Visual cycle-derived bisretinoids as endogenous natural products: enzyme-free formation, photochemical reactivity, and retinal degeneration

Abstract

Covering: up to 2026

Bisretinoids are a chemically distinct class of endogenous natural products formed by the non-enzymatic condensation of visual-cycle retinoids. Derived from dietary provitamin A carotenoids via retinaldehyde intermediates, these pigments form spontaneously within the photoreceptor disc membranes through Schiff base chemistry with phosphatidylethanolamine, generating structurally diverse pyridinium, dihydropyridine and retinal dimer species. In contrast to enzyme-directed biosynthesis, bisretinoid biogenesis is governed by the intrinsic electrophilicity of the conjugated retinaldehydes within a lipid-dense environment. Their extended polyene systems endow them with distinctive excited-state properties, enabling efficient intersystem crossing and photosensitized generation of singlet oxygen under visible light. Subsequent oxidative fragmentation produces reactive electrophilic carbonyl species, including methylglyoxal and glyoxal, which covalently modify biomolecules and contribute to retinal pigment epithelium dysfunction and drusen formation. Despite their well-documented pathological roles, bisretinoids have not been systematically examined within a natural product framework. Here, we integrate the current knowledge of their biogenesis, electronic structure, and photochemical reactivity and consider how factors such as retinaldehyde flux, membrane composition, and iron homeostasis modulate their accumulation and reactivity. By framing bisretinoids as autochthonous natural products governed by intrinsic chemical principles, this review highlights new opportunities for mechanistically informed therapeutic intervention in retinal degeneration.

This review reframes retinal bisretinoids as autochthonous natural products, demonstrating how polyene electrophilicity and non-enzymatic aza-electrocyclization dictate their inevitable biogenesis and subsequent photochemical toxicity.

1. Introduction

Carotenoids are a major class of dietary natural products that serve as the primary source of vitamin A (retinoids) in humans and other vertebrates.^1–4 As C40 terpenoid pigments biosynthesized via the methylerythritol phosphate (MEP) pathway in plants, they embody a conserved polyene architecture whose electronic and chemical properties are preserved across biological systems.^5–7 In vertebrates, provitamin A carotenoids, such as β-carotene, undergo central oxidative cleavage by β-carotene 15,15′-dioxygenase (BCO1) to yield retinaldehyde, while xanthophylls, including lutein and zeaxanthin, are asymmetrically processed by β-carotene 9′,10′-oxygenase (BCO2).^1–3,8,9 Through these transformations, plant-derived terpenoid scaffolds are converted into reactive retinoid intermediates that sustain the visual cycle, establishing a direct chemical lineage from dietary carotenoids to retinal chromophores and their downstream reactivity (Fig. 1).^4,9,10

Fig. 1. Biosynthetic origin of the macular xanthophylls and visual cycle-derived bisretinoids. Dietary carotenoids enter the vertebrate metabolism through two principal routes. Provitamin A carotenoids (left) undergo central oxidative cleavage by beta-carotene 15,15-prime-dioxygenase (BCO1) to yield retinal, which is subsequently reduced and esterified to form the circulating retinoids that supply the vitamin A-dependent visual cycle. Non-provitamin A xanthophylls (right) are transported selectively to the retina as intact pigments. In humans, BCO2 (beta-carotene oxygenase 2) expression is suppressed in the retina, enabling intact xanthophyll accumulation; in rodents, active BCO2 expression prevents this accumulation. Among the three principal macular xanthophylls, lutein (3R, 3′R, 6′R) and zeaxanthin (3R, 3′R) are delivered directly from the circulation, while meso-zeaxanthin (3R, 3′S) is generated locally within the RPE through enzymatic isomerization of lutein (RPE, in situ); the reverse reaction has not been experimentally demonstrated. The carbon numbering on the lutein and meso-zeaxanthin structures indicates the positions relevant to this stereochemical transformation: the double-bond shift from the C4′–C5′ to the C5′–C6′ position converts the ε-ionone ring of lutein to a β-ionone ring, altering the C-3′ configuration from R to S. Within the photoreceptor disc membrane, retinaldehydes generated by the visual cycle condense non-enzymatically with the ethanolamine headgroup of phosphatidylethanolamine (PE) to form N-retinylidene-PE (NRPE) via a reversible Schiff-base equilibrium. A second retinaldehyde molecule then participates in subsequent non-enzymatic reactions, leading to A2-PE. R denotes the shared polyene backbone shown in the inset.

Within the macula, the xanthophylls lutein, zeaxanthin and meso-zeaxanthin are selectively accumulated to form macular pigment, a concentrated reservoir of dietary natural products that filter high-energy blue light and quenches reactive oxygen species (Fig. 1, upper right, xanthophyll structures).^11–17 The abundance of these carotenoids is regulated by tissue-specific transport processes, systemic carotenoid metabolism, and long-term dietary exposure, reflecting a balance between selective uptake, local retention, and metabolic turnover.^{4,11,12,18–23}

Concurrently, the vitamin A-dependent visual cycle continuously regenerates the 11-cis-retinal chromophore required for phototransduction.²⁴ Within a tightly regulated circuit, spanning photoreceptors and the retinal pigment epithelium (RPE), the photoactivation of rhodopsin generates all-trans-retinal as a reactive aldehyde that is transiently retained within the photoreceptor disk membranes.^24–27 These membranes provide a lipid-dense, phosphatidylethanolamine (PE)-rich environment that favors spontaneous Schiff-base formation and non-enzymatic condensation. Upon illumination, accumulated bisretinoids within the disk membrane act as photosensitizers, promoting the conversion of molecular oxygen to singlet oxygen (¹O₂) and thereby initiating photooxidative reactions.^28,29

Visual cycle-derived retinaldehydes can undergo non-enzymatic condensation with PE to generate a structurally diverse family of bisretinoids (Fig. 1, lower panel, non-enzymatic condensation pathway).^28–32 The prototypical member, A2E, was identified as a major fluorophore of RPE lipofuscin, and arises from the coupling of two retinal molecules with an ethanolamine moiety via the intermediate A2-phosphatidylethanolamine (A2-PE).^28,30,31 Subsequent structural and biosynthetic studies have established that A2E represents only one member within a broader chemical landscape, encompassing geometric isomers, partially reduced dihydropyridine species, and retinal dimer-derived congeners.^29,33–36

Bisretinoid-containing lipofuscin accumulates progressively within RPE lysosomes with age, a process markedly accelerated in genetic contexts that increase bisretinoid formation, including loss-of-function mutations in the ABC transporter ABCA4.^29,37,38 Consistent with this chemical burden, Abca4-deficient or disease-associated models exhibit elevated bisretinoid levels, enhanced fundus autofluorescence, and increased susceptibility to light-induced damage.^37–40 Clinical imaging further associates abnormal bisretinoid lipofuscin distribution with the development of geographic atrophy and related macular dystrophies, whereas epidemiologic and supplementation data link higher macular pigment optical density with reduced risk or slower progression of AMD (age-related macular degeneration).^{11,12,39,41,42}

From a natural products chemistry perspective, bisretinoids reside at the interface of metabolism and spontaneous organic reactivity. Unlike canonical natural products assembled by enzyme-directed cascades, they arise from the intrinsic reactivity of conjugated retinaldehydes and membrane-associated primary amines within the low-water, lipid-dense environment of photoreceptor disc membranes.^28,29,33,43 In this context, bisretinoids are best viewed as autochthonous natural products: structurally small molecules generated in situ from a dietary vitamin A precursor through the interplay of physiological flux and non-enzymatic reactions.^{29–31,36,44} Their extended conjugation and amphiphilic, often cationic, architecture confer visible light absorption, susceptibility to photooxidation, and a propensity for the covalent modification of biomolecules.^28,45–47

Conceptually, bisretinoids of RPE lipofuscin can be considered alongside other endogenous pigmentary or Maillard-type natural products, including advanced glycation end-products, porphyrins, and neuromelanin, which arise from normal metabolism but acquire pathogenic significance upon overproduction or impaired clearance.^29,48 Upon light-driven oxidation, bisretinoids generate reactive electrophilic species that modify RPE proteins and lipids, disrupt membrane organization, and initiate downstream stress responses in RPE cells.^42,45–57 However, in the retina, this chemistry is uniquely bifurcated: dietary carotenoids give rise to both photoprotective macular pigments and essential vitamin A derivatives that generate potentially toxic bisretinoid species, with the balance between these fates governed by carotenoid cleavage, retinoid flux, membrane composition, and environmental light exposure.^4,12,28,31

