Engineering

Throughout our project, we rigorously applied the Design-Build-Test-Learn (DBTL) engineering cycle to develop and optimize biological systems for heavy metal bioremediation and biosensing. Beginning with rational gene design and codon optimization, we systematically constructed and validated expression vectors for key proteins (AtPCS and PseMT) that can chelate and reduce heacy metal ions into nanoparticles in vivo, iteratively optimizing expression conditions in E. coli and ensuring robust solubility and activity. We then advanced to co-expression strategies and demonstrated functional synergy in metal tolerance and sequestration, verified both quantitatively and visually. This iterative approach extended to plant and cyanobacterial systems, where transient expression and transformation protocols were rapidly tested and refined—enabling fast cycles of hypothesis-driven experimentation. Finally, we built modular screening platforms for synthetic promoters, ready for high-throughput functional testing. At each stage, our cycles of design, construction, empirical validation, and data-driven learning enabled us to identify bottlenecks, adapt protocols, and achieve key milestones. This workflow has established a strong foundation for expanding synthetic biology applications in environmental remediation and smart biosensing.

As a seventh-year iGEM team, we are committed to the long-term impact and continuity of our project. Recognizing that several critical modules—such as stable transformation and functional demonstration in cyanobacteria, as well as the identification and validation of high-performing i-motif promoter biosensors—remain works in progress, we will carry forward these unfinished goals into iGEM 2026. Building on the strong technical foundation and engineering platforms established this year, our team will leverage the experience and data accumulated to further advance synthetic biology solutions for environmental remediation and biosensing in the next competition cycle.



1. Optimize the expressing condition of AtPCS and pseMTs in Ecoli

A key foundation of our project was the successful and efficient expression of two core proteins: phytochelatin synthase (AtPCS) from Arabidopsis thaliana and metallothionein (PseMT) from Pseudomonas in E. coli. These proteins are essential for our system’s heavy metal binding and sequestration as nano particles, so optimizing their expression was a critical early milestone.

Primary structures of PCs (A) and PC-Cd complexes (B). The structure of PCs is generally indicated as (γ-Glu-Cys)n-Gly, where n=2 to 11. Heavy metal ions such as Cd coordinately bind one, two, three or, at maximum capacity, four sulfur atoms from either single or multiple PC molecules, resulting in amorphous complexes.

Structure of MT from P. fluorescens Q2-87. (A) Representative structure of the protein backbone fold. (B) Topology of Zn3Cys9His cluster. (C) Topology of Cd4Cys9His cluster. (B,C) are reproduced from Habjanic et al. (2018) with permission from The Royal Society of Chemistry. (Habjanič, J., et al 2020)

To begin, we performed codon optimization of both AtPCS and PseMT to match E. coli’s preferred codon usage, ensuring high translation efficiency and protein yield. The optimized genes were synthesized commercially and cloned into the pET28a expression vector, which provides a strong T7 promoter for IPTG-inducible protein expression and adds a His-tag for easy detection and purification. The constructs were sequence-verified after transformation into E. coli Top10 cells for plasmid amplification, ensuring accuracy before moving to protein expression experiments.

Plasmid map for pET28a-AtPCS

Plasmid map for pET28a-pseMT

We then transformed the validated plasmids into E. coli BL21(DE3), a strain commonly used for protein production. To systematically optimize expression, we set up a protocol where single colonies were first grown overnight in selective LB medium. The next day, cultures were diluted and grown to mid-log phase (OD₆₀₀ ≈ 0.6), then induced with a range of IPTG concentrations (from 0 to 1 mM). Importantly, we chose an induction temperature of 30°C—not only to balance protein expression and solubility, but also to simulate the environmental conditions under which our target cyanobacterium, Synechocystis sp. PCC 6803, naturally grows. Cultures were incubated at this temperature for 12 hours. After induction, we harvested the cells, lysed them under neutral conditions, and separated soluble and insoluble protein fractions by centrifugation. (details refer to: https://2025.igem.wiki/sz-shd/protocol)

SDS-PAGE analysis of AtPCS expression in E. coli BL21 (DE3) under different IPTG induction concentrations at 30°C for 12 hours. Lane 1: protein marker; Lane 2: 0 mM IPTG; Lane 3: 0.2 mM IPTG; Lane 4: 0.4 mM IPTG; Lane 5: 0.6 mM IPTG; Lane 6: 0.8 mM IPTG; Lane 7: 1 mM IPTG. The arrow indicates the induced AtPCS protein band at ~55 kDa.

