* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Nuclear-cytoplasmic distribution of Inh2
Survey
Document related concepts
Cytoplasmic streaming wikipedia , lookup
Extracellular matrix wikipedia , lookup
Cell growth wikipedia , lookup
Endomembrane system wikipedia , lookup
Tissue engineering wikipedia , lookup
Cytokinesis wikipedia , lookup
Signal transduction wikipedia , lookup
Cellular differentiation wikipedia , lookup
Organ-on-a-chip wikipedia , lookup
Cell culture wikipedia , lookup
Cell encapsulation wikipedia , lookup
Cell nucleus wikipedia , lookup
Transcript
Research Article 3739 Domains of type 1 protein phosphatase inhibitor-2 required for nuclear and cytoplasmic localization in response to cell-cell contact Craig Leach*, Masumi Eto and David L. Brautigan Center for Cell Signaling, University of Virginia School of Medicine, Box 800577-MSB7225, Charlottesville VA 22908, USA *Author for correspondence (e-mail: [email protected]) Accepted 10 July 2002 Journal of Cell Science 115, 3739-3745 © 2002 The Company of Biologists Ltd doi:10.1242/jcs.00052 Summary Inhibitor-2 of type 1 protein phosphatase is a phosphoprotein conserved among all eukaryotes, and it appears in both the nucleus and cytoplasm of tissue culture cells. We discovered that endogenous inhibitor-2 is concentrated in the nucleus of cells cultured at low density, whereas cells growing at high density excluded inhibitor-2 from the nucleus. There was rapid redistribution of inhibitor-2 when cells were replated at low or high density. Localization of myc-tagged forms of inhibitor-2 showed that residues 119-197 were required for nuclear accumulation in low-density cells and residues 78-119 were required for cytoplasmic localization in high-density cells. Fusion of inhibitor-2 residues 78-119 to green fluorescent protein was sufficient to produce cytoplasmic retention. Inhibitor-2 fused to triple tandem green fluorescent protein (100 kDa) was imported into the nucleus of low-density cells but was not excluded from the nucleus when cells reached high density, implying that inhibitor-2 was actively imported into the nucleus but exited by passive diffusion instead of active export. We conclude that inhibitor-2 contains two separate domains that control its localization in the nucleus or cytoplasm. This change in inhibitor-2 localization may direct inhibitor-2 to different forms of protein phosphatase 1 or change the localization of protein phosphatase, as part of the cellular response to cell-cell contacts at high density. Introduction Type 1 protein phosphatase (PP1) is an essential and highly conserved enzyme, with over 80% amino-acid sequence identity between baker’s yeast (called GLC7) and human (Cannon et al., 1994) PP1. PP1 is the major enzyme that dephosphorylates serine and threonine residues on proteins. It has been implicated in the control of a range of cellular processes, including the cell cycle, gene expression, cell adhesion and glycogen metabolism (Cohen, 1989; Shenolikar, 1994). To achieve this wide range of effects, PP1 exists as a series of holoenzymes, with its catalytic subunit, PP1C, bound to a regulatory subunit, of which there are dozens. Evidence suggests that the subcellular localization, activity and substrate specificity of PP1C are controlled by its interaction with these regulatory subunits (Bollen, 2001; Egloff et al., 1997). Inhibitor-2 (Inh2) was discovered as a heat stable protein, purified from rabbit skeletal muscle, which could inhibit PP1C activity in vitro (Huang and Glinsmann, 1976). Inh2, like PP1, is conserved among eukaryotes, with Inh2 homologues identified in baker’s yeast (GLC8), Drosophilia, rats, rabbits and humans (Cannon et al., 1994; Helps and Cohen, 1999; Osawa et al., 1996; Sanseau et al., 1994; Zhang et al., 1992). Inh2 forms a 1:1 complex with PP1C and inhibits the activity of PP1 towards several substrates in vitro; it also binds to and inhibits the activity of the PP1G holoenzyme (Resink et al., 1983; Yang et al., 1981). Previous studies have suggested that there are multiple sites in Inh2 that interact with the catalytic subunit of PP1 (Huang et al., 1999; Yang et al., 2000). It is unclear if these different sites target Inh2 to different PP1 