Characterization of Atpase Activity of the Aaa Arc from Bifidobacterium Longum Subsp. Infantis

Bifidobacteria are considered to be probiotics that exist in the large intestine and are helpful to maintain human health. Oral administration of bifidobacte-ria may be effective in improving the intestinal flora and environment, stimulating the immune response and possibly preventing cancer. However, for consistent and positive results, further well-controlled studies are urgently needed to describe the basic mechanisms of this microorganism. Analysis of the proteas-ome-lacking Bifidobacterium longum genome reveals that it possesses a gene, IPR003593 AAA ATPase core, which codes a 56 kDa protein containing one AAA ATPase domain. Phylogenetic classification made by CLANS, positioned this sequence into the ARC divergent branch of the AAA ATPase family of proteins. N-terminal analysis of the sequence indicates this protein is closely related to other ATPases such as the Rhodococcus erythropolis ARC, Archaeoglobus fulgidus PAN, Mycobacterium tuberculosis Mpa and the human proteasomal Rpt1 subunit. This gene was cloned, the full-length recombinant protein was overexpressed in Escherichia coli, purified as a high-molecular size complex and named Bl-ARC. Enzymatic characterization showed that Bl-ARC ATPase is active, Mg +2-dependent and sensitive to N-ethylmaleimide. Gene organization positions bl-arc in a region flanked by a cluster of genes that includes pup, dop and pafA genes. These findings point to a possible function as a chaperone in the degradation pathway via pupylation. INTRODUCTION Lactic acid bacterium Bifidobacterium longum was isolated and described over a century ago from human infant feces and was associated with a healthy intestinal tract in humans (Tissier, 1900). This flora protects the intestinal tract from proliferation or infection by harmful bacteria; recent research has focused on bifidobacteria to establish the importance of these organisms in influencing certain normal functions of the intestinal tract and in exploring their role in the human health. Evidence of the previous observations was described in a report where dietary administration of lyophilized cultures of B. longum suppressed colon and mamma-ry carcinogenesis in laboratory animal models (Reddy, 1999), while another describes that bifidobacteria have effects on the treatment of rotavirus diarrhea (Marteau et al., 1990). Daily oral administration of bifidobacteria, oli-gofructose and inulin also produced an anti-inflammato-ry effect in a rat model of acute colitis induced by DSS (Osman et al., 2006). Lastly, ingestion of bifidobacteria culture may stimulate the immune response (Hiramatsu et al., 2011). The relationship of B. longum with its host provides a myriad of unexplored biological mechanisms. Since experimental evidence of the effectiveness of bifido-bacteria is limited, …


INTRODUCTION
Lactic acid bacterium Bifidobacterium longum was isolated and described over a century ago from human infant feces and was associated with a healthy intestinal tract in humans (Tissier, 1900).This flora protects the intestinal tract from proliferation or infection by harmful bacteria; recent research has focused on bifidobacteria to establish the importance of these organisms in influencing certain normal functions of the intestinal tract and in exploring their role in the human health.
Evidence of the previous observations was described in a report where dietary administration of lyophilized cultures of B. longum suppressed colon and mammary carcinogenesis in laboratory animal models (Reddy, 1999), while another describes that bifidobacteria have effects on the treatment of rotavirus diarrhea (Marteau et al., 1990).Daily oral administration of bifidobacteria, oli-gofructose and inulin also produced an anti-inflammatory effect in a rat model of acute colitis induced by DSS (Osman et al., 2006).Lastly, ingestion of bifidobacteria culture may stimulate the immune response (Hiramatsu et al., 2011).
