bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1 Dynlt3, a novel fundamental regulator of melanosome movement,

2 distribution, maturation and transfer

3

4

5

6 Zackie Aktary1-3, Alejandro Conde-Perez1-3, Florian Rambow1-3, Mathilde Di 7 Marco6,7, François Amblard3-5, Ilse Hurbain6,7, Graça Raposo6,7, Cédric 8 Delevoye6,7, Sylvie Coscoy3,4, & Lionel Larue1-3* 9

10

11

12 1 INSERM U1021, Normal and Pathological Development of Melanocytes. Institut 13 Curie, PSL Research University, Orsay, France 14 2 CNRS UMR3347, Paris-Sud University, University Paris-Saclay, Orsay, France 15 3 Equipes Labellisées – Ligue Contre le Cancer, Orsay, France 16 4 Laboratoire Physico-Chimie Curie, Institut Curie, PSL Research University - 17 Sorbonne Universités, UPMC - CNRS. Paris, France. 18 5 Center for Soft and Living Matter, Institute for Basic Science (IBS), Departments of 19 Bioengineering and Physics, Ulsan National Institute of Science and Technology, 20 Ulsan, South Korea. 21 6 CNRS UMR144, Structure and Membrane Compartments. Institut Curie, Paris 22 Sciences & Lettres Research University, Centre National de la Recherche 23 Scientifique, Paris, France 24 7 CNRS UMR144, Cell and Tissue Imaging Facility. Institut Curie, Paris Sciences & 25 Lettres Research University, Centre National de la Recherche Scientifique, Paris, 26 France 27

28

29 *Address correspondence to Lionel LARUE 30 e-mail: [email protected] - Tel: (33) 1 69 86 71 07 - Fax: (33) 1 69 86 71 09 31 Institut Curie – Bat 110 - 91405, Orsay Cedex, France 32

33

1 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

34

35

36 ABSTRACT

37

38 Skin pigmentation is dependent on cellular processes including melanosome

39 biogenesis, transport, maturation and transfer to keratinocytes. However, how the

40 cells finely control these processes in space and time to ensure proper pigmentation

41 remains unclear. Here, we show that a component of the cytoplasmic dynein

42 complex, Dynlt3, is required for efficient melanosome transport, maturation and

43 transfer. In melanocytes with decreased levels of Dynlt3, pigmented melanosomes

44 undergo a more directional (convective) motion leading to their peripheral location in

45 the cell, and are not fully matured. Stage IV melanosomes are more acidic, but still

46 heavily pigmented, resulting in a less efficient melanosome transfer. Finally, the level

47 of Dynlt3 is dependent on β-catenin activity, revealing a novel function of the Wnt/β-

48 catenin signalling pathway during melanocyte and skin pigmentation, by coupling the

49 transport, position and maturation of melanosomes required for their transfer.

50

51

2 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

52 INTRODUCTION

53 Melanocytes are neural crest derived cells that are primarily found in the

54 epidermis and hair follicles, but are also found in the inner ear, eye, heart and

55 meninges 1. These cells contain specialized organelles, known as melanosomes,

56 which are responsible for pigmentation. Melanosomes belong to the lysosome

57 related organelle (LRO) family and exist in four different stages 2. Only stage III and

58 IV melanosomes are pigmented. Stage I melanosomes, the most primitive

59 melanosomes, are membrane bound early endosomal structures that initiate the

60 intraluminal Pmel fibrillation. The Pmel fibrils are fully formed in stage II

61 melanosomes, following which melanin synthesis begins. In stage III, different

62 melanogenesis enzymes (e.g. Tyrp1, Dct) are transported and inserted into the

63 melanosome limiting membrane to produce melanin, which deposits on the pre-

64 existing Pmel fibrils. Melanin accumulates in stage IV melanosomes that will become

65 heavily pigmented. Melanosomes/melanin are then transferred to keratinocytes

66 through an exocytosis/phagocytosis process to fulfill their primary role in protecting

67 the skin from UV irradiation 3. Prior to melanosome exocytosis and melanin transfer,

68 melanosomes must acquire a secretory signature through maturation steps that are

69 still poorly understood. Maturation is a complex process that includes the

70 melanosome lumen deacidification or neutralization, the acquisition of specific

71 endosomal components by the melanosomes, and the concomitant removal of

72 melanosome-associated membrane proteins (Vamp7 and most likely Tyrp1) through

73 the generation and release of melanosomal membrane tubules 4-6.

74 The intracellular trafficking of melanosomes, similar to other LROs, involves

75 the action of different families of motor proteins. Along the microtubules,

76 melanosomes are transported by both kinesin and dynein motors 7. Most kinesins

3 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

77 move their respective cargoes from the nucleus towards the periphery, whereas

78 dynein motors move their cargoes in the opposite direction 8. At the cell periphery,

79 melanosomes are transferred to the actin microfilaments, where the complex

80 consisting of Myosin Va, Rab27a and Melanophilin move the melanosomes

81 centrifugally to be transferred. While mutations in these three genes lead to

82 pigmentation defects in humans and in mice, to date, no mutation in dynein and

83 kinesin genes has been described in pigmentation.

84 In humans and mice, abnormal skin pigmentation results from a reduction/lack

85 of melanocytes, a defect in melanin synthesis, and/or disrupted melanosome

86 transport/maturation/transfer 9. We have previously shown that alterations of β-

87 catenin activity (gain [bcat*] and loss of [∆ex2-6]) result in the reduction of

88 melanocyte proliferation and a coat color phenotype in mice 10,11.

89 In the current study, using melanocyte cell lines established from the skin of

90 wild-type and bcat* C57BL/6J mice in culture, we observed a marked diminution of

91 the number of pigmented melanosomes in the perinuclear area of bcat*

92 melanocytes. We identified Dynlt3, a member of the cytoplasmic dynein complex as

93 a downregulated component in bcat* melanocytes compared to wild-type, and as a

94 regulator of melanosome movement, distribution, maturation and transfer. Our data

95 demonstrate, for the first time, the fundamental role of Dynlt3 in coat and skin

96 coloration.

97

4 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

98 RESULTS

99

100 Melanocytes expressing stable β-catenin display peripheral melanosome

101 distribution

102 We established melanocyte cell lines in culture from the skin of wild-type (9v)

103 and bcat* (10d) C57BL/6J pups. As expected, WT pigmented melanosomes (mainly

104 stages III and IV) were distributed uniformly throughout the cell (Fig. 1a,b, left).

105 However, melanosome localization was markedly different in bcat* melanocytes,

106 which displayed a notable absence of pigmented melanosomes in the perinuclear

107 area (Fig. 1a,b, right), while the total number of melanosomes in WT and bcat* cells

108 was similar (Fig. 1c). The quantification of the number of perinuclear melanosomes

109 in a 15 μm area around the nucleus revealed that ≈ 20% of the pigmented

110 melanosomes were perinuclear in WT cells, compared to only 5% in bcat* cells (Fig.

111 1d). Immunofluorescence microscopy using anti-Tyrp1 antibodies- mainly labeling

112 stages III and IV melanosomes 12- further confirmed the reduction of pigmented

113 melanosomes in the perinuclear area of bcat* cells as compared to WT melanocytes

114 (Fig. 1e).

115 To confirm that the observed reduction of perinuclear melanosomes in this

116 mutant cell line was not limited to the specific bcat* construct, we expressed another

117 active β-catenin mutant, the βcat-∆ex3-GFP mutant, in WT melanocytes 13. This

118 mutant construct encodes for a β-catenin protein that lacks the serine and threonine

119 residues that allow for its proteasomal degradation. Similar to the bcat* cells,

120 expression of βcat-∆ex3-GFP in WT melanocytes (WT-βcatGFP, fluorescent cells)

121 resulted in a reduced number of perinuclear melanosomes compared to a GFP

122 control (WT-GFP, fluorescent cells) (Supplementary Fig. 1).

5 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

123 Pigmented melanosomes consist of stage III and IV melanosomes.

124 Ultrastructurally, stage III melanosomes contain dark and thick intraluminal melanin-

125 positive fibrils, while stage IV melanosomes are fully filled with pigment. Melanosome

126 maturation was evaluated in WT and bcat* cells by transmission electron microscopy

127 (TEM) assay. Intriguingly, pigmented bcat* melanosomes were significantly larger

128 than in WT melanosomes (≈370 nm vs. ≈300 nm) (Supplementary Fig. 2).

129

130 β-catenin affects Dynlt3 expression

131 The notable peripheral melanosome distribution in bcat* cells very likely

132 reflected a disruption in their proper intracellular trafficking, due to the exogenous

133 expression of β-catenin. To examine this possibility, we looked at the transcriptome

134 of WT and bcat* melanocytes and specifically focused on the expression of 67 genes

135 known to be involved in Lysosomal Related Organelle (LRO) biogenesis, transport

136 and maturation (Fig. 2a and Table 1). From these analyses, we identified Dynlt3, a

137 member of the cytoplasmic dynein complex, as a gene whose expression was

138 decreased two-fold in bcat* cells compared to WT cells.

139 Transcriptomic data were confirmed by RT-qPCR and western blot analyses

140 and showed that both Dynlt3 mRNA and protein levels were decreased in bcat* cells,

141 relative to WT cells (Fig. 2b,c). These results were corroborated using independent

142 melanocyte cell lines established from independent WT and bcat* pups

143 (Supplementary Fig. 3a). The regulation of Dynlt3 levels by β-catenin was then

144 validated using different approaches. We first overexpressed βcat-∆ex3-GFP in WT

145 cells, following which Dynlt3 mRNA and protein levels were both significantly

146 decreased (Fig. 2d-g). Subsequently, we knocked down β-catenin in WT cells and

147 observed an increase in both Dynlt3 mRNA and protein (Fig. 2h-k). Finally, the

6 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

148 knockdown of β-catenin in bcat* melanocytes resulted in a similar increase in Dynlt3

149 mRNA and protein levels (Fig. 2l-o), thereby clearly demonstrating that β-catenin

150 expression appears to have an inhibitory effect on the levels of Dynlt3.

151

152 Knockdown of Dynlt3 in WT melanocytes phenocopies melanosome

153 distribution in bcat* melanocytes

154 Given that Dynlt3 is a member of the Tctex-type 3 family of Dynein motor light

155 chain implicated in retrograde transport 14,15 and that its expression was decreased in

156 bcat* cells, we wondered whether Dynlt3 was a player involved in melanosome

157 distribution. To this end, we reduced the levels of Dynlt3 in WT melanocytes. Similar

158 to bcat* cells, knockdown of Dynlt3 phenocopied the exclusion of pigmented

159 melanosomes from the perinuclear area (Fig. 3a-c). The number of perinuclear

160 melanosomes was decreased two-fold in WT-siDynlt3 cells as compared to control

161 (Fig. 3d), while the total number of melanosomes was unchanged (Fig. 3e). As

162 previously, WT-siDynlt3 cells stained with an anti-Tyrp1 antibody revealed a notable

163 absence of staining in the perinuclear area, which was not the case in WT-siCtrl

164 melanocytes (Fig. 3f). To confirm these results, we overexpressed Dynlt3 in bcat*

165 melanocytes, which resulted in the redistribution of the melanosomes in the

166 perinuclear area (Fig. 4).

167

168 Decreased Dynlt3 amounts increased melanosome motility.

169 We next examined potential alterations in melanosome motility using time-

170 lapse video microscopy to track pigmented melanosome movement. Direct

171 observation of melanosome movements revealed that generally motion is composed

7 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

172 of variable and intermittent movement, with high level of pauses followed by short

173 bursts of movements, as previously reported 16,17.

