Functional Identification of Acetyl-CoA C-Acetyltransferase Gene from Fritillaria unibracteata

Abstract

Fritillaria unibracteata is a rare and endangered medicinal plant in the Liliaceae family, whose bulbs have been used in traditional Chinese traditional medicine for over 2000 years. The mevalonate (MVA) pathway is involved in the growth, development, response to environmental stress, and active ingredient production of plants; however, the functional characterization of MVA-pathway genes in the Liliaceae family remains poorly documented. In this study, an Acetyl-CoA C-acetyltransferase gene (FuAACT) was first cloned from F. unibracteata. It exhibited structural features of the thiolase family and showed the highest sequence identity with the Dioscorea cayenensis homolog. The K_m, V_max, and K_cat of the recombinant FuAACT were determined to be 3.035 ± 0.215 μM, 0.128 ± 0.0058 μmol/(min·mg), and 1.275 ± 0.0575 min⁻¹, respectively. The optimal catalytic conditions for FuAACT were ascertained to be 30 °C and pH 8.9. It was stable below 50 °C. His361 was confirmed to be a key amino acid residue to enzymatic catalysis by site-directed mutagenesis. Subsequent subcellular localization experiments demonstrated that FuAACT was localized in chloroplasts and cytoplasm. FuAACT-overexpressing transgenic Arabidopsis thaliana plants showed higher drought tolerance than wild-type plants. This phenotypic difference was corroborated by significant differences in seed germination rate, lateral root number, plant height, and leaf number (p < 0.05). Furthermore, the FuAACT transgenic plants resulted in the formation of a more developed fibrous root system. These results indicated that the FuAACT gene revealed substantial biological activity in vitro and in vivo, hopefully providing the basis for its further research and application in liliaceous ornamental and medicinal plants.

1. Introduction

The Liliaceae family is composed of a variety of economically ornamental and medicinal plants. Among them, Fritillaria unibracteata Hsiao et K.C.Hsia (F. unibracteata) is a perennial herb that belongs to the genus Fritillaria, which is a resource of ornamental plants and pharmaceutical metabolites [1]. It is one of the original plants of the important medicinal herb “Chuan Beimu” [2]. It was originally distributed in alpine areas of western Sichuan, southern Qinghai, and southern Gansu with their altitudes mainly ranging from 3200 to 4500 m [3,4]. Furthermore, temperature and water supply play significant roles in the growth, development, and distribution of F. unibracteata [4]. F. unibracteata is rich in diverse steroidal alkaloids, terpenoids, and steroidal saponins, which are associated with significant pharmacological activities, such as the ability to promote expectoration, suppress coughing, and relieve asthma [5,6]. Subject to long-term wild resource exploitation, specific environmental needs, and a 4-5-year perennial life cycle, F. unibracteata has become endangered and been designated a Level II protected species in China. Although artificial cultivation has been applied to partly alleviate the pressure on wild resources, the limited increase in supply has not met the tremendous clinical demand [7]. Therefore, the functional identification of genes related to the biosynthesis of active ingredients in F. unibracteata will provide candidate genes for the metabolic engineering of active ingredients, with which wild resources of F. unibracteata will be eventually protected.

The biosynthesis of steroidal alkaloids, terpenoids, and steroidal saponins relies on the mevalonate (MVA) pathway in the cytoplasm and the 2-methyl-D-erythritol 4-phosphate (MEP) pathway in the plastid [5,8], which are involved in the formation of skeleton. Acetyl-CoA C-acetyltransferase (AACT), a member of the thiolase family, is responsible for the initial step of the MVA pathway, which catalyzes the condensation of two acetyl-CoA molecules to form acetoacetyl-CoA. The expression of AACT genes from various species has revealed species-specific expression patterns. For instance, Bacopa monniera AACT (BmAACT) is enriched in roots and petals [9], A. thaliana AACT2 (AtAACT2) is strongly expressed in seedlings, roots and inflorescences, while it is weaker in stems and leaves [10], and Houttuynia cordata AACT (HcAACT) has the highest transcriptional abundance in stems and rhizomes [11]; however, Ginkgo biloba AACT (GbAACT) is enriched in fruits and leaves [12]. More importantly, the thiolase family directly affects the biosynthesis of sterols, terpenoids, and phytohormones [13,14,15]. Positive correlation between AACT levels and saponin production was observed in Platycodon grandiflorus and Rehmannia glutinosa [16,17]. In Ganoderma lucidum, AACT transcript levels showed a similar significant positive correlation with triterpenoid production. This finding suggests that AACT plays a central role in regulating terpenoid metabolic pathways [18]. Moreover, Sando et al. [19] found that overexpression of the HbAACT gene promoted yeast growth, and Soto et al. [20] found that overexpression of the MsAACT1 gene enhanced both salt tolerance and squalene production in transgenic plants. Another study confirmed that the overexpression of AACT could improve the stress tolerance of B. monniera [9]. Collectively, these results suggest that AACT displays divergent expression profiles and functional roles in different plant species, making functional investigations of AACT across various plant lineages essential.

