Specificity of acyl-CoA binding protein to acyl-CoAs: influence on the lipid metabolism in Aspergillus oryzae

Abstract

Acyl-CoA binding protein (ACBP) is involved in lipid metabolism and regulation of gene expression in eukaryotic cells, however, the specific functional roles of this important class of proteins remain to be elucidated. We have cloned and expressed a recombinant Aspergillus oryzae ACBP (AoACBP) that has a preference for binding relatively long chain acyl-CoAs such as palmitoyl (C16:0)-CoA (K_d = 82 nM) and stearoyl (C18:0)-CoA (K_d = 59 nM) by microscale thermophoresis binding assay. The high affinities for binding C16:0-CoA and C18:0-CoA were also conformed to the composition of fatty acids in A. oryzae 3.042. Moreover, the increased AoACBP levels were analyzed by Western blot procedure. And the expression levels and distribution patterns of AoACBP in this fungal life cycle were also investigated by RT-PCR and immunofluorescence. The accumulation of AoACBP have an effect on lipid metabolism, and it was significantly increased from 24 to 48 h suggest that AoACBP plays an important role in the later growth stage of A. oryzae 3.042 for fatty acid synthesis.

---

Introduction

The commercial value of the koji mold, Aspergillus oryzae, is reflected in its wide use for the industrial production of certain traditional fermented Asian foods such as soy sauce and soybean paste.^1,2 And the lipids of soy are always consumed and transformed by A. oryzae or yeast in the fermentation. For example, 83–90% oil of soy sauce are fatty acids that composed of palmitic acid (11–17%), oleic acid (15–20%), linoleic acid (55–60%), linolenic acid (6–9%) and stearic acid (2–3%).³ Some reports also showed that the primary sources of n−3 fatty acids like α-linolenic acid (ALA, 18:3) were vegetable oils for human dietary intake, principally soybean in the United States and Japan.^4,5

Microbial fatty acid biosynthesis and degradation is very important in fermentation. The mycelium of A. oryzae is usually rich in lipids and contains visible lipid droplets in tissues.⁶ Fatty acids are major constituents of phospholipids and neutral lipids, and fatty acyl-CoAs play a vital part in the oxidation and synthesis of fatty acids. Acyl-CoA binding protein (ACBP), one candidate cytosolic protein, is believed to play an important role in intracellular acyl-CoA transport and thus the pool formation of acyl-CoA esters, which modulate diverse cellular functions, such as regulation of energy metabolism and cell signaling.^7,8 Some studies have also shown that ACBP binds long-chain acyl-CoA esters (LCA) (C₁₂–C₂₂) with high specificity and affinity.⁹ LCA act both as intermediates in cellular metabolism and as regulators of intracellular functions.¹⁰ ACBP is involved in donating LCA for gene regulation.¹⁰ LCA can bind to the fadR gene and thereby inhibit DNA-binding activity of fadR gene, which coordinately regulated the fatty acid biosynthesis and degradation.^11,12 The acyl-CoA/ACBP complex can donate acyl-CoA for β-oxidation.¹³

ACBP was first identified in mammals, and now it had been found in eukaryotic organisms and a number of pathogenic bacteria for its high conserved sequences. Trypanosome ACBP has a high affinity for binding myristoyl-CoA.¹⁴ Bovine hepatic ACBP has the same binding affinity of hexadecanoyl-CoA and cis-9-octadecenoyl-CoA, but the rat was more tend to the hexadecanoyl-CoA.¹⁵ There is no report about ACBP from filamentous fungi up to now. Previously we showed that the A. oryzae ACBP (AoACBP) displayed much greater affinity for palmitoyl-CoA than for myristoyl-CoA, and the acyl-CoA binding domain of AoACBP contains evolutionarily conserved residues Tyr and Lys which are critical to acyl-CoA binding activity.¹⁶

While the conserved nature of AoACBP suggests its essential function in cellular metabolism and development, its specific roles in the life cycle of A. oryzae remains to be elucidated. However, the essentiality of AoACBP gene for growth and development prevents the use of a conventional genetic approach to assign specific functions to various developmental stages of A. oryzae. In this study, the binding affinities of AoACBP and acyl-CoAs, and the effect on lipid metabolism to be proposed by microscale thermophoresis binding analysis and the transcriptome analysis. As a result, these efforts will accelerate the translation of fundamental knowledge into practical applications and thus lead to the development of efficient and robust industrial strains of this important fungus.

