SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Listeria monocytogenes, a food-borne bacterial pathogen causing significant human mortality, propagates by expressing genes in response to environmental signals, such as temperature and pH. Listeria gene (lmo0799) encodes a protein homologous to the Bacillus subtilis YtvA, which has a flavin-light, oxygen or voltage (LOV) domain and a Sulfate Transporters Anti-Sigma factor antagonist (STAS) output domain that regulates transcription-initiation factor Sigma B in the bacterial stress response upon exposure to light. This could be significant for the pathogenesis of listeriosis because Sigma B has been linked to virulence of Listeria, and the Listeria Lmo0799 protein has recently been identified as a virulence factor activated by blue light. We have cloned, expressed heterologously in Escherichia coli and purified the full-length LM-LOV-STAS protein. Although it exhibits photochemical activity similar to that of YtvA, LM-LOV-STAS lacks an almost universally conserved arginine in the flavin-binding site, as well as another positively charged residue, a lysine in YtvA. The absence of these positive charges was found to destabilize retention of the flavin mononucleotide (FMN) chromophore in the LM-LOV-STAS protein, particularly at higher temperatures. The unusual sequence of the LM-LOV-STAS protein alters both spectral features and activation/deactivation kinetics, potentially expanding the sensory capacity of this LOV domain, e.g. to detect light plus cold.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Light acts as an environmental signal in a wide range of organisms, including many bacteria [1, 2]. LOV domains are ca 10 kD polypeptides that belong to the Per-Arnt-Sim (PAS) super family [3]. The LOV domain studied here is a module apparently sensing only light, although it shares structural similarity to modules sensing oxygen or voltage. As is the case for those other sensors, here light signals are relayed to an output domain in the same molecule, activating specific biochemical functions. The nature of the sensory signal detected depends on the cofactor bound to the LOV domain. LOV domains binding FMN or flavin adenine dinucleotide (FAD) [4-7] function as light sensors in plant, algal and fungal blue-light receptors [8]. Our group recently showed that light activation of a LOV domain containing histidine kinase from the pathogen Brucella abortus is involved in infection of mammalian cells [9]. Light-driven physiological effects were also documented for Caulobacter crescentus [10]. The role of light in bacterial pathogens is not currently understood, but the LOV domain studied here has partial homology to one which has been characterized in Bacillus subtilis. It contains a flavin–LOV-domain protein, YtvA, which has a STAS (Sulfate Transporters Anti-Sigma factor antagonist) output domain [11]. YtvA has been shown to control the bacterial stress response through the control of gene expression by the transcription-initiation factor sigma B when the bacterium is exposed to light [1].

Listeria monocytogenes is a food-borne bacterial pathogen that can cause serious infections in the human host [12]. Environmental signals, such as temperature, control the expression of Listeria genes involved in growth/survival both outside and inside a host [13]. Genomic search has identified a Listeria gene (lmo0799) homologous to YtvA, predicted to encode a LOV-STAS protein; we suspect this protein might mediate light regulation of gene expression through Sigma B. That mechanism could be of great significance in the pathogenesis of listeriosis (the infectious disease caused by Listeria) because Sigma B has been linked to virulence of Listeria where it has been shown to be involved in diverse functions such as invasion of intestinal epithelial cells, oral infection of guinea pigs and acid tolerance [14, 15, 2]. A recent publication reports that the lmo0799 Listeria gene that codes for the LM-LOV-STAS protein is a virulence factor activated by blue light and that the knockout mutant both has decreased virulence and is unaffected by light [16].

LOV-domain crystal and solution NMR structures show a typical PAS-domain protein fold with the flavin chromophore held noncovalently in a protein pocket via hydrogen bonding and hydrophobic interactions [17-20]. Most LOV domains bind FMN as chromophore [4, 5, 8, 9, 21]. However, in the fungus Neurospora crassa the LOV domain from the White Collar-1 (WC-1) photoreceptor binds FAD [6], whereas its VIVID LOV photoreceptor binds either FAD or FMN [7] and is equal physiological activity with either chromophore. In all photoactive LOV domains, the sulfur of a reactive cysteine is located within a few angstroms of carbon C(4a) of the flavin chromophore. Upon illumination, the LOV-domain chromophore undergoes a cyclic photoreaction which generates a cysteinyl-flavin adduct between the thiol sulfur of the reactive cysteine and the flavin [22-26]. In most examples, the adducts revert spontaneously in the dark, on timescales ranging from minutes to hours, depending on the system [9, 26, 27]. However, several Arabidopsis LOV-domain proteins and the LOV histidine kinase from Brucella exhibit truncated photocycles in which the cysteinyl-flavin adduct is stable and does not regenerate the ground state in the dark. Formation of adduct (a signaling state) causes a local chromophore/protein structural perturbation that subsequently propagates within the full-length proteins, ultimately activating an output domain that controls specific cellular functions [9, 10, 28, 29].

The photochemically active Listeria LOV-STAS protein (the lmo0799 gene product) has not been previously studied, and the mechanism of interaction with its putative target, the antisigma factor antagonist, remains unknown. We have cloned, expressed heterologously in E. coli and purified the full-length LM-LOV-STAS protein. We show here that the LM-LOV-STAS protein exhibits photochemical activity similar to that of YtvA, yet has a distinctly different active site structure resulting in unusual absorbance features, perturbed activation/deactivation kinetics and reduced stability, possibility expanding the role of this LOV domain as a light plus cold sensor.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

LM-LOV-STAS expression and purification

We cloned the LM-LOV-STAS gene (GenBank accession number AE017262) from the L. monocytogenes strain 4b F2365 genomic DNA. In brief, the DNA segment corresponding to the LOV-STAS domain was PCR amplified and cloned into the pET/100D TOPO plasmid (Invitrogen) at the 3′ end with an enterokinase site and an N-terminal 5xHis tag. Growth conditions were as previously described [24]. Although most LOV proteins are sufficiently stable to allow storage of cell pellets for months, the LM-LOV-STAS protein shows poor stability at −20°C, requiring purification from cell pellet within approximately a week of initial production.

