Available online at www.sciencedirect.

com

ScienceDirect

Mining for novel antibiotics

Justin R Randall and Bryan W Davies
Recent work has highlighted the diversity of compounds
Address that can be found in native microbiomes. Investigation of
Department of Molecular Biosciences, University of Texas at Austin,
Austin, TX, USA bacteria from the nematode microbiome led to the dis-
covery of darobactin, a compound with a new mechanism
Corresponding author: Davies, Bryan W ([email protected]) of action that targets the essential Gram-negative outer
membrane protein BamA [8]. Staphylococcus lugdunensis
Current Opinion in Microbiology 2021, 63:66–69
from the human nose was shown to produce lugdunin,
which inhibits the growth of Staphylococcus aureus [9]. Our
This review comes from a themed issue on Antimicrobials
own microbiome has also been found to produce a wide
Edited by David Weiss and Marcin Grabowicz spectrum of natural products, many with antibacterial
activity [10]. In addition to exploration of microbiomes
and terrestrial niches [11,12], the marine environment is
also proving a rich source of antibiotic leads [13,14],
though in all cases issues of cultivability remain.
https://doi.org/10.1016/j.mib.2021.06.001
The vast majority of bacteria are difficult or impossible to
1369-5274/ã 2021 Elsevier Ltd. All rights reserved.
cultivate under standard laboratory conditions historically
used for antibiotic identification. Furthermore, the gene
clusters responsible for antibiotic production frequently
remain silent until they receive proper stimuli. Both of
these hurdles prevent facile investigation of potentially
new antibiotics. Thus, while we push to explore new
Keeping pace with the development and spread of drug
environments we also need to advance methodologies
resistant bacteria will require a strong and diverse antibi-
used to grow and trigger bacteria to produce any hidden
otic pipeline. Currently, clinical phase development
antibiotics they may contain. New techniques are being
remains quite limited and narrow, consisting largely of
implemented to grow and test uncultivable samples in
derivatives from known antibiotics [1,2]. The modifica-
their natural environment both alone and in association
tion of existing antibiotics increases the lifetime and
with other microbes. The development of methods to
useful breadth of known antibiotics, but its ability to
grow bacteria in situ, such as the iCHIP diffusion cham-
keep up with antibiotic resistance is finite. Thus, discov-
ber, has facilitated the cultivation of previously unculti-
ery and clinical development of novel classes of antibio-
vable bacteria by allowing them access to the natural
tics with new targets is essential.
compounds required to meet their nutritional needs
[15,16]. This general approach aided in the discovery
Finding and advancing new drugs to treat disease is
of lassomycin, which targets the ATP-dependent prote-
shared by all therapeutic sectors. However, relatively
ase ClpC1P1P2 to kill Mycobacterium tuberculosis [12], and
poor financial returns on antibiotics make them a less
teixobactin, which has potent activity against Gram-
attractive investment [3,4]. Thus improving the effi-
positive bacteria through inhibition of cell wall synthesis
ciency with which we identify, select, and validate new
[11]. Microbes may only produce new compounds when
antibiotic leads for clinical investigation will be important
stimulated by other prokaryotic or eukaryotic organisms.
for reviving and maintaining a robust antibiotic pipeline.
Co-culturing approaches can provide the necessary sig-
naling molecules to help coax expression of latent gene
Natural reservoirs
clusters and their associated natural products [17–19].
Historically, the discovery of novel antibiotics has come
The advancement of approaches that closer approximate
from mining microorganisms (predominantly from soil)
native microbial environments is likely to continue to
for natural products with antimicrobial activity [5–7].
help identify novel antibiotics.
During the mid 20th century, this approach identified
the majority of antibiotic classes we use today. After an
initial burst of discovery, this approach quickly had Integration of omics data in antibiotic
diminished returns, as rediscovery of previously identi- discovery
fied antibiotics became the norm. However, this early As we expand our exploration of the natural world,
work investigated only a fraction of potential natural approaches that can focus our efforts will help increase
habitats and microbial sources for antibiotics. New the throughput and efficiency of antibiotic discovery.
microbe hunters are probing underexplored niches and Advances in DNA sequencing technology have been
microbes with encouraging results. instrumental in identifying new bacteria and gene

