Shewanella oneidensis JG3631 strain was obtained from Jeff Gralnick’s lab (University of Minnesota, Minneapolis, MN, United States). The strain has been engineered to express the multi-heme cytochrome complex MtrCAB under control of a native promoter PtorF that responds to inducer molecule trimethylamine N-oxide (TMAO). Additional information about the strain is available in Supplementary Figure S1. Previously, strain JG3631 reduced iron oxide in proportion to the concentration of TMAO inducer added to the culture. The MtrCAB expression level plateaued at 1,000 μM TMAO (West et al., 2017). Here, we induce cells with 0, 50, 100, and 1,000 μM TMAO. Shewanella oneidensis JG1486 was used for control experiments reported in the Supplementary Information, containing deletions of mtrB, mtrE, mtrC, mtrF, mtrA, mtrD, omcA, dmsE, SO4360, cctA, and recA (Coursolle and Gralnick, 2012).

Cultures of Shewanella oneidensis JG3631 were inoculated from a bacterial frozen stock into Luria-Bertani medium and grown overnight (14–16 h) at 30°C under aerobic conditions. Cultures were then transferred to Shewanella minimal media prepared from the recipe from Bretschger et al. (2007) with 15 mM lactate as electron donor, 30 mM fumarate as electron acceptor, and TMAO (inducer) at 0, 50, 100, or 1,000 μM. Cultures were grown under anaerobic conditions. After 24 h, the cells grown in minimal medium were centrifuged, washed with 7 mM HEPES buffer, and suspended in 7 mM HEPES buffer to a final OD_600nm of 0.8–1.0. Control experiments showed that more dilute cell cultures were also capable of forming ZnS:Mn(II) particles (Supplementary Figure S2). A stoichiometric excess of lactate (10 mM) was used as electron donor in the culture, and the culture was made anaerobic by bubbling sterile nitrogen gas into the bottle. This culture was then used in the experiments for nanomaterial synthesis. Nanomaterial synthesis experiments were performed under anaerobic conditions.

Solid manganese (IV) oxide was prepared using the protocol described in an earlier work (Burdige and Nealson, 1985). Prepared manganese (IV) oxide was mixed with HEPES buffer and injected into anaerobic bacterial cultures described above to a final concentration of 750 μM manganese with 10 mM lactate was the electron source. After 24 h of manganese reduction by the bacteria, a filter sterilized stock solution of 2.5 mM zinc sulfate was added to the culture followed by 2.5 mM sodium sulfide. Extended manganese reduction, for 48 h total, did not result in additional manganese reduction or change the photoluminescence of the resulting particles (Supplementary Figure S3). Manganese reduction occurred at 303°K, and samples were moved to room temperature (approximately 295°K) after the addition zinc and sulfide for the remainder of the synthesis reaction. Samples were thoroughly mixed via vortex. The precipitation of nanomaterials started immediately and proceeded for 16 h. Continued manganese reduction was not detected after the addition of zinc sulfate and sodium sulfide, as shown in Supplementary Figure S4. As shown in Supplementary Figure S5, cell viability was maintained throughout the manganese reduction step, however, no live cells were detected 16 h after the addition of zinc sulfate and sodium sulfide.

Chemical synthesis of Mn:ZnS nanomaterial was accomplished by adding precursors to a sterilized, anaerobic serum bottle containing 25 ml 7 mM HEPES buffer. Sterile nitrogen gas was bubbled through the buffer solution and precursor stock solutions to make them anaerobic. To synthesize Mn:ZnS, first 2.5 mM zinc sulfate was added followed by different manganese acetate at a concentration of either 0, 0.1, 0.5, 1, or 5 mM. Finally, 2.5 mM sodium sulfide was added and the bottle was thoroughly mixed using a vortex. Chemical synthesis was performed at room temperature. The addition of Shewanella oneidensis MR-1 cells during chemical synthesis did not appear to impact photoluminescent properties of the chemically synthesized cells, see Supplementary Figure S6.

The reduction of manganese by Shewanella oneidensis JG3631 was quantified using the leucoberbelin blue (LBB) assay (Francis et al., 2002). Five hundred microliters was collected directly from well mixed anoxic serum bottles using a sterile, 20G syringe. Sample was added to LBB [0.04% (w/v) LBB in 45 mM acetic acid] to react in the dark for 15–20 min, and then centrifuged to separate the cellular material and insoluble fractions. A standard curve for concentration of was made by preparing serial dilutions of KMnO₄ and measuring the absorbance at 620 nm to quantify the concentration of Mn. To calculate the amount of Mn(II) at a given time, we subtract the initial amount of Mn(IV) from the amount of Mn(IV) remaining in solution. A calibration curve was made using solutions of potassium permanganate, see Supplementary Figure S7.

