Paenibacillus pabuli US132 strain, previously isolated [18], was used as source of chromosomal DNA. E. coli DH5α (F⁻supE44 ϕ80 δlacZΔ M15 Δ(lacZYA-argF) U169 endA1 recA1 hsdR17 (r_k⁻, m_k⁺) deoR thi-1 λ⁻ gyrA96 relA1) was used for all plasmid constructions. E. coli ER2566 (F⁻ λ⁻fhu A2 [lon] ompT lacZ:T7 gene1 gal sulA11 Δ(mcrC-mrr)114:IS10 R(mcr-73:miniTn10-TetS)2 R(zgb-210:Tn10) (TetS) endA1 [dcm]) and E. coli JM109 (F′traD36 proA⁺B⁺ lacI^qz ΔM15/Δ(lac-proAB e14⁻ (McrA⁻) recA1 endA1 gyrA96 thi-1 hsdR17 (r_k⁻m_k⁺) supE44 relA1) were used for protein expression. The plasmids pSJ8 and pSJ9 (this work) carrying the cgtase US132 gene, derived from pCR2.1 (Invitrogen) and pTrc99a vector [19], respectively. E. coli recombinant strains were grown on LB, M9, 2TY [20], or M9ZB medium [16] containing 100 μg/mL ampicillin. Paenibacillus pabuli US132 strain was grown as previously described [18].

To amplify an internal cgtase gene fragment, we used the highly conserved aa regions previously reported for CGTase enzymes as members of the α-amylase family [4, 21]. Furthermore, we aligned several cgtases DNA sequences and we located the corresponding CGTases conserved regions. Hence, a degenerate pair of primers, S209 (5′-GAYTTTGCRCCCAAYCAT-3′) and S211 (5′-ATCATGATTGTCRATRAA-3′), corresponding respectively to the conserved regions I (DFAPNH) and IV (FIDNHD), were designed. The chromosomal DNA isolated from Paenibacillus pabuli US132 strain was used as template for gene amplification by Pfu DNA polymerase (Fermentas). The PCR parameters were 94°C for 120 seconds followed by 35 cycles of 94°C for 30 seconds, 50°C for 45 seconds, and 72°C for 120 seconds. The PCR product was purified using Wizard SV Gel and PCR Clean-Up System (Promega). To amplify the entire US132 cgtase gene, we have conceived two primers, S260 (5′-TCCAATATTTCTTACGAT-3′) and S262 (5′-GTTACGTTACTCCGGG-3′), flanking the closely matching cgtase gene (accession no. L25256). The resulting PCR fragment was cloned into the pCR2.1 vector to give pSJ8 plasmid harboring the US132 cgtase gene under the control of the IPTG inducible T7 promoter. The pSJ8 was digested with HindIII and EcoRV and the resultant cgtase gene was subcloned in the pTrc99a, under the control of the Trc promoter (IPTG inducible), to give pSJ9 plasmid. Nucleotide sequences of the US132 cgtase gene carried by three independent pSJ8 plasmids obtained from different PCR reactions were determined. DNA sequencing was carried out using the BigDye Terminator v3.1 Cycle Sequencing Kit and the automated ABI Prism3100-Avant Genetic Analyser (Applied Biosystems). Sequence analysis and comparison were performed with BioEdit and CLUSTAL W programs. Homology search was performed using BLAST search algorithm. For the promoter determination, the prokaryotic promoter prediction program NNPP2.2A (http://www.fruitfly.org/cgi-bin/seqtools/promoter.pl) was used. The US132 cgtase sequence was submitted to the EMBL data bank under Accession no. AM748796.

The heterologous expression of the US132 CGTase was investigated using different recombinant strains (DH5α/pSJ8, ER2566/pSJ8, DH5α/pSJ9, ER2566/pSJ9 or JM109/pSJ9), various temperature cultivation (37°C or 19°C during all operating time and 37°C followed by a shift to 19°C after IPTG induction) and different medium composition (LB, M9, 2TY or M9ZB). An overnight culture of the studied strain, cultivated at 37°C, was used to inoculate 50 mL of basal medium (in 250 mL shake flask) with an initial OD_{600 nm} of 0.1. Protein expression was induced by addition of different IPTG concentration (16–240 μg/mL) at OD_{600 nm} between 0.8 and 1. Effects of the postinduction time on the US132 CGTase production were also investigated. All experiments were performed at least twice.

