Important amino acid residues of potato plant uncoupling protein ( St UCP )

Chemical modifications were used to identify some of the functionally important amino acid residues of the potato plant uncoupling protein (StUCP). The proton-dependent swelling of potato mitochondria in K+-acetate in the presence of linoleic acid and valinomycin was inhibited by mersalyl (Ki = 5 μM) and other hydrophilic SH reagents such as Thiolyte MB, iodoacetate and 5,5'-dithio-bis-(2-nitrobenzoate), but not by hydrophobic N-ethylmaleimide. This pattern of inhibition by SH reagents was similar to that of brown adipose tissue uncoupling protein (UCP1). As with UCP1, the arginine reagent 2,3-butadione, but not N-ethylmaleimide or other hydrophobic SH reagents, prevented the inhibition of StUCP-mediated transport by ATP in isolated potato mitochondria or with reconstituted StUCP. The results indicate that the most reactive amino acid residues in UCP1 and StUCP are similar, with the exception of N-ethylmaleimide-reactive cysteines in the purine nucleotide-binding site. Correspondence

The protein chemistry of StUCP has not been studied to the same extent as has UCP1.A 32-kDa StUCP has been characterized as a hydrophobic protein which is not retained on hydroxylapatite in the detergent micellar solution (1,6,7).Chemical modifications of reactive amino acid residues, the cleavage pattern produced by proteases, and ligand binding (except for studies with 8azido-ATP (13)) have not been studied in StUCP.In the present study, we examined the effects of several chemical modifiers on StUCP-mediated transport as well as StUCP inhibition by purine nucleotides.Our results clearly show that the pattern of reactive amino acid residues in StUCP is similar to that of UCP1, with the exception that no N-ethylmaleimide (NEM)-reactive cysteines were found in the purine nucleotide-binding site of StUCP.

Isolation of mitochondria and protein determination
Potato mitochondria were isolated as described previously (5,8,9) in medium containing 250 mM sucrose, 10 mM HEPES, pH 7.2, and 0.3 mM EGTA.The protein concentration was 30-40 mg/ml, as determined by the biuret method.A crude fraction was used for swelling studies and for most of the isolations.For some isolations, a Percoll gradient centrifugation was used to remove contamination by plastid proteins, starch and other substances.Qualitatively, transport measurements using the crude fraction gave identical results as those performed with Percoll-purified mitochondria.

Swelling assay of StUCP transport function
Proton-dependent swelling of potato mi-tochondria (0.2 mg protein/ml) in K + -acetate (55 mM K + -acetate, 5 mM K + -HEPES, 0.2 mM Tris-EDTA, 0.1 mM Tris-EGTA, pH 6.9) initiated by valinomycin in the presence of linoleic acid (16 µM) has been used as a standard assay for StUCP-mediated transport (5).Since valinomycin allows the uniport uptake of K + and neutral acetic acid is able to penetrate the lipid bilayer, an efflux of H + is necessary to induce swelling.In our assay, this H + efflux was concomitant with linoleic acid cycling which allowed swelling since StUCP mediated the uptake of linoleic acid anion, while protonated linoleic acid passed spontaneously through the lipid bilayer by a flip-flop mechanism and released H + externally.Hexanesulfonate uniport was assayed as valinomycin-induced swelling in medium containing 51.1 mM Na + -hexanesulfonate, 30.8 mM K + -HEPES, pH 7.2, 190 µM Tris-EDTA and 95 µM Tris-EGTA.The side effects caused by the chemical modifiers used, including the induction of mitochondrial swelling without the addition of ionophore and membrane stiffening, were controlled by performing a swelling assay in K +acetate containing nigericin, which does not depend on protein carriers.When a decrease in this rate (v Nig [c]) was observed at a given concentration [c] of modifier, the rates of valinomycin-induced StUCP-mediated swelling were corrected by multiplying this decrease by the factor v Nig

Chemical modifications of potato mitochondria
For carrying reactions, mitochondria were resuspended in the sucrose isolation medium (5 mg protein/ml) and aliquots of stock solutions (aqueous or in dimethylsulfoxide) of various reagents were added and incubated for 1 h (unless otherwise indicated) at 0 o C. For NEM, DTNB and phenylglyoxal, pH was raised to 8.2 by adding 20 mM Tris-HEPES, pH 8.4, to the stock solution and 2 µM propranolol was added.