174 The movement of melanosomes in bcat* cells was enhanced when compared

175 with WT melanosomes. On average, the total distance that bcat* melanosomes

176 traveled was longer than WT melanosomes (24.0 μm vs. 11.8 μm, Fig. 5a,b).

177 Similarly, βcat-Δex3-GFP WT cells showed increased total distance travelled by

178 melanosomes relative to WT cells expressing GFP alone (21.4 μm vs. 11.1 μm, Fig.

179 5a,b). Additionally, following transfection with GFP or βcat-Δex3-GFP, we monitored

180 the trajectories of melanosomes in GFP negative cells (i.e. that were unsuccessfully

181 transfected). The total distance that the melanosomes travelled on average in these

182 two non-GFP cells (GFP or βcat-Δex3-GFP) was similar to WT cells (11.8 μm vs.

183 11.0 μm vs. 11.3 μm, respectively, Fig. 5a,b). Taken together, β-catenin exogenous

184 expression increased movement of melanosomes.

185 Similarly, downregulation of Dynlt3 in WT melanocytes stably expressing

186 shRNA targeting this gene resulted in a significant decrease of Dynlt3 and an

187 increase in the total distance melanosomes travelled compared to shCtrl cells

188 (17.9 μm vs. 11.9 μm, Fig. 5a,b), accompanied by an increased average distance

189 and average velocity of melanosomes (Supplementary Fig. 4). Consistently,

190 Euclidean distance travelled by the melanosomes was increased after

191 overexpression of β-catenin or downregulation of Dynlt3 (Fig. 5b,c).

192 Finally, the percentage of pausing melanosomes was determined by

193 considering the number of frames where no movement was recorded. Tracked

194 melanosomes in bcat*, WT-βcat-Δex3 or WT-shDynlt3 cells paused less frequently

195 than controls (Fig. 5b,d), showing together that Dynlt3, through bcat expression,

196 contributes to at least the intracellular dynamic of pigmented melanosomes.

8 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

197

198 Mathematical and statistical analyses reveal a convective melanosome

199 movement after reduction of Dynlt3 levels

200 Displacement of an object can be generated by a superposition of various

201 elementary processes: Brownian movement (diffusion) as a basis process, with the

202 potential addition of velocity-driven movement (convection or directed movement),

203 elastic tethering (flexible attachment to a fixed point) and/or confinement. “Not

204 convective” conditions refer here to populations of trajectories that could be

205 generated without addition of a velocity-driven movement (Fig. 6a,b). In order to

206 identify the elementary physical processes involved in melanosome movement, we

207 compared experimental melanosome trajectories (x, y, t) with simulated trajectories

208 integrating these different types of movements 18. This comparison is based on eight

209 spatio-temporal descriptors (see supplementary information). Here, we focus on a

210 descriptor referring to the classical analysis of the Mean-square displacement (MSD

211 or ) during a time interval (t).

212 In the case of purely 2D diffusive motion, MSD(t) = 4Dt (D, diffusion

213 coefficient), while for purely directed movement MSD(t) = V2t2 (V, velocity). Here, the

214 MSD was fit with , and α, the MSD exponent, was taken as our main

215 descriptor (Fig. 6c, α cumulative distribution function, cdf). α gives access to the

216 convective (α shifts towards 2) or not convective (α centered about 1) characteristics

217 of the movement.

218 Compared to a Brownian reference process (black), half of WT-shCtrl

219 trajectories (blue) were shifted to the left and correspond to non-convective

220 trajectories, while the other half were shifted to the right, indicative of convective

221 movement (Fig. 6d). The vast majority of WT-shDynlt3 trajectories (red) were

9 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

222 convective (Fig. 6d), with a statistically significant difference between shCtrl and

223 shDynlt3 α distributions (p<0.005, Wilcoxon test). Similarly, there was a statistically

224 significant difference between the α distributions of WT and bcat* cells and between

225 WT-GFP and WT- βcat-Δex3-GFP cells (Supplementary Fig. 5), with the

226 melanosomes in cells overexpressing β-catenin displaying more convective

227 movement. Accordingly, the populations were subdivided in two subpopulations,

228 small [S] and large [L] amplitudes, as assessed by the second moments of

229 trajectories (Fig. 6d). The frontier was chosen such that S mainly corresponded to a

230 homogeneous not convective (NC) population, and L to a convective (C) population

231 (see supplementary information), with trajectories between the two lines defining an

232 error bar. The number of non-convective trajectories (out of 75 trajectories) was

233 estimated to be 56-62 for WT, compared to 14-18 for bcat*; 60-64 for WT-GFP

234 compared to WT- βcat-Δex3-GFP (Supplementary Fig. 5); and 36-43 for WT-shCtrl

235 compared to 14-22 for WT-shDynlt3 (Fig. 6e). This illustrates that the downregulation

236 of Dynlt3 or the expression of active β-catenin both lead to a significant increase of

237 convective trajectories.

238

239 Overexpression of β-catenin or reduction of Dynlt3 diminishes melanosome

240 maturation and transfer

241 Melanosome maturation was evaluated in cells expressing different amount of

242 Dynlt3 by assessing melanosome acidity. Non-fully mature melanosomes are more

243 acidic than mature melanosomes. In this context, the number of acidic and

244 pigmented melanosomes was approximately two-fold higher in bcat* cells relative to

245 WT cells (12.4 vs. 6.4 acidic melanosomes/field), as well as in WT-siDynlt3 vs. WT-

246 siCtrl (13.0 vs. 7.3) (Fig. 7a).

10 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

247 Melanocytes with increased β-catenin or decreased Dynlt3 levels harboured

248 more pigmented and peripheral melanosomes, which would represent an ideal

249 subpopulation of pigment granules to be transferred to keratinocytes. However, their

250 change in pH might reflect a failure to acquire transfer capacity 5. We thus examined

251 whether these cells would have a deficiency in melanosome transfer and

252 consequently on coat color. Indeed, the hairs of bcat* C57BL/6J mice were lighter

253 than WT C57BL/6J mice (Fig. 7b). Melanosome transfer was evaluated by co-

254 culturing WT, bcat* WT-shCtrl, and WT-shDynlt3 melanocytes in the presence of

255 Balb/c MK keratinocytes for 10 days prior to immunofluorescence analysis using

256 plakoglobin (keratinocyte marker) and Hmb45 (a melanosome marker labeling the

257 intraluminal melanin-positive fibrils) (Fig. 7c,e). We then counted the number of

258 keratinocytes in which we could detect at least one Hmb45-positive structure; this

259 number was significantly lower when Dynlt3 was lower than normal (Fig. 7d,f). In

260 addition, we further confirmed that melanosome transfer was decreased in

261 melanocytes with decreased Dynlt3 by FACS (Supplementary Fig. 6).

262 Taken together, these data suggest that in melanocytes with stable, activated

263 β-catenin or in those with diminished expression of Dynlt3, both melanosome

264 maturation and transfer are negatively affected.

265

11 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

266 DISCUSSION

267 In this study, we showed that in melanocytes derived from transgenic mice

268 expressing active β-catenin (bcat* cells), pigmented melanosomes were primarily

269 absent from the perinuclear area and localized at the cell periphery. Notably, the

270 expression of Dynlt3, a member of the cytoplasmic dynein complex, was

271 downregulated in bcat* cells and knockdown of Dynlt3 in WT melanocytes

272 phenocopied the peripheral localization of pigmented melanosomes and their

273 absence from the perinuclear area. Melanosome movement was increased and

274 melanosomes moved in a more convective (directed) manner in cells with diminished

275 levels of Dynlt3. Decreased levels of Dynlt3 resulted in more acidic melanosomes

276 that was associated with a reduction of melanosome transfer to keratinocytes.

277 Altogether, these results show that Dynlt3 is a key player in regulating melanosome

278 movement, maturation and transfer.

279 Dynlt3 is a component of the cytoplasmic dynein complex, a macromolecular

280 protein complex that is essential for the transport of cargos from the cell periphery

281 towards the nucleus 14,15. While the heavy chain (Dync1h1) of this complex has an

282 ATPase activity and interacts with the microtubules, the light intermediate (Dync1li1

283 and Dync1li2) and the light chains, including Dynlt3, along with dynactin, a dynein-

284 interacting cofactor and activator, are thought to play a role in cargo recognition.

285 Despite previous studies having shown that various virus capsid proteins (e.g.

286 papilloma and herpes) interact with Dynlt3 and hijack the Cytoplasmic Dynein

287 complex to reach the nucleus of infected cells 19,20, this work is the first to

288 demonstrate functionally that Dynlt3 is involved in the intracellular trafficking of an

289 organelle.

12 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

290 To date, transgenic mice with melanocyte-specific mutations in cytoplasmic

291 dyneins have neither been identified nor produced, and therefore the role of Dynlt3 in

292 mouse pigmentation remains unknown. However, transgenic mice containing

293 heterozygous mutations in Dync1h1, which encodes the dynein heavy chain

294 Dync1h1, have been produced. In general, these mice show defects in motor

295 functions, with motor neuron degeneration and/or sensory neuropathy observed 21-24.

296 Yet, no obvious coat color phenotypes were reported in these heterozygous mice.

297 Importantly, homozygous Dync1h1 mice are not viable and die during embryonic

298 development or shortly after birth. Therefore, whether a coat color phenotype would

299 be observable in homozygous mice but would not apparent in heterozygous mice

300 remains unknown. Conditional mouse mutants would be required to answer this

301 point, using for instance the Tyr::Cre mice 25. While mutations in DYNLT3 are not

302 significantly detailed in human disease, reduction of DYNLT3 protein has been

303 observed in late stage Parkinson’s disease that did not develop melanoma 26.

304 Microarray analyses of LRO trafficking genes demonstrated that among the

305 components associated with cytoplasmic dynein, only Dynlt3 levels were altered in

306 response to β-catenin expression, suggesting that only Dynlt3 is a target of β-

307 catenin. Importantly, our results identify Dynlt3 as a novel β-catenin target, and one

308 that whose expression is negatively regulated by β-catenin. However, we cannot

309 exclude that knockdown of other dynein complex members may have a similar effect

310 on melanosome localization. Importantly, in WT and bcat* cells, Dynlt1, the other

311 member of the Tctex light chain family, was not expressed (Supplementary Fig. 2b),

312 in agreement with previous studies showing that Dynlt1 and Dynlt3 are not

313 expressed in the same tissue 27,28. Whether other cytoplasmic dynein subunits affect

13 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

314 melanosome localization, and which ones in particular do so, are currently under

315 investigation.

316 Video microscopy and manual tracking showed that melanosome movement

317 is not directional (non-convective). Instead, the direction that melanosomes moved

318 changed consistently, with melanosomes often going forwards and backwards

319 repeatedly. Downregulation of Dynlt3 resulted in a more directed movement with

320 changes in direction occurring less frequently. On microtubules, this back and forth

321 movement of melanosomes can be thought of as competitions between the kinesin

322 and dynein motors. While this random movement would be energetically costly and

323 seemingly inefficient, this back and forth movement may occur to allow

324 melanosomes to mature properly without passing onto the actin microfilaments and

325 being transferred before they are fully mature.

326 To our knowledge, this work is the first indication that defects in a cytoplasmic

327 dynein has an impact on melanosome transfer. That decreased Dynlt3 levels

328 diminished melanosome transfer is somewhat perplexing since pigmented

329 melanosomes localized primarily at the cell periphery, and were ideally positioned to

330 be further transferred. However, in cells that expressed less Dynlt3, melanosomes

331 were more acidic, as more pigmented melanosomes accumulated acridine orange.