In our previous work, a total of five candidate FuAACT unigenes were obtained from F. unibracteata by de novo sequencing [5], and therefore, it is essential to carry out functional characterization of these candidate members. In this study, our main aims were to clone one FuAACT gene from F. unibracteata and to investigate its expression pattern in various tissues first. Second, the heterologous expression of the FuAACT gene in E. coli was measured, and the kinetic parameters of recombinant FuAACT were determined. Third, His361 was substituted with Ala, thereby confirming the essential role of His361 in condensation reaction. In addition, the subcellular localization of FuAACT was determined. Finally, transgenic A. thaliana plants overexpressing the FuAACT gene were constructed, and their drought tolerance was subsequently evaluated. To our knowledge, this paper is among the few studies concerning the MVA pathway genes in liliaceous plants. This work will contribute to providing a basis for further verification of other FuAACTs and other genes encoding MVA pathway enzymes.

2. Materials and Methods

2.1. Materials

2.1.1. Strain and Plasmid

E. coli strains (BL21, DH5α™) harboring pET28a-FuAACT, pGreenII0229-GFP, and pCambia2301 plasmids, along with Agrobacterium tumefaciens GV3101, were stored in the laboratory. All strains were purchased from Sangon Biotech (Shanghai, China), while study primers designed by Primer Premier 5 were synthesized by Tsingke Biotechnology (Beijing, China).

2.1.2. Reagents and Plant Materials

This study used a Plasmid Mini Kit I D6943 (OMEGA Bio-tek, Guangzhou, China), 1-5^TM 2× High Fidelity Master Mix (Tsingke, Beijing, China), Mut Express II Fast Mutagenesis Kit V2 (Vazyme, Nanjing, China), Gel Extraction Kit D2500 (OMEGA Bio-tek, Guangzhou, China), BCA Protein Assay Kit (Sangon, Shanghai, China), and Prime STAR HS DNA Polymerase (Takara, Dalian, China).

Fresh F. unibracteata samples from Qinghai Lvkang Biol. Dev. Co., Huzhu, Xining, Qinghai, China, were rinsed with distilled water and stored at −80 °C for analyses. Wild-type A. thaliana (ecotype Columbia-0) and Nicotiana benthamiana (N. benthamiana, ecotype LAB) seeds were kept in our laboratory.

2.1.3. Experimental Apparatus

Thus study employed a Veriti^TM96 Well Fast Thermal Cycler (Thermo Fisher, Waltham, MA, USA) and Laser confocal microscopy A1R+ (Nikon 1774059S, Tokyo, Japan).

2.2. Experimental Methods

2.2.1. Cloning and Bioinformatics Analysis of FuAACT

Transcriptome data of F. unibracteata tissues were previously obtained [5]. Based on the de novo sequence of FuAACT from the transcriptome data, the coding region of FuAACT was amplified using forward primer FuAACT—F1: 5′-GGATCCATGGCAAGCCTGCA-3′ (BamHI site in italics) and reverse primer FuAACT—R1: 5′-GAATTCTTAAATCAGTTCCAGAACCAGTGC-3′ (EcoRI site in italics). The amplified products were separated by agarose gel electrophoresis, gel-purified, digested with BamHI and EcoRI, and cloned into pET28a. The recombinant plasmid pET28a-FuAACT was sequenced to confirm that no substitutions or deletions occurred.

