PLOS ONE

RESEARCH ARTICLE

Development and validation of reliable

astaxanthin quantification from natural
sources
Inga K. Koopmann ID, Annemarie Kramer, Antje Labes ID*

ZAiT, Center for Analytics in Technology Transfer of Bio and Food Technology Innovations, Flensburg
University of Applied Sciences, Flensburg, Schleswig-Holstein, Germany

* [email protected]

a1111111111
a1111111111
a1111111111 Abstract
a1111111111
a1111111111 Astaxanthin derived from natural sources occurs in the form of various esters and stereo-
mers, which complicates its quantitative and qualitative analysis. To simplify and standard-
ize astaxanthin measurement with high precision, an enzymolysis-based astaxanthin
quantification method was developed to hydrolyze astaxanthin esters and determine free
OPEN ACCESS
astaxanthin in all its diastereomeric forms. Astaxanthin standards and differently pro-
cessed Haematococcus pluvialis biomass were investigated. Linear correlation of stan-
Citation: Koopmann IK, Kramer A, Labes A (2022)
Development and validation of reliable astaxanthin dards of all-E-astaxanthin was observed in a measurement range between extract
quantification from natural sources. PLoS ONE concentrations of 1.0 μg/mL and 11.2 μg/mL with a coefficient of variation below 5%. The
17(12): e0278504. https://doi.org/10.1371/journal. diastereomers 9Z-, and 13Z-astaxanthin, and two di-Z-forms were detected. In contrast to
pone.0278504
the measurement of standards, the observed measurement range was extended to 30 μg/
Editor: Vandana Vinayak, Doctor Harisingh Gour mL in extracts from H. pluvialis. The nature of the sample had to be taken into account for
Vishwavidyalaya: Dr Hari Singh Gour University,
measurement, as cell, respectively, sample composition altered the optimal concentration
INDIA
for astaxanthin determination. The measurement precision of all-E-astaxanthin quantifica-
Received: March 10, 2022
tion in dried H. pluvialis biomass (1.2–1.8 mg dried biomass per sample) was calculated
Accepted: November 9, 2022 with a coefficient of variation of maximum 1.1%, whereas it was below 10% regarding the
Published: December 2, 2022 diastereomers. Complete enzymolysis was performed with 1.0 to 2.0 units of cholesterol
Peer Review History: PLOS recognizes the esterase in the presence of various solvents with up to 2.0 mg biomass (dry weight). The
benefits of transparency in the peer review method was compared with other astaxanthin determination approaches in which astax-
process; therefore, we enable the publication of anthin is converted to acetone in a further step before measurement. The developed
all of the content of peer review and author
method resulted in a higher total astaxanthin recovery but lower selectivity of the diaste-
responses alongside final, published articles. The
editorial history of this article is available here: reomers. The reliability of photometric astaxanthin estimations was assessed by compar-
https://doi.org/10.1371/journal.pone.0278504 ing them with the developed chromatographic method. At later stages in the cell cycle of
Copyright: © 2022 Koopmann et al. This is an open H. pluvialis, all methods yielded similar results (down to 0.1% deviation), but photometry
access article distributed under the terms of the lost precision at earlier stages (up to 31.5% deviation). To optimize sample storage, the
Creative Commons Attribution License, which shelf life of astaxanthin-containing samples was investigated. Temperatures below -20˚C,
permits unrestricted use, distribution, and
excluding oxygen, and storing intact H. pluvialis cells instead of dried or disrupted biomass
reproduction in any medium, provided the original
author and source are credited. reduced astaxanthin degradation.
Data Availability Statement: All relevant data are
within the paper and its Supporting information
files.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 1 / 34

PLOS ONE Astaxanthin extraction and determination

Funding: IKK, AK, AL: The ZAiT is part of the Introduction

project Grenzland INNOVATIV Schleswig-Holstein
[innovative border region Schleswig-Holstein] Astaxanthin (3,3´-dihydroxy-β,β´-carotene-4,4´-dione) is a secondary ketocarotenoid. It has
funded by the Ger-man Federal Ministry of a hydrocarbon backbone that comprises a central, delocalized π-electron system. β-ionone
Education and Research (BMBF) in context of rings terminate the hydrocarbon chain at both ends [1, 2]. The presence of one hydroxy- and
“Inno-vative Hochschule” [innovative university], one oxo-group at each of these terminal rings further classifies it as xanthophyll. Its biosyn-
https://www.bmbf.de/bmbf/en/home/home_node.
thesis has been observed mainly in microalgae [3–12] but also in a few protists [13–15], bac-
html. We acknowledge financial support by Land
Schleswig-Holstein (federal state Schleswig- teria [16–20], archaea [21] as well as yeasts [22, 23], and very few plant species [24, 25].
Holstein) within the funding program Open Access- Astaxanthin exhibits properties against photooxidative stress by scavenging free radicals
Publikationsfonds. The funders had no role in [26–31]. These antioxidative effects were also observed in mammalian cells in vitro and in
study design, data collection and analysis, decision vivo [32–37] and led to its application as a nutritional supplement, food and feed additive,
to publish, or preparation of the manuscript.
and in cosmetics. Astaxanthin from natural sources has been authorized for human con-
Competing interests: IK has worked for Sea and sumption in many countries worldwide [38]. Few minor adverse effects of dosed astaxanthin
Sun Technology, a provider of H. pluvialis bio- consumption were observed [38]. Allergic reactions to the consumption of H. pluvialis pro-
mass for the study. All authors have declared that
teins cannot be excluded, but their likelihood has been considered low in tested astaxanthin-
no other competing interests exist. This does not
alter our adherence to PLOS ONE policies on
rich novel food ingredients by the European Food Safety Authority (EFSA) [39]. It can also
sharing data and materials. be produced synthetically, but its natural form has gained interest with respect to consumer
demands [40, 41]. A major source for the biotechnological production of astaxanthin is the
green alga Haematococcus pluvialis [11, 42]. It contains 1.9 to 7.0% astaxanthin of its dry
weight in certain life cycle stages and under stress conditions [43–48]. As in other producing
organisms, astaxanthin in H. pluvialis is derivatized at one or both hydroxyl groups with var-
ious fatty acids, mainly leading to the formation of monoesters (76–94%), besides diesters
(2–23%) and free astaxanthin (0.3–4%) [49–53]. The most common fatty acids occurring in
these combinations are C18:1 and C18:2, amongst many others [49–52]. The conjugated
double-bound system can theoretically form a variety of diastereomers [54]. The most abun-
dant form in H. pluvialis is the all-E configuration followed by the sterically unhindered 9Z-,
13Z- and 15Z-forms [55–59] as well as corresponding di-Z-forms [50, 58, 60]. Due to the two
stereogenic carbon atoms C-3 and C-3’, three or four enantiomers are possible when a sym-
metrical or asymmetrical Z-form is considered, respectively (3S,3’S, 3R,3’R, 3S,3’R / 3R,3’S).
In H. pluvialis, the ratio of the astaxanthin enantiomers was reported to be 88:10:2 and 99:0:1
(3S,3’S, 3R,3’R and 3S,3’R) [57, 61, 62].
Determining total astaxanthin in biological samples is crucial on the analytical and pro-
duction scale. Most studies use estimating, photometric quantification methods instead of
precise determination of total astaxanthin including its isomers for the means of hyphenated
chromatography. Photometry enables simple and fast determination without the need for
sophisticated equipment [23, 43, 63–65]. The biological variability of astaxanthin with all its
different isomeric forms and esters [49, 51] complicates the exact quantification of astax-
anthin. Moreover, astaxanthin is sensitive to treatment with solvents [54, 66, 67]. Access to
sophisticated hyphenated chromatography protocols is limited. Here we demonstrate a fast,
robust, and reliable method to get reproducible and comparable results of an overall astax-
anthin content.
Many available methods to determine the astaxanthin content in H. pluvialis have been
described. The most sophisticated ones extract, identify, and quantify a variety of esters and
isomers with liquid chromatographic and spectrophotometric/mass spectrometric methods
[46, 49–51, 64, 68–74]. However, this is time-consuming, as liquid chromatographic meth-
ods that differentiate the varying esters require much time for separation. Correct evalua-
tion is elaborate because absorption spectra of carotenoids are similar. Moreover, correct
quantification is difficult due to the different absorption coefficients of the geometric
isomers.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 2 / 34

PLOS ONE Astaxanthin extraction and determination

A faster and more general approach is to extract all carotenoids and determine astaxanthin
photometrically by estimating its proportion of total carotenoids [23, 43, 63]. The methods
optionally include or destroy residual chlorophylls with a chemical treatment before measure-
ment [45, 64, 65, 75, 76]. Recent approaches suggested the use of a higher wavelength (530
nm) than the absorbance maximum of astaxanthin for quantification to avoid overlapping
absorption by chlorophylls and other carotenoids [48, 59, 77]. These photometric methods are
cost-extensive, though the estimation may lead to unpredictable deviations, and the chloro-
phyll destruction may also impact the carotenoids. The geometrical isomers of astaxanthin
have similar absorption spectra [50, 58–60, 66, 78], though different extinction coefficients
[79], which impedes their individual quantification. To refine those photometric methods,
astaxanthin can be fractionated by chromatographic methods (thin layer or column chroma-
tography) beforehand and the corresponding fractions can be measured photometrically,
applying the respective absorption coefficients [80–82]. However, the correct merging of the
many esters and isomers is still difficult, leading to inaccuracy and deviation.
An option is to saponify the esters and isolate and quantify the resulting free astaxanthin
with a shorter chromatographic method. Here, different methods for deesterification have
been applied. Alkaline treatments with NaOH [55, 70, 83–88] or KOH [86, 89] are possible but
can cause changes in the structural conformation and/or degradation of carotenoids [77, 85,
86, 90]. Thus, another approach is enzymolysis with esterases [61, 83, 85, 91, 92], lipases [93–
95] or whole-cell catalysts [96].
The enzymolysis of astaxanthin is a promising compromise to quantify astaxanthin pre-
cisely while minimizing time consumption. Based on the enzymolysis of carotenoid esters,
developed by Jacobs et al. [91], methods for quantification of astaxanthin were adapted by
various studies and protocols [61, 83, 87, 92, 97–99]. However, to our knowledge, an in-
depth evaluation of the process boundaries, detection limits, accuracy, and applicability of
various astaxanthin-containing natural sources has not yet been performed. Therefore, a
method for robust and reliable determination of astaxanthin from H. pluvialis was developed
and validated, which was still faster than methods determining the various esters of astax-
anthin. The five step-method comprises the preparation of biological samples and astax-
anthin extraction, enzymolysis, liquid-liquid extraction of astaxanthin, processing of the
resulting extract, and measurement with ultra-high performance liquid chromatography
(UHPLC) and UV/VIS spectrometry. Particular emphasis was set to relevant factors:
Enzyme and biomass amount, shelf life, incubation time, quantification limits, linearity, pre-
cision, systematic errors, robustness, and isomerization effects were tested. The final method
was compared to photometric astaxanthin determination approaches. We aimed to develop
a method that is applicable at all different stages of astaxanthin production and works with
different formulations of astaxanthin. Therefore, it was tested with commercially available
H. pluvialis and astaxanthin samples.

Materials and methods

Chemicals and reagents
Analytical grade acetone (SupraSolv), petroleum ether, and acetonitrile (hypergrade) were
obtained from Merck (Darmstadt, Germany). Ethanol and Tris(hydroxymethyl)aminomethan
(TRIS) (� 99.9%) were provided by Carl Roth (Karlsruhe, Germany) and formic acid (99%
ULC/MS) by Biosolve (Valkenswaard, Netherlands). Hydrochloric acid for pH value adjust-
ment was purchased from Merck (Darmstadt, Germany). Cholesterol esterase from Pseudomo-
nas sp. for enzymolysis was obtained from MP Biomedicals (Eschwege, Germany). All-E-
astaxanthin standard in its free form (SML0982, � 97%, 3S,3’S, from Blakeslea trispora) was

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 3 / 34

PLOS ONE Astaxanthin extraction and determination

provided by Sigma-Aldrich (Taufkirchen, Germany) and astaxanthin monopalmitate (1017,

3RS,3’RS) by CaroteNature (Münsingen, Swiss).

H. pluvialis biomass and astaxanthin containing extracts

Various commercially available sources of Haematococcus pluvialis biomass containing
astaxanthin were used: Dried and disrupted cysts were obtained from Golden Peanut (Gar-
stedt, Germany). Sea & Sun Technology GmbH (Trappenkamp, Germany) provided con-
centrated H. pluvialis aplanospores from seven different batches in nutrient depleted media.
Sea & Sun Technology also provided oleoresins of H. pluvialis obtained from supercritical
CO2 (SC-CO2) extraction either pure or diluted in ethanol. Cultivation and harvest parame-
ters are not available on those samples, but for two batches of concentrated, fresh biomass
obtained from Sea & Sun Technology. Those were taken from day 22 to 28 of a light and
nutrient-induced stress phase. All had been cultivated in BG11 medium (Culture Collection
of Algae at the University of Göttingen): 17.6 mM NaNO3, 0.18 mM K2HPO4 x 3H2O, 0.3
mM MgSO4 x 7H20, 0.25 mM CaCl2 x 2H2O, 0.031 mM citric acid, 0.023 mM ferric ammo-
nium citrate, 0.003 mM Na2EDTA x H2O, 0.19 mM NaCO3 and 1 mL/L trace metal solution
made of 1.0 mM H3BO3, 1.0 mM MnSO4 x H2O, 1.00 mM ZnSO4 x 7H2O, 0.01 mM
(NH4)6Mo7O24 x 4H2O and 0.1 mM CuSO4 x 5H2O. These were used exclusively for the
experiments in which the developed method was to be compared with photometric
methods.

Disruption of H. pluvialis and astaxanthin extraction

Undisrupted biomass was bead-milled to ensure astaxanthin accessibility during further pro-
cessing. Therefore, 0.2–4.0 mg cell dry mass were weighed into lysis tubes type C (ceramic
beads, diameter 0.4–0.6 mm) by Analytik Jena (Jena, Germany). 500 μL acetone were added
for astaxanthin extraction. When analyzing aplanospores in liquid medium, volumes of equal
to or less than 300 μL of well-homogenized samples were filled into the same lysis tubes and
matched to a final amount of 0.5–2.0 mg of biomass. The tubes were filled up to 500 μL with
acetone accordingly. The cysts were broken mechanically in a swing mill (MM 2000, Retsch,
Haan, Germany) at 27 Hz for 3 minutes without cooling and centrifuged at 10,000 x g. The
supernatant of both sample types was transferred to a centrifuge tube, and 500 μL of fresh ace-
tone were added to the lysis tube. This procedure was repeated twice until the supernatant and
the residual biomass were colorless.

Astaxanthin deesterification by enzymolysis

The preparation for enzymolysis varied with the sample type. Using concentrated fresh or
dried biomass, the combined supernatants obtained from astaxanthin extraction were filled up
to 3 mL with acetone (≙ 90–100% v/v). Highly viscous oleoresins were weighed directly into
centrifuge tubes with 0.4 to 15.8 mg and diluted in 3 mL acetone. Oleoresins dissolved in etha-
nol were used at 0.1 to 3 mL and filled up with acetone to 3 mL. 2 mL of 50 mM TRIS buffer
(pH 7 at 21˚C) and 600 μL cholesterol esterase solution at a concentration of 3.3 U/mL sus-
pended in the same buffer were added. Accordingly, final cholesterol esterase concentration in
the sample was 2.0 units or 0.36 units per mL. The tubes were incubated at 37˚C in a water
bath and mixed gently every 10 minutes. Astaxanthin was recovered by liquid-liquid extraction
with 2 mL (≙ 26% v/v) of petroleum ether. The mixture was shaken vigorously for 10 s to ame-
liorate astaxanthin transfer into the petroleum ether. Subsequent phase separation was
enhanced by centrifugation at 3,000 x g for 1 minute. The upper, astaxanthin-containing phase

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 4 / 34

PLOS ONE Astaxanthin extraction and determination

was filtered (0.45 μm, PET) into amber vials. It was either measured directly or stored at -21˚C
for one night before analysis.

