Food Anal.

Methods
DOI 10.1007/s12161-016-0553-z

High-Performance Liquid Chromatography with Fluorescence

Detection for Simultaneous Analysis of Retinoids (Retinyl
Palmitate, Retinyl Acetate, and Free Retinol) and α-, β-, γ-,
and δ-Tocopherols in Foods
Mami Ishimaru 1 & Masato Haraoka 1 & Hideo Hatate 1 & Ryusuke Tanaka 1

Received: 12 February 2016 / Accepted: 25 May 2016

# Springer Science+Business Media New York 2016

Abstract A method based on high-performance liquid chro- Introduction

matography with fluorescence detection for the simultaneous
analysis of retinoids (vitamin A) and tocopherols (vitamin E) Retinoids (also known as vitamin A) and tocopherols (also
was developed. This method consists of an isocratic solution known as vitamin E) are lipid-soluble vitamins that are found
using hexane/ethyl acetate (85:15, v/v) as the mobile phase in various foods, including seafood. The term Bretinoid^ refers
and fluorescence detection using a time program that sets the to any compound possessing the biological activity of retinol
excitation (Ex) and emission (Em) wavelengths at adequate (retinyl esters, retinol, retinal, retinoic acid, and the oxidated
elution times for retinoids (Ex 342 nm, Em 476 nm) and and conjugated forms of both retinol and retinal) (Sporn et al.
tocopherols (Ex 298 nm, Em 325 nm), respectively. The 1976), and retinoids have many functions, such as vision
separation of three retinoids (retinyl palmitate, retinyl acetate, (Palczewski 2012; Von Lintig 2012) and regulation of the
and free retinol) and four tocopherol homologs was growth and differentiation of cells (Gudas and Wagner 2011;
achieved with sufficient reproducibility and quantitative abil- Morriss-Kay and Sokolova 1996). Retinoids are found in
ity. Additionally, the necessity of saponification was consid- vegetables and fruits as carotenoid (provitamin A) or in eggs,
ered. As a result, saponification was not used in this method liver, and dairy products as retinyl ester (palmitate) (Harrison
because of the complexity of the procedure and the loss of free 2012). Furthermore, in the case of seafood, eel muscle is rich
retinol. The retinoid and tocopherol contents of various foods in retinyl palmitate.
were evaluated using the developed method. Our method The tocopherols include four compounds: α-, β-, γ-, and
could evaluate the retinoid and tocopherol contents of fish δ-tocopherols (IUPAC-IUB Joint Commission on
(eel, Anguilla japonica, and amberjack, Seriola dumerili) Biochemical Nomenclature 1982). These compounds play
muscle and liver, roasted soybean (Glycine max) flour, and many important roles in animal reproduction (Elvehjem
Japanese torreya seed (Torreya nucifera). Additionally, our et al. 1944), antioxidant activity (Brownlee et al. 1977), and
method could be applied to the determination of retinoids prevention of chronic diseases, especially those believed to
and tocopherols not only in foods but also in supplements have an oxidative stress component, such as cardiovascular
and cosmetics. diseases, atherosclerosis, and cancer (Rimm et al. 1993;
Stampfer et al. 1993). α-Tocopherol is the most abundant
form in nature and has the highest biological activity.
Keywords HPLC . Retinoid . Tocopherol . Fluorescence Otherwise, γ-tocopherol is suspected to scavenge electrophil-
detection . Food ic mutagens and prevent cardiovascular diseases by lowering
blood pressure (Wechter et al. 1996). α-Tocopherol is found in
fish meat and vegetable oil, and γ-tocopherol is found in seeds
* Ryusuke Tanaka
[email protected]
and nuts such as soybean.
Retinoids and tocopherols are analyzed using high-
1
Department of Marine Biology and Environmental Sciences, Faculty
performance liquid chromatography (HPLC) with UV-Vis or
of Agriculture, University of Miyazaki, fluorescence detection from a saponified sample and are usu-
Gakuen-Kibanadai-nishi-1-1, Miyazaki 889-2192, Japan ally determined by different HPLC methods (Hosotani and
 Food Anal. Methods

