1. Introduction
Astaxanthin (3,3′-dihydroxy-β-carotene-4,4′-dione) is a carotenoid pigment, for which multifunctional health benefits and safety are reported widely [
1,
2]. While synthetic preparations dominate in the feed market, a natural astaxanthin is being exclusively used in pharmaceutical, cosmetic, and food industries [
3]. Only a few sources of natural astaxanthin are known, among which a fresh water green algae
Haematococcus pluvialis is the most preferred choice [
4].
Astaxanthin oleoresin derived from the algal biomass is represented by the mix of all-E-isomer and 9Z- and 13Z-isomers with all-E-form being a main component [
5]. There are multiple approaches to the quantitation of astaxanthin in samples by high-performance liquid chromatography (HPLC) using isocratic or gradient mobile phases in either normal-phase or reversed-phase mode [
6,
7,
8,
9,
10]. The availability of different analytical methods increases the measurement uncertainty; thus, the same sample can be measured as containing ±20% of astaxanthin, which understandably disrupts operations within the astaxanthin industry and the same astaxanthin level stated on a label varies significantly between producers.
The commonly used validated analytical procedure described in the United States Pharmacopoeia (USP) monograph (USP 42) suggests using a reversed-phase HPLC with a three-component gradient mobile phase (81% methanol, 15% t-butylmethylether, 4% phosphoric acid (1% aqueous)) for astaxanthin quantification [
11]. The procedure also defines different response factors for astaxanthin Z isomers—1.3 for 13Z and 1.1 for 9Z—at wavelength 474 nm. However, the verification of this method performed in the quality department showed low reproducibility of the method. On the one hand, low reproducibility is related to the presence of different astaxanthin isomers (13Z, 9Z, All-E, 15Z) and their variable relative ratio in the samples, as well as other carotenoids, which absorb light at wavelength 474 nm [
12]. Thus, a HPLC chromatogram contains many carotenoid peaks, which are located close to each other and can be partially merged depending on the HPLC system settings. This decreases a resolution between adjacent peaks and impairs astaxanthin quantification. On the other hand, the complexity of the analytical procedure—enzymatic hydrolysis of esters, extraction with petroleum ether—adds to the method uncertainty. However, laboratory trials recently demonstrated that light and heat are the main factors affecting astaxanthin analysis, with light being the most important. It was found that the same oleoresin sample dissolved in acetone in a transparent glass flask, therefore being exposed to ambient light, possesses higher UV/Vis absorbance than the sample protected from light by covering the flask with aluminium foil. Interestingly, this change happens only in an acetone solution of astaxanthin since direct illumination of oleoresin does not impact its absorbance. Moreover, it was noticed that the level of astaxanthin isomers is not the same within repeated measurements of the same sample and might be a key to understanding the observed light-induced increase of astaxanthin level. The observations are supported by scientific evidence about astaxanthin isomerization reactions [
5,
13,
14]. The conversion of trans-astaxanthin to cis-astaxanthin under heat impact [
13,
14,
15], as well as isomerization reactions of other carotenoids [
16,
17] have been demonstrated. However, the information regarding the light-induced increase of astaxanthin absorbance and transformation of its isomers in acetone solution is lacking.
Therefore, the aim of this study is to evaluate all effects of light on astaxanthin analysis, both spectrometric and HPLC, namely the growth of astaxanthin absorbance in an acetone solution and isomerization events depending on the carotenoid composition of samples, thus, collecting data for further revision, improvement, and eventually, harmonization of the astaxanthin analytical procedure within the industry.
