Phosphorylation of Ser28 in Histone H3 Mediated by Mixed Lineage Kinase-like Mitogen-activated Protein Triple Kinase α*

The mitogen-activated protein kinase cascades elicit modification of chromatin proteins such as histone H3 by phosphorylation concomitant with gene activation. Here, we demonstrate for the first time that the mixed lineage kinase-like mitogen-activated protein triple kinase (MLTK)-α phosphorylates histone H3 at Ser28. MLTK-α but neither a kinase-negative mutant of MLTK-α nor MLTK-β interacted with and phosphorylated histone H3 in vivo and in vitro. When overexpressed in 293T or JB6 Cl41 cells, MLTK-α phosphorylated histone H3 at Ser28 but not at Ser10. The interaction between MLTK-α and histone H3 was enhanced by stimulation with ultraviolet B light (UVB) or epidermal growth factor (EGF), which resulted in the accumulation of MLTK-α in the nucleus. UVB- or EGF-induced phosphorylation of histone H3 at Ser28 was not affected by PD 98059, a MEK inhibitor, or SB 202190, a p38 kinase inhibitor, in MLTK-α-overexpressing JB6 Cl41 cells. Significantly, UVB- or EGF-induced phosphorylation of histone H3 at Ser28 was blocked by small interfering RNA of MLTK-α. The inhibition of histone H3 phosphorylation at Ser28 in the MLTK-α knock-down JB6 Cl41 cells was not due to a defect in mitogen- and stress-activated protein kinase 1 or 90-kDa ribosomal S6 kinase (p90RSK) activity. In summary, these results illustrate that MLTK-α plays a key role in the UVB- and EGF-induced phosphorylation of histone H3 at Ser28, suggesting that MLTK-α might be a new histone H3 kinase at the level of mitogen-activated protein kinase kinase kinases.

The status of a eukaryotic cell is determined by endogenous and exogenous signals, and the end point of the pathways that transduce these signals is DNA. DNA is organized into a nucleoprotein complex termed chromatin, which not only facilitates packaging of DNA within the nucleus but also serves as an important factor in the regulation of genome function (1). The basic unit of chromatin is the nucleosome, which consists of 146 base pairs of DNA wrapped around an octamer of core histone proteins, including H2A, H2B, H3, and H4. The identification of enzymes that post-translationally modify histones and the discovery of ATP-dependent chromatin remodeling complexes were milestones in the elucidation of the role of chromatin in genome function and regulation (2,3).
Cellular signaling pathways influence histones by a number of methods, including phosphorylation of the histone tails, thereby controlling gene expression (4,5). In Tetrahymena, phosphorylation of histone H3 at Ser 10 and Ser 28 is very important for chromosome condensation and segregation during mitosis (6). The phosphorylation of histone H3 at Ser 10 appears to be involved in the initiation of chromosome condensation but not the maintenance of it (7). Although phosphorylation of histone H3 at Ser 10 is related to chromosome assembly, congregation at the metaphase plate, and segregation (8 -10), phosphorylation of histone H3 at Ser 28 occurs in chromosomes predominantly during early mitosis and coincides with the initiation of mitotic chromosome condensation in various mammalian cell lines (11). However, the molecular mechanism is not understood clearly yet. Growing knowledge has demonstrated that phosphorylation of histone H3 at Ser 10 and Ser 28 is involved in different signal transduction pathways and is dependent on the specific stimulation or stress. For example, histone H3 at Ser 10 is phosphorylated by RSK2 when cells are stimulated with epidermal growth factor (EGF) 1 (12), and by mitogen-and stress-activated protein kinase 1 (MSK1) when cells are stimulated with EGF or 12-O-teteradecanoylphorbol-13-acetate (TPA) (13). In addition, our previous studies demonstrated that ultraviolet B (UVB) irradiation markedly induces phosphorylation of histone H3 at Ser 10 and Ser 28 , and the phosphorylation was inhibited by a MEK1 inhibitor or a p38 kinase inhibitor as well as dominant negative mutant ERK or dominant negative mutant p38 kinase in the cells (14,15). These results clearly demonstrated that MAPKs play a key role to phosphorylate histone H3 at Ser 10 and Ser 28 following stimulation by tumor promoters such as EGF, TPA or UVB.
