A dynein loading zone for retrograde endosome motility at microtubule plus-ends

Abstract

In the fungus Ustilago maydis, early endosomes move bidirectionally along microtubules (MTs) and facilitate growth by local membrane recycling at the tip of the infectious hypha. Here, we set out to elucidate the molecular mechanism of this process. We show that endosomes travel by Kinesin-3 activity into the hyphal apex, where they reverse direction and move backwards in a dynein-dependent manner. Our data demonstrate that dynein, dynactin and Lis1 accumulate at MT plus-ends within the hyphal tip, where they provide a reservoir of inactive motors for retrograde endosome transport. Consistently, endosome traffic is abolished after depletion of the dynein activator Lis1 and in Kinesin-1 null mutants, which was due to a defect in targeting of dynein and dynactin to the apical MT plus-ends. Furthermore, biologically active GFP-dynein travels on endosomes in retrograde and not in anterograde direction. Surprisingly, a CLIP170 homologue was neither needed for dynein localization nor for endosome transport. These results suggest an apical dynein loading zone in the hyphal tip, which ensure that endosomes reach the expanding growth region before they reverse direction.

Introduction

Fungal pathogenicity is based on the ability to invade the host tissue by polarized growth of the hypha. It is thought that this process is based on motor-mediated transport of growth supplies along the cytoskeleton to the growing hyphal apex (Steinberg, 2000), where polarized exocytosis occurs (Gow, 1995). In addition, evidence exist that endosome-based membrane recycling supports hyphal tip growth (Wedlich-Söldner et al, 2000). Interestingly, in yeast-like cells and hyphae of the dimorphic plant pathogenic fungus Ustilago maydis, early endosomes (EE) show rapid microtubule (MT)-dependent motility (Wedlich-Söldner et al, 2000). In yeast-like cells, this bidirectional traffic is mediated by Kinesin-3 and cytoplasmic dynein (Wedlich-Söldner et al, 2002), and similar motors might mediate the motility of endosomes in vertebrates (Hoepfner et al, 2005).

Currently, there are several concepts for the activity of kinesins and dynein in bidirectional organelle traffic. Most experimental evidence from animal cells suggests that both motors interact on the same organelle, thus controlling each other's activity (Gross et al, 2002; Welte, 2004). This notion is supported by the finding that Kinesin-1 directly binds the dynein complex (Ligon et al, 2004). However, such physical interaction could also reflect that one motor is a cargo of the other. This notion is supported by recent results showing that fungal dynein, which accumulates at the plus-ends of MTs (Han et al, 2001; Lee et al, 2003; Sheeman et al, 2003), requires Kinesin-1 for its plus-ends targeting in Aspergillus nidulans (Zhang et al, 2003).

In Saccharomyces cerevisiae growing MT plus-ends take dynein to the cortex, where it becomes activated and ‘off-loaded' in order to migrate towards the minus-ends, thereby mediating spindle positioning (Lee et al, 2003; Sheeman et al, 2003). A key component in this process is Pac1 (Sheeman et al, 2003), which is an NudF/LIS1 homologue (Xiang et al, 1999). These factors are thought to activate dynein for minus-end directed motility (Faulkner et al, 2000; Smith et al, 2000; Coquelle et al, 2002; Zhang et al, 2003; Lansbergen et al, 2004; Mesngon et al, 2006). Another factor involved in dynein regulation is dynactin (Schroer, 2004), which also affects dynein activity and its MT affinity at plus-ends (Valetti et al, 1999; Zhang et al, 2003). Moreover, together with the cytoplasmic linker protein 170 (CLIP-170), dynactin is thought to recruit endosomes to MT plus-ends (Pierre et al, 1992; Vaughan, 2005), and similar mechanisms might also account for other cellular cargoes (Mimori-Kiyosue and Tsukita, 2003). Taken together, these findings emphasize the importance of MT plus-ends for intracellular organelle motility.

Based on these studies, it is tempting to speculate that plus-end located fungal dynein represents a reservoir for minus-end directed transport in interphase (Yamamoto and Hiraoka, 2003; Zhang et al, 2003). Here, we provide experimental evidence for such a concept. We show that cytoplasmic dynein and homologues of the dynein activators LIS1 and dynactin localize to plus-ends of MTs. A Kinesin-3 delivers EE to the hyphal apex, where they are loaded onto dynein for retrograde transport. This ensures that endosomes reach the hyphal tip, where they participate in cell expansion by endocytic membrane recycling.

Results

Dynein and a Kif1A-like kinesin support bidirectional endosome traffic in hyphae of U. maydis

It was previously shown that EE move along MTs in yeast-like cells and hyphae of U. maydis (Wedlich-Söldner et al, 2000). As a first step towards an understanding of this motility, we determined the directionality of MTs in hyphae by the observation of a YFP-tagged EB1-homologue Peb1 that binds only to growing MT plus-ends (Straube et al, 2003). Therefore, any moving signal indicates the orientation of a MT. Infectious hyphae of U. maydis consist of a single tip cell of ∼120–150 μm in length and 2–3 μm width (Figure 1A), with the nucleus being located close to the center of the cell (Figure 1B [II]; chromosomal DNA labeled by CFP-histone 4 is given in blue; strain AB33Peb1Y_H4C, for all strains see Table I; see Supplementary Movie 1). Hyphae contain long MTs that often form bundles (Steinberg et al, 2001). The majority of Peb1-YFP signals that, based on their consistent signal intensity, marked individual MT plus-ends moved towards the growing apex and the basal septum, suggesting that plus-ends are directed to the cell poles (Figure 1B [I]+[III] and Figure 1C; Peb1 in red). In the middle of the cell, Peb1-YFP signals grow towards each other (Figure 1B [II], 1C) and passed the nucleus in opposite directions, indicating that MTs have an antipolar orientation in the central region of the hypha. Consistent with previous results (Wedlich-Söldner et al, 2000), individual MTs and bundles of MTs served as tracks for bidirectional motility of EE that were tagged with the specific t-SNARE Yup1-GFP (Figure 2A; strain AB33RT_YG; see Supplementary Movie 2), but which also localizes to stationary vacuoles, albeit to a minor extend (Wedlich-Söldner et al, 2000). Interestingly, most EE reached the MT ends in the hyphal tip before they reversed direction (Figure 2B). This was a first indication that factors in the hyphal apex participate in retrograde transport. In yeast-like cells, EE traffic depends on Kinesin-3 and dynein (Wedlich-Söldner et al, 2002). Consistently, deletion of Kinesin-3 in hyphae almost abolished EE motility (Figure 2C, ΔKin3; strain AB33ΔKin3_YG; for ‘control' see Figure 2D, strain AB33YG; for Figure 2D–F; see Supplementary Movie 3 on EMBO web site) and led to clustering of EE around the nucleus in the middle of the shorter hyphae (Figure 2E), whereas EE were equally distributed in control hyphae (Figure 2D). Although these Δkin3 hyphae were bipolar (Figure 2E, Schuchardt et al, 2005), MT orientation was not altered (89% plus-ends towards tip; n=59; strain AB33ΔKin3_Peb1Y), indicating that EE are trapped at minus-ends of MTs near the cell center. In contrast, depletion of dynein shifted EE to the cell poles (Figure 2F, Dyn1↓; strain AB33rDyn1_YG), where MT plus-ends were concentrated. Unfortunately, Western analysis revealed that expression of dynein was not completely abolished in these mutants (not shown). Therefore, we analyzed EE traffic in pheromone-induced hyphae of temperature-sensitive dynein mutants (strain FB1Dyn2^ts_YG). Such dynein mutant hyphae show growth defects (Fuchs et al, 2005) and contained longer MTs (not shown). However, MT orientation was not affected, as 95.5% of all Peb1-RFP signals were directed to the tip (n=40; strain FB1Dyn2^tsPeb1Y). Again, EE were trapped at plus-ends in the hyphal tip (not shown) and almost all motility was abolished (Figure 2C, Dyn2ts). These observations are most consistent with the notion that a balance of Kinesin-3 and dynein activity underlies long-distance EE transport in U. maydis hyphae.

