Arabidopsis YEATS domain proteins facilitate DNA double-strand break repair via homology-directed pathways Authors: Neeraja Vegesna, Laura Bouza-Morcillo, Clara Bourbousse, Yasaman Jami-Alahmadi, En Li, Maherun Nisa, Ana Marie S. Palanca, James A. Wohlschlegel, Julie A. Law Published: 2026 ABSTRACT SignificanceDetecting and repairing damaged DNA is critical for maintaining cellular function. These processes occur within chromatin, yet how chromatin effectors (proteins that modulate the properties and compositions of chromatin) facilitate this process in plants remains poorly understood. Here we identified two chromatin readers, YAF9A and YAF9B, that are critical for double-strand break repair in the plant model, Arabidopsis thaliana. We further connect these activities to distinct chromatin remodeling complexes and demonstrate that the YAF9B reader domain is required for its role in promoting genome stability. These findings bridge a major gap in our understanding of chromatin mediated DNA repair in plants and highlight the importance of chromatin readers in coordinating different steps of the repair process. To maintain genome integrity in the face of both endogenous and exogenous sources of DNA damage, organisms have evolved cellular programs to detect and repair different kinds of lesions (1–3). These programs are critical to prevent the accumulation of deleterious mutations that can reduce organismal fitness, resulting in developmental defects (2, 4) or the progression of diseases, including cancer (3). These programs unfold within the context of chromatin—a biopolymer in which DNA is wrapped around histone proteins to form repeating arrays of nucleosomes (5). For example, during double-strand break (DSB) repair, the DNA damage response (DDR) results in dynamic alterations of the chromatin surrounding the break as well as extensive transcriptional changes, both of which are critical for proper DNA repair (1, 3, 6–9). While recent studies of plant responses to DSBs have shed light on the kinetics of gene expression changes in response to DNA damage (10) and have demonstrated roles for several chromatin remodeling complexes (7, 11), our knowledge of how chromatin facilitates DNA repair in plants remains poorly understood, lagging far behind our understanding of these processes in yeast and animals. While chromatin serves as the backdrop for DNA repair, it is not a uniform landscape. The positions, compositions, and compaction states of nucleosome arrays, as well as the histone proteins forming each nucleosome, can be modified by chromatin effectors to influence the accessibility and behavior of the underlying genetic material (5). These effectors can be broadly categorized into four groups: “Writers” that catalyze the addition of diverse chemical tags to specific residues within histone proteins, collectively termed histone modifications (e.g., methylation, acetylation, phosphorylation, and ubiquitination) (12). “Readers” that bind these modifications, “erasers” that catalyze their removal, and “remodelers” that use the energy of ATP to remove, slide, or exchange nucleosomes within chromatin (12). In the context of DNA repair, some pathways are more reliant on chromatin effector complexes and chromatin modifications than others. For example, the repair of DSBs via canonical nonhomologous end joining (NHEJ) pathways uses ligases to reseal unresected, broken DNA ends and thus, after the initial detection of DNA damage, this pathway requires relatively minimal alterations of chromatin (1, 3, 4). On the other hand, chromatin modifications and chromatin remodeling complexes play important roles during the initial steps of repair via noncanonical NHEJ pathways and at many steps important for repair via homologous recombination (HR) and homology-directed repair (HDR) pathways. In addition to facilitating the recognition and resection of DNA at breaks, specific chromatin modifications have been shown to mediate different repair steps, for example the recruitment of BRCA1 and RAD51 to promote HR (7, 9, 13–15). While much less is known about roles of chromatin in these downstream repair steps in plants (7, 15), many aspects of the initial recognition of DSBs and signaling of the DDR are highly conserved. Briefly, in plants, animals, and fungi, DSBs are recognized by conserved protein complexes that sense DNA damage and recruit serine/threonine kinases [i.e., ATAXIA-TELANGIECTASIA MUTATED (ATM) and/or ATAXIA-TELANGIECTASIA MUTATED AND RAD3-RELATED (ATR)] (1–3, 7). These kinases initiate signaling cascades through the phosphorylation of target proteins that regulate different aspects of the DNA damage response. These include, but are not limited to: phosphorylation of sensor proteins, which aid in the activation of ATM’s kinase activity (16), phosphorylation of the histone variant H2A.X, which marks DNA repair foci and is one of the best characterized histone modifications involved in the DNA damage response (2, 14), and phosphorylation of important transcription factors (TFs), namely SUPPRESSOR OF GAMMA-RESPONSE 1 (SOG1) in Arabidopsis (2, 17) and p53 in mammals (18–20). Although not phylogenetically related, these master regulators of the DNA damage response share some target genes (10) and serve similar functions in promoting genome stability by coordinating gene expression programs that promote DNA repair, modulate the cell cycle, and, in cases of extreme damage, initiate cell death (17, 19, 20). Another conserved feature of the DNA damage response is the involvement of complexes that remodel chromatin and alter histone acetylation levels at DSBs to facilitate DNA repair. In particular, the Nucleosomal Acetyltransferase of histone H4 (NuA4) and SWi2/snf2-Related (SWR1) complexes in yeast, as well as their SNF-2-related CREB-binding Activator Protein (SRCAP) and/or TIP60 counterparts in humans and drosophila, have interconnected and well characterized roles in regulating gene expression (21, 22) and coordinating DSB repair (23). These multisubunit complexes contain some distinct subunits, including the Essential SAS2-related Acetyltransferase 1 (Esa1) histone acetyltransferase in NuA4 and the SWi2/snf2-Related (Swr1) chromatin remodeler in SWR1 (23). However, they also have several shared accessory components, including Actin 1 (Act1), Actin Related Protein 4 (Arp4), Swr1 Complex 4 (Swc4), and Yeast ALL1-Fused gene from chromosome 9 (Yaf9) (23). Together these unique and shared components facilitate the recruitment and activities of these remodeling complexes at chromatin. In the case of DSB repair in yeast, both the NuA4 and SWR1 complexes are recruited to break sites via the interaction of one of their shared subunits, ARP4, with γ-H2AX (23, 24). Once targeted, current models posit that the NuA4 complex acetylates H4 and H2A, as well as the histone variant H2AZ, which is incorporated by the remodeling activity of the SWR1 complex. These acetylation and histone exchange activities facilitate the generation of an open chromatin environment, promoting end resection and the recruitment of additional