In situ imaging of bacterial membrane projections and associated protein complexes using electron cryo-tomography

The ability to produce membrane projections in the form of tubular membrane extensions (MEs) and membrane vesicles (MVs) is a widespread phenomenon among bacteria. Despite this, our knowledge of the ultrastructure of these extensions and their associated protein complexes remains limited. Here, we surveyed the ultrastructure and formation of MEs and MVs, and their associated protein complexes, in tens of thousands of electron cryo-tomograms of ∼ 90 bacterial species that we have collected for various projects over the past 15 years (Jensen lab database), in addition to data generated in the Briegel lab. We identified MEs and MVs in 13 species and classified several major ultrastructures: 1) tubes with a uniform diameter (with or without an internal scaffold), 2) tubes with irregular diameter, 3) tubes with a vesicular dilation at their tip, 4) pearling tubes, 5) connected chains of vesicles (with or without neck-like connectors), 6) budding vesicles and nanopods. We also identified several protein complexes associated with these MEs and MVs which were distributed either randomly or exclusively at the tip. These complexes include a secretin-like structure and a novel crown-shaped structure observed primarily in vesicles from lysed cells. In total, this work helps to characterize the diversity of bacterial membrane projections and lays the groundwork for future research in this field.


Introduction 49
Membrane extensions and vesicles (henceforth referred to as MEs and MVs) have been described 50 in many types of bacteria. They are best characterized in diderms, where they stem mainly from 51 the outer membrane (OM; we thus refer to OMEs and OMVs) and perform a variety of functions 52 [1][2][3][4]. For example, the OMEs of Shewanella oneidensis (aka nanowires) are involved in 53 extracellular electron transfer [5,6]. The OM tubes of Myxococcus xanthus are involved in the 54 intra-species transfer of periplasmic and OM-associated material between different cells that is 55 essential for the complex social behavior of this species [7][8][9]. The OMVs of Vibrio cholerae act 56 as a defense mechanism, helping the bacterium circumvent phage infection [10]. A marine 57 Flavobacterium affiliated with the genus Formosa (strain Hel3_A1_48) extrudes membrane tubes 58 and vesicles that contain the type IX secretion system and digestive enzymes [11]. OMVs often 59 function in pathogenesis. The OM blebs and vesicles of Flavobacterium psychrophilum have 60 proteolytic activities that help release nutrients from the environment and impede the host immune 61 system [12]. The OMVs of Francisella novicida contain virulence factors, suggesting they are 62 involved in pathogenesis [13]. Similarly, the virulence of Flavobacterium columnare is associated 63 with the secretion of OMVs [14], and membrane tubes and secreted vesicles have been observed 64 in other, human pathogens like Helicobacter pylori and Vibrio vulnificus [15,16]. an inner scaffold and lateral ports [15]. V. vulnificus produces tubes from which vesicles ultimately 78 pinch off by biopearling, forming a regular concentric pattern surrounding the cell [16]. Cells with 79 an external surface layer (S-layer) can produce structures known as "nanopods," which consist of 80 membrane vesicles inside a sheath of S-layer. These have been reported in the soil-residing 81 bacterium Delftia sp.  and archaea of the order Thermococcales [26]. Finally, some 82 diderms produce DNA-containing MVs consisting of both inner and outer membranes (see Ref. 83 [4] and references therein). 84 85 Different models have been proposed for how MEs and MVs form. In diderms, membrane 86 blebbing may occur due to changes in the periplasmic turgor pressure, lipopolysaccharide 87 repulsion or alterations in the contacts between the OM and the peptidoglycan cell wall [4]. Chains 88 of interconnected vesicles are often observed, either as a result of direct vesicular budding from 89 the OM or due to biopearling of membrane tubes [6,11]. Formation of tubes is thought to be a 90 stabilizing factor as it results in smaller vesicles, with tubes pearling into distal chains of vesicles 91 that eventually disconnect [27]. Other extensions may be formed by dedicated machinery. 92 Interestingly, nanotubes involved in cytoplasmic exchange have been reported to be dependent on 93 a conserved set of proteins involved in assembly of the flagellar motor known as the type III 94 To understand what membrane extensions exist in bacterial cells and how they might form, we 114 undertook a survey of ~90 bacterial species, drawing on a database of tens of thousands of electron 115 cryo-tomograms of intact cells collected by our group for various projects over the past 15 years 116 [31,32], in addition to data generated in the Briegel lab. Our survey revealed membrane projections 117 in 13 bacterial species. These projections took various forms: 1) tubes with a uniform diameter 118 and with an internal scaffold, 2) tubes with a uniform diameter and without a clear internal scaffold, 119 3) tubes with a vesicular dilation at their tip (teardrop-like extensions), 4) tubes with irregular 120 diameter or pearling tubes, 5) interconnected chains of vesicles with uniform neck-like connectors, 121 6) budding or detached OMVs, and 7) nanopods. We also identified protein complexes associated 122 with MEs and MVs in these species. These complexes were either randomly distributed on the 123 MEs and MVs or exhibited a preferred localization at their tip.