Several authoritative reviews have addressed macular carotenoid biology, bisretinoid chemistry, and RPE lipofuscin in isolation.^{11,12,29,44,49} However, the literature has focused predominantly on descriptive photophysics, genetic models, and clinical correlations, while paying comparatively limited attention to the natural-product lineage of bisretinoids, the chemical logic underlying their formation as phospholipid-retinal conjugates, and the origins of their structural diversity. Recent advances, including biomimetic and total synthesis, revision of previously misassigned structures, and emerging connections between bisretinoid chemistry and regulated cell death, highlight the need for an integrated chemical framework that connects formation, structure, and function.^50–53

In this review, we frame visual cycle-derived bisretinoids as endogenous natural products arising from dietary carotenoids and retinoids and analyze their biogenesis, structural diversification, chemical reactivity, and retinal toxicity through the lens of contemporary natural products chemistry (Fig. 1). By emphasizing non-enzymatic polyene-aldehyde chemistry, biomimetic parallels between in vivo formation and synthetic routes, and the chemical basis of retinal toxicity, we present a mechanistically grounded perspective that complements existing biological narratives and defines new opportunities for therapeutic intervention targeting retinaldehyde flux, chemical persistence, and bisretinoid burden.

2. Carotenoids as the natural product origin of retinoids

2.1. Carotenoids as terpenoid natural products: biosynthesis and structural logic

Building on the chemical lineage outlined above, carotenoids can be understood as terpenoid natural products whose biosynthesis and structure preconfigure their downstream conversion to retinoids. Carotenoids comprise one of the most structurally diverse families of terpenoid natural products, originating from the universal C5 building blocks isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP).^5–7 In photosynthetic organisms, these precursors are supplied predominantly by the plastidial 2-C-methyl-d-erythritol-4-phosphate (MEP) pathway, which furnishes IPP and DMAPP for the biosynthesis of carotenoids and other plastidial isoprenoids, whereas the cytosolic mevalonate (MVA) pathway supplies sterols and other cytosolic isoprenoids.^5–7 This compartmentalization establishes a biosynthetic logic in which extended conjugated polyenes are produced as specialized natural products with intrinsic photophysical and chemical reactivity.^5–7

Chain extension of IPP and DMAPP yields geranylgeranyl diphosphate (GGPP), a C20 allylic precursor to phytoene.⁵⁴ Phytoene synthase catalyzes the dimerization of two GGPP molecules via a presqualene diphosphate-type intermediate to form phytoene, the first committed C40 intermediate that defines the carotenoid scaffold.^5,54 Phytoene is essentially colorless owing to its limited conjugation, but sequential desaturation reactions progressively extend the conjugated π-system, while geometric isomerization converges on all-trans lycopene, a highly conjugated polyene that serves as a central branching point for carotenoid diversification.^5,54

From a natural products chemistry standpoint, the defining feature of carotenoids is their extended, delocalized π-electron system, which shifts absorption into the visible range and enables efficient excited-state energy dissipation.^55–58 Critically, this same polyene architecture renders carotenoids inherently susceptible to electrophilic addition, radical-mediated hydrogen abstraction, and oxygen-dependent oxidative cleavage; thus, photoprotection and oxidative degradation emerge as complementary consequences of a shared structural design.^55–59

All-trans-lycopene serves as the principal precursor to cyclic carotenes and xanthophylls through the action of lycopene cyclases that install β-ionone and/or ε-ionone end groups.^5,54 The identity and functionalization state of these terminal rings divide carotenoids into two broad classes: carotenes, which are hydrocarbon polyenes such as β-carotene, and xanthophylls, which carry oxygenated substituents such as hydroxyl, epoxy, or keto groups, as exemplified by lutein, zeaxanthin, and astaxanthin.^5,54 This end-group chemistry governs polarity, membrane partitioning, and susceptibility to oxidative transformation, thereby encoding functional specialization directly into scaffold design.⁵⁹

In animals, carotenoid end-group structures dictate metabolic fate by controlling compatibility with carotenoid oxygenases. Provitamin A carotenoids that retain at least one unsubstituted β-ionone ring undergo central oxidative cleavage by BCO1 to yield retinaldehyde, whereas many xanthophylls are preferentially processed by BCO2 or evade productive cleavage and persist as intact pigments in tissues.^1–4,8–10 This substrate selectivity reflects a structural rule: modifications at the ring terminus, including hydroxylation, epoxidation, and ketolation, alter steric and electronic complementarity to the enzyme active site and thereby redirect the chemical fate of the polyene.^1–4,8,9

Non-enzymatic oxidative scission of the carotenoid polyene backbone under conditions of oxidative stress generates shorter apocarotenoid aldehydes and ketones, distinct from the regioselective enzymatic cleavage carried out by BCO1 and BCO2.^54,60,61 This apocarotenoid chemistry establishes a chemical precedent for retinoid reactivity: retinal and related retinoids can be viewed as chemically preorganized terpenoid fragments whose aldehyde functionality confers privileged reactivity in downstream condensation and redox processes.^1–4,9

Trophic transfer illustrates that carotenoid scaffolds function as portable natural products, synthesized in photosynthetic or microbial organisms and subsequently redeployed in higher organisms. For example, astaxanthin accumulation in marine animals is derived primarily from dietary acquisition from microalgae and, to a lesser extent, certain yeasts rather than de novo biosynthesis, underscoring the evolutionary persistence and cross-kingdom continuity of carotenoid scaffold chemistry.^62–64

The terpenoid origin, polyene electronic structure, and oxygenase-governed cleavage logic of carotenoids define a continuous chemical lineage from plant-derived C40 pigments to vertebrate retinoids. This lineage furnishes the retinaldehyde substrates that enter non-enzymatic bisretinoid-forming chemistry within photoreceptor disk membranes, as discussed in the following section.^{1–4,28–31,36}

2.2. Structural continuity and tissue-specific biotransformation of dietary carotenoids in vertebrates

As outlined in Section 2.1, carotenoids enter animal biology as structurally preorganized terpenoid natural products whose polyene architecture and end-group chemistry determine their downstream fate. Therefore, in vertebrates, dietary carotenoids are selectively preserved, cleaved, or remodeled according to their structural identity, rather than treated as generic antioxidants.

Following dietary intake, carotenoids are released from the food matrix and incorporated into mixed micelles, after which their uptake by enterocytes is influenced by physicochemical properties such as polarity and micellar solubility.^18,20–22 Xanthophylls such as lutein and zeaxanthin are preferentially solubilized owing to their hydroxylated end groups, whereas more hydrophobic carotenes exhibit greater dependence on lipid co-ingestion.^20,21 After incorporation into chylomicrons, carotenoids are redistributed among circulating lipoproteins, including VLDL, LDL, and HDL, which mediate their tissue-selective delivery.^47,65

Systemic distribution is further shaped by carotenoid oxygenases that partition polyene scaffolds into chemically distinct outcomes. BCO1 centrally cleaves provitamin A carotenoids bearing an unsubstituted β-ionone ring to yield retinaldehyde as the universal C20 retinoid precursor.^1,3,9 BCO2 performs asymmetric cleavage of a broader range of carotenoids, particularly xanthophylls, limiting excessive mitochondrial accumulation and redirecting substrates into shorter apocarotenoids.^3,8 Together, these enzymes divide dietary carotenoids into retinoid precursors, apocarotenoid fragments, and intact xanthophylls preserved for tissue delivery.

