New Results

Combinatorial design of chemical-dependent protein switches for controlling intracellular electron transfer

Bingyan Wu, Joshua T. Atkinson, Dimithree Kahanda, George. N. Bennett, Jonathan J. Silberg
doi: https://doi.org/10.1101/645978

ABSTRACT

One challenge with controlling electron flow in cells is the lack of biomolecules that directly couple the sensing of environmental conditions to electron transfer efficiency. To overcome this protein component limitation, we randomly inserted the ligand binding domain (LBD) from the human estrogen receptor (ER) into a thermostable 2Fe-2S ferredoxin (Fd) from Mastigocladus laminosus and used a bacterial selection to identify Fd-LBD fusion proteins that support electron transfer from a Fd-NADP reductase (FNR) to a Fd-dependent sulfite reductase (SIR). Mapping LBD insertion sites onto structure revealed that Fd tolerates domain insertion adjacent to or within the tetracysteine motif that coordinates the 2Fe-2S metallocluster. With both classes of the fusion proteins, cellular ET was enhanced by the ER antagonist 4-hydroxytamoxifen. In addition, one of Fds arising from ER-LBD insertion within the tetracysteine motif acquires an oxygen-tolerant 2Fe-2S cluster, suggesting that ET is regulated through post-translational ligand binding.

INTRODUCTION

In redox synthetic biology, non-native electron transfer (ET) pathways are constructed to control electron flow within cells and between cells and extracellular materials1–3. Within the cytosol, synthetic control over ET can aid in optimizing the flux through metabolite pathways by altering the cofactor specificity of oxidoreductases4–6, increasing electron transfer (ET) efficiency between pairs of proteins through fusion7, 8, and generating synthetic cells that insulate ET pathways from off-pathway reactions9–11. Regulating ET is also critical for developing the next generation of bioelectronics and biosensors12. Recent studies have shown that diverse bacteria can couple their metabolism with extracellular materials13–15. Photosensitive extracellular materials have been used to create light-driven enzymatic synthesis through direct coupling with oxidoreductases or by precipitating these materials on cells to create shells that harvest light for cell metabolism16–18. Additionally, proteins that mediate extracellular ET have been transplanted into non-native hosts to enable synthetic electrical communication between cells and materials19, 20. While hybrid cells and materials have been reported that leverage synthetic biology for in situ sensing and chemical synthesis, these studies typically control ET at the transcriptional level, limiting the speed at which ET can be regulated for different applications.

One challenge with controlling ET within cells is our limited understanding of the mechanisms by which this process is regulated in nature. A wide range of protein electron carriers (PECs) have been described that control ET21. In the low potential range, ferredoxins (Fds) and flavodoxins (Flds) are frequently used to regulate ET22, 23. Many organisms contain multiple low potential PECs, with some organisms containing >50 Fds and Flds24. Biochemical and proteomic studies have provided evidence that PEC paralogs can differ in their specificity for their partner oxidoreductases25, 26, allowing them to function as ET hubs, directing different proportions of electrons between the coexpressed partner proteins. Cells can gain additional control over ET by differentially regulating the gene expression of PEC paralogs27, 28. Recently, evidence emerged that Fds may at times use post-translational control to regulate ET in cells. Nitrogenase-protecting Fds use oxidation to alter their conformation and binding to nitrogenase as a mechanism to protect their partner protein from oxidative damage29, 30. Additionally, phosphorylated and calcium-bound Fds have been reported, suggesting that fast regulation of ET has arisen during evolution31, 32. However, very few natural proteins have been discovered that use post-translational mechanisms to control ET. This limits our ability to program ET for metabolic engineering and bioelectronic applications.

Recently, a protein electrical switch was engineered through rational design that exhibits chemical-dependent ET3. In this study, Fd-family sequence information was used to identify backbone locations in a thermostable 2Fe-2S Fd from Mastigocladus laminosus (Ml-Fd) that tolerate fission without losing the ability to support ET in cells33. All four of the rationally-designed split Fds (sFds) were unable to support ET transfer unless they were fused to pairs of peptides that stabilized their interaction3. Additionally, the fragments from one variant could be fused to the ends of a ligand-binding domain (LBD) to create a chemical-dependent protein switch3. This topological mutant arose from insertion of the human estrogen receptor (ER) LBD after Gly35 in Ml-Fd, a location that is outside of the tetracysteine motif that coordinates the 2Fe-2S metallocluster33. While this study provided evidence that domain insertion can be used to create electrical protein switches, using an approach similar to that described for creating fluorescent biosensors for estrogenic compounds34, it only examined one of the possible designs for a Fd containing a ER LBD.

To better understand where Fds tolerate insertion of LBDs, we used transposon mutagenesis to randomly insert the ER LBD at different locations within Ml-Fd35–37, and we subjected this library to a cellular selection for Fd ET. We identified three novel topological mutants that support cellular ET. These proteins all presented enhanced ET in cells when exposed to the ER antagonist 4-hydroxytamoxifen (4-HT). Purification of one variant arising from ER-LBD insertion within the Ml-Fd tetracysteine motif revealed that this protein acquires an oxygen-tolerant 2Fe-2S cluster in the absence of 4-HT, suggesting that some of these proteins regulate ET through chemical-induced changes in the holoprotein structure.

Materials and Methods

Materials

DNA oligos were obtained from Integrated DNA Technologies. Phusion DNA polymerase was purchased from Thermo Fisher. Restriction enzymes and T4 DNA ligase were purchased from New England Biolabs. Chloramphenicol, kanamycin, isopropyl β-D-1-thiogalactopyranoside (IPTG), and dithiothreitol (DTT) were purchased from Research Products International. Anhydrotetracycline (aTc) and 4-hydroxytamoxifen (4-HT) were purchased from Sigma-Aldrich.

Transposon Design

To create a transpon for inserting unique BsaI sites at different locations within the Fd gene, we PCR amplified the Entranceposon M1-CamR (F760, Thermo Fisher Scientific) using primers that modify the R1 sequences to encode BsaI restriction sites. The amplicon was cloned into pUC19 using HindIII and XbaI, and the chloramphenicol resistance cassette (chlR) was replaced with a kanamycin resistance cassette (kanR) using Gibson Assembly38. The transposon-propagation plasmid (pUC19-Mu-BsaI-KanR) encodes a transposon (Mu-BsaI-KanR) that is similar in sequence to a previously described transposon48. In our transposon, the 4 base pair single-stranded overhangs generated by BsaI digestion (CTGC and GCTG) are distinct, since they were designed to be compatible with the N-terminal linker sequence in our synthetic gene encoding the ER-LBD, which encoded an Ala-Ala dipeptide at the beginning. In addition, our transposon contains kanR rather than chlR. To generate Mu-BsaI-KanR for MuA reactions, it was excised from transposon-propagation plasmid using HindIII and BglII.

