bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

KCH kinesin drives nuclear transport and cytoskeletal coalescence for tip cell growth

Moé Yamada and Gohta Goshima# Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan #[email protected]

Long-distance transport along including organelles, proteins, and RNA, microtubules (MTs) is critical for are transported to their appropriate intracellular organisation. In animals, positions where they specifically function in antagonistic motor proteins kinesin response to internal and external signals. (plus end-directed) and dynein (minus Although it had been believed that plants end-directed) drive cargo transport. In predominantly utilize actin and myosin to land plants, however, the identity of move cellular components, recent studies motors responsible for transport is have uncovered the prevalence of poorly understood, as genes encoding microtubule (MT)-dependent transport as cytoplasmic dynein are missing. How well (Kong et al., 2015; Miki et al., 2015; other functions of dynein are brought Nakaoka et al., 2015; Zhu et al., 2015; about in plants also remains unknown. Yamada et al., 2017). However, a unique Here, we show that a subclass of the feature of plant motor systems is that the kinesin-14 family, KCH—which can also genes encoding cytoplasmic dynein, the bind actin—drives MT minus sole MT minus end-directed transporter in end-directed nuclear transport in the animals, have been lost during plant moss Physcomitrella patens. When all evolution. Moreover, dynein function is not four KCH genes were deleted, the limited to cargo transport, as a variety of nucleus was not maintained in the cell fundamental cellular processes requires centre but was translocated to the dynein, such as MT-based force apical end of protonemal cells. In the generation at the cortex (Grill and Hyman, knockout (KO) line, apical cell tip 2005; Gonczy, 2008; McNally, 2013), growth was also severely suppressed. MT-MT crosslinking (Ferenz et al., 2009; KCH was localised on MTs, including at Tanenbaum et al., 2013), and MT-actin the MT focal point near the tip where crosslinking (Grabham et al., 2007; Perlson MT plus ends coalesced with actin et al., 2013; Coles and Bradke, 2015). filaments. MT focus was not persistent However, how plants execute these in KCH KO lines, whereas actin functions without dynein remains destabilisation also disrupted the focus unanswered. despite KCH remaining on unfocused The moss Physcomitrella patens is an MTs. Functions of nuclear transport emerging model plant of cell and and tip growth were distinct, as a developmental biology, in part due to the truncated KCH construct restored applicability of homologous recombination nuclear transport activity but not tip and high-resolution live imaging (Cove, growth retardation of the KO line. Thus, 2005; Cove et al., 2006; Vidali and our study identified KCH as a Bezanilla, 2012). The protonemal apical long-distance retrograde transporter as cell of P. patens is an excellent system to well as a cytoskeletal crosslinker, study MT-based transport. MTs are reminiscent of the versatile animal predominantly aligned along the cell dynein. longitudinal axis with a characteristic overall polarity depending on cell cycle INTRODUCTION stage (illustrated in Figure 2A). Nucleus, chloroplasts, and newly formed MTs have Intracellular transport is a critical been identified as cargo that is transported cellular mechanism for cell organisation in on MT tracks in protonemal cells, wherein eukaryotic cells. Many cellular components, nuclear movement was shown to be

1 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

independent of actin (Miki et al., 2015; et al., 2017). KAC/KCA is a class V Nakaoka et al., 2015; Yamada et al., 2017). kinesin-14 that no longer possesses MT Using this model system, kinesin-ARK affinity but has acquired an actin-binding (armadillo repeat-containing kinesin) was region and regulates actin-dependent first identified as a plus end-directed chloroplast photo-relocation movement nuclear transporter; upon RNAi knockdown and anchorage to the plasma membrane of this plant-specific, plus end-directed (Suetsugu et al., 2010; Suetsugu et al., motor protein, the nucleus migrated 2012). Class III kinesin-14 is localised to towards the cell centre after cell division as the spindle in moss but appears to have normal but then moved back to the cell lost MT-based motor activity (Miki et al., plate, i.e. the nucleus showed an abnormal 2014; Jonsson et al., 2015). The class IV minus end-directed motility (Figure 2A; Miki kinesin-14 TBK has a weak MT motor et al., 2015). It was also revealed that the activity and localises to cortical MTs yet its non-processive, minus end-directed KCBP cellular function is unknown (Goto and (kinesin-like calmodulin-binding protein)—a Asada, 2007; Jonsson et al., 2015). Class member of the kinesin-14 protein II kinesin-14 genes form a large clade in family—is required for minus end-directed the plant kinesin family, where 9 out of 61 A. nuclear transport. In the absence of KCBP, thaliana kinesin genes are classified into the nucleus could not move to the cell this clade (Figure 1A) and multiple activities centre immediately after cell division, i.e. and cellular functions have been reported. minus end-directed motility was inhibited Kinesin14-II possesses the calponin (Figure 2A; Yamada et al., 2017). Although homology (CH) domain in its a single dimeric KCBP cannot take multiple amino-terminal region followed by steps along the MT (non-processive), dimerisation and motor domains (hereafter clustered motors exhibit processive motility called KCH, which stands for kinesin with in vitro and in vivo; thus, multiple KCBP CH domain; Figure 1B). Adjacent to the motors associated with the nuclear surface motor domain, there exists an can transport the nucleus towards MT uncharacterised C-terminal extension in minus ends (Jonsson et al., 2015; Yamada this subfamily that is not found in ATK or et al., 2017). However, nuclear transport KCBP (Preuss et al., 2004; Frey et al., function of KCBP is limited during the latest 2009; Shen et al., 2012). Mutant analyses stage of cell division, as KCBP is no longer have uncovered divergent functions of necessary for maintaining central KCH, such as cell size regulation positioning of the nucleus during (OsKCH1; Frey et al., 2010), mitochondrial interphase. It is plausible that an additional respiration (Arabidopsis KP1; (Yang et al., minus end-directed motor protein that 2011)), and cell-to-cell movement of a antagonises kinesin-ARK and possibly transcription factor (Arabidopsis KinG; other plus end-directed kinesins is Spiegelman et al., 2018). However, the expressed in moss cells. complete picture of KCH function has not Minus end-directed kinesin-14 is been elucidated, since loss-of-function duplicated uniquely in the land plant analysis using complete null mutants has lineage and constitutes six subfamilies; in not been conducted for this highly Arabidopsis, it is the most expanded family duplicated gene subfamily in flowering among the kinesin superfamily (Zhu and plants. In contrast, P. patens possesses Dixit, 2011; Shen et al., 2012). KCBP only four KCH genes that are highly belongs to class VI of kinesin-14 and homologous (Figure 1A; red), suggesting transports not only the nucleus but also that they redundantly exhibit basal chloroplasts in moss (Yamada et al., 2017). functions of this kinesin subfamily. In Arabidopsis, the cytoskeletal In this study, we generated a plant with organisation of the trichome cell is a complete deletion of the KCH gene of P. defective in kcbp mutants, suggesting an patens, and provide evidence that KCH additional function to nuclear/chloroplast drives minus end-directed nuclear transport (Tian et al., 2015). The class I transport. Furthermore, KCH contributes to kinesin-14 ATK (HSET/XCTK2/Ncd) cell tip growth likely via crosslinking MTs at conserved in animals, is essential for the apical tip. These two functions are mitotic spindle coalescence, and drives distinct, as a KCH fragment that lacks the minus end-directed transport of newly unusual C-terminal extension fulfils the formed MTs along other MTs in the moss function of nuclear transport but not tip cytoplasm (Ambrose et al., 2005; Yamada growth. In contrast, the CH domain, which