332 Since melanosomes de-acidify their lumen before being transferred to keratinocytes,

333 and since increased acidity is an indication of immature melanosomes, our data

334 suggests that Dynlt3 plays some role in the regulation of melanosome maturity, or at

335 the very least, in the de-acidification of the melanosome lumen. Recently, it has been

336 shown that the maturation of melanosomes before transfer involves the formation

337 and release of membrane tubules containing proteins (e.g. Vamp7) that are needed

338 to tune the melanosome homeostasis (including the pH and melanin content) and

14 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

339 ultimately their secretory capacity 5. The separation of these tubules from the mature

340 melanosome requires proteins involved in membrane scission (e.g. Myosin 6,

341 Optineurin), whose downregulation results in increased melanosome size and

342 acidity. One should note that Myosin 6 was upregulated in bcat* vs. WT melanocytes

343 (Fig. 2a). Our hypothesis is that since the level of Dynlt3 is reduced, the cell tries to

344 compensate for this reduction by increasing the amount of Myosin 6 to allow the

345 separation of these tubules from the melanosomes. Therefore, it is conceivable that

346 Dynlt3 plays a role in this process, possibly in the pulling, fission and/ or the

347 retrograde transport of the melanosomal tubules, therefore contributing to

348 melanosome maturation and secretion.

349 Taken together, our results have identified Dynlt3 as a novel and important

350 player in melanosome biology (transport, maturation and transfer) and in coat

351 coloration and pigmentation. Furthermore, our results suggest that cytoplasmic

352 dynein proteins may have other roles in the biology of different organelles, in addition

353 to their already established functions in organelle trafficking.

354

15 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

355 MATERIALS AND METHODS

356

357 Cell culture

358 Mouse primary melanocyte cell lines were established from the skin of wild-type

359 (WT) and bcat* C57BL/6J transgenic mice as previously described 10. Melanocyte

360 cell lines were grown in Ham’s F12 media (Gibco, 21765-029) supplemented with

361 10% FBS (Gibco, 10270-106), antibiotics (100 U/mL penicillin and 100 μg/mL

362 streptomycin; Gibco, 15140), and 200nM TPA (Sigma, P1585). Balb/c MK mouse

363 epidermal keratinocytes were grown in Joklik Modified MEM (Sigma, M8028)

364 supplemented with 10% FBS, antibiotics (as above) and 5 ng/mL EGF (Sigma,

365 E9644). All cell lines were maintained at 37°C in a humidified atmosphere containing

366 5% CO2.

367 For co-culture experiments, the melanocyte and keratinocyte cell lines were

368 cultured in melanocyte growth media in the presence of 5 ng/mL EGF and 50 nM

369 CaCl2. Co-culture experiments were performed with a ratio of 2:1, keratinocytes to

370 melanocytes and the cells were left in co-culture for 10 days, after which they were

371 processed as described below. For immunofluorescence experiments, the cells were

372 plated on glass coverslips coated with 1 mg/mL collagen (Sigma, C7661).

373 Animal care, use, and experimental procedures were conducted in

374 accordance with recommendations of the European Community (86/609/EEC) and

375 Union (2010/63/UE) and the French National Committee (87/848). Animal care and

376 use were approved by the ethics committee of the Curie Institute in compliance with

377 the institutional guidelines.

378

379

16 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

380 Transfection and infection of melanocyte cell lines

381 All cells were transfected using the Amaxa Cell Line Nucleofector Kit L (Lonza, VCA-

382 1005) according to the manufacturer’s recommendations. Briefly, cells were

383 trypsinized and counted, with 7.5x105 cells resuspended in the Amaxa transfection

384 reagent. Four micrograms of each respective plasmid or 100 pmol of siRNA was

385 added to the mixture, which was then transferred to a cuvette and incubated in the

386 Amaxa nucleofector. Following nucleofection, the cells were added to a six-well dish

387 containing pre-warmed growth media. The following day, the cells were washed and

388 incubated in fresh growth media supplemented with TPA. Transfections were

389 assayed 48 hours following nucleofection. Control GFP and β-catenin-GFP tagged

390 plasmids have been previously described 13. The Dynlt3-GFP (pCMV3-mDynlt3-

391 GFPSpark) and the control GFP vectors (pCMV3-GFP) were purchased from Sino

392 Biological (MG5A2492-ACG and CV026, respectively).

393 9v cells expressing a control shRNA or an shRNA against Dynlt3 (targeting

394 sequence: 5’-ATC AGA TGT TGT ATC CCA A-3’) were produced using viral

395 particles and Dharmacon SMARTvector plasmids (VSC11715 and V3SM11241-

396 233912870, respectively). Following infection, cells were selected using 1 μg/mL

397 puromycin.

398

399 siRNA sequences

400 siRNA targeting mouse β-catenin and Dynlt3 were purchased from Dharmacon as a

401 SMARTpool mix of 4 sequences. The sequences of the siRNA are as follows: for β-

402 catenin: 5'-GAA CGC AGC AGC AGU UUG U-3', 5'-CAG CUG GCC UGG UUU GAU

403 A-3', 5'- GCA AGU AGC UGA UAU UGA C-3', 5'- GAU CUU AGC UUA UGG CAA U-

404 3'; and for Dynlt3: 5’-CCC AUA AUA UAG UCA AAG A-3’, 5’-GGU GGU AAC GAU

17 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

405 UAU AAU G-3’, 5’-GGG GAA AGC UUA CAA GUA C-3’, 5’-CAG AGG AGC CCG

406 UAU GGA U-3’. A random, scrambled sequence (5’-AAU UCU CCG AAC GUG UCA

407 CGU-3’) was used as a control.

408

409 Semi-quantitative Real-time PCR

410 Total RNA was extracted from cells using the miRNeasy kit from Qiagen (217004)

411 according to the manufacturer’s protocol. Reverse transcription reactions were

412 performed with 500 ng of total RNA and M-MLV reverse transcriptase (Invitrogen),

413 according to the manufacturer’s instructions. The cDNA generated was then

414 subjected to quantitative real-time PCR for the analysis of gene expression using ABI

415 7900HT. Oligonulceotides used for PCR were as follows: for β-catenin, 5’-CGT

416 GGA CAA TGG CTA CTC AA-3’ (F) and 5’-TGT CAG CTC AGG AAT TGC

417 AC-3’ (R); for Dynlt3 5’-GCG ATG AGG TTG GCT TCA ATG CTG-3’ (F) and 5’-

418 CAC TGC ACA GGT CAC AAT GTA CTT G-3’ (R); and for Gapdh 5’-ACC CAG AAG

419 ACT GTG GAT GG-3’ (F) and 5’-CAC ATT GGG GGT AGG AAC AC-3’ (R). F means

420 forward. R means reverse.

421

422 Western blot analysis

423 Whole cell lysates were prepared from melanocyte cell lines using ice-cold RIPA

424 buffer supplemented with complete protease inhibitor cocktail and PhosStop

425 phosphatase inhibitor cocktail (Roche). For western blotting, 25 μg total protein was

426 separated on 15% denaturing acrylamide SDS-PAGE gels and the proteins

427 transferred to nitrocellulose membranes. Membranes were blocked in 5% non-fat

428 milk in Tris-buffered saline supplemented with 0.05% Tween-20 (TBST) and probed

429 with primary antibodies overnight. The signal was detected using peroxidase-

18 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

430 conjugated anti-mouse or anti-rabbit secondary antibodies (refs) and enhanced

431 chemiluminescence (ECL; Thermo Fisher). The primary antibodies were used as

432 follows: Dynlt3, 1:200 (Sigma, hpa003938), β-catenin, 1:2000 (Abcam, ab6302),

433 GFP, 1:500 (Institut Curie), and β-actin, 1:10,000 (Sigma, A5441).

434

435 Immunofluorescence microscopy

436 Melanocyte cells were seeded on 18mm glass cover slips and grown to confluence,

437 after which they were washed twice with ice-cold PBS and fixed-permeabilized with

438 ice-cold methanol-acetone (1:1 solution) on ice for 5 minutes. Next, cells were

439 washed twice with cold PBS and blocked with a blocking solution containing 1% BSA

440 (w/v) and 10% FBS in PBS for 1 hour at room temperature. Cells were washed again

441 twice with PBS and then incubated with primary antibodies against anti-Tyrp1

442 (Abcam, ab3312) and anti-GFP (Institut Curie) at 4°C overnight. The following day,

443 the cells were washed again three times with PBS and then incubated with Alexa

444 555 anti-rabbit (ThermoFisher, A231572) or Alexa 488 anti-mouse (ThermoFisher,

445 A21202) secondary antibodies for 1 hour at room temperature in the absence of

446 light. Cells were then again washed two times before being counterstained with 0.5

447 µg/µL DAPI to visualize the nucleus. Coverslips were then mounted on glass slides

448 using ProLong Gold antifade reagent (Invitrogen, P36934). Images were then taken

449 using an Upright widefield Leica Microscope at 63X.

450 For co-culture experiments, mouse-keratinocyte co-cultures were washed

451 twice with cold PBS and then fixed with 4% paraformaldehyde (VWR 1.04005.1000)

452 for 20 min at room temperature. Following two washes with PBS, the cells were

453 permeabilized with 0.2% v/v Triton X-100 in PBS for 10 min at room temperature.

454 The cells were once again washed and then processed as described above, using

19 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

455 anti-Hmb45 (Abcam, 787) and anti-plakoglobin (Abcam, 184919) primary antibodies.

456 Co-cultures were examined using a CLSM SP5 Leica confocal microscope.

457

458 Video microscopy and melanosome tracking

459 Cells were seeded in 6-well glass plates (MatTek Corporation P06G-1.5-20-F) and

460 were imaged using the Ratio 2 inverted video microscope. Cells were maintained at

461 37°C in a humidified atmosphere of 5% CO2 while videos were being taken. Videos

462 were taken at 100X objective over a period of 5 minutes, with one image taken every

463 0.5 sec, for a total number of 601 images. Videos were taken from a minimum of

464 nine melanocytes per cell line/transfection from three independent experiments.

465 From these melanocytes, the trajectories of 75 pigmented melanosomes were

466 manually followed using the Manual Tracking plugin for ImageJ 29. Each

467 melanosome path was chosen at random. The Manual Tracking allowed provided a

468 number of different parameters that were used to characterize the movement of the

469 individual melanosomes. Namely, we calculated the total distance the melanosomes

470 moved, which was the sum of values for all of the 601 frames. The Euclidean

471 distance was calculated as the distance between the melanosome position at the

472 first frame and the last frame. The average distance was calculated as the total

473 distance divided by the number of frames (601). Similarly, the average velocity was

474 calculated by taking the average distance per frame and dividing by the time (0.5

475 sec). Finally, the percentage of time the melanosomes were paused was calculated

476 by assessing the number of frames (out of the 601 total frames) where a “0” value

477 was recorded for the distance travelled. Distribution of spatiotemporal descriptors of

478 trajectories, in particular Mean Square Displacement characteristics, was computed

20 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

479 and compared to simulated trajectories in order to distinguish convective vs. non-

480 convective movements (18 and supplementary information).

481

482 Melanosome number and localization quantification

483 Melanosome quantification was performed using ImageJ Analyze Particle feature. To

484 exclude melanosomes nearby the nuclear region, a circle of 15μm was designated

485 around the nucleus, as an arbitrary region of separation.

486 Melanosome localization in the different cell lines was quantified using

487 ImageJ. A line was drawn along the length of each individual cell, with the width of

488 the line extending to cover the entire cell. The pixel intensity along the length of the

489 line was used to give a quantitative measure of melanosome distribution. Pixel

490 intensity was measured from one end of the cell until reaching the nucleus, where

491 the measurement stopped and was then continued on the other side of the nucleus.

