Acyl-ACP thioesterases from Camelina sativa: Cloning, enzymatic characterization and implication in seed oil fatty acid composition Author Rodríguez-Rodríguez, Manuel Fernando Author Salas, Joaquín J. Author Garcés, Rafael Author Martínez-Force, Enrique text Phytochemistry 2014 2014-11-30 107 7 15 http://dx.doi.org/10.1016/j.phytochem.2014.08.014 journal article 10.1016/j.phytochem.2014.08.014 1873-3700 10491218 2.2. Genomic organization of C. sativa acyl-ACP thioesterase genes To analyze the genomic organization of the CsFatA and CsFatB genes, two genomic DNA fragments at the locus were amplified using two different primer pairs. Clones of 1214 and 1427 nucleotides were obtained and sequenced for CsFatA and CsFatB , respectively. The intron and exon organization of the three CsFatA and CsFatB alleles were found by comparing their cDNA and genomic DNA sequences (Supplementary data, Table 2 ). The CsFatA1 allele was 1743 bp long, the CsFatA2 allele was 1745 bp and the CsFatA3 allele was 1773 bp. All CsFatA alleles had six introns and thus, they each contained seven exons, which were of similar length (Supplementary data, Table 2 ). Intron 3 of the CsFatA3 allele differed most in length as it contained an insertion of 32 nucleotides at the beginning of the sequence (169 bp in i3CsFatA1; 171 bp in i3CsFat-A2; and 200 bp in i3CsFatA3: Supplementary data, Table 2 ). The CsFatB1 allele was 1790 bp long, the CsFatB2 allele 1872 bp and the CsFatB3 allele 1870 bp. All have four introns and five exons with a high degree of identity and homology between them, except for introns 2 and 3 that display significant differences in both sequence and length (221 bp for i2CsFatB1; 305 bp for i2CsFatB2; 314 bp for i2CsFatB3; 139 bp for i3CsFatB1; 141 bp for i3CsFatB2; and 131 bp for i3CsFatB3: Supplementary data, Table 2 ). Fig. 1. Alignment of the deduced amino acid sequences of acyl-ACP thioesterase A enzymes from C. sativa ( Cs FatA1, AFQ60947.1; Cs FatA2, AFQ60948.1; Cs FatA3, AFQ60946.1), Arabidopsis thaliana ( At FatA1, NP_189147.1; At FatA2, NP_193041.1) and Zea mays ( Zm FatA, DAA40472.1). Identical amino acids are shaded in black, whereas conserved residues are shaded in grey. The amino acids considered to constitute the signal peptide are boxed. The three conserved residues that constitute the catalytic triad are indicated with an star (Asn-273; His-275; Gln-311), and the residues involved in specific substrate recognition and related with the thioesterase activity are indicated by the arrowheads (Gly-135; Ala-137; Arg-143; Lys-144; Thr-182; Arg-184; Arg-212; Arg-213; Lys-216). The conservative changes in the amino acid sequence between the three Cs FatA alleles are marked by closed circles and the semi-conservative changes between them by open circles. The genetic map of C. sativa , comprised of 157 amplified fragment length polymorphisms (AFLPs) and 3 single sequence repeat markers were published in 2006 ( Gehringer et al., 2006 ). This study showed that C. sativa has 20 chromosomes, a figure found only among known alloploids or plant species. A triplication of the C. sativa genome might have resulted from two allopolyploidy events, first resulting in tetraploidy (4n) and then hexaploidy (6n), or it could also be derived from the combination of an autotetraploid (4n) and diploid (2n) species in an autopolyploidized (6n) genome ( Hutcheon et al., 2010 ). The alignment of our intron sequences with the recently available sequences from the C. sativa Genome Project (http://www.camelinadb.ca/), which is included in The Prairie Gold Project, allowed the different alleles of the thioesterase genes to be located in the Camelina genome: CsFatA1 , CsFatA2 and CsFatA3 on chromosomes 15, 19 and 1, respectively; and CsFatB1 , CsFatB2 and CsFatB3 on chromosomes 14, 17 and 3, respectively. Analyzing the intron sequences suggests the existence of two groups of sequences for each thioesterase. Thus, s CsFatA1 and CsFatA2 presented strong identity with obvious differences from CsFatA3 . Similarly the CsFatB3 sequences showed much more variability with respect to the other two alleles, corresponding to the complementation group of chromosomes previously reported ( Gehringer et al., 2006 ). These results are more consistent with the Camelina genome being autopolyploid due to the combination of an autotetraploid and a diploid species. The diploid parent could have contributed to the C. sativa genome with two possible combinations, 7 + 7 + 6 or 6 + 6 + 8 giving the total of 20 chromosomes. 