Edited by Alexander Namiot Glazer, University of California, Berkeley, CA, and approved October 4, 2012 (received for review July 10, 2012)
The marine cyanobacterium Synechococcus is the second most abundant phytoplanktonic organism in the world's oceans. The ubiquity of this genus is in large part due to its use of a diverse set of photosynthetic light-harvesting pigments called phycobiliproteins, which allow it to efficiently exploit a wide range of light colors. Here we uncover a pivotal molecular mechanism underpinning a widespread response among marine Synechococcus cells known as “type IV chromatic acclimation” (CA4). During this process, the pigmentation of the two main phycobiliproteins of this organism, phycoerythrins I and II, is reversibly modified to match changes in the ambient light color so as to maximize photon capture for photosynthesis. CA4 involves the replacement of three molecules of the green light-absorbing chromophore phycoerythrobilin with an equivalent number of the blue light-absorbing chromophore phycourobilin when cells are shifted from green to blue light, and the reverse after a shift from blue to green light. We have identified and characterized MpeZ, an enzyme critical for CA4 in marine Synechococcus. MpeZ attaches phycoerythrobilin to cysteine-83 of the α-subunit of phycoerythrin II and isomerizes it to phycourobilin. mpeZ RNA is six times more abundant in blue light, suggesting that its proper regulation is critical for CA4. Furthermore, mpeZ mutants fail to normally acclimate in blue light. These findings provide insights into the molecular mechanisms controlling an ecologically important photosynthetic process and identify a unique class of phycoerythrin lyase/isomerases, which will further expand the already widespread use of phycoerythrin in biotechnology and cell biology applications.
Cyanobacteria within the Synechococcus spp. are found in marine environments from the equator to the polar circles, and members of this genus contribute significantly to the total phytoplankton biomass and productivity of the oceans (1⇓–3). Their ubiquity is due in part to their wide pigment diversity (4), which arises mainly from differences in the composition of their light-harvesting antennae or phycobilisomes (PBS). PBS consist of a core and six or eight rods radiating from the core that contain the phycobiliprotein phycocyanin (PC) and one or two types of phycoerythrins (PEs), PEI and PEII (5). All phycobiliproteins are α/β heterodimers that are assembled into hexamers by linkers. PEs may bind two different types of chromophores, green light (GL)-absorbing phycoerythrobilin (PEB) and blue light (BL)-absorbing phycourobilin (PUB). These chromophores are ligated to PE by PEB lyases (6, 7) or PEB–lyase–isomerases, which both attach the chromophore and isomerize it to PUB (8). No PE-specific PEB–lyase–isomerase has been described to date. PUB predominates in Synechococcus found in nutrient-poor open ocean waters, vast areas where blue light penetrates deeper than any other color (9).
Marine Synechococcus are divided into three major pigment types, with type 1 PBS rods containing only PC, type 2 containing PC and PEI, and type 3 containing PC, PEI, and PEII (4). Type 3 can be further split into four subtypes (3 A–D) on the basis of the ratio of the PUB and PEB chromophores bound to PEs. For all pigment types and subtypes, the size and number of PBS may vary with irradiance (10), but only pigment subtype 3d is able to vary its pigmentation in response to changes in ambient light color through a process called type IV chromatic acclimation (hereafter called CA4) (4, 11, 12).
Other CA types, such as CA2 and CA3, have been studied in freshwater cyanobacteria (13, 14). Like CA4, these processes are photoreversible, but they involve very different protein and bilin changes. For example, CA3 in Fremyella diplosiphon, which occurs when cells are shifted between red light and GL, involves switching between PC and PE and their corresponding chromophores in the PBS rods (14, 15). In contrast, CA4 occurs when marine Synechococcus cells are shifted between GL and BL, and during this process there is no change in the phycobiliprotein composition of the PBS rods (11). Instead, CA4 involves changes in the chromophores associated with two different cysteines within the α-PEII subunit (12). In GL, PEB is bound to these sites, whereas in BL, PUB is bound. The mechanism(s) controlling these changes is unknown. Here, we use biochemical and molecular genetic approaches to describe MpeZ, an enzyme involved in the ligation and isomerization of a PEII-linked chromophore, and demonstrate its pivotal role in the poorly understood but globally important process of CA4. The discovery of this class of enzymes has the potential to further expand the current broad use of phycoerythrin in biotechnology and cell biology applications.
