Elucidating tissue and subcellular specificity of the entire SUMO network reveals how stress responses are fine-tuned in a eukaryote

Abstract

SUMOylation is essential in plant and animal cells, but it remains unknown how small ubiquitin-like modifier (SUMO) components act in concert to modify specific targets in response to environmental stresses. In this study, we characterize every SUMO component in the Arabidopsis root to create a complete SUMO Cell Atlas in eukaryotes. This unique resource reveals wide spatial variation, where SUMO proteins and proteases have subfunctionalized in both their expression and subcellular localization. During stress, SUMO conjugation is mainly driven by tissue-specific regulation of the SUMO E2-conjugating enzyme. Stress-specific modulation of the SUMO pathway reveals unique combinations of proteases being targeted for regulation in distinct root tissues by salt, osmotic, and biotic signals. Our SUMO Cell Atlas resources reveal how this posttranslational modification (PTM) influences cellular- and tissue-scale adaptations during root development and stress responses. To our knowledge, we provide the first comprehensive study elucidating how multiple stress inputs can regulate an entire PTM system.

Fig. 1.. Description and validation of pMDS plasmid system for dual analysis of transcription and translation in plants.

(A) Organization of pMDS1 vector showing reporters for transcription [mTurquoise (mTurQ)], translation (C-terminal mVenus), and a 2A self-cleaving peptide. (B) Organization of pMDS2 vector showing reporters for transcription (mTurQ), translation (N-terminal mVenus), and a 2A self-cleaving peptide. (C) Confocal image of pMDS1_SHRpro:SHR:mVenus:mTurQ showing gene expression (mTurQ) in the stele region and protein (mVenus) translocating to endodermis in the root meristem. (D) Confocal image of pMDS2_VAM3pro:mVenus:VAM3:mTurQ showing subcellular expression (mTurQ) in the nucleus and protein (mVenus) moving to the vacuole in the root epidermis. Red channel shows mCherry expression. nu, nucleus; vac, vacuole; *, endodermis of root meristem. Scale bar, 10 μM. (E) Immunoblot of pSHR::SHR pMDS1 showing efficient ribosomal skipping induced by the 2A peptides. Bands corresponding to mTurquoise are labeled as mTurQ. (F) Immunoblot of pVAM3::VAM3 pMDS2 showing efficient ribosomal skipping induced by the 2A peptides. Bands corresponding to mTurquoise are labeled as mTurQ.

Fig. 2.. SUMO components exhibit distinct domains of expression and protein accumulation in Arabidopsis root tips.

(A) Schematic representation of the root tip zones and tissues measured in the Arabidopsis root. (B) Confocal images displaying gene expression (cyan) and protein localization (yellow) of all 32 SUMO components in the root tip of A. thaliana. The cell outline is shown in red by a genetic membrane marker. Scale bars, 100 μm. (C) Heatmap representing the semiquantitative level of gene expression and protein abundance of all SUMO components in the varying root zones and tissues (n = 5). Zero indicated no fluorescence observed, while level 4 or 5 defines a high level of fluorescence. For (B), image settings have been optimized for individual lines to allow comparison of expression patterns; for (C), image settings are identical between lines to allow quantitative comparisons between components.

Fig. 3.. Salt stress modulates the SUMO system in a tissue-specific manner.

(A) Heatmap representing the log₂ fold changes in gene expression within the SUMO components 3 hours after 150 mM salt stress in several root tissues and zones (n = 4). The top bar graph displays the number of SUMO genes per tissue type in a specific zone that change in a statistically significant manner (P ≤ 0.05). Dark gray boxes indicate combinations that were not analyzed, as no expression or protein was observed in this tissue. (B) Similar heatmap as in (A) representing the log₂ fold changes in protein abundance within the SUMO components 3 hours after 150 mM salt stress in several root tissues and zones. (C) Confocal microscopy images showing changes in expression (mTurquoise: cyan) and protein abundance (mVenus: yellow) of three major changers during salt stress. Scale bars, 100 μm. (D) Western blot of three major SUMO changers during salt stress, confirming the trend as seen in image analysis. UTP--glucose-1-phosphate uridylyltransferase (UGPase) was used as a loading control. (E) Schematic representation of the SUMO cycle and its components. The cycle represents the changes in protein abundance in the meristematic stele. Components with modest changes (fold change < 0.5) are shown as transparent, while those with the largest changes (log change > 0.5) are shown in solid color based on the scale of the heatmap in (B). Components with statistically significant changes are also outlined.

