All the known dynamin family proteins perform cellular functions by associating with a target membrane. Classical dynamins contain a pleckstrin homology (PH) domain responsible for membrane binding (Figure 1a). Drp6 associates with nuclear envelope and regulates nuclear remodeling in Tetrahymena. However, Drp6 like other DRPs lacks a PH domain or any recognizable membrane-binding domain (Figure 1a). Sequence alignment revealed the presence of a region in Drp6 that is located at the position of the PH domain of classical dynamin (Figure 1b, Figure 1—figure supplement 1a) and that was earlier named the Drp-targeting determinant (DTD) (Elde et al., 2005).

Figure 1 with 1 supplement see all

Download asset Open asset

To assess whether the DTD is important for recruitment of Drp6 to target membranes, full-length DRP6 and drp6∆DTD were expressed separately as N-terminal GFP-fusion proteins and their cellular distributions were compared by confocal microscopy. As expected, Drp6 was chiefly present on the nuclear envelope and also on some cytoplasmic puncta (Figure 1c, Figure 1—figure supplement 1b). These cytoplasmic puncta of Drp6 are endoplasmic reticulum derived vesicles (Rahaman et al., 2008). When confocal images of GFP-drp6ΔDTD-expressing cells were analyzed, it was observed that the deletion of DTD resulted in complete loss of nuclear localization, and it was mainly associated with cytoplasmic puncta (Figure 1c, Figure 1—figure supplement 1b). These results clearly demonstrate that DTD is necessary for recruiting Drp6 to the nuclear envelope, but not to cytoplasmic puncta. We next evaluated if the DTD is sufficient for nuclear envelope targeting, by expressing it as a GFP-fusion protein. The confocal images of cells expressing GFP-drp6-DTD showed that GFP-DTD often appeared as cytoplasmic puncta, but that nuclear envelope localization in a subset of cells is detectable albeit less prominently as compared to that of GFP-drp6 (Figure 1c, Figure 1—figure supplement 1b). This suggests that the DTD is able to interact with the nuclear envelope. This interaction appears weaker than for the full-length protein, suggesting that other domains also contribute to nuclear recruitment of Drp6. Similarly, other dynamin family members rely for their targeting on a membrane-binding domain but also depends on other domains such as GD (Vallis et al., 1999; von der Malsburg et al., 2011; Bramkamp, 2012). We conclude that DTD is essential but not sufficient for Drp6 recruitment to the nuclear envelope. In contrast, Drp6 does not require its DTD for targeting to the ER-derived cytoplasmic vesicles.

Recruitment of a protein to a target membrane is achieved either by interaction with membrane lipids or by forming a complex with another membrane protein. The DTD is essential for recruitment of Drp6 to the nuclear envelope and may represent a membrane-binding domain. To test this idea, we generated N-terminal histidine-tagged drp6-DTD and DRP6 for bacterial expression and purification and then used the purified proteins in lipid overlay assays. Drp6 was purified to near homogeneity (Figure 2a) and incubated with total Tetrahymena lipid spotted on nitrocellulose membrane either in the presence or absence of GTP. Drp6 interacts with Tetrahymena lipid with or without GTP (Figure 2b), suggesting that it harbors a lipid-binding domain. To identify the lipids with which Drp6 interacts, we performed the overlay assay using commercially available strips spotted with fifteen different lipids (Echelon Biosciences, USA). The results demonstrated that Drp6 specifically interacts with three phospholipids, namely phosphatidylserine (PS), phosphatidic acid (PA), and cardiolipin (CL) (Figure 2b). In order to find out whether lipid binding is a property of the DTD, we partially purified DTD as an N-terminal his tagged protein (Figure 2a) and used it for the overlay assay. Like the full-length protein, DTD also interacts with all three phospholipids (Figure 2b).

Figure 2

Download asset Open asset

We then looked at lipid binding in the physiologically relevant context of a bilayer, using an in vitro binding assay to liposomes containing 10% PA, 10% PS, or 10% CL. All the liposomes also contained 70% PC and 20% PE. In sucrose density gradients, the recombinant Drp6 co-migrated with all the three types of liposomes, appearing in the top (light) fractions (Figure 2c). Drp6 was not found in the gradient top fractions in the absence of added liposomes (Figure 2c). Similarly, Drp6 also failed to co-migrate with liposomes that contained only PC and PE (Figure 2c). Taken together, our results indicate that Drp6 interacts with membranes in vitro and that this depends on the presence of either PS, PA, or CL. DTD by itself interacts with the same three phospholipids as holo-Drp6, consistent with the idea that DTD is the membrane-binding domain. To further test this idea, we performed sub-cellular fractionation of cells expressing GFP-drp6-DTD or GFP-drp6. As shown in Figure 2d, Drp6 appeared in both soluble and membrane fractions. DTD also appeared in the membrane fraction, suggesting that it can bind to membranes in vivo (Figure 2d). Taken together, these results lead us to conclude that DTD is a membrane-binding domain and requires PS, PA, or CL for association with the membrane.

