Improved high performance liquid chromatographic method for determination of carotenoids in the microalga Chlorella pyrenoidosa

Journal of Chromatography A, 1102 (2006) 193–199

Improved high performance liquid chromatographic method for determination of carotenoids in the microalga Chlorella pyrenoidosa B. Stephen Inbaraj, J.T. Chien, B.H. Chen ∗ Department of Nutrition and Food Sciences, Fu Jen University, Taipei 242, Taiwan Received 11 May 2005; received in revised form 23 September 2005; accepted 26 October 2005 Available online 18 November 2005

Abstract Microalgae have become an important commercial source of carotenoids and microalgae-derived functional foods are consumed by people worldwide. Therefore, an HPLC method was developed to discern the variety and content of carotenoids in the microalga Chlorella pyrenoidosa. The microalga sample was powdered, extracted, saponified and subjected to HPLC analysis. A mobile phase of methanol–acetonitrile–water (84:14:2, v/v/v) (A) and methylene chloride (100%) (B) with the following gradient elution was developed: 100% A and 0% B in the beginning, maintained for 14 min, decreased to 95% A in 25 min, 75% A in 30 min, 74% A in 35 min, 45% A in 50 min and returned to 100% A in 55 min. A total of 32 carotenoids were resolved within 49 min by using a C30 column with flow rate at 1 mL/min and detection at 450 nm. An internal standard ␤-apo-8 -carotenal was used to quantify all the carotenoids. All-trans-lutein was present in exceptionally large amount (125034.4 ␮g/g), followed by cis isomers of lutein (27975.3 ␮g/g), all-trans-␣-carotene (2465.8 ␮g/g), zeaxanthin (2170.3 ␮g/g), cis isomers of ␤-carotene (2159.3 ␮g/g), alltrans-␤-carotene (2155.0 ␮g/g), cis isomers of ␣-carotene (1766.7 ␮g/g), ␤-cryptoxanthin (334.9 ␮g/g), neoxanthin and its cis isomers (199.7 ␮g/g), neochrome (65.2 ␮g/g), auroxanthin (38.5 ␮g/g) and violaxanthin and its cis isomers (38.1 ␮g/g). © 2006 Elsevier B.V. All rights reserved. Keywords: Microalgae; Chlorella pyrenoidosa; Carotenoids; HPLC

1. Introduction Chlorella tablet derived from a chlorophycean alga, Chlorella pyrenoidosa, has been the choice of over 10 million people worldwide for promotion of overall health and healing. Significant attention has recently been drawn to the use of microalgae for deriving functional food, as microalgae produces a great variety of metabolites that are essential for human health. Among the metabolites that include proteins, vitamins, minerals, enzymes and fatty acids, xanthophylls and carotenes are produced in rich amounts in microalgae during their normal growth phase and when exposed to environmental stimuli (carotenogenesis) [1]. For example, chlorophycean algae have been shown to contain major carotenoids (as in higher plants), namely, lutein, zeaxanthin, ␤-carotene, ␣-carotene, violaxanthin and neoxanthin during their normal growth phase [2]. In addition, some microalgae, under favorable conditions, synthesize very high amounts

∗

Corresponding author. Tel.: +886 2 29053626; fax: +886 2 29021215. E-mail address: [email protected] (B.H. Chen).

0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2005.10.055

of secondary carotenoids like astaxanthin, canthaxanthin and echinenone (carotenogenesis) [3]. The role of carotenoids in the prevention of chronic diseases and their health promoting mechanisms has been well documented [4–6]. Therefore, it is imperative to learn the variety and amount of carotenoids in microalgae-derived chlorella tablet. The analysis of carotenoids has been routinely performed by reversed-phase HPLC because of its improved separation efficiency [7,8]. Mostly, analysis of carotenoids in different chlorella species has been done using a C18 column, reporting inadequate carotenoid profile or lack in resolving geometrical isomers. For instance, using a C18 column, only seven carotenoids that include neoxanthin, violaxanthin, lutein, ␤carotene, astaxanthin, canthaxanthin and astaxanthin esters were isolated in both green and orange cells of chlorella vulgaris [9], while in another study on 15 chlorophycean microalgae, Del Camp et al. [10] have determined only five carotenoids namely astaxanthin, canthaxanthin, ␤-carotene, lutein and violaxanthin. None of the above studies have reported either full spectrum of carotenoids or their isomers. Several reports have demonstrated that a C30 column could provide better resolution of carotenoids

