Crystal Structure of Candida antarctica Lipase B with a Putative Pro-Peptide Region

Abstract

There are 25 crystal structures of Lipase B from Candida antarctica (CalB) that have been previously reported. In this study, we report the first CalB crystal structure that shows the assumed pro-peptide region at the N-terminus (Ala19–Arg25). This 1.45 Å structure shows that this segment of seven amino acids is an extension of the N-terminal loop and that it does not interact with or effect conformational changes in the flexible lid domain, which covers the active site of the enzyme. As such, this region is unlikely to be a classical pro-peptide.

1. Introduction

Lipase B from Candida antarctica (CalB) has been extensively studied both functionally and structurally due to its vital role in research and industry as a biocatalyst [1,2,3,4,5,6,7,8,9]. CalB has an α/β hydrolase fold [10], which is structurally stabilized by three disulfide bonds. These disulfide bonds are formed between Cys47–Cys89, Cys241–Cys283, and Cys318–Cys336. A catalytic triad of Ser130–Asp212–His249 creates the active site of the enzyme [1]. The entrance of this active site is guarded by a flexible region (Val164-Ala176), also known as a lid domain. Due to the pliable nature of this lid domain, it can have different loop conformations, such as an open or closed loop [5,6,7,8,11]. Traditionally, the open conformation has an alpha helix (α5) formation [1,2,6,7,9,12]. The lid domain is mainly hydrophobic in nature, having only one polar residue (Asp170). Asp170 plays a significant role in observed conformational changes, as it can form a salt bridge with Lys315 to form a fully closed conformation [6], or, when not interacting with Lys333, it can result in a fully open conformation [11].

CalB has been extensively structurally characterized, with 25 crystal structures published over the past 30 years. All of these structures represent the mature enzyme (Leu26–Pro342), lacking both the cleavable signal sequence (Met1–Ala18) and the seven-amino-acid sequence (Ala19–Arg25) from the N-terminus of the protein, which is annotated as being a pro-peptide region [1,13,14].

Pro-peptides are regions that are cleaved by proteolysis after either assisting in the folding of the mature protein or maintain the enzyme in an inactive state [15,16]. There are many lipases that are reported to have a pro-peptide region, but their sequence is generally longer than the seven-amino-acid sequence at the N-terminus of CalB. For example, (i) lipase from Rhizomucor miehei (RmL) comprises an N-terminal pro-peptide region with 70 residues, followed by 269 residues of mature protein [17]; (ii) another lipase from Rhizopus oryzae (RoL) has a pro-peptide of 97 residues [18]; (iii) a 207-residue-long pro-peptide region is reported for Staphylococcus hyicus lipase (SHyL) [19]; (iv) Rhizopus chinensis lipase (RcL) has 94-amino-acid-long pro-peptide residues [20].

All previous structural work, as well as virtually all biochemical studies and uses of CalB, is based on the mature protein. This suggests that the “pro-peptide” does not have a significant function in folding to the native state and, hence, implies that if it is a pro-peptide, it may maintain the enzyme in an inactive state. However, we have reported high-level production of CalB that includes the N-terminal putative pro-peptide along with the mature protein (Ala19–Pro 342), with the enzyme having higher activity than commercially sourced CalB [21]. This raises questions as to whether this segment of seven amino acids is a pro-peptide region. If so, what is the function of this segment, and does it interact with the lid domain or the active site in any way to modulate activity? To address these questions, we report here the first crystal structure of CalB that includes the assumed pro-peptide region. This structure, determined at a 1.45 Å resolution, reveals that this segment is not structured and that it does not interact with the lid domain or the active site.

2. Materials and Methods

2.1. CalB Production

A gene corresponding to UniProt entry P41365 encoding CalB residues Ala19–Pro342 (lacking the N-terminal signal peptide) was purchased from GenScript (GenScript, Piscataway, New Jersey, United States). The gene was inserted using the restriction site pairs NdeI/BamHI into a modified pET23-based vector with a pTac promoter replacing the T7 promoter. This resulted in a protein with an N-terminal hexa-histidine tag followed by a TEV cleavage site and MetAla (encoding the NdeI restriction site) prior to the first amino acid of the protein sequence. The resulting plasmid was named pAS88. The plasmid was sequenced to confirm the correct insertion of the gene.