Dynamin binds to membrane lipids via interaction between its PH domain and the PIP2 head group. A hydrophobic patch in the PH domain of classical dynamin is important for membrane association/insertion (Ramachandran et al., 2009). The sequence similarity between PH domain of human dynamin 1 and corresponding region of Drp6 is very low (Figure 1b). However, the structural similarity is very high among all the dynamin family proteins whose structures are known. Therefore, we generated a three-dimensional model of Drp6 in order to identify the corresponding hydrophobic region in the DTD. The 3-D modeling of Drp6 shows that the structure of DTD is not related to PH domain, but that nonetheless has a hydrophobic patch (aa 553–555) (Figure 3a). To test the importance of this patch, we substituted the first residue, I553, with M. We expressed this mutant allele as an N-terminal GFP-fusion (GFP-Drp6-I553M), and in parallel expressed a GFP-fusion of the wild-type protein (GFP-drp6), and characterized their localization in Tetrahymena by confocal microscopy. While the latter localized mainly on the nuclear envelope with few cytoplasmic puncta, the mutant GFP-drp6-I553M was not visible on the nuclear envelope but was instead exclusively present at cytoplasmic puncta (Figure 3b, Figure 3—figure supplement 1). This difference suggests that the isoleucine at 553rd position is important for nuclear localization of Drp6.

Figure 3 with 1 supplement see all

Download asset Open asset

Dynamin and DRPs require binding and hydrolysis of GTP for target membrane localization and membrane remodeling functions. Drp6 hydrolyses GTP in vitro (Kar et al., 2018). To understand whether the mutation at I553 affects GTP binding and/or hydrolysis, we expressed and purified Drp6-I553M and Drp6 as N-terminal histidine-tagged proteins (Figure 4a) and compared their GTPase activities. The GTPase activity of Drp6-I553M (0.061 ± 0.002 nmol/μM/min) was not significantly different from that of Drp6 (0.056 ± 0.003 nmol/μM/min) (Figure 4b). The GTPase activities of both wild-type and mutant proteins were also found to be similar when reactions were carried out for 0–20 min (Figure 4b). We also assessed the Michaelis–Menten constant (K_m) and maximum velocity (V_max) for both these proteins (Figure 4c). The K_m of Drp6-I553M (384 µM) was found to be slightly higher than that of Drp6 (180 µM), suggesting a marginal decrease in GTP binding affinity. The mutation did not affect V_max (0.089 nmol/µM/min for Drp6; 0.0893 nmol/µM/min for Drp6-I553M). Taken together, these results suggest that the defect in nuclear localization of Drp6-I553M is not due to defective GTPase activity.

Figure 4

Download asset Open asset

Figure 4—source data 1

Mutation at I553 does not affect GTPase activity and self-assembly of Drp6.

https://cdn.elifesciences.org/articles/64416/elife-64416-fig4-data1-v2.xlsx

Download elife-64416-fig4-data1-v2.xlsx

Another property of DRPs is their ability to assemble and dis-assemble at their target membranes. This involves self-assembly of helical spirals and ring structures and is important for membrane association and membrane remodeling functions. The self-assembly of Drp6 and Drp6-I553M was evaluated by size exclusion chromatography. Drp6 eluted in the void volume as an oligomer containing at least six monomers, as previously observed (Kar et al., 2018; Figure 4d). Similarly, Drp6-I553M also formed higher order structures, as the majority of the protein eluted in the void volume (Figure 4d). A small peak of material eluting near the 150 kDa marker might correspond to a dimer. The breadth of this smaller peak suggests that it consists of a mixture of monomeric and oligomeric structures, with a dimer at the peak fraction. Since the mutant protein was able to form higher order oligomeric structure, we suggest that mutation does not inhibit its self-assembly. However, there might be a difference in the assembly products formed by the wild-type and mutant proteins. To examine this possibility, we compared the ultrastructure of the wild-type and mutant proteins by electron microscopy of negatively stained preparations of the purified recombinant proteins. The Drp6 appeared mostly as large helical spirals, which are similar to structures found in other DRPs (Figure 4e). Ring-like structures were also present, as also found in other members of the family. Similarly, in Drp6-I553M samples, we observed both helical spirals and ring-like structures (Figure 4e). These results suggest that the I553M mutation does not block in vitro oligomerization of Drp6. Taken together, our results suggest that defective nuclear localization of Drp6-I553M is not due to defects in GTPase activity or perturbation of self-assembly.

