Thirty different marine macro algae (seaweed) belonging to three divisions were tested for quorum sensing inhibition using Chromobacterium violaceum CV026. Among them, A. taxiformis showed QS inhibition (Table 1). Violacein production is a quorum sensing regulated behavior in strain CV026. The white colored, opaque zone of inhibition with intact bacteria represents the QS inhibition. Antibacterial activity, represented by transparent zone/growth inhibition, was also observed. Cinnamaldehyde was used as a positive control, because at low concentrations, it does not inhibit the growth of the reporter strain and inhibits AHL-mediated QS [26,27]. The predicted mechanism of QS inhibition involves the interference of three carbon aliphatic side chains, with the binding of the smaller AHLs to their cognate receptors [26]. The biofilm formation is one of the important means of fouling in marine habitat. It is a quorum sensing-mediated process. QS controls bacterial biofilm differentiation and maturation, and its disruption may prevent microbial biofouling [20]. A. taxiformis belongs to red algae (Bonnemaisoniales, Bonnemaisoniaceae and Rhodophyta). It has been reported earlier that red algae show the highest antifouling activity among three groups of marine macro algae [28]. It is worth mentioning that Delisea pulchra, belonging to red algae, is known to synthesize halogenated furanones, which inhibit quorum sensing through accelerated LuxR turnover [17,18,19].

The cinnamaldehyde (positive control) showed only QSI, turbid growth and inhibition of violacein (Figure 1a). Standardization of cinnamaldehyde concentration showed maximum QS inhibition at 75 mM (data not shown). Methanol was used as negative control and did not have any significant effect on violacein production (Figure 1a). Antibacterial activity is revealed through a zone of clearance at the center, while QSI is seen at the periphery (Figure 1b). Some of the gorgonian corals from the Caribbean reef showed antibacterial and QS inhibitory effects [24]. Lyngbyoic acid obtained from a marine cyanobacterium is known to inhibit quorum sensing [29]. Since, N-acyl-homoserine lactone is considered as a signaling molecule of widely present Gram-negative bacteria in marine environment, we have targeted the LuxI/LuxR-type QS system of Gram-negative bacteria as a reporter system for the study of QS inhibitors from some common seaweeds. Though the antimicrobial potential of Asparagopsis had been reported earlier [30,31,32,33,34,35], we are reporting the quorum sensing inhibition for the first time.

Five different SPE cartridges were used to fractionate the extract, and 50 different fractions were collected. C₂ and C₁₈ cartridges showed clear separation of two activities, one showing QSI and another antimicrobial activity (Table 2). In contrast, the fractions obtained using CN-E cartridge showed only antimicrobial activity, and those from modified styrene-divinylbenzene polymer (PPL) showed only QSI (with finger-like projections). The PH cartridge was most efficient in separating the fractions into QSI and antimicrobial activity (Table 2). The properties of the sorbents present in different cartridges help in predicting the structure (aliphatic/aromatic) and nature (polar/non polar) of the possible active compound (Table 3). Using PH cartridge, it was also possible to differentiate the fraction showing QSI into distinct finger-like projections (Figure 2a). The results clearly show that with antimicrobial activity, QSI and QSI with finger-like projections could be assigned to different fractions of the extract (Figure 2a–c). We presume that QSI and QSI with finger-like projections may be due to two different active compounds or the derivative of the same with a different diffusion. We could separate these two effects using different cartridges (Table 2). C2 cartridge gave only QSI, while in PPL, only QSI with finger-like projection was seen. This needs further validation by identifying the compounds. The variations obtained with different cartridges may be linked to the compounds with different polarities and their interactions with the sorbent (as detailed in the Experimental section). The Bond Elut PH exhibits slightly different selectivity from other non polar sorbents. This added selectivity results from the electron density of the aromatic ring present in it. Retention of planar conjugated organic molecules is enhanced as compared to aliphatic bonded phases.

