Encapsulation of dimer or tandem-enzymes within capsid proteins, in particular within engineered encapsulins

Abstract

The present invention relates to a nanoparticle, comprising an assembly of bacterial encapsulin protein monomers, and a protein located inside the nanoparticle, wherein the protein is a dimer or tandem-enzyme wherein each enzyme of the dimer or tandem is fused to at least one terminal targeting peptide (TP). The invention further relates to nanoparticles comprising modified encapsulin proteins, and methods and uses of the nanoparticles in medicine, industry, cleaning and biotechnology.

Description

[0001] ENCAPSULATION OF DIMER OR TANDEM-ENZYMES WITHIN CAPSID PROTEINS, IN PARTICULAR WITHIN ENGINEERED ENCAPSULINS

[0002] The present invention relates to a nanoparticle, comprising an assembly of bacterial encapsulin protein monomers, and a protein located inside the nanoparticle, wherein the protein is a dimer or tandem-enzyme wherein each enzyme of the dimer or tandem is fused to at least one terminal targeting peptide (TP). The general structure of a dimer is monomer-TP, while the general structure of a tandem is monomer-linker-monomer-TP. The invention further relates to nanoparticles comprising modified encapsulin proteins, and methods and uses of the nanoparticles in medicine, industry, cleaning, and biotechnology.

[0003] Background of the invention

[0004] Many applications of enzymes in biotechnology necessitate the use of an enzyme that remains stable over extended periods, even in challenging environments. Various encapsulation techniques have been developed to protect enzymes, offering potential benefits such as functionalization for purposes like immobilization onto solid supports or targeted delivery to specific tissues (17-20). Typically, these techniques involve post-purification modification of the enzymes using structures like liposomes, polysaccharides or organic frameworks. Such manipulations may affect enzyme folding and activity in spite of some encouraging results.

[0005] Nitroreductases, versatile enzymes found across various organisms, play pivotal roles in diverse applications spanning bioremediation, cancer therapy, antibiotic production, genetic engineering, biosensors, biocatalysis, and beyond (1). These enzymes are typically FMN- dependent and catalyze the reduction of nitro aromatic or nitro heterocyclic compounds by via a ping-pong bi-bi mechanism to yield corresponding hydroxylamines or amines.

[0006] In the first of two sequential reactions, NADH or NADPH reduces the coenzyme FMN, releasing NAD(P)+. In the second reaction, the reduced FMN transfers electrons to the substrate's nitro groups. Historically, Escherichia coli (E. coll) nitroreductases have been classified into oxygeninsensitive (type I) and oxygen-sensitive (type II) enzymes, which mediate two-electron and single-electron transfers, respectively. NfsB from E. coli is a key representative for the type I enzyme class. Although the physiological roles of nitroreductases are still unknown, these enzymes exhibit broad substrate specificity and catalytic activity toward both natural and synthetic compounds, which makes them highly attractive for various applications. In environmental bioremediation, they detoxify nitroaromatic pollutants, (2, 3) while in cancer therapy, they selectively activate prodrugs within tumor cells, or NfsB and NfsA can also activate antimicrobials like Eeyarestatin 1 (ESI), which is one the rare inhibitors of the bacterial Sec protein channel (4-11). Additionally, these enzymes are integral in the biosynthesis of antibiotics and serve as selectable markers in genetic engineering (12, 13). Nitroreductases also find utility in biosensors for pollutant detection and are employed as biocatalysts for organic synthesis, contributing to pharmaceutical intermediates and fine chemicals (6, 14-16). Their multifaceted functions underscore their significance in biotechnology, medicine, environmental science, and chemical synthesis.

[0007] Cytosine-desaminases may be of use in diagnosis and therapy.

[0008] Many enzymes are used in laundry detergents for an efficient cleaning, such as amylases, lipases, or cellulases. In order to be most effective, these enzymes need to be protected in the harsh environment of washing, e.g. from proteases that also form part of laundry detergents. Encapsulins, self-assembling icosahedral protein nanocages derived from prokaryotes, represent a versatile set of tools for nanobiotechnology. Uniquely, encapsulins selectively selfassemble around cargo enzymes tagged with short, conserved peptide sequences at the termini of cargo proteins — referred to as targeting peptides (TPs) packaging them, also designated as small encapsulation signal peptides (ESig) or encapsulin localization sequence (ELS). In contrast to all other known microbial protein compartments or organelles, the encapsulin shell protein shares the HK97 (Hong Kong 97) phage-like fold, pointing to an evolutionary connection with the world of bacteriophages and viruses.

[0009] Studies show that purified encapsulin derived from Thermotoga maritima offers key advantages, including colloidal stability, storability, and compatibility with blood upon intravenous injection into animal models. Furthermore, it demonstrates an excellent nanosafety profile in mice, with no significant weight loss or adverse pathology observed (70). Encapsulin, like other intravenously applied protein nanocages (PNCs), can induce the production of PNC-specific antibodies, which play a key role in their clearance from the body. Enhancing the therapeutic viability of these nanocages can be achieved by using site-directed mutagenesis to modify their immunogenicity, specifically by removing certain B- and T-cell epitopes (71). Encapsulin, in particular, shows a remarkable ability to tolerate various amino acid modifications, making it a promising candidate for developing more effective and less immunogenic therapeutic agents (68, 72-77).

[0010] Beyond this, the exceptional amenability of encapsulins to genetic engineering allows researchers to modify both the interior and exterior of the protein shell for diverse biotechnological and therapeutic applications. By incorporating alternative cargo proteins through an ELS, encapsulins can be repurposed to compartmentalize enzymes employed in ADEPT, providing a protective environment that enhances enzyme stability and function. Meanwhile, their exterior surface can be functionalized with targeting moieties to enable tumorspecific delivery.

[0011] Among the various targeting moieties, affibodies stand out for their compact size, high target affinity, and efficient bacterial expression, making them ideal for functionalizing protein nanocages to achieve tumor specificity. Derived from the IgG-binding domains of Staphylococcal protein A, affibodies are small (~58 amino acids) proteins arranged in a stable three-helix bundle (78).

[0012] Human Epidermal Growth Factor Receptor 2 (HER2) is a critical member of the epidermal growth factor receptor (EGFR) family involved in the regulation of cell growth, proliferation and differentiation (52). HER2 is overexpressed in a subset of cancers (53-56), particularly breast and ovarian cancer, which significantly impacts prognosis and treatment strategies (57, 58). Studies indicate that patients with HER2-positive tumors experience poorer outcomes due to increased invasiveness, higher recurrence rates, and a greater propensity for metastasis compared to HER2-negative cases (57, 59). In clinical practice, the identification of HER2 overexpression plays a crucial role in guiding therapeutic decisions. Trastuzumab (Herceptin), a monoclonal antibody that specifically targets the HER2 receptor, has been pivotal in the management of HER2-positive breast cancer. Its use has significantly improved patient outcomes by inhibiting HER2-mediated signaling pathways and promoting antibody-dependent cellular cytotoxicity (60). However, resistance to trastuzumab is a significant clinical challenge, as some patients experience primary resistance while others develop secondary resistance during treatment. This necessitates the exploration of alternative targeted therapies to overcome these limitations.

[0013] HER2-specific affibodies have been explored for targeted delivery of small drug molecules, (79, 80) photosensitizers (81), imaging reagents (82-84), enzymes (85, 86), and organometallic catalysts (87), as well as for the delivery of nanoparticles and other synthetic carriers (74, 88- 90).

[0014] One approach of enhancing the efficacy of targeted therapies involves the covalent conjugation of active drug molecules to antibodies or other targeting moieties, creating antibody-drug conjugates (ADCs) (61, 62). While ADCs hold promise their delivery is inherently limited by the finite number of drugs that can be covalently bound to the targeting molecule (63, 64). This limitation can be overcome by using enzyme-activated prodrugs as a viable alternative (65). Cancer-targeted enzymes can convert a prodrug into a greater quantity of active drug molecules at the tumor site, thereby maximizing therapeutic efficacy and potentially minimizing the systemic toxicity associated with traditional chemotherapy. However, a systemic application of enzymes presents challenges, particularly regarding enzyme stability and the immunogenicity of non-human enzymes used in Antibody -Directed Enzyme Prodrug Therapy (ADEPT). Recent advances indicate that encapsulating enzymes in nanoscale carriers may help overcome these challenges (66, 67), with encapsulins emerging as a promising candidate.

[0015] WO 2017 / 182760A1 discloses encapsulin-based nanoparticles and discloses an assembly of encapsulins fused to a peptide located at the outside of the nanoparticle and a protein inside the nanoparticle.

[0016] US 2019-0276501A1 discloses an encapsulin microcompartment which differs in its features from the encapsulin particles of the present invention.

[0017] WO 2002 / 062937 relates to a washing and cleaning agent, whereby the washing and cleaning agent comprises fine microparticles, said fine microparticles themselves comprising one or several washing and cleaning agent components. US 2023-0145060A1 discloses self-assembling nanocage monomers, including encapsulin, fused to a bioactive moiety.

[0018] EP3661968 discloses a fusion protein comprising a nanocage monomer; and an antibody or fragment thereof linked to the nanocage monomer.

[0019] Finally, US7687474B2 discloses a nucleotide-based method to treat cancer locally by administering a chromium reductase and a prodrug to be activated.

[0020] It is an object of the present invention to provide improved methods for the encapsulation to maintain and improve the properties, in particular the stability and activity, of enzymes. The encapsulated enzymes can be used in biotechnology, medicine, environmental science, and chemical synthesis. Other objects and advantages will readily become apparent for the person of skill from studying the following more detailed description and examples.

[0021] In a first aspect thereof, the present invention solves the above object by providing a nanoparticle, such as a capsid, comprising an assembly of bacterial encapsulin protein monomers, and an internal protein located inside the nanoparticle, wherein the internal protein is a dimer of an enzyme and / or a tandem-enzyme, wherein each enzyme of the dimer or tandem is fused to at least one terminal targeting peptide (TP).

[0022] In the context of the present invention the terms “capsid” and “nanoparticle” are used interchangeably to designate a hollow particle of more or less spherical shape that is assembled exclusively or essentially of bacterial encapsulin protein monomers and comprises pores. The lumen of the particle may contain a protein or proteins as a payload, such as the nitroreductase enzyme.

