Abstract

The SCP2 (sterol carrier protein 2)-thiolase (type-1) functions in the vertebrate peroxisomal, bile acid synthesis pathway, converting 24-keto-THC-CoA and CoA into choloyl-CoA and propionyl-CoA. This conversion concerns the β-oxidation chain shortening of the steroid fatty acyl-moiety of 24-keto-THC-CoA. This class of dimeric thiolases has previously been poorly characterized. High-resolution crystal structures of the zebrafish SCP2-thiolase (type-1) now reveal an open catalytic site, shaped by residues of both subunits. The structure of its non-dimerized monomeric form has also been captured in the obtained crystals. Four loops at the dimer interface adopt very different conformations in the monomeric form. These loops also shape the active site and their structural changes explain why a competent active site is not present in the monomeric form. Native mass spectrometry studies confirm that the zebrafish SCP2-thiolase (type-1) as well as its human homolog are weak transient dimers in solution. The crystallographic binding studies reveal the mode of binding of CoA and octanoyl-CoA in the active site, highlighting the conserved geometry of the nucleophilic cysteine, the catalytic acid/base cysteine and the two oxyanion holes. The dimer interface of SCP2-thiolase (type-1) is equally extensive as in other thiolase dimers; however, it is more polar than any of the corresponding interfaces, which correlates with the notion that the enzyme forms a weak transient dimer. The structure comparison of the monomeric and dimeric forms suggests functional relevance of this property. These comparisons provide also insights into the structural rearrangements that occur when the folded inactive monomers assemble into the mature dimer.

Introduction

Fatty acids and fatty acid metabolism are essential for all forms of life. In the fatty acid degradation pathway, the fatty acids are conjugated to CoA. In this pathway, also known as the β-oxidation cycle, four reactions (dehydrogenation/oxidation, hydration, dehydrogenation and thiolysis) remove a two-carbon fragment in a cyclic fashion from the activated acyl-CoA. The last reaction of these four steps is catalyzed by thiolase. The end products of the thiolytic cleavage reaction catalyzed by thiolases are the chain shortened substrate CoA-derivative and acetyl-CoA (when the substrate does not have a 2-methyl group) or propionyl-CoA (when the substrate is a 2-methyl-branched acyl-CoA molecule) (Figure 1). In vertebrates, the β-oxidation degradative pathway is located in mitochondria and peroxisomes. In the peroxisomes, there are two sets of enzymes catalyzing the degradation of very long-chain unbranched acyl chains and very long-chain 2-methyl-branched acyl chains, respectively. The latter set of enzymes is also important in the bile acid synthesis pathway, leading to the conversion of cholesterol into cholic acid [1]. First cholesterol is converted into trihydroxycholestanoic acid and subsequently conjugated with CoA forming 25S-3α,7α,12α-trihydroxy-5β-cholestanoyl-CoA (THC-CoA). Subsequently, THC-CoA is degraded by the four β-oxidation steps. In the thiolytic step, 24-keto-THC-CoA is converted into choloyl-CoA and propionyl-CoA (Figure 1). This reaction is catalyzed by the dimeric peroxisomal SCP2 (sterol carrier protein 2)-thiolase (E.C. 2.3.1.176) [2,3].

The SCP2-thiolase reaction.

Figure 1.
The SCP2-thiolase reaction.

(A) The overall reaction catalyzed by SCP2-thiolase (type-1). Only the main intermediate complexes of the reaction cycle are shown. The substrates are 24-keto-THC-CoA and CoA, and they are converted into propionyl-CoA and choloyl-CoA. The conserved sequence fingerprints of the active site loops (Table 1) indicate that Cys87 is the nucleophilic cysteine and Cys300 is the acid/base cysteine, whereas His348 activates the nucleophilic cysteine. (B) The covalent structure of the steroid moiety of 24-keto-THC-CoA.

Figure 1.
The SCP2-thiolase reaction.

(A) The overall reaction catalyzed by SCP2-thiolase (type-1). Only the main intermediate complexes of the reaction cycle are shown. The substrates are 24-keto-THC-CoA and CoA, and they are converted into propionyl-CoA and choloyl-CoA. The conserved sequence fingerprints of the active site loops (Table 1) indicate that Cys87 is the nucleophilic cysteine and Cys300 is the acid/base cysteine, whereas His348 activates the nucleophilic cysteine. (B) The covalent structure of the steroid moiety of 24-keto-THC-CoA.

Table 1
The six human thiolases
Class CT-thiolase T1-thiolase T2-thiolase TFE-thiolase AB-thiolase SCP2-thiolase (type-1) 
Functional properties 
 Location Cytosol Mitochondria Mitochondria Mitochondria Peroxisomes Peroxisomes 
 Assembly Tetramer Tetramer Tetramer Dimer Dimer Dimer 
 Function Biosynthetic Biosynthetic/degradative Biosynthetic/degradative Degradative Degradative Degradative 
 Physiological substrates 2 molecules of acetyl-CoA 2 molecules of acetyl-CoA/CoA and long-chain, unbranched 3-ketoacyl-CoA 2 molecules of acetyl-CoA/CoA and 2-methyl-branched or unbranched 3-ketobutyryl-CoA CoA and long-chain, 2-methyl-branched or unbranched 3-ketoacyl-CoA CoA and long-chain, unbranched 3-ketoacyl-CoA CoA and long-chain 2-methyl-branched or -unbranched 3-ketoacyl-CoA or 24-keto-THC-CoA 
Sequence fingerprints of the four active site loops 
 Nβ3–Nα3 CxS/T CxS/T CxS/T CxS/T CxS/T CxS/T 
 Cβ2–Cα2 NEAF NEAF NEAF HEAF NEAF HDCF 
 Cβ3–Cα3 GHP GHP GHP GHP GHP GHP 
 Cβ4–Cβ5 CxG CxG CxG CxA CxG xxx 
Active site properties 
 Nucleophile C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) 
 Acid/base C(CxG) C(CxG) C(CxG) C(CxA) C(CxG) C(HDCF) 
 Activation of nucleophile H(GHP) H(GHP) H(GHP) H(GHP) H(GHP) H(GHP) 
 OAH1 hydrogen bond donors Wat, H(GHP) Wat, H(GHP) Wat, H(GHP) H(HEAF), H(GHP) Wat, H(GHP) H(HDCF), H(GHP) 
 OAH2 main chain NH hydrogen bond donors N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) 
PDB codes 
 PDB codes of structures of human thiolases 1WL4 [754C2J [762IBW [775ZQZ [782IIK (unpublished) 6HRV, 6HSJ, 6HSP (these studies) 
 PDB codes of corresponding structures1     1AFW [4], 2WUA [51], 2WU9 [51], 2C7Y [52 
Class CT-thiolase T1-thiolase T2-thiolase TFE-thiolase AB-thiolase SCP2-thiolase (type-1) 
Functional properties 
 Location Cytosol Mitochondria Mitochondria Mitochondria Peroxisomes Peroxisomes 
 Assembly Tetramer Tetramer Tetramer Dimer Dimer Dimer 
 Function Biosynthetic Biosynthetic/degradative Biosynthetic/degradative Degradative Degradative Degradative 
 Physiological substrates 2 molecules of acetyl-CoA 2 molecules of acetyl-CoA/CoA and long-chain, unbranched 3-ketoacyl-CoA 2 molecules of acetyl-CoA/CoA and 2-methyl-branched or unbranched 3-ketobutyryl-CoA CoA and long-chain, 2-methyl-branched or unbranched 3-ketoacyl-CoA CoA and long-chain, unbranched 3-ketoacyl-CoA CoA and long-chain 2-methyl-branched or -unbranched 3-ketoacyl-CoA or 24-keto-THC-CoA 
Sequence fingerprints of the four active site loops 
 Nβ3–Nα3 CxS/T CxS/T CxS/T CxS/T CxS/T CxS/T 
 Cβ2–Cα2 NEAF NEAF NEAF HEAF NEAF HDCF 
 Cβ3–Cα3 GHP GHP GHP GHP GHP GHP 
 Cβ4–Cβ5 CxG CxG CxG CxA CxG xxx 
Active site properties 
 Nucleophile C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) C(CxS/T) 
 Acid/base C(CxG) C(CxG) C(CxG) C(CxA) C(CxG) C(HDCF) 
 Activation of nucleophile H(GHP) H(GHP) H(GHP) H(GHP) H(GHP) H(GHP) 
 OAH1 hydrogen bond donors Wat, H(GHP) Wat, H(GHP) Wat, H(GHP) H(HEAF), H(GHP) Wat, H(GHP) H(HDCF), H(GHP) 
 OAH2 main chain NH hydrogen bond donors N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) N(Nβ3–Nα3), N(Cβ4–Cβ5) 
PDB codes 
 PDB codes of structures of human thiolases 1WL4 [754C2J [762IBW [775ZQZ [782IIK (unpublished) 6HRV, 6HSJ, 6HSP (these studies) 
 PDB codes of corresponding structures1     1AFW [4], 2WUA [51], 2WU9 [51], 2C7Y [52 
1

Other structures that have been used for comparisons are the dimeric Mycobacterium tuberculosis TFEL1 FadA5-thiolase structures 4UBT [21] and 5ONC [53], as well as the dimeric parasitic SCP2-thiolase (type-2) structures 3ZBG[18], 5LNQ [17] and 4BI9 [18].

Thiolases are homo-dimers or homo-tetramers (dimer of dimers) and, in vertebrates, they are found in the cytosol, mitochondria and peroxisomes (Table 1). The thiolase monomer is composed of the topologically similar N-terminal and C-terminal subdomains and an insertion subdomain, known as the loop-domain. Each of these three subdomains is ∼120 residues. The N- and C-terminal subdomains together form the thiolase core being two β-sheets sandwiched between α-helical layers. The N-terminal subdomain mediates the assembly of the thiolase monomers into the canonical dimer [4]. The loop-domain extends out from the N-terminal subdomain [between β-strands 4 (Nβ4) and 5 (Nβ5)] containing mainly α-helices and folding over the thiolase core. This domain forms the binding pocket for the CoA moiety of the substrate and its tetramerization loop, which is an insertion of ∼20 residues at the beginning of the loop-domain, mediates the tetramerization of the thiolase dimers [5]. The tetramerization loop is missing in the dimeric thiolases.

The vertebrate peroxisomal SCP2-thiolase has long been poorly characterized. This dimeric thiolase is also the most divergent member of the thiolase family of enzymes (sharing ∼20% sequence identity with thiolases of any other thiolase subfamily) and, unlike any other thiolase subfamily, its gene also encodes for a subsequent ∼20-residue long linker region followed by the C-terminal domain (∼120 residues), known as the sterol carrier protein 2 (SCP2-protein) [6] (Figure 2). The full-length protein is also referred to as the SCPx-protein [vertebrate protein with the SCP2-thiolase (N-terminal part), being extended at its C-terminus by a 20-residue linker and subsequently its SCP2-domain] [3], whereas the thiolase part is known as the SCP2-thiolase. The SCP2-thiolase is obtained from SCPx by proteolytic cleavage by a peroxisomal protease, which also generates the SCP2-protein. This small basic ∼13 kDa protein has high affinity for acyl-CoA [7,8]. The SCP2-protein is also expressed as a pro-protein from an alternate transcription site of the same gene that encodes the full-length SCPx-protein [9]. The vertebrate SCP2-thiolase occurs in vivo predominantly as the 46 kDa dimeric peroxisomal thiolase [2,10,11]. The C-terminal SCP2-domain is needed for the targeting of full-length SCPx into the peroxisomes since the peroxisomal targeting signal, type-1 (PTS1) [12] is located at the C-terminal end of the SCPx-protein. In vitro enzymatic assays and in vivo animal studies have indicated that the SCP2-thiolase participates in the β-oxidation of very long-chain fatty acyl-CoA and 2-methyl-branched long-chain fatty acyl-CoA [2,13]. The latter conversion is a critical step in the biosynthesis of bile acids from cholesterol. SCP2-thiolase is also proposed to play a role in the degradation of long-chain dicarboxylic acids [14].

The zebrafish and human SCP2-thiolase (type-1) sequences.

Figure 2.
The zebrafish and human SCP2-thiolase (type-1) sequences.

(A) The domains of zebrafish (UniProt: Q6P4V54) and human (UniProt: P22307) SCPx. The ordered part of the structure of the zebrafish SCP2-thiolase (type-1) (Dr-SCP2-thiolase-1) is colored yellow. The domain borders of the SCP2-domain of the human SCPx are as proposed previously [8]. (B) The sequence alignment of Dr-SCP2-thiolase-1 and human SCP2-thiolase (type-1) (Hs-SCP2-thiolase-1). The secondary structure elements of chain B of Dr-SCP2-thiolase-1 are indicated above the sequences. The N-domain, the loop-domain and the C-domain secondary structure elements are labeled in green, orange and red, respectively. The N-terminus and the C-terminus of the ordered part of Dr-SCP2-thiolase-1 are labeled with a >(Asn6) and a # (Pro401), respectively. The C-termini of the three constructs used in these studies are identified as ¤ (Ser406), * (Thr413, Thr424). The sequence fingerprints of the four catalytic loops (Table 1) are identified below the sequences. The CxS/T loop provides the nucleophilic cysteine, Cys87 and the HDCF loop provides the acid/base cysteine, Cys300. His348 of the GHP loop activates the nucleophilic cysteine. The CxG loop contributes to OAH2.

Figure 2.
The zebrafish and human SCP2-thiolase (type-1) sequences.

(A) The domains of zebrafish (UniProt: Q6P4V54) and human (UniProt: P22307) SCPx. The ordered part of the structure of the zebrafish SCP2-thiolase (type-1) (Dr-SCP2-thiolase-1) is colored yellow. The domain borders of the SCP2-domain of the human SCPx are as proposed previously [8]. (B) The sequence alignment of Dr-SCP2-thiolase-1 and human SCP2-thiolase (type-1) (Hs-SCP2-thiolase-1). The secondary structure elements of chain B of Dr-SCP2-thiolase-1 are indicated above the sequences. The N-domain, the loop-domain and the C-domain secondary structure elements are labeled in green, orange and red, respectively. The N-terminus and the C-terminus of the ordered part of Dr-SCP2-thiolase-1 are labeled with a >(Asn6) and a # (Pro401), respectively. The C-termini of the three constructs used in these studies are identified as ¤ (Ser406), * (Thr413, Thr424). The sequence fingerprints of the four catalytic loops (Table 1) are identified below the sequences. The CxS/T loop provides the nucleophilic cysteine, Cys87 and the HDCF loop provides the acid/base cysteine, Cys300. His348 of the GHP loop activates the nucleophilic cysteine. The CxG loop contributes to OAH2.

SCP2-thiolase deficiency, caused by a premature stop codon in the thiolase-coding region, has been reported to cause severe disease symptoms and, in the peroxisomes of the affected patient, both full-length 58 kD SCPx and the 46 kDa SCP2-thiolase could not be detected, whereas the SCP2-domain, having its own promoter, is normally present [11]. Studies of SCPx knockout mice revealed similar metabolic changes as observed in this patient [15].

