Initial characterization of a blue-light sensing, phototropin-related protein from Pseudomonas putida: a paradigm for an extended LOV construct

The open reading frame PP2739 from Pseudomonas putida KT2440 encodes a 151 amino acid protein with sequence similarity to the LOV domains of the blue-light sensitive protein YtvA from Bacillus subtilis and to the phototropins (phot) from plants. This sensory box LOV protein, PpSB2-LOV, comprises a LOV core, followed by a C-terminal segment predicted to form an alpha-helix, thus constituting a naturally occurring paradigm for an extended LOV construct. The recombinant PpSB2-LOV shows a photochemistry very similar to that of YtvA and phot-LOV domains, yet the lifetime for the recovery dark reaction, taurec=114 s at 20 degrees C, resembles that of phot-LOV domains (5-300 s) and is much faster than that of YtvA or YtvA-LOV (>3000 s). Time-resolved optoacoustics reveals phot-like, light-driven reactions on the ns-micros time window with the sub-nanosecond formation of a flavin triplet state (PhiT=0.46) that decays into the flavin-cysteine photoadduct with 2 micros lifetime (Phi390=0.42). The fluorescence spectrum and lifetime of the conserved W97 resembles the corresponding W103 in full-length YtvA, although the quantum yield, PhiF, is smaller (about 55% of YtvA) due to an enhanced static quenching efficiency. The anisotropy of W97 is the same as for W103 in YtvA (0.1), and considerably larger than the value of 0.06, found for W103 in YtvA-LOV. Different to YtvA and YtvA-LOV, the fluorescence for W97 becomes larger upon photoproduct formation. These data indicate that W97 is located in a similar environment as W103 in full-length YtvA, but undergoes larger light-driven changes. It is concluded that the protein segment located C-terminally to the LOV core (analogous to an interdomain linker) is enough to confer to the conserved tryptophan the fluorescence characteristics typical of full-length YtvA. The larger changes experienced by W97 upon light activation may reflect a larger conformational freedom of this protein segment in the absence of a second domain.     

Modulation of the photocycle of a LOV domain photoreceptor by the hydrogen-bonding network

An extended hydrogen-bonding (HB) network stabilizes the isoalloxazine ring of the flavin mononucleotide (FMN) chromophore within the photosensing LOV domain of blue-light protein receptors, via interactions between the C(2)═O, N(3)H, C(4)═O, and N(5) groups and conserved glutamine and asparagine residues. In this work we studied the influence of the HB network on the efficiency, kinetics, and energetics of a LOV protein photocycle, involving the reversible formation of a FMN-cysteine covalent adduct. The following results were found for mutations of the conserved amino acids N94, N104, and Q123 in the Bacillus subtilis LOV protein YtvA: (i) Increased (N104D, N94D) or strongly reduced (N94A) rate of adduct formation; this latter mutation extends the lifetime of the flavin triplet state, i.e., adduct formation, more than 60-fold, from 2 μs for the wild-type (WT) protein to 129 μs. (ii) Acceleration of the overall photocycle for N94S, N94A, and Q123N, with recovery lifetimes 20, 45, and 85 times faster than for YtvA-WT, respectively. (iii) Slight modifications of FMN spectral features, correlated with the polarization of low-energy transitions. (iv) Strongly reduced (N94S) or suppressed (Q123N) structural volume changes accompanying adduct formation, as determined by optoacoustic spectroscopy. (v) Minor effects on the quantum yield, with the exception of a considerable reduction for Q123N, i.e., 0.22 vs 0.49 for YtvA-WT. The data stress the importance of the HB network in modulating the photocycle of LOV domains, while at the same time establishing a link with functional responses.     

