CHAPTER 1: Structures, Nomenclature and General Chemistry of Carotenoids and Their Esters

Carotenoids are widespread isoprenoids in nature that intervene in many actions ranging from the collection of light and photoprotection to the regulation of gene expression and communication within or between species, to mention some examples.¹  Interestingly, they can be chemically or enzymatically converted into other derivatives that can act as compounds with vitamin activity, phytohormones or aromas, among others.² 

Carotenoids are, therefore, much more than natural pigments providing mainly yellow, orange or red colours. Indeed, they are very versatile compounds that are eliciting increasing interest in different disciplines, such as plant science, agriculture, food science and technology, nutrition and health, among others. In relation to these three latter disciplines, although their roles as natural pigments and precursors of retinoids with vitamin A activity have long been known, the renewed interest in these compounds is mainly due to a large body of evidence accumulated in the last 30 years indicating that they may be health-promoting compounds and may be important in the context of functional foods.³  Thus, carotenoids are thought to contribute to reducing the risk of developing certain types of cancers, as well as cardiovascular, eye, skin or bone diseases, and are even thought to be beneficial to cognitive function.^4–8  Furthermore, carotenoids can provide cosmetic benefits.⁹  There is evidence that such cosmetic benefits can lead certain populations to increase their intake of carotenoid-containing products, which can be used in the context of public health as a strategy to promote the consumption of fruits and vegetables.¹⁰  Although the beneficial health-promoting effects of carotenoids are often attributed to their possible role as antioxidants, other underlying mechanisms should be considered (e.g. pro-oxidant or anti-inflammatory actions or modulation of membrane properties, among others).^4,11–17  Interestingly, carotenoids and/or their derivatives can play essential roles in cell signalling pathways, such as by interacting with transcription factors such as nuclear factor erythroid 2-related factor 2 (Nrf2)^18,19  or nuclear factor-κB (NF-κB).^20,21 

Although carotenoids in human fluids and tissues are found almost exclusively free (although sometimes carotenoid esters have been reported in plasma and skin^22,23  at levels markedly lower compared to unesterified ones, and recently in the colostrum, but not in mature human milk²⁴ ), in foods they can be associated with other molecules, such as sugars, proteins or fatty acids. Such associations can result in significant changes in their properties. Indeed, it is very common for food xanthophylls to be in the form of esters, especially in many fruits. Although esterification does not have an impact on the carotenoid chromophore and therefore on its colour, it can markedly modify its solubility and susceptibility to oxidation, which in turn can have an impact on relevant aspects that explain their levels in foods and humans, such as their biosynthesis, deposition, stability and bioavailability, among others.²⁵  The study of the esterification of carotenoids is undoubtedly gaining popularity and has been greatly facilitated by important advances in analytical techniques that make possible the separation and identification of a great variety of carotenoid esters.²⁶ 

In this chapter, nomenclature and structural aspects related to fatty acids, isoprenoids, carotenoids and, finally, carotenoid associations with other molecules, with an emphasis on carotenoid acyl esters, are dealt with, referring to some physical–chemical properties and their importance at different levels.

Fatty acids (FAs) are ancient and ubiquitous molecules present in all living matter. Both free and as part of complex lipids, they play a number of key roles in metabolism as critical structural components of phospholipids and other complex lipids in cellular membranes, as gene regulator and as parts of triacylglycerols, a major metabolic fuel (for the storage and transport of energy).

The quantitative proportion and qualitative composition of FAs in various organisms are characteristic for every species and genus, and depending on the environment, they can be used for identification purposes.²⁷  In bacteria, more than 300 FAs and related compounds have been found. Large microbial libraries are available, such as the Sherlock System library, which has over 1500 bacterial species, along with 200 species of yeast, identified by the composition of FA methyl esters (FAMEs).²⁸  FAs have also been used for chemotaxonomic purposes in cyanobacteria²⁹  and microalgae.³⁰  Microalgae are the primary producers of long-chain polyunsaturated FAs that accumulate through the various trophic levels. The FAs derived from microalgae are gaining interest because of their potential application in the food³¹  and biofuel industries.³² 

In the plant kingdom, an amazing variety of FA structures are found, particularly in angiosperm seed oils. An electronic database of seed oil FA (SOFA) composition (http://sofa.mri.bund.de) at the Max Rubner-Institut has been made available to researchers for different purposes, from biochemical systematics and plant phylogeny chemotaxonomy to the search for genes useful for tailor-made industrial fats.³³ 

In animal tissues, the FA profile reflects the tissue biosynthesis and the FA composition of ingested lipids, especially for monogastrics (pigs, poultry and rabbits), since ruminants can hydrogenate FAs in the rumen.³⁴  Some FAs also have important metabolic roles as biosynthetic precursors of oxylipins, including the eicosanoids (prostaglandins, leukotrienes, thromboxanes and lipoxins) and docosanoids (protectins, resolvins and maresins), while in plants, hormones, such as the jasmonates, are derived from α-linolenic FA.

FAs are defined by the International Union of Pure and Applied Chemistry (IUPAC) as “aliphatic monocarboxylic acids derived from or contained in esterified form in animal or vegetable fat, oil or wax which may be saturated or unsaturated. By extension, the term is sometimes used to embrace all acyclic aliphatic carboxylic acids.”³⁵ 

Classification of FAs in classes and subclasses can be done according to the LIPID MAP, available at the LIPID MAPS-Nature Lipidomics Gateway.³⁶  This is a free and comprehensive online resource, providing tutorials and instructional material and experimental data for lipids and genes, along with protocols and standards, databases of lipid structures and lipid-associated genes or proteins and a variety of lipidomics tools. The database is accessible through any web browser (http://www.lipidmaps.org).