Phylogenetic analysis of thiolase sequences encoded by eubacterial and eukaryotic genomes [16] has identified two subgroups of SCP2-thiolases. The vertebrate SCP2-thiolases [also referred to as the SCP2-thiolase (type-1)] have the additional SCP2-domain (Figure 2), whereas the other subgroup, without the SCP2-domain, is referred to as the SCP2-thiolase (type-2) [16]. Thiolases of the latter subgroup, of trypanosomes (Tb-SCP2-thiolase-2) and leishmania (Lm-SCP2-thiolase-2), have been characterized [17]. These thiolases occur in the mitochondria and are also referred to as the parasitic SCP2-thiolases. Both SCP2-thiolase subgroups have distinct sequences (Supplementary Figure S1). The enzymological characterization of the trypanosomal SCP2-thiolases (type-2) revealed that this thiolase is more active in the biosynthetic reaction than in the degradative reaction [18]. Indeed, this thiolase is proposed to be involved in the sterol biosynthetic pathway in the parasites [18,19]. Vertebrate SCP2-thiolase has low enzymatic activity with short-chain 3-ketoacyl-CoAs [2], but it accepts the bile acid intermediate with the cholesterol end group [2] as a substrate, in line with its function [11]. Another thiolase that utilizes 3-keto-steroid-CoA esters as a substrate is the dimeric Mycobacterium tuberculosis FadA5-thiolase [20]. The SCP2-thiolase (type-1) substrate, 24-keto-THC-CoA, is much more polar than the substrate (3-oxo-pregn-4-ene-20-acyl-CoA) of the FadA5-thiolase [21], and in the phylogenetic analysis, the FadA5-thiolase is classified as a TFEL1-thiolase [16], which is a different thiolase subfamily when compared with the SCP2-thiolases.

The thiolase reaction mechanism has been extensively studied by structural enzymological approaches using the bacterial biosynthetic tetrameric thiolase [5,2226]. Four catalytic loops have been identified (Table 1), and each of these loops has its own characteristic sequence fingerprint [27] (Figure 2). The thiolase reaction mechanism involves a nucleophilic cysteine activated by a histidine, a general acid/base cysteine and two oxyanion holes (OAHs). The nucleophilic cysteine attacks the substrate and becomes covalently modified. In the degradative direction (Figure 1), the general acid/base cysteine subsequently provides a proton for the leaving propionyl-CoA/acetyl-CoA molecule and, in the next step, it activates the incoming CoA by abstracting a proton. The two OAHs (Table 1) bind and stabilize the negatively charged reaction intermediates formed during the reaction cycle. OAH1 [23] binds the CoA-thioester oxygen atom, stabilizing the negative charge of the enolate intermediate, whereas OAH2 binds the thioester oxygen atom of the acetylated cysteine, stabilizing the negatively charged tetrahedral intermediate [27].

The high-resolution protein crystallographic structures reported here reveal the active site geometry of the zebrafish (Danio rerio) dimeric SCP2-thiolase (type-1) (Dr-SCP2-thiolase-1) both in its unliganded and in its liganded form. The structure of the monomeric form of Dr-SCP2-thiolase-1 has also been captured in this crystal form, suggesting that the Dr-SCP2-thiolase-1 dimer in solution is in equilibrium with its monomer. Native mass spectrometry experiments with Dr-SCP2-thiolase-1 and with its human homolog (Hs-SCP2-thiolase-1) confirm that the SCP2-thiolase (type-1) is a weak transient dimer in solution. The protein–protein interactions at the dimer interface of this thiolase are compared with those of other dimeric thiolases.

Materials and methods

Generation of the zebrafish SCP2-thiolase-Δ406, zebrafish SCP2-thiolase-Δ413 and human SCP2-thiolase-Δ424 expression constructs

The codon-optimized full-length cDNAs of zebrafish SCP2-thiolase (UniProt: Q6P4V5) and human SCP2-thiolase (UniProt: P22307) [28] were purchased from GenScript (U.S.A.), being inserted between the NdeI and BamHI sites of the pGS21a plasmid. The 5′-end of the designed cDNAs encodes the MHHHHHHAM-tag and the MHHHHHHAMA-tag of the zebrafish and human protein, respectively (the latter M denotes the starting Met of the gene). The pGS21a–ZfSCP2 plasmid served as a template in PCR for the synthesis of the zebrafish SCP2-thiolase Δ-variants lacking the SCP2-domain. The PCR was carried out using the Phusion DNA polymerase (Finnzymes) in HighFidelity buffer supplemented with 3% DMSO (dimethyl sulfoxide) and 200 µM dNTPs in a final volume of 50 µl. T7 forward primer was used as a 5′-primer, and 5-gccgcgatacgggatcctcatcaggaactggtttcc and 5-cgatgatgccgaggatcctcatcacgttgagacggc were used as 3′-primers for the zebrafish SCP2-thiolase-Δ406 variant (residues 407–538 deleted) and for the zebrafish SCP2-thiolase-Δ413 variant (residues 414–538 deleted), respectively. The PCR products were digested with the NdeI and BamHI restriction enzymes and ligated into the similarly digested pGS21a vector and transformed into Turbo Escherichia coli (New England Biolabs, U.S.A.) competent cells.

The pGS21a–Hs-SCP2 plasmid was used to make the construct of the human SCP2-thiolase Δ-variant. The codon-optimized full-length human SCP2-thiolase cDNA had been designed to preserve a specific restriction site, which allowed a deletion of the cDNA fragment encoding residues Ala363–Leu547 using NheI and BamHI restriction enzymes followed by ligation of a short synthetic cDNA fragment encoding the residues Ala363–Thr424 into the pGS21a–HsSCPx vector, generating the human SCP2-thiolase-Δ424 variant (residues 425–547 deleted). The DNA sequences of the inserts were confirmed by DNA sequencing.

Protein expression of the zebrafish and human SCP2-thiolase Δ-variants

The zebrafish and human SCP2-thiolase Δ-variants were expressed in the E. coli BL21(DE3) strain (Novagen, Merck Millipore, U.S.A.) containing the pGro7 (Takara Inc., Japan) plasmid encoding the GroEL–GroES complex. Transformed bacterial cells were grown at 37°C in M9ZB media supplemented with 34 µg/ml chloramphenicol, 100 µg/ml ampicillin or 50 µg/ml carbenicillin and 0.3 mg/ml l-arabinose (for induction of the GroEL–GroES complex) until the OD600 was 0.5–0.6. IPTG was added to a final concentration of 0.4 mM and the growth was continued 4–5 h for the zebrafish SCP2-thiolase Δ-variants and 2 h 30 min for the human SCP2-thiolase-Δ424 variant at 30°C. The cells were harvested by centrifugation at 4°C. The cell pellets were suspended in Ni-NTA (Ni2+-nitrilotriacetate) lysis buffer (50 mM NaH2PO4, 0.3 M NaCl, 10 mM imidazole, pH 8.0), and cell suspensions were stored at −70°C for later use.

Purification of the zebrafish SCP2-thiolase Δ-variants

The buffers used in the purification of the zebrafish SCP2-thiolase Δ-variants were supplemented with 10% glycerol. The frozen cell suspensions were thawed at room temperature and diluted with Ni-NTA lysis buffer to give 0.5–1 g of cells/10 ml of buffer. EDTA-free protease inhibitor cocktail (Roche) and β-mercaptoethanol were added to the final concentration of 20 mM. Cells were lysed by 30–45 min incubation with 100 µg/ml lysozyme, 20 µg/ml DNaseI and 2 µg/ml RNaseA at room temperature followed by sonication (10 × 10 s pulse, 30 s cooling), and the homogenate was cleared by centrifugation at 30 000×g for 1 h at 4°C. The supernatant was mixed with Ni-NTA matrix (Qiagen GmbH, Germany) and rotated for 1 h at 4°C. The Ni-NTA matrix was washed with 50–100 bed volumes of Ni-NTA wash buffer (50 mM NaH2PO4, 0.3 M NaCl, 20 mM imidazole, pH 8.0), and the bound proteins were eluted with 7–10 bed volumes of Ni-NTA elution buffer (50 mM NaH2PO4, 0.3 M NaCl, 250 mM imidazole, pH 8.0). Fractions containing the Δ-variants were pooled, dialyzed over night against 2 l of Resource Q anion exchange buffer (50 mM Tris–HCl, 5 mM DTT, 2 mM EDTA, pH 8.5) at 4°C and applied to a 6 ml Resource Q (GE Healthcare, U.S.A.) column. Bound proteins were eluted with a linear gradient of NaCl from 0 to 10% in anion exchange buffer. Pooled fractions were concentrated and applied onto the HiLoad 16/60 prep grade (GE Healthcare, U.S.A.), and gel filtration column was equilibrated with 50 mM Tris–HCl, 5 mM DTT, 2 mM EDTA, 100 mM NaCl (pH 7.5). For the preparation of the sample for mass spectrometry analysis, the glycerol was omitted from the gel filtration buffer of the zebrafish SCP2-thiolase-Δ406 sample. The protein in the peak fractions was concentrated using Amicon Ultra (Merck Millipore Ltd.) centrifugal filters. Protein concentrations were determined by absorption at 280 nm with a Nanodrop spectrophotometer (ThermoFisher Scientific) using the specific extinction coefficient calculated by Protparam [29] of the ExPASy portal [30]. The purified protein samples showed single bands on SDS–PAGE gels and were stored in gel filtration buffer at −70°C for later use. The specific activity of ZfSCP2-thiolase-Δ406 and ZfSCP2-thiolase-Δ413 in the degradative direction was determined at 25°C in 0.5 ml reaction volume using a Jasco V-660 spectrophotometer (Jasco, MD, U.S.A.). The degradation of acetoacetyl-CoA was measured by detecting the formation of the new thioester bond [31], measured at 232 nm (ε = 4500 M−1 cm−1). The reaction mixture consisted of 50 mM Tris–HCl, pH 8.1, 60 μM CoA, 50 μM acetoacetyl-CoA and 2.8 μg of enzyme. The data obtained were analyzed with the Spectra Manager software (Jasco, MD, U.S.A.). The specific activity for the degradation of acetoacetyl-CoA was found to be 0.2 ± 0.1 µmol min−1 mg−1. Such low specific activity for this short-chain substrate was also reported for the mammalian rat SCP2-thiolase (type-2) [2].

Purification of the human SCP2-thiolase-Δ424 variant

For the purification of the human SCP2-thiolase-Δ424 variant, all the buffers, as used for the zebrafish SCP2-thiolase Δ-variants, were supplemented with 1 mM DTT. The frozen cell suspension was thawed and diluted with Ni-NTA lysis buffer to yield ∼0.5 g/10 ml buffer. Lysozyme was added to a final concentration of 0.1 µg/ml, and the cells were disrupted by sonication (6 × 10 s pulse and 30 s cooling). The cell lysate was centrifuged at 30 000×g at 4°C for 45 min. Ni-NTA purification was carried out as for the zebrafish SCP2-thiolase Δ-variants. The eluted human SCP2-thiolase-Δ424 protein was dialyzed against 2 l of 25 mM Tris–HCl, 1 mM EDTA (pH 8.6) overnight at 4°C. The dialyzed sample was loaded onto a 5 ml HiTrap Q (GE Healthcare, U.S.A.) anion exchange column equilibrated with A buffer (25 mM Tris–HCl, 1 mM EDTA, pH 8.6) and eluted with a linear gradient of NaCl from 0 to 40% in A buffer. Fractions containing the human SCP2-thiolase-Δ424 were pooled, concentrated using an Amicon Ultra centrifugal filter and loaded onto a Superdex 16/60 HiLoad preparatory grade gel filtration column equilibrated with 25 mM Tris–HCl, 0.1 M NaCl and 1 mM EDTA (pH 8.6). The purified protein showed a single band as visualized on SDS–PAGE gels. For the preparation of the sample used in the mass spectrometry analysis, the glycerol was omitted from the gel filtration buffer. The protein samples of the peak fractions were pooled and subsequently concentrated and stored in the gel filtration buffer at −70°C as described above for the zebrafish SCP2-thiolase Δ-variants.

Expression and purification of the trypanosomal SCP2-thiolase (type-2)

Tb-SCP2-thiolase-2 (Q57XD5) was expressed and purified as described earlier [32] with minor modifications. For the expression, an overnight culture was used for inoculation of M9ZB media supplemented with 34 µg/ml chloramphenicol, 30 µg/ml kanamycin and 0.3 mg/ml l-arabinose. Bacterial cells were grown until the OD600 was 0.5–0.6. IPTG was added to a final concentration of 0.1 mM and the growth was continued overnight at 20°C. The cell pellets were suspended in lysis buffer (25 mM Tris–HCl, 0.15 M (NH4)2SO4, 15% glycerol, pH 7.8). Before sonication (6 × 10 s pulse and 30 s cooling), ATP, lysozyme and β-mercaptoethanol were added to final concentrations of 5 mM, 0.1 µg/ml and 10 mM, respectively. The cell lysate was centrifuged at 30 000×g at 4°C for 60 min. The supernatant was mixed with Ni-NTA matrix equilibrated with lysis buffer and rotated for 1 h at 4°C. The Ni-NTA matrix was washed with 30 bed volumes of wash buffer (100 mM Tris–HCl, pH 7.8, 0.3 M NaCl, 10% glycerol, 1 mM DTT and 40 mM imidazole), and the bound proteins were eluted with elution buffer (100 mM Tris–HCl, 0.1 M NaCl, 10% glycerol, 300 mM imidazole, 1 mM DTT, pH 8.5). Protein containing fractions were pooled, concentrated and applied onto the HiLoad 16/60 gel filtration column equilibrated with gel filtration buffer (25 mM Tris–HCl, 50 mM NaCl, 1 mM EDTA, 50 mM DTT, pH 8.5). Protein in the peak fractions was concentrated using Amicon Ultra (Merck Millipore Ltd.) centrifugal filters. Protein concentrations were determined by measuring the absorption at 280 nm using the specific extinction coefficient calculated by Protparam. The pure protein samples, as seen on SDS–PAGE gels, were stored at −70°C for later use.

Mass spectrometry

Prior to the mass spectrometric measurements, the protein samples were desalted/buffer exchanged to 400 mM ammonium acetate (pH 7.0) solution, using a PD-10 (Sephadex G-25) column (GE Healthcare, Sweden). Protein concentrations of the eluted fractions were determined by UV absorbance at 280 nm using the sequence-derived extinction coefficient. The eluted fractions were pooled and concentrated using Vivaspin-500 5K MWCO centrifugal concentrators (GE Healthcare). All the reagents and solvents were HPLC quality.

All mass spectra were measured using a 12-T Bruker solariX XR Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (Bruker Daltonics, Bremen, Germany), equipped with a dynamically harmonized Paracell and Apollo-II electrospray ionization (ESI) source, operated in the positive-ion mode. The instrument was calibrated externally using sodium perfluoroheptanoic acid (NaPFHA) clusters. The instrument was operated and data were acquired using ftmsControl 2.0 software, and the mass spectra were further processed and analyzed using DataAnalysis 4.4 software. The protein solution, at 40 µM monomer concentration, was directly infused into the ion source at a flow rate of 2 µl/min. Coenzyme concentration was 120 µM. Drying and nebulizing gas were at 4 l/min and 1 bar, respectively. The ion source temperature was kept at 200°C. For native MS (mass spectrometry) measurements, the instrument parameters were optimized for detecting high mass protein complexes. The time of flight was set to 2.2 ms, and the RF frequencies were set as follows: octopole: 5 MHz; collision cell: 2 MHz; transfer optics: 2 MHz. The Skimmer 1 was set to 50 V. The ions were detected within the mass range of m/z 387–7000. The ion accumulation time was set to 0.1 s and, for each mass spectrum, a total of 256 time-domain transients were summed using 1 Mword dataset size and were zero-filled once to obtain the final 2 MWord magnitude mode data. No spectra apodization was applied.