holoenzymes in order to respond to changes in cell signaling. Inh2 is a 23 kDa phosphoprotein, small enough to passively diffuse through nuclear pores, yet it contains both a putative nuclear localization signal (137KKRQFEMKRK147) and a sequence resembling a leucine-rich nuclear export signal (155LNIKLARQLI165) (Gerace, 1995; Quimby and Corbett, 2001). Previous reports have shown that Inh2 can differentially localize to the cytoplasm or nucleus, depending on the phase of the cell cycle, in human HS68 fibroblast cells (Kakinoki et al., 1997). Mutation of two lysine residues present in the putative bipartite NLS domain diminished the cell-cycledependent nuclear accumulation of Inh2, suggesting that Inh2 utilized an active import mechanism (Kakinoki et al., 1997). Inh2 levels fluctuate during the cell cycle, starting from a low level and increasing during G1, peaking during S phase, dropping during G2 and then reaching its highest peak during mitosis (Brautigan et al., 1990; Kakinoki et al., 1997). Progression through the cell cycle can be halted by the formation of cell-cell contacts in a process known as contact inhibition (Bunge, 1979). This process is poorly understood and in some instances, such as tissue formation, cells must overcome contact inhibition. On the other hand, aberrant control of this process may contribute to the type of uncontrolled growth seen in cancer. In this study we discovered that in both HeLa and PC3 cells, Key words: Nuclear import, Green fluorescent protein, Cell density 3740 Journal of Cell Science 115 (19) the subcellular localization of Inh2 changes in response to cell density and is independent of the cell cycle. We mapped the regions of Inh2 involved in its subcellular distribution in response to cell density and have identified a domain capable of causing the cytoplasmic retention of a heterologous protein. Materials and Methods Tissue culture and general reagents Tissue culture media and sera were obtained from Gibco BRL. PC3 cells were cultured in RPMI media supplemented with 10% FBS, and HeLa cells were cultured in DMEM supplemented with 10% NCS. Cells were cultured at 37°C in 5% CO2. Mouse antibodies to both the myc (clone 9E10) and HA (clone 12CA5) epitopes were obtained from the antibody core facility at the University of Virginia. The PP1δ antibody was obtained from QCB. Chemicals were from Sigma unless otherwise noted. Immunofluorescence microscopy PC3 and HeLa cells were cultured to confluency and then replated at different cell densities onto coverslips coated with fibronectin. After 12-18 hours, or other times as specified in the figure legends, the samples were fixed with 3% paraformaldehyde [2% for green fluorescent protein (GFP) fluorescence] and permeabilized with 0.1% Triton-X100. Specimens were blocked with 3% bovine serum albumin/1% ovalbumin and incubated with a mouse antibody to Inh2 (Transduction Laboratories). A sheep antibody to mouse IgG conjugated with fluorescein (Amersham) was used for immunofluorescence. Coverslips were mounted onto glass slides with Vectashield mounting medium (Vector Laboratories). Slides were visualized using a Nikon Microphot-5A microscope with 20× objective and were captured with a Hamamatsu C4742 digital camera operated by OpenLab software (Improvision) and processed in Adobe Photoshop for printing. Subcellular fractionation HeLa cells were cultured at either low density (LD), ~60 cells per mm2 or high density (HD), ~600 cells per mm2, and permeabilized by scraping with 0.25 M sucrose, 10 mM Tris-HCl (pH 7.5), 5 mM MgCl2 and 0.01% digitonin. Samples were incubated on ice for 10 minutes and then centrifuged at 5000 g for 5 minutes. The pellets containing cell nuclei were washed once in the same buffer and then dissolved in buffer containing 1% SDS. Samples of the supernatant (cytosol, C) and the dissolved pellets (nuclear, N) were normalized for total protein by a Coomassie-dye-binding assay from BioRad. Samples were resolved by SDS-PAGE and analyzed by western blotting using antibodies to Inh2 (Transduction Laboratories) and RCC1 (Santa Cruz). HRP-conjugated secondary antibodies from Pierce were used in conjunction with Supersignal reagent (Pierce) and used to expose X-ray