194

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

and their geometrical isomers than a C18 column [7,8,11]. Due to unavailability of comprehensive carotenoid profile and complexity of carotenoids in the microalgae (C. pyrenoidosa)derived functional food, this study was undertaken to develop an appropriate HPLC method, employing a C30 column, for determination of a complete array of carotenoids including their isomers. 2. Experimental 2.1. Materials Microalgae (C. pyrenoidosa)-derived tablets, which are available commercially from Vedan Enterprise Corp. (Taichung, Taiwan) was procured. The tablets were then ground into fine powder and used for extraction. Reference samples of neoxanthin and violaxanthin were isolated from spinach obtained from a local supermarket. Fresh spinach leaves were cut into small pieces, freeze-dried and ground into fine powder prior to extraction. The semi-preparative TLC plates (silica gel 60 F254, Merck, Germany) coated with silica gel (0.5 mm thickness) were used for preparation of neoxanthin and violaxanthin. All-trans forms of lutein, ␣-carotene, ␤-carotene, ␤cryptoxanthin and zeaxanthin standards were purchased from Sigma (St. Louis, MO, USA). Chemicals like potassium hydroxide and anhydrous sodium sulfate were obtained from Riedel-de H¨aen (Barcelona, Spain). The HPLC-grade and analytical-grade solvents (Mallinckrodt, Paris, KY, USA) were used without further purification. Deionized water was obtained using a MilliQ water purification system (Millipore, Bedford, MA, USA). Analytical separation was done using a YMC C30 column (250 mm × 4.6 mm I.D., 5 ␮m) obtained from Waters (Milford, MA, USA). 2.2. Instrumentation The HPLC instrument encompasses an Agilent G1312A binary pump (Agilent 1100 series, Agilent Technology, Palo Alto, CA, USA), a Rheodyne model 7161 injector (Rheodyne, Rohnert Park, CA), a DP-4010 degasser (Sanwa Tsusho, Tokyo, Japan), a Jasco MD-915 photodiode-array detector and a Borwin computer software. The rotary evaporator (model N-1) obtained from Eyela (Tokyo, Japan) was used for evaporation of carotenoid extracts. The minced spinach leaves were freeze dried in a freeze-dryer (model FD-24) from Chin-Ming (Taipei, Taiwan). The sonicator (model 2210R-DTH) used was from Branson (Danbury, CT, USA). The absorbance of neoxanthin and violaxanthin fractions isolated from spinach was measured using a spectrophotometer (model CE3021) from Cecil (Cambridge, UK). 2.3. Extraction of carotenoids from chlorella tablet Carotenoids were extracted from chlorella tablet by adopting a method described by Chen et al. [12]. In short, 1 g of the chlorella tablet was treated with a 30 mL mixture of hexane–ethanol–acetone–toluene (10:6:7:7, v/v) in a 100 mL

volumetric flask and shaken for 1 h. To the contents, 2 mL of 40% methanolic KOH was added for saponification at 25 ◦ C in the dark under nitrogen gas for 16 h. After saponification, 30 mL of hexane was added for partition of carotenoids, shaken for 1 min and 10% sodium sulfate solution was added and diluted to volume. The mixture was allowed to stand until two phases separated clearly. The upper layer containing carotenoids was collected, while the residue was repeatedly extracted until no trace of carotenoids was left. The extracts of upper layer were then pooled, evaporated to dryness, redissolved in 1 mL methanol–methylene chloride (50:50, v/v), and filtered through a 0.2 ␮m membrane filter for HPLC analysis. The whole extraction procedure was carried out under dimmed light and nitrogen gas was flushed into vials to avoid isomerization or degradation of carotenoids. 2.4. Preparation of neoxanthin and violaxanthin standards Owing to the absence of commercial neoxanthin and violaxanthin standards, they were obtained from spinach according to a method described by Chen et al. [12]. The TLC plates (20 cm × 20 cm) were initially activated at 100 ◦ C for 1 h, placed in glass tanks lined with filter paper and 150 mL methanol–acetone–hexane (1:29:70, v/v/v) was poured into the tank for saturation until 30 min. About 10 spots of carotenoid extract (10 ␮L each) were applied to each TLC plate using a 10 ␮L microcapillary pipette and the chromatograms were developed to a distance of about 16 cm. The individual bands were scrapped separately and transferred immediately to a sintered crucible attached with a side-arm filtration flask. As many as 200 spots of neoxanthin and violaxanthin bands were collected separately from 20 TLC plates and eluted with acetone for spectrometric measurement of absorbance at 443 nm for violaxanthin and 439 nm for neoxanthin. Knowing the extinction coefficient of violaxanthin (2550) and neoxanthin (2243), their concentrations were calculated in accordance with the Beer’s law. The concentrations of violaxanthin and neoxanthin were calculated to be 304.4 and 200.6 ␮g/mL, respectively. 2.5. Photoisomerization of lutein, α-carotene and β-carotene standards The reference standards of lutein, ␣-carotene and ␤-carotene (1 mg each) were dissolved in 10 mL methylene chloride so that volumes of 1 mL distributed in separate vials contains 100 ␮g/mL each. The vials were then placed in an incubator at 25 ◦ C under four fluorescent tubes (20 W each) at a distance of 30 cm. The vials were illuminated for 24 h with an intensity of 2000–3000 lx. The illuminated standards were evaporated to dryness, reconstituted in 1 mL methanol–methylene chloride (50:50, v/v) and filtered through a 0.2 ␮m membrane filter for subsequent HPLC analysis. 2.6. Chromatographic analysis and peak identification At the outset, several binary and ternary solvent systems in isocratic or gradient mode were experimented and their sepa-

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

ration efficiency in terms of retention factor (k ) and separation factor (α) was evaluated and compared. Also, the separation efficiency with regard to different sample solvents was compared. The most suitable mobile phase system comprised of methanol–acetonitrile–water (84:14:2, v/v/v) (A) and methylene chloride (100%) (B) with the following gradient condition: 100% A and 0% B in the beginning, maintained for 14 min, decreased to 95% A in 25 min, 75% A in 30 min, 74% A in 35 min, 45% A in 50 min and returned to 100% A in 55 min. The most suitable sample solvent was methanol–methylene chloride (50:50, v/v). The C30 column resolved a total of 32 carotenoids within 49 min when 20 ␮L of the extract was injected. The mobile phase was pumped at a flow rate of 1 mL/min and the response was detected at 450 nm. The identification of various carotenoids was accomplished by comparison of retention times, absorption spectra of unknown peaks with reference standards and co-chromatography with added standards. The cis isomers of carotenoids were tentatively identified based on the spectral characteristics and Q-ratios reported in the literature. The epoxy-containing carotenoids were identified by comparing the absorption characteristics of peaks before and after adding 10 ␮L methanolic HCl (0.1 N) to the sample extract. The peak purity was automatically determined by a Jasco photodiode-array detector.