2.3. Fatty acid analysis of Escherichia coli expressing C. sativa acyl-ACP thioesterases Mature Cs FatA and Cs FatB proteins were overexpressed in E. coli after removing the hydrophobic domain of Cs FatB. The hydrophobic domain of these enzymes is often removed for expression in bacteria to increase their concentration in the soluble phase ( Jones et al., 1995 ; Facciotti and Yuan, 1998 ). Only one allele of each acyl-ACP thioesterase type , CsFatA and CsFatB , was cloned into the pQE-80 L vector because the Cs FatA alleles only have three differences in their amino acid sequence, one conservative (Asp-167- Arg) and two semi-conservative changes (Ser-128-Gly; Asp-294- His), and the different CsFatB alleles have four purely conservative changes (Leu-134-Ile; Val-305-Phe; Lys-343-Arg; Ser-370-Ala). The amino acid residues involved in substrate recognition and those related to the hydrolase activity of Cs FatA and Cs FatB were identical in the different alleles ( Figs. 1 and 2 ) . Fig. 2. Alignment of the deduced amino acid sequences of the acyl-ACP thioesterase B enzymes from C. sativa ( Cs FatB1, AFQ60949.1; Cs FatB2, AFQ60950.1; Cs FatB3, AFQ60951.1), Arabidopsis thaliana ( At FatB, CAA85388.1) and Zea mays ( Zm FatB, AFW85914.1). Identical amino acids are shaded in black, whereas conserved residues are shaded in grey. The amino acids considered to constitute the signal peptide are boxed and the hydrophobic region in FatB is underlined. The three conserved residues that constitute the catalytic triad are indicated by an star (Asn-319; His-321; Cys-383), and the residues involved in the specific substrate recognition and related with the thioesterase activity are indicated by the arrowheads (Asp-186; Gly-189; Met-233; Arg-235; Lys-267; Glu-270). The conservative changes in the amino acid sequence between the three CsFatB alleles are indicated by open circles. In plants, the substrate specificity of thioesterases determines the oil composition because these enzymes are involved in the export of acyl-ACP from the plastid to cytosol. In E. coli , thioesterases cleave acyl-ACPs producing the free fatty acids necessary for regulatory signals, export or degradation ( Lennen and Pfleger, 2012 ). The fatty acid composition of E. coli expressing Camelina acyl-ACP thioesterase genes were analyzed and compared with control cells transformed with the empty pQE-80L vector ( Table 1 ). The expression of Cs FatA and Cs FatB produced a 45% and 68% decrease in the total fatty acid content of E. coli , respectively. These results showed that C. sativa thioesterases alter E. coli fatty acid metabolism, diverting the acyl chains away from the fatty acid and lipid biosynthetic pathways. These free fatty acids would later be secreted or degraded in the β- oxidation pathway ( Lennen and Pfleger, 2012 ). The main change in the E. coli fatty acid composition when Cs FatA was expressed was the reduction in unsaturated fatty acids, mainly cis-vaccenic acid (18:1ω7). This reduction was compensated for by an increase of palmitoleic acid (16:1ω7). However, the expression of Cs FatB caused the opposite effect, a decrease in saturated fatty acids and in particular, that of palmitic acid (16:0) that is compensated for with an increase in stearic acid (18.0). The effect of the expression of acyl-ACP thioesterases on E. coli depends on the phenotype of the recipient strain. Thus, regular E. coli strains like the Bluescript one used in this work experiment a diminution of the fatty acid content, and a lower proportion of the fatty acids hydrolyzed by the enzyme ( Voelker and Davies, 1994 ; Sánchez-García et al., 2010 ). This is caused because the hydrolyzed fatty acids are degraded and recycled via β- oxidation. On the contrary, when thioesterases are expressed in strains deficient on fatty acid activation or degradation, as it is the case of FadD88, the fatty acids hydrolyzed (16:0 or 18:1) are accumulated or excreted in the culture medium ( Huynh et al., 2002 ). 