Photosensory and Signaling Properties of Cryptochromes

The blue-light protein sensors, cryptochromes, compose the widespread class of photoreceptors that regulate processes of development in plants and circadian rhythms in animals and plants. These photoreceptors can also function as magnetoreceptors. During the past decade, cryptochromes were discovered and characterized in several photosynthesizing algae, where they may act not only as regulatory photoreceptors but also as photolyases, catalyzing the repair of ultraviolet-induced DNA lesions. Cryptochrome proteins bind flavin adenine dinucleotide (FAD) as the chromophore in the photolyase homology region (PHR) domain and contain the cryptochrome C-terminal extension (CCE), which is attached to PHR near the FAD-binding site. The chromophore is responsible for photosensory properties of cryptochromes, while CCE is essential for their signaling activities. Photosensory processes are initiated by photochemical FAD conversions involving electron/proton transfer and the formation of redox forms. These reactions lead to changes in chromophore–protein interactions. The resulting conformational transitions in protein structure provide the molecular foundation of cryptochrome signaling activity in living systems. In plants, cryptochrome protein with photoreduced FAD undergoes conformational changes, causing disengagement of the PHR domain and CCE, which is accompanied by the formation of functionally active dimers/tetramers of cryptochrome molecules. Photooligomerization is considered a key process necessary for cryptochrome signaling activity, inasmuch as oligomers ensure the formation of their complexes with a variety of proteins that are the components of photoreceptor signaling pathways. Cryptochrome–protein interactions in such complexes change the protein-signaling activities, leading to gene expression alteration and photomorphogenesis. The review discusses current knowledge on photosensory and signaling properties of cryptochromes.

Ultraviolet light, visible and of near infrared regions of electromagnetic radiation spectrum (290-800 nm), is a key environmental stimulus, which is accepted by the organisms of the entire biological kingdom. In addition to its role as a major source of energy in photosynthesis, the light carries vitally important information about space and time, while contributing to adaptation of the organisms to the habitat conditions. Behavioral responses and processes of development and launching of circadian rhythms are common types of physiological adaptation to absorption of light.
Living systems detect light stimuli by mediation of sensory photoreceptors, which convert a physical signal into biochemical signaling cascades and subsequent photobiological responses. Photoreceptor proteins commonly contain molecules of chromophores sensitive to photons of different energies. Absorption of photon by the chromophore photoreceptor in a state adapted to the dark initiates a series of photochemical reactions ("photocycle"), which bind the chromophore to the surrounding protein structure.
The occurring changes manifest themselves in the transition of a photoreceptor from a dark (inactive) state into a state adapted to light or, in other words, a "signaling" conformational state. This transition marks the photoactivation of a photoreceptor. Commonly, the photocycle is fully reversible, that is, metastable signaling state is spontaneously disengaged in a thermal reaction into the initial dark (light-independent) state. Several different classes are distinguished in photoreceptors based on specific nature and photocycle of a chromophore. They include phytochromes, which are sensors of red/far-red light (600-750 nm), cryptochromes, and light-oxygen-voltage (LOV) proteins, which are sensors of ultraviolet light (UV-light) in the blue/ultraviolet А light range (UVA) (320-400 nm/400-500 nm), as well as UV resistance locus 8 (UVR8) protein, which is a sensor of UV-light photons in the B range (UVB) (290-320 nm) [1]. Except for the UVR8 protein not containing the special chromophore, each photoreceptor can be functionally divided into a photosensory module containing a chromophore and providing for light absorption and REVIEW an effector (signaling) module mediating a transduction of signal and physiological response. This review considers flavoprotein photoreceptors of a cryptochrome with flavin adenine dinucleotide (FAD) as the chromophore.