SDS-PAGE analysis of soluble PseMT protein expressed in E. coli BL21(DE3) following induction with increasing concentrations of IPTG at 30°C for 12 hours. After cell lysis under neutral conditions, only the soluble protein fractions were analyzed. Lane 1: protein marker; Lane 2: 0 mM IPTG; Lane 3: 0.2 mM IPTG; Lane 4: 0.4 mM IPTG; Lane 5: 0.6 mM IPTG; Lane 6: 0.8 mM IPTG; Lane 7: 1 mM IPTG. The arrow indicates the soluble PseMT protein band (~10 kDa).

The expression levels and solubility of AtPCS and PseMT were evaluated by SDS-PAGE. Distinct bands at approximately 54.5 kDa (AtPCS) and 7.9 kDa (PseMT) were visible in the soluble fractions, and the intensity of these bands increased with higher IPTG concentrations. These results confirmed that both proteins were not only highly expressed but also remained soluble under the chosen conditions—an essential requirement for their downstream metal chelation activity.

Through this systematic optimization process, we determined that inducing protein expression at 30°C with 0.5 mM IPTG for 12 hours provided the best yield and solubility for both AtPCS and PseMT. This protocol was then adopted for all subsequent co-expression and functional assays. Our work demonstrates the importance of rational gene design, careful construct validation, and systematic parameter testing—key elements of the Design-Build-Test-Learn (DBTL) cycle in synthetic biology.

2. Co-expression of AtPCS and PseMT in Ecoli

After successfully optimizing the individual expression of AtPCS and PseMT in E. coli, our next goal was to achieve their simultaneous co-expression, enabling the engineered cells to harness the synergistic heavy metal binding and detoxification properties of both proteins. To accomplish this, we constructed a dual-expression vector by assembling the codon-optimized AtPCS and PseMT genes into the same pET28a backbone using Golden Gate assembly. The resulting co-expression plasmid, pET28a-AtPCS-PseMT, was verified by nanopore whole plasmid sequencing by Genewiz.

We transformed this construct into E. coli BL21(DE3) cells, and induced protein expression with 0.5 mM IPTG at 30°C for 12 hours, following conditions previously optimized for solubility and protein yield. To evaluate whether both proteins were produced simultaneously, we extracted total soluble protein from induced and uninduced cultures and performed Western blot analysis using an anti-His tag antibody. The results revealed two distinct bands in the induced samples: one at approximately 54.5 kDa, corresponding to His-tagged AtPCS, and another at around 7.9 kDa, corresponding to His-tagged PseMT. Importantly, no such bands were observed in the negative control (cells with empty vector or uninduced cultures), confirming that the observed expression was both specific and IPTG-dependent.

Western blot analysis confirming co-expression of His-tagged AtPCS (~55 kDa) and PseMT (~10 kDa) in E. coli BL21(DE3). Lane 1: protein molecular weight marker; Lane 2: negative control (nc, uninduced); Lane 3: cells induced with 0.5 mM IPTG. Blots were probed with an anti-His tag antibody. Bands corresponding to AtPCS and PseMT are visible only in the induced sample, indicating successful co-expression of both proteins. GAPDH served as internal reference

The successful detection of both AtPCS and PseMT in the same sample demonstrated effective co-expression in E. coli. This achievement was further validated by the use of GAPDH as a housekeeping gene, which confirmed uniform protein loading across samples. The establishment of a robust co-expression system not only streamlines downstream functional studies but also provides a strong foundation for engineering microbial platforms capable of enhanced heavy metal remediation. By confirming that both proteins can be efficiently and specifically produced together in E. coli, we have taken a critical step forward in realizing our synthetic biology solution for environmental heavy metal pollution.

3. Function characterization of AtPCS and PseMT in Ecoli

To ensure that our assays reflected biologically meaningful stress—challenging homeostasis but not overwhelming or killing the cells outright—we grounded our metal concentration choices in published studies on E. coli metal tolerance and adaptation. For Fe²⁺, concentrations were chosen based on findings that E. coli can adapt to millimolar levels of FeSO₄, displaying both stress responses and survival (Thomas et al., 2021). For Co²⁺, we referenced Ranquet et al. 2007 (JBC) and Majtan et al. 2010, which show that cobalt at sub-millimolar to low millimolar concentrations is biologically active, disrupts iron metabolism, and induces cellular stress without causing instant lethality. For Zn²⁺ and Cu²⁺, we consulted studies demonstrating that moderate excess (typically ~0.1–0.5 mM Zn²⁺) perturbs metal homeostasis and triggers stress responses but allows for cell growth and adaptation (Osman et al., 2017, Nat Commun; Xu et al., 2019). These studies collectively justify our use of metal concentrations that are high enough to induce clear stress phenotypes and measurable biological responses, yet within the plausible range for experimental evolution and physiological challenge.