Kitagawa 2003; Indyk 1988; Panfili et al. 2003) or same 1000 μg mL−1 of the respective compound. One milliliter of
HPLC method for simultaneous analysis of them (Turner each of the seven stock solutions was mixed with 3 mL of
and Burri 2012). Saponification is effective for hydrolyzing n-hexane/ethyl acetate (85:15, v/v); thus, each mixed solution
any ester linkage and removing the overwhelming proportions contained 100 μg mL−1 of compound. These were used as the
of triglyceride lipid but require complicated procedure for standard solutions.
saponification such as 30-min heating and twice partition for
collection of unsaponified matter. Furthermore, their methods Materials
could not separate retinoids and tocopherols simultaneously
so determination of both retinoids and tocopherols by using Amberjack (Seriola dumerili) (n = 6) and fish feed were gifted
their methods require a lot of equipment such as an HPLC from Miyazaki Prefectural Fisheries Research Institute. Each
system and time. To resolve these problems, some methods amberjack was made to bleed by cutting its gills, and its inter-
which are conducted without saponification or which separate nal organs were removed. The six dressed fishes were cut into
retinoids and tocopherols simultaneously were reported. Casal loins and the dorsal muscles were mixed. The experimental
et al. (2001) reported a method for analysis of retinol, β-car- protocol was approved by the Institutional Animal Care and
otene, and α-tocopherol in adipose tissue simultaneously Use Committee, University of Miyazaki. Roasted soybean
without saponification, and Amaral et al. (2005) reported an (Glycine max) flour, eel (Anguilla japonica), Japanese torreya
analytical method of eight tocols (four tocopherols and four seed (Torreya nucifera), and supplement capsules were pur-
tocotrienols) in hazelnuts without saponification. Krčmova chased at a local store. One eel was treated the same way as
et al. (2009) separated free retinol, 13-cis and all-trans retinoic amberjack. Japanese torreya seed was cracked and shells were
acid, α-tocopherol, and γ-tocopherol in human serum simul- removed. And then, some pieces of seed were cut fine and
taneously by using HPLC without saponification. McMahon mixed. Supplement capsules contained lamprey (Lethenteron
et al. (2013) reported a method for analysis of retinyl palmi- japonicum) oil, soybean oil, and olive (Olea europaea) oil.
tate, retinyl acetate, and α-tocopherol without saponification. Toning lotion that contained retinyl acetate was purchased at
However, there are no methods that can simultaneously sepa- a local store.
rate and determine the four homologs of tocopherols and the
three retinoids (retinyl palmitate, retinyl acetate, and free ret-
inol). Furthermore, many methods of simultaneous analysis HPLC System
use a UV-Vis detector, which can be difficult because the
sensitivity and selectively of this detector are relatively low. The retinoids and tocopherols were analyzed using an HPLC
In this study, we developed an HPLC method for the si- system that included a PU-2080 plus pump, an 860-CO col-
multaneous analysis of retinoids and tocopherols in foods that umn oven (40 °C) (JASCO Corporation, Tokyo, Japan), an
does not require a complicated saponification procedure and L-7480 fluorescence detector (Hitachi High-Technologies
uses an isocratic system with a fluorescence detector; the Corporation, Tokyo, Japan), and a 7725i Rheodyne sample
method has high sensitivity and selectivity. injector connected to a 10-μL sample loop (Rheodyne,
Rohnert Park, CA, USA). This system was equipped with an
Inertsil NH2 column (250 mm × 4.6 mm, 5-μm particle size,
Materials and Methods GL Science, Tokyo, Japan) connected to a Guard-Pak guard
column containing μBondapak NH2 (Waters, Milford, MA,
Chemicals USA). The mobile phase was n-hexane/ethyl acetate (85:15,
v/v) and the flow rate was 1.0 mL min−1. The detector was
Ethyl acetate, n-hexane, chloroform, methanol, sodium chlo- programmed for excitation (Ex) at 342 nm and emission (Em)
ride, pyrogallol, and potassium hydroxide, all JIS special at 476 nm from 0 to 5.5 min for the detection of retinyl pal-
grade, were obtained from Wako Pure Chemical Industries, mitate and retinyl acetate, Ex at 298 nm and Em at 325 nm
Ltd. (Tokyo, Japan). Retinyl palmitate, retinyl acetate, and from 5.5 to 17.0 min for the detection of tocopherols, and Ex
retinol were obtained from Sigma-Aldrich Japan, Ltd. at 342 nm and Em at 476 nm from 17.0 to 22.0 min for the
(Tokyo, Japan). D-α-tocopherol, D-β-tocopherol, D-γ-tocoph- detection of free retinol. New samples were injected at inter-
erol, and D-δ-tocopherol were obtained from Eisai Co., Ltd. vals of 25.0 min. HPLC was conducted at room temperature.
(Tokyo, Japan).
Stock solutions of the three retinoids (retinyl palmitate, Method Validation
retinyl acetate, and free retinol) and four tocopherols (α-to-
copherol, β-tocopherol, γ-tocopherol, and δ-tocopherol) were Validation of the HPLC method included precision and spec-
prepared by dissolving the compounds in n-hexane/ethyl ace- ificity, linearity, range, limits of detection and quantification,
tate (85:15, v/v). Each of these seven solutions contained and recovery.
Food Anal. Methods