According to this classification, the fatty acids (FA) are a diverse group of molecules synthesised by chain elongation of an acetyl-CoA primer with malonyl-CoA (or methylmalonyl-CoA) groups that may contain a cyclic functionality and/or are substituted with heteroatoms. Thirteen subclasses are known, indicating the large variety of structures that can be found in nature. The first subclass includes the most common straight-chain saturated FAs (SFAs) containing a terminal carboxylic acid. Although several hundred FA forms have been identified in nature, the number frequently occurring in the common lipids is much fewer (from 10 in plants to about 20 in animal tissues). Most common FAs consist of a straight chain with an even number of carbon atoms (from 12 to 22) because the biosynthetic pathway common to all organisms involves chemical linkage of two-carbon units together, though FAs with shorter, longer and odd-numbered chain also exist in nature.³⁷ 

The FAs can be classified into three categories: SFAs, which lack unsaturated linkages between carbon atoms, and unsaturated FAs, which can be further divided into monounsaturated FAs (MUFAs) and polyunsaturated FAs (PUFAs), depending on the number of unsaturated bonds (one or more, respectively). In PUFAs, double bonds are usually separated by a single methylene group, which is known as a methylene-interrupted pattern.³⁸  Although much less common, ‘conjugated FAs’, where double bonds are not separated by a methylene group, are also found, most of them derived from the unconjugated structure of linoleic acid (LA; 18:2n-6). Conjugated FAs are now gaining much interest because some of them have shown special nutritional properties. Depending on the FA chain length, the FAO/WHO Expert Consultation³⁹  recommends a further division into four subgroups in the SFA group: short chain (3–7 carbons), medium chain (8–13 carbons), long chain (14–20 carbons) and very long chain (>21 carbons). Similarly, unsaturated FAs are also further classified into three subgroups: short chain (≤19 carbons), long chain (20–24 carbons) and very long chain (≥25 carbon atoms).

Unsaturated FAs show isomerism, which can be either positional or geometrical. The positional isomers occur when the double bond(s) are located at a different position in the carbon chain. Geometrical (cis/trans) stereoisomerism occurs when substituents are arranged differently in space due to restricted rotation of a double bond in the molecule. When the two ligands attached to separate atoms connected by the double bond lie on the same side of a plane, they are considered to be located at the cis position to each other. If they are on opposite sides, their relative position is described as trans. For alkenes, the terms cis and trans may be ambiguous and have therefore been replaced by the E/Z convention, which is based on the application of sequence rules (Figure 1.1).³⁵ 

The geometry of the double bonds in eukaryotes is exclusively cis and usually positioned at the third, sixth or ninth carbon from the terminal methyl group. The significance of the ubiquitous cis structural feature of the unsaturated lipid double bond is due to its contribution to the organisation of phospholipids in one of the most important units of living organisms: the cell membrane.⁴⁰  Membrane homeostasis is obtained by a precise balance between saturated and cis unsaturated structures as the key feature of the homeoviscous adaptation.⁴¹  Because of its essentiality in living organisms, the cis geometry is strictly controlled by the regiospecific and stereoselective enzymatic activity of desaturases during MUFA and PUFA biosynthesis.^42,43 

FAs with double bonds of trans (or E) configuration are occasionally found in natural lipids from ruminant animal tissues, where they are either naturally formed by biohydrogenation in the rumen or during industrial processing (hydrogenation) and so enter the food chain, but in minor amounts. The suitability of trans-FA (TFA) for human nutrition has been widely discussed and related to an increase in all-cause mortality.⁴⁴  In fact, both industrial and ruminant sources of TFA consumption have been positively associated with an increased risk of cardiovascular disease (CVD).⁴⁵  For this reason, the elimination of industrial TFA in foods has been proposed, and limits to their content have been legislated in many countries.⁴⁶ 

The terminology of FAs can be confusing due to the different nomenclature systems in use. Whatever the system used, it must clearly identify the structure and indicate the different aspects described above: carbon chain, number of unsaturation, stereoisomeric configuration and the position of the first double bond in relation to the carboxylic or methyl end groups, along with the presence of other substituents (e.g. branched chains, ring systems and oxygen groups).

Although trivial names give no clue as to the structure, they are most frequently used in the case of FAs. These names are derived from a common source of the compound or the source from which it was first isolated (e.g. myristic acid was first identified in seed oils from the Myristicaceae family and oleic acid is the major constituent of olive oil [oleum]). Similar names may correspond to very different structures (e.g. arachidonic and arachidic acids; both have 20 carbons but different structures, since one is a PUFA [four double bonds] and the other is a SFA). Similarly, isomeric forms have different trivial names (e.g. oleic [18:1 cis] and elaidic [18:1 trans] acids).

Systematic nomenclature is more technically clear and descriptive. It is derived from the official chemical nomenclature established by the IUPAC.⁴⁷  FAs are numbered with the carbon atom of the carboxyl group as C-1. Shorthand and abbreviation notations, like trivial names, save space and contribute to rapid understanding. For example, FAs are named by their number of carbon atoms and their number of double bonds after a colon (e.g. 18:0 stands for stearic acid and 18:1 for oleic acid). For unsaturated FAs, two different abbreviations are used to make clear where the double bonds are located in molecules. In the case of cis/trans-Δx, the double bond is located on the xth carbon–carbon bond, counting from the carboxyl end group. The cis or trans (Z/E) notation indicates whether the molecule is arranged in a cis or trans conformation. This is the IUPAC systematic recommended notation.⁴⁷  In the ‘n minus’ (n-x) system, also known as the omega system (this notation is discouraged today, although it is widely used), n is the number of carbon atoms in the chain and x is the (lower) locant of the double bond closest to the methyl end of the molecule. This system easily defines the different metabolic series, such as n-9, n-6 and n-3, etc. The ‘n minus’ system is applicable only to cis unsaturated FAs and to those cis PUFAs whose double bonds are arranged in a methylene-interrupted manner.

Different denominations of common FAs and information about some of their sources are shown in Tables 1.1–1.3.

The physical properties of the different FAs, such as solubility, melting point and susceptibility to oxidation, will depend on the number of carbon atoms of the molecule and the number of double bonds. The FAs will also determine the physical properties of the molecules in which they take part as components.

FAs are amphipathic molecules containing a hydrophobic (the aliphatic moiety) and a hydrophilic (carboxyl group) part. Short-chain FAs (C < 4) are freely soluble in water, either in their protonated or ionised forms, with pKa values of about 4.5; however, long-carbon-chain FAs are poorly soluble in water. As the aliphatic chain length of the FA increases, the protonated FA becomes much less soluble; thus, the aqueous solubility of FAs with more than 12 carbons is quite low. On the other hand, at high pH values, a negatively charged carboxylate group (COO⁻) is formed, and the ionised FA is fairly water soluble; this property gives ionised FAs their detergent properties. Thus, the actual water solubility, particularly of FAs with the longest chains, is often very difficult to determine since it is markedly influenced by media pH and temperature, and also because FAs tend to associate, leading to the formation of monolayers or micelles.⁴⁸ 

