By Tim Sandle, Ph.D.
As has been widely reported, the rise in antimicrobial resistant microorganisms presents a major global health challenge. Many pathogens have developed multi-drug resistant strains, partly through antibiotic overuse and partly due to a slowdown in drug innovation. Consequently, new generations of antimicrobials are urgently needed.
A major challenge with antimicrobial drug development is the necessity to test millions of drug compounds in the hope of finding one that will kill a given microbial pathogen. As part of the search for new antimicrobial compounds designed to be effective against emerging drug-resistant bacteria, laboratories are using robotic chemical-synthesizing machines to screen metabolites of both natural and synthetic origin, together with automated solutions to try and grow previously unculturable organisms that might be sources of antimicrobial compounds. These instruments are designed to be rapid, able to assemble dozens of compounds at a time and allowing them to be tested as potential drugs or to use a combination of methods to promote growth.
The advantage of these technologies is that they overcome the more tedious aspects of chemical experimentation, freeing scientists to spend time on the more analytical aspects of their work. Another advantage is their ability to speed up the analysis of multiple compounds. A third advantage is that they enable scientists to reproduce reactions that others have developed without having to reoptimize every step of the synthesis.
Automated Synthesis Systems
There are three potential sources of novel antimicrobials: what is found in the natural world (such as from plants or microorganisms isolated from soil) that can be synthesized; new compounds that can be developed from scratch; and reassessment of previously developed compounds or drugs to determine if they have antimicrobial action (such as reassessing an anti-cancer medication).
Assessing these sources and making meaningful progress is challenging. Taking just one aspect of the natural world — plants — there are millions of plant species, and each species has its own metabolome (the product of the plant's metabolism). Only around 5% of all these natural products have been characterized. To advance analyses forward, new instrumentation is required, together with considerable computing power.
Automated synthesis systems are types of laboratories that work through the testing of compounds or that grow and screen microorganisms through a linear combination of steps. Each of the individual steps can be modularized into hardware that accomplishes the specific analysis. The analysis is often supported by innovations with artificial intelligence.
Some instruments enable continuous flow. With this approach, chemical reagents flow through a series of tubes, and new chemicals can be added at different points. This presents an advantage over traditional "batch chemistry," where each step needs to be undertaken separately and human intervention is needed to move the process along to the next stage.1 Many of the more recent systems have decreased in size, enabling smaller quantities of chemicals to be used and creating instrumentation that can be adopted by a larger range of laboratories.
A considerable part of the research involves examining previously identified compounds. Screening large "libraries" of compounds to find those with a desired biological activity is a powerful method for discovering new drugs. Many of these compounds are applied for purposes different than what they were originally designed for. Such compounds present advantages for widening the pipeline for antimicrobial drug discovery since many compounds will work differently than traditional antibiotics.2
Once candidate compounds are selected, they need to be screened against pathogens of concern. High-throughput screening (HTS) has become the preferred method to assess a selection of drug candidates against different microorganisms. These require automation, given the scale of testing required. The recent miniaturization of screening platforms has provided new opportunities for screening, saving time and cost. Robotics that take just a few microliters of working solution per well enable high-density microplates to be used. Some technologies now permit the use of pico-droplet platforms. Microbial growth, inhibition, or death against different compounds can be assessed by using techniques like fluorescent microcolony screening, optical density assessments, or through the detection of galactosidase-producing microcolonies.
Harnessing Artificial Intelligence
The automation and analysis process is enhanced through innovations in artificial intelligence. Large and complex data sets are analyzed using algorithms, which can be enhanced through the improvements realized by machine learning.3 Algorithms can predict the molecular formula of metabolites and classify them by type and identity. This form of AI also can predict the substructure of unknown metabolites and based on similarities in structure, connect them to previously characterized metabolites. This process helps predict their functions and to identify candidate materials for further assessment, possibly leading toward drug development.
