Download Aminolaevulinic acid synthase of Rhodobacter capsulatus: high

Biochem. J. (2013) 451, 205–216 (Printed in Great Britain)
205
doi:10.1042/BJ20121041
Aminolaevulinic acid synthase of Rhodobacter capsulatus : high-resolution
kinetic investigation of the structural basis for substrate binding and
catalysis
Anna-Lena KAUFHOLZ*, Gregory A. HUNTER†, Gloria C. FERREIRA†1 , Thomas LENDRIHAS†, Vanessa HERING*,
Gunhild LAYER*, Martina JAHN* and Dieter JAHN*1
*Institute of Microbiology, Technical University of Braunschweig, Spielmannstraße 7, D-38106 Braunschweig, Germany, and †Department of Molecular Medicine, University of South
Florida, Tampa, FL 33612, U.S.A.
The first enzyme of haem biosynthesis, ALAS (5-aminolaevulinic
acid synthase), catalyses the pyridoxal 5 -phosphate-dependent
condensation of glycine and succinyl-CoA to 5-aminolaevulinic
acid, CO2 and CoA. The crystal structure of Rhodobacter
capsulatus ALAS provides the first snapshots of the structural
basis for substrate binding and catalysis. To elucidate the
functional role of single amino acid residues in the active site
for substrate discrimination, substrate positioning, catalysis and
structural protein rearrangements, multiple ALAS variants were
generated. The quinonoid intermediates I and II were visualized
in single turnover experiments, indicating the presence of an αamino-β-oxoadipate intermediate. Further evidence was obtained
by the pH-dependent formation of quinonoid II from the product
5-aminolaevulinic acid. The function of Arg21 , Thr83 , Asn85
and Ile86 , all involved in the co-ordination of the succinylCoA substrate carboxy group, were analysed kinetically. Arg21 ,
Thr83 and Ile86 , all of which are located in the second subunit
to the intersubunit active site, were found to be essential. Their
location in the second subunit provides the basis for the required
structural dynamics during the complex condensation of both
substrates. Utilization of L-alanine by the ALAS variant T83S
indicated the importance of this residue for the selectiveness of
binding with the glycine substrate compared with related amino
acids. Asn85 was found to be solely important for succinyl-CoA
substrate recognition and selectiveness of binding. The results of
the present study provide a novel dynamic view on the structural
basis of ALAS substrate-binding and catalysis.
INTRODUCTION
ALAS catalysis was complicated. Obviously, the static view
provided by the ALAS crystal structure had to be supplemented by
functional kinetic data. Since the PLP cofactor undergoes multiple
changes in its electronic properties during substrate binding and
catalysis, the kinetics of the different partial reactions can be
followed spectroscopically [9,10]. Consequently, we employed
the R. capsulatus ALAS system to answer the open questions
concerning the ALAS enzymatic mechanism and the contribution
of active-site residues.
Besides ALAS, there are only three other PLP-dependent
enzymes known to catalyse the mechanistically unusual
cleavage of two α-carbon bonds, namely AONS (8-amino-7oxononanoate synthase) [11], SPT (serine palmitoyltransferase)
[12,13] and KBL (2-amino-3-ketobutyrate-CoA ligase) [14,15].
These few enzymes constitute the α-oxoamine synthase family
of PLP-dependent enzymes [1]. AONS catalyses the first four
steps in biotin biosynthesis, the decarboxylative condensation
of L-alanine and pimeloyl-CoA to (7S)-AONS [16]. SPT
catalyses the condensation of L-serine and palmitoyl-CoA to
form 3-oxodihydrosphingosine, the initial step of sphingolipid
biosynthesis [17,18]. Finally, KBL catalyses the conversion of
2-amino-oxobutyrate, the product of threonine degradation in
the presence of CoA, into glycine and acetyl-CoA [14,15].
The current knowledge of the exact chemistry of α-oxoamine
synthase catalysis comes from a combination of spectroscopic,
PLP (pyridoxal 5 -phosphate)-dependent enzymes catalyse a
broad variety of reactions such as decaboxylations, transaminations, racemizations, eliminations, retro-added cleavages
and Claisen-type condensations [1,2]. The first enzyme of
tetrapyrrole biosynthesis in non-plant eukaryotes and the αsubclass of purple bacteria, the homodimeric ALAS (5aminolaevulinic acid synthase; EC 2.3.1.37), is a PLP-dependent
enzyme [3–6]. Multiple mutations of the human enzyme are
connected with the detrimental disease X-linked sideroblastic
anaemia [7]. In a Claisen-type condensation reaction, involving
the untypical cleavage of two amino acid α-carbon bonds,
succinyl-CoA and glycine are converted by ALAS into CO2 ,
CoA and the general haem precursor ALA (aminolaevulinic
acid) [8].
A few years ago we solved the, as of yet, only crystal structure
of ALAS from Rhodobacter capsulatus in various complexes
with its cofactor PLP and substrates [3]. A detailed molecular
view of the ALAS active site at different stages of catalysis
was therefore then accessible. Owing to the high flexibility of
the enzyme, the existence of multiple potential catalytic amino
acid residues, highly complex substrate co-ordination and an
intersubunit active site composed of amino acid residues from
both subunits, prediction of the step-by-step chemistry underlying
Key words: 5-aminolaevulinic acid synthase (ALAS), haem
biosynthesis, quinonoid intermediate formation, steady-state
kinetics, stopped-flow spectroscopy.
Abbreviations used: ALA, aminolaevulinic acid; ALAS, 5-ALA synthase; AONS, 8-amino-7-oxononanoate synthase; DTT, dithiothreitol; hemA,
5-aminolaevulinic acid synthase; IPTG, isopropyl β-D-thiogalactopyranoside; KBL, 2-amino-3-oxobyturate-CoA ligase; α-KGD, α-oxoglutarate
dehydrogenase; mALAS2, mouse ALAS 2, erythroid; PLP, pyridoxal 5 -phosphate; SPT, serine palmitoyltransferase.
1
Correspondence may be addressed to either of these authors (email [email protected] and [email protected]).
c The Authors Journal compilation c 2013 Biochemical Society
206
Figure 1
A.-L. Kaufholz and others
Postulated reaction mechanism
The catalytic path of the R. capsulatus enzyme with all postulated intermediates is shown. The ALAS reaction follows a Bi Bi kinetic mechanism. The substrate glycine binds first followed by the
second substrate, succinyl-CoA. The products CoA and CO2 are released prior to ALA. The last step is the regeneration of the holoenzyme.
radiolabelling, kinetic and structural biology studies using
various members of the enzyme family [11–13,15,16,19–
21].
Whereas for some α-oxoamine synthase family members, like
AONS, a detailed view of the molecular basis of catalysis is
available, major open questions remain for ALAS catalysis.
The known part of the ALAS reaction mechanism follows
the typical α-oxoamine synthase catalysis. It is an ordered Bi
Bi kinetic mechanism with glycine binding prior to succinylCoA, and the release of ALA after carbon dioxide and CoA
[3,22,23]. As shown in Figure 1 the PLP cofactor is covalently
bound to an active site lysine residue (Lys248 ) as a Schiff-base
linkage at its ε-amino group [9]. This form of the enzyme is
called the internal aldimine. The first step of ALAS catalysis
is the formation of the external aldimine via the breakage of
the Schiff-base bond of PLP to Lys248 and the association of the
substrate glycine with the enzyme-bound PLP. The internal
c The Authors Journal compilation c 2013 Biochemical Society
aldimine formed via transaldimination can follow two different
potential routes for the generation of the PLP-bound product ALA,
both involving quinonoid intermediates. Decarboxylation of PLPbound glycine would lead to the formation of a transient quinonoid
intermediate followed by Claisen condensation with succinylCoA. Release of CoA would lead to the product aldimine.
Alternatively, a quinonoid intermediate I can be generated by
stereo-specific, Lys248 -catalysed abstraction of the pro-R-proton
from the PLP-bound glycine. Most likely, the condensation of
the resulting quinonoid intermediate I with the second substrate
succinyl-CoA would lead to the formation of a 2-amino-3oxoadipate intermediate. The release of CoA from this tetrahedral
intermediate and decarboxylation of the generated α-amino-βoxoadipate–ALAS aldimine would then lead to the formation
of quinonoid intermediate II. The last steps are protonation of
the potential quinonoid intermediate II with the formation of the
ALAS–ALA aldimine. ALA release and finally regeneration of
Aminolaevulinic acid synthase
the internal aldimine in a transaldimination reaction is common
to both pathways [9,23].
