Design

Last year, the CNPEM-BRAZIL team participated in the iGEM Design League, the Latin American spin-off of the iGEM competition. Our project, B.A.R.B.I.E. (Bioengineered Approach for Removal of Microplastics through Bioremediation and Innovative Electromagnetics), was inspired by the Kyoto 2019 team, which used plastic-binding proteins to address plastic pollution.

Our approach involved a bio-based system designed to detect and remove microplastics (MPs) from water treatment plants (WTPs). We combined encapsulin, a protein from Quasibacillus thermotolerans capable of mineralizing iron, with magnetic particles. This allowed us to create a magnetic complex capable of being pulled out by external magnetic fields. Attached to the encapsulin was BaCBM2, a plastic-binding protein that adheres to various plastics, and the SpyTag-SpyCatcher system to link the components.

The idea was to employ this magnetic protein complex (MBPE) to water treatment plants, where it would bind to MPs. Using magnetic fields, we could then remove the MPs along with the encapsulin complex. While innovative, this strategy proved too complex and hard to implement in real life, leading us to a completely new approach this year.

Design

Cycle 1 of B.A.R.B.I.E. 4.0 started this year, focusing on simplifying and improving the performance of plastic-binding proteins (PBPs) to enhance microplastic removal. Based on lessons learned from last year’s complex encapsulin-based system, we chose to streamline the approach by replacing encapsulin with Mad10, a small peptide that still allowed magnetic manipulation with a great reduction of the system's complexity. At this stage, our focus remained on water treatment plants (WTPs), as we had not yet transitioned to household filters.

A key part of our design was the inclusion of specific elements in the protein circuit: a 6xHis tag, the plastic-binding protein (PBP), miRFP (a red fluorescent protein), and three repeats of the Mad10 peptide.

• 6xHis tag: This was added to simplify the protein purification process. By including a 6xHis tag, we ensured that the proteins could be easily purified using nickel affinity chromatography. This tag binds strongly to nickel ions, allowing us to isolate PBPs from other cellular proteins in a single purification step, enhancing the efficiency and speed of the protein production process.
• PBP (Plastic-Binding Protein): The core of the system is the PBPs (Bacillus subtilis BaCBM2 and our designed protein BARBIE1), which are responsible for binding to various types of plastics. The goal was to enhance BARBIE1 affinity to different plastic types through our design pipeline, ensuring that the protein could efficiently capture microplastics in solution.
• miRFP (Red Fluorescent Protein): We included miRFP as a visual marker to monitor the expression and localization of the protein. The red fluorescence would allow the tracking and confirm the presence of the protein in the system, ensuring that it was being correctly expressed and localized where expected.
• 3xMAD10: The three repeats of the Mad10 peptide were included to retain the magnetic properties that were present in the original encapsulin design, but in a much simpler form. Mad10 is a peptide known to bind to magnetic particles, allowing the protein to be manipulated by an external magnetic field. By including three copies of Mad10, we aimed to increase the magnetic responsiveness of the system, enabling more efficient capture and removal of microplastics when combined with magnetic separation techniques.

In this cycle, we also developed a novel pipeline for designing PBPs with enhanced affinity for pollutants, especially microplastics. The steps involved in creating these engineered proteins included:

• Generating the structure of BaCBM2 using AlphaFold2.
• Performing docking assays with six common plastics (PP: Polypropylene, PE: Polyethylene, PET: Polyethylene Terephthalate, NY: Nylon, PVC: Polyvinyl Chloride and PS: Polystyrene) using Gnina software.
• Removing overlaps and visualizing sequences using ChimeraX.
• Using reverse folding with LigandMPNN to generate unique protein sequences.
• Retrieving the consensus sequence from the 36,000 designed proteins to identify structural features shared across all tested plastics. This consensus sequence led to the creation of our most optimized plastic-binding protein, BARBIE1.

BARBIE1 was then validated through additional docking assays, which confirmed its increased affinity for all plastics compared to the original BaCBM2, representing a significant advancement in our protein design capabilities.

Build

In this step, the focus was on building the newly designed plastic-binding protein, BARBIE1, generated from our protein design pipeline. The protein was derived from BaCBM2 and optimized for enhanced plastic affinity, as described in the design phase. To validate this new protein, we needed to ensure its successful expression, purification, and structural integrity.

