H+- and Na+-elicited swift changes of the microtubule system in the biflagellated green alga Chlamydomonas

The microtubule cytoskeletal system is integral to diverse cellular processes. Although microtubules are known for dynamic instability, the system is tightly controlled in typical interphase animal cells. In contrast, diverse evidence suggests that the system is mercurial in the unicellular fresh water green alga, Chlamydomonas, but intense autofluorescence from photosynthesis pigments has hindered the investigation. By expressing a bright fluorescent reporter protein at the endogenous level, we demonstrate in real time discreet sweeping changes in algal microtubules elicited by fluctuation of intracellular H+ and Na+. These results suggest disparate sensitivity of this vital yet delicate system in diverse organisms; and illuminate how pH may drive crucial cellular processes; how plants respond to, and perhaps sense stresses; and how many species could be susceptible to accelerated changes in global environments.

fluorescent protein, NeonGreen (NG) that is 2.7 X brighter than EGFP (Shaner et al., 2013), and 76 the relative abundant plus end-binding protein, EB1, as the NG carrier (Harris et al., 2016). 77 EB1 plays central roles in eukaryotes (Su et al., 1995; reviewed by Akhmanova and 78 Steinmetz, 2010; Kumar and Wittmann, 2012). Its N-terminal domain preferentially binds to the 79 lattice among tubulins at the plus end of MTs, whereas its C-terminal domain can associate with 80 a wide array of proteins. The two domains operate in concert to accelerate MT dynamics (Rogers 81 et What causes the switch remains uncertain. 89 Using EB1-NG as a reporter, we captured in real time unexpected changes in EB1-NG 90 patterns and MT dynamics signaled through H + and Na + . The remarkable sensitivity and the 91 distinct responses in wild type (WT) cells and mutants shed critical insight on the divergence of 92 the MT system, pH regulated processes and the vulnerability of organisms subjected to 93 environmental stresses.

EB1-NG reports remarkable sensitivity of the MT system in Chlamydomonas 98
The The birth of new comets from BBs appeared stochastic. We did not measure the birth 118 rates, hindered by substantial fluctuations and the narrow apical area. Instead we analyzed comet 119 length and speed from the side view. Line scans along the lengths of comets show the typical 120 feature of EB1 comets -the brightest spot corresponds to the area where tubulins are primarily at 121 the transitional state, slightly behind the leading edge of plus ends with GTP-tubulins ( Figure  122 1c). The distribution of comet speeds shows that MT growth rates varies nearly two folds ( Figure  123 1d). The dataset from cells in the TAP medium (black bars) skews toward the slow end relative 124 to the Na + /HEPES dataset (hatched bars). The average velocities are signifantly different (Mann-125 Whitney U test, p < 0.001), at 5.8 ± 0.26 and 7.9 ± 0.42 µm/sec respectively (Figure 1e), which 126 are within the normal range measured in diverse eukaryotic cells (Harris et al., 2016). 127 Curiously, in some long recordings, comets suddenly gave way to a bird cage-like pattern 128 (Figure 1f1,  comets (t4 and t5). These observations demonstrate that Chlamydomonas MT system is highly 137 mercurial; and suggests that excitation illumination creates a condition that is unfavorable for 138 MT dynamics, but is reversed in the dark. As illumination opens channelrhodopsins that conduct 139 Time-lap fluorescent images were taken 10 sec apart from the top (T), side (S) and rear (R) of cells resuspended in the TAP culture medium. EB1-NG appeared like typical comets (arrowheads), emerging from the BB area, coursing along the contour of the cell body and then vanishing as approaching the rear end. The frame rate is 1 frame/sec. (c) Normalized line scans along the length of MT plus ends showed a similar EB1 intensity profile in the TAP medium (n = 18 comets from 6 cells) and the Na + /HEPES buffer (n=11 comets from 3 cells). The position with peak intensity was designated as 0. The value was negative toward plus end; positive toward BBs. AU, arbitrary unit of fluorescence intensity. (d) The distribution and (e) the mean and the SEM of EB1 comet speed in the TAP medium (n = 36 comets from 6 cells in 6 recordings) and 5 mM Na + /HEPES buffer (n = 22 comets from 3 cells in 3 recording) are significant different (Mann-Whitney U test, P < 0.001). (f) Altered MT patterns during fluorescence microscopy. The EB1 comet pattern occasionally switched to a bird cage pattern (f1). Comets returned while the bird cage receded in ~ 1 min. In flattened cells that were compressed by the cover slip gradually, both MTs and comets became explicit (f2, top panel). Comets disappeared after ~ 100 sec (bottom panel, t2), but returned after illumination was switched off for 30 sec (t3). The process was repeatable after another 100 sec illumination and then another light off period (t4 and t5). The alternate white and black bars illustrate the scheme of alternate illumination and dark periods. Scale bars, 5 µm. a number of cations and Cl - (Nigel et al., 2002; ; reviewed by Hegemann and Berthold,  140 2009), we hypothesize that fluctuations of electrolyte concentrations modulate the MT system in 141 Chlamydomonas. Considering the light sensitivity, we elected to use wide field microscopy and 142 minimal light intensity to test this.