Current Opinion in Microbiology 2021, 63:66–69 www.sciencedirect.com

 Mining for novel antibiotics Randall and Davies 67

clusters from uncultured communities likely to produce hidden patterns associated with antibiotic activity. Unlike
new antibiotics. The use of genome-mining approaches previous computational exploration of drug design, these
can help prioritize communities for screening, and pro- new approaches do not require human theories of antibi-
vide routes for synthetic investigation. A recent method otic action and can find patterns in nature itself. Impor-
that incorporated phylogenic analysis of biosynthetic tantly, these methods can factor in aspects of toxicity and
genes along with a lack of known co-associated resistance large-scale synthesis before discovery, which is not possi-
determinates was used to predict glycopeptide antibiotics ble with traditional antimicrobial screening alone.
that were likely to have unique biological activities [20].
This process led to the identification of complestatin and A recent study developed a neural network that identified
corbomycin, which were shown to have a novel mecha- antibacterial compounds from a set of more than ten
nism of action that targets autolysins in Gram-positive million molecules [28]. Importantly, some of the iden-
bacteria. tified compounds were structurally distinct from known
antibiotics. This work demonstrates the power of com-
Once identified, heterologous expression of latent bio- bining experimental and in silico approaches to probe new
synthetic gene clusters can be used to produce new areas of antibiotic chemical space. Investigation of anti-
compounds in sufficient quantities for study and devel- bacterial peptides has also benefited greatly from new
opment, rather then relying on production from the computer-aided approaches. Ribosomally synthesized
parent organism [21,22]. Alternatively, as computational antimicrobial peptides have been advocated for years
prediction of gene cluster products improves, putative as potent antibacterials, but clinical success has been
antibiotics may be synthesized de novo based solely on largely elusive. However, databases of thousands of nat-
genomic information. Humimycins were identified by ural and synthetic antibacterial peptides that are now
genome mining of human-associated bacteria [10]. Based available can facilitate computationally aided de novo
on a bioinformatics prediction, select humimycins were generation and optimization of peptide antibiotics [29].
chemically synthesized and shown to be active against Recent studies have shown that design principles can be
several bacteria including Streptococcus and Staphylococcus applied to identify stable, active, and non-toxic peptides
species. to treat infection [30]. Furthermore, computational
approaches are also being developed to perform in silico
All of these advances offer fresh hope of identifying evolution and further increase the efficiency with which
antibiotic leads from natural sources. However, it remains we can modify leads for improved antibacterial activity
unclear how deep this well of clinically viable natural [31]. The distinction between small molecule and pep-
products goes. Increasing screening efficiency will be tide antibiotics is fluid as many old and newly discovered
important as some estimates suggest thousands if not antibiotics are based on a macrocyclic peptide core struc-
millions of bacteria will need to be investigated to find ture. However, our understanding of rules supporting
each new viable lead [5,6]. Once identified, production at macrocyclic peptide antibacterial activity is relatively
scale of new natural product scaffolds necessary for opti- small compared to their linear counterparts. Expansion
mization and eventually distribution can also be challeng- of peptide databases to fill this knowledge gap may
ing. Clearly, natural sources will play an important role in enable the development of more potent macrocyclic
identifying novel leads going forward. However, devel- peptide-based antibiotics. New approaches to generate
opment of additional strategies will both bolster natural and test diverse macrocyclic sequences will help meet
product discovery, as well as offer further paths to antibi- this need [32–34].
otic development that are needed to restore and maintain
a robust clinical pipeline. Beyond small molecules
Treatments in other therapeutic areas, such as oncology,
Computation provides new paths to discovery have expanded from small molecules, to biologics, to cell-
Past attempts to discover new antibiotics using high- based treatments in recognition that a one size fits all
throughput screening of small molecule chemical libraries model is not appropriate to target all factors of a disease.
against defined targets proved largely unsuccessful While antibiotic development has largely been confined
[23,24]. However, the data from screens like these, and to small molecules, it also appears to be expanding to
the ever-growing list of active and inactive compounds include new modalities. Monoclonal antibodies have
generated over the past several decades, may provide been wildly successful in treating a range of diseases.
information needed to advance new routes to identify They have several unique properties compared to small
novel antibiotics. This importantly includes datasets molecules including generally low toxicity, high specific-
describing factors influencing the ability to bypass the ity, an ability to engage the immune system, and
bacterial membrane barrier in addition to antimicrobial extended half-life [35]. While efforts to identify antibo-
activity [25–27]. New computational approaches, such as dies with direct antibacterial activity have been challeng-
machine learning, now allow us to churn through the ing [36,37], other antibody-based approaches have shown
massive antimicrobial datasets at our disposal to look for promise. A bispecific antibody from MedImmune

