APE1 recruits ATRIP to ssDNA in an RPA-independent manner to promote the ATR DNA damage response

Cells have evolved the DNA damage response (DDR) pathways in response to DNA replication stress or DNA damage. In the ATR-Chk1 DDR pathway, it has been proposed that ATR is recruited to RPA-coated single-strand DNA (ssDNA) by direct ATRIP-RPA interaction. However, it remains elusive whether and how ATRIP is recruited to ssDNA in an RPA-independent manner. Here, we provide evidence that APE1 directly associates ssDNA to recruit ATRIP onto ssDNA in an RPA-independent fashion. The N-terminal motif within APE1 is required and sufficient for the APE1-ATRIP interaction in vitro and the distinct APE1-ATRIP interaction is required for ATRIP recruitment to ssDNA and the ATR-Chk1 DDR pathway activation in Xenopus egg extracts. In addition, APE1 directly associates with RPA70 and RPA32 via two distinct motifs. Taken together, our evidence identifies APE1 as a direct recruiter of ATRIP onto ssDNA independent of RPA in the activation of ATR DDR pathway. Summary APE1 associates with ssDNA and ATRIP directly via distinct motifs and thereby recruits ATRIP onto ssDNA independent of RPA to promote the ATR DDR. APE1 interacts with RPA via distinct two motifs in vitro but such RPA-APE1 interaction is dispensable for ATRIP recruitment onto ssDNA.


Introduction
gaps. We chose to test two 100bp-dsDNA structures with either 30nt or 80nt ssDNA gap 154 covalently linked with 5'-biotin on top strand (designated as "30nt gap" and "80nt gap") for 155 subsequent Streptavidin magnetic bead-bound isolation and analysis from Xenopus egg extracts 156 (Top of the panel of Figure 1B). When equal moles of dsDNA with 30nt or 80nt ssDNA gap were 157 added to HSS, more RPA70 and RPA32 as well as APE1 and ATRIP are recruited to the 80nt-158 ssDNA gap and Chk1 phosphorylation was also enhanced ( Figure 1B). This enhanced Chk1 159 phosphorylation is likely due to increased RPA complex recruitment onto the 80nt-ssDNA gap 160 ( Figure 1B). Whereas APE1 depletion led to compromised Chk1 phosphorylation, the recruitment 161 of ATRIP but not RPA70 nor RPA32 was compromised in APE1-depleted HSS ( Figure 1B). Our 162 observations so far suggest that APE1 is important for the recruitment of ATRIP onto RPA-coated 163 ssDNA and Chk1 phosphorylation in Xenopus HSS. 164 Next, to determine whether APE1 plays any role in the RPA-dependent ATRIP recruitment onto 165 ssDNA, we tested whether His-tagged ATRIP recombinant protein can be recruited to 30nt or 166 80nt ssDNA gap in vitro in the absence or presence of equal moles of recombinant RPA complex. 167 Consistent with previously reported RPA-dependent ATRIP recruitment to ssDNA (Zou and 168 Elledge, 2003), 30/80nt-ssDNA coated with RPA70 and RPA32 significantly enhanced the 169 recruitment of His-ATRIP in vitro, although almost no binding of ATRIP onto ssDNA (30nt and 170 80mt) was observed in the absence of recombinant RPA complex ( Figure 1C). We noticed more 171 binding of His-ATRIP onto 80nt ssDNA gap compared with 30nt ssDNA gap ( Figure 1C). As 172 expected, the recruitment of His-ATRIP onto 30nt and 80nt ssDNA was similar to each when 173 same amount of ssDNA gap structures was coupled to beads ( Figure 1-figure supplement 1A). 174 Furthermore, the addition of GST-APE1 but not GST protein increased the recruitment of His-175 ATRIP onto ssDNA with the presence of His-RPA complex in vitro ( Figure 1D). It is worth to note 176 that the presence of GST-APE1 had almost no noticeable effect on the recruitment of His-RPA70 177 and His-RPA32 to ssDNA gap structures ( Figure 1D). Similarly, the presence of GST-APE1 but 178 not GST increased the recruitment of endogenous ATRIP but not endogenous RPA70/RPA32 to 179 ssDNA gap structures in the Xenopus HSS (Figure 1-figure supplement 1B). Whereas RPA itself 180 is sufficient for ATRIP recruitment onto ssDNA in vitro, our observations here suggest that APE1 181 may stimulate the RPA-dependent ATRIP recruitment onto ssDNA in vitro. Alternatively, it is 182 possible that APE1 may play an additional but direct role in the recruitment of ATRIP onto ssDNA 183 in vitro that is independent of RPA. 184

