Dry Lab

---

    Transient (co)expression of isoflavone biosynthetic genes in Nicotiana benthamiana

    Puerarin, daidzein, and genistein are isoflavonoids, a class of natural products that act as phytoestrogens in mammals and that are synthesised in an isoflavone pathway. These compounds are commonly found in household plants like soybeans, whereas puerarin is only found in the Pueraria genus. Nicotiana benthamiana has been used as an effective and easy plant for the functional characterisation of various plant proteins and enzymes (Sainsbury & Lomonossoff, 2014). Specifically, N. benthamiana is “amenable to protein overexpression through well-established Agrobacterium infiltration methods for DNA delivery... [due to] its native capacity to secrete heme peroxidases” (Goodin et al. 2008). However, because it lacks isoflavone synthase (IFS), N. benthamiana is unable to produce isoflavones. Thus, we performed transient co-expression of N. benthamiana leaves with a set of six isoflavone biosynthetic genes from P. mirifica: chalcone synthase (PmCHS), chalcone reductase (PmCHR), chalcone isomerase (PmCHI), isoflavone synthase (PmIFS), 2-hydroxyisoflavanone (PmHID), and C-glycosyltransferase (PmC-UGT) genes. This set of genes would allow N. benthamiana to produce puerarin, daidzein, and genistin. To augment isoflavone production, this collection of isoflavone biosynthetic genes was co-expressed with the Arabidopsis R2R3 MYB12 transcription factor (AtMYB12), thus amplifying the expression of flavonoid biosynthetic genes. Along with the six isoflavone biosynthetic genes from P. mirifica and AtMYB12, we also used Aequorea victoria green fluorescent protein (GFP) as a control group in our experiment.

Fig. 1 (the right hand image is a zoomed in version to the left pathway of the left image): Isoflavonoid and proposed miroestrol biosynthetic pathways in P. mirifica. Modified from Suntichaikamolkul et al. (2022, p. 63). Enzyme abbreviations: PAL, phenylalanine ammonia-lyase; C4H, cinnamate 4-hydroxylase; 4CL, coumarate-CoA ligase; CHS, chalcone synthase; CHI, chalcone isomerase; IFS, isoflavone synthase; HID, 2- hydroxyisoflavanone dehydratase; C-UGT, C-glycosyltransferase; O- UGT, O-glycosyltransferase; CYP, cytochrome P450; IFR, isoflavone reductase; PT, prenyltransferase

    In Fig. 1, daidzein and puerarin follow the same ‘left’ pathway. Genistein, on the other hand, follows the ‘right’ pathway. However, both pathways utilised the same sequence of enzymes to catalyse each substrate. Our experiment will not only be able to showcase N. benthamiana’s ability to produce isoflavones, but also to identify its preferred metabolic pathway: puerarin or genistin.

    Agrobacterium Vector-Mediated Infiltration and Vector Construction

    To allow for transient (co)expression of isoflavone biosynthetic genes in N. benthamiana, we utilised Agrobacterium vector-mediated infiltration (commonly referred to as agro-infiltration) as a method to perform transient transformation. This method was chosen as it leverages the natural ability of Agrobacterium tumefaciens to transfer DNA (T-DNA) into plant cells, facilitating rapid analysis of gene function and protein interactions without the need for stable transformation that allows for genetic modification to be present in future generations, but risks cross contamination with wildtypes and takes a longer experimental period.

    Vector Design

    We first needed to design the plasmids required for transient expression of the isoflavonoid biosynthesis pathway. Since we knew we were going to achieve our results from N. benthamiana through agroinfiltration, the team decided that it would be best to use an existing plasmid backbone as a template and build on it.

    We selected the pEAQ-HT-DEST1 vector because it was specifically designed for transient expression in plants as opposed to other vectors such as pCAMBIA which are designed for stable transformation.

There were also other features in the pEAQ DEST1 plasmid that appealed to us:

1. It contains a strong CaMV 35S promoter for high-level expression of our compound.
2. It is equipped with elements such as the LB and RB T-DNA repeats, which are necessary for T-DNA transfer during Agrobacterium-mediated plant transformation.
3. It suppresses RNA silencing of the transfer complex via the P19 suppressor gene⁴.
4. It has kanamycin resistance (KanR) for selection of transformed bacteria and plants.

