The chemical genetic strategy in which mutational enlargement of the ATP-binding site sensitises of a protein kinase to bulky ATP analogues has proved to be an elegant tool for the generation of conditional analogue-sensitive kinase alleles in a variety of model organisms. Here, we describe a novel substitution mutation in the kinase domain that can enhance the sensitivity of analogue-sensitive kinases. Substitution of a methionine residue to phenylalanine in the +2 position after HRDLKxxN motif of the subdomain VIb within the kinase domain markedly increased the sensitivities of the analogue-sensitive kinases to ATP analogues in three out of five S. pombe kinases (i.e. Plo1, Orb5 and Wee1) that harbor this conserved methionine residue. Kinome alignment established that a methionine residue is found at this site in 5–9% of kinases in key model organisms, suggesting that a broader application of this structural modification may enhance ATP analogue sensitivity of analogue-sensitive kinases in future studies. We also show that the enhanced sensitivity of the wee1.as8 allele in a cdc25.22 background can be exploited to generate highly synchronised mitotic and S phase progression at 36°C. Proof-of-principle experiments show how this novel synchronisation technique will prove of great use in the interrogation of the mitotic or S-phase functions through temperature sensitivity mutation of molecules of interest in fission yeast.
Many cell biology studies use genetic or pharmacological approaches to stop, promote or accelerate molecular function. Complete removal of protein function through gene knockout (KO), knockdown or point mutation is the most widely applied genetic approach; however, it is a chronic manipulation that is only applicable to non-essential genes. Furthermore, compensatory suppressor mutations can readily emerge, or the physiology of the cell can adapt to lessen or alter the severity of the phenotype. Consequently, conditional mutations are often preferred because their function is retained until a switch to the restrictive conditions, such as altered temperature or growth medium composition, inactivates the molecule (Aguilera, 1994; Bartel and Varshavsky, 1988; Jimenez and Oballe, 1994; Poloni and Simanis, 2002). Such traditional point mutant approaches are now being complemented by the powerful method of acute destruction of fusion proteins through portable heat-inducible or auxin-inducible degradation tags (Dohmen et al., 1994; Labib et al., 2000; Nishimura et al., 2009).
Pharmacological inhibition also provides an acute, and often reversible, inactivation. Inhibitor docking at the active or allosteric site of an enzyme can achieve a high level of specific inhibition, however, inhibitor development is labour-intensive and can be plagued by off-target effects. The highly conserved structure of the ATP-binding pocket of eukaryotic kinases has made off-target effects a particular issue for the development of kinase inhibitors. This significant issue has been ingeniously turned to the investigators advantage by Kevan Shokat and colleagues in an approach that combines genetics with pharmacology to generate a kinase that is genetically sensitised to inhibition by generic ATP analogues (Bishop et al., 2000).
These inhibitor-sensitive kinase alleles are created by mutating the ‘gatekeeper’ residue in the catalytic domain to smaller glycine [analogue-sensitive 1 (as1)] or alanine [analogue-sensitive 2 (as2)] residues to ‘enlarge’ the ATP-binding pocket (Alaimo et al., 2001; Bishop et al., 2000). A generic bulky ATP analogue can then competitively inhibit the activity of the sensitised kinase, without impeding the function of any other kinase in the cell because their ATP-binding pockets remain too small to accommodate the bulky analogue (Bishop et al., 2000; Cipak et al., 2011). It is sometimes necessary to introduce a further mutation in the −1 position of DFG motif to enhance the sensitivity of either as1 or as2 alleles [generating so-called as3 and as4 alleles, respectively (Bishop et al., 2000; Blethrow et al., 2004; Zhang et al., 2005)] or a mutation to restore catalytic activity when it has been compromised by sensitising mutations (Zhang et al., 2005). Collectively, these options generate an array of ‘kinase.as’ alleles that can be tested with a bank of eight generic ATP analogues to identify an effective combination for a particular kinase (Cipak et al., 2011). This chemical–genetic approach to acute abolition of function has proved highly effective in a range of eukaryotes (Bishop et al., 2000; Brodersen et al., 2006; Burkard et al., 2007; Fleissner, 2013; Grallert et al., 2012; Hochegger et al., 2007; Kim et al., 2008; Wan et al., 2004; Wang et al., 2003). Furthermore, analogue-sensitive alleles can also deliver kinase activity at inappropriate times or locations (in genetic terms, an acute dominant gain-of-function allele). In this approach, analogue-sensitising mutations are combined with mutations that confer constitutive activation [such as acidic mutation of activating phosphorylation sites in the T loop mutation of AGC kinases (Pearce et al., 2010)]. Constitutive activity is restrained through analogue inhibition until being abruptly released upon transition to an analogue-free environment (Grallert et al., 2013b).