[0023] In principle any suitable encapsulin can be used that is able to form capsids / nanoparticles with pores. Examples may be found in the literature, such as in, for example, Andreas, M.P., Giessen, T.W. Large-scale computational discovery and analysis of virus-derived microbial nanocompartments. Nat Commun 12, 4748 (2021). Preferred is the nanoparticle according to the present invention, wherein the encapsulin protein is selected from an encapsulin protein of Mycobacterium smegmatis, Thermotoga marilima. Myxococcus xanlhus. Quasibacillus thermololerans. Haliangium ochraceum. Acinetobacter baumannii. Mycobacterium leprae, Mycobacterium avium, Synechococcus elongatus, Burkholderia cepacia, and Klebsiella pneumoniae.

[0024] The encapsulin protein of the nanoparticle according to the present invention may be modified and may comprise a mutated pore-forming loop increasing the diffusion rate through the pores in the nanoparticle, for example by introducing amino acid substitutions and / or deletions.

[0025] The dimer or tandem-enzyme according to the present invention comprises enzymes selected from the group consisting of cytosine-desaminases, nitroreductases, and enzymes used in laundry detergents, such as amylases, lipases, or cellulases, wherein the enzymes in particular are selected from wildtype and recombinant bacterial enzymes, such as, for example, the NfsB of E. coli. The present inventors engineered a tandem enzyme by linking two copies of the enzyme, such as NfsB or cytosine desaminase with a flexible linker peptide sequence (as illustrated in Figure 1 A, and see, for example, SEQ ID NOs: 7 or 8) to ensure dimer formation and thus proper dimer incorporation into the encapsulin structure.

[0026] Another aspect of the present invention relates to the nanoparticle according to the present invention fused to an affibody, such as a Human Epidermal Growth Factor Receptor 2 (HER2)- specific affibody. The affibody may be fused at its N- or C-terminus to the nanoparticle according to the present invention. The construct may be used in an affibody-directed enzyme- prodrug therapy targeting cancers, such as HER2-overexpressing cancers, by delivering enzymes, such as cytosine deaminase to convert 5 -Fluorocytosine (5 -FC) to 5 -Fluorouracil (5- FU) in situ. Another aspect of the present invention then relates to an at least bicistronic vector comprising a first nucleic acid sequence encoding an encapsulin protein as defined in the context of the present invention, and a second nucleic acid sequence encoding at least one enzyme of the dimer or the tandem-enzyme as defined in the context of the present invention.

[0027] Another aspect of the present invention then relates to a host cell comprising the at least one bicistronic vector according to the present invention.

[0028] Yet another aspect of the present invention then relates to a method for producing the nanoparticle according to the present invention, comprising recombinantly expressing the encapsulin and the enzyme of the dimer or the tandem-enzyme in the cell according to the present invention, and purifying the resulting nanoparticles or capsids comprising the incorporated dimer or tandem-enzyme.

[0029] Yet another aspect of the present invention then relates to a method for converting a substrate in vitro or in vivo, comprising providing the nanoparticle according to the present invention, comprising a dimer or tandem-enzyme, contacting the nanoparticle with at least one suitable substrate to be converted, and converting said substrate. Preferably, the method provides a substrate conversion that is similar or higher than a conversion of the respective nonencapsulated enzyme, in particular the dimer or tandem-enzyme thereof.

[0030] Another aspect of the present invention then relates to a pharmaceutical composition, comprising the nanoparticle according to the present invention, and at least one suitable pharmaceutically acceptable carrier.

[0031] Another aspect of the present invention then relates to the pharmaceutical composition according to the present invention for use in the treatment of diseases, such as for the use of treating cancer, comprising administering the pharmaceutical composition in combination with at least one anti-cancer prodrug, such as, for example, Nbzp (see, for example, Tugba Gungor, et al. PRODRUGS FOR NITROREDUCTASE BASED CANCER THERAPY- 2: Novel amide / Ntr combinations targeting PC3 cancer cells, European Journal of Medicinal Chemistry, Volume 171, 2019, Pages 383-400) that can be enzymatically reduced into the pharmaceutically active drug. The cancer may be lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

[0032] Another aspect of the present invention then relates to a method for treating diseases, such as cancer, comprising administering the pharmaceutical composition according to the present invention in combination with at least one anti-cancer prodrug, such as, for example, Nbzp or 5 -fluorocytosine (5-FC) that can be enzymatically converted into the pharmaceutically active drug to a patient or subject in need for such a treatment. The cancer may be lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

[0033] Another aspect of the present invention then relates to the use of the nanoparticle according to the present invention or the pharmaceutical composition according to the present invention for detoxifying environmental nitroaromatic pollutants, for cleaning of starches, fats, or cellulose, the biosynthesis of antibiotics, selectable marker in genetic engineering, as a biosensor for pollutant detection or a biocatalyst for the organic synthesis of pharmaceutical intermediates or fine chemicals.

[0034] As mentioned, the problem of the present invention is solved by providing a nanoparticle comprising an assembly of bacterial encapsulin protein monomers. Furthermore, an internal protein is located inside the nanoparticle that is selected from a dimer of an enzyme and / or a tandem-enzyme. Each enzyme of the dimer or tandem is preferably fused to at least one terminal targeting peptide (TP).

[0035] Protein nanoparticles or capsids, such as encapsulins, can be viewed as provided a genetically encoded enzyme encapsulation, and are particularly suitable for immobilization of selected enzymes within physiological conditions (21-24). Encapsulins are prokaryotic protein-based nanocompartments. These structures self-assemble into protein shells ranging from 24 to 42 nm in diameter, characterized by the viral HK97-fold of their shell protein. In nature, these nanocompartments can encapsulate specific cargo proteins, such as ferritin-like proteins, peroxidases, and desulfurases.

[0036] The capsids, comprised of 60 - 240 subunits arranged in an icosahedral structure, engage with a conserved C-terminal encapsulin localization sequence (ELS), or a more extended N-terminal encapsulation-mediating domain, found on the guest protein. This sequence, also referred to as a targeting peptide (TP) or cargo-loading peptide, effectively guides the guest protein into the interior of the capsid during its self-assembly process (25, 26).

[0037] Protein nanocompartments are present across various bacterial and archaeal phyla, and they are apparently involved in functions such as iron storage, resistance to oxidative stress, anaerobic ammonium oxidation, and sulfur metabolism. Recent genome-mining research has categorized encapsulins into four distinct families (type I - IV), which differ in their sequences, structures, and operon organizations.

[0038] Encapsulins are natural protectors against harsh conditions such as proteases, heat, and pH change (21, 22, 27-31). The inventor’s recent research had demonstrated the remarkable ability of encapsulins to safeguard chemical catalysts within cellular environments (32, 33). Moreover, the versatility of encapsulins extends to their outer surface, which can be engineered for diverse applications, such as affinity chromatography, immobilization, or cellular targeting (34-38).

[0039] Also, importantly, engineered encapsulins can be completely genetically encoded and produced in abundance via recombinant expression in E. coli.

[0040] Nitroreductases are important enzymes for a variety of applications, including cancer therapy and bioremediation. Encapsulation was used in order to improve stability and activity. The inventors focused on genetically encoded encapsulation of nitroreductases within protein capsids, like encapsulins. The inventor’s experiments showcase the preferred encapsulation of nitroreductase NfsB and a cytosine-desaminase as functional dimers, such as tandem enzymes, within selected encapsulins, which enhances protein activity and stability in diverse conditions.

[0041] Building on the inventor’s previous experiments on the encapsulation of organometallic catalysts, the inventors utilized encapsulin (UniProt ID: I7G8Y9) from the soil organism Mycobacterium smegmatis (M. smegmatis). This encapsulin is a type I family member with T1 symmetry.

[0042] In the inventor’s experiments, tandem NfsBs and cytosine-desaminases were designed that exhibited enhanced activity and stability across challenging conditions. Using cryogenic electron microscopy (cryo-EM), the inventors extensively characterized the encapsulated nitroreductase and demonstrated its activity against various substrates, which can diffuse through the pores of the capsids to reach NfsB. Furthermore, the inventors showed that encapsulated enzymes, such as tandem nitroreductases or cytosine-desaminases can activate cancer prodrugs in the extracellular environment, providing a promising strategy for enzyme- directed prodrug activation in future therapeutic applications.

[0043] Kwon et al. (in: Pore Engineering as a General Strategy to Improve Protein-Based Enzyme Nanoreactor Performance, ACS Nano 2024, 18, 25740-25753) disclose a pore-engineered universal enzyme nanoreactor platform based on encapsulins-microbial self-assembling protein nanocompartments with programmable and selective enzyme packaging capabilities. They propose a pore engineering strategy focused on increasing shell porosity. By increasing pore size, they aimed to optimize molecular flux of a broad range of enzyme substrates and cofactors across the protein shell. At the same time, the pore design is said to take into account that very large pores (>3 nm) can lead to the loss of many advantages of enzyme sequestration. They introduced a loop deletion (red) connecting A-domain helices a6 and a7 (amino acids 195 to 201) and replaced amino acids 194 and 202 with glycine in the encapsulin of Myxococcus xanthus.

[0044] In general, any suitable encapsulin protein can be used for the present invention. Nevertheless, in particular in view of the above, preferred is the nanoparticle according to the present invention, wherein the bacterial encapsulin protein is selected from a type I encapsulin, preferably forming a T1 symmetric particle.

[0045] Further preferred is the nanoparticle according to the present invention, wherein the encapsulin protein is selected from an encapsulin protein of Mycobacterium smegmatis, Thermotoga marilima. Myxococcus xanthus, Quasibacillus thermotolerans, Haliangium ochraceum, Acinetobacter baumannii, Mycobacterium leprae, Mycobacterium avium, Synechococcus elongatus, Burkholderia cepacia, and Klebsiella pneumoniae.

[0046] Mutations within the pore region were found to be beneficial for activity of the encapsulated enzyme by increasing diffusion rates, e.g. of products, substrates and cofactors. Cryogenic electron microscopy revealed the overall architecture of the encapsulated dimeric NfsB within the nanoreactor environment and identified multiple pores in the shell. These findings highlighted the potential of encapsulins as versatile tools for enhancing enzyme performance across various fields.

[0047] Therefore, in a particular aspect of the present invention, the nanoparticle according to the present invention comprises an encapsulin protein comprising a mutated pore-forming loop leading to an increase of the diffusion rate through the pores in the nanoparticle. A mutated pore-forming loop may also lead to an increase of the size of the pore and may be achieved as described herein, for example by introducing amino acid substitutions and / or deletions.