repair factors (23). While the precise roles of many NuA4 and SWR1 components remain to be determined, mutants in key factors, including the catalytic subunits (24–28) and several of the shared proteins [e.g., the YAF9 (26–29), SWC4 (29), and ARP4 (24) chromatin readers], are sensitive to various genotoxic stresses, demonstrating their important functions in the DNA damage response. In plants, NuA4 and SWR1 complexes were identified using affinity purification approaches and shown to possess similar, though not identical, protein subunits as yeast, including orthologs of the catalytic subunits and all four of the shared proteins (30–37). Interestingly, many of these subunits have been duplicated in plants (30), suggesting they could act redundantly or serve specialized functions within their respective complexes. Like their yeast and animal counterparts, these plant complexes catalyze histone acetylation (30, 38, 39) and H2AZ deposition (40–43), respectively, and have demonstrated roles in regulating gene expression (30, 42, 43). Specifically, they play important roles in controlling diverse biological processes including flowering time, reproduction, cell proliferation, and plant responses to the environment (30, 42, 43). However, roles for these complexes in DNA repair are less well understood in plants. For the NuA4 complex, the histone acetyltransferase of the myst family 1 (ham1) and ham2 mutants are sensitive to UV-B damage (44), but roles for these proteins and the NuA4 complex in DSB repair remain unexplored in plants. For the SWR1 complex, mutations in several SWR1-specific subunits, including the chromatin remodeler, photoperiod-independent early flowering 1 (pie1), are sensitive to DSB-inducing agents and show defects in HDR repair assays and in meiosis (45). These findings demonstrate roles for chromatin remodeling complexes in plant DNA repair. However, many aspects of their functions in DSB repair remain to be explored, including how chromatin readers, like YAF9A and YAF9B, contribute to repair. To identify chromatin effectors important for DNA repair in plants, a reverse genetic approach was utilized. While powerful, previous forward genetic screens for repair factors mainly identified Arabidopsis proteins important for NHEJ pathways, which are the preferred pathways utilized in somatic plant tissues (46). This bias has resulted in a large gap in our understanding of how chromatin facilitates DNA repair via HR and HDR pathways in plants. To fill this gap, chromatin effectors coexpressed with genes important for DNA repair by HR, including BRCA1 and RAD51, were identified using a recently published transcriptional network of the Arabidopsis DNA damage response (10). Among the coexpressed genes, YAF9B was identified. In Arabidopsis there are two YAF9 genes: YAF9A, which is constitutively expressed (32, 47), and YAF9B, which is induced by DNA damage (10). Both YAF9A and YAF9B contain YEATS domains that bind acetylated histones (32), confirming their roles as chromatin readers. While these proteins have redundant roles in regulating flowering time and plant development (31, 32), neither YAF9A nor YAF9B have known role(s) in plant DNA repair. In the current work, we demonstrate that the induction of YAF9B expression in response to DNA damage is DSB specific and relies on a cis-regulatory promoter element bound by the SOG1 TF. We also show that both YAF9A and YAF9B are required for DSB repair using several different assays. In a “true-leaf” assay, which is most sensitive to defects in various NHEJ pathways, we found that YAF9A and YAF9B act redundantly to repair DSBs and promote meristem health (i.e., its ability to produce new leaves). However, when we specifically assessed DNA repair via HDR pathways using reporter lines, we found strong defects in yaf9a and yaf9b single mutants, revealing nonredundant roles. Based on these results, we hypothesize that YAF9A and YAF9B act redundantly in early steps of repair that are shared by many pathways (i.e., end resection) but have distinct roles in later steps that are specific to HDR pathways. To better understand how YAF9A and YAF9B function, we identified the protein complexes containing these chromatin readers. Consistent with previous reports (33, 35), YAF9A copurified with components of both the NuA4 and SWR1 chromatin remodeling complexes. In contrast, YAF9B only copurified with components of NuA4, defining a DNA damage-specific version of this complex. These findings not only provide evidence for a NuA4-specific component in mediating DSB repair in plants, but also shows that YAF9A and YAF9B are part of different complexes which, akin to the independent roles of the NuA4 and SWR1 complexes in DSB repair in yeast (23, 48), is in line with their nonredundant roles in HDR pathways. As both the plant NuA4 and SWR1 complexes have roles in transcriptional regulation under normal growth conditions, we assessed the effects of the yaf9 mutants on the transcriptional response to DNA damage but found no defects. This supports our hypothesis that YAF9A and YAF9B play direct roles in DNA repair rather than indirect roles in regulating gene expression. Finally, we demonstrated that mutation of key residues within the reader domain of YAF9B are critical for its role in the DNA damage response. These results establish roles for YAF9A and YAF9B in DSB repair, highlighting the importance of histone readers and expanding our knowledge of how chromatin-associated factors govern DSB repair in plants. Results and Discussion YAF9B Expression Is DSB-Specific, Requires the SOG1 Motif, and Is the Highest in Meristematic Tissues. The YAF9 family of YEATS domain chromatin readers (49–51) were identified as candidates for regulating DSB repair in plants based on the expression pattern of the Arabidopsis YAF9B gene as well as roles for yeast YAF9 in promoting DNA repair and genome stability (52). In Arabidopsis, the transcriptional response to treatment with DSB-inducing gamma-irradiation (γ-IR) was recently characterized over a 24 h time course (10). While YAF9A expression was high even under control conditions (− γ-IR) and was not altered by exposure to gamma-irradiation (+ γ-IR) (Fig. 1A), the expression pattern of YAF9B was similar to many genes important for DSB repair (10): it showed low expression under control conditions but was strongly induced after exposure to γ-IR (Fig. 1A). Furthermore, SOG1, the master transcriptional regulator of the DNA damage response (10, 53, 54), was enriched at its consensus motif within the YAF9B promoter (Fig. 1A) (10), and YAF9B induction in response to DNA damage was lost in the sog1-1 mutant (Fig. 1A) (10). Together these observations suggest YAF9B is directly regulated by SOG1 during the DNA damage response. YAF9B expression is induced after DNA damage in a SOG1-dependent manner. (A) Genome browser tracks showing the expression levels of the YAF9A and YAF9B genes in 6-d-old wt or sog1-1 mutant seedlings at different time points, specified in minutes (’) and/or hours (h), after mock (− γ-IR) or gamma-irradiation (+ γ-IR; 100 