Biochemistry. 2017 Sep 26;56(38):5112-5124. doi: 10.1021/acs.biochem.7b00689. 

Phospho-Priming Confers Functionally Relevant Specificities for Rad53 Kinase Autophosphorylation.

Chen ES1,2, Weng JH1,3, Chen YH1, Wang SC1, Liu XX1, Huang WC1, Matsui T4, Kawano Y5, Liao JH1, Lim LH1,2, Bessho Y1, Huang KF1, Wu WJ1, Tsai MD1,2.

1 Institute of Biological Chemistry, Academia Sinica , Taipei 115, Taiwan.

2 Institute of Biochemical Sciences, National Taiwan University , Taipei 106, Taiwan.

3 Institute of Biochemical Sciences, Department of Chemistry, National Tsing Hua University , Hsinchu 300, Taiwan.

4 Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Stanford University , Menlo Park, California 94025, United States.

5 RIKEN SPring-8 Center , 1-1-1 Kouto, Sayo, Hyogo 679-5148, Japan.



The vast majority of in vitro structural and functional studies of the activation mechanism of protein kinases use the kinase domain alone. Well-demonstrated effects of regulatory domains or allosteric factors are scarce for serine/threonine kinases. Here we use a site-specifically phosphorylated SCD1-FHA1-kinase three-domain construct of the serine/threonine kinase Rad53 to show the effect of phospho-priming, an in vivo regulatory mechanism, on the autophosphorylation intermediate and specificity. Unphosphorylated Rad53 is a flexible monomer in solution but is captured in an asymmetric enzyme:substrate complex in crystal with the two FHA domains separated from each other. Phospho-priming induces formation of a stable dimer via intermolecular pT-FHA binding in solution. Importantly, autophosphorylation of unprimed and phospho-primed Rad53 produced predominantly inactive pS350-Rad53 and active pT354-Rad53, respectively. The latter mechanism was also demonstrated in vivo. Our results show that, while Rad53 can display active conformations under various conditions, simulation of in vivo regulatory conditions confers functionally relevant autophosphorylation.

PMID: 28858528



Despite decades of extensive studies, it remains unclear why so many proteins (perhaps up to 50% of all human proteins) are phosphorylated, and why most of the phosphorylated proteins are phosphorylated at multiple sites. While more and more phosphorylations have been demonstrated to play important functional roles, the majority of phosphorylation sites still have unknown functions. The question “whether every phosphorylation has a specific functional role” still has no definitive answer for most proteins.


In fact, our knowledge about phosphorylation in biology is far less than we think. For example, we still don’t understand the molecular mechanism of most protein kinases, which are responsible for phosphorylation of other proteins. Many protein kinases need to be activated by phosphorylating itself (autophosphorylation). Many in vitro studies of autophosphorylation make use of the kinase domain alone, which is often robust in autophosphorylating itself. However, whether authophosphorylation of the kinase domain alone occurs at the correct site is often not verified.


Rad53 is a checkpoint kinase that plays a key role in the DNA damage response cascade of yeast. It consists of two FHA domains (ref 1) and a kinase domain. In addition, there is a cluster of four threonine sites at the N-terminus. Our previous study showed that phosphothreonine (pThr) containing peptides from the N-terminus can bind to the FHA1 domain of Rad53 (ref 2). In addition, we verified that phosphorylation at these N-terminal threonine sites actually occur in vivo (ref 3).


Based on the above studies reported previously, we formulated a hypothesis that phospho-priming at the N-terminal Thr sites is a prerequisite step (thus termed “phospho-priming”) for autophosphorylation at the correct site of the kinase domain of Rad53, leading to activation of the Rad53 kinase. In addition, we predicted that without phospho-priming, the kinase can still autophosphorylate itself, but may not be at the correct site. A major challenge of this study is the preparation of “phospho-primed Rad53 protein”, which was achieved by semi-synthetic methods. Subsequently we used mass spectrometry analyses to demonstrate that, in fact, phospho-primed Rad53 autophosphorylates itself at Thr354 of the activation loop, leading to active Rad53. On the other hand, in the absence of phospho-priming, autophosphorylation can also take place, but at Ser350 instead, leading to an inactive form of Rad53. The results are illustrated in Figure 1.


Our work did not stop here. Due to our long term interest in structures, we proceeded to solve the structure of the “non-phospho-primed Rad53 dimer” by X-ray crystallography, and the structure of “phospho-primed Rad53 dimer” by small angle X-ray scattering (SAXS) models. The results provide structural insight for the different site specificities of the two cases. It is clear from the structures that the binding between the FHA1 domain and the pThr site plays an important role in properly orienting the conformation of the dimer, which in turn ensures autophosphorylation at Thr354.


The potential significance of this work goes beyond the Rad53 kinase. It has important implications in several aspects: (1) For in vitro structural and mechanistic studies, it is important to use not just a single domain, but instead multi-domain protein constructs should be used, in order to mimic in vivo conditions as much as we can. (2) Likewise, we should use proteins that are specifically modified according to in vivo conditions. (3) Phosphorylation is only one of the many possible modifications and is perhaps the simplest one. However, even the preparation of site-specifically, homogeneously phosphorylated protein requires advanced chemical and biochemical approaches. It will take even more effort to prepare proteins with other modifications, such as glycosylation, sumoylation, etc. (4) Without making such special effort, it could be a significant over simplification to extrapolate the results of in vitro biophysical studies to in vivo conditions.



Figure 1. Phospho-priming leads to a dimer involving reciprocal binding between the FHA1 domain and one of the phosphothreonine at the N-terminus, which ensures functionally relevant autophosphorylation at Thr354. Without phospho-priming, autophosphorylation occurs at Ser350 which is a catalytically inactive form.



1.      “Structure and Function of the Phosphothreonine-Specific FHA Domain”.  Anjali Mahajan, Chunhua Yuan, Hyun Lee, Eric S.-W. Chen, Pei-Yu Wu, and Ming-Daw Tsai, Science Signaling 1, re12 (2008).  (Review)

2.       “Diphosphothreonine-specific interaction between SQ/TQ cluster and an FHA domain in the Rad53-Dun1 kinase cascade”.  Hyun Lee, Chunhua Yuan, Andrew Hammet, Anjali Mahajan, Eric S.-W. Chen, Ming-Ru Wu, Mei-I Su, Jörg Heierhorst, Ming-Daw Tsai, Mol. Cell 30, 767-778 (2008).

3.       “Use of quantitative mass spectrometric analysis to elucidate the mechanisms of phospho-priming and auto-activation of the checkpoint kinase Rad53 in vivo”. Eric S.-W. Chen, Nicolas C. Hoch, Shun-Chang Wang, Achille Pellicioli, Jörg Heierhorst, and Ming-Daw Tsai, Mol. Cell. Proteomics 13, 551-565 (2014).