T3SS are usually encoded on pathogenicity islands (see chapter 12), and their genes are induced only when the bacterium encounters its host or if they are cultivated under conditions that are designed to mimic the host. The effector proteins they inject are also encoded by the same DNA element, and their genes are turned on at the same time. The part of the injectisome that traverses the outer membrane is composed of a secretin protein related to those of the T2SS. It also forms a β-barrel composed of about a dozen secretin subunits. Like the secretins of the T2SS, these might require normal bacterial lipoproteins, as well as other components of the secretion machinery, to assemble the channel in the membrane.
Effector proteins to be secreted by at least some T3SS contain a short sequence located on the N terminus of the protein; unlike the cleavable signal sequences used by the Sec and Tat systems, this signal is not cleaved off when the protein is secreted. Many of the effector proteins injected into eukaryotic cells are involved in subverting the host defenses against infection by bacteria. This can be illustrated by Yersinia pestis, the bacterium that causes bubonic plague and in which T3SS were first discovered. In amimals, one of the first lines of defense against infecting bacteria is the macrophages, phagocytic white blood cells that engulf invading bacteria and destroy them by emitting a burst of oxidizing compounds. However, when a macrophage binds to a Yersinia cell, the bacterium injects effectors called Yop proteins into the macrophage cell before it can be engulfed. Once in the eukaryotic cell, these effectors disarm the cell by interfering with its signaling systems and thus preventing the macrophage from engulfing the bacterium. For example, one of the Yop proteins is a tyrosine phosphatase, which removes phosphates from proteins in a signal transduction system in the macrophage, blocking the signal to take up the bacterium and preventing the burst of oxidizing compounds. Some T3SS even inject proteins that provide receptors on the cell surface to which the bacterium can adsorb in order to enter the eukaryotic cell.
TYPE IV SECRETION SYSTEMS
Type IV secretion systems (T4SS) utilize a secretin-like protein (VirB9) that forms a β-barrel channel in the outer membrane and extends into the periplasm, where it makes contact with proteins in the inner membrane. However, unlike true secretins, it seems to require another outer membrane protein, VirB7, to make a channel. The VirB9 protein is covalently attached to the VirB7 protein, which in turn is covalently attached to the lipid membrane, making the structure very stable. A coupling protein (VirD4) binds specific proteins and targets them to the channel. The energy for secretion comes from the cleavage of ATP or GTP in the cytoplasm by channel-associated proteins (Figure 2.39).
T4SS are discussed in chapters 5 and 6, because they are also involved in DNA transfer during conjugation and transformation. The T-DNA transfer system of Agrobacterium tumefaciens has served as the prototype T4SS and is the one about which the most is known and to which all others are compared. Accordingly, the genes and proteins of other T4SS are numbered after their counterparts in the T-DNA transfer system, named the vir genes because of their role in virulence in plants. Some of the genes in the T-DNA transferred into plant cells cause growth of the plant cell, leading to the formation of tumors called crown galls. Others trigger the plant cells to make unusual compounds, called opines, that can be used by the bacterium as a carbon, nitrogen, and energy source (see Box 5.1). Like other T4SS, the Agrobacterium system also directly injects proteins into the plant cell, which makes it a bona fide protein secretion system.
TYPE V SECRETION SYSTEMS: AUTOTRANSPORTERS
All of the secretion systems discussed above use some sort of structure formed of β-sheets assembled into a ring called a β-barrel to get them through the outer membrane. Some of these β-barrels are part of the secretion apparatus itself, while others, like TolC, are recruited from other functions in the cell. However, in type V systems, secreted proteins carry their own β-barrel with them in the form of a domain of the protein that can create a β-barrel when it gets to the outer membrane. These proteins are called autotransporters because they transport themselves.
The mechanism used by autotransporters is illustrated in Figure 2.40, which also shows their basic structure. Most autotransporters consist of four domains, the translocator domain at the C terminus that forms a β-barrel in the outer membrane, an adjacent flexible linker domain (not shown) that may extend into the periplasm, a passenger domain that contains the functional part of the autotransported protein, and sometimes a protease domain that may cleave the passenger domain off the translocator domain after it passes through the channel formed by the translocator domain.
Figure 2.40 Structure and function of a typical autotransporter. A Haemophilus influenzae adhesin is shown; the length in amino acids of each domain, where known, is indicated by the number above the structure, as are some of the important amino acids in the protease domain. The transporter domain at the C terminus that forms a β-barrel in the outer membrane is shown in orange; the passenger domain and the protease domain that cleaves the passenger domain off the transporter domain outside the cell are shown in purple. The flexible linker domain is not indicated. The signal sequence that is cleaved off when the protein passes through the SecYEG (Sec) channel in the inner membrane is shown in black. OM, outer membrane; IM, inner membrane. Modified from Surana NK, Cotter SE, Yeo H-J, et al, in Waksman G, Caparon M, Hultgren S (ed), Structural Biology of Bacterial Pathogenesis (ASM Press, Washington, DC, 2005).
Autotransporters are typically transported through the inner membrane using the SecYEG channel, so they have a signal sequence that is cleaved off as they pass into the periplasm. Their translocator domain then enters the outer membrane, where it forms a 12-stranded β-barrel. This assembly does not occur by itself but requires the same accessory factors used for the assembly of many secretins, including the periplasmic chaperone, Skp, and the Bam complex (see above). The flexible linker domain guides the passenger domain into and through the channel to the outside of the cell. The passenger domain can then be cleaved off by its own protease domain or remain attached to the translocator domain and protrude outside the cell, depending on the function of the passenger domain. The source of the energy for autotransportation is unclear, since, as mentioned, there is no ATP or GTP in the periplasm and the outer membrane does not have a membrane potential. One possibility is that the autotransporter arrives at the periplasm in a “cocked” or high-energy state that drives its own transport.
The prototypical autotransporter is the immunoglobulin A protease of Neisseria gonorrhoeae. It is involved in evading the host immune system by cleaving IgA antibodies on mucosal surfaces. Most known autotransporters are large virulence proteins that perform various roles in bacterial pathogenesis or in helping to evade the host immune system. The IcsA protein of Shigella flexneri, a cause of bacterial dysentery, is localized to the outer membrane, where it recruits a host actin-regulating protein, which in turn recruits another host complex that polymerizes host actin into filaments, pushing the bacterium through the eukaryotic cell cytoplasm as part of the infection mechanism, a process called actin-based motility.