The relationship of B. longum with its host provides a myriad of unexplored biological mechanisms.Since experimental evidence of the effectiveness of bifidobacteria is limited, it is urgent to study the basic molecular mechanisms underlying the health-promoting features, especially in complex phenomena as anticarcinogenic or anti-inflammatory effects.Due to the importance of the aforementioned characteristics, the probiotic B. longum is subject to global analyses such as genome sequencing, genome variations (Bar-Nun & Glickman, 2012;Hao et al., 2011), Delcenserie et al., 2008;Guy et al., 2005;Schell et al., 2002;Sela et al., 2008) (Shkoporov et al., 2013;Yu et al., 2013;Zhurina et al., 2013) and proteomics (Olivares et al., 2011;Ruiz et al., 2009;Sun et al., 2008;Yuan et al., 2006;Zhao & Cheung, 2013).
B. longum belongs to a distinctive subclass of actinobacteria (bifidobacteria) that are proteasome-lacking or-ganisms (Zhang et al., 2004).In general, bifidobacteriales conserve the set of chaperones such as the trigger factor, small Hsp, GroEL/GroES, DnaK/DnaJ/GrpE, and ClpB (Sugimoto & Sonomoto, 2011) that substitute the UPS.We found that B. longum genome has the arc and accessory genes for the mechanism of pupylation (pup, pafA and dop genes).
In this work we describe the cloning, expression and biochemical characterization of the AAA ATPase ARC from B. longum with the intention to know if the coding sequence of the arc gene is active for subsequent functional studies.

MATERIALS AND METHODS
Sequence analysis and 3D structure.Homologues of the B. longum arc sequenced open reading frames were identified using the BLAST tool server at the NCBI (www.blast.ncbi.nlm.nih.gov/).Sequence alignments were done using the Clustal W application program.CLustal ANalysis of Sequences (CLANS) Program was used to detect families of related proteins in a large set of AAA sequences (Frickey & Lupas, 2004).3D structure prediction was performed by using the I-TASSER server (Roy et al., 2010;2012;Zhang, 2008) and the UCSF-Chimera program for interactive analyses of the structures (Meng et al., 2006).
Bacterial strain and cell culture.The B. longum subsp.infantis strain (DSMZ 20088) was purchased from the Leibniz-Institute DSMZ-German Collection of Microorganisms and Cell Cultures.B. longum was grown in a CO 2 chamber at 37°C in MRS medium containing 1 g peptone, 1 g beef extract, 0.5 g yeast extract, 0.1 g Tween 80, 0.2 g ammonium citrate, 0.5 g sodium acetate, 0.01 g MgSO 4 , 0.005 g MnSO 4 , 0.2 g K 2 HPO 4 , 2 g glucose and 0.05 g cysteine in 100 mL distillated water for three days.
DNA isolation.B. longum culture was harvested by centrifugation at 6 000 × g for 15 min at room temperature.Cell pellet was resuspended in the presence of 10 µg/µL lysozyme, 20 mg/ml proteinase K and 2 mg/ml zymolyase, for 30 min on ice.After incubation, phenol/ chloroform/isoamyl alcohol was added and stirred for 10 min.To separate the two phases (aqueous/organic) the samples were harvested at 5 400 g for 10 min.The aqueous layer was transferred to a fresh tube and 70% ethanol was added.Centrifugation was then performed at 5 400 g for 6 min to pellet the DNA.This washing step was repeated once.The pellet was air-dried for 5 min and resuspended in 30 µL sterile Milli-Q-water.
