Overexpression of the groESL Operon Enhances the Heat and Salinity Stress Tolerance of the Nitrogen-Fixing Cyanobacterium Anabaena sp. Strain PCC7120

The bicistronic groESL operon, encoding the Hsp60 and Hsp10 chaperonins, was cloned into an integrative expression vector, pFPN, and incorporated at an innocuous site in the Anabaena sp. strain PCC7120 genome. In the recombinant Anabaena strain, the additional groESL operon was expressed from a strong cyanobacterial P_psbA1 promoter without hampering the stress-responsive expression of the native groESL operon. The net expression of the two groESL operons promoted better growth, supported the vital activities of nitrogen fixation and photosynthesis at ambient conditions, and enhanced the tolerance of the recombinant Anabaena strain to heat and salinity stresses.Nitrogen-fixing cyanobacteria, especially strains of Nostoc and Anabaena, are native to tropical agroclimatic conditions, such as those of Indian paddy fields, and contribute to the carbon (C) and nitrogen (N) economy of these soils (22, 30). However, their biofertilizer potential decreases during exposure to high temperature, salinity, and other such stressful environments (1). A common target for these stresses is cellular proteins, which are denatured and inactivated during stress, resulting in metabolic arrest, cessation of growth, and eventually loss of viability. Molecular chaperones play a major role in the conformational homeostasis of cellular proteins (13, 16, 24, 26) by (i) proper folding of nascent polypeptide chains; (ii) facilitating protein translocation and maturation to functional conformation, including multiprotein complex assembly; (iii) refolding of misfolded proteins; (iv) sequestering damaged proteins to aggregates; and (v) solubilizing protein aggregates for refolding or degradation. Present at basal levels under optimum growth conditions in bacteria, the expression of chaperonins is significantly enhanced during heat shock and other stresses (2, 25, 32).The most common and abundant cyanobacterial chaperones are Hsp60 proteins, and nitrogen-fixing cyanobacteria possess two or more copies of the hsp60 or groEL gene (http://genome.kazusa.or.jp/cyanobase). One occurs as a solitary gene, cpn60 (17, 21), while the other is juxtaposed to its cochaperonin encoding genes groES and constitutes a bicistronic operon groESL (7, 19, 31). The two hsp60 genes encode a 59-kDa GroEL and a 61-kDa Cpn60 protein in Anabaena (2, 20). Both the Hsp60 chaperonins are strongly expressed during heat stress, resulting in the superior thermotolerance of Anabaena, compared to the transient expression of the Hsp60 chaperonins in Escherichia coli (20). GroEL and Cpn60 stably associate with thylakoid membranes in Anabaena strain PCC7120 (14) and in Synechocystis sp. strain PCC6803 (15). In Synechocystis sp. strain PCC6803, photosynthetic inhibitors downregulate, while light and redox perturbation induce cpn60 expression (10, 25, 31), and a cpn60 mutant exhibits a light-sensitive phenotype (http://genome.kazusa.or.jp/cyanobase), indicating a possible role for Cpn60 in photosynthesis. GroEL, a lipochaperonin (12, 28), requires a cochaperonin, GroES, for its folding activity and has wider substrate selectivity. In heterotrophic nitrogen-fixing bacteria, such as Klebsiella pneumoniae and Bradyrhizobium japonicum, the GroEL protein has been implicated in nif gene expression and the assembly, stability, and activity of the nitrogenase proteins (8, 9, 11).Earlier work from our laboratory demonstrated that the Hsp60 family chaperonins are commonly induced general-stress proteins in response to heat, salinity, and osmotic stresses in Anabaena strains (2, 4). Our recent work elucidated a major role of the cpn60 gene in the protection from photosynthesis and the nitrate reductase activity of N-supplemented Anabaena cultures (21). In this study, we integrated and constitutively overexpressed an extra copy of the groESL operon in Anabaena to evaluate the importance and contribution of GroEL chaperonin to the physiology of Anabaena during optimal and stressful conditions.Anabaena sp. strain PCC7120 was photoautotrophically grown in combined nitrogen-free (BG11⁻) or 17 mM NaNO₃-supplemented (BG11⁺) BG11 medium (5) at pH 7.2 under continuous illumination (30 μE m⁻² s⁻¹) and aeration (2 liters min⁻¹) at 25°C ± 2°C. Escherichia coli DH5α cultures were grown in Luria-Bertani medium at 37°C at 150 rpm. For E. coli DH5α, kanamycin and carbenicillin were used at final concentrations of 50 μg ml⁻¹ and 100 μg ml⁻¹, respectively. Recombinant Anabaena clones were selected on BG11⁺ agar plates supplemented with 25 μg ml⁻¹ neomycin or in BG11⁻ liquid medium containing 12.5 μg ml⁻¹ neomycin. The growth of cyanobacterial cultures was estimated either by measuring the chlorophyll a content as described previously (18) or the turbidity (optical density at 750 nm). Photosynthesis was measured as light-dependent oxygen evolution at 25 ± 2°C by a Clark electrode (Oxy-lab 2/2; Hansatech Instruments, England) as described previously (21). Nitrogenase activity was estimated by acetylene reduction assays, as described previously (3). Protein denaturation and aggregation were measured in clarified cell extracts containing ∼500 μg cytosolic proteins treated with 100 μM 8-anilino-1-naphthalene sulfonate (ANS). The pellet (protein aggregate) was solubilized in 20 mM Tris-6 M urea-2% sodium dodecyl sulfate (SDS)-40 mM dithiothreitol for 10 min at 50°C. The noncovalently trapped ANS was estimated using a fluorescence spectrometer (model FP-6500; Jasco, Japan) at a λ_excitation of 380 nm and a λ_emission of 485 nm, as described previously (29).The complete bicistronic groESL operon (2.040 kb) (GenBank accession no. {"type":"entrez-nucleotide","attrs":{"text":"FJ608815","term_id":"222354886","term_text":"FJ608815"}}FJ608815) was PCR amplified from PCC7120 genomic DNA using specific primers (Table ​(Table1)1) and the amplicon cloned into the NdeI-BamHI restriction sites of plasmid vector pFPN, which allows integration at a defined innocuous site in the PCC7120 genome and expression from a strong cyanobacterial P_psbA1 promoter (6). The resulting construct, designated pFPNgro (Table ​(Table1),1), was electroporated into PCC7120 using an exponential-decay wave form electroporator (200 J capacitive energy at a full charging voltage of 2 kV; Pune Polytronics, Pune, India), as described previously (6). The electroporation was carried out at 6 kV cm⁻¹ for 5 ms, employing an external autoclavable electrode with a 2-mm gap. The electroporation buffer contained high concentrations of salt (10 mM HEPES, 100 mM LiCl, 50 mM CaCl₂), as have been recommended for plant cells (23) and other cell types (27). The electrotransformants, selected on BG11⁺ agar plates supplemented with 25 μg ml⁻¹ neomycin by repeated subculturing for at least 25 weeks to achieve complete segregation, were designated AnFPNgro.

TABLE 1.

Plasmids, strains, and primers used in this study

Molecular Cloning of Cyanobacterial Pteridine Glycosyltransferases That Catalyze the Transfer of either Glucose or Xylose to Tetrahydrobiopterin