144
Sequential changes in the MT system elicited by a short HA pulse and subsequent wash 145 We first used the well-defined pH shock, recording EB1-NG signals in cells exposed to HA in 146 two complementary devices, perfusion chamber and diffusion chamber (Figure 2a) shortening speed was not analyzed because of few motionless cells and few shortening MTs with 180 a definitive plus end. Tubulin reporters will be more appropriate for shortening analysis. flagella (arrow) were amputated (right panel). EB1 signals remained at the BB area but were 186 static (black arrowhead). Contrary to deflagellation within seconds upon HA perfusion (Wheeler 187 et al., 2008), the deflagellation in the diffusion chamber takes more than one minute due to 188 gradual acidification. Thus when cells are exposed to HA, shank binding increases, comets 189 disappear, endwise resorption becomes evident and then flagella become amputated. The 190 sequential events occurring in the diffusion chamber are summarized in figure 2f.

191
Lowering intracellular pH elicits Ca 2+ influx, whereas Ca 2+ prevents MT formation and 192 promotes MT disassembly (Weisenberg, 1972  Schematics depicting an open-ended perfusion chamber (left panels) and a diffusion chamber (right panels) for capturing the HA-induced rapid changes. (b) A 10-µl aliquot of cells resuspended in the TAP medium was placed in a perfusion chamber. The images (b1, 2) were captured before and after perfusion with 20 mM HA/TAP (pH 4.5, t=0, black arrow). The following recordings (b3-5) captured the events right after the TAP medium (pH 7) was injected to wash away HA (t=0, clear arrow). B3 is the first clear image after fluid and cells stopped flowing. Comets already disappeared within 87 seconds after HA perfusion. They started emerging 43 sec after wash. (c) The process preceding HA-induced disappearance of EB1 comets in diffusion chambers. A 40-µl aliquot of cells resuspended in HEPES was placed in a diffusion chamber encircled by Vaseline, under the coverslip and an objective lens. HA was injected to the other side of the chamber and diffused toward cells that were being imaged. During the gradual acidification process, both comets (white arrowheads) and shank binding MTs (black arrowheads) were evident first and then both patterns vanished. (d) Time lapse images and kymographs revealed endwise resorption of EB1-decorated MTs (white arrowheads). (e) Comets (white arrowheads) in the cell body vanished first before the excision of flagella (arrows). Following deflagellation, EB1 diffused away from the tip. EB1 signals remained at BBs but was static (black arrowhead homeostasis. We reasoned that altering HA treatment might also change other cations such as 209 Na + or K + . To test this, we extended HA exposure -resuspending cells in pH3, 10 mM 210 HA/double distilled water (ddw) for 5 mins. As expected, EB1 patterns were absent except for 211 the static signal at the BB area (   In an attempt to decelerate pH-induced resorption in the acidification phase, we took advantage 259 of a tubulin mutant, tub2. A missense mutation near the colchicine binding site in β-tubulin 260 increases MT stability since tub2 cells are colchicine-resistant, and have more acetylated MTs 261 (Schibler and Huang, 1991). In the EB1-NG transgenic tub2 cells, the comet pattern (Figure 4a  replicates the light-induced sporadic transient appearance of the bird cage pattern in WT cells 295 (Figure 1, f1). Changes elicited by 7.5 mM HA partially mimics HA-induced responses in 296 diffusion chambers (Figure 2b-e). 297