PHOTOEXCITED STATE PROPERTIES OF THE FLAVIN CHROMOPHORES
Photophysical, photochemical, and spectroscopic properties of flavins are known to be governed by a system of conjugated double bonds from an isoalloxazine ring (Fig. 1a). Absorption of a blue UVA photon triggers charge redistribution in the isoalloxazine ring and changes the flavin redox potential, which initiates its photochemical conversions, involving electron/proton transfer and formation of the radical forms. Results of theoretical isoalloxazine ring studies suggest that both the first singlet (S 1 ) and triplet (T 1 ) excited flavin states correspond to the π-π*-transitions. Properties of these states, however, differ in FAD and flavin mononucleotide (FMN), acting as a chromophore in LOV-photoreceptors. In the latter, the T 1 -state is formed with a high quantum yield (ϕ = 0.5-0.7). In these conditions, effective occupation of T 1 -level by FMN responsible for photochemical activity of this chromophore is exclusively governed by endogenous properties of the isoalloxazine ring. Chromophore of FAD cryptochromes differs from FMN in that photoexcited isoalloxazine displays its properties under the influence of spatial proximity of the second heterocyclic compound, that is, adenine. The resulting stacking (stacked U-like configuration) provides for photoinduced intramolecular electron transfer, leading to quenching of the S 1 -state, decrease in fluorescence intensity with the maximum at 520-530 nm, and considerably weakened T 1 -state formation in FAD [2]. FAD in cryptochromes of plants can exist in four forms; specifically, fully oxidized (FAD), an anionic radical form (FAD •-), a neutral radical form (FADН • ), and anionic reduced form (FADН -). Reduction of photoexcited flavin (FAD*) includes the following reactions: FAD* + е -→ FAD •-+ Н + → FADН • + е -→ FADН - [3]. Molecular orbital  involved in FAD → FADНtransformation after the sequential second electron attachment is an antibonding between N 5 -C 4a and bonding between C 4a -C 10a . This transforms N 5 =C 4a -C 10a =N 1 bonds of the fully oxidized flavin form in the N 5 -C 4a =C 10a -N 1 bond of the anion-reduced form (FADН -) [3]. According to the theoretically estimated absorption electronic spectra with vibrationally resolved structure, the lowest energy peak for all redox forms of FAD is represented by a single electronic transition of the type π-π* centered on the isoalloxazine ring. Analysis of these spectra demonstrates that they overall match experimental absorption spectra of FAD (Fig. 1b). The peak in absorption spectra for all redox forms of FAD is seen in the UVA range of 360-370 nm. Anionic reduced form (FADН -) displays a single peak in this range (FADНpoorly absorb visible light). The peak is found in the visible range of the spectrum at 450 nm in fully oxidized form of FAD; the peaks occur at 410 and 470 nm in anion-radical FAD •-. The peak of FADН • is strongly displaced from the blue range to green and red ranges of the spectrum between 500 and 650 nm; this is a critical distinction of neutral radical from FAD and FAD •- [4]. The reported data shows that each redox form of FAD is characterized by specific peaks in absorption spectra. Therefore, the absorption spectra in the UVA and visible light ranges provide important information about redox-states of FAD chromophores in photoreceptors. As noted above, FAD of cryptochromes undergoes redox reactions at photon absorption. These reactions induce photosensory process, which includes changes in chromophore-protein interactions. The resulting conformational transitions in the structure of a protein immediately surrounding the chromophores are distributed inside photoreceptors, which form a molecular basis of their signaling activity in living systems. Multiple papers, reporting novel, more in-depth information about primary mechanisms of photosensory and signaling processes mediated by cryptochromes in various organisms were published in recent years. The present review largely focuses on the analysis of data contained in these publications.
PROTEINS OF THE CRYPTOCHROME FAMILY/PHOTOLYASES: OVERVIEW Cryptochromes CRY1/CRY2 of Arabidopsis thaliana were identified first. Later, genes of the related proteins were found in genomes of multiple organisms from various phylogenetic groups, such as microbes, algae, fungi, and animals [2,[5][6][7]. All cryptochrome proteins share a significant structural similarity with photolyases. These light-sensitive enzymes contain cofactor FAD in the form of FADН -. In the photoexcited state, FADНis directly involved in repair of two major UV-induced DNA lesions, specifically, cyclobutane pyrimidine dimers (CPD) or pyrimidine 6-4 pyrimidone photoproducts (6-4PP) (Fig. 2). Enzymes repairing CPD or 6-4PP have been named CPD-photolyases or (6-4)-photolyases, respectively [8]. Quantum yield of photorepair by (6-4)-photolyases (ϕ = 0.1) is considerably smaller than quantum yield of photorepair by CPD-photolyases (ϕ = 0.7-0.9), despite structural similarity of these two types of photolyases and basic photochemistry of their cofactor FAD. This can be attributed to 6-4PP repair being a more complex reaction compared with CPD photorepair. We will discuss it in some detail in the next section.