195

ibration curves were obtained by plotting concentration against area. The LODs and LOQs calculated using a formula described in the previous article [15] were 0.06 and 0.18, 0.02 and 0.06, 0.03 and 0.09, 0.002 and 0.007, 0.008 and 0.02 ␮g/mL, respectively, for all-trans-lutein, all-trans-zeaxanthin, all-trans␤-cryptoxanthin, all-trans-␣-carotene and all-trans-␤-carotene. 2.9. Determination of recovery Recoveries of carotenoid standards from the C30 column were determined by adding two concentrations, 5 and 10 ␮g/mL each of zeaxanthin, all-trans-␣-carotene and alltrans-␤-carotene, 50 and 100 ␮g/mL of all-trans-lutein and 2 and 5 ␮g/mL of ␤-cryptoxanthin separately to the powdered chlorella tablet (1 g) before extraction. On comparing the concentration of each standard added initially and that calculated after HPLC analysis, the respective carotenoid recovery was calculated. The mean recovery obtained from duplicate analyses was 85.5, 94.5, 82.0, 91.8 and 86.8% for alltrans-lutein, all-trans-zeaxanthin, all-trans-␤-cryptoxanthin, all-trans-␣-carotene and all-trans-␤-carotene, respectively. Due to the unavailability of commercial standards and similarity in extinction coefficient, the percent recoveries of cis isomers of carotenoids were considered equivalent to those of their parent trans forms.

2.7. Preparation of standard curves An appropriate internal standard, ␤-apo-8 -carotenal, was used for quantification of carotenoids. A fixed concentration of ␤-apo-8 -carotenal (IS) (10 ␮g/mL) was added to various concentrations of zeaxanthin, all-trans-␣-carotene and all-trans␤-carotene (1, 5, 8, 10 and 20 ␮g/mL), all-trans-lutein (0.2, 0.6, 1, 10, 50, 100, 240 and 420 ␮g/mL) and ␤-cryptoxanthin (0.2, 0.6, 1, 5, 10 ␮g/mL) separately, and the standard curves for each were prepared by plotting concentration ratio against area ratio. The regression analysis was performed in Microsoft Excel XP software and a high correlation coefficient (r2 > 0.99) was found for all the standard curves. The standard curves for neoxanthin and violaxanthin were not prepared because of purity problem, which was revealed by the presence of several peaks when the individual fractions collected from TLC was analysed using HPLC. Therefore, both neoxanthin and violaxanthin were quantified in relation with the internal standard by calculating the area ratio of each to IS and multiplying with the concentration of IS. The data were subjected to analysis of variance and Duncan’s multiple range test, using SAS [13]. 2.8. Determination of limits of detection (LOD) and quantification (LOQ) In accordance with a method described by the International Conference on Harmonization [14], both LOD and LOQ were determined by preparing standard curves with three concentrations, 1.0, 5.0 and 8.0 ␮g/mL each of all- transzeaxanthin, all-trans-␣-carotene and all-trans-␤-carotene, and 0.2, 0.6 and 1.0 ␮g/mL each of all-trans-lutein and all-trans-␤cryptoxanthin. Analysis was performed in triplicate and the cal-

3. Results and discussion Considering the complexity of carotenoid profile in microalgae [1,9,10], a gradient solvent system of methanol– acetonitrile–water (84:14:2, v/v/v) and methylene chloride (100%), as described in the method section, was developed to resolve a range of carotenoids in chlorella tablet (Fig. 1). A total of 33 carotenoids, including the internal standard (␤-

Fig. 1. HPLC chromatogram of carotenoids extracted from chlorella tablet. Chromatographic conditions are described in the text. Peaks: (1) auroxanthin; (2) auroxanthin; (3) violaxanthin; (4) neochrome; (5) cis-neoxanthin; (6) neoxanthin; (7) neochrome; (8) neoxanthin; (9) cis-neoxanthin; (10) neoxanthin; (11) cis-lutein; (12) 13- or 13 -cis-lutein; (13) cis-lutein; (14) 13- or 13 -cis-lutein; (15) all-trans-lutein; (16) zeaxanthin; (17) 9- or 9 -cis-lutein; (IS) internal standard (␤-apo-8 carotenal); (18) 9- or 9 -cis-lutein; (19) cis-lutein; (20) cis-lutein; (21) cis-lutein; (22) cis-␤-carotene; (23) ␤-cryptoxanthin; (24) 13- or 13 -cis-␣carotene; (25) 13- or 13 -cis-␣-carotene; (26) 13- or 13 -cis-␤-carotene; (27) 9or 9 -cis-␤-carotene; (28) all-trans-␣-carotene; (29) 9- or 9 -cis-␣-carotene; (30) 9- or 9 -cis-␣-carotene; (31) all-trans-␤-carotene; (32) 9- or 9 -cis-␤-carotene.

