PTCL specimens were collected in Fukushima Medical University and The Cancer Institute. A human embryonic kidney 293T (HEK293) cell line was obtained from American Type Culture Collection (ATCC: Manassas, VA, USA), and maintained in Dulbecco’s modified Eagle’s medium-F12 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine (both from Invitrogen). An IL-2-dependent mouse cytotoxic T-cell line, CTLL-2, and a human T-cell line, Jurkat, were both obtained from ATCC, and maintained in RPMI1640 medium (Invitrogen) supplemented with 10% FBS. Mouse recombinant IL-2 (Peprotech, Rocky Hill, NJ, USA) was added to the culture medium of CTLL-2 at the concentration of 2 ng/ml. Total RNA was extracted from cell lines and a PTCL specimen with an RNeasy mini kit (Qiagen, Valencia, CA, USA), and was subjected to reverse transcriptase (RT) with an oligo-dT primer. Written informed consent was obtained from the subjects who provided cancer specimens, and the study was approved by the human ethics committee of the University of Tokyo, Jichi Medical University, Fukushima Medical University, and The Cancer Institute of the Japanese Foundation for Cancer Research.

Resequencing with a custom cDNA-capture system was performed as previously described (8). In brief, RNA probes of 120 bases were designed to interrogate cDNAs of 906 human protein-coding genes, and were synthesized by Agilent Technologies (Santa Clara, CA, USA). Hybridization of cDNA fragments isolated from a PTCL specimen to the RNA probes was performed according to the protocols recommended for the SureSelect Target Enrichment system (Agilent Technologies). Purified cDNA fragments were then subjected to deep sequencing with a Genome Analyzer IIx (GAIIx; Illumina, San Diego, CA, USA) for 76 bases from both ends by the paired-end sequencing system.

The full-length cDNA for the wild-type or the R634H mutant of STK10 was amplified by PCR from cDNA of a PTCL specimen. The cDNAs for other mutant forms of STK10 were further generated by a polymerase chain reaction (PCR)-based mutagenesis approach. The cDNA for PLK1, IKK-α, IKK-β or IKK-γ was amplified by PCR from our cancer cell lines.

HEK293 cells were transfected with the expression plasmids, a promoter-less Renilla luciferase plasmid pGL-4.70 (Promega, Madison, WI, USA), and firefly luciferase-based reporter plasmid for Fos (pFL700) (9), Myc (pHXL) (10), NF-κB (Stratagene, La Jolla, CA, USA), Bcl-xL (11), Notch (pGa981-6) (12), Wnt (TOP-flash, Upstate Biotechnology, Lake Placid, NY, USA) or Rho (pSRE.L) (13) signaling pathway. Luciferase activities were determined with the Dual-Luciferase Reporter Assay System (Promega), and the firefly luciferase activities were normalized to the Renilla luciferase activities.

Quantitative real-time RT-PCR was performed with QuantiTect SYBR Green PCR kit (Qiagen) and an Applied Biosystems 7900HT Fast Real-Time PCR System. The PCR conditions include the first incubation at 50°C for 2 min and then 95°C for 15 min, followed by 60 cycles of 94°C for 15 sec and 60°C for 30 sec and 72°C for 1 min. Relative expression levels of target mRNAs to GAPDH were normalized to those in the mock-transfected cells. The primer sequences used for the PCR reactions were: 5′-TTCATTCCTGGCAAGTGGATCATT-3′ and 5′-ATGGCAGCATCATTGTTCTCATCA-3′ for TLR2, 5′-TGACAACCTTCTGGTTGGTAGGGA-3′ and 5′-CCAAGGTCATGGT TGTCCAAAGAC-3′ for BCL2, 5′-AAGAATCACCAGCA GCAAGTGTCC-3′ and 5′-TTGGGTTGTGGAGTGAGTGTTCAA-3′ for CCL2, and 5′-CCAGGTGGTCTCCTCTGACTTCAA-3′ and 5′-CACCCTGTTGCTGTAGCCAAATTC-3′ for GAPDH.

Antibodies used in this study were: anti-FLAG M2 (Sigma-Aldrich, St. Louis, MO, USA), anti-PLK1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-phospho-PLK1 (BD Pharmingen, San Jose, CA, USA), anti-ACTB (Cell Signaling Technology, Danvers, MA, USA), anti-mouse IgG and anti-rabbit IgG (both from GE Healthcare, Piscataway, NJ, USA). HEK293 cells were transfected with the appropriate expression vectors, and then lysed in lysis buffer [1% NP-40, 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM NaF, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride and aprotinin]. Target proteins were purified by incubation of the cell lysates with appropriate antibodies and Protein G Sepharose Fast Flow (Sigma-Aldrich) for 3 h at 4°C. Immunoblot analyses were visualized with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific, Waltham, MA, USA).

