Extracellular vesicles as therapeutics for inflammation and infection

Abstract

Extracellular vesicles (EVs) are an emergent next-generation biotechnology with broad application potential. In particular, immunomodulatory bioactivity of EVs leading to anti-inflammatory effects is well characterized. Cell source and culture conditions are critical determinants of EV therapeutic efficacy, while augmenting EV anti-inflammatory bioactivity via diverse strategies including RNA cargo loading and protein surface display has proven effective. Yet, translational challenges remain. Additionally, the potential of direct anti-microbial EV functionality has only recently emerged but offers the possibility of overcoming drug-resistant bacterial and fungal infections through novel, multifactorial mechanisms. As discussed herein, these application areas are brought together by the potential for synergistic benefit from technological developments related to EV cargo loading and biomanufacturing.

Graphical abstract

Extracellular vesicles (EVs) are implicated as therapeutics for many applications, and as with many other modalities, the COVID-19 pandemic spurred interest in expansion of capabilities and translation for urgent health problems related to infection as well as harmful inflammatory sequelae. Anti-inflammatory properties of EVs are well described, especially those sourced from human mesenchymal stem/stromal cells (MSCs), dating back to the first therapeutic use in humans to treat graft-versus-host disease [1]. More recently, MSC EVs have been investigated in clinical trials for treatment of COVID-19-related pneumonia and acute respiratory distress syndrome (ARDS) among other applications. However, while most translational focus to date has been on primary MSCs as the source for anti-inflammatory EVs, recent evidence has emerged to support the exploration and development of alternative sources that are more amenable to scalable biomanufacturing and the type of safety and reproducibility profiles that will support ultimate regulatory approval. Additionally, recent evidence has emerged to suggest the potential for EVs as treatments for infectious diseases, which remain a critical global problem for which new treatment options are desperately needed. The potential of EVs to fight infection via immunomodulatory mechanisms follows from prior data in anti-inflammatory applications, but studies from the last several years have now established a paradigm of direct anti-microbial effects of human cell-sourced EVs on pathogenic organisms. This offers the possibility of adding to the infectious disease therapeutic arsenal while avoiding common mechanisms of resistance that plague most currently-approved antimicrobial drugs. Here, the progress and potential of EVs as both anti-inflammatory and anti-microbial effectors will be discussed as well as challenges that remain to be overcome to enable clinical translation of EVs for these applications.

EVs as Anti-Inflammatory Effectors – Cell Source

To date, MSCs have been the primary source cell for production of anti-inflammatory EVs in literature. This has led to the investigation of primary cell-derived MSC EVs in clinical trials for disease applications such as Crohn’s disease (ClinicalTrials.gov ID: NCT05499156), diabetic wound healing (ClinicalTrials.gov ID: NCT05243368), and a recently announced phase 3 trial investigating COVID-19-associated ARDS [2,3]. However, when considering large scale EV biomanufacturing, limitations associated with primary MSCs arise in the form of limited parental cell expansion, heterogeneity, and donor variability [4,5]. These same limitations apply to sources such as cardiac and neural progenitor cells, which also produce EVs with significant anti-inflammatory bioactivity [6,7].

To circumvent these issues, immortalization strategies have been utilized to produce a scalable therapeutic EV source cell. EVs obtained from MSC lines immortalized via alterations to human telomerase reverse transcriptase (hTERT) have been employed in a variety of cell-based models as well as preclinical animal models of neuroinflammatory hypoxia-ischemia, hypersensitivity, cardiac injury, and colitis, among others [8–14]. Additionally, immortalization of MSCs by insertion of a MYC gene has been achieved for scalable EV production, and these EVs have undergone preliminary safety studies for tumorigenic potential [15].

Alternatively, utilization of self-renewing pluripotent stem cell (PSC) sources, in combination with differentiation, can yield renewable sources for EV production. Preclinical animal studies have shown that induced pluripotent stem cell (iPSC)-derived MSC EVs can alleviate inflammation in tendinopathy, Sjogren’s syndrome, atopic dermatitis, and sepsis [16–21]. EVs obtained from PSC-derived cardiac progenitor cells, neural stem cells, and monocytes have successfully abrogated inflammatory responses in animal models for myocardial infarction, status epilepticus, and humanized HIV infection, respectively [6,22–25]. However, PSC differentiation can be time consuming, labor intensive, and result in a heterogenous population of differentiated cells, confounding the quality control process of resulting EV batches [26]. Thus, undifferentiated iPSCs have begun to be explored for EV production, with anti-inflammatory properties evident in animal models of wound healing, osteoarthritis, and stroke in aged mice [27–31]. Key results are summarized in Table 1.

