Phycocyanin extraction from Limnospira spp.: sustainable source of natural blue color for the food industry

Abstract

Phycocyanin (PC) is a blue protein pigment whose interest increased due to its potential as alternative to synthetic food colorants. The cyanobacteria of the genus Limnospira spp. are the main source of commercially available PC. The extraction and purification processes showed high variability in yield, purity, quality and environmental and economic sustainability. During the last years, different extraction methods like freezing and thawing, ultrasound assisted extraction and enzyme assisted extraction have been improved to increase PC yield and reduce purification steps once crude extract is obtained. This review aims to provide a comprehensive analysis of PC extraction techniques from Limnospira biomass. Conventional and innovative technologies are compared to highlight their performances, in terms of yield, purity, feasibility, cost, and environmental impact. Furthermore, separation and purification are reported to elucidate which processes are adopted nowadays and which are the possible improvement to increase the purity grade of protein. This comparative study also aims to provide the basis for the development of alternative methods for the optimization of phycocyanin extraction, using also neural networks and artificial intelligence.

Graphical abstract

Highlights

• •Freeze-thawing gives high C-PC yield, but low scalability for industrial application.
• •L. platensis is the top source for natural blue pigment.
• •Ultrasound plus freeze-thawing improves C-PC extraction yield.
• •ANN and AI have potential in optimizing C-PC extraction methods.
• •C-PC market value ranges from 0,3 US$/g to $33/mg for food and analytical grade respectively.

1. Introduction

Arthrospira spp. is a filamentous nontoxic species of cyanobacteria (commonly referred as blue-green microalgae), widely used in the food industry as protein supplement and as a natural source of functional nutrients, but also used in the nutraceutical sector, and animal feed (Pan-utai et al., 2022).

Although Arthrospira is historically sold under the common name “Spirulina”, according to official taxonomy classification Arthrospira and Spirulina are two distinct genera, categorized in separate orders and even if both include helically-coiled trichomes, they exhibit poor physical and ecological similarities (Nowicka-Krawczyk et al., 2019). Inside Arthrospira genus there are 23 species, of which three are marine, three are from high pH habitats, while the remaining ones are from freshwater. Species like A. maxima, A. fusiformis and A. platensis, recognized as commercial strains, have been isolated from extremely alkaline environment in tropical and subtropical areas, whereas species like A. jenneri from freshwaters with low salinity (European Commission, 2024). Phylogenetic analyses based on the 16S rRNA gene demonstrated that besides their ecological differences, A. jenneri and commercial species belong to different genera, determining a new classification in which the commercial ones were assigned under the name of Limnospira genus. At the beginning, this classification produced several doubts about the collocation of A. platensis species due its ecological characteristics close to A. janneri rather than L. fusiformis, L. maxima, and L. indica (Roussel et al., 2023). However, with the establishment of Limnospira genus in 2019, whole-genome sequencing of all different strains were compared and it was demonstrated that Limnospira is monospecific genus and it is represented by L. platensis PCC 7345 (number in Pasteur Culture Collection) (Papapanagiotou and Gkelis, 2019). Although genomic analysis highlighted large intra-specific diversity inside this genus, it was also suggested that strains included and formerly designated as L. fusiformis, L. indica, L. maxima and other related strains, should be regarded as substrains of L. platensis (V Kupriyanova and Los, 2024). Therefore, in this review, we are going to use the name Spirulina referring to Limnospira genus, that is L. platensis and its subspecies.

L. platensis biomass is particularly rich in proteins (55–70 % of dry weight, with all essential amino acids), carbohydrates (12–25 %) and lipids (4–8.2 %) with health-beneficial fatty acids, such as eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), 3,6 γ-linolenic acid, α-linolenic acid, stearidonic acid, and arachidonic acid (Thangsiri et al., 2024). However, among different bioactive compounds that can be extracted, a special interest had been focused on C-phycocyanin (C-PC), a water-soluble phycobiliprotein which represents one of the major high added value molecules and the most abundant protein present in this genus (Russo et al., 2024a). C-PC is the primary pigment produced by this cyanobacteria species, containing up to 20 % phycocyanin of dry biomass in optimized growing condition and environment, followed by allophycocyanin (A-PC, light-blue pigment) and phycoerythrin (PE, red pigment) (Nikolova et al., 2024). The majority of C-PC on industrial scale is extracted from L. platensis or L. maxima and the expanded knowledge on its extraction is mainly due to the efficient application as a natural colorant for food and cosmetics, even though recently most researches focused on antioxidant, anti-inflammatory, hepatoprotective and anticarcinogenic properties of this pigment (Fernandes et al., 2023). Currently, the different techniques which have been developed for the extraction of C-PC from microalgae are usually divided in physical, chemical and enzymatic methods. In the last decades, several improvements have been made in order to minimize waste and to obtain a high production, high-grade purification and long-term storage of pigments (Russo et al., 2023). However, an appropriate approach to extract C-PC from microalgal species able to preserve and stabilize its bioactivity is strictly dependent on the disintegration of their cell wall, which consists mainly in peptidoglycan and an outer membrane composed of lipopolysaccharides, similar to gram negative bacteria. In fact, although some of these methods, like freezing and thawing, result in relatively high purity extracts (Tavanandi et al., 2018), still there are several limitations that involve many conventional techniques as providing low-purity extracts or pigments with poor stability when stored and used in food processing (Di Salvo et al., 2023).

For this reason, beyond C-PC, several studies investigated the extraction of bioactive compounds from the residual biomass of L. platensis, by using solvent extraction and/or enzymatic hydrolysis. For example, enzyme-assisted extraction is a green process showing high potential use for the downstream applications. In fact, enzyme-assisted extraction can yield a wide range of secondary metabolites such as minerals, antioxidants, vitamins and sterols used for different applications (Otero and Verdasco-Martín, 2023). However, the high cost of enzymes that increases the process economics and the requirement of further processing to separate and purify the products set limitations of its use only for extracting only high-value products (Wang et al., 2019).

Thus, this review aims to compare extraction methods applied on Spirulina species and to evaluate the C-PC yield, purity, and selectivity obtained by these methods. Moreover, a particular focus will be addressed to the advantages and drawbacks of each of them from an economic, environmental and feasibility point of view.

2. Bibliometric analysis

A bibliometric analysis was conducted using data extracted from Scopus database. The search query used was the following: TITLE-ABS-KEY (Spirulina) OR TITLE-ABS-KEY (Arthrospira) OR TITLE-ABS-KEY (Limnospira) AND TITLE-ABS-KEY (phycocyanin) AND TITLE-ABS-KEY (extraction) OR TITLE-ABS-KEY (pigment AND extraction) OR TITLE-ABS-KEY (enzyme AND assisted AND extraction) AND PUBYEAR > 2018 AND PUBYEAR < 2025 AND (EXCLUDE (DOCTYPE, “cr”) OR EXCLUDE (DOCTYPE, “ch”) OR EXCLUDE (DOCTYPE, “er”) OR EXCLUDE (DOCTYPE, “sh”) AND (LIMIT-TO (LANGUAGE, “English"))

The analysis of recent literature on C-PC extraction methods from cyanobacteria, particularly Spirulina, reveals a growing interest and significant advancements in the field over the past five years. This surge in research is driven by the increasing commercial demand for natural compounds with strong antioxidant, anti-inflammatory, and therapeutic properties.

2.1. Keywords analysis

To generate the keyword co-occurrence network map, we conducted a bibliometric analysis using Python on the dataset from Scopus, focusing on research articles about C-PC extraction from Spirulina. Keywords from ‘Author Keywords' and ‘Index Keywords' were combined and cleaned, and relevant papers were filtered to include those mentioning both “phycocyanin” and “extraction.” We constructed a co-occurrence matrix and created a "NetworkX" graph, applying a threshold to filter less frequent keyword pairs. Community detection was performed using the Louvain method, followed by K-means clustering to reduce the clusters to seven. The network was visualized using a spring layout with custom colors for distinct clusters, and a cluster report was generated to identify representative and top keywords for each cluster. The network visualization and cluster report were saved as image and CSV files, respectively, providing insights into research trends and key focus areas in phycocyanin extraction techniques.

2.2. C-PC extraction bibliometric analysis results

The scientific production obtained indicates a steady increase in publications related to C-PC extraction. The search query focused on terms such as “Spirulina,” “Arthrospira,” “phycocyanin,” and “extraction,” spanning from 2018 to 2023.