For purification, cells were thawed and resuspended with the wash buffer (300 mm NaCl, 50 mm phosphate, pH 7) used in the TALON cobalt-based resin affinity purification column (Clonetech). Phenylmethylsulfonyl fluoride (final concentration 1 mm) was added to the buffer to minimize serine protease activity. Cells were broken by two passages through a French press cell at 20 000 psi (lb in−2). Unbroken cells and debris were removed by centrifugation at 33 300 g for 25 min. Protein purification procedures using the TALON affinity resin in a batch process were as described previously [9]. Presence of the protein was verified by its absorption spectrum. Exchange to a buffer containing 5 mm phosphate and 5 mm NaCl at pH 8 was carried out using a Sephadex G25 column (GE Healthcare). Depending on the concentration required for a particular measurement, the protein sample was concentrated with Vivaspin concentrators (GE Healthcare). Due to the instability of the protein, all experiments were carried out immediately following protein purification. The H39R mutant was generated using a QuickChange site-directed mutagenesis kit (Stratagene).

SDS-PAGE gel electrophoresis

Protein purity was confirmed by gel electrophoresis using PhastGel PhastSystem (GE Healthcare), which was programmed for resolving a polyacrylamide gradient Phast gel of 8–25% (GE Healthcare). SeeBlue Plus2 Pre-Stained protein standard (Invitrogen) was used to estimate the molecular size.

Chromophore identification

We employed high performance liquid chromatography coupled to tandem mass spectrometry (Thermo Finnigan LC/MS/MS) to identify the extracted chromophore. The protein sample was dialyzed twice against Nanopure water (Barnstead NANOpure Diamond) for a total time of 16 h, and was then subjected to a 2% tricholoacetic acid treatment for protein precipitation. The sample was centrifuged at 4°C and 18 100 g in a Sorvall SS-34 rotor for 15 min to remove the precipitated proteins. Supernatant containing the chromophore was lyophilized and resuspended in 35% ethanol to a concentration of flavin chromophores adequate for LC/MS/MS (only very faint yellow color). FAD (Sigma-Aldrich), FMN (flavin mononucleotide; MP Biomedicals) and riboflavin (Sigma-Aldrich) were used for comparison.

Fluorescence spectroscopy

Fluorescence spectra were acquired at room temperature using a Varian Cary Eclipse fluorescence spectrophotometer (version 3), with a slow scanning speed (120 nm min−1) and medium detector sensitivity. Emission spectra excited at 270 nm (the third singlet excited state, S3) were recorded from 280 to 400 nm.

Light-induced absorption changes on the microsecond timescale

Data acquisition and analysis were carried out using equipment of local construction and computational routines previously described [30]. Briefly, absorbance difference spectra at time delays between 1 and 30 μs after photoexcitation were collected using white light from a flashlamp dispersed by a spectrograph and detected with an Andor DH520 intensified charge coupled device detector whose gate was ~100 ns. A dye laser pumped by the third harmonic of an Nd:YAG laser provided vertically polarized 477 nm actinic light pulses of approximately 10 ns duration, with an energy flux of 80 μJ mm−2. A flow system using a computer-controlled, stepper motor-driven syringe pump delivered 1 μL of fresh LM-LOV-STAS sample for each laser flash. That volume exactly matches the irradiated volume as described by the optical path dimensions given below. Absorbance data from ~50 samples at each time delay after photoexcitation were averaged. Probe light from a xenon flashlamp (FXQ-856, approximately 5 μs duration) was collimated before it passed through a linear polarizer. The beam was then refocused onto an opaque mask with a 1 mm × 0.5 mm rectangular aperture, which confined the beam to the photolyzed region under the cuvette insert [31]. The laser pulse traversed the sample perpendicular to the path of the flashlamp probe light. The optical path lengths for the probe and laser light were 2 and 0.5 mm, respectively. The white light used for probing the absorbance was linearly polarized at the magic angle (54.7°) relative to the laser polarization axis to prevent rotational diffusion artifacts, although these were not expected on the timescale used here for LM-LOV-STAS [32]. The temperature for all measurements was 20 ± 0.1°C. Adduct formation caused by the probe light sources is insignificant as the system has been shown to photolyze <0.5% of a bovine rhodopsin sample, which has a quantum yield on the order of unity.

Light-induced changes at longer times

Following photolysis of LM-LOV-STAS to form the adduct, time-resolved optical difference spectra were collected on the seconds-to-hours timescale with a Hewlett Packard 8452A diode array spectrometer, driven via a GPIB interface (IEEE 488) and LabVIEW software (LabVIEW; National Instruments) for automated control of data acquisition, which was developed previously in our laboratory [9]. The optical path length was 1 cm. A continuous, broad-band excitation beam was provided by a fiber optic illuminator (V-LUX 1000; Volpi Mfg.), filtered through a Corning Glass (Model 5–56) filter, with maximum transmission at 420 nm and approximately 80 nm half bandwidth. The sample was illuminated for 1 min to reach near saturation, creating as much of the excited population as possible. The temperature was controlled using a circulating water bath (Isotemp Refrigerated Circulator, Model 900; Fisher Scientific), with fluctuations of ±0.5°C. The temperature was measured inside the cuvette with a thermocouple (Fluke). Adduct formation contributed by the diode array irradiation is minimal (<1%), as consecutive absorption spectra of the LM-LOV-STAS protein sample show virtually no loss in the 450 nm feature.

Data analysis using global fitting

Data were analyzed using Matlab-based programs (The Mathworks) previously described [33]. In brief, absorption data were presented in a matrix format with times in a 1, 3, 6, 10 and 30 μs progression and wavelengths (ranging from 300 to 700 nm) at 4 nm intervals (100 wavelengths). The data were subsequently subjected to singular value decomposition to reduce noise. Global nonlinear least-squares fitting was used to decompose kinetic changes at all measured wavelengths into a sum of exponential components,

  • display math(1)

Each term after the first in Eq. (1) contains an apparent rate constant, 1/τi, for an observed kinetic change and the wavelength-dependent amplitudes, bi(λ), which represent the spectral change associated with each rate constant. Together the bi(λ) constitute the so-called b-spectra. Spectral modeling was applied to extract kinetic information from the b-spectra and the apparent rate constants [34, 35]. In general, the b-spectra reflect spectral changes associated with each exponential process and provide information to model kinetic schemes. It should be noted that the bi(λ), for i  1 describe changes that are time dependent, i.e. they are associated with a lifetime. However, the change described by b0(λ), if one is observed, is time independent and represents incomplete return of the sample to the prephotolysis state [33, 36].

