The breadth of macropinocytosis research

1. Introduction

Macropinocytosis is a form of endocytosis in which cells ingest extracellular fluid and solutes into relatively large endocytic vesicles called macropinosomes. The membrane of macropinosomes derives from plasma membrane following protrusive movements of actin-rich membrane folds called ruffles, which close at their distal margins to enclose extracellular fluids. Nascent macropinosomes either recycle to the cell surface or are delivered into lysosomes, where the enclosed solutes may be degraded by acid hydrolases and transported into the cytoplasm for anabolic metabolism. In some cells, macropinocytosis occurs continuously whereas in others it is initiated by receptor–ligand interactions at the cell surface.

A growing number of cellular activities and pathologies have been attributed to macropinocytosis, and additional roles for micropinocytosis in cell physiology remain possible and untested. Therefore, this is a good time to review current understanding of its mechanism, regulation and relevance to physiology and disease. The articles of this issue of Philosophical Transactions B grew out of a discussion meeting entitled, ‘Macropinocytosis’, held in May 2018 and sponsored by the Royal Society. In this introductory article, we summarize current research on macropinocytosis, suggest areas for future work and guide readers to more in-depth reviews of these topics in this special issue or elsewhere in the literature.

2. Macropinocytosis as a mechanism for acquiring nutrients

Macropinocytosis was discovered in the 1930s by Warren Lewis, who used time-lapse microcinematography to observe the characteristic movements of ruffling and fluid engulfment by macrophages and sarcoma cells [1,2]. The essentials of the phenomenon are evident in Lewis' original movies of mouse macrophages (electronic supplementary material, Movie S1) and sarcoma cells (electronic supplementary material, Movie S2). Lewis named this activity pinocytosis and proposed that it provided nutrients for cell growth. This idea was supported by studies of Zanvil Cohn and colleagues during the 1960s and 1970s, who described the role of lysosomes in the digestion of proteins and sugars internalized by pinocytosis in macrophages [3–6]. Nutritional functions for pinocytosis were also described in studies of amoeba in the 1950s and 1960s [7,8]. The application of electron microscopy to cell biology led to the discovery of pinosomes too small to resolve by light microscopy, which were later shown to be the products of receptor-mediated endocytosis and clathrin-mediated endocytosis [9]. These ‘micropinosomes’ led to the renaming of the larger pinosomes seen by Lewis and Cohn as macropinosomes. Now macropinocytosis is considered a subset of activities collectively referred to as clathrin-independent endocytosis (see the article by Donaldson [10]).

The large capacity of macropinosomes suggests that they mediate non-selective internalization and degradation of extracellular nutrients supporting cell metabolism and growth. This concept has been supported in recent years by studies in the soil amoeba Dictyostelium discoideum [11,12] and in tumour cells transformed by oncogenic Ras [13]. The upregulation of macropinocytosis by dysregulation of Ras explains how axenic strains of Dictyostelium acquire nutrients in liquid medium [12] (see the article by Williams et al. [14]). In cancer cells transformed by mutations of Ras, phosphoinositide (PI) 3-kinase or PTEN (phosphatase and tensin homolog), elevated macropinocytosis provides a mechanism for cell growth in nutrient-poor tumour environments. Oncogene-mediated upregulation of macropinocytosis and other mechanisms of nutrient scavenging in cancer cells have become a major focus of current research [15] (see the articles by Palm [16] and Commisso [17]). Indeed, metabolic status regulates rates of extracellular protein scavenging by macropinocytosis in coordination with the nutrient-dependent regulation of the cytoplasmic salvaging pathway called autophagy (see the article by Florey & Overholtzer [18]).

3. Macropinosomes as delivery vehicles

In addition to its general and likely original role in nutrient acquisition, macropinocytosis can also be exploited for additional, more specialized roles. The best understood of these is in antigen-presenting immune cells such as macrophages and immature dendritic cells (DC) that undergo constitutively high rates of macropinocytosis to sample their environment for antigens and subsequently stimulate immune responses by naive T cells [19].