Comparative genomics analysis showed that all sequenced marine Synechococcus strains that undergo CA4 possess a specific gene, called mpeZ (4). In Synechococcus sp. RS9916 (hereafter 9916), mpeZ is downstream of a gene of unknown function and overlaps a gene putatively encoding a truncated form of the photosystem II core protein PsbA (Fig. 1A). RNA blot analysis demonstrated that mpeZ transcript accumulation was CA4 regulated, being six times more abundant in cells grown in BL than in GL (Fig. 1B). Primary and secondary structure analyses of the encoded protein, MpeZ, revealed a large domain belonging to the PBS lyase HEAT-like repeat family (Fig. S1A), suggesting that this protein could be a bilin lyase–isomerase involved in mediating the shift between PEB- and PUB-enriched PBS rods during CA4.
mpeZ genome localization and expression. (A) mpeZ genome context: “psbA” denotes a fragment of psbA; “unk” denotes an unknown gene. (B) (Left) Representative RNA blot of transcripts from cells acclimated to GL or BL using mpeZ and 16S rRNA (ribo) probes. Numbers are lengths in kbp. (Right) Relative mean transcript levels of mpeZ in 9916 cells grown in GL or BL. Values expressed as a percentage of transcripts from BL-grown cells after ribo normalization. Data are from three independent RNA blot analyses; error bars show SE.
MpeZ was tested for lyase–isomerase activity by producing it in Escherichia coli cells expressing ho1 and pebS, which encode the proteins needed for the synthesis of PEB (16), along with six-histidine–tagged (HT) versions of wild-type (WT) and mutant forms of either MpeA (PEII α-subunit) or CpeA (PEI α-subunit). Spectral analyses of purified wild-type HT-MpeA from MpeZ-containing E. coli cells revealed absorbance and fluorescence emission maxima at 495 and 510 nm, respectively (Fig. 2A), which matched the spectral properties of PUB attached to protein (8, 18, 19). HT-MpeA was detectable on protein gels and contained an attached bilin (Fig. 2C). As expected, no absorbance or fluorescence was detectable from HT-MpeA expressed in cells lacking MpeZ (Fig. 2 A and C) because nonchromophorylated recombinant PE subunits are generally insoluble in E. coli (6, 20). There are three canonical chromophore-binding cysteines at positions 75, 83, and 140 within MpeA. These cysteines were mutated to alanine in various combinations and expressed in E. coli cells producing MpeZ and PEB. The spectral properties of purified HT-MpeA-C75A,C140A matched those of HT-MpeA, whereas HT-MpeA-C83A showed no absorbance or fluorescence, indicating that the latter form was nonchromophorylated (Fig. 2 B and C) (6, 20). When MpeZ was coexpressed with HT-CpeA in the PEB-producing E. coli strain, the HT-CpeA protein showed no absorbance or fluorescence, indicating that, in this E. coli system, MpeZ does not chromophorylate CpeA (Fig. 2B). From these data we conclude that, when expressed in E. coli, MpeZ functions as a phycobilin lyase–isomerase, attaching PEB at Cys-83 of MpeA and isomerizing it to PUB (Fig. 2D).
Analyses of recombinant HT-MpeA and HT-CpeA produced in presence or absence of MpeZ. (A) Absorbance (solid lines) and fluorescence emission (dashed lines) spectra for (i) HT-MpeA purified from cells containing MpeA, PEB (17), and MpeZ (orange); (ii) HT-MpeA purified from cells containing MpeA and PEB only (no MpeZ; black); and (iii) HT-CpeA purified from cells containing CpeA and PEB and MpeZ (aqua). (B) Absorbance (solid lines) and fluorescence emission (dashed lines) spectra for MpeA with cysteinyl-binding sites replaced by alanines as (i) HT-MpeA-C75A,C140A purified from cells containing MpeA-C75A,C140A and PEB and MpeZ (orange) and (ii) HT-MpeA-C83 purified from cells containing MpeA-C83A, PEB, and MpeZ (black). (C) (Upper) Coomassie-stained SDS polyacrylamide gel with HT-MpeA purified from cells containing MpeA, PEB with (lane 1) or without (lane 2) MpeZ, from cells with MpeA-C83A, PEB, and MpeZ (lane 3) or from cells containing MpeA-C75A, C140A, PEB, and MpeZ (lane 4). The molecular mass of the standard loaded in lane “S” is indicated on the right. (Lower) Zinc-enhanced fluorescence of bilins within the above gel. (D) The chemical reaction catalyzed by MpeZ is the attachment of PEB (red) to a cysteine residue of a MpeA apoprotein (black) and its isomerization to PUB (blue).
