The Secret Life of Exosomes: What Bees Can Teach Us About Next-Generation Therapeutics

Abstract

Mechanistic exploration has pinpointed nanosized extracellular vesicles, known as exosomes, as key mediators of the benefits of cell therapy. Exosomes appear to recapitulate the benefits of cells, and more. As durable azoic entities, exosomes have numerous practical and conceptual advantages over cells. Will cells end up just being used to manufacture exosomes, or will they find lasting value as primary therapeutic agents? Here, a venerable natural process—the generation of honey—serves as an instructive parable. Flowers make nectar, which bees collect and process into honey. Cells make conditioned media, which laboratory workers collect and process into exosomes. Unlike flowers, honey is durable, compact and nutritious, but these facts do not negate the value of flowers themselves. The parallels suggest new ways of thinking about next-generation therapeutics.

Introduction

All eukaryotic cells secrete, and take up, exosomes. These tiny extracellular vesicles (EVs) are laden with genetic instructions that influence recipient cells, sometimes subtly, sometimes dramatically. Over the last two decades, scientific attention to exosomes has exploded: the number of annual citations jumped from 28 in 1996 to 24,765 in 2016 (1). The result has been nothing less than a revolution in our understanding of how cells communicate in health and disease. We now recognize that exosome secretion and uptake create a dynamic, complex signaling network linking cells near and far. Think of the extracellular space as a sea containing trillions of messages in a bottle, quickly read and answered, always turning over, and you begin to get a sense of what is going on inside us every moment of every day.

The recognition that progenitor cells secrete exosomes, and that those exosomes are bioactive, ushered in the concept of exosomes not just as innate signaling entities but also as next-generation cell-free therapeutic candidates (TCs). The general idea is to place isolated cells in defined media, collect the conditioned media, purify the exosomes and then store the exosomes for future use. Here I review the development of exosome-based TCs for the treatment of heart failure; the reader is referred elsewhere for reviews of the mechanistic biology of exosomes(2) or their use in diagnostics (3,4). I conclude by drawing analogies between exosome manufacturing and the process whereby bees make honey. The parallels are more striking than may be evident at first glance.

Cell therapy for heart failure

Despite major advances in pharmacologic and device therapy, heart failure (HF) remains one of the major public health challenges in the modern world. HF prevalence is increasing, not only in the USA(5) but also worldwide(6). No therapy approved for use in HF reverses the disease at a fundamental level; on rare occasions, short-term mechanical circulatory support results in an apparent remission, but, even then, relapses often occur(7). The promise of cell therapy, thus far unfulfilled, is the possibility of regenerating sufficient healthy myocardium to enable stabilization or even regression of heart failure. (The focus here is on HF with reduced ejection fraction [HFrEF], where the growth of new healthy heart tissue is desirable. Tissue regeneration is not likely to be helpful in HF with preserved ejection fraction [HFpEF], where the heart tends to be hypertrophic; nevertheless, other properties of cell therapy, such as anti-inflammatory effects, might be salutary in HFpEF (8).)

A number of cell types have been studied clinically in HF patients(9), including bone marrow mononuclear cells, CD34+ circulating endothelial progenitors, mesenchymal stem cells and their derivatives (cardiopoietic cells and mesenchymal precursor cells), and cardiosphere-derived cells (CDCs). These are all, at best, adult progenitor cells; they are not pluripotent(10). Among classical pluripotent cell types, embryonic stem cell-derived cardiomyocyte sheets have been transplanted epicardially onto the heart, but not injected into the heart itself (11). Progress has been slow, with many false starts. Of the cell types tested to date, only two appear to be in active commercial development for HF. Allogeneic mesenchymal precursor cells are in phase 3 testing for HFrEF(12), and allogeneic CDCs are being developed for various specialized types of heart failure, notably the cardiomyopathy associated with Duchenne muscular dystrophy (13).