Here, we report cloning of cyanobacterial genes encoding pteridine glycosyltransferases that catalyze glucosyl or xylosyl transfer from UDP-sugars to tetrahydrobiopterin. The genes were cloned by PCR amplification from genomic DNA which was isolated from culture and environmental samples and overexpressed in Escherichia coli for an in vitro activity assay.Tetrahydrobiopterin (BH4) is well known among pteridine compounds as a cofactor for aromatic amino acid hydroxylases and nitric oxide synthases in animals (19). Pteridine glycosides such as biopterin and 6-hydroxymethylpterin glycosides have been found in cyanobacteria and anaerobic photosynthetic bacteria (2, 4, 5, 8, 11, 13, 15, 17, 21). Although the function of these glycosides remains unknown, they are abundant and ubiquitous in cyanobacteria, implying some essential role (3, 6, 16-18, 21). There is a group of enzymes, named pteridine glycosyltransferases (PGTs), known to catalyze a variety of glycosyl transfers to pteridines. The first PGT isolated from the cyanobacterium Synechococcus sp. PCC 7942 was shown to catalyze a glucosyl transfer from UDP-glucose to BH4 and was therefore named UDP-glucose:BH4 glucosyltransferase (BGluT) (7). After cloning of the gene encoding BGluT (6), a PGT that catalyzes the transfer of glucuronic acid for cyanopterin synthesis was identified (12). In addition, there are many putative PGT homologs encoded in bacterial genomes, although their exact catalytic functions have not been determined. We recently found that BGluT is useful for the simultaneous detection of oxidized and reduced forms of BH4 in animal samples (14). Glycosyltransferases are also being studied intensively for applications in the design of novel pharmaceutical derivatives (1, 10). We were thus encouraged to find PGTs with new substrate specificities or enzymatic properties not only for study of protein structure and function but also for application in BH4 research. In this study, we succeeded in cloning four cyanobacterial genes encoding PGTs with either glucosyl- or xylosyltransferase activity, and here we report the results.PGT genes were cloned from Arthrospira platensis CY-007 (obtained through Hawaii Oceanic Institute sampling) and Arthrospira maxima CY-049 (UTEX 2342), which were cultured in the Korea Marine Microalgae Culture Center, and from environmental DNA sequences (designated UCNR-001 and UCNR-002) isolated from wild algal mats in the Nakdong River, South Korea. In order to amplify conserved internal sequences of the unknown PGT genes, degenerate PCR primers were designed from the nucleotide sequences of cyanobacterial PGT homologs using GeneFisher2 (9). A protein homology search with BGluT against the bacterial genome database in NCBI revealed more than a hundred PGT homologs. When a phylogenetic tree was constructed from the putative sequences, there was a separate group comprising cyanobacterial PGTs. Figure ​Figure11 shows the cyanobacterial cluster, in which members shared sequence identities of more than 34%. Because the degenerate primers designed from all of the cyanobacterial PGTs were too highly degenerate, the cluster was divided into four subgroups, as shown in Fig. ​Fig.1:1: this division allowed primers to be designed for each of the four subgroups. The PGTs in subgroup I were clearly distinguishable from the others, because they all originated from marine picocyanobacteria, which are abundant in the pelagic realm. Subgroup I could be divided further into two groups comprising PGTs from either Prochlorococcus species (CIA) or marine Synechococcus species (CIB). Subgroup II was also divided into two groups, CIIA, consisting mostly of PGTs from Synechococcus species, and CIIB, containing the other PGTs. Among the primers designed for each subgroup, those for the CIIA and CIIB subgroups successfully amplified DNA sequences of the expected sizes. The primer sequences were 5′-GTTCAGGAWTAGGAGGTGGAGT-3′ (CIIA-forward)/5′-CGCYTCAATWGCTACATTTCCA-3′ (CIIA-reverse) and 5′-ACGACTGGCTMYCGYTTTAYCTGA-3′ (CIIB-forward)/5′-GCYTCCACCCAYTTRGGGGTCA-3′ (CIIB-reverse). Based on the determined partial gene sequences, additional sets of primer pairs were designed for the inverse PCR method (20). The sequences were 5′-GATGAACTACAACAGGGTCTGCGTC-3′ (CY-007 forward)/5′-CGGCTTTTTAAGGCTTTTGCCATATTC-3′ (CY-007 reverse), 5′-GTCTGCGTGAATGTCGAGG-3′ (CY-047 forward)/5′-ATGACCTCGGCTGTGTAAG-3′ (CY-047 reverse), and 5′-CCTACAAAAAGAGCTAGGCGACTGTTTTG-3′ (UCNR forward)/5′-CCAAAGAAACGGAAGCCATGCTG-3′ (UCNR reverse). Total genomic DNA samples were partially digested with RsaI and then self-ligated to be used as templates for PCR amplification with the primer pairs. The amplified DNA sequences revealed the missing 5′- and 3′-end sequences of the genes.Open in a separate windowFIG. 1.Neighbor-joining phylogenetic tree of cyanobacterial PGT protein sequences, identified by NCBI accession numbers. Bootstrap values are presented at the nodes. The names of strains whose PGTs are characterized are in bold.The deduced protein sequences were multiply aligned with BGluT (Fig. ​(Fig.2).2). Amino acid identities for all sequences in pairwise comparisons are given as percentages in Fig. ​Fig.2.2. Recently, draft assemblies of the genome sequences of Arthrospira platensis strain Paraca and Arthrospira maxima CS-328 (UTEX 2342) were announced. The annotated PGT (GenBank accession no. {"type":"entrez-protein","attrs":{"text":"EDZ91868","term_id":"209491491","term_text":"EDZ91868"}}EDZ91868) of Arthrospira maxima CS-328 was identical to the PGT of CY-049 at both the amino acid and nucleotide levels, proving that the two organisms originated from the same UTEX stock (UTEX 2342). On the other hand, the PGTs of Arthrospira platensis strains Paraca and CY-007 were different at nine individual nucleotides, resulting in seven amino acid differences. A phylogenetic analysis showed that CY-007 and CY-049 PGTs belonged to the CIIB subgroup and that UCNR-001 and UCNR-002 PGTs clustered in the CIIA subgroup (data not shown).Open in a separate windowFIG. 2.Alignment of multiple PGT sequences. Conserved sequences are shaded at four levels using GeneDoc software. At the end of the alignment, amino acid identities in percentages are given for all sequences in pairwise comparisons.In order to identify the catalytic function of the putative PGTs, the recombinant proteins were produced in Escherichia coli. The complete open reading frame (ORF) sequences were amplified by PCR from the genomic DNA samples, cloned into the pGEM-T vector, and subsequently cloned as NdeI/BamHI restriction fragments into pET-28b (for CY-007 and CY-049 sequences) or pET-15b (for UCNR-001 and UCNR-002 sequences). E. coli BL21(DE3)/pLysS transformants were induced with 0.05 to 0.2 mM isopropyl-β-d-thiogalactopyranoside and were cultured for 8 h at 22°C. The recombinant proteins were purified by chromatography on Ni-nitrilotriacetic acid gel according to the instructions of the manufacturer (Qiagen). The proteins were eluted with 250 mM imidazole, dialyzed against a mixture of 20 mM Tris-HCl (pH 7.5) and 30% (vol/vol) glycerol, and stored in aliquots at −70°C until use. Purification of the proteins was confirmed by electrophoresis on an SDS-polyacrylamide gel (Fig. ​(Fig.3A).3A). BGluT from a previous purification was used (6). Aliquots of PGT were assayed at 37°C for 10 min in a reaction mixture of 100 μl containing 50 mM sodium phosphate, pH 7.5, 10 mM MnCl₂, 0.2% ascorbic acid, 1 μM BH4 (Schircks Lab, Switzerland), and 100 μM UDP-glucose or UDP-xylose. The reaction mixture was combined with an equal volume of acidic iodine solution (2% KI and 1% I₂ in 1 N HCl) for 1 h in the dark. After centrifugation, the supernatant was mixed in a 10:1 volume ratio with 5% ascorbic acid and subjected to high-performance liquid chromatography (HPLC). HPLC was performed with a Gilson 321 pump equipped with an Inertsil ODS-3 column (150 by 2.3 mm; particle size, 5 μm [GL Science, Japan]) and a fluorescence detector (Shimadzu RF-10AXL). Pteridines were eluted with 10 mM potassium phosphate buffer (pH 6.0) at a flow rate of 1.2 ml/min and were monitored at excitation and emission wavelengths of 350 and 450 nm, respectively.Open in a separate windowFIG. 3.Analysis of purified recombinant PGTs on an SDS-12.5% polyacrylamide gel (A) and HPLC analysis of the enzymatic products (B).The enzymatic products of PGTs (Fig. ​(Fig.3)3) appeared only when enzymes were incubated with BH4 as a sugar acceptor and either UDP-glucose (for CY-007, UCNR-001, and UNCR-002 PGTs) or UDP-xylose (for CY-049 PGT) as a sugar donor. HPLC analysis of cultured CY-007 and CY-049 cells confirmed the presence of the corresponding biopterin glycosides (data not shown), supporting the conclusion that the PGTs exhibited genuine in vivo activities. This is the first report of a gene encoding a PGT that catalyzes xylosyl transfer to BH4. Although the data are not shown here, we found additional xylosyl transfer PGTs in Anabaena sp. PCC 7120, Gloeobacter violaceus PCC 7421, and Thermosynechococcus elongatus BP-1, whose genomic sequences were determined. The putative PGT genes (represented in Fig. ​Fig.1)1) were amplified by PCR from the genomic DNA, which was a kind gift from the Kazusa DNA Research Institute (http://genome.kazusa.or.jp/cyanobase/). The recombinant proteins for the in vitro activity assay were prepared by cloning the genes into pET-28b and overexpressing the proteins in E. coli according to the same procedures performed for the other PGTs. Interestingly, CY-007 and CY-049 PGTs exhibited different substrate specificities, although they share 93% protein sequence identity, and they also had higher specific activities than the other PGTs (Fig. ​(Fig.4).4). The three-dimensional structures of the proteins are currently being investigated to further understanding of the structural properties involved. Considering the cyanobacterial PGTs hitherto identified, there seems to be little correlation between their substrate preferences and phylogenetic classification. However, the CI group PGTs, which diverged early from the CII group PGTs, might have some distinctive features. Finally, the successful cloning of PGT genes from environmental DNA allows for potentially new PGTs to be isolated from cyanobacteria, which are abundant in nature.Open in a separate windowFIG. 4.Comparative analysis of PGT activities. The maximal activity (100%) corresponds to complete glycosylation of 1 μM BH4 in the reaction mixture, which contained 0.5 mM UDP-xylose for CY-049 PGT or 0.5 mM UDP-glucose for the other PGTs. The mixtures were incubated for 10 min with the indicated amounts of proteins.     

Structural Study of the Serratia entomophila Antifeeding Prophage: Three-Dimensional Structure of the Helical Sheath

The sheath of the Serratia entomophila antifeeding prophage, which is pathogenic to the New Zealand grass grub Costelytra zealandica, is a 3-fold helix formed by a 4-fold symmetric repeating motif disposed around a helical inner tube. This structure, determined by electron microscopy and image processing, is distinct from that of the other known morphologically similar bacteriophage sheaths.The antifeeding prophage (Afp) of Serratia entomophila and Serratia proteamaculans is a naturally occurring virus tail-like structure which delivers a putative toxin molecule that leads to starvation of the New Zealand grass grub Costelytra zealandica (5). Afp is composed of 18 different gene products (molecular masses of 6.5 to 263 kDa). The first 16 open reading frames have orthologues (Photorhabdus virulence cassettes [PVC]) in the insecticidal bacterium Photorhabdus luminescens TTO1 genome (5). Afp and PVCs morphologically resemble a typical R-type bacteriocin (6, 12, 16) However, Afp is the only known phage tail-like protein complex that is not a bacteriocin-protein complex of considerable medical relevance that targets the same or closely related bacterial strains (1, 3, 8, 12). The major component of Afp is a contractile cylindrical outer sheath encasing an inner tube speculated to house the toxin molecule (6). A dome-shaped “head” defines one extremity of the tube, while the other end is attached to a “bell-shaped” structure with a base morphologically similar to the base plate of the T4 bacteriophage tail (9).Transmission electron micrographs of two-dimensional (2D) projections of negatively stained (Fig. ​(Fig.11 A) or frozen-hydrated and vitrified (Fig. ​(Fig.1B)1B) recombinant Afp particles (see Fig. S1 in the supplemental material) were used for computational image analysis. A globally averaged image of the Afp particle in the major configuration (called E here) (Fig. ​(Fig.1C),1C), generated using negatively stained specimens, clearly distinguished the morphologies of the various constituent structural parts. Thus, the cylindrical sheath appears to be formed by a periodic structure harboring a distinctive, inverted-V-shaped feature. A minor population of Afp particles displays an alternate configuration (called C here) where, concomitant with contraction of the sheath (averaged axial compression of ∼52% [see Table S1 in the supplemental material]), the inner tube, shorn off the bell-shaped structure, is revealed (Fig. ​(Fig.1A)1A) (6). Several other bacteriocins undergo such a high degree of compression, which has been characterized in detail for the tail sheath of bacteriophage T4 (9). We also generated individual global averages for the periodic sheath structure, for the bell-shaped structure, and for the inner tube (Fig. ​(Fig.1C)1C) which provide more accurate dimensions of these different sections (see Table S1 in the supplemental material) than those reported earlier (6).Open in a separate windowFIG. 1.(A) Electron micrograph of a negatively stained preparation of partially purified recombinant Afp particles. The gray, white, and black arrows point to an Afp particle in the extended (E) configuration, to an Afp particle with the sheath contracted, exposing the inner tube in the contracted (C) configuration, and to an inner tube with the bell-shaped structure attached at one end, respectively. (B) Cryoelectron micrograph recorded from a preparation similar to that shown in panel A. (C) Globally averaged images of Afp particles (3,026 images) in the E state and the three distinguishable parts visualized by negative staining. Bars, 200 Å. (D) Averaged power spectrum of the sheath of vitrified Afp particles. The black and white arrows indicate reflections delineating the axial rise (1/78.5 Å) and the helical pitch (1/118 Å), respectively. In panels A and C, lighter regions represent protein and the contrast is reversed in panel B. Global averages were created using classalign2 of the EMAN suite (10) and visualized in bshow (4).For a better insight, we determined the 3D structure of the central periodic section of the Afp particle in the E state. A global power spectrum derived from the cryoimages established the structure to be helical with a clear first meridional reflection at 1/78.5 Å and the first, strongest nonmeridional reflection at 1/118 Å (Fig. ​(Fig.1D).1D). These reflections correspond to the axial rise (Δx) and the pitch of the helix, respectively, and reflect a turn angle (Δψ) of about ±240° (3₂ helix) for the repeating motif. The correct sign, i.e., the hand, of the helix remains to be determined. Computationally excised overlapping segments of this helical section from images of vitrified and negatively stained Afp particles were subjected to 3D reconstruction using the iterative helical real-space reconstruction (IHRSR) algorithm (2) using the determined helical parameters (see Fig. S2 in the supplemental material). After a few iterations, the presence of an in-plane 4-fold symmetry (C4) was apparent in the density map (see Fig. S2 in the supplemental material), which was then imposed in the subsequent reconstruction exercises. However, no stable solution was forthcoming, even after many (e.g., 30) iterative cycles. This is generally indicative of the presence of heterogeneity in the form of variations in helix translation and/or twist angle (15) in the structure. As a first step, we focused our attention on the pitch value, and following classification (see Fig. S3 in the supplemental material), we found that the majority of the image segments correspond to a pitch of 120 Å. These segments were then selected out of the full data set and led to a stable and refined 3D reconstruction. We also obtained very comparable results for the helical section when images of negatively stained Afp sheath sections were used, thus supporting our computational approach (see Fig. S4 in the supplemental material) and general conclusions about the E state described below.Figure ​Figure22 A is an isosurface representation of the density map of the helical Afp sheath in the E state calculated at ∼21.5-Å resolution (see Fig. S5 in the supplemental material). To the best of our knowledge, a 4-fold rotational symmetry has not been seen for any other contractile T4 bacteriophage taillike structure, which points to the unique architecture of the Afp sheath. A power spectrum generated using the 2D projection from the final density map, compared to the experimental global power spectrum (Fig. ​(Fig.2D),2D), showed strong agreement, confirming the fidelity of the computational image analysis. The density map displays protein layers, ∼80 Å thick, that are stacked on each other in a periodic fashion. The uneven outer surface of the sheath is perforated and decorated with ∼35-Å protrusions. When rendered with a raised threshold, a characteristic feature of the map is a contiguous, high-electron-density region having an inverted-Y-shaped structure (Fig. 2C and E; see Fig. S4 in the supplemental material). At the modest ∼21.5-Å resolution, the boundary of the repeating subunit cannot be defined. A 25 ± 3-Å-wide central lumen is seen clearly when viewed along the helix axis (Fig. ​(Fig.2B)2B) and likely represents the pore of the inner tube (see also below). Using scanning transmission electron microscopy (STEM) (see Fig. S6 in the supplemental material), we estimated the averaged molecular mass of the central helical section of an Afp particle to be 9.8 ± 0.4 kDa/Å (Fig. ​(Fig.3)3) (14) and that of only the inner tube to be ∼2.5 kDa/Å, based on a relatively small pool of such images. These values translate to a mass contribution of approximately 145 kDa of the subunit whose periodic arrangement forms the outer component of the sheath (i.e., excluding the inner tube) (see Fig. S6 in the supplemental material). This value is not very different from the cumulative mass of the different proteins, i.e., homologous afp2, afp3, and afp4, thought to be involved in Afp sheath formation (5) (see Fig. S7 in the supplemental material).Open in a separate windowFIG. 2.Orthogonal isosurface rendering, at 1 σ (standard deviation) of the computed ∼21.5-Å density map of the helical sheath of the vitrified Afp particle viewed normal (A) and parallel (B) to the helix. The images were generated using the software package CHIMERA (13). The arrow indicates a surface perforation. (C) The Afp density map rendered at 3.5 σ to highlight the largest contiguous high-electron-density regions; one circumscribed by a black ellipse is computationally extracted and shown in panel E. (D) Comparison of the experimental, averaged power spectrum of the helical sheath of Afp (left) with that computed (right) from the 2D projection of the calculated density map. (F) Global average of the inner tube of the Afp particle and a plot of the surface density variation (scaled from 0 to 1) along the helical (y) axis. The dimension along the tube is plotted on the x axis.Open in a separate windowFIG. 3.(A) Dark-field micrograph of a freeze-dried, unstained preparation of Afp particles used in STEM measurements. An Afp particle in the E state, an Afp inner tube with the attached bell-shaped structure, and a tobacco mosaic virus particle, used as a calibration standard, are marked by the arrowhead and the gray and white arrows, respectively. (B) Histogram plot of the measured distributions of mass per unit length corresponding to the uniform periodic section of the Afp particles overlaid with a fitted Gaussian curve produced by using the ORIGIN6 software package.A paucity of images of the C state (∼5% of the complete data set) precluded a full, refined 3D reconstruction, but based on the available 2,774 overlapping image segments of the isolated inner tube, a global average was calculated. A plot of contrast variation (Fig. ​(Fig.2F)2F) indicates that the surface is characterized by ∼40-Å spaced elevated crests and invaginated grooves, in agreement with the calculated axial rise of ∼39 Å for the subunit (see Fig. S8 in the supplemental material) comprising the tube. Based on these preliminary results, it appears that the helical symmetry of the inner tube is markedly different from that of the sheath.Our observation that the pitch of the helix in the E state can vary by as much as ∼50 Å attests to the flexible nature of the sheath, which is required for compressibility and may be facilitated by the somewhat porous nature of the sheath (Fig. ​(Fig.2).2). Preliminary deductions (data not shown) based on a small pool of images of the C state suggest profound rearrangement of the elements of the sheath. How that translates to extrusion of the toxin remains to be revealed.      

