Introduction

Synthetic colourants are widely used in the beverages industry to: (i) enhance naturally occurring colours, (ii) to correct natural variations in colour during processing or storage, and (iii) to add colour to beverages that would otherwise be colourless (or coloured differently). The colour of foods and beverages has a striking effect on taste thresholds, acceptance, and palatability and is used by consumers to judge the safety, quality, and palatability of a product before consumption (Suzuki et al. 2017). Colour can also modulate palatability and comfort and even the thermal perception of foods and beverages (Suzuki et al. 2017). Moreover, food processors are increasingly interested in changing the colour of foods and beverages not only for the above-mentioned effects of colour on the product but also to surprise their clients or as a marketing strategy to attract new consumers (Spence 2019).

Food colourants can be classified according to different criteria; for example, they can be classified according to their origin as natural or synthetic. Synthetic food colourants are generally more stable than their natural counterparts are, but some of them have been associated with negative health outcomes in the past. Today, food colourants are subjected to a wide range of toxicity tests and strict legislative provisions (Amchova et al. 2015); their use is regulated by the Food and Drug Administration in the United States and by the European Food Safety Authority in the European Union. Those approved can be considered as safe. However, consumers try to avoid synthetic or artificial additives and ingredients and prefer foods and ingredients that are organic or natural (Asioli et al. 2017). Moreover, natural colourants are (sometimes) associated with positive health outcomes; for example, consumption of astaxanthin derived from natural sources has been associated with the a decrease in oxidative stress and inflammation and a promotion of the immune response of humans (Villaró et al. 2021). For this reason, the search for novel sustainable, safe and natural food colourants is a research line with great potential.

Microalgae are a valuable and trendy source of natural colourants. Natural pigments derived from microalgae are already commercially available. For example, β-carotene from Dunaliella salina and astaxanthin from Haematococcus pluvialis are sold not only for their colouring properties but also for their antioxidant capacity and potential for being used as functional foods (Lafarga 2019). Microalgae are not only rich in carotenoids but also chlorophylls (green) and phycobiliproteins which are proteins with linear tetrapyrrole prosthetic groups that absorb light and are involved in photosynthesis. Depending on their absorption properties, phycobiliproteins can be divided into three major groups: (i) phycoerythrins (PEs, λmax = 540–570 nm), (ii) phycocyanins (λmax = 610–620 nm), and (iii) allophycocyanins (λmax = 650–655 nm). All together these molecules form a macromolecular complex known as the phycobilisome. Their potential use as colourants in foods and beverages is very wide. The colour of phycocyanins varies from purple to deep blue, allophycocyanins are generally blue-green, and phycoerythrins generally appear pink or red (Bermejo et al. 2002) and they have the added advantage of being water-soluble. Phycoerythrins can be divided into R-phycoerythrins, B-phycoerythrins, or C-phycoerythrins depending on their absorption characteristics (Bermejo et al. 2003). These valuable pigments have applications in biomedical research, as bioactive molecules with antioxidant, immunomodulatory and anticarcinogenic activities, or as colourants because of their pinkish and reddish colour (Mercier et al. 2022). Pink and red colours are associated with sweet flavours (Spence 2019), probably because the association of redness and ripeness or maturity of fruits (Wadhera and Capaldi-Phillips 2014). This is one of the reasons why several naturally uncoloured products are coloured pink/red before commercialisation. For example, light pink tonic waters, isotonic drinks, and gins are common is supermarket shelves.