Materials and methods

Strains, plasmids, RNA extraction and transcriptome sequencing

Escherichia coli JM109, E. coli Rosetta (DE3), and A. oryzae 3.042 were obtained from the strain collection center of Tianjin University of Science & Technology (China). Vector PMAL-c4x (Invitrogen, USA) was used to construct recombinant plasmid for the maltose-binding protein (MBP) fusion expression of AoACBP protein. Total RNAs of A. oryzae 3.042 were extracted using Trizol reagents (Promega, USA), DNaseI, and Sera-mag Magnetic Oligo(dT) Beads (Illumina). The integrity and quantity of total RNAs were assessed by the Nano Drop (NanoDrop Technologies, USA) and synthesis of the first cDNA strand was performed using 1 μg total RNAs from each sample with PrimeScript Reverse Transcriptase in a 20 μL reaction volume according to the manufacturer's instructions. The cDNA was end-repaired, amplified, denatured, and sequenced with an Illumina Genome Analyzer IIx using proprietary reagents. RNA-Seq libraries were constructed using SOLiD Total RNA-Seq Kit, and the reads were mapped to the genomes. Fragments per kilobase of exon model per million mapped reads (FPKM) were measured and represented the gene expression levels.

Cloning, expression and purification of AoACBP

AoACBP from A. oryzae 3.042 was up to 77% identical with A. fumigatus, A. clavatus, A. nomius, and 99% identical with A. flavus (Table 1). The cDNA of AoACBP (GenBank accession number Ao3042_01296) was cloned and expressed as an MBP-AoACBP fusion protein, and purified with a step gradient of maltose.¹⁶ MBP-AoACBP fusion protein and MBP-tag was estimated to be 82 kDa and 42 kDa.

Table 1 Percentage identity of the nucleotide sequence of Ao3042_01296 with other fungal ACBPs

Species	Gene sequence	Identity (%)
Aspergillus oryzae RIB40	AO090001000299	100
Aspergillus flavus NRRL3357	AFLA_067050	99
Aspergillus nomius NRRL 13137	ANOM_009680	86
Aspergillus clavatus NRRL	ACLA 023160	79
Aspergillus fumigatus Af293	AFUA 1G12300	77
Neosartorya fischeri NRRL 181	NFIA 013220	78
Aspergillus niger CBS 513.88	XM_001392573.2	No

---

Microscale thermophoresis binding assay and fatty acids analysis

The binding affinities of AoACBP and fatty acyl-CoAs were measured using the microscale thermophoresis (MST) method with a NanoTemper monolith NT.115 (US, California). The fatty acyl-CoAs were listed in Table 2. Different concentrations of test ligands (ranging from 62.5 μM to 0.0153 μM) were respectively incubated with 10 mM purified MBP-AoACBP or MBP protein for 5 min in assay buffer (10 mM KH₂PO₄, 10 mM K₂HPO₄, pH 7.4). Then the samples were loaded into the NanoTemper glass capillaries and microthermophoresis was carried out at a light emitting diode (LED) power of 10% and a MST power of 80%. Finally, K_d values were calculated using the mass action equation via the NanoTemper software from duplicate reads of triplicate experiments.¹⁷

Table 2 The attinity of K_d between MBP-AoACBP and each fatty acyl-CoA tested by MST