The native thiolase monomers appeared at m/z ∼2800–3500 and the dimer at m/z ∼4000–5000. The isotopic resolution was obtained for the monomer and the most abundant isotopic mass is therefore reported. For the dimer, isotopic peaks are unresolvable, and the mass is therefore reported as an average mass instead. The Kd values for the protein dimerization were estimated based on the equation Kd = [P]0/(2R2 + R), where [P]0 is the total protein concentration (expressed as monomers) and R = [D]/[M] is a solution concentration ratio of protein dimer (D) and monomer (M). R can be directly obtained from the integrated signal intensities ID and IM using the following equation: R = IM/(IM + 2ID), based on the mass balance equation [P]0 = [M] + 2[D] at equilibrium.

Crystallization of unliganded zebrafish SCP2-thiolase-Δ413 and data collection

All crystallization experiments (Table 2) were carried out using the sitting drop vapor diffusion method with 96-well crystallization plates. Initial screening of the crystallization conditions for SCP2-thiolase-Δ413 was carried out using Proplex (Molecular Dimensions, U.K.), JCSG-Plus (Molecular Dimensions, U.K.), PACT (Molecular Dimensions, U.K.), Crystal Screen I&II (Hampton Research, CA, U.S.A.) and in-house Factorial screen [33] at 20 and 4°C. The crystallization condition, which produced crystals in PACT premier screen, was further optimized. SCP2-thiolase-Δ413 protein was crystallized at room temperature by mixing 3.5 mg/ml protein solution (in gel filtration buffer) in an equal volume with well solution containing 100 mM HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], pH 6.4, 18% PEG (polyethylene glycol)3350 and 250 mM sodium acetate trihydrate in a total volume of 2 µl on a Corning 3556 crystallization plate (Corning Incorporated, NY, U.S.A.). Crystals appear in a few days. For data collection, the crystal was transferred directly from the drop to liquid nitrogen and diffraction data were collected at the I03 beamline of Diamond Light Source (DLS), using a PILATUS 6M detector.

Table 2
Crystallization conditions, crystal handling and statistics of data collection, data processing and refinement
Dataset Unliganded With CoA bound (soaking) With octanoyl-CoA bound (cocrystallization) 
Crystallization 
 Crystallization well solution 100 mM HEPES, pH 6.4, 18% PEG3350, 250 mM sodium acetate 100 mM HEPES, pH 7.0, 15% PEG3350, 250 mM sodium acetate 100 mM MOPS, pH 7.0, 10% PEG6000, 20% glycerol 
 Soaking – 100 mM HEPES, pH 6.9, 300 mM sodium acetate, 20% PEG3350, 12% glycerol, 0.7 mM acetoacetyl-CoA – 
 Cocrystallization – – In the presence of the reaction cocktail of 2E-decenoyl-CoA and NAD+ incubated with MFE1 
 Cryo-cooling Direct transfer in liquid N2 Direct transfer in liquid N2 Direct transfer in liquid N2 
Data collection 
 Beamline DLS I03 DLS I04-1 DLS I04 
 Detector PILATUS 6M PILATUS 2M PILATUS 6M 
 Wavelength (Å) 0.9763 0.9174 1.2822 
Data processing1 
 Space group CCC
a, b, c (Å) 133.16, 81.94, 96.81 132.96, 80.88, 96.47 132.18, 81.32, 96.63 
α, β, γ (o90.00, 125.81, 90.00 90.00, 126.29, 90.00 90.00, 126.08, 90.00 
 Resolution (Å) 29.64–1.95 (2.05–1.95) 47.59–1.46 (1.56–1.46) 42.88–1.73 (1.83–1.73) 
Rmerge (%)2 5.2 (42.4) 11.0 (67.9) 9.1 (63.0) 
 Rpim (%) 3.2 (22.2) 4.0 (19.4) 5.7 (38.1) 
CC1/2 (%) 99.9 (87.1) 99.9 (85.6) 99.6 (40.3) 
 I/σ(I)3 15.7 (2.9) 13.7(2.9) 8.4 (1.3) 
 Completeness (%) 98.4 (99.3) 98.2 (98.2) 99.1 (97.6) 
 Redundancy 3.4 (3.4) 7.5 (7.5) 3.2 (2.6) 
 Observed reflections 205 110 1 055 865 273 378 
 Unique reflections 60 799 140 415 85 694 
 Wilson B-factor (Å234.8 18.4 28.0 
Refinement 
 Rwork4/Rfree5(%) 16.94/20.70 15.65/17.42 19.01/22.73 
Number of atoms 
 Protein atoms in chain A/B 2896/2964 2941/3015 2914/2965 
 CoA/octanoyl-CoA (in chain B) –/– 48/– 48/57 
 Waters 242 556 225 
Average B-factor (Å2
 Protein atoms in chain A/B 34.1/35.1 15.4/13.5 35.1/32.9 
 CoA/octanoyl-CoA (in chain B) –/– 46.2/– 131.5/55.0 
 Waters 37.0 25.1 38.3 
RMS deviations 
 Bond lengths (Å) 0.012 0.009 0.013 
 Bond angles (o1.5 1.6 1.6 
Ramachandran plot6 
 Favored region (%) 98.8 98.8 98.7 
 Allowed region (%) 1.2 1.2 1.3 
 Outlier region (%) 0.0 0.0 0.0 
PDB Code 6HRV 6HSJ 6HSP 
Dataset Unliganded With CoA bound (soaking) With octanoyl-CoA bound (cocrystallization) 
Crystallization 
 Crystallization well solution 100 mM HEPES, pH 6.4, 18% PEG3350, 250 mM sodium acetate 100 mM HEPES, pH 7.0, 15% PEG3350, 250 mM sodium acetate 100 mM MOPS, pH 7.0, 10% PEG6000, 20% glycerol 
 Soaking – 100 mM HEPES, pH 6.9, 300 mM sodium acetate, 20% PEG3350, 12% glycerol, 0.7 mM acetoacetyl-CoA – 
 Cocrystallization – – In the presence of the reaction cocktail of 2E-decenoyl-CoA and NAD+ incubated with MFE1 
 Cryo-cooling Direct transfer in liquid N2 Direct transfer in liquid N2 Direct transfer in liquid N2 
Data collection 
 Beamline DLS I03 DLS I04-1 DLS I04 
 Detector PILATUS 6M PILATUS 2M PILATUS 6M 
 Wavelength (Å) 0.9763 0.9174 1.2822 
Data processing1 
 Space group CCC
a, b, c (Å) 133.16, 81.94, 96.81 132.96, 80.88, 96.47 132.18, 81.32, 96.63 
α, β, γ (o90.00, 125.81, 90.00 90.00, 126.29, 90.00 90.00, 126.08, 90.00 
 Resolution (Å) 29.64–1.95 (2.05–1.95) 47.59–1.46 (1.56–1.46) 42.88–1.73 (1.83–1.73) 
Rmerge (%)2 5.2 (42.4) 11.0 (67.9) 9.1 (63.0) 
 Rpim (%) 3.2 (22.2) 4.0 (19.4) 5.7 (38.1) 
CC1/2 (%) 99.9 (87.1) 99.9 (85.6) 99.6 (40.3) 
 I/σ(I)3 15.7 (2.9) 13.7(2.9) 8.4 (1.3) 
 Completeness (%) 98.4 (99.3) 98.2 (98.2) 99.1 (97.6) 
 Redundancy 3.4 (3.4) 7.5 (7.5) 3.2 (2.6) 
 Observed reflections 205 110 1 055 865 273 378 
 Unique reflections 60 799 140 415 85 694 
 Wilson B-factor (Å234.8 18.4 28.0 
Refinement 
 Rwork4/Rfree5(%) 16.94/20.70 15.65/17.42 19.01/22.73 
Number of atoms 
 Protein atoms in chain A/B 2896/2964 2941/3015 2914/2965 
 CoA/octanoyl-CoA (in chain B) –/– 48/– 48/57 
 Waters 242 556 225 
Average B-factor (Å2
 Protein atoms in chain A/B 34.1/35.1 15.4/13.5 35.1/32.9 
 CoA/octanoyl-CoA (in chain B) –/– 46.2/– 131.5/55.0 
 Waters 37.0 25.1 38.3 
RMS deviations 
 Bond lengths (Å) 0.012 0.009 0.013 
 Bond angles (o1.5 1.6 1.6 
Ramachandran plot6 
 Favored region (%) 98.8 98.8 98.7 
 Allowed region (%) 1.2 1.2 1.3 
 Outlier region (%) 0.0 0.0 0.0 
PDB Code 6HRV 6HSJ 6HSP 
1

Values in parentheses refer to the highest resolution shell.

2

Rmerge = ΣhklΣi|Ii(hkl) − 〈I(hkl)〉|)/ΣhklΣiIi(hkl), where Ii(hkl) is the intensity of the ith measurement of reflection (hkl) and 〈I(hkl)〉 is its mean intensity.

3

I is the integrated intensity and σ(I) is its estimated standard deviation.

4

Rwork = (Σhkl|Fo − Fc|)/ΣhklFo, where Fo and Fc are the observed and calculated structure factors, respectively.

5

Rfree is calculated as for Rwork but from a randomly selected subset of the data (5%), which were excluded from the refinement calculations.

6

Calculated by PROCHECK [46].

Soaking of a zebrafish SCP2-thiolase-Δ406 crystal with acetoacetyl-CoA and data collection

For soaking experiments, the zebrafish SCP2-thiolase-Δ406 variant was crystallized at 4°C in 100 mM HEPES, pH 7.0, 15% PEG3350, 250 mM sodium acetate trihydrate by mixing an equal volume of 4.5 mg/ml protein solution (in gel filtration buffer) and well solution on a TTP Labtech iQ plate (TTP Labtech, U.K.) in a total volume of 2 µl. For soaking with acetoacetyl-CoA (Table 2), the crystals were transferred into a soaking solution containing 100 mM HEPES, pH 6.9, 300 mM sodium acetate trihydrate, 20% PEG3350, 12% glycerol, 0.7 mM acetoacetyl-CoA for 25 min at 4°C. In a 0.7 mM acetoacetyl-CoA solution, CoA is also present, as determined using Ellman's reagent corresponding to ∼40 µM CoA. After incubation for 25 min, the crystals were directly transferred into liquid nitrogen. Diffraction data were collected at the I04-1 beamline of DLS, using a PILATUS 2M detector.

Cocrystallization of zebrafish SCP2-thiolase-Δ406 with substrate mixture and data collection

For the cocrystallization experiment of zebrafish SCP2-thiolase-Δ406 with a substrate mixture (Table 2), 74 nmol of 2E-decenoyl-CoA was converted into 3-ketodecanoyl-CoA by incubating with the recombinant rat multifunctional enzyme, type-1 (MFE1) [34] in 50 mM Tris–HCl, 50 mM KCl, pH 8.0, supplemented with 400 nmol of NAD+ in a final volume of 50 µl, at room temperature for 8 min. MFE1 is a multifunctional enzyme, having two active sites. Its hydratase activity will catalyze the hydration of 2E-decenoyl-CoA into 3S-hydroxydecanoyl-CoA and its dehydrogenase active site will subsequently catalyze the NAD+-dependent dehydrogenation of 3S-hydroxydecanoyl-CoA into 3-ketodecanoyl-CoA with the formation of NADH. The final equilibrium mixture will contain NAD+, NADH, 2E-decenoyl-CoA, 3S-hydroxydecanoyl-CoA, 3-ketodecanoyl-CoA and CoA. CoA can be present as an impurity (in the 2E-decenoyl sample) from the beginning or can have come from the hydrolysis of the CoA derivatives in the incubation buffers, or both. MFE1 was removed from the reaction mixture with 25 µl of Ni-NTA matrix (Qiagen, Germany) equilibrated with the reaction buffer and filtered through a 0.2 µm centrifugal filter (Millipore). The filtrate (theoretical maximal concentration of 3-ketodecanoyl-CoA is 1.5 mM) was mixed in a 1 : 1.6 ratio with 11.3 mg/ml zebrafish SCP2-thiolase-Δ406 protein sample (in gel filtration buffer), and crystallization experiments were set up at 4°C with the in-house Factorial screen [33], Proplex and JCSG-Plus screens with 300 nl drop size with an equal ratio of well and protein solution on a TTP Labtech iQ plate. Crystals appeared in the Factorial screen condition composed of 100 mM MOPS [3-(N-morpholino)-propanesulfonic acid], pH 7.0, 10% PEG6000 and 20% glycerol, and they were transferred directly into liquid nitrogen on the fifth day from setting up the crystallization experiments. Diffraction data were collected at the I04 beamline of DLS, using a PILATUS 6M detector.

Data processing and structure determination

All crystals belonged to the monoclinic C2 space group and the unit cell parameters for the unliganded crystal structure are 133.16 Å, 81.94 Å, 96.81 Å, 90.00°, 125.81° and 90.00°. The Matthew's coefficient, calculated with MATTHEWS_COEF [35,36], which is part of the CCP4 package [37], is 2.3 Å3/Da with two subunits in the asymmetric unit giving 47% solvent content.

The diffraction data of the unliganded crystal and the reaction mixture cocrystallization crystal were processed using the XDS program package [38]. For the acetoacetyl-CoA, soaked crystal XDSAPP2 [39] was used for diffraction data processing. The data collection statistics of the three data sets are summarized in Table 2.

The structure of the unliganded crystal form was solved by the molecular replacement method with PHASER [40] using one subunit of the Leishmania mexicana SCP2-thiolase (PDB code: 3ZBG) [18] as a search model. Non-conserved amino acids were pruned to the gamma atom using CHAINSAW [41], and only the most conserved part of the thiolase fold was used for molecular replacement. The molecular replacement software PHASER gave a single solution with reasonable Z-scores (RTF Z-score 4.8 and TF Z-score 5.8 for the first monomer and 11.5 for the second monomer). Crystal symmetry generates the classical thiolase dimer from one of these monomers, whereas the other monomer is not part of a dimer. The molecular replacement solution was restrain refined with REFMAC5 applying jelly-body restraints [42]. The refined molecular replacement model was autotraced with SHELXE [43], with 0.5 solvent fraction, to give a pseudo-free correlation coefficient of 71.4% with 26 built peptides consisting of 650 residues. The obtained poly-alanine model was refined using REFMAC5, and the phase information of the output reflection data file was combined with the amino acid sequence information using the automatic model building features of BUCCANEER [44], which uniquely allocated 775 residues to two chains with 94% completeness.