Upon the completion of nanomaterial synthesis, the contents of the bottle were transferred to a 50 ml conical tube for rinsing and cleaning. Nanomaterial solutions were centrifuged at 3,800 ×g for 30 min to collect the nanomaterials, the supernatant was discarded, and the nanomaterial pellet was re-suspended in DI water. This washing step was repeated four times to remove salts and cellular materials from the solution of nanomaterials. In samples where we observed excess aggregation, the final solution of nanomaterials was sonicated in an ice bath for 20 min prior to AFM and SEM.

By utilizing a genetically engineered bacteria, we designed a biological system to synthesize semi-conductive ZnS nanomaterials doped with Mn(II). Moreover, the degree to which the gene circuit responds to an outside signal modulates the concentration of available Mn(II) for doping in ZnS, thereby adjusting the optoelectronic properties of these biogenically fabricated nanomaterials. Estimated manganese concentrations using EPMA are presented in Supplementary Table S1. These nanomaterials of biogenic origin exhibit almost identical properties to those made by non-biological, chemical methods. To begin our study we synthesized ZnS nanoparticles via chemical means. In bulk, these particles exhibited a characteristic blue emission upon excitation with UV light (Figure 1A). Next, we introduced variable concentrations of Mn(II) during chemical synthesis and the soluble Mn(II) was passively integrated into the ZnS nanoparticles. These doped nanoparticles exhibited a characteristic orange hue upon excitation with UV light (Figure 1A) (Beerman, 2005; Cao et al., 2009; Deng et al., 2011). The photoluminescent intensity of the Mn(II) doped nanoparticles depended on the level of doping.

To control the optical properties of Mn:ZnS nanoparticles through biogenic route, we used an engineered strain (JG3631) of Shewanella oneidensis MR-1 because of its metabolic versatility, whole genome sequence availability, and a library of characterized, engineered strains (Bouhenni et al., 2005; Nealson, 2005; Bretschger et al., 2007; Fredrickson et al., 2008). Shewanella naturally respire insoluble metal oxides of iron and manganese via extracellular electron transport protein complex MtrCAB, a multiheme cytochrome complex. This protein complex moves electrons from the periplasmic space to the exterior of the cell during respiration (Myers and Myers, 2001; Nealson et al., 2002; Bretschger et al., 2007). Expressed under anaerobic conditions, the Mtr pathway is composed of three components: MtrA, a periplasmic decaheme c-cytochrome; MtrB, an outer membrane porin; and MtrC, an outer membrane decaheme c-type cytochrome. Here, we utilize an engineered strain JG3631 in which an inducible promotor PtorF regulates expression of the mtrCAB operon (Figure 1B) (West et al., 2017). Previously it was shown that this strain reduced external iron oxide in proportion to the amount of the inducer (TMAO) added to the culture (West et al., 2017).

First, we tested the ability of the strain JG3631 to reduce manganese in the presence of different TMAO concentrations. As shown in Figure 1C, the amount of Mn(IV) reduced by the cell culture was proportional to the concentration of inducer molecule TMAO. However, at 0 mM TMAO, some manganese reduction was observed, potentially due to a combination of leaky expression of mtrCAB and additional biochemical pathways involved in low levels of manganese reduction, such as MtrDEF cytochrome (Coursolle and Gralnick, 2010). Additional information on the engineered strain and the experimental outline is presented in the Supplementary Figure S1.

After confirming that we could control manganese reduction, and therefore the available concentration of Mn(II) via the concentration of TMAO inducer molecule, we investigated the biogenic synthesis of Mn:ZnS. Because Shewanella prefer to express metal reduction protein complexes in the absence of oxygen, all biogenic reactions took place under anaerobic conditions, as outlined in the “Materials and Methods” section. The nanoparticles produced without Mn(IV) added to the culture appear blue under UV excitation (Figure 1D, left). By adding TMAO, however, the MtrCAB pathway is modulated in a manner reflecting the concentration of inducer molecule (Figure 1C), i.e., more TMAO results in more insoluble Mn(IV) being reduced to soluble Mn(II). As expected, nanoparticles produced in the presence of both Mn(IV) and TMAO exhibit a characteristic red-shifted emission upon UV excitation, similar to chemical synthesis. The buffer solutions and the precursor solutions did not exhibit any UV associated luminescence (Supplementary Figure S8). Altogether, by connecting a naturally occurring metal reduction route in an engineered strain of Shewanella with an inducible promotor, we made possible the controllable synthesis of nanoparticles with tunable optoelectronic properties.