film. Replating assay and quantification HeLa cells growing at either low or high density were trypsinized for 5 minutes, centrifuged for 2 minutes at 1000 g and replated to the opposite condition onto fibronectin-coated coverslips for either 20, 40 or 60 additional minutes. Adherent cells were washed with PBS, fixed with 2% paraformaldehyde, permeabilized with 0.1% Triton-X100 and immunostained for Inh2 using an antibody from Transduction Laboratories. On the basis of microscopic examination, cells were divided into three categories: (1) nucleus brighter than cytoplasm; (2) cytoplasm brighter than nucleus; and (3) indistinguishable (owing to cell shape or equivalent staining). A total of 100 cells for each time point was counted for each of three separate experiments. Electroporation HeLa cells were trypsinized, pelleted and resuspended in DMEM supplemented with 10% neonatal calf serum and 15 mM HEPES to a concentration of 2-5×106 cells/ml. This suspension was electroporated with 20-40 µg of the plasmid diluted in 210 mM NaCl using a BioRad Genepulser set on 300 mV and 960 µF. For electroporation of myctagged Inh2, cells were replated to low density for 12-18 hours and then replated to low or high density onto fibronectin-coated coverslips for an additional 4 hours. To compare Inh2-GFP3 with the NLS mutant, electroporated cells were replated to low density on fibronectin-coated coverslips for 6-10 hours. Analysis of Inh2HA3 and Inh2GFP3 at high cell density was done by initially plating electroporated cells to low density overnight and then replating to high density for 4 hours. Calculation of import index Total cell fluorescence intensity and nuclear fluorescence were determined by a digital CCD camera using Openlab software. Cytoplasmic staining was determined by subtracting the nuclear fluorescence from the total fluorescence. The import index was calculated by the equation [(Ci-Ni)/Ni], which subtracted the average cytoplasmic pixel intensity from the average nuclear pixel intensity to create a value that represented how much brighter the nucleus was than the cytoplasm. This number was then divided by the average nuclear pixel intensity to normalize for differences in protein expression between cells. Results Distribution of Inh2 and PP1δ in the nucleus and cytoplasm of low- and high-density cells Endogenous Inh2 in both HeLa and PC3 cells cultured at low cell density was concentrated in the nucleus relative to the cytoplasm (Fig. 1A,B). By contrast, Inh2 was cytoplasmic in HeLa and PC3 cells grown to high cell density (Fig. 1B), with no detectable staining of Inh2 in the nucleus. Subcellular fractionation confirmed the Inh2 distribution seen by immunofluorescence microscopy. HeLa cells cultured at low and high density were permeabilized with 0.01% digitonin to selectively disrupt the plasma membrane but not the nuclear envelope. The cell suspension was centrifuged to yield a supernatant consisting of the cytoplasm and a pellet containing the intact nuclei and cytoskeleton that were dissolved in 1% SDS. Western blotting of these fractions is shown in Fig. 1C. In low-density cells, Inh2 was predominantly in the nuclear fraction, with a much lower level detected in the cytoplasmic fraction. On the other hand, in high-density cells Inh2 was present mostly in the cytoplasmic fraction, with much lower levels in the nuclear fraction. Therefore the results of immunofluorescence and subcellular fractionation were consistent, showing nuclear versus cytoplasmic localization of Inh2 in response to cell density. The results reinforce each other and exclude possible artifacts of immunofluorescence, such as masking of epitopes or interference by fixation. A constitutively nuclear protein, RCC1, was used as a marker and as expected was only present in the nuclear fraction. The same amounts of total protein were analyzed, and indeed samples from both low- and high-density cells displayed the same level of RCC1 staining. We noted thatInh2 staining was more intense in the low-density cells. This result showed that low-density cells had a higher concentration of Nuclear-cytoplasmic distribution of Inh2 3741 Fig. 2. Endogenous PP1δ has a similar localization to Inh2. HeLa cells were grown