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

196

apo-8 -carotenal), were resolved within 49 min. The resolution of peaks after 20 min was relatively better than that before 20 min, which is probably because of the complex nature of the epoxy-containing compounds in chlorella tablet. Nevertheless, the separation factor (α) values (Table 1) greater than 1 for all the peaks shows that a good selectivity of mobile phase to carotenoid composition in chlorella tablet was attained. Moreover, the retention factor (k ) values for all the peaks ranged from 0.53 to 13.25 (Table 1), which also signifies that a proper solvent strength was maintained. This complies with the k values reported between 0.5 and 20 for separation of complicated components [16]. Except for peaks 3 and 13, the purities of all the other peaks were higher than 90% (Table 1). Peaks 1–10 were identified as epoxy-containing compounds by comparing their retention behavior with those of neoxanthin and violaxanthin standards prepared from spinach, and observing a hypsochromic shift of about 40 nm for violaxanthin and 20 nm for neoxanthin after addition of 0.1 N methanolic HCl (Table 2). It is reported that under acidic condition, violaxanthin can be converted to auroxanthin because of formation of two 5,8-epoxides from their 5,6-epoxy groups [23]. Peak 3 was

tentatively identified as violaxanthin, because a hypsochromic shift of 40 nm occurred after epoxide test. Peaks 1 and 2 were tentatively identified as auroxanthin due to a hypsochromic shift of 40 nm shown when compared to violaxanthin before epoxide test and no wavelength change observed after epoxide test. Similarly, peaks 5, 6 and 8–10 were identified as neoxanthin or its cis isomers, as a hypsochromic shift of about 20 nm was shown after epoxide test, revealing the formation of neochrome from neoxanthin in the presence of HCl. Peaks 4 and 7 were tentatively identified as neochrome on the basis of a hypsochromic shift of 20 nm when compared to neoxanthin before epoxide test and no wavelength change observed after epoxide test. Further identification of peaks was based on the comparison of retention behavior and absorption spectra of unknown peaks with the peaks of isomerized standards. Fig. 2 shows the HPLC chromatogram of illuminated carotenoid standards with unified numbering of peaks as in Fig. 1 for comparison. The peaks in Fig. 2 were identified by comparing the spectral characteristics and Q-ratios reported in the literature (Tables 3–5). The order of elution of peaks and their assignments in Fig. 2B are consistent with those of unambiguously identified geomet-

Table 1 Retention time, retention factor (k ), separation factor (α), purity and content of carotenoids in chlorella tablet Peak no.

Compound

Retention time (min)

ka

αb

Peak purity (%)

Content (␮g/g)c

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Auroxanthin Auroxanthin Violaxanthin Neochrome cis-Neoxanthin Neoxanthin Neochrome Neoxanthin cis-Neoxanthin Neoxanthin cis-Lutein 13- or 13 -cis-Lutein cis-Lutein 13- or 13 -cis-Lutein All-trans-lutein All-trans-zeaxanthin 9- or 9 -cis-Lutein ␤-apo-8 -Carotenal (IS) 9- or 9 -cis-Lutein cis-Lutein cis-Lutein cis-Lutein cis-␤-Carotene All-trans-␤-cryptoxanthin 13- or 13 -cis-␣-Carotene 13- or 13 -cis-␣-Carotene 13- or 13 -cis-␤-Carotene 9- or 9 -cis-␤-Carotene All-trans-␣-Carotene 9- or 9 -cis-␣-Carotene 9- or 9 -cis-␣-Carotene All-trans-␤-Carotene 9- or 9 -cis-␤-Carotene

5.15 5.82 7.00 8.69 10.00 10.86 12.47 12.93 14.78 16.02 17.09 21.90 22.70 24.59 29.13 33.02 33.49 34.25 35.24 36.52 37.20 37.77 39.01 40.20 41.35 42.28 43.99 44.63 45.19 45.67 46.78 47.29 48.02

0.53 0.73 1.08 1.58 1.97 2.22 2.70 2.84 3.39 3.75 4.07 5.50 5.74 6.30 7.64 8.80 8.94 9.16 9.46 9.84 10.04 10.21 10.58 10.93 11.27 11.55 12.05 12.24 12.41 12.55 12.88 13.03 13.25

1.13 (1,2)d 1.20 (2,3) 1.13 (3,4) 1.15 (5,6) 1.09 (6,7) 1.15 (7,8) 1.04 (8,9) 1.14 (9,10) 1.08 (10,11) 1.07 (11,12) 1.06 (12,13) 1.04 (14,15) 1.08 (15,16) 1.19 (16,17) 1.13 (17,18) 1.01 (18,19) 1.02 (19,IS) 1.03 (IS,20) 1.04 (20,21) 1.02 (21,22) 1.02 (22,23) 1.03 (23,24) 1.03 (24,25) 1.03 (25,26) 1.02 (26,27) 1.04 (27,28) 1.02 (28,29) 1.01 (29.30) 1.01 (30,31) 1.02 (31,32) 1.01 (32,33) 1.02 (33,34) 1.02 (33,34)

