Roberto Favilla

Reversible unfolding of bovine odorant binding protein induced by guanidinium hydrochloride at neutral pH.

An analysis of the unfolding and refolding curves at equilibrium of dimeric bovine odorant binding protein (bOBP) has been performed. Unfolding induced by guanidinium chloride (GdnHCl) is completely reversible as far as structure and ligand binding capacity are concerned. The transition curves, as obtained by fluorescence and ellipticity measurements, are very similar and have the same protein concentration-independent midpoint (C1/2 approximately 2.6 M). This result implies a sequential, rather than a concerted, unfolding mechanism, with the involvement of an intermediate. However, since it has not been detected, this intermediate must be present in small amounts or have the same optical properties of either native or denatured protein. The thermodynamic best fit parameters, obtained according to a simple two-state model, are: deltaG degrees un,w = 5.0 +/- 0.6 kcal mol(-1), m = 1.9 +/- 0.2 kcal mol(-1) M(-1) and C1/2 = 2.6 +/- 0.1 M. The presence of the ligand dihydromyrcenol has a stabilising effect against unfolding by GdnHCl, with an extrapolated deltaG degrees un,w of 22.2 +/- 0.9 kcal mol(-1), a cooperative index of 3.2 +/- 0.3 and a midpoint of 4.6 +/- 0.4 M. The refolding curves, recorded after 24 h from dilution of denaturant are not yet at equilibrium: they show an apparently lower midpoint (C1/2 = 2.2 M), but tend to overlap the unfolding curve after several days. In contrast to chromatographic unfolding data, which fail to reveal the presence of folded intermediates, chromatographic refolding data as a function of time clearly show a rapid formation of folded monomers, followed by a slower step leading to folded dimers. Therefore, according to this result, we believe that the preferential unfolding/refolding mechanism is one in which dimer dissociation occurs before unfolding rather than the reverse.

Alkaline denaturation and partial refolding of pepsin investigated with DAPI as an extrinsic probe.

The binding parameters of DAPI to porcine stomach pepsin have been described in the previous article in this issue (A. Mazzini et al.). Here we exploit the differences in the spectroscopic (fluorescence and circular dichroism) properties of DAPI bound to either native or alkali denatured pepsin. We follow the kinetics of pepsin denaturation around neutrality (pH range 6.8-7.4), at several phosphate buffer ionic strengths (range 0.02-0.25). The dependence of the apparent dissociation rate constant on pH clearly shows that the rate limiting step follows the dissociation of about three acidic protein residues. The accelerating effect by ionic strength we observed can be accounted for by a simple treatment based on both transition state theory and Debye-Hueckels limiting law. Furthermore, when a solution of pepsin, rapidly denatured at pH 7, is reacidified to a pH between 4.5 and 5.5, a substantial recovery of protein secondary structure, with no enzymatic activity, is observed, judging by the far UV circular dichroism of the protein. This process of partial refolding can easily be followed using DAPI as an extrinsic reporter group, able to monitor the kinetics of formation and decay of a highly fluorescent intermediate. This process becomes faster at a lower pH, at least in the limited range investigated (pH 4.5-5.5), in which the refolded protein does not aggregate, but, in contrast to unfolding, is almost independent in ionic strength.

Guanidinium chloride induced unfolding of a hemocyanin subunit from Carcinus aestuarii