Table 1:

Anti-inflammatory EVs – Cell Source

Animal model	EV source	Effector Cargo	Route of administration	Citation
Hypoxic-ischemic brain injury	hTERT immortalized MSCs	N/A	Intranasal	8
Hypersensitivity	hTERT immortalized MSCs	HIF1α	Subcutaneous injection	9
Cardiac hypertrophy + interstitial fibrosis	hTERT immortalized MSCs	Oncostatin M	IP	10
Colitis	hTERT immortalized MSCs	HIF1α	IP	11
Tendinopathy	iMSCs	N/A	Local injection	16, 18
Tendinopathy	iMSCs	DUSP2/3	Local injection	17
Sjo gren’s Syndrome	iMSCs	TGFβ1, miR-21, miR-125b	IV	19
Atopic dermatitis	iMSCs	IDO1	Subcutaneous injection	20
Sepsis/acute lung injury	iMSCs	miR-125b-5p	IP	21
Myocardial infarction	ESC-derived cardiovascular progenitor cells	N/A	Local injection	22
Status epilepticus	iNSCs	miR-320a, 103a-3p, 21-5p, 26a-5p, 320b, 30a-3p, 181a-5p, 191-5p, agrin, PTX3, hemopexin, Gal-3BP, nidogen-1	Intranasal	23
HIV infection	iPSC-derived monocytes	miR-155 antagomiR	Intraocular	25
Diabetic wound healing	iPSCs	N/A	Local injection	29
Osteoarthritis	iPSCs	N/A	Local injection	31
Ischemic stroke	iPSCs	AKT1, CALM	IV	32

EVs as Anti-Inflammatory Effectors – Augmenting Potency

In addition to selecting an optimal cell source for anti-inflammatory EV production, there has been a coordinated effort from researchers to enhance EV anti-inflammatory efficacy. One commonly utilized method is to biochemically prime EV-producing cells using either small molecules or proteins. Interferon-ɣ priming of MSCs has been reported to enhance the anti-inflammatory efficacy of secreted EVs in animal models of colitis, liver fibrosis, tendon repair, and atopic dermatitis, among others [20,32–34]. However, addition of protein supplements such as Interferon-ɣ can be expensive when considering scaled production of therapeutic EVs. A cheaper and therefore more scalable strategy is to alter the physical culture conditions of EV producing cells [35]. A common strategy that has demonstrated increased vascularization potential of MSC EVs is to culture them in hypoxic conditions [36]. Dong et al. recently demonstrated that hypoxic preconditioning of MSCs yielded EVs with superior anti-inflammatory properties in a murine chronic asthma model [37]. Additionally, culturing MSCs as 3D spheroids enhanced EV anti-inflammatory effects in a murine lung fibrosis model [38].

While strategies to alter culture conditions rely on the resultant change of EV cargo compositions, another avenue to enhance EV anti-inflammatory potency is to load them with specific bioactive cargos. RNA content of EVs is purportedly key to their bioactivity, and thus RNAs are some of the most studied biomolecules for EV loading [39,40]. Specifically, microRNA (miRNA) EV loading studies are some of the most common; overexpression of miR-155–5p, miR-320a, miR-223, and miR-200b-3p in MSCs have generated EVs that possess enhanced anti-inflammatory efficacy in periodontitis, rheumatoid arthritis, hepatitis, and myocardial infarction respectively [41–44]. However, in general, overexpression approaches rely on an overall increase in miRNA content in the cell translating to an increase within secreted EVs and thus are potentially highly inefficient. Therefore, genetic engineering techniques to actively sort miRNA to secreted EVs are being investigated using short nucleotide motif sequences [39,45,46]. Similarly, for bioactive protein cargo, genetic engineering techniques have been utilized to load proteins into EVs via protein fusions or using protein motif sequences [47,48]. Additionally, protein display on EVs via genetic engineering is a highly effective strategy to enhance achieve anti-inflammatory outcomes [49]. Key results are summarized in Table 2.