Figure 1.

Microtubule orientation in hyphae of U. maydis. (A) Overview of a hypha of U. maydis. Boxed regions (I–III) correspond to series given in 1B. Note that tip cells leave ‘empty' cell wall section behind (arrow) while they grow at their tip (right end). Bar: 10 μm. (B) Peb1-YFP (given in red) marks MT plus-ends that elongate towards the basal septum (I, arrows), as well as the hyphal tip (III, arrows), indicating that the plus-ends are orientated towards both poles of the hyphal cell. Around the nucleus (II; labeled by histone 4-CFP is given in blue), Peb1-signals move in both directions (arrow and arrowhead), indicating that MTs have an antipolar orientation. Elapsed time is given in seconds. Bars: 5 μm. (C) Quantitative analysis of growing MT ends demonstrates that the plus-ends (indicated by ‘plus') are focused to the tip (indicated by red arrows) and the subapical septum (indicated by blue arrows). An antipolar MT orientation in the middle of the cell suggests that minus-ends are located near the nucleus. Numbers indicate the percentage of Peb1-YFP signals that move in the direction indicated. Supplementary movie for Panel B is given on the EMBO web site.

Table 1.

Strains and plasmids used in this study

Strains/plasmid	Genotype	Reference
AB33RT_YG	a2 PnarbW2 PnarbE1, bleR/potefRFPTub1/potefYup1GFP	This study
AB33YG	a2 PnarbW2 PnarbE1, bleR/potefYup1GFP	This study
AB33ΔKin3_YG	a2 PnarbW2 PnarbE1 Δkin3∷hygR, bleR/potefYup1GFP	This study
FB1Dyn2ts_YG	a1b1 dyn2ts, natR/potefYup1GFP	This study
AB33ΔKin1_Peb1Y	a2 PnarbW2 PnarbE1 Δkin1∷hygR, Ppeb1-peb1-yfp, bleR, natR	This study
AB33ΔKin3_Peb1Y	a2 PnarbW2 PnarbE1 Δkin3∷hygR, Ppeb1-peb1-yfp, bleR, natR	This study
FB1Dyn2tsPeb1Y	a1b1 dyn2∷dyn2ts Ppeb1-peb1-yfp, natR, bleR	This study
AB33rDyn1_YG	a2 PnarbW2 PnarbE1 Pcrg1-dyn1, bleR, natR/potefYup1GFP	This study
AB33ΔKin1_YG	a2 PnarbW2 PnarbE1 Δkin1∷hygR/potefYup1GFP	This study
AB33Peb1Y_H4C	a2 PnarbW2 PnarbE1 Ppeb1-peb1-yfp, bleR, natR/pH4CFP	This study
AB33G3Dyn2	a2 PnarbW2 PnarbE1 3xgfp-dyn2, bleR, natR	This study
AB33G3Dyn2_RT	a2 PnarbW2 PnarbE1 3xgfp-dyn2, bleR, hygR/potefRFPTub1	This study
AB33ΔKin1_G3D2_RT	a2 PnarbW2 PnarbE1 Δkin1∷hygR 3xgfp-dyn2, natR, bleR/potefRFPTub1	This study
AB33ΔKin1_G3D2_YR2	a2 PnarbW2 PnarbE1 Δkin1∷hygR 3xgfp-dyn2, natR, bleR/potefYup1RFP2	This study
AB33G3Dyn2_YR2	a2 PnarbW2 PnarbE1 3xgfp-dyn2, hygR, bleR/potefYup1RFP2	This study
FB2GLis1	a2b2 PLis1-gfp-lis1, hygR	This study
FB2GLis1_RT	a2b2 PLis1-gfp-lis1, hygR/potefRFPTub1	This study
FB2nGLis1	a2b2 Pnar-gfp-lis1, natR	This study
FB2G3Dyn2_nLis1	a2b2 Pdyn2-3xgfp-dyn2 Pnar-lis1, hygR natR	This study
FB2G3Dyn2	a2b2 Pdyn2-3xgfp-dyn2, hygR	This study
FB2G3Dyn2_nLis1_RT	a2b2 Pdyn2-3xgfp-dyn2 Pnar-lis1, hygR natR/potefRFPTub1	This study
FB2G3Dyn2_nLis1_YR2	a2b2 Pdyn2-3xgfp-dyn2 Pnar-lis1, hygR natR/potefYup1RFP2	This study
FB2YR2	a2b2/potefYup1RFP2	This study
FB2nLis1_Peb1R_GT	a2b2 Pnar-lis1 Ppeb1-peb1-rfp, natR, bleR/potefGFPTub1	This study
FB2G3Lis1_YR2	a2b2 Plis1-3xgfp-lis1, natR /potefYup1RFP2	This study
FB2G3Dya1_RT	a2b2 Pdya1-3xgfp-dya1, bleR/potefRFPTub1	This study
FB2rGDya1	a2b2 Pcrg1-gfp-dya1, bleR	This study
FB2rDya1_G3D	a2b2 Pcrg1-dya1 Pdyn2-3xgfp-dyn2, bleR, hygR	This study
FB2rDya1_G3D_RT	a2b2 Pcrg1-dya1 Pdyn2-3xgfp-dyn2, bleR, hygR/potefRFPTub1	This study
FB2rDya1_YG	a2b2 Pcrg1-dya1, bleR/potefYup1GFP	This study
FB2rDya1_Peb1R_GT	a2b2 Pcrg1-dya1 Ppeb1-peb1-rfp, natR, bleR/potefGFPTub1	This study
FB2G3Dya1_YR2	a2b2 Pdya1-3xgfp-dya1, bleR/potefYup1RFP2	This study
FB2YG	a2b2/potefYup1GFP	This study
FB2ΔKin1_G3Dya1_RT	a2b2 Δkin1∷hygR Pdya1-3xgfp-dya1, bleR/potefRFPTub1	This study
FB2rKin1_GLis1	a2b2 Plis1-gfp-lis1 Pcrg1-kin1, hygR natR	This study
FB2GClip1_RT	a2b2 Pclip1-gfp-clip1, hygR/potefRFPTub1	This study
FB2ΔClip1	a2b2 Δclip1∷bleR	This study
FB2ΔClip1_YG	a2b2 Δclip1∷bleR/potefYup1GFP	This study
FB2ΔClip1_G3Dyn2_RT	a2b2 Δclip1∷bleR Pdyn2-3xgfp-dyn2, hygR/potefRFPTub1	This study
FB2GT	a2b2/potefGFPTub1	(Steinberg et al, 2001)
FB2ΔClip1_GT	a2b2 ΔClip1∷bleR/potefGFPTub1	This study
FB2ΔClip1 _Peb1R_GT	a2b2 Δclip1∷bleR, Ppeb1-peb1-rfp, natR/potefGFPTub1	This study
RWS9	a1b1/pKin3YFP/pPX10-146CFP	(Wedlich-Söldner et al, 2002)
potefGFPTub1	Potef-egfp-tub1, cbxR	(Steinberg et al, 2001)
potefRFPTub1	Potef-mrfp-tub1, cbxR	(Straube et al, 2005)
potefYup1GFP	Potef-yup1-sgfp, cbxR	(Wedlich-Söldner et al, 2000)
potefYup1RFP2	Potef-yup1-2xmrfp, cbxR	This study
pKin3YFP	Potef-kin3-yfp, cbxR	(Wedlich-Söldner et al, 2002)
pPX10-146CFP	Potef-yup1PX1–146-cfp, bleR	(Wedlich-Söldner et al, 2002)
PH4CFP	Phis4-h4-cfp, hygR	(Straube et al, 2005)
a, b, mating type loci; Δ, deletion; P, promoter; ∷, homologous replacement; -, fusion; hphR, hygromycin resistance; bleR, phleomycin resistance; natR, nourseothricin resistance; cbxR, carboxin resistance; /, ectopically integrated; otef, constitutive promoter; nar, conditional nitrate reductase promoter; crg1, conditional arabinose-induced promoter; E1, W2, genes of the b mating type locus; gfp, enhanced green fluorescent protein; yfp, yellow-shifted fluorescent protein; mrfp, monomeric red fluorescent protein; yup1: endosomal t-SNARE; kin1, kinesin-1 family member; kin3, kinesin-3 family member; dyn1: N-terminal half of the dynein heavy chain; dyn2: C-terminal half of the dynein heavy chain; dyn2ts: temperature-sensitive allele of dyn2; tub1, α-tubulin; lis1, LIS1 homologue; clip1, CLIP-170 homologue; dya1, p150glued homologue; yup1PX1−146, localization domain of Yup1; his4, histone 4.