AACTs from various plants were downloaded from the National Center for Biotechnology Information (NCBI) database. The isoelectric point and molecular weight were calculated on the ExPASy website (https://web.expasy.org/protparam/, accessed on 28 January 2025). The homology modeling of FuAACT was followed with the SWISS-MODEL tool (http://swissmodel.expasy.org/, accessed on 10 April 2025). The final model of FuAACT was docked with the substrates using Autodock vina 1.5.6 [21]. Secondary structure prediction was performed using the Prabi online tool (https://npsa-prabi.ibcp.fr/, accessed on 10 June 2025). Multiple sequence alignment was performed on the AACTs from various species using MEGA 12.0 software. Following, a phylogenetic tree was constructed using the maximum likelihood (ML) approach on MEGA 12.0 software. The domains of the predicted encoded proteins were analyzed using InterPro (https://www.ebi.ac.uk/interpro/, accessed on 28 January 2025). Shannon entropy allows for the quantification of protein variation, whose value (H(X)) was calculated using the following equation in Python 3.13 software:

$$H\left(X\right)=-\sum _{i=1}^{n}p\left(x_i\right){log}_{2}p\left(x_i\right)$$

where p(x_i) represents the probability of a random variable x_i.

2.2.2. Protein Expression and Purification

The recombinant pET28a-FuAACT plasmid was transformed into E. coli BL21 (DE3) using the heat shock protocol. A single colony was subjected to rotation culturation at 37 °C in LB liquid medium containing 50 μg/mL kanamycin until OD600 reached 0.6. After induction with 0.1 mM Isopropyl-β-d-thiogalactoside (IPTG) overnight at 23 °C at 100 rpm, cells were pelleted, resuspended in binding buffer (10 mM Tris-HCl, pH 7.9, 500 mM NaCl, 20 mM imidazole), and sonicated at 30% power for 45 min (3 s on/3 s off). The supernatant was loaded onto Ni²⁺-NTA resin and eluted with 200 mM imidazole buffer to yield purified recombinant FuAACT.

The eluted recombinant proteins underwent sequential dialysis against 10% glycerol, 250 mM NaCl, and 2 mM Tris-HCl (pH 7.4) at 4 °C for 4 h, followed by 150 mM NaCl (same buffer) overnight. Crude and purified protein samples (10 μg each) were resolved by 12.5% SDS-PAGE and visualized via Coomassie Blue staining.

2.2.3. Enzymatic Assay of FuAACT

FuAACT (1 mL) was incubated with acetyl-CoA in 100 mM Tris-HCl (pH 8.8) containing MgCl₂ at 30 °C for 12 min, and acetoacetyl-CoA formation was monitored spectrophotometrically at 303 nm (the Mg²⁺-enolate complex; Thermo Fisher). Various temperatures, including 27 °C, 30 °C, 37 °C, 43 °C, and 50 °C, were adopted to measure the optimal temperature. Meanwhile, the relative activities of FuAACT at various pH environments, such as 6.4, 7.8, 8.4, 8.9, 9.4, and 9.9, were measured, and the given pH exhibiting the highest activity was determined as the optimal pH.

Thermal stability was assessed by pre-incubating FuAACT at 40, 50, 60, and 70 °C for 1 h, followed by residual activity assays under optimal conditions (pH 8.8, 30 °C).

2.2.4. Site Mutagenesis

To ascertain the role of His361 in the catalytic function of FuAACT, the His361 residue was replaced with an Ala residue using a pair of primers FuAACT-361H-A:5′-TAGTCTGGGTGCACCGCTGGGTTGCA-3′ and FuAACT-361H-A: 5′-CAGCGGTGCACCCAGACTAACTGCA-3′. According to the procedure of the Point Mutation Kit (Tiandz, China), pET28-FuAACT was used as a template to generate the mutated pET28-FuAACT^H361A plasmid [22]. Transformation, expression, purification, and activity assays of the mutant were conducted as previously described (Section 2.2.2 and Section 2.2.3).

2.2.5. Subcellular Localization of FuAACT

The prediction of protein subcellular localization was performed using various tools, including SignalP v4.1 (https://services.healthtech.dtu.dk/services/SignalP-4.1/, for the prediction of signal peptides, accessed on 28 July 2025), TMHMM Server v2.0 (https://services.healthtech.dtu.dk/services/TMHMM-2.0/, for the prediction of transmembrane helices, accessed on 28 July 2025) and WoLF PSORT (https://wolfpsort.hgc.jp/, for the prediction of subcellular localization, accessed on 28 July 2025) [23,24].