Gy) treatments. SOG1 enrichment levels (+ γ-IR; 100 Gy; 20’) from ChIP (IP) and input (IN) samples are shown below, including the SOG1 peak region in the YAF9B promoter, as well as the position and sequence (CTT-N7-AAG) of the SOG1 motif. The YAF9A and YAF9B gene models are blue, the neighboring genes are gray, and the position of the YEATS domains are black. The − γ-IR and + γ-IR mRNA-seq tracks and the IP and IN SOG1 ChIP-seq tracks are colored as indicated (Left), superimposed, and shown on a scale of 0 to 75 reads per 10 million (RP10M). All data are from ref. 10. (B) Images showing the histochemical detection of GUS in 7-d-old seedlings 24 h after mock (− γ-IR) or gamma-irradiation (+ γ-IR; 100 Gy) treatments. (C) Genome browser tracks showing the expression levels of the indicated genes 1h30’ after mock (− γ-IR) or gamma-irradiation (+ γ-IR; 100 Gy) treatments from either undissected 6-d-old seedlings or the indicated dissected tissues. The − γ-IR and + γ-IR mRNA-seq tracks are colored as indicated (Left), superimposed, and shown in RP10M based on the indicated scales (Upper Left). (D–G) Images showing the histochemical detection of GUS in 7-d-old seedlings 24 h after mock (− γ-IR) or gamma-irradiation (+ γ-IR; 100 Gy) treatments. For each group of images in B and D–G, the transgenes [insert (ins) numbers represent independent transgenic lines] and genetic backgrounds for the seedlings shown are as labeled. In panels F and G, the mutations introduced into the SOG1 motif in YAF9B promoter are indicated below the motif. (H) Images showing the histochemical detection of GUS in 7-d-old seedlings 24 h after no treatment, exposure to gamma-irradiation (+ γ-IR; 100 Gy), bleomycin (20 µg/mL), cisplatin (50 µM), ultraviolet light (UV-C radiation, 6,000 J/m2), hydroxyurea (80 mM), salt (NaCl, 400 mM), cold (4 °C), or heat (37 °C) treatments. For each group of images, the transgenes (including the insert number) and genetic backgrounds for the seedlings shown are as labeled. To further support the connections between YAF9B expression, the DNA damage response, and the SOG1 transcription factor, β-glucuronidase (GUS) reporter constructs were generated. These reporters included either the endogenous YAF9A or YAF9B promoter and their respective 3’UTRs flanking the GUS gene. Consistent with previous Reverse Transcriptase PCR (RT-PCR) data showing widespread expression of YAF9A under nonstressed conditions (32, 47), several independent pYAF9A::GUS::3’UTR reporter lines [insertions 1, 2, and 3 (ins1-3)] showed strong expression irrespective of γ-IR treatment across all tissues in both 7-d-old seedlings (ins1; Fig. 1B and ins2-3; SI Appendix, Fig. S1A) and 14-d-old seedlings (ins1-3; SI Appendix, Fig. S1A). In contrast, the pYAF9B::GUS::3’UTR reporter lines showed expression only after γ-IR treatment, with the strongest GUS expression detected in the shoot apex and root tips of 7-d-old seedlings (ins1; Fig. 1B and ins2; SI Appendix, Fig. S1B) and 14-d-old seedlings (ins1-2; SI Appendix, Fig. S1B). Notably, the tissue specific GUS expression patterns observed for both the pYAF9::GUS::3’UTR reporter lines were consistent with the in vivo expression patterns for YAF9A and YAF9B, which were assessed directly via mRNA-seq experiments using dissected tissues from similarly staged wild-type (wt) (Col-0) seedlings either with or without exposure to γ-IR (Fig. 1C). While YAF9A was uniformly expressed across tissues and conditions, YAF9B expression was induced by γ-IR and was the highest in samples containing meristematic tissues (i.e., the shoot apex and root+ hypocotyl; Fig. 1C). These expression analyses confirm the broad, constitutive expression pattern of YAF9A and demonstrate that in addition to being induced by DNA damage, YAF9B expression is tissue-specific, mimicking both the expression pattern of the constitutively expressed SOG1 gene (55) and the damage responsiveness of several other DNA repair genes (56–59) (Fig. 1C). To assess the role of the SOG1 cis-regulatory motif in controlling the tissue-specific expression of YAF9B, the pYAF9B::GUS::3’UTR reporter was first crossed into two different sog1 mutant backgrounds [sog1-1 (54) and sog1-101 (59)] to confirm that the reporter recapitulates the SOG1-dependent expression pattern observed at the endogenous YAF9B locus (Fig. 1A). In both cases, the pYAF9B::GUS::3’UTR reporter was induced in meristematic tissues after γ-IR in the wt cosegregants and this expression was lost in the sog1 mutants (Fig. 1 D and E). Next, the role of the SOG1 cis-regulatory motif was assessed using mutated versions of the pYAF9B::GUS::3’UTR reporter where conserved bases within the SOG1 motif were swapped (purines for pyrimidines or vice versa) (Fig. 1F) or the spacing within the SOG1 motif was altered by deleting 3 bases (Fig. 1G). For each construct, several independent lines were tested, and in all cases mutations within the SOG1 motif resulted in strongly reduced GUS expression in both 7-d-old seedlings (ins1; Fig. 1 F and G and ins2-3; SI Appendix, Fig. S1 C and D) and 14-d-old seedlings (ins1-3; SI Appendix, Fig. S1 C and D) as compared to the pYAF9B::GUS::3’UTR reporter with the unaltered SOG1 motif (Fig. 1B and SI Appendix, Fig. S1B). These findings further support a direct regulation of YAF9B by SOG1 by demonstrating that the SOG1 cis-regulatory motif within the YAF9B promoters is required for GUS induction in meristematic tissues after DNA damage. To assess the responsiveness of the YAF9A and YAF9B promoters to different types of DNA damage, the pYAF9::GUS::3’UTR reporter lines, along with wt lines as controls, were exposed to a variety of genotoxic conditions. For the YAF9A reporter, the GUS staining patterns were similar under all conditions (Fig. 1H), as was the expression of the endogenous YAF9A gene from wt plants (SI Appendix, Fig. S1E), demonstrating that YAF9A expression is not significantly altered in response to any of these genotoxic stresses. For the YAF9B reporter lines, the GUS staining was specific to the shoot apex and root tips and was most prominent in response to DSB-inducing agents [γ-IR and bleomycin (57, 60)] (Fig. 1H), with little to no expression observed in response to other stresses [i.e., UV-C radiation for crosslinking (60–63), cisplatin and hydroxyurea for replication stress (57, 64) and cold, salt, and heat for environmental stress (57, 65, 66)] (Fig. 1H). Consistent with these results, expression of endogenous YAF9B was also the most strongly induced in response to DSB-inducing agents in wt plants (SI Appendix, Fig. S1E). Although these stresses likely result in different levels of DNA damage, the elicitation of a stress response under our conditions was verified by assessing the induction of previously characterized stress-responsive genes (10, 62, 67–72) (SI Appendix, Fig. S1F). Taken together, these findings highlight the distinct expression patterns, cis-regulatory elements, and environmental responsiveness of the YAF9A and YAF9B genes. yaf9 Mutants Are Sensitive to Genotoxic Stress. To investigate the roles of YAF9A and YAF9B in DNA