2 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

has been assumed to be the cargo (i.e. KO line. Unlike the KCBP KO line, sister actin) binding site, was not required for nuclei moved towards the cell centre after either function. We propose that plant KCH chromosome segregation, indicating that is a versatile cargo transporter that also minus end-directed motility during fulfils other MT-based functions, analogous telophase was not impaired in the absence to the dynein motor in animals. of KCH (Figure 2B, Movie 1). However, unlike the control cells that maintained RESULTS cell-centre positioning of the nucleus during tip growth, the nucleus did not stop moving Complete KCH deletion affects moss at the cell centre but migrated further growth and morphology towards the cell tip (apical cell) or moved The expression database (http:// back towards the cell plate (subapical cell). http://bar.utoronto.ca/efp_physcomitrella/cg Consequently, the apical cells positioned i-bin/efpWeb.cgi) (Ortiz-Ramirez et al., the nucleus near the apical cell wall in the 2016) and our previous localisation KCH KO line; this phenotype was rescued analysis of KCH-Citrine fusion protein by ectopic Cerulean-KCHa expression (Citrine is a YFP variant) (Miki et al., 2014) (Figure 2C, D). During subsequent mitosis suggested that KCHa, b, and c, but not of apical cells, spindle assembly took place KHCd, are expressed in protonemal cells. 10% more apically compared with the However, since C-terminal tagging might control line, resulting in an apical shift of the perturb the function of the kinesin-14 cell division site (p < 0.05, n = 13 [KO] and subfamily, the motor domain of which is 7 [control]); this suggests a physiological generally located closer to the C-terminus, role for nuclear positioning. Given the we inserted the Citrine gene in front of known MT polarity during interphase of endogenous KCHa, b, and c (elements apical cells (plus-ends predominantly face other than the Citrine ORF were not the apex; Hiwatashi et al., 2014, Yamada integrated) (Figure S1A, B). With the newly et al., 2017; Figure 2A) and the selected Citrine-KCH lines, we confirmed appearance of the nuclear migration that KCHa, b, and c are indeed expressed defects contrary to kinesin-ARK depletion in protonemal cells, and furthermore, (the nucleus moves back to the cell plate in exhibit MT localisation at the cell tip (Figure apical cells; Miki et al., 2015), it was 1C). suggested that KCH drives MT minus To test the contribution of KCH to end-directed transport of the nucleus nuclear positioning and other intracellular during interphase. processes, we sequentially deleted four To determine whether the distribution KCH genes by means of homologous of other organelles is perturbed in the recombination in the moss lines expressing absence of KCH, we assessed the GFP-tubulin and histoneH2B-mRFP intracellular distribution of chloroplasts (by (Figure S1C, D). The KCHacd triple autofluorescence), mitochondria (the knockout (KO) line grew in an N-terminal 78 amino acids of the γ-subunit indistinguishable manner to wild-type moss. of the Arabidopsis mitochondrial F1Fo However, when all four KCH genes were ATPase) (Uchida et al., 2011; Nakaoka et deleted, moss colony growth was severely al., 2015), and vacuoles (the retarded (Figure 1D, E). In addition, the GFP-tubulin-excluded areas). Unlike the gametophore leaf was curly and the rhizoid nucleus, we did not observe any abnormal was much shorter than in the control line distribution or morphology of these (Figure 1F). These phenotypes were organelles, suggesting that the effect of suppressed when Cerulean-tagged KCHa KCH deletion is specific to the nucleus was expressed by a constitutively active (Figure 2E, Figure S2). promoter in the quadruple KO line (Figure In Arabidopsis root and mesophyll 1D–F). These results indicate that KCH is a cells, the nucleus is transported along actin critical motor in moss development, albeit filaments by the myosin XI-i motor, where not essential for moss viability. its mutant exhibited not only nuclear dynamics defects but also nuclear KCH is required for minus end-directed deformation (the nucleus becomes nuclear transport along MTs during rounder) (Tamura et al., 2013). interphase Interestingly, when we observed the KCH We performed live imaging of the KO line with spinning-disc confocal protonemal apical cells of the quadruple microscopy, we detected stretching and

3 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

invagination of the nucleus in 55% (n = 29) and Nick, 2012). Since the microscopy of apical cells (Figure 2F). The discrepancy technique employed is not sensitive in nuclear shape in the Arabidopsis myoXI-i enough to detect individual Citrine mutant and moss KCH KO is possibly molecules (Jonsson et al., 2015), the because of the additional force applied on diffraction-limited spots represent clustered the nuclear surface of moss, for example, Citrine-KCHa. This observation is by kinesin-ARK. Nevertheless, this consistent with the notion that KCH observation further supports the idea that functions as a minus end-directed KCH is responsible for nuclear motility. transporter of the nucleus.

Processive motility of KCH in vivo KCH promotes polarised tip growth Cytoskeletal motor proteins that drive In addition to defects in nuclear cargo transport are generally processive, transport, we observed severe tip growth where a single dimeric motor that attaches retardation in the complete KCH KO line to a cytoskeletal filament (MT or actin) (Figure 4A, Movie 3). Moreover, takes multiple steps towards one direction abnormally branched tip cells were before dissociation. Some non-processive occasionally observed in the KO line, motors can also be transporters when reflecting improperly polarised tip growth multiple dimers participate in cargo (Figure 4B). These defects were transport. KCBP represents the latter suppressed by Cerulean-KCHa expression, example; multiple KCBP molecules bind to confirming that tip growth defects in the KO MTs via the motor domain and to vesicular line were due to the loss of KCH proteins. cargo via the tail domain and execute The slow tip growth phenotype raised long-distance transport (Yamada et al., a possibility that nuclear mislocalisation 2017). The purified OsKCH1 motor was may be a secondary effect of growth also shown to be non-processive but its retardation in the KCH KO line. However, cohort action drives actin motility along this was unlikely as the nucleus was MTs (Walter et al., 2015). On the other mispositioned in the subapical cell, which hand, there have been contradictory hardly grows even in wild-type cells (Figure reports as to whether full-length KCH 2B, C). Nevertheless, to exclude this shows processive motility in vivo. When possibility, we examined nuclear tobacco GFP-NtKCH was expressed in positioning under another slow-growth tobacco BY-2 cells, motility of GFP signals condition generated by PRFa (profilin; an (i.e. clustered GFP signals) along MTs was actin regulator) RNAi (Vidali et al., 2007; detected (Klotz and Nick, 2012). In contrast, Nakaoka et al., 2012). We confirmed that GFP-AtKinG was observed only as static the nucleus was more centrally localised punctae on MTs in Nicotiana benthamiana when tip growth was suppressed following leaf epidermal cells (Spiegelman et al., RNAi of PRFa (Figure 2C, D), suggesting 2018). In these studies, however, that nuclear mistranslocation was not a GFP-tagged constructs were ectopically secondary effect of the growth defect overexpressed and might not represent associated with KCH KO. native KCH dynamics. To test if endogenous P. patens KCH MT focus formation at the tip requires exhibits processive motility in moss cells, actin and KCH we acquired time-lapse images of There are several tip-growing cells in Citrine-KCHa, which was expressed by the plants, such as moss protonemata, root native promoter at the endogenous locus, hairs, and pollen tubes. Actin filaments are using oblique illumination fluorescence essential for tip growth in these cells microscopy. This microscopy allows for (Rounds and Bezanilla, 2013), whereas visualisation of the cortex-proximal region, direction of growth is defined by MTs in which is largely devoid of some cases. In moss protonemata, MT auto-fluorescence derived from disruption by inhibitors or depletion of key chloroplasts. We observed punctate Citrine MT regulators leads to skewed or signals on MTs, and interestingly, the branched tip growth (Doonan et al., 1988; signals moved towards minus ends at a Hiwatashi et al., 2014). Since KCH binds to velocity of 441 ± 226 nm/s (± SD, n = 90) both MT and actin in vitro (Frey et al., 2009; for 1.6 ± 1.5 µm (± SD, n = 74; Figure 3, Xu et al., 2009; Umezu et al., 2011; Walter Movie 2); this velocity was 8-folds faster et al., 2015; Tseng et al., 2018), we than previously reported for NtKCH (Klotz hypothesised that P. patens KCH might