Using the pET28a-FuAACT plasmid as a template, primers FuAACT -F2: 5′-GAATTCATGGCAGCCGGGAGCGTT-3′ (EcoRI site in italics) and FuAACT -R2: 5′-CGCGGATCCGCGAATCAGTTTATGAATGG-3′ (BamHI site in italics) were applied to amplify the FuAACT fragment. Following digestion with EcoRI and BamHI, the PCR product was ligated into the corresponding sites of the pGreenII0229-GFP vector to generate the pGreenII0229-FuAACT-GFP translational fusion. The pGreenII0229-FuAACT-GFP plasmid was transformed into A. tumefaciens GV3101 using the freeze–thawing method and incubated in LB solid medium (containing 50 μg/mL gentamicin and 50 μg/mL kanamycin) to generate recombinant A. tumefaciens GV3101.

A. tumefaciens GV3101 carrying pGreenII0229-FuAACT-GFP was infiltrated into N. benthamiana leaves (1 mL) and incubated under low light for 2 days prior to confocal imaging. Chloroplast autofluorescence and GFP were excited at 640 nm (emission at 675 nm) and 488 nm (emission at 510 nm), respectively.

2.2.6. Expression Pattern of FuAACT Gene

The quantitative expression of the FuAACT gene in flowers, leaves, stems, and bulbs of F. unibracteata was analyzed by real-time PCR. Gene-specific primer pairs (5′-GCTGAAGAAACTCCGACC-3′ and 5′-CAAGAATGCGAGCACCACT-3′) were designed, adopting the 18S rRNA gene as the reference gene [25]. Each reaction consisted of 1 μL cDNA template in 14 μL reaction buffer, with thermal cycling conditions as follows: 1 cycle at 95 °C for 60 s, followed by 40 cycles at 95 °C for 10 s, 50 °C for 10 s, and finally at 72 °C for 10 s. Each reaction was repeated three times. Following the reaction, fluorescence curves were analyzed, and the relative expression of the FuAACT gene was calculated using the 2^−ΔΔCt method.

2.2.7. Construction of Transgenic A. thaliana Overexpressing FuAACT

The FuAACT gene fragment was amplified with primers FuAACT-F4: 5′-GCGGGTCGACGGTACCATGGCAAGCCTGCAGGCC-3′ and FuAACT-R4: 5′-TAGACATATGGGTACCTTAAATCAGTTCCAGAACCAGTG-3′ (KpnI site in italics). The amplified product and pCambia2301 plasmid were ligated via the homologous recombination method to generate the recombinant pCambia2301- FuAACT plasmid.

The plasmid pCambia2301-FuAACT was then transformed into A. tumefaciens GV3101 followed by the transformation of Arabidopsis thaliana using the leaf dip method. Transgenic lines were screened on half-strength MS medium supplemented with 50 μg/mL kanamycin. The DNA-based identification of transgenic lines was performed by PCR using primers 5′-GACGCACAATCCCACTATCC-3′ and 5′-GCTTCCGGCTCGTATGTTG-3′. The RNA-based identification of transgenic lines was performed by PCR using primers 5′-TATCTGGCAGAAGCACGCAA-3′ and 5′-TCTGCACACACACCCATACC-3′.

2.2.8. Statistical Analysis

All experimental values represent the mean ± SD of triplicates. An independent t-test or one-way ANOVA test was used to determine whether there was a statistically significant difference between the two groups. Statistical significance was defined as p < 0.05.

3. Results

3.1. Sequence Analysis of FuAACT Gene

The FuAACT gene was successfully cloned from the rhizome of F. unibracteata and submitted to NCBI under accession number PV706214. The FuAACT gene consists of 1218 bp with a GC content of 46%, encoding a 405-amino-acid peptide with a molecular weight of 41.79 kD. To evaluate the evolutionary relationship of FuAACT, 60 AACTs from various species were collected, including 54 sequences from angiosperms, 1 sequence from moss, 1 sequence from algae, and 4 sequences from fungi. The maximum likelihood (ML) method was used to construct a phylogenetic tree, with bootstrap values exceeding 70 for 80.05% of nodes, confirming its high reliability (Figure 1). In the ML tree, fungi AACTs formed a root branch with high bootstrap (99 BP), indicating that they might represent ancestral genes for AACTs in algae, muscus, and spermatophyte. The ingroup corresponding to spermatophyte was strongly supported (100 BP), in which angiosperm was monophyletic (99 BP) and a sister to Ginkgo biloba from gymnosperm, suggesting that AACT from angiosperms might originate from AACT from gymnosperm. Furthermore, dicots, as the largest plant group, were paraphyletic, with the majority falling into a moderately supported clade (95 BP). FuAACT clustered closely with monocot AACTs, supported by relatively high bootstrap values (100 BP). Since FuAACT and Dioscorea cayenensis AACTs formed the closest clade (98 BP), a close genetic relationship between the two species was proposed.