Analysis and quantification of free astaxanthin and lutein by

UHPLC-PDA-MS
Extracts were vortexed (Vortex 3, IKA-Werke GmbH und Co. KG, Staufen, Germany) and
ultrasonicated for 20–30 s (Sonorex Super RK 103 H, Bandelin electronic GmbH und Co. KG,
Berlin, Germany) before analysis if they had been stored overnight. Qualification and quantifi-
cation of astaxanthin were performed by UHPLC using an ACQUITY Arc system by Waters
(Milford, MA, USA) equipped with a sample manager (FTN-R), a quaternary solvent manager
(R), an UV/Vis detector (2998 PDA Detector), and a mass spectrometer (Acquity QDa Detec-
tor). A C18-column (Cortecs C18 2.7 μm, 90 Å, 3.0 x 100 mm) was operated at 40˚C. The
injection volume was 5 μL. Starting conditions were 70% millipore water and 30% acetonitrile,
both containing 0.1% formic acid. Over four minutes, the gradient increased linearly to 90%
acetonitrile and 10% millipore water. This ratio was kept isocratic for five minutes. A rinsing
step on 100% acetonitrile was attached for a further 2.5 minutes. 3.5 minutes were used for
regeneration of the starting conditions. Flow velocity was 0.5 mL/min. Optical spectra were
measured in a range of 200 to 800 nm, and astaxanthin data were analyzed and quantified at
its determined absorbance maximum of 474 nm. The absorption maximum of lutein was
shown to be at 448 nm. This wavelength was used for quantification. The mass spectrometer
with electrospray ionization (ESI) was operated in positive mode with a cone voltage of 15 V
and a probe temperature of 600˚C, measuring in a range of m/z 150 to 1250. For further accu-
racy, the mass of astaxanthin was observed by selected ion recording (SIR) at m/z 598 [M+H]+
and lutein at m/z 570 [M+H]+. Besides the major all-E-astaxanthin peak, peaks with UV/Vis
absorption spectra corresponding to Z-astaxanthin isomers [50, 58–60, 66, 78] that were addi-
tionally accompanied by peaks with the mass of astaxanthin in SIR were assigned to 9Z- and
13Z-astaxanthin and several di-Z-isomers. However, as the latter are difficult to differentiate
without further validation, they were summed and are consecutively termed di-Z-isomers.
Quantities of all diastereomers were estimated by using the quantification of all-E-astaxanthin,
corrected by factors adjusting the different extinction coefficients published by Bjerkeng et al.
[79], namely 1.20 for 9Z-astaxanthin, 1.56 for 13Z-astaxanthin, and 1.11 for the di-Z-isomers.

Calibration curve—Astaxanthin
For identification and quantification, free all-E-astaxanthin was used at concentrations of 0.5–
45.0 μg/mL in acetone. Blank acetone was applied for zero value determination. Samples of 0.5
to 54.8 μg free astaxanthin and 1.0 to 62.6 μg astaxanthin monopalmitate were subjected to the
deesterification process to determine their recoveries. Both were diluted in acetone and treated
as described above. In the processing of free astaxanthin, the cholesterol esterase solution was
replaced with the same amount of TRIS buffer. For quantification of monopalmitate-ester
derived free astaxanthin, a proportion of 69.4% w/w astaxanthin was assumed by molecular
weight calculation.

Calibration curve—Lutein
To identify and quantify lutein from H. pluvialis, free lutein standard in concentrations of 1.0
to 61.5 μg/mL dissolved in acetone was used for calibration. Linear regression was performed
by ordinary least squares and forced through zero for optimal approximation to standards
recovered in petroleum ether.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 5 / 34

PLOS ONE Astaxanthin extraction and determination

Optimization of enzymolysis and liquid-liquid extraction

Enzyme amount and duration of enzymolysis. The amount of cholesterol esterase was
varied between 0.05 to 2.0 units per reaction for equal amounts of astaxanthin esters in 0.4 mg
H. pluvialis powder (Golden Peanut) to pinpoint optimal enzyme concentration for astax-
anthin conversion during 0.75 hours of incubation. In addition, the incubation time of enzy-
molysis was doubled to 1.5 hours using 0.5 units and 2.0 units of cholesterol esterase.
Influence of ethanol on astaxanthin enzymolysis. The recovery of astaxanthin in
samples processed in the presence and absence of ethanol was compared. Therefore, two
different ethanolic SC-CO2 extracts (n = 7 and n = 5) of H. pluvialis with volumes between
150 and 1000 μL were used. They were either fed directly to the deesterification process
(n = 5 and n = 3) or ethanol was evaporated at 40˚C under nitrogen, and the sample was
subsequently resuspended in acetone and TRIS buffer containing 2.0 units of cholesterol
esterase for deesterification (n = 2 and n = 2). All samples were enzymolyzed, extracted,
and astaxanthin content was quantified based on UHPLC-PDA measurements as described
above.
Influence of solvents used in liquid-liquid extraction on astaxanthin quantification.
The volume of the upper petroleum ether layer was determined after liquid-liquid extraction
to determine extract concentrations. Therefore, blank samples (n = 8) consisting of 3 mL ace-
tone and 2.6 mL TRIS buffer were heated to 37˚C for 20 minutes in a water bath. Afterward,
the samples were shaken vigorously with 2 mL petroleum ether, centrifuged at 3,000 x g for 1
minute, and cooled to 21˚C in a water bath. The complete upper layer was transferred into
reaction tubes and weighed (n = 4). Its density was determined by measuring the upper layer
of the remaining samples with a pycnometer (n = 4). The corresponding volume was
calculated.
To quantify the influence of ethanol on the volume of the upper extraction phase, various
volumes of ethanol (0 to 2010 μL) were filled up to 3 mL with acetone and treated similarly to
the previous experiment.

Extract processing
After liquid-liquid extraction, various processing procedures of the petroleum ether fraction
were compared to reach the maximum recovery of astaxanthin. Therefore, 20 μg of free
astaxanthin standard were deesterified and extracted with petroleum ether as described
above. The upper layer was treated in three ways: (1) The extraction phase was not treated.
(2) An aliquot of the extraction phase was dried under nitrogen and redissolved in the same
volume of acetone. (3) Most of the extraction phase was transferred to another tube, and
fresh petroleum ether was added to the original sample. The extraction and transfer steps
were repeated twice until the upper layer was colorless. The combined fractions were dried
at 40˚C under nitrogen and redissolved in 2 mL of acetone. The samples were filtered, and
astaxanthin was measured by the UHPLC-PDA-MS method. All experiments were per-
formed in triplets or quartets.

Evaluation of detection limits and linearity of astaxanthin determination

Detection limits and linearity of astaxanthin measurement were determined by varying the
amount of H. pluvialis powder (Golden Peanut) from 0.04 mg to 2.0 mg in triplets. 3.3 and 4.0
mg were tested without replicates. They were enzymolyzed with a constant concentration of
2.0 units cholesterol esterase in 3 mL of acetone and 2.6 mL of TRIS buffer and 0.75 hours of
incubation.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 6 / 34

PLOS ONE Astaxanthin extraction and determination

Determination of measurement precision

The astaxanthin content of two samples from different packages of dried H. pluvialis biomass
(Golden Peanut) was measured using the standard enzymolytic UHPLC-PDA procedure
described above. Measurements were performed once on three and four different days to
determine the variation between experiments carried out independently. Between 1.2 and 1.8
mg of sample were applied.

Method adjustment
Liquid cultures and oleoresins. Various liquid culture batches of H. pluvialis biomass
(Sea & Sun Technology) concentrated at 7.1, 39.5, 182, and 262 g/L were examined. For dis-
ruption, volumes of 1 to 45 μL of these samples were transferred to lysis tubes to get final
amounts of 0.1 to 2.0 mg of biomass. Further processing was accomplished as described above.
Additionally, four oleoresins of H. pluvialis extracted with SC-CO2 were investigated.
Therefore, 0.15 to 1.8 mg of oleoresin were directly added to the deesterification solution.
Enzymolysis was carried out with 2.0 units and 4.0 units of cholesterol esterase per reaction.
Extraction and quantification were performed as described above.
Sample mixtures. Various astaxanthin-containing samples were deesterified individually
and merged to examine cross-interactions. The individual samples contained either 240 μg of
H. pluvialis powder (Golden Peanut), 16 μg of astaxanthin monopalmitate, or 12 μg of free
astaxanthin. Maximum 4.5% w/w of the H. pluvialis powder were considered to be free astax-
anthin by measurements. Astaxanthin monopalmitate consists of about 69.4% w/w of free
astaxanthin. Thus, the final extracts of the samples contained 4.2 to 4.7 μg/mL of free astax-
anthin. Moreover, three samples were prepared, two containing mixtures of two of the single
solutions, and one sample containing all three. In these, the individual samples were repre-
sented at identical amounts as in the samples containing only one of the analytes.
After enzymolysis and recovery in petroleum ether, three aliquots of 600 μL were taken
from each sample. They were prepared either with 63 μL pure acetone, 31.5 μL pure acetone
and 31.5 μL of astaxanthin standard solution containing 3.6 μg free astaxanthin, or with 63 μL
of astaxanthin standard solution containing 7.2 μg free astaxanthin. Furthermore, 600 μL pure
petroleum ether were enriched with acetone and astaxanthin like the samples. The experiment
was performed once.

Method comparison to photometric astaxanthin estimation

To compare the developed method to simpler photometric approaches, the astaxanthin con-
tent of several H. pluvialis samples (Sea & Sun Technology) taken towards the end of a cultiva-
tion phase were measured with UHPLC-PDA-MS and photometrically. From two individual
culture batches, samples were taken on days 22–24 and 27–28 of the cultivation. For each
UHPLC measurement, 1.0 mg of H. pluvialis biomass was used. Extraction, enzymolysis, and
quantification were carried out as described above. All samples were measured once. For pho-
tometric measurements, samples were extracted similarly. The extracts were measured at λ =
470 nm and with a wavelength scan from 300 to 700 nm in steps of 2 nm. Astaxanthin content
was calculated using two different approaches: (1) Lambert-Beer was applied with the
concentration of astaxanthin (Cax) in g/mL, the optical path length d in cm, the absorption
at the wavelength represented with the subscript number and the 1%-absorption coefficient
1cm εax =1980 [100mL/(g cm)] of a carotenoid mixture in 80% acetone and 20% water [100].
A1% �

A factor of 0.8 for the astaxanthin proportion of total carotenoids was assumed based on vari-
ous literature data [45, 53, 59, 101]. This (Eq 1) will be termed “general equation” hereafter. (2)

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 7 / 34

PLOS ONE Astaxanthin extraction and determination

The calculations of Lichtenthaler [100] for chlorophylls and total carotenoids were used to
determine the carotenoid fraction. See Eqs 2.1–2.3 with the concentration of chlorophyll a
(Ca), chlorophyll b (Cb), and of total carotenoids (Cx+c) in μg/mL in acetone with 20% water,
and the absorptions at corresponding wavelengths represented with the subscript numbers.
Again, the astaxanthin proportion was calculated by multiplying the total carotenoid content
with a factor of 0.8 (Eq 2.4).
0:8 A470
Cax ¼ ð1Þ
εax d

Ca ¼ 12:2 A663:2 2:79 A646:8 ð2:1Þ

Cb ¼ 21:50 A646:8 25:10 A663:2 ð2:2Þ

ð1000 A470 1:82 Ca 85:02 Cb Þ

Cxþc ¼ ð2:3Þ
198

Cax ¼ 0:8 Cxþc ð2:4Þ

Shelf life experiments

Different samples were prepared to determine the shelf life of astaxanthin in various matrices
and environmental conditions. (1) Free all-E-astaxanthin at a concentration of 1 μg/mL in ace-
tone was stored in amber vials at -80˚C, -20˚C, 4˚C, and room temperature for 833 days. Opti-
cal density was measured initially, after 11, 42, and 833 days by UV/Vis photometry at λ = 474
nm. (2) Astaxanthin concentration of a concentrated H. pluvialis culture (40 g/L) (Sea & Sun
Technology) was determined directly and after storage at 4˚C in the dark for 104 and 489 days.
(3) The same sample was freeze-dried. Therefore, aliquots were poured into small glass vessels
with a filling height of approximately 0.5 cm. They were placed into a freeze drier (Alpha 1–4,
Christ, Osterode, Germany) for 24 hours at � 37 Pa. After the lyophilization was finished, the
vessels were sealed under vacuum and stored at 4˚C in the dark. However, one sample was
measured directly using the described enzymolytic UHPLC-PDA standard method. The other
samples were exposed to ambient air after 7, 108, and 489 days and measured. Afterward, the
samples were closed without vacuum sealing and stored under the same conditions until 489
days after lyophilization. Hereafter, all samples were measured again. (4) Two sealed packages
of H. pluvialis biomass (Golden Peanut) were stored as purchased at -21˚C in the dark over
the whole experiment. One was opened and closed again tightly at the beginning of the experi-
ment. Both packages were opened after two and a half years and measured using the
UHPLC-PDA standard method.

Statistics
Linear regression was performed by ordinary least squares method, and the significance of the
deviation of the y-intercepts from zero was evaluated by t-tests. To determine the device-
related measurement deviation, free all-E-astaxanthin standards dissolved in acetone were
measured regularly prior to or after sample measurements. A total of 45 samples with astax-
anthin concentrations between 0 and 45.0 μg/mL were taken into account. Tests on significant
deviations were calculated using mean difference tests with a level of significance of σ = 0.05.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 8 / 34

PLOS ONE Astaxanthin extraction and determination

Results and discussion

An enzymolysis-based process for astaxanthin deesterification from natural samples was estab-
lished to quantify astaxanthin easily without having to identify its various esters. Interactions
of sample amount and enzyme quantity were evaluated to find optimal conditions for maxi-
mum astaxanthin recovery. These were verified with different astaxanthin-containing samples.
Moreover, the shelf life was evaluated to minimize losses prior to analysis. Finally, the method
was compared to simple photometric approaches for astaxanthin determination to assess their
applicability in process monitoring (Fig 1).

Calibration curve
Detection limits and linearity of astaxanthin standards. For astaxanthin quantification,
a calibration was established with free all-E-astaxanthin standard dissolved in acetone, filtered,
and injected directly into the UHPLC-PDA-MS system in various concentrations. The result-
ing peak areas were integrated for calibration curve determination. The retention time of the
all-E-astaxanthin peak was 7.50 minutes. Linear regression resulted in the correlation f(x) =
1693.4x − 990.57 with a coefficient of determination of 0.9985. The coefficient of variation was
below 4% for all data points, except for the zero value. The calculated y-intercept was signifi-
cantly different from zero (Table 1). Measurement of all-E-astaxanthin dissolved in acetone
has a greater number of replicates than the following measurements because it was also used as
a parallel check on the stability of the instrument.
As the process should be applied for natural samples that have to be deesterified, the cali-
bration was repeated with free all-E-astaxanthin and astaxanthin monopalmitate. Standards of
both substances were subjected to enzymolysis conditions to verify the behavior and linearity
of the calibration (Fig 1, a1). The retention time of all-E-astaxanthin was 6.43 minutes.

Fig 1. Schematic overview of the method development for astaxanthin quantification.

https://doi.org/10.1371/journal.pone.0278504.g001

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 9 / 34

PLOS ONE Astaxanthin extraction and determination

Table 1. Calibration curves of astaxanthin standards.

Sample Concen-tration limits Number of measure- Linear regression Linear regression
(μg/mL) ments through zero
Standard Solvent Processing Integrated Equation R2 Equation R2
isomers
All-E-Axa acetone direct all-E-Ax 0–32.3 112 1693.4x- 0.9985 1643.0x 0.9966
990.57�
All-E-Ax acetone direct all-E, 9Z, 13Z, 0–32.3 27 1718.2x- 0.9993 1724.4x 0.9982
diZ-Ax 676.84�
All-E-Ax acetone direct all-E-Ax 0–11.2 78 1580.9x- 0.9964 1493.7x 0.9917
548.95
All-E-Ax acetone direct all-E, 9Z, 13Z, 0–11.2 21 1626.2x- 0.9962 1566.1x 0.9933
diZ-Ax 429.73
All-E-Ax PEc like in all-E-Ax 0–11.2 10 1624.4x-65.64 0.9996 1615.3x 0.9995
+ acetone enzymolysis
Ax-Mpb PE like in all-E-Ax 0–7.5 9 1700.8x+5.50 0.9994 1702.0x 0.9994
+ acetone enzymolysis
All-E-Ax & Ax- PE like in all-E-Ax 0–11.2 18 1633.6x 0.9982 1638.9x 0.9982
Mp + acetone enzymolysis +33.13
All-E-Ax & Ax- PE like in all-E, 9Z, 13Z, 0–11.2 18 1753.3x 0.9985 1759.9x 0.9985
Mp + acetone enzymolysis diZ-Ax +40.56

Regressions used for quantification are highlighted bold.