The word cloud of keywords present inside the bibliometric database (Fig. 1) provided a visual representation of the most frequently occurring terms in the literature on phycocyanin extraction. Dominant keywords include “extraction,” “phycocyanin,” “Spirulina,” “Arthrospira platensis,” and “biomass,” reflecting the primary focus of the research on optimizing extraction processes and characterizing the biomass sources. Other significant terms such as “ultrasound,” “protein,” “purification,” and “antioxidant” indicate the common methodologies and applications being explored. The prominence of terms like “evaluation,” “potential,” and “efficiency” underscores the ongoing efforts to enhance the extraction yield and functional quality of phycocyanin. This visual analysis corroborates the bibliometric findings, highlighting key themes and technological innovations in the field. In Fig. 2 the keyword co-occurrence network generated from the bibliometric analysis of the literature related to phycocyanin extraction was reported. Based on the keyword co-occurrence network map, several key themes emerged in the literature on extraction techniques of phycocyanin from Spirulina. The central theme is “Extraction” which is prominently highlighted and closely linked with a multitude of other keywords, indicating its significance and widespread research interest. Clustered around “Extraction” we found several interconnected groups, each representing distinct but related research areas.

Fig. 1.

Word cloud of keywords obtained after the bibliographic research on C-PC extraction from Spirulina.

Fig. 2.

Keyword Co-occurrence Network related to phycocyanin extraction from Spirulina, highlighting the key thematic areas within the field.

The largest cluster, labeled as “#1 Extraction”, signifies the central focus of current research efforts, exploring various methodologies and optimization techniques for extracting phycocyanin. The second cluster, "#2 Phycocyanin”, highlights the chemical compound of interest, underscoring studies that delve into its properties, benefits, and potential applications. “#3 Spirulina”, forming another significant cluster, indicates the biological source of phycocyanin, with studies focusing on its cultivation, processing, and enhancement of phycocyanin yield.

Other notable clusters include “#4 Microalgae”, which expands the scope beyond Spirulina, suggesting comparative studies or alternative sources for phycocyanin. The cluster “#5 Solvents” implies research into the different solvents used in the extraction processes, examining their efficiency and impact on the yield and purity of phycocyanin. “#6 Bioactive Compounds” indicates an interest in the broader spectrum of compounds extracted from Spirulina and other microalgae, emphasizing the multifunctionality of these organisms. Lastly, “#7 Antioxidants” reflects studies that explore the antioxidant properties of phycocyanin, potentially linking its health benefits to its extraction processes. The interconnected nature of these clusters suggests a multidisciplinary approach in the research, combining aspects of biochemistry, biotechnology, environmental science, and health sciences. The network map provides a comprehensive overview of the current state of research and highlights the primary areas of interest and investigation.

Geographical distribution of corresponding authors' countries in publications related to C-PC extraction topic showed that India leads the field with the highest number of publications, followed by Indonesia, Brazil, China, and France. The collaboration between countries, as indicated by the multiple-country publications (MCPs), highlights the international effort in advancing extraction technologies and methodologies. Countries like Germany, Portugal, and Malaysia also show a balanced contribution between single-country publications and MCPs, indicating active domestic research alongside international collaborations.

3. Phycocyanin

L. platensis as all the cyanobacteria can produce energy from the oxygenic photosynthesis facilitated by structures called phycobilisomes (PBS), light-harvesting antenna complexes distributed on the thylakoid membrane that contain phycobiliproteins, protein pigments which absorb light and transfer energy to the photosynthetic reaction centers. PBS structures are relatively simple, composed of several phycobiliproteins (PBPs) connected to the reaction center of photosystem I (PSI) or photosystem II (PSII) through the core membrane linker (LCM). PBPs are natural, water-soluble pigment proteins responsible in cyanobacteria for around 50 % of light uptake (Hsieh-Lo et al., 2019; Park et al., 2018). As accessory pigments of chlorophyll, they improve the ability to harvest light energy in a wide range of wavelengths respect to chlorophyll which has a low extinction coefficient, especially in red and blue regions of the spectrum (Hotos and Antoniadis, 2022).

PBPs are represented by the dark blue C-PC, with a maximum absorption spectra ranging from 610 to 625 nm and maximum fluorescence emission at 645–650 nm, by the red phycoerythrin (PE) (peak of light absorption at ∼565 nm) both composing the rod of PBS, and by A-PC (light blue pigment with absorbance peak at ∼650 nm) that represents the core of PBS; all the PBPs are closely connected through linkers. C-PC gathers light energy from the environment and transmits it to the A-PC core (Liu et al., 2024). Fig. 3 reports a schematic flow of the phycobilisomes-mediated photosynthetic process that occurs in the cyanobacteria.

Fig. 3.

Photosynthetic Electron Transport Chain in Cyanobacteria. PS I, Photosystem I; PSII, Photosystem II; PQ, plastoquinone pool; Cyt b6f, cytochrome b6f complex; PC, plastocyanin; PE, Phycoerythrin; C-PC, C-Phycocyanin; APC, Allophycocyanin (Figure created in the Mind the Graph platform, www.mindthegraph.com).

Thus, C-PC role in the photosystem is to boost photosynthesis efficiency by absorbing light energy between 495 and 650 nm and transfer it to chlorophyll in the photosystem, increasing the efficiency of energy production in cyanobacteria.

C-PC is composed of a heterodimeric protein component (α and β subunits with a molecular weight of 17 and 21 kDa, respectively) and a non-protein component called phycocyanobilin attached through a thioester bond catalyzed by phycobiliprotein lyases to cysteine residues at position 84 in the α subunit and cysteines 84 and 155 of the β subunit (Roy et al., 2024). The heterodimer assemble into a ring-shaped trimer to produce a hexameric structure which provides phycocyanin stability (Fernandes et al., 2023). Moreover, the close proximity of phycocyanobilins results in intra- and intermolecular electron interactions, responsible for the unique spectroscopic characteristics of each aggregation state (Böcker et al., 2020). Using mass spectroscopy, nuclear magnetic resonance (NMR) and infrared analysis, it has been revealed that the structure of phycocyanobilin (whose molecular weight is approximately 588 Da) exhibits a variety of intermolecular hydrogen bonds and one of these is tetrapyrrole chain. Therefore, the numerous chemical changes that occur, such as the esterification process, which is linked to dehydrogenation, are caused by these H-bonding (Chittapun et al., 2020).

Consequently, behind the choice of disruptive method adopted, C-PC extraction is highly influenced by different chemical and physical parameters like temperature, pH, biomass form and biomass/solvent ratio that could compromise C-PC purity and its application in different sectors. Temperature has a strong influence in the extraction of intracellular compounds since alters cell membrane shape, increasing mass transfer of internal compounds to the extraction medium (Pez Jaeschke et al., 2021). The impact of temperature on C-PC degradation has been the subject of numerous investigations, demonstrating the degradation rate of phycocyanin with increasing temperature at different times. For this reason, extraction methods applied for phycocyanin extraction are usually performed in a range of temperatures of 25–47 °C, since above 50 °C many studies have reported an increase in the degradation rate (Böcker et al., 2019). In particular, physical extraction methods are carried out in an optimum range between 25°C and 30 °C to avoid the degradation of protein and the decrease in C-PC purity (Tavanandi and Raghavarao, 2020).

As mentioned before, pH represents another important parameter that has an influence on stability of C-PC and, in particular, on its spectral properties changing the color of the protein. Generally, aqueous buffer solutions are set at pH range of 6–7, to prevent phycocyanin instability that is reported at pHs below 5 and above 8 (Moreira et al., 2018). Sodium phosphate (pH 6–8) is the most common buffer solution used for C-PC extraction from L. platensis, even though several studies have reported better extraction yield with CaCl₂ (1.5 %, w/v) or NaCl (0.15 M) used as solvent (İlter et al., 2018; Wang et al., 2023; Benucci et al., 2023).

High yield and purity grade of C-PC extraction from L. platensis have been investigated also considering the biomass/solvent ratio according to the biomass form and solvent used. Most of the research results are based on dried biomass of Spirulina respect to wet biomass form, which is used less frequently as a starting material, while sodium phosphate buffer is the preferred solvent (Jaeschke et al., 2019). Nevertheless, extraction yields increase with the biomass/solvent ratio, on the other hand, a high biomass/solvent ratio causes reduction of C-PC purity. The biomass/solvent ratio for C-PC extraction from Spirulina has only been examined in a small number of studies and the findings are unclear. The ratio of 1:10 (w:v) produced the maximum extraction yield with sodium phosphate buffer (pH 6.8) among the tested 1:6, 1:8 and 1:10 ratios. Although the ratio of 1:6 produced the purest extract, an increase in solvent availability could result in the extraction of additional molecules, including interfering proteins (Tavanandi et al., 2018).