Staging Library

The gene encoding Mastigocladus laminosus (Ml-Fd) was PCR amplified to create an amplicon encoding the Ml-Fd gene flanked by BbsI restriction sites. This DNA was cloned into pFd007, which uses an aTc-inducible promoter for expression3, and the resulting chlR vector (pFd007-staging) was used for the transposition reaction. Transposition was performed in a 20 µL reaction by mixing MuA reaction buffer, 100 ng linear Mu-BsaI-KanR, 500 ng pFd007-staging, and 0.22 μg MuA (Thermo Fisher Scientific, F750). The reaction was incubated at 30 °C for 16 hours and then 75 °C for 10 minutes to heat inactivate MuA. The reaction product was purified using a DNA Clean and Concentrator-5 kit (Zymo Research), and the entire sample was transformed into electrocompetent E. coli XL1-Blue. Cells were allowed to recover in SOC medium at 37 °C for 1 hour prior to plating onto LB-agar plates containing 34 μg/mL chloramphenicol and 50 μg/mL kanamycin. This antibiotic combination selects for the antibiotic resistance cassettes in pFd007-staging (chlR) and Mu-BsaI-KanR. Approximately 25,000 cfu were obtained. Assuming that the cells doubled one time during the SOC recovery, this indicates that our library contains 12,500 unique vectors. To generate the staging library, cells were harvested from plates by adding LB, scraping colonies into the LB, and pooling the cell slurries from each plate. The vector ensemble was purified from the cell slurry using a Miniprep Kit (Qiagen).

Size-Selected Library

The staging library contains vectors with transposons inserted in the vector and in the gene of interest39. To isolate Fd genes containing a transposon, the staging library was digested with BbsI and size selected using a 0.8% agarose gel. Linear DNA corresponding to a Fd gene containing a single transposon (1,700 bp) was excised from the gel and purified using a Gel DNA recovery kit (Zymo Research). The Fd-transposon gene fusions were cloned into pFd007-staging using Golden Gate Assembly40. The product of the reaction was purified using a DNA Clean and Concentrator kit (Zymo Research), the entire product was transformed into electrocompetent E. coli XL1-Blue, cells were allowed to recover in 1 mL of SOC at 37 °C for 1 hour, and cells were plated onto LB agar plates containing 34 μg/mL chloramphenicol and 50 μg/mL kanamycin. The colonies obtained after overnight growth (∼13,000) were harvested from plates, and the resulting library was purified.

Domain-Insertion Library

The gene encoding the ER-LBD (residues 302-552) flanked by linkers (AAAGGGGSGGGGS at the N-terminus, GGGGSGGGGSAAA at the C-terminus) was PCR amplified from a previously described template using primers that add BsaI restrictions sites at the ends of the amplicon3. The size-selected library was digested with BsaI, the vector (∼3,500 bp) was separated from the transposon using using gel electrophoresis, and the vector DNA was purified using a Zymo clean gel DNA recovery kit (Zymo Research). This purified DNA and ER-LBD gene were assembled using Golden Gate Assembly40. The product of this reaction was purified using a DNA Clean and Concentrator kit (Zymo Research) and transformed into E. coli XL1-Blue using electroporation. Following recovery in SOC medium at 37 °C for 1 hour, cells were plated onto LB agar containing 34 μg/mL chloramphenicol. Following a 16 hour incubation at 37°C, ∼1,000 cfu were obtained. These colonies were harvested and pooled. The final domain-insertion library was obtained by purifying plasmids from this cell slurry using a QIAprep Spin Miniprep Kit (Qiagen, Inc.).

Library Selection

E. coli EW11, a strain lacking a functional sulfite reductase gene11, was transformed using electroporation with pSAC01 (100 ng), a plasmid that constitutively expresses Zea mays FNR and SIR3, and the domain-insertion library (100 ng). Cells were recovered in SOC at 37 °C for 1 hour, pelleted by centrifugation, and resuspended in M9sa medium3, which contains sulfate as the only sulfur source. Resuspended cells were spread onto four types of agar plates, including plates containing M9sa, M9c, which is made by adding methionine and cysteine to M9sa, M9sa containing 200 ng/mL aTc, and M9sa containing 200 ng/mL aTc and 50 μM 4-HT. All plates contained 34 μg/mL of chloramphenicol and 100 μg/mL of streptomycin to select for library vectors and pSAC01, respectively. Plates were incubated at 37 °C for 72 hours prior to assessing growth. The topological mutants obtained from the selection are made up of a sFd whose fragments are fused to the ER LBD termini. For this reason, we name each topological mutant using the residue after which the ER LBD was inserted, e.g., sFd-25-ER.

Complementation Analysis

Electroporation was used to transform E. coli EW11 with pSAC01 and a plasmid that expresses either Ml-Fd3, a C42A mutant of Ml-Fd that lacks a cysteine required to coordinate a 2Fe-2S cluster3, or a topological Fd mutant. Single colonies containing both plasmids were obtained by growing cells on LB-agar plates containing 34 μg/mL of chloramphenicol and 100 μg/mL of streptomycin. Individual colonies obtained from overnight incubations were used to inoculate liquid cultures (100 μL) containing M9c with 0.5% DMSO, 34 μg/mL of chloramphenicol, and 100 μg/mL of streptomycin. These cells were grown for 16 hours while shaking in an Infinite m1000 Pro plate reader (Tecan) at 37 °C with shaking at 217 rpm and an amplitude of 1.5 mm in double-orbital mode. An aliquot (2 μL) of each starting culture was diluted 1:50 in fresh M9sa medium (100 µL) in a flat bottom Costar 96-well plate (#3595, Corning). These cultures were then grown in the presence and absence of aTc (200 ng/mL) and 4-HT in an Infinite m1000 Pro plate reader (Tecan) at 37 °C with shaking at 217 rpm and an amplitude of 1.5 mm in double-orbital mode. Cell growth was assessed every 10 minutes by reading optical density (OD) at 600 nm for 48 hours. To minimize evaporation, the 96-well plate was wrapped with parafilm. Control experiments (Ml-Fd and C42A) were performed in wells that were separated from each by at least one row to avoid cross contamination.

