Isolation, Crystallization, and Properties of Indolyl-3-alkane mHydroxylase A NOVEL TRYPTOPHAN-METABOLIZING ENZYME*

Indolyl-3-alkane alpha-hydroxylase, a novel tryptophan-metabolizing enzyme, was prepared in crystalline form from soil isolate organism Pseudomonas XA. Emission spectroscopy and atomic absorption analyses of purified enzyme revealed the presence of iron (0.8 mol/mol of protein), and a number of observations supported the presence of heme prosthetic group (1.1 mol/mol of protein). The S20,w value of indolyl-3-alkane alpha-hydroxylase is 10.2 S, and the molecular weight by sedimentation equilibrium ultracentrifugation is 250,000. The E1%280 of the enzyme is 21, and the isoelectric point by isoelectric focusing on ampholine polyacrylamide gel plates is 4.8. The enzyme catalyzes hydroxylation on the side chain of a variety of 3-substituted indole compounds, including certain tryptophan-containing oligopeptides. The reaction product from tryptamine was identified by proton nuclear magnetic resonance and gas chromatography/mass spectroscopy analyses. While the indole ring remained intact, hydroxylation occurred at the side chain carbon adjacent to the ring. Nuclear magnetic resonance studies indicated that hydroxylation always took place at the same position when the substrate was tryptophan methyl ester, tryptophol, indole-3-propionate, or indole-3-butyrate. No other chemical change occurred when these substrates were incubated with the enzyme. The Km value of indolyl-3-alkane alpha-hydroxylase for L-tryptophan is 2.4 X 10(-6) M, at pH 7.2. The enzyme is inhibited by potassium cyanide (0.1 mM) or hydroxylamine (1mM), but not by NaBH4 (25 mM), aminooxyacetic acid (7mM), quinacrine (1 mM), chlortetracycline (1 mM), p-mercuribenzoate (0.1 mM), or ethylenediaminetetraacetate (1 mM). The plasma half-life (t1/2) of indolyl-3-alkane alpha-hydroxylase in tumor-bearing mice is approximately 25 h.

Indolyl-3-alkane cu-hydroxylase, a novel tryptophan-metabolizing enzyme, was prepared in crystalline form from soil isolate organism Pseudomonas XA. Emission spectroscopy and atomic absorption analyses of purified enzyme revealed the presence of iron (0.8 mol/mol of protein), and a number of observations supported the presence of heme prosthetic group (1.1 mol/mol of protein). The s~,,.~ value of indolyl-3-alkane cY-hydroxylase is 10.2 S, and the molecular weight by sedimentation equilibrium ultracentrifugation is 250,000. The E.$& of the enzyme is 21, and the isoelectric point by isoelectric focusing on ampholine polyacrylamide gel plates is 4.8.
The enzyme catalyzes hydroxylation on the side chain of a variety of 3-substituted indole compounds, including certain tryptophan-containing oligopeptides. The reaction product from tryptamine was identified by proton nuclear magnetic resonance and gas chromatography/mass spectroscopy analyses. While the indole ring remained intact, hydroxylation occurred at the side chain carbon adjacent to the ring. Nuclear magnetic resonance studies indicated that hydroxylation always took place at the same position when the substrate was tryptophan methyl ester, tryptophol, indole-3propionate, or indole-3-butyrate. No other chemical change occurred when these substrates were incubated with the enzyme.

