High temporal resolution reveals simultaneous plasma membrane recruitment of the TPLATE complex subunits

The TPLATE complex (TPC) is a key endocytic adaptor protein complex in plants. TPC contains six evolutionary conserved subunits and two plant specific subunits, AtEH1/Pan1 and AtEH2/Pan1, which are not associated with the hexameric subcomplex in the cytoplasm. To investigate the dynamic assembly of the octameric TPC at the plasma membrane (PM), we performed state-of-the-art dual-color live cell imaging at physiological and a lowered temperature. Our data show that lowering the temperature slows down endocytosis and thereby enhances the temporal resolution of the differential recruitment of endocytic components. Under both normal and lowered temperature conditions, the core TPC subunit TPLATE, and the AtEH/Pan1 proteins, exhibited simultaneous recruitment at the PM. These results, together with our co-localization analysis of different TPC subunits, allow us to conclude that in plant cells, TPC is not recruited to the PM sequentially but as an octameric complex. One sentence summary Lowering the temperature increases spatiotemporal resolution of protein recruitment at the plasma membrane.

shift was transmitted to the seedling almost instantaneously (Fig. S1).

95
To examine the capacity of lowering the temperature to slow down endocytosis, etiolated hypocotyl cells 96 expressing endocytic markers CLC2-mKO and TPLATE-GFP the temperature was shifted during image 97 acquisition and no obvious defect in the recruitment was observed. Density analysis of early-arriving 98 (TPLATE and TML) as well as late arriving DRP1A furthermore showed that lowering the temperature did 99 not visually affect the PM recruitment of these endocytic markers while it prolonged their lifetime at the PM 100 instantly (Figure 1 and Figure 2A).

101
To further evaluate how lowering the temperature affects the dynamic behavior of endocytic proteins, we  (Fig. S3). These results together, show that lowering the temperature in etiolated seedlings using 114 our experimental setup slows down CME efficiently and rapidly.

116
Lowering the temperature enhances the temporal resolution of plasma membrane recruitment.

117
Having established the effect of lowering the temperature on endocytic dynamics in Arabidopsis cells, we 118 then examined whether we would be able to enhance the temporal difference between an early (TPLATE)

128
We performed a similar experiment comparing TPLATE and CLC2, which were previously shown only to   promoting autophagy (Wang et al., 2019), suggest that TPC could be in essence a hexameric complex, 140 which temporarily gains two additional subunits. To reveal the spatiotemporal relationship among the TPC 141 subunits, we visualized their dynamic behavior at the PM using multiple dual-color labeled lines. To 142 evaluate the functional association of TPC subunits at PM, we crossed complemented mutant lines of core 143 subunits (TPLATE, TML and TWD40-1) and generated double complemented homozygous mutant lines.

144
We also combined complemented mutants of the core subunit TPLATE with AtEH1/Pan1 or AtEH2/Pan1 145 and generated the respective double complemented, double homozygous mutant lines. As a first 146 indication for differential recruitment, we monitored the steady-state percentage of co-localization between 147 the markers using a flattened projection of five consecutive frames. Comparing an early endocytosis 148 marker such as TPLATE, with a late marker such as DRP1A revealed that the differential arrival between 149 those makers leads to a substantial non-colocalizing fraction (roughly 40%, Fig. 4A).

155
In Arabidopsis seedlings, AtEH1/Pan1 and AtEH2/Pan1 have so far been shown functionally associated 156 with TPLATE at autophagosomes and to be delivered to the vacuole after carbon starvation and Conc A

159
Similar to the combination of the core subunits, both AtEH1/Pan1-mRuby3 and AtEH2/Pan1-mRuby3 160 showed a very high degree of colocalization with TPLATE-GFP at the PM ( Fig. 4D and E). To exclude the 161 possibility that TPLATE and AtEH/Pan1 foci overlapped due to the high density of endocytic foci, we 162 compared the colocalization percentage between TPLATE and AtEH1/Pan1 on images where the 163 TPLATE channel was rotated 90 degrees. This rotation resulted in only around 10% of foci colocalizing 164 ( Fig. 5F), confirming that the observed high degree of colocalization is not random. These results also 165 suggest a tight association between TPLATE and the AtEH/Pan1 proteins at the PM.

167
AtEHs/Pan1 and the core TPC subunits are simultaneously recruited to the PM

168
To further investigate TPC assembly at the PM at the level of the different subunits, we compared the 169 recruitment and disassociation behavior among pairs of TPC subunits at 25 and 12 °C. As a benchmark,

170
we compared the behavior of two closely related TPC subunits, TPLATE and TML, which by homology to 171 other adaptor complexes are presumed to be part of the core of TPC.
6 Time-lapse imaging and kymograph analyses showed that dual-labeled TML-GFP and TPLATE-TagRFP 173 foci appear and disappear simultaneously at the PM when imaged at 25 °C ( Figure 5A, 5B and S6).