to either low (A) or high density (B) on coverslips and examined by immunofluorescence using an antibody against PP1δ. Fig. 1. Inh2 localization corresponding to cell density. PC3 (A,B) and HeLa (C,D) cells were grown to either low (A,C) or high density (B,D) on fibronectin-coated coverslips and examined by immunofluorescence using an antibody against human Inh2. To confirm Inh2 distribution by western blotting, high-density (HD) or low-density (LD) HeLa cells were permeabilized with 0.01% digitonin to yield a cytosolic fraction, labeled C, and a nuclear fraction, labeled N. The top panel shows western blotting using an antibody to Inh2 and the bottom panel shows blotting with an antibody to RCC1 the Ran activator protein, a constitutively nuclear protein. Inh2 than high-density cells, consistent with the difference in intensity seen by immunofluorescent staining (Fig. 1). As Inh2 binds to the catalytic subunit of PP1, we investigated the localization of PP1δ by immunofluorescence. PP1δ localized mostly to the nucleus of low-density HeLa cells (Fig. 2A) and was excluded from the nucleus of high-density cells (Fig. 2B) in a manner similar to Inh2. Kakinoki et al. reported nuclear accumulation of Inh2 in cells during the S phase of the cell cycle, so we wanted to determine if low- and high-density cells had obvious differences in cell cycle distribution (Kakinoki et al., 1997). FACS analysis of propidium-iodide-stained low- and highdensity HeLa cells showed similar distribution of cells in the G1/S/G2-M phases of the cell cycle (Fig. 3). Therefore high cell density did not arrest HeLa cell cycle progression nor did the different localization of Inh2 in HeLa cells grown to low or high density seem to depend on the phase of the cell cycle. Redistribution of Inh2 in response to replating at different cell densities The nuclear-cytoplasmic distribution of Inh2 was reversed within minutes in high-density cells replated at low density and low-density cells replated at high density. HeLa cells cultured at high density were trypsinized and replated within 10 minutes at low density onto coverslips and examined by immunofluorescence for Inh2 at 20 (Fig. 4A,B,C), 40 or 60 (Fig. 4D,E,F) minutes after replating. Nuclear accumulation of Inh2 was evidently completed within 60 minutes (Fig. 4F), as seen by the appearance of overlap (yellow) of Inh2 staining (green) and DAPI staining (red) that were initially separate (Fig. 4C). The graph in Fig. 4 displays the kinetics of Inh2 import into the nucleus. In a reciprocal experiment, HeLa cells cultured at low density were trypsinized, collected by centrifugation and replated at high density onto coverslips. The kinetics of Inh2 redistribution from the nucleus to the cytoplasm was the same as import into the nucleus (data not shown). Addition of Leptomycin B, an inhibitor of Crm1mediated export, did not prevent the translocation of Inh2 from the nucleus to the cytoplasm when low-density cells were replated at high density (data not shown). The data show that the redistribution of Inh2 from the nuclear to cytoplasmic compartments occurs rapidly and does not depend on active nuclear export. Separate domains of Inh2 control subcellular distribution We prepared a series of Inh2 proteins that were myc-tagged at the N-terminus. These defined four major domains of Inh2, a C-terminal region, an N-terminal region, a central domain and a NLS-containing region (Fig. 5K). The Inh2 proteins were transiently expressed in HeLa cells and visualized by anti-myc immunofluorescence (Fig. 5). The myc-Inh2[0-197] behaved exactly like the endogenous full-length Inh2 (205 residues), showing accumulation in the nucleus of low-density cells (Fig. 5D) and exclusion from the nucleus of high density cells (Fig. 5C). The myc-Inh2 displayed the same localization when expressed in HEK293, COS7 and NIH3T3 cells grown at low or high density (data not shown). Without the N-terminal domain (residues 0-77), myc-Inh2[78-197] showed the same cell-density-dependent nuclear localization as full-length Inh2 (Fig. 5E,F). By contrast, deletion of both the C-terminal domain (residues 150-197) and the NLS-containing domain (residues 120-150) yielded a truncated form of myc-Inh2[0119] that