To evaluate their ability to tolerate and interact with heavy metals, we exposed E. coli strains expressing either AtPCS, PseMT, or both (PCS-MT), alongside a negative control (empty vector), to LB media supplemented with a mixture of copper (Cu), zinc (Zn), nickel (Ni), and cobalt (Co) ions at defined concentrations. After induction of protein expression with IPTG, the cultures were incubated with metals, and we monitored both cell growth and changes in medium appearance.

Escherichia coli cells were incubated in LB medium supplemented with 1.5 mM Cu, 2.5 mM Zn, 3 mM Ni, and 1.5 mM Co at 30°C with shaking at 220 rpm for 24 hours. NC: cells with the empty pET28a vector; pseMTs: cells expressingpseMT; AtPCS: cells expressing AtPCS.

Growth of E. coli expressing the pET28a-PCS-MT co-expression vector under different metal concentrations over time. NC: negative control with empty pET28a vector. 1X: pET28a-PCS-MT with 1.5 mM Cu, 2.5 mM Zn, 3 mM Ni, and 1.5 mM Co. 5X: pET28a-PCS-MT with five times the 1X metal concentrations. Cultures were incubated at 30°C and sampled at 0, 24, and 48 hours.

Strikingly, cells co-expressing AtPCS and PseMT demonstrated markedly improved tolerance to heavy metal stress compared to strains expressing either protein alone or the control. While higher concentrations of metals (5X) inhibited growth in all strains, at 1X concentration, the co-expressing strain maintained robust growth and displayed less medium discoloration, indicating reduced metal toxicity. In contrast, strains expressing only AtPCS or PseMT showed only moderate improvement over the control, underscoring the enhanced protective effect of the combined system.

To directly measure metal uptake, we quantified the residual concentrations of Cu, Zn, Ni, and Co in the culture supernatant using inductively coupled plasma optical emission spectrometry (ICP-OES) after 24 and 48 hours of incubation. The results revealed that the co-expressing strain removed approximately 50% more metal ions from the solution after 48 hours compared to the negative control, with significantly greater reduction than either single-gene strain. This confirmed that the combined action of AtPCS and PseMT confers superior metal sequestration capacity, likely due to their complementary mechanisms of chelation and binding.

Workflow for sample preparation and analysis of metal ion concentration by ICP-OES.(1) Preparation of reagents and dilution of culture supernatant samples under a fume hood.(2) Aliquoting and organization of diluted samples into tubes for analysis.(3) Measurement of metal ion concentrations using an Agilent 720ES inductively coupled plasma optical emission spectrometer (ICP-OES).

Concentration of Cu, Zn, Ni, and Co ions remaining in the culture supernatant as measured by mass spectrometry (mg/L). Orange bars indicate control samples (without plasmid), gray bars indicate samples with the pET28a-PCS-MT co-expression plasmid after 24 hours, and blue bars indicate samples with the plasmid after 48 hours incubation. A significant reduction(~50%) in metal ion concentration was observed in cultures expressing the plasmid, particularly after 48 hours, demonstrating the enhanced metal uptake by engineered E. coli.

Further visual confirmation of functional metal sequestration was obtained using transmission electron microscopy (TEM). Engineered E. coli cells expressing both AtPCS and PseMT accumulated dense intracellular nanoparticles following metal exposure, as evidenced by electron-dense particles within the cytoplasm. These nanoparticles were far less prevalent or absent in the control strain, providing direct ultrastructural evidence of enhanced metal accumulation by the co-expressing cells.

Cell pellets of E. coli after 24-hour incubation in LB medium containing 1.5 mM Cu, 2.5 mM Zn, 3 mM Ni, and 1.5 mM Co at 30°C 24hr, followed by centrifugation at 8000 × g for 1 minute. NC: cells with the empty pET28a vector; 1X: cells expressing the pET28a-PCS-MT co-expression vector. The brown coloration of the 1X pellet indicates the synthesis of metal nanodots by the engineered cells.

