Scalable and Consolidated Microbial Platform for Rare Earth Element Leaching and Recovery from Waste Sources

Chemical methods for the extraction and refinement of technologically critical rare earth elements (REEs) are energy-intensive, hazardous, and environmentally destructive. Current biobased extraction systems rely on extremophilic organisms and generate many of the same detrimental effects as chemical methodologies. The mesophilic methylotrophic bacterium Methylobacterium extorquens AM1 was previously shown to grow using electronic waste by naturally acquiring REEs to power methanol metabolism. Here we show that growth using electronic waste as a sole REE source is scalable up to 10 L with consistent metal yields without the use of harsh acids or high temperatures. The addition of organic acids increases REE leaching in a nonspecific manner. REE-specific bioleaching can be engineered through the overproduction of REE-binding ligands (called lanthanophores) and pyrroloquinoline quinone. REE bioaccumulation increases with the leachate concentration and is highly specific. REEs are stored intracellularly in polyphosphate granules, and genetic engineering to eliminate exopolyphosphatase activity increases metal accumulation, confirming the link between phosphate metabolism and biological REE use. Finally, we report the innate ability of M. extorquens to grow using other complex REE sources, including pulverized smartphones, demonstrating the flexibility and potential for use as a recovery platform for these critical metals.

DNA manipulation, Molecular Cloning, and Mutagenesis.The Δppx mutant strain was generated using the counter selection marker sacB.The donor plasmid was constructed as follows: ~800 bp regions of genomic DNA flanking the ppx gene (META1p2050) were amplified using primers designed with 20 bp overlaps for homology-based assembly as previously reported [46].Linearized pCM433KanT was also produced by PCR with 20 bp overlaps for the 5' and 3' flanking regions.The final construct was assembled as previously described [40], Sanger sequenced for verification and transformed into M. extorquens AM1.Counterselection was conducted with 5% sucrose as reported [47].The deletion was confirmed via PCR and Sanger sequencing.
To prepare plasmid pAZ1 for overproduction of the methylolanthanin biosynthetic gene cluster (mll), a yeast in vivo DNA assembly strategy was used based on the DNA assembler method. 1 First, the genes META1p4132 through META1p4133 and META1p4134 through META1p4138 were PCR-amplified as two 4.5 kb fragments with 400 bp of overlap on each end from M. extorquens AM1 gDNA isolated using the DNeasy PowerMax Soil Kit (Qiagen, Germantown, MD, USA).Two expression elements were obtained using pRES (unpublished) as template: a 4.0 kb E. coli helper fragment (derived from pCC1Fos (Genbank accession EU140751, Epicentre Biotechnologies, Madison, WI, USA) and a 3.3 kb S. cerevisiae helper fragment (derived from pYJKSD,(60)).A third fragment (5.0 kb) to integrate elements of expression for M. extorquens AM1 was obtained by using pCM66T as a template.To prevent competition with the E. coli backbone, the colE locus of pCM66T was removed with OE-PCR through amplification of the components up and downstream of colE locus.Four different 800 bp joint fragments with 400 bp overlaps with each main fragment were obtained by amplifying expression elements from the E. coli fragment, the S. cerevisiae fragment, the mll gene fragments, and the M. extorquens AM1 fragment.Following electrophoresis, PCR products were purified from a 1% agarose gel using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, Waltham, MA, USA).150 ng of each PCR product was combined, dried under N2, and the final mixture was resuspended in 4 μL of Milli-Q double-deionized water.Electrocompetent, uracil auxotrophic S. cerevisiae HZ848 were freshly prepared and transformed with this mixture and spread on synthetic complete medium minus uracil (SC-ura) plates to select for homologous recombination of the DNA mixture.Eight prototrophic colonies were grown in liquid SC-ura and lysed using the Zymoprep Yeast Plasmid MiniPrep II Kit (Zymo Research, Irvine, CA, USA).
The plasmid was purified from lysate using the Zymo Research BAC DNA Miniprep Kit (Zymo Research, Irvine, CA,USA) according to the manufacturer's protocol.The plasmid was then transformed into electrocompetent TransforMax EPI300TM E. coli (Lucigen, Middleton, WI, USA), and plated on LB with chloramphenicol for selection.Resultant colonies were grown in liquid SOB media overnight followed by induction through passaging into fresh media with chloramphenicol and CopyControl Induction Solution (Lucigen, Middleton, WI, USA).After five hours of growth, the plasmid was purified using BAC DNA Miniprep Kit (Zymo Research, Irvine, CA) and verified via Sanger sequencing (UC Berkeley DNA Sequencing Facility, Berkeley, CA, USA).The plasmid was then used in the electroporation of ΔmxaF to generate strain ΔmxaF/pAZ1.