Phycocyanin concentration can be assessed spectrophotometrically in aqueous extracts taking into account the spectral overlap of C-PC and A-PC.

Many papers report for quantification of C-PC in Limnospira, the methods and the relative extinction coefficients developed by Bennett and Bogobad (1973) (Billy et al., 2023) for the filamentous cyanobacteria Fremyella diplosiphon.

Yoshikawa and Belay (2008), instead, developed a more accurate method for the spectrophotometric determination of A-PC and C-PC in the aqueous extract (100 mM phosphate buffer, pH 6.0) of Spirulina (A. platensis) providing linear responses for C-PC and A-PC in the range of 25–250 μg/mL.

Their final equations for calculation of phycobiliprotein concentrations were as follows:

$$CPC,mg/mL=\left(0.162\times OD620\right)-\left(0.098\times OD650\right)$$

$$APC,mg/mL=\left(0.180\times OD650\right)-\left(0.042\times OD620\right)$$

The purity grade of C-PC is usually expressed as the A620/A280 ratio of C-PC extract solution. According to the purity grade reached, the potential application of the C-PC changes: a C-PC purity index ≥ 3.9 will be taken into account for the reactive grade, instead for food grade application a purity grade greater than 0.7 is sufficient. Consequently, increasing the purity grade of the pigment, the market value of phycocyanin increases too, also considering the challenges encountered during the extraction and purification processes, that affect the quality of C-PC protein pigment (Freire Balseca et al., 2024). The commercial selling price of C-PC fluctuates according on the purity index starting from about 0,3 US$/g for the lowest quality food grade product up to 5 US$/mg for reagent grade, and 33 US$/mg for analytical grade product with a purity ratios exceeding 4 (Roy and Pabbi, 2022). The global phycocyanin market was estimated to be worth USD 203.33 million in 2023. Over the period from 2024 to 2030, this market is expected to develop at a compound annual growth rate (CAGR) of 9.6 %, reaching approximately USD 391.72 million by 2030 (Thevarajah et al., 2022).

Due to growing demand, technologies and strategies to improve Limnospira biomass productivity, concentration and purity grade of this blue protein, are in continuous development. In fact, to enable mass manufacturing, especially in the food sector, the biomass and productivity of Spirulina have to be enhanced. Recent research studies have reported that, among different strains of this genus, L. platensis, Spirulina subsalsa and L. maxima have been recognized as high-quality varieties. For instance, it has been demonstrated that Spirulina subsalsa, a species of Spirulina grown in seawater, has the highest C-PC content compared to the widely utilized L. platensis (Jiang et al., 2021). In Limnospira genus, different strains have been evaluated and among them, Limnospira CCM-UdeC 040 and UTEX 2342 showed the highest C-PC content, respectively 69.90 ± 9.32 mg/g and 69.15 ± 3.97 mg/g (López-Rodríguez et al., 2023). L. platensis is recognized as one of the main sources of natural commercial phycobiliprotein C-PC, determining an increasing interest in its cultivation. Thus, it is important to consider the optimal conditions by which L. platensis can grow, to improve the content of C-PC as well as its purity grade. There are different factors that can affect growth and pigment accumulation in Spirulina, such as nitrogen source, carbon source, initial biomass concentration, pH, salt concentration, and light. In particular, for L. platensis productivity, it has been observed that the optimal temperature range is around 30–35 °C, maintaining a pH between 8.5 and 10.5, with a light intensity of about 300 μmol photons m²/s, which is optimal for maximizing C-PC production. Moreover, the use of red and blue light wavelengths, with an adequate nitrogen supply, can further enhance pigment synthesis (Thevarajah et al., 2022; Batista et al., 2019).

4. Extractive methods

Although several extraction methods like freezing and thawing and bead milling are the most commonly used for C-PC recovery, these techniques require long processing time, produce unpurified extracts with high concentration of cell debris and most of the time they are not easily scaled-up for food industry (İlter et al., 2018). Moderate selectivity, high solvent consumption and low extraction efficiency (50–60 % yield) are the primary issues with the traditional extraction procedures (Tavanandi and Raghavarao, 2020). Another significant disadvantage that involves these procedures is using fresh raw materials, which have several drawbacks like quick deterioration and contamination by microorganisms. This is mainly attributed to the first critical step in the extraction such as the disruption of microalgae cell wall that, according to processing time required, could increase the yield. However, although increasing processing time may raise C-PC yield, shear-sensitive natural pigments like C-PC could be denatured.

Instead, physiologically active compounds are preserved when green extraction techniques are used since they shorten the extraction period, boost the yield of the targeted protein, and reduce extra purification steps. In this context, these new emerging technologies (microwaves, pulsed electric fields, ultrasounds and enzymes) have been recently evaluated to overcome these limitations, particularly to increase C-PC recovery yield, separating from other soluble chromophores or proteins.

Fig. 4 shows the effects on the cyanobacterial cell wall of the main used techniques for assisting C-PC extraction.

Fig. 4.

Schematic representation of the cyanobacterial cell wall permeabilization/destruction techniques for C-PC extraction: pulsed electric field, freeze-thawing; enzyme extraction, ultrasound, bead milling and high-pressure homogenization.

As a result, “green” extraction techniques employed with cyanobacteria are becoming more and more popular (Nikolova et al., 2024). Thus, many studies have reported that the combination of physical, chemical or enzymatic methods to improve the extraction of the pigment could be a solution, avoiding yield loss and increasing purity grade of phycocyanin after extraction. The main results of the updated literature review regarding extraction methods C-PC extraction yields and solvents are summarized in Table 1.

Table 1.

Comparative results of Phycocyanin Extraction Methods from Limnospira spp.

Extraction Method	Solvent	C-PC yield (mg/g)	Reference
Freeze-thawing	CaCl2 solution	170.3 ± 5.3	Kuhnholz et al. (2024)
	20 mM sodium acetate and 50 mM sodium chloride	250.13 ± 25.72	Athiyappan et al. (2024)
	Phosphate buffer (pH 5.1)	159.33 ± 39.97	Athiyappan et al. (2024)
	Distilled water	139.24 ± 55	Athiyappan et al. (2024)
	Phosphate buffer (pH 6.5)	83.4	Yu (2017)
	Distilled water	18.159 ± 0.017	Blanco-Vieites et al. (2023)
	20 mM sodium acetate and 50 mM NaCl buffer (pH 5.1)	217.18 ± 21.47	Chentir et al. (2018)
	0.1 M phosphate buffer pH (6.8)	74.51	Tavanandi et al. (2018)
	10 mM Tris-HCl buffer	41.90 ± 0.23	Ferreira-Santos et al. (2020)
	1 % CaCl2	27.7	Dianursanti (2021)
	Phosphate buffer (6.8 pH, 0.1 M)	35.31
	10 mM Tris-HCl buffer (pH 8.3)	101 ± 0.247	Ores et al. (2016)
	Distilled water	151.8	Ruiz-Domínguez et al. (2019)
	Na-phosphate buffer (pH 7.0) 100 mM	232.8
	Na-phosphate buffer 50 nM	177.9 ± 5.9
Bead milling	Sodium-phosphate buffer (pH 7.2)	94.90	Jaeschke et al. (2019)
	Distilled water	119.48 ± 6.7	Käferböck et al. (2020)
	Distilled water	73.1 ± 0.76	Pan-utai et al. (2022)
	Ethanol	6.44 ± 0.41	Martí-Quijal et al. (2023)
	Deionized water + 1 % chitosan	109.2 ± 7.4	Kuhnholz et al. (2024)
High pressure homogenization	Distilled water	58	Chen et al. (2022)
	Phosphate buffer (pH 7, 0.1M)	44	Kuhnholz et al. (2024)
	Na-phosphate buffer	291.9 ± 6.7	Ruiz-Domínguez et al. (2019)
	Distilled water and 1 % chitosan	101.1 ± 1.3	Kuhnholz et al. (2024)
Ultrasound assisted extraction	Phosphate buffer	180.36 ± 26.62	Athiyappan et al. (2024)
	Acetate buffer	228.17 ± 6.60	Athiyappan et al. (2024)
	Water	163.31 ± 35.84	Athiyappan et al. (2024)
	Britton-Robinson	132.59	Sharma et al. (2020)
	Distilled water	122.0 ± 4.5	Chen et al. (2022)
	Phosphate buffer + lisozyme	98.24	Tavanandi and Raghavarao (2020)
	Phosphate buffer + Tween 80	92.73	Tavanandi and Raghavarao (2020)
	Phosphate buffer 0.1M	44.5	Devi et al. (2020)
	Distilled water + 1 % chitosan	152.1 ± 4.1	Kuhnholz et al. (2024)
Pulsed electric field	Distilled water	26	Carullo et al. (2021)
	Distilled water	21	Carullo et al. (2020)
	Deionized water	∼120	Ricós-Muñoz et al. (2023)
Microwave assisted extraction	Sodium phosphate buffer 0.1 M	8.4	Rodrigues et al. (2020)
	Distilled water	28.90	Fratelli et al. (2021)
	Distilled water	215.0 ± 5.5	Ruiz-Domínguez et al. (2019)
Enzyme assisted extraction	Citrate buffer (50 mM, pH 5.0)	3.1	Lee et al. (2022)
	Phosphate buffer	23.43	Tavanandi et al. (2019)
	Phosphate buffer	80.16	Tavanandi and Raghavarao (2020)
	Phosphate buffer 0.1M	33.98	Devi et al. (2020)