APE1 recognizes and binds with ssDNA directly in a length-dependent manner in vitro 185
Although APE1 is known as a DNA repair protein to specifically recognize and process AP site, it 186 remains unclear whether and how APE1 interacts with ssDNA. To identify the possible direct role 187 of APE1 in ATRIP recruitment onto ssDNA, we first performed systematic analysis of APE1 188 association with ssDNA. Our bead-bound experiments showed that GST-APE1 but not GST was 189 recruited onto 30nt and 80nt ssDNA gap structures in vitro (Figure 2A-B). We also determined 190 that GST-APE1 but not GST was recruited onto beads coupled with 70nt ssDNA in vitro ( Figure  191 2- figure supplement 1). Furthermore, we demonstrated that GST-APE1 but not GST was recruited 192 to beads coupled with 40nt, 60nt, and 80nt ssDNA, but not 10nt nor 20nt ssDNA ( Figure 2C). 193 Furthermore, the longer ssDNA is, the more GST-APE1 is recruited ( Figure 2C). Collectively,194 these observations suggest an APE1-ssDNA interaction in a length-dependent manner in vitro 195 (30-80nt) regardless the ssDNA is alone or in gapped structures. 196 To further dissect domain requirements within APE1 for ssDNA association, we generated a 197 series of deletion GST-tagged APE1 and found that WT GST-APE1 and AA101-316 GST-APE1 198 but not any other deletion GST-APE1 tested (i.e., AA35-316, AA1-100, AA1-34, AA35-100, 199 AA101-200) associated with beads coupled with 70nt-ssDNA in vitro (Figure 2A and 2D). Intriguingly, AA101-316 but not AA35-316 GST-APE1 associated with ssDNA ( Figure 2D). We 201 speculate that the fragment of AA35-100 within APE1 may somehow inhibit the APE1-ssDNA 202 association due to a currently unknown mechanism. In addition, our EMSA assays revealed that 203 WT GST-APE1 but not GST formed protein-ssDNA complex in vitro ( Figure 2E). Notably, neither 204 AA35-316 nor AA1-34 GST-APE1 formed protein-ssDNA complex in EMSA assays ( Figure 2E). 205 These observations suggest that AA1-34 within APE1 is required but seems not sufficient for 206 ssDNA association at least under our tested conditions, and that APE1 AA35-316 is deficient for 207 ssDNA association while APE1 AA101-316 is proficient in ssDNA interaction ( Figure 2). 208 What are the effects of N-terminal motif of APE1 for its 3'-5' exonuclease and AP endonuclease 209 activities? Similar to our previous report (Lin et al., 2020), WT GST-APE1 but neither ED (E95Q-210 D306A) GST-APE1 nor GST displayed 3'-5' exonuclease and AP endonuclease activities ( Figure  211 2- figure supplement 2A-B). Notably, AA101-316 GST-APE1 is defective for 3'-5' exonuclease and 212 AP endonuclease activities (Figure 2-figure supplement 2A-B); however, AA35-316 GST-APE1 213 is proficient in AP endonuclease activity but deficient for 3'-5' exonuclease activity (Figure 2-figure  214 supplement 2C-D). These observations suggest the importance of the AA1-34 motif of APE1 for 215 its 3'-5' exonuclease activity and the AA35-100 motif within APE1 for its AP endonuclease activity. 