Despite the structural diversity of circulating carotenoids, the human retina selectively accumulates three principal xanthophylls, lutein, zeaxanthin, and meso-zeaxanthin, reaching concentrations that greatly exceed those present in plasma (Fig. 1, upper right, xanthophyll structures).^15,44 Notably, meso-zeaxanthin is essentially absent from circulation, implying local ocular generation rather than direct dietary delivery.^15,44 Current evidence indicates that meso-zeaxanthin arises primarily from lutein through enzymatic isomerization within the RPE, a transformation involving a double-bond shift from the 4′–5′ to the 5′–6′ position of the epsilon-ionone ring that converts the (3R, 3′R, 6′R) configuration of lutein to the (3R, 3′S) meso form while leaving the extended polyene backbone intact (Fig. 1, upper right, xanthophyll structures).¹⁵ Although the enzymatic machinery responsible for this conversion remains incompletely characterized, the reaction represents a clear example of localized structural editing, in which a dietary natural product is remodeled in situ to generate a functionally distinct metabolite rather than being simply absorbed or degraded. Retinal sequestration and spatial patterning of these pigments are reinforced by intracellular binding proteins that function as molecular sinks: glutathione S-transferase Pi 1 (GSTP1) binds zeaxanthin and meso-zeaxanthin, while StAR-related lipid transfer domain protein 3 (StARD3) preferentially binds lutein through its START domain, which accommodates structural differences at the carotenoid end group.^36,39 These interactions stabilize intact carotenoid scaffolds, prevent uncontrolled diffusion, and establish steep concentration gradients across retinal layers.

This retina-specific handling of carotenoids exemplifies a broader principle of animal secondary metabolism, in which dietary natural products undergo localized structural transformation to generate functionally specialized metabolites within a defined tissue microenvironment. The contrast between the xanthophyll and provitamin A pathways is particularly instructive in this regard. While lutein is structurally edited and spatially organized as an intact pigment, provitamin A carotenoids such as β-carotene are irreversibly converted by BCO1 into retinaldehyde, a reactive polyene aldehyde whose subsequent chemistry is governed primarily by physicochemical environment rather than enzymatic control. This retinaldehyde pool constitutes the immediate chemical precursor for the non-enzymatic condensation reactions leading to bisretinoid formation, as discussed in Section 3.

3. Formation of visual cycle-derived bisretinoids: non-enzymatic polyene-aldehyde chemistry

Bisretinoids are a structurally diverse class of endogenous natural products formed non-enzymatically from visual cycle-derived retinaldehydes within photoreceptor disc membranes. In contrast to terpene or polyketide biosynthesis, which proceeds through enzyme-mediated carbon–carbon bond formation, bisretinoid assembly arises from spontaneous reactions between electrophilic polyene aldehydes and primary amines in a lipid-dense membrane environment. This chemistry integrates Schiff-base formation, conjugated polyene reactivity, and protonation-dependent electrophilicity, giving rise to an ensemble of fluorophores that accumulate in the RPE as bisretinoid lipofuscin. The chemical basis of this reactivity is rooted in the quantum-mechanical properties of extended polyene systems (Section 3.5) and its chemical inevitability is underscored by the biomimetic synthesis of A2E from retinal and ethanolamine, demonstrating that the bisretinoid scaffold can arise solely from intrinsic polyene-aldehyde chemistry in the absence of enzymatic catalysis.⁶⁶

3.1. Reactive aldehyde chemistry in the outer retina: chemical preconditions for bisretinoid formation

The outer retina constitutes a chemically privileged microenvironment in which visual cycle-derived retinaldehydes encounter a dense phospholipid matrix that favors non-enzymatic condensation reactions. Retinal isomers, including all-trans- and 11-cis-retinal, contain an electrophilic C15 aldehyde conjugated with a polyene chain, rendering them susceptible to nucleophilic addition by primary amines.⁶⁷ Among membrane-associated nucleophiles, the ethanolamine headgroup of phosphatidylethanolamine (PE) is highly abundant in photoreceptor disc membranes, positioning it as the dominant reaction partner for retinaldehyde condensation. This reaction requires the ethanolamine nitrogen in its unprotonated free amine form (–NH₂); the protonated ammonium species (–NH₃⁺) lacks the nucleophilicity required for Schiff-base formation with retinaldehyde.^28,30,68

Model-membrane studies indicate that photoreceptor disc membranes exhibit low polarity, limited water accessibility, and tight lipid packing, conditions that enhance aldehyde electrophilicity and stabilize imine formation.^69,70 Under these constraints, retinal isomers bearing an intact C15 aldehyde readily form Schiff-base adducts with PE, establishing N-retinylidene-PE (NRPE) as the thermodynamically and kinetically favored initial intermediate in bisretinoid biogenesis.^28,31,36 This condensation arises from intrinsic carbonyl reactivity within the physicochemical landscape of disc membranes, without requiring enzymatic catalysis. NRPE is also the ligand recognized by ABCA4 (ATP-binding cassette subfamily A member 4), a membrane transporter that assists in the detoxification of retinaldehyde (in the form of NRPE).⁷¹ Mutations in the ABCA4 gene are responsible for recessive Stargardt disease (STGD1).⁷²

Both all-trans- and 11-cis-retinal readily undergo condensation with ethanolamine or the ethanolamine headgroup of PE to form bisretinoid species, as demonstrated by in vitro studies and their accumulation in dark-reared mice in which 11-cis-retinal is the predominant chromophore.^19,73 The relative contributions of these retinaldehyde isomers under physiological conditions still need to be established. Nevertheless, these observations indicate that bisretinoid formation is governed primarily by retinaldehyde availability rather than by isomeric geometry.

The essential chemical conditions for bisretinoid biosynthesis are defined by retinaldehyde electrophilicity, PE nucleophilicity, and a membrane environment that promotes imine formation. Subsequent steps, including NRPE protonation, addition of a second aldehyde to form A2PE, hydrolysis to A2E, and further diversification into a range of bisretinoids, follow directly from this initial reactivity and are discussed in the following sections (Fig. 2).

Fig. 2. Non-enzymatic reaction network of the bisretinoid formation from all-trans-retinal and phosphatidylethanolamine. Structures of the key substrates are shown at the top: all-trans-retinal, phosphatidylethanolamine (PE; P denotes the phospholipid headgroup), and the first condensation product N-retinylidene-PE (NRPE). The central intermediate tautomer x arises from NRPE via a [1,5] H-shift and serves as the branching point for two divergent pathways. Path a: nucleophilic addition of a second retinal molecule to tautomer x, followed by aza-6π electrocyclization and autooxidation yields dihydropyridinium-PE, which loses 2H⁺ to form A2-PE and lysoA2PE, or loses 1H⁺ to form A2-DHP-PE. Phosphate hydrolysis of A2-PE releases A2E; photoisomerization of A2E (hν) yields iso-A2E. Glycerophosphoethanolamine headgroup substitution of A2-PE produces A2GPE. Path b: [1,4] addition of tautomer x to a second retinal molecule generates a precursor of the all-trans-retinal dimer through a Mannich-type ring closure and elimination; subsequent reaction with PE yields all-trans-retinal-dimer-PE. The tautomer x label indicates this intermediate as the chemically activated species whose electronic properties govern pathway selectivity.

3.2. Formation, protonation, and chemical behavior of NRPE

NRPE is the first discrete molecular intermediate in bisretinoid biosynthesis, which is formed through the intrinsic reactivity of retinaldehyde with the ethanolamine headgroup of PE. Early structural studies of A2E and related bisretinoids demonstrated that all-trans-retinal readily condenses with primary amines, establishing aldehyde-amine chemistry as the basis of NRPE formation in photoreceptor membranes.^43,67 The identification of A2-PE in photoreceptor outer segments provided in vivo evidence for retinal-PE Schiff-base intermediates, supporting NRPE as the central precursor to all downstream bisretinoid species.^31,37

Mechanistically, NRPE formation proceeds via classical Schiff-base chemistry: nucleophilic addition of the ethanolamine nitrogen to the C15 aldehyde yields a short-lived carbinolamine intermediate, which dehydrates to form the imine. Although these intermediates have not been isolated directly from retinal tissue, their existence is strongly supported by synthetic bisretinoid studies and by the behavior of retinal-ethanolamine adducts in solution.^28,43 The resulting imine can be protonated to form an iminium species with enhanced electrophilicity and extended conjugation, facilitating subsequent reactions with additional retinal molecules.^28,43

The high PE content of photoreceptor disc membranes, together with their enrichment in polyunsaturated fatty acids, creates a low-polarity low-water-activity environment that favors imine formation and stabilization. Under these conditions, the equilibrium is shifted toward dehydration, suppressing hydrolytic reversal and stabilizing NRPE within the membrane bilayer.^68,74 Although direct measurements of water activity or imine hydrolysis kinetics in native discs are lacking, observations in model membranes are consistent with this mechanism.^53,75–77

Redistribution of NRPE across membrane leaflets by ATP-binding cassette transporter A4 (ABCA4) modulates its accessibility to hydrolytic processes (retinol dehydrogenases) and thus the balance between NRPE retention and conversion to A2-PE.⁷¹ ABCA4 does not participate in NRPE formation, which arises from intrinsic aldehyde-amine reactivity rather than enzyme catalysis. Therefore, NRPE formation represents the first chemically driven step in bisretinoid biosynthesis, occurring whenever reactive retinaldehydes encounter ethanolamine-containing lipids within a suitably hydrophobic membrane environment.