Effect of hydrophilic SH reagents on StUCPmediated transport in mitochondria
Proton-dependent swelling of potato mitochondria initiated by valinomycin in K + -acetate containing linoleic acid was reversibly inhibited by the organomercurial SH reagent mersalyl with an apparent K i of 5 µM (Figure 1A, only 10-s preincubations).This type of swelling reflected the ability of StUCP to translocate linoleic acid anions (5).The effect of mersalyl can be considered as a specific inhibition, since swelling independent of a protein carrier, i.e., the nigericin-induced swelling in K + -acetate, was not affected up to 100 µM mersalyl (Figure 1A).Above 100 µM, and above 40 µM in the presence of linoleic acid, mersalyl induced nonspecific permeability changes which were observed as mitochondrial swelling without the ionophore.Some mitochondrial preparations were more sensitive to mersalyl and this made measurements with them more difficult.
To avoid the interference of nonspecific permeability changes, we used Thiolyte MB, a covalently interacting SH modifier.Mitochondria were preincubated for 1 h with increasing Thiolyte MB doses (Figure 2).The IC 50 for Thiolyte MB was around 500 nmol/mg protein.Carrier-independent swelling was not significantly affected by Thiolyte MB, indicating that the modification of the SH groups in StUCP inhibits the transport activity of this protein.Carboxymethylation by iodoacetate (which also affects SH groups) also inhibited StUCP transport activity at higher doses (IC 50 of 100 µmol/mg protein), but only with 10-s preincubations (Figure 1B).Ellmans reagent (DTNB) inhibited the activity by 18 and 31% at 1000 and 3000 nmol/mg protein, respectively, after a 2-h incubation, as calculated from the rates corrected for the nonspecific effect (incubations at pH >8 lead to preswelling after a few hours).In contrast, NEM and other hydrophobic SH reagents (eosinmaleimide, phenylarsineoxide) were not inhibitory up to 10 µmol/mg protein.Hexanesulfonate uniport via StUCP was partially inhibited by hydrophilic SH reagents, e.g., by 1000 nmol Thiolyte MB/mg protein.

Effect of arginine reagents on ATP inhibition of StUCP-mediated transport
Reagents specific for other amino acid residues did not inhibit transport or prevent the inhibition by ATP at doses up to 10 µmol/mg protein.The reagents tested included DIDS, TNBS and lysine-specific pyridoxalphosphate.Only an arginine-specific reagent, 2,3-butadione, completely prevented the inhibition of linoleic acid transport by 4 mM ATP (Figure 3) at doses above 100 nmol/mg protein (see inset in Figure 3).Thus, a 1-h incubation with 4000 nmol/mg protein 2,3-butadione shifted the ATP dose-response curve so that the extrapolated apparent K i was much greater than 10 mM (Figure 3).Surprisingly, phenylglyoxal, a more bulky arginine reagent, had no effect at doses up to 10 µmol/mg protein.NEM, which prevented nucleotide inhibition of UCP, also had no effect on ATP inhibition of StUCP (data not shown).

Confirmation of the effects of 2,3-butadione and Thiolyte MB using reconstituted StUCP
The effect of 2,3-butadione on StUCP reconstituted into proteoliposomes after premodification by 2,3-butadione in mitochondria was identical to that found in potato mitochondria.2,3-Butadione prevented purine nucleotide inhibition of H + efflux, including inhibition by GTP, when the H + efflux was monitored with the fluorescent probe PBFI concomitant with K + influx (Figure 4).The inhibitory effect of Thiolyte MB was also confirmed for isolated StUCP reconstituted into liposomes for which linoleic acid uniport or concomitant H + efflux was detected by TES quenching of the fluorescent probe SPQ (Figure 5).Reconstituted Thiolyte MB-modified StUCP showed no transport activity (Figure 5).