Tryptophan Fluorescence in the Bacillus subtilis Phototropin‐related Protein YtvA as a Marker of Interdomain Interaction¶

The Bacillus subtilis protein YtvA, related to plant phototropins (phot), binds flavin mononucleotide (FMN) within the N‐terminal light, oxygen and voltage (LOV) domain. The blue light‐triggered photocycle of YtvA and phot involves the reversible formation of a covalent photoadduct between FMN and a cysteine (cys) residue. YtvA contains a single tryptophan, W103, localized on the LOV domain and conserved in all phot‐LOV domains. In this study, we show that the fluorescence parameters of W103 in YtvA‐LOV are markedly different from those observed in the full‐length YtvA. The fluorescence quantum yields are ca 0.03 and 0.08, respectively. In YtvA‐LOV, the maximum is redshifted (ca 345 vs 335 nm) and the average fluorescence lifetime shorter (2.7 vs 4.7 ns). These data indicate that W103 is located in a site of tight contact between the two domains of YtvA. In the FMN‐cys adduct, selective excitation of W103 at 295 nm results in minimal changes of the fluorescence parameters with respect to the dark state. On 280 nm excitation, however, there is a detectable decrease in the fluorescence emitted from tyrosines, with concomitant increase in W103 fluorescence. This effect is reversible in the dark and might arise from a light‐regulated energy transfer process from a yet unidentified tyrosine to W103.     

On the binding of BODIPY-GTP by the photosensory protein YtvA from the common soil bacterium Bacillus subtilis

The YtvA protein, which is one of the proteins that comprises the network carrying out the signal transfer inducing the general stress response in Bacillus subtilis, is composed of an N-terminal LOV domain (that binds a flavin [FMN]) and a C-terminal STAS domain. This latter domain shows sequence features typical for a nucleotide (NTP) binding protein. It has been proposed (FEBS Lett., 580 [2006], 3818) that BODIPY-GTP can be used as a reporter for nucleotide binding to this site and that activation of the LOV domain by blue light is reflected in an alteration of the BODIPY-GTP fluorescence. Here we confirm that BODIPY-GTP indeed binds to YtvA, but rather nonspecifically, and not limited to the STAS domain. Blue-light modulation of fluorescence emission of YtvA-bound BODIPY-GTP is observed both in the full-length YtvA protein and in a truncated protein composed of the LOV-domain plus the LOV-STAS linker region (YtvA(1-147)) as a light-induced decrease in fluorescence emission. The isolated LOV domain (i.e. without the linker region) does not show such BODIPY-GTP fluorescence changes. Dialysis experiments have confirmed the blue-light-induced release of BODIPY-GTP from YtvA.     

Natural abundance solution 13C NMR studies of a phototropin with photoinduced polarization

Strongly spin-polarized 13C NMR lines have been observed upon photoexcitation of FMN-binding LOV domains from the blue-light receptor phototropin. Their origin can be rationalized in terms of intermediate radical-pair spin chemistry. Due to hyperfine-selective branching into singlet and triplet products of different lifetime, nuclear spin polarization builds up on nuclei that possess high electron-spin density in the radical state. By examining point-mutated LOV domains of phototropin, spin-polarized 13C NMR signals in emission arising from 13C nuclei at natural abundance in the apoprotein can be unambiguously assigned to a tryptophan residue that is located at a distance of about 14 A from the FMN cofactor and that undergoes photoinduced electron transfer to the flavin. This demonstrates the potential of photo-CIDNP in unraveling reactive intermediates in protein function.     

Three putative photosensory light, oxygen or voltage (LOV) domains with distinct biochemical properties from the filamentous cyanobacterium Anabaena sp. PCC 7120

Light, oxygen or voltage (LOV) domains function as blue-light sensors in the phototropin family of photoreceptors found in plants, algae and bacteria. We detected putative LOV domains (Alr3170-LOV, All2875-LOV and Alr1229-LOV) in the genome of a filamentous cyanobacterium, Anabaena sp. PCC 7120. These cyanobacterial LOV domains are closely clustered with the known LOV domains. Alr3170-LOV and A112875-LOV carry the conserved cysteine residue unique to the photoactive LOV, whereas Alr1229-LOV does not. We expressed these three LOV domains in Escherichia coli and purified them. In fact, Alr3170-LOV and A112875-LOV that are conserved in Nostoc punctiforme, a related species, bound flavin mononucleotide and showed spectral changes unique to known LOV domains on illumination with blue light. Alr3170-LOV was completely photoreduced and dark reversion was slow, whereas A112875-LOV was slowly photoreduced and dark reversion was rapid. For comparison, AvA112875-LOV in a closely related A. variabilis was also studied as a homolog of A112875-LOV. Finally, we observed that Alr1229-LOV that is not conserved in N. punctiforme showed no flavin binding.     

