Identification of putative effectors of the Type IV secretion system from the Wolbachia endosymbiont of Brugia malayi.

Wolbachia is an unculturable, intracellular bacterium that persists within an extremely broad range of arthropod and parasitic nematode hosts, where it is transmitted maternally to offspring via vertical transmission. In the filarial nematode Brugia malayi, a causative agent of human lymphatic filariasis, Wolbachia is an endosymbiont, and its presence is essential for proper nematode development, survival, and pathogenesis. While the elucidation of Wolbachia:nematode interactions that promote the bacterium's intracellular persistence is of great importance, research has been hampered due to the fact that Wolbachia cannot be cultured in the absence of host cells. The Wolbachia endosymbiont of B. malayi (wBm) has an active Type IV secretion system (T4SS). Here, we have screened 47 putative T4SS effector proteins of wBm for their ability to modulate growth or the cell biology of a typical eukaryotic cell, Saccharomyces cerevisiae. Five candidates strongly inhibited yeast growth upon expression, and 6 additional proteins showed toxicity in the presence of zinc and caffeine. Studies on the uptake of an endocytic vacuole-specific fluorescent marker, FM4-64, identified 4 proteins (wBm0076 wBm00114, wBm0447 and wBm0152) involved in vacuole membrane dynamics. The WAS(p)-family protein, wBm0076, was found to colocalize with yeast cortical actin patches and disrupted actin cytoskeleton dynamics upon expression. Deletion of the Arp2/3-activating protein, Abp1p, provided resistance to wBm0076 expression, suggesting a role for wBm0076 in regulating eukaryotic actin dynamics and cortical actin patch formation. Furthermore, wBm0152 was found to strongly disrupt endosome:vacuole cargo trafficking in yeast. This study provides molecular insight into the potential role of the T4SS in the Wolbachia endosymbiont:nematode relationship.

Introduction

After leprosy, lymphatic filariasis is the leading cause of permanent disability, afflicting at least 150 million people worldwide [1, 2]. The World Health Organization reports that 1.23 billion people in 58 countries are at risk for developing the disease [1]. Lymphatic filariasis results from the mosquito-borne transmission of three distinct pathogenic nematodes: Wuchereria bancrofti, Brugia timori and Brugia malayi. In the early stages of the infection, anti-parasitic agents such as diethylcarbamazine and ivermectin have proved to be effective in eliminating immature worms and have been used in mass treatment programs [3]. Unfortunately, adult worms which are largely responsible for the pathology associated with infection cannot be easily treated by early-stage anti-parasitic drugs [4, 5]. With the discovery that these filarial nematodes require the presence of an obligate intracellular, Gram-negative bacterial endosymbiont of the Wolbachia genus to survive and reproduce, effective clearance of filarial infections has been achieved with antibiotic treatment [4-8]. Accordingly, significant efforts to understand the interactions of Wolbachia with nematodes at the molecular level have been undertaken to identify additional potential drug targets [9].

In B. malayi, Wolbachia exists and replicates within membrane-bound vacuoles in the lateral hypodermal chord and female germline [10-12]. These bacteria-laden compartments are often seen in close association with the endoplasmic reticulum and Golgi compartments [13], suggesting that Wolbachia may directly interact with host organelles to support its survival and co-opt normal host pathways in order to obtain nutrients and to prevent autophagic degradation in the host [14-16]. Only recently, however, have researchers identified a small number of secreted proteins from Wolbachia that appear to directly interact with host nematode proteins [17, 18]. Identification and characterization of the full complement of Wolbachia:host interactions will likely provide important information regarding the ability of Wolbachia, and other endosymbionts, to be maintained in the host.

Many Gram-negative intracellular bacteria utilize type III or type IV secretion systems (T3SS; T4SS) to deliver bacterial proteins into host cells, ensuring their successful invasion, survival, and replication within the host [19, 20]. Recent genomic studies have shown that Wolbachia from the filarial nematode B. malayi (wBm) contain the necessary gene products to produce an active T4SS which is regulated by the wBmxR1 and wBmxR2 transcriptional regulators [16]. It is thought that Wolbachia utilizes its T4SS to secrete a number of effector proteins to modulate bacteria:host interactions, much like the related human pathogenic bacteria from the order Rickettsiales, such as Rickettsia conorii, Ehrlichia chaffeensis, and Anaplasma phagocytophilium [21]. However, identifying and characterizing these putative effectors has proven difficult in the filarial system due to the fact that the full life cycle of neither Wolbachia nor B. malayi can be supported in culture for facile genetic or biochemical manipulation. With modern bioinformatics, molecular biology, and a surrogate eukaryotic host system capable of genetic manipulation, the activities of individual putative effector proteins from these bacteria can be evaluated.

The budding yeast Saccharomyces cerevisiae has routinely been used as a “host” organism to screen for T3SS/T4SS effectors, as many of its core physiological pathways, such as protein trafficking, endolysosmal membrane dynamics, and cytoskeletal dynamics, are conserved in higher eukaryotes [22-24]. By expressing individual genes encoding predicted effector proteins in yeast, researchers have successfully identified and characterized several secreted effectors from a diverse set of bacterial pathogens including Shigella flexneri [25, 26], Legionella pneumophila [27, 28], and Chlamydia trachomatis [29]. Oftentimes, these effector proteins contain eukaryotic-like motifs, such as ankyrin repeat and coiled-coil domains–known protein-protein interaction motifs–suggesting the ability of these effectors to directly interact with and modulate eukaryotic cell biology. Secreted bacterial effectors with activities on eukaryotic physiology are often inhibitory to yeast growth [26] and can therefore be rapidly screened to identify potential candidates for further study. Additionally, powerful biochemical assays to measure protein trafficking pathways and the visualization of intracellular organelle morphologies can be used to identify effectors that may modulate intracellular membrane or general organelle dynamics during infection or symbiosis [30, 31].

By using some of these classical eukaryotic protein motifs as an identifier, and combining this information with genes known to be associated with, or co-transcribed with, the components of the wBm T4SS [16, 32], we have identified and cloned forty-seven potential effector proteins from wBm into yeast expression vectors. These putative effectors were then screened for the ability to inhibit yeast growth, alter normal vacuole protein sorting pathways, and to disrupt normal yeast vacuolar morphology. Five candidates strongly inhibit yeast growth upon expression, and six additional proteins show toxicity in the presence of zinc and caffeine. Studies on the uptake of an endocytic vacuole-specific fluorescent marker, FM4-64, identify 4 proteins (wBm0076, wBm00114, wBm0152, and wBm0447) capable of altering vacuole membrane dynamics. Additionally, we find that wBm0076 disrupts normal actin cytoskeleton dynamics by inducing the aberrant formation of cortical actin patches in vivo. wBm0152 expression strongly inhibits the delivery of representative endosomal protein cargo to the yeast vacuole, suggesting the ability of a Wolbachia protein to manipulate eukaryotic membrane traffic. These findings provide insight into the potential role of these proteins in the nematode:Wolbachia symbiotic relationship.