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Sequence comparison

Three light-sensing proteins: B. subtilis YtvA (NCBI accession number NP_390912), Oceanobacillus iheyensis LOV-STAS (a protein kinase of a deep-sea halobacterium with NCBI accession number NP_691509) and oat LOV2 (a protein extensively studied in our laboratory) were chosen to compare for homology with LM-LOV-STAS's LOV domain. Sequence homology comparison was done by LALIGN (ExPASy). Fig. 1 shows a sequence homology comparison between the LOV domains of B. subtilis YtvA, L monocytogenes, O. iheyensis and oat LOV2.

image

Figure 1. Sequence comparison of LOV domains between LM-LOV-STAS, YtvA, O. Iheyensis LOV-STAS (OI LOV-STAS) and oat LOV2. Residues of the chromophore binding pocket are highlighted in blue. The glutamic acid and lysine residues forming the conserved salt bridge are highlighted in magenta. Underlined amino acids interact with the chromophore.

Download figure to PowerPoint

LM-LOV shares 54% homology with YtvA, 48% with O. iheyensis LOV domain and 46% with oat LOV2. All LOV domains that have been previously studied have the reactive cysteine in the highly conserved flavin binding pocket, with the sequence GXNCRFLQ [8, 37]. However, in LM-LOV-STAS, the highly conserved arginine that occurs in the flavin binding pocket of other LOV domains is replaced by a histidine. The X group in the sequence, which is not highly conserved among LOV domains, is lysine in YtvA and O. iheyensis LOV-STAS, and arginine in oat LOV2. These two potentially positively charged side chains have significantly different character from the neutral serine which occurs there in the LM-LOV-STAS X position.

Full-length LM-LOV-STAS has 253 amino acids (Fig. 2) and a predicted molecular mass of 28.8 kD without the purification tag.

image

Figure 2. Complete amino acid sequence of LM-LOV-STAS from the N-terminus to the C-terminus. The LOV domain extends from V19 to V121 (underlined).

Download figure to PowerPoint

The tag adds about 0.6 kD. LM-LOV-STAS has only one cysteine in addition to the one that is reactive in adduct formation. This second cysteine is located within the STAS domain (C153). In addition, unlike most LOV-domain proteins, the highly conserved tryptophan (W79 in the YtvA sequence in Fig. 1) is replaced by methionine in LM-LOV-STAS [37]. In fact, unlike B. subtilis YtvA, LM-LOV-STAS is completely devoid of tryptophan residues. The highly conserved arginine in the flavin binding pocket (R39) is part of the hydrogen-bonding network that stabilizes the chromophore [37]. Having it replaced by histidine—a basic, polar, yet smaller amino acid—is likely to affect the chromophore stability, photochemistry and perhaps even functional properties. The LM-LOV-STAS protein contains 12 tyrosines, which in the absence of any tryptophan become the main contributors to protein luminescence. The light emission originating from tyrosine/tyrosinate could yield useful structural information, e.g. ionization state of side chain and also polarity of its surroundings.

Spectroscopic properties of LM-LOV-STAS protein

The absorption spectrum of LM-LOV-STAS (Fig. 3, solid line) in the ground state has the typical vibronic features near 450 nm (first singlet excited state, S1) that have been observed among all the LOV domains published so far.

image

Figure 3. Absorption spectrum of the ground state of LM-LOV-STAS (solid) and oat LOV2 (dashed). Vibronic structures of the band at 450 nm are similar for both proteins. However, spectral features in the 370 nm region are different, presumably due to different electrostatics in the chromophore-binding environment.

Download figure to PowerPoint

The 370 nm band (the second excited singlet state, S2) is markedly different from that of YtvA and other LOV domains. Typically, the vibronic structure of the 370 nm band of other LOV domains shows a “rounded” absorption maximum (round top) composed of a lower and higher amplitude vibrational peaks at 350 and 370 nm, respectively (oat LOV2 spectrum, dashed line). In contrast, the 370 nm band from LM-LOV-STAS has two vibronic peaks of equal intensity, leading to a broad band (a flat top). To investigate whether this spectral difference was due to substitution of histidine for arginine in the binding site, we explored the possibility of restoring the original electrostatic state. At pH 8, the typical environmental conditions used in our work, arginine is still positively charged, but histidine is predominantly neutral, resulting in the loss of a positive charge in the chromophore-binding region. We attempted to titrate the protein into the acidic pH range to recover the positive charge in the histidine, and possibly the “rounded” absorption maximum at 350 and 370 nm. Due to the instability of the protein, a significant amount of the chromophore was released, which masked the true features of the protein at acidic pH. Alternatively, we created the LM-LOV-STAS mutant H39R to probe any spectroscopic, adduct formation and/or decay perturbation that could possibly be due to the histidine. However, the mutation resulted in expressed protein with no chromophore (as verified with BCA assay after purification). This could be due to the net positive charge carried by the arginine and a lack of counter ion within the H39R mutant LM-LOV-STAS protein.

Fluorescence emission in the aromatic region of LM-LOV-STAS is unique among LOV domains described so far in that there is no contribution from tryptophan fluorescence. Fig. 4 shows the emission spectra of LM-LOV-STAS in the 300–400 nm range (excitation at 270 nm) at pH 7.7.

image

Figure 4. Emission spectra of tyrosine fluorescence of LM-LOV-STAS excited at 280 nm. The peak at 310 nm is likely to be due to the protonated tyrosine, whereas the broad feature at 340 nm could be due to tyrosinate.

Download figure to PowerPoint

At near neutral pH the emission is dominated by the band characteristic of neutral tyrosine at around 300 nm, but also contains a significant contribution of longer wavelength luminescence, possibly originating from ground-state tyrosinate and/or from tyrosine that has undergone transient excited state proton transfer to neighboring acceptor groups, thus emitting from its ionized excited tyrosinate state [38].