Protein Purification

Ml-Fd was purified as previously described21. The gene encoding sFd-55-ER was PCR amplified and cloned into pET28b using Golden Gate Assembly40. E. coli Rosetta(DE3) transformed with this plasmid were grown at 37 °C in 4 L of LB containing 50 μg/mL kanamycin and 100 μM ferrous ammonium sulfate to mid log phase (OD ∼0.7). Protein expression was induced by adding 0.1 mM IPTG. Following induction, cells were grown overnight at 25 °C while shaking at 250 rpm, harvested by centrifugation (4,000 xg) and resuspended in lysis buffer, which contained 50 mM Tris pH 8, 5 mM dithiothreitol (DTT), and cOmplete Mini EDTA-free protease inhibitor tablet (Roche Diagnostics GmbH). Resuspended cells were split into 30 mL fractions, sonicated on ice with a Qsonica sonicator, 4435 Q55 Sonicator Microprobe (1/4“), 70% amplitude, 1s on 1s off pulse for 8 minutes twice. Clarified lysates were generated by centrifugation (20,000 xg) and loaded onto a DE52 anion exchange column (Whatman) that had been equilibrated with TED buffer (50 mM Tris pH 8, 1 mM EDTA, and 5 mM DTT). The column was washed with TED, and TED containing 50, 100, 150, 200, 250, 300 and 500 mM NaCl was used to elute protein. The brown eluent, which eluted at 250 mM NaCl, showed a band of expected size (41 kDa) on an SDS-PAGE. This fraction was diluted five times with TED buffer and loaded onto HiTrap Q XL column using an AKTA Start FPLC system (GE Healthcare). The column was washed with TED buffer, and a linear NaCl gradient in TED was used to elute the protein. Those fractions appearing brown were analyzed for purity using SDS-PAGE. All fractions contained a major band having the expected molecular weight (41 kDa). Those fractions appearing homogeneous were pooled and concentrated using Amicon Ultra 10k MWCO spin column. This protein was further purified using a size exclusion column (HiLoad16/600 Superdex 200pg, GE Healthcare) equilibrated with TED containing 100 mM NaCl. Protein appearing homogeneous by SDS-PAGE was pooled and used for all subsequent analysis.

Whole Cell Fluorescence

The RFP gene was PCR amplified and cloned using Golden Gate Assembly into aTc-inducible vectors that express Ml-Fd (pFd007), sFd-25-ER (pFd-ER-25), sFd-30-ER (pFd-ER-30), sFd-35-ER (pFd-ER-35), sFd-54-ER (pFd-ER-54) and sFd-55-ER (pFd-ER-55). This yielded vectors that express Fds with RFP fused to their C-terminus through 12 amino acid linker (GGSGGSGGSGGS). Each vector was transformed into E. coli EW11 with pSAC01, cells were grown on LB-agar plates containing 34 μg/mL of chloramphenicol and 100 μg/mL of streptomycin, and individual colonies were used to inoculate liquid cultures (100 μL) containing M9c, 0.5% DMSO, 34 μg/mL of chloramphenicol, and 100 μg/mL of streptomycin. These cells were grown for 16 hours in a Spark plate reader (Tecan) at 37 °C with shaking at 90 rpm at an amplitude of 3 mm in double-orbital mode. An aliquot (2 μL) of each starting culture was diluted 1:50 in fresh M9c medium (100 µL) in a Nunc™ Edge 2.0 96-Well Plates (Thermo Fisher). These cultures were grown in the presence and absence of aTc (200 ng/mL) and 4-HT in a Spark plate reader (Tecan) at 37 °C using a similar procedure. Cell growth was assessed every 5 minutes by reading optical density (OD) at 600 nm for 24 hours as well as whole cell red fluorescence (λex = 560 nm, λem = 650 nm). The ratio of emission to OD was calculated for stationary phase cultures. To minimize evaporation, 96-well plates were wrapped with parafilm, and grown in a Spark humidity cassette.

Spectroscopic Characterization

Purified sFd-55-ER was dialyzed into TED buffer containing 100 mM NaCl. Protein concentration was determined spectrophotometrically using the calculated extinction coefficient (ℇ280nm = 32,890 M-1cm-1). A Varian Cary 50 spectrophotometer and a JASCO J-815 spectropolarimeter were used to obtain absorbance and circular dichroism spectra, respectively. Samples contained 40 µM sFd-55-ER in a cuvette having a 1 cm path. For comparison, absorbance and circular dichroism spectrums were obtained for 40 μM Ml-Fd (ℇ280nm = 9,190 M-1cm-1).

Midpoint potentials

Electrochemical measurements were carried out anaerobically using a DropSens Stat 8000P Potentiostat. Glass vial with a three-electrode system was continuously purged with Ar until the time of the electrochemical measurement, which used a 1 M, KCl Ag/AgCl electrode as the reference electrode and a platinum wire as the counter electrode. Edge-plane graphite electrode (EPG) was used as the working electrode for thin film electrochemistry and potentials were reported relative to the standard Hydrogen electrode (SHE). Electrodes were cleaned and placed in the glass cell containing 23.5°C mixed buffer solution (5 mM sodium acetate/MES/MOPS/TAPS/CHES/CAPS) pH 8.0 containing 100 mM NaCl. Aliquot of 431 μM Ml-Fd (7 µL) or 302 μM sFd-ER-55 (10 µL) were applied directly on the PGE electrode surface with or without 1 μL of 50 mM 4-HT, which was saturated in TED buffer. Thin film was incubated for 5 minutes, while blowing the surface with Ar gas. Square wave voltammograms were collected at 23.5°C at 10 Hz, immediately following placement of the PGE electrode in the glass cell containing buffer solution. Electrochemical signals were analyzed using QSOAS open-source software. Two experiments were performed on each protein with or without 4-HT, under identical conditions.

Statistics

The results from growth complementation and whole cell fluorescence measurements represent the mean and standard deviation of biologically independent samples (n = 3). All reported P values were obtained using two-tailed, independent t-tests.

RESULTS AND DISCUSSION

Library construction

The pioneering studies by Frances Arnold and coworkers revealed that artificial proteins can be created in the lab by mimicking evolution41 In this directed evolution approach42, proteins with novel functions are discovered by creating combinatorial libraries of mutants and mining those libraries for biomolecules with the desired functional properties using a selection or a screen. In cases where large fitness jumps are desired, libraries encoding topological mutants of proteins are increasingly used, such as random domain insertion libraries for creating protein switches43, 44. Several library approaches have been described for generating domain insertion libraries, including nuclease-based and transposase-mediated methods35, 36, 45, 46. Among these approaches, transposase mutagenesis is appealing to use for creating protein switches because it avoids deletions that can arise with nuclease-based methods37. Transposase mutagenesis has recently been used to discover chemical-dependent protein switches by targeting lactamase35, CRISPR-Cas947, GFP48, and an amino-acyl tRNA synthetase36 for random domain insertion.

