Production of Genistein in Amaranthus tricolor var. tristis and Spinacia oleracea by Expression of Glycine max Isoflavone Synthase

Isoflavonoids, the diverse group of secondary metabolites derived from the phenylpropanoid pathway, are distributed predominantly in leguminous plants. It has received considerable attention in recent days due to its health promoting benefits and is known to prevent certain diseases in humans. These isoflavonoids are synthesized from flavonoid intermediates of phenylpropanoid pathway by the enzyme isoflavone synthase. Metabolic engineering of isoflavonoid biosynthesis in non-legume crop plants could offer the health benefits of these compounds in diverse plant species further contributing for crop improvement. The transient expression of heterologous genes in the host is considered as an alternative to stable expression, that can provide a rapid way of studying the pathway engineering for metabolite production and could also act as a production platform for nutraceuticals and biopharmaceuticals. In this study, isoflavone genistein was produced in Amaranthus tricolor var. tristis and Spinacia oleracea by transiently expressing Glycine max isoflavone synthase (GmIFS). The GmIFS gene was cloned in plant expression vector pEarleyGate 102 HA and pEAQ-HT-DEST 3 and transformed into plants by agroinfiltration. The presence of transgene in the agroinfiltrated leaves was confirmed by semiquantitative reverse-transcription polymerase chain reaction. The flavonoid substrate naringenin and isoflavonoid genistein were quantified using high performance liquid chromatography in both wild-type and infiltrated leaf samples of both the plants. The naringenin content varied in the range of 65.5–338.5 nM/g fresh weight, while the accumulation of genistein was observed with varying concentrations from 113 to 182.6 nM/g fresh weight in the agroinfiltrated leaf samples of both A. tricolor var. tristis and S. oleracea. These results indicate that the transient expression of GmIFS gene has led to the synthesis of isoflavonoid genistein in A. tricolor var. tristis and S. oleracea providing an insight that stable expression of this gene could enrich the nutraceutical content in the crop plants. To the best of our knowledge, this is the first report on transient expression of GmIFS gene for the production of genistein in A. tricolor var. tristis and S. oleracea.


Introduction
Isoflavonoids, a class of natural plant polyphenols attribute to many biological activities such as antioxidant [1], antimutagenic [2], anticarcinogenic [3][4][5], antiproliferative [6,7] effects are playing an imperative role in human health and nutrition [8][9][10]. These compounds are naturally present in Papilionoideae, a subfamily of Leguminosae or Fabaceae members [11]. Apart from the other vital bioactivities, it is evident that the isoflavonoid genistein in dietary consumption showed a wide spectrum of health protective abilities such

Results
The seeds of A. tricolor var. tristis and S. oleracea procured from Tamil Nadu Agricultural University seed center were sown in commercially available soil manure mixture for germination and kept in greenhouse. The germinated 15-day-old plantlets were transferred individually to the pots and used for further studies. The schematic representation and timeline for the production of secondary metabolites in plants by transient gene expression are depicted in Figure 2.

Results
The seeds of A. tricolor var. tristis and S. oleracea procured from Tamil Nadu Agricultural University seed center were sown in commercially available soil manure mixture for germination and kept in greenhouse. The germinated 15-day-old plantlets were transferred individually to the pots and used for further studies. The schematic representation and timeline for the production of secondary metabolites in plants by transient gene expression are depicted in Figure 2.

Confirmation of Cloned IFS in Plant Expression Vector
The IFS gene procured from Donald Danforth Plant Science Center was used as template to clone the gene in plant expression vector pEarleyGate 102 HA ( Figure 3A) and pEAQ-HT-DEST 3 ( Figure 3B) by gateway cloning. Both the vectors were used for the transient expression of IFS gene. The IFS gene was cloned into entry vector pDONR/Zeo and further moved into the destination vector pEarleyGate 102 HA and pEAQ-HT-DEST 3. The generated clones were confirmed by polymerase chain reaction (PCR) and restriction digestion.

Confirmation of Cloned IFS in Plant Expression Vector
The IFS gene procured from Donald Danforth Plant Science Center was used as template to clone the gene in plant expression vector pEarleyGate 102 HA ( Figure 3A) and pEAQ-HT-DEST 3 ( Figure 3B) by gateway cloning. Both the vectors were used for the transient expression of IFS gene. The IFS gene was cloned into entry vector pDONR/Zeo and further moved into the destination vector pEarleyGate 102 HA and pEAQ-HT-DEST 3. The generated clones were confirmed by polymerase chain reaction (PCR) and restriction digestion.