�
y-intercept significantly different from zero.
a
Ax = Astaxanthin.
b
Ax-Mp = Astaxanthin monopalmitate.
c
PE = Petroleum ether.

https://doi.org/10.1371/journal.pone.0278504.t001

Subsequently, calibration was repeated for the obtained astaxanthin UHPLC-PDA peaks and
compared to the previous based on the standard dissolved in acetone. Calculations with the
previously examined line of best fit did not properly depict the lower concentrations of enzy-
molyzed astaxanthin in a concentration range between 0.0 to 13.6 μg/mL and resulted in an
overestimation of all-E-astaxanthin content of a maximum of 131% at a concentration of
0.4 μg/mL and 57% at 1.0 μg/mL astaxanthin. Thus, linear regressions were calculated from
the newly obtained values. Regressions resulted in similar linear correlations below 11.2 μg/
mL astaxanthin in the liquid-liquid extraction phase, which was the upper detection limit. It
was f(x) = 1624.4x − 65.639 (R2 = 0.9996, n = 10) for free all-E-astaxanthin from free all-E-
astaxanthin standard and f(x) = 1700.8x + 5.500 (R2 = 0.9994, n = 9) for free all-E-astaxanthin
from astaxanthin monopalmitate (Fig 2). The y-intercept was not significantly different from
zero for both. As the aim was to quantify astaxanthin derived from natural samples that have
to be prepared by enzymolysis, quantification should be performed by using an appropriate
calibration curve. The calibration using astaxanthin and astaxanthin monopalmitate subjected
to the standard deesterification procedure was the most trustworthy. For higher accuracy, a
calibration curve combining the previous two, including the later described isomers and
forced through zero, was used. It was f(x) = 1759.9x (R2 = 0.9985, n = 18). This curve repre-
sented astaxanthin concentrations between 1.0 μg/mL and 11.2 μg/mL with less than 5% error
in both directions.
At 0.5 μg/mL the calculated recovery decreased by 11%. Thus, a concentration between 0.5
and 1.0 μg/mL should be recognized as the lower detection limit. Moreover, the linear correla-
tion ended when the astaxanthin concentration exceeded 11.2 μg/mL. Higher concentrations
resulted in reduced recoveries down to 55% at 27 μg/mL. This behavior was correlated to the

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 10 / 34

PLOS ONE Astaxanthin extraction and determination

Fig 2. Calibration curves of differently processed astaxanthin standards. B close-up from A including the linear
regressions through zero of the standards in petroleum ether and acetone mixture. ● all-E-astaxanthin standard in
acetone ┉ linear f(x) = 1693.4x-990.57, R2 = 0.9985, ◆ all-E-astaxanthin standard in petroleum ether-acetone mixture
╸━ ╸ linear f(x) = 1615.3x, R2 = 0.9995, ◼ deesterified astaxanthin monopalmitate in petroleum ether-acetone mixture
╸╸╸ linear f(x) = 1702.0x, R2 = 0.9994.
https://doi.org/10.1371/journal.pone.0278504.g002

precipitation of astaxanthin, which was observed at the phase boundary and glass wall when
more than 11.2 μg/mL of astaxanthin were present during the liquid-liquid extraction step.
The liquid-liquid extraction of astaxanthin into petroleum ether is a delicate step as pure astax-
anthin is poorly soluble in pure petroleum ether. During liquid-liquid extraction, acetone is
absorbed into the petroleum ether, shifting the equilibrium and accumulating astaxanthin in
the upper phase. The solubility of astaxanthin in the petroleum ether-acetone mixture is still
limited at an upper boundary of approximately 11.2 μg/mL. This phenomenon was more pro-
nounced for the processed free astaxanthin than processed astaxanthin monopalmitate. E.g.,
astaxanthin derived from the free standard resulted in a 58% recovery at 22.2 μg/mL, whereas
astaxanthin derived from astaxanthin monopalmitate resulted in an 80% recovery at 22.3 μg/
mL. This difference indicates an alteration of astaxanthin solubility in the upper layer by the
presence of cleaved palmitic acid. Further testing of the type of solvent and ratio might
increase astaxanthin solubility, but the functionality of the deesterification process has to be
ensured. In the current experimental setup, a concentration of 11.2 μg/mL astaxanthin in the
liquid-liquid extraction phase can be considered as the upper detection limit by restricting the
amount of biomass for optimal astaxanthin recovery.
The observed shifts in retention time can be attributed to the different solvents. The astax-
anthin peak was more than one minute later in pure acetone than in petroleum ether-acetone.
This may be due to altered binding behavior on the column at starting conditions and/or mis-
cibility with the mobile phase, which started at a high water content (70% v/v). However,
decreasing the water content at the beginning affected peak shape negatively.
Selectivity. Considering free all-E-astaxanthin standard dissolved in acetone and injected
directly, diastereomers were detected besides the main all-E-astaxanthin peak (S1 Table). 9Z-
astaxanthin, 13Z-astaxanthin and one di-Z-isomer were observed with a medium proportion of
total astaxanthin of 2.4±0.2% (n = 29) at 7.90 minutes, 0.4±0.1% (n = 26) at 8.84 minutes, and
0.2±0.03% (n = 26) at 8.56 minutes, respectively. The proportion of 9Z- and di-Z-astaxanthin
remained constant, whereas 13Z-astaxanthin rose with prolonged storage and multiple mea-
surements of the same sample up to 2.2%, indicating isomerization reactions during storage in
acetone. Organic solvents have been reported to cause the isomerization of carotenoids [54, 66,
67, 102]. These were described to favor the 13Z-isomer [67], which can be confirmed here.
In standards processed similarly to enzymolyzed samples, all-E-astaxanthin was detected,
and two minor peaks were assigned to 9Z- (7.51 minutes) and 13Z-astaxanthin (8.50 minutes).

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 11 / 34

PLOS ONE Astaxanthin extraction and determination

The medium proportions of 9Z-astaxanthin, relative to total astaxanthin, were 5.9±0.5%

(n = 9) and 4.9±0.3% (n = 8), obtained from free astaxanthin and astaxanthin monopalmitate,
respectively. The peak areas of 13Z-astaxanthin were 1.3±0.5% (n = 9) and 2.1±0.9% (n = 8)
obtained from free astaxanthin and astaxanthin monopalmitate, respectively. Thus, 9Z- and
13Z-astaxanthin were detected at significantly higher quantities in processed free astaxanthin
and astaxanthin monopalmitate than in free astaxanthin dissolved in acetone. Again, this indi-
cates isomerization reactions of the standards during storage and processing. However, 13Z-
astaxanthin was not the most abundant diastereomer observed in these experiments, indicat-
ing that stereolability is dependent on the specific isomer and conditions [103]. Enzymolysis
was performed at 37˚C; elevated temperatures have been reported to enforce isomerization in
carotenoids [58, 67, 102, 104–107]. Therefore, the enzymolytic reaction itself bears the poten-
tial for further isomerization. These significant differences between the abundance of various
isomers under different conditions, especially the solvents used, indicate that a proper calcula-
tion of the isomers is complicated. Most astaxanthin determination methods require its solu-
tion in one or more solvents and multiple reaction steps, which might shift the proportion of
the isomers. Consequently, isomer ratios can be compared within one method to estimate vari-
abilities, but statements beyond cannot be made without additional tests. Changes in the pro-
portions of stereoisomers might be interesting for various applications as diastereomers have
been described to exhibit different antioxidant activity in vitro [108] and variable bioavailabil-
ity [79]. Therefore, their exact determination should be studied further.
For astaxanthin quantification, in the following experiments, the linear regression of the
combined results of the enzymolyzed free-astaxanthin and astaxanthin monopalmitate was
used as described above. To account also for the isomers, their peak areas were included.
Therefore, their peak areas were multiplied by correction factors to integrate their different
extinction properties. These were 1.20 for 9Z-astaxanthin, 1.56 for 13Z-astaxanthin, and 1.11
for the di-Z-astaxanthin isomers based on the extinction coefficients of Bjerkeng et al. [79].
All obtained areas were summed and the resulting calibration curve was f(x) = 1759.9x (R2 =
0.9983, n = 18). All different diastereomers were also found in extracts from H. pluvialis
(Fig 3).
It can be concluded that the amount of astaxanthin is the limiting factor for enzymolysis or
liquid-liquid extraction. A sensible measurement range is 1.0 to 11.2 μg/mL of total astax-
anthin concentration in the extraction phase. Moreover, applying the respective calibration
curve is essential to keep the quantification error below 5%.

Method development
Method development was based on dried H. pluvialis powder. All relevant process steps, i.e.,
disruption and extraction, enzymolysis, liquid-liquid extraction, and the processing of the liq-
uid-liquid extracts, were investigated and adapted if needed. Subsequently, detection limits,
linearity, and precision of the obtained method were determined (Fig 1).
Disruption of H. pluvialis and astaxanthin extraction. Astaxanthin was extracted from
H. pluvialis biomass by disrupting the cells with a bead mill in the presence of acetone. This
process was superior to grinding and ultrasonication under various conditions. A complete
decolorization of the cell pellet, as an indicator for complete astaxanthin extraction, was only
achieved by bead milling.
Enzymolysis. The enzymolysis is the core of the established method. Proper enzyme func-
tion and turnover must be ensured for complete deesterification of the various astaxanthin
esters and free astaxanthin recovery. To find an optimal enzyme concentration, its amount
and incubation time were varied, and free astaxanthin was measured (Fig 1b).

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 12 / 34

PLOS ONE Astaxanthin extraction and determination

Fig 3. UV/Vis and SIR chromatograms of an enzymolyzed H. pluvialis extract and corresponding absorption spectra of astaxanthin diastereomers.
https://doi.org/10.1371/journal.pone.0278504.g003

The variation of cholesterol esterase (Fig 1, b1) from 0.05 to 2.00 units added to a constant
quantity of 0.4 mg H. pluvialis biomass per sample resulted in calculated total astaxanthin pro-
portions between 0.87±0.30% w/w (n = 3) and 4.20±0.03% w/w (n = 3) (Fig 4 and S1 Table).
Respectively, 0.003 units to 0.12 units had been present per μg total free astaxanthin. The
resulting all-E-astaxanthin concentrations in the petroleum ether-acetone phase were between
1.33 and 6.29 μg/mL, thus in the linear calibration range. The increasing astaxanthin content
with a higher enzyme concentration indicates substrate conversion by cholesterol esterase in
the presence of acetone. Cholesterol esterase has been shown to work in a wide pH- and tem-
perature range and to maintain its activity in the presence of solvents [109]. Furthermore, an
activity increase was reported when minor proportions of up to 10% v/v of organic solvents
were added [110]. Thus, the enzyme seems appropriate for astaxanthin deesterification in the
described solvent-rich environment. The lowest astaxanthin values, accompanied by a distinct
increase in the calculated astaxanthin concentration and comparatively high standard devia-
tions (0.13–0.57% w/w), were observed between 0.05 and 0.50 units of cholesterol esterase.
Here, enzymolysis was incomplete. Astaxanthin content recovered from treatment with 1.0
unit cholesterol esterase did not differ significantly from 1.5 units. Still, 2.0 units resulted in
significantly more total astaxanthin, however, with a small effect size compared to 1.5 units
(0.16 percentage points). Therefore, 1.0 to 2.0 units hydrolyzed the majority of astaxanthin
esters. Su et al. reported good deesterification results with 4.0 units of cholesterol esterase per
reaction. However, they did not specify the used biomass amount [85]. For lipases, slower

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 13 / 34

PLOS ONE Astaxanthin extraction and determination

Fig 4. Astaxanthin content in 0.4 mg H. pluvialis biomass, enzymolyzed with varying amounts of cholesterol
esterase. Standard deviation is indicated for total astaxanthin in triplicates; 1.0 unit was a duplicate. Standard
deviations for the individual isomers can be found in S1 Table.
https://doi.org/10.1371/journal.pone.0278504.g004

conversion rates and smaller efficiencies have been reported. 4.6 U/μg carotenoids have
resulted in 63.2% free astaxanthin recovery after 7 hours of incubation. Moreover, the recovery
even decreased when more enzyme was applied [93]. Huang et al. showed the highest free
astaxanthin yields of 80% with 80 units per μg of astaxanthin esters with a recombinant lipase
after one hour of incubation [94]. Working with higher enzyme amounts may secure complete
conversion, especially at higher astaxanthin concentrations. Nevertheless, astaxanthin solubil-
ity in subsequent liquid-liquid extraction is limited, and higher enzyme application is not nec-
essarily more beneficial. Consequently, 0.06 to 0.12 units per μg of total free astaxanthin were
considered sufficient for most samples. This corresponded to 1.0 to 2.0 units of cholesterol
esterase per sample or 2.5 to 5.0 units per mg of H. pluvialis biomass in these experiments. A
further optimization might be achieved by using a higher enzyme concentration and simulta-
neously repeating the liquid-liquid extraction process as performed by Moretti et al. [61].
9Z-, 13Z- and two di-Z-isomers of astaxanthin were observed in all samples. Their biomass
proportions were 0.21±0.01% w/w (n = 3), 0.19±0.02% w/w (n = 3), and 0.18±0.01% w/w
(n = 3), respectively, at 2.0 units. Regarding their proportions to total astaxanthin, no signifi-
cant difference was observed in the di-Z-isomers when cholesterol esterase concentration was
varied. Increasing cholesterol esterase from 0.05 to 2.0 units caused the proportions of 9Z- and
13Z- to total astaxanthin to rise from 4.35 to 4.89% and from 3.56 to 4.47%, respectively. 9Z-
astaxanthin was equally or less abundant than in the enzymolyzed standards, whereas 13Z-
astaxanthin and the di-Z-isomers were detected in higher proportions. Higher or equal diaste-
reomer concentrations should be assumed in natural samples than in pure all-E-astaxanthin
standards. The lack of 9Z-astaxanthin in this natural sample might indicate that sample com-
position influences the isomer equilibria, as it is more complex than the standards. Further
isomerization to other isomers or even reverse isomerization to all-E-astaxanthin [103, 111]
during enzymolysis and extraction is possible. Compared to the standards, a higher proportion
of 13Z-astaxanthin and its di-Z-isomers might be assumed in this sample. However, their
abundance might partially also arise from isomerization reactions. Besides the already

