A phenylpropanoid bearing cyclobutane ring, compound 1, was isolated as a yellow oil from the ethyl acetate fraction. Thin-layer chromatography (TLC) analysis revealed an Rf value of 0.3 in a solvent system of n-hexane:ethyl acetate (1:1), with absorption observed under UV Light at 254 nm. Upon heating post-derivatization with 10% H2SO4, the compound exhibited an orange coloration under 366 nm UV light. High-resolution ESI-Orbitrap-MS in positive ion mode displayed a molecular ion peak at m/z 401.15958 [M + H]+, consistent with a molecular formula of C22H24O7.

Infrared (IR) spectroscopy indicated the presence of a hydroxyl group through a broad absorption band at 3421 cm⁻1 and a carbonyl group at 1714 cm⁻1. The 1H NMR spectrum (Table 1) of compound 1 showed signals attributable to four aromatic protons [δH 7.19 (2H, J = 8.5 Hz), 6.90 (2H, J = 8.5 Hz)], consistent with a 1,4-disubstituted benzene ring; a pair of trans-olefinic protons [δH 7.38, 6.26, each J = 16 Hz]; a single olefinic proton (δH 6.41); two oxymethine protons (δH 5.15, 4.46); two methoxy groups (δH 3.81, 3.40); an oxygenated ethyl moiety (δH 4.26, q, 2H; 1.32, t, 3H); and several other resonances corresponding to methine protons.

The 13C NMR spectrum (Table 1) exhibited 20 carbon signals (with two overlapping), including two ester carbonyls (δC 175.1, 166.2), eight sp2 carbons (including overlapping aromatic carbons), and five oxygenated carbons [δC 77.9, 63.2, 60.9; and methoxy carbons at δC 55.4, 51.9]. The HMQC experiment confirmed direct proton-carbon correlations, as listed in Table 1.

The COSY spectrum demonstrated proton coupling within the aromatic system [H-1/H-5 (δH 6.90), H-2/H-4 (δH 7.19)], between oxymethine protons H-12 (δH 5.15) and H-15 (δH 6.41), between the ethyl protons H-20 (δH 4.26) and H-21 (δH 1.32), as well as between the trans-olefinic protons H-17 (δH 7.38) and H-18 (δH 6.26). Evidence for the presence of a cyclobutane ring was indicated by the deshielded resonances and correlations involving H-8 (δH 2.67), H-10 (δH 3.29), and H-9 (δH 3.22), with further support from HMBC correlations between H-9/C-11 and H-10/C-11 (Fig. 1A). Further HMBC correlations supported the structural connectivity between the cyclobutane ring, a lactone, and a cyclohexane moiety. Specifically, correlations were observed between H-8/C-16, H-9/C-12, H-12/C-16, H-12/C-11, H-12/C-14, H-13/C-8, and H-10/C-12. The first methoxy group [δH 3.81, δC 55.4] was attached to the aromatic ring, evidenced by HMBC correlations between H-7 and C-6. The second methoxy group [δH 3.40, δC 51.9] was linked to the cyclobutane ring, as shown by the HMBC correlation between H-22 and C-11.

Compound 1. A 1H–1H COSY and HMBC correlations; B 1H–1H NOESY correlation; C Comparison of compounds 1 and 1a

The planar structure of compound 1 was elucidated by extensive 2D NMR analysis (Fig. 1A). HMBC correlations between H-8/C-3, H-8/C-4, and H-2/4/C-8 further confirmed the linkage of the cyclobutane to the methoxyphenyl moiety. The ethyl acrylate unit was found to be appended to the cyclohexane ring at its olefinic side, supported by HMBC correlations involving H-18/C-14, H-17/C-19, H-18/C-19, H-17/C-15, and H-20/C-19.

A structurally similar compound, 1a, was previously reported by A. Schrader et al. as a photochemical product [16]. The primary structural divergence between 1 and 1a lies in the nature of their alkyl ester substituents (Fig. 1B). NOESY spectra provided additional stereochemical insight, revealing cross-peaks between H-7 and H-1/5, H-12 and H-22, H-13 and H-18, and H-15 and H-17. However, due to overlapping resonances within the cyclobutane region, a 1D-difference NOE experiment was performed to supplement the configuration assignment.