Growth and Harvesting of Bacteria
The bacterial organism was isolated from a local stagnant pond by selective enrichment on L-arginine/salts medium. Bacteria were routinely grown at 30" with aeration in Erlenmeyer flasks on a rotary shaker or in a fermentor.
The cells were harvested by centrifugation in the stationary phase of growth and stored at -75" until needed. Each liter of basic salts medium contained 0. It is known that certain enzymes (10, 11) are cleared at a considerably slower rate from the peripheral blood of mice injected with the lactate dehydrogenase-elevating virus. Most transplantable mouse tumors have been found to be infected with this virus (12). Therefore, the plasma half-life of indolyl-3-alkane a-hydroxylase was determined in methylcholanthrene-induced fibrosarcoma tumor-bearing (BALB/c x C57BL/6)F, mice which had been infected with the lactate dehydrogenase-elevating virus (12). Enzyme was injected intraperitoneally at a dose of 175 units/kg of body weight, and blood for determination of plasma indolyl-3-alkane cr-hydroxylase was obtained from mice by the orbital bleeding technique (11). Enzyme activity was assayed by measuring *'CO2 evolution from L-[l-"Cltryptophan. IndolylS-alkane a-Hydroxylase: Novel Enzymatic Activity vested cells were suspended in 2 to 3 volumes of 0.04 M potassium phosphate buffer, pH 6.5, and sonically disrupted with a Bronwill Biosonik IV sonifier with three 3-min bursts. The sonicate. was centrifuged 1 h at 9000 x g. L-Tryptophan was added to the cell extract to a concentration of 0.5 mM, and the enzyme solution was heated at 55" for 15 min with gentle agitation. After coolbg to 4", the precipitate was removed by centrifugation.

Characteristics of the Organism
Solid ammonium sulfate (263 g/liter) was added slowly to the supernatant enzyme solution while maintaining pH 6.5 by dropwise addition of ammonium hydroxide. After 30 min at 4", the precipitate was removed by centrifugation. Ammonium sulfate (125 g/liter) was then added to the supernatant, and the resulting precipitate was collected by centrifugation, suspended in a minimal volume of 0.02 M potassium phosphate buffer, pH 6.7, and dialyzed against the suspending buffer.
The dialyzed enzyme solution was applied to a DEAE-Sephadex column (5 x 150 cm), which had been equilibrated with 0.02 M potassium phosphate, pH 6.7. The column was eluted with a Miter linear gradient of 0 to 0.75 M KC1 in phosphate buffer. The flow rate was 50 to 60 ml/h and 25ml fractions were collected. The enzyme was eluted at between 0.13 to 0.16 M KCl, as determined by a conductivity bridge. The active fractions were combined, concentrated, and dialyzed against 0.02 M potassium phosphate buffer (pH 6.7) in an Amicon ultrafiltration cell. The enzyme from the DEAE-Sephadex step was chromatographed on a Sephadex G-200 column (2.5 x 90 cm). The column was eluted with 0.02 M potassium phosphate buffer, pH 6.7. The flow rate was about 20 ml/h. Appropriate fractions (7 ml/tube) were combined, concentrated, and dialyzed against 5 mM potassium phosphate buffer, pH 6.7. The enzyme solution was then applied to a hydroxylapatite column (1.5 x 20 cm), which had been equilibrated with 5 mM potassium phosphate, pH 6.7. The column was eluted with 10 mM potassium phosphate buffer, pH 6.7. The flow rate was 20 to 30 ml/h. The enzyme pool of fractions which appeared near the front was concentrated by ultrafiltration and stored at -75".
Crystallization of Indolyl-3-alkane a-Hydroxylase Solid ammonium sulfate was added to 40% saturation to a cold solution of chromatographically purified enzyme (3 to 8 mg of protein/ml) in 5 mM potassium phosphate buffer. The solution was clarified by centrifugation.
Ammonium sulfate was then gradually added until a slight turbidity appeared (about 52% saturation). This faint precipitate was immediately removed by centrifugation and crystallization was allowed to proceed at 5" for 24 h or longer. The crystals were needle-shaped as shown in Fig. 1.