216 APE1 interacts and recruits ATRIP onto ssDNA in an RPA-independent manner in vitro and 217 promotes the ATR DDR pathway in Xenopus egg extracts using a non-catalytic mechanism 218 We next tested whether and how APE1 might interact with ATRIP directly by protein-protein 219 interaction assays. GST pulldown assays showed that GST-APE1 but not GST directly interacted 220 with His-ATRIP in vitro (Figures 2A and 3A). Domain dissection experiments revealed that both 221 AA35-316 GST-APE1 and AA1-100 GST-APE1 associated with His-ATRIP to the similar capacity 222 as WT GST-APE1 ( Figure 3A). However, AA101-316 GST-APE1 and other fragments of APE1 223 tested (i.e., AA1-34, AA35-100, and AA101-200) were deficient for interaction with His-ATRIP 224 ( Figure 3A). In addition, neither of the point mutants within GST-APE1's active sites (i.e., ED, 225 D306A, and C92A-C98A) affected the APE1-ATRIP interaction ( Figure 3B), although they are 226 deficient for 3'-5' exonuclease as shown previously (Lin et al., 2020). Thus, our findings indicate 227 that AA35-100 within APE1 is required for ATRIP interaction and AA1-100 is the minimum 228 fragment within APE1 sufficient for ATRIP association in vitro (Figures 2A and 3A). 229 Based on the observation of direct APE1-ATRIP interaction ( Figure 3A), we intended to test 230 whether APE1 could recruit ATRIP onto ssDNA directly in the absence of RPA in vitro. We found 231 that His-ATRIP protein was recruited onto 30nt and 80nt ssDNA gap structures in the presence 232 of WT GST-APE1 but not GST (Compare Lanes 4-6 and Lane 1-3 in "Bead-bound", Figure 3C). 233 Due to its deficiency in ssDNA interaction ( Figures 2D and 3A), AA35-316 GST-APE1 was not 234 recruited to 30nt and 80nt ssDNA gap structures, which led to the insufficient recruitment of His-235 ATRIP onto ssDNA (Lanes 10-12 in "Bead-bound", Figure 3C). Notably, AA101-316 GST-APE1 236 was recruited to 30nt and 80nt ssDNA gap structures but could not recruit ATRIP to ssDNA, due 237 to deficiency in ATRIP association (Lanes 7-9 in "Bead-bound", Figure 3C). These observations 238 strongly support that APE1 interacts with ssDNA via its AA1-34 fragment and recruits ATRIP onto 239 ssDNA via its AA1-100 in in vitro reconstitution systems, and that such APE1-mediated ATRIP 240 recruitment onto ssDNA is independent of RPA. 241 To test the biological significance of APE1-mediated ATRIP onto ssDNA, we performed rescue 242 experiments in APE1-depleted HSS. WT GST-APE1 but not AA101-316 GST-APE1 rescued the 243 recruitment of endogenous ATRIP onto 30nt and 80nt ssDNA gap structures and subsequent 244 Chk1 phosphorylation, although endogenous RPA70 and RPA32 as well as WT/AA101-316 GST-245 APE1 associated with ssDNA gap structures in APE1-depleted HSS (Compare Lanes 4-6 and 246 Lanes 7-9, Figure 3D). This observation indicates the significance of the APE1-ATRIP interaction 247 for ATRIP recruitment onto ssDNA and subsequent ATR DDR pathway activation in Xenopus egg 248

