56
3 Single Molecule Non-cleavable Multiply Active Antibacterials
Structure—based design
Structural information on targets can be used in a number of areas to inform the
design of multi-mechanism hybrid antibacterials. As for other medicinal agents, this
design process is a rational one, if the target structures are known or can be modelled.
Quite often though a combination of a rational and more intuitive design approach
is employed, or an intuitive/analogue based approach. While rational target-based
design has many positive aspects, it is also worth going beyond this and include
intuitive or ‘counter-intuitive’ aspects of design alongside computer-based target
design. Using both biologically as well as non-biologically-sourced structural toolboxes to express design ideas can also be very powerful resulting in what might be
termed biological—biological, biological—non-biological, or non-biological—nonbiological-type hybrids while tapping into a wider chemical structural space. Finding
or confirming putative targets for the new hybrids or prodrug-derived compounds
might also be assessed in the first instance using the Scifinder
® database looking
for similarities in structures between the proposed compounds and structures of
compounds in the database with known bacterial targets or modes of action. Careful
antibacterial testing would then need to follow to confirm activity modes or otherwise. Although in a different area, the structural similarity check aspect of this type
of strategy has been used to successfully find presumed cellular pathways or receptors for hit compounds from a cell-based ROS assay, which subsequently identified
the sigma-1-receptor as a potential new drug target for the treatment of the rare but
devastating vanishing white matter disease (Atzmon et al. 2018). The Scifinder
®
database was used in this work for the similarity checking. Automated procedures
are being developed to predict promiscuity which could be useful in complementing
and refining the similarity searching (Hopkins 2009).
In addition, the automated design of ligands with polypharmacological profiles
using adaptive design procedures is gaining ground and could be applied in principle
to poly-active antibacterials where good structures for the bacterial targets are available. Besnard et al. (2012) have discussed this evolutionary approach starting from
a clinically used anticholinesterase inhibitor (Donepezil) and evolving a range of
ligands which could access the brain and display either specific polypharmacology
or highly selective G-protein-coupled receptor features.
Fragment-based design
A knowledge of drug structures in combinations can be used to underpin at least the
initial design of single multi-action agents. Simpler structural fragments containing
the essential pharmacophoric units can then be identified and checked for activity
prior to subsequent combining of the fragments and further testing. In this way
potential multi-targeting hybrids can be accessed (Bolognesi 2013). The combination
of two fragments may involve linking, fusion or merging processes. Linking is a
common approach in general and one example of this in the non-antibacterial area is
through use of an exocyclic double bond to link two separate pharmacophoric units
(Gandini et al. 2018).
3 Single Molecule Non-cleavable Multiply Active Antibacterials
Structure—based design
Structural information on targets can be used in a number of areas to inform the
design of multi-mechanism hybrid antibacterials. As for other medicinal agents, this
design process is a rational one, if the target structures are known or can be modelled.
Quite often though a combination of a rational and more intuitive design approach
is employed, or an intuitive/analogue based approach. While rational target-based
design has many positive aspects, it is also worth going beyond this and include
intuitive or ‘counter-intuitive’ aspects of design alongside computer-based target
design. Using both biologically as well as non-biologically-sourced structural toolboxes to express design ideas can also be very powerful resulting in what might be
termed biological—biological, biological—non-biological, or non-biological—nonbiological-type hybrids while tapping into a wider chemical structural space. Finding
or confirming putative targets for the new hybrids or prodrug-derived compounds
might also be assessed in the first instance using the Scifinder
® database looking
for similarities in structures between the proposed compounds and structures of
compounds in the database with known bacterial targets or modes of action. Careful
antibacterial testing would then need to follow to confirm activity modes or otherwise. Although in a different area, the structural similarity check aspect of this type
of strategy has been used to successfully find presumed cellular pathways or receptors for hit compounds from a cell-based ROS assay, which subsequently identified
the sigma-1-receptor as a potential new drug target for the treatment of the rare but
devastating vanishing white matter disease (Atzmon et al. 2018). The Scifinder
®
database was used in this work for the similarity checking. Automated procedures
are being developed to predict promiscuity which could be useful in complementing
and refining the similarity searching (Hopkins 2009).
In addition, the automated design of ligands with polypharmacological profiles
using adaptive design procedures is gaining ground and could be applied in principle
to poly-active antibacterials where good structures for the bacterial targets are available. Besnard et al. (2012) have discussed this evolutionary approach starting from
a clinically used anticholinesterase inhibitor (Donepezil) and evolving a range of
ligands which could access the brain and display either specific polypharmacology
or highly selective G-protein-coupled receptor features.
Fragment-based design
A knowledge of drug structures in combinations can be used to underpin at least the
initial design of single multi-action agents. Simpler structural fragments containing
the essential pharmacophoric units can then be identified and checked for activity
prior to subsequent combining of the fragments and further testing. In this way
potential multi-targeting hybrids can be accessed (Bolognesi 2013). The combination
of two fragments may involve linking, fusion or merging processes. Linking is a
common approach in general and one example of this in the non-antibacterial area is
through use of an exocyclic double bond to link two separate pharmacophoric units
(Gandini et al. 2018).