Structure refinement

Subsequently, the unliganded structure was completely checked against the electron density maps and refined at 1.95 Å resolution using COOT [45], REFMAC5, PROCHECK [46] and MOLPROBITY [47] via the PHENIX graphical user interface [48]. Waters and solvent molecules were added iteratively, and the refinement was stopped when the remaining peaks (above 5 sigma) in the (Fo − Fc) maps did not provide significant structural information. The refinement statistics are summarized in Table 2.

The zebrafish SCP2-thiolase-Δ406 unliganded crystal that was soaked in the presence of 0.7 mM acetoacetyl-CoA diffracted to 1.46 Å resolution. The initial structure of the acetoacetyl-CoA-soaked crystal structure was obtained by refining the unliganded structure without any solvent molecules against the new data set. The positive electron density (Supplementary Figure S2) in the active site was modeled as CoA, as there was no density for the acyl-moiety. This structure will therefore also be referred to as the structure with bound CoA. The co-ordinate set of the idealized structure of CoA was obtained from the HIC-Up server [49], and the dictionary files used in the refinement and in COOT were generated using the PRODRG2 server [50]. The model building, refinement and validation were done as described above for the unliganded structure. The refinement statistics are summarized in Table 2.

The crystal obtained from the reaction mixture cocrystallization experiment of zebrafish SCP2-thiolase-Δ406 appeared in a few days in 100 mM MOPS, pH 7.0, 10% PEG6000, 20% glycerol condition at 4°C, and a 1.73 Å resolution data set was collected. The space group and the packing of the molecules in the cocrystallization crystal are the same as in the unliganded crystal. The initial model of the cocrystallization structure was obtained by the molecular replacement method with PHASER using the unliganded structure without any solvent molecules as a search model. The molecular replacement solution was refined with REFMAC5 and waters were added before manual fitting of the ligands into the (Fo − Fc) electron density map. The positive electron density in the active site was modeled as octanoyl-CoA. The co-ordinates of the idealized structure of octanoyl-CoA were obtained from the HIC-UP server [49], and the dictionary files used in the refinement and in COOT were generated using the PRODRG2 server [50]. The model was manually rebuilt, refined and validated iteratively as described for the unliganded structure. The active site ligand eventually was modeled as a double conformation of octanoyl-CoA and CoA (adopting the same conformation as observed in the CoA complexed structure) with occupancies of 0.7 and 0.3, respectively. This structure will also be referred to as the structure with bound octanoyl-CoA. The fit of the active site ligands in the omit electron density map is shown in Supplementary Figure S3. The refinement statistics are summarized in Table 2.

Structure analysis

The peroxisomal zebrafish SCP2-thiolase (type-1) crystal structures have been compared with (i) other dimeric thiolase structures that have an open active site: the peroxisomal AB-thiolases of yeast (PDB code: 1AFW) [4], of sunflower (PDB code: 2WUA) [51] and of Arabidopsis thaliana (PDB codes: 2WU9 [51] and 2C7Y [52]) and with (ii) the bacterial M. tuberculosis dimeric TFEL1 FadA5-thiolase (PDB codes: 4UBT [21] and 5ONC [53]), which also has an open active site, as well as with (iii) the dimeric SCP2-thiolases (type-2) of L. mexicana (PDB codes: 3ZBG [18], 5LNQ [17]) and of Trypanosoma brucei brucei (PDB code: 4BI9 [18]), that have closed active sites. The superpositions were carried out with the SSM [54] option in COOT.

Amino acid sequences of zebrafish and human SCP2-thiolases (type-1) were aligned with Clustal O [55]. Secondary structure elements were assigned according to DSSP [56] and amino acid sequence alignment was output with ESPript [57].

Results

Presence of monomeric and dimeric forms of the zebrafish SCP2-thiolase in the crystal and in solution

Three crystal structures of Dr-SCP2-thiolase-1 have been obtained and refined, respectively, to 1.95 Å (the unliganded crystal structure), to 1.46 Å (with bound CoA; soaked) and to 1.73 Å (with bound octanoyl-CoA; cocrystallized) (Table 2). In each of these three structures, the crystal packing of the protein molecules is the same. There are two chains in the asymmetric unit, referred to as chain A and chain B. Chain A exists as a monomer, whereas chain B forms the classical thiolase dimer with its crystallographic symmetry mate. Despite having tested a wide range of other crystallization and cocrystallization conditions, no other crystal forms were obtained. The crystals were grown in sitting drops in which the initial protein monomer concentration varied between 1.7 and 3.5 mg/ml (38 and 78 μM (of the monomer), respectively). The crystals appeared in HEPES, HEPES and MOPS buffer, pH 6.4, 7.0 and 7.0, in the presence of 18% PEG3350, 20% PEG3350 and 10% PEG6000, respectively (Table 2). In all structures, the N-terminal and C-terminal residues are disordered, such that the built models start from Asn6 and extend to Pro401 (Figure 2), except in the soaked crystal structure where chain A ends at Phe400. The N-terminal and C-terminal ends are at the backside of the thiolase dimer, far away from the active sites that are located at its front side (Figure 3). In all three structures, the electron density is not visible for the disordered region at the beginning of the loop-domain of both chains. In chain B, 6, 8 and 7 residues are disordered starting at Arg119, Glu118 and Arg119, respectively, in the unliganded, soaked and cocrystallized structures. In chain A, 13, 14 and 13 residues are disordered, starting at Arg119, Glu118 and Glu118, respectively, in the unliganded, soaked and cocrystallized structures. The Nβ3–Nα3 loop is well ordered in chain B of each of the three structures, whereas in chain A it is a high B-factor loop and in the unliganded structure Gly90 of this loop has not been modeled. The other region with high B-factors, near Arg237, is in the helical part between Nβ5 and Cβ1, at the backside of the molecule, of each of the chains in each of the structures.

The structure of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Figure 3.
The structure of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Top view, along the dimer two-fold axis, being identified as a black oval. In this view, both active sites are in the front, whereas the N- and C-termini are at the backside (furthest from this point of view). The octanoyl-CoA molecules are shown as gray (green subunit) and blue (brown subunit) sticks. Tyr144 marks the C-terminal region of the Lα(134–144)-helix of the green subunit. The * marks the hinge region (at Pro150, at the beginning of the Lα2-helix) of the movement that rotates the C-terminal tip of Lα(134–144) towards the pantetheine-binding pocket (labeled as ‘p’). The locations of the active sites with respect to the dimer interface are also visualized in Supplementary Movie S1.

Figure 3.
The structure of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Top view, along the dimer two-fold axis, being identified as a black oval. In this view, both active sites are in the front, whereas the N- and C-termini are at the backside (furthest from this point of view). The octanoyl-CoA molecules are shown as gray (green subunit) and blue (brown subunit) sticks. Tyr144 marks the C-terminal region of the Lα(134–144)-helix of the green subunit. The * marks the hinge region (at Pro150, at the beginning of the Lα2-helix) of the movement that rotates the C-terminal tip of Lα(134–144) towards the pantetheine-binding pocket (labeled as ‘p’). The locations of the active sites with respect to the dimer interface are also visualized in Supplementary Movie S1.

The existence of monomeric and dimeric Dr-SCP2-thiolase-1 in the same crystal form suggests that the protein also exists both as monomer and as dimer in solution, which is typical for so-called weak transient dimers [58]. Native MS was therefore used to determine the oligomeric state of Dr-SCP2-thiolase-1 in solution. The native mass spectrum of 40 μM (of the monomer) Dr-SCP2-thiolase-1 measured in 400 mM ammonium acetate (pH 6.9) clearly indicates the presence of both monomeric and dimeric protein forms (Figure 4), with an approximate Kd of 38 μM, calculated on the basis of the monomer–dimer intensity ratio (see Materials and Methods for details). This solution behavior of Dr-SCP2-thiolase-1 explains why its monomeric and dimeric structures have been captured in this crystal form. Furthermore, the native mass spectrum of Dr-SCP2-thiolase-1 measured in the presence of 120 μM CoA showed that both the monomer and the dimer are able to bind the ligand, the dimer having a slightly higher affinity (Figure 4), and that the ligand binding does not shift very much the monomer–dimer equilibrium. The CoA-binding affinity is estimated to be in the high micromolar range. Since the protein is a mixture of monomers and dimers and both protein forms are able to bind the ligand, it was not attempted to further determine the ligand-binding affinities, for example, by titration experiments.

ESI FT-ICR mass spectra of Dr-SCP2-thiolase-1.

Figure 4.
ESI FT-ICR mass spectra of Dr-SCP2-thiolase-1.

(A) In denaturing solution conditions (MeCN/H2O, 1:1 v/v + 4% HOAc; protein concentration 10 µM). (B,C) In native solution conditions (400 mM NH4OAc, pH 7.0; protein concentration 40 µM) in the absence (B) and presence (C) of 120 µM CoA ligand. The numbers denote different protein ion charge states, and M and D refer to protein monomer and dimer, respectively.

Figure 4.
ESI FT-ICR mass spectra of Dr-SCP2-thiolase-1.

(A) In denaturing solution conditions (MeCN/H2O, 1:1 v/v + 4% HOAc; protein concentration 10 µM). (B,C) In native solution conditions (400 mM NH4OAc, pH 7.0; protein concentration 40 µM) in the absence (B) and presence (C) of 120 µM CoA ligand. The numbers denote different protein ion charge states, and M and D refer to protein monomer and dimer, respectively.

Presence of monomeric and dimeric forms of human SCP2-thiolase (type-1) and trypanosomal SCP2-thiolase (type-2) in solution

The observation that Dr-SCP2-thiolase-1 occurs as a weak transient dimer prompted us to investigate also the possible dimerization of two other thiolases from the SCP2-thiolase family. The native mass spectrum of Hs-SCP2-thiolase-1 (having 72% sequence identity with Dr-SCP2-thiolase-1) measured at 40 μM protein concentration showed clear signals for both monomeric and dimeric species, suggesting weak transient dimer formation (Figure 5A). However, the dimeric fraction was significantly higher, which indicates stronger dimeric association when compared with Dr-SCP2-thiolase-1. The estimated Kd of the Hs-SCP2-thiolase-1 dimer is 0.8 μM. In contrast, Tb-SCP2-thiolase-2 showed practically complete dimerization at 40 μM, suggesting a very stable dimer (Figure 5C). Similar to Dr-SCP2-thiolase-1, each of the forms of human and trypanosomal SCP2-thiolase were observed to bind CoA (Figures 5B,D).

ESI FT-ICR mass spectra of Hs-SCP2-thiolase-1 (A,B) and Tb-SCP2-thiolase-2 (C,D).

Figure 5.
ESI FT-ICR mass spectra of Hs-SCP2-thiolase-1 (A,B) and Tb-SCP2-thiolase-2 (C,D).

In native solution conditions (400 mM NH4OAc, pH 7; protein concentration 40 μM), in the absence (A,C) and presence (B,D) of 120 μM CoA ligand. The numbers denote different protein ion charge states, and M and D refer to protein monomer and dimer, respectively.

Figure 5.
ESI FT-ICR mass spectra of Hs-SCP2-thiolase-1 (A,B) and Tb-SCP2-thiolase-2 (C,D).

In native solution conditions (400 mM NH4OAc, pH 7; protein concentration 40 μM), in the absence (A,C) and presence (B,D) of 120 μM CoA ligand. The numbers denote different protein ion charge states, and M and D refer to protein monomer and dimer, respectively.

Competent active site is located at the dimer interface

Chain B forms in the crystal the classical thiolase dimer [4] with a neighboring chain B, because of a crystallographic two-fold axis. The overall structure of the dimer is shown in Figure 3 and Supplementary Movie S1. The two active sites of the thiolase dimer are at the dimer interface. Nevertheless, the catalytic residues of each active site are provided by the same monomer, but the residues at the N-terminus of Nα2 of the other subunit also contribute to the shape of the substrate-specificity pocket. The center of this dimer interface is formed by the two Nβ3 β-strands (Asn82–Val83–Asn84), which are running antiparallel to each other, being related by the nearby dimer two-fold axis (Figure 6). The Nβ3-strand is the edge β-strand of the β-sheet of the N-terminal core domain of each monomer, and these interactions therefore generate an extended β-sheet of β-strands coming from both subunits. The Nβ3-strand continues in the Nβ3–Nα3 loop, which provides the nucleophilic cysteine. Other extensive monomer–monomer hydrophobic and hydrogen bond protein–protein interactions are ‘above’ and ‘below’ (Figure 6) the Nβ3 antiparallel β-strands. The latter interactions are formed by the C-terminal regions of the Nα3 helices of both monomers (residues 96– 106, which are at the backside of the molecule, far away from the active site). The former interactions are formed by the Nα2′-helix (the ′ indicates that this residue belongs to the opposing subunit) of the opposing chain and the Nβ3–Nα3 loop at the active site. All these interactions are direct protein–protein interactions, not mediated via waters. Near the active site, the hydroxyl group of Ser62′, located just before the Nα2′ helix, is hydrogen bonded to OD1(Asn84) and the main chain N(Asn86) in the Nβ3–Nα3 loop. Gln66′ in Nα2′ is also interacting with the Nβ3–Nα3 catalytic loop and also with the nearby Cβ4–Cβ5 catalytic loop, forming hydrogen bonds with its NE2 group to OD1(Asn85) as well as to Gly389(O). Arg67′ in Nα2′ forms a salt bridge with Asp134 at the beginning of the Lα(134–144)-helix of the loop-domain. Tyr70′, located immediately after the Nα2′ helix, forms with its OH group hydrogen bonds to Gly389(O) in the Cβ4–Cβ5 loop and OD2(Asp255) and NH(Asp255) in the Cβ1–Cα1 loop (Figure 6 and Supplementary Figure S4). The side chains of Tyr70′ and Leu75′ are in van der Waals contact with Phe260 of the Cβ1–Cα1 loop region (Supplementary Figure S4). The residues involved in these interactions are very well conserved in the vertebrate SCP2-thiolase sequences (Supplementary Figure S1). Other interactions of side chains of the opposing subunit with residues near the active site region concern Arg129′ (just before the Lα(134–144) helical region of the loop-domain) that interacts with the main chain of Lys116 (beginning of the loop-domain).

The dimer interface of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Figure 6.
The dimer interface of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Top view (stereo), along the two-fold axis. The interactions with the Nα2’ helix of the other subunit (brown) are highlighted. (A) The Nβ3–Nα3 loop (including Asn84 and Asn87) is at the center of the monomer–monomer interactions of the dimer interface. The mode of binding of octanoyl-CoA is shown in gray sticks. (B) Zoomed-in view of the active site. The mode of binding of octanoyl-CoA (CO8) with respect to the four catalytic loops, being the CxS/T loop (87–89), the HDCF loop (298–301), the GHP loop (347–349) and the CxG loop (386–389), is visualized. Also shown is the Lα2-helix (Ala151, Phe156). Cys87 is the nucleophilic cysteine. Cys300 is the acid/base cysteine. OAH1 is formed by the NE2 atoms of His298 and His348, and OAH2 is formed by N(Cys87) and N(Leu387).

Figure 6.
The dimer interface of the zebrafish SCP2-thiolase dimer, complexed with octanoyl-CoA.

