channels in the outer membrane presents some of the same problems associated with having channels in the cytoplasmic membrane, such as the SecYEG channel. For example, how do they select the proteins that are to go through without letting others through, and how do they keep smaller molecules from going in and out? This process is called channel gating; the gate is open only when the protein being exported passes through. A second issue is the source of the energy to export a protein through the outer membrane. There is no ATP or GTP in the periplasmic space to provide energy, and the outer membrane is not known to have a proton gradient across it to create an electric field. In this section, we describe mechanisms used by the various secretion systems for solving these problems and mention some examples of proteins exported by each of the systems.
TYPE I SECRETION SYSTEMS
Type I secretion systems (T1SS) secrete a protein directly from the cytoplasm to the outside of the cell (Figure 2.39). They are different from the other types of secretion systems and more closely related to a large family of ATP-binding cassette (ABC) transporters that export small molecules, including antibiotics and toxins, from the cell. The ABC transporters tend to be more specialized, exporting only certain molecules from the cell. To get the protein through the inner membrane, T1SS use a dedicated system that consists of two proteins, an ABC-type protein in the inner membrane and an integral membrane protein that bridges the inner and outer membranes. To get through the outer membrane, T1SS use a multiuse protein, TolC, that forms the β-barrel channel in the outer membrane. Because the TolC channel has other uses and also exports other molecules, including toxic compounds, from the cell, it is recruited to the T1SS only when the specific protein is to be secreted. When the molecule to be secreted binds to the ABC protein, the integral membrane protein recruits TolC, which then forms the β-barrel in the outer membrane. The cleavage of ATP by the ABC protein presumably provides the energy to push the secreted protein all the way through the TolC channel to the outside of the cell.
Figure 2.39 Schematic representation of the type I, II, III, and IV protein secretion systems. The examples shown are for type I (hemolysin A [HlyA] of Escherichia coli), type II (pullulanase of Klebsiella oxytoca), type III (Yop of Yersinia), and type IV (vir of Agrobacterium tumefaciens).
The classical example of a protein secreted by a T1SS is the HylA hemolysin protein of pathogenic E. coli. This toxin inserts itself into the plasma membrane of eukaryotic cells, creating pores that allow the contents to leak out. It uses a dedicated T1SS composed of HylB (the ABC protein) and HylD (the integral membrane protein). Because HylA is not transported through the inner membrane by either the SecYEG channel or the Tat system, it does not contain a cleavable N-terminal signal sequence. Instead, like all proteins secreted by T1SS, it has a sequence at its carboxyl terminus that is recognized by the ABC transporter but, unlike a signal sequence, is not cleaved off as the protein is exported.
The TolC channel has been crystallized and its structure determined (see Koronakis et al., Suggested Reading). This structure has provided interesting insights into the structure of β-barrels in general and how they can be gated and opened to transport specific molecules. Briefly, three TolC polypeptides come together to form the channel through the outer membrane. Each of these monomers contributes four transmembrane domains to form a β-barrel that is always open on one side of the outer membrane, the side on the outside of the cell. In addition, each monomer has four longer α-helical domains that are long enough to extend all the way across the periplasm. These four α-helical domains contribute to the formation of a second channel that is aligned with the first channel and traverses the periplasm. Because of these two channels, the secreted protein can be transported all the way from the inner membrane to the outside of the cell. In addition, the channel in the periplasm can open and close and therefore “gate” the channel. When a protein is being transported and the TolC channel is recruited, the α-helical domains of the periplasmic channel may rotate, which untwists them and opens the gate on the periplasmic side. The molecule is then secreted all the way through both channels to the outside of the cell.
TYPE II SECRETION SYSTEMS
Type II secretion systems (T2SS) are very complex, consisting of as many as 15 different proteins (Figure 2.39). Most of these proteins are in the inner membrane and periplasm, and only 1 is in the outer membrane, where 12 of the secretin polypeptides come together to form a large β-barrel with a pore large enough to pass already folded proteins. The formation of this channel requires the participation of normal cellular lipoproteins that may become part of the structure. The secretin protein has a long N terminus that extends through the periplasm to make contact with other proteins of the T2SS in the inner membrane. This periplasmic portion of the secretin may also gate the channel, as with the TolC channel.
Even though many of the components of the T2SS are in the inner membrane, they use either the SecYEG channel or the Tat pathway to get their substrates through the inner membrane. Therefore, proteins transported by this system have either the Sec type or the Tat type of cleavable signal sequences at their N termini. Protein folding is usually completed in the periplasm before transport through the outer membrane. Some of the periplasmic and inner membrane proteins of the secretion system are related to components of pili and have been called pseudopilin proteins (see chapter 4). It has been proposed that the formation and retraction of these pseudopili work like a piston to push the protein through the secretin channel in the outer membrane to the outside of the cell. In this way, the energy for secretion could come from the inner membrane or the cytoplasm, as shown in the figure, since, as mentioned above, there is no source of energy in the periplasm. In support of this model, the pseudopili have been seen to produce pili outside an E. coli cell when the gene for the pilin-like protein was cloned and overproduced in E. coli.
Some examples of proteins secreted by T2SS are the pullulanase of Klebsiella oxytoca and the cholera toxin of Vibrio cholerae. The pullulanase degrades starch, and the cholera toxin is responsible for the watery diarrhea associated with the disease cholera (see chapter 12). The cholera toxin is composed of two subunits, A and B, and after transport by the SecYEG channel, one of the A and five of the B subunits assemble in the periplasm, followed by secretion through the secretin channel and into the intestine of the vertebrate host. The associated B subunit then assists the A subunit into mucosal cells, where it ADP-ribosylates (adds ADP) to a membrane protein that regulates the adenylate cyclase. This disrupts the signaling pathways and causes diarrhea. T2SS are also related to some DNA transfer systems used in transformation (see chapter 6) and are closely related to the systems that assemble type IV pili on the cell surface (see below).
TYPE III SECRETION SYSTEMS
Type III secretion systems (T3SS) are probably the most impressive of the secretion systems (see Galan and Waksman, Suggested Reading). They form a syringe-like structure composed of about 20 proteins, which takes up virulence proteins called effectors from the cytoplasm of the bacterium and injects them directly through both membranes into a eukaryotic cell (Figure 2.39). For this reason, they are sometimes called injectisomes. They exist in many Gram-negative animal