Engineering chromatin to encode transcriptional immune memory in Arabidopsis

Article activity feed

1. 

   PREreview

   Oct 30, 2026

   This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/23810923.

   In this manuscript, Xu and colleagues investigate the chromatin-associated mechanisms underlying transcriptional memory following immune stimulation in Arabidopsis thaliana. Through an integrated approach combining time-resolved RNA-seq, single-nucleus transcriptomics, chromatin profiling, reverse genetics and targeted epigenome engineering the authors identify a subset of flg22-responsive genes that exhibit enhanced transcriptional responsiveness following an initial immune challenge.

   The study reveals that these memory-associated genes possess distinctive chromatin features including relatively low basal H3K4me3, enrichment of H3K27me3 and increased H3K4me3 following priming. The …

   This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/23810923.

   In this manuscript, Xu and colleagues investigate the chromatin-associated mechanisms underlying transcriptional memory following immune stimulation in Arabidopsis thaliana. Through an integrated approach combining time-resolved RNA-seq, single-nucleus transcriptomics, chromatin profiling, reverse genetics and targeted epigenome engineering the authors identify a subset of flg22-responsive genes that exhibit enhanced transcriptional responsiveness following an initial immune challenge.

   The study reveals that these memory-associated genes possess distinctive chromatin features including relatively low basal H3K4me3, enrichment of H3K27me3 and increased H3K4me3 following priming. The authors further show that the H3K4 demethylase JMJ14 is involved in regulating transcriptional memory. Most notably, using a CRISPR/dCas9-based SunTag system to recruit the H3K4 methyltransferase SDG2 they demonstrate that targeted chromatin modification at the WRKY29 locus enhances its subsequent transcriptional response and is associated with improved resistance to Pseudomonas syringae.

   This study advances our understanding the functional relationship between chromatin modification and transcriptional memory in plants. The use of targeted epigenome engineering to enhance the inducibility of endogenous immune genes is compelling and moves the field beyond purely correlative associations between histone modifications and stress-responsive transcription.

   The major outstanding questions concern the molecular specificity of the engineered H3K4 methylation, the role of JMJ14 in memory establishment, and the extent to which transcriptional memory operates in a cell-type-specific manner. Addressing these points by clarifying the interpretation of the existing evidence would strengthen the manuscript and provide a more comprehensive understanding of the chromatin mechanism underlying immune memory.

   Overall, this manuscript makes a strong conceptual contribution to plant epigenetics and provides a foundation for future investigations into chromatin-based transcriptional memory.

   Major comments:

   1. Distinguishing the specific contribution of H3K4me3 from other consequences of SDG2 recruitment

   A major advance of this study is the successful application of epigenome-engineering. Recruitment of SunTag:SDG2 to the WRKY29 promoter enhances transcriptional responsiveness following flg22 stimulation. The catalytically inactive and alternative-targeting controls strengthen the interpretation. (Fig. 6 and Supp FIg. 22)

   However, the authors report that targeted SDG2 recruitment increases both H3K4me2 and H3K4me3 at the engineered locus (Page 19, lines 412-415, Fig. 6a-b for H3K4me3, Supp Fig. 21a for H3K4me2). Thus, the observed transcriptional enhancement cannot be attributed exclusively to H3K4me3.

   Although the natural priming experiments suggest a stronger association with sustained H3K4me3 accumulation and the authors acknowledge the possible contribution of H3K4me2 in their Discussion (lines 542-549), the relative contribution of H3K4me2 and H3K4me3 remain unresolved. An orthogonal approach distinguishing these methylation states would strengthen the mechanistic conclusion. Alternatively, the authors could qualify their central claim to emphasize that targeted SDG2-mediated H3K4 methylation enhances transcriptional responsiveness.

   1. The mechanistic role of JMJ14 in transcriptional memory requires further clarification

   The identification of JMJ14 as a regulator of transcriptional memory raises an interesting mechanistic question, given its established role as an H3K4 demethylase. Enhanced transcriptional memory correlates with persistent accumulation of H3K4me3, yet loss of the H3K4 demethylase JMJ14 substantially compromises the memory response, presenting an apparent mechanistic paradox.

   The authors propose that JMJ14 maintains relatively low H3K4me3 levels in the resting state and thereby facilitates subsequent chromatin reconfiguration during immune priming. Although plausible, this model is not directly established by the current data. It remains unclear whether the impaired memory phenotype in jmj14 reflects defective chromatin-state transition, altered basal transcriptional activity, or indirect changes in immune signaling.

   While Figure 4 provides pre-challenge expression measurements, a more detailed analysis relating basal transcription, H3K4me3 abundance and subsequent priming responses in wild-type and jmj14 plants would help distinguish these possibilities. Examining H3K4me3 accumulation and persistence during priming and recovery in the mutant would provide especially valuable insight.

   Clarifying this relationship would strengthen the JMJ14-dependent chromatin-turnover model proposed in Figure 7.

   1. Cell-type-specific expression of memory-associated genes should be distinguished from cell-type-specific transcriptional memory

   The single-nucleus RNA-seq analysis represents an interesting dimension of the study that shows immune memory-associated genes exhibit more spatially restricted expression patterns than the broader population of flg22-responsive genes. The authors strengthen this observation through comparisons with independent single-cell datasets.

   However, the snRNA-seq experiments primarily characterize expression following acute flg22 treatment rather than directly comparing transcriptional responses in the primed and naive plants following recurrent stimulation. The data establish that genes identified as memory-associated in bulk transcriptomic experiments show preferential expression in particular cell populations, but do not demonstrate that these populations exhibit stronger transcriptional memory.

   Although the authors acknowledge this limitation in the Discussion, the interpretation of cell-type-specific immune memory could be more cautious. Comparing primed and naive transcriptional responses at single-nucleus resolution following recurrent flg22 stimulation would help determine whether transcriptional memory is indeed restricted to specific cell populations.

   Minor comments

   1. Definition of memory genes: The identification of 213 memory genes relies primarily on the 30-minute response (line 102-108). Could the authors clarify whether additional genes exhibit memory-like transcriptional kinetics at other time points? This would help determine whether the identified gene set captures the full extent of the transcriptional memory program.

   2. Biological reproducibility: While the authors provide ChIP-seq replicate correlations (Supp. Fig 14), showing whether priming-induced H3K4me3 enrichment at memory genes is consistent across biological replicates would strengthen confidence that these changes represent reproducible chromatin responses rather than variation between samples (Fig. 5a).

   3. Pathogen resistance: Independent bacterial growth measurements, such as CFU assays would strengthen the evidence for enhanced resistance phenotype observed using bioluminescence (Fig 6d).

   I reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor.

   Competing interests

   The author declares that they have no competing interests.

   Use of Artificial Intelligence (AI)

   The author declares that they did not use generative AI to come up with new ideas for their review.

   Read the original source
2. 

   Version published to 10.64898/2026.09.10.750729 on bioRxiv

   Sep 14, 2026