Regulated activity of PP2A-B55 delta is crucial for controlling entry into and exit from mitosis in Xenopus egg extracts

Abstract

Entry into mitosis depends on the activity of cyclin-dependent kinases (CDKs). Conversely, exit from mitosis occurs when mitotic cyclins are degraded, thereby extinguishing CDK activity. Exit from mitosis must also require mitotic phosphoproteins to revert to their interphase hypophosphorylated forms, but there is a controversy about which phosphatase(s) is/are responsible for dephosphorylating the CDK substrates. We find that PP2A associated with a B55 delta subunit is relatively specific for a model mitotic CDK substrate in Xenopus egg extracts. The phosphatase activity measured by this substrate is regulated during the cell cycle--high in interphase and suppressed during mitosis. Depletion of PP2A-B55 delta (in interphase) from 'cycling' frog egg extracts accelerated their entry into mitosis and kept them indefinitely in mitosis. When PP2A-B55 delta was depleted from mitotic extracts, however, exit from mitosis was hardly delayed, showing that other phosphatase(s) are also required for mitotic exit. Increasing the concentration of PP2A-B55 delta in extracts by adding recombinant enzyme inhibited the entry into mitosis. This form of PP2A seems to be a key regulator of entry into and exit from mitosis.

Figure 1

CDK substrate-directed protein phosphatase is regulated during cell cycle. (A) Interphase egg extract containing 0.1 mg/ml cycloheximide was supplemented with DMSO, okadaic acid (2.5 μM), p27^Kip1 (2 μM) or cyclin Δ90 (500 nM) as indicated. The extracts were incubated at 23°C. Aliquots were taken at 20 min intervals for immunoblotting with anti-Apc3 and anti-phospho CDK target antibodies. (B) Interphase egg extract was supplemented with cyclinΔ90 (500 nM) in the presence or absence of U0126 (300 μM). The extracts were incubated at 23°C and aliquots were taken for the Fizzy-Ser 50 phosphatase assay. Open triangles, +cyclinΔ90; closed triangles, +cyclinΔ90+U0126 (C) CSF-arrested egg extract was supplemented with p27^Kip1 (2 μM), followed by incubation at 23°C. Histone H1 kinase and Fizzy-Ser 50 phosphatase activities were measured at each time point. Open squares, H1 kinase of mock extract; open circles, phosphatase activity of mock extract; closed squares, H1 kinase of p27 ^Kip1-added extract; closed circles, phosphatase activity of p27 ^Kip1-added extract (D) Aliquots from each time point of (C) were blotted with anti-Apc3 and anti- phospho CDK target antibodies.

Figure 2

PP2A is the major phosphatase that dephosphorylates CDK substrates in interphase egg extracts. (A) Titration of okadaic acid on the Fizzy-Ser 50 phosphatase activity. Error bars denote the range from three independent experiments. The activities are expressed in percentage of phosphatase activity of okadaic acid-free extract. (B) Phosphatases were removed from interphase egg extract by immunodepletion. The depletion efficiency shown on the right side was estimated by comparing the signal intensities from immunoblotting a dilution series of mock-depleted extract. PP1, PP4, PP5 and PP6 were depleted using antibodies against their catalytic subunits. Depletion of PP2A was carried out with anti-A subunit monoclonal antibody 6F9 and confirmed with immunoblotting with anti-PP2A catalytic subunit antibody. (C) Fizzy-Ser 50 phosphatase activities were measured in each phosphatase-depleted extract shown in (B). The activities are expressed in percentages of the phosphatase activity of mock-depleted extract.

Figure 3

B55δ is the major B subunit for PP2A that dephosphorylates CDK substrates in interphase. (A) B subunits of PP2A were removed from interphase egg extract by immunodepletion. The depletion efficiency shown on the right side was estimated by comparing the signal intensities from immunoblotting of a dilution series of mock-depleted extract. (B) The Fizzy-Ser 50 phosphatase activity of each B subunit-depleted extracts shown in (A). Activities are expressed as a percent of phosphatase activity of mock-depleted extract.

Figure 4

PP2A–B55δ controls mitotic progression in Xenopus egg extracts. (A) Coomassie brilliant blue staining of the purified recombinant PP2A heterotrimer complex. Aα and Cα subunits were expressed from Xenopus cDNA clones, whereas B55δ is a rat protein. Left lane shows molecular weight markers and their size in kDa (kilodaltons) (B) Immunoblots of mock, B55δ-depleted and added-back extracts using anti-Xenopus B55δ (top) and catalytic (bottom) subunits. Note that the amount of the recombinant rat B55δ is equivalent to the endogenous protein but shows a weaker signal (right lane) because of weak reactivity of the rat protein with the anti-Xenopus B55δ antibody used for blotting. (C) Cell-cycle progression of B55δ-depleted extract. Fresh interphase extract was treated with protein-G Sepharose beads conjugated with unrelated rabbit serum or anti-B55δ serum. To one half of a B55δ-depleted extract, recombinant PP2A–B55δ complex was added back to give the same level of phosphatase activity as the original extract. These extracts were incubated at 23°C to monitor progress through the cell cycle. Aliquots were taken at 7-min intervals for immunoblotting of Apc3, Cyclin B2 and CDK target phosphorylation. Numbers on the right side indicate molecular weight in kDa. (D) Histone H1 kinase activity of each extract shown in (C). Black squares, mock-depleted extract; red circles, B55δ-depleted extract; blue triangles, B55δ-depleted/added-back extract.

Figure 5

Additional PP2A–B55δ changes the timing of mitotic progression in a dose-dependent manner. Fresh cycling extract was supplemented with 50% extra, 2, 3 or 5 times the amount of endogenous enzyme by adding recombinant PP2A-B55δ (as measured by phosphatase activity). Aliquots were taken at 10-min intervals for immunoblotting of Apc3, cyclin B2, CDK target phosphorylation and Cdk1-tyr 15 phosphorylation.

Figure 6

(A) PP2A–B55δ is required throughout preceding interphase but not after mitotic entry for the dephosphorylations of CDK substrates. Interphase egg extracts supplemented with 0.1 mg/ml of cycloheximide were immunodepleted for B55δ before (lanes 7–12) or after (lanes 19–24) mitotic induction by adding recombinant GST–cyclin B1. These extracts were then mixed on ice with anti-GST beads to remove the added cyclin, followed by incubation at 23°C to induce mitotic exit. The samples in lanes 1, 7, 13 and 19 were taken before cyclin depletion. Samples were collected at the indicated times after incubation at 23°C, analyzed by SDS–PAGE and blotted with the indicated antibodies. (B) Interphase egg extract was depleted for B55δ (lanes 11–20) before adding GST–cyclin B1, then CDK activity was inhibited either by adding 2 μM of p27^Kip1 or by removing GST–cyclin B1 with anti-GST beads. Samples were processed as described in (A).