The slow progress may be explained, at least in part, by inadequacies in prevailing doctrine. The initial rationale for cell therapy HF trials was based on such doctrine: transplanted cells would engraft, proliferate and differentiate into new myocardium. However, preclinical studies have taught us that progenitor cells do not work that way. Few (≪1% of injected) cells are measurable 3–4 weeks after transplantation, but functional and structural benefits persist for at least 6 months (14,15). During the weeks that appreciable numbers of transplanted cells persist in the tissue, they induce lasting effects in the host myocardium (15). Attempts to pinpoint the key paracrine signals initially focused on growth factors and chemokines such as SDF-1 (14). The emphasis shifted in recent years, when exosomes were implicated as the mediators of many paracrine benefits.

Exosomes

Extensive evidence now supports the concept that EVs are vital for the benefits of cell therapy. For some time, it has been recognized that exosomes are secreted by neural progenitors (16), MSCs (17), CD34⁺ cells (18) and other progenitor cell types (Figure 1, summary of exosome biology). In that context, we demonstrated that blocking exosome production renders CDCs ineffective, while CDC exosomes reproduce the benefits of the parent CDCs (19). Multiple other cell types also appear to rely on EVs as mediators of therapeutic bioactivity (20). Table 1 lists the sources of EVs which have been associated with cardiovascular bioactivity, and their potential applications in disease. Seven of the 18 cell sources listed are stem or progenitor cells. Now, the entire field acknowledges exosomes to be the key mediators of the benefits of cell therapy. Work from many labs, converging on similar issues contemporaneously, radically shifted the prevailing dogma, and did so remarkably rapidly. Even cell types which clearly engraft and differentiate to some degree (e.g., pluripotent stem cell-derived cardiomyocytes) are increasingly believed to act largely by secreting exosomes which improve the viability and function of host heart tissue (11,21–23).

Figure 1. Exosome biology.

Left panel: Schematic of exosome biogenesis. Exosomes arise from the fusion of surface membrane invaginations (multivesicular bodies) and products of the Golgi apparatus. The resulting vesicles are either degraded by lysosomes or secreted as exosomes. Right panel: List of cardinal features of exosomes.

Table 1.

Diverse cardiovascular sources of EVs: bioactivity and therapeutic implications

Cell source of EVs	Bioactivity of EVs	Potential therapeutic indications	First citation
Bone marrow dendritic cells	Decreased T cell activity	Allograft rejection	Pêche et al., 2003(48)
HUVECs	Soluble TNF receptor release	Inflammatory, immune, and stress responses	Hawari et al., 2004(49)
Cardiomyocytes	Release of heat shock proteins	Cardioprotection	Gupta & Knowlton, 2007(50)
Platelets	Myocardial dysfunction	Antagonism in sepsis	Azevedo et al., 2007(51)
Sca-1+ heart cells	Enhanced endothelial cell migration	Wound healing	Vrijsen et al., 2010(52)
KLF2-expressing endothelial cells	Reduced atherosclerotic lesions in ApoE(−/−) mice	Atherosclerosis	Hergenreider et al., 2012(53)
CD34+ cells	Increased endothelial cell viability, proliferation, and tube formation	Angiogenesis	Sahoo et al, 2011(18)
Mesenchymal stromal cells	Cytoprotection of pulmonary artery endothelial cells	Pulmonary hypertension	Lee et al, 2012(54)
Macrophages	Induced differentiation in naive monocytes	Inflammatory disorders	Ismail et al, 2013(55)
Cardiospheres	EVs shown by electron microscopy	Unclear	Barile et al, 2012(56)
Cardiac fibroblasts	Cardiomyocyte hypertrophy	Antagonism in hypertrophy	Bang et al, 2014(57)
Cardiosphere- derived cells	EV blockade prevents, and EVs mimic, CDC benefits	Myocardial infarction, heart failure	Ibrahim et al, 2014(19)
c-kit+ heart cells	Enhanced endothelial cell tube formation, decreased profibrotic genes in fibroblasts	Ischemia-reperfusion injury, heart failure	Gray et al, 2015(58)
Vascular smooth muscle cells	Autocrine calcification	Vascular calcification	Kapustin et al, 2015(59)
AT1R-expressing HEK cells	AT1R functionally reconstituted via exosomes	Modulation of GPCR signaling	Pironti et al, 2015(60)
Pluripotent stem cells	Cytoprotection of H9C2 cells	Ischemia-reperfusion injury	Wang et al, 2015(21)
Cardiac progenitors from human ESCs	Enhanced recovery post myocardial infarction	Myocardial infarction, heart failure	Kervadec et al, 2016(22)