2.4. Substrate specificity and kinetic parameters of Camelina acyl-ACP thioesterases The kinetic parameters of recombinant Cs FatA and Cs FatB were investigated after purification by metal ion affinity chromatography (IMAC). This method allowed us to obtain highly purified enzymes in a single step (see Fig. 3 ). The substrate specificity of the Cs FatA and Cs FatB enzymes was determined by assaying their activity on different acyl-ACP substrates at a constant concentration ( Fig. 4 ). The Cs FatA enzyme displayed a high level of activity on unsaturated fatty acid derivatives, mainly with 18:1-ACP, and it was 14-fold less active towards 18:0-ACP. Cs FatB had preference for 16:0-ACP with lower activities towards 18:0-ACP and 18:1- ACP. These results are similar to those reported previously for thioesterases from other plants, such Garcinia mangostana ( Hawkins and Kridl, 1998 ) , Carthamus tinctorius ( Knutzon et al., 1992 ) , Brassica campestri ( Pathak et al., 2004 ) , A. thaliana ( Salas and Ohlrogge, 2002 ) , H. annuus ( Serrano-Vega et al., 2005 ) and R. communis ( Sánchez-García et al., 2010 ) . Table 1 Fatty acid composition of E. coli cells containing recombinant plasmid. The date are the averages of 3 independent samples.

16:0	16:1ω7	17:0Δ a	18:0	18:1ω7	19:0Δ b	UFA c	STA d	UFA/STA	µg FAs/mg cell
pQE80	52.6 ± 0.6	9.9 ± 0.5	21.3 ± 0.4	1.8 ± 0.4	11.2 ± 0.2	3.2 ± 0.1	45.58 ± 0.97	54.42 ± 0.97	0.84 ± 0.15	6.24 ± 0.90
pQE80_CsFatA	55.9 ± 1.3	21.7 ± 0.7	13.1 ± 1.3	3.7 ± 0.5	5.2 ± 0.1	0.5 ± 0.1	40.36 ± 0.73	59.64 ± 0.73	0.68 ± 0.01	3.49 ± 0.06
pQE80_CsFatB	43.1 ± 1.8	29.3 ± 0.7	7.1 ± 0.9	7.0 ± 1.0	12.5 ± 0.7	1.1 ± 0.1	49.89 ± 0.74	50.11 ± 0.94	1.00 ± 0.0	2.01 ± 0.38

a cis-9,10-Methylen-hexadecanoic acid, cyclopropane derivative from 16:1. b cis-11,12-Methylen-octadecanoic acid, cyclopropane derivative from 18:1. c Unsaturated fatty acids and derivatives; 16:1ω7 + 17:0Δ + 18:1ω11 + 19:0Δ. d Saturated fatty acids; 16:0 + 18:0. Kinetic parameters were also calculated for both enzymes acting on different substrates, displaying similar Km values for all of them, all in the micromolar order. These values were slightly higher than those reported previously for acyl-ACP thioesterases from H. annuus ( Serrano-Vega et al., 2005 ) , R. communis ( Sánchez-García et al., 2010 ) or M. tetraphylla ( Moreno-Pérez et al., 2011 ) . The V max of Cs FatA for 18:1-ACP was 85.3 nkat/mg prot, one order of magnitude higher the V max found for the other substrates assayed. The K cat and catalytic efficiency ( K cat / Km ) values were also highest for this substrate ( Table 2 ), which is in good agreement with the kinetic parameters described for most FatAs investigated to date. Similarly, Cs FatB displays a typical profile of FatB enzymes, showing greater catalytic efficiency towards 16:0- ACP, which displayed a V max value that was 5-fold higher than that for 18:1-ACP. These V max and K cat values were lower than those described by Sánchez-García et al. (2010) for R. communis , yet they were higher than those reported for M. tetraphylla ( Moreno-Pérez et al., 2011 ) . Fig. 4. Substrate specificity of C. sativa acyl-ACP thioesterases expressed in E. coli . The activity was measured with the purified His-tagged Cs FatA (black columns) and His-tagged Cs FatB (white columns) enzymes, testing different acyl-ACP substrates. The data represent the mean (±SD) from three independent assays. Fig. 3. Coomassie blue stained SDS–PAGE showing recombinant C. sativa acyl-ACP thioesterase A (panel A) and acyl-ACP thioesterase B (panel B). Lane 1, soluble fraction, 15 µg protein; lane 2, soluble fraction not retained on the Ni–NTA Agarose column (Qiagen), 15 µg protein; lane 3, Ni–NTA Agarose wash, 1.5 µg protein; lane 4, purified acylACP thioesterase 1.5 µg protein. Table 2 Kinetic parameters of the purified recombinant Cs FatA and Cs FatB proteins acting on different acyl-ACP substrate.