appeared in the cytoplasm but not the nucleus, regardless of cell density (Fig. 5G,H). The results suggest that the region 120-197 was required for accumulation of Inh2 in the nucleus of low-density cells. By truncation of both the Nterminal domain and the C-terminal domain, we created a minimal protein containing both the central domain and the 3742 Journal of Cell Science 115 (19) Fig. 3. FACS analysis of cells cultured at different densities. HeLa cells growing at low or high density were trypsinized and stained with propidium iodide as described in Materials and Methods. The phase of the cell cycle was determined on the basis of the intensity of the propidium iodide staining using CellQuest software. NLS domain: myc-Inh2[78-150]. This truncated form exhibited both nuclear accumulation in low-density cells and cytoplasmic retention in high-density cells in a manner identical to the full-length protein (Fig. 5I,J). Residues 78-119 are sufficient for cytoplasmic retention On the basis of size, both myc-Inh2[0-119] and myc-Inh2[78197] would be expected to pass through nuclear pores and be uniformly distributed throughout the cell. Because these proteins were both predominantly localized to the cytoplasm in high-density cells, we propose that residues 78-119, common to both proteins, are involved in retention of these proteins in the cytoplasm. We created a fusion protein of Inh2[78-119] with GFP. Transient expression of this protein in HeLa cells revealed that whereas GFP alone was localized to both the nucleus and cytoplasm of low-density cells (Fig. 6A), the fusion protein containing Inh2 residues 78-119 was retained in the cytoplasm (Fig. 6B). We conclude that residues 78-119 are sufficient to retain a heterologous protein in the cytoplasm. Localization of Inh2 fusion proteins To examine transit in and out of the nucleus, Inh2 was produced as two different fusion proteins; one with a triple HA (HA3) tag resulting in a protein small enough (<30 kDa) for diffusion and the other with three tandem GFPs (GFP3) to create a protein of >100 kDa, which is too large to traverse nuclear pores by passive diffusion. HeLa cells expressing these fusion proteins were plated at low density for 12 hours and then replated at either low (Fig. 7) or high density (Fig. 8). After allowing for cell spreading, the cells were fixed for microscopy and the nuclear and cytoplasmic fluorescence (anti-HA or GFP) quantified. In low-density cells that have endogenous Inh2 in the nucleus, we found that both HA3- and GFP3-Inh2 fusion proteins also accumulated in the nucleus (Fig. 7A). When two lysines (K144, K146) in the putative NLS sequence were mutated to alanine in Inh2-GFP3, nuclear accumulation was abrogated (Fig. 7B). Mutation of a highly conserved phosphorylation site (Thr72) to alanine was previously reported to abrogate the nuclear accumulation of Inh2 (Kakinoki et al., 1997). We tested this mutation in the context of the Inh2-GFP3 protein but did not see a significant reduction in nuclear import (Fig. 7). Fig. 4. Inh2 relocalization occurs within minutes of altering cell density. HeLa cells growing at high density were trypsinized and replated at low density on fibronectincoated coverslips. After 20 (A,B,C), 40 or 60 minutes (D,E,F), cells were fixed with 3% paraformaldehyde and examined by immunofluorescence microscopy using an antibody to Inh2. Panels A and D show the staining of endogenous Inh2 in green. Panels B and E show the DAPI staining of DNA in red. Panels C and F show the merged image of the Inh2 staining with DAPI staining in red with the overlay of the two stains yielding yellow. Cells were scored as having nuclear staining brighter than, equal to, or dimmer than cytoplasmic staining. The bottom graph shows the combined result of three independent experiments with more than 100 cells being tallied in each experiment. The dashed line shows the percentage of cells with predominant nuclear staining and the solid line shows the percentage of cells with predominant cytoplasmic staining. Nuclear-cytoplasmic distribution of Inh2 3743 Fig. 6. Residues 78-119 are sufficient for cytoplasmic retention of a heterologous protein. HeLa cells cultured at low density and expressing