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 14 / 34

PLOS ONE Astaxanthin extraction and determination

mentioned effects of temperature and solvents on isomerization, the sample matrix is more
complex due to other cell components from H. pluvialis. NaCl and iodine have also been
described to catalyze the isomerization behavior of carotenoids [67, 79, 103]. Therefore, reli-
able quantification of their total amounts is not possible. These results can only be seen as an
estimate for the calculation of total astaxanthin.
The duration of enzymolysis was extended from 0.75 to 1.5 hours. 0.5 and 2.0 units of cho-
lesterol esterase were applied to constant biomass of 0.4 mg dried H. pluvialis (Fig 1, b2). No
significant difference in the measured all-E-astaxanthin amount was observed compared to the
shorter incubation time at both enzyme concentrations. This indicates that increasing enzyme
concentration is more efficient than elongating incubation time. Additionally, studies showed
a tendency of free astaxanthin to decrease when enzymolysis was prolonged [85], which fur-
ther encourages shorter incubation at higher enzyme concentrations. With longer incubation
time, the proportion of 9Z-astaxanthin and the di-Z-isomers decreased by 9–18%, whereas
13Z-astaxanthin remained constant. Degradation and/or isomerization of astaxanthin isomers
during and after deesterification is possible but seems less severe compared to alkaline saponi-
fication [85]. Although longer enzymolysis duration may still result in good all-E-astaxanthin
recoveries, as also shown by Su et al. [85], relatively higher cholesterol esterase levels at shorter
incubation time can result in sufficient total astaxanthin yields without the risk of further
isomerization reactions. 2.0 units of cholesterol esterase and 0.75 hours of incubation time can
be used as an initial approach toward the astaxanthin measurement of an unknown sample,
which can be adapted if necessary. Further improvement of enzymolysis might be achieved by
surface-active detergents or bile acids, as indicated by Uwajima et al. [109], but might be lim-
ited due to interaction with the used solvents.
Liquid-liquid extraction. After enzymolysis, free astaxanthin had to be recovered from
the buffer solution. Therefore, liquid-liquid extraction with 2 mL of petroleum ether was per-
formed. After phase separation, the upper layer increased in volume (Fig 1, c2). This volume
was determined at 2.31+0.03 mL (n = 4). Acetone was dissolved in petroleum ether during
phase mixing, resulting in an excess volume [112]. Various studies encountered similar prob-
lems by drying and resolving the combined extracts in solvent [61, 83]. This procedure is more
laborious and prone to astaxanthin losses. As the volume of the petroleum ether phase is cru-
cial for the exact quantification of astaxanthin, the increased volume was used to calculate
astaxanthin content in all samples processed likewise. Moreover, sodium sulfate has been
used [61, 83] to remove residual water from the petroleum ether phase and enhance solvent
strength. However, this is unnecessary when the described boundaries for astaxanthin solubil-
ity are observed and dry or highly concentrated samples are used.
Processing of liquid-liquid extracts. The extract obtained from liquid-liquid extraction
can be injected directly into UHPLC-PDA-MS or further processed (Fig 1d). Various authors
evaporated the upper extraction phase and redissolved the extracts in organic solvents prior to
measurement [61, 83]. These methods were compared to evaluate recoveries and find the most
cost and time extensive procedure. Direct analysis of this phase resulted in a recovery of 98.24
±1.43% (n = 4) total astaxanthin. Further processing resulted in a decrease in recovery. An ali-
quot was evaporated and redissolved in the same volume of acetone. Its astaxanthin content
was quantified with the respective calibration for astaxanthin in acetone. Total astaxanthin
recovery was significantly lower with 91.68±1.32% (n = 4). A repeated extraction, evaporation,
and resumption in acetone resulted in 96.34±0.68% (n = 3) total astaxanthin recovery. The
generally lower recovery might be due to losses during processing. The proportion of 9Z-astax-
anthin was lower, whereas 13Z-astaxanthin and the di-Z-isomers were more abundant when
the samples were redissolved in acetone (S1 Table). This direct comparison and comparison to
the standards for calibration curve determination demonstrate changing isomer recoveries in

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 15 / 34

PLOS ONE Astaxanthin extraction and determination

different solvents, indicating once more that a valid determination is not possible due to
changes during processing. Altogether, these experiments point to direct processing as the
most correct and time and cost-reducing procedure. Still, the quantification of the diastereo-
mers in natural samples of H. pluvialis, especially in the presence of other carotenoids, is more
precise when the samples are dissolved in acetone, as the chromatographic resolution of the
diastereomers and lutein was higher.
Detection limits and linearity of astaxanthin determination. The complex composition
of H. pluvialis biomass was assumed to influence enzymolysis. Moreover, the previously deter-
mined detection limits should be verified when applying natural samples. Therefore, the bio-
mass quantity of H. pluvialis was varied from 0.04 to 3.98 mg at constant saponification time
(0.75 hours) and cholesterol esterase concentration (2.0 units) [99] to find an optimal astax-
anthin transformation range (Fig 1, a2). A maximum yield of 3.74±0.01% w/w (n = 3) all-E-
astaxanthin was observed at 0.8 mg of H. pluvialis powder (≙ 13.0 μg/mL all-E-astaxanthin
concentration in liquid-liquid extract) (Fig 5 and S1 Table). Between 0.2 and 2.0 mg (≙ 3.18
and 32.4 μg/mL all-E-astaxanthin concentration in liquid-liquid extract), � 97% of the maxi-
mum recovery were still reached. This implies an extended measurement range compared to
the standards, whose recovery decreased significantly, starting at approximately 11.2 μg/mL
astaxanthin concentration. The sample matrix of whole-cell biomass is much more complex
than the standards. Lipophilic cellular components such as fatty acids or other carotenoids are
dissolved in the hydrophobic solvent, altering its composition and enhancing solubility for
other constituents like astaxanthin. Z-isomers of many carotenoids, including astaxanthin,
have a higher solubility in various solvents than the all-E-isomer [104, 113, 114]. Thus, the cali-
bration curve might be applied at astaxanthin concentrations above 11.2 μg/mL. However, a
definite range cannot be specified because biomass composition, especially in terms of fatty
acids and pigments, is variable, particularly when H. pluvialis is exposed to stress conditions
[88, 115–117]. The measured all-E-astaxanthin content decreased significantly, applying either
more or less biomass. It was 3.33±0.05 (n = 3), 3.24, and 2.86% w/w for 0.04, 3.26, and 3.98 mg
of biomass, respectively (≙ 0.64, 51.81, and 63.25 μg/mL all-E-astaxanthin concentration in

Fig 5. Astaxanthin content in H. pluvialis biomass, deesterified with 2.0 units of cholesterol esterase. Standard
deviation is indicated for total astaxanthin in triplicates. Standard deviations for the individual isomers can be found in
S1 Table. The figure demonstrates that the method gets similar results for a broad range of biomass inputs.
https://doi.org/10.1371/journal.pone.0278504.g005

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 16 / 34

PLOS ONE Astaxanthin extraction and determination

liquid-liquid extract). The decrease at 0.04 mg is likely due to measuring inaccuracy at the
lower border of the measurement range, which was also determined in the standards. The sig-
nificant decrease in recovery when using more than 2.0 mg of biomass might be explained by a
relatively too low enzyme concentration. Moreover, other carotenoid and cholesterol esters
and phospho- and triacylglycerides are in direct competition with the deesterification of astax-
anthin-esters and may decelerate the process [118–121]. Jacobs et al. reported higher conver-
sion for carotenoid esters that contain a cyclopentenyl terminal ring rather than a
cyclohexenyl terminal ring, implying other carotenoid esters might be hydrolyzed preferably
[91]. Moreover, on the fatty acid side, a higher hydrolysis rate has been shown for longer chain
and polyunsaturated fatty acids esterified with cholesterol [109]. This might further favor reac-
tions of the enzyme with other molecules. Additionally, proteins might inhibit proper enzyme
function [122–124]. Another reason might be insufficient solubility of astaxanthin in the
extraction phase. However, for none of the samples, precipitation was visible. The proportion
of 9Z- and 13Z-astaxanthin increased significantly only when relating the samples with the
lowest and highest biomass (0.04 and 2.0 mg). However, it did not vary significantly when
samples with similar concentrations were compared. Moreover, the proportion of all diaste-
reomers increased when the biomass amount was raised to 3.3 and 4.0 mg. These corre-
sponded to all-E-astaxanthin concentrations of 51.81 and 63.25 μg/mL, which are well above
the upper detection limit. Thus, the higher solubility of the Z-isomers [104, 113, 114] might
change the equilibrium in their favor.
This experiment expanded the outlined upper detection limit of 11.2 μg/mL of all-E-astax-
anthin in the extraction phase to approximately 30 μg/mL. The lower detection limit was
approved at approximately 1.0 μg/mL.
Precision of astaxanthin determination. In order to demonstrate measuring precision
and repeatability, astaxanthin content was measured multiple times in two different samples.
Each measurement was performed on a different day to prove the similarity of the indepen-
dent experiments. The samples had an all-E-astaxanthin content of 3.46±0.04% w/w (n = 5)
and 3.98±0.04% w/w (n = 4) (Fig 6 and S1 Table). This equaled coefficients of variation of

Fig 6. Astaxanthin content in two H. pluvialis powders. Standard deviation is indicated for each astaxanthin
diastereomer. Sample A: n = 5, sample B: n = 4.
https://doi.org/10.1371/journal.pone.0278504.g006

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 17 / 34

PLOS ONE Astaxanthin extraction and determination

1.1% and 1.0%, respectively. The isomers 9Z-, 13Z-astaxanthin, and two di-Z-isomers were
observed in both samples with coefficients of variation between 1.9 and 9.0%. Thus, the preci-
sion of all-E-astaxanthin measurement was high, whereas it fluctuated for the diastereomers. It
might be improved by evaporating and resolving the extracts in acetone prior to analysis.
However, such treatment would be most likely at the expense of total astaxanthin recovery.

Method adjustment
The resulting method was further validated regarding linearity, detection limits, precision, and
robustness when various samples of astaxanthin containing biomass or extracts were used.
Therefore, fresh H. pluvialis cultures and either pure oleoresins or oleoresins diluted in ethanol
were applied at various concentrations. The sample type specific method adjustments and vali-
dations are explained in the following.
Sample type: H. pluvialis liquid cultures. H. pluvialis cells concentrated in nutrient-
depleted medium were disrupted in the presence of acetone. The first passage in the bead mill
had only little effect on the extraction of carotenoids, as hardly any color change of the
medium-acetone phase was observed. Major cell disruption and extraction were achieved only
in the second and third passage. This might be due to an increased acetone proportion during
the passages, which has a stronger dewatering effect on membranes and cell walls.
Biomass of four batches of H. pluvialis was processed according to the standard method
(Fig 1, a4). All-E-astaxanthin content was determined ranging from 0.73±0.04% w/w (n = 3) to
3.82±0.15% w/w (n = 3) (Fig 7 and S1 Table). Independently from the actual astaxanthin con-
centration, the obtained all-E-astaxanthin recovery in three of the four batches was above 90%
of the maximum achieved concentration when using 0.2 to 1.5 mg of biomass, respectively 2.5
to 15.0 μg/mL of all-E-astaxanthin extract concentration. These results confirm the applicabil-
ity of the linear calibration correlation of all-E-astaxanthin at concentrations above 11.2 μg/
mL for natural samples. However, the maximum all-E-astaxanthin recovery was observed at
different extract concentrations in each sample. These extended over the whole measurement

Fig 7. Astaxanthin content of four batches of H. pluvialis suspended in residual medium. M1 to M4 derive from
different batches. Standard deviations are indicated for total astaxanthin. Replicates of M1 to M4 vary between three
and five. Standard deviations for the individual isomers and replicates can be found in S1 Table.
https://doi.org/10.1371/journal.pone.0278504.g007

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 18 / 34

PLOS ONE Astaxanthin extraction and determination

range and were found at 2.5, 8.2, 11.3, and 19.7 μg/mL. Thus, the previously determined mea-
surement range should be condensed to an optimal measurement range, which is highly
dependent on the sample. Throughout its life cycle, H. pluvialis exhibits different characteris-
tics, e.g., highly stressed cells with a high astaxanthin level also indicate an altered composition
regarding fatty acids, carotenoids, and other cell compounds [44, 68, 88, 115–117, 125], which
might further change enzyme activity, solvent equilibria, and astaxanthin solubility.
In most samples, the measured isomers essentially showed the same proportion when
between 0.75 and 1.75 mg of H. pluvialis biomass (dry weight) were used. Applying less than
0.75 mg biomass resulted in higher deviations of the measurements. This reinforces the
assumption that the proportion of isomers measured is largely stable within a constant experi-
mental setup. In relation to the total astaxanthin content, 2.65 to 14.54% di-Z-astaxanthin, 2.96
to 5.03% 9Z-astaxanthin, and 3.38 to 4.17% 13Z-astaxanthin were observed. Yuan and Chen,
who compared the different isomers in extracts of H. pluvialis during and after saponification
under alkaline conditions, observed a higher peak area ratio of 9Z- to 13Z-astaxanthin [55–
57]. Gong et al. also observed higher 9Z- than 13Z-astaxanthin proportions after enzymolysis
[125], indicating a higher abundance of 9Z-astaxanthin in general. However, changes in the
composition of isomers have been observed during H. pluvialis cultivation [125], leading to
the suggestion that strain, cultivation, and stress conditions influence the abundance of the dif-
ferent isomers, as evidenced by the high variance of di-Z-isomers in these experiments. As
already discussed, storage and processing conditions might also have an influence, and the
comparison to other methods is presumably biased.
Compared to the share of geometrical isomers observed in the deesterified all-E-astaxanthin
standards, 9Z-astaxanthin was less or equally abundant, and 13Z- as well as the di-Z-astax-
anthin isomers were more abundant in samples of H. pluvialis suspended in medium. Thus,
the same principle that biomass, temperature, and solvents affect isomerization behavior
applies as already described previously for dried H. pluvialis samples.
Sample type: Oleoresins. Oleoresins from H. pluvialis are mostly obtained by SC-CO2
extraction. Here, modifiers are often used for enhanced extractability [89, 126–128] and due to
system requirements. Ethanol, as a solvent, is very common and can be found in the resulting
extracts. For astaxanthin determination, it can be evaporated prior to enzymolysis. For cost
and time reduction, direct measurement of ethanolic extracts was investigated using the
method developed here. Its robustness and effects on further processing and astaxanthin
recovery were examined.
The proper function of the enzymolysis in the presence of ethanol had to be ensured (Fig
1, c1). No significant difference was observed between the samples processed with ethanol
and those processed without ethanol with respect to all-E-astaxanthin. An extract with a high
astaxanthin concentration resulted in 52.9±0.4μg/mL (n = 5) of all-E-astaxanthin when equal
or less than 18% v/v ethanol were present during enzymolysis and extraction. 53.1±0.5 μg/
mL (n = 2) of all-E-astaxanthin were detected when ethanol was evaporated prior to process-
ing (S2 Table). Another extract with a lower astaxanthin concentration resulted in 8.6
±0.2 μg/mL (n = 3) of all-E-astaxanthin when equal or less than 18% v/v ethanol was present
during deesterification. Comparably 8.8±0.2 μg/mL (n = 2) of all-E-astaxanthin were
obtained when ethanol was evaporated prior to processing. As already described, cholesterol
esterase has been shown to maintain its activity in the presence of solvents and, more specifi-
cally, ethanol [109]. It might even work better if small amounts of up to 10% v/v of organic
solvents are added [110]. Ethanol concentrations above 18% v/v resulted in a sharp decline
of the measured all-E-astaxanthin content. It decreased to 50% and 16% of the previously cal-
culated astaxanthin quantities for 36% v/v and 54% v/v of ethanol, respectively. A decrease in
activity and stability with � 10% v/v ethanol concentrations has been observed [110]. A