extracts. 249
In light of the significance of APE1 and its 3'-5' exonuclease in the initial end resection of defined 250 SSB structures and subsequent ATR DDR pathway in Xenopus HSS system (Lin et al., 2020), 251 we sought to test whether APE1's catalytic function plays a vital role in the direct ATRIP regulation 252 and chose to use ED GST-APE1 lacking 3'-5' exonuclease and AP endonuclease (Figure 2-figure  253 supplement 2A-B) (Lin et al., 2020). We found that similar to WT GST-APE1, ED GST-APE1 254 bound with ssDNA and recruited endogenous ATRIP onto ssDNA and rescued Chk1 255 phosphorylation in APE1-depleted HSS (Lanes 10-12, Figure 3D). This observation suggests that 256 APE1's nuclease activity is dispensable for its direct recruitment of ATRIP onto ssDNA gap 257 structures and subsequent ATR-Chk1 DDR pathway activation in the Xenopus HSS system. 258 Together, these findings demonstrate that APE1 directly associates with and recruits ATRIP onto 259 ssDNA in vitro and that APE1 recruits ATRIP onto ssDNA via a non-catalytic function to promote 260 the ATR-Chk1 DDR pathway activation in the Xenopus HSS system. 261

APE1 interacts with RPA70 and RPA32 via two distinct binding motifs 262
We have shown that both RPA complex and APE1 can independently recruit recombinant ATRIP 263 onto ssDNA in vitro ( Figures 1C and 3C), and that APE1 is required for the recruitment of ATRIP 264 onto RPA-coated ssDNA in Xenopus HSS system ( Figure 1B). Next, we sought to determine 265 whether APE1 interacts with RPA directly, and, if so, whether RPA plays a role for APE1-mediated it is not feasible technically to test whether RPA depletion directly or indirectly affects APE1-269 mediated ATRIP recruitment and the subsequent ATR-Chk1 DDR pathway in the HSS system. 270 Our strategy is to identify RPA-interaction motifs within APE1 in vitro, and to determine whether 271 such a mutant APE1 deficient in RPA-interaction still recruits ATRIP onto ssDNA in the HSS 272

system. 273
First, we tested the possibility that recombinant GST-APE1 might interact with purified 274 recombinant His-RPA complex (RPA70, RPA32, and RPA14) by protein-protein interaction 275 assays. Our GST pulldown assays showed that WT GST-APE1 but not GST interacted with His-276 RPA70 in vitro ( Figure 4A). Almost no noticeable effects were observed for the interaction of ED, 277 DA, and CA GST-APE1 with His-RPA70, suggesting that the E95, D209, D306, C92, and C98 278 residues in APE1 are not critical for ATRIP interaction (Figure 4-figure supplement 1A). Domain 279 dissection experiments revealed that AA35-316 and AA1-100 GST-APE1 associated with His-280 RPA70 in a similar capacity to WT GST-APE1 ( Figure 4A). The binding to His-RPA70 was 281 decreased but nevertheless not completely eliminated in other deletion fragments of GST-APE1 282 tested (i.e., AA101-316, AA1-34, AA35-100, AA101-200) ( Figure 4A). These observations 283 suggest (I) that the first 100 amino acids of APE1 are important for RPA association and (II) that 284 more than one binding sites within APE1 may mediate interaction with RPA complex. 285 Second, to further test our hypothesis of multiple bindings sites of APE1 for its interaction with 286 RPA complex, we performed amino acid sequence alignments of APE1 (Xenopus APE1 and 287 human APE1) to several human RPA70-interacting proteins (e.g., ETAA1, ATRIP, RAD9A, NBS1, 288 and Mre11) and RPA32-interacting proteins (e.g., ETAA1, XPA, SMARCAL1, and TIPIN) and 289 found that APE1 contains a putative RPA70-binding motif (15AA, designated as RBM1) and a 290 putative RPA32-binding motif (41AA, designated as RBM2) ( Figure 4B  RPA70 and RPA32 in vitro using previously uncharacterized two distinct motifs within APE1 (i.e., 303 RBM1 and RBM2) ( Figure 4B). 304 Our earlier result showed that APE1-depletion led to defective ATRIP recruitment to ssDNA at 305 gap structures and Chk1 phosphorylation in Xenopus HSS system ( Figure 1B). Our rescue 306 experiments showed that similar to WT GST-APE1, the RPA-interaction-deficient mutant GST-307 APE1 (RBM1-M, RBM2-M, and RBM-DM) rescued the recruitment of endogenous ATRIP and 308 RPA70/RPA32 onto ssDNA and subsequent Chk1 phosphorylation in APE1-depleted HSS 309 ( Figure 4E). This result suggests that the RPA-APE1 interaction may be dispensable for the 310 APE1-mediated ATRIP recruitment onto ssDNA in Xenopus egg extracts. In addition, we tested 311 whether RPA plays any role in the APE1-ssDNA interaction in reconstitution system. Based on 312 the length-dependent APE1 association with ssDNA ( Figure 2C), we added excess recombinant 313 His-RPA complex and found that APE1 interaction with longer ssDNA (40nt, 60nt, and 80nt) was 314 ATRIP onto ssDNA for subsequent ATR-Chk1 activation in Xenopus egg extracts ( Figure 3D). 332 Notably, the nuclease deficient ED APE1 is still proficient for ATRIP recruitment and ATR DDR in 333 Xenopus egg extracts, similar to WT APE1 ( Figure 3D). Together, these evidences support a non-334 catalytic role of APE1 via the direct recruitment of ATRIP onto ssDNA in the ATR-Chk1 DDR 335