Dynamin family proteins including Drp6 associate with their target membranes by interacting with specific lipids. To ask whether the I553M mutation might affect these interactions, we used in vitro membrane-binding assays. Recombinant Drp6 and Drp6-I553M proteins were incubated separately with liposomes containing either 10% CL, or PS, or PA. The association of the proteins with liposomes was then judged based on their co-flotation in sucrose density gradients. As can be seen in Figure 5a, Drp6 co-floated with all the three liposomes and appeared on the top fractions, whereas Drp6-I553M co-floated with PS- or PA-containing liposomes but failed to co-float with CL-containing liposomes. Since the mutant retains the ability to associate with PS- and PA-containing liposomes, the overall membrane-binding activity of Drp6 does not depend on I553. Instead, the mutation of I553 specifically inhibits interaction with CL. Based on these results, we infer that I553 in the membrane-binding domain is necessary for CL interactions and that these interactions are important for membrane targeting in vitro.

Figure 5

Download asset Open asset

We then asked whether the interaction between Drp6 and CL was important for nuclear targeting in vivo. Importantly, while the nuclear envelope of animals (Keenna et al., 1970; Kleinig et al., 1971; Sato et al., 1972; Jarasch et al., 1973) lacks CL, it is present in the nuclear membrane of Tetrahymena (Nozawa et al., 1973). To evaluate whether CL is required for nuclear localization of Drp6, we depleted CL from GFP-drp6-expressing cells using pentachlorophenol (PCP), a polychlorinated aromatic compound. PCP is a respiratory uncoupler and a potent inhibitor of CL synthesis (Ono and White, 1971). Within 30 min of PCP treatment, GFP-drp6 dissociated from nuclear envelopes in the majority of cells while remaining associated with cytoplasmic puncta (Figure 5b). Quantitative analysis showed that while GFP-drp6 was localized at the nuclear envelope of all untreated cells, more than 80% of the cells completely lost nuclear localization of GFP-drp6 with the remaining cells showing decreased nuclear localization upon PCP treatment. To check that the nuclear envelope itself remains intact under these conditions, we localized the nuclear pore protein GFP-Nup3/MacNup98B. We found that GFP-Nup3/MacNup98B was clearly associated with nuclear envelopes before or after PCP treatment (Figure 5b). Moreover, PCP treatment does not disrupt membrane structure in general since the distribution of a cortical membrane-binding protein GFP-Nem1D (Shukla et al., 2018) was also not affected by this treatment (Figure 5b). These results therefore suggest that the delocalization of GFP-drp6 upon PCP treatment is due to loss of CL.

If the defect in localization of Drp6-I553M is due to a defect in CL-dependent targeting, one might expect that the defect would be suppressed in the presence of the wild-type protein, since the mutant protein would co-assemble with the correctly targeted wild type. To test this idea, we co-expressed GFP-drp6-I553M and mCherry-drp6. mCherry-drp6 colocalized almost entirely with GFP-drp6-I553M and was targeted to nuclear envelopes as well as cytoplasmic puncta, strongly suggesting that the mutant protein is able to co-assemble with the wild-type protein (Figure 5c). This result also reinforces the earlier conclusion that the mutation does not affect the overall structure of the protein.