4.1. Freeze and thawing

Freeze and thawing method for C-PC extraction is one of the most valuable techniques applied, reported as being the best for yielding the most phycobiliproteins and in particular for the phycocyanin. Indeed, in order to obtain the desired bioproduct, like thermosensitive bio-compounds, freeze-thawing involves one or more cycles of freezing at very low temperatures (often < −20 °C) followed by thawing the biomass, causing phase separation after protein's denaturation. The freeze-thawing procedure can be used as a pretreatment to help solubilize and disperse biomass in preparation for subsequent steps such as bead milling, ultra-sonication, and high-pressure homogenization (de Souza Celente et al., 2023). This method is one of the most reliable and validate for the C-PC extraction, even though several authors have investigated different conditions to improve the extraction yield. For this extraction method, there are several factors that could influence the efficiency of the yield and purity of the extracted phycobiliprotein. Many authors reported that the ideal condition for the extraction of C-PC with a high extraction yield and good purity is running four cycles of freezing and thawing. Throughout serial extraction, Tavanandi et al. (2018) demonstrated that after 4 cycles of freezing and thawing no further improvement of C-PC yield was obtained with a cumulative value of 74.51 mg/g dw biomass. Moreover (Ores et al., 2016), comparing results of C-PC yield with freeze and thawing extraction with two and four cycles, demonstrated that the extraction yield increased from 58 ± 5 to 101 ± 0.2 mg/g. Beside extraction yield efficiency, Chittapun et al. (2020), also reported that increasing the number of cycles could compromise purity by breaking down the crude extract's soluble C-PC protein. Increasing the frequency of freeze-thaw cycles can cause damage to proteins and cells by introducing multiple ice forming events. These findings confirmed that the number of cycles in freeze and thawing extraction depends on cyanobacterial strain since each of them has a unique structure affecting the yield and purity of C-PC. Moreover, according to Pispas et al. (2024) a lower biomass/solvent ratio may increase the solvent or buffer's diffusivity into the cells and extract more C-PC, whereas a greater ratio may result in a lower C-PC extraction. Therefore, the degree of cell rupture and C-PC extraction may be enhanced by combining freezing and low biomass/solvent ratio.

As reported in the Table 1, for freezing and thawing method, distilled water, acetate buffer, CaCl₂ and sodium phosphate buffer are most popular solvent used. Phosphate buffer, in the pH range of 6.8–7, provides higher stability, gives better results in terms of extraction yield compared to the extraction yield in phosphate buffer and CaCl₂ (Dianursanti, 2021). Indeed, phosphate buffer 0.1 M allowed to recover higher phycocyanin than 1 % CaCl₂ (respectively 35.69 mg/g and 27.7 mg/g yield). Higher values in terms of yield (113.4 mg/g of dry weight) were observed by Pispas et al. (2024) using 1 M Tris-HCl buffer (pH 7) respect to phosphate buffer at 0.1 M at pH 6.8 and 6.5 tested by Tavanandi et al. (2018). Differently in L. maxima, Kuhnholz et al. (2024) demonstrate calcium chloride solution superiority over other extractants like deionized water and phosphate buffers, highlighting how solvent extract selection may contribute to higher C-PC content through this method (170 mg/g dw). However these results are in contrast with those of Ruiz-Domínguez et al. (2019) who reported the highest C-PC yield (233 mg/g dry biomass) with Na-phosphate buffer (100 mM; pH 7), revealing sodium phosphate buffer as the best solvent for freeze and thawing method. Nowadays, freeze and thawing process is considered one of the most efficient extraction methods for C-PC despite its simplicity. However, the repetitive freeze-thaw process is not appropriate for industrial applications, as it is time and energy consuming at large biomass throughputs. (Kuhnholz et al., 2024).

4.2. Bead milling

Considering its high efficiency of disruption and the commercially accessible devices at large scale, mechanical cell disruption, namely bead milling, is a very promising technology for industrial scale application (Zinkoné et al., 2018). This method is based on the functionality of a rotating agitator which acquiring enough kinetic energy in a bead-filled vessel, ensure the collision of algae cell walls and consequently their fragmentation by colliding with beads (Suarez Garcia et al., 2019). Several factors can affect the efficiency of bead milling on the destruction of microalgae cell wall like the dimension and the load of the bead, agitation speed and material properties which still require optimization for energy efficiency and protein yield, since this technique has been widely applied mainly for protein extraction (Mear et al., 2023). Glass (low-density beads) and zirconium (high-density beads) are the most often utilized materials for the beads. Glass beads perform better with low viscosity media, while zirconium is the preferred processing material for high viscosity media (Corrêa et al., 2021). Previous studies on the extraction and purification of C-PC from Limnospira species generally employed dried and bead milled powder to achieve high yields, thus bead milling is mostly used as positive control with other techniques. Indeed, Jaeschke et al. (2019) compared the extract of L. platensis cells after bead milling and after pulsed electric field (PEF) treatment followed by 6 h of incubation at 42 ± 2 °C. C-PC concentration of 94.9 ± 6.0 mg/g dw was achieved from bead milled biomass using sodium phosphate buffer as the extraction solvent, while PEF treatment with highest energy resulted in a lower C-PC recover (71.9 ± 1.36 mg/g dw). Thus, in terms of yield, bead milling extraction is highly effective method in recovery C-PC and for this reason most of the time is used as a single-operation cell disintegration (Alavijeh et al., 2020). Compared to freeze-thawing method, bead milling extraction is more exploitable at an industrial level, showing a high disruption efficiency of up to 98 %, a substantial biomass loading capacity ranging from 60 to 150 g/L, and a conservative energy input of 0.43 kWh/kg biomass (Soto-Sierra et al., 2018). However, as the other mechanical physical methods, one of the main drawback is low purity grade. Jaeschke et al. (2019) observed that chlorophyll content in the extract by using bead milling was higher than PEF treatment, indicating that a poor selective extraction of soluble compounds as phycocyanin can be obtained (Jaeschke et al., 2019). Due to the lack of selectivity, bead milling method should require further downstream processes to remove interfering compounds and fine cell debris.

4.3. Supercritical fluid extraction (SFE)

Supercritical carbon dioxide (SC-CO₂) extraction is an eco-friendly technique widely employed as a separation technology in the food processing and pharmaceutical sectors for the extraction of bioactive compounds. Unlike solvent extraction, CO₂ separation is simple since the transfer of analytes from the matrix is made possible by the expansion of microalgal cells, with a small amount of cosolvents. This is possible due to the fact that this green technology exploits pressures and temperatures higher than the CO₂ critical point (Molino et al., 2020). In Pan-utai et al. (2022) study, L. platensis oven dried biomass was investigated for pigment extraction using SFE technology carrying out two main extraction experiments with and without ethanol as a cosolvent during SFE extraction. After the extraction of non-polar bioactive compounds using an SC-CO₂, residues of L. platensis biomass were collected for phycobiliprotein extraction. Thus, these residues from each SFE trial were successively subjected to ultrasonication assisted extraction in distilled water and phosphate buffer 0.01 M. This study revealed that L. platensis residues, without the use of ethanol as a cosolvent, and phosphate buffer extracts produced the highest C-PC yield, ranging from 29.18 to 56.09 mg/g (Table 1). However, the maximum extract purity (0.61) concerning C-PC was obtained with SFE with ethanol followed by ultrasonication extraction in phosphate buffer. Nevertheless, in comparison to the control, the concentration, extraction yield, and extract purity improved when phycobiliproteins were successively extracted after SFE.