The existence of the α-amino-β-oxoadipate intermediate has
not been experimentally confirmed for ALAS catalysis as yet, but
it is known that the corresponding intermediates occur in AONS
reactions [21]. The enzyme–glycine complex crystal structure
revealed that Asn54 of R. capsulatus ALAS forms hydrogen
bonds to one of the oxygen atoms of the carboxy group of the
glycine substrate. After condensation with succinyl-CoA and the
indicative formation of an α-amino-β-oxoadipate intermediate,
this carboxy group is released as CO2 . Consequently, this active
site asparagine residue might be the crucial residue controlling
ALAS catalysis.
We generated two variants of R. capsulatus ALAS, N54Q
and N54D, with the aim of trapping the quinonoid I as well
as the α-amino-β-oxoadipate intermediate in the active site.
Moreover, enzyme assays using O-methylglycine as the first
substrate were performed. After the addition of succinyl-CoA,
the resulting β-oxoacidmethylester aldimine cannot undergo
enzymatic decarboxylation and should accumulate. Beside the
question regarding the oxoadipate intermediate, questions of
the functional dynamic contribution of amino acid residues
within the active site to binding and discrimination against nonsubstrate compounds of similar structure still remain. Again, the
crystal structure of R. capsulatus ALAS only provides a static
view of the active site of the final substrate co-ordination. The first
substrate glycine residue is tightly co-ordinated via a network of
bonds to Arg374 , Asn54 and Ser189 . The adenine ring of succinylCoA binds near the enzyme surface and its pantothenate portion
extends down a cleft towards the PLP cofactor, over 20 Å (1
Å = 0.1 nm) away from the surface. This solvent-excluded activesite environment provides highly specific molecular recognition of
succinyl-CoA. The succinyl-CoA carboxylate group is tightly coordinated via a salt bridge to Arg21 , two hydrogen bonds to Thr83
and Asn85 and van der Waals interactions with Ile86 (Figure 2)
[3]. However, it remains to be determined which amino acids
are functionally involved in the discrimination against other nonsubstrate compounds and which are important for catalysis.
To investigate the contribution of altered ALAS amino acids
to substrate recognition, catalysis and product release, several
ALAS variants carrying conservative and non-conservative amino
acid exchanges (N54D, N54Q, R21K, R21E, T83S, N85Q,
N85F and I86H) were constructed, recombinantly produced,
purified and kinetically characterized. Furthermore, multiple
substrate analogues were tested with the wild-type ALAS
and selected variants. The amino acids required for glycine
(Thr83 ) and succinyl-CoA (Asn85 ) substrate discrimination against
structurally related compounds were functionally identified. The
essential contribution of Arg21 , Thr83 and Ile86 , all located in the
second subunit of the intersubunit active site, underscored
the importance of the structural flexibility required for the
complex catalyses performed by ALAS.
EXPERIMENTAL
Materials
Ampicillin, PLP, BSA, succinyl-CoA sodium salt, ALA
hydrochloride, α-oxoglutaric acid, α-KGD (α-oxoglutarate
dehydrogenase), Hepes free acid, Mops, TPP (thiamine pyrophosphate), L-alanine, D-alanine, L-cysteine, L-threonine, L-serine,
octanoyl-CoA, butyryl-CoA and NAD + were purchased from
Sigma–Aldrich. Glucose, glycerol, glycine, magnesium chloride
hexahydrate, O-methylglycine and potassium hydroxide were
from Fisher Scientific or Roth. Benzonase was purchased from
Figure 2
207
Active site of R. capsulatus ALAS
The substrates succinyl-CoA (turquoise) and glycine (salmon), the PLP cofactor bound to the
essential active site lysine (light brown) and selected amino acids (grey or green) are shown as
sticks. Monomer 1 is highlighted in light green and monomer 2 in light grey. Additional residues
from the second monomer are marked with an asterisk. Hydrogen bonds are displayed by broken
lines. The cofactor PLP and Lys248 , the lysine residue involved in PLP-binding and catalysis,
are marked in light brown. The residue Arg21 from monomer 1 and the residues Thr83 , Asn85
and Ile86 from monomer 2 are all involved in the co-ordination of the succinyl-CoA substrate
carboxylate group. This co-ordination is carried out via a salt bridge by Arg21 , two hydrogen
bonds to Thr83 and Asn85 , and van der Waals interaction with Ile86 . Asn54 from monomer 1 can
stabilize or destabilize the α-amino-β-oxoadipate intermediate. The formation of the quinonoid
intermediate II is acid-catalysed by His142 (from monomer 1), which is located directly above the
cofactor ring. The Figure was prepared using PyMOL (http://www.pymol.org) and PDB codes
2BWN, 2BWO and 2BWP.
Merck, whereas the QuikChange® site-directed mutagenesis kit
was provided by Agilent Technologies. CompleteTM Mini protease
inhibitor was purchased from Roche and chloramphenicol and
glutathione from Roth. The SDS/PAGE reagents, Econo-Pac
Chromatography columns and the Bradford protein assay buffer
were from Bio-Rad Laboratories. T4 DNA ligase was from
New England Biolabs and Phusion DNA polymerase was
from Finnzymes. The oligonucleotides for sequencing were
purchased from MWG-Biotech. PreScissionTM Protease and
Glutathione SepharoseTM 4 Fast Flow were purchased from
GE Healthcare. IPTG (isopropyl β-D-thiogalactopyranoside) and
DTT (dithiothreitol) were from Gerbu Biotechnik. The Lysing
matrix for cell digestion, Lysing Matrix B Bulk and the digestion
machine FastPrep® 24 were purchased from MP Biomedicals.
Protein concentration cells were purchased from Sartorius Stedim.
Bacterial strains and genomic DNA
The Escherichia coli strains used for cloning experiments and
protein production were DH10β (Invitrogen), BL21(DE3)pLysS
(Stratagene) and BL21(DE3)RIL (Merck).
Cloning of the R. capsulatus hemA (5-aminolaevulinic acid
synthase)
The genomic DNA of R. capsulatus DSM 938 was from the
DSMZ. Primers with two different restriction sites, BamHI
for the forward primer and XhoI for the reverse primer,
were ordered from Metabion with the aim of subsequent
cloning of the generated PCR fragment into the pGEX-6P1 vector. The primers sequences were: 5 -CGCGGATCCGCGATGGACTACTACAATCTCGCG-3 (forward) and 3 -CCGCTCGAGCGGTCACGCACAGCGCGCCCA-5 (reverse). The
amplified hemA sequence was cloned into the expression vector
pGEX-6P-1 and the accuracy of the resulting plasmid was verified
by DNA sequencing.
c The Authors Journal compilation c 2013 Biochemical Society
208
A.-L. Kaufholz and others
Site-directed mutagenesis
Mutagenesis experiments for R. capsulatus were performed
according to manufacturer’s instructions with the Qiagen
site-directed mutagenesis kit. The employed oligonucleotides
for the ALAS variants were: 5 -CGAGGGACGTTACAAGACGTTCATCG-3 for R21K, 5 -CGAGGGACGTTACGAGACGTTCATCG-3 for R21E, 5 -GGTTCGGGCGGCAGCCGCAACATCTC-3 for T83S, 5 -GCGGCACCCGCCAGATCTCGGGCAC-3 for N85Q, 5 -GCGGCACCCGCTTCATCTCGGGCAC-3 for N85F, 5 -GCACCC GCAACCACTCGGGCACCAC-3 for I86H, 5 -CTGGTGCGGCCAGGACTATCTGGGC-3 for N54Q and 5 -CTGGTGCGGCGACGACTATCTGGGC-3 for N54D (underlined residues represent the
introduced mutated codons for the desired amino acid exchanges
upon protein production). DNA sequencing was used for the
verification of the desired exchanges.
Protein purification, SDS/PAGE and protein concentration
determination
Recombinant R. capsulatus wild-type ALAS and its variants were
purified from E. coli BL21(DE3)pLysS or BL21(DE3)RIL cells.
A total of 10 litres (20 × 500 ml) of vapour-sterilized LB (Luria–
Bertani) medium containing 100 μg/ml ampicillin and 34 μg/ml
chloramphenicol in 1 litre Erlenmeyer flasks were each inoculated
with 5 ml of an overnight culture of either E. coli BL21(DE3)RIL
or BL21(DE3)pLysS cells, carrying pGEX-6P-1-Rc-A (pGEX6P-1 vector with the R. capsulatus wild-type hemA gene) and
incubated under vigorous aeration at 37 ◦ C. After the culture
reached a D600 of 0.6 the expression of hemA was initiated by
the addition of 200 μM IPTG. The optimal growth temperatures
were either 25 ◦ C or 17 ◦ C for 22 h and at 180 rev./min. Cells were
harvested by centrifugation for 20 min at 3000 g and 4 ◦ C. The
cell sediment was suspended in 15–40 ml of lysis buffer [20 mM
Hepes (pH 7.5), 200 mM NaCl, 10 mM DTT, 20 μM PLP, 20 μl
of benzonase (250 units/μl) and 1 tablet of CompleteTM protease
inhibitor]. Lysing Matrix B Bulk (125 mg/ml) was added into a
2-ml-reaction tube and filled up with suspended cell sediment.