    Then, we cloned six pEAQ-HT-DEST1 vectors using In-Fusion Cloning to contain the six aforementioned transient expression factors obtained from P. mirifica. We chose to do this in six different vectors rather than one because it would be easier for us to identify problems, if there are any later in the engineering cycle, if we had six vectors to alternate infiltration of rather than one giant vector.

    We also cloned 2 additional pEAQ-HT-DEST1 vectors to with GFP to use as a control for determining if agroinfiltration was successful and AtMYB12 to bolster production of substrates required for the isoflavonoid biosynthetic pathway so our modifications wouldn’t have any effects on other processes in the plant and that our target compounds would have increased production.

    Fig. 2: Vector diagram for each of the eight transformed Agrobacterium; AtMYB12, PmCHI, PmIFS, PmCHS, PmCHR, PmCGT, PmHID, GFP

 Kill Switch

---

    Our team also wanted to create safety measures to prevent the spread of our modified plant outside of controlled environments. After some consideration we chose to create a kill switch that would activate depending on specific conditions.

    We first wanted to choose a condition based on which our kill switch would activate. We originally thought about making the plant dependent on specific nutrients, which would only be available in controlled environments. The kill switch would activate and kill the plant without these nutrients. Further consideration showed that this would not be cost-effective for producing our supplements in the long run. Thus we decided to go with the atPfr-promoter, which is light-inducible and responds to red wavelengths of light, so our plant is dependent on a red light filtered growing environment. Furthermore, using our method to produce the isoflavonoids would only require a one-time investment in a red light filter, and if the plant went outside, the kill switch would activate almost immediately with red light being everywhere.

    Then we had to choose a terminator sequence. We chose Barnase, which is a ribonuclease that causes degradation of single stranded RNAs. We chose this for our killing mechanism because the expression of this protein without its inhibitor Barstar would result in cell death and cause the target plant to die.

    We then selected the pEAQ-HT-DEST1 vector as a base for the kill switch because its parameters fit the most with what we needed. The vector diagram in Figure 3 shows the structure of the kill switch.

Fig.3: Vector diagram for the kill switch

 Wet Lab Design

---

    Building and Testing Round 1

1.     1. Cloning of our synthesised plasmid

       Our project's wet lab began with cloning our bacteria via In-Fusion Cloning method. Unlike the traditional cloning method of using ligation enzymes that depends on specific restriction sites, limiting the choice of vectors and inserts, In-Fusion Cloning allows for a directional cloning of multiple fragments into any vector. First, we used restriction enzymes to linearise pHEAQ-HT-DEST1. Then, we combined the linearised vector with the PCR-amplified insert (referring to the T-DNA section on each of the eight diagrams as shown in Fig. 2)⁵ and the In-Fusion Cloning Master mixed to perform the cloning reaction (as per the kit’s instructions). From then on, the resulting solution (from the combination phase of In-Fusion Cloning) was transformed into E. coli strain DH5-α via the heat shock transformation method.

       Out of eight transient expression factors, six from P. mirifica-PmCHR, PmCHI, PmIFS, PmHID, PmCHS, and PmCGT—were cloned by us. However, AtMYB12 and GFP were already available in the lab and were thus not cloned.

       ⁵Before transformation, we performed PCR and RNA gel electrophoresis to confirm that the length of our recombinant vector is relatively the same as the theoretical length.

2.     2. Growing bacteria

       We grew the transformed E. coli strain DH5-α on a petri dish filled with Luria-Bertani (LB) agar, and kanamycin.⁶ The base LB agar includes four grams of peptone, two grams of yeast extract, four grams of table salt (NaCl), and three grams of agar. E. coli colonies verified by colony PCR were then sent for DNA sequencing. The plasmids identified as positive clones were then transformed into A. tumefaciens, which was left to grow in a new petri dish with LB agar (same base formulae) supplemented with not only kanamycin but also rifamycin and gentamicin.

       ⁶KanR is added as a safety step. Regular E. coli are susceptible to kanamycin. However, the insert that we cloned is specifically designed to include the kanamycin-resistant gene. Therefore, the bacteria's survival in this medium confirms the successful transformation of our recombinant plasmid into E. coli.