Further, the toxic effect of the extract and fractions on the reporter, C. violaceum CV026, was investigated using the disc diffusion method. The results are depicted in Figure 3. The extract gave both GI, a clear zone around the disc and QS inhibition at the periphery (Figure 3a). The fraction of the extract, which showed GI in QSI bioassay (Figure 2c), showed a clear zone, confirming the GI in the disc assay (Figure 3b). As expected, another fraction gave a white opaque zone around the disc (Figure 3c) without showing any GI. The result confirms the findings of QSI assay and also supports the notion that the QSI is based on the interference of bacterial signaling, rather than antibacterial activity.

The QSI activity of the individual fraction of the extract was further confirmed using an isogenic AHL-negative strain (MG44) of Serratia liquefaciens. In the biosensor, S. liquefaciens, the sole AHL synthase gene swrI is mutated by gene replacement, and it gives a green fluorescent protein in presence of external AHL [36]. The absence of green color fluorescence indicates the quorum sensing inhibition. The positive fraction of the extract inhibited the AHL-induced green fluorescence (Figure 4). This confirmed the QSI of the extract as observed in the C. violaceum CV026 bioassay.

ICR-FT/MS has a molecular mass resolution enabling to overlay all spectra of fractions showing antibacterial activity from the various SPE systems in one color (red) together with the adjacent non-active (QS inhibition) fractions (in blue) to constrain the positive hits corresponding to the active compounds (Figure 5a–c). Overlaid ICR-FT/MS spectra showed considerable analogies between A. taxiformis fractions, which were tested positive for antibacterial activity, and those who were for QS inhibition, except for a few signals. For some detected masses, the intensities were significantly increased for many active fractions. The ultrahigh resolution of the high field ICR-FT/MS system enabled a direct assignment of elementary composition to the exact masses of interest with a 200 ppb precision. Figure 5c showed a peak corresponding to a molecular mass of 307.58. The likely chemical formula of the compound is predicted to be C₁₄H₂₇O₅S, and the expected active compound could be 2-dodecanoyloxyethanesulfonate. The calculated molecular mass of the compound is 307.4, which is close to the observed mass. To our best of knowledge, the molecular weight does not resemble any of the previously reported inhibitors. However, sulfur-containing AHL-analogues, N-(propylsulfanylacetyl)-L-homoserine lactone and N-(pentylsulfanylacetyl)-L-homoserine lactone are known to inhibit QS [37].

Thirty different seaweed samples were collected from intertidal zone of Bay of Bengal and Arabian Sea at six different locations from Tamil Nadu, India (Thonithurai, Nochiurani, Kovalam, Erwadi, Rameshwaram and Sadamunian Valasai with N 09°17′16.9″; E 79°11′11.4″, N 09°16′16.0″; E 79°01′02.0″, N 08°05′05.3″; E 77°33′33.3″, N 09°12′12.4″; E 78°43′43.5″, N 09°09′09.6″; E 78°39′39.5″, N 09°11′11.4″; E 78°43′43.1″, respectively). The herbaria have been deposited in the Taxonomic Reference Center for Seaweeds at CSIR-CSMCRI, Bhavnagar, India. The fresh samples were washed twice with seawater. The samples were sterilized using ethanol (50%) and sodium hypochloride (1%) [38]. Thereafter, washed samples were blot dried and methanol extract was prepared from each freeze-dried sample (1 g/10 mL). The powdered algal samples were stirred for 24 h. The seaweed extracts were filtered twice (Whatman filter paper no. 1) and then concentrated (1 g/10 mL) under vacuum using a rotary evaporator (BÜCHI, Switzerland) at low temperature (35 °C) to avert evaporation of volatile compounds [33]. The concentrated extracts were stored at −20 °C for further analysis.

