Unfolding intermediate in the peroxisomal flavoprotein D-amino acid oxidase

The flavoenzyme D-amino acid oxidase (DAAO) from Rhodotorula gracilis is a peroxisomal enzyme and a prototypical member of the glutathione-reductase family of flavoproteins. DAAO is a stable homodimer with a FAD molecule tightly bound to each 40 kDa subunit. In this work the urea-induced unfolding of dimeric DAAO was compared with that of a monomeric form of the same protein, a deleted dimerization loop mutant. By using circular dichroism spectroscopy, protein and flavin fluorescence, 1-8anilinonaphtalene sulfonic acid (ANS) binding and activity assays, we demonstrated that the urea-induced unfolding of DAAO is a three state process, yielding an intermediate, and that this process is reversible. The intermediate specie lacks the catalytic activity and the characteristic tertiary structure of native DAAO, but has significant secondary structure and retains flavin binding. Unfolding of DAAO proceeds through formation of an expanded, partially unfolded inactive intermediate, characterized by low solubility, by increased exposure of hydrophobic surfaces, and by increased sensitivity to trypsin of the β -strand F5 belonging to the FAD-binding domain. The oligomeric state does not modify the inferred folding process. The strand F5 is in contact with the C-terminal α -helix containing the Ser-Lys-Leu sequence corresponding to the type 1 peroxisomal targeting signal, and this structural element interacts with the N-terminal βαβ flavin binding motif (Rossmann fold). The expanded conformation of the folding intermediate (and in particular the higher disorder of the mentioned secondary structure elements) could match the structure of the inactive holoenzyme required for in vivo trafficking of DAAO through the peroxisomal membrane. electrode assay on protein samples incubated at 15 °C for 40 in the presence of different concentrations of urea. Data were fitted to Eq. 4. Inset. Refolding reaction: the activity was determined on protein samples unfolded as above and refolded by 10-fold dilution in 50 mM potassium phosphate, pH 7.5, 2 mM EDTA, 5% glycerol, for 10 min (open or for 24 hours (filled symbols) at 15 °C. The urea concentration reported is that in the refolding solution, after dilution. The point at 3 M urea was obtained by a 10-fold dilution of a protein sample at 8 M urea with a 2.75 M urea containing refolding buffer.

Unfolding of D-amino acid oxidase 3 Protein folding/unfolding is a highly cooperative process. It has been shown that the folding/unfolding of small globular proteins occurs via a two-state process, whereas the folding/unfolding of larger proteins (> 100 amino acids) is complex and often involves the formation of intermediate(s) (1)(2)(3). The most thorough investigations of protein folding and stability have been done with unusually small proteins, which are folded into single domains and display simple two-state unfolding processes. It is of interest, however, to extend studies to larger, more complex, and therefore more typical proteins. D-amino acid oxidase (EC 1.4.3.3, DAAO) has attracted our attention as a suitably more complex subject because it is considered the paradigm of the dehydrogenase/oxidase class of flavoproteins (4), and in particular of those in which the flavin is non-covalently bound. In fact, many proteins in nature require the non-covalent binding of cofactors to perform their biological activity, and these molecules fold in a cellular environment where their cognate cofactors are present.
However, the manner in which cofactors affect the folding pathway remains poorly understood, because kinetic folding studies are frequently conducted in the absence of potentially complicating ligands. Furthermore, flavoproteins are often multi-subunit proteins constituted either by identical or by different polypeptide chains. Up to now, deep and complete investigations have been restricted to small flavoproteins, such as flavodoxin (5).
To have insights on the relationships between cofactor uptake, the stable interaction between identical subunits (and on its significance), folding, and intracellular trafficking, we undertook a study of the stability of structural elements in the peroxisomal flavoenzyme DAAO from the yeast Rhodotorula gracilis. Proteins destined for the peroxisome are synthesised on free ribosomes of the cytoplasm and transported into peroxisomes posttranslationally. There are two types of peroxisomal targeting signals (PTS): the first one (PTS1) is the SKL C-terminal sequence (or some conservative variants of it) (6,7). Proteins belonging to this group are made at their mature size, do not undergo cleavage of the targeting by guest on March 23, 2020 http://www.jbc.org/ Downloaded from Unfolding of D-amino acid oxidase 4 sequence upon transport into peroxisome, and have cytosolic receptors which mediate their association with the peroxisomal import machinery (6,7). It has also demonstrated that stably folded proteins are substrates for peroxisomal import (8).
