Brachypodium distachyon 21 (Bd21) was used in this study. Bd21 seeds were surface sterilized with 20% sodium hypochlorite for 10 min, rinsed with sterile water five times, incubated in water for 12 h at room temperature, and then transferred to a wet filter paper to germinate in darkness for 24 h (22–25°C). The seeds that germinated uniformly were selected and spread in plastic pots containing ½ MS (Murashige and Skoog), which was changed every 2 days, and grown in a greenhouse with a daily photo cycle of 16 h light/8 h dark (26°C/18°C) and 65–75% air humidity.

The yeast strains used in this study are listed in Supplementary Table S1. Yeast cells were cultured at 30°C in YPAD or in a synthetic dextrose (SD) medium (0.67% Bacto-yeast nitrogen base without amino acids, 2% glucose) supplemented with necessary nutrients as recommended (Amberg et al., 2005). The lithium acetate method of yeast transformation was carried out as described (Ito et al., 1983). The sources and preparation of the ubc13Δ::HIS3 (Brusky et al., 2000) cassette was as described previously.

To clone B. distachyon UBC13 genes, total RNA was extracted from B. distachyon seedlings using a TRIzol reagent (Invitrogen), and the ThermoScript RT-PCR kit (Invitrogen) was used to synthesize first-strand cDNA. The BdUBC13 ORFs were amplified by PCR from the above cDNA preparation using gene-specific primers (Supplementary Table S2). The yeast two-hybrid vectors pGAD424Bg and pGBT9Bg were derived from pGAD424 and pGBT9 (Bartel and Fields, 1995). Genes cloned into plasmid pCAMBIA1300 are behind a CaMV 35S promoter.

The yeast two-hybrid strain PJ69-4a (James et al., 1996) was co-transformed with different Gal4_AD and Gal4_BD vectors. The co-transformed colonies were selected on an SD-Leu-Trp plate. Protein interaction was determined on synthetic complete medium lacking Trp, Leu and His, and supplemented with 3-amino-1,2,4-triazole (3-AT, Sigma-Aldrich).

Full-length BdUBC13A and BdUBC13B were cloned in plasmid pET30a(+). Full-length AtUEV1A, AtUEV1B, AtUEV1C, and AtUEV1D were cloned in plasmid pGEX-6p-1. The His₆ and GST fusion constructs were transformed into E. coli strain BL21 (DE3). The recombinant proteins were purified with Ni Sepharose and glutathione (Amersham Pharmacia) according to the manufacturer’s protocol.

For the pull-down assay, crude cell extracts were loaded on Glutathione Sepharose^TM 4B beads and then 10 μg of purified His₆-BdUbc13A or His₆-BdUbc13B was later added. After incubation and washing, the GST beads were boiled with SDS-PAGE loading buffer for 10 min before western blotting analysis.

In vitro Ub conjugation reactions were performed by using the purified Ubc13A/B and GST-AtUev1A/C proteins as described above. Ub thioester conjugation initiation reagents were purchased from Boston Biochem. The 20 μl reaction mixture contained 225 nM E1 enzyme, 200 μM Ub, 1 mM MgATP, 1 mM Ubc13, and 1 mM Uev1 in the supplied reaction buffer. The K63R mutant Ub proteins were purchased from Boston Biochem (UM-K63R). The conjugation reactions were performed at 37°C for 2 h. Samples were subjected to SDS-PAGE (12%), and Ub and poly-Ub were detected by western blotting using mouse monoclonal anti-ubiquitin (P4D1; Cell Signaling).

Yeast strain HK578-10D and its isogenic ubc13Δ were transformed with pGAD-Ubc13A or pGAD-Ubc13B. Transformants were selected on SD-Leu plates. The gradient plate and serial dilution assays were performed as described previously (Xu et al., 2014). A gradient plate was made by pouring molten YPD agar with indicated MMS concentration into a tilted square Petri dish. After solidification, the Petri dish was returned flat and equal volume of YPD agar without MMS was poured to form the top layer. Cells from an overnight culture were mixed with molten YPD agar and immediately imprinted onto freshly made gradient plates via a microscope slide.

