Archives
Mass spectrometry Matrix assisted laser desorption ionizatio
Mass spectrometry. Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) MS experiment was performed in linear and positive-ion mode on a DE-STR instrument equipped with a nitrogen laser (PE Biosystems, Foster City, CA). Each spectrum was calibrated in the default mode in which 0.1% mass accuracy was typically obtained. Purified protein samples were dissolved in an aqueous solution of 0.1% TFA and the sample aliquots were mixed with an equal volume of matrix solution, which was a saturated solution of α-cyano-4-hydroxycinnamic Methicillin sodium salt receptor in a solvent mixture of CH3CN, H2O, and TFA (50/50/0.1, v/v).
On-line LC–tandem mass spectrometry (MS/MS) was employed for peptide sequencing, and a 0.32×150mm C18 capillary column (300Å in pore size, 5μm in particle size, Micro-Tech Scientific, Vista, CA) was used. The Surveyor HPLC system was employed for the LC–MS/MS experiment, and a home-made pre-column splitter was used. The flow rate was 4–6μL/min after splitting, and a 63-min gradient of 2–65% CH3CN in 0.6% aqueous solution of acetic acid was employed. The effluent from the HPLC column was directed to an LCQ Deca XP ion-trap mass spectrometer (ThermoFinnigan, San Jose, CA). Mass calibration was carried out by using caffeine, a tetrapeptide MRFA, and Ultramark 1621 that were supplied by the instrument vendor. The spray voltage was 4.0kV, and the capillary temperature was maintained at 225°C. MS/MS was done in data-dependent scan mode by selecting the most abundant protonated ions observed in MS mode for collisional activation. The mass width for precursor ion isolation was 2.5 m/z units and the collision gas was helium. To achieve better signal-to-noise ratio in MS/MS, LC–MS/MS experiments were also carried out to monitor the fragmentation of one or several pre-selected precursor ions.
Results
Discussion
Bovine HMGN1 was reported to be phosphorylated by protein kinase CK2 in vitro, and Ser89 and Ser99 were identified as major and minor recognition sites, respectively [22], [25]. Here we confirmed that Ser88 and Ser98 in human HMGN1 protein, which correspond to Ser89 and Ser99 in the bovine protein, are two in vitro recognition sites of recombinant human protein kinase CK2. In addition, we identified three new sites, i.e., Ser6, Ser7, and Ser85, in human HMGN1 protein that can be phosphorylated by the CK2 kinase in vitro. These results parallel our observations with the HMGN1 protein isolated from MCF-7 human breast cancer cells and suggest that protein kinase CK2 could be the enzyme catalyzing the phosphorylation of these five serine residues in HMGN1 protein in MCF-7 cells in vivo [23].
A recent analysis of 308 sites that are phosphorylated by protein kinase CK2 highlights the paramount importance of amino acids with negatively charged side chains that are at positions n+3 (the most crucial one), n+1, and n+2, where n is the position of phosphorylation [26]. In this context, all the sites that we identified here can be considered as bona fide CK2 recognition sites (Fig. 1).
HMGN proteins play important roles in de-compacting the chromatin fiber and stabilizing the unfolded chromatin [5]. The C-terminal region of the HMGN1 protein contains the chromatin unfolding domain (CHUD), which unfolds the chromatin via interaction with histone H1 and core histone tails [27], [28]. The phosphorylation of the C-terminus of HMGN1 protein, due to its resulting increased negative charge, could displace DNA from positively charged histone H1 and core histone tails thereby de-compacting the chromatin. The unfolded chromatin provides the gateway through which transcription factors could access the target genomic DNA region and initiate gene transcription.
Ser6 in HMGN1 was previously identified as major in vitro substrate of A-kinase and G-kinase, with higher affinity to the G-kinase [19]. On the other hand, Ser6 and Ser7 are followed by a Glu at the n+3 and n+2 positions, respectively, which makes these two serines consensus substrates for protein kinase CK2. This is indeed confirmed by our in vitro phosphorylation experiments, though it remains unclear which kinase(s) might be involved in the phosphorylation of this residue in vivo. Ser6 and Ser7 are in the unique sequence region of HMGN1 protein which is close to the N-terminal nuclear-localization signal (NLS) region of the protein. The NLS region, which contains 2–4 basic amino acid residues such as Arg or Lys, is highly conserved in HMGN proteins and plays an important role in transporting HMGN proteins in and out of the nucleus via its interaction with the importin proteins [29], [30]. X-ray structure of the complex formed between importin β and the NLS peptide of parathyroid hormone-related protein suggested that phosphorylation at the close proximity of the NLS domain could destabilize the binding of NLS peptide to importin β[29]. In this respect, the phosphorylated side chain may clash onto the planer aromatic ring of Trp residues in the importin β-binding domain thereby weakening the interaction of the aromatic amino acid(s) with the basic residues in the NLS domain [29]. Therefore, the phosphorylation of Ser6 and Ser7 may compromise the binding of HMGN1 protein to the binding domain of importin proteins, which in turn affects the nuclear transport and sub-cellular localization of HMGN1 protein. Protein kinase CK2 could potentially be an enzyme that regulates this process.