Although, analogue sensitisation is a highly effective approach, some kinases remain refractory to strong inhibition even after the introduction of all of the canonical sensitising mutations identified to date (Cipak et al., 2011), suggesting that further mutation could boost the efficacy of these canonical mutations. We previously identified a mutation that converted the marginally sensitive plo1.as3 mutation into a highly sensitive plo1.as8 allele (Grallert et al., 2013b). We were led to this mutation because the catalytic domain of the fission yeast polo kinase Plo1 has 90% similarity (47% identity) to that of the human polo kinase PLK1 and yet PLK1as is highly sensitive to analogue inhibition whereas plo1.as3 is not (Burkard et al., 2007). We therefore exploited this high level of similarity to seek differences that could account for the intractability of the highly related Plo1. The crystal structure of PLK1 solved in the presence of the ATP analogue AMPPNP suggested that the hydrophobic interactions made by a phenylalanine two residues after the HRDLKxxN motif in subdomain VIb (Hanks and Hunter, 1995) made an important contribution to ATP docking (Kothe et al., 2007). Mutation of the methionine (M170) found at this position in Plo1 to phenylalanine generated the plo1.as8 allele that was effectively inhibited by 3-MB-PP1 (Grallert et al., 2013b). We now extend these analyses to show that this M→F mutation confers sensitivity upon the three out of the five S. pombe kinases that harbour methionine at this position that we have tested. We exploit the enhanced sensitisation conferred by the M→F mutation in the wee1.as8 allele to generate highly synchronised mitoses in which protein function in either mitosis or S phase can be readily interrogated through temperature-sensitive mutation in the molecule of interest.
The M→F mutation enhances the sensitivity of casein kinase II to analogue inhibition
We previously established that acute administration of the ATP analogue 3MB-PP1 to asynchronous plo1.as8 cultures blocked mitotic progression as cells arrested cell cycle progression in mitosis with the monopolar spindle phenotype that is the hallmark of extensive or complete loss of Plo1 function (Grallert et al., 2013b; Ohkura et al., 1995). Chronic exposure to analogue through growth on solid medium reiterated the acute impact of analogue inhibition in liquid culture (Fig. 1A) and established that plo1.as8 is most effectively inhibited by 3BrB-PP1 (Fig. 1B). In log-phase liquid culture Plo1.as8 protein levels were indistinguishable from those of wild-type Plo1 3 hours after analogue inhibition at which time kinase activity is severely perturbed (Fig. 1C) (Grallert et al., 2013b).
The striking impact of the M170F mutation upon Plo1 analogue sensitivity prompted us to ask whether the same substitution would enhance the analogue sensitivity of other S. pombe kinases that harboured a methonine residue at this position: Orb5, Pat1, Srb10 and Wee1 (Fig. 2A). As Srb10 kinase is a non-essential kinase for which testable genetic interactions are currently lacking, we mutated the remaining three; however, the pat1.as8 mutant was inviable (supplementary material Fig. S1) limiting our analysis to Orb5 and Wee1. Orb5 encodes fission yeast casein kinase II. It is an essential kinase that establishes growth polarity (Snell and Nurse, 1994). orb5.as2 displayed moderate sensitivity to 30 µM of 3BrB-PP1, whereas orb5.as1 showed none (Fig. 2B). The M167F mutation in either the orb5.as1 or orb5.as2 backbone generated the orb5.as8 and orb5.as9 alleles, which were more sensitive to analogue inhibition than the respective parental allele (Fig. 2B). All alleles were most effectively inhibited by either 3BrB-PP1 or NM-PP1 (Fig. 2C).
orb5.as8 inhibition did not produce the ‘orb’ phenotype observed in the original orb5.ts mutants at the restrictive temperature (data not shown), despite the fact that orb5.as8 cells show sensitivity to 30 µM 3BrB-PP1 (Snell and Nurse, 1994). This is consistent with the observation in Cipak et al. and suggests that either the level of inhibition exerted by the inhibitors is insufficient or that the orb phenotype observed in the temperature-sensitive mutant might not be the direct consequence of the withdrawal of the Orb5 kinase activity (Cipak et al., 2011). To test this latter possibility we shifted orb5.as8 cells to 36°C as we added the inhibitor to assess whether the Orb phenotype would emerge as a combined consequence of Orb5 inhibition and heat shock, however, no Orb phenotype was seen. We therefore favour the interpretation that residual Orb5 activity persists in orb5.as8 analogue-treated cells.
