Chemical synthesis of HMGA1a proteins enables elucidation of potential roles of phosphorylation in regulating the protein's DNA binding ability

Abstract

HMGA1a is a member of ‘High Mobility Group’ (HMG) protein family, among which HMGA family is named after the featured DNA-binding domain, AT-hook. HMGA1a takes active parts in nuclear activities via its flexible structure and its diverse post-translational modifications (PTMs). It is well acknowledged that HMGA1a adopts a flexible (rather than defined) structure when free in solution, whereas it readily undergoes disordered-to-ordered structure change upon binding events. Unfortunately, on the other hand, the study into the how the PTMs enable its versatile roles remains elusive. It is largely due to the heterogeneous and combinatory features inherent to protein PTMs, making it difficult to clearly correlate structure to function. Currently known PTMs of HMGA1a include Ser/Thr phosphorylation, Arg/Lys methylation, Lys acetylation, and ADP-ribosylation. To provide a tool towards further understanding of its PTMs, we developed a synthetic strategy via Hmb-assisted Ser/Thr ligation. This protein chemical ligation method allowed us to assembly the full-length HMGA1a in highly convergent manner, meanwhile to introduce modifications at desired sites at stoichiometric level. During the synthesis, the introduction of phosphoThr to Pro-bound peptidyl resin was found not an easy task; it was overcome by preparation of phosphoThr-Pro dipeptide as a building block. Moreover, the utility of Hmb (2-hydroxy-4-methoxybenzyl) turned out to be crucial to this synthetic strategy. The presence of Hmb not only facilitated the solid phase peptide synthesis of the Glu-rich C-terminal fragment of HMGA1a, but also contributed to the improvement of the otherwise poor ligation efficiency encountered in protein synthesis. With the established synthetic protocol, the full length HMGA1a can be obtained in around 14 working days on a multimilligram scale. The protocol enables the preparation of HMGA1a analogs with phosphorylations at flanked AT-hooks, as well as at C-terminal acidic tail; methylation at the 1st AT-hook was also achieved. Furthermore, we examined the interaction of these synthetic HMGA1a proteins with an AT-rich DNA sequence. It was found that co-existence of multiple phosphorylations at AT hooks (known as DNA-binding domains) of HMGA1a is necessary to decrease its DNA binding affinity, whereas phosphorylations at its C-terminal acidic tail (non DNA-binding domain) exert a surprisingly more pronounced effect. On the other hand, when the acidic tail of HMGA1a was deleted, the protein-DNA binding was slightly increased. These findings indicate that the acidic tail of HMGA1a did contribute to the interaction with DNA. One possibility is that the phosphorylation at the C-terminal tail shields intramolecularly the positively charged DNA-binding domain via the arginine (at AT hooks)-phosphate (at the acidic tail) ionic interaction. To test this hypothesis, we prepared another two HMGA1a analogs with photoaffinity labels at the C-terminus, which were subjected to photocrosslinking followed by proteomic digestion and mass spectrometry. The result showed some promising clue of such a hypothesis; meanwhile more direct evidence is needed to reach a solid conclusion. Taken together, this study not only provides an efficient and robust method to synthesize HMG protein family, but also paves way for comprehensive analysis of regulatory roles of diverse PTMs of HMG proteins in chemical epigenetics.