1 Basic Silinker
1.1 Modification of LPG
We expect the designed Silinker to be a dual-connector protein that can efficiently bind to the surface of mesoporous silica and the functional proteins we want to modify on silicon dioxide.
figure 1| Basic Silinker Functional Mode Diagram (gray squares represent silicon dioxide surface)
So we turned our attention to a dual-functional connector protein called linker protein G(LPG) that can also bind to the surface of silicon dioxide.
LPG is a bioconjugation platform based on SBP. LPG consists of two functionally distinct regions - the silicon dioxide binding peptide SBP (referred to as the “linker”) and the antibody binding protein (protein G, PG). In this platform technology, LPG serves as an anchor to rapidly and specifically immobilize antibodies on silicon dioxide surfaces, ensuring the functional display of the coupled antibodies without the need for complex conjugation chemistry. This connector technology has been applied in biotechnological and biomedical applications [1].
PG is a 25kDa cell wall protein derived from Streptococcus group G, which is a trimeric Fc receptor. Protein G specifically binds to the Fc region of IgG antibodies.
However, this protein linker can only bind IgG antibodies to silicon dioxide, which limits its applicability to functional proteins. We aim to develop a standardized connector with stronger affinity and broader applicability, capable of addressing most protein modification needs. Therefore, we modified LPG, both in terms of the SBP region and by replacing the PG protein [2].
1.2 Selection of SBP
Silica-binding peptide (SBP) greatly facilitates the binding of proteins to silicon surfaces through non-specific physical adsorption. The main mechanisms involved in SBP binding include electrostatic interactions, hydrophobic forces, and hydrogen bonding.
1. Electrostatic interactions:
The silicon dioxide surface carries a large negative charge, while the peptide residues carry positive charges, leading to electrostatic interactions between the two.
2. Hydrophobic interactions:
SBPs are designed to recognize and adsorb onto differently charged peptide ligands on silicon dioxide under different pH conditions. This is especially useful when the silicon dioxide surface is uncharged and requires positive charges for binding, particularly for neutral sequences.
3. Hydrogen bonding:
Polar groups in the peptide form hydrogen bonds with the silanol and siloxane groups on the silicon dioxide surface (residues such as Ser, His, and Arg).
Experimental evidence indicates that the relative orientation of the imidazole side chain on SBP and positively charged residues significantly influences the binding of SBP to silicon dioxide. Among various SBPs developed by researchers, we have chosen SBP5 for use in Silinker. SBP5 is derived from Car9 and the modification involves changing the orientation of the imidazole plane and introducing a positively charged histidine-rich region to mitigate the negative charge effects. This modification greatly enhances the binding capacity of SBP5.
Furthermore, compared to other SBPs, SBP5 exhibits binding affinity within an intermediate range, avoiding excessive binding strength that would render it insensitive to solution conditions [3].
1.3 SBP Oligomerization Modification
As a multivalent binding peptide, the silicon dioxide affinity of SBP is influenced not only by its monomeric structure but also by SBP oligomerization. Experimental studies have shown that 3×LPG outperforms 2×LPG in terms of binding strength, while 1×LPG does not bind to silicon dioxide. Moreover, substituting the spacer sequence in the LPG sequence with GGGGS has been found to improve the binding affinity. Literature suggests that replacing the three repeated spacer sequences with (GGGGS)12 significantly enhances the binding affinity to silicon dioxide, almost equivalent to the level of 4×LPG.
Therefore, we utilize the oligomerization effect of SBP and introduce a linker in LPG to increase its affinity for silicon dioxide [4].
We changed the original flexible linker to a rigid linker(EAAAK sequence, rigid structure, linked to SBP, reduces steric hindrance with silica dioxide, facilitating their binding), in an attempt to separate the SBP and the Streptavidin through the rigid structure of the linker, so that the SBP can effectively bind to the hydroxyl group on the silica surface and fully display the affinin domain.
1.4 Affinity Selection
Biotin-streptavidin interaction is known to be the strongest non-covalent binding interaction between a protein and ligand, exhibiting a binding affinity ten thousand times higher than that of antigen-antibody interactions, and the binding is rapid.
Biotinylation is the process of covalently attaching biotin to molecules, such as amino acids or proteins. Due to the small size of biotin, biotinylation is typically fast and specific, without interfering with the inherent functionality of the molecule. Biotin binding to streptavidin or anti-biotin proteins also exhibits high affinity, rapid binding kinetics, and high specificity. Therefore, biotinylated proteins are widely used in biomedicine and serve as valuable tools in biotechnology.