3.2. Conserved Motif Analysis of AACT

InterPro analysis identified two conserved domains in the FuAACT protein: IPR020616 (N-terminal; 15–274 amino acids) and IPR020617 (C-terminal; 283–403 amino acids). Two motifs involved in the catalytic function, NVHGGAVSLGHPLGCSG (blue box in Supplementary Material S2) and GVASVCNGGGASA (green box in Supplementary Material S2), were significant and were also detected in the C-terminal regions of FuAACT. The core heptapeptide CNGGGAS catalyzes the Claisen condensation of acetyl-CoA to acetoacetyl-CoA [26,27]. Moreover, several key residues such as Cys99, His361, and Cys391 were absolutely conserved in all AACTs (Supplementary Material S2). Specifically, Cys99 is involved in the formation of acylase intermediates, while Cys391 and His361 are responsible for the deprotonation and stabilization of the transition state, respectively [28]. Conserved motif analysis indicated that a total of 11 motifs (Figure 2) were distributed in AACTs with p-values ranging from 1.69 × 10⁻⁹⁹ to 3.48 × 10⁻¹⁹⁴ (Supplementary Material S3), confirming high reliability of this analysis. Among them, motifs 2, 3, 4, 5, 6, 9, and 10 were conserved across all AACTs. Two key functional motifs, NVHGGAVSLGHPLGCSG and GVASVCNGGGASA, were distributed in motif 10 and motif 11, respectively. Thus, whether D. catenatum AACT lacks thiolase activity due to the absence of motif 11 needs to be verified in the future. Overall, these conserved motifs and key residues observed in FuAACT imply that FuAACT might share enzymatic activity similar to that of other AACTs.

3.3. Protein Variability Analysis of AACT

The quantification of protein variation was achieved by calculating Shannon entropy between FuAACT and AACT proteins from monocot and dicot plants. The mean Shannon entropy of AACTs in monocots (0.256 ± 0.222) was significantly lower than in dicots (0.469 ± 0.362), reflecting greater sequence divergence between FuAACT and dicot AACTs compared to monocot AACTs (Supplementary Material S4). Whereas monocots showed minimal variability (Shannon entropy ≥ 2.0), dicot AACTs possessed six highly variable N-terminal loci (Shannon entropy ≥ 2.0), demonstrating that the N-terminal region underlies sequence variability. The top five conserved regions within the monocotyledonous and dicotyledonous AACT proteins were identified with an average Shannon entropy ranging from 0.034 to 0.157 (

Table 1. Analysis of top five conserved regions of AACT proteins.

No	Start	End	Sequence	Average Shannon Entropy
1	373	393	TLLGVLRQRNGKTGVASVCNG	0.034
2	92	111	CTTINKVCASGMKATMLAAQ	0.085
3	146	165	GHDTIVDGMLKDGLWDAYND	0.085
4	205	217	AFSWEIVPVEVSA	0.150
5	254	273	PALAIPKALSNADLE	0.157

), where residues from Thr373 to Gly393 contained the core catalytic Cys391-Asn392-Gly393 tripeptide.

3.4. Homology Modeling and Important Catalytic Sites of FuAACT

The Prabi online tool predicted that α-helix was the main secondary structure in FuAACT, comprising 178 residues (43.95%), followed by random coils (31.85%), extended strand (15.06%), and β-turn (9.14%). This conformational feature is consistent with those of typical thiolases [29]. To identify key residues critical for enzymatic activity, homology modeling was conducted using SWISS-MODEL with template 5xyj.1.A (53.92% identity to FuAACT). Structural visualization using PyMOL software confirmed that α-helix was the predominant secondary structure in FuAACT, consistent with the secondary structure prediction. FuAACT composed four identical monomers (Figure 3A), and its monomer structure was similar to the 3D conformation of the thiolase reported by Wiesenborn et al. [30]. Each monomer contained canonical N-terminal (15–273, purple) and C-terminal (283–403, green) regions (Figure 3B).

