation of PP1 activity. Results We have established a cell line for inducing protein expression of the type 1, alpha-isotype, serine/threonine protein phosphatase (PP1α). A plasmid encoding a fusion protein of the catalytic subunit of PP1α with a 6-histidine peptide (6His) and a peptide from hemagluttinin (HA) was transfected into the UMUC3 transitional cell carcinoma cell line, previously transfected with the reverse tetracycline transactivator plasmid pUHD172-1neo. A stable cell line designated LLWO2F was established by selection with hygromycin B. 6His-HA-PP1α protein appeared in cell lysates within two hours following addition of doxycycline to the culture medium. This protein localizes to the nucleus as does endogenous PP1α, and was shown to associate with PNUTS, a PP1-nuclear targeting subunit. Like endogenous PP1α, immunocomplexed 6His-HA-PP1α is active toward phosphorylase a and the product of the retinoblastoma susceptibility gene, pRB. When forcibly overexpressing 6His-HA-PP1α, there is a concomitant decrease in endogenous PP1α levels. Conclusions These data suggest the existence of an autoregulatory mechanism by which PP1α protein levels and activity remain relatively constant. RT-PCR analyses of isolated polysome fractions support the notion that this putative autoregulatory mechanism is exerted, at least in part, at the translational level. Implications of these findings for the study of PP1α function in vivo are discussed.

fractions support the notion that this putative autoregulatory mechanism is exerted, at least in part, at the translational level. Implications of these findings for the study of PP1α function in vivo are discussed.

Background
One of the major cellular serine/threonine protein phosphatases is protein phosphatase type 1 (PP1) [1]. An abundant enzyme expressed in all cells, complex regulation of PP1 is thought to be essential for proper temporal and spatial regulation of PP1 catalytic activity towards individual substrates [2]. Three different isotypes of the catalytic subunit, designated PP1α, PP1δ, and PP1γ1, are expressed in many different cell types [3]. Studies employing many eukaryotic systems all point to a crucial role for PP1 activity in controlling cell cycle progression, and an absolute requirement of this activity for mitotic exit [4][5][6][7]. There is also evidence to suggest that phosphorylation of PP1 and its associated proteins by the cyclin-dependent kinases may regulate PP1 activity in a cell cycle stagedependent manner [8][9][10]. Hence, current efforts are directed towards identification of cell cycle-dependent substrates for PP1, and how regulation of PP1 activity towards these substrates controls the cell division cycle.
One physiological substrate for PP1 appears to be the product of the retinoblastoma susceptibility gene, pRB [11][12][13], a demonstrated tumor suppressor. The cell growth and tumor suppressive activity of pRB is regulated by its phosphorylation state. pRB phosphorylation varies as a function of cell cycle phase; during G 1 the hypophosphorylated form predominates, while the hyperphosphorylated form accumulates during S, G 2 and M phase [14][15][16]. It is this hypophosphorylated form of pRB present during early and mid G 1 which functions to modify gene expression. This is achieved by complexing with transcription factors resulting in repression [17] or stimulation [18] of transcription. Under conditions favoring proliferation, the phosphorylation of pRB via cyclin-dependent kinases (CDKs) in mid-to-late G 1 phase results in liberation of E2F and other pRB-bound transcription factors, which then activate the transcription of S-phase genes [19]. In late M phase, pRB returns to its growth suppressive, hypophosphorylated form due to the action of a pRB-directed protein phosphatase belonging to the type 1 class of serine/threonine protein phosphatases [11,20,21].
Of critical importance is the question of the in vivo effect of PP1 activity on pRB and growth regulation. To date, several techniques have been employed in this regard to demonstrate PP1 actions in intact cells. Berndt et. al. [22] used electroporation to introduce PP1 protein into tissue culture cells having noted that previous genetic attempts to accomplish overexpression were not readily accomplished. Using this technique, they showed a G1 block by PP1 that was dependent upon the presence of pRB. In another report, micro-injection of PP1 decreased extractability of pRB from nuclei, presumably by increasing dephosphorylation and increasing association with nuclear proteins thereby playing a role in G1/S transition inhibition [13]. Micro-injection of antibodies to the PP1 catalytic subunit has been shown to block cells in metaphase [23]. Since this is the time during mitosis when pRB is normally dephosphorylated [11,24], it is tempting to speculate that prevention of pRB dephosphorylation by inhibiting PP1 activity impedes an orderly progression through mitosis, perhaps contributing towards a transformed phenotype.
Here, using the tetracycline induction system [25], we developed an inducible PP1 expression system to investigate the regulation of PP1 activity. We view this as a first step towards addressing the question of the in vivo effect of PP1 activity on pRB and growth regulation. Our results support the notion that induced PP1α protein functions identically to endogenous PP1α with respect to non-catalytic subunit binding and activity towards phosphorylase a and the tumor suppressor pRB. Localization of this protein to the nucleus also mirrors that of endogenous PP1α in vivo. When using this system to forcibly overexpress 6His-HA-PP1α, there is a concomitant decrease in endogenous PP1α levels, suggesting the existence of an autoregulatory mechanism by which PP1α protein levels remain constant. RT-PCR analyses of isolated polysome fractions supports the notion that this putative autoregulatory mechanism is exerted, at least in part, at the translational level. Figure 1 shows construct of the tetracycline-inducible plasmid used to express 6His-HA-PP1α. As shown in Figure 2A, 24 hr incubation with doxycycline in the medium has no appreciable effect on endogenous PP1α protein expression in the untransfected parent cell line (UMUC3) or a cell line transfected with the reverse tetracycline transactivator plasmid pUHD172-1neo only (LLWO1). A time course of induction by doxycycline was then carried out in stable clone LLWO2F. Harvested at various time intervals, whole-cell lysates were prepared and the proteins were separated by SDS-PAGE for western blotting using antibody to the hemagglutinin tag. As shown in Figure 2B, a protein with the anticipated molecular weight of 40 kDa was detected. A clear differential of expression of 6His-HA-PP1α (indicated at the left of the panel) was observed at 1.5 hours after addition of doxycycline, which became more pronounced at later time points. The doxycyclineindependent increase in 6His-HA-PP1α is attributed to background expression when using this system [25]. As predicted for a protein under the control of the doxycycline-inducible promoter, the abundance of 6His-HA-PP1α decreases in the absence of the inducer over time ( Figure 2C). Subsequent time course experiments revealed peak abundance occurring between 16 and 24 hr, and remaining constant thereafter for at least 72 hr, at which time the experiment was terminated. As such, we have chosen the 24 hr induction time for all subsequent experiments to ensure peak abundance of the induced 6His-HA-PP1α.

