|Home | About | Journals | Submit | Contact Us | Français|
Sumoylation (a covalent modification by Small Ubiquitin-like Modifiers or SUMO proteins) has been implicated in the regulation of various cellular events including cell cycle progression. We have recently identified CDK1, a master regulator of mitosis and meiosis, as a SUMO target both in vivo and in vitro, supporting growing evidence concerning a close cross talk between sumoylation and phosphorylation during cell cycle progression. However, any data regarding the effect of sumoylation upon CDK1 activity have been missing.
In this study, we performed a series of in vitro experiments to inhibit sumoylation by three different means (ginkgolic acid, physiological levels of oxidative stress, and using an siRNA approach) and assessed the changes in CDK1 activity using specific antibodies and a kinase assay. We have also tested for an interaction between SUMO and active and/or inactive CDK1 isoforms in addition to having assessed the status of CDK1-interacting sumoylated proteins upon inhibition of sumoylation. Our data suggest that inhibition of sumoylation increases the activity of CDK1 probably through changes in sumoylated status and/or the ability of specific proteins to bind CDK1 and inhibit its activity.
Cell cycle progression is regulated by various post-translational modifications of which phosphorylation has been the most studied. Cyclin-dependent kinases, or CDKs, are serine-threonine kinases that regulate key events in cell cycle progression. CDKs are catalytically active only when bound to a regulatory cyclin. CDK inhibitors, or CKIs, on the other hand, block kinase activity in the cyclin-CDK complex [1,2,3]. Furthermore, for many CDKs to be completely active, an activating phosphorylation or dephosphorylation must take place on a specific tyrosine, serine, and/or threonine residue(s). For example, a highly conserved phosphorylation on tyrosine (Tyr) 15 by Wee1 kinase leads to inhibition of the activity of CDK1; whereas members of the CDC25 phosphatase family counteract this activity . Antibodies against either activating or inhibitory phosphorylation are helpful in the assessment of the activity status of specific kinases.
Sumoylation is yet another type of post-translational modification by Small Ubiquitin-like Modifiers (or SUMO proteins) that has been identified as an important regulatory mechanism of several cellular processes including cell cycle progression [5,6,7,8,9]. Four SUMO paralogs have been identified: SUMO1, 2, 3 (often termed SUMO2/3 because of their 95% sequence identity), and 4. SUMO1, 2, and 3 are abundantly expressed in different tissues, but SUMO4 is restricted to the kidney, liver, and lymph nodes [10,11,12]. The process of sumoylation involves maturation, activation, conjugation, and ligation steps mediated by an E1 activating enzyme (the Aos1-UBA2 heterodimer), an E2 conjugating enzyme (UBC9), and one of the SUMO E3 ligases, respectively. Sumoylation often occurs on a lysine residue within a consensus sequence: ψ-K-X-D/E, where ψ indicates a hydrophobic amino acid and X indicates any amino acid. However, not all consensus sequences are sumoylated, and sumoylation often occurs outside of the consensus sequences. Sumoylation is reversed through the action SENPs that cleave the isopeptide bond between SUMO and its substrate. SUMO targets include factors that regulate cell cycle progression, transcription, translation, and protein transport. In addition to the numerous identified targets of sumoylation, there is a growing list of proteins that interact with SUMO non-covalently [13,14,15,16,17].
It was discovered that ginkgolic acid (GA), an extract from Ginkgo biloba leaves, is capable of inhibiting sumoylation without significantly affecting other cellular processes. The GA directly binds SUMO-activating enzyme (E1) and inhibits the formation of the E1-SUMO intermediate . This inhibitor has been used in previous studies of sumoylation [19, 20]. It has also been established that SUMO-conjugating enzymes are highly sensitive to oxidative stress. Very low/physiological concentrations of H2O2 specifically affect SUMO-conjugating machinery, causing desumoylation before other processes are activated in the cells  . Several studies have provided evidence that sumoylation and phosphorylation interact at multiple levels. A sumoylation-dependent phosphorylation and phosphorylation-dependent sumoylation have been identified [23,24,25], and inhibition of sumoylation by the sumoylation inhibitor GA significantly decreased tyrosine phosphorylation of multiple proteins . In accordance with these data, using immunoprecipitation followed by mass spectrometry identification, we have recently identified several kinases as targets of sumoylation in mouse germ cells (meiotic spermatocytes and spermatids) . We have also observed significant changes in germ cell phosphorylation patterns (including specific phosphorylation events required for meiotic progression) upon inhibition of sumoylation with GA (unpublished data). One of the interesting SUMO targets identified at our published screen was CDK1 kinase, a crucial and indispensable regulator of both mitotic and meiotic G2/M progression [26,27]. An in vitro sumoylation assay supported possible sumoylation of CDK1; and co-immunoprecipitation experiments using mouse germ cell and human HEK cell lysates confirmed possible covalent and non-covalent interactions between CDK1 and SUMO [26,27]. A bioinformatics analysis revealed the presence of the consensus sumoylation site in the amino acid sequence of the mouse but not the human CDK1; However, the alignment of the two sequences revealed a difference in only one amino acid, with a possible target lysine still present at the same position [26,27]. Interestingly, another important cell cycle regulator, CDK2 (not identified by our screen), contained no such sequence. Notably, CDK1 was also identified as a SUMO target in Drosophila embryos, supporting our finding and suggesting a possible conserved role of sumoylation in the regulation of CDK1 activity . Furthermore, in a phosphoproteome screen, CDK1 phosphorylation levels had seemingly decreased upon inhibition of sumoylation with GA . Given the importance of both activating and inhibitory phosphorylation in the regulation of the CDK1 activity, these findings suggest that sumoylation can regulate CDK1 activity during cell cycle progression. However, how the activity of CDK1 is affected by sumoylation is not currently known.
