{"paper_id":"0e8f0e6b-84ae-46eb-acd7-9f61b1126a71","body_text":"Structural Basis of a Novel Heme Binding Bacterial One-Component \nSwitch \n \nJames J. Siclari1,2, Malvin Forson1,3, Cullen Roeder1,3, Eta A. Isiorho1, Denize C. \nFavaro1, Rinat R. Abzalimov1, Stephen S. Gisselbrecht4, Alec H. Follmer5, Martha L. \nBulyk4,6, Kevin H. Gardner1,7,8,* \n \n*: direct correspondence to kgardner@gc.cuny.edu \n \n1: Structural Biology Initiative, CUNY Advanced Science Research Center, New York, NY 10031 \n2: Ph.D. Program in Biology, The Graduate Center – City University of New York, New York, NY \n10016 \n3: Ph.D. Program in Biochemistry, The Graduate Center – City University of New York, New \nYork, NY 10016 \n4: Division of Genetics, Department of Medicine, Brigham and Women’s Hospital and Harvard \nMedical School, Boston, MA 02115 \n5: Department of Chemistry, University of California -Davis, Davis, CA 95616, USA \n6: Department of Pathology, Brigham and Women’s Hospital and Harvard Medical School, \nBoston, MA 02115 \n7: Ph.D. Programs in Biochemistry, Biology, and Chemistry, The Graduate Center – City \nUniversity of New York, New York, NY 10016 \n8: Department of Chemistry and Biochemistry, City College of New York, New York, NY  10031 \n \nORCIDs: \nJames J. Siclari: 0000-0001-5275-4515 / Malvin Forson: 0000-0001-8810-4039 / Cullen Roeder: \n0009-0007-8791-2668 / Eta A. Isiorho: 0000-0002-6242-9297 / Denize C. Favaro: 0000-0001-\n5563-2548 / Rinat R. Abzalimov: 0000-0003-2110-1532 / Stephen S. Gisselbrecht: 0000-0001-\n8723-902X / Alec H. Follmer: 0000-0002-6244-6804 / Martha L. Bulyk: 0000-0002-3456-4555 / \nKevin H. Gardner: 0000-0002-8671-2556 \n \nKeywords:  Per-ARNT-Sim domains, Bacterial Signaling, Heme Binding Proteins, One-\nComponent Systems \n \nRunning title: FG214, A Novel Heme-Regulated DNA-binding Protein \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n2\nAbstract \nOne-component systems (OCSs) integrate sensory and effector functions within \na single protein, enabling rapid gene expression changes in response to environmental \ncues. Here, we characterized a novel putative OCS protein, FG214, from Fimbriimonas \nginsengisoli, which drew our attention as a potential redox or O2-regulated helix-turn-\nhelix (HTH)-Per-ARNT-Sim (PAS) transcription factor. Data supporting this included our \nobservation of the FG214 PAS domain binding a hexacoordinate heme b in oxidized \nconditions and undergoing a slate of redox and ligand-dependent conformational \nchanges, transitioning from a monomer to a homodimer. Spectroscopic and structural \ndata revealed that oxidation stabilizes the likely HTH-PAS intramolecular domain \ninterface, while reduction of the heme iron dissociates the HTH, freeing previously-\nsequestered homodimerization surfaces. Similar effects were seen by addition of a \nsmall molecule ferric heme ligand, as directly visualized with a 1.47 Å crystal structure \nof an imidazole-bound truncated construct. Using in vitro DNA-binding assays, we \nidentified an artificial promoter sequence and demonstrated ligand-enhanced protein-\nDNA binding. Finally, we performed proof of concept experiments exploring the ability of \nFG214 to homodimerize in vivo, setting the stage for a redox or gas sensitive biosensor. \nTogether, these findings define FG214 as a novel heme-binding PAS DNA binding \nprotein and potential transcription factor, complementing known heme-PAS two-\ncomponent signaling switches.  \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n3\nSignificance Statement \nStudying sensory protein structure-function relationships is paramount to fully \nunderstanding cellular adaptation to environmental cues. Here, we have discovered a \nheme-regulated DNA binding switch, FG214, that undergoes a monomeric to dimeric \ntransition upon redox-triggered changes. The nature of this transition is reminiscent of \nother proteins with similar domain architecture, although those have been observed to \ninstead sense changes in the presence of other distinct stimuli. Here we not only \ndescribe the activation mechanisms of FG214, but also provide support for its ability to \nbe used as a regulatory gene expression tool. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n4\nIntroduction \nBacteria rely on an array of sensory and signal transduction proteins to detect \nand respond to environmental fluctuations. Many such responses are mediated by one-\ncomponent systems (OCS), in which the sensory and effector domains reside within a \nsingle polypeptide(1). Among these proteins, Per-ARNT-SIM (PAS) domains are often \nfound playing sensory roles given their remarkable adaptability, binding diverse small \nmolecules to allosterically regulate partner domains differentially (1-4). \n \nA subset of bacterial PAS domains coordinate heme prosthetic groups, enabling \ndirect sensing of gases, redox potential, and metabolic state(5). Much of our \nunderstanding of this regulation stems from one such type of system, the Bradyrhizobia \nFixL histidine kinase-FixJ response regulator pair, where changes in O\n2 occupancy at \nthe distal position alters net kinase activity of FixL, along with subsequent FixJ \nphosphorylation and transcriptional activation. However, questions remain open about \nboth FixL signaling – particularly how signals are transduced from the hemes to activate \nthis enzyme – and more broadly about their applicability to other heme-PAS proteins(6-\n9). \n \nBeyond their physiological roles, PAS-containing OCS proteins offer powerful \nplatforms for biosensing and synthetic regulation due to their compact architecture and \nmodular signaling logic. The blue-light responsive activator EL222, for example, \nundergoes a conformational switch upon illumination that releases its helix-turn-helix \n(HTH) effector, activating transcription(10-13). Previous work from our lab and others \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n5\nhave engineered this system for uses in prokaryotic and eukaryotic model systems (14-\n20). Similar sensory logic could, in principle, be harnessed for small molecule or redox \ndriven systems. \n \nHere we describe FG214, a HTH-PAS transcription factor from Fimbriimonas \nginsengisoli(21), that binds a heme b within its PAS domain using two protein histidine \nsidechains to an oxidized Fe(III) as purified in vitro. We show that redox changes or \nexogenous imidazole induce structural rearrangements which convert FG214 from a \nmonomer into a homodimer, enabling