The M→F mutation enhances Wee1 sensitivity to analogue inhibition
The timing of mitotic commitment is regulated by the activity of the Cdk1–Cyclin-B kinase (Nurse, 1990). The complex is maintained in an inactive state throughout interphase through phosphorylation of the tyrosine 15 of the catalytic Cdk1 subunit by Wee1 kinase (Featherstone and Russell, 1991; Gould and Nurse, 1989). Mitotic commitment is then triggered by the removal of this inhibitory phosphate by the phosphatase Cdc25 (Kumagai and Dunphy, 1991; Millar et al., 1991; Russell and Nurse, 1986). The timing of mitotic commitment is therefore determined by the balance of Wee1 and Cdc25 activities. Inappropriate withdrawal of Wee1 kinase activity by either gene deletion, point mutation or temperature-dependent inactivation of the wee1.50 allele accelerates the timing of mitotic entry to reduce cell length at division (Fig. 3A) (Nurse, 1975; Nurse and Thuriaux, 1980; Russell and Nurse, 1987). Analogue addition to a wee1.as1 allele reduces cell size at division (Masuda et al., 2011), indicating that kinase activity is inhibited; however, the limited reduction in cell length suggests that wee1.as1 cells retain considerable kinase activity in the presence of ATP analogue (Fig. 3A; Table 1). We therefore asked whether the introduction of the M→F mutation two residues after the HRDLKxxN motif in wee1.as1 (Fig. 2A), to create wee1.as8, would enhance the severity of Wee1 inhibition.
The indicated strains were grown to mid-log phase at 25°C in EMM2 before the addition of either methanol or 3BrB-PP1 for 3 hours. Underlined letters denote the native residue; bold letters indicate the substituted residue in the respective mutant allele. Cell lengths are means ± s.d.
Addition of 20 µM 3BrB-PP1 to wee1.as8 cells reduced the size threshold for cell division below that of either wild type or wee1.as1 (Table 1; Fig. 3A). Wee1.as8 protein levels were indistinguishable from wild-type Wee1 levels over the first 3 hours of analogue addition (data not shown). We exploited the classical genetic antagonism between wee1+ and cdc25+ (Fantes, 1979) to compare the analogue sensitivities of wee1.as1 and wee1.as8. Loss of cdc25.22 phosphatase activity at 36°C results in the accumulation of Cdk1 that is phosphorylated on tyrosine 15, which blocks mitotic commitment (Gould et al., 1990; Hagan and Hyams, 1988). Because the essential function of Cdc25 is to remove this phosphate, the temperature sensitivity of cdc25.22 strain is alleviated by simultaneous removal of Wee1 activity (Fantes, 1979; Russell and Nurse, 1987) (Fig. 3B). We therefore compared the ability of wee1.as1 and wee1.as8 to suppress cdc25.22 lethality at 36°C. Inhibition of Wee1.as8 but not Wee1.as1 activity with analogue addition suppressed cdc25.22 lethality at 36°C (Fig. 3B). A comparison of four ATP analogues revealed that the suppression (and therefore Wee1 inhibition) was most effective with 3BrB-PP1 (Fig. 3C).
In summary, by enhancing a hydrophobic interaction with a methionine to phenylalanine mutation, we enhanced the analogue sensitivity of all three fission yeast kinases that were tested: Plo1, casein kinase II and Wee1.