Autophosphorylation, electrophoretic mobility and immunoreaction of oat phototropin 1 under UV and blue Light

Phototropins are UV-A/blue light photoreceptors containing two flavin mononucleotide (FMN)-binding domains, light, oxygen and voltage (LOV)1 and LOV2, of which LOV2 is more sensitive toward light and more important for the physiological response compared with LOV1. Some physiological responses are plant phototropism, chloroplast migration and stomatal opening. Oat phototropin 1 together with light-dependent autophosphorylation shows a reduced electrophoretic mobility and reduced immunoreaction against a heterologous antiserum; both effects were suggested to be caused by phosphorylation at the same sites (M. Salomon, E. Knieb, T. von Zeppelin and W. Rudiger [2003] Biochemistry 42, 4217-4225). In this study, we show that both effects can be separated from each other: at low temperature, reduced immunoreaction preceded the mobility shift, and irradiation with UV-C light led to the mobility shift without the loss of immunoreactivity. We demonstrated that UV-C light at 280 nm, which does not match any absorption maximum of FMN, leads to autophosphorylation of phototropin. It is hypothesized that UV-C light causes differential activation of the LOV domains via energy transfer from aromatic amino acids.     

Primary processes during the light-signal transduction of phototropin

Phototropin is a blue-light photoreceptor in plants that mediates phototropism, chloroplast relocation, stomata opening and leaf expansion. Phototropin molecule has two photoreceptive domains named LOV1 (light-oxygen-voltage) and LOV2 in the N-terminus and a serine/threonine kinase domain in the C-terminus, and acts as a blue light-regulated kinase. Each LOV domain binds a flavin mononucleotide as a chromophore and undergoes unique cyclic reactions upon blue-light absorption that comprises a cysteinyl-flavin adduct formation through a triplet-excited state and a successive adduct break to revert to the initial ground state. The molecular reactions underlying the photocycle are reviewed and one of the probable molecular schemes is presented. Adduct formation alters the secondary protein structure of the LOV domains. This structural change could be transferred to the linker between the kinase domain and involved in the photoregulation of the kinase activity. The structural changes as well as the oligomeric structures seem to differ between LOV1 and LOV2, which may explain the proposed roles of each domain in the photoregulation of the kinase activity. The photoregulation mechanism of phototropin kinase is reviewed and discussed in reference to the regulation mechanism of protein kinase A, which it resembles.     

Molecular dynamics simulation of the LOV2 domain from Adiantum capillus-veneris

The mechanism for signal transduction from the LOV-domains toward the kinase region of phototropin is still not well understood. We have performed molecular dynamics (MD) simulations and CONCOORD calculations on the LOV2 domain of Adiantum capillus-veneris, with the goal to detect possible differences between the two forms of the LOV domain which may not show up in the static crystal structures. Since no such clear differences are found in the MD simulations also, we suggest that the real, biologically active conformation of the LOV domain within the whole phototropin is different from the crystal structure of the isolated LOV domains. The MD simulations do offer, however, insight into details of the dynamics of the dark and illuminated LOV domains, which are discussed in the light of recent experiments.     

Conformational dynamics of phototropin 2 LOV2 domain with the linker upon photoexcitation

Conformational dynamics of LOV2 domain of phototropin, a plant blue light photoreceptor, is studied by the pulsed laser induced transient grating (TG) technique. The TG signal of LOV2 without the linker part to the kinase domain exhibits the thermal grating signal due to the heat releasing from the excited state and a weak population grating by the adduct formation. The diffusion coefficients of the adduct product after forming the chemical bond between the chromophore and Cys residue are found to be slightly smaller than that of the reactant, which implies that the core shrinks slightly on the adduct formation. After that change, no significant conformational change was observed. On the other hand, the signal of LOV2 with the linker part to the kinase domain clearly shows very different diffusion coefficients between the original and the adduct species. The large difference indicates significant global conformational change of the protein moiety upon the adduct formation. More interestingly, the diffusion coefficient is found to be time-dependent in the observation time range. The dynamics representing the global conformational change is a clear indication of a spectral silent intermediate between the excited triplet state and the signaling product. From the temporal profile analysis of the signal, the rate of the conformational change is determined to be 2 ms.     

