Lanthanoid-Binding Peptides and their Surface Display in Yeast

Our lab work was organized to lay the biological foundation for our novel approach of recycling Rare-Earth Elements. We used cutting edge technology to create the world's first Saccharomyces cerevisiae colony, that expresses lanthanide-binding proteins on its cell surface. We were able to successfully prove the capabilities of our recombinant yeast, as well as synthetically immobilized peptides, to effectively isolate Rare-Earths from leachate-like solutions. Additionally, we were able to utilize phage display as a state-of-the-art technology for the identification of new lanthanide-binding peptides.


To facilitate the filtration of lanthanides under realistic recycling conditions, we chose filamentous fungi as an optimal organism for the creation of a biological adsorption matrix. These filamentous fungi could be genetically modified to display lanthanide binding-peptides on their cell surface and therefore enhance their adsorption ability. Because we wanted to help future projects in this field as well, our lab was designed to present a logical and reproducible process to identify and test a multitude of different proteins and expression systems.

We have achieved this functionality by modularly organizing our lab work. Now, our procedures can be easily adjusted to the respective project by simply exchanging peptide sequences, expression systems or organisms.

Fig 1. | Picking of transformed colonies for verification of transformation through sequencing.

In this sense, we describe the production of lanthanide-binding peptides in E. coli and the subsequent testing of said sequences under bioleaching conditions. One of the most important parts of our lab work has been world’s first cell surface expression of the lanthanide-binding protein Lanmodulin on Saccharomyces cerevisiae, followed by the evaluation of the adsorption capabilities of our novel recombinant yeast colony. We also tested phage display as an effective and reproducible method for the identification of new lanthanide binding-peptide sequences. These sequences could in turn be used to continue the previously mentioned process of protein evaluation.

Production of Lanthanide Binding Proteins in E. coli

For our production of lanthanoid-binding peptides in E. coli, we have selected peptides from scientific literature according to our criteria of selectively binding lanthanoids at low pH as in an acidic bioleaching procedure while exhibiting high reusability.

These peptides are the synthetic peptides Nd01, Nd02 1 and “Most common” (MC) 2 and non-synthetic peptides EF1 and EF4 3 – EF-like hand domains of the Lanmodulin protein found in Methylorubrum extorquens. These peptides were to be produced in E. coli as it is a model organism with widespread common use. Therefore, methods of genetic manipulation and expression systems are well established, which allows for a cost-effective expression with high purity which can easily be scaled up. Also E. coli has well defined growth and expression conditions and can be used safely in a laboratory setting making it a secure and reliable choice for our mostly undergraduate laboratory team.

For efficient peptide purification and protein attachment, we chose a Strep-tag rather than a His-tag due to its diminished propensity for protonation within the lower pH range, as would be in a possibly envisioned bioleaching process 4.

Furthermore, to employ the peptides on a matrix besides a fungal surface – for example through immobilization on a hydrophobic bead – we decided to undertake experimentation on a polystyrene microtiterplate using an LCI tag to immobilize the peptides.

The pET28a-Strep-17x-LCI vector, containing an expression cassette, a kanamycin resistance and both Strep-tag and an LCI-anchor, fused with a 17 amino acid long linker to ensure flexibility and freedom of the fused peptides, was obtained from our supervisor.

Fig 2. | Construct pET28a_mostcomm-Strep-17x (5534 bp). Construct codon optimized for gene expression of peptides in E. coli.

We implemented our genes of interest Nd01, Nd02, MC and EF1 and EF4 through a two-step overlap PCR with primers designed on benchling into the pET28a-Strep-17x-LCI vector to produce five new vectors validating the amplification through gel electrophoresis and sequencing of the PCR product.

Following the chemical induction of competency in E. coli BL21 we were able to transform them to generate five analog strains selected through plating on kanamycin containing plates. The expression of the construct after induction with IPTG was checked and confirmed by SDS-page and western blot. Peptides were purified from large scale expression cultures using Strep-tag affinity chromatography on ÄKTA and resulting fractions of respective peptides prepared for following tests through rebuffering and aliquotation using BCA.

Assessing Peptides for a Process under Bioleaching Conditions

Our project deals with the sustainable recycling of e-waste, with a focus on the recovery of lanthanides. For this we explore a biological process using bioleaching in order to mobilize components of interest from e-waste and to selectively capture these solubilized components with peptides on a matrix. For this we characterize the previously successfully produced peptides Nd01, Nd02, MC, EF1 and EF4 and assess their applicability for an envisioned combined process of bioleaching to mobilize lanthanoids from e-waste into solution and selectively extracting these using peptides.