Protein and chromophore characterization

The electrophoretic mobility of the LM-LOV-STAS in SDS-PAGE gels (see Supplementary Material, Figure S1) agrees with the expected molecular size of the protein (29.4 kDa). However, SDS-PAGE shows the presence of both monomers and dimers, with monomers being the dominant species at low protein concentrations (450 nm absorbance in 1 cm path <0.1), and with increasing oligomerization being observed at higher protein concentrations (>0.3 OD at 450 nm). The observation of protein aggregates, even in the presence of SDS, suggests strong protein–protein interactions in the full LM-LOV-STAS construct, which could have functional implications. This observation differs from what was reported in YtvA, where aggregation was observed in the isolated LOV domain but not in the full construct [39]. When the full LM-LOV-STAS construct was eluted with 80 mm imidazole buffer, the fractions (OD450 = 0.07) were more homogeneous, containing primarily the monomer and a trace amount of proteolytic fragments.

Chromophore analysis

While most of the LOV domains, like the “Phot” protein family [5], have FMN as their chromophore, some instead bind FAD [6, 7]. Work in our lab (M. Frederickson, private communication) shows the presence of both FMN and FAD in the bacterial LOV histidine kinase of Brucella melitensis. Most of the LOV-domain chromophore identification was previously done using ethanol extraction at neutral pH followed by thin layer chromatography [5]. Application of this technique consistently yielded FMN as the flavin bound to the LOV domain. Those experimental conditions were not ideal, especially for FAD, because the phosphodiester bond is not stable and its cleavage into FMN and adenosine monophosphate is catalyzed by hydroxide ion under physiological conditions (~pH 7) [40]. As a result, we chose the low-pH trichloroacetic acid extraction method, which stabilizes the FAD chromophore, followed by the sensitive HPLC/MS detection method as our means for identifying the chromophore. FAD, FMN and riboflavin have the isoalloxazine ring as their basic structure, but each flavin has a different side chain. Table 1 summarizes the results from HPLC/MS measurements. While all the flavin standards were recovered and detected using identical analytical conditions, no FAD was detected in the protein extracts. The predominant chromophore is clearly FMN. The presence of a small amount of riboflavin may result from hydrolysis or from the breaking up of FMN in the mass spectrometer.

Table 1. Mass spectrometry results of the identification of the extracted chromophore, showing that the LM-LOV-STAS chromophore is FMN. The trace amount of riboflavin observed is likely to be from the breakdown of FMN, either during extraction process or in the mass spectrometer. The difference between the FAD from the extracted sample and the standard could possibly be due to the protonation of the phosphate. Total amount of flavin detected was normalized to 100%
FlavinExtracted sample MS peak (m/z)Standard MS peak (m/z)
Riboflavin377.29 (1. 2%)377.40
FMN457.04 (98%)457.08
FAD784.32 (<1%)785.97

Photochemical characterization of recombinant LM-LOV-STAS

Early photocycle

As observed in other LOV domains previously studied, photoexcitation of the LM-LOV-STAS produces a triplet species, indicated by a broad absorption band in the difference spectrum measured at 1 μs (Fig. 5, upper panel) beginning near 500 nm and extending beyond 700 nm that decays away on the microsecond timescale. The lower panel of Fig. 5 shows the b-spectra resulting from global exponential fitting.

image

Figure 5. Time-resolved absorbance changes during LM-LOV-STAS adduct formation. Upper panel. The absorption difference spectra of LM-LOV-STAS after excitation with a 477 nm laser pulse. Spectra were collected at 1, 3, 6, 10 and 30 μs. Lower panel. Results of global multiexponential fitting of difference absorption spectra from LM-LOV-STAS. b1 describes the absorbance change taking place with an apparent time constant of 5.4 μs, and b0 is the difference spectrum seen after the 5.4 μs process is complete.

Download figure to PowerPoint

Assignment of this spectral feature to a triplet state was first reported for the early intermediate species of oat LOV2 [26]. In LOV2, the intersystem crossing that leads to triplet formation has a time constant of approximately 2 ns [41]. Here we were not able to resolve that step kinetically. The triplet-species decay is accompanied by the formation of the flavin-cysteinyl adduct, with a signature absorption peak around 380 nm (LM-LS380). As in other LOV domains, not all the triplet species of LM-LOV-STAS decays through the adduct formation process. From the absorption difference spectra of LM-LOV-STAS, we estimate that the branching ratio is 40–60%, to adduct and to ground state, respectively, which is a lower adduct yield than for the other similar systems previously investigated in our lab (oat LOV2, Brucella melitensis LOV histidine kinase) [9]. The branching ratio was calculated as the ratio of the persistent 450 nm bleach at long times to the full bleach at early times. The difference is presumably caused by the return of some triplet state species directly to the 450 nm ground state. The absorbance decrease at 450 nm in the 1 μs trace measures total ground-state population converted to the triplet state, whereas the long-lived bleach at 450 nm measures the fraction of the ground state that converts into the 380 nm adduct at long times. The data show that this latter fraction is only 40% of the total, which indicates that 60% must have decayed from the triplet state back to the ground state. Thus, the reaction scheme requires the inclusion of a thermal branching, 2:3, (0.4:0.6), forward into the LM-LOV-STAS 380 nm adduct state (τF) and backward to the ground state (τR). The oat LOV2 domain has a branching ratio of 1:1, and the adduct formation shows an apparent time constant (τapp) of 2 μs in the branching scheme. That apparent time constant was interpreted as the sum of equal forward and backward branching rates 4 μs (1/τapp = 1/τF + 1/τR). Unlike most of the LOV domains previously studied, our global exponential fitting shows that LM-LOV-STAS has a significantly slower apparent time constant of adduct formation of 5.4 μs (adding a second exponential did not improve the fitting quality significantly). The apparent formation rate constant in this branching pathway becomes

  • display math(2)

As τapp is known, using the branching ratio kF/kR = 0.4/0.6 we determined that τF = 15 μs, τR = 8 μs.