95.7 94.9 82.0 97.7 98.8 99.8 99.8 96.6 98.7 98.8 95.9 97.2 84.6 91.8 97.3 96.1 97.0 98.6 97.6 99.9 99.9 99.9 99.9 99.8 99.9 99.7 99.3 99.6 92.2 99.7 99.9 92.2 99.4

19.8 18.7 38.1 37.2 18.7 24.0 28.0 40.4 103.3 13.3 3731.5 9795.6 1038.7 7763.7 125034.4 2170.3 2825.2

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.1 0.1 0.2 0.1 0.3 0.2 0.2 0.4 0.7 0.4 8.8 18.4 6.0 13.8 42.3 13.7 4.7

1921.6 478.3 203.8 216.9 254.9 334.9 362.8 545.9 727.7 579.6 2465.8 515.6 342.4 2155.0 597.1

± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

6.0 5.2 2.5 1.7 5.2 6.3 2.9 2.6 7.2 4.1 10.9 3.5 4.2 13.3 4.4

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 a b c d

k : retention factor. α: selectivity (separation factor). Average of duplicate analyses ± standard deviation. Numbers in parentheses represent peak numbers.

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

197

Table 2 Tentative identification data for all-trans plus cis forms of carotenoids in chlorella tablet Peak no.

Compound

Retention time (min)

λ (nm, inline)a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Auroxanthin Auroxanthin Violaxanthin Neochrome cis-Neoxanthin Neoxanthin Neochrome Neoxanthin cis-Neoxanthin Neoxanthin cis-Lutein 13- or 13 -cis-Lutein cis-Lutein 13- or 13 -cis-Lutein All-trans-lutein All-trans-zeaxanthin 9- or 9 -cis-Lutein ␤-apo-8 -Carotenal (IS) 9- or 9 -cis-Lutein cis-Lutein cis-Lutein cis-Lutein cis-␤-Carotene All-trans-␤-Cryptoxanthin 13- or 13 -cis-␣-Carotene 13- or 13 -cis-␣-Carotene 13- or 13 -cis-␤-Carotene 9- or 9 -cis-␤-Carotene All-trans-␣-Carotene 9- or 9 -cis-␣-Carotene 9- or 9 -cis-␣-Carotene All-trans-␤-Carotene 9- or 9 -cis-␤-Carotene

5.15 5.82 7.00 8.69 10.00 10.86 12.47 12.93 14.78 16.02 17.09 21.90 22.70 24.59 29.13 33.02 33.49 34.25 35.24 36.52 37.20 37.77 39.01 40.20 41.35 42.28 43.99 44.63 45.19 45.67 46.78 47.29 48.02

– 338 326 326 326 332 332 326 326 326 332 332 332 326 332 344 332 – 332 344 344 344 344 344 332 332 344 344 344 344 344 350 344

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 a b c d e f g h i j

380 374 410 398 404 416 398 418 411 416 416 415 422 416 423 427 416 – 421 423 421 421 427 426 416 416 422 422 426 421 421 430 428

398 398 434 422 428 440 422 446 434 440 440 440 446 440 446 452 440 464 446 446 446 446 452 452 446 440 446 452 449 446 446 458 452

λ (nm, reported)

422 422 465 440 458 470 446 470 464 470 464 464 470 464 470 476 470 – – – 476 476 476 478 470 464 476 476 476 470 470 482 476

– – – – – – – – – – – – – – – – –

– – – – 411 415 – 416 411 415 – 419 423 419 426 425 420

398 398 435 417 429 438 417 440 429 438 434 439 446 439 448 454 442

422b 422b 465c 441c 459c 468d 441c 464b 459c 468d 464b 465e 470b 465e 472f 478f 467e

– – – – – – 332 332 – – – – – – –

420 423 423 423 429 428 416 416 421 – 421 421 421 426 –

442 440 446 446 447 454 438 438 443 452 444 442 442 454 452

467e 470b 470b 470b 465c 480g 465h 465h 470i 477b 472e 468e 468e 478f 477b

Epoxide test hypsochromic shift

Q-ratioj found

Q-ratio reported

380 374 380 398 398 398 398 398 398 398

0.45 0.52 0.20 0.28 0.28 0.27 0.22 0.18 0.11 0.19 0.16 0.44 0.32 0.56 0.12 0.11 0.13

– – 0.22c 0.25c 0.28c – 0.25c – 0.28c – 0.19b 0.38e 0.34b 0.38e 0.06f 0.06f 0.12e

0.16 0.26 0.34 0.39 0.45 0.40 0.52 0.40 0.46 0.26 0.10 0.23 0.24 0.12 0.20

0.12e 0.20b 0.34b 0.34b 0.44c – 0.47h 0.41h 0.43i 0.13b 0.09e 0.12e 0.12e 0.08f 0.13b