We consider different aspects of interest for the assessment of usability of the peptides starting already with the ease, scalability and safety of production of the peptides.

Fig 3. | Microtiterplate containing Arsenazo III assay sample solutions.

Central to the evaluation is the binding capability and affinity of the designated peptides to lanthanoids. To measure this, we chose neodymium (in solution as Nd3+) as a stand in for lanthanoids in general for multiple reasons. Testing for multiple (or all) lanthanoids would have been very extensive, proof of concept is sufficient for us, and we also have reason to believe affinity within the periodic group is similar for all lanthanoids, since all enter solution as trivalent cations 5 and have similar atomic radii and affinity of the same peptide for different lanthanoids has already been shown. 6 Also, neodymium has more real-world use and higher demand than some other representatives of lanthanoids 7. We measured neodymium concentration through the already previously used Arsenazo III assay 8,9 which we derived for use on microtiterplates.

For experimental setup, we decided to immobilize the peptides so that the lanthanoid concentration could be measured through without interference of the AS III with peptide ion interaction. We evaluated the immobilization density of our LCI-anchored fusion peptides on polystyrene microtiterplates using HRP-ELISA. Once immobilized, we quantified binding capability and selectivity of the previously produced peptides.

Absorbance at 650 nm as optical density (OD) of peptide supernatant is plotted against original concentration. The neodymium solution applied to the peptides had a starting neodymium concentration of c0(Nd3+) = 25 µM and c0(Nd3+) = 50 µM. 94 µL of this was treated with immobilized peptides (applied to each well were 25 µM *100 µL respectively) Nd01, Nd02, MC, EF1 and EF4. For the calibration curve neodymium solutions of 6,25 µM, 12,5 µM, 25 µM and 50 µM were used.

All absorbance data plotted has been deducted the absorbance of the respective blank. The measurements for the calibration curve were done in duplicates while the measurements for the peptide supernatants were done in triplicates. Affinity of peptides was underestimated so that standard curve was extended by the 0 µM absorbance measurement for the calculation of neodymium concentration of the peptide treated supernatant.

Fig 4. | Absorbance data of microtiterplate containing calibration curve and peptide treated neodymium solution supernatant following Arsenazo III-assay.

The above results show our highest binding achieved with all peptides capturing over 95 % of neodymium from two different sample solutions. Yet for our proof of concept, we aim to show the above observed binding capability by extracting neodymium from a real-world industry e-waste sample.

Fig 5. | Leaching solutions obtained from STENA Recycling e-waste sample.

Before this, we had to answer the question what neodymium concentration to expect from using an industry e-waste sample and if acidic leaching would even be feasible for mobilizing components of interest. For this we wanted to bioleach a sample using acids directly produced by Aspergillus niger or Acidithiobacillus ferrooxidans, yet, due to safety concerns we abandoned this idea. Instead, we conducted acidic leaching on an e-waste sample from industry provided from STENA Recycling and a NdFeB-Magnet at different pHs. Our results show a significant amount and rising tendency of neodymium being mobilized at lower pH.

Selectivity of the peptides is also a factor which we wanted to investigate, since it is essential for peptides to capture the specific component of interest out of very heterogenous solutions resulting from bioleaching. Regarding this we conducted a selectivity experiment using iron, a common component in e-waste and in the oxidizing environment of an acidic bioleaching solution it also goes into solution as a trivalent cation, Fe3+. First, we verified that the AS III assay was not influenced by the iron also in solution. We found that the peptides are highly selective in this context denoting no significant interference in neodymium binding with iron present.

Fig 6. | Raw e-waste solutions during bioleaching process.

Having successfully immobilized the peptides, validated selectivity and produced an authentic e-waste leachate, for our proof of concept, we wanted to recover neodymium from a real world, heterogeneous industry e-waste leachate. For this we chose two reliably performing peptides, again immobilized them and treated them with our previously quantified leachate. Our results show that we were able to recover over 68 % of neodymium from the industry leachate using our experimental setup.

Surface Expression of Lanmodulin

To realistically test the functionality of our envisioned cell surface display system for the isolation of rare earth elements from leaching solutions, we chose to express Lanmodulin (LanM) on Saccharomyces cerevisiae EBY100. LanM is a native lanthanide-binding protein containing four EF-domains capable of specifically binding to lanthanides 10 for the expression on a fungal surface.