[0048] More preferred is the nanoparticle according to the present invention, wherein the encapsulin protein is selected from AT. smegmatis comprising a mutated pore-forming amino acid sequence VSETTAHGYPIRE (Seq ID NO: 2) wherein the mutated pore-forming amino acid sequence is selected from the amino acid sequences VSETTAAAIRE (Seq ID NO: 3), VSETTAAIRE (Seq ID NO: 4), VSETAAAIRE (Seq ID NO: 5), and VSETAAIRE (Seq ID NO: 6). Included in the present invention are also mutated sequences comprising one or two amino acid exchanges with amino acids that have similar properties and / or characteristics, such as glycine for alanine, and the like.

[0049] According to the present invention, preferred is the nanoparticle according to the present invention, wherein at least one marker or label sequence is fused to the N- and / or C-terminus of the encapsulin protein monomer, preferably selected from a His-tag, a Strep-tag, and a maltose binding protein (MBP) affinity tag.

[0050] According to the present invention, preferred is the nanoparticle according to the present invention, wherein the dimer or tandem-enzyme comprises enzymes selected from the group consisting of cytosine-desaminases, nitroreductases, and enzymes used in laundry detergents, such as amylases, lipases, or cellulases, wherein the enzymes in particular selected from wildtype and recombinant bacterial enzymes, such as, for example, the NfsB of E. coli.

[0051] According to the present invention, any suitable nitroreductase protein or enzyme may be used in the nanoparticle according to the present invention. Preferably, the nitroreductase protein is selected from wildtype and recombinant bacterial nitroreductase proteins, such as, for example, the NfsB of E. coli as described herein. Preferably, the enzyme is used as a dimer or a linked tandem NfsB (tdNfsB) form thereof. According to the present invention, a preferred tandem enzyme was created by linking two NfsB copies with a flexible linker sequence as illustrated in Figure 1 A and SEQ ID NO: 7 to ensure obligate dimer formation and thus incorporation into the encapsulin structure. Furthermore, a preferred tandem enzyme was created by linking two cytosine-desaminases as illustrated in SEQ ID NO: 8 to ensure obligate dimer formation and thus incorporation into the encapsulin structure. The flexibility of the linker is determined by length and sequence content and glycine-serine (GS) linkers are commonly preferred based on their highly flexible profiles.

[0052] According to the present invention, any suitable targeting peptide (TP) sequence may be used with the present enzymes of the dimer or tandem-enzyme. Preferred is the nanoparticle according to the present invention, wherein the at least one terminal targeting peptide (TP) sequence is fused to the N- and / or C-terminus of the enzymes of the dimer or tandem-enzyme. A preferred TP is the ELS sequence from smegmatis and T. marilima. as described herein. It was found in the context of the present invention that the protein concentration inside the nanoparticle according to the present invention is higher than 2 mM, preferably higher than 3 mM, such as between 3 and 3.5 mM. This is achieved, for example, because of the tandem enzymes as designed according to the present invention.

[0053] Another aspect of the present invention then relates to an at least bicistronic vector comprising a first nucleic acid sequence encoding an encapsulin protein as defined in the present invention, and a second nucleic acid sequence encoding at least one enzyme of the dimer or the tandemenzyme as defined in the present invention. An example of such a vector is shown in Figure 1, and based on the pETDuet™-l vector. Another aspect of the present invention then relates to a host cell comprising the at least one bicistronic vector according to the present invention. Preferred are bacterial or mammalian host cells.

[0054] Another aspect of the present invention then relates to a method for producing the nanoparticle according to the present invention, comprising recombinantly expressing the encapsulin and the enzyme of the dimer or the tandem-enzyme in the host cell according to the present invention, and purifying the resulting nanoparticles or capsids comprising the incorporated dimer or tandem-enzyme. Respective methods are described herein, and described in the art.

[0055] Yet another aspect of the present invention then relates to a method for converting a substrate in vitro or in vivo, comprising providing the nanoparticle according to the present invention comprising a dimer or tandem-enzyme, contacting the nanoparticle with at least one suitable substrate to be converted, and converting said substrate.

[0056] Preferred is the method according to the present invention, wherein the non-reduced substrate is a prodrug, such as, for example, Nbzp that can be enzymatically reduced into the pharmaceutically active drug or 5 -fluorocytosine (5-FC) that can be enzymatically converted into the pharmaceutically active drug 5 -fluorouracil (5-FU), or the substrate is a nitroaromatic environmental pollutant or the substrate is a selectable marker in genetic engineering or a starch, fat, or cellulose.

[0057] Yet another aspect of the present invention then relates to a method for reducing nitroaromatic compounds in vitro or in vivo, comprising providing the nanoparticle according to the present invention in vitro or in vivo, contacting the nanoparticle with at least one nitroaromatic substrate suitable for reduction, and reducing said nitroaromatic substrate. Respective exemplary methods and substrates are also described herein, and also described in the art.

[0058] It was found in the context of the present invention that the inventive method can provide an enzymatic substrate conversion that is similar or even higher than a conversion of the respective non-encaspulated dimer or tandem-enzyme(s).

[0059] Preferred is the method according to the present invention, wherein the non-reduced substrate is a prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug, or the substrate is a nitroaromatic environmental pollutant or the substrate is a selectable marker in genetic engineering.

[0060] In a further aspect, provided are pharmaceutical compositions comprising the nanoparticle according to the present invention, and at least one suitable pharmaceutically acceptable carrier. In some embodiments, the composition comprises an aqueous formulation. In some embodiments, the composition comprises the nanoparticle at a concentration in the range of about 1 mg / ml to about 2 mg / ml, 3 mg / ml, 4 mg / ml, 5 mg / ml, 6 mg / ml, 7 mg / ml, 8 mg / ml, 9 mg / ml, lOmg / ml, 11 mg / ml, 12 mg / ml, 13 mg / ml, 14 mg / ml, 15 mg / ml, 16 mg / ml, 17 mg / ml, 18 mg / ml, 19 mg / ml or 20 mg / ml. In some embodiments, the composition is lyophilized. In some embodiment, the composition further comprises one or more additional therapeutic agents, e.g., second, third or fourth therapeutic agents. The composition may further comprise one more cofactors as required for the enzymatic conversion, as described herein.

[0061] Another aspect of the present invention relates to the medical use of the pharmaceutical composition according to the present invention in the treatment of diseases, such as for the use in treating cancer, comprising administering the pharmaceutical composition in combination with at least one anti-cancer prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug. The use may further comprise one more cofactors as required for the enzymatic conversion, as described herein. The cancer may be lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

[0062] Further provided are methods for treating diseases, such as cancer, comprising administering the pharmaceutical composition according to the present invention in combination with at least one anti-cancer prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug to a patient or subject in need for such a treatment. The use may further comprise one more cofactors as required for the enzymatic conversion, as described herein. The cancer may be lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

[0063] Further provided is the non-medical use of the nanoparticle according to the present invention or the pharmaceutical composition according to the present invention for detoxifying environmental nitroaromatic pollutants, for cleaning of starches, fats, or cellulose, the biosynthesis of antibiotics, as a selectable marker in genetic engineering, as a biosensor for pollutant detection or a biocatalyst for the organic synthesis of pharmaceutical intermediates or fine chemicals. This aspect also includes the use in diagnostics. Respective methods are described in the art (see, for example, https: / / doi.org / 10. l l l l / pce.14099)

[0064] The inventors have successfully demonstrated the encapsulation of an active NfsB enzyme that relies on two co-factors for its activity. This achievement involved cloning a tandem enzyme to facilitate dimer formation, which is essential for co-factor binding, and subsequent activity. Through thorough characterization using TEM, DLS, and cryo-EM, the inventors gained deep insights into the protein assembly. Moreover, the inventor’s investigation revealed that the increasing pore size of the encapsulation system could enhance enzyme turnover, possibly due to less restricted diffusion whilst maintaining the protective encapsulated environment. The encapsulated nitroreductase exhibited activity on various structurally distinct substrates and demonstrated improved performance compared to the free enzyme, particularly under challenging conditions. Furthermore, encapsulated NfsB efficiently activated prodrugs and profluorophores in the extracellular environment. Overall, the inventor’s findings further contribute to a deeper understanding of enzyme encapsulation and hold promise for stabilizing enzymes for various applications including diagnosis and diagnostics, prodrug activation, and beyond. Further, the inventors have successfully demonstrated the encapsulation of yeast cytosine deaminase, which will generate 5-FU in situ, a standard chemotherapy regimen for many HER2-overexpressing cancers.

[0065] The present invention relates to the following items:

[0066] Item 1 : A nanoparticle comprising an assembly of bacterial encapsulin protein monomers, and an internal protein located inside the nanoparticle, wherein the internal protein is a dimer of an enzyme and / or a tandem-enzyme, wherein each enzyme of the dimer or tandem is fused to at least one terminal targeting peptide (TP).

[0067] Item 2. The nanoparticle according to Item 1, wherein the bacterial encapsulin protein is selected from a type I encapsulin, preferably forming a T1 symmetric particle.

[0068] Item 3. The nanoparticle according to Item 1 or 2, wherein at least one marker or label sequence is fused to the N- and / or C-terminus of the encapsulin protein monomer, preferably selected from a His-tag, a Strep-tag, and an MBP affinity tag.

[0069] Item 4. The nanoparticle according to any one of Items 1 to 3, wherein the encapsulin protein is selected from an encapsulin protein of Mycobacterium smegmatis, Thermotoga marilima. Myxococcus xanlhus. Quasibacillus thermotolerans, Haliangium ochraceum. Acinetobacter baumannii. Mycobacterium leprae, Mycobacterium avium, Synechococcus elongatus, Burkholderia cepacia, and Klebsiella pneumoniae.

[0070] Item 5. The nanoparticle according to any one of Items 1 to 4, wherein the encapsulin protein comprises a mutated pore-forming loop increasing the diffusion rate through the pores in the nanoparticle, for example by introducing amino acid substitutions and / or deletions.

[0071] Item 6. The nanoparticle according to Item 5, wherein the encapsulin protein is selected from M. smegmatis comprising a mutated pore-forming amino acid sequence VSETTAHGYPIRE (Seq ID NO: 2) wherein the mutated pore-forming amino acid sequence is selected from the amino acid sequences VSETTAAAIRE (Seq ID NO: 3), VSETTAAIRE (Seq ID NO: 4), VSETAAAIRE (Seq ID NO: 5), and VSETAAIRE (Seq ID NO: 6).

[0072] Item 7. The nanoparticle according to any one of Items 1 to 6, wherein the dimer or tandemenzyme comprises enzymes selected from the group consisting of cytosine-desaminases, nitroreductases, and enzymes used in laundry detergents, such as amylases, lipases, or cellulases, wherein the enzymes in particular selected from wildtype and recombinant bacterial enzymes, such as, for example, the NfsB of E. coli. Item 8. The nanoparticle according to any one of Items 1 to 7, wherein the at least one terminal targeting peptide (TP) sequence is fused to the N- and / or C-terminus of the enzymes of the dimer or tandem-enzyme.