DNA binding. Our work identifies a novel one-\ncomponent heme binding PAS protein and establishes a mechanistic framework for \nredox-regulated “effector release” while also providing proof of concept experiments for \nits utility as an in vivo biosensor. These findings expand the functional landscape of PAS \ndomains and suggest new strategies for engineering redox-sensitive systems. \n \nResults\n \nRedox-dependent global conformational changes \nFG214 was identified in a bioinformatics screen for PAS domain-containing \ntranscription factors with potential chemosensory triggers(22). These analyses predicted \nan N-terminal LuxR-type DNA-binding tetrahelical helix-turn-helix (HTH) DNA-binding \ndomain and a C-terminal PAS domain, connected by a predicted helical linker. While \nthese two types of domains are commonly found together (over 6000 occurrences in \nSMART(23) as of Feb 2026), the vast majority of these have a domain architecture with \nan N-terminal PAS sensor and C-terminal LuxR HTH DNA-binding domain, as \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n6\npreviously seen in EL222(10) (Fig. 1a,b). Following overexpression and purification \nfrom E. coli, FG214 was visibly red, consistent with metal binding. Subsequent UV-\nvisible absorbance spectroscopy and LC-MS revealed a heme b ligand(24) (Fig. \n1c,d,e), and redox titrations demonstrated reversible spectral changes consistent with a \nredox-active heme(25-27) (Fig. 1d). \n \nReduction of the heme iron induced widespread structural changes detected by \n1H and 15N/1H TROSY NMR (Fig. 1f,g). In the oxidized state, the spectra exhibited well-\ndispersed resonances of uniform intensity, suggesting a folded protein. We observed 1H \npeaks of oxidized FG214 distinctively upfield of 0 ppm, likely arising from protons \ninteracting with a low-spin ferric Fe(III) state in the heme(28). These peaks were greatly \nperturbed by reduction, typical of a paramagnetic Fe(III) to diamagnetic Fe(II) spin \nsystem transition(29).  \n \nMore globally, reduction caused extensive chemical shift perturbations and peak \nbroadening in FG214 \n15N/1H TROSY spectra, suggestive of a large-scale protein \nconformational change. While we lack the chemical shift assignments needed to fully \nestablish the nature of this event using solely solution NMR, we can establish two key \nfeatures by comparing spectra acquired from a series of FG214 truncations (1-214, 73-\n214 [Δ72], 87-214 [Δ86]) (Fig. S1). Several peaks in this series that are present in all \nthree spectra – and hence must arise from residues in the PAS domain or an N-terminal \n25 residue section – show a progressive linear change in chemical shift, strongly \nsuggesting that the truncations alter an equilibrium. Coupled with AlphaFold3 models of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n7\nthe variants, this pattern was highly reminiscent of the light-triggered helix release \n \nFigure 1: FG214 is a redox-sensitive heme-binding protein. a) Domain architecture of \nFG214 and another PAS-containing transcription factor, EL222. b) AlphaFold3 model of \nFG214, highlighting locations of HTH and PAS domains, as well as the connecting 4α helix. \nc) Superdex S200 SEC chromatogram of FG214 compared to standards of known molecular \nweight. (inset) SDS-PAGE gel verifying expected MW (24 kDa predicted from sequence) and \nimage of purified protein. d) UV-visible absorbance spectra of oxidized (black) and reduced \n(red) FG214. e) Electrospray ionization LC-MS analysis confirming the presence of heme b \nwithin FG214 isolated from E. coli. f) 1H NMR spectra of oxidized (black) and reduced (red) \nFG214. g) 15N/1H TROSY spectra of oxidized (black) and reduced (red) FG214. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n8\nequilibrium seen in the photosensitive AsLOV2 system with an FMN-bound PAS \ndomain(30, 31). These same peaks exhibited reduction-triggered chemical shifts as \nwell, strongly suggesting that a heme-based mechanism also alters this equilibrium in \nvitro (Fig. S1). Taken together, our solution NMR data indicate substantial redox-\ntriggered changes in FG214 structure suggestive of a triggered release of N-terminal \nregions before the PAS domain. \n \nReduction promotes release of the HTH domain\n \nHydrogen-deuterium exchange mass spectrometry (HDX-MS) was used to \nelucidate how heme reduction shifts the conformational landscape of FG214. HDX-MS \nprotection patterns of both the oxidized and reduced form of the protein are generally \nconsistent with AlphaFold3 predictions, including placements of all 2° structures in both \ndomains. The predicted long 4α helix is also visible in these protection patterns, but \nnotably, it shows a marked increase in deuteration levels upon reduction (Fig. 2a,b,c). \nThis interdomain 4α helix appears to make direct contact with the FG214 PAS domain \nin the oxidized monomer AlphaFold3 model. Our experimental data clearly suggests \nreduction destabilizes the 4α helix from the PAS core, consistent with an allosteric \nrelease of the effector domain. This is a mechanism seen multiple times previously in \nprokaryotic and eukaryotic signaling proteins(31-34). We also see reduction influenced \nchanges in the PAS domain around the known ligand binding regions, mainly in the Eα \nhelix and H\nβ  strand, consistent with a redox-triggered shift in the geometry or volume of \nthe heme binding pocket. These features suggest that the oxidized state adopts a stable \nmonomeric “off” conformation, whereas reduction destabilizes the fold. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n9\n \nHeme coordination chemistry is linked to FG214 dimerization \nThe iron in heme b is typically coordinated by a combination of histidine and/or \nmethionine residues in the oxidized state(35). Identifying the coordinating residues \nwithin FG214 allows for targeted mutagenesis to probe structure-function relationships \nand potentially shift the equilibrium between inactive and active conformations. Guided \nby AlphaFold3 structural predictions (Fig. 3a), we identified four candidate residues (His \n \nFigure 2: HDX-MS indicates that FG214 employs an effector-release mediated mode of \nactivation. a) Peptide deuteration uptake plots of selected regions of FG214 under oxidized \n(black) or reduced (red) conditions. b) Heat map showing percent change in deuteration over \ntime between oxidized and reduced states. Secondary structure diagram shows HTH domain \nin orange, A/i1 α in pink, and PAS domain in cyan. c) Heat map of reduction-triggered \nchanges in HDX protection as data mapped onto FG214 AlphaFold3 model. Red color \nindicates less protection in reduced form; blue indicates more protection in reduced form; \nblack indicates no data due to gaps in peptide coverage. Asterisks denote statistical \nsignificance derived from a two-tailed Welch’s t-test. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n10 \n156, His 159, Met 172, His 175) and sought to experimentally determine their roles in \nheme coordination. \n \nAs an initial examination, we acquired X-band electron paramagnetic resonance \n(EPR) spectra of oxidized FG214, revealing two rhombic S=1/2 populations (Fig 3b). \nBoth species had g-values supporting the assignment of 6-coordinate low spin-ferric \nheme, sometimes referred to as highly anisotropic/axial low-spin (HALS) systems(36). \nThe first, less rhombic species exhibited g-values (3.01, 2.26, 1.38) similar to many \n \nFigure 3: FG214 binds heme with two histidine residues in the oxidized state. a) \nAlphaFold3 model of full length FG214 heme binding pocket. b) X-band EPR spectra of \noxidized FG214 showing two different low spin ferric states. Asterisk indicates a small \nbackground Cu signal. c) 1H NMR spectra of oxidized (black) and reduced (red) wild type \n(bottom, reproduced from Figure 1f) and point mutants of potential liganding residues. d) \nSEC-MALS using a Superdex 200 column of WT (black) and H175I (blue) FG214 proteins. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n11\nother HALS proteins where the heme cofactors are bound by two histidine residues \narranged in an antiparallel geometry (Fig. 3b), consistent with our 1H NMR spectra \ndisplaying upfield-shifted resonances (Fig 1f). The second species is more rhombic and \nonly two g-values are observed in this magnetic field range (g-values: 3.15, 2.2). While \nwe cannot rule out that this second species arises from a second orientation of the bis-\nHis coordination, it is also possible that this more rhombic population indicates a \npotential His-Met coordination in the oxidized state with similar g-values observed for \nthese systems(36). \n \nTo further test this model, we generated a series of single point mutations \ntargeting the four histidine and methionine residues predicted in the binding pocket and \nsubstituting each with isoleucine to maintain hydrophobicity and approximate sidechain \nvolume. We used UV-visible absorbance spectroscopy to compare relative heme \nloading across mutants, using samples with matched protein concentrations (as judged \nby A\n280 values) and, focusing on changes in the Q- and Soret band regions. While \nM172I exhibited near-wildtype ability to bind heme as judged by the Q- and Soret band \npeak intensities, the other three mutants were impaired. This was most evident with the \nH156I mutant, particularly with the near absence of a Soret band in the reduced state \n(Fig. S2, Table S1). The H159I mutant also exhibited a distinct spectral signature, \nincluding a new absorbance maximum at 662 nm and an altered color in solution, \nreflecting perturbed coordination geometry (Fig. S2, Table S1). The persistence of \npartial binding among all mutants suggests compensatory heme coordination, \nunderscoring some degree of pliability in the FG214 heme binding site.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n12 \n \n1H NMR spectra and size exclusion chromatography of the mutated proteins \nfurther revealed structural changes caused by these substitutions. Focusing first on the \nNMR signals upfield of 0 ppm, H156I and H159I displayed substantial peak losses in \nthis region, consistent with global destabilization of the protein (Fig. 3c). M172I and \nH175I were subtler in their changes, but notably with oxidation-specific effects, with \nM172I being most strongly perturbed in the reduced form and H175I in the oxidized \nstate. Coupled with UV-visible absorbance data and the AlphaFold3 prediction, we \nassign His 156 (proximal) and His 175 (distal) as the heme-coordinating residues, with \nsome potential role for Met 172 as an alternative distal participant.  \n \nTo explore the roles of these heme-coordinating residues in the FG214 \nquaternary structure, we used size-exclusion chromatography with inline light scattering \n(SEC-MALS) to characterize changes in solution shape and mass. While the lack of \nheme binding by H156I precluded us from acquiring meaningful SEC-MALS data on this \nvariant, we obtained data clearly showing that wildtype FG214 is monomeric in solution \nand H175I is markedly shifted towards a dimeric species (Fig. 3d). Taken together, our \nsolution data clearly support a model where changes near the heme site affect both an \nintramolecular HTH-PAS interaction predicted in the oxidized state, leading to protein \ndimerization. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n13 \nCrystal structure of the activated dimer \nGuided by truncation analysis (Fig. S1), we successfully crystallized an \nimidazole-bound construct lacking the first N terminal 72 residues (Δ 72), encompassing \npart of the HTH 4α helix through the PAS domain. In solution, this construct shares \nmany features in common with the full-length protein, including being monomeric by \nSEC-MALS when oxidized, having nearly identical UV-visible absorbance \ncharacteristics, and similar 1H NMR shifts – with a notable shift in the location of heme \nvinyl chemical shifts near –8 ppm, showing that changes outside the PAS domain \nimpact the environment near the heme (Fig. S3). During crystallization trials for FG214, \nwe observed that addition of imidazole was required to obtain crystals, which we \nattribute to imidazole’s well-characterized ability to ligand ferric heme proteins. In doing \nso, it is known to bind the distal coordination site of heme, outcompeting native protein \nsidechain interactions, reproducing the electrostatic properties of Fe(II)-O\n2 \ncomplexes(37, 38). \n \nFrom these imidazole-bound FG214(Δ72) crystals, we obtained a 1.47 Å \nstructure of an imidazole-bound homodimer (Fig. 4a), mediated by both HTH-HTH and \nPAS-PAS interactions. Within the PAS domain, H156 serves as the proximal heme-\ncoordinating residue, while an imidazole molecule occupies the distal site, fully \ncorroborating our spectroscopic and mutagenesis data. Meanwhile, the H175 sidechain, \nwhich we expect to normally serve as the distal ligand, was