398 398 398 422 422 422 422 422 422 422

422 422 422 440 441 440 446 440 446 446

A gradient mobile phase of methanol–acetonitrile–water (84:14:2, v/v/v) and methylene chloride (from 100:0, v/v to 45:55, v/v) was used. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0, v/v to 70:30, v/v) was used by Liu et al. [8]. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0 to 70:30, v/v) was used by Chen et al. [7]. A mobile phase of acetone and water (from 70:30, v/v to 100:0, v/v) was used by Razungles et al. [17]. A mobile phase of methanol–methylene chloride (99:1, v/v) was used by Chen et al. [18]. A mobile phase of methanol–methylene chloride–2-propanol (89:1:10, v/v/v) was used by Tai and Chen [19]. A mobile phase of acetonitrile–methanol–methylene chloride–hexane (from 85:10:2.5:2.5, v/v to 45:10:22.5:22.5, v/v) was used by Khachik et al. [20]. A mobile phase of methanol–methyl-tert-butyl ether (75:25, v/v) was used by Bohm et al. [21]. A mobile phase of 1-butanol–acetonitrile (30:70, v/v) and methylene chloride (from 99:1 to 90:10, v/v) was used by Lin and Chen [22]. Q-ratio is defined as the height ratio of the cis peak to the main absorption peak.

rical isomers of ␣-carotene by NMR spectroscopy [24]. Peaks 11, 13, 19–22 were identified as cis isomers, however, no cis position was assigned due to the absence of any reported Qratios. The chromatographic data and spectral characteristics of all the peaks in Fig. 2 were compared with those of unknown peaks in Fig. 1. In addition, illuminated standards solutions were also added to the sample extract separately and injected into HPLC for co-chromatography. Thus, the peaks 11–32 in Fig. 1 were indeed found to be same as those identified from Fig. 2. Accordingly, peaks 15, 16, 23, 28 and 31 were positively identified as all-trans forms of lutein, zeaxanthin, ␤-cryptoxanthin, ␣-carotene and ␤-carotene, respectively, while the other peaks (peaks 11–14, 17–22, 24–27, 29, 30 and 32) were tentatively identified as their cis isomers based on the criteria described

above. Table 2 presents the spectral data and the Q-ratios on comparison with those of the reported ones for the complete array of carotenoids including cis isomers identified in the chlorella tablet. The content of various carotenoids in chlorella tablet is shown in Table 1. The content of the carotenoids followed the order: all-trans-lutein and its cis isomers (93.1%)  all-trans-␤carotene and its cis isomers (2.6%) ≈ all-trans-␣-carotene and its cis isomers (2.6%) > zeaxanthin (1.3%) > epoxy-containing compounds (0.2%) ≈ ␤-cryptoxanthin (0.2%). Obviously, the xanthophyll lutein, being one of the prominent structural and functional components essential for cellular survival, is present in very large amount when compared to the other xanthophylls and carotenes. Several epidemiological studies reported that

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

198

Fig. 2. HPLC chromatogram of carotenoid standards after illumination at 25 ◦ C for 24 h. Chromatographic conditions are described in the text. (A) Illuminated lutein standard—peaks: (11) cis-lutein; (12) 13- or 13 -cis-lutein; (13) cis-lutein; (14) 13- or 13 -cis-lutein; (15) all-trans-lutein; (17) 9- or 9 -cis-lutein; (18) 9or 9 -cis-lutein; (19) cis-lutein; (20) cis-lutein; (21) cis-lutein. (B) Illuminated ␣-carotene standard—peaks: (24) 13- or 13 -cis-␣-carotene; (25) 13- or 13 cis-␣-carotene; (28) all-trans-␣-carotene; (29) 9- or 9 -cis-␣-carotene; (30) 9or 9 -cis-␣-carotene. (C) Illuminated ␤-carotene standard—peaks: (22) cis-␤carotene; (26) 13- or 13 cis-␤-carotene; (27) 9- or 9 -cis-␤-carotene; (31) alltrans-␤-carotene; (32) 9- or 9 -cis-␤-carotene.

lutein is not only effective against age-related macular degeneration (AMD) [25] but also effective in delaying chronic diseases [26] and progression of early atherosclerosis [27]. The cis isomers of lutein make up about 17.0% of total lutein content, but there are no specific reports on biological activity of cis isomers of lutein. Nevertheless, lesser amounts of 9-cis-lutein, 9 -cislutein, 13-cis-lutein and 13 -cis-lutein were found in the human

retina [28]. Zeaxanthin is the next xanthophyll present in single largest amount (1.32%) that can be supplemented along with lutein as neutraceuticals against AMD. However, its lesser concentration when compared to lutein suggests that the microalga C. pyrenoidosa is not subjected to any external stress during its growth, as zeaxanthin content in microalgae is regulated by light irradiance [1]. Other xanthophylls, though contributes only 0.2% to the total content, are present in a wide variety. Carotenes (␣-carotene and ␤-carotene) and their cis isomers constitute about 5.2% of total carotenoids. Normally, the separation of ␣-carotene and its isomers is pursued with less importance. However, in the present study all-trans-␣-carotene with most of its predominant geometrical isomers were adequately resolved. Unlike ␤-carotene, ␣-carotene is often overlooked for possessing only one half of the provitamin A activity. However, ␣-carotene is more effective as an antioxidant and more potent anti-carcinogen than ␤-carotene [29]. In this study, all-trans-␣carotene (2465.8 ␮g/g) is present in higher amounts when compared to all-trans-␤-carotene (2155.0 ␮g/g), while the reverse is true with regard to their cis isomers (Table 2). Though the cis isomers of ␣-carotene are less efficiently converted to Vitamin A [24], they are present in several biological samples [30]. On the other hand, 9-cis-␤-carotene was reported to possess higher antioxidant potency when compared to its all-trans-isomer [31]. Thus, ␤-carotene and ␣-carotene along with ␤-cryptoxanthin (0.2%) present in chlorella tablet can act as vitamin A precursors and antioxidants. The absence of any secondary carotenoids like astaxanthin, canthaxanthin and echinenone further confirms that the microalga (C. pyrenoidosa) is not subjected to any environmental stimuli during its growth to undergo carotenogenesis [1]. Based on the results it may be concluded that the range of carotenoids present in chlorella tablet can have synergistic effect on promoting overall human health. The HPLC method developed can be used as a concise method to determine an array of carotenoids in microalgae and microalgae-derived products. Further research is warranted to extend the study for determination of carotenoids content in several representative commercial