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 19 / 34

PLOS ONE Astaxanthin extraction and determination

similar effect has been shown for a lipase from P. aeruginosa [95]. It can be explained by sta-
bilization and destabilization of hydrophobic interactions depending on ethanol concentra-
tion [129–131], suggesting an inhibition of the enzyme with ethanol concentrations � 18%
v/v in these experiments. There was no significant difference in the proportion of 13Z-astax-
anthin comparing the deesterification in the presence and without ethanol. However, with-
out ethanol, the proportion of 9Z-astaxanthin was 18 and 40% higher in the samples with a
higher or lower total astaxanthin content, respectively. In the sample with less astaxanthin,
the proportion of the di-Z-astaxanthin isomers was 60% higher in the absence of ethanol.
Solvent-related isomerization might account for this. Overall, the all-E-astaxanthin amount
did not change significantly. Thus, minor ethanol proportions in the enzymolytic process
can be regarded as unproblematic.
After enzymolysis, in the liquid-liquid extraction, the upper phase expansion was studied in
the presence of various quantities of ethanol (Fig 1, c2). After processing and extracting with
petroleum ether, the upper-phase volume decreased to 2.24, 2.17, 2.11, 2.04, and 2.00 mL
when 3, 7, 11, 14, and 18% v/v ethanol were used, respectively. Linear regression resulted in a
line of best fit with f(x) = -0.0003x + 2.3052 (R2 = 0.9943, n = 6). Binary mixtures of hexane
and ethanol cause a volume dilatation [132], whereas mixtures of acetone and ethanol result in
volume contraction [133, 134]. This second phenomenon might also impact the ternary mix-
ture of petroleum ether, acetone, and ethanol, resulting in a smaller excess volume than in the
absence of ethanol [135]. When 36% v/v ethanol were initially added to the sample, this linear
correlation no longer applied as the volume of the upper phase decreased to 1.95 mL. Here the
influence of a higher ethanol proportion resulted in a volume contraction.
It can be concluded from both experiments that the presence of ethanol generally does not
influence the total astaxanthin recovery as long as its effect on the volumetric change in liquid-
liquid extraction is taken into account during quantification and the volumetric maximum of
18% v/v is considered.
To study the optimal concentration range, detection limits, and linearity of astaxanthin
determination in oleoresins (Fig 1, a4), three different samples were deesterified with 2.0 units
to 4.0 units of cholesterol esterase at quantities between 0.15 and 15.8 mg. Above 90% of the
maximum measured astaxanthin content were recovered when using 0.3 to 1.2 mg of oleo-
resin. This corresponded to 1.8 μg/mL to 20.4 μg/mL of all-E-astaxanthin in the respective
extract. There was a sample-specific optimal measuring range. Exceeding or falling below it
resulted in reduced recoveries.
In oleoresin O1, the highest all-E-astaxanthin value of 5.8% w/w was observed at 0.4 mg
oleoresin (≙ 10.0 μg/mL all-E-astaxanthin in the liquid-liquid extract) enzymolyzed with 4.0
units of cholesterol esterase. Stepwise increase of oleoresin to 1.6 mg (≙ 41.2 μg/mL all-E-astax-
anthin) and reducing the enzyme to 2.0 units resulted in a steady 17% decrease of overall all-E-
astaxanthin recovery. This indicates an excess of enzyme capacity and/or solubility at higher
oleoresin amounts. Oleoresins mainly contain fatty acids and other lipophilic compounds
such as carotenoids and their esters [39, 101, 136], which can act as competitive substrates.
In oleoresin O2, which was oleoresin O1 diluted in sunflower oil, the highest all-E-astax-
anthin levels were observed between 0.75 and 1.00 mg of oleoresin (≙ 4.4 to 5.9 μg/mL all-E-
astaxanthin in the liquid-liquid extract). Less astaxanthin was detected when applying more or
less oleoresin. A doubling of enzyme concentration in sample O2 resulted in a minor increase
of maximum 5.5% all-E-astaxanthin or 3.9% total astaxanthin when directly comparing samples
with equal or less than 1.0 mg (≙ 5.9 μg/mL in the liquid-liquid extract) oleoresin. Therefore,
the enzyme capacity was sufficient to deesterify astaxanthin in the range below 5.9 μg/mL.
Conversely, in sample O3, the highest all-E-astaxanthin value of 3.6% w/w was observed at
the highest applied oleoresin per sample, which was 1.65 mg (≙ 25.9 μg/mL all-E-astaxanthin

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 20 / 34

PLOS ONE Astaxanthin extraction and determination

in the liquid-liquid extract) and decreased to 3.0% w/w at 0.78 mg oleoresin (≙ 12.3 μg/mL all-
E-astaxanthin in the liquid-liquid extract). This supports the hypothesis that complex lipo-
philic samples can be used at concentrations exceeding the previously determined measure-
ment range. Possibly the higher abundance of lipophilic substances results in enhanced
solubility of astaxanthin. For proper enzymatic activity, the previously outlined maximum
astaxanthin concentration of approximately 30 μg/mL should still be considered, or enzyme
quantity has to be increased.
No severe influence of the amount of oleoresin used on the proportion of the isomers was
observed in the range below 1.8 mg of sample. 7.51 to 19.69% of the total astaxanthin were the
di-Z-forms, 5.14 to 13.01% 9Z-astaxanthin, and 3.99 to 7.24% 13Z-astaxanthin. Their abun-
dance was higher than in the alga cultures suspended in medium or dried cells. Two effects
might account for this. First, elevated temperatures and high pressures, also present during
SC-CO2 extraction, lead to a higher isomerization rate and altered isomer profiles [58, 102,
104–107]. However, SC-CO2 extraction is a process that is considered mild. Álvarez et al. did
not find significant differences in isomer proportions between different extraction conditions
[137]. Second, Z-isomers might have higher extraction rates in SC-CO2 extraction a priori due
to their higher solubility in solvents [104, 113, 114]. This might be the reason for the selective
accumulation of Z-isomers [138, 139].
In this study, significant higher proportions of 9Z- than 13Z-astaxanthin were detected in
all oleoresins (Fig 8 and S1 Table). Various authors have also reported this in enzyme deesteri-
fied supercritical fluid extracts of H. pluvialis [85, 137]. All-E-astaxanthin and all-E-astaxanthin
diacetate exposed to isomerization inducing conditions showed higher levels of 9Z- and 13Z-
astaxanthin than other di-Z-isomers in the resulting mixtures [58, 60].
Sample mixtures. To further quantify the observed effects of natural samples on the solu-
bility and thus linear measurement range of astaxanthin in liquid-liquid extracts, different
standards and a natural sample with equal astaxanthin amount were examined independently
and in mixtures (Fig 1, a5). The all-E-astaxanthin recoveries of the independent samples were

Fig 8. Astaxanthin content of three different oleoresins of H. pluvialis. Standard deviations are indicated for total
astaxanthin. O1 (n = 3) and O3 (n = 3) derive from different batches. O2 (n = 7) is a dilution of O1. Standard
deviations for the individual isomers can be found in S1 Table.
https://doi.org/10.1371/journal.pone.0278504.g008

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 21 / 34

PLOS ONE Astaxanthin extraction and determination

88% from astaxanthin monopalmitate, 83% from free astaxanthin, and 98% from H. pluvialis.
The calculated astaxanthin concentration in the final extraction phase was between 4 and 5 μg/
mL. The addition of acetone after liquid-liquid extraction might have caused the unexpected
reduced recovery of the standards. Mixtures of those samples yielded recoveries of 95% to
109%, even at a total all-E-astaxanthin concentration of 13.3 μg/mL, indicating that the pres-
ence of biomass components or acetone facilitates the solution of astaxanthin in the extraction
phase. Evaluating all these samples, a proportional increase of the measured astaxanthin con-
centration to the prepared astaxanthin concentration was observed: a linear regression forced
through zero resulted in a proportion of 1.0205 of measured to prepared total astaxanthin con-
centration with a coefficient of determination of 0.9654. However, the recovery decreased
when free all-E-astaxanthin standard dissolved in acetone was added after the enzymolysis.
The same samples prepared with 3.6 μg and 7.2 μg free all-E-astaxanthin standard resulted in a
correlation of 0.9408 (R2 = 0.9690) and 0.8836 (R2 = 0.9441), respectively. The addition of
astaxanthin and acetone after extraction might change the phase equilibria, density, and astax-
anthin solubility, resulting in the observed measurement inaccuracy. Apart from these rela-
tions, all-E-astaxanthin extract concentrations up to 23.6 μg/mL were detected linearly without
precipitation effects. Moreover, those samples containing the whole-cell biomass or mixtures
resulted in higher recoveries in all experiments. Thus, fatty acids and other cellular compo-
nents are the only other differing factors in these experiments that might influence astaxanthin
solubility. It is difficult to discriminate between the two observed effects of reduced and
enhanced recovery due to acetone/standard addition and the presence of cellular components,
respectively, due to the experimental setup. To understand the circumstances affecting all-E-
astaxanthin solubility, recovery, and thus measurement boundaries for its quantification, a
broader concentration range without further astaxanthin addition after processing needs to be
considered, as all previous tests showed that the applicable measurement concentration of all-
E-astaxanthin in H. pluvialis biomass is higher than in measurements of standards.

Method comparison to photometric astaxanthin estimation

Many methods for fast and easy astaxanthin determination are based on a simple photometric
measurement of H. pluvialis extracts. Two such approaches using different mathematical mod-
els were compared to the developed UHPLC-PDA method to evaluate the discrepancy
between them and conclude whether photometric techniques can still lead to a reasonable esti-
mate of astaxanthin (Fig 1e). For an even broader impression of the possible validity of those
methods, two H. pluvialis cultures were tested at different times towards the end of a cultiva-
tion period to account for possible changes in the carotenoid and chlorophyll composition.
Astaxanthin production had been induced by exposure to high light intensities and nitrogen
starvation. Besides astaxanthin, the lutein content was determined by UHPLC-PDA measure-
ments, and spectra of the extracts were compared. UHPLC data revealed steadily increasing
all-E-astaxanthin content from 0.99 to 2.47% w/w and 1.31 to 2.70% w/w in batch A and B,
respectively (Fig 9 and S1 Table). Total astaxanthin concentration rose from 1.13 to 2.79% w/w
and from 1.50 to 3.07% w/w in batch A and B, respectively. Photometric data evaluated with
the general equation (Eq 1) resulted in total astaxanthin contents between 1.34 and 2.85% w/w
and between 1.97 and 2.95% w/w in batch A and B, respectively. Astaxanthin content calcu-
lated from the same photometric data but with the equations of Lichtenthaler et al. [100]
increased from 1.31 to 2.78% w/w in batch A and from 1.91 to 2.93% w/w in batch B. Astax-
anthin amounts obtained from the photometric methods were generally similar. Though, here
a higher astaxanthin content was estimated compared to UHPLC-PDA measurements in the
first half of both experiments. On the 27th and 28th day, photometric and UHPLC-PDA

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 22 / 34

PLOS ONE Astaxanthin extraction and determination

Fig 9. Astaxanthin content in two different batches of H. pluvialis cultures towards the end of a cultivation
period. Left batch A, right batch B. All-E-astaxanthin calculated from UHPLC-PDA measurements (◆), total
astaxanthin calculated from UHPLC-PDA measurements (◼), total astaxanthin calculated from a photometric
approach using Eq 1 (▲) and Eq 2 (●).
https://doi.org/10.1371/journal.pone.0278504.g009

approaches resulted in similar or slightly smaller astaxanthin concentrations calculated from

photometric data.
The initial deviation may be due to a misestimation of other carotenoids and chlorophylls
by the photometric approach. It has been shown that carotenoid composition changes during
cultivation; astaxanthin increases faster than other carotenoids, and lutein and chlorophylls
even decrease [44, 68, 125]. This was confirmed for lutein, as concentration fell from 0.09 to
0.06% w/w and 0.13 to 0.08% w/w in batch A and B, respectively, during the trial period. In
addition, the wavelength scans of the whole extracts revealed the presence of chlorophyll a by
maxima at 661 and 662 nm as well as a shoulder in the carotenoid peak at 434 nm [100], which
decreased over the measurement period. The extract spectra also revealed a shifting absorption
maximum from 466 to 472 nm and 468 to 478 nm in batch A and B, respectively. This λmax
shift was probably also due to a reduction of chlorophylls and possibly other carotenoids that
absorb at shorter wavelengths while astaxanthin levels increased. This experiment was made
towards the end of two cultivation periods and is in accordance with measurements of Bous-
siba et al. They measured absorbance spectra of extracts of H. pluvialis at different cell cycle
stages but observed a much greater variance when comparing green to red cells [75]. Lich-
tenthaler et al. especially considered and corrected for the chlorophylls in their equations, but
they are based on the analysis of various plants with another carotenoid composition and
without astaxanthin. Both photometric approaches assume that astaxanthin accounts for about
80% of all carotenoids, which is not true for various stress stages of H. pluvialis cells, and an
absorption coefficient A1% �
1cm of 1980 [100mL/(g cm)] for the total xanthophylls [100] was
assumed. Apart from the mentioned unsuitability for precise measurement of astaxanthin, it
might result in a further misestimation as it also does not distinguish between the diastereo-
mers. UHPLC-PDA analysis showed an increase in all diastereomers over the measurement
periods. The relative proportion of 9Z- and 13Z-astaxanthin increased slightly initially,
whereas the proportion of the di-Z-isomers decreased constantly. Gong et al. observed similar
correlations and reported a steady increase of total astaxanthin and its isomers during cultiva-
tion but no significant difference in the relative proportion of 9Z- and 13Z-astaxanthin during
the astaxanthin accumulation phase [125]. Both photometric approaches include the geometric
isomers. However, their different proportions and absorption coefficients are neglected. Maxi-
mum extinction is generally reduced and shifted to shorter wavelengths in Z-isomers of carot-
enoids [54, 79]. Therefore, the exact determination of astaxanthin, including the absorption
characteristics of its isomers, in complex matrices, with other carotenoids and chromophores

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 23 / 34

PLOS ONE Astaxanthin extraction and determination

of unknown portions and extinction properties is a delicate task. Recent methods circumvent
these problems by measuring at higher wavelengths (530 nm in DMSO), where the absorption
of most carotenoids, except for astaxanthin, is near zero [48, 59, 77]. Minor deviations might
still occur because of the named issues with geometrical isomers. Another possibility is to use
Gauss peak spectra methods to estimate carotenoids and astaxanthin [140, 141]. The different
isomer spectra can be taken into account here.
Photometric methods provided a reasonably good estimate of the total astaxanthin content
at a given time point near the end of a stress period. Only minor deviations between 0.2 and
4.0% from the UHPLC-PDA-based measurements occurred. However, comparison with sam-
ples taken at an earlier stage of cultivation resulted in deviations up to 30%, with the potential
for even greater discrepancy. Without further insights into the cell composition, such methods
should only be used to estimate differences in astaxanthin content of the same cultures. Astax-
anthin in differently cultivated H. pluvialis algae, strains, or even processed samples should not
be compared because photometric measurements cannot specifically distinguish between iso-
mers, other carotenoids, and chlorophylls.

Shelf life of astaxanthin containing samples

Astaxanthin-rich biomass is often stored prior to extraction and quantification. Generally,
astaxanthin is highly reactive. It reacts with oxygen and forms various auto-oxidation products
with shifted absorption maxima [142] and different absorption coefficients, leading to lower
extinction. Moreover, isomerization reactions of carotenoids are induced by incubation at ele-
vated temperatures also in the absence of solvents [106]. Although this effect generally depends
on the height of the temperature [58, 67, 102, 104, 105, 107], longer storage at lower tempera-
tures may result in similar isomerization effects. To achieve reproducible results, astaxanthin
losses should be minimized during storage. Therefore, the stability of free astaxanthin standard
in acetone was examined at various temperatures to characterize its vulnerability without fur-
ther protection. Furthermore, dried and non-dried H. pluvialis biomass was investigated at dif-
ferent temperatures and in the presence and absence of oxygen to find storage conditions that
reduce astaxanthin losses and storage costs.
Free astaxanthin standard. Free all-E-astaxanthin standard with a concentration of 1 μg/
mL in acetone was stored at -80˚C, -20˚C, 4˚C, and room temperature for 833 days in the
dark. At room temperature, a decrease of 6% and 12% after 10 days and 41 days was observed,
respectively. The concentration of samples stored in cooler conditions remained constant dur-
ing this period. After 833 days, the sample stored at room temperature had lost 93% of its ini-
tial concentration, and the sample stored at 4˚C had lost 27%. For storage at -80˚C and -20˚C
no major changes in optical density were observed. These results indicate partial prevention of
astaxanthin degradation by temperature reduction. This assumption is supported by other
studies in which increased degradation of astaxanthin was observed at elevated temperatures
[76, 143, 144].
Non-disrupted H. pluvialis biomass concentrated in residual medium. H. pluvialis bio-
mass suspended in a nutrient depleted liquid medium with a concentration of 40 g/L was
stored at 4˚C in the dark. The measured all-E-astaxanthin content decreased to 93 and 83% of
its initial value of 0.99±0.01% w/w (n = 6) when the samples were measured after 104 and 489
days, respectively. Similar protection of astaxanthin in frozen cells of H. pluvialis after 672 days
was observed by Miao et al., who reported a loss of less than 15% astaxanthin [76]. H. pluvialis
aplanospores have a rigid, multilayered cell wall that protects the alga from unfavorable envi-
ronmental conditions [145, 146], which might impair oxygen diffusion into the cell and thus
astaxanthin oxidation.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 24 / 34

PLOS ONE Astaxanthin extraction and determination

Lyophilized and non-disrupted H. pluvialis biomass exposed to ambient atmosphere.