MAPK kinase kinases (MAPKKKs) are a family of serine/ threonine protein kinases that can activate one or several of the MAPK kinase/MAPK cascades (16). MAPKKKs phosphorylate two serine/threonine residues in the activation/phosphorylation sites of MAPK kinases (17). The well known MAP-KKKs include the Raf, the MEK kinase (MEKK) and the mixed lineage kinase (MLK) family. The Raf family kinases selectively activate the ERK pathway (18). On the other hand, MEKK1, MEKK2, and MEKK3 activate both the ERK pathway and the JNK/SAPK pathway (19). MLK3, MUK, TAK1, and ASK1 activate the JNK/SAPK and p38 kinase pathways (16).
Some of the MAPKKKs are activated by distinct extracellular stimuli. For example, Raf-1 is activated by mitogens and phorbol esters, TAK1 is activated by transforming growth factor-␤ and cytokines (17,20), and ASK1 is activated by tumor necrosis factor-␣ (21). Very recently, we demonstrated that MLTK-␣, a member of the MLK family, induced neoplastic cell transformation and tumorigenesis in athymic nude mice (22). In addition, MLTK-␣ translocates into the nucleus after stimulation of EGF or TPA and also induces phosphorylation of c-Myc, Elk-1, c-Jun, and ATF2 (22). However, even though the role of histone H3 is well characterized in chromatin structure and function, very little is known about the protein kinases that are responsible for the phosphorylation of histones.
Here, we demonstrated that MLTK-␣ is a new kinase that can phosphorylate histone H3 at Ser 28 , but not Ser 10 , in vitro and in vivo after stimulation by EGF or UVB. This phosphorylation of histone H3 by MLTK-␣ may play an important role in chromatin structure remodeling mediating MLTK-␣-induced cell transformation or tumorigenesis.
Construction of Mammalian Expression Vectors-For the mammalian two-hybrid (M2H) system, the cDNAs of 60 human kinases were amplified by PCR and introduced into pBIND two-hybrid system vectors. The cDNA encoding histone H3.3 (a generous gift from Aichi Cancer Center Research Institute, Nagoya, Japan) was recombined into the BamHI/KpnI site of the pACT vector. To produce stably transfected JB6 Cl41 cells, a cDNA fragment from MLTK-␣ was introduced into the BamHI/KpnI site of pcDNA4-HisMAXA (pHisG-MLTK-␣) (Invitrogen). The point mutation of histone H3 at Ser 10 (S10A) or Ser 28 (S28A) was generated by using the QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA) and introduced into the pcDNA3.1/V5-His vector (pV5-H3-S10A or -S28A), respectively. The cDNAs of MLTK-␣, MLTK-␤, MLTK-␣-KM, and MLTK-␤-KM cloned into pSR␣-HA were kindly provided by Dr. Eisuke Nishida (Kyoto University, Kyoto, Japan). All of the positive clones containing cDNA inserts were confirmed by restriction enzyme mapping and DNA sequence analysis (Genewiz Inc., North Brunswick, NJ).
Transfection-The pHisG-MLTK-␣ stably expressed cells were established using 293T and JB6 Cl41 cells selected by 200 g/ml of zeocin and pooled. To analyze the interaction of proteins by M2H, 3T3 cells (1.5 ϫ 10 4 ) were seeded into 48-well plates and incubated with 10% FBS-DMEM for 18 h before transfection. Transient transfection was carried out using 293T cells to analyze the in vitro kinase activity. The cells were harvested at 48 h of incubation after transfection.
Mammalian Two-hybrid Assay-The DNAs, pACT-phistone H3.3, pBIND kinases, and pG5-luciferease, were combined in the same molar ratio, and the total amount of DNA was not more than 100 ng/well. The transfection was performed with Lipofectamine Plus reagent and following the manufacturer's recommended protocols. The cells were disrupted by addition of 200 l of cell lysis buffer directly into each well of the 48-well plate, and then aliquots of 100 l were added to each well of a 96-well luminescence plate. The luminescence activity was measured automatically by a computer program from MTX Lab, Inc. (Vienna, VA). The relative luciferase activity was calculated and normalized based on the pG5-luciferase basal control. For assessment of transfection effi-ciency and protein amount, the luciferase assay, Renilla luciferase activity assay, or Lowry protein assay was used.
UVB Irradiation and Chemical Treatment-The JB6 Cl41 overexpressing cells were seeded in 60-mm dishes and cultured in 5% FBS/ MEM until they reached 70% confluence and then starved in serumdeprived MEM for 36 h. The cells were then incubated for 2 h in fresh serum-deprived MEM, after which time they were exposed to 4 kJ/m 2 UVB and then cultured for an additional 30 min. Because the normal UVB lamp also generates a small amount of UVC light, the UVB irradiation was carried out in a UVB exposure chamber fitted with a Kodak Kodacel K6808 filter that eliminates all wavelengths below 290 nm. For the M2H assay, pBIND kinases and pACT-H3 plasmids were transiently co-transfected in Swiss 3T3 cells. After starvation in serumdeprived DMEM for 24 h, the cells were exposed to 4 kJ/m 2 UVB and then cultured for 12 h more. The cells were incubated with 10 ng/ml EGF in culture medium at 37°C for 30 min or 12 h for the in vivo kinase assay or for the M2H assay, respectively.