Extracts rich in B-phycoerythrin obtained from the red alga Porphyridium cruentum have been used as colouring agents in milk products mimicking the colour of strawberry-flavoured beverages and demonstrate a very high stability at a wide range of pH values and during storage (García et al. 2021b). Phycoerythrin can be considered safe; different studies have evaluated the potential toxicity of phycoerythrin using animal models and reported no mortality or treatment-related major clinical signs (Soni et al. 2010). The number of studies assessing the colouring capacity of natural B-phycoerythrin is very limited. The goal of this work was to assess the potential utilisation of a phycoerythrin-rich extract obtained from Porphyridium cruentum as a natural colourant in four different beverages, namely isotonic drinks, tonic water, gin, and wine. The selected beverages have a pH generally lower than 4.0 and are currently being commercialised with a pink colour. Their colour is achieved using synthetic colourants such as Ponceau 4R (E-124), amaranth (E-123), erythrosine (E-127) or Allura Red (E-129) and carminic acid (E-120). Despite being obtained from a natural resource, the dietary rules of vegans and several religions authorities do not allow consuming carminic acid which is derived from insects (Schweiggert 2018).

Materials and methods

Materials and reagents used

Ammonium sulphate, monosodium phosphate, disodium phosphate, and sodium azide were from Sigma Aldrich (Spain). The 3.5 kDa SnakeSkin dialysis tubing was from Thermo Scientific (USA). Commercial isotonic drinks, tonic water, gins, and wines were purchased from local retailers. The microalga used was Porphyridium cruentum UTEX 161, produced at the SABANA Demonstration Plant located (Almería, Spain) at the University of Almeria in Spain.

Extraction of phycoerythrin from Porphyridium cruentum

The phycobiliproteins were recovered following the method of García et al. (2021a) with minor modifications. Briefly, 2.0 ± 0.0 g of freeze-dried biomass of P. cruentum were suspended in 100 mL of 0.1 M phosphate buffer (pH 5.5) and stirred at 350 rpm for 15 min. Then, the suspension was centrifuged using a Medtronic BL-S centrifuge (JP Selecta, Barcelona, Spain) operating at 12,000 rpm and 4 ºC for 15 min and the supernatant was recovered. The supernatant was saturated to 60% using ammonium sulphate and stirred at 4 ºC for 24 h. The mixture was centrifuged and the precipitate was dialyzed using 3.5 kDa SnakeSkin dialysis tubing) against ultrapure water. 1% sodium azide (w/v) was added to the purified extract which was stored at 4 ºC in the dark until further use. Sodium azide was not added to the extracts used as food colourants.

Characterisation of the extract

To assess the effect of the pH on the stability of phycoerythrin, the extract was resuspended in 20 mM phosphate buffer (pH ranging from 1.0 to 13.0) at a concentration of 0.015 mg·mL−1. The lower and higher pH ranges were achieved by adding 1 M HCl or NaOH, respectively. The pH of the extracts was constant throughout the determinations with variations lower than 0.5% in all cases. The optical density of the solutions was measured at 200–700 nm and its fluorescence at was measured 550–700 nm using a FP-6500 spectrofluorometer (Jasco, Spain).

The concentrations of phycobiliproteins, namely phycocyanin \([R-PC]\), allophycocyanin \([APC]\), and phycoerythrin \([B-PE]\), were estimated using the equations:

$$[R-PC] \left(mg\;mL^{\mathit{-1}}\right)=\frac{{A}_{620}-0.7\cdot {A}_{650}}{7.38}$$
$$[APC]\left(mg\;mL^{\mathit{-1}}\right)=\frac{{A}_{650}-0.19\cdot {A}_{620}}{5.65}$$
$$[B-PE] \left(mg\;mL^{\mathit{-1}}\right)=\frac{{A}_{565}-2.8\cdot \left[R-PC\right]-1.34\cdot [APC]}{12.7}$$

where \({A}_{620}\), \({A}_{650}\), and \({A}_{565}\) is the absorbance measured at 620, 650, and 565 nm using a Lambda 20 spectrophotometer (Perkin Elmer, USA). Three technical determinations were taken per natural replicate.

Sample preparation

Four different types of commercial beverages were evaluated: isotonic drinks, tonic water, gin, and wine. The selected products were commercialised by the same brand either pink-coloured or colourless. The products were not processed upon reception except for the tonic water and the wine, which were degassed by agitation for 20 min using a HB502 magnetic stirrer (Bibby Scientific, UK) operated at 350 rpm. The beverages were stored at 4 ± 1 ºC until further use.