Table 2:

Anti-inflammatory MSC EVs – Augmented Potency

Animal model	Type of Augmentation	Active molecule/stimuli	Route of administration	Citation
Colitis	Biochemical priming	TNFa, IFNg	IP	32
Liver fibrosis	Biochemical priming	IFNg	IV	33
Tendon injury	Biochemical priming	IFNg	Local biomaterial diffusion	34
Chronic asthma	Hypoxic preconditioning	Hypoxia	IV	37
Lung fibrosis	Physical	3D culture	Intranasal	38
Rheumatoid arthritis	Overexpression	miR-320a	IV	42
Autoimmune hepatitis	Overexpression	miR-223	IP	43
Myocardial infarction	Overexpression	miR-200b-3p	Intracardiac injection	44
Systemic inflammation	Genetic engineering	TNFR1, IL-6ST	IV	49

EVs as Anti-Microbial Effectors – Targeting Bacteria

EVs have long been known to possess immunoregulatory properties, and these often manifest in the anti-inflammatory capabilities discussed above. However, EVs can also exhibit immunoregulatory capacity during infection states, with interesting antimicrobial therapeutic implications. An early example is the effect of antimicrobial activity of EVs from human neutrophils when stimulated by opsonized Staphylococcus aureus [50]. More recently, increasing evidence has demonstrated that host EVs can interact directly with bacterial pathogens transferring nutrients [51], miRNA [52], and neutralizing toxins [53] as shown in Figure 1. Work by Koeppen et al. reported that human airway epithelial cell (AEC)-derived EVs inhibited Pseudomonas aeruginosa biofilm formation. MiRNA let-7b-5p was identified as a component of the EV cargo and was shown to repress expression of biofilm associated proteins in Pseudomonas aeruginosa. Additionally, this effect was blunted by cotreatment with a let-7b-5p antagomir. The authors showed that the effect of the EV treatments could work synergistically with beta-lactam antibiotics aztreonam and carbenicillin through blunting expression of aztreonam induced proteins [52]. Another study by Keller et al. identified that A549 cells, when stimulated by bacterial or CpG DNA, released EVs with ADAM10 on their surface. This allowed the EVs to function as decoys, binding Staphylococcus aureus alpha toxin to protect host cells in vitro as well as enhancing survival of mice infected with Staphylococcus aureus [53].

Figure 1: Schematic of modes of interaction of host cell-derived extracellular vesicles with pathogenic microbes.

Created with BioRender.com.

In addition to these studies, which have primarily focused epithelial cell-derived EVs, immune cell-derived vesicles have also been examined. Macrolets - 10–30 um, discoid, organelle-containing vesicles, derived from macrophages - were produced via exposure of THP-1 macrophages to Escherichia coli LPS. These macrolets were found to be capable of killing Escherichia coli via reactive oxygen species production [54]. Another study explored the differences between two classifications of neutrophil vesicles called neutrophil-derived trails and neutrophil-derived microvesicles, both of which were found to possess bactericidal activity against Staphylococcus aureus and Escherichia coli via reactive oxygen species and granule-dependent pathways [55].

EVs as Anti-Microbial Effectors – Targeting Fungi

Much of the initial research demonstrating cross-kingdom transfer of RNA between host organisms and pathogens was initially developed in the field of plant EVs and their interactions with plant fungal pathogens [56,57]. However, in recent years some of these concepts have been recapitulated in human host-fungal pathogen systems [3,58]. Shopova et al. found that neutrophil-derived EVs, in response to infection with Aspergillus fumigatus, were enriched with antimicrobial peptides cathepsin G, azurocidin, and defensin 1. These EVs were found to be taken up by Aspergillus fumigatus and inhibit its growth by preventing hyphal extension [59] as shown in Figure 1. A study by Zhao et al. examined the antifungal effects of the human oral epithelial cell line Leuk-1 against the fungal pathogen Candida albicans [58]. Leuk-1 EVs were shown to decrease hyphal formation and growth in a dose-dependent manner. Furthermore, in a mouse model of oral candidiasis, EV treatment decreased the size and amount of oral lesions and decreased invasion of Candida albicans into the oral epithelium [58].

EVs as Anti-Microbial Effectors – Targeting Parasites

The majority of host-parasite EV studies have focused on EVs modulating the host immune response, while a few studies have shown anti-parasitic effects of host EVs. An early example by Hu et al. demonstrated that H69 biliary epithelium cells upon infection with Cryptosporidium parvum released EVs containing beta defensin 2 and LL-37. These EVs were shown to decrease the viability of Cryptosporidium parvum and decrease their infectivity as well as bind their surface directly [60]. Additionally, a recent study found that in Plasmodium falciparum infected red blood cell EVs decreased parasitemia specifically in the second invasion cycle of Plasmodium falciparum [61].

