Evolving a Generalist Biosensor for Bicyclic Monoterpenes

Simon d’Oelsnitz, Vylan Nguyen, Hal S. Alper, Andrew D. Ellington

TL;DR

Researchers have engineered a versatile biosensor capable of detecting a range of bicyclic monoterpenes, crucial compounds in the flavor, fragrance, and pharmaceutical industries. This was achieved by evolving a bacterial transcription factor using a novel high-throughput directed evolution method, paving the way for more efficient metabolic engineering and chemical analysis.

What Problem Does This Paper Address?

Monoterpenes are a large and commercially important class of molecules, but their detection, especially within living cells, is challenging. Traditional methods like gas chromatography-mass spectrometry are low-throughput and not suitable for real-time cellular monitoring. While genetic biosensors offer a promising alternative for high-throughput analysis and regulation, finding or engineering sensors for specific molecules, particularly those lacking a chromophore like monoterpenes, is difficult. Existing natural biosensors are limited in scope, and engineering new ones can be time-consuming and yield only modest improvements. This paper tackles the challenge of creating a generalist biosensor that can detect multiple bicyclic monoterpenes, overcoming the limitations of specificity and throughput in current analytical methods.

Methods

The core of this work lies in the development and application of a directed evolution technique called Seamless Enrichment of Ligand-Inducible Sensors (SELIS). SELIS combines a high-throughput positive screen (detecting ligand-dependent gene expression) with a negative selection (ensuring repression in the absence of the ligand), allowing for rapid identification of functional biosensors from large mutant libraries.

The authors started with the CamR repressor from Pseudomonas putida, a transcription factor known to respond to camphor. They engineered this system to function in Escherichia coli and then systematically optimized the reporter circuit components: the promoter, ribosome binding site (RBS), and the CamR operator sequence. This optimization significantly improved the signal-to-noise ratio of the biosensor.

Following circuit optimization, SELIS was employed to evolve CamR to recognize a panel of four non-cognate monoterpenes: borneol, fenchol, eucalyptol, and camphene. This involved generating mutant libraries of CamR through error-prone PCR and then selecting for variants that showed ligand-dependent activity.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • CamR as a versatile platform: The study demonstrates that CamR, originally a camphor sensor, can be effectively evolved into a generalist biosensor for a range of bicyclic monoterpenes. This highlights the potential of repurposing existing transcription factors for new applications.
  • SELIS efficiency: The SELIS method proved highly effective, enabling the rapid screening of large mutant libraries and the identification of functional biosensor variants in a matter of days.
  • Convergent evolution: Interestingly, different evolutionary trajectories for sensing various monoterpenes often converged on common mutations, suggesting that certain amino acid changes are particularly beneficial for broad monoterpene recognition.
  • Generalists emerge first: The evolution process surprisingly yielded generalist sensors rather than highly specific ones, even when attempts were made to "push" for specificity towards a single monoterpene. This suggests that a generalist intermediate might be a common step in the evolution of ligand-binding proteins.
  • Synthetic biology advantage: The optimized synthetic reporter circuit in E. coli outperformed the native system, underscoring the benefits of modular design and orthogonality in synthetic biology.

Key Results

  • The optimized CamR biosensor system achieved a 150-fold signal-to-noise ratio, a significant improvement over the original construct.
  • SELIS successfully evolved CamR variants that responded to four distinct monoterpenes: borneol, fenchol, eucalyptol, and camphene.
  • A majority of the recovered functional variants contained single amino acid substitutions, with Y58F and A139T being particularly prevalent.
  • These mutations, when analyzed individually, showed that many substitutions that increased response to one monoterpene also increased response to others, confirming the generalist nature of the evolved sensors.
  • Even attempts to specialize the sensor towards fenchol resulted in variants that remained broadly responsive to other monoterpenes.

Limitations and Future Work

  • Generalist vs. Specialist: The study notes the surprising lack of highly specific variants, even after targeted evolution. This suggests that either CamR has a limited evolutionary landscape for specialization, or the SELIS method's negative selection might be too stringent, favoring generalists. Future work could explore modifications to SELIS or alternative directed evolution strategies to isolate specialists.
  • Understanding Mechanisms: While key mutations were identified and mapped onto a homology model, the precise molecular mechanisms by which these mutations confer broad monoterpene recognition are not fully elucidated. Further structural and biochemical studies could provide deeper insights.
  • Broader Applications: The authors envision these generalist biosensors as a starting point for high-throughput screening and dynamic regulation of monoterpene production strains. Future work could involve integrating these sensors into microbial production systems and testing their performance in more complex biological contexts. The authors also suggest that the SELIS approach could be applied to evolve other transcription factors for different ligand classes.

Summary updated

Abstract

Prokaryotic transcription factors can be repurposed as analytical and synthetic tools for precise chemical measurement and regulation. Monoterpenes encompass a broad chemical family that are commercially valuable as flavors, cosmetics, and fragrances, but have proven difficult to measure, especially in cells. Herein, we develop genetically encoded, generalist monoterpene biosensors by using directed evolution to expand the effector specificity of the camphor-responsive TetR-family regulator CamR from <i>Pseudomonas putida</i>. Using a novel negative selection coupled with a high-throughput positive screen (Seamless Enrichment of Ligand-Inducible Sensors, SELIS), we evolve CamR biosensors that can recognize four distinct monoterpenes: borneol, fenchol, eucalyptol, and camphene. Different evolutionary trajectories surprisingly yielded common mutations, emphasizing the utility of CamR as a platform for creating generalist biosensors. Systematic promoter optimization driving the reporter increased the system's signal-to-noise ratio to 150-fold. These sensors can serve as a starting point for the high-throughput screening and dynamic regulation of bicyclic monoterpene production strains.
BibTeX
@inproceedings{doelsnitz2022evolving,
  author = {Simon d’Oelsnitz and Vylan Nguyen and Hal S. Alper and Andrew D. Ellington},
  title = {{Evolving a Generalist Biosensor for Bicyclic Monoterpenes}},
  booktitle = {ACS},
  year = {2022},
  doi = {10.1021/acssynbio.1c00402},
}
Figure