The colour of both pink-coloured and colourless products was determined according to the CIELAB colour space. Colour recordings were taken using a CM-5 chroma-meter (Konica Minolta, Japan) and the illuminant/observer pair D65/10°, which approximates to daylight. A few microliters of the phycobiliprotein-rich extract were added to the colourless samples and the colour change was registered. The addition of the extract was carried out under continuous agitation using a HB502 magnetic stirrer (Bibby Scientific) operated at 350 rpm. The colour difference was determined after the addition of the natural pigments using the equation:

$$\Delta E=\sqrt{{\Delta L}^{2}+{\Delta a}^{2}+{\Delta b}^{2}}=\sqrt{{({L}_{0}-{L}_{i})}^{2}+{({a}_{0}-{a}_{i})}^{2}+{({b}_{0}-{b}_{i})}^{2}}$$

where \({L}_{0}\), \({a}_{0}\), and \({b}_{0}\) refer to the colour parameters of the commercial pink products and \({L}_{i}\), \({a}_{i}\), and \({b}_{i}\) refer to the colour parameters of the beverages containing phycobiliproteins at each time point. The process was repeated until the pink colour was comparable to that of the commercial pink products and \(\Delta E\) was minimum. All the colour recordings were taken in triplicate per natural replicate. The staining factor was calculated as the total amount of protein per litre needed to obtain minimise \(\Delta E\).

The viscosity of the beverages was measured using a Cannon Fenske 50 U-tube viscometer (UK) as described elsewhere (García et al. 2021a). The viscosity measurements were carried out in triplicate per natural replicate on the same day that the samples were prepared.

Kinetic study of colorant stability as a function of temperature

The absorbance spectra of each sample was daily recorded for one month; then, the B-PE concentration as estimated. The first-order kinetic equation was fitted using these data and the kinetic parameters were obtained. The temperature-dependent degradation of the B-PE was calculated by a first-order exponential equation to give a degradation rate (K) at three different temperatures (4, 25, and 42 ºC).

$$Ln \left[B-PE\right]={Ln \left[B-PE\right]}_{0}-K\;t$$

where [B-PE] is the concentration of B-PE at the time t, [B-PE]0 is the initial concentration, and K is the degradation rate (day−1). The stability of the B-PE extract was analysed by using the half-life (t1/2) equation:

$${t}_{1/2 }=\frac{0.693}{K}$$

The samples were stored at the selected temperatures and in the dark for 30 days.

Sensory analysis

The commercial acceptance of the colour was carried out using 330 mL bottles in a CAC 60 colour assessment cabinet (VeriVide Limited, UK) and the D65 illuminant, which approximates to daylight. Both, the commercial products sold pink and the phycoerythrin-containing beverages were assessed by 60 semi-trained panellists who scored the acceptability of the colour using a 5-point hedonic scale (1: extremely dislike, 2: dislike, 3: nor like nor dislike, 4: like, and 5: extremely like).

Statistical analysis

The data were analysed using analysis of variance using SPSS Statistics 22 (IBM, USA). A Tukey HSD test was carried out to identify where the sample differences occurred. The criterion for statistical significance was p < 0.05.

Results

Characterisation of the B-PE extracts

Figure 1A shows the absorption and fluorescence spectra of the P. cruentum-derived extract. The absorbance showed a main peak at 545 nm, which indicates that B-PE was the main component in the extract. A smaller peak at 620 nm was also observed, demonstrating that the extract was not pure and R-PC was present in the sample. In the present work, the phycobiliprotein-rich extract contained mainly B-PE (79%) and R-PC (21%). Allophycocyanins were not detected in the extract. The B-PE concentration of the extract was 0.75 gB-PE·L−1, and the purity grade, defined as the ratio between the peak absorbance at 565 nm and the absorbance of proteins measured at 280 nm (\({A}_{565}/{A}_{280}\)), was 2.50. In terms of fluorescence, a maximum fluorescence intensity emission was observed at 575 nm (Fig. 1A).

