Biological synthesis and anti-inflammatory activity of arylalkylamine

Hydroxycinnamic acid amides (HCAAs) are natural compounds with antifungal, anticancer, and anti-inflammatory activities. Extraction from plants and chemical synthesis have been the major approaches to obtain these compounds. We used a biological method to synthesize HCAA derivatives (arylalkylamines). Two genes, SHT encoding serotonin N-hydroxycinnamoyl transferase and 4CL encoding 4-coumaroyl-CoA ligase, were introduced into Escherichia coli. Using this E. coli transformant as a biocatalyst, 24 arylalkylamines were synthesized. The anti-inflammatory activities of five synthesized compounds, including N-p-coumaroyl phenethylamine, N-caffeoyl phenethylamine, N-p-coumaroyl 3-phenylpropylamine, N-p-coumaroyl 4-phenylbutylamine, and N-p-coumaroyl 4-methoxyphenethylamine, were measured. Among them, N-p-coumaroyl 4-phenylbutylamine showed the best anti-inflammatory activity.


Introduction
Hydroxycinnamic acids (HCAs; also known as phenylpropanoids) are the major constituents of diverse phenolic compounds found in plants, including lignin, flavonoids, and tannins [1]. HCAs also form conjugates with other molecules such as amines, sugars, and organic acids. HCA amides (HCAAs) are formed by the reaction between HCAs and amines such as phenethylamine, tyramine, tryptamine, or dopamine [2,3]. HCAAs have been isolated from plants and their structures were determined [2,4]. Some phenolic amides, such as dihydro-N-caffeoyltyramine trans-N-feruloyloctopamin, trans-N-caffeoyltyramine, and cis-N-caffeoyltyramine, exhibit antifungal activity [5]. Furthermore, the production of some phenolic HCAAs is increased upon infection of pathogens and wounding. Therefore, they are considered to mediate host defenses [6,7].

Synthesis of chemicals
4CL and SHT were cloned into the pCDF duet vector as reported previously (pC-4CL-SHT) [19]. The resulting construct was transformed into E. coli BL21 (DE3). After overnight culture, the transformant was inoculated into fresh Luria-Bertani broth containing 50 lg/mL spectinomycin. The transformant was grown at 37°C to an absorbance value of 1.0 at 600 nm. Genes in the transformant were induced by adding 1 mM isopropyl b-D-1thiogalactopyranoside to the culture medium, which was then incubated at 18°C for 16 h. The cells were harvested and resuspended in M9 medium. To examine the relative conversion rate of each compound, 1 mM substrate was added to the medium and it was incubated at 30°C.
The reaction products were analyzed using high-performance liquid chromatography [19,20]. The structures of N-HC phenethylamine derivatives were determined using proton nuclear magnetic resonance (NMR) spectroscopy [21]. Spectral features were: N-p-coumaroyl 4-methoxyphenethylamine, 1  Quantification of each product was carried out using p-coumaric acid as a standard. Triplicate experiments were performed, and error bars indicate the standard deviations.

Cell viability
The murine macrophage cell line (RAW 264.7) was maintained in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum and penicillin/streptomycin at 37°C under a humidified atmosphere of 5% CO 2 . RAW 264.7 cells were plated in 96-well plates at a density of 3 9 10 4 cells per well 24 h prior to treatment. Five different compounds [N-p-coumaroyl phenethylamine, N-caffeoyl phenethylamine, N-p-coumaroyl 3-phenylpropylamine, Np-coumaroyl 4-phenylbutylamine, and N-p-coumaroyl 4-methoxyphenethylamine; (Table 1)] were added to the cells at 0, 1, 10, and 25 lM. After incubation for 24 h, the culture medium was removed completely and CCK-8 solution was added 1:10 (v/v) with fresh culture medium. After further incubation for 1 h at 37°C, the absorbance of the supernatant was measured at 450 nm using a SpectraMAX microplate reader (Molecular Devices, Sunnyvale, CA, USA) according to manufacturer's protocol. To examine cell viability in detail, two compounds were selected (N-p-coumaroyl 4-methoxyphenethylamine and N-p-coumaroyl 4-phenylbutylamine) and added to cells at 0, 1, 2, 5, and 10 lM. After 24 h incubation, cell viability was assessed as described above.
Cytokine release study RAW 264.7 cells were plated in 6-well plates at a density of 6 9 10 5 cells per well 24 h prior to treatment. After cells were pre-incubated with each of six compounds (curcumin, N-p-coumaroyl phenethylamine, N-caffeoyl phenethylamine, N-p-coumaroyl 3-phenylpropylamine, Np-coumaroyl 4-phenylbutylamine, and N-p-coumaroyl 4-methoxyphenethylamine) for 6 h at different concentrations as shown in Table 1, LPS (1 lg/mL) was added to the cells for 18 h. The cells were then pelleted by centrifugation at 13,000 rpm for 10 min at 4°C. The amount of TNF-a released into the supernatants was measured using a mouse TNF-a ELISA kit according to manufacturer's protocol. To examine TNF-a induction in detail, two compounds (N-p-coumaroyl 4-methoxyphenethylamine and N-p-coumaroyl 4-phenylbutylamine) were added to cells at 1, 2, 5, and 10 lM for 6 h prior to co-incubation with LPS (100 ng/mL) for a further 18 h. Released TNF-a levels were measured quantitatively by an ELISA as described above.