Batch fermentation was carried out in a 7 L fermentor INFORS AG CH-4103 (Bottmingen, Switzerland) containing 4.5 L of 2TY medium. The fermentor was inoculated, with an initial OD_{600 nm} of 0.1, by an overnight culture of the ER2566/pSJ8 strain grown at 37°C in 2TY broth. The pH, aeration and agitation were maintained constant at 7.4, 1.5 vvm and 500 rpm, respectively, during all cultivation. The temperature was shifted from 37 to 19°C after induction with 16 mg/L of IPTG when OD_{600 nm} reached 0.8–1.

The CGTase extract was prepared from the periplasm by the modified osmotic-shock procedure of Ausubel et al. [22]. Cells were harvested by centrifugation at 8000 × g for 10 minutes and the pellets were suspended in 30 mM Tris-HCl buffer (pH 8) containing 20% sucrose and 1 mM EDTA. After agitation for 10 minutes at 25°C, cells removed by centrifugation were rapidly suspended in ice-cold water, incubated for 10 minutes at 0°C and clarified by centrifugation at 12000 × g for 30 minutes. The supernatant was recovered as periplasmic fraction and used for CGTase assay. Residual cells were sonicated at 4°C for 6 minutes (pulsations 3 s, amplify 90) using a Vibra-Cell 72405 Sonicator and insoluble fraction (cells debris) was recuperated by centrifugation at 12000 × g for 30 minutes.

For the purification of the recombinant US132 CGTase, the ER2566/pSJ8 periplasmic fraction was heat-treated at 60°C for 15 minutes in presence of 10 mM calcium followed by centrifugation at 16000 × g for 30 minutes at 4°C. The supernatant was then purified using hydrophobic interaction chromatography and starch adsorption as previously described for the native enzyme [18].

The CGTase activity, determined as dextrinisation activity, was monitored at 60°C for 10 min as described in a previous work [18].

The kinetic parameters (K_m and V_max) were determined by incubating the purified CGTase at 60°C in 50 mM sodium acetate buffer (pH 6.5) and using various concentrations of soluble starch ranging from 0.4 to 1.6 g/L. Samples were taken at regular time intervals and the reaction was stopped by boiling for 5 minutes. The amount of generated β-CDs was detected according to a phenolphthalein method [23]. The initial speed was calculated for each substrate concentration and then represented according to the Lineweaver-Burk method [24]. The K_m and V_max were determined graphically using Hyper32 program (http://homepage.ntlworld.com/john.easterby/hyper32.html).

To clone the US132 cgtase gene, an internal PCR fragment of about 600 bp was firstly amplified using US132 genomic DNA and two primers corresponding to the conserved regions I and IV in CGTases as described in Materials and Methods section. The nucleotide sequence of this internal fragment (582 bp) showed the highest homology (95% identity) with homologous region of the putative Bacillus Q cgtase gene (accession no. L25256). This result incited us to conceive two primers from the 5′ and 3′ flanking regions of the Bacillus Q cgtase gene in order to amplify the entire US132 cgtase gene using the chromosomal DNA of P. pabuli US132 as template. The PCR amplification, using Pfu DNA polymerase, gave a nucleotide fragment of approximately 2400 bp. The cloning of this fragment in pCR2.1 vector, under the control of the T7 promoter, provided pSJ8 plasmid and conferred starch degrading activity for E. coli host strains.

The nucleotide sequence analysis of the fragment containing the US132 cgtase gene revealed the presence of a single open reading frame (ORF) with two potential initiation ATG codons (Figure 1). A putative Shine-Dalgarno site (5′-AGAAGGGTGG-3′), exhibiting a good rationally complementarity with the 3′ end of the US132 16S rRNA gene [18], was 7 bases upstream the first ATG codon. Consequently, this latter is most likely the true initiator codon. Thus, the US132 cgtase ORF, consisted of 2157 bp, encoded a protein having 718 amino acids. By using the promoter prediction program, a potential −10 region (5′-TCTGCA-3′) and −35 region (5′-TTGGCG-3′) were found at 101 and 124 bp upstream the initiator codon, respectively. The first 34 amino acids are considered as signal peptide according to the SignalP server (http:/www.cbs.dtu.dk/services/signalP). Therefore, the mature enzyme was consisted of 684 amino acids with an estimated molecular mass of about 74.3 kDa.