Efficient Production of 2-Oxobutyrate from 2-Hydroxybutyrate by Using Whole Cells of Pseudomonas stutzeri Strain SDM

2-Oxobutyrate is an important intermediate in the chemical, drug, and food industries. Whole cells of Pseudomonas stutzeri SDM, containing NAD-independent lactate dehydrogenases, effectively converted 2-hydroxybutyrate into 2-oxobutyrate. Under optimal conditions, the biocatalytic process produced 2-oxobutyrate at a high concentration (44.4 g liter⁻¹) and a high yield (91.5%).2-Oxobutyrate (2-OBA) is used as a raw material in the synthesis of chiral 2-aminobutyric acid, isoleucine, and some kinds of medicines (1, 8). There is no suitable starting material for 2-OBA production by chemical synthesis; therefore, the development of innovative biotechnology-based techniques for 2-OBA production is desirable (12).2-Hydroxybutyrate (2-HBA) is cheaper than 2-OBA and can be substituted for 2-OBA in the production of isoleucine, as reported previously (9, 10). The results of those studies also indicated that it might be possible to produce 2-OBA from 2-HBA by a suitable biocatalytic process. In the presence of NAD, NAD-dependent 2-hydroxybutyrate dehydrogenase can catalyze the oxidation of 2-HBA to 2-OBA (4). However, due to the high cost of pyridine cofactors (11), it is preferable to use a biocatalyst that directly catalyzes the formation of 2-OBA from 2-HBA without any requirement for NAD as a cofactor.In our previous report, we confirmed that NAD-independent lactate dehydrogenases (iLDHs) in the pyruvate-producing strain Pseudomonas stutzeri SDM (China Center for Type Culture Collection no. M206010) could oxidize lactate and 2-HBA (6). Therefore, in addition to pyruvate production from lactate, P. stutzeri SDM might also have a potential application in 2-OBA production.To determine the 2-OBA production capability of P. stutzeri SDM, the strain was first cultured at 30°C in a minimal salt medium (MSM) supplemented with 5.0 g liter⁻¹ dl-lactate as the sole carbon source (5). The whole-cell catalyst was prepared by centrifuging the medium and resuspending the cell pellet, and biotransformation was then carried out under the following conditions using 2-HBA as the substrate and whole cells of P. stutzeri SDM as the biocatalyst: 2-HBA, 10 g liter⁻¹; dry cell concentration, 6 g liter⁻¹; buffer, 100 mM potassium phosphate (pH 7.0); temperature, 30°C; shaking speed, 300 rpm. After 4 h of reaction, the mixture was analyzed by high-performance liquid chromatography (HPLC; Agilent 1100 series; Hewlett-Packard) using a refractive index detector (3). The HPLC system was fitted with a Bio-Rad Aminex HPX-87 H column. The mobile phase consisted of 10 mM H₂SO₄ pumped at 0.4 ml min⁻¹ (55°C). Biotransformation resulted in the production of a compound that had a retention time of 19.57 min, which corresponded to the peak of authentic 2-OBA (see Fig. S1 in the supplemental material).After acidification and vacuum distillation, the new compound was analyzed by negative-ion mass spectroscopy. The molecular ion ([M − H]⁻, m/z 101.1) signal of the compound was consistent with the molecular weight of 2-OBA, i.e., 102.1 (see Fig. S2 in the supplemental material). These results confirmed that 2-HBA was oxidized to 2-OBA by whole cells of P. stutzeri SDM.To investigate whether iLDHs are responsible for 2-OBA production in the above-described biocatalytic process, 2-HBA oxidation activity in P. stutzeri SDM was probed by native polyacrylamide gel electrophoresis. After electrophoresis, the gels were soaked in a substrate solution [50 mM Tris-HCl buffer (pH 8.0) containing 0.1 mM phenazine methosulfate, 0.1 mM 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide, and 1 mM l-lactate, dl-lactate, or dl-2-HBA] and gently shaken. As shown in Fig. ​Fig.1,1, d- and l-iLDH migrated as two bands with distinct mobilities. The activities responsible for d- and l-2-HBA oxidation were located at the same positions as the d- and l-iLDH activities, respectively. No other bands responsible for d- and l-2-HBA oxidation were detected. Moreover, the dialysis of the crude cell extract did not lead to loss of 2-HBA oxidation activity and the addition of 10 mM NAD⁺ could not stimulate the reaction (see Table S1 in the supplemental material). These results implied that in the biocatalytic system, 2-HBA was oxidized to 2-OBA by iLDHs present in P. stutzeri SDM.Open in a separate windowFIG. 1.Activity staining of iLDHs after native polyacrylamide gel electrophoresis with lactate or 2-HBA as the substrate.Although the SDM strain could not use 2-HBA or 2-OBA for growth (see Fig. S3 in the supplemental material), 2-HBA might induce some of the enzymes responsible for 2-OBA production in the biocatalytic process. To exclude this possibility, the SDM strain was cultured in MSM containing dl-lactate or pyruvate as the sole carbon source. As shown in Fig. ​Fig.2,2, the enzyme activities that catalyzed lactate and 2-HBA oxidation were simultaneously present in the cells cultured on lactate and were absent in those cultured on pyruvate. After the lactate or pyruvate was exhausted, 5.05 g liter⁻¹ dl-2-HBA was added to the medium. It was observed that dl-2-HBA was efficiently converted to 2-OBA in the medium containing dl-lactate (Fig. ​(Fig.2a).2a). No 2-OBA production was detected in the medium containing pyruvate. Because 2-HBA addition did not induce the enzymes involved in 2-HBA oxidation (Fig. 2a and b), we concluded that the iLDHs induced by dl-lactate catalyzed 2-HBA oxidation in this biocatalytic process.Open in a separate windowFIG. 2.Time course of P. stutzeri SDM growth on media containing dl-lactate (a) and pyruvate (b). 2-HBA was added to the medium after the exhaustion of lactate or pyruvate. Symbols: ▴, lactate; ▵, pyruvate; •, 2-HBA; ○, 2-OBA; ▪, cell density; ▧, iLDHs activity with dl-lactate as the substrate; ▒, iLDHs activity with dl-2-HBA as the substrate.iLDHs could catalyze the oxidation of the substrate in a flavin-dependent manner and might use membrane quinone as the electron acceptor. Unlike the oxidases, which directly use the oxygen as the electron acceptor, this substrate oxidation mechanism could prevent the formation of H₂O₂ (see Fig. S4 in the supplemental material). The P. stutzeri SDM strain efficiently converted dl-2-HBA to 2-OBA with high yields (4.97 g liter⁻¹ 2-OBA was produced from 5.05 g liter⁻¹ dl-2-HBA); therefore, 2-OBA production by this strain can be a valuable and technically feasible process. To increase the efficiency of P. stutzeri SDM in the biotechnological production of 2-OBA, the conditions for biotransformation using whole cells of P. stutzeri SDM were first optimized. The influence of the reaction pH and 2-HBA concentration on 2-OBA production was determined in 100 mM phosphate buffer containing whole cells harvested from the medium containing dl-lactate as the sole carbon source. The reaction was initiated by adding the whole cells and 2-HBA at 37°C, followed by incubation for 10 min. After stopping the reaction by adding 1 M HCl, the 2-OBA concentration was determined by HPLC.As shown in Fig. ​Fig.3a,3a, ​,2-OBA2-OBA production was highest at pH 7.0. Under acidic or alkaline conditions, the transformation of 2-HBA to 2-OBA decreased. The optimal 2-HBA concentration was found to be 0.4 M, as shown in Fig. ​Fig.3b.3b. 2-OBA production increased as the 2-HBA concentration increased up to about 0.4 M and decreased thereafter. The concentration of the whole-cell catalyst was then optimized using 0.4 M 2-HBA as the substrate at pH 7.0. As shown in Fig. ​Fig.3c,3c, the highest 2-OBA concentration was obtained with 20 g (dry cell weight [DCW]) liter⁻¹ of P. stutzeri SDM. The 2-OBA concentration decreased with any increase beyond this cell concentration.Open in a separate windowFIG. 3.Optimization of the biocatalysis conditions. (a) Effect of pH on 2-OBA production activity. (b) Effect of 2-HBA concentrations on 2-OBA production activity. (c) Effect of the concentration of P. stutzeri SDM on biotransformation. OD, optical density.After optimizing the biocatalytic conditions, we studied the biotechnological production of 2-OBA from 2-HBA by using the whole-cell catalyst P. stutzeri SDM. As shown in Fig. ​Fig.4,4, when 20 g (DCW) liter⁻¹ P. stutzeri SDM was used as the biocatalyst, 48.5 g liter⁻¹ 2-HBA was biotransformed into 44.4 g liter⁻¹ 2-OBA in 24 h.Open in a separate windowFIG. 4.Time course of production of 2-OBA from 2-HBA under the optimum conditions. Symbols: ▪, 2-OBA; •, 2-HBA.Biocatalytic production of 2-OBA was carried out using crotonic acid, propionaldehyde, 1,2-butanediol, or threonine as the substrate (2, 7, 8, 12). Resting cells of the strain Rhodococcus erpi IF0 3730 produced 15.7 g liter⁻¹ 2-OBA from 20 g liter⁻¹ 1,2-butanediol, which is the highest reported yield of 2-OBA to date (8). By using the whole-cell catalyst P. stutzeri SDM, it was possible to produce 2-OBA at a high concentration (44.4 g liter⁻¹) and a high yield (91.5%). Due to the simple composition of the biocatalytic system (see Fig. S5 in the supplemental material), 2-HBA and 2-OBA could be easily separated on a column using a suitable resin. Separation of 2-OBA from the biocatalytic system was relatively inexpensive. The biocatalytic process presented in this report could be a promising alternative for the biotechnological production of 2-OBA.      