4 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

regulate cytoskeletal organisation at the (Movie 6), and nuclear positioning (Figure cell tip. 7B, C) were restored by ∆CH expression. We first performed live imaging of In contrast, we observed recovery of GFP-tubulin and lifeact-mCherry (actin nuclear positioning, but not tip growth or marker) in control apical cells. As reported MT focus formation, in the ∆C lines, previously, MT and actin focal points were indicating that the C-terminal extension is detected at the tip (Vidali et al., 2009; dispensable for nuclear transport function Hiwatashi et al., 2014); we also observed but is essential for tip growth. that they were largely—though not completely—colocalised with each other DISCUSSION (Figure 5A, Movie 4). These focal points were not maintained when either MT or We generated, for the first time, a plant actin was disrupted by specific drugs completely lacking in KCH proteins, which (Figure 5A, Movie 4); thus, MT and actin exhibited several noticeable phenotypes. focal points at the tip are mutually We focused our study on two prominent dependent. Interestingly, in the KCH KO phenotypes associated with protonemal line, the MT focus was smaller and much apical cells, nuclear mispositioning and tip less persistent (Figure 5B, C, Movie 5). growth retardation. This study also Furthermore, actin still accumulated at the represents the first report, to our transiently-formed MT focal point, knowledge, on intracellular dynamics of indicating the presence of other factor(s) function-verified KCH protein expressed that crosslink MTs and actin at the tip from the native locus. Our data elucidated (Figure 5D). the function of KCH as a long-distance KCHa localisation at the tip retrograde transporter and its role in disappeared following MT destabilisation cytoskeletal coalescence, during which an by oryzalin treatment, whereas actin unanticipated molecular mechanism might destabilisation by latrunculin A did not be involved (Figure 7D). affect KCHa MT association (Figure 5E); thus, KCH and MT co-localise independent KCH is a potent retrograde transporter of actin, but focal points cannot be formed The dimeric, truncated form of rice without actin. KCH1 or moss KCHa was shown to be non-processive in in vitro motility assays The C-terminal region of KCH is (Jonsson et al., 2015; Walter et al., 2015), required for tip growth but not for whereas a recent report showed nuclear transport, whereas the CH processive motility of rice KCH2 as a domain is dispensable for either activity dimer; OsKCH2 possesses unique To functionally dissect the KCH sequences adjacent to the C-terminus of protein, we constructed several truncation the motor domain which ensure constructs tagged with Cerulean, processivity (Tseng et al., 2018). However, transformed them into the KCH KO line, a cohort action of rice KCH1 can transport selected for transgenic lines, and assessed actin filaments along MTs over long if the phenotypes were rescued (Figure 6A, distances, suggesting that KCH family S1E). Surprisingly, the truncated KCHa proteins can generally function as deleted of the canonical CH domain, which long-distance cargo transporters (Walter et has been assumed to be the actin-binding al., 2015). Our observation of endogenous site, restored protonemal colony growth KCHa dynamics in the cytoplasm indicated and rhizoid development (Figure 6B, C). that this kinesin forms a cluster and indeed The gametophore leaves also showed moves processively towards the minus considerable recovery of morphology ends of MTs. Its run velocity (~440 nm/s) (Figure 6B). In contrast, we could not was approximately 3-folds faster than that obtain a transgenic line that showed of ATK (kinesin-14-I) and comparable to normal protonemal growth or KCBP (kinesin-14-VI, 413 nm/s)—two gametophore/rhizoid development after other kinesin-14 family proteins for which transformation of the KCH fragment with a processive motility in clusters and cargo deleted C-terminal extension (Figure 6B, transport function have been identified C). (Jonsson et al., 2015; Yamada et al., 2017). At the cellular and intracellular levels These results suggest that KCH proteins in the protonemata, tip growth (Figure 7A, are potent minus end-directed transporters. Movie 3), persistent MT focus formation Paradoxically, a minus-end-directed