Late photocycle

The kinetics of LM-LOV-STAS adduct decay back to the ground state were measured by monitoring the recovery of the absorption from 300 to 550 nm in the dark at 20, 21.5, 23 and 24.5°C. The upper panel of Fig. 6 shows the difference spectra recorded as the sample returns spontaneously to the ground state at 20°C. The lower panel of Fig. 6 shows the b-spectra resulting from global exponential fitting of the data in the upper panel.

image

Figure 6. Absorbance changes associated with LM-LOV-STAS adduct decay. The upper panel shows the time-resolved light-induced absorption changes for LM-LOV-STAS following light excitation at 20°C. Spectra were taken every 7.5 min. The time indicated above the traces is the last delay time, and the time indicated at the bottom of the traces corresponds to the first delay time. Arrows indicate the direction of the spectral changes in the indicated region with time. The lower panel shows the b-spectra resulting from global exponential fitting, with a lifetime of 91 min for the b1 component.

Download figure to PowerPoint

One exponential produced satisfactory fits of the data at these temperatures. The time constants associated with b1 (τrec) at the various temperatures are listed in Table 2. b0 is the difference spectrum between the final product extrapolated to infinite time and the protein before photoexcitation. By definition, b0 describes changes that are time independent and will become zero if the recovery is a completely cyclic process. The nonzero b0 found under these conditions suggests that some of the protein might be permanently bleached upon illumination, which was also observed in similar experiments conducted with YtvA, for which slow recovery kinetics were also reported [27].

Table 2. Recovery lifetimes and thermodynamic data of LM-LOV-STAS
 Without imidazoleWith 10 mm imidazole
Lifetime (τrec), min, at 20°C914.5
Enthalpy of activation (kJ mol−1)140100
Entropy of activation (cal mol−1·K)−70−0.07

At temperatures above 26°C, an additional process besides simple adduct decay began to contribute to the observed kinetics. Beginning at 26°C (Fig. 7), chromophore release became significant as shown by the small positive absorption near 500 nm in the difference spectra at later delay times.

image

Figure 7. Time-resolved light-induced absorption changes for LM-LOV-STAS after light excitation at 26.5°C. Spectra were taken every 5 min. The positive absorption around 500 nm is indicative of the formation of a new species that had not been present prior to photoexcitation, possibly the free chromophore released during the adduct breakage.

Download figure to PowerPoint

This increase is suggestive of the formation of a new species not present in the dark state. Global analysis with one exponential was no longer sufficient. This was demonstrated by larger residuals (data not shown), which are the differences between experimental data and data calculated using the parameters from exponential fitting. Fig. 8 shows the full photocycle of LM-LOV-STAS at pH 8 and room temperature.

image

Figure 8. Photocycle scheme of the LM-LOV-STAS. Blue-light photolysis produces a triplet species on the submicrosecond timescale, which decays through a branching scheme into the both ground state and the adduct state. All the time constants in this scheme are based on measurements done at 20°C. Numbering is based on the full amino acid sequence.

Download figure to PowerPoint

Enthalpy of activation

The observed long decay to the ground state suggests the presence of a thermally activated process. Eyring's transition theory states that

  • display math(3)

where kb is the Boltzmann constant, h is the Plank constant, ΔH is the enthalpy of activation, R is the gas constant and ΔS is the entropy of activation. We determined from the temperature dependence of the kinetics both the enthalpy and entropy of activation for the adduct decay process, which were 140 kJ mol−1 and −70 cal mol−1·K, respectively.

As with other LOV domains the LM-LOV-STAS decay kinetics were significantly accelerated by the presence of imidazole as an exogenous catalyst [42]. We used 10 mm imidazole to catalyze the reaction and because of the accelerated recovery rate, measured the kinetics at 13, 16, 18 and 20°C. Table 2 compares the recovery lifetimes observed at 20°C and the thermodynamic properties with and without imidazole.

The spectral features resulting from measurements of the exogenously catalyzed reactions are the same as those of the endogenously catalyzed ones below 26°C. The enthalpy of activation of the catalyzed reaction (Ea,cat) from the Arrhenius plot is about 100 kJ mol−1, which is significantly lower than that for the endogenously catalyzed reaction. It should be noted that the entropy of activation for the imidazole-catalyzed reaction is less negative, indicating that the base makes the recovery more entropically favorable.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Sequence difference and UV–Vis absorption spectrum lineshape

Because our interest here is the photochemically active component of the various blue-light sensitive proteins, only sequences of their LOV domains were compared (Fig. 1). One of the uncommon features of the LOV domain from L. monocytogenes is that both arginines in the highly conserved chromophore-binding pocket are replaced, one by serine (S36) and the other one by histidine (H39). This is unusual because most LOV domains have at least one of the highly conserved arginines, if not both. [37]. Serine does not have a titratable side chain, whereas histidine has a pKa of 6. At pH 8 or 7, the pH at which most of the spectroscopic measurements were carried out, the loss of both arginines means that two positive charges are absent in the binding region, which affects the band shape and frequencies of the second singlet exited state transition band (S2) in the near-UV absorption spectrum, which, for flavin, is known to be more sensitive to the local electrostatic environment than S1 [43, 44]. The difference in dipole moment, Δμ, between the ground state and S1 is very small, so as ΔE =− Δμ·F, where F is an externally applied electric field (or local electrostatic field in the molecular pocket), the effect on the S1 transition energy is negligible. However, the dipole moment of S2 is greater than that of S1 by 60% [45]. Therefore, change in the dipole moment during S0 to S2 will result in a much larger shift than occurs in S0 [RIGHTWARDS ARROW] S1, as observed. In addition, it is apparent from the amplitudes of the vibronic structures for LM-LOV-STAS, as compared with oat LOV2, that they exhibit a large variation in their Franck-Condon factors. This is an additional argument for the effect of the unique amino acid residues in LM-LOV-STAS on the local environment in the FMN binding pocket.