Intestinal Endocellular Symbiotic Bacterium of the Macaque Louse Pedicinus obtusus: Distinct Endosymbiont Origins in Anthropoid Primate Lice and the Old World Monkey Louse

A symbiotic bacterium of the macaque louse, Pedicinus obtusus, was characterized. The symbiont constituted a gammaproteobacterial lineage distinct from the symbionts of anthropoid primate lice, localized in the midgut epithelium and the ovaries and exhibiting AT-biased genes and accelerated molecular evolution. The designation “Candidatus Puchtella pedicinophila” was proposed for it.Sucking lice (Insecta: Phthiraptera: Anoplura), ectoparasitic insects that feed exclusively on the blood of their specific mammalian hosts, are generally associated with endosymbiotic bacteria (2, 13). Recent molecular studies have demonstrated that symbiotic bacteria of sucking lice are of multiple evolutionary origins (1, 8, 10, 16).From primates, three louse genera, Pediculus, Pthirus, and Pedicinus, have been recorded. Three Pediculus and two Pthirus species are known from anthropoid primates, harboring gammaproteobacterial “Candidatus Riesia spp.” in the stomach disc (1, 16). Meanwhile, 14 Pedicinus species, whose symbiotic bacteria have not been characterized, have been recorded from Old World monkeys (3).Samples of Pedicinus obtusus were collected in 2008 at the Jigokudani Monkey Park, Nagano, Japan, where a wild population of the Japanese macaque, Macaca fuscata, is exhibited to the public under careful control. When animals were caught and inspected, samples of P. obtusus were collected from the anesthetized animals under the supervision of a veterinary doctor (permission number 19-26-24, Nagano Prefecture, Japan). Samples of P. obtusus were collected in November 2008 by courtesy of Kiyoyasu Kowaki from a subspecies of the Japanese macaque, M. fuscata yakui, that is endemic in Yakushima Island, Kagoshima, Japan. Each of the insect samples was subjected to DNA extraction, and a 1.5-kb segment of the 16S rRNA gene (5) and a 1.6-kb segment of the groEL gene (7) were amplified by PCR, cloned, and sequenced. Molecular phylogenetic analyses were performed using the programs PAUP 4.0b10 (Sinauer Associates, Sunderland, MA), RAxML version 7.0.0 (17), and MrBayes 3.1.2 (15).The 1,490-bp 16S rRNA gene sequences formed a distinct lineage in the Gammaproteobacteria, exhibiting no phylogenetic affinity to the symbionts of other louse species, including those of human lice and the chimpanzee louse (see Fig. S1 in the supplemental material). The 1,601-bp groEL gene sequences also constituted a gammaproteobacterial lineage, exhibiting no phylogenetic affinity to the symbiont of human lice (Fig. ​(Fig.1)1) These results indicated that the endosymbiotic bacterium of the Old World monkey louse evolved independently of the endosymbiotic bacteria of the anthropoid primate lice. Considering that the date of divergence of Old World monkeys and anthropoid primates has been inferred as 23 to 31 million years ago, the endosymbiotic evolution in the primate lice must have occurred within this time scale (12).Open in a separate windowFIG. 1.groEL gene sequence-based molecular phylogenetic analysis of the symbiont of P. obtusus in the Gammaproteobacteria. A maximum-likelihood tree inferred from 1,040 unambiguously aligned nucleotide sites is shown. Bayesian and neighbor-joining analyses gave essentially the same results (data not shown). Statistical support values higher than 70% are indicated at the nodes in the order of maximum-likelihood/Bayesian/neighbor-joining values. Asterisks indicate statistical support values lower than 70%. Sequence accession numbers are shown in brackets. AT contents of the sequences are also shown. The sequences from the symbionts of human and monkey lice are highlighted in boldface. P-symbiont, primary symbiont; S-symbiont, secondary symbiont.The molecular evolutionary rates of the symbiont gene sequences were analyzed with a relative rate test using the program RRTree (14). The evolutionary rates of the 16S rRNA and groEL gene sequences in the lineage of the P. obtusus symbiont were significantly higher than those in the lineages of related free-living gammaproteobacteria (see Table S1 in the supplemental material).The AT contents of the 16S rRNA and groEL gene sequences of the P. obtusus symbiont were 53.5% and 64.8%, respectively. These values were equivalent to those of obligate endosymbionts of other insects (>50% for the 16S rRNA gene and >60% for the groEL gene) and were remarkably higher than those of allied free-living gammaproteobacteria (∼45% for the 16S rRNA gene and 45 to 50% for the groEL gene) (Fig. ​(Fig.1;1; see Fig. S1 in the supplemental material).Obligate endosymbiotic bacteria that cospeciate with their host insects commonly exhibit peculiar genetic traits, including AT-biased nucleotide composition, an accelerated rate of molecular evolution, and significant genome reduction (18). The AT-biased nucleotide composition and the accelerated evolution observed with the P. obtusus symbiont (Fig. ​(Fig.1;1; see Fig. S1 and Table S1 in the supplemental material) are suggestive of a stable and intimate host-symbiont association over evolutionary time.Fluorescent in situ hybridization targeting 16S rRNA of the symbiont was performed using the Alexa Fluor 555-labeled oligonucleotide probes Al555-ML439 (Al555-5′-ATAATATCTTCTTTCCTACCGA-3′) and Al555-ML1256 (Al555-5′-GCTAATTCTTGCGAATTTGCTT-3′) as described previously (9). In first-, second-, and third-instar nymphs, the symbiont signals were detected in the posterior half of the stomach in the abdomen (Fig. ​(Fig.2A).2A). In the posterior stomach, the signals exhibited a periodical and striated pattern (Fig. ​(Fig.2B).2B). Magnified images located the symbiont signals endocellularly in the intestinal wall tissue (Fig. ​(Fig.2C).2C). In some of the third-instar nymphs, the symbiont signals were found not only in the posterior stomach but also in the ovaries (Fig. ​(Fig.2A).2A). In adult females, the symbiont signals were restricted to the lateral oviducts (Fig. ​(Fig.2A),2A), where many bacteriocytes formed a pair of symbiotic organs called ovarial ampullae (Fig. ​(Fig.2D).2D). The ovarial ampullae were located adjacent to developing oocytes in the ovarioles, where the symbiont was passed to the developing eggs (Fig. ​(Fig.2E2E).Open in a separate windowFIG. 2.Localization of the symbiont in nymphs and adults of P. obtusus. (A) General localization of the symbiont in nymphal and adult insects. The symbiont signals are seen in the posterior stomach (green arrows) and the ovarial ampullae (yellow arrows). (B) An image of the posterior stomach of a second-instar nymph. Periodical and striped regions are densely populated by the symbiont. (C) An image of the posterior stomach of a third-instar nymph. The symbiont signals are restricted to the gut epithelial cells and not detected in the stomach lumen. Bacteriocytes are intercalated with uninfected cells, forming a striated pattern. (D) An enlarged image of the ovarial ampulla, consisting of a number of uninucleated bacteriocytes. (E) An image of the ovary in an adult female, wherein developing oocytes are found in the ovarioles. Ovarial ampullae (yellow arrows) are located at the anterior tip of the lateral oviducts, where symbiont transmission to oocytes occurs (white arrow). Panel A shows epifluorescent images, while panels B to E are confocal optical sectioning images. Red and blue signals reflect the symbiont 16S rRNA and the host nuclear DNA, respectively. Abbreviations: lu, stomach lumen; oc, oocyte.These results suggested that in third-instar female nymphs of P. obtusus, the symbiont localization drastically changes, from the posterior stomach to the ovarial ampullae (Fig. ​(Fig.2A).2A). Presumably, the endocellular symbiont escapes the host cells and somehow moves to the female reproductive organ, establishing a new endocellular association and finding a way to the next host generation. Interestingly, such symbiont migrations from a symbiotic organ to the ovaries in third-instar female nymphs have been reported for the human body louse Pediculus humanus (4, 13) and the slender pigeon louse Columbicola columbae (6, 13). Here it is notable that the symbiotic bacteria of P. obtusus, P. humanus, and C. columbae are phylogenetically not related to each other (see Fig. S1 in the supplemental material). The mechanisms underpinning the symbiont migration are currently unknown. Eberle and McLean (4) suggested by a series of experiments that the ovary of the human body louse might emanate a humoral factor that attracts the symbiotic bacteria to induce the migration. Whether or not this hypothesis is true deserves future experimental studies of these louse species and their symbiotic bacteria.On the basis of these results, we propose the designation “Candidatus Puchtella pedicinophila” for the novel endosymbiont lineage. The generic name honors Otto Puchta, who identified the biological role of the human louse symbiont as the provision of B vitamins (11). The specific name indicates association with a Pedicinus monkey louse. Whether the other monkey lice harbor symbiotic bacteria allied to the symbiont of P. obtusus deserves future studies.     

High-Resolution Cryo-Electron Microscopy Structures of Murine Norovirus 1 and Rabbit Hemorrhagic Disease Virus Reveal Marked Flexibility in the Receptor Binding Domains