These result have been confirmed also by (Martí-Quijal et al., 2023) by comparing the conventional C-PC extraction by stirring with SFE. In this case, under stirring (control), the value obtained for PC showed a considerable increase, rising from 2.838 ± 0.081 mg/g dw (control) to 6.438 ± 0.411 mg/g dw (SFE), or a 126.8 % increase. Moreover, new evidences have been reported by Aleksić et al. (2024) using SFE as pretreatment in L. maxima. For the C-PC extraction, the best conditions to obtain the greatest amount of mg C-PC/g of biomass resulted at 17 MPa and 75 °C in the presence of cosolvent. In this case the yield was 11.10 % ± 0.40 with a C-PC content of 133.98 ± 0.48 mg/g of biomass. As mentioned above, one of the main advantages of this method is being environmentally safe since no harmful organic solvent's residues are produced. In addition, compared to traditional technologies, supercritical CO₂ operates at relatively low temperatures to preserve the bioactivity of heat-sensitive compounds like phycocyanin, besides to be easily scaled up for industrial production. However, high energy and economic effort are needed to support C-PC extraction, also taking into account that cosolvent like ethanol or water are required, thus involving additional purification step to remove them (Pinto et al., 2022).

4.4. High pressure homogenization

High pressure homogenization is another physical method, considered as a powerful technique to obtain a significant high energy efficiency (0.4 kWh/kg biomass) protein yields, without compromising their natural properties in the final extract (Pereira et al., 2020). This process forces a liquid suspension through a narrow nozzle, which undergoes an intense shear force. In the narrow gap, a pressure drop is applied, resulting in high fluid velocities, thus not only shear forces but also the pressure drop between the environment and the nozzle contributes to cell disruption. Besides pressure range employed (≈20–120 MPa) to promote turbulence, liquid-shear stress and friction, there are some other parameters that can influence the correctly functioning of this method like the number of homogenization passages, dry cell weight and microalgal species (Corrêa et al., 2021). Additionally, temperature constitutes another important variable, since the pressure drop interferes with the temperature of an aqueous solution involved, forcing samples to be refrigerated between homogenization passages.

However, regarding C-PC extraction, few studies investigated the potential of this technique, moreover, most of the time it is coupled with pulsed electric field and ultrasound assisted extraction. Ruiz-Domínguez et al. (2019) considered two important factors to optimize this technique for the phycocyanin recovery yield from Limnospira maxima. Five pressure levels (800, 1000, 1200, 1400 and, 1600 bar) and three different solvents (distilled water, 100 mM Na-phosphate buffer (pH 7.0) and, Na-phosphate buffer:water 1:1) were tested to find out the effects of these factors in relation to C-PC concentration in the extract. Results indicated that by applying a pressure of 1400 bar and employing Na-phosphate buffer, a high C-PC yield (∼291.9 mg/g in wet biomass) was achieved. These findings suggested a direct correlation between increasing pressure and C-PC recovery, although with pressure not above 1400 bar, as confirmed also by Viana Carlos et al. (2021) in L. platensis, since protein can be denatured at higher pressures. Recently, Zhou et al. (2024) reported a combination of ultrasonic extraction with high pressure homogenization, which led to an improved yield of 135.84 mg/g of dry biomass although the purity of crude C-PC was only 0.892, remaining at the food-grade level (>0.7). An inventive tool for solid-liquid extraction, the Naviglio Extractor, speeds up the release of bioactive substances into solvents, such as C-PC from L. platensis. Unlike conventional extraction techniques like maceration, which entail static soaking over long periods of time, this device uses alternating high and low pressure (Naviglio et al., 2019). During high-pressure phases, the cyclical pressure variations of the Naviglio system (up to 10 bar) facilitate the penetration of solvents into the cell matrix and the effective extraction of solutes. During the low-pressure phase, mechanical forces push the solute compounds out of the damaged cells, including unbound cellular contents like phycocyanin, causing their quick elution into the solvent. Room temperature operation of the Naviglio Extractor is advantageous for heat-sensitive chemicals like phycocyanin, because it reduces thermal degradation. This technology is quicker and uses less energy than maceration because of its controlled cycles. Thus, improved kinetics of the Naviglio extractor ensures excellent efficiency for delicate or time-sensitive biomolecules, making it a useful tool for food, pharmaceutical, and biotechnological applications. Research has shown that this method works especially well for removing bioactive substances from natural sources, cutting down on time by 1:20 when compared to maceration while preserving excellent yields and quality (Naviglio et al., 2019). As for bead milling extraction method, also high-pressure homogenization, although high energy demanding, is a highly recommended technique for industrial use, being an easily scalable process. However, the non-selectivity and the solubilization of undesired compounds (i.e. the green color development due to co-extracted chlorophyll, which contaminates the C-PC extract) represent the main drawbacks requiring intensive purifications steps after the extraction process.

4.5. Ultrasound assisted extraction

In the last decades, the potential of C-PC as a natural safe food product has reflected an increasing market demand that should achieve two important goals: high yield extraction and high purity grade of the phycobiliprotein. In the industrial context, these two target are strictly linked to several aspects like cost effectiveness, time consuming processes and ensuring quality and stability of C-PC (Chini Zittelli et al., 2022). For these reasons, new green technologies for the extraction of bioactive compounds from microalgae have been developed to improve human and environmental health processes in the food sector (Russo et al., 2024b). Among these technologies, ultrasound assisted extraction method has been recognized as one of the most efficient method for cell disruption and extraction of many compounds from various matrices on laboratory scale (Nahid et al., 2023). This method uses ultrasonic waves with frequencies between 20 kHz and 100 MHz to break down microalgae cell walls. The propagation of these ultrasonic waves into the medium produces areas where the compression and rarefaction determine the formation of bubbles which can change their shape until collapse. This mechanism is carried out by a driving force known as acoustic cavitation which promote turbulence, collisions and agitation phenomena, allowing cell disruption and consequently a faster extraction of bioactive compounds (Vernès et al., 2019). For the ultrasound assisted extraction method, time exposure and intensity represent the two main parameters that can influence the efficiency of extraction. In most of the studies, this technology is applied by using ultrasonic probes, usually carried out with low extraction time (2.5–16.2 min) but high intensity (>1 W/cm²). Although in this way a higher extraction yield is ensured, these conditions can induce rising temperatures, affecting protein quality through denaturation (Tavanandi et al., 2018). In fact, although this process induces cavitation capable of disrupting cell walls, uneven wave distribution can create localized hot spots, potentially damaging the target compounds.

Concerning the application of ultrasound for C-PC extraction from Limnospira spp., many studies revealed that ultrasound assisted extraction method in combination with conventional methods, especially freeze-thawing, produced better results in terms of yield extraction and purity grade. The latest studies of ultrasound assisted extraction method from Spirulina species are reported in Table 1. In Teixeira et al. (2024) study, the optimization of ultrasound method with saline solutions, supported the hypothesis that ions from CaCl₂ and NH₄Cl solutions promote an easier C-PC extraction with higher yield (57 mg/g), by increasing hydrogen bonds between protein and water. However, it has been demonstrated that the coupling ultrasonication with freeze and thawing method provides best results, even if purity decreased with increasing number of freeze/thaw cycles, due to the release of additional molecules at each cycle (Chen et al., 2022; Pan-utai and Iamtham, 2019). İlter et al. (2018) evidenced that the highest C-PC content (98.84 mg/g) was obtained from frozen L. platensis biomass, using 1.5 % CaCl₂ (w/v) as extraction medium, after 16 min of extraction time.

Hardiningtyas et al. (2022) demonstrated that after 10 min of ultrasonication with a 10 mM sodium phosphate buffer, 1:8 (w/v) at 40 kHz and 200 W, a higher C-PC extraction yield was obtained from wet biomass of L. platensis compared to the freeze-thawing method.

In general, ultrasound assisted extraction provides several advantages in the C-PC extraction, above all shortening extraction time and reduced solvent consumption. Moreover, due to its easy operational mechanism and low operating cost, this method appears to be the most convenient to be applied on an industrial scale. However, concerning purity grade of C-PC in the extract, further studies on optimization of this technology are needed, in order to comprehend which parameters could affect its extraction efficiency (Shen et al., 2023).