[0073] Item 9. The nanoparticle according to any one of Items 1 to 8, wherein the protein concentration inside the nanoparticle is higher than 2 mM, preferably higher than 3 mM, such as between 3 and 3.5 mM.

[0074] Item 10. An at least bicistronic vector comprising a first nucleic acid sequence encoding an encapsulin protein as defined in any one of Items 1 to 9, and a second nucleic acid sequence encoding at least one enzyme of the dimer or the tandem-enzyme as defined in any one of Items 1 to 9.

[0075] Item 11. A host cell comprising the at least bicistronic vector according to Item 10.

[0076] Item 12. A method for producing the nanoparticle according to any one of Items 1 to 9, comprising recombinantly expressing the encapsulin and the enzyme of the dimer or the tandem-enzyme in the cell according to Item 11, and purifying the resulting nanoparticles or capsids comprising the incorporated dimer or tandem-enzyme.

[0077] Item 13. A method for converting a substrate in vitro or in vivo, comprising providing the nanoparticle according to any one of Items 1 to 9 comprising a dimer or tandem-enzyme, contacting the nanoparticle with at least one suitable substrate to be converted, and converting said substrate.

[0078] Item 14. The method according to Item 13, wherein the non-reduced substrate is a prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug, or the substrate is a nitroaromatic environmental pollutant or the substrate is a selectable marker in genetic engineering or a starch, fat, or cellulose.

[0079] Item 15. The method according to Item 13 or 14, wherein the substrate is a nitroaromatic compound to be reduced in vitro or in vivo, comprising providing the nanoparticle according to any one of Items 1 to 9 comprising a dimer or tandem-enzyme of nitroreductases, contacting the nanoparticle with at least one nitroaromatic substrate suitable for reduction in vitro or in vivo, and reducing said nitroaromatic substrate.

[0080] Item 16. The method according to any one of Items 13 to 15, wherein the method provides a substrate conversion that is similar or higher than a conversion of the respective nonencapsulated enzyme, in particular the dimer or tandem-enzyme thereof.

[0081] Item 17. A pharmaceutical composition, comprising the nanoparticle according to any one of Items 1 to 9, and at least one suitable pharmaceutically acceptable carrier.

[0082] Item 18. The pharmaceutical composition according to Item 17 for use in the treatment of diseases, such as for the use of treating cancer, comprising administering the pharmaceutical composition in combination with at least one anti-cancer prodrug, such as, for example, Nbzp that can be enzymatically reduced into the pharmaceutically active drug.

[0083] Item 19. The pharmaceutical composition for use according to Item 18, wherein the cancer is lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

[0084] Item 20. Use of the nanoparticle according to any one of Items 1 to 9 or the pharmaceutical composition according to Item 17 for detoxifying environmental nitroaromatic pollutants, for cleaning of starches, fats, or cellulose, the biosynthesis of antibiotics, selectable marker in genetic engineering, as a biosensor for pollutant detection or a biocatalyst for the organic synthesis of pharmaceutical intermediates or fine chemicals.

[0085] The invention will now be described further in the following examples with reference to the accompanying figures, nevertheless, without being limited thereto. For the purposes of the present invention, all references as cited are incorporated by reference in their entireties.

[0086] Figure 1 shows the design and characterization of Enc{NfsB} and Enc{tdNfsB}. A) Schematic outline of the cloned constructs. B) Blue Native PAGE analysis of purified Enc{NfsB} and Enc{tdNfsB}. Each sample contained 3 pg of protein per lane in NativePAGE Sample Buffer. C) SDS-Page analysis of Enc{NfsB} and Enc{tdNfsB}, as well as NfsB and tdNfsB. D) DLS analysis ofEnc{NfsB} and Enc{tdNfsB}. Figure 2 shows the cryo-EM analysis of Enc{tdNfsB}. A) High resolution map with a single subunit shown as cartoon rendering. B) A composite map of the high-resolution icosahedral encapsulin map (grey) and the interior map (yellow) resulting from particle subtraction of the cage densities and modelled density within the interior. Inset: 2-D classes following particle subtraction showing clear areas of density that correspond to the packing of tdNsfB within the encapsulin. C) Five-fold pore in the open and closed states as determined by local refinement. The principle shift in pore size is reliant on the movement of Hisl88 from an occluded position in the open to an extended-outwards position in the closed state.

[0087] Figure 3 shows the enzymatic activity of Enc{NfsB} and Enc{tdNfsB}. A) Schematic outline of the coupled luciferase assay using caged-hydroxyl-CBT as aNfsB substrate. B) Comparison of the enzymatic activity of NfsB, tdNfsB, Enc{NfsB} and Enc{tdNfsB} as analyzed by the coupled luciferase assay. The conversions shown have been normalized to a no-enzyme control. C) HPLC analysis of Enc{NfsB}, Enc{tdNfsB}, four pore mutants of Enc {tdNfsB}, as well as NfsB and tdNfsB, in turning over the substrate Nbzp. Gray: NfsB; Red: tdNfsB; Blue: Enc{tdNfsB}; Green: Enc{tdNfsB} A4Ala2; Purple: Enc{tdNfsB} A4Alal; Yellow: Enc{tdNfsB} A6Ala3; Turquoise: Enc{tdNfsB} A6Ala2. Assay conditions: FMN: 400 nM, NADH: 400 pM, Nbzp: 25 pM, enzyme: 20 nM tdNfsB or 3.33 nM Enc{tdNfsB} in 50 mM HEPES, 0.05% Tween 20, pH 7.4. The data points shown have been normalized to a no-enzyme control. D) Reaction velocity Vo of the different constructs shown in C).

[0088] Figure 4 shows the design and characterization of four pore mutants of Enc{tdNfsB}. A) Exterior view of the 5-fold pores of encapsulin shell of M. smegmatis shown in electrostatic surface representation. Surface representations where generated using APBS electrostatics plugin in PyMOL. Red color represents negative charge potential (-5 eV), while blue color represents positive charge potential (5 eV). We show computational models of surface representations. B) DLS analysis of the pore mutants.

[0089] Figure 5 shows a comparison of Enc{tdNfsB} and free tdNfsB in terms of stability and enzymatic activity under challenging conditions. The profluorphore CoNO? was applied as a nitroreductase substrate. A) Enzymatic activity in lysis buffer and cell lysate. Incubation time with CoNO? was 10 min. B) Enzymatic activity in the presence and absence of protease. Treatment time with protease was 10 min, followed by a 10 min incubation with CONO2. C) Enzymatic activity after heat treatment at 55 °C for 1 min, followed by a 10 min incubation with CoNO?.

[0090] Figure 6 shows the Enc{tdNfsB} activation of a nitroaromatic prodrug and a profluorophore in live cells. A) Effect of extracellular prodrug activation on H1299 lung carcinoma cell viability. B) Fluorescence microscopy and bright-field images of HeLa cells treated for 2 h with either 40 pM CONO2, 400 pM NADH, 400 nM FMN, or 40 pM CONO2, 400 pM NADH, 400 nM FMN, and 20 nM Enc{tdNfsB}. Fluorescence microscopy images of HeLa cells treated with Enc{tdNfsB} show green fluorescence from uncaged C0NH2.

[0091] Figure 7 shows the design and characterization of Enc{tdCD} and Enc{eGFP}. A) Schematic outline of the cloned constructs. B) Blue Native PAGE analysis. C) SDS-PAGE analysis after purification via affinity chromatography on a StrepTrap HP column, followed by size exclusion chromatography on a HiLoad 16 / 600 Superdex 200 PG column. D) Representative chromatogram of a size-exclusion column chromatography run (HiLoad 16 / 600 Superdex 200 PG column). E) AF4 fractionation profile using UV (280 nm). F) TEM analysis of Enc{tdCD} and Enc{eGFP}. Bar: 50 nm.

[0092] Figure 8 shows the prodrug activation by Enc{tdCD}. A) HPLC analysis of in vitro Enc{tdCD}-mediated conversion of 5-fluorocytosine (5-FC) to 5 -fluorouracil (5-FU). B) Impact of in cellulo prodrug activation by Enc{tdCD} on J774A.1 cell viability assessed by MTS assay.

[0093] Figure 9 shows the design and characterization ofZHEREnc{tdCD} andZHEREnc{eGFP}. A) Schematic outline of the cloned constructs. B) Blue Native PAGE analysis. C) SDS-PAGE analysis after purification via affinity chromatography on a HisTrap HP column, followed by size exclusion chromatography on a HiLoad 16 / 600 Superdex 200 PG column. D) Representative chromatogram of a size-exclusion column chromatography run (HiLoad 16 / 600 Superdex 200 PG column). E) AF4 fractionation profile using UV (280 nm). F) TEM analysis of / ni REnc[tdCD } andZHEREnc{eGFP}. Bar: 50 nm.

[0094] Figure 10 shows the fluorescence microscopy and bright-field images of SKOV-3 cells treated for 90 min with either 20 nMZHEREnc{eGFP}, 20 nM Enc{eGFP}, or left untreated. Figure 11 shows the flow cytometry analysis ofZHEREnc{eGFP} and Enc{eGFP} binding to HER2 on the surface of SKOV-3 and OV7 cells. Cells were treated withZHEREnc{eGFP} or Enc{eGFP} at 3.3 pmol (red), 10 pmol (orange), or 20 pmol (green), and compared to untreated controls (blue). A) Percentage of GFP-positive SKOV-3 cells. B) Percentage of OV7 cells. C) Mean fluorescence intensity (MFI) of SKOV-3 cells. D) Mean fluorescence intensity (MFI) of OV7 cells. (E-H) Representative flow cytometry histograms showing fluorescence intensity distribution in SKOV-3 (E, G) and OV7 (F, H) cells treated withZHEREnc{eGFP} (E, F) or Enc{eGFP} (G, H). Statistical significance: * p < 0.05, ** p < 0.01, *** p < 0.001.

[0095] Figure 12 shows the impact of in cellulo prodrug activation byZHEREnc{tdCD} on SKOV-3 cell viability assessed by MTS assay.

[0096] Figure 13 shows DLS analysis of encapsulin constructs: Enc (A), Enc{tdCD} and Enc{eGFP} (B), ZHEREnc{tdCD} and ZHEREnc{eGFP} (C).