It is generally assumed that cryptochromes evolved from photolyases as a precursor, while genes of cryptochromes and photolyases split prior to the prokaryote-eukaryote divergence. This is evidenced by data on the presence of both photolyases' and chryptochromes' genes in bacterial genomes. Discovery of proteins with dual function, that is, the photolyase and regulatory, in fungi, diatoms, and chlorophytes confirms common evolutionary roots of photolyases and cryptochromes [9][10][11].
Both cryptochromes and photolyases contain FAD as a basic chromophore; however, FAD primarily governs cryptochromes' photosensory and signaling properties in contrast to its catalytic function in photolyases. The majority of cryptochromes fail to catalyze DNA photorepair. Their particular members, however-belonging to the cryptochrome/photolyase family (CPF)-retain this capability. According to the contemporary phylogenetic and functional classification [7,9], CPF proteins are divided into four classes. Of them, two major classes are the plant cryptochromes and plant-like proteins (pCRY) and cryptochromes of animals and animal-like proteins (aCRY). Cryptochromes of plants and chlorophytes, e.g., CPH1 Chlamydomonas reinhardtii, primarily act as sensory photoreceptors, whereas the animal cryptochromes can perform functions of either photoreceptors (type I) or regulators of circadian rhythms not responsive to light (type II). Type I cryptochromes were found in insects (cryptochrome of Drosophila melanogaster, dCRY, is considered their prototype); type II cryptochromes were found in mammals. Another type of cryptochromes was recently identified in birds, fish, amphibians, and reptiles; it contains FAD and is photochemically active [12].
Additionally, animal-like cryptochromes were discovered and described outside of the animal kingdom, specifically, in photosynthesizing algae, where they can act not only as signaling photoreceptors but also as (6-4)-photolyases [10,11]. Along with pCRY (CPH1) protein, chlorophyte C. reinhardtii contains aCRY (CraCRY) protein, which is 40-50% homologous in the sequence to aCRY proteins of chlorophyte Ostreococcus tauri (OtCPF1) and diatom Phaeodactylum tricornutum (PtCPF1) as well as (6-4)-photolyase of chlorophyte Dunaliella salina. Cryptochromes PtCPF1, OtCPF1, and CraCRY are bifunctional proteins, exhibiting (6-4)-photolyase activity and the function of gene expression regulator. Interestingly, the regulatory function of the CraCRY protein was identified not only with the exposure to blue light absorbed by FAD chromophore in oxidized form but also with the exposure of red light absorbed by FADН • [11].
Nearly all CPF proteins undergo light-induced reaction, so-called photoactivation, in which catalytically inactive photoexcited FAD is reduced by a nearby tryptophan (removes an electron from a nearby tryptophan). In the reduced state, flavin chromophore either initiates transduction of signal (FADН • ) or, after further absorbing a photon, catalyzes DNA repair (FADН -). Rapid transport of an electron from the protein surface on FAD is provided by tryptophan triad. This intracellular electron transfer pathway is extended in CraCRY and cryptochromes of animals of type I on account of the fourth aromatic residue [13,14], that is, tyrosine (CraCRY) or tryptophan (dCRY). The extended electron-transport pathway is essential both for FAD photoreduction reaction in proteins of aCRY and functioning of (6-4)-photolyases.