www.sciencedirect.com Current Opinion in Microbiology 2021, 63:66–69

68 Antimicrobials

Figure 1 Acknowledgements
B.W.D. is supported by the National Institutes of Health (R01 AI125337,
New Sources Computational R01 AI148419, R21 AI159203), Defense Threat Reduction Agency
Novel niches (HDTRA1-17-C0008), the Welch Foundation (F-1870), and Tito’s
In situ cultivation
Approaches
Genome mining Handmade Vodka.
Co-culture
Expanded datasets
Bioloigcs
Design principles
Clinical References and recommended reading
Antibiotic Papers of particular interest, published within the period of review,
have been highlighted as:
Pipeline
 of special interest
 of outstanding interest

1. Theuretzbacher U, Gottwalt S, Beyer P, Butler M, Czaplewski L,

Lienhardt C, Moja L, Paul M, Paulin S, Rex JH et al.: Analysis of
Screening and Production the clinical antibacterial and antituberculosis pipeline. Lancet
High-throughput methods Infect Dis 2019, 19:e40-e50.
Heterologous expression
de novo synthesis 2. Shore CK, Coukell A: Roadmap for antibiotic discovery. Nat
Microbiol 2016, 1:16083.
Current Opinion in Microbiology
3. Projan SJ: Why is big Pharma getting out of antibacterial drug
discovery? Curr Opin Microbiol 2003, 6:427-430.
Exploration of new antimicrobial sources and advances in
computational approaches can synergize with advanced methods for 4. Renwick MJ, Brogan DM, Mossialos E: A systematic review and
critical assessment of incentive strategies for discovery and
high-throughput screening and production to help generate novel
development of novel antibiotics. J Antibiot (Tokyo) 2016,
antibiotics and expand a limited clinical pipeline. 69:73-88.
5. Baltz RH: Marcel Faber Roundtable: is our antibiotic pipeline
unproductive because of starvation, constipation or lack of
targeting the virulence factor PcrV and exopolysaccharide inspiration? J Ind Microbiol Biotechnol 2006, 33:507-513.
Psl of Pseudomonas aeruginosa has shown efficacy in mouse 6. Lewis K: The science of antibiotic discovery. Cell 2020,
models, and is preceding through clinical development 181:29-45.
[38,39]. Still other approaches combine the benefits of 7. Hutchings MI, Truman AW, Wilkinson B: Antibiotics: past,
antibody and small molecule approaches. Genentech present and future. Curr Opin Microbiol 2019, 51:72-80.
developed a novel antibody-antibiotic conjugate to com- 8. Imai Y, Meyer KJ, Iinishi A, Favre-Godal Q, Green R, Manuse S,
bat S. aureus. The antibody binds the bacterium, but the  Caboni M, Mori M, Niles S, Ghiglieri M et al.: A new antibiotic
selectively kills Gram-negative pathogens. Nature 2019,
antibiotic only becomes active once inside host cells 576:459-464
that have taken up the bacterium [40]. This antibody- This manuscript describes the discovery of a novel antibiotic from the
nematode microbiome.
antibiotic conjugate is also proceeding through clinical
development [41]. The natural repertoire of infected and 9. Zipperer A, Konnerth MC, Laux C, Berscheid A, Janek D,
Weidenmaier C, Burian M, Schilling NA, Slavetinsky C, Marschal M
convalescent patients may also prove a fertile ground for et al.: Human commensals producing a novel antibiotic impair
identifying protective antibodies [42] and help reveal new pathogen colonization. Nature 2016, 535:511-516.
bacterial epitopes that can be targeted to control infec- 10. Chu J, Vila-Farres X, Inoyama D, Ternei M, Cohen LJ, Gordon EA,
tion. A major hurdle in pursuing innovative new  Reddy BVB, Charlop-Powers Z, Zebroski HA, Gallardo-Macias R
et al.: Discovery of MRSA active antibiotics using primary
approaches is the lack of standardized proof-of-concept sequence from the human microbiome. Nat Chem Biol 2016,
assays [2]. In this regard it is encouraging to observe two 12:1004-1006
This manuscript describes the direct synthesis and activities of antibiotics
respected companies pursuing this new direction and it predicted from microbiome sequence analysis.
will be exciting to see how they validate and position
11. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I,
monoclonal treatments in the market. Conlon BP, Mueller A, Schäberle TF, Hughes DE, Epstein S et al.: A
new antibiotic kills pathogens without detectable resistance.
Nature 2015, 517:455-459.
The unceasing emergence and spread of antibiotic resis-
tant bacteria is beginning to leave the clinical antibiotic 12. Gavrish E, Sit CS, Cao S, Kandror O, Spoering A, Peoples A,
Ling L, Fetterman A, Hughes D, Bissell A et al.: Lassomycin, a
cupboard bare. Previously stalled developmental strate- ribosomally synthesized cyclic peptide, kills Mycobacterium
gies using natural sources and synthetic screening have tuberculosis by targeting the ATP-dependent protease
ClpC1P1P2. Chem Biol 2014, 21:509-518.
been reinvigorated by new technologies and computa-
tional approaches providing renewed hope (Figure 1). 13. Fenical WB: Marine microbial natural products: the evolution of
a new field of science. J Antibiot (Tokyo) 2020, 73:481-487.
Still, continued advancement and investment in the field
will be required to restore the clinical antibiotic pipeline 14. Schinke C, Martins T, Queiroz SCN, Melo IS, Reyes FGR:
Antibacterial compounds from marine bacteria, 2010-2015. J
back to its needed levels. Nat Prod 2017, 80:1215-1228.
15. Nichols D, Cahoon N, Trakhtenberg EM, Pham L, Mehta A,
Belanger A, Kanigan T, Lewis K, Epstein SS: Use of ichip for high-
Conflict of interest statement throughput in situ cultivation of “uncultivable” microbial
Nothing declared. species. Appl Environ Microbiol 2010, 76:2445-2450.

