Expression of full-length human alkylglycerol monooxygenase and fragments in Escherichia coli

Matthias Mayer; Markus A. Keller; Katrin Watschinger; Gabriele Werner-Felmayer; Ernst R. Werner; Georg Golderer

doi:10.1515/pterid-2013-0014

Article Open Access

Expression of full-length human alkylglycerol monooxygenase and fragments in Escherichia coli

Matthias Mayer , Markus A. Keller , Katrin Watschinger , Gabriele Werner-Felmayer , Ernst R. Werner and Georg Golderer

Published/Copyright: May 11, 2013

Published by

Become an author with De Gruyter Brill

Author Information Explore this Subject

From the journal Pteridines Volume 24 Issue 1

Abstract

Alkylglycerol monooxygenase (AGMO; EC 1.14.16.5) is the only enzyme known to cleave the O-alkyl ether bond of alkylglycerols in humans. It is an integral membrane protein with nine predicted transmembrane domains. We attempted to express and purify full-length and truncated forms of AGMO in Escherichia coli. Full-length AGMO could not be expressed in three different E. coli expression strains, three different expression vectors and several induction systems. We succeeded, however, in expression of three N-terminally strep-tagged truncated forms, named active sites 1, 2 and 3, with 205, 134 and 61 amino acids, respectively. Active site 1 fragment, containing two predicted transmembrane regions, a membrane associated region and all known amino acid residues important for catalytic activity, was not fully soluble even in 8 M urea. Active site 2 containing only one predicted membrane associated domain required 8 M urea for solubilisation and eluted in gel filtration in 1 M urea as a trimer. Active site 3 with no hydrophobic domain eluted in gel filtration in 1 M urea as monomer and dimer. These results show that even truncated forms of AGMO are barely soluble when expressed in E. coli and show a high tendency for aggregation.

Keywords: alkylglycerol monooxygenase; hydrophobicity; oligomerisation; tetrahydrobiopterin; Enzymes: alkylglycerol monooxygenase, gene symbol AGMO, systematic name 1-alkyl-sn-glycerol, tetrahydrobiopterin:oxygen oxidoreductase (EC 1.14.16.5)

Introduction

Ether lipids are abundant in higher eukaryotic membranes [1]. They have been shown to play an important role in the inhibition of cataract formation and signalling processes. Furthermore, they are a major component of brain membranes and are also involved in spermatogenesis [1]. Alkylglycerol monooxygenase (AGMO) is found in the endoplasmic reticulum of cells and shows highest activity in rat liver. It is the only enzyme known to cleave the O-alkyl ether bond of alkylglycerols in a tetrahydrobiopterin-dependent reaction yielding a free aldehyde and a glycerol derivative [2]. Fatty aldehyde hydroxylase (FALDH; EC 1.2.1.48) detoxifies the reactive free aldehyde through a NAD-dependent conversion to the corresponding fatty acid [3]. Although its function was first described in 1964, the AGMO sequence remained unknown until 2010 because it was impossible to purify this labile integral membrane enzyme without loss of activity. Only by selecting candidate genes by bioinformatic approaches and by proteomic analysis of partially purified fractions containing AGMO enzymatic activity and transiently transfecting 10 of them into mammalian cells, it could be shown that transmembrane protein 195 (TMEM195), an open reading frame with unknown function, was able to confer AGMO activity to cells [4]. Human AGMO shows no sequence homology to the two other classes of tetrahydrobiopterin-dependent enzymes (aromatic amino acid hydroxylases and nitric oxide synthases), and therefore it forms a third class of tetrahydrobiopterin-dependent enzymes [4]. AGMO contains a fatty acid hydroxylase motif and nine predicted transmembrane regions [4, 5]. It is not known whether the enzyme is active as a monomer or an oligomer. A secondary structure prediction and an ab initio calculation of the catalytic centre of the reaction [5] facilitated the design of three different N-terminally strep-tagged fragments with decreasing hydrophobicity (Figure 1). Here, we report our attempts to express full-length and truncated N-terminally strep-tagged peptides of AGMO in Escherichia coli, solubilise and purify them with Strep-Tactin affinity chromatography and gel filtration.