Cells were disrupted using FastPrep® 24 at 4 ◦ C (three times for
45 s). Cell debris and insoluble protein fractions were removed by
centrifugation for 30–45 min at 4 ◦ C and 14 000 g. The resulting
supernatant was loaded on to a glutathione–Sepharose column and
eluted with a glutathione-containing buffer. Purity was assessed
by SDS/PAGE [24] and protein concentration was determined by
the bicinchoninic acid or Bradford assay method using BSA as
standard [25,26].
Steady-state kinetic analysis
Enzyme activity was determined using a continuous
spectrophotometric enzyme-coupled assay with ALAS and αKGD as the interacting enzymes at a constant temperature of
30 ◦ C [10]. The observed rates were fitted to the Michaelis–
Menten equation using the non-linear regression analysis software
program SigmaPlot (Systat Software). The NADH concentration
was calculated with the Beer–Lambert law as described previously
[27]. The α-KGD couples oxidation of CoA to succinyl-CoA
to the reduction of NAD + to NADH. The NADH production
is equivalent to the production of ALA and could be followed
spectrophotometrically at 340 nm [10].
Stopped-flow spectroscopy
The R. capsulatus ALAS reactions were performed in
100 mM AMPSO {3-[(1,1-dimethyl-2-hydroxyethyl)amino]-2
c The Authors Journal compilation c 2013 Biochemical Society
hydroxypropanesulfonic acid} (pH 9.5), containing 10 % (v/v)
glycerol at 30 ◦ C. The final concentrations of the reactants for the
R. capsulatus ALAS-catalysed reactions were 60 μM bacterial
ALAS, 20 mM glycine (O-methylglycine, L-alanine, D-alanine,
L-cysteine, L-threonine or L-serine) and 30 μM succinyl-CoA
(octanoyl-CoA or butyryl-CoA). Rapid-scanning stopped-flow
kinetic measurements were conducted using an OLIS model
RSM-1000 stopped-flow spectrophotometer. The dead time of this
instrument was approximately 2 ms, and the observation chamber
optical path length was 4.0 mm. Scans covering the wavelength
region 270–550 nm were acquired at a rate of 1000/s and averaged
to either 62 or 31 scans/s in order to condense the resulting
data files to a tractable size for data-fitting analysis. An external
water bath was utilized to maintain a constant temperature for the
syringes and observation chamber. The pre-steady-state kinetics
of quinonoid intermediate II formation and decay at 510 nm
were modelled from single-wavelength traces with KinTecSim
simulation software as described previously [28,29]. A fourkinetic-step mechanism as described by eqn (1) was utilized:
k1
k2
k3
k4
EG+SCoA ⇔ EGSCoA ⇔ EQ2 ⇔ EALA ⇔ E+ALA
(1)
where EG and SCoA represents the enzyme–glycine (alanine)
complex and succinyl-CoA prior to mixing, EGSCoA is an initial
collision complex, EQ2 is the quinonoid II intermediate, which is
in bold to denote that this intermediate was the input responsible
for the observed signal, EALA is the enzyme–product complex
and E + ALA represents the dissociated enzyme and product.
Single-wavelength traces with O-methylglycine as the amino
acid substrate were also modelled according to eqn (1), with the
mathematically inconsequential caveat that EQ2 was replaced
by EQ1, to signify that in this case it was considered to be the
quinonoid intermediate I that is being observed.
Substrate specificity
Substrate-specificity determinations were carried out at 30 ◦ C. The
experimental set-up was similar to that for the studies involving
steady-state kinetics. In one set of experiments, the specificity
towards glycine was examined by testing other amino acids while
maintaining succinyl-CoA as the second substrate. In another set
of experiments, glycine was kept as the first substrate, whereas
different acyl-CoA derivatives were tested as second substrate.
RESULTS AND DISCUSSION
Evidence for an α-amino-β-oxoadipate intermediate during ALAS
catalysis
First, we approached the open questions for the existence of an αamino-β-oxoadipate intermediate with combined mutational and
spectroscopic experiments. We intended to follow the formation
and the decay of the quinonoid forms I and II to collect evidence
for the proposed oxoadipate intermediate. Finally, experiments
for the isolation of the α-amino-β-oxoadipate intermediate are
outlined.
Detection of the quinonoid I intermediate
In this context, our initial expectation was to ‘trap’ the
reaction intermediate. To focus on the formation and decay of
the quinonoid intermediate I (Figure 1), the O-methylglycine
analogue was used instead of glycine. The substrate analogue
O-methylglycine carries a methyl group on the oxygen atom
Aminolaevulinic acid synthase
Figure 3
209
Pre-steady-state reactions of R. capsulatus ALAS
(A) Reaction of the R. capsulatus wild-type ALAS–glycine complex with succinyl-CoA under single turnover conditions. (B) Reaction of the R. capsulatus ALAS N54Q–glycine complex with succinyl-CoA
under single turnover conditions. (C) Reaction of R. capsulatus wild-type ALAS–O -methylglycine complex with succinyl-CoA. (D) Reaction of R. capsulatus ALAS N54Q–O -methylglycine complex
with succinyl-CoA. (E) Reaction of the R. capsulatus ALAS wild-type and variants involved in substrate recognition and co-ordination. Green, ALAS R21K; cyan, ALAS R21E; pink, ALAS T83S; blue,
ALAS N85F; dark green, wild-type ALAS. (F) Reaction of R. capsulatus ALAS T83S variant with glycine and L-alanine as the first substrates. Green, ALAS T83S reaction with L-alanine; red, ALAS
T83S reaction with glycine. In all cases the modelled fits are included as black lines.
which in glycine is hydrogen-bonded to Asn54 . Nevertheless,
O-methylglycine carries all determinants required for external
aldimine formation; however, the external aldimine cannot
decarboxylate to yield a quinonoid (pathway 1, see the
Introduction section). Analogously, after succinyl-CoA addition,
the generated methylester of the β-oxoacid–aldimine complex
cannot decarboxylate to yield the quinonoid intermediate II.
Succinyl-CoA binding to the ALAS protein was described
previously to be important for quinonoid I formation [30]. Thus
we monitored the formation of the quinonoid intermediate I in
the R. capsulatus ALAS-catalysed reaction using the substrate
analogue O-methylglycine and transient kinetics under presteady-state conditions and following the changes in absorbance at
510 nm (Figure 3 and Table 1). The observation of the quinonoid
intermediate I in the R. capsulatus ALAS-catalysed reaction is in
good agreement with similar results obtained from experiments
performed with E. coli and Mycobacterium tuberculosis AONS
[16,31] and SPT [32]. Similar to the reaction of the ALAS–
glycine complex with succinyl-CoA, the kinetic trace for the
reaction of the ALAS–O-methylglycine complex with succinylCoA was best described as a four-step process [eqn (1); k1 = 2.96
s − 1 , k2 = 0.19 s − 1 , k3 = 32.10 s − 1 and k4 = 0.64 s − 1 ; Table 1).
Overall, using O-methylglycine it was possible to visualize for the
first time the quinonoid intermediate I during the R. capsulatus
ALAS-catalysed reaction.
Analysis of the Asn54 variants
The active-site residue Asn54 of R. capsulatus ALAS is
responsible for the hydrogen bonds to the carboxy group of
PLP-bound glycine. After condensation with succinyl-CoA a
possible α-amino-β-oxoadipate intermediate is formed, and Asn54
c The Authors Journal compilation c 2013 Biochemical Society
210
Table 1
A.-L. Kaufholz and others
Rates of quinonoid intermediate formation and decay under single-turnover conditions
The R. capsulatus ALAS reaction was fitted to a four-step reaction with four observable rates. k 1 , rate of the formation of the initial collision complex; k 2 , rate of the formation of quinonoid intermediate
I/II; k 3 , rate of the formation of the enzyme–product complex; k 4 , rate of the dissociation of the enzyme–product complex into the enzyme and product; k − 1 , decay of initial collision complex; k − 2 ,
decay of the quinonoid intermediates I/II; k − 3 , decay of the enzyme–product complex; k − 4 , reverse rate of the enzyme and product formation; ND, non-detectable under the experimental conditions.
