OptoLacI: optogenetically engineered lactose operon repressor LacI responsive to light instead of IPTG.

Optogenetics' advancement has made light induction attractive for controlling biological processes due to its advantages of fine-tunability, reversibility, and low toxicity. The lactose operon induction system, commonly used in Escherichia coli, relies on the binding of lactose or isopropyl ß-d-1-thiogalactopyranoside (IPTG) to the lactose repressor protein LacI, playing a pivotal role in controlling the lactose operon. Here, we harnessed the light-responsive light-oxygen-voltage 2 (LOV2) domain from Avena sativa phototropin 1 as a tool for light control and engineered LacI into two light-responsive variants, OptoLacIL and OptoLacID. These variants exhibit direct responsiveness to light and darkness, respectively, eliminating the need for IPTG. Building upon OptoLacI, we constructed two light-controlled E. coli gene expression systems, OptoE.coliLight system and OptoE.coliDark system. These systems enable bifunctional gene expression regulation in E. coli through light manipulation and show superior controllability compared to IPTG-induced systems. We applied the OptoE.coliDark system to protein production and metabolic flux control. Protein production levels are comparable to those induced by IPTG. Notably, the titers of dark-induced production of 1,3-propanediol (1,3-PDO) and ergothioneine exceeded 110% and 60% of those induced by IPTG, respectively. The development of OptoLacI will contribute to the advancement of the field of optogenetic protein engineering, holding substantial potential applications across various fields.

CUREs for high-level Galectin-3 expression.

Galectins are a large and diverse protein family defined by the presence of a carbohydrate recognition domain (CRD) that binds ß-galactosides. They play important roles in early development, tissue regeneration, immune homeostasis, pathogen recognition, and cancer. In many cases, studies that examine galectin biology and the effect of manipulating galectins are aided by, or require the ability to express and purify, specific members of the galectin family. In many cases, E. coli is employed as a heterologous expression system, and galectin expression is induced with isopropyl ß-galactoside (IPTG). Here, we show that galectin-3 recognizes IPTG with micromolar affinity and that as IPTG induces expression, newly synthesized galectin can bind and sequester cytosolic IPTG, potentially repressing further expression. To circumvent this putative inhibitory feedback loop, we utilized an autoinduction protocol that lacks IPTG, leading to significantly increased yields of galectin-3. Much of this work was done within the context of a course-based undergraduate research experience, indicating the ease and reproducibility of the resulting expression and purification protocols.

Expression of soluble moloney murine leukemia virus-reverse transcriptase in Escherichia coli BL21 star (DE3) using autoinduction system.

Autoinduction systems in Escherichia coli can control the production of proteins without the addition of a particular inducer. In the present study, we optimized the heterologous expression of Moloney Murine Leukemia Virus derived Reverse Transcriptase (MMLV-RT) in E. coli. Among 4 autoinduction media, media Imperial College resulted the highest MMLV-RT overexpression in E. coli BL21 Star (DE3) with incubation time 96 h. The enzyme was produced most optimum in soluble fraction of lysate cells. The MMLV-RT was then purified using the Immobilized Metal Affinity Chromatography method and had specific activity of 629.4 U/mg. The system resulted lower specific activity and longer incubation of the enzyme than a classical Isopropyl ß-D-1-thiogalactopyranoside (IPTG)-induction system. However, the autoinduction resulted higher yield of the enzyme than the conventional induction (27.8%). Techno Economic Analysis revealed that this method could produce MMLV-RT using autoinduction at half the cost of MMLV-RT production by IPTG-induction. Bioprocessing techniques are necessary to conduct to obtain higher quality of MMLV-RT under autoinduction system.

LactoSpanks: A Collection of IPTG Inducible Promoters for the Commensal Lactic Acid Bacteria Lactobacillus gasseri.