All cryptochromes are defined based on their similar two-domain structure, including the N-terminal α/β-domain containing the β-sheet and the C-terminal α-helix domain, binding a FAD chromophore. In combination, they compose a highly conservative photolyase homology region (PHR) also referred to as PHR-domain. Additionally, the structure of cryptochromes includes the C-terminal extension (CRY C-terminal extension, CCE), also referred to as CCE-domain, differing in the number of amino acid residues [15]. In CRY1 and CRY2, the CCE-domain consists of 180 and 110 residues, respectively. The CCE-domain strongly binds to the PHR-domain in darkness but escapes the PHR nucleus after photoactivation and transits into a disordered state. Next, the destructured part of CCE of 80 residues accommodates interactions with some proteins that are the components of signaling pathways of cryptochromes. The CCE-domain is absent in photolyases and plays a role of effector domain in cryptochromes. The PHR of cryptochromes was earlier believed to act exclusively as a photosensory domain. Recent data on physical interaction of the majority of signaling proteins with PHR-domains of cryptochromes, however, suggest that, similar to CCR, PHR also can perform function of the effector domain [16]. This is the first time when a particular crystal structure of PHR-domain CRY1 appears to be surprisingly similar to a structure of CPD-photolyase in Escherichia coli, despite their evolutionary distance. A crystal structure of plant cryptochrome of the full length has not, however, been obtained due to a technical difficulty of crystallization of protein with large intrinsically disordered CCE-domain. Not only the full-length crystal structure but also the complexes with corresponding signaling proteins were successfully identified in Drosophila cryptochrome dCRY, containing a comparatively small CCE-domain. Findings of this study demonstrate a key role of FAD-binding pocket and physical interactions between PHRand CCE-domains in functioning of animal cryptochromes [17]. It is generally agreed that light-dependent changes in interaction between PHR-and CCEdomain can explain photoinduced conformational changes in plants as well. The particular PHR and CCE interaction sites, however, have yet to be identified.
According to contemporary ideas, photoactivation of plant cryptochromes begins with blue-light photon  Recent studies of the abovementioned processes generated novel data discussed below.

PHOTOCHEMICAL PROCESSES IN CPF PROTEINS AND PHOTOACTIVATION OF CRYPTOCHROMES
Structural similarity between the PHR-domain of cryptochromes and photolyases in f lavin-binding protein pocket, in particular, defines the common principle of their response to light. Specifically, photoinduced excitation of chromophore initiates reactions of electron transfer in both types of proteins. Mechanisms of the photochemical processes, however, fundamentally differ between cryptochromes and photolyases. This is particularly evident in redox-forms of FAD chromophore in the course of its photocycles. Light-driven activation of photolyases takes place with involvement of two chromophores, that is, catalytic FAD, existing in its basic state in an anionic reduced form (FADН -), and antenna chromophore. The majority of photolyases has the function of antenna chromophore performed by 5,10-methenyltetrahydrofolate (MTHF) or 8-hydroxydeazaf lavin (8-HDF). These particular chromophores feature high molar absorption coefficients within the UVA range (ε 370nm = 29 000 M -1 cm -1 in MTHF or ε 400nm = 25 000 M -1 cm -1 in 8-HDF) compared to FADН -(ε 370nm = 8000 M -1 cm -1 or ε 400nm = 2800 M -1 cm -1 ). This increases the number of absorbed photons utilized for photophysical resonance excitation energy transfer to FADНwhile enhancing its potential to reduce CPD or 6-4PP for their subsequent repair. Additionally, the rate of DNA repair may be significantly increased, which is important in low light conditions. Photolyase, binding a DNA region that contains CPD or 6-4PP, forms a stable complex in which these photoproducts come into close (within a range of action of the van der Waals forces) contact with FADН -. The catalytic act involves electron transfer from FADНin a singlet-excited state to DNA lesion, which results in the formation of a complex with a charge transfer between a flavin neutral radical (FADН • ) and anion-radical of CPD or 6-4PP. Next, a CPD anion radical undergoes redistribution of elec-tron density and cyclobutane ring cleavage into two initial pyrimidines, while FADН • is reduced through reverse excess electron transfer to an active FADНform involved in subsequent photocatalytic cycles.