Yeast strain DBY747 derivative WXY849 bears a trp1-289 amber mutation that can be reverted to Trp⁺ by several different mutation events (Xiao and Samson, 1993). DBY747 cells with the ubc13Δ mutation were transformed with pGAD-Ubc13A or pGAD-Ubc13B. The transformants were selected on SD-Leu plates. Each set of experiments contained five independent cultures of each strain. Overnight yeast cultures were counted using a hemocytometer and five replicate cultures of 10 ml of SD-Leu medium for each strain with a start concentration of 20 cells/ml were incubated at 30°C until the cell number reached 2 × 10⁷ cells/ml. Cells were collected by centrifugation at 4000 rpm, washed and plated onto YPAD in triplicate to score the total survivors and onto SD-Trp plates to score Trp⁺ revertants. Spontaneous mutation rates (number of revertants per cell per generation) were calculated as previously described (Williamson et al., 1985).

GFP-UBC13A and GFP-UBC13B fusion genes were constructed by cloning UBC13A and UBC13B ORFs into the binary vector pCAMBIA1300-GFP downstream of GFP at the SacI and KpnI sites. The resulting constructs were introduced into Agrobacterium tumefaciens PMP90 then the PMP90 was co-infiltrated with A. tumefaciens P19 into Nicotiana benthamiana leaves (Waadt and Kudla, 2008). GFP and DAPI fluorescence were observed 2 days after transformation using a Zeiss 5 live confocal microscope.

Different development phase samples were homogenized in liquid nitrogen. Total protein was extracted using a buffer containing 50 mM Tris-HCl pH 8.0, 0.3 M NaCl, 2 mM EDTA, 10% Glycerol, 0.1% Triton X-100, 10 mM PMSF, 3 mM DDT, and 1x protease inhibitor cocktail (Sigma-Aldrich). The extract was centrifuged at 16,000 g for 10 min at 4°C. The supernatant was then transferred to a new tube and boiled with SDS-PAGE loading buffer for 5 min before western blotting analysis. The BdUbc13 protein in each sample was detected by anti-hUbc13 monoclonal antibody 4E11 (Andersen et al., 2005).

To identify B. distachyon UBC13 genes, two highly conserved Arabidopsis UBC13 genes were used to blast the B. distachyon genomic database. Two genes, BRADI2G16770 and BRADI2G46290, were obtained and named BdUBC13A and BdUBC13B, respectively. The two deduced BdUbc13 proteins contain only three amino-acid variations. In comparison with Ubc13s from other organisms, the active site Cys87 (green asterisk) for the Ub thioester formation (McKenna et al., 2001), Met64 (blue asterisk) for the physical interaction with RING finger E3 ligases (Wooff et al., 2004), and Glu55, Phe57, and Arg70 (red asterisks) for the physical interaction with Mms2 (Pastushok et al., 2005) are all conserved (Figure 1A). To further investigate the evolution of B. distachyon UBC13, the nucleotide sequences of BdUBC13A and BdUBC13B were compared with UBC13 coding sequences from other organisms. There is 87% nucleotide sequence identity between BdUBC13A and BdUBC13B. The phylogenetic tree shows that BdUBC13s have evolved from a common plant ancestor and are closely related to OsUBC13 (Figure 1B).

In budding yeast, Ubc13 and Mms2 form a heterodimer (Brusky et al., 2000) that synthesizes Lys63-linked polyubiquitin chains, which are required for error-free DDT (Broomfield et al., 2001). In human (Hofmann and Pickart, 1999), Arabidopsis (Wen et al., 2008) and Zebrafish (Wen et al., 2012), Ubc13s also interacts with Uev1s. A yeast two-hybrid assay (Fields and Song, 1989) was used to determine whether BdUbc13s interact with Uev1s. As seen in Figure 2A, the two BdUbc13s interact with yeast Uev1 (yMms2) and two human Uev1s (hMms2 and hUev1) in the presence of 5 mM 3-AT. They also have strong interaction with all four Arabidopisis Uev1s (Figure 2B). To further confirm the physical interaction between BdUbc13s and Uev1s in vitro, a GST-affinity pull-down assay was performed. As shown in Figure 2C and Supplementary Figure S1A, recombinant GST-AtUev1s fusion proteins can pull down purified recombinant His₆-BdUbc13 fusion proteins, but the GST alone cannot. Based on the above observations, we conclude that BdUbc13s can form stable heterodimers with Uev1s in vivo and in vitro.