Notably, the C-terminal domain protruded from the N-terminal domain and contributed to most of the active sites (Figure 3C). Multiple sequence alignment and structural analysis revealed that Cys99, His361, and Cys391 formed a fully conserved catalytic triad, with His361 situated in the loop region containing the active site. Additionally, His361 was part of the thiolase-conserved motif NVHGGAVSLGHPLGCSG (residues 351 to 367) in the C-terminal region. Molecular dynamics simulation showed that the binding free energy of the FuAACT–substrate complex decreased significantly from −7.0 kcal/mol to −5.1 kcal/mol, and the key hydrogen bond interaction between His361 and the substrate was completely disrupted (Figure 3D). These results highlight the potential role of His361 in the catalytic function of FuAACT.

3.5. Heterologous Expression, Purification, and Activity Assay of FuAACT

The recombinant FuAACT protein was successfully expressed in E. coli BL21 and purified by Ni-agarose resin. Its molecular weight was 45 kDa, higher than the predicted mass (41.79 kDa) due to the N-terminal His-tag fusion peptide (Figure 4). Before determining kinetic parameters, the optimal temperature, pH, and thermal stability of purified FuAACT were determined. FuAACT exhibited optimal activity at 30 °C and pH 8.9. After thermal treatment at 50 °C for 3 h, FuAACT retained 63.08% of the initial activity, whereas activity decreased significantly to 6.04% at 60 °C, indicating that FuAACT exhibits good thermal stability below 50 °C.

Then, under the optimal reaction conditions, the K_m, V_max, and K_cat/K_m values of FuAACT were determined as 3.035 μM, 0.128 μmol/(min·mg), and 0.421 μM⁻¹·min⁻¹, respectively. To investigate the role of His361 in FuAACT catalysis, this residue underwent site-directed mutagenesis. As shown in

Table 2. Kinetic parameters of FuAACT and FuAACT-H361A.

	Vmax (μmol/(min·mg))	Km (μM)	Kcat (min−1)	Kcat/Km (μM−1·min−1)
FuAACT	0.128 ± 0.0058	3.035 ± 0.215	1.275 ± 0.0575	0.421 ± 0.0109
FuAACT-H361A	0.006 ± 0.00496 ***	2.41 ± 1.95398	0.064 ± 0.0496 ***	0.029 ± 0.00251 ***

Note: Significant differences in kinetic parameters between FuAACT and FuACAT-H361A were analyzed with a t-test. *** represents p < 0.001.

, this point mutation significantly affected FuAACT catalysis. For FuAACT^H361A, the K_m, V_max, and K_cat/K_m values changed to 2.41 μM, 0.006 μmol/(min·mg), and 0.029 μM⁻¹·min⁻¹, respectively (Table 2). Compared with wild-type FuAACT, FuAACT^H361A exhibited decreases of 95.3% and 93.1% in V_max and K_cat/K_m, respectively. This indicates that the mutant has increased substrate affinity but significantly decreased catalytic efficiency. His361 is a conserved residue in the catalytic core of the thiolase family, and its nitrogen-containing heterocycle structure is essential for maintaining the active site conformation. Upon mutation to alanine, loss of the heterocycle structure may disrupt π-π interactions with neighboring residues (e.g., Cys99 and Cys391), leading to the distortion of the active site structure and subsequent impairment of catalytic efficiency.

The results show that FuAACT catalyzes the condensation of two molecules of acetyl-CoA to acetoacetyl-CoA, and its activity depends on the integrity of the conserved residue His361. Through functional validation and site-directed mutagenesis, these findings clarify the key role of His361 in the enzymatic catalytic mechanism, providing a theoretical foundation for further investigating the molecular regulation of steroid alkaloid synthesis in F. unibracteata bulbs.

3.6. Analysis on Sucellular Localization of FuAACT

SignalP v4.1 and TMHMM Server v2.0 predicted that the FuAACT protein lacked signal peptides and transmembrane domains, suggesting its role as an intracellular protein. To confirm its subcellular localization, the amino acid sequence of FuAACT was analyzed using WoLF PSORT. The results indicate that FuAACT was predicted to localize to the chloroplast, cytoplasm, and nucleus.