DAAO catalyses the dehydrogenation of D-isomers of amino acids, to give the corresponding α-keto acids, ammonia and hydrogen peroxide. During the years DAAOs have been the object of extensive investigation (9,10). In solution, DAAO from R. gracilis is a stable 80 kDa homodimer, with a molecule of FAD tightly (K d = 2 × 10 -8 M), but noncovalently bound to each 40 kDa subunit. The 3D structure of DAAO has been resolved at very high resolution allowing to find the rationale of its high catalytic efficiency (11,12). In the "side to tail" model of monomer-monomer interaction (with a high buried surface area, 3049 Å 2 ) (12), a large contribution to the interaction between monomers is given by a long (21 amino acids) loop connecting β-strands F5 and F6, that is unique to yeast DAAO.
Recently, by rational design and site directed mutagenesis, a stable monomeric holoenzyme form of DAAO has been obtained by partial elimination of this loop (from Ser 308 to Lys 321 : SPLSLGRGSARAAK) (13). The ∆loop mutant DAAO shows slightly altered spectral and kinetic properties, a lower temperature stability and a 5-fold increase in the K d for FAD binding compared to the wild-type enzyme. We also demonstrated the possibility of obtaining a monomeric form of yeast DAAO by treatment with 0.5 M NH 4 SCN, without deletion of the βF5-βF6 loop (14), as well as by removal of the coenzyme to yield the corresponding apoprotein (10). This latter result suggests a structural relationship between the FADharboring domain and the regions involved in dimerization. Recently, temperature ramp experiments following different probes allowed the identification of a clear sequence of events in the course of thermal unfolding of wild-type and ∆loop protein forms (14).
Apparently, a first, low-temperature energetic domain relates to the unfolding of tertiary by guest on March 23, 2020 http://www.jbc.org/

Downloaded from
Unfolding of D-amino acid oxidase 5 structure regions, whereas a second energetic domain relates to the loss of secondary structure elements and to the release of the cofactor at higher temperatures.
In this paper, we expand the results obtained until now, and report a novel, reversible step in the urea-induced unfolding pathway of R. gracilis DAAO, that we believe may be relevant to intracellular trafficking of DAAO. Furthermore, to provide a basis for understanding the structure-function relationships in flavoproteins and the determinants of their stability, we have attempted to compare the chemical unfolding of dimeric DAAO with that of a monomeric form obtained by site directed mutagenesis.

EXPERIMENTAL PROCEDURES
Materials -Recombinant wild-type and ∆loop mutant DAAO were expressed and purified from E. coli cells as described previously (13,15). Starting from a 10-liters fermentation broth, 180 mg and 80 mg of pure enzyme with a specific activity of 110 and 86 U/mg protein were obtained for wild-type and ∆loop DAAO, respectively. The enzyme concentration was determined by using extinction coefficients at 455 nm of 12.6 mM -1 cm -1 for wild-type and 11.3 mM -1 cm -1 for ∆loop DAAOs (10,13). Urea was from Pierce Chemical Co., and the other reagents were of analytical grade.
DAAO activity assay -DAAO activity was assayed with an oxygen electrode at pH 8.5, air saturation, and 25 °C, using 28 mM D-alanine as substrate in the presence of 0.2 mM FAD (10). The effect of urea concentration on enzyme activity of DAAO was determined using the oxygen-electrode assay on protein samples previously incubated at 15 °C for 40 min in the presence of different concentrations of urea.  Spectroscopy -All fluorescence measurements were performed by using a 1-ml cell in a Jasco FP-750 instrument equipped with a thermostated cell holder. Tryptophan emission spectra were taken from 300 to 400 nm using excitation wavelengths of 280 nm and 298 nm.
Flavin emission spectra were recorded from 475 to 600 nm using an excitation wavelength of 450 nm; 10 and 20 nm bandwidths were used for excitation and emission, respectively.
Steady-state fluorescence measurements were performed at 15 °C, and at 0.02 mg/ml protein concentration. All spectra were corrected by subtracting the emission of the buffer.  Data analysis -Unfolding curves were usually analyzed using a two-state mechanism.
Unfolding curves for the N↔D transition were normalized to the apparent fraction of the unfolding form, F D , using the following equation (16): where Y is the observed variable parameter, and Y N and Y U are the values characteristic of the native and fully unfolded conformations, respectively. The difference in free energy between the folded and the unfolded state, ∆G, was calculated by the following equation: where K is the equilibrium constant, R is the gas constant, and T is the absolute temperature.
