Biomanufacturing Biotinylated Magnetic Nanomaterial via Construction and Fermentation of Genetically Engineered Magnetotactic Bacteria

Biosynthesis provides a critical way to deal with global sustainability issues and has recently drawn increased attention. However, modifying biosynthesized magnetic nanoparticles by extraction is challenging, limiting its applications. Magnetotactic bacteria (MTB) synthesize single-domain magnetite nanocrystals in their organelles, magnetosomes (BMPs), which are excellent biomaterials that can be biologically modified by genetic engineering. Therefore, this study successfully constructed in vivo biotinylated BMPs in the MTB Magnetospirillum gryphiswaldense by fusing biotin carboxyl carrier protein (BCCP) with membrane protein MamF of BMPs. The engineered strain (MSR−∆F−BF) grew well and synthesized small-sized (20 ± 4.5 nm) BMPs and were cultured in a 42 L fermenter; the yield (dry weight) of cells and BMPs reached 8.14 g/L and 134.44 mg/L, respectively, approximately three-fold more than previously reported engineered strains and BMPs. The genetically engineered BMPs (BMP−∆F−BF) were successfully linked with streptavidin or streptavidin-labelled horseradish peroxidase and displayed better storage stability compared with chemically constructed biotinylated BMPs. This study systematically demonstrated the biosynthesis of engineered magnetic nanoparticles, including its construction, characterization, and production and detection based on MTB. Our findings provide insights into biomanufacturing multiple functional magnetic nanomaterials.


Introduction
The heightened awareness of actively mitigating climate change, reducing pollution, and improving food security has driven the need for biomanufacturing. Biosynthesis provides a key platform for bioproduction [1], including green synthesis of biodiesel [2][3][4], ethanol [5,6], organic acids [7,8], and biopolymers [9]. Compared with those, fossil-based to bio-based material production represents a more significant challenge in biosynthesis technology and high-volume output [1].
Among various materials, magnetic nanomaterials display a higher specific area and increased magnetic, optical, and electrical stability and are widely applied in imaging [10,11], cancer treatment [12], biosensors [13], gene delivery [14], and agriculture sectors [15,16]. Traditional chemical synthesis of Fe 3 O 4 nanoparticles (NPs) is performed by precipitating Fe 2+ /Fe 3+ [17,18]. Organic solvents (e.g., methacrylic acid, esters, oleyl alcohol, and oleic acid) were used as stabilizers or surfactants in the manufacturing process, possessing severe adverse environmental effects and eventually causing harm to human health. Compared with traditional synthesis, green synthesis of nanomaterials using different biological entities, such as plant debris, and microorganism cells, overcomes many destructive effects
The cell growth and magnetic response of MSR−∆F−BF were measured, and wildtype MSR-1 and MSR−∆F were used as controls. The cell density (OD 565 ) and magnetic response (Cmag) were measured every 2 h using a UV-visible spectrophotometer in the bacteria culture process (UNICO2100; UNICO Instrument Co., Shanghai, China). MSR−∆F−BF cell growth and Cmag curves were constructed to detect the status of strain growth and magnetosome synthesis. The recombinant plasmids were constructed as shown in Figure 1A. The mamF and bccp genes were amplified from the genomic DNA of MSR-1 using corresponding primers. The initiation codon of mamF and termination codon of bccp were removed during amplification. mamF and bccp were then fused into bccp-mamF using fusion PCR, with a hydrophilic polypeptide linker (Gly 4 Ser) 3 between fusion genes. bccp-mamF was cloned into a broad host-vector pBBR1MCS-2 to construct the recombinant plasmids pBBR−bccp−mamF. The pBBR−bccp−mamF was transformed into MSR−∆F by biparental conjugation to construct a recombinant strain, MSR−∆F−BF. Selecting MSR−∆F as the host strain can increase the amount of biotin displayed on engineered BMPs. The authigenic biotin ligase link biotin to a lysine residue of biotin carboxyl carrier protein (BCCP) displayed on BMP provided a recombinant magnetosome, BMP−∆F−BF. The MSR−∆F native mamF gene in the genome was replaced by the gentamycin-resistance gene. To detect the expression of fusion genes bccp−mamF, the gfp−marked expression plasmids pBBR−bccp−mamF-gfp were constructed. Western blotting was performed to evaluate the green fluorescent protein (GFP) expression level.