To verify the prediction, plasmids pGreenII0229-GFP and pGreenII0229-FuAACT-GFP were constructed and transiently introduced into N. benthamiana. Under laser confocal scanning (Figure 5), the control GFP showed an indiscriminate green fluorescence, while the chloroplast revealed red fluorescence. The overlapping of FuAACT-GFP fluorescence with chloroplast autofluorescence produced yellow fluorescence, indicating co-localization. This result indicates that the FuAACT protein is predominantly localized to chloroplasts and the cytoplasm, where the MVA pathway exists.

3.7. Expressional Pattern of FuAACT Gene

To further investigate the expression pattern of the FuAACT gene, real-time PCR was used to detect its expression in the stem, leaf, flower, and bulb of F. unibracteata. The results (Figure 6) show that FuAACT expression was tissue-specific, with the highest levels in the leaf (10.56), followed by the stem (4.7), bulb (1.0), and flower (0.99) (p < 0.001). These results suggest that high FuAACT expression in the leaf may supply sufficient intermediates for the synthesis of downstream metabolites.

3.8. Phenotypic Analysis of Transgenic FuAACT A. thaliana Under Drought Stress

The PCR verification of transgenic plants was performed using DNA as a template (Figure 7A). Meanwhile, real-time PCR was carried out to evaluate the expression data of the FuAACT gene in transgenic Arabidopsis, by which substantial expression of the FuAACT gene was observed in the transgenic Arabidopsis, while a negative result was observed in the wild-type Arabidopsis (Figure 7B). The results confirm that the transgenic plants overexpressing the FuAACT gene were constructed successfully. In the seed germination experiment, seeds of both transgenic and wild-type A. thaliana began to germinate on day 1, followed by a rapid germination phase between days 1 and 3, reaching maximal germination on day 5. Statistical results showed that the final germination rate of transgenic A. thaliana was 67.20%, which was significantly higher than that of the wild type (57.21%) (Figure 8A), suggesting that overexpression of the FuAACT gene may promote germination. It is hypothesized that FuAACT may promote acetoacetyl-CoA by activating the MVA pathway, thereby upregulating the accumulation of downstream plant hormones, such as cytokinins and gibberellin (GA), to improve seed germination efficiency.

Under normal conditions, the transgenic A. thaliana plants grew slightly better than the wild type, while the transgenic plants showed stronger adaptability to drought stress (Figure 8). Under drought stress, the leaves of transgenic plants remained green, while those of wild-type plants exhibited slight etiolation. Additionally, transgenic plants exhibited a significantly higher number of lateral roots, greater plant height, and increased leaf number compared to wild-type plants (p < 0.05). The improved roots were conducive to nutrient and water absorption to absorb nutrients and water from soil, while the improved leaves might enhance photosynthesis to provide sufficient carbohydrates. Overall, these results identify FuAACT as a candidate gene for drought tolerance engineering, despite the underlying molecular mechanisms remaining unclear.

4. Discussion

Vollack and Bach (1996) first cloned the AACT gene in the plant cytoplasm and confirmed its enzymatic activity [31], which promoted the subsequent in-depth research on plant metabolic engineering and stress tolerance. Current studies have highlighted the role of AACTs in isoprenoid synthesis (e.g., squalene, natural rubber, triterpene saponins) by regulating precursors in dicotyledonous plants such as Medicago sativa [20], Helianthus annuus L. [14], H. brasiliensis [19], and B. monniera [9]. However, the functional identification of AACT in the Liliaceae family has not been reported to date.

In this study, the FuAACT gene was cloned for the first time. Its amino acid length was similar to those of BmAACT [9], HbAACT [19], SaAACT [32], TaAACT [33], and other dicots (365–437 AA). Sequence alignment showed the conserved key catalytic residues (Cys99, His361, and Cys391) along with the motif NVHGGAVSLGHPLGCSG (motif 10) and an active site GVASVCNGGGGASA in the C-terminal region. These conserved sites and motifs implied that the construction of the active site of AACT remained canonical during plant evolution, consistent with the report of Xiong et al. [33]. Compared with AtAACT2 [10] and BmAACT [9], FuAACT exhibited higher affinity (lower K_m) but lower catalytic efficiency (K_cat/K_m), indicating that the enzymatic properties of AACTs varied in different plants. Such differences in kinetic properties might result from the variation in amino acids in the binding sites, including His353, Ala388, Ser389, and Val390. Furthermore, the results also suggest that AACTs in different plants might control the metabolic flux through the mutation of key residues at precise positions of key residues. In comparison with F. unibracteata mevalonate kinase (FuMK) [5], FuAACT exhibited a lower catalytic rate due to its lower V_max, suggesting that FuAACT may play a more critical role than FuMK in steroid alkaloid biosynthesis in F. unibracteata.