Structures affecting LM-LOV-STAS stability

We attempted to model the structure of the binding pocket of the LM-LOV domain using the crystal structure of the dark-state YtvA (PDB 2PR5) in Pymol. This is a plausible approach because of the excellent sequence homology (54%) between these proteins without any gaps in the number of residues. However, we cannot identify any ionizable residues in the chromophore binding pocket of LM-LOV-STAS that could provide the negative counter ion to a positively charged histidine, suggesting that the local environment has a high effective pH or is highly hydrophobic. In YtvA, the distance between the Arg39, adjacent to the reactive cysteine, and the phosphate oxygen of FMN is 2.88 Å (numbering is based on the LOV-domain sequence, as shown in Fig. 1), which is within hydrogen-bonding distance and is believed to participate in hydrogen-bonding stabilization of the FMN chromophore [37]. However, in LM-LOV-STAS, H39 in the binding pocket could compromise the stability of the protein. We observed that the flavin is slowly released and a large fraction of the free chromophore is observed within a few days of storage. In addition, there is significant release of chromophore upon photoexcitation. This stability issue was not reported for YtvA [37] and is not observed in other LOV domains studied in our laboratory, whether they are isolated LOV domains or part of a full protein construct. It is possible that losing those arginine-related hydrogen bonds leaves the FMN ribityl chain and the neighboring glutamine “dangling,” resulting in a weakly bound, vulnerable interaction between the chromophore and the protein pocket, which could account for the low stability of the protein and a decreased chromophore-binding affinity. The relative instability of the chromophore retention in LM-LOV-STAS may arise from such a weak (or nonexistent) hydrogen-bonding network.

Given the absence of arginines in its chromophore binding pocket, which appear to be important for stabilizing FMN binding, and the observed temperature dependence of FMN release, LM-LOV-STAS seems to be a poorly designed light sensor. The fact that specific residues have been changed to reduce stability of bound FMN raises the possibility that an additional, unappreciated functional role exists for LM-LOV-STAS, such as signaling a particular combination of conditions like light and cold. That would be of considerable interest as Listeria is a pathogen often associated with refrigerated foods such as cheese. The existence of new sensing modalities in LM-LOV-STAS may not be surprising given the already broad range of phenomena detected by LOV proteins [8].

LM-LOV-STAS photocycle

The LOV domain from LM-LOV-STAS appears to follow the same mechanism as that of the other previously studied LOV domains, leading to the formation of the flavin-cysteine adduct like oat LOV2 and phot-LOV1 from Chlamydomonas reinhardtii [26, 46], with an absorption peak at 390 nm. However, those of full-length proteins, like LM-LOV-STAS and YtvA [21] are slightly blue shifted to 380 nm, and the kinetics of the LM-LOV-STAS adduct formation is much slower compared with that of the other LOV domains [21, 26, 46]. Whereas most of the latter have an apparent time constant of 2 μs, that of LM-LOV-STAS has a time constant of 5.4 μs, and it also has a low yield with only 40% of the triplet state proceeding to form the adduct. The b-spectra of the intermediates do not show any unusual features that can account for such slow kinetics. It is not apparent to us whether there is a correlation between the loss of two positive charges and the slow adduct formation kinetics. One way to probe the effect would be through further mutational analysis. Although both YtvA and LM-LOV-STAS feature slow adduct decay kinetics, as suggested by the difference spectra (Fig. 7), there is chromophore release at high temperature in LM-LOV-STAS. This was unexpected because YtvA and LM-LOV-STAS share a high level of homology, and YtvA has been reported to be very stable, allowing temperature-dependent adduct decay measurements at temperatures as high as 37°C to be made [27]. We have estimated the FMN binding constant by measuring the fraction of free and bound chromophore at successive dilutions as described previously [47]. This method is based on the estimation of free and bound chromophores from the potassium iodide collision quenchable and nonquenchable fluorescence fractions, respectively. This preliminary measurements of the FMN binding equilibrium constant (data not shown) suggests that the equilibrium fractions of bound and free FMN at the concentrations used in our spectroscopic experiments are nearly equal (approximately 10–20 μm), giving an approximate binding constant in the 10 micromolar range. This is considerably weaker binding than in oat LOV2 for which we observe at least one order of magnitude stronger binding (1 micromolar range). It is likely that at higher temperature, the protein becomes more flexible which combined with a loosely bound chromophore, expedites its release as the adduct state decays.

The quantum efficiency for the production of the adduct signaling state by the isolated oat LOV2 domain of phototropin 1 is approximately 0.45 [24]. This value is consistent with the 1:1 branching ratio observed, which suggests that about 90% of the singlet excited state undergoes intersystem crossing into the triplet. This value is also consistent with the rather low fluorescence quantum yield of FMN in the protein pocket (~9%) when compared with that of free FMN in solution (~20%) [26]. Essentially little thermal loss exists in this system (0. 45 forward, 0. 45 backward and ~0. 1 radiation). If the early relaxation processes of LM-LOV-STAS are similar to those of the other LOV-domain proteins, one would expect, based on the forward branching ratio, that the overall quantum efficiency for this photoreceptor protein would be somewhat lower than that of the phototropin chromophore. The chromophore loss upon photoexcitation and the instability of LM-LOV-STAS did not allow us to carry out meaningful measurements of either the fluorescence and/or the photochemical quantum yields, due to the simultaneous presence of free and bound FMN in the preparations. This free FMN fraction contributes to both absorption and fluorescence, yet is photochemically inactive. Its concentration also varies from sample to sample and within a given sample as it ages. If this problem can be surmounted in future work, it would be of interest to determine the intersystem-crossing yield, the fluorescence quantum yield and the photochemical quantum yield for the production of the adduct in LM-LOV-STAS, to obtain a full description of the relaxation processes following the molecular excitation to the S1 state.

Spontaneous LM-LOV-STAS adduct decay involves a significant molecular conformation change

The activation enthalpy was 140 kJ mol−1 and the entropy of activation is −70 cal mol−1·K. This indicates that the process is markedly thermally activated and that it is entropically unfavorable, presumably requiring ordering of the active site structure or a specific electrostatic configuration to generate the transition state. LOV-domain adduct recovery is known to be base catalyzed, involving a proton transfer reaction [49]. The high enthalpy of activation is consistent with the involvement of hydrogen-bond breaks and/or protonation changes in rate-limiting steps of adduct recovery.