5 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

motor is enriched at the MT plus end at the (Tanenbaum et al., 2010; Tsai et al., 2010). cell tip (Figure 1C, 5E). This might be Consistent with this model, the nucleus achieved by interacting with other MT plus unidirectionally moved towards the cell end-tracking proteins or actins. However, plate or cell tip in the absence of the results obtained using oblique kinesin-ARK or KCH, respectively. We illumination fluorescence microscopy could not detect enrichment of KCH or indicate that many KCH molecules kinesin-ARK (Miki et al., 2015) on the associate only transiently with the MT and nuclear surface during microscopy; then dissociate before moving along the however, failure in detection does not MT (Movie 2). Therefore, it is possible that necessarily preclude the possibility that both non-motile and motile proteins can be KCH acts as a transporter, as the visualized using confocal microscopy at the fluorescence signals might not be MT-rich cell tip. Consistent with this idea, distinguishable from cargo-free motors in the mCherry-tubulin signals increased at the cytoplasm or cytoplasmic background the tip similar to Citrine-KCH (Figure 1C), signals. This is particularly a likely scenario and the latter localised uniformly to the MT in moss protonemata, since when the MT focus was disrupted by auto-fluorescence derived from chloroplasts latrunculin A treatment (Figure 5E, right). is dominant in the cytoplasm. Using oblique Whether minus end-directed motility of illumination microscopy, we observed KCH is required for cytoskeletal Citrine-KCHa signals for each organisation at the tip remains to be cortex-proximal MT, suggesting that KCH determined. also associates with nucleus-proximal MTs (note that the nucleus cannot be located in KCH drives nuclear transport the focal plane using this microscopy Our study identified a defect in nuclear technique). Our truncation/rescue positioning in the absence of KCH. Nuclear experiments suggest that the region translocation during the cell cycle of downstream of the CH domain and protonemal apical cells can be divided into upstream of the motor domain is required four phases (Figure 2A), each driven by for nuclear attachment. It would be MT-based transport. KCBP is a minus interesting to search for the nuclear end-directed kinesin required for nuclear envelope-associated adaptor of KCH in transport specifically during phase I, the future studies. post-mitotic phase (Yamada et al., 2017). In contrast, kinesin-ARK is responsible for KCH and actin ensure MT focus plus end-directed transport during phase III formation for tip growth and possibly also phase IV, which Apical tip growth was severely correspond to the majority of the interphase suppressed in the KCH KO line. Based on (Miki et al., 2015). KCH is the third kinesin previous studies, this phenotype could be identified required for nuclear positioning; attributed to defects in the actin its KO phenotype indicated that KCH also cytoskeleton, MT cytoskeleton, and/or lipid functions during IV. The following three biogenesis at the tip (van Gisbergen et al., pieces of data strongly suggest that KCH 2012; Vidali and Bezanilla, 2012; Hiwatashi actually transports the nucleus, rather than et al., 2014; Saavedra et al., 2015). Defects indirectly affecting nuclear positioning by, in general housekeeping processes such for example, altering overall MT polarity; (1) as protein translation would also perturb KCH showed minus end-directed, cell growth. Although involvement of KCH processive motility in cells (Figure 3), (2) the in lipid production or other general cellular nucleus moved all the way to the cell tip in processes was not excluded, our the KCH KO line (Figure 2B), where plus localisation and phenotypic data more ends of MTs are enriched regardless of strongly suggest that KCH regulates MT KCH presence (Figure 5B), and (3) in the and possibly also actin cytoskeletons for tip absence of KCH, abnormal localisation was growth. Most compellingly, the detected for the nucleus, but not for the characteristic focus composed of other three cytoplasmic organelles (Figure MTs—which we observed to largely 2E). Thus, we envisage that kinesin-ARK coincide with the actin focal point—did not and KCH execute bi-directional transport of persistently form in KCH KO cells. the nucleus, reminiscent of animal cells However, unlike for actin or MT where antagonistic cytoplasmic dynein and destabilisation, tip growth was not kinesin-1 (or kinesin-3) motors are involved completely inhibited in KO cells.

6 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

Phenotypically, this was consistent with the observation that smaller and transient Conservation of KCH function MT/actin foci were still detectable in the KO In flowering plants, KCH family line. Residual MT bundling activity may be proteins show more divergence in amino mediated by other proteins such as acid sequences than in moss (Figure 1A), plant-specific plus end-directed kinesins and each member appears to play a KINIDa and b, whose double deletion distinct role. Interestingly, several reported resulted in curved growth accompanied phenotypes associated with KCH depletion with MT focus impersistence (Hiwatashi et and overexpression in flowering plants al., 2014). In addition, the mislocalised appear to be consistent with our findings nucleus also dampened tip growth, as this for moss. For example, coleoptile cells large organelle occasionally occupied the were shorter in the rice kch1 mutant, apical space where MT and actin normally whereas tobacco BY-2 cells were form clusters (e.g. Figure 2B). Thus, elongated upon overexpression of rice another significant role of nuclear KCH (Frey et al., 2010). Tobacco positioning in plants—other than division GFP-KCH, when ectopically expressed, site determination—may be ensuring decorates cortical MT arrays but is also proper organisation of the cytoskeletal abundantly present around the nucleus network for cell function. Nevertheless, (Klotz and Nick, 2012). Furthermore, nuclear mislocalisation is not the sole factor overexpression of rice KCH in the tobacco contributing to tip growth suppression, BY-2 cell line induced nuclear positioning since MT foci were not stably maintained in defects (Frey et al., 2010). Although these KO cells whose nuclei resided fairly distant studies did not elucidate the basis of the from the tip (e.g. Figure 5B). Furthermore, phenotypes, our study suggests that KCH the C-terminal deletion construct restored may transport the nucleus and promote nuclear positioning but not growth defects. MT-actin interaction in flowering plants as Thus, cytoskeletal disorganisation well. However, KCH functions might not be independent of nuclear positioning is likely limited to nuclear transport and MT-actin the major factor responsible for growth interactions during moss development, as retardation in KCH KO cells. many KCH proteins not associated with the The mechanism via which KCH nucleus were observed running along MTs promotes the formation of a large and (Figure 3). The complete KO line persistent ensemble of the two cytoskeletal generated in this study is a potentially filaments remains unclear. MT-actin valuable resource for uncovering the list of coalescence in the KCH KO line indicates KCH functions throughout plant the presence of other protein(s) that bridge development. the two filaments. KCH might support the coalescence via MT crosslinking. METHODS Alternatively, the model that KCH, independent of the CH domain, coalesces Moss culture and transformation MTs with actin by directly binding to both Moss lines used in this study are listed in Table filaments remains viable; intriguingly, an in S1; all lines originated from the Physcomitrella vitro study indicates that actin interacts also patens Gransden2004 strain. with the motor domain of rice KCH-O12 GFP-tubulin/histoneH2B-mRFP and (Umezu et al., 2011). Furthermore, the mCherry-tubulin strains were used as mother C-terminal extension of KCH was critical strains for gene disruption and Citrine tagging, for MT coalescence. Long (> 150 a.a.) respectively (Nakaoka et al., 2012; Kosetsu et C-terminal extension from the motor al., 2013). Methodologies of moss culture, domain is unique to the plant kinesin14-II– transformation, and transgenic line selection V families, and apart from the coiled-coil, were previously described thoroughly (Yamada no sequences have been identified in this et al., 2016). Briefly, cells were cultured on BCD region from which protein function can be agar medium for imaging. Transformation was deduced. Our study showed that this performed by the standard PEG-mediated unusual extension is a critical element of method and stable lines were selected with KCH, yet its exact function remains antibiotics (blasticidin S, nourseothricin, and unclear; it may be required for KCH hygromycin). Gene knockouts were obtained by accumulation at the tip or it may constitute replacing endogenous KCH genes with a an additional MT or actin interaction drug-resistant marker flanked by lox-P interface. sequences. After deleting KCH-a and b, the two

