Introduction

The cell membrane binding affinity and/or permeability of compounds are often critical factors in drug development [1]. If the compounds cannot bind to the cell membrane and/or penetrate the cell membrane, then they may not be able to exhibit biological activity, even if they contain a structure of interest in drug development [2]. This would be the same problem for compounds with low cell membrane binding affinity and/or permeability, as they would be judged to have weak activity. Therefore, many potentially bioactive compounds may have been overlooked due to their inability to bind and/or penetrate the target cells, even though their structures inherently possess biological activities. This means that such compounds could be promising drug seeds, if their cell membrane binding affinity and/or permeability are improved.

Olivetolic acid (1b) (Fig. 1) is a common intermediate produced by olivetolic cyclase (OAC) in the biosynthesis of ∆9-tetrahydrocannabinol (∆9-THC) and cannabidiol (CBD) [3]. To create a platform for the development of therapeutically beneficial cannabinoids, we previously generated 1b-analogs with n-heptyl- (1c), n-nonyl- (1d), or n-undecyl- (1e) moieties, by exploiting engineered OAC and tetraketide synthase (TKS) (Fig. 1) [4]. In nature, olivetolic acid is prenylated to form cannabigerolic acid (CBGA) (2b), which is further converted to ∆9-tetrahydrocannabinolic acid (∆9-THCA) and cannabidiolic acid (CBDA) by cyclizing its prenyl moiety (Fig. 1). The pharmaceutically important cannabigerol (3b) (Fig. 1), ∆9-THC, and CBD are heat-decarboxylated derivatives of 2b, ∆9-THCA, and CBDA, respectively [5].

Fig. 1
figure 1

Structures of compounds 1a–f, 2a–f, 3a–f, and 4a–f

Full size image

A recent study revealed that despite its low brain penetration, the biosynthetic precursor molecule 1b exhibited a moderate anticonvulsant effect in a mouse model of Dravet syndrome, with comparable potency to the known anticonvulsant cannabinoid, CBD [6]. To the best of our knowledge, no other activity has been reported for 1b. In contrast, 3b reportedly showed weak binding affinity to human cannabinoid receptor type 1 (CB1) and type 2 (CB2) as compared with ∆9-THC, but displayed relatively potent activities toward several ligand-gated cation channels of the transient receptor potential (TRP) superfamily, and bound to the 5-hydroxytryptamine receptor subtype 1A (5-HT1A) for serotonin stimulation [7,8,9]. Furthermore, 3b is reportedly among the potent cannabinoids possessing strong antibacterial activities against some clinically relevant, drug-resistant Staphylococcus aureus strains [10]. Recent study suggested that these antibacterial activities are derived from the action on the cytoplasmic membrane [11]. The antibacterial effect of 3b against Bacillus subtilis has also been reported [12]. Although its activity is moderate, olivetol (4b) (Fig. 1), a decarboxylated derivative of 1b, reportedly has antibacterial activities against these S. aureus strains [10]. Importantly, due to the lack of undesirable psychoactive effects, such as those from ∆9-THC, compound 3b is currently attracting keen attention as more appropriate candidate for novel drug development [13,14,15,16,17]. Meanwhile, 2b could also be a promising cannabinoid with similar antibacterial activities against the S. aureus strains [10].