either GFP (A) or a fusion of residues 78-119 of Inh2 with GFP (B) were visualized for GFP fluorescence. On the other hand, in high-density cells only Inh2-HA3 (Fig. 8B) behaved in a manner identical to endogenous Inh2, becoming mostly cytoplasmic, whereas Inh2-GFP3 (Fig. 8A) remained concentrated in the nucleus. The results imply that Inh2 passively diffuses out of the nucleus and that the size of the GFP3 fusion prevented this diffusion, but the much smaller HA tag did not. Discussion This study shows that endogenous Inh2 as well as overexpressed Inh2 accumulates in the nucleus of both HeLa and PC3 cells when cultured at low density. When a higher cell density is reached, Inh2 is retained in the cytoplasm and Fig. 5. Separate domains of Inh2 are required for cytoplasmic retention and nuclear import. HeLa cells expressing myc-tagged Inh2 proteins were cultured at either low (B,D,F,H,J) or high density (A,C,E,G, I) on fibronectin-coated coverslips. A and B show a phase image for I and J and are representative of all panels. C and D show a near full-length Inh2 [0-197]; E and F are an N-terminally truncated form (residues 78-197); G and H are a C-terminal truncation (residues 0-119) and I and J are of a central construct consisting of residues 78-150. K shows a schematic of the truncations used. Fig. 7. Inhibitor-2 is actively imported into the nucleus. A and B show low-density HeLa cells expressing either an Inh2GFP3 fusion (A) or an Inh2GFP3 fusion in which lysines 142 and 144 have been mutated to alanine, termed NLS mutGFP3 (B). Cells were visualized with the fluorescence of GFP and the nuclear, cytoplasmic and total fluorescence were quantified. The graph shows the import index for GFP3, Inh2GFP3 fusion, NLS MutGFP3 or Inh2[Thr72Ala]GFP3 fusion. 3744 Journal of Cell Science 115 (19) Fig. 8. Inhibitor-2 exits the nucleus by passive diffusion. HeLa cells were electroporated with either Inh2GFP3 or Inh2HA and grown at low density for 6 hours. The cells were then trypsinized and plated onto fibronectin-coated coverslips for 4 hours. The coverslips were then fixed and visualized either for GFP fluorescence or by indirect immunofluorescence techniques directed against the triple HA tag (orange). A shows the GFP fluorescence and B shows a merge of the HA immunofluorescence (orange) and DAPI staining of the DNA (Blue). excluded from the nucleus. Truncation of the C-terminal 80 amino acids of Inh2 prevented the nuclear accumulation of Inh2. Within this region, which is required for import, there are two clusters of basic amino acids between residues 137-146, similar to known bipartite NLSs (Quimby and Corbett, 2001). HeLa cells expressing an Inh2GFP3 fusion in which K144 and K146 were mutated to alanines did not show nuclear import in low-density cells. The results indicate that Inh2 contains a functional NLS and enters the nucleus by active import. Nuclear pores allow passive diffusion of proteins with a mass up to 70 kDa, which raises the question of why would Inh2, a 23 kDa protein, contain a functional NLS? One possible explanation is that Inh2 exists in the cell as part of a protein complex and that the NLS of Inh2 is responsible for mediating the nuclear translocation of the entire complex. PP1 is present as holoenzymes consisting of the catalytic subunit bound to various targeting subunits, such as GM or neurabin. Inh2 is capable of binding and inhibiting the activity of at least one such PP1 holoenzyme (Strålfors et al., 1985). Thus, Inh2 could be controlling the localization of one or more PP1 holoenzymes in response to changes in cell density. Kakinoki et al. described a nuclear-cytoplasmic redistribution of overexpressed Inh2 correlating to the phase of the cell cycle in HS68 cells that depended on an NLS and on Inh2 phosphorylation (Kakinoki et al., 1997). However, the localization of Inh2 discovered in this study, although it requires an NLS, does not rely on the cell cycle or Inh2 phosphorylation. Thus, there seem to be multiple mechanisms controlling Inh2 localization. Western blotting of Inh2 showed that low-density cells had more Inh2 than high-density cells. This raises the question of whether or not protein degradation or synthesis is responsible for