Immunoblotting-The proteins were resolved by SDS-PAGE and transferred onto polyvinylidene difluoride membranes. The membranes were blocked and then hybridized with the appropriate primary antibody overnight at 4°C. The protein bands were visualized by a chemiluminescence detection kit (ECL; Amersham Biosciences) after hybridization with the horseradish peroxidase-conjugated secondary antibody from rabbit or mouse.
Immunofluorescence Assay-For translocation of exogenous MLTK-␣, the cells were fixed in 4% paraformaldehyde, and pHisG-MLTK-␣ was detected with a monoclonal anti-HisG antibody and a Texas Red-conjugated secondary antibody. Nuclei were stained with 4,6-diamidino-2-phenylindole. The cells were incubated for 24 h, were starved in serum-free medium for an additional 24 h, and then were or were not irradiated with 4kJ/m 2 UVB or 10 ng/ml EGF. The samples were analyzed using a fluorescence microscope system (Leica) and Image-Pro software program (Silver Spring).

Histone H3 Interacts with MLTK-␣-Recently, interest has
increased in identifying the kinases that modulate chromatin through their mediation of the phosphorylation of histone H3. To identify kinases that interact with histone H3, we used a M2H assay system. In this system, histone H3 was cloned into the pACT expression vector, and individual candidate human protein kinases were cloned into the pBIND expression vector in combination with the pG5 luciferase reporter gene. An interaction between histone H3 and kinases brings together the Gal4-binding domain (from pBIND) and the VP16 transactivation domain (from pACT) of the fusion proteins and activates the luciferase reporter gene in Swiss 3T3 cells. We used the known interaction between histone H3 and DYRK3 as a positive control (pACT-H3/pBIND-DYRK3). DYRK3 is a dual specificity protein kinase that catalyzes phosphorylation on serine/ threonine and tyrosine residues and can phosphorylate histone H3 and H2B on serine and threonine residues (24). From the results of this assay in which several interesting interactions were found, we confirmed that MLTK-␣ interacted with histone H3 (Fig. 1A). Fig. 1A shows that more than a 40-fold increase in luciferase activity was observed in MLTK-␣ and histone H3 co-transfected cells (pACT-H3/pBIND-MLTK-␣) compared with only pACT-H3 transfected cells.
We next examined the in vitro interaction between MLTK-␣ and purified histone H3. First, the cDNA coding sequence of MLTK-␣ or DYRK3 was subcloned into the pcDNA4His/MaxA vector to generate pHisG-MLTK-␣ or -DYRK3, and the fusion protein was in vitro translated using the TNT Quick-coupled transcription/translation system (Promega). Histone H3 immunoprecipitated with protein A-Sepharose 4B beads was incubated with [ 35 S]methionine-labeled MLTK-␣ or DYRK3. The bound protein eluted from the beads was separated by SDS-PAGE and then detected by autoradiography. Fig. 1B shows that MLTK-␣ or DYRK3 efficiently interacted with histone H3 in the in vitro binding assays. Input represents 5% of material used for the in vitro binding assays.
MLTK-␣, but Not MLTK-␤, Phosphorylates Histone H3 (Ser 28 ) in Vitro-MLTK-␣ and MLTK-␤ were identified by 5Јand 3Ј-RACE analyses, which showed at least two mRNA variants with different 3Ј-sequences (25). cDNA cloning of the two variants and open reading frame prediction revealed that the two proteins have identical amino acid sequences in the NH 2 -terminal region (residues 1-311), which contains a kinase domain and a leucine zipper motif. The COOH-terminal region of MLTK-␣ has a sterile ␣-motif domain that has been shown to be involved in protein-protein interactions and dimer formation of several signaling molecules and transcription regulators (26).