The effects of guanidinium hydrochloride (GuHCl) on the functional and structural properties of a 75-kDa, functionally active hemocyanin (Hc) subunit isolated from the crab Carcinus aestuarii (holo-CaeSS2) were investigated. The holo form of the protein contains two copper ions in the active site and is capable of reversibly binding dioxygen. The present results are compared with those previously described for the corresponding functionally inactive subunit (apo-CaeSS2), devoid of the two active site copper ions (accompanying paper [R. Favilla, M. Goldoni, A. Mazzini, M. Beltramini, P. Di Muro, B. Salvato, paper published in this issue]). As with apo-CaeSS2, both equilibrium and kinetic unfolding measurements were carried out using light scattering (LS), circular dichroism, intrinsic and extrinsic fluorescence (IF and EF, respectively). In addition here, absorbance spectroscopy was exploited to evaluate oxygen binding by holo-CaeSS2. These data were combined with those relative to the protein copper content to obtain information on the stability of the active site as a function of denaturant concentration. The different techniques used revealed several unfolding transitions. At GuHCl<1 M, an appreciable increase of LS intensity was observed, about an order of magnitude lower than with apo-CaeSS2, suggesting some reversible protein aggregation. A first cooperative transition as a function of GuHCl was detected as an increase of intensity of the protein IF (C(1/2)=1 M), followed by a second transition, characterised by a small intensity decrease and a red shift of the emission maximum (C(1/2)=1.4 M). Cooperative transitions with C(1/2) values near 1.4 M GuHCl were also detected by following the decrement of: (a) EF intensity of anilino-1-naphtalenesulphonate (ANS) bound to the protein; (b) absorbance at 340 nm, typical of oxy holo-CaeSS2; (c) copper-to-protein stoichiometry. A transition at higher GuHCl (C(1/2)=1.8 M) was also observed by far UV circular dichroism (far UV CD) and related to global unfolding. Unfolding kinetics was also studied using the fluorescence stopped-flow technique. All traces were best fitted by a sum of three or four exponential terms, depending on GuHCl concentration. A comprehensive unfolding model is proposed, which accounts for most of the complex behaviour of this protein subunit, including oxy and deoxy native and aggregation-prone intermediates, a highly fluorescent intermediate, molten globule-like apo and unfolded species.

Fluorescence quenching of tryptophan and related compounds by hydrogen peroxide.

The fluorescence of tryptophan and some related derivatives was found to be quenched by hydrogen peroxide. The quenching mechanism was shown to be essentially dynamic in nature, without any ground-state complex formation, and it is interpreted as resulting through an electron transfer from the excited indole ring to hydrogen peroxide.

The binding of 1,N6-etheno-nad to bovine liver glutamate dehydrogenase

The binding of 1,N6-etheno-NAD (epsilon NAD) to bovine liver glutamate dehydrogenase (L-glutamate:NAD(P)+ oxidoreductase (deaminating), EC 1.4.1.3) saturated with glutarate has been investigated at pH 7.0, 0.05 M phosphate buffer at 20 degrees C, by fluorescence titrations. epsilon NAD binds to the protein in a simple fashion: one molecule of coenzyme per enzyme polypeptide chain in the range of enzyme concentrations investigated (from above 50 to a few micromoles of enzyme polypeptide chains/liter). The fluorescence enhancement factor, Q, of bound epsilon NAD relative to free epsilon NAD is independent of the saturation degree, as deduced from the constant value of the long fluorescence decay lifetime (about 21 ns), and is about 17, as deduced from Fmax/F0 ratio values obtained after extrapolation from double reciprocal plots of 1/delta F vs. 1/[glutamate dehydrogenase]. This value for the Q factor is also independent of enzyme concentration, as well as of the presence of either GTP or ADP. At low enzyme concentrations (below 20 mumol polypeptide chains/liter), the dissociation constant of epsilon NAD increases progressively from a plateau value of about 50 microM to about 100 microM at infinite dilution. This is interpreted as being due to a minor affinity of glutamate dehydrogenase hexamers, with respect to higher aggregation states of the enzyme, towards epsilon NAD. As expected, GTP and ADP change the affinity of glutamate dehydrogenase towards epsilon NAD in an opposite manner: GTP strongly increases it, whereas ADP strongly decreases it (Kappd around 6 microM with saturating GTP and around 300 microM with saturating ADP). Furthermore, in the case of GTP, both GTP and epsilon NAD bind to glutamate dehydrogenase with positive cooperativity, with a Hill coefficient of approx. 1.8 for both and a Kappd approximately equal to 30 microM for the binding of GTP to glutamate dehydrogenase saturated with epsilon NAD and glutarate. The value of the Q factor remains the same, even in the presence of the effectors (again from lifetime measurements), as well as the number of epsilon NAD binding sites per enzyme polypeptide chain. These results are interpreted in terms of independent active sites, in the case without effectors. With ADP the binding appears to be simple, but no careful investigation has been attempted at low enzyme concentrations because of the low saturation degree achievable, whereas with GTP the cooperativity can be explained as due to a shift towards hexamers from higher aggregation states.