Lactic acid bacteria (LAB) are important for many biotechnological applications such as bioproduction and engineered probiotics for therapy. Inducible promoters are key gene expression control elements, yet those available in LAB are mainly based on bacteriocin systems and have many drawbacks, including large gene clusters, costly inducer peptides, and little portability to in vivo settings. Using Lactobacillus gasseri, a model commensal bacteria from the human gut, we report the engineering of synthetic LactoSpanks promoters (Pls), a collection of variable strength inducible promoters controlled by the LacI repressor from E. coli and induced by isopropyl ß-d-1-thiogalactopyranoside (IPTG). We first show that the Phyper-spank promoter from Bacillus subtilis is functional in L. gasseri, albeit with substantial leakage. We then construct and screen a semirational library of Phyper-spank variants to select a set of four IPTG-inducible promoters that span a range of expression levels and exhibit reduced leakages and operational dynamic ranges (from ca. 9 to 28 fold-change). With their low genetic footprint and simplicity of use, LactoSpanks will support many applications in L. gasseri, and potentially other lactic acid and Gram-positive bacteria.

Mechanistic aspects of IPTG (isopropylthio-ß-galactoside) transport across the cytoplasmic membrane of Escherichia coli-a rate limiting step in the induction of recombinant protein expression.

Coupling transcription of a cloned gene to the lac operon with induction by isopropylthio-ß-galactoside (IPTG) has been a favoured approach for recombinant protein expression using Escherichia coli as a heterologous host for more than six decades. Despite a wealth of experimental data gleaned over this period, a quantitative relationship between extracellular IPTG concentration and consequent levels of recombinant protein expression remains surprisingly elusive across a broad spectrum of experimental conditions. This is because gene expression under lac operon regulation is tightly correlated with intracellular IPTG concentration due to allosteric regulation of the lac repressor protein (lacY). An in-silico mathematical model established that uptake of IPTG across the cytoplasmic membrane of E. coli by simple diffusion was negligible. Conversely, lacY mediated active transport was a rapid process, taking only some seconds for internal and external IPTG concentrations to equalize. Optimizing kcat and KM parameters by targeted mutation of the galactoside binding site in lacY could be a future strategy to improve the performance of recombinant protein expression. For example, if kcat were reduced whilst KM was increased, active transport of IPTG across the cytoplasmic membrane would be reduced, thereby lessening the metabolic burden on the cell and expediating accumulation of recombinant protein. The computational model described herein is made freely available and is amenable to optimize recombinant protein expression in other heterologous hosts. ONE-SENTENCE SUMMARY: A computational model made freely available to optimize recombinant protein expression in Escherichia coli other heterologous hosts.

Outer membrane protein N expressed in Gram-negative bacterial strain of Escherichia coli BL21 (DE3) Omp8 Rosetta strains under osmoregulation by salts, sugars, and pHs.

This study presented the expression of the outer membrane protein N in E. coli BL21 (DE3) Omp8 Rosetta under its growth condition and by osmoregulation. The effects of osmotic stress caused by salts, sugars, or pH values on the survival of the target Gram-negative bacterial strain of E. coli BL21 (DE3) Omp8 Rosetta and OmpN expression remain unknown. Here, tryptone yeast extract with varied salts and concentrations was initially used to generate an LB broth medium. To show how salts and concentration affect bacterial growth, the optical density at 600 nm was measured. The findings supported the hypothesis that salts and concentrations control bacterial growth. Moreover, a Western blotting study revealed that OmpN overexpression was present in all tested salts after stimulation with both glucose and fructose after being treated individually with anti-OmpN and anti-histidine tag polyclonal antibodies on transferred nitrocellulose membrane containing crude OmpN. Following the presence of the plasmid pET21b(+)/ompN-BOR into E. coli BL21 (DE3) Omp8 Rosetta, which was expressed in the recombinant OmpN protein (BOR), OmpN expression was demonstrated for all monovalent cations as well as MgCl2. All of the tested salts, except for BaCl2, NaH2PO4, and KH2PO4, showed overexpression of recombinant BOR after Isopropyl ß-D-1-thiogalactopyranoside (IPTG) induction. Using CH3COONa, both with and without IPTG induction, there was very little bacterial growth and no OmpN expression. With NaCl, a pH value of 7 was suitable for bacterial development, whereas KCl required a pH value of 8. According to this research, bacterial growth in addition to salts, sugars, and pH values influences how the OmpN protein is produced.

IPTG-induced high protein expression for whole-cell biosynthesis of L-phosphinothricin.