Transmission Electron Microscope (TEM) Observation
A TEM was used to observe the morphology of cell and magnetosomes of MSR−∆F−BF, MSR−WT, and MSR−∆F. The strains were cultured in LAY media for 24 h at 30 • C and collected by centrifugation at 12,000 rpm for 2 min at 4 • C. Pellets were rinsed three times with ddH 2 O. Cultured strains (1 mL) were suspended in 50 µL ddH 2 O. Then, 10 µL of the suspension was dropped onto a copper mesh for 10 min, dried for 30 min using infrared light, and observed under a TEM (model JEM-1230, JEOL; Tokyo, Japan).
Structural details of the magnetosomes in MSR−∆F−BF were determined using a high-resolution transmission electron microscope (HRTEM) at ESPCI (Paris, France) on a JEOL 2010F microscope at 200 kV, equipped with a Gatan cooled CCD camera (1024 × 1024 pixels). For each particle, the structural parameters of the crystals were determined from the two-dimensional fast Fourier transform of the HRTEM image. The magnetic properties of BMP−∆F−BF were determined using a vibrating sample magnetometer (Model 3900; Princeton Measurements Corporation, Princeton, USA, sensitivity 5.0 × 10 -10 Am 2 ) from the Paleomagnetism and Geochronology Laboratory. . Cell and magnetosome yields were calculated, as described previously [44]. The inoculum (5 mL) stored at −80 • C was activated in a 50 mL LAY medium containing gentamycin (5 µg/mL), kanamycin (5 µg/mL), and nalidixic acid (5 µg/mL) at 100 rpm at 30 • C for 24 h. After subculturing thrice, 10% (v/v) inoculum was transferred to the 42 L fermenter for submerged culture.
The 42 L fermenter was filled with a 30 L culture medium containing 20 g (70-80%) sodium lactate, 5.0 g NH 4 Cl, 6.0 g MgSO 4 .7H 2 O, 15.0 g K 2 HPO 4 .3H 2 O, 15.0 g yeast extract, and 17.5 mL Wolfe mineral solution. The feeding medium contained 500 g lactic acid, 90 mL NH 3 .H 2 O, 6.0 g MgSO 4 .7H 2 O, 15.0 g yeast extract, 15.0 g K 2 HPO 4 .3H 2 O, 10.0 g FeCl 3 .6H 2 O, and 17.5 mL Wolfe mineral solution was dissolved in 700 mL ddH 2 O. The culture and feeding medium were sterilized for 30 min at 121 • C. Gentamycin (5 µg/mL), kanamycin (5 µg/mL), and nalidixic acid (5 µg/mL) were added before inoculation. The pH was maintained at 6.8 by automated supplementation of the feeding medium supplemented with high concentration of lactate, and the temperature was maintained at 30 • C. The initial airflow was 0.5 L/min, and the stirring rate was 100 rpm. Once the dissolved oxygen (dO 2 ) decreased to 15%, we increased the airflow to 1 L/min and maintained dO 2 between 0% and 1% by regulating agitation (10 rpm every 2 h). When the dO 2 decreased to 5%, 7.5 mL 1M isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to induce the bccp-mamF gene expression. The cell density and magnetic response were measured every 4 h during the fermentation process.