Transformation of Agrobacterium tumefaciens with GmIFS and Plant Transformation
Plant expression vector pEarleyGate 102 HA and pEAQ-HT-DEST 3 harboring IFS gene was mobilized into A. tumefaciens AGL1 by electroporation. Electroporation uses short high-voltage pulses which help to beat the capacitance of the cell membrane inducing reversible breakdown of the membrane. This transient, permeabilized state of the membrane allows the passage way of the DNA into the cell [48]. After electroporation, the transformed A. tumefaciens colonies were selected on yeast extract and nutrient broth agar plates supplemented with 75 mg L −1 kanamycin and 60 mg L −1 rifampicin, and the colonies were confirmed by PCR using gene specific primers. Positive clone was further subcultured and used for the agroinfiltration.
We set out to investigate the effect of IFS expression in A. tricolor var. tristis and S. oleracea by transient expression using agroinfiltration. Leaves of four-week-old A. tricolor var. tristis and S. oleracea ( Figure 4A,B) were infiltrated with A. tumefaciens suspensions containing either one of the vectors pEarleyGate 102 HA or pEAQ-HT-DEST 3 harboring IFS gene ( Figure 5). The infiltrated leaves were harvested at day 4 after infiltration. Total RNA was isolated, and cDNA was synthesized, which was used as the template for the semiquantitative reverse-transcriptase PCR. The expected band size of 1500 bp confirmed the expression of IFS gene in the infiltrated leaves of A. tricolor var. tristis and S. oleracea ( Figure 6A). The results showed that transcription of the IFS gene was similar in the leaves infiltrated either with pEAQ-HT-DEST 3:IFS or pEarleyGate 102 HA:IFS. β-Actin (~114 bp) was used as an internal control ( Figure 6B).

Transformation of Agrobacterium tumefaciens with GmIFS and Plant Transformation
Plant expression vector pEarleyGate 102 HA and pEAQ-HT-DEST 3 harboring IFS gene was mobilized into A. tumefaciens AGL1 by electroporation. Electroporation uses short high-voltage pulses which help to beat the capacitance of the cell membrane inducing reversible breakdown of the membrane. This transient, permeabilized state of the membrane allows the passage way of the DNA into the cell [48]. After electroporation, the transformed A. tumefaciens colonies were selected on yeast extract and nutrient broth agar plates supplemented with 75 mg L −1 kanamycin and 60 mg L −1 rifampicin, and the colonies were confirmed by PCR using gene specific primers. Positive clone was further subcultured and used for the agroinfiltration.
We set out to investigate the effect of IFS expression in A. tricolor var. tristis and S. oleracea by transient expression using agroinfiltration. Leaves of four-week-old A. tricolor var. tristis and S. oleracea ( Figure 4A,B) were infiltrated with A. tumefaciens suspensions containing either one of the vectors pEarleyGate 102 HA or pEAQ-HT-DEST 3 harboring IFS gene ( Figure 5). The infiltrated leaves were harvested at day 4 after infiltration. Total RNA was isolated, and cDNA was synthesized, which was used as the template for the semiquantitative reverse-transcriptase PCR. The expected band size of 1500 bp confirmed the expression of IFS gene in the infiltrated leaves of A. tricolor var. tristis and S. oleracea ( Figure 6A). The results showed that transcription of the IFS gene was similar in the leaves infiltrated either with pEAQ-HT-DEST 3:IFS or pEarleyGate 102 HA:IFS. β-Actin (~114 bp) was used as an internal control ( Figure 6B).

HPLC Detection of Naringenin and Genistein in Infiltrated Samples
HPLC was performed to determine the presence of naringenin in wild-type A. tricolor var. tristis and S. oleracea followed by genistein detection in the infiltrated leaves. The results showed that the naringenin peaks were observed in both the plant species corresponding to that of naringenin standard. Wild-type A. tricolor var. tristis and S. oleracea showed 265.5 and 338.5 nM/g FW (nanomolar/gram fresh weight) of naringenin in the plant extracts. In the control samples, S. oleracea showed higher levels of naringenin content when compared to A. tricolor var. tristis. (Figure 7). Consequently, the naringenin content was also determined in the agroinfiltrated A. tricolor var. tristis and S. oleracea with pEarleyGate 102 HA and pEAQ-HT-DEST 3 vectors harboring IFS. As shown in Figure 7, the leaves of A. tricolor var. tristis and S. oleracea infiltrated with pEarleyGate 102 HA and pEAQ-HT-DEST 3 showed considerably less concentration of naringenin varying in the range of 65.5-134 nM/g FW compared to control.

