A cryptic natural variant allele of BYPASS2 suppresses the bypass1 mutant phenotype

Abstract

The Arabidopsis (Arabidopsis thaliana) BYPASS1 (BPS1) gene encodes a protein with no functionally characterized domains, and loss-of-function mutants (e.g. bps1-2 in Col-0) present a severe growth arrest phenotype that is evoked by a root-derived graft-transmissible small molecule that we call dalekin. The root-to-shoot nature of dalekin signaling suggests it could be an endogenous signaling molecule. Here, we report a natural variant screen that allowed us to identify enhancers and suppressors of the bps1-2 mutant phenotype (in Col-0). We identified a strong semi-dominant suppressor in the Apost-1 accession that largely restored shoot development in bps1 and yet continued to overproduce dalekin. Using bulked segregant analysis and allele-specific transgenic complementation, we showed that the suppressor is the Apost-1 allele of a BPS1 paralog, BYPASS2 (BPS2). BPS2 is one of four members of the BPS gene family in Arabidopsis, and phylogenetic analysis demonstrated that the BPS family is conserved in land plants and the four Arabidopsis paralogs are retained duplicates from whole genome duplications. The strong conservation of BPS1 and paralogous proteins throughout land plants, and the similar functions of paralogs in Arabidopsis, suggests that dalekin signaling might be retained across land plants.

Figure 1

Dalekin signaling model. A, Dalekin is a metabolite whose synthesis in roots is regulated by BPS1 (Van Norman et al., 2004). Dalekin affects root development and also is transported to the shoot where it arrests growth. B, Dalekin’s growth-inhibiting activity in shoots is linked to reduced cytokinin (CK) responsiveness and downregulation of WUSCHEL (WUS) in the shoot apical meristem.

Figure 2

A natural variant suppressor screen revealed diverse impacts on the bps1-2 phenotype. A, Col-0 and the bps1-2 mutant (Col-0). bps1 single mutants have developmental defects in their shoots and roots. B, Modified bps1-2 phenotypes derived from crosses with six different accessions. Size bars = 1 mm, seedling age = 13 dpi.

Figure 3

The Apost-1 suppressor of bps1-2 is semi-dominant. From left to right, phenotypes of Col-0, bps1-2 homozygous for the Apost-1 suppressor, bps1-2 heterozygous for the suppressor, and bps1-2 that lacks the suppressor. Size bars = 2 mm, seedling age = 11 dpi.

Figure 4

The Apost-1 suppressor has a stronger impact on shoot tissue. We compared phenotypes of Col-0, Red4A, bps1-2, and Red4A bps1-2 in 4-day seedlings (A), 6-day seedlings (B), and 10-day seedlings (C). The bps1-2 and Red4A bps1-2 roots always appeared abnormal (e.g. root hairs approached the RAM) whereas leaf development in Red4A bps1-2 largely matched that of Col-0 and Red4A, unlike the leaves of bps1-2. Size bars = 1 mm.

Figure 5

Dalekin is still produced in Apost-1–suppressed bps1-2. A, Examples of GUS staining in root meristems of pCYCB1;1::GUS after 17 h of metabolite or control treatments (dalekin bioassay). Scale bars = 100 μm. B, Bioassay results indicate that dalekin levels in bps1-2 and Red4A bps1-2 are indistinguishable. Nine roots tested for each treatment. Statistically different outcomes are indicated with different letters (Mann–Whitney U test, P < 0.05). The box limits are the 25% and 75% quartiles, the center line is the median, and the whiskers represent the full data range.

Figure 6

BPS2 is the Apost-1 suppressor of bps1-2. A, Bulked segregant mapping of a bps1-2 modifier encoded by the Apost-1 accession in selfed progeny of the Orange6 bps1-2/+ introgression line (Supplemental Figure S1). Alleles polymorphic between Apost-1 and Col-0 were scored for relative frequencies by taking the ratio of Apost-1 SNPs in suppressed versus unsuppressed bps1-2 mutants (y-axis) in 10 kb bins across all five chromosomes of Arabidopsis (x-axis). The highest value indicates the position with the greatest frequency of Apost-1 SNPs. The table below shows the number of Col-0 (reference) and Apost-1 (non-reference) reads for each phenotypic category within the top 10 kb bin on Chr2. B, BPS2 gene map and adjacent genomic sequences (2733 bp). Gray boxes represent flanking genes. Apost-1 SNPs are shown along with their positions relative to the transcription start site. Coding sequence SNPs also show their corresponding amino acid alterations. Diagram is to scale. C, Pictures of the BPS2 transformant phenotypes and controls at 7 dpi. Insertion of BPS2^Apost-1 into bps1-2 bps2-2 double mutants reinstates the Apost-1–suppressed bps1-2 phenotype, confirming suppressor gene identity. Scale bars = 2 mm.

Figure 7

The BPS gene family is highly conserved in land plants. Phylogenetic analysis of the BYPASS homologs. Maximum likelihood tree of BYPASS1 homologs from fully sequenced genomes. Branch supports were estimated by 1000 replicates. Taxa mentioned in the results are indicated by red text.