An aliquot of the same biomass was lyophilized and stored under ambient air atmosphere at
4˚C. The all-E-astaxanthin content was 0.95±0.01% w/w (n = 3) directly after freeze-drying,
but it decreased to 69% and 32% of this initial value when the samples were measured 108 and
489 days later, respectively. These findings are similar to those of Ahmed et al., who reported
35% astaxanthin degradation in lyophilized H. pluvialis after 20 weeks in non-vacuum condi-
tions [144]. Again, the relatively high temperature and direct oxygen exposure probably pro-
moted astaxanthin degradation. The protective sheath of the cell wall might have failed due to
its desiccation compared to the previous experiments.
Lyophilized and non-disrupted H. pluvialis biomass partially exposed to ambient atmo-
sphere. Aliquots of the same lyophilized samples as before were vacuum-sealed directly after
freeze-drying. They were exposed to ambient air immediately, 7, 108, and 489 days later. These
samples were measured altogether on day 489. The all-E-astaxanthin recoveries were 32, 44, 70,
and 83% of the initial value, respectively (Fig 10 and S1 Table). The di-Z-isomers were generally
most abundant, followed by 9Z- and 13Z-astaxanthin. Their total amount decreased the longer
the sample was exposed to ambient air. However, their proportion in relation to the total astax-
anthin content was similar. Exposure to oxygen has been reported to negatively affect astax-
anthin during storage of H. pluvialis [76, 143, 144], e.g., Raposo et al. observed an improvement
in astaxanthin degradation in spray-dried samples when stored under nitrogen or vacuum
atmosphere compared to storage under air [71]. The small losses still observed in this experi-
ment might be due to reactions with residual oxygen in the cells or the atmosphere of the pack-
aging. Lyophilization was performed at 37 Pa, and no atmosphere change was applied.
Dried and disrupted H. pluvialis biomass sealed and exposed to ambient air. Two
sealed bags of H. pluvialis powder were stored at -21˚C in the dark. One of them was opened,
exposed to ambient air, and closed tightly again without changing the atmosphere. The astax-
anthin content of both samples was measured after two and a half years. The all-E-astaxanthin
content was 3.98±0.04% w/w (n = 4) and 3.46±0.04% w/w (n = 5) in the sealed and the opened
sample, respectively (S1 Table). Compared to the closed sample, 9Z- and 13Z-astaxanthin

Fig 10. Astaxanthin content in lyophilized H. pluvialis biomass. It was exposed to ambient atmosphere for different
periods of time and measured after 489 days.
https://doi.org/10.1371/journal.pone.0278504.g010

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 25 / 34

PLOS ONE Astaxanthin extraction and determination

decreased by 12% and 22%, respectively, in the opened sample, while the sum of the di-Z-iso-
mers increased by 14%. Related to total astaxanthin, all-E-, 9Z-, and 13Z-astaxanthin were not
significantly different, and only the proportion of the di-Z-isomers increased from 5.9±0.1
(n = 4) to 7.7±0.3% (n = 5). As described above, oxidation of astaxanthin is probably the reason
for its general decrease in the opened sample. Compared to the previously analyzed lyophilized
samples stored without a protective atmosphere at 4˚C, a beneficial effect of the reduced tem-
perature was observed, as total astaxanthin content decreased less.

Conclusion
A method for astaxanthin quantification was developed to accurately determine astaxanthin
from a variety of H. pluvialis biomass, extracts, and formulations. Specifically, the following
method parameters considered: Enzymolysis, extraction, and extract processing. Besides all-E-
astaxanthin, the diastereomers 9Z-, 13Z- and two di-Z-isomers of astaxanthin were detected.
In natural samples, the measurement precision of all-E-astaxanthin was determined with a
maximum coefficient of variation of 1.1%, whereas it was below 10% regarding the diastereo-
mers. Generally, linear correlations of biomass to astaxanthin content were determined in the
extraction phase between 1.8 μg/mL and up to 30 μg/mL all-E-astaxanthin. It was demon-
strated that an optimal concentration for quantification depended on the sample type and
composition. Optimal cholesterol esterase concentration was dependent on astaxanthin con-
centration and biomass composition, but 2.0 units were generally sufficient in the outlined
quantification range. The robustness of the method was demonstrated for ethanolic extracts of
H. pluvialis obtained from SC-CO2 extraction. Direct quantification from liquid-liquid
extracts was corrected for volume aberrations and dilatations during solvent mixing.
Based on our research, we recommend starting with between 0.5 and 2.0 mg of astax-
anthin-containing biomass. The initial experiment can be performed with 2.0 units of choles-
terol esterase and an incubation time of 0.75 hours. The settings of the enzymolysis can be
adapted depending on the type of the sample, i.e., fresh or dried biomass or extracts. For H.
pluvialis samples, the stress level and thus estimated astaxanthin content must be considered.
Different sample amounts should be processed to examine the approximate astaxanthin con-
centration and the influence of cellular components and to assure compliance with extraction
limits. Enzyme amount and concentration can be adapted in a second measurement set if nec-
essary. Special attention has to be paid to the correct calibration and quantification.

Supporting information
S1 Table. Overview of the astaxanthin content in the various experiments.
(PDF)
S2 Table. Overview of the astaxanthin content determined in ethanolic SC-CO2 extracts.
(PDF)

Acknowledgments
This study was supported by Sea & Sun Technology GmbH, especially Dr. Stefan Hindersin
and Clemens Elle, who provided various batches of differently processed H. pluvialis biomass.

Author Contributions
Conceptualization: Inga K. Koopmann, Annemarie Kramer, Antje Labes.
Data curation: Inga K. Koopmann.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 26 / 34

PLOS ONE Astaxanthin extraction and determination

Formal analysis: Inga K. Koopmann.

Funding acquisition: Antje Labes.
Investigation: Inga K. Koopmann.
Methodology: Inga K. Koopmann, Annemarie Kramer.
Project administration: Annemarie Kramer.
Supervision: Annemarie Kramer, Antje Labes.
Validation: Inga K. Koopmann.
Visualization: Inga K. Koopmann.
Writing – original draft: Inga K. Koopmann.
Writing – review & editing: Inga K. Koopmann, Annemarie Kramer, Antje Labes.

References
1. Britton G, Liaaen-Jensen S, Pfander H, editors. Carotenoids: Handbook. Basel: Springer Basel AG;
2004. https://doi.org/10.1007/978-3-0348-7836-4
2. National Center for Biotechnology Information. PubChem Compound Summary for CID 5281224,
Astaxanthin. https://pubchem.ncbi.nlm.nih.gov/compound/Astaxanthin. Accessed 07.06.2022.
3. Del Campo JA, Rodrı́guez H, Moreno J, Vargas MÁ, Rivas J, Guerrero MG. Accumulation of astax-
anthin and lutein in Chlorella zofingiensis (Chlorophyta). Appl Microbiol Biotechnol. 2004; 64: 848–
854. https://doi.org/10.1007/s00253-003-1510-5 PMID: 14689249
4. Abe K, Hattori H, Hirano M. Accumulation and antioxidant activity of secondary carotenoids in the
aerial microalga Coelastrella striolata var. multistriata. Food Chem. 2007; 100: 656–661. https://doi.
org/10.1016/j.foodchem.2005.10.026
5. Orosa M, Torres E, Fidalgo P, Abalde J. Production and analysis of secondary carotenoids in green
algae. J Appl Phycol. 2000; 12: 553–556.
6. Zhang DH, Lee YK, Ng ML, Phang SM. Composition and accumulation of secondary carotenoids in
Chlorococcum sp. J Appl Phycol. 1997; 9: 147–155.
7. Czeczuga B. Carotenoids in Euglena rubida mainx. Comp Biochem Physiol. 1974; 48B: 349–354.
8. Czeczuga B. Characteristic carotenoids in some phytobentos species in the coastal area of the Adri-
atic Sea. Acta Soc Bot Pol. 1986; 55: 601–609.
9. Müller T, Bleiß W, Martin CD, Rogaschewski S, Fuhr G. Snow algae from northwest Svalbard: their
identification, distribution, pigment and nutrient content. Polar Biol. 1998; 20: 14–32.
10. Procházková L, Remias D, Holzinger A, Řezanka T, Nedbalová L. Ecophysiological and ultrastructural
characterisation of the circumpolar orange snow alga Sanguina aurantia compared to the cosmopoli-
tan red snow alga Sanguina nivaloides (Chlorophyta). Polar Biol. 2021; 44: 105–117. https://doi.org/
10.1007/s00300-020-02778-0 PMID: 33519055
11. Shah MR, Liang Y, Cheng JJ, Daroch M. Astaxanthin-Producing Green Microalga Haematococcus
pluvialis: From Single Cell to High Value Commercial Products. Front Plant Sci. 2016; 7: 1–28. https://
doi.org/10.3389/fpls.2016.00531 PMID: 27200009
12. Kopecký J, Schoefs B, Loest K, Štys D, Pulz O. Microalgae as a source for secondary carotenoid pro-
duction: a screening study. Algol Stud. 2000; 98: 153–168. https://doi.org/10.1127/algol_stud/98/
2000/153
13. Pawar PR, Velani S, Kumari S, Lali AM, Prakash G. Isolation and optimization of a novel thraustochy-
trid strain for DHA rich and astaxanthin comprising biomass as aquafeed supplement. 3 Biotech. 2021;
11: 1–10. https://doi.org/10.1007/s13205-020-02616-4 PMID: 33489688
14. Aki T, Hachida K, Yoshinaga M, Katai Y, Yamasaki T, Kawamoto S, et al. Thraustochytrid as a poten-
tial source of carotenoids. J Am Oil Chem Soc. 2003; 80: 789–794. https://doi.org/10.1007/s11746-
003-0773-2
15. Park H, Kwak M, Seo J, Ju J, Heo S, Park S, et al. Enhanced production of carotenoids using a Thraus-
tochytrid microalgal strain containing high levels of docosahexaenoic acid-rich oil. Bioprocess Biosyst
Eng. 2018; 41: 1355–1370. https://doi.org/10.1007/s00449-018-1963-7 PMID: 29948212

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 27 / 34

PLOS ONE Astaxanthin extraction and determination

16. Yokoyama A, Izumida H, Miki W. Production of Astaxanthin and 4-Ketozeaxanthin by the Marine Bac-
terium, Agrobacterium aurantiacum. Biosci Biotechnol Biochem. 1994; 58: 1842–1844. https://doi.org/
10.1271/bbb.58.1842
17. Tsubokura A, Yoneda H, Mizuta H. Paracoccus carotinifaciens sp. nov., a new aerobic Gram-negative
astaxanthin-producing bacterium. Int J Syst Biol. 1999; 277–282.
18. Yokoyama A, Miki W, Izumida H, Shizuri Y. New trihydroxy-keto-carotenoids isolated from an astax-
anthin-producing marine bacterium. Biosci Biotechnol Biochem. 1996; 60: 200–203. https://doi.org/10.
1271/bbb.60.200 PMID: 27299394
19. Osanjo GO, Muthike EW, Tsuma L, Okoth MW, Lünsdorf H, Abraham W-R, et al. A salt lake extremo-
phile, Paracoccus bogoriensis sp. nov., efficiently produces xanthophyll carotenoids. Afr J Microbiol
Res. 2009; 3: 426–433.
20. Iizuka H, Nishimura Y. Microbiological studies on petroleum and natural gas. X. Carotenoid pigments
of hydrocarbon-utilizing bacteria. J Gen Appl Microbio. 1969; 15: 127–134.
21. Calo P, de Miguel T, Sieiro C, Velazquez JB, Villa TG. Ketocarotenoids in halobacteria: 3-hydroxy-
echinenone and trans-astaxanthin. J Appl Bacteriol. 1995; 79: 282–285. https://doi.org/10.1111/j.
1365-2672.1995.tb03138.x
22. Andrewes AG, Phaff HJ, Starr MP. Carotenoids of Phaffia rhodozyma, a red-pigmented fermenting
yeast. Phytochemistry. 1976; 1003–1007.
23. Tran TN, Tran Q-V, Huynh HT, Hoang N-S, Nguyen HC, Ngo D-N. Astaxanthin Production by Newly
Isolated Rhodosporidium toruloides: Optimization of Medium Compositions by Response Surface
Methodology. Not Bot Horti Agrobot Cluj-Napoca. 2018; 47: 320–327. https://doi.org/10.15835/
nbha47111361
24. Seybold A, Goodwin T. Occurrence of astaxanthin in the flower petals in Adonis annua. Nature. 1959;
184: 1714–1715.
25. Li Y, Gong F, Guo S, Yu W, Liu J. Adonis amurensis Is a Promising Alternative to Haematococcus as a
Resource for Natural Esterified (3S,30 S)-Astaxanthin Production. Plants. 2021; 10: 1–15. https://doi.
org/10.3390/plants10061059 PMID: 34070556
26. Di Mascio P, Kaiser S, Sies H. Lycopene as the most efficient biological carotenoid singlet oxygen
quencher. Arch Biochem Biophys. 1989; 274: 532–538. https://doi.org/10.1016/0003-9861(89)90467-
0 PMID: 2802626
27. Conn PF, Schalch W, Truscott TG. The singlet oxygen and carotenoid interaction. J Photochem
Photobiol B Biol. 1991; 11: 41–47. https://doi.org/10.1016/1011-1344(91)80266-k PMID: 1791493
28. Miki W. Biological functions and activities of animal carotenoids. Pure Appl Cem. 1991; 63: 141–146.
29. Shimidzu N, Goto M, Miki W. Carotenoids as Singlet Oxygen Quenchers in Marine Organisms. Fish
Sci. 1996; 62: 134–137. https://doi.org/10.2331/fishsci.62.134
30. Liu J, Zhang X, Sun Y, Lin W. Antioxidative capacity and enzyme activity in Haematococcus pluvialis
cells exposed to superoxide free radicals. Chin J Oceanol Limnol. 2010; 28: 1–9. https://doi.org/10.
1007/s00343-010-9244-6
31. Rodrigues E, Mariutti LRB, Mercadante AZ. Scavenging Capacity of Marine Carotenoids against
Reactive Oxygen and Nitrogen Species in a Membrane-Mimicking System. Mar Drugs. 2012; 10:
1784–1798. https://doi.org/10.3390/md10081784 PMID: 23015774
32. Liu X, Shibata T, Hisaka S, Osawa T. Astaxanthin inhibits reactive oxygen species-mediated cellular
toxicity in dopaminergic SH-SY5Y cells via mitochondria-targeted protective mechanism. Brain Res.
2009; 1254: 18–27. https://doi.org/10.1016/j.brainres.2008.11.076 PMID: 19101523
33. Chen Y-Y, Lee P-C, Wu Y-L, Liu L-Y. In Vivo Effects of Free Form Astaxanthin Powder on Anti-Oxida-
tion and Lipid Metabolism with High-Cholesterol Diet. PLOS ONE. 2015; 10: e0134733. https://doi.org/
10.1371/journal.pone.0134733 PMID: 26262684
34. Jiang X, Chen L, Shen L, Chen Z, Xu L, Zhang J, et al. Trans-astaxanthin attenuates lipopolysaccha-
ride-induced neuroinflammation and depressive-like behavior in mice. Brain Res. 2016; 1649: 30–37.
https://doi.org/10.1016/j.brainres.2016.08.029 PMID: 27559013
35. Xue Y, Qu Z, Fu J, Zhen J, Wang W, Cai Y, et al. The protective effect of astaxanthin on learning and
memory deficits and oxidative stress in a mouse model of repeated cerebral ischemia/reperfusion. Brain
Res Bull. 2017; 131: 221–228. https://doi.org/10.1016/j.brainresbull.2017.04.019 PMID: 28479214
36. Farruggia C, Kim M-B, Bae M, Lee Y, Pham TX, Yang Y, et al. Astaxanthin exerts anti-inflammatory
and antioxidant effects in macrophages in NRF2-dependent and independent manners. J Nutr Bio-
chem. 2018; 62: 202–209. https://doi.org/10.1016/j.jnutbio.2018.09.005 PMID: 30308382
37. Sharma K, Sharma D, Sharma M, Sharma N, Bidve P, Prajapati N, et al. Astaxanthin ameliorates
behavioral and biochemical alterations in in-vitro and in-vivo model of neuropathic pain. Neurosci Lett.
2018; 674: 162–170. https://doi.org/10.1016/j.neulet.2018.03.030 PMID: 29559419

