Downloads

High-throughput metabolite quantification for synthetic biology

Sensitive and robust results using the Echo^® MS System

Mackenzie Pearson¹, Rahul Baghla¹, Robert Proos¹, Neal Liddle¹, Baljit Uhbi¹, Chia-Wei Lu², Tony Wu², Mona Elbadawi², Michael Leavell², and Hanxiao Jiang^2

1SCIEX, USA; ²Amyris, USA

Abstract

Synthetic biology is a rapidly growing field driven by recent technology developments in gene editing and others. Large strain libraries can be developed that the must be screened for metabolomic changes. Current LC-MS/MS separation techniques are used but are often not rapid enough for efficient library screening. Here, a quantitative approach for the rapid screening of yeast strains using the Echo MS System was demonstrated, monitoring over 60 metabolites. The amount of analysis time saved was considerable, with the Echo MS system analysis time being 5.6x faster than the traditional LC-MS analysis, while still demonstrating sufficient reproducibility and sensitivity for strain profiling.

Introduction

Synthetic biology is a rapidly maturing interdisciplinary field that supports delivery of sustainably-produced products across many application areas, including human health (therapeutics, adjuvants), nutrition (vitamins, zero-calorie sweeteners, meatless meat), personal care (cosmetics, fragrances), agriculture (fertilizers), industrial chemicals (detergents, monomers), fuels and many others.^1,2 Products that are currently inaccessible due to cost, quality or supply constraints are now accessible at higher quality, at lower cost, and from an unconstrained supply.³

These developments are in large part due to technological improvements in the rapid genetic engineering of multiple types of host organisms, high-throughput culture, and analytical screening.^4,5,6 Currently, chromatography-based separation techniques (such as LC-MS/MS) are the most utilized approaches for strain analysis. However, these technologies are often not rapid enough to analyze large biological libraries, which can contain 10⁵ entities or more, on a practical time scale.⁷ With this advancing field comes the demand for an advanced technology to rapidly profile and quantify several analytes across a very large number of samples in a high-throughput approach.⁸ The Echo MS System combines Acoustic Ejection Mass Spectrometry (AEMS) with the SCIEX Triple Quad™6500+ system to provide an integrated solution with the capability of quantifying approximately 1 sample per second.⁹The system delivers reproducible nanoliter-sized droplets from a complex matrix, in a 384- or 1536-well plate format, into a solvent stream for direct analysis by MS. This system provides the speed and accuracy to help circumvent the chromatographic bottleneck of high-throughput analysis.

Here, a quantitative approach for the rapid screening of 90 yeast strains using the Echo MS System was demonstrated, monitoring over 60 metabolites. The goal of publishing this workflow is to demonstrate the ability of the Echo MS System to rapidly screen a biological matrix for key metabolites of interest with significantly reduced analysis times (Figure 1).

Key features of the Echo MS System for high throughput metabolomic quantification

Rapid analysis of 67 endogenous metabolites in yeast extract in a 384-well plate in under 6 hours
One sample analyzed every 3 seconds
Low acoustic ejection sample volumes used - one 50 nL ejection per well for each method
Good linearity demonstrated with excellent accuracy and reproducibility
Minimal carryover due to AEMS and open port interface (OPI)

Methods

Sample preparation: Saccharomyces cerevisiae was grown in 4% sucrose media for 30 hours, quenched in cold methanol to halt metabolism, then spun down to separate extra- and intra-cellular matrices. Intracellular metabolites were extracted with 75:25 organic solvent/buffer solution. Both matrices were further diluted with water for a final organic solvent content of 50%, then filtered via a 3kD molecular cutoff membrane. Samples were stored at -70^oC until analysis.

Authentic standards were prepared separately in 50:50 methanol/water, then pooled per acquisition method and diluted in 50:50 methanol/water to 3 µg/mL, the top calibration concentration. Standard solutions were then serially diluted in the same diluent across a concentration range of 0.003 – 3 µg/mL. A total of 15 methods were used for the targeted panel.

Acoustic ejection method: Methanol was used as the carrier solvent at a flow rate of 425 µL/min. A volume of 50 nL (20 droplets) was acoustically ejected into the system.

Mass spectrometry: The Echo MS System coupled with the SCIEX Triple Quad 6500+ system. SCIEX OS Software 1.6.10 was used to control the system.

Data processing: Peak areas and calculated concentrations were generated using SCIEX OS Software.