Top view (stereo), along the two-fold axis. The interactions with the Nα2’ helix of the other subunit (brown) are highlighted. (A) The Nβ3–Nα3 loop (including Asn84 and Asn87) is at the center of the monomer–monomer interactions of the dimer interface. The mode of binding of octanoyl-CoA is shown in gray sticks. (B) Zoomed-in view of the active site. The mode of binding of octanoyl-CoA (CO8) with respect to the four catalytic loops, being the CxS/T loop (87–89), the HDCF loop (298–301), the GHP loop (347–349) and the CxG loop (386–389), is visualized. Also shown is the Lα2-helix (Ala151, Phe156). Cys87 is the nucleophilic cysteine. Cys300 is the acid/base cysteine. OAH1 is formed by the NE2 atoms of His298 and His348, and OAH2 is formed by N(Cys87) and N(Leu387).

Properties of the active site of the dimer, complexed with CoA and with octanoyl-CoA

The crystal structures with bound CoA and with bound octanoyl-CoA were determined using, respectively, a soaking protocol and a cocrystallization protocol (Table 2). Both protocols result in the same crystal form as was also obtained for the unliganded Dr-SCP2-thiolase-1. In the soaking protocol, the crystal was transferred into a mother liquor that includes 0.7 mM acetoacetyl-CoA for 25 min, before cryo-cooling. In this incubation, mixture CoA is also present. The electron density map shows that CoA is bound in the dimer active site of chain B (Supplementary Figure S2). In the cocrystallization experiment, Dr-SCP2-thiolase-1 is cocrystallized with a reaction mixture that is made by incubating 2E-decenoyl-CoA with MFE1 and NAD+. This reaction mixture, which includes NAD+, NADH, 2E-decenoyl-CoA, 3S-hydroxydecanoyl-CoA, 3-ketodecanoyl-CoA and CoA, after removal of MFE1, was mixed with Dr-SCP2-thiolase-1, which then can catalyze the thiolytic cleavage of 3-ketodecanoyl-CoA to acetyl-CoA, resulting in an enzyme with the octanoyl-group attached to the nucleophilic Cys87. However, a covalent modification of the catalytic cysteine was not seen in the final electron density maps. The positive density in the active site (Supplementary Figure S3) could best be modeled as a double conformation of CoA (occupancy 0.3) and octanoyl-CoA (occupancy 0.7). It is possible that the zebrafish SCP2-thiolase has catalyzed the first step of the thiolase reaction cycle and produced acetyl-CoA and octanoylated cysteine. Subsequently, the octanoyl moiety is transferred to CoA and the octanoyl-CoA remains bound. Alternatively, or both, 2E-decenoyl-CoA is bound, but the terminal two-carbon atoms are disordered. In the soaked and in the cocrystallization structures, the ligand is only bound in the active site of chain B. Soaking of the unliganded crystals with acetoacetyl-CoA does not disrupt the crystal packing nor does the cocrystallization with substrate cause any conformational differences in the molecules and/or rearrangement of the molecules in the crystal lattice. The latter observation indicates that in the presence of CoA and these acyl-CoA molecules, the dimer is also in equilibrium with the monomer.

The CoA complexed structures show that the CoA moiety is bound in a binding pocket that is completely open to the bulk solvent (Figure 3), and which is mainly formed by residues of the loop-domain, as also described previously [59]. The adenine moiety is wedged between Leu214, Leu215 and Phe193, Pro219 (Figure 7 and Supplementary Figure S5) and the phospho-groups interact with Arg206 and Lys20. Ala152, Phe156 of Lα2, Phe300 (of the HDCF loop) and Cys218 (of the pantetheine loop, just before Nβ4) point towards the pantetheine moiety (Figure 7). The acyl-moiety of octanoyl-CoA is bound in such a way that it points away from the catalytic pocket, being bent towards its pantetheine moiety as well as to Glu118 (end of Nβ4) and Pro150 at the beginning of Lα2 (Figure 6). The fatty acyl tail has only van der Waals interactions with the protein but the thioester oxygen is hydrogen bonded to OH(Tyr59). The substrate-specificity pocket is also completely solvent exposed. Its precise shape is unknown as the loop residues Arg119–Gly120–Ser121–Leu122–Ser123–Ser124–Lys125, just after Nβ4, are disordered. Such disorder at the beginning of the loop-domain has not been observed in any other thiolase structure. These disordered loop residues line the substrate-specificity pocket, which is completed by residues of the Lα(134–144)-helix and the Lα2-helix (Figure 6), plus residues of the Nβ3–Nα3 loop as well as by the Nβ2–Nα2 loop of the other subunit of the dimer (Figure 6). Ser62′ of the latter loop interacts with Asn84 and Asn86 of the NNN-sequence motif (residues 84–86) of the Nβ3–Nα3 loop, just before the nucleophilic cysteine, Cys87. The sequences of these two loops are conserved in the sequence alignment of the vertebrate SCP2-thiolases (Supplementary Figure S1).

The mode of binding of CoA.

Figure 7.
The mode of binding of CoA.

(Top view) The adenine ring is buried and wedged between Phe193, Leu214 and Pro219. The phospho-groups interact with Lys20 and Arg206. The pantetheine moiety interacts with Phe156 (of Lα2), Phe301 (of the HDCF motif) and Cys218 (of the pantetheine loop). Also shown are residues of the four catalytic loops, labeled as CxS/T, GHP, HDCF and CxG. The pantetheine loop is labeled with ‘p’. The red sphere is the water molecule bound in OAH1, shaped by the NE2 atoms of His298 (of the HDCF motif) and His348 (of the GHP motif).

Figure 7.
The mode of binding of CoA.

(Top view) The adenine ring is buried and wedged between Phe193, Leu214 and Pro219. The phospho-groups interact with Lys20 and Arg206. The pantetheine moiety interacts with Phe156 (of Lα2), Phe301 (of the HDCF motif) and Cys218 (of the pantetheine loop). Also shown are residues of the four catalytic loops, labeled as CxS/T, GHP, HDCF and CxG. The pantetheine loop is labeled with ‘p’. The red sphere is the water molecule bound in OAH1, shaped by the NE2 atoms of His298 (of the HDCF motif) and His348 (of the GHP motif).

The catalytic site, at the bottom of the binding pocket of the acyl-moiety, is formed by the four catalytic loops: the Nβ3–Nα3 loop (CxS/T motif), the Cβ4–Cβ5 loop, the Cβ2–Cα3 loop (HDCF motif) and the Cβ3–Cα3 loop (the GHP loop) (Figures 6 and 7). These loops provide the hydrogen bond donors of the two OAHs and also the nucleophilic cysteine (Cys87 of the CxS/T motif) and the acid/base cysteine (Cys300 of the HDCF motif). The hydrogen bond donors of OAH1 are the NE2 atoms of the histidine side chains of the HDCF and GHP loops, like in the parasitic SCP2-thiolases Lm-SCP2-thiolase-2 and Tb-SCP2-thiolase-2. Similar to these parasitic SCP2-thiolases, both histidine side chains are anchored via hydrogen bond networks to their ND1 moieties [17]. A water molecule has been modeled in OAH1 in each of the three structures. The hydrogen bond donors of OAH2 are N(Cys87) and N(Leu387) (Figure 6). Leu387 is in the Cβ4–Cβ5 loop, like in the parasitic SCP2-thiolases, but the sequence fingerprint of this loop [27] (Table 1) and its conformation is not conserved. In the unliganded crystal structure, the nucleophilic cysteine, Cys87, is fully oxidized to its sulfenic acid derivative. In the CoA complexed structure, Cys87 is partially oxidized to its sulfenic acid derivative. In both cases, the hydroxyl group of the sulfenic acid group points into OAH2. In the octanoyl-CoA complexed structure, Cys87 of chain B appears not to be oxidized and a bound water molecule is modeled in OAH2. Cys87 of chain A is not oxidized in any of the three structures.

Structural differences between the folds of the single monomer and the dimerized monomer of the zebrafish SCP2-thiolase

The overall structures of the single monomer and the monomer complexed in the dimer are the same, but there are striking differences in the regions involved in the dimer interface (Figure 8). In both monomers, the beginning of the loop-domain is disordered. There are four loop regions that undergo major conformational changes, being (i) Nβ3–Nα3 (residues Val83 till Thr92), (ii) Cβ4–Cβ5 (residues Asn384–Ala390) (Figure 8), (iii) Cβ1–Cα1 (residues Asp255–Ser264) and (iv) Nβ4–Lα2, (the beginning of the loop-domain, Met117–Pro150). The Nβ3–Nα3 loop, the Nβ4–Lα2-region and the Cβ4–Cβ5 region also generate the geometry of the active site. In the dimer, these loops interact with the Nβ2′–Nα2′ loop of the opposing monomer, but only small conformational changes are observed for the latter loop when comparing the single monomer and the dimerized monomer. Also, the residues 96–106 at the C-terminal regions of the Nα3-helices, at the backside of the molecule (Figure 6), which have significant intra-dimer interactions with each other, do not show conformational changes.

Structural differences between the folds of the single monomer and the dimerized monomer of the zebrafish SCP2-thiolase (type-2).

Figure 8.
Structural differences between the folds of the single monomer and the dimerized monomer of the zebrafish SCP2-thiolase (type-2).

Stereo view, rotated about the vertical axis with respect to Figure 6. (A) Comparison of the superimposed chain B (green) and chain A (gray). Key residues are labeled in both chains to identify the loops that change conformation, being Cys87 (after Nβ3), Met117 (end of Nβ4), Tyr144 [C-terminal region of Lα(134–144), Phe260 (after Cβ1) and Leu387 (after Cβ4)]. (B) The rotation of Lα(134–144) from the dimer interface (green) to the monomer conformation (gray). Viewed from the center of the other subunit. ‘p’ identifies the pantetheine loop (residues Cys217–Thr220), just before Nβ5. The hydrogen bond interactions of the Gln154 side chain with the two conformations of the Lα(134–144)-helix are visualized by dotted lines. These structural differences are also visualized in Supplementary Movie S2.

Figure 8.
Structural differences between the folds of the single monomer and the dimerized monomer of the zebrafish SCP2-thiolase (type-2).

Stereo view, rotated about the vertical axis with respect to Figure 6. (A) Comparison of the superimposed chain B (green) and chain A (gray). Key residues are labeled in both chains to identify the loops that change conformation, being Cys87 (after Nβ3), Met117 (end of Nβ4), Tyr144 [C-terminal region of Lα(134–144), Phe260 (after Cβ1) and Leu387 (after Cβ4)]. (B) The rotation of Lα(134–144) from the dimer interface (green) to the monomer conformation (gray). Viewed from the center of the other subunit. ‘p’ identifies the pantetheine loop (residues Cys217–Thr220), just before Nβ5. The hydrogen bond interactions of the Gln154 side chain with the two conformations of the Lα(134–144)-helix are visualized by dotted lines. These structural differences are also visualized in Supplementary Movie S2.

The conformational differences between the monomeric chain A and the dimerized chain B can be described as follows: (i) the N-terminus of the Nα3 helix (Gly89–Gln102) unfolds and the region Asn83 to Ser92 of the active site Nβ3–Nα3 loop (CxS/T motif) rotates outwards, away from the compact thiolase core domain. The Nα3 helix (Leu93–Gln102) in chain A is four residues shorter than in chain B (Gly89–Gln102). The nucleophilic cysteine (Cys87), located in this loop, points to the bulk solvent in chain A, together with Asn86. In chain A, the residues Ser88–Gly90 adopt a more extended conformation with high B-factors (60–70 vs. 30 Å2) compared with chain B, where the OG(Ser88) side chain points inwards, having hydrogen bonds with O(Asn85) and O(Ala390). (ii) In the Cβ4–Cβ5 loop, the cis-peptide bond of Gly388–Gly389 in chain B flips into a trans-peptide bond in chain A and residues Gly386–Leu387–Gly388 rotate outwards in a concerted movement with the movement of the Nβ3–Nα3 loop (Figure 8 and Supplementary Movie S2). (iii) The small helix of Leu256 to Glu261 of the Cβ1–Cα1 loop in chain B (unique for the vertebrate SCP2-thiolase) has adopted, in chain A, an extended unstructured conformation pointing into bulk solvent, being stabilized by crystal contacts. (iv) In chain B, residues Lys125 to His136 are involved in interactions at the dimer interface and residues Asp134 to Tyr144 fold into an α-helix, referred to as the Lα(134–144)-helix. The latter helix is via a short linker region connected to the next helix of the loop-domain, being Lα2(150–165). In chain A, the region, Asn131–Pro150, has rotated as a rigid body towards the ligand-binding site, away from its position at the dimer interface. A major rotation happens at Pro150 (a unique, conserved residue in vertebrate SCP2-thiolases; Supplementary Figure S1) just at the beginning of Lα2. The torsion angle of N(Pro150)-Cα-C-N(Ala151), ψ(Pro150), is 157° in the dimeric form and −32° in the monomeric form. Also, the residues in the linker region between Lα(134–144) and Lα2(150–165) show altered phi/psi angles. The Cα of Gly145 [at the C-terminal region of the Lα(134–144)-helix] moves over a distance of 23 Å between the two conformations. In both chains, the conformation of this region is well defined by its electron density. In chain B, the tip of this region, near Leu146–Ala147, anchored by hydrogen bond interactions to the side chain of Gln154 of Lα2 (Figure 8), points into bulk solvent. In chain A, these two tip residues are buried, having main chain hydrogen bond interactions with Cys218 and Thr220 of the pantetheine-binding loop, just before Nβ5. In addition the side chains of these two tip residues, Leu146 and Ala147, in chain A, make hydrophobic interactions with side chains that shape the pantetheine-binding pocket.

Structural differences between the active sites of the single monomer and the complexed monomer of the SCP2-thiolase dimer

Considering the four catalytic loops, it can be noted that there are no conformational changes for the Cβ2–Cβ3 loop (HDCF loop) and the Cβ3–Cα3 loop (the GHP loop); however, the architecture of the A-chain catalytic site is remarkably different from the B-chain catalytic site due to the concerted conformational changes in the Nβ3–Nα3 and Cβ4–Cβ5 loops (Figure 8 and Supplementary Movie S2). The HDCF loop and the GHP loop form OAH1; therefore, OAH1 is still intact in chain A and a water molecule has been built at this site in some of its structures. However, due the conformational changes of the Nβ3–Nα3 and the Cβ4–Cβ5 loops, OAH2 is disrupted and the nucleophilic cysteine, Cys87, of the Nβ3–Nα3 loop is now pointing away from the catalytic site, towards bulk solvent. The structure of the active site of the monomeric thiolase, chain A, as captured in this crystal form, is not in a catalytically competent conformation, because (i) the pantetheine-binding pocket interacts with the Leu146–Ala147 region just after the Lα(134–144)-helix, blocking therefore the binding of the CoA moiety of the substrate. In addition, (ii) the side chain of the nucleophilic cysteine points away from the catalytic site and (iii) OAH2, shaped by Nβ3–Nα3 and Cβ4–Cβ5, is not formed.