The long-lived EB1-decorated MTs are due to the rise of intracellular [Na + ] but not [K + ] 298
To identify the ion causing the formation of static thick EB1-decorated MT bundles in cells 299 recovering from HA bath, HA-bathed cells were washed with different solutions. Interestingly, 300 as shown by time-lapse images separated by a 10-second interval, EB1-decorated MT fibers 301 formed only if the wash solution contained Na + , such as the NaOH-buffered HEPES (5 mM Na + ) 302 or 5 mM NaCl in ddw (Figure 5a, top two panels). On the other hand, comets resumed profusely 303 if the wash solution lacked Na + , such as the KOH-buffered HEPES buffer, 5 mM KCl in ddw or 304 plain ddw (bottom three panels). Therefore, Na + accounts for the reformation of thick, long-lived 305 static MTs in the recovery phase of HA-bathed cells.

306
Since Na + has low permeability compared to K + ( in Na + /EGTA for 5 min (Figure 5b, left panels). Thick MTs appeared static after 10 mins. In 315 contrast, cells resuspended in KOH-buffered EGTA had vibrant comet activities (right panels).

316
The static EB1-binding MTs in Na + /EGTA treated cells were also cold-resistant. 317 As The changes of algal MT system elicited by intracellular acidification are swift, stunning and 357 novel (Figure 2b-e, 3a and 4). Among heightened shank binding, diminished comet activity, 358 paused MT growths, and MT shortening, the bird cage pattern of shank binding is the most 359 sensitive, elicited reliably by 5 mM HA (Figure 4d). They are unlikely signaled through cell 360 death pathways, since these changes are reversible, even after 5 min HA bath (Figure 2b, 3). The 361 recovery occurs within a minute, either completely ( Figure 2b) or protractedly (Figure 3e), 362 depending on exposure and wash procedures. Extracellular Ca 2+ is not required for the HA-363 induced changes, since they still occur after K + /EGTA treatment. However, we cannot rule out 364 the involvement of Ca 2+ released from intracellular storages and other signaling pathways. 365 Although the pH of [HA]ex that triggers these changes is ~3, we expect that the resulting 366 intracellular pH is close to or higher than 6.3. HA-induced changes in diffusion chambers appear 367 before deflagellation (Figure 2e) that occurs at pH 6.3 (Wheeler et al., 2008;Braun and 368 Hegemann, 1999). Consistent with this, the bird cage pattern could be transiently triggered 369 merely by illumination that may open H + -selective channelrhodopsin (Figure 1f1). Conversely, 370 comets return to compressed cells when illumination is turned off for 30 seconds (Figure 1f2); or 371 Figure 5. Na + -dependent changes of the MT system. (a) MTs in cells were largely frozen after 5-min 10 mM pH3 HA bath and 3 min in the wash solution, such as 5 mM pH7.4 Na + / HEPES buffer or 5 mM NaCl solution (black arrowheads). In contrast, growing MTs with a comet (white arrowheads) returned if the wash buffer lacked Na + , such as 5 mM K + /HEPES buffer, 5 mM KCl solution, or the double distilled water (ddw). (b) Thick MTs in cells resuspended in 21 mN Na + /EGTA for 5 min or 10 min (left panel), contrary to comets in cells in 21 mM K + / EGTA (right panel). Thick MTs were still growing after 5 min incubation but static after 10 min incubation (c) High [Na + ]ex, without preexposure to HA, was sufficient to alter comet patterns. Contrary to typical comets in cells resuspended in the HEPES buffer with 5 mM Na + , long comets were thick in cells resuspended in 55 mM Na + for 5 min (panel I). Normalized linescans confirmed little tapered intensity (panel II, n=36 comets from 11 cells in 5 mM Na + ; n=13 comets from 4 cells in 55 mM Na + ). As shown in the range of speed (panel III) the long comets were moving, and the mean speeds of short and long comets were significantly different (panel IV, n=51 from 11 cells in 5 mM Na + ; n=18 from 4 cells in 55 mM Na + ) (P < 0.05). (d) Two representative cells after 5 min in 150 mM Na + /HEPES. Some cells still retained a few thick MTs (cell I). Some only had static EB1 signals at the BB area (cell II, arrowhead). Scale bars, 5 µm.