Macropinocytosis also provides a route into the cell that is exploited by several viruses and pathogenic bacteria. Salmonella enterica var. Typhimurium stimulates cell surface ruffling in epithelial cells and macrophages and enters inside spacious phagosomes resembling macropinosomes [20,21]. Macropinocytosis also facilitates cell invasion by the intracellular pathogens Legionella pneumophila, Shigella flexneri and some protozoa [22–24]. Many viruses enter host cells via macropinocytosis, including vaccinia, adenovirus, echovirus, human immunodeficiency virus and Ebola virus [25,26]. Ebola infection can be inhibited by calcium transporter inhibitors, which act by inhibiting macropinosome traffic to late endocytic compartments, indicating that targeting macropinocytosis may be therapeutically useful [27]. Some viruses enter cells through non-macropinocytic mechanisms, and the contributions of macropinocytosis to viral infection may vary with different target cells or with the growth state of the target cells. For example, macropinocytosis is often more prominent in growing cells than quiescent cells, and host cell growth state may influence which pathway a virus uses to infect.

The efficiency of macropinocytosis-mediated delivery of extracellular macromolecules to intracellular targets suggests that the process can be used to deliver drugs into cells with high rates of macropinocytosis. Numerous studies implicate macropinocytosis in the uptake of therapeutic proteins, nucleic acids and liposomes that are too large or charged to enter cells efficiently by other pathways [28]. Also, Mycobacteria BCG used for therapeutic inhibition of bladder cancer enters tumour cells by macropinocytosis [29]. In this issue, Desai et al. [30] review the potential applications of macropinocytosis for delivery of therapeutic nucleic acids into target cells.

4. Signalling to and from macropinosomes

The role for macropinocytosis in cell growth and its requirement for many proteins and phospholipids implicated in growth regulation suggest that macropinocytosis is an evolutionarily conserved mechanism for regulating cell growth that in metazoans has been shackled to growth factor receptor signalling [31], a concept developed by King and Kay in this issue [32]. The extent to which macropinocytosis is required for growth of non-transformed metazoan cells is not known. Other forms of endocytosis may provide all the nutrients that slowly growing cells need. Moreover, unlike free-living amoebae, which must be wary of the environmental molecules they absorb, metazoan cells may acquire extracellular nutrients entirely by transport proteins in the plasma membrane, obviating the need for ingestion of nutrients by endocytosis. Despite this, growth factor receptor signalling to the metabolic regulator mechanistic target-of-rapamycin complex-1 (mTORC1) often requires macropinocytosis of extracellular amino acids or the actin-rich macropinocytic cup domains of plasma membrane, which suggests that macropinocytosis remains essential to growth control in mammalian cells (see the article by Swanson & Yoshida [33]). Thus, although macropinocytosis supports cell growth when upregulated by mutations, the ubiquity of macropinocytosis and its requirement for the growth of all metazoan cells remain uncertain.

Growth factor-stimulated macropinocytosis increases delivery of amino acids into lysosomes, thereby signalling activation of mTORC1 at the cytosolic face of the lysosomal membranes [34]. In addition, signals that are initiated at the plasma membrane during macropinosome formation reach target molecules on endocytic compartment membranes by as yet unknown routes, which may include macropinosome membranes. Indeed, macropinosomes formed in response to lipopolysaccharide (LPS) or activated H-Ras are enriched in Akt [35,36], which suggests that macropinosome membranes serve as vehicles for transport of signals from plasma membrane receptors to targets on intracellular membranous compartments. The concept of macropinosomes as platforms for signal transduction was introduced by Porat-Schliom et al. [35], who showed that macropinosomal membranes contain proteins and phospholipids not typically associated with more conventional endosomes, including H-Ras, Akt and phosphatidylinositol (3,4,5)-trisphosphate. In addition to activities associated with macropinosomes that are qualitatively distinct from more conventional endosomes, the large size of macropinosomes could provide quantitatively significant increases in those signalling activities common to all endosomes. For example, total intracellular levels of phosphatidylinositol 3-phosphate or phosphatidylinositol (3,5)-bisphosphate on membranes may increase significantly during macropinocytosis, possibly exceeding concentration thresholds for qualitatively distinct kinds of signalling.