Induction of 6His-HA-PP1 in LLW02F cells
As shown in Figure 2C, a significant reduction in 6His-HA-PP1α abundance can be observed between 3 and 12 hrs following removal of doxycycline. This reduction continued through the 24 hr time point at which time the experiment ended. Taken together, these data confirm the dependence of 6His-HA-PP1α protein production on the The 6His-HA-PP1α-pTEP4m plasmid Figure 1 The 6His-HA-PP1α-pTEP4m plasmid. The plasmid contains the 6His-HA-PP1α fusion gene under the control of the TetO/ CMV promoter. This promoter is active in the presence of the rtTA-nls protein and doxycycline. The plasmid contains an ampicillin resistance gene (Amp r) for selection in bacteria and the hygromycin resistance gene for selection in eukaryotic cells.
tet-O/CMV promoter, thus demonstrating inducible expression of PP1α. As expected [25], some low-level expression of 6His-HA-PP1α in the absence of the inducer doxycycline can also be detected.

Biological and Biochemical Characterization of Inducible 6His-HA-PP1α
To begin addressing the biological and biochemical characterization of inducible PP1α, we next localized this protein in situ using immunohistochemistry. As shown in Figure 3, there is a predominant nuclear localization of the hemagglutinin epitope (panel A) compared to the low-level cytoplasmic and nuclear background reactivity seen in the uninduced LLWO2F control cells (panel B). Thus, not only can doxycycline-induced expression of PP1α be detected immunohistochemically, but localization of this protein appears to be identical to that of endogenous PP1 [26].
Targeting of PP1 to the nucleus is due in part to a PP1associated nuclear targeting subunit (PNUTS; [27]). To further investigate the properties of inducible PP1α, we performed coprecipitation studies to determine if PNUTS can also be found associated with 6His-HA-PP1α. Towards this goal, a fusion protein of GST and PNUTS was tested for the ability to capture induced PP1α. Lysates from induced LLWO2F cells were combined with GSTalone-or GST-PNUTS-loaded glutathione Sepharose beads. Following separation by SDS-PAGE and western blotting, antibody to hemagglutinin was used to detect any associated 6His-HA-PP1α. Antibody reactivity would indicate a link from the GST epitope to the hemagglutinin epitope via a PNUTS to PP1α interaction. As shown in Figure 4, the hemagglutinin epitope was captured by the GST-PNUTS fusion protein, but not the GST-alone protein. These data support the idea that induced PP1α and the nuclear targeting PP1-associated protein PNUTS can form a complex. These results further support the view that induced PP1α behaves similarly if not identically to endogenous PP1α, and provides a possible mechanistic explanation for 6His-HA-PP1α localization to the nucleus.
Two different approaches were taken to address the biochemical activity of inducible PP1α. First, immunocomplexed 6His-HA-PP1α was tested for its ability to dephosphorylate phosphorylase a. Phosphorylated on a single serine residue, this substrate is often used to monitor the biochemical activity of PP1 in vitro [28]. As shown in Figure 5, greater phosphatase activity is detected in immunoprecipitates from induced LLWO2F cell lysates compared to uninduced. This is consistent with the notion that inducible PP1α is enzymatically active. Our second approach involved testing immunopurified 6His-HA-PP1α for its ability to dephosphorylate the growth A. Doxycycline has no effect on PP1α protein expression in control cell lines Figure 2 A. Doxycycline has no effect on PP1α protein expression in control cell lines. Equal quantities (50 ug) of whole-cell lysates harvested in the absence (indicated above the lanes with a (-) symbol), or presence (at 24 hr following addition, indicated above the lanes with a (+) symbol), of doxycycline were separated by SDS-PAGE and subjected to immunoblotting using antibody specific for PP1α. UMUC3 -parent cell line from which stable expressor LLWO2F was selected; LLWO1 -cell line generated from UMUC3 which was stably transfected with the reverse tetracycline transactivator