Over

view

We have developed a series of measurement methods targeting lipids for our project. We employed two different methods to measure intracellular lipids, and both methods can measure all types of intracellular lipids.

We found that gas chromatography-mass spectrometry (GC-MS) is a mature measurement technique. However, different samples require different detection methods. Upon reviewing various literature sources, we discovered inconsistencies in the detection methods for the same substance. Therefore, we conducted a detailed exploration of GC-MS measurement methods for MEL and decanoic acid, aiming to provide guidance for future researchers.

Measuring growth curves is a commonly used experimental approach that often involves establishing multiple experimental groups and continuously culturing and sampling over a period of time. The most common method is using shaker flasks, which may require a large quantity of drugs and equipment. On the other hand, when using a 96-well plate for cultivation, it gets contaminated upon detection, preventing further cultivation. We have developed a new method to achieve contamination-free detection using a 96-well plate without compromising our ability to analyze bacterial growth trends.

During the measurement process, we also discovered a correlation between fermentation substrates and products. Based on this finding, we developed an algorithm for inferring the products.

To ascertain whether the carbon flux in the XM01 strain has been successfully altered by knocking out the DGA1 and ARE1 genes, we need to evaluate the intracellular lipid content of the cells. It is essential to exclude the influence of lipids in the culture medium when measuring the intracellular lipid content. Therefore, we have employed two methods, one for quantitative analysis and the other for qualitative analysis, to detect the intracellular oil in the XM01 strain.

1.Staining Method for Qualitative Measurement of Intracellular Lipids

The cell culture of the XM01 strain ΔDGA1ΔARE1 double knockout strain was mixed with a solution of Nile Red (5μg/mL in dimethyl sulfoxide). The mixture was incubated for 10 minutes at room temperature under light-shielded conditions. Visualizations were performed using an Olympus U-LH100HG fluorescence microscope at a magnification of 100x, using both brightfield and fluorescence modes (excitation wavelength of 530 nm and emission wavelength of 626 nm). The images of the cells were captured and recorded using the CELL Sens software (Figure 1).



Figure 1 Intracellular lipid staining of original XM01 strain and ΔDGA1ΔARE1 strain

2.Weighing Method for Quantitative Measurement of Intracellular Lipids

The fermentation broth of the XM01 strain ΔDGA1ΔARE1 double knockout strain was fermented for different durations: 36 hours, 48 hours, and 168 hours. After fermentation, the fermentation broth was collected, and centrifuged at 8000rpm for 10 minutes to separate the cells. The cells were then dried and weighed. Following the weighing process, 5ml of concentrated hydrochloric acid and 5ml of water were added. The mixture was sealed and heated at 80℃ for 6 hours. Next, 7ml of 60% methanol and 7ml of dichloromethane were added, and the mixture was vigorously shaken to ensure thorough mixing. The mixture was then centrifuged at 8000rpm for 10 minutes, and the lower dichloromethane phase was separated. The separated phase was dried at 80℃ and weighed to obtain the weight of intracellular oil.

To ensure accurate measurement of the fermentation results for each module, we opted to utilize gas chromatography-mass spectrometry (GC-MS) for the detection of the expected products. In order to obtain the best measurement outcomes, we made several adjustments to the experimental procedure.

1.Measurement of MEL production

The fermentation broth to be tested was extracted using ethyl acetate, resulting in the ethyl acetate phase containing MEL. This phase was then evaporated to obtain purified MEL.

A portion of the MEL sample was taken and mixed with 1ml of chloroform. To this mixture, 2.5ml of methanolic sulfuric acid solution was added. The mixture was sealed and heated at 80°C for 2.5 hours. After cooling, 1ml of saturated sodium chloride solution and 1ml of hexane were added. The mixture was shaken for 30 seconds and allowed to settle, and the upper layer was collected. It was then filtered through a 0.22μm organic phase filter membrane and prepared for GC-MS analysis.

Separation of MEL was carried out using an HP-INNOWAX capillary column (inner diameter 30 m × 0.25 mm, film thickness 0.25 μm). Initially, the oven temperature was set to 100°C for 1 minute, then ramped up to 250°C at a rate of 250°C/min, and held at 250°C for 5 minutes. The injector and detector temperatures were set at 275°C and 300°C, respectively.

2.10-hydroxydecanoic acid

Initially, we attempted to detect decanoic acid using the same method employed for MEL detection. However, the results were extremely poor, and we struggled to obtain the necessary data. We attributed this to the extraction and detection parameters, as the MEL detection method might not be suitable for decanoic acid.

In the second attempt, we adopted a new sample processing method and adjusted the detection parameters.

First, 2ml of the supernatant was collected after centrifugation, and its pH was adjusted to 2.0 using 6 mol/L HCl. Chloroform was used for extraction, and the lower chloroform phase was collected. Then, 2ml of methylation reagent (sulfuric acid/methanol, volume ratio of 1:4) was added, and the mixture was heated at 60°C for 20 minutes. After cooling to room temperature, 2ml of hexane and saturated sodium chloride solution were added. The mixture was vortexed for 20 seconds, left to stand for 30 minutes, and then centrifuged at 3000 rpm for 10 minutes. The upper layer, consisting of hexane (1.5 ml), was collected, filtered through a 0.22μm organic phase filter membrane, and prepared for analysis using GC-MS.