repair, several yaf9 single and double mutants were characterized. Consistent with previous RT-PCR assays (32, 47), mRNA-seq experiments (with mock or γ-IR treatments) confirmed that the expression of YAF9A and YAF9B are clearly disrupted in the yaf9a-1 and yaf9b-2 mutants, respectively (SI Appendix, Fig. S2A). In the yaf9a-1 mutant, the transcript is truncated, and in the yaf9b-2 mutant, the transcript is truncated and the expression is no longer responsive to γ-IR treatment (SI Appendix, Fig. S2A). In contrast to these null mutants, the yaf9b-3 mutant is weaker and condition specific, which is consistent with the position of the T-DNA in the YAF9B promoter between the SOG1 motif and the transcription start site (SI Appendix, Fig. S2A). In a wt background YAF9B expression is in fact strongly reduced in the yaf9b-3 mutant after γ-IR treatment (SI Appendix, Fig. S2A), suggesting that the proximity of the SOG1 motif to the YAF9B gene is important for YAF9B upregulation in response to DNA damage. However, the yaf9a-1,b-3 double mutant (SI Appendix, Fig. S2B) did not display the same developmental defects (e.g., reduced plant size, infertility, etc.) previously observed in the yaf9a-1,b-2 double mutant (31, 32) or another yaf9a,b mutant combination (73), suggesting that the yaf9b-3 mutant is weaker than the other yaf9b mutants. Indeed, in the yaf9a-1,b-3 background, increased expression of the yaf9b-3 allele was observed under both mock and γ-IR conditions (SI Appendix, Fig. S2 A and C). PCR amplification (SI Appendix, Fig. S2C) and sequencing (SI Appendix, Fig. S2D) of the YAF9B transcript in the yaf9b-3 single and yaf9a-1,b-3 double mutants confirmed the presence of a full length, wt transcript, demonstrating that yaf9b-3 is a knock down rather than a loss of function mutant, explaining its weaker developmental phenotype (SI Appendix, Fig. S2B). Taken together, these yaf9 alleles provide a series of mutants with variable strengths and expression patterns that can be leveraged to assess the roles of YAF9A and YAF9B in DNA repair. Using the aforementioned yaf9 mutants, defects in DNA repair were first assessed using the “true leaf” assay. In this assay, the accumulation of DNA damage is quantified by counting the number of deformed or missing true leaves that arise in young seedlings exposed to genotoxic stresses due to cell cycle arrest and/or cell death in the apical meristem (45, 74). Compared to both wt and no treatment controls, a higher proportion of the yaf9a-1,b-2 double mutant plants showed true leaf defects after exposure to 100 Gy of γ-IR across three independent experiments (Fig. 2A and SI Appendix, Fig. S2E) and these defects arose in a dose-dependent manner (SI Appendix, Fig. S2F). The other single and double yaf9 mutants behaved similar to the wt control (Fig. 2A and SI Appendix, Fig. S2 E and F), demonstrating redundancy between YAF9A and YAF9B and revealing that even the low level of YAF9B expression in the yaf9a-1,b-3 double is sufficient to promote plant tolerance to genotoxic stress. To confirm that the damage sensitivity observed in the yaf9a-1,b-2 mutant is due to disruption of the YAF9A and YAF9B genes, complementation assays were conducted using transgenes containing the YAF9A or YAF9B genes driven by their endogenous promoters and tagged with either 3XFLAG- or GFP. To avoid the fertility defects observed in the homozygous yaf9a-1,b-2 double mutant, these transgenes were introduced into either yaf9a-1(ho),b-2(het) or yaf9a-1(het),b-2(ho) lines, respectively, as diagrammed in SI Appendix, Fig. S2G. After selecting for homozygous YAF9A and YAF9B transgenic lines, and genotyping for the yaf9a-1 and yaf9b-2 alleles, true leaf assays were conducted. For these assays, six independent YAF9A and YAF9B tagged lines [three with 3XFLAG (ins1-3) and three with GFP (ins1-3)] were utilized (ins1; SI Appendix, Fig. S2H and ins2-3; SI Appendix, Fig. S2 I and J). As is evidenced by the reductions in seedlings with deformed or missing true leaves after γ-IR exposure, these assays demonstrated that both the 3XFLAG- and GFP-tagged YAF9A and YAF9B proteins were able to largely rescue the DNA damage defects observed in the yaf9a-1,b-2 mutant (SI Appendix, Fig. S2 H–J), confirming that YAF9A and YAF9B act redundantly to promote plant tolerance to genotoxic stress. YAF9A and YAF9B are required for meristem health and homology-directed repair after DSB damage. (A) True leaf assay showing the percentage of 11-d-old seedlings categorized as having proper true leaves (green), deformed true leaves (orange), or one/none true leaves (red) after either no treatment or treatment with 100 Gy. Images of representative seedlings for each category are shown above the bar plots. For each bar plot, the genotypes are labeled on the x-axis. On the y-axis, the color-coded dots represent the percentage of seedlings belonging to each category calculated based on sets of ~30 seedlings that represent biological replicates (n = 3 or 4). The stacked bars are the average of all the data points for each category. (B and C) Upper: Transgenic DSB reporter constructs (DGU.US/SSA and IU.GUS/SDSA) (75, 76) in which a GUS gene is interrupted by a unique sequence including the 18-bp I-SceI recognition site (red bar). DSBs can be introduced into theses GUS reporters via 2 methods: They can be induced precisely at the unique I-SceI sequence by introduction of the I-SceI enzyme on a separate transgene or they can be introduced at a lower rate by exposure to genotoxic agents, like γ-IR (lightning bolt), that introduce DSBs randomly throughout the genome. In either case, repair of the DSB via HDR mechanisms generates a functional GUS gene that can be detected as blue spots after histochemical staining. Lower: Representative images showing the recombination events via SSA/DGU.US and SDSA/IU.GUS in 15-d-old F1 seedlings homozygous for the genotypes indicated below each image and heterozygous for the reporter and inducer constructs indicated above. Detailed crossing schemes are shown in SI Appendix, Fig. S3A. Multiple crosses were performed and the cross number is indicated on the top-left corner of the image, with images from additional crosses shown in SI Appendix, Fig. S3 B and C. (D and E) Violin plots showing the recombination events via SSA/DGU.US and SDSA/IU.GUS after 100 Gy γ-IR in seedlings with the genotypes indicated below and the reporters indicated above each graph. For the DGU.US and IU.GUS reporters recombination events were scored in 14-d-old or 19-d-old seedlings, respectively. In D, the data points for F3 seedlings from different lineages (i.e., different F2 parents of the indicated genotypes; see SI Appendix, Fig. S3A for crossing details) are represented by the different color-coded shapes. In E, the data points for F1 seedlings from different lineages (i.e., the progeny of different genetic crosses; see SI Appendix, Fig. S3D for crossing details) are represented by the different color-coded shapes. For each epitope-tagged YAF9 construct, two independent inserts (ins) are shown. For D and E, the asterisks (*) represent the