Protein expression (

Table 1. Summary of CalB expression.

Source Organism	Candida antarctica (Yeast)
DNA source	Codon optimized by GenScript
Cloning vector	Modified pET23-based with a pTac promoter
Expression vector	pAS88 and pMJS205 1
Expression host	Escherichia coli BL21 (DE3)

¹ A CyDisCo (cytoplasmic disulfide bond formation) variant that assists in protein folding [23].

) and purification were carried out as described earlier for disulfide-rich proteins [22]. Two plasmids, pAS88 and pMJS205, were co-transformed into the chemically competent Escherichia coli (E. coli) strain BL21 (DE3) from Stratagene. The main culture was grown in terrific broth autoinduction media (Formedium) with trace elements and 0.8% v/v glycerol at 30 °C. A three-step purification strategy, including immobilized metal affinity chromatography (IMAC), reverse IMAC after TEV cleavage, and size exclusion chromatography (SEC), was carried out for protein production [22]. The cleavage of the N-terminal hexa-histidine tag was carried out after IMAC purification. The IMAC-purified protein was buffer-exchanged with 20 mM phosphate buffer, pH 7.4, and 150 mM NaCl using a PD-10 desalting column. The protein was incubated with TEV protease (1:50 molar ratio) overnight at 4 °C. The protein sample was then loaded onto a HiTrap™ 5 mL chelating HP column (GE Healthcare, Little Chalfont, UK), and the same protocol was followed for reverse IMAC as for IMAC purification. The histidine-tag-cleaved CalB protein was found, as expected, in the flowthrough. The protein sample was concentrated to 1 mL volume and was further purified using a pre-equilibrated Superdex 200 Increase 10/300 column (Merck, Darmstadt, Germany) with 20 mM phosphate buffer, pH 7.4, and 150 mM NaCl for SEC purification. SEC-purified protein was pooled and concentrated for protein crystallization. The protein was concentrated using Amicon ultra-4 centrifugal filters (Merck, Darmstadt, Germany) with a 10 kDa molecular weight cutoff. The extinction coefficient for full-length protein is 41,285 M⁻¹ cm⁻¹ at 280 nm.

2.2. CalB Crystallization

The final concentration of purified CalB was 10 mg/mL in 20 mM phosphate buffer, pH 7.4, and 150 mM NaCl. It was used for crystallization through the sitting-drop vapor diffusion method at 22 °C (295.15 K) using a PACT commercial crystallization screen (molecular dimensions). A crystal seed stock (1:1000 v/v) dilution was prepared from PACT crystallization conditions (200 mM potassium sodium tartrate tetrahydrate; 100 mM bis-tris propane, pH 8.5, 20% w/v PEG 3350). Crystal quality was optimized by changing the buffer pH and PEG 3350 concentration. The optimization screen was prepared with 200 mM potassium sodium tartrate tetrahydrate; 100 mM bis-tris propane, ranging from pH 6.5 to 8.5; and a PEG 3350 concentration (w/v) ranging between 20% and 25%. The crystallization drop was prepared on a 96-well triple sitting-drop plate (sptlabtech) using a Mosquito LCP nanodispenser (sptlabtech, Melbourn, United Kingdom) with a drop volume of 200 nL of protein, 160 nL of reservoir solution, and 40 nL of crystal seed. The crystallization parameters are summarized in

Table 2. Summary of CalB crystallization.

Method	Sitting-Drop Vapor Diffusion
Plate type	96-well TTP Lab Tech triple sitting-drop
Temperature	295.15 K
Protein concentration	10 mg/mL
Protein storage solution	20 mM phosphate buffer, pH 7.4; 150 mM NaCl
Crystallization solution	200 M potassium sodium tartrate tetrahydrate 100 M bis-tris propane, pH 7.5; 25% w/v PEG 3350
Crystallization drop volume	Crystal drop volume is 400 nL, where 200 nL is protein, 160 nL of crystallization solution, and 40 nL is crystal seed

. A single well-ordered 150-micron-sized crystal was stored in liquid nitrogen and was used for data collection.