Assayed intermediate
Rates
Glycine quinonoid intermediate II pH 9.0 (s − 1 )
O -methylglycine quinonoid intermediate I pH 9.0 (s − 1 )
Wild-type ALAS
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
k1
k2
k3
k4
k −1
k −2
k −3
k −4
−6
30.44 +
− 5.04 × 10− 5
3.09 +
2.58
×
10
−
−5
6.75 +
− 3.46 × 10 − 9
0.07 +
3.45
×
10
−
−4
1.59 +
− 4.99 × 10 − 8
0.03 +
4.51
×
10
−
−7
7.42 +
− 2.02 × 10
7.98 × 10 − 4 +
1.67
× 10 − 9
−
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
−5
0.17 +
− 4.64 × 10 − 8
1.76 +
3.87
×
10
−
− 10
2.39 +
− 8.86 × 10 − 3
0.06 +
3.11
×
10
−
−4
10.06 +
− 3.97 × 10− 6
0.03 +
− 4.59 × 10 − 6
0.33 +
− 2.12 × 10
−8
2.81 × 10 − 4 +
− 6.95 × 10
279.02 +
556.91
−
−2
3.07 +
− 3.77 × 10 − 2
4.67 +
2.26
×
10
−
8.05 × 10 − 3 +
× 10 − 6
− 4.92
−4
4.42 +
− 4.3 × 10
32.74 +
− 4.41
−5
1.00 × 10 − 10 +
− 2.23 × 10
2.80 × 10 − 7
−7
0.39 +
− 3.31 × 10 − 3
4.24 +
1.37
×
10
−
−5
5.74 +
− 5.37 × 10 − 7
0.14 +
− 1.17 × 10 − 3
24.15 +
− 2.15 × 10
7.38 × 10 − 2 +
× 10 − 8
− 3.37
−6
0.79 +
1.65
×
10
−
− 10
6.74 × 10 − 4 +
− 4.99 × 10
ND
ND
ND
ND
ND
ND
ND
ND
−4
2.96 +
− 2.65 × 10 − 9
0.19 +
9.48
×
10
−
−2
32.10 +
− 1.14 × 10− 6
0.64 +
4.23
×
10
−
26.43 +
− 0.47
4.49 × 10 − 2 +
× 10 − 6
− 1.15
−4
7.69 +
− 5.3−×4 10
− 11
4.31 × 10 +
− 4.21 × 10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
−5
3.01 +
− 1.26 × 10 − 7
0.19 +
2.70
×
10
−
−3
30.39 +
− 5.55 × 10− 7
0.63 +
− 7.93 × 10 − 4
27.49 +
− 4.26 × 10
4.45 × 10 − 2 +
× 10 − 10
− 7.66
−5
7.59 +
3.19
×
10
−
− 11
4.26 × 10 − 4 +
− 2.65 × 10
ND
ND
ND
ND
ND
ND
ND
ND
R21K
R21E
T83S
N85F
N54Q
N54D
is most likely to be involved in the onset of the following
decarboxylation reaction. Conceivably, the glycine carboxylate
is released from the Asn54 -co-ordinated α-amino-β-oxoadipate
intermediate yielding the quinonoid intermediate II. To assess
the proposed role of the R. capsulatus Asn54 in stabilizing the
α-amino-β-oxoadipate intermediate [29], this residue was
c The Authors Journal compilation c 2013 Biochemical Society
substituted with either glutamine (N54Q) or aspartate (N54D)
using site-directed mutagenesis. The desired N54K variant was
found to be insoluble during recombinant production in E. coli
cells. The N54D variant would be expected to highly destabilize
the α-amino-β-oxoadipate intermediate by ionic repulsion
between the negatively charged aspartate and the carboxy group.
Aminolaevulinic acid synthase
Table 2
211
Kinetic parameters for R. capsulatus wild-type ALAS and its variants involved in substrate recognition and co-ordination
The shown parameters were obtained as outlined in the Experimental section. ND, non-detectable enzymatic activity under the assay conditions used.
ALAS variants
K mGly (mM)
K mS−CoA (μm)
k cat (s − 1 )
k cat / K mGly (mM/s)
k cat / K mS−CoA (μM/s)
Wild-type
R21K
R21E
T83S
N85Q
N85F
I86H
N54Q
N54D
−3
2.5 × 10 − 1 +
− 9 × 10
12.27 +
6.66
−
ND
−2
9.6 × 10 − 2 +
− 3.4 × 10
2.49 +
1.29
−
3.56 +
− 1.69
4.68 +
− 1.83
1.7 +
− 0.7
ND
−1
3.6 × 10 − 1 +
− 1.4 × 10
9.19 +
2.97
−
ND
−1
4 × 10 − 1 +
− 2.5 × 10
13.48 +
4.22
−
4.06 +
− 2.94
−1
8.8 × 10 − 1 +
− 4.3 × 10
0.85 +
− 0.3
ND
−2
2.7 × 10 − 1 +
− 9 × 10 − 2
9.2 × 10 − 2 +
5.5
×
10
−
ND
−2
3.2 × 10 − 2 +
− 1.2 × 10
−2
2.5 × 10 − 1 +
7
×
10
−
−1
9.5 × 10 − 1 +
− 5.4 × 10 − 2
2.3
×
10
3.3 × 10 − 2 +
−
0.32 +
− 0.03
ND
10.8 × 10 − 1
1 × 10 − 2
ND
8 × 10 − 2
1.9 × 10 − 2
2.3 × 10 − 1
1.5 × 10 − 2
1.9 × 10 − 1
ND
7.5 × 10 − 1
7.5 × 10 − 3
ND
3.3 × 10 − 1
10.1 × 10 − 1
2.7 × 10 − 1
2 × 10 − 3
3.8 × 10 − 1
ND
This might prevent the reaction from occurring altogether,
accelerate the formation of quinonoid intermediate II or change
the conformation in the active site enough that a quinonoid
intermediate is not observed even if the variant were active.
The results of the present study indicate that N54D mutation
prevents the reaction from occurring. As expected, the asparagine
to aspartate residue substitution (N54D) led to an ALAS variant
with no detectable enzymatic activity, indicating the importance
of the exchanged residue in substrate binding and catalysis. We
postulated that the conservative N54Q exchange would result in
an ALAS enzyme of reduced activity since the glutamine residue
side chain would partly substitute for the asparagine one. The
R. capsulatus enzyme was found to have a K m value for glycine of
∼0.25 mM (Table 2), which, while in the same order of magnitude
as for the R. sphaeroides enzyme [33], is approximately 100-fold
lower than that reported for mALAS2 (mouse ALAS 2, erythroid)
[34]. Similar values were also obtained for the physiological
substrate L-alanine for AONS [11]. The conservatively mutated
N54Q ALAS variant enzyme was active. It showed a 7-fold
increase in the K mGly value, a 2-fold increase in the K mS−CoA value
and a 1.2-fold increase in kcat value compared with the wildtype ALAS. Thus the catalytic efficiency for glycine was only
17.5 % of that of the wild-type enzyme. Nevertheless, the catalytic
efficiency of N54Q towards succinyl-CoA (kcat / K mS−CoA ) was about
50 % of the wild-type ALAS (Table 2). The introduction of
the N54D mutation in R. capsulatus ALAS abolished enzyme
activity and consequently quinonoid intermediate II formation.
Obviously, the inhibition of overall substrate binding prevented
the following catalytic steps including quinonoid intermediates
I and II formation. The N54Q substitution in R. capsulatus
ALAS only slightly decreased the rates assigned to the quinonoid
intermediate II lifetime and decay (Figures 3A and 3B, and
Table 1). The rates of quinonoid intermediate II formation
(k1 and k2 ) for wild-type ALAS were k1 = 30.44 s − 1 and
k2 = 3.09 s − 1 , whereas the N54Q variant reached 0.39 s − 1 for
k1 and 4.23 s − 1 for k2 . The rate of the two steps in quinonoid
intermediate II decay for wild-type ALAS enzyme are nearly
6.75 s − 1 (k3 ) and 7.37 × 10 − 2 s − 1 (k4 ), and 5.73 s − 1 (k3 ) and
0.14 s − 1 (k4 ) for the N54Q mutant. Clearly, the quinonoid
intermediate II decay steps in the N54Q variant are similar to
the wild-type ALAS, only the rate of the formation of the first
initial collision complex (k1 ) differed. The glutamine residue
obviously substituted efficiently for the active site asparagine.
Overall, detection of quinonoid II is strong evidence for catalysis
proceeding via α-amino-β-oxoadipate.
The subsequent condensation of the quinonoid intermediate
I with succinyl-CoA should lead to the accumulation of the
methyl ester of the β-oxoacid–aldimine complex and CoA release.
This is a similar reaction to that reported for AONS with
its corresponding substrates and analogues [21]. However, no
additional absorption maximum was observed between 30 s and
1 h after the start of the reaction. Instead, the quinonoid I form
of the enzyme disappeared over time. Multiple enzyme variants
(see below) tested at different pH values (between 5 and 10), with
different buffers, salt concentration, temperatures and substrate
concentrations did not yield methylated α-amino-β-oxoadipate.