BACKGROUND: Biocatalytic production of L-phosphinothricin (L-PPT) is currently the most promising method. In this work, we use an Escherichia coli strain coexpressing of D-amino acid oxidase and catalase (E. coli DAAO-CAT) to oxidation biocatalytic D-PPT to PPO, then use the second E. coli strain coexpressing glutamate dehydrogenase and formate dehydrogenase (E. coli GluDH-FDH) to reduce biocatalytic PPO to L-PPT. MAIN METHODS AND MAJOR RESULTS: We compared the effects of different concentrations of IPTG or lactose on protein expression and enzyme activity in 5 L fermenter. The best induction conditions for E. coli DAAO-CAT were 0.05 mM IPTG, induction for 18 h at 28°C. The specific enzyme activities of DAAO and CAT were 153.20 U g-1 and 896.23 U g-1 , respectively. The optimal induction conditions for E. coli GluDH-FDH were 0.2 mM IPTG, induction for 19 h at 28°C. The specific enzyme activities of GluDH and FDH were 41.72 U g-1 and 109.70 U g-1 , respectively. The 200 mM D-PPT was biocatalyzed by E. coli DAAO-CAT for 4 h with space-time yield of 9.0 g·L-1 ·h-1 and conversion rate of over 99.0%. Then 220 mM PPO was converted to L-PPT by E. coli GluDH-FDH for 3 h with space-time yield of 14.5 g·L-1 ·h-1 and conversion rate of over 99.0%. To our knowledge, this is the most efficient biocatalytic reaction for L-PPT production. CONCLUSIONS AND IMPLICATIONS: We found that IPTG has advantages compared with lactose in the enzyme activity and biomass of E. coli DAAO-CAT and E. coli GluDH-FDH, and IPTG is more environmentally friendly. Our data implicated that IPTG can replace lactose in terms of economic feasibility and effectiveness for scaled-up industrial fermentations.

Potent IPTG-inducible integrative expression vectors for production of recombinant proteins in Bacillus subtilis.

The IPTG-inducible promoter family, Pgrac, allows high protein expression levels in an inducible manner. In this study, we constructed IPTG-inducible expression vectors containing strong Pgrac promoters that allow integration of the transgene at either the amyE or lacA locus or both loci in Bacillus subtilis. Our novel integrative expression vectors based on Pgrac promoters could control the repression of protein production in the absence and the induction in the presence of an inducer, IPTG. The ß-galactosidase (BgaB) protein levels were 9.0%, 15% and 30% of the total cellular protein in the B. subtilis strains carrying single cassettes with the Pgrac01, Pgrac100 or Pgrac212 promoters, respectively. The maximal induction ratio of Pgrac01-bgaB was 35.5 while that of Pgrac100-bgaB was 7.5 and that of Pgrac212-bgaB was 9. The inducible expression of GFP and BgaB protein was stably maintained for 24 h, with the highest yield of GFP being 24% of cell total protein while the maximum amount of BgaB was found to be 38%. A dual integration of two copies of the gfp+ gene into the B. subtilis genome at the lacA and amyE loci resulted in a yield of about 40% of total cellular protein and a 1.74-fold increase in GFP compared with single-integrated strains containing the same Pgrac212 promoter. The capability of protein production from low to high levels of these inducible integrative systems is useful for fundamental and applied research in B. subtilis.

Engineering transcriptional regulation in Escherichia coli using an archaeal TetR-family transcription factor.

Synthetic biology requires well-characterized biological parts that can be combined into functional modules. One type of biological parts are transcriptional regulators and their cognate operator elements, which enable to either generate an input-specific response or are used as actuator modules. A range of regulators has already been characterized and used for orthogonal gene expression engineering, however, previous efforts have mostly focused on bacterial regulators. This work aims to design and explore the use of an archaeal TetR family regulator, FadRSa from Sulfolobus acidocaldarius, in a bacterial system, namely Escherichia coli. This is a challenging objective given the fundamental difference between the bacterial and archaeal transcription machinery and the lack of a native TetR-like FadR regulatory system in E. coli. The synthetic σ70-dependent bacterial promoter proD was used as a starting point to design hybrid bacterial/archaeal promoter/operator regions, in combination with the mKate2 fluorescent reporter enabling a readout. Four variations of proD containing FadRSa binding sites were constructed and characterized. While expressional activity of the modified promoter proD was found to be severely diminished for two of the constructs, constructs in which the binding site was introduced adjacent to the -35 promoter element still displayed sufficient basal transcriptional activity and showed up to 7-fold repression upon expression of FadRSa. Addition of acyl-CoA has been shown to disrupt FadRSa binding to the DNA in vitro. However, extracellular concentrations of up to 2 mM dodecanoate, subsequently converted to acyl-CoA by the cell, did not have a significant effect on repression in the bacterial system. This work demonstrates that archaeal transcription regulators can be used to generate actuator elements for use in E. coli, although the lack of ligand response underscores the challenge of maintaining biological function when transferring parts to a phylogenetically divergent host.

