Proteomic analysis of bacterial (outer) membrane vesicles: progress and clinical potential

KEYWORDS:

• Bacteria
• immunization
• membrane vesicles
• outer membrane vesicles
• proteomics
• vaccine

1. Introduction

In recent years, the discovery and in depth analysis of bacterial membrane vesicles (MV) has opened up a new field in bacterial proteomics, in terms of both basic research on a hitherto neglected subcellular fraction and in clinical potential of these biological spheres. MVs are meanwhile known to be constitutively released by a large fraction of Gram-positive and Gram-negative commensal and pathogenic bacteria. The vesicles, which consist mostly of a lipid membrane bilayer, are formed as blebs, varying in sizes between 20 and 300 nm. The cargo of MVs identified so far differs amongst bacterial species and comprises different putatively biologically active entities like nucleotides (DNA, RNA), proteins (enzymes, toxins) as well as immunogenic peptidoglycan in Gram-negative bacteria. By considering the reported cargo and experimentally determined results by exposition of MVs, e.g., to cell lines, the assigned functions of MVs are versatile as proteins found in the vesicles are involved in intercellular communication, nutrient acquisition, stress responses, and virulence [1]. Global proteome studies of outer membrane vesicles (OMVs, found in Gram-negative bacteria) revealed enrichment of proteins of the outer membrane, while proteins of the inner membrane were lacking [2]. This gave first evidence that the formation of bacterial vesicles must be regarded as a directed process rather than a random event. In this context, environmental factors such as bacterial growth phase, lifestyle, media composition, antibiotic exposure, iron, and oxygen availability may alter the biogenesis and cargo of MVs [3]. It is further assumed that there is not only a single mechanism regarding the biogenesis of MVs as discussed for Gram-negative bacteria, but several may occur simultaneously [4]. In the same paper, in particular, accumulation of misfolded proteins in the periplasm, disturbances of crosslinks between the outer/inner membrane and the peptidoglycan, as well as alterations in the peptidoglycan have been described as important components in the biogenesis of OMVs. Additionally, a prophage-derived endolysin has recently been described as a mediator for the production of MVs by generating holes in the peptidoglycan cell wall [5].

1.1. Clinical potential of MVs

Membrane proteins and proteins that are exposed on the surface play a major role in nutrient acquisition, immune modulation, adhesion, and virulence. Therefore, these proteins are crucial for bacterial infections. Because MVs carry bacterial proteins, both on the surface and in the lumen, they can be exploited as vaccine candidates with a high clinical potential. The most striking advantage of proteins being presented on the surface of MV is their native conformational state as well as their native local environment. This makes these proteins particularly suitable as immunogens, and hence, they are perfectly suited for vaccination approaches by presence and presentation alike. The last notion has been corroborated by the fact that both purified PorA and PorA incorporated in MV have induced experimentally comparable IgG titers, with only the porins residing in their natural local environment provoked the production of antibodies that were able to kill meningococci [6]. These adjuvant properties have been used successfully in vaccination for over 20 years against meningococcal disease by Neisseria meningitidis [7]. Anyhow, due to strain variabilities, the vaccines cannot be applied to specific meningococcal outbreaks, as the immune response against N. meningitidis OMVs is almost exclusively directed against the variable PorA. Thus, the presence of highly conserved membrane protein species is mandatory for vaccine development. Nevertheless, native MV vaccines against Bordetella pertussis, Acinetobacter baumannii, and Staphylococcus aureus and other pathogens seem to be promising for future developments, as reviewed in [8] and [9].

1.2. Bioengineered MVs: designing of next-generation vaccines

As we drift into a postantibiotic era, variable solutions for fast vaccine development are of vital importance. Due to the facilitated display of genetically modified proteins on their surface, MVs have the potential to be the vaccine platform of choice, as the development in the last years indicates [10]. Recently, it has been shown that overexpressed outer membrane proteins were present in high density and sufficient purity in their correct folding on OMVs of Escherichia coli [11]. Thus, the MVs are used as small transport units, which carry one or more heterologous proteins on the outside or in the lumen. This strategy has already been proven with successful immunogenicity of such proteins when coupled to abundant (outer) membrane proteins on vesicles [12]. Different concepts of the design of recombinant OMVs and their application are reviewed in detail in [10]. Similar results were acquired when OMVs were decorated with a conserved β-(1–6)-linked poly-N-acetyl-D-glucosamine glycopolymer [13]. Glycoengineered OMVs represent another path for the development of OMV vaccine approaches [14]. Often, the lipopolysaccharide (LPS) pathway was modified to reduce endotoxicity of the vesicles explicitly in the Gram-negative hosts [15]. This kind of modifications has gained importance, because detergent-free MV extraction protocols are getting popular, especially as the LPS content is not reduced compared to detergent based extraction. One reason for this trend is due to proteomic studies that revealed the enrichment of membrane proteins in MVs that are extracted with detergent-free methods, illustrating different immunogenic properties of the MVs depending on their extraction strategy [16]. Additionally, several mutants are reported that exhibit a hypervesiculation phenotype, causing an increase in vesiculation up to 200-fold compared to the wild type [17]. Most of the proteins, in which gene deletions lead to a hypervesiculating phenotype, are involved in cellular stress responses or constitute important cellular components linking the inner membrane, the peptidoglycan, and the outer membrane. Due to the increased vesiculation, higher yields of MV can be achieved with only a single batch of cultivation, resulting in a more efficient bioproduction process. Also, it has been shown that the MVs is of one of the hypervesiculation mutants, which lacks the protein NlpI (restricts the activity of peptidoglycan endopeptidase), contained a higher amount of the outer membrane protein W, which is coupled to green fluorescent protein in comparison to the wild-type MVs of E. coli [18]. This suggests that at least some of the described hypervesiculation mutants may alter or even suppress the MV-associated sorting mechanisms as this would explain the increased presentation of the recombinant protein on the MVs of the NlpI mutant. Therefore, hypervesiculation mutants might not only be interesting for a higher yield during the bioprocess, but also when specific recombinant proteins shall be presented in a higher abundance on the surface of the vesicles. If the reported results hold true in further studies for other recombinant proteins for membrane display, this would be of great interest for the community regarding the development of recombinant MV vaccines in the future.