We began by purchasing the DNA synthesis of the BARBIE1 sequence, which was the consensus from our pool of 36,000 designed proteins. The synthesized gene is cloned into the pET-15b expression vector, which utilizes the T7 promoter for high-level protein expression. This vector also included a 6xHis tag for ease of purification by affinity chromatography.

This construction was transformed it into Escherichia coli BL21(DE3), a strain commonly used for recombinant protein production due to its ability to express proteins under the control of the T7 promoter.

Test

After constructing the BARBIE1 protein, we conducted a series of in silico and experimental assays to evaluate its structural integrity, functionality, and plastic-binding efficiency, comparing it to BaCBM2.

In Silico Tests

We began by using AlphaFold 3 to predict BARBIE1's structural properties. The prediction indicated that BARBIE1 is likely to exist as a monomer, with a folding pattern highly similar to BaCBM2. This was crucial for ensuring that the modifications made during the design phase did not compromise the protein’s overall structure or its ability to function as a plastic-binding protein.

Next, we performed an electrostatic and hydrophobicity assessment in silico. These analyses revealed that BARBIE1 had increased hydrophobicity and a more negative charge compared to BaCBM2. Both characteristics are favorable for enhancing its affinity for hydrophobic plastic surfaces, suggesting that BARBIE1 would be more effective in binding plastics under various environmental conditions.

Experimental Tests

We used size exclusion chromatography with multi-angle light scattering (SEC-MALS) to analyze the oligomeric states of both BaCBM2 and BARBIE1. The results for BaCBM2 were consistent with the literature, showing predominantly monomeric forms, as well as smaller populations of dimers and tetramers. In contrast, the SEC-MALS results for BARBIE1 revealed a tendency for protein aggregation, with values that did not correspond to dimers, trimers, or tetramers, but rather large protein aggregates. This tendency for aggregation highlighted a potential issue with BARBIE1's stability in solution.

We also used dynamic light scattering (DLS)to study BARBIE1’s interaction with plastic particles. The data showed that BARBIE1 formed protein-corona structures around the plastic particles, indicating an increase of the hydrodynamic radius, which could enhance its ability to capture microplastics but also raised concerns about solubility under certain conditions.

Molecular dynamics simulations were performed to test BARBIE1’s plastic-binding affinity under different pH levels and temperatures. The simulations demonstrated that BARBIE1 maintained its plastic-binding affinity across a range of environmental conditions, indicating a possible robustness as a plastic-binding protein.

Finally, circular dichroism (CD) was used to assess the similarity between BARBIE1 and BaCBM2 as well as its thermal stability. The results indicated that while both proteins exhibited some thermal resistance, BARBIE1 showed reduced ability to refold after unfolding compared to BaCBM2, which retained a higher degree of refolding capacity.

Learn

The most significant insight from Cycle 1 came through Integrated Human Practices, where discussions with experts revealed that focusing on water treatment plants (WTPs) might not be the most viable path for implementation. The complexities of incorporating new filtration technologies in WTP infrastructure, combined with regulatory challenges, made us pivot towards household filters. This shift empowers individuals to improve their water quality without waiting for large-scale governmental changes.

From a technical perspective, the SEC-MALS results showed that BARBIE1 had a tendency to aggregate, forming protein clusters instead of remaining in its intended monomeric form. This unexpected behavior raised concerns about the protein’s effectiveness in binding microplastics, as aggregation could hinder its functionality. We realized that to move forward, we needed to address this issue directly.

In response to these findings, we added a new step to our protein design pipeline specifically to prevent protein aggregation. This modification aimed to refine BARBIE1 and ensure that it would remain stable in solution, improving its functionality in real-world applications. Additionally, we began exploring new strategies to prevent BARBIE1 from aggregating, ensuring it would be an efficient and reliable component of our filtration system.

These insights and technical adjustments were critical for guiding the next iteration of the project. By addressing the aggregation issue and refining our approach, we could optimize BARBIE1 for its intended use in household filters, making the system more practical and effective.

Design

With the valuable insights gained from Cycle 1, we shifted our focus towards developing a filtration system for household use. The initial idea was to create a filter capable of efficiently capturing microplastics. We realized that the filter would need a matrix to hold the plastic-binding proteins (PBPs) in place. Initially, we considered immobilizing our proteins on activated charcoal, which is commonly used in filtration systems. However, to ensure that our strategy could be applied to a wider variety of filter types, we opted for a biological alternative.