Results

Identification of a set of putative effectors from the wBm genome

In a previous study [16], the existence of an active bacterial T4SS was confirmed in the Wolbachia endosymbiont of the filarial parasite Brugia malayi. Due to the probability that the T4SS acts as a major communication channel between Wolbachia and B. malayi, we set out to identify potential substrates/effectors of the wBm T4SS. Using a variety of bioinformatic approaches (Materials and Methods), we identified 47 candidate effectors (Table 1) that displayed one or more of the following characteristics: i) eukaryotic-like protein possessing domains which share higher levels of homology to eukaryotic proteins than bacterial ones. These regions of homology include ankyrin repeats, Ser/Thr kinase motifs and coiled-coil domains, which have been previously identified in the T4SS effector proteins from Legionella pneumophila [33], Bartonella henselae [34], and Coxiella burnetii [35], ii) protein sequences that are potentially related to T4SS machineries, based on sequence homology to Rickettsia sibirica proteins shown to interact with components of its T4SS [36], or genes that are co-transcribed with components of the wBm T4SS [16], and iii) bacterial proteins identified in excretory-secretory products from B. malayi adult worms and microfilaria [37]. Each of the 47 open reading frames was cloned into the galactose-inducible yeast expression vector pYES2/NT A (see Materials and Methods), and further studied for their ability to modulate yeast physiology.

Table 1

Wolbachia putative Type IV effector proteins screened in this study.

wBm gene designation	Predicted function	Reasona	wBm gene designation	Predicted function	Reasona
WBM0014	ABC transporter	4	WBM0430	invasion associated protein B	S
WBM0032	tRNA modification GTPase	4	WBM0432	WSP family	S, Bm
WBM0044	amino acid transporter	Bm	WBM0447	ankyrin repeat protein	S, Bm, E
WBM0057	ZapA superfamily, hypothetical		WBM0452	SAM-dependent methyltransferase	E, 4
WBM0064	FAD-dependent oxidoreductase		WBM0482	hypothetical, tropomyosin-like	Bm, E
WBM0070	Mg2+/Co2+ transporter	4	WBM0484	hypothetical protein	S
WBM0076	WAS(p) family protein, proline-rich	E	WBM0491	cytochrome C oxidase assembly protein	S
WBM0100	Wolbachia Surface Protein (WSP) family	S	WBM0506	outer membrane protein	E
WBM0114	peptide deformylase	4	WBM0582	ankyrin repeat protein	S
WBM0152	PAL-like	S, E	WBM0665	ankyrin repeat protein	4
WBM0164	hypothetical, tropomyosin-like	E	WBM0666	dihydrolipoamide acetyltransferase	4
WBM0165	hypothetical protein	U	WBM0671	DNA uptake lipoprotein
WBM0181	ATP-binding chaperone	Bm, 4	WBM0672	hypothetical protein	Bm, U
WBM0193	hypothetical, tropomyosin-like	Bm, E	WBM0709	coprophyrinogen III oxidase	4
WBM0209	pyruvate phosphate dikinase	Bm	WBM0711	DNA recombination rmuC-like	S, Bm, 4
WBM0213	hypothetical protein	E, U	WBM0736	secreted FK506 binding protein-like	Bm
WBM0222	preprotein translocase YajC	4	WBM0748	hypothetical protein	4
WBM0277	calcineurin-like phopshodiesterase	4	WBM0749	hypothetical protein	Bm, 4
WBM0284	WSP family	4	WBM0751	hypothetical protein	4
WBM0287	ankyrin repeat protein	E	WBM0752	hypothetical protein	4
WBM0290	D-alanyl D-alanine carboxypeptidase	S	WBM0772	hypothetical protein	Bm, E
WBM0307	cytochrome C oxidase subunit	4	WBM0791	N6-adenine methylase
WBM0384	ankyrin repeat protein, metalloprotease	E	WBM0792	hypothetical protein	E
WBM0394	ankyrin repeat protein	E

aCharacteristics of gene or protein product that supported selection: S, detected in Wolbachia secretome; Bm, detected in dissected Brugia malayi; E, predicted eukaryotic-like protein motifs; 4, co-regulated with wBm Type IV secretion apparatus or similar to predicted Type IV effectors from related organisms; U, unique to Wolbachia. ORFs noted in bold text were visualized to express in yeast via immunoblot (S1 Fig).

Putative wBm effectors inhibit yeast growth

In an effort to identify those wBm proteins that manipulate eukaryotic physiology, we tested the ability of each of the putative effectors to inhibit yeast growth; many bacterial effector proteins have been identified in such a manner [25-29]. Each galactose-inducible plasmid harboring an individual wBm “effector” was transformed into a yeast strain and gene expression was induced via a modified Gal4p transcriptional regulator which responds to the hormone β-estradiol, instead of the usual galactose (Materials and Methods), thus allowing GAL promoter induction with the same carbon source as uninduced conditions (glucose) and preventing any confounding results that may be introduced by changing carbon sources across expression conditions. Under growth conditions containing β-estradiol, we were able to detect full-length protein expression of 24 of the 47 putative effector proteins via the N-terminal epitope provided by the pYES2/NT A expression plasmid (S1 Fig, also noted on S1 Table). It is likely that the total number of expressed proteins is underrepresented using this detection method, however, since N-terminal protein processing or degradation events may occur upon protein overproduction in vivo; further possibilities that could result in translation defects of these putative effectors were not explored. When strains harboring these plasmids were grown on minimal medium with or without β-estradiol, we found wBm0014, wBm0032, wBm0044, wBm0076, and wBm0284 strongly inhibit yeast growth in the presence of 1 μM β-estradiol (determined by a reduction in cell titers by three orders of magnitude), but not in the absence of inducer (Fig 1A). Slight inhibition in yeast growth (as determined by a reduction in cell titers by two orders of magnitude) was observed upon expression of wBm0070, wBm0114, and wBm0222 (Fig 1A). Other expressed wBm genes did not strongly inhibit the growth of yeast under these conditions (S2 Fig).

Fig 1

wBm genes inhibit yeast growth upon expression.

