1 A mitochondria-anchored isoform of the actin-nucleating Spire protein 2 regulates mitochondrial division

3 Uri Manor*1, Sadie Bartholomew*2, Gonen Golani3, Eric Christenson4, Michael 4 Kozlov3, Henry Higgs5, James Spudich2, and Jennifer Lippincott-Schwartz1

5 *Contributed equally

6 1Cell Biology and Metabolism Program, Eunice Kennedy Shriver National 7 Institute of Child Health and Human Development, Bethesda, MD 20892

8 2Department of Biochemistry, Stanford University School of Medicine, Stanford, 9 CA 94305, USA.

10 3Department of Physiology and Pharmacology, Tel Aviv University, Tel Aviv, Israel. 11 12 4Unit on Structural and Chemical Biology of Membrane Proteins, Eunice 13 Kennedy Shriver National Institute of Child Health and Human Development, 14 Bethesda, MD 20892

15 5Department of Biochemistry, Geisel School of Medicine at Dartmouth, Hanover, 16 NH 03755, USA

17 Mitochondrial division, essential for survival in mammals, is enhanced by 18 an inter-organellar process involving ER tubules encircling and 19 constricting mitochondria. The force for constriction is thought to involve 20 actin polymerization by the ER-anchored isoform of the formin protein 21 INF2. Unknown is the mechanism triggering INF2-mediated actin 22 polymerization at ER-mitochondria intersections. We show that a novel 23 isoform of the formin-binding, actin-nucleating protein Spire, Spire1C, 24 localizes to mitochondria and directly links mitochondria to the actin 25 cytoskeleton and the ER. Spire1C binds INF2 and promotes actin assembly 26 on mitochondrial surfaces. Disrupting either Spire1C actin- or formin- 27 binding activities reduces mitochondrial constriction and division. We 28 propose Spire1C cooperates with INF2 to regulate actin assembly at ER- 29 mitochondrial contacts. Simulations support this model’s feasibility and 30 demonstrate polymerizing actin filaments can induce mitochondrial 31 constriction. Thus, Spire1C is optimally positioned to serve as a molecular 32 hub that links mitochondria to actin and the ER for regulation of 33 mitochondrial division.

34 35 Introduction 36 Mitochondrial division is a complex process that is essential for survival in 37 mammals1,2, and is facilitated by the actin cytoskeleton3-8. Two distinct steps 38 define mitochondrial division - an initial constriction of mitochondrial membranes, 39 followed by final membrane scission5,7,9,10. Scission is mediated by the dynamin-

1 40 related protein, Drp1, which self-assembles on the surface of the mitochondrial 41 outer membrane into helices that drive final mitochondrial division2,11. The initial 42 constriction step narrows the mitochondrial tube diameter, which is necessary for 43 Drp1 helix assembly9,10,12-16. This step is independent of Drp1 and occurs at ER- 44 mitochondria intersection zones where ER tubules associate with and wrap 45 around the mitochondrial outer membrane along the plane of mitochondrial 46 division5,7,9,10. Along these zones, actin filaments polymerize, providing the force 47 needed for constriction that floppy ER tubules lack5-7. 48 49 Inverted formin 2 (INF2) is a formin family protein that promotes actin filament 50 polymerization in a regulated fashion5,7. An ER-anchored splice isoform of INF2 51 (usually referred to as INF2-CAAX)7,17 has been shown to facilitate mitochondrial 52 constriction and division via its actin polymerization activity7. Given INF2, but not 53 actin assembly, is localized throughout the ER, how INF2-mediated actin 54 assembly is specifically triggered at ER-mitochondria intersections to ensure 55 mitochondrial division remains an open central question. 56 57 Spire proteins are membrane-binding actin-nucleators that interact with and 58 regulate formin proteins18-27. Synergistic promotion of actin assembly by Spire 59 and formin proteins has been implicated in driving a variety of processes ranging 60 from vesicle trafficking to DNA repair in the nucleus21,27-29. In these systems, 61 membrane-associated Spire proteins nucleate actin filaments, which are then 62 further polymerized by formin proteins, ultimately leading to actin-dependent 63 translocation. Given these characteristics of Spire proteins, we set out to 64 investigate whether any Spire proteins could be involved in helping promote 65 INF2- and actin-dependent constriction and division of mitochondria at ER- 66 mitochondria association zones. 67 68 Vertebrates have two known Spire genes, Spire1 and Spire2. Each Spire protein 69 contains highly conserved domains with specific capabilities, including: four actin- 70 monomer binding WH2 domains necessary for nucleating actin filaments; an 71 mFYVE domain that binds to intracellular membranes and facilitates 72 oligomerization26,30; and an N-terminal KIND domain that serves to bind to and 73 regulate formin proteins18,19,21-27. Although no Spire proteins have been shown 74 previously to localize to mitochondria or the ER30, here we report a previously 75 uncharacterized alternate splice-isoform of Spire1 (named Spire1C) that localizes 76 to mitochondria, promotes actin assembly on mitochondrial surfaces, and 77 interacts with ER-anchored INF2 to regulate mitochondrial constriction and 78 division. Our results support a model where Spire1C and INF2 coordinately drive 79 actin- and ER-dependent mitochondrial division. They also reveal that Spire1C 80 directly links mitochondria to both the actin cytoskeleton and the ER. 81 82 Results 83 Spire1C’s previously uncharacterized alternate exon, ExonC, is highly conserved 84 We identified and characterized a novel alternate splice-isoform of Spire1 that 85 contains KIND and WH2 domains common to all Spire proteins, as well as a

2 86 previously uncharacterized unique 58 amino acid alternate exon sequence (Exon 87 C) (Figure 1A and see Materials and Methods and Figure 1-figure supplements 88 1-3 for details on Spire1C cloning, probe generation, and sequence information). 89 Because of its alternative ExonC sequence, we named the Spire1 isoform 90 Spire1C (Figure 1A). After determining that Spire1C mRNA was present in 91 multiple mouse tissues (Figure 1-figure supplement 3), we examined its presence 92 and conservation among species. Using the UCSC Genome Browser, we 93 searched for additional DNA or mRNA sequences that contain Spire1C31. When 94 compared to mouse, we found striking identity in ExonC for rat, rabbit, human, 95 dog, elephant, opossum, platypus, and chicken. It is not found in the annotated 96 zebrafish sequence. Compared to our amplified mouse sequence of 58 amino 97 acids, human ExonC differs by 2 residues (96% identical), platypus by 8 residues 98 (86% identical), and chicken by 12 residues (79% identical). Sequence homology 99 is strongest among mammals, but chicken maintains conservation that is unlikely 100 to be solely due to chance. Conservation among species extends beyond the 101 coding exon and into the upstream and downstream intronic regions of the gene, 102 stretching ~150 bases in the 3’ direction (data not shown). These conserved 103 extensions are likely involved in splicing regulation of the exon. The high level of 104 conservation within and surrounding ExonC suggests that its role is indispensible 105 for the health of the organism. 106 107 Spire1C’s ExonC is necessary and sufficient for localization to mitochondria 108 When cells were transfected with a myc-tagged Spire1C construct, the protein 109 showed extensive co-distribution with the mitochondrial marker mitoRFP (Figure 110 1B, myc-Spire1C). A polyclonal antibody generated against a peptide containing 111 the unique 58 amino acids in ExonC (Figure 1B and Figure 1-figure supplements 112 1-3) also showed extensive mitochondria-specific labeling within cells (Figure 1B, 113 α−ExonC) (see Figure 1-figure supplements 2-3 and Materials and Methods for 114 additional information on α-ExonC). Testing the role of ExonC in mitochondrial 115 targeting of Spire1C, we found that a GFP-tagged ExonC fusion protein robustly 116 localized to mitochondria in expressing cells (Figure 1B, GFP-ExonC). By 117 contrast, a myc-tagged Spire1C construct lacking ExonC never showed specific 118 mitochondrial localization (Figure 1B, myc-Spire1ΔC). These results suggest that 119 Spire1C is an endogenously expressed mitochondria-associated protein that 120 targets to mitochondria via its ExonC domain. 121 122 Spire1C localizes to the mitochondrial outer membrane with its formin- and actin- 123 binding domains facing the cytoplasm 124 To examine whether Spire1C distributes on the surface or interior of 125 mitochondria we compared the distribution of GFP-ExonC (marking Spire1C) 126 with mitoRFP (marking the mitochondrial matrix) using structured illumination 127 microscopy, which gives a two-fold resolution improvement over conventional 128 confocal imaging32. GFP-ExonC labeling on mitochondria surrounded that of 129 mitoRFP labeling (Figure 2A), suggesting Spire1C localizes to the periphery of 130 mitochondria, most likely on the mitochondrial outer membrane. 131 To confirm Spire1C’s mitochondrial outer membrane localization, we employed a

3 132 fluorescence protease protection (FPP) assay33, which can determine a protein’s 133 membrane topology (Figure 2B). GFP was fused to the N-terminus of Spire1C 134 (GFP-Spire1C), the N-terminus of ExonC (GFP-ExonC), or to the C-terminus of 135 ExonC (ExonC-GFP) (Figure 2C). The constructs were then co-expressed in 136 cells with OMI-mCherry, a mitochondrial intermembrane space (IMS) protein34. 137 Thereafter, the plasma membrane of the cells was gently permeabilized with 138 digitonin, followed by treatment with trypsin to extinguish cytoplasmic GFP 139 fluorescence. We reasoned that only if the fluorescent tag from these constructs 140 faced the cytoplasm would their fluorescence be abolished by the trypsin. Both 141 GFP-Spire1C and GFP-ExonC lost nearly all their fluorescence within 60 142 seconds of trypsin treatment. By contrast, fluorescence from ExonC-GFP was 143 protected, similar to that seen for co-expressed OMI-mCherry, which as a 144 mitochondrial IMS protein should be insensitive to trypsin34 (Figure 2D; Videos 1- 145 3). These results suggest that the WH2 and KIND domains of Spire1C face the 146 cytoplasm (since both reside N-terminal to ExonC) (Figure 2C), a topology where 147 they could participate in formin-binding and actin-nucleation activity. The data 148 further suggest that the C-terminus of ExonC is not exposed to the cytoplasm. 149 This raises the possibility that ExonC is embedded in the outer membrane, either 150 as a transmembrane or hairpin protein. 151 152 In support of this notion, transmembrane domain prediction software35 indicated 153 that residues 26-46 within ExonC form an α-helix characteristic of prokaryotic 154 transmembrane domains. Secondary structure prediction software PHYRE36 also 155 predicted a second α-helix within ExonC (Figure 1-figure supplement 3). 156 Interestingly, when we expressed a full-length Spire1C construct with GFP fused 157 to the C-terminus (Spire1C-GFP), the protein remained mostly cytoplasmic and 158 no longer properly localized to mitochondria (data not shown). The protein also 159 rapidly escaped cells treated with digitonin, suggesting it was not able to target 160 efficiently to mitochondria. Taken together, our data suggests that Spire1C is a 161 mitochondrial outer membrane protein, possibly with a hairpin conformation given 162 ExonC’s two predicted α-helix domains (Figure 1-figure supplement 3). 163 164 We next performed photobleaching experiments to investigate the dynamics of 165 Spire1C’s association with mitochondrial membranes. Photobleaching of a 166 portion of a mitochondrial element that expressed GFP-Spire1C resulted in rapid 167 recovery of fluorescence, with replenishment arising first in regions close to the 168 bleach site and later at regions further away (Figure 2E and Figure 2-figure 169 supplement 1), similar to that seen for GFP-tagged proteins that feely diffuse 170 along membranes37. In contrast, very little recovery during the same time period 171 occurred when an entire mitochondrial element expressing GFP-Spire1C was 172 photobleached (Figure 2F). Thus, Spire1C appears to diffuse laterally along 173 mitochondrial membranes, rather than rapidly bind and dissociate from the 174 membrane, further supporting the idea of Spire1C being a mitochondrial 175 transmembrane protein. 176 177 Spire1C promotes actin assembly on mitochondrial surfaces