The decision to use Saccharomyces cerevisiae as a proof-of-concept organism was made because filamentous fungi are less accessible and require specific selection and testing for particular applications. S. cerevisiae was chosen due to its widespread availability of genetic modification tools, the abundance of relevant strains, and the ability to induce the formation of pseudohyphae 11, which could mimic the functionality of fungal filaments in our project.

Fig 7. | Grown colonies of Saccharomyces cerivisiae EBY100 on plates without tryptophane after successful transformation of pYD1-LanM.

For the surface display, we chose the inducible pYD1 vector system where LanM is fused to a subunit of the α-Agglutinin receptor AGA-2P and linked to AGA-1P by disulfide bridges. AGA-1P is anchored at the yeast cell surface. The fusion protein also contains a linker peptide, an Xpress epitope, a V5 epitope and a His6-tag. The expression is controlled by the promoter GAL1, enabling inducible expression by cultivating with galactose.12

We obtained the gene for LanM from a modified pET19b plasmid from the iGEM Bonn 2021 team. After successful assembly of the pYD1-LanM expression vector and the transformation in S. cerevisiae, the binding capability of the recombinant strain containing pYD1-LanM was tested using an adapted Arsenazo III assay.

Fig 8. | Microscoped yeast colony after growing on minimal agar plates without tryptophane after successful transformation of pYD1-LanM.

To differentiate between surface-binding through LanM and natural biosorption, a negative control only containing pYD1 was used. A defined number of cells were incubated in neodymium solutions of different concentrations for two hours. Afterwards, the cells were separated via centrifugation and the neodymium concentration in the supernatant was determined using a calibration curve.

The minimal variance in absorbance between the supernatant of the negative control and the calibration curve suggests that, at these low cell concentrations, the negative control is incapable of adsorbing substantial quantities of Nd3+.

In contrast, the supernatant of the positive control containing pYD1-LanM exhibited noticeably lower absorbance for each initial concentration. Therefore, it appears that the Lanmodulin surface construct has allowed the mutant cells to bind neodymium ions in significantly larger quantities compared to natural biosorption of cells not expressing LanM despite the low number of cells being present during the assay.

The percentage of bound neodymium ions varied from 24% up to 60 %, the mean value amounted to 40%. For the highest initial concentration of 50 µM, 64 mg neodymium was bound per 1 g of cell dry weight during incubation. It is expected that higher concentrations of neodymium would lead to the mass of bound neodymium to be elevated even further as saturation did not seem to occur yet. However, the calibration curve stops showing linearity for higher concentrations.

In conclusion, the mass of bound neodymium ions exhibits a direct correlation with the initial neodymium concentration and notably varies between cells that express LanM on their surface and those that do not. Mutant cells expressing LanM on their surface are capable of binding higher amounts of Nd than cells only expressing the pYD1 construct.

Fig 9. | Absorbance shown against neodymium concentration including calibration curve (blue) and supernatants after incubation with cells containing pYD1-LanM (orange) and cells containing pYD1 (grey).

Identifying new Binding Sequences with Phage Display

To identify new lanthanide-binding sequences, we performed iterative biopanning experiments using the Ph.D.12 phage library from New England Biolabs. Similar experiments were conducted to identify gallium binding peptides for wastewater treatment.13 Random sequences of 12 amino acids are inserted into the viral envelope proteins. 14 During biopanning experiments phages are exposed to ionized neodymium bound to NTA-Agarose, washed and then eluted chemically or biologically. The phages eluting last indicate neodymium-binding capabilities. Lastly, the amino acid sequences of bound phages are determined by sequencing isolated clones which were beforehand amplified in E. coli.

Fig 10. | Schematic representation of biopanning experiments using phage display, the phages were exposed to neodymium-agarose.

We were able to identify one sequence potentially able to bind neodymium. The relatively low number of new sequences is only due to our limited capacity of sequencing new colonies. We have isolated a great number of additional positive colonies which could be used to gain additional lanthanide-binding sequences, e.g., via next generation sequencing methods.

These new sequences could prove to be a great source of novel lanthanide-binding peptides potentially applicable for rare earth elements recycling. To confirm and characterize their ability to bind to neodymium and other lanthanides effectively, additional experiments that specifically assess their binding capabilities, similar to those showcased in our evaluation of peptides for bioleaching applications, are necessary to be carried out.

Future Outlook

Our project lies the groundstone for decentralized, community-based e-waste recycling. Within that framework, this lab work provided the first tools from synthetic biology for sustainable and efficient extraction procedures for Rare-Earth elements. On the basis of these tools, other groups may now expand on our work to recycle and profit from their own e-waste or even other valuable materials.