5. Macropinosome formation and maturation

Because macropinosomes are relatively large organelles of different sizes, the mechanisms that underlie their formation cannot be explained by a stereotyped assembly of scaffolding molecules, such as occurs in the assembly of the uniformly small clathrin-coated vesicles. Rather, the construction of a macropinosome requires localized cell motility: a crawling action that builds a 0.2 to 6 µm diameter cup-shaped cell protrusion, which then contracts at its distal margin to form a similarly sized intracellular vacuole derived from plasma membrane. Like the extension of phagocytic cups that engulf extracellular particles, macropinosome formation requires actin polymerization beneath the plasma membrane necessary to project a cup-shaped ruffle, followed by contraction of the actin network at the rim to close the cup into an intracellular organelle [37]. Recent applications of lattice light sheet microscopy to macropinocytosis have revealed the dynamics of ruffles and cups, as well as unexpected tent-pole-like filopodia in macrophages that suggest different mechanisms of macropinosome formation [38] (see the article by Wall et al. [39]). Unlike phagocytosis, the forming macropinosome has no particle surface to guide morphogenesis of the organelle. Rather, it is formed by a self-organizing sequence of spatially arranged chemical reactions. The essential ingredients of these self-organized biochemical cascades are membrane phospholipids, small GTPases and GTPase effector enzymes that control the activities of actin and myosin [40]. Considerable effort is directed toward understanding which molecules regulate the various chemical activities and how those molecules interact to form the sequence of different cytoplasmic movements that build the macropinosome [41,42] (see reviews by Swanson & Yoshida [33], and Williams et al. [14]).

The fate of nascent macropinosomes varies between cell types but, in general, they either recycle directly to the plasma membrane or fuse with other macropinosomes or endolysosomes, which include early endosomes, late endosomes, recycling endosomes and lysosomes. Additionally, nascent macropinosomes migrate centripetally along cytoplasmic microtubules and extend tubulovesicular extensions that mediate membrane recycling and vesicle fission. The underlying chemistry for this macropinosome maturation also entails localized GTPase activities and associated changes in phospholipids (see the articles by Donaldson [10] and Wall et al. [39]).

The chemistries required for macropinosome formation are themselves regulated either by signalling from cell surface receptors or by receptor-independent activities in cytoplasm. Receptor-mediated stimulation of macropinocytosis is understood in outline form: it usually requires phosphatidylinositol 3′-kinase (PI3 K), other phospholipid-modifying enzymes and the small GTPases Rac, Ras and Rab5a. Constitutive macropinocytosis in macrophages also entails receptor-mediated signalling and PI3 K (see the article by Doodnauth et al. [43]). However, the constitutive macropinocytosis of Dictyostelium and possibly also DC and Ras-transformed cancer cells use many of the same chemistries—PI3 K, Ras and Rac—but their regulation is independent of cell surface receptors. In Dictyostelium, cups assemble and close continuously and spontaneously, indicative of an intrinsic excitable behaviour (see the articles by Williams et al. [14] and King & Kay [32]).

The regulation of macropinosome maturation is important but largely uncharacterized. In macrophages, macropinosomes deliver their contents to lysosomes, whereas in other cells macropinosomes recycle to the plasma membrane without fusing with lysosomes [44,45]. This suggests that the enzymes necessary for macropinosome formation and maturation are subject to feedback regulation by other conditions inside the cell, such as metabolic status, or have obtained specialized properties important for distinct roles such as antigen presentation. Such regulation could modulate either the receptors that trigger macropinocytosis or the signalling molecules necessary for macropinosome closure or fusion with endolysosomes.

Overall rates of macropinocytosis are affected by the metabolic status of the cell. mTORC1 and AMP kinase (AMPK) are responsive to intracellular metabolic states and affect rates of macropinocytosis [46]. Feedback inhibition of macropinocytosis by mTORC1 acts not at the level of macropinosome formation but rather at some later step that determines whether internalized solutes are degraded in lysosomes [47,48]. Scavenging of extracellular molecules by macropinocytosis is modulated by chemistries similar to those that regulate ingestion of other cells by entosis or the scavenging of intracellular molecules by autophagy, indicating integrated regulation of all lysosomal feeding pathways (see the article by Florey & Overholtzer [18]).