492 The data is presented as a measured pixel intensity over the percentage of the cell

493 covered (as such, all measurements end at 100, regardless of the cell length). A

494 trendline is presented (in red) to emphasize the trend.

495

496 Flow Cytometry

497 For all flow cytometry experiments, cells were trypsinized and the resulting

498 cell pellets were resuspended and fixed in 4% paraformaldehyde for 20 minutes at

499 room temperature. Following centrifugation, the cell pellets were then washed twice

500 with PBS containing 1% BSA and 0.1% saponin (Sigma, S4521). Cells were then

501 incubated in primary antibodies against Hmb45 and plakoglobin (both at 1:100) for 2

502 hours at 4°C in the absence of light. Next, cell pellets were washed once with

503 PBS/BSA/saponin and incubated in Alexa 647 mouse (ThermoFisher A31573) and

21 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

504 Alexa 488 mouse secondary antibodies (at 1:500) for 1 hour. Cells were then

505 washed in PBS and resuspended in PBS + 0.1% azide. Flow cytometric analyses

506 were performed using a BD FACSCanto II machine.

507

508

509 Melanosome acidity

510 Cells were seeded in 60 mm dishes and allowed to reach approximately 75%

511 confluence. Acridine orange (Sigma, A6014) was added to the cells at 2.5 μg/mL and

512 incubated for 20 minutes at 37°C. The cells were then incubated in warm PBS and

513 imaged using a confocal microscope. To look at acridine orange staining, excitation

514 wavelengths between 450-490 nm and emission between 590-650 nm were used 30.

515 Melanosome acidity was assessed as by counting the number of pigmented

516 melanosomes that were positive for acridine orange staining. This was done by

517 merging the brightfield and acridine orange signals. The number of melanocytes

518 assessed per field of view was always kept constant.

519

520 Electron microscopy

521 Melanocytes seeded on coverslips were chemically fixed in 2.5 % (v/v)

522 glutaraldehyde, 2 % (v/v) paraformaldehyde in 0.1M cacodylate buffer for 24h at 4°C,

523 post-fixed with 1% (w/v) osmium tetroxide supplemented with 1.5% (w/v) potassium

524 ferrocyanide, dehydrated in ethanol and embedded in Epon as previously described

525 31. Ultrathin sections of cell monolayers or tissue were prepared with a Reichert

526 UltracutS ultramicrotome (Leica Microsystems) and contrasted with uranyl acetate

527 and lead citrate. Electron micrographs were acquired on a Tecnai Spirit electron

22 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

528 microscope (FEI, Eindhoven, The Netherlands) equipped with a 4k CCD camera

529 (EMSIS GmbH, Münster, Germany).

530

23 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

531

532 ACKNOWLEDGEMENTS

533 ZA was supported by the Fondation ARC and Labex ANR Labex CelTisPhyBio

534 (ANR-11-LABX-0038 and ANR-10-IDEX-0001-02 PSL). ACP was supported by a

535 fellowship from Institut Curie. We are grateful to the Rimbaud and Delmas families

536 for their donations to our laboratory. This work was supported by Ligue Contre le

537 Cancer, Fondation ARC, Institut Carnot, INCa, and ITMO Cancer, CNRS, INSERM

538 and the French National Research Agency through the "Investments for the Future"

539 program (France-BioImaging, ANR-10-INSB-04), and is under the program

540 «Investissements d’Avenir» launched by the French Government and implemented

541 by ANR Labex CelTisPhyBio (ANR-11-LABX-0038 and ANR-10-IDEX-0001-02 PSL).

542 We acknowledge the PICT-IBiSA, member of the France-BioImaging national

543 research infrastructure.

544

545

24 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

546 REFERENCES

547

548 1 Colombo, S., Berlin, I., Delmas, V. & Larue, L. in Melanins and melanosomes 549 (eds P.A. Riley & J. Borovansky) 21-51 (Wiley-VCH Verlag & Co, 2011). 550 2 Delevoye, C., Marks, M. S. & Raposo, G. Lysosome-related organelles as 551 functional adaptations of the endolysosomal system. Curr Opin Cell Biol 59, 552 147-158, doi:10.1016/j.ceb.2019.05.003 (2019). 553 3 D'Alba, L. & Shawkey, M. D. Melanosomes: Biogenesis, Properties, and 554 Evolution of an Ancient Organelle. Physiol Rev 99, 1-19, 555 doi:10.1152/physrev.00059.2017 (2019). 556 4 Tarafder, A. K. et al. Rab11b mediates melanin transfer between donor 557 melanocytes and acceptor keratinocytes via coupled exo/endocytosis. The 558 Journal of investigative dermatology 134, 1056-1066, 559 doi:10.1038/jid.2013.432 (2014). 560 5 Ripoll, L. et al. Myosin VI and branched actin filaments mediate membrane 561 constriction and fission of melanosomal tubule carriers. J Cell Biol 217, 2709- 562 2726, doi:10.1083/jcb.201709055 (2018). 563 6 Dennis, M. K. et al. BLOC-1 and BLOC-3 regulate VAMP7 cycling to and from 564 melanosomes via distinct tubular transport carriers. J Cell Biol 214, 293-308, 565 doi:10.1083/jcb.201605090 (2016). 566 7 Hume, A. N. & Seabra, M. C. Melanosomes on the move: a model to 567 understand organelle dynamics. Biochem Soc Trans 39, 1191-1196, 568 doi:10.1042/BST0391191 (2011). 569 8 Hirokawa, N., Noda, Y., Tanaka, Y. & Niwa, S. Kinesin superfamily motor 570 proteins and intracellular transport. Nat Rev Mol Cell Biol 10, 682-696, 571 doi:10.1038/nrm2774 (2009). 572 9 Lamoreux, M. L., Delmas, V., Larue, L. & Bennett, D. The colors of mice : A 573 model genetic network. 297 (Wiley-Blackwell, 2010). 574 10 Delmas, V. et al. Beta-catenin induces immortalization of melanocytes by 575 suppressing p16INK4a expression and cooperates with N-Ras in melanoma 576 development. Genes & Development 21, 2923-2935, doi:10.1101/gad.450107 577 (2007). 578 11 Luciani, F. et al. Biological and mathematical modeling of melanocyte 579 development. Development 138, 3943-3954, doi:10.1242/dev.067447 (2011). 580 12 Raposo, G., Tenza, D., Murphy, D. M., Berson, J. F. & Marks, M. S. Distinct 581 protein sorting and localization to premelanosomes, melanosomes, and 582 lysosomes in pigmented melanocytic cells. J Cell Biol 152, 809-824, 583 doi:10.1083/jcb.152.4.809 (2001). 584 13 Yajima, I. et al. A subpopulation of smooth muscle cells, derived from 585 melanocyte-competent precursors, prevents patent ductus arteriosus. PLoS 586 ONE 8, e53183, doi:10.1371/journal.pone.0053183 (2013). 587 14 Reck-Peterson, S. L., Redwine, W. B., Vale, R. D. & Carter, A. P. The 588 cytoplasmic dynein transport machinery and its many cargoes. Nat Rev Mol 589 Cell Biol 19, 382-398, doi:10.1038/s41580-018-0004-3 (2018). 590 15 Liu, J. J. Regulation of dynein-dynactin-driven vesicular transport. Traffic 18, 591 336-347, doi:10.1111/tra.12475 (2017).

25 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

592 16 Palmisano, I. et al. The ocular albinism type 1 protein, an intracellular G 593 protein-coupled receptor, regulates melanosome transport in pigment cells. 594 Hum Mol Genet 17, 3487-3501, doi:10.1093/hmg/ddn241 (2008). 595 17 Lopes, V. S. et al. The ternary Rab27a-Myrip-Myosin VIIa complex regulates 596 melanosome motility in the retinal pigment epithelium. Traffic 8, 486-499, 597 doi:10.1111/j.1600-0854.2007.00548.x (2007). 598 18 Coscoy, S., Huguet, E. & Amblard, F. Statistical analysis of sets of random 599 walks: how to resolve their generating mechanism. Bulletin of mathematical 600 biology 69, 2467-2492, doi:10.1007/s11538-007-9227-8 (2007). 601 19 Schneider, M. A., Spoden, G. A., Florin, L. & Lambert, C. Identification of the 602 dynein light chains required for human papillomavirus infection. Cell Microbiol 603 13, 32-46, doi:10.1111/j.1462-5822.2010.01515.x (2011). 604 20 Apcarian, A., Cunningham, A. L. & Diefenbach, R. J. Identification of binding 605 domains in the herpes simplex virus type 1 small capsid protein pUL35 606 (VP26). J Gen Virol 91, 2659-2663, doi:10.1099/vir.0.019984-0 (2010). 607 21 Chen, X. J. et al. Proprioceptive sensory neuropathy in mice with a mutation in 608 the cytoplasmic Dynein heavy chain 1 gene. J Neurosci 27, 14515-14524, 609 doi:10.1523/JNEUROSCI.4338-07.2007 (2007). 610 22 Hafezparast, M. et al. Mutations in dynein link motor neuron degeneration to 611 defects in retrograde transport. Science 300, 808-812, 612 doi:10.1126/science.1083129 (2003). 613 23 Courchesne, S. L., Pazyra-Murphy, M. F., Lee, D. J. & Segal, R. A. 614 Neuromuscular junction defects in mice with mutation of dynein heavy chain 615 1. PLoS One 6, e16753, doi:10.1371/journal.pone.0016753 (2011). 616 24 Sabblah, T. T. et al. A novel mouse model carrying a human cytoplasmic 617 dynein mutation shows motor behavior deficits consistent with Charcot-Marie- 618 Tooth type 2O disease. Sci Rep 8 , 1739, doi:10.1038/s41598-018-20081-1 619 (2018). 620 25 Delmas, V., Martinozzi, S., Bourgeois, Y., Holzenberger, M. & Larue, L. Cre- 621 mediated recombination in the skin melanocyte lineage. Genesis 36, 73-80 622 (2003). 623 26 Chu, Y. et al. Alterations in axonal transport motor proteins in sporadic and 624 experimental Parkinson's disease. Brain 135, 2058-2073, 625 doi:10.1093/brain/aws133 (2012). 626 27 King, S. M. et al. Cytoplasmic dynein contains a family of differentially 627 expressed light chains. Biochemistry 37, 15033-15041, 628 doi:10.1021/bi9810813 (1998). 629 28 Chuang, J. Z., Milner, T. A. & Sung, C. H. Subunit heterogeneity of 630 cytoplasmic dynein: Differential expression of 14 kDa dynein light chains in rat 631 hippocampus. J Neurosci 21, 5501-5512 (2001). 632 29 Cordelieres, F. P. et al. Automated cell tracking and analysis in phase- 633 contrast videos (iTrack4U): development of Java software based on combined 634 mean-shift processes. PLoS ONE 8, e81266, 635 doi:10.1371/journal.pone.0081266 (2013). 636 30 Pierzynska-Mach, A., Janowski, P. A. & Dobrucki, J. W. Evaluation of acridine 637 orange, LysoTracker Red, and quinacrine as fluorescent probes for long-term 638 tracking of acidic vesicles. Cytometry A 85, 729-737, 639 doi:10.1002/cyto.a.22495 (2014).

26 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

640 31 Hurbain, I., Romao, M., Bergam, P., Heiligenstein, X. & Raposo, G. Analyzing 641 Lysosome-Related Organelles by Electron Microscopy. Methods Mol Biol 642 1594, 43-71, doi:10.1007/978-1-4939-6934-0_4 (2017). 643

644

645

27 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

646 Figure legends

647

648 Figure 1. Melanosomes are distributed at the periphery of melanocytes

649 expressing activated β-catenin.