Extraction and Characterization of BMP−∆F−BF
Fermenter−cultured MSR−∆F−BF was collected by centrifugation at 10,000 rpm at 4 • C for 15 min. The collected cells were resuspended in 20 mM phosphate−buffered saline (PBS) with a mass volume ratio of 1:10 (pH 7.4) and disrupted using an ultrasonic homoge-nizer (Ningbo Scientz Biotechnology Co., Ltd, Ningbo, China) for 20 min at 3 s/time with a time interval of 5 s at 400 W to lyse the cells. BMPs were collected from the disrupted cell suspension using a columnar neodymium−boron magnet. The supernatant was discarded, and the separated magnetosomes were resuspended in 10 mM PBS (pH 7.4), sonicated for 15 min at 3 s/time with a time interval of 5 s at 150 W, and washed twice. Then, we reduced the power to 100 W and washed the sample several times until the supernatant protein concentration was lower than 0.1 mg/mL. The purified BMP−∆F−BF was resuspended in 25% glycerol and stored at 4 • C.
Purified BMP−∆F−BF and wild-type BMP were observed under a TEM. For this purpose, a small number of BMPs was suspended in 1 mL of deionized water and thoroughly dispersed by ultrasonication for 10 min. Then, 10 µL of the suspension was dropped onto a copper mesh for 10 min, air-dried, and observed under a TEM. Hydrated radii and zeta potential of BMP−∆F−BF were detected at Nanjing Dongna Biological Technology Co., Ltd., Nanjing, China.

Preparation of Biotin−Labeled BMPs with NHS−Biotin
BMPs are membrane-enclosed magnetite nanocrystals. The primary amines of the BMP membrane were biotinylated by NHS−Biotin to achieve biotin-labeled BMPs (BMP−biotin). Then, 2 mg wild-type BMP was suspended in 1 mL 10 mM PBS (pH 7.4), and 13.5 µL 20 mM NHS−Biotin was added. The reaction mixture was ultrasonicated (70 W) for 1 min, incubated with shaking (200 rpm) for 30 min at 25 • C, and placed over a magnet to isolate BMP−biotin. BMP−biotin was suspended in a 25% glycerol and stored at 4 • C.

Detection of the Streptavidin−HRP Linkage Rate of BMP−∆F−BF and BMP−biotin
The linkage rate of BMP−∆F−BF and BMP−biotin with HRP−streptavidin was compared with commercial biotin-labeled magnetic beads (MB−biotin; superparamagnetic iron oxide; particle size 20-50 nm; Nanocs, Natick, MA, USA). BMP−∆F−BF, BMP−biotin, and MB−biotin were incubated with 1% bovine serum albumin (BSA) and stored at 4 • C. One−step ELISA was used to detect the linkage rate. A 96−well microtiter plate coated with 300 µL 2% BSA was incubated at 4 • C overnight and washed thrice with 250 µL 10 mM PBST (pH 7.4). Then, 2 mg BMP−∆F−BF, BMP−biotin, and MB−biotin each were added (each BMPs and MB three parallel) and captured using a magnetic rack. The sample was then washed thrice with PBST, and 100 µL 1:2000 diluted HRP-streptavidin (Beijing Solarbio Science & Technology, Beijing, China) was added to each well, with coupling for 1 h at 600 rpm at room temperature, captured by a magnetic rack and washed five times with PBST. The color was developed using 100 µL tetramethylbenzidine (TMB) for each well, reacted for 3 min at room temperature, and then 50 µL 2 M H 2 SO 4 was added to stop the reaction. The supernatant was transferred to a new 96−well microtiter. Absorbance at 450 nm was measured on a microplate reader (blank control, only BMPs and MB incubated, no HRP-streptavidin). The OD 450 value was calculated to compare the BMPs and MB linkage rates.

Construction of the Engineered Strain MSR−∆F−BF
Previously, we adopted the "BMP surface display technique" to exhibit the functional molecule (protein A) on BMPs [40] used for the detection or diagnostic immunoassays, having a high Ab-conjugation and antigen-adsorption capability [44]. Here, we fused BCCP expression with BMPs abundant membrane protein (MamF) to display biotin on BMPs ( Figure 1A). Firstly, a fusion expression vector was constructed. The bccp and mamF genes were fused into the bccp-mamF gene by fusion PCR then cloned into a broad host-vector pBBR1MSC-2 to construct fusion expression vector pBBR−bccp−mamF. Then, pBBR−bccp−mamF was transformed into MSR−∆F by biparental conjugation to construct an engineered strain, MSR−∆F−BF. The authigenic biotin ligase can link biotin to a lysine residue of BCCP protein, obtained biotin-labeled engineered BMPs (BMP−∆F−BF). MSR−∆F native mamF gene in the genome was replaced by the gentamicin-resistance gene. Choosing MSR−∆F as the host strain can increase the amount of biotin displayed on engineered BMPs.