(A) BY4742 yeast strains genetically modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods) and harboring the GAL-inducible control plasmid pYES2/NT A, or pYES2/NT A containing the specified wBm open reading frame were grown to saturation in CSM-uracil medium containing 2% glucose, and each culture was diluted to OD600 = 1.0 in sterile 0.9% NaCl. 10-fold serial dilutions were spotted onto CSM-uracil containing 2% glucose with and without 1 μM β-estradiol. Plates were incubated for 48 h at 30°C. (B) Strains and dilutions from (A) were spotted to CSM-uracil media containing 2% glucose supplemented with 7.5 mM ZnCl2, 5 mM caffeine, and with or without 1 μM β-estradiol. Plates were incubated for 72 h at 30°C. Images are representative of three independent experiments.

Due to the ability of Wolbachia to live inside membrane-bound compartments in B. malayi cells [13], we next hypothesized that Wolbachia could secrete proteins that subvert normal endolysosomal membrane dynamics, potentially as a method of avoiding autophagic degradation [38, 39]. As yeast strains are known to require normal endolysosomal or vacuolar function to resist high levels of zinc and caffeine [40, 41], we tested the ability of these wBm genes to induce growth sensitivity to 7.5 mM ZnCl2 and 5 mM caffeine. As expected, putative effectors that were toxic on standard minimal media showed enhanced β-estradiol-dependent toxicity on zinc/caffeine plates (Fig 1B). Interestingly, strains harboring the wBm0076 construct were sensitive to zinc/caffeine even under non-induction conditions (Fig 1B, lower left), suggesting that even low-level expression of wBm0076 is sufficient to inhibit yeast growth under these conditions. Because expression of each of the eight above identified putative effectors alone is sufficient to inhibit yeast growth under normal growth conditions, no conclusions can be drawn from the similarly defective growth observed in the presence of both β-estradiol and zinc/caffeine; it is not known if this observed toxicity is a result of protein-induced endolysomal membrane trafficking defects or other protein-induced defects to yeast biology. Of note, however, was the discovery that wBm0057, wBm0064, wBm0213, wBm0277, wBm0287 and wBm0447 were only strongly toxic in the presence of zinc and caffeine (as determined by an observable reduction in cell titers by three orders of magnitude) (Fig 1B), suggesting that these strains may have defects in normal endolysosomal homeostasis, such as in defects in the TOR kinase pathway [42, 43] or protein trafficking pathways [44], although defects in DNA repair pathways may also contribute to caffeine sensitivity [45]. No other plasmids were found to induce additional sensitivity of yeast to zinc/caffeine treatment upon expression, though several others show slight inhibition of yeast growth (S2 Fig). Taken together, we find several putative effector proteins from Wolbachia that inhibit the growth of yeast, suggesting these proteins may function to modulate essential eukaryotic pathways during Wolbachia symbiosis.

Putative Wolbachia effectors disrupt yeast vacuole membrane dynamics

As reported previously, Wolbachia likely avoids both lysosomal and autophagic degradation pathways to survive within a membrane-bound organelle since chemical activation of autophagy results in clearance of Wolbachia from maturing nematodes [39]. Due to the fact that wBm must avoid normal host autophagic pathways, it is likely that wBm secretes proteins that alter normal host endolysosomal membrane trafficking pathways. The degradative vacuole of yeast has long been used as a model organelle for mammalian lysosome function and eukaryotic endolysosomal fusion dynamics [46], therefore we assayed the effects of wBm putative effector expression on yeast vacuole dynamics.

The yeast vacuole is a highly dynamic organelle that undergoes constant rounds of fission and fusion in response to the osmolarity of the extracellular environment [47]. The homotypic fusion of yeast vacuoles, like the majority of eukaryotic intracellular membrane fusion events, is governed by a conserved set of protein machinery consisting of soluble NSF attachment protein receptors (SNAREs), a Rab-family small GTPase, and a Rab/SNARE-binding multisubunit tethering factor, the homotypic vacuole fusion and protein sorting (HOPS) complex (reviewed in [48]). Yeast strains defective in the proper regulation of homotypic fusion, either directly or indirectly, generally display aberrant vacuolar morphologies. These phenotypic changes have long been used to identify important genes and regulators of endolysosomal membrane fusion and protein trafficking machineries in yeast [41, 49]. Six different classes of vacuoles have been described, classified as A through F [50]; class A vacuoles are considered a wild type morphology. The remaining classes B-F are considered abnormal vacuole morphologies and are phenotypically described briefly: class B vacuoles represent highly fragmented vacuoles containing more than twenty vacuoles per cell, class C vacuoles are highly fragmented with no discernable vacuolar structure, class D vacuoles have a single enlarged vacuole, class E vacuoles have an enlarged prevacuolar compartment with small vesicles present, and class F vacuoles have many small vacuoles surrounding a larger prominent vacuole. Importantly, the expression of some secreted effectors from pathogenic bacteria in yeast show the ability to both fragment the yeast vacuole in vivo and inhibit the fusion of yeast vacuoles in vitro [40, 51, 52].

Yeast strains harboring a control plasmid under induction conditions (1 μM β-estradiol) show 1–5 large vacuoles per cell, as stained by the endocytic uptake of the vacuole-specific fluorescent marker, FM4-64 [30] (Fig 2). Under the same growth conditions, strains expressing wBm0076 display an apparent highly fragmented, class C vacuole morphology in approximately 74% of the cells observed, often indicative of either a complete block in vacuole biogenesis, including defects in homotypic vacuole fusion, or disruptions in the delivery of FM4-64-containing endocytic vesicles to the vacuole [53, 54]. In strains expressing wBm0114, approximately 35% of the observed cells display a large number of small, vacuole-like, class B compartments that may indicate defects in vacuolar protein sorting pathways [53, 55]. In the presence of wBm0152, more than half of the yeast cells contain large vacuolar compartments with accumulations of small membrane compartments immediately adjacent to the vacuolar limiting membrane; these accumulations are likely enlarged pre-vacuolar compartments (class E compartments) that arise in strains defective in endosome:vacuole fusion or endosomal sorting complexes required for transport (ESCRT) function [50, 56]. Finally, approximately 30% of cells expressing wBm0447 display an aberrant vacuole morphology that was not easily classified: most cells were multi-vacuolar with small membrane accumulations at the plasma membrane (Fig 2). No other strains harboring wBm genes displayed a significant defect in yeast vacuole morphology (> 85% wild type vacuoles, S3 Fig); the ability of the selected wBm proteins to strongly manipulate yeast endolysosomal membrane dynamics is specific to a select few putative effectors.

Fig 2

Normal vacuole dynamics are altered upon wBm gene expression.

BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods), and harboring GAL-inducible pYES2/NT A control plasmid or pYES2/NT A containing an individual wBm open reading frame were grown to saturation in CSM-uracil medium, subcultured to CSM-uracil and grown for 6h at 30°C with or without the addition of 1 μM β-estradiol. Cells were stained for 20 minutes with 10 μM FM4-64 at 30°C followed by a 1.5 h chase in CSM-uracil at 30°C. Vacuoles were visualized via microscopy, and cells displaying abnormal vacuole morphologies were counted. Percentage and standard deviation of cells displaying the corresponding abnormal vacuolar morphology is noted and determined from three independent experiments; n ≥ 100 cells per experiment. Bar = 3 μ.

wBm0152 disrupts endosome:Vacuole cargo trafficking in yeast

To further characterize the potential defects in endolysosomal trafficking caused by the expression of putative wBm effector proteins, we decided to identify wBm proteins that caused aberrant trafficking of endosome:vacuole cargo upon expression. We employed both GFP-tagged carboxypeptidase S (CPS), a membrane-bound vacuolar protease [57], and GFP-tagged Sna3p, a transmembrane ubiquitin ligase adapter protein involved in vacuole protein sorting [58]. It is known that both proteins cargo traffic to the yeast vacuole via an endosomal/multivesicular body (MVB) intermediate compartment [31, 57]. Under normal growth conditions of wild type strains, both GFP-CPS and Sna3-GFP localize to the yeast vacuole regardless of the presence of β-estradiol (Fig 3A and 3B). VPS33 encodes for the Sec1/Munc18-family protein, which is a member of the multisubunit class C core vacuole/endosome tethering (CORVET) complex required for endosome:endosome fusion events. It is also a member of the multisubunit homotypic fusion and vacuole protein sorting (HOPS) complex required for both homotypic vacuole and late endosome:vacuole fusion [59-62]. In the strain lacking VPS33, both GFP-CPS and Sna3-GFP were found in aberrant intracellular compartments, as expected (Fig 3A and 3B, vps33Δ).

Fig 3

wBm0152 expression inhibits endosomal traffic to the yeast vacuole.

BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods), and harboring (A) GFP-CPS plasmid or (B) Sna3-GFP plasmid in addition to pYES2/NT A control plasmid or pYES2/NT A wBm0152 were grown to saturation in CSM-lysine-uracil medium. Cells were subcultured to CSM-lysine-uracil with and without the addition of 1 μM β-estradiol and grown for 6h at 30°C. Cells were stained with 10 μM FM4-64 dye for 20 minutes at 30°C and chased 1.5h in CSM-lysine-uracil medium at 30°C. Cells were visualized, and representative images are shown. Bar = 3 μ. Inset in (A) magnifies the cell denoted by the arrow to better visualize the GFP-CPS trafficking defect in cells expressing wBm0152; images are representative of two independent experiments.

Upon induction with β-estradiol, most putative wBm effector proteins did not alter the normal vacuolar localization of either GFP-CPS (S4 Fig) or Sna3-GFP (S5 Fig), suggesting that expression of these proteins does not strongly alter normal endosome:vacuole traffic in yeast. Expression of the predicted peptidoglycan associated lipoprotein (PAL) wBm0152, however, prevented both GFP-CPS and Sna3-GFP from accumulating within the vacuole lumen (Fig 3A and 3B, respectively), suggesting that this protein is capable of modulating the MVB:vacuole fusion event, and may play some role in either preventing the degradation of Wolbachia in nematode lysosomes, or supporting the synthesis of the Wolbachia-containing vacuole from post-Golgi membranes.

Expression of wBm0076 disrupts the yeast actin cytoskeleton

Expression of the wBm0076 gene is lethal to yeast (Fig 1). Analysis of the protein sequence via BLAST searching identifies orthologs in closely related bacteria, including: Wolbachia of Onchocerca volvulus (WP_025264345.1), Wolbachia of Onchocerca ochengi (WP_014868845.1), Wolbachia of Drosophila melanogaster (AAS14498), and RickA from Rickettsia conorii (WP_041471735.1). Wbm0076 is a protein of 392 amino acids in length, containing a putative WH2 (verprolin, V) domain, an ~18 amino acid actin-binding motif, a central α-helical domain (C), an acidic region (A), and 2 proline-rich domains with homology to proteins of the eukaryotic neural Wiskott-Aldrich syndrome protein (N-WASP/WAS) family [63], including the bacterial RickA and actin-polymerizing ActA from Listeria monocytogenes [64] (Fig 4A and 4B). N-WASP proteins are known to bind actin monomers and the conserved eukaryotic Arp2/3 complex to stimulate branched actin filament formation in an Arp2/3-dependent manner [65, 66]. In addition, the RickA protein was determined to polymerize actin ‘comet tails’ on the surface of Rickettsia to aid in intracellular motility during infection, and to polymerize branched actin structures in a Arp2/3-dependent manner in vitro [67]. Based on the homology of wBm0076 protein to both eukaryotic and bacterial proteins known to modulate actin structures, we hypothesized that wBm0076 was likely modifying actin structures in yeast upon expression.

Fig 4

wBm0076 is a WAS(p)-family protein.

Domain structure of wBm0076 orthologs, including: human N-WASp (hu N-WASp; BAA20128.1), Brugia malayi WSP-1 (Bma WSP-1; CRZ22528.1), S. cerevisiae Las17p (Sce Las17p, NP_014824.1), wBm0076 (WP_011256278.1), wMel WD_0811 (WD_0811; AAS14498), Rickettsia conorii RickA (Rc RickA; WP_041471735.1), and Listeria monocygenes ActA (Lm ActA; ABC40914.2). Domains including poly-proline motifs (PP), the WH2/verprolin-like (V), central α-helical (C), and acidic regions (A) are depicted. (B) Sequence alignment of the conserved VCA domains of the proteins in (A). Completely conserved residues are highlighted with a black box and residues conserved in more than half of the sequences are denoted with an asterisk. Only the first WH2 domains of hu N-WASP and Bma WSP-1 are aligned for simplicity.