In an attempt to search for small therapeutic agents for the treatment of coronavirus disease 2019 (COVID-19), a recent study demonstrated that 2b could block the live severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) and the emerging variants, B.1.1.7 and B.1.351 [18]. It should be noted here that most results obtained with 2b and 3b were obtained from assay systems involving their cell membrane binding affinity and/or permeability. It is also relevant that both 2b and 3b are 1b derivatives with a highly hydrophobic geranyl moiety. Prenyl modifications of compounds reportedly enhanced the biological activities of the mother molecules by increasing their binding affinity to the cell membrane [19, 20]. As of now, the importance of the prenyl chain for the pharmacological actions of 2b and 3b has not yet been fully elucidated. However, in light of the role of the prenylation, the activities observed in 2b and 3b can be attributed primarily to the structure excluding the geranyl moiety. Indeed, despite having different side chains (geranyl for 2b and cyclohexene terpene ring for CBDA), the equal effectiveness of CBDA and 2b against SARS-CoV-2 was also observed in a previous study [18]. These findings suggested that the geranyl side-chain in 2b and the cyclohexane ring in CBDA mainly function as enhancers to increase their hydrophobicity, and thereby, both compounds could show equal effectiveness against SARS-CoV-2. As mentioned above, previous studies successfully generated the 1b-analogs with longer alkyl-chains, 1c–e [4]. Due to the hydrophobicity gained with longer alkyl-chains, these 1b-analogs may also be potential active compounds like 2b, even though they lack the prenyl moiety. Herein, we report the antibacterial activities of 1b and its analogs 1c–e, and those with n-propyl- (1a) and n-tridecyl- moieties (Fig. 1), against Gram-positive bacteria, S. aureus and B. subtilis, as a preliminary evaluation of their biological activities. Remarkably, 1c–f newly gained the ability to inhibit the growth of S. aureus and B. subtilis in a chain-length-dependent manner. Furthermore, comparisons of the antibacterial activities of 1b, compounds 2b and 3b, olivetol (4b), and their respective analogs with n-propyl- (2a, 3a, 4a), n-heptyl- (2c, 3c, 4c), n-nonyl- (2d, 3d, 4d), n-undecyl- (2e, 3e, 4e), and n-tridecyl- (2f, 3f, 4f) moieties (Fig. 1), against the two tested positive bacteria indicated that 1e and 1f showed strong activities against B. subtilis, which were higher than that of 3b. Interestingly, the antibacterial assay also revealed that the n-heptyl modification at C-6 of 2b effectively enhanced its activity against B. subtilis.

Materials and methods

Chemicals and reagents

All chemical reagents, including 1b, were purchased from Fujifilm Wako Pure Chemical Corporation unless otherwise specified. Compound 4b was purchased from Sigma-Aldrich.

General experimental procedures

HRMS (ESI) data were obtained in the positive and negative modes on a SHIMADZU LC–MS-IT-TOF spectrometer. NMR spectra were recorded on ECA500II and ECX400P spectrometers (JEOL). Compounds 1a–f were dissolved in dimethyl sulfoxide-d6 (DMSO-d6) and 2a–f, 3a–f, and 4a–f in chloroform-d (CDCl3), and their spectra are expressed in ppm) based on the residuals of these solvents at 2.49 and 7.26 for 1H NMR and 39.5 and 77.0 for 13C NMR.

NphB G286S/Y288A mutant expression plasmid construction and protein purification

A modified pQE80L vector (Qiagen) for expression as an N-terminal glutathione S-transferase (GST) fusion protein with a PreScission Protease (Amersham Biosciences) cleavage site (LEVLFQGP) was linearized by PCR, using KOD-Plus-Neo DNA Polymerase (Toyobo) and the pair of primers: 5′-GTCGACCTGCAGCCAAGCTTAATTAGCTGC-3′ as the sense primer, and 5′-TCCCGGGCCCTGGAACAGAACTTCCAGAT-3′ as the antisense primer (the 15 bp sequences overlapping with those of primers used for the amplification of the NphB G286S/Y288A DNA fragment are underlined). The full-length cDNA encoding the NphB G286S/Y288A mutant was synthesized by Integrated DNA Technologies, and served as a template to amplify the cDNA fragment of the mutant by PCR, using TTCCAGGGCCCGGGATGAGCGAAGCGGCGGATGTGGAACGTGTG-3′ as the sense primer, and 5′-TGGCTGCAGGTCGACTTAATCTTCCAGCGAATCAAACGCTTTCAG-3′ as the antisense primer. The amplified mutant cDNA fragments were connected by an In-Fusion HD cloning kit (Clontech), according to the manufacturer’s protocol. After confirmation of the sequences, the resultant expression plasmid (pQE-GSTNphB G286S/Y288A) was transformed into Escherichia coli M15 (pREP4). The cells harboring the plasmid were cultured to an OD600 of 0.6 in LB medium, containing ampicillin (100 μg/mL) and kanamycin (25 mg/mL) at 25 °C. IPTG was then added to a final concentration of 0.5 mM to induce gene expression, and the culture was incubated further at 16 °C for 16 h.