Numbers	Carbon numbers	Species	Kd value
1	C4:0	Butyryl-CoA	644 ± 32 nM
2	C6:0	Hexanoyl-CoA	556 ± 41 nM
3	C8:0	Octanoyl-CoA	551 ± 36 nM
4	C10:0	Decanoyl-CoA	460 ± 27 nM
5	C12:0	Dodecanoyl-CoA	212 ± 18 nM
6	C14:0	Myristoyl-CoA	365 ± 23 nM
7	C16:0	Palmitoyl-CoA	82 ± 7 nM
8	C18:0	Stearoyl-CoA	59 ± 6 nM
9	C20:0	Eicosanoyl-CoA	23 ± 5 nM
10	C22:0	Behenoyl-CoA	33 ± 6 nM

---

The mold mycelia was freeze dried and crushed using mortar and pestle. The internal standard C17:0 (3 mg mL⁻¹) was added before the methyl esterification by chlorhydric acid–methanol (10%, 1 mL) for 3 h at 60 °C. Then 1 mL n-hexane was added for extraction, and the mixture was shocked and centrifuged at 6000 × g for 3 min. The upper layer was extracted. This procedure was operated twice. The upper layer was blasted with N₂. Finally the mixture was dissolved by 1 mL n-hexane, then analyzed by GC/MS. GC/MS was performed with a 30 m × 250 μm × 0.25 μm DB-WAXETR column (Agilent Technologies, Santa Clara, California) mounted in an Agilent GC-MS 7890A/5975C (Agilent Corp., USA). The temperature program was om agreement with those reported by Wang et al.¹⁸

Immunization of rabbits with recombinant AoACBP

A pathogen-free rabbit was used to raise antibodies to AoACBP. Initial immunization used 0.2 mg affinity-purified MBP-AoACBP protein emulsified in an equal volume of complete Freund's adjuvant. Two subsequent booster immunizations (0.1 mg) were injected at 30 and 60 days, respectively, after the primary immunization. Rabbit sera were collected prior to and after the immunization protocol. The anti-MBP portion of the polyclonal antibodies was removed by absorbing antiserum with an equal volume of amylose-resin conjugated with MBP. The antibody titer and specificity were evaluated by Western blot analysis.

Western blot analysis

Western blot analysis was performed to examine the presence and levels of AoACBP protein in A. oryzae cells after growth on rice juice agar plates for 24, 30, 36, 42, and 48 h. Mycelia from the above plates were harvested respectively and grinded into the powder in liquid nitrogen. Total proteins were extracted from 100 mg of mycelia power according to a previously described protocol.^19,20 Extracted proteins were fractionated in a 12% SDS-PAGE gel and transferred onto a nitrocellulose membrane. The membrane was first blocked with 5% BSA in TBS (20 mM Tris, pH 7.5, 50 mM NaCl) for 1 h and then incubated with rabbit anti-AoACBP antibodies and a monoclonal anti-rabbit IgG antibody conjugated to alkaline phosphatase in 1% BSA in TTBS (TBS with 0.05% Tween-20). The blot was washed three times with TTBS after each incubation step. Finally, the labeled proteins were developed using 5′-bromo-4-chloro-3-indolyl phosphate (BCIP). The β-actin protein was used as a loading control.

Real-time quantitative PCR (qRT-PCR)

To quantify expression levels of ACBP gene at different growth points, we conducted real-time quantitative PCR (qRT-PCR) reactions using the primer pair AoACBP-F (5′-CTTTGACCTGGCTATTCTGGGGAT-3′) and AoACBP-R (5′-GCGGAGGTGAAGACGCG-3′). The total volume of PCR reaction is 20 μL containing 10 μL of MasterMix with SYBR (Solarbio, Beijing, China), 300 nmol liter⁻¹ of both primers and 1 μL of cDNA template. The qRT-PCR amplification parameters comprise denaturation (95 °C, 10 min), 40 cycles of denaturation (95 °C, 30 s) and annealing (60 °C, 30 s). PCR amplifications were performed on an Applied Biosystems 7500 Real Time Quantitative PCR System workstation (Applied Biosystems, USA). The transcript level of glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH) as an internal control was PCR-amplified using the primer set: GADPH-F 5′-GCTGCCCATCAAGCACGG-3′ and GADPH-R 5′-CCTACAGAGCGGGTGACAAAG-3′. Differential gene expression levels were calculated using the 2^−△△c_t method.

