Nepeta nuda plants were collected from natural habitats (Pirin; Rhodopes; and two samples Rila #1 and Rila #2 collected from the same location with 3-week difference) in Bulgaria during the flowering period (Supplementary Table S1). Samples of vouchers were deposited to the Herbarium of Sofia University “St. Kliment Ohridski,” Sofia, Bulgaria (the vouchers are listed in Supplementary Table S1). In vitro plants were maintained in the collection of the Department of Plant Physiology at the Faculty of Biology, as described by Nedelkova et al. (2012) and Dragolova et al. (2015). The plants were grown under sterile controlled conditions on MS (Murashige and Skoog, 1962) medium with 3% sucrose and 0.7% agar under white light (80 μmol m⁻²s⁻¹ photosynthetic active radiation, cool white fluorescent TL-D 36W/54-765 1SL/25 Philips), photoperiod 16 h light/8 h dark, 25±1°C, 60–70% humidity. The plants were passaged every 4 weeks via shoot tips. The ex vitro plants were adapted for 3 months in a greenhouse and transferred to soil for 1 year. Wild-grown N. nuda (in situ Pirin and ex vitro) and in vitro cultured plants (Dragolova et al., 2015) were used for comparative analyses (n ≥ 15; Supplementary Table S1). For dry weight (DW), the plants were dried in the dark at room temperature. The plant extracts preparation is described in Supplementary Material and summarized in Supplementary Table S2.

For the light treatments, the internodes from in vitro cultivated plants were grown for 5 weeks under different light quality and intensity as discussed in the following: White (W) light (80 μmol m⁻² s⁻¹; fluorescent) as a control, and a combination of blue and red light (blue:red:far red 15%:75%:10%) with two different light intensities: BR (240 μmol m⁻² s⁻¹; LED), and BRS (40 μmol m⁻² s⁻¹, LED, i.e., as “shadowed”) (Zhiponova et al., 2020a). The similar-looking plants were selected for analysis (n ≥ 20).

The procedures for spectrophotometric quantification of phenolics and flavonoids are described by Zhiponova et al. (2020b), and performed with spectrophotometer Shimadzu UV 1800 (Kyoto, Japan). The content of polyphenols was determined according to Singleton et al. (1999) by a standard curve created using the known concentrations of gallic acid (GA). The content of flavonoids was measured according to the protocol of Chang et al. (2002) by a standard curve based on the known concentrations of quercetin (Q).

The extracts obtained by different solvents (Supplementary Table S2A) were tested to determine 50% effective concentration (EC₅₀) for radical scavenging activity using stable 2,2'-diphenyl-1-picrylhydrazyl (DPPH) free radical (Blois, 1958). The extract concentrations between 50 and 200 μg ml⁻¹ were prepared in 1.5 ml methanol followed by the addition of 0.5 ml of 0.1 mM DPPH. Methanol was used as a blank and a mixture of methanol (1.5 ml) and DPPH (0.5 ml) was utilized as a negative control. The samples were incubated for 30 min in the dark, and the absorbance was measured at λ = 517 nm with spectrophotometer Shimadzu UV 1800 (Kyoto, Japan). The results are given as EC₅₀ in μg ml⁻¹ extract.

The crude methanol extracts from plant variants were prepared as described in Supplementary Material and presented in Supplementary Table S2B. The DPPH-radical-scavenging activity was measured using Trolox as a standard, according to the method of Brand-Williams et al. (1995). The reaction included 80 μl extract (1 mg DW ml⁻¹) and 1920 μl 0.06 mM DPPH solution in methanol, and the samples were incubated at room temperature in the dark for 30 min. After the calibration against methanol, the absorbance was detected at λ = 515 nm with spectrophotometer Shimadzu UV 1800 (Kyoto, Japan). The radical–scavanging activity was measured in μM Trolox ml⁻¹ via a standard curve. The DPPH–scavanging activity was expressed as mM Trolox per gram DW (mM Trolox gDW⁻¹).

