**2.1 Synthesis and structure of HlyA**

The synthesis, maturation, and secretion of *E. coli* HlyA are determined by the *hlyCABD*  operon ((Felmlee *et al*, 1985), (Issartel *et al.*, 1991), (Koronakis *et al.,* 1997), (Nieto *et al.* ,1996)). The membrane-associated export proteins are synthesized at a lower level than the cytosolic HlyC and pro-HlyA, in part because of transcription termination within the *hlyCABD*  operon (Felmlee *et al*, 1985). This termination is suppressed by the elongation protein RfaH and a short 59-bp, *ops* (operon polarity suppressor) (Bailey *et al.,* 1992, 1996), (Cross *et al.,* 1990), (Nieto *et al*., 1996) that act in concert to allow the transcription of long operons such as *hly*, *rfa*, and *tra* encoding the synthesis and export of extracellular components key in the virulence and fertility of Gram-negative bacteria (Bailey *et al.*, 1992, 1997).

The structural gene *hlyA* produces a single 110-kDa polypeptide. The estimated pI of the toxin is 4.5, with this characteristic being common among the RTX toxins. The N-terminal hydrophobic domain is predicted to contain nine amphipathic α-helices (Soloaga *et al.,* 1999). Using photoactivable liposomes, Hyland *et al* (2001) demonstrated that the region comprised between residues 177-411 is the one that becomes inserted into membranes. The C-terminal calcium-binding domain contains 11-17 of the glycine- and aspartate-rich nonapeptide β-strand repeats. Although the membrane interaction of HlyA is assumed to occur mainly through the amphipathic α-helical domain, that both major domains of HlyA are directly involved in the membrane interaction of HlyA has recently been proposed, with the calcium-binding domain in particular being responsible for the early stages of the HlyA's docking to the target membrane (Sanchez-Magraner *et al.*, 2007).

The topic of the existence of a receptor for the toxin in erythrocytes remains quite controversial. Nevertheless, Cortajarena *et al* (2003) observed that a short sequence from the C-terminal domain (between residues 914–936) was the main HlyA segment that bound to the glycophorin A on erythrocytes.

The last 60 C-terminal amino acids consist of 2 α-helices separated by 8-10 charged residues. This domain is implicated in the transport of the toxin to the extracellular medium (Hui *et al.,* 2000). Fig. 1 shows a scheme of the HlyA structure.

Fig. 1. A scheme of the HlyA structure.

The more relevant domains of HlyA are indicated.

### **2.2 The posttranslational activation of HlyA**

The proHlyA protoxin is matured in the cytosol to the active form by HlyC-directed fatty acylation before export from the toxin-producing bacteria. This process consists in a posttranslational modification of the ε-amino groups of internal lysine residues by covalent attachment of amide-linked fatty-acyl residues. This reaction is catalyzed by the HlyC acyltransferases expressed together with the protoxins (Goebel & Hedgpeth, 1982). The mechanism of this novel type of protein acylation was extensively analyzed for HlyA (Issartel *et al.,* 1991), (Stanley *et al.,* 1994). HlyC uses the fatty-acyl residues carried by acylcarrier protein (ACP) to form a covalent acyl-HlyC intermediate, which species then transfers the fatty-acyl residues to the ε-amino groups of the Lys 564 and Lys 690 residues of proHlyA (Worsham *et al*., 2001, 2005). ACPs carrying various fatty-acyl residues including palmitate (16:0) and palmitoleate (16:1), the most common in *E. coli*—could be efficiently used *in vitro* as acyl donors for the modification of HlyA ((Issartel *et al.*, 1991), (Trent *et al.*, 1998)). *In vivo*, however, HlyC exhibits a high selectivity for myristic acid (14:0), which species was found to constitute about 68% of the acyl chains covalently linked to Lys 564 and Lys 690 of the native HlyA (Lim *et al.*, 2000). Contrary to expectations, the extremely rare odd-carbon saturated fatty-acyl residues 15:0 and 17:0 were found to constitute the rest of the *in-vivo* acylation of HlyA in two different clinical *E. coli* isolates (Lim *et al.*, 2000). Both acylation sites in the HlyA genome function independently of one another with respect to the kinetics of their interaction with acyl-HlyC (Langston *et al.*, 2004). By using deleted protoxin variants and protoxin peptides as substrates in an *in-vitro* maturation reaction dependent on only HlyC and acyl-ACP, two independent HlyC-recognition domains were identified on the HlyA protoxin, each of which spanned one of the target lysine residues (Stanley *et al.*, 1996). Each domain required 15 to 30 amino acids for basal recognition and 50 to 80 for full wild-type acylation, but HlyC recognized a large topology rather than a linear sequence. The loss of the Lys 564 acylation site either by mutation or structural deletion affected the thermodynamics of the acylation reaction at Lys 690, implying an undefined connectivity between the two acylation sites (Worsham *et al.*, 2005). Nevertheless, the intact acylation at Lys 690 is essential for HlyA activity.