After establishing the necessity of the I553 residue and CL in Drp6 nuclear recruitment, we addressed if the loss of nuclear recruitment of Drp6-I553M is specifically due to the presence of methionine at this position. To test this, we replaced I553 with alanine in the GFP-drp6-I553A protein and expressed it in the cell. Confocal microscopic analysis demonstrated that, similar to GFP-drp6-I553M, GFP-drp6-I553A also lost its association with nuclear envelope and was detected mostly as cytoplasmic puncta (Figure 6a, Figure 6—figure supplement 1). As expected, the wild-type GFP-drp6 was chiefly localized in the nuclear envelope under the same experimental conditions. This result reiterates our earlier conclusion that the I553 residue is critical for nuclear recruitment of Drp6. We further confirmed that the loss of Drp6-I553A recruitment to the nuclear envelope is specifically due to loss of its interaction with CL by performing floatation assay (Figure 6b). For this purpose, we purified recombinant Drp6-I553A protein and tested its binding to liposomes containing either CL, PS, or PA. Western blot analysis showed that Drp6-I553A lost interaction with CL liposomes while retaining its interaction with PS and PA liposomes (Figure 6b). To rule out that the loss of nuclear association is not due to the defect in GTPase activity or self-assembly, we performed size exclusion chromatography and GTP hydrolysis assay. When the purified recombinant Drp6-I553A protein was subjected to size exclusion chromatography, it eluted primarily in the void volume as higher order oligomeric structures (Figure 6c). A small fraction appearing as a mixture of monomers and oligomers with a peak around the dimer was also observed in the chromatogram. The GTPase assay results suggested that the I→A substitution does not affect its activity as assessed by the rate of GTP hydrolysis (Figure 6d,e). Based on these results, we conclude that the loss of Drp6-I553A nuclear recruitment is not due to the defect in GTPase activity or in the self-assembly of Drp6. The results also establish that Ile at the 553rd position is essential for Drp6-CL interactions and is a critical determinant of Drp6 nuclear recruitment.

Figure 6 with 1 supplement see all

Download asset Open asset

Figure 6—source data 1

Mutations at I553 results in loss of nuclear localization and cardiolipin interactions without affecting GTPase activity, self-assembly, overall folding and 3-D conformation of Drp6.

https://cdn.elifesciences.org/articles/64416/elife-64416-fig6-data1-v2.xlsx

Download elife-64416-fig6-data1-v2.xlsx

We next examined whether the loss of interaction between the mutant proteins and CL is due to a change in the overall folding or conformation of Drp6. To assess overall folding, circular dichroism (CD) spectra of the mutants were compared with that of wild-type protein (Figure 6f). The analysis of far-UV CD spectra showed that Drp6 contains both α-helix and β-sheets as secondary structures with an ellipticity minima at 222 nm. Similar spectra were observed for I553M and I553A mutants. These results suggest that the mutant proteins were able to fold correctly. To rule out any change in the three-dimensional conformation due to the amino acid substitutions, we performed fluorescence spectroscopic analysis using intrinsic fluorescence of the lone tryptophan present in the Drp6 protein. The fluorescence spectra of the wild-type Drp6 showed a peak at around 332 nm, suggesting that the tryptophan residue is mostly buried (Figure 6g). The fluorescence spectra of I553M and I553A mutants did not show significant difference from that of wild type, suggesting that the overall structure of the proteins did not change due to mutations. To further assess the conformation around the I553 residue, acrylamide quenching experiments were performed using intrinsic tryptophan fluorescence. Since the single tryptophan (W548) is located only five amino acid residues apart from the I553, it allowed us to monitor the effect of mutations on the local conformation around this isoleucine residue. The accessibility of the tryptophan was measured by determining the quenching constant (k_q) for wild-type protein as well as the mutant proteins (Figure 6h). The results showed comparable values of k_q (1.01 M⁻¹ ns⁻¹ for Drp6; 1.11 M⁻¹ ns⁻¹ for Drp6-I553M and 1.04 M⁻¹ ns⁻¹ for Drp6-I553A) for all the three proteins, suggesting that there is no major change in the local conformation due to mutations. These results lead us to conclude that the loss of CL interaction with I553A or I553M mutant is not due to a change in overall folding or local conformation of the mutant proteins. Taken together, it can be concluded that a single isoleucine residue in the membrane-binding domain is essential for interaction with CL and targeting Drp6 to the nuclear membrane.

We next investigated the role of two amino acid residues (E552 and M554) at the vicinity of I553 in CL binding and nuclear association. For this purpose, two independent mutants drp6-E552D and drp6-M554L were expressed either as N-terminal GFP-fusion proteins to assess their localizations in Tetrahymena cells or as N-terminal histidine fusion proteins for determining CL binding in vitro. Confocal microscopic analysis of the Tetrahymena cells expressing either GFP-drp6-E552D or GFP-drp6-M554L showed that, in addition to cytoplasmic puncta, both the mutant proteins also associated with nuclear envelope (Figure 7a). Floatation assay was performed to evaluate whether these two mutants also bind to CL-containing membrane in vitro. Both Drp6-E552D and Drp6-M554L appeared on the top fractions in the floatation experiments, suggesting their binding to CL (Figure 7b). Therefore, we conclude that the amino acid residues neighboring I553 do not determine the specificity for CL binding and nuclear recruitment of Drp6. Thus, the isoleucine at the 553rd position confers specificity for CL mediated nuclear recruitment of Drp6.