Production of violet color pigment is a quorum sensing regulated behavior in C. violaceum strain CV026. The extracts were screened for QSI using C. violaceum CV026 as the reporter strain, as per Milton et al., with modification [23,39]. Five hundred microliters of overnight grown culture of C. violaceum CV026 was inoculated in 20 mL of nutrient broth medium and incubated at 30 °C, shaking at 170 rpm for 2–3 h until the optical density reached up to 0.7 at 600 nm. Nutrient broth soft agar (0.8% agar, 150 mL) was maintained at 45 °C, and 10 mL of culture (OD_600nm = 0.7), 10 μL from 1 mg/mL stock solution of hexanoyl homoserine lactone (Sigma-Aldrich, St. Louis, MO, USA) were added before plating. Five microliter extract was spotted on top of the agar. After overnight incubation of the plates at 30 °C, the surroundings of the spots were examined for inhibition of reporter violacein. Cinnamaldehyde (Sigma-Aldrich, St. Louis, MO, USA) was used as positive control, and the optimum concentration showing maximum QS inhibition was determined. Equal volume (5 μL) of methanol was used as negative control. Bioassay was performed in triplicate.

The methanol extract of A. taxiformis was dried and redissolved in purified, acidified water (pH-2.0). This exchange was needed to reverse the polarity of the samples and transfer them to solid phase extraction (SPE). The extracts were fractionated using five different SPE cartridges in changing polarity (Table 3), namely Bond Elut C₂, C₁₈, CN-E, PPL and PH (Agilent, Santa Clara, CA, USA) [40]. These cartridges have a volume of 1 mL, including 100 mg sorbent material of 40 μm particle size. SPE materials were chosen to be orthogonal in their sorption behavior to obtain a wide spectrum of analytes and classify them according to their hydrophobicity.

Each cartridge was conditioned with 1 mL methanol, 5 mL purified water and 1 mL purified, acidified water (pH of 2.0) before loading with 200 μL of sample. The cartridges were than eluted consequently 10 times with 200 μL eluent of different methanol concentrations (from 10% to 100% v/v methanol in water). In total, 50 fractions (10 per cartridge) were collected.

Each fraction was re-screened for QS inhibition using C. violaceum CV026, as given in Section 3.2 above. Results were analyzed by observing the surrounding of the spot. Veselova et al. [41] found that production of the purple pigment violacein is controlled by the quorum sensing molecule N-hexanoyl-L-homoserine lactone (HHL). C. violaceum 026 (CV026) is a mutant strain that is incapable of producing HHL and violacein. However, it possesses the capacity of violacein production in the presence of external AHLs, which can be evaluated by the color change of colony (white to purple) [41].

Quorum sensing inhibition activity of A. taxiformis was also confirmed using green fluorescence protein (GFP)-tagged Serratia liquefaciens MG44 as the reporter strain [36]. Overnight grown culture (150 μL) of Serratia liquefaciens MG44 in 5 mL Luria-Bertani (LB) medium, containing ampicillin (100 μg/mL), kanamycin (50 μg/mL) and tetracycline (20 μg/mL), was mixed with 50 μL of N-hexanoyl-L-homoserine lactone (0.5 mM) and spread on LB plate and kept for 30 min. Methanolic algal extract of A. taxiformis (5 μL) was spotted and kept for overnight incubation at 30 °C. The antiquorum sensing activity of the extract was observed under UV light.

Selected SPE fractions were analyzed by a 12 Tesla Ion Cyclotron Resonance Fourier Transformation Mass Spectrometer (ICR-FT/MS; Bruker, Bremen, Germany) coupled to an Apollo II Electrospray Ionization source (ESI; Bruker, Bremen, Germany) in negative mode. The samples were injected at a flow rate of 120 μL/h with a nebulizer gas pressure of 17.4 psi and a dry gas flow rate of 4 L/min (200 °C). All spectra were acquired with a collision energy of 1.5 V and a time domain size of 4 MWord within a mass range of 100–1000 m/z. For each spectrum, 100 scans were accumulated. All fractions showing antibacterial activity and each fraction, extracted with a methanol concentration less and above, were diluted 1 to 50 with methanol and transferred to a Gilson Auto sampler 223 (Gilson Incorporated, Middleton, Wisconsin, WI, USA). In total, spectra of 17 extracts were acquired.