Figure 2.

Endosome organization and motility in the Kinesin-3 and dynein mutants. (A) EE that are labeled by a fusion protein of GFP and the membrane receptor Yup1 (arrowhead, arrow; Wedlich-Söldner et al, 2000) migrate bidirectioally along MTs, which are visualized by expression of RFP-α-tubulin. Elapsed time is given in seconds. Bar: 3 μm. (B) A quantitative analysis of the behavior of individual EE confirmed that most organelles reach the MT plus-ends before they reverse direction. (C) Quantitative analysis of the EE motility frequencies at 10 μm distance to the tip of control and mutant hyphae. Almost identical numbers of vesicles move towards the tip (anterograde) and the basal nuclear region (retrograde) of control hyphae. This motility is abolished in Kinesin-3 mutants (Δkin3) and in dynein mutants (Dyn2ts). (D) In control hyphae, Yup1-GFP labeled EE are motile and equally distributed along the length of the hyphae. Occasionally, transient endosome concentrations are found in the apex. Bar: 10 μm. (E) Δkin3 mutant hyphae are usually bipolar and much shorter (Schuchardt et al, 2005). The EE cluster near the nucleus, where the minus-ends of MTs are concentrated. Note that the MT orientation in Δkin3 hyphae was confirmed by analysis of Peb1-YFP motility (not shown). Bar: 10 μm. (F) In the conditional dynein mutants, depletion of dynein strengthens anterograde EE motility and leads to accumulations of EE at the MT plus-ends near the cell poles. Bar: 10 μm. Supplementary movie for Panels A and D–F is given on the EMBO web site.

Dynein accumulates at MT plus-ends in the hyphal tip

Next, we set out to visualize dynein by fusing a triple GFP-tag to the N-terminus of the endogenous dyn2 gene (strain AB33G₃Dyn2), which encodes the C-terminal half of the dynein heavy chain (Straube et al, 2001). The fusion protein did not cause any morphological phenotypes, indicating that it is fully functional. GFP₃-Dyn2 formed comet-like motile structures (Figure 3A, arrowhead; see Supplementary Movie 5) and was significantly concentrated at MT ends in the tip of growing hyphae (Figure 3A, arrow and Figure 3B; strain AB33G₃Dyn2_RT). From there, small dynein particles could be detected that move in retrograde direction (see below), suggesting that the dynein accumulation is a reservoir for retrograde motor proteins. Quantitative analysis of the fluorescence intensity revealed that the dynein signals in the first 3 μm of the hyphae were significantly stronger than those at MT tips in subapical regions (Figure 3C). A more detailed analysis revealed that the amount of dynein remained constant over long distances, but rapidly increased in the apical 5 μm and reached maximum values in the close proximity to the hyphal tip (Figure 3D), suggesting that dynein is actively concentrated in the apex. In vivo observation of EE in strain AB33G₃Dyn2_YR₂ revealed that EE reach the dynein accumulation before they reverse direction (Figure 3E, arrow; see Supplementary Movie 4).