Purity of Indolyl-d-alkane a-Hydroxylase
The purified enzyme showed a single protein component when it was analyzed by disc gel electrophoresis ( Fig. 2A), isoelectric focusing on ampholine polyacrylamide gel plates (Fig. 2B), and ultracentrifugation. Identification of Iron and Heme in Indolyl-d-alkane a-Hydroxylase Emission spectroscopy analysis of exhaustively dialyzed enzyme revealed the presence of iron, but no copper or other metal. We thank Dr. Donald Swanson of American Cyanamid for providing the analysis. Atomic absorption analysis (also by Dr. Swanson) indicated an iron content of 160 ppm c 10%. On the basis of an enzyme molecular weight of 250,000, the upper range of the iron value is equivalent to 0.8 mol of iron/m01 of enzyme. There was no stimulation of enzyme activity when Fez+ or Fe3+ salts (1 mM) were added to the enzyme reaction mixture.
A number of observations support the presence of heme prosthetic group in the enzyme. The enzyme was inhibited by potassium cyanide (Table II). In polyacrylamide gels the enzyme was strongly stained by benzidine reagent (0.2% in 0.5% acetic acid with 0.06% H,O,). Crystalline enzyme exhibited the characteristic absorption spectrum for a porphyrin group (a major peak at approximately 425 nm and much smaller peaks at 528 and 558 nm) (Fig. 3), as well as essentially the same pyridine hemochromogen spectra. The content of heme was determined from the extinction coefficient of the pyridine hemochrome (E,, = 33 at 556 nm). The enzyme was found to contain 1.1 mol of hemelmol of enzyme protein.  reaction since preincubation of enzyme with NaBH4 (25 mM) or aminooxyacetic acid (7 mM), both potent inhibitors of pyridoxal phosphate action, resulted in no inhibition of enzyme activity. Similarly, the absence of any effect by quinacrine or chlortetracycline suggests that flavin nucleotide is not involved in enzyme activity (Table II).

Isoelectric Point and Molecular Weight oflndolyl-3-alkane a-Hydroxylase
The isoelectric point of purified enzyme by isoelectric focusing on LKB Instruments ampholine-polyacrylamide gel plates was 4.8 & 0.1 with the pH 3.5 to 9.5 carrier.
The sedimentation coefficient of purified enzyme was determined with a Beckman-Spinco model E analytical ultracentrifuge. Schlieren patterns showed a single protein component. Thes 20,w value of the enzyme was 10.2 S. The molecular weight of indolyl-3-alkane a-hydroxylase from sedimentation equilibrium studies was 250,000. We thank Dr. Ralph Barclay of the CIBA-GEIGY Corp. and Dr. Marion Barclay of the Sloan-Kettering Institute for providing the ultracentrifugal analyses.

Amino Acid Composition and Extinction Coefficient
An aqueous solution of homogeneous indolyl-3-alkane CYhydroxylase containing 1 mg of protein (Lowry et al. (7) has an absorbance of 2.6 at 260 nm in a Gilford model 240 spectrophotometer at 25". Aliquots of homogeneous enzyme were hydrolyzed with p-toluenesulfonic acid at 110" in sealed evacuated glass tubes for 22, 48, and 72 h (13). Other aliquots were oxidized by performic acid for cysteic acid determination (14). Amino acid analyses were performed on a Beckman model 119 amino acid analyzer (15,161. The data are presented in Table III. The sum of individual amino acid weights (corrected for the weight of water eliminated in forming peptide bonds) was used to calculate the absorbancelmg of protein. The E@ of indolylS-alkane a-hydroxylase from Pseudomonas XA is 21.

Substrate Specificity of IndolylS-alkane a-Hydroxylase
Results summarized in Table IV illustrate the relative activity of the enzyme with a wide variety of 3-substituted indole compounds. It is noteworthy that the enzyme showed activity with tryptophan-containing di-and oligopeptide.

Effect of pH and Substrate Concentration on Activity
The effect of pH on enzymatic activity is shown in Fig. 4. Similar pH activity profiles were obtained when enzyme activity was measured by oxygen consumption or '*CO, evolution Indolyl-3-alkane a-Hydroxylase: Novel Enzymatic Activity from L-[1-Yltryptophan. The pH activity profiles for tryptamine and L-tryptophan methyl ester were also determined and found to be very similar to that of L-trypotophan.
The initial reaction rates of WOy evolution from L-[l-Wltryptophan were determined at varying substrate concentrations. The average K,,{ values of indolyl-3-alkane cu-hydroxylase in duplicate determinations are 5.8 ? 0.4 x lo-" M at pH

Stoichiometry of Enzyme Reaction
The stoichiometry of the enzyme reaction was studied at 37", in 0.1 M sodium acetate buffer, pH 5.5. Oxygen uptake was determined polarographically, NH,, production with Nessler's reagent and CO, evolution by counting '%O, release from L-