P-value from Wilcoxon tests comparing each genotype to the wt control [not significant (ns): P > 0.05; *P <= 0.05; **P <= 0.01; ***P <= 0.001; ****P <= 0.0001]. YAF9A and YAF9B Have Nonredundant Roles in DSB Repair via Homology-Directed Repair Pathways. Given the coexpression of YAF9B with genes required for HR (10), two previously described HDR reporters (75, 76) were utilized to directly assess the roles of YAF9A and YAF9B in DSB repair. Both these reporters (Fig. 2 B and C) contain a GUS gene that is nonfunctional until a DSB is generated at a specific location (red tick mark) and then repaired either by single strand annealing (SSA) at the DGU.US reporter or by synthesis-dependent strand annealing (SDSA) at the IU.GUS reporter. To achieve high rates of HDR at these reporters, an inducer construct that expresses the I-SceI restriction enzyme and generates DSBs in the GUS reporter at the desired location is required. However, lower rates of repair are also observed after exposure to random mutagenesis, including γ-IR exposure. In all scenarios, restoration of the GUS gene through HDR allows rates of SSA and SDSA to be determined by counting the number of blue sectors in plants harboring these reporters. To test the effects of yaf9 mutants on DSB repair using the aforementioned reporters, the required genetic materials were generated via the crossing schemes outlined in SI Appendix, Fig. S3A. Using the inducer system, the number of blue sectors resulting from repair of either the DGU.US/SSA or IU.GUS/SDSA reporters were visibly lower in the yaf9a-1 and yaf9b-2 seedlings when compared to seedlings from the wt control crosses (Fig. 2 B and C). Furthermore, this phenotype was consistent across several independent crosses (SI Appendix, Fig. S3 B and C). To assess the effects in additional mutant backgrounds, these assays were also conducted in the absence of the inducer, with damage generated by exposure to γ-IR (see SI Appendix, Fig. S3A for crossing scheme). In these assays, the number of blue spots resulting from HDR were significantly reduced in all the yaf9 mutants compared to the wt controls in both reporter backgrounds (Wilcoxon tests ****P ≤ 0.0001; Fig. 2D). In line with yaf9b-3 being a knockdown rather than a null allele, this mutant displayed a weaker phenotype than yaf9b-2 (Fig. 2D). By comparison, the recombination rates in the yaf9a-1 mutant were lower than both yaf9b single mutants (Fig. 2D), revealing a stronger repair defect in this mutant. However, the repair defects in the yaf9a-1,b-2 double mutant were only slightly enhanced compared to the yaf9a-1 single mutant Fig. 2D). These findings demonstrate that, unlike for the true leaf assay (Fig. 2A and SI Appendix, Fig. S2 E and F), YAF9A and YAF9B cannot fully compensate for each other and instead act in a largely nonredundant manner to promote HDR at DSBs. To confirm that the DSB repair defects observed in the yaf9 mutants are due to reduced recombination rates, rather than, for example, epigenetic silencing of the GUS reporter lines (76, 77), complementation assays were conducted. For these assays, the DGU.US/SSA and IU.GUS/SDSA reporter lines, in either the wt, yaf9a-1, or yaf9b-2 backgrounds used in Fig. 2D, were crossed with the previously generated YAF9 complementing lines (SI Appendix, Fig. S2G), and recombination rates were assessed in the resulting F1 seedlings after exposure to γ-IR Fig. 2E, see SI Appendix, Fig. S3D for crossing scheme). Compared to the uncomplemented yaf9a-1 and yaf9b-2 controls, recombination rates were significantly increased in the presence of multiple, independent tagged YAF9 proteins Fig. 2E), demonstrating a partial rescue of the recombination defects at both HDR reporters. This complementation affirms that the GUS reporters were not epigenetically silenced and remain competent for recombination and GUS detection in these yaf9 mutant backgrounds. Although confounding roles for the yaf9 mutants in promoting full gene expression at these reporters cannot be formally excluded, these findings suggest direct roles for YAF9A and YAF9B in DSB repair via HDR mechanisms. YAF9A and YAF9B Facilitate DSB Repair in a Chromatin-Mediated but Transcription-Independent Manner. The DNA damage response involves the regulation of thousands of genes, many of which are critical for DSB repair itself or for other repair-associated biological processes like cell cycle regulation (Fig. 3A) (10, 66, 67, 78, 79). To determine whether the roles for YAF9A and/or YAF9B in DNA repair are due to changes in gene regulation upon exposure to DNA damage, three replicate mRNA-seq experiments [batch (b) 1-3] were performed using wt or yaf9 mutant seedlings 3 h after either mock (− γ-IR) or γ-irradiation treatments (+ γ-IR; 100 Gy) (Dataset S1). As an initial control, differentially expressed genes (DEG; |Fold change (FC)| ≥2 and P-value ≤0.01) were identified using just the wt seedling data +/− γ-IR treatments ([wt (+ γ-IR)]b1-3 vs. [wt (− γ-IR)]b1-3; Fig. 3B and Dataset S2). These analyses recapitulated the previously published transcriptional changes (10), with similar patterns for the up- and downregulated genes at the 3 h time point (i.e., path 1 genes are the most upregulated in response to γ-IR, followed by path 2, etc.). Furthermore, these control samples were robustly separated by condition (+/− γ-IR) in a principal component analysis (PCA) across all three experiments (b1-3; Fig. 3C). In this PCA, a clear separation between the +/− γ-IR samples was also observed for all the yaf9 single and double mutants (Fig. 3C), suggesting the DNA damage response remains largely intact. Furthermore, with the exception of the yaf9a-1,b-2 mutant, which displays strong developmental defects even under normal conditions (31, 32), all the other yaf9 mutants clustered together with their respective wild-type controls (Fig. 3C). Together, these analyses demonstrate wild-type-like expression profiles under both mock and DNA damage conditions for these yaf9 mutants. YAF9A and YAF9B are present in distinct chromatin remodeling complexes and facilitate DNA repair via mechanisms that are independent from roles in transcriptional regulation. (A) Reproduction of the DREM model from ref. 10, which is licensed under CC BY-NC-ND 4.0, summarizing the transcriptional response of Arabidopsis seedlings exposed to DNA damage. The y-axis shows the log2 FC in expression of genes after exposure to γ-irradiation (+ γ-IR; 100 Gy) or mock (− γ-IR) treatments. The genes are grouped together into 11 differently colored paths (1-11) based on their expression patterns across a time course (x-axis) where the time is shown in minutes (’) and hours (h). The major gene ontology terms for select paths (indicated by their corresponding colors) are shown below. The 3 h time point, which was selected for analysis of the roles for the YAF9 proteins in the DNA damage response, is highlighted in gray. (B) Volcano plot showing the changes in gene expression observed in wild-type seedlings 3 h after exposure to DNA damage. The x- and y-axes show the log2 FC and −log10 P values for