Cloning, heterologous expression in E. coli and immunoblotting.Amplification of the bl-arc gene (IPR003593 AAA+ ATPase core ORF) was performed with a pair of the following oligonucleotides: 5'-AT-GAGCGACACCGAAGATCTCG-3' and 5'-TTCG-GCGGTGCGGATGCGGC-3' and B. longum genomic DNA.To add a coding C-terminally fused His6-tag to the bl-arc gene into the phage T7 promoter-based expression pET28a(+) vector, we used the NdeI and XhoI sites at the 5' and 3' terminal ends: 5'-GGGGAATTCCAT-ATGAGCGACACCG-3' and 5'-CTCTCGAGTTCG-GCGGTGCGGATG-3'.PCR was done in a 50 µL reaction in the presence of Pfx polymerase (Invitrogen) using the following conditions: 94°C for 5 min, 30 cycles of 94°C for 30 s, 55°C for 30 s, 68°C for 2 min; and a final extension at 68°C for 10 min.The PCR product was cloned into the pCR4-TOPO TA vector and transformed into the host strain E. coli XL1 blue (Invitrogen).This construct was digested with NdeI and XhoI (Invitrogen) and ligated into a NdeI/XhoI pET28a(+) vector sites.The pET28a(+)-bl-arc construct was used to transform E. coli BL21(DE3) pLysS cells.Cell culture was grown in 500 mL of LB medium in the presence of kanamycin (50 mg/mL) and chloramphenicol (35 mg/ mL) at 28°C and induced with 1 mM IPTG until the optical density reached 0.6 at a wavelength of 600 nm.The cells were harvested at 13 000 rpm at 4°C for 20 minutes.The cell pellet was resuspended in lysis buffer (50 mM Na 2 H 2 PO 4 , 300 mM NaCl, pH 8.0), sonicated and pelleted again at 13 000 rpm for 20 minutes.The recombinant Bl-ARC protein was purified by affinity chromatography on Ni-NTA spin columns (Qiagen) under native conditions.Protein concentration was determined by using the Bio-Rad Protein Assay Solution and bovine serum albumin was used as a protein standard.To confirm the theoretically calculated molecular mass and the state of oligomerization, proteins were analyzed on a 10% SDS/PAGE and 5-15% native gradient gels.Bands were stained using the colloidal Coomassie method (R250, BioRad).The crude extract and purified protein were blotted from SDS or native gels onto nitrocellulose.Membranes were probed with an anti-His mouse monoclonal antibody conjugated with alkaline phosphatase (AbD Serotec) and antigen-antibody complexes were visualized by using alkaline phosphatase-conjugated anti-mouse IgG antibodies (Invitrogen) developed with NBT and BCIP (Amersham Bioscience).
Mass spectrometry.Bl-ARC was identified by mass spectrometry analysis (LC-ESI-MS/MS) in the LINAN laboratory facility at IPICYT.Data analysis was performed on the MS/MS spectra datasets using the MAS-COT search algorithm (Matrix Science, London, UK).Searches were conducted using the NCBI database.
ATPase activity assay.ATPase characterization was assayed by measuring the release of inorganic phosphate from ATP, by the method of Lanzetta and coworkers (Lanzetta et al., 1979).Triplicate reactions containing 10 μg of the recombinant Bl-ARC in a buffer containing 50 mM MES, pH 5.0, 100 mM NaCl, 10 mM of several divalent cation metals, and 1 mM nucleotide were incubated for 15 min at 37°C.Negative control reactions in the absence of the enzyme, also in triplicate, were performed in parallel.Kinetic parameters, K m and V max were determined within the first ten minutes of the assay by measuring the initial velocity of ATP hydrolysis.These parameters were calculated using the KaleidaGraph program (Synergy Software).