4 178 Given that Spire proteins can nucleate actin via their highly conserved WH2- 179 repeat domain38-40, we next investigated whether overexpression of Spire1C 180 promotes actin assembly on mitochondria. In non-transfected cells, only low 181 levels of actin co-localized with mitochondria, without any apparent specificity 182 (Figure 3 and Figure 3-figure supplement 1, control). Upon Spire1C 183 overexpression, however, actin accumulated to high levels specifically on 184 mitochondria (Figure 3 and Figure 3-figure supplement 1, Spire1C 185 overexpression). This actin enrichment on mitochondrial surfaces was not 186 dependent on Spire’s formin-binding KIND domain, since overexpression of 187 Spire1C lacking the KIND domain (Spire1CΔKIND) still induced actin enrichment 188 on mitochondrial surfaces (Figure 3, Spire1CΔKIND overexpression). In contrast, 189 actin enrichment on mitochondria was muted upon overexpression of Spire1C 190 mWH2 (Figure 3 and Figure 3-figure supplement 1, Spire1C mWH2 191 overexpression), which contains mutations in its WH2 domains that block Spire- 192 mediated actin nucleation39,40 (see Materials and Methods). These results 193 demonstrated that Spire1C promotes actin assembly on mitochondria, most likely 194 through Spire1C’s ability to nucleate actin filaments. 195 196 Spire1C modulates mitochondrial fission via its formin-binding KIND and actin- 197 nucleating WH2-repeat domains 198 Given that actin assembly has been shown to play an important role in regulating 199 mitochondrial fission3-8, the observation that Spire1C localizes to mitochondria 200 and promotes actin assembly on mitochondrial surfaces raised the possibility that 201 Spire1C plays a role in mitochondrial fission. To test this directly, we assessed 202 the effect of Spire1C overexpression, mutation and depletion on mitochondrial 203 morphology, length, and fission. While all cells in all conditions in this study 204 displayed a combination of fragmented and tubular mitochondria, overexpression 205 of Spire1C resulted in a significant shift towards fragmented mitochondria 206 compared to control cells (Figure 4A, +Spire1C). In contrast, overexpression of 207 Spire1C mWH2 or Spire1CΔKIND resulted in a shift towards tubular mitochondria 208 (Figure 4A). Depletion of Spire1C using shRNA also resulted in a shift towards 209 tubular mitochondria (Figure 4B). To quantify these changes, we measured 210 mitochondrial lengths in each of these conditions, and found that Spire1C 211 overexpression resulted in shorter mitochondria on average, whereas mutation or 212 depletion of Spire1C resulted in longer mitochondria (Figure 4C, left graph, and 213 Figure 4-figure supplement 1), resembling the effect of disrupting mitochondrial 214 fission due to the depletion of Drp1 or INF2 from cells5-7,15,41,42. To determine 215 whether these morphological changes were a result of altered mitochondrial 216 fission dynamics, we counted the number of mitochondrial fission events in each 217 of these conditions. We found that fission events increased in frequency when 218 cells were overexpressing Spire1C, whereas overexpression of Spire1C mWH2 219 or Spire1CΔKIND decreased the frequency of fission events (Figure 4C, right 220 graph). Similarly, shRNA-mediated knockdown of Spire1C decreased the 221 frequency of fission events (Figure 4C, right graph). Taken together, our results 222 show Spire1C promotes mitochondrial fission via Spire1C’s actin-nucleating and 223 formin-binding capabilities.

5 224 225 Disrupting the Spire1C KIND domain decreases ER-mitochondria overlaps 226 Since ER tubules have been implicated in mitochondrial constriction and 227 division5,7,9,10, we investigated whether Spire1C influences ER-mitochondria 228 association. Upon overexpression of Spire1C or Spire1C mWH2, we observed 229 no significant change in ER-mitochondrial overlap or crossover sites of ER 230 tubules and mitochondria compared to control cells (Figure 5A, B). However, in 231 cells overexpressing Spire1CΔKIND, we observed a significant decrease in ER- 232 mitochondria overlap, as well as a decrease in the number of ER tubules 233 crossing over mitochondria (Figure 5A,B). This suggested that the KIND domain 234 of Spire1C might play a role in regulating the extent of ER-mitochondria 235 intersections within cells, and that Spire1CΔKIND-mediated disruption of 236 mitochondrial fission (see Figure 4C, right panel) could be due to a reduction in 237 ER-mediated mitochondrial constriction in these cells7,9,10. 238 239 The Spire1C KIND domain directly interacts with ER-anchored INF2 to promote 240 mitochondrial fission 241 One way the KIND domain of Spire1C could affect ER-mediated mitochondrial 242 division is by binding to ER-anchored INF2. To test this possibility, we performed 243 in vitro GST pull-down assays. We found that the C-terminal half of INF2 (INF2- 244 CT), but not the N-terminal half (INF2-NT), associated with a GST-tagged 245 Spire1C KIND domain, but not GST alone (Figure 6A). Addition of the N-terminal 246 half of INF2 (INF2-NT) inhibited the interaction between Spire1C KIND and INF2- 247 CT in our GST pulldown assays (Figure 6A; last well), suggesting that INF2’s 248 ability to self-interact43-45 can regulate its association with Spire1C. We further 249 confirmed the interaction between Spire1C’s KIND domain and INF2 using 250 fluorescence anisotropy43 (Figure 6B), which showed a specific interaction 251 between Spire1C’s KIND domain and INF2. Taken together, these data 252 demonstrate that the Spire1C KIND domain directly binds to INF2. 253 254 We next tested whether disrupting Spire1C’s interaction with INF2 inhibits 255 mitochondrial fission. To test this hypothesis, we first asked whether removing 256 the KIND domain from Spire1C blocks the increase in mitochondrial division 257 associated with overexpressing a constitutively active INF2 mutant, INF2 A1495,7. 258 Consistent with previous reports5,7, we found that overexpressing INF2 A149 259 resulted in significant shortening of mitochondria (Figure 6C,E; A149 alone). 260 Similarly, co-overexpression of INF2 A149 and Spire1C or INF2 A149 and 261 Spire1C mWH2 resulted in very short, fragmented mitochondria (Figure 6C,E 262 +Spire1C or +Spire1C mWH2). By contrast, overexpression of INF2 A149 and 263 Spire1CΔKIND significantly disrupted INF2 A149-mediated mitochondrial 264 fragmentation (Figure 6C, +Spire1CΔKIND). These results suggest that while 265 Spire1C actin-nucleating activity may be unnecessary for mitochondrial fission 266 when INF2 is constitutively active, INF2 binding to the Spire1C KIND domain is 267 necessary for INF2 to maximally induce mitochondrial fission. In further 268 confirmation of the hypothesis that Spire1C and INF2 jointly work to drive 269 mitochondrial fission, we found that siRNA-mediated knockdown of INF2

6 270 disrupted Spire1C-mediated upregulation of mitochondrial fission (Figure 6D). 271 Taken together, the results suggest that Spire1C interacts with INF2 via 272 Spire1C’s KIND domain, and that this interaction promotes mitochondrial fission. 273 274 Spire1C promotes ER-mediated mitochondrial constriction in an INF2-dependent 275 fashion 276 Mitochondrial fission would be co-dependent on Spire1C and INF2 if Spire1C’s 277 interaction with INF2 drives ER-mediated mitochondrial constriction. To test this 278 hypothesis, we employed confocal fluorescence imaging of ER and mitochondria 279 to examine mitochondrial constriction sites in cells co-expressing the ER marker 280 Ii33-mCherry and different variants of Spire1C. Mitochondria constriction was 281 visible at sites of ER-mitochondria crossover slightly more frequently in cells 282 overexpressing Spire1C compared to cells not overexpressing the construct 283 (Figure 7A-B, Spire1C, see arrows for ER-mediated mitochondrial constrictions, 284 and arrowheads for ER-mitochondria intersections that didn’t result in 285 constrictions). Notably, cells overexpressing Spire1C mWH2 displayed 286 significantly fewer mitochondrial constrictions at ER-mitochondria intersections 287 (Figure 7A-B, Spire1C mWH2). Similarly, cells overexpressing Spire1CΔKIND 288 showed a decrease in the frequency of constrictions at ER-mitochondria 289 intersections (Figure 7A-B, Spire1CΔKIND). RNA-mediated Spire1C knockdown 290 also resulted in decreased constrictions (Figure 7A-B, Spire1C shRNA). Finally, 291 overexpressing Spire1C in cells treated with INF2 siRNA showed a significant 292 reduction in mitochondrial constrictions (Figure 7A-B, INF2 siRNA+Spire1C). 293 Taken together, our results suggest that Spire1C and INF2 work together to 294 promote mitochondrial division by driving ER-mediated mitochondrial 295 constriction, and that this process is dependent on Spire1C’s ability to nucleate 296 actin filaments on mitochondrial surfaces, as well as the ability for Spire1C and 297 INF2 to interact via the Spire1C KIND domain. 298 299 Simulations indicate that pressure from actin polymerization and actomyosin 300 contraction forces are sufficient for driving mitochondrial constriction. 301 Given the above experimental results, we used in silico simulations to test a 302 potential model in which forces mediated by the actin cytoskeleton induce 303 mitochondrial constriction. In this scenario, actin filament polymerization within 304 the gap between the ER tubule surrounding the mitochondria and the 305 mitochondrial outer membrane (Figure 8A) results in localized pressure that 306 drives mitochondrial constriction to diameters required for Drp1 helix formation7. 307 This pressure could originate either from forces exerted by actin polymerization 308 against the mitochondrial outer membrane7, or by myosin-II dimer mediated 309 contraction of actin filaments lying between the ER and mitochondrial 310 membranes5,6, or by a concerted action of these two complementary 311 mechanisms. 312 313 To substantiate this, we created a simulation of mitochondrial constriction in 314 response to a localized pressure generated by the above-mentioned 315 mechanisms (Figure 8B, Figure 8-figure supplement 1, and Figure 8-source data