the changes in Inh2 localization. One possibility is that lowdensity cells produce more Inh2 and saturate the system responsible for cytoplasmic retention leading to nuclear import and accumulation. This is unlikely because overexpression of Inh2[0-197] does not result in nuclear accumulation. Protein degradation might control localization by degrading nuclear Inh2 in high density but not low-density cells. However, this does not appear to be the case, because Inh2-GFP3 expressed in high-density cells still accumulates in the nucleus. We defined residues 78-150 as the minimal domain of Inh2 that is able to undergo nuclear-cytoplasmic redistribution in response to changes in cell density. Conspicuously absent from this domain are residues 10IKGI13, which interact with the catalytic subunit of PP1 (Huang et al., 1999). We further defined this region by demonstrating that residues 78-119 were sufficient to cause cytoplasmic retention of GFP, which is present in both the cytoplasm and nucleus when expressed in HeLa cells. Previous studies suggest that this region of Inh2 does not bind to PP1 (Huang et al., 1999). Although PP1 binding is not required for Inh2 localization, there is the possibility that PP1 may be bound to Inh2 when it redistributes between the nucleus and cytoplasm in response to cell density. There have been reports of other proteins altering their subcellular distribution in response to cell density (Baba et al., 2001; Trompeter et al., 1996). One example is ZnBP, which localizes to the nucleus of low-density cells and is excluded from the nucleus in high-density cells (Trompeter et al., 1996). ZnBP, like Inh2, contains a bipartite NLS responsible for its nuclear localization in low-density cells (Trompeter et al., 1999). Another similarity is that the cytoplasmic retention domain of Inh2 [78-119] contains a cluster of acidic residues (7/13); ZnBP also contains clusters of acidic residues. It is possible that these acidic clusters interact with a similar cytoplasmic protein that retains Inh2 and ZnBP in the cytoplasm. Additional work with ZnBP has led to the proposal of a developmental function of ZnBP. Overexpression of ZnBP in HL-60 cells leads to an increase in cell proliferation accompanied by a decrease in differentiation (Rodriguez et al., 1998). PP1 has previously been shown to activate transcription factors, such as Sp1, by dephosphorylation (Daniel et al., 1996). There are several examples of genes whose transcription is altered in response to cell density (e.g. VHL protein, fibronectin and collagen) (Baba et al., 2001; Wolthuis et al., 1993). One potential model is that Inh2 transports a PP1 holoenzyme into the nucleus of low-density cells to alter gene transcription by dephosphorylating certain transcription factors. Investigations of how cell density and gene transcription interrelate are fundamental for understanding both cancer and embryonic development. Cell-density-related control of Inh2 localization opens new avenues of study into the role of PP1 function in cell-cell signaling. This work was supported by grant GM 56362 from the National Institute of General Medical Sciences. C.A.L. was supported by training grant GF-10461 to the University of Virginia Cell and Molecular Biology Program. M.E. was supported by a post-doctoral fellowship from the American Heart Association Mid-Atlantic Affiliate. References Baba, M., Hirai, S., Kawakami, S., Kishida, T., Sakai, N., Kaneko, S., Yao, M., Shuin, T., Kubota, Y., Hosaka, M. et al. (2001). Tumor suppressor protein VHL is induced at high cell density and mediates contact inhibition of cell growth. Oncogene J. 20, 2727-2736. Bollen, M. (2001). Combinatorial control of protein phosphatase-1. Trends Biochem. Sci. 26, 426-431. Brautigan, D. L., Sunwoo, J., Labbe, J. C., Fernandez, A. and Lamb, N. J. C. (1990). Cell cycle oscillation of phosphatase inhibitor-2 in rat fibroblasts coincident with p34CDC2 restriction. Nature 344, 74-78. Nuclear-cytoplasmic distribution of Inh2 Bunge, R. G. L., Lieberman, M., Raben, D., Salzer, J., Whittenberger, B. and Woolsey, T. (1979). Growth control by cell to cell contact. J. Supramol. Struct. 