3.3. Biomimetic synthesis and mechanistic correspondence with in vivo bisretinoid formation

The chemical logic underlying bisretinoid formation in vivo closely parallels reaction pathways familiar form synthetic organic chemistry. The non-enzymatic assembly of A2E and related bisretinoids within photoreceptor membranes mirrors principles demonstrated in biomimetic total synthesis, supporting the view that bisretinoid biogenesis proceeds through intrinsically favored polyene-aldehyde chemistry rather than through specialized enzymatic catalysis.

The first total synthesis of A2E, reported in 1997, demonstrated that condensation of retinal-derived aldehydes with primary amines under mildly acidic conditions is sufficient to generate the pyridinium bisretinoid scaffold.⁶⁶ In this route, formation of a retinal Schiff base followed by protonation to an iminium species activates the conjugated polyene toward intramolecular cyclization, yielding a dihydropyridinium intermediate that undergoes spontaneous oxidation to A2E.^43,66 These transformations proceed without enzyme catalysis and instead reflect intrinsic features of retinal-derived polyenes, including proton-assisted iminium activation, extended π-conjugation, and pericyclic electrocyclization.

These same reaction elements are recapitulated during bisretinoid formation in vivo. In photoreceptor disc membranes, retinaldehyde condenses non-enzymatically with the ethanolamine headgroup of PE to form N-retinylidene-PE (NRPE), the initial imine intermediate.^28,31 Subsequent protonation of NRPE within the low-dielectric, membrane-embedded environment increases iminium character and activates the conjugated polyene toward nucleophilic addition by a second retinal molecule, yielding A2-PE as the immediate biosynthetic precursor to A2E (Fig. 2, path a).^29,31,78

Both synthetic and biological pathways converge on a common mechanistic step: formation of a di-retinal iminium intermediate that undergoes aza-6π electrocyclization to generate a dihydropyridinium ring system.^28,71,73In vitro studies demonstrated that these dihydropyridinium intermediates readily aromatize via autooxidation to yield the fully conjugated pyridinium core of A2E, a transformation that also occurs spontaneously under physiological oxygen tension.³³ This correspondence underscores that the in vivo pathway exploits the same orbital symmetry and charge-delocalization principles that govern biomimetic polyene cyclization in solution.

Enzymatic involvement in this system is confined to upstream processes that regulate retinal availability rather than bisretinoid assembly itself. ABCA4 modulates NRPE distribution and lifetime by facilitating its translocation across disc membranes, thereby influencing its access to hydrolytic (retinol dehydrogenases) or condensation pathways.^37,71 However, ABCA4 does not participate directly in C–C or C–N bond formation, indicating that bisretinoid construction assembly is fundamentally chemically driven, rather than enzyme-templated.

3.4. Structural diversification from a common NRPE-derived reactive intermediate

As described in Section 3.3, A2E formation in vivo proceeds through intrinsically favored polyene-iminium chemistry. This process does not yield a single defined product; rather, the initial NRPE-derived iminium or enamine intermediate constitutes a chemically promiscuous reactive manifold from which multiple bisretinoid scaffolds emerge via competing, non-enzymatic pathways (Fig. 2).

The best-characterized outcome of this chemistry is A2E, generated by the addition of a second-retinal molecule to protonated NRPE, followed by aza-6π electrocyclization and oxidative aromatization.^28–31 Variations in retinal isomer composition, protonation state, and reaction sequence can divert this common intermediate toward alternative products. One such product is iso-A2E, a geometric isomer that retains the pyridinium core but differs in polyene double-bond configuration, highlighting that product distribution is governed by reaction conditions and retinal isomerism rather than enzymatic control (Fig. 2, path a, A2E/iso-A2E equilibrium).^30,36

A second major branch yields A2-dihydropyridine phosphatidylethanolamine (A2-DHP-PE), a reduced bisretinoid species in which the dihydropyridinium ring does not undergo full aromatization.³⁴ Structural characterization of A2-DHP-PE from retinal tissue demonstrated that incomplete oxidation of the cyclized intermediate is a physiologically relevant outcome, reinforcing the role of oxygen availability and redox microenvironment in modulating product distribution at late stages of bisretinoid assembly without the need to invoke alternative biosynthetic machinery (Fig. 2, path a, upper right, A2-DHP-PE branch).³⁴

In parallel, tautomer x, the NRPE-derived intermediate that feeds aza-electrocyclization toward A2-PE, can engage in an alternative reaction pathway. Nucleophilic attack at C20 of this intermediate on C13 of free all-trans-retinal proceeds via a Michael-type addition, followed by Mannich-type ring closure and elimination of the phosphoethanolamine headgroup to yield an all-trans-retinal dimer (atRALdi).^52,79 Isotopic labeling and stereochemical analyses show that atRALdi formation shares early intermediates with A2E biosynthesis, diverging at the stage of second-retinal engagement.^35,79 Ethanolamine-linked forms of this dimer (atRALdi-E) are also detected in retinal tissue, confirming that this alternative coupling pathway operates under physiological conditions (Fig. 2, path b, left panel).⁸⁰

Further structural diversification arises through variation in the phospholipid backbone. In addition to diacyl A2-PE, substitution of the phosphoethanolamine headgroup with glycerophosphoethanolamine yields A2-GPE, demonstrating that changes in phospholipid composition are directly translated into bisretinoid structural diversity.⁸¹ Moreover, a substantial fraction of lyso-A2-PE exists as alkyl ether-linked glycerophosphoethanolamine adducts rather than monoacyl phospholipids.⁸² Ether-linked phospholipids exhibit increased resistance to hydrolysis and oxidation, altered membrane partitioning, and prolonged intracellular persistence relative to their diacyl species, indicating that bisretinoid formation intersects selectively with ether lipid pools in photoreceptor membranes (Fig. 3).

Fig. 3. Structural classification of the bisretinoid natural products of the retinal pigment epithelium. Bisretinoids are organized into three structural classes. Pyridinium bisretinoids share a fully aromatic pyridinium core (bold ring) and extended conjugated polyene arms. A2E and iso-A2E are geometric isomers (diastereomers), differing in the configuration of the C13′ C14′ double bond on the longer polyenyl arm. i-A2E, enclosed in a dashed box, was originally reported as iiso-A2E; total synthesis by Vidal et al.⁵² confirmed a positional isomer of A2E, bearing the pentaenyl chain at C3 rather than C2 of the pyridinium ring. Phospholipid-linked bisretinoids remain attached to the glycerophospholipid backbone. In A2PE (diacyl A2PE), two acyl chains (R1 and R2, highlighted in blue) are esterified at the glycerol. In lysoA2PE (alkyl ether-linked), the sn-1 position bears an alkyl ether linkage (O–R, highlighted in green) rather than an acyl ester; this ether linkage confers increased resistance to hydrolysis and alters membrane partitioning relative to the diacyl species. A2GPE arises from glycerophosphoethanolamine headgroup substitution. A2-DHP-PE contains a partially reduced dihydropyridine core (shown in grey) rather than the fully aromatic pyridinium ring, reflecting incomplete oxidation during biosynthesis. Retinal dimer-derived bisretinoids (all-trans-retinal-dimer-PE, all-trans-retinal-dimer-E, and all-trans-retinal dimer) share a cyclohexadiene ring system (purple) formed through Michael-type addition and Mannich-type cyclization rather than aza-electrocyclization.

Together, these species, including A2E, iso-A2E, A2-DHP-PE, atRALdi, and related lipid-conjugated congeners, define a structural landscape that originates from a common chemically activated precursor rather than multiple enzyme-specific biosynthetic pathways. The diversity reflects the intrinsic reactivity of conjugated polyene iminium and enamine intermediates operating within a membrane-restricted, mildly acidic, and oxygenated environment. Analogous to polyketide and terpene pathways that generate families of related metabolites through scaffold diversification, bisretinoid formation yields a spectrum of endogenous pigments via non-enzymatic branching chemistry. In this context, the retina functions as a reaction chamber in which a single retinal-derived intermediate gives rise to multiple chemically distinct natural products that collectively shape lipofuscin composition in retina and downstream reactivity.