[0097] Figure 14 shows HPLC chromatograms and APCI-MS spectra of 5-FC (A), 5-FU (B), and Enc{tdCD}-mediated conversion of 5-FC to 5-FU (C). 5-FC expected mass = 128.03 Da (M - H+); 5-FU expected mass = 129.02 Da (M - H+).

[0098] Figure 15 shows purification and assembly analysis of ZHEREnc{tdCD} and Full- ZHEREnc{tdCD}. (A, B) Representative chromatograms of affinity chromatography runs on a HisTrap HP column for ZHEREnc{tdCD} (A) and Full-ZHEREnc{tdCD} (B). Stepwise elution was performed to enhance sample homogeneity by fractionating protein subpopulations according to their His-tag content, which directly correlates with affibody incorporation. For ZHEREnc{tdCD}, whose shell consists of a mixed population of wild-type and affibody-fused encapsulin monomers, a three-step elution was applied (20%, 60%, and 100% elution buffer) to fractionate protein populations with varying His-tag density. In contrast, Full- ZHEREnc{tdCD}, whose shell consists exclusively of affibody-fused encapsulin monomers, was purified using a two-step elution (20% and 100%), as no heterogeneity in shell composition was expected. (C, D) SDS-PAGE analysis of elution fractions from affinity chromatography for ZHEREnc{tdCD} (C) and Full-ZHEREnc{tdCD} (D). At 20% elution, non-target proteins and impurities are removed, while protein populations with a lower proportion of affibody- fused encapsulin monomers in the shell of ZHEREnc{tdCD} elute at 60%. The remaining target protein, enriched in affibody-fused encapsulin monomers, elutes at 100%, ensuring higher sample homogeneity. The differential elution behavior of ZHEREnc{tdCD} reflects the variability in affibody incorporation, as only a fraction of the shell protomers carries a His-tag. In contrast, the shell of Full-ZHEREnc{tdCD} is composed entirely of His-tagged subunits, warranting the use of a two-step elution strategy. Protein bands are indicated as follows: affibody-fused encapsulin monomer (purple arrow, 37.8 kDa), wild-type encapsulin monomer (blue arrow, 28.8 kDa), and tdCD (yellow arrow, 38.4 kDa). (E) BN-PAGE analysis: lane 1 - Enc{tdCD}, lane 2 - ZHEREnc{tdCD}, lane 3 - Full-ZHEREnc{tdCD} 20% elution, lane 4 - Full-ZHEREnc{tdCD} 100% elution. The absence of high-molecular-weight species in Full- ZHEREnc{tdCD} (black arrows indicate expected positions) suggests that steric hindrance introduced by the affibody moiety disrupts proper nanocompartment assembly.

[0099] Figure 16 shows densitometric analysis of shell-modified encapsulin constructs. A) Representative SDS-PAGE gel of ZHEREnc{tdCD} and ZHEREnc{eGFP}. B) Densitometry analysis of the encapsulin shell composition, which consists of 60 identical subunits, a mixture of affibody-modified encapsulin (ZHEREnc) and wild-type encapsulin (Enc). The table presents the calculated number of each subunit type per shell for ZHEREnc{tdCD} and ZHEREnc{eGFP} constructs.

[0100] SEQ ID NO. 1 shows the C-terminal ELS sequence from smegmatis (31), Ser-Leu-Gly-Ile- Gly-Ser-Leu-Lys-Gly-Thr-Arg.

[0101] SEQ ID NOs. 2 to 6 show the wildtype and modified loop sequences, respectively, of the encapsulin of M. smegmatis according to the present invention.

[0102] SEQ ID NO: 7 shows the tandem-enzyme of the enzyme NfsB according to the present invention, bold: linker between monomers; underlined: encapsulin localization sequence. Strep- Tag-tdNfsB (nitroreductase) MWSHPQFEKGGGGSDIISVALKRHSTKAFDASKKLTPEQAEQIKTLLQYSPSSTNSQP WHFIVASTEEGKARVAKSAAGNYVFNERKMLDASHVVVFCAKTAMDDVWLKLVV DQEDADGRFATPEAKAANDKGRKFFADMHRKDLHDDAEWMAKQVYLNVGNFLLG VAALGLDAVPIEGFDAAILDAEFGLKEKGYTSLVVVPVGHHSVEDFNATLPKSRLPQ NITLTEVLEGSAGQGAQAGQGAQAGSSAGDIISVALKRHSTKAFDASKKLTPEQAE QIKTLLQYSPSSTNSQPWHFIVASTEEGKARVAKSAAGNYVFNERKMLDASHVVVFC AI<TAMDDVWLI<LVVDQEDADGRFATPEAI<AANDI<GRI<FFADMHRI<DLHDDAEW MAKQVYLNVGNFLLGVAALGLDAVPIEGFDAAILDAEFGLKEKGYTSLVVVPVGHH SVEDFNATLPKSRLPQNITLTEVLESLGIGSLKGTR

[0103] SEQ ID NO: 8 shows the tandem-enzyme of the enzyme tdCD (cytosin desaminase) according to the present invention, bold: linker between monomers; Red: encapsulin localization sequence. Strep-Tag -tdCD (cytosin desaminase)

[0104] MVTGGMASKWDQKGMDIAYEEAALGYKEGGVPIGGCLINNKDGSVLGRGHNMRF QKGSATLHGEISTLENCGRLEGKVYKDTTLYTTLSPCDMCTGAIIMYGIPRCVVGENV

[0105] NFKSKGEKYLQTRGHEVVVVDDERCKKIMKQFIDERPQDWFEDIGELEGSAGQGAQ AGQGAQAGSSAGTVTGGMASKWDQKGMDIAYEEAALGYKEGGVPIGGCLINNKD GSVLGRGHNMRFQKGSATLHGEISTLENCGRLEGKVYKDTTLYTTLSPCDMCTGAII MYGIPRCVVGENVNFKSKGEKYLQTRGHEVVVVDDERCKKIMKQFIDERPQDWFED IGELAGGGSASLGIGSLKGTR

[0106] Examples

[0107] MATERIALS AND METHODS

[0108] Protein production and purification

[0109] Corresponding plasmids were transformed into chemically competent A. coli BL21 Star (DE3) cells, and transformants were selected on LB-agar plates containing ampicillin or streptomycin. Single colonies were transferred into LB-medium (supplemented with ampicillin) and grown overnight at 37 °C, 180 rpm. 5 mL pre-culture was transferred into 1 L auto-induction medium (supplemented with ampicillin) and grown at 19 °C for 65 h.

[0110] Cells were harvested by centrifugation (5000 x g, 4 °C, 30 min) and resuspended at a concentration of 5 - 6 mL / g wet weight in lysis buffer (for Strep-tagged proteins: 100 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, pH 8, for His-tagged proteins: 20 mM sodium phosphate, 500 mM NaCl, and 20 mM imidazole, pH 7.4, both supplemented freshly before use with 1 mM DTT, 1 mM PMSF and 1 mg / mL lysozyme; incubation on ice for 30 - 45 min). Cells were lysed by sonication on ice using Bandelin Sonopuls HD 2070 ultrasonic homogenizer (Bandelin Electronics, Berlin, Germany, microtip of 3 mm in diameter) for 3 min on / off 2 times at 40% amplitude. Cell debris was removed via centrifugation at 15000 x g for 30 min. Proteins were then purified using 5 mL StrepTrap HP column (Cytiva, Germany) and eluted with elution buffer (100 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 2.5 mM desthiobiotin, pH 8). A typical encapsulin purification from 1 L culture yielded approximately 25 - 30 mg of protein after pooling all fractions from the 5 mL StrepTrap HP column (4.5 - 5 mg / mL). A typical purification of unencapsulated cargo proteins NfsB or tdNfsB from 1 L culture yielded approximately 20 - 25 mg of protein after pooling all fractions from the 5 mL StrepTrap HP column (2.5 - 3 mg / mL).

[0111] Proteins bearing a His-tag were purified using a 5 mL His-Trap HP column. Following initial washing, proteins were eluted using a stepwise gradient of elution buffer (20 mM sodium phosphate, 500 mM NaCl, 500 mM imidazole, pH 7.4) at concentrations of 20%, 60%, and 100%. Only the fractions eluted with 100% elution buffer were retained for subsequent applications. A typical purification of His-tagged encapsulin from 1 L of culture yielded approximately 10-15 mg of protein, with all fractions eluted at 100% elution buffer from the 5 mL His-Trap HP column pooled (3-3.5 mg / mL).

[0112] Further purification was achieved by size exclusion chromatography on a HiLoad 16 / 600 Superdex 200 PG column (Akta Protein Purification System, GE Healthcare Germany). Proteins were eluted from the column using an elution buffer. For in vitro experiments, proteins were eluted with 50 mM HEPES, 0.05% Tween 20, pH 7.4. For cellular assays, proteins were eluted with 50 mM HEPES, pH 7.4. Following elution, the proteins were reconstituted with a two-fold excess of FMN at 4 °C overnight. The excess unbound FMN was then separated using a PD-10 desalting column with the identical buffer. Fractions were analyzed by SDS-PAGE, pooled, subdivided, and stored at - 80 °C until further use.

[0113] Negative stain-transmission electron microscopy (TEM)

[0114] Nanocompartment samples were diluted to an initial concentration of 0.4 mg / mL in HEPES buffer (50 mM, pH 7.4) and adsorbed for 20 s on carbon-coated grids (Electron Microscopy Sciences, CF300-CU). The grid was stained with 1% uranyl acetate for 20 s. The excess stain was removed using filter paper. TEM images were recorded with a Thermo Scientific Talos L200C transmission electron microscope, operated at 120 kV. The images were analyzed using ImageJ software.

[0115] Dynamic light scattering (DLS)

[0116] DLS data were collected on a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK). Measurements were performed at 25 °C using ultra-low volume quartz cuvette (ZEN2112) containing 1 mg / mL of nanocompartment (Enc{NfsB}, Enc{tdNfsB}, Enc{tdCD} or pore mutants) in 50 mM HEPES, 0.05% Tween 20, pH 7.4. Three measurements were performed with the following set up: attenuator: 8; mean count rate (kcps): 260 - 420. The data were analyzed and presented with Zetasizer Nanoseries software (Malvern Instruments, Malvern, UK) using the general purpose analysis model. The intensity-weighted mean hydrodynamic diameter for each measurement was calculated and recorded.