To detect the expression of the fusion gene, a gfp-labeled three-gene fusion expression vector was constructed. The expression of fusion protein BCCP-MamF-GFP on engineered BMPs was confirmed by Western blotting (Figures 1B and S1). Compared with wild-type BMPs (BMP−WT), engineered BMPs displayed a specific band at 55 kDα. TEM observation indicated that no BMPs were synthesized in MSR−∆F, whereas new small-size BMPs were synthesized in MSR−∆F−BF ( Figure 1C). The fusion gene expression complemented the disruption of the mamF gene in MSR−∆F−BF, confirming that the fusion gene was successfully expressed in MSR−∆F−BF.
The MSR−∆F−BF cell growth and BMP synthesis states were detected in the flask and compared with MSR−∆F and wild-type MSR-1. Flask-cultured MSR−∆F−BF growth was slower than that of MSR−WT but faster than that of MSR−∆F. OD 565 reached the highest (~0.98) at 22 h, approaching MSR−WT (~1.01) at 20 h ( Figure 1D). As no BMPs were synthesized in MSR−∆F, no Cmag value was detected. The highest Cmag value of MSR−∆F−BF was~0.3 at 16 h, 3.33-fold lower than that of MSR−WT (~1.0) at 12 h ( Figure 1E), because the BMP size in MSR−∆F−BF was smaller than that in MSR−WT.
These results indicate that the constructed MSR−∆F−BF grows well and can synthesize BMPs. MSR−∆F−BF can be used for producing of biotinylated BMPs.

Magnetic Properties Analysis of MSR−∆F−BF
MSR−∆F−BF displayed a lower Cmag value, and the synthesized BMP size was smaller than that of MSR−WT. TEM showed that MSR−∆F−BF synthesized BMPs displayed sufficient spaces between the particles. In this section, we firstly adopted highresolution TEM and fast Fourier transform to analyze the BMP composition of MSR−∆F−BF. The BMP−∆F−BF crystal shape remained a hexagonal octahedron, and a few possessed an irregular shape (Figure 2A). Despite the small size of BMP−∆F−BF, it exhibited a typical ferroferric oxide lattice structure. To elucidate the magnetic properties of BMPs in MSR−∆F−BF, the hysteresis loops were measured (at room temperature) and compared with that of MSR−WT. Hysteresis loops for MSR−WT displayed a bigger gap between two lines than that for MSR−∆F−BF ( Figure 2B). The coercivity and saturation magnetization values for MSR−∆F−BF were 2.24 and 4.26, respectively, much smaller than those for MSR−WT (7.65 and 12.55, respectively) ( Table 2). Compared with MSR−WT, MSR−∆F−BF synthesized small-size BMPs, which decreased its magnetic properties, but the core constituent was still noted as Fe3O4. This property displays better applicability, and the dispersibility of BMP−∆F−BF will be better than that of BMP−WT. Subsequent experiments will attest to this viewpoint.   To elucidate the magnetic properties of BMPs in MSR−∆F−BF, the hysteresis loops were measured (at room temperature) and compared with that of MSR−WT. Hysteresis loops for MSR−WT displayed a bigger gap between two lines than that for MSR−∆F−BF ( Figure 2B). The coercivity and saturation magnetization values for MSR−∆F−BF were 2.24 and 4.26, respectively, much smaller than those for MSR−WT (7.65 and 12.55, respectively) ( Table 2). Compared with MSR−WT, MSR−∆F−BF synthesized small-size BMPs, which decreased its magnetic properties, but the core constituent was still noted as Fe 3 O 4 . This property displays better applicability, and the dispersibility of BMP−∆F−BF will be better than that of BMP−WT. Subsequent experiments will attest to this viewpoint.