Current Opinion in Microbiology 2021, 63:66–69 www.sciencedirect.com

 Mining for novel antibiotics Randall and Davies 69

16. Kaeberlein T, Lewis K, Epstein SS: Isolating “uncultivable” 31. Porto WF, Irazazabal L, Alves ESF, Ribeiro SM, Matos CO,
microorganisms in pure culture in a simulated natural Pires ÁS, Fensterseifer ICM, Miranda VJ, Haney EF, Humblot V
environment. Science 2002, 296:1127-1129. et al.: In silico optimization of a guava antimicrobial peptide
enables combinatorial exploration for peptide design. Nat
17. Traxler MF, Watrous JD, Alexandrov T, Dorrestein PC, Kolter R: Commun 2018, 9:1490.
Interspecies interactions stimulate diversification of the
Streptomyces coelicolor secreted metabolome. mBio 2013, 4. 32. Tucker AT, Leonard SP, DuBois CD, Knauf GA, Cunningham AL,
Wilke CO, Trent MS, Davies BW: Discovery of next-generation
18. Ueda K, Kawai S, Ogawa H, Kiyama A, Kubota T, Kawanobe H, antimicrobials through bacterial self-screening of surface-
Beppu T: Wide distribution of interspecific stimulatory events displayed peptide libraries. Cell 2018, 172:618-628.e13.
on antibiotic production and sporulation among
Streptomyces species. J Antibiot (Tokyo) 2000, 53:979-982. 33. Sohrabi C, Foster A, Tavassoli A: Methods for generating and
screening libraries of genetically encoded cyclic peptides in
19. Ueda K, Beppu T: Antibiotics in microbial coculture. J Antibiot drug discovery. Nat Rev Chem 2020, 4:90-101.
(Tokyo) 2017, 70:361-365.
34. Deschuyteneer G, Garcia S, Michiels B, Baudoux B, Degand H,
20. Culp EJ, Waglechner N, Wang W, Fiebig-Comyn AA, Hsu Y-P, Morsomme P, Soumillion P: Intein-mediated cyclization of
Koteva K, Sychantha D, Coombes BK, Van Nieuwenhze MS, randomized peptides in the periplasm of Escherichia coli and
Brun YV et al.: Evolution-guided discovery of antibiotics that their extracellular secretion. ACS Chem Biol 2010, 5:691-700.
inhibit peptidoglycan remodelling. Nature 2020, 578:582-587.
35. Casadevall A, Dadachova E, Pirofski L: Passive antibody therapy
21. Li L, Maclntyre LW, Brady SF: Refactoring biosynthetic gene for infectious diseases. Nat Rev Microbiol 2004, 2:695-703.
clusters for heterologous production of microbial natural 36. Storek KM, Auerbach MR, Shi H, Garcia NK, Sun D, Nickerson NN,
products. Curr Opin Biotechnol 2021, 69:145-152. Vij R, Lin Z, Chiang N, Schneider K et al.: Monoclonal antibody
targeting the b-barrel assembly machine of Escherichia coli is
22. Bekiesch P, Basitta P, Apel AK: Challenges in the heterologous
bactericidal. Proc Natl Acad Sci U S A 2018, 115:3692-3697.
production of antibiotics in Streptomyces. Arch Pharm
(Weinheim) 2016, 349:594-601. 37. Storek KM, Chan J, Vij R, Chiang N, Lin Z, Bevers J, Koth CM,
Vernes J-M, Meng YG, Yin J et al.: Massive antibody discovery
23. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL: Drugs for bad used to probe structure-function relationships of the essential
bugs: confronting the challenges of antibacterial discovery. outer membrane protein LptD. eLife 2019, 8.