An HP-5 capillary column (inner diameter 30 m × 0.32 mm, film thickness 0.25 μm) was used for separation. The oven temperature was initially set at 80°C for 2 minutes, followed by a temperature ramp of 5°C per minute up to 250°C, which was maintained for 10 minutes. The injector and detector temperatures were set at 250°C and 300°C, respectively.

These improvements yielded favorable results during the second analysis. However, unfortunately, we did not detect decanoic acid, and it was observed that decanoic acid, as a fermentation substrate, had hardly diminished. We speculated that the fermentation time was insufficient, and the high glucose content in the fermentation medium favored the consumption of glucose by the engineered bacteria rather than decanoic acid. Therefore, we decided to extend the fermentation time and reduce the glucose content in the fermentation medium for the next experiment.

Unfortunately, our third fermentation experiment also yielded unsatisfactory results. The good news was that decanoic acid was observed to have decreased, but we still did not find decanoic acid in its hydroxylated form (10-HDDA).

During the fermentation process, the quantity of bacteria plays a crucial role and has a significant impact on the yield. Therefore, we aimed to minimize the impact on bacterial growth by knocking out specific genes. Additionally, to prevent the escape of engineered bacteria, we designed a suicide switch. To evaluate the effectiveness of these two modules, we needed to measure the bacterial growth status. We selected the widely used indicator OD600, but tailored our experimental methods based on the specific characteristics of each experiment.

1.ΔDGA1ΔARE1 Growth Rate Assessment of Double Knockout Bacterial Strains

The wild-type strain and the ΔDGA1ΔARE1 double knockout strain were cultured in seed culture medium for 48 hours. They were then transferred to fermentation medium and cultured for 96 hours. Samples were taken every 3 hours, and the concentration of the fermentation broth was measured using a UV spectrophotometer at OD600. A growth curve was generated based on these measurements.

2.measurement of Kill switch

Due to the need to explore the effect of NeuAc concentration in the context of the kill switch, we planned to establish seven gradient concentration experiments. Shaking flask cultivation at this scale is labor-intensive and requires a substantial amount of NeuAc, which we did not have in abundance. We considered using a 96-well plate cultivation method as an excellent option, as it significantly reduces the amount of NeuAc required and eliminates the need for numerous conical flasks. However, sampling from a 96-well plate and measuring absorbance at OD600 using a UV spectrophotometer posed challenges.

Using a plate reader to measure the OD600 absorbance of each well in a 96-well plate requires opening the lid, potentially contaminating the bacterial culture and preventing continuous cultivation. In the end, we opted for 96-well plate cultivation, measurement using a plate reader, but without opening the lid during measurements, ensuring contamination-free assessment of the 96-well plate. We observed that the difference in OD600 absorbance between measurements with and without opening the lid was a constant value for the same sample. This indicates that lid-closed measurements do not affect our assessment of bacterial growth trends.

This improvement in the cultivation method significantly reduced our workload and the consumption of various reagents, allowing us to obtain a substantial amount of data at minimal cost (Figure 2).



Figure 2 A 96-well plate was used to continuously measure OD600 during culture

We conducted the fermentation to produce MEL (Mannosylerythritol lipids) in shake flasks using natural plant oils such as soybean oil, peanut oil, rapeseed oil, corn oil, and sunflower oil as carbon sources. The resulting fermentation broth was subjected to extraction using ethyl acetate to obtain the ethyl acetate phase, which was further subjected to liquid-liquid distillation to isolate MEL. Subsequently, we performed a washing step using a hexane-methanol-water mixture (1:6:3) to remove residual oil and fatty acids. The methanol phase was collected and subjected to rotary evaporation to obtain preliminarily purified MEL, and the yield was calculated.



Figure 3 Fatty acid composition of natural vegetable oil and MEL

We conducted methylation on the natural plant oil carbon sources and the corresponding synthesized MEL. Subsequently, we performed gas chromatography/mass spectrometry (GC/MS) analysis to determine the fatty acid composition of the carbon sources and the produced MEL.

We observed that the ratio of linoleic acid (9,12-octadecadienoic acid) in different carbon sources determined the proportion of decenoic acid (C10:1) in MEL, while the composition of oleic acid (9-octadecenoic acid) determined the composition of decanoic acid (C10:0). Additionally, the proportion of palmitic acid (hexadecanoic acid) in the carbon source significantly influenced the proportion of caprylic acid (C8:0) in the fatty acid composition of MEL.

Sunflower oil showed the highest proportion of linoleic acid at 59.59% among the natural plant oil carbon sources, resulting in the highest proportion of unsaturated C10 acid (51.81%) in the MEL produced through sunflower oil fermentation. Rapeseed oil exhibited the highest proportion of oleic acid, leading to a significant percentage of capric acid (C10:0) at 63.20% in the MEL fraction. Among all carbon sources, cottonseed oil demonstrated the highest proportion of palmitic acid, thereby yielding the highest proportion of caprylic acid (C8:0) in the MEL fraction.

Based on our findings, we believe that the proportions of oleic acid, linoleic acid, and palmitic acid in the carbon sources determine the proportions of C10 and C8 acids in MEL. Therefore, we aim to utilize modeling techniques to predict product composition based on the choice of carbon source.

In the hardware aspect, we designed multiple impeller blades. To select the most suitable impeller blade, we need to test the efficiency of each blade in mixing oil and water. We measure the time required for the oil and water to be thoroughly mixed using the impeller blade as an indicator of its mixing efficiency.