Abstract:
Glycopeptide antibiotics (GPAs) - such as vancomycin or teicoplanin - are valuable natural
products that have been used to fight bacterial infections by inhibiting cell wall biosynthesis.
Numerous members of the genera Amycolatopsis and Streptomyces within the phylum of
Actinomycetota produce GPAs as part of their secondary metabolism. GPAs usually consist
of a heptapeptide backbone decorated with a variety of different modifications, including
glycosylation, halogenation, or acylation, which modulate their biological activity. The
backbone consists of various proteinogenic and non-proteinogenic amino acids, the
composition of which provides the basis for the general classification of GPAs into five
subgroups. Biosynthesis is carried out by non-ribosomal peptide synthetases (NRPS) and
modifying enzymes. These are encoded in a biosynthetic gene cluster (BGC) together with
enzymes that mediate the supply of precursor building blocks, resistance, and export. Although
the biosynthesis of some GPAs has been well studied, the nature of their export is still poorly
understood.
Here, I characterized several GPA transporters and identified them as type IV ABC
transporters using bioinformatic analysis and structure prediction. Despite high similarity in
sequence and structure, phylogenetic analysis suggested that GPA transporters do exhibit
differences, reflecting the diversity of their predicted substrates. This was confirmed using an
in vivo export assay, which showed that Tba, the transporter of the type I GPA balhimycin, Tri,
the putative transporter of the type III GPA ristomycin, and Tva, the putative transporter of the
type I GPA vancomycin, exhibit selective specificity for their cognate substrates based on
backbone composition rather than modifications. Moreover, analysis of the local cellular
environment of Tba in Amycolatopsis balhimycina showed that many biosynthetic enzymes
are in close proximity to the transporter and potentially form a microcompartment. Coimmunoprecipitation of biosynthetic enzymes with the transporter Tba supports this hypothesis
and indicates that there are even specific protein-protein interactions. It is conceivable that
these interactions are crucial for the transport-dependent biosynthesis of balhimycin. To
characterize the binding and transport kinetics of Tba and Tri in vitro, I also optimized their
expression and purification in E. coli. Due to challenges, e.g., in achieving high expression
levels, incorrect membrane insertion, proteolysis, and inefficient detergent extraction, the
expression host was changed. It became apparent that the native host of Tba,
Amycolatopsis balhimycina, or a closely related organism, Streptomyces lividans, was more
suitable for the expression of the GPA transporters Tba and Tri than the commonly used host
Escherichia coli.
In conclusion, these findings add to the limited knowledge of GPA export and demonstrate that
GPA transporters are functionally integrated into the biosynthetic process of their cognate
substrates. This work provides the basis for further single-protein studies of GPA ABC
transporters, which can significantly broaden our knowledge of the substrate specificity of
transport proteins. In addition, a detailed analysis of the interactions between Tba and the
biosynthetic enzymes could shed light on the role of membrane-associated bacterial
microcompartments. A better understanding of the transport process will furthermore help to optimize the microbial production of GPAs in an industrial setting to increase the export of medically relevant GPAs.
ABC transporter