The conserved His361 residue in the active site of FuAACT was proposed to be the catalytic residue [28]. Notably, in existing thiolase studies, the His348 residue forms a hydrogen bond with nucleophilic cysteine (Cys89) to enhance its nucleophilicity. Mutants H348A or H348N lost 99–99.9% of their original activities [34]. A similar hydrogen bond was also observed in the FuAACT-acetyl CoA complex using homology modeling, and the subsequent substitution of H361 residue by Ala residue caused substantial loss in catalytic efficiency but improved affinity. In combination with site-directed mutagenesis and model observation, it was speculated that the mutation enhanced the hydrophobicity in the substrate pocket, thereby improving affinity. However, the disruption of the hydrogen bond between His361 and Cys89 residues destroyed the electron transfer and reduced the nucleophilicity of Cys89 residue. In addition, based on the results of amino acid sequence alignment (Supplementary Material S2) and molecular docking (Figure 3D), the conserved Phe28, Ileu262, and Ser358 residues also participate in the catalytic process, whose roles in catalysis will be verified in future work.

Diverse subcellular localization of AACT has been observed in different plants, although the MVA pathway is canonically localized to the cytoplasm. Indeed, the cytoplasm, endoplasmic reticulum, nucleus, and peroxisomes have been reported as AACT localization sites [1,10,35,36]. In this study, FuAACT was found to be localized in both plastids and cytoplasm, suggesting that various subcellular compartmentalization was proposed to be involved in the biosynthetic pathway of terpenoids. This finding echoes the reported involvement of AACT in chloroplast metabolism in A. thaliana [10], strengthening the role of chloroplast in the isoprenoid synthesis. Another possible reason is that AACT might carry non-enzymatic functions, similar to glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH). GAPDH is abundant in the cytosol but is also found in the nucleus where it plays roles via non-enzymatic binding [37]. However, this simultaneous localization of FuAACT must be further verified, and hence, a plastid-localized marker, such as ribulose bisphosphate carboxylase, will be used in future work [38]. Tissue-specific expression analysis further revealed that FuAACT expression was significantly higher in leaves than in roots, bulbs, and flowers. A similar leaf-enriched distribution of AACTs was observed in G. biloba [12] and Isodon rubescens [39]. Combined with its subcellular localization, we speculate that FuAACT may play a significant role in leaf tissue. Previous studies have confirmed that the enhancing expression of MVA pathway genes in the leaf tissue drove the accumulation of terpenoids such as carotenoids [40], an important pigment in photosynthesis. Therefore, the role of FuAACT in the biosynthesis of steroidal alkaloids in bulbs of F. unibracteata might be indirect, or its role is being replaced with other FuAACT homologs. However, the above speculation needs further experimental verification.

The overexpressed FuAACT gene promoted seed germination and drought tolerance of A. thaliana, which might be due to the increased synthesis of phytohormones. However, the accumulation of various phytohormones needs to be measured in future studies to validate this hypothesis. Additionally, to elucidate the role of FuAACT in stress tolerance, future studies should determine (1) whether subcellular localization of FuAACT changes under drought and salinity stress; (2) the expression pattern of FuAACT under these stresses; (3) the change in metabolites, such as isoprenoid precursor and downstream terpenoids, between transgenic and wild-type plants; and (4) the crystal structure of FuAACT to clarify its catalytic mechanism.

5. Conclusions

In this study, the FuAACT gene was first cloned and characterized in vitro and in vivo. The His361 residue was verified as a key catalytic residue, which lays the foundation for understanding the catalytic mechanism and modification of FuAACT. Overexpression of FuAACT improved tolerance to drought and salinity stress, although the detailed mechanism remained unclear. In conclusion, this work extended the functional AACT members of the Liliaceae family, rich with economically ornamental and medicinal plants, but also enlarged the pool of candidate genes for plant stress resistance.