The yeast actin cytoskeleton consists of two main conformations: tight bundles of many actin filaments (cables) [68], and concentrated patches of highly branched actin that accumulate near the periphery of the yeast cell; these cortical actin patches are linked to newly-formed endocytic vesicles, thus enabling their retrograde trafficking [69]. To visualize cortical actin patches in the presence of wBm0076, we used a yeast strain expressing RFP-tagged Abp1p. Abp1p is an actin filament binding protein which interacts with Arp2/3 to stimulate Arp2/3 activity [70], and regulates the capping of barbed-end actin filaments when bound to Aim3p [71]. In a yeast strain harboring a vector control, Abp1-RFP localizes to several cortical punctae per cell, as expected (Fig 5A). The presence of the wBm0076 plasmid under non-inducing conditions does not alter the punctate localization of Abp1-RFP (Fig 5A). Upon addition of 1 μM β-estradiol, however, yeast expressing wBm0076 display an increase in both in the number of Abp1-RFP punctae and cell size after 6 hours, when compared to the vector control strain (Fig 5A and 5B). While the overall number of cortical actin patches in wBm0076-expressing strains did not appear to increase when correcting for the estimated cell volume (Fig 5B, right panel), it is not yet clear if the increased number of cortical actin patches increased the cell volume, or if the increased cell volume induced a corresponding increase in cortical actin patches. When left in inducing conditions overnight, however, Abp1-RFP-containing punctae in the wBm0076 strain are no longer detected and cells appear to lyse through a local disruption in the cell wall.

Fig 5

wBm0076 expression alters yeast actin dynamics in vivo.

S288C yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods), expressing (A) Abp1-RFP and harboring either the pYES2/NT A control plasmid or pYES2/NT A wBm0076 were grown to saturation in CSM-uracil medium, subcultured to CSM-uracil, with or without 1 μM β-estradiol, for 6h at 30°C. Cells were harvested at indicated time points and visualized. Bar = 5 μ. (B-D) The average ± standard deviation Abp1-containing punctae are plotted per cell (B), cell diameter (C), and estimated cell volume / punctae (D, assume cell is a sphere during volume calculation) were measured from cells grown in (A), and ≥100 cells each per three independent experiments were measured; ns = P > 0.05 (not significant); (*) = P ≤ 0.05; (**) = P ≤ 0.01; (****) = P ≤ 0.0001. (E) As in (A), except yeast cells harbored Abp140-3xGFP instead of Abp1-RFP.

To visualize actin cables in vivo, we used GFP-tagged Abp140p, the actin-binding tRNA methyltransferase responsible for the addition of 3-methylcytidine in the anticodon loop of yeast tRNAs [72]. The N-terminal actin-binding domain of Abp140p is known to localize to both actin cable and cortical actin patches; control yeast expressing Abp140-3xGFP clearly mark actin cables, with a higher local concentration around the bud neck, as expected (Fig 5C). Upon a 6-hour induction with 1 μM β-estradiol, cells harboring the wBm0076 plasmid no longer show elongated actin cables, but rather a high concentration of discrete punctae similar to those seen in Abp1-RFP cells (Fig 5C); extended incubations in β-estradiol confirm the ability of wBm0076 to lyse yeast cells (Fig 5C). Taken together, expression of wBm0076 in yeast leads to an extensive disruption of the yeast actin cable network, an increase in the number of cortical patches containing branched actin, and cell lysis.

Deletion of the Arp2/3-activating protein ABP1 provides resistance to wBm0076 expression

The isolation of yeast mutant derivatives that are no longer sensitive to wBm0076 expression could provide insight into not only the cellular processes modulated by wBm0076, but also help elucidate the protein targets or biochemical activity of wBm0076. Because wBm0076 protein displays homology to WAS(p)-family proteins and disrupts the yeast actin cytoskeleton in vivo, we hypothesized that deletions of genes encoding for proteins known to be involved in regulating actin dynamics may provide some resistance to wBm0076 expression in yeast.

In yeast, Las17p is the sole WAS(p)-family protein present that interacts with a large number of proteins via its conserved WH1 and proline rich domains and plays an important role in regulating endocytosis [73-76]. We therefore assayed the toxicity of wBm0076 expression in thirty-one single gene deletion backgrounds, each of which were selected by their ability to produce proteins that either directly interact with Las17p, or otherwise regulate actin dynamics. Deletion strains harboring wBm0076 or vector only controls were serially diluted and grown on minimal media containing 1% raffinose plus either 2% glucose (uninduced condition) or 2% galactose (induced). As expected, wBm0076 was toxic to yeast upon galactose induction (Fig 6A; WT/0076). We noted that some of these deletion strains (arf1Δ, sho1Δ, vrp1Δ, las17Δ, bbc1Δ, myo5Δ) grew very poorly on media containing galactose (S6 Fig), even in the absence of wBm0076, and we therefore could not easily interpret these results. Importantly, expression of wBm0076 did not restore the ability of the las17Δ strain to grow on galactose (S6 Fig); it is therefore unlikely that wBm0076 directly complements Las17p activity in Saccharomyces. However, of the thirty-one deletion strains tested, only abp1Δ strains displayed resistance to expression of wBm0076 on media containing galactose (Figs 6A and S6; abp1Δ/0076).

Fig 6

Deletion of ABP1 reduces the toxicity of wBm0076 expression.

(A) BY4742 yeast strains deleted for the indicated gene and harboring either pYES2/NT A or pYES2/NT A wBm0076 (0076) were grown overnight in CSM medium lacking uracil. Cultures were diluted to an OD600 = 1.0 in sterile 0.9% NaCl, then spotted in 10-fold dilutions on plates containing 1% raffinose and either 2% glucose or 2% galactose to induce wBm0076 expression. Plates were incubated for 72 h at 30°C and imaged. (B) BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters and harboring either pYES2/NT A (pYES), pYES2/NT A wBm0076 (0076), or pYES2/NT A wBm0076-mRuby2 (0076-mRuby) were grown spotted as in (A), except that plates either contained or lacked 1 μM β-estradiol to induce wBm0076 expression. Plates were incubated for 48 h and imaged. (C) Yeast strains harboring Abp140-3xGFP and pYES2/NT A wBm0076-mRuby2 were grown overnight in CSM medium lacking uracil and subcultured to fresh CSM-Ura containing 1 μM β-estradiol. After 5 h outgrowth at 30°C, cells were harvested and visualized. Images are representative of three independent experiments; bar = 5μ.

Because Abp1p localizes to cortical actin patches and its deletion limits the toxicity of wBm0076 expression, we considered the possibility that wBm0076 would localize to cortical actin patches to regulate their formation in vivo. Accordingly, we constructed a wBm0076-mRuby expression construct that still maintained toxicity (and thus, activity) upon expression (Fig 6B). In a yeast strain harboring Abp140-3xGFP, which also localizes to cortical actin patches (Fig 4B), expression of wBm0076-mRuby induces the formation of cortical actin patches and concomitant loss of actin cables previously observed (Fig 6C, compare to Fig 4B). Additionally, we observe colocalization of wBm0076 with these Abp140-3xGFP-positive punctae, suggesting that wBm0076 localizes to cortical actin patches to regulate actin dynamics and may do so through direct interactions with yeast proteins directly involved in actin patch formation.