Live Cell Imaging Using Riboswitch-Spinach tRNA Fusions as Metabolite-Sensing Fluorescent Biosensors.

The development of fluorescent biosensors is motivated by the desire to monitor cellular metabolite levels in real time. Most genetically encodable fluorescent biosensors are based on receptor proteins fused to fluorescent protein domains. More recently, small molecule-binding riboswitches have been adapted for use as fluorescent biosensors through fusion to the in vitro selected Spinach aptamer, which binds a profluorescent, cell-permeable small molecule mimic of the GFP chromophore, DFHBI. Here we describe methods to prepare and analyze riboswitch-Spinach tRNA fusions for ligand-dependent activation of fluorescence in vivo. Example procedures describe the use of the Vc2-Spinach tRNA biosensor to monitor perturbations in cellular levels of cyclic di-GMP using either fluorescence microscopy or flow cytometry. In this updated chapter, we have added procedures on using biosensors in flow cytometry to detect exogenously added compounds. The relative ease of cloning and imaging of these biosensors, as well as their modular nature, should make this method appealing to other researchers interested in utilizing riboswitch-based biosensors for metabolite sensing.

Heterologous Overexpression of Human FAD Synthase Isoforms 1 and 2.

The study of human FAD synthase enzymes requires a recombinant strategy to produce large amount of purified proteins in a soluble form. E. coli was exploited to this aim. To achieve the production of FAD synthase in a large scale, E. coli strains, plasmids (promoter, tags), growth temperature, inducer concentration, medium composition, and osmotic pressure were optimized. To date there is no universal protocol for protein expression, but for each protein a specific combination of "expression parameters" can be selected in order to maximize the results. An experimental protocol for the expression of two isoforms of the human FAD synthase was set up. The final procedures are based on the use of E. coli Rosetta(DE3) strain. Two different plasmids were used to obtain optimal amount of the two protein isoforms. In both cases, following the addition of the IPTG inducer, the growth temperature was lowered to increase the solubility of the recombinant protein. The detailed procedures for FAD synthase isoform 1 and isoform 2 overproduction are described in this protocol.

Expression of Cloned Genes in E. coli Using IPTG-Inducible Promoters.

Many Escherichia coli expression vectors make use of the lac operon. In general, the lac operator (lacO) is located downstream from the promoter of the target gene, so that binding of the lac repressor blocks transcription initiation until lactose or the isopropyl-ß-d-thiogalactopyranoside (IPTG) analog is added. The protocol given here is intended for use with IPTG-inducible vectors. l-Arabinose-inducible systems derived from the ara operon offer an alternative to expression systems based on the lac operon; guidance for their use is also provided.

Tunable expression rate control of a growth-decoupled T7 expression system by L-arabinose only.

BACKGROUND: Precise regulation of gene expression is of utmost importance for the production of complex membrane proteins (MP), enzymes or other proteins toxic to the host cell. In this article we show that genes under control of a normally Isopropyl ß-D-1-thiogalactopyranoside (IPTG)-inducible PT7-lacO promoter can be induced solely with L-arabinose in a newly constructed Escherichia coli expression host BL21-AI, a strain based on the recently published approach of bacteriophage inspired growth-decoupled recombinant protein production. RESULTS: Here, we show that BL21-AI is able to precisely regulate protein production rates on a cellular level in an L-arabinose concentration-dependent manner and simultaneously allows for reallocation of metabolic resources due to L-arabinose induced growth decoupling by the phage derived inhibitor peptide Gp2. We have successfully characterized the system under relevant fed-batch like conditions in microscale cultivation (800 µL) and generated data proofing a relevant increase in specific yields for 6 different Escherichia coli derived MP-GFP fusion proteins by using online-GFP signals, FACS analysis, SDS-PAGE and western blotting. CONCLUSIONS: In all cases tested, BL21-AI outperformed the parental strain BL21-AI, operated in growth-associated production mode. Specific MP-GFP fusion proteins yields have been improved up to 2.7-fold. Therefore, this approach allows for fine tuning of MP production or expression of multi-enzyme pathways where e.g. particular stoichiometries have to be met to optimize product flux.