Submerged Culture of MSR−∆F−BF in Bioreactor
Large-scale cultivation of the engineered strains is the basis for engineered BMP application. However, MTB large-scale cultivation is challenging, especially the cultivation of engineered strains. Similarly, MTB such as Magnetospirillum gryphiswaldense MSR-1, Magnetospirillum magnetotacticum MS-1, Magnetospirillum magneticum AMB-1, Magnetovibrio blakemorei MV-1, and Desulfovibrio magneticus RS-1 can only be cultivated on the laboratory scale. The BMP production level is low. The order of the BMP yield is MSR-1 > AMB-1 > MS-1 [53]. Previously, Zhang et al. achieved the highest cell yield of 9.16 g/L and BMP yield of approximately 356.52 mg/L by optimizing the MSR-1 fed-batch culture feeding medium [54]. Later, we reported the yield cultivation of an engineered strain (M. gryphiswaldense ∆F-FA) on a large scale and achieved the highest cell and BMP dry weight of 2.26 g/L and 62.29 mg/L, respectively [44].
We attempted to culture M. gryphiswaldense MSR−∆F−BF in a 42 L fermenter and detected the cell and magnetosome yield. BMP synthesis began when dO 2 decreased to~1% ( Figure 3A). Low dO 2 concentration inhibited cell growth; however, high dO 2 concentration influenced BMP synthesis. Thus, dO 2 concentration control is significant for MSR-1 fermentation. The gradual increase in dO 2 caused a significant increase in biomass. Therefore, we controlled the dO 2 value between 0% and 1% by regulating the airflow and agitation speed to ensure bacterial growth and BMP synthesis. Cmag reached 1.3 after 48 h of culturing and then began to decline. MSR−∆F−BF fermentation was complete when the Cmag value decreased to~0.8 ( Figure 3B). The maximum cell and BMP yield were 8.14 g/L and 134.44 mg/L, respectively,~3-fold higher than those of M. gryphiswaldense ∆F-FA. These findings indicate that MSR−∆F−BF is potentially useful for industrial fermentation processes, although further optimization of culture conditions is necessary. Large-scale cultivation of the engineered strains is the basis for engineered BMP application. However, MTB large-scale cultivation is challenging, especially the cultivation of engineered strains. Similarly, MTB such as Magnetospirillum gryphiswaldense MSR-1, Magnetospirillum magnetotacticum MS-1, Magnetospirillum magneticum AMB-1, Magnetovibrio blakemorei MV-1, and Desulfovibrio magneticus RS-1 can only be cultivated on the laboratory scale. The BMP production level is low. The order of the BMP yield is MSR-1 > AMB-1 > MS-1 [53]. Previously, Zhang et al. achieved the highest cell yield of 9.16 g/L and BMP yield of approximately 356.52 mg/L by optimizing the MSR-1 fed-batch culture feeding medium [54]. Later, we reported the yield cultivation of an engineered strain (M. gryphiswaldense ∆F-FA) on a large scale and achieved the highest cell and BMP dry weight of 2.26 g/L and 62.29 mg/L, respectively [44].