Processing of smartphones.
Postconsumer ZTE Quest N817 and Nokia 6136 smartphones including lithium ion batteries were purchased from eBay.Phones were pulverized using a Blendtec Total Blender (Blendtec, Orum, UT) in pulses of ~20 seconds for 5 to 10 min.The resulting debris was sifted through a set of geological sieves (American Geo, Poughkeepsie, NY).Fragments below 0.15 mm were retained and autoclaved.
Growth analysis with ores, minerals, and smartphones.For growth with ores and minerals, cultures of M. extorquens AM1 were grown overnight in MP succinate medium without exogenous lanthanides.When cultures reached mid-log phase (~OD 0.8), a 300 µL aliquot was sub-cultured to 10 mL of fresh MP methanol supplemented with 0.5% ore.As controls, 10 mL MP methanol with ores and without or with the addition of 2 µM lanthanum chloride (LaCl 3 ) was inoculated with 300 µL bacterial aliquots or with MP methanol lacking cells.Every 3 hours, a 200 µL aliquot of culture was removed and OD 600 was recorded using a Spectramax M2 plate reader (Molecular Devices, Sunnyvale, CA).The no cell control containing the respective ore was used as the blank.For growth with smartphones, 2 mL of MP succinate culture was inoculated into 100 mL MP methanol media with and without 0.5% smartphone powder and LaCl 3 in 250 mL shake flasks.Growth was measured by serial dilution and CFU analysis.Calcium is normally added to Hypho minimal media as an essential metal for MxaF methanol dehydrogenase.However, the strains of M. extorquens AM1 used in this study do not have mxaF and rely on REE for methanol dehydrogenase activity via XoxF and ExaF.We tested if removing calcium from the growth medium would impact REE uptake in M. extorquens AM1 but observed no significant differences (Figure S2).

Flexibility of M. extorquens for REE bioleaching and bioaccumulation from complex sources
Robust growth and bioaccumulation of Nd from magnet swarf indicated that M. extorquens AM1 can effectively leach and acquire Ln from poorly soluble sources.Therefore, we tested the ability of M. extorquens AM1 to grow with unrefined REE sources with very low solubility in the growth medium.Monazite ore, for example, contains up to ~45% REEs, with the Ce 2 O 3 content comprising as much as 17% of the total REEs.However, the REEs occur as poorly soluble phosphates and fluorocarbonates in these ores.When challenged with the ores monazite or bastnäsite as the sole REE source, M. extorquens AM1 grew as well as with soluble LaCl 3 (Table 2).Similar growth results were also observed with bastnäsite crystal (Table S2).
Complex debris like E-waste represents virtually untapped reservoirs of valuable REEs.
Cellular smartphones contain several species of REE, including yttrium, lanthanum, terbium, neodymium, gadolinium and praseodymium, making smartphone E-waste a valuable, unutilized source for REE recovery.However, smartphone E-waste poses two major challenges for bioleaching.The metals in this E-waste are highly insoluble in oxide form, and smartphone batteries contain other metals (e.g., iron, boron, copper, tellurium, manganese, mercury, lead, tungsten, lithium, cobalt), all of which can be toxic.We tested the feasibility of using smartphone E-waste as a REE source by assessing the capacity of M. extorquens AM1 to grow on methanol.
Somewhat surprisingly, with 0.5% pulp density (w/v) of blended smartphones, M. extorquens AM1 grew as well as with soluble LaCl 3 (Table S2).In fact, along with the ores and crystal REE sources, the insolubility of the metals and complexity of the source did not significantly impact growth rate or growth yield of the culture (Table S2) showing the flexibility and robustness of M.
extorquens AM1 as a biological platform for the bioaccumulation of REE from complex sources.

Figure S1 .
Figure S1.Impact of inorganic phosphate on strain performance.A, Growth rate of M.

Figure S2 .
Figure S2.Effect of calcium in growth medium on Nd uptake.M. extorquens AM1 ΔmxaF and

Figure S5 .
Figure S5.Bioleaching and bioaccumulation during growth with organic acids.Amount of Fe,

Figure S6 .
Figure S6.Effect of citrate on bioleaching and bioaccumulation.(A) Growth of M. extorquens

Figure S7 .
Figure S7.Selective bioconcentration of REE from culture grown in a 0.75 L benchtop

Table S2 .
Growth of M. extorquens AM1 with pure and complex REE sources