Immunoprecipitates were washed with kinase buffer [50 mM NaCl, 50 mM Tris-HCl (pH 7.4), 10 mM MgCl₂, 10 mM MnCl₂ and 0.1 mM Na₃VO₄], and then incubated in kinase buffer containing [γ-³²P]ATP (Perkin-Elmer, Boston, MA, USA) and histone H2A protein (New England BioLabs, Ipswich, MA, USA) for 30 min at room temperature.

CTLL-2 cells were infected with retrovirus expressing STK10 and the blasticidin-resistant gene, and cultured under the presence of 10 μg/ml blasticidin (InvivoGen, San Diego, CA, USA). Blasticidin-resistant cells were then treated with 1 μM dexamethasone (Sigma-Aldrich), and cell number was determined every 24 h with CellTiter-Glo Luminescent Cell Viability Assay (Promega). At 72 h after treatment with dexamethasone, CTLL-2 cells were collected and subjected to apoptosis assay with flow cytometry (FACSCanto II; BD Biosciences, San Jose, CA, USA) using Annexin V/propidium iodide (PI) staining kit (eBioscience, San Diego, CA, USA).

To identify transforming genes in PTCL, cDNAs for cancer-related genes (n=906) (8) were isolated from a PTCL specimen obtained from a 25 year-old male negative for EBV and HTLV-I infection, and subjected to deep sequencing with GAIIx. We thus obtained a total of 69,552,278 independent, high-quality reads that mapped to 868 cDNAs (95.8%). Screening of missense mutations, insertions, deletions and fusion genes through our in-house computational pipeline (8) revealed a total of 19 nonsynonymous mutations but no indels or gene fusions, which were further confirmed by Sanger sequencing (Table I). It should be noted, however, that 9 of the 19 nonsynonymous mutations were recently registered as single nucleotide polymorphisms (SNPs) in the 1000 genome project (http://www.1000genomes.org) after our initial analysis.

One of the 19 nonsynonymous mutations was a nucleotide substitution of G-to-A at position 2201 of human STK10 cDNA (GenBank accession number, NM_005990) that was later deposited as an SNP (rs115974403). Sequencing of this position in genomic DNA of the same specimen further confirmed this substitution, which results in replacement of an arginine residue at amino acid position 634 with a histidine residue (R634H) (Fig. 1A). We also searched STK10 mutations among our human cancer specimens (n=76), cancer cell lines (n=56) and the COSMIC database for cancer genome mutations (Release v58, http://cancer.sanger.ac.uk/cancergenome/projects/cosmic/), revealing additional 11 nonsynonymous mutations; E68K in a CML cell line K562, L85P, T256I, R411-frameshift (R411fs), K670N and K898N in various specimens of ovarian cancer, K277E in a testicular cancer specimen, N442S in a pancreatic cancer cell line Pa21C, E511Q in a specimen of upper aerodigestive tract cancer, R567* in Jurkat, and E863Q in a lung cancer specimen (Fig. 1B).

To examine whether STK10(R634H) affects intracellular signaling pathway related to Fos, Myc, NF-κB, Bcl-xL, Notch, Wnt or Rho, we examined the reporter activity for each pathway under the expression of the wild-type or the R634H mutant of STK10. As shown in Fig. 2A, the R634H mutant significantly elevated the NF-κB reporter activity compared to the wild-type (P<0.05, Student’s t-test), but did not affect the other signaling systems. To further investigate the role of STK10 in the NF-κB signaling, we generated cDNA for a kinase-dead mutant (K65I) of STK10, and examined its effects on the NF-κB signaling. Of note, the wild-type STK10 rather suppressed the NF-κB pathway, but the R634H mutation abolished such effect, suggesting that STK10 is a negative regulator for intracellular NF-κB activity (Fig. 2B). This hypothesis was further reinforced by the fact that the kinase-inactive K65I mutant markedly elevates the NF-κB reporter signaling. We further asked the effects of other STK10 mutants on NF-κB, and revealed that the L85P and K277E mutants found in the COSMIC database clearly induced NF-κB (Fig. 2C). Noteworthy, these mutations were somatically acquired in ovarian and testicular cancers, respectively. The other STK10 mutations did not affect NF-κB signaling (Fig. 2C and data not shown).