We expressed HA-tagged MLTK-␣ (pHA-MLTK-␣) and -MLTK-␤ (pHA-MLTK-␤), or the kinase negative mutants of MLTKs, made by replacing Lys 45 with Met (pHA-MLTK-␣-KM and -MLTK-␤-KM) in Swiss 3T3 cells. We then determined whether either or both of the MLTKs phosphorylated histone H3 in vitro and whether the kinase-negative mutants of the MLTKs blocked the phosphorylation of histone H3 in vitro. After immunoprecipitation with an anti-HA antibody, we analyzed MLTKs kinase activity for the phosphorylation of histone H3 in vitro. Very little phosphorylation of histone H3 was detected in MLTK-␣-KM, MLTK-␤, or MLTK␤-KM in vitro (Fig. 3, first panel), although a clear phosphorylation of histone FIG. 1. Histone H3 is identified as an MLTK-␣ substrate. A, direct interaction of MLTK-␣ with histone H3 in a mammalian twohybrid assay. For a negative control, the pACT or pBIND plasmids were transfected with the reporter plasmid, pG5-luc, into Swiss 3T3 cells (18,000 cells/ml). For the positive control, histone H3 and DYRK3 were transfected with the pG5-luc reporter plasmid into the cells. Histone H3 and/or pBIND-MLTK-␣ were co-transfected with the pG5-luc plasmid to confirm the interaction of MLTK-␣ with histone H3. After a 36-h incubation, the firefly luciferase activity was determined in cell lysates and normalized against Renilla luciferase activity. All of the experiments were performed at least twice with triplicate samples and are depicted as the means Ϯ S.E. The data were recorded as a relative fold luciferase activity using a Luminoskan Ascent plate reader (Thermo Electron Corp., Helsinki, Finland). *, p Ͻ 0.005. B, MLTK-␣ interacts with histone H3 in an in vitro binding assay. The cDNA of MLTK-␣ or DYRK3 was translated in vitro, and then 35 S-MLTK-␣ or -DYRK3 proteins bound with histone H3 immunoprecipitated beads were visualized by autoradiography. Input represents 5% of material used for the in vitro binding assays. IP, immunoprecipitation.
MLTK-␣ Plays an Important Role in the Phosphorylation of Histone H3 at Ser 28 in Vivo-To study whether MLTK-␣ regulates the phosphorylation of histone H3 at Ser 28 in vivo, we examined changes in endogenous histone H3 phosphorylation in 36-h serum-starved 293T cells, which stably express the pHisG-MLTK-␣. The results demonstrated that overexpression of MLTK-␣ in 293T cells was accompanied by increased phosphorylation of histone H3 at Ser 28 with serum deprivation (Fig. 4A).
The Accumulation of HisG-MLTK-␣ in the Nucleus Was Induced by UVB or EGF Stimulation-MLTK-␣ contains two putative nuclear export signal-like sequences in the NH 2 and COOH termini. MLTK-␣ is mainly localized in the cytosol and is known to accumulate in the nucleus when cells are treated with leptomycin, an inhibitor of the nuclear export signal receptor, EGF, or TPA (22,25). To examine whether MLTK-␣ accumulates in the nucleus after stimulation of cells with UVB or EGF, JB6 Cl41 cells stably transfected with the mock vector or pHisG-MLTK-␣ were treated with UVB or EGF. HisG-MLTK-␣ was detected with an anti-HisG antibody and an horseradish peroxidase-conjugated secondary antibody (Fig. 5A) or Texas-Red-conjugated secondary antibody (Fig. 5B). The results showed that HisG-MLTK-␣ accumulated in the nucleus 30 min after stimulation with 4 kJ/m 2 UVB (Fig. 5, A and B, left panels) or 10 ng/ml EGF (Fig. 5, A and B, right panels). These results suggested that accumulation of MLTK-␣ in the nucleus might be important in the phosphorylation of histone H3 at Ser 28 .
The Interaction Between Histone H3 and MLTK-␣ Is Enhanced after UVB or EGF Stimulation-Despite our previous data showing direct evidence that MLTK-␣ can interact with and phosphorylate histone H3 Ser 28 in vitro and in vivo, we could not exclude the possibility of another pathway leading to phosphorylation of histone H3 at Ser 28 by MLTK-␣, which is mediated through ERKs, p38, p90 RSK , and/or MSK1. To examine whether the MLTK-␣-mediated histone H3 phosphorylation occurs through well established signal transduction pathways, we first assessed whether UVB or EGF stimulation would affect the interaction between MLTK-␣ and histone H3 using the M2H assay. Swiss 3T3 cells were seeded in 48-well plates in 10% FBS/DMEM and cultured until they reached 70% confluence. The cells were then transfected with pACT-H3 and pBIND-MLTK-␣ or -DYRK3. The results confirmed that after treatment with 4 kJ/m 2 UVB (Fig. 6A) or 10 ng/ml EGF (Fig.   6B), the interaction of histone H3 with MLTK-␣ was greater than in unstimulated cells. These results suggest that the interaction between MLTK-␣ and histone H3 was enhanced by UVB or EGF and mediates the UVB-or EGF-induced phosphorylation of histone H3 at Ser 28 in vivo.