Table 3 Identification data for all-trans and cis forms of lutein after illumination of all-trans-lutein standard at 25 ◦ C for 24 h Peak no.

Compound

Retention time (min)

ka

Peak purity (%)

λ (nm, in-line)b

11 12 13 14 15 17 18 19 20 21

cis-Lutein 13- or 13 -cis-Lutein cis-Lutein 13- or 13 -cis-Lutein All-trans-lutein 9- or 9 -cis-Lutein 9- or 9 -cis-Lutein cis-Lutein cis-Lutein cis-Lutein

17.28 21.74 22.81 24.48 28.88 33.21 34.72 35.93 37.03 37.73

4.18 5.61 6.02 6.52 7.83 9.04 9.47 9.75 10.02 10.17

99.6 95.2 90.3 97.8 94.7 95.0 92.9 99.1 99.9 99.9

332 332 338 332 332 332 332 344 344 344

a b c d e f

410 415 422 416 422 416 421 422 416 421

λ (nm, reported) 434 440 446 440 446 440 446 446 446 446

458 464 470 464 470 470 470 476 470 470

339 – – – – – – – – –

405 419 423 419 426 420 420 422 423 423

429 439 446 439 448 442 442 440 446 446

447c 465d 470e 465d 472f 467d 467d 470e 470e 470e

k : retention factor. A gradient mobile phase of methanol–acetonitrile–water (84:14:2, v/v/v) and methylene chloride (from 100:0, v/v to 45:55, v/v) was used. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0 to 70:30, v/v) was used by Chen et al. [7]. A mobile phase of methanol–methylene chloride (99:1, v/v) was used by Chen et al. [18]. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0, v/v to 70:30, v/v) was used by Liu et al. [8]. A mobile phase of methanol–methylene chloride–2-propanol (89:1:10, v/v/v) was used by Tai and Chen [19].

Q-ratio found

Q-ratio reported

0.31 0.41 0.28 0.39 0.09 0.13 0.12 0.18 0.37 0.39

0.35c 0.38d 0.34e 0.38d 0.06f 0.12d 0.12d 0.20e 0.34e 0.34e

B.S. Inbaraj et al. / J. Chromatogr. A 1102 (2006) 193–199

199

Table 4 Identification data for all-trans and cis forms of ␣-carotene after illumination of all-trans-␣-carotene standard at 25 ◦ C for 24 h Peak no.

Compound

Retention time (min)

ka

Peak purity (%)

λ (nm, in-line)b

24 25 28 29 30

13- or 13 -cis-␣-Carotene 13- or 13 -cis-␣-Carotene All-trans-␣-carotene 9- or 9 -cis-␣-Carotene 9- or 9 -cis-␣-Carotene

41.33 42.31 45.19 45.79 46.99

10.67 10.93 11.73 11.88 12.17

94.5 97.5 90.7 94.9 97.9

332 332 344 344 344

a b c d

416 417 426 421 421

442 442 448 446 446

λ (nm, reported) 470 470 476 470 470

332 332 – – –

416 416 421 421 421

438 438 444 442 442

465c 465c 472d 468d 468d

Q-ratio found

Q-ratio reported

0.44 0.38 0.09 0.11 0.10

0.47c 0.41c 0.09d 0.12d 0.12d

k : retention factor. A gradient mobile phase of methanol–acetonitrile–water (84:14:2, v/v/v) and methylene chloride (from 100:0, v/v to 45:55, v/v) was used. A mobile phase of methanol–methyl-tert-butyl ether (75:25, v/v) was used by Bohm et al. [21]. A mobile phase of methanol–methylene chloride (99:1, v/v) was used by Chen et al. [18].

Table 5 Identification data for all-trans and cis forms of ␤-carotene after illumination of all-trans-␤-carotene standard at 25 ◦ C for 24 h Peak no.

