The RUS gene family is found throughout eukaryotes and was expanded in algae
We previously reported that ROOT UV-B SENSITIVE1 (RUS1) and RUS2 are required for post-germination growth in Arabidopsis, and that they likely play a role in vitamin B6 (pyridoxal-5’-phophate) homeostasis. The RUS1 and RUS2 genes both encode proteins that contain a DOMAIN OF UNKNOWN FUNCTION 647 (DUF647) . The Arabidopsis genome encodes for six DUF647-containing proteins (RUS1 through RUS6). RUS proteins are found in most eukaryotic species, including all plants, and most fungus and animals. We previously identified RUS3 as the clear ortholog to the single RUS gene found in most animal genomes . All plant genomes analyzed were found to encode for multiple RUS proteins, usually six or more. Protein sequence analyses identified clear orthologs of RUS1, RUS2, RUS3, and RUS6 in all plant genomes. The genomes of rice and the moss Physcomitrella patens each contained recent duplications of RUS6, but the rice genome lacked a clear RUS4 ortholog, and the P. patens genome lacked a clear RUS5 ortholog. The genome of the gymnosperm Pinus sylvestrus contained orthologs for all six RUS genes (Figure S1). Interestingly, we also identified orthologs for all six RUS genes in the genome of a Charophyte algae, Klebsormidium nitens, which branched from the plant lineage at least 700 million years ago . Therefore, the expansion of the RUS gene family into the current set of six genes occurred long before the evolution of the embryophytes began.
Identification and analysis of knockout mutants for RUS3, RUS4, and RUS5
In an effort to further understand the functional roles for all RUS members, we screened and identified knockout mutants for RUS3 (AT1G13770), RUS4 (AT2G23470), RUS5 (AT5G01510) and RUS6 (AT5G49820). Potential T-DNA insertional lines were identified in the public database and verified by gene-specific PCR markers. Homozygous knockout mutants were identified for RUS3 (two lines: SALK_135717C and SALK_042033C), RUS4 (one line: GK-447F02-024530) and RUS5 (one line: SALK_038772C) (Fig. 1). All mutant lines contain T-DNA insertions in exons (Fig. 1) (Figure S2). Homozygous mutants for all three genes (RUS3, RUS4, and RUS5) were isolated, suggesting that mutations in these three genes do not cause embryo lethality (Fig. 1). No noticeable morphological differences were observed between these mutants and the WT (Col-0) plants when grown under standard growth conditions.
Loss of function in RUS6 (AT5G49820) results in embryo lethality
Two T-DNA insertion lines (GK278G06 and emb1879/cs16037) were identified for RUS6 (AT5G49820), and verified by PCR markers and direct sequencing (Fig. 2A, B; Figure S2). GK278G06 was obtained from Gabi-Kat  (https://www.gabi-kat.de/) and confirmed to have a pAC106/pAC116 T-DNA insertion in exon 11 (Fig. 2A). The cs16037/EMB1879 line was obtained from ABRC (Arabidopsis Biological Resource Center). The cs16037/EMB1879 line has a deletion from the promoter region until intron 6, which was replaced by the pCSA104 T-DNA insertion (Fig. 2B). The deletion/insertion was verified by PCR markers and DNA sequencing (Figure S2). After confirming the mutations, we named GK278G06 and cs16037/emb1879 as rus6-1 and rus6-2, respectively (Fig. 2).
We were unable to identify any homozygous rus6 mutants in either of the initial seed stocks for rus6-1 or rus6-2. In order to produce homozygous rus6 plants, rus6-1/+ and rus6-2/+ were each self-fertilized and their progenies were grown. PCR-based genotyping was used to genotype individual progeny, but no homozygous rus6 mutants were identified in the offspring of either heterozygous parent (rus6-1/+, n=121; rus6-2/+, n=14), suggesting that homozygous rus6 mutants are embryo lethal.
The rus6-1 T-DNA insertion contains a sulfadiazine (Sul) resistance gene. We grew the offspring of self-fertilized rus6-1/+ plants in the presence of Sul, and observed that 65.27% of the seedlings displayed Sul resistance and 34.73% displayed Sul sensitivity (n=262) (Table 1; Fig. 2C). These numbers were consistent with a 2:1 ratio of rus6-1/+ to +/+ plants, which is expected if the homozygous rus6-1 plants are absent. A subset (n=17) of the Sul-resistant plants were PCR genotyped and were all identified as rus6-1/+; no homozygous rus6-1 plants were found.