To test whether transposase-mediated protein engineering can be used to discover novel sFd-ER that support ET in cells, transposase-mutagenesis was used to generate a library of sFd-ER (Figure 1a). We first created a synthetic transposon (Figure 1b) that contains a selectable marker (kanR) and transposase-binding sites (R1R2/R2R1) that are flanked by BsaI restriction site. This DNA is similar to a recently described synthetic transposon48, with one minor difference. The single stranded DNA overhangs adjacent to the transposase binding sites (R1R2 and R2R1), which encode a portion of the linkers that ultimately connect the ER-LBD and Fd, were modified so that they encode an Ala-Ala dipeptide. A three-step protocol was then used to create the vector library. In this protocol, MuA was initially used to randomly insert the synthetic transposon into a vector containing the Ml-Fd gene. To obtain the staging library, the product of this reaction was transformed in E. coli XL1-Blue, and vectors containing an inserted transposon were selected by growing cells on LB-agar plates containing kanamycin and chloramphenicol. This selection yielded ∼12,500 unique colonies, which exceeds the number of possible variants in our library (n = 6,482). The Fd-transposon gene hybrids were excised from this ensemble of vectors and subcloned into a vector that uses TetR to control Fd gene expression3. This size-selected library was transformed into E. coli XL1-Blue, and cells were grown on LB-agar plates to select for the desired vectors. This cloning step yielded ∼6,500 unique colonies, which exceeds the number of Fd-transposon gene hybrids (n = 296 unique insertion sites) that our method can create. To generate the final domain insertion library, we subcloned the gene encoding the ER-LBD (residues 302-552) in place of the synthetic transposon. This final step cloning yielded more than 500 unique colonies. In cases where this gene is inserted in frame, it is fused to the Fd through flexible linkers at its N- and C-termini (AAAGGGGSGGGGS and GGGGSGGGGSAAA, respectively).

Figure 1.
Figure 1. Library creation using random domain insertion.

(a) Transposon-mediated domain insertion of the ER-LBD into Ml-Fd generates a library of topological mutants. (b) Three-step protocol for library generation. In step 1, MuA inserts a synthetic transposon at different position of the staging plasmid. Restriction digestion of the resulting staging library with BbsI yields Ml-Fd containing and lacking a transposon insertion. In step 2, the Ml-Fd-transposon gene hybrids are isolated and cloned into a vector to create the size-selected library. In step 3, the transposon is replaced with a gene encoding ER-LBD to generate the domain-insertion library.

Selecting for protein electron carriers

To identify sFd-ER that coordinate a 2Fe-2S cluster and are able to support ET from a donor to an acceptor protein, we transformed the domain insertion library into E. coli EW1111, a strain with a chromosomal mutation within the gene (cysI) that encodes sulfite reductase. Because this strain is unable to grow on minimal medium containing sulfate as a sulfur source11, we transformed our library of vectors into cells along with a vector that constitutively express Zea mays FNR and SIR3. We then selected for growth on agar plates containing sulfate as the only sulfur source, aTc to induce sFd-ER expression, and 4-HT. This latter chemical was included in our growth medium because structural studies have shown that this and other antagonists stabilize an ER-LBD conformation with the N- and C-termini in closer proximity than that observed in the absence of bound ligand or in the presence of bound agonists49–51. With this selection, growth complementation was expected to only occur in cases where a sFd-ER was able to support ET from FNR to SIR (Figure 2a).

Figure 2.
Figure 2. Selecting for ferredoxins that exhibit electron transfer.

(a) Cells are only able to grow on minimal medium containing sulfate as the sulfur source if they express a Fd that supports ET from FNR to SIR. Selection of the domain-insertion library in presence of 4-HT at 37°C yielded five topological mutants, including ER-LBD insertions after Ml-Fd residues Gln25, Ala30, Gly35, Leu54, and Ile55. (b) Growth complementation observed after 48 hours with E. coli EW11 expressing FNR, SIR and the different Fd topological mutants in the presence of different 4-HT concentrations. After a 48 hour aerobic incubation at 37°C, absorbance at 600 nm was measured to quantify complementation. As a frame of reference, experiments were performed using Ml-Fd and Ml-Fd-C42A. Error bars represent ±1σ calculated from 3 biologically independent samples. Compared to the C42A mutant, growth was significantly enhanced in Ml-Fd under all conditions (P = 6.38×10-3), in sFd-55-ER, sFd-54-ER, and sFd-35-ER in the presence of ≥ 0.781 µM 4-HT (P = 3.31 × 10-4, 8.32 × 10-3, 8.68 × 10-3, respectively), in sFd-30-ER in the presence of ≥ 25 µM 4-HT (P = 8.85 × 10-3). sFd-25-ER did not show enhanced growth under any condition (P > 0.05).

Selection of our domain-insertion library on medium containing sulfate as the only sulfur source yielded 36 colonies that complemented growth in the presence of 4-HT, which represented ∼3% of the colonies plated under these conditions. Growth complementation performed in the absence of 4-HT yielded a smaller percentage of colonies. Sequencing colonies obtained from this selection (+4-HT) identified 5 unique topological mutants, which arose from insertion of the ER-LBD within Ml-Fd following residues 25 (sFd-25-ER), 30 (sFd-30-ER), 35 (sFd-35-ER), 54 (sFd-54-ER), and 55 (sFd-55-ER). These variants all arose from in frame insertions. Terminal fusions (LBD-Fd and Fd-LBD) were also observed. Additionally, one variant encoded Ml-Fd with the ER LBD inserted in the reverse orientation following Fd residue 33.

To establish whether the variants arising from LBD insertion within the Fd gene support ET in cells, we compared their growth complementation with native Ml-Fd and a mutant (C42A) that is unable to coordinate a 2Fe-2S metallocluster and support ET from FNR to SIR3. These experiments were performed in cells constitutively expressing Zea mays FNR and SIR and in liquid medium containing only sulfate as a sulfur source. Four out of the five sFd-ER variants presented growth complementation that is significantly higher than that observed with Ml-Fd-C42A in the presence of 4-HT, including sFd-30-ER, sFd-35-ER, sFd-54-ER, and sFd-55-ER (Figure 2b). This finding suggests that these fusion proteins all acquire a 2Fe-2S metallocluster and are capable of supporting ET from FNR to SIR. sFd-35-ER is related to a variant that was rationally designed in a previous study3. The sFd-35-ER discovered in our library has two additional residues (Ile-Asp) before the linker that connects the N-terminus of the ER LBD to Ml-Fd. These extra residues are added because MuA insertion of the transposon leads to a short sequence duplication (5 bp) at the insertion site 52.