Since the enzyme revealed its highest activity at pH 9.5, fast
hydrolysis of the methylester was concluded which would lead to
decarboxylation and product formation. The results of the present
study are in good agreement with the results for the mouse enzyme
[10]. Experiments with O-methylglycine aimed at trapping the
α-amino-β-oxoadipate intermediate were only possible with
the N54Q variant, since the N54D variant was inactive. However,
similar to wild-type ALAS, only the formation of quinonoid I was
observed (N54Q k1 = 3.01 s − 1 , k2 = 0.19 s − 1 , k3 = 30.39 s − 1 and
k4 = 0.63 s − 1 , Table 2).
Detection of quinonoid II
The energy associated with succinyl-CoA-binding drives
decarboxylation of the α-amino-β-oxoadipate intermediate and
ultimately the formation of ALA via quinonoid intermediate II
[29]. Since the release of the product was proposed as the ratelimiting step, quinonoid intermediate II should be observed during
catalysis. The formation and decay of the quinonoid intermediate
II with an absorbance maximum at approximately 510 nm were
monitored under single-turnover conditions at 30 ◦ C and pH of
9.0 using stopped-flow absorption spectroscopy (Figures 3A and
3B, and Table 1). The kinetic traces for the R. capsulatus ALAS
reaction at pH 9.0 (Figures 3A and 3B) were best described by
the four-step sequential mechanism represented by eqn (1). At a
lower pH, less quinonoid intermediate II was detectable in the R.
capsulatus ALAS reaction. Quinonoid intermediate II formation
with ALA was found to be pH dependent, with an optimal signal at
pH 9.0 (results not shown). In Figure 3, the time courses at 510 nm
were overlaid with the best fits of the data to eqn (1). The first of
the four steps was assigned to an initial collision complex
between the reactants, the second step was assigned to the
formation of the quinonoid intermediate II and the other two
to the decay steps of this intermediate. Overall, our pre-steadystate measurements permitted us to clearly identify quinonoid
intermediates I and II as in the R. capsulatus ALAS-catalysed
reaction. In the context of successful quinonoid intermediate
II detection a transient α-amino-β-oxoadipate is the only
possible intermediate between the two quinonoid intermediates.
Analogously to murine ALAS quinonoid intermediate II decay
c The Authors Journal compilation c 2013 Biochemical Society
212
A.-L. Kaufholz and others
Figure 5 Active site of R. capsulatus ALAS with mutated residues as
chemical structures
Figure 4 pH-dependent formation of quinonoid II from the product ALA by
wild-type ALAS
Purified enzyme was incubated with 5 mM ALA at pH 9.5 (A), 7.5 (B) and 6 (C), and the
absorption spectra were recorded as outlined in the Experimental section.
in single-turnover experiments follows two kinetic steps with
involvement of Lys248 . Obviously, ALAS does not directly utilize
the electron sink of the cofactor, which further sustains the
outlined stereoelectronic control hypothesis. Instead, an enol
derivative that is in equilibrium with quinonoid intermediate II
and the product ALA-bound external aldimine is possible [27].
Finally, we observed a two-step quinonoid intermediate decay
(Table 1) for the R. capsulatus ALAS enzyme and some of
its derivatives. The first step might again be acid-catalysed by
Lys248 and consequently be pH dependent. However, for the R.
capsulatus ALAS and its variants stable quinonoid intermediate
II formation was optimally observed at pH 9.0 (Table 1). Finally,
Lys248 protonates the enol which abolishes quinonoid absorbance
and leads to the fast step of quinonoid decay. The second step
of quinonoid intermediate decay should be pH independent and
represents the rate limiting step of the overall ALAS reaction, the
release of the substrate ALA. This behaviour is clearly reflected
by the data of quinonoid intermediate II formation and decay
studies presented in Table 1.
pH-dependent quinonoid II production from ALA
To obtain further evidence for the α-amino-β-oxoadipate
intermediate the ALAS reaction was performed backwards with
the product ALA at different pH values. A basic pH would
favour the deprotonation of PLP-bound ALA to yield quinonoid
II, thus providing further evidence for a catalytic path including
the α-amino-β-oxoadipate intermediate. As shown in Figure 4
with increasing pH, more quinonoid II was formed from ALA,
confirming the initial assumption. Ile86 also comes from the
adjacent subunit of the dimer and forms a van der Waals
interaction with the carboxylate group of the succinyl-CoA
substrate. The amino acid substitution of isoleucine residue
with histidine (I86H) (Figure 5) resulted in a less-active protein
(kcat = 3.3 × 10 − 2 s − 1 ). Here, the measured turnover number of
the ALAS variant was so low that the enzyme could be denoted
as de facto inactive. Surprisingly, binding of the succinyl-CoA
c The Authors Journal compilation c 2013 Biochemical Society
The substrates succinyl-CoA (turquoise) and glycine (salmon), the PLP cofactor bound to the
essential active site lysine (light brown) and selected amino acids (grey or green) are shown
as sticks. Furthermore selected amino acids are overlaid with the chemical structures of their
mutated analogues (pink). Monomer 1 is highlighted in light green and monomer 2 in light
grey. Additionally, residues from monomer 2 are marked with an asterisk. The cofactor PLP and
Lys248 , the lysine residue involved in PLP-binding and catalysis, are marked in light brown.
The residue Arg21 from monomer 1 and the residues Thr83 , Asn85 and Ile86 from monomer
2 are all involved in co-ordination of the succinyl-CoA substrate carboxylate group. This
co-ordination is carried out via a salt bridge by Arg21 , two hydrogen bonds to Thr83 and Asn85 ,
and van der Waals interaction with Ile86 . Asn54 from monomer 1 can stabilize or destabilize
the α-amino-β-oxoadipate intermediate. The formation of the quinonoid intermediate II is
acid-catalysed by His142 (from monomer 1), which is located directly above the cofactor ring.
The figure was prepared using PyMOL (http://www.pymol.org) and PDB codes 2BWN, 2BWO
and 2BWP.
substrate was not drastically hindered (K mS−CoA = 8.8 × 10 − 1
μM). However, the glycine residue affinity to the enzyme was
drastically reduced. The introduced histidine residue lacks the
methyl groups necessary for the formation of important van
der Waals interactions. However, this amino-acid exchange also
introduces an additional positive charge into the active site. The
overall structural arrangement with respect to the position of
neighbouring amino acids including Thr83 , which was in contact
to the PLP-bound glycine substrate, might have been disturbed.
This might have caused problems with the initial glycine-substrate
binding. In contrast the missing van der Waals contact of the
mutant enzyme to succinyl-CoA might be negligible.
Thr83 from subunit 2 is responsible for active-site flexibility and mediates
selectivity for glycine as a substrate
R. capsulatus ALAS exists as a dimer with the active site located
at the subunit interface. Thr83 of the second ALAS monomer coordinates the carboxylate group of the succinyl-CoA substrate
mainly bound to the first monomer via a hydrogen bond (2.6 Å)
and forms van der Waals interactions to the glycine substrate.
The conservative amino acid exchange of threonine to serine
(T83S) (Figure 5) yielded an enzyme variant with an 8.5-fold
lower kcat value (3.2 × 10 − 2 s − 1 ) compared with the wild-type
enzyme. Interestingly, the ALAS variant T83S revealed a higher
affinity to glycine residues compared with the wild-type ALAS
(K mGly = 9.6 × 10 − 2 mM). It required, at the same time, a similar
succinyl-CoA concentration (K mS−CoA = 4 × 10 − 1 μM) for half
saturation. The hydroxy group of the serine residue might have
contributed further to the co-ordination of the glycine substrate,
whereas at the same time the hydrogen bond to succinyl-CoA was
missing. As a consequence catalysis was slower. Consequently,
quinonoid intermediate II formation required double the time of
Aminolaevulinic acid synthase
the wild-type enzyme and its decay was over 10-fold higher.
Clearly, Thr83 localized between the two substrates is involved in
the structural movement required to co-ordinate the condensation
of both the substrate with the subsequent α-amino-β-oxoadipate
and quinonoid intermediate II formation. Its location on the other
subunit might be essential to fulfil this function.
The active-site Thr83 is localized directly at the border of the
two bound substrates glycine and succinyl-CoA. Thr83 interacts
with both substrates, and spatially constrains the active site so
that larger alternative substrates, differing in chain length and
chemical composition, might not bind. Consequently, we tested
how this crucial residue tolerates substrate alternatives to glycine.