Our previous structural studies on intact, infectious murine norovirus 1 (MNV-1) virions demonstrated that the receptor binding protruding (P) domains are lifted off the inner shell of the virus. Here, the three-dimensional (3D) reconstructions of recombinant rabbit hemorrhagic disease virus (rRHDV) virus-like particles (VLPs) and intact MNV-1 were determined to ∼8-Å resolution. rRHDV also has a raised P domain, and therefore, this conformation is independent of infectivity and genus. The atomic structure of the MNV-1 P domain was used to interpret the MNV-1 reconstruction. Connections between the P and shell domains and between the floating P domains were modeled. This observed P-domain flexibility likely facilitates virus-host receptor interactions.Murine norovirus 1 (MNV-1) (3, 14, 15) and rabbit hemorrhagic disease virus (RHDV) are members of the genera Norovirus and Lagovirus of the family Caliciviridae that offer a comparison to recombinant human norovirus (rNV) virus-like particles (VLPs) for assessing the structures and roles of domains within the capsid proteins of this family of viruses. Calicivirus particles contain 180 copies of the 56- to 76-kDa major capsid protein (Orf2), which is comprised of the internal/buried N terminus (N), shell (S), and protruding (P) domains (9, 10). The S domain, an eight-stranded β-barrel, forms an ∼300-Å contiguous shell around the RNA genome. A flexible hinge connects the shell to a “protruding” (P) domain at the C-terminal half of the capsid protein, which can be further divided into a globular head region (P2) and a stem region (P1) that connects the shell domain to P2. The accompanying article (13) describes the determination of the structure of the P domain of MNV-1 to a resolution of 2.0 Å.We recently determined the cryo-transmission electron microscopy (TEM) structure of MNV-1 to ∼12-Å resolution (4) and found that, compared to rNV VLPs (10) and San Miguel sea lion virus (SMSV) (1, 2), the protruding domains are rotated by ∼40° in a clockwise fashion and lifted up by ∼16 Å. To better understand the unusual conformation of MNV-1 and whether it is unique to this particular member of the calicivirus family, the ∼8-Å cryo-TEM structures of infectious MNV-1 and the VLPs of RHDV were determined.MNV-1 was produced as previously described (4). Three liters of cell culture yielded 0.5 to 1.0 mg of purified virus with a particle/PFU ratio of less than 100. Baculovirus expression and purification of recombinant RHDV (rRHDV) VLPs were performed as previously described (8). Cryo-electron microscopy (EM) data were collected at the National Resource for Automated Molecular Microscopy (NRAMM) facility in San Diego, CA (4). Images were collected at a nominal magnification of ×50,000 at a pixel size of 0.1547 nm at the specimen level using Leginon software (12) and processed with Appion software (5). The contrast transfer function for each set of particles from each image was estimated and corrected using ACE2 (a variation of ACE [7]). Particle images were automatically selected (11). The final stacks of particle images contained 20,425 MNV virions and 7,856 rRHDV VLPs, and EMAN 3D (6) was used for the reconstructions. Resolutions were estimated by Fourier shell correlations (FSC) of the three-dimensional (3D) reconstructions and application of a cutoff of 0.5. An amplitude correction of the final electron density was performed using GroEL small-angle X-ray scattering (SAXS) data.3D reconstructions of MNV-1 and rRHDV were calculated to resolutions of 8 Å and 8.1 Å, respectively (Fig. ​(Fig.1).1). The P domains of rNV VLPs rest directly on top of the shell domain (10) (Fig. ​(Fig.1A).1A). In contrast, the P domains of MNV-1 are lifted and rotated above the shell of the capsid (4) (Fig. ​(Fig.1B).1B). At this higher resolution, there was a clear connection between the P1 domain and the shell domain in all three capsid subunits (Fig. ​(Fig.1B,1B, arrow A). Unlike the smooth protruding domains of rNV, MNV-1 has two clear “horns” (arrow B), not dissimilar to those observed for the sapoviruses (1, 2). There also are islands of density in the interior of the shell, directly beneath the 5-fold axes, that may represent ordered regions of RNA.Open in a separate windowFIG. 1.Stereo diagrams (left) and thin sections (right), with radius coloring, of rNV (A), MNV-1 (B), and an rRHDV VLP (C). For rNV, the atomic coordinates (10) were used. In MNV, arrow A indicates the thin connector between the P1 and S domains. Arrow B denotes the horns found at the tips of the P2 domains. Arrow C denotes the large gap between the P1 and S domains in the rRHDV VLP. Arrow D denotes the false connectivity in rRHDV VLPs between the P1 domain and the S domain near the 5-fold axes.As with MNV-1, there is a marked gap between the P and S domains in the rRHDV VLP (Fig. ​(Fig.1C,1C, arrow C). This gap is not as pronounced as in MNV-1 because the P domains are not rotated as in MNV-1. In this electron density map, the A/B dimers appear to be touching the shell domain near the 5-fold axes. This contact difference between the A/B dimers and the C/C dimers could be the reason why the tops of the C/C dimers appear to be markedly disordered compared to the A/B dimers in rRHDV and the C/C dimers in MNV-1.Shown in Fig. ​Fig.22 is the fitting of the atomic structures of the MNV-1 P domain (13) and the rNV S domains into the MNV-1 3D reconstruction electron density. The horns (arrow A, loops A′-B′ and E′-F′) observed at the tips of the P domain match exceedingly well with the electron density. As discussed in the accompanying publication (13), the A′-B′ and E′-F′ loops displayed two discrete conformations, a closed structure, where the two loops were tightly associated, and an open structure, where the loops were splayed apart. The horns of the closed conformation fit better into the reconstruction, as the E′-F′ loop in the open form jutted out of the density at the base of the horns. The unmodified density in the lower panel of Fig. ​Fig.22 shows fine features in the shell domain and a very clear connection between the shell and P1 domains. The connections between the P1 and S domains were of sufficient quality to build a basic backbone model by uncoiling the linker region (arrow B). The P domain in the unfiltered 3D reconstruction was far less ordered than the S domain (Fig. ​(Fig.2).2). This was likely due to movement of the entire P domain with respect to the shell.Open in a separate windowFIG. 2.Fitting of the MNV-1 P domain and the rNV shell domain into the MNV-1 electron density. A, B, and C subunits are represented by blue, green, and red, respectively. The electron density is shown in transparent gray. The top panel is the 8.0-Å-resolution 3D reconstruction modified using a low-pass filter. The bottom panel is the reconstruction without modification. The horns on the tops of the P domains are denoted by arrow A. Arrow B denotes the connection between the S and P domains.Using the structure of rNV VLP P domains for modeling, the rRHDV P domains are lifted off the surface of the shell, but not rotated as with MNV-1. This places the bottom edge of the A subunit P1 domain near the S domain at the 5-fold axes. The P-domain dimers of rNV and rRHDV have a more “arch-like” shape than MNV-1. Unlike in MNV-1, the electron densities of the C/C dimers in rRHDV are far more diffuse than those of the A/B dimers (Fig. ​(Fig.3B)3B) and the connector between the S and P1 domains is not clear. During fitting, the connector region was not as extended as with MNV-1. This may afford greater flexibility, leading to more diffuse electron density.Open in a separate windowFIG. 3.Fitting of the rNV atomic structure into the rRHDV VLP electron density. The upper stereo image shows the 8.1-Å-resolution 3D reconstruction after modification by a low-pass filter. Below is the same reconstruction prior to density modification.When the atomic models for the MNV-1 P domains (13) were placed into the cryo-TEM electron density (Fig. ​(Fig.4),4), the C termini extended deep into the cores of adjacent P domains. Possible connections not accounted for by the P-domain structures were also observed in the electron density between the P domains. A bulge between the P1 and P2 domains in the 3D reconstruction indicated a possible interaction between the C termini and the adjacent P domains. These same interactions were observed in the crystal lattice. This highly mobile C terminus may be a flexible tether between the P domains in the intact virion.Open in a separate windowFIG. 4.Possible carboxyl-terminus interactions between the P domains of MNV-1. (A) Stereo image of MNV-1 calculated to 12-Å resolution with (red) and without (yellow) the last 10 residues of the P domain. (B) The calculated MNV-1 density with the carboxyl terminus removed (yellow) overlaid onto the 3D reconstruction of MNV-1 (blue). Note the strands of difference density that roughly correspond to the C terminus in panel A. (C) The C-terminus interactions observed in the structure of the MNV-1 P domains. Shown in blue and green are ribbon diagrams of an A/B P-domain dimer. In mauve is a surface rendering of the C terminus from a crystallographically related dimer. (D) Surface rendering of the final MNV-1 model with possible interactions between the P domains in MNV-1. The carboxyl termini of the A subunits (blue) interact with the counterclockwise-related B subunits around the 5-fold axes (white arrows). Around the 3-fold (quasi-6-fold) axes, the C subunits interact with the A subunits and the B subunits interact with the C subunits (orange arrows).It is absolutely clear that the hinge region between the S and P domains affords a remarkable degree of flexibility in the P domains that is not genus specific or related to differences between rVLPs and authentic virions. The simplest explanation for the role of this transition is that it gives the P domains flexibility that may be used to optimize interactions with cell receptors during attachment and entry. In this way, the P domains can increase their avidity for the cell surface by being more facile in adapting to the presentation of cellular recognition motifs.     

On Relevance of Codon Usage to Expression of Synthetic and Natural Genes in Escherichia coli

A recent investigation concluded that codon bias did not affect expression of green fluorescent protein (GFP) variants in Escherichia coli, while stability of an mRNA secondary structure near the 5′ end played a dominant role. We demonstrate that combining the two variables using regression trees or support vector regression yields a biologically plausible model with better support in the GFP data set and in other experimental data: codon usage is relevant for protein levels if the 5′ mRNA structures are not strong. Natural E. coli genes had weaker 5′ mRNA structures than the examined set of GFP variants and did not exhibit a correlation between the folding free energy of 5′ mRNA structures and protein expression.IN genomes, natural selection may act on silent sites of codons to make translation of highly expressed genes more efficient, an effect linked primarily to abundances of tRNA isoacceptor molecules (Ikemura 1985; Bulmer 1987; Kanaya et al. 1999). Codon choice may also be linked to formation of secondary structures in mRNA that reduce protein levels, as has been shown with haplotypes of the human COMT gene (Nackley et al. 2006). Kudla et al. (2009) have recently reported an experiment that contributes toward understanding how synonymous codon usage shapes gene expression. They have constructed a library of 154 synthetic variants of a green fluorescent protein (GFP) gene that varied randomly at synonymous sites while retaining the original amino acid sequence. The authors concluded that codon usage (CU) bias did not correlate with protein levels measured as fluorescence of the GFP, but also that the minimum free energy of a mRNA secondary structure in a 42-nucleotide region at [−4,37] that overlaps the start codon (“hairpin stability”) bears a great significance. CU bias was quantified by the widely used codon adaptation index (CAI) method (Sharp and Li 1987), essentially a measure of the distance of a gene''s codon usage to the codon usage of a predefined set of highly expressed genes. The CAI and some of its more recent alternatives, such as measure independent of length and composition (MILC) (Supek and Vlahovicek 2005), have been shown to be a viable surrogate for gene expression in various unicellular organisms. Additionally, in a multiple linear regression of rank fluorescence against a number of sequence-derived attributes, including CAI and the abovementioned hairpin stability, Kudla et al. (2009) did not find CAI to contribute significantly toward the prediction of protein levels, in contrast to the hairpin stability.

Both the codon adaptation index and the 5′ mRNA secondary structures influence protein levels in the Kudla et al. data:

The described statistical analyses, however, failed to address the case in which a nonlinear three-way dependency between hairpin stability, codon usage, and fluorescence might exist; data are visualized in Figure 1, A–C, and in figure 2B in Kudla et al. Such complex patterns in data are readily captured by the support vector machines (SVM) algorithm, reviewed in Noble (2006) and Ben-Hur et al. (2008). We have employed the SVM with a radial basis function kernel to regress fluorescence against both hairpin stability and CAI simultaneously (Figure 1B) and computed the Pearson''s correlation coefficient in cross-validation (here denoted as Q) between true and predicted values of fluorescence (See File S1). A linear model based solely on hairpin stability as employed by Kudla et al. (Figure 1A) can explain Q² = 38.6% of variance in protein levels, while the nonlinear SVM regression that takes CAI into account explains Q² = 52.2% of variance. The difference in Q is statistically significant at P = 10⁻¹⁹⁰ (paired t-test). Note that Kudla et al. utilize the Spearman rank correlation coefficient (ρ) in their article; the hairpin stability would explain ρ² = 44.6% of the variance in expression levels if the requirement for a linear relationship was abandoned in this manner.Open in a separate windowFigure 1.—Regression of protein levels against folding free energy of an mRNA hairpin at nucleotides −4 through 37 (A), against the hairpin free energy and the codon adaptation index (Sharp and Li 1987) (B and C), or against the hairpin free energy and the codon frequencies (D and E). The colors show the measured protein levels, while the background shading reflects the protein levels predicted by the specific model. (A) Predictions by linear regression. (B and E) Predictions by a support vector machine with a radial basis function kernel. (C) Predictions by an M5′ regression tree. (D) A schematic of the M5′ model, where coefficients in the terminal nodes are derived from data where protein levels, all codon frequencies, and hairpin free energies were normalized to [0,1] to facilitate comparison between the influence of codons, the hairpin stability, and the constant in the regression equation. All coefficients ≥0.1 are in boldface type. In the plots (A–C and E), a slight amount of random “jitter” was introduced to the point positions (at most, 3% of the range of each axis) to better visualize overlapping points. In the plot in E, a single outlying point is not shown. See Figure S2 for the same plots without jitter and with the outlier in E included. R² is the squared Pearson''s correlation coefficient between actual and model-predicted protein levels; Q² is similar, but obtained in cross-validation (10-fold, 100 runs), and is a more conservative estimate of regression accuracy.Open in a separate windowFigure 2.—The distributions of RNA folding free energies of a 42-nucleotide window in the mRNA between positions −4 and +37, where the “A” in the “AUG” start codon has index zero. The distributions are shown separately for the 154 gene variants from Kudla et al. (2009) and for the genes from the E. coli K12 genome. The dotted line indicates the 5th percentile of the E. coli values at −10.9 kcal/mol.Compared to the SVM, a more interpretable generalization of the data can be achieved by a different nonlinear regression approach, the M5′ tree (Wang and Witten 1997), which recursively divides the data to reduce the variance of the dependant variable within each partition and then builds separate linear models for the partitions. The resulting regression tree (Figure 1C; supporting information, Figure S1) better explains the correlation between protein levels on one side and hairpin stability and CAI on the other side when compared to a linear model employed by Kudla et al. that regresses protein levels against hairpin stability only [see figure 2B in Kudla et al. (2009) and Figure 1A]; 9.3% more variance is explained by the M5′, P = 10⁻⁹¹ (paired t-test). An interpretation that follows from the general structure of the M5′ tree (Figure S1) is that, at high mRNA hairpin stability, protein levels will generally be quite low and not dependant on CAI; in contrast, with less stable mRNA hairpins, both hairpin stability and CAI play a role in determining protein levels. In the interpretation of the M5′ tree structure, we would place less emphasis on the exact coefficients of the linear models in the leaves because the reliability of these fine-grained features of the M5′ model can strongly depend on the good coverage of all parts of the mRNA–CAI space data points.

