JOURNAL OF BIOLOGICAL 268, No. 12, 8541-8546, 1993 0 1993 ... · 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol.

0 1993 by The American Society for Biochemistry and Molecular Biology, Inc THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, No. 12, Issue of April 25, pp. 8541-8546, 1993

Printed in U. S. A.

Structure-Function Studies of Yeast Ferrochelatase IDENTIFICATION AND FUNCTIONAL ANALYSIS OF AMINO ACID SUBSTITUTIONS THAT INCREASE V,,, AND THE KM FOR BOTH SUBSTRATES*

(Received for publication, October 5 , 1992)

Abdelhamid Abbas and Rosine Labbe-Bois$ From the Laboratoire de Biochimie des Porphyrines, Institut Jacques Monod, Uniuersite Paris 7, 2 Place Jussieu, 75251 Paris Cedex 05, France

The molecular basis of the ferrochelatase defects was investigated in two “protoporphyric” and partially heme-deficient yeast mutants. Ferrochelatase, a mitochondrial inner membrane-bound enzyme, catalyzes the incorporation of ferrous iron into protoporphyrin, the last step in protoheme biosynthesis. The mutant cells made normal amounts of normal-sized ferrochelatase, as detected by immunoblotting. The mutations were identified by sequencing the mutant hem15 alleles amplified in vitro from mutant strains genomic DNA. A single nucleotide change, causing an amino acid substitution, was found in each mutant. Substitu- tion of the conserved Ser-169 by Phe caused a 10-fold increase in V,, and a 45- and 35-fold increase in the KM for protoporphyrin and metal, respectively. Re- placement of Ser-174 by Pro produced the same ef- fects, but to a lesser degree. There was a good corre- lation between the ferrochelatase defects measured in vitro and the heme synthesis deficiencies estimated in vivo. The decreased in vivo heme synthesis is probably due to the lower affinity of the mutant enzymes for iron. We propose that the region identified by the two close mutations contributes to the binding domains of metal and protoporphyrin.

The chelation of ferrous iron by the tetrapyrrolic macro- cycle protoporphyrin IX to form protoheme is the final step in the heme biosynthetic pathway and is catalyzed by the membrane-bound enzyme ferrochelatase (protoheme ferro- lyase EC 4.99.1.1). In eukaryotic cells, where it has been most studied, the enzyme is associated with the inner mitochondrial membrane, with the active site facing the matrix compart- ment, (Dailey, 1990, and references therein). The ferrochelatases isolated from rat (Taketani and Tokunaga, 1981) and bovine (Taketani and Tokunaga, 1982) liver and from the yeast Saccharomyces cereuisiae (Camadro and Labbe, 1988) have very similar physicochemical and catalytic properties. Ferrochelatase cDNAs or genes have been isolated and sequenced from mouse (Taketani et al., 1990; Brenner and Frasier, 1991), human (Nakahashi et al., 1990), S. cerevkiae (Labbe-Bois, 1990), Escherichia coli (Miyamoto et al., 1991) and Bradyrhizobium japonicum (Frustaci and O’Brian, 1992). The derived amino acid sequences show from 22 to 88% identity when analyzed in pairs and 10% identity when taken

* This work was funded by the Centre National de la Recherche Scientifique and by the Universitb Paris 7. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

143540479; Fax: 33-144275716. $To whom correspondence should be addressed: Tel.: 33-

together. The hydropathy profile reveals no apparent membrane-spanning segment, indicating that ferrochelatase is pe- ripheral.

A model has been proposed for the mechanism of action of ferrochelatase, on the basis of kinetic studies, chemical modifications of sulfhydryl and arginyl residues, and from the analysis of the strong inhibition by N-alkylporphyrins (Dailey, 1990 and references therein). This model implicates cysteine and arginine residues in the binding of metal and porphyrin propionate(s), respectively, and involves distortion of the porphyrin ring as a transition-state intermediate. It has recently received strong support with the finding that antibodies elicited to a distorted N-methylporphyrin (transition-state analog) could catalyze metal ion chelation by the planar porphyrin (Cochran and Schultz, 1990). But the fact that no cysteine is conserved in the known ferrochelatase sequences makes it unlikely that cysteine(s) may form part of the binding site for metal substrate.

No mutational analysis of ferrochelatase has yet been reported which could help investigate the structural and functional requirements of the enzyme. Ferrochelatase activity is decreased in the human and bovine hereditary disease, eryth- ropoietic protoporphyria, in which excessive protoporphyrin accumulates in various tissues, causing cutaneous photosen- sitivity (Nordmann and Deybach, 1990). Some enzymic defects have been identified in protoporphyria, but the mutational events responsible for them are unknown (Bloomer et al., 1987; Straka et al., 1991; Blom et al., 1990). On the other hand, a number of mutations causing amino acid substitutions have been reported in human (Lamoril et al., 1991) and E. coli (Miyamoto et al., 1991; Nakahigashi et al., 1991) ferrochelatases, but the nature of the enzyme dysfunction was not analyzed.