The proHlyA protoxin is matured in the cytosol to the active form by HlyC-directed fatty acylation before export from the toxin-producing bacteria. This process consists in a posttranslational modification of the ε-amino groups of internal lysine residues by covalent attachment of amide-linked fatty-acyl residues. This reaction is catalyzed by the HlyC acyltransferases expressed together with the protoxins (Goebel & Hedgpeth, 1982). The mechanism of this novel type of protein acylation was extensively analyzed for HlyA (Issartel *et al.,* 1991), (Stanley *et al.,* 1994). HlyC uses the fatty-acyl residues carried by acylcarrier protein (ACP) to form a covalent acyl-HlyC intermediate, which species then transfers the fatty-acyl residues to the ε-amino groups of the Lys 564 and Lys 690 residues of proHlyA (Worsham *et al*., 2001, 2005). ACPs carrying various fatty-acyl residues including palmitate (16:0) and palmitoleate (16:1), the most common in *E. coli*—could be efficiently used *in vitro* as acyl donors for the modification of HlyA ((Issartel *et al.*, 1991), (Trent *et al.*, 1998)). *In vivo*, however, HlyC exhibits a high selectivity for myristic acid (14:0), which species was found to constitute about 68% of the acyl chains covalently linked to Lys 564 and Lys 690 of the native HlyA (Lim *et al.*, 2000). Contrary to expectations, the extremely rare odd-carbon saturated fatty-acyl residues 15:0 and 17:0 were found to constitute the rest of the *in-vivo* acylation of HlyA in two different clinical *E. coli* isolates (Lim *et al.*, 2000). Both acylation sites in the HlyA genome function independently of one another with respect to the kinetics of their interaction with acyl-HlyC (Langston *et al.*, 2004). By using deleted protoxin variants and protoxin peptides as substrates in an *in-vitro* maturation reaction dependent on only HlyC and acyl-ACP, two independent HlyC-recognition domains were identified on the HlyA protoxin, each of which spanned one of the target lysine residues (Stanley *et al.*, 1996). Each domain required 15 to 30 amino acids for basal recognition and 50 to 80 for full wild-type acylation, but HlyC recognized a large topology rather than a linear sequence. The loss of the Lys 564 acylation site either by mutation or structural deletion affected the thermodynamics of the acylation reaction at Lys 690, implying an undefined connectivity between the two acylation sites (Worsham *et al.*, 2005). Nevertheless, the intact acylation

Fig. 1. A scheme of the HlyA structure.

The more relevant domains of HlyA are indicated.

**2.2 The posttranslational activation of HlyA** 

at Lys 690 is essential for HlyA activity.

No other HlyA sequences are required for toxin maturation, including the immediately Cterminal Ca+2 binding repeats. Indeed, *in vitro*, Ca+2 ions prevent acylation at both sites (Stanley *et al.*, 1996). The extreme sensitivity of the proHlyA activation reaction to free Ca+2 supports the view that intracellular Ca+2 levels in *E. coli* are too low to affect toxin activity and that Ca+2 binding does not occur until the toxin is outside the cell.

This posttranslational modification is remarkable because the behavior of the protein is changed by lipid modification from a benign protein to a frank toxin—part of this transformation being an exclusive mechanism in prokaryotes since in only a few eukaryotic proteins is this type of acylation found (for example, in the nicotinic acetylcholine receptor; the insulin receptor; and cytokines such as TNF-α, IL-1α and IL-1β) (Stanley *et al.*, 1998). In the following section we discuss the role that these covalently bound fatty acids play in the toxin's mechanism of action.