Following chemical and biogenic synthesis, we assessed the optoelectronic properties (e.g., photoluminescence and absorbance) of the Mn:ZnS. Both methods yielded nanomaterials with the characteristic 600 nm photoluminescence (PL) peak associated with the presence of dopant metal Mn(II) in a ZnS lattice. The shift in PL emission due to Mn(II) doping into the ZnS nanomaterial is caused by an additional electronic transition between excited electrons and the energy levels of the Mn dopant (Bhargava et al., 1994; Karar et al., 2004). Chemically synthesized nanomaterials produced PL peaks in the 602–604 nm range and the biogenic nanomaterials produced peaks in the 604–608 nm range (Figure 2A,B). The difference in the PL peak emission between the two methods may be a result of biogenic moieties altering the emission spectrum. Earlier work showed that PL emission wavelength of ZnS nanoparticles varied according to the capping agent used during the reaction (Warad et al., 2005; Wanjari et al., 2015).

Buffers, and precursor solutions did not produce any absorbance or PL emission peaks (Supplementary Figure S8). Manganese may react with sulfide to produce manganese sulfide (MnS) precipitate. Some groups have observed PL emission on MnS films near 640 nm, which could explain the slight red shift in the PL peak with increasing Mn(II) concentration (Goede et al., 1986). We do not observe any PL optical signature of MnS nanoparticles in this region (Supplementary Figure S8). This finding leaves Mn:ZnS as the only candidate emitter at 600 nm during PL characterization. Next, we examined the effect of biologically regulated Mn(II) concentration on the PL emission of the nanoparticles. In the chemical method, the PL intensity increased up to 2 mM Mn(II) and subsequently decreased. In the biogenic method, the PL intensity increased up to 100 μM TMAO induction and then decreased with increasing concentrations of TMAO. Recall that TMAO concentrations dictates the concentrations of soluble Mn(II). This observed effect of dopant concentrations on PL emission peak intensity — an increase followed by a decrease upon reaching a critical concentration (Figure 2C,D) — is consistent with previous results (Son et al., 2007; Cao et al., 2009). This effect, known as luminesce quenching, is a common feature among doped semi-conductors, of which Mn(II) doped ZnS is a classic example (Hurd and King, 1979; Sasakura et al., 1981; Katiyar and Kitai, 1992). Briefly, the probability of an electron indirectly transitioning through a Mn(II) site to the ground state increases with Mn concentration up to 1–4 wt%. As the Mn concentration increases beyond a few percent by weight, however, unfavorable interactions between adjacent Mn(II) sites, interruptions in the ZnS crystallinity, an increase in non-radiative transition processes, and a decrease the in the excitation efficiency from the ZnS lattice diminish the effect of doping. Finally, it is worth noting that where the Mn(II) ions subsume with the nanomaterial dramatically affects its optoelectronic properties. It is known that incorporating Mn(II) on the surface of ZnS quantum dots instead of the bulk results in an ultraviolet PL emission. That our nanomaterial emits in the visible region near 600 nm suggests that the majority of Mn(II) is embedded within the ZnS structure rather than the surface (Xiao and Xiao, 2008). The PL emission intensity was generally higher in chemically synthesized particles, and it is currently unclear if synthesis in the presence of cellular compounds impacts the intensity of the PL emission. Overall, the PL properties of the chemically and biogenically synthesized Mn:ZnS nanomaterials are similar to each other and with literature precedent.

Next, we measured the absorption spectrum for biogenically and chemically synthesized nanomaterials in the 300–700 nm wavelength windows. Figure 2E,F show the absorption spectrum for the nanoparticles is in the UV wavelength regions of 310–320 nm, which agrees with earlier work (Kole and Kumbhakar, 2012). Chemically synthesized nanoparticles had an absorption peak of 310 nm and the biogenic nanoparticles had an absorption peak of 315 nm. From the absorbance, one may find the band gap of the nanomaterials. The band gap energies of nanomaterials made by chemical and biogenic methods are 4.0 eV and 3.9 eV respectively. These values are higher than for bulk ZnS material, which is 3.7 eV. An inverse relationship exists between the size of the nanoparticle and its bandgap due to quantum mechanical confinement, i.e., the system resembles a particle-in-a-box. The bandgap is the energetic difference between the ground state and excited state (the bands) of the electron in an atom or bulk material. A single atom possesses a large bandgap because the allowed electron states are precisely defined. That is, the band is extremely narrow and the gap between bands (the bandgap) is large. A continuous, bulk material, however, is composed of many, overlapping electron orbitals. The effect of overlapping increases the band width with a concomitant decrease between the ground and excited state of electrons in the material. As a result, the band is wider and the gap between bands decreases (Son et al., 2007; Begum and Chattopadhyay, 2014; Marusak et al., 2016).