For the precision and specificity tests, analytical solutions Consideration of the Necessity of Saponification
were prepared by diluting the standard solution with n--
hexane/ethyl acetate (85:15, v/v) at 3 μg mL−1 of each com- To consider the necessity of saponification of the extracted lipids
pound. Five separate 10-μL aliquots of each analytical solu- for retinoid and tocopherol analysis, the lipid content of the eel
tion were injected and analyzed using HPLC. The retention muscle was saponified according to Matsushita et al.’s (2010)
time (tR), peak area, capacity factor (K′), resolution factor method. The eel muscle lipid was extracted using the recovery
(Rs), and separation factor (α) were estimated using the results test method. The extracted eel lipid solution was taken into two
of five replicate analyses. The parameters were defined using 15-mL glass test tubes (5 mg lipid tube−1) and evaporated using a
the following standard formulas: K′ = (tRn − t0) / t0, Rs = 2(tRn+ centrifugal concentrator. One eel lipid sample was redissolved
1 − tRn) / (Wn+1 + Wn), α = K’n+1/K’n, where tRn is the retention with 0.50 mL of n-hexane/ethyl acetate (85:15, v/v). The other
time of each peak, t0 is the retention time of the first solvent eel lipid sample was reconstituted with 0.10 mL of 1.0 % sodium
peak (t0 = 3.26 min), and Wn is the width of the peak at the chloride in water and 2.0 mL of 3.0 % pyrogallol in ethanol. This
baseline. The subscript n refers to the order of elution of the solution was saponified by heating at 70 °C for 30 min after the
retinoids and tocopherols. addition of 0.20 mL of 60 % potassium hydroxide in water. After
To test the linearity, range, limit of detection (LOD), and heating, 4.5 mL of 1.0 % sodium chloride in water and 3.0 mL of
limit of quantification (LOQ) of the method, analytical solu- n-hexane/ethyl acetate (85:15, v/v) were added to the reaction
tions were prepared by diluting standard solutions in grada- mixture, which was centrifuged at 1500g for 10 min. The upper
tions from 5-fold to 10,000-fold with n-hexane/ethyl acetate layer was transferred into another tube and the lower layer was
(85:15, v/v). A 10-μL aliquot of each solution was injected mixed again with 3.0 mL of n-hexane/ethyl acetate (85:15, v/v)
and analyzed using HPLC. The linearity, range, LOD, and and centrifuged at 1500g for 10 min. Both upper layers were
LOQ were estimated using the results of these analyses. combined, and the solvents were removed using a centrifugal
LOD was calculated as the concentration at which the retinoid concentrator. The residue was redissolved in 0.50 mL of n-hex-
and tocopherol peaks could be detected without any baseline ane/ethyl acetate (85:15, v/v). The samples with and without
noise disturbance (>3 times the baseline noise). LOQ was saponification were used for the determination of retinoid and
calculated as the concentration at which the responses of the tocopherol contents using HPLC. All measurements were per-
analytes were >10 times the baseline noise. formed in triplicate by using one sample. The results of all mea-
For the recovery tests, fortified samples were prepared by surements are expressed as means ± the standard deviation (SD).
adding a 1- or 2-mL aliquot of mixed standard solution (0, 50,
and 100 μg mL−1 of each compound) to amberjack muscle or The Retinoid and Tocopherol Contents of Foods
to roasted soybean flour. First, their lipids were extracted ac-
cording to Folch et al.’s (1957) method. Amberjack muscle The retinoid and tocopherol contents of various foods (amber-
(5 g each) and roasted soybean flour (2 g each) were weighed jack, fish feed, eel, roasted soybean flour, Japanese torreya seed,
into separate 50-mL glass centrifuge tubes, and then, a 1-mL and supplement capsule) were analyzed using this HPLC meth-
(for amberjack) and a 2-mL (for roasted soybean flour) aliquot od. Additionally, to confirm that our developed method can de-
of standard solution were added to each. Then, 30 mL of termine the retinoid contents in a non-food substance, a toning
chloroform/methanol (2:1, v/v) was added to each tube and lotion that contained retinyl acetate was analyzed. Lipids from
these samples were homogenized. Then, fragments of these eel muscle and liver, amberjack muscle, roasted soybean flour,
samples remaining on the blade were rinsed in the test tube Japanese torreya seed, and fish feed were extracted by the recov-
with 4 mL (for amberjack) and 7.5 mL (for roasted soybean ery test method. Lipid from the supplement was obtained by
flour) of distilled water, and the samples were centrifuged for cutting the supplement and dissolving it with chloroform/
10 min at 1500 g. The upper layer was removed by an aspi- methanol (2:1, v/v). The lipid of the lotion was extracted by the
rator and the remaining lower layer was filtered with a cotton method below. One milliliter of the lotion was added to 4 mL of
plug. The solution was evaporated with a rotary evaporator. chloroform/methanol (2:1, v/v) and centrifuged at 1500g for
The residues were redissolved and adjusted to 25 mL (for 10 min. The upper layer was removed by an aspirator and the
amberjack) and 50 mL (for roasted soybean flour) by the remaining lower layer was adjusted to 10 mL by the addition of
addition of chloroform/methanol (2:1, v/v). The extracted lipid chloroform/methanol (2:1, v/v). The extracted lipid solutions of
solutions of amberjack and roasted soybean flour (5 mg lipid these samples were taken into separate 15-mL glass test tubes
tube−1) were evaporated using a centrifugal concentrator and (5 mg lipid tube−1) and evaporated using a centrifugal concen-
redissolved in 0.50 mL of n-hexane/ethyl acetate (85:15, v/v); trator. The residues were redissolved with 0.50 mL of n-hexane/
the retinoid and tocopherol contents were determined using ethyl acetate (85:15, v/v) and used for the determination of the
HPLC. Recovery indicates the medium recoveries and the retinoid and tocopherol contents by HPLC. All measurements
relative standard deviation (RSD) for the three analysis. were performed in triplicate by using one sample. The results of
RSD were estimated using the analysis results. all measurements are expressed as means ± SD.
 Food Anal. Methods