Accuracy and reproducibility of quantification

The MRM transitions for the selected metabolites of this study were optimized and then monitored in the yeast samples using the Echo MS System. Amino acids, nucleosides, organic acids, and other classes were included in the panel. A total of 15 methods were created in both positive and negative polarity to cover the 62 MRM transitions that monitored 67 metabolites. Isobars (leucine and isoleucine, for example) were monitored by a single MRM transition. There is no LC separation on the Echo MS System. Because the 62 MRM transitions account for a total of 67 metabolites, 5 transitions were reported as a sum for the isobar pair.

The total run time required to analyze the 67 metabolites for each 384-well plate was just over 5.7 hours (including acquisition overhead) when using the Echo MS System. In contrast, the same analysis required 32 hours when using a 5 minute UHPLC-MS/MS screening method. Thus, the Echo MS approach significantly reduced the amount of time needed for this metabolite screen, ~5.6x faster than a UHPLC-MS approach (Figure 1).

Good sensitivity of detection of the standards was observed, with 35 out of the 67 metabolites detected at 0.003 µg/mL, the lowest concentration tested (Figure 2). An example of the quantitative accuracy is shown in Table 1 for ornithine across the 3 orders of linear dynamic range that was tested. All %CV values reported are less than ten percent for 3 replicate ejections from a well. Example extracted ion chromatograms (XICs) for ornithine are shown in Figure 3. Accuracy for the assay was measured at each concentration for each analyte, and the acceptance criteria was set to +/- 20 percent.

To illustrate reproducibility for other compounds, an example of sequential triplicate ejections at 0.003 µg/mL is shown in Figure 4. Valine, leucine/isoleucine, cysteine, and homoserine/threonine represent four MRM transitions that were monitored in a single method.

Yeast strain analysis

Separate 384-well plates of intracellular and extracellular yeast matrices were screened for this study. Plates were provided in eight of the 96-well format and transferred into two of the 384-well format. 90 different yeast strains were each analyzed for all 67 metabolites. Figure 5 illustrates a heat map of metabolite peak areas from both matrices for valine, homoserine/threonine, cysteine, and leucine/isoleucine shown in a 96-well plate format for reporting. At 3 seconds per sample, the dataset for these chosen analytes for one 384-well plate took just 19.2 minutes. Concentrations for these metabolites are reported in µg/mL. Blank samples are shown in wells H12.

Conclusions

The utility of the Echo MS System to rapidly profile and quantify synthetic biology samples has been demonstrated here. The intracellular and extracellular matrices for 90 different yeast strains were analyzed and a total of 67 metabolites were monitored in about 5.7 hours. This high-throughput, robust workflow consumed a total of 750 nL of sample volume per well. The amount of analysis time saved was considerable. The Echo MS System analysis was completed 5.6x faster than the traditional LC-MS analysis. The reproducibility and sensitivity demonstrated are sufficient for strain profiling, making the Echo MS System workflow a powerful solution for synthetic biology.

References

Voigt, C.A. (2020) Synthetic biology 2020–2030: six commercially-available products that are changing our world. Nature Commun 11, 6379.
Tateno, M., Stone, B.J., Srodulski, S.J. et al. Synthetic Biology-derived triterpenes as efficacious immunomodulating adjuvants. Sci Rep 10, 17090 (2020).
Payen C, Thompson D. (2019) The renaissance of yeasts as microbial factories in the modern age of biomanufacturing. Yeast. Dec;36(12): 685-700.
Meng, F., and Ellis, T. (2020) The second decade of synthetic biology: 2010-2020. Nature Commun., 11(1), 5174.
Shaw, J. (2020) Engineering Life, Synthetic Biology and the frontiers of technology. Harvard Magazine 120-37, 37-41(74-75).
Hill, C. B., Czauderna, T., et al. (2015) Metabolomics, Standards, and Metabolic Modeling for Synthetic Biology in Plants. Frontiers in bioengineering and biotechnology, 3, 167.
Leavell, M.D., Singh, A.H., & Kaufmann-Malaga, B.B. (2020) High-throughput screening for improved microbial cell factories, perspective and promise. Current Opinion in Biotechnology. 62 (22-28).
Rienzo M, Jackson S.J., Chao L.K., Leaf T, Schmidt T..J, Navidi A.H., Nadler D.C., Ohler M, Leavell M.D. (2021) High-throughput screening for high-efficiency small-molecule biosynthesis. Metab Eng. 63:102-125.
Rapid MS/MS analysis with Acoustic Ejection Mass Spectrometry (AEMS). SCIEX technical note, RUO-MKT-02-11385-A.