Figure 7

Download asset Open asset

Previously, it has been shown that Drp6 is essential for MAC development in Tetrahymena (Rahaman et al., 2008). We have now established that interaction of Drp6 with CL is important for nuclear recruitment. Therefore, we hypothesized that inhibition of CL-Drp6 interaction would inhibit Drp6 function in MAC development. We took two independent approaches to perturb the interaction between CL and Drp6 and assessed the effect on MAC development. In the first approach, CL was depleted by treating cells at a stage prior to MAC development with PCP and then measuring the efficiency of new MAC formation in conjugating cells (Figure 8a). Quantitative analysis showed that while 71 ± 3.6% of the conjugants developed MACs in the control pairs, only 24 ± 1.7% developed MACs in the PCP-treated pairs (Figure 8a). In the second approach, we perturbed CL–Drp6 interaction by treating conjugants with nonyl acridine orange (NAO). NAO interacts with CL with very high affinity and has been used in mammalian cells to block interactions between CL and mitochondrial proteins involved in electron transport (Maftah et al., 1990). Exposing conjugants to NAO significantly inhibited new MAC development (Figure 8a).

Figure 8

Download asset Open asset

Figure 8—source data 1

Inhibition of Drp6-CL interaction inhibits macronuclear development in Tetrahyemena.

https://cdn.elifesciences.org/articles/64416/elife-64416-fig8-data1-v2.xlsx

Download elife-64416-fig8-data1-v2.xlsx

In conjugating Tetrahymena, MAC development is not the only phenomenon requiring nuclear expansion. At a prior stage, the germline micronuclei (MICs) show dramatic elongation (Cole et al., 1997). We found that PCP treatment did not affect the frequency of elongation, that is, the percentage of pairs showing elongated MICs, but did produce a decrease in the extent of MIC elongation (Figure 8a). This inhibition of elongation had no detectable consequences for the subsequent stage of MIC meiosis. Treatment with NAO had no measurable effect on MIC elongation (Figure 8a) or the subsequent meiosis. These results are consistent with the idea that the key requirement for CL is during MAC expansion.

We next reasoned that if interaction of Drp6 with CL is important for MAC expansion, then over-expression of the isolated Drp6 membrane-binding domain might competitively inhibit the interaction and block Drp6 function during MAC expansion. To test this possibility, we expressed GFP-drp6-DTD in Tetrahymena and then allowed the cells to conjugate with a Drp6 wild-type strain. We measured MAC development in these pairs, and in pairs from a parallel WT x WT cross, at 8 hr during conjugation. In the control cross, more than 75% of pairs developed new MACs. In striking contrast, only 4–5% of pairs developed normal MACs in the pairs that included GFP-drp6-DTD-expressing cells (Figure 8b). Therefore, MAC development was almost completely blocked when GFP-drp6-DTD-expressing cells comprised one of the conjugation partners (Figure 8b).

In similar experiments, we also asked whether the expression of GFP-drp6∆DTD might have a dominant-negative inhibitory effect on Drp6 function. Indeed, we found that in pairs where one cell over-expressed the ∆DTD construct, the pairs showed inhibition of MAC development that was similar to that induced by expression of the isolated DTD domain (Figure 8b). Neither the expression of GFP-drp6-DTD nor GFP-drp6∆DTD significantly reduced the fraction of pairs showing elongated MICs (Figure 8b).

In conclusion, prior experiments with DRP6 gene knockout pointed to a specific function in MAC development. Our current results from over-expression of mutant alleles are consistent with this idea. Taken together, our results support a model in which interaction of Drp6 with CL, for which a single isoleucine residue acts as a key determinant, is critical for nuclear targeting and therefore for MAC expansion.