Evaluating ligand regulation

Our selection was performed in the presence of 4-HT to identify sFd-ER that support ET in the presence of antagonist. To determine whether any of these variants require antagonist for efficient ET in cells, we examined the effects of different 4-HT concentrations on the growth of E. coli EW11 coexpressing each sFd-ER variant, Zea mays FNR, and Zea mays SIR. As controls, we performed similar experiments with Ml-Fd, which supports ET from FNR to SIR, and Ml-Fd-C42A, which does not support ET from FNR to SIR3. In the absence of 4-HT, these experiments revealed that all of the sFd-ER variants present growth complementation that is significantly lower than that of Ml-Fd (Figure 2b), but not significantly higher than the C42A mutant (P > 0.05, using a two-tailed, independent t-test). At the lowest 4-HT concentration tested (0.781 µM), sFd-54-ER and sFd-55-ER presented complementation that was significantly higher than that observed with the C42A mutant. At the highest 4-HT concentration (50 µM), these variants exhibited complementation that was not significantly different from Ml-Fd (P > 0.05, using a two-tailed, independent t-test). Taken together, these findings show that sFd-54-ER and sFd-55-ER exhibit chemical-dependent ET. sFd-30-ER and sFd-35-ER also displayed 4-HT induced complementation. However, the level of complementation observed at each 4-HT concentration was lower than that observed with sFd-54-ER and sFd-55-ER.

The chemical-dependent complementation observed with the sFd-ER fusion proteins could arise because 4-HT affects protein expression or because this ligand binds each sFd-ER and alters ET. To evaluate whether 4-HT influences protein expression, we created constructs that express the different sFd-ER with red fluorescent protein fused at their C-terminus through a twelve amino acid linker (GGSGGSGGSGGS) and evaluated the ratio of whole cell red fluorescence in the presence and absence of 4-HT (Figure 3). As a control, we performed identical experiments with a Ml-Fd-RFP fusion protein. With each RFP fusion protein, this analysis revealed similar red fluorescence in the presence and absence of 4-HT (P > 0.05, using a two-tailed, independent t-test). These findings provide evidence that the increased complementation observed with each fusion protein in the presence of 4-HT arises because this chemical alters the ET of each sFd-ER, rather than expression.

Figure 3.
Figure 3. The effect of 4-HT on the fluorescence of cells expressing Fd-RFP fusions.

The OD-normalized fluorescence obtained over a range of 4-HT concentrations was calculated for each Fd-RFP fusions analyzed, including Ml-Fd-RFP, sFd-25-ER-RFP, sFd-30-ER-RFP, sFd-35-ER-RFP, sFd-54-ER-RFP, and sFd-55-ER-RFP. As a negative control, cells containing a vector without RFP were evaluated (open circles shown with Ml-Fd-RFP). Circles and error bars represent the average and ±1σ, respectively, of n=3 biologically independent samples. Upon addition of 4-HT in DMSO, there was no significant increase in fluorescence compared to DMSO alone for any of the Fds tested (P > 0.05 using a two-tailed, independent t-test).

In vitro measurements

In a previous study3, we found that recombinant sFd-35-ER contains a 2Fe-2S following purification under aerobic conditions. In this protein, domain insertion occurred outside of the Ml-Fd tetracysteine motif that coordinates the 2Fe-2S metallocluster33. To evaluate whether domain insertion within the tetracysteine motif affects the ability of a sFd-ER to acquire a 2Fe-2S cluster, we overexpressed sFd-55-ER in E. coli Rosetta(DE3) and purified this protein aerobically using a combination of ion exchange and size exclusion chromatography. Recombinant sFd-55-ER was brown in color following the purification protocol, like Ml-Fd. In addition, this protein presented a visible circular dichroism (CD) spectrum that is similar to the spectrum observed with Ml-Fd (Figures 4a) and other plant-type 2Fe-2S Fds3, 53, 54. The presence of a 2Fe-2S cluster on purified recombinant sFd-55-ER following expression shows that this protein can acquire a metallocluster in the absence of 4-HT, and it suggests that this protein functions as a chemical-dependent ET switch by regulating the conformation of the protein rather than regulating the acquisition of the redox active iron-sulfur cluster.

To investigate if domain insertion affects the reduction potential of Ml-Fd, we compared the electrochemical properties of sFd-55-ER with Ml-Fd using protein thin film voltammetry. Experiments with sFd-55-ER could not be performed at pH = 7.0 in the presence of neomycin, as in a previous study3, because sFd-55-ER precipitated when concentrated under these conditions. For this reason, we performed experiments under more basic conditions (pH = 8.0) in the absence of neomycin. Under this condition, the midpoint potential of Ml-Fd was −428 mV (Figure 4b), a value that is 92 mV more negative than that observed at pH = 7.0 previously3. The midpoint potential of sFd-55-ER (−418 mV) was 10 mV more positive than Ml-Fd (Figure 4c). Experiments performed in the presence of 4-HT revealed that the Ml-Fd midpoint potential was not affected by this chemical, while the midpoint potential of sFd-55-ER shifted 45 mV more negative.

Figure 4.
Figure 4. Characterization of a purified Fd-ER-LBD fusion.

(a) Visible CD spectra of sFd-55-ER (red) and Ml-Fd (black) acquired under aerobic conditions at 20°C. (b) Ml-Fd and (c) sFd-55-ER were subjected to protein thin-film voltammetry in the absence (dashed line) and presence (solid line) of 4-HT. Baseline-subtracted data are plotted at the bottom, while raw square wave voltammetry data are shown at the top.

Mapping domain insertion onto structure

The sFd-ER variants that exhibit chemical-dependent ET in cells arise from domain insertion at different distances from the 2Fe-2S cluster in Ml-Fd (Figure 5a). Within the Ml-Fd structure33, the backbone cleavage sites occur at 14.7 Å (sFd-30-ER), 18.6 Å (sFd-35-ER), 15.8 Å (sFd-54-ER), and 19.4 Å (sFd-55-ER) from the center of mass of the 2Fe-2S cluster and the alpha carbon of the residue where cleavage occurs as measured using PyMOL. We also compared the domain insertion sites with the cysteines that coordinate the iron-sulfur cluster (Figure 5b-c). This analysis shows that two of the protein variants arise from LBD insertion within the tetracysteine motif (sFd-54-ER and sFd-55-ER), and the other two arise from domain insertion between the N-terminus of Ml-Fd and the tetracysteine motif (sFd-30-ER and sFd-35-ER). Mapping the backbone fission sites onto the structure of the Zea mays Fd-SIR complex reveals that three of the variants that display chemical-dependent ET (sFd-35-ER, sFd-54-ER, and sFd-55-ER) have SIR contacting residues divided among the Fd fragments55. In contrast, the FNR-contacting residues observed in the Zea mays Fd-FNR complex are not divided among the different Fd fragments in any of the active sFd-ER that display chemical-dependent ET in cells56.