Fig. 4.. Osmotic stress regulates a distinct set of genes compared to salt stress.

(A) Heatmap representing the log₂ fold changes in gene expression within the SUMO components 3 hours after 300 mM mannitol stress in several root tissues and zones (n = 4). The top bar graph displays the number of SUMO genes per tissue type in a specific zone that change in a statistically significant manner (P ≤ 0.05). Dark gray boxes indicate combinations that were not analyzed, as no expression or protein was observed in this tissue. Boxes with a marked corner indicate significant changes. (B) Similar heatmap as in (A) representing the log₂ fold changes in protein abundance within the SUMO components 3 hours after 300 mM mannitol stress in several root tissues and zones. (C) Confocal microscopy images showing changes in expression (mTurquoise: cyan) and protein abundance (mVenus: yellow) of three major changers during salt stress. Scale bars, 100 μm. (D) Western blot of three major SUMO changers during salt stress, confirming the trend as seen in image analysis. UGPase was used as a loading control. (E) Schematic representation of the SUMO cycle and its components. The cycle represents the changes in protein abundance in the meristematic stele. Components with modest changes (fold change < 0.5) are shown as transparent, while those with the largest changes (log change > 0.5) are shown in solid color based on the scale of the heatmap in (B). Components with statistically significant changes are also outlined.

Fig. 5.. Biotic stress targets a distinct subset of SUMO system genes compared to abiotic stress.

(A) Heatmap representing the log₂ fold changes in gene expression within the SUMO components 3 hours after 1 μM flagellin treatment in several root tissues and zones (n = 4). The top bar graph displays the number of SUMO genes per tissue type in a specific zone that change in a statistically significant manner (P ≤ 0.05). Dark gray boxes indicate combinations that were not analyzed, as no expression or protein was observed in this tissue. Boxes with a marked corner indicate significant changes. (B) Similar heatmap as in (A) representing the log₂ fold changes in protein abundance within the SUMO components 3 hours after 1 μM flagellin treatment in several root tissues and zones. (C) Confocal microscopy images showing changes in expression (mTurquoise: cyan) and protein abundance (mVenus: yellow) of three major changers during 1 μM flagellin treatment. Scale bars, 100 μm. (D) Western blot of three major SUMO changers during 1 μM flagellin treatment, confirming the trend as seen in image analysis. UGPase was used as a loading control. (E) Schematic representation of the SUMO cycle and its components. The cycle represents the changes in protein abundance in the meristematic stele. Components with modest changes (fold change < 0.5) are shown as transparent, while those with the largest changes (log change > 0.5) are shown in solid color based on the scale of the heatmap in (B). Components with statistically significant changes are also outlined.

Fig. 6.. Stress-specific SCE1 interactome characterized by IP-MS analysis.

(A) Schematic representation of the experimental design and proteomics workflow used for SCE1–IP-MS analysis under distinct stress conditions. (B to D) Venn diagrams illustrating overlapping and stress-specific SCE1 interactors identified under mannitol, salt, and flg22 treatments, respectively. (E) Venn diagram summarizing the overlap of SCE1 interactors across all tested stress conditions relative to the control. (F) Expression analysis of SCE1 interactors identified from the salt stress experiment across root tissues and their pseudotime trajectories. The heatmap displays the proportion z-score (−3 to 3) of standardized gene expression profiles across tissue types and zones. The red-blue colormap represents relative expression levels, with red indicating up-regulation (+1 to +3 SDs above the mean) and blue indicating down-regulation (−1 to −3 SDs below the mean), with the mean represented as 0. Black boxes highlight qualitative cluster groups, and dendrograms depict hierarchical clustering of genes (rows) and zones (columns) based on expression similarity.