Discussion
The pattern of reactive amino acid residues in StUCP was surprisingly similar to that of mammalian UCP1 (29-33; for reviews, see [17][18][19].This suggests that the structures of StUCP and UCP1 are very likely to be closely related, despite only about 40% identity in their sequences (14,15).
The chemical modification of reactive amino acid residues in proteins has been widely used to study protein structure/function relationships.Site-directed mutagenesis has shown that the identification of a residue as essential for a given function is not a straightforward task.In many cases, the effects of modifiers differ from the phenotypes of the corresponding substitution mutants.Interference by the reagent and/or the mutation with the protein function may indicate that i) the residue is essential for that function, i.e., is involved in the required functional interactions (in this case, the substitution mutants have an identical phenotype), ii) the modification of the residue produces steric hindrances which are the actual cause of the altered function (substitution mutations show no such effect), or iii) the residue is important for maintaining a proper conformation of the protein and cannot retain this position after being modified or mutated.With UCP1, case (i) is valid for its Arg 276, whereas case (ii) has been indicated for its cysteine residues.When Arg 276 was either substituted in a mutated UCP1 protein (34) or modified by phenylglyoxal and 2,3-butadione (32), purine nucleotide binding and gating were absent.Since the proximal third matrix segment was photolabeled at three different positions with 8-azido-, 2-azido-and 3-O-(5-fluoro-2,4-dinitrophenyl) adenosine 5triphosphate (FNDP-ATP) (35), and since the deletion of residues 261-269 resulted in the lack of nucleotide inhibition (36), it was concluded that the main location of the nucleotide-binding site in UCP1 was located between the fifth and sixth transmembrane segments.This site probably forms a waterfilled cavity which penetrates deeply into the membrane close to the opposite surface (35).This cavity in UCP1 is lined with SH residues (C213, C224, C253, C287, C304, and possibly C188).Studies on these residues identified the case (ii) described above, since SH substitution mutants of UCP1 have no disrupted binding or transport (33).
The modification of UCP1 by hydrophobic and hydrophilic SH reagents drastically reduces inhibition by GDP (31).In contrast to UCP1, NEM did not prevent ATP inhibition of transport in StUCP.However, transport was inhibited by the arginine reagent 2,3-butadione.These findings suggest a probable difference between the purine nucleotide-binding site of UCP1 and StUCP and indicate that StUCP does not contain modifiable SH groups at or close to the nucleotidebinding site.Alternatively, SH groups may not be important for maintaining the integrity of StUCP conformation.These findings agree with the amino acid sequence of potato plant UCP (14,16).Thus, C188 of UCP1 is conserved in UCP2 and UCP3, but is substituted by A197 in StUCP (14).Of the two cysteines conserved in the fifth a-helix of UCP1, 2 and 3, the first, C234, is shifted two residues towards the matrix in StUCP such   Hydrophilic, but not hydrophobic, SH reagents were good inhibitors of UCP1-mediated FA-induced H + transport (30).Similarly, in StUCP only hydrophilic SH reagents inhibited StUCP-mediated transport of linoleic acid and hexanesulfonate, while hydrophobic SH reagents, arginine, lysine and other modifiers had no effect.Hence, inhibition by hydrophilic SH reagents is common to StUCP and UCP1.This inhibitory effect on UCP1 has not yet been fully explained.The SH groups which maintain the integrity of the translocation pathway or, alternatively, participate directly in the translocation mechanism, are probably distinct from those interacting with NEM (in UCP1) and interfere with nucleotide binding after modification (31).These SH groups are probably located at yet unknown similar positions in the StUCP sequence.In addition, the type of interference by SH reagents with the StUCP translocation mechanism is likely to be the same as for UCP1.A possible candidate for such a residue is C90, located in the second a-helix of StUCP, which does not have any counterpart in the sequences of UCP1, 2 and 3. Residue C24 of UCP1, absent in StUCP, may serve a similar function for C90 in StUCP. ^

Figure 1 -[Figure 2 -
Figure 1 -Inhibition of protondependent swelling of potato mitochondria by mersalyl (A) and iodoacetic acid (B) in K + -acetate buffer.The inhibition by mersalyl of StUCP-mediated transport (filled circles) and nigericin-mediated, protein-independent swelling (open squares) are specific and nonspecific effects of mersalyl, respectively.The solid line represents the fit of the data using the Hill equation with a Hill coefficient of 2, yielding an apparent K i of 5 µM.B, The iodoacetate dose-response curve, yielding an IC 50 around 100 µmol/mg protein, has already been corrected for the nonspecific effect produced by this compound.The correction and other details of the measurements are described in Material and Methods.

Figure 3 -
Figure3-Prevention of ATP inhibition of StUCP-mediated transport following modification of potato mitochondria with 2,3butadione.The inhibition by ATP of StUCP-mediated proton-dependent swelling in K + -acetate buffer vs log [ATP] is shown for unmodified potato mitochondria (triangles) and mitochondria premodified with 4000 nmol/mg protein 2,3-butadione (diamonds).Inset, Inhibition by 4 mM ATP vs butadione dose in the preincubations.The assay conditions are described in Material and Methods.

Figure 5 -
Figure 5 -Lack of H + efflux in proteoliposomes containing Thiolyte MB-modified StUCP.StUCP from mitochondria treated with Thiolyte MB (1000 nmol/mg protein) were isolated and reconstituted into vesicles (trace a).The response of normal reconstituted StUCP (control) is shown in trace b.H + efflux was monitored by TES quenching of the fluorescent probe SPQ.The addition of 53 µM linoleic acid (LA) caused internal acidification of the vesicles, resulting in the flip-flop of neutral fatty acids into the inner lipid leaflet and subsequent dissociation in the internal medium.StUCP function was seen as an H + efflux (internal alkalinization, indicated by the decrease in SPQ fluorescence), initiated by 1.3 µM valinomycin (val).This efflux was suppressed in Thiolyte MB-modified samples.Vesicles (25 µl per assay) contained 84.4 mM TEA sulfate, 28.85 mM TEA-TES, pH 7.2, ([TEA] was 9.2 mM) and 0.6 mM Tris-EGTA.In the external medium, 84.4 mM K 2 SO 4 replaced TEA sulfate.
in the a-helix is occupied by F231.The second SH (C213 of UCP1) is not conserved in StUCP and is substituted by T220.The similarity of the purine nucleotide-binding site in StUCP and UCP1 is reflected by the effect of 2,3-butadione, which probably interacts with the conserved arginines in UCPs (and in the mitochondrial carrier gene family as a whole), such as R276 of UCP1(37), which corresponds to R281 and R278 in StUCP and AtUCP, respectively(14,15).