Tetrahymena thermophila CU428 and B2086 strains were obtained from Tetrahymena Stock Center (Cornell University, USA). Cells were cultured in SPP medium (2% proteose peptone [BD, USA], 0.2% glucose, 0.1% yeast extract, and 0.003% ferric EDTA) at 30°C under shaking at 90 rpm. For conjugation, mating type cells were grown in SPP media to a density of 3 × 10⁵ cells/ml, washed and resuspended in DMC media (0.17 mM sodium citrate, 0.1 mM NaH₂PO₄, 0.1 mM Na₂HPO₄, 0.65 mM CaCl₂, 0.1 mM MgCl₂), and incubated at 30°C, 90 rpm for 16–18 hr. To initiate conjugation, starved cells of two different mating types were mixed and incubated at 30°C without shaking. All the reagents were purchased from Sigma–Aldrich unless mentioned otherwise.

The Drp6∆DTD lacking aa 517–600 was created by overlap PCR using a set of four oligonucleotides in a two-step PCR method. Drp6-E552D, Drp6-I553M, Drp6-I553A, and Drp6-M554L were generated by site-directed mutagenesis using Quick Change protocol (Stratagene). For expression in Tetrahymena, rDNA-based vector pVGF or pIGF was used. While the PCR products of Drp6 (aa 1–710) and Drp6-DTD (aa 517–600) were inserted between XhoI and ApaI restriction sites of pVGF, the Drp6-E552D, Drp6-I553M, Drp6-I553A, Drp6-M554L, and Drp6∆DTD were introduced into pIGF using Gateway cloning strategy (Invitrogen) using the manufacturer’s protocol and were expressed as N-Ter GFP-tagged fusion proteins. All the constructs were confirmed by sequencing. Conjugating wild-type Tetrahymena cells were transformed with these constructs by electroporation, and the transformants were selected using 100 µg/ml paromomycin sulfate. For the co-expression studies, mCherry-drp6 was generated by introducing mCherry sequences between PmeI and XhoI sites followed by Drp6 sequences between XhoI and ApaI sites of NCVB vector. Co-transformants were generated by biolistic transformation of linearized mCherry-drp6 nucleotide sequences into the cells expressing GFP-drp6-I553M and were selected in the presence of 60 µg/ml blasticidine and 120 µg/ml paromomycin sulfate supplemented with 1 µg/ml cadmium chloride.

Cells were grown to a log phase (2.5–3.5 × 10⁵ cells/ml), and expression was induced by adding cadmium chloride at concentration of 1 µg/ml for 4 hr. Cells were harvested at 1100 g, fixed with 4% paraformaldehyde for 20 min at RT, washed and resuspended in 10 mM HEPES, pH 7.5, and DAPI stained before imaging.

For expression in Escherichia coli, the amplified PCR products of Drp6 and Drp6-DTD were cloned into pRSETB using BamHI and EcoRI sites. The Drp6-E552D, Drp6-I553M, Drp6-I553A, and Drp6-M554L were cloned in pRSETB vector using Quick Change protocol (Stratagene). The resulting constructs were transformed into chemically competent E. coli C41(DE3) cells and transformants were inoculated into LB broth supplemented with 100 μg/ml ampicillin and grown at 37°C till the OD₆₀₀ reached 0.4. The cultures were then shifted to 18°C and expression was induced after 1 hr by adding 0.5 mM IPTG (Sigma) and kept for 16 hours at the same temperature before harvesting the cells. The harvested cells were resuspended inice-cold buffer A (25 mM HEPES pH 7.5, 300 mM NaCl, 2 mM MgCl₂, 2 mM β-mercaptoethanol, and 10% glycerol) supplemented with EDTA-free protease inhibitor cocktail (Roche) and 100 mM phenyl methyl sulfonyl fluoride, lysed by sonication, and the lysates were centrifuged at 52,000 g for 45 min at 4°C. The supernatants were incubated with Ni-NTA agarose resin (Qiagen, Germany) for 2 hr before washing with 100 bed volume buffer A supplemented with 50 mM imidazole. The bound proteins were eluted with 250 mM imidazole in buffer A. The purified proteins were checked by Coomassie-stained SDS–PAGE gels, and the purity was assessed by Image J analysis (NIH, USA). The fractions containing the purified proteins were pooled, dialyzed with buffer A, and concentrated using Amicon ultra-15 filters (Merck-Millipore, Germany). Protein concentration was estimated by Bradford assay (Bio-Rad Laboratories, USA).

Samples were subjected to SDS–PAGE gel, and the proteins were transferred to PVDF membrane. Membrane was blocked with 2% bovine serum albumin (BSA) in TBST (50 mM Tris–Cl, 150 mM NaCl, pH 8.0%, and 0.05% Tween 20) for 1 hr. The blot was then incubated with HRP-conjugated anti-His monoclonal antibody (1:5000) and detected with supersignalfemto substrate (Thermoscientific, USA) using ChemiDoc imaging system (Bio-Rad Laboratories, USA).