3. Discussion
The study findings demonstrate the intrinsic behaviour of astaxanthin isomers in an acetone solution and confirm that its exposure to light changes the UV absorbance of the solution quickly and significantly, which results in the increase of the recorded astaxanthin level. Apparently, this change of total absorbance is related to the isomerization events in an acetone solution caused by light, namely, the transformation of 13Z astaxanthin to All-E and 9Z astaxanthin. Indeed, the capacity of 13Z astaxanthin to absorb light is lower compared to All-E astaxanthin, meaning that the same amount of 13Z astaxanthin has lower absorbance than All-E. Thus, the sample which contains a higher level of 13Z astaxanthin is tested spectrometrically as having a lower astaxanthin percentage even if both samples have the same quantity of astaxanthin molecules. A similar but lower difference exists between 9Z and All-E astaxanthin. The HPLC method accounts for this difference by multiplying the HPLC peak areas of 13Z and 9Z astaxanthin by response factors 1.3 and 1.1 respectively [
11]. Therefore, because of the presence of different astaxanthin isomers with different absorbance rates and in different ratios in oleoresin, the UV method cannot be as precise as the HPLC method. Additionally, the isomerization reaction of astaxanthin under light adds a significant amount of overall uncertainty in the UV method since the astaxanthin level might fluctuate to up to 10% (
Figure 1). Indeed, there is no accurate UV/Vis approach here. Measurement under “no light” gives lower numbers because of the high level of 13Z astaxanthin. The light approach (analysis after at least 2 h of light exposure) is closer to the true value since at least 80% of the isomers are All-E astaxanthin and the calibration curve is made using the All-E astaxanthin standard. At the same time, the “no-light” UV/Vis approach is faster and gives more stable and reproducible numbers when the carotenoid matrix is the same. The carotenoid matrix influence on the analysis is clearly visible in
Figure 2, where the light-induced growth of astaxanthin level is different between the oleoresin samples.
The presence of astaxanthin isomers with different absorbance rates makes an HPLC method preferable for astaxanthin quantification, but at the same time, it is also significantly affected by light exposure of the acetone solution. Despite similar transformation events in all oleoresin samples, namely a decrease of the 13Z astaxanthin and an increase of the All-E and 9Z astaxanthin HPLC peak area, the light-induced changes are not equal in Astalif 10 samples and the reference standard. Since the standard absorbance increase speed is lower than in Astalif 10 samples, their quantification against the standard shows higher values than in the “no-light” approach (
Figure 3). For instance, the absolute absorbance increase in the oleoresin sample O10-200 was 17%, and the standard increased by 7% at the same time, thus, the percentage of O10-200 appears to be higher by approx. 10% after 3-h light exposure.
The differences in the carotenoid matrices must be the key factor of the isomerization reactions’ variability in the oleoresin samples. However, the only pronounced feature of the reference standard is significantly lower 9Z astaxanthin. It is not clear how the makeup of the carotenoid matrix makes the standard more resilient to light exposure.
It was also shown that the drop (300–400 absorbance units) in 13Z astaxanthin after light exposure does not align with the total increase (500–700 absorbance units) of All-E and 9Z astaxanthin in the Astalif 10 samples, which is almost two times higher. This may suggest that either the response coefficients of cis astaxanthin isomers are not correct, or there are other isomerization reactions in place. Indeed, the USP monograph method requires the use of response factors 1.3 and 1.1 for 13Z and 9Z astaxanthin, respectively, however, other alternative approaches state 1.6 and 1.13 as response factors. Thus, to improve the method, the response factors must be reassessed taking into account light-induced isomerization events.
It must be noted that while the increase of UV/Vis absorbance was 8–11%, the HPLC absorbance increased by 7–17%. A higher maximum increase of HPLC values is related not only to an increase of the total astaxanthin absorbance, but also to light-induced degradation of apocarotenal. Since total absorbance is calculated by dividing by the apocarotenal absorbance, lower apocarotenal values make total absorbance higher. Surprisingly, the apocarotenal level, which must be very similar in all samples, was, in fact, lower by 4% in the standard (
Figure 4). This makes its final values higher and consequently underestimates the Astalif 10 samples. It is unclear why this happens, but a plausible explanation may be the matrix effect on the degradation/isomerization of apocarotenal and/or on the extraction coefficient by petroleum ether. Moreover, the degradation profile of apocarotenal is similar between the Astalif 10 samples and the pure apocarotenal solution, but the curve slope of apocarotenal in the standard is lower (
Figure 6).