Figure 3.

Dynein localization and EE movement. (A) In U. maydis hyphae dynein, visualized by a fusion of 3xGFP and the endogenous Dyn2 forms motile comet-like structures (arrowheads). The strongest signals are always found in the hyphal apex (arrow, inset). Bar: 10 μm. (B) Colocalization of GFP₃Dyn2 (Dyn, green in overlay) and RFP-α-tubulin labeled MTs (red) in control hyphae. Dynein localizes to the MT plus-ends that are reaching into the hyphal apex. Bar: 5 μm. (C) A quantitative analysis of the GFP₃-Dyn2 signal intensity at MT plus-ends in the first 3 μm of the hyphal tip (apical) and in at least 12 μm distance from the tip (basal) reveals that the dynein accumulations at the hyphal tip is more than twice as strong as in basal parts of the cell. Note that the signal intensity has been normalized to a single GFP. (D) The average intensity of dynein signals at MT plus-ends was measured in relation to distance to the tip, demonstrating that the amount of dynein remains constant over most of the length of the hypha, but rapidly increases at the hyphal apex. (E) Colocalization of GFP₃Dyn2 (Dyn, green in overlay series) and Yup-RFP₂ (EE, red in overlay series) demonstrates that EE (arrows) migrate into the dynein signal before they reverse direction (colocalization results in yellow). Elapsed time is given in seconds. Bar: 2 μm. Supplementary movie for Panel E is given on the EMBO web site.

Retrograde EE movement depends on activation of dynein by an LIS1 homologue

The accumulation of dynein in the hyphal tip suggested that the motor stays inactive until its cargo reaches the MT plus-ends. Therefore, we analyzed the mechanism of dynein activation and its impact on EE motility. A good candidate for dynein regulation is LIS1, which is thought to activate dynein (Faulkner et al, 2000; Smith et al, 2000; Zhang et al, 2003; Mesngon et al, 2006). The LIS1 homologue in U. maydis shares 47% amino-acid identity with its closest homologue NudF from A. nidulans (Xiang et al, 1999) and, like other LIS1-members, is predicted to contain an LisH motif, a coiled-coil region and seven WD40 domains (Figure 4A). We fused 3xGFP to the N-terminus of the endogenous lis1 gene, and this did not result in any defects in nuclear migration or growth (see below), indicating that the protein is fully functional. GFP₃-Lis1 localized to MT tips (Figure 4B; strain FB2GLis1_RT), and concentrates at plus-ends in the hyphal tip (Figure 4B and C), suggesting that the putative dynein activator is part of the apical dynein complex. Next, we analyzed the role of Lis1 in dynein activation. In strain FB2G₃Dyn2_nLis1, lis1 is expressed under the control of the inducible nar1-promoter (Brachmann et al, 2001). Upon growth in complete medium (CM) containing ammonium, this promoter is repressed and Lis1 is depleted (see Supplementary Figure 1 on EMBO web site), whereas it is induced in the presence of nitrate (NM). Reduction of Lis1 levels led to nuclear migration defects in yeast-like cells and longer MTs in hyphae (not shown). However, 87.5% of all plus-ends were still oriented to the hyphal tip (n=48; strain FB2nLis1_Peb1R_GT). Neither of these treatments affected dynein protein levels (see Supplementary Figure 1 on EMBO web site). However, the amount of dynein at MT plus-ends in the hyphal tip was significantly increased in the absence of Lis1, while it was slightly decreased upon overexpression of Lis1 (Figure 4D and E; Lis1↑: high protein levels, Lis1: no Lis1; FB2G₃Dyn2_nLis1_RT; see Supplementary Movie 5). Consistent with the concept that EE are loaded onto dynein at the hyphal apex and that Lis1 activates dynein, EE were not able to leave the tip in the absence of Lis1 (Figure 4F; overlay of two time points, stationary organelles appear in yellow; FB2G₃Dyn2_nLis1_YR₂; for Figures 4F, 5F and 6E see also Supplementary Movie 1). Consequently, EE motility was significantly impaired (Figure 4G; control strain: FB2YR₂; see Supplementary Movie 6). In summary, these results indicate that the LIS1-homologue Lis1 is crucial for the activation of dynein for retrograde EE transport.

Figure 4.

The role of a LIS1-homologue in dynein localization and endosome motility. (A) Lis1 from U. maydis (UmLis1, Accession Number: EAK84394) shares a similar domain structure with other LIS1-homologues, including those from A. nidulans (AnNudF, EAA57983), S. cerevisiae (ScPac1p, AAB00685) and human (HsLis1, NP 000421). (B) GFP-tagged endogenous Lis1 (green) localizes to plus-ends of RFP-α-tubulin containing MTs (red). Note that the tip contains the strongest signal. Bar: 5 μm. (C) Quantification of the GFP-Lis1 intensity at MT plus-ends in the apical 3 μm of the hyphal tip (apical) and in at least 12 μm distance from the tip (basal) demonstrates that Lis1 is concentrated at MT plus-ends in the tip. (D) The absence of Lis1 (Lis1↓) results in an increase of dynein at plus-ends within the hyphal tip. Overexpression of Lis1 (Lis1↑) reduces the apical dynein signal. Bar: 2 μm. (E) Quantitative analysis of GFP₃Dyn2 intensity at MT plus-ends revealed that the absence of Lis1 (↓Lis1) significantly increases the amount of dynein at MT plus-ends. Consistently, Lis1 overexpression (↑Lis1) has the opposite effect. (F) Overlay of two images at different time points illustrates the motility of RFP₂-tagged endosomes. Moving organelles appear in green or red, while stationary signals result in yellow color. Depletion of Lis1 led to an immobile accumulation of EE in the hyphal apex (strain FB2G₃Dyn2_nLis1_YR₂). Bar: 5 μm. Supplementary movies for Panel D and F are given on the EMBO web site. (G) A quantitative analysis of EE traffic reveals that overexpression of Lis1 (↑Lis1) affects EE motility only slightly, while nearly all motility is abolished in the absence of Lis1 (↓Lis1).

Figure 5.

