Biosynthesis of porphyrins in Rhodopseudomonas palustris--VI. The effect of metals, thiols and other reagents on the activity of uroporphyrinogen decarboxylase.

The effect of several metals and reagents on the decarboxylation rate of uroporphyrinogen I by using a 16-fold purified preparation of Uroporphyrinogen Decarboxylase from Rhodopseudomonas palustris, was studied. 1 mM Hg2+ and Cu2+ were strong inhibitors, 1 mM Zn2+ and Fe2+ under certain conditions and 1 mM Fe3+ and Cr3+ also inactivated the enzyme, but Pb2+, Cd2+ and Al3+ did not. Metals inhibition was reversed by 1 mM GSH or CySH. 0.1 mM DTNB and PCMB, 1 mM pyridoxal phosphate and 100 mM chloral hydrate, as well as 1 mM 2-methoxy-5-nitrotropone and 0.2 mM diethylpyrocarbonate inhibited Uroporphyrinogen Decarboxylase; while GSH, CySH, N-ethylmaleimide, sodium thioglycolate, 1,4-dithioerythritol, EDTA and O-phenantroline did not modify activity. Data obtained would indicate that one cysteine, one or two histidine residues and probably a lysine group are required for enzyme activity.

activity in porphyria cutanea tarda (PCT) , research in this area has attracted great interest. Uro-D from Rhodopseudomonas palustris (Koopmann et al., 1986) exhibits a number of properties similar to the enzyme from other sources; but it also shows some characteristics which are rather unique, making this organism very convenient to carry out mechanistic studies. As it was found in all sources so far examined, decarboxylation of urogen is a two stage process; however, with the Pp. palustris enzyme, a differential behaviour towards isomers III and I of the urogen was observed. Decarboxylation of heptaporphyrinogen is the rate-limiting step when urogen III is the substrate, while removal of the final carboxyl group of pentaporphyrinogen becomes the rate controlling step when urogen I is the substrate, with practically no accumulation of other intermediates. For this reaction it was proposed the scheme: urogen I v~ ,pentagen I v2 ~coprogen I, where v I >>v 2 (Koopmann et al., 1986). This paper describes the results of studies undertaken to obtain evidence as to the type of essential residues at the active site of Rp. palustris Uro-D. To this end, the effect of several metals and various reagents on the enzymic decarboxylation rate of urogen I, was investigated.

Enzyme preparation
A fraction purified 16-fold from Pp. palustris, following the procedure already described (Koopmann et al., 1986) was used throughout this study. 374 GERARDO ENRIQUE KOOPMANN and ALCI~ MARIA DEL C. BATLLE

Uro-D assay
Uro-D activity was assayed as previously detailed (Koopmann et al., 1986). Unless otherwise stated, the standard incubation conditions were as follows: the reaction mixture contained 0.1 ml of the enzyme preparation (0.20mg protein) in 0.05 M Tris-HC1 buffer pH 6.8; 15/,M urogen I, and buffer to make a final volume of 2 ml. Studies to ascertain the effect of metals, thiols and other reagents were performed by adding appropriate concentration of assay buffer (up to 0.2 ml) of the test substance. Solutions of the chemicals were either preincubated for 20 min at 37°C or without preincubations; the reaction was allowed to proceed for 60 min at 37°C after the addition of the substrate, with constant mechanical shaking, in the dark and under freeoxygen nitrogen. Controls were run containing no reagents. When the effect of 2-methoxy-5-nitrotropone and diethylpyrocarbonate was tested, the experimental conditions were different, following the procedures described by Tamaoki et al. (1967) and Setlow and Mansour (1970) respectively.
The rest of experimental conditions were already described (Koopmann et al., 1986).
Enzyme activity was quantitated and expressed on the basis of nmoles of coprogen I (C) and total decarboxylated products (TP) formed per hour. Activity of all controls was taken as 100%.
Other conditions or any added or omitted component are indicated in the text or in the legends to tables. Metals were chloride salts, except lead, copper and zinc acetates and ferrous ammonium sulphate. In all instances a wider concentration of the different compounds was tested, however only data obtained at the maximum are here reported.
Other materials or methods not specified here were those previously reported (Koopmann et al., 1986). Table 1 shows the inhibitory effect of some metals on the activity of Uro-D. Hg 2+ and Cu 2÷ are the most potent inhibitors of the enzyme, as already demonstrated (Woods et aL, 1981;De Verneuil et al., 1983;Straka and Kushner, 1983). Zn 2÷ and Fe 2÷, under certain conditions, are also inhibitors. Hg 2÷, Zn 2÷ and Fe 2÷ inhibitions can be reversed by GSH or CySH, suggesting that these metals do not denature the enzyme but probably bind reversibly to sulphydryl groups at the active site. Confirming previous findings (Woods et al., 1981) Pb 2÷, a well known inhibitor of 6-aminolevulinate dehydratase (ALA-D) and various other thiol enzymes, did not affect Uro-D activity; Cd 2+ did not inhibit the enzyme either. In most cases greater inhibition was attained when the enzyme was previously incubated with the metal; this is particularly noticeable with I mM Hg 2+, Zn 2+ and Fe z+ .