Figure 1

Domains, location of fragments and hydrophobicity of human alkylglycerol monooxygenase. (A) Schematic secondary structure of full-length alkylglycerol monooxygenase and truncations (active site 1, active site 2 and active site 3) based on Watschinger et al. [5]. Full-length alkylglycerol monooxygenase (amino acid stretch 1–445) has nine proposed transmembrane (TM) regions and a membrane-associated domain (MAD). Active site 1 (amino acid stretch 130–334) has two proposed TM regions (TM4, TM5) and a MAD. Active site 2 (amino acid stretch 201–334) lacks TM regions but incorporates a proposed MAD. Active site 3 (amino acid stretch 205–265) lacks TM regions and MADs and was proposed to be cytoplasmic. (B) Hydrophobicity blot of full-length alkylglycerol monooxygenase performed with Phobius [6]. Proposed TM motifs are shaded in grey. Proposed cytoplasmic regions are shown with a black line.

Materials and methods

Materials

Most chemicals were purchased from Sigma-Aldrich (Schnelldorf, Germany), Roth (Karlsruhe, Germany), Merck (Darmstadt, Germany), Serva (Heidelberg, Germany) and Lactan (Graz, Austria). Chemicals purchased by other companies are mentioned in the text.

Methods

Cloning

Full-length human AGMO (gene bank accession number: NM_001004320, NP_001004320, 445 amino acids, 53.4 kDa) was cloned into three expression vectors: pET28b (N-terminal His and T7 tag; Merck), pPR-IBA101 (C-terminal double Strep-tag; IBA GmbH, Göttingen, Germany) and pPR-IBA2 (N-terminal Strep-tag; IBA GmbH). Amino acid stretches 130–334 (active site 1, 25.9 kDa), 201–334 (active site 2, 17.0 kDa) and 205–265 (active site 3, 8.5 kDa) were cloned into the pPR-IBA2 vector. Standard procedures for PCR, transformation, cloning and analysis of DNA were used. All constructs were tested by restriction enzyme digest and sequencing (Microsynth AG, Balgach, Switzerland). Recombinant plasmids were transformed into E. coli expression strains Lemo21(DE3) (New England Biolabs Inc., Frankfurt, Germany), KRX (Promega, Mannheim, Germany) and BL21-CodonPlus^®-RIL (Stratagene, Vienna, Austria).

Expression

Full-length human AGMO in pPR-IBA2 was tested in all three E. coli strains. Full-length human AGMO in vectors pET28b, pPR-IBA2 and pPR-IBA101 was tested in Lemo21(DE3). The different strains were induced with isopropyl-β-D-1-thiogalactopyranoside (IPTG) according to the manufacturer’s instructions. Expression was tested up to 24 h. Cultures were shaken at 225 rpm and 37°C. In some experiments, 15°C was used instead to induce chaperone expression in E. coli [7]. An autoinduction protocol was tested for full-length human AGMO expression in all three strains transformed with human alkylglycerol monooxygenase-pPR-IBA2 [8] at 37°C. Expression of active site 1 was tested in all three expression strains [6 h of expression at 37°C, 225 rpm in 100 mL Terrific Broth (TB) medium]. Then, 1 mL samples (0 h, 1 h, 2 h, 4 h, 6 h, 8 h, 10 h, 24 h after induction) were collected by centrifugation, the pellets digested with lysozyme [100 mM phosphate buffer pH 7.0, 1 mM ethylenediamine-tetraacetic acid (EDTA), 1 mM dithiothreitol (DTT), 1 mM phenylmethanesulfonylfluoride (PMSF), 0.1 mg/mL lysozyme, 10 U/mL DNAseI] for 20 min at 25°C shaken with 650 rpm, homogenised with an ultra-turrax (T8, Ika GmbH, Staufen, Germany). Protein content was measured by Bradford assay using bovine serum albumin as standard [9, 10] and tested by Western blot (10 μg protein per lane) using anti-Strep-Tactin horse radish peroxidise (IBA GmbH). Larger cell pellets were homogenised using a French press (SLM Alminco, SLM Instruments, Corston, UK) after the lysozyme digest procedure.