Discussion

Interconversion between the two states of Dr-SCP2-thiolase-1

The characterization of Dr-SCP2-thiolase-1 has shown that it forms a weak transient dimer, having a Kd of ∼38 µM and therefore it exists in two states, being the dimeric and monomeric state. In two other dimeric thiolases, two different states have also been characterized. It concerns thiolases of different subfamilies, being the plant peroxisomal AB-thiolase [51,52] and the bacterial FadA5-thiolase of M. tuberculosis [53]. In both examples, the two states are dimers, being in a catalytically competent state and an inactive state. In both inactive states, the nucleophilic cysteine becomes involved in an SS-bridge. In the plant peroxisomal thiolase, the SS-bridge is with another cysteine located in the loop-domain of the same subunit. This SS-bridge formation is the result of a conformational change in the beginning of the loop-domain yielding a wider cleft between the two thiolase monomers. In the oxidized FadA5-thiolase, the conformational changes of the catalytic loops allow for the formation of disulfide bonds between the nucleophilic Cys93 and the acid/base Cys377 as well as between Cys59 and Cys91 [53] within the same subunit. Such an SS-bridge formation between the two catalytic cysteines was recently also reported for a tetrameric thiolase [60]. In these three examples, the dimeric, respectively the tetrameric, assembly remains intact, and it was shown in these cases that the equilibrium between the oxidized and reduced forms is regulated by the redox potential of the protein buffer. In the oxidized forms, intra chain SS-bridges are formed, which need to be reduced to achieve conversion into the competent, catalytically active dimer or tetramer, respectively. In the SCP2-thiolase (type-1), the conversion between the dimer and monomer does not involve the breaking/formation of SS-bridges. It is not known if the latter interconversion between the two forms is a fast or a slow process. In any case, it requires the interconversion between the cis-peptide Gly388–Gly389 (in chain B) and the corresponding trans-peptide (in chain A).

Structural comparisons with other dimeric thiolases

The dimeric SCP2-thiolases have been divided into two classes, type-1 (vertebrates) and type-2 (parasitic) [16]. The vertebrate Dr-SCP2-thiolase-1 crystal structure shows that its catalytic site is bulk solvent exposed (Figure 3), whereas in the parasitic Lm-SCP2-thiolase-2, the active site is shielded from bulk solvent by a helical region at the beginning of the loop-domain. Another difference between the vertebrate and parasitic SCP2-thiolases concerns the disordered region at the beginning of the loop-domain (just after Nβ4) of Dr-SCP2-thiolase-1 (Figure 6), which is well ordered in other thiolases. The sequence identity between Dr-SCP2-thiolase-1 and the parasitic T. brucei and L. mexicana SCP2-thiolases (type-2) is 22% and 23%, respectively. The latter thiolase is proposed to catalyze the Claisen condensation reaction for the synthesis of acetoacetyl-CoA from two molecules of acetyl-CoA in the sterol biosynthesis pathway in parasites [19], whereas the substrate of the SCP2-thiolase (type-1) is the bulky, polar 24-keto-THC-CoA (Figure 1) [2]. The SCP2-thiolase (type-2) is predicted to accept only short-chain substrates [17], which correlates well with its tight, buried catalytic site.

Crystal structures are also known of dimeric thiolases of two other subfamilies, being of the peroxisomal AB-thiolases (such as the yeast thiolase [61]) and of the bacterial TFEL1 FadA5-thiolase [21]. The sequence identity between Dr-SCP2-thiolase-1 and the yeast and bacterial dimeric thiolases is 13% and 13%, respectively. The substrates for both of these thiolases have bulky acyl moieties, being long-chain linear fatty acids (AB-thiolase) and acyl chains with a steroid moiety at its ω-end (the FadA5-thiolase). As seen in the Dr-SCP2-thiolase-1 structure, also in these structures, the catalytic site is bulk solvent exposed. These structural comparisons highlight the structural variability of the N-terminal region of the loop-domain, immediately after the Nβ4 loop and before the Lα2-helix (Figure 9). This loop region is near the catalytic site (Figure 6), and it determines the shape of the substrate-specificity-binding pocket for the acyl chain. The structure of the subsequent Lα2–Lα3 region of the loop-domain is well conserved. However, the subsequent regions of the loop-domain that form the binding pocket of the CoA moiety (from Lα3 till the pantetheine loop, just before Nβ5) also adopt different structures in different thiolases, with the consequence that the mode of binding of the ADP-part of the CoA moiety is not conserved (Figure 9). In the Dr-SCP2-thiolase-1 structure, the adenine ring is almost totally buried, whereas in the Lm-SCP2-thiolase-2 and FadA5-thiolase structures, the adenine ring is much more bulk solvent exposed.

Comparison of the loop-domain structures.

Figure 9.
Comparison of the loop-domain structures.

Top view, like Figure 3. Superposition of the loop-domain of zebrafish SCP2-thiolase (type-1) (orange) complexed with octanoyl-CoA and (A) leishmania SCP2-thiolase (type-2) (gray), complexed with acetoacetyl-CoA (5LNQ), and (B) Mycobacterium tuberculosis FadA5-thiolase (cyan), complexed with CoA and a steroid product (3G6)(4UBT) and (C) yeast peroxisomal AB-thiolase (blue), complexed with 2-methylpentane-2,4-diol (MPD) (1AFW). The ligands are shown as sticks. Met117 (Nβ4) and Thr220 (Nβ5) identify the beginning and end of the loop-domain, respectively. Residues Ala152 and Thr186 identify the beginning of Lα2 and the C-terminal region of Lα3, respectively. It can be noted that the structures of the Lα2 and Lα3 helices are well conserved. Asp134 is at the beginning of the Lα(134–144)-helix.

Figure 9.
Comparison of the loop-domain structures.

Top view, like Figure 3. Superposition of the loop-domain of zebrafish SCP2-thiolase (type-1) (orange) complexed with octanoyl-CoA and (A) leishmania SCP2-thiolase (type-2) (gray), complexed with acetoacetyl-CoA (5LNQ), and (B) Mycobacterium tuberculosis FadA5-thiolase (cyan), complexed with CoA and a steroid product (3G6)(4UBT) and (C) yeast peroxisomal AB-thiolase (blue), complexed with 2-methylpentane-2,4-diol (MPD) (1AFW). The ligands are shown as sticks. Met117 (Nβ4) and Thr220 (Nβ5) identify the beginning and end of the loop-domain, respectively. Residues Ala152 and Thr186 identify the beginning of Lα2 and the C-terminal region of Lα3, respectively. It can be noted that the structures of the Lα2 and Lα3 helices are well conserved. Asp134 is at the beginning of the Lα(134–144)-helix.

Properties of the thiolase protein–protein dimer interfaces

The importance of transient protein–protein interactions for the function [58,62] and consequently for the evolution [63,64] of proteins has been reviewed recently. Transient protein–protein interactions can be divided into strong and weak interactions [58]. Weak transient interactions have been suggested to play a role in biochemical pathways, for example, for facilitating the transfer of unstable intermediates between active sites [65], as well as to provide a mechanism to regulate the enzymatic function such as (i) the formation of inactive oligomers from active monomers [66] or (ii) the formation of active dimers from inactive monomers [67,68]. Several properties of protein–protein interfaces that contribute to their stability, such as size, polarity and shape complementarity, have been studied [63,6971].

In the thiolase dimer, the center of the dimer interface is formed by the edge Nβ3-strands (Figures 6 and 8) of both subunits, forming two antiparallel β-strands of a mixed β-sheet that extends from one subunit to the second subunit. The length of the Nβ3-strand is shorter in Dr-SCP2-thiolase-1, when compared with any other dimeric thiolase (Table 3). Various properties that characterize the dimer interface have been calculated for the available dimeric thiolase structures with the PISA program [69]. The size of the surface area of the Dr-SCP2-thiolase-1 dimer interface is similar as in any of the other dimeric thiolases, being 2555 Å2. The surface area is more polar than the dimer interface of any other thiolases, as is quantified by the number of interface hydrogen bonds (70, whereas in the other thiolases this number is 52 or less) as well as by its P-value of the solvation-free energy gain (ΔGi), which is 0.27. For the other dimer interfaces, the value is less than 0.18. The P-value is calculated by PISA, and it is a measure of the relative hydrophobicity for the given interface (compared with surfaces in contact with bulk solvent), being lower for more hydrophobic surfaces, and indeed, the calculated solvation-free energy of the SCP2-thiolase interface, ΔGi, (−15.3 kcal/mol), is less favorable than for any of the other interfaces, being −17.4 kcal/mol or lower (Table 3). With the SURFNET program [72] the volumes of the gap region at the dimer interface, as well as its ‘roughness’, were calculated (Table 3), showing that these two properties are within the same range as calculated for the other thiolase dimer interfaces.

Table 3
Properties1 of the dimer interface of dimeric thiolases
PDB code2 Nres Surface area (Å2ΔiG (kcal/mol) ΔiG P-value NHB NSB Gap-volume (Å3Gap- index (Å) Planarity (Å) Percentage of side chain (SC) and main chain (MC)-mediated hydrogen bonds NHB /100 Å2 
SC–SC MC–MC SC–MC 
6HRV 63/62 2555 −15.3 0.27 70 1793 0.7 4.4 28.6 11.4 60.0 2.7 
3ZBG 70/70 2325 −18.7 0.17 45 1915 0.8 4.2 24.4 26.7 48.9 1.9 
1AFW 67/70 2388 −25.5 0.07 39 2412 1.0 5.1 35.9 25.6 38.5 1.6 
2WUA 74/72 2586 −26.2 0.15 52 18 1426 0.6 4.9 38.5 23.1 38.5 2.0 
2WU9 69/68 2508 −24.3 0.18 48 15 1646 0.7 4.6 37.5 25.0 39.6 1.9 
4UBT3 72/70 2330 −17.4 0.14 28 2688 1.2 4.2 28.6 28.6 42.9 1.2 
PDB code2 Nres Surface area (Å2ΔiG (kcal/mol) ΔiG P-value NHB NSB Gap-volume (Å3Gap- index (Å) Planarity (Å) Percentage of side chain (SC) and main chain (MC)-mediated hydrogen bonds NHB /100 Å2 
SC–SC MC–MC SC–MC 
6HRV 63/62 2555 −15.3 0.27 70 1793 0.7 4.4 28.6 11.4 60.0 2.7 
3ZBG 70/70 2325 −18.7 0.17 45 1915 0.8 4.2 24.4 26.7 48.9 1.9 
1AFW 67/70 2388 −25.5 0.07 39 2412 1.0 5.1 35.9 25.6 38.5 1.6 
2WUA 74/72 2586 −26.2 0.15 52 18 1426 0.6 4.9 38.5 23.1 38.5 2.0 
2WU9 69/68 2508 −24.3 0.18 48 15 1646 0.7 4.6 37.5 25.0 39.6 1.9 
4UBT3 72/70 2330 −17.4 0.14 28 2688 1.2 4.2 28.6 28.6 42.9 1.2 
1

These properties have been calculated by PISA [69], except for the Gap-volume, the Gap-index and the Planarity, which have been calculated by SURFNET [72,79]. Gap-index is the Gap-volume divided by the surface area. Planarity has been calculated as the root mean square deviation (rmsd) of the interface atoms from the least squares plane fitted through these atoms; a high rmsd identifies a more ‘rough’ interface. Abbreviations: Nres: number of residues of the two chains involved in dimer interactions, respectively; NHB: the number of hydrogen bonds; NSB: the number of hydrogen bonds involved in salt bridges.

2

6HRV, Dr-SCP2-thiolase-1 (these studies: the B/B dimer interface has been used). 3ZBG, Lm-SCP2-thiolase-2; 1AFW, yeast Saccharomyces cerevisiae AB-thiolase; 2WUA, sunflower Helianthus annuus AB-thiolase; 2WU9, plant A. thaliana 3-ketoacyl-CoA AB-thiolase; 4UBT, Mycobacterium tuberculosis TFEL1 FadA5-thiolase: in each of these analyses, the A/B dimer interface has been used.

3

Expression tag residues from chain A and chain B were removed before doing the structural analysis.

The number of interface salt bridge hydrogen bond interactions of the thiolase dimer interface is highly variable (Table 3). In Dr-SCP2-thiolase-1, there are two salt bridges, between the side chains of Arg67 (in Nα2) and Asp134 [located in the mobile Lα(134–144)-helix] and between their corresponding two-fold related residues. The occurrence of hydrophobic, polar and charge atoms at interfaces has been studied using the YRB-coloring scheme [73]. In Dr-SCP2-thiolase-1, the salt bridges occur at the rim of the dimer interface (Figure 10A), whereas in Lm-SCP2-thiolase-2 a salt bridge is observed in the middle of the dimer interface (Figure 10B). Salt bridges are also found at the center of the dimer interface of the yeast AB-thiolase (Figure 10C), whereas in the FadA5-thiolase (Figure 10D) a charged residue (Asp90) interacts with a nearby main chain peptide NH-group of the adjacent subunit as well as with water molecules buried at the dimer interface.

Properties of the dimer interface surface.

Figure 10.
Properties of the dimer interface surface.

Viewed from the center of the other subunit, like in Figure 8B. The YRB-coloring code [73] is used, such that the hydrophobic atoms are in yellow, the polar, uncharged atoms are in white, whereas the charged oxygen atoms of glutamate and aspartate are highlighted in red and the charged nitrogen atoms of arginine and lysine are highlighted in blue. (A) The YRB-colored surface of the dimer interface of the B-chain of Dr-SCP2-thiolase-1. (B) The YRB-colored surface of the dimer interface of Lm-SCP2-thiolase-2 (3ZBG). (C) The YRB-colored surface of the dimer interface of the yeast peroxisomal thiolase (1AFW). (D) The YRB-colored surface of the dimer interface of the Mycobacterium tuberculosis FadA5-thiolase (4UBT).

Figure 10.
Properties of the dimer interface surface.

Viewed from the center of the other subunit, like in Figure 8B. The YRB-coloring code [73] is used, such that the hydrophobic atoms are in yellow, the polar, uncharged atoms are in white, whereas the charged oxygen atoms of glutamate and aspartate are highlighted in red and the charged nitrogen atoms of arginine and lysine are highlighted in blue. (A) The YRB-colored surface of the dimer interface of the B-chain of Dr-SCP2-thiolase-1. (B) The YRB-colored surface of the dimer interface of Lm-SCP2-thiolase-2 (3ZBG). (C) The YRB-colored surface of the dimer interface of the yeast peroxisomal thiolase (1AFW). (D) The YRB-colored surface of the dimer interface of the Mycobacterium tuberculosis FadA5-thiolase (4UBT).