displaced into an \nunstructured loop that meets the other monomer at the dimerization interface. Of note, \nthe nearby M172 sidechain is similarly outside of the PAS domain in one of the two \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n14 \nchains but is not resolved in the other. These structural changes clearly implicate \nlinkage between heme coordination and protein oligomerization state together. \n \nExamining the FG214 dimer, we saw potential interactions between the two \nchains mediated by both PAS/PAS and 4α/4α contacts. At the PAS domains, these were \nchiefly involving A\n/i3 α helix/β-sheet and β-sheet/β-sheet interactions; of note, the β-sheet \nresidues are also predicted by AlphaFold3 to make an intramolecular monomeric \ninterface with 4α (Fig. 4b). At the HTH domains, hydrophobic sidechains of the two 4α \n \nFigure 4: Crystal structure of an N-terminally truncated FG214 protein reveals \nimidazole driven dimerization. a) 1.47 Å crystal structure of FG214(Δ 72). PDBID:10JX. \nMesh shows electron density. Insets show heme binding cavity and hydrophobic interface of \nHTH 4α helix. b) FG214 PAS domain highlighting residues 4 Å away from the 4α helix in the \nAF3 monomer model (left) and 4 Å from the other PAS domain in the dimeric crystal structure \n(right). c) AlphaFold3 model of oxidized full length FG214 highlighting the locations of the 4α  \nresidues involved in the dimeric structure of Figure 4a. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n15 \nhelices interact to form the dimer; reminiscent of coiled-coil interfaces of transcription \nfactors(39-41). And again, the 4α residues at the dimer interface are predicted to play \nan integral role in stabilizing the monomeric form (Fig. 4c). \n \nSoaking these same crystals with sodium dithionite (DT) yielded datasets leading to a \n1.67 Å structure of FG214(Δ72), revealing a homodimer with nearly identical 4° \nstructural arrangements as previously seen. However, there are distinct changes at the \nheme – which has a His 156-Met 172 iron coordination pair and His 175 on an extended \nloop pointing into solution (Fig. S4).This raises the potential that alternative His-His and \nHis-Met heme coordination pairs – as suggested by our EPR and mutagenesis results \nin the full length protein (Fig. 3) – may exist and be biased in different crystal forms of \nFG214 variants. Indeed, AlphaFold3 modeling suggests that removal of the HTH \ndomain favors His-Met coordination in the monomer (Fig. S5), perhaps due to altered \nstates of the outer \nβ -sheets within the PAS core due to the shortened 4α helix. \nRegardless, these structural data strongly implicate changes at the heme site being \namplified by the surrounding protein to affect the monomer:dimer equilibrium of FG214, \nas is commonly seen as a regulatory mechanism in other proteins. \n \nIdentification of an artificial DNA binding site \n \nTo explore the functional importance of these changes, we aimed to identify a \nsuitable artificial DNA sequence for FG214 binding, both to enable tests of our \nhypothesis of imidazole-driven activation and inform downstream functional assays and \nfuture bioengineering. To do so, we used universal protein binding microarrays (PBMs) \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n16 \nto screen FG214 for in vitro sequence-specific DNA binding activity against all possible \n8 bp sequences(42, 43). Analysis of PBM-derived scores from these experiments of \nthese data yielded a DNA binding specificity motif for FG214 (Fig. 5a), with a notable \nbias towards a purine-rich strand on one side of an 8 bp sequence. \n \n \nFigure 5: FG214 binds a GC-rich palindromic promoter region enhanced by imidazole \nin the ferric state. a) DNA half-site for FG214 binding identified by universal protein-binding \nmicroarray (PBM). b) Median intensity of specifically-bound 8-mers in PBM for FG214 WT \nand H175I in the absence and presence of 500 mM imidazole. c) Fluorescence polarization \nmeasurements of 1 nM FAM-labeled DR2 or IR2 sequences with increasing concentrations \nof FG214 in the absence and presence of 10 mM imidazole. All individual measurements \nwere collected in triplicate; points shown are average ± 1 standard deviation. d) \nβ -\ngalactosidase activity quantified following bacterial two hybrid (BTH) analysis of split adenylyl \ncyclase vectors (NC) or fused to leucine zippers (PC), FG214 (Δ 72), FG214 full-length, or \nFG214 full-length H175I. Statistical significance derived from a two-sample t-test, with *** = \n0.0001 < p < 0.001 and * = 0.01 < p < 0.05. Individual datapoints are shown, with top bar \ndrawn at mean value and error bars showing ± one standard deviation from the mean. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n17 \nGiven the role of the distal His 175 in maintaining the FG214 monomer, we also \ntested whether ligand-induced displacement of this sidechain in a wild-type background \ncould promote DNA binding. Analysis of PBM fluorescence intensities across the array \nrevealed enhanced binding by the H175I mutant relative to wild-type FG214, with further \nincreases for both constructs in the presence of 500 mM imidazole (Fig. 5b). These \nobservations suggest that imidazole-induced displacement of His 175 facilitates domain \nrearrangements sufficient for dimer-mediated DNA binding, thereby partially activating \nFG214 under ferric conditions. These findings align with prior studies showing that \nexogenous imidazole can act as a functional mimic for in vivo heme ligands(44-46). In \nthis context, imidazole binding appears to trigger an active conformation of FG214 \nwithout the need for heme reduction. \n \nTo develop an optimal high-affinity DNA binding site for in vitro assays, we \nconstructed a series of DNA sequences containing direct or inverted repeats of the \nPBM-identified motif (GGGGCGGGG), assuming that this sequence is a likely half-site \nfor binding a FG214 dimer. We generated 14 FAM labeled binding substrates with \nvarying half-site orientation and spacer length from 1-7 bp (Fig. S6). Each DNA \nconstruct was labeled with FAM (fluorescein derivative) on the 5\n/i3  end to enable \nquantitative binding analysis via fluorescence polarization/anisotropy. Oligos containing \ndirect repeats of the motif bound FG214 weakly and non-specifically, and without any \ndependence on mutation or imidazole concentration (Fig. S6). In contrast, oligos with \ninverted repeats of the motif displayed stronger