[0117] Asymmetric flow field flow fractionation (AF4)

[0118] The AF4 system consisted of a flow controller (Eclipse AF4, Wyatt, Dernbach, Germany), a MALS detector (DAWN Heleos II, Wyatt), a differential refractometer (Optilab T-rEX, Wyatt) and the separation channel (SC channel, regenerated cellulose membrane, cut-off 10 kDa, 350 pm spacer, wide type, Wyatt). The elution buffer used was 50 mM HEPES adjusted to pH 7.4 and filtered through 0.1 pm. AF4 runs were conducted at 22 °C with a channel flow rate of 1.0 mL / min and the following flow sequence (Vx = cross flow in mL / min): (a) elution (2 min, Vx: 1.0); (b) focus (1 min, Vx: 1.5), focus + inject (1 min, Vx: 1.5, inject flow: 0.2 mL / min); focus (2 min, Vx: 1.5); (c) elution (30 min, Vx: 1.0); elution (30 min, linear Vx gradient: 1.0 to 0.05); (d) elution (5 min, Vx: 0.0); (e) elution + inject (2 min, Vx: 0.0). A total protein mass of 23 ± 2.0 pg was injected. The eluted sample concentration was calculated from the refractive index signal using a specific refractive index increment of 0.185 mL / g. Mass-weighted mean values of the molar masses and the radii of gyration were calculated from MALS data using the ASTRA 8 software package (Wyatt).

[0119] SDS-PAGE and gel densitometry

[0120] Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Fisher). Following dilutions were prepared in HEPES buffer (50 mM, pH 7.4): Enc{tdNfsB} at 500 pg / mL; Enc at 90, 225, 360, 495, and 630 pg / mL; and tdNfsB at 26.15, 78.45, 130.75, 183.05, and 235.35 pg / mL. Each protein dilution (20 pL) was mixed with an equal volume (20 pL) of 2* SDS sample buffer (125 mM Tris-HCl (pH 6.8), 4% SDS, 20% (v / v) glycerol, 0.02% bromophenol blue, 200 mM DTT) and then denatured at 95 °C for 10 minutes.

[0121] For SDS-PAGE, a 15% gel was prepared in-house and mounted in the SDS-PAGE apparatus. SDS-PAGE running buffer (25 mM Tris, 186 mM Glycine, 0.1% SDS) was added to the bottom and top reservoirs, with the wells washed to remove any trapped air bubbles before sample loading. A volume of 10 pL from each sample was carefully loaded into the wells. The BlueStar Plus Prestained Protein Marker (MWP04, Nippon), which covers a molecular weight range of 10-240 kDa, was included as a standard. The electrodes were connected to a power supply, and electrophoresis was conducted at 200 V and 60 mA for approximately 60 minutes until the bromophenol blue dye front had migrated to the lower edge of the gel.

[0122] After electrophoresis, the gels were stained overnight in a solution containing 0.1% Coomassie R-250 in 10% (w / v) ammonium sulfate, 1% (v / v) phosphoric acid, and 10% (v / v) methanol, followed by destaining with water. The destained gels were imaged using the UVP ChemStudio Touch 815 (Analytik Jena) and analyzed with ImageJ software. Background values were subtracted, and calibration curves were plotted for Enc in the range of 45-315 pg / lane and for tdNfsB in the range of 13.08-117.68 pg / lane, correlating the band areas to the respective protein amounts per lane.

[0123] The amounts of Enc and tdNfsB in the Enc{tdNfsB} sample were determined using the corresponding calibration curves. The molar concentrations of Enc (1804.2 kDa) and tdNfsB (52.3 kDa) were calculated, and their ratio was assessed to quantify the amount of encapsulated tdNfsB within the Enc nanocompartment. The experiment was repeated twice to ensure reproducibility.

[0124] For Enc,ZHEREnc,ZHEREnc{tdCD} andZHEREnc{eGFP}, background signal was subtracted, and regions of interest (ROIs) were defined around the bands corresponding to encapsulin monomers (Enc, 28.8 kDa) and affibody-fused encapsulin monomers (ZHEREnc, 37.8 kDa) in lanes containingZHEREnc{tdCD} andZHEREnc{eGFP} samples. The signal intensity for each band was measured and plotted. To accurately quantify the ratio between Enc andZHEREnc, a correction factor was applied to account for the difference in molecular weights, as larger proteins bind more Coomassie dye and therefore produce a stronger signal. Specifically, the signal intensity for the Enc band was multiplied by a factor of 1.3125 (calculated as 37.8 kDa / 28.8 kDa). This adjustment allowed for accurate comparison of the relative abundance of the two subunit types. Based on the corrected signal ratio, the number of Enc andZHEREnc subunits per shell was calculated, considering that each encapsulin nanocompartment is composed of exactly 60 monomers. The experiment was performed in duplicate to ensure reproducibility. Blue Native PAGE

[0125] Blue native PAGE analysis was performed using NativePAGE 3-12% Bis-Tris 1.0 mm Mini Protein Gels (Thermo Fisher) in a XCell Sure-LockTM Mini-Cell gel system (Invitrogen, Darmstadt, Germany). Each sample contained of 2.5 or 3 pg of protein and was mixed with 8 pL NativePAGE Sample Buffer (0.5 M 6-aminohexanoic acid, 0.05 g / L Coomassie Brilliant Blue G250, 37.5% glycerol). Electrophoresis was performed using the following buffer systems: anode buffer (25 mM imidazole / HCl, pH 7.0), dark blue cathode buffer (7.5 mM imidazole / HCl, 50 mM Tricine, 0.2% Coomassie Brilliant Blue G250, pH 7.0) and light blue cathode buffer (7.5 mM Imidazole / HCl, 50 mM Tricine, pH 7.0). Electrophoresis started with the dark blue cathode buffer at 120 V for 25 min, followed by a transition to the light blue cathode buffer for an additional 12 hours at 50 V. The voltage was then increased to 120 V for 1 hour, and finally, the gel was subjected to 200 V for the last 30 minutes. After electrophoresis, the gel was stained with Coomassie Brilliant Blue G250 (80 mg dissolved in 1 L of water with 30 mM HC1) and destained with water.

[0126] Grid preparation and data collection

[0127] Protein was exchanged into 20 mM HEPES pH 7.5, 100 mM NaCl and concentrated to 20 mg / mL before application on to Quantifoil 1.2 / 1.3 300 mesh grids. 3 pL were applied to a glow discharged grid in an FEI vitrobot at 8 °C and 95% humidity. Grids were blotted at force 4 for 4 seconds prior to plunge freezing in liquid ethane. Samples were screened on a JEOL F2 microscope prior to extended data collection on a JEOL CRYOARM300 at SCMI (Glasgow, UK). Movies were recorded at a dose of 60.2 e' / A2using a DE-APOLLO with pixel size 0.7832 A in super-resolution mode.

[0128] Map reconstruction and model building

[0129] Movies were processed in CryoSPARCv.4.3. Briefly, raw movies were motion corrected and 2 x 2 binned using Patch Motion correction prior to CTF estimation in PatchCTF. Particles were then picked using a low-resolution template produced from screening data on the F2 and resultant particles were extracted 4 x 4 binned in an extraction box size of 512 pixels. The initial particle stack was 2-D classified and run through ab initio model building with icosahedral symmetry to produce a low-resolution starting model. Particles were subsequently sorted through hetero-refinement before re-extraction 2 x 2 binned and further hetero-refinements with “bad” volumes used as sinks to remove poor particles. A final particle stack was reextracted uncropped following reference-based motion correction and run through homogenous refinement with per-particle defocus and CTF-refinement with EWS correction with icosahedral symmetry. The resulting map was sharpened and monomeric PDB 7boj was docked before refinement and icosahedral expansion in Phenix.

[0130] A second map and model were produced by focused local refinement on the pentameric encapsulin around the 5-fold symmetry. A mask was produced in ChimeraX to cover the pentameric unit and local refinement was carried out without symmetry restraints. The resulting model produced a closed pore structure. The model was then refined and underwent symmetry expansion in Phenix.

[0131] Finally, efforts were made to determine the structure of tdNfsB dimers within the encapsulin shell. Particle subtraction was carried out using the final masked volume and 2D classification was used to align the interior components. Density was observed, and a low-resolution model produced with defined densities for each of six tdNfsB units inside the capsid. The resulting map was sharpened and a model was built de novo before refinement and icosahedral expansion in Phenix.

[0132] A second map and model were produced by focused local refinement on the pentameric encapsulin around the point of 5-fold symmetry. A mask was produced in ChimeraX to cover the pentameric unit and local refinement was carried out without symmetry restraints. An encapsulin model was then built as a monomer prior to being expanded along map determined icosahedral parameters in Phenix. The resulting model demonstrated a closed pore structure.

[0133] Finally, efforts were made to determine the structure of tdNfsB dimers within the encapsulin shell. Particle subtraction was carried out using the final masked volume and 2D classification was used to align the interior components. Density was observed but no high resolution reconstruction could be generated.

[0134] Model building of pore mutants

[0135] To predict the structures of the generated encapsulin pore mutants, their amino acid sequences were submitted to the I-TASSER server. I-TASSER employs threading and ab initio modeling to generate structure predictions. For alignment, PDB ID 7BOJ, which represents the wild-type encapsulin, was used as the reference template. The model with the highest C-score, reflecting the highest confidence, was chosen for further analysis. This high-confidence model was then aligned to the pentameric subunit of the 7BOJ template using PyMOL. Surface representations of these pentameric subunits were generated using the APBS plugin in PyMOL to visualize electrostatic potentials. To characterize the structural changes in the five-fold pore, its diameter was measured using MoleOnline (https: / / mole.upol.cz / ).

[0136] Luminescence-based assay for NfsB enzymatic activity

[0137] The assay was performed in a white OptiPlate384 microtiter plate (PerkinElmer, Germany) with a total volume of 20 pL per well using an assay buffer containing 50 mM HEPES, 0.05% Tween 20, pH 7.4. The final enzyme concentrations in the reaction mixtures were set to 20 nM NfsB, 10 nM tdNfsB, 3.3 nM Enc{NfsB}, or 3.3 nM Enc{tdNfsB}. For NADH concentration screening, the final reaction mixture contained 25 pM caged-hydroxy-CBT (final DMSO concentration 2.5%), 400 nM FMN, and varying concentrations of NADH at 0, 50 pM, 100 pM, 200 pM, 300 pM, or 400 pM. In the FMN concentration screening, the reaction mixture included 25 pM caged-hydroxy-CBT (final DMSO concentration 2.5%), 400 pM NADH, and varying concentrations of FMN at 0, 50 nM, 100 nM, 200 nM, 300 nM, or 400 nM. For each experimental setup, control wells containing FMN, NADH, and hydroxy-CBT at the same concentrations as the experimental wells were included. All reactions were initiated by adding enzyme dilutions or assay buffer for the negative control wells. The reactions were incubated at 37 °C for 1 hour, after which 20 pL of Luciferin detection reagent (Promega, Germany), supplemented with D-cysteine to a final concentration of 5 mM, was added. After a 20-minute incubation at room temperature, the luminescent signal was measured using a microplate reader (Tecan Spark® Multimode Microplate Reader). All conditions were assessed in triplicate, and the experiment was repeated twice. During the subsequent data evaluation, the readouts from the experimental wells were corrected by subtracting the readouts from the negative control wells before plotting the results.