Madine and Darby bovine kidney (MDBK) cells (purchased from ATCC) were used and grown in Dulbecco's Modified Eagle Medium (DMEM, Sigma–Aldrich) supplemented with 8% fetal calf serum (FCS, Sigma–Aldrich) (with 8 μg ml⁻¹ gentamycin, SANDOZ, and 10 mM HEPES buffer, Sigma–Aldrich). The maintenance medium contained FCS with reduced concentration of 4%. The strain F of Human alphaherpesvirus type 1 (HHV-1) (purchased from ATCC) was propagated in MDBK cells and stored at −70°C (BINDER GmbH, Germany) until used. This HHV-1 strain is sensitive to the antiviral agent acyclovir (Hinkov et al., 2020). The virus titer was determined by cytopathic effect (CPE) assay using the method of Reed and Muench (1938) and plaque assay (Dulbecco, 1952) in MDBK.

The following bacterial strains were purchased from the Bulgarian National Bank for Industrial Microorganisms and Cell Cultures (NBIMCC): The Gram-negative bacteria Acinetobacter calcoaceticus NBIMCC 3730 and Klebsiella pneumoniae NBIMCC 3670, and the Gram-positive bacteria Bacillus cereus NBIMCC 1085 and Staphylococcus aureus ATCC 25923. The bacterial strains were cultured overnight at 37°C on Muller–Hinton agar (MHA) medium.

All measurements were performed in the period 2019–2021 and included plant material from 3 to 6 biological repeats and technical replicates (n ≥ 3). Each biological repeat contained plant material from an average 15 plants (n ≥ 15). The results are presented as mean values ± standard errors (SE). To evaluate statistical differences among the treatment variants, one-way ANOVA followed by Holm–Sidak test was performed using Sigma Plot 11.0 software. The differences were considered significant at p < 0.05. The data processing for the antiviral assay was done by the Origin 9.1 program. In Sigma Plot, the Pearson's correlation coefficient (R) was calculated using average parameter values of the biological repeats for the variants (n ≤ 6) with a level of significance 0.05. The principal component analysis (PCA) of the metabolite content in N. nuda light variants was performed in R 3.6.3 using prcomp function after logarithmic normalization of the data set. PCA plot was done with ggbiplot R package.

The genetic discrimination of N. nuda was a prerequisite step for characterization and validation of its taxonomic position. To achieve this goal, we applied the DNA barcoding method based on the conserved regions ITS and chloroplast (rbcL, matK, trnH-psbA) genomes (Supplementary Table S4). The generated DNA barcode sequences were submitted to the BOLD database² (accession number BUL002-22³). It should be noted the lack of other accessions in databases belonging to the species N. nuda. For the construction of phylogenetic trees, we retreived Nepeta accessions from the National Center for Biotechnology Information (NCBI) and BOLD databases, which resulted in a set of 56 sequences for ITS, 44 for rbcL, 44 for matK and 5 for trnH-psbA. Among these sequences, we selected accessions with high homology to N. nuda and covered all the available in genebanks species diversity, as the very short fragments were discarded. To make the tree less redundant, one to two accessions per Nepeta species from databases were retained, while other accessions with identical sequences and names were discarded. Among the used markers, the ITS-derived sequences for Nepeta species were the most enriched in BOLD and NCBI databases, and the phylogenetic tree included 47 Nepeta species. The studied N. nuda appeared to be most similar to N. sheilae (98.7% identity) in a subclade together with N. schiraziana, N. deflersiana, N. isaurica, N. congesta, N. heliotropifolia, N. crassifolia, and N. cataria (Figure 1). The number of Nepeta species for rbcL, and matK regions was 15 and 10, respectively (Figures 2A,B). For trnH-psbA, only NCBI accessions corresponding to six Nepeta species were found (Figure 2C). The taxonomic assignment based on chloroplast markers was not congruent with the data for the ITS marker. The data for all three chloroplast DNA barcodes revealed that N. nuda falls into one cluster with N. italica, N. grandiflora, N. hemsleyana, N. cataria, and N. tuberosa (Figure 2D). The DNA barcode rbcL showed 100% identity of N. nuda to N. italica, N. grandiflora, and N. cataria. Furthermore, N. nuda was very close to N. italica, N. grandiflora (99.8%), and N. cataria (99.6%) when considering the data from matK marker (data are not shown). The trnH-psbA marker showed 97.9% and 97.3% similarity to N. italica and N. cataria, respectively. The results showed the highest identity between N. nuda and N. italica. The consensus tree of all three chloroplast markers supported the assignment of analyzed N. nuda to N. italica, and N. grandiflora with a bootstrap value of 84% (Figure 2D).