1.3. Proteomic analysis of membrane vesicles

Proteomic analyses of MVs are challenging as MVs have to be enriched due to the low concentrations of MVs in the bacterial supernatant and body fluids. This can be accomplished either by sedimentation methods such as precipitation, ultracentrifugation, or density gradient centrifugation or by filtration methods like ultrafiltration or tangential flow filtration. Most commonly, a combination of ultracentrifugation and density gradient centrifugation is used. Even though they are time-consuming, those methods are easy to perform and the resulting vesicles are of high purity, which is not only important for possible vaccine application, but also because of the interference with downstream processing, e.g. protein determination. Together with the developments in proteomics in general, and in membrane proteomics specifically, MV proteomics is progressing rapidly. For instance, the acquisition speed and the sensitivity of the mass spectrometers are continuously increased. In modern proteomics, gel-free approaches for bottom-up proteomics are getting more popular. The most commonly applied methods are in-solution digest, filter-aided sample preparation, and suspension traps (S-Traps). A current comparison showed that S-Trap provides the most efficient digestion with the highest number of unique protein identifications [19]. The biggest advantage of the gel-free compared with in-gel digest approaches is the opportunity for decreased measuring time needed but omitting a pre-fractionation of the sample. Even though state-of-the-art mass spec instruments can handle complexities of OMV samples without pre-fractionation, e.g. by increasing the liquid chromatography gradient length, methods like high pH reversed-phase chromatography can be performed before the liquid chromatography-mass spectrometry run if a pre-fractionation of the sample is necessary [20]. Due to the orthogonality to the second RPLC separation, this leads to increased number of identifications of peptides and proteins. For quantification, both label-free and label-based methods may be applied. While label-free quantification is based on spectral counting, peak area, or intensity of the peptides, label-based methods are relying on metabolic or chemical labeling. Whereas label-free methods are very easy to apply, label-based methods give more accurate results, but are work-intensive and work-expensive. By employing metabolic labeling techniques, quantitative comparisons of the OMVs and the outer membrane regarding their protein composition lead to insights into the dynamic functions of the OMVs in the context of pathogenesis [21] and within bacterial communities. Further, the utilization of proteomic techniques and determination the OMV proteome is of interest for the understanding of the physiology of their biogenesis or in case of vaccine development for the identification of putative immunogens [22]. Moreover, it is helpful to check the successful incorporation of heterologous or hybridized proteins in bioengineered vesicles. Additionally, possible targets for mutations, change in composition or coupling (chimera proteins) can be selected. Furthermore, OMVs generate an immune response in the host cells as reviewed in [23]. Therefore, proteomics is also applied to investigate the impact of OMVs on the host‒pathogen interaction. Recently, it has been reported that OMVs, apart from carrying protein, LPS and peptidoglycan, also transfer DNA into eukaryotic cells [24]. Consequently, further studies are required to guarantee a safe application of OMVs as vaccines, because it is unclear if an integration of the DNA occurs. Next to proteomics, lipidomics represents an important tool, e.g. for the understanding of the lifecycle of OMVs, as the entry into the host cell is dependent on the LPS structure [25] and might also influence the biogenesis of OMVs [26]. Together, the combination of some of the described techniques will help with the dosing, bioengineering, and characterization of MV vaccines as not only the concentration of the vesicles is of interest, but also their purity, the concentration of the proteins on the surface of the vesicles, and the vesicle proteome in general. Anyhow, it still remains a challenge to investigate the differences of the proteome within a purified vesicle population as this would require single-vesicle proteomics. Further, by performing proteomic studies of host‒pathogen interaction, more insights into the adaption of the host will be gained. This will lead to a better understanding of the biological roles of OMVs, the putative moonlight functions of their protein cargo, and the mechanism of action when OMVs are applied as a vaccine.

2. Summary

Even though there are some challenges to overcome, the progress in the development of MV vaccines is highly promising. This is especially due to the presentation of the antigens in their natural environment, which seems to have a significant impact on their immunogenicity. Furthermore, MV vaccines benefit from their high stability, the low production costs, and their biological safety. Bioengineering of MVs offers new opportunities for the vaccine development, e.g. having the possibility to present the proteins of different bacterial species as well as subtypes simultaneously and thus increasing the vaccine coverage compared to native OMVs. Although not being the driver of this development, proteomics may contribute essential information toward this development. Especially in the upcoming postantibiotic era, where facilitated and timely vaccine development will be required, vaccine platforms such as MVs, which shorten the development time, will be of great importance and could contribute to overcome the antibiotic resistance crisis [10,27].

Declaration of interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This work is supported by the European Maritime and Fisheries Fund and the federal state Mecklenburg-Vorpommern [grant MV-II.1-LM-003/730217000003].