PD 98059 and/or SB 202190 Have No Effect on UVB-or EGF-induced Phosphorylation of Histone H3 at Ser 28 in MLTK-
␣-overexpressing JB6 Cl41 Cells-We hypothesized that the signal through MLTK-␣ could activate a well established MAPK pathway to phosphorylate histone H3 following stimulation by UVB or EGF. Therefore, we determined whether PD 98059 and/or SB 202190 affected UVB-or EGF-induced phosphorylation of histone H3 at Ser 28
is a specific inhibitor of p38 kinase, and pretreatment of cells with 4 M SB 202190 almost totally blocked UVB-induced phosphorylation of histone H3 at Ser 28 (31). EGF activates ERKs more strongly than JNK/SAPK, whereas UV radiation activates JNK/SAPKs much more strongly than ERKs (32).
Our results also showed that 4 kJ/m 2 UVB-and 10 ng/ml EGF-induced phosphorylation of histone H3 at Ser 28 in mock vector transfected JB6 Cl41 cells was blocked by inhibitors (25 M PD 98059 and/or 1 M SB 202190) of the MAPK pathways (Fig. 7A). Surprisingly, these inhibitors had no effect on histone H3 phosphorylation in MLTK-␣-overexpressing JB6 Cl41 cells (Fig. 7B, first panels). However, the inhibitors effectively suppressed ERKs and p38 kinase phosphorylation in MLTK-␣overexpressing cells (Fig. 7B, third and fifth panels), indicating that MLTK-␣-mediated phosphorylation of histone H3 at Ser 28 is independent of ERK and p38 MAPK pathways.
We also used the siRNA method to knock down the endogeneous MLTK-␣ expression level and then determined the effects on UVB-or EGF-induced phosphorylation of histone H3 at Ser 28 . First, we confirmed the efficiency of siRNA for knocking down the endogeneous MLTK-␣ mRNA level using reverse transcription PCR. The psi-MLTK-␣ construct was co-trans-

FIG. 5. Stimulation by EGF or UVB resulted in a nuclear localization of the HisG-MLTK-␣ fusion protein.
A, nuclear localization induced by treatment with 4 kJ/m 2 UVB (left panels) or 10 ng/ml EGF (right panels) was assessed in JB6 Cl41 cells stably transfected with mock vector or pHisG-MLTK-␣. The cells (2ϫ10 5 /ml) were seeded in 60-mm plates and cultured 24 h in 5% FBS/MEM in a 37°C, 5% CO 2 incubator. The cells were starved in serum-deprived MEM for 36 h, treated with 10 ng/ml EGF (30 min) or 4 kJ/m 2 UVB, and harvested after incubation for an additional 30 min. The expression of HisG-MLTK-␣ was assessed by immunoblotting (IB) using 25 g of cytoplasmic or nuclear proteins and an anti-HisG antibody. For visualizing equal loading of protein, ␣-tubulin, or ␤-actin was immunoblotted and detected with specific antibodies. B, accumulation of MLTK-␣ induced by UVB or EGF was visualized using immunofluorescence assay as described under "Experimental Procedures." DAPI, 4Ј,6Ј-diamino-2-phenylindole.