All of the following procedures were performed at 4 °C. In brief, the E. coli cells were harvested by centrifugation at 5,000 g and resuspended in 50 mM Tris–HCl buffer (pH 8.0), containing 0.2 M NaCl, 10% (v/v) glycerol, and 2 mM DTT (buffer A). The cells were disrupted by sonication, and the lysate was centrifuged at 12,000 g for 30 min. The supernatant was loaded onto a Glutathione Sepharose 4B affinity column (GE Healthcare) equilibrated with buffer A, and the column was then washed with buffer A. The GST-tag was cleaved on the column overnight by PreScission Protease, and then the recombinant NphB G286S/Y288A was eluted with buffer A. The resultant NphB G286S/Y288A, thus, contains three additional residues (Gly-Pro-Gly) at the N-terminal flanking region, derived from the PreScission Protease recognition sequence. After fivefold dilution of the protein solution with 20 mM Tris–HCl buffer (pH 8.0), containing 10% (v/v) glycerol, 25 mM NaCl, and 2 mM DTT, the protein solution containing the NphB G286S/Y288A mutant was further purified to homogeneity by chromatography on a HiLoad 16/60 Superdex 200 column (GE Healthcare). The peak fractions were concentrated to 10 mg/mL in 20 mM Tris–HCl buffer (pH 8.0) buffer, containing 25 mM NaCl and 2 mM DTT.

Synthesis of 1a–f, 2a–f, 3a–f, and 4a–f

Compounds 1a–f and 4a–f were synthesized as previously reported [4, 21, 22]. Compounds 3a–f were synthesized according to the published methods [23]. To synthesize 2a–f, 2 mM 1a–f were incubated in an enzyme reaction mixture consisting of 50 mM HEPES (pH 7.5), with 5 mM MgCl2, 2 mM geranyl pyrophosphate (GPP), and 0.5 mg/mL purified NphB G286S/Y288A in a final volume of 20 mL, according to the previously reported method with some modifications [24]. The reaction mixtures were incubated at 25 °C and extracted with ethyl acetate. The organic solvent was evaporated, and 2a–f were purified using an Agilent Infinity II 1260 HPLC system coupled with a TSK-gel ODS-80Ts column (4.6 × 150 mm, TOSOH) (Flow rate: 0.6 mL/min; mobile phase, water and acetonitrile, both containing 0.1% trifluoroacetic acid; 0‒1 min: 50%, 1‒30 min: 50–100% acetonitrile; UV: 260 nm or 280 nm). The purities of all tested compounds used for the antibacterial assay were estimated based on their peak areas on the HPLC chromatograms, using an Agilent Infinity II 1260 HPLC system coupled with a TSK-gel ODS-80Ts column (4.6 × 150 mm, TOSOH) (Flow rate, 0.6 mL/min; mobile phase, water and acetonitrile, both containing 0.1% trifluoroacetic acid; 0‒1 min: 50%, 1‒30 min: 50–100% acetonitrile; UV: 260 nm or 280 nm).

Varinolic acid (1a)

White solid (10.4 mg, Yield 8.7%); 1H NMR (500 MHz, DMSO-d6) data, see Table S1; 13C NMR (125 MHz, DMSO-d6) data, see Table S2; HRMS (ESI) m/z: [M–H]− Calcd. for C10H11O4 195.0663; Found 195.0666.

2,4-Dihydroxy-6-tridecylbenzoic acid (1f)

White solid (20.4 mg, Yield 14%); 1H NMR (500 MHz, DMSO-d6) data, see Table S1; 13C NMR (125 MHz, DMSO-d6) data, see Table S2; HRMS (ESI) m/z: [M–H]− Calcd. for C20H31O4 335.2228; Found 335.2255.

Cannabigerovarinic acid (2a)

White solid (2.2 mg, Yield 13%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C20H27O4 331.1915; Found 331.1910.

Cannabigerolic acid (2b)

White solid (2.0 mg, Yield 41%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C22H31O4 359.2228; Found 359.2229.

Cannabigerophorolic acid (2c)

White solid (7.3 mg, Yield 35%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C24H35O4 387.2541; Found 387.2549.

(E)-3-(3,7-Dimethylocta-2,6-dien-1-yl)-2,4-dihydroxy-6-nonylbenzoic acid (2d)

White solid (4.6 mg, Yield 31%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C26H39O4 415.2854; Found 415.2867.

(E)-3-(3,7-Dimethylocta-2,6-dien-1-yl)-2,4-dihydroxy-6-undecylbenzoic acid (2e)

White solid (2 mg, Yield 11%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C28H43O2 443.3167; Found 443.3179.

(E)-3-(3,7-Dimethylocta-2,6-dien-1-yl)-2,4-dihydroxy-6-tridecylbenzoic acid (2f)

White solid (1.6 mg, Yield 9%); 1H NMR (400 MHz, CDCl3) data, see Table S3; HRMS (ESI) m/z: [M–H]− Calcd. for C30H47O4 471.3480; Found 471.3467.