It is well known that Ubc13 and Mms2 (Uev) play a key role in the DDT pathway by polyubiquitinating PCNA with K63-linked chains (Hoege et al., 2002). Ubc13 and Uev1 together can catalyze K63-linked free Ub chains (Hofmann and Pickart, 1999). However, hUbc13-K92 serves as the site of Ub conjugation and a hUbc13-K92A mutation does not affect the free chain formation (McKenna et al., 2001). To avoid Ub conjugation to Ubc13 itself, we made the corresponding BdUbc13-K94A mutation (Figure 1A, purple asterisk) and found that BdUbc13A together with AtUev1A or AtUev1C can indeed generate free Ub chains while either BdUbc13 or AtUev1 alone cannot (Figure 3). More importantly, if Ub-K63R was used in the same reaction, no poly-Ub chains were observed, confirming that they are indeed K63-linked poly-Ub chains. The same result was also obtained for BdUbc13B together with AtUev1A or AtUev1C (Supplementary Figure S1B). Hence, B. distachyon Ubc13s and AtUev1s catalyze Lys63-linked Ub chains in vitro.

Yeast Ubc13 and Mms2 function in the error-free DDT pathway (Broomfield et al., 1998; Brusky et al., 2000). To determine whether the BdUbc13s also play critical roles in error-free DDT, we took advantage of the UBC13 gene conservation within eukaryotes and performed a gradient plate assay (Xiao et al., 1996) and spontaneous mutagenesis assay (Xiao and Samson, 1993). As shown in Figure 4A, the expression of BdUBC13A or BdUBC13B genes can protect the ubc13 null mutant from killing by MMS, while the pGAD424 vector cannot.

The DDT pathway differs from other DNA repair pathways in that its defect induces hypersensitivity to a wide range of DNA-damaging agents that cause replication blocks. To further confirm that BdUBC13s function in the yeast DDT pathway, we performed a serial dilution assay in the presence of several representative DNA-damaging agents including MMS (an alkylating agent), UV irradiation, and 4-nitroquinoline 1-oxide (4NQO, causes bulky adducts). As shown in Figure 4B, both BdUbc13A and BdUbc13B can rescue the ubc13 null mutant phenotypes while the vector transformant has no effect in the phenotypic rescue.

One of the hallmarks of error-free DDT is its ability to limit spontaneous and DNA-damage-induced mutagenesis (Xu et al., 2015). Under our experimental conditions, the ubc13 mutant displayed a 13-fold increase in the spontaneous mutation rate. The expression of BdUBC13A or BdUBC13B reduced the spontaneous mutation rate of the ubc13 null mutant to nearly the wild type level (Figure 4C), and the slight difference with the wild-type rate could be accounted for by plasmid loss in the transformants. Taken together, our results indicate that BdUbc13 can replace the function of yUbc13 in the yeast error-free DDT pathway.

In order to assess possible BdUbc13 functions in its native host, we examined the subcellular distribution of BdUbc13s in tobacco. GFP-BdUbc13A and GFP-BdUbc13B fusion constructs were created, transiently transfected to tobacco (N. benthamiana) leaves by A. tumefaciens and their subcellular localization was monitored by fluorescence microscopy. As shown in Figure 5, both GFP-BdUbc13A and GFP-BdUbc13B were found in the nucleus and cytoplasm, but appear to be enriched in the nucleus. In addition, the fluorescent signals were also readily detected in the cytoplasmic membrane. Since, the DDT function occurs exclusively in the nucleus, the above observations suggest that BdUbc13A and BdUbc13B are also involved in other cellular pathways in addition to DDT.

It was reported that the Arabidopsis UBC13 gene has a relatively uniform expression in different tissues or during biotic or abiotic stresses (Wen et al., 2006), and OsUBC13 expression remains remarkably constitutive during development (Zang et al., 2012). To investigate BdUBC13 expression pattern during development, total protein was extracted from 7-days seedling as well as different tissues and subjected to western blot analysis using a monoclonal antibody raised against human Ubc13 (Andersen et al., 2005). As seen in Figure 6A, total BdUbc13 levels (Ubc13A + Ubc13B) do not appear to fluctuate dramatically during development and in different tissues.

We also analyzed relative BdUBC13 expression in response to abiotic stress from an available database (Priest et al., 2014). It is of great interest to note that while salt and drought stresses do not significantly alter BdUBC13A expression, cold stress induces BdUBC13A over time and heat treatment dramatically reduces BdUBC13A expression (Figure 6B), suggesting the involvement of BdUBC13A in response to environmental temperature stresses. Unfortunately there was no data on BdUBC13B in the same database.