The alignment of the mature US132 CGTase sequence to the data bank showed 93, 92, 74, and 73% identity with the CGTases of Bacillus sp. (accession no. CAA46901), Bacillus licheniformis (accession no. CAA33763), Bacillus circulans strain 251 (accession no. CAA55023), and Bacillus sp.1011 (accession no. 1I75A), respectively. The highest homology (97% similarity and 94% identity) was found with the CGTase of Bacillus circulans strain no. 8 (accession no. 1CGT). The alignment showed a difference of 39/684 amino acids which does not radically affect the product specificity. Indeed, both enzymes are specific for the β-CD production and are classified as β-CGTases since they generated a mixture of CDs composed of α-, β-, and γ-CD 20 : 58 : 13 and 10 : 64 : 20 ratios for the CGTase of strain US132 and B. circulans strain no. 8 [25], respectively. However, as far as we know, no data was available concerning biochemical properties of the CGTase of B. circulans strain no. 8 (pH, thermoactivity, thermostability) allowing the comparison of these two enzymes.

From the multiple sequence alignment (Figure 2), the five structural domains (A, B, C, D, and E) [11, 26] could be identified in the US132 CGTase. Furthermore, four highly conserved regions labelled I–IV and located in the catalytic domain were found. These conserved regions would constitute the active center of the enzyme since it contained the three catalytic residues Asp229, Glu257, and Asp328 referring to the CGTase of B. circulans strain no. 8 numbering. The conserved amino acids in the acceptor binding site, namely Lys47, Tyr89, Asn94, Phe183, Asn193, Leu194, Tyr195, Asp196, Phe259, Phe283, and Asp371, reported as the main determinants for cyclization reaction [12, 27], were also identified.

The purification of the US132 CGTase from the native strain (Paenibacillus pabuli) was extremely hampered by the viscosity of the crude supernatant, probably due to the presence of polysaccharides of unknown nature (data not shown). In the aim to overcome this problem and also to overproduce the US132 CGTase, we have tested the enzyme production from the initial recombinant E. coli strain (DH5α/pSJ8). This study showed that this latter strain, cultivated in LB medium at 37°C and induced by 16 μg/mL IPTG for 18 hours, exhibited a very low activity of about only 0.3 U/mL. For this reason, we have studied the effect of several parameters on the production enhancement of the active recombinant CGTase.

For the purification of the recombinant US132 CGTase, we have used the periplasmic fraction of the ER2566/pSJ8 strain cultivated in the fermentor in the cultivation conditions described above. This periplasmic fraction was firstly heat-treated at 60°C in order to remove E. coli thermolabile proteins. The extract was then purified using hydrophobic interaction chromatography (HIC) and starch adsorption as described for the native enzyme [18]. The recombinant enzyme was purified to homogeneity 25-fold with a yield of 20% (Table 3, Figure 6). This recombinant US132 CGTase retained the same biochemical properties (data not shown) as the native enzyme (thermoactivity, thermostability, pH stability, CDs production) [18].

The kinetic parameters were determined by incubating the purified recombinant US132 CGTase in presence of various soluble starch concentrations and using the Lineweaver-Burk method. The V_max and K_m values were evaluated to be 253 ± 36 μmol of β-cyclodextrin/mg/min and 0.36 ± 0.18 g/L, respectively. K_m values determined, using also soluble starch as substrate, for CGTase from Bacillus agaradhaerens [33], Bacillus circulans E192 [34], Bacillus firmus [35], and Bacillus sp. TS1-1 [11] were 21.2, 5.7, 1.21, and 0.52 g/L, respectively. Since a small value of K_m exhibited a high affinity for the substrate, the values shown above suggested that US132 CGTase was more specific towards starch than the other reported CGTases.