Compound

Retention time (min)

ka

Peak purity (%)

λ (nm, in-line)b

22 26 27 31 32

cis-␤-Carotene 13- or 13 -cis-␤-Carotene 9- or 9 -cis-␤-Carotene All-trans-␤-carotene 9- or 9 -cis-␤-Carotene

39.99 44.13 45.27 47.47 48.31

10.36 11.49 11.79 12.33 12.52

99.9 96.3 99.9 90.3 96.2

344 344 344 354 344

a b c d e

410 423 423 427 425

446 446 452 458 452

λ (nm, reported) 470 476 476 482 476

– 339 – – –

– 420 – – –

442 445 452 458 452

471c 470d 477e 482e 477e

Q-ratio found

Q-ratio reported

0.39 0.43 0.14 0.09 0.16

0.44c 0.37d 0.13e 0.07e 0.13e

k : retention factor. A gradient mobile phase of methanol–acetonitrile–water (84:14:2, v/v/v) and methylene chloride (from 100:0, v/v to 45:55, v/v) was used. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0 to 70:30, v/v) was used by Chen et al. [7]. A mobile phase of methanol–methyl-tert-butyl ether (75:25, v/v) was used by Bohm et al. [21]. A mobile phase of methanol–2-propanol (99:1, v/v) and methylene chloride (from 100:0, v/v to 70:30, v/v) was used by Liu et al. [8].

microalgae-derived functional food samples. Furthermore, stability of carotenoids under different processing conditions and kinetics of isomerization and degradation of carotenoids in such microalgae-derived functional foods will be investigated. References [1] E. Jin, J.E.W. Polle, H.K. Lee, S.M. Hyun, M. Chang, J. Microbiol. Biotechnol. 13 (2003) 165. [2] T. Goodwin, G. Britton, in: T.W. Goodwin (Ed.), Plant Pigments, Academic Press, Padstow, Cornwall, 1988, p. 61. [3] M. Rise, E. Cohen, M. Vishkautsan, M. Cojocaru, H.E. Gottlieb, S.M. Arad, J. Plant Physiol. 144 (1994) 287. [4] H.D. Sesso, J.E. Buring, E.P. Norkus, J.M. Gaziano, Am. J. Clin. Nutr. 79 (2004) 47. [5] A.V. Rao, S. Agarwal, Nutr. Res. 19 (1999) 305. [6] H. Tapiero, D.M. Townsend, K.D. Tew, Biomed. Pharmacother. 58 (2004) 100. [7] J.P. Chen, C.Y. Tai, B.H. Chen, J. Chromatogr. A 1054 (2004) 261. [8] H.L. Liu, T.H. Kao, B.H. Chen, Chromatographia 60 (2004) 411. [9] L. Gouveia, V. Veloso, A. Reis, H. Fernandes, J. Novais, J. Empis, Biores. Technol. 57 (1996) 157. ´ Vargas, J. Rivas, M.G. [10] J.A. Del Camp, J. Moreno, H. Rodr´ıguez, M.A. Guerrero, J. Biotechnol. 76 (2000) 51. [11] L.C. Sander, K.E. Sharpless, M. Pursch, J. Chromatogr. A 880 (2000) 189. [12] B.H. Chen, S.H. Yang, L.H. Han, J. Chromatogr. 543 (1991) 147. [13] SAS/STAT Guide for Personal Computers, version 8.1, SAS Instruments, Cary, NC, 2003.

[14] International Conference on Harmonization (ICH), Guideline on the Validation of Analysis Procedures: Methodology Q2B, 1996. [15] M.T. Lee, B.H. Chen, Chromatographia 54 (2001) 613. [16] J.W. Dolan, LC–GC 5 (1987) 1030. [17] A.J. Razungles, I. Babic, J.C. Sapis, C.L. Bayonove, J. Agric. Food Chem. 44 (1996) 3821. [18] B.H. Chen, H.Y. Peng, H.E. Chen, J. Agric. Food Chem. 43 (1995) 1912. [19] C.Y. Tai, B.H. Chen, J. Agric. Food Chem. 48 (2001) 5962. [20] F. Khachik, C.J. Spangler, J.C. Smith, L.M. Canfield, A. Steck, H. Pfander, Anal. Chem. 69 (1997) 1873. [21] V. Bohm, N.L.P. Nienaber, M.G. Ferruzzi, S.J. Schwartz, J. Agric. Food Chem. 50 (2002) 221. [22] C.H. Lin, B.H. Chen, J. Chromatogr. A 1012 (2003) 103. [23] B.H. Davies, in: T.W. Goodwin (Ed.), Chemistry and Biochemistry of Plant Pigments, second ed., Academic Press, London, 1976. [24] C. Emenhiser, G. Englert, L.C. Sander, B. Ludwig, S.J. Schwartz, J. Chromatogr. A 719 (1996) 333. [25] N.I. Krinsky, J.T. Landrum, R.A. Bone, Annu. Rev. Nutr. 23 (2003) 171. [26] J.A. Mares-Perlman, A.E. Millen, T.L. Ficek, S.E. Hankinson, J. Nutr. 132 (2002) 5185. [27] J.H. Dwyer, M. Navab, K.M. Dwyer, K. Hassan, P. Sun, A. Shircore, S. Hama-Levy, G. Hough, X. Wang, T. Drake, N.B. Merz, A.M. Fogelman, Circulation 103 (2001) 2922. [28] F. Khachik, P.S. Bernstein, D.L. Garland, Invest. Ophthalmol. Vis. Sci. 38 (1997) 1802. [29] M. Murakoshi, H. Nishino, Y. Satomi, J. Takayasu, T. Hasegawa, H. Tokuda, A. Iwashima, J. Okuzumi, H. Okabe, H. Kitano, R. Iwasaki, Cancer Res. 52 (1992) 6583. [30] F. Khachik, C.J. Spangler, J.C. Smith, Anal. Chem. 69 (1997) 1873. [31] G. Levin, S. Mokady, Free Radic. Biol. Med. 17 (1994) 77.