FIG. 6. UVB or EGF stimulation causes an enhanced association between histone H3 and MLTK-␣. Swiss 3T3 cells were cotransfected with pACT-H3, pBIND kinases (MLTK-␣, DYRK3), and pG5-luc plasmid and then starved in serum-deprived DMEM for 24 h. The cells were then incubated for 2 h in fresh serum-deprived DMEM. The cells were then either exposed or not exposed (W/O) to 4 kJ/m 2 UVB (A) or 10 ng/ml EGF (B) and incubated for an additional 12 h at 37°C. After incubation, the luciferase activity was determined in cell lysates and normalized against Renilla luciferase activity. *, p Ͻ 0.005. All of the experiments were performed at least twice with triplicate samples and are depicted as the means Ϯ S.E. fected into wild type JB6 Cl41 cells with pcDNA3.1neo and selected with G418. Total RNA was isolated from pmU6 or psi-MLTK-␣ stably transfected cells, and mRNA levels were compared. The magnitude of gene knock-down was analyzed by reverse transcription PCR (Fig. 8C). MLTK-␣ mRNA was detected after 30 cycles in pmU6 stably transfected cells (Fig. 8C). In contrast, the internal control, ␤-actin, showed no difference in density, indicating that expression of psi-MLTK-␣ specifically knocks down the MLTK-␣ mRNA level (Fig. 8C). Further results indicated that UVB-or EGF-induced phosphorylation of histone H3 at Ser 28 was almost totally blocked in the psi-MLTK-␣ transfected cells but not in pmU6pro vector stably transfected cells (Fig. 8D, first panel). Total histone H3 levels were unchanged (Fig. 8D, second panel). We examined MSK1, p90 RSK kinase, ERK1/2, or p38 kinase activity in mock and endogeneous MLTK-␣ knock-down cells to evaluate whether the knock-down of MLTK-␣ could have an effect on the well known signaling pathway to phosphorylate histone H3 induced by UVB or EGF. The results showed that 4 kJ/m 2 UVB and 10 ng/ml EGF can induce the phosphorylation of MSK1 (S376), p90 RSK (T359/S360), ERK1/2, or p38 not only in the pmU6 transfected cells but also in the MLTK-␣ knock-down cells (psi-MLTK␣) compared with the control group (Fig. 8D, third through sixth panels, respectively). Finally, our results indicate that MLTK-␣ plays a key role in mediating phosphorylation of histone H3 at Ser 28 independent of MSK1, p90 RSK , ERK1/2, or p38 kinase in vivo.

DISCUSSION
This study elucidates a signal transduction pathway involved in UVB-or EGF-induced phosphorylation of histone H3 at serine 28 by focusing particularly on a new role for MLTK-␣ by which the phosphorylation of histone H3 is mediated. His-tones are now clearly known as integral and dynamic components of the machinery responsible for regulating gene transcription (34 -37). The same is probably true for other DNArelated processes such as replication, repair, recombination, and chromosome segregation (6,9). Many types of cancer are associated with translocations or mutations in chromatin-modifying enzymes and regulatory proteins (38 -42). For example, mutations in the histone H3-Ser 10 kinase, RSK2, are associated with Coffin-Lowry syndrome, and increased aurora B histone kinase activity has been associated with colorectal cancer (12,43,44). However, integrated information regarding kinases that modulate chromatin by post-transcriptional modification is lacking. To identify new kinases that might be associated with chromatin modification, we screened interactions between histone H3 and 60 kinases using the mammalian twohybrid assay in vivo and verified the interactions in vitro. The results presented here show that MLTK-␣ interacted with histone H3, resulting in phosphorylation of histone H3 at serine 28.
The MLKs are a family of serine/threonine protein kinases that function in a phospho-relay module to control the activity of specific MAPKs, ERK1/2, p38 kinase, JNKs, and ERK5 (45). On the basis of domain arrangements and sequence similarity within their catalytic domains, the MLKs cluster into three subgroups, including the MLKs, the dual leucine zipper-bearing kinases, and the zipper sterile-␣-motif kinase (ZAK) (45). ZAKs ZAK-␣ and ZAK-␤ are alternative spicing products and are also referred to as MLTK-␣ and MLTK-␤ (25,45). In mammalian MLTK, the sterile ␣-motif kinase domain of the MLTK-␣ is in the middle of the protein sequence (46). The MLTK-␤ splice variant form is identical to MLTK-␣ from the amino terminus to the zipper domain but then diverges and terminates shortly thereafter, thus lacking a sterile ␣-motif domain (25). MLTK-␣ has two putative nuclear export signal domains and has been shown to accumulate in the nucleus when treated with 2 ng/ml leptomycin, an inhibitor of the nuclear export signal receptor (25). We observed that UVB or EGF can induce MLTK-␣ accumulation in the nucleus. These results suggested that UVB or EGF can influence the export of MLTK-␣ from the nucleus to the cytosol. NH 2 -terminal histone tails may be targets for ATP-dependent chromatin remodeling factors such as Swi/Snf and NURF, thus participating in the modulating chromatin architecture (47)(48)(49). Another way in which they can modulate chromatin may be through a diverse array of post-translational modifications, such as acetylation and phosphorylation. With a potential functional role of histone H3 phosphorylation in chromosome condensation, the identification of the kinase(s) responsible for this phosphorylation is of great importance. Ipl1/AIR-2 kinases in yeast and nematodes and the Never In Mitosis A (NIMA) kinase in Aspergillus nidulans have been identified as mitotic H3 kinases (50,51). Interestingly, recent data demonstrated that a member of the Survivin/BIR family has an indirect role in histone H3 phosphorylation in Caenorhabditis elegans and was originally identified as a family of proteins that have an inhibitory effect on apoptosis (52). Using RNA-mediated interference to suppress expression of BIR-1 in C. elegans, these mutant embryos were found to lack histone H3 phosphorylation and exhibit problems with chromosome condensation and segregation (53). Our results showed that an overexpressed MLTK-␣ phosphorylated histone H3 at Ser 28 , but not Ser 10 . Interestingly, MLTK-␤ and MLTK mutants (MLTK-␣KM) had very little kinase activity to phosphorylate histone H3 compared with MLTK-␣. This suggests that the difference in their kinase activity results mainly from the serine/threonine kinase domain (residues 16 -277) and the sterile ␣-motif (residues 337-408) only present in MLTK-␣. In addi-  A and B, right panels) and incubated for an additional 30 min. The media were removed, and the cells were washed twice with cold phosphate-buffered saline. Phosphorylation of histone H3 at Ser 28 , ERK1/2, or p38 kinase was determined by immunoblot (IB) analysis of cell lysates with an anti-phospho-H3 (Ser 28 ), anti-phospho-ERK1/2, or anti-phospho-p38 kinase antibody, respectively. Total histone H3, ERK1/2, or p38 kinase was used as an internal control to monitor equal protein loading. tion, the introduction of siRNA to suppress expression of MLTK-␣ almost totally blocked phosphorylation of histone H3 at Ser 28 in JB6 Cl41 cells.