7 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

markers were removed by transient expression was performed as previously described of Cre recombinase, followed by replacement of (Jonsson et al., 2015; Nakaoka et al., 2015); KCH-d and b with the same markers. The cells were cultured in BCDAT medium covered Citrine gene was inserted into the N-terminus of with cellophane and imaging was performed KCH-a via homologous recombination. with a Nikon Ti microscope with a TIRF unit, a Drug-resistant genes were not integrated into 100× 1.49 NA lens, GEMINI split view the genome, as non-linearized, unstable (Hamamatsu), and EMCCD camera Evolve plasmids containing a drug resistant marker (Roper). All imaging was performed at 24–25°C were co-transformed with the linearized plasmid in the dark, except for the protonema growth containing Citrine tag constructs for transient assay that utilises light. drug selection. Gene disruption and Citrine tag insertion were confirmed by PCR. Drug assay Mosses were plated on agar gel in 35-mm Plasmid construction dishes following standard procedure and Plasmids (and primers for plasmid construction) incubated for 4–5 d (Yamada et al., 2016). Prior used for gene disruption, protein expression, to drug treatment, gels were excised from the and Citrine tagging are listed in Table S2. Gene dish except the central ~1 cm2 area on which knockout constructs were designed to replace moss colonies grew. Then, 1.5 mL water was endogenous KCH genes with a drug-resistant added to the dish and incubated for 30 min (i.e. marker flanked by lox-P sequences. One kb of equilibrated), followed by addition of 1.5 mL drug genomic DNA sequences solution. Cells were treated with 10 µM oryzalin upstream/downstream of start/stop codons (AccuStandard), 25 µM latrunculin A (Wako) or were amplified and cloned into the vectors 0.5–1% DMSO as control. containing lox-P sequences and drug-resistant markers. To generate a plasmid for Citrine Colony growth assay tagging, 1 kb of genomic DNA sequences Protoplasts prepared by the standard driselase upstream/downstream of the KCH gene start treatment were washed three times with 8% codon were cloned into the pKK138 vector mannitol solution, followed by overnight (Kosetsu et al., 2013; Yamada et al., 2016). To incubation in protoplast liquid medium in the generate truncation/rescue plasmids, KCH-a dark (Yamada et al., 2016). Protoplast solution sequences were amplified from a cDNA library was mixed with PRM-T and plated onto a and ligated into the pENTR/D-TOPO vector cellophane-covered PRM plate (cellophane containing Cerulean or Citrine sequences, was a gift from Futamura Chemical, Co., LTD.) followed by a Gateway LR reaction into a and incubated for 4 d. The cellophane was pTM153 vector that contains the EF1α transferred to a fresh BCDAT plate, incubated promoter, blasticidin-resistance cassette, and for 5 d, and then colonies were picked and used PTA1 sequences designated for homologous to inoculate a yet new BCDAT plate. After recombination-based integration (Miki et al., incubation for a further 10–11 d, colonies were 2016). Lifeact-mCherry was expressed by the imaged with a commercially-available digital actin promoter. camera (Olympus C-765) or stereoscopic microscope (Nikon SMZ800N and Sony In vivo microscopy ILCE-QX1α). Methods for epifluorescence and spinning-disc confocal microscopy were previously described Phylogenetic tree construction thoroughly (Yamada et al., 2016). Briefly, Following the method described in (Miki et al., protonemal cells were plated onto glass-bottom 2014; Miki et al., 2015), kinesin sequences plates coated with BCD agar medium and were aligned with MAFFT and the gaps were cultured for 4–5 d. The KO line was pre-cultured removed from the alignment using MacClade. on BCD medium covered with cellophane for 2– The phylogenetic tree was constructed using 3 d before being placed onto glass-bottom neighbour joining (NJ) methods using the plates. Long-term imaging by a wide-field Molecular Evolutionary Genetics Analysis microscope (low magnification lens) was (MEGA) software. Statistical support for internal performed with a Nikon Ti (10× 0.45 NA lens branches by bootstrap analyses was calculated and EMCCD camera Evolve [Roper]). using 1,000 replications. High-resolution imaging was performed with a spinning-disc confocal microscope (Nikon Data analysis TE2000 or Ti; 100× 1.45 NA lens, CSU-X1 To quantify the relative position of the nucleus in [Yokogawa], and EMCCD camera ImagEM the apical cell, microscope images that showed [Hamamatsu]). Oblique illumination microscopy the cell wall, nucleus, and cell tip were analysed

8 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

manually with ImageJ. Non-mitotic cells were Cove, D., Bezanilla, M., Harries, P., and randomly selected. The velocity and run length Quatrano, R. (2006). Mosses as model of Citrine-KCHa motility along endoplasmic MTs systems for the study of metabolism and were quantified based on kymographs, which development. Annu Rev Plant Biol 57, were generated from the images acquired with 497-520. oblique illumination fluorescence microscopy. Doonan, J.H., Cove, D.J., and Lloyd, C.W. To measure the area of moss colonies, cultured (1988). Microtubules and microfilaments in tip moss colony images were acquired by a digital growth: evidence that microtubules impose camera (Olympus C-765). All acquired images polarity on protonemal growth in were aligned side-by-side and the generated Physcomitrella patens. J Cell Sci 89, 533-540. single image was analysed with ImageJ; Ferenz, N.P., Paul, R., Fagerstrom, C., specifically, the image was processed by ‘Make Mogilner, A., and Wadsworth, P. (2009). binary’ and the colonies in the processed image Dynein antagonizes eg5 by crosslinking and were automatically detected and measured by sliding antiparallel microtubules. Curr Biol 19, ‘Analyze particles’ (size, 1.0-Infinity [mm2]; 1833-1838. circularity, 0.00-1.00). To quantify the duration Frey, N., Klotz, J., and Nick, P. (2009). of MT focus formation at the tip, GFP-tubulin Dynamic bridges--a calponin-domain kinesin was imaged every 3 s and the presence or from rice links actin filaments and absence of MT foci was manually judged at microtubules in both cycling and non-cycling each time frame; the focus was recognised cells. Plant Cell Physiol 50, 1493-1506. when two or more MT ends were converged. Frey, N., Klotz, J., and Nick, P. (2010). A kinesin with calponin-homology domain is Accession numbers involved in premitotic nuclear migration. J Exp Sequence data used in this article can be found Bot 61, 3423-3437. in the Phytozome database under the following Gonczy, P. (2008). Mechanisms of asymmetric accession numbers, KCHa, Pp3c14_19550; cell division: flies and worms pave the way. KCHb, Pp3c2_9150; KCHc, Pp3c17_21780; Nat Rev Mol Cell Biol 9, 355-366. KCHd, Pp3c24_19380. Goto, Y., and Asada, T. (2007). Excessive expression of the plant kinesin TBK5 converts ACKNOWLEDGMENTS cortical and perinuclear microtubules into a We thank Mitsuyasu Hasebe for providing the radial array emanating from a single focus. plasmids, Momoko Nishina, Tomohiro Miki, and Plant Cell Physiol 48, 753-761. Rie Inaba for technical assistance, as well as Grabham, P.W., Seale, G.E., Bennecib, M., Tomomi Kiyomitsu, Elena Kozgunova, Peishan Goldberg, D.J., and Vallee, R.B. (2007). Yi, and Tomohiro Miki for helpful comments Cytoplasmic dynein and LIS1 are required for regarding the manuscript. This work was microtubule advance during growth cone funded by the TORAY Science Foundation and remodeling and fast axonal outgrowth. J JSPS KAKENHI 15K14540 and 17H06471 (to Neurosci 27, 5823-5834. G.G). M.Y. is a recipient of a JSPS pre-doctoral Grill, S.W., and Hyman, A.A. (2005). Spindle fellowship. positioning by cortical pulling forces. Dev Cell 8, 461-465. AUTHOR CONTRIBUTIONS Hiwatashi, Y., Sato, Y., and Doonan, J.H. M.Y. and G.G. conceived and designed the (2014). Kinesins have a dual function in research project. M.Y. performed the organizing microtubules during both tip growth experiments and analysed the data. M.Y. and and cytokinesis in Physcomitrella patens. G.G. wrote the paper. Plant Cell 26, 1256-1266. Jonsson, E., Yamada, M., Vale, R.D., and REFERENCES Goshima, G. (2015). Clustering of a kinesin-14 motor enables processive Ambrose, J.C., Li, W., Marcus, A., Ma, H., retrograde microtubule-based transport in and Cyr, R. (2005). A minus-end-directed plants. Nat Plants 1. kinesin with plus-end tracking protein activity Klotz, J., and Nick, P. (2012). A novel is involved in spindle morphogenesis. Mol Biol actin-microtubule cross-linking kinesin, Cell 16, 1584-1592. NtKCH, functions in cell expansion and Coles, C.H., and Bradke, F. (2015). division. New Phytol 193, 576-589. Coordinating neuronal actin-microtubule Kong, Z., Ioki, M., Braybrook, S., Li, S., Ye, dynamics. Curr Biol 25, R677-691. Z.H., Julie Lee, Y.R., Hotta, T., Chang, A., Cove, D. (2005). The moss Physcomitrella Tian, J., Wang, G., and Liu, B. (2015). patens. Annu Rev Genet 39, 339-358. Kinesin-4 Functions in Vesicular Transport on