7 316 1). We modeled the constriction site of the mitochondrial outer membrane as a 317 membrane tubule whose resistance to deformations is characterized by a 318 bending modulus of 8⋅10-20 Joules, typical for a lipid bilayer46. The pressure 319 deforming the membrane tubule was applied in the middle of the constriction 320 zone along a strip of 50-nm thickness, corresponding to that of a typical ER 321 tubule, while the computed shapes of the mitochondria constriction region 322 corresponded to those of three different constriction events imaged with electron 323 tomography10 (Figure 8B). The computed pressure values required for generation 324 of these three degrees of mitochondrial constriction (Figure 8-figure supplement 325 1 and Figure 8-source data 1) enabled us to calculate the numbers of 326 polymerizing actin filaments, Nf, or the tension, γm, which has to be developed 327 within the actin contractile system. Assuming that the force developed by one 47 328 polymerizing actin filament is about 1 pN , the estimated filament number, Nf, 329 varies between 10-20. The actomyosin tension values, γm, range from 2 to 3 pN. 330 The obtained estimations for both Nf and γm are perfectly reasonable 331 physiologically, which supports the feasibility of the suggested mechanisms. 332 Thus, our results and model are fully consistent with previous studies suggesting 333 that tightly regulated actin assembly at ER-mitochondria intersection sites 334 facilitates mitochondrial membrane scission by Drp14-7,9,10. 335 336 Discussion 337 A key event in the mitochondrion’s life cycle is its division into distinct 338 mitochondrial elements. Prior work studying this process demonstrated that 339 division occurs at sites where ER wraps around mitochondria9,10, with the ER 340 providing a platform for actin polymerization mediated by the ER-anchored formin 341 INF25-7. This actin meshwork is proposed to provide the force that drives 342 mitochondrial constriction prior to Drp1-mediated mitochondrial division. Missing 343 from this picture has been a molecular player that regulates INF2-mediated actin 344 polymerization, ensuring that polymerization occurs specifically at ER- 345 mitochondria contact sites to drive mitochondrial constriction and division. Here, 346 we demonstrate that Spire1C, a novel mitochondrial outer membrane protein, 347 can serve this role by both binding to INF2 as well as by acting as an actin- 348 nucleator. 349 350 Spire proteins are membrane-binding actin-nucleators that interact with and 351 regulate formin proteins18-24. Given this, Spire proteins are potential candidates 352 for regulating the actin polymerization activity of INF2 on mitochondrial 353 membranes. In searching for such a protein, we identified a specific isoform of 354 Spire1, Spire1C, which resides on mitochondria and interacts with INF2. Spire1C 355 is distinct from other Spire proteins in having mitochondrial outer membrane 356 localization. This localization is a result of Spire1C’s unique ExonC domain, 357 which serves as a mitochondria-targeting sequence. Spire1C undergoes lateral 358 diffusion on the mitochondrial outer membrane, and is oriented with its formin- 359 binding and actin-nucleating domains facing the cytoplasm. Spire1C promotes 360 actin assembly on mitochondrial outer membranes; when Spire1C is 361 overexpressed a massive buildup of actin around mitochondria is observed. The

8 362 actin buildup is dependent on Spire1C’s actin-nucleating WH2 domain, but not its 363 formin-binding KIND domain. Therefore, Spire1C’s canonical actin-nucleating 364 domain drives actin accumulation on mitochondria independently of its 365 interactions with formin proteins. 366 367 Given Spire1C’s ability to assemble actin filaments on mitochondrial membranes, 368 we examined whether modulating Spire1C’s activity could affect mitochondrial 369 lengths or their division rates, and we found that it can. Overexpressing Spire1C 370 increases mitochondrial division rates while depleting Spire1C has the opposite 371 effect, causing mitochondria to become highly elongated. Because the increased 372 fission seen in cells overexpressing Spire1C depends not only on Spire1C’s 373 actin-nucleating WH2 domain but also its formin-binding KIND domain, we 374 reasoned that Spire1C-mediated mitochondrial fission also depended on formin 375 proteins. Given INF2’s previously established role as a formin protein involved in 376 ER-mediated mitochondrial constriction and division5-7, we hypothesized that 377 Spire1C could be working together with INF2. We postulated that Spire1C could 378 be promoting mitochondrial division by interacting with ER-anchored INF2, in 379 order to enable mitochondria to come into close proximity with the ER so that 380 actin-nucleation by Spire1C could enhance actin assembly mediated by INF2. 381 Supporting this possibility, we found in GST pulldown and fluorescence 382 anisotropy assays that the Spire1C KIND domain directly binds to INF2. In cells 383 overexpressing Spire1C lacking its KIND domain-mediated INF2-binding activity 384 (Spire1CΔKIND) or its actin-nucleating activity (Spire1C mWH2), ER- 385 mitochondria associations leading to mitochondrial constrictions were 386 significantly decreased. Moreover, overexpressing Spire1CΔKIND prevented 387 mitochondria from dividing in cells expressing a constitutively active INF2 mutant 388 (INF2 A149) that normally induces dramatic mitochondrial fission5,7. Finally, 389 Spire1C overexpression in cells lacking INF2 failed to induce mitochondrial 390 fission. 391 392 All these observations suggest a model in which mitochondrial Spire1C and ER- 393 anchored INF2 conspire to mediate mitochondrial constriction via actin filament 394 assembly. In this scheme, Spire1C:actin complexes on mitochondria associate 395 with INF2 on the ER, acting together with other ER-mitochondria tethering 396 complexes48 to draw the two organelles together. This results in the ER wrapping 397 around the mitochondria. Once this occurs, actin filaments nucleated by Spire1C 398 are elongated by the actin polymerization activity of INF2, similar to the “rocket 399 launcher” mechanism shown for other actin nucleating and formin proteins49. 400 Because INF2 can both polymerize and sever actin filaments, a complex 401 meshwork of actin grows between the ER and mitochondria, which may be 402 further impacted by myosin-II dimer recruitment5,6 as well as perhaps other actin- 403 regulatory proteins such as cofilin or Arp2/34,50,51. The growing actin meshwork 404 between the ER and mitochondria then exerts pressure on the mitochondrial 405 outer membrane causing its constriction. Our computational modeling of this 406 process confirmed that polymerizing actin filaments from this meshwork has 407 sufficient force to bend and constrict the mitochondrial membrane once the

9 408 filaments abut the mitochondria surface. 409 410 Clearly, further work is needed to clarify the mechanism by which Spire1C and 411 INF2 facilitate mitochondrial division. First, a better understanding of how 412 Spire1C and INF2 interact with and regulate one another’s activities during 413 mitochondrial constriction is required. The relatively low affinity between 414 Spire1C’s KIND domain and INF2 detected in our assays, if applicable to living 415 cells, is consistent with Spire1C:INF2 dissociation once INF2 begins to elongate 416 actin filaments. Second, as several proteins are known to tether ER- 417 mitochondrial membranes48, the role of these proteins in promoting or disrupting 418 Spire1C’s interaction with INF2 also needs to be studied. Such interactions may 419 underlie differences in the mode of mitochondrial division seen in cells 420 undergoing apoptosis, mitophagy, mitosis, or in response to toxins such as LLO 421 from Listeria52-55. Finally, the precise organization of the actin meshwork 422 responsible for constricting mitochondria needs to be characterized at higher 423 resolution. This will help determine whether the actin meshwork constricts 424 mitochondria by myosin-mediated contraction, by elongating filaments pushing, 425 or by a combination of both. 426 427 While we have focused on Spire1C’s role in mitochondrial constriction, the 428 establishment of Spire1C as a mitochondrial outer membrane protein suggests 429 that Spire1C is optimally positioned to serve as a molecular hub that links 430 mitochondrial dynamics to the actin cytoskeleton as well as to the ER. While our 431 appreciation of the role of actin in mitochondrial division is rapidly growing4-8, 432 there are other important functions for actin in mitochondrial dynamics, such as 433 mitochondrial motility in neurons56,57, mitochondrial partitioning prior to cell 434 division in fibroblasts58,59, and perhaps also the partitioning of mitochondrial 435 DNA60-62. This list is almost certainly not exhaustive; there may yet be other 436 known and unknown roles for the actin cytoskeleton in mitochondrial biology, and 437 vice versa. Interestingly, Spire1C directly interacts with the tail domain of myosin 438 Va (data not shown), an actin-binding motor protein that has been shown to be 439 involved in both mitochondrial and ER movement in neurons63. In other cellular 440 systems myosin Vb, Rab11a, and Spire proteins cooperate to drive actin-based 441 vesicle movements and dynamics20,27 – perhaps similar mechanisms exist for 442 mitochondrial movements. Along these lines, it is interesting to note that Rab11a 443 has also been implicated in mitochondrial dynamics64 – exploring these findings 444 in the context of Spire1C function may provide new insight towards mitochondrial 445 movements and dynamics, and perhaps the relationship between actin- 446 dependent motility and actin-dependent fission. Finally, the recent discovery of a 447 role for the ER in mediating endosomal constriction and division raises the 448 possibility that endosomal isoforms of Spire51,65 are playing a similar role in 449 promoting ER/actin/INF2-mediated endosomal fission. In fact, results from 450 previous studies suggest that overexpression of the endosomal Spire2 protein 451 lacking its KIND domain may result in endosome elongation30, which would be 452 analogous to what we have observed for Spire1CΔKIND overexpression and 453 mitochondria.

10 454 455 In conclusion, our identification and characterization of Spire1C as an ER- and 456 actin-binding mitochondrial outer membrane protein opens the door for novel 457 avenues towards understanding the regulation of myriad roles of actin, 458 mitochondria, and the ER in cellular function and disease66. 459 460 Materials and Methods 461 Cell culture and transfections 462 U2OS and Cos-7 cells were purchased from ATCC. U2OS cells stably 463 expressing GFP-INF2 was described in Chhabra et al. 200917. All cells were 464 grown in DMEM (Invitrogen) with 10% fetal bovine serum. For imaging, 465 fibronectin coated coverslips ranging between 168 and 172 μm (for fixed cell 466 imaging) or #1.5 LabTek chambers (for live cell imaging) were incubated with 10 467 μg/mL of fibronectin in PBS at 37ºC for 30 min prior to plating the cells. Transient 468 transfections were performed using FuGene 6 (Promega) according to the 469 manufacturer’s recommendations. For overexpression experiments, 1 μg of DNA 470 per coverslip was used. For minimal perturbation while imaging ER and 471 mitochondria, 50 ng of DNA was used as described in Friedman et al. 201110. 472 For siRNA transfections, cells were treated as in Korobova et al., 20137. Briefly, 473 U2OS cells stably expressing GFP-INF217 were plated on 6-well plates, treated 474 with 63 pg of siRNA per well, and analyzed 72 hours post-transfection. siRNA or 475 shRNA-mediated knockdown was confirmed by loss of GFP-INF2 or GFP- 476 Spire1C fluorescence. 477 478 Plasmids and siRNA oligonucleotides 479 Ii33-mCherry was a generous gift from P. Satpute (National Institutes of Health, 480 Bethesda, MD). MitoEmerald and mitoRFP were gifts from A. Rambold (National 481 Institutes of Health, Bethesda, MD). Spire1C was amplified from mouse brain 482 cDNA using the sequence of NM_194355 as a reference. We expected to obtain 483 a sequence yielding protein corresponding to GI 37595748, however, it contained 484 an additional 58 residues (ExonC). Thorough examination of all available 485 mammalian Spire1 isoforms revealed at least 3 alternatively-spliced exons, which 486 we refer to as exons A (majority of KIND domain), B (protein sequence 487 AVRPLSMSHSFDLS), and C (protein sequence 488 VPRITGVWPRTPFRPLFSTIQTASLLSSHPFEAAMFGVAGAMYYLFERAFTSRW 489 KPSK). 490 491 To obtain a full-length Spire1C construct, we amplified the mouse Spire1 gene 492 AK129296, which contains the full KIND domain through the first 3 WH2 493 domains, along with NM_194355, which contains a partial KIND domain and 494 ExonC without Exon B. A series of amplifications of partial gene sequences was 495 then performed to obtain versions of mouse Spire1 that were +/- each of exons 496 A, B, and C (Figure 1-figure supplement 1). Each variant of the Spire1 gene was 497 cloned into the AscI and PacI sites of modified pCS2+ vectors containing epitope 498 tag sequences adjacent to the multiple cloning site (MCS), creating Spire1C 499 constructs with either N-terminal 6x-myc or fluorescent protein tags. For the FPP

11 500 assay and knockdown experiments, human Spire1C ORF XM_005258122 501 (acquired from Genscript USA Inc.) was cloned into the pEGFP-C1/pmApple-C1 502 and pEGFP-N1/pmApple-N1 vectors (Clontech) using the XhoI-BamHI and NheI- 503 AgeI restriction sites, respectively. A nucleation-deficient mutant of Spire1C 504 (Spire1C mWH2) was generated by utilizing internal PstI and AfeI restriction sites 505 in the Spire1C gene. A sequence of 822 nucleotides of the Spire1C gene 506 between internal PstI and AfeI sites was synthesized (Genscript USA Inc.) that 507 contained alanines in place of the key hydrophobic residues required for 508 nucleation in all four WH2 domains38,39,67. Insertion of the alanine-mutated WH2 509 domains was confirmed with DNA sequencing. All primers used are shown below 510 along with the WH2 mutant insertion. 511 Primers for spire1 gene amplification and plasmid construction