Sequence analysis, phylogenetic classification and 3D structure
Deduced amino acid sequence analysis of the IPR003593 AAA+ ATPase core ORF showed 40% identity to the ARC AAA ATPase from R. erythropholis (Genbank AAC6869.1) and Streptomyces coelicor (Genbank AAC64282.1).Database searches displayed a wide number of ARC homologs (Zhang et al., 2004).BLAST searches and multiple sequence alignment confirmed high sequence similarity against chaperones involved in the degradation pathway via proteasome such as archaeal PAN (Smith et al., 2005) and the human Rpt1 proteasomal subunit (Bar-Nun & Glickman, 2012) (Fig. 1).The analysis revealed presence of a single AAA family domain where the Walker A region covers the residues Gly232 to Leu240, the Walker B from Iso311 to Glu316 and the SRH from Ala358 to Asp376.As depicted in Fig. 1, the sequence contains an N-terminal coiled coil considered important for possible binding and recognition of target proteins in the degradation process (Djuranovic et al., 2009) (Met1 to Pro46).This region contains three conserved leucine residues in positions 7, 14, 21 and one Ala28.Likewise, other conserved amino acids, Ala35, Lys36, Leu39, Leu42, Pro45 and Pro46 are present at the N-terminal substrate recognition domain of proteasomal ATPases (Djuranovic et al., 2009).As expected, sequence analysis of the C-termini clearly shows the two missing regions required for the interaction with proteasome (box I and II).This feature has been reported for Corynebacterium efficiens, C. diphtheriae, and C. glutamicum (Zhang et al., 2004) (Fig. 1).CLANS' analysis (Frickey & Lupas, 2004) was computed against 10 000 AAA proteins covering all subfamilies of the AAA ATPase.The tree confirmed the IPR003593 AAA+ ATPase core deduced amino acid sequence to be a part the ARC subgroup in the phylogenetic dendrogram; thus we termed this molecule as Bl-ARC.
A 3D model of Bl-ARC was constructed based on the homology of AAA published structures from M. tuberculosis ARC and Mus musculus p97.Homology modeling was predicted against the Protein Data Bank (http:// www.rcsb.org)using the I-TASSER server.The server returned a model using the X-ray crystal structure of the AAA protein p97 (PDB ID: 3CF1) as a template.We obtained the best hit: 0.52 TM-score against Bl-ARC that indicates the global structural similarity between both structures (a TM-score > 0.5 indicates a model of correct topology vs TM-score < 0.17 means random similarity) (Xu & Zhang, 2010).Moreover, we obtained a C-score of 0.33; this value confirmed the quality of the predicted model (range 0 to 1; higher value indicates more confidence).Lastly, the RMSD score (root-mean-square deviation of atomic positions) of 5.35 denoted a high average distance between the atoms of the superimposed proteins (Roy et al., 2010).
The visualization of the structure was obtained with the USFC-Chimera program.We found that the N-terminal domain in Bl-ARC is composed of five parallel beta strands (β5, β1, β4, β3, β2) flanked by alpha helices as reported for other AAA proteins (Ogura & Wilkinson, 2001;Reuter et al., 2004).In the ARC proteins this region assists the chaperone activity by mediating interactions with Pup proteins in the pupylation mechanism (Sutter et al., 2009).
Crystal structures of hexameric AAA proteins such as HsIU, FtsH, LonB and ClpX have been reported by their importance in protein processing.This conformation plays a crucial role in mechanistic events such as upon ATP hydrolysis that allows substrates to adjust and unfold; the substrates can be then translocated through the pore of the hexameric ring (Vale, 2000).
The hexameric structure of Bl-ARC was therefore modeled, considering the interdomain crystal structure described for M. tuberculosis (PDB 3M9B) that is 28% similar to Bl-ARC (Fig. 2).Results of this model suggest a close functional relationship between Bl-ARC and proteins that target substrates for degradation.

Expression of recombinant Bl-ARC and state of oligomerization
Sequencing of the positive clones with the bl-arc gene confirmed an inserted DNA of 1566 bp in length corresponding to the expected size by agarose gel electrophoresis (Fig. 3a).The calculated molecular mass was also confirmed when the overexpressed soluble C-His-tagged recombinant Bl-ARC appeared as a band of approximately 56 kDa by SDS/PAGE (Fig. 3b).Western blot was performed (Fig. 3c) to detect the recombinant protein.Gradient native electrophoresis gel, 5-15%, was performed to detect the migration of the native protein.An apparent oligomeric species of 336 kDa (Fig. 3d) resemble a possible homohexameric complex.We expected Bl-ARC to form a ring conformation such as ReARC, therefore preliminary transmission electron microscopy has been performed on poly-L-lysine grids to confirm the structure (personal communication, Timo Merloo, Department of Cellular and Molecular Medicine, UCSD) at 68 000× magnification.We have so far observed small aggregations and few particle ring patterns, however experiments through particle averaging are needed to confirm the structure.