Fig. 1
figure 1

(A) Absorption and fluorescence spectra of the B-PE extract from P. cruentum at pH = 7.0, (B) variation of absorbance at 545 nm and relative fluorescence intensity at 575 nm of the B-PE extract solution at different pH values, and (C) absorbance spectra of the B-PE extract at pH values of 2 and 11. The protein concentration was 15 mgB-PE·L−1

In the present work, the pH of the studied beverages ranged between 3.0 and 4.0, except for gin that had a pH of 6.5 (Table 1). The stability of the extract at different pH values is shown in Fig. 1B. Overall, the spectra maintained their profile and high values of absorbance and fluorescence intensity were measured at pH values between 4.0 and 9.0. At pH values lower than 3.0 and higher than 10.0, the stability of the proteins was significantly affected. This same effect was observed when the extract was assessed at highly acid and alkaline conditions (Fig. 1C). The extracts showed a peak absorbance at 545 nm and a second peak at 280 nm, which was caused by the total protein content of the extract. Although the absorbance measured at 545 nm was lower at pH values in the range 3.0–4.0 (p < 0.05), the spectrum remained the same than that measured at pH 7.0. The pH of the solution also affected the colour of the extract, which was brownish at pH values lower than 3.0 and colourless at pH values higher than 12. The L*, a*, and b* values of the extracts at different pH values are listed in Table 2. Overall, the L* and b* values of the B-PE extract were higher at pH values lower than 3.0 and higher than 10.0 and the opposite trend was observed for the a* value, which decreased at pH values lower than 2.5 and higher than 11.0 (p < 0.05).

Table 1 Staining factors of the studied beverages ([B-PE] extract = 0.75 mgB-PE·mL−1)
Table 2 Colour parameters of the PE-rich extract at different pH values. Results are expressed as mean values ± SD. The protein concentration was 0.015 mgB-PE·mL−1

Assessment of the pigmentation capacity of B-PE in commercial beverages

The pigmentation capacity of the generated extract was assessed in different commercial products, listed in Table 1. The pH of these products ranged between 3 and 4, except for the gin which had a pH of 6.5. The uncoloured commercial products were pigmented using different amounts of the B-PE extract (Fig. 2) and the amount of extract needed to minimise \(\Delta E\), the staining factor, was recorded (Table 1). Overall, the staining factors of the tonic waters and the gin were the lowest, being in both cases below 2 mg·L−1. The staining factor of both wines and of the isotonic drink was around 3 and 4.5 mg·L−1, respectively. The minimum \(\Delta E\) values reached depended on the beverage evaluated, being lower than 3 in the isotonic drink, the gin, and one of the wines. The maximum value achieved was 4.0, obtained after adding 3 mg of B-PE per litre of wine.

Fig. 2
figure 2

Colouring curves of the beverages resulting from the colouring assays. The protein concentration of the extract was 750 mgB-PE·L−1

The stability of the extracts was assessed at different temperatures, namely 4, 25 and 42 ºC. The stability of the B-PE extract was analysed by the decline in concentration from the initial value (day 0) over 30 days of incubation (Table 3). The greater stability was obtained for the gin sample (D). Overall, in the rest of the samples, the behaviour observed for the different samples was similar, observing average degradation times of around 18 and 3 days at 4 and 42 °C, respectively. When the samples were stored at 25 °C, the second most stable sample after the above-mentioned gin was the tonic water with a stability of 17.9 days. The stability of the colour of the B-PE-containing products was also evaluated during an 11-day cold storage period. Results, shown in Fig. 3 revealed that the L* value was not significantly affected, except for a slight increase after 11 days of storage in the wine and isotonic drink samples (p < 0.05). Similarly, the b* value of all the studied samples increased during storage and the opposite trend was observed between a* values, which decreased during storage. The \(\Delta E\) value, which refers to the change of colour of the samples during storage showed an increase with time, suggesting that the colour difference with respect to day 0 was higher every day. The maximum \(\Delta E\) values were achieved at day 11 and were lower than 3 for all the studied samples and lower than 1 for the tonic water and the gin (Table 4). Finally, the visual acceptance of the B-PE-containing products was assessed by 60 semi-trained consumers. No differences were observed between the acceptability score of the products commercialised pink and the B-PE-containing products. The acceptability index of the wines, tonic waters, isotonic drinks, and gins were 3.45, 3.63, 3.48 and 3.52, respectively (Fig. 4).