To further analyze the role of MpeZ, we created an mpeZ insertion mutant in 9916 (Fig. S2) and tested it for its ability to carry out CA4 by recording the Ex495 nm:Ex550 nm fluorescence excitation ratio, with emission set at 580 nm (hereafter Ex495:550), which has been used previously as a proxy for assessing the in vivo PUB:PEB ratio (12). For “control” cultures (cells with normal CA4, carrying the same antibiotic resistance marker as the mpeZ insertion mutant) acclimated to GL and then switched to BL, the Ex495:550 increased from 0.7 to 1.5 over a 6-d period and subsequently remained constant (Fig. 3A). In contrast, this ratio steadily rose from 0.7 to 0.9 for mpeZ mutant cultures over the 11-d experimental period. Complementary responses were obtained for control and mpeZ mutant cultures when BL-acclimated cells were shifted to GL (Fig. 3B). Thus, compared with the control, the loss of MpeZ activity resulted in a 75% decrease in the difference between the Ex495:550 value in BL versus GL, and this was attributable to the lower Ex495:550 value in BL. Control and mpeZ mutant cell growth was measured at three BL irradiances (Fig. 3C). Growth was similar for the two cultures at 15 µmol photons m−2⋅s−1 but was much slower in the mutant at 5 µmol photons m−2⋅s−1. At 1 µmol photons m−2⋅s−1, the control cells grew slowly whereas mpeZ mutant cells showed virtually no growth. Thus, in BL, the disruption of mpeZ affected both the fluorescence characteristics of the PBS and growth, especially at low irradiances.
Effect of mpeZ disruption on spectral properties and growth of 9916 cells. (A) Ex495:550 from control (closed circles) and mpeZ mutant (open circles) cells grown in GL and then shifted to BL at time 0. (B) Ex495:550 from the same cell cultures grown in BL and then shifted to GL at time 0. (C) Growth curves for control (closed symbols) and mpeZ mutant (open symbols) cells grown at different BL irradiances: circles, squares, and triangles correspond to 15, 5, and 1 µmol photons m−2⋅s−1, respectively. (D) Absorption spectra of the MpeA protein purified from WT (Left) or mpeZ mutant cells (Right) grown in BL (blue line) and GL (green line).
To identify the rod proteins that are acted upon by MpeZ in vivo, the PEI and PEII α and β subunits (CpeA, CpeB, MpeA and MpeB) were isolated from 9916 wild type and mpeZ mutant cells grown in BL and GL (Fig. S3 A and B). The identity of proteins in each of the major peaks was confirmed by mass spectrometry (MS). No difference was observed between the HPLC profiles of phycobiliproteins from wild-type and mpeZ mutant cells. Comparison of the 280 nm absorbance chromatograms (Fig. S3 A and B), did not reveal significant differences in the CpeA:MpeA and CpeB:MpeB ratios in WT and mpeZ mutant cells in either light condition. Spectral analysis of isolated MpeA demonstrated that the absorption spectra were the same for wild-type and the mpeZ mutant in GL but differed in BL, where PEB absorbance was detectable in the mutant but not in wild-type cells (Fig. 3D). Similar analyses of isolated CpeA showed that the PUB:PEB absorbance ratios were the same in the wild-type and mpeZ mutant cultures in BL and GL (Fig. S3 C and D). These data demonstrate that MpeZ is involved in the attachment of PUB to MpeA, but not CpeA, in BL-grown wild-type 9916 cells.
The type of bilins attached to the major PBS rod proteins of wild-type and mpeZ mutant cells grown in BL and GL was determined using parallel UV-visible (UV-VIS) spectroscopy and tandem MS of HPLC-purified proteins. The MS peak intensities (extracted ion chromatograms, or EICs) and UV-VIS spectra for MpeA-C83 tryptic peptides are provided in Fig. 4, and those for MpeA-C75 and MpeA-C140 tryptic peptides are provided in Fig. S4. Similar data for peptides containing CpeA-C82 and CpeA-C139 are presented in Figs. S5 and S6. The results for the HPLC purification and spectral analysis of CpeB and MpeB from wild type and the mpeZ mutant are shown in Fig. S7; the UV-VIS spectra and MS peak intensities for tryptic peptides containing the four chromophore-binding cysteines of CpeB, the two chromophore-binding cysteines of MpeB, and the single chromophore-binding cysteine of RpcA, encoding the PC α subunit, are provided in Fig. S8. The data extracted from all of these analyses is summarized in Table 1 and show that CpeA-C139, MpeA-C83, and MpeA-C140 are the three amino acids that undergo changes in bilin composition during CA4. In wild-type cells, each has PEB attached in GL and PUB attached in BL, whereas in the mpeZ mutant MpeA-C83 fails to attach PUB in BL. These data confirm that the role of MpeZ in CA4 is to ligate PUB to MpeA-C83 during growth in BL.