9 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

Cortical Microtubules and Regulates Cell Wall Growth mechanisms in tip-growing plant cells. Mechanics during Cell Elongation in Plants. Annu Rev Plant Biol 64, 243-265. Mol Plant 8, 1011-1023. Saavedra, L., Catarino, R., Heinz, T., Kosetsu, K., de Keijzer, J., Janson, M.E., Heilmann, I., Bezanilla, M., and Malho, R. and Goshima, G. (2013). (2015). Phosphatase and Tensin Homolog Is MICROTUBULE-ASSOCIATED PROTEIN65 a Growth Repressor of Both Rhizoid and is essential for maintenance of phragmoplast Gametophore Development in the Moss bipolarity and formation of the cell plate in Physcomitrella patens. Plant Physiol 169, Physcomitrella patens. Plant Cell 25, 2572-2586. 4479-4492. Shen, Z., Collatos, A.R., Bibeau, J.P., Furt, F., McNally, F.J. (2013). Mechanisms of spindle and Vidali, L. (2012). Phylogenetic analysis positioning. J Cell Biol 200, 131-140. of the Kinesin superfamily from physcomitrella. Miki, T., Nishina, M., and Goshima, G. (2015). Front Plant Sci 3, 230. RNAi screening identifies the armadillo Spiegelman, Z., Lee, C.M., and Gallagher, repeat-containing kinesins responsible for K.L. (2018). KinG Is a Plant-Specific Kinesin microtubule-dependent nuclear positioning in That Regulates Both Intra- and Intercellular Physcomitrella patens. Plant Cell Physiol 56, Movement of SHORT-ROOT. Plant Physiol 737-749. 176, 392-405. Miki, T., Nakaoka, Y., and Goshima, G. Suetsugu, N., Yamada, N., Kagawa, T., (2016). Live cell microscopy-based RNAi Yonekura, H., Uyeda, T.Q., Kadota, A., and screening in the moss Physcomitrella patens. Wada, M. (2010). Two kinesin-like proteins Methods Mol Biol 1470, 225-246. mediate actin-based chloroplast movement in Miki, T., Naito, H., Nishina, M., and Goshima, Arabidopsis thaliana. Proc Natl Acad Sci U S G. (2014). Endogenous localizome identifies A 107, 8860-8865. 43 mitotic kinesins in a plant cell. Proc Natl Suetsugu, N., Sato, Y., Tsuboi, H., Kasahara, Acad Sci U S A 111, E1053-1061. M., Imaizumi, T., Kagawa, T., Hiwatashi, Y., Nakaoka, Y., Kimura, A., Tani, T., and Hasebe, M., and Wada, M. (2012). The KAC Goshima, G. (2015). Cytoplasmic nucleation family of kinesin-like proteins is essential for and atypical branching nucleation generate the association of chloroplasts with the endoplasmic microtubules in Physcomitrella plasma membrane in land plants. Plant Cell patens. Plant Cell 27, 228-242. Physiol 53, 1854-1865. Nakaoka, Y., Miki, T., Fujioka, R., Uehara, R., Tamura, K., Iwabuchi, K., Fukao, Y., Kondo, Tomioka, A., Obuse, C., Kubo, M., M., Okamoto, K., Ueda, H., Nishimura, M., Hiwatashi, Y., and Goshima, G. (2012). An and Hara-Nishimura, I. (2013). Myosin XI-i inducible RNA interference system in links the nuclear membrane to the Physcomitrella patens reveals a dominant role cytoskeleton to control nuclear movement and of augmin in phragmoplast microtubule shape in Arabidopsis. Curr Biol 23, generation. Plant Cell 24, 1478-1493. 1776-1781. Ortiz-Ramirez, C., Hernandez-Coronado, M., Tanenbaum, M.E., Akhmanova, A., and Thamm, A., Catarino, B., Wang, M., Dolan, Medema, R.H. (2010). Dynein at the nuclear L., Feijo, J.A., and Becker, J.D. (2016). A envelope. EMBO Rep 11, 649. Transcriptome Atlas of Physcomitrella patens Tanenbaum, M.E., Vale, R.D., and McKenney, Provides Insights into the Evolution and R.J. (2013). Cytoplasmic dynein crosslinks Development of Land Plants. Mol Plant 9, and slides anti-parallel microtubules using its 205-220. two motor domains. Elife 2, e00943. Perlson, E., Hendricks, A.G., Lazarus, J.E., Tian, J., Han, L., Feng, Z., Wang, G., Liu, W., Ben-Yaakov, K., Gradus, T., Tokito, M., and Ma, Y., Yu, Y., and Kong, Z. (2015). Holzbaur, E.L. (2013). Dynein interacts with Orchestration of microtubules and the actin the neural cell adhesion molecule (NCAM180) cytoskeleton in trichome cell shape to tether dynamic microtubules and maintain determination by a plant-unique kinesin. Elife synaptic density in cortical neurons. J Biol 4. Chem 288, 27812-27824. Tsai, J.W., Lian, W.N., Kemal, S., Kriegstein, Preuss, M.L., Kovar, D.R., Lee, Y.R., Staiger, A.R., and Vallee, R.B. (2010). Kinesin 3 and C.J., Delmer, D.P., and Liu, B. (2004). A cytoplasmic dynein mediate interkinetic plant-specific kinesin binds to actin nuclear migration in neural stem cells. Nat microfilaments and interacts with cortical Neurosci 13, 1463-1471. microtubules in cotton fibers. Plant Physiol Tseng, K.F., Wang, P., Lee, Y.J., Bowen, J., 136, 3945-3955. Gicking, A.M., Guo, L., Liu, B., and Qiu, W. Rounds, C.M., and Bezanilla, M. (2013). (2018). The preprophase band-associated