The present work describes two amino acid substitutions in the S. cerevisiae ferrochelatase which lead to increased K M

for both metal and protoporphyrin and to an increased Vmax. We have also analyzed the phenotypic consequences of these enzymic defects. The implication of the mutated region in the binding of substrates at the active site of ferrochelatase is discussed.

MATERIALS AND METHODS

Strains, Media, and Cultivation-The S. cereuisiae mutant strains Sm12 (Mata leul arg4 cttl-1 hem15-3) and Sm41 (Mata leul arg4 ctal-1 hem15-4) were isolated from the parental strains DCT1-3D and DCT3-4A, respectively (Kurlandzka and Rytka, 1985). Since the strain DCT3-4A no longer exists, the strain DCT1-3D was used as reference wild-type HEM15 strain for both mutant strains. The cells were grown at 30 “C with vigorous agitation and aeration in complete medium (1% yeast extract, 1% bactopeptone) with either 2% glucose (YPG) or 2% ethanol + 2% glycerol (YPE) as carbon source. The medium was occasionally supplemented with 0.2% Tween 80 (YPGT,

854 1

a542 Yeast Ferrochelatase Mutations YPET). Iron-rich medium was supplemented with 0.18 mM FeCb + 4 mM sodium citrate.

E. coli strain DH5a (Bethesda Research Laboratories) was used as the bacterial host for plasmid cloning and proliferation.

Isolation and Sequencing of the hem15 Mutant Alleles-Genomic DNA was prepared from the parental and mutant strains as described (Sherman et al., 1986). It was used as template (1 fig) for amplification of the HEM15 alleles by the polymerase chain reaction employing the GeneAmp kit (Perkin-Elmer Cetus). Two oligonucleotides, cor-

to the A + l of the initiator ATG), 5"CCGGAATTCACACA- responding to nucleotides -54 to -34 of the coding strand (relative

TACCTGCTATTTGGAC, and to nucleotides 1207-1187 of the non- coding strand, 5'-CCCAAGCTTTGAGATTGTGGGATGAATGG, were synthesized with the addition of an EcoRI or a Hind111 site (underlined) to the 5' end, respectively. Amplification was carried out with 30 cycles of 1 min denaturation at 90 "C, 1 min annealing at 55 "C, and 3 min extension at 70 "C. The resulting amplified DNA fragments (1.26 kilobases) were cloned into pBluescript (Stratagene). Plasmid DNA isolated from 10 individual transformants for each strain was combined to alleviate possible errors introduced by the Tacq polymerase and sequenced using the Sequenase polymerase (U. S. Biochemicals), [ w ~ ~ S ] ~ A T P (Amersham), and HEM15-specific oligonucleotide primers.

Ferrochelatase Immunodetection and Activity Assay-Total proteins extracted from trichloroacetic acid-treated cells and proteins from the membrane fraction in 0.05 M NaOH were resolved by SDS- polyacrylamide gel electrophoresis and electrophoretically transferred to nitrocellulose membranes. An alkaline phosphatase-coupled secondary antibody (Promega) was used to visualize the anti-yeast ferrochelatase antibody prepared by J. M. Camadro (Camadro and Lahbe, 1988).

Membrane-bound ferrochelatase was assayed spectrofluorometri- cally by measuring the rate of zinc-protoporphyrin formation (Ca- madro and Labbe, 1988). Mitochondrial membranes were prepared as follows. Cells were harvested in the mid-log phase of growth in YPET medium, washed with water, suspended in 0.1 M Tris-HC1, pH 7.6, and mechanically disrupted with glass beads. The resulting cell homogenate was cleared of debris by centrifugation at 5,000 X g for 5 min. The membrane fraction was pelleted at 100,000 X g for 60 min. Membranes were washed once with 0.1 M Tris-HC1, pH 7.6, resuspended at a protein concentration of 20-30 mg/ml, and stored at -70 "C. Proteins were determined by the method of Lowry using bovine serum albumin as standard.