Immunofluorescence localization of AoACBP in A. oryzae

To examine the presence and/or localization of AoACBP in the developmental stages of A. oryzae, slides were inserted into the rice juice agar plates at an angle of about 45°, and A. oryzae spores were inoculated near the insertion positions. Individual slides were removed at 30, 36, and 42 h. Fungal cells on each slide were fixed with 4% polyformaldehyde for 20–30 min, rinsed with PBS, permeablized with 0.2% Triton X-100 for 10 min, blocked in 3% BSA-PBS (20–30 min), labeled with primary antibodies (1 h in 0.5% BSA-PBS) and incubated with secondary antibody conjugated with FITC in the dark at room temperature (2 h in 0.5% BSA-PBS). Samples were washed after each incubation step (three times, 5 min each) using PBS, and observed under immunofluorescence microscopy (Olympus, Japan).

Results

Binding preference of fatty acyl-CoAs and fatty acids analysis of A. oryzae

The binding preferences between AoACBP and the fatty acyl-CoAs were investigated at room temperature using MST. During the experiment, a constant concentration of MBP-AoACBP was fixed at 10 mM. By titrating MBP-AoACBP with an increasing concentration of fatty acyl-CoAs (C4:0–C22:0), the binding curves were obtained (Fig. 1). The purified MBP tag was used for the control binding assay at the same time. MST yielded a K_d value for the interaction of MBP-AoACBP and each kind of fatty acyl-CoA (Table 2). The K_d values for MBP-AoACBP to butyryl (C4:0)-CoA, hexanoyl (C6:0)-CoA, octanoyl (C8:0)-CoA, decanoyl (C10:0)-CoA, dodecanoyl (C12:0)-CoA and myristoyl (C14:0)-CoA were over 200 nM. Remarkably, the K_d values for palmitoyl (C16:0)-CoA, stearoyl (C18:0)-CoA, icosanoyl (C20:0)-CoA and behenoyl (C22:0)-CoA were 82 ± 7, 59 ± 6, 23 ± 5, 33 ± 6 nM respectively, which showed the better binding preference for LCAs. And no significant binding was observed between MBP tag and the fatty acyl-CoAs.

Fig. 1 The K_d values for ligand to binding to AoACBP using the microscale thermophoresis method.

The fatty acid composition of A. oryzae 3.042 was shown in Fig. 2. The contents of the fatty acids were raised under accelerated growth conditions. Palmitic acid (C16:0) and stearinic acid (C18:0) were particularly abundant in these kinds of fatty acids, and the contents were 11.6 ± 0.2 and 14.7 ± 0.8 mg g⁻¹ respectively. The results were in concert with the binding preferences above. Although the binding affinities of C20:0-CoA and C22:0-CoA with AoACBP were higher than C18:0-CoA, the contents of C20:0 and C22:0 were much lower than that of C16:0 and C18:0.

Fig. 2 The contents of fatty acids in A. oryzae 3.042 at different growth stages.

Western blot and expression analysis of AoACBP

To determine the expression levels of AoACBP at the life cycle stages of A. oryzae 3.042, polyclonal antibodies to AoACBP were used (as described under experimental procedures). As shown in Fig. 3a, the expression levels of AoACBP were significantly increased under accelerated growth conditions (24–48 h). This findings were further reinforced the idea that AoACBP plays a critical role in the life cycle.