10 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

kinesin-14 OsKCH2 is a processive microtubule-based motor proteins. minus-end-directed microtubule motor. Nat Protoplasma 249, 887-899. Commun 9, 1067. Zhu, C., Ganguly, A., Baskin, T.I., McClosky, Uchida, M., Ohtani, S., Ichinose, M., Sugita, D.D., Anderson, C.T., Foster, C., Meunier, C., and Sugita, M. (2011). The PPR-DYW K.A., Okamoto, R., Berg, H., and Dixit, R. proteins are required for RNA editing of rps14, (2015). The fragile Fiber1 kinesin contributes cox1 and nad5 transcripts in Physcomitrella to cortical microtubule-mediated trafficking of patens mitochondria. FEBS Lett 585, cell wall components. Plant Physiol 167, 2367-2371. 780-792. Umezu, N., Umeki, N., Mitsui, T., Kondo, K., and Maruta, S. (2011). Characterization of a novel rice kinesin O12 with a calponin MOVIE LEGENDS homology domain. J Biochem 149, 91-101. van Gisbergen, P.A., Li, M., Wu, S.Z., and Movie 1. Nuclear migration defects in the KCH Bezanilla, M. (2012). Class II formin targeting KO line to the cell cortex by binding PI(3,5)P(2) is Images were acquired every 3 min with essential for polarized growth. J Cell Biol 198, epifluorescence microscopy. Green, MT; 235-250. magenta, histone. Scale bar, 50 µm. Vidali, L., and Bezanilla, M. (2012). Physcomitrella patens: a model for tip cell Movie 2. Processive motility of clustered growth and differentiation. Curr Opin Plant Citrine-KCH-a Biol 15, 625-631. Citrine (green) and mCherry-tubulin (magenta) Vidali, L., Augustine, R.C., Kleinman, K.P., images were acquired every 0.24 s in the and Bezanilla, M. (2007). Profilin is essential endoplasm of a protonemal cell. Note that for tip growth in the moss Physcomitrella auto-fluorescence derived from chloroplasts are patens. Plant Cell 19, 3705-3722. also visible (green). Scale bar, 5 µm. Vidali, L., Rounds, C.M., Hepler, P.K., and Bezanilla, M. (2009). Lifeact-mEGFP reveals Movie 3. Tip growth retardation in the KCH a dynamic apical F-actin network in tip KO line growing plant cells. PLoS One 4, e5744. Images were acquired every 3 min (10× lens, Walter, W.J., Machens, I., Rafieian, F., and transmission light). Expression of the Diez, S. (2015). The non-processive rice CH-deleted construct restored normal growth, kinesin-14 OsKCH1 transports actin filaments whereas C-terminal deletion did not. Scale bar, along microtubules with two distinct velocities. 50 µm. Nat Plants 1, 15111. Xu, T., Qu, Z., Yang, X., Qin, X., Xiong, J., Movie 4. MT and actin focal points at the cell Wang, Y., Ren, D., and Liu, G. (2009). A tip cotton kinesin GhKCH2 interacts with both Images were acquired every 30 s with microtubules and microfilaments. Biochem J spinning-disc confocal microscopy. Drugs were 421, 171-180. added at time 0. Green, MT; magenta, actin. Yamada, M., Miki, T., and Goshima, G. (2016). Scale bar, 5 µm. Imaging Mitosis in the Moss Physcomitrella patens. Methods Mol Biol 1413, 263-282. Movie 5. KCH-dependent formation of the Yamada, M., Tanaka-Takiguchi, Y., Hayashi, MT focus at the apical cell tip M., Nishina, M., and Goshima, G. (2017). GFP-tubulin images were acquired every 3 s Multiple kinesin-14 family members drive with spinning-disc confocal microscopy. Scale microtubule minus end-directed transport in bar, 5 µm. plant cells. J Cell Biol 216, 1705-1714. Yang, X.Y., Chen, Z.W., Xu, T., Qu, Z., Pan, Movie 6. MT focus at the apical cell tip after X.D., Qin, X.H., Ren, D.T., and Liu, G.Q. expression of truncated KCH (2011). Arabidopsis kinesin KP1 specifically GFP-tubulin images were acquired every 3 s interacts with VDAC3, a mitochondrial protein, with spinning-disc confocal microscopy. Note and regulates respiration during seed that signals derived from Citrine-KCH fragments germination at low temperature. Plant Cell 23, are also merged in the rescue samples, since 1093-1106. the emission filter associated with our Zhu, C., and Dixit, R. (2011). Functions of the microscope could not separate Citrine Arabidopsis kinesin superfamily of fluorescence from GFP. Scale bar, 5 µm.

11 bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

Table S1. Moss lines used in this study Line # Clone # Name Plasmid Resistance References 0002 3 GFP-tubulin/histone H2B-mRFP (GFP-tub/H2B-RFP) GFP-tubulin, pGG616 G418, Zeo Nakaoka et al., 2012 0291 1 GFP-tubulin/Mitochondria-tagRFP GFP-tubulin, pMT-RFP G418, Zeo Nakaoka et al., 2015 0367 52 mCherry-tubulin pTM426 BSD Kosetsu et al., 2013 0459 5, 6 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO GFP-tubulin, pGG616, pMN508, pMN509, pMY159 Zeo, G418, NTC 0465 2,5,9 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215 Zeo, G418, NTC, BSD 0466 1 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215 Zeo, G418, NTC, BSD 0480 24, 38 mCherry-tubulin/Citrine-KCHa pTM426, pMN570 BSD 0481 49 mCherry-tubulin/Citrine-KCHb pTM426, pMY177 BSD 0482 242, 261, 275 mCherry-tubulin/Citrine-KCHc pTM426, pMY187 BSD 0483 35,145 mCherry-tubulin/Citrine-KCHd pTM426, pMY186 BSD 491 1, 2 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Mitochondria-tagRFP GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMT-RFP Zeo, G418, NTC, BSD, Hyg 0493 1, 2 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Lifeact-mCherry GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY252 Zeo, G418, NTC, BSD, Hyg 0505 2, 4 GFP-tub/H2B-RFP/Lifeact-mCherry GFP-tubulin, pGG616, pMY252 Zeo, G418, Hyg 0638 0106, 0301 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Citrine-KCHa FL GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY383 G418, Zeo, BSD 0639 0108, 0301 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Citrine-KCHa ∆CH (199-1396) GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY388 G418, Zeo, BSD 0640 0102, 0103, 0304 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Citrine-KCHa ∆C (1-1024) GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY389 G418, Zeo, BSD 0641 0108, 0301 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Citrine-KCHa N (1-598) GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY390 G418, Zeo, BSD 0642 0103, 0303 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Citrine-KCHa C (978-1396) GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY391 G418, Zeo, BSD 0645 0303, 0307 GFP-tub/H2B-RFP/KCHa KO/KCHc KO/KCHd KO/KCHb KO/Cerulean-KCHa FL GFP-tubulin, pGG616, pMN508, pMN509, pMY159, pMY215, pMY329 G418, Zeo, BSD Zeo: Zeocin, Hyg: Hygromycin, BSD: Blasticidin S, NTC: Nourseothricin bioRxiv preprint doi: https://doi.org/10.1101/308775; this version posted April 26, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC 4.0 International license.