In this study, we performed a series of in vitro experiments to inhibit sumoylation by three different means (GA, physiological levels of oxidative stress, and using an siRNA approach) and assessed the changes in CDK1 activity using specific antibodies and a kinase assay. We have also tested for an interaction between SUMO and active or inactive CDK1 isoforms, and, additionally, assessed the status of CDK1-interacting sumoylated proteins upon inhibition of sumoylation.
Our data suggest that inhibition of sumoylation increases the activity of CDK1 probably through changes in the sumoylation status and/or ability of specific sumoylated proteins to bind CDK1 and inhibit its activity.
HEK 293 (ATCC® CRL-1573™) cells and the type B spermatogonia-derived GC1 (ATCC® CRL2053™; ) cells were purchased from ATCC (Manassas, VA) and grown in DMEM media (11995-065, Life Technologies, Carlsbad, CA) with 5% fetal bovine serum (FBS, 16140-071, Life Technologies), 5% bovine growth serum (SH3054103HI, Fisher Scientific, Carlsbad, CA), 1% penicillin/streptomycin (15140-122, Life Technologies), and 0.5% Fungizone (15290-018, Life Technologies) at 37°C with 5% CO2.
Cells were treated with 200 μM of freshly reconstructed and diluted ginkgolic acid (75741, Sigma-Aldrich, St. Louis, MO or DMSO (Sigma-Aldrich) for 6 hours and 2 μM of H2O2 for 1 hour. The dosage and timing of the treatment were chosen based on previously published data on inhibition of sumoylation and previous studies in our laboratory [19,22].
UBC9 and control siRNAs were purchased from Santa Cruz Biotechnology Inc. (sc-36773 & sc-36869, Dallas, TX). 0.5 × 106 GC1 cells were seeded onto separate 10-cm tissue culture dishes and grown overnight at 5% CO2 and 37°C. 80 pmol of siRNA were transfected into each dish of GC1 cells using Lipofectamine® RNAiMAX Transfection Reagent (13778030, Life Technologies) at a ratio of 0.3 μl of reagent per pmol of siRNA. The transfection procedure followed the manufacturer’s instructions. The cells were subjected to 6 hours of transfection followed by a 48-hour recovery period prior to analysis.
Whole cell protein lysates were prepared as previously described, using the whole cell extraction kit and protease inhibitor from Millipore (2910, Billerica MA) complemented with 2.5 mg/ml of N-ethylmaleimide (NEM (a de-sumoylation inhibitor), E3876-100G, Sigma-Aldrich, St. Louis, MO) and phosphatase inhibitors (88667, Thermo Fisher Scientific, Rockford, IL) according to the manufacturer’s instructions.