The primary photophysics of the Avena sativa phototropin 1 LOV2 domain observed with time-resolved emission spectroscopy

The phototropins are blue-light receptors that base their light-dependent action on the reversible formation of a covalent bond between a flavin mononucleotide (FMN) cofactor and a conserved cysteine in light, oxygen or voltage (LOV) domains. The primary reactions of the Avena sativa phototropin 1 LOV2 domain were investigated by means of time-resolved and low-temperature fluorescence spectroscopy. Synchroscan streak camera experiments revealed a fluorescence lifetime of 2.2 ns in LOV2. A weak long-lived component with emission intensity from 600 to 650 nm was assigned to phosphorescence from the reactive FMN triplet state. This observation allowed determination of the LOV2 triplet state energy level at physiological temperature at 16600 cm(-1). FMN dissolved in aqueous solution showed pH-dependent fluorescence lifetimes of 2.7 ns at pH 2 and 3.9-4.1 ns at pH 3-8. Here, too, a weak phosphorescence band was observed. The fluorescence quantum yield of LOV2 increased from 0.13 to 0.41 upon cooling the sample from 293 to 77 K. A pronounced phosphorescence emission around 600 nm was observed in the LOV2 domain between 77 and 120 K in the steady-state emission.     

On the reaction mechanism of adduct formation in LOV domains of the plant blue-light receptor phototropin

The blue-light sensitive photoreceptor, phototropin, is a flavoprotein which regulates the phototropism response of higher plants. The photoinduced triplet state and the photoreactivity of the flavin-mononucleotide (FMN) cofactor in two LOV domains of Avena sativa, Adiantum capillus-veneris, and Chlamydomonas reinhardtii phototropin have been studied by time-resolved electron paramagnetic resonance (EPR) and UV-vis spectroscopy at low temperatures (T < or = 80 K). Differences in the electronic structure of the FMN as reflected by altered zero-field splitting parameters of the triplet state could be correlated with changes in the amino acid composition of the binding pocket in wild-type LOV1 and LOV2 as well as in mutant LOV domains. Even at cryogenic temperatures, time-resolved EPR experiments indicate photoreactivity of the wild-type LOV domains, which was further characterized by UV-vis spectroscopy. Wild-type LOV1 and LOV2 were found to form an adduct between the FMN cofactor and the functional cysteine with a yield of 22% and 68%, respectively. The absorption maximum of the low-temperature photoproduct of wild-type LOV2 is red-shifted by about 15 nm as compared with the FMN C(4a)-cysteinyl adduct formed at room temperature. In light of these observations, we discuss a radical-pair reaction mechanism for the primary photoreaction in LOV domains.     

Perturbation of the ground-state electronic structure of FMN by the conserved cysteine in phototropin LOV2 domains

