Bioadsorption of Rare Earth Elements through Cell Surface Display of Lanthanide Binding Tags

Abstract

With the increasing demand for rare earth elements (REEs) in many emerging clean energy technologies, there is an urgent need for the development of new approaches for efficient REE extraction and recovery. As a step toward this goal, we genetically engineered the aerobic bacterium Caulobacter crescentus for REE adsorption through high-density cell surface display of lanthanide binding tags (LBTs) on its S-layer. The LBT-displayed strains exhibited enhanced adsorption of REEs compared to cells lacking LBT, high specificity for REEs, and an adsorption preference for REEs with small atomic radii. Adsorbed Tb(3+) could be effectively recovered using citrate, consistent with thermodynamic speciation calculations that predicted strong complexation of Tb(3+) by citrate. No reduction in Tb(3+) adsorption capacity was observed following citrate elution, enabling consecutive adsorption/desorption cycles. The LBT-displayed strain was effective for extracting REEs from the acid leachate of core samples collected at a prospective rare earth mine. Our collective results demonstrate a rapid, efficient, and reversible process for REE adsorption with potential industrial application for REE enrichment and separation.

Figure 1. Bioengineering of Caulobacter crescentus for REE adsorption

(A). Diagram of engineered S-layer gene (rsaA) constructs with dLBT insertions. A muc1B spacer, encoding the human mucin protein, was appended to the C-terminal end of dLBT. The copy number of the resulting dLBT-mucR1 peptide was increased exponentially. The number labels of the constructs correspond to the lanes described in (B). (B) SDS-PAGE of S-layer extracted from the following strains: 1) wild type CB2A, 2) CB2A rsaA (control), 3) dLBTx1, 4) dLBTx2, 5) dLBTx4 and 6) dLBTx8. dLBTx4 and dLBTx8 cells were grown in PYE medium supplemented with additional Ca²⁺ (2.5 μM) and Ca²⁺ (2.5 μM) with trace metals, respectively. MW; molecular weight (kDa) markers. Arrows on the right indicate the engineered RsaA protein expressed from each strain.

Figure 2. Tb³⁺ adsorption to LBT-displayed cells

(A) Tb³⁺ titration of dLBT constructs with no added Ca²⁺, measured by luminescence (ex/em 280/544). ICP-MS quantitation of Tb³⁺ adsorption by dLBTx4 and control cells at increasing Tb³⁺ concentration in the absence (B) or presence (C) of 100 mM Ca²⁺. The adsorption contribution of LBT in dLBTx4 was approximated by subtracting the total adsorbed Tb³⁺ by the control strain from that adsorbed by dLBTx4, yielding 12.9 ± 4.6 μM Tb³⁺. The uncertainty in this expression was determined using error propagation. (D) Tb³⁺ titration of dLBT constructs with 100 mM Ca²⁺. The data from (A) were plotted as dotted lines for comparison. (E) ICP-MS quantitation of Tb³⁺ (10 μM added) adsorption by dLBTx4 and control cells at different Ca²⁺ concentrations. (F) ICP-MS quantitation of Tb³⁺ (10 μM added) adsorption by dLBTx4 and control cells within the pH range of 4–6 in the presence of 100 mM Ca²⁺.

Figure 3. REE adsorption specificity

(A) Competition binding experiments with dLBTx4 cells preloaded with 10 μM Tb³⁺ followed by addition of various metal ions at concentrations up to 10 mM. Normalized luminescence was calculated as described in methods. (B) Tb³⁺ adsorption to dLBTx4 and control cells at increasing Cu²⁺ concentrations. The fraction of Tb³⁺ adsorbed was determined by quantifying the soluble Tb³⁺ concentrations before and after incubation with cells using ICP-MS. (C) Competition experiments with dLBTx4 cells preloaded with 10 μM Tb³⁺ followed by addition of REE ions up to 352 μM. Normalized luminescence was calculated as described in methods. (D) Tb³⁺ and La³⁺ (20 μM each) adsorption to dLBTx4 and control cells. The fraction of REE bound was determined by ICP-MS. All experiments were performed in the presence of 150 mM Ca²⁺. Error bars represent standard deviations of three replicates.

Figure 4. Citrate-mediated REE desorption

(A) REE desorption and recovery were performed by pre-loading dLBTx4 cells with 10 μM Tb³⁺ in the presence or absence of 150 mM Ca²⁺ followed by the addition of increasing concentrations of citrate, gluconate or acetate. Normalized luminescence was calculated as described in methods. (B) Three cycles of Tb³⁺ adsorption and desorption were performed with citrate (5 mM) in the presence of 150 mM Ca²⁺. Gray bars depict the normalized luminescence signal for Tb³⁺ loading and blue bars depict the fraction of Tb³⁺ eluted using 5 mM citrate during each cycle as quantified by ICP-MS. (C) and (D) The predicted fraction of Tb³⁺ that was not complexed with acetate or citrate, respectively, using the thermodynamic model. Results are shown for pH 5 and 6.1 within the range of acetate and citrate concentrations used in (A). Note that the concentration scale in Fig. 4D is expanded to focus on the rapid decline in uncomplexed Tb at low citrate concentrations. The individual Tb³⁺ species present in the aqueous solution in the presence of acetate or citrate are shown in Figure S3 or Figure S4, respectively.