Ferrochelatase activity was monitored by directly recording the rate of zinc-protoporphyrin formation at 30 "C with a Jobin-Yvon JY3-D spectrofluorometer equipped with a thermostated cell-holder allowing magnetic stirring in the cuvette and a red-sensitive Hama- matsu R928H4 photomultiplier tube. The maximum excitation and emission wavelengths for zinc-protoporphyrin under the assay conditions were 418 and 588 nm, respectively. The reaction mixture (3 ml final) contained 0.1 M Tris-HC1, pH 7.6, 0.2 mg/ml Tween 80 (including the amount of Tween 80 added with the protoporphyrin solution), protoporphyrin, and zinc. The reaction was initiated by adding 0.1-0.5 mg of membrane protein. Stock solutions of 1 mM protoporphyrin IX disodium salt (Serva) in 0.1 M Tris-HC1, pH 7.6, containing 1% (w/v) Tween 80 were diluted with the same buffer just before use; their absorption spectra were run to check for monomer- ization of protoporphyrin. Zinc was prepared as a 3 mM stock solution of ZnSOl. 5H20 in water. Pure zinc-protoporphyrin (Porphyrin Prod- ucts, Logan, UT) was used to calibrate the assay. Ferrochelatase activity was expressed as nanomoles of zinc-protoporphyrin/hour/ milligram of protein. The enzyme kinetic data were analyzed graphically and using the EZ-FIT curve-fitting program on an IBM-PC microcomputer (Perrella, 1988).

Parameters of the Heme Biosynthesis Pathway-Published proce- dures were used to 1) record low temperature (liquid nitrogen) absorption spectra of whole cells, 2) quantify whole cell total heme by the pyridine hemochrome spectra, 3) analyze the cytochrome content by reduced-minus-oxidized difference spectra, 4) identify by high performance liquid chromatography and quantify the porphyrins accumulated in the cells and excreted in the culture medium, 5) measure the intracellular content of 5-aminolevulinic acid (ALA),' and 6) measure the activity of ALA synthase and coproporphyrinogen oxidase in cell-free extracts (Urban-Grimal and Labbe-Bois, 1981; Rytka et al., 1984; Zagorec and Labbe-Bois, 1986). Total non-heme

The abbreviations used are: ALA, 5-aminolevulinic acid; Mes, 4- morpholineethanesulfonic acid zinc-proto, zinc-protoporphyrin.

iron was quantified spectrophotometrically after its chelation with bathophenanthroline sulfonate as described by Tangeras et al. (1980), with minor modifications. Briefly, the samples were "solubilized in 10 mM Mes buffer, pH 4.5, containing 1% (w/v) SDS and incubated at 95 "C for 1 h in the case of whole cells. Any oxidized Fe(II1) was reduced with dithionite, and the spectra of the iron complexes were recorded from 500 to 600 nm, and an absorption coefficient EmM ( A A , 535-600 nm) = 19.4 was used.

RESULTS

Phenotypic Characteristics of the Mutant Sm12 and Sm41 Strains-The two mutant Sm12 and Sm41 strains were isolated from two different haploid parental strains in a system- atic search for mutants partially defective in the last steps of the heme biosynthetic pathway. These mutants accumulate porphyrins, are photosensitive, fluoresce under UV light, but maintain sufficient heme synthesis to grow on non-ferment- able carbon sources such as ethanol or glycerol (Kurlandzka and Rytka, 1985). Sm12 and Sm41 were shown, by genetic analysis, to carry single recessive mutations at the HEM15 locus, the structural gene for ferrochelatase. The growth rate of the mutant cells was lower than that of the parent strain (Table I). Heme synthesis in the mutant Sm12 was about 2.5- fold lower than in its parent strain, and it accumulated large amounts of porphyrins, mainly protoporphyrin (90-95% protoporphyrin, 2-5% coproporphyrin, and 2-5% pentacarboxy- porphyrin) (Fig. 1, Table I). The heme content of mutant Sm41 was also lower than normal, but the relative decrease could not be quantified since its parent strain is no longer available. Both mutant cells produced less heme and protoporphyrin when growing fermentatively on glucose than when they were growing respiratively on ethanol; the reason for the glucose effect is unclear at present. The majority of protoporphyrin accumulated by the mutant cells was, in fact, excreted at the cell surface and appeared in the culture medium as small "globules" of (sedimentable) protoporphyrin aggregates. The addition of Tween 80 or other detergent (Kurlandzka and Rytka, 1985) to the culture medium pseudosolubilized these aggregates, and the excreted protoporphyrin was then present as the monomer in the detergent micelles. The spectra of mutant cells grown in the absence of detergent show two large absorption bands at 585-590 and 640-645 nm charac- teristic of aggregated protoporphyrin (Fig. 1). Cells grown in the presence of 0.2% Tween 80 showed only small peaks, besides the cytochrome peaks, near 575 and 630 nm for the protoporphyrin monomer and near 585 nm for zinc-protoporphyrin (Fig. 1). Thus, almost all the protoporphyrin made by the mutant cells in excess of that used for heme synthesis is excreted out of the cells and very little, if any, is used by ferrochelatase to make zinc-protoporphyrin.