Identification of Reaction Products
The structure assigned to the tryptamine reaction product was based on 'H NMR spectral analysis in D,O (Fig. 5). The low field multiplet was centered at 6 = 7.45 and integrated for 5 protons. The shape of the signal was almost identical to that of the aromatic signal of authentic tryptamine.
Thus, the indole ring of the reaction product was intact. Two signals (a triplet at 6 = 5.35 and a doublet at fi = 3.49, integrated for 1 and 2 protons, respectively) coupled to each other (J = 6.4 HZ) clearly established the -CH-CH,-structure in the molecule. Since there was no other signal, the above evidence proves unequivocally that hydroxylation has taken place on the side chain carbon adjacent to the indole ring. Essentially identical results were obtained for the other reaction products, using as substrate tryptophol, tryptophan methyl ester, indole-3-propionic acid, or indole-3-butyric acid. The latter two compounds were not decarboxylated since no 'H NMR signal for a methyl group appeared. Therefore, the primary reaction catalyzed by indolyl-3-alkane cu-hydroxylase is hydroxylation of the side chain carbon adjacent to the indole ring, whereas decarboxylation (of tryptophan only) is secondary. Appar- ently, deamination is also secondary since enzymatic action on tryptophan methyl ester did not liberate NH:,. An attempt was made to isolate side chain-hydroxylated tryptophan by gently hydrolyzing the ester group of side chain-hydroxylated tryptophan methyl ester, which was fairly stable. During the hydrolysis with hog liver esterase, a product formed which migrated slower than the esterified compound on thin layer chromatography.
Upon addition of indolyl-3-alkane c*-hydroxylase to the reaction mixture, this newly visualized compound disappeared with the concomitant formation of NH,, and a brown color. Thin layer chromatography confirmed that the brown color was the same as that produced when tryptophan was incubated with indolyl-3-alkane oc-hydroxylase. This experiment suggests that the compound formed after esterase action was a-hydroxytryptophan.
Elemental analyses of the reaction product from tryptamine. HCl (Galbraith Laboratories) revealed a good correlation between actual and calculated carbon, hydrogen, and nitrogen content assuming that the product remained as the hydrochloride salt. The carbon, hydrogen, and nitrogen content was found to be essentially unchanged after the sample was dried in uacuo overnight at 50". Any firmly bound H,O could not be successfully removed at higher temperatures without incurring product decomposition.
Analysis of the tryptamine product for an alcohol group was performed by the vanadium-oxine color test, as outlined under "Experimental Procedures." Two milligrams of tryptophol, which contains a primary alcohol group, produced appreciable Two milligrams of tryptamine product gave a color reading of 0.91 which is indicative of the presence of an alcohol group.
Gas chromatography/mass spectroscopy of silylated tryptamine revealed a major peak at 377 (M+ + l), which corresponds to the mass of tryptamine with three trimethylsilyl moieties, as expected. The silylated tryptamine product showed a major peak at 465 (M+ + l), which corresponds to tryptamine product with four trimethylsilyl moieties, also as expected for hydroxytryptamine.
This evidence, together with the results from 'H NMR analyses, firmly establish the structure of the product.
The proposed tryptamine product has appeared previously in the literature without any description of its preparation (17). Attempts in our laboratory to chemically synthesize the enzyme reaction product of tryptamine or tryptophol (a diol), however, have been unsuccessful. The failure to accomplish the desired synthesis was due to the reduction with LiBH, of the carbonyl function adjacent to the indole ring beyond the alcohol stage to a methylene group.