We attempted to culture M. gryphiswaldense MSR−∆F−BF in a 42 L fermenter and detected the cell and magnetosome yield. BMP synthesis began when dO2 decreased to ~1% ( Figure 3A). Low dO2 concentration inhibited cell growth; however, high dO2 concentration influenced BMP synthesis. Thus, dO2 concentration control is significant for MSR-1 fermentation. The gradual increase in dO2 caused a significant increase in biomass. Therefore, we controlled the dO2 value between 0% and 1% by regulating the airflow and agitation speed to ensure bacterial growth and BMP synthesis. Cmag reached 1.3 after 48 h of culturing and then began to decline. MSR−∆F−BF fermentation was complete when the Cmag value decreased to ~0.8 ( Figure 3B). The maximum cell and BMP yield were 8.14 g/L and 134.44 mg/L, respectively, ~3-fold higher than those of M. gryphiswaldense ∆F-FA. These findings indicate that MSR−∆F−BF is potentially useful for industrial fermentation processes, although further optimization of culture conditions is necessary. As the cell grew, the dO2% in the broth was consumed continuously. dO2% decreased to ~5%, and IPTG was added to introduce fusion gene expression. dO2% decreased to ~1%; the agitation speed was added at 10 rpm/2 h to maintain microaerobic condition. (B) MSR−∆F−BF cell growth and Cmag value curve. The highest OD565 and Cmag reached ~22 and ~1.3, respectively. As the cell grew, the dO 2 % in the broth was consumed continuously. dO 2 % decreased to~5%, and IPTG was added to introduce fusion gene expression. dO 2 % decreased to~1%; the agitation speed was added at 10 rpm/2 h to maintain microaerobic condition. (B) MSR−∆F−BF cell growth and Cmag value curve. The highest OD 565 and Cmag reached~22 and~1.3, respectively.

Preparation and Characterization of BMP−∆F−BF
Fermented cells were stored at −80 • C and used for engineered magnetosome purification. Cells at 1 g wet weight were suspended in 10 mL PBS, high-power ultrasonication was applied to disrupt the cell, and BMP−∆F−BF was separated using a magnet ( Figure 4A). Then, BMP−∆F−BF was washed several cycles by a low-power ultrasonic bath to remove cell debris on BMPs. After washing six times, the supernatant protein concentration was lower than 0.1 mg/mL and nearly stable, indicating that no associated proteins dropped from BMP−∆F−BF ( Figure 4B). TEM showed the extracted BMP−∆F−BF; no protein fragment was observed in the background, providing purified engineered BMPs ( Figure 4C).
Bioengineering 2022, 9, x FOR PEER REVIEW 9 of 16 displayed weak magnetism compared with wild type. Therefore, BMP−∆F−BF showed good dispersibility in the solution. First, we measured the purified BMP−∆F−BF particle size using the ImageJ software (National Institutes of Health, Bethesda, MD, USA). The average particle size was 20 ± 4.5 nm, smaller than that of MSR−WT (35 ± 7.8 nm). BMP−∆F−BF and BMP−WT displayed a narrow diameter distribution ( Figure 4D). Then, the hydrated radii and zeta potential of BMP−∆F−BF in a water-based suspension system were detected and compared with BMP−WT and commercial MB−biotin ( Table 3). The BMP−∆F−BF hydrated radii value was 53.52 ± 10.23 nm, much smaller than the wild type (113.9 ± 62.19 nm) and MB−biotin (669.3 ± 288.5 nm). The zeta potential reflected the particle surface charge, with a value above +30 mV or below −30 mV indicating good particle dispersity. BMP−∆F−BF's zeta potential was −23.6 ± 6.12 mV, close to −30 mV, lower than wild type (−17.7 ± 6.78 mV) and MB−biotin (−17.5 ± 3.84 mV). These results indicated that the smaller particle sizes and lower BMP-∆F-BF magnetic properties exhibit outstanding dispersity in water-based suspension systems and render them suitable candidates for further BMP-∆F-BF applications.   The dispersibility of nanomaterial in solution played a key role in applications. As mentioned above, the engineered strain MSR−∆F−BF synthesized small-sized BMPs and displayed weak magnetism compared with wild type. Therefore, BMP−∆F−BF showed good dispersibility in the solution. First, we measured the purified BMP−∆F−BF particle size using the ImageJ software (National Institutes of Health, Bethesda, MD, USA). The average particle size was 20 ± 4.5 nm, smaller than that of MSR−WT (35 ± 7.8 nm). BMP−∆F−BF and BMP−WT displayed a narrow diameter distribution ( Figure 4D). Then, the hydrated radii and zeta potential of BMP−∆F−BF in a water-based suspension system were detected and compared with BMP−WT and commercial MB−biotin ( Table 3). The BMP−∆F−BF hydrated radii value was 53.52 ± 10.23 nm, much smaller than the wild type (113.9 ± 62.19 nm) and MB−biotin (669.3 ± 288.5 nm). The zeta potential reflected the particle surface charge, with a value above +30 mV or below −30 mV indicating good particle dispersity. BMP−∆F−BF's zeta potential was −23.6 ± 6.12 mV, close to −30 mV, lower than wild type (−17.7 ± 6.78 mV) and MB−biotin (−17.5 ± 3.84 mV). These results indicated that the smaller particle sizes and lower BMP-∆F-BF magnetic properties exhibit outstanding dispersity in water-based suspension systems and render them suitable candidates for further BMP-∆F-BF applications. Table 3. Hydrated radii, zeta potential and polydispersity of BMP-∆F-BF, BMP-WT and MB-biotin.