TEM image of engineered E. coli cells showing intracellular electron-dense metal nanoparticles (red arrows). Cells were processed according to standard TEM protocols, including fixation with glutaraldehyde and osmium acid, dehydration, resin embedding, ultrathin sectioning (70–90 nm), and staining with uranyl acetate and lead citrate. Imaging was performed on a JE 1200-EX TEM (JEOL, Tokyo, Japan).

Taken together, these functional assays conclusively demonstrate that E. coli strains co-expressing AtPCS and PseMT exhibit much greater resilience to heavy metal stress and a higher capacity for metal uptake and nanoparticle formation than strains expressing either protein alone. This superior performance highlights the effectiveness of our synthetic biology approach and confirms that combining phytochelatin synthases with metallothioneins offers a powerful strategy for biological heavy metal remediation.

4. Expression and function validation of AtPCS and pseMTs in Moss Physcomitrella patens

To evaluate the potential of Physcomitrella patens as a chassis for heavy metal bioremediation, we focused on establishing a rapid and efficient transient expression system to test the function of two key metal-binding proteins: phytochelatin synthase (AtPCS) and Pseudomonas metallothionein (pseMT). Rather than generating stable transgenic lines—a process that is time-consuming and labor-intensive in moss—we adopted a strategy based on transient transformation, aiming for fast cycles of design, testing, and optimization. All constructs for this work were assembled in the broad-host-range pMMB67EH backbone, which supports high-level gene expression in plant systems.

Our workflow began with the preparation of sterile Physcomitrella tissue. To eliminate any background contamination, non-sterile moss from laboratory stock was transferred onto BCD medium containing 100 mg/L timentin and incubated at 25°C under continuous light. Over the course of two weeks, small fragments of moss proliferated into healthy, contaminant-free colonies, providing a reliable substrate for transformation. This step was essential to avoid microbial interference in both the transformation process and subsequent protein expression analysis.

Establishment of sterile Physcomitrella patens cultures from non-sterile laboratory stock. Moss was transferred to BCD medium containing 100 mg/L timentin and incubated at 25°C with 5000 LUX continuous light. Left: Moss fragments at Day 0. Right: Proliferation of sterile moss after 14 days.

For gene delivery, we constructed the pMMl8-35S-pcs-mt vector by cloning the codon-optimized AtPCS and pseMT sequences under the control of the cauliflower mosaic virus 35S(Cmv35s) promoter into the pMMB67EH plasmid using Golden Gate assembly. The vector was confirmed nanopore whole plasmid sequencing by genewiz, ensuring the fidelity of the assembled construct. Rather than using traditional PEG-mediated or biolistic transformation methods, we opted for a carbon nanodot-mediated transformation protocol, which allowed us to deliver plasmids efficiently into moss cells and achieve rapid, transient gene expression. In this protocol, approximately 200 mg of sterile moss tissue was incubated in a transformation solution containing MES buffer, 0.5% glycerol, and a suspension of carbon nanodots pre-complexed with the pmml8-35S-pcs-mt plasmid.(Details in https://2025.igem.wiki/sz-shd/protocol) The moss was gently agitated in this solution for twelve hours to maximize uptake, then transferred to liquid BCD medium and cultured under continuous light at 25°C.

To assess whether the introduced genes were expressed, we harvested moss samples at 0, 24, and 48 hours post-transformation and extracted total protein for western blot analysis. Membranes were probed with an anti-His tag antibody to detect the recombinant AtPCS (~54.5 kDa) and pseMT (~7.9 kDa) proteins, and then reprobed with anti-plant β-actin antibody as a loading control. Robust expression of both AtPCS and pseMT was observed at 24 and 48 hours, with negligible signal at 0 hours, confirming the effectiveness of the transient transformation system and the activity of our pMMB67EH-based construct in Physcomitrella cells.

Western blot analysis of AtPCS and pseMTs expression in Physcomitrella patens following transformation with carbon nano dots. Total protein was extracted at 0, 24, and 48 hours post-transformation. Blots were probed with anti-His tag antibody to detect AtPCS (~54.5 kDa) and pseMTs (~7.9 kDa), then reprobed with anti-β-actin antibody as internal reference (~39 kDa). Strong bands at 24 and 48 hours indicate successful uptake and expression of both recombinant proteins in moss tissue.