Fig. 3 The expression levels of AoACBP by Western blot (a) and RT-PCR (b). Rabbit anti-AoACBP antibodies were used for immunoblot.

To gain insights into the physiological role of AoACBP gene in the life cycle stages of A. oryzae 3.042, we used qRT-PCR to measure the transcript level of fungal cells grown on rice juice agar plates for 24–48 h. The expression levels of AoACBP was increased steadily from 24 h to 48 h, after normalizing of the expression levels of the GAPDH gene (Fig. 3b).

Immunolocalization of AoACBP

We explored the function of AoACBP via immunolocalization considering its essential role in lipid metabolism. We took advantage of the simple technique to allow the fungus to grow onto a slide and complete its life cycle there. AoACBP antibodies were used to localize the AoACBP protein in the fungal cells by immunofluorescence microscopy. Immunostaining of the spore-containing structure of old mycelium and spores were observed at 42 h, whereas no strong immunostaining of young mycelium was observed at 30 and 36 h (Fig. 4). The immunolocalization of AoACBP clearly indicates that AoACBP plays a critical role in the late stage for the accumulation of long-chain fatty acids in A. oryzae 3.042.

Fig. 4 Immunolocalization of AoACBP in the life cycle stages of A. oryzae by immunofluorescence. Fungal tissues were labeled with primary antibodies against AoACBP and incubated with secondary antibodies conjugated with FITC.

Transcriptome analysis of the lipid metabolism

The transcriptomes of A. oryzae 3.042 (grow 30, 36, 42 h) were sequenced and produced 2.8 × 10⁷, 1.7 × 10⁷ and 2.9 × 10⁷ reads (about 100 bp per read). And the mapping rates were 62%, 55% and 54%, respectively. The expression levels of FPKM values were shown in Table S1.† Differential gene expression diversity controls fatty acid secretion during development from 30 h to 42 h. Our results now show that high expression levels of AoACBP can promote β-oxidation and metabolic flexibility to help maintain the fatty acids balance. Importantly, the biosynthesis of the long-chain fatty acids, sterols and phospholipids were relatively inhibited (Fig. 5). The complete list of lipid metabolic reactions and the corresponding enzymes is found in Table S2.†

Fig. 5 The pathways of lipid metabolism. Those genes were expressed at a continued greater level and lower level in A. oryzae are indicated by yellow triangles and red circles respectively. The gene expression levels were listed in Table S1.† The numbers were on behalf of the metabolic reactions which were listed in Table S2.†

Discussion

Fatty acid biosynthesis and metabolism in microorganism is very important for fermented food industry, especially for the fat conversion of the raw materials.²¹ Lipodieresis has beneficial effects on flavor quality during the fermentation, and can also increase its costs.^22,23 The flavor of fermented food was significantly influenced by the lipid contents in fermentation. Fatty acids, the fundamental ingredient of soy sauce, is an important contributor for aroma in fermented food. Kinds of fatty acids were formed and then transferred to some volatile substances. ACBP is known to be an essential regulator in intracellular acyl-CoA transport during fatty acid synthesis and β-oxidation. The function of ACBP is critical to lipid metabolism and relevant cell signaling of eukaryotic cells.^24,25

The results presented in this study demonstrate the binding preferences of fatty acyl-CoAs and the influence of the lipid metabolism in A. oryzae. We cloned and successfully expressed MBP-tagged AoACBP in E. coli. Subsequently, MBP-AoACBP was proved to have a preference for binding C16:0-CoA, C18:0-CoA, C20:0-CoA and C22:0-CoA by the MST assay. Palmitic acid (C16:0) and stearinic acid (C18:0) were detected to be the major long-chain fatty acids with high content in A. oryzae. The preference of C16:0-CoA and C18:0-CoA may indicate that C16:0 and C18:0 play an active role in fatty acid metabolism in A. oryzae. It is well documented that lipid metabolism plays an essential role in the development of many groups of organisms, including the filamentous fungi.^26–28