To further confirm the effects of STK10 mutants on the NF-κB activation, expression levels of its target genes were quantified among cells expressing the wild-type or mutant forms of STK10 (Fig. 2D). Consistent with the results of luciferase assay, L85P, K277E, R634H and K65I mutants each induced the expression of NF-κB target genes including TLR2, BCL2 and CCL2, compared to the wild-form. These results indicate that polymorphism as well as somatic mutations of STK10 may regulate NF-κB signaling in various human cancers.

It is not clear yet how STK10 regulates the NF-κB pathway. When HEK293 cells were stimulated with tumor necrosis factor (TNF)-α, K65I amino acid change cancelled the STK10-mediated NF-κB suppression (Fig. 2E). However, such regulation was not observed on the IKK complex-driven NF-κB activation, indicating that STK10 may interact with the NF-κB pathway between TNF receptors and the IKK complex.

To examine if the STK10 mutations directly affect its enzymatic activity, FLAG-tagged wild-type or mutant forms of STK10 was expressed in HEK293 cells, and subjected to immunoprecipitation with antibodies to FLAG, followed by an in vitro kinase assay with histone H2A as an exogenous substrate. As shown in Fig. 3A, while the wild-type STK10 clearly phosphorylates histone H2A, the L85P, K277E and K65I mutants remarkably attenuate the substrate phosphorylation. We could not, however, observe a notable difference between the wild-type and the R634H mutant in their ability of H2A phosphorylation, suggesting that R634H mutation may regulate the NF-κB signaling not through suppressing its enzymatic activity but through affecting its interaction to other proteins.

Of note, the R411fs and R567* mutants did not phosphorylate histone H2A, indicating the coiled-coil domain and/or carboxyl terminal end of STK10 (Fig. 1B) is indispensable for histone phosphorylation. Immunoblot analysis with antibodies to FLAG revealed that protein amounts of the L85P, K277E and K65I mutants were markedly reduced, which may account for the decreased phosphorylation of histone H2A.

Since STK10 is also known to phosphorylate PLK1 (14), we further investigated the phosphorylation level of PLK1 when co-expressed with the wild-type or mutant forms of SKT10 (Fig. 3B). Consistent with the result of the in vitro kinase assay, phosphorylation of PLK1 was reduced under the presence of the L85P, K277E or K65I mutant. Again, R634H mutation did not significantly affect the PLK1 phosphorylation. Strikingly, however, both of R411fs and R567* mutants could phosphorylate PLK1 equally to, or greater than, the wild-type. The carboxyl-terminal half of STK10 may, thus, be dispensable for the phosphorylation of PLK1 in vivo.

NF-κB plays pivotal roles in inflammatory and anti-apoptotic responses in cells, and constitutive NF-κB activation has been recognized as a hallmark of several lymphoid malignancies (15). Given the NF-κB-activating potential of STK10 mutants, we then asked whether such mutants also regulate cell apoptosis. Glucocorticoids are known to be a potent apoptosis-inducer in lymphocytes, and have been used in therapeutic regimens for lymphoid malignancies (15). As shown in Fig. 4A, treatment of a mouse T-cell line, CTLL-2, with dexamethasone inhibited cell proliferation, and, importantly, introduction of the wild-type STK10 markedly augmented such effect, again confirming its tumor-suppressive role. In contrast, introduction of NF-κB-activating mutants restored cell growth, in a parallel manner to their NF-κB-activating potential.

To determine whether the growth-inhibitory effect thus observed was related to apoptosis, we conducted the Annexin V/PI staining assay (Fig. 4B). Although an increase in the Annexin V-positive fraction was apparent in dexamethasone-treated, wild-type STK10-expressing CTLL-2 cells, introduction of the L85P, K277E, R634H or K65 mutant of STK10 significantly suppressed apoptosis. Thus, a part of polymorphism/somatic mutations within STK10 not only activate the NF-κB pathway, but directly exert anti-apoptotic function.

As it was demonstrated that some STK10 mutations induce NF-κB activation and render resistance to dexamethasone-induced apoptosis, we further searched for STK10 mutations in genomic DNAs of 92 PTCL specimens. It should be noted, however, that PTCL specimens frequently contain numerous normal cells in addition to neoplastic cells, which makes it difficult to determine the presence or absence of STK10 mutations with conventional Sanger sequencing. We therefore chose a deep sequencing approach with the GAIIx system (16), leading to the isolation of one specimen carrying STK10(R634H) (data not shown).