Table I also shows that the mutant Sm12 contained more ALA than its parental wild-type strain. This correlates well with a 2-fold increase in ALA synthase activity. A similar increase was reported in a mutant partially defective in heme formation due to an altered uroporphyrinogen decarboxylase activity (Rytka et al., 1984). The regulatory event(s) and the physiological signal(s) underlying this increase in ALA synthase activity are not clear at present since 1) it is observed only in partially heme-defective mutants and not in totally heme-deficient mutants, and 2) the expression of ALA synthase has been shown to be regulated transcriptionally by a composite array of activation and repression (Keng and Guar- ente, 1987). In contrast, the elevated coproporphyrinogen oxidase activity in Sm12, especially when grown in glucose was not surprising, since it has been shown that synthesis of the enzyme is transcriptionally regulated by heme in a nega-

Yeast Ferrochelatase Mutations 8543

TABLE I Characteristics of the heme biosynthetic pathway in the parent (wild-type) and mutant Sm12 and Sm41 strains

The cells were grown in either 2% glucose (Glu) or 2% ethanol + 2% glycerol (Eth) medium and harvested in their exponential phase. Hemes and ALA were measured on whole cells. Porphyrins were measured both in whole cells and in the culture medium, and the total (intracellular + excreted) is reported. Porphyrins were 90-95% protoporphyrin in the mutants and 8045% coproporphyrin in the parent and are expressed as such. ALA synthase and coproporphyrinogen oxidase activities were measured in cell-free extracts. All measurements were performed with the same cell culture. Results given are the averages for duplicate determinations made in a t least three separate experiments. The ranges were 20% of the average values. ND: not determined.

Intracellular (/g dry wt) Activity

Strain Carbon source Doubling time Total

Total hemes porphyrins (+excreted)

ALA Coproporphyrinogen ALA synthase oxidase

h nmol pmol nmollhlmg Wild-type Glu

Eth Sm12 Glu

Eth

1.9 80 3.4 235 2.2 11 5 105

Sm41 Glu 1.8 70 Eth 3.7 160

n d

a

W T I Sm 12

, . " . I . . . . I " . I

5 00 550 600 nm

FIG. 1. Low temperature absorption spectra of whole cells of the parent DCT1-3D (WT) and mutant Sm12 and Sm41 strains. The cells were grown in ethanol + glycerol medium, supplemented when noticed with 0.2% Tween 80 (T), and harvested during exponential growth. Spectra were recorded at liquid-nitrogen temperature with 40 mg dry weight of cells. Reduction was achieved by endogenous substrates.

tive fashion (Zagorec and Labbe-Bois, 1986; Zagorec et al., 1988).

Immunodetection of the Mutant Ferrochelatases-The two mutants Sm12 and Sm41 made normal amounts of normal mature-sized ferrochelatase, as estimated by Western blot analysis of total cell proteins or membrane-bound proteins (Fig. 2). Normal amounts of full-sized HEM15 mRNA were also found by Northern blot analysis (data not shown). This indicates that there was no appreciable alteration in the processing, structure, or amount of the mutant enzyme proteins and that point mutations in the coding region of the

6 0.9 2.85 0.25 19 1.6 5.5 0.25 88 1.6 4.6 2.4

211 4.5 10.5 0.8

32 1.7 ND ND 60 1.2 ND ND

A B

FIG. 2. Immunodetection of ferrochelatase in total and membrane proteins prepared from the parent and mutant strains. Total proteins from trichloroacetic acid treated cells (20 pg) (panel A ) and membrane proteins (20 pg) (panel B ) were fractionated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membranes were reacted first with yeast ferrochelatase antiserum and then with alkaline phosphatase-conju- gated anti-rabbit secondary antibodies. Lane 1 , wild-type strain; lane 2, mutant Sm12; lane 3, mutant Sm41; lane F, isolated ferrochelatase.

HEM15 gene are probably the cause of the enzyme defects. Identification of the Mutations in the Mutant hem15 Al-

leles-The mutations were identified by sequencing the mutant hem15 alleles in mutant strain DNA after in vitro amplification by polymerase chain reaction. The 1.26-kilobase amplified DNA fragments encompassed the entire HEM15 coding region plus 54 and 26 nucleotides of the 5'- and 3'- flanking regions, respectively. A single nucleotide change was found in each mutant allele. The heml5-3 allele of Sm12 strain contained a C to T transition a t codon 169 that caused a Ser to Phe amino acid substitution. The hem15-4 allele of Sm41 had a T to C transition at codon 174 resulting in a Ser to Pro amino acid change (Fig. 3). The two mutations are located near each other and lie in an evolutionarily conserved region of the protein rich in hydroxylated amino acid residues. In particular, the mutation in mutant Sm12 substitutes a Phe for an invariant Ser (Fig. 3B). These results would suggest that the two mutations affect ferrochelatase functioning in the same manner, but to a different extent.