The rus6-2 T-DNA insertion confers Basta (glufosinate) resistance. In agreement with the rus6-1 results, we again observed results consistent with a lack of homozygous rus6-2 plants. 66.20% of samples displayed basta resistance, and 33.80% displayed basta lethality (n=213) (Table 1). The basta resistant plants (n=22) that were PCR genotyped were all rus6-2/+.
The Sul and Basta resistance results conformed to the expected 66.7% to 33.3% (2:1) segregation ratio for heterozygous to WT seedlings if homozygous progeny were missing (Table 1). Additionally, seed germination rates were comparable between the mutant lines and wild-type controls, suggesting that the seeds of homozygous rus6 embryos were not produced. Thus, we hypothesized that the lack of rus6 homozygotes was caused by early embryo lethality leading to seed abortion, rather than failed germination. Taken together, these results suggested that homozygous mutations in rus6 result in embryo lethality.
Loss of function in RUS6 disrupts embryo development, leading to a white developing seed phenotype.
We observed rus6/+ mutant plants from germination through maturity, and found that all vegetative parts of the plant were indistinguishable from WT (Fig. 2D, E). To investigate the lack of homozygous rus6 seeds, we opened rus6/+ siliques and characterized the developing seeds inside. While most of the developing seeds were green, similar to wild-type plants, we also observed a high percentage of developing seeds that were white, or brown and wrinkled, depending on the age of the silique (Fig. 3). We suspected that the white developing seeds contained the rus6 homozygotes, and predicted that they represented 25% of the seeds in the silique, to fit a 3:1 ratio of green to white seeds . A more extensive phenotypic analysis of developing seeds found 25.59% white or brown seeds in the siliques of rus6-1/+ plants (n=895) and 23.88% white or brown seeds in the siliques of rus6-2/+ plants (n=356) (Table 2). These results suggested that rus6 homozygotes are embryo lethal, and are the cause of the white developing seed phenotype.
Complementation abolishes the rus6 embryo lethal phenotype
Our initial analyses of the rus6-1 and rus6-2 mutations strongly suggested that at least one functioning copy of RUS6 is required in Arabidopsis plants. To reduce the possibility of an additional T-DNA insertion somewhere in genome being fully or partially responsible for the rus6 phenotype, we twice backcrossed rus6-1/+ plants to wild-type Col-0. The rus6 phenotype remained consistent in the purified backcrossed line, which led strong support to the rus6-1 mutation being the cause of the phenotype.
In order to genetically complement the rus6-1 mutation, we created a chimeric pZP222 construct containing RUS6-GFP driven by the native RUS6 promoter (RUS6::RUS6-GFP). The GFP tag was included for later analysis with fluorescence microscopy. rus6-1/+ plants were transformed using Agrobacterium tumefaciens, and T1 seeds were harvested and plated on antibiotic selection MS media. Two resistant T1 plants were identified and PCR analysis confirmed that they contained the RUS6::RUS6-GFP transgene. T2 seeds were collected from each line, and antibiotic selection and PCR genotyping was performed. We identified rus6-1 homozygous plants in the T2, which contained at least one copy of the RUS6:RUS6-GFP transgene (Fig. 4A, B). The complementation of the rus6 lethality phenotype by RUS6:RUS6-GFP demonstrated that the rus6-1 mutation was responsible for the rus6 mutant phenotype (Fig. 4C).
rus6 homozygous mutations prevent embryo development past the globular phase
To examine differences in embryo development between the white and green seeds in rus6-1/+ siliques, we performed Differential Interference Contrast microscopy (DIC) on developing seeds. Seeds from the same rus6-1/+ silique were removed, cleared, and examined together, and the results were consistent across siliques analyzed. We initially performed microscopy on seeds from late stage siliques of rus6-1/+ plants, and observed that the white seeds completely lacked a detectable embryo. We then examined seeds from siliques of decreasing maturity, which resulted in an increase in the number of white seeds that contained embryos, which were never observed to be past the globular stage. Finally, we observed that in very young siliques all of the white seeds contained globular phase or earlier embryos. These results suggested that the rus6 embryos were in fact initiated, but degraded and became undetectable after failing to advance past the globular phase.
