Fig 1.
Extracellular HMGB3 activates innate immune responses.
A. HMGB3- or Pep1-induced MAPK activation in Arabidopsis. Leaves were collected 15 min after infiltration with water containing either 1 μM recombinant HMGB3 (HMGB3) or Pep1 peptide (Pep1) for the MAPK activation assay. Activation of MPK3, MPK4, and MPK6 by MAPK kinase-mediated phosphorylation of the TEY sequence was detected with α-pTEpY antibody using immune-blot (IB) analyses. Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) large subunit protein stained with Coomassie Brilliant Blue (CBB) served as a loading control. 1 and 2 denote independent biological replicates. B. RT-PCR analysis of WRKY33 (upper panel) and PDF1.2 (lower panel) expression at the indicated times after infiltration with water containing either 0.1 μM HMGB3 or Pep1. Expression levels were plotted relative to the expression in untreated leaves. Data are the mean ± SD (n = 4). C. HMGB3- or Pep1-induced callose deposition in Arabidopsis. Leaves were stained with aniline blue 15 h after infiltration with water containing either 0.1 μM HMGB3 or Pep1. Representative pictures are shown in the left panel. Bars = 100 μm. Data are the mean ± SD (right panel, n = 20). D. HMGB3- or Pep1-induced resistance to B. cinerea. Leaves were infiltrated with water containing the indicated concentrations of HMGB3 or 1 μM Pep1 one day before B. cinerea inoculation. Representative disease symptoms at 3 days post infection (dpi) are shown in the upper panel. Data corresponding to this time point are presented as the mean ± SD (lower panel, n = 6). Leaves infiltrated with water served as mock control in all experiments. Asterisks in B, C, and D indicate significant differences from the mock-treated leaves (t test, P < 0.05).
Fig 2.
BAK1 and BKK1 are required for extracellular HMGB3-induced innate immune responses.
A. HMGB3- or Pep1-induced MAPK activation was compromised in bak1-5 and bak1-5/bkk1-1. Leaves were collected 15 min after infiltration with water containing 1 μM HMGB3 or Pep1 for the MAPK activation assay. pTEpY phosphorylation of the TEY motif of MPK3, MPK4, and MPK6 was detected with α-pTEpY antibody. Rubisco large subunit protein stained with CBB served as the loading control. B. HMGB3-induced defense-related gene expression was compromised in bak1-5/bkk1-1. Leaves were collected 30 min after infiltration with water containing 0.1 μM HMGB3. Following RT-PCR, expression levels were plotted relative to the expression in water-treated (mock) wt leaves, which was set at 1. Data are the mean ± SD (n = 4). C. HMGB3-induced callose deposition was compromised in bak1-5/bkk1-1. Leaves were stained with aniline blue 15 h after infiltration with water containing 1 μM HMGB3. Representative pictures are shown in the left panel. Bars = 100 μm. Data are the mean ± SD (right panel, n = 20). D. HMGB3- or Pep1-induced resistance to B. cinerea was compromised in bak1-5/bkk1-1. Leaves were infiltrated with water containing 1 μM HMGB3 or Pep1 one day before B. cinerea infection. Representative disease symptoms at 3 dpi are shown in the left panel and the data corresponding to this time point are presented as the mean ± SD (right panel, n = 6). Leaves infiltrated with water served as mock control in experiments B-D. Asterisks in B and C indicate significant differences from the mock-treated leaves or between HMGB3-treated wt vs bak1/bkk1 (t test, P < 0.05).
Fig 3.
HMGB3 is released into apoplast by B. cinerea infection.
HA-HMGB3 was detected in apoplast only after B. cinerea inoculation. HMGB3 localized in nucleus and cytoplasm, but not in apoplast, in the absence of the pathogen. Total leaf extract, the apoplastic fraction (Apo), cytoplasmic (Cyto), and nuclear (Nuc) fractions were size fractionated on 12% SDS-PAGE followed by IB analyses with α-HA or α-histone H3 antibody. Histone H3 level were monitored to assess contamination by nuclear proteins. N. benthamiana leaves were spray inoculated with a spore suspension of B. cinerea one day after infection with Agrobacterium carrying an empty vector (EV) or the HA:HMGB3 expression vector. Total and subcellular protein fractions were prepared one day after B. cinerea inoculation (two days after agro-infection). The band detected at the expected size for HA-HMGB3 (~20 kDa) is marked with red asterisk. CBB stained gel indicates that loading of three apoplastic fractions were very similar, as was loading of the two cytoplasmic fractions and the two nuclear fractions.
Fig 4.
Silencing HMGBs increases susceptibility to B. cinerea.
A. Relative expression of HMGBs (HMGB1-HMGB14) in 4-week-old plant from three different amiR-hmgbs lines (#3, #5 and #9) was compared to their expression levels in wt plants, which was set at 1 and denoted by a red line. Data are the mean ± SD (n = 4). B. amiR-hmgbs lines showed enhanced susceptibility to B. cinerea infection. Representative disease symptoms at 3 dpi are shown in the upper panel, and the data corresponding to this time point are presented in the lower panel as the mean ± SD (n = 6). C. Extracellular HMGB3 restored resistance of amiR-hmgbs transgenic plants (line # 3) to B. cinerea. Leaves were infiltrated with water containing 1 μM HMGB3 one day before B. cinerea infection. Representative disease symptoms at 3 dpi are shown in the upper panel. Data corresponding to this time point are presented as the mean ± SD (lower panel, n = 6). Asterisks indicate significant differences from the untransformed wt plants in A and B or mock-treated amiR-hmgbs transgenic plants in C (t test, P < 0.05).