In LOV2, the blue-light sensitive domain of phototropin, the primary photophysical event involves intersystem crossing (ISC) from the singlet-excited state to the triplet state. The ISC rate is enhanced in LOV2 as compared to flavin mononucleotide (FMN) in solution, which likely results from a heavy-atom effect of a nearby conserved cysteine, C450. Here, we applied fluorescence line narrowing (FLN), resonance Raman (RR) and Fourier-transform infrared (FTIR) spectroscopy to investigate the electronic structure of FMN bound to Avena sativa LOV2 (AsLOV2), its C450A mutant and Adiantum LOV2 (Phy3LOV2). We demonstrate that FLN is the method of choice to obtain accurate vibrational spectra on highly fluorescent flavoproteins. The vibrational spectrum of AsLOV2-C450A showed small but significant shifts with respect to those of wild type AsLOV2 and Phy3LOV2, with a systematic down-shift of Ring I vibrations, upshifts of Ring II and III vibrations and an upshift of the C2=O mode. These trends are similar to those in FMN model systems with an electron-donating group substituted at Ring I, known to induce a quinoid character to the electronic structure of oxidized flavin. Thus, enhancement of the ISC rate in LOV2 is induced through weak electron donation by the cysteine which mixes the FMN pi-electrons with the heavy sulfur orbitals, manifesting itself in a quinoid character of the ground electronic state of oxidized FMN. The proximity of the cysteine to FMN thus not only enables formation of a covalent adduct between FMN and cysteine, but also facilitates the rapid electronic formation of the reactive FMN triplet state.     

Crystal Structure of Human CD47 in Complex with Engineered SIRPα.D1(N80A)

Background: Targeting the CD47/SIRPα signaling pathway represents a novel approach to enhance anti-tumor immunity. However, the crystal structure of the CD47/SIRPα has not been fully studied. This study aims to analyze the structure interface of the complex of CD47 and IMM01, a novel recombinant SIRPα-Fc fusion protein. Methods: IMM01-Fab/CD47 complex was crystalized, and diffraction images were collected. The complex structure was determined by molecular replacement using the program PHASER with the CD47-SIRPαv2 structure (PDB code 2JJT) as a search model. The model was manually built using the COOT program and refined using TLS parameters in REFMAC from the CCP4 program suite. Results: Crystallization and structure determination analysis of the interface of IMM01/CD47 structure demonstrated CD47 surface buried by IMM01. Comparison with the literature structure (PDB ID 2JJT) showed that the interactions of IMM01/CD47 structure are the same. All the hydrogen bonds that appear in the literature structure are also present in the IMM01/CD47 structure. These common hydrogen bonds are stable under different crystal packing styles, suggesting that these hydrogen bonds are important for protein binding. In the structure of human CD47 in complex with human SIRPα, except SER66, the amino acids that form hydrogen bonds are all conserved. Furthermore, comparing with the structure of PDB ID 2JJT, the salt bridge interaction from IMM01/CD47 structure are very similar, except the salt bridge bond between LYS53 in IMM01 and GLU106 in CD47, which only occurs between the B and D chains. However, as the side chain conformation of LYS53 in chain A is slightly different, the salt bridge bond is absent between the A and C chains. At this site between chain A and chain C, there are a salt bridge bond between LYS53 (A) and GLU104 (C) and a salt bridge bond between HIS56 (A) and GLU106 (C) instead. According to the sequence alignment results of SIRPα, SIRPβ and SIRPγ in the literature of PDB ID 2JJT, except ASP100, the amino acids that form common salt bridge bonds are all conserved. Conclusion: Our data demonstrated crystal structure of the IMM01/CD47 complex and provides a structural basis for the structural binding interface and future clinical applications.     

Solid-state photo-CIDNP effect observed in phototropin LOV1-C57S by (13)C magic-angle spinning NMR spectroscopy

Until now, the solid-state photo-CIDNP effect, discovered in 1994 by Zysmilich and McDermott, has been observed selectively in photosynthetic systems. Here we present the first observation of this effect in a nonphotosynthetic system, the blue-light photoreceptor phototropin LOV1-C57S using (13)C magic-angle spinning (MAS) NMR.     

Light‐Regulated Tetracycline Binding to the Tet Repressor

Elucidation of the signal‐transmission pathways between distant sites within proteins is of great importance in medical and bioengineering sciences. The use of optical methods to redesign protein functions is emerging as a general approach for the control of biological systems with high spatiotemporal precision. Here we report the detailed thermodynamic and kinetic characterization of novel chimeric light‐regulated Tet repressor (TetR) switches in which light modulates the TetR function. Light absorbed by flavin mononucleotide (FMN) generates a signal that is transmitted to As‐LOV and YtvA‐LOV fused TetR proteins (LOV=light–oxygen–voltage), in which it alters the binding to tetracycline, the TetR ligand. The engineering of light‐sensing protein modules with TetR is a valuable tool that deepens our understanding of the mechanism of signal transmission within proteins. In addition, the light‐regulated changes of drug binding that we describe here suggest that engineered light‐sensitive proteins may be used for the development of novel therapeutic strategies.     