RESULTS
Hg ~+ and Fe 2+ also affected more strongly decarboxylation of pentagen I, that is v2, the rate-limiting step of the reaction (Koopmann et aL, 1986), was more sensible to inhibition than Vl. Here the pattern of porphyrinogen decarboxylated products showed a greater proportion of pentagen at the expenses of coprogen; in the presence of GSH however, not only inhibition was prevented but the original distribution of copro and pentagen of the controls was again obtained.
It is interesting that the inhibitory action of 1 mM Fe :+ was observed using either Tris-HC1 or phosphate buffer, although it was somehow greater when preincubation was done with the former.
Trivalent ions Fe 3+ and Cr 3+ also diminished Uro-D activity; AI 3+ instead, did not. Table 2 describes the effects of several compounds on Uro-D activity. As already noted (Koopmann et al., 1986), the presence of different thiols with or without preincubation, did not affect the enzyme. As expected, DTNB strongly inhibited, and it must be noted that here too the second step in the decarboxylation of urogen I was more sensible than the first. On titration with this reagent (data not shown), approximately five free SH groups per 46,000 of in R. palustris--VI 375 molecular weight were found; however, enzymic activity was rapidly lost after disappearance of only one thiol group. Sulphydryl-inactivating chemicals such as PCMB and chloral hydrate, clearly inhibited Uro-D activity, while N-ethylmaleimide only produced about 15% reduction; anyhow these findings were confirming previous results (Kardish and Woods, 1980;De Verneuil et al., 1983;Elder et aL, 1983;Kawanishi et al., 1983;Straka and Kushner, 1983) and supporting the view that a thiol group would play an important role in the enzyme. Neither EDTA nor o-phenantroline had any effect (Mauzerall and Granick, 1958;Tomio et al., 1970;Romeo and Levin, 1971;Woods et al., 1981;De Verneuil et al., 1983;Elder et al., 1983;Kawanishi et al., 1983;Koopmann et al., 1986), indicating that the purified Uro-D contains no metal (Kawanishi et al., 1983).
Interestingly, inactivation with diethylpyrocarbonate is suggesting that histidine residues may reside at the active site of the enzyme (Kawanishi et al., 1983). Moreover, the presence of one or more essential amino groups was also indicated by the inactivating effect of 2-methoxy-5-nitrotropone and pyridoxal phosphate.
Additional support for the existence of cysteine and histidine rests as part of the active center was found analyzing the pH profile of Pp. palustris Uro-D (Koopmann et al., 1986), that showed a maximum of 6.9 and 6.8 for urogens III and I as substrate, and inflection points at 6.40 and 8.45, which are known Experimental conditions are the same as those indicated in Table I  to correspond with pK values observed for histidine (6.00) and cysteine (8.33) (Lehninger,t975). The rabbit (Mauzerall and Granick, 1958) and the avian (Kawanishi et al., 1983) erythrocyte enzymes also had inflection points at pHs 8.2 and 6.0 respectively.

DISCUSSION
The inhibition pattern of the metal ions studied appears to support the presence of both cysteine and histidine at the active site of Rp. palustris Uro-D. Depending on the oxidation state, metals are preferentially bound to different types of ligands, involving atoms which can extend from highly electronegative or hard donor atoms such as oxygen, to the more polarizable or soft atoms such as sulfur. Ions binding hard donors are classified as hard cations while those binding soft donors are called soft cations, and there seem to be a good correlation between the biochemical activity of a metal ion and some parameters measuring their hardness or softness (Jones and Vaughn, 1978).
From the softest metal ions tested, Pb 2+, Cd 2+, Zn 2+, Cu 2+, Hg 2+ and Fe:+; Hg 2+ and Cu 2+ were strongly inhibitory and to a lesser extent Zn 2÷ and Fe:+; however, Pb 2+ and Cd 2+ were not. These metals are usually found bound to soft ligands, in this case most likely to sulfur donor portion of Uro-D. This observation is in agreement with the involvement of a eysteine residue, from the inflection point at pH 8.45 in the pH profile of enzyme activity, the reversal of metal inhibition by thiols and the inactivation by DTNB and SH reagents.
Confirming previous findings (Mukerji et at., 1984;Mukerji and Pimstone, 1986), Fe 2+ and Zn 2+ were found to inactivate Uro-D to different degree depending on the experimental conditions; GSH reversed inactivation, supporting further the view that these metals should bc blocking the active site by reversibly binding essential SH groups.
In spite of its well known affinity for sulfhydryl residues, neither Pb 2+ nor Cd 2+ inhibited Uro-D. These results can be explained if a monothiol rather than a dithiol group is at the active site, as already suggested (Woods et al., 1981), because these metals are strong inhibitors of enzymes possessing more than one SH residue or vicinal dithiols at the active center (Seehra et al., 1981); additional support for this view came from results of inhibition by arsenate (Batlle et al., 1965;Woods et al., 1981), and titration with DTNB (De Verneuil et al., 1983); indicating that at least one SH in the Uro-D is involved in the decarboxylation reaction mechanism.
According to Klopman (1968) metal hardness classification, Fe 3+ and Cr 3+ are intermediate ions, capable of accepting electrons from tertiary aromatic nitrogen as found in histidine; these results, together with diethylpyrocarbonate inactivation and the inflection point at pH 6.4 in the pH profile of enzyme activity, are strongly indicative of histidine at or near the active site. Furthermore, amino acid composition of human (De Verneuil et al., 1983) and avian (Kawanishi et al., 1983) erythrocytes showed between 11 and 12 histidine residues per mol of the native enzyme, and results on photoinactivation of human red blood cell Uro-D by porphyrins  would also imply a mechanism involving blocking or a conformational change of the active site through binding of a histidine functional residue. Specific studies on the role of histidine in the reaction mechanism of urogen decarboxylase are in progress.
Some indication of essential amino groups for catalysis was also obtained from the action of 2-methoxy-5-nitrotropone and pyridoxal phosphate.
In conclusion, present data would indicate that one cysteine, one or two histidine residues and possibly one lysine group are required for enzyme activity and therefore are involved in the reaction mechanism of urogen decarboxylation by Rp. palustris Uro-D.