Fig 5.
SA binds to HMGB3 and inhibits its DAMP activity.
A-B. SPR analysis of HMGB3’s SA-binding activity. A. Sensorgrams obtained with different concentrations of HMGB3 using a 3AESA-immobilized sensor chip. B. Sensorgrams obtained with 1 μM HMGB3 in the presence of indicated concentrations of SA using a 3AESA-immobilized chip. The signals detected from a mock-coupled control chip were subtracted. C. Photoaffinity labeling of HMGB3 using 4-AzSA. HMGB3 was incubated with 50 μM 4-AzSA in the presence of different concentrations of SA, and then exposed to UV light (30 mJ). HMGB3 labeled with 4-AzSA was detected by IB analysis with α-SA antibodies. HMGB3 stained with CBB served as a loading control. The results are expressed as a percentage of inhibition in the presence of the indicated fold excess of SA as compared to the amount 4-AzSA crosslinked HMGB3 formed in the absence of SA, which was assigned 0% inhibition. Data are the mean ± SD (n = 2). D. SA inhibited HMGB3-induced, but not Pep1-induced, MAPK activation. Leaves were collected 15 min after infiltration with water (mock) or with water containing either 1 μM HMGB3 or Pep1 in the presence of the indicated concentrations of SA. Leaves not infiltrated served as an untreated control. Phosphorylated, and thus activated, of MPK3, MPK4, and MPK6 were detected with α-pTEpY antibody. Rubisco large subunit protein stained with CBB served as a loading control. E. SA inhibited HMGB3’s, but not Pep1’s, ability to induce callose deposition. Leaves were stained with aniline blue 15 h after infiltration with water containing 0.1 μM HMGB3 or Pep1 in the presence or absence of 1 μM SA. Representative pictures are shown in upper panel. Bars = 100 μm. Data are the mean ± SD (lower panel, n = 20). F. SA inhibited enhanced resistance to B. cinerea induced by HMGB3. Leaves were infiltrated with water containing 100 nM HMGB3 with or without 1 μM SA one day before B. cinerea infection. Representative disease symptoms at 3 dpi are shown in the upper panel. Data corresponding to this time point are presented as the mean ± SD (lower panel, n = 6). Asterisks indicate a significant difference from the mock-treated leaves (t test, P < 0.05). Leaves infiltrated with water served as controls in experiments D-F.
Fig 6.
Arg50 and Lys54 are required for SA to bind and inhibit HMGB3’s DAMP activity.
A. Sensorgrams obtained with 1 μM wild type (WT) HMGB3 and the R50A/K54A mutant using a 3AESA-immobilized sensor chip. The signals detected from a mock-coupled control chip were subtracted. B. The ability of the R50A/K54A mutant to activate MAPKs was not suppressed by SA. Leaves were sampled 15 min after infiltration with water containing 1 μM R50A/K54A with or without 1 μM SA. Phosphorylated MPK3, MPK4, and MPK6 were detected by α-pTEpY antibody. Rubisco large subunit protein stained with CBB served as the loading control. C. R50A/K54A-induced callose deposition was not suppressed by SA. Leaves were sampled 15 h after infiltration with water containing 0.1 μM HMGB3 (WT) or 0.1 μM R50A/K54A with or without 1 μM SA. Representative pictures for callose staining are shown in the upper panel. Bars = 100 μm. Data are the mean ± SD (lower panel, n = 20). Leaves infiltrated with water served as mock control. Asterisks indicate a significant difference from the mock-treated leaves (t test, P < 0.05). D. SA failed to inhibit the enhanced resistance to B. cinerea induced by mutant HMGB3 (R50A/K54A). Leaves were infiltrated with water containing 0.1 μM mutant HMGB3 with or without 1 μM SA one day before B. cinerea infection. Representative disease symptoms at 3 dpi are shown in the upper panel. Data corresponding to this time point are presented as the mean ± SD (lower panel, n = 6). Asterisks indicate a significant difference from the mock-treated leaves (t test, P < 0.05). Leaves infiltrated with water served as a control.
Fig 7.
Schematic of the mechanisms through which SA inhibits the DAMP activities of HMGBs in plant and animal cells.
In plant cells, extracellular HMGB3 is recognized by a yet to be identified receptor complexed with the regulatory LRR RLKs BAK1 and/or BKK1. Recognition induces a series of pattern-triggered immune (PTI) responses, including MAPK activation (MPK3 and MPK6), defense-related gene expression (WRKYs, PR-1, and PDF1.2), and callose deposition. SA binding to extracellular HMGB3 inhibits its DAMP activity. In animal cells, extracellular HMGB1 is recognized by multiple cell surface receptors, including C-X-C chemokine receptor 4 (CXCR4) and toll-like receptor 4 (TLR4), depending on its redox state. SA inhibits chemo-attractant activity of reduced HMGB1 (hHMGB1RE), which is recognized by the cell surface receptor CXCR4 after heterocomplex formation of hHMGB1RE with C-X-C motif-containing chemokine 12 (CXCL12) [24,42]. SA also inhibits activation of Cox-2 and pro-inflammatory cytokine genes (IL-6 and TNF-α) by the disulfide-bonded form of HMGB1 (hHMGB1SS), which is mediated by heterocomplex formation of hHMGB1SS with myeloid differentiation factor 2 (MD-2), followed by signal transduction through the cell surface receptor TLR4 and transcription factor NF-κB [23,24].