Figure 5.
Figure 5. ER-LBD insertion sites mapped onto Fd structure.

(a) A comparison of the Ml-Fd primary structure with the domain insertion sites (red lines) and the residues that coordinate the 2Fe-2S cluster (yellow), make residue-residue contacts SIR (green), and make contacts with FNR (purple). (b) A comparison of the tertiary structure of Ml-Fd (PDB: 1RFK) with the domain insertion sites in red. (c) The Fd fragments generated by domain insertion illustrate the proximity of the insertion sites and the 2Fe-2S cluster. Insertion after Leu54 or Ile55 occurs within the tetracysteine motif while the other domain insertion sites (after Gln25, Ala30, and Gly35) yield a pair of fragments with all of the cysteines on one fragment.

Conclusion

Our combinatorial experiment identified three novel chemical-dependent ET proteins, illustrating the benefit of mining a library of topological mutants with a selection for protein design37, 44. Fluorescence analysis of protein expression in the presence and absence of ligand provided evidence that all three of these sFds regulate ET at the post-translational level. This ligand-induced ET could arise because of changes in: (i) iron-sulfur cluster biogenesis, (ii) iron-sulfur cluster stability, (iii) the electrochemical potential of the Fd, (iv) Fd partner binding, and/or (v) ET rates. In vitro studies of sFd-55-ER revealed that this protein acquires an oxygen-tolerant 2Fe-2S cluster in the absence of an ER ligand. This finding suggests that domain insertion does not regulate ET through iron-sulfur cluster biogenesis or stability. Additionally, these observations suggest that ligand does not alter the proximity of the Fd fragments fused to each end of the ER-LBD, since both fragments contain ligands required for 2Fe-2S cluster binding.

Electrochemical measurements revealed that sFd-55-ER has a midpoint potential that is more positive than Ml-Fd, and these measurements found that ligand binding induces a change in the midpoint potential of sFd-55-ER so that it becomes 45 mV more negative. A previous study examining the in vitro properties of sFd-35-ER observed a similar midpoint potential shift but in the opposite direction3. Our finding with sFd-55-ER suggests that ligand-induced changes in the midpoint potential of this protein is likely not responsible for the ligand-induced ET observed in cells, and it implicates changes in partner binding as the mechanism of ET regulation. To directly test this idea, additional in vitro studies will be required in the future that examine the effects of ligand binding on sFd-ER affinities for FNR and SIR as well as ET rates. Assays for monitoring FNR affinity for Fd interactions have been described previously57, and SIR affinity for Fd can be quickly measured by monitoring the effect of Fd concentration on sulfide production using a fluorescent reporter, sulfidefluor-758.

Previous studies examining sequence diversity in libraries created using transposon mutagenesis have revealed large biases52, with the abundance of insertion sites varying over three orders of magnitude59. This observation suggests that our library sampling was relatively sparse, and it highlights the importance of sampling millions of vectors at each step during library construction that involves bacterial transformations. With improved sampling, the selection described herein could be coupled to deep sequencing to generate a more comprehensive profile of the locations where Ml-Fd tolerates domain insertion44, 60 and other topological mutations, like circular permutation59 and fission39. Such experiments will aid in developing design rules for creating ferredoxin switches that regulate ET for diverse bioelectronic and metabolic engineering applications.

ACKNOWLEDGEMENTS

This research was supported by Office of Naval Research grant N00014-17-1-2639 (to J.J.S.). Additionally, JTA was supported by a Lodieska Stockbridge Vaughn Fellowship.