HPLC Detection of Naringenin and Genistein in Infiltrated Samples
HPLC was performed to determine the presence of naringenin in wild-type A. tricolor var. tristis and S. oleracea followed by genistein detection in the infiltrated leaves. The results showed that the naringenin peaks were observed in both the plant species corresponding to that of naringenin standard. Wild-type A. tricolor var. tristis and S. oleracea showed 265.5 and 338.5 nM/g FW (nanomolar/gram fresh weight) of naringenin in the plant extracts. In the control samples, S. oleracea showed higher levels of naringenin content when compared to A. tricolor var. tristis. (Figure 7). Consequently, the naringenin content was also determined in the agroinfiltrated A. tricolor var. tristis and S. oleracea with pEarleyGate 102 HA and pEAQ-HT-DEST 3 vectors harboring IFS. As shown in Figure 7, the leaves of A. tricolor var. tristis and S. oleracea infiltrated with pEarleyGate 102 HA and pEAQ-HT-DEST 3 showed considerably less concentration of naringenin varying in the range of 65.5-134 nM/g FW compared to control. Quantification analysis showed that the leaves infiltrated with GmIFS accumulated comparable amounts of genistein. The peaks in the infiltrated leaves were identified with that of the standard peak of genistein. The wild-type control plants showed a negligible amount of genistein in comparison with the infiltrated leaves. The transiently transformed Quantification analysis showed that the leaves infiltrated with GmIFS accumulated comparable amounts of genistein. The peaks in the infiltrated leaves were identified with that of the standard peak of genistein. The wild-type control plants showed a negligible amount of genistein in comparison with the infiltrated leaves. The transiently transformed lines of A. tricolor var. tristis and S. oleracea also showed significant amounts of genistein accumulation with varying concentrations from 113 to 182.6 nM/g FW (Figure 7). It was observed that almost 65% of the concentration of naringenin was utilized as substrate for genistein production in A. tricolor var. tristis and about 50% in case of S. oleracea (Figure 8). The leaves of A. tricolor var. tristis and S. oleracea infiltrated with pEarleyGate 102 HA, and pEAQ-HT-DEST 3 harboring IFS showed significant genistein biosynthesis (Figure 8). Quantification analysis showed that the leaves infiltrated with GmIFS accumulated comparable amounts of genistein. The peaks in the infiltrated leaves were identified with that of the standard peak of genistein. The wild-type control plants showed a negligible amount of genistein in comparison with the infiltrated leaves. The transiently transformed lines of A. tricolor var. tristis and S. oleracea also showed significant amounts of genistein accumulation with varying concentrations from 113 to 182.6 nM/g FW (Figure 7). It was observed that almost 65% of the concentration of naringenin was utilized as substrate for genistein production in A. tricolor var. tristis and about 50% in case of S. oleracea ( Figure  8). The leaves of A. tricolor var. tristis and S. oleracea infiltrated with pEarleyGate 102 HA, and pEAQ-HT-DEST 3 harboring IFS showed significant genistein biosynthesis ( Figure  8).