Multiplexed characterization of rationally designed promoter architectures deconstructs combinatorial logic for IPTG-inducible systems.

A crucial step towards engineering biological systems is the ability to precisely tune the genetic response to environmental stimuli. In the case of Escherichia coli inducible promoters, our incomplete understanding of the relationship between sequence composition and gene expression hinders our ability to predictably control transcriptional responses. Here, we profile the expression dynamics of 8269 rationally designed, IPTG-inducible promoters that collectively explore the individual and combinatorial effects of RNA polymerase and LacI repressor binding site strengths. We then fit a statistical mechanics model to measured expression that accurately models gene expression and reveals properties of theoretically optimal inducible promoters. Furthermore, we characterize three alternative promoter architectures and show that repositioning binding sites within promoters influences the types of combinatorial effects observed between promoter elements. In total, this approach enables us to deconstruct relationships between inducible promoter elements and discover practical insights for engineering inducible promoters with desirable characteristics.

Optogenetic control of the lac operon for bacterial chemical and protein production.

Control of the lac operon with isopropyl ß-D-1-thiogalactopyranoside (IPTG) has been used to regulate gene expression in Escherichia coli for countless applications, including metabolic engineering and recombinant protein production. However, optogenetics offers unique capabilities, such as easy tunability, reversibility, dynamic induction strength and spatial control, that are difficult to obtain with chemical inducers. We have developed a series of circuits for optogenetic regulation of the lac operon, which we call OptoLAC, to control gene expression from various IPTG-inducible promoters using only blue light. Applying them to metabolic engineering improves mevalonate and isobutanol production by 24% and 27% respectively, compared to IPTG induction, in light-controlled fermentations scalable to at least two-litre bioreactors. Furthermore, OptoLAC circuits enable control of recombinant protein production, reaching yields comparable to IPTG induction but with easier tunability of expression. OptoLAC circuits are potentially useful to confer light control over other cell functions originally designed to be IPTG-inducible.

Molecular Cloning and Differential Gene Expression Analysis of 1-Deoxy-D-xylulose 5-Phosphate Synthase (DXS) in Andrographis paniculata (Burm. f) Nees.

Andrographis paniculata 1-deoxy-D-xylulose-5-phosphate synthase (ApDXS) gene (GenBank Accession No MG271749.1) was isolated and cloned from leaves for the first time. Expression of ApDXS gene was carried out in Escherichia coli Rosetta cells. Tissue-specific ApDXS gene expression by quantitative RT-PCR (qRT-PCR) revealed maximum fold expression in the leaves followed by stem and roots. Further, the differential gene expression profile of Jasmonic acid (JA)-elicited in vitro adventitious root cultures showed enhanced ApDXS expression compared to untreated control cultures. A. paniculata 3-hydroxy-3-methylglutaryl-coenzyme A reductase (ApHMGR) gene expression was also studied where it was up-regulated by JA elicitation but showed lower expression compared to ApDXS. The highest expression of both genes was found at 25 µm JA elicitation followed by 50 µm. HPLC data indicated that the transcription levels were correlated with increased andrographolide accumulation. The peak level of andrographolide accumulation was recorded at 25 µM JA (9.38-fold) followed by 50 µM JA (7.58-fold) in elicitation treatments. The in silico generated ApDXS 3D model revealed 98% expected amino acid residues in the favored and 2% in the allowed regions of the Ramachandran plot with 92% structural reliability. Further, prediction of conserved domains and essential amino acids [Arg (249, 252, 255), Asn (307) and Ser (247)] involved in ligand/inhibitor binding was carried out by in silico docking studies. Our present findings will generate genomic information and provide a blueprint for future studies of ApDXS and its role in diterpenoid biosynthesis in A. paniculata.

Heterologous expression, purification, and refolding of SRY protein: role of L-arginine as analyzed by simulation and practical study.