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 28 / 34

PLOS ONE Astaxanthin extraction and determination

38. Brendler T, Williamson EM. Astaxanthin: How much is too much? A safety review. Phytother Res.
2019; 33: 3090–3111. https://doi.org/10.1002/ptr.6514 PMID: 31788888
39. European Food Safety Authority. Scientific Opinion on the safety of astaxanthin-rich ingredients
(AstaREAL A1010 and AstaREAL L10) as novel food ingredients. EFSA J. 2014; 12: 1–34.
40. Capelli B, Bagchi D, Cysewski GR. Synthetic astaxanthin is significantly inferior to algal-based astax-
anthin as an antioxidant and may not be suitable as a human nutraceutical supplement. Nutrafoods.
2013; 12: 145–152. https://doi.org/10.1007/s13749-013-0051-5
41. Edwards JA, Bellion P, Beilstein P, Rümbeli R, Schierle J. Review of genotoxicity and rat carcinogenic-
ity investigations with astaxanthin. Regul Toxicol Pharmacol. 2016; 75: 5–19. https://doi.org/10.1016/j.
yrtph.2015.12.009 PMID: 26713891
42. Lorenz RT, Cysewski GR. Commercial potential for Haematococcus microalgae as a natural source of
astaxanthin. Trends Biotechnol. 2000; 18: 160–167. https://doi.org/10.1016/S0167-7799(00)01433-5
43. Olaizola M. Commercial production of astaxanthin from Haematococcus pluvialis using 25,000-liter
outdoor photobioreactors. J Appl Phycol. 2000; 12: 499–506.
44. Torzillo G, Goksan T, Faraloni C, Kopecky J, Masojı́dek J. Interplay between photochemical activities
and pigment composition in an outdoor culture of Haematococcus pluvialis during the shift from the
green to red stage. J Appl Phycol. 2003; 15: 127–136. https://doi.org/10.1023/A:1023854904163
45. Aflalo C, Meshulam Y, Zarka A, Boussiba S. On the relative efficiency of two- vs. one-stage production
of astaxanthin by the green alga Haematococcus pluvialis. Biotechnol Bioeng. 2007; 98: 300–305.
https://doi.org/10.1002/bit.21391 PMID: 17318905
46. Rao R, Sarada AR, Baskaran V, Ravishankar GA. Identification of Carotenoids from Green Alga Hae-
matococcus pluvialis by HPLC and LC-MS (APCI) and Their Antioxidant Properties. J Microbiol Bio-
technol. 2009; 19: 1333–1341.
47. Wang J, Han D, Sommerfeld MR, Lu C, Hu Q. Effect of initial biomass density on growth and astax-
anthin production of Haematococcus pluvialis in an outdoor photobioreactor. J Appl Phycol. 2013; 25:
253–260. https://doi.org/10.1007/s10811-012-9859-4
48. Liyanaarachchi VC, Nishshanka GKSH, Premaratne RGMM, Ariyadasa TU, Nimarshana PHV, Malik
A. Astaxanthin accumulation in the green microalga Haematococcus pluvialis: Effect of initial phos-
phate concentration and stepwise/continuous light stress. Biotechnol Rep. 2020; 28: e00538. https://
doi.org/10.1016/j.btre.2020.e00538 PMID: 33294401
49. Miao F, Lu D, Li Y, Zeng M. Characterization of astaxanthin esters in Haematococcus pluvialis by liquid
chromatography–atmospheric pressure chemical ionization mass spectrometry. Anal Biochem. 2006;
352: 176–181. https://doi.org/10.1016/j.ab.2006.03.006 PMID: 16597431
50. Holtin K, Kuehnle M, Rehbein J, Schuler P, Nicholson G, Albert K. Determination of astaxanthin and
astaxanthin esters in the microalgae Haematococcus pluvialis by LC-(APCI)MS and characterization
of predominant carotenoid isomers by NMR spectroscopy. Anal Bioanal Chem. 2009; 395: 1613–
1622. https://doi.org/10.1007/s00216-009-2837-2 PMID: 19466394
51. Grewe C, Griehl C. Time- and media-dependent secondary carotenoid accumulation in Haematococcus
pluvialis. Biotechnol J. 2008; 3: 1232–1244. https://doi.org/10.1002/biot.200800067 PMID: 18683169
52. Miao F, Lu D, Zhang C, Zuo J, Geng Y, Hu H, et al. The synthesis of astaxanthin esters, independent
of the formation of cysts, highly correlates with the synthesis of fatty acids in Haematococcus pluvia-
lis. Sci China C Life Sci. 2008; 51: 1094–1100. https://doi.org/10.1007/s11427-008-0141-6 PMID:
19093083
53. Harker M, Tsavalos AJ, Young AJ. Autotrophic growth and carotenoid production of Haematococcus
pluvialis in a 30 liter air-lift photobioreactor. J Ferment Bioeng. 1996; 82: 113–118. https://doi.org/10.
1016/0922-338X(96)85031-8
54. Zechmeister L. Cis-trans Isomerization and Stereochemistry of Carotenoids and Diphenyl-polyenes.
Chem Rev. 1944; 34: 267–344. https://doi.org/10.1021/cr60108a004
55. Yuan J-P, Chen F. Purification of trans-astaxanthin from a high-yielding astaxanthin ester-producing
strain of the microalga Haematococcus pluvialis. Food Chem. 2000; 68: 443–448.
56. Yuan J-P, Chen F. Hydrolysis Kinetics of Astaxanthin Esters and Stability of Astaxanthin of Haemato-
coccus pluvialis during Saponification. J Agric Food Chem. 1999; 47: 31–35. https://doi.org/10.1021/
jf980465x PMID: 10563844
57. Yuan J-P, Chen F. Chromatographic Separation and Purification of trans-Astaxanthin from the
Extracts of Haematococcus pluvialis. J Agric Food Chem. 1998; 46: 3371–3375. https://doi.org/10.
1021/jf980039b
58. Euglert G, Vecchi M. trans/cis Isomerization of Astaxanthin Diacetate/Isolation by HPLC. and Identifi-
cation by 1H-NMR. Spectroscopy of Three Mono-cis- and Six Di-cis-Isomers. Helv Chim Acta. 1980;
63: 1711–1718. https://doi.org/10.1002/hlca.19800630640

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 29 / 34

PLOS ONE Astaxanthin extraction and determination

59. Casella P, Iovine A, Mehariya S, Marino T, Musmarra D, Molino A. Smart Method for Carotenoids
Characterization in Haematococcus pluvialis Red Phase and Evaluation of Astaxanthin Thermal Sta-
bility. Antioxidants. 2020; 9: 1–17. https://doi.org/10.3390/antiox9050422 PMID: 32414186
60. Subramanian B, Tchoukanova N, Djaoued Y, Pelletier C, Ferron M, Robichaud J. Investigations on the
geometrical isomers of astaxanthin: Raman spectroscopy of conjugated polyene chain with electronic
and mechanical confinement: Investigations on the geometrical isomers of astaxanthin. J Raman
Spectrosc. 2014; 45: 299–304. https://doi.org/10.1002/jrs.4459
61. Moretti VM, Mentasti T, Bellagamba F, Luzzana U, Caprino F, Turchini GM, et al. Determination of
astaxanthin stereoisomers and colour attributes in flesh of rainbow trout (Oncorhynchus mykiss) as a
tool to distinguish the dietary pigmentation source. Food Addit Contam. 2006; 23: 1056–1063.
62. Renstrøm B, Borch G, Skulberg OM, Liaaen-Jensen S. Optical purity of (3S,3’S)-astaxanthin from
Haematococcus pluvialis. Phytochemistry. 1981; 20: 2561–2564.
63. Zhao X, Liu H, Zhang X, Zhang G, Zhu H. Astaxanthin from Haematococcus pluvialis Microencapsu-
lated by Spray Drying: Characterization and Antioxidant Activity. J Am Oil Chem Soc. 2019; 96: 93–
102. https://doi.org/10.1002/aocs.12170
64. Sarada R, Vidhyavathi R, Usha D, Ravishankar GA. An Efficient Method for Extraction of Astaxanthin
from Green Alga Haematococcus pluvialis. J Agric Food Chem. 2006; 54: 7585–7588. https://doi.org/
10.1021/jf060737t PMID: 17002425
65. Boussiba S, Vonshak A. Astaxanthin Accumulation in the Green Alga Haematococcus pluvialis. Plant
Cell Physiol. 1991; 32: 1077–1082.
66. Kulikov EA, Kulikova IS, Vasilov RG, Selishcheva AA. The Effect of the Solvent Nature and Lighting on
Isomerization and Oxidative Degradation of Astaxanthin. Biophysics. 2020; 65: 433–442. https://doi.
org/10.1134/S0006350920030112
67. Yuan J-P, Chen F. Isomerization of trans-Astaxanthin to cis-Isomers in Organic Solvents. J Agric Food
Chem. 1999; 47: 3656–3660. https://doi.org/10.1021/jf981319u PMID: 10552699
68. Yuan J-P, Gong X-D, Chen F. Separation and Analysis of Carotenoids and Chlorophylls in Haemato-
coccus lacustris by High-Performance Liquid Chromatography Photodiode Array Detection. J Agric
Food Chem. 1997; 45: 1952–1956. https://doi.org/10.1021/jf970002b
69. Cifuentes AS, González MA, Vargas S, Hoeneisen M, González N. Optimization of biomass, total
carotenoids and astaxanthin production in Haematococcus pluvialis Flotow strain Steptoe (Nevada,
USA) under laboratory conditions. Biol Res. 2003; 36: 343–357. https://doi.org/10.4067/S0716-
97602003000300006 PMID: 14631867
70. Kang CD, Sim SJ. Selective Extraction of Free Astaxanthin from Haematococcus Culture Using a Tan-
dem Organic Solvent System. Biotechnol Prog. 2007; 23: 866–871. https://doi.org/10.1021/
bp0700354 PMID: 17567038
71. Raposo MFJ, Morais AMMB, Morais RMSC. Effects of spray-drying and storage on astaxanthin con-
tent of Haematococcus pluvialis biomass. World J Microbiol Biotechnol. 2012; 28: 1253–1257. https://
doi.org/10.1007/s11274-011-0929-6 PMID: 22805845
72. Dong S, Huang Y, Zhang R, Wang S, Liu Y. Four Different Methods Comparison for Extraction of
Astaxanthin from Green Alga Haematococcus pluvialis. Sci World J. 2014; 2014: 1–7. https://doi.org/
10.1155/2014/694305 PMID: 24574909
73. Jaime L, Rodrı́guez-Meizoso I, Cifuentes A, Santoyo S, Suarez S, Ibáñez E, et al. Pressurized liquids
as an alternative process to antioxidant carotenoids’ extraction from Haematococcus pluvialis microal-
gae. LWT—Food Sci Technol. 2010; 43: 105–112. https://doi.org/10.1016/j.lwt.2009.06.023
74. Reyes FA, Mendiola JA, Ibañez E, del Valle JM. Astaxanthin extraction from Haematococcus pluvialis
using CO2-expanded ethanol. J Supercrit Fluids. 2014; 92: 75–83. https://doi.org/10.1016/j.supflu.
2014.05.013
75. Boussiba S, Fan L, Vonshak A. Enhancement and determination of astaxanthin accumulation in green
alga Haematococcus pluvialis. Methods Enzymol. 1992; 213: 386–391.
76. Miao F, Geng Y, Lu D, Zuo J, Li Y. Stability and changes in astaxanthin ester composition from Hae-
matococcus pluvialis during storage. Chin J Oceanol Limnol. 2013; 31: 1181–1189. https://doi.org/10.
1007/s00343-013-2105-3
77. Li Y, Miao F, Geng Y, Lu D, Zhang C, Zeng M. Accurate quantification of astaxanthin from Haemato-
coccus crude extract spectrophotometrically. Chin J Oceanol Limnol. 2012; 30: 627–637. https://doi.
org/10.1007/s00343-012-1217-5
78. de Bruijn WJC, Weesepoel Y, Vincken J-P, Gruppen H. Fatty acids attached to all-trans-astaxanthin
alter its cis–trans equilibrium, and consequently its stability, upon light-accelerated autoxidation. Food
Chem. 2016; 194: 1108–1115. https://doi.org/10.1016/j.foodchem.2015.08.077 PMID: 26471660

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 30 / 34

PLOS ONE Astaxanthin extraction and determination

79. Bjerkeng B, Følling M, Lagocki S, Storebakken T, Olli JJ, Alsted N. Bioavailability of all-E-astaxanthin
and Z-isomers of astaxanthin in rainbow trout (Oncorhynchus mykiss). Aquaculture. 1997; 157: 63–
82. https://doi.org/10.1016/S0044-8486(97)00146-4
80. Orona-Navar A, Aguilar-Hernández I, Cerdán-Pasarán A, López-Luke T, Rodrı́guez-Delgado M, Cár-
denas-Chávez DL, et al. Astaxanthin from Haematococcus pluvialis as a natural photosensitizer for
dye-sensitized solar cell. Algal Res. 2017; 26: 15–24. https://doi.org/10.1016/j.algal.2017.06.027
81. Chen H-M, Meyers SP. A rapid quantitative method for determination of astaxanthin pigment concen-
tration in oil extracts. J Am Oil Chem Soc. 1984; 61: 1045–1047. https://doi.org/10.1007/BF02636215
82. An G-H, Schuman DB, Johnson EA. Isolation of Phaffia rhodozyma Mutants with Increased Astax-
anthin Content. Appl Environ Microbiol. 1989; 55: 116–124. https://doi.org/10.1128/AEM.55.1.116-
124.1989
83. Gomez PI, Inostroza I, Pizarro M, Perez J. From genetic improvement to commercial-scale mass cul-
ture of a Chilean strain of the green microalga Haematococcus pluvialis with enhanced productivity of
the red ketocarotenoid astaxanthin. AoB Plants. 2013; 5: plt026. https://doi.org/10.1093/aobpla/plt026
PMID: 23789055
84. Jin H, Lao YM, Zhou J, Zhang HJ, Cai ZH. Simultaneous determination of 13 carotenoids by a simple
C18 column-based ultra-high-pressure liquid chromatography method for carotenoid profiling in the
astaxanthin-accumulating Haematococcus pluvialis. J Chromatogr A. 2017; 1488: 93–103. https://doi.
org/10.1016/j.chroma.2017.01.088 PMID: 28179080
85. Su F, Xu H, Yang N, Liu W, Liu J. Hydrolytic efficiency and isomerization during de-esterification of nat-
ural astaxanthin esters by saponification and enzymolysis. Electron J Biotechnol. 2018; 34: 37–42.
https://doi.org/10.1016/j.ejbt.2018.05.002
86. Toomey MB, McGraw KJ. Modified Saponification and HPLC Methods for Analyzing Carotenoids from
the Retina of Quail: Implications for Its Use as a Nonprimate Model Species. Investig Opthalmology
Vis Sci. 2007; 48: 3976–3982. https://doi.org/10.1167/iovs.07-0208 PMID: 17724175
87. Galarza JI, Arredondo Vega BO, Villón J, Henrı́quez V. Deesterification of astaxanthin and intermedi-
ate esters from Haematococcus pluvialis subjected to stress. Biotechnol Rep. 2019; 23: e00351.
https://doi.org/10.1016/j.btre.2019.e00351 PMID: 31312607
88. Doria E, Temporiti MEE, Damiani MC, Popovich CA, Leonardi PI, Nielsen E. Influence of Light Stress
in the Accumulation of Xanthophylls and Lipids in Haematococcus Pluvialis Grown under Autotrophic
or Mixotrophic Conditions. J Mar Biol Aquac. 2018; 4: 30–35. https://doi.org/10.15436/2381-0750.18.
1799
89. Nobre B, Marcelo F, Passos R, Beirão L, Palavra A, Gouveia L, et al. Supercritical carbon dioxide
extraction of astaxanthin and other carotenoids from the microalga Haematococcus pluvialis. Eur
Food Res Technol. 2006; 223: 787–790. https://doi.org/10.1007/s00217-006-0270-8
90. Lietz G, Henry CJK. A modified method to minimise losses of carotenoids and tocopherols during HPLC
analysis of red palm oil. Food Chem. 1997; 60: 109–117. https://doi.org/10.1016/S0308-8146(96)00319-
6
91. Jacobs PB, LeBoeuf RD, McCommas SA, Tauber JD. The cleavage of carotenoid esters by choles-
terol esterase. Comp Biochem Physiol Part B Comp Biochem. 1982; 72: 157–160. https://doi.org/10.
1016/0305-0491(82)90027-X
92. Denery JR, Dragull K, Tang CS, Li QX. Pressurized fluid extraction of carotenoids from Haematococ-
cus pluvialis and Dunaliella salina and kavalactones from Piper methysticum. Anal Chim Acta. 2004;
501: 175–181. https://doi.org/10.1016/j.aca.2003.09.026
93. Zhao Y, Guan F, Wang G, Miao L, Ding J, Guan G, et al. Astaxanthin Preparation by Lipase-Catalyzed
Hydrolysis of Its Esters from Haematococcus pluvialis Algal Extracts. J Food Sci. 2011; 76: C643–
C650. https://doi.org/10.1111/j.1750-3841.2011.02119.x PMID: 22417348
94. Huang J, Yang Z, Zhu R, Qian X, Wang Y, Li Y, et al. Efficient heterologous expression of an alkaline
lipase and its application in hydrolytic production of free astaxanthin. Biotechnol Biofuels. 2018; 11: 1–
12. https://doi.org/10.1186/s13068-018-1180-2 PMID: 29983744
95. Nagao T, Fukami T, Horita Y, Komemushi S, Sugihara A, Shimada Y. Enzymatic enrichment of astax-
anthin from Haematococcus pluvialis cell extracts. J Am Oil Chem Soc. 2003; 80: 975–981. https://doi.
org/10.1007/s11746-003-0806-x
96. Dong H, Li X, Xue C, Mao X. Astaxanthin preparation by fermentation of esters from Haematococcus
pluvialis algal extracts with Stenotrophomonas species. Biotechnol Prog. 2016; 32: 649–656. https://
doi.org/10.1002/btpr.2258 PMID: 26949202
97. Lorenz T. HPLC and spectrophotometric analysis of carotenoids from Haematococcus algae powder.
2001 pp. 1–10. Report No.: BioAstin/NatuRose TM Technichal Bulletin.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 31 / 34