Discussion
The phenylpropanoid pathway is one of the most studied metabolic frameworks that fulfills several physiological functions necessary for plant growth and development. This pathway involves an array of enzymes that are responsible for the production of different metabolites. IFS is one among them, as it catalyzes the synthesis of isoflavonoids genistein, daidzein from flavonoids naringenin and liquiritigenin, respectively, by an aryl ring migration [18,49,50]. As these isoflavonoids are reported to benefit plethora of beneficial activities in humans [3,8,11,14,18,50,51], the present research emphasized on engineering the metabolic pathway by transient expression of GmIFS in non-leguminous plants A. tricolor var. tristis and S. oleracea for the production of genistein which is not normally produced by this species.
Metabolic engineering holds a promising hope for the viable systematization of production of biologically active compounds that are necessary for human health and generation of improved crop varieties involving the redirection of enzymatic reaction(s) in endogenous pathways to synthesize novel products or enrich the existing ones [52][53][54][55]. Two approaches employed for the functional gene studies in plants are stable and transient expression. The development of stable transgenic lines by the stable integration of genome of plants is a time-consuming and laborious process. By contrast, the transient expression of heterologous genes in plants has considerable advantages over the stable expression, which is quick and inexpensive for the production of recombinant biopharmaceuticals and valuable metabolites [26,46]. Further, this approach could be utilized to study the activity of endogenous and exogenous enzymes in metabolic engineering, which provides insight to understand the mechanisms underlying in the complex gene regulatory and metabolic pathways. The transient gene expression is a rapid, flexible and cost-effective approach to study the expression of heterologous genes without the integration of desired gene(s) in the host genome [56,57]. The transient expression can be performed either by syringe infiltration or vacuum infiltration [58][59][60]. The high level of foreign gene expression can be obtained by co-expression of a viral-encoded suppressor of gene silencing which prevents the post-transcriptional gene silencing in the infiltrated tissues [25,30].
Here, we have introduced the GmIFS gene in pEarleyGate HA 102 and pEAQ-HT-DEST 3 by gateway cloning [21,61] and mobilized the vectors into A. tumefaciens for agroinfiltration in A. tricolor var. tristis and S. oleracea leaves. Further, the gene expression was confirmed by RT-PCR. The expected amplicon size of~1.5 kb corresponding to IFS gene was observed which confirmed the heterologous expression of IFS in the infiltrated leaf samples of A. tricolor var. tristis and S. oleracea. The results suggested that the isoflavonoid synthesis is catalyzed by the presence of IFS using the flavonoid intermediates [18,19,21]. The earlier reports showed that the stable expression of IFS gene in non-legume plants such as rice, tomato, Arabidopsis, Petunia and onion can produce genistein in transgenic plants [19,21,[62][63][64]. The HPLC results revealed that levels of naringenin in the leaf samples of infiltrated A. tricolor var. tristis and S. oleracea decreased significantly compared to that of control, indicating that this compound was utilized by the transformed IFS gene for genistein production. However, the bioactivity of genistein produced transiently in both the plants was not examined. The future work will focus on the elucidation of genistein activity and its production in crop plants. The present study demonstrated that both the plants could be used as hosts for the analysis and pathway engineering of metabolic enzymes. It also depicts that the transient system can be employed to study the function of new vector and validate the expression of both endogenous and exogenous enzymes in metabolic pathways in short time.

Plant Materials
The seeds of A. tricolor var. tristis and S. oleracea were procured from the seed center, Tamil Nadu Agricultural University, Coimbatore, India. These seeds were grown in soil under greenhouse conditions. The plantlets were allowed to grow for 4-5 weeks.

Construction of Plant Expression Vector
IFS cDNA (Genbank accession number: AF195798) was kindly provided by Dr. Oliver Yu and Dr. Brian Mc Gonigle, Donald Danforth Plant Science Center, USA. The IFS gene was cloned into plant expression vector pEarleyGate 102 HA and pEAQ-HT-DEST 3 by gateway cloning technology (Invitrogen, Carlsbad, CA, USA) as per manufacturer's protocol. The IFS gene was amplified by PCR, and purified PCR products were cloned into entry vector pDONR/Zeo and moved to destination vectors pEarleyGate 102 HA and pEAQ-HT-DEST 3, as described previously [21]. The transformed colonies in the destination vector were confirmed by gene-specific PCR and restriction digestion. Subsequently, the plant expression vectors were transformed to A. tumefaciens strain AGL1 by electroporation.

Agroinfiltration
A. tumefaciens strain AGL1 either containing pEarleyGate 102 HA or pEAQ-HT-DEST 3 plasmid harboring IFS gene under the control of CaMV 35S promoter were used for transient expression. The A. tumefaciens cultures were inoculated in 5 mL of yeast extract, and nutrient broth medium was supplemented with 60 mg L −1 rifampicin and 75 mg L −1 kanamycin and incubated at 28 • C for 2 days. The grown culture was centrifuged at 5000× g rpm for 15 min at 4 • C. The supernatants were discarded, and pellets were resuspended in Murashige and Skoog liquid medium. Acetosyringone (150 µM) was added to this suspension and incubated at room temperature for 2 h. Agrobacterium suspension was infiltrated into the leaves of A. tricolor var. tristis and S. oleracea leaf surfaces using the syringe infiltration technique [65]. The healthy leaves were infiltrated using 1 mL syringe without the needle on the abaxial leaf surfaces with Agrobacterium suspension to spread