Cannabigerovarin (3a)

Yellow oil (2.1 mg, Yield 2.1%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C19H27O2 287.2017; Found 287.2017.

Cannabigerol (3b)

Yellow oil (32.1 mg, Yield 12%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C21H31O2 315.2330; Found 315.2302.

Cannabigerophorol (3c)

Yellow oil (2.4 mg, Yield 2%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C23H35O2 343.2643; Found 343.2649.

(E)-2-(3,7-Dimethylocta-2,6-dien-1-yl)-5-nonylbenzene-1,3-diol (3d)

Yellow solid (30.5 mg, Yield 3%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C25H39O2 371.2956; Found 371.2942.

(E)-2-(3,7-Dimethylocta-2,6-dien-1-yl)-5-undecylbenzene-1,3-diol (3e)

Yellow solid (16.8 mg, Yield 3%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C27H43O2 399.3269; Found 399.3266.

(E)-2-(3,7-Dimethylocta-2,6-dien-1-yl)-5-tridecylbenzene-1,3-diol (3f)

Yellow solid (28.3 mg, Yield 5.7%); 1H NMR (400 MHz, CDCl3) data, see Table S4; HRMS (ESI) m/z: [M–H]− Calcd. for C29H47O2 427.3582; Found 427.3583.

5-Propylbenzene-1,3-diol (4a)

White solid (126.5 mg, Yield 88.3%); 1H NMR (500 MHz, CDCl3) data, see Table S1; 13C NMR (125 MHz, CDCl3) data, see Table S2; HRMS (ESI) m/z: [M–H]− Calcd. for C13H19O2 151.0765; Found 151.0785.

Grevillol (4f)

White solid (176,1 mg, Yield 38%); 1H NMR (500 MHz, CDCl3) data, see Table S1; 13C NMR (125 MHz, CDCl3) data, see Table S2; HRMS (ESI) m/z: [M–H]− Calcd. for C19H31O2 291.2330; Found 291.2318.

Antibacterial assays

The antibacterial activities were evaluated using an MTT assay according to the previously published protocol, against Gram-positive bacteria (S. aureus NBRC 100910 and B. subtilis NBRC 13719), with slight modifications [25]. Bacterial strains were inoculated on YP agar plates [1% polypeptone (Nihon Pharmaceutical, Tokyo, Japan), 0.2% yeast extract (Difco Laboratories, Franklin Lakes, NJ, USA), 0.1% MgSO4⋅7H2O, and 1.5% agar (Fujifilm Wako Pure Chemical Corporation)] and incubated at 37 °C for 12 h. Solutions of compounds were prepared at 10 mM in DMSO and diluted to sixteen sample concentrations (0.078125, 0.15625, 0.3125, 0.625, 1.25, 1.5625 2.5, 3.125, 5, 6.25, 10, 12.5, 25, 50, 100, and 200 μM) in 96-well plates containing the microbial strains incubated in YP medium. A DMSO control was also included. The plates were incubated further at 37 °C for 12 h. Ampicillin (Nacalai Tesque, Kyoto, Japan) and kanamycin (Nacalai) were used as reference reagents. Finally, 10 µL of MTT solution (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (0.5 mg/mL in iPrOH- HCl) was added to each well and incubated further at 37 °C for 1 h. All MIC values were determined in triplicate. The MIC was defined as the lowest concentration at which no growth was observed.

Results and discussion

The present study was designed to gain insights into the prospective therapeutic properties of 1b and its analogs. Only one phytochemical investigation of 1b was found in the literature [6]. In contrast, the antibacterial activities of 2b and 3b against S. aureus have been well studied [10, 26]. Therefore, we first sought to identify the potential activities of 1a–f as antibacterial agents against S. aureus in the preliminary investigation. To this end, we also assessed the antibacterial activities of 1a–f against B. subtilis, since 3b reportedly exhibited antibacterial activity against this Gram-positive bacterium [10]. The antibacterial assay was performed with the MTT method, using ampicillin as a positive control. In our previous report, we produced 1b–e using our engineered OAC and TKS [4]. Here, we obtained 1b–e by chemical synthesis according to previously reported method [21], as shown in Fig. 2a, to obtain the desired amounts in a more cost-effective way for antibacterial activity evaluations. To further evaluate the effects of chain lengths on the antibacterial activity, we also synthesized 1a and 1f using the same method as for 1b–e (Fig. 2a).