650 a) Brightfield (top) and inverse brightfield (bottom) images of WT (9v) and bcat* (10d)

651 melanocytes. Bar, 20 μm.

652 b) Quantitation of pixel intensity (in black) across the length of the WT and bcat*

653 melanocytes shown in Panel A. Trendlines are shown in red.

654 c) Quantification of the total number of melanosomes in WT and bcat* melanocytes.

655 Results represent the mean ± SD of pooled data from three independent

656 experiments with a total of nineteen WT and fourteen bcat* cells. ns signifies no

657 statistical significance (p > 0.05) as determined by the two-sided Mann-Whitney

658 test.

659 d) Quantification of the number of perinuclear melanosomes (in a 15 µm diameter

660 circle centered around the nucleus) in WT and bcat* melanocytes. Results

661 represent the mean ± SD of pooled data from three independent experiments

662 with a total of nineteen WT and fourteen bcat* cells. Statistical significance was

663 determined by the two-sided Mann-Whitney test and *** signifies p < 0.001.

664 e) Immunofluorescence analysis of WT and bcat* melanocytes. Cells were

665 processed for immunofluorescence and stained with anti-Tyrp1 (red) antibodies

666 and the nuclei were stained with DAPI (blue). Note that Tyrp1 staining was also

667 observed in compartments adjacent to the nucleus, which is most likely the

668 protein that is being synthesized and processed in the Golgi/ER. Bar, 20 μm.

669 670 Figure 2. Dynlt3 is downregulated after β-catenin expression.

28 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

671 a) Expression profile of 67 genes associated with trafficking of melanosomes and

672 other lysosome related organelles (LRO). Data was obtained from Affymetrix

673 mRNA transcriptomic analyses and is presented as the ratio of expression of

674 bcat* (10d) cells to WT (9v) cells. The raw data set is available upon request.

675 b) RT-qPCR analysis of Dynlt3 mRNA levels in WT and bcat* melanocytes.

676 c) Western blot analysis of Dynlt3 protein levels in WT and bcat* melanocytes.

677 Quantification of the data is presented below from four independent experiments.

678 d,e) RT-qPCR analysis of β-catenin (D) and Dynlt3 (E) expression in WT

679 melanocytes following transfection with the βcat-∆ex3-GFP (βcat) expression

680 vector or a GFP control.

681 f) Western blot of βcat and Dynlt3 levels in WT melanocytes following transfection

682 with the βcat-∆ex3-GFP (βcat) expression vector or a GFP control.

683 g) Quantification of the Dynlt3 western blot data presented in Panel F.

684 h,i) RT-qPCR analysis of β-catenin (H) and Dynlt3 (I) expression in WT melanocytes

685 following transfection with either a control siRNA or siRNA targeting β-catenin.

686 j) Western blot of βcat and Dynlt3 levels in WT melanocytes following transfection

687 with either a control siRNA or siRNA targeting β-catenin.

688 k) Quantification of the Dynlt3 western blot data presented in Panel J.

689 l,m) RT-qPCR analysis of β-catenin (L) and Dynlt3 (M) expression in bcat*

690 melanocytes following transfection with either a control siRNA or siRNA targeting

691 β-catenin.

692 n) Western blot of βcat and Dynlt3 levels in bcat* melanocytes following transfection

693 with either a control siRNA or siRNA targeting β-catenin.

694 o) Quantification of the Dynlt3 western blot data presented in Panel N.

29 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

695 For western blot analysis, all statistical significance was determined using unpaired

696 two-sided T-tests. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

697

698 Figure 3. Melanosomes are distributed primarily at the periphery of

699 melanocytes when Dynlt3 levels are decreased.

700 a) Brightfield (top) and inverse brightfield (bottom) images of WT (9v) cells

701 transfected with either siControl (left) or siDynlt3 (right). Bar, 20 μm. Ctrl means

702 control.

703 b) Western blot analysis of Dynlt3 levels in WT cells transfected with siControl (Ctrl)

704 or siDynlt3. Relative quantification of the western blot analysis was performed

705 from three independent experiments. Statistical significance was determined

706 using unpaired two-sided T-tests. *** p<0.002.

707 c) Quantitation of pixel intensity (in black) across the length of the WT cells

708 transfected with siControl or siDynlt3 shown in Panel A. Trendlines are shown in

709 red.

710 d) Quantification of the number of perinuclear melanosomes (in a 15 µm diameter

711 circle centered around the nucleus) in WT melanocytes transfected with siCtrl or

712 siDynlt3. Results represent the mean ± SD of pooled data from three independent

713 experiments with a total of nineteen WT-siCtrl and twenty-one WT-siDynlt3 cells.

714 Statistical significance was determined by the two-sided Mann-Whitney test, ***: p

715 < 0.001.

716 e) Quantification of the total number of melanosomes in WT melanocytes transfected

717 with siCtrl or siDynlt3. Results represent the mean ± SD of pooled data from three

718 independent experiments with a total of nineteen WT-siCtrl and twenty-one WT-

30 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

719 siDynlt3 cells. Statistical significance was determined by the two-sided Mann-

720 Whitney test, ns: not significant.

721 f) Immunofluorescence analysis of WT melanocytes transfected with siCtrl or or

722 siDynlt3. Cells were processed for immunofluorescence 48 hours after

723 transfection and stained with anti-Tyrp1 (red) antibodies. Nuclei were stained with

724 DAPI (blue). Note that Tyrp1 staining was also observed in compartments

725 adjacent to the nucleus, which is most likely the protein that is being synthesized

726 and processed in the Golgi/ER. Scale bar, 20 μm.

727

728 Figure 4. Overexpression of Dynlt3 redistributes melanosomes in bcat*

729 melanocytes.

730 a) Brightfield (top), inverse brightfield (middle) and GFP (bottom) images of bcat*

731 (10d) melanocytes transiently expressing GFP or GFP-tagged Dynlt3. Bar,

732 20 μm.

733 b) Quantitation of pixel intensity (in black) across the length of the bcat*-GFP and

734 bcat*-Dynlt3GFP melanocytes shown in Panel A. Trendlines are shown in red.

735 c) Immunofluorescence analysis of bcat* melanocytes transiently expressing GFP or

736 GFP-tagged Dynlt3. Cells were processed for immunofluorescence and stained

737 with anti-Tyrp1 (red) antibodies and the nuclei were stained with DAPI (blue). Bar,

738 20 μm.

739

740 Figure 5. Melanosome movement is increased in cells producing active β-

741 catenin or with diminished levels of Dynlt3.

742 Melanosome movement (corresponding mainly to stage IV melanosome) was

743 assessed by brightfield video microscopy over a period of 5 minutes and

31 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

744 melanosome trajectories were followed using ImageJ software. For each cell line,

745 75 melanosomes were followed from a minimum of 5 independent cells. WT (9v)

746 cell lines were transfected with control GFP or β-catenin-GFP expression vectors,

747 respectively. Analyses were performed on both transfected (i.e. GFP-positive)

748 and non-transfected (i.e. GFP-negative) cells. In addition, WT cells were

749 transfected with either control or Dynlt3 shRNA vectors and analyses were

750 performed on the resulting red cells. Each dot represents one melanosome. (a)

751 The total distance indicates the sum of all of the individual tracks for each

752 melanosome, (c) The Euclidean distance refers to the distance between the start

753 and end position of each melanosome. (d) The percentage of time that each

754 melanosome spent stationary/paused refers to the proportion of the number of

755 tracks where no movement was measured compared to the total number of

756 tracks. (b) The level of Dynlt3 in these different cell lines was evaluated after

757 western blot analysis. Relative quantification of the western blot analysis was

758 performed from three to four independent experiments. In panel c, one outlier

759 was removed from WT+βcat+, its value was 20 μm. For western blot analysis, all

760 statistical significance was determined using unpaired two-sided T-tests.

761 **:p <0.01; *** p<0.001. For the other analyses, statistical significance was

762 measured using two-sided Mann-Whitney test. **** p < 0.0001.

763

764 Figure 6. Increased convective movement of melanosomes in cells with

765 increased levels of β-catenin or decreased Dynlt3.

766 a,b) Schematic representation of not convective (a) and convective (b) movement.

767 The purple ellipse represents the cargo. The arrows represent the direction and

32 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

768 the orientation of the movement, and the length of the arrows are directly

769 proportional to the magnitude of the movement.

770 c) α MSD distributions for WT-shCtrl (blue) and WT-shDynlt3 (red) populations of

771 trajectories, compared to the distribution for simulated Brownian (purely diffusive)

772 random walks (nwalks=10 000, with the same nsteps (601) as for experimental

773 trajectories). Theoretical shifts compared to Brownian process when adding

774 convection, noise, tethering or confinement are indicated with arrows. **,

775 statistically significant difference between WT-shDynt3 and WT-shCtrl, p<0.005.

776 Statistical analyses were done using a Wilcoxon test.

777 d) μ2S, μ2L plots reflecting the extent of WT-shCtrl (b) and WT-shDynlt3 (c)

778 trajectories. Each point corresponds to one trajectory. Lines correspond to

779 frontiers between non-convective (NC) and convective (C) trajectories: see SI for

780 criteria and validation.

781 e) Number of non-convective (NC) trajectories for melanosomes in WT, bcat*, WT-

782 GFP, WT- βcat-∆ex3-GFP, WT-shCtrl and WT-shDynlt3 cells, out of 75

783 trajectories for each population of trajectories. The same frontiers were used for all

784 populations (see SI). Error bars correspond to trajectories between the two

785 frontiers (see Fig. 5d, and supplementary Fig. 5S).

786

787 Figure 7. Melanosome maturation and transfer are altered in cells producing

788 active β-catenin or with diminished level of Dynlt3.

789 a) Quantitation of acridine orange positive melanosomes in WT, bcat*, WT + siCtrl

790 and WT + si Dynlt3 cells.

791 b) Photograph of wild-type and bcat* adult mice (left) and of their respective dorsal

792 hairs (right).

33 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

793 c) Immunofluorescence analysis of melanosome transfer between melanocytes WT

794 (9v) or bcat* (10d) and Balb/c MK keratinocytes. Melanocytes and keratinocytes

795 were co-cultured for 10 days, fixed and processed for immunofluorescence

796 analyses using Plakoglobin and HMB45 antibodies. The yellow arrows indicate

797 Pmel positive staining localized outside of melanocytes.

798 d) Quantitation of the percentage of Pmel and Plakoglobin double positive cells is

799 presented (right). Pmel positivity is assessed as being either positive or negative

800 within each respective keratinocyte. Quantitation was done from a minimum of

801 three independent co-cultures.

802 e) Immunofluorescence analysis of melanosome transfer between melanocytes (WT

803 + sh Ctrl or WT + shDynlt3) and Balb/c MK keratinocytes. Melanocytes and

804 keratinocytes were co-cultured for 10 days, fixed and processed for

805 immunofluorescence analyses using Plakoglobin and HMB45 antibodies. The

806 yellow arrows indicate Pmel positive staining localized outside of melanocytes.

807 f) Quantitation of the percentage of Pmel and Plakoglobin double positive cells is

808 presented (right). Pmel positivity was assessed as being either positive or

809 negative within each respective keratinocyte. Quantitation was done from a

810 minimum of three independent co-cultures.