Inspired by the iGEM UCopenhagen 2022 team, which used spider silk protein (Nt2RepCt) to create sustainable fishing nets, we decided to explore the use of spidroin for our matrix. Spidroin has a high sustainability footprint and is cost-effective compared to other membranes commonly used for water purification1. Additionally, we found in the literature that nanofibers made from similar materials can be used for water filtration, and spidroin can form hydrogels that allow for the immobilization of proteins2. Furthermore, spider silk has been previously used in water purification for the batch adsorption of copper and nickel, providing a basis for its application in our system³.

For this iteration, we designed a construct combining several key elements: the Nt2RepCt spidroin protein, the SpyTag-SpyCatcher system, and the PBP (either BaCBM2 or BARBIE1). The goal was to create a strong and sustainable matrix that could incorporate our plastic-binding proteins and be integrated into household filters.

• Nt2RepCt Spidroin Protein:Chosen for its high sustainability and lower cost, this dual-repeat spider silk protein forms nanofibers that provide the structural backbone of the filtration matrix. The protein’s ability to form hydrogels also allowed us to immobilize the PBPs efficiently within the matrix. Literature also supports the use of spider silk in water purification, particularly for the adsorption of metals like copper and nickel, adding further value to its potential use.
• PBP (Plastic-Binding Protein): As in Cycle 1, the PBPs (BaCBM2 or BARBIE1) were at the core of the filtration system, responsible for binding microplastics as water passed through the filter. Their enhanced affinity for plastics, validated in the previous cycle, was expected to perform effectively when attached to the spidroin matrix.
• SpyTag-SpyCatcher System: This system was included to covalently link the PBPs to the spidroin matrix. The irreversible bond between SpyTag and SpyCatcher ensured that the plastic-binding proteins would remain securely attached to the matrix, even under varying flow conditions.

This design aimed to create a biologically-based filtration system capable of efficiently capturing microplastics while maintaining the sustainability and cost advantages of spidroin compared to conventional filtration membranes.

Build

In the build phase of cycle 2, we made multiple attempts to assemble the BaCBM2-SpyCatcher and BARBIE1-SpyCatcher constructs using Golden Gate assembly. Despite repeated efforts, these assemblies were unsuccessful. During this time, we found new literature that provided a better strategy for ensuring that our proteins would adhere to the spider silk matrix, which led us to abandon the current assembly approach.

For the Nt2RepCt-SpyTag, the assembly was successful. The plasmids were transformed via electroporation into E. coli DH5α, and the correct assembly was confirmed by Sanger sequencing.

Additionally, we worked on the cloning and transformation of Barbie1-Cys and BaCBM2-Cys. Both proteins were modified with a cysteine residue to enhance their interaction with the microplastics sensor surface, crucial for amplifying signals in electrochemical measurements. The constructs were assembled using the Golden Gate method and transformed into E. coli DH5α via electroporation. The correct assembly of both constructs was confirmed by Sanger sequencing, but we did not proceed with the purification of BaCBM2-Cys in Cycle 2. Instead, we shifted our focus to the next phase, where BaCBM2-Cys will be fused to the N-terminal of spidroin for further development in Cycle 3.

Test

In this phase, we focused on evaluating the expression and purification of Nt2RepCt-SpyTag and Barbie1-Cys, as well as analyzing the structural properties of Nt2RepCt using Circular Dichroism (CD), the heating ramp analysis, and Scanning Electron Microscopy (SEM).

Expression and Purification of Nt2RepCt-SpyTag:

The expression of Nt2RepCt-SpyTag was confirmed in E. coli DH5α using SDS-PAGE. However, during the initial purification process using Immobilized Metal Affinity Chromatography (IMAC) on a Ni-column, a significant amount of Nt2RepCt appeared in the flow-through, indicating inefficient binding to the column. This suggested the need for optimization in the purification protocol to improve yield and binding efficiency.

Improved Purification of Nt2RepCt-SpyTag:

In response to the poor initial purification, we optimized the purification strategy by removing imidazole from Buffer A and reducing the imidazole concentration in Buffer B. This adjustment, combined with a slower flow rate on the ÄKTA system (0.75 mL/min), greatly improved protein binding to the Ni-column. SDS-PAGE analysis confirmed a higher yield of Nt2RepCt in the elution fractions. A second round of purification was performed to further purify the protein, followed by Size Exclusion Chromatography (SEC), which confirmed the homogeneity of the sample.