The role of a p150^glued homologue in dynein localization and endosome motility. (A) The p150^glued homologues from U. maydis (UmDya1, Accession Number: EAK84712) shares a similar domain structure with other p150^glued homologues from A. nidulans (AnNudM, EAA58707), S. cerevisiae (ScNip100p, NP 015151) and human (Hsp150, NP 004073). (B) GFP₃-tagged endogenous Dya1 (green) accumulates at the plus-ends of RFP-α-tubulin labeled MTs (red). The strongest signal is present in the hyphal apex. Bar: 5 μm. (C) Fluorescence intensity measurements of GFP₃Dya1 at MT plus-ends in the first 3 μm of the hyphal tip (apical) and in basal parts of the cell (in >12 μm distance from the tip) reveals that dynactin was ∼2-times more concentrated at MT plus-ends in the hyphal tip than in subapical regions. (D) Quantification of the GFP₃Dyn2 signals at MT plus-ends and in the apical cytoplasm shows that high levels of Dya1 (↑Dya1) has no significant effect on dynein localization (P=0.306), whereas dya1 repression (↓Dya1) leads to a dramatic decrease of the dynein signal. (E) No dynein is present at apical MT plus-ends after depletion of Dya1 (Dya1↓). Bar: 2.5 μm. (F) Overlay of two images at different time points illustrates the motility of GFP-tagged endosomes in strain FB2rDya1_YG. Moving organelles appear in green or red, while stationary signals result in yellow. Depletion of Dya1 leads to an immobile accumulation of EE in the hyphal apex. Bar: 5 μm. Supplementary movies for Panels E and F are given on the EMBO web site. (G) The expression of Dya1 (↑Dya1) has no effect on EE transport, whereas the absence of Dya1 (↓Dya1) abolishes almost all EE motility.

Figure 6.

The role of a CLIP-170 homologue in dynein and endosome targeting. (A) The CLIP-170 homologues from U. maydis (UmClip1, Accession Number: EAK82811) shares a similar domain structure with other CLIP-170 members from S. cerevisiae (ScBik1p, NP 009901), Schizosaccharomyces pombe (SpTip1, P79065) and human (HsCLIP170, A43336). (B) GFP-tagged endogenous Clip1 (green) localizes to plus-ends of RFP-α-tubulin labeled MTs (red). Bar: 5 μm. (C) In contrast to dynein, dynactin and Lis1 intensities GFP-Clip1 fluorescence signals at MT plus-ends in the first 3 μm of the hyphae (apical) do not differ from those in basal parts of the cells (in >12 μm distance from tip). (D) In Δclip1 cells, the intensity of the dynein signal at MT plus-ends in the tip is not different from those of control cells. (E) In the clip1 deletion strain, GFP₃-tagged dynein (green) still accumulates at plus-ends of RFP-labeled MTs (red). Bar: 2.5 μm. (F) Deletion of clip1 does not affect bidirectional EE motility. (G) In Δclip1 mutants, the organization of GFP-α-tubulin labeled MTs is not altered compared to control cells (strain FB2GT). Bar: 5 μm. Supplementary movies for Panels E and G are given on the EMBO web site. (H) Parameters of MT dynamic instability in control cells and Δclip1 mutant hyphae.

Dynactin is essential for dynein plus-ends localization and retrograde EE transport

In order to further characterize the regulation of dynein, we investigated the role of the dynein activator dynactin (Schroer, 2004) in EE transport. We identified dya1, a putative component of the dynactin complex that shares 19% amino acid identity with its homologue from A. nidulans. Similar to other p150^glued homologues, Dya1 is predicted to contain a CAP-Gly motif, as well as numerous coiled-coil regions (Figure 5A). We fused a triple GFP-tag to the N-terminus of the endogenous dya1 gene and found that the fusion protein is functional, as it did not cause the dya1 mutant phenotype (see below). The fusion protein localized to the plus-ends of MTs (Figure 5B; strain FB2G₃Dya1_RT), with most dynactin being concentrated at the apical MT ends (Figure 5B and C), which corresponds well with the data raised for dynein and Lis1 (see above). Next, we generated conditional mutants that expressed dya1 under the control of the strong crg1-promoter (Bottin et al, 1996). In the presence of arabinose, this promoter allowed expression of GFP-Dya1, whereas in the presence of glucose it was repressed and no fusion protein was detected (see Supplementary Figure on EMBO web site; strain FB2rGDya1). These growth conditions neither affected the orientation of MTs (86.2% of plus-ends grow towards the hyphal tip, n=29; strain FB2rDya1_Peb1R_GT) nor changed the levels of dynein in cell extracts of strain FB2rDya1_G₃D (see Supplementary Figure on EMBO web site). Depletion of Dya1 abolished dynein accumulation in the tip (Figure 5D and E; Dya1↓: no Dya1, Dya1↑: high levels of Dya1; strain FB2rDya1_G₃D_RT; see Supplementary Movie 5) and this led to an EE aggregation in the hyphal apex (Figure 5F; overlay of two time points, stationary signals appear in yellow) and a significant reduction in EE motility (Figure 5G; strain FB2rDya1_YG; see Supplementary Movie 6). These results demonstrate that the dynactin subunit Dya1 is essential for setting up a dynein reservoir at the hyphal tip, which in turn is required for retrograde EE traffic. Hence, these results add further support to our concept that organelles are loaded onto dynein in the hyphal tip.

Clip1 is not necessary for dynein plus-ends binding and EE movement

In vertebrates, the cytoplasmic linker protein CLIP-170 links endocytic membranes (Pierre et al, 1992) and the dynactin complex, and thereby indirectly dynein to MT plus-ends (Valetti et al, 1999; Lansbergen et al, 2004). Thus, we investigated the role of Clip1, the U. maydis CLIP-170 homologue, in EE traffic and dynein targeting. Although sequence similarity between Clip1 and human CLIP-170 is low (15% amino acid identity), Clip1 is predicted to have approximately the same size and contains a CAP-Gly motif flanked by two Serine-rich regions, as well as numerous coiled coil regions and two C-terminal zinc-finger like domains (Figure 6A). In addition, a fusion protein of GFP and Clip1 localized to MT plus-ends (Figure 6B; strain FB2GClip1_RT), which is consistent with its predicted role in targeting organelles or dynein/dynactin to MT tips. However, in contrast to dynein, Lis1 and Dya1, the CLIP-170 homologue showed similar signal intensities at apical and subapical MT plus-ends (Figure 6C), which argues against a role in dynein/dynactin anchoring. In order to characterize the role of Clip1 at the plus-ends, we constructed a clip1 deletion mutant (strain FB2ΔClip1) and confirmed the absence of the gene by Southern blot and RT–PCR (not shown). The absence of Clip1 had no effect on the amount of dynein at MT plus-ends in the tip (Figure 6D and E; strain FB2ΔClip1_G₃Dyn2_RT; see Supplementary Movie 5) and neither anterograde nor retrograde EE traffic was inhibited (Figure 6F; FB2ΔClip_YG; see Supplementary Movie 6). Surprisingly, deletion of Clip1 also did not have a significant effect on MT organization and polarity either (Figure 6G; strains FB2ΔClip_GT; FB2ΔClip_Peb1R_GT) and only pausing rates of MTs were affected (Figure 6H). Taken together, these data show that the CLIP-170 homologue is neither involved in dynein anchoring nor in endosome targeting. In addition, Clip1 has only minor roles in MT dynamic instability.