[0138] For the quantification of NfsB enzymatic activity, a luminescence-based assay was conducted using a standard curve of hydroxy-CBT (final DMSO concentration of 2.5%) prepared in the same assay buffer at the following concentrations: 0 pM, 1 pM, 5 pM, 10 pM, 15 pM, 20 pM, and 30 pM. A volume of 20 pL of each standard was pipetted into the wells and incubated at 37 °C for 1 hour. Subsequently, 20 pL of Luciferin detection reagent (Promega, Germany), supplemented with D-cysteine to a final concentration of 5 mM, was added. After a 20-minute incubation at room temperature, the luminescent signal was measured using a microplate reader (Tecan Spark® Multimode Microplate Reader). Each concentration was assessed in triplicate, and the entire experiment was repeated twice. During data evaluation, the readouts from the experimental wells were corrected by subtracting the readouts from the negative control wells (0 pM hydroxy-CBT) before plotting the results.

[0139] In vitro prodrug activation

[0140] In vitro prodrug activation was performed in an assay mixture containing final concentrations: 400 pM NADH, 400 nM FMN, 25 pM of each of the prodrugs (Nbzp (5), CB1954 (39), MA60, MA63, AV2, AV4 (4, 40)), and either 20 nM tdNfsB, 3.3 nM Enc{tdNfsB} or corresponding pore size mutants. The assay was carried out in HEPES buffer (50 mM, pH 7.4) containing 2.5% DMSO as co-solvent at 37 °C at 350 rpm. At designated time points, aliquots from control and reaction samples were collected and the reactions stopped by mixing 1 :1 with cold acetonitrile (ACN).

[0141] Chromatographic separation was achieved on UltiMate 3000 (Thermo Scientific, Germany) using a Polaris 5 C18-A 250 x 4.6 mm column (Agilent, USA). For the analysis of MA60, MA63, CoNO?, and CB1954, water was used as mobile phase A, and ACN as mobile phase B. For the analysis of Nbzp, AV2, and AV4 compounds, mobile phase A consisted of water supplemented with 0.1% TFA, and mobile phase B consisted of ACN supplemented with 0.1% TFA. Analysis was performed with a flow rate of 1 mL / min using the following program: 0 - 0.5 min at 5% mobile phase B, 0.5 - 3 min gradient of 5-35% mobile phase B, 3 - 5 min gradient of 35 - 95% mobile phase B, 5 - 8 min at 95% mobile phase B, 8 - 8.5 min gradient of 95 - 5% mobile phase B, 8.5 - 11 min at 5% mobile phase B. The mass spectrometry analysis was performed with an Agilent 6545 LC / Q-TOF system operating in the positive ion mode. During the subsequent quantitative evaluation of prodrug conversion, the area under the curve (AUC) for the prodrug peaks in enzyme-containing samples was divided by the AUC for the corresponding negative controls (without enzyme) and then multiplied by 100% to obtain the percentage conversion.

[0142] In vitro prodrug activation of 5-FC was conducted in HEPES buffer (50 mM, pH 7.4) at 37 °C and 350 rpm. The reaction was initiated by adding a dilution of Enc{tdCD} enzyme to achieve a final concentration of 188 pg / mL (corresponds to 75 nM) or by adding assay buffer for the negative control, both mixed with 5-FC dilution at a final concentration of 3 mM. After 1 hour of incubation, aliquots from both control and reaction samples were collected, and the reactions were terminated by mixing 1 : 1 with cold acetonitrile (ACN).

[0143] Chromatographic separation was performed using an Agilent 1290 Infinity II UHPLC system (Agilent, Germany) equipped with an ACQUITY UPLC BEH Amide column (130 A, 1.7 pm, 150 x 2.1 mm, Waters, USA). The mobile phases included A: water, B: 0.5% ammonium hydroxide in water, and C: ACN. The analysis was conducted at a flow rate of 0.5 mL / min and a column temperature of 40 °C, following a gradient program that began with 0% A, 5% B, and 95% C, transitioning to 45% A, 5% B, and 50% C, over the first 6 minutes. The system was maintained at 45% A, 5% B, and 50% C from 6 to 7 minutes. Mass spectrometric analysis was performed using an Agilent 6545 LC / Q-TOF system operating in negative ion mode.

[0144] Kinetics analysis of nitroreductase activity

[0145] The nitroreductase activity of free and encapsulated tdNfsB was assessed by measuring the consumption of NADH at 37 °C. The reactions were performed in HEPES buffer (50 mM, pH 7.4) containing 2.5% DMSO as a co-solvent. The reaction mixtures contained 1.5 mM NADH and variable concentrations of the substrate Nbzp (100, 250, 500, 750, 1000, or 1500 pM). Reactions were initiated by adding the enzyme to achieve final concentrations of 180 nM tdNfsB, 30 nM Enc{tdNfsB}, or corresponding pore size mutants. The assay was conducted in transparent 96-well plates (Sarstedt), with a final reaction volume of 50 pL. Absorbance at 340 nm (A340) was measured in duplicate over a total duration of 30 minutes, with readings taken at 15-second intervals using a Tecan Spark® Multimode Microplate Reader. Background control reactions, lacking protein, were performed and the absorbance values for these controls were subtracted from those of the enzymatic samples to obtain corrected values. The NADH concentration was calculated using the Beer-Lambert law, employing the molar absorption coefficient (extinction coefficient) of NADH at 340 nm (6220 M1cm ') and a path length of 0.134 cm. Initial reaction rates for the different concentrations of Nbzp were determined by calculating the slope of the linear portion of the initial phase of the time course plots. Subsequently, nonlinear regression analysis was performed on the substrate concentration versus initial rate data using the Michaelis-Menten model. Data were plotted in a Lineweaver- Burk plot.

[0146] Assessment of protease susceptibility of free and encapsulated tdNfsB

[0147] To evaluate the susceptibility of free and encapsulated tdNfsB to proteolytic degradation, both forms were incubated with Protease from Streptomyces griseus (Sigma- Aldrich, P5147). The incubation was conducted using 1 U of protease per 200 pmol of free tdNfsB or per 33.3 pmol of Enc{tdNfsB}, at 37°C for 10 minutes in 50 mM HEPES, pH 7.4. Following incubation, aliquots were taken and added to the assay mixture to initiate the nitroreductase reaction. The final assay mixture contained 400 pM NADH, 400 nM FMN, and 25 pM CoNO?. The final enzyme concentrations in assay were adjusted to either 20 nM for free tdNTR or 3.3 nM for Enc{tdNfsB}, with a final volume of 100 pL in a black 96-well plate. Fluorescence measurements were performed after an incubation time of 10 min at 37 °C, using a Tecan Spark 10M plate reader at 37 °C. The excitation wavelength was set to 380 nm, and emission was measured at 510 nm.

[0148] Assessment of thermal stability of free and encapsulated tdNfsB

[0149] To evaluate the thermal stability of free and encapsulated tdNfsB, both forms were incubated for 1 min at 55 °C. Following incubation, aliquots were taken and added to the assay mixture to initiate the reaction. The final assay mixture contained 400 pM NADH, 400 nM FMN, and 25 pM CoNO?. The final enzyme concentrations in assay were adjusted to either 20 nM for free tdNfsB or 3.3 nM for Enc {tdNfsB}, with a final volume of 100 pL in a black 96-well plate. Fluorescence measurements were performed after an incubation time of 10 min at 37 °C, using a Tecan Spark 10M plate reader at 37 °C. The excitation wavelength was set to 380 nm, and emission was measured at 510 nm.

[0150] Assessment of activity in HeLa cell lysate of free and encapsulated tdNfsB

[0151] A confluent cell culture dish was placed on ice and washed twice with 10 mL of ice-cold PBS. After draining the PBS, 1 mL of ice-cold Pierce IP lysis buffer was added per 150 mm dish. The HeLa cells were scraped off with a cold plastic scraper and transferred into pre-cooled microfuge tubes. The cell suspension was agitated at 4 °C for 30 minutes and then centrifuged at 13,000 g for 10 minutes at 4 °C. The supernatant was carefully aspirated into fresh pre-cooled tubes and kept on ice, while the pellet was discarded. Aliquots were taken and added to the assay mixture to initiate the reaction.

[0152] The enzyme activity assay was conducted either in Pierce IP lysis buffer or in HeLa cell lysate. The final assay mixture contained 400 pM NADH, 400 nM FMN, and 25 pM CONO2. The enzyme concentrations were adjusted to 20 nM for free tdNTR or 3.3 nM for Enc{tdNfsB}, in a final volume of 100 pL within a black 96-well plate. Fluorescence measurements were performed at 37°C using a Tecan Spark 10M plate reader, with excitation at 380 nm and emission measured at 510 nm.

[0153] Cell lines and culture conditions

[0154] HeLa and H1299 lung carcinoma cell lines were maintained under standard conditions at 37 °C and 5% CO2. HeLa cells were cultured in DMEM (high glucose) supplemented with 10% fetal calf serum, while H1299 cells were cultured in RPMI-1640 medium containing 10% fetal calf serum.

[0155] Mouse monocyte cell line J774A.1 and human ovarian cancer cell lines SKOV-3 and OV7 were maintained under standard culture conditions at 37 °C and 5% CO2. SKOV-3 and J774A.1 cells were cultured in high-glucose DMEM supplemented with 10% fetal calf serum and 10% heat- inactivated fetal calf serum, respectively. The OV7 cell line was cultured in a DMEM:HAMS F-12 medium supplemented with 5% fetal calf serum, 0.5 pg / mL hydrocortisone, and 10 pg / mL insulin.

[0156] Assessment of activity in HeLa cell lysate of free and encapsulated tdNfsB

[0157] A confluent cell culture dish was placed on ice and washed twice with 10 mL of ice-cold PBS. After draining the PBS, 1 mL of ice-cold Pierce IP lysis buffer was added per 150 mm dish. The HeLa cells were scraped off with a cold plastic scraper and transferred into pre-cooled microfuge tubes. The cell suspension was agitated at 4 °C for 30 minutes and then centrifuged at 13,000 g for 10 minutes at 4 °C. The supernatant was carefully aspirated into fresh pre-cooled tubes and kept on ice, while the pellet was discarded. The protein concentration, as measured with a Nanodrop instrument, was 4-4.4 mg / mL.