Table 3 Rates of degradation (K) and half-life (t½) of B-PE extract after 30 days incubation at 4, 25 and 42 ºC of the different beverages
Fig. 3
figure 3

Changes in (A) ΔE, (B) L*, (C) a*, and (D) b* during refrigerated storage at 4 ± 1 ºC in the dark

Table 4 Colour recordings after cold storage. Results are expressed as mean values ± SD
Fig. 4
figure 4

(A) Photographs of the beverages commercialised pink (commercial) and the products developed using B-PE (generated) and (B) acceptance scores of the test beverages assessed using 44 semi-trained panellists and a 5-point hedonic scale

The viscosity of the different beverages evaluated is described in Supplementary Material 1 (Table 1MS). Overall, the addition of the B-PE extracts at the concentrations evaluated in the present work did not affect the density or the viscosity of any of the studied beverages. The viscosity of the B-PE-containing products was also comparable to that of the products pigmented using synthetic pigments.

Discussion

Characterisation of the B-PE extract

One of the most important attributes of a B-PE or R-PC extract is its purity, which determines its potential industrial use. Purity ratios higher than 0.7 are generally considered as sufficient for food applications, while \({A}_{565}/{A}_{280}\) values higher than 4.0 are desired for analytical purposes (Hsieh-Lo et al. 2019). In the present work, the purity of the protein-rich extract was 2.50, which suggests its potential for use in food applications as a natural pigment. Indeed, the high purity of the extract would allow its utilisation in other industrial applications such as cosmetics or as a colourant of pharmaceuticals (Hsieh-Lo et al. 2019). This value is in line with that obtained by Marcati et al. (2014), who achieved a purity index of 2.3 after membrane purification and compares well with previous reports that obtained \({A}_{565}/{A}_{280}\) values around 0.9 (Martínez et al. 2019) and 2.0 (García et al. 2021b). Higher purity ratios could be achieve by using chromatographic techniques; for example, a purity of 4.25 was recently achieved after purification of phycoerythrin from Pyropia yezoensis (Zang et al. 2020) and a purity of 5.6 was achieved after purification of phycoerythrin from Polysiphonia urceolata (Liu et al. 2005) in both cases using chromatographic techniques. To achieve this high purity at the commercial scale remains a challenge, as column chromatography is difficult to apply for industrial production. These high purities are not required for food applications. A purity index lower than 4.0 implies the presence of other proteins; indeed, the P. cruentum-derived extract also contained R-PC (21%). However, despite the blue colour of R-PC, because of the much higher concentration of B-PE, the colour of the extract was pink.

Previous work has suggested that the use of phycobiliproteins as food-grade pigments could be limited because of their low stability to light or acidic conditions (Newsome et al. 2014). Figure 1B shows the stability of the colour of the extract at different pH values. The proteins were denatured at pH values lower than 3.0 and higher than 10.0, which is in line with previous observations for B-PE obtained from other algae. The stability of B-PE was high at pH values ranging from 4.0 to 9.0 (García et al. 2021a) and from 4.0 to 10.0 (Liu et al. 2009). Extreme pH values do not only affect the optical density or fluorescence of the B-PE extract but also its colour; indeed, a previous study observed how the colour of a B-PE extract obtained from P. cruentum changed from violet at pH values lower than 3 to pink at pH values in the range 3.7–12.1 and turned colourless at pH values higher than 12.6 (González-Ramírez et al. 2014). In the present work, the pH also had a striking effect on the colour of the extract (p < 0.0001; Table 2). When the pH was lower than 3.0, a* values were lower and b* values were higher, indicating a decrease in the pink hue. This was consistent with visual observations as the colour of the extract changed from its characteristic pink colour to a brown/violet colour. The extract turned almost colourless at pH values higher than 10.