To infer the impact of the habitat on the level of phenolic compounds in flowers and leaves, a comparison was performed between N. nuda populations collected from three different regions of Bulgaria—Pirin, Rhodopes, and Rila mountains (two samples) (Figure 3A; Supplementary Table S1). The quantity of phenols varied significantly in plants from the distinct habitats. The two samples from Rila were collected at the same place with a difference of growth by 3 weeks which was reflected by a noticeable variation between the respective phenolic content. The plants from Rila were enriched in phenolic compounds compared to the other two populations. As a general trend, more phenolic compounds (up to 30 mg gDW⁻¹) were found in flowers compared to leaves (Figure 3B). The opposite trend was observed for the flavonoid content with higher values in the leaves than in the flowers (Figure 3B). The largest quantity of flavonoids (~10 mg gDW⁻¹) was detected in the leaves of populations from Rhodopes and Rila. Next, we compared the fluctuation in phenolic content between the flowers and leaves of wild-grown and in vitro cultivated plants (Figure 4A). The phenol content was lowest in the in vitro control (15.9 mg gDW⁻¹) and highest in the flowers of ex vitro adapted plants (38.2 mg gDW⁻¹) (Figure 4B). Flavonoids were detected in all samples in larger quantities than in the in vitro plants (Figure 4B).

Before testing the biological activities of the extracts, a comparative analysis of the yield using different solvents (water, methanol, ethanol, acetone, and chloroform) was carried out. The highest yield was achieved when water was used for extraction (nearly 37%), and the lowest with acetone (4%) as a solvent (Supplementary Table S2A). The extraction with water and methanol produced the highest phenolic content (nearly 100 mg gExtract⁻¹), and the lowest was detected with chloroform (7-mg gExtract⁻¹). Flavonoid content was enriched in acetone and ethanol (68- and 24-mg gExtract⁻¹, respectively) extracts. The EC₅₀ for DPPH-scavenging demonstrated the largest antioxidant activity in water and methanol extracts followed by ethanol, acetone and chloroform (in chloroform the activity was almost lacking). Pearson correlation analysis in Supplementary Table S2A shows a significant correlation between the amount of phenolics and DPPH scavenging activity (R = −0.706 with p = 0.001). Furthermore, a comparison between the crude methanol extracts of N. nuda variants revealed the most effective DPPH radical scavenging (mM Trolox gDW⁻¹) in the flowers of wild-grown plants followed by the leaves, whereas a pronounced decline in the in vitro cultivated plants was observed (Table 1). In accordance, significant correlation (R = 0.943, p = 0.0004) was confirmed between the DPPH activity and the phenolic content.

Next, the biological activity of N. nuda water extracts from the ex vitro flowers and leaves and the in vitro plants was compared. To evaluate the cytotoxicity of the extracts, the concentrations in the range 0.125–9.0 mg ml⁻¹ were applied on MDBK cells (Table 2). The most toxic was the extract from the leaves of ex vitro plants (CC₅₀ = 5.2 mg ml⁻¹) and the least toxic was the water extract from in vitro grown plants (CC₅₀=7.5 mg ml⁻¹). The strong antiviral potential of water extracts from the flowers of ex vitro grown N. nuda manifested by the inhibition of viral replication and cell survival was dose dependent. The treatment of HHV-1 strain F with the flower water extract was performed simultaneously and 1 h after cell inoculation reaching antiviral cell protection at MNC 81.37% (EC₅₀ 0.599 mg ml⁻¹) and 65.62% (EC₅₀ 1.155 mg ml⁻¹), respectively (Table 2). The water extract from the leaves of ex vitro grown N. nuda also showed substantial antiviral potential (Table 2) but only after simultaneous cell inoculation. At MNC, the cell protection reached 75.18% (EC₅₀ 0.725 mg ml⁻¹). After the addition of leaf water extracts 1 h after cell inoculation at MNC, the cell protection hardly reached 43.87%. Water extracts from in vitro grown plants showed no antiviral activity (Table 2).