Kinesin-1-based targeting of dynein/dynactin to MT plus-ends promotes minus-end directed endosome motility

The results presented so far were consistent with the notion that EE are loaded onto dynein at the apex. Therefore, we next asked how this apical dynein loading zone is established. In A. nidulans Kinesin-1 participates in plus-ends targeting of dynein (Zhang et al, 2003), and we considered it possible that U. maydis Kinesin-1 (Lehmler et al, 1997) also delivers dynein/dynactin to MT plus-ends. In Kinesin-1 null mutants, MT orientation was not altered (91% of all plus-ends directed towards hyphal apex; n=131; AB33ΔKin1_Peb1Y). However, no GFP-tagged dynein was found at the MT tips in the apex of these hyphae, but concentrated at subapical regions (Figure 7A, inset shows hyphal tip, strain AB33ΔKin1_G₃D2_RT; see Supplementary Movie 7). A similar distribution was found for dynactin (Figure 7B; strain FB2ΔKin1_G₃Dya1_RT), indicating that Kinesin-1 delivers a complex of dynein and dynactin to MT plus-ends in the hyphal tip. In contrast, the Lis1 accumulation at MT tips was only partially reduced in Kinesin-1 null mutants (not shown), suggesting that Lis1 is targeted by other mechanisms, which might also involve Kinesin-1. Consistent with the concept that dynein has to accumulate at the hyphal tip for retrograde transport, EE were trapped at apical MT plus-ends in Δkin1 hyphae (Figure 7C; strain AB33ΔKin1_YG), and almost no EE motility was observed (Figure 7D). This demonstrates that the apical localization of dynein/dynactin is required for retrograde EE transport.

Figure 7.

Kinesin-1-based targeting of dynein and dynactin to MT plus-ends in the hyphal tip. (A) Colocalization of GFP₃Dyn2 (Dyn, green in overlay) and RFP-α-tubulin labeled MTs (Tub, red in overlay) in Δkin1 mutants reveals that dynein colocalizes with MTs in the subapical region. However, dynein does not reach the plus-ends in the hyphal tip (insets). Bar: 5 μm. (B) No GFP₃Dya1 (green) appears at apical MT plus-ends (red) in Kinesin-1 null mutants, demonstrating that dynactin requires Kinesin-1 for long-distance plus-end targeting. Bar: 5 μm. (C) In Δkin1 hyphae EE cluster in the hyphal tip, indicating that deletion of Kinesin-1 strengthened plus-end directed transport, which is due to a lack of dynein in the hyphal tip. Bar: 5 μm. (D) Quantitative analysis of EE motility in the apical 10 μm of control and Kinesin-1 null mutant hyphae. Almost all motility is abolished in the absence of Kinesin-1. (E) Linescan analysis of intensities of GFP₃-Dyn2 and Yup1-RFP₂ in Δkin1 mutant hyphae (strain AB33ΔKin1_G₃D2_YR₂). In the absence of Kinesin-1, EE travel to the tip but are trapped there. Dynein is not enriched in the endosome cluster, which argues against a direct binding of the motor to anterogradely moving organelles. Supplementary movies for Panels A and C are given on the EMBO web site.

Dynein locates only on retrogradely moving endosomes

In a final set of experiments, we set out to obtain direct evidence for our hypothesis that EE are loaded onto dynein in the hyphal tip. Consistent with our hypothesis that dynein does not travel on EE towards the hyphal apex, we found no increase in dynein signals in the apical cluster of EE that are trapped in the hyphal tip of Δkin1 hyphae (Figure 7E). In addition, we did colocalization studies of EE and the functional GFP-fusion proteins. We analyzed the existing strains AB33G₃Dyn2_YR₂ and RWS9, which contains Kinesin-3YFP and CFP-labeled EE (Wedlich-Söldner et al, 2002). In addition, we generated mutant strains for Dya1 and Lis1 (FB2G₃Dya1_YR₂ and FB2G₃Lis1_YR₂; see Table I). The experiments demonstrated that 88.5% of all faint GFP₃-dynein signals (n=52) leave the dynein accumulation for retrograde traffic (Figure 8A, colored arrows indicate individual moving signals; see Supplementary Movie 8). Consistently, dynein and dynactin were found to move on retrogradely travelling EEs (Figure 8B; Dynein, Dynactin). A quantitative analysis demonstrated that 100% of the backward moving EE carried dynein (n=20; 10 cells), while GFP₃-Dyn2 was not found on anterogradely moving EE (n=14; 12 cells; Figure 8C). In contrast, GFP₃-labeled Lis1 was found on 10% of the tipward moving EE, but localized to only 18.2% of the retrograde EE (Figure 8C) and often left the EE immediately after it started moving (Figure 8B; Lis1). Finally, we found that YFP-labeled Kinesin-3 colocalized with all moving EE in both directions (Figure 8C). Taken together, these results support our hypothesis and strongly suggest that dynein and dynactin bind to endosomes at the MT plus-ends and move organelles retrogradely. In contrast, Lis1 appears to have only transient contact with retrograde endosomes.

Figure 8.

Dynein, dynactin, Lis1 on moving early endosomes. (A) In vivo observation of functional GFP₃-Dyn2 at the hyphal tip demonstrates that dynein is continuously released from its apical accumulation. Individual signals marked by colored arrows. Interval between frames is 1 s. Bar: 1 μm. (B) Colocalization studies on strains expressing Yup1-RFP₂ and GFP-tags fused to dyn2 (dynein), dya1 (dynactin) and lis1 (Lis1) demonstrate that the dynein/dynactin complex localizes to retrogradely moving organelles, whereas Lis1 only transiently appears on EE. Bars: 1 μm. (C) Quantitative analysis of EE that move for at least 3 μm. Kinesin-3 locates on all EE, whereas dynein was only found on retrogradely moving organelles. Supplementary movie for Panel B is given on the EMBO web site.