EICs and UV-VIS absorption spectra for tryptic peptides containing C83 of MpeA isolated from WT 9916 and mpeZ mutant cells grown in BL or GL. (A) EIC for the peptide EKC83KR (M+2H)2+ at m/z 625.8 derived from WT cells grown in GL. (Inset) UV-VIS absorption spectrum for the peak at retention time 57.5 min (“1” on the chromatogram) indicates PEB on C83. abs., absorbance. (B) EIC for the peptide EKC83KR (M+2H)2+ at m/z 625.8 derived from BL-grown WT cells. (Inset) UV-VIS absorption spectrum for the peak at 57.4 min (“2” on the chromatogram) indicates PUB on C83. (C) EIC for the peptide C83KR (M+2H)2+ at m/z 496.7 derived from GL-grown mpeZ mutant cells. (Inset) UV-VIS absorption spectrum for the peak at retention time 44.6 min (“3” on the chromatogram) indicates PEB on C83. (D) EIC for the peptide C83K (M+2H)2+ at m/z 418.7 derived from BL-grown mpeZ mutant cells. (Inset) UV-VIS absorption spectrum at 50.0 min (“4” on the chromatogram) indicates PEB on C83.
Chromophores found at different cysteinyl sites for phycobiliproteins examined in WT and mpeZ cultures grown in BL and GL
CA4 is a sophisticated physiological mechanism by which some marine Synechococcus strains can finely tune the absorption properties of their antenna complexes to the ambient light color (4, 11). Here we show that MpeZ, present in all Synechococcus strains sequenced to date that carry out CA4, is a key enzyme in this process. This enzyme is unique among phycobilin lyase–isomerases described so far because it chromophorylates phycoerythrin, specifically PEII (MpeA). The only other such enzymes known are PecE/PecF, which bind phycocyanobilin at C84 of the phycoerythrocyanin α-subunit and isomerize it to a phycoviolobilin (21, 22), and RpcG, which ligates PEB at C84 of a phycocyanin α-subunit and isomerizes it to PUB (8). All three enzymes belong to the E/F clan of phycobilin lyases, characterized by the presence of an α/α-superhelix fold and Armadillo repeat motifs (23⇓–25), although MpeZ is only distantly related to PecE (27–32% identity and 45–47% homology) and the N terminus of RpcG (23–25% identity and 42–43% homology) and matches just a short part of PecF (29% identity and 53–55% homology over 57 residues) (Fig. S1B). It is particularly interesting that the conserved motif of PecF and the PecF-like C terminus of RpcG that is involved in isomerization, NHCQGN (underlined in Fig. S1B), is absent in MpeZ (8, 26).
Our results are consistent with MpeZ’s role in the isomerization of PEB to PUB and its attachment at MpeA-C83 in BL. MS analyses revealed that the mpeZ mutant possesses a PEB at MpeA-C83 in both BL and GL, indicating that PEB lyase activity is retained at this position. Although unlikely, we cannot rule out the possibility that the mpeZ mutant is producing a form of MpeZ that has kept its PEB ligation activity but lost its isomerase activity. It is also possible that another lyase is adding PEB to MpeA-83 in BL when MpeZ is absent. Recently, it was demonstrated that CpeY was involved in binding a PEB to CpeA-C82 from Fremyella diplosiphon (6) and that this reaction was facilitated by the presence of CpeZ. Because orthologs of CpeY/CpeZ are present in 9916, they very likely catalyze binding of PEB to CpeA-C82. We hypothesize that CpeY (+CpeZ) may also add PEB to MpeA-C83 in GL while MpeZ adds PUB in BL, and these may be the lyases/lyase–isomerases that collectively control the CA4-regulated change of chromophorylation at C83 on MpeA.
CA4-mediated changes in chromophorylation at the other two sites, CpeA-C139 and MpeA-C140 (Table 1), are likely mediated by one or two additional lyase/lyase–isomerase pair(s) that have not been identified yet. Alternatively, a separate PUB synthesis pathway could exist that, in concert with the PEB synthesis pathway, increases the PUB:PEB ratio in BL and decreases it in GL. Also, although we cannot exclude the possibility that chromophore attachment occurs autocatalytically, we believe that this hypothesis is less likely because such in vivo chromophore attachment has not been reported for phycobiliproteins.