Concluding remarks

In this crystal form of Dr-SCP2-thiolase-1, the structures of its monomeric and dimeric forms have been captured. These structures show that the assembly of the inactive monomers into the catalytically competent, mature dimer causes striking structural rearrangements. Native mass spectrometry studies with Hs-SCP2-thiolase-1 and Dr-SCP2-thiolase-1 have shown that in solution, these thiolases are weak transient dimers. The cell biological studies on human SCPx have shown that the truncated form that has been studied here is also predominantly present in vivo in the peroxisome [11]. The structural comparisons of the monomeric and dimeric forms show that four loop regions at the dimer interface, which are well defined in the dimeric form, adopt completely different conformations in the monomeric form. Each of these loops is near the active site. One of those regions, the Nβ3–Nα3 loop, becomes a high B-factor loop in the structure of the monomeric form. The mass spectrometry solution-binding studies with CoA show that CoA binds to the monomeric form, suggesting that the conformation of the region of the swapped-out Lα(134–144)-helix (Figure 8 and Supplementary Movie S2), as captured in the chain A structure, is flexible in solution, allowing for the binding of CoA. In any case, crystal contacts stabilize the observed, swapped-out conformations of the Lα(134–144)-helix and the Cβ1–Cα1 loop of chain A. Two of the four loop regions of changed conformation (Nβ3–Nα3 and Cβ4–Cβ5) generate in the dimeric form the OAH2 oxyanion hole, which is critically important for catalysis. Clearly, dimerization is required to shape a catalytically competent active site. In two other dimeric thiolases, it has been found that the formation of the catalytically competent dimeric form is regulated by the redox potential of the protein buffer [51,53], and it is suggested that this property is of functional importance in vivo. For these thiolases, only the reduced form is competent for catalysis. For the SCP2-thiolase (type-1), it is not known if in vivo the dimer (active) or monomer (inactive) form is predominantly present and how its monomer–dimer equilibrium could be regulated. Since the active site is at the dimer interface, it is possible that the binding of the 24-keto-THC-CoA substrate, having a bulky acyl chain, shifts this equilibrium towards the active dimeric form. It is known that the nucleophilic cysteine in the thiolase competent active site is sensitive to oxidation, being frequently observed in thiolase crystal structures in its sulfenic acid oxidation state. Since the peroxisomal milieu can be highly oxidative [74], the nucleophilic cysteine can become oxidized in its dimeric form, whereas in its monomeric form this cysteine is much less reactive and therefore less sensitive for oxidation. Further studies are required to establish if the intraperoxisomal in vivo conditions also favor the presence of a similar dimer–monomer equilibrium as observed in these studies and how this equilibrium is regulated.

Abbreviations

     
  • DLS

    Diamond Light Source, Oxfordshire, U.K.

  •  
  • Dr-SCP2-thiolase-1

    Danio rerio (zebrafish) SCP2-thiolase (type-1) (only the thiolase part)

  •  
  • ESI

    electrospray ionization

  •  
  • FadA5-thiolase

    Mycobacterium tuberculosis TFEL1-thiolase encoded by the fadA5 gene

  •  
  • FT-ICR

    Fourier transform ion cyclotron resonance

  •  
  • HEPES

    4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

  •  
  • Hs-SCP2-thiolase-1

    Homo sapiens SCP2-thiolase (type-1) (only the thiolase part)

  •  
  • Lm-SCP2-thiolase-2

    Leishmania mexicana SCP2-thiolase (type-2)

  •  
  • MFE1

    multifunctional enzyme, type-1

  •  
  • MOPS

    3-(N-morpholino)-propanesulfonic acid

  •  
  • MS

    mass spectrometry

  •  
  • Ni-NTA

    Ni2+-nitrilotriacetate

  •  
  • OAH

    oxyanion hole

  •  
  • PEG

    polyethylene glycol

  •  
  • SCP2-protein

    sterol carrier protein 2 (only the SCP2 part of SCPx)

  •  
  • SCP2-thiolase (type-1)

    vertebrate SCP2-thiolase (only the thiolase part of SCPx)

  •  
  • SCP2-thiolase (type-2)

    leishmanial and trypanosomal SCP2-thiolase

  •  
  • SCPx

    vertebrate protein with the SCP2-thiolase (N-terminal part), being extended at its C-terminus by a 20-residue linker and subsequently its SCP2-domain

  •  
  • Tb-SCP2-thiolase-2

    Trypanosoma brucei brucei SCP2-thiolase (type-2)

  •  
  • THC-CoA

    25S-3α,7α,12α-trihydroxy-5β-cholestanoyl-CoA

Author Contribution

T.-R.K. and R.K.W. conceived the experiments. T.-R.K. carried out all structural studies. T.-R.K. and C.J.T. performed the expression, purification and enzyme kinetic studies. M.L., M.M.M. and J.J. carried out all the mass spectrometry experiments. T.F. provided the original human SCP2-thiolase (type-1) construct and participated in the discussions on the results. W.S. synthesized the 2E-decenoyl-CoA. T.-R.K. and R.K.W. performed the structural analysis. T.-R.K., R.K.W., J.R. and J.J. participated in the discussion of results and wrote the manuscript.

Funding

The FT-ICR MS facility of the University of Eastern Finland is also supported by the European Regional Development Fund (grant no. A70135) and EU's Horizon 2020 Research and Innovation Program (grant no. 731077).