2.3. Data Collection, Processing, Structure Solution, and Refinement of CalB Crystal

The X-ray diffraction data from the frozen crystal was collected with the I04 beamline at Diamond Light Source (DLS), United Kingdom. The crystal diffracted to a maximum resolution of 1.19 Å, but considering the completeness parameter (i.e., >95%), the cutoff resolution was 1.45 Å during refinement. The auto-processed data from the automated Xia2 DIALS pipeline was of good quality for structure solving via molecular replacement using Phenix.Phaser version 1.17.10 [24] in the P2₁ space group. The model template for molecular replacement used was chain A of CalB (PDB entry 1TCA [1]), with 100% structure identity and 97% sequence coverage. Structure refinement and manual model building were carried out using Phenix.Refine version 1.17.10 of the PHENIX suite [25] and COOT version 0.9.8.95 [26], respectively. As the data was near atomic resolution (1.45 Å), the hydrogen atoms were added to the model, and the temperature factors (B-factors) were refined anisotropically. The quality of the final models was validated with MolProbity version 4.5.2 [27]. The data-processing and structure refinement statistics are reported in

Table 3. Data collection and statistics of the data-processing and refinement of the structures of CalB (Ala19–Pro342).

Data Collection Parameters
Diffraction source	I04-DLS-UK
Wavelength (Å)	0.9795
Detector	Eiger2 XE 16 M
Temperature (K)	100
Data-processing software	XDS/Aimless
Data-processing statistics
Unit cell parameters (Å,°)	a = 47.34, b = 80.98, c = 73.99 α = 90.00, β = 98.30, γ = 90.00
Space group	P21
Resolution range (Å)	54.37–1.45 (1.45–1.41) 1
Molecules per asymmetric unit	2
Number of observations (total)	503,254 (27,177) 1
Number of observations (unique)	104,810 (7629) 1
Redundancy	4.4 (3.56) 1
Completeness (%)	99.86 (86.60) 1
Rmerge (%)	4.7 (41.4) 1
I/σ (I)	14.5 (1.3) 1
CC (1/2)	0.999 (0.905) 1
Refinement statistics
Resolution (Å)	46.85–1.45 (1.51–1.45) 1
Rwork (%)	13.59 (19.45) 1
Rfree (%)	16.79 (24.78) 1
Number of unique reflections	97,613
Number of non-hydrogen atoms	4700
Average B-factors (Å2)	23.12
r.m.s.d. bond length (Å)	0.005
r.m.s.d. bond angle (°)	0.81
Ramachandran plot
Favored (%)	97.17
Allowed (%)	2.83
Outliers (%)	0.00
PDB code	9EVI

¹ The number in parentheses refers to the outer shell.

. The coordinates have been deposited in the Protein Data Bank with the accession code 9EVI. CCP4MG version 2.10.11 was used to prepare structural figures and an animation video [28]. The superposition of models and the root mean square deviations (RMSDs) were calculated using SUPERPOSE-COOT version 0.9.8.95 [29]. Buried surfaces and residues at the intermolecular contacts in the crystals were identified with the PISA server (http://www.ebi.ac.uk/msd-srv/prot_int/pistart.html (accessed on 1 April 2025) [30].

3. Results and Discussion

3.1. Overall Structure of CalB

CalB was crystallized in space group P2₁ with two molecules in the crystallographic asymmetric unit (Figure 1a). The RMSD values of the carbon-alpha (Cα) atoms for CalB chains A and B were solved in this study against chain A of the previously solved structure (PDB entry 1TCA [1]), and the values are 0.47 Å and 0.50 Å, respectively. Access to the active site of the protein, as governed by the flexible lid domain (Val164–Ala176), is different between the two chains (Figure 1a), with chain A in the closed state and chain B in the open state conformation. The conformational changes in the lid domain (closed and open states), as reported earlier [6], are due to the contacts between alpha helix α5 (Pro293–Ala312) and the beta hairpin motif (Lys333–Pro342). Polyethylene glycol (PEG) molecules can be observed near the active site in chain B of our structure, which has the open lid conformation, whereas they are not found in chain A, with the closed lid conformation.