A major factor that affects the melting point of FAs is the geometric shape of the molecules. SFAs are more linear than unsaturated FAs with cis double bonds, allowing them to closely pack together with the consequently high potential for attractive intermolecular interactions. Therefore, SFAs have higher melting points than unsaturated FAs with the same carbon number. However, ‘even’ FAs show higher melting points than those of the ‘odd’ FAs immediately below and above them.⁴⁹  In the case of unsaturated FAs, the presence of double bonds produces a large decrease in the melting point. Therefore, oils with a high proportion of unsaturated FAs have a low melting temperature. Animals that live at low temperatures (i.e. fish) have a large proportion of long-chain unsaturated FA to prevent their fats from solidifying. Table 1.4 shows the melting point values of some FAs.⁵⁰  Although unsaturation has a pronounced lowering effect on the melting point, in isomeric pairs, the trans form of the acid shows a higher melting temperature.⁵¹ 

The susceptibility of FAs to oxidation is dependent on their degree of unsaturation. SFAs are very stable, but as the number of double bonds increases, the susceptibility to oxidation also increases. The relative oxidation rates of oleic (18:1), linoleic (18:2) and linolenic (18:3) acids were reported to be 1:27:77.⁵²  The carbon–hydrogen bond strength in double bonds is reduced in comparison to the aliphatic chain of stearic acid (99 kcal mol⁻¹ vs. 80 kcal mol⁻¹ in oleic and 69 kcal mol⁻¹ in LAs). This reduction of bond strength allows hydrogen to be more easily abstracted from the FA, leading to the formation of free radicals—the first step in the lipid oxidation cascade.⁵³ 

However, there is evidence indicating that the kinetics of FA oxidation depends upon the milieu in which it reacts with oxidants. For example, aqueous environments, such as those at the cell membrane/plasma and cytosol/cell membrane interfaces, yield different oxidation profiles than organic ones. In fact, some experiments in cell cultures show that some FAs might indirectly act as anti- rather than pro-oxidants in vascular endothelial cells, hence decreasing inflammation and, in turn, the risk of atherosclerosis and CVD.⁵⁴  Enzyme-catalysed oxidation is the initial step in the production of eicosanoids and jasmonates, which are biologically active metabolites in animals and plants, respectively.⁵⁵ 

The biosynthesis of FAs occurs in all living organisms in the cytosol and requires NADPH and acetyl-CoA. The standard way for cells to achieve this is through the FA synthesis cycle. This cycle includes eight enzymes (acyl-CoA synthase, acyl-CoA carboxylase, acyltransferase, ketoacyl synthase, ketoacyl reductase, hydroxyacyl dehydratase, enoyl reductase and thioesterase) and is initiated with acetic acid, CoA and ATP to yield acetyl-CoA, with acyl-CoA synthase as the catalyst. Acetyl-CoA is converted to malonyl-CoA by a biotin-dependent acetyl-CoA carboxylase. This irreversible reaction is the limiting step in the biosynthesis of FA Malonyl-CoA and NADPH are used by the multi-enzyme FA synthase to yield palmitate. This mechanism leads to a wide variety of lipids that contain the fatty acyl chain, including FAs, phospholipids and glycerolipids. In animals, biosynthesis occurs primarily in the liver, adipose tissues, central nervous system and lactating mammary gland.⁵⁶ 

The enzymes of FA biosynthesis are divided into two groups. While in animals and fungi the FA synthase (FASI) is a multifunctional protein encoded by a single gene,⁵⁷  in plants, bacteria and lower eukaryotes the enzyme is encoded by two genes (FASII), and their polypeptide products coalesce to form a multifunctional complex.⁵⁸  While FASI produces only palmitate, FASII is capable of producing a large diversity of FAs, with different chain lengths from unsaturated FAs, iso- and anteiso-branched-chain FAs, to hydroxy FAs.

FAs can further be elongated into very long chains by individual membrane-bound enzymes named elongases, located in the endoplasmic reticulum. The synthesis of very-long-chain FAs is a ubiquitous system found in different organisms and cell types.⁵⁹  FA desaturase, an enzyme in the endoplasmic reticulum, introduces double bonds between carbons 9 and 10 in palmitate and stearate, producing palmitoleic (16:1:Δ9) and oleic (18:1:Δ9) acids, respectively. Mammals lack the Δ12- and Δ15-desaturase enzymes necessary for desaturation of 18-carbon FA at the n-3 (or Δ15) or n-6 (or Δ12) positions. Thus, LA (18:2:Δ9,12) and α-linolenic acid (ALA; 18:3:Δ9,Δ12, Δ15) are essential FAs (EFAs) that must be supplied by the diet because the body cannot synthesise them. By comparison, plants and algae contain the enzymes Δ12- and Δ15-desaturase, and as a result, LA and ALA are two of the most prevalent FAs found in plant tissues and oils.⁶⁰  EFAs are metabolised to their respective long-chain metabolites: dihomo-γ-linolenic acid (DGLA) and arachidonic acid (AA) from LA; and eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) from ALA. Some of these long-chain metabolites form precursors to respective prostaglandins (PGs), thromboxanes (TXs) and leukotrienes (LTs), lipoxins (LXs) and resolvins.⁶¹ 

Among lipids, FAs are of crucial relevance to the structure and physiology of the body because, either free or as part of a complex, lipids have a range of biological activities, from storage to energy transport, as essential components of all membranes or as gene regulators.⁶² 

As part of triglycerides, FAs are the primary source of energy (9 kcal g⁻¹ or 37.62 kJ g⁻¹), and in infants provide more than 50% of the daily energy requirements. All FAs can be oxidised by most aerobic tissues, but not the brain; however, the specific energy yield depends on the structure of the FA, providing an alternative to glucose. FA oxidation becomes important in times of limited glucose availability.

As components of cell membranes, phospholipids influence the physical nature of the membrane (called ‘fluidity’), which in turn influences the function of membrane proteins and the movement of proteins within the membrane. PUFAs confer distinctive properties on the membranes, particularly decreasing their rigidity, while SFAs and MUFAs ensure that there is a balance between rigidity and flexibility. Indeed, SFAS and 2-hydroxy FAs in sphingolipids appear to give additional rigidity and hydrogen-bonding stability to the sub-domains of membranes termed ‘rafts’. The FA composition of phospholipids may be influenced by diet, metabolism, hormonal milieu, state of cell activation and genetics, among other factors.⁶³ 

FAs released from membrane lipids or taken up into cells can have specific metabolic, functional or signalling roles, such as diacylglycerols, ceramides, lysophospholipids and endocannabinoids, and there is evidence that the FA composition of those signalling molecules influences their biological activity.⁶⁴  Some FAs are of an essential character and are precursors of powerful, locally acting metabolites (i.e. the eicosanoids and docosanoids [of 20 and 22 carbon atoms, respectively], such as LTs, PGs, TXs, prostacyclins, protectins and resolvins).^64,65 