The data were analyzed assuming the free energy of unfolding or refolding, ∆G, to be linearly dependent on the urea concentration (denoted here by C), as described in detail previously (17): in which ∆G w and ∆G represent the free energy of unfolding or refolding in the absence and presence of urea, respectively; C m is the midpoint concentration of urea required for unfolding or refolding; and m stands for the slope of the unfolding or refolding curve at C m . A leastsquares curve fitting analysis was used to calculate the values of ∆G w , m, and C m by a The urea-unfolding curves corresponding to a two-state model were analysed using Eq. (4), derived by (18), that incorporates Eq. (1-3): 1 + e -(∆Gw -mC)/RT The analysis of the equilibrium unfolding transition using the flavin fluorescence data was performed according to a three-state denaturation pathway (N ↔ I ↔ U) according (19) in which C m1 and C m2 are the midpoint concentration of urea for the N ↔ I and I ↔ U transitions, respectively; F N-I and F I-U represent the percentage change of flavin fluorescence associated to the N ↔ I and I ↔ U transitions, respectively; exponents n 1 and n 2 reflect the steepness of the transition between states as a function of urea concentration.  (20). Mass spectra were elaborated using the software MassLynx 2.0, furnished with the spectrometer. Mass values were reported as average masses.  Table I). For both proteins, unfolding by urea is accompanied by an increase in both flavin and tryptophan fluorescence.

Spectral properties of wild-type and ∆loop
Far-UV CD spectra of monomeric and dimeric DAAOs did not reveal any major difference in the features related to the secondary structure of the two proteins while the near-UV CD spectra of the wild-type and ∆loop mutant DAAO were different. The differences have been ascribed to a different contribution from aromatic amino acid residues, which are responsible for most transitions in the near-UV spectral region. Since the deleted portion does not contain aromatic residues, the different spectroscopic features of the two proteins have been explained in terms of an altered mutual relationships between nearby structural elements (14). Following the addition of ≥ 6 M urea, the Far-and Near-UV spectra of both wild-type and ∆loop DAAOs were superimposible.
Equilibrium unfolding studies -The stability of wild-type and ∆loop DAAO forms was studied at first by equilibrium unfolding measurements. In the experiments reported in the following sub-sections, different spectroscopic signals, catalytic activity, and the associative behaviour of both proteins were monitored after equilibration in the presence of increasing urea concentration. nm at 5-6 M urea, a change that does not parallel the change in fluorescence intensity (Fig. 1).
The fluorescence red shift stems from transfer of tryptophan side chains to a more polar environment upon protein unfolding (21).
Plots of changes in the intensity protein fluorescence at equilibrium as a function of the urea concentration apparently suggest a simple two-state transition (Fig. 1). The free energy of unfolding, ∆G, was calculated according to Eq. 2. The free energy of unfolding in the absence of the denaturant (∆G w ) can be obtained by extrapolation of ∆G to zero denaturant concentration by using Eq. 3. The energy value determined for the monomeric ∆loop enzyme is lower than the value for wild-type DAAO ( Table I) Table I.  (Table I) Table I).  Table I).
Associative behavior -The effects of urea on the oligomerization state and/or the molecular size of wild-type and ∆loop DAAOs were monitored by size-exclusion chromatography (Fig. 5). In the absence of denaturant, the two enzymes were, respectively, fully monomeric (∆loop, Ve = 14.8 ml) and fully dimeric (wild-type, Ve = 13.4 ml).
However, upon pre-incubation with urea, both DAAO forms eluted as multiple peaks at In fact, return to tryptophan fluorescence values appropriate for the urea concentration present after dilution took about 24 hours for both DAAO forms (see Fig. 7A for wild-type DAAO).
Changes in flavin fluorescence showed a different behaviour. Still the largest part of the decrease in fluorescence was still observed during the experiment dead-time, but the final absolute values (at ≥ 24 hours) were higher than those observed during protein unfolding experiments at the urea concentration present after dilution. This indicates that cofactor uptake during the refolding process was somewhat impaired (see Fig. 7B).
All together, these results show that the refolding of urea-treated wild-type and ∆loop DAAOs is reversible considering the tertiary protein structure, although some alterations are evident between the native and refolded DAAO in the microenvironment surrounding the flavin cofactor.