While 25 previous crystal structures are available for CalB (Leu26–Pro342), this is the first structure for the protein that also includes Ala19–Arg25, the designated putative pro-peptide region of CalB [1,13,14]. The sequence of this fragment consists of seven amino acids (ATPLVKR). This region was visible in chain A due to its conformation being fixed by crystal contact sites. These interactions were not present for chain B; hence, this region in chain B was flexible, and therefore, the N-terminal seven residues of chain B are not visible in the electron density map. A polder electron density map [31] at sigma 3 was created for the chain A residues (Figure 1b). This fragment is 16.6 Å long and has very high isotropic B-factor values ranging from 51 to 102 for the Cα main of chain A (Figure 1c). This segment is dominated by hydrophobic residues except for two basic amino acid residues (Lys24 and Arg25). The crystal structure reveals that this segment sequence should simply be viewed as a flexible loop extension region of the traditional N-terminal of mature CalB (Leu26–Ser35), which is reported in all previous structures.

The N-terminal regions for both chains are anti-parallel to each other, with the presence of solvent accessible residues (Figure 2). A PISA interface analysis shows that the interface area for both chains of the asymmetric unit is 664.8 Å. Weak hydrophobic interactions can be seen between the chain A residues (Thr20, Pro21, and Val23) and the C-terminal region chain B residues (Pro324, Pro328, and Phe329), as shown in Figure 2b. Apart from these weak contacts, no interactions were observed for the N-terminal segment (Ala19–Arg25), including with the active site and/or the lid domain, as might be expected for a pro-peptide. This structural analysis, combined with the successful production and crystal structures of the Leu26–Pro342 form of the protein created by multiple groups [1,2,4,5,6,7,8,9,11,12], suggests that these seven amino acids are not involved in folding or in the activity of the enzyme. Furthermore, the N-terminal loop extension is far away from the active site and the lid domain in both the open and closed conformations, and the enzyme in this region is highly active [21]. Hence, it is unlikely that this region plays any role in keeping the protein in an inactive state, the other traditional role assigned to pro-regions.

3.2. Comparative Analysis of Conformational Changes in the Lid Domain

The lid domain (Val164–Ala176) plays an important role in the activity of CalB. Three distinct loop conformations and an α-helical conformation can be observed in different CalB structures (Figure 3). The solved structure was structurally aligned with the crystal structure of PDB entry 5A71 [6] and PDB entry 1TCA [1]. The loop in chain A of this structure is slightly bent toward the α-helix (Figure 3), and this conformation can be considered a partially closed lid conformation. In contrast, this region of chain B moves toward the C-terminal β-hairpin domain, resulting in a fully open lid conformation (Figure 3).

A DALI analysis for structural alignments of chain A of 9EVI, showing the extended N-terminal loop with 25 previously solved crystal structures of mature CalB, was performed (

Table 4. A summary of all the previous crystal structures of CalB. The Z-score and RMSD values are reported via DALI analysis using chain A of 9EVI (solved structure) with an N-terminal extended region (Ala19–Arg25). Different conformations of the lid domain per chain per structure are also reported. The structure resolution, cell parameters, and symmetry groups for these structures can be found in Supplementary Table S1.