G2-arrested wee1.as8 cdc25.22 cells execute mitosis immediately after Wee1 inhibition
Transient inhibition of Cdc25 function by incubation of cdc25.22 cells at 36°C for 4.25 hours is a widely used approach to synchronise mitotic progression in fission yeast (Hagan, 1988; Moreno et al., 1989). To date, the need for a shift back to 25°C to induce release from arrest and the synchronous wave of mitosis has meant that molecular functions can only be interrogated in the ensuing mitosis by pharmacological intervention, and not by any of the large range of temperature-sensitive mutations previously isolated by the community. This limitation prompted us to ask whether wee1.as8 analogue inhibition might be used as an alternative means to release the cdc25.22-imposed cell cycle arrest at 36°C (when temperature-sensitive alleles would be inactivated).
An early log phase wee1.as8 cdc25.22 culture was arrested by increasing the temperature from 25°C to 36°C for 4.25 hours before addition of 30 µM 3BrB-PP1-inhibited Wee1.as8. Microtubule staining revealed a highly synchronous progression through mitosis that was completed within 60 minutes after analogue addition (Fig. 4A). Blotting samples to detect the spindle pole body (SPB) component Cut12 and an epitope-tagged version of the anaphase-promoting complex (APC) component Cut9 revealed characteristic waves of Cut12 and Cut9 phosphorylation as cdc25.22 wee1.as8 cut9.HA cells transited this highly synchronous mitosis (Grallert et al., 2013a; Yamada et al., 1997) (Fig. 4B). To directly compare the kinetics of mitotic activation in the temperature- and analogue-release procedures, an early log phase wee1.as8 cdc25.22 culture was arrested by increasing temperature from 25°C to 36°C for 4.25 hours before being split in two. One half was returned to 25°C, while 30 µM 3BrB-PP1 was added to the other. Samples were removed for histone H1 kinase assays following precipitation of active Cdc2–cyclin-B complexes with p13Suc1 beads. The kinetics of histone H1 kinase activity in the two approaches were strikingly similar (Fig. 4C) with the analogue release displaying a slightly sharper peak indicative of greater synchrony.
We assessed the utility of this synchronisation approach for functional analysis by live-cell imaging with a wee1.as8 cdc25.22 strain in which the SPB component Sid4 had been tagged with tdTomato and the α-tubulin Atb2 with GFP. To monitor the specificity of the response, otherwise isogenic wee1+ cdc25.22 cells that had been labelled by transient immersion in lectin-TRITC were mixed with the wee1.as8 cdc25.22 cells before mounting for microscopy. As shown in Fig. 5 and supplementary material Movie 1, all wee1.as8 cdc25.22 cells initiated spindle formation 10 minutes after ‘analogue release’, and completed mitosis with the appearance of post-anaphase array (Hagan and Hyams, 1988) within 55 minutes of analogue addition (Fig. 5; supplementary material Movie 1). In contrast, neighbouring lectin-labelled control cdc25.22 cells (marked by * in Fig. 5) retained the cytoplasmic microtubule array characteristic of G2 arrested cells (Hagan and Hyams, 1988) throughout the experiment.
Perturbation of mitotic progression in a wee1.as8 cdc25.22 analogue-release synchronised division by temperature sensitive mitotic mutations
We next asked whether the high level of synchrony generated by this wee1.as8 cdc25.22 ‘analogue-release’ approach could be utilised to assess the mitotic function of the molecule of interest with a temperature-sensitive loss-of-function mutation. The compromised APC activity arising from incubation of cut9.665 cells at 36°C blocks the metaphase–anaphase transition leading to an accumulation of cells with a metaphase plate of condensed chromosomes (Hirano et al., 1986; Samejima and Yanagida, 1994). Analogue-released wee1.as8 cdc25.22 cut9.665 cells transiently accumulated much higher levels of metaphase spindles than wild-type cells before they ‘leaked’ through this mitotic arrest to execute telophase and cytokinesis with the classic ‘cut’ phenotype that originally led to the identification of the cut9.665 mutation (Fig. 6A,B) (Hirano et al., 1986). Mitotic progression of analogue-released wee1.as8 cdc25.22 cells was also blocked at 36°C by perturbation of the kinesin 5 function that drives the inter-digitation of the two halves of the bipolar spindle microtubules. Inactivation of kinesin 5 with the temperature-sensitive cut7.24 mutation arrested mitotic progression with the monopolar spindle phenotype and condensed chromosomes that are the hallmarks of a deficiency in Cut7 function (Fig. 6C) (Hagan and Yanagida, 1990; Tallada et al., 2009).
