FEBS 15694 FEBS Letters 368 (1995) 117-12l Redox properties of the sulfhydrogenase from Pyrococcus furiosus Alexander F. Arendsen, Peter Th.M. Veenhuizen, Wilfred R. Hagen* Department of Biochemistry, WageningenAgricultural University, Dreijenlaan 3, NL-6703 HA Wageningen, The Netherlands Received 24 April 1995; revised version received 26 May 1995 of a [2Fe-2S] and a [4Fe-4S] cluster in a EPR redox titration Abstract The sulfhydrogenase from the extreme thermophile [6]. Reduction potentials of -410 mV of the [2Fe-2S] cluster has been re-investigated. The a13-/6 hetero- Pyrococcusfuriosus and of-210 mV of the [4Fe~,S] cluster were determined. In this tetrameric enzyme of 153.3 kDa was found to contain 17 Fe, 17 paper we present data on the EPR properties of P furiosus S 2-, and 0.74 Ni. The specific activity of the purified protein was 80 U/mg. Three EPR signals were found. A rhombic S = 1/2 sulfhydrogenase that are consistent with the presence of three, signal (g = 2.07, 1.93, 1.89) was observed reminiscent in its shape rather than two signals. We also compare differences in metal and temperature dependence of spectra from [4Fe-4S] °+;~÷) clus- content and reduction potentials. The possibility that the rhom- ters. However, in reductive titrations the spectrum appeared at bic signal represents a two-electron transferring cluster is dis- the unusually high potential Emas = -90 mV. Moreover, the sig- cussed. nal dissappeared again at Ema.s = -328 mV. Also, two other signals appear upon reduction: a near-axial (g = 2.02, 1.95, 1.92) S = 1/2 spectrum (Em,7.s = -303 mV) indicative for the presence 2. Materials and methods of a [2Fe-2S] ~2+;1+) cluster, and a broad spectrum of unknown 2.1. Growth of organism origin with effective g-values 2.25, 1.89 (Em,7.s = -310 mV). We P furiosus (DSM 3638) was grown on potato starch (5 g/l) in a stirred hypothesize that the latter signal is caused by magnetic interac- (150 rpm) 200 liter fermentor (lp 300, Bioengineering, Wald, Switzer- tion of the rhombic signal and a third cluster. land) at 90°C, in a medium described in [7]. NaC1 p.a. was replaced by table salt. The medium also contained yeast extract (1 g/l) and cysteine Key words: Hydrogenase; EPR; Pyrococcus furiosus (0.5 g/l). Trace elements were Na2WO4, 10 ktM; Fe(NH4)SO2, 25/.tM; NiCI2, 2.1 pM; H3BOs, 0.2 tiM; zinc acetate, 1/.tM; CuSO4, 0.04 ~tM; MnC12"4H20, 1 pM; COSO4, 1.5//M; Na2MoO4.2H20, 0.03/IM. Vita- mins used were biotin, 2/~M; folic acid, 2/.tM; nicotinamide, 5/~M; thiamin HC1, 5/lM; riboflavin, 5/tM; pyridoxine HCI, l0/~M; cyano- 1. Introduction cobalamine, 5/tM; p-aminobenzoic acid, 5/tM; lipoic acid, 5/tM; and pantothenic acid, 5/IM. No elemental sulfur (S °) was present. The The sulfhydrogenase of the anaerobic archaeon Pyrococcus medium was continuously flushed with nitrogen to remove evolved furiosus was first isolated by Bryant and Adams [1]. P furiosus hydrogen gas. Growth was followed by measuring the absorbance at 660 nm. Cells were frozen and stored at -20°C until use. Potato starch grows near 100°C by fermentation of carbohydrates, resulting and table salt were obtained from the local grocery. in CO 2 and H 2 as sole products. However, H 2 is only found when the organism is grown in the absence of elemental sulfur 2.2. Isolation procedure (S°). When the culture medium contains sulfur, H2S instead of Sulfhydrogenase was purified according to a modification of the procedure of Bryant and Adams [1]. Ethanol was emitted from the H2 is produced. Since H2 inhibits growth, it was hypothesized buffers. Because of a higher capacity a Sephacryl PG 200 molecular that sulfur reduction is a means of detoxification [2]. Attempts sieve (2.5 × 100 cm) was used instead of a Superose 6 column. As the to isolate the sulfur reductase revealed that sulfur reduction first purification step an ammonium sulfate precipitation was per- activity coincided with hydrogenase activity. Therefore, it was formed. Ammonium sulfate was added to the extract to 55% saturation. concluded that both H2S and H 2 production are catalyzed by After a 10 min spin at 5,000 x g hydrogenase activity was located in the pellet. The pellet was resuspended in standard buffer and dialyzed the same enzyme, and by concequence the enzyme was called against standard buffer. As an extra purification step a 50 ml Phenyl sulfhydrogenase [3]. The protein is encoded by a chromosomal Sepharose hydrophobic interaction column was used. Fractions con- operon of four open-reading frames [4] for four polypeptides taining hydrogenase activity eluting from the