Some FAs can regulate the expression or activity of transcription factors, so they play a role in controlling gene expression and protein production by cells. These effects enable FAs to regulate metabolic processes such as FA synthesis and oxidation, lipoprotein assembly and clearance, insulin sensitivity and inflammation.⁶⁶ 

Isoprenoids, which have also been termed terpenes or terpenoids, are considered the biggest family of natural compounds in living organisms. These compounds are ancient and widespread; they have been identified in sediments dating back ca. 2.5 billion years, and over 23 000 members of the family have been identified in the most diverse organisms.^67,68  Being a vast group of compounds, they can be classified into different groups depending on the number of such isoprene units (Figure 1.2, Table 1.5).^68–70 

They are built by consecutive condensations of building blocks of five carbon atoms, namely isopentenyl diphosphate (IPP) and its isomer, dimethylallyl diphosphate (DMAPP) (Figure 1.3).

These isoprene building units are biosynthesised via two distinct metabolic routes, namely the mevalonate (in eukaryotes, Archaebacteria and the cytosol of higher plants) and the non-mevalonate pathways (eubacteria, green algae and the plastids of higher plants), discussed in details in Chapter 4.

Being ancient and widespread natural compounds, it is not surprising that isoprenoids are involved in key processes. In fact, they can act as regulators of gene expression, modulators of membrane properties, vitamins, antimicrobial agents, hormones, pheromones, electron transporters, pigments, aromas, flavour compounds, etc.^67,70 

Carotenoids are biosynthesised by all photosynthetic organisms and by some non-photosynthetic bacteria and fungi.⁷¹  In general, animals cannot synthesise them de novo, although they are able to modify them. As an example, humans and other mammalians can express at least two carotenoid-cleavage oxygenases (usually termed as β,β-carotene 15,15′-monooxygenase and β,β-carotene 9′,10′-dioxygenase) that cleave carotenoids into oxidised derivatives, respectively, retinoids and apocarotenoids.^72,73  However, it was recently demonstrated that some arthropods, such as aphids and adelgids, among others, can biosynthesise carotenoids thanks to the presence in their genomes of carotenogenic genes laterally transferred from fungi. As a result, these arthropods could be regarded as ‘natural transgenic organisms’.²⁵ 

Carotenoids are widespread in nature. It is estimated that they are present in ca. 700 organisms belonging to the three domains of life, more specifically in ca. 10 organisms within archaea, ca. 170 bacteria and ca. 500 eukaryotes.⁷⁴  Moreover, their occurrence is well described in different plant structures (photosynthetic tissues, petals, anthers, stigmas, fruits, seeds, roots), land and water animals (sponges, jellyfish, fish, molluscs, arthropods, reptiles, mammals, birds, etc.), macroscopic algae and fungi and a myriad of microorganisms including cyanobacteria, one of the first inhabitants of our planet.³  Indeed, it is noteworthy that carotenoids are found in organisms adapted to the most disparate environmental conditions, from the bottom of the ocean to glaciers, thermal ponds, hypersaline waters or even very dry, oxidising or radioactive conditions.^75–79 

The main structural feature of carotenoids is their system of conjugated double bonds (c.d.b.), which is usually termed the ‘polyene chain’. The colourless carotenoids phytoene (three c.d.b.) and phytofluene (five c.d.b.) are rarities within the carotenoid family as their systems of c.d.b. are much shorter than those of most carotenoids (Figure 1.4). Such clear differences concerning other carotenoids are expected to have an impact on the properties and actions of these colourless carotenoids, and this is eliciting much interest at different levels, including in the promotion of health and cosmetics.^9,80 

Although there are exceptions, a typical carotenoid is a tetraterpenoid containing eight isoprenoid building blocks; hence, it usually has 40 atoms of carbon. Depending on the presence or absence of rings in their molecules, they can be classified into two main groups: cyclic or acyclic carotenoids, respectively. The different basic end groups described in carotenoids are shown in Figure 1.5 and further information is given in Table 1.6.

The numbering of the atoms of carbon goes from the ends to the centre of the molecule, from 1 to 15 on one side of the molecule and from 1′ to 15′ on the other. The methyl groups are counted from 16 to 20 and from 16′ to 20′, respectively (Figure 1.6).^81,82 

Similarly, carotenoids can be classified into two groups regarding the presence or absence of oxygen in their molecules. Hydrocarbon carotenoids—that is, those exclusively containing carbon and hydrogen atoms in their molecules (e.g. α-carotene, β-carotene, γ-carotene, ζ-carotene, lycopene, neurosporene, phytoene or phytofluene)—are termed carotenes (Figure 1.7).

Carotenoids also containing oxygen are termed xanthophylls. Among the most common oxygenated functional groups in carotenoids from food are hydroxyl groups, present for instance in β-cryptoxanthin, lutein and zeaxanthin (Figure 1.8).

Other oxygenated functional groups in xanthophylls are the epoxides (antheraxanthin, neoxanthin, violaxanthin) or furanoids (auroxanthin) (Figure 1.9).

Carbonyl groups can also be found in carotenoids (e.g. astaxanthin, canthaxanthin, capsanthin, β-citraurin; Figure 1.10). Other oxygenated groups that can be naturally found in carotenoids are carboxylic, acetate, lactone or sulphate groups.^82,83 

Apart from these two general classifications of carotenoids as cyclic or acyclic or carotenes and xanthophylls, other subgroups of carotenoids can be distinguished by their structure. Thus, there are bacterial carotenoids that have one or two additional isoprenoid units and contain, respectively, 45 or 50 atoms of carbon. One typical example is decaprenoxanthin (Figure 1.11).

There are also two distinct subgroups of carotenoids that contain fewer than 40 atoms of carbon. Norcarotenoids (e.g. peridinin; Figure 1.12) are carotenoids that lack one to three atoms of carbon in the central hydrocarbon backbone. On the other hand, apocarotenoids lack fragments at one or both ends of the molecules (e.g. the saffron carotenoid crocetin, which contains 20 atoms of carbon; Figure 1.12).

In addition, carotenoids in which a bond between adjacent carbons (except carbons 1 and 6 in rings) has been broken (e.g. semi-β-carotenone) are termed secocarotenoids (Figure 1.13).