RNA Isolation and cDNA Synthesis
All the materials used for the RNA isolation were treated with diethyl pyrocarbonate (Sigma-Aldrich, Saint Louis, MO, USA). The total RNA was isolated from the infiltrated leaves of A. tricolor var. tristis and S. oleracea after 4 days post infiltration (dpi). The infiltrated leaves and wild-type leaves were harvested and immediately frozen in liquid nitrogen. The total RNA was isolated from the leaves using total RNA isolation kit (Macherey-Nagel, Düren, Germany). The cDNA was synthesized from the total RNA with random hexamer primer using the M-MuLV reverse transcriptase (M-MuLV RTase) according to the manufacturer's instructions (Fermentas, Waltham, MA, USA). The reaction components used for the cDNA synthesis consisted of 1 µg RNA, 0.5 µL of random hexamer, nuclease-free water, and the reaction mixture was incubated at 65 • C for 2 min for denaturation. Immediately, it was kept in ice for 2 min, and the enzyme mixture containing 1U of M-MuLV RTase, buffer, dNTPs were added and kept at 37 • C for one hour and then inactivated at 95 • C for 5 min. The synthesized cDNA was used as template for PCR using the IFS gene specific primers. A reaction mixture containing the PCR master mix, primers and DNA template was prepared as mentioned in Table 1. β-Actin was used as internal control for the semiquantitative expression analyses.

HPLC Detection
The naringenin standard was obtained from Sigma Aldrich for HPLC analysis. After 4 days post infiltration, infiltrated leaf samples of A. tricolor var. tristis and S. oleracea were collected. Wild-type leaves from both the plants were collected and used as control. One gram of fresh leaves was homogenized in HPLC grade methanol. The homogenized mixture was incubated in a concentrator plus vacuum evaporator (Eppendorf, Germany) at 55 • C for 6-8 h. The dried samples were again extracted in HPLC methanol and filtered through 0.45 µm filter. Waters HPLC system equipped with a model 515 HPLC pump, Model 2998 photodiode-array detector and auto sampler was used for analysis. Separations were performed on a C 18 column with acetonitrile-water as mobile phase, and the reaction conditions were followed as described previously [21]. Naringenin in the samples was detected at 280 nm and identified by comparison of their retention times with those of standard.
The genistein standard was obtained from Sigma Aldrich for HPLC detection. Wildtype leaves from A. tricolor var. tristis and S. oleracea were used as controls. After 4 dpi, leaf samples of A. tricolor var. tristis and S. oleracea leaves were collected, homogenized in HPLC grade methanol and evaporated in concentrator plus vacuum evaporator (Eppendorf, Germany) at 55 • C for 6-8 h. The obtained dried samples were reconstituted in methanol, and 0.45 µL filtration was performed for HPLC analysis. The mobile phase was 80% methanol programmed at a flow rate of 1.0 mL/minute [51]. The injection volume was 20 µL, and the column was maintained at ambient temperature (24-26 • C). Genistein in the samples was detected at 260 nm and identified by comparison of their retention times with those of standard.

Statistical Analysis
Two biological replicates were used for HPLC quantification. The results were expressed as mean ± standard deviation.

Conclusions
The present study demonstrated that the transient expression of GmIFS gene in A. tricolor var. tristis and S. oleracea can produce the isoflavone genistein from the intermediate naringenin. The transient approach has been utilized to introduce the isoflavone pathway in these green leafy vegetables. To the best of our knowledge, this is the first report on transient expression of GmIFS in the A. tricolor var. tristis and S. oleracea for genistein production. The phenyl propanoid pathway being a complex metabolic pathway can be further explored in detail by the transient approach with infiltration or co-infiltration of multiple genes for complete understanding of the role of other enzymes and substrates involved in biosynthesis of isoflavonoids and other secondary metabolites. This proofof-concept study deciphers light on the production of health-beneficial isoflavones, and other pharmaceutically and nutraceutically relevant compounds by metabolic engineering, which can be employed to produce biofortified crops with better nutritional value for human health.

Funding:
The authors would also like to thank UGC-SAP and DST-FIST for the financial support.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.