40000
(a) separated free retinol, β-carotene, and α-tocopherol using
0- 5.5 min 5.5- 17.0 min 17.0- 22.0 min
normal-phase HPLC with a silica gel column; however, this
Detector response (µV) 30000 Ex 342 nm Ex 298 nm Ex 342 nm
Em 476 nm Em 325 nm Em 476 nm
method used a gradient system and required two detectors (a
2
20000
diode array detector for free retinol and β-carotene and a fluo-
1 3
rescence detector for α-tocopherol). Also, Nimalaratne et al.
10000 7 (2014) separated lipid-soluble vitamins (retinoid, tocopherols,
4 5 6 and vitamin D3) by using reversed-phase fast liquid chroma-
0
tography with a C30 column; however, their method could not
0 5 10 15 20 separate β-tocopherol and γ-tocopherol sufficiently. Kamal-
Retention time (min)
Eldin et al. (2000) examined many types of HPLC for the
separation of four tocopherol homologs. As a result, four to-
60000 (b) copherols were successfully separated using an NH2 column.
Detector response (µV)

50000 2 Following this, we attempted to separate four tocopherols and

40000 3 4 retinoids by using an NH2 column. A hexane and ethyl acetate
6 7
1 5
30000 mixed solvent, which is used for the extraction of tocopherols
20000
from the samples, was used as the mobile phase; consequent-
ly, three retinoids and four tocopherols were separated. Next,
10000
the proportions of the solvent were examined. As a result,
0
0 5 10 15 20
those compounds were separated sufficiently using hexane/
Retention time (min) ethyl acetate (85:15, v/v).
Fig. 1 HPLC chromatogram obtained with mixed standard solution (a) To confirm the precision and specificity of the method de-
and each standard sample (b) of retinoids and tocopherols. The numbers on veloped here, the tR, peak area, K′, Rs, and α were investigat-
the chromatogram represent the following compounds: 1, retinyl palmitate;
ed; the results are displayed in Table 1. The RSDs of the tRs for
2, retinyl acetate; 3, α-tocopherol; 4, β-tocopherol; 5, γ-tocopherol; 6, δ-
tocopherol; 7, retinol. The arrows in this figure show the time program of all retinoids and tocopherols were less than 0.40 %, and the
the fluorescence detector. The detector was programmed for excitation (Ex) RSDs of the peak areas for all retinoids and tocopherols were
at 342 nm and emission (Em) at 476 nm from 0 to 5.5 min for the detection less than 4.00 %. The separation of each retinoid and tocoph-
of retinoids, Ex at 298 nm and Em at 325 nm from 5.5 to 17.0 min for the
erol exhibited good specificity. The RSDs of the peak areas
detection of tocopherols, and Ex at 342 nm and Em at 476 nm from 17.0 to
22.0 min for the detection of retinol obtained using the HPLC method developed here were less
than 4.00 % for each compound, which suggests that the
Results and Discussion method exhibits high reproducibility. The linearity, range of
detection, LOD, and LOQ were investigated; the results are
Method Validation shown in Table 2. The correlation factor for each of the reti-
noids and tocopherols was greater than 0.998. This suggests
A typical chromatogram of mixed retinoids and tocopherols is that the method developed here exhibits good linearity. The
shown in Fig. 1a, and that of each retinoids and tocopherols is linear range was from 0.05 to 15 μg mL−1 for retinyl palmi-
shown in Fig. 1b. tate, retinyl acetate, and α-tocopherol; from 0.10 to
Many methods for the simultaneous analysis of retinoids 15 μg mL −1 for β- and γ-tocopherol; from 0.20 to
and tocopherols have been reported. Casal et al. (2001) 10 μg mL−1 for δ-tocopherol; and from 0.20 to 15 μg mL−1