Measurement of the Mutant Ferrochelatase Actiuity-Pre- liminary measurements of the activity of the membrane- bound mutant ferrochelatases were made by quantifying the protoheme synthesized under anaerobic conditions from ferrous iron (150 PM) and protoporphyrin IX (40 pM) by its pyridine hemochromogen spectrum (Camadro and Labbe, 1981). The mutant enzymes were unexpectedly more active than the wild-type enzyme, 2-3-fold for Sm12 and 1.2-1.5- fold for Sm41. But assays with varying concentrations of

a544 Yeast Ferrochelatase Mutations A

Sm12 Sm41 W T

A C G T A C G T A C G T

B . t t . * t * * * . . * Human 189 T Q - - Q Y - C - - T - S S L N A I Y - Y Y

E.coli 126 P L - - Q F - C - - V - A V W D E L A - I L

Yeast 161 S Q Y P H F S Y S T T G S S I N E L W R Q I + + sm12 Sm4 1

FIG. 3. Amino acid substitutions in the mutant ferrochelatases are located in an evolutionarily conserved region. Panel A, nucleotide sequences of the polymerase chain reaction-amplified DNA of the mutants showing the single C to T (codon 169 TCC to TTC) and T to C (codon 174 TCC to CCC) transitions in the ferrochelatase hem25 alleles of mutant Sm12 and Sm41 strains, respectively. Panel B, the sequences of the human and E. coli ferrochelatases are from Nakahashi et al. (1990) and Miyamoto et al. (1991), respectively. Identical amino acids (-) are marked by an asterisk (*) and conservative ones by a dot (0) above the sequences. Amino acid substitutions in the yeast mutant enzymes are indicated by the arrows.

substrates gave inconsistent results. We therefore used a more sensitive test to analyze the defects of the mutant enzymes which permitted direct spectrofluorimetric recording of zinc- protoporphyrin formation (Camadro and Labbe, 1988). Zn2+ can be used in place of Fez+ since 1) the yeast enzyme has a similar affinity for both metals, 2) there is reciprocal competitive inhibition between them, and 3) there is no change in the protoporphyrin affinity whatever metal is used (Camadro and Labbe, 1982, 1988). Thus, both metals appear to bind to the same site and are used in the same fashion by ferrochelatase.

The initial velocity of zinc-protoporphyrin formation was constant for a t least 2-3 min (Fig. 4) and was proportional to protein concentration in the range used (0.1-1 mg protein/ assay). Tween 80 used to pseudosolubilize protoporphyrin was kept at 0.2 mg/ml, since higher amounts (0.8 mg/ml) inhibited both mutant enzymes: 3-4-fold for Sm12 and 1.2-1.5-fold for Sm41. Both mutant enzymes were totally inhibited by 1 mM palmitic acid (20 pl/assay of 20 mg of palmitic acid/ml of dimethyl sulfoxide). This was suerising, and yet unexplained, since palmitic acid is required for full activity of purified yeast and rat ferrochelatases (Camadro and Labbe, 1988; Taketani and Tokunaga, 1981) and is often added to ferrochelatase assays. Last, we verified that the non-heme ferrous iron present in the membranes (14-17 pg/lOO mg of protein), which can be used in part by ferrochelatase but very slowly (Camadro and Labbe, 1982; Tangeras, 1985), was not inhibitory to the zinc-chelatase activity; the activity did not change in presence of 30 p~ EDTA and excess Zn2+ (100 p ~ ) .

Kinetic Analysis of the Mutant Ferrochelatase-The two

Sm12 Sm41 0.4 ,

I ~ ~ ~ ~ . ~ - ~ . ~ . . ~ . ~ ~ ~ - ~ . ~ ~ ~ .

T i m e (min)

FIG. 4. Direct spectrofluorimetric assay for measuring the Zn2+-protoporphyrin chelatase activity of ferrochelatase in membranes of the parent DCT1-3D (WT) and mutant Sm12 and Sm41 strains. The membranes were prepared and the assays carried out as under “Materials and Methods.” The reaction mixture (3 ml final) contained 0.1 M Tris-HCI, pH 7.6,0.02% Tween 80, 1 pM protoporphyrin IX, 0.5, 0.2, or 0.4 mg of protein for the wild-type, Sm12, or Sm41 strain, respectively. The reaction was initiated by adding membrane protein ( E ) . Zn2+ (Zn) was added at a final concentration of 30 p~ for Sm12 and 2 p~ for Sm41. The excitation and emission wavelengths were 418 and 588 nm, respectively.