DEGs identified by comparing the genotypes and conditions indicated above the plot across three independent mRNA-seq experiments [batches (b)1-3]. DEGs present in the DREM model shown in A are color coded based on the gene groups in which they reside. DEGs not present in the DREM model are shown in gray. The dotted lines represent |log2 FC| ≥1 and P-value ≤ 0.01. (C) PCA showing data points from all 3 batches mRNA-seq experiments. The different experiments (batch 1-3) are shown in different shapes, the different genotypes are shown in different colors, and the different conditions (3 h after 100 Gy; + γ-IR or 3 h after mock; − γ-IR) are shown as filled or open shapes, respectively. For each genotype and condition, two technical replicates were included. (D) Cartoons based on Bieluszewski et al. (80), which is licensed under CC BY 4.0, summarizing the current understanding of the Arabidopsis SWR1 (35–37) and NuA4 (33) complexes, color-coded and labeled with subcomponents. Filled subunits were enriched via YAF9A (Upper) or YAF9B (Lower) IP-MS experiments. (E) Volcano plots showing the enrichment or depletion of proteins from a combination of several independent YAF9A or YAF9B IP-MS experiments compared to nontransgenic, wild-type IP-MS experiments. The red hashed lines demarcate a |log2 FC| ≥ 1 and a P-value ≤ 0.05. Enriched proteins previously identified as components of the SWR1 and NuA4 remodeling complexes are labeled and colored based on the legend below. The global observations from the PCA were also confirmed via DEG analyses using the yaf9 mutants (|FC| ≥2 and P-value ≤ 0.01; Dataset S2): Under mock conditions [mutant (− γ-IR) vs. wt (− γ-IR); SI Appendix, Fig. S4A], very few genes were misregulated in the yaf9 single or yaf9a-1,b-3 double mutants, none of which play known roles in DNA repair (Dataset S2). Consistent with the PCA, many more genes were misregulated in the yaf9a-1,b-2 double mutant (SI Appendix, Fig. S4A; 61 down, 513 up DEGs and Dataset S2). However, after DNA damage [mutant (+ γ-IR) vs. mutant (− γ-IR); SI Appendix, Fig. S4B], all the yaf9 single and double mutants showed a transcriptional response similar to the wt control (SI Appendix, Fig. S4B vs. Fig. 3B). In fact, these responses were so similar that comparisons between the mutant and wild-type γ-irradiation datasets [mutant (+ γ-IR) vs. wt (+ γ-IR); SI Appendix, Fig. S4C] revealed very few misexpressed genes. Among all the yaf9 mutants, only the yaf9a-1,b-2 double showed a significant difference in its DNA damage response (SI Appendix, Fig. S4C; 132 down, 1,131 up DEG), but this difference corresponds to a slightly more, rather than less, robust response when compared to the wild-type control. This is evident by the larger scale of expression changes observed for DNA damage response genes in the yaf9a-1,b-2 mutant in both the volcano plot shown in SI Appendix, Fig. S4C and in the heatmaps shown in SI Appendix, Fig. S4D [i.e., some genes normally upregulated after DNA damage (paths 1-7) are more upregulated in the yaf9a-1,b-2 mutant, and some genes normally downregulated (paths 8-11) are more downregulated in the yaf9a-1,b-2 mutant, but the converse situation (genes normally up, that are now down, etc.) is not observed]. Taken together, these analyses demonstrate that the transcriptional response to DNA damage is not negatively impacted in either the yaf9 single or double mutants. Furthermore, since the yaf9a and yaf9b single mutants have strong defects in DSB repair (Fig. 2) but show little to no changes in gene expression under normal conditions or after exposure to DNA damage (SI Appendix, Fig. S4), these data support direct roles for YAF9A and YAF9B in DNA repair, akin to roles proposed for yeast YAF9 (52), rather than indirect roles in regulating the transcriptional response to DNA damage. To determine whether YAF9B, like YAF9A (33, 35), interacts with NuA4 and SWR1 chromatin-modifying complexes (Fig. 3D), which could facilitate their roles in DNA repair (45), immunoprecipitation and mass spectrometry (IP-MS) experiments were performed. These experiments utilized 8-d-old seedlings from nontransgenic wild-type controls and representative 3XFLAG-tagged YAF9A or YAF9B complementation lines, in their respective single mutant backgrounds (pYAF9A:YAF9A:3XFLAG/yaf9a-1 (ins1) and pYAF9B:YAF9B:3XFLAG/yaf9b-2 (ins1); SI Appendix, Fig. S2G and Fig. 2E). For YAF9A one IP-MS experiment was conducted under mock conditions and another 6 h after exposure to 100 Gy of γ-IR. For YAF9B, two IP-MS experiments were also conducted, both 6 h after exposure to 100 Gy of γ-IR. As the replicate experiments for YAF9A and YAF9B gave similar results (Dataset S3), they were analyzed together to identify enriched proteins copurifying with either YAF9A or YAF9B, respectively (Fig. 3 D and E). Consistent with previous reports, YAF9A copurified with nearly all the previously identified NuA4 and SWR1 components (Fig. 3 D and E and Dataset S3). However, the YAF9B purification suggests a preferential, if not specific, interaction with the NuA4 complex. When considering the most significant copurifying factors (|log2 FC > 1| and adjusted P-value <= 0.05), as well as those that are highly enriched (log2 FC > 3) despite less significant adjusted P-values, YAF9B only interacts with shared NuA4/SWR1 proteins and NuA4-specific proteins, but no SWR1-specific proteins (Fig. 3 D and E and Dataset S3). These include SWC4 and ARP4 from the shared targeting module, but notably not YAF9A. Indeed, even in the YAF9A experiment conducted after exposure to γ-IR, YAF9B was not detected (Dataset S3). These findings suggest that each complex only contains a single YAF9 subunit, which is consistent with previous results showing both YAF9A and YAF9B interact with the same subunit, SWC4, by Y2H (31). In addition to these shared components, YAF9B also copurified with another protein previously determined by Y2H, EAF1B (31). This NuA4-specific protein serves as a scaffold for the shared NuA4/SWR1 targeting module as well as the NuA4-specific piccolo catalytic complex, for which EPL1B, HAM1, and ING2 were all enriched in the YAF9B purification. In contrast, the SWR1-specific proteins showed little to no enrichment. These interaction data demonstrate that YAF9B forms a distinct, DNA damage-specific NuA4 complex. Taken together with the absence of transcriptional defects in yaf9a and yaf9b mutants (Fig. 3C and SI Appendix, Fig. S4), the roles for YAF9A and YAF9B in DNA repair (Figs. 1 and 2 and SI Appendix, Figs. S2 and S3), and past roles demonstrated for other SWR1-complex components in DSB repair (45), these findings link YAF9A and YAF9B, and their respective chromatin remodeling complexes, to DSB repair in a manner that is chromatin-mediated but transcription-independent. Conserved Residues within the YEATS Domain Are Required for YAF9B Function. To investigate whether the acetylation reader function of the YAF9B YEATS domain (32) is important in mediating its role in DNA repair, a mutagenesis approach was employed. The Arabidopsis YAF9A and YAF9B proteins share homology