From the Producer Microorganisms to the Lasso Scaffold - Lasso Peptides: Bacterial Strategies to Make and Maintain Bioactive Entangled Scaffolds - page 21

sequences of four cysteines that are assembled into two disulfide bonds that link
the ring with the loop and the tail. Class II peptides do not contain disulfides and
class III has a single disulfide that links the ring with the tail. The N-terminal amino
acid that is engaged in the macrolactam is Cys in class I, while it is Gly in classes II
and III. Among the 35 representatives of the lasso peptide family described to date,
the class II is the most largely represented, with 30 members either from Proteo-
bacteria (23) or Actinobacteria (7). Class III contains only one representative cur-
rently. Worth of note, class I and class III lasso peptides have never been identified
in Proteobacteria and are exclusively exemplified in Actinobacteria (six peptides)
(Table 2.5 ). The Cys connectivity appears to be invariable in class I peptides: the
disulfide bonds are formed between Cys1/Cys13 and Cys7/Cys19. In BI-32169 the
disulfide linkage is between Cys6 and Cys19 (Table 2.4 ).
The first lasso peptide to be identified was anantin and was discovered in 1991,
when Weber et al. described the formation of the macrolactam ring (Weber et al.
1991 ; Table 2.4 ). However, the first description of the 3D lasso topology was only
published in 1994 for the class I lasso peptide RP-71955 (also called aborycin)
isolated from a Streptomyces sp. as an anti-HIV agent (Frechet et al. 1994 ). The
macrolactam linkage and cysteine bond pairing stabilizing the lasso topology were
clearly demonstrated by NMR. A few other class I lasso peptides were further iden-
tified, such as siamycins I and II (Constantine et al. 1995 ; Katahira et al. 1996 ;
Yano et al. 1996 ), and sviceucin (Ducasse et al. 2012a ) (also termed as SSV 2083
(Kersten et al. 2011 )).
Most lasso peptides currently known are of class II (30 out of 35 representa-
tives). Sequence alignments do not reveal conserved elements, except the Gly1 and
Glu/Asp residues participating in the cyclization. The first class II lasso peptides
have been isolated since 1991 with anantin (Weber et al. 1991 ), followed by mi-
crocin J25 (MccJ25; Salomón and Farías 1992 ) and RES-701 peptides (Morishita
et al. 1994 ; Ogawa et al. 1995 ; Katahira et al. 1996 ), but their lasso topologies were
not established MccJ25 was studied in the course of microcin research. Microcins
are potent antibacterial peptides produced by Enterobacteria that are active against
closely related bacteria and contribute to bacterial competitions in the intestinal
tract (for a review see Duquesne et al. 2007 ). The MccJ25 structure was subject to
debate before its lasso topology was firmly established (Blond et al. 1999 , 2001 ;
Bayro et al. 2003 ; Rosengren et al. 2003 ; Wilson et al. 2003 ; Rebuffat et al. 2004 ).
Because of its production by Escherichia coli and the only characterized gene clus-
ter until 2008, MccJ25 became and still remains the archetype for lasso peptides.
Lasso peptides identified to date contain 15-24 proteinogenic amino acid resi-
dues assembled by conventional peptidic bonds and one isopeptide bond that closes
the macrolactam ring by linking the Glu/Asp side-chain carboxylate to the N-ter-
minus (Table 2.4 ). The N-terminal residue of the currently known lasso peptides is
either Cys or Gly. However, genome analyses permitted to propose that lasso pep-
tides with other amino acids at position 1, such as serine, may exist (Severinov et al.
2007 ). The side-chain carboxylate that is involved in ring closure may arise either
from an Asp at positions 8 or 9 or a Glu at positions 7, 8 or 9. Interestingly, predic-
tions from genome analyses indicated the only suitable amino acid for macrolactam