plasmid pUHD172-1neo only. B -Doxycycline induces a protein in LLW02F cells that is recognized by antibody to hemagglutinin. Equal quantities (50 ug) of whole-cell lysates harvested at the various time points in hours after doxycycline addition (indicated above each lane) were separated by SDS-PAGE and subjected to immunoblotting as described in Materials and Methods. C -Time-course of recovery from doxycycline induction. LLWO2F cells were induced for 24 hr. At the end of this time, equal quantities (50 ug) of whole-cell lysates harvested at the various time points in hours following removal of doxycycline (indicated above each lane) were separated by SDS-PAGE and subjected to immunoblotting with antibody to hemagglutinin.

UMUC3 LLWO1
DOX -+ -+ PP1α suppressor protein pRB. Having previously reported on the ability of immunocomplexed endogenous PP1 to dephosphorylate pRB [20], we followed the same experimental approach. Figure 6 shows a significant decrease in the 32 P radiolabel intensity for pRB following incubation with the anti-hemagglutinin immunoprecipitate compared to the normal mouse IgG control immunoprecipitate. In addition, okadaic acid, a potent inhibitor of PP1 activity [29], significantly inhibits pRB dephosphorylation in this reaction. Taken together, these data indicate that inducible PP1α is biochemically active.

Upregulation of inducible 6His-HA-PP1α coincides with downregulation of endogenous PP1α
One prediction would be that by increasing expression of the catalytic subunit of PP1α, overall PP1-specific activity within the cell would increase. To address this, we tested the phosphatase activity of LLWO2F whole-cell lysate using phosphorylase a as the substrate. To inhibit any endogenous PP2A activity, which can also dephosphorylate phosphorylase a and thus interfere with PP1 activity measurements, these assays were carried out in the presence of 5 nM okadaic acid, which does not inhibit PP1 activity [30]. Although reproducible, we observed only modest increases in PP1-specific activity found in lysates from induced cells compared to the uninduced controls ( Figure 7). One possible explanation for this finding is that the overall level of PP1α within the cell remains relatively constant regardless of induction. Indeed, a time Immunohistochemistry with the antibody to hemagglutinin shows doxycycline-dependent expression of 6His-HA-PP1α in vivo To further test this hypothesis, immunoprecipitation and western blotting experiments were carried out using PP1 isoform-specific antibodies. As shown in Figure 9, top panel, induced (approximate molecular weight of 40 kDa) and endogenous (approximate molecular weight of 37 kDa) PP1α are clearly resolved by SDS-PAGE. As predicted, both proteins are recognized by antibody specific for PP1α when performing immunoprecipitation or western blotting. Immunoprecipitation using PP1α antibody followed by western blotting using the same anti-PP1α reveals a reciprocal relationship in PP1α abundance following doxycycline induction; endogenous PP1α levels decrease in response to increasing levels of 6His-HA-PP1α (top panel, compare left two lanes). This difference in abundance is more striking in the whole-cell lysate lanes; endogenous PP1α fails to be detected in lysates prepared from induced cells (compare middle two lanes). Parallel experiments performed using the same lysate and antibody to hemagglutinin for immunoprecipitation and western blotting show the position of 6His-HA-PP1α and the fact that endogenous PP1α is not recognized by this antibody (bottom panel). Western blotting for additional isoforms of PP1 reveals that this reduction in PP1α following induction is isotype-specific; when 6His-HA-PP1α levels increase, concomitant with a decrease in PP1α levels, PP1δ and PP1γ1 levels remain relatively unchanged from those found in uninduced cell lysate ( Figure 10). Taken together, these data suggest that an as yet undescribed negative-feedback or autoregulatory mechanism exists for PP1α which contributes towards maintaining a constant level of protein expression and enzyme activity.