Table 1 Result from evaluation

of the precision and specificity of tR (min) tR-RSD (%) AREA AREA-RSD (%) K′ Rs α
the LC methoda
Retinyl palmitate 3.368 0.271 329,422 3.227 0.035 – –
Retinyl acetate 4.407 0.218 458,812 3.208 0.354 3.919 10.347
α-Tocopherol 6.334 0.240 368,532 3.060 0.946 6.504 2.674
β-Tocopherol 8.699 0.283 154,161 2.587 1.672 6.652 1.768
γ-Tocopherol 10.247 0.269 175,638 3.295 2.147 3.772 1.284
δ-Tocopherol 13.796 0.305 198,506 3.848 3.238 7.197 1.508
Retinol 17.638 0.394 438,218 3.202 4.418 4.702 1.364
a
Chromatographic parameters of retinoid and tocopherols with fluorescence HPLC (Detection: fluorescence at
342 nm (excitation, Ex)/ 476 nm (emission, Em) for retinoid, and at 298 nm (Ex)/325 nm (Em) for tocopherols),
mixed retinoid and tocopherol concentration, 30 ng each per 10-μL injection (n = 5). Retention time (tR), capacity
factor (K′), resolution factor (Rs), and separation factor (α)
Food Anal. Methods

Table 2 Result from evaluation of the linearity and range of the LC method

Regression equation Correlation Linear range (μg mL−1) LOD (μg mL−1) LOQ (μg mL−1)
factor

Retinyl palmitate ya = 1.00 × 10−5xb − 1.93 × 10−1 0.9994 0.05–15.00 0.1500 0.4995
Retinyl acetate y = 5.82 × 10−6x + 1.66 × 10−1 0.9994 0.05–15.00 0.1500 0.4995
α-Tocopherol y = 8.79 × 10−6x + 2.43 × 10−1 0.9992 0.05–15.00 0.1500 0.4995
β-Tocopherol y = 1.87 × 10−5x-2.14 × 10−1 0.9994 0.10–15.00 0.1500 0.4995
γ-Tocopherol y = 1.65 × 10−5x + 2.96 × 10−1 0.9990 0.10–15.00 0.1500 0.4995
δ-Tocopherol y = 1.35 × 10−5x + 2.07 × 10−1 0.9985 0.20–10.00 0.1500 0.4995
Retinol y = 6.11 × 10−6x + 3.83 × 10−1 0.9990 0.20–15.00 0.3000 0.9990
a
y = Mass concentration (μg mL−1 )
b
x = HPLC Peak area

for free retinol. The LODs and LOQs for the various retinoids 10.60 ± 0.13 mg 100 g−1 sample of α-tocopherol. The retinoid
and tocopherols were found to be from 0.1500 to 0.3000 and and tocopherol recoveries from the saponified sample, based on
from 0.4995 to 0.9990 μg mL−1, respectively. The mean re- molecular weight, were 87.61 % of the retinoids (from retinyl
covery for each of the retinoids and tocopherols was investi- palmitate to free retinol) and 100.61 % of α-tocopherol (data not
gated; the results are shown in Table 3. The average recoveries shown). This suggests that the retinoid content decreased by
of all the retinoids and tocopherols were 82.91 (12.03 %) to about 10 % due to the saponification treatment.
128.93 % (17.0 %) for amberjack and 100.72 (7.11 %) to Retinol is easily oxidized by heat and UV irradiation because
137.25 % (7.69 %) for roasted soybean flour. This indicates of its chemical structure, which contains many double bonds
that our developed method has sufficient quantitative ability. (Ihara et al. 1999). In this study, the approximately 10 % loss
of retinol recovery might be caused by the decomposition of
Consideration of the Necessity of Saponification retinol that was exposed to the heat process during saponifica-
tion. In this study, we choose not to saponify the extracted lipids
To consider the necessity of saponification of the extracted lipids for retinoid analysis because of the complexity of the saponifi-
for retinoid and tocopherol analysis, eel muscle lipid was treated cation procedure and the loss of retinol during saponification.
with and without saponification. Eel muscle contains a large
amount of retinyl palmitate. Typical chromatograms of the sam- Retinoid and Tocopherol Contents of Foods
ples are shown in Fig. 2a, b, respectively, and the retinoid and
tocopherol contents are shown in Table 4. The eel muscle lipid The retinoid and tocopherol contents of foods were investigat-
without saponification contained 4.43 ± 0.03 mg 100 g−1 sample ed using the method developed here (Table 4). Typical chro-
of retinyl palmitate and 10.54 ± 0.21 mg 100 g−1 sample of α- matograms of amberjack and fish feed are shown in Fig. 3a, b
tocopherol. The saponified eel muscle lipid contained and typical chromatograms of roasted soybean flour and lo-
2.12 ± 0.03 mg 100 g−1 sample of retinyl palmitate and tion are shown in Fig.4a, b, respectively.