4.6. Microwave assisted extraction (MAE)

Microwave assisted extraction is a quick, easy, and environmentally friendly way to extract bioactive compounds from microalgae species that are highly mechanically resistant. This technology provides several advantages like short operation time, low solvent consumption, high extraction yield and high purity of the final extract (Sarkarat et al., 2023). MAE employs microwave with frequencies ranging from 300 MHz to 300 GHz that, migrating though biological matrices, produce ionic conduction and dipole rotation. Consequently, these two phenomena under the influence of electric field convert their energy into heat, which start to dissipate inside the medium. As soon as the temperature rises, water evaporates and the inner pressure changes, causing collapse of cell wall (Fratelli et al., 2021). MAE has been wildly used for lipid, polysaccharides and phenolic compounds, and the efficiency of this method has been tested also for pigment extraction from microalgae (Rathnasamy et al., 2019). Nevertheless, this technology is still poorly explored for the C-PC extraction in L. platensis. İlter et al. (2018) carried out MAE with 150 W and 120 s of extraction time and obtained the lowest C-PC content (9.34 mg/g) compared to ultrasound and freeze and thawing processes. Also, in Chia et al. (2019) study, microwave showed lower phycocyanin recovery compared to ultrasound, homogenization, and freeze and thawing extraction method (Table 1). Other studies on MAE, reported higher extraction yields with 1400 W, 2.5 GHz, 120 s, compared to conventional technologies like bead milling (Larrosa et al., 2018). Rodrigues et al. (2020) evaluated the C-PC recovery, optimizing MAE by using protic ionic liquids (PIL), instead of the aprotic liquids and buffer. PIL are ionic liquid characterized by high thermal stability, low volatility and high solvent capacity. In MAE, they enhance the extraction efficiency by absorbing microwave energy more effectively, leading to faster heating and improved solubility of target compounds, making the process more sustainable and efficient compared to traditional solvents (Fan et al., 2019). Indeed, testing the efficiency of PILs as solvents for each extraction method used, Rodrigues et al. (2020) showed that microwave performed better than ultrasound assisted extraction, reporting higher C-PC yield (8.40 vs 5.95 mg/g), but not comparable to the conventional extraction method(10.87 mg/g).

Instead, Ruiz-Domínguez et al. (2019) investigated MAE optimization in L. maxima by testing three different powers, extraction times and solvents and comparing the process with conventional extraction and one of green methodology like High pressure homogenization. Higher C-PC content (∼215.0 ± 5.5 mg/g) was obtained using distilled water as solvent and applying a power of 100 W for 30 s, but it was still a lower yield compared to high pressure homogenization using the same solvent.

Thus, MAE seems to be an innovative technology for phycocyanin extraction, being an easy and fast method from which high yield can be obtained on an industrial scale. However, as a drawback, higher temperatures reached during the process could lead to loss of biosensitive compounds and it is still too expensive compared to the other green technologies.

4.7. Pulsed electric field (PEF)

Pulsed electric fields is an extraction method based on the application of short electrical pulses (less than 150 μs) with high voltage (>1 kV/cm) through biological matrices placed between two electrodes. This technology has been recognized as an environmentally safe and energy-efficient way to induce cell disruption through electroporation (Geada et al., 2018). The electroporation is a phenomenon which produces changes in membrane permeability that, through the formation of pores on membrane, allows cellular material diffusion into the medium (Käferböck et al., 2020). For this reason, PEF is a process that induces an irreversible electroporation according to microalgal membrane characteristics, in order to extract bioactive compounds. Recently the use of pulsed electric field has become quite diffuse for C-PC extraction, since respect to other methods, this technology presents the advantages of being selective and suitable for thermolabile compounds extraction. The rate by which the pigment is released though the membrane is determined by pore sizes whose shape is dependent from PEF intensity, temperature, time, and agitation conditions (Jaeschke et al., 2019).

Ricós-Muñoz et al. (2023) proposed PEF as pre-treatment for C-PC extraction from Spirulina, achieving 120 mg/g (dw) of C-PC with low electric field strength of 3 kV/cm at room temperature. These results showed a high efficiency of the PEF also associated with other technologies (Ricós-Muñoz et al., 2023). Jaeschke et al. (2019) combined PEF treatment with freeze and thawing extraction to optimize C-PC extraction from L. platensis dry biomass, compared to bead milled control (Jaeschke et al., 2019). The application of PEF with freeze and thawing method resulted in a higher extraction yield (147.33 ± 2.45 mg/g dw) of C-PC respect to the control (119.48 ± 6.7 mg/g dw) enhancing an easier and faster strategy to extract phycobiliprotein. These results were corroborated by Käferböck et al. (2020), for the extraction of C-PC from L. maxima biomass. Also in this study, PEF treatment associated to freeze and thawing extraction produced much higher release of phycocyanin (147.33 ± 2.45 mg/g dw) respect to untreated biomass (21.04 ± 0.76 and mg/g dw) (Käferböck et al., 2020). Furthermore, another important benefit of utilizing PEF is the high level of purity obtained in the crude extract, which makes further purification processes easier.

Thus, PEF is often associated with other extraction methods, in order to allow the release of C-PC from microalgal biomass.

Nevertheless, despite the low operational costs, some drawbacks are associated with PEF treatment like development of gas bubbles that affect the uniformity of the electric field application, high capital cost of the equipment and heat energy peaks in high-conductivity products (İlter et al., 2018; Alkanan et al., 2024).

4.8. Enzyme assisted extraction (EAE)

Physical extractive methods above described, although their drawbacks, have been widely used for the extraction of different bioactive compounds from L. platensis. Among them, freeze-thawing, high homogenization pressure and ultrasound-assisted extraction methods resulted in higher yields, especially when combined, despite their low purity extract (Tavanandi et al., 2018). Therefore, a new strategy for the extraction of bioactive compounds from microalgae is highly required. Enzyme assisted extraction method is highly selective, based on the activity of specific enzymes, as carbohydrases, lipases and proteases, to allow the cell membrane permeabilization of microalgae, working directly on the biomass before the extraction (Otero and Verdasco-Martín, 2023; Callejo-López et al., 2020). As already described, the assessment for C-PC yield is based on the quantity present in crude extract, after down-streaming processes. However, just 50–60 % of the total C–PC in the dry biomass is effectively extracted by using most of the primary extraction methods, probably due to high resistance of cell membrane during disruption (Tavanandi et al., 2018). Most of the studies based on the cell membrane permeabilization of Limnospira spp. by using enzymatic method have been carried out on wet biomass. Thus, the combination of ultrasound assisted extraction with enzymatic extraction produced positive results in dry biomass, even though further research is necessary to define the best protocol for primary extraction of C-PC. For the enzymatic extraction, usually lysozyme is used to digest the peptidoglycan layer under certain temperature and pH, since the reaction conditions influence the enzymatic activity (Primo et al., 2018). Proteases (Alcalase®,and Flavourzyme®), endo-glucanase (Ultraflo®) and exo-glucanase (Vinoflow®) are usually exploited in degradation of Spirulina cell wall membrane. In the study of Verdasco-Martín et al. (2019), these four different enzymatic treatments were evaluated due to their efficient breakdown of membrane proteins, lipoproteins, and peptidoglycan. In particular the authors observed that the addittion of Alcalase® produced the best extraction yield of C-PC (Verdasco-Martín et al., 2019). Instead, Devi et al. (2020) standardized different parameters which affect the efficiency of enzymatic extraction combined with ultrasonication in L. maxima. In this study, temperature of 35 ± 2 °C, pH 7, S/L (1:4 w/v) and an enzyme concentration of 0.4 mg mL⁻¹ resulted the optimal condition to carry out extraction experiment using lysozyme, giving 25.95 mg/g dw in dry biomass (Devi et al., 2020). Results obtained by Tavanandi & Raghavarao (2020) in L. platensis dry biomass, showed higher C-PC yield (98.24 mg/g) after extraction with lysozyme (Tavanandi and Raghavarao, 2020). Moreover, C-PC was extracted in another study (Lee et al., 2022) by employing the protease Collupulin®. In this case, the extraction with Collupulin® produced a C-PC recovery of 3.1 mg/g dw biomass, a very low value probably caused by the protease activity or by the intensive thermal treatment to deactivate the enzyme. Thus, enzymatic hydrolysis is a promising technique for improving the extraction efficiency of intracellular compounds. Despite some challenges about the process parameters, the enzymatic hydrolysis present significant opportunities for further optimization, with numerous untapped potentials in enzymatic treatments that could drive the future development of the methodology (Karabulut et al., 2024).