Protein purification

(1) Membrane preparation. Membranes were prepared with two consecutive centrifugation steps at 4°C [3000×g (Megafuge 2.0 RS, Heraeus, Kendro Laboratory Products, Osterode, Germany) and 55,000×g (Beckman Coulter Avanti J-26XP, Beckman, Fullerton, CA, USA) at 4°C for 30 min and 60 min, respectively]. The insoluble fractions after the 55,000×g centrifugation step were washed twice with 0.5% (w/v) deoxycholate and then centrifuged at 55,000×g again. The insoluble fraction was resolubilised with the chaotropic agent urea (6 or 8 M) at a pH of 9.0 for 1 h at 25°C shaking at 225 rpm. Solubility was tested by an additional centrifugation step at 100,000×g (Beckman Optima TLX Ultracentrifuge, Beckman) for 15 min at 4°C. (2) Strep-Tactin affinity chromatography. For Strep-Tactin affinity chromatography, the resolubilised insoluble fraction (see above) was diluted 1:16 or 1:20 to reduce the urea concentration under 1 M by addition of 100 mM Tris HCl, pH 9.0, containing 150 mM NaCl, and 1 mM EDTA according to the compatibility between reagents and resin and centrifuged again at 17,000×g for 15 min at 4°C. Affinity chromatography was performed with Strep-Tactin Superflow (Iba, Goettingen, Germany). After application of the protein the column was washed with four to five column volumes of 100 mM Tris HCl, pH 9.0, containing 150 mM NaCl, and 1 mM EDTA and the protein was eluted with 2.5 mM desthiobiotin (100 mM Tris HCl, 150 mM NaCl, 1 mM EDTA, pH 9.0). The OD₂₈₀ peak fractions (Pharmacia Uvicord SII, GE Healthcare) were collected and the pH was shifted to 12 with sterile filtered 3 N NaOH, and the urea concentration was increased to 1 M using 8 M urea solution, pH 12, containing 150 mM NaCl and 1 mM EDTA. Pooled elution peak fractions were concentrated by ultrafiltration (Vivaspin 4 column, Vivascience) at 3000×g and 4°C. (3) Gel-filtration. Gel-filtration with a Superose 12/10/300 column (GE Healthcare) was used as an analytical step to investigate aggregation and as a final preparative step. Three conditions were tested for separation of both active site 2 and active site 3. Running buffer (100 mM Tris HCl, pH 12.0, 150 mM NaCl, 1 mM EDTA) was supplemented with either 1 M urea, 6 M urea or 0.1% (w/v) sodium dodecyl sulphate (SDS) at a flow of 0.5 mL/min. Eluting protein was detected by UV absorption at 280 nm (Ti Series 1050, Agilent Technologies, Vienna, Austria). Molecular mass markers used were thyreoglobulin (670 kDa), β-amylase (200 kDa), equine serum albumin (67 kDa), myoglobin (18 kDa) and FAD (0.8 kDa). Peak fractions were analysed by Coomassie stained SDS-PAGE, Western blot analyses (anti-Strep-Tactin horse radish peroxidase, IBA GmbH) and by mass spectrometry (MALDI-TOF-TOF, performed at the Division of Clinical Biochemistry, Innsbruck Medical University).

Results

Expression of full-length human AGMO

In all tested conditions which include three bacterial strains, three different expression vectors and different induction protocols (15°C and 37°C, induction by IPTG or lactose) full-length human AGMO could not be expressed as determined by anti-Strep-Tactin Western blotting. We therefore attempted to express shorter fragments of human AGMO in E. coli.