The CAI may not be an optimal summary of codon usage for predicting expression of overexpressed genes:

Regarding use of CAI in the present context, it should be noted that CAI''s original purpose was to serve as a proxy for gene expression in conditions of abundance that result in fast growth in the organism''s environmental niche. The CAI or related approaches (Supek and Vlahovicek 2005) may not, however, be an ideal representation of codon usage when examining overexpression of a foreign protein at levels that exceed the natural abundances of the host''s most highly expressed proteins. This was indeed shown to be the case in a recent article by Welch et al. (2009) in which the authors reported an experiment with heterologous expression of variants of two proteins in E. coli: an antibody fragment and a phage DNA polymerase. Welch et al. found that codon frequencies in general, but not CAI specifically, correlated well with protein levels and postulated that for overexpressed proteins optimal codons would correspond to the codons translated efficiently under amino acid starvation (Elf et al. 2003; Dittmar et al. 2005). Analogously to Welch et al., we now apply our regression algorithms not to the CAI, but directly to the codon frequencies that CAI attempts to summarize in the Kudla et al. data (See File S2). An M5′ regression tree trained on the hairpin stability and codon frequencies (Figure 1D) explains 10.6% more variance (P = 10⁻⁸³, paired t-test) in protein levels than an M5′ tree trained on hairpin stability and CAI (Figure 1C, Figure S1). A SVM regression model trained on the hairpin stability and a simple linear combination of selected codon frequencies (Figure 1E) explains 8.8% more variance (P = 10⁻⁸², paired t-test) than the SVM that uses CAI (Figure 1B). An SVM trained on the hairpin stability and the full set of codon frequencies (not shown in Figure 1) explains Q² = 65.0% of variance in the protein abundances, a sizable increase (P ≈ 10⁻²⁶⁰, paired t-test) compared to a linear regression on solely the [−4,37] hairpin stability (Q² = 38.6%) as originally employed by Kudla et al. and also as compared to a set of randomized controls (Q² = 20.1–30.7%; Table S1). Therefore, not relying on a predefined notion of codon optimality—as embodied in the CAI—further strengthens the argument that the correlation of CU and protein levels is far from negligible in this data set.Additionally, we found some correlation between codon frequencies and 5′ mRNA hairpin stability in the Kudla et al. gene variants (Figure S4). The fact that the two factors were not completely independent adds weight to the relevance of CU to protein levels since one could not be certain that even the variance in protein levels explained by 5′ mRNA structures is wholly due to the structures themselves and not to the confounding variables—here, the codon frequencies.The M5′ tree trained on codon frequencies (Figure 1D) follows the same general structure as the M5′ tree trained on the CAI (Figure S1) where the codon frequencies become relevant with mRNA hairpins weaker than −9.75 kcal/mol, while with stronger [−4,37] mRNA hairpins protein levels are generally low. Our interpretation is that the lack of a stable secondary structure that could obstruct translational initiation is a necessary but not a sufficient condition for high protein expression. When the initiation phase is unhindered, the bottleneck would shift to the elongation phase in which codon optimality plays an important role. In the literature, theoretical models of translation may consider either the initiation (Bulmer 1991) or the elongation phase (Xia 1998) as the rate-limiting step of translation under physiological conditions; we are not aware of such analyses describing translation of artificially overexpressed genes.The codons identified as relevant by our M5′ model of the Kudla et al. data are different from, but not inconsistent with, those proposed by Welch et al. (Table S2). We anticipate that the rules for codon optimality for overexpression in an Escherichia coli host will become better defined as more large-scale experiments, such as the two discussed here (Kudla et al. 2009; Welch et al. 2009), are carried out.

The “RNA structure + codon usage” model agrees with independent experimental data and is robust to removal of extreme values:

Our reanalysis of the Kudla et al. data should be viewed in light of the conclusions of Welch et al. (2009) who find that codon usage, but not the 5′ hairpin stability, correlates with protein levels in their data, while noting that their gene variants generally have considerably weaker 5′ mRNA hairpins than the sequences in Kudla et al. Welch et al. reconcile the different outcomes of the two experiments by noting that “inhibition of initiation by especially strong mRNA structure would obscure effects resulting from factors that influence elongation, such as codon usage” (page 9). Here we propose that precisely the same model can be derived solely from the Kudla et al. data. Furthermore, we find that the 154 gene variants from Kudla et al. indeed do have unusually stable 5′ mRNA hairpins (mean free energy = −9.68 kcal/mol) in comparison to natural E. coli genes (mean free energy = −6.15 kcal/mol) (P = 10⁻³⁸ by Mann–Whitney U-test; see Figure 2). The part of the distribution of Kudla et al. gene variants that overlaps with the bulk of the E. coli genes, with 5′ mRNA hairpin free energies lower than ∼ −10 kcal/mol, corresponds to the range where our M5′ model indicates a stronger influence of CU on protein levels (Figure S1, Figure 1D).We investigate to what extent the presence of a group of sequences extreme in their 5′ mRNA hairpin stabilities in the Kudla et al. data set (left peak in Figure 2) influenced the authors'' conclusion that the hairpin stabilities have an overarching influence on protein levels. After removing the sequences below the 5th percentile of the E. coli natural hairpin stabilities (−10.9 kcal/mol), we were left with 109 of the original 154 Kudla et al. sequences. The accuracy of regressing protein levels against mRNA hairpin stability deteriorates greatly (Q² = 18.5%) after removing the 45 sequences, but less so with SVM and M5′ regression that take into account both CU and the hairpin stability (udla et al. basically captured the difference between these extreme cases—in which very strong 5′ mRNA secondary structures blocked expression—and all other sequences. However, to explain the variation in protein levels within the nonextreme set, hairpin stabilities by themselves are not sufficient and need to be complemented with CU.

TABLE 1

Accuracy of the regression of protein levels against 5′ mRNA hairpin stability or against 5′ mRNA hairpin stability and codon frequencies

Data set	Linear regression, hairpin stability only (%)	SVM, hairpin stability + codon frequencies (%)	M5′, hairpin stability + codon frequencies (%)
Full (n = 154)	38.6	65.0	56.7
No strong hairpins (n = 109)	18.5	53.0	40.4

Open in a separate windowThe cross-validation correlation coefficient squared (Q²) is compared with the full Kudla et al. data set (154 proteins) and the reduced data set (109 proteins) where mRNA hairpin folding energies are ≥ −10.9 kcal/mol, the 5th percentile of natural E. coli genes.In addition to measuring protein levels in the 154-sequence data set, Kudla et al. performed an additional experiment where an unstructured 28-codon tag was fused to 5′ ends of 72 (of 154) GFP sequence variants. Adding the tag was found to enhance protein levels, supporting the conclusion of Kudla et al. that 5′ structure of mRNA had a strong influence on protein production. After an analysis of the data, we found (see File S3) that data from this specific experiment are not well suited to serve as a direct verification of our existing M5′ and SVM regression models. Still, we can compare the protein level predictions of our existing SVM model on the same set of sequences before and after adding the unstructured tag. We found that the predicted expression levels have increased for 67 of 72 sequences (Table S3) after adding the tag that fixes 5′ mRNA folding energy at a weak −6.1 kcal/mol, a result consistent with the Kudla et al. experiment. Additionally, we have trained a new SVM regression model only on the tagged 72-sequence set (See File S2) and found that, within this set, SVM regression can again predict GFP levels solely from codon usage (5′ mRNA structure is invariant among these sequences) at Q² = 37.7%. This amount of variance is similar, or even somewhat larger than, the difference in the variance explained by mRNA vs. mRNA+codons (38.6% vs. 65.0%) in the original data. Therefore, codon usage is of similar importance in shaping the protein levels within the tagged 72-sequence set, as it was in the original 154-sequence set.

mRNA 5′ end secondary structure stabilities do not correlate with protein levels for natural E.

coli genes: To further verify our proposed model, we analyzed the relative contributions of mRNA hairpin stabilities and CU on expression levels of natural E. coli genes (See File S2). If the hairpin stabilities were limiting for expression in the range of folding free energies spanned by the E. coli mRNAs, one would expect to see a correlation between the free energy of mRNA 5′ end folding and the abundance of the corresponding protein. We found no such correlation using the folding free energies of the [−4,37] mRNA region (Figure 3) or equal-sized regions centered around the start codon at [−20,21] or on the expected location of a Shine–Dalgarno sequence (Shultzaberger et al. 2001) at [−30,11] (see Figure S3). Unsurprisingly, CAI correlated well with protein levels (Figure 3) in all examined experimental data sets (Lopez-Campistrous et al. 2005; Lu et al. 2007; Ishihama et al. 2008). Therefore, within the boundaries of the mRNA folding free energies spanned by E. coli genes, the CU plays a dominant role in shaping gene expression (or the CU may possibly be shaped by the expression; see Concluding remarks). As for the stronger mRNA hairpins with < −11 kcal/mol, they are present in the Kudla et al. data, but are very rare in the E. coli genome, which could be explained by one of two scenarios: (i) Above a certain threshold, the mRNA hairpin stability may become so detrimental to expression that all the mutants having such hairpins are subject to very strong negative selection and therefore are absent from the genome. And/or (ii) the Kudla et al. data set may not be representative of the genes in the E. coli genome or the mutational processes they undergo; for example, the amino acid sequence of the GFP''s beginning might be unusually conducive to forming RNA hairpins. Unless further analyses prove differently, it seems reasonable to surmise that in natural E. coli genes mRNA secondary structures would shape expression if they were highly stable, consistent with the finding of a universal (albeit not particularly strong) trend toward avoidance of 5′ mRNA structures in genomes (Gu et al. 2010). However, it can also be concluded at this point—and with more confidence—that at lower secondary structure stabilities the CU has an overarching influence on expression. Such a model of expression-related gene sequence determinants in E. coli is fully consistent with our interpretation of the M5′ regression tree that we have derived from the Kudla et al. data.Open in a separate windowFigure 3.—Correlations between the E. coli absolute protein abundances measured in three independent experiments (Lopez-Campistrous et al. 2005; Lu et al. 2007; Ishihama et al. 2008) and the codon adaptation index (CAI) or the free energy of folding of a secondary structure in the mRNA [−4,37] region (in kcal/mol; more negative values denote a more stable RNA secondary structure). “ρ” is the Spearman''s rank correlation coefficient.

Concluding remarks:

We argue that Kudla et al. worked with a set of gene sequences in which strong mRNA secondary structures (that effectively abolished expression) were frequent enough to mask the relevance of codon frequencies on protein levels when examined only with linear regression methods. While mRNA secondary structures can certainly occur when designing synthetic genes, it is highly questionable to what extent Kudla et al.''s conclusion that CU is of little importance for expression would be generally valid for biotechnological applications, especially since we have shown that the influence of CU is nevertheless present even in the Kudla et al. data. What is beyond doubt, however, is that a strong 5′ mRNA secondary structure can be a roadblock in heterologous expression, and therefore the synthetic gene variants harboring such structures should be avoided. The more specific rules regarding the exact location of the hairpin on the gene sequence, the hairpin''s length, or the tolerable levels of folding free energy will have to be established by further experimentation.A recent algorithm for estimating the efficiency of ribosomal binding sites from the mRNA sequence (Salis et al. 2009) explicitly takes into account the folding free energy of RNA secondary structures, along with other factors. When protein overexpression is desired, the conclusions of Welch et al. and (by our reanalysis) the Kudla et al. data indicate that CU should be optimized in addition to the ribosome binding site sequence to ensure that both initiation and elongation phases of translation are free of impediments.On the basis of their results, Kudla et al. also discuss the evolutionary link between the CU of natural genes and the expression levels of proteins for which they code. They propose that selection for translational efficiency acts at a global level in cells; the codons that accelerate elongation would be preferred in a highly expressed gene not because they facilitate production of that particular protein, but to free up ribosomes for the rate-determining initiation phase of translation of the total cellular mRNA pool. Effectively, the flow of causality between CU and expression would be reversed in comparison to the established view. This hypothesis should be critically reevaluated because it depends on the assertion that manipulating a gene''s CU cannot cause protein levels to increase, an assertion poorly supported by the Kudla et al. data.     