Further, we tested the antibacterial activity of N. nuda extracts against selected Gram (–) and Gram (+) bacteria by three different methods: Disk diffusion, microdilution with the establishment of MIC in liquid bacterial suspension, and lethal MBC by simultaneous plating of extract and bacteria on an agar plate (Table 3; Supplementary Figure S1). The antibacterial activity of N. nuda extracts against the tested bacterial strains was confirmed by all the three approaches. Interestingly, the extracts from in situ and in vitro plants showed similar activity, which suggests that the antibacterial activity of N. nuda was not affected by environmental conditions. The disk diffusion assay was suitable for testing of phenolic standards, as its application indicated that the phenolic acids FA and GA had antibacterial activity against the tested bacterial strains, while the flavonoid Q did not have any effect (Supplementary Figure S2).

The profiling of polyphenols and iridoids in N. nuda methanol extracts resulted in the identification of 63 compounds that were divided into the following four major groups: (i) Phenolic acids and derivatives (30 compounds); (ii) flavonoid glycosides and aglycones (21 compounds); (iii) iridoid glycosides (11 compounds); and (iv) one compound belonging to another class (Table 4; Supplementary Table S6). In the studied samples, 12 derivatives of hydroxybenzoic acids (1–6, 8, 12, 15, 17, 21, 25) were identified including GA hexosides (1, 3), hydroxybenzoic acid-related compounds (5, 6, 8, 12, 15, 17, 21, 25), protocatechuic acid (2) and vanillic acid (4). The hydroxycinnamic acids were represented by 10 compounds as caffeic acid (20) and its derivatives (7, 9, 10, 13, 27), FA (16) and its derivative (14), RA (18), and its derivative (24). The caffeic acid esters nepetoidin B1 (22) and nepetoidin B2 (28), and the caffeic acid oligomer clinopodic acid A (23) were also found. In addition, two hydroxycoumarins, aesculin (11) and aesculetin (19), were detected. Three polyphenols that are combination of two Danshensu molecules were detected: salvianolic acid C (26), and the derivatives methyl salvianolate C1 (29) and methyl salvianolate C2 (30). The identified flavonoids belonged entirely to the subgroup of flavones with eight luteolin-related compounds (31, 34–37, 41–43), 10 apigenin-related compounds (32, 33, 38–40, 44–47, 49), thymusin (48), cirsimaritin (50), and xanthomicrol (51). Among the identified flavonoid glycosides, except apigenin 7-O-hexoside (32), all other derivatives were glucuronides of the flavones apigenin and luteolin. A total of 11 iridoid glycosides were identified as listed in the following: Two nepetanudosides (52, 56), dihydrocornic acid (53), secologanin (54), six epideoxyloganic acid-related compounds (eDLA) (55, 57–59, 61, 62), and geniposidic acid (60). The compound 12-O-hexosyl-jasmonate (63) was placed in the group “Other”.