Results: 141
We examined tens of thousands of electron cryo-tomograms of ~ 90 bacterial species collected in 142 the Jensen lab for various projects over the past 15 years together with tomograms collected in the 143 Briegel lab. Most cells were intact, but some had naturally lysed. Note that we make this 144 classification based on the cells' appearance in tomograms; intact cells have an unbroken cell 145 envelope, uniform periplasmic width, and consistently dense cytoplasm. In addition to cryo-146 tomograms of cells, this dataset also included naturally-shed vesicles purified from S oneidensis. 147 In all, we identified OMEs and OMVs in 13 bacterial species (summarized in Table S1). 148

I-The diverse forms of bacterial membrane structures 150
Based on their features, we classified membrane projections into the following categories: 1) 151 tubular extensions with a uniform diameter and with an internal scaffold ( Fig. 1 a & b); 2) tubular 152 extensions with a uniform diameter and without a clear internal scaffold (Fig. 1 c-g ); 3) tubular 153 extensions with a vesicular dilation at the tip (a teardrop-like structure) and irregular dark densities 154 inside (Fig. 1h); 4) tubular extensions with irregular diameter or pearling tubes (Fig. 2 Table S1 for a summary of these 159 observations. 160 161 Scaffolded membrane tubes were observed only in H. pylori and had a uniform diameter of 40 nm. 162 The H. pylori strain imaged (fliP * ) contains a naturally-occurring point mutation that disrupts the 163 function of FliP, the platform upon which other CORE proteins assemble [33][34][35]. In addition, the 164 dataset contained other mutants in this fliP * background including additional CORE proteins (ΔfliO 165 and ΔfliQ), flagellar basal body proteins (ΔfliM and ΔfliG), and the tyrosine kinase required for 166 expression of the class II flagellar genes (ΔflgS) [36] (Figs. 1 a-b and 4). This suggests that the H. 167 pylori membrane tubes are unrelated to the CORE-dependent nanotubes that mediate cytoplasmic 168 exchange in B. subtilis and other species [18,24]. 169 170 Previously, H. pylori tubes were described as forming in the presence of eukaryotic host cells [15]. 171 Here, however, we observed tubes on H. pylori grown on agar plates in the absence of eukaryotic 172 cells, suggesting that they also form in the absence of host cells. We observed some differences, 173 though, from the tubes formed in the presence of host cells: the tube ends were closed, no clear 174 lateral ports were seen, and the tubes were usually straight. While some of these tubes extended 175 more than 0.5 µm, we never observed pearling. However, in some tubes, the internal scaffold did 176 not extend all the way to the tip, and its absence caused the tube to dilate (from 40 nm in the 177 presence of the scaffold to 66 nm in its absence, see Fig. 4f). In some cases we also observed tubes 178 stemming from vesicles resulting from cell lysis (Figs. 4f and S1). 179

180
In Flavobacterium anhuiense and Chitinophaga pinensis, which are both endophytic species 181 extracted from sugar beet roots, in addition to tubes with irregular diameter and OMVs, tubular 182 extensions with a uniform diameter and a vesicular dilation (teardrop-like structure) were observed 183 stemming from the sides of the cell (Fig. 1h). Interestingly, irregular dark densities were observed 184 inside these teardrop-like extensions (Fig. 1h). Chains of vesicles connected by neck-like bridges 185 were similarly observed in a single species: Borrelia burgdorferi. The bridges were consistently 186 ~14 nm in length and ~8 nm in width. Where chains were seen attached to the outer membrane, a 187 neck-like connection was present at the budding site (Fig. 2h). Vesicles in each chain were of a 188 uniform size, usually 35-40 nm wide (e.g. Fig. 2i), but occasionally larger (e.g. Fig. 2h). 189

190
When both tubes and vesicles were observed in the same species, the tubes generally had a more 191 uniform diameter than the vesicles, which were of variable sizes and often had larger diameters ~45 nm and vesicles exhibited diameters ranging from ~13-25 nm. The nanopods were seen either 208 detached from the cell (Fig. 3 e-g), or budding from the pole of C. crescentus (Fig. 3h). 209