174
Lowering the temperature maintained this simultaneous appearing and disappearing behavior. However,

175
there were very small changes in average lifetime, which were found to be negligible by their effect size 176 (expressed as the Hedge's g value). These small changes could be explained by there being more signal 177 noise and differential bleaching effects of the fluorophores during the image acquisition at the lower 178 temperature conditions ( Figure 5A-5C and S6).

179
We next compared the PM appearance and disappearance behavior of AtEH/Pan1 and TPLATE.

187
Clathrin-mediated endocytosis is a highly dynamic process. It is best understood in mammalian and yeast

197
Here, we report that lowering the temperature using a microfluidics-based on-slide approach can be used  Our data also show that the temporal resolution of the differential recruitment of endocytic players could 206 be enhanced at lower temperatures, which was most apparent between early and late arriving endocytic 7 proteins. Lowering the temperature also enabled us to slightly enhance the temporal difference between 208 two early arriving proteins, TPLATE and CLC2, indicating that our approach worked. Given the 209 temperature effect on the dynamics of plant processes, awareness should be increased to perform live-210 cell imaging in plants under accurate temperature control. No doubt this will substantially increase the 211 robustness and reproducibility of future image data collection.

212
The TPC is an evolutionary ancient adaptor complex, which was not retained in yeast or animal cells (Hirst imaging times under low temperature conditions failed to generate sufficient SNR images for automatic 248 quantification. Actually, the only combination which yielded acceptable data when automatically quantified 249 was our TPLATE-DRP1A combination, as the DRP1A is actually 35S-driven.

250
We therefore generated kymographs of our dual-color endocytic spinning disc movies and measured 251 the dwell-times of the proteins manually. We aimed at measuring several hundred events per experiment 252 and combined measurements of independent persons to achieve an unbiased assessment of the data.

253
We were unable to identify any recruitment of AtEH/Pan1 proteins at the PM which was independent of 254 another TPC subunit. Our findings therefore show that the TPLATE adaptor protein complex is likely 255 recruited to the PM as an octameric unit, both at normal and at lowered temperature conditions. We can 256 however not exclude the possibility that the differential recruitment between the individual TPC subunits 257 are too dynamic to monitor under our limitating, one-second temporal resolution, conditions. However, due 258 to the fact that we were able to detect differences in the CLC2 and TPLATE dynamics, which are of the 259 order of a few seconds, we hypothesize that TPC is recruited as an octameric unit at the PM during CME.

260
The continuous development of novel microscopy as well as labeling techniques will help to overcome our

306
The CherryTemp system (Cherrybiotech), which enables ultra-fast temperature shifts between 5 °C and 307 45°C was used to modulate and maintain the temperature during the spinning disk imaging 308 (https://www.cherrybiotech.com/heater-cooler-for-microscopy). Prior to imaging, a single etiolated seedling 309 was laid between two coverslips with 1/2 strength MS liquid medium and incubated with the CherryTemp

314
Prior to imaging, whole 6-day-old Col_0 seedlings were incubated with 4 μM FM4-64 (Invitrogen) solution 315 in 1/2 strength MS liquid medium between 2 coverslips at 25 °C or 12 °C for 30 minutes.    As a control, one of the channels was rotated 90°C (no interpolation) and analyzed similarly as described  The results were analyzed with the estimation method to calculate mean, mean differences, confidence 394 intervals, and Hedges' g (Claridge-Chang and Assam, 2016; Ho et al., 2019). 95% confidence intervals for 395 the mean differences were calculated using bootstrap methods (re-sampled 5000 times, bias-corrected, 396 and accelerated). Effect size was measured using Hedges' g and accordingly to the standard practise is 397 referred as 'negligible' (g < 0.2), 'small' (0.2 < g < 0.5), 'medium' (0.5 < g < 0.8) or 'large' (g > 0.8) (Cumming, 398 2012). Hedges' g is a quantitative measurement for the difference between mean indicating how much two 399 groups differ from each other, if Hedges' g equals 1, the two groups differ by 1 standard deviation. R                     (4μM, 30mins) in Col-0 root epidermal cells at both temperatures. The red circle represents the mean. The p-value was calculated by the Welch two sample T-test. The mean difference with the bootstrap 95% confidential interval (green circle and green line) is shown on the right side of the plot. Hedges' g value is a standardized effect size. n represents the number of measurements (n = 43 cells for 25 °C and n = 48 cells for 12 °C) from 11 individual roots respectively. Scale bar, 10 µm. The green circle in the bottom graph represents the paired mean difference with the bootstrap 95% confidential interval (green line). Hedges' g value is a standardized effect size. n represents the number of events.