3.5. Quantum-mechanical basis of bisretinoid reactivity

The efficient assembly of retinal-derived aldehydes into bisretinoid scaffolds arises from the quantum-mechanical properties of conjugated polyenes (Fig. 4). The electronic architecture of retinal and its Schiff-base intermediates promotes iminium formation, charge delocalization, and pericyclic reactivity, making bisretinoid formation intrinsically favored under physiological conditions.

Fig. 4. Electronic architecture of A2E and conceptual excited-state deactivation scheme. Left: structure of A2E with the three electronic features governing its excited-state behavior: extended pi-conjugation across the retinoid polyene backbone (orange), a permanent pyridinium cation at the heterocyclic core, and an amphiphilic architecture. Quantum-chemical calculations on retinal protonated Schiff base models demonstrate that protonation delocalizes the positive charge across the polyene chain, yielding non-integer C–C bond orders and a LUMO that extends along the full conjugated system rather than localizing at the iminium nitrogen.^114,115 This LUMO delocalization lowers the activation barrier for nucleophilic addition at positions distal from the iminium and provides the orbital continuity required for aza-6π electrocyclization during bisretinoid biosynthesis. In A2E, the permanent pyridinium cation sustains this delocalization without requiring protonation-dependent ionization, predisposing the molecule to efficient intersystem crossing and singlet oxygen sensitization. Right: Jablonski-type diagram depicting the deactivation pathways from the first singlet excited state (S₁) following visible light absorption (blue arrow). Intersystem crossing (ISC; green bold arrow) to the lowest triplet state (T₁) is the dominant pathway, enabling energy transfer to ground-state molecular oxygen and generation of singlet oxygen (¹O₂) through triplet sensitization. Fluorescence (grey dashed arrow) and internal conversion (IC; orange dashed arrow) represent the minor competing deactivation pathways from S₁ to S₀; IC is non-radiative. Phosphorescence (pink dashed arrow) is indicated as inferred, as direct measurement of bisretinoid phosphorescence has not been reported.

Retinaldehyde is distinguished from most biological aldehydes by its extended pentaenal system, which enables extensive π-electron delocalization. Quantum-chemical calculations have demonstrated that, upon imine formation with a primary amine, protonation generates a retinal Schiff-base iminium whose lowest unoccupied molecular orbital (LUMO) is delocalized along the polyene chain rather than localized at the iminium carbon.^83,84 This delocalization lowers the activation barrier for nucleophilic addition and enables electrophilic reactivity at positions distal from the original aldehyde.

LUMO delocalization has two key consequences for bisretinoid biogenesis. First, it stabilizes cationic intermediates formed during NRPE protonation, enabling iminium species to persist long enough for bimolecular reactions with a second retinal molecule.^28,73 Second, it provides the orbital continuity required for aza-6π electrocyclization, a symmetry-allowed pericyclic process that converts linear di-retinal iminium intermediates into cyclic dihydropyridinium scaffolds.^28,73,84 This transformation, central to A2-PE and A2E formation, arises directly from polyene orbital topology rather enzyme-specific catalysis.

Computational studies of retinal and related retinoid chromophores further reveal that protonation reshapes the potential energy surface of the polyene chain. In protonated Schiff bases, charge redistribution increases electrophilicity at multiple conjugated positions and stabilizes transition states associated with bond rotation, cyclization, and rearrangement.^77,83,84 These effects account for the readiness of bisretinoid formation, which proceeds under mildly acidic conditions, and for its resistance to suppression once retinal flux and proton availability coincide.

The same electronic features that enable bisretinoid assembly also govern the structural diversification described in Section 3.4. Small perturbations in protonation state, retinal isomer geometry, and local dielectric environment bias the reaction trajectory toward full aromatization (A2E), incomplete oxidation (A2-DHP-PE), or alternative bond-forming pathways such as all-trans-retinal dimer formation.^34,35,52,79 Therefore, chemical diversity emerges naturally from a shared electronics platform rather than discrete biosynthetic instructions.

Within a natural products chemistry framework, retinal-derived bisretinoids exemplify how quantum-mechanical design principles embedded in primary metabolites give rise to complex secondary chemistry without enzymatic direction. The retina exploits the delocalized π-electron system of retinal to generate a family of endogenous polyene natural products governed by orbital symmetry, charge stabilization, and membrane-constrained reaction dynamics. These reactions unfold continuously over decades, linking fundamental electronic structure to the gradual accumulation of bisretinoid lipofuscin-associated pigments. This chemistry establishes the foundation for the reactivity and toxicity mechanisms examined in Section 4.

4. Chemical and photochemical reactivity of bisretinoids

As detailed in Section 3, bisretinoids arise from non-enzymatic polyene-aldehyde reactions that generate structurally complex, electronically activated natural products. Once formed, these molecules are not chemically inert. Their extended conjugation, cationic character, and amphiphilicity render them highly reactive under conditions of light exposure, oxygen availability, and membrane confinement. In this section, retinal toxicity is examined as a direct consequence of bisretinoid chemistry operating with a biological context. Photooxidation, electrophile formation, and membrane disruption are analyzed as the primary chemical pathways through which bisretinoid reactivity is translated into cellular injury.

4.1. Photooxidation, epoxidation, and polyene scission

The most extensively characterized chemical transformation of bisretinoids is light-driven oxidation of their conjugated polyene systems. Bisretinoids such as A2E absorb strongly in the visible range and efficiently populate excited singlet and triplet states, enabling photosensitized reactions with molecular oxygen.^28,46,47 Excited bisretinoids generate singlet oxygen (¹O₂) via Type II electron-transfer and superoxide anion radicals via Type I electron-transfer, with the relative contributions of each pathway depending on the local chemical environment and oxygen availability (Fig. 4).^28,46,47 These reactive intermediates promote oxygen addition and oxidative modification along the conjugated polyene backbone.

Photochemical studies demonstrated that irradiation of A2E in the presence of oxygen produces a spectrum of oxidized derivatives, including mono- and poly-epoxides, endoperoxides, and peroxy intermediates.^45–47 These oxygenated species retain the core bisretinoid scaffold but exhibit an altered electronic structure and increased chemical lability. Importantly, these transformations occur at wavelengths and oxygen tensions relevant to physiological retinal illumination, underscoring their biological relevance.^45,47

Additional photochemical complexity arises from site-selective reactions of singlet oxygen along the polyene chain. Under appropriate conditions, radical oxidation initiated by the triplet-state A2E or by electron-transfer reactions generate oxidized congeners such as peroxy-A2E and furan-A2E.^85,86 Formation of furan-A2E in particular reflects a rearrangement sequence in which epoxidation, ring opening, and intramolecular attack yield a furan ring embedded into the bisretinoid scaffold. This transformation illustrates the chemical plasticity of the A2E chromophore and its capacity to channel singlet oxygen addition and subsequent rearrangements into structurally divergent products.⁸⁵

These oxidative modifications destabilize the conjugated polyene backbone promoting subsequent scission reactions. Structural characterization of photooxidation products shows that these processes generate shorter aldehydes and ketones, including highly reactive carbonyl fragments such as methylglyoxal and glyoxal, which can diffuse and react with cellular nucleophiles (Fig. 5).^33,47,87 These fragmentations reflect the intrinsic instability of polyene peroxides, which undergo homolytic or heterolytic O–O bond cleavage, followed by breakdown of the conjugated system.^57,58

Fig. 5. Triplet-state driven photochemical reactivity of A2E and the downstream formation of sub-RPE deposits. Top: reaction sequence initiated by visible light absorption. Triplet-state A2E (T₁) transfers energy to ground-state molecular oxygen via a Type II photosensitization mechanism, generating singlet oxygen (¹O₂), which drives polyene photooxidation. Type I electron-transfer pathways generating superoxide radical anion represent a secondary contribution and are not shown for clarity. Middle: two representative photooxidized A2E intermediates. In MG-generating oxidation (left), singlet oxygen addition to the C5–C6 and C7–C8 double bonds of the polyene (green peroxide and epoxide moieties) leads to oxidative scission and release of methylglyoxal (MG) as a reactive dicarbonyl fragment. In GO-generating oxidation (right), addition at the C9–C10 and C11–C12 positions (blue peroxide moieties) yields glyoxal (GO) upon scission. Red dashed lines indicate the sites of oxidative scission. Bottom: downstream consequences of MG and GO release. Both dicarbonyls react with the ε-amino group of lysine and the guanidinium group of arginine residues on RPE proteins via Schiff base formation. While the MG pathway is shown as a representative model, GO undergoes analogous reactions through a similar mechanism to generate GO- and MG-adducts. These modified proteins undergo extensive protein crosslinking, leading to the accumulation of insoluble aggregates. This chemical transformation, initiated by the reactive dicarbonyl fragments of A2E, collectively drives the formation of sub-RPE protein deposits (drusen), a hallmark of age-related macular degeneration.