with human GAS41 and yeast Yaf9 (30, 52). The residues required for the YEATS domain to recognize lysine acetylation in GAS41 have been identified and shown to form an aromatic cage (81, 82) (SI Appendix, Fig. S5A; green triangles). Based on an alignment with GAS41 (SI Appendix, Fig. S5A) and a structural prediction of YAF9B (83, 84) (Fig. 4 A and B), amino acids H89, S91, F92, W111, G112, and F114 were identified as candidate residues important for the YAF9B YEATS domain–histone interaction. To assess the roles of these amino acids in DNA repair, alanine substitutions of the aforementioned residues, either alone or in combinations of up to four residues, were introduced into both the 3XFLAG- and GFP-tagged YAF9B constructs and transformed into yaf9a-1(het),b-2(ho) plants. Proper protein expression and accumulation of the wild-type or mutated 3XFLAG-tagged YAF9B proteins were verified using western blots (SI Appendix, Fig. S5 B and C). True leaf assays were then performed in yaf9a-1,b-2 double mutants homozygous for either the wild-type or the mutated YAF9B transgenes (Fig. 4C). Consistent with previous assays (Fig. 2A and SI Appendix, Fig. S2 E and F), the sensitivity in yaf9 single mutants after γ-IR was comparable to the wild-type control, the yaf9a-1,b-2 double mutant was more sensitive, and addition of functional YAF9B rescued the true leaf defects (Fig. 4C). Among the single amino acid substitutions, the S91A, F92A, G112A, and F114A single mutants all complemented the yaf9a-1,b-2 double mutant to a similar level as observed for the wild-type version of YAF9B (Fig. 4C), demonstrating that alone, these mutations are not sufficient to disrupt YAF9B function in vivo. In contrast, the YAF9B constructs harboring the W111A mutation all showed lower levels of complementation, with the defects in the W111A, G112A, F114A triple and the H89, W111A, G112A, F114A quadruple mutant combinations nearly equivalent to the yaf9a-1,b-2 double mutant (Fig. 4C). To directly test the roles of the YEATS domain in HDR, the wildtype YAF9B-3XFLAG transgene, or transgenes harboring the triple (HWF) or quadruple (HWGF) mutations, were introduced into a background homozygous for the IU.GUS reporter and the yaf9b mutant by crossing. As previously observed Fig. 2E), the rates of HDR at the reporter were reduced in the yaf9b mutant and were partially rescued in lines harboring a wild-type YAF9B-3XFLAG transgene Fig. 4D). However, lines harboring point mutations within the YAF9B YEATS domain failed to rescue the defects observed in the yaf9b mutant (Fig. 4D). These findings demonstrate an essential role for the YAF9B YEATS domain in DNA repair. Notably, the sensitivities observed in the YAF9B substitutions are in agreement with the previously identified importance of their counterparts in GAS41 (85). Moreover, the importance of the Tryptophan (W) in the aromatic cage for YEATS domain function in both Arabidopsis YAF9B and Human GAS41 (85), suggests a conservation of the histone-binding mechanism. Altogether, these data demonstrate that a functional YEATS domain is required for the role of YAF9B in DNA repair. The YEATS reader domain is required for YAF9B function. (A) AlphaFold predicted 3D-structure of YAF9B (83, 84) where the YEATS domain, A-Box, and B-Box are colored as in (SI Appendix, Fig. S5A). The residues predicted to interact with histone acetyl marks are shown in green (ball and stick). These residues are labeled with the amino acid symbol and number in a zoomed in view depicted in (B). (C) True leaf assays showing the rescue capacity of YAF9B with key acyl mark interactors mutated using 3XFLAG- (Left) and GFP-tagged lines (Right). The plot shows the percentage of seedlings with proper true leaves (green), deformed true leaves (orange), and one/none true leaves (red) in 12-d-old seedlings after treatment with 100 Gy in yaf9 single, double mutants, and rescue lines. Each rescue line is represented by 2 to 3 biological replicates (independent insertions, represented by different point shapes) and 2 to 3 technical replicates (seeds from the same seed packet grown on different plates and randomized positions in the growth chamber). Each color-coded dot represents the percentage of seedlings belonging to one of the three true leaf phenotype categories calculated based on ~30 seedlings. Each stacked bar is the mean of 3 to 4 data points represented by dots. (D) Violin plots showing recombination events measured using the SDSA/IU.GUS reporter following 125 Gy of X-ray irradiation in 21-d-old seedlings of the genotypes indicated below. All three YAF9B transgenes correspond to insertion line 1 (ins1) and data points from two independent experiments are represented by circles and triangles, respectively. The asterisks (*) indicate significance levels from pairwise Wilcoxon tests. [not significant (n.s.): P > 0.05; *P <= 0.05; **P <= 0.01; ***P <= 0.001; ****P <= 0.0001]. Conclusion In this work, we expand the roles of YAF9A and YAF9B to include not just plant development and reproduction (31, 32, 47) but also DSB repair. First, we determined the DNA damage- and tissue-specific regulation of YAF9B by the SOG1 transcription factor via a cis element present in the YAF9B promoter (Fig. 1 and SI Appendix, Fig. S1). Second, we demonstrated roles for YAF9A and YAF9B in promoting plant meristem health in response to genotoxic stress (Fig. 2 and SI Appendix, Fig. S2) and facilitating DSB repair via HDR mechanisms (Fig. 2 and SI Appendix, Fig. S3). Third, we found that YAF9A and YAF9B denote distinct chromatin remodeling complexes (Fig. 3 D and E and Dataset S3), with YAF9B defining a DNA damage-specific NuA4 complex while YAF9A, like its yeast counterpart, associates with both the SWR1 and NuA4 complexes (23, 30). This finding represents the first link between the NuA4 complex and DSB repair in plants and suggests a plant-specific innovation to cope with DNA damage—the formation of a specialized NuA4 complex that is enriched in meristematic tissues. Finally, we connected the function of YAF9B in DNA repair to its YEATS reader domain (Fig. 4 and SI Appendix, Fig. S5). Together, these findings expand our understanding of the chromatin effectors that regulate DSB repair in plants and reveal both conserved and plant-specific aspects of YAF9 functions. Mechanistically, our studies revealed that the yaf9a and yaf9b mutants have no effect on the transcriptional response to DNA damage (Fig. 3 and SI Appendix, Fig. S4) but show defects in DNA repair assays (Fig. 2 and SI Appendix, Figs. S2 and S3), supporting direct roles in DNA repair rather than indirect roles through gene regulation. While the precise roles of YAF9A and YAF9B during DSB repair require further investigation, their associations with the SWR1 and NuA4 complexes (Fig. 3 D and E and Dataset S3), and the behaviors of the yaf9a and yaf9b mutants in different types of repair assays, offer clues to their functions. For example, in the true leaf assay, which scores many different types of repair (e.g., both NHEJ, HR, and HDR pathways), YAF9A and YAF9B act in a largely redundant