Actin-Based Motility of Burkholderia thailandensis Requires a Central Acidic Domain of BimA That Recruits and Activates the Cellular Arp2/3 Complex

Burkholderia species use BimA for intracellular actin-based motility. Uniquely, Burkholderia thailandensis BimA harbors a central and acidic (CA) domain. The CA domain was required for actin-based motility, binding to the cellular Arp2/3 complex, and Arp2/3-dependent polymerization of actin monomers. Our data reveal distinct strategies for actin-based motility among Burkholderia species.In common with selected species of Listeria, Shigella, Rickettsia, and Mycobacterium, some members of the genus Burkholderia are capable of intracellular actin-based motility (reviewed in reference 9). Such motility promotes cell-to-cell spread in the absence of immune surveillance and, in some cases, escape from autophagy. The melioidosis pathogen Burkholderia pseudomallei forms actin-rich bacterium-containing membrane protrusions (5), in a manner that requires BimA (Burkholderia intracellular motility A) (11). B. pseudomallei BimA exhibits C-terminal homology to the Yersinia autosecreted adhesin YadA and possesses motifs associated with actin binding, including WASP homology 2 (WH2) domains and proline-rich motifs (11). Actin-based motility is also a feature of infection by the glanders pathogen Burkholderia mallei and the avirulent B. pseudomallei-like species Burkholderia thailandensis (10). BimA homologs exist in these species that can compensate for the actin-based motility defect of a B. pseudomallei bimA mutant (10), and mutation of B. mallei bimA results in loss of function (7). BimA homologs from B. mallei and B. thailandensis (BimA_th) differ markedly in primary sequence from the B. pseudomallei protein (BimA_ps) and each other (10), raising the possibility that they may act in distinct ways. Intraspecies conservation of BimA is high in natural populations of B. pseudomallei, with the exception of a geographically restricted B. mallei-like BimA variant (8).Mechanisms of bacterial actin-based motility converge on activation of the Arp2/3 (actin-related protein 2/3) complex. Activation of Arp2/3 requires cellular nucleation-promoting factors (NPFs) such as Wiskott-Aldrich syndrome protein (WASP) family members, and pathogens capable of actin-based motility often mimic the activity of NPFs or recruit and activate them at the bacterial pole (reviewed in references 3 and 9). Arp2/3 is localized throughout B. pseudomallei-induced actin tails (2); however, the role that it plays in actin-based motility is unclear, as expression of an inhibitory domain of the cellular NPF Scar1 does not interfere with actin-based motility of B. pseudomallei (2). Moreover, BimA_ps can polymerize actin in vitro in an Arp2/3-independent manner (11).Recruitment and activation of the Arp2/3 complex by cellular and pathogen-associated NPFs require one or more WH2 domains and an amphipathic central and acidic (CA) domain (3, 6). Analysis of the primary sequence of BimA homologs revealed a CA domain in B. thailandensis BimA that was absent in the BimA proteins of other Burkholderia species and conserved relative to WASP family members, Listeria ActA, and Rickettsia RickA (10). Here we surveyed the conservation of the CA domain and probed its role in actin-based motility and the binding and activation of the Arp2/3 complex.Primers were designed to amplify an 87-bp region of the CA domain (corresponding to amino acid residues 102 to 130 inclusive) of the bimA gene of the sequenced B. thailandensis strain E264 (5′-AGGCGGGTAATCGACTCA-3′ and 5′-TTCGTCGTCCGACCATGA-3′). These primers, and universal bimA-specific primers (8), were used to screen 203 Burkholderia isolates for bimA and the CA domain by PCR using Platinum Taq DNA polymerase (Invitrogen, Paisley, United Kingdom) and boiled lysates of single colonies as template. The strain collection was described previously (8), supplemented by an additional 52 B. pseudomallei, 23 B. thailandensis, and 19 B. mallei isolates and 10 other isolates representing 5 other Burkholderia species. A bimA amplicon was detected for all isolates of species B. pseudomallei, B. thailandensis, B. mallei, and Burkholderia oklahomensis. Consistent with analysis of sequenced genomes (8), the CA domain was restricted to, and always found in, B. thailandensis isolates. Such a test may prove useful to rapidly discriminate between avirulent B. thailandensis and the closely related biothreat agents B. pseudomallei and B. mallei.To investigate the function of the CA domain, we deleted the region corresponding to residues 96 to 130 of the strain E264 BimA by PCR-ligation-PCR (1) using the cloned bimA gene of B. thailandensis strain E30 as a template (10). Pfu proofreading DNA polymerase was used to separately amplify the region 5′ of the CA domain with primers B_th-comp forward (5′-CATGAATTCCCATGCGTGCAACAGTTGCT-3′) and 5′-CGAGCCGCCCGCGCCTCGCGTGTT-3′ and the region 3′ of the CA domain with B_th-comp reverse (5′-CTTCTCGAGTCACCATTGCCAGCTCATGCCTACGC-3′) and 5′-TCCCCTCCGCCGACGCCGATCGCAA-3′ from pME6032- bimA_th (10). The PCR amplicons were ligated, and the desired recombinant was amplified by a further round of PCR with primers B_th-comp forward and B_th-comp reverse. The product was subcloned under the Ptac promoter in pME6032 (4) via EcoRI and XhoI sites incorporated in the primers (underlined), yielding pME6032-bimA_th-ΔCA. Faithful amplification and deletion of the CA domain were confirmed by nucleotide sequencing (data not shown).To evaluate the role of the CA domain in actin-based motility, pME6032-bimA_th and the ΔCA variant were introduced into a B. pseudomallei strain 10276 bimA::pDM4 mutant (11) by electroporation with selection for tetracycline resistance. Strains were amplified in Luria-Bertani (LB) medium, and J774.2 murine macrophage-like cells were infected at a multiplicity of 10 in RPMI medium containing 10% (vol/vol) fetal calf serum at 37°C in a 5% CO₂ atmosphere. After 30 min cells were washed and overlaid with medium containing 250 μg/ml kanamycin to kill extracellular bacteria. During bacterial culture and cell infection, expression of BimA proteins was induced with 0.25 mM isopropyl-β-d-thiogalactoside (IPTG). Eight hours postinfection cells were washed, fixed, permeabilized, and stained for bacteria, polymerized F-actin, and nuclei essentially as described previously (10). Images were captured using a Leica confocal laser scanning microscope with LAS AF v.2.0 software. B. pseudomallei 10276 and 10276 bimA::pDM4 were included as positive and negative controls, respectively (Fig. ​(Fig.11 A and B). As expected B. thailandensis E30 BimA restored the ability of the 10276 bimA::pDM4 mutant to form actin tails (10) (Fig. ​(Fig.1C).1C). Deletion of the CA domain of B. thailandensis BimA abolished this activity (Fig. ​(Fig.1D),1D), indicating that it is required for actin-based motility.Open in a separate windowFIG. 1.Representative confocal laser scanning micrographs of J774.2 cells infected with B. pseudomallei strain 10276 (A), an isogenic bimA::pDM4 mutant (B), or 10276 bimA::pDM4 trans complemented with pME6032-bimA_th (C) or pME6032-bimA_th-ΔCA (D) and induced to express the proteins under IPTG induction. Bacteria (red) were stained using mouse monoclonal anti-B. pseudomallei lipopolysaccharide antibody (Camlab, Cambridge, United Kingdom) detected with anti-mouse Ig-Alexa Fluor⁵⁶⁸ (Molecular Probes, Leiden, Netherlands). F-actin (green) was stained with Alexa Fluor⁴⁸⁸-conjugated phalloidin. DNA (blue) was stained with 4′,6-diamidino-2-phenylindole. Bars, 5 μm. Bacteria forming actin tails are marked with arrows.Monoclonal antibodies raised against BimA_ps (11) failed to recognize BimA_th on the bacterial pole (data not shown); therefore, we were unable to conclude that loss of actin-based motility upon deletion of the CA domain may be a consequence of failed secretion or polar localization. To investigate the role of the CA domain in actin binding and polymerization, the ΔCA variant of B. thailandensis BimA was PCR amplified from pME6032-bimA_th-ΔCA with primers 5′-GGGCCCGGATCCGCCGCTGACGAGACG-3′ and 5′-GGGCCCGAATTCTCACGCTCGCGCGTCG-3′. The product was first cloned by a topoisomerase-mediated process into pCR2.1-TOPO (Invitrogen, Paisley, United Kingdom) and then subcloned as a BamHI and EcoRI fragment into similarly digested pGEX-2T-1 (Amersham Biosciences, Buckinghamshire, United Kingdom), creating a fusion to the carboxyl terminus of glutathione S-transferase (GST). The subcloned region corresponds to amino acids 47 to 386 of BimA_th and was confirmed by sequencing to be identical to the pGEX-BimA_th plasmid described previously (10), except for the deleted CA region. The region encoding B. pseudomallei BimA residues 54 to 455 (lacking the signal peptide and membrane anchor) was amplified from pME6032-bimA_ps (10) with primers 5′-GCGCGCGGATCCATGAATCCCCCCGAACCGCCGGGC-3′ and 5′-GCGCGCGAATTCTTAGCGCGCGGTGTCGGTG-3′ and fused to GST in pGEX-4T-1 as described above. Plasmids encoding GST or GST fusions to BimA_ps, BimA_th, or BimA_th-ΔCA domain were separately introduced into Escherichia coli K-12 Rosetta2(DE3)pLysS cells expressing rare aminoacyl tRNAs (Merck Biosciences, Nottingham, United Kingdom). LB cultures were induced to express the proteins at late logarithmic phase for 3 h at ambient temperature, which proved optimal for recovery of intact proteins using glutathione Sepharose 4B beads (Fig. ​(Fig.2A2A).Open in a separate windowFIG. 2.SDS-PAGE analysis of fusions of BimA_ps (residues 54 to 455), BimA_th (residues 47 to 386), or BimA_th lacking the CA domain to the carboxyl terminus of GST (A). Sepharose beads coated with these proteins, or GST alone, were examined for their ability to sequester actin (B), p34-Arc/ARPC2 (C), or Arp3 (D) from murine splenic extracts by Western blotting of bead-associated proteins using specific antibodies. Coated beads were also examined for their ability to bind highly purified rhodamine-labeled actin in PBS (red) by confocal laser scanning microscopy (E). Mw, molecular weights in thousands.Beads coated with GST, GST-BimA_ps, GST-BimA_th, or GST-BimA_th-ΔCA were incubated with murine splenic cell lysate as described previously (10). After 15 min of incubation at ambient temperature, beads were washed with ice-cold Tris-buffered saline and analyzed by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis and Western blotting with antiactin antibody (10). As reported previously, beads coated with GST-BimA_ps or GST-BimA_th, but not GST, sequestered actin from the splenic extract (Fig. ​(Fig.2B)2B) (10). Deletion of the CA domain did not impair actin binding, consistent with the fact that a predicted WH2 domain remains intact (10). It may be inferred that such binding is direct, since coated beads also bound highly purified rhodamine-labeled actin in phosphate-buffered saline (PBS) as assessed by confocal microscopy of beads treated as described previously (Fig. ​(Fig.2E)2E) (10). The ability of coated beads to sequester the Arp2/3 complex from murine splenic extracts was examined using rabbit anti-p34-Arc/ARPC2 (Millipore, Watford, United Kingdom) and goat anti-Arp3 (Autogen Bioclear, Wiltshire, United Kingdom) detected with species-specific Ig-horseradish peroxidase conjugates by a chemoluminescence method (Amersham Biosciences, Buckinghamshire, United Kingdom). GST-BimA_th, but not GST-BimA_ps, sequestered p34-Arc/ARPC2 (Fig. ​(Fig.2C)2C) and Arp3 (Fig. ​(Fig.2D).2D). Deletion of the CA domain abolished the ability to sequester Arp2/3 components (Fig. 2C and D), indicating that it plays a role in recruitment of the complex.The requirement for Arp2/3 and the CA domain in actin polymerization by B. thailandensis BimA was evaluated in vitro using pyrene-labeled actin. Polymerization of such monomers leads to an emission of fluorescence that can be sensitively recorded over time. Lyophilized pyrene-actin, Arp2/3, and the verprolin-like central and acidic (VCA) domain of WASP were obtained from Cytoskeleton (Universal Biologicals, Cambridge, United Kingdom) and prepared per the manufacturer''s instructions. Assay conditions were essentially as described previously (12). Briefly, 90-μl reaction mixtures were assembled in black opaque 96-well plates containing 100 nM GST or GST fusion protein, 2 μM pyrene-actin, and 30 nM Arp2/3 as required. Reactions were initiated by the addition of 10 μl of 10× polymerization buffer (100 mM Tris, pH 7.5, 500 mM KCl, 20 mM MgCl₂, 10 mM ATP), and the emission of fluorescence at 407 nm, after excitation at 365 nm, was followed every 30 s for 90 min using a Tecan Infinite M200 fluorescent plate reader with i-control software. The GST, GST-BimA_ps, GST-BimA_th, and GST-BimA_th-ΔCA proteins were prepared as described above but eluted from beads with 10 mM reduced glutathione. Triplicate determinations were performed with two independently purified sets of protein. Data from a representative assay are shown in Fig. ​Fig.3.3. Rates of polymerization were calculated as the rise in fluorescence units per second during the linear phase of polymerization, and the means of six values per protein ± standards of the means are recorded in Table ​Table1.1. Results were analyzed by pairwise Student t tests using R software (version 2.11), and P values of ≤0.05 were taken as significant.Open in a separate windowFIG. 3.Polymerization of pyrene-labeled actin monomers by GST (green lines), GST-BimA_ps (blue lines), GST-BimA_th (red lines), and GST-BimA_th-ΔCA (yellow lines) in the absence (solid lines) or presence (dotted lines) of the Arp2/3 complex. The Arp2/3 complex was confirmed to be active in the presence of the purified VCA domain of WASP (black line). The graph shows the increase in fluorescence units (at 407 nm) over time during a single representative experiment. Triplicate determinations were performed with two sets of independently purified proteins, and the mean rates of fluorescence are recorded in Table ​Table11.