Assessment of the pigmentation capacity of B-PE in commercial beverages

As shown in Table 2 the pink colour of the extract was stable at pH values ranging from 3.0 to 9.0. The pH of the test beverages ranged between 3 and 7, suggesting that the colour of the B-PE once introduced into the beverage would be pink. Indeed, after adding the B-PE extracts the colour of the beverages was pink and the amount of extract needed to mimic the colour of the commercial products was very low. The staining factors for the different beverages were lower than 4.5 mg·L−1 in all the studied beverages. This value refers to the amount of pigment that is needed to minimise the colour difference between the B-PE-containing product and the one that is already commercialised pink. Results demonstrated that, for example, to achieve a pink colour comparable to that of pink gin, approximately 1.5 mg·L−1 was required. These results demonstrate the potential of B-PE for being used a natural colourant in the beverages industry. In the first place, because such low amounts are unlikely to cause an adverse effect on the physicochemical properties of the products or the health of the consumers. Indeed, the viscosity of the samples was not modified after the addition of the B-PE extract. In the second place, because one of the main challenges of the utilisation of microalgae at industrial scale is their high production costs, which limits the utilisation of microalgal biomass to a high-end market (Lafarga 2020). The low staining factors achieved suggests that the process could be economically viable. The content of B-PE in the biomass of P. cruentum depends on the strain used and on the production conditions (e.g. photobioreactor used, culture medium, environmental conditions). A recent study quantified the concentration of B-PE in the biomass of P. cruentum as 32 mg·g−1 (Martínez et al. 2019). Considering this value, 1 kg of P. cruentum would be enough to produce approximately 7.3, 20.0, 21.3, or 11.0 m3 of isotonic drinks, tonic water, gin, or wine, respectively, with a pink colour comparable to that of currently commercialised products. In addition, the leftovers of B-PE extraction could still be used in other applications such as food or feed production or as feedstock for the production of agricultural biostimulants. For example, Marcati et al. (2014) developed a process based on membrane technologies to extract both, phycoerythrin and polysaccharides with potential applications in cosmetics, nutraceuticals, or pharmaceuticals.

The colour of the product is important because it has a striking effect on its acceptability and commercial success. Indeed, a recent work where the authors developed a dairy product based on skimmed milk and phycobiliproteins concluded that the fortification of the product using phycoerythrin not only improved the products antioxidant capacity but also the perceived organoleptic properties (Galetović et al. 2020). The colour difference between the beverages commercialised pink-coloured and the products pigmented using B-PE is shown in Table 1. Overall, the minimum \(\Delta E\) values obtained in all the tested products were equal or lower than 4.0, suggesting that the colour obtained was very similar to that of the commercial product but the differences were distinguished by the human eye. Wibowo et al. (2015) reported that \(\Delta E\) values between 0.5 and 1.5 are slightly noticeable but values between 1.5 and 3.0 are noticeable by a human observer. This is not negative as the difference was minimal and the current trend is to substitute synthetic colourants such as Allura Red (E-129) with natural pigments generated from renewable sources. These include, for example, the phycobiliproteins used in the present work of the natural pigment anthocyanin (E-163), which is found in red fruits (Lafarga et al. 2019). One of the current challenges of using phycobiliproteins as natural pink/red pigments in beverages is that they are not yet accepted as food additives. In the EU, for example, novel food products must comply with Regulation (EU) 2015/2283, known as the novel foods regulation. Scientific evidence must demonstrate that the novel ingredient does not pose a safety risk to human health and this process is slow and expensive. However, other microalgae-derived products have been accepted as food additives including the natural red pigment astaxanthin which is obtained from H. pluvialis (Turck et al. 2020). Phycoerythrin has been demonstrated to be non-toxic and to provide protection against permanganate-mediated DNA damage (Soni et al. 2010). This study, together with many other that have been carried out during the last decade and those that are ongoing will certainly promote the acceptance of B-PE as a food ingredient.