Among the biotin-binding protein family, unglycosylated streptavidin exhibits low levels of non-specific binding and minimizes interference in module assembly. Streptavidin is a tetrameric protein with a size of 66 kDa. One monomer of streptavidin binds to the imidazolone ring of biotin with high affinity and specificity, forming a complex with four biotin molecules. The affinity between streptavidin and biotin is extremely strong, with a dissociation constant in the range of 10-14 mol/L.
In the amino acid composition of streptavidin, glycine and alanine are abundant, and the active group that binds biotin is a tryptophan residue in the peptide chain. Under protein hydrolase action, streptavidin can break between amino acids 10-12 at the N-terminus and 19-21 at the C-terminus. The resulting core streptavidin retains the ability to bind biotin. The activity unit of streptavidin is also expressed as the amount required to bind 1 μg of biotin, with the highest activity of 18 U for 1 mg of streptavidin.
Considering the degree of streptavidin-biotin binding, expected experimental results, and the properties of the protein itself, we have selected streptavidin as the protein for binding SBP.
Figure 2| Structure of Streptavidin
Figure 3| Biotin-Streptavidin Conjugation Scheme
1.5 Streptavidin Modification
In subsequent experimental processes, we discovered that the tetrameric streptavidin could interfere with the function of the Twisted Silinker.
On the other hand, monomeric streptavidin possesses unique structural and functional characteristics. It allows for the detection of biotinylated ligands while avoiding target aggregation caused by multivalent binding. Aggregation of bioactive molecules often leads to functional changes, but monomeric streptavidin facilitates imaging and tracking studies of live cells by ensuring monovalent biotin binding. Therefore, to ensure the widespread application of Basic Silinker in the field of biomedicine, we eventually decided to use monomeric streptavidin as the connector (mSA).
Streptavidin (SA) and its homologs are widely used in research and biotechnology. However, traditional SA is a tetrameric protein that may have limitations in certain applications [5]. Moreover, modifications of traditional divalent or monovalent streptavidin often result in instability or reduced binding affinity. Through literature review, we found that scientists have developed a monovalent streptavidin, mSA, with a biotin affinity of 2.8 nM, which is currently the highest among monovalent streptavidins.
However, the expression of mSA in Escherichia coli forms inclusion bodies, limiting its usability [6]. Adding a protein tag can promote the soluble expression of mSA. It was experimentally found that mbp (maltose-binding protein) and trx (thioredoxin) were the most effective tags and resulted in higher expression levels.
Therefore, we decided to replace SA with mSA and add TrxA as our protein tag to enhance the soluble expression of mSA.
1.6 the overall structure of Basic Silinker
With the above design, we obtained the overall structure of Basic Silinker.
In pathway construction, we incorporated several elements to optimize the functionality and properties of the system. Firstly, TrxA was used as a solubility-enhancing tag for mSA [7]. This tag facilitated the soluble expression of mSA, overcoming the issue of inclusion body formation typically associated with mSA expression in Escherichia coli.
Additionally, we included a His tag, which served as a protein affinity tag. This tag enabled the purification and isolation of the recombinant protein using metal ion-affinity chromatography.
To facilitate downstream processing and protein manipulation, we incorporated a thrombin cleavage site. This site allowed for specific and precise cleavage of the fusion protein at the desired location [8]. It provided flexibility in protein engineering and modification after purification.
The use of monomeric streptavidin (mSA) was a crucial modification in the pathway. Unlike regular streptavidin with four biotin-binding sites, mSA has only one site, which greatly reduces the probability of inclusion body formation during expression [9]. This modification ensures the proper folding and functionality of mSA.
A rigid linker sequence, EAAAK, was introduced to connect the mSA and the silica binding peptide (SBP). This linker minimized steric hindrance and allowed for optimal binding between mSA and silica dioxide [10]. This interaction between mSA and silica is essential for subsequent applications using silica-based materials.
In summary, the pathway construction involved the strategic incorporation of TrxA as a solubility-enhancing tag, a His tag for protein affinity, a thrombin cleavage site for specific protein manipulation, monomeric streptavidin to prevent inclusion body formation, a rigid linker to optimize binding between mSA and SBP, and the SBP for binding to silica dioxide. These modifications and elements were carefully chosen to enhance the overall performance and utility of the constructed pathway.
Figure 4| Basic Silinker Pathway (TrxA: mSA solubility-enhancing tag [7]; His tag: protein affinity tag; thrombin site: thrombin cleavage site [8]; mSA: monomeric streptavidin with only one biotin-binding site, avoids inclusion body formation [9]; linker: EAAAK sequence, rigid structure, linked to SBP, reduces steric hindrance with silica dioxide, facilitating their binding [10]; SBP: silica binding peptide.)