Protein concentrations were determined via bicinchoninic acid (BCA) protein assay (23225, Thermo Fisher Scientific, Rockford, IL) using bovine serum albumin (BSA) as the standard. Gel electrophoresis was performed under reducing conditions using NuPAGE 4–12% gradient bis-tris polyacrylamide gels (Life Technologies, Carlsbad, CA, USA), and MOPS running buffer (Life Technologies, Carlsbad, CA, USA). Protein electrophoresis (and transfer) was performed using the Invitrogen XCell SureLock Mini-Cell electrophoresis system (Life Technologies, Carlsbad, CA, USA) at a constant voltage (200 V). After electrophoresis, proteins were transferred to a nitrocellulose membrane (Novex nitrocellulose membrane, 0.45 μm pore size, Life Technologies, Carlsbad, CA, USA) using NuPAGE transfer buffer (Life Technologies, Carlsbad, CA, USA). The membrane was first blocked with 2% membrane blocking agent (RPN2125V, GE Healthcare UK Limited, Little Chalfont, UK) in PBS-T: PBS + 0.02% (v/v) Tween 20 (00-3005, Life Technologies Carlsbad, CA) for 1 hour at room temperature. The membrane was then incubated with primary antibodies in PBS containing 2% BSA and 0.1% sodium azide for either 2 hours at room temperature or overnight at 4°C. Following three washes with PBS-T, the membrane was further incubated with secondary antibodies that were diluted to 1:5000 in PBS-T for 1 hour at room temperature. Rabbit polyclonal anti-SUMO1 antibody (ab32058) was purchased from Abcam (Cambridge, MA) and used in a 1:500 dilution; mouse monoclonal anti-CDC2 antibody (sc-54) was purchased from Santa Cruz Biotechnology and used in a 1:100 dilution; rabbit polyclonal anti-phosphorylated CDC2 (Tyr15) antibody (9111S, an inhibitory phosphorylation) was purchased from Cell Signaling Technology (Danvers, MA) and used in a 1:500 dilution; equal loading was ensured with monoclonal anti-β-actin antibody (sc-1615, Santa Cruz) in a 1:1000 dilution. The secondary antibodies used in this study included the following: ECLTM anti-rabbit IgG HRP linked (NA934V, GE Healthcare UK Limited, Little Chalfont, UK), and goat anti-mouse IgG (H + L) HRP (AP308P, EMD Millipore Corporation, Billerica MA). Western blot detection was performed using Luminata™ Forte (EMD Millipore Corporation, Billerica, MA), in accordance with the manufacturer’s instructions. Quantitative densitometry analyses were performed using the Quantity One software (Bio-Rad Laboratories, Hercules, CA) and the density values were normalized to actin. In each experiment, controls (untreated samples) were considered as 1 and other samples were normalized to the controls. To calculate the difference between samples, Student’s paired t-test was used. P values <0.05 were considered statistically significant.
Anti-SUMO (sc-5308 AC) and anti-CDC2 agarose conjugates (sc-54 AC) were purchased from Santa Cruz.
The Lysates were pre-cleared by a 1-hour agitation with 40 μl of Mouse IgG-Agarose (A0919-2ML, Sigma-Aldrich, St. Louis, MO) at 4°C on an orb ital rocker, and their concentrations were determined by BCA assay (see above). According to their concentrations, individual lysates, which contained 0.8 mg of protein, were diluted to a final volume of 1 ml with lysis buffer and mixed with either 80 μl of anti-CDC2 agarose conjugate (sc-54 AC, Santa Cruz) or 40 μl of Mouse IgG-Agarose (to achieve the same final IgG concentration). The lysate/antibody-agarose mixtures were subjected to agitation at 4°C overnight on an orbital rocker and then individually transferred into spin columns (IP/crosslink kit, 26147, Thermo Scientific, Rockford, IL) for a 1 min centrifugation at 4°C. The lysates were discarded as flow through from centrifugation and the antibody-agarose was retrieved by washing twice with 300 μl of 1 × kinase buffer (40 mM of Tris pH 7.5, 0.1 mg/ml of BSA, complemented with NEM at the final concentration of 2.5 mg/ml) at 4°C. Then 50 μl of kinase assay reaction mix was set up by adding 37.5 μl of ddH2O, 5 μl of 10 × kinase buffer (400 mM of Tris pH 7.5, 1 mg/ml of BSA, complemented with NEM at the final concentration of 2.5 mg/ml), 2.5 μl of CDC2 substrate CSH103 (sc-3065, Santa Cruz), and 5 μl of magnesium/ATP cocktail (20-113, EMD Millipore corporation) to each column at room temperature. The reaction mixtures were incubated at 30°C for 30 min followed by immediate chilling on ice and 1 min of centrifugation at 4000 rpm at room temperature for sample collection. 40 μl of each kinase assay reaction mixture was then transferred into a 96-well plate and individually mixed with Kinase-Glo® Plus Substrate (V378A, Promega corporation, Madison, WI) at a 1:1 ratio, incubated at room temperature for 10 min, and covered from light. Signals were detected by the GLOMAX multi detection system (Model#: E7031, Promega corporation instrument) using the Kinase-Glo protocol. Kinase activity was inversely proportional to the level of free/unused ATP. After the kinase assay, the antibody-agarose within the individual columns was washed twice with 200 μl of lysis buffer and once with 100 μl of conditioning buffer (IP/crosslink kit, 26147, Thermo Scientific, Rockford, IL) at 4°C. Proteins that were pulled down by the antibody-agarose were eluted with elution buffer (IP/crosslink kit, 26147, Thermo Scientific) at room temperature and analyzed by western blot to confirm successful IP. IP with anti-SUMO antibody was performed using a procedure described in our recent publication .