#	PDB	Chain	Lid Domain chain	Z-Score	RMSD	Reference
1	1LBS	A–F	Helical A–F	57.2–57.1	0.5	[2]
2	1LBT	A–B	Helical A–B	57.5	0.6
3	1TCA	A	Helical A	57.2	0.6	[1]
4	1TCB	A–B	Helical A–B	57.3–57.2	0.6
5	1TCC	A–B	Helical A–B	57.8–57.7	0.5–0.6
6	3ICV	A	Unmodelled A	46.2	1.0	[4]
7	3ICW	A	Unmodelled A	45.9	1.0
8	3W9B	A–D	Helical A–D	57.1–56.3	0.6–0.7	Unpublished
9	4K6G	A–B	Unmodelled A, Open B	58.3–58.1	0.4–0.8	[5]
10	4K6H	A–B	Open A, Closed B	59.4–58.2	0.4–0.9
11	4K6K	A–B	Unmodelled A–B	57.6–57.1	0.4–0.5
12	4K5Q	A	Open A	57.9	0.9
13	4ZV7	A	Helical A	57.9	0.6	[12]
14	5A6V	A–B	Helical A, Closed B	57.4	0.6–0.7	[6]
15	5A71	A–B	Helical A, Closed B	57.9–57.7	0.6–0.7
16	5GV5	A–H	Helical A,B,F,G,H, Unmodelled D,E, Open C	57.7–56.7	0.6–0.9	[7]
17	6ISP	A–D	Closed A–D	57.3–57.2	0.7–1.0	[8]
18	6ISQ	A–B	Open A–B	55.3–55.2	1.3
19	6ISR	A–B	Open A–B	53.6–53.6	1.2–1.3
20	6J1P	A–B	Closed A, Open B	59.5–57	0.4–0.9	[11]
21	6J1Q	A–B	Closed A, Open B	59.3–57.6	0.4–0.8
22	6J1R	A–B	Open A, Closed B	58.9–58	0.5–0.9
23	6J1S	A–B	Closed A, Open B	59.2–57.3	0.4–0.9
24	6J1T	A–B	Open A, Closed B	58.2–57.8	0.7–0.9
25	6TP8	A–C	Helical A–C	58.1	0.6	[9]

). The high Z-score and low RMSD values show high structural similarity, with the only significant differences being in the lid domain due to the different conformations. Structural alignment shows that chain A of this structure has a similar lid domain conformation to the crystal structure of PDB entries 4K6G chain B, 4K6H chain A, 4K5Q [5], 5GV5 chain C [7], 6J1P chain B, and 6J1Q chain B [11]. In contrast, chain B of this structure has a similar lid domain structure to PDB entries 4K6H chain B [5], 6J1Q chain B, 6J1R chain B, 6J1S chain A, and 6J1T chain B [11]. This further confirms that the extended N-terminal loop does not have any influence on the lid domain conformation.

3.3. Comparison with the Pro-Peptide Region of Other Lipases

The structure of RmL (PDB entry 6QPP [32]) reveals that its pro-peptide region wraps around the surface of the mature protein, burying the active site and binding area for the substrate before maturation, resulting in the inhibition of the enzyme [32]. Similarly, the 27-amino-acid pro-peptide of RcL, in a structural study by Zhang et al., was reported to have the 27 residues of the pro-peptide, also known as the N-terminal polypeptide segment (NTPS), which inhibits the activity of the protein by interacting with both the closed lid domain and the core protein structure [33]. In contrast, the pro-peptide region of SHyL has been reported to have the other classical function of a pro-peptide, i.e., folding, in that it is responsible for the translocation of mature protein and stabilization against proteolytic degradation [19]. The pro-sequence for RoL is reported to have both classical pro-peptide functions, in that it is responsible for the proper folding of the mature protein and inhibits enzyme activity [18].

4. Conclusions

The structure presented in this study shows that this region (Ala19–Arg25) is the continuation of the N-terminal rather than an annotated pro-peptide domain. This sequence is small compared to other lipases with longer structured pro-peptide domains. The region appears to play no role in folding or in its activities, the latter being consistent with it not being in the vicinity of the active site or the lid domain. Hence, it is unlikely to be a pro-peptide.

5. Patents

A patent for the production system used to make the protein for structural studies using sulfhydryl oxidases in the cytoplasm of E. coli is held by the University of Oulu: Method for producing natively folded proteins in a prokaryotic host (Patent number 9238817; date of patent, 19 January 2016). Inventor: Lloyd Ruddock.