Background

In recent years, many studies have been conducted on membrane vesicles (MVs) produced by Gram-negative bacteria (Kulp and Kuehn, 2010). MVs are spherical membranous structures with diameters ranging between 20 and 300 nm, and they enable a protected secretion of proteins, lipids, RNA, DNA and other effector molecules (Beveridge, 1999; Mashburn-Warren and Whiteley, 2006; Ellis and Kuehn, 2010). The ability of bacterial MVs to interact with and enter host cells has prompted their exploration for novel clinical and biotechnological applications (Chen et al., 2010; Gujrati et al., 2014; Robbins and Morelli, 2014; van der Pol et al., 2015).

Different methods have been used to characterize and quantify bacterial MVs such as protein or lipopolysaccharide quantification, nanoparticle tracking analysis, flow-cytometry or proteomic analysis among others, but none of them allow MV visualization or clarify MV structure (Kulp and Kuehn, 2010). In many studies, MV-producing cells and vesicle morphology have been visualized using conventional electron microscopy techniques such as scanning electron microscopy or thin-section transmission electron microscopy (TEM) of samples chemically fixed and dehydrated at room temperature (Fiocca et al., 1999; McBroom and Kuehn, 2007; Deatherage et al., 2009; Alves et al., 2015; Ercoli et al., 2015). However, in both methods the extracellular matter in which MVs are found has a propensity to collapse and be removed during sample preparation (Beveridge, 1999; Fiocca et al., 1999; Nevot et al., 2006; Mashburn-Warren and Whiteley, 2006; McBroom and Kuehn, 2007; Deatherage et al., 2009; Ellis and Kuehn, 2010; Chen et al., 2010; Gujrati et al., 2014; Robbins and Morelli, 2014; Alves et al., 2015; Ercoli et al., 2015; van der Pol et al., 2015). TEM observation of negatively stained MVs is the most commonly used technique to assess MV presence and morphology. In this case, MVs are dried on a grid at room temperature and most of them are visualized as irregular and wrinkled round structures (Figure 1A). Although useful for confirming the presence of MVs, this technique does not provide enough resolution to visualize their exact shape or the presence of atypical MVs or artifacts, which may arise from the preparation process or when working with new or genetically manipulated strains (Bernadac et al., 1998; Kolling and Matthews, 1999; Wai et al., 2003; Lee et al., 2007; McBroom and Kuehn, 2007; Roier et al., 2016). The variability created by these features can hamper studies evaluating the application of MVs in different fields (Gujrati et al., 2014; van der Pol et al., 2015). Furthermore, in conventional electron microscopy techniques, samples have to be observed under a high vacuum and the electron beam leads to the destruction of native biological structure (Nevot et al., 2006).

In cryo-transmission electron microscopy (cryo-TEM), biological specimens are immobilized or fixed using physical procedures based on freezing. The goal is to obtain vitreous or amorphous ice from the sample water after ultra-rapid freezing, avoiding ice crystal formation and thus preserving the cellular structures. In contrast with conventional methods, ultra-rapid freezing immobilizes all molecules in a sample within milliseconds. After this instantaneous cryo-immobilization, it is very important to maintain this state throughout the imaging acquisition in the cryo-electron microscope (Cavalier et al., 2008).

In recent years, improvements in cryo-TEM techniques have enabled the imaging of biological specimens with greatly enhanced resolution and close to their native state, which has refined our knowledge of bacterial structures, including bacterial MVs (Renelli et al., 2004; Jensen and Briegel, 2007; Studer et al., 2008; Palsdottir et al., 2009; Frias et al., 2010; Perez-Cruz et al., 2013 and 2015).
There are several methods for freezing samples. Plunging samples into a pre-cooled cryogen, such as ethane or propane, is a common technique for freezing suspensions (viruses, liposomes, bicelles, micelles...), molecular assemblies (proteins, nucleic acids...), isolated organelles and cell structures and even small cells such as bacteria (Dobro et al., 2010). This method, known as plunge freezing, has been used to visualize bacterial MVs (Renelli et al., 2004), and has revealed, for example, the existence of a new type of Gram-negative MVs produced by environmental and pathogenic bacteria, named outer-inner membrane vesicles (O-IMVs) (Figures 1B, 2C and 2) (Perez-Cruz et al., 2013 and 2015).


Figure 1. Different samples of isolated MVs observed by (cryo-)TEM. A. MVs from Shewanella vesiculosa M7^T prepared by negative staining and observed by TEM. Vesicles appear collapsed and as irregular and wrinkled round structures. B and C. MVs from Shewanella vesiculosa M7^T and (C) Acinetobacter baumannii AB41 prepared by plunge freezing and observed by cryo-TEM. MVs appear as regular round structures and different types can be observed: conventional bacterial outer membrane vesicles with one membrane layer and a new type called outer-inner membrane vesicles (O-IMV) with two membrane layers (black arrows). D and E. Exosomes produced by melanoma cancer cells and (E) liposomes synthesized from lipids. All types of vesicles appear as regular round structures and membrane layers are clearly visualized. Scale bars = 500 μm.

Plunge freezing of MVs allows us to observe their exact shape, size and integrity, the presence of one or more membrane layers, and attached surface-associated structures like viruses, among other features (Perez-Cruz et al., 2016).

Bacterial MVs are not the only focus of investigation. Another expanding research area is the study of extracellular vesicles (EVs), which are released by cells from all domains of life-Eukarya, Archaea and Bacteria-and are considered intercellular communicasomes, acting as a mechanism for distance delivery of active compounds between cells (Yoon et al., 2014). EVs have demonstrated great promise as natural drug delivery systems loaded with therapeutic molecules (Armstrong and Stevens, 2018). Other candidates for the improvement of drug delivery systems are liposomes and other lipidic formulations (Alavi et al., 2017), varying in size and lamellar number, and cryo-TEM observation could also be applied to advance their development. The described protocol therefore has potential application for other types of vesicles currently under active research (Figures 1D and 1E).


Figure 2. Cryo-TEM of isolated MVs, a general overview of the whole protocol