Although LM-LOV-STAS and YtvA share a high level of homology, they differ in many aspects. Specifically, the instability issue reported here, to our knowledge, has not been reported for any of the photoreceptor LOV domains. It is intriguing to speculate whether it is a property with functional significance, i.e. a light plus cold sensor. With regards to its physiological function, the virulence factor of Listeria has been shown to be activated by blue light [16]. However, further studies will be required to elucidate both the molecular mechanism of activation and the signal transduction system involved in virulence.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

This work was supported by NSF Grant MCB-0843662 (R.A.B). We thank Dr. Daniel Portnoy at University of California, Berkeley for kindly providing us the genomic DNA used to clone the L. monocytogenes LOV domain. We are grateful to Dr. Winslow Briggs for very helpful suggestions on the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Avila-Perez, M., K. J. Hellingwerf and R. Kort (2006) Blue light activates the sigma(B)-dependent stress response of Bacillus subtilis via YtvA. J. Bacteriol. 188, 64116414.
  • 2
    Wiedmann, M., T. J. Arvik, R. J. Hurley and K. J. Boor (1998) General stress transcription factor sigma(B) and its role in acid tolerance and virulence of Listeria monocytogenes. J. Bacteriol. 180, 36503656.
  • 3
    Taylor, B. L. and I. B. Zhulin (1999) PAS domains: Internal sensors of oxygen, redox potential, and light. Microbiol. Mol. Biol. Rev. 63, 479506.
  • 4
    Christie, J. M., P. Reymond, G. K. Powell, P. Bernasconi, A. A. Raibekas, E. Liscum and W. R. Briggs (1998) Arabidopsis NPH1: A flavoprotein with the properties of a photoreceptor for phototropism. Science 282, 16981701.
  • 5
    Christie, J. M., M. Salomon, K. Nozue, M. Wada and W. R. Briggs (1999) LOV (light, oxygen, or voltage) domains of the blue-light photoreceptor phototropin (nph1): Binding sites for the chromophore flavin mononucleotide. Proc. Natl. Acad. Sci. U. S. A. 96, 87798783.
  • 6
    He, Q., P. Cheng, Y. Yang, L. Wang, K. Gardner and Y. Liu (2002) White collar-1, a DNA binding transcription factor and a light sensor. Science 297, 840843.
  • 7
    Schwerdtfeger, C. and H. Linden (2003) VIVID is a flavoprotein and serves as a fungal blue light photoreceptor for photoadaptation. EMBO J. 22, 48464855.
  • 8
    Crosson, S., S. Rajagopal and K. Moffat (2003) The LOV domain family: Photoresponsive signaling modules coupled to diverse output domains. Biochemistry 42, 210.
  • 9
    Swartz, T. E., T. S. Tseng, M. A. Frederickson, G. Paris, D. J. Comerci, G. Rajashekara, J. G. Kim, M. B. Mudgett, G. A. Splitter, R. A. Ugalde, F. A. Goldbaum, W. R. Briggs and R. A. Bogomolni (2007) Blue-light-activated histidine kinases: Two-component sensors in bacteria. Science 317, 10901093.
  • 10
    Purcell, E. B., D. Siegal-Gaskins, D. C. Rawling, A. Fiebig and S. Crosson (2007) A photosensory two-component system regulates bacterial cell attachment. Proc. Natl. Acad. Sci. U. S. A. 104, 1824118246.
  • 11
    Aravind, L. and E. V. Koonin (2000) The STAS domain - a link between anion transporters and antisigma-factor antagonists. Curr. Biol. 10, R53R55.
  • 12
    Freitag, N. E., G. C. Port and M. D. Miner (2009) Listeria monocytogenes - from saprophyte to intracellular pathogen. Nat. Rev. Microbiol. 7, 623628.
  • 13
    Gahan, C. G. M. and C. Hill (2005) Gastrointestinal phase of Listeria monocytogenes infection. J. Appl. Microbiol. 98, 13451353.
  • 14
    Garner, M. R., B. L. Njaa, M. Wiedmann and K. J. Boor (2006) Sigma B contributes to Listeria monocytogenes gastrointestinal infection but not to systemic spread in the guinea pig infection model. Infect. Immun. 74, 876886.
  • 15
    Kim, H., K. J. Boor and H. Marquis (2004) Listeria monocytogenes sigmaB contributes to invasion of human intestinal epithelial cells. Infect. Immun. 72, 73747378.
  • 16
    Ondrusch, N. and J. Kreft. (2011) Blue and red light modulates SigB-dependent gene transcription, swimming motility and invasiveness in Listeria monocytogenes. PLoS ONE 6, 116.
  • 17
    Crosson, S. and K. Moffat (2001) Structure of a flavin-binding plant photoreceptor domain: Insights into light-mediated signal transduction. Proc. Natl. Acad. Sci. U. S. A. 98, 29953000.
  • 18
    Fedorov, R., I. Schlichting, E. Hartmann, T. Domratcheva, M. Fuhrmann and P. Hegemann (2003) Crystal structures and molecular mechanism of a light-induced signaling switch: The Phot-LOV1 domain from Chlamydomonas reinhardtii. Biophys. J . 84, 24742482.
  • 19
    Moglich, A. and K. Moffat (2007) Structural basis for light-dependent signaling in the dimeric LOV domain of the photosensor YtvA. J. Mol. Biol. 373, 112126.
  • 20
    Zoltowski, B. D., C. Schwerdtfeger, J. Widom, J. J. Loros, A. M. Bilwes, J. C. Dunlap and B. R. Crane (2007) Conformational switching in the fungal light sensor vivid. Science 316, 10541057.
  • 21
    Losi, A., E. Polverini, B. Quest and W. Gartner (2002) First evidence for phototropin-related blue-light receptors in prokaryotes. Biophys. J . 82, 26272634.
  • 22
    Caetano-anolles, G., D. K. Cristestes and W. D. Bauer (1988) Chemotaxis of Rhizobium-meliloti to the plant flavone luteolin requires functional nodulation genes. J. Bacteriol. 170, 31643169.
  • 23
    Crosson, S. and K. Moffat (2002) Photoexcitd structure of a plant photoreceptor domain reveals a light-driven molecular switch. Plant Cell 14, 10671075.
  • 24
