Extraction and partial characterisation of antioxidant pigment produced by Chryseobacterium sp. kr6

Abstract Pigments synthesised by Chryseobacterium sp. kr6 growing on feather waste were extracted and characterised. The pigment extract was characterised by KOH test, UV–vis, CIELAB colour system, HPLC-DAD-MS, FTIR and its antioxidant capacity was evaluated. A positive bathochromic shift was observed when kr6 colonies or pigment extracts were subjected to alkaline solution (20% KOH) and a λmax at 450 nm was detected for acetone extracts, although no typical fine structure of carotenoids was detected in the electomagnetic spectra. The HPLC profile of the extracted pigment showed that the compound has three different peaks with λmax near 450 nm. The FTIR analysis shows some principal functional groups from a flexirubin-like molecule. The pigmented compound also presents antioxidant activity evaluated by the scavenging of the ABTS radical.


Introduction
Pigments have been widely used in different types of industries, such as food, pharmaceutical, cosmetic and textile, offering features that make products more attractive and pleasing in appearance. The inputs, products and waste generated in the industrial processes of synthetic dyes can generate carcinogenic effects, cause disorder in the nervous, immune and digestive system of the workers and consumers. In addition, they present slow or no degradation in the environment and accumulate in living beings causing damage to ecosystems. Although toxic and dangerous, they are still produced and marketed on a large scale (Forgacs et al. 2004).
Pigments from natural sources such as plants, insects and other organic materials have been used by humans since prehistory. The growing demand for natural and safe products for human consumption and awareness for conserving the environment has led to increased research on pigments produced by bacteria . The use of pigmented bacteria has increased because they are widely distributed in soil, air or water and are considered easy to isolate and cultivate in the laboratory. Synthetic culture media are often used for the cultivation of bacteria but due to their high cost, available agroindustrial residues have been considered as bacterial growth substrates (Riffel et al. 2011). There is also the possibility of applying genetic techniques to improve the pigment yields and to date it is known that bacterial pigments are biodegradable in the environment and have no toxicity at all (Tuli et al. 2014;Venil et al. 2017). In addition, some natural pigments such as carotenoids or flexirubin-type (Figure 1), naturally produced by bacteria of the family Flavobacteriaceae, have been associated to prevention of various diseases such as cancer, age-related macular degeneration, cataracts, cardiovascular illness, treatment of chronic skin diseases, eczema, gastric ulcers among others (Jomova and Valko 2013).
In this context, the present study had as objective to produce and characterise the pigments produced by Chryseobacterium sp. kr6, originally isolated from decomposing chicken feathers, using a synthetic nutrient medium or mineral medium with feather meal as growth substrate.

Pigment production and extraction
After 48 h cultivation, an average 0.251 g of Chryseobacterium sp. kr6 biomass generated 0.0778 g of total pigment extract from liquid BHI medium. The extraction method employed was easy to perform and no pigment was left in the biomass. Acetone is the most common solvent for extracting such pigments, followed by methanol and ethyl acetate (Wang et al. 2013;Schöner et al. 2014;Venil et al. 2016). In this work, it was evidenced that the initial maceration process facilitates the cell rupture, resulting a more homogeneous powder, providing a higher contact surface of the biomass with the solvent and a rapid and effective extraction ( Figure S1). The extraction of carotenoids from shrimp waste was improved by alkaline treatment as compared with the use of solvents (Jeddi et al. 2013), suggesting that the pigment source may have influence on recovery.

KOH 20% test
The pigment extracted from kr6 was positive for KOH test. There was a rapid shift from yellow to reddish when the drops of the alkaline solution were added to the extracted pigment in acetone ( Figure S1). When tested directly on the colonies of kr6 they also had a colour shift from orange (original colony colour) to red ( Figure S1). The same colour change has been reported by other authors, which concluded that it indicates the presence of a flexirubin-type pigment (Oren 2011;Venil et al. 2014). The reversible shift of the λ max occurs by the presence of phenolic hydroxyl group in ring A, which is possibly oxidised by the highly basic pH, and the polyene chromophore. Although the KOH test was positive, it should be considered that some types of sulphonolipids such as seen in Xanthomonas spp., or phenolic carotenoids may have similar structural elements and give a misleading positive reaction (Chaudhari et al. 2009).