Figure 5| Basic Silinker alpha fold 2 model
2 Cut Silinker(CS)
The Cut Silinker incorporates specific cleavage sites into the Basic Silinker, enabling the enzymatic release of the target protein. Furthermore, to achieve the replaceability of the cleavage sites, we employed inteins, self-cleaving and self-joining peptides, to fulfill this function. We constructed three fusion proteins, named N-terminal linker (mSA-linker), cut-linker, and C-terminal linker (SBP-linker), and subsequently verified their connectivity and enzymatic cleavage efficacy in subsequent experiments. The composition and functionalities of each fusion protein are elaborated below.
Figure 6| Functional schematic diagram of Cut Silinker
Figure 7| Cut Silinker Pathway (TrxA: mSA solubility-enhancing tag [7]; His tag: protein affinity tag; thrombin site: thrombin cleavage site [8]; TEV: TEV cleavage site; mSA: monomeric streptavidin with only one biotin-binding site, avoids inclusion body formation [9]; linker: EAAAK sequence, rigid structure, linked to SBP, reduces steric hindrance with silica dioxide, facilitating their binding [10]; SBP: silica binding peptide.)
2.1 Inteins
Inteins are sequences inserted within host proteins that resemble introns and exons in DNA, which can be cleaved and ligated with the flanking peptides on both ends (Figure 8, 2019 iGEM Pittsburgh). Considering the dynamics of splicing reactions and the orthogonality of inteins, we chose two inteins, GP41-1 and NrdJ, from the 2019 iGEM Pittsburgh team, which are mutually compatible and have excellent kinetic data, to achieve splicing connections between recombinant proteins.
Figure 8| Inteins schematic diagram (2019 iGEM Pittsburgh)
2.2 N-terminal linker (mSA-linker)
The N-terminal linker is a modified version of the Basic linker, divided into two parts while retaining the mSA portion, and still plays a role in constructing an avidin-biotin affinity system. The C-terminus of the recombinant protein contains a segment with the N-terminal sequence of GP41-1 (BBa_K3308067), which can be connected to the N-terminal of GP41-1C of the Cut linker (BBa_K3308068). Additionally, a TrxA solubility-enhancing protein tag is added to the N-terminus to improve protein solubility for expression. The His-tag is used for purification and separation, while the thrombin site is employed to remove the tag after purification.
2.3 Cut linker
The Cut linker consists of the C-terminal sequence of the GP41-1 intein (BBa_K3308068), the N-terminal sequence of the NrdJ intein (BBa_K3308069), and a variable peptide segment. The C-terminal sequence of the GP41-1 intein is used to connect with the N-terminal linker (mSA-linker), while the N-terminal sequence of the NrdJ intein is used to connect with the C-terminal linker (SBP-linker). The variable peptide segment enables recognition and cleavage by protein cutting enzymes in specific environments, releasing the biotin-modified functional protein that has been cleaved and bound with mSA. To facilitate experimental verification, we have chosen a matrix metalloproteinase 2 (MMP2) cleavable short peptide (PLGVR) as the variable peptide segment. MMPs are a family of zinc-dependent endopeptidases that are overexpressed in the extracellular environment of certain tumors. The Cut Silinker, modified with PLGVR, can be recognized and cleaved by MMP2, releasing the functional protein. In addition, we have incorporated a TEV recognition site and a His tag for fusion protein purification [11].
2.4 C-terminal linker (SBP-linker)
The C-terminal linker is a modified version of the Basic linker, divided into two parts while retaining the SBP portion, which interacts with mesoporous silica surfaces.
Since the connection of arbitrary enzymatic cleavage sites requires the use of intein self-cleavage, the design of the CS linker includes two types of intein sequences. In the SBP-linker, the N-terminal of the NrdJ-C intein (BBa_K3308070) is connected to the N-terminus of the recombinant protein and can be linked to the C-terminus of the NrdJ-N intein (BBa_K3308069), present in the Cutting linker.
The design of NrdJ is based on the Pittsburg 2019 Intein Separation Logic Gate System (BBa_K3308007), which is found in our team’s parts library. NrdJ-C is essential for the self-splicing of the NrdJ intein and retains the functional splicing ability of the GP41-1 intein. It works in collaboration with GP41-1 to introduce the intermediate cleavage site into the silinker structure. The rigid linker and SBP are designed at the N-terminus of NrdJ-C, while the C-terminus of NrdJ-C includes a TEV recognition site and a HisTag purification tag.