Upon selective irradiation of H-8 (δH 2.67), enhancements were observed for H-9 (δH 3.22), H-13 (δH 4.46), and H-2/4 (δH 7.19). Comparative 1D-difference NOE data between 1 and 1a are summarized in Table 2. Notably, compound 1 showed reciprocal NOE enhancements between H-8 and H-9, although at low intensities, which may be attributed to differences in NMR instrumentation-compound 1 being measured on a 600 MHz spectrometer, while 1a was analyzed on a 300 MHz system. Consequently, minor signal enhancements may have arisen from interactions with neighboring protons. Overall, the relative stereochemistry of compound 1 mirrors that of compound 1a, as depicted in the complete structural representation (Fig. 1B) and further supported by the three-dimensional conformational model (Fig. 1C).

Several known compounds were also isolated and identified through spectroscopic comparisons with Literature data, including 3-caren-5-one-2-ol (2) [17], ethyl-syn-2,3-dihydroxy-3-(4-methoxyphenyl) propanoate (3) [18], ethyl-anti-2,3-dihydroxy-3-(4-methoxyphenyl) propanoate (4) [18], ethyl-4-methoxycinnamate (5) [19], 4-methoxybenzoic acid (6) [20].

The photochemical generation of 1a [16] provided a conceptual framework for proposing a plausible biosynthetic pathway for compound 1, a new natural product derived from K. galanga. Given that ethyl 4-methoxycinnamate is a major constituent of K. galanga, it is postulated as the precursor (Fig. 2). Exposure to natural sunlight during plant growth or post-harvest drying processes may facilitate [2 + 2] cycloaddition [21] between the trans double bond of compound 5 and the aromatic ring, followed by epoxidation and subsequent ring opening, ultimately forming the lactone through double bond migration and carboxylate conjugation. Based on its structural features and proposed biosynthesis, compound 1 can be classified as a phenylpropanoid dimer rather than a lignan, since its formation involves photochemical dimerization rather than oxidative coupling of monolignols [22].

Proposed biosynthetic pathway of 1

Inflammation is a vital physiological defense mechanism involving intricate networks of cellular and molecular mediators; however, its dysregulation is implicated in the onset and progression of numerous chronic diseases, highlighting the need for safe and effective anti-inflammatory therapies [23]. Among these mediators, nitric oxide (NO) plays a dual role—functioning as an anti-inflammatory agent under normal physiological conditions, but contributing to pathological inflammation when produced in excess [24]. K. galanga demonstrates potent analgesic and anti-inflammatory effects, mediated by its bioactive constituents that suppress inflammatory signaling pathways [25]. Previous studies demonstrated that ethyl 4-methoxycinnamate (5), the principal compound in K. galanga, exhibited dual COX-1 and COX-2 inhibitory activity at 1.12 µM and 0.83 µM, respectively [19]. Moreover, another reported its inhibitory activity on IL-1 and TNF-α with IC50 values of 166.4 µg/mL and 96.84 µg/mL, respectively [26]. In alignment with these reports, compound 5 in this study inhibited NO production with an IC50 of 12.2 ± 4.0 µM (Table 3). Compounds 3–4 and 5 showed notably different anti-inflammatory activities, with compound 5 being the most potent. This enhanced activity is likely due to its conjugated double bond, which may improve interaction with inflammatory targets such as iNOS. Structural features like unsaturation appear to significantly influence anti-inflammatory potency, as supported by this result. Intriguingly, the newly identified phenylpropanoid bearing cyclobutane ring, compound 1, also exhibited significant NO inhibitory activity with an IC50 of 23.1 ± 6.4 µM, suggesting promising anti-inflammatory potential (Fig. 3).

Effect of isolated compounds on NO secretion in RAW264.7 cells. Concentration of sample is 40 μM except for L–NMMA (100 μM). All the results are presented as the mean ± SD (n = 3). Statistical analysis was conducted using one way ANOVA (Tukey’s test). Significance difference (####p < 0.0001) compared with vehicle control group, whereas (**** p < 0.0001, ** p < 0.01, * p < 0.05) compared with LPS treatment group