It has been reported that around 90–95% of B-PE might be degraded due to its high sensitivity to physical factors such as light, temperature, pH, and oxygen (Sun et al., 2003). Regarding the thermokinetic stability of the B-PE extract after 30 days of incubation at 4, 25 and 40 ºC, the rate of degradation (K) of B-PE rapidly increased with an increase of the storage temperature (Table 3). The maximum stability (t½) was observed in the gin sample (D) with 42.50 days at 4 °C. For the rest of the temperatures, gin was also the matrix that showed the greatest stability. The second position in terms of stability was achieved by the tonic water. The samples with the lowest stability were the isotonic drink and the wine. These results can be explained in terms of the pH of the different beverages. The pH of the gin (6.5) promotes the stability of the protein. On the other hand, the lowest stability was observed in the two matrices with the most acidic pH values, 3.3 and 3.2 for the isotonic drink and the wine, respectively. Kannaujiya and Sinha (2016) reported similar results.

The stability of the colour is also important as natural pigments are generally less stable than their synthetic counterparts (Schweiggert 2018). The colour of the B-PE-containing beverages evaluated in the present work was demonstrated to be stable during an 11-day storage period. During the first week of storage, the colour differences were lower than 1.5 for all samples, which suggests that no differences could be observed. At the end of the storage period, the \(\Delta E\) values were lower than 1.0 in wine and tonic water samples and in the range 1.5–3.0 for the wine and isotonic drink samples (Table 4). The higher \(\Delta E\) values in the latter could be attributed to the higher complexity of these matrices. The colour differences could not been seen by the participants of this work; as mentioned before, samples with \(\Delta E\) values lower than 3.0 are slightly noticeable by the human eye (Wibowo et al. 2015). However, the colour of the samples did change during storage and the colour differences might have been higher after a longer storage period. Previous studies revealed that the addition of polyphenols, polysaccharides, or proteins could improve the stability of natural colourants (Gençdağ et al. 2022). Citric acid and sucrose were also found to maintain the stability of phycoerythrin at 40 ºC while benzoic acid was the best additive to maintain the stability at 4 ºC (Kannaujiya and Sinha 2016). Studies aiming at improving the stability of the developed beverages are ongoing.

Phycobiliproteins have demonstrated huge potential for being used in the manufacture of foods and beverages. Not just as pigments but also as more complex and high-value ingredients including health-enhancing illuminated colourants (Mahanil et al. 2021). Most of the health-promoting properties of phycoerythrin and the other phycobiliproteins have been attributed to their antioxidant capacity. Different works have validated the antioxidant capacity of this molecule in vitro and in animal models (Sonani et al. 2017). For example, Sonani et al. (2014) revealed anti-ageing activity of phycoerythrin as well as increased resistance to cellular stress (resulting in improved life span and health span) using Caenorhabditis elegans, which is considered as a good model to study ageing in humans. Similarly, Cano-Europa et al. (2010) concluded that phycoerythrin isolated from Pseudoanabaena tenuis prevented the increase of oxidative markers and protected mice cells against HgCl2-induced cell damage in mice. Other bioactivities that have been attributed to phycoerythrin include antibacterial activity against both Gram positive and Gram negative bacteria and anticancer activity against HepG2 cell lines (Hemlata et al. 2018). Overall, the extract isolated in this work showed huge potential for being used as a colouring additive in isotonic drinks, gins, wines, and tonic water and allowed achieving colours that were comparable to those of commercial products. The colour of the phycoerythrin-rich extract generated from the biomass of P. cruentum was pink and stable at pH values in the range 3.0–9.0. Its use is not recommended in beverages with a pH below or above this range. The stability of the colour of the tonic water, isotonic drink, gin and wine pigmented using phycoerythrin was stable during 11 days; although further studies are needed to assess its stability during longer periods. A visual acceptability test revealed that the colour of the products was well accepted and the acceptability of the products was comparable to that of commercial pink products.