In summary, the Cutting Silinker is a standardized adapter designed to meet the diverse protein cleavage requirements in various microenvironments. It enables interchangeable connections with the mSA-linker and SBP-linker by designing proteins with intein sequences at both ends.
3 Twisted Silinker(TS)
In addition to the versatile Basic Silinker and the Cut Silinker for cleavage purposes, we aim to design an intelligent Silinker capable of sensing environmental changes and interacting with the surroundings. Hence, we have developed a Twisted Silinker, which possesses the ability to bind specific ions. When the Silinker binds with these ions, its structure undergoes a conformational change, resulting in a twisted form. As the name suggests, the Twisted Silinker aims to bring the target protein closer to the silica dioxide carrier, facilitating the release of the delivered cargo in response to different environmental stimuli.
Figure 9| Functional schematic diagram of Twisted Silinker
3.1 Calmodulin
In our research, we found that calmodulin (CaM) can be used for our Twisted Silinker (TS). When the concentration of Ca2+ increases in the environment, the conformation of calmodulin is induced to change. The peptide region of calmodulin is more prone to folding upon binding with Ca2+, which brings the target protein closer to mesoporous silica and affects the release of substances in mesoporous silica.
Similarly, this calcium ion sensing segment can be replaced with other ion-sensing segments to achieve a wider range of functions. For example, sensing pH changes in the tumor microenvironment (such as pHLIPs), which is a classic sensing approach for nanodrug delivery.
3.2 mSA
Streptavidin (SA) and its variants are widely used in research and biotechnology. There are many choices for SA currently available, but when selecting components, we prefer a biotin-avidin complex system that has minimal impact on the target protein and Silinker functionality. However, we found that traditional SA is a tetrameric protein that can bind to calmodulin, causing functional abnormalities. Similarly, bivalent or monovalent streptavidin modifications often face issues of instability or low affinity. In the end, we chose a monovalent streptavidin variant called mSA, which has been developed by scientists through amino acid mutations and has the highest affinity for biotin among current monovalent streptavidins, with an affinity of 2.8nM. As a monomeric streptavidin, we believe it has great potential in the Silinker system.
3.3 Rigid linker
We chose a rigid linker instead of a flexible linker because during the modeling process, we found that the flexible linker tends to fold in its natural state, which would bring the calmodulin, which we want to be in an extended state, into a tightly bound state.
Figure 10| Twisted Silinker alpha fold 2 model
Figure 11| Twisted Silinker Pathway (TrxA: mSA solubility-enhancing tag [7]; His tag: protein affinity tag; thrombin site: thrombin cleavage site [8]; mSA: monomeric streptavidin with only one biotin-binding site, avoids inclusion body formation [9]; linker: EAAAK sequence, rigid structure, linked to SBP, reduces steric hindrance with silica dioxide, facilitating their binding [10]; SBP: silica binding peptide.)
In addition to this, we created TS-GFP to help us visualize protein function during the design session for functional verification. It replaced the linker sequence with GFP, which can be excited to emit green fluorescence when there is a conformational change between the calcium-binding peptide and calcium-binding protein [16]. This can facilitate the easier verification of whether other components of Twisted Silinker can successfully perform their functions.
Figure 11plus｜Twisted Silinker-GFP Pathway (TrxA: mSA solubility-enhancing tag [7]; His tag: protein affinity tag; thrombin site: thrombin cleavage site [8]; mSA: monomeric streptavidin with only one biotin-binding site, avoids inclusion body formation [9]; GFP: verify the proper folding of structure[15]; SBP: silica binding peptide
4 Pairing Silinker(PS)
Within cells, protein dimerization is common and plays a significant role in substance conversion, energy transformation, and signal transduction. Therefore, we aim to design a Pairing Silinker that can bind to silica nanoparticles and achieve diverse functionalities through dimerization. Each Pairing Silinker is equipped with specific ion binding sites, allowing the distance between two proteins to be regulated by changes in substance concentration.
Taking into account the application scenarios, calcium ions are widely present in living organisms and serve various functions as intracellular signaling ions. Therefore, we chose a calcium ion binding peptide as the functional element [13] to construct a dimeric structure resembling “P1…Ca2+…P2” (non-covalent binding between calcium ions and peptides) to achieve the primary functionality of the Pairing Silinker.
