WRKY70 (At3g56400, Q9LY00) curation notes

2026-10-15: genome-wide WRKY70 binding at SA-biosynthesis, PR and senescence promoters (DAP-seq)

Open question: is WRKY70 a direct transcriptional regulator of SA biosynthesis genes
(especially ICS1/SID2, At1g74710) and of PR genes, as opposed to senescence genes?
The question matters because a module on an open PR models a WRKY70 -| SA-biosynthesis
feedback edge. The literature disagrees on direction:

• Wang et al. 2006 found ICS1 and SA elevated in wrky54 wrky70 PMID:17096590.
• Saleh et al. 2015 proposed WRKY70 as a repressor at PR1 together with non-sumoylated NPR1 PMID:26269953.
• Li et al. 2004 placed WRKY70 downstream of NPR1 as an SA-pathway activator PMID:14742872.

Analysis: genes/ARATH/WRKY70/WRKY70-bioinformatics/ (reproducible; just all). Full
results and limitations are in [file:ARATH/WRKY70/WRKY70-bioinformatics/RESULTS.md].

Data: the O'Malley et al. 2016 Plant Cistrome DAP-seq peaks for WRKY70 (dataset
WRKY_tnt/WRKY70_col; 8,244 peaks; the MEME motif is a W-box), plus 37 other WRKY DAP-seq
datasets for comparison. DAP-seq is in vitro PMID:27203113.
No in vivo WRKY70 ChIP-seq was analysed.

Result (2 kb upstream window, robust to a -3 kb/+0.5 kb window):

• No WRKY70 peak at ICS1/SID2. The nearest WRKY70 summit is about 11 kb away. Twelve
  of the 37 other WRKY DAP-seq datasets do bind the ICS1 promoter window.
• No WRKY70 peak at PR1, PR2, ICS2, PBS3 or PAD4. The PR1 promoter has 3 WT-box
  (YGACTTTT) motifs and is bound by 11 other WRKY datasets, but not by WRKY70.
• WRKY70 peaks at EDS5, CBP60G, SARD1 and PR5. These are promoter peaks. There are also
  peaks just downstream of the TSS at NPR1 (+126) and EDS1 (+75). The EDS5 and CBP60G
  sites are bound by most WRKY datasets (31 and 29 of 37), so they are not WRKY70-selective.
• No WRKY70 peak at the senescence markers. This covers SAG12, SAG13, SEN1 and ORE1.
  Leaf-senescence genes are not enriched.
• The control gene ACT2 has a WRKY70 peak.
• Weak, non-specific enrichment in SA-response genes. WRKY70 promoter targets are
  modestly enriched among GO "response to salicylic acid" genes (OR 1.66, p = 7.3e-3,
  uncorrected) and "defense response" genes. WRKY33 DAP-seq shows the same enrichment, so
  it is a generic WRKY/W-box property and not WRKY70-specific.

Interpretation:

• The in vitro binding data do not support a direct WRKY70 -| ICS1 promoter edge. The ICS1
  de-repression in wrky54 wrky70 may be indirect, or may depend on partners, chromatin or
  modification states that DAP-seq lacks.
• The data say nothing about activator versus repressor direction.
• A missing DAP-seq peak does not exclude in vivo occupancy.

Curation consequence:

• No GO annotation changes. A suggested question and a suggested experiment were added
  to the review, asking for WRKY70 ChIP-qPCR/ChIP-seq at ICS1 and PR1 under SA and
  pathogen conditions, with wrky54 wrky70 and the NPR1 sumoylation mutants as context.
• For the open module PR: the WRKY70 -| SA-biosynthesis edge should be treated as
  genetic (indirect) unless in vivo binding at ICS1 is shown.