0.5 1 1.5 [Protoporphyrin] ( P M )

FIG. 5. The mutant ferrochelatases have lower affinity for protoporphyrin and higher V,, than the wild-type enzyme. Initial velocities (nanomoles of zinc-proto/hour/milligram of protein) were measured as under “Materials and Methods” and Fig. 4, a t different concentrations of protoporphyrin with a constant concentration of Zn2+: 30 p~ for Sm12,lO p~ for Sm41, and endogenous for the parent ( WT) strain. The curves were obtained with the EZ-FIT program. For clarity, only a few experimental points are shown.

mutant enzymes had lower affinities for Zn2+ than the normal enzyme (Figs. 4 and 6A). Zn2+ had to be added to the reaction mixture for maximal activity of the mutant enzymes, while the wild-type enzyme was already saturated with the endogenous Zn2+ present in the assay. The two mutant enzymes also had lower affinities for protoporphyrin, and they were considerably more active, especially Sm12, than the wild-type enzyme (Figs. 5 and 6B) .

The individual kinetic parameters were obtained from the secondary plots derived from the primary double-reciprocal plots of initial rates versus concentration of one substrate at various fixed concentrations of the second substrate. They were identical to the values calculated using the “EZ-FIT” program (Perrella, 1988). The results are reported in Table 11. The high activities of the mutant enzyme were due to increased V,,,,,. The KM for protoporphyrin was increased to

Yeast Ferrochelatase Mutations 8545

A

V -1 Om6 t

0.5 1 1.5 Ef n ++ (A M)]-l

/// Sm41

Sm12

10 20 30 [Protoporphyrin (AM)]"

FIG. 6. Double-reciprocal plots of zinc-chelatase activities of wild-type and mutant ferrochelatases as function of substrate concentration. Initial velocities (nanomoles of zinc-proto/ hour/milligram of protein) were measured at different concentrations of ZnZ+ with a fixed concentration of protoporphyrin (2 WM) (panel A ) , and at different concentrations of protoporphyrin with a fixed concentration of Zn2+ (30, 5, and endogenous for Sm12, ,311141, and wild-type ( WT) enzymes, respectively) (panel B) .

TABLE I1 Kinetic parameters of mutant ferrochelatases

The zinc-protoporphyrin activity of wild-type (WT) and mutant Sm12 and Sm41 ferrochelatases was measured as in Fig. 6, at different fixed concentrations of one substrate while varying the concentration of the second substrate. The KM ( p M ) and V,, (nmol Zn-proto/h/mg of protein) values were calculated graphically from the secondary double-reciprocal plots and by using the EZ-FIT program. The results are the averages of a t least three determinations with the standard deviations. The apparent (app) values for the wild-type enzyme were measured in this work, while the KM values for ZnZ+ were taken from Camadro and Labbe (1982,1988) for the yeast membrane-bound and purified enzyme, respectively. Numbers in brackets represent the increase with respect to the wild-type values.

Strain KM

Protoporphyrin Zn2+ V-

WT 0.03 app 0.15, 0.25 Sm12

9.5 app 1.4 k 0.2 7.5 f 3 100 f 20

Sm41 0.3 f 0.05 0.8 f 0.2 30 f 5 (-45) (-35) (-10)

(-10) (-4) (-3)

nearly the same relative extent as was Vmax for each mutant. But the KM for Zn2+ was relatively more increased in Sm12 than in Sm41. Although these changes of the mutant parameters are expressed with respect to wild-type values which are

not directly comparable or related, they correspond surpris- ingly well with what had been anticipated: similar defects should affect both mutant ferrochelatases but to different extents.

The catalytic efficiency (Vm,,/KM) of the Sm12 ferrochelatase was decreased in almost the same proportion for the two substrates. Since protoporphyrin accumulates in uiuo, it should not be limiting and it is likely that the concentration of iron is decisive for heme production. We tried to increase intracellular iron by growing the cells in the presence of excess ferric iron (180 PM, which is 18-fold the iron concentration of YPE medium). The cells made only half as much protoporphyrin and two-thirds of normal ALA, but there was no change in heme formation when the whole cells or membrane preparations were tested for cytochromes and heme content (data not shown). Heme synthesis in Sm41 cells was also not increased by iron supplementation. In fact, the concentration of intracellular non-heme iron was only slightly increased in the mutant and wild-type cells growing in iron-enriched medium (2.5-3 gmol/g dry weight) as compared to normal medium (2-2.5 wmol/g dry weight), indicating that the cells did not accumulate significantly more iron when challenged with higher extracellular iron concentration. This is because iron uptake is limited by the activities of both the ferrireductase (KM for ferric iron, 3 PM)* and the ferrous iron transporter (KM for ferrous iron, 0.15 g~ (Eide et al., 1992)) which are involved in the reductive assimilation of iron in S. cereuisiae (Lesuisse and Labbe, 1989).