Tetrahymena cells expressing GFP-drp6 or GFP-DTD were lysed in 500 µl of ice-cold lysis buffer containing 25 mM Tris–Cl pH 7.5, 300 mM NaCl, 10% glycerol supplemented with E-64, pepstatin, aprotinin, PMSF, and protease inhibitor cocktail (Roche) by passing through a ball-bearing homogenizer with a nominal clearance of 0.0007 in. The resulting lysates were centrifuged at 16,000 g for 15 min at 4°C, the supernatant was collected as soluble protein fraction, and the pellet containing the membrane fraction was resuspended in 500 µl lysis buffer. The proteins in both soluble fraction and membrane fraction were separated in 12% SDS–PAGE gel and analyzed by western blotting using anti-GFP polyclonal antibody (1:4000; Sigma–Aldrich).

Total Tetrahymena lipid was extracted from growing Tetrahymena cells (5 × 10⁵ cells/ml) following the method by Bligh and Dyer, 1959. Drops of 5 µl in chloroform were spotted on the nitrocellulose membrane and incubated with His-drp6 (90 µg/ml) in GTPase assay buffer in the presence or absence of 1 mM GTP for 1 hr. In control experiments, BSA was used in place of His-drp6. The assay using membrane lipid strips (P-6002, Echelon Biosciences, USA) spotted with 100 pmol of 15 different lipids were used according to the manufacturer’s instruction. The binding of proteins was detected by western blot analysis using anti-His monoclonal antibody.

Lipids (Avanti Polar) were dissolved in analytical grade chloroform, and liposomes were prepared using 2.5 mg total lipid in 1 ml chloroform. The liposomes contained 70% PC and 20% PE along with either 10% CL or 10% PA or 10% PS. A thin dry film was obtained by drying the solution in a round bottom flask, and solvent was completely removed in a lyophilizer. Liposomes were made by rehydrating the film in buffer A (25 mM HEPES pH 7.5, 2 mM MgCl₂, 150 mM NaCl) pre-warmed at 37°C. The resuspended solution was extruded 17–21 times through extruder (Avanti Polar) using filter with 100 nm pore, and the size distribution was measured by DLS in Malvern Zetasizer Nano. For floatation assay, 1 µM protein was incubated with 0.5 mg liposomes in buffer A supplemented with 1 mM GTP for 1 hr at RT. Sucrose was added to the reaction mixture (final sucrose concentration 40%), placed at the bottom of a 13 ml ultra-centrifugation tube, and overlaid with 2 ml each of 35%, 30%, 25%, 20%, 15%, and 0% sucrose solutions in the same buffer. The gradient was subjected to ultra-centrifugation in Beckman Coulter ultra-centrifuge at 35,000 RPM for 15 hr at 4°C using SW41 rotor. Fractions (1 ml each) were collected from top and detected by western blotting using anti-His HRP-conjugated monoclonal antibody (1:5000).

The GTP hydrolysis activity of recombinant Drp6, Drp6 I553M, and Drp6-I553A was measured in a colorimetric assay using Malachite Green-based phosphate assay reagent (BIOMOL Green, Enzo Life Sciences, USA). The GTPase assay (20 µl in 25 mM HEPES pH7.5, 15 mM KCl, 2 mM MgCl₂) was performed in the presence of 1 mM GTP (Sigma) using 1 μM protein for 0–30 min at 37°C. The reaction was stopped by adding 5 μl of 0.5 M EDTA, and absorbance was measured at 620 nm. For measuring K_M and K_cat, reactions were performed for 10 min at 37°C in triplicate using varying concentrations of GTP (50 μM–2000 μM). The values obtained from three independent experiments were plotted and analyzed using GraphPad Prism7 software. The statistical analysis was performed by using unpaired t-test.

Size exclusion chromatography was performed on the Superdex 200 GL 10/300 column (GE Life Sciences, USA) using Akta Explorer FPLC system (GE Healthcare, USA), which was calibrated with standard molecular weight markers (Sigma–Aldrich, USA). Five hundred microliters of protein (0.5 mg/ml) in buffer A was loaded onto the pre-equilibrated column and was run at 0.5 ml/min. The chromatogram was recorded by taking absorbance at 280 nm.