return within ~ 45 sec after HA is washed away (Figure 2b). These observations strongly suggest 372 that a slight imbalance of pH homeostasis is sufficient to elicit changes in algal MT system. 373 Although pH affects proteins' ionization and thus their functions and protein-protein 374 interactions in general (Hepler, 2016), we speculate that the MT system is particularly sensitive 375 to declining pH because of the acidic pI of tubulins and EB1. For example, the respective pI of 376 Chlamydomonas α-tubulin, β-tubulin and EB1 is 5.01, 4.82, and 5.7. A decrease of pH from the 377 resting level will make these proteins less negatively charged, especially at their C-terminal leading to nearly immediate appearance of the bird cage pattern in tub2 cells (Figure 4a-b) or in 382 WT cells exposed to 5 mM HA ( Figure 4d). As pH descends further, additional changes in 383 protein conformation may inhibit the growth of MTs and EB1 binding to plus ends, leading to 384 comet reduction or ultimate disappearance. 385 These HA-induced changes explain long standing questions regarding pH variations. For 386 example, the pollen tube tip has overlapping regions. The MT zone, in particular, lags behind the 387 F-actin, Ca 2+ and acid zone (Gibbon and Kropf, 1994; reviewed by Hepler, 2016). Likewise, a 388 basic shift directs MT-supported fertilization processes of sea urchins, whereas depressing pH 389 inhibits the processes and triggers MT disassembly (Schatten et al., 1985). This and an increase 390 of 0.3-0.5 pH unit in mitosis inspired the pH clock hypothesis for cell cycle control (Gagliardi 391 and Shain, 2013). In line with this, EB1 preferentially binds to MT plus ends in arrested mitotic 392 phase extract of Xenopus oocytes, but uniformly decorates MTs in interphase extract (Tirnauer et 393 al., 2002) perhaps with a lower pH, analogous to HA-induced bird cage pattern (Figure 4c). 394 Given the role of EB1 in recruiting effector molecules and the swiftness of pH-induced changes 395 in MTs and EB1 patterns, tuning pH may indeed control cell cycle at least for certain organisms. continues rising, comets lengthen; cortical MTs undergo ectopic nucleation, splitting, bundling, 401 decelerate and stop eventually (Figure 3 and 5). Although it takes longer to elicit Na + responses, 402 this is likely due to low Na + permeability. Likewise, Ca 2+ permeability is tightly controlled. As 403 such [Na + ]in and [Ca 2+ ]in cannot be adjusted as nimbly as [H + ]in. 404 The degree of changes correlates with [Na + ]in. These changes aggravating with time in 21 405 mM Na + /EGTA (Figure 5b) indicates that they are occurring before [Na + ]in reaches 21 mM, 406 which is made possible by EGTA treatment. As EGTA chelates Ca 2+ , this further rules out the 407 involvement of extracellular Ca 2+ . Similarly, Na + -dependent responses emerge within ~45 408 seconds once HA bath is replaced with various solutions containing only 5 mM Na + (Figure 5a), 409 suggesting accelerated rise of [Na + ]in due to the activity of Na + / H + exchangers (Pittman et al.,410 2009). On the other hand, by simply relying on limited passive diffusion through the normal 411 plasma membrane, 55 mM [Na + ]ex is sufficient to change comet length but only slows down 412 growth rate slightly (Figure 5c). Based on the incremented responses, rather than all-or-none 413 responses to a threshold, we speculate that algal MT system is also sensitive to [Na + ]in, perhaps 414 in a linear manner. Contrary to [Na + ]ex, raising [K + ]ex had no evident effect (Figure 5 a-c). This is 415 reasonable, given high [K + ]in, ~ 70 mM in Chlamydomonas cells (Malhotra, 1995). It highlights 416 the selective sensitivity of the algal MT system to Na + and rules out mere ionic effects. One 417 interesting possibility is that Na + binds to particular sites in algal tubulins, analogous to Ca 2+ 418 binding sites in mammalian tubulins (Solomon, 1977;Serrano et al., 1986).