Based on the identified compounds in this work and additional phenolic-related compounds (labeled 64^* to 75^*), 21 compounds in total were selected for comparative quantification between the studied N. nuda samples by the UHPLC/DAD/qqqMS method (Figure 5). These 21 compounds were presented by 7 phenolic acids, 12 flavonoids, 1 iridoid, and quinic acid. The following phenolic acid-related compounds were chosen from Supplementary Table S5 and quantified: Protocathechuic acid (2), aesculin (11), FA (16), RA (18), caffeic acid (20), and in addition GA (64^*) and chlorogenic (5-O-caffeoylquinic) acid (65)^*. The quinic acid (75^*) was measured as a conjugate with caffeic acid forming the chlorogenic acid. The data highlighted RA (18) and FA (16) as the most abundant compounds followed by caffeic acid (20). Besides identifying flavones shown in Supplementary Table S5, such as luteolin (42); apigenin (47); and cirsimaritin (50), apigetrin (66^*) and hispidulin (67^*) were also measured. To confirm whether other flavonoid classes were present in N. nuda, more compounds from the class of flavanones were investigated: eriodictyol (68^*) and naringenin (69^*); and from the class of flavonols: Q (70^*), rutin (71^*), isoquercetin (72^*), astragalin (73^*), and isorhamnetin (74^*). The results indicated that in the used experimental conditions, the flavones were the predominant flavonoids represented mostly by apigenin (47), apigetrin (66^*), and cirsimaritin (50). The metabolite accumulation under in situ and in vitro conditions was hardly affected. The significant differences between the flowers and leaves were observed, as the metabolite accumulation in the flowers, particularly of the analyzed flavonoids, was generally enhanced. Most of the detected metabolites had highest levels in the flowers, while in the in vitro plants, a general reduction was observed. Strikingly, under in vitro conditions, FA (16) was the only compound showing an increased accumulation.

To explore the impact of light on metabolite accumulation, we cultivated in vitro plants under three different lights: White with normal intensity as a control (W), blue–red with high intensity (BR), and blue–red with low intensity (BRS). The compounds from Figure 5 were quantified and subjected to PCA analysis (Figure 6). The total contribution of both principal components (PC) in the variation of the data set was 83.1%: 43.5% for PC1 and 39.6% for PC2. The formation of PC1 was caused mainly by the differences in light spectrum (W vs.BR). The PC2 formation was closely related to the light intensity (BR vs. BRS). The metabolites were distributed according to the light variant, as the following components were significantly upregulated: The W-specific substances aesculin (11) and FA (16); the BR-specific RA (18), cirsimaritin (50), naringenin (69^*), and 1,5,9-eDLA (59); and the BRS-specific chlorogenic acid (65^*). There were no significant light-dependent changes in caffeic acid (20), luteolin (42), eriodictyol (68^*) and Q (70^*). Rutin (71^*) and isoquercetin (72^*) were significantly upregulated by BR in comparison to W, but the effect of BRS was rather variable. Astragalin (73^*) was upregulated by the blue–red light spectrum range independently from the light intensity. The metabolites protocatechuic acid (2), GA (64^*), apigenin (47), apigetrin (66^*), hispidulin (67^*), isorhamnetin (74^*), and quinic acid (74^*) were not detected in the in vitro N. nuda samples.

The principle of DNA barcoding is to create a shared source of DNA sequences that can be used to identify taxonomic features (Hollingsworth et al., 2011). The DNA barcoding markers used for the members of Lamiaceae include chloroplast DNA regions, such as rbcL (a conserved, coding region), matK (a rapidly evolving, coding region) and trnH-psbA (a rapidly evolving and length variable intergenic spacer) (Fazekas et al., 2008; Theodoridis et al., 2012). The plastome data has been successfully used to estimate family-wide relationships within Lamiaceae with sufficient resolution at subfamily and tribe levels (Li et al., 2016; Zhao et al., 2021). The previous studies highlight the relevance of nuclear non-coding ITS DNA sequences to infer evolutionary relationships at intrageneric taxonomic ranks of Nepeta species often correlating with floral and pollen morphological characters (Jamzad et al., 2003a; Al-Qurainy et al., 2014; Özcan, 2019; Alzeibr et al., 2020). Based on ITS markers, Jamzad et al. (2003a) have shown that Nepeta is defined as monophyletic consisting of two large clades with four subclades. In our study, we showed that N. nuda belongs to a subclade with the closely related species N. congesta, N. heliotropifolia, N. deflersiana and N. cataria, and with the closest phylogenetic distance to the recently assigned N. sheilae Hedge & R.A. King (Al-Qurainy et al., 2014; Alzeibr et al., 2020). While N. nuda belongs to the Orthonepeta section, the morphology of the phylogenetically close N. sheilae, endemic to Saudi Arabia, resembles with section Oxynepeta (Alzeibr et al., 2020). The other species from this clade also belong to the section Oxynepeta except for N. cataria. They not only have close genetic structure but also share common morphological features, such as corolla shape pattern (Jamzad et al., 2003a). According to Fazekas et al. (2008), three-region DNA barcodes from plastid genome allow sufficient genetic discrimination. The available information from the chloroplast markers allowed us to relate N. nuda to N. italica, N. cataria, and N. grandiflora subclade.