Shoolingin-Jordan et al. [26] showed that the T83S variant of
R. sphaeroides ALAS accepted amino acids other than glycine
as a substrate including threonine, serine and alanine. For the
R. capsulatus wild-type ALAS it was proposed, on the basis of
its crystal structure, that substrates other than glycine can be
bound in the active site. However, they would block the access
for the second substrate, succinyl-CoA [24]. To investigate this
hypothesis we conducted activity assays for steady-state kinetic
determination with the wild-type and the T83S ALAS variant with
the amino acids L-serine, L-cysteine, L-threonine, L-alanine and
D-alanine as alternative substrates to glycine. Substrate-specificity
assays revealed that the R. capsulatus wild-type ALAS was highly
specific for glycine. Even with a high amount of succinyl-CoA
(20 μM), no notable enzyme activity was observed with the other
tested amino acids (results not shown). In contrast the ALAS T83S
variant showed enzyme activity with L-alanine, but not with any of
the other tested amino acids (Table 3). Interestingly, the reaction
of the R. capsulatus ALAS T83S mutant enzyme with L-alanine
showed an almost identical kcat value (3.8 × 10 − 2 s − 1 ) compared
with the reaction with the physiological substrate glycine.
However, the K m value for alanine was higher (K mL−Alanine =
2.7 × 10 − 1 mM) than that for glycine (K mGly = 1.4 × 10 − 1 mM).
The affinity of the mutant enzyme for succinyl-CoA was
approximately the same in both reactions. Furthermore, we
performed pre-steady-state reactions with the ALAS T83S variant
as with the physiological substrate glycine in comparison with Lalanine. Figure 3(F) illustrates the results of these experiments. As
expected from the kinetic characterization described above, wildtype ALAS did not react with L-alanine. Consequently, no stable
quinonoid intermediate II was detectable. In contrast, ALAS T83S
reacted with glycine and L-alanine during quinonoid intermediate
II formation. The reaction of ALAS T83S with glycine was
∼6.1-fold higher than with L-alanine. As expected the major
difference between wild-type ALAS and the T83S variant was
the binding of the first substrate, whereas the condensation with
the second one was not so much affected by the amino acid
exchange. In summary, Thr83 plays an important role in glycine
binding and discrimination as assumed by Heinz and co-workers
[3]. Clearly, Thr83 localized in-between the two substrates is also
important for the discrimination of the glycine substrate against
other structurally related amino acids.
Recognition of succinyl-CoA
Structural determinants of acyl-CoA substrates for ALAS recognition
The second substrate, succinyl-CoA, binds in a hydrophobic
pocket at the entrance of the channel leading to the active
site. Its carboxylate group is co-ordinated by Thr365 and Arg21
of the first monomer as well as by Thr83 of the second
monomer [24]. Shoolingin-Jordan et al. [35] and we [8]
previously tested different acyl-CoA derivatives with ALAS from
R. sphaeroides and the mouse respectively. Acyl-CoA chain
213
Table 3 Substrate specificity tested for the R. capsulatus ALAS T83S variant
with glycine and L-alanine
Different amino acids (L-alanine, D-alanine, L-cysteine, L-threonine and L-serine) were tested as
glycine alternative with the R. capsulatus wild-type ALAS and the ALAS T83S variant. Wild-type
ALAS and its Thr83 variant showed no detectable enzyme activity with D-alanine, L-cysteine,
L-threonine and L-serine. Although in the T83S reaction glycine as well as L-alanine served as
substrates, there was no detectable enzyme activity for the wild-type enzyme under the assay
conditions used. Given values were obtained via steady-state kinetics analysis as outlined in the
Experimental section. ND, non-detectable enzymatic activity under the assay conditions used.
Parameter
Reaction with glycine
Reaction with L-alanine
K mGly (mM)
K mS−CoA (μm)
K mL−Alanine (mM)
k cat (s − 1 )
k cat / K mGly (mM/s)
k cat / K mS−CoA (μM/s)
k cat / K mL−Alanine (μM/s)
−1
1.4 × 10 − 1 +
− 1.2 ×−10
1
5 × 10 − 1 +
− 1.6 ×
NA
−2
3.4 × 10 − 2 +
− 1.3 × 10
2.4 × 10 − 1
6.8 × 10 − 2
ND
ND
−1
6.5 × 10 − 1 +
− 1.6 × 10 − 1
1.3
×
10
2.7 × 10 − 1 +
−
−3
3.8 × 10 − 2 +
− 8 × 10
ND
5.8 × 10 − 2
1.4 × 10 − 1
Table 4 Stopped-flow single turnover rates for R. capsulatus wild-type
ALAS and the ALAS T83S variant with L-alanine or glycine
The R. capsulatus ALAS reaction was fitted to a four step reaction with four observable
rates. k 1 , rate of the formation of the initial collision complex; k 2 , rate of the formation of
quinonoid intermediate I/II; k 3 , rate of the formation of the enzyme–product complex; k 4 , rate
of the dissociation of the enzyme–product complex into the enzyme and product; k − 1 , decay
of initial collision complex; k − 2 , decay of the quinonoid intermediates I/II; k − 3 , decay of
the enzyme–product complex; k − 4 , reverse rate of the enzyme and product formation; ND,
non-detectable under the experimental conditions.
Enzyme
Glycine (s − 1 )
L-Alanine (s − 1 )
ALAS wild-type
ND
ND
ND
ND
ND
ND
ND
ND
−5
0.17 +
− 4.64 × 10 − 8
1.76 +
3.87
×
10
−
− 10
2.39 +
− 8.86 × 10 − 3
0.06 +
− 3.11 × 10 − 4
10.06 +
− 3.97 × 10− 6
0.03 +
− 4.59 × 10 − 6
0.33 +
− 2.12 × 10
−8
2.81 × 10 − 4 +
− 6.95 × 10
ND
ND
ND
ND
ND
ND
ND
ND
2.77 × 10 − 2 +
× 10 − 9
− 9.01
−6
0.29 +
5.10
×
10
−
−7
0.39 +
− 1.19 × 10− 6
9.48 +
− 5.2 × 10 − 6
1.67 +
− 1.86 × 10
5.12 × 10-3 +
− 0.42
−3
5.14 × 10 − 2 +
− 1.08 × 10 − 7
4.68 × 10 − 5 +
1.59
×
10
−
T83S
lengths and hydrophobicity influenced ALAS activity. Previously,
we tested four different acyl-CoA derivatives (octanoyl-CoA,
butyryl-CoA, β-hydroxybutyryl-CoA and glutaryl-CoA) with
three different murine ALAS2 variants (R85L, R85K and
R85L/T430V). The results showed that acyl-CoA substrates
of increased hydrophobicity (e.g. octanoyl-CoA and butyrylCoA) showed greater affinity for the mALAS2 variants with a
substituted aliphatic amino acid (R85L and R85L/T430V). We
concluded that the chemical characteristics of the CoA-derived
tail and the hydrogen-bonding potential of the invariant acyl-CoAbinding residues are responsible for reaction specificity. In the
present study we executed substrate-specificity assays with two
different acyl-CoA derivatives (butyryl-CoA and octanoyl-CoA)
with the R. capsulatus wild-type ALAS and the R21K variant. The
obtained results are summarized in Table 4 and compared with
those for the physiological acyl-CoA substrate succinyl-CoA.
c The Authors Journal compilation c 2013 Biochemical Society
214
A.-L. Kaufholz and others
Table 5
Comparison of steady-state kinetic constants for R. capsulatus wild-type ALAS and its variant R21K with acyl-CoA derivatives as a second substrate
Steady-state kinetic analyses were performed with octanoyl-CoA and butyryl-CoA and compared to the kinetic constants for the physiological substrate succinyl-CoA. The different acyl-CoA molecules
differ in their length of the acyl residue.
Parameter
Wild-type ALAS
ALAS R21K
Substrate structure
(a) Succinyl-CoA as substrate
−1
K mS−CoA (μm)
3.6 × 10 − 1 +
− 1.4 × 10
−3
K mGly (mM)
2.5 × 10 − 1 +
9
×
10
−
−2
k cat (s − 1 )
2.7 × 10 − 1 +
− 9 × 10
S−CoA
−1
k cat / K m
(μM/s)
7.5 × 10
k cat / K mGly (mM/s)
10.8 × 10 − 1
7.05 +
− 2.58
11.02 +
− 2.20
−2
4.21 × 10 − 2 +
− 3.5 × 10
5.9 × 10 − 3
3.8 × 10 − 3
(b) Butyryl-CoA as substrate
K mbutryl−CoA (μm)
K mGly (mM)
k cat (s − 1 )
k cat / K mbutryl−CoA (μM/s)
k cat / K mGly (mM/s)
1.89 +
− 1.04
22.18 +
− 7.75
−3
2.3 × 10 − 2 +
− 3.2 × 10
1.2 × 10 − 2
1.02 × 10 − 3
−1
9.8 × 10 − 1 +
− 2.0 × 10
−1
7.4 × 10 − 1 +
2.9
×
10
−
−1
1.2
×
10
2.3 × 10 − 1 +
−
2.3 × 10 − 1
3.1 × 10 − 1
(c) Octanoyl-CoA as substrate
−1
K moctanoyl−CoA (μm)
6.4 × 10 − 1 +
− 2.8 × 10
−1
K mGly (mM)
4.9 × 10 − 1 +
3.6
×
10
−
k cat (s − 1 )
8.9 × 10 − 2 +
k
10 − 2
3.6
×
cat
−
octanoyl−CoA
−1
k cat / K m
(μM/s)
1.4 × 10
−2
8.6 × 10 − 1 +
− 4.9 × 10
9.35 +
3.85
−
1.7 × 10 − 2 +
− 4.6 × 10 − 3
1.9 × 10 − 2
The results of the steady-state kinetics experiments revealed
enzyme activity independent of the type of acyl-CoA substrate
used. The observed substrate specificity for butyryl-CoA
(kcat /K mbutyryl−CoA = 2.3 × 10 − 1 μM/s) was comparable with those
of the wild-type enzyme (kcat /K mS−CoA = 7.5 × 10 − 1 μM/s).
