Our research sits at the intersection of microbial biotechnology, quantitative synthetic biology and fundamental molecular bioscience. We believe that synthetic biology is enhanced by perspectives from a broad range of disciplines, and that tools required for cellular engineering also have the power to illuminate biology.



Research Themes


Quantitative Biology of Fluorescence

As illustrated by the Design-Build-Test-Learn framework of classical Engineering Biology, the effective generation of engineered cells requires iterative cycles of variant generation and testing. Moreover, most projects require platforms capable of screening the functions of large libraries of genetic variants in high throughput. Fluorescence is one of the most powerful methods for functional library screening, as its detection does not require cell lysis or processing, it can be detected effectively across many orders of magnitude, and enables detection spatially, kinetically, and in bulk or single cell formats. In addition, a wide range of fluorescent molecules, from small molecule fluorophores to fluorescent proteins and fluorogenic RNA molecules are now available for tracking a wide variety of cellular activities. However, classical fluorescence-based screening technologies typically neglect to ascertain key quantitative features of their platforms, which limits their usefulness.

Absolute protein quantification with FPCountR

One of the key features missing from typical fluorescence assays is calibration. Typically carried out by measuring the fluorescence of a dilution series of well-known small molecule fluorophores (such as fluorescein or sulforhodamine), calibration allows the relationship between the arbitrary ‘relative fluorescence units (RFU)’ exported by an instrument to be converted into a standardised units, that allow experiments to be comparable across instruments. This approach has been suggested to assist with the interpretation of fluorescent protein (FP) based assays, on the basis that fluorescent protein levels produced by engineered circuits (in promoter or RBS optimisation libraries) can be expressed in terms of standardised units of fluorescein or sulforhodamine.

However, three aspects of fluorescent protein quantification remain elusive:

  • Units of fluorescein or sulforhodamine are not biologically meaningful; we need units of protein.
  • As such, typical experiments cannot even reveal the order of magnitude at which FPs are being expressed (i.e. we cannot tell whether there are 10 or 100 or even 100,000 proteins per cell).
  • While the green fluorescent fluorescein may allow the comparison of GFP levels between instruments, it cannot address the comparison between two different fluorescent proteins in the same cell.

Using FPs directly as calibrants would allow us to quantify fluorescence in absolute units of protein copies per cell, but it is rarely attempted as it requires FP purification, which is seen as laborious and expensive. We developed a simple laboratory method for such a calibration without the use of purified protein standards. The resultant method requires only the use of cell lysates from E. coli that have overexpressed the FP in question. Protein quantification is carried out using an absorbance assay that allows high throughput quantification (compatible with 96-well plates) at very low cost (UVclear plates, £2/plate, rather than quartz plates, £2000/plate) by relying on a validated analytical conversion that carefully normalises and subtracts background signals to achieve accuracies within -/+10%. The analytical method has been described in full and can be applied by the use of an open source R package ( fpcountr ).

Numerous other aspects of the analytical workflow, including the conversion to protein molar concentrations rather than protein copies per cell, and quantification of the extent to which cell density obscures fluorescence were also considered and included, where no similar tool had ever done so previously. As an analytical biochemistry protocol co-developed with an accessible software package, this work illustrates the power of combining approaches from disparate fields for microbial synthetic biology.

Read more about this work here and here .



Translation Engineering for Microbial Biotechnology

A promising route to tackle the trade-off in cellular resources between synthetic protein production and cellular growth is to use a separate dedicated pool of orthogonal ribosomes to produce synthetic proteins. However, the optimisation of strains containing two ribosomal pools – native for the host cell’s proteome and orthogonal for synthetic proteins – has yet to be thoroughly explored. We addressed this by creating orthogonal ribosomes that fluoresce by inserting fluorescent RNA aptamers into tethered orthogonal ribosomal RNA (TO-rRNA).

To study the tolerance of the engineered ribosomes to aptamer insertion, we have assembled and screened a library of candidate insertion sites, identifying several sites in both the 16S and 23S TO-rRNA that enables ribosome labelling with minimal effect on translation activity. Serendipitously, we also identifed sites in the rRNA where insertion appears to not only be tolerated but to enhance orthogonal ribosome activity, across multiple bacterial strains and RNA insertions. Using bulk and single cell assays, we have demonstrated that these variants allow us to label orthogonal ribosomes for dynamic tracking and across populations, making it a promising tool for optimising orthogonal translation in engineered cells.

Ribosome engineering offers great potential, both for the development of next-generation microbial cell factories, as well as a tool to expand our understanding of ribosome function in living cells.

Read more about our recent paper here.



Data, Software and Reproducibility


In combination with our experimental work, we develop software tools to automate basic analytical workflows and assist newcomers to programming languages utilise existing software tools with the least effort and difficulty.

Data extraction with Parsley

Multiwell plate readers are an important tool in the life sciences. They are frequently used to measure fluorescence, absorbance and luminescence, among other measurement types. As they are heavily used for high-throughput assays and screens, their analysis benefits from automated (programmatic) data analysis.

However, most plate readers export raw data in formats that software packages cannot work with, and that do not contain the necessary metadata for downstream analysis. A necessary initial step in every analysis is therefore extracting the data, reformatting it into the correct ‘tidy’ data structure, and joining it with any required metadata: we call this process ‘parsing’.

While a few parser functions for certain export formats from certain plate readers have been written, there is no generic tool that can handle data parsing from any plate reader: this is why Parsley was built.

Read more about Parsley here and here.