RT-PCR of polysome and total RNA specific for endogenous PP1α and 6His-HA-PP1α
To address the possible autoregulatory mechanism(s) of PP1α expression and activity, we analyzed the RNA levels of both endogenous PP1α and induced 6His-HA-PP1α. For uninduced cells, only endogenous PP1α RNA can be found associated with polysomes ( Figure 11, second lane). While some 6His-HA-PP1α RNA can be detected in the total RNA, the majority PP1α RNA detected in the total preparation is endogenous (fourth lane). In contrast, Time (hrs) Dox induction 6His-HA-PP1α PP1α it appears that for the doxycycline-induced cells (third lane), there is more 6His-HA-PP1α RNA associated with polysomes than endogenous PP1α. However, both messages appear to be present at the same level in the total RNA preparation (fifth lane). This apparent preference for 6His-HA-PP1α RNA association with polysomes suggest that autoregulation takes place, at least in part, at the translational level.

Conclusions
We observed that endogenous PP1α protein levels decrease in response to increasing levels of induced 6His-HA-PP1α protein. This observation suggests that an as yet undescribed negative-feedback or autoregulatory mecha-nism exists for PP1α which contributes towards maintaining a constant level of this enzyme. Given the difficulty in achieving overexpression of PP1 protein and activity in mammalian cells, this suggestion provides an attractive explanation. This putative mechanism appears to be isotype-specific, since the protein levels of PP1δ and PP1γ1 remained relatively unchanged in response to induced expression of PP1α. Precedent for an autoregulatory mechanism comes from a report by Baharians and Schonthal [31], in which they demonstrate endogenous PP2A protein levels decrease in the presence of increasing amounts of forcibly-expressed exogenous PP2A. Their report also provides evidence that this control is exerted at the translational level and does not involve regulation of transcription or RNA processing. Using a similar approach of analyzing mRNA association with ribosomes as a measure of translation efficiency, a similar level of translational control also appears to exist for PP1α. If mRNA association with ribosomes is truly a measure of translation efficiency, one conclusion from these data is that in doxycycline-induced cells, 6His-HA-PP1α RNA is more efficiently translated than endogenous PP1α. This would help to explain why we observed a decrease in PP1α isoform-specific antibody immunoprecipitation and detection of endogenous and induced PP1α by western blotting Figure 9 PP1α isoform-specific antibody immunoprecipitation and detection of endogenous and induced PP1α by western blotting. PP1α isoform-specific antibody was used to immunoprecipitate proteins from equal quantities (50 ug) of uninduced (first lane from the left, top panel) and induced LLWO2F cell lysate (second lane from the left, top panel). The precipitated proteins were then separated by SDS-PAGE and immunoblotted with the same isoform-specific antibody. The positions of 6His-HA-PP1α and PP1α are indicated to the left of the top panel. Whole-cell lysates (10 ug) were separated in the next two lanes. Last lane contains 1 ug of purified recombinant PP1α as a control for endogenous PP1α (37 kDa). Bottom panel shows identical experiment using antibody to hemagglutinin for immunoprecipitation and immunodetection. Although low levels of 6His-HA-PP1α can be detected in cell lysate from uninduced cells following western blotting, this protein is almost undetectable in the same lysate following immunoprecipitation and western blotting. This observation is most likely due to low efficiency of immunoprecipitation.
Isoform-specific antibody detection of PP1 by western blot-ting induced and uninduced LLWO2F whole-cell lysates Figure 10 Isoform-specific antibody detection of PP1 by western blotting induced and uninduced LLWO2F whole-cell lysates. Different amounts (5 ug for the first and third lanes, 10 ug for the second and fourth lanes) of lysate from induced and uninduced cells were separated by SDS-PAGE in parallel and then immunoblotted using PP1α-, PP1δ-, and PP1γ1-specific antibody. Results illustrate downregulation of endogenous PP1α in response to increasing amounts of induced 6His-HA-PP1α, while the levels of PP1δ and PP1γ1 remain relatively constant. Lane C -control lane of isoform-specific antibody immunoprecipitation of respective isoforms from 50 ug of uninduced cell lysate.