Fig. 2
figure 2

Synthesis of a 1a–f and 2a–f and b 3a–f and 4a–f

Full size image

As shown in Table 1, the assay revealed that neither 1a nor 1b showed antibacterial activities against both tested bacteria, even at 200 μM concentrations. However, 1c possessed antibacterial activities against both bacteria, and these activities were further increased in a chain-length-dependent manner. Remarkably, both 1e and 1f had robust activities with MIC values of 6.25 μM and 2.5 μM against S. aureus and B. subtilis, respectively. These observations suggested that the elongation of the alkyl side-chain was effective to confer the antibacterial activity to the previously inactive 1b, especially against B. subtilis.

Table 1 Antibacterial activities of 1a–f, 2a–f, 3a–f, and 4a–f
Full size table

It is remarkable that the non-prenylated 1e and 1f possessed strong activities similar to those reported for 2b and 3b. However, a previous study revealed that the structurally simpler 4b, as compared to 1b, exhibited moderate antibacterial activity against S. aureus, despite having the same pentyl side-chain: 4b exhibited MIC values ranging from 64 to 128 μg/mL against the S. aureus strains in the previous study [10], while 1b showed no activity in this study. The differences suggested that the carboxylate moiety may be an unfavorable functional group for the antibacterial property of 1. To clarify this, and in the absence of the literature describing the alkyl-chain-length dependency of 4b, we synthesized a series of 4b-analogs according to the previously reported methods (Fig. 2b) and evaluated their antibacterial activities using the same method as described above.

In contrast to the previous report [10], 4b did not show any activity against S. aureus in our assay (Table 1). The use of different strains in our study and the previously reported study may be the cause of this different result. Furthermore, the assay revealed that the 4b-analogs with alkyl moieties ranging from n-propyl to n-undecyl showed the same chain-length-dependent patterns of antibacterial activity against S. aureus as those of the 1-series. However, 4f showed significantly decreased antibacterial activity against S. aureus, with a MIC value of 200 μM, as compared with that of 4e. Thus, the 4b-analog with the n-undecyl moiety was the most potent against S. aureus, with a MIC value of 6.25 μM in the tested 4-series, and its potency was consistent with those of 1e and 1f. A similar chain-length-dependent pattern was also found in the antibacterial activities of the 4b-analogs against B. subtilis. Among the tested 4b-series, compound 4e with the n-undecyl moiety exhibited the highest activity against B. subtilis with a MIC value of 3.125 μM, but its activity was slightly weaker than those of 1e and 1f. Despite having the same alkyl side-chain length as 1f, compound 4f showed lower activity, which might result from the loss of the carboxylic group in its structure. The lower activities of the 4b-series than the 1b-series against B. subtilis suggested that the carboxylate moiety could be an important functionality for enhancing the antibacterial activity against this bacterium. It is also noteworthy that the 4b-series achieved the optimum length at the n-undecyl moiety. A similar case was also found in previous structure–activity relationship studies, in which the n-octyl moiety was the best length for the activities of ∆8-THC and ∆9-THC at C-3, rather than longer ones [27], suggesting that the alkyl side-chain length is also one of the factors regulating the activities.

In this context, it is particularly interesting to determine whether the length of the alkyl side-chain influences the bioactivities of 2b and 3b, since only the antibacterial activities of 2b and 3b with an n-pentyl moiety have been studied. Compound 3b and its analogs were thus synthesized from the respective 4 compounds as the starting material (Fig. 2b). Presumably, the 2-series can be synthesized from the 1-series using this prenyl moiety-direct incorporation method, after their carboxyl groups were protected. However, the undesired 5-geranylated and 3,5-digeranylated 1-series will be also synthesized even if we use the methyl ester of the 1-series, resulting in the low yield of the 2-series. Similar cases were also observed in our chemical synthesis of 3b and 3d, where we also detected undesired 6-geranylated 3b and 6-geranylated and 2,6-digeranylated 3d analogs as by-products, although we did not check all products in the prenylation reaction of the 4-series. In previous studies, the other chemical methods of the prenylation of the resorcylic acids or the carboxylation of the prenylated resorcinols have been also reported [28,29,30]. However, these methods are also performed by the protection of the carboxylate or both carboxyl and hydroxyl groups on the structure to yield the geranylated resorcylic acid, leading to the necessity of the multistep reaction. On the other hand, a recent study reported that the NphB G286S/Y288N mutant can catalyze regiospecific geranylations of 1a and 1b to yield 2a and 2b, respectively, with high efficiency [24]. Therefore, we exploited this mutant to directly incorporate the geranyl moiety into 1, to synthesize 2b and its analogs (Fig. 2a). Our study revealed that the mutant enzyme can also produce 2c–f from 1c–f, respectively. As a side note, the regiospecific geranylation of the 4-series might not be possible, even with the use of this mutant enzyme for the 3-series synthesis: our preliminary NphB mutant reaction with 4b–e followed by LC–MS analyses revealed that the enzyme produced not only the desired 3b–e, but also undesired mono- and di-geranylated 3b–e analogs.