EVs as Anti-Viral Effectors

There has been significant research demonstrating that viral infection can lead to changes in composition and production levels of host EVs. Additionally, EVs can modulate the immune response of recipient cells which can be pro-infection – suppressing the host’s antiviral response – or be anti-viral, boosting the immune system response [62,63]. For example, human placental trophoblast EVs confer viral resistance to recipient cells through C19MC-Associated miRNA due to the subsequent increase in autophagy [64]. It is important to note that the anti-viral activity of host EVs is enacted primarily through modulating the immune and host cell response and not through direct interactions with viruses, with the exception of decoy EVs. It has been shown that EVs from CD4+ T cells can act as decoys preventing HIV infection, which can be abolished by HIV virulence factor Nef [65,66]. Additionally, ACE2-containing EVs derived from the 293FT cell line can inhibit SARS CoV-2 infection [67]. In addition to the ability of EVs to act as decoys in SARS CoV-2 infection, BM-MSC EVs have been tested for treatment of COVID-19 respiratory failure because of their anti-inflammatory effects, as discussed above [68]. Key results for the preceding 4 sections are summarized in Table 3.

Table 3:

Anti-microbial EVs

Pathogen	EV source	Stimulation/Infection Prior to EV Collection	Effector Cargo	Model	Citation
Bacteria
S. aureus	Neutrophilic granulocytes	Opsonized S. aureus	N/A	In vitro	50
S. aureus	A549 cells	CpG DNA or bacterial DNA	ADAM10	In vivo mouse model of S. aureus infection	53
S. aureus and E. coli	Neutrophil	Opsonized E. coli and S. aureus	NADPH oxidase	In vitro	55
E. coli	THP-1 macrophages	E. Coli LPS	Mitochondria, Lysosomes, ER Structures	In vitro	54
P. aeruginosa	Airway Epithelial Cells (AEC)	N/A	miRNA let-7b-5p	In vitro biofilm models	52
Fungi
C. albicans	Leuk-1 human oral epithelial cell line	N/A	N/A	In vivo mouse model of oral candidiasis	58
A. fumigatus	Neutrophils	A. fumigatus	Cathepsin G, azurocidin, defensin 1.	In vitro	59
Parasites
C. parvum	H69 biliary epithelium cells	C. parvum	Beta defensin 2, LL-37	In vitro	60
P. falciparum	Red blood cells	P. falciparum	N/A	In vitro	61
Viruses
HIV	CD4+ T cells	N/A	CD4 Receptor	In vitro infectivity assay	65
SARS CoV-2	293FT	N/A	ACE2	In vitro infectivity assay	67

Challenges to Translating EVs for Anti-Inflammatory and Anti-Microbial Applications

Current roadblocks to the clinical translation of EV-based therapeutics for treatment of infection-associated inflammatory disease include scalable production, low potency, and a lack of knowledge of mechanisms of action. While developments towards scalable EV production are underway in the form of optimizing culture conditions and bioreactor utilization, much work remains. As increasingly innovative genetic engineering strategies for EV cargo loading emerge, it will be crucial to appropriately integrate these approaches with existing optimized culture conditions to maximize EV potency.

Underlying all this, a lack of understanding of the mechanisms of action behind anti-inflammatory and anti-microbial effects of host-derived EVs limits the potential for rational improvements and could be a critical barrier to clinical translation. While the diverse cargo of EVs make them an attractive, multifunctional option to treatment of infection-associated inflammation, it also makes it difficult to tease apart their mechanisms of action. For example, there is no data to truly support uptake of host EVs by bacteria, though initial work has shown uptake of host EVs by fungal pathogens [3,59]. Additionally, there is little knowledge of how human noncoding RNA can affect gene expression in microbes. This lack of knowledge as it pertains to mechanism may be a critical barrier to FDA approval as well as funding for antimicrobial EV-based research.

More broadly, limitations in the design of clinical trials for antimicrobial drugs in general also limit the ability for translation of EV-based therapeutics for these applications. Small domestic patient populations make clinical trials in the U.S. challenging to carry out with appropriate power. Additionally, industrial incentives for new antimicrobial agent development are poor due to low profit potential, resulting from novel agents only being utilized for select, highly resistant cases [69].

Conclusions

EVs from a variety of sources with various modifications have shown substantial activity as anti-inflammatory agents, and emerging evidence suggests a significant potential role for EVs in the anti-microbial therapeutic landscape. Synergistic potential between these application spaces is also possible, for example via employment of protein loading and/or display strategies to augment inherent anti-inflammatory properties of EVs with anti-microbial activity to treat an infected non-healing wound. However, further translation to clinical applications will depend on addressing challenges related to EV biomanufacturing and increasing knowledge of EV mechanisms of action.

Highlights.

• Overcoming source cell limitations is key to biomanufacturing anti-inflammatory EVs.

• Methods to enhance EV bioactivity must integrate appropriately with scalability.

• Interkingdom communication between pathogens and hosts via EVs can modulate infection

• EV therapies can synergize their anti-inflammatory effect with anti-pathogenic action