Primer 1 GCGCGGCGCGCCATGGAACTGCATACATTTCTGACCAAAATTAAGAG Primer 2 GCGCTTAATTAATCAGATCTCGTTGATAGTCCGTTCTGAAG Primer 3 GAGCAGGCGCGCCATGGCCAATACCGTGGAGGCTG Primer 4 GGCGTTAATTAATCTAGTCTGCTCCGTCTAATTTCTTC Primer 5 GACGGCGCGCCATGGCGCAGCCCTCCAG Primer 6 GGCGTTAATTAATCTAGTCTGCTCCGTCTAATTTCTTC Primer 7 CCATGTGCTCCAGGAAGAAGCC Primer 8 CTGCCTTCCAAGCCATACTCTACTCTAC

512 All primers are 5' to3'. AscI or PacI restriction sites are underlined. 513 514 WH2 mutant insertion sequence: 515 516 CGAGGCTGCAGATGAAGGCCCGGAAGATGAAGACGGAGAG 517 AAGAGAAGCATCTCAGCCATCCGGTCCTATCAGGACGTTATGAAG 518 ATCTGTGCTGCTCACCTCCCAACTGAGTCGGAGGCACCCAATCAT 519 TATCAGGCAGTATGTCGGGCCCTGTTCGCAGAAACCATGGAACTG 520 CATACATTTCTGACCAAAATTAAGAGTGCAAAGGAGAACCTTAAG 521 AAGATTCAAGAAATGGAAAAGGGTGATGAATCTAGCACAGATCTG 522 GAGGACCTGAAAAATGCAGACTGGGCCCGGTTCTGGGTACAAGCG 523 GCGAGGGATTTGCGAAATGGGGTAAAAGCTAAGAAAGTCCAGCAG 524 CGGCAGTACAACCCTCTGCCCATTGAGTACCAACTGACCCCTTAT 525 GAGATGGCCGCGGACGACATTCGGTGCAAAAGATACACCGCGAGA 526 AAAGTAATGGTAAATGGTGACGTCCCCCCTCGGTTGAAAAAGAGT 527 GCTCATGAGGTCGCCGCTGACTTTATCAGATCAAGACCCCCTGCA 528 AATCCAGTTTCAGCCAGAAAACTGAAACCAACCCCACCACGGCCA 529 CGGAGCCTCCATGAAAGAGCAGCAGAAGAAATTAAAGCAGAAAGA 530 AAGGCTCGGCCTGTGTCACCAGAAGAAATTAGACGGAGCAGACTA 531 GCAGTGCGGCCACTTAGCATGTCTCACAGTTTTGACTTGTCAGAT 532 GTCACTACGCCAGAATCTCCAAAGAATGTTGGAGAATCATCTATG

12 533 GTGAATGGAGGCTTAACATCTCAAACAAAAGAAAATGGGCTGAGC 534 GCTGCCCAGCAGGGGTC 535 536 The entire amino acid sequence of Spire1C is below: 537 538 MAQPSSPGGEGPQLGAAGGPRDA 539 LSLEEILRLYNQPINEEQAWAVCFQCCGSLRAAAARRQPHRRVRSAAQIRVWR 540 DGAVTLAPAAAAAAEGEPPPASGQLGYSHCTETEVIESLGIIIYKALDYGLKENE 541 ERELSPPLEQLIDQMANTVEADGSKDEGYEAADEGPEDEDGEKRSISAIRSYQ 542 DVMKICAAHLPTESEAPNHYQAVCRALFAETMEL 543 HTFLTKIKSAKENLKKIQEMEKGDESSTDLEDLKNA 544 DWARFWVQVMRDLRNGVKLKKV 545 QQRQYNPLPIEYQLTP 546 YEMLMDDIRCKRYTLRKV 547 MVNGDVPPRLK 548 KSAHEVILDFIRSRPPLNPV 549 SARKLKPTPPRPRS 550 LHERILEEIKAERKLRPV 551 SPEEIRRSRL 552 AVRPLSMSHSFDLS 553 DVTTPESPKNVGESSMVNGGLTSQTKENGLSAAQQGSAQRKRLLKAPTLAELD 554 SSDSEEEKSLHKSTSSSSASPSLYEDPVLEAMCSRKKPPKFLPISSTPQPERRQ 555 PPQRRHSIEKETPTNVRQFLPPSRQSSRSL 556 VPRITGVWPRTPFRPLFSTIQTASLLSSHPFEAAMFGVAGAMYYLFERAFTSRW 557 KPSK 558 EEFCYPVEC 559 LALTVEEVMHIRQVLVKAELE 560 KYQQYKDVYTA 561 LKKGKLCFCCRTRRFSFFTWSYTCQFCKRPVCSQCC 562 KKMRLPSKPYSTLPIFSLGPSALQRGESCSRSEKPSTSHHRPLRSIARFSTKSR 563 SVDKSDEELQFPKELMEDWSTMEVCVDCKKFISEIISSSRRSLVLANKRARLKR 564 KTQSFYMSSAGPSEYCPSERTINEI 565 566 KIND domain 567 WH2 domains (mutated to alanine to make nucleation deficient) 568 Alternate exon B 569 Alternate ExonC 570 Spire box 571 mFYVE domain 572 573 Spire1 shRNA constructs were generated by cloning the sequences into 574 Clontech’s pSingle-tTS-shRNA vectors using the HindIII/XhoI restriction 575 sites. The sequences cloned into the vector to knockdown Spire1C were 576 5’- TCGAGGGATTAGACGTAGCAGATTATTCAAGAG 577 ATAATCTGCTACGTCTAATCTTTTTTACGCGTA-3’ (Spire1C 3’ UTR, 578 used for fixed cell imaging) and 5’-

13 579 TCGAGGCGAATAATCTCCTGACTAATTCAAGAGATTAGTCAGGAGATT 580 ATTCGTTTTTTACGCGTA-3’ (Spire1C ORF, used for live cell imaging). 581 Oligonucleotides for human INF2 siRNA were previously described in Korobova 582 et al., 20137. Briefly, the sequence used to knockdown INF2 was 5’- 583 ACAAAGAAACTGTGTGTGA-3’, and as a control, Silencer Negative Control #1 584 (Ambion) was used. 585 586 Amplification of alternate ExonC DNA from mouse tissue panel 587 Oligos flanking ExonC were designed to amplify the spire1C gene. A mouse 588 tissue cDNA panel (Clontech) was used as a template for amplification using 589 Primers 7 and 8. Amplified DNA containing ExonC was ~400 bp, while DNA 590 lacking this exon was ~200 bp. Multiple oligomer sets were utilized to eliminate 591 non-specific amplification while capturing as many on-target amplifications as 592 possible. PCR reactions were run on a 2% agaorse gel, and bands of the 593 appropriate size were excised from the gel and purified using a QIAquick gel 594 extraction kit (Qiagen). Purified DNA was cloned into the pCR II-TOPO vector 595 using the TOPO-TA cloning kit (Invitrogen). Sequence analysis was used to 596 confirm the sequence of the amplified and inserted DNA. 597 598 Plasmids for antibody generation 599 The Spire1C gene was amplified from mouse cDNA as described above. Vectors 600 used for protein purification include a modified avidin-6x his-MBP-TEV-3x FLAG- 601 Precision construct under the P1 promoter, as well as a modified pGEX vector for 602 N-terminal GST fusion proteins. All vector backbones were gifts of Dr. Aaron 603 Straight and are described in Figure 1-figure supplement 2. 604 605 Antibody production and affinity purification 606 ExonC and the C-terminal 50 residues of mouse Spire1 were cloned into the avi- 607 his-MBP-TEV-3xFLAG-precision vector described above or a modified pGEX 608 vector (for N-terminal GST fusion proteins) using standard techniques (Figure 1- 609 figure supplement 2). Avi-his-MBP-TEV-3xFLAG-precision constructs were 610 expressed and purified as described above with the following modifications. For 611 avi-his-MBP-TEV-3xFLAG-precision constructs, protein was purified over Ni-NTA 612 resin, and the eluate was further purified on an S-200 gel filtration column, 613 followed by a HiTrap Q column to remove any contaminating DNA. Protein 614 samples were sent to Cocalico Biologicals and injected into rabbits to produce 615 antisera. 616 617 GST fusion proteins were used for affinity column construction for affinity 618 purification of antibodies. These proteins were purified by using single colonies of 619 transformed Rosetta (DE3) cells to inoculate 400 mL Terrific Broth (TB; 620 Invitrogen) cultures containing 100 μg/mL carbenicillin, 34 μg/mL 621 chloramphenicol, which was grown overnight at 37°C. This culture was diluted 622 into 2L of fresh TB with antibiotics and grown to an O.D. of 0.8-0.9, at which time 623 it was moved to 23°C. After 1 hour at 23°C, cultures were induced with 0.5 mM 624 isopropyl β-D-1-thiogalactopyranoside (IPTG) for 3-4 hours and harvested as

14 625 described for purification of avi-his-MBP-TEV-3xFLAG-Precision proteins above. 626 Cells were thawed in lysis buffer (50 mM Tris, 1 M NaCl, 1 mM EDTA, 1 mM 627 DTT, and protease inhibitor cocktail, pH 7.8), sonicated, and lysates were 628 centrifuged for 30 minutes at 125,000 xg 4°C. Supernatant was applied to 629 hydrated glutathione resin (2 mL bed volume per liter culture), protein bound for 1 630 hour at 4°C, and resin was washed extensively with lysis buffer. Protein was 631 eluted with elution buffer (50 mM Tris, 150 mM NaCl, 1 mM EDTA, 20 mM 632 reduced glutathione, 1 mM DTT, and protease inhibitor cocktail, pH 7.8) and 633 loaded onto a HiTrapQ anion exchange column. A salt gradient of 150 mM to 1M 634 was used for protein elution. 635 636 Affinity column construction 637 Spire1 affinity columns were made using GST-fusion proteins following the 638 method of Finan, Hartman, and Spudich, 201168. Proteins were coupled to Affi- 639 Gel 10 by washing with 5 resin/column volumes (CV) of 0.2 M glycine, pH 2 and 640 quickly equilibrating with PBS. Antisera was filtered through a 0.2 μm filter and 641 flowed over the column continuously overnight. The column was washed with 50 642 CV wash buffer (PBS with 500 mM NaCl and 0.1% Tween 20), followed by 2.5 643 CV 0.2x PBS. Antibody was eluted with 1 CV 0.2 M glycine, pH 2 directly into 1M 644 Tris, pH 8.5 to neutralize the solution. Concentration of elution fractions was 645 checked on a Nanodrop spectrophotometer using the IgG setting. The column 646 was washed with 20 CV PBS, the antisera was re-filtered, and the purification 647 process was repeated to isolate additional antibody. Fractions with an O.D. > 0.2 648 were pooled, dialyzed into PBS containing 50% glycerol, and stored at -20°C. 649 Antisera to ExonC yielded no IgG after affinity purification. Instead, whole 650 antisera were used for subsequent experiments to probe ExonC function and 651 localization. 652 653 Antibody characterization 654 Two rabbit polyclonal antibodies described above and three commercially 655 available antibodies were used for examining expression patterns of Spire1 656 protein in various cell types. The antibodies discussed are: 1) Rabbit polyclonal 657 anti-Spire1 C-term (affinity-purified), 2) Rabbit polyclonal anti-Spire1 ExonC 658 (whole antisera), 3) Sigma mouse monoclonal anti-Spire1, 4) Abcam mouse 659 monoclonal anti Spire1, and 5) Abnova rabbit antisera to Spire1. Notably, all of 660 the commercially-available antibodies were targeted to residues 482-584 of the 661 Spire1 isoform lacking ExonC (NP_064533), and thus could only detect non- 662 ExonC containing isoforms. 663 Western blots were performed with 5-50 μg cell lysate and antibodies/antisera 664 was tested at various concentrations, temperatures, and lengths of time for best 665 conditions. Optimized conditions for all antibodies used in this work are described 666 below. HRP-conjugated goat anti-rabbit secondary antibody was used at 667 1:20,000 in all cases. 668 IB IF Source Species Type Target region Dilution Time Temperature Dilution Sigma mouse monoclonal 482-584 of NP_064533; 1: 2,000 1 h 23°C n/a