In the anion-radical, the 6-4PP C6 5'-end base is bonded to the C4 3'-end base with an OH-group at the C4 3'-end base transposed to the C5 5'-end base (Fig. 2). In contrast to CPD, in which breakage of 5-6-с bonds reduces bases to their initial forms, breaking the C6-C4 and C5-OH in anion-radical of 6-4PP would result in the formation of two damaged bases; in other words, no DNA repair would occur. Therefore, the enzyme needs to catalyze not only the reactions of bond breakages but also a reaction of OH-group transfer to repair 6-4PP. The intermediate in a form of 6-4PP anion-radical cannot, however, undergo these two concerted reactions; therefore, a model was proposed according to which the formation of undamaged pyrimidines (after breaking the C4-C6 and C5-OH bonds) is possible under the condition of conversion of the anion-radical into another intermediate, specifically, oxetane. This mechanism was recently confirmed by findings of the crystallographic analysis of the CraCRY PHR/6-4PP-DNA complex [18]. It follows from the analysis that a key role in the formation of oxetane intermediate is played by two histidine residues in the active site that are involved in proton transfer. According to the obtained results, a mechanism of DNA repair by (6-4)-photolyase requires a second photon to be absorbed to proceed from the oxetane intermediate to breaking the C4-C6 and C5-OH bonds. This second step is analogous to the single-photon-driven repair of pyrimidine dimers by CPD-photolyases.
Another pathway of electron transfer exists in photolyases in addition to cyclic electron transfer from photoexcited FADН -, during which DNA lesions undergo repair. It involves three or four conservative tryptophan residues. This pathway called photochemical activation is not required for DNA repair but is assumed to be important for maintaining occupancy of a FADНexcited state [9].
The majority of data on photochemistry of cryptochromes was obtained as a result of studying the CRY1/CRY2 in A. thaliana, CPH1 and CraCRY in C. reinhardtii, and dCRY in D. melanogaster [14,[19][20][21][22]. In contrast to photolyases, in the ground state, these cryptochromes contain FAD in a fully oxidized form and do not commonly (except CraCRY, which additionally inserts chromophore 8-HDF) bind an antenna chromophore. As seen from cryptochrome structures, this is determined by replacement of the key amino acid residues in protein pockets, binding an antenna chromophore, which prevents its recognition and insertion.
Study of CRY1 and CPH1 using infrared Fourier spectroscopy in combination with time-resolved UV-vis spectroscopy established that FAD, absorbing the blue light in excited singlet state, is reduced to anion-radical FAD •by mediation of electron transfer from a nearby tryptophan residue (Trp 1 Н), which is part of a conserved tryptophan triad. This ultrafast reaction of electron transfer to FAD proceeds with time constants of 0.4 ps [23]; while the reaction of electron transfer between terminal tryptophan (Trp 3 Н) and Тrp 1 Н takes 31 ps. They are accompanied by the formation of terminal tryptophan (Тrp 3 Н •+ ) cation radical, which transitions into radical after deprotonation. The subsequent interaction of with tyrosine residue (TyrОН) causes its reduction to Тry 3 Н and formation of tyrosine radical (ТyrО • ) within several ms. FAD anion-radical (FAD •-) formed at photoreduction becomes protonated in CRY1 and CPH1 within several microseconds to neutral radical (FADН • ). According to the received data, aspartic acid residue located in close proximity to isoalloxazine ring of flavin can act in this reaction as a proton donor. Disappearance of the FAD •radical is obviously separated from the processes inside tryptophan triads since its protonation and transition into a FADН • form is delayed in time by six or seven orders of magnitude in relation to the initial electron transfer. The distinct separation of two processes appears surprising, considering that the proton donor is located in close proximity to flavin. This discordance may be explained by the fact that the structural reorganization of aspartic acid is a prerequisite for electron transfer, inasmuch as it forms a hydrogen bond with an oxygen atom of the protein backbone [23]. Interestingly, only plant cryptochromes conserved aspartic acid residue, whereas the latter is replaced with cysteine residue in animal cryptochromes of type I, e.g., in dCRY. This can possibly be attributed to deprotonation associated with electron transfer at FAD photoreduction in dCRY and formation of the FAD •anion-radical only. This assumption is supported by data according to which the replacement of aspartic acid with cysteine in PHR-domain of CPH1 completely blocks electron transfer. The reaction of protonation is functionally important for plant cryptochromes since the lifetime of FADН • increases to several milliseconds in vitro and several minutes in vivo [20,24].