Table 3 Result from evaluation of the recovery of the LC methoda

Recovery (%) Amberjack Roasted soybean flour

Add solution (μg mL−1) 50 100 Average 50 100 Average

Retinyl palmitate 99.80 (1.49) 78.09 (5.12) 88.94 (13.7) 108.98 (1.46) 98.65 (4.36) 103.82 (6.13)
Retinyl acetate 91.53 (2.14) 74.28 (6.23) 82.91 (12.03) 104.35 (1.39) 96.53 (4.76) 100.44 (5.23)
α-Tocopherol 147.77 (5.47) 110.09 (8.29) 128.93 (17.09) 145.23 (2.54) 129.27 (6.64) 137.25 (7.69)
β-Tocopherol 125.66 (5.43) 106.02 (7.12) 115.84 (10.82) 129.16 (0.85) 124.25 (10.99) 126.70 (7.16)
γ-Tocopherol 126.16 (3.19) 106.07 (6.06) 116.12 (10.33) 146.49 (7.36) 123.77 (12.61) 135.13 (12.79)
δ-Tocopherol 123.35 (2.22) 106.36 (5.79) 114.85 (8.91) 130.74 (47.03) 119.32 (10.66) 125.03 (32.16)
Retinol 99.44 (1.97) 79.99 (7.31) 89.72 (12.65) 105.82 (3.00) 95.62 (6.61) 100.72 (7.11)

Recovery data indicates the means and (RSD%) for the three analysis
a
One or 2-mL aliquot of mixed retinoid and tocopherols standard solutions (50 and 100 μg mL−1 of each compounds) was added to 5 g of amberjack
meat or 2 g of roasted soybean flour, respectively, which was then mixed and treated as described in the BMaterial and Methods^ section
 Food Anal. Methods

60000 40000
2 (a) (a)
50000
30000
Detector response (µV)

Detector response (µV)

40000

30000 20000
3
1
20000
10000
10000
2 5 6
0 0
0 5 10 15 20 0 5 10 15 20
Retention time (min) Retention time (min)

25000
60000 (b)
2
(b)
50000 20000

Detector response (µV)

Detector response (µV)

40000
15000

30000
10000

20000
1
5000
2
10000 6
3 5
4
0
0 0 5 10 15 20
0 5 10 15 20
Retention time (min)
Retention time (min)
Fig. 3 HPLC chromatograms of retinoids and tocopherols in the lipids of
Fig. 2 HPLC chromatograms of retinoids and tocopherols in eel muscle amberjack muscle (a) and fish feed (b). The numbers on the
lipid without saponification (a) and with saponification (b). The numbers chromatograms represent the following compounds: 1, retinyl
on the chromatograms represent the following compounds: 1, retinyl palmitate; 2, retinyl acetate; 3, α-tocopherol; 4, β-tocopherol; 5, γ-
palmitate; 2, α-tocopherol; 3, retinol tocopherol; 6, δ-tocopherol

Table 4 Retinoid and tocopherol contents of foods

Contents of retinoid and tocopherols (mg 100 g−1 sample)

Retinyl palmitate Retinyl acetate α-Tocopherol β-Tocopherol γ-Tocopherol δ-Tocopherol Retinol

Eel (muscle) 1a 4.43 ± 0.03 nd 10.54 ± 0.21 nd nd nd nd

Eel (muscle) 2b nd nd 10.60 ± 0.13 nd nd nd 2.12 ± 0.03
Eel (liver) 152.98 ± 3.80 nd 6.73 ± 0.46 nd nd nd nd
Amberjack 0.04 ± 0.00 0.08 ± 0.00 0.87 ± 0.07 0.05 ± 0.00 0.17 ± 0.02 0.10 ± 0.01 nd
Feed for fishes 2.82 ± 0.14 1.13 ± 0.06 0.17 ± 0.01 0.51 ± 0.01 0.69 ± 0.03 0.24 ± 0.01 nd
Roasted Soybean flour nd nd 3.77 ± 0.17 1.17 ± 0.06 17.72 ± 1.10 10.68 ± 0.88 nd
Japanese torreya seed nd nd 5.99 ± 0.21 45.11 ± 0.70 2.49 ± 0.05 3.92 ± 0.02 nd
Lotionc nd 0.68 ± 0.04 nd nd nd nd nd
Supplementc 70.19 ± 2.74 nd 246.78 ± 8.84 1.90 ± 0.06 7.46 ± 0.30 40.23 ± 1.46 nd

All values were presented as means ± standard deviation and were estimated by using the triplicate analysis results
nd not detected
a
Lipid of eel muscle treated without saponification
b
Lipid of eel muscle treated with saponification
c
The contents of lotion were presented as mg mL−1 lotion. The contents of retinoid supplement were presented as mg mg−1 lipid
Food Anal. Methods