A critical question was raised as to whether these events were mediated through the well known MAPK pathways. The MSK1 cascades are known to elicit modification of chromatin proteins such as histone H3 by phosphorylation and/or acetylation concomitant with gene activation (33). The phosphorylation of histone H3 that occurs during immediate-early gene induction does so on only a subset of nucleosomes (54). Recently, MSK2 and, to a lesser extent, MSK1, were reported as the major protein kinases required for the phosphorylation of histone H3 at both Ser 10 and Ser 28 after stimulation of primary embryonic fibroblasts by TPA or anisomycin (33). Also, EGF, TPA (13), or UVB irradiation (55) activates endogenous MSK1. A previous study showed that PD 98059, SB 202190, or H89 inhibited EGF-or UVB-induced phosphorylation of histone H3 at Ser 10 and Ser 28 , and the effects were mediated by MSK1 (14,15). In this report, UVB induces activation of ERKs, p38 kinase, JNKs, AKT, p70/85 S6K , and p90 RSK . However, H89, an inhibitor of MSK1, further enhances UVB-induced activation of these kinase but inhibits MSK1 activity resulting in inhibition of UVB-induced phosphorylation of histone H3 at serine 28 induced by UVB. Until now, only MSK1 was reported to play a key role as a histone H3 kinase to mediate signal transduction induced by UVB or EGF.
When overexpressed in cells, MLTK-␣ is capable of activating the four major MAPK pathways: ERK, JNK/SAPK, p38, and ERK5 (25). We reported previously that MLTK-␣ is an "onco-kinase" that can regulate activity of onco-transcription factors such as ATF2, c-Jun, Elk-1, and c-Myc through MAPK signal cascades and thereby plays a key role in neoplastic cell transformation and cancer development (22). In our results, we further elucidated a new role for MLTK-␣ in mediating the Mock, MLTK-␣-overexpressing, and MLTK-␣ knock-down JB6 Cl41 cells were starved for 36 h by incubating in serum-deprived MEM at 37°C in a 5% CO 2 atmosphere. The cells were then incubated for 2 h in fresh serum-deprived MEM and exposed to 4 kJ/m 2 UVB (A) or treated with 10 ng/ml EGF (B) and incubated for an additional 30 min. After removal of media and harvesting, phosphorylation of histone H3 at Ser 28 , MSK1 at Ser 376 , or p90 RSK at Thr 359 -Ser 360 (T359/S360) was determined by immunoblot (IB) analysis of cell lysates with an anti-phospho-H3 (Ser 28 ), anti-phospho-MSK1 (Ser 376 ), or anti-phospho-p90 RSK (T359/S360), respectively. Total histone H3 or ␤-actin was used as an internal control to monitor equal protein loading. C, endogeneous MLTK-␣ mRNA expression was suppressed by siRNA directed against MLTK-␣. pmU6pro control or psi-MLTK-␣ stably transfected cells were co-transfected with pcDNA3.1neo and selected with geneticin (800 g/ml) for 10 days and then pooled. Total RNA was isolated and reverse transcribed using an oligo(dT) primer. Panel C shows the amplification of MLTK-␣ mRNA by reverse transcription-PCR and visualization by ethidium bromide agarose gel electrophoresis. ␤-Actin was used as an internal control. D, pmU6pro or psi-MLTK-␣ stably transfected cells were starved for 36 h by incubating in serum-deprived MEM at 37°C in a 5% CO 2 atmosphere. The cells were then incubated for 2 h in fresh serum-deprived MEM and exposed to 4 kJ/m 2 UVB or treated with 10 ng/ml EGF. Phosphorylation of histone H3 at Ser 28 , MSK1, p90RSK, ERK1/2, or p38 kinase was determined by immunoblot analysis with a specific antibody, respectively. E, MAPKs, including ERK1/2, p38 MAPKs, or MSK1 are involved in UVB-or EGF-induced phosphorylation of histone H3 at Ser 28 . Herein, we provide evidence showing that MLTK-␣, which is located upstream of the MAPKs cascade, plays a key role in UVB-or