As noted above, FADН • absorbs in green and red ranges (500-650 nm) in contrast to all other flavin redox-forms, exclusively absorbing in the blue/UVA region. This FADН • property proves its role as a signaling form of flavin chromophore in plant cryptochromes. The addition of green light was shown to inhibit a number of responses to blue light controlled by cryptochrome. The antagonistic action of green light is correlated with a decrease in FADН • concentration as a result of its photoreduction to inactive form FADН − . Photochemical transformations of flavin chromophore are completed with the reaction of FADН − with oxygen and regeneration of an initial oxidized FAD state [19]. Altogether, a redox photocycle of FAD-chromophore of plant cryptochromes can be presented in the following basic outline: At the absorption of blue-light photon in a 400-500 nm (hν-1) range, FAD transitions to a photoexcited state FAD* and an electron is transferred to it from residue of tryptophan included in tryptophan triad. The resultant cation-radical FAD •− becomes protonated to neutral radical FADН • . A reduced anion FADН − is formed after the absorption by the neutral radical of a long-wave visible light photon in the range of 500-650 nm (hυ-2) and an electron transfer onto it. The cycle is completed with oxidation of this flavin form by molecular oxygen to an initial FAD form. Radical pair [FADН • + ] can be a mediator at the first oxidation step. At the second step, superoxide in the radical pair is replaced with its protonated form , with which the former is at equilibrium and which oxidizes FADН • to FAD, yielding H 2 O 2 [25].
As previously noted, signaling activity of cryptochromes is provided by their CCE-domain, whereas the PHR-domain performs regulated control over the CCE function. After the photoactivation, the CCEdomain is released and can interact with components of cryptochrome signaling pathways. The question on how the CCE release is related to cryptochrome photochemistry, however, remained unclear until recently.
A study using infrared time-resolved spectroscopy found an appreciable change in a β-sheet structure after the formation of signaling FADН • form in the N-terminal α/βsubdomain of CPH1-PHR photoreceptor within 500 μs. This fact is considered as evidence to a key role of β-sheet in PHR and CCE interaction as well as its role in the loss of its structure in the signaling process [23]. Interestingly, in CRY1, the CCE transitions into a disordered state within 100 μs after the FADН • formation, in other words, much later. A model proposed based on these data shows [23] how the structural changes are synchronized with photochemistry of flavin chromophore in plant cryptochromes.

STRUCTURAL ASPECTS OF SIGNAL TRANSDUCTION BY CRYPTOCHROMES: PHOTOOLIGOMERIZATION AND INTERACTION WITH SIGNALING PROTEINS
It follows from the data considered above that cryptochromes with photoreduced FAD undergo conformational changes leading to disengagement of PHR-and CCE-domains. It has been recently shown that this process may induce changes in interaction between cryptochrome molecules accompanied by their homo-oligomerization. Photooligomerization is required for physiological activity of plant cryptochromes since it ensures an increase in their affinity to signaling proteins, such as transcription factors of CRY-interacting basing helix-loop-helix (bHLHs) (CIBs), constitutive photomorphogenic1 (COP1)/suppressors of phytochrome A (SPAs), and blue-light inhibitors of CRYs (BICs). Cryptochrome-protein interactions in such complexes change the protein signaling activities, leading to gene expression alteration and photomorphogenesis [26]. It should be noted that key details of this model have not been precisely defined, particularly with respect to structural changes in cryptochrome homo-dimer, based on which the latter could be distinguished from a monomer structure.
Genetic study on the structure-function relationship in plant cryptochromes was conducted to find a solution to this question [27,28]. Mutations of Trptriad residues in CRY1 or CRY2 were shown to block an electron transfer cascade and FAD photoreduction in vitro but have no effect on physiological activity of mutants in vivo. Some mutants, namely, W374A in CRY2 or its equivalent W377A in CRY1, exhibited constitutive, that is, light-independent physiological activity. These mutants benefited the establishment of a structure of cryptochrome active homo-oligomers. In contrast to the wild-type CRY2, which undergoes photooligomerization, mutant W374А of CRY2 exhibits homo-oligomerization in vitro. This result is consistent with the data that photooligomerization is required for functional activity of CRY2 and that mutant W374A of CRY2 is constitutively active in vivo [26].