15 "last 100 a.a.”; flanks (but O/N 4°C n/a does not contain) z` Abcam mouse monoclonal 1:1,000 O/N 4°C n/a Abnova rabbit antisera 1:1,000 O/N 4°C n/a

This polyclonal, study rabbit affinity 1:2,000 O/N 4°C 1 to 100 purified last 51 residues 1:3,000 1 h 23°C This study rabbit antisera ExonC 1:1,000 O/N 4°C 1 to 500 669 670 GST Pulldown Assay 671 Spire-KIND (amino acids 1-234) was expressed as a GST fusion protein in 672 bacteria, and purified on glutathione-sepharose (GE Biosciences) followed by 673 Superdex200 gel filtration (GE Biosciences) of the glutathione-eluted GST-fusion 674 protein. GST-KIND or GST alone was re-bound to glutathione-sepharose in 675 binding buffer (50 mM KCl, 1 mM MgCl2, 1 mM EGTA, 10 mM Hepes-HCl pH 7.4, 676 1 mM DTT, 0.02% thesit (Sigma), 10 μg/mL aprotinin, 2 μg/mL leupeptin, 0.5 mM 677 benzamidine). INF2-CT (amino acids 469-1249, containing FH1, FH2 and C- 678 terminal regions) and INF2-NT (amino acids 1-420, containing DID and 679 dimerization region) were purified as described44. Proteins were mixed at 20 μM 680 GST protein, 1 μM INF2-CT and 10 μM INF2-NT in binding buffer and incubated 681 overnight, then quickly washed once in binding buffer. Proteins in glutathione 682 sepharose-bound pellet were resolved by SDS-PAGE. 683 684 Anisotropy Binding Assay 685 INF2 C-term (human CAAX variant, amino acids 941-1249) was expressed in 686 bacteria, purified and labeled on its N-terminal amine with tetramethylrhodamine 687 succinimide as described43. Labeled INF2-C-term (20 nM) was mixed with 688 varying concentrations of Spire-KIND or BSA in 10 mM Hepes pH 7.4, 50 mM 689 KCl, 1 mM MgCl2, 1 mM EGTA, 1 mM DTT, 0.5 mM Thesit detergent 690 (nonaethylene glycol monododecyl ether) at 23˚C for 1 hr before reading 691 fluorescence anisotropy in an M-1000 fluorescence plate reader (Tecan Inc) at 692 530 nm excitation and 585 nm emission. 693 694 Immunofluorescence 695 Cells were washed in phosphate buffered saline (PBS; pH 7.4) then fixed with 696 4% paraformaldehyde (PFA) for 30 minutes. Cells were then permeabilized with 697 0.1% Triton X-100 for 30 minutes before being blocked overnight with 4% bovine 698 serum albumin (BSA) at 4ºC. The next day, cells were incubated with primary 699 antibody for 2 hours, rinsed 3 times with PBS for 10 minutes each, then 700 incubated with secondary antibodies (Invitrogen) for 1 hour, rinsed 3 times with 701 PBS for 10 minutes each, then counterstained with phalloidin (Invitrogen) for 30 702 minutes, then rinsed with PBS three times, then mounted using ProLong Gold 703 antifade reagent. 704 705 Confocal Imaging 706 Confocal images were acquired with an Apochromat 63x 1.4 NA objective lens 707 (Carl Zeiss) on a Marinas spinning disk confocal imaging system (Intelligent

16 708 Imaging Innovations) using an EM charge-coupled device camera (Evolve; 709 Photometrics), or a 100x Apo TIRF 1.49 NA objective (Nikon Instruments) on a 710 Yokogawa CSU-X1 spinning disk system using an EM charge-coupled device 711 camera (Evolve; Photometrics). Cells were imaged in HEPES-buffered growth 712 media. Confocal z-stacks were taken using 200 nm steps. Images were 713 deconvoluted using Slidebook 6. Individual 16-bit tiff image files were exported, 714 then processed using ImageJ. 715 716 Photobleaching of GFP-Spire1C in specific cellular regions. 717 GFP-Spire1C photobleaching experiments presented in Figure 2 and Figure 2- 718 figure supplement 2 were carried out on a Marianas spinning disc confocal 719 microscope equipped with a Mosaic Digital Illumination System. The laser power 720 entering the Mosaic was 9 mW. Image acquisition and photobleaching of GFP- 721 Spire1C (including region selection and 405 laser exposure control) were carried 722 out using Slidebook 6 software. 723 724 Structured Illumination Microscopy (SIM) Imaging 725 SIM imaging of fixed cells was performed using an ELYRA SIM (Carl Zeiss) with 726 an Apochromat 63x 1.4 NA oil objective lens. Five angles of the excitation grid 727 with five phases each were acquired for each channel and each z-plane, which 728 were spaced at 110 nm each. SIM processing was performed using the SIM 729 module of the Zeiss Zen software package. 16-bit grayscale tiffs were 730 subsequently exported to ImageJ for quantification and processing into rendered 731 colored images. Channels in maximum projection images were aligned in the xy- 732 plane using maximum projection images of fluorescent beads. 733 734 Image Processing and Analysis 735 All image analysis and processing was performed using ImageJ. Mitochondria 736 lengths were measured manually by first setting the scale according to pixel size, 737 drawing a line along the length of the mitochondria, then using ImageJ’s 738 “measure” function. Colocalization analysis and rendering was performed using 739 the colocalization plugin included in the MacBiophotonics ImageJ plugin bundle 740 (http://rsb.info.nih.gov/ij/plugins/mbf/index.html). When calculating Pearson’s 741 values, the mitochondria channel was used as a mask for colocalization. ER- 742 mitochondria intersection sites were visually identified as regions where ER 743 tubules could be clearly visualized crossing mitochondria – these regions were 744 always in the periphery of the cell, significantly restricting the total number of 745 intersections that could be reliably identified. Mitochondria constriction sites were 746 visually identified as regions defined by relative narrowing of mitochondria 747 diameter or reduced fluorescence. Magnifications of boxed regions were 748 generated using ImageJ. Color images of merged 16-bit tiffs exported from the 749 microscope were generated using the ImageJ “merge channels” function. 750 Statistical analysis was performed using Excel (Microsoft). P-values were 751 determined using the unpaired Student’s t-test or ANOVA, as appropriate. 752 753 Modeling

17 754 The deformed shapes of the membrane tubule representing the constriction site 755 of the mitochondrial outer membrane were determined by minimizing the energy 756 of the membrane bending upon the condition of a given pressure P acting on a 757 limited region in the middle of the tube (Figure 8B). 758 759 The value of the bending energy, FB, was determined by 760 761 = , (S1) 762 763 where κ = 8 10-20 Joule is the lipid bilayer bending modulus, J is the local total 764 curvature of the membrane surface changing along the membrane surface and 765 equal at each point to the sum of the local principal curvatures46,69. The 766 integration in (S1) is performed over the whole surface of the deformed tubule. 767 768 The boundary conditions for the energy minimization consisted in the 769 requirements that at the tubule left and right edges (i) the tubule cross-sectional 770 radius, r, remains equal to its initial (preceding the deformation) value r=R, and 771 (ii) the tubule profile remains parallel to the tubule axis. While the tubule length 772 L=680nm was required to remain constant during the deformation, the tubule 773 surface area was free to change. This means that the membrane lateral tension 774 was taken to be zero, which guaranteed that the membrane bending energy was 775 the sole contribution to the membrane elastic energy. 776 777 The energy minimization and the shape determination for each pressure value 778 were performed using Brakke’s “Surface Evolver” program70. 779 780 Acknowledgements 781 We would like to thank Margot Quinlan for purified Spire1C-KIND protein and for 782 helpful suggestions. Richard Youle, Chunxin Wang, John Hammer, and the 783 members of the Lippincott-Schwartz lab all provided helpful comments and 784 suggestions on the manuscript. 785 786 787 References 788 789 1 Nunnari, J. & Suomalainen, A. Mitochondria: in sickness and in health. Cell 790 148, 1145-1159, doi:10.1016/j.cell.2012.02.035 (2012). 791 2 Archer, S. L. Mitochondrial fission and fusion in human diseases. The New 792 England journal of medicine 370, 1074, doi:10.1056/NEJMc1316254 (2014). 793 3 DuBoff, B., Gotz, J. & Feany, M. B. Tau promotes neurodegeneration via DRP1 794 mislocalization in vivo. Neuron 75, 618-632, 795 doi:10.1016/j.neuron.2012.06.026 (2012). 796 4 Li, S. et al. Transient assembly of F-actin on the outer mitochondrial 797 membrane contributes to mitochondrial fission. J Cell Biol 208, 109-123, 798 doi:10.1083/jcb.201404050 (2015).

18 799 5 Korobova, F., Gauvin, T. J. & Higgs, H. N. A role for myosin II in mammalian 800 mitochondrial fission. Curr Biol 24, 409-414, doi:10.1016/j.cub.2013.12.032 801 (2014). 802 6 Hatch, A. L., Gurel, P. S. & Higgs, H. N. Novel roles for actin in mitochondrial 803 fission. J Cell Sci, doi:10.1242/jcs.153791 (2014). 804 7 Korobova, F., Ramabhadran, V. & Higgs, H. N. An actin-dependent step in 805 mitochondrial fission mediated by the ER-associated formin INF2. Science 806 339, 464-467, doi:10.1126/science.1228360 (2013). 807 8 De Vos, K. J., Allan, V. J., Grierson, A. J. & Sheetz, M. P. Mitochondrial function 808 and actin regulate dynamin-related protein 1-dependent mitochondrial 809 fission. Curr Biol 15, 678-683, doi:10.1016/j.cub.2005.02.064 (2005). 810 9 Murley, A. et al. ER-associated mitochondrial division links the distribution of 811 mitochondria and mitochondrial DNA in yeast. eLife 2, e00422, 812 doi:10.7554/eLife.00422 (2013). 813 10 Friedman, J. R. et al. ER tubules mark sites of mitochondrial division. Science 814 334, 358-362, doi:10.1126/science.1207385 (2011). 815 11 Lackner, L. L. & Nunnari, J. M. The molecular mechanism and cellular 816 functions of mitochondrial division. Biochim Biophys Acta 1792, 1138-1144, 817 doi:10.1016/j.bbadis.2008.11.011 (2009). 818 12 Mears, J. A. et al. Conformational changes in Dnm1 support a contractile 819 mechanism for mitochondrial fission. Nat Struct Mol Biol 18, 20-26, 820 doi:10.1038/nsmb.1949 (2011). 821 13 Yoon, Y., Pitts, K. R. & McNiven, M. A. Mammalian dynamin-like protein DLP1 822 tubulates membranes. Mol Biol Cell 12, 2894-2905 (2001). 823 14 Legesse-Miller, A., Massol, R. H. & Kirchhausen, T. Constriction and Dnm1p 824 recruitment are distinct processes in mitochondrial fission. Mol Biol Cell 14, 825 1953-1963, doi:10.1091/mbc.E02-10-0657 (2003). 826 15 Labrousse, A. M., Zappaterra, M. D., Rube, D. A. & van der Bliek, A. M. C. 827 elegans dynamin-related protein DRP-1 controls severing of the 828 mitochondrial outer membrane. Molecular cell 4, 815-826 (1999). 829 16 Ingerman, E. et al. Dnm1 forms spirals that are structurally tailored to fit 830 mitochondria. J Cell Biol 170, 1021-1027, doi:10.1083/jcb.200506078 831 (2005). 832 17 Chhabra, E. S., Ramabhadran, V., Gerber, S. A. & Higgs, H. N. INF2 is an 833 endoplasmic reticulum-associated formin protein. J Cell Sci 122, 1430-1440, 834 doi:10.1242/jcs.040691 (2009). 835 18 Quinlan, M. E. Direct interaction between two actin nucleators is required in 836 Drosophila oogenesis. Development 140, 4417-4425, 837 doi:10.1242/dev.097337 (2013). 838 19 Vizcarra, C. L. et al. Structure and function of the interacting domains of Spire 839 and Fmn-family formins. Proc Natl Acad Sci U S A 108, 11884-11889, 840 doi:10.1073/pnas.1105703108 (2011). 841 20 Schuh, M. An actin-dependent mechanism for long-range vesicle transport. 842 Nat Cell Biol 13, 1431-1436, doi:10.1038/ncb2353 (2011).