The only difference of these two molecules is the terminal
carboxy group of succinyl-CoA lacking in the butyryl-CoA
structure (Table 4). In contrast, substrate specificity was clearly
decreased using octanoyl-CoA as second substrate. Here,
kcat /K moctanoyl−CoA (1.4 × 10 − 1 μM/s) was approximately five times
lower compared with the physiological substrate (kcat /K mS−CoA =
7.5 × 10 − 1 μM/s). The tail of octanoyl-CoA consists of seven
methyl groups, two more than in succinyl-CoA, but lacks the
carboxy group (Table 4). A consequence of this longer tail
is a positive charge owing to the electron-shifting properties
of the methyl groups towards the oxygen atom. However, an
overall tolerance of acyl-CoA chain length in R. capsulatus
ALAS catalysis was observed. This might be owing to unspecific
hydrophobic co-ordination of the hydrocarbon backbone.
Co-ordinating amino acids and experimental approach
Next, we focused on the functional analysis of active-site aminoacid residues which co-ordinate the succinyl-CoA substrate.
Arg21 forms a salt bridge to the carboxylate group of succinylCoA, whereas Thr83 , Asn85 and Ile86 co-ordinate the same group
via hydrogen bonds. In most cases a conservative amino-acid
exchange was combined with a non-conservative one. For the
steady-state kinetic measurements of the various enzyme variants
a coupled enzyme assay with the NADH-dependent α-KGD was
used. The K m and kcat values for glycine and succinyl-CoA were
determined. Quinonoid intermediate I and II formation and decay
were tested as described above. Ile86 from subunit 2 mediates
the structural rearrangements required for glycine binding and
catalysis.
c The Authors Journal compilation c 2013 Biochemical Society
Arg21 co-ordinates succinyl-CoA binding and structural rearrangements on
ALAS
The non-conservative exchange from Arg21 to a glutamate
residue (R21E) (Figure 5) led to an inactive enzyme (Table 2).
In this variant the positively charged guanidinium side chain
[HN = C(NH2 )-NH-R] of arginine is replaced by the negatively
charged side chain of glutamate. The glutamate carboxy group
was expected to probably reject the carboxylate group of
succinyl-CoA, resulting in an inactive enzyme and therefore
succinyl-CoA binding to the enzyme was expected to be
prevented. The conservative amino-acid exchange of arginine to
lysine (R21K) yielded an enzyme with an approximately three
times lower kcat value (9.2 × 10 − 2 s − 1 ) than the wild-type ALAS
(kcat = 2.7 × 10 − 1 s − 1 ). High amounts of both substrates were
necessary for half saturation of the ALAS R21K variant (K mGly
= 12.27 mM and K mS−CoA = 9.19 μM). Moreover, we failed to
stabilize and detect quinonoid intermediates I and II with this
mutant (Table 1). Clearly, this amino-acid exchange, even though
it was conservative, yielded an enzyme where the active site
and thus the binding of both substrates and subsequent catalysis
were negatively affected. Enzyme activity might suffer once the
hydrogen bond triad responsible for the correct succinyl-CoA
co-ordination in the active site cannot be formed owing to the
inserted lysine residue in the ALAS R21K variant. Furthermore,
Arg21 is located close to Thr365 , a residue positioned at the tip of the
dynamic active-site loop which is involved in the conformational
changes between the open and closed form of the enzyme [3]. Consequently, Arg21 is essential for ALAS catalysis in several ways.
Asn85 from subunit 2 mediates succinyl-CoA recognition
Asn85 also comes from the second ALAS monomer, and coordinates the succinyl-CoA carboxylate group via a hydrogen
bond with a distance of 3.1 Å [3]. Its contribution to succinyl-CoA
Aminolaevulinic acid synthase
binding has been investigated recently for mouse ALAS [8]. The
analysis of the conservatively exchanged N85Q ALAS variant
revealed the contribution of this amino acid residue to mainly
substrate recognition and less to catalysis. The K m values for
succinyl-CoA were 0.36 μM for the wild-type ALAS, which is
in the range of the value for AONS [11]. Both the K m values
for glycine (K mGly = 2.49 mM) and especially for succinyl-CoA
(K mS−CoA = 13.48 μM) were in dimensions different to the wildtype enzyme, whereas the rate of catalyses was found to be
similar (kcat = 2.5 × 10 − 1 s − 1 ). Obviously, the distance of the
hydrogen bond between the succinyl-CoA substrate and Asn85 is
crucial for efficient substrate recognition. However, after substrate
binding catalysis is not affected by the amino-acid exchange. In
agreement, no stable quinonoid intermediate I, but a quinonoid
intermediate II, with similar behaviour as observed for the wildtype enzyme was detected. At first sight the results for ALAS
carrying a non-conservative exchange (N85F) (Figure 5) seemed
somewhat surprising. The amino-acid exchange yielded a protein
with a 3.5-fold increase in the kcat value (9.5 × 10 − 1 s − 1 ). Even the
overall catalysis of the mutant enzyme was better than observed
for the wild-type protein, poor substrate binding for glycine
(K mGly = 3.56 mM) and again succinyl-CoA (K mS−CoA = 4.06 μM)
remained. Obviously, the substitution of a polar with a non-polar
amino acid resulted in a higher catalytic activity, most probably
owing to a less tight co-ordination of the carboxylate group of
succinyl-CoA. If the succinyl-CoA carboxylate group was bound
less tightly in the mutant ALAS enzyme, all following steps after
succinyl-CoA binding might have proceeded faster. As a result a
faster ALA release was possible and thus the enzyme showed a
higher turnover number than the wild-type protein. In summary, an
ALAS enzyme with a higher activity than the naturally occurring
enzyme, however, with less affinity for both substrates (for N85F
K mGly = 3.56 mM and K mS−CoA = 4.06 μM) was generated. Both
amino-acid exchanges (N85Q and N85F) revealed the importance
of Asn85 for succinyl-CoA substrate recognition and binding.
Conclusions
The present study raised four main conclusions: (i) ALAS
catalysis proceeds via quinonoid I and quinonoid II intermediates,
strongly indicating that ALAS catalysis proceeds via quinonoid
I and quinonoid II intermediates strongly suggesting an αamino-β-oxoadipate intermediate. At the pH of 9.5 necessary for
optimal R. capsulatus catalysis a ‘trapped’ methylated α-aminoβ-oxoadipate derived from O-methylglycine and succinyl-CoA
is too unstable for further biophysical analysis; (ii) amino acids
of the intersubunit active site derived from subunit 2 provide the
required structural flexibility for catalysis; (iii) Thr83 of subunit 2
mediates glycine substrate specificity; and (iv) Asn85 of
subunit 2 mediates succinyl-CoA substrate recognition.
AUTHOR CONTRIBUTION
Gregory Hunter assisted in performing the steady-state and pre-steady-state kinetic
analyses and analysed the experimental data. Gregory Hunter and Gloria Ferreira helped
to write the paper. Thomas Lendrihas also assisted in performing pre-steady-state kinetic
analysis. Vanessa Hering helped to produce and purify all necessary enzymes. Gunhild
Layer helped in designing Figures of the ALAS active site. Gregory Hunter, Gloria Ferreira,
Martina Jahn and Dieter Jahn contributed to the analysis, interpretation and discussion of
all experimental data.
FUNDING
This work was supported by the Deutsche Forschungsgemeinschaft [grant number DFG-Ja
470/10-1].