Machine learning also can be used to further predict the mechanism of action and potency of different antibiotics.4
These computational approaches are leading to interesting developments. For example, a new class of antibiotics (benzoxazinoids) has been characterized from rice and a second class (glycoalkaloids) isolated from allium.5 Perhaps the most important new class of antimicrobials to be discovered in recent years — through automated miniaturized screening methods — is teixobactin. This compound can inhibit bacterial cell wall synthesis by binding to important pathways, lipid II (precursor of peptidoglycan) and lipid III (precursor of cell wall teichoic acid), and, hence, prevent bacterial growth. The discovery was facilitated by using a multichannel device called the iChip10 (isolation chip version 10).6 This enabled researchers to simultaneously isolate and grow uncultured bacteria isolated from soil samples.7
One of the more interesting of the novel antimicrobials are peptides and peptoids (synthetic versions of peptides). These mimic the ability of the human body to fight infections, and many more recent antimicrobial drugs are based on peptides. High-density peptide arrays enable up to 1 million reactions to be tested using on-chip solid-phase synthesis, and such technology has pushed forward the hunt for antimicrobial peptides.
Developing and manufacturing peptides is time-consuming, and testing different iterations has slowed the development of new peptide drugs. Automated synthesis systems accelerate this process by forming links between amino acids (the buildings blocks of proteins). This enables the generation of complete peptide molecules in under one hour. Technologies capable of delivering novel peptides within short time periods are based on flow chemistry. With such systems, chemicals flow through a series of modules that each perform one step of the overall synthesis.
A scientist, or more commonly the machine algorithm itself, selects a target amino acid sequence. The automated process releases the amino acids in the desired order into a module where they are briefly heated to make them more chemically reactive. The activated amino acids flow into a chamber where they are added to the growing peptide chains. After synthesis, small peptides can be joined together to form larger proteins.
Although automated processing solutions present many advantages, there remain issues that affect the reliability and robustness of the instruments. These include the impact of working with small volumes, including evaporation and liquid adhesion, compatibility with laboratory equipment, and synergies around transferring compounds and microbial libraries to different systems. There are also opportunities to improve the algorithms used to set up and review experimental data as well as the overarching issue of the hunt for novel candidate compounds to test against the pathogens of greatest risk and, hopefully, through to clinical trials.
- Bédard, A-C. et al. Reconfigurable system for automated optimization of diverse chemical reactions. Science, 2018; 361 (6408): 1220
- Salam, A. et al. From the Leaves of Castanea sativa Inhibits Virulence in Staphylococcus aureus. Frontiers in Pharmacology, 2021; 12 DOI: 10.3389/fphar.2021.640179
- Ide, Y. et al. Machine Learning-Based Analysis of Molar and Enantiomeric Ratios and Reaction Yields Using Images of Solid Mixtures. Industrial & Engineering Chemistry Research, 2023; DOI: 10.1021/acs.iecr.3c01882
- Fields F.R. et al. Novel antimicrobial peptide discovery using machine learning and biophysical selection of minimal bacteriocin domains. Drug Dev. Res. 2020;81:43–51
- Tsugawa, H. et al. A cheminformatics approach to characterize metabolomes in stable-isotope-labeled organisms. Nature Methods, 2019; 16 (4): 295
- Nichols D et al. Use of ichip for high-throughput in situ cultivation of “uncultivable” microbial species. Appl. Environ. Microbiol 2010, 76, 2445–2450
- Ling, L. L. et al. A new antibiotic kills pathogens without detectable resistance. Nature 2015, 517, 455–459
About The Author:
Tim Sandle, Ph.D., is a pharmaceutical professional with wide experience in microbiology and quality assurance. He is the author of more than 30 books relating to pharmaceuticals, healthcare, and life sciences, as well as over 170 peer-reviewed papers and some 500 technical articles. Sandle has presented at over 200 events and he currently works at Bio Products Laboratory Ltd. (BPL), and he is a visiting professor at the University of Manchester and University College London, as well as a consultant to the pharmaceutical industry. Visit his microbiology website at https://www.pharmamicroresources.com.