Peroxidase activity of horse liver alcohol dehydrogenase in the presence of β-NAD+ and hydrogen peroxide

Horse liver alcohol dehydrogenase (LADH) has been extensively studied, mainly with respect to its catalytic ability to interconvert alcohols to aldehydes, but its main physiological function appears to be still unknown, since ethanol is not formed in the animal’s body. During the last twenty years, other alternative activities have been found for LADH; the first, cronologically speaking, concerns the dismutation reaction of aldehydes [ 1,2] . An isomerase activity of LADH, concerning the isomerization of several aldehydes to ketones, was found first by Van Eys [3] and later allosterically correlated with the dehydrogenase activity [4] . In 1963 Waksman and Roberts [S] described a transaminase activity for yeast alcohol dehydrogenase and other dehydrogenases. A steroid activity of LADH, discovered in 1960 [6] , could be solely ascribed to the more basic subfraction of the enzyme preparation only in 1966 [7] and later associated to an isoenzyme of LADH [8]. It has also been found that vitamin A is oxidized to retinene by LADH in the retina [9a,9b] and more recently that methanol can function as substrate [lo]. In this paper we report evidence for a peroxidase activity of LADH, which originates when hydrogen peroxide is added to a solution containing /3-NAD’ and LADH.

Synchrotron SAXS studies on the structural stability of Carcinus aestuarii hemocyanin in solution.

The effect of GuHCl and of NaCl on the structural properties of the hemocyanin (Hc) from Carcinus aestuarii has been studied by small angle x-ray scattering (SAXS) using synchrotron radiation. SAXS data collected as a function of perturbant concentration have been used to analyze conformational states of hexameric holo and apoHc as well as the holo and apoforms of the monomeric subunit CaeSS2. In the case of the holoprotein in GuHCl, two concentration domains were identified: at lower concentration, the perturbant induces aggregation of Hc molecules, whereas at higher concentration the aggregates dissociate with concomitant denaturation of the protein. In contrast, with apoHc the denaturation occurs at rather low GuHCl, pointing to an important effect of the active site bound copper for the stabilization of Hc tertiary structure. The effects of NaCl are similar to those of GuHCl as far as CaeSS2 is concerned, namely oligomerization precedes denaturation, whereas in the case of the hexameric form no aggregation occurs. To improve data analysis, on the basis of the current models for Hc monomers and oligomers, the fraction of each aggregation state and/or unfolded protein has been determined by fitting experimental SAXS curves with form factors calculated from Monte Carlo methods. In addition, a global analysis has been carried out on the basis of a thermodynamic model involving an equilibrium between a monomer in a nativelike and denatured form as well as a class of equilibria among the monomer and other aggregates.

Dissociation kinetics of hemocyanin from Octopus vulgaris

The native form of hemocyanin (Hc) from Octopus vulgaris can be completely dissociated, at alkaline pH and in the presence of EDTA, from 49S decamers to 11S monomers. The kinetics of this process was studied, using a Bio-Logic stopped flow system, by following the time dependence of the 450-nm light intensity, scattered at 90 degrees, in the 7.9-8.8 pH range. All experimental traces were best fitted by a sum of three exponential decay functions. We then tried to best fit these decay functions with a series of kinetic models, the best of them resulting in one whose dissociation of decamers to monomers takes place in three consecutive and irreversible steps, with a highly cooperative step concerning dissociation of octamers to dimers, which appears to be the only intermediate species. This model was preferred over several others, not only for the best norm value but also for the best accordance between each calculated and experimental kinetic parameter (rate constants and amplitudes). Although other more complex models may be considered, our best fit model represents the simplest one, which is able to describe the observed dissociation kinetics.