Plasma Half-life of Indolyl-3-alkane a-Hydroxylase
The plasma half-life (tliZ) of indolyl-3-alkane a-hydroxylase injected into tumor-bearing (BALB/c x C57BL/6)F, mice was about 25 h (Fig. 6). DISCUSSION A variety of bacteria was tested in our laboratory for production of ammo acid-degrading enzymes that would be suitable for studying the effect of amino acid deprivation on growth of tumors. One of our soil isolate organisms (Pseudomonas XA) was found to contain an enzyme which metabolizes tryptophan by a mechanism which is different from any previously described enzyme. This enzyme catalyzes hydroxylation on the side chain of a variety of 3-substituted indole compounds. color in this test (A,,, = l.O), whereas 2 mg of tryptamine gave Nuclear magnetic resonance spectroscopy studies with reaca color reading equal to that of the H,O blank (A,,, = 0.07). tion products of tryptamine, tryptophol, tryptophan methyl Indolyl-3-alkane a-Hydroxylase: Novel Enzymatic Activity ester, indole-3-butyrate, and indole-3-propionate revealed that the site of hydroxylation is at the carbon adjacent to the indole moiety (a carbon). The structural requirements for the side chain at position 3 of indole appear to be broad. Even certain tryptophan-containing oligopeptides act as substrates. Although certain substitutions on the ring seem to be tolerated (Table IV), only compounds with an intact indole moiety are attacked by the enzyme.
It was not possible to identify the principal product formed by the action of indolyl-3-alkane a-hydroxylase on L-tryptophan. The principal, indole-containing tryptophan product is apparently unstable since efforts to isolate it with preparative layer chromatography, silica gel, or Bio-Gel P2 column chromatography resulted in the appearance of ten or more compounds. Relatively stable reaction products were obtained from tryptamine, tryptophol, 5-hydroxytryptamine, tryptophan methyl ester, indole-3-propionate, and indole-3-butyrate, and they were used for analyses. The product from tryptamine was most extensively studied. Nuclear magnetic resonance spectroscopy indicated that the reaction product was a-hydroxytryptamine.
When this product was derivatized by silylation and subjected to gas chromatography/mass spectroscopy, the resulting mass fragments pattern was in excellent agreement with the structure deduced from nuclear magnetic resonance spectroscopy. Thus, the action of indolyl-3-alkane a-hydroxylase on (a) tryptamine, (b) tryptophol, (c) indole-3-butyrate, and (d) L-tryptophan may be illustrated as follows. Indolyl-3-alkane a-hydroxylase contains heme prosthetic group. Pyridoxal phosphate or flavin nucleotide do not appear to be involved in enzyme activity. Emission spectroscopy analysis revealed the presence of iron, but no copper or other metal. The enzyme was not significantly inhibited by carbon monoxide. Addition of sodium dithionite to a solution of the enzyme resulted in a shift in the Soret band towards a longer wavelength. Potassium ferricyanide had no effect on the absorption spectrum or enzyme activity. These data suggest that the active enzyme is in a ferric form.
The presently described enzyme, which demonstrates a novel activity, possesses certain characteristics in common with other aromatic amino acid-metabolizing enzymes. The well studied tryptophan oxygenase from Pseudomonas fluerescens (181, which converts L-tryptophan to formylkynurenine, also is an iron-porphyrin enzyme. However, unlike the presently described enzyme, the tryptophan dioxygenase was shown to be inhibited by carbon monoxide and ferricyanide, indicating that the reduced enzyme was the active species. Tryptophan oxidative decarboxylase from Pseudomonas savastanoi produces indole-3-acetamide plus CO,. No cofactor has been demonstrated for that enzyme other than an essential sulfhydryl group (19). Aromatic L-amino acid decarboxylase has been isolated from bovine brainstem (20) and from Micrococcus percitreus (21). Unlike the presently studied enzyme, the aromatic L-amino acid decarboxylase enzymes exhibited an absolute requirement for pyridoxal phosphate and yielded tryptamine and CO, from L-tryptophan. The conversion of dopamine to norepinephrine in rat adrenal gland (22) may be regarded as analogous to the reaction under investigation. The enzyme dopamine /3-hydroxylase which has been shown to require 0,, ascorbate, and fumarate, oxidizes the side chain carbon adjacent to the aromatic ring. That enzyme also has broad specificity, since it was shown to hydroxylate numerous structural analogues of dopamine at the same position.
While our work was in progress, Takai et al. (23) reached essentially the same conclusions studying the novel tryptophan-metabolizing enzyme which was isolated from our laboratory organism Pseudomonas XA.