Expression and Purification of Barbie1-Cys:

Barbie1-Cys expression was also confirmed using SDS-PAGE. However, during the purification process, we encountered challenges regarding the co-elution along with bacterial proteins. Despite using a linear gradient of imidazole for elution, the separation between Barbie1-Cys and E. coli proteins was inefficient, hindering further characterization of Barbie1-Cys in this cycle. This result highlighted the need for additional optimization in the purification protocol, particularly to resolve the co-elution issue.

Circular Dichroism (CD) Analysis:

Following purification, CD analysis was used to examine the secondary structure of Nt2RepCt before and after heating. The CD spectrum showed significant changes after the heating ramp, indicating a structural transition, likely associated with the formation of a hydrogel. The heating process altered the secondary structure, consistent with the expected behavior of hydrogel formation, making it a promising candidate for use in water filtration applications.

Nt2RepCt Heating Ramp:

To further explore the hydrogel formation properties of Nt2RepCt, we conducted a heating ramp analysis, heating the protein from 25°C to 60°C. The results indicated that the conformational changes observed during heating were non-reversible, which is a key characteristic for stable hydrogel formation. This stability at ambient temperature without reverting to the original protein state is crucial for its use in filtration systems.

Nt2RepCt Hydrogel Scanning Electron Microscope (SEM):

Finally, the hydrogel formed by Nt2RepCt was analyzed using Scanning Electron Microscopy (SEM). The SEM images revealed an irregular surface topology with potential pore formation, supporting its functionality in filtering applications. This irregularity and roughness of the surface suggest that the protein can create compartments and fibers, which are beneficial for particle filtration.

Learn

Cycle 2 provided crucial insights that shaped the direction of our project, particularly in terms of protein expression, purification, and the functional behavior of our constructs.

1- Nt2RepCt-SpyTag Purification: The initial purification attempts for Nt2RepCt-SpyTag revealed that our initial buffer conditions were suboptimal, leading to significant protein loss in the flow-through. Through trial and error, we learned that modifying the purification buffers, especially removing imidazole from Buffer A and adjusting the imidazole concentration in Buffer B, significantly improved the binding efficiency of Nt2RepCt to the Ni-column. This allowed us to achieve a purer and more concentrated sample of the protein, demonstrating the importance of buffer composition in maximizing protein yield. Additionally, we discovered that the flow rate on the purification system played a crucial role in avoiding premature fiber formation during purification.

2- Challenges with Barbie1-Cys Purification:The purification of Barbie1-Cys faced challenges regarding co-elution along bacterial proteins, making it difficult to separate Barbie1-Cys from contaminants. Despite various attempts to optimize the purification protocol, we were unable to obtain a sufficiently pure sample for further characterization. This highlighted the need to rethink our approach to producing and purifying Barbie1-Cys. We learned that more stringent purification strategies will be necessary in the next iteration to address the co-elution issues and to ensure we can proceed with functional testing.

3-Nt2RepCt Hydrogel Formation: One of the major successes of Cycle 2 was confirming the ability of Nt2RepCt to form hydrogels. Using Circular Dichroism (CD) and the heating ramp analysis, we observed that Nt2RepCt undergoes a non-reversible conformational change when heated, leading to the formation of a stable hydrogel. This discovery is significant for the development of our filtration system, as the formation of a hydrogel matrix provides a solid foundation for immobilizing plastic-binding proteins and ensuring their functionality in water filtration applications. The non-reversible nature of this conformational change suggests that the hydrogel will maintain its structure under real-world conditions, making it an ideal matrix for long-term use in filtration.

4- Surface Morphology of Nt2RepCt Hydrogel: The Scanning Electron Microscopy (SEM) images provided valuable insights into the surface structure of the hydrogel formed by Nt2RepCt. The irregular surface and pore formation observed in the SEM images are promising features for filtration applications. The creation of a rough, fibrous surface with potential compartments suggests that Nt2RepCt can effectively filter microplastics by trapping particles within its structure. This finding reinforces the potential of Nt2RepCt as a robust matrix for our filtration system.