215
REFERENCES
1 Eliot, A. C. and Kirsch, J. F. (2004) Pyridoxal phosphate enzymes: mechanistic, structural,
and evolutionary considerations. Annu. Rev. Biochem. 73, 383–415
2 Toney, M. D. (2011) Controlling reaction specificity in pyridoxal phosphate enzymes.
Biochim. Biophys. Acta 1814, 1407–1418
3 Astner, I., Schulze, J. O., van den Heuvel, J., Jahn, D., Schubert, W. D. and Heinz, D. W.
(2005) Crystal structure of 5-aminolevulinate synthase, the first enzyme of heme
biosynthesis, and its link to XLSA in humans. EMBO J. 24, 3166–3177
4 Heinemann, I. U., Jahn, M. and Jahn, D. (2008) The biochemistry of heme biosynthesis.
Arch. Biochem. Biophys. 474, 238–251
5 Layer, G., Reichelt, J., Jahn, D. and Heinz, D. W. (2010) Structure and function of enzymes
in heme biosynthesis. Protein Sci. 19, 1137–1161
6 Tan, D. and Ferreira, G. C. (1996) Active site of 5-aminolevulinate synthase resides at the
subunit interface. Evidence from in vivo heterodimer formation. Biochemistry 35,
8934–8941
7 Camaschella, C. (2009) Hereditary sideroblastic anemias: pathophysiology, diagnosis,
and treatment. Semin. Hematol. 46, 371–377
8 Lendrihas, T., Zhang, J., Hunter, G. A. and Ferreira, G. C. (2009) Arg85 and Thr430 in
murine 5-aminolevulinate synthase coordinate acyl-CoA-binding and contribute to
substrate specificity. Protein Sci. 18, 1847–1859
9 Ferreira, G. C., Neame, P. J. and Dailey, H. A. (1993) Heme biosynthesis in mammalian
systems: evidence of a Schiff base linkage between the pyridoxal 5 -phosphate
cofactor and a lysine residue in 5-aminolevulinate synthase. Protein Sci. 2,
1959–1965
10 Hunter, G. A. and Ferreira, G. C. (1995) A continuous spectrophotometric assay for
5-aminolevulinate synthase that utilizes substrate cycling. Anal. Biochem. 226,
221–224
11 Mann, S. and Ploux, O. (2011) Pyridoxal-5 -phosphate-dependent enzymes involved in
biotin biosynthesis: structure, reaction mechanism and inhibition. Biochim. Biophys. Acta
1814, 1459–1466
12 Ikushiro, H. and Hayashi, H. (2011) Mechanistic enzymology of serine
palmitoyltransferase. Biochim. Biophys. Acta 1814, 1474–1480
13 Lowther, J., Naismith, J. H., Dunn, T. M. and Campopiano, D. J. (2012) Structural,
mechanistic and regulatory studies of serine palmitoyltransferase. Biochem. Soc. Trans.
40, 547–554
14 Mukherjee, J. J. and Dekker, E. E. (1987) Purification, properties, and N-terminal amino
acid sequence of homogeneous Escherichia coli 2-amino-3-ketobutyrate CoA ligase, a
pyridoxal phosphate-dependent enzyme. J. Biol. Chem. 262, 14441–14447
15 Schmidt, A., Sivaraman, J., Li, Y., Larocque, R., Barbosa, J. A., Smith, C., Matte, A.,
Schrag, J. D. and Cygler, M. (2001) Three-dimensional structure of
2-amino-3-ketobutyrate CoA ligase from Escherichia coli complexed with a
PLP-substrate intermediate: inferred reaction mechanism. Biochemistry 40,
5151–5160
16 Webster, S. P., Alexeev, D., Campopiano, D. J., Watt, R. M., Alexeeva, M., Sawyer, L. and
Baxter, R. L. (2000) Mechanism of 8-amino-7-oxononanoate synthase: spectroscopic,
kinetic, and crystallographic studies. Biochemistry 39, 516–528
17 Hanada, K. (2003) Serine palmitoyltransferase, a key enzyme of sphingolipid metabolism.
Biochim. Biophys. Acta 1632, 16–30
18 Ikushiro, H., Hayashi, H. and Kagamiyama, H. (2004) Reactions of serine
palmitoyltransferase with serine and molecular mechanisms of the actions of serine
derivatives as inhibitors. Biochemistry 43, 1082–1092
19 Alexeev, D., Alexeeva, M., Baxter, R. L., Campopiano, D. J., Webster, S. P. and Sawyer, L.
(1998) The crystal structure of 8-amino-7-oxononanoate synthase: a bacterial
PLP-dependent, acyl-CoA-condensing enzyme. J. Mol. Biol. 284, 401–419
20 Hunter, G. A. and Ferreira, G. C. (2009) 5-Aminolevulinate synthase: catalysis of the first
step of heme biosynthesis. Cell. Mol. Biol. 55, 102–110
21 Kerbarh, O., Campopiano, D. J. and Baxter, R. L. (2006) Mechanism of α-oxoamine
synthases: identification of the intermediate Claisen product in the
8-amino-7-oxononanoate synthase reaction. Chem. Commun. 1, 60–62
22 Fanica-Gaignier, M. and Clement-Metral, J. (1973) 5-Aminolevulinic-acid synthetase of
Rhodopseudomonas spheroides Y. Purification and some properties. Eur. J. Biochem. 40,
13–18
23 Nandi, D. L. (1978) δ-Aminolevulinic acid synthase of Rhodopseudomonas spheroides .
Binding of pyridoxal phosphate to the enzyme. Arch. Biochem. Biophys. 188,
266–271
24 Astner, I. (2007) Klonierung, Reinigung, Kristallisation und Strukturlösung der
5-Aminolevulinsäuresynthase und Untersuchung der Auswirkung und Behandelbarkeit
von XLSA-Mutationen. Ph.D Thesis, TU-Braunschweig, Braunschweig, Germany
25 Lam, H., Oh, D. C., Cava, F., Takacs, C. N., Clardy, J., de Pedro, M. A. and Waldor, M. K.
(2009) D-Amino acids govern stationary phase cell wall remodeling in bacteria. Science
325, 1552–1555
c The Authors Journal compilation c 2013 Biochemical Society
216
A.-L. Kaufholz and others
26 Shoolingin-Jordan, P. M., Al-Daihan, S., Alexeev, D., Baxter, R. L., Bottomley, S. S.,
Kahari, I. D., Roy, I., Sarwar, M., Sawyer, L. and Wang, S. F. (2003) 5-Aminolevulinic acid
synthase: mechanism, mutations and medicine. Biochim. Biophys. Acta 1647, 361–366
27 Zhang, J. and Ferreira, G. C. (2002) Transient state kinetic investigation of
5-aminolevulinate synthase reaction mechanism. J. Biol. Chem. 277, 44660–44669
28 Barshop, B. A., Wrenn, R. F. and Frieden, C. (1983) Analysis of numerical methods for
computer simulation of kinetic processes: development of KINSIM: a flexible, portable
system. Anal. Biochem. 130, 134–145
29 Hunter, G. A., Zhang, J. and Ferreira, G. C. (2007) Transient kinetic studies support
refinements to the chemical and kinetic mechanisms of aminolevulinate synthase. J. Biol.
Chem. 282, 23025–23035
30 Hunter, G. A. and Ferreira, G. C. (1999) Pre-steady-state reaction of 5-aminolevulinate
synthase. Evidence for a rate-determining product release. J. Biol. Chem. 274,
12222–12228
Received 27 June 2012/2 January 2013; accepted 31 January 2013
Published as BJ Immediate Publication 31 January 2013, doi:10.1042/BJ20121041
c The Authors Journal compilation c 2013 Biochemical Society
31 Bhor, V. M., Dev, S., Vasanthakumar, G. R., Kumar, P., Sinha, S. and Surolia, A. (2006)
Broad substrate stereospecificity of the Mycobacterium tuberculosis
7-keto-8-aminopelargonic acid synthase: spectroscopic and kinetic studies. J. Biol.
Chem. 281, 25076–25088
32 Ikushiro, H., Fujii, S., Shiraiwa, Y. and Hayashi, H. (2008) Acceleration of the substrate
Cα deprotonation by an analogue of the second substrate palmitoyl-CoA in serine
palmitoyltransferase. J. Biol. Chem. 283, 7542–7553
33 Jordan, P. M. and Laghai-Newton, A. (1986) Purification of 5-aminolevulinate synthase.
Methods Enzymol. 123, 435–443
34 Gong, J., Hunter, G. A. and Ferreira, G. C. (1998) Aspartate-279 in aminolevulinate
synthase affects enzyme catalysis through enhancing the function of the pyridoxal
5 -phosphate cofactor. Biochemistry 37, 3509–3517
35 Shoolingin-Jordan, P. M., LeLean, J. E. and Lloyd, A. J. (1997) Continuous coupled
assay for 5-aminolevulinate synthase. Methods Enzymol. 281, 309–316