Fluorescence stopped-flow studies on the binding of 1,N6-etheno-NAD to bovine liver glutamate dehydrogenase

Fluorescence stopped-flow techniques have been used to investigate the binding of the oxidised coenzyme eNAD to bovine liver glutamate dehydrogenase (L-glutamate:NAD(P)+ oxidoreductase (deaminating), EC 1.4.1.3) saturated with glutarate, a substrate analogue, by following the transient kinetics of fluorescence intensity changes associated with changes in the binding of 1,N6-etheno-NAD (eNAD) to the enzyme, using displacement by NAD, NADP, ADP or GTP. Computer simulations of the various kinetic models provide a detailed picture of the molecular interactions between the active site (site I) and regulatory sites (sites II and III), specific for adenine and guanine nucleotides, respectively. The observed enhancement of the eNAD dissociation rate constant from site I can satisfactorily be accounted for as being due to the effect of ADP or NAD (and to a lesser extent NADP) binding to site II. This provides a mechanism for the allosteric activation of this enzyme via a predominantly intrasubunit interaction. By contrast the isomerisation of the enzyme induced by ADP alone is markedly slowed down by the occupancy of site I by eNAD in the presence of glutarate. The inhibitory effect of the allosteric effector GTP correlates with a tightening of eNAD binding, causing a decrease of the coenzyme dissociation rate constant followed by a slow isomerisation of the enzyme complexed with eNAD and glutarate.

The binding of 1,N6-ethenoNAD to bovine liver glutamate dehydrogenase: studies using the time-correlated single photon counting fluorescence technique

The time-correlated single photon counting (TCPC) fluorescence technique has been used as a novel approach to investigate ligand-protein interaction, for the case of the binding of the fluorescent coenzyme analogue 1,N6-ethenoNAD (epsilon NAD) to bovine liver glutamate dehydrogenase in the presence of glutarate, a substrate analogue which stabilizes the complex. System calibration was performed using solutions of epsilon ADP and carefully purified epsilon NAD mixed at variable molar ratios (pH 7.0, 0.05 M sodium phosphate buffer, 20 degrees C). The fluorescence lifetimes obtained after deconvolution were 2.4 ns (for epsilon NAD) and 23 ns (for epsilon ADP), in good agreement with literature values obtained under similar conditions. epsilon NAD binds to glutamate dehydrogenase in the presence of 50 mM glutarate, with a fluorescence quantum yield enhancement factor, Q, of about 17-fold, as previously reported (Favilla, R. and Mazzini, A. (1984) Biochim. Biophys. Acta 48-57). For this system, fluorescence lifetime values were obtained after deconvolution as 2.4 ns for free epsilon NAD and 21 ns for bound epsilon NAD. These values did not vary appreciably with enzyme concentration nor with degree of saturation, thus reflecting the existence of only one spectroscopically relevant type of complex. Addition of either GTP or ADP did not affect the lifetime of epsilon NAD bound to the enzyme, but only its affinity, thus allowing calculations of binding strengths. In the case of a simple binding (i.e., in the absence of GTP) the dissociation constant of the complex could be derived from a simple relationship, in which only the ratio between the pre-exponential factors and the parameter gamma, which represents the molar fraction of epsilon NAD molecules free in solution in the open conformation, are to be taken into account. The results are in good agreement with those reported by some of us (reference above) using a steady-state fluorescence technique, which by itself is, however, unable to resolve the number of relevant species present in the system.