UV-visible spectrophotometry
The colour change was analysed by UV-visible spectrophotometry to monitor changes in wavelength of maximum absorption (λ max ). The λ max value for the pigment prior to the addition of KOH was 450 nm, and shifted to 465 nm after addition of the alkaline solution ( Figure  S2 A). The increment in λ max value after addition of a strong base may be an indicative for estimating the chain size of the polyene associated with the pigment structure. Flexirubintype pigments showing a λ max increase of about 25 nm would indicate that the molecule has 6-8 conjugated double bonds in its structure (Achenbach 1987).
The spectrophotometric analysis of the kr6 pigment did not shown the typical fine structure of carotenoid pigments. It presented a λ max at 450 nm in acetone, similar to other works that analyse orange pigments extracted from Chryseobacterium species or bacteria belonging to the Flavobacteriaceae family (Wang et al. 2013;Schöner et al. 2014Schöner et al. , 2016. For example, the pigment of Chryseobacterium sp. UTM-3T also exhibited maximum absorbance at 450 nm when extracted with acetone (Venil et al. 2014).
Pigmented compounds of strain kr6 extracted from each of the cultivation media tested in this work had also a λ max at 450 nm. The spectra of each extracted compound and a more concentrated sample from BHI medium (Figure S2 B). In other studies, the strain kr6 was demonstrated to grow and produce keratinolytic enzymes using feather meal as unique source of nutrients (Riffel and Brandelli 2006;Riffel et al. 2011). In this work, it was observed a decrease in the absorbance signal in feather meal media, possibly due to the small amount of biomass produced in this condition, which resulted in less pigment produced.

Determination of CIELAB colour parameters
The extracted pigment was set at an absorbance value of 1.5 at 450 nm and analysed by colorimetry. The values obtained for the colorimetric parameters of the CIELAB system are presented in Table 1. The L* value was high, indicating a high luminosity, close to white. The negative value of a* indicates a colour close to yellow and green, and the value of b* was positive, which confirms that the compound is yellow shade. The Chroma or saturation was 43.14 and the Hue angle was 78.59, which also represents colours in the yellow region. Previous works evaluating the colour of flexirubin-type pigments extracted from Chryseobacterium sp. used an absorbance range between 1.0 and 2.0 at 450 nm for the analysis (Venil et al. 2014;Aruldass et al. 2016). The values of L*, b* and chroma were low but the value of a* was also negative (Table 1). For the Hue angle they obtained a value of 77.39 and 84.37, similar to that observed in this work. The L* value may vary by the previously established absorbance range of the colour analysis as well as saturation or purity (Chroma value).

HPLC-DAD-MS
The pigments extracted with acetone were analysed by HPLC-DAD. The pigments obtained from each cultivation had three major peaks with identical retention times, which indicate that in the different culture media analysed, the same type of pigmented compound is generated. The three major peaks appeared in acetone-extracted pigments obtained from different cultivation media ( Figure S3), although their corresponding area decreased when cultivation was performed in feather meal. This smaller amount of extracted pigment suggests that cultivation with few nutrients generated less biomass resulting in less pigment. In addition, each of the peaks obtained from different cultivation media had very similar UV-visible spectra with λ max at 450 nm or near this value. This small variation in wavelength depends on the number of conjugated carbons in the main chain of the chemical structure.
The m/z values obtained as [M + H] + for the three major peaks were 557, 608 and 569, respectively. Even with these data, it was not possible to define the molecular structure for the pigmented components. Thus, further efforts are needed to improve the purification process and elucidate the exact chemical structures.

Fourier transform infrared spectroscopy
The FTIR analysis was performed with lyophilised biomass and extracted pigment from Chryseobacterium sp. kr6. The spectra obtained from the biomass ( Figure S4 A) did not show specific peaks of flexirubin due to the large number of compounds present in the sample and its paste characteristic, which made it difficult to distribute in the KBr support and its measurement. However, it was possible to identify the peaks of the main functional groups of flexirubin in the spectrum of partially purified pigment (Figure S4 B). The peaks at 2921 and 2848 cm −1 are characteristic of the C-H bond found in the polyunsaturated chain and the aromatic ring. The peak obtained at 1733 cm −1 corresponds to the stretching of the C=O bond of the ester linking the phenol with the polyunsaturated chain and the peak at 1466 cm −1 is characteristic of the aromatic ring of the phenol. In addition, a peak at 718 cm −1 was observed, which could correspond to the C-Cl bond, suggesting a flexirubin-type pigment with a Cl radical on the resorcinol ring. Venil et al. (2014) characterised a flexirubin-type pigment from Chryseobacterium sp. UTM-3T, and identified in the infrared spectrum of the pigment intense peaks at 1737 and 1217 cm −1 , characteristic of the C-O bonds. In addition, they found an absorption band between 893 and 526 cm −1 , which according to the authors, is characteristic of flexirubin pigments. The basic chemical structure of this pigment can be modified by varying the length and branching of the hydrocarbon chains in resorcinol and by introducing additional phenyl and chlorine substituents specifically on the dialkylresorcinol ring (Oren 2011;Jehlička et al. 2013). Different chlorination in the aliphatic rings is found in each microorganism producing flexirubin. Thus, a wide variety of flexirubin-type pigments may arise (Kim and Stafsnes 2013). Due to the limited existing literature on infrared spectra of this type of pigment, it is difficult to make a more accurate comparison. In addition, the degree of purification of the compound affects the identification of specific bands by overlapping with the lipid bands present in the sample.