EGF-induced phosphorylation of histone H3 at Ser 28 independent of ERK1/2, p38 MAPK, or MSK1. phosphorylation of histone H3 at Ser 28 resulting from MLTK-␣ activation following stimulation with UVB or EGF. Evidence that MLTK-␣ functions in the pathway leading to histone H3 phosphorylation comes in part from overexpression experiments or gene knock-down experiments with si-MLTK-␣. Specifically, the overexpression of MLTK-␣ in JB6 Cl41 cells enhanced phosphorylation of histone H3 at Ser 28 induced by UVB or EGF and, moreover, overcame the general inhibition of the MAPK pathway by the MAPK inhibitors PD 98059 and SB 202190. Other potential candidates for mediating phosphorylation of histone H3 include p90 RSK and MSK1, which are downstream of Akt or ERK, or p38, respectively. Our results indicated that although UVB or EGF induced the activation of p90 RSK and MSK1 not only in mock and MLTK-␣-overexpressing cells but also in MLTK-␣ knock-down cells, UVB-or EGFinduced phosphorylation of histone H3 (Ser 28 ) is blocked in the MLTK-␣ knock-down cells compared with control groups. These data suggest that MLTK-␣ may act on MSK1-independent pathways or as a histone H3 kinase to promote the UVB-or EGF-induced phosphorylation of histone H3 at serine 28.
Finally, we showed that MLTK-␣, one of the member MAP-KKKs, is indeed not only a direct histone H3 kinase that catalyzes the phosphorylation of histone H3 at serine 28 but also up-regulates ERK1/2, p38 kinase and thus causes the activation of MSK1 (Fig. 8E). However, these results raise the intriguing question as to how MSK1 and MLTK-␣ are both essential for the phosphorylation of histone H3 at serine 28. Previously, Soloaga et al. (33) as well as our group (15) showed that MSK1 mediates phosphorylation of histone H3 at serine 28 induced by TPA, EGF, or UVB. In Fig. 8 (A and B), we showed that overexpressing MLTK-␣ in JB6 cells enhances the EGF-or UVB-induced phosphorylation of MSK1, whereas si-MLTK-␣ down-regulated the level of phosphorylation MSK1 as well as phosphorylation of histone H3 at serine 28. On the other hand, knock-down of MLTK-␣ in wild type JB6 Cl41 cells significantly blocked EGF-or UVB-induced histone H3 phosphorylation at serine 28, but the level of phosphorylation of MSK1 (S376) was still at a relatively high level. This suggests not only that MLTK-␣ is an upstream kinase of ERK1/2 and p38 kinase and thereby can mediate the activity of MSK1 but also that MLTK-␣ can act in a pathway independent of the MAPK cascades. The phosphorylation of histone H3 occurs during mitosis and at late G 2 (7). Moreover, Soloaga et al. (33) reported that a small proportion of both wild type and MSK1/ MSK2 knock-out cells did not arrest in G 0 /G 1 after prolonged serum starvation but were still in G 2 as determined by fluorescence-activated cell sorter analysis. Most likely, the phosphorylation of histone H3 might be not regulated by only one kinase, because histone H3 is also considered to be one of many "housekeeping proteins" involved in maintenance of the cell. Therefore, we suggest that MLTK-␣ as well as MSK1 may play an important role in the phosphorylation of histone H3 at serine 28 induced by either EGF or UVB.
Because phosphorylation of histone H3 plays a critical role in both the chromosome and the nucleus assembly in mitosis, the identification of kinases responsible for the regulation of histone phosphorylation is of primary importance. In our study, we provided evidence showing that MLTK-␣ is new kinase that can phosphorylate histone H3 at Ser 28 in vitro and in vivo after stimulation with UVB or EGF, suggesting that this phosphorylation of histone H3 by MLTK-␣ may play an important role in chromatin structure remodeling.