The antibacterial assays of 3b and its other alkyl-chain length analogs showed that the highest potencies were achieved with the n-pentyl and n-heptyl moieties with MIC values of 2.5 μM, and the n-pentyl moiety with a MIC value of 3.125 μM, against S. aureus and B. subtilis, respectively (Table 1). In contrast, compounds 3d and 3e exhibited decreased antibacterial properties with MIC values of 6.25 μM and 25 μM, respectively, and the activity against S. aureus was completely lost in 3f with the n-tridecyl moiety. In the case of B. subtilis, the activity loss started from 3d. However, 3a with the n-propyl moiety exhibited antibacterial properties against S. aureus, which are different from the antibacterial properties of 1a and 4a. A previous study showed that the cannabinoids, ∆9-THCV and CBDV, bearing n-propyl lipophilic side-chains at C-3 instead of n-pentyl moieties, were less active against methicillin-resistant S. aureus (MRSA) as compared to ∆9-THC and CBD, respectively [11]. Overall, the optimum chain length on the 3b-skeleton for the antibacterial activities was thus shifted toward the n-pentyl and n-heptyl moieties against S. aureus and the n-pentyl moiety against B. subtilis, respectively, in sharp contrast to those of the 1- and 4-series. Remarkably, the antibacterial activities against S. aureus showed that 3b and 3c were more potent than the 1- and 4-series. However, for 3b and 3c, the antibacterial activities against B. subtilis were slightly lower than those against S. aureus. This observation is different from the 1- and 4-series, in which 1e and 1f with long n-undecyl and n-tridecyl moieties, and 4e with an n-undecyl moiety, showed more potent activities.

The 2-series also exhibited the shifted antibacterial optimum active points on 2b and 2c against S. aureus and on 2c against B. subtilis, which were quite similar to those of 3b (Table 1). As in the 3-series, compound 2f showed negative effects on the activities against both tested Gram-positive bacteria. The negative effects of increasing the chain length of 2 were observed from the n-nonyl moiety against both Gram-positive bacteria, in a manner similar to those of 3. It can be postulated that the excessively elongated alkyl-chains could result in lower binding affinity to the target molecule due to the steric hindrance, which could lead to a decrease or loss in the antibacterial activities of 4 derivatives, such as 2f, 3f, and 4f. Finally, comparisons of the activities of all tested compounds demonstrated that 3b/3c and 2b/2c are the most and second-most potent cannabinoids as antibacterial agents against S. aureus, while 2c and 1e/1f are the most and second-most potent compounds as antibacterial agents against B. subtilis. These findings suggest that further improvements of the hydrophobic balance of the C-6 alkyl side-chain in combination with the C-3 position could be one of the strategies to enhance the bioactivities of these 1b derivatives.

In conclusion, the antibacterial activities described here have corroborated that the incorporation of a longer alkyl chain at the C-6 position in the resorcylic acid structure confers antibacterial properties against S. aureus and B. subtilis to the resultant 1b derivatives, even if the hydrophobic geranyl moiety was absent from the C-3 position. In particular, our results suggested that the longer n-undecyl and n-tridecyl alkyl-chains were most effective for enhancing the activities of 1b-type compounds against B. subtilis. The study also demonstrated that the n-heptyl modification at C-6 of 2b effectively enhanced the activity against B. subtilis. However, simultaneously, the study also indicated that further elongation of the alkyl moieties decreased the activities of the 2b-series. Similar cases were also observed in the antibacterial assays for the 3- and 4-series. As mentioned above, 3b is thought to achieve its antibacterial activity by targeting the cytoplasmic membrane of S. aureus. Although further investigation of the role of the elongated alkyl moiety for the antibacterial activities should be conducted, these observations may suggest the importance of the balance of the binding affinities of the compounds between the cytoplasmic membrane and the unidentified target molecules for the activity. Thus, further evaluations of the antibacterial activities, as well as other various biological activities of 1b and 2b derivatives, especially with optimized hydrophobicities at both the alkyl and prenyl side-chain positions of the core skeleton, would be beneficial for the discovery of novel drug seeds.