binding, particularly for the H175I \nvariant, and often with imidazole dependence (Fig. S6). From these experiments, a \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n18 \nparticularly clear exemplar of FG214 DNA binding characteristics are evident from \noligos with a 2 bp central spacer: Inverted copies of this motif (IR2; Table S3) exhibited \nthe most pronounced imidazole-dependent binding response (without imidazole: Kd ≈  \n950 nM; with 10 mM imidazole, affinity increased by approximately an order of \nmagnitude to an apparent K\nd ≈  100 nM) (Fig. 5c). In contrast, direct repeats around the \nsame 2 bp spacer (DR2; Table S3) showed moderate binding of Kd ≈  600 nM with no \nimidazole dependence. \n \nIn Vivo Validation of FG214 homodimerization \nTo explore the potential for FG214 to adopt the homodimeric state that appears \nto be required for DNA binding, we employed a bacterial two hybrid (BTH) assay to \nprobe FG214 homodimerization in E. coli. By fusing FG214 (full length, mutant, or the \nΔ 72 truncation) to fragments of a split adenylyl cyclase we can analyze FG214 \nhomodimerization following co-transforming into E. coli lacking its native adenylyl \ncyclase(47, 48). Here, complementation of the split cyclase will only occur with sufficient \nFG214 homodimerization, facilitating cyclic AMP (cAMP) production from ATP and \nactivating CAP-dependent expression of a \nβ -galactosidase which can be easily assayed \non plates or in solution as a proxy for dimerization. Employing this BTH assay by \nquantitatively measuring β -galactosidase activity in cell lysates in the presence of ortho-\nnitrophenyl-β -D-galactopyranoside (ONPG), we saw significant increases in β -\ngalactosidase activity in FG214(Δ 72) and the H175I full length fusions compared to the \nnegative control (NC), confirming of FG214 homodimerization in E. coli grown under \nstandard aerobic conditions (Fig. 5d). No significant homodimerization was seen for full-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n19 \nlength FG214, indicating that these growth conditions are insufficient on their own to \ntrigger homodimerization. Taken together, these cell-based data are all consistent with \nour in vitro biochemical and structural results. \n \nDiscussion \nOur work provides structural and mechanistic insight into a heme-binding PAS-\nHTH one-component DNA-binding protein, FG214. By characterizing the monomeric \ninactive oxidized state, the intermediate reduced state, and the active DNA-bound \nhomodimer triggered by imidazole, we define three likely signaling states of the protein \n(Fig. 6). Together, these data support a model in which FG214 employs an effector-\nrelease activation mechanism to sense and respond to environmental gases, a \nframework previously observed in other PAS-containing signaling systems. In addition \nto expanding our knowledge of bacterial signaling proteins, we have laid the foundation \nfor the development of FG214-based transcriptional reporter systems. \n \nEffector-release mechanism\n \nStructural and dynamic analyses of the oxidized FG214 monomer validate the \nAlphaFold prediction for the HTH 4α helix to form key contacts with the PAS domain β -\nsheet, particularly via HDX-MS analysis showing this helix to be well ordered. This \nconformation masks HTH and PAS surfaces required for dimerization, as demonstrated \nby our solution measurements. Upon Fe(III) reduction or ligand binding, however, this \nhelix is markedly destabilized, suggesting that helix unwinding facilitates dimerization \nand activation. Indeed, our crystal structures of reduced and imidazole-bound \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n20 \nFG214(Δ78) confirm that the newly-freed PAS β-sheet and HTH 4α helix contribute \nsubstantial homodimer interfaces, supporting a model in which allosteric changes \ninitiated at the heme cofactor reorganize nearby FG214 sheet/helix interactions as a key \ntrigger to transition from a monomeric “off” state to a dimeric DNA-binding “on” state. \n \nThis activation mechanism parallels those used by certain other PAS sensors, \nincluding the light-activated AsLOV2 and EL222 from the LOV subfamily(10, 12, 30) \n(Fig. 6). In those cases, allosteric signals triggered by illumination or redox changes at \nan internal flavin cofactor relay across the LOV β-sheet, displacing the Jα helix \n(AsLOV2 (30, 31)) or the 4α helix (EL222 (10)) as part of the activation process. We \n \nFigure 6: Proposed signaling model for FG214, with a signal-induced monomer to \ndimer transition activating DNA binding, like other PAS domain sensors. a) Proposed \nschematic of FG214 activation. b) EL222 and AsLOV2, PAS domain containing proteins that \nrely on interactions between auxiliary helices and β -sheet surfaces to maintain inactive \nstates. Note that auxiliary helices can originate either N- or C-terminal of the PAS domain \nand run in either direction across PAS β-sheet surface. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n21 \nunderscore that the helices involved in these stimuli-dependent binding/unbinding \nequilibria bind to equivalent spots on the PAS β-sheet, despite coming from very \ndifferent locations in the primary sequence compared to the PAS domain itself – from \nimmediately C-terminal (AsLOV2 Jα) to 70-80 residues C-terminal (EL222 4α) to 20-40 \nresidues N-terminal (FG214 4α) – and binding in different orientations (Fig. 6b). This \nsuggests evolutionary conservation of a common signaling mechanism despite marked \ndifferences as substantial as the order of domains(49, 50). Upon freeing the 4α helix, \nthe FG214 β-sheet becomes accessible for homodimerization, utilizing the same \nresidues just liberated to do so (Fig. 4b,c). This is very reminiscent of the proposed \nsignaling mechanism for EL222(10, 12, 13), despite differences in the controlling \ncofactor (heme b vs. FMN) and domain orientation (HTH-PAS vs. PAS-HTH). \n \nDNA binding specificity and implications for dimerization\n \nOur DNA-binding data across constructs with varying spacer lengths provide key \ninformation into the monomer-dimer equilibrium of FG214 as well as functional data of \nthe active state. We interpret the binding to direct-repeat sequences – relatively weak, \nrelatively low change in fluorescence polarization, and chiefly unaffected by changes in \nspacer, addition of imidazole or the H175I mutation – likely reflects low-affinity \nrecruitment of monomeric FG214 to individual half-sites. In