We were also interested in whether AoACBP participates in the growth and development of A. oryzae. Expression levels by Western blot, RT-PCR and distribution patterns by immunofluorescence showed that AoACBP fluorescence molecules accumulated on the old mycelium and asexual spores in the later growth stage, which may indicate that long-chain fatty acids were preserved for the next generation. Several studies of filamentous fungi have suggested a role between polyunsaturated fatty acids and spore formation.²⁹

Most microbial regulatory networks play important roles in adjusting their concentrations of membrane components. This change was regulated by metabolites including lipids, which promote their biosynthesis, degradation, export and storage or inhibit some biosynthetic and degrading pathways. LCAs are always recognized as critical regulators of signaling pathways. The synthesis of LCAs is carried by acyl-CoA syntheses. It may increase the β-oxidation of fatty acids, TAG synthesis, and alter phospholipid fatty acid composition,³⁰ or even regulate transcription.³¹ Long-chain fatty acids are activated on the cytosolic side of the mitochondrial outer membrane, and transported into the mitochondrial matrix by carnitine palmitoyl transferase for β-oxidation (Fig. 5).³² The reaction from sn-glycerol-3-phosphate to 1,2-diacyl-sn-glycerol-3-phosphate was activated after the accumulation of LCAs. And then glycerol-3-phosphate acyltransferase was proposed to promote TAG synthesis. The expression levels of the enzymes for the synthesis of phospholipids varied unstable at these growth stages. As shown in Fig. 5, sterol biosynthesis was inhibited under certain conditions.

In summary, our study presented new insights on the binding affinity and specificity of AoACBP in fatty acyl-CoA utilization. It also shows that the important relationship between the distribution pattern and expression levels of AoACBP. The information significantly increased our knowledge about ACBP functions, and the synthesis of fatty acids. The microbial fatty acid biosynthesis, transferring and degradation should be further explored for its industrial use in traditional fermented foods.

Conflict of interest

The authors of this manuscript have no conflicts of interest to declare.

Ethical approval

The protocols of the study were approved by the Ethics Committee of Jiangnan University, China (JN no. 52014) and procedures were carried out in strict accordance with European Community guidelines (Directive 2010/63/EU) for the care and use of experimental animals.

Acknowledgements

This study was financially supported by these projects in China (31171731, 31460447, 31401682, 20142BDH80003, 2013-CXTD002, 31401682, BK20140146, 3000035402, 00001384, 30000411, 300098020110, 300098030105, 2014QNBJRC0010, 2015T80498, “555 talent project” of Jiangxi Province), Jiangxi Provincial Key Laboratory of Bioprocess Engineering and Co-Innovation Center for In Vitro Diagnostic Reagents and Devices of Jiangxi Province.