There are at least six more HPLC peaks on the chromatograms apart from the three mentioned astaxanthin peaks and apocarotenal. At the same time, the astaxanthin oleoresin sample’s carotenoid matrix includes other astaxanthin isomers (15Z, 9,9′-di-cis, 9,13′-di-cis, etc.) and other carotenoids, like β-carotene, lutein, zeaxanthin, and canthaxanthin [
12,
18,
19]. Their total level, however, does not exceed 10% in most cases. A slight decrease by 2–4% of these carotenoids’ absorbance in Astalif 10 samples after light exposure was established, thus, the samples appeared to have a higher astaxanthin purity (95% vs. 92%) if analyzed in ambient light. Individual carotenoid peaks changed even more drastically: peaks at 7.3–8.8 min reduced significantly, but peaks at 12.2–12.8 increased instead. Chemical reactions and their constants associated with these changes are yet to be established.
The study’s results explain the causes of the UV/Vis and HPLC method uncertainty and point out at flaws in the USP method. It also appears unclear which of the HPLC approaches yields results closer to a true value. Protecting samples from light reduces variability and makes it possible to record the true isomeric distribution, which is not affected by light. Therefore, the “no-light” approach is validated and used in the quality control department. At the same time, it increases the chance of underestimation of the oleoresin samples. Moreover, the EU regulation regarding astaxanthin quality establishes the level 13Z isomer as lower than 7% and all-E form as higher than 75% from other isomers [
20,
21]. Such a level is recorded only by “light” approach since the average 13Z astaxanthin relative level is 18–24%, and all-E is 60–72% from all astaxanthin isomers if measured in “no-light” conditions. Thus, the revision and improvement of an astaxanthin analysis is especially important for the harmonization between companies and regulation of the astaxanthin industry.
Underlying chemical mechanisms associated with light exposure need to be elucidated in further studies. Contrarily, heat-induced astaxanthin isomerization processes have already been published [
13,
14]. Interestingly, heat provokes opposite isomerization processes—conversion of trans-astaxanthin to cis-astaxanthin mainly to 13Z isomer associated with a drop of total absorbance [
22,
23]. Whether heating the samples to 37 °C for 45 min for hydrolysis of astaxanthin esters during sample preparation impact the isomerization process is presently unknown. However, only few studies attempted to investigate the light impact on E/Z isomerization [
5,
24]. Viazau et al. [
5] showed that under prolonged illumination, the content of both Z-isomers decreases. However, the study does not mention the conversion of cis-astaxanthin to trans-form instead explained the drop in Z-isomer by oxidation and the formation of epoxy- and apo-products [
5]. It must be noted that in our study, the illumination of the acetone solution was performed after hydrolysis of the esters during the HPLC preparation procedure. Contrarily, the esterified astaxanthin solution was exposed to light for the UV/Vis analysis. Esterification might have an impact on cis-trans equilibrium, as it was shown recently that free astaxanthin produces larger amounts of 9Z isomer and monopalmitate esterification results in increased 13Z isomerization instead [
25]. However, this study was performed on separate astaxanthin molecules, whereas algal-derived natural oleoresin contains a mixture of carotenoid molecules, thus, additional studies are needed to discover any difference in the light-induced cis to trans-astaxanthin conversion between free and esterified astaxanthin.
Developing a better astaxanthin quantification method requires the identification of all mechanisms of transformation between astaxanthin isomers, a revision of response factors for 13Z and 9Z astaxanthin, and elucidating why different carotenoid matrices behave differently after exposure to light. The conventional USP or similar approaches, however, lead to a high deviation of the measurement, thus, the difference of more than 10% between the laboratories is not surprising.