Next, we characterized the crystalline structure of Mn:ZnS synthesized by the chemical and biogenic methods via XRD, an X-ray scattering technique which measures coherent diffraction from crystalline domains within nanomaterials across an entire sample. The XRD pattern indicated there were three distinct diffraction peaks (28.7°, 47.9°, and 56.7°) with 2𝜃 values corresponding to three planes (111), (220), and (311). These peaks confirm that the synthesized nanomaterial has a cubic phase of zinc blende consistent with previous reports and zinc sulfide standard (Supplementary Figure S9). By analyzing the XRD data using the Scherrer model (Supplementary Equation S1), one may extract the size of the crystalline domains within the nanomaterials. Briefly, the width of the diffraction peak is inversely proportional to the size of the crystalline domain. For both the chemically and biogenically synthesized nanomaterials, the crystalline domain is between 5 and 8 nm in diameter (Supplementary Table S2), slightly smaller than the particle size derived from optical methods. This is not unexpected, as the geometry and crystallinity of nanoparticles may differ due to the presence of amorphous, X-ray diffusive regions, which XRD cannot detect. Finally, as further evidence that ZnS and Mn:ZnS made the bulk of the nanomaterial, no peaks associated with MnS crystals appear in the XRD pattern of our purified nanomaterials (Figure 3 and Supplementary Figure S9). To determine if the presence of Mn(IV) during synthesis influenced nanoparticle properties, Mn(IV) oxide was added during chemical synthesis. The addition of Mn(IV) oxide did not inhibit the formation of ZnS and ZnS:Mn(II) nanoparticles, see Supplementary Figure S9B. Also, Mn(IV) oxide alone was not sufficient for manganese doping during chemical synthesis, see Supplementary Figure S10. These results are in line with earlier work and show good agreement between the chemical and biogenic route (Son et al., 2007).

Next, we leveraged energy-dispersive X-ray spectroscopy (EDX) measurements to confirm the elemental composition of the nanomaterial. In an EDX experiment, X-ray excitation induces emission spectra unique to specific atomic nuclei in the material, thereby furnishing information about the specific elemental make-up of the sample. The peaks of zinc and sulfur indicate the material formed was zinc sulfide (Supplementary Figures S11, S12) in agreement with the XRD data. The primary peaks in these data correspond to Si and O and likely arise from the sample substrate, a Si₂ wafer. The secondary peaks, however, correspond to Zn, S, and Mn (in doped samples). It is worth noting that the peaks are unlikely from left over substrates as the samples were thoroughly washed to remove the initial reactants used for nanomaterial synthesis.

Next, we used scanning electron microscopy (SEM) and AFM to identify the size and shape of the synthesized ZnS and Mn:ZnS nanomaterials. SEM images showed that particles were quasi-spherical (Figure 4A) in good agreement with earlier work (Warad et al., 2005; Chandrakar et al., 2015). SEM analysis was used for identifying the morphology of the nanoparticles, while AFM was used to quantify the size of the nanoparticles.

We measured the size of the ZnS and Mn:ZnS nanomaterials using AFM. AFM size analysis (Figure 4B,C) reveals that the size of Mn:ZnS nanomaterials synthesized via chemical method was slightly larger and polydisperse compared to nanomaterials synthesized via biogenic method. Researchers have measured the size of chemically synthesized Mn:ZnS nanomaterials in the range of 2–70 nm (Warad et al., 2005; Cao et al., 2009; Kole and Kumbhakar, 2012; Komada et al., 2012). The mean value for the chemically synthesized particles is 8.7 nm [0 mM Mn(II)], 11.5 nm [1 mM Mn(II)] and the mean value for the biogenically synthesized particles is 4.0 nm (0 TMAO, no induction) and 5.0 nm (1 mM TMAO, full induction). The difference in size and polydisperity may result from different reaction kinetics between the chemical and biogenic methods, an advantage in nanomaterial sysnthesis that merits further investigating. Altogether, EDX, SEM, and AFM reveal that the nanomaterials synthesized via chemical and biogenic routes were pure, spherical, and approximately 2–20 nm.