11, 175-187. Cannon, J. F., Pringle, J. R., Fiechter, A. and Khalil, M. (1994). Characterization of glycogen-deficient glc mutants of Saccharomyces cerevisiae. Genetics J. 136, 485-503. Cohen, P. (1989). The structure and regulation of protein phosphatases. Annu. Rev. Biochem. 58, 453-508. Daniel, S., Zhang, S., DePaoli-Roach, A. A. and Kim, K. H. (1996). Dephosphorylation of Sp1 by protein phosphatase 1 is involved in the glucose-mediated activation of the acetyl-CoA carboxylase gene. J. Biol. Chem. 271, 14692-14697. Egloff, M., Johnson, D. F., Moorhead, G., Cohen, P. T. W., Cohen, P. and Barford, D. (1997). Structural basis for the recognition of regulatory subunits by the catalytic subunit of protein phosphatase 1. EMBO J. 16, 1876-1887. Gerace, L. (1995). Nuclear export signals and the fast track to the cytoplasm. Cell 82, 341-344. Helps, N. R. and Cohen, P. T. W. (1999). Drosophila melanogaster protein phosphatase inhibitor-2: identification of a site important for PP1 inhibition. FEBS Lett. 463, 72-76. Huang, F. L. and Glinsmann, W. H. (1976). Separation and characterization of two phosphorylase phosphatase inhibitors from rabbit skeletal muscle. Euro. J. Biochem. 70, 419-426. Huang, H.-B., Horichi, A., Watanabe, T., Shih, S.-R., Tsay, H.-J., Li, H.C., Greengard, P. and Nairn, A. C. (1999). Characterization of the inhibition of protein phosphatase-1 by DARPP-32 and inhibitor-2. J. Biol. Chem. 274, 7870-7878. Kakinoki, Y., Somers, J. and Brautigan, D. L. (1997). Multisite phosphorylation and the nuclear localization of phosphatase inhibitor 2green fluorescent protein (Inh2-GFP) fusion protein during S phase of the cell growth cycle. J. Biol. Chem. 272, 32308-32314. Osawa, Y., Nakagama, H., Shima, H., Sugimua, T. and Nagao, M. (1996). Identification and characterization of three isotypes of protein phosphatase inhibitor-2 and their expression profiles during testis maturation in rats. Eur. J. Biochem. 242, 793-798. Quimby, B. B. and Corbett, A. H. (2001). Nuclear transport mechanisms. Cell. Mol. Life Sci. 58, 1766-1773. 3745 Resink, T. J., Hemmings, B. A., Tung, H. Y. L. and Cohen, P. (1983). Characterization of a reconstituted Mg-ATP-dependent phosphatase. Eur. J. Biochem. 133, 455-461. Rodriguez, P., Vinuela, J. E., Alvarez-Fernandez, L., Buceta, M., Vidal, A., Dominguez, F. and Gomez-Marquez, J. (1998). Overexpression of prothymosin alpha accelerates proliferation and retards differentiation in HL-60 cells. Biochem. J. 331. Sanseau, P., Jackson, A., Alderton, R. P., Beck, S., Senger, G., Sheer, D., Kelly, A. and Trowsdale, J. (1994). Cloning and characterization of human phosphatase inhibitor-2 (IPP-2) sequences. Mamm. Genome 5, 490-496. Shenolikar, S. (1994). Protein serine/threonine phosphatases – new avenues for cell regulation. Annu. Rev. Cell Biol. 10, 55-86. Strålfors, P., Hiraga, A. and Cohen, P. (1985). The protein phosphatases involved in cellular regulation. Purification and characterisation of the glycogen-bound form of protein phosphatase-1 from rabbit skeletal muscle. Eur. J. Biochem. 149, 295-303. Trompeter, H., Schiermeyer, A., Blankenburg, G., Hennig, E. and Soling, H. (1999). Factors involved in the cell density-dependent regulation of nuclear/cytoplasmic distribution of the 11.5-kDa Zn2+binding protein (parathymosin-alpha) in rat hepatocytes. J. Cell Sci. 112, 4113-4122. Trompeter, H.-L., Brankd, I. A. and Soling, H.-D. (1996). Variable nuclear cytoplasmic distribution of the 11.5 kDa zinc-binding protein (parathymosin-alpha) and identification of a bipartite nuclear localization signal. J. Biol. Chem. 271, 1187-1193. Wolthuis, A., Boes, A. and Grond, J. (1993). Cell density modulates growth, extracellular matrix, and protein synthesis of cultured rat mesangial cells. Am. J. Pathol. 143, 1209-1219. Yang, J., Hurley, T. D. and DePaoli-Roach, A. (2000). Interaction of inhibitor-2 with the catalytic subunit of type 1 protein phosphatase. J. Biol. Chem. 275, 22635-22644. Yang, S.-D., Vandenheede, J. R. and Merlevede, W. (1981). Identification of inhibitor-2 as the ATP-Mg-dependent protein phosphatase modulator. J. Biol. Chem. 256, 10231-10234. Zhang, Z., Bai, G. and Lee, E. Y. C. (1992). PCR cloning of the cDNA of rabbit skeletal muscle protein phosphatase inhibitor-2. Biochem. Biophys. Res. Commun. 186, 1168-1170.