Sedimentation and gel retardation studies show a stronger interaction of HMG 1 and 2 with negatively supercoiled DNA than with linear, nicked-circular, or positively supercoiled ds-DNA. An apparent unwinding angle of 58 degrees was obtained for HMG 1 and 2 when assayed by protection of negatively supercoiled DNA from topoisomerase I relaxation or when assayed by the supercoiling of nicked-circular DNA with T4 DNA ligase. The protection of negatively supercoiled DNA was linear up to molar ratios of about 250:1. There was little change in binding reactions or in the protection of supercoiled DNA at ratios above 250:1, indicating that both activities saturate and that HMG 1 and 2 have binding site sizes of about 20 bp. P1, the major tryptic fragment of HMG 1 or 2 which retains the two DNA binding HMG 1/2 boxes, displays a 2-fold increase in binding to all types of ds-DNA compared to intact HMG 1 or 2. However P1 protects negatively supercoiled DNA from topoisomerase I relaxation about 5-fold less than intact HMG 1 or 2. Complete protection with P1 occurs at a molar ratio 1040:1, indicating a DNA binding site size of about 4 bp and an apparent unwinding angle of 10 degrees. P1 binding to closed-circular ss-DNA also involves a binding site of about 4 bp. Adding the acidic C-terminal fragment to P1 reversed its binding and allowed topoisomerase I to relax supercoiled DNA. These findings highlight the importance of the acidic C-terminal domains of HMG 1 and 2 in limiting electrostatic interactions of the HMG 1/2 boxes with ds- or ss-DNA. N-Ethylmaleimide inhibited the binding of intact HMG 1 or 2 to negatively supercoiled DNA, but did not inhibit the electrostatic binding of HMG 1 or 2 to ss-DNA, or of P1 to any form of DNA (ds or ss). These results suggest that cysteine residues are involved in the specific interaction of HMG 1 or 2 with negatively supercoiled DNA and that the acidic C-terminal domains modulate an intramolecular conformational change involving sulfhydryls within the HMG 1/2 boxes.

The high mobility group protein HMG 14, which is preferentially associated with nucleosomes containing active gene sequences, is phosphorylated on different sites according to the tissue and stimulus being studied. In the thyroid, HMG 14 displays TSH-dependent phosphorylation that is mediated by cAMP-dependent protein kinase (A-kinase). We have, therefore, studied how phosphorylation of HMG 14 on its major and minor A-kinase sites (Ser-6 and -24) affects its interactions with nucleosomes and various forms of DNA, since this could reflect a means of regulating its function of binding to active chromatin. Approximately twice as much Ser-6 phospho- and 4 times as much Ser-6,24 diphospho-HMG 14 were required to produce the same degree of nucleosome band displacement as that caused by native unphosphorylated HMG 14. Phosphorylation also reduced the ability of HMG 14 to protect the ends of nucleosomal DNA from thermal denaturation. When the electrophoretic mobility of naked DNA was examined, the Ser-6 phospho-HMG 14 was about half as effective as native HMG 14 in retarding the various forms of double stranded DNA, and Ser-6,24 diphospho-HMG 14 was even less effective. Our data demonstrate that electrostatic interactions between DNA and basic amino acids in two highly conserved regions (residues 1-5 and 16-27) can be modulated by phosphorylation at Ser-6 and Ser-24. The ability of mammalian HMG 14, but not HMG 17, to display hormone-dependent phosphorylation may indicate a route for differentially modulating their binding to transcriptionally active chromatin.

Treating mice with T4 increases the level of immunoreactive epidermal growth factor (EGF) in the thyroid. In order to establish whether this response might reflect a T4-dependent increase in the levels of messenger RNA (mRNA) in the thyroid, we prepared an internal standard which permits us to quantitate EGF mRNA levels by reverse transcription plus polymerase chain reaction amplification (RT-PCR). Our synthetic EGF mRNA construct contains the same flanking primer sequences used to amplify mature EGF message, but 70 bases were eliminated from the center of the 277-base EGF sequence to permit the PCR product of this internal standard to be distinguished by its smaller size (EGF 207). This synthetic mRNA also contains a poly(A)tail, which permits it to be reverse transcribed. We then added a range of concentrations of this internal standard mRNA to aliquots of total RNA from each pair of thyroid lobes and determined the concentration of EGF 207 at which the PCR primers were incorporated equally into the 277 and 207 bands after RT-PCR. Thyroid RNA from male Balb/c mice treated with T4 (0.25 micrograms/g.day) for 14 days contained 2.8-fold more EGF mRNA than RNA from control mice (P < 0.01). Competitive RT-PCR EGF mRNA levels were determined for thyroid RNA samples from mice treated with T4 for various times up to 14 days. The most significant increase occurred after 1 day's treatment (P < 0.005). This demonstration of a thyroid hormone-dependent increase in the level of thyroidal EGF mRNA adds support to the concept that EGF may function as an autocrine/paracrine regulator of thyroid function.

No abstract

No abstract