70000
was determined by analysis with the method developed in this
60000
5 (a) study (45.11 ± 0.70 mg 100 g−1 sample).
To evaluate our developed method, retinoids and tocopherols
50000
Detector response (µV)

not only in foods but also in supplements and cosmetics were

40000 analyzed. As a result, oil from a supplement capsule was found
4
to contain large amounts of retinyl palmitate and α-tocopherol
30000
(70.19 ± 2.74 and 246.78 ± 8.84 mg mg−1 lipid, respectively).
20000 The lotion contained 0.68 ± 0.04 mg mL−1 lotion of retinyl acetate.
2
10000
3

0 Conclusion
0 5 10 15 20
Retention time (min)
In this study, three retinoids and four tocopherols were sepa-
rated and sufficiently detected using our developed HPLC
400000 (b) method. Furthermore, our method required no saponification
1
of the lipid samples and used a fluorescence detector for the
detection of both retinoids and tocopherols. Thus, using our
Detector response (µV)

300000
method, simultaneous analysis of retinoids and tocopherols
can be achieved more readily and cheaply than by the usual
200000 method. Additionally, the developed method showed suffi-
cient reproducibility and quantitative ability.
100000
Using this method, retinoids and tocopherols in various
foods that contain large amounts of these compounds, such
as fish muscle, Japanese torreya seed, and roasted soybean
0
0 5 10 15 20
flour, were analyzed. Furthermore, the retinoid contents of a
Retention time (min) supplement and toning lotion were determined using our
Fig. 4 HPLC chromatograms of retinoids and tocopherols in the lipids of method. Therefore, our developed method is applicable not
roasted soybean flour (a) and toning lotion (b). The numbers on the only to the determination of retinoids and tocopherols in foods
chromatograms represent the following compounds: 1, retinyl acetate; but also to component analysis of medicines or dynamic stud-
2, α-tocopherol; 3, β-tocopherol; 4, γ-tocopherol; 5, δ-tocopherol
ies of these compounds in live organisms.

Eel muscle contains a large amount of α-tocopherols and Acknowledgments This study was supported by a grant from the
retinoid. Especially, saponified eel muscle contains University of Miyazaki, Japan.
2.12 ± 0.13 mg 100 g−1 sample of retinol. This value agrees with
the result of Seo et al. (2013). Additionally, eel liver contains a Compliance with Ethical Standards
large amount of retinyl palmitate (152.98 ± 3.80 mg 100 g−1
Conflict of Interest Mami Ishimaru declares that she has no conflict of
sample). Amberjack contains small amounts of retinyl acetate, interest. Masato Haraoka declares that he has no conflict of interest.
α-tocopherol, β-tocopherol, γ-tocopherol, and δ-tocopherol. Hideo Hatate declares that he has no conflict of interest. Ryusuke
Similarly, the fish feed contained small amounts of the four Tanaka declares that he has no conflict of interest.
tocopherols. α-Tocopherol was 0.87 ± 0.07 mg 100 g−1 sample
Ethical Approval This article contains some studies with fish subjects.
and this value agrees with the contents of the Standard Tables of
All procedures performed in studies involving animals (fish) were in
Food Composition in Japan (2015). The other tocopherols in the accordance with the ethical standards of the institution or practice at
feed may be derived from vegetable oils, such as soybean oil, which the studies were conducted.
which were added to the feed. The tocopherols in amberjack
were then derived from their feed and accumulated in their adi- Informed Consent Not applicable.
pose tissue. Roasted soybean flour is known to contain large
amounts of γ-tocopherol (11.47 ± 1.44 mg 100 g−1 sample)
(Oikawa 1982), and a larger amount of γ-tocopherol was deter-
mined by analysis with the method developed in this study
References
(17.72 ± 1.10 mg 100 g−1 sample). Additionally, Japanese
Amaral JS, Casal S, Torres D, Seabra RM, Oliveira BPP (2005)
torreya seed is known to contain a large amount of β- Simultaneous determination of tocopherols and tocotrienols in
tocopherol (67.5 mg 100 g−1 sample) (Standard Tables of Food hazelnuts by a normal phase liquid chromatographic method. Anal
Composition in Japan 2015); a smaller amount of β-tocopherol Sci 21:1545–1548
 Food Anal. Methods

