No abstract

The redox regulation of Janus kinase 2 (JAK2) is poorly understood, and there are contradictory reports as to whether the enzyme's activity is inhibited or stimulated by oxidizing conditions in the cell. Here we demonstrate that multiple cysteine residues within the JAK2 catalytic domain may be crucial for enzymatic activity. The enzyme is catalytically inactive when oxidized; activity can be restored via reduction to the thiol state. A series of recombinant variants of JAK2 were overproduced using the baculoviral expression vector system. A truncated variant of JAK2, GST/(NDelta661)rJAK2, provided evidence that the amino-terminal autoinhibitory domain was not essential for direct redox regulation and that only nine cysteine residues were potentially involved. The effect of individually and combinatorially altering these nine cysteines was examined via cysteine-to-serine mutagenesis. This identified four cysteine residues in the catalytic domain (Cys866, Cys917, Cys1094, and Cys1105) that cooperatively maintain JAK2's catalytic competency. Our data are consistent with a direct mechanism for redox regulation of JAK2 via oxidation and reduction of critical cysteine residues.

Equilibrium binding data were analyzed to characterize the interaction of the linker histone H1 degrees with unmodified T4 phage DNA. Data were cast into the Scatchard-type plot described by McGhee and von Hippel and fit to their eponymous model for nonspecific binding of ligand to DNA. The data were not fit by the simple McGhee-von Hippel model, nor fit satisfactorily by the inclusion of a cooperativity parameter. Instead, the interaction appeared to be well described by Crothers' allosteric model, in which the higher affinity of the protein for one conformational form of the DNA drives an allosteric transition of the DNA to the conformational form with higher affinity (form 2). At 214 mM Na(+), the observed affinity K for an isolated site on unmodified T4 bacteriophage DNA in the form 2 conformation is 4.5 x 10(7) M(-1). The binding constant for an isolated site on DNA in the conformation with lower affinity, form 1, appears to be about 10-fold lower. Binding affinity is dependent on ion concentration: the magnitude of K is about 10-fold higher at 14 mM (5.9 x 10(8) M(-1) for form 2 DNA) than at 214 mM Na(+) concentration.

Histone H1 proteins bind to DNA and are important in formation and maintenance of chromatin structure. Little is known about differences among variant H1 histones in their interactions with DNA. We examined the effects of histones H1(0) and H1t on thermal denaturation of several DNA species. One of the DNA molecules was a 214-base-pair fragment from the plasmid pBR322, which contains an AT-rich and a GC-rich region. Both H1(0) and H1t bound preferentially to one region of the DNA fragment, a region that is relatively GC-rich. This result indicates that histones H1(0) and H1t are not totally nonspecific but rather bind with some sequence preference to DNA. This conclusion was supported by studies of other DNA species, including two 92-base-pair fragments derived from the two regions of the 214-mer, and several synthetic homocopolymers of DNA. Data obtained with the homocopolymers suggested that the binding preference was not simple preference for GC base pairs. The binding of the two H1 variants was not identical: there appear to be differences in binding site sizes, affinities, and sequence selectivities between H1t and H1(0).