Many, if not most, of the effects of exosomes are seemingly mediated by their RNA contents, specifically miRs and other noncoding RNAs (ncRNAs(2)). Much of the focus has been on miRs, which have a known mechanism of action associated with specific cell signaling pathways. A single miR may be implicated in any given effect of EVs. For example, endothelially-derived EVs containing miR-503 regulate pericyte-endothelial crosstalk and thus may contribute to microvascular diabetic complications (24). In MSCs, exosomal miR-223 contributes to cardioprotection in sepsis (25). However, miRs constitute a small fraction of the ncRNAs within exosomes (26). It is becoming increasingly clear that non-miR ncRNAs also underlie some forms of bioactivity (27). The emerging paradigm is as follows: transplanted progenitor cells secrete exosomes which transfer payloads into target cells, inducing transcriptomic and phenotypic changes that underlie the benefits of cell therapy. The fact that exosomes reproduce key benefits of cells suggests that exosomes may themselves be attractive next-generation cell-free TCs.

Cells versus exosomes as therapeutic agents

There is ample evidence for efficacy of cardiac cell therapy in preclinical models (28) and, to a lesser degree, in patients (29). However, cells are fragile living entities which can be difficult to manufacture and to handle (30). Exosomes offer the potential to overcome key limitations of cell therapy. Potential advantages over cells include: 1) product stability: exosomes are tough lipid bilayer vesicles which tolerate lyophilization and other extremes of handling, while retaining bioactivity (31); 2) immune tolerability: human exosomes work well in non-immunosuppressed mice (32), rats (33) and pigs (34), even with repeat administration (32); 3) exosomes are efficacious after systemic delivery (32,35,36); 4) dosing is not limited by microvascular plugging or loss of transplanted cell viability (2); 5) there are multiple approaches to enhancing efficacy, including, but not limited to, genetic engineering of the parent cells(37); 6) ability to cross the blood-brain barrier (38): though not particularly germane to HF, this is a notable distinguishing feature between exosomes and their parent cells.

Despite these potential advantages, exosomes remain at a primitive stage of development as TCs. In terms of clinical applications, the only studies which have been completed are small academic trials of exosomes from dendritic cells in cancer immunotherapy, with mixed results (39). One isolated case report used intravenous MSC exosomes to treat refractory graft-versus-host disease, with apparent short-term success (40). As of September 18, 2017, clinicaltrials.gov lists 42 active trials, but most are diagnostic studies of exosomes in biological fluids; of the 4 therapeutic trials, 3 are nutraceutical and the 4^th is a single-center study of plasma-derived exosomes in wound healing(41) (ClinicalTrials.gov Identifier NCT02565264). Therefore, this search yields no evidence of a concerted program to prepare clinical-grade exosomes for therapeutic applications. Unlike cell manufacturing, where standard operating procedures for mass production have been in development for decades(29,30), exosome manufacturing is in its infancy. Fundamental issues which remain largely unresolved include batch-to-batch variation, purity, quality control, delivery (including targeted delivery), off-target effects, pharmacokinetics and pharmacodynamics. Thus, enthusiasm should be tempered by the recognition that the field is in its infancy, and problems will inevitably arise over time. Nevertheless, it is worth reflecting on how exosomes are made, and exploring natural precedents to that manufacturing process.

How to make therapeutic exosomes

To make exosomes, we start by plating or suspending isolated cells in serum-free media, and allowing the cells to condition the media (19). (Serum-free media are preferred for exosome manufacturing, as serum contains trillions of exosomes per milliliter(42), making it impossible to distinguish newly-secreted versus pre-existing EVs when serum is included.) The conditioning step can be as brief as hours or as long as 15 days. The media are then collected for further processing. Raw conditioned media contain not only exosomes but also proteins and other products of metabolism. Further processing, by laboratory personnel, typically involves ultrafiltration to concentrate the EV-media and to remove non-EV contents (e.g., by molecular weight exclusion). Ultracentrifugation may be used to separate exosomes from larger EVs such as microvesicles or apoptosomes (43). Diafiltration is used to exchange the culture media for the desired EV storage media. Finally, the purified product is packaged into vials or other dosage-defined units, for later use.