Lastly, there are carotenoids in which the system of c.d.b. is shifted, such that there is a simple bond between carbons 15 and 15′ rather than the usual double bond. Carotenoids with this structural feature (e.g. rhodoxanthin) are termed retrocarotenoids (Figure 1.10).⁸³ 

Apart from these natural carotenoids, more than 150 carotenoids that contain heteroatoms have been chemically synthesised (Figure 1.14).⁸⁴ 

Different spatial isomers of carotenoids can exist. Featuring many double bonds in their molecules, carotenoids can exist as all-trans (all-E) or cis (Z) isomers, which can markedly differ in shape. Furthermore, many carotenoids have chiral centres in their molecules, so that different optical isomers can exist. Concerning the stereochemistry of carotenoids, it is important to note that, although a specific carotenoid isomer can adopt many different shapes in space, a specific preferred conformation of low energy is expected to exist or predominate.⁸² 

Geometrical isomerism refers to the relative position of substituents around a planar carbon–carbon double bond. Carotenoid geometrical isomers are often designated using the cis/trans designations, although the E/Z designation is considered to be more precise. This is based on the application of sequence rules (Figure 1.1).⁸⁵  Different geometrical isomers of a carotenoid can significantly differ in size and shape, as can be readily observed in Figure 1.15.

In general, the (all-E) isomers of carotenoids are the most stable and therefore the most abundant. Most theoretical Z isomers are not detected as there are important steric hindrances in the carotenoid structure (e.g. in positions 7 and 7′). On the other hand, some Z isomers are commonly found in nature and can readily be formed in carotenoid extracts.^82,86  Some typical examples of Z isomers of carotenoids that occur naturally in certain sources are (15Z)-phytoene (usually the major isomer in carotenogenic organisms),⁸⁷  bixin (a 9Z isomer) in the seeds of Bixa orellana⁸³  or the highly sterically hindered (7Z,9Z,7′Z,9′Z)-lycopene (prolycopene), a major carotenoid in tangerine tomatoes (Figure 1.16).⁸⁸ 

In any case, the presence of Z isomers of carotenoids in any source should always be interpreted with the understanding that they may be formed as a result of diverse industrial or culinary treatments or during carotenoid handling in the laboratory.^89–91 

A molecule that contains carbon atoms to which four different substituents are attached (that is, asymmetric carbon atoms, which constitute a chiral or stereogenic centre) can exist in different configurations. A classic example of a carotenoid with a chiral centre is zeaxanthin, which can exist as different isomers, namely (3R,3′R)-zeaxanthin, (3S,3′S)-zeaxanthin and (3R,3′S)-zeaxanthin (meso-zeaxanthin) (Figure 1.17). The discernment of the different optical isomers of zeaxanthin is necessary as two of them, (3R,3′R)-zeaxanthin and (3R,3′S)-zeaxanthin (meso-zeaxanthin), are found in the macula lutea of the human retina.⁹² 

Carotenoids, like other molecules, can form aggregates as a result of weak and reversible bonding by hydrogen bonds, van der Waals interactions, dipole forces and the hydrophobic effects of hydrophobic molecules, their polar groups and the surrounding solvent. Obviously, the aggregates have different physical–chemical properties as compared with the individual molecules.^93,94 

Carotenoids are known to form two different kinds of aggregates when they are in hydrated polar solvents. In one of them, the carotenoid molecules are stacked such that their unsaturated backbones are parallel to each other and tightly packed (the so-called H-aggregates). These self-assemblies are characterised by a pronounced blue shift (hypsochromic shift) of their absorption maxima as compared to the single molecule. Contrastingly, in the so-called J-aggregates, the backbones of c.d.b. are organised in a more head-to-tail fashion, producing a loose association of the carotenoid molecules. These aggregates are characterised by exhibiting a red shift (bathochromic shift) of their absorption maxima in comparison to the single carotenoid molecules. It is thought that both types of aggregates can form assemblies resembling ladders, brickwork or staircases (Figure 1.18) and that the formation of carotenoid J- or H-aggregates in hydrated solvents (e.g. hydroalcoholic mixtures) depends on the pH, the initial concentration of the carotenoid and the ratio ethanol to water.⁹⁴  This subject is further explored in Chapter 2.

The electron-rich polyene backbone of carotenoids makes them very susceptible to oxidative enzymatic or non-enzymatic breakdown. As a result of such transformations, many different compounds that retain some characteristic carotenoid structural features are formed. Like carotenoids, these compounds can be involved in important biological actions and have important applications. Thus, within this miscellaneous group, there are compounds with vitamin A activity, hormones or aroma/flavour compounds.²  The derivatives resulting from the breakdown of carotenoids can be either symmetric or asymmetric and can undergo further cleavage.⁶⁹ 

Carotenoids can be oxidatively cleaved by means of non-enzymatic, non-specific reactions, such as via singlet oxygen, lipoxygenase co-oxidation or photooxidation.⁹⁵  On the other hand, carotenoid cleavage oxygenases (CCOs) are a family of non-haem iron enzymes that intervene in the oxidative breakdown of carbon–carbon double bonds in different locations of the polyene backbone, producing derivatives containing carbonyl functions (aldehyde or ketone groups) in the cleaving ends. Some of these enzymes, which can be found in plants, algae, fungi, bacteria, mammals and other animals, act specifically on apocarotenoids and are usually named apocarotenoid cleavage oxygenases.⁹⁶ 

Retinoids are diterpenes formed by four isoprene units joined in a head-to-tail manner. Hence, they have 20 atoms of carbon, and some retinoids exhibit vitamin A activity. In mammals, the cleavage of carotenoids into retinoids is catalysed by a cytoplasmatic non-haem iron oxygenase enzyme usually referred to as β,β-carotene 15,15′-monooxygenase 1 (hereafter CCO1). This CCO can centrally cleave β-carotene to produce two molecules of all-trans-retinal (all-E-retinal), which can be irreversibly oxidised into retinoic acid by retinal dehydrogenase or reversibly reduced into retinol by a retinal reductase (Figure 1.19). Apart from β-carotene, CCO1 can also cleave other carotenoids as long as they have at least one unsubstituted β-ring, a condition met by other common dietary carotenes (α-carotene) and xanthophylls (e.g. β-cryptoxanthin and some β-apo-carotenals).^72,73,97,98 

Besides CCO1, mammalian genomes can also encode at least another non-haem iron oxygenase enzyme that catalyses the cleavage of carotenoids. This is usually named β,β-carotene 9′,10′-dioxygenase and can cleave β-carotene, as well as other provitamin A and non-provitamin A carotenoids (e.g. lycopene, zeaxanthin and lutein) as it exhibits a wider substrate specificity at both the 9,10 and 9′,10′ double bonds. As a result, both non-volatile apocarotenoid and volatile compound cleavage products are formed (Figure 1.20).^72,73 

Apocarotenoids derived from the cleavage of dietary carotenoids in humans are eliciting increasing interest as they may have health-promoting effects through their interaction with cell signalling pathways, such as in relation to carcinogenesis or by protecting against oxidative stress.^18,99,100 

Carotenoids can be cleaved at different asymmetrical locations, giving a series of carbonylic odorant compounds with 9–13 atoms of carbon (Figure 1.21).