An unexpected result from this study is that, although MpeZ appears to be responsible for the chromophorylation of only one of the three sites that change chromophores during CA4, there was a 75% decrease in the difference between the Ex495:550 value in BL versus in GL (Fig. 3 A and B). This is a more dramatic decrease than might have been expected for a single chromophore change, but may be due in part to the position of the MpeA-C83 chromophore in the energy transfer flow within a PE hexamer. The structure of R-PE of Polysiphonia urceolata allowed distance measurements between bilins within a PE hexamer and estimates of likely energy transfer pathways (27). PEB at CpeA-C83 played a critical role in transferring energy from the chromophores located on the outside of the PE hexamer (i.e., β50/61-PUB and α140-PEB in P. urceolata) to the terminal PEB acceptor located at β82 (5, 27). In 9916 cells grown in BL, the two external chromophores are β50/61-PUB and α140-PUB. In the mpeZ mutant, PEB at MpeA-C83 instead of the PUB in wild-type cells (Table 1) may alter relaxation constraints within PEII and/or result in different spectral overlaps with the other bilins present within the hexamer, allowing for dissipation of the excited state by mechanisms other than fluorescence. Quantum yield and fluorescence lifetime measurements for PEII from BL-grown 9916 wild-type and mpeZ mutant cells should resolve this issue.
By allowing marine Synechococcus strains to alter their pigment ratios to match the ambient light color environment, CA4 is likely to confer a fitness advantage over those strains that have fixed pigmentation in habitats where the ratio of blue to green light varies frequently (4). Such an advantage appears to be conferred by CA3, which is beneficial in environments where the red- to green-light ratio varies over time periods longer than the CA3 acclimation time (28). Given the remarkable ubiquity and abundance of marine Synechococcus in the world’s oceans, CA4 must be a globally significant light color acclimation process. The discovery of the first lyase–isomerase controlling CA4 confirms previous proposals that such an enzyme(s) is critical for this response (4, 12). Two other forms of chromatic acclimation that have been analyzed, CA2 and CA3, are complex responses that involve changes in the expression of genes encoding phycobiliprotein and bilin biosynthetic enzymes (14). The fact that MpeZ is a PEII PEB lyase–isomerase, together with data showing that the composition of phycobiliproteins in the rods does not change during CA4 (12), demonstrates that CA4 is fundamentally different from other forms of CA and is likely to be regulated through different light sensing and signal transduction mechanisms (14).
The discovery of MpeZ provides a valuable addition to the array of phycobilin lyases available for producing natural or artificial phycobiliproteins for medical and biological research and industry (29, 30). Because PEB-containing PE conjugated to antibodies or other proteins is currently widely used in bioimaging and cell-sorting applications due to its superior fluorescent properties, MpeZ will be a valuable tool for producing PUB-containing PE for in vivo biotechnological applications.
RS9916, isolated from 10 m deep in the Gulf of Aqaba (31), was obtained from the Roscoff Culture Collection (strain no. RCC555) (4). Wild-type or mpeZ mutant Synechococcus RS9916 cells were grown at 22 °C in PCR-S11 (32) with or without 50 µg/mL kanamycin in polycarbonate Nalgene culture flasks in continuous light using Chroma75 T12 fluorescent bulbs (General Electric). Cultures were acclimated for at least 7 d in BL or GL using filters (LE716 Mikkel Blue, LE738 Jas Green; LEE Filters) at 15 µmol photons m−2⋅s−1 unless noted. Photon flux was measured with a Li-Cor LI-250 light meter connected to a LI109SA quantum sensor. Fluorescence excitation spectra were measured using a Biotek Synergy-Mx spectrofluorimeter and used to calculate the Ex495:550.