5. Purification of phycocyanin

Following cell wall disruption during extraction, all the intracellular components are released into surrounding solvent including C-PC due to its hydrophilic nature. Indeed, solvent also affects the solubility of C-PC due to its ionic strength, which affects the protein structure (Pez Jaeschke et al., 2021). Since the molecular structure is susceptible to changes in the environment, by preserving the protein structure and color intensity of C-PC, an aqueous buffer solution with a pH of about 7 optimizes its solubility and stability. For instance, phosphate buffer is frequently employed because it can maintain the stability of C-PC and avoid protein aggregation, which could reduce yield and quality. As a result, the pigment effectively diffuse into the solvent, creating a high-quality extract that can be further purified using other techniques like ammonium sulfate precipitation or centrifugation (Julianti et al., 2019; Li et al., 2020). Therefore, solvents play a crucial role in C-PC extraction, either as dissolution reagents or just to maintain the proper molecular interactions in transformations of biomolecule (Alam et al., 2021). For the C-PC recovery, Deep Eutectic Solvents (DES) were suggested as one of the best alternatives to organic solvents, since they are biodegradable, not very toxic, readily available as well as easy to prepare (Hessel et al., 2022; Zhuang et al., 2023). Indeed, at a moderate to high temperature, DES are synthesized by mixing two substances that may form hydrogen bonds: one that behaves as a hydrogen bond acceptor (HBA) and the other that functions as a hydrogen bond donor (HBD). The combination of these components leads to a decrease in the entropy related to phase transition, significantly lowering the melting temperature of the mixture compared to the melting temperatures of the separate components (Florindo et al., 2019). DES have been recognized as excellent extractants for hydrophobic compounds, such as lipids, while their hydrophilic counterparts exhibit significant efficacy in extracting hydrophilic molecules, highlighting the versatility of DESs in various extraction methods (Gomes et al., 2020). The optimal extraction efficiency and yield for C-PC were determined to be 94.2 % and 92.0 %, respectively, using ChCl-Urea/K₂HPO₄.Thus, since their properties remain unchanged, potentially facilitating further recycling, this suggested that this method for phycocyanin extraction is environmentally friendly and has the potential to be easily scaled up (Zhuang et al., 2023). In this context, also ionic liquids (ILs) were proposed by Sánchez-Laso et al. (2021) as an alternative way for phycobiliprotein extraction due to their properties as green solvents. In this study, sonication and an imidazolium-based IL were investigated for the extraction of phycobiliproteins, optimizing the use of a factorial experimental method to determine C-PC extracted content. The results have shown that IL recovered by separating extracted phycobiliproteins using a dialysis-based procedure, can be utilized for seven successive extraction cycles with a suitable amount of C-PC yield, ranging from 75 mg/g (fresh IL) to 60 mg/g. However, although ILs are expensive and present low biodegradability when they can no longer be reused, they are efficient and ecological solvents especially with aqueous two-phase system (ATPS) separation technique for the extraction of C-PC from L. platensis (Chang et al., 2018). Usually after that, proteins are further isolated and refined using a variety of purification techniques, such as centrifugation, filtration, two-phase aqueous extraction and chromatography methods like ion exchange, size exclusion, affinity, and hydrophobic interaction chromatography (Lauceri et al., 2018). Among them, one of the most popular high resolution methods for protein separation is ion exchange chromatography (IEC), and it has recently been demonstrated that IEC produces C-PC with high recovery and purity grade (de Amarante et al., 2020). Nowadays, one of the main challenges of traditional purification methods is excluding impurities like chlorophyll, other proteins, lipids, and polysaccharides to achieve high purity without degrading the phycocyanin (Marzorati et al., 2020). For that, molecular docking simulation provides an alternative way to carry out a single step purification of the protein, through the selection of specific ligand. Molecular docking is a computational simulation technique used to predict how two molecules, such as a protein and a ligand, will interact with each other, as in the case of phycocyanin. Specifically, in the work of Shi et al. (2024), ursolic acid was identified as the optimal affinity ligand for binding C-PC, minimizing the binding of other proteins like A-PC, by comparing the docking scores of small molecules with C-PC and A-PC. After the optimization of purification conditions, this one-step process resulted in a C-PC purity index of 4.53 with a yield of 69 %, exceeding the standard for reagent-grade C-PC. Although this technique is simple, fast and easily scalable respect to conventional ones, two major drawbacks need to be considered: in this study, only small-volume affinity resin was optimized, making difficult a production scaling up, and the inability to reuse the resin for multiple cycles due to its instability. Lauceri et al. (2023), instead, proposed a two-step purification-extraction procedure where fresh biomass cell lysis though ultrasound assisted extraction is coupled with ammonium sulfate solution (39 %–50 %) to minimize the phycobiliprotein in solution, followed by extraction phase after the removal of ammonium. With respect to the conventional ultrasonication procedures where biomass cell lysis and phycobiliprotein extraction occur simultaneously, here the two processes occur separately. Indeed, just after the cell lysis phase, the clean biomass is suspended in CaCl₂ extracting solutions for phycocyanin extraction, getting a purity grade > 2.5 and 22 % C-PC yield, and allowing to obtain highly concentrated bright blue phycobiliprotein (Lauceri et al., 2023). However, the use of activated charcoal, still represent a sustainable alternative to chromatography techniques, which confirms the purity of phycocyanin with food-grade outcome, although it could be further improved by combining this method with additional low expensive purification processes (Aoki et al., 2021).

Moreover Nisticò et al. (2022) proposed membrane-based technologies like microfiltration (MF) and ultrafiltration (UF) providing a practical method for fractionation, purification, and concentration steps in L. maxima with aqueous extraction. These procedures prevent the C-PC molecule from potential thermal denaturation and deactivation, thanks to their mild operating conditions (Nisticò et al., 2022). Additionally, other benefits over traditional technologies include minimal energy consumption, no phase shift, great selectivity, modularity, and the ability to easily scale up without the need for chemical additives. It is worth to underline that, besides the efficiency of UF during the purification process, this mechanism affects the permeate flux which declines until reaching a steady-state value. For these reasons, further investigations are necessary, especially for C-PC yield enhancement, in order to reduce production time, cost and environmental impact.

6. Phycocyanin stability

The increasing interest of using food-approved blue phycobiliprotein to replace synthetic colorant shifts the attention on the optimization of extraction methods, to ensure the stability of protein. Indeed, during extraction, purification and storage, phycocyanin can be altered, compromising its application as a blue natural dye (Chini Zittelli et al., 2023).

The hexameric architecture which characterizes this protein is essential to keep not only its biological activity but also the intensity and position of the maximum absorption of the pigment. Thus, this structure must be preserved to avoid C-PC degradation that can be influenced by temperature, pH, salt concentration and light. Different studies have investigated the stability of protein under different conditions, in order to define the optimum one for each parameter. It is widely known that, at pH 5-6 C-PC is quite stable. Nevertheless, according to the temperature, the rate of C-PC degradation can change. According to Gavrailov et al. (2023), C-PC was most stable at a pH of 4.8 than at higher between 50 and 70 °C (Gavrailov et al., 2023). In addition, among those factors that influence the stability of C-PC, salt concentration plays an important role, since it stabilizes the structure by covering the surface of phycocyanin. It has been observed that by adding NaCl at a concentration of more than 1 % w/w into the solvent, it is possible to prevent C-PC degradation by more than 50–70 % (Shafieiyoun et al., 2024).

Therefore, the stability of C-PC is essential to avoid the loss of chromophores and consequently losing their color into the solution, since it is one of the main sensory factors that customers consider when selecting food and food supplements. For food products and additives, Gavrailov et al. (2023) proposed the ideal temperatures and pH to maximize phycocyanin stability during incorporation in food products are respectively up to 50 °C, and pH between 4.5 and 5.5 (Gavrailov et al., 2023).

New strategies to improve the stability of phycocyanin, such as the use of stabilizing agents, have been developed. In the study of Huo et al. (2022) after 60 min at 65 °C, samples treated with 40 % glucose, mannose, mannitol, galactose, and maltose, retained from 48.6 ± 2.1 % to 83.1 ± 0.7 % of C-PC, indicating that sugars had a beneficial effect on enhancing the thermal stability (Huo et al., 2022). Some studies confirmed that the stabilizing effect produced from sugars is attributable to the glycosidic bond between sugar and protein (Hadiyanto et al., 2018; Stanic-Vucinic et al., 2018). Sodium azide and dithiothreitol are commonly used as preservatives for phycocyanin for analytical purposes, instead food grade preservatives such as citric, ascorbic and benzoic acids have been efficiently used (Kannaujiya and Sinha, 2016). Anyway, further investigation of alternative preservatives should be improved in order to improve C-PC application in food sector (Hsieh-Lo et al., 2019).