DISCUSSION

We have identified two point mutations in the yeast ferrochelatase gene that change two amino acids lying close together, and we have analyzed their functional consequences on enzyme behavior i n uiuo and in uitro. The mutant cells synthesized less heme than normal and accumulated and excreted large amounts of protoporphyrin. The i n vivo rate of heme synthesis can be estimated from the heme content of whole cells, the generation time, and knowing that -150 mg of membrane protein is recovered from 1 g dry weight of cells. These rates are 0.45, 0.15, and 0.3 nmol heme/h/mg membrane protein for the wild-type, Sm12 and Sm41 mutant strains, respectively. They represent 4.5, 0.15, and 1% of the maximal velocity measured i n uitro for the zinc-chelatase activity of the wild-type and mutant enzymes (Table 11). Therefore, the mutant enzymes in Sm12 and Sm41 strains function in vivo 30 and 4.5 times less efficiently than the wild- type enzyme, which correlates remarkably well with the relative increase in their KM for Zn2+ (35 and 4, Table 11). This suggests that iron is limiting for heme production in the mutant cells and that no compensatory mechanism increases the concentration of iron in the mitochondrial membrane or matrix. Iron metabolism and heme synthesis are not directly coupled in Rhodopseudomonas sp. (Moody and Dailey, 1985) and the mouse liver (Tangeras, 1986), but inhibition of heme synthesis stimulates iron uptake mediated by transferrin receptor in rat reticulocytes (Adams et al., 1989).

The amino acid changes, Ser-169 to Phe in Sm12 and Ser- 174 to Pro in Sm41 mutants, are located near each other and affect ferrochelatase function in the same manner but to different degree. The functional consequences of the change of Ser-169, which is conserved among the five ferrochelatases known to date, are much more severe than those of the change of Ser-174, which is replaced by Val in the bacterial enzymes. The two mutant enzymes have higher KM for both substrates,

P. Labbe, unpublished results.

8546 Yeast Ferrochelatase Mutations

protoporphyrin and metal. This suggests that the binding sites of the two substrates are not independent of each other, in agreement with the model of an ordered sequential enzyme mechanism where iron binds prior to porphyrin (Dailey and Fleming, 1983). The mutations may have identified a region of ferrochelatase that is involved in the binding of both substrates and, probably, they have affected the structure of the binding sites. This region (Fig. 3B) consists of a central segment rich in hydroxylated residues which might be part of the metal-binding domain, flanked on both sides by regions rich in conservative aliphatic or aromatic residues that might be involved in hydrophobic interactions with porphyrin.

The increased V,, of the mutant ferrochelatases, almost proportional to their increased KM for protoporphyrin, was surprising. However, a similar situation has been described in the analysis of porphyrin specificity of ferrochelatase: hydrophobic substituents at positions 1, 2, 3, and 4 on pyrroles A and B lowered both the Vmax and the KM for porphyrin of the enzyme (Honeybourne et al., 1979; Dailey and Fleming, 1983). This inverse relationship between the affinity for porphyrin and the rate of heme formation suggests that the release of heme might control the overall reaction rate. The natures of the substituents at positions 2 and 4 are important for pro- ducing a porphyrin either substrate or competitive inhibitor of ferrochelatase, regardless of its binding ability (Dailey and Fleming, 1983; Dailey et al., 1989). These substituents also affect the inhibitory activity of the N-alkyl porphyrins (De Matteis et al., 1985; Mccluskey et al., 1989). It is significant that an antibody raised to N-methylmesoporphyrin (2-, 4- ethyl) did not catalyze the metallation of protoporphyrin (2-, 4-vinyl) or deuteroporphyrin (2-, 4-H) (Cochran and Schultz, 1990). All these results indicate a role for the 2- and 4-vinyl groups in the optimal binding of the porphyrin into the ferrochelatase active site. We propose that the region of the enzyme affected by the mutations contributes to the hydrophobic interaction(s) and van der Waals contact(s) of the active site with the porphyrin vinyl group(s).

Detailed analysis of these mutant enzymes for their affinity for porphyrins substituted at the 2- and 4-positions and for their inhibition by different isomers of N-alkyl porphyrins, together with analysis of mutant enzymes obtained by site- directed mutagenesis of invariant amino acids, will help clarify the way in which the enzyme binds its substrates, distorts the porphyrin ring, and releases heme after chelation of iron.

Acknowledgments-We thank J. Rytka for providing the yeast mutant strains, J.M. Camadro for assistance with immunological methods, P. Labbe for helpful advice on the ferrochelatase and iron assays and 0. Parkes for his help in preparing the manuscript.