Key Takeaways for Next Cycle:

• Optimization of Purification Protocols:e learned that adjusting purification protocols is critical, particularly for difficult-to-purify proteins like Barbie1-Cys. More focused efforts on refining the purification strategy will be necessary in Cycle 3.
• Hydrogel Formation:The successful hydrogel formation by Nt2RepCt offers a solid foundation for immobilizing plastic-binding proteins, making it a crucial element in our filtration system.
• Continued Exploration of Protein Aggregation:Further work is needed to prevent protein aggregation, especially for the PBPs, to ensure their functionality when immobilized on the hydrogel matrix.

Design

Building on the lessons from Cycle 2 and new findings from the literature, we decided to simplify the design of our filtration system. Instead of using the SpyTag/SpyCatcher system to attach BaCBM2 and BARBIE1 to the spidroin matrix, we opted for a more straightforward strategy: fusing the N-terminal of spidroin directly to these plastic-binding proteins. This decision was driven by evidence that the N-terminal of spidroin is already sufficient to promote hydrogel formation, making the additional attachment mechanism unnecessary.

Why This Strategy?

• Hydrogel Formation:Literature indicates that the N-terminal domain of spidroin naturally supports hydrogel formation, which is essential for creating a stable filtration matrix. By fusing this domain directly to BaCBM2 and BARBIE1, we expect to maintain the protein's ability to form a hydrogel while also immobilizing the plastic-binding proteins within the matrix. This simplifies the system and reduces the complexity of the assembly.
• Reducing Aggregation:One of the major challenges we encountered in Cycle 2 was the aggregation of BARBIE1. By fusing BARBIE1 with the N-terminal spidroin, we anticipate that this new configuration will help reduce aggregation, providing a more stable and functional protein within the hydrogel.
• Efficiency and Simplification:This new design eliminates the need for extra components, such as SpyTag and SpyCatcher, and allows us to produce a single protein that can both form the hydrogel and bind microplastics. This will streamline production and purification, making the overall system more efficient and scalable.

Our expectation with this revised approach is to create a robust, simplified filtration system that reduces the complexity of previous designs and also enhances performance through a more stable and efficient protein-hydrogel integration.

Build

In the build phase, we constructed the fusion proteins Nt-BaCBM2-Cys and Nt-Barbie1-Cys using Gibson Assembly, which allowed us to integrate the N-terminal of spidroin directly with the plastic-binding proteins BaCBM2 and BARBIE1, both modified with a cysteine for enhanced interaction with the filtration matrix.

• Nt-BaCBM2-Cys: The biobricks BBa_K5396009 (basic) and BBa_K5396005 (composite) encode the N-terminal of spidroin (Nt2RepCt) fused to BaCBM2-Cys. The construct was generated using Gibson Assembly with the previously constructed plasmids BBa_K5396007 and BBa_K5396011 as templates. After assembly, the product was transformed into E. coli DH5α via electroporation, and successful plasmid construction was confirmed through Sanger sequencing.
• Nt-Barbie1-Cys: Similarly, the biobricks BBa_K5396010 (basic) and BBa_K5396006 (composite) encode the N-terminal of spidroin fused to Barbie1-Cys. This construct was also generated using Gibson Assembly with the plasmids BBa_K5396008 and BBa_K5396011 as templates. The assembled product was transformed into E. coli DH5α via electroporation, and the successful construction was confirmed by Sanger sequencing.

Test

During Cycle 3, several tests were performed to evaluate the expression, purification, and structural characterization of Nt-BaCBM2-Cys and Nt-Barbie1-Cys:

Purification of Nt-BaCBM2-Cys:

The first attempt to purify Nt-BaCBM2-Cys using Immobilized Metal Affinity Chromatography (IMAC) on a Ni-column resulted in co-elution with E. coli proteins. A single peak was observed on the chromatogram, and SDS-PAGE confirmed the presence of both the target protein and bacterial contaminants.

Refinement of Purification:

In the second purification attempt, the flow rate was reduced to 0.75 mL/min, leading to improved peak separation. SDS-PAGE analysis of the eluted fractions confirmed higher concentrations of purified Nt-BaCBM2-Cys in peaks 3 and 4, with fewer contaminants.

Expression Tests:

To optimize the expression of Nt-BaCBM2-Cys and Nt-Barbie1-Cys, small-scale induction tests were conducted. Samples were taken at 0h, 3h, 4h, and 5h after IPTG induction. SDS-PAGE results showed a slight increase in protein expression in the samples collected after 3-5 hours.