Expression of active sites 1, 2 and 3

Three truncated N-terminal strep-tagged^® fragments were cloned (for details see Methods section) and expression was tested in all three E. coli strains (Figure 2). Highest expression of active site 1 was achieved in the KRX strain (Figure 2), detectable levels of active site 1 were obtained in BL21-CodonPlus(DE3)-RIL (Figure 2). Lemo21(DE3), a BL21(DE3) derivative, showed no expression of active site 1 (Figure 2). When testing expression of active sites 2 and 3 in KRX cells, similar levels were obtained as for active site 1 and therefore further studies were performed with the KRX strain.

Figure 2

Western blots of N-terminally strep-tagged active site 1 in different E. coli strains.

AS1, bacteria transformed with an expression plasmid carrying a 205 amino acids fragment of human alkylglycerol monooxygenase; co, bacteria transformed with empty expression vector. Expression was induced with IPTG (and L-rhamnose for expression in the KRX strain) according to the manufacturer’s instructions of each individual strain. Pellets were harvested 6 h after induction (for details see Methods section) and mechanically homogenised. Then, 10 μg protein of each homogenate was separated on a 15% Tris-tricine gel and stained with anti-Strep-Tactin horse radish peroxidase antibody conjugate. The lower panels represent Coomassie staining of the membrane to confirm comparable protein loading. Molecular mass (kDa) is indicated by arrows.

Solubilisation and purification of active sites 1, 2 and 3

Preparative expression of all three human AGMO fragments was performed in KRX cells. After membrane preparation, all three fragments were insoluble in detergent-free buffer and had to be washed with deoxycholate and resolubilised with urea (8 M urea for active sites 1 and 2, 6 M urea for active site 3; Figure 3). Active site 1 was not fully soluble even in 8 M urea. By contrast, active site 2 was fully soluble after membrane preparation at 8 M urea. Active site 3 could be fully solubilised in 6 M urea. Therefore, we were able to further purify active site 2 and active site 3 with Strep-Tactin affinity chromatography. As a final purification step, we then used gel filtration at pH 12.0 with 1 M urea as an eluent (Figure 4). To obtain a monomeric peak of active sites 2 and 3, we tried to increase the molarity of urea to 6 M, or to use the detergent SDS (0.1% w/v). Both conditions did not lead to monomerisation (not shown). Thus, even at pH 12 and 1 M urea active site 2 elutes as a trimer, and active site 3 as an approximately equal mixture of monomers and dimers (Figure 3). We checked gel-filtration peaks for purity by Coomassie stained SDS-PAGE gel (Figure 5). After concentrating those fractions containing the highest amount of protein obtained during gel filtration by ultrafiltration, a 17-kDa band with immunoreactivity to the anti-Strep-Tactin antibody was identified for the active site 2 gel filtrate, corresponding to the calculated molecular mass (17 kDa) of this fragment (Figure 4A). In the active site 3 gel filtrate, however, three different bands were found, at approximate molecular masses of the monomer (8.5 kDa), the dimer (17 kDa) and the trimer (25.5 kDa; Figure 5B), which all contained the strep tag as indicated by Western blotting (Figure 4C). Bands were excised and analysed by mass spectrometric analysis. All bands were identified as human AGMO fragments. This led us to the conclusion that even small fragments of human AGMO partly aggregated when incubated with SDS and β-mercaptoethanol in SDS-PAGE.

Figure 3

Western blot of resolubilised active site 1, active site 2 and active site 3.

The three fragments were expressed as N-terminally strep-tagged proteins in KRX cell membranes prepared and resolubilised with urea. Resolubilised samples (10 μL) were separated on a 15% Tris-tricine (for active site 1 and active site 3) or 20% polyacrylamide (for active site 2) gel and stained with anti-Strep-Tactin horse radish peroxidase antibody conjugate. SUP, supernatant after resolubilisation and centrifugation at 100,000×g; PELLET, pellet after resolubilisation and centrifugation at 100,000×g. The migration of molecular mass markers is shown by arrows. The urea concentration used is indicated at the bottom of the figure.