Nat Rev Drug Discov 2007, 6:29-40.
38. Ali SO, Yu XQ, Robbie GJ, Wu Y, Shoemaker K, Yu L,
24. Payne DJ, Miller LF, Findlay D, Anderson J, Marks L: Time for a DiGiandomenico A, Keller AE, Anude C, Hernandez-Illas M et al.:
change: addressing R&D and commercialization challenges Phase 1 study of MEDI3902, an investigational anti-
for antibacterials. Philos Trans R Soc Lond B Biol Sci 2015, Pseudomonas aeruginosa PcrV and Psl bispecific human
370:20140086. monoclonal antibody, in healthy adults. Clin Microbiol Infect
2019, 25:629.e1-629.e6.
25. Durrant JD, Amaro RE: Machine-learning techniques applied
to antibacterial drug discovery. Chem Biol Drug Des 2015, 39. DiGiandomenico A, Keller AE, Gao C, Rainey GJ, Warrener P,
85:14-21. Camara MM, Bonnell J, Fleming R, Bezabeh B, Dimasi N et al.: A
multifunctional bispecific antibody protects against
26. de la Fuente-Nunez C: Toward autonomous antibiotic Pseudomonas aeruginosa. Sci Transl Med 2014, 6:262ra155.
discovery. mSystems 2019, 4.
40. Lehar SM, Pillow T, Xu M, Staben L, Kajihara KK, Vandlen R,
27. Richter MF, Drown BS, Riley AP, Garcia A, Shirai T, Svec RL,  DePalatis L, Raab H, Hazenbos WL, Morisaki JH et al.: Novel
Hergenrother PJ: Predictive compound accumulation rules antibody-antibiotic conjugate eliminates intracellular S.
yield a broad-spectrum antibiotic. Nature 2017, 545:299-304. aureus. Nature 2015, 527:323-328
This manuscript describes a novel antibody-antibiotic approach to target
28. Stokes JM, Yang K, Swanson K, Jin W, Cubillos-Ruiz A, intracellular bacteria.
 Donghia NM, MacNair CR, French S, Carfrae LA, Bloom-
Ackermann Z et al.: A deep learning approach to antibiotic 41. Peck M, Rothenberg ME, Deng R, Lewin-Koh N, She G,
discovery. Cell 2020, 181:475-483 Kamath AV, Carrasco-Triguero M, Saad O, Castro A, Teufel L et al.:
This manuscript describes the use of deep learning to screen chemical A phase 1, randomized, single-ascending-dose study to
libraries and identifies compounds with antibacterial activity. investigate the safety, tolerability, and pharmacokinetics of
DSTA4637S, an anti-Staphylococcus aureus thiomab
29. Der Torossian Torres M, de la Fuente-Nunez C: Reprogramming antibody-antibiotic conjugate, in healthy volunteers.
biological peptides to combat infectious diseases. Chem Antimicrob Agents Chemother 2019, 63.
Commun (Camb) 2019, 55:15020-15032.
42. Wrammert J, Koutsonanos D, Li G-M, Edupuganti S, Sui J,
30. Mourtada R, Herce HD, Yin DJ, Moroco JA, Wales TE, Engen JR, Morrissey M, McCausland M, Skountzou I, Hornig M, Lipkin WI
Walensky LD: Design of stapled antimicrobial peptides that are et al.: Broadly cross-reactive antibodies dominate the human
stable, nontoxic and kill antibiotic-resistant bacteria in mice. B cell response against 2009 pandemic H1N1 influenza virus
Nat Biotechnol 2019, 37:1186-1197. infection. J Exp Med 2011, 208:181-193.

www.sciencedirect.com Current Opinion in Microbiology 2021, 63:66–69