Detection of BMP−∆F−BF Coupling Streptavidin Rate
The linkage rate of BMP−∆F−BF to streptavidin was detected and compared with BMP−biotin, constructed through chemical modification and commercial MB−biotin. First, the conditions for the linkage between BMP−∆F−BF and streptavidin were optimized, including the coupling time, buffer, and BMP−∆F−BF and streptavidin weight rate ( Figure 5A). The bonding between biotin and avidin displayed the strongest affinity (Kd = 10~15 M), leading to the linkage time ( Figure 5A1) and buffer ( Figure 5A2) having no significant influence on the streptavidin linkage rate. BMP−∆F−BF (0.1 mg) was used to bind with different amounts of streptavidin (ranging from 0.1-0.5 mg) to optimize the linkage weight rate of BMP−∆F−BF and streptavidin ( Figure 5A3). As streptavidin's weight increased, BMP−∆F−BF linkage with streptavidin increased, and the streptavidin nonspecific absorption of BMP−WT increased. When the BMP−∆F−BF and streptavidin weight rate was 1:3, the rate of BMP−∆F−BF linkage with streptavidin was considerably higher than BMP−WT nonspecific absorption and 1:1 and 1:2 conditions, indicating suitable conditions for BMP−∆F−BF linkage with streptavidin.
BMP−∆F−BF linkage with the streptavidin rate was compared with chemically modified BMP−biotin. The schematic diagram of BMP−biotin construction is shown in Figure 5B. BMP in vitro biotinylation was constructed by incubating with NHS-biotin. The -NHS group of NHS-biotin can react with the -NH 2 group of the BMP membrane and modify biotin onto BMP, forming biotin-labeled BMP. BMP−∆F−BF linkage with the HRPlabeled streptavidin (Streptavidin−HRP) rate was compared with those of BMP−biotin and commercial MB−biotin ( Figure 5C). BMP−∆F−BF and BMP−biotin linkage with streptavidin rates were almost the same, lower than MB−biotin. However, as the storage time increased to 75 d, BMP−biotin and MB−biotin linkage with the Streptavidin−HRP rate markedly decreased, and the BMP−∆F−BF linkage rate became more stable. One possible explanation is that nonspecific chemical modification helped the reaction with all amino groups on BMP. Biological modification is specific to the BMPs transmembrane protein (e.g., MamF) as an anchoring protein, ensuring stable protein expression on BMPs.

Discussion
Interest in nanomaterials and especially nanoparticles has exploded in the past decades primarily due to their novel or enhanced physical and chemical properties compared to the bulk material. These extraordinary properties have created a multitude of innovative applications in the fields of medicine and pharma, electronics, agriculture, chemical catalysis, the food industry, and many others. More recently, nanoparticles are also being synthesized "biologically" through the use of plant− or microorganism−mediated

Discussion
Interest in nanomaterials and especially nanoparticles has exploded in the past decades primarily due to their novel or enhanced physical and chemical properties compared to the bulk material. These extraordinary properties have created a multitude of innovative appli-cations in the fields of medicine and pharma, electronics, agriculture, chemical catalysis, the food industry, and many others. More recently, nanoparticles are also being synthesized "biologically" through the use of plant− or microorganism−mediated processes, as an ecofriendly alternative to the expensive, energy−intensive, and potentially toxic physical and chemical synthesis methods [55]. Among these organisms, MTB are unique as they synthesize magnetic nanoparticles in their organelles (magnetosome) under strict and fine genetic control [24][25][26].