6. Macropinocytosis and cell motility

The cytoskeletal activities that create macropinosomes are also needed for oriented cell migration, and these activities compete for the limited cytoplasmic resources needed to effect both kinds of motility. In Dictyostelium, the chemoattractants cyclic AMP and folic acid stimulate both migration and macropinocytosis. However, the two activities are mutually exclusive: macropinosome formation inhibits oriented movement [49]. Similarly, the movement of DC, which is mediated by contractile activities of actin and myosin at the rear of the cell, is slowed by macropinocytosis occurring at the front of the cell [50–52]. Immature DC are actively macropinocytic and migrate slowly through tissues. Mature DC downregulate macropinocytosis and coordinately increase rates of migration [51]. Macropinocytosis at the advancing edge of immature DC allows them to move against hydrostatic pressure gradients that appear when cells enter blind-ended capillaries, thereby permitting immature DC to explore tissues more effectively than mature DC [53].

7. Goals for ongoing research

One of the most important goals for future research is to identify more specific inhibitors of macropinocytosis, ensuring that observed phenotypes are specifically owing to loss of macropinocytosis. This would further advance understanding of the process and the design of therapeutic strategies. Determining the ubiquity of macropinocytosis in growing metazoan cells and its importance for receptor signalling or cell growth should identify features of macropinocytosis that are common to neoplastic and normal cells and therefore inappropriate for therapeutic targeting [15]. The assembly of a comprehensive parts list obtained by genetic screens will likely identify new macropinocytosis-specific activities and molecules. If a protein can be identified that is essential for macropinocytosis, but not much else, then it would make a good target for inhibitor screens. As the relationships between macropinocytosis and autophagy, phagocytosis, entosis and other forms of nutrient scavenging are better defined, we can then ask what is distinct about macropinocytosis that may serve as targets for specific manipulation of the process.

Almost nothing is known about macropinocytosis in vivo. Understanding this will be essential to put cellular studies in proper context and to determine how much and where macropinocytosis occurs within a whole animal. Applications of advanced light microscopy to macropinocytosis in vitro and in vivo will likely lead to new concepts and avenues of investigation. We may find prescient the early reaction to Lewis' first descriptions of pinocytosis, which suggested that it provides a mechanism for ‘purification of body fluids’ that augments phagocytosis [1].

How do some cells drink so much? How does water flow through them? Engulfing fluid by macropinocytosis is energetically expensive. In addition to the energy required to construct macropinosomes, the ingestion of extracellular fluid in macropinosomes carries water and ions that cells otherwise work to exclude. Surprisingly, Dictyostelium mutants completely blocked in macropinosome exocytosis still take up normal amounts of fluid, but do not become swollen [54]. Water and ions must be extruded independently—most probably owing to transport proteins at the plasma membrane or the contractile vacuoles that consume energy to counteract cell swelling by osmosis [54]. In mammalian cells, the failure to process macropinocytosed fluid efficiently can lead to an exaggerated vacuolation that kills the cell, a process named methuosis [55]. Such vacuolation suggests that macropinosomes form and accumulate but fail to dispose of the ingested water. More importantly, perhaps, it indicates the importance of water flow into and out of cells. The flux of ions and water in cells and across macropinosome membranes is presently difficult to measure, but understanding these movements will likely be important for manipulating the process experimentally or therapeutically.

The following articles of this issue consider these topics and more in fascinating detail. Collectively, they show that macropinocytosis has significant roles in physiology and disease and that although much has been accomplished in explaining the process, much more remains to be learned.

Data accessibility

This article has no additional data.

Authors' contributions

Both authors wrote, discussed and revised the manuscript.

Competing interests

The authors have no competing interests.

Funding

This work was supported by NIH grant no. R01 GM110215 to J.A.S. and by a grant to J.K. from the Royal Society.