contrast, inverted repeats \ndisplayed higher affinity, more stimulus dependence, and more variation by spacer \nlength. Coupled with the higher \nΔ mP for the saturated inverted repeats when compared \nto the direct, these data are consistent with engagement by an activated dimeric form of \nFG214. Given that GC rich regions are implicated in transcriptional control of oxidative \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n22 \nstress genes(51-53), which appear throughout the native Fimbriimonas ginsengisoli \ngenomes, we speculate that FG214 may have an in vivo role controlling the expression \nof these genes, but this remains to be experimentally validated. While such cellular \nexperiments are outside the scope of this work, we believe that the correlation we see \nbetween in vitro and E. coli dimerization of FG214 truncation or the H175I mutation – a \nstep required for DNA binding – provides both useful reagents and frameworks for those \nstudies. \n \nComparison with other heme-binding PAS sensors\n \nHeme-binding PAS domains have been previously reported in other proteins, but \nnever in an OCS context. Of these, several gas-sensing bacterial PAS proteins have \nbeen best studied, including FixL, Aer2, and EcDos/DosP(54-57). Other proteins, like \nthe FlrB histidine kinase, use the PAS fold to detect levels of heme directly as opposed \nto using heme as a prosthetic group to sense secondary ligands, as bacteria often \nsequester heme from symbionts or host organisms as an iron source(58-60). \n \nAmong known heme-binding PAS proteins, the oxygen-sensing kinase FixL \nremains the best characterized and is often regarded as the canonical model. In \nBradyrhizobium japonicum, FixL forms part of a two-component system in which the \nheme-binding PAS domain modulates kinase activity in response to oxygen levels. In its \noxidized state, FixL binds heme in a pentacoordinate geometry through a single \nproximal histidine(46, 61, 62). Researchers also used ligands like imidazole or cyanide \nto mimic O\n2 binding under oxidized conditions in vitro(54, 63). Such binding events \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n23 \nreshape a preformed PAS homodimer within FixL(64), leading to conformational \nchanges within a coiled-coil linker to a downstream histidine kinase, controlling \nenzymatic activity as it does. \n \nIn contrast, FG214 coordinates heme via a chiefly bis-histidine ligation in the \nferric state and undergoes distal coordination rearrangement upon activation. \nSubsequently triggered protein conformational changes are also distinctly different, as \nFG214 uses heme sensing to control a monomer:dimer equilibrium for activation. These \ndifferent coordination chemistries and signaling mechanism illustrate how PAS domains \nhave evolved diverse strategies to couple similar environmental cues and relay them \ninto downstream signaling. The comparison underscores the remarkable modularity and \nadaptability of PAS domains across species and regulatory contexts.  \n \nImplications for redox signaling and synthetic applications\n \nBy uncovering the molecular basis of heme-based sensing in FG214, we have \nrevealed how structural modularity of PAS domains within one-component systems \nenables evolutionary diversification of signaling logic to respond to different signals with \ndifferent domain orientations while retaining some mechanistic similarity with effector \nrelease. We suggest that FG214 thus represents a new prototype for heme-regulated \ntranscriptional switches, expanding the known repertoire of OCS PAS signaling \nmechanisms beyond light and ligand-sensing systems, and laying the foundation for \nnovel engineered redox- or gas-sensitive proteins. Such systems could be valuable by \nboth informing aspects of natural biological signaling and in a range of biotech or \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n24 \nenvironmental applications where control of biological processes by redox or heme-\nbinding ligands may be useful. \n \nMaterials and Methods\n \nProteins were expressed in E. coli BL21(DE3) cells and purified through nickel-\naffinity chromatography and size exclusion chromatography. UV-visible absorbance \nspectra were collected on a Varian Cary 60 spectrophotometer. NMR data were \ncollected using a Bruker Avance III HD 800 MHz (18.8 T) spectrometer with a 5 mm TCI \nCryoProbe at 298K. Crystallographic data were collected at National Synchrotron Light \nSource II (NSLS-II) light at Brookhaven National Laboratory on beamline 19-ID (NYX) \nand processed using the autoPROC toolbox(65). DNA binding was assessed by \nfluorescence polarization using FAM-labeled dsDNA collected with a Spectramax I3 \nequipped with a fluorescence polarization cartridge (Molecular Devices) using a \nsequence identified from a protein binding microarray(42). Detailed descriptions of all \nmethods are available in SI Appendix. \n \nData Availability\n \nThe X-ray structure coordinates for truncated FG214 with and without Na2S2O4 \nare available from the Protein Data Bank under accession codes 10JY and 10JX \nrespectively. PBM data are deposited in the NIH Gene Expression Omnibus (GEO) \nunder accession GSE319048. All other data are available upon request to K.H.G. on \nbehalf of the authors. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n25 \nAcknowledgements \nWe thank members of the Gardner lab as well as Prof. Elizabeth Boon and Jason \nWithorn (Stony Brook University) for helpful discussions. We thank Prof. Anum Glasgow \n(Columbia University) and members of her lab assistance running PIGEON and PFNet \nfor HDX-MS peptide list curation and validation, as well as Dr. Paul H. Oyala (Caltech) \nfor EPR access and advice. This work was supported by grants from the NIH (R01 \nGM106239 and R35 GM156296 to K.H.G.; UC Davis start-up funds for A.H.F.). Use of \nthe NYX beamline (19-ID) at the National Synchrotron Light Source II (NSLS II) is \nsupported by the New York Structural Biology Center. NSLS II is a U.S. Department of \nEnergy (DOE) Office of Science User Facility operated for the DOE Office of Science by \nBrookhaven National Laboratory under contract DE-SC0012704. This manuscript is the \nresult of funding in whole or in part by the National Institutes of Health (NIH). It is \nsubject to the NIH Public Access Policy. Through acceptance of this federal funding, \nNIH has been given a right to make this manuscript publicly available in PubMed \nCentral upon the Official Date of Publication, as defined by NIH. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n26 \nReferences \n1. L. E. Ulrich, E. V. Koonin, I. B. Zhulin, One-Component Systems Dominate Signal \nTransduction in Prokaryotes. Trends in Microbiology 13, 52-56 (2005). \n2. J. T. Henry, S. Crosson, Ligand-Binding PAS Domains in a Genomic, Cellular, and \nStructural Context. Annual Review of Microbiology 65, 261-286 (2011). \n3. B. L. Taylor, I. B. Zhulin, PAS Domains: Internal Sensors of Oxygen, Redox Potential, \nand Light. Microbiology and Molecular Biology Reviews 63, 479-506 (1999). \n4. E. C. Stuffle, M. S. Johnson, K. J. Wa tts, PAS Domains in Bacterial Signal Transduction. \nCurrent Opinion in Microbiology 61, 8-15 (2021). \n5. M.-A. Gilles-Gonzalez, G. Gonzalez , Signal Transduction by Heme-Containing PAS-\nDomain Proteins. Journal of Applied Physiology 96, 774-783 (2004). \n6. A. E. Gallio  et al., An Extended Network for Regulation of Heme Homeostasis in Cells. \nProceedings of the National Academy of Sciences 122 (2025). \n7. A. Farhana, V. Saini, A. Kumar, J. R. Lancaster, A. J. C. Steyn, Environmental Heme-\nBased Sensor Proteins: Implications for Understanding Bacterial Pathogenesis. \nAntioxidants & Redox Signaling 17, 1232-1245 (2012). \n8. C. Y. Taabazuing, J. A. Hangasky, M. J. Knapp, Oxygen Sensing Strategies in Mammals \nand Bacteria. Journal of Inorganic Biochemistry 133, 63-72 (2014). \n9. N. Chim, P. M. Johnson, C. W. Goulding, Insights into Redox Sensing Metalloproteins in \nMycobacterium tuberculosis. Journal of Inorganic Biochemistry 133, 118-126 (2014). \n10. A. I. Nash  et al., Structural Basis of Photosensitivity in a Bacterial Light-Oxygen-\nVoltage/Helix-Turn-Helix (LOV-HTH) DNA-Binding Protein. Proceedings of the National \nAcademy of Sciences 108, 9449-9454 (2011). \n11. G. Rivera-Cancel, L. B. Motta-Mena, K. H. Gardner, Identification of Natural and Artificial \nDNA Substrates for Light-Activated LOV–HTH Transcription Factor EL222. Biochemistry \n51, 10024-10034 (2012). \n12. B. D. Zoltowski, L. B. Motta-Mena, K. H. Gardner, Blue Light-Induced Dimerization of a \nBacterial LOV–HTH DNA-Binding Protein. Biochemistry 52, 6653-6661 (2013). \n13. A. S. Chaudhari  et al., Light-Dependent Flavin Redox and Adduct States Control the \nConformation and DNA-Binding Activity of the Transcription Factor EL222. Nucleic Acids \nResearch 53 (2025). \n14. L. B. Motta-Mena  et al., An Optogenetic Gene Expression System with Rapid Activation \nand Deactivation Kinetics. Nature Chemical Biology 10, 196-202 (2014). \n15. A. Reade et al., TAEL: A Zebrafish-Optimized Optogenetic Gene Expression System \nwith Fine Spatial and Temporal Control. Development 144, 345- 355 (2016). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n27 \n16. M. Rullan, D. Benzinger, G. W. Schmidt, A. Milias-Argeitis, M. Khammash, An \nOptogenetic Platform for Real-Time, Single-Cell Interrogation of Stochastic \nTranscriptional Regulation. Molecular Cell 70, 745-756.e746 (2018). \n17. E. M. Zhao  et al., Design and Characterization of Rapid Optogenetic Circuits for \nDynamic Control in Yeast Metabolic Engineering. ACS Synthetic Biology 9, 3254-3266 \n(2020). \n18. E. M. Zhao  et al., Optogenetic Regulation of Engineered Cellular Metabolism for \nMicrobial Chemical Production. Nature 555, 683-687 (2018). \n19. M. M. Cleere, K. H. Gardner, Optogenetic Control of Phosphate-Responsive Genes \nUsing Single-Component Fusion Proteins in Saccharomyces cerevisiae. ACS Synthetic \nBiology 13, 4085-4098 (2024). \n20. S. M. Hoffman, S. Espinel ‐ Ríos, S. K. Kwartler, M. A. Lalwani, J. L. Avalos, Balancing \nDoses of EL222 and Light Improves Optogenetic Induction of Protein Production in \nKomagataella phaffii. Biotechnology and Bioengineering 10.1002/bit.29027 (2025). \n21. W.-T. Im  et al., Description of Fimbriimonas ginsengisoli gen. nov., sp. nov. within the \nFimbriimonadia Class nov., of the Phylum Armatimonadetes. Antonie van Leeuwenhoek \n102, 307-317 (2012). \n22. J. J. Siclari, D. C. Favaro, R. H. Huang, K. H. Gardner, A Pipeline for Screening Small \nMolecule\n‐ Enhanced Protein Stability in a Bacterial Orphan Receptor. Protein Science 34 \n(2025). \n23. I. Letunic, S. Khedkar, P. Bork, SMART: Recent Updates, New Developments and \nStatus in 2020. Nucleic Acids Research 49, D458-D460 (2021). \n24. A. Wißbrock, A. A. Paul George, Hans, T. Kühl, D. Imhof, The Molecular Basis of \nTransient Heme-Protein Interactions: Analysis, Concept and Implementation. Bioscience \nReports 39, BSR20181940 (2019). \n25. B. San Francisco, E. C. Bretsnyder, R. G. Kranz, Human Mitochondrial Holocytochrome \nC Synthase’s Heme Binding, Maturation Determinants, and Complex Formation with \nCytochrome C. Proceedings of the National Academy of Sciences 110, E788-E797 \n(2013). \n26. K. Ghosh  et al., Spectroscopic and Biochemical Characterization of Heme Binding to \nYeast Dap1p and Mouse PGRMC1p. Biochemistry 44, 16729-16736 (2005). \n27. Y. Ishitsuka et al., Arg97 at the Heme-Distal Side of the Isolated Heme-Bound PAS \nDomain of a Heme-Based Oxygen Sensor from Escherichia coli (Ec DOS) Plays Critical \nRoles in Autoxidation and Binding to Gases, Particularly O2. Biochemistry 47, 8874-\n8884 (2008). \n28. V. Bondarenko, S. Dewilde, L. Moens, G. N. La Mar, Solution 1H NMR Characterization \nof the Axial Bonding of the Two His in Oxidized Human Cytoglobin. Journal of the \nAmerican Chemical Society 128 (2006). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n28 \n29. M. Gochin, H. Roder, Protein Structure Refinement Based on Paramagnetic NMR Shifts: \nApplications to Wild‐ Type and Mutant Forms of Cytochrome c. Protein Science 4, 296-\n305 (1995). \n30. S. M. Harper, L. C. Neil, K. H. Gardner, Structural Basis of a Phototropin Light Switch. \nScience 301, 1541-1544 (2003). \n31. X. Yao, M. K. Rosen, K. H. Gardner, Estimation of the Available Free Energy in a LOV2-\nJα  Photoswitch. Nature Chemical Biology 4, 491-497 (2008). \n32. J. A. Imelio, F. Trajtenberg, A. Buschiazzo, Allostery and Protein Plasticity: the \nKeystones for Bacterial Signaling and Regulation. Biophysical Reviews 13, 943-953 \n(2021). \n33. C. Kiel, D. Matallanas, W. Kolch, The Ins and Outs of RAS Effector Complexes. \nBiomolecules 11, 236 (2021). \n34. N. Zhang, Y. Dong, H. Zhou, H. Cui, Effect of PAS-LuxR Family Regulators on the \nSecondary Metabolism of Streptomyces. Antibiotics 11, 1783 (2022). \n35. T. Li, H. L. Bonkovsky, J.