II-Protein complexes associated with membrane structures. 210
Next, we examined protein complexes associated with OMEs and OMVs that we could identify in 211 our cryo-tomograms. These complexes fell into three categories: 1) randomly-located complexes 212 found on OMEs, OMVs and cells; 2) randomly-located complexes observed only on OMEs and 213 OMVs, and 3) complexes exclusively located at the tip of OMEs/OMVs. 214

215
In the first category, we observed what appeared to be the OM-associated portion of the empty 216 basal body of the type IVa pilus (T4aP) machinery in OMEs of M. xanthus. These complexes, 217 which were also found in the OM of intact cells, did not exhibit a preferred localization within the 218 tube ( Fig. 5a & b). 219 220 The second category of protein complexes, observed only on MEs and not on cells, contained two 221 structures. The first was a trapezoidal structure observed on purified OMVs of S. oneidensis. The 222 structure was ~11 nm wide at its base at the membrane and was seen sometimes on the outside 223 ( Fig. 5c) and sometimes the inside of vesicles (Fig. 5 d). The second structure was a large crown-224 like complex. We first observed these complexes on the outer surface of membrane vesicles 225 associated with lysed M. xanthus cells (Fig. 6a). Occasionally, they were also present on what 226 appeared to be the inner leaflet of the inner membrane of lysed cells (Fig. 6b). The exact topology 227 is difficult to determine, however, since the arrangement of inner and outer membranes can be 228 confounded by cell lysis. The structure of this complex was consistent enough to produce a 229 subtomogram average from nine examples, improving the signal-to-noise ratio and revealing 230 greater detail (Fig. 6c). These crown-like complexes are ~40 nm tall with a concave top and a base 231 ~35 nm wide at the membrane (Fig. 6c). No such complexes were seen on OMEs and OMVs 232 associated with intact M. xanthus cells. We identified a morphologically similar crown-like 233 complex on the outside of some tubes and vesicles purified from S. oneidensis ( Fig. 6d-f). However, 234 this complex was smaller, ~15 nm tall and ~20 nm wide at its base. As these were purified 235 OMEs/OMVs, we cannot know whether they stemmed from lysed or intact cells. Interestingly, we 236 found a similar large crown-like structure associated with lysed cells of two other species in which 237 we did not identify MEs, namely Pseudomonas flexibilis and Pseudomonas aeruginosa ( However, we also observed differences between the two structures. In L. pneumophila, the widest 266 part of the secretin (15 nm) is located near the plug close to the OM, and the lower end of the 267 complex is narrower (12 nm). In F. johnsoniae, this topology is reversed, with the narrowest part 268 near the plug and OM (Fig. 7i-l). Additionally, the lowest domain of the L. pneumophila secretin 269 did not resolve into a distinct ring as we saw in F. johnsoniae and no extracellular density was 270 observed in L. pneumophila, either in the subtomogram average or single particles [38].

Discussion: 279
Our results highlight the diversity of MEs and MVs structures that bacteria can form even within 280 a single species (Fig. 8). For example, we saw two types of membrane tubes in lysed P. 281 luteoviolacea cells: one narrower with a uniform diameter of 20 nm which did not pearl into 282 vesicles, and one wider with a variable diameter that did pearl into vesicles ( Fig. 1g and Fig. 2c), show that the tubes of H. pylori are CORE-independent, indicating that they are different from the 299 CORE-dependent nanotubes described in other species. 300 301 A recent study showed that the formation of bacterial tubes significantly increases when cells are 302 stressed or dying [28]. Consistent with this, in our cryo-tomograms we saw many MEs and MVs 303 associated with lysed cells (such as in H. pylori, H. hepaticus, and P. luteoviolacea). We also saw 304 tubes and vesicles stemming from intact cells. Given the nature of cryo-ET snapshots, we cannot 305 tell whether a cell that appears intact is stressed, nor can we know whether MEs/MVs formed 306 before or after a cell lysed. One observation which might be related to this issue comes from F. 307 johnsoniae where tubes with regular diameters were seen stemming mainly from cells with a 308 noticeably wavy OM (45 examples that conferred fitness advantages on bacterial species in various environments. Today, the ability 365 of bacteria to extend their membranes to form tubes or vesicles is a widespread phenomenon with 366 many important biological functions. We hope that the structural classification we present here 367 will serve as a helpful reference for future studies in this growing field.