Plasmids used are listed in Table S1 and primers in Table S2. pASmpeZ was made by PCR amplification of an ∼800-nucleotide internal region of mpeZ using primers mpeZ-internal-for and mpeZ-internal-rev, cutting with BamHI and inserting into similarly cut pMUT100 (33). The cloning junctions and inserted mpeZ fragment were sequenced. One expression vector used was previously described (16). RS9916 mpeA and RS9916 cpeA were amplified using the corresponding primers listed in Table S2. Amplified fragments were cloned in the pCOLA-Duet (Novagen) vector using BamHI-SalI to generate pCOLADuet-RS9916mpeA and into the BamHI and HindIII sites to create pCOLADuet-RS9916cpeA. RS9916 mpeZ was PCR-amplified and cloned into BglII/XhoI-cut pCDF-Duet (Novagen) to create pCDF-RS9916mpeZ. The mpeA and cpeA sequences were inserted into pCOLADuet in frame with the sequence encoding a HT. Single amino acid changes in mpeA were made using fusion PCR amplification and the primers listed in Table S2. All cloning junctions and PCR-amplified regions were sequenced.
pASmpeZ was transformed into E. coli MC1061 (34) containing pRK24 (35) and pRL528 (36). Biparental mating of exponentially growing RS9916 and E. coli cells was conducted as described (33), except that 9916 cells were grown in BL and then kept in darkness for 2 d before mating for a minimum of 72 h at 30 °C. Cells were plated as previously described (33), except that plates were kept at 22 °C at 5 µmol photons m−2⋅s−1 for the first 3 d and then transferred to 15 µmol photons m−2⋅s−1. Individual colonies were picked and tested for mpeZ disruption using PCR amplification, nucleotide sequencing, and DNA blot analysis using a probe for mpeZ.
One hundred milliliters of wild-type 9916 cells at a density of ∼108 cells mL−1 and grown in BL or GL were used for RNA analysis as previously described (37), using 10 µg/lane of RNA and a mpeZ probe radiolabeled as for the DNA blot.
Expression plasmids were cotransformed into E. coli BL21 (DE3) cells and colonies were selected on Luria Bertani (LB) plates with the appropriate antibiotics as described in ref. 6. To produce recombinant proteins, a single colony was inoculated into a 200-mL overnight culture in LB medium with the appropriate antibiotics and shaken at 20 °C at 180 × g for 30–48 h until the optical density reached OD600 nm = 0.6. Production of T7 RNA polymerase was induced by the addition of 0.5 mM isopropyl β-D thiogalactoside. Cells were incubated with shaking at 180 × g at 20 °C for another 48 h before harvest by centrifugation. Cell pellets were immediately processed for protein purification as previously described (38). The entire purification process was carried out in the dark at 4 °C. Following dialysis to remove imidazole, spectroscopic measurements were taken immediately.
Polypeptides were resolved by SDS/PAGE (15%, wt/vol), and polypeptides were visualized by staining with Coomassie Brilliant Blue R-250. Fluorescence from bilins linked to proteins was detected with excitation at 488 nm as described in ref. 6.
Fluorescence emission and absorbance spectra were recorded as described in ref. 6.
PBS were purified as described (39). HPLC was used to separate each phycobiliprotein as described in the legend for Fig. S3. LC/MS/MS analyses were performed on fractions collected from a C4 column and digested with trypsin as described previously (6).
HPLC-separated and trypsin-digested phycobiliprotein samples from WT or mpeZ mutant cells grown in BL or GL were separated by capillary HPLC as described in the legend for Fig. S4. The UV-VIS detector recorded absorption spectra from 250 to 750 nm at 2.5 Hz. Tandem mass spectra were recorded and analyzed as described in the legend for Fig. S4.
We thank David Scanlan, Bianca Brahamsha, and Brian Palenik for providing materials and suggesting methods for growing and transforming 9916. This work was supported by grants to F.P. and L.G. from the Agence Nationale de la Recherche program PELICAN (contract ANR-09-GENM-030) and the European program MicroB3 (Seventh Framework Program contract 287589); an International Projects and Activities Grant from the Office of International Programs at Indiana University (to D.M.K.); and National Science Foundation Grants MCB-1029414 (to D.M.K.) and MCB-0843664 (to W.M.S.). The METACyt Biochemical Analysis Center was supported by a grant from The Lilly Foundation.
Author contributions: A.S., A.B., N.B., F.P., J.A.K., L.A.H., L.G., A.G., W.M.S., and D.M.K. designed research; A.S., A.B., N.B., J.A.K., L.A.H., A.G., W.M.S., and D.M.K. performed research; A.S., N.B., J.A.K., L.A.H., and D.M.K. contributed new reagents/analytic tools; A.S., A.B., N.B., F.P., J.A.K., L.A.H., L.G., W.M.S., and D.M.K. analyzed data; and A.S., A.B., F.P., J.A.K., L.A.H., L.G., W.M.S., and D.M.K. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1211777109/-/DCSupplemental.