wee1.as8 cdc25.22 arrest–release in the analysis of DNA replication
In fission yeast, the size threshold for the G2/M transition determines the timing of cell division. The size control for transition of the G1/S boundary is lower than that at which cells are born (Nurse, 1975). Consequently S phase is completed before cytokinesis so that the execution point for genes that act at G1/S is −0.1 of a cell cycle [i.e. at the end of the previous cycle (Nurse et al., 1976)] and DNA replication factors first associate with chromatin mid-way through anaphase (Kearsey et al., 2000). The cdc25.22 temperature arrest–release approach is therefore highly effective in the analysis of DNA replication and associated checkpoints (Heichinger et al., 2006). We therefore asked whether the analogue arrest–release approach would prove similarly effective.
We used flow cytometry to compare the timing of DNA replication following temperature-dependent release of the same wee1.as8 cdc25.22 culture. An early log phase wee1.as8 cdc25.22 culture was shifted from 25°C to 36°C for 4.25 hours before being split into two. One half was returned to 25°C, while the ATP analogue, 3BrB-PP1 (30 µM) was added to the other half. Flow cytometric measurement of DNA content revealed that the portion of the culture that was released by conventional temperature-shift-initiated S phase 50 minutes after release from G2 arrest (Fig. 7A). A strikingly similar profile was seen 40 minutes after the suppression of Wee1 activity by addition of 30 µM 3BrB-PP1 (Fig. 7B). Addition of 10 mM hydroxyurea, to deplete the nucleotide pool and stall replication fork progression, completely arrested cell cycle progression of analogue-released cells in S-phase (Fig. 7B central panel), indicating that the integrity of the S-phase checkpoint was preserved following analogue addition. We then asked whether the cdc10.v50 mutation, which blocks the transcription of DNA replication factors at 36°C (Lowndes et al., 1992; Marks et al., 1992), would prove as effective at compromising DNA replication in this wee1.as8 cdc25.22 analogue-release approach as it does in a simple shift from 25°C to 36°C (Marks et al., 1992). The static FACS profiles established that replication was indeed inhibited in analogue-released cdc10.v50 wee1.as8 cdc25.22 cells after analogue addition (Fig. 7B, right panel).
Collectively, these results indicate that the wee1.as8 cdc25.22 analogue-release strategy provides a highly effective approach with which to exploit temperature-sensitive mutations to interrogate molecular functions within either S phase or mitosis.
The chemical genetic strategy introduced by the Shokat laboratory has created a powerful approach with which to study protein kinase function in different model systems. The approach has been more widely embraced by the genetic than the biochemical communities. The basis for this discrepancy probably lies in the assays employed in each case. With genetic analyses the assessment is based on an in vivo analysis of protein function. The question is simple: does the cell execute the functions controlled by the mutated kinase when the analogue is absent, but not when it is present? By these criteria, seven of the eight kinases we have tested have been rendered analogue sensitive (data not shown). A comparable degree of success has been achieved through the systematic introduction of as1 and as2 mutations into genes encoding fission yeast kinases (Cipak et al., 2011). In contrast, attempts to exploit this approach in biochemical assays fail more often than they succeed because the sensitising mutations depress activity below detectable thresholds. Such a radical fall in our ability to detect activity might stem from the fact that most kinase assays are on non-physiological substrates in non-physiological buffers, and at non-physiological pH and concentrations. They are therefore far from the native conditions that are assessed in the genetic exploitation of the analogue-sensitising approach. In other words, a cell can probably tolerate a reduction of kinase activity by an order of magnitude, or more, whereas crude and artificial in vitro kinases assays cannot. Indeed, the budding yeast cdc28.as1 mutation reduces the kcat/Km of the in vitro kinase activity 50 fold and yet cell cycle progression is largely unimpaired (Bishop et al., 2000). Similarly, in fission yeast, the sid2.as4 and ark1.as3 mutations support cellular function in the absence of analogue (Grallert et al., 2012; Hauf et al., 2007) and yet activity in established kinase assays is virtually undetectable (data not shown).
In this paper, we show that mutation of a methionine residue two residues after the consensus HRDLKxxN of the ATP-binding pocket to phenylalanine enhances the sensitivity of as1 or as2 kinases to inhibitors in all the cases that we have tested in fission yeast (plo1, orb5 and wee1) (Fig. 8). While this strategy has proved highly effective in fission yeast, we note that a methionine is found at this position of a number of kinases in other systems in which analogue-sensitising approaches are being successfully applied (supplementary material Table S1). Should any of these kinases prove refractory to inhibition after introduction of the canonical sensitising mutations, we propose that it might be profitable to add the M→F mutation.