Circular Dichroism (CD) Analysis:

With the newly purified Nt-BaCBM2-Cys, CD analysis was performed to characterize its secondary structure. The spectrum showed a notable similarity between Nt-BaCBM2-Cys and Nt2RepCt-SpyTag, particularly in the range of 210 to 250 nanometers, indicating a structural contribution from both domains. In contrast, BaCBM2-RFP-3xMad10 displayed significant differences in the CD spectrum, particularly at lower wavelengths. Between 205 and 210 nm, Nt-BaCBM2-Cys showed lower absorbance compared to both Nt2RepCt and BaCBM2, suggesting that the fusion of the BaCBM2-Cys and Nt domains affects the overall secondary structure.

Learn

Cycle 3 provided valuable insights into the challenges of purifying and expressing both Nt-BaCBM2-Cys and Nt-Barbie1-Cys, as well as offering a deeper understanding of their structural properties through Circular Dichroism (CD) analysis.

• Purification Challenges: The initial difficulty in separating Nt-BaCBM2-Cys from E. coli contaminants highlighted the need for optimization of the purification process. By adjusting the flow rate during IMAC purification, we improved peak separation, but overall protein yield remained low. This indicates that further refinements are necessary, possibly involving modifications to the buffer composition or the use of different purification strategies.
• Expression Optimization: Expression tests conducted for both Nt-BaCBM2-Cys and Nt-Barbie1-Cys showed a slight increase in protein production after 3-5 hours of IPTG induction. However, the constitutive activity of the AB_T7_lacO promoter suggests that it may not be fully responsive to IPTG, limiting the increase in expression levels for both proteins. This insight suggests that alternative expression systems or promoters may be needed to achieve higher yields in future cycles.
• Structural Insights from Circular Dichroism (CD): The CD analysis of Nt-BaCBM2-Cys provided key insights into its secondary structure. The similarities observed between Nt-BaCBM2-Cys and Nt2RepCt-SpyTag suggest that the N-terminal spidroin domain contributes significantly to the protein’s overall structure. However, the contrast between Nt-BaCBM2-Cys and BaCBM2-RFP-3xMad10, especially at lower wavelengths, indicated that the fusion of BaCBM2-Cys with the N-terminal of spidroin altered the structural profile of the protein. This finding is crucial for understanding how the fusion impacts the behavior of the protein in the hydrogel matrix and its binding capacity to microplastics.
• Aggregation Comparison Between Cycles 1 and 3: Upon comparing the aggregation results from Cycle 1 and Cycle 3, we observed that the aggregation score for BARBIE1 did not improve significantly between these cycles. This indicated that despite the design modifications, the fusion did not resolve the aggregation issue as much as expected. To address this, the next step is to add an MBP (Maltose-Binding Protein) tag, which has been validated in silico to improve solubility and could potentially reduce aggregation in future experiments.
• Hydrogel Formation: Although some purified Nt-BaCBM2-Cys was obtained, the yield was still insufficient to perform hydrogel formation tests. This reinforced the importance of optimizing the expression and purification steps to generate enough protein for functional characterization and hydrogel testing in the future.

Next Steps:

• Further Refinement of Purification: Additional experiments will focus on fine-tuning the purification protocol to achieve higher yields and better separation of the target protein from contaminants.
• Promoter Investigation: We will explore alternative promoters or expression systems to improve the responsiveness to IPTG and boost protein production for both Nt-BaCBM2-Cys and Nt-Barbie1-Cys.
• MBP Tag Addition: To improve solubility and reduce protein aggregation, we plan to incorporate an MBP (Maltose-Binding Protein) tag in future iterations, which has already shown promising results in in silico analysis.
• Pipeline for Multiple Pollutants: The success of our protein engineering pipeline has inspired the development of a new approach for the B.A.R.B.I.E project: a toolbox for customizable protein design. This pipeline, capable of designing proteins with enhanced affinity for specific pollutants, offers a versatile solution not only for iGEM teams but also for addressing real-world environmental challenges. It can be adapted and tested for filtering a wide range of pollutants from water, extending beyond microplastics.
• Gene Library for High-Throughput Testing: As a result of our novel protein design pipeline, we have synthesized a gene library containing over 10,000 different proteins suitable for expression in Saccharomyces cerevisiae. Over the next few months, we will perform high-throughput screening to evaluate the efficiency of these proteins in binding and removing different types of pollutants, marking a significant step forward in the development of this customizable filtration system.