Figure 4

Gel-filtration chromatogram of active site 2 (full line) and active site 3 (dashed line) at pH 12.0 and 1 M urea.

Elution volume of molecular mass markers is shown on top of the chromatogram. The main peak of active site 2 (full line, calculated monomeric mass 17 kDa) elutes at roughly trimeric size of approximately 51 kDa. Active site 3 (dashed lined, calculated molecular mass of 8.5 kDa) elutes with a double peak approximately at the size of the dimer (17 kDa) and the monomer (8.5 kDa).

Figure 5

Migration of purified active sites 2 and 3 in standard SDS-PAGE.

(A) Analysis of ultrafiltrate of gel filtration of active site 2 by SDS-PAGE and Coomassie staining: 160 μg protein of active site 2 was separated on a 20% polyacrylamide gel. (B) Analysis of ultrafiltrate of gel filtration of active site 3 by SDS-PAGE and Coomassie staining: 100 μg protein of active site 3 was separated on a 15% Tris- tricine gel. (C) Western blot of the corresponding active site 3 sample: 2 μg protein was separated on the same gel as in (B), blotted and the membrane stained with anti-Strep-Tactin horse radish peroxidase antibody conjugate. The full arrows indicate the sites where samples to be analysed by mass spectrometry were taken from. The migration of molecular mass markers is shown by arrowheads.

Discussion

This study shows for the first time expression and purification of a truncated form of human AGMO in amounts suitable for immunisation and other biochemical approaches. Whereas full-length human AGMO was resistant to expression in different bacterial hosts and from different vectors, a smaller fragment containing all or at least parts of the proposed catalytically active site [5] could be expressed in E. coli. It was, however, very surprising to see that even though these fragments were thoroughly chosen to harbour less and less hydrophobic segments, we were not able to fully solubilise the two longer fragments, and the smaller fragment still showed a very high tendency to aggregate. Human AGMO active site 1 (235 amino acids) is predicted to contain only two of the nine transmembrane segments found in full-length protein and also has one membrane associated domain and a hydrophilic loop. This protein could not be solubilised even in high urea concentrations. By contrast, human AGMO active site 2 (134 amino acids), which only has a single predicted membrane associated domain, and the hydrophilic loop could be solubilised in 8 M urea, but in gel filtration it appeared oligomerised. Oligomerisation was less in the shorter human AGMO active site 3 (61 amino acids, only the hydrophilic loop was present), but still it eluted as a dimer-monomer mixture in gel filtration. Not even the stringent conditions in SDS-PAGE sample preparation (5%, v/v, β-mercaptoethanol) and during the SDS-PAGE run (0.2%, w/v, SDS) were able to break these interactions fully. Although fairly rare, there are other proteins mentioned in the literature which are SDS resistant. Intermediate II [11] oligomers of α-synuclein (14 kDa) have been described to be resistant to 0.2% SDS and synaptogamin I [12] is insensitive to β-mercaptoethanol and 1% (w/v) SDS. Further attempts will be needed to optimise expression of longer AGMO fragments and also the full-length protein to allow future biochemical and structural studies. Changing the expression host from bacteria to a eukaryotic cell system might provide a more adequate setting for expression, folding and processing of such a highly hydrophobic enzyme.

Corresponding author: Georg Golderer, Division of Biological Chemistry, Biocentre, Innsbruck Medical University, Innrain 80–82, 6020 Innsbruck, Austria

We thank Petra Loitzl, Rita Holzknecht and Nina Madl (all Innsbruck Medical University) for excellent technical support. We would like to especially thank the Division of Clinical Biochemistry for mass spectrometric analyses of active site 2 and active site 3. This work was supported by the Austrian Research Funds (FWF), project P22406.