Substrate	Km (µM)	V max (nkat/mg prot)	K (s ― 1) cat	K / K( s ― 1 µM ― 1) cat m
Cs FatA	16:0-ACP	1.5	3.8	0.3	0.2
16:1-ACP	0.5	1.6	0.1	0.3
18:0-ACP	0.5	4.8	0.4	0.75
18:1-ACP	3.9	85.3	7.0	1.8
Cs FatB	16:0-ACP	6.6	68.1	4.4	0.6
16:1-ACP	2.2	0.8	0.06	0.03
18:0-ACP	3.9	7.2	0.6	0.2
18:1-ACP	3.1	12.8	1.0	0.3

Data represent the mean of 3 independent determinations; SD <10% of mean value. Fig. 5. Expression of the CsFatA (black columns) and CsFatB (white columns) genes in vegetative tissues and developing seeds from C. sativa determined by QRT-PCR. The data represent the mean values ± SD of three independent assays. 2.5. Expression profiles of C. sativa acyl-ACP thioesterases The expression of the acyl-ACP thioesterase genes in developing seeds and vegetative tissues of C. sativa was studied by quantitative real time PCR (QRT-PCR). The profile of transcript accumulation was temporally regulated during the development of the embryo ( Fig. 5 ), with the strongest expression of the CsFatA and CsFatB genes occurring in developing seeds at 12, 18 and 24 days after flowering (DAF), the phase of oil accumulation ( He et al., 2004 ). In seed tissue, CsFatA was always expressed more strongly than CsFatB , which fits well with the composition of Camelina oil in which linoleic and linolenic acids predominate, fatty acids derived from oleic acid that is mainly exported via FatA. The expression of these genes is significantly weaker in vegetative tissue, with the exception of CsFatB in leaves that could be involved in the production of saturated fatty acids used for surface lipid biosynthesis. Indeed, stronger expression of CsFatB in leaves was also reported in R. communis ( Sánchez-García et al., 2010 ) . Nevertheless, the expression patterns observed in this analysis suggests that C. sativa acyl-ACP thioesterases are important for oil deposition in the seed. 3. Conclusions The cloning and sequencing of CsFatA and CsFatB thioesterases from developing Camelina seeds shows that they are encoded by a single copy gene, three different alleles existing of each gene. In both cases, the differences found in the coding region between these alleles are not important, accumulating mostly single nucleotide polymorphisms (SNP), insertions and deletions in the introns. Indeed, the highly conserved papain-like catalytic triad, asparagine, histidine and glutamine, are maintained in Cs FatA and Cs FatB. The heterologous expression of these enzymes in E. coli produced a contrasting effect on bacterial fatty acid composition, Cs FatA causing a decrease in the unsaturated fatty acids, mainly 18:1ω7, and Cs FatB augmenting these fatty acids. The substrate specificity of these enzymes is similar to that reported previously in other plants, Cs FatA showing a strong preference for 18:1-ACP and Cs FatB for 16:0-ACP. The kinetic parameters of both enzymes differ only slightly from those described in H. annuus , R. communis or M. tetraphylla .