endogenous PP1α protein expression in doxycyclineinduced cells. Since we did not perform rigorous analyses on PP1α regulation of transcription or RNA processing, we cannot rule out the possibility that some level of transcriptional control also exists. Indeed, the abundance of endogenous PP1α RNA found in uninduced cells appears equal to or greater than the combined abundance of endogenous and 6His-HA-PP1α RNA (Figure 9), supporting the notion that some level of translational control may also exist. These findings, provided by the development of this inducible PP1 expression system, will afford new research opportunities aimed towards addressing the question of how PP1 activity is regulated. This putative autoregulatory mechanism for the catalytic subunit may help to explain why the overall abundance of PP1 during the various phases of the cell cycle remains relatively constant [32]. Historically, regulation of PP1 catalytic activity has resulted from associated subunits inhibiting or targeting the catalytic subunit towards substrates [33]. While the PP1 catalytic subunit is not found alone in the cell, such unassociated catalytic subunits have been shown to be active, at least in vitro. Although speculative, there may be a threshold amount of catalytic subunit in vivo above which associated proteins are limiting. Such a situation may compromise the ability of associated proteins to effectively regulate PP1 catalytic activity. To guard against this scenario, control over the amount of PP1 catalytic subunit present via an autoregulatory mechanism would be beneficial.
In summary, we report on the development of an inducible mammalian cell expression system for the catalytic subunit of the α-isotype for PP1. This protein behaves similarly if not identically to endogenous PP1α with respect to nuclear localization, complex formation to a recently described PP1-associated nuclear targeting protein, and in vitro enzymatic activity towards phosphorylase a and the cell growth and tumor suppressor protein pRB.
The data presented here validate a useful system with which to manipulate PP1 activity in vivo for the purpose of functional studies. Just as is the case for PP2A [31], it seems clear that efficient overall increases in PP1 protein abundance may not be achieved due to a putative negative feed-back or autoregulatory mechanism. However, induced expression of exogenous PP1 which behaves identically to endogenous PP1 while remaining physically distinguishable can facilitate the use of mutants to address functional questions, particularly in the area of cell cycle regulation and tumor cell growth suppression. Indeed, the use of a constitutively active mutant of PP1 has already been shown to result in pRB-dependent G1 arrest in human cancer cells [22]. Placing this mutant into the inducible expression system described here has the potential to down-regulate expression of wild-type endogenous PP1. Doing so in a timed, controlled manner will permit in vivo studies addressing PP1 function during discrete phases of the cell cycle, and will be the subject of future endeavors.

Tetracycline Induction System
The rabbit PP1α cDNA in pDR540 [34] was ligated into pcDNA3 (Invitrogen) to create a fusion encoding 6 histidines (6-His) followed by an epitope (YPYDVPDYA) from the Hemophilus influenza hemagglutinin protein (HA). The sequence and reading frame was validated by BigDye DNA sequencing (PE Applied Biosystems). The fusion gene was then cut from the 6His-HA-PP1α-pcDNA3 plasmid and ligated into pTEP4m, which contains a tetracycline response element for inducible expression [25], and a hygromycin resistance gene for selection in eukaryotic cells (see Figure 1). pUHD172-1neo [25] encodes the reverse tetracycline transactivator protein fused with a nuclear localization signal (rtTA-nls). In addition to the rtTA-nls gene under control of the CMV promoter, the plasmid contains neomycin and ampicillin antibiotic resistance genes.