Our previously published sumoylation data were obtained in primary germ and human HEK cells, therefore, germ cell derived (GC1) and HEK cell lines were used in the experiments. In order to determine how inhibition of sumoylation affects the activity of CDK1, we first incubated the HEK and GC1 cells with sumoylation inhibitor ginkgolic acid (GA, 200 μM, 6 hours ) followed by a western blot analysis using anti-SUMO and -an inhibitory (Tyr 15) phosphorylated CDK1 isoform antibodies. Upon inhibition of global sumoylation, evident by the decrease in the high molecular weight (HMW) SUMO conjugates (Fig. 1A, three different experiments are shown), a significant decrease was observed in the level of inhibitory phosphorylation, suggesting an increase in CDK1 activity following GA treatment (Fig. 1A). In a different manner, a slight decrease in the level of total CDK1, assessed using an antibody that recognizes all isoforms of CDK1, was not statistically significant (Fig. 1A). An anti-actin antibody was used to confirm equal loading. Together, these data suggest that sumoylation inhibits CDK1 activity. As another means to inhibit sumoylation, we used low levels of oxidative stress (H2O2, 2mM; 1 hour [21,22]) and tested the levels of phosphorylated and total CDK1 isoforms. Similar to the results obtained with GA, when sumoylation was inhibited the level of the inhibitory phosphorylation significantly decreased (Fig. 1B, three different experiments are shown). Again, the slight decrease in the total level of CDK1 was not statistically significant.
Although GA and low levels of H2O2 have been suggested to specifically affect sumoylation machinery, their effects upon other pathways in the cells cannot be discounted. To confirm the obtained results, we have also performed siRNA experiments to inhibit SUMO-conjugating enzyme UBC9, and, therefore, sumoylation. Control and down-regulated samples from different experiments were run on western blot to assess the activity of CDK1 upon inhibition of sumoylation. Consistent with the results obtained from the use of GA and H2O2, upon inhibition of sumoylation with siRNA, only the inhibitory phosphorylation isoform of CDK1 decreased; again, suggesting that sumoylation inhibits CDK1 (Fig. 1C, three different experiments are shown).
We have also performed an immunoprecipitation with anti-CDK1 antibody followed by a kinase assay using CDK substrate to confirm the western blot results. Successful specific enrichment of CDK1 was confirmed using western blot (Fig. 2A). The kinase assay measures the amount of free ATP left after the reaction (Fig. 2B, Y-axis) and is thus inversely related to kinase activity. As can be seen in Fig. 2B, upon siRNA of sumoylation, there was a significant increase in the kinase activity, supporting our finding that sumoylation inhibits CDK1 kinase activity.
Based on our data, we also sought to test whether in vivo SUMO preferably interacts with active or inactive isoforms of CDK1. In order to determine this, an immunoprecipitation with anti-SUMO antibodies was followed by a western blot with either anti-total or -an inhibitory (Tyr 15) phosphorylated CDK1 isoform antibodies. The results show that SUMO interacts with both total and inactive isoforms of CDK1, but the phosphorylated/inactive isoform is more enriched in the SUMO pool-down fraction as compared to the amount of the protein found in the whole cell lysate (Fig. 3A).
We next wanted to test whether the activation of CDK1 activity is directly caused by the lower level of covalent or non-covalent interaction of CDK1 with SUMO, or if the effect is a result of the inhibition of sumoylation of other proteins interacting with CDK1, which may then regulate its activity. For that, we performed an immunoprecipitation with CDK1 antibody using the control lysate and the lysate produced after an inhibition of sumoylation, followed by a western blot with anti-SUMO antibody. As can be seen in Fig. 3B, the level of CDK1 sumoylaion does not change upon inhibition of global sumoylation (arrow on the SUMO blot), but the level of sumoylation of other high molecular weight (HMW) proteins interacting (and co-immunoprecipitated) with CDK1 significantly decreased. As expected, the decrease in the level of inhibitory phosphorylation can be noticed upon the inhibition of sumoylation. These findings suggest that sumoylation of specific proteins interacting with CDK1 inhibits the kinase’s activity, and that the inhibitory effect is being reversed upon inhibition of sumoylation. Identification of the CDK1-interacting proteins and the effect of sumoylation on their activity and the activity of CDK1 and its down-stream targets should be the focus of further studies. Although this study focused on the effect of sumoylation on CDK1 activity, it has been shown that CDK1, in turn, can phosphorylate SUMO-conjugating enzyme UBC9, thus enhancing its sumoylation activity ; Together, these data suggest a complex interplay between the two post-translational modifications involved in cell cycle regulation that should be further unraveled.
This study was supported by the NIH, NICHD, and Academic Research Enhancement Award 1R15HD067944-01A1 (MV, PI). Undergraduate student research was supported by Selma and Jacques H. Mitrani Foundation.