Purified recombinant Drp6 or Drp6 I553M (1 μM) was incubated with 0.5 mM GTPγS in 25 mM HEPES pH 7.5, 150 mM NaCl, and 2 mM MgCl₂ for 20 min at room temperature and was adsorbed for 2 min onto a 200 mesh carbon coated Copper grid (Ted Pella, Inc, USA). The grid was stained with a drop of 2% freshly prepared uranyl acetate (MP Biomedicals, USA) for 2 min and dried at room temperature for 10 min. The electron micrographs were collected on a FEI Tecnai G2 120 kV electron microscope.

Tetrahymena cells were fixed with 4% paraformaldehyde (PFA) in 50 mM HEPES pH 7.5 for 20 min at RT and were collected in 10 mM HEPES pH 7.5 after centrifugation at 1100 × g. The fixed cells were stained with DAPI (0.25 ug/ml) (Invitrogen, USA) and washed with 10 mM HEPES pH 7.5 before imaging. The images were collected in a Zeiss LSM780 or Leica DMI8 confocal microscope.

CD measurements were recorded at 145 µg/ml protein in 25 mM HEPES pH 7.5, 300 mM NaCl, 2 mM MgCl₂, 2 mM β-mercaptoethanol using a Jasco J-1500 instrument (Jasco Inc, USA). Ellipticity (millidegrees) was measured between 260 and 205 nm. Mean residue ellipticity ([θ]_MRW in deg cm² dmol⁻¹) for each wavelength was calculated by using the following formula:

where θ is the ellipticity in mdeg, M_r is the molecular weight of the protein in Da, c is protein concentration in mg/ml, l is the path length in centimeter, and N_A is the number of amino acid residues of the protein.

For the fluorescence spectroscopy experiments, 300 µl of 0.2 µM protein in 25 mM HEPES pH 7.5, 300 mM NaCl, 2 mM MgCl₂, and 2 mM β-mercaptoethanol were measured by excitation at 295 nm (2 nm band pass) in a 10 mm quartz cuvette using FLS-1000 (Edinburgh Instruments Ltd., UK). For the fluorescence quenching experiments, different acrylamide concentrations (0–400 mM) were added to 300 µl of 1.5 µM protein sample, and the spectra were recorded in the Cary Eclipse fluorescence spectrophotometer (Agilent Technologies, USA). The emission spectra were collected from 310 to 410 nm (10 nm band pass), and the values at 332 nm (the emission peak) were analyzed using Stern–Volmer equation, expressed as follows:

where F₀ is the fluorescence in the absence of quencher and F is the fluorescence in the presence of quencher. Q denotes the concentration of the quencher in M. K_sv is the Stern–Volmer constant. The obtained values were plotted using GraphPad Prism 7, and K_sv was calculated from the slope of the graph. The quenching constant was calculated using the equation:

Where k_q is the quenching constant (in Mol⁻¹ ns⁻¹) and τ is the life time (in ns) of tryptophan. The value of τ is considered to be 2.7 ns as reported earlier (Ronda et al., 2018).

The growing Tetrahymena cells either expressing GFP-drp6 or GFP-Nup3/MacNup98B (a kind gift from Prof. Douglas Chalker, Washington University in St. Louis; Malone et al., 2008; Iwamoto et al., 2009) or GFP-Nem1D were treated with 10 µM PCP in DMSO for 30 min before fixing with 4% paraformaldehyde. The conjugation pairs were treated with either 30 µM PCP or 0.5 µM NAO (Invitrogen, USA) after 2.5 hr, 4.5 hr, or 7.5 hr post-mixing and were fixed after 30 min of addition. The cells were stained with DAPI (0.25 µg/ml) before imaging. The statistical analysis was performed by using unpaired t-test.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

We thank Prof. Aaron Turkewitz from the University of Chicago, Dr Kausik Chakraborty from IGIB, Prof. Jacek Gaertig from University of Georgia, Dr Utpal Nath from Indian Institute of Science, and Dr Renjith Mathew from National Institute of Science Education and Research for critical evaluation and useful comments on the manuscript. For other help, we thank members of this laboratory Sakti Ranjan Rout and Soham Mukhopadhyay. The work was partly funded by DBT grant (BT/PR14643/BRB/10/862/2010) to AR. The funders have no role in study design and interpretation, or the decision to submit the work for publication.

1. Received: October 29, 2020
2. Accepted: March 3, 2021
3. Accepted Manuscript published: March 4, 2021 (version 1)
4. Version of Record published: March 9, 2021 (version 2)

© 2021, Kar et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.