Having established successful protein expression, we next sought to determine whether these recombinant proteins conferred enhanced heavy metal uptake and nanoparticle biosynthesis. Transformed moss tissues were exposed to solutions containing copper, zinc, nickel, and cobalt ions for 24–48 hours. Following metal exposure, tissues were fixed with glutaraldehyde and prepared for analysis by environmental scanning electron microscopy (ESEM). ESEM imaging revealed clear aggregates of electron-dense metal nanoparticles distributed across the surface of the moss at both 1,000× and 2,500× magnification, as indicated by red arrows in the micrographs. These aggregates were not present in untransformed control moss, demonstrating that transiently expressed AtPCS and pseMT enabled the moss to synthesize and accumulate metal nanoparticles extracellularly under metal stress.

Environmental scanning electron microscope (ESEM) images of Physcomitrella patens tissue after metal nanoparticle biosynthesis. Moss samples were fixed with 2.5% glutaraldehyde and imaged using a Quattro ESEM at 5.00 kV accelerating voltage, 57 pA beam current, 5.4 mm working distance, and secondary electron detection (SE mode). Nanoparticle aggregates (red arrows) are visible on the moss surface at 1,000× magnification (left, scale bar: 40 μm) and 2,500× magnification (right, scale bar: 10 μm), confirming successful metal nanoparticle formation.

5. Expression of AtPCS and pseMTs in Cyanobacteria PCC6803 and integration into bioreactor

To further expand the versatility and sustainability of our heavy metal bioremediation platform, we sought to harness the unique advantages of the cyanobacterium Synechocystis sp. PCC6803. This photosynthetic microorganism is attractive for environmental applications due to its ability to grow autotrophically under light, its ease of cultivation in aqueous systems, and its suitability for scale-up in controlled bioreactors. Our goal was to engineer PCC6803 to co-express Arabidopsis phytochelatin synthase (AtPCS) and Pseudomonas metallothionein (pseMTs), and to lay the groundwork for future bioprocess development.

Abstract of SZ-SHD 2025 project

The engineering process began with the design and assembly of an expression vector tailored for use in cyanobacteria. We first validated the function of a theophylline-inducible promoter system by constructing a reporter plasmid expressing blue fluorescent protein (BFP), using BBa_K592100 from the iGEM 2023 distribution kit. After confirming that the promoter permitted gene induction in Synechocystis cells, we replaced the BFP reporter with codon-optimized pcs and mt genes, enabling potential co-expression of AtPCS and pseMTs in response to theophylline. The use of the pMMB67EH backbone facilitated modular cloning and broad-host-range compatibility, supporting our synthetic biology approach.

Transformation of cyanobacteria presents unique challenges, particularly with respect to efficiency and stability. We employed the natural transformation protocol for Synechocystis PCC6803, leveraging its innate competence. Exponentially growing cells were incubated with our constructed vector and plated on BG11 agar containing kanamycin as a selective agent. Despite repeated attempts, we observed no colony formation on plates containing the transformed cells, while control plates also showed no growth, confirming the stringency of selection. The lack of transformants suggested that our protocol may require further optimization or that alternative transformation strategies—such as conjugation or electroporation—may be needed to achieve successful integration and expression of the target genes.

Natural transformation of Synechocystis sp. PCC 6803. Left: Control BG11 plate containing kanamycin shows no growth. Right: BG11 plate with transformed cells also shows an absence of colony formation after incubation, indicating that no kanamycin-resistant transformants were obtained.

Although we were not able to generate engineered PCC6803 strains within this cycle, we proceeded to characterize the growth and biomass production of wild-type cultures in a laboratory bioreactor as a foundation for future system integration.

Cultures were grown at 30°C under continuous illumination of 5000 LUX, and OD600 was measured daily to monitor cell density. The growth curve closely followed the Gompertz model, with an excellent fit (R² = 0.9905), providing a robust quantitative framework for predicting culture performance and planning scale-up .

Growth curve of wild-type Synechocystis sp. PCC 6803 in a light bioreactor at 30°C with continuous 5000 LUX illumination. Blue dots represent observed OD600 values measured daily. The orange solid line shows the Gompertz model fit (R² = 0.9905), and the yellow dashed line represents the model extension beyond the observation period. The fitted equation and parameters are indicated at the bottom of the plot.

In parallel, we measured the wet biomass yield by harvesting 10 mL culture samples over three days, with three replicates per day. The average biomass yield was determined to be 0.0087 g per OD per mL of medium, a key metric for future assessments of heavy metal uptake and process throughput .

Measurement of wet biomass for Synechocystis sp. PCC 6803 cultures from the bioreactor. Representative image showing a pre-weighed tube containing harvested cyanobacterial cells. Biomass was measured from 10 mL culture samples across three days (three replicates per day). The average yield was determined to be 0.0087 g/OD/mL medium.