Escherichia coli is a widely-used cell factory for recombinant protein production, nevertheless, high amount of produced protein is seen in aggregated form. The purpose of this study was to improve the solubility of recombinant bovine sex-determining region Y protein (rbSRY) by exploring the effect of temperature, inducer, and water-arginine mixed solvent. Codon-optimized rbSRY expressed in Rosetta-gami B (DE3) pLysS and purified by NI-NTA His-select affinity chromatography in the native and denaturing conditions. A three-dimensional model of SRY was built and studied through molecular dynamics simulations in water and in the presence of L-arginine as co-solvent. Results indicated the significant effects of temperature and IPTG concentration (P < 0.001) on the solubility of rbSRY. The binding activity of native, inclusion bodies and refolded fractions to anti-rbSRY monoclonal antibody were concentration-dependent (P < 0.001). Based on molecular modeling results, the propensity of fragments in the N-terminal domain to form ß-sheet and the relative instability of α-helices in terminal domains are the probable reasons for the high aggregation potential of SRY, which are mitigated in the presence of L-arginine. Altogether, our rbSRY protein was properly produced and applying appropriate culture conditions could help enhance its solubility, refold inclusion bodies, and improve its activity upon refolding.

Controlling spatiotemporal pattern formation in a concentration gradient with a synthetic toggle switch.

The formation of spatiotemporal patterns of gene expression is frequently guided by gradients of diffusible signaling molecules. The toggle switch subnetwork, composed of two cross-repressing transcription factors, is a common component of gene regulatory networks in charge of patterning, converting the continuous information provided by the gradient into discrete abutting stripes of gene expression. We present a synthetic biology framework to understand and characterize the spatiotemporal patterning properties of the toggle switch. To this end, we built a synthetic toggle switch controllable by diffusible molecules in Escherichia coli. We analyzed the patterning capabilities of the circuit by combining quantitative measurements with a mathematical reconstruction of the underlying dynamical system. The toggle switch can produce robust patterns with sharp boundaries, governed by bistability and hysteresis. We further demonstrate how the hysteresis, position, timing, and precision of the boundary can be controlled, highlighting the dynamical flexibility of the circuit.

Development of the Mammalian Expression Vector System that can be Induced by IPTG and/or Lactose.

Techniques used for the regulation of gene expression facilitate studies of gene function and treatment of diseases via gene therapy. Many tools have been developed for the regulation of gene expression in mammalian cells. The Lac operon system induced with isopropyl ß-D-1- thiogalactopyranoside (IPTG) is one of the employed inducible systems. IPTG mimics the molecular structure of allolactose and has a strong affinity for the corresponding repressor. IPTG is known to rapidly penetrate into mammalian cells and exhibits low toxicity. In the present study, we developed a new inducible expression system that could regulate the expression of genes in mammalian cells using IPTG. Here we confirm that unlike other vector systems based on the Lac operon, this expression system allows regulation of gene expression with lactose in the mammalian cells upon transfection. The co-treatment with IPTG and lactose could improve the regulatory efficiency of the specific target gene expression. The regulation of gene expression with lactose has several benefits. Lactose is safe in humans as compared to other chemical substances and is easily available, making this technique very cost-effective.

Lactic acid containing polymers produced in engineered Sinorhizobium meliloti and Pseudomonas putida.

This study demonstrates that novel polymer production can be achieved by introducing pTAM, a broad-host-range plasmid expressing codon-optimized genes encoding Clostridium propionicum propionate CoA transferase (PctCp, Pct532) and a modified Pseudomonas sp. MBEL 6-19 polyhydroxyalkanoate (PHA) synthase 1 (PhaC1Ps6-19, PhaC1400), into phaC mutant strains of the native polymer producers Sinorhizobium meliloti and Pseudomonas putida. Both phenotypic analysis and gas chromatography analysis indicated the synthesis and accumulation of biopolymers in S. meliloti and P. putida strains. Expression in S. meliloti resulted in the production of PLA homopolymer up to 3.2% dried cell weight (DCW). The quaterpolymer P (3HB-co-LA-co-3HHx-co-3HO) was produced by expression in P. putida. The P. putida phaC mutant strain produced this type of polymer the most efficiently with polymer content of 42% DCW when cultured in defined media with the addition of sodium octanoate. This is the first report, to our knowledge, of the production of a range of different biopolymers using the same plasmid-based system in different backgrounds. In addition, it is the first time that the novel polymer (P(3HB-co-LA-co-3HHx-co-3HO)), has been reported being produced in bacteria.