#### **2.3 The secretion of HlyA into the extracellular medium**

Maturation increases the hydrophobicity of the protein, but that property is not required for export (Ludwig et al. 1987). *E. coli* HlyA-related toxins are all secreted across both membranes by the type-I export process employing an uncleaved C-terminal recognition signal (Nicaud *et al.*, 1986), (Stanley *et al.,* 1991), but no N-terminal leader peptide (Felmlee *et al.*, 1985) or periplasmic intermediate (Felmlee & Welch, 1988), (Koronakis *et al.,* 1989). The HlyA secretory apparatus comprises HlyB (an inner-membrane traffic ATPase, the ATPbinding cassette), HlyD (a membrane-fusion protein), and TolC (an outer-membrane protein) (Schulein *et al.*,1992), (Wandersman & Delepelaire, 1990), (Wang *et al.,* 1991). In *E. coli* and most other pathogens, TolC is encoded by a separate gene from *hlyCABD.* As mentioned before (*cf*. **Section 2.0**) the type-I-secretion–signal sequences have been located within the last 60 C-terminal amino acids, consisting of 2 α-helices separated by 8-10 charged residues (Hui *et al.*, 2000).

The mechanism of exportation of HlyA is as follows: The trimeric accessory protein HlyD has been proposed to form a substrate-specific complex with the inner-membrane protein HlyB, which latter species subsequently recognizes the C-terminal signal peptide of HlyA. Upon the binding of HlyA, the HlyD trimer interacts with the trimeric TolC protein of the outer membrane, inducing a conformational change and the consequent export of HlyA. This assembly between the complex HlyB-HlyD with TolC very likely occurs because, as has been demonstrated by X-ray crystallography, the trimeric complex of TolC is very similar in size to the trimeric structure of HlyD, thus facilitating the formation of a continuous transperiplasmic export channel through which HlyA can pass (Koronakis *et al.,* 2000). This complex appears to be transient, with it disengaging and reverting to a resting state once the substrate has been transported (Thanabalu *et al.,* 1998). The energy necessary for the secretion process depends not only on ATP hydrolysis mediated by HlyB but also on the proton motive force exerted on the inner membrane (Koronakis *et al.,* 1991, 1995). Type-I secretion is generally assumed to involve the translocation of unfolded proteins (Young & Holland, 1999), although Pimenta *et al* (2005) have suggested that contact with HlyD directly or indirectly affects the folding of HlyA either during the latter's transit through the translocator or afterwards.

In the last decade many researchers have been interested in this type of secretion machinery because of its potential use in the export of chimeric proteins and in vaccine production (Gentschev *et al.,*1996, 2002).

Although HlyA has its own machinery for export from the bacteria, the presence of a physiologically active HlyA in the outer-membrane vesicles (OMVs) of clinical-hemolytic (Balsalobre *et al.*, 2006) as well as laboratory-recombinant strains of *E. coli* (Herlax *et al.,* 2010) has recently been demonstrated.

OMVs are constantly being discharged from the surface of Gram-negative bacteria during bacterial growth. All Gram-negative bacteria studied to date, including *E. coli*, produce OMVs; and their release is increased when the bacteria are exposed to stressful conditions such as antibiotics or serum. Even though the release of OMVs could not be demonstrated *in vivo*, the presence of particles resembling those vesicles has, in fact, been detected in plasma from patients with different infectious processes (Beveridge, 1999). OMVs serve as secretory vehicles for the proteins and lipids of Gram-negative bacteria and in this manner play roles in establishing a colonization niche for carrying or transmitting virulence factors into host cells or otherwise modulating the host defense and response, thus acting as well as longrange virulence factors that can protect luminal cargo from extracellular host proteases and so penetrate into tissues more readily than the larger bacteria (Kuehn & Kesty, 2005). In addition to toxin-protein delivery, other roles have been characterized for OMVs—namely, interspecies interaction and communication during multispecies infections plus DNA uptake and transfer (Mashburn-Warren & Whiteley, 2006). In the particular example of HlyA, we have demonstrated that the toxin secreted in this way is transferred to the target cell in a concentrated manner and as such is more hemolytically efficient than the free HlyA (Herlax *et al.*, 2010). Moreover, Balsalobre *et al.* (2006) demonstrated that the HlyA associated with OMVs is protected from the attack of proteases, thus facilitating the survival of the toxin within the adverse medium of a patient's plasma.