Nepeta italica L. is considered phylogenetically and chemotaxonomically close to N. nuda, as both species contain 1,8-cineole, which is a major constituent of the essential oil (Baser et al., 2000). Caffeic acid esters, predominantly RA (Janicsák et al., 1999; Pedersen, 2000), nepetoidins A and B (Grayer et al., 2003), and leaf surface flavonoids from the flavone group (cirsimaritin, apigenin, luteolin) (Jamzad et al., 2003b) represent important phenolic chemomarkers for the genus Nepeta and the whole subfamily Nepetoidae. In addition, chemotaxonomic phenolic profiling supported the close relation of N. nuda to N. cataria, N. grandiflora M. Bieb. “Dawn to Dusk,” and N. mussinii Spreng. ex Henckel, all belonging to the section Cataria (Mišić et al., 2015). Our results support the data reported by Mišić et al. (2015) and allow us to consider chloroplast markers as more reliable for accurate taxonomic assignment of representatives of the genus Nepeta, as compared to ribosomal markers. Furthermore, N. nuda displays the highest identity (the mean value of three markers was 99.3%) to N. italica. The accessions of the latter that are presented in the phylogenetic trees have Greek origin (Theodoridis et al., 2012). This strongly supports the presence of a distinct genetic pool of Nepeta species on the Balkan Peninsula but also implies its evolutionary divergence with the presence of species and/or ecotype variability influenced by eco–geographical and demographic factors. However, further work at larger population scale is required to validate this notion.

The production of iridoids and caffeic acid esters as dominant phenolics in Nepeta species was assumed to have evolved simultaneously under natural selection toward providing ecological advantages for efficient adaptation to abiotic and biotic factors (Mišić et al., 2015; Mint Evolutionary Genomics Consortium, 2018). Nepeta nuda plants are found mostly in open forest areas and meadows, in the mountains and at subalpine altitudes up to 2100 m (Baden, 1991). Kofidis and Bosabalidis (2008) demonstrated that increasing the altitude led to the development of N. nuda glandular trichomes and reduction of the respective essential oils. Besides, the epidermal cells of summer leaves contained phenols, while the autumn leaves did not have these compounds. The flavones (cirsimaritin, luteolin, apigenin, and thymusin) are known to function as UV screens for heat reduction, as antimicrobial agents or insect-feeding deterrents (Jamzad et al., 2003b; Hostetler et al., 2017). In our study, differences in the antioxidant parameters between N. nuda plants collected from Pirin (1,850 m), Rhodopes (1,750 m), and Rila (1,150 m) could be linked to different altitudes. Interestingly, the two samples from Rila were collected from the same place, but with a difference of 3 weeks, also showed an obvious change in antioxidant parameters. The phenolic compounds and iridoids play an important role in plant defense, as in N. nuda these metabolites act as antioxidants and agents with insect antifeeding, phytotoxic, antiproliferative, antiviral, and antibacterial activities (Table 5). Our data support the presence of these secondary metabolites in the wild-grown plants since phenolics were enriched in natural environment, where the exposure to abiotic and biotic factors considerably affected plant immunity. In general, environmental stimuli induce generation of reactive oxygen species (ROS) that can damage plant cells and their function. As antioxidants, the phenolics neutralize free radicals and prevent lipid peroxidation (Pacifico et al., 2015). Compared to the wild-grown plants, during in vitro cultivation plant growth and metabolic activity are reduced (Kapchina-Toteva et al., 2014; Yordanova et al., 2017). In the in vitro conditions, the herbicide paraquat is used as an elicitor causing oxidative stress and resulting in subsequent accumulation of N. nuda phenolic compounds including RA (Cvetković et al., 2015). ROS can also act in signaling cascades to transmit environmental and developmental signals (Xia et al., 2015). Flavonoids can also modulate the auxin-mediated signaling cascades (Mouradov and Spangenberg, 2014; Hostetler et al., 2017). Upon herbivore attack, wounding-generated ROS induce phytoimmune responses including release of jasmonic acid (JA) and synthesis of salicylic acid (SA) for systemic resistance (Roossinck, 2015). The JA induces production of antifeeding compounds for insects (e.g., phenolic acids and iridoids) and essential oil volatile constituents (monoterpenoids) with repellent effect (Table 5).