19 843 21 Pfender, S., Kuznetsov, V., Pleiser, S., Kerkhoff, E. & Schuh, M. Spire-type actin 844 nucleators cooperate with Formin-2 to drive asymmetric oocyte division. 845 Curr Biol 21, 955-960, doi:10.1016/j.cub.2011.04.029 (2011). 846 22 Pechlivanis, M., Samol, A. & Kerkhoff, E. Identification of a short Spir 847 interaction sequence at the C-terminal end of formin subgroup proteins. J 848 Biol Chem 284, 25324-25333, doi:10.1074/jbc.M109.030320 (2009). 849 23 Quinlan, M. E., Hilgert, S., Bedrossian, A., Mullins, R. D. & Kerkhoff, E. 850 Regulatory interactions between two actin nucleators, Spire and Cappuccino. 851 J Cell Biol 179, 117-128, doi:10.1083/jcb.200706196 (2007). 852 24 Bosch, M. et al. Analysis of the function of Spire in actin assembly and its 853 synergy with formin and profilin. Molecular cell 28, 555-568, 854 doi:10.1016/j.molcel.2007.09.018 (2007). 855 25 Zeth, K. et al. Molecular basis of actin nucleation factor cooperativity: crystal 856 structure of the Spir-1 kinase non-catalytic C-lobe domain (KIND)*formin-2 857 formin SPIR interaction motif (FSI) complex. J Biol Chem 286, 30732-30739, 858 doi:10.1074/jbc.M111.257782 (2011). 859 26 Kerkhoff, E. Actin dynamics at intracellular membranes: the Spir/formin 860 nucleator complex. European journal of cell biology 90, 922-925, 861 doi:10.1016/j.ejcb.2010.10.011 (2011). 862 27 Montaville, P. et al. Spire and Formin 2 synergize and antagonize in 863 regulating actin assembly in meiosis by a ping-pong mechanism. PLoS biology 864 12, e1001795, doi:10.1371/journal.pbio.1001795 (2014). 865 28 Belin, B. J., Lee, T. & Mullins, R. D. DNA damage induces nuclear actin filament 866 assembly by formin-2 and Spire-1/2 that promotes efficient DNA repair. 867 (2015). 868 29 Schuh, M. An actin-dependent mechanism for long-range vesicle transport. 869 Nat Cell Biol 13, 1431-1436, doi:10.1038/ncb2353 (2011). 870 30 Dietrich, S., Weiss, S., Pleiser, S. & Kerkhoff, E. Structural and functional 871 insights into the Spir/formin actin nucleator complex. Biol Chem 394, 1649- 872 1660, doi:10.1515/hsz-2013-0176 (2013). 873 31 Kent, W. J. et al. The human genome browser at UCSC. Genome research 12, 874 996-1006, doi:10.1101/gr.229102. Article published online before print in 875 May 2002 (2002). 876 32 Allen, J. R., Ross, S. T. & Davidson, M. W. Structured illumination microscopy 877 for superresolution. Chemphyschem : a European journal of chemical physics 878 and physical chemistry 15, 566-576, doi:10.1002/cphc.201301086 (2014). 879 33 Lorenz, H., Hailey, D. W. & Lippincott-Schwartz, J. Fluorescence protease 880 protection of GFP chimeras to reveal protein topology and subcellular 881 localization. Nat Methods 3, 205-210, doi:10.1038/nmeth857 (2006). 882 34 Muñoz-Pinedo, C. et al. Different mitochondrial intermembrane space 883 proteins are released during apoptosis in a manner that is coordinately 884 initiated but can vary in duration. Proceedings of the National Academy of 885 Sciences 103, 11573-11578 (2006). 886 35 Claros, M. G. & von Heijne, G. TopPred II: an improved software for 887 membrane protein structure predictions. Computer applications in the 888 biosciences : CABIOS 10, 685-686 (1994).

20 889 36 Kelley, L. A. & Sternberg, M. J. Protein structure prediction on the Web: a case 890 study using the Phyre server. Nature protocols 4, 363-371, 891 doi:10.1038/nprot.2009.2 (2009). 892 37 Cole, N. B. et al. Diffusional mobility of Golgi proteins in membranes of living 893 cells. Science 273, 797-801 (1996). 894 38 Quinlan, M. E., Heuser, J. E., Kerkhoff, E. & Mullins, R. D. Drosophila Spire is an 895 actin nucleation factor. Nature 433, 382-388, doi:10.1038/nature03241 896 (2005). 897 39 Loomis, P. A. et al. Targeted wild-type and jerker espins reveal a novel, WH2- 898 domain-dependent way to make actin bundles in cells. J Cell Sci 119, 1655- 899 1665, doi:10.1242/jcs.02869 (2006). 900 40 Salles, F. T. et al. Myosin IIIa boosts elongation of stereocilia by transporting 901 espin 1 to the plus ends of actin filaments. Nat Cell Biol 11, 443-450, 902 doi:10.1038/ncb1851 (2009). 903 41 Wakabayashi, J. et al. The dynamin-related GTPase Drp1 is required for 904 embryonic and brain development in mice. J Cell Biol 186, 805-816, 905 doi:10.1083/jcb.200903065 (2009). 906 42 Bleazard, W. et al. The dynamin-related GTPase Dnm1 regulates 907 mitochondrial fission in yeast. Nat Cell Biol 1, 298-304, doi:10.1038/13014 908 (1999). 909 43 Ramabhadran, V., Hatch, A. L. & Higgs, H. N. Actin monomers activate 910 inverted formin 2 by competing with its autoinhibitory interaction. J Biol 911 Chem 288, 26847-26855, doi:10.1074/jbc.M113.472415 (2013). 912 44 Ramabhadran, V., Gurel, P. S. & Higgs, H. N. Mutations to the formin homology 913 2 domain of INF2 protein have unexpected effects on actin polymerization 914 and severing. J Biol Chem 287, 34234-34245, doi:10.1074/jbc.M112.365122 915 (2012). 916 45 Chhabra, E. S. & Higgs, H. N. INF2 Is a WASP homology 2 motif-containing 917 formin that severs actin filaments and accelerates both polymerization and 918 depolymerization. J Biol Chem 281, 26754-26767, 919 doi:10.1074/jbc.M604666200 (2006). 920 46 Helfrich, W. Elastic properties of lipid bilayers: theory and possible 921 experiments. Zeitschrift fur Naturforschung. Teil C: Biochemie, Biophysik, 922 Biologie, Virologie 28, 693-703 (1973). 923 47 Footer, M. J., Kerssemakers, J. W., Theriot, J. A. & Dogterom, M. Direct 924 measurement of force generation by actin filament polymerization using an 925 optical trap. Proc Natl Acad Sci U S A 104, 2181-2186, 926 doi:10.1073/pnas.0607052104 (2007). 927 48 Rowland, A. A. & Voeltz, G. K. Endoplasmic reticulum-mitochondria contacts: 928 function of the junction. Nature reviews. Molecular cell biology 13, 607-625, 929 doi:10.1038/nrm3440 (2012). 930 49 Breitsprecher, D. et al. Rocket launcher mechanism of collaborative actin 931 assembly defined by single-molecule imaging. Science 336, 1164-1168, 932 doi:10.1126/science.1218062 (2012).

21 933 50 Derivery, E. et al. The Arp2/3 activator WASH controls the fission of 934 endosomes through a large multiprotein complex. Developmental cell 17, 935 712-723, doi:10.1016/j.devcel.2009.09.010 (2009). 936 51 Liu, R. et al. Wash functions downstream of Rho and links linear and 937 branched actin nucleation factors. Development 136, 2849-2860, 938 doi:10.1242/dev.035246 (2009). 939 52 Stavru, F., Palmer, A. E., Wang, C., Youle, R. J. & Cossart, P. Atypical 940 mitochondrial fission upon bacterial infection. Proc Natl Acad Sci U S A 110, 941 16003-16008, doi:10.1073/pnas.1315784110 (2013). 942 53 Youle, R. J. & van der Bliek, A. M. Mitochondrial fission, fusion, and stress. 943 Science 337, 1062-1065, doi:10.1126/science.1219855 (2012). 944 54 Hoppins, S. & Nunnari, J. Cell Biology. Mitochondrial dynamics and apoptosis- 945 -the ER connection. Science 337, 1052-1054, doi:10.1126/science.1224709 946 (2012). 947 55 Chan, D. C. Fusion and fission: interlinked processes critical for mitochondrial 948 health. Annual review of genetics 46, 265-287, doi:10.1146/annurev-genet- 949 110410-132529 (2012). 950 56 Hollenbeck, P. J. & Saxton, W. M. The axonal transport of mitochondria. J Cell 951 Sci 118, 5411-5419, doi:10.1242/jcs.02745 (2005). 952 57 Pathak, D., Sepp, K. J. & Hollenbeck, P. J. Evidence that myosin activity 953 opposes microtubule-based axonal transport of mitochondria. J Neurosci 30, 954 8984-8992, doi:10.1523/JNEUROSCI.1621-10.2010 (2010). 955 58 Rohn, J. L. et al. Myo19 ensures symmetric partitioning of mitochondria and 956 coupling of mitochondrial segregation to cell division. Curr Biol 24, 2598- 957 2605, doi:10.1016/j.cub.2014.09.045 (2014). 958 59 Quintero, O. A. et al. Human Myo19 is a novel myosin that associates with 959 mitochondria. Curr Biol 19, 2008-2013, doi:10.1016/j.cub.2009.10.026 960 (2009). 961 60 Reyes, A. et al. Actin and myosin contribute to mammalian mitochondrial 962 DNA maintenance. Nucleic Acids Res 39, 5098-5108, doi:10.1093/nar/gkr052 963 (2011). 964 61 Boldogh, I. R., Ramcharan, S. L., Yang, H. C. & Pon, L. A. A type V myosin 965 (Myo2p) and a Rab-like G-protein (Ypt11p) are required for retention of 966 newly inherited mitochondria in yeast cells during cell division. Mol Biol Cell 967 15, 3994-4002, doi:10.1091/mbc.E04-01-0053 (2004). 968 62 Boldogh, I. R. et al. A protein complex containing Mdm10p, Mdm12p, and 969 Mmm1p links mitochondrial membranes and DNA to the cytoskeleton-based 970 segregation machinery. Mol Biol Cell 14, 4618-4627, doi:10.1091/mbc.E03- 971 04-0225 (2003). 972 63 Wagner, W., Brenowitz, S. D. & Hammer, J. A., 3rd. Myosin-Va transports the 973 endoplasmic reticulum into the dendritic spines of Purkinje neurons. Nat Cell 974 Biol 13, 40-48, doi:10.1038/ncb2132 (2011). 975 64 Landry, M. C. et al. A functional interplay between the small GTPase Rab11a 976 and mitochondria-shaping proteins regulates mitochondrial positioning and 977 polarization of the actin cytoskeleton downstream of Src family kinases. J Biol 978 Chem 289, 2230-2249, doi:10.1074/jbc.M113.516351 (2014).