Table S2. List of PCR primers and plasmids used in this study Plasmid Gene 5' primer 3' primer Selection marker targeting site pMN508 KCH-a AAActcgagGGTGGCTGCGGTCCCTGGTGAG AAAggatccTAGAGGGTAAGTCCTCAGCATG BSD Endogenous KO AAAaagcttAACTTCAGCGTTCAGACGCGAC AAAAgcggccgcCTGCTTCCTGAGCTGTACCAGG pMY215 KCH-b AAAgggcccGAGTTTGCGTATCCTTGGTAGG AAAcatatgTTGGAATATCATTCTATTTACCG BSD Endogenous KO AAActcgagAAAAAcatatgAACTACGGCGTTCAAACAAAATTACC AAAccgcggAGAACCGCCAAATGGGTTGTG pMN509 KCH-c AAActcgagGATGCTGTGGTGGTTGCGGTCG AAAggatccTTATGACAAATCTTAGTTATCATAG Hyg Endogenous KO AAAaagcttTCTTGCCACATCCATAACTTCAG AAAAgcggccgcCACTGTTTATGATACAAGGTAC pMY159 KCH-d AAAgtcgacAGTACACAGTGTGATTG AAAcatatgTCGTCAGTAAGTCTTCCTTCATC NTC Endogenous KO AAAgatatcAAAAAcatatgGGCCCCTTCGCAGCTCTAAAC AAAccgcggGTGATTTAAGGCCCTGTTTGC pMY329 Cerulean AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC AAAAgcggccgcCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a AAAAgcggccgcccccttcaccATGGATGTGGCAAGAATGGGG AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY383 Citrine AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC AAAAgcggccgcCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a FL AAAAgcggccgcccccttcaccATGGATGTGGCAAGAATGGGG AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY388 Citrine AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC AAAactagtactacctccacctccCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a (∆CH) AAAactagtTCCATGTCCTCAAGTTCATC AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY389 Citrine AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a (∆C) AAAactagtGATGTGGCAAGAATGGGGTATG AAAAggcgcgccCACCCTTTTAACCGCTCTCGCCTCTATCGTGTC pMY390 Citrine AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a (N) AAAactagtGATGTGGCAAGAATGGGGTATG AAAAggcgcgccCACCCTTTTACCTCATAAACTCATCTTTGGT pMY391 Citrine AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC BSD PTA1 KCH-a (C) AAAactagtGCTCGAAGTAACAAGGAAAG AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY427 Cerulean AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC NTC PTA1 KCH-a FL AAAAgcggccgcccccttcaccATGGATGTGGCAAGAATGGGG AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY417 Cerulean AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC NTC PTA1 KCH-a (∆CH) AAAactagtTCCATGTCCTCAAGTTCATC AAAAggcgcgcccacccttTTACCTCCAAGAGGTTGAGGACC pMY418 Cerulean AAAAgcggccgccaccATGGTGAGCAAGGGCGAGGAGC aaaactagtactacctccacctccCTTGTACAGCTCGTCCATGC NTC PTA1 KCH-a (∆C) AAAactagtGATGTGGCAAGAATGGGGTATG AAAAggcgcgccCACCCTTTTAACCGCTCTCGCCTCTATCGTGTC pMY252 Lifeact AAAcccgggCACCATGGGTGTCGCAGATTTGAT AAAggtaccCTTGTACAGCTCGTC Hyg MADS2 mCherry pMN570 KCH-a AAActcgagGGTGGCTGCGGTCCCTGGTGAG AAAactagttATGGATGTGGCAAGAATGGGGTATG Endogenous Tagging AAAatcgatAACTTCAGCGTTCAGACGCGACTA AAAAgcggccgcGCAACCACATAATCGCATGCTGC pMY177 KCH-b GAGTTTGCGTATCCTTGGTAGG AACTACGGCGTTCAAACAAAATTACC Endogenous Tagging AAAactagttATGGATGTGGCTAGAATGGGTTATG AAAAgcggccgcGCTGAGTTCGCATCACATTGACAG pMY187 KCH-c AAActcgagTGGTGTCGATGCTGTGGTGGTTG AAAaagcttTCTTGCCACATCCATAACTTCAG Endogenous Tagging AAAtctagagATGGGGTATGAGAACGGTTCATCC AAAAgcggccgcCATGTTATTCATATTCAGACATG pMY186 KCH-d AAAggtaccGTCGACAGTACACAGTGTGATTG AAAaagcttGGCCCCTTCGCAGCTCTAAAC Endogenous Tagging AAAactagttATGCATGACTCCACAGCGAGTTC AAAAgcggccgcGGTTGAACAATGTTGATGACC Tagging KCH-a GTGCTGATGCAATGCTGGTAGG GAGACCCACGTTTCCAACAATC Genotyping KCH-b GTTAGGGAGATCGGATTGGGAAAGG GACAGACCGAGATTTCCAACC Outer PCR KCH-c GGTCTGTGAGGTCGCCTCATGCACGG CCATATCCGATAACAGCCGCAAG KCH-d CCGTATTCAGCGCGAATCACAAAAG GTTCCTACCTTTGGCACTGAGC Knock out KCH-a CCTGGCGAGTGAGTGAGAGAGTG CGCTTGTGCTTCCACAACACTGAG Genotyping KCH-b GTTAGGGAGATCGGATTGGGAAAGG GACTGAATCAGCAGTTTGTATTGAG Outer PCR KCH-c GGTCTGTGAGGTCGCCTCATGCACGG CAATGGAATACAATGGCTTCTACG KCH-d CCGTATTCAGCGCGAATCACAAAAG GCAACCGAGTTTGAAGTGTTACAGC Knock out KCH-a GTACAACCTGGCGCTGTACACAAGGTTAG GCACACGTACATCGAGTGCAGTGGTCG Genotyping KCH-b GGTGTACGATAAGTGCAGGAATGACGAC GAAATGGAGTTAGCAGGTTCGAAACTTGC Inner PCR KCH-c GGTTGTGGCATTAGGCTACCTTATGACATG CCATATCCGATAACAGCCGCAAG KCH-d CCTGTAACGCAGGTATCTTCTGCGCCTC GGAGAGGGATATGCAATGCTGATTAGC Lowercase letters indicate the restriction enzyme site or overlapping region for INFUSION cloning (TAKARA)