Results and discussion
Biological synthesis of arylalkylamines E. coli harboring pC-4CL-SHT synthesized six different HCA-phenethylamine derivatives from eight HCAs tested [19]. This indicated that SHT and 4CL used diverse substrates. We investigated whether SHT could use phenethylamine derivatives. To examine whether E. coli harboring pC-4CL-SHT could use 4-methoxyphenethylamine, 3-phenylpropylamine, and 4-phenylbutylamine as acyl group acceptors, and p-coumaric acid as an acyl group donor to produce arylalkylamines, the culture filtrate was analyzed using highperformance liquid chromatography. A new peak was observed from each reaction combination (Fig. 1). The molecular masses of the products of 4-methoxyphenethylamine, 3-phenylpropylamine, and 4-phenylbutylamine were 297.666, 281.621, and 295.643 Da, respectively. These were the expected molecular masses of each product. We also determined the structure of each product using NMR (see below). Twelve acyl group donors (HCA derivatives) along with three phenethylamine derivatives were also tested as substrates. One acyl group acceptor and one acyl group donor were fed at a time to E. coli expressing pC-4CL-SHT. Among the 36 donor and acceptor combinations tested, over 27 products were synthesized. Each acyl group acceptor showed a similar preference; HCA derivatives containing 4-hydroxy or 4-methoxy were better substrates than those containing 2-hydroxy, 3-hydroxy, or 3-methoxy groups. HCAs containing more than two methoxy groups (2,4-dimethoxycinnamic acid, 3,4-dimethoxycinnamic acid, sinapic acid, and 3,4,5-trimethoxycinnamic acid) were poor substrates ( Table 2).
To confirm that the reaction products had the expected structures, the structures of three compounds (each acyl acceptor and p-coumaric acid conjugate) were determined using NMR. Three compounds were synthesized by feeding p-coumaric acid and each amine at 1 mM. We synthesized 799.1 lM N-p-coumaroyl 4-methoxyphenethylamine, 730.2 lM N-p-coumaroyl 3-phenylpropylamine, and 868.3 lM N-p-coumaroyl 4-phenylbutylamine.
The synthesized compounds were purified as described in An et al. [22]. The structure of each compound was determined using NMR. In the 1 H NMR spectrum of N-pcoumaroyl 4-methoxyphenethylamine, there were eight proton peaks. One was a methoxy signal (3.75 ppm, s, 3H) and two showing an aryl region were H-9 and H-10. Five peaks were evident in the aromatic region. Among them, two at 7.45 and 6.46 ppm showed doublets with approximately 16 Hz of coupling constants. Therefore, they were easily assigned as H-5 and H-6. Integration of the other three peaks at 7.41, 7.15, and 6.85 ppm revealed they were 2H, 2H, and 4H, respectively. All eight protons of the phenyl groups were evident. The 1 H NMR assignments of Table 1 Concentrations of compounds used for assessing cell viability and anti-inflammatory activity Therefore, structures of the three reaction products were determined to be N-p-coumaroyl 4-methoxyphenethylamine, N-p-coumaroyl 3-phenylpropylamine, and N-pcoumaroyl 4-phenylbutylamine, respectively.  Toxicity and anti-inflammatory activity of five synthesized compounds We tested the toxicity and anti-inflammatory activity of the five synthesized compounds (N-p-coumaroyl phenethylamine, Ncaffeoyl phenethylamine, N-p-coumaroyl 3-phenylpropylamine, N-p-coumaroyl 4-phenyl butylamine, and N-p-coumaroyl 4-methoxyphenethylamine) by the CCK-8 assay and measuring TNF-a release, respectively. None of the compounds showed any significant cytotoxicity up to 10 lM ( Fig. 2A). N-p-Coumaroyl phenethylamine, N-caffeoyl phenethylamine, and N-p-coumaroyl 4-methoxyphenethylamine showed negligible inhibition of TNF-a release following treatment with 1 lg/mL LPS at concentrations of 10 and 1 lM. However, N-p-coumaroyl 3-phenylpropylamine and Np-coumaroyl 4-phenylbutylamine showed approximately 24.6 and 22.6% inhibition at 10 and 1 lM, respectively (Fig. 2B).
We further investigated the effects of N-p-coumaroyl 4-phenylbutylamine and N-p-coumaroyl 4-methoxyphenethylamine using four different concentrations of each compound to test their cytotoxicity and the anti-inflammatory activity at a lower LPS concentration (100 ng/mL). Neither compound was cytotoxic up to 10 lM (Fig. 2C). The amount of TNF-a released was decreased to 68.0 ± 3.2 and 82.1 ± 3.9% by treatment with 10 lM N-p-coumaroyl 4-phenylbutylamine and N-p-coumaroyl 4-methoxyphenethylamine, respectively (Fig. 2D).
To date, most studies of HCAA involved changing the side chain in phenolic amides and using these compounds to test several activities. For example, Lee et al. [18]  synthesized HCAAs showed anti-inflammatory activity [17]. These HCAAs had diverse amine moieties such as tyramine, dopamine, and phenethylamine, and higher concentrations (25 and 50 lM) of each compound were used. Biological synthesis of arylalkylamines has been tried previously [19,23], but these studies were limited to synthesizing arylalkylamines that had diverse phenolic side chains. In the current study, we changed not only the phenolic side chain in the amines, but also the carbon chain length of the amines using engineered E. coli, and tested their anti-inflammatory activity. This was the first attempt to alter the chain length of the phenolic moiety in the amine.
Previous studies showed that caffeic and ferulic acid moieties of HCAAs showed better anti-inflammatory activity than p-coumaric acid. Our study revealed that increasing the length of the alkyl group of the phenyl amine, such as 4-phenylbutylamine, yielded a better antiinflammatory compound. Our initial screening involved synthesizing N-feruloyl 4-phenylbutylamine, and N-cafferoyl 4-phenylbutylamine. The anti-inflammatory activity of these compounds is intriguing and should be examined further.