Configurations of homo-dimer and homotetramer of mutant W374A of CRY2, in which these oligomers are formed by mediation of interactions of PHR-domain surfaces, were identified using lowtemperature electron microscopy with visualization. Mutations at the CRY2 interface resulted in a decrease in homo-dimer affinity to the signaling protein of CIB1. Based on this fact, CIB1 was assumed to interact with CRY2 at the homo-dimer interface and, therefore, precisely the oligomers rather than monomers appear to be the active forms of cryptochromes [29]. It should be noted that the majority of residues found in either (one of the two) interfaces of the CRYtetramer are conserved in plant cryptochromes but not in cryptochromes of animals. In this regard, it would be interesting to determine whether homo-oligomerization is required for functioning of this group of cryptochromes, which could develop other interface structures, or if the dimerization-dependent mechanism of cryptochrome photoactivation in plants is unique.
Photoactivated plant cryptochromes are known to undergo inactivation by three mechanisms, including spontaneous dark reversion of homo-oligomers to monomers, interaction with inhibitors (BICs), and ubiquitin-dependent proteolysis. Of these mechanisms, only inhibition of chryptochromes induced by BICs is a light-sensitive process. This suggests a more dynamic role of CRY-BIC interaction in regulation of cryptochrome activity in plants growing in the light. The recently performed analysis of crystal structure of the PHR-CRY2 complex with BIC2 identified two potential mechanisms, shedding light on a question as to how BIC proteins can inactivate plant cryptochromes [30]. First, BIC can inhibit FAD photoreduction. The CRY2 residues of the order of ten are located in the immediate proximity to FAD. Binding of BIC2 and PHR-CRY2 increases the distance between electron donor (W397) and acceptor (FAD isoalloxazine ring), which may hamper electron transfer. Additionally, this binding may lead to a rotation of еру carboxyl side chain of a presumed proton donor (D393) by 50°, which increases a distance between proton donor and proton acceptor in еру chromophore. This change makes the pronation practically impossible. Consequently, BIC2 may block photoreduction of FAD to FAD •− and its protonation to FADН • . Second, BICs may act as competitive inhibitors of CRY2 homo-oligomerization. In the PHR-CRY2 and BIC2 complex, a structure of BIC2 fragment governs its capability to wrap around the groove between two subdomains of еру CRY2 PHR-domain. PHR-CRY2 and BIC2 each have 16 residues that are involved in forming the complex between them. Individual mutations of several residues at the complex interface lowered the apparent affinity between PHR-CRY2 and BIC2 in vitro. The CRY-BIC heterodimer and CRY-CRY homodimer contain two CRY2 residues, namely, W349 and arginine R208. In еру heterodimer, W349 residue hydrophobically interacts with isoleucine I57 of BIC2, whereas R208 forms a salt bridge and hydrogen bond with glutamine E50 of BIC2. In the CRY2-CRY2 homodimer, W349 and R208 are at the interface. These data argue strongly that binding of BICs to CRYs competitively inhibits photooligomerization of CRY, while blocking its photoactivation.
In summary it can be said that the knowledge about primary mechanisms of light perception by cryptochromes and principles of their functioning in various organisms significantly advanced over recent years. A special consideration should be given to findings in regard to cryptochrome photoreceptors recently discovered in chlorophyte C. reinhardtii (CPH1, Cra-CRY) and in vertebrates (CRY4). These photoreceptors were characterized with respect to the structure and versatile functions; the studies examined in detail the photocycles of flavin chromophores, which differ due to the nature and number of the aromatic residues and other amino acids involved in electron/proton transfer cascade at flavin photoreduction [11,31]. Light absorption by еру chromophore initiates electron transfer to flavin through еру tryptophan triad in plant cryptochromes. This electron-transport pathway includes the fourth aromatic residue, which can be either tryptophan (dCRY) or tyrosine (CraCRY) in the type I animal and animal-like cryptochromes. In еру cryptochrome of pigeon Columba livia (ClCRY4), еру Trp-triad chain is extended to include the fourth tryptophan and tyrosine residue. This imparts the highest quantum yield of flavin chromophore photoreduction in ClCRY4 compared to other cryptochromes.
Despite the great length at which FAD photoreduction mechanism was studied in vitro, a question on specifically how electron transfer to FAD is involved in photoactivation of cryptochromes remains unresolved. Mutations of Trp-triad residues in CRY1/CRY2 that block electron transfer and FAD photoreduction in vitro were shown not to affect their biochemical and physiological activity in vivo [27,28]. In other words, Trp-triad-dependent FAD photoreduction is not a prerequisite for cryptochrome functioning. Further studies on alternative pathways of electron transfer are apparently necessary to uncover photoactivation mechanism of cryptochrome proteins applying the novel and innovative approaches. Another issue associated with FAD photocycle applies to a magnetic sensing function of cryptochromes.