The rus6 embryos in white seeds were severely delayed, and unable to develop past the globular phase (Fig. 4A, 4B, 4C, and 4D). In contrast, the embryos inside developing green seeds, which were either rus6/+ or wild-type, had normal developmental morphology (Fig. 5E and F). Additionally, the embryos in green seeds in each silique examined were all at a similar stage of development. The rus6 embryos displayed altered morphology, and careful observations determined that they were unable to reach the transition phase. The hypophysis or columella cells were either absent or distorted in such a way as to appear as part of the lower tier (Fig. 5B). Additionally, the suspensors of mutant embryos at this stage were difficult to detect and often appeared to be absent. Our analyses suggest that rus6 mutant embryo development stalled at the mid- to late- globular phase, and that the embryos subsequently deteriorated leading to failed seed development.
RUS6 is expressed in the embryo
To observe RUS6 expression in vivo, we analyzed GFP fluorescence in homozygous rus6 mutants complemented by our RUS6::RUS6-GFP construct. To minimize the auto-florescence that comes with more developed tissues, we performed laser scanning confocal microscopy on embryos in the late heart to early torpedo stages. We detected GFP fluorescence in complemented embryos that was significantly above the background auto-fluorescence seen in the wild-type control (Fig. 6). Observation at higher magnifications revealed that RUS6 is not specifically localized to the cell wall, nucleus, mitochondria, or any diffuse organelle. In contrast, fluorescence patterns suggested that RUS6 was localized to either the cytosol, chloroplasts, or other plastid [15,16]. This was consistent with TargetP 1.1 prediction, which predicted that RUS6 localizes to either the chloroplast or other cellular location, but lacked strong support for one over the other .
RUS6 expression in vegetative and reproductive organs
In order to further evaluate RUS6 expression patterns, we use a pBI101 construct to generate a RUS6::GUS reporter gene. The RUS6 promoter used in this reporter was the same region that was successfully used in the complementation of the rus6-1 mutation. Following Agrobacteria-mediated transformation, selection and PCR analysis confirmed twelve primary (T1) transformants. We performed preliminary GUS staining on all twelve lines, and selected the two with the highest GUS expression levels for further imaging and analyses. T2 plants from line 12 yielded the highest GUS activity in the flowers, while T2 plants from line 1 had the highest expression for all other tissues. RUS6::GUS expression was observed to be subtle, surprisingly dynamic, and was only detected at specific stages of development.
We stained one- through six-days-old RUS6::GUS light-grown seedlings grown vertically on M.S. plates. No GUS activity was observed in one day old seedlings, but two days old seedlings showed some degree of GUS activity in the cotyledons (Fig. 7A, B, C). We were unable to detect GUS activity in three- to six-days-old seedlings. Moreover, the GUS activity in two-days-old seedlings was only present in approximately 50% of the seedlings. This suggests that RUS6 expression was dynamic and temporally specific to a precise stage of development. Dark-grown seedlings did not show GUS activity in the cotyledons until day three, which was sustained through day five, and showed no activity by day six. The dark-grown experiment suggested that RUS6 expression was required for a longer period of time in the dark.
In ten-days-old seedlings, GUS activity was observed in the root junction during lateral root formation. In 20-days-old seedlings, GUS activity became clearly defined to the edges of the developing root primordia. GUS activity was also observed at this time in some lateral roots (Fig. 7D, E), and very faintly at mid-length in the primary root. Interestingly, some lateral roots, root tips, and root junctions showed GUS activity, while others did not. GUS expression did not appear to be based on the length of the lateral root, or any other observable marker of development.
GUS activity was not detected in leaves at any stage of development. However, flowers had the highest detected GUS activity in the plant, which was especially high in the anther (Fig. 7F, G, H, I). GUS activity was uniformly highest in the flower at stage 11, (as defined by Smyth et al., 1990 ). However, some flowers at later stages showed GUS activity, while others at the same stage of development did not. This pattern persisted even in flowers attached to the same inflorescence stem. Further investigation of dissected anthers revealed that GUS activity was especially high in the tapetum (Fig. 7I).