REFERENCES

  1. 1.↵
    TerAvest MA, Ajo-Franklin CM. Transforming exoelectrogens for biotechnology using synthetic biology. Biotechnol Bioeng. 2016;113(4):687–697.
    OpenUrlCrossRef
  2. 2.
    Schuergers N, Werlang C, Ajo-Franklin CM, Boghossian AA. A synthetic biology approach to engineering living photovoltaics. Energy Environ Sci. 2017;10(5):1102–1115.
    OpenUrl
  3. 3.↵
    Atkinson JT, Campbell IJ, Thomas EE, et al. Metalloprotein switches that display chemical-dependent electron transfer in cells. Nat Chem Biol. 2019;15(2):189–195.
    OpenUrl
  4. 4.↵
    Khoury GA, Fazelinia H, Chin JW, Pantazes RJ, Cirino PC, Maranas CD. Computational design of Candida boidinii xylose reductase for altered cofactor specificity. Protein Sci. 2009;18(10):2125–2138.
    OpenUrlCrossRefPubMedWeb of Science
  5. 5.
    Wang Y, San K-Y, Bennett GN. Cofactor engineering for advancing chemical biotechnology. Curr Opin Biotechnol. 2013;24(6):994–999.
    OpenUrlCrossRefPubMed
  6. 6.↵
    Chen X, Li S, Liu L. Engineering redox balance through cofactor systems. Trends Biotechnol. 2014;32(6):337–343.
    OpenUrl
  7. 7.↵
    Rumpel S, Siebel JF, Farès C, et al. Enhancing hydrogen production of microalgae by redirecting electrons from photosystem I to hydrogenase. Energy Environ Sci. 2014;7(10):3296–3301.
    OpenUrl
  8. 8.↵
    Yacoby I, Pochekailov S, Toporik H, Ghirardi ML, King PW, Zhang S. Photosynthetic electron partitioning between [FeFe]-hydrogenase and ferredoxin:NADP+− oxidoreductase (FNR) enzymes in vitro. Proc Natl Acad Sci USA. 2011;108(23):9396–9401.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    Agapakis CM, Ducat DC, Boyle PM, Wintermute EH, Way JC, Silver PA. Insulation of a synthetic hydrogen metabolism circuit in bacteria. J Biol Eng. 2010;4:3.
    OpenUrlCrossRefPubMed
  10. 10.
    Agapakis CM, Silver PA. Modular electron transfer circuits for synthetic biology: insulation of an engineered biohydrogen pathway. Bioeng Bugs. 2010;1(6):413–418.
    OpenUrlPubMed
  11. 11.↵
    Barstow B, Agapakis CM, Boyle PM, Grandl G, Silver PA, Wintermute EH. A synthetic system links FeFe-hydrogenases to essential E. coli sulfur metabolism. J Biol Eng. 2011;5:7.
    OpenUrlPubMed
  12. 12.↵
    Selberg J, Gomez M, Rolandi M. The Potential for Convergence between Synthetic Biology and Bioelectronics. Cell Syst. 2018;7(3):231–244.
    OpenUrl
  13. 13.↵
    Hernandez ME, Newman DK. Extracellular electron transfer. Cell Mol Life Sci. 2001;58(11):1562–1571.
    OpenUrlCrossRefPubMedWeb of Science
  14. 14.
    Weber KA, Achenbach LA, Coates JD. Microorganisms pumping iron: anaerobic microbial iron oxidation and reduction. Nat Rev Microbiol. 2006;4(10):752–764.
    OpenUrlCrossRefPubMedWeb of Science
  15. 15.↵
    Fredrickson JK, Romine MF, Beliaev AS, et al. Towards environmental systems biology of Shewanella. Nat Rev Microbiol. 2008;6(8):592–603.
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    Liu C, Gallagher JJ, Sakimoto KK, et al. Nanowire-bacteria hybrids for unassisted solar carbon dioxide fixation to value-added chemicals. Nano Lett. 2015;15(5):3634–3639.
    OpenUrlCrossRefPubMed
  17. 17.
    Sakimoto KK, Wong AB, Yang P. Self-photosensitization of nonphotosynthetic bacteria for solar-to-chemical production. Science. 2016;351(6268):74-77.
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    Kornienko N, Sakimoto KK, Herlihy DM, et al. Spectroscopic elucidation of energy transfer in hybrid inorganic-biological organisms for solar-to-chemical production. Proc Natl Acad Sci USA. 2016;113(42):11750–11755.
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    Jensen HM, Albers AE, Malley KR, et al. Engineering of a synthetic electron conduit in living cells. Proc Natl Acad Sci USA. 2010;107(45):19213–19218.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    Jensen HM, TerAvest MA, Kokish MG, Ajo-Franklin CM. CymA and Exogenous Flavins Improve Extracellular Electron Transfer and Couple It to Cell Growth in Mtr-Expressing Escherichia coli. ACS Synth Biol. 2016;5(7):679–688.
    OpenUrl
  21. 21.↵
    Liu J, Chakraborty S, Hosseinzadeh P, et al. Metalloproteins containing cytochrome, iron-sulfur, or copper redox centers. Chem Rev. 2014;114(8):4366–4469.
    OpenUrlCrossRefPubMed
  22. 22.↵
    Pierella Karlusich JJ, Lodeyro AF, Carrillo N. The long goodbye: the rise and fall of flavodoxin during plant evolution. J Exp Bot. 2014;65(18):5161–5178.
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.↵
    Atkinson JT, Campbell I, Bennett GN, Silberg JJ. Cellular Assays for Ferredoxins: A Strategy for Understanding Electron Flow through Protein Carriers That Link Metabolic Pathways. Biochemistry. 2016;55(51):7047–7064.
    OpenUrlCrossRefPubMed
  24. 24.↵
    Kim S-H, Harzman C, Davis JK, et al. Genome sequence of Desulfitobacterium hafniense DCB-2, a Gram-positive anaerobe capable of dehalogenation and metal reduction. BMC Microbiol. 2012;12:21.
    OpenUrlCrossRefPubMed
  25. 25.↵
    Peden EA, Boehm M, Mulder DW, et al. Identification of global ferredoxin interaction networks in Chlamydomonas reinhardtii. J Biol Chem. 2013;288(49):35192–35209.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    Burkhart BW, Febvre HP, Santangelo TJ. Distinct Physiological Roles of the Three Ferredoxins Encoded in the Hyperthermophilic Archaeon Thermococcus kodakarensis. MBio. 2019;10(2).
  27. 27.↵
    Terauchi AM, Lu S-F, Zaffagnini M, et al. Pattern of expression and substrate specificity of chloroplast ferredoxins from Chlamydomonas reinhardtii. J Biol Chem. 2009;284(38):25867–25878.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    Cassier-Chauvat C, Chauvat F. Function and Regulation of Ferredoxins in the Cyanobacterium, Synechocystis PCC6803: Recent Advances. Life (Basel). 2014;4(4):666-680.
    OpenUrl
  29. 29.↵
    Schlesier J, Rohde M, Gerhardt S, Einsle O. A Conformational Switch Triggers Nitrogenase Protection from Oxygen Damage by Shethna Protein II (FeSII). J Am Chem Soc. 2016;138(1):239–247.
    OpenUrl
  30. 30.↵
    Milton RD, Cai R, Sahin S, et al. The In Vivo Potential-Regulated Protective Protein of Nitrogenase in Azotobacter vinelandii Supports Aerobic Bioelectrochemical Dinitrogen Reduction In Vitro. J Am Chem Soc. 2017;139(26):9044–9052.
    OpenUrlPubMed
  31. 31.↵
    Moscatiello R, Zaccarin M, Ercolin F, et al. Identification of ferredoxin II as a major calcium binding protein in the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. BMC Microbiol. 2015;15:16.