Encapsulation during the last years became one of the most used technologies to ensure the stability of sensitive compounds from temperature, light, oxygen and free radicals. Encapsulation is the process of covering or coating active compounds on a micrometric or nanometric scale with a carrier substance to create ultrafine particles or capsules. In particular, the use of spray drying has been identified as one of the most efficient and less expensive way to incapsulate sensitive products (İlter et al., 2021). This technology is based on the formation of microparticles using a mechanical centrifugal force or a pressure difference and a consequent removal of water using hot air avoiding thermal damages to phycocyanin (Pan-utai and Iamtham, 2020). In spray drying, during the rapid evaporation period the droplet temperature remains low, since them remains at the wet bulb temperature. However, heat damage can occur when the powder is exposed to hot air until it separates in the cyclone. Providing additional protection through double encapsulation while drying is one potential way to reduce heat damage to C-PC (Chandralekha et al., 2021). Among commonly used coadjuvants in this process, maltodextrin acts as a principal wall material thanks to its low cost and excellent film-forming capacity, cyclodextrins can form inclusion complexes that protect C-PC from oxidative degradation, and chitosan provides an extra coating layer with antimicrobial, antioxidant and film-forming properties (Yang et al., 2017), further stabilizing the encapsulated product during both spray drying and long-term storage. The combined action of these carriers not only helps maintain protein structural integrity and color brightness but also enhances the overall shelf life of the final encapsulated product.

7. Artificial neural networks for the optimization of phycocyanin extraction

Recent studies have designed a range of long-term bioprocesses to support the industrialization of high-density biomass cultivation, bioenergy generation, and high value bioproducts. At the end of extraction and purification processes, C-PC content obtained on industrial scale is even less than half of their maximum content observed in short-term experiments. Therefore, to overcome this challenge process optimization has to be executed to maximize bioprocess productivity and efficiency of this high-value bioproduct (del Rio-Chanona et al., 2016). Recently, new computational techniques have been applied to bioprocesses to improve their production yields, among which artificial intelligence (AI), machine learning (ML) and artificial neural networks (ANNs) stand out. AI encompasses computational methods and algorithms that enable machines to perform tasks typically requiring human intelligence, such as reasoning, learning, decision-making, and pattern recognition. A key subset of AI is the ML, where models learn from data (detecting patterns and relationships) rather than following hard-coded instructions. Within ML, ANNs are computational architectures inspired by biological neurons. In an ANN, interconnected nodes (neurons) process information simultaneously, adjusting internal “weights” as they receive input data and error feedback (Hilali et al., 2022). This iterative process, called backpropagation, allows the network to discover complex, often non-linear relationships that might be difficult or impossible to capture using traditional statistical methods (Imamoglu, 2024). As a result, ANNs can make high-fidelity predictions and optimize processes across various scientific and industrial fields.

In a recent study, Chong et al. (2024) developed a novel system for predicting the concentration of C-PC from L. platensis using ML and deep learning (DL) techniques. This study applied a non-invasive, vision-based system using RGB (red-green-blue), HSL (hue-saturation-lightness) and CMYK (cyan-magenta-yellow-black) color models combined with ML/DL techniques to predict C-PC concentration. Indeed, unlike traditional methods, which require chemical extraction and UV–VIS spectrophotometry, this approach eliminates the need for pre-extraction processes, providing a real-time and cost-effective alternative. The system integrates digital imaging, ML, and DL models to correlate color features with C-PC concentration, representing a unique application in microalgae bio-molecule detection (Chong et al., 2024).

However, the application of advanced methods like image augmentation and the ability to distinguish subtle color changes across growth stages could boost the system's accuracy, scalability, and practical use.

Hilali et al. (2022) instead, compared Response Surface Methodology (RSM) and Artificial Neural Networks (ANNs) to optimize ultrasound-assisted extraction of phycocyanin from Spirulina using different glycerol-based Natural Deep Eutectic Solvents (NaDES). By adopting a three-layer ANN with one hidden layer and training data via the Levenberg–Marquardt algorithm, they demonstrated that ANNs exhibit higher predictive accuracy (R² up to 0.994) than RSM under various extraction conditions (temperature, time, water ratio). Remarkably, glycerol:glucose (2:1) provided both superior extraction yields and enhanced phycocyanin stability (Hilali et al., 2022). Their findings highlight ANN as a highly efficient modeling tool, paving the way for greener, more effective approaches to microalgal pigment recovery.

Emerging research increasingly points to the feasibility of AI-based sensor integration and predictive modeling for identifying the best combination of physical (e.g., ultrasonication, bead milling), chemical (e.g., ionic liquids, deep eutectic solvents), and enzymatic treatments to disrupt cell membranes efficiently. ML-driven multi-objective optimization would further ensure that the process remains cost-effective and energy-efficient by balancing factors such as solvent usage, protein denaturation thresholds, and final extract purity. Additional refinements could be achieved by coupling ANNs with evolutionary algorithms (such as genetic algorithms) or fuzzy inference systems to enable a closed-loop, adaptive approach, continuously updating operational conditions based on real-time data. Integrating AI, ML, and ANN in this manner could accelerate process development, delivering highly reproducible, scalable, and sustainable methods for extracting this valuable natural blue pigment from Limnospira spp.

8. Challenges and future work

Main challenges related to the phycocyanin extraction from Spirulina are related to the development of new strategies to pursuit high yield and purity grade in the extract. Nevertheless, increasing improvements, as described for ultrasound assisted extraction as well as enzymatic hydrolysis, there are some aspects which still limit these technologies. In addition, small production scale and cost effectiveness struggle the application of the most efficient technologies like freeze and thawing method, but also enzyme-assisted extraction. Moreover in this context, the improvement of these methods could provide significant results also in the recovery of phycocyanin extraction by-products, which could provide an alternative protein source for functional food development (Sefrienda et al., 2023). Thus, further research and development costs may be incurred in the process of creating and deploying more effective extraction techniques, especially in terms of environmental sustainability.

9. Conclusion

In recent years, the rising demand for C-PC, particularly as a blue pigment in the food industry, has sparked growing interest in research focused on optimizing its production and processing efficiency. Cyanobacteria of the genus Limnospira are currently the main biomass for the production of this precious pigment and its large-scale cultivation is now a consolidated reality in the context of biotechnological cell farming. The results of the research show that C-PC extraction methods must be specific to the strain and that in addition to the efficiency and quality of the extract, it is also necessary to look both at the safety and sustainability aspects of processing.

In this review, we have surveyed different extraction methods used to obtain C-PC, highlighting both the varying approaches to data reporting and the broad discrepancies in yields and concentrations achieved As summarized in Table 1, it is difficult to identify a single “best” method because of wide methodological variations (such as strain differences, biomass pretreatments, and solvent systems) that can drastically influence reported yields. While Freeze-thawing gave the best results in terms of C-PC recovery, in particular when coupled with ultrasound assisted extraction, ultrasound, PEF and high-pressure homogenization showed an easier C-PC extraction. It is worth to note that for each extractive method, results from studies were accomplished by optimum condition of temperature, pH, incubation time and other important parameters which could affect the extraction. Also, this review identified the main factors affecting the chemical degradation of C-PC (i.e. temperature, pH, oxidation and light) and the methods that can be used to increase its stability. Improving extraction and purification technologies in Limnospira spp., could be a starting point to explore the same strategies also in other cyanobacteria characterized by a highest C-PC productivity.

Finally, research effort must be made to optimize the methods to achieve high C-PC recovery and purity at lowest energetic, environmental and economic costs tacking in consideration also the residual biomass processing and utilization in a frame of biorefinery approach. In this context, the potential of enzyme-based approaches warrants deeper investigation, as it offers a green, sustainable alternative that can reduce solvent use and harsh processing conditions.

CRediT authorship contribution statement

Mariacristina D'Ascoli: Writing – original draft, Writing – review & editing. Antonio L. Langellotti: Conceptualization, Writing – review & editing. Giovanni L. Russo: Conceptualization, Writing – original draft, Writing – review & editing. Angela Sorrentino: Writing – review & editing. Prospero Di Pierro: Supervision, Resources, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Professor Aiqian Ye

Data availability

No data was used for the research described in the article.