References

1. Gorgas K, Teigler A, Komljenovic D, Just WW. The ether lipid-deficient mouse: tracking down plasmalogen functions. Biochim Biophys Acta 2006;1763:1511–26.10.1016/j.bbamcr.2006.08.038Search in Google Scholar PubMed

2. Tietz A, Lindberg M, Kennedy EP. A new pteridine-requiring enzyme system for the oxidation of glyceryl ethers. J Biol Chem 1964;239:4081–90.10.1016/S0021-9258(18)91137-3Search in Google Scholar

3. Kelson TL, Secor McVoy JR, Rizzo WB. Human liver fatty aldehyde dehydrogenase: microsomal localization, purification, and biochemical characterization. Biochim Biophys Acta 1997;1335:99–110.10.1016/S0304-4165(96)00126-2Search in Google Scholar

4. Watschinger K, Keller MA, Golderer G, Hermann M, Maglione M, Sarg B, et al. Identification of the gene encoding alkylglycerol monooxygenase defines a third class of tetrahydrobiopterin-dependent enzymes. Proc Natl Acad Sci USA 2010;107:13672–7.10.1073/pnas.1002404107Search in Google Scholar PubMed PubMed Central

5. Watschinger K, Fuchs JE, Yarov-Yarovoy V, Keller MA, Golderer G, Hermetter A, et al. Catalytic residues and a predicted structure of tetrahydrobiopterin-dependent alkylglycerol mono-oxygenase. Biochem J 2012;443:279–86.10.1042/BJ20111509Search in Google Scholar PubMed PubMed Central

6. Kall L, Krogh A, Sonnhammer EL. Advantages of combined transmembrane topology and signal peptide prediction: the Phobius web server. Nucleic Acids Res 2007;35: W429–32.10.1093/nar/gkm256Search in Google Scholar PubMed PubMed Central

7. Phadtare S. Unwinding activity of cold shock proteins and RNA metabolism. RNA Biol 2011;8:394–7.10.4161/rna.8.3.14823Search in Google Scholar PubMed PubMed Central

8. Gordon E, Horsefield R, Swarts HG, de Pont JJ, Neutze R, Snijder A. Effective high-throughput overproduction of membrane proteins in Escherichia coli. Protein Expr Purif 2008;62:1–8.10.1016/j.pep.2008.07.005Search in Google Scholar PubMed

9. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248–54.10.1016/0003-2697(76)90527-3Search in Google Scholar

10. Compton SJ, Jones CG. Mechanism of dye response and interference in the Bradford protein assay. Anal Biochem 1985;151:369–74.10.1016/0003-2697(85)90190-3Search in Google Scholar PubMed

11. Kostka M, Högen T, Danzer KM, Levin J, Habeck M, Wirth A. Single particle characterization of iron-induced pore-forming alpha-synuclein oligomers. J Biol Chem 2008;283:10992–1003.10.1074/jbc.M709634200Search in Google Scholar PubMed

12. Fukuda M, Kanno E, Ogata Y, Mikoshiba K. Mechanism of the SDS-resistant synaptotagmin clustering mediated by the cysteine cluster at the interface between the transmembrane and spacer domains. J Biol Chem 2001;276:40319–25.10.1074/jbc.M105356200Search in Google Scholar PubMed

Received: 2013-3-11

Accepted: 2013-3-29

Published Online: 2013-05-11

Published in Print: 2013-06-01

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Articles in the same Issue

https://doi.org/10.1515/pterid-2013-0014

Keywords for this article

alkylglycerol monooxygenase; hydrophobicity; oligomerisation; tetrahydrobiopterin; Enzymes: alkylglycerol monooxygenase, gene symbol AGMO, systematic name 1-alkyl-sn-glycerol, tetrahydrobiopterin:oxygen oxidoreductase (EC 1.14.16.5)

Creative Commons

BY-NC-ND 3.0