TABLE 1.

Rates of polymerization of pyrene-labeled actin monomers by GST and GST-BimA fusion proteins in the absence or presence of the Arp2/3 complex

Streptomyces Development in Colonies and Soils

Streptomyces development was analyzed under conditions resembling those in soil. The mycelial growth rate was much lower than that in standard laboratory cultures, and the life span of the previously named first compartmentalized mycelium was remarkably increased.Streptomycetes are gram-positive, mycelium-forming, soil bacteria that play an important role in mineralization processes in nature and are abundant producers of secondary metabolites. Since the discovery of the ability of these microorganisms to produce clinically useful antibiotics (2, 15), they have received tremendous scientific attention (12). Furthermore, its remarkably complex developmental features make Streptomyces an interesting subject to study. Our research group has extended our knowledge about the developmental cycle of streptomycetes, describing new aspects, such as the existence of young, fully compartmentalized mycelia (5-7). Laboratory culture conditions (dense inocula, rich culture media, and relatively elevated temperatures [28 to 30°C]) result in high growth rates and an orderly-death process affecting these mycelia (first death round), which is observed at early time points (5, 7).In this work, we analyzed Streptomyces development under conditions resembling those found in nature. Single colonies and soil cultures of Streptomyces antibioticus ATCC 11891 and Streptomyces coelicolor M145 were used for this analysis. For single-colony studies, suitable dilutions of spores of these species were prepared before inoculation of plates containing GYM medium (glucose, yeast extract, malt extract) (11) or GAE medium (glucose, asparagine, yeast extract) (10). Approximately 20 colonies per plate were obtained. Soil cultures were grown in petri dishes with autoclaved oak forest soil (11.5 g per plate). Plates were inoculated directly with 5 ml of a spore suspension (1.5 × 10⁷ viable spores ml⁻¹; two independent cultures for each species). Coverslips were inserted into the soil at an angle, and the plates were incubated at 30°C. To maintain a humid environment and facilitate spore germination, the cultures were irrigated with 3 ml of sterile liquid GAE medium each week.The development of S. coelicolor M145 single colonies growing on GYM medium is shown in Fig. ​Fig.1.1. Samples were collected and examined by confocal microscopy after different incubation times, as previously described (5, 6). After spore germination, a viable mycelium develops, forming clumps which progressively extend along the horizontal (Fig. 1a and b) and vertical (Fig. 1c and d) axes of a plate. This mycelium is fully compartmentalized and corresponds to the first compartmentalized hyphae previously described for confluent surface cultures (Fig. 1e, f, and j) (see below) (5); 36 h later, death occurs, affecting the compartmentalized hyphae (Fig. 1e and f) in the center of the colony (Fig. ​(Fig.1g)1g) and in the mycelial layers below the mycelial surface (Fig. 1d and k). This death causes the characteristic appearance of the variegated first mycelium, in which alternating live and dead segments are observed (Fig. 1f and j) (5). The live segments show a decrease in fluorescence, like the decrease in fluorescence that occurs in solid confluent cultures (Fig. ​(Fig.11 h and i) (5, 9). As the cycle proceeds, the intensity of the fluorescence in these segments returns, and the segments begin to enlarge asynchronously to form a new, multinucleated mycelium, consisting of islands or sectors on the colony surfaces (Fig. 1m to o). Finally, death of the deeper layers of the colony (Fig. ​(Fig.1q)1q) and sporulation (Fig. ​(Fig.1r)1r) take place. Interestingly, some of the spores formed germinate (Fig. ​(Fig.1s),1s), giving rise to a new round of mycelial growth, cell death, and sporulation. This process is repeated several times, and typical, morphologically heterogeneous Streptomyces colonies grow (not shown). The same process was observed for S. antibioticus ATCC 11891, with minor differences mainly in the developmental time (not shown).Open in a separate windowFIG. 1.Confocal laser scanning fluorescence microscopy analysis of the development-related cell death of S. coelicolor M145 in surface cultures containing single colonies. Developmental culture times (in hours) are indicated. The images in panels l and n were obtained in differential interference contrast mode and correspond to the same fields as in panels k and m, respectively. The others are culture sections stained with SYTO 9 and propidium iodide. Panels c, d, k, l, p, and q are cross sections; the other images are longitudinal sections (see the methods). Panels h and i are images of the same field taken with different laser intensities, showing low-fluorescence viable hyphae in the center of the colonies that develop into a multinucleated mycelium. The arrows in panels e and s indicate septa (e) and germinated spores (s). See the text for details.Figure ​Figure22 shows the different types of mycelia present in S. coelicolor cultures under the conditions described above, depending on the compartmentalization status. Hyphae were treated with different fluorescent stains (SYTO 9 plus propidium iodide for nucleic acids, CellMask plus FM4-64 for cell membranes, and wheat germ agglutinin [WGA] for cell walls). Samples were processed as previously described (5). The young initial mycelia are fully compartmentalized and have membranous septa (Fig. 2b to c) with little associated cell wall material that is barely visible with WGA (Fig. ​(Fig.2d).2d). In contrast, the second mycelium is a multinucleated structure with fewer membrane-cell wall septa (Fig. 2e to h). At the end of the developmental cycle, multinucleated hyphae begin to undergo the segmentation which precedes the formation of spore chains (Fig. 2i to m). Similar results were obtained for S. antibioticus (not shown), but there were some differences in the numbers of spores formed. Samples of young and late mycelia were freeze-substituted using the methodology described by Porta and Lopez-Iglesias (13) and were examined with a transmission electron microscope (Fig. 2n and o). The septal structure of the first mycelium (Fig. ​(Fig.2n)2n) lacks the complexity of the septal structure in the second mycelium, in which a membrane with a thick cell wall is clearly visible (Fig. ​(Fig.2o).2o). These data coincide with those previously described for solid confluent cultures (4).Open in a separate windowFIG. 2.Analysis of S. coelicolor hyphal compartmentalization with several fluorescent indicators (single colonies). Developmental culture times (in hours) are indicated. (a, e, and i) Mycelium stained with SYTO 9 and propidium iodide (viability). (b, f, and j) Hyphae stained with Cell Mask (a membrane stain). (c, g, and l) Hyphae stained with FM 4-64 (a membrane stain). (d, h, and m) Hyphae stained with WGA (cell wall stain). Septa in all the images in panels a to j, l, and m are indicated by arrows. (k) Image of the same field as panel j obtained in differential interference contrast mode. (n and o) Transmission electron micrographs of S. coelicolor hyphae at different developmental phases. The first-mycelium septa (n) are comprised of two membranes separated by a thin cell wall; in contrast, second-mycelium septa have thick cell walls (o). See the text for details. IP, propidium iodide.The main features of S. coelicolor growing in soils are shown in Fig. ​Fig.3.3. Under these conditions, spore germination is a very slow, nonsynchronous process that commences at about 7 days (Fig. 3c and d) and lasts for at least 21 days (Fig. 3i to l), peaking at around 14 days (Fig. 3e to h). Mycelium does not clump to form dense pellets, as it does in colonies; instead, it remains in the first-compartmentalized-mycelium phase during the time analyzed. Like the membrane septa in single colonies, the membrane septa of the hyphae are stained with FM4-64 (Fig. 3j and k), although only some of them are associated with thick cell walls (WGA staining) (Fig. ​(Fig.3l).3l). Similar results were obtained for S. antibioticus cultures (not shown).Open in a separate windowFIG. 3.Confocal laser scanning fluorescence microscopy analysis of the development-related cell death and hyphal compartmentalization of S. coelicolor M145 growing in soil. Developmental culture times (in days) are indicated. The images in panels b, f, and h were obtained in differential interference contrast mode and correspond to the same fields as the images in panels a, e, and g, respectively. The dark zone in panel h corresponds to a particle of soil containing hyphae. (a, c, d, e, g, i, j, and k) Hyphae stained with SYTO 9, propidium iodide (viability stain), and FM4-64 (membrane stain) simultaneously. (i) SYTO 9 and propidium iodide staining. (j) FM4-64 staining. The image in panel k is an overlay of the images in panels i and j and illustrates that first-mycelium membranous septa are not always apparent when they are stained with nucleic acid stains (SYTO 9 and propidium iodide). (l) Hyphae stained with WGA (cell wall stain), showing the few septa with thick cell walls present in the cells. Septa are indicated by arrows. IP, propidium iodide.In previous work (8), we have shown that the mycelium currently called the substrate mycelium corresponds to the early second multinucleated mycelium, according to our nomenclature, which still lacks the hydrophobic layers characteristic of the aerial mycelium. The aerial mycelium therefore corresponds to the late second mycelium which has acquired hydrophobic covers. This multinucleated mycelium as a whole should be considered the reproductive structure, since it is destined to sporulate (Fig. ​(Fig.4)4) (8). The time course of lysine 6-aminotransferase activity during cephamycin C biosynthesis has been analyzed by other workers using isolated colonies of Streptomyces clavuligerus and confocal microscopy with green fluorescent protein as a reporter (4). A complex medium and a temperature of 29°C were used, conditions which can be considered similar to the conditions used in our work. Interestingly, expression did not occur during the development of the early mycelium and was observed in the mycelium only after 80 h of growth. This suggests that the second mycelium is the antibiotic-producing mycelium, a hypothesis previously confirmed using submerged-growth cultures of S. coelicolor (9).Open in a separate windowFIG. 4.Cell cycle features of Streptomyces growing under natural conditions. Mycelial structures (MI, first mycelium; MII, second mycelium) and cell death are indicated. The postulated vegetative and reproductive phases are also indicated (see text).The significance of the first compartmentalized mycelium has been obscured by its short life span under typical laboratory culture conditions (5, 6, 8). In previous work (3, 7), we postulated that this structure is the vegetative phase of the bacterium, an hypothesis that has been recently corroborated by proteomic analysis (data not shown). Death in confluent cultures begins shortly after germination (4 h) and continues asynchronously for 15 h. The second multinucleated mycelium emerges after this early programmed cell death and is the predominant structure under these conditions. In contrast, as our results here show, the first mycelium lives for a long time in isolated colonies and soil cultures. As suggested in our previous work (5, 6, 8), if we assume that the compartmentalized mycelium is the Streptomyces vegetative growth phase, then this phase is the predominant phase in individual colonies (where it remains for at least 36 h), soils (21 days), and submerged cultures (around 20 h) (9). The differences in the life span of the vegetative phase could be attributable to the extremely high cell densities attained under ordinary laboratory culture conditions, which provoke massive differentiation and sporulation (5-7, 8).But just exactly what are “natural conditions”? Some authors have developed soil cultures of Streptomyces to study survival (16, 17), genetic transfer (14, 17-19), phage-bacterium interactions (3), and antibiotic production (1). Most of these studies were carried out using amended soils (supplemented with chitin and starch), conditions under which growth and sporulation were observed during the first few days (1, 17). These conditions, in fact, might resemble environments that are particularly rich in organic matter where Streptomyces could conceivably develop. However, natural growth conditions imply discontinuous growth and limited colony development (20, 21). To mimic such conditions, we chose relatively poor but more balanced carbon-nitrogen soil cultures (GAE medium-amended soil) and less dense spore inocula, conditions that allow longer mycelium growth times. Other conditions assayed, such as those obtained by irrigating the soil with water alone, did not result in spore germination and mycelial growth (not shown). We were unable to detect death, the second multinucleated mycelium described above, or sporulation, even after 1 month of incubation at 30°C. It is clear that in nature, cell death and sporulation must take place at the end of the long vegetative phase (1, 17) when the imbalance of nutrients results in bacterial differentiation.In summary, the developmental kinetics of Streptomyces under conditions resembling conditions in nature differs substantially from the developmental kinetics observed in ordinary laboratory cultures, a fact that should be born in mind when the significance of development-associated phenomena is analyzed.