Photooxidation does not simply degrade bisretinoids into inert debris but instead, converts a photoreactive chromophore into a chemically diversified ensemble of secondary products, many of which retain or even enhance biological reactivity. This transformation accounts for the increased fluorescence yield of photooxidized bisretinoids while simultaneously generating toxic reaction products, directly linking fundus autofluorescence changes to underlying chemical instability.^39,42,45 However, the magnitude of this photochemical burden is not uniform across the bisretinoid family. All-trans-retinal dimers generate singlet oxygen more efficiently than A2E, a difference reflecting the greater number of unconjugated double bonds in their polyene system, which facilitates energy transfer to molecular oxygen and likely contributes to the differential oxidative burden exerted by individual bisretinoid species within the heterogeneous RPE lipofuscin mixture.⁸⁰

4.2. Electrophilic fragments and covalent biomolecular targeting

A defining consequence of bisretinoid photooxidation is the generation of electrophilic carbonyl-bearing fragments that react covalently with cellular nucleophiles. Aldehydes and α,β-unsaturated carbonyls produced by polyene scission readily engage thiol and amine functionalities in biomolecules through Michael addition or Schiff-base formation.^33,47,88

Experimental studies using irradiated A2E and lipofuscin extracts show the formation of stable adducts with glutathione and protein cysteine residues, leading to depletion of intracellular redox buffers and irreversible modification of lysosomal proteins.^86,88 These reactions proceed without enzymatic activation and are driven by the inherent electrophilicity of the cleavage-derived aldehydes, placing them within the broader framework of carbonyl stress chemistry^48,89,90

PE, the same lipid that serves as the initial nucleophile in bisretinoid formation, is also a major downstream target of electrophilic fragments. Covalent modification of PE alters membrane packing, increases permeability, and disrupts lipid–protein interactions.^87,91 This chemistry directly links bisretinoid photochemistry to loss of membrane integrity at the organelle level.

These electrophilic reactions persist beyond the period of illumination. Unlike intact bisretinoids, whose reactivity is largely light-dependent, carbonyl fragments remain chemically active in the dark, extending the temporal window of toxicity and enabling diffusion-driven propagation of damage within cellular compartments.^47,88 This persistence converts transient photochemical events into sustained chemical injury.

4.3. Membrane perturbation as physicochemical toxicity

In addition to covalent reactivity, bisretinoids exert toxicity through direct physicochemical interactions with biological membranes. Their amphiphilic architecture, comprising a hydrophobic polyene core and a cationic or zwitterionic headgroup, promotes stable insertion into phospholipid bilayers rather than transient surface association.^28,36

Model membrane studies show that A2E intercalates into lipid bilayers through its amphiphilic architecture, altering lipid organization and increasing membrane permeability even in the absence of light.^69,70 These effects, attributable to the detergent-like properties of the A2E scaffold, are particularly pronounced in acidic environments where electrostatic interactions between the permanently cationic pyridinium headgroup and negatively charged membrane lipids promote stable bilayer insertion and lysosomal accumulation.^87,92 In RPE cells, A2E and related bisretinoids accumulate progressively within lysosomes through mechanisms independent of protonation-mediated trapping. A2E bears a permanent positive charge on its pyridinium nitrogen that does not undergo pH-dependent ionization, and other bisretinoids, including A2-DHP-PE and all-trans-retinal dimer, likewise accumulate in lysosomal storage bodies despite lacking an ionizable nitrogen.⁴⁴ Instead, lysosomal retention reflects the amphiphilic architecture of these molecules and their intrinsic resistance to lysosomal degradation.²⁹ This accumulation compromises degradative capacity and disrupts organelle homeostasis, sensitizing RPE cells to oxidative and inflammatory stress.⁹³

The physicochemical toxicity of bisretinoids extends beyond lysosomes to mitochondria, reflecting mechanisms distinct from lysosomal membrane disruption and is supported by direct experimental evidence. A2E inhibits cytochrome c oxidase (complex IV of the mitochondrial respiratory chain), an effect observed in both solubilized and reconstituted systems and attributed to interactions between its amphiphilic, cationic scaffold and the cardiolipin-rich inner mitochondrial membrane.⁹⁴ In A2E-laden RPE cells exposed to blue light, this inhibition is accompanied by mitochondrial dysfunction, including network fragmentation, imbalanced fission–fusion dynamics, decreased ATP production, elevated reactive oxygen species, and cytochrome c release, culminating in apoptosis.^95,96 In parallel, bisretinoid-dependent photooxidation induces endoplasmic reticulum stress, as indicated by upregulation of the unfolded protein response markers GRP78 and CHOP, an effect attenuated by antioxidant pretreatment and linked to the oxidative burden generated by photosensitization.⁹⁷ Taken together, these findings demonstrate that bisretinoid accumulation imposes a multi-compartment physicochemical burden on RPE cells, in which lysosomal membrane disruption, mitochondrial respiratory inhibition, and endoplasmic reticulum stress converge to progressively erode cellular homeostasis.

5. Bisretinoids as autochthonous natural products: chemical persistence and context-dependent retinal degeneration

The chemical transformations described in Sections 3 and 4 culminate in a defining feature of bisretinoid biology: progressive intracellular accumulation. The preceding sections have established that bisretinoid formation arises from non-enzymatic retinal-PE chemistry and the resulting pigments are intrinsically reactive under illumination. The remaining question is why this chemistry leads to degeneration over decades rather than acute injury, and which properties of the bisretinoid scaffold govern this temporal behavior. The answer lies in chemical persistence. Bisretinoids accumulate not because cellular clearance has failed, but because no enzymatic machinery exists to disassemble their stable pyridinium and di-retinal scaffolds.

5.1. Accumulation as a emergent outcome of enzymatic and non-enzymatic chemistry

Bisretinoids arise downstream of the visual cycle, an enzymatically controlled pathway essential for phototransduction. However, once retinaldehyde is generated, its reactions with PE and other nucleophiles proceed without enzymatic guidance.^28–31,36 The coexistence of regulated enzymatic flux and spontaneous chemical reactivity renders bisretinoid formation chemically favored but biologically unmanaged.

No dedicated enzymatic pathway exists for bisretinoid degradation. Once formed, A2E and related congeners resist lysosomal hydrolases and persist as stable amphiphilic pigments due to their permanently charged pyridinium cores and membrane-associated architecture, which limit enzymatic accessibility.^87,92,98 Therefore, their accumulation reflects intrinsic chemical stability rather than metabolic dysfunction. This property is shared with other endogenous pigments, including neuromelanin and advanced glycation end-products, which likewise arise from normal metabolism but accumulate due to resistance to enzymatic degradation.^48,89 However, bisretinoids are distinguished by continuous biosynthetic input: as long as the visual cycle operates and retinaldehyde is generated, new bisretinoid molecules are continuously delivered to RPE cells through daily phagocytosis of shed photoreceptor outer segments, where they accumulate as stable amphiphilic deposits in a uniquely permissive cellular environment.^29,31 A2E bears a permanent positive charge on its quaternary pyridinium nitrogen and does not undergo pH-dependent ionization; therefore, its intracellular retention reflects intrinsic structural stability rather than a specific failure of degradative capacity.⁴⁴

