Fig 1.
A high number of genes are highly expressed in the trigger hair.
Of note, the expression of 810 DEGs is elevated by a factor of at least 2 in the trigger hairs compared to all of the other tissues such as trap, rim, flower, petiole, root, and gland tissues (log2FC > 1, BH adjusted p-value < 0.001, trigger hair normalized counts > 50). Enriched GO terms for subset02 (810 genes, red bar, representing trigger hair vs. all tissues) of the intersection analysis between DEGs groups are shown (GO enrichment, BH adjusted p-value < 0.05). Underlying full raw dataset is provided in S1 Data. BH, Benjamini–Hochberg; DEG, differentially expressed gene; GO, gene ontology; UDP-glucose, uridine diphosphate glucose.
Fig 2.
Ion channels are among the most trigger hair–specific genes.
The dots represent the trigger hair Qgene|hair-value calculated by the Shannon entropy method for tissue specificity. A very low Qgene|hair-value indicates high tissue specificity. The stacked bars show the proportion to which each gene is expressed across all investigated tissues. The numbers on the red bars represent the mean FPKM expression values for the trigger hair tissue. The genes marked in bold letters represent ion channels. Shown are the top 10 highly expressed genes (trigger hair FPKM > 20) out of the bona fide trigger hair–specific genes. Genes that passed the Shannon entropy calculated Qgene|hair-value < 3.9 bits, comprising 1% of the total Laplace distribution area, and were at the same time trigger hair up-regulated DEGs at least 2-fold in the trigger hair compared to all the other tissues (log2FC > 1, BH adjusted p-value < 0.001), were considered bona fide trigger hair–specific genes. Underlying full raw dataset is provided in S2 Data. BH, Benjamini–Hochberg; DEG, differentially expressed gene; FPKM, Fragments Per Kilobase of transcript per Million mapped reads.
Fig 3.
Functional properties of trigger hair–specific KDM1.
(A) Localization of YFP::KDM1 fusion constructs in the PM of A. thaliana mesophyll protoplasts (top) and X. laevis oocytes (bottom). A merged picture of YFP fluorescence and the chloroplast autofluorescence is presented in the upper panel. The lower panel shows a magnified image from an enlarged section of an oocyte to demonstrate the YFP signal on the surface of the membrane. Representative images are shown. (B) KDM1 cDNA confers functional expression of hyperpolarization-activated currents in X. laevis oocytes. Currents of up to −15 μA were elicited in response to hyperpolarizing pulses from a holding potential of 0 mV in 100 mM KCl bath solution. (C) Inward K+ currents elicited by a KDM1 expressing Xenopus oocyte markedly increased upon acidification of the external solution (Vm = −80 mV). The dotted line represents 0 current. (D) Comparison of the ISS at −150 mV of the indicated KDM1 channel mutants. ISS for each mutant were normalized to −150 mV at pH 4, which displays 100% activity. Note that mutants containing H147S are less affected by an external pH change at these voltages, whereas the other histidine mutants mediate WT-like K+ influx (n ≥ 6; mean ± SE). (E) The V1/2 of KDM1 WT and the H147S mutant was plotted against the applied pH values and fitted as described in the Materials and methods section with the following parameters: for mutant and WT, a = 0.6 and RT/F = 25.2 mV resulted in Vs = RT/(aF) = 42mV; for WT, V1/2_inf = −177.048 mV, pKO = 5.1328, pKC = 4.04; and for H147S, V1/2_inf = −153.632, pKO = 4.6285, pKC = 4.1914. The V1/2-pH curve of the mutant is compressed compared to that of the WT pointing to fundamental alterations in the pH-sensing process. (n ≥ 14; mean ± SE). The full raw dataset and the statistical analysis is provided in S3 Data. PM, plasma membrane; SE, standard error; WT, wild-type; YFP, yellow fluorescent protein.
Fig 4.
Cs+ blocks KDM1 and trigger hair–dependent excitability: Cs+ reduces the AP restoring force.
(A) KDM1-mediated steady currents (normalized to −150 mV and 30 mM K+) at 100 mM or 30 mM external K+ with or without 30 mM Cs+. The ISS-V curves show inward-directed potassium currents, which ceased when Cs+ was applied (n = 17; mean ± SD). Under the same conditions, the efficiency of the cationic blocker appeared to be reduced, when Cs+ was replaced by TEA+ or Ba2+ (c.f. S9I Fig). (B) Trigger hairs were immersed in 1.5% low melting agarose without (left/control) and with (right) 30 mM Cs+ for 100 minutes. The surface potential electrode was inserted into the trap tissue for AP recordings. (C) Frequency-dependent surface potential recording of a D. muscipula trap as shown in (B). Incomplete AP block was measured after 50 minutes incubation in Cs+ (red). Twenty mechanical stimulations (arrows/¼ Hz) were applied to the trigger hairs. Under control conditions (black), 10 simulations were translated into APs (green arrows), while 10 (purple arrows) were not. Cs+ treatment reduced AP firing to just 4 (red trace). (D) Plot of the number of APs evoked in traps treated as defined in (C). Fifteen traps were each stimulated 20 times. The bottom and top of the boxes denote the first and third quartiles, respectively, the middle line is the median, and the whiskers are the most extreme values within 1.5× the interquartile distance below the first or above the third quartile. Cs+ treatment reduced the number of fired APs significantly (p = 6.0 × 10−8; F1,29 = 134.3, 1-way ANOVA, n = 15). The full raw dataset and statistical analysis is provided in S3 Data. ANOVA, analysis of variance; AP, action potential; SD, standard deviation.
Fig 5.
Reduced K+in channel activity increases recovery lag time after AP firing.
(A) A mechanical stimulus (green arrow) initiates an AP (upper panel) during which the accompanied ion and water fluxes reduce the cell turgor (lower panel). In the following recovery phase, the initial condition is slowly reestablished (inset shows the recovery on a larger timescale). Reduced activity of K+in channels significantly slows down the recovery process (red curve). (B) When the THs are stimulated with low frequency (green arrows), the interval is insufficient for a full recovery of the cell. Over time, however, a new working range is established that is still above the turgor threshold (green) as seen in the inset. (C) Juvenile trap with the corresponding juvenile TH and adult trap with the corresponding adult and electrically excitable TH. (D) Normalized expression values of KDM1 quantified by qPCR in trap and THs in juvenile and adult stages (mean normalized to 10,000 actin (± SE), n = 3–6; t test between whole adult trap and adult TH p-value = 0.00038). The proportion of KDM1 expression within the dissected adult TH shows a dominant (up to 70%) expression level in the base of the TH (containing the sensory cells), (n = 3; mean ± SE). The full raw dataset and statistical analysis is provided in S3 Data. (E) Increasing the frequency of the stimulus causes the turgor to fall below the threshold. In this condition, a mechanical stimulus may not initiate an AP and sporadic dropouts occur (purple arrows, purple dots in the enlarged view in the inset). (F) High-frequency stimulation in conditions with reduced K+in channel activity results in larger dropout intervals. Initially, however, as long as the cell still has reserves, APs are initiated as in the control case. AP, action potential; qPCR, quantitative polymerase chain reaction; SE, standard error; TH, trigger hair.