811

34 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

812 Dynlt3, a novel fundamental regulator of melanosome movement,

813 distribution, maturation and transfer

814

815 Zackie Aktary1-3, Alejandro Conde-Perez1-3, Florian Rambow1-3, Mathilde Di Marco6,7, 816 François Amblard3-5, Ilse Hurbain6,7, Graça Raposo6,7, Cédric Delevoye6,7, Sylvie 817 Coscoy3,4, & Lionel Larue1-3* 818

819

820

821 Supplementary information

822

823

824

825 This part is associated with Figure 5. Eight descriptors were defined: α, ρ, μ2L, μ2S,

826 μ3L, μ3S, μ4L and μ4S. The comparison between a set of experiments and a set of

827 simulated trajectories was performed by fitting the distributions of eight descriptors

828 (MSD exponent, anisotropy, the second [q=2], third [q=3, skewness] and fourth [q=4,

829 kurtosis] normalized moments projected on the two main L and S axis (Fig. 6 and

830 Supplementary Fig. 5). The seven other descriptors refer only to 2D spatial

831 characteristics and were obtained by projecting trajectory points on its own axis

832 system, giving two unidimensional distributions which were statistically described by

833 the computation of moments (E) of order q (2-4) (l corresponds to the center scalar

834 distribution of the projection of the long axis, n to sample size, the second moment

835 (when q=2) E2 [long axis] is named μ2L-abs), and by anisotropy defined as the ratio of

836 the second moments μ2S-abs / μ2L-abs (s means short axis).

837 Small amplitude WT melanosome subpopulation was non convective (fitting

838 with anisotropic Brownian movement + elastic tethering + noise), estimated to 39%

35 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

839 of WT melanosome population. In contrast, only half of the population of the bcat*

840 "small" was non-convective, and the other half was convective, estimated to 7% of

841 bcat* population.

842 Large amplitude WT and bcat* subpopulations correspond to convective

843 behavior of melanosomes. Convective behavior may correspond to two types of

844 movement: intermittent or uniform. We consider the movement uniform when the

845 velocity of each melanosome is constant during the whole observation period, and

846 intermittent if the conditions are not fulfilled. We cannot discriminate between the

847 uniform and intermittent behaviors of the wt or bcat* melanosomes. This is due to the

848 fact the velocity is too low compared to the diffusion.

849

36 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

850 Supplementary Figure Legends

851

852 Supplementary Figure 1. Melanosomes are distributed at the periphery of

853 melanocytes expressing exogenous activated β-catenin.

854 a) Brightfield (top), inverse brightfield (middle) and GFP (bottom) images of WT (9v)

855 melanocytes transiently expressing GFP or GFP-tagged β-catenin. Bar, 20 μm.

856 b) Quantitation of pixel intensity (in black) across the length of the WT-GFP and WT-

857 β-catGFP melanocytes shown in Panel A. Trendlines are shown in red.

858 c) Immunofluorescence analysis of WT melanocytes transiently expressing GFP or

859 GFP-tagged b-catenin. Cells were processed for immunofluorescence and

860 stained with anti-Tyrp1 (red) antibodies and the nuclei were stained with DAPI

861 (blue). Bar, 20 μm.

862

863 Supplementary Figure 2. bcat* melanosomes are larger than WT melanosomes.

864 a,b) Transmission electron microscopy of WT (a) and bcat* (b) melanocytes. Scale

865 bar: 2 μm.

866 c) Length of melanosome was determined from four WT and five bcat* melanocytes

867 with a total of at least 200 melanosomes per cell line. Statistical significance was

868 measured using two-sided Mann-Whitney test. **** p < 0.0001.

869

870 Supplementary Figure 3. Dynlt3, and not Dynlt1, is expressed in WT and

871 downregulated in bcat* melanocytes.

872 Ct values for RT-qPCR analysis of Dynlt3 (A) and Dynlt1 (B) mRNA expression in

873 two cell lines derived from WT mice (9v and 13d) and two from bcat* mice (10d and

874 14d).

37 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

875

876 Supplementary Figure 4. Melanosome movement is increased in cells

877 producing active β-catenin or with diminished level of Dynlt3.

878 Melanosome movement (corresponding mainly to stage IV melanosome) was

879 assessed by brightfield video microscopy over a period of 5 minutes and

880 melanosome trajectories were followed using ImageJ software. For each cell line, 75

881 melanosomes were followed from a minimum of five independent cells. WT (9v) cell

882 lines were transfected with control GFP or β-catenin-GFP expression vectors,

883 respectively. Analyses were performed on both transfected (i.e. GFP-positive) and

884 non-transfected (i.e. GFP-negative) cells. In addition, WT cells were transfected with

885 either control or Dynlt3 shRNA vectors and analyses were performed on the resulting

886 red cells. Each dot represents one melanosome. The average distance (a) and the

887 average velocity (b) for each melanosome are shown, Statistical significance was

888 measured using two-sided Mann-Whitney test. **** p < 0.0001.

889

890 Supplementary Figure 5. Mathematical modeling of melanosome trajectories.

891 a,d) α MSD distributions for 9v and 10d (a), and 9v-Ctrl and 9v-bcat (d) populations

892 of trajectories, compared to the distribution for simulated Brownian (purely

893 diffusive – black line) random walks (nwalks=10 000, with the same nsteps (601)

894 as for experimental trajectories). Theoretical shifts compared to Brownian process

895 when adding convection, noise, tethering or confinement are indicated with

896 arrows.

897 b,c,e,f) μ2S, μ2L plots reflecting the extent of 9v (b), 10d (c), 9v Ctrl (e), and 9v bcat (f)

898 trajectories. Each point corresponds to one trajectory. Lines correspond to

38 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

899 frontiers between non-convective (NC) and convective (C) trajectories: see SI for

900 criteria and validation.

901 g,h,i) α MSD, anisotropy, L and S second moments, L and S skewness, and L and S

902 kurtosis distributions for 9v and 10d (g), 9v-Ctrl and 9v-bcat (h), and 9v Sh Ctrl

903 and 9v Sh Dynlt3 (i) populations of trajectories, compared to the distribution for

904 simulated Brownian (purely diffusive – black line).

905 j,k,l) Criteria to establish the frontiers for non-convective (NC) and convective (C)

906 movement for α MSD, anisotropy, and L and S second moments distributions. On

907 the top and the bottom the lower and upper subdivisions are given, respectively.

908

909 Supplementary Figure 6. FACS analysis of melanosome transfer.

910 a) FACS analysis of Hmb45 and plakoglobin positive cell populations in WT and

911 bcat* melanocytes.

912 b) Quantitation of melanosome transfer assayed by FACS. WT (9v), bcat* (10d), WT

913 + sh Ctrl or WT + sh Dynlt3 melanocytes were co-cultured with Balb/c MK mouse

914 keratinocytes for 10 days, after which the co-cultures were fixed and processed

915 for FACS analyses using Hmb45 and Plakoglobin antibodies. The percentage of

916 Plakoglobin and Hmb45 double positive cells within each respective co-culture is

917 presented. Quantitation was done from a minimum of five independent co-culture

918 experiments.

919 c) Gating strategy for FACS experiments. Melanocyte and keratinocyte cells were

920 stained with either control IgG, plakoglobin or Pmel antibodies. The live cell

921 population was chosen for gating. IgG control antibodies were used to establish a

922 baseline of background signal. Following this, the established gate was applied to

923 samples stained with the respective antibodies of interest. The gates were kept

39 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

924 constant for all cell lines (melanocyte and keratinocyte) and for the co-culture

925 experiments.

926

40 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

927 928 929 930 931 Supplementary Table 1. Mouse models associated with mutations in 67 932 genes involved in trafficking of melanosomes 933 and other lysosome related organelles (LRO). 934 935 936 937 938 Gene Gene Product Human Mouse Model Stage Disease TYR TYR OCA1 Albino Maturation OCA2 OCA2/P OCA2 Pink eye dilution Maturation TYRP1 TYRP-1 OCA3 Brown Maturation SLC45A2 SLC45A2/MATP OCA4 Underwhite Maturation DCT DCT/TRP-2 Slaty Maturation PMEL PMEL/Pmel17/Silver/ Silver Maturation gp100 GPR143 GPR143/OA1 OA1 Oa1 Biogenesis MLANA MART1/melan-a Melan-a Maturation AP3B1 AP-3β3a subunit HPS2 Pearl Transport AP3D1 AP3 δ subunit Mocha Transport AP3M1 AP-3 µ3a subunit no2 Transport AP3S1 AP-3 σ3a subunit no2 Transport BLOC1S5 Muted Muted Transport BLOC1S4 Cappuccino HPS1 Cappuccino Transport DTNBP1 Dysbindin HPS7 Sandy Maturation BLOC1S3 BLOS3 HPS8 Reduced Biogenesis pigmentation PLDN Pallidin HPS9 Pallid Maturation BLOC1S1 BLOS1 No reported coat Transport color BLOC1S2 BLOS2 no2 Transport SNAPIN Snapin Perinatal lethality1 Transport HPS1 HPS1 HPS1 Pale ear Biogenesis HPS3 HPS3 HPS3 Cocoa Biogenesis HPS4 HPS4 HPS4 Light ear Biogenesis HPS5 HPS5 HPS5 Ruby-eye 2 Biogenesis HPS6 HPS6 HPS6 Ruby-eye Biogenesis VPS11 VPS11 no2 Transport VPS16 VPS16 no2 Transport VPS18 VPS18 no reported coat Transport color3 VPS33A VPS33A Buff Maturation VPS39 VPS39/Vam6 no2 Transport VPS41 VPS41 Embryonic lethal3 Transport VPS33B VPS33B ARCS1 no2 Maturation C14ORF13 VIPAR/VPS16B ARCS2 no2 Transport 3 RABGGTA Subunit of Rab Gunmetal Biogenesis geranylgeranyl transferase II LYST LYST CHS Beige Transport MYO5A Myosin VA GS1 Dilute Transport RAB27A RAB27A GS2 Ashen Transport 41 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

MLPH Melanophilin GS3 Leaden Transport STX17 Syntaxin 17 no2 Elimitation TRAPPC6 TRAPPC6A Mhyp Transport A VAC14 VAC14/ARPIKFYVE Inglis (infantile Transport gliosis) FIG4 Fig/SAC3 CMT4J Pale tremor Maturation AP1S1 AP-1 σ1 subunit no2 Transport AP1S2 AP-1 σ2 subunit no2 Transport AP1S3 AP-1 σ3 subunit no2 Transport AP1M2 AP-1 µ2 subunit no2 Transport ATG7 Autophagy related 7 Perinatal lethality3 Elimination ATG13 Autophagy related 13 no1 Elimination BACE2 Bace2/ASP1 No phenotype Biogenesis CD63 Cd63 No coat color Biogenesis phenotype DCTN1 Dynactin 1 Embryonic lethal3 Transport F2RL1 F2RL1 No coat color Transfer phenotype GPNMB GPNMB Eye color Transfer phenotype KIF13A KIF13A No coat color Maturation/Transport phenotype KIF3A KIF3A Embryonic lethal3 Transport KIF3B KIF3B Embryonic lethal3 Transport KIF5B KIF5B Embryonic lethal3 Transport MREG Melanoregulin Dilute suppressor Transfer RAB32 RAB32 no2 Biogenesis RAB11B RAB11B no2 Transfer RAB38 RAB38 chocolate Biogenesis RILP RILP no2 Transport SLC24A5 OCA6 hypopigmentation Maturation TSG101 TSG101/VPS23 Embryonic lethal3 Maturation MYO6 Myosin VI no2 Biogenesis DYNLT3 DYNLT3 no2 Transport RAB3D RAB3D No coat color Elimination phenotype 939 940 941 1: no conditional mutant mice available 942 2: absence of mouse model 943 3: conditional mutant mice available. Inactivation in the melanocyte lineage was not 944 performed yet. 945

42

a WT bcat*

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. brightfield

inverse brightfield

b WT bcat* 150 150 ty (a.u.) ensi ty ty (a.u.) ensi ty 100 100 in t in t el ixel P Pix 50 50 0 50 100 0 50 100