pathway. 336
A previous study has demonstrated that APE1 can incise the AP site within ssDNA in a sequence-337 and secondary-structure-dependent manner (Fan et al., 2006). Although this finding implies that 338 that APE1 may associate with ssDNA, the molecular determinants of APE1 for ssDNA interaction 339 is elusive. Our data in this study indicate that APE1 directly associates with ssDNA (>30nt length) Future studies are needed to distinguish whether APE1 associates with ssDNA and ssRNA using 347 similar or distinct mechanisms. 348 To the best of our knowledge, it is the first time showing that APE1 direclty associates with RPA 349 via its two distinct motifs (Figure 4). A previous study mentioned that His-tagged human APE1 350 protein did not associate with purified untagged RPA protein using Ni-NTA-bead-based pulldown The data in this study support that APE1 directly associates and recruits ATRIP onto ssDNA unknown Protein X may negatively regulate ATRIP binding to RPA complex especially RPA70 N-368 terminal domain, which has been involved in several proteins' recruitment such as ETAA1, Mre11, 369 Nbs1, and Rad9. The Protein X's negative regulation of ATRIP interaction may be counteracted 370 by APE1 in the HSS system. Future studies will test this hypothesis. Taken together, we propose 371 APE1 as a critical recruiter of ATRIP onto ssDNA in an RPA-independent manner to promote the 372 ATR-Chk1 DDR pathway. 373

Materials and methods 374
Key resources The preparation of Xenopus HSS and immunodepletion of target proteins in HSS were described The 39bp FAM-labeled dsDNA-AP structure for APE1 endonuclease assays was prepared as 392 previously described (Lin et al., 2020). As shown previously (Lin et al., 2020), the 70bp FAM-393 dsDNA structure was prepared and treated with Nt.BstNBI and CIP to make the FAM-dsDNA-394 SSB for APE1 exonuclease assays. The FAM-dsDNA-SSB structure was purified from agarose 395 via QIAquick gel extraction and then phenol-chloroform extraction. The 70nt FAM-ssDNA 396 structure in Figure 2E was synthesize as Oligo#1. 397 The 100bp Biotin-dsDNA structure with a 30nt or 80nt ssDNA gap (30nt gap or 80nt gap) in the

Electrophoretic mobility shift assays (EMSA) 436
The EMSA assays for testing DNA-Protein interaction were similar to methods described

Material availability statement 441
Materials generated in this study can be accessed by contacting the corresponding author. 442

Acknowledgments 443
We thank Drs. Matthew Michael, Karlene Cimprich, and Howard Lindsay for reagents. The Yan 444 lab was supported, in part, by grants from the NIH/NCI (R01CA225637) and the NIH/NIEHS 445 (R21ES032966), and funds from UNC Charlotte. 446 The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. Hus1-Rad1 (9-1-1) clamp activates checkpoint signaling via TopBP1. Genes Dev.   Biotin-labeled ssDNA with different lengths (10nt, 20nt, 40nt, 60nt, or 80nt) were added to an 688 interaction buffer containing GST or GST-APE1. After incubation for 30 min at room temperature, 689

Author contributions 450
the DNA-bound fractions and the input were examined via immunoblotting analysis. (D) 690 Streptavidin beads coupled with Biotin-labeled ssDNA (70nt) were added to an interaction buffer 691 containing (70nt) GST or WT or fragment of GST-APE1. After incubation for 30 min at room 692 temperature, the DNA-bound fractions and the input were examined via immunoblotting analysis.

693
(E) An EMSA assay shows the interaction between WT, AA35-316 and AA1-34 GST-APE1 and 694 the 70nt-ssDNA structure in vitro. 695 The online version of this article includes the following source data and figure supplements for 696 Streptavidin beads coupled with Biotin-labeled dsDNA with ssDNA gap structures (30nt or 80nt) 723 were added to APE1-depleted HSS, which was supplemented with GST or GST-tagged proteins 724 (WT, AA101-316, or ED GST-APE1) as indicated. DNA-bound fractions and total extract samples 725 were examined via immunoblotting analysis as indicated.   in Xenopus egg extracts. APE1 interacts with RPA via two distinct motifs; however, APE1 771 interaction with ATRIP, but not its interaction with RPA nor APE1 nuclease, is important to recruit 772 ATRIP onto ssDNA in Xenopus egg extracts to promote the ATR DDR activation. 773