3. Infiltration of N. benthamiana

After two days, we again verified the Agrobacterium using colony PCR and agarose gel electrophoresis. We then selected positive clones for overnight inoculation. Subsequently, we washed the overnight culture three times with sterilised distilled water. The pallets were then resuspended in a resuspension solution of 10 mM MgCl₂, 10 mM MES-K, and 100 μM acetosy-ringone, pH 5.6, before being mixed into 3 different solutions:

1. Control: Agrobacterium with >pEAQ-GFP vector.
2. Agrobacterium harbouring the AtMYB12, along with an upstream array of isoflavone biosynthetic genes, including PmCHR, PmCHI, PmIFS, PmHID, and PmCHS which culminate the pathway at daidzein or genistein.
3. Agrobacterium harbouring the AtMYB12, along with an upstream array of isoflavone biosynthetic genes including PmCHR, PmCHI, PmIFS, PmHID, PmCHS, and PmCGT which culminate the pathway for puerarin or daidzein.

We mixed Agrobacterium in equal amounts for solutions 2 and 3 (adjusting the OD to 0.5).

After each of the solutions has rested at room temperature for two-three hours, we infiltrate each solution into the top four leaves of roughly 4-week-old N. benthamiana trees. We then leave these trees to grow for an additional five days.

4. Extraction and analysis

Before extracting the isoflavones, we freeze-dried the collected leaves of the modified N. benthamiana. Then we perform extraction using our sample and 500 ul of 50% methanol and 20 ug/ml of isovitexin as an internal standard. After that, using only the extraction's supernatant, we performed HPLC analysis. We used a reversed-phase C18 column (50 x 4.6 mm) filled with 5 M core-shell silica. The linear gradient of the mobile phase was set from 95:5 to 85:15 of 1% acetic acid (v/v) in water and 1% acetic acid (v/v) in acetonitrile, with a flow rate of 1 ml/min for 45 min. The column temperature during the test was set at 45 degrees Celsius. Using UV absorbance (254 nm and 280 nm), we monitored the eluate. Calibration curves of chromatogram peak areas for the standard compounds were produced.

Building and Testing Round 2

After our first round of engineering, the peak of expected isoflavones did not obviously show based on HPLC results. Thus, we repeated the aforementioned process for Step 2: Growing Bacteria and Step 3: Infiltration again, though with some variation in our methodology for extraction and analysis. These variations include the following:

1. Shifts from using dry weight samples to fresh weight samples in step 4 - extraction and analysis.

   1. The fresh weight sample was done by grinding the sample into fine powder directly after collection instead of freeze-drying the sample first before grinding.
   2. Justification: Following the utilisation of dry weight as the sample, HPLC analysis revealed that the calibration curves of the chromatogram exhibit peaks corresponding to the standard compounds daidzein, genistin, daidzein, and genistein. However, the aforementioned peaks are insufficient. Thus, we switched from dry weight to fresh weight samples.

2. Change in HPLC gradient

   1. The mobile phase was set to 100% solvent A and 0% solvent B (at 0 mins) to 55% solvent A and 45% solvent B (at 45 mins). Solvent A, 0.1% of acetic acid in distilled water, and solent B, 0.1% acetic acid in acetonitrile with a flow rate of 1 ml/min while maintaining the column temperature at 45°C and the sample injection at 20µl. However, the parameters used in the first round of HPLC were similar, but there was a change in the gradient: 95% solvent A and 5% solvent B (at 5 min) and 15% solvent B and 85% solvent A (at 40 min). In addition, the sample injected was increased from 10µl to 20µl.
   2. Justification: Altering the gradient rate enables the attenuation of noise surrounding the peak of each compound while concurrently enhancing the prominence of other peaks.

Fig. 4: Graph showing the flow rate of each solvent (solvent A and solvent B) for the first HPLC analysis; 95% solvent A and 5% solvent B (at 5 min) and 15% solvent B and 85% solvent A (at 40 min).

Fig. 5: Graph showing the flow rate of each solvent (solvent A and solvent B) for the second analysis; 100% solvent A and 0% solvent B (at 0 mins) and 55% solvent A and 45% solvent B (at 45 mins).