Distance (a.u.) Distance (a.u.)

nb of melanosomes/cell Perinuclear melanosomes c ns d 6 40 *** 30 3 4 20 %

x 10 2 10

0 0 WT bcat* WT bcat* e WT bcat*

Tyrp1 DAPI

Figure 1 Aktary et al 0.0 0.5 1.0 1.5 2.0 2.5 3.0 k j i h d o n m l a.u. a.u. a.u. 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 20 40 60 MYO6 iβcat si iβcat si 0 D2ERTD391E Ctnnb1 mRNA LYST a MYO5A BLOC1S1 Ctnnb1 mRNA GFP Ctrl bioRxiv preprint Ctnnb1 mRNA Ctrl SLC45A2 RABGGTA BLOC1S2 bcat* HPS5 WT WT ***

*** OCA2 HPS3 *** KIF13A MLANA (which wasnotcertifiedbypeerreview)istheauthor/funder.Allrightsreserved.Noreuseallowedwithoutpermission. βcat RAB38 BLOC1S3

TRAPPC6A doi: FIG4 RAB32

STX17 https://doi.org/10.1101/2021.01.13.426541 VPS33A

GPNMB bcat* (10d)vs.WT MLPH

e HPS4 DCT GPR143 a.u. a.u. 0.0 0.5 a.u. 1.0 TYRP1

si si BACE2 0 1 2 3 0 1 2 3 VPS39 RAB27A

Dynlt3 mRNA AP3D1 VPS18 GFP Dynlt3 mRNA Ctrl Dynlt3 mRNA Ctrl AP3B1 AP1S3

bcat* PMEL WT WT AP1M2 *** F2RL1 *** VAC14

*** KIF5B VPS33B βcat βcat βcat VPS11 HPS6 CNO VPS41 ATG7

Figure 2 PLDN

KIF3B ;

AP1S2 this versionpostedJanuary13,2021. Dynlt3 Dynlt3 Dynlt3 RAB11B (9v) f actin actin actin VPS16 GFP βcat βcat βcat SNAPIN TYR

s s RILP

i i AP3M1 DTNBP1 TSG101 GFP

Ctrl AP1S1 Ctrl MREG bcat* HPS1 WT WT DCTN1 KIF3A

β AP3S1 βcat β DNAIC1 c

c DYNLT3 at

at FHIT MUTED RAB31 RAB3D 42 13 88 kDa 42 13 88 kDa 42 13 29 120 88 120 kDa

The copyrightholderforthispreprint β-actin c b g Dynlt3

a.u. a.u. GFP a.u. a.u a.u. 0.0 0.5 1.0 1.5 0.0 0.5 1.0 si si 0.0 0.5 1.0 0 1 2 3 0 1 2 3 Dynlt3 protein Dynlt3 protein Dynlt3 mRNA Ctrl Ctrl WT WT WT Dynlt3 protein - bcat* Aktary etal WT WT ** * ** ** * βcat βcat βcat bcat* bcat* bcat* 13 kDa 42 a WT si Ctrl si Dynlt3 bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

brightfield

inverse brightfield b si Ctrl Dynlt3 c WT siCtrl siDynlt3 Dynlt3 120 100 Actin ensi ty (au) 80 *** in t

1.0 el 60 a.u. 0.5 Pix 0 50 100 0 50 100 0.0 Distance (au) si Ctrl Dynlt3 Distance (au) d Perinuclear melanosomes e nb melanosomes/cell 5 ns 40 *** 4 3 30 3

20 x 10 % 2 10 1 0 0 si Ctrl Dynlt3 si Ctrl Dynlt3 f WT si Ctrl si Dynlt3

Tyrp1 DAPI

Figure 3 Aktary et al a bcat* GFP Dynlt3-GFP bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

brightfield

inverse brightfield

GFP

b bcat* GFP Dynlt3-GFP

(a .u.) 150

ty ensi ty 100 in t el 50 Pix 0 50 100 0 50 100 Distance (a.u.) Distance (a.u.)

c bcat* GFP Dynlt3-GFP

Tyrp1 DAPI

Figure 4 Aktary et al bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

WT bcat* ** a Total Distance b 1.0 Dynlt3 **** **** a.u 0.5 ns **** Actin 0.0 60 WT bcat* WT + **

GFP βcat 1.0 40 Dynlt3

a.u 0.5 Actin 0.0 GFP β-catGFP

Distance ( μ m) 20 WT + sh *** Ctrl Dynlt3 1.0 0 Dynlt3

a.u 0.5 -- GFP- GFP+ βcat+ βcat- shCtrl shDynlt3 Actin 0.0 sh Ctrl Dynlt3 bcat* WT

Euclidean Distance Pausing Melanosome s c d

15 **** **** **** 100 pau sed m) 10 μ

f t ime 50 o 5 Distance ( Distance ercen t P 0 0 -- GFP- GFP+ βcat+ βcat- shCtrl shDynlt3 -- GFP- GFP+ βcat+ βcat- shCtrl shDynlt3

bcat* WT bcat* WT

Figure 5 Aktary et al. bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1.0 a Not convective c e (NC) 100 ** 75 0.5 cdf WT + WT + 50 sh Ctrl sh Dynlt3 % C 25 (C) 0.0 0 0.4 0.6 0.8 1.0 1.2 1.4 α WTbcat* WT + bcat WT + sh MSD neg pos Ctrl Dynlt3 b Convective 104 d L WT + sh Ctrl C L C 2 103 S S

2

2S 10 μ -abs ( m )

NC WT + sh Dynlt3 101 NC NC 0 0 10 101 102 103 10 101 102 103 2 μ-abs (μm 2 ) μ-abs (μm )

Figure 6 Aktary et al bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint a (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 25 ** 25 ** 20 20

15 15

10 10

5 5

0 0 nb of acidic melanosomes/field WT bcat* nb of acidic melanosomes/field WT + siCtrl WT + siDynlt3 b c Plakoglobin Hmb45 Plakoglobin / Hmb45

WT WT

bcat* WT bcat* bcat* 50 * d 40 30 20 e Plakoglobin Hmb45 Plakoglobin / Hmb45 10 % Ker with Hmb45 0 WT bcat* WT + sh * Ctrl 50 f 40 30

20 WT + sh 10 Dynlt3 % Ker with Hmb45 0 WT + sh Ctrl Dynlt3

Figure 7 Aktary et al.

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 Dynlt3, a novel fundamental regulator of melanosome movement,

2 distribution, maturation and transfer

3

4 Zackie Aktary1-3, Alejandro Conde-Perez1-3, Florian Rambow1-3, Mathilde Di Marco6,7, 5 François Amblard3-5, Ilse Hurbain6,7, Graça Raposo6,7, Cédric Delevoye6,7, Sylvie 6 Coscoy3,4, & Lionel Larue1-3* 7

8

9

10 Supplementary information

11

12

13

14 This part is associated with Figure 5. Eight descriptors were defined: α, ρ, µ2L, µ2S,

15 µ3L, µ3S, µ4L and µ4S. The comparison between a set of experiments and a set of

16 simulated trajectories was performed by fitting the distributions of eight descriptors

17 (MSD exponent, anisotropy, the second [q=2], third [q=3, skewness] and fourth [q=4,

18 kurtosis] normalized moments projected on the two main L and S axis (Fig. 6 and

19 Supplementary Fig. 5). The seven other descriptors refer only to 2D spatial

20 characteristics and were obtained by projecting trajectory points on its own axis

21 system, giving two unidimensional distributions which were statistically described by

22 the computation of moments (E) of order q (2-4) (l corresponds to the center scalar

23 distribution of the projection of the long axis, n to sample size, the second moment

24 (when q=2) E2 [long axis] is named µ2L-abs), and by anisotropy defined as the ratio of

25 the second moments µ2S-abs / µ2L-abs (s means short axis).

26 Small amplitude WT melanosome subpopulation was non convective (fitting

27 with anisotropic Brownian movement + elastic tethering + noise), estimated to 39%

1

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 of WT melanosome population. In contrast, only half of the population of the bcat*

2 "small" was non-convective, and the other half was convective, estimated to 7% of

3 bcat* population.

4 Large amplitude WT and bcat* subpopulations correspond to convective

5 behavior of melanosomes. Convective behavior may correspond to two types of

6 movement: intermittent or uniform. We consider the movement uniform when the

7 velocity of each melanosome is constant during the whole observation period, and

8 intermittent if the conditions are not fulfilled. We cannot discriminate between the

9 uniform and intermittent behaviors of the wt or bcat* melanosomes. This is due to the

10 fact the velocity is too low compared to the diffusion.

11

2

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 Supplementary Figure Legends

2

3 Supplementary Figure 1. Melanosomes are distributed at the periphery of

4 melanocytes expressing exogenous activated β-catenin.

5 a) Brightfield (top), inverse brightfield (middle) and GFP (bottom) images of WT (9v)

6 melanocytes transiently expressing GFP or GFP-tagged β-catenin. Bar, 20 µm.

7 b) Quantitation of pixel intensity (in black) across the length of the WT-GFP and WT-

8 β-catGFP melanocytes shown in Panel A. Trendlines are shown in red.

9 c) Immunofluorescence analysis of WT melanocytes transiently expressing GFP or

10 GFP-tagged b-catenin. Cells were processed for immunofluorescence and

11 stained with anti-Tyrp1 (red) antibodies and the nuclei were stained with DAPI

12 (blue). Bar, 20 µm.

13

14 Supplementary Figure 2. bcat* melanosomes are larger than WT melanosomes.

15 a,b) Transmission electron microscopy of WT (a) and bcat* (b) melanocytes. Scale

16 bar: 2 µm.

17 c) Length of melanosome was determined from four WT and five bcat* melanocytes

18 with a total of at least 200 melanosomes per cell line. Statistical significance was

19 measured using Mann-Whitney test. **** p < 0.0001.

20

21 Supplementary Figure 3. Dynlt3, and not Dynlt1, is expressed in WT and

22 downregulated in bcat* melanocytes.

23 Ct values for RT-qPCR analysis of Dynlt3 (A) and Dynlt1 (B) mRNA expression in

24 two cell lines derived from WT mice (9v and 13d) and two from bcat* mice (10d and

25 14d).

3

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1

2 Supplementary Figure 4. Melanosome movement is increased in cells

3 producing active β-catenin or with diminished level of Dynlt3.

4 Melanosome movement (corresponding mainly to stage IV melanosome) was

5 assessed by brightfield video microscopy over a period of 5 minutes and

6 melanosome trajectories were followed using ImageJ software. For each cell line, 75

7 melanosomes were followed from a minimum of five independent cells. WT (9v) cell

8 lines were transfected with control GFP or β-catenin-GFP expression vectors,

9 respectively. Analyses were performed on both transfected (i.e. GFP-positive) and

10 non-transfected (i.e. GFP-negative) cells. In addition, WT cells were transfected with

11 either control or Dynlt3 shRNA vectors and analyses were performed on the resulting

12 red cells. Each dot represents one melanosome. The average distance (a) and the

13 average velocity (b) for each melanosome are shown, Statistical significance was

14 measured using Mann-Whitney test. **** p < 0.0001.

15

16 Supplementary Figure 5. Mathematical modeling of melanosome trajectories.

17 a,d) α MSD distributions for 9v and 10d (a), and 9v-Ctrl and 9v-bcat (d) populations

18 of trajectories, compared to the distribution for simulated Brownian (purely

19 diffusive – black line) random walks (nwalks=10 000, with the same nsteps (601)

20 as for experimental trajectories). Theoretical shifts compared to Brownian process

21 when adding convection, noise, tethering or confinement are indicated with

22 arrows.