Brownlee NR, Huttner JJ, Panganamala RV, Cornwell DG (1977) Role of formula and adult nutritionals by normal phase HPLC: first action
vitamin E in glutathione-induced oxidant stress: methemoglobin, 2012.10. J AOAC Int 96:1073–1081
lipid peroxidation, and hemolysis. J Lipid Res 18:635–644 Morriss-Kay GM, Sokolova N (1996) Embryonic development and pat-
Casal S, Macedo B, Oliveira MBPP (2001) Simultaneous determination tern formation. FASEB J 10:961–968
of retinol, beta-carotene and alpha-tocopherol in adipose tissue by Nimalaratne C, Sun C, Wu J, Curtis JM, Schieber A (2014)
high-performance liquid chromatography. J Chromatogr B 763:1–8 Quantification of selected fat soluble vitamins and carotenoids in
Elvehjem CA, Gonce JE Jr, Newell GW (1944) The effect of vitamin E infant formula and dietary supplements using fast liquid chromatog-
on reproduction in dogs on milk diets. J Pediatr 24:436–441 raphy coupled with tandem mass spectrometry. Food Res Int 66:69–
Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the 77
isolation and purification of total lipides from animal tissues. J Biol Office for Resources, Policy Division Science and Technology Policy
Chem 226:497–509 Bureau, Ministry of Education, Culture, Sports, Science and
Gudas LJ, Wagner JA (2011) Retinoids regulate stem cell differentiation. Technology (2015) Standard tables of food composition in
J Cell Physiol 226:322–330 Japan—2015—(seventh revised version). National Printing
Harrison EH (2012) Mechanisms involved in the intestinal absorption of Bureau, Tokyo
dietary vitamin a and provitamin a carotenoids. BBA-Mol Cell Biol Oikawa K (1982) Vitamin E content in foods (part 1) rice and pulses. J
Lipids 1821:70–77 Home Econ Jpn 33:628–632
Palczewski K (2012) Chemistry and biology of vision. J Biol Chem 287:
Hosotani K, Kitagawa M (2003) Improved simultaneous determination
1612–1619
method of β-carotene and retinol with saponification in human se-
Panfili G, Fratianni A, Irano M (2003) Normal phase high-performance
rum and rat liver. J Chromatogr B 791:305–313
liquid chromatography method for the determination of tocopherols
Ihara H, Hashizume N, Hirase N, Suzue R (1999) Esterification makes
and tocotrienols in cereals. J Agric Food Chem 51:3940–3944
retinol more labile to photyolysis. J Nutr Sci Vitaminol 45:353–358
Rimm EB, Stampfer MJ, Ascherio A, Giovannucci E, Colditz GA,
Indyk HE (1988) Simplified saponification procedure for the routine de- Willett WC (1993) Vitamin E consumption and the risk of coronary
termination of total vitamin E in dairy products, foods and tissues by heart disease in men. New Engl J Med 328:1450–1456
high-performance liquid chromatography. Analyst 113:1217–1221 Seo JS, Choi JH, Seo JH, Ahn TH, Chong WS, Kim SH, Cho HS, Ahn JC
Iupac-Iub Joint Commission on Biochemical Nomenclature (1982) (2013) Comparison of major nutrients in eels Anguilla japonica
Nomenclature of tocopherols and related compounds. Eur J cultured with different formula feeds or at different farms. Fish
Biochem 123:473–475 Aquat Sci 16(2):85–92
Kamal-Eldin A, Görgen S, Pettersson J, Lampi AM (2000) Normal-phase Sporn MB, Dunlop NM, Newton DL, Smith JM (1976) Prevention of
high-performance liquid chromatography of tocopherols and chemical carcinogenesis by vitamin a and its synthetic analogs (ret-
tocotrienols—comparison of different chromatographic columns. J inoids). Fed Proc 35:1332–1338
Chromatogr A 881:217–227 Stampfer MJ, Hennekens CH, Manson JE, Colditz GA, Rosner B, Willett
KrčMova L, Urbánek L, Solichová D, Kašparová M, VlčKová H, WC (1993) Vitamin E consumption and the risk of coronary disease
Melichar B, Sobotka L, Solich P (2009) HPLC method for simulta- in women. New Engl J Med 328:1444–1449
neous determination of retinoids and tocopherols in human serum Turner T, Burri BJ (2012) Rapid isocratic HPLC method and sample
for monitoring of anticancer therapy. J Sep Sci 32:2804–2811 extraction procedures for measuring carotenoid, retinoid, and to-
Matsushita T, Inoue SI, Tanaka R (2010) An assay method for determin- copherol concentrations in human blood and breast milk for inter-
ing the total lipid content of fish meat using a 2-thiobarbituric acid vention studies. Chromatographia 75:241–252
reaction. JAOCS, J Am Oil Chem Soc 87:963–972 Von Lintig J (2012) Metabolism of carotenoids and retinoids related to
Mcmahon A, Christiansen S, Shine L, Loi C, Dowell D (2013) vision. J Biol Chem 287:1627–1634
Simultaneous determination of 13-cis and all-trans vitamin a palmi- Wechter WJ, Kantoci D, Murray ED Jr, D’amico DC, Jung ME, Wang
tate (retinyl palmitate), vitamin a acetate (retinyl acetate), and total WH (1996) A new endogenous natriuretic factor: Llu-α. Proc Natl
vitamin E (α-tocopherol and dl-α-tocopherol acetate) in infant Acad Sci U S A 93:6002–6007