Based upon the finding that abp1Δ strains limit the toxicity of wBm0076 expression, wBm0076-mRuby localizes to cortical actin patches, and the knowledge that Abp1p is intimately involved in activating Arp2/3 during endocytosis, we find it likely that the toxicity of wBm0076 in yeast is directly related to its ability to alter actin dynamics in vivo. Given the conservation of the regulation of actin dynamics across eukaryotes, these data suggest that wBm0076 may be important for the reorganization of actin structures in the nematode.

Discussion

Though some progress has been made in understanding the Wolbachia:host relationship at the molecular level, much remains unknown about the mechanisms by which Wolbachia persists in its nematode or insect hosts. The use of S. cerevisiae as a highly tractable surrogate eukaryotic host provides an avenue to rapidly screen wBm putative T4SS effectors that may manipulate eukaryotic cell physiology and membrane trafficking pathways likely required for the persistence of Wolbachia in host cells. Historically, the S. cerevisiae model has successfully identified virulence factors from human pathogens that, upon expression in yeast, induce a growth defect and alter eukaryotic physiology, as is the case with the identification of secreted proteins from a variety of pathogenic bacteria, including Chlamydia trachomatis, Legionella pneumophila, Pseudomonas aeruginosa, and Salmonella enterica [27, 29, 77, 78]. Recently, S. cerevisiae was used to identify fourteen candidate effectors of the Wolbachia endosymbiont of Drosophila melanogaster, wMel, with each candidate effector inducing growth defects in S. cerevisiae [79]. Collectively, these data have verified that S. cerevisiae is a valuable tool to evaluate the molecular function of a number of Wolbachia-secreted proteins. Therefore, we have employed S. cerevisiae to screen 47 predicted wBm effectors to identify those that manipulate eukaryotic physiology. Due to the ability of some of these proteins to induce growth defects and alter the cellular processes in S. cerevisiae (Table 2), we propose that they may function as secreted effectors of wBm and may be necessary for germline cell invasion and persistence of wBm in its nematode host. Strikingly, two of these proteins–wBm0076 and wBm0152—are particularly strong modulators of yeast physiology.

Table 2

Summary of aberrant yeast phenotypes observed in this study.

Growtha	Growth with Zn/caffeinea,b	Vacuole morphologyc	Endosome:vacuole cargo trafficking
wBm0014 (strong)	wBm0057 (strong)	wBm0076	wBm0152
wBm0032 (strong)	wBm0064 (strong)	wBm0114
wBm0044 (strong)	wBm0213 (strong)	wBm0152
wBm0070 (weak)	wBm0277 (strong)	wBm0447
wBm0076 (strong)	wBm0287 (strong)
wBm0114 (weak)	wBm0447 (strong)
wBm0222 (weak)
wBm0284 (strong)

aInhibition of cell growth on indicated medium denoted as strong (cell titer three orders of magnitude lower in a serial dilution) or weak (two orders of magnitude), when compared to no induction controls.

bInhibition of cell growth on medium containing ZnCl2 and caffeine (Fig 1B), but not on medium lacking these additions.

cProteins capable of producing aberrant vacuolar morphologies in ≥30% of the cells examined.

wBm0152, conserved across Wolbachia isolates of diverse hosts, is a predicted peptidoglycan associated lipoprotein (PAL) that was previously identified as a component of the wBm secretome [80]. PAL proteins are thought to be intimately involved in regulating Gram-negative cell envelope integrity via interactions with both periplasmic peptidoglycan and the outer membrane [81]. Relatively recent work has found, however, that PAL proteins can exist in a so-called ‘dual conformation,’ where a population of these proteins are surface-exposed [82]. This particular conformation creates the possibility that these proteins may also be involved in extracellular contacts. Indeed, previous work by Melnikow and coworkers have shown that wBm0152 was found to bind actin and was hypothesized to be an important link between Wolbachia cells and host actin/tubulin during endosymbiosis, even though direct binding between wBm0152 and actin in solution was not directly assayed [17]. In our study, we find that expression of wBm0152 in S. cerevisiae induces a striking endosomal trafficking defect, as seen by aberrant “class E” vacuole morphologies, and in the mislocalization of the endosomal cargo proteins (CPS and Sna3p) to an enlarged prevacuolar compartment. This phenotype is reminiscent of yeast strains defective in endosomal sorting complex required for transport (ESCRT) complex proteins, which are involved in vacuolar protein sorting, Golgi protein recycling, and membrane repair/remodeling pathways [56]. Because the membrane comprising Wolbachia replicative vesicles in Drosophila appear to originate from the host Golgi [38], and because wBm0152 seems to delay or block the delivery of Golgi-derived cargo to the vacuole in yeast, it is tempting to speculate that wBm0152 may play a role in either the biogenesis of the wBm membrane-bound compartment, or preventing the degradation of the Wolbachia-containing vacuole by interfering with its fusion with the nematode lysosome.