Antioxidant activity
The pigmented compound extracted from kr6 biomass presented antioxidant activity. An ABTS radical scavenging activity of 85% was obtained. Values calculated as Trolox equivalents were similar to those obtained for carotenoids and xanthophylls evaluated by their antioxidant activities by eliminating the ABTS radical cation (Table S1).
A flexirubin pigment from Chryseobacterium sp. UTM-3T also showed antioxidant activity by the DPPH method, which was attributed to the amphiphilic properties of the phenolic constituents . Another study confirms the antioxidant activity of the flexirubin-type pigment isolated from Fontibacter flavus YUAB-SR-25. The pigment showed significant antioxidant activity by eliminating DPPH and hydroxyl radicals, inhibition of lipid peroxidation and reduction of nitric oxide (Prabhu et al. 2013). The antioxidant activity of the pigments under study, considered as an alkylpolyene-dialkylresorcinol hybrid, is attributed to the OH phenolic groups present in their structures and the alkyl chain containing conjugated double bonds. At least in vitro, these compounds must be protected against photo oxidative damage and lipid peroxidation because each contains a conjugated double bond system, which is primarily responsible for their antioxidant function.

Bacterium and culture conditions
Chryseobacterium sp strain kr6 was isolated from chicken feathers (Riffel and Brandelli 2006). The lineage was stored at the Laboratory of Biochemistry and Applied Microbiology of ICTA -UFRGS (Porto Alegre, Brazil) and maintained in Brain Hearth Infusion (BHI) broth containing 20% (v/v) glycerol at −20 °C. Specifically, kr6 colonies are bright orange in BHI medium.
The liquid culture media used for the production of biomass and pigments from strain kr6 were BHI broth, or medium containing feather meal (10 g/L), NaCl (0.5 g/L), KH 2 PO 4 (0.4 g/L) and K 2 HPO 4 (0.3 g/L), supplemented or not with 10 g/L BHI. The media were esterilised at 121 °C for 15 min. The inoculum (1 mL) adjusted to an optical density of 1.0 at 600 nm, was transferred to 70 mL of culture medium in a 250 mL Erlenmeyer flask and incubated in an orbital shaker at 250 rpm for 48 h. The culture temperature was 30 °C and the pH of the medium was adjusted to 8.0.

Pigment extraction
The cultures were centrifuged at 16,000g for 10 min at 10 °C and the supernatant was discarded. The pellets were washed three times with distilled water followed by centrifugation at the previous conditions. The supernatant was discarded and the pellet was lyophilised and stored in a desiccator covered from light until the extraction processes. The total lyophilised biomasses were initially macerated with 3 mL acetone until getting a dry powder and were placed in ultrasonic bath (Ultrasonic Cleaner USC 700, Unique) suspended again in 2 mL acetone at room temperature (23 ± 2 °C). The extraction was repeated until the biomass was colourless. To confirm that the pigment was fully extracted, 2 mL of ethyl acetate and 2 mL of methanol were used in the last extraction. The samples were centrifuged at 16,000g for 5 min at 10 °C. The supernatant containing the pigment was collected with a glass pipette in a test tube, dried with nitrogen and stored at −10 °C. Biomasses and dry pigment were previously weighed to calculate the yield of pigment production by the bacteria.

KOH test
The test was performed on the extracted pigment collected in glass tubes and directly on the bacterial colonies placed on a glass slide. An amount of 3-4 drops of 20% (w/v) KOH was added to the samples. The rapid reversible shift from bright orange to a reddish pink, purple or brown colour indicates that might be a flexirubin-type pigment (Oren 2011). When there is no colour shift it may suggest that it is a carotenoid-type pigment.