Bl-ARC is active in a temperature range of 25-65ºC, however the maximum ATPase activity was observed at 37ºC; this corresponds to the mesophilic nature of B. longum.Moreover, the enzyme remained stable at a wide pH range, with preferential activity in the acidic pH (1-6) and a maximum activity was reached at pH 5.0.Interestingly, evidence of pH measurements of feces from bottle-fed infants was found to be > 7.0 versus breast-fed infants with < 6.0 during the first 7 weeks after birth, thus producing an important inhibition of undesirable intestinal bacteria by the presence of bifidobacteria (Lee & O'Sullivan, 2010).Adults possess intraluminal pH of small intestine of 6 to about pH 7.4 in the terminal ileum (Fallingborg, 1999); therefore it has been proposed the use bifidobacterial supplementation in the diet.
Michaelis-Menten plot showed a K m value of 327 μM and V max of 492 pmol of ATP hydrolyzed per μg of Bl-ARC per min.These parameters are comparable to AAA ReARC and MjPAN (Table 3).
We found that Bl-ARC is an active AAA ATPase.The bl-arc gene in B. longum genome (Sela et al., 2008) was found in a close proximity to the accessory genes pup, pafA and dop involved in the mechanism of pupylation.This suggests a possible AAA unfoldase activity for target substrates prior to degradation.The functional mechanism of ARC in B. longum needs to be further investigated at a molecular and cellular level, along with its role in the maintenance of homeostasis within the host.

Figure 1 .
Figure 1.Alignment of B. longum ARC with representative members of proteasomal ATPases.(Top) For the coiled-coil structures, the positions of the hydrophobic core residues are marked (a, d), and the regions predicted to assume coiled-coil structure in the individual sequences are underlined; the sequence of the ScGCN4 leucine zipper is included for reference, and the location of the trigger sequence essential for folding is indicated.(Middle) The consensus secondary structure is shown as s = beta strand and h = alpha helix.(Bottom) The AAA domain is shown Walker A/B and Sensor 1 motif, the arginine residue predicted to act as an arginine finger, as well as the pore loop, are indicated in the alignment with arrows; Second Region of Homology (SRH) is shown in the upper text with a black line.Residues conserved in the aligned sequences are highlighted in black and similar residues are highlighted in grey.A box at the C-terminal end of the Bl-ARC sequence is depicted indicating the lacking-proteasome interacting region.AfPAN (Archaeoglobus fulgidus gi 3122632), MjPAN (Methanocaldococcus jannaschii gi 2492524), ReARC (Rhodococcus erythropolis gi 3790601), MtARC (Mycobacterium tuberculosis gi 15841607), KrARC (Kineococcus radiotolerans gi 69285208), ApPAN (Aeropyrum pernix gi 20532220) and BlARC (Bifidobacterium longum subsp.infantis gi 213524442.(b) ARC proteins from proteasome-lacking organisms: Blon, Bifidobacterium longum subsp.infantis; Cef, Corynebacterium efficiens; Cdi, C. diphtheriae; Cglu, C. glutamicum.Microorganisms containing proteasome genes contain regions in ARC that interact with the proteasome α-subunit (boxes I and II): Mlep, M. leprae; Mtub, M. tuberculosis; Rery, R. erythropolis; Scoe, S. coelicolor.

Figure 2 .
Figure 2. The overall 3D structure prediction of the AAA Bl-ARC obtained by homology modeling.(a) Composition of the ATP binding domains.The Walker motif A and B segments are colored green and cyan respectively, and the SRH is in magenta.(b) Complete assembly of the hexameric ring.This prediction was conducted using the Mtb ARC crystal interdomain.

Table 2 . Divalent metal ion specificity at 1 mM ATP
Nucleotide hydrolyzing activity is expressed as a percentage of the control reaction *