    OpenUrl
  32. 32.↵
    Angeleri M, Zorina A, Aro E-M, Battchikova N. Interplay of SpkG kinase and the Slr0151 protein in the phosphorylation of ferredoxin 5 in Synechocystis sp. strain PCC 6803. FEBS Lett. 2018;592(3):411–421.
    OpenUrl
  33. 33.↵
    Fish A, Danieli T, Ohad I, Nechushtai R, Livnah O. Structural basis for the thermostability of ferredoxin from the cyanobacterium Mastigocladus laminosus. J Mol Biol. 2005;350(3):599–608.
    OpenUrlCrossRefPubMedWeb of Science
  34. 34.↵
    McLachlan MJ, Katzenellenbogen JA, Zhao H. A new fluorescence complementation biosensor for detection of estrogenic compounds. Biotechnol Bioeng. 2011;108(12):2794–2803.
    OpenUrlPubMed
  35. 35.↵
    Edwards WR, Busse K, Allemann RK, Jones DD. Linking the functions of unrelated proteins using a novel directed evolution domain insertion method. Nucleic Acids Res. 2008;36(13):e78.
    OpenUrlCrossRefPubMed
  36. 36.↵
    Thomas EE, Pandey N, Knudsen S, Ball ZT, Silberg JJ. Programming Post-Translational Control over the Metabolic Labeling of Cellular Proteins with a Noncanonical Amino Acid. ACS Synth Biol. 2017;6(8):1572–1583.
    OpenUrl
  37. 37.↵
    Atkinson JT, Wu B, Segatori L, Silberg JJ. Overcoming component limitations in synthetic biology through transposon-mediated protein engineering. In: Methods in Enzymology. Elsevier; 2019.
  38. 38.↵
    Gibson DG, Young L, Chuang R-Y, Venter JC, Hutchison CA, Smith HO. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods. 2009;6(5):343–345.
    OpenUrlCrossRefPubMedWeb of Science
  39. 39.↵
    Segall-Shapiro TH, Nguyen PQ, Dos Santos ED, et al. Mesophilic and hyperthermophilic adenylate kinases differ in their tolerance to random fragmentation. J Mol Biol. 2011;406(1):135–148.
    OpenUrlCrossRefPubMed
  40. 40.↵
    Engler C, Kandzia R, Marillonnet S. A one pot, one step, precision cloning method with high throughput capability. PLoS ONE. 2008;3(11):e3647.
    OpenUrlCrossRefPubMed
  41. 41.↵
    Frances A. Nobel Lecture. https://www.nobelprize.org/prizes/chemistry/2018/arnold/lecture/. Published December, 2018. Accessed May 4, 2019.
  42. 42.↵
    Romero PA, Arnold FH. Exploring protein fitness landscapes by directed evolution. Nat Rev Mol Cell Biol. 2009;10(12):866–876.
    OpenUrlCrossRefPubMedWeb of Science
  43. 43.↵
    Ribeiro LF, Warren TD, Ostermeier M. Construction of Protein Switches by Domain Insertion and Directed Evolution. Methods Mol Biol. 2017;1596:43–55.
    OpenUrl
  44. 44.↵
    Higgins SA, Savage DF. Protein Science by DNA Sequencing: How Advances in Molecular Biology Are Accelerating Biochemistry. Biochemistry. 2018;57(1):38–46.
    OpenUrl
  45. 45.↵
    Baird GS, Zacharias DA, Tsien RY. Circular permutation and receptor insertion within green fluorescent proteins. Proc Natl Acad Sci USA. 1999;96(20):11241–11246.
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    Tullman J, Guntas G, Dumont M, Ostermeier M. Protein switches identified from diverse insertion libraries created using S1 nuclease digestion of supercoiled-form plasmid DNA. Biotechnol Bioeng. 2011;108(11):2535–2543.
    OpenUrlCrossRefPubMed
  47. 47.↵
    Oakes BL, Nadler DC, Flamholz A, et al. Profiling of engineering hotspots identifies an allosteric CRISPR-Cas9 switch. Nat Biotechnol. 2016;34(6):646–651.
    OpenUrlCrossRefPubMed
  48. 48.↵
    Nadler DC, Morgan S-A, Flamholz A, Kortright KE, Savage DF. Rapid construction of metabolite biosensors using domain-insertion profiling. Nat Commun. 2016;7:12266.
  49. 49.↵
    Tanenbaum DM, Wang Y, Williams SP, Sigler PB. Crystallographic comparison of the estrogen and progesterone receptor’s ligand binding domains. Proc Natl Acad Sci USA. 1998;95(11):5998–6003.
    OpenUrlAbstract/FREE Full Text
  50. 50.
    Brzozowski AM, Pike AC, Dauter Z, et al. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature. 1997;389(6652):753-758.
    OpenUrlCrossRefPubMedWeb of Science
  51. 51.↵
    Shiau AK, Barstad D, Loria PM, et al. The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell. 1998;95(7):927–937.
    OpenUrlCrossRefPubMedWeb of Science
  52. 52.↵
    Haapa-Paananen S, Rita H, Savilahti H. DNA transposition of bacteriophage Mu. A quantitative analysis of target site selection in vitro. J Biol Chem. 2002;277(4):2843–2851.
    OpenUrlAbstract/FREE Full Text
  53. 53.↵
    Stephens PJ, Thomson AJ, Dunn JB, et al. Circular dichroism and magnetic circular dichroism of iron-sulfur proteins. Biochemistry. 1978;17(22):4770–4778.
    OpenUrlCrossRefPubMedWeb of Science
  54. 54.↵
    Ta DT, Vickery LE. Cloning, sequencing, and overexpression of a [2Fe-2S] ferredoxin gene from Escherichia coli. J Biol Chem. 1992;267(16):11120–11125.
    OpenUrlAbstract/FREE Full Text
  55. 55.↵
    Kim JY, Nakayama M, Toyota H, Kurisu G, Hase T. Structural and mutational studies of an electron transfer complex of maize sulfite reductase and ferredoxin. J Biochem. 2016;160(2):101–109.
    OpenUrlCrossRefPubMed
  56. 56.↵
    Kurisu G, Kusunoki M, Katoh E, et al. Structure of the electron transfer complex between ferredoxin and ferredoxin-NADP(+) reductase. Nat Struct Biol. 2001;8(2):117–121.
    OpenUrlCrossRefPubMedWeb of Science
  57. 57.↵
    Brandt ME, Vickery LE. Charge pair interactions stabilizing ferredoxin-ferredoxin reductase complexes. Identification by complementary site-specific mutations. J Biol Chem. 1993;268(23):17126–17130.
    OpenUrlAbstract/FREE Full Text
  58. 58.↵
    Lin VS, Lippert AR, Chang CJ. Cell-trappable fluorescent probes for endogenous hydrogen sulfide signaling and imaging H2O2-dependent H2S production. Proc Natl Acad Sci USA. 2013;110(18):7131–7135.
    OpenUrlAbstract/FREE Full Text
  59. 59.↵
    Atkinson JT, Jones AM, Zhou Q, Silberg JJ. Circular permutation profiling by deep sequencing libraries created using transposon mutagenesis. Nucleic Acids Res. 2018;46(13):e76.
    OpenUrl
  60. 60.↵
    Wrenbeck EE, Faber MS, Whitehead TA. Deep sequencing methods for protein engineering and design. Curr Opin Struct Biol. 2017;45:36–44.
    OpenUrl
Back to top
Posted May 23, 2019.
Download PDF
Email
Combinatorial design of chemical-dependent protein switches for controlling intracellular electron transfer
Bingyan Wu, Joshua T. Atkinson, Dimithree Kahanda, George. N. Bennett, Jonathan J. Silberg
bioRxiv 645978; doi: https://doi.org/10.1101/645978
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area