Figure 12| Functional mode diagram: When the calcium ion concentration reaches a certain threshold, the calcium ion binding peptide will undergo dimerization, thereby bringing the target proteins closer together. When the calcium ion concentration is below this threshold, the calcium ion binding peptide cannot dimerize, and the target proteins move away from each other.
Figure 13| Pathway diagram:(TrxA: TrxA solubility-enhancing tag, increasing protein solubility [7]; His tag: Protein affinity tag; thrombin site: thrombin cleavage site [8]; mSA: While conventional streptavidin has four biotin binding sites, mSA has only one, which avoids the formation of inclusion bodies [9]; CBP1: Calcium ion binding peptide; linker: EAAAK sequence, rigid structure, connecting with SBP, reducing steric hindrance between linker and silica dioxide, facilitating their binding [10]; SBP: Silica-binding peptide.)
Meanwhile, the potential of Pairing Silinker is limitless, depending on the type of target protein. Here, we will briefly discuss two examples: enzyme activation and fluorescence resonance.
4.1 Enzyme Activation [14]
Under physiological conditions, caspase-9 exists as an inactive monomer. During apoptosis, caspase-9 forms a complex with the cofactor Apaf-1 in the presence of cytochrome c and ATP, generating the “apoptosome”. This oligomeric complex co-localizes with multiple caspase-9 molecules, leading to an increase in the local concentration of caspase-9 to its dissociation constant (Kd), triggering homodimerization through a “self-siphoning” mechanism and inducing enzyme activation. This interaction stabilizes the activation loop and allows the active site residues to adopt a substrate-binding conformation [1]. Subsequently, caspase-9 can provide proteolytic activity.
Therefore, the use of Pairing Silinker can facilitate caspase-9 dimerization, promoting apoptotic cell death and providing a novel approach for cancer treatment.
4.2 Fluorescence Resonance [15]
Based on the principle of fluorescence resonance energy transfer (FRET), when the distance between a donor fluorophore and an acceptor chromophore is within a certain range, the fluorescence emitted by the excited donor can be transferred to the acceptor molecule. Thus, utilizing Pairing Silinker can bring proteins on the surface of silica nanoparticles closer, activating fluorescence. This can optimize various fluorescence-based applications such as fluorescent probes.
5 Nucleotide Silinker(NS)
During the design process of the Silinker family, we have observed that for Silinker variants responding to different signals, SBP and mSA remain constant, with only the linker region between them being changed. Therefore, modular design that enables orthogonal connections between SBP, linker, and mSA can simplify the design and production process of Silinker. Moreover, in the expression and purification of Silinker proteins, we have found that using a protein as a linker necessitates the construction of plasmids, bacterial transformation, and extraction and purification of the target protein. This process becomes challenging for proteins with high hydrophobicity or an inability to form specific spatial structures (e.g., inclusion bodies or low expression levels) in Escherichia coli, thereby hindering cost-effective production on a large scale.
Nucleic acids, as biologically produced molecules that can be rapidly and efficiently generated, hold great promise as ideal linkers for Silinker. Nucleic acids possess excellent programmability, with the combination of only four nucleotide bases enhancing their predictable structural properties. Additionally, the increasing number of nucleic acid aptamers and nucleic acid molecular switches signifies the potential of these molecules for signal-responsive applications.
To achieve the design of nucleic acids as signal-responsive linkers, it is essential to establish specific and orthogonal connections between functional nucleic acids and both the SBP and new protein adapters. This implies the need to explore nucleic acid-protein conjugation methods beyond biotin, with relatively strong binding affinity. We have identified a nucleic acid-protein conjugation approach: TrwC, a type of relaxase found in bacteria, possesses an N-terminal sequence that can covalently link to the 5’ end of specific DNA sequences. Therefore, we have engineered a fusion protein of mSA-TrwC.
Figure 14| Pathway diagram:(mSA: While conventional streptavidin has four biotin binding sites, mSA has only one, which avoids the formation of inclusion bodies [9]; TrwC, a type of relaxase found in bacteria, possesses an N-terminal sequence that can covalently link to the 5’ end of specific DNA sequences. )
Figure 15| Nucleotide Silinker alpha fold2 model
The functionality of DNA can be harnessed as a signal-responsive linker by connecting it to the 3’ end of biotinylated streptavidin (BS) and using the mSA-TrwC fusion protein to recognize a specific sequence at the 5’ end of the DNA. By reintroducing the mSA adapter into Silinker, the formation of a BS-DNA-NS complex can occur, allowing the functional nucleic acid to exert its specific function as a signal-responsive linker.
Figure 16| Functional schematic diagram of Nucleotide Silinker
