The ataxia telangiectasia and Rad3-related (ATR)-dependent DNA damage checkpoint is one of the major DNA damage responses. The ATR pathway is required to maintain genome integrity and ensure faithful duplication of the genomic DNA in the S phase. Once activated by DNA lesions or replication stress, ATR kinase phosphorylates downstream substrates to activate various safeguarding mechanisms including cell cycle arrest and cellular senescence. Cancer cells suffer from high levels of DNA replication stress and are more dependent than normal cells on the ATR pathway. The components of the ATR pathway have also attracted attention as therapeutic targets in cancer treatment [1], [2].

Rad17 initiates the ATR pathway and recognizes damaged DNA to activate the DNA damage checkpoint. Rad17 shows sequence and structural homology to replication factor C (RFC) proteins [3], and Rad17 and small subunits of the canonical RFC complex, RFC2–5 proteins, compose an RFC-like Rad17–RFC2–5 complex in which RFC1 is replaced by Rad17 [4]. Rad17 and RFC2–5 proteins share similar domain structures. The N-terminal halves comprise the AAA+ ATPase domains, and the C-terminal halves contain several α-helices that are involved in the interaction between the RFC subunits and the RFC-like complex formation [5]. The Rad9–Hus1–Rad1 (9–11) complex is a PCNA-like heterotrimeric complex [6], [7], and the Rad17–RFC2–5 complex loads the 9–11 complex onto damaged chromatin to activate ATR kinase [8], [9], [10]. The Rad17–RFC2–5 complex is often called a clamp loader because it loads the 9–11 clamp on damaged DNA.

The structures of the Rad17–RFC2–5 and 9–11 complexes in human [5] and those of orthologous RAD24–RFC2–5 and DDC1–MEC3–RAD17 complexes in Saccharomyces cerevisiae (S.c.) [11], [12] were recently reported. The AAA+ ATPase domains of human Rad17 and S.c. RAD24 substantially interacted with the 9–11 complex and occupied the largest interaction surface. The main interface was composed of the KYxxL motif [5], [12] that we previously reported as an essential amino acid sequence for the interaction between Rad17 and the 9–11 complex in vivo [13].

In the canonical RFC and PCNA complexes, RFC binds PCNA in such a way that four of its five subunits bind ATP, and ATP hydrolysis releases PCNA from the RFC complex [14]. The Rad17–RFC2–5 complex is also regulated by ATP binding as an AAA+ ATPase complex. The interaction between Rad17 and the 9–11 complex requires nucleotide binding, not hydrolysis, in the ATPase domains of Rad17 and the RFC subunits [15], [16], [17]. The interaction is promoted by ATP, ADP, ATPγS, and dATP [8]. ATP hydrolysis by the RFC2–5 subunits results in a conformational change in the Rad17–RFC2–5 complex that may disrupt the interaction between the Rad17–RFC2–5 and 9–11 complexes [11].

The function of the Rad17 protein is regulated by several loops between or outside the globular domains: The N-terminal loop encodes tandem destruction boxes [18], and the loop between the ATPase domain and the C-terminal α-helical domain, which we named the central basic domain, encodes nuclear and nucleolar localization signals [19]. However, the reported structure of the human Rad17–RFC2–5 complex lacks those loops, including the C-terminal unstructured region [5]. We previously reported that a C-terminal tail of human Rad17 is embedded with an amino acid motif named iVERGE (IxxYxS in Vertebrates at the Edge of Rad17 with Glutamate/aspartate Extension) that is essential for the interaction between the Rad17–RFC2–5 and 9–11 complexes [20]. At the C-terminal end of Rad17, a stretch of acidic residues extends to the terminal carboxyl group. In the polyanionic tail, an amino acid sequence IxxYxS is conserved among Rad17 proteins in vertebrates, and the IxxYxS motif is essential for the interaction between Rad17 and the 9–11 complex. The conserved serine residue in the IxxYxS motif, S667 in human Rad17, is phosphorylated by CK2, and the phosphorylation is crucial for the interaction with the 9–11 complex [21]. The C-terminal T670 is also phosphorylated by CK1δ/ε, and the phosphorylation promotes the S667 phosphorylation by CK2 [22]. The iVERGE was not visualized in the reported structure of the human Rad17–RFC2–5 and 9–11 complexes [5] and is not conserved in Saccharomyces cerevisiae. The iVERGE peptide is conserved in vertebrates, but the underlying molecular mechanism remains unknown. Here, we show that iVERGE directly interacts with the Hus1 subunit of the 9–11 complex independently of the AAA+ ATPase domains in vivo. The binding conformation was predicted by de novo modeling in silico and was confirmed in vivo. The binding energy calculation showed that the iVERGE peptide and the KyxxL motif are the main interaction surfaces between the Rad17–RFC2–5 and 9–11 complexes. Our data reveal another crucial molecular interface between the Rad17–RFC2–5 and 9–11 complexes in vertebrates that connects both complexes independently of the ATPase domains.