REFERENCES Adams, M. L., Ostapiuk, I., and Grasso, J. A. (1989) Biochim. Biophys. Acta

Blom, C., Klasen, E. C., and Van Steveninck, J. (1990) Biochim. Biophys. Acta

Bloomer, J. R., Hill, H. D., Morton, K. O., Anderson-Burnham, L. A,, and

Brenner, D. A., and Frasier, F. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 849-

1012,243-253

1039,339-342

Straka, J. G. (1987) J. Biol. Chem. 262,667-671

A511 Camadro, J. M., and Labbe, P. (1981) Bioehimie 63,463-465 Camadro, J. M., and Labbe, P. (1982) Biochim. Biophys. Acta 707,280-288 Camadro, J. M., and Labbe, P. (1988) J. Biol. Chem. 263,11675-11682 Cochran, A. G., and Schultz P. G. (1990) Science 249,781-783 Dailey, H. A. (1990) in Biosynthesis of Heme and Chlorophylk (Dailey, H. A,,

Dailey, H. A,, Jones, C. S., and Karr, S. W. (1989) Biochim. Biophys. Acta 999, Dailey, H. A,, and Fleming, J. E. (1983) J. Biol. Chem. 258,11453-11459

ed) pp. 123-163, McGraw-Hill, New York

1 - 1 1 1 - 1 1

De Matteis, F., Gibbs, A. H., and Harvey, C. (1985) Biochem. J. 226,537-544 Eide, D., Davis-Kaplan, S., Jordan, I., Sipe, D., and Kaplan, J. (1992) J. Biol.

Frustaci, J. M., and O'Brian, M. R. (1992) J. Bacteriol. 174,4223-4229 Honevbourne. C. L.. Jackson. J. T.. and Jones. 0. T. G. (1979) FEBS Lett. 98.

Chem. 267,20774-20781

207"210 , , I , . ,

Keng, T., and Guarente, L. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 9113-

Kurlandzka, A., and Rytka, J. (1985) J. Gen. Microbiol. 131,2909-2918

Lamoril, J., Boulechfar, S., de Verneuil, H., Grandchamp, B., Nordmann, Y., Labbe-Bois, R. (1990) J. Biol. Chem. 266,7278-7283

and Deybach, J. C. (1991) Biochem. Biophys. Res. Commun. 181,594-599 Lesuisse, E., and Labbe, P. (1989) J. Gen. Microbiol. 136,257-263 Mccluskey, S. A., Whitney, R. A., and Marks, G. S. (1989) Mol. Pharmacol. 36,

Mivamoto. K.. Nakahieashi, K.. Nishimura. K.. and Inokuchi. H. (1991) J. Mol.

9117

608-614

Biol. 219,393-398 ' '

. .

Moody, M. D., and Dailey, H. A. (1985) J. Bacteriol. 161,1074-1079 Nakahashi, Y., Taketani, S., Okuda, M., Inoue, K., and Tokunaga, R. (1990)

Nakahigashi, K., Nishimura, K., Miyamoto, K., and Inokuchi, H. (1991) Proc.

Nordmann, Y., and Deybach, J. C. (1990) in Biosynthesis of Heme and Chloro-

Perrella, F. W. (1988) Anal. Biochem. 174,437-447 Rytka, J., Bilinski, T., and Labbe-Bois, R. (1984) Biochem. J. 218,405-413 Sherman, F., Fink, G. R., and Hicks, J. B. (1986) Laboratory Course Manual

for Methods in Yeast Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Straka, J. G., Hill, H. D., Krikava, J. M., Kools, A. M., and Bloomer, J. R. (1991) Am. J. Hum. Genet. 48,72-78

Taketani, S., and Tokunaga, R. (1982) Eur. J. Bzochem. 127,443-447 Taketani, S., and Tokunaga, R. (1981) J. Biol. C k m . 256,12748-12753

Taketani, S., Nakahashi, Y., Osumi, T., and Tokunaga, R. (1990) J. Bzol. Chem.

Tangeras, A. (1985) Biochim. Biophys. Acta 843,199-207 Tangeras, A. (1986) Biochim. Biophys. Acta 882, 77-84 Tangeras, A., Flatmark, T., Backstrom, D., and Ehrenberg, A. (1980) Biochim.

Urban-Grimal, D., and Labbe-Bois, R. (1981) Mol. Gen. Genet. 183,85-92 Zagorec, M., and Labbe-Bois, R. (1986) J. Biol. Chem. 261,2506- 2509 Za orec, M., Buhler, J. M., Treich, I., Keng, T., Guarente, L., and Labbe-Bois,

Biochem. Biophys. Res. Commun. 173,748-755

Natl. Acad. Sci. U. S. A. 88,10520-10524

phylls (Dailey, H. A., ed) pp. 491-542, McGraw-Hill, New York

265,19377-19380

Bwphys. Acta 589,162-175

i. (1988) J. Biol. Chem. 263,9718-9724

JOURNAL OF BIOLOGICAL 268, No. 12, 8541-8546, 1993 0 1993 ... · 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol.

Documents