We have successfully carried out the experiments for the various stages of our laboratory modules, demonstrating the functionality of all our proof of concepts. Hence, our next step would be the modification of a suitable filamentous fungus and the cultivation of a real mycelial network as filter prototype.

Simultaneously, we would need to continue our protein evaluation process, to identify new lanthanide binding sequences. The first logical step would be the testing of the novel sequences, we were able to identify with our phage display. To develop a fully functional fungal filtration system, it would be crucial to find suitable peptide candidates capable of withstanding prolonged exposure to acidic environments.

In summary, we have achieved all major goals we set for our laboratory work, laying the necessary groundwork for the creation of a unique approach to the recycling of rare earth elements from electronic wastes. Besides this major achievement, we were also able to create an experimental framework for other groups to expand and conduct their own experiments upon.

Fig 11. | In the future: More pipetting.

  1. T. Sawada, M. Asada, T. Serizawa (2016)
    Selective Rare Earth Recovery Employing Filamentous Viruses with Chemically Conjugated Peptides
    ChemistrySelect 2016, 1, 2712
  2. L. E. Rosaleny, S. Cardona-Serra, L. Escalera-Moreno, J. Baldoví, V. Gołȩbiewska, K. Wlazło, P. Casino, H. Prima-García, A. Gaita-Ariño, and E. Coronado (2018)
    Peptides as Versatile Platforms for Quantum Computing
    Chem. Lett. 2018, 9, 16, 4522–4526
  3. S. M. Gutenthaler, S. Tsushima, R. Steudtner, M. Gailer, A. Hoffmann-Röder, B. Drobot, L. J. Daumann (2022)
    Lanmodulin peptides - unravelling the binding of the EF-Hand loop sequences stripped from the structural corset
    Inorg Chem Front. 2022 Jun 30;9(16):4009-4021
  4. M. E. Kimple, A. L. Brill, R. L. Pasker (2018)
    Overview of Affinity Tags for Protein Purification
    Current Protocols in Protein Science Volume 73, Issue 1
  5. N. N. Greenwood, A. Earnshaw (1997)
    Chemistry of the Elements (2nd ed.)
    Butterworth-Heinemann. pp. 1230–1242.
  6. J. A. Cotruvo Jr., E. R. Featherston, J. A. Mattocks, J. V. Ho, T. N. Laremore (2018)
    Lanmodulin: A Highly Selective Lanthanide-Binding Protein from a Lanthanide-Utilizing Bacterium
    . Am. Chem. Soc. 2018, 140, 44, 15056–15061
  7. X. Deng, J. Ge (2020)
    Global wind power development leads to high demand for neodymium praseodymium (NdPr): A scenario analysis based on market and technology development from 2019 to 2040
    Journal of Cleaner Production, Volume 277
  8. C. Hogendoorn, P. Roszczenko-Jasińska, N. C. Martinez-Gomez, J. de Graaff, P. Grassl, A. Pol, H. J. M. Op den Camp, L. J. Daumann (2018)
    Facile Arsenazo III-Based Assay for Monitoring Rare Earth Element Depletion from Cultivation Media for Methanotrophic and Methylotrophic Bacteria
    Applied and Environmental Microbiology
  9. Team Calgary 2021
    iGEM Competition 2021
  10. E. C. Cook, E. R. Featherston, S. A. Showalter, J. A. Cotruvo Jr. (2019)
    Structural Basis for Rare Earth Element Recognition by Methylobacterium extorquens Lanmodulin
    Biochemistry 2019 58 (2), 120-125
  11. J. M. Gancedo (2001)
    Control of pseudohyphae formation in Saccharomyces cerevisiae
    FEMS Microbiology Reviews Volume 25, Issue 1
  12. Invitrogen Life Technologies (2002)
    pYD1 Yeast Display Vector Kit Catalog no. V835-01 Version D
  13. N. Schönberger, C. Taylor, M. Schrader, B. Drobot, S. Matys, F. L. Lederer, K. Pollmann (2021)
    Gallium-binding peptides as a tool for the sustainable treatment of industrial waste streams
    Journal of Hazardous Materials, 414, 125366
  14. NEW ENGLAND BioLabs (2022)
    Ph.D.™ Phage Display Libraries, NEB #E8100S, #E8101S, #E8102L, #E8110S, #E8111L, #E8120S, #E8121L