Overall, while stable expression of AtPCS and pseMTs in Synechocystis PCC6803 remains an ongoing challenge, our work has established the molecular tools and bioreactor framework necessary for the next stage of synthetic biology-driven bioremediation. The validation of the inducible promoter system and the baseline bioreactor data provide a strong platform for iterative optimization. In future cycles, we aim to improve transformation efficiency, explore alternative gene delivery methods, and ultimately demonstrate the light-driven, scalable removal of heavy metals from aqueous environments using engineered cyanobacteria. This integrated approach will advance the realization of sustainable, biotechnological solutions for environmental remediation.

6. Engineering a Screening Platform for Carbon Dot-Inducible Promoters as Sensors of Cellular Saturation with Reduced Metal Nanoparticles

As the field of plant synthetic biology advances toward increasingly precise and responsive genetic circuits, the need for biosensors that can report on cellular status in real time becomes ever more pressing. In response, we have engineered a modular screening system to develop synthetic promoters that can indicate when a plant cell is saturated with reduced metal nanoparticles—a common endpoint in phytoremediation and plant nanobiotechnology.

Building on previous discoveries that carboxyl-modified carbon dots can modestly activate native i-motif-containing promoters in rice, we hypothesized that synthetic promoters incorporating these motifs could serve as sensitive elements for detecting changes in intracellular nanoparticle concentrations.

CDs activate the expression of chromomethylase (CMT) genes by binding to i-motifs within these genes (Huang, R et al 2025)

To enable rapid functional screening, we constructed a robust dual-reporter platform based on the pGreen-mscarlet-lv0 vector, which includes both luciferase and fluorescent reporters, alongside a Golden Gate-compatible entry system. This modular setup allows us to efficiently clone individual i-motif sequences upstream of a minimal promoter and quantitatively assess their responsiveness within plant cells.

With this testing system in place, our next step is to systematically screen the library of i-motif sequences by introducing them into the reporter construct and evaluating their activation in response to carbon dot exposure and intracellular accumulation of reduced metal nanoparticles. By measuring luciferase activity and fluorescence output, we aim to identify specific i-motif configurations that act as effective sensors for when the cell approaches or reaches saturation with metal nanoparticles.

Although the optimal i-motif-based promoter has not yet been identified, the establishment of this screening platform represents a significant engineering milestone. It provides a direct and scalable route to discovering new biosensors that can be integrated into regulatory circuits for plant-based nanomaterial production, environmental monitoring, or phytoremediation. Once validated, such promoters could enable plants to autonomously report on their nanoparticle load, paving the way for feedback-regulated expression systems and safer, more efficient metallophyte applications.

References:

  • Majtan, T., Frerman, F. E., & Kraus, J. P. (2011). Effect of cobalt on Escherichia coli metabolism and metalloporphyrin formation. Biometals, 24(2), 335–347. https://doi.org/10.1007/s10534-010-9400-7
  • Osman, D., Foster, A. W., Chen, J., Svedaite, K., Steed, J. W., Lurie-Luke, E., Huggins, T. G., & Robinson, N. J. (2017). Fine control of metal concentrations is necessary for cells to discern zinc from cobalt. Nature Communications, 8(1), Article 1884. https://doi.org/10.1038/s41467-017-02085-z
  • Ranquet, C., Ollagnier-de-Choudens, S., Loiseau, L., Barras, F., & Fontecave, M. (2007). Cobalt stress in Escherichia coli: The effect on the iron-sulfur proteins. The Journal of Biological Chemistry, 282(42), 30442–30451. https://doi.org/10.1074/jbc.M702519200
  • Thomas, M. D., Ewunkem, A. J., Boyd, S., Williams, D. K., Moore, A., Rhinehardt, K. L., Van Beveren, E., Yang, B., Tapia, A., Han, J., Harrison, S. H., & Graves, J. L., Jr. (2021). Too much of a good thing: Adaption to iron (II) intoxication in Escherichia coli. Evolution, Medicine, and Public Health, 9(1), 53–67. https://doi.org/10.1093/emph/eoaa051
  • Xu, Z., Wang, P., Wang, H., Yu, Z. H., Au-Yeung, H. Y., Hirayama, T., Sun, H., & Yan, A. (2019). Zinc excess increases cellular demand for iron and decreases tolerance to copper in Escherichia coli. The Journal of Biological Chemistry, 294(45), 16978–16991. https://doi.org/10.1074/jbc.RA119.010023