5.2. From chemical persistence to regulated cellular stress responses

The chemical stability that enables the accumulation of bisretinoids also determines the conditions under which their photochemical reaction products become biologically consequential. As biosynthetic replenishment continuously adds to the pigment pool over decades, the accumulated burden is increasingly sensitized to photooxidation and electrophilic fragmentation. Although bisretinoids resist enzymatic degradation, these photochemical reaction products are not biologically inert.²⁹ As established in Section 4, photooxidation generates reactive carbonyl species that propagates across cellular compartments, linking bisretinoid persistence to broader cellular stress responses through three converging pathways.^99,100

First, electrophilic fragments generated by polyene scission, including methylglyoxal and glyoxal, deplete glutathione and form AGE adducts with cellular proteins, contributing to carbonyl stress within the RPE.^95,99 These dicarbonyls can modify mitochondrial proteins and impair respiratory chain function, elevating reactive oxygen species and lowering the threshold for apoptotic signaling; however, direct bisretinoid-specific evidence for this mechanism remains limited.^101,102

Second, bisretinoid-driven membrane permeabilization releases cathepsins into the cytosol, triggering inflammasome activation and complement pathways associated with the para-inflammatory state of dry AMD.^73,102

Third, iron released from stored ferritin under conditions of iron dysregulation in AMD promotes oxidation of bisretinoid by generating highly reactive hydroxyl radicals (˙OH) through Fenton-type reactions.^{88,101,103,104} Photosensitization of bisretinoid resulting in singlet oxygen generation and hydroxyl radical generated by Fenton chemistry can also lead to lipid peroxidation.^88,105

Each of these pathways is initiated by a specific chemical event traceable to bisretinoid reactivity, rather than by a generalized oxidative stress signal.

These findings demonstrate that bisretinoid-associated retinal degeneration is not the consequence of a single toxic event, but an emergent property of long-lived, chemically persistent natural products operating within a constrained cellular microenvironment. The slow kinetics of degeneration reflect the time required for bisretinoid burden to exceed the adaptive capacity of RPE stress responses, with this threshold governed by retinaldehyde flux, membrane composition, light exposure history, and individual genetic background.

5.3. Bisretinoids as autochthonous natural products: convergence of evidence

The preceding sections collectively establish that bisretinoids qualify as autochthonous natural products in a chemically rigorous sense. Section 2 traced their origin to dietary carotenoids through an enzymatic lineage culminating in retinaldehyde formation. Section 3 defined their assembly in terms of non-enzymatic polyene-aldehyde condensation, iminium activation, and pericyclic cyclization, mirroring mechanisms observed in biomimetic synthesis. Section 4 linked their biological effects directly to intrinsic chemical properties, including photoexcitation, electrophile generation, and amphiphilic membrane disruption. Sections 5.1 and 5.2 further demonstrated that their pathological significance emerges from chemical persistence coupled with continuous biosynthetic input, rather than acute toxicity.

Taken together, bisretinoids fulfill the defining criteria of autochthonous natural products. They are structurally well-defined small molecules generated endogenously from dietary precursors, assembled through recognizable chemical logic without enzymatic templating, capable of exerting biological effects that arise from intrinsic physicochemical properties, and sufficiently persistent to accumulate to biologically relevant concentrations within a defined cellular niche.^{28–31,36,44} Their diversity arises from non-enzymatic branching chemistry rather than parallel biosynthetic pathways, mirroring the scaffold diversification observed in classical polyketide and terpenoid natural product families.

This perspective reframes retinal degeneration associated with bisretinoid accumulation as a consequence of chemical accumulation rather than genetic predisposition or oxidative insult alone. In the retina, pathology emerges from the gradual accumulation of membrane-active, photoresponsive natural products whose chemical properties overwhelm cellular homeostasis over decades. Recognizing bisretinoids as autochthonous natural products with defined chemical identities, rather than as amorphous oxidative byproducts, identifies specific molecular targets for intervention, including retinaldehyde flux, bisretinoid reactivity, metal availability, and pigment burden. These therapeutic implications are discussed in Section 6.

6. Outlook and emerging directions

The preceding sections establish bisretinoids as chemically persistent, autochthonous natural products whose formation and reactivity are governed by intrinsic polyene-aldehyde chemistry rather than specialized biosynthetic enzymes. This perspective reframes retinal degeneration as a problem of long-term chemical burden and context-dependent reactivity, extending inquiry beyond descriptive pathology or single-enzyme inhibition. Current advances converge on four interrelated themes: (i) regulated cell death pathways sensitized by bisretinoid chemistry, (ii) modulation of retinal and retinoid flux that controls chemical burden, (iii) strategies to neutralize or remove accumulated bisretinoid material, and (iv) structural precision as a basis for mechanistic disambiguation and targeted intervention.

6.1. Regulated cell death pathways sensitized by bisretinoid chemistry

An important conceptual shift is the recognition that bisretinoid-associated toxicity engages regulated cell death programs rather than inducing nonspecific oxidative damage. Evidence links bisretinoid burden to iron-dependent oxidative injury in RPE cells, implicating iron-catalyzed lipid peroxidation cascades with features of ferroptosis in disease progression.^88,105 Iron accumulation in the aging retina creates a permissive redox environment in which photooxidized bisretinoids and their electrophilic fragments amplify lipid peroxidation and membrane damage.⁹⁷

This convergence is mechanistically grounded in bisretinoid chemistry. Bisretinoid photooxidation generates electrophiles and lipid-reactive species (Section 4) that sensitize membranes to iron-catalyzed peroxidation, lowering the threshold for ferroptotic injury.^88,106 Thus, iron-dependent cell death arises as a downstream response to sustained chemical stress rather than a primary toxic trigger, reinforcing the role of bisretinoids as upstream chemical determinants of regulated degeneration.

6.2. Modulating chemical burden through flux control

A second emerging direction targets the flux of reactive aldehydes that drive bisretinoid formation. Classical approaches have focused on inhibition of the visual cycle enzyme RPE65 to reduce all-trans-retinal generation, and both genetic and pharmacological studies show that limiting retinal availability suppresses bisretinoid accumulation.^25,27 The chemical perspective developed here extends flux modulation beyond RPE65 to upstream processes governing vitamin A delivery, carotenoid metabolism, and retinoid handling.

Recent work shows that dietary vitamin A supplementation and systemic retinoid homeostasis influence the long-term bisretinoid load, underscoring that bisretinoid chemistry is governed by substrate supply over decades rather than acute photic events.¹⁰⁷ The proportion of retinaldehyde diverted into bisretinoid formation under physiological conditions remains unquantified. Given that bisretinoid accumulation is slow and cumulative, even a small fractional diversion of retinaldehyde flux could be sufficient to generate pathologically relevant pigment burdens over decades, but direct measurements of this conversion efficiency in vivo are currently lacking and represent an important unresolved question. Therefore, flux modulation represents a chemical strategy to reduce the probability of non-enzymatic condensation reactions rather than a disease-specific intervention. Importantly, these approaches slow bisretinoid formation but do not address the persistence of pigments already deposited within lysosomes, thereby demonstrating the need for the complementary strategies discussed in Section 6.3.

6.3. Chemical mitigation and burden reduction

A third strategy targets bisretinoids directly as chemical entities. Iron chelation attenuates bisretinoid-driven oxidative injury by limiting metal-catalyzed redox cycling, demonstrating that modification of the chemical environment can blunt downstream toxicity without altering bisretinoid structure.^88,105 In parallel, the incorporation of deuterated polyunsaturated fatty acids reduces lipid peroxidation and dampens secondary damage initiated by bisretinoid photochemistry, illustrating that modification of the membrane substrate can modulate the consequences of bisretinoid accumulation.⁵¹

Recent efforts have shifted toward burden reduction, focusing on mobilization or clearance of existing bisretinoid lipofuscin material. Pharmacological agents such as soraprazan (remofuscin) reduce the lipofuscin load in preclinical and early clinical settings, marking a transition from flux suppression to direct removal of accumulated material.¹⁰⁸ Enzymatic degradation offers a mechanistically distinct alternative: exogenous horseradish peroxidase, delivered intracellularly to RPE cells, degrades A2E through sequential epoxidation and oxidative scission of the polyene arms, generating aldehyde-bearing fragments at lysosomal pH, and reduces cellular A2E content by approximately 40%.¹⁰⁹ Although both pharmacological and enzymatic approaches remain at early stages, these studies demonstrate that bisretinoid pigments, despite their chemical stability and resistance to endogenous lysosomal hydrolases, are not immutable end products but pharmacologically addressable targets.