    Salomon, M., J. Christie, E. Knieb, U. Lempert, U. and W. Briggs (2000) Photochemical and mutational analysis of the FMN-binding domains of the plant blue light receptor, phototropin. Biochemistry 39, 94019410.
  • 25
    Salomon, M., W. Eisenreich, H. Durr, E. Schleicher, E. Knieb, V. Massey, W. Rudiger, F. Muller, A. Bacher and G. Richter (2001) An optomechanical transducer in the blue light receptor phototropin from Avena sativa. Proc. Natl. Acad. Sci. U. S. A. 98, 1235712361.
  • 26
    Swartz, T., S. Corchnoy, J. Christie, J. W. Lewis, I. Szundi, W. Briggs and R. Bogomolni (2001) The photocycle of a flavin-binding domain of the blue light photoreceptor phototropin. J. Biol. Chem. 276, 3649336500.
  • 27
    Losi, A., B. Quest and W. Gartner (2003) Listening to the blue: The time-resolved thermodynamics of the bacterial blue-light receptor YtvA and its isolated LOV domain. Photochem. Photobiol. Sci. 2, 759766.
  • 28
    Froehlich, A. C., Y. Liu, J. J. Loros and J. C. Dunlap (2002) White Collar-1, a circadian blue light photoreceptor, binding to the frequency promoter. Science 297, 815819.
  • 29
    Idnurm, A. and S. Crosson (2009) The photobiology of microbial pathogenesis. PLoS Pathog. 5, e1000470. DOI: 10. 1371/journal. ppat. 1000470
  • 30
    Lewis, J. W. and D. S. Kliger (1993) Microliter flow cell for measurement of irreversible optical absorbency transients. Rev. Sci. Instrum. 64, 28282833.
  • 31
    Lewis, J. W. and D. S. Kliger (2000) Absorption spectroscopy in studies of visual pigments: Spectral and kinetic characterization of intermediates. Methods Enzymol.. 315, 164178.
  • 32
    Kliger, D. S., J. W. Lewis and C. E. Randall (1990) Polarized Light in Optics and Spectroscopy, Academic Press, Boston.
  • 33
    Szundi, I., J. W. Lewis and D. S. Kliger (1997) Deriving reaction mechanisms from kinetic spectroscopy. Application to late rhodopsin intermediates. Biophys. J . 73, 688702.
  • 34
    Shrager, R. I. and R. W. Hendler (1982) Titration of individual components in a mixture with resolution of difference spectra, pKs, and redox transitions. Anal. Chem. 54, 11471152.
  • 35
    Thorgeirsson, T. E., J. W. Lewis, S. E. Wallace-Williams and D. S. Kliger (1992) Photolysis of rhodopsin results in deprotonation of its retinal Schiff's base prior to formation of metarhodopsin II. Photochem. Photobiol. 56, 11351144.
  • 36
    Georgiadis, K. E., N. I. Jhon and O. Einarsdottir (1994) Time-resolved optical-absorption studies of intramolecular electron-transfer in cytochrome-c-oxidase. Biochemistry 33, 92459256.
  • 37
    Losi, A. (2004) The bacterial counterparts of plant phototropins. Photochem. Photobiol. Sci. 3, 566574.
  • 38
    Lakowicz, J. R. (1999) Principles of Fluorescence Spectroscopy, 2nd ed., Kluwer Academic/Plenum Publishers, New York.
  • 39
    Buttani, V., A. Losi, E. Polverini and W. Gartner (2006) Blue news: NTP binding properties of the blue-light sensitive YtvA protein from Bacillus subtilis. FEBS Lett. 580, 38183822.
  • 40
    Kaukinen, U., T. Venalainen, H. Lonnberg and M. Perakyla (2003) The base sequence dependent flexibility of linear single-stranded oligoribonucleotides correlates with the reactivity of the phosphodiester bond. Org. Biomol. Chem. 1, 24392447.
  • 41
    Kennis, J., S. Crosson, M. Gauden, I. van Stokkum, K. Moffat and R. van Grondelle (2003) Primary reactions of the LOV2 domain of phototropin, a plant blue-light photoreceptor. Biochemistry 42, 33853392.
  • 42
    Alexandre, M., J. Arents, R. Van Grondelle, K. Hellingwerf and J. Kennis (2007) A base-catalyzed mechanism for dark state recovery in the Avena sativa phototropin-1 LOV2 domain. Biochemistry 46, 31293137.
  • 43
    Song, P. S., T. A. Moore and W. E. Kurtin (1972) Molecular luminescence studies of flavins. 2. Interactions involving excited-states. Z. Naturforsch., B: Chem. Sci. B 27, 10111015.
  • 44
    Sun, M., P. S. Song and T. A. Moore (1972) Molecular luminescence studies of flavins. 1. Excited-states of flavins. J. Am. Chem. Soc. 94, 17301740.
  • 45
    Stanley, R. J. and H. Jang (1999) Electronic structure measurements of oxidized flavins and flavin complexes using stark-effect spectroscopy. J. Phys. Chem. A 103, 89768984.
  • 46
    Kottke, T., J. Heberle, D. Hehn, B. Dick and P. Hegemann (2002) Phot-LOV1: Photocycle of a blue-light receptor domain from the green alga Chlamydomonas reinhardtii. Biophys. J . 84, 11921201.
  • 47
    Tseng, T. S., M. A. Frederickson, W. R. Briggs and R. A. Bogomolni (2010) Light-activated bacterial LOV-domain histidine kinases. Methods Enzymol. 471, 125134.
  • 48
    Holzer, W., A. Penzkofer, M. Fuhrmann and P. Hegemann (2002) Spectroscopic characterization of flavin mononucleotide bound to the LOV1 domain of Phot1 from Chlamydomonas reinhardtii. Photochem. Photobiol. 75, 479487.
  • 49
    Corchnoy, S. B., T. E. Swartz, J. W. Lewis, I. Szundi, W. R. Briggs and R. A. Bogomolni (2003) Intramolecular proton transfers and structural changes during the photocycle of the LOV2 domain of phototropin 1. J. Biol. Chem. 278, 724731.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
php12004-sup-0001-FigS1.pdfapplication/PDF31KFigure S1. SDS-PAGE gel of the LM-LOV-STAS eluted with 150 mm imidazole (Lane 7). It contains mostly monomer and a small amount of dimer. Lanes 1 and 2 are prestained protein molecular weight markers—SeeBlue Plus2 Pre-Stained Standard (Invitrogen) and ColorBurst Electrophoresis Marker (Sigma Aldrich), respectively. Lane 3 is the lysate after French-pressing E. coli. Lane 4 is the column flow through. Lanes 5 and 6 are the equilibration buffer wash and 20 mm imidazole wash, respectively.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.