The depressive effects of the N. sibthorpii Bentham iridoid epinepetalactone provoke alterations in general behavioral pattern of several animal species, and the role of additional compounds is suggested (Galati et al., 2004). In the N. nuda extracts, high content of the iridoid glucoside 1,5,9-eDLA (59) in the flowers was detected, which could suggest that this compound is the major iridoid in this species and might be attractive for cats, as well. As an indirect support for this assumption is the inhibition of human alphaherpesvirus type 1 by the N. nuda extracts. This antiviral effect could be associated with the feline herpesvirus type 1 (FHV-1) that is a cat alphaherpesvirus causing acute ocular surface disease, dermatitis, respiratory disease, and potentially intraocular disease (Thomasy and Maggs, 2016). However, many antiviral drugs developed for treatment of human herpesviruses have been also used for cat treatment after infection with FHV-1. This suggests that N. nuda could produce metabolites with a cat healing power. The metagenomic studies have shown that viruses are abundant in the wild plants, but they are typically asymptomatic, as insects and other herbivores are common vectors in virus transmission (Roossinck, 2015). The existing relationships between plants and viruses have shaped their evolution, and strikingly, the viruses could be mutualists rather than pathogens. For instance, the plant viruses can have positive impacts on plants' ability to cope with biotic and abiotic stress factors. Notably, some viruses use plants as vectors and do not replicate in plants but they are transmitted horizontally through insect feeding on plants. Interestingly, virus-infected plants produce the hormone SA, which counteracts the JA response, and makes plants a better host (Roossinck, 2015). We identified 12-O-hexosyl-jasmonate (63) in N. nuda, which inhibits JA signaling (Miersch et al., 2008). However, the potential mechanism of N. nuda effects and herpes-related viruses requires further investigation. In addition to the possible antiviral properties of iridoids, our data pointed to N. nuda phenolic compounds accumulating mainly in flowers of the wild-grown plants.