wBm0076 is a WAS family protein that likely promotes actin cytoskeletal rearrangements and has sequence homology to RickA, a protein expressed by Rickettsia conorii that is known to activate Arp2/3 and drive the polymerization of host actin to support bacterial intracellular motility during infection, similar to that of Listeria monocytogenes [67, 83]. In eukaryotes, Arp2/3 is a protein complex that is associated with cortical actin patches and aids in actin arrangement for the nucleation of branched actin filaments that drive processes like endocytosis and cellular motility [65, 66]. In yeast, it has been shown that the conserved Abp1p protein interacts with both actin and Arp2/3 complex to initialize actin polymerization in S. cerevisiae [70, 84], specifically localizing to cortical actin patches and initiating endocytic vesicle formation and actin patch disassembly through its recruitment of other kinases [85, 86]. Our data shows wBm0076 expression induces yeast cell lysis, in a manner similar to strains lacking F-actin [87]. Therefore, wBm0076 may induce yeast cell lysis through the elimination of F-actin structures by either enhanced turnover of F-actin, or through the enhanced polymerization of branched actin structures and the concomitant decrease in monomeric actin required to maintain F-actin. Due to the fact that we have also identified a genetic interaction between Abp1p and wBm0076, it is possible limiting excess Arp2/3 activation (and hence, excessive branched actin formation) by both Abp1p and wBm0076 independently, is sufficient to limit cell lysis upon wBm0076 expression. Alternatively, Abp1p may be the direct target of wBm0076 in vivo to regulate actin polymerization since the ability of Abp1 and N-WASP to directly interact to stimulate Arp2/3 activity has previously been shown in mammalian neuronal cells [88]. Interestingly, it has been shown that Wolbachia induces a reorganization of cellular F-actin structures during invasion of the germline in female nematodes, thus ensuring the vertical transmission of Wolbachia to offspring [89]. It is therefore possible that wBm0076 plays a vital role in this particular process, as actin “comet tails” have not, as of yet, been detected to play a role in the mobility of Wolbachia in B. malayi [90]; B. malayi contains the Abp1p homolog Bm4914 (20% end-to-end identity; 32% similar to yeast Abp1p), which could serve as the target of wBm0076 in the nematode. Interestingly, profilin mutants of Drosophila are incapable of efficient maternal passage of Wolbachia to offspring [90], and the Wolbachia parasite of Drosophila (wMel) produces at least one other putative secreted effector, WD0830, capable of bundling actin filaments in vivo and in vitro via its α-synuclein domain [91]. It is also important to note that the wBm0076 homolog in wMel, WD0811, while containing a clear WH2 domain, lacks the conserved C and A domains usually found in WASP-family proteins (Fig 4); WD0811 therefore may manipulate actin polymerization in Drosophila in a manner distinct from that of wBm0076. Additionally, WD0811 has been previously been shown to be translocated through a surrogate T4SS [92], thus providing additional evidence to support the hypothesis that wBm0076 in an authentic T4SS effector of wBm. The data presented in this study provide additional mounting evidence that Wolbachia modulates host actin for successful transmission to offspring.

Recently, researchers have carried out a similar screen of eighty-four putative wMel effectors in yeast [79]. Of those eighty-four, twelve (based on sequence homology) were also screened in our study and three (wBm0076/WD0811; 36% identical, wBm0287/WD0566; 40% identical, wBm0447/WD0438; 44% identical) were moderately to severely toxic to yeast cells as determined by induced growth defects in the presence or absence of stressors. Notably, both WD0811 and WD0438 were identified by Rice et al. as candidate effectors of wMel due to induced growth defects in yeast [79]. wMel proteins are more similar to Wolbachia endosymbionts of arthropods within its same or similar clade than it is to wBm [79]; wBm and wMel have 696 orthologous proteins and similar metabolic capacities that likely differ due to host and the relationship Wolbachia maintains with such host [14]. The use of yeast to identify putative effectors from Wolbachia endosymbionts of distinct hosts, and specifically homologous putative effectors, confirms the validity of such a screen for use as an initial characterization of protein functions in a genetically intractable system. In this work, we have furthered the characterization of two putative effectors of wBm that appear to have drastic effects on yeast cell biology, thus providing additional insight into their physiological function to support the intracellular survival of Wolbachia in B. malayi. Although we find individual wBm proteins can manipulate yeast growth and trafficking pathways, it is most likely a concerted effort of many secreted proteins is required to maintain both the intracellular colonization of wBm and survival of B. malayi. Our studies demonstrate the successful use of S. cerevisiae as a host system to identify putative effectors that may shed light on their roles in manipulating the cellular physiology of B. malayi to help maintain this essential relationship.

Materials and methods

Yeast strains and plasmid constructions

For all growth analyses and microscopic visualization studies, yeast strain BY4742 (MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) was used for all studies. In order to create yeast strains that activate GAL1 promoters via the addition of β-estradiol, strains were transformed with linearized pAGL (a gift from Dr. Daniel Gottschling, University of Washington), which introduces the gene encoding for the Gal4-estrogen receptor-VP16 (GEV) chimeric protein into the leu2Δ0 locus [93]. S288C yeast strains expressing either Abp1p-RFP or Abp140-3xGFP were a kind gift from Dr. Bruce Goode (Brandeis University).

Hypothetical or eukaryotic-domain containing Wolbachia proteins such as ankyrin repeats or Ser/Thr kinase motifs were searched for using BLAST against the NCBI database. Proteins highly related to eukaryotes rather than bacteria were selected for this study. Coiled coil domain containing proteins were predicted using an online tool at: https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=npsa_lupas.html [94]. Rickettsia sibirica proteins that interact with different components of its T4SS identified in a bacterial two-hybrid system [36] and proteins that potentially co-transcribed with components of T4SS were used to as query to search using TBLASTN against the wBm genome to identify the potential homolog in filarial Wolbachia. Wolbachia proteins previously identified in excretory-secretory (ES) products from B. malayi adult worms and microfilaria [37] were also selected. Domain structures of wBm0076 and orthologs were determined with HHpred (MPI bioinformatics Toolkit, https://toolkit.tuebingen.mpg.de/#/tools/hhpred).

Wolbachia genes encoding for putative T4SS-secreted proteins were cloned into the galactose-inducible yeast expression vector, pYES2/NT A via the USER (uracil-specific excision reagent) cloning system developed at NEB. Briefly, gene-specific PCR primers were designed to clone the predicted wBm open reading frame into the vector expressing an amino-terminal Xpress epitope tag; these primers (S1 Table) contained a linker complementary to pYES2/NT A, with an additional uracil, and were used to amplify the target gene with PfuTurbo Cx hotstart DNA polymerase (Stratagene). PCR products were then treated with the USER Enzyme (New England Biolabs) to create unique 3´ single-stranded extensions which anneal and ligate to linearized pYES2/NT A, which had been previously amplified with vector-specific primers containing a complementary linker to the insert gene of interest.

In order to utilize the URA3 pGO45 and pMM134 plasmids, which express GFP-CPS or Sna3p-GFP, respectively [57, 95], for yeast endosome:vacuole transport studies in the presence of another URA3 yeast expression plasmid harboring the putative wBm effector, pGO45 and pMM134 were first converted from URA3 to LYS2 via homologous recombination methods by digesting pM2660 (ATCC) with HindIII [96], and transforming yeast strains harboring either pGO45 or pMM134 with this digest. Growth was selected on CSM medium + 2% glucose lacking lysine. Colonies from these plates were screened for lack of growth on CSM + 2% glucose medium without uracil. Plasmids from phenotypically correct strains were isolated, reintroduced into the appropriate yeast background, and confirmed for function via fluorescence microscopy.