420
Common changes elicited by high extracellular HA, Na + and Ca 2+ 421 EB1 signals largely vanish at 150 mM [Na + ]ex (Figure 5d), as in 10 mM HA (Figure 2c, 3a) and 422 75 mM Ca 2+ (Figure 2g) except residual static signals at the BB area. They are likely caused by 423 synergistic disassembly and paused new growth, and an immotile EB1 population underneath 424 BBs respectively (Yan et al., 2006;Pedersen et al., 2003). The similar outcomes caused by 425 distinct ions and obvious shrinkage of the cell body at even higher concentrations of Na + and 426 Ca 2+ suggest that hypertonicity is involved. We envisage that high concentration responses could 427 be caused by one cation exceeding a threshold concentration; and/or simultaneous rises of 428 multiple electrolytes as H2O moves out of cells. Hypertonicity may evoke additional pathways. 429 The capture of endwise resorption only in low concentration conditions (Figure 2d and  430 4d) suggests that increased concentrations of these ions will heighten shortening-the incidence 431 and/or speed. This is reminiscent to high Ca 2+ effects. In vitro, Ca 2+ blocks MT formation 432 (Weisenberg, 1972 in HA-bathed algae that takes ~55 mins (Figure 3f). 448 The electrolyte sensitivity of algal MT system is contrary to the perceived stable MT 449 system in interphase mammalian cells (Lieuvin et al., 1994)  The sensitivity of algal MT system to Na + is consistent with enlarged or clustered algal 460 cells cultured in high salt media (Takouridis et al., 2015). This could be caused by anomalies in 461 the MT-supported processes in the cell cycle, such as mitosis and trafficking-dependent release 462 of hatching enzymes (Kubo et al., 2009). Yet this fresh water green alga has several strategies to 463 adapt to salinity (Perrineau et al., 2014), such as glycerol production (Husic and Tolbert, 1986), 464 switches in gene expression (Gao et al., 2016), sexual reproduction and mutations (Takouridis et 465 al., 2015). Salinity adaptation and the incredible H + and Na + sensitivity of algal MT system that 466 bears semblance to that in both animal and plant cells demand a fresh look at how environmental 467 stresses affect diverse organisms. 468 One is ocean acidification by anthropogenic CO2 (Raven et al., 2005). triggers distinct changes of Chlamydomonas MT (Figure 4d). Notably, dictated by CO2 474 chemistry, the ratios of permeant CO2 and H2CO3 to non-permeant ionic forms will increase 475 further as solutions acidify, aggravating intracellular acidification. Therefore, intracellular 476 acidification could be equally, if not more, insidious to marine species that are not equipped to 477 cope with this stress; and may poise to shape aqueous landscapes and drive evolution (Cannon et 478 al., 1985).