Discussion

Bidirectional motility of organelles is a very common phenomenon, which was found in various organelles, including endosomes, secretory vesicles, lysosomes and mitochondria (Gross 2004; Welte, 2004). Bidirectional motility was also described for endosomes in yeast-like cells of U. maydis (Wedlich-Söldner et al, 2000, 2002). In this plant pathogenic fungus EE participate in hyphal growth (Wedlich-Söldner et al, 2000), and hyphal growth is essential for host invasion and pathogenic development (Weber et al, 2003). Thus, we set out to understand the molecular mechanism of bidirectional EE motility in hyphae. We used the tSNARE Yup1, which was found to be located on rapidly moving organelles that were initially identified as EE based on their affinity to the endocytic dye FM4-64 (Wedlich-Söldner et al, 2000). Recent studies demonstrated that endosome specific Rab5-GTPases colocalize with Yup1-GFP in U. maydis (U Fuchs, I Schuchardt, G Steinberg, unpublished), which confirms that the Yup1-RFP₂ positive organelles are EE.

The role of plus-end binding proteins in regulation of dynein function

The results presented here support the idea that dynein at MT plus-ends is inactive until an organelle reaches the tip (Figure 9). A well-studied dynein interacting protein is NudF/Lis1 (Xiang et al, 1999), which is thought to regulate dynein activity in various cell types (Faulkner et al, 2000; Smith et al, 2000; Coquelle et al, 2002; Zhang et al, 2003; Lansbergen et al, 2004; Mesngon et al, 2006). Here, we provide in vivo evidence for such a function. High levels of Lis1 reduced the dynein signal at the tip, whereas the depletion of Lis1 significantly increased the amount of dynein at apical MT plus-ends, which coincided with an inhibition of retrograde EE motility. Interestingly, Lis1 accumulates at the plus-ends in amounts similar to that of dynein, indicating that Lis1 is part of the inactive dynein complex. This suggests that the effects of Lis1 on dynein are regulated by other upstream factors. Such components could travel on the arriving EE and trigger the retrograde transport machinery (Figure 9), which might involve phosphorylation of Lis1 (Coquelle et al, 2002).

Figure 9.

Dynein in minus-end directed traffic of early endosomes. (1) In the current model, Kinesin-1 takes dynein/dynactin to the MT plus-ends in the hyphal tip. Preliminary evidence indicates that a portion of Lis1 might hitchhike on the transported dynein/dynactin complex (not shown), but other mechanisms are also likely. (2) An inactive complex of dynein, dynactin and Lis1 accumulates at the plus-ends close to the growth region of the hypha. (3) Kinesin-3 transports EE along MTs to the inactive dynein/dynactin/Lis1 complex, thereby delivering an unknown activator of Lis1. (4) When EE reach the plus-ends at the hyphal apex they quickly exchange material with the growing tip. This might involve the uptake of material for transport towards the cell body as well as local membrane recycling processes. (5) Subsequently, the unknown activator triggers Lis1-dependent activation of the dynein/dynactin complex, which results in retrograde motility of the EE.

Another activator of dynein is the dynactin complex (Schroer, 2004). Initially discovered as a factor involved in dynein-dependent organelle traffic in vitro (Gill et al, 1991), it recently emerged as a factor that increases the processivity of dynein (King and Schroer, 2000) and participates in linking organelles to MTs via its MT binding region (Vaughan et al, 2002). In mammalian cells, dynactin concentrates at MT plus-ends (Valetti et al, 1999; Vaughan et al, 1999), where it is thought to mediate loading of endosomes to dynein (Vaughan, 2005). Indeed, our results demonstrate that an intact dynactin complex is required for the accumulation of dynein at plus-ends and, consequently, for retrograde EE transport. In contrast, dynactin is not required for plus-ends targeting of dynein in S. cerevisiae (Sheeman et al, 2003). This contradiction might result from a different targeting mechanism, as both dynein and dynactin are delivered to the hyphal apex by Kinesin-1, whereas this kinesin is not even present in bakers yeast. However, further studies on the mechanism of plus-ends targeting of both proteins are required to elucidate the molecular details of delivery of dynein/dynactin to MT plus-ends.

We found that Clip1, a homologue of mammalian cytoplasmic linker protein 170 (CLIP-170), is neither involved in dynein targeting nor in EE traffic. This finding is most surprising, as CLIP-170 takes part in anchoring of dynein/dynactin in mammals (Coquelle et al, 2002; Lansbergen et al, 2004) and yeast (Sheeman et al, 2003), but also anchors endocytic membranes to MT plus-ends (Pierre et al, 1992; Vaughan, 2005). Members of the CLIP-170 protein family have also been shown to regulate MT dynamics as anticatastrophe factors in fungi (Berlin et al, 1990; Brunner and Nurse, 2000; Efimov et al, 2006), and are rescue factors in mammalians (Komarova et al, 2002). However, the deletion of Clip1 had no effect on MT organization and only increased pausing rates of MTs in U. maydis hyphae. Thus, it appears that CLIP-170 like proteins perform various cellular roles in the different systems, and it is tempting to speculate that this variation reflects the complex interaction of the numerous plus-ends binding proteins in organizing or regulating the plus-ends complex.

A dynein-loading site at the plus-ends of MTs in the hyphal tip

We previously reported that EE travel to the tip of U. maydis hyphae, where they are thought to support tip growth by membrane recycling (Wedlich-Söldner et al, 2000). The major finding of this study is that EE reach the plus-ends in the hyphal tip by the activity of Kinesin-3, where they are apparently loaded onto an inactive dynein/dynactin/Lis1 complex that becomes activated. The results presented here demonstrate that dynein/dynactin takes endosomes, and presumably inactive Kinesin-3, towards the cell body, while most Lis1 does not bind endosomes or leaves early after reversal of transport direction (Figure 9). This model is supported by several lines of evidence: (1) MT plus-ends in the hyphal tip carry the strongest dynein, dynactin and Lis1 signals; (2) most EE move into the dynein/dynactin/Lis1 accumulation at MT plus-ends before they reverse direction; (3) EE are trapped at plus-ends in the hyphal tip when dynein is not activated by Lis1; (4) EE cluster at plus-ends in the tip when dynein targeting to the tip is impaired in Kinesin-1 null mutants or when dynactin is depleted; and (5) dynein/dynactin colocalizes with retrogradely but not anterogradely moving EE. Taken together, these results strongly suggest that dynein and associated factors are targeted to the hyphal tips by Kinesin-1, where dynein binds to Kinesin-3 delivered EE (Figure 9). Recent evidence from Drosophila oocytes provide good indication that a similar mechanism of dynein targeting exists in animal cells. Here, dynein/dynactin localization and function also depends on the activity of Kinesin-1 (Duncan and Warrior, 2002; Januschke et al, 2002). In mammals, plus-ends located dynactin is thought to mediate targeting of membranous cargoes to the MT tip, which is followed by the recruitment of dynein and initiation of retrograde transport (Vaughan et al, 2002; Vaughan, 2005). Thus, mammalian dynein is usually absent from MT plus-ends (Vaughan et al, 1999), whereas an already assembled complex of dynein/dynactin/Lis1 is waiting at plus-ends of interphase MTs in U. maydis (this study) and A. nidulans (Han et al, 2001; Zhang et al, 2003). These results suggest that some of the basic principles of bidirectional organelle traffic differ between cell types. The reason for the situation in fungi might lie in the way fungal hyphae expand.