UV-vis spectrophotometery
The maximum absorption wavelength (λ max ) of the yellowish pigment was determined by UV-vis scanning spectrophotometry between 200 and 800 nm. The electronic spectrum was obtained on a Shimadzu-UV-mini 1240 equipment (Shimdzu, Kyoto, Japan) at 5 nm intervals. The dried pigment was diluted in 5 mL acetone for the measurement.

Determination of CIELAB colour parameters
The pigment was set at an absorbance of 1.5 at the maximum wavelength of 450 nm using acetone as the solvent for the analysis. L*, a* and b * values were measured using a. colorimeter (Minolta, CR-400, Osaka, Japan) with the CIELAB colour system (Hunter Associates Laboratory Inc., Virginia, USA). Standard values refer to the white calibration plate (L = 94.23, a = −0.55 and b = 9.68). L* indicates the specific brightness of the pigment from 0 (black) to 100 (white). Values of b* greater than 0 represent yellow colour and values smaller than 0 blue colour, for values of a* greater than 0 represent green colour and values smaller than 0 red or purple colour. Values equal to 0 (a* = b* = 0) represent achromatic colour, that is, grey colour. With these values, the values of hue angle (Hue°) and Chroma or saturation of the pigment were calculated.
The Chroma value denotes the saturation or purity of the pigment colour and was calculated with the formula: The value of the hue angle (Hue °) denotes 0 for red tones, 90 for yellows, 180 for green and 270 for blue and was calculated with the formula:

HPLC-DAD-MS
The identification of pigments was performed in a Shimadzu HPLC (Kyoto, Japan) equipped with quaternary pump (LC-20AD), degassing unit (DGU-20A5), automatic injector and diode array detector (DAD) (SPD-M20A) connected in series to a Bruker Daltonics mass spectrometer (Esquire 6000 model, Bremen, Germany), using APCI (atmospheric pressure chemical ionisation) as the source of ionisation. Prior to analysis, the dry pigment was dissolved in methanol and injected into HPLC-DAD-MS. The compounds were separated on a Merck C18 column (5 μm, 250 × 4.6 mm) with flow rate of 0.9 mL/min at 29 °C, mobile phase consisting of water: formic acid (99.9: 0.1% v/v) (solvent A) and methanol: formic acid (99.9: 0.1 v/v) (solvent B) in a linear gradient from 40:60 (v/v) A:B to 0:100 (v/v) over 20 min; this composition (0:100, v/v) was maintained for 10 min. The spectra were obtained between 200 and 600 nm and the chromatograms processed at 450 nm. The mass spectra were acquired with a scan range of m/z 100 to 1000; and the MS parameters were as follows: APCI source in positive ionisation mode; capillary voltage: 4000 V, end plate offset: −500 V, drying gas temperature (N 2 ): 310 °C, flow: 8 L/min, nebulizer: 30 psi; MS/MS fragmentation energy: 1.4 V.

Fourier transform infrared spectroscopy
The spectra of the lyophilised biomass of Chryseobacterirum sp. kr6 and the dry extracted pigment were obtained using a Shimadzu 8300 FTIR spectrophotometer (Shimadzu, Kyoto, Japan). Samples were placed on KBr supports for analysis in the range of 4000 to 400 cm −1 with 132 scans and the resolution was adjusted to 4 cm −1 .

Antioxidant activity
The analysis was performed according to the methodology described by Nenadis et al. (2004) with some modifications. ABTS and potassium persulphate were dissolved in distilled water to give a final concentration of 7 mM and 140 mM, respectively. The two solutions were mixed and left at room temperature in the dark for 16 h prior to their use to produce the ABTS radical (ABTS• + ). For the study of the antioxidant activity of the pigment the solution of the ABTS radical was diluted with distilled water to an absorbance of 0.70 ± 0.05 nm at 750 nm. Trolox was prepared in ethanol for use as standard antioxidant. The Trolox standard (concentration range 0.1-2.0 mM) was added to the diluted ABTS• + solution and the absorbance readings were performed after 6 min of the reaction using spectrophotometer.

Conclusions
Chryseobacterium sp. strain kr6 produces a pigment easily extracted with acetone, presenting colour shift in basic pH, λ max at 450 nm and compatible FTIR with orange/yellow flexirubin-type pigments. There is no difference between the pigmented compounds synthesised in different media, just the amount of the biomass and pigment due to the few nutrients. Further purification process is necessary to elucidate the precise molecular structure of the pigments.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work received financial support from CNPq and CAPES (Brazil).