To create the wBm0076-mRuby expressing pYES2NTA plasmid, the yomRuby2 gene was amplified from the plasmid pFA6a-link-yomRuby2-SpHis5 [97] using primers 5’- AGCTTTTCTTATAAAACAATTGATGGTGTCCAAAGGAGAGGAG and 5’- AGGGATAGGCTTAGCTGCAATTTACTTATACAATTCATCCATA, containing homology to both the C-terminus of the wBm0076 gene and the pYES2NTA-wBm0076 vector. BY4742 was co-transformed with pYES2NTA-wBm0076, previously digested with PmeI, and the mRuby2 amplicon and were plated to CSM-uracil to select for gap-repaired plasmids. Transformants were screened for red fluorescence via microscopy, and the resultant plasmid was purified and sequenced for confirmation (Eton Bioscience Inc).

Cell lysis and western blotting

β-estradiol responsive yeast strains harboring GAL-inducible pYES2/NT A control plasmid or pYES2/NT A containing an individual wBm open reading frame were grown to saturation in CSM-uracil medium, subcultured to CSM-uracil supplemented with 1 μM β-estradiol and grown for 6h at 30°C. 5.0 OD600 units were harvested and protein was extracted in lysis buffer (0.1 M NaOH, 2% SDS, 2% β-mercaptoethanol, 0.05 M EDTA-NaOH pH 8.0) and boiled 5 minutes prior to the addition of 0.1M acetic acid. Equal volumes of extracts were separated on 13% SDS-PAGE gels and immunoblotted for wBm proteins using the commercially-available anti-Xpress monoclonal antibody (1:1000, Thermo Fisher Scientific) for 1 h, using blocking buffer containing 0.5% Tween and 5% dry non-fat milk. Goat anti-mouse IgG secondary antibody conjugated to HRP (1:20000, Thermo Fisher Scientific) was applied to the membrane for 1 h, washed, and developed with SuperSignal West Pico PLUS chemiluminescent substrate (Thermo Scientific).

Microscopy

For evaluation of vacuole morphologies, β-estradiol responsive yeast strains harboring either control pYES2/NT A vector or pYES2/NT A WSPs were grown to saturation in selective medium at 30°C, subcultured to fresh media with or without 1 μM β-estradiol and grown for an additional 6 hours. After 6 hours, the entire culture was harvested via centrifugation, suspended in 50 μL CSM lacking uracil, and FM4-64 was added to 3 μM. Cells were incubated at 30°C for 20 minutes, harvested, washed, then suspended in 5 mL fresh medium. This culture was incubated for an additional 90 minutes at 30°C, harvested, washed, and suspended in 25 μL CSM medium lacking uracil. Cell suspensions were mounted to slides that had been pre-treated with a 1:1 mixture of polylysine (10% w/v):concanavalin A (2 mg/ml) solution. Cells were visualized using a Nikon Ti-U fluorescence microscope, and images were processed using the Fiji software package [98, 99].

Statistical analysis

Statistical analysis was performed within the Prism software package (GraphPad Software, v. 6.0b). Column statistics were performed via a 1-way ANOVA Repeated Measures test and Holm-Bonferroni post-test. Where noted in figures, ns = P > 0.05 (not significant); (*) = P ≤ 0.05; (**) = P ≤ 0.01; (****) = P ≤ 0.0001.

Primers used in this study.

(DOCX)

Click here for additional data file.

Analysis of wBm effector protein expression in yeast.

BY4742 yeast strains genetically modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods) were assayed for individual putative wBm effector expression via anti-Xpress immunoblot (Materials and Methods). Images shown are representative of three independent replicates.

(PDF)

Click here for additional data file.

Most individual wBm effector genes do not inhibit yeast growth upon expression.

BY4742 yeast strains genetically modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods) and harboring the GAL-inducible control plasmid pYES2/NT A, or pYES2/NT A containing the specified wBm open reading frame were grown to saturation in CSM-uracil medium containing 2% glucose, and each culture was diluted to OD600 = 1.0 in sterile 0.9% NaCl. 10-fold serial dilutions were spotted onto CSM-uracil containing 2% glucose with and without 1 μM β-estradiol or 7.5 mM ZnCl2/5 mM caffeine. Plates were incubated for 48 or 72 h at 30°C; results are representative of three independent experiments.

(PDF)

Click here for additional data file.

Effects of wBm effector gene expression on yeast vacuole morphology.

BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods), and harboring GAL-inducible pYES2/NT A control plasmid or pYES2/NT A containing an individual wBm open reading frame were grown to saturation in CSM-uracil medium, subcultured to CSM-uracil supplemented with 1 μM β-estradiol, and grown for 6h at 30°C. Cells were stained for 20 minutes with 10 μM FM4-64 at 30°C, followed by a 1.5 h chase in CSM-uracil at 30°C. Cells were visualized and representative crops from three independent experiments were generated; bar = 5 μ.

(PDF)

Click here for additional data file.

CPS-GFP is generally not mislocalized in yeast strains expressing wBm genes.

BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods), and harboring pGO45 GFP-CPS plasmid in addition to pYES2/NT A control plasmid or pYES2/NT A harboring an individual wBm open reading frame were grown to saturation in CSM-lysine-uracil medium. Cells were subcultured to CSM-lysine-uracil with 1 μM β-estradiol and grown for 6 h at 30°C. Cells were stained with 10 μM FM46-4 for 20 minutes at 30°C, chased for 1.5 h in CSM-lysine-uracil medium at 30°C, then visualized. Representative crops from two independent experiments are shown; bar = 3 μ.

(PDF)

Click here for additional data file.

GFP-Sna3 is generally not mislocalized in yeast strains expressing wBm genes.

BY4742 yeast strains modified with GEV for β-estradiol-dependent induction of GAL promoters (Materials and Methods) and harboring the Sna3-GFP plasmid in addition to pYES2/NT A control plasmid, or pYES2/NT A harboring an individual wBm open reading frame, were grown to saturation in CSM-lysine-uracil medium. Cells were subcultured to CSM-lysine-uracil with 1 μM β-estradiol and grown for 6 h at 30°C. Cells were stained with 10 μM FM46-4 for 20 minutes at 30°C, chased for 1.5 h in CSM-lysine-uracil medium at 30°C, then visualized. Representative images from two independent experiments are shown; bar = 3 μ.

(PDF)

Click here for additional data file.

Deletions of genes involved in yeast actin dynamics do not reduce the toxicity of wBm0076 expression.

BY4742 yeast strains deleted for the indicated gene and harboring either pYES2/NT A or pYES2/NT A wBm0076 (0076) were grown overnight in CSM medium lacking uracil. Cultures were diluted to an OD600 = 1.0 in sterile 0.9% NaCl, then spotted in 10-fold dilutions on plates containing 1% raffinose and either 2% glucose or 2% galactose to induce WBM0076 expression. Plates were incubated for 72 h at 30°C and imaged; results are representative of three independent experiments.

(PDF)

Click here for additional data file.