The role of the dynein loading zone in hyphal tip growth

In U. maydis, a strong dynein accumulation is found in the growing tip of the hypha that waits for a Kinesin-3-delivered organelle in order to be activated and ‘off-loaded' onto the organelle. In contrast, in mammalian systems, kinesin and dynein motors are often located on the same organelle (Gross, 2004; Ligon et al, 2004; Welte, 2004). Having both motors on a single organelle appears to be an economic way to mediate bidirectional motility, which raises the question why filamentous fungi have established an apical dynein loading zone. The answer to this question might be the key to an understanding of fungal tip growth. Hyphae are surrounded by a rigid cell wall that is synthesized and modified by polarized exocytosis (Gow, 1995) and local membrane recycling (Wedlich-Söldner et al, 2000) at the growing tip. One way to ensure that organelles reach the tip is that retrograde motors wait there to take organelles backwards. If this is true, hyphal tip growth and the existence of an apical dynein loading zone are intimately linked and most likely co-evolved. During evolution, bidirectional transport along MTs was lost in the yeast S. cerevisiae (Madden and Snyder, 1998), but the same principal of dynein waiting at plus-ends to be activated in order to move towards the minus-end is still used for nuclear migration (Lee et al, 2003; Sheeman et al, 2003). Taken together, we speculate that the establishment of a dynein loading zone in the hyphal apex was a crucial step in fungal evolution, which is essential for tip growth and, therefore, the evolutionary success of this important organismic group.

Materials and methods

Strains and plasmids

MTs were labeled with ectopically expressed RFP-α-tubulin or GFP-α-tubulin (Steinberg et al, 2001), while EE were marked by a fusion protein of GFP or a 2xRFP protein fused to an endosome specific t-SNARE (Wedlich-Söldner et al, 2000), respectively. Dynein and dynactin were visualized with a 3xGFP-tag fused to the N-terminus of the endogenous dyn2 or dya1 gene, while the endogenous Clip1 and Lis1 were labeled by a single GFP. In order to deplete dya1 or dyn1, both genes were expressed under the control of the crg1 promoter that is repressed in glucose (Bottin et al, 1996), while lis1 was expressed under control of the nar1 promoter (Brachmann et al, 2001). Deletion of clip1 was performed by homologous replacement of the whole open reading frame with a phleomycin resistance cassette. Further details on U. maydis strains and plasmids are listed in Table I.

Growth condition

Strains were grown over night at 28°C in CM (Holliday, 1974) supplemented with 1% glucose (CM-G) or 1% arabinose (CM-A). Solid medium contained 2% (w/v) bacto-agar. Induction of hyphal growth in AB33-derivatives, in which the filament-inducing bE/bW transcription factor is under control of the nitrate inducible nar1 promoter, was done as previously described (Fuchs et al, 2005). Hyphal growth of strains derived from FB1 or FB2 was induced by synthetic pheromone (Fuchs et al, 2005). AB33rDyn1_YG was grown over night in CM-A, and the crg1-promoter was repressed and hyphal growth induced in glucose containing nitrate minimal medium (Holliday, 1974) for 16–22 h at 28°C, 200 r.p.m. Strains that allowed controlled expression of lis1 (FB2nGLis1 and FB2G₃Dyn2_nLis1 and derivates) and dya1 (FB2rDya1_G₃D and derivates, FB2rDya1_YG and FB2rGDya1) were grown over night in CM-G, washed in fresh medium, followed by resuspension in 0.5 ml NM-G or CM-A, respectively. Hyphal growth was induced by synthetic pheromone 6–8 h at 22°C and 200 r.p.m.

Sequence analysis

U. maydis homologues of LIS1, CLIP-170 and p150^glued were identified in the public databases by using PubMed (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi). Domain analysis was performed using SMART, Coils and MotifScan (all accessible on ExPASy web site http://us.expasy.org/).

Microscopy, image processing and quantitative analysis

Logarithmically growing cells were embedded in low melt agarose or placed on top of a 2% agar cushion, and were immediately observed using a Zeiss Axioplan II imaging microscope (Oberkochen, Germany) and filter sets for YFP (BP500/20, FT515, BP535/30), CFP (BP436, FT455, BP480-500), FITC (BP470/20, FT510, BP515-565), and DsRed (BP565/30, FT585, BP620/60). Image aquisition was carried out using a CoolSNAP-HQ CCD camera (Photometrics, Tocsen, USA) controlled by the imaging software MetaMorph (Universal Imaging, Downing Town, USA). All image processing, including adjustment of brightness, contrast and gamma-values, was performed with MetaMorph and Photoshop (Adobe). Two-tailed t-tests were carried out using Prism (GraphPad). Signal intensities of GFP₃-Dyn2, GFP-Lis1, GFP₃-Dya1 and GFP-Clip1 were measured using MetaMorph. The average intensity within at MT plus-ends was corrected for the cytoplasmic background or the extracellular background. Endosome motility was analyzed in image sequences of 20–60 frames, taken with an exposure time of 500–1000 ms. The anterograde and retrograde EE movements were counted in a distance of about 10 μm from the hyphal tip. Only organelles that were moving for at least 3 μm were considered. Quantitative analysis of MT dynamics was performed essentially as described (Steinberg et al, 2001). The pausing rate was defined as the time in percent that an MT plus-ends does neither elongate nor shorten. MT orientation in different areas of the hyphae was determined essentially as described (Straube et al, 2003).

Supplementary Material

Supplementary Figure on EMBO web site

Supplementary Figure on EMBO web site; strain FB2rGDya

Movie 1

Movie 2

Movie 3

Movie 4

Movie 5

Movie 6

Movie 7

Movie 8

Movie Legends