Synchronisation of cell cycle progression in a population of cells is an indispensable technique that is widely employed to interrogate molecular function in cell cycle control and execution. Three major synchronisation approaches are frequently employed in S. pombe: nda3.KM311 and cdc25.22 arrest–release approaches and centrifugal elutriation. Centrifugal elutriation relies upon the direct relationship between the sedimentation rate during centrifugation and cell size (Creanor and Mitchison, 1982). As DNA replication is completed before cytokinesis, selection of the small cells from an asynchronous culture generates a culture of early G2 cells that progress synchronously through the cell cycle.
Transient cell cycle arrest with the cold-sensitive nda3.KM311 mutation offers a simpler alternative to elutriation. nda3+ encodes the sole β-tubulin of fission yeast. nda3.KM311 cells are unable to form microtubules at 20°C. Cell cycle progression of an asynchronous nda3.KM311 culture is therefore arrested within mitosis through the activation of the spindle assembly checkpoint (SAC) (Hiraoka et al., 1984; Vanoosthuyse and Hardwick, 2009). A return to temperatures above the permissive temperature of 30°C restores β-tubulin function enabling rapid spindle formation, checkpoint silencing and a highly synchronous and uniform progression through mitosis throughout the population. This technique is highly effective for interrogating the molecular functions of proteins with temperature-sensitive mutations (Tamm et al., 2011; Uemura et al., 1987); however, the temperature-sensitive mutation in the second protein must permit full growth at 30°C (the permissive temperature for nda3.KM311). An alternative mode of release from the nda3.KM311 arrest point has proved equally effective in the analysis of recovery from SAC-dependent cell cycle arrest. In this approach, an analogue-sensitive allele (Hauf et al., 2007) of the aurora kinase Ark1 that triggers SAC activation when microtubules fail to form (Petersen and Hagan, 2003) is inhibited by analogue addition to ark1.as3 nda3.KM311 cells to drive them through mitosis into interphase in the absence of spindle formation (Vanoosthuyse and Hardwick, 2009).
Cell cycle arrest in the cdc25.22 arrest–release protocol is imposed by incubation at 36°C for 4.25 hours to prevent the removal of an inhibitory phosphate from Cdc2, thereby arresting cell cycle progression in G2 phase. A return to 25°C re-activates Cdc25 to remove the phosphate and so promotes a highly synchronous mitotic commitment and progression. This protocol has not been of great use as a means to address functional questions with temperature-sensitive mutants because the mitotic release is performed at 25°C. Because the G2 arrest can be released by inactivating Wee1 through analogue addition, wee1.as8 cdc25.22 now removes this barrier to permit the exploitation of temperature-sensitive mutants that can be inactivated at 36°C. Presumably either Pyp3 (Millar et al., 1992) or residual Cdc25 activities, or a combination of the two, are sufficient to remove the inhibitory phosphate from Cdc2 upon Wee1 inhibition. Proof of principle experiments established that the synchronous mitotic progression induced by analogue addition to arrested wee1.as8 cdc25.22 cells was completely blocked when the function of either the APC or kinesin 5 were abolished with temperature-sensitive mutations.
The inherent coupling of mitosis and S phase that stems from the size thresholds for transition of the G2/M and G1/S boundaries (Nurse, 1975) means that the synchronisation of cell cycle progression through transitory inactivation of Cdc25 has proved as useful for the analysis of DNA replication as for mitotic functions (Heichinger et al., 2006). We show that analogue release proved as effective as the traditional temperature-dependent release at triggering a synchronous S phase that displayed a robust replication checkpoint. The analogue-release approach will therefore pose a highly effective tool with which to use temperature-sensitive mutations in the analysis of both genome replication and segregation. The generation of a temperature-sensitive mutation in a gene of interest through marker switch, marker reconstitution and recombinase-mediated cassette exchange (RCME) technologies is well established (MacIver et al., 2003; Tang et al., 2011; Watson et al., 2008), setting the scene for a rapid pipeline for the detailed interrogation of molecular function within either S or M phase using the wee1.as8 cdc25.22 analogue arrest release approach.