PLOS ONE Astaxanthin extraction and determination

98. Fuji Chemical Industry Co., Ltd. Spectrophotometric and HPLC Analysis Method for Determining
Astaxanthin Content in AstaRealR-P2AF. 1918 pp. 1–18.
99. Cyanotech Corporation. Analysis of Natural Astaxanthin Derived from Haematococcus Microalgae in
Astaxanthin Oleoresin, Astaxanthin Gelcaps, Astaxanthin Beadlets, and Haematococcus
Biomass. pp. 1–22.
100. Lichtenthaler HK. Chlorophylls and Carotenoids: Pigments of Photosynthetic Biomembranes. Methods
Enzymol. 1987; 148: 350–382.
101. Ruiz-Domı́nguez MC, Espinosa C, Paredes A, Palma J, Jaime C, Vı́lchez C, et al. Determining the
Potential of Haematococcus pluvialis Oleoresin as a Rich Source of Antioxidants. Molecules. 2019;
24: 1–17. https://doi.org/10.3390/molecules24224073 PMID: 31717936
102. Honda M, Takahashi N, Kuwa T, Takehara M, Inoue Y, Kumagai T. Spectral characterisation of Z-iso-
mers of lycopene formed during heat treatment and solvent effects on the E/Z isomerisation process.
Food Chem. 2015; 171: 323–329. https://doi.org/10.1016/j.foodchem.2014.09.004 PMID: 25308676
103. Zechmeister L. Cis-Trans Isomeric Carotenoids Vitamins A and Arylpolyenes. Vienna: Springer;
1962. https://doi.org/10.1007/978-3-7091-5548-6
104. Murakami K, Honda M, Takemura R, Fukaya T, Kubota M, Wahyudiono, et al. The thermal Z-isomeri-
zation-induced change in solubility and physical properties of (all-E)-lycopene. Biochem Biophys Res
Commun. 2017; 491: 317–322. https://doi.org/10.1016/j.bbrc.2017.07.103 PMID: 28735868
105. Murakami K, Honda M, Wahyudiono, Kanda H, Goto M. Thermal isomerization of (all-E)-lycopene and
separation of the Z-isomers by using a low boiling solvent: Dimethyl ether. Sep Sci Technol. 2017; 52:
2573–2582. https://doi.org/10.1080/01496395.2017.1374412
106. Aman R, Schieber A, Carle R. Effects of Heating and Illumination on Trans-Cis Isomerization and Deg-
radation of β-Carotene and Lutein in Isolated Spinach Chloroplasts. J Agric Food Chem. 2005; 53:
9512–9518. https://doi.org/10.1021/jf050926w PMID: 16302770
107. Honda M, Sowa T, Kawashima Y. Thermal- and Photo-Induced Isomerization of All-E - and Z-Isomer-
Rich Xanthophylls: Astaxanthin and Its Structurally-Related Xanthophylls, Adonirubin, and Adonix-
anthin. Eur J Lipid Sci Technol. 2020; 1900462: 1–9. https://doi.org/10.1002/ejlt.201900462
108. Liu X, Osawa T. Cis astaxanthin and especially 9-cis astaxanthin exhibits a higher antioxidant activity
in vitro compared to the all-trans isomer. Biochem Biophys Res Commun. 2007; 357: 187–193. https://
doi.org/10.1016/j.bbrc.2007.03.120 PMID: 17416351
109. Uwajima T, Terada O. Purification and Properties of Cholesterol Esterase from Pseudomonas fluores-
cens. Agric Biol Chem. 1976; 40: 1957–1964. https://doi.org/10.1080/00021369.1976.10862345
110. Takeda Y, Aono R, Doukyu N. Purification, characterization, and molecular cloning of organic-solvent-
tolerant cholesterol esterase from cyclohexane-tolerant Burkholderia cepacia strain ST-200. Extremo-
philes. 2006; 10: 269–277. https://doi.org/10.1007/s00792-005-0494-8 PMID: 16463077
111. Yuan J-P, Chen F. Kinetics for the reversible isomerization reaction of trans-astaxanthin. Food Chem.
2001; 73: 131–137. https://doi.org/10.1016/S0308-8146(01)00107-8
112. Qin A, Hoffman DE, Munk P. Excess volumes of mixtures of alkanes with carbonyl compounds. J
Chem Eng Data. 1992; 37: 55–61. https://doi.org/10.1021/je00005a018
113. Honda M, Kodama T, Kageyama H, Hibino T, diono W, Kanda H, et al. Enhanced solubility and
reduced crystallinity of carotenoids, β-carotene and astaxanthin, by Z-isomerization. Eur J Lipid Sci
Technol. 2018; 120: 1–8. https://doi.org/10.1002/ejlt.201800191
114. Honda M, Watanabe Y, Murakami K, Takemura R, Fukaya T, Wahyudiono, et al. Thermal isomeriza-
tion pre-treatment to improve lycopene extraction from tomato pulp. LWT. 2017; 86: 69–75. https://doi.
org/10.1016/j.lwt.2017.07.046
115. Saha SK, McHugh E, Hayes J, Moane S, Walsh D, Murray P. Effect of various stress-regulatory fac-
tors on biomass and lipid production in microalga Haematococcus pluvialis. Bioresour Technol. 2013;
128: 118–124. https://doi.org/10.1016/j.biortech.2012.10.049 PMID: 23196231
116. Sarada R, Tripathi U, Ravishankar GA. Influence of stress on astaxanthin production in Haematococ-
cus pluvialis grown under different culture conditions. Process Biochem. 2002; 37: 623–627. https://
doi.org/10.1016/S0032-9592(01)00246-1
117. Recht L, Zarka A, Boussiba S. Patterns of carbohydrate and fatty acid changes under nitrogen starva-
tion in the microalgae Haematococcus pluvialis and Nannochloropsis sp. Appl Microbiol Biotechnol.
2012; 94: 1495–1503. https://doi.org/10.1007/s00253-012-3940-4 PMID: 22361859
118. Howles PN, Hui DY. Cholesterol Esterase. 1st ed. In: Mansbach CM, Tso P, Kuksis A, editors. Intesti-
nal Lipid Metabolism. 1st ed. Boston, MA: Springer; 2001. pp. 119–134.
119. Chen Q, Sternby B, Nilsson A. Hydrolysis of triaclyglycerol arachidonic and linoleic acid ester bonds by
human pancreatic lipase and carboxyl ester lipase. Biochim Biophys Acta BBA—Lipids Lipid Metab.
1989; 1004: 372–385.

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 32 / 34

PLOS ONE Astaxanthin extraction and determination

120. Sugihara A, Shimada Y, Nomura A, Terai T, Imayasu M, Nagai Y, et al. Purification and Characteriza-
tion of a Novel Cholesterol Esterase from Pseudomonas aeruginosa, with Its Application to Cleaning
Lipid-stained Contact Lenses. Biosci Biotechnol Biochem. 2002; 66: 2347–2355.
121. Hui DY, Howles PN. Carboxyl ester lipase: structure-function relationship and physiological role in lipo-
protein metabolism and atherosclerosis. J Lipid Res. 2002; 43: 2017–2030. https://doi.org/10.1194/jlr.
r200013-jlr200 PMID: 12454261
122. Ben Ali Y, Carrière F, Verger R, Petry S, Muller G, Abousalham A. Continuous monitoring of choles-
terol oleate hydrolysis by hormone-sensitive lipase and other cholesterol esterases. J Lipid Res. 2005;
46: 994–1000. https://doi.org/10.1194/jlr.M400509-JLR200 PMID: 15716583
123. Gargouri Y, Julien R, Pieroni G, Verger R, Sarda L. Studies on the inhibition of pancreatic and micro-
bial lipases by soybean proteins. J Lipid Res. 1984; 25: 1214–1221. https://doi.org/10.1016/S0022-
2275(20)34465-5 PMID: 6542933
124. Gargouri Y, Julien R, Sugihara A, Verger R, Sarda L. Inhibition of pancreatic and microbial lipases by
proteins. Biochim Biophys Acta BBA—Lipids Lipid Metab. 1984; 795: 326–331. https://doi.org/10.
1016/0005-2760(84)90082-1
125. Gong F, Zhang C, Zhang L, Liu J. Changes of carotenoids contents and analysis of astaxanthin geo-
metrical isomerization in Haematococcus pluvialis under outdoor high light conditions. Aquac Res.
2020; 51: 770–778. https://doi.org/10.1111/are.14427
126. Krichnavaruk S, Shotipruk A, Goto M, Pavasant P. Supercritical carbon dioxide extraction of astax-
anthin from Haematococcus pluvialis with vegetable oils as co-solvent. Bioresour Technol. 2008; 99:
5556–5560. https://doi.org/10.1016/j.biortech.2007.10.049 PMID: 18068354
127. Machmudah S, Shotipruk A, Goto M, Sasaki M, Hirose T. Extraction of Astaxanthin from Haematococ-
cus pluvialis Using Supercritical CO2 and Ethanol as Entrainer. Ind Eng Chem Res. 2006; 45: 3652–
3657. https://doi.org/10.1021/ie051357k
128. Pan J-L, Wang H-M, Chen C-Y, Chang J-S. Extraction of astaxanthin from Haematococcus pluvialis
by supercritical carbon dioxide fluid with ethanol modifier: Supercritical CO2 fluid extraction of astax-
anthin from microalgae. Eng Life Sci. 2012; 12: 638–647. https://doi.org/10.1002/elsc.201100157
129. Amin MdA, Halder R, Ghosh C, Jana B, Bhattacharyya K. Effect of alcohol on the structure of cyto-
chrome C: FCS and molecular dynamics simulations. J Chem Phys. 2016; 145: 1–8. https://doi.org/10.
1063/1.4972065 PMID: 28010091
130. Nandi S, Parui S, Halder R, Jana B, Bhattacharyya K. Interaction of proteins with ionic liquid, alcohol
and DMSO and in situ generation of gold nano-clusters in a cell. Biophys Rev. 2018; 10: 757–768.
https://doi.org/10.1007/s12551-017-0331-1 PMID: 29147940
131. Deshpande A, Nimsadkar S, Mande SC. Effect of alcohols on protein hydration: crystallographic analy-
sis of hen egg-white lysozyme in the presence of alcohols. Acta Crystallogr D Biol Crystallogr. 2005;
61: 1005–1008. https://doi.org/10.1107/S0907444905009364 PMID: 15983424
132. Ormanoudis C, Dakos C, Panayiotou’ C. Volumetric Properties of Binary Mixtures. 2. Mixtures of n-
Hexane with Ethanol and I-Propanol. J Chem Eng Data. 1991; 36: 39–42.
133. Qin A, Hofmann DE, Munk P. Excess volumes of mixtures of some alkyl esters and ketones with alka-
nols. Collect Czech Chem Commun. 1993; 58: 2625–2641.
134. Miyano Y, Hayduk W. Solubilities of n-butane gas and densities for acetone-methanol, acetone-etha-
nol, and acetone-propanol solvent solutions. J Chem Eng Data. 1986; 31: 81–83. https://doi.org/10.
1021/je00043a023
135. Dortmund Data Bank, 2022, www.ddbst.com.
136. Schoefs B, Rmiki N-E, Rachadi J, Lemoine Y. Astaxanthin accumulation in Haematococcus requires a
cytochrome P450 hydroxylase and an active synthesis of fatty acids. FEBS Lett. 2001; 500: 125–128.
https://doi.org/10.1016/S0014-5793(01)02596-0
137. Álvarez CE, Vardanega R, Salinas-Fuentes F, Ramı́rez JP, Muñoz WB, Jiménez-Rondón D, et al.
Effect of CO2 Flow Rate on the Extraction of Astaxanthin and Fatty Acids from Haematococcus pluvia-
lis Using Supercritical Fluid Technology. Molecules. 2020; 25: 1–17.
138. Gamlieli-Bonshtein I, Korin E, Cohen S. Selective separation of cis-trans geometrical isomers of β-car-
otene via CO2 supercritical fluid extraction. Biotechnol Bioeng. 2002; 80: 169–174. https://doi.org/10.
1002/bit.10357 PMID: 12209772
139. Watanabe Y, Honda M, Higashiura T, Fukaya T, Machmudah S, Wahyudiono, et al. Rapid and Selec-
tive Concentration of Lycopene Z-isomers from Tomato Pulp by Supercritical CO2 with Co-solvents.
Solvent Extr Res Dev Jpn. 2018; 25: 47–57. https://doi.org/10.15261/serdj.25.47
140. Küpper H, Seibert S, Parameswaran A. Fast, Sensitive, and Inexpensive Alternative to Analytical Pig-
ment HPLC: Quantification of Chlorophylls and Carotenoids in Crude Extracts by Fitting with Gauss
Peak Spectra. Anal Chem. 2007; 79: 7611–7627. https://doi.org/10.1021/ac070236m PMID: 17854156

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 33 / 34

PLOS ONE Astaxanthin extraction and determination

141. Thrane J-E, Kyle M, Striebel M, Haande S, Grung M, Rohrlack T, et al. Spectrophotometric Analysis of
Pigments: A Critical Assessment of a High-Throughput Method for Analysis of Algal Pigment Mixtures
by Spectral Deconvolution. PLOS ONE. 2015; 10: e0137645. https://doi.org/10.1371/journal.pone.
0137645 PMID: 26359659
142. Etoh H, Suhara M, Tokuyama S, Kato H, Nakahigashi R, Maejima Y, et al. Auto-Oxidation Products of
Astaxanthin. J Oleo Sci. 2012; 61: 17–21. https://doi.org/10.5650/jos.61.17 PMID: 22188802
143. Niamnuy C, Devahastin S, Soponronnarit S, Vijaya Raghavan GS. Kinetics of astaxanthin degradation
and color changes of dried shrimp during storage. J Food Eng. 2008; 87: 591–600. https://doi.org/10.
1016/j.jfoodeng.2008.01.013
144. Ahmed F, Li Y, Fanning K, Netzel M, Schenk PM. Effect of drying, storage temperature and air expo-
sure on astaxanthin stability from Haematococcus pluvialis. Food Res Int. 2015; 74: 231–236. https://
doi.org/10.1016/j.foodres.2015.05.021 PMID: 28411988
145. Hagen C, Siegmund S, Braune W. Ultrastructural and chemical changes in the cell wall of Haemato-
coccus pluvialis (Volvocales, Chlorophyta) during aplanospore formation. Eur J Phycol. 2002; 37:
217–226. https://doi.org/10.1017/S0967026202003669
146. Damiani MC, Leonardi PI, Pieroni OI, Cáceres EJ. Ultrastructure of the cyst wall of Haematococcus
pluvialis (Chlorophyceae): wall development and behaviour during cyst germination. Phycologia.
2006; 45: 616–623. https://doi.org/10.2216/05-27.1

PLOS ONE | https://doi.org/10.1371/journal.pone.0278504 December 2, 2022 34 / 34