22 979 65 Kerkhoff, E. Cellular functions of the Spir actin-nucleation factors. Trends in 980 cell biology 16, 477-483, doi:10.1016/j.tcb.2006.07.005 (2006). 981 66 Rappold, P. M. et al. Drp1 inhibition attenuates neurotoxicity and dopamine 982 release deficits in vivo. Nature communications 5, 5244, 983 doi:10.1038/ncomms6244 (2014). 984 67 Quinlan, M. E., Hilgert, S., Bedrossian, A., Mullins, R. D. & Kerkhoff, E. 985 Regulatory interactions between two actin nucleators, Spire and Cappuccino. 986 J Cell Biol 179, 117-128, doi:10.1083/jcb.200706196 (2007). 987 68 Finan, D., Hartman, M. A. & Spudich, J. A. Proteomics approach to study the 988 functions of Drosophila myosin VI through identification of multiple cargo- 989 binding proteins. Proc Natl Acad Sci U S A 108, 5566-5571, 990 doi:10.1073/pnas.1101415108 (2011). 991 69 Spivak, M. A comprehensive introduction to differential geometry. 2d edn, 992 (Publish or Perish, inc., 1979). 993 70 Brakke, K. A. The surface evolver. Experimental mathematics 1, 141-165 994 (1992). 995 996

23 997 Figure Legends 998 999 Figure 1: The Spire1 alternate exon ExonC is necessary and 1000 sufficient for localization to mitochondria. (A) Full length Spire1C 1001 domain structure: The number ranges indicate the amino acid regions of 1002 the conserved domains probed in this study. (B) Spire1C localizes to 1003 mitochondria. myc-Spire1C: U2OS cells cotransfected with myc-Spire1C 1004 and mitoRFP show robust localization of myc-Spire1C to mitochondria. α- 1005 ExonC: U2OS cells stained with an antibody raised against ExonC (α- 1006 ExonC) and expressing mitoRFP show endogenous Spire1C labeling on 1007 mitochondria. GFP-ExonC: U2OS cells cotransfected with GFP-ExonC 1008 and mitoRFP show robust targeting of GFP-ExonC to mitochondria. myc- 1009 Spire1ΔC: U2OS cells cotransfected with myc-Spire1ΔC and mitoRFP 1010 show no specific targeting of myc-Spire1ΔC to mitochondria. All cells were 1011 fixed and primary antibodies were counterstained with Alexa-488 1012 secondary antibody before imaging with confocal fluorescence 1013 microscopy. Scale bars: 10 μm. Inserts are magnifications of the boxed 1014 regions.

1015 Figure 2. Spire1C localizes to the mitochondrial outer membrane with its 1016 formin and actin binding domains facing the cytoplasm. 1017 (A) ExonC localizes to the peripheral region of mitochondria. Left: SIM 1018 image of a U2OS cell transfected with GFP-ExonC and mitoRFP reveals 1019 localization of GFP-ExonC to the periphery of mitochondria. Right: 1020 Magnification of boxed region on left. The insert on the lower right corner 1021 is a fluorescence intensity linescan of the rectangular boxed region 1022 indicating the inversely related profiles of GFP-ExonC vs. mitoRFP. Scale 1023 bar: 10 μm. (B) Illustration of the principle of the fluorescence protease 1024 protection (FPP) assay performed on mitochondrial outer membrane 1025 proteins. If a fluorescent protein tag faces the cytoplasm (green circle), it is 1026 degraded by trypsin and its fluorescence is depleted. If the protein tag 1027 faces the interior of the mitochondria (blue triangle), it is protected from 1028 trypsin in the cytoplasm, and thus its fluorescence remains after trypsin 1029 addition. An intermembrane space (IMS) marker, OMI-mCherry, serves as 1030 a control to verify that the mitochondrial outer membrane has not been 1031 permeabilized by the digitonin treatment, and furthermore to confirm that 1032 trypsin is not degrading proteins in the intermembrane space. (C) 1033 Schematic of the constructs used in our FPP assays. (D) Cells 1034 cotransfected with OMI-mCherry and GFP-Spire1C (N-terminal GFP tag, 1035 left) or GFP-ExonC (N-terminal GFP-tag, middle) were treated with 20μM 1036 digitonin and 4mM trypsin. In both cases OMI-mCherry fluorescence 1037 remained, whereas the N-terminal GFP tags were mostly depleted within 1038 60 seconds after trypsin treatment. In contrast, in cells transfected with 1039 ExonC-GFP (C-terminus GFP tag), GFP fluorescence remains unchanged 1040 after 60 seconds of trypsin treatment. Scale bar: 10 μm. (E) GFP-Spire1C 1041 laterally diffuses along the mitochondrial outer membrane. A small region

24 1042 of a mitochondrion labeled with GFP-Spire1C was photobleached (white 1043 boxed region). The rapid, directional recovery from the unbleached region 1044 into the bleached region (see also Figure 2-figure supplement 2) suggests 1045 GFP-Spire1C is stably associated with and laterally diffuses along the 1046 mitochondrial outer membrane. (F) GFP-Spire1C does not readily 1047 exchange with the cytoplasm or neighboring mitochondria since 1048 photobleaching of an entire mitochondrion resulted in very low 1049 fluorescence recovery over the same period of time as in (E). Scale bar: 1 1050 μm.

1051 Figure 3: Spire1C promotes actin assembly on mitochondrial 1052 surfaces Overexpression of Spire1C causes actin accumulation on 1053 mitochondria. Control: SIM image of a Cos7 cell expressing mitoEmerald 1054 and stained with phalloidin-568 to visualize actin shows low amounts of 1055 overlap between mitochondria and actin (Mander’s: 0.43±0.020, ncells=26). 1056 Spire1C overexpression: A Cos7 cell expressing mitoEmerald and 1057 overexpressing myc-Spire1C and stained with phalloidin-568 reveals 1058 significantly increased actin accumulation on mitochondria (Mander’s: 1059 0.64±0.066, ncells=19, p<0.05) compared to control cells. GFP- 1060 Spire1CΔKIND: A Cos7 cell overexpressing the formin-binding deficient 1061 GFP-Spire1CΔKIND stained with phalloidin-568 reveals significant 1062 accumulation of actin on mitochondria compared to control cells 1063 (Mander’s: 0.55±0.017, ncells=18, p<0.05). A Cos7 cell overexpressing 1064 GFP-Spire1C mWH2 displays no increased accumulation of actin 1065 (Mander’s: 0.41±0.040, ncells=15, p=0.32) compared to control cells. Scale 1066 bar: 5 μm.

1067 Figure 4: Spire1C promotes mitochondrial fission via its formin- 1068 binding KIND and actin-nucleating WH2 domains. (A) U2OS cells 1069 overexpressing Spire1C display shorter mitochondria (2nd panel, 2.2±0.5 1070 μm, nmitochondria=211, ncells=14, p<0.0001), whereas cells overexpressing rd 1071 Spire1C mWH2 (3 panel, 6.2±1.52 μm, nmitochondria=332, th 1072 ncells=15,p<0.0001) or Spire1CΔKIND (4 panel, 9.0±1.50 μm, 1073 nmitochondria=232, ncells=16, p<0.0001) display longer, more tubulated 1074 mitochondria compared to control cells (1st panel, 3.57 ± 0.45 μm, 1075 nmitochondria=322, ncells=34). Scale bar: 10 μm. (B) Cells transfected with 1076 mitoEmerald (and neighboring non-transfected cells) stained with α- 1077 ExonC (upper row) showed robust colocalization of mitoEmerald and α- 1078 ExonC, with a mixture of tubulated and fragmented mitochondria. Cells 1079 cotransfected with Spire1 shRNA and mitoEmerald with no detectable α- 1080 ExonC labeling (lower row) display long, tubulated mitochondria (6.1±1.26 1081 μm, nmitochondria=222, ncells=17). All primary antibodies were counterstained 1082 with Alexa-568 secondary antibody. Scale bar: 15 μm. (C) Left: Average 1083 mitochondrial lengths for control cells and cells overexpressing Spire1C, 1084 Spire1C mWH2, Spire1 shRNA or Spire1CΔKIND. Right: Average number 1085 of mitochondrial fission events in one cell in a timespan of 10 minutes for

25 1086 control (ncells=17), Spire1C overexpressing (ncells=10, p<0.0001), Spire1C 1087 mWH2 overexpressing (ncells=12, p<0.0001), Spire1 knockdown (ncells=25, 1088 p<0.0001) and Spire1CΔKIND overexpressing (ncells=22, p<0.0001) cells. 1089 At least 3 separate experiments were performed for all conditions. Error 1090 bars represent standard error of the mean.

1091 Figure 5: Spire1CΔKIND overexpression reduces the amount of ER- 1092 mitochondria overlap. (A) Confocal images of cells expressing Ii33- 1093 mCherry and overexpressing GFP-Spire1C (2nd row) or GFP-Spire1C 1094 mWH2 (3rd row), but not GFP-Spire1CΔKIND (4th row), display significant 1095 overlap of mitochondria with ER, similar to control cells expressing 1096 mitoEmerald. The images on the right-hand side show a magnified view of 1097 the boxed region in the merge image, with overlapping pixels in displayed 1098 in white. Scale bar: 15 μm. (B) Bar graph representing the average 1099 number of ER-mitochondria intersections per cell. We were able to resolve 1100 an average of 14.3±3.5 intersections in control cells (ncells=12). GFP- 1101 Spire1C overexpressing cells had 14.7±1.93 (ncells=15) ER-mitochondria 1102 intersections per cell. GFP-Spire1C mWH2 expressing cells had an 1103 average of 14.7±1.62 (ncells=11) ER-mitochondria intersections per cell. 1104 Spire1CΔKIND expressing cells had 6.8±1.33 (ncells=14, p<0.05) ER- 1105 mitochondria intersections per cell. 1106 1107 Figure 6: Spire1C and INF2 directly interact and work together to 1108 regulate mitochondrial fission. (A) INF2-CT directly binds to Spire1- 1109 KIND in vitro. GST pull-down assays in actin polymerization buffer, 1110 containing combinations of the following: 20 μM GST or GST-Spire1 KIND 1111 bound to glutathione-sepharose beads; 1 μM INF2-CT; and 10 μM INF2- 1112 NT. Co-incubation of the GST-Spire1 KIND domain pulls down INF2-CT 1113 (3rd to last lane), but not INF2-NT (2nd to last lane). INF2-CT pulldown is 1114 inhibited by the addition of the INF2-NT (last lane). STDS lane represents 1115 0.2 μM INF2-CT. (B) Fluorescence anisotropy binding curve of purified 1116 INF2-CT (20 nM) labeled with tetramethylrhodamine succinimide mixed 1117 with varying concentrations of Spire1 KIND or BSA reveals a direct 1118 interaction between Spire1 KIND and INF2-CT. (C) Cells overexpressing a 1119 constitutively active INF2 mutant (INF2 A149 alone, ncells=16, 1120 nmitochondria=232) display very short, fragmented mitochondria compared to 1121 control cells (p<0.0001). Cells overexpressing A149 and Spire1C 1122 (+Spire1C, ncells=14, nmitochondria=461) or Spire1C mWH2 (+Spire1C mWH2, 1123 ncells=20, nmitochondria=377) similarly display very short mitochondria. In 1124 contrast, cells overexpressing A149 and Spire1CΔKIND display longer, 1125 more tubulated mitochondria (+Spire1CΔKIND, ncells=18, nmitochondria=379, 1126 p<0.0001). (D) Cells overexpressing Spire1C and treated with scrambled 1127 siRNA display shorter, more fragmented mitochondria (+scramble siRNA, 1128 ncells=20, nmitochondria=434, p<0.05). In contrast, cells overexpressing 1129 Spire1C and treated with INF2 siRNA display significantly longer, more 1130 tubulated mitochondria (Spire1C+INF2 siRNA, ncells=22, nmitochondria=627,