-T. Guo, Structural Analysis of Heme Proteins: Implications for \nDesign and Prediction. BMC Structural Biology 11, 13 (2011). \n36. G. Zoppellaro  et al., Review: Studies of Ferric Heme Proteins with Highly \nAnisotropic/Highly Axial Low Spin (S = 1/2) Electron Paramagnetic Resonance Signals \nwith Bis‐ Histidine and Histidine‐ Methionine Axial Iron Coordination. Biopolymers 91, \n1064-1082 (2009). \n37. A. Tanaka, T. Shimizu, Ligand Binding to the Fe(III)-Protoporphyrin IX Complex of \nPhosphodiesterase from Escherichia coli (EcDOS) Markedly Enhances Catalysis of \nCyclic di-GMP: Roles of Met95, Arg97, and Phe113 of the Putative Heme Distal Side in \nCatalytic Regulat. Biochemistry 47, 13438-13446 (2008). \n38. V. Fojtikova  et al., Kinetic Analysis of a Globin-Coupled Histidine Kinase, Af GcHK: \nEffects of the Heme Iron Complex, Response Regulator, and Metal Cations on \nAutophosphorylation Activity. Biochemistry 54, 5017-5029 (2015). \n39. M. G. Surette, J. B. Stock, Role of \nα -Helical Coiled-Coil Interactions in Receptor \nDimerization, Signaling, and Adaptation during Bacterial Chemotaxis. Journal of \nBiological Chemistry 271, 17966-17973 (1996). \n40. P. Macheboeuf et al., Streptococcal M1 Protein Constructs a Pathological Host \nFibrinogen Network. Nature 472, 64-68 (2011). \n41. R. Hurme, K. D. Berndt, E. Namork, M. Rhen, DNA Binding Exerted by a Bacterial Gene \nRegulator with an Extensive Coiled-coil Domain. Journal of Biological Chemistry 271, \n12626-12631 (1996). \n42. M. F. Berger  et al., Compact, Universal DNA Microarrays to Comprehensively Determine \nTranscription-Factor Binding Site Specificities. Nature Biotechnology 24, 1429-1435 \n(2006). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n29 \n43. M. F. Berger, M. L. Bulyk, Universal Protein-Binding Microarrays for the Comprehensive \nCharacterization of the DNA-Binding Specificities of Transcription Factors. Nature \nProtocols 4, 393-411 (2009). \n44. A. K. Bidwai, A. J. Ahrendt, J. S. Sullivan, L. B. Vitello, J. E. Erman, pH Dependence of \nCyanide and Imidazole Binding to the Heme Domains of Sinorhizobium meliloti and \nBradyrhizobium japonicum FixL. Journal of Inorganic Biochemistry 153, 88-102 (2015). \n45. X. Huang, J. T. Groves, Oxygen Activation and Radical Transformations in Heme \nProteins and Metalloporphyrins. Chemical Reviews 118, 2491-2553 (2018). \n46. W. Gong, B. Hao, M. K. Chan, New Mechanistic Insights from Structural Studies of the \nOxygen-Sensing Domain of Bradyrhizobium japonicum FixL. Biochemistry 39, 3955-\n3962 (2000). \n47. M. G. Olson, M. Goldammer, E. Gauliard, D. Ladant, S. P. Ouellette, \"A Bacterial \nAdenylate Cyclase-Based Two-Hybrid System Compatible with Gateway® Cloning\". \n(Springer New York, 2018), 10.1007/978-1-4939-7871-7_6, pp. 75-96. \n48. G. Karimova, N. Dautin, D. Ladant, Interaction Network among Escherichia coli \nMembrane Proteins Involved in Cell Division as Revealed by Bacterial Two-Hybrid \nAnalysis. Journal of Bacteriology 187, 2233-2243 (2005). \n49. A. Möglich, R. A. Ayers, K. Moffat, Structure and Signaling Mechanism of Per-ARNT-Sim \nDomains. Structure 17, 1282-1294 (2009). \n50. S. T. Glantz  et al., Functional and Topological Diversity of LOV Domain Photoreceptors. \nProceedings of the National Academy of Sciences 113, E1442-E1451 (2016). \n51. M. Chawla et al., Mycobacterium tuberculosis WhiB4 Regulates Oxidative Stress \nResponse to Modulate Survival and Dissemination in vivo. Molecular Microbiology 85, \n1148-1165 (2012). \n52. P. Mondragon  et al., TrmB Family Transcription Factor as a Thiol-Based Regulator of \nOxidative Stress Response. mBio 13 (2022). \n53. Y. Li, Z.-G. He, The Mycobacterial LysR-Type Regulator OxyS Responds to Oxidative \nStress and Negatively Regulates Expression of the Catalase-Peroxidase Gene. PLoS \nONE 7, e30186 (2012). \n54. E. K. Monson, M. Weinstein, G. S. Ditta, D. R. Helinski, The FixL Protein of Rhizobium \nmeliloti can be Separated into a Heme-Binding Oxygen-Sensing Domain and a \nFunctional C-Terminal Kinase Domain. Proceedings of the National Academy of \nSciences 89, 4280-4284 (1992). \n55. E. Orillard, S. Anaya, M. S. Johnson, K. J. Watts, Oxygen-Induced Conformational \nChanges in the PAS-Heme Domain of the Pseudomonas aeruginosa Aer2 Receptor. \nBiochemistry 60, 2610-2622 (2021). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\nSiclari et al. – Heme-binding one component sensor \n \n30 \n56. V. M. Delgado-Nixon, G. Gonzalez, M.-A. Gilles-Gonzalez, Dos, a Heme-Binding PAS \nProtein from Escherichia coli, Is a Direct Oxygen Sensor. Biochemistry 39, 2685-2691 \n(2000). \n57. D. Garcia, E. Orillard, M. S. Johnson, K. J. Watts, Gas Sensing and Signaling in the \nPAS-Heme Domain of the Pseudomonas aeruginosa Aer2 Receptor. Journal of \nBacteriology 199, JB.00003-00017 (2017). \n58. P. Mukherjee, S. Agarwal, S. B. Mallick, J. Dasgupta, PAS domain of flagellar histidine \nkinase FlrB has a unique architecture and binds heme as a sensory ligand in an \nunconventional fashion. Structure 32, 200-216.e205 (2024). \n59. F. Noya, A. Arias, E. Fabiano, Heme Compounds as Iron Sources for Nonpathogenic \nRhizobium Bacteria. Journal of Bacteriology 179, 3076-3078 (1997). \n60. E. M. Zygiel, A. O. Obisesan, C. E. Nelson, A. G. Oglesby, E. M. Nolan, Heme Protects \nPseudomonas aeruginosa and Staphylococcus aureus from Calprotectin-Induced Iron \nStarvation. Journal of Biological Chemistry 296, 100160 (2021). \n61. W. Gong et al., Structure of a Biological Oxygen Sensor: A New Mechanism for Heme-\nDriven Signal Transduction. Proceedings of the National Academy of Sciences 95, \n15177-15182 (1998). \n62. M. F. Reynolds, New Insights into the Signal Transduction Mechanism of O2-Sensing \nFixL and Other Biological Heme-Based Sensor Proteins. Journal of Inorganic \nBiochemistry 259, 112642 (2024). \n63. A. Mokdad  et al., Photoacoustic Calorimetry Studies of O2-Sensing FixL and (R200, \nI209) Variants from Sinorhizobium meliloti Reveal Conformational Changes Coupled to \nLigand Photodissociation from the Heme-PAS Domain. Biochemistry 63, 116-127 \n(2024). \n64. G. S. A. Wright  et al., Architecture of the Complete Oxygen-Sensing FixL-FixJ Two-\nComponent Signal Transduction System. Science Signaling 11, eaaq0825 (2018). \n65. C. Vonrhein et al., Data processing and analysis with the autoPROC toolbox. Acta \nCrystallographica Section D Biological Crystallography 67, 293-302 (2011). \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.15.711900doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}