These oxidised breakdown derivatives of carotenoids are called norisoprenoids and can be formed in reactions either catalysed or not by enzymes, and either by direct degradation of carotenoids or via glycosylated intermediates.^69,101 

An important derived aroma compound is safranal, a potent aroma compound mainly responsible for the scent of the saffron spice, which can be produced via the enzymatic cleavage of zeaxanthin or by thermal treatments.^102,103  Other norisoprenoids that are potent aromatic compounds are β-ionone, β-damascenone and β-cyclocitral (Figure 1.22), among many others. In this way, carotenoids are considered to be very important for the production of the typical aromas of not only flowers, such as violets or roses,⁶⁹  but also widely consumed products such as tomatoes, grapes, raspberries, wines, tea and watermelon.^{2,101,104–107} 

The glycosylated monoterpene picrocrocin (Figure 1.23) is the main contributor to the sour taste of saffron spice and is the precursor of safranal.^102,103 

Trisporic acid (Figure 1.2) is a carotenoid breakdown derivative that is formed from β-carotene via retinol. It acts as a regulator of fungal sexual reproduction in some moulds, which is often accompanied by elevated biosynthesis of β-carotene. Trisporic acid is harnessed for the commercial production at industrial scale of this carotene from Blakeslea trispora.² 

The grasshopper ketone (Figure 1.24) was first found in the frothy exudate of the grasshopper Romalea microptera when it was disturbed.¹⁰⁸ 

The scented compound β-ionone (Figure 1.22) has recently been shown also to have a repellent effect against certain insects in plants.¹⁰⁹ 

Abscisic acid (ABA; Figure 1.25) is a phytohormone derived from the cleavage of the (9Z)-isomers of the epoxycarotenoids neoxanthin and violaxanthin. ABA intervenes in a broad range of actions including senescence and abscission of leaves, dormancy of buds and seeds, stomatal closure, seedling development and tolerance to diverse kinds of stress, among others.^2,110,111 

Interest in the study of the production and actions of strigolactones, a ‘new’ kind of carotenoid-derived plant hormone, has been expanding in recent years. The reasons for this are that these compounds favour the establishment of arbuscular mycorrhizae, the parasitisation of roots by other plants and the adaptation of plant architecture to the availability of nutrients, among others.^112–114  The structures of some members of the family—carlactone and strigol—are shown in Figure 1.26.

Carotenoids are best known by their trivial names, usually deriving from the sources from which they were first described or where they are particularly abundant. For instance, β-carotene owes its name to the scientific name of the carrot (Daucus carota).¹¹⁵ 

Apart from this traditional designation, there is also a common semi-systematic nomenclature that has the advantage of providing information about the structure of the carotenoid. In this case, the two halves of the carotenoid are considered, and the compound is named as a derivative of the corresponding carotene. For this purpose, references to the end groups are made using Greek letters (Figure 1.5, Table 1.6). In addition, numbers and suffixes are used to denote the presence of substituents at specific locations, chiral centres, etc.⁸⁶  The official rules for carotenoid nomenclature were approved by the IUPAC in 1974.⁸¹  Some examples of the semi-systematic nomenclature of common food carotenoids are shown in Table 1.7.

The main structural feature of carotenoids is their long system of c.d.b. This polyene backbone is the key contributor to their general properties (colour, reactivity, photochemical properties, shape, etc.), which in turn are ultimately key to understanding the diverse actions and applications of these versatile compounds.⁸²  The acyclic carotenes phytoene and phytofluene, which are the precursors of the other carotenoids, are rarities in the sense that they have much fewer c.d.b. (three and five, respectively; Figure 1.4) as compared to most carotenoids, as a result of which they are colourless and have distinctive properties and actions.⁹ 

The high number of c.d.b. characteristic of carotenoids endows the molecules with great rigidity, such that carotenoid molecules in their (all-E) configurations typically exhibit a rod-like shape. The corresponding Z isomers have an angular shape (Figure 1.15).⁸²  These differences in shape can have important consequences. For instance, Z isomers are thought to be less susceptible to aggregation in biological milieus, which in turn can have an impact on their solubility and bioavailability. On the other hand, it is reasonable to suggest that the ability of different geometrical isomers to fit into cellular structures or to interact with enzymes and other proteins may be markedly different.¹¹⁶ 

With very few exceptions, carotenoids are hydrophobic compounds. Therefore, the extraction of carotenoids requires solvents like acetone, hexane, diethyl ether, chloroform, etc. (see Chapter 8).¹¹⁷  Thus, in biological systems, carotenoids are usually found in lipidic milieus as membranes, mixed micelles (structures where lipid-soluble digested dietary components are incorporated for their uptake by enterocytes), lipid droplets or lipoproteins.^82,118–120  Carotenoids with carboxylic groups (e.g. bixin; Figure 1.16) can form water-soluble sodium or potassium salts.^121,122 

Coloured carotenoids absorb maximally in the range 400–500 nm, depending on the number and arrangement of the c.d.b. present in their chromophore. Their absorption spectra differ significantly from those of chlorophylls, and since both pigments form part of the photosynthetic apparatus, carotenoids contribute to the harvesting of light for this key process in nature.¹²³  The relationship between the chemical structure of carotenoids and the features of their spectra has long been known and is dealt with in great detail in classical texts¹²⁴  and in Chapter 10.

The UV–visible spectrum of a typical carotenoid usually contains three absorption maxima whose wavelengths depend on the number and arrangement of c.d.b. and the solvent used to obtain the spectra.^124,125  Regardless of the solvent, the values of λ_max increase with the number of c.d.b. (for more detail, see Chapter 10).