Following the unfolding of wild-type DAAO with 2 M urea (at which ~ 80% of the initial activity is lost, see Fig. 4), and its refolding at 0.2 M urea for 10 min at 15 °C, the refolded enzyme showed ≥ 90% of the initial specific activity. Under comparable conditions, the recovery of enzymatic activity was lower for the ∆loop mutant (Fig. 4, inset). Starting from fully denatured (8 M urea) wild-type and ∆loop DAAOs, about 30% and 15% of the initial activity was recovered, respectively, after 10 min of refolding at 0.8 M urea. Activity recovery figures increased significantly when refolding was allowed to occur for longer times, surface (see Fig. 8). This site is not accessible to attack by trypsin under native conditions, and becomes exposed (and therefore sensitive to proteolysis) in the partially unfolded Chemical denaturation experiments show that most of the characteristic tertiary structure is destroyed at 3 M urea, where most of the secondary structure is almost retained.
Therefore, at 2-3 M urea, a stable equilibrium intermediates are formed. These protein forms still binds the flavin cofactor, although with a significant increase in dissociation constant, but their catalytic activity is totally lost. A model for the equilibrium unfolding of DAAO based on a three-state model is shown in Fig. 9. Folding intermediates with these characteristics, often referred to as molten globule-like species, have been reported for a few proteins (27).
Our ANS binding studies confirm that the unfolding intermediate observed at 2-3 M urea has exposed hydrophobic patches on structural sites that are accessible to ANS in both the wild- This may be consequent to the exposure of "sticky" surface hydrophobic sites/patches, as detected by ANS binding studies, that favor hydrophobic interaction between surfaces belonging to distinct proteins (23). At increasing urea concentration, the chaotropic effect of this molecule disrupts the solvating water structure of DAAO and leads to stabilization of an expanded, fully unfolded, and soluble multimeric apoprotein aggregate (see Fig. 9). Sizeexclusion chromatography shows that both wild-type and ∆loop DAAOs at > 3 M urea are in a multimeric conformation, thus the intermediate should correspond to aggregated molten globules. An expansion of the protein before full denaturation has been reported for several multimeric enzymes (see (28), and references therein). In spite of the formation of aggregates, unfolding of both DAAO forms may be considered as reversible. Only slight alterations in the flavin binding regions are indicative of differences between the native and the refolded proteins. The reversibility of DAAO unfolding is a prerequisite to enable the future application of hydrogen exchange pulse labeling methods in combination with NMR spectroscopy or mass spectroscopy, and real time NMR experiments to study DAAO folding in atomic detail.
A further conclusion of the present study is that the dimerization increases the structural stability of DAAO. All our equilibrium spectroscopic measurements show that the urea concentration required for the unfolding of monomeric ∆loop mutant is lower than that of dimeric wild-type DAAO, and that different urea concentrations are required for complete unfolding of each protein (see Table I). It should be noted that the region susceptible to trypsin cleavage in the unfolding intermediate is far away from the monomer-monomer interaction area. Previous studies indicated a significant lower thermal stability of the ∆loop mutant with respect to wild-type DAAO (13). Thermal unfolding of both DAAO forms was characterized by two distinct steps (14) (29). The C-terminal α-helix is highly susceptible to proteolysis (24,25) and contains the SKL sequence corresponding to the PTS1 signal for peroxisomal import (see Fig. 8).
Recently, it has been demonstrated that at least 12 C-terminal residues of a given substrate protein are implicated in PTS1 signal recognition: 1) the Cterminal tripeptide, 2) four residues upstream interacting with the surface of Pex5 receptor, and 3) a polar, solvent-accessible and unstructured region (five residues) with linker function (30). The expanded structure of the C-terminal sequence in the DAAO folding intermediate could represent the optimal conformation for interaction of DAAO with its cognate PTS receptor during in vivo trafficking to peroxisome.
The investigations of the folding process of DAAO should give also a clue as for the rules that govern folding of other flavoproteins, an delivery of peroxisomal proteins to the peroxisome in an "inactive and partially folded state". Recent experimental evidence demonstrated that proteins which are incapable of assuming their native conformation are also substrates for peroxisomal import (31). When combined with previous reports demonstrating the import of stably folded proteins (8), these results were used to support the model in which tertiary structure was immaterial as for protein import into the peroxisomal matrix (31).
Such a conclusion may be not true in the case of "dangerous" proteins such as DAAO.  Values similar to those reported were also obtained using the protein fluorescence values determined following the excitation at 298 nm. In parenthesis are reported the values determined following the shift of protein fluorescence emission maximum (see Fig. 1) b These values have been estimated using a three-state model according to (19).