479
The other is salt stress and osmostress caused by draught, which is exacerbating due to  Typically cells were resuspended in solutions for 5 min unless indicated otherwise. An 526 aliquot of 5 µl cell suspension was placed on a slide and then covered by an 18 X 18-mm 2 cover 527 slip. The edges were sealed with nail polish before imaging. For compression experiments, a 3 µl 528 aliquot of cell suspension was placed on a glass slide and then covered with a 22 X 22-mm 2 529 cover slip. Cells became gradually compressed by the coverslip as evident by flattened cell body. 530 For pH pulse in a perfusion chamber, an aliquot of 10 µl cells in the TAP medium was placed on 531 a cover slip pre-coated with 5 µl 0.001% poly-L-lysine. The cover slip was then inverted to 532 assemble a perfusion chamber as shown in Figure 2a. process was recorded in two consecutive live-stream clips. For this long recording duration, 535 excitation light intensity was reduced to 25% with a neutral density filter. For HA pulse in a 536 diffusion chamber, 40 µl cells in 5 mM pH7.4 Na + /HEPES was placed at one side of a diffusion 537 chamber underneath a 40X objective lens (Figure 2a). A live-streaming video was recorded 538 following the injection of 20 µl 100 mM pH2.8 HA through the Vaseline wall to the opposite 539 side of the chamber. For HA bath, a cell pellet from 50 µl liquid culture was resuspended in 50 540 µl 10 mM pH3 HA. An aliquot of 10 µl cell mixture was placed on a cover slip. The cell-loaded 541 cover slip was inverted to create a perfusion chamber. After a total 5-min exposure to HA, HA 542 was flushed away with an aliquot of 200 µl-indicated fluid and then a video was recorded. To 543 test MT cold lability after recovery from HA bath, a perfusion chamber with treated cells was 544 chilled by ice for 3 minutes. A video was taken immediately afterwards, ~ 20 seconds after the 545 chamber was removed from ice. Each treatment was repeated at least twice in each experiment.

546
Individual experiments were repeated independently 3 times at least. 547 548 Image Analysis 549 To measure EB1 comet speed, a 40-second substack containing side views of cells were first 550 made by the open source image process software, ImageJ (https://imagej.nih.gov/ij/index.html); 551 and individual comets were analyzed with a Matlab-based particle tracking software, 552 plusTipTracker (Applegate et al. 2011). In each cell that maintained completely quiescent for the 553 tracking period, all tractable comets which transverse at least one third of the cell length were 554 analyzed. The numbers of qualified cells and comets from numerous recordings were indicated. 555 To generate line scans of EB1 intensity at microtubule plus ends, a line tool in ImageJ was used 556 to measure gray values along the length of comets. Relative fluorescence intensity was 557 normalized after calculation by subtracting a background gray value measured next to the comet 558 with the line tool. Histograms were generated with the Microsoft program, Excel. Kymographs 559 were generated with an ImageJ plug-in multiple kymograph 560 (https://www.embl.de/eamnet/html/body_kymograph.html). 561 562 Statistical Analysis 563 All data are given as mean±SEM (standard error of the mean) and analyzed with Sigmaplot 13.0 564 (Systat Software, Inc., San Jose, CA). Sample sizes for comet speed measurement are limited by 565 the fact that few cells are entirely quiescent, which is necessary for digital tracking. 566 567 Acknowledgement 568 This work is supported by Marquette University Startup for P. Yang. 569 570 Competing Interests 571 The authors declare that no competing interests exist. 572 573 Supplemental data 574 Video 1-1 (for Figure 1b) EB1-NG comets in WT cells. 575 Video 1-2 (for Figure 1f1) Transient bird-cage pattern in WT cells that occurred sporadically 576 during imaging. 577 Video 1-3 (for Figure 1f2) Disappearance and return of comets in compressed cells following 578 alternate periods of illumination and darkness. 579 Video 3 (for Figure 3a) WT cells in Na + /HEPES after 5-min HA bath. 580 Gao, X., Zhang, F., Hu, J., Cai, W., Shan, G., Dai, D., Huang, K., and Wang, G. (2016). 627 MicroRNAs modulate adaption to multiple abiotic stresses in Chlamydomonas reinhardtii. 628 Sci Rep 6, 38228. 629