Cell Culture, Transfection, and Selection
The UMUC3 cell line was obtained from the American Type Culture Collection. The cells were maintained in Dulbecco's minimal essential medium(Life Technologies) RT-PCR of polysome and total RNA specific for endogenous PP1α and 6His-HA-PP1α Figure 11 RT-PCR of polysome and total RNA specific for endogenous PP1α and 6His-HA-PP1α. PCR products from polysome and total RNA from induced and uninduced cell lysates were prepared and analyzed as described in Materials and Methods. MW -molecular mass standard ladder. supplemented with 10% fetal bovine serum, 20 units/ml penicillin and 20 ug/ml streptomycin. Cells were incubated at 37C, in a 5% carbon dioxide containing atmosphere.
Plasmid DNA was introduced into cultured cells using Lipofectamine (Life Technologies). 2 ug of plasmid DNA and 25 ul of Lipofectamine were combined in 200 ul of Opti-mem media (Life Technologies) and allowed to stand at room temperature for 30 minutes. This mixture was diluted to 1 ml with Opti-mem and applied to cells in one well of 6-well tissue culture plates. After two hours, 1 ml of complete medium was added to each well. After twenty-four hours the medium was replaced with complete medium.
Transfection of UMUC3 cells with pUHD172-1neo was followed 48 hr later by selection using geneticin. Isolated colonies were obtained at 800 µg/ml. These cells were passaged and expanded before transfection with pBI-EGFP (Clontech), which contains the enhanced green fluorescent protein coding sequence under control of a bi-directional tetracycline response element. At the end of the 2 hr transfection (as described above) the medium was replaced with complete medium with and without doxycycline at 1 ug/ml final concentration. 24 hr later the cells were examined by fluorescence microscopy. The presence of doxycycline dependent green fluorescence was indicative of stable transfection with pUHD172-1neo encoding the rtTA-nls protein and transient transfection with pBi-EGFP. Parallel cultures were then passaged and transfected with 6His-HA-PP1α-pTEP4m. Selection with hygromycin was begun 48 hr later. Isolated colonies were obtained at 100 ug/ml hygromycin and were passaged and diluted to extinction. Six cell lines, designated LLW02A through LLW02F, were established. The LLW02F cell line was used for all subsequent experiments. This choice was based on high levels of induced PP1α expression as assayed by immunoblotting. Doxycycline was used at 1 ug/ml for induction.

SDS-PAGE and Immunoblotting
Cells were lysed for 15 min at 4°C in EBC buffer (50 mM Tris-HCl, pH 8.0, 120 mM NaCl, 0.5% Nonidet P-40) containing 10 ug/ml of the protease inhibitors aprotinin, leupeptin, and phenylmethylsulfonylfluoride (PMSF). The lysates were cleared by centrifugation at 14,000 × g for 10 minutes. Electrophoresis was performed in SDS-polyacrylamide gels [35] using the indicated amount of total cell protein [36]. After electrophoresis, the proteins were transferred to nitrocellulose paper in buffer containing 25 mM Tris-HCl, 192 mM glycine, 20% v/v methanol, and 0.01% SDS, pH 8.5 [37]. Residual protein binding sites on the nitrocellulose were blocked by incubation for 30 minutes in TBST (25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% Tween-20) containing 4% non-fat dry milk. Next, the nitrocellulose was incubated in TBST containing 2% non-fat dry milk containing primary antibody directed towards the indicated protein. Following three washes of 10 min each with TBST, the nitrocellulose was probed with horse-radish peroxidase-conjugated anti-IgG (Promega, Madison, WI) and developed using chemiluminescence detection (Pierce, Rockford, IL) according to the manufacturer's instructions.

Immunohistochemistry
Cell monolayers were rinsed 3 times with PBS, fixed for 10 minutes in -20 o C methanol, and then air dried for 1 hr. Hematoxylin was used as the nuclear counterstain. Following dehydration in ethanol and air drying, the cell monolayers were flooded with glycerol for subsequent observation, photography, and storage.

