Endocytosis is the process by which cells internalize extracellular material. It is crucial to cell survival and the proper functioning of tissues, being involved in processes as diverse as growth, neural transmission and pathogen clearance. But several opportunistic molecules (bacterial and plant toxins) and pathogens (viruses and bacteria) can exploit the endocytic machinery of a host cell for their own gain1,2. Writing in Science, Mercer and Helenius3 report how vaccinia virus (Fig. 1) — a cousin of variola virus, which causes smallpox — deceives host cells into taking it up through endocytosis.

Figure 1: Source of infection.
figure 1


False-colour electron microscopy image of vaccinia virus.

When cells are damaged or dying, for example during programmed cell death (apoptosis), they show several characteristic features. For instance, phosphatidylserine, a lipid that is abundant in the inner (cytoplasmic) layer of the cell membrane, is redistributed to the outer layer. This phospholipid is thus available to bind to receptors on the surface of phagocytic cells that initiate the apoptotic cell's destruction and engulf it4. Another sign of apoptosis is membrane blebbing, or the formation of irregular bulges on the cell membrane. Blebbing also occurs during other processes, such as cell migration and division, but its function is unclear.

When enveloped viruses bud off from their host cell, they inherit a lipid coating (envelope) that has the same composition as the host cell membrane. Mercer and Helenius3 report that the outer layer of the vaccinia virus envelope contains phosphatidylserine and that this is crucial for infection. They propose that the underlying mechanism is as follows. When a cell becomes infected with the virus, it displays apoptotic features, including the presence of phosphatidylserine in its outer membrane layer. Thus, when the virus buds off from the cell, it inherits this as part of its envelope. Consequently, cells probably 'mistake' the unusually large vaccinia virus for an apoptotic body (the debris of dying cells) and engulf it.

Mercer and Helenius3 find that vaccinia virus seems to enter its host cell through an endocytic process called macropinocytosis, which normally mediates fluid uptake. Like virus budding, virus uptake also exploits apoptotic mechanisms. The authors show that vaccinia virus initially binds to cytoplasmic protrusions called filopodia that extend from the surface of the target cell. It moves along them towards the cell body, and then somehow sends signals into the cell, stimulating extensive membrane blebbing. During blebbing, the actin network that forms the scaffolding of the cell beneath the cell membrane becomes detached (Fig. 2). Blebbing is transient, and soon after, the network reassembles in the same cellular location, and the blebs retract5.

Figure 2: Vaccinia virus chooses to be different.
figure 2

a, Most viruses, as well as plant and bacterial toxins, enter host cells through the classic method of endocytosis, which involves membrane invagination and pinching off of the membrane to form an intracellular transport vesicle. Three different forms of this type of endocytosis are shown. b, Mercer and Helenius3 find that vaccinia virus enters by an endocytic mechanism resembling macropinocytosis. On the cell surface, the virus triggers membrane blebbing, which might also lead to the formation of membrane invaginations that will evolve into transport vesicles. The actin network, which is normally present beneath the cell membrane and is involved in various endocytic processes, is absent in the bleb, but re-forms during bleb retraction.

The authors find that virus internalization by this macropinocytosis-like process is mediated by the blebs, as bleb retraction and re-formation of the actin network coincide with virus entry. Moreover, the drug blebbistatin, which inhibits blebbing6, blocks virus entry. In addition, virus internalization requires several proteins (including actin, PAK1, Rac1 and various lipid- and protein-kinase enzymes) that are involved in membrane blebbing3. Thus, blebbing might participate in endocytosis, probably when the bleb is retracting and the actin system is re-forming, as the bleb could fold over, or invaginate — a process that would resemble macropinocytosis.

The possibility that blebbing and macropinocytosis are two entirely independent processes, both of which are stimulated by vaccinia virus, is equally valid. Specifically, blebbistatin inhibits the myosin II protein, which is required for blebbing. But it can also inhibit myosin-II-independent processes such as macropinocytosis7. So Mercer and Helenius's observations raise the question of whether macropinocytosis should be subdivided into at least two types: the traditional type in which membrane ruffling (small, dynamic folds of the membrane; Fig. 2b) precedes vesicle formation; and the type that involves blebbing. Multi-modal macropinocytosis would fit well with the increasing number of other types of endocytosis that are being identified8.

That opportunistic pathogens exploit various mechanisms for entry and replication within host cells is also documented in a study9 of the bacterium Pseudomonas aeruginosa. This pathogen induces the formation of very large membrane blebs in epithelial cells, entering the blebs and replicating there. The blebs are quite translucent, and do not seem to contain cytoskeletal elements such as actin. Moreover, the bacteria are highly motile within the blebs. But the exact mechanism of bacterial entry into them remains elusive.

Discoveries often raise new questions, and Mercer and Helenius's work3 is no exception. First, what is the exact relationship between blebbing and macropinocytosis? Cholesterol, for example, is required for both virus infection and macropinocytosis. Is it also required for blebbing? Is virus-induced blebbing cell-type-specific? What happens in polarized cells, in which membrane components and structural elements vary in different parts of the cell, as opposed to the non-polarized cell lines that were studied here? As the active form of the Arf6 protein inhibits virus infection, one might also wonder how Arf6 is involved in this process. Much is to be learnt about the mechanisms and pathways underlying the internalization of opportunistic pathogens such as vaccinia virus. Ironically, further knowledge about endocytosis itself is likely to come from studies of pathogens.