6.4. Photoprotective interception of bisretinoid photochemistry

A complementary strategy targets the photochemical consequences of bisretinoid accumulation rather than bisretinoid burden itself. Because singlet oxygen generated by bisretinoid photosensitization is the primary initiator of downstream oxidative injury, antioxidants capable of singlet oxygen quenching or radical scavenging represent a mechanistically orthogonal intervention.

Macular xanthophylls provide the most physiologically relevant line of defense. Lutein and zeaxanthin suppress the photooxidation of both A2E and its biosynthetic precursor A2-PE, with zeaxanthin consistently outperforming lutein owing to its greater number of conjugated double bonds, and both carotenoids exceeding α-tocopherol in singlet oxygen quenching efficiency.¹¹⁰ Protection operates primarily through physical quenching via triplet–triplet energy transfer with minimal consumption of the carotenoid, a feature that enables catalytic regeneration over multiple quenching cycles. Bilberry-derived anthocyanins act through a related but distinct mechanism, with the quinonoidal anhydrobase formed at neutral pH providing a conjugated diene system competent for singlet oxygen addition; among the nine monoglycosides characterized, malvidin 3-glucoside exhibited the greatest potency.¹¹¹

More recent work has extended protection further along the photodegradation cascade. Quercetin and cyanidin-3-glucoside not only suppress A2E photooxidation in cell-free and cellular assays, but also reduce methylglyoxal-protein adduct formation, attenuate receptor for advanced glycation end-product (RAGE) mRNA upregulation, and decrease 4-hydroxynonenal release from bisretinoid-laden photoreceptor outer segments, connecting antioxidant intervention directly to the protein carbonylation and drusen-associated pathology outlined in Section 4.¹¹² Rational scaffold modification of quercetin by Mannich reaction afforded water-solubilizing dimers with enhanced potency relative to the parent flavonol, demonstrating that synthetic optimization of natural product scaffolds can improve both activity and drug-like properties.¹¹³ Across these studies, the structural requirement for efficacy converges on ortho-dihydroxyl substitution and extended conjugation. Whether sufficient concentrations reach the RPE and photoreceptor outer segment in vivo remains a prerequisite for therapeutic translation.

6.5. Structural precision as a basis for mechanistic disambiguation and targeted intervention

A fifth direction concerns the need for structural precision defining individual bisretinoid species and resolving their distinct contributions to toxicity. Recent total synthesis efforts have enabled stereochemical reassignment of bisretinoid congeners previously misidentified by spectroscopic analysis alone. A notable example is iiso-A2E, a bisretinoid fluorophore from human, bovine, and murine RPE, which has been revised by total synthesis as the C13′ C14′ E isomer, now correctly designated i-A2E (Fig. 3). These studies demonstrate that positional and geometric isomers within the bisretinoid family, which differ in the attachment point or geometry of their polyenyl chains on the pyridinium core, cannot be reliably distinguished by NMR alone and requires synthetic confirmation. Because UV absorption and photochemical reactivities differ among these isomers, structural precision is essential for accurate interpretation of photobiological data.⁵² These findings also expose a limitation of studies that treat RPE bisretinoid mixtures as chemically uniform: observed biological effects may reflect the properties of specific molecular species present in different proportions rather than a generic bisretinoid class.

Therefore, structural precision has direct implications for mechanistic research and therapeutic design. Resolving the individual contributions of A2E, iso-A2E, A2-DHP-PE, atRALdi, and related congeners to photooxidative damage, organelle stress, and regulated cell death requires analytical and synthetic tools capable of defining stereochemically and constitutionally distinct compounds. In parallel, the design of neutralizing agents or photochemical quenchers will benefit from precise understanding of the electronic and structural features responsible for singlet oxygen sensitization and electrophile generation. Viewed in this light, chemical synthesis and structural analysis of bisretinoids are not merely academic exercises but enabling technologies for next-generation mechanistically informed therapeutic strategies.

An unresolved question concerns the origin of selectivity within the bisretinoid chemical space. Although current evidence strongly supports a non-enzymatic basis for bisretinoid formation, the possibility that specific proteins influence product distributions by stabilizing intermediates, altering membrane organization, or regulating retinal flux warrants further investigation. Elucidating whether these factors act merely as environmental modulators or as more direct participants in reaction pathways will be critical for defining the boundary between spontaneous chemical reactivity and regulated biosynthetic control in this system.

7. Conclusions

This review develops a coherent chemical account of bisretinoid natural products, tracing their origin from dietary carotenoids through enzymatic retinoid metabolism to non-enzymatic polyene-aldehyde condensation within photoreceptor disk membranes. It further connects their intrinsic physicochemical properties to the mechanisms by which they gradually erode RPE homeostasis. Across these sections, a unifying principle has emerged: bisretinoid formation, diversification, persistence, and toxicity are governed by a common chemical logic embedded in the extended conjugated polyene system of retinaldehyde.

Enzymatic processes, including the visual cycle, carotenoid cleavage, and ABCA4-mediated transport, define the context and rate of bisretinoid biosynthesis, but the chemistry itself from Schiff-base formation to aza-electrocyclization to singlet oxygen sensitization proceeds without enzymatic direction. Recognizing this distinction reframes retinal degeneration as the cumulative consequence of intrinsic chemical reactivity operating over decades within a specialized cellular niche, rather than as a failure of a single biological pathway.

Several priorities define future investigation. The individual contributions of structurally distinct bisretinoid species to RPE dysfunction remain incompletely resolved and will require both synthetic access to stereochemically defined compounds together with analytical methods capable of quantifying them in tissue. Mechanistic links between bisretinoid chemistry and regulated cell death pathways, including iron-dependent lipid peroxidation and inflammasome activation, require delineation at the level of specific chemical intermediates rather than bulk oxidative endpoints. The relationship between systemic carotenoid and retinoid status and long-term bisretinoid burden deserves prospective investigation as a modifiable determinant of degeneration risk. Finally, the demonstration that bisretinoid lipofuscin burden can be pharmacologically reduced establishes a framework in which bisretinoids are treated as chemically addressable targets rather than irreversible deposits.

Progress in these areas depends on recognizing bisretinoids as structurally defined endogenous natural products whose chemistry, rather than their mere presence, governs pathological outcome.

8. Conflicts of interest

There are no conflicts to declare.

Biographies

Biography

Hye Jin Kim.

Hye Jin Kim is an Assistant Professor at the College of Pharmacy, Keimyung University. She received her PhD degree in Pharmacognosy from the Kyung Hee University in 2012. She then joined the laboratory of Janet R. Sparrow at the Columbia University Medical Center, where her research focused on the structural characterization of novel bisretinoid compounds in RPE lipofuscin, as well as the photochemical and oxidative mechanisms underlying the bisretinoid-mediated retinal degeneration. Since joining the faculty at the Keimyung University in 2022, her research has centered on natural product-based therapeutic approaches for retinal degenerative diseases.

Biography

Janet R. Sparrow.

Janet R. Sparrow is the Anthony Donn Professor of Ophthalmic Science at the Columbia University, with a joint appointment in the Department of Pathology and Cell Biology. She received her PhD degree from the University of Toronto. Her laboratory pioneered the structural and biochemical characterization of RPE lipofuscin, demonstrating that it consists of visual cycle-derived bisretinoids whose toxicity arises from photochemical reactivity and iron-catalyzed oxidative damage. She has published more than 250 papers and received numerous honors, including the Lew R. Wasserman Merit Award, the Alcon Research Institute Award, and an American Academy of Ophthalmology Achievement Award.

Biography

Young Pyo Jang.

Young Pyo Jang is a Professor at the College of Pharmacy, the Kyung Hee University, Seoul. He received his PhD degree in Pharmacognosy from the Seoul National University. Between 2002 and 2006, he conducted postdoctoral research at the Columbia University (New York, NY), focusing on anthocyanins and aging pigments in the laboratories of Professor Koji Nakanishi and Professor Janet Sparrow. Since joining the faculty at Kyung Hee University in 2007, his research has centered on the standardization of herbal medicines and natural raw materials using various metabolomics tools. He served as the President of the Korean Society of Pharmacognosy in 2024.

9. Data availability

No primary research results, software or code have been included, and no new data were generated or analysed as part of this review.