Absorbing mostly blue and violet light, carotenoids typically exhibit yellowish, orange or reddish colours.^124,126,127  At least seven c.d.b. are needed for a carotenoid to exhibit colour, although this attribute also depends on other factors. One of them is the concentration, but there are others, like the aggregation of carotenoid molecules or the association with other molecules like proteins.^82,128,129  Thus, a carotenoid-containing solution can vary from transparent to light yellow, yellow, orange, red and even beyond upon concentration, a phenomenon that can be commonly observed when working on the isolation and concentration of carotenoids. As already discussed in an earlier section, the aggregation of carotenoid molecules can be accompanied by red (bathochromic) or blue (hypsochromic) shifts, with consequent colour modifications.⁹⁴ 

Natural pigments in general and carotenoids in particular have enormous ecological importance as they are essential for communication between and within species. The colours of flowers or fruits are essential for pollination and seed dispersal, and therefore for propagation.^1,130  In animals, colour can convey key information of great value for species recognition, warning, mimicry, crypsis, sexual signalling and other processes. More specifically, the colour afforded by carotenoids in animals can inform about parasite load, nutritional and immunological states, fecundity, genetic quality and photoprotection.^131–133 

As far as foods are concerned, colour is one of the attributes that is related to acceptability by consumers. The effect of specific structural differences on the colours of some common food carotenoids within the parameters of the CIELAB uniform space¹³⁴  has been studied in detail.¹²⁸  The results revealed that the carotenoids clustered in the a*b* plane as a function of their c.d.b. number. As far as hue is concerned, the decrease in the conjugation of the molecules is accompanied by an increase in hue values and the aperture of the end ring leads to clearly decreased values.

The long polyene chain of carotenoids is rich in electrons; hence, it is important to explain the antioxidant or pro-oxidant activities of carotenoids.⁸²  The relationship of the chemical structure of carotenoids with the in vitro antioxidant and pro-oxidant properties of carotenoids has been the subject of many papers over the last 20 years,^135–138  in which different experimental conditions (oxidants, concentrations of carotenoids, solvents, etc.) have been used. There is also a wealth of revisions on the subject, concluding that, usually, carotenoids act as antioxidants, but also as pro-oxidants under certain conditions.^139–143 

In any case, it is well established that carotenoids act in plants by protecting them from photooxidation phenomena through different mechanisms, such as the prevention of the formation of singlet oxygen or its quenching. Without these and other essential functions of carotenoids in photosynthesis, life may not have developed on our planet.^144,145  In addition, carotenoids can protect humans from photooxidation in the skin¹⁴⁶  and probably in the eye.¹⁴⁷ 

Carotenoids can be either free or associated with other molecules, which undoubtedly can have an impact on some of their properties. Typical molecules that carotenoids can be associated with are sugars, proteins and, more frequently in common foods, FAs.

Some carotenoids can be associated with sugar moieties, such as glucose or gentiobiose, thus forming glucosides. As an example, crocetin is found glycosylated in saffron (Crocus sativus) stigmas and gardenia (Gardenia jasminoides) fruits.^148,149  The compounds resulting from the association of crocetin and sugar moieties are termed crocins (Figure 1.27), which are indeed glycosidic esters of crocetin, and can exist as (all-E) or Z isomers.¹⁰³ 

In addition, carotenoid glycosides can also be found in microorganisms (e.g. decaprenoxanthin diglucoside from Arthrobacter sp.¹⁵⁰  and astaxanthin glucoside from Agrobacterium aurantiacum).¹⁵¹  Furthermore, glucoside carotenoids acylated with FAs (e.g. thermocryptoxanthins and thermozeaxanthins) occur in the thermophilic eubacterium Thermus thermophilus.^79,152 

Some carotenoids can form complexes with diverse proteins (carotenoproteins) in several animals (see Chapter 6). The study of carotenoproteins has been particularly prolific in invertebrate animals, such as crustaceans, where many carotenoproteins, frequently containing astaxanthin and canthaxanthin, have been described in ovaries, eggs, exoskeleton and haemolymph.¹⁵³ 

The flesh of salmonid fish can exhibit pink hues as a result of the accumulation of carotenoid pigments like astaxanthin and canthaxanthin, acquired from their diets (for more detail, see Chapter 6). After absorption, these carotenoids reach the muscle, where they can bind to the actomyosin complex through non-specific hydrophobic bonds.¹¹⁸  Carotenoids are also associated with proteins in the mature feathers of many birds.¹⁵⁴ 

Carotenoproteins have also been described in microbes. A notorious example is the water-soluble modular cyanobacterial orange carotenoid protein, which has a role in the photoprotection of these microorganisms.¹⁵⁵ 

Apart from microbes and animals, carotenoproteins have also been reported in carrots.^156,157 

Carotenoid sulphate derivatives are mostly restricted to bacteria (e.g. caloxanthin 3-sulphate from the bacterium Erythrobacter longus)¹⁵⁸  and animals (e.g. bastaxanthin D from the sponge Ianthella basta and ophioxanthin from the starfish Ophioderma longicauda).^159–161  The structures of these derivatives are shown in Figure 1.27.

In general, either of the terms ‘carotenoid ester’ or ‘xanthophyll ester’ is used for referring to carotenoid acyl esters, although other less frequent ester forms can also be included under this terminology. Carotenoids can be found in nature either free or conjugated with other molecules to produce sulphate esters, glycosides, glycoside esters, glycosyl esters and acyl esters, among others.^83,162,163  Although carotenoid sulphate esters are restricted to few bacteria and animals, glycoside esters of carotenoids (e.g. thermocryptoxanthins and thermozeaxanthins from the thermophilic eubacterium T. thermophilus),^79,152  glycosyl esters of carotenoic acids (e.g. crocin from the stigmas of C. sativus)¹⁰³  and acyl esters (e.g. zeaxanthin dipalmitate from Physalis fruits, also known as physalien)^164,165  are more widely distributed in bacteria, algae, animals and plants. The structures of some of the carotenoids mentioned above are shown in Figures 1.27 and 1.28.

Undoubtedly, carotenoid acyl esters (xanthophyll esters) involving medium- and long-chain FAs are the most ubiquitous carotenoid-derived forms distributed among living organisms. Consequently, xanthophyll esters are frequently found in foods of plant and animal origin. Interestingly, peridinin, one of the most abundant carotenoids in nature, has an acetyl group in its molecule at position 3′ (Figure 1.12).⁸³  Some recent papers,^26,166,167  as well as Chapter 7, have compiled and critically reviewed the occurrence of xanthophyll esters in foods.

The acylation reaction (Figure 1.29) must necessarily be performed over a hydroxyl group; thus, only hydroxy-xanthophylls may derive into xanthophyll esters. Figure 1.8 shows the structures of various hydroxyl-xanthophylls, some of which are very commonly found esterified with FAs as part of the carotenoid profile in foods.