Characterization of the dynamics of an essential helix in the U1A protein by time-resolved fluorescence measurements

The RNA recognition motif (RRM), one of the most common RNA-binding domains, recognizes single-stranded RNA. A C-terminal helix that undergoes conformational changes upon binding is often an important contributor to RNA recognition. The N-terminal RRM of the U1A protein contains a C-terminal helix (helix C) that interacts with the RNA-binding surface of a beta-sheet in the free protein (closed conformation), but is directed away from this beta-sheet in the complex with RNA (open conformation). The dynamics of helix C in the free protein have been proposed to contribute to binding affinity and specificity. We report here a direct investigation of the dynamics of helix C in the free U1A protein on the nanosecond time scale using time-resolved fluorescence anisotropy. The results indicate that helix C is dynamic on a 2-3 ns time scale within a 20 degrees range of motion. Steady-state fluorescence experiments and molecular dynamics simulations suggest that the dynamical motion of helix C occurs within the closed conformation. Mutation of a residue on the beta-sheet that contacts helix C in the closed conformation dramatically destabilizes the complex (Phe56Ala) and alters the steady-state fluorescence, but not the time-resolved fluorescence anisotropy, of a Trp in helix C. Mutation of Asp90 in the hinge region between helix C and the remainder of the protein to Ala or Gly subtly alters the dynamics of the U1A protein and destabilizes the complex. Together these results show that helix C maintains a dynamic closed conformation that is stable to these targeted protein modifications and does not equilibrate with the open conformation on the nanosecond time scale.     

Thermodynamic and structural basis of phosphorylation-induced disorder-to-order transition in the regulatory light chain of smooth muscle myosin

We have performed molecular dynamics simulations of the phosphorylation domain (PD) of the regulatory light chain (RLC) of smooth muscle myosin, to gain insight into the thermodynamic principles governing the phosphorylation-induced disorder-to-order transition. Simulations were performed in explicit water under near-physiological conditions, starting with an ideal alpha-helix. In the absence of phosphorylation, the helical periodicity of the peptide was disrupted at residues T9-K11, while phosphorylation significantly favored the helical periodicity, in agreement with experimental data. Using the MM/PBSA approach, we calculated a relative free energy of -7.1 kcal/mol for the disorder-to-order transition. A large enthalpic decrease was compensated by a large loss of conformational entropy, despite the small helical increase (no more than three residues) upon phosphorylation. Phosphorylation decreased the conformational dynamics of K and R side chains, especially R16, which forms a salt bridge with pS19. Mutation of R16 to A or E prevented this phosphorylation-dependent ordering. We propose that phosphorylation balances the enthalpy-entropy compensation of the disorder-to-order transition of RLC via short and long-range electrostatic interactions with positively charged residues of the phosphorylation domain. We suggest that this balance is necessary to induce a disorder-to-order conformational change through a subtle energy switching.     

Nanosecond folding dynamics of a three-stranded beta-sheet

We studied conformational stability and folding kinetics of a three-stranded beta-sheet containing two rigid turns. Static infrared measurements indicate that this beta-sheet undergoes a broad but cooperative thermal unfolding transition with a midpoint at approximately 53 degrees C. Interestingly, time-resolved infrared experiments show that its relaxation kinetics in response to a temperature-jump (T-jump) occur on the nanosecond time scale (e.g., the relaxation time is approximately 140 ns at 35.0 degrees C), thereby suggesting that the conformational relaxation encounters only a small free energy barrier or even proceeds in a downhill manner. Further Langevin dynamics simulations suggest that the observed T-jump relaxation kinetics could be modeled by a conformational diffusion process along a single-well free energy profile, which allowed us to determine the effective diffusion constant and also the roughness of the folding energy landscape.