Our data on phytochemical potential of N. nuda provide additional information about the phytotoxic activity of N. nuda extracts (Table 5). The inhibition of plant growth occurs upon competition for common resources, such as nutrients and light (Cheng and Cheng, 2015). Phenolic acids and coumarins act as direct allelopathic agents by inhibiting cell proliferation and the respective growth processes (Table 5). The high antiproliferative activity of N. nuda extracts is associated with growth inhibition of different plant species (wheat, cucumber, corn, radish, lettuce, cress, purslane; Table 5). Allelopathic effect could be also indirect and mediated by the inhibition of plant growth-promoting bacteria (PGPB). For instance, A. calcoaceticus annotated as PGPB (Kang et al., 2009; Ishizawa et al., 2020), could act through production of extracellular gibberellic acid that stimulates plant elongation as a shade avoidance response (Yang and Li, 2017). It could be assumed that the “sun-plant” N. nuda can inhibit such light-competing plant species. Besides, A. calcoaceticus can improve cucumber growth by assisting degradation of phenolic allelochemicals, and activating antioxidant enzymes (Wu et al., 2018). Other PGPB, such as K. pneumoniae promotes plant growth inducing systemic resistance (Liu et al., 2018), and the soil saprophyte B. cereus ensures fungal protection of tomato seedlings (Ramírez et al., 2022). The potentially lethal animal and human pathogen S. aureus is directly affected by the phenolic compound SA that reduces the pathogen virulence and attachment to the root surface (Prithiviraj et al., 2005). Our data support the notion that phenolic compounds in N. nuda extracts directly inhibited the studied bacterial strains by impeding their growth. Interestingly, in contrast to the antioxidant and antiviral activities, where the N. nuda extracts from wild-grown plants showed significantly stronger activities than the in vitro plants, the antibacterial activities were nearly identical and did not depend on the environmental conditions. Recent report by Aničić et al. (2021) highlights Nepeta iridoids (nepetalactones and 1,5,9-eDLA) and the phenolic acid RA in methanol extracts, as potent antimicrobial agents. In our study, the in vitro plants had an increased FA content. Besides the literature reports discussing the growth inhibition of K. pneumoniae by FA (Lo and Chung, 1999), we tested the pure compound and confirmed its activity against all the four bacterial strains (Supplementary Figure S3). Accordingly, methanol extracts from other in vitro cultivated Nepeta species also inhibit the growth of B. cereus and S. aureus (Nestorović et al., 2010).

The light composition and intensity are important factors in stimulating plant flowering (Zhiponova et al., 2020a). The observed metabolite enrichment in N. nuda flowers led to the use of the previously reported light formula for boosting of flowering (Zhiponova et al., 2020a) in our in vitro experiment. Consistently with the natural environmental adaptation of N. nuda, the high light intensity stimulated RA (18), cirsimaritin (50), naringenin (69^*), rutin (71^*), isoquercetin (72^*), as well as 1,5,9-eDLA (59). Lower light intensity significantly upregulated the chlorogenic acid (65^*), which supports the existence of a positive correlation between the increased production of this metabolite and the low intensity and combination of blue, red and far–red lights (Chen et al., 2016). Aesculin (11) and FA (16) were specific for the W control variant, which could reflect lack of importance for flowering. Indeed, respective decrease (for aesculin) and slightly higher levels (for FA) of these compounds were detected in the flowers, as compared to the leaves.

This study characterized the impact of environmental factors in modulating the content of bioactive compounds in N. nuda that is largely distributed in Bulgaria. DNA barcoding enabled us to genetically discriminate and determine the precise phylogenetic position of N. nuda using the available Nepeta sequence records in public databases. We provided the first DNA barcode records for N. nuda in the BOLD database, thus contributing to the enrichment of global catalogs and specific genetic diversity for the genus Nepeta in Bulgaria. These data can be a valuable basis for further species identification and support future studies regarding Nepeta genetic variability. Organ specificity of phytoimmunity features of N. nuda, such as the antioxidant and antiviral activities, phenolic acids, flavonoids and iridoid glycosides, were analyzed and discussed from an ecophysiological perspective, and their dependence on environmental conditions. Furthermore, this study showed that the light spectrum and intensity are crucial factors affecting the differential accumulation of phenolic acids, flavonoids and iridoids in N. nuda. Therefore, these studies could strengthen and facilitate the understanding of N. nuda ecology and targeted modulation of its productivity under controlled conditions, thus entailing their potential benefits for agriculture and pharmacology.

The current work provides the platform for further detailed studies on Nepeta species in multiple research directions. In point of view of genetic diversity, the study of ecological and taxonomic dynamics within the genus Nepeta is of particular interest especially when correlating a genetic ecotype and metabolite activities. It would be of interest to define more aspects of the molecular mechanism regulating phenolics and iridoids levels—effect of environmental signals, gene expression, parallel profiles of other N. nuda metabolites. Regarding the biological activities, it would be of interest to identify specific related metabolites with importance for food quality and human health. The application of biotechnological approaches would assist targeted accumulation of compounds by elicitors, as well as enhanced metabolic production by cell cultures and bioreactor.