manner as only the yaf9a,b double mutants show strong defects upon exposure to DNA damage (Fig. 2 and SI Appendix, Fig. S2). However, in the GUS reporter assays, which specifically assess repair via HDR pathways (86), YAF9A and YAF9B are largely nonredundant as both the yaf9a and yaf9b single mutants show reduced rates of repair (Fig. 2 and SI Appendix, Fig. S3). While these results are seemingly contradictory, our understanding of the SWR1 and NuA4 chromatin remodeling complexes associated with these YAF9 proteins suggests a plausible explanation. The activities of the SWR1 and NuA4 complexes are interconnected and associated with the promotion of open chromatin states in yeast, mammals, and plants (23, 30, 42). The NuA4 complex catalyzes the acetylation of lysine residues in the N-terminal tails of H4, H2A, and H2A.Z (23, 30), which neutralizes their positive charge and thus weakens their interactions with the negatively charged DNA backbone (87). Likewise, the SWR1 complex facilitates the incorporation of the H2A.Z variant (23, 42, 43), which confers lower nucleosome stability in some conditions (88). Within the context of DNA repair, the formation of open chromatin promotes end resection, an early DNA processing step that is important for noncanonical NHEJ, HR, and HDR pathways (9, 86, 89). Based on these functions, we hypothesize that the roles of Arabidopsis YAF9A and YAF9B are redundant for early DNA repair steps, like end resection, that are shared between noncanonical NHEJ, HR, and HDR pathways. For the latter steps in repair, the yeast NuA4 and SWR1 complexes (or the combined mammalian TIP60 complex) also play important roles, as the acetylation of histone tails and the incorporation of the H2A.Z variant facilitate the recruitment of additional factors to DSBs to promote HDR (89–91). Since loss of either YAF9A or YAF9B in plants results in strong repair defects, we hypothesize these YAF9 proteins, and their associated SWR1 and NuA4 complexes, are either required at high levels, such that loss of either YAF9A or YAF9B results in a haploinsufficiency-like phenotype, or they play nonredundant roles in these later steps of HDR (Fig. 2 and SI Appendix, Fig. S3). In the latter scenario, our finding that both YAF9A and YAF9B associate with NuA4 complexes (Fig. 3 D and E and Dataset S3), yet the yaf9b single mutant shows strong defects in HDR (Fig. 2 and SI Appendix, Fig. S3), would support a neofunctionalization model in which the YAF9A-NuA4 and YAF9B-NuA4 complexes have distinct roles in DNA repair. To further support these hypotheses, it will be important to determine what specific roles each of these YAF9 proteins/complexes play in early and late stages of DSB repair and how they regulate chromatin to promote HDR mechanisms. As an initial step toward understanding the functions of the Arabidopsis YAF9 proteins at chromatin, we investigated the importance of the YAF9B YEATS domain for DNA repair. YEATS domains are histone binding domains initially characterized as readers of lysine acetylation. However, more recent work in yeast and mammals has shown YEATS domains can also bind, and even prefer, other types of lysine acylations, including propionylation, butyrylation, and crotonylation, that fit within the same binding cleft (50, 51). In Arabidopsis, both YAF9A and YAF9B were found to bind acetylated histone tails (32). However, their preferences for binding acetylation at different lysine positions or within different histone tails (H3, H4, H2AZ) remains limited and their potential to bind other types of acylations remains completely unknown. Despite many questions remaining about the specificity of the Arabidopsis YAF9 YEATS domains, mutation of highly conserved aromatic residues predicted to form π-stacking interactions with acyl-lysines residues revealed that the YAF9B YEATS domain is critical for its role in DNA repair. This finding marks a major advance in our understanding of how chromatin modifications coordinate DNA repair in plants and represents one of the few direct links between DNA repair, in this case DSB repair, and a chromatin reader function (i.e., the YAF9B acyl-lysine activity). Although the importance of their reader domains in repair has not been explicitly tested, other notable examples of such connections include PDS5C and MSH6, which have Tudor binding specificities (84, 92) and repair defects (93, 94) that strongly support direct roles in mismatch and homology-dependent repair at genes marked with H3K4me1, respectively. Likewise, BCP4 and XIP, which have BRCA1 C-terminal (BRCT) domains shown to bind γ-H2A.X (95, 96), play important roles in DSB repair as revealed by true leaf and HDR assays, respectively (95, 96). This growing list of connections between chromatin readers and DNA repair pathways/steps represent initial, but critical advances toward understanding how chromatin modifications coordinate repair to ensure genome stability. In the case of YAF9B, the importance of its YEATS domain for DNA repair suggests a role either in the targeting or retention of the NuA4 complex at chromatin in response to DNA damage. To fulfill this role, several nonmutually exclusive scenarios can be envisioned. For example, the YEATS domain of YAF9B could be required to recruit the NuA4 complex to chromatin surrounding DSBs with preexisting acylated histones. Alternatively, NuA4 recruitment could instead depend on ARP4-mediated recognition of phosphorylated H2AX [a mark that rapidly accumulates at DSBs via the kinase activities of ATM and ATR (2, 14)], as previously demonstrated in yeast (24). Once recruited the NuA4 complex could then acetylate the surrounding chromatin, generating binding sites for the YAF9B YEATS domain and facilitating the stable accumulation and/or spreading of NuA4 at DSBs. Or, a combination of these two mechanisms could be required to facilitate NuA4 targeting and efficient DNA repair. Likewise, the YAF9A YEATS domain could be required for the recruitment and/or retention of the SWR1 complex at DSBs to facilitate the incorporation of H2A.Z and the subsequent repair of DSBs. In moving forward, gaining a better understanding of the specificities of the YAF9 YEATS domains and the order of action for the different YAF9 complexes will be especially important in distinguishing between the proposed scenarios and revealing how the YAF9 chromatin readers coordinate different steps of DSB repair. Materials and Methods All work was conducted using Arabidopsis lines obtained from ABRC or generated by transformation with plasmids created using gateway cloning. Full experimental procedures, including plant growth conditions, plasmid construction and plant transformation, GUS staining, expression analysis, western blotting, assessment of DNA repair defects, and mass spectrometry are provided in SI Appendix. Raw data for the expression analyses and mass spectrometry are proved in the GEO (GSE308666) and MassIVE (MSV000099158) repositories. Supplementary Material Appendix 01 (PDF) Dataset S01 (XLSX) Dataset S02 (XLSX) Dataset S03 (XLSX) Dataset S04 (XLSX) Dataset S05 (XLSX)