Acknowledgements

We thank V.V. Suresh Babu for his contributions at the initial stage of these studies. We are grateful for the expert support at the DLS beam lines I03, I04-1 and I04 and to Andrey Lebedev for his suggestions concerning the molecular replacement calculations. The use of the instruments and expertise of the Biocenter Oulu proteomics and protein analysis, DNA sequencing and protein X-ray crystallography core facilities (which are part of Biocenter Finland and Instruct-FI) is gratefully acknowledged. The FT-ICR MS facility of the University of Eastern Finland also acknowledges the support of the Biocenter Kuopio and the Instruct-FI structural biology network of Biocenter Finland. The authors acknowledge the support and the use of resources of Instruct-ERIC. We thank Ville Ratas for providing excellent help with the crystallization experiments.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Russell
,
D.W.
and
Setchell
,
K.D.
(
1992
)
Bile acid biosynthesis
.
Biochemistry
31
,
4737
4749
2
Antonenkov
,
V.D.
,
Van Veldhoven
,
P.P.
,
Waelkens
,
E.
and
Mannaerts
,
G.P.
(
1997
)
Substrate specificities of 3-oxoacyl-CoA thiolase A and sterol carrier protein 2/3-oxoacyl-CoA thiolase purified from normal rat liver peroxisomes. Sterol carrier protein 2/3-oxoacyl-CoA thiolase is involved in the metabolism of 2-methyl-branched fatty acids and bile acid intermediates
.
J. Biol. Chem.
272
,
26023
26031
3
Seedorf
,
U.
,
Brysch
,
P.
,
Engel
,
T.
,
Schrage
,
K.
and
Assmann
,
G.
(
1994
)
Sterol carrier protein X is peroxisomal 3-oxoacyl coenzyme A thiolase with intrinsic sterol carrier and lipid transfer activity
.
J. Biol. Chem.
269
,
21277
21283
PMID:
[PubMed]
4
Mathieu
,
M.
,
Modis
,
Y.
,
Zeelen
,
J.P.
,
Engel
,
C.K.
,
Abagyan
,
R.A.
,
Ahlberg
,
A.
et al.  (
1997
)
The 1.8 Å crystal structure of the dimeric peroxisomal 3-ketoacyl-CoA thiolase of Saccharomyces cerevisiae: implications for substrate binding and reaction mechanism
.
J. Mol. Biol.
273
,
714
728
5
Modis
,
Y.
and
Wierenga
,
R.K.
(
1999
)
A biosynthetic thiolase in complex with a reaction intermediate: the crystal structure provides new insights into the catalytic mechanism
.
Structure
7
,
1279
1290
PMID:
[PubMed]
6
Ossendorp
,
B.C.
,
Van Heusden
,
G.P.
,
De Beer
,
A.L.
,
Bos
,
K.
,
Schouten
,
G.L.
and
Wirtz
,
K.W.
(
1991
)
Identification of the cDNA clone which encodes the 58-kDa protein containing the amino acid sequence of rat liver non-specific lipid-transfer protein (sterol-carrier protein 2). Homology with rat peroxisomal and mitochondrial 3-oxoacyl-CoA thiolases
.
Eur. J. Biochem.
201
,
233
239
7
Stolowich
,
N.J.
,
Frolov
,
A.
,
Atshaves
,
B.
,
Murphy
,
E.J.
,
Jolly
,
C.A.
,
Billheimer
,
J.T.
et al.  (
1997
)
The sterol carrier protein-2 fatty acid binding site: an NMR, circular dichroic, and fluorescence spectroscopic determination
.
Biochemistry
36
,
1719
1729
8
Stolowich
,
N.J.
,
Petrescu
,
A.D.
,
Huang
,
H.
,
Martin
,
G.G.
,
Scott
,
A.I.
and
Schroeder
,
F.
(
2002
)
Sterol carrier protein-2: structure reveals function
.
Cell. Mol. Life Sci.
59
,
193
212
9
Ohba
,
T.
,
Holt
,
J.A.
,
Billheimer
,
J.T.
and
Strauss
, III,
J.F.
(
1995
)
Human sterol carrier protein x/sterol carrier protein 2 gene has two promoters
.
Biochemistry
34
,
10660
10668
10
Antonenkov
,
V.D.
,
Van Veldhoven
,
P.P.
and
Mannaerts
,
G.P.
(
2000
)
Isolation and subunit composition of native sterol carrier protein 2/3-oxoacyl-coenzyme A thiolase from normal rat liver peroxisomes
.
Protein Expr. Purif.
18
,
249
256
11
Ferdinandusse
,
S.
,
Kostopoulos
,
P.
,
Denis
,
S.
,
Rusch
,
H.
,
Overmars
,
H.
,
Dillmann
,
U.
et al.  (
2006
)
Mutations in the gene encoding peroxisomal sterol carrier protein X (SCPx) cause leukencephalopathy with dystonia and motor neuropathy
.
Am. J. Hum. Genet.
78
,
1046
1052
12
Gould
,
S.J.
,
Keller
,
G.A.
and
Subramani
,
S.
(
1988
)
Identification of peroxisomal targeting signals located at the carboxy terminus of four peroxisomal proteins
.
J. Cell Biol.
107
,
897
905
13
Seedorf
,
U.
,
Raabe
,
M.
,
Ellinghaus
,
P.
,
Kannenberg
,
F.
,
Fobker
,
M.
,
Engel
,
T.
et al.  (
1998
)
Defective peroxisomal catabolism of branched fatty acyl coenzyme A in mice lacking the sterol carrier protein-2/sterol carrier protein-x gene function
.
Genes Dev.
12
,
1189
1201
14
Ferdinandusse
,
S.
,
Denis
,
S.
,
Van Roermund
,
C.W.
,
Wanders
,
R.J.
and
Dacremont
,
G.
(
2004
)
Identification of the peroxisomal beta-oxidation enzymes involved in the degradation of long-chain dicarboxylic acids
.
J. Lipid Res.
45
,
1104
1111
15
Kannenberg
,
F.
,
Ellinghaus
,
P.
,
Assmann
,
G.
and
Seedorf
,
U.
(
1999
)
Aberrant oxidation of the cholesterol side chain in bile acid synthesis of sterol carrier protein-2/sterol carrier protein-x knockout mice
.
J. Biol. Chem.
274
,
35455
35460
16
Anbazhagan
,
P.
,
Harijan
,
R.K.
,
Kiema
,
T.R.
,
Janardan
,
N.
,
Murthy
,
M.R.
,
Michels
,
P.A.
et al.  (
2014
)
Phylogenetic relationships and classification of thiolases and thiolase-like proteins of Mycobacterium tuberculosis and Mycobacterium smegmatis
.
Tuberculosis
94
,
405
412
17
Harijan
,
R.K.
,
Kiema
,
T.R.
,
Syed
,
S.M.
,
Qadir
,
I.
,
Mazet
,
M.
,
Bringaud
,
F.
et al.  (
2017
)
Crystallographic substrate binding studies of Leishmania mexicana SCP2-thiolase (type-2): unique features of oxyanion hole-1
.
Protein Eng. Des. Sel.
30
,
225
233
18
Harijan
,
R.K.
,
Kiema
,
T.R.
,
Karjalainen
,
M.P.
,
Janardan
,
N.
,
Murthy
,
M.R.
,
Weiss
,
M.S.
et al.  (
2013
)
Crystal structures of SCP2-thiolases of trypanosomatidae, human pathogens causing widespread tropical diseases: the importance for catalysis of the cysteine of the unique HDCF loop
.
Biochem. J.
455
,
119
130
19
Millerioux
,
Y.
,
Mazet
,
M.
,
Bouyssou
,
G.
,
Allmann
,
S.
,
Kiema
,
T.
,
Bertiaux
,
E.
et al.  (
2018
)
De novo biosynthesis of sterols and fatty acids in the Trypanosoma brucei procyclic form: carbon source preferences and metabolic flux redistributions
.
PLoS Pathog.
14
,
e1007116
20
Nesbitt
,
N.M.
,
Yang
,
X.
,
Fontan
,
P.
,
Kolesnikova
,
I.
,
Smith
,
I.
,
Sampson
,
N.S.
et al.  (
2010
)
A thiolase of Mycobacterium tuberculosis is required for virulence and production of androstenedione and androstadienedione from cholesterol
.
Infect. Immun.
78
,
275
282
21
Schaefer
,
C.M.
,
Lu
,
R.
,
Nesbitt
,
N.M.
,
Schiebel
,
J.
,
Sampson
,
N.S.
and
Kisker
,
C.
(
2015
)
Fada5 a thiolase from Mycobacterium tuberculosis: a steroid-binding pocket reveals the potential for drug development against tuberculosis
.
Structure
23
,
21
33
22
Kursula
,
P.
,
Ojala
,
J.
,
Lambeir
,
A.M.
and
Wierenga
,
R.K.
(
2002
)
The catalytic cycle of biosynthetic thiolase: a conformational journey of an acetyl group through four binding modes and two oxyanion holes
.
Biochemistry
41
,
15543
15556
PMID:
[PubMed]
23
Merilainen
,
G.
,
Poikela
,
V.
,
Kursula
,
P.
and
Wierenga
,
R.K.
(
2009
)
The thiolase reaction mechanism: the importance of Asn316 and His348 for stabilizing the enolate intermediate of the claisen condensation
.
Biochemistry
48
,
11011
11025
24
Thompson
,
S.
,
Mayerl
,
F.
,
Peoples
,
O.P.
,
Masamune
,
S.
,
Sinskey
,
A.J.
and
Walsh
,
C.T.
(
1989
)
Mechanistic studies on beta-ketoacyl thiolase from Zoogloea ramigera: identification of the active-site nucleophile as Cys89, its mutation to Ser89, and kinetic and thermodynamic characterization of wild-type and mutant enzymes
.
Biochemistry
28
,
5735
5742
25
Williams
,
S.F.
,
Palmer
,
M.A.
,
Peoples
,
O.P.
,
Walsh
,
C.T.
,
Sinskey
,
A.J.
and
Masamune
,
S.
(
1992
)
Biosynthetic thiolase from Zoogloea ramigera. Mutagenesis of the putative active-site base cys-378 to ser-378 changes the partitioning of the acetyl S-enzyme intermediate
.
J. Biol. Chem.
267
,
16041
16043
PMID:
[PubMed]
26
Modis
,
Y.
and
Wierenga
,
R.K.
(
2000
)
Crystallographic analysis of the reaction pathway of Zoogloea ramigera biosynthetic thiolase
.
J. Mol. Biol.
297
,
1171
1182
27
Haapalainen
,
A.M.
,
Meriläinen
,
G.
and
Wierenga
,
R.K.
(
2006
)
The thiolase superfamily: condensing enzymes with diverse reaction specificities
.
Trends Biochem. Sci.
31
,
64
71
28
He
,
Z.
,
Yamamoto
,
R.
,
Furth
,
E.E.
,
Schantz
,
L.J.
,
Naylor
,
S.L.
,
George
,
H.
et al.  (
1991
)
cDNAs encoding members of a family of proteins related to human sterol carrier protein 2 and assignment of the gene to human chromosome 1 p21—pter
.
DNA Cell Biol.
10
,
559
569
29
Gasteiger
,
E.
,
Hoogland
,
C.
,
Gattiker
,
A.
,
Duvaud
,
S.
,
Wilkins
,
M.R.
,
Appel
,
R.D.
et al.  (
2005
) Protein identification and analysis tools on the ExPASy server. In
The Proteomics Protocols Handbook
(
Walker
,
J. M.
, ed.), pp.
571
607
,
Humana Press
,
Totowa, NJ
30
Artimo
,
P.
,
Jonnalagedda
,
M.
,
Arnold
,
K.
,
Baratin
,
D.
,
Csardi
,
G.
,
de Castro
,
E.
et al.  (
2012
)
ExPASy: SIB bioinformatics resource portal
.
Nucleic Acids Res.
40
,
597
31
Antonenkov
,
V.D.
,
Van Veldhoven
,
P.P.
and
Mannaerts
,
G.P.
(
1999
)
The commonly used Mg2+-enolate assay can lead to underestimation of thiolase activity
.
Anal. Biochem.
267
,
418
420
32
Mazet
,
M.
,
Harijan
,
R.K.
,
Kiema
,
T.
,
Haapalainen
,
A.M.
,
Morand
,
P.
,
Morales
,
J.
et al.  (
2011
)
The characterization and evolutionary relationships of a trypanosomal thiolase
.
Int. J. Parasitol.
41
,
1273
1283
33
Zeelen
,
J.P.
,
Hiltunen
,
J.K.
,
Ceska
,
T.A.
and
Wierenga
,
R.K.
(
1994
)
Crystallization experiments with 2-enoyl-CoA hydratase, using an automated ‘fast-screening’ crystallization protocol
.
Acta Crystallogr. D Biol. Crystallogr.
50
,
443
447
34
Kasaragod
,
P.
,
Venkatesan
,
R.
,
Kiema
,
T.R.
,
Hiltunen
,
J.K.
and
Wierenga
,
R.K.
(
2010
)
Crystal structure of liganded rat peroxisomal multifunctional enzyme type 1: a flexible molecule with two interconnected active sites
.
J. Biol. Chem.
285
,
24089
24098
35
Matthews
,
B.W.
(
1968
)
Solvent content of protein crystals
.
J. Mol. Biol.
33
,
491
497
36
Kantardjieff
,
K.A.
and
Rupp
,
B.
(
2003
)
Matthews coefficient probabilities: improved estimates for unit cell contents of proteins, DNA, and protein-nucleic acid complex crystals
.
Protein Sci.
12
,
1865
1871
37
Winn
,
M.D.
,
Ballard
,
C.C.
,
Cowtan
,
K.D.
,
Dodson
,
E.J.
,
Emsley
,
P.
,
Evans
,
P.R.
et al.  (
2011
)
Overview of the CCP4 suite and current developments
.
Acta Crystallogr. D Biol. Crystallogr.
67
,
235
242
38
Kabsch
,
W.
(
2010
)
Xds
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
125
132
39
Sparta
,
K.M.
,
Krug
,
M.
,
Heinemann
,
U.
,
Mueller
,
U.
and
Weiss
,
M.S.
(
2016
)
Xdsapp2.0
.
J. Appl. Crystallogr.
49
,
1085
1092
40
McCoy
,
A.J.
,
Grosse-Kunstleve
,
R.W.
,
Adams
,
P.D.
,
Winn
,
M.D.
,
Storoni
,
L.C.
and
Read
,
R.J.
(
2007
)
Phaser crystallographic software
.
J. Appl. Crystallogr.
40
,
658
674
41
Stein
,
N.
(
2008
)
CHAINSAW: a program for mutating pdb files used as templates in molecular replacement
.
J. Appl. Crystallogr.
41
,
641
643
42
Murshudov
,
G.N.
,
Skubak
,
P.
,
Lebedev
,
A.A.
,
Pannu
,
N.S.
,
Steiner
,
R.A.
,
Nicholls
,
R.A.
et al.  (
2011
)
REFMAC5 for the refinement of macromolecular crystal structures
.
Acta Crystallogr. D Biol. Crystallogr.
67
,
355
367
43
Sheldrick
,
G.M.
(
2010
)
Experimental phasing with SHELXC/D/E: combining chain tracing with density modification
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
479
485
44
Cowtan
,
K.
(
2006
)
The Buccaneer software for automated model building. 1. Tracing protein chains
.
Acta Crystallogr. D Biol. Crystallogr.
62
,
1002
1011
45
Emsley
,
P.
,
Lohkamp
,
B.
,
Scott
,
W.G.
and
Cowtan
,
K.
(
2010
)
Features and development of coot
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
486
501
46
Laskowski
,
R.A.
,
MacArthur
,
M.W.
,
Moss
,
D.S.
and
Thornton
,
J.M.
(
1993
)
PROCHECK: a program to check the stereochemical quality of protein structures
.
J. Appl. Crystallogr.
26
,
283
291
47
Davis
,
I.W.
,
Leaver-Fay
,
A.
,
Chen
,
V.B.
,
Block
,
J.N.
,
Kapral
,
G.J.
,
Wang
,
X.
et al.  (
2007
)
Molprobity: all-atom contacts and structure validation for proteins and nucleic acids
.
Nucleic Acids Res.
35
,
375
383
48
Adams
,
P.D.
,
Afonine
,
P.V.
,
Bunkoczi
,
G.
,
Chen
,
V.B.
,
Davis
,
I.W.
,
Echols
,
N.
et al.  (
2010
)
PHENIX: a comprehensive python-based system for macromolecular structure solution
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
213
221
49
Kleywegt
,
G.J.
(
2007
)
Crystallographic refinement of ligand complexes
.
Acta Crystallogr. D Biol. Crystallogr.
63
,
94
100
50
Schuttelkopf
,
A.W.
and
van Aalten
,
D.M.
(
2004
)
PRODRG: a tool for high-throughput crystallography of protein-ligand complexes
.
Acta Crystallogr. D Biol. Crystallogr.
60
,
1355
1363
51
Pye
,
V.E.
,
Christensen
,
C.E.
,
Dyer
,
J.H.
,
Arent
,
S.
and
Henriksen
,
A.
(
2010
)
Peroxisomal plant 3-ketoacyl-CoA thiolase structure and activity are regulated by a sensitive redox switch
.
J. Biol. Chem.
285
,
24078
24088
52
Sundaramoorthy
,
R.
,
Micossi
,
E.
,
Alphey
,
M.S.
,
Germain
,
V.
,
Bryce
,
J.H.
,
Smith
,
S.M.
et al.  (
2006
)
The crystal structure of a plant 3-ketoacyl-CoA thiolase reveals the potential for redox control of peroxisomal fatty acid beta-oxidation
.
J. Mol. Biol.
359
,
347
357
53
Lu
,
R.
,
Schaefer
,
C.M.
,
Nesbitt
,
N.M.
,
Kuper
,
J.
,
Kisker
,
C.
and
Sampson
,
N.S.
(
2017
)
Catabolism of the cholesterol side chain in Mycobacterium tuberculosis is controlled by a redox-sensitive thiol switch
.
ACS Infect. Dis.
3
,
666
675
54
Krissinel
,
E.
and
Henrick
,
K.
(
2004
)
Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions
.
Acta Crystallogr. D Biol. Crystallogr.
60
,
2256
2268
55
Sievers
,
F.
,
Wilm
,
A.
,
Dineen
,
D.
,
Gibson
,
T.J.
,
Karplus
,
K.
,
Li
,
W.
et al.  (
2011
)
Fast, scalable generation of high-quality protein multiple sequence alignments using clustal omega
.
Mol. Syst. Biol.
7
,
539
56
Kabsch
,
W.
and
Sander
,
C.
(
1983
)
Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features
.
Biopolymers
22
,
2577
2637
57
Gouet
,
P.
,
Courcelle
,
E.
,
Stuart
,
D.I.
and
Metoz
,
F.
(
1999
)
ESPript: analysis of multiple sequence alignments in PostScript
.
Bioinformatics
15
,
305
308
PMID:
[PubMed]
58
Perkins
,
J.R.
,
Diboun
,
I.
,
Dessailly
,
B.H.
,
Lees
,
J.G.
and
Orengo
,
C.
(
2010
)
Transient protein-protein interactions: structural, functional, and network properties
.
Structure
18
,
1233
1243
59
Janardan
,
N.
,
Harijan
,
R.K.
,
Kiema
,
T.R.
,
Wierenga
,
R.K.
and
Murthy
,
M.R.N.
(
2015
)
Structural characterization of a mitochondrial 3-ketoacyl-CoA (T1)-like thiolase from Mycobacterium smegmatis
.
Acta Crystallogr. D Biol. Crystallogr.
71
,
2479
2493
60
Kim
,
S.
,
Jang
,
Y.
,
Ha
,
S.
,
Ahn
,
J.
,
Kim
,
E.
,
Lim
,
J.H.
et al.  (
2015
)
Redox-switch regulatory mechanism of thiolase from Clostridium acetobutylicum
.
Nat. Commun.
6
,
8410
61
Mathieu
,
M.
,
Zeelen
,
J.P.
,
Pauptit
,
R.A.
,
Erdmann
,
R.
,
Kunau
,
W.H.
and
Wierenga
,
R.K.
(
1994
)
The 2.8 A crystal structure of peroxisomal 3-ketoacyl-CoA thiolase of Saccharomyces cerevisiae: a five-layered alpha beta alpha beta alpha structure constructed from two core domains of identical topology
.
Structure
2
,
797
808
PMID:
[PubMed]
62
Acuner Ozbabacan
,
S.E.
,
Engin
,
H.B.
,
Gursoy
,
A.
and
and Keskin
,
O.
(
2011
)
Transient protein-protein interactions
.
Protein Eng. Des. Sel.
24
,
635
648
63
Dey
,
S.
,
Ritchie
,
D.W.
and
Levy
,
E.D.
(
2018
)
PDB-wide identification of biological assemblies from conserved quaternary structure geometry
.
Nat. Methods
15
,
67
72
64
Levy
,
E.D.
,
De
,
S.
and
Teichmann
,
S.A.
(
2012
)
Cellular crowding imposes global constraints on the chemistry and evolution of proteomes
.
Proc. Natl Acad. Sci. U.S.A.
109
,
20461
20466
65
Kappock
,
T.J.
,
Ealick
,
S.E.
and
Stubbe
,
J.
(
2000
)
Modular evolution of the purine biosynthetic pathway
.
Curr. Opin. Chem. Biol.
4
,
567
572
66
López-Pelegrín
,
M.
,
Cerdà-Costa
,
N.
,
Cintas-Pedrola
,
A.
,
Herranz-Trillo
,
F.
,
Bernadó
,
P.
,
Peinado
,
J.R.
et al.  (
2014
)
Multiple stable conformations account for reversible concentration-dependent oligomerization and autoinhibition of a metamorphic metallopeptidase
.
Angew. Chem. Int. Ed. Engl.
53
,
10624
10630
67
Zühlsdorf
,
M.
,
Werten
,
S.
,
Klupp
,
B.G.
,
Palm
,
G.J.
,
Mettenleiter
,
T.C.
and
Hinrichs
,
W.
(
2015
)
Dimerization-induced allosteric changes of the oxyanion-hole loop activate the pseudorabies virus assemblin pUL26N, a herpesvirus serine protease
.
PLoS Pathog.
11
,
e1005045
68
Niemi
,
M.H.
,
Rytkönen-Nissinen
,
M.
,
Miettinen
,
I.
,
Jänis
,
J.
,
Virtanen
,
T.
and
Rouvinen
,
J.
(
2015
)
Dimerization of lipocalin allergens
.
Scientific Rep.
5
,
13841
69
Krissinel
,
E.
and
Henrick
,
K.
(
2007
)
Inference of macromolecular assemblies from crystalline state
.
J. Mol. Biol.
372
,
774
797
70
Nooren
,
I.M.A.
and
Thornton
,
J.M.
(
2003
)
Diversity of protein-protein interactions
.
EMBO J.
22
,
3486
3492
71
Nooren
,
I.M.A.
and
Thornton
,
J.M.
(
2003
)
Structural characterisation and functional significance of transient protein-protein interactions
.
J. Mol. Biol.
325
,
991
1018
72
Laskowski
,
R.A.
(
1995
)
SURFNET: a program for visualizing molecular surfaces, cavities, and intermolecular interactions
.
J. Mol. Graph.
13
,
323
330
73
Hagemans
,
D.
,
van Belzen
,
I.A.
,
Moran Luengo
,
T.
and
Rudiger
,
S.G.
(
2015
)
A script to highlight hydrophobicity and charge on protein surfaces
.
Front. Mol. Biosci.
2
,
56
74
Walker
,
C.L.
,
Pomatto
,
L.C.D.
,
Tripathi
,
D.N.
and
Davies
,
K.J.A.
(
2018
)
Redox regulation of homeostasis and proteostasis in peroxisomes
.
Physiol. Rev.
98
,
89
115
75
Kursula
,
P.
,
Sikkilä
,
H.
,
Fukao
,
T.
,
Kondo
,
N.
and
Wierenga
,
R.K.
(
2005
)
High resolution crystal structures of human cytosolic thiolase (CT): a comparison of the active sites of human CT, bacterial thiolase, and bacterial KAS I
.
J. Mol. Biol.
347
,
189
201
76
Kiema
,
T.R.
,
Harijan
,
R.K.
,
Strozyk
,
M.
,
Fukao
,
T.
,
Alexson
,
S.E.
and
Wierenga
,
R.K.
(
2014
)
The crystal structure of human mitochondrial 3-ketoacyl-CoA thiolase (T1): insight into the reaction mechanism of its thiolase and thioesterase activities
.
Acta Crystallogr. D Biol. Crystallogr.
70
,
3212
3225
77
Haapalainen
,
A.M.
,
Merilainen
,
G.
,
Pirila
,
P.L.
,
Kondo
,
N.
,
Fukao
,
T.
and
Wierenga
,
R.K.
(
2007
)
Crystallographic and kinetic studies of human mitochondrial acetoacetyl-CoA thiolase: the importance of potassium and chloride ions for its structure and function
.
Biochemistry
46
,
4305
4321
78
Liang
,
K.
,
Li
,
N.
,
Wang
,
X.
,
Dai
,
J.
,
Liu
,
P.
,
Wang
,
C.
et al.  (
2018
)
Cryo-EM structure of human mitochondrial trifunctional protein
.
Proc. Natl Acad. Sci. U.S.A.
115
,
7039
7044
79
Jones
,
S.
and
Thornton
,
J.M.
(
1996
)
Principles of protein-protein interactions
.
Proc. Natl Acad. Sci. U.S.A.
93
,
13
20

Author notes

*

Present address: Department of Biological and Environmental Science, PO Box 35, University of Jyväskylä, Jyväskylän yliopisto FI-40014, Finland.

Present address: FBMM, University of Oulu, PO Box 5400, University of Oulu, FI-90014 Oulu, Finland.