26 1131 p<0.0001). Scale bars: 5 μm. (E) Bar graph displaying average 1132 mitochondria lengths for each of the conditions in this figure. Error bars 1133 represent standard error of the mean. 1134 1135 Figure 7: Spire1C overexpression enhances mitochondrial 1136 constriction via its WH2 and KIND domains in cooperation with INF2. 1137 (A) Representative confocal images of U2OS cells expressing Ii33- 1138 mCherry in order to visualize ER tubules crossing over mitochondria in 1139 cells expressing mitoEmerald (1st row) or overexpressing GFP-Spire1C 1140 (2nd row), GFP-Spire1C mWH2 (3rd row), GFP-Spire1CΔKIND (4th row), or 1141 GFP-Spire1C while treated with INF2 siRNA (5th row). Arrows indicate ER- 1142 mitochondria intersection points associated with mitochondrial 1143 constriction. Arrowheads indicate ER-mitochondria intersections not 1144 resulting in mitochondrial constriction. Scale bar: 1 μm. (B) Bar graphs 1145 representing the average percentage of ER-mitochondria intersections 1146 associated with mitochondrial constriction for each construct used. In cells 1147 expressing mitoEmerald, 55.2±5.5% (nintersections=172, ncells=12) of ER- 1148 mitochondria intersections appeared to result in mitochondrial constriction. 1149 In GFP-Spire1C overexpressing cells, 71.5±4.5% (nintersections=221, 1150 ncells=15, p<0.05) of ER-mitochondria intersections resulted in 1151 mitochondrial constriction. In GFP-Spire1C mWH2 overexpressing cells, 1152 24.1±2.4% (nintersections=162, ncells=11, p<0.01) of ER-mitochondria 1153 intersections appeared to result in mitochondrial constriction. In Spire1C 1154 knockdown cells, 37.2±5.7% (nintersections=123, ncells=11, p<0.001) of ER- 1155 mitochondria intersections appeared to result in mitochondrial constriction. 1156 In GFP-Spire1CΔKIND overexpressing cells, 29.5±4.7% (nintersections=95, 1157 ncells=14, p<0.01) of ER-mitochondria intersections appeared to result in 1158 mitochondrial constriction. In GFP-Spire1C overexpressing cells treated 1159 with INF2 siRNA, 27.5±5.3% (nintersections=178, ncells=16, p<0.01) of ER- 1160 mitochondria intersections appeared to result in mitochondrial constriction.

1161 Figure 8: Putative model for how mitochondrial Spire1C and ER-anchored 1162 INF2 could mediate mitochondrial constriction via actin filament assembly. 1163 (A) Spire1C:actin complexes on mitochondria associate with INF2 on the ER. 1164 Actin filaments nucleated by Spire1C are elongated by the actin polymerization 1165 activity of INF2. The actin filament elongation activity exerts pressure on the 1166 mitochondrial outer membrane, thereby driving constriction of the latter. 1167 Tethering complexes may play a role in maintaining association between ER and 1168 mitochondrial membranes. Myosin-II dimers and the related contractile actin ring, 1169 which may also be involved in mitochondrial constriction, are not shown for 1170 simplicity. (B) Computational results showing mitochondrial shapes resulting from 1171 deformation by constricting pressure P developed by the actin polymerization 1172 and/or actin contractile based mechanisms (see also Figure 8-figure supplement 1173 1, Figure 8-source data 1, and Materials and Methods for more information). The 1174 mitochondrial constriction site was modeled as a tubular membrane of about 1175 680nm length and with initial radius R = 230nm. The dark blue strip in the middle

27 1176 represents the 50nm wide zone of the pressure application. The images 1177 correspond to three degrees of the mitochondria constriction characterized by 1178 cross-sectional radii r in the narrowest place of 145nm, 110nm and 65nm. The 1179 corresponding values of the pressure P, the required numbers of the 1180 polymerizing actin filaments, Nf, and the required tensions in the actin contractile 1181 ring, γm, are presented in Figure 8-figure supplement 1 and Figure 8-source data 1182 1).

1183

1184 Supplementary Data Legends

1185 Figure 1-figure supplement 1: Construction of the complete Spire1C protein 1186 sequence as explained in detail in the Materials and Methods. Spire1C gene 1187 amplified from mouse brain cDNA. Initial construct (top) based on NM_194355 1188 was later combined with the sequence based on AK129296 to form full-length 1189 Spire1C protein of 802 amino acids with all known exons. Spire1C diagram 1190 (bottom) showing both characterized (red, blue, yellow, and orange) and 1191 uncharacterized (purple, green, and black) domains (indicated in the legend on 1192 the bottom).

1193 Figure 1-figure supplement 2: Constructs used to probe Spire1 function. 1194 (A) Spire1C constructs used in this study. Residue numbers are for the 802 a.a. 1195 full-length mouse Spire1C protein. (B) Coomassie-stained gels of purified 1196 proteins used for antibody production (pET constructs) and affinity purification 1197 (GST constructs).

1198 Figure 1-figure supplement 3: Spire1C contains a previously 1199 uncharacterized alternate exon of 58 amino acids. (A) Spire1 ExonC was 1200 found to be present in the tissues listed. *PCR from cDNA did not yield 1201 conclusive results, but the ExonC-containing gene was originally amplified from 1202 mouse brain cDNA. (B) Sequencing results from TOPO pCR2.1 vector using the 1203 M13 reverse primer. ExonC sequence is shown in red. (C) Translated protein 1204 sequence from red portion in C. (D) Representative immunoblots comparing 1205 commercial antibodies to Spire1 used to detect endogenous Spire1 in HeLa and 1206 PC12 cell lysates, 25 ug per lane. (E) Affinity purified Spire1 C-terminal antibody 1207 and ExonC antisera (both generated in this study) tested on HeLa and PC12 cell 1208 lysates, as in (A). Strikingly different protein bands are observed for the same cell 1209 type with different antibodies and between cell types with the same antibody. 1210 Molecular weight protein markers are the same size for each blot, though the 1211 positions vary slightly (as indicated). (F) The Phyre server was used to predict 1212 any secondary structural elements in ExonC. Phyre compiles structure 1213 predictions from three different algorithms (psipred, jnet, and sspro) to form a 1214 consensus prediction and probability for each structural element. Probability 1215 ranges from 1-10, with 10 being the most probable. Yellow amino acids are 1216 small/polar, green are hydrophobic, red are charged, and purple are 1217 aromatic+cysteine.

28 1218 Figure 2-figure supplement 1: GFP-Spire1C laterally diffuses on the 1219 mitochondrial outer membrane. (A) After partial bleaching of mitochondria 1220 labeled with GFP-Spire1C, rapid fluorescence recovery occurs in a directional 1221 fashion. The fluorescence intensity in the region closest to the unbleached region 1222 (green rectangle) increases more rapidly than the fluorescence intensity in the 1223 region farther away (red rectangle) from the unbleached region. Scale bar: 2 μm. 1224 (B) Graph of the fluorescence intensity of the green and red regions in (A) shows 1225 the differential rate of recovery over time for each region.

1226 Figure 3-figure supplement 1: Spire1C promotes actin assembly near 1227 mitochondria in a WH2-dependent fashion. (A) Non-transfected (α-ExonC): 1228 SIM image of a U2OS cell stained with phalloidin-488 and α-Spire1C to visualize 1229 endogenous Spire1C localization relative to actin shows very low amounts of 1230 overlap between mitochondria and actin. myc-Spire1C (α-myc): A U2OS cell 1231 overexpressing myc-Spire1C, stained with anti-myc and phalloidin-488 reveals 1232 increased actin accumulation on mitochondria compared to non-transfected cells. 1233 myc-Spire1C mWH2 (α-myc): A U2OS cell overexpressing the actin-nucleation 1234 deficient myc-Spire1C mWH2 labeled with anti-myc and phalloidin-488 reveals 1235 muted accumulation of actin on mitochondria compared to wild-type myc-Spire1C 1236 overexpression. Scale bars: 15 μm. The yellow-boxed regions are magnified 1237 underneath each image. All cells were counterstained with Alexa-568 secondary 1238 antibody. 1239 1240 Figure 4-figure supplement 1: Distribution of mitochondrial lengths 1241 measured in each condition. Histogram showing distribution of frequencies of 1242 mitochondrial lengths for each of the conditions shown in the bar graphs in 1243 Figure 4. 1244 1245 Figure 8-figure supplement 1: Computational results of simulations of 1246 mitochondrial constriction mediated by actin polymerization and actin constriction 1247 mechanisms. (A) The cross-sectional radius in the narrowest place of the 1248 mitochondria shape, r, as a function of the pressure, P, exerted on the limited 1249 region in the middle of the constriction region (see Figure 8 of the main text). The 1250 radius, r, and the pressure, P, are presented in the universal dimensionless 1251 forms, r/R, and PR3/2κ, where R is the initial (preceding the deformation) 1252 mitochondrial radius and κ is the membrane bending modulus. The dashed lines 1253 indicate the specific deformations presented in Figure 8 of the main text. (B) The 1254 number of the actin filaments, Nf, and the tension in the actin contractile ring, γm, 1255 providing the pressure as functions of the resulting mitochondria deformation. 1256 The deformation is quantified by r/R (see (A) for definition). 1257 1258 Figure 8-source data 1: Specific values of the system parameters and the 1259 computational results for the three specific extents of mitochondrial constriction 1260 presented in Figure 8, Figure 8-figure supplement 1, and discussed in the main 1261 text. 1262

29 1263 Video 1: A U2OS cell coexpressing GFP-Spire1C (N-terminus tag) and OMI- 1264 mCherry displays rapid loss of GFP fluorescence signal after the addition of 10 1265 μM digitonin and 4 mM trypsin, yet persistent mCherry fluorescence, indicating 1266 that trypsin is degrading the GFP tag on the N-terminus of Spire1C in the 1267 cytoplasm, but not OMI-mCherry, which resides in the mitochondria 1268 intermembrane space. Scale bar: 10 μm.

1269 Video 2: A U2OS cell coexpressing GFP-ExonC (N-terminus tag) and OMI- 1270 mCherry displays rapid loss of GFP fluorescence signal after the addition of 10 1271 μM digitonin and 4 mM trypsin, yet persistent mCherry fluorescence, indicating 1272 that trypsin is degrading the GFP tag on the N-terminus of Spire1C in the 1273 cytoplasm, but not OMI-mCherry, which resides in the mitochondria 1274 intermembrane space. Scale bar: 10 μm.

1275 Video 3: A U2OS cell coexpressing ExonC-GFP (C-terminus tag) and OMI- 1276 mCherry displays no loss of GFP fluorescence signal after the addition of 10 μM 1277 digitonin and 4 mM trypsin, indicating the GFP tag on the C-terminus of ExonC is 1278 protected within the mitochondrial lumen. Scale bar: 10 μm.

1279 1280

1281

1282 1283 1284 1285

1286

1287

1288

30