Preparation of [ 32 P]pRB substrate and pRB-directed phosphatase activity assays
To assay pRB-directed phosphatase activity, [ 32 P]-labeled pRB substrate was prepared by immunoprecipitation with monoclonal antibody to pRB, PMG3-245 (PharMingen), from [ 32 P]-labeled CV-1P cells as described previously [20]. For the source of enzyme, immunocomplexed 6His-HA-PP1α from induced LLWO2F cells were mixed and incubated with [ 32 P]-labeled immunoprecipitated pRB for 30 minutes at 30°C as previously described [20]. Phosphatase reactions were terminated by the addition of SDS-PAGE sample buffer and boiling. Proteins were separated on SDS-polyacrylamide gels. Gels were fixed, dried subjected to autoradiography. Immunoprecipitation using PP1 Isoform-specific antibody LLWO2F cell extracts (50 ug) were immunoprecipitated with approximately 1 ug of isoform-specific rabbit antiserum to PP1α as previously described [38]. Immunoprecipitation with 1 ug of normal rabbit serum served as a negative control.

GST-PNUTS Affinity Chromatography
Induction of GST-PNUTS fusion protein synthesis and isolation by affinity chromatography using glutathione-Sepharose beads (Pharmacia) was performed as described previously [32]. Induced LLWO2F cell extracts were then incubated for 1 hour at 4°C with recombinant GST or GST-PNUTS fusion protein bound to glutathione-Sepharose. Bound proteins were washed with PBS, separated on 10 % SDS-polyacrylamide gels, and transferred to nitrocellulose. Protein blots were then developed using antibody to hemagglutinin for 6His-HA-PP1α detection.

Polysome isolation
Uninduced and induced LLWO2F cells (10 × 100 mm dishes) were washed 3-times with ice-cold serum-free DMEM containing 10 ug/ml cycloheximide. After the last wash, cells were scraped into a final volume of 5 ml of this same medium for each 10-dish treatment. The cells were harvested with a brief centrifugation. The cell pellet was resuspended in 0.5 ml lysis buffer (125 mM KCl, 12.5 mM mgCl2, 10 mM Hepes, pH 6.8, 0.1 mM dithiothreitol, 10 ug/ml cycloheximide, 0.5% Triton X-100, 0.5% deoxycholate) followed by 10 passages through a 22 gauge needle to lyse the cells. The resulting extracts were then layered on the top of a 10-40% sucrose gradient (125 mM KCl, 12.5 mM mgCl2, 10 mM Hepes, pH 6.8, 0.1 mM dithiothreitol, 10 ug/ml cycloheximide) and centrifuged at 4°C in a SW50.1 rotor at 48,000 rpm for 60 min. After centrifugation, the gradient was fractionated in 200 ul aliquots and analyzed at A 260 using a Bio-Rad SmartSpec 3000. A single peak of A 260 absorbance in the bottom portion of the gradient was analyzed as polysome fraction.

RT-PCR analyses of PP1α mRNA
LLWO2F cells were induced with doxycycline (2 ug/ml) for 24 hr. Negative control LLWO2F cells were not induced with doxycycline. For total RNA experiments, RNA was harvested from monolayered cells using QIAshredder and Qiagen RNeasy mini-kit. Polysome RNA was recovered from polysome fractions by first diluting fractions in 3.5 volumes of lysis buffer from the Qiagen RNeasy mini-kit and then continuing with the manufacturer's instructions. RNA was quantitated by UV spectrometry. 1 ug total RNA (or 200 ng polysome RNA) was used in first strand cDNA synthesis using SUPERSCRIPT II reverse transcriptase (Life Technologies) following manufacturer's instructions. cDNA was extracted by one phenol chloroform:isoamyl alcohol extraction followed by one chloroform isoamyl alcohol extraction and then ethanol precipitated. cDNA was resuspended in 20 ul water and used in a 30-cycle PCR reaction with 1 uM of each of the following four primers: {CCACGCTGTTTTGACCTCCAT-AGAAGACAC, CACATAGTCCCCCAGAAAGAGGTAGTT-GCT}, in which product only forms from 6His-HA-PP1α cDNA, and {GACGCGGGCAAGCAGTCCCTCGAGAC-CATTGCCTGCTG, CTGGAGACCCACGACCTGGCCT-GCCGTTG}, in which product only forms from endogenous PP1α cDNA. Reaction also contained 200 uM dNTPs, 1 mM Mg ++ , and 5 units Taq DNA polymerase (Life Technologies) with supplied reaction buffer. 2 ul of the PCR product was loaded onto a 1% agarose gel containing ethidium bromide for UV visualization.

Authors' Contributions
JER, JWL, and EMM conceived the study and participated in its design and execution. MPS participated in plasmid constructions. PK performed RT-PCR and polysome analysis. EV-M provided PP1-isoform specific antibodies and participated in PP1 immunoprecipitations and western blots. Authors read and approved the final manuscript.