In general, hydroxy-xanthophylls have one or two hydroxyl groups that can be esterified with a range of FAs. Hydroxyl groups are usually located at carbon 3 of the end groups (β, ε, κ, etc.; e.g. zeaxanthin, lutein, capsanthin, violaxanthin and astaxanthin), although other positions, such as carbon 4, are also possible (e.g. isozeaxanthin; Figures 1.8–1.10).

Thus, for a particular xanthophyll, the diversity of derived xanthophyll esters will depend on the number of hydroxyl groups in the carotenoid molecule and the number of FAs involved (usually from four to six). Therefore, a monohydroxy-xanthophyll (such as β-cryptoxanthin and β-citraurin) will render only monoesters (e.g. β-cryptoxanthin laurate and β-citraurin myristate), while a dihydroxy-xanthophyll (such as zeaxanthin and lutein) will derive into monoesters or diesters when one or two hydroxy groups are acylated, respectively (Figure 1.28). Hydroxyl groups in the central chain of the carotenoid structure can also be esterified, as in the case of vaucheriaxanthin 3-acetate 19′-octanoate from the alga Nannochloropsis salina.⁸³  Consequently, the number of possible acylated forms is greater for dihydroxy-xanthophylls.

Depending on the nature of the acyl groups, diesters can be grouped into two classes, namely homodiesters when both acyl moieties are identical (e.g. zeaxanthin dipalmitate) and heterodiesters when the acyl moieties are different (e.g. zeaxanthin myristate-palmitate). This naming system was first used by Mellado-Ortega and Hornero-Méndez,¹⁶⁸  and was widely adopted after that,^26,167,169  although other similar systems have also been used (i.e. homogeneous diesters for homodiesters and heterogeneous or mixed diesters for heterodiesters).^166,170  Additionally, for the case of a dihydroxy-xanthophyll with an asymmetrical structure (e.g. lutein), the two hydroxyl groups are not equivalent, and subsequently two different regioisomers will be possible for each monoester and heterodiester. For instance, lutein 3-O-palmitate and lutein 3′-O-palmitate are the two possible regioisomers for lutein palmitate (monoester), having the acyl moieties in the hydroxyl group located at position 3 of the β-ring or at position 3′ of the ε-ring. Analogously, two regioisomers are possible for lutein myristate-palmitate, namely lutein 3′-O-myristate-3-O-palmitate and lutein 3′-O-palmitate-3-O-myristate. Figure 1.30 exemplifies the great diversity of different esterified forms that can be derived for a single xanthophyll (e.g. zeaxanthin for symmetrical structures and lutein for unsymmetrical structures) and three common SFAs (palmitic, myristic and stearic acids). Taking all of this together, it is easy to appreciate the great analytical complexity of the natural extracts containing xanthophyll esters. In fact, the presence of carotenoid esters has been overlooked in many studies, mainly due to the extensive use of saponification as a routine step in carotenoid analysis. Fortunately, the continuous improvement in the performance of modern analytical techniques, in particular high-performance liquid chromatography with UV–visible and mass spectrometry detectors, has facilitated recent developments in the analysis, identification and characterisation of the xanthophyll esters from natural sources (for more details, see Chapters 7–11).^{26,169,171,172} 

Obviously, the association of carotenoids with other molecules has an impact on the size and shape of the carotenoid and consequently on their interactions with other structures.

The association with sugars can lead to carotenoid solubilisation in water, as is the case of saffron crocins.^3,103  Carotenoproteins are also water soluble.^154,173,174 

The esterification of xanthophylls renders them more lipophilic and can have an impact on their biosynthesis and accumulation, as this increased lipophilicity could favour their sequestration by chromoplastic structures.¹⁷⁵  Free carotenoids (i.e. non-esterified) are located in the chloroplasts of green plant cells, together with chlorophylls, as well as in the chromoplasts of other plant tissues such as fruits, tubers and flower petals. However, it is in the chromoplasts where the carotenoids, both free and esterified with FAs, are massively synthesised and accumulated. The esterification of xanthophylls takes place during the ripening of most fruits and the senescence of leaves, coinciding with the transformation of the chloroplasts into chromoplasts.^176,177  Large amounts of xanthophyll esters are found in chromoplastic tissues, which constitutes direct evidence of the important role of esterification in the carotenoid accumulation capacity of plant cells.^178–181  Through the esterification mechanism, fruits and flowers enhance their external colour in order to increase the attraction of animals as pollinators and seed dispersion vectors.^182,183  At present, there is great interest in deciphering the biochemistry and genetics of the xanthophyll esterification process, including the identification and characterisation of the responsible gene(s) and enzyme(s) (see Chapters 4–6).^{26,168,180,184–188} 

From a nutritional point of view, it should be taken into consideration that an important proportion of the carotenoids present in our diet are in esterified form (see Chapter 7). The increased solubilisation and extractability of xanthophyll esters during food digestion in the presence of dietary fat has been shown to enhance carotenoid bioavailability;¹⁸⁹  however, this aspect has been ignored in most studies. Moreover, xanthophyll esters have also been shown to be more stable than free carotenoids.^189–198 

The association of carotenoids with proteins can extend the palette of carotenoid colours to grey, black, brown, green, blue or purple. Interestingly, thermal treatments such as those used to cook the meats of some of these animals cause the dissociation of the carotenoproteins and the appearance of the typical carotenoid hues.^153,154 

Carotenoids in the mature feathers of many birds are thought to be strongly bound to keratin, which is a structural, inert and insoluble protein. Interestingly, these associations can have a substantial impact on the colour provided by the same carotenoid both in different bird species and in different feathers of the same species. In other words, the same carotenoid can appear yellow, orange or red in feathers of different species or different feathers of the same species. Furthermore, as a result of these associations with keratin, the extraction of carotenoids from feathers is particularly difficult.¹⁵⁴  On the other hand, the acylation reaction does not modify the carotenoid chromophore; thus, their colour and UV–visible spectra remain quite similar to those of their free counterparts (see Chapter 10).

Carotenoproteins are thought to stabilise carotenoids.^154,173,174  On the other hand, it is thought that the esterification of carotenoids with FAs can modify the immediate environment and lead to modifications of their reactivity towards oxidising agents; such modifications are dependent on the type of FA bound to the xanthophyll.¹⁸⁹  In general, it is thought that esterification with FAs increases the stability of carotenoids.²⁵  The greater stability of esterified xanthophylls seems to be related to increases in their liposolubility compared to free xanthophylls, providing better integration into membrane structures and therefore reducing their susceptibility to adverse conditions in their environments.¹⁷⁶