Design

Cycle 1 consisted of a sensing proposal based solely on the electrochemical response of micro- and nanoplastics. The design stages were divided as follows:

• Sensor architecture: A three-electrode architecture was chosen: working, counter, and reference electrodes. This configuration allows for various electrochemical measurements to be performed quickly and reproducibly.
• Detection method: The initial idea was to explore impedance spectroscopy measurements in a solution of Na₂SO₄. Later, the focus shifted to exploring square-wave voltammetry measurements in a solution of K₂NO₃ with K₃[Fe(CN)₆]/K₄[Fe(CN)₆].

Build

Therefore, based on the established design, the electrochemical devices were microfabricated in a cleanroom environment at the Brazilian Nanotechnology National Laboratory (LNNano) through the following fabrication steps:

• Glass substrate preparation
• Hexamethyldisilazane (HMDS) adhesion layer
• Photoresist deposition
• Photolithography and development
• Metal layer deposition
• Liftoff process
• Surface cleaning and hydrophilization
• SU-8 photoresist deposition
• Photolithography for SU-8
• Development of SU-8
• Final cleaning

Additionally, the experimental setup for the measurements was assembled using the portable PalmSens4 potentiostat, connectors, and a notebook for data storage and monitoring. Tests with different concentrations of polystyrene (PS) particles were prepared to evaluate the sensor’s performance.

Test

Therefore, the following tests were conducted:

• Electrochemical Impedance Spectroscopy (EIS) measurements for Na₂SO₄ and SWV for K₃[Fe(CN)₆]/K₄[Fe(CN)₆]: It was possible to evaluate the behavior of the solution using impedance spectroscopy.
• Analytical curve with EIS and Na₂SO₄, and with Square Wave Voltammetry SWV and K₃[Fe(CN)₆]/K₄[Fe(CN)₆]: The linearity in detecting PS particles was assessed.
• Scanning Electron Microscopy (SEM): The microfabricated device was thoroughly examined to understand the precision of the fabrication process.

Learn

From the results obtained in the tests, it was noted that there was some difficulty in finding a correlation between the impedance measurements and the PS concentration. Based on this, our team learned that square wave voltammetry (SWV) might be more advantageous for quantifying microplastics in our system due to its precision, simplicity, and speed. Moreover, for the students involved in the practical work, this cycle was extremely important as it provided a first experimental experience with several electrical measurements, which facilitated the development of subsequent cycles.

Design

After the first cycle, it was observed that the electrochemical sensor without biological modification showed limitations in the precise quantification of microplastics (MNPs), especially at lower concentrations. Thus, in the second cycle, the inclusion of a biological element was proposed to increase the sensor’s selectivity: the BaCBM2 protein, known for its high affinity for plastics. The objective was to test two main approaches for protein immobilization on the working electrode surface:

• Dropcasting: A simple method of protein adsorption, where a drop of the BaCBM2 solution is deposited directly on the electrode, allowing fixation through physical interaction.
• Self-Assembled Monolayers (SAMs): Structured modifications on the electrode surface using L-cysteine, cysteamine, and graphene oxide, with the goal of promoting a stronger and more stable protein attachment, thereby amplifying the electrochemical response.

Based on the protein-plastic binding characteristics, the use of the Langmuir model was planned to determine the dissociation constant (Kd), which allows quantification of the affinity between the proteins and plastic particles.

Build

The construction phase involved the preparation and modification of the sensors using the two immobilization methods:

• Drop casting: The BaCBM2 protein was adsorbed directly onto the electrode surface by depositing the protein solution followed by an incubation period. The sensor was then washed and subjected to electrochemical measurements using square wave voltammetry (SWV).
• Self-Assembled Monolayers (SAMs): The electrode surface was modified using three different agents (L-cysteine, cysteamine, and graphene oxide) to form self-assembled monolayers that facilitate the immobilization of BaCBM2. Each agent was incubated with the sensor for 15 hours at 4°C, followed by protein deposition, activation and the application of SuperBlock, to block non-specific binding sites.

Additionally, assays with progressively increasing concentrations of polystyrene nanobeads (PSNBs) (0.01 to 100 mg/L) were prepared to create analytical curves and evaluate the sensor’s performance.