Materials and Methods
Strains used in this study are listed in supplementary material Table S2. Cells were cultured in liquid Edinburgh minimal medium number 2 (EMM2) or on solid YES medium as indicated (Moreno et al., 1991). ATP analogues (Toronto Research Chemicals, Dalton Pharma Services) were dissolved in methanol to generate 50 mM stock solutions that were subsequently added to cultures or cooled molten YES agar.
Mutations were generated with the Phusion® site-directed mutagenesis (New England Biolabs) protocol on DNA fragments that had been amplified by PCR and cloned into TOPO® vector (Life Technologies). Mutant alleles were introduced into the native loci by the marker switch approach (MacIver et al., 2003) in a pku80.Δ background. Correct integration events were identified by PCR analysis and backcrossed using tetrad dissection to eliminate pku80.Δ.
Tubulin immunofluorescence, DAPI, lectin and Calcofluor White staining were conducted using established procedures (Grallert et al., 2013a). Tissue culture supernatant from the TAT1 monoclonal line (Woods et al., 1989) (gift from K. Gull, University of Oxford) was used at 1∶80. All live-cell microscopy was conducted on a DeltaVision system (Applied Precision) fitted with a Zeiss ×100, 1.45 NA objective in conjunction with Softworx (Applied Precision) and Imaris (Bitplane) software. Cells were mounted on glass-based culture (IWAKI) plates. Images were maximal projections of 33 sections with 0.3 µm between the slices; images were taken every 5 minutes.
DNA content analysis was performed by FACS analysis and used Cytox Green (Invitrogen) according to published procedures (Knutsen et al., 2011).
Western blots were probed with affinity purified anti-Cut12 antibodies (Bridge et al., 1998) (1∶100), the 12CA5 monoclonal antibody that recognises the HA epitope (Roche; 1∶500), the PN24 polyclonal Cdc2 antibody (1∶500) (Simanis and Nurse, 1986), polyclonal anti-Plo1 antibodies (1∶100) (Mulvihill et al., 1999) or the α tubulin monoclonal antibody TAT1 (1∶2000) (Woods et al., 1989).
For the histone H1 kinase assays, 15 µl p13Suc1 beads (Millipore) were collected by centrifugation at 3000 g for 2 minutes and washed twice with 0.5 ml immunoprecipitation (IP) buffer [50 mM Tris-HCl pH 8.0, 250 mM NaCl, 50 mM NaF, 5 mM EDTA, 0.1 mM Na3VO4, 0.1% Triton X-100, 1 mM DTT, 1 mM PMSF and protease inhibitor cocktail (Roche)]. Cell pellets (2×108 cells) were lysed by chilled glass beads in 0.5 ml IP buffer using cell disrupter (Fastprep FP120, Qbiogene). Cleared supernatants were incubated with washed p13Suc1 beads for 1 hour at 4°C. p13Suc1 beads were washed three times with IP buffer and twice with kinase buffer (50 mM Tris-HCl pH 7.5, 10 mM MgCl2, 20 mM EGTA, 5 mM KCl, 1 mM sodium β-glycerolphosphate, 1 mM DTT). 10 µl of kinase reaction (2 µg histone H1, 0.1 mM ATP, 0.4μ Ci P32-ATP in kinase buffer) was added to the washed beads and incubated at 30°C for 30 minutes. The kinase reaction was stopped by adding 5 µl Laemmli sample buffer and heating at 70°C for 10 minutes.
We thank: Viesturs Simanis (EPFL/ISREC, Switzerland) and Kayoko Tanaka (University of Leicester, UK) for strains, and Keith Gull (University of Oxford, UK) and Nimesh Joseph (CRUK Cell Division Group) for antibodies, Ben Hodgson for assistance with FACS analysis, and both Danny Bitton (University College London, UK) and Prasad Jallepalli (Memorial Sloan-Kettering Cancer Center, USA) for highly stimulating and informative discussions.
Y.D.T. did all the work presented after initial background discussions with A.P. D.K. constructed some further as alleles that are not included in the manuscript. Y.D.T. and I.H. conceived the study and wrote the manuscript.
This work was supported by Cancer Research UK (CRUK) [grant number C147/A6058 to Prof. Richard Marais].