hydroxyapatite column of 48.7, 41.8, 33.2, and 29.6 kDa [5]. Unlike other hydrogenases were pooled and adjusted to the appropriate ionic strenght of 0.25 M by adding crystals of (NH4)2SO4. A 1 liter linear gradient of 0.254) M the P. furiosus enzyme catalyzed preferably H 2 production [1]. (NH4)2SO4 in standard buffer was applied. Hydrogenase eluted at 0 M Bryant and Adams reported the enzyme to contain 31 Fe, 24 (NH4)2SO4. S 2-, and 1 Ni per 185 kDa. Despite the high metal content only two EPR signals were found. In the reduced state the enzyme 2.3. Activity measurements, analytical chemistry exhibited resonances at g = 2.03, 1.93, and 1.92, which could Hydrogenase activity was measured in the hydrogen production assay according to Bryant and Adams [1] using a gas chromatograph been observed from 70 to 20 K. This signal integrated to only (model 3400, Varian). one spin/mol. Upon lowering the temperature a complex signal Protein was measured using the microbiuret method [8]. Iron was was found with fast spin-lattice relaxation rate, which ac- determined colorimetrically as the ferene complex [9]. Nickel was deter- counted for another spin/mol. No Ni signals were found. The mined by atomic absorption spectroscopy on a Hitachi 180-80 Polar- ized Zeeman Atomic Absorption Spectrophotometer equipped with a former signal was assigned to a [2Fe-2S] cluster, while the latter pyrocuvette in the laboratory of Dr. S.P.J. Albracht (The University was proposed to arise from at least two interacting iron-sulfur of Amsterdam). clusters. In a later publication Adams reports the monitoring SDS polyacrylamide electrophoresis was performed with a midget system (Pharmacia) holding 8-× 5-× 0.75-cm gels, according to the method of Laemmli [10]. The composition (mass/volume) of the stack- *Corresponding author. Fax (31) (8370) 84 801. ing gel was 4% acrylamide and 0.1% bisacrylamide; the running gel was 0014-5793/95l$9.50 © 1995 Federation of European Biochemical Societies. All rights reserved. SSD1 0014-5793(95)00622-2 118 A.E Arendsen et al./FEBS Letters 368 (1995) 117-121 17.5% acrylamide and 0.07% bisacrylamide. The native molecular mass Table 1 of the protein was determined with a Superose 6 HR 10/30 column Physico-chemical data on P Juriosus sulfhydrogenase (Pharmacia) equilibrated with standard buffer + 0.15 M NaCI using a flow rate of 0.5 ml/min. Glucose oxidase (152 kDa), lactate dehydroge- Previous reports This paper nase (109 kDa), transferrin (74 kDa), BSA (67 kDa), and ovalbumine Ni 0.98 + 0.05 [1] 0.74 _+ 0.21 (43 kDa) were used as the markers for the calibration. The void volume Fe 31 + 3 [1] 17+2 was determined with Dextran blue. S2 24 + 4 [1] 17+2 Spec. activity (U/mg) 360 [1] 80 2.4. Spectroscopy E m [2Fe-2S] (mV) -410 [6] -303 EPR measurements and mediated redox titrations were as in [11]. In E m [4Fe~tS] (mV) nr -90 titration experiments protein concentration was typically 4 ,uM in 100 E m 'broad signal' (mV) -210 [6] -310 mM HEPES pH 7.5, and the following EPR signals were monitored: the g = 1.89 peak of the broad signal at the wings; the g = 1.89 peak nr, not reported. of the rhombic S = 1/2 signal; the g = 1.92 peak of the near-axial S = 1/2 signal. Titrations were done at 23°C. ucts (~, fl, Z, and 6) [5]. The sum of the masses of the individual 3. Results subunits is 153.3 kDa. To obtain the subunit composition we performed a gel-filtration experiment. On a Superose 6 column 3.1. Isolation the native protein ran with an apparent molecular mass of 135 Our first attempts to isolate the sulfhydrogenase from Pyro- kDa (Fig. 2). This value fits reasonably with a ~flg~ subunit coccus Juriosus using the procedure described by Bryant and composition. Moreover, on a SDS gel all four subunits have Adams [1] did not result in a pure hydrogenase sample. Instead, similar intensities (Fig. 1). We therefore conclude that P. furio- more then four bands were observed on SDS gel. By use of a sus sulfhydrogenase is an ~flg~ heterotetramer of 153.3 kDa. modified procedure, i.e. by using ammonium sulfate precipita- Table 1 lists the iron, sulfide, and nickel contents of three tion as the first step, and a Phenyl-Sepharose hydrophobic different preparations. The enzyme contained 17 + 2 Fe, 17 + 2 interaction column, a preparation with a higher apparent purity S 2-, and 0.74 + 0.21 Ni atoms per 153.3 kDa. on SDS gel was obtained. Clearly, four subunits can be ob- served of 45, 43, 31, and 28 kDa (Fig. 1). These values fit 3.3. EPR spectroscopy reasonably with the values of the subunits deduced from the In the reduced state P.