Test

The modified sensors underwent a series of tests:

• SWV after modification: Square wave voltammetry (SWV) measurements were taken before and after the immobilization of the BaCBM2 protein. A noticeable decrease in current was observed after protein adsorption, indicating that the protein successfully adhered to the electrode surface.
• XPS (X-ray Photoelectron Spectroscopy): Surface characterization was performed to confirm the presence and homogeneous distribution of the proteins and SAMs on the electrode.
• Analytical curve: Both the drop casting method and the SAMs approach generated analytical curves based on microplastic concentrations. The SAM modified with cysteamine resulted in a curve with a coefficient of determination (R²) of 0.99, showing excellent reproducibility and low variations between measurements.
• Determination of Kd: The dissociation constant (Kd) was obtained using the Langmuir model to quantify the affinity between the BaCBM2 protein and polystyrene particles.

Learn

The results demonstrated a significant improvement in the quantification of microplastics compared to the previous cycle. The SAM modification, particularly using cysteamine, proved to be the most efficient, resulting in a more precise and reproducible electrochemical response, with an R² of 0.99. The use of immobilized proteins allowed for more accurate quantification of microplastics and opened the possibility of determining the dissociation constant (Kd), which could be valuable for more detailed studies of protein-particle affinity. Furthermore, this cycle highlighted the importance of the protein-plastic interaction in the sensor efficiency, suggesting that future adjustments to immobilization conditions and sensor design could further enhance performance.

Design

Cycle 3 explored a new approach for the sensor, utilizing the process of protein corona formation on micro and nano plastic particles (MNPs). The main concept involved the interaction of MNPs with the BaCBM2 protein in solution before deposition onto the working electrode. This bulk interaction process allowed for the formation of protein layers around the MNPs, creating what is known as a protein corona.

The choice of this methodology was motivated by the potential to enhance the selectivity and detection of MNPs with pre-adsorbed proteins, which can help capture the target particles more effectively. Additionally, the expression and purification of the BaCBM2 protein with a cysteine coating were considered to improve interaction with the surface of the MNPs.

Build

To construct this biosensor, the following steps were carried out:

• Solution Preparation: MNPs and the BaCBM2 protein were mixed in PBS solution and agitated for 10 minutes to ensure the formation of protein coronas around the particles.
• Dynamic Light Scattering (DLS) Measurements: DLS was conducted to evaluate the particle size and confirm the formation of the corona around the MNPs. These measurements allowed for the identification of changes in hydrodynamic diameter, suggesting the involvement of the protein.
• Computational and Analytical Modeling: A theoretical model was developed to predict the formation of the coronas and the interaction between the MNPs and the proteins in solution, based on the experimental conditions.
• Electrochemical Measurements: After the formation of the coronas, part of this solution was deposited directly onto the sensors, followed by the addition of an electrolyte to establish contact between the electrodes. Square wave voltammetry (SWV) measurements were then performed.

Test

The tests conducted yielded the following results:

• DLS Results: The DLS measurements confirmed the formation of protein coronas around the MNPs, with a significant increase in hydrodynamic diameter, demonstrating that the adsorption process was effective.
• Computational Analysis: The theoretical model developed to describe the formation of the coronas was consistent with the results observed in the DLS measurements, indicating that the formation of the protein layers can be reliably predicted.
• SWV Measurements: The analytical curves obtained from the SWV measurements revealed a decrease in the coefficient of determination (R² = 0.40), indicating a weak correlation between the concentration of MNPs and the electrochemical response. Additionally, the standard deviations in the measurements increased, negatively impacting the reproducibility of the results.

Learn

Despite the unsatisfactory results for the quantification of MNPs, Cycle 3 provided valuable insights into the formation of protein coronas. The technique was confirmed by DLS, demonstrating effective interaction between the proteins and the MNPs, which can be beneficial for future sensor optimizations.

Although the initial strategy of expressing and purifying the BaCBM2 protein with a cysteine solution was not tested, there is significant potential to integrate this approach in upcoming cycles to ensure greater accuracy in the corona deposition process.

The primary limitation identified was the lack of precision and reproducibility in quantifying MNPs through electrochemical measurements. This suggests that adjustments in the interaction time between the proteins and the MNPs, as well as the incubation time on the sensor surface, could enhance sensor performance in future cycles. In summary, this cycle paved the way for new investigations and optimizations, particularly concerning the behavior of protein coronas and their influence on electrochemical outcomes.