RADICAL PAIRS IN FAD PHOTOCYCLE
AND MAGNETIC SENSING OF CRYPTOCHROMES Formations of radical pairs that can be affected by weak magnetic fields are generally considered to underlie a magnetic sensing of cryptochrome proteins [32]. This may result in a change in chemical constants of a rate of FAD redox reactions, causing difference in the photoreceptor activated state concentration and, consequently, changed biological activity. The natures of magnetic sensing radical pairs and stages at which they act during cryptochrome redox-cycle have not been, however, precisely defined and remain a debatable issue. Two alternative models of different radical pairs governing a magnetic sensing of cryptochromes were proposed based on multiple experimental and theoretical studies. According to model one, radical pair [FAD •-Trp •+ ] is formed by an initiating reaction of electron transfer to photoexcited FAD from tryptophan residues included in the tryptophan triad (tetrad). This radical pair undergoes a magnetically sensitive interconversion between singlet and triplet states: . FAD •can transition to FADН • form when becoming protonated and to an inactive FAFНform after photoreduction.
FADНreoxidation in the darkness by reaction with oxygen forms the basis of model two assuming that a radical pair [FADН • ] may be an intermediate in this process. Additionally, it may undergo a magnetically sensitive interconversion between singlet and triplet states with the singlet state transitioning to a FAD initial form after a release of hydrogen peroxide as shown on the FAD-chromophore redox-cycle diagram above. This mechanism was partially confirmed by the recent study [33]. CRY1 responses to light in vivo were shown to amplify in magnetic field, even including cases of exposure to it during the dark intervals between light pulses. Therefore, the magnetically sensitive reaction in cryptochrome photocycle is considered to take place at a FADНreoxidation stage and possibly involve reactive oxygen species. Their specific nature has not yet, however, been defined. Additionally, theoretical analysis data argues against a flavin/superoxide radical pair as the magnetosensing intermediate in cryptochromes due to very rapid superoxide spin relaxation. Further studies are obviously necessary for identifying the magnetosensing intermediates.
The major problematic issue with the foregoing model one lies in that it includes some findings that remain debatable due the existing controversies between photochemical and biological data. Additionally, some studies on bird orientation in a magnetic field generated data inconsistent with the discussed model. One of them demonstrates that birds can orient in a magnetic field when affected by green light, whose photons are not absorbed by FAD but by FADН • only. Consequently, as a result, it can be reduced to FADН -, which excludes a possibility to form a flavin/tryptophan radical pair and its modification by a magnetic field. The in-depth photochemical and structural analysis of cryptochrome from C. livia (ClCRY4) was performed in a recent study for clarification of the question concerning the inconsistency of the photochemical and biological data. The No. 2 2022 FRAIKIN obtained crystal structures exhibited modifications in evolutionary conservation of tryptophan triad; specifically, the tryptophan chain is extended on account of the fourth tryptophan and tyrosine residue in ClCRY4 [31]. This ensures a high quantum yield of FAD photoreduction reaction (ϕ ~ 0.4), exceeding the comparable quantum yields in other cryptochromes (ϕ 0 .2). High quantum yield increases the sensitivity of ClCRY4 to low intensity light, which is consistent with observations of behavioral responses in birds during their orientation in a magnetic field. Overall, the study results are in line with night migratory behavior in vertebrates at low lights levels and predict photochemical pathways allowing magnetosensing in the living systems. A number of questions in this field, however, remain unresolved. Novel approaches to the investigation of behavioral responses of the organisms are required for their clarification as well as further photochemical studies of cryptochromes in migratory birds, which will provide deeper insight into mechanisms of biological adaptations in various organisms.

FUNDING
This study was performed under the state assignment of Moscow State University, project no. 121032500058-7.

COMPLIANCE WITH ETHICAL STANDARDS
Conflict of Interest. The author declares that he has no conflicts of interest.
Statement on the Welfare of Animals. The article does not contain any studies involving animals performed by any of the authors.