Magnetosomes are excellent biomaterials, but only a few MTBs could achieve tractability and straightforward cultivation because of their restricted living conditions. Only two strains in the genus Magnetospirillum-M. gryphiswaldense MSR−1 and M. magneticum AMB-1-could be cultivated in bulk. Unfortunately, the magnetosome productions of these strains were much lower than the chemical synthesized magnetic nanoparticles [29], and the productions by those engineered stains were even lower [44]. In this study, we used MSR-1 as a host strain for the bioproduction of biotinylated magnetosomes; the engineered magnetosomes yield~134.44 mg/L, over twofold higher than the latest report by another recombinant strain [44]. In addition, there are several advantages to this recombinant strain and bioengineered magnetosomes. First, the strain MSR−∆F−BF was cultivated in a 42 L fermenter, while stains in most other reports (including wild−type and recombinant strains) were cultivated in a shaking flask or 5 L fermenter [29], which implied that MSR−∆F−BF could be cultivated more suitably at an industrial scale. Second, in terms of the size of the bioengineered particle, BMP−∆F−BF (20 ± 4.5 nm) was smaller than BMP−WT (35 ± 7.8 nm), which meant better dispersity in the water−based suspension system. Third, compared with commercial biotin−labeled magnetic beads and chemically constructed biotinylated magnetosomes, BMP−∆F−BF has better storage stability.
The process of magnetosome biotinylation in this work could also be considered an effective biotin surface display technique. The surface display is a recombinant technology that expresses target proteins on the surface of biomembranes, biostructures, or even biomolecules, and can be applied to almost all types of biological entities from viruses to mammalian cells. Phage display, mRNA display, ribosome display, bacteria display, and yeast display are among the most commonly used methods. These techniques have been used for various biotechnical and biomedical applications such as therapy of disease, drug screening, biocatalysts, library screening, quantitative assays, and biosensors [56][57][58]. However, most functional molecules expressed on the surface of display platforms are proteins or peptides. The successful display of biotin on the magnetosome surface suggested the possibility and strategies display biomolecules beyond proteins by bioengineering methods, which would greatly increase the applicable scopes of the above systems.
Up until now, biotin has been the only nonprotein biomolecule employed in the magnetosome surface display system. Biotin is widely distributed in animal and plant tissues and has a molecular weight of 244.31 Da with two ring structures: imidazolone ring, the main site noncovalently binding with avidin/streptavidin or special receptors with high specificity and affinity [59]; and thiophene ring, with the pentanoic acid as the side chain at C2, whose carboxyl terminal is the only structure to bond covalently with other biomolecules [60]. Biotin carboxyl carrier proteins (BCCPs) are molecules of 69~73 amino acid residues, to which a biotin group is covalently attached through a lysine residue [61]. The avidin/biotin couple is popular in the conjugation of nanosystems. Biotinylated nanoparticles were widely used for biological research and biomedical applications, including the sensing of analytes or the delivery of drugs [62,63]. In this work, fusion expression of mamF and bccp genes resulted in magnetosome surface modification by functional biotins in vivo, without grafting and conjugation steps of artificial nanoparticles.
Particularly, bioengineered biotinylation of magnetosomes brings an important new function to the display system. It is known that biotin receptors are overexpressed in different tumor cells, including breast, ovarian, and lung cancer [64,65]. Thus, the approach of biotin functionalization has been explored as an attractive strategy for tumor nuclear magnetic resonance imaging, targeted-drug, and gene−delivery systems [66][67][68][69]. Conse-quently, BMP−∆F−BF could be an excellent potential candidate as an imaging agent and targeted-delivery carrier for tumors.
In brief, we systematically introduced the bioengineering and biofabrication process of biotinylated magnetosomes. Both the strategy and resulting engineered magnetosome could be promising examples for preparing multiple functional biogenic magnetic nanomaterials.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available from the first author (J.X.) upon reasonable request.