References

1. M. Machida, O. Yamada and K. Gomi, DNA Res., 2008, 15, 173–183.
2. K.-J. Hong, C.-H. Lee and S. W. Kim, J. Med. Food, 2004, 7, 430–435.
3. Y. Fukushima, H. Itoh, T. Fukase and H. Motai, Appl. Microbiol. Biotechnol., 1991, 34, 586–590.
4. P. M. Kris-Etherton, D. S. Taylor, S. Yu-Poth, P. Huth, K. Moriarty, V. Fishell, R. L. Hargrove, G. Zhao and T. D. Etherton, Am. J. Clin. Nutr., 2000, 71, 179S–188S.
5. Y. Kokubo, H. Iso, J. Ishihara, K. Okada, M. Inoue, S. Tsugane and J. S. Group, Circulation, 2007, 116, 2553–2562.
6. T. Kikuma, M. Ohneda, M. Arioka and K. Kitamoto, Eukaryotic Cell, 2006, 5, 1328–1336.
7. B. B. Kragelund, J. Knudsen and F. M. Poulsen, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids, 1999, 1441, 150–161.
8. L. Lee, C. A. DeBono, D. R. Campagna, D. C. Young, D. B. Moody and M. D. Fleming, J. Invest. Dermatol., 2007, 127, 16–23.
9. J. Knudsen, M. Burton and N. Faergeman, Adv. Mol. Cell Biol., 2004, 33, 123–152.
10. J. Knudsen, T. B. Neergaard, B. Gaigg, M. V. Jensen and J. K. Hansen, J. Nutr., 2000, 130, 294S–298S.
11. C. DiRusso, T. L. Heimert and A. K. Metzger, J. Biol. Chem., 1992, 267, 8685–8691.
12. N. Raman and C. C. DiRusso, J. Biol. Chem., 1995, 270, 1092–1097.
13. J. T. Rasmussen, N. J. FÆrgeman, K. Kristiansen and J. Knudsen, Biochem. J., 1994, 299, 165–170.
14. K. G. Milne and M. A. Ferguson, J. Biol. Chem., 2000, 275, 12503–12508.
15. J. T. Rasmussen, T. Börchers and J. Knudsen, Biochem. J., 1990, 265, 849–855.
16. Q. Hao, X. Liu, G. Zhao, L. Jiang, M. Li and B. Zeng, Biotechnol. Lett., 2015, 1–7.
17. J. L. Parker and S. Newstead, Nature, 2014, 507, 68–72.
18. L. Wang, W. Chen, Y. Feng, Y. Ren, Z. Gu, H. Chen, H. Wang, M. J. Thomas, B. Zhang and I. M. Berquin, PLoS One, 2011, 6, e28319.
19. P. Melin, J. Schnürer and E. Wagner, Mol. Genet. Genomics, 2002, 267, 695–702.
20. N. Osherov and G. May, Fungal Genet. Newsl., 1998, 38–40.
21. C. Siciliano, E. Belsito, R. De Marco, M. L. Di Gioia, A. Leggio and A. Liguori, Food Chem., 2013, 136, 546–554.
22. J.-X. Guinard, C. Wee, A. McSunas and D. Fritter, Food Qual. Prefer., 2002, 13, 129–137.
23. N. Asuming-Bediako, M. Jaspal, K. Hallett, J. Bayntun, A. Baker and P. Sheard, Meat Sci., 2014, 96, 187–194.
24. N. J. Faergeman and J. Knudsen, Biochem. J., 1997, 323, 1–12.
25. J. Knudsen, M. V. Jensen, J. Nils, T. B. Neergaard and B. Gaigg, Mol. Cell. Biochem., 1999, 192, 95–103.
26. A. M. Calvo, L. L. Hinze, H. W. Gardner and N. P. Keller, Appl. Environ. Microbiol., 1999, 65, 3668–3673.
27. Z.-Y. Wang, D. M. Soanes, M. J. Kershaw and N. J. Talbot, Mol. Plant-Microbe Interact., 2007, 20, 475–491.
28. Y. Oh, N. Donofrio, H. Pan, S. Coughlan, D. E. Brown, S. Meng, T. Mitchell and R. A. Dean, Genome Biol., 2008, 9, 1–24.
29. A. M. Calvo, H. W. Gardner and N. P. Keller, J. Biol. Chem., 2001, 276, 25766–25774.
30. L. O. Li, J. M. Ellis, H. A. Paich, S. L. Wang, N. Gong, G. Altshuller, R. J. Thresher, T. R. Koves, S. M. Watkins, D. M. Muoio, G. W. Cline, G. I. Shulman and R. A. Coleman, J. Biol. Chem., 2009, 284, 27816–27826.
31. S. Y. Bu, M. T. Mashek and D. G. Mashek, J. Biol. Chem., 2009, 284, 30474–30483.
32. D. Neess, S. Bek, H. Engelsby, S. F. Gallego and N. J. Faergeman, Prog. Lipid Res., 2015, 59, 1–25.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20532a

---