Exosomes are to parent cells as honey is to flowers

As summarized above, cells make conditioned media, which human beings collect and process into exosomes. This highly artificial process has only been operative for a few decades: exosomes were first recognized in the 1980s (see ref (2) for a review). Meanwhile, a natural workflow which originated ~100 million years ago is strikingly analogous. Flowers make nectar, which bees collect and process into honey. In both cases, the starting point is a fragile living entity (cells or flowers); the intermediate is a biological liquid (conditioned media or nectar); and the ultimate deliverable is a stable cell-free biological product.

As depicted in the Central Illustration, reflection on the way bees make honey opens up new ways to conceptualize our manufacturing of exosomes. Not surprisingly, the process whereby bees make honey is highly evolved and standardized, leading to a much more reproducible product than is typical with current technologies of exosome manufacturing. Bees collect nectar from multiple blossoms (44) and store it in a specialized stomach where invertase and other enzymes begin to break down the complex sugars to monosaccharides. The processing of nectar continues in the hive, with repeated cycles of regurgitation and reingestion among worker bees, further breaking down the sugars and progressively decreasing water content. Eventually bees place the processed nectar in honeycomb cells, further condense the liquid into honey by fanning it with their wings, and seal each cell with a wax lid. (45). Thus, bees don’t just gather and store the nectar: they also alter it chemically. When we make exosomes using current technology, the general goal is to purify, but not to alter, the vesicles. What honeybees do routinely is akin to ongoing efforts to alter exosomes chemically to enhance their efficacy, e.g., by covalent modifications designed to achieve selective targeting (46,47).

Central Illustration. Parallels between exosome manufacturing and the making of honey.

Top row: Cells condition the media in which they are grown. Humans collect conditioned media and extract exosomes. Lower row: Flowers make nectar. Bees collect nectar and process it to make honey. The starting points (cells, flowers) are both fragile living entities. The intermediates are fluids (conditioned media, nectar) arising from those living entities. The end products (exosomes, honey) are durable and azoic, with traces of their origins but distinct properties.

Conclusions

Cells, like flowers, are fragile life forms. They die easily, and their properties are exquisitely dependent on how they are grown and maintained. Cell therapy is akin to horticulture: the goal is to grow precious living entities, then to protect them from the many stresses that will predictably lead to their degradation. Exosome manufacturing, in contrast, resembles the secret life of bees. The entire focus is on processing a biological liquid into a distinct cell-free product. Exosomes, like honey, contain the “active ingredient” of their sources, but differ fundamentally from those sources. Both exosomes and honey are durable, azoic entities. They have unique uses and applications; neither is just a proxy for the living things from which they are derived.

Ironically, our increasing mechanistic understanding of cell therapy has led us away from cells, and towards exosomes. Why not ditch cells altogether and focus solely on exosomes as next-generation TCs? Although exosomes come with their own limitations, detailed above, they contain all the promise of cell therapy, and more, without the myriad hassles of having deal with living cells as a drug product. In this view, cells may simply become a manufacturing intermediate to produce exosomes. Nevertheless, just as flowers have their unique uses, it is possible that cells might be better-suited than their exosomes for certain indications. Cells, delivered intravascularly, can lodge in capillaries and serve as natural slow-release platforms for exosomes (34). Time will tell if cell therapy will survive as such, or if it will turn out to be entirely supplanted by exosomes and, eventually, by defined factors.

Abbreviations

EV

extracellular vesicle

TC

therapeutic candidate

HF

heart failure

HFrEF

HF with reduced ejection fraction

HFpEF

HF with preserved ejection fraction

CDC

cardiosphere-derived cell

ncRNA

noncoding RNA

MSC

mesenchymal stem cell