23 b,c,e,f) µ2S, µ2L plots reflecting the extent of 9v (b), 10d (c), 9v Ctrl (e), and 9v bcat (f)

24 trajectories. Each point corresponds to one trajectory. Lines correspond to

4

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 frontiers between non-convective (NC) and convective (C) trajectories: see SI for

2 criteria and validation.

3 g,h,i) α MSD, anisotropy, L and S second moments, L and S skewness, and L and S

4 kurtosis distributions for 9v and 10d (g), 9v-Ctrl and 9v-bcat (h), and 9v Sh Ctrl

5 and 9v Sh Dynlt3 (i) populations of trajectories, compared to the distribution for

6 simulated Brownian (purely diffusive – black line).

7 j,k,l) Criteria to establish the frontiers for non-convective (NC) and convective (C)

8 movement for α MSD, anisotropy, and L and S second moments distributions. On

9 the top and the bottom the lower and upper subdivisions are given, respectively.

10

11 Supplementary Figure 6. FACS analysis of melanosome transfer.

12 a) FACS analysis of Hmb45 and plakoglobin positive cell populations in WT and

13 bcat* melanocytes.

14 b) Quantitation of melanosome transfer assayed by FACS. WT (9v), bcat* (10d), WT

15 + sh Ctrl or WT + sh Dynlt3 melanocytes were co-cultured with Balb/c MK mouse

16 keratinocytes for 10 days, after which the co-cultures were fixed and processed

17 for FACS analyses using Hmb45 and Plakoglobin antibodies. The percentage of

18 Plakoglobin and Hmb45 double positive cells within each respective co-culture is

19 presented. Quantitation was done from a minimum of five independent co-culture

20 experiments.

21

5

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 4 5 Supplementary Table 1. Mouse models associated with mutations in 67 6 genes involved in trafficking of melanosomes 7 and other lysosome related organelles (LRO). 8 9 10 11 12 Gene Gene Product Human Mouse Model Stage Disease TYR TYR OCA1 Albino Maturation OCA2 OCA2/P OCA2 Pink eye dilution Maturation TYRP1 TYRP-1 OCA3 Brown Maturation SLC45A2 SLC45A2/MATP OCA4 Underwhite Maturation DCT DCT/TRP-2 Slaty Maturation PMEL PMEL/Pmel17/Silver/ Silver Maturation gp100 GPR143 GPR143/OA1 OA1 Oa1 Biogenesis MLANA MART1/melan-a Melan-a Maturation AP3B1 AP-3β3a subunit HPS2 Pearl Transport AP3D1 AP3 δ subunit Mocha Transport AP3M1 AP-3 µ3a subunit no2 Transport AP3S1 AP-3 σ3a subunit no2 Transport BLOC1S5 Muted Muted Transport BLOC1S4 Cappuccino HPS1 Cappuccino Transport DTNBP1 Dysbindin HPS7 Sandy Maturation BLOC1S3 BLOS3 HPS8 Reduced Biogenesis pigmentation PLDN Pallidin HPS9 Pallid Maturation BLOC1S1 BLOS1 No reported coat Transport color BLOC1S2 BLOS2 no2 Transport SNAPIN Snapin Perinatal lethality1 Transport HPS1 HPS1 HPS1 Pale ear Biogenesis HPS3 HPS3 HPS3 Cocoa Biogenesis HPS4 HPS4 HPS4 Light ear Biogenesis HPS5 HPS5 HPS5 Ruby-eye 2 Biogenesis HPS6 HPS6 HPS6 Ruby-eye Biogenesis VPS11 VPS11 no2 Transport VPS16 VPS16 no2 Transport VPS18 VPS18 no reported coat Transport color3 VPS33A VPS33A Buff Maturation VPS39 VPS39/Vam6 no2 Transport VPS41 VPS41 Embryonic lethal3 Transport VPS33B VPS33B ARCS1 no2 Maturation C14ORF13 VIPAR/VPS16B ARCS2 no2 Transport 3 RABGGTA Subunit of Rab Gunmetal Biogenesis geranylgeranyl transferase II LYST LYST CHS Beige Transport MYO5A Myosin VA GS1 Dilute Transport RAB27A RAB27A GS2 Ashen Transport

6

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. MLPH Melanophilin GS3 Leaden Transport STX17 Syntaxin 17 no2 Elimitation TRAPPC6 TRAPPC6A Mhyp Transport A VAC14 VAC14/ARPIKFYVE Inglis (infantile Transport gliosis) FIG4 Fig/SAC3 CMT4J Pale tremor Maturation AP1S1 AP-1 σ1 subunit no2 Transport AP1S2 AP-1 σ2 subunit no2 Transport AP1S3 AP-1 σ3 subunit no2 Transport AP1M2 AP-1 µ2 subunit no2 Transport ATG7 Autophagy related 7 Perinatal lethality3 Elimination ATG13 Autophagy related 13 no1 Elimination BACE2 Bace2/ASP1 No phenotype Biogenesis CD63 Cd63 No coat color Biogenesis phenotype DCTN1 Dynactin 1 Embryonic lethal3 Transport F2RL1 F2RL1 No coat color Transfer phenotype GPNMB GPNMB Eye color Transfer phenotype KIF13A KIF13A No coat color Maturation/Transport phenotype KIF3A KIF3A Embryonic lethal3 Transport KIF3B KIF3B Embryonic lethal3 Transport KIF5B KIF5B Embryonic lethal3 Transport MREG Melanoregulin Dilute suppressor Transfer RAB32 RAB32 no2 Biogenesis RAB11B RAB11B no2 Transfer RAB38 RAB38 chocolate Biogenesis RILP RILP no2 Transport SLC24A5 OCA6 hypopigmentation Maturation TSG101 TSG101/VPS23 Embryonic lethal3 Maturation MYO6 Myosin VI no2 Biogenesis DYNLT3 DYNLT3 no2 Transport RAB3D RAB3D No coat color Elimination phenotype 1 2 3 1: no conditional mutant mice available 4 2: absence of mouse model 5 3: conditional mutant mice available. Inactivation in the melanocyte lineage was not 6 performed yet. 7

7

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. a WT-GFP WT-βcatGFP brightfield

inverse brightfield

GFP

9v 10d

b WT-GFP ) WT-βcatGFP (a .u.) 150 (a .u. 150 ty ty ensi ty en si 100 100 in t in t el el Pix Pix 50 50 0 50 100 0 50 100 Distance (a.u.) Distance (a.u.) β c WT-GFP WT- catGFP

Tyrp1 DAPI

Supplementary Figure 1 Aktary et al bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

a WT b bcat*

c 800 ****

600

400

200 melanosome length (nm) 0 WT bcat*

Supplementary Figure 2 Aktary et al bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

a Dynlt3 mRNA level 23

22 val ue t 21 C

20 9v 13d 10d 14d

WT bcat*

b Dynlt1 mRNA level 32

31 val ue

t 30 C

29 9v 13d 10d 14d

WT bcat*

Supplementary Figure 3 Aktary et al bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

a Average Distance

**** **** **** 0.15

0.10

0.05 Distance ( μ m)

0.00 -- GFP- GFP+ βcat+ βcat- shCtrl shDynlt3

10d 9v

b Average Velocity

0.25 **** **** ****

0.20

0.15

0.10 Velocity ( μ m/sec) Velocity 0.05

0.00 -- GFP- GFP+ βcat+ βcat- shCtrl shDynlt3

bcat* WT

Supplementary Figure 4 Aktary et al. 4 1.0 10 a b WT c bcat*

bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.4265412 3 ; this version posted January 13, 2021. The copyright holder for this preprint (NC)(which was not certified by peer review) is the author/funder.10 All rights reserved. No reuse allowed without permission. 0.5 cdf WT bcat* 2 10

(C) 2S

μ -abs ( m ) 1 10 0.0

4 1.0 10 d e L WT Ctrl C f WT bcat

2 3 (NC) 10 WT WT S 0.5 cdf 2 Ctrl bcat 10

(C) 2S

μ -abs ( m ) 1 0.0 10 NC 0 1 2 3 0 1 2 3 0.4 0.6 0.8 1.0 1.2 1.4 10 10 10 10 10 10 10 10 α MSD μ-abs (μm 2 ) μ-abs (μm 2 )

g WT bcat* h WT Ctrl WT bcat i WT Sh Ctrl WT Sh Dynlt3 MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution 1.0 1.0 1.0

0.5 0.5 0.5

0.0 0.0 0.0 -3 -2 -1 0 -3 -2 -1 0 -3 -2 -1 0 0.6 1.0 1.4 10 10 10 10 0.6 1.0 1.4 10 10 10 10 0.6 1.0 1.4 10 10 10 10 L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution 1.0 1.0 1.0

0.5 0.5 0.5

0.0 0.0 0.0 0 0 0 0 0 0 10 10 10 10 10 10 L skewness distribution S skewness distribution L skewness distribution S skewness distribution L skewness distribution S skewness distribution 1.0 1.0 1.01.0

0.5 0.5 0.5

0.0 0.0 0.0 -1-2 210 -1 0 21 -1-2 210 -1 0 21 -1-2 210 -1 0 21 L kurtosis distribution S kurtosis distribution L kurtosis distribution S kurtosis distribution L kurtosis distribution S kurtosis distribution 1.0 1.0 1.0

0.5 0.5 0.5

0.0 0.0 0.0 2 34 5 6 2 4 6 8 2 34 5 6 2 4 6 8 2 34 5 6 2 4 6 8

j WT bcat* k WT Ctrl WT bcat l WT Sh Ctrl WT Sh Dynlt3 lower subdivsion 56NC / 19C lower subdivsion 14NC / 61C lower subdivsion 60NC / 15C lower subdivsion 25NC/ 50C lower subdivsion 36NC / 39C lower subdivsion 14NC / 61C

MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution 1.0 1.0 1.0 1.0 1.0 1.0

0.5 0.5 0.5 0.5 0.5 0.5

0.0 0.0 0.0 0.0 0.0 0.0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution 1.0 1.0 1.0 1.0 1.0 1.0

0.5 0.5 0.5 0.5 0.5 0.5

0.0 0.0 0.0 0.0 0.0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 10 10 10 10 10 10 10 10 10 10 10 10

upper subdivsion 62 NC / 13C upper subdivsion 18NC/ 57C upper subdivsion 64NC/ 11C upper subdivsion 27NC/ 48C upper subdivsion 43NC / 32C upper subdivsion 22NC / 53C

MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution MSD exponent distribution Anisotropy distribution 1.0 1.0 1.0 1.0 1.0 1.0

0.5 0.5 0.5 0.5 0.5 0.5

0.0 0.0 0.0 0.0 0.0 0.0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 -4 -2 0 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 0.6 1.0 1.4 10 10 10 L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution L 2nd moment distribution S 2nd moment distribution 1.0 1.0 1.0 1.0 1.0 1.0

0.5 0.5 0.5 0.5 0.5 0.5

0.0 0.0 0.0 0.0 0.0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 10 10 10 10 10 10 10 10 10 10 10 10

Supplementary Figure 5 Aktary et al. a WT bcat* bioRxiv preprint doi: https://doi.org/10.1101/2021.01.13.426541; this version posted January 13, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. Plakoglobin

Pmel Pmel

10 10 b * * s 8 8

6 6 mel + c el l mel + c el ls

4 4 and P and P g g g g P

P 2 2 % % 0 0 WT bcat* WT + sh Ctrl Dynlt3 c WT melanocyte population Rabbit IgG Mouse IgG

Pmel Plakoglobin

Supplementary Fig. 6 Aktary et al.