[0158] The enzyme activity assay was conducted either in Pierce IP lysis buffer or in HeLa cell lysate. The final assay mixture contained 400 pM NADH, 400 nM FMN, and 25 pM CONO2. The enzyme concentrations were adjusted to 20 nM for free tdNTR or 3.3 nM for Enc{tdNfsB}, in a final volume of 100 pL within a black 96-well plate. Fluorescence measurements were performed after an incubation time of 10 min at 37°C, using a Tecan Spark 10M plate reader, with excitation at 380 nm and emission measured at 510 nm.

[0159] Fluorescence microscopy

[0160] For live cell imaging, 300 pL of a 5xl04HeLa cell suspension were seeded per well in p-Slide 8-well chambers with polymer coverslip bottoms (Ibidi) and allowed to attach overnight. After attachment, the medium was replaced with 300 pL of fresh medium containing either 40 pM CoNO?, 400 pM NADH, and 400 nM FMN (negative control) or 40 pM CoNO?, 400 pM NADH, 400 nM FMN, and 20 nM Enc{tdNfsB}. Cells were incubated for 2 h. Following incubation, the medium was removed, and the cells were washed twice with serum-free DMEM. The plasma membrane was stained with CellMask Orange (2.5 pg / mL) prior to imaging (red color).

[0161] Fluorescence imaging was performed using an inverted Zeiss LSM 880 laser scanning microscope equipped with an LD LCI Plan- Apochromat 63x / 1.40 Oil DIC M27 objective. Green fluorescence (C0NH2) was collected between 446-535 nm upon excitation at 405 nm. Red fluorescence (CellMask Orange) was excited at 561 nm and emission was collected between 567 - 735 nm. All images were acquired under consistent settings to ensure comparability across different experimental conditions.

[0162] For live cell imaging, a 300 pL suspension of 1 x 105SKOV-3 cells was seeded into each well of p-Slide 8-well chambers with polymer coverslip bottoms (Ibidi, Germany) and allowed to attach overnight. After the attachment period, the medium was replaced with 300 pL of fresh medium containing either 20 nM Enc{eGFP}, 20 nM ZHEREnc{eGFP}, or was left untreated. The cells were then incubated for 90 minutes. Following the incubation, the medium was removed, and the cells were washed twice with serum-free DMEM. The plasma membrane was subsequently stained with CellMask Orange (2.5 pg / mL) prior to imaging (red fluorescence). Fluorescence imaging was conducted using an inverted Zeiss LSM 880 laser scanning microscope equipped with an LD LCI Plan- Apochromat 63x / 1.40 Oil DIC M27 objective. Green fluorescence (eGFP) was detected between 499-552 nm upon excitation at 488 nm, while red fluorescence (CellMask Orange) was excited at 561 nm and emission was collected between 566-735 nm. All images were captured under consistent settings to ensure comparability across different experimental conditions.

[0163] Flow cytometry

[0164] SKOV-3 and OV7 cells were cultivated in T75 flasks and subsequently harvested for flow cytometry. Initially, each flask was rinsed twice with 6 mL of PBS, followed by the addition of 3 mL of Trypsin-EDTA solution and incubation at 37 °C for 5 minutes to facilitate cell detachment. Next, 7 mL of cell culture medium was added, and the resulting cell suspension was transferred to 15 mL Falcon tubes. The cells were pelleted by centrifugation at 200xg at room temperature, washed twice with warm PBS, and counted. The final cell pellets were resuspended in ice-cold FACS buffer (2% v / v FBS and 0.08% w / v sodium azide in PBS).

[0165] Aliquots of 1 x 106cells in 70 pL of FACS buffer were incubated on ice for 60 minutes with 30 pL of protein solution (final volume: 100 pL). Cells were treated with either Enc{eGFP} orZHEREnc{eGFP} at three different amounts (3.3 pmol, 10 pmol, and 20 pmol), or left untreated as a control. Protein dilutions were prepared in 50 mM HEPES (pH 7.4) to achieve the desired concentrations. Following incubation, the cells were washed twice with 1 mL of FACS buffer, resuspended in 400 pL of FACS buffer, and stained with 4',6-diamidin-2-phenylindole (DAPI, final concentration: 0.15 pg / mL) for 5 minutes to exclude non-viable cells. All samples were analyzed within one hour on a BD LSRFortessa flow cytometer (BD, Germany) with 50 000 events being recorded per sample. Data processing was performed using FlowJo vlO software. Single cells were gated by plotting FSC-W versus FSC-A to eliminate cell aggregates, and DAPI-positive events were excluded to isolate the live-cell population. This final population was then used for subsequent analysis of protein binding.

[0166] Extracellular prodrug activation

[0167] H1299 cells were seeded at a density of 2500 cells per well in 96-well plates. After allowing the cells to attach overnight, the medium was removed, and 100 pL of cell culture medium containing 400 pM NADH, 400 nM FMN, and one of the prodrugs (100 pM Nbzp, 100 pM MA60) was added to the wells. Additionally, either 20 nM Enc{eGFP}, Enc{tdNfsB} or 3.33 nM tdNsfB was included in the medium. DMSO (0.5%), 400 pM NADH, 400 nM FMN was used as a negative control. After 72 hours of incubation, the MTS assay was performed as follows: CellTiter 96® AQueous MTS Reagent Powder (Promega, Germany) was prepared as a 1 mg / mL solution in Dulbecco's phosphate-buffered saline (DPBS), which was then filter- sterilized using a 0.2 pm filter, aliquoted, and stored at -20 °C. Phenazine methosulfate (PMS; Sigma) was prepared at a concentration of 0.92 mg / mL in DPBS, also filter-sterilized through a 0.2 pm filter, aliquoted, and stored at -20 °C. Immediately before performing the MTS assay, 1.524 mL of the MTS solution was combined with 76 pL of the PMS solution. An additional 8 mL of medium was added to this mixture. In the assay plate, the old medium was replaced with 120 pL of the MTS / PMS solution mixture. Furthermore, 120 pL of the same MTS / PMS mixture in medium was added to empty wells as a negative control. The cells were incubated for 2.5 to 3 hours under standard culture conditions, after which the plate was briefly shaken and measured using a microplate reader (Tecan Spark® Multimode Microplate Reader) at an optical density of 490 nm. All conditions were assessed in triplicate, and the experiment was repeated twice. The average absorbance at 490 nm from the empty “no cell” control wells was substracted from all other absorbance values to obtain corrected absorbance values. The absorbance of untreated cells was set at 100% viability. The viability of experimental wells was calculated by dividing their absorbance by the absorbance of untreated cells and multiplying by 100%.

[0168] In cellulo prodrug activation

[0169] To investigate prodrug activation in the J774A.1 cell line, 2000 cells / well were plated in 96- well plates. After allowing the cells to attach overnight, the medium was replaced with either 20 nM Enc{eGFP} or 20 nM Enc{tdCD} protein dilutions in complete culture medium, or with fresh culture medium for the untreated control group. After 18 hours of incubation, the cells were washed twice with 200 pL PBS, and freshly prepared 0.5 mM 5-FC in complete culture medium (filter-sterilized) was added. At identical intervention time points, the control wells received complete culture medium instead of protein dilutions, were washed twice with PBS, and had either 0.5 mM 5-FC (negative control) or 0.5 mM 5-FU (positive control) added. After 72 hours of incubation, cell viability was assessed using the CellTiter 96® AQueous Cell Proliferation Assay (MTS assay, Promega, Germany) according to the manufacturer’s instructions. Briefly, CellTiter 96® AQueous MTS Reagent Powder was dissolved in DPBS at 1 mg / mL, filter-sterilized, aliquoted, and stored at -20 °C. Phenazine methosulfate (PMS; Sigma) was prepared at a concentration of 0.92 mg / mL in DPBS, also filter-sterilized through a 0.2 pm filter, aliquoted, and stored at -20 °C. Immediately before performing the MTS assay, 1.524 mL of the MTS solution was combined with 76 pL of the PMS solution. An additional 8 mL of medium was added to this mixture. In the assay plate, the old medium was replaced with 120 pL of the MTS / PMS solution mixture. Furthermore, 120 pL of the same MTS / PMS mixture in medium was added to empty wells as a negative control. The cells were incubated for 2.5 to 3 hours under standard culture conditions, after which the plate was briefly shaken and measured using a microplate reader (Tecan Spark® Multimode Microplate Reader) at an optical density of 490 nm. All conditions were assessed in triplicate, and the experiment was repeated twice. The average background absorbance (from “no cell” control wells) was subtracted from all values to obtain corrected absorbance values. The absorbance of untreated cells was set at 100% viability. The viability of experimental wells was calculated by dividing their absorbance by the absorbance of untreated cells and multiplying by 100%. SKOV-3 cells were seeded in a 96-well plate at a density of 6000 cells / well and allowed to attach overnight. The following day, the medium was replaced with either 50 nMZHEREnc{eGFP} or 50 nMZHEREnc{tdCD} protein dilutions in complete culture medium and incubated for 1 hour. After incubation, the cells were washed twice with 200 pL PBS, and freshly prepared 4 mM 5 -FC in complete culture medium (filter-sterilized) was added. At the same intervention time points, control wells received complete culture medium instead of protein dilutions, were washed twice with PBS, and then had either 4 mM 5-FC (negative control) or 2 mM 5-FU (positive control) added. Dilutions exceeding 2 mM of 5-FU could not be prepared in the medium due to solubility limitations. After 120 hours of incubation, cell viability was determined using the MTS assay as described above.

[0170] Results

[0171] Construction of the Encapsulated Nitroreductase Enzyme as a Single or Tandem Entity

[0172] The inventors inserted the nitroreductase NfsB gene (41) from E. coli into the pETDuet™-l vector and appended the previously identified C-terminal ELS sequence from M. smegmatis (31) (Ser-Leu-Gly-Ile-Gly-Ser-Leu-Lys-Gly-Thr-Arg, (SEQ ID NO. 1)). Recognizing that nitroreductase operates as an active dimer, the inventors engineered a tandem enzyme by linking two NfsB copies with a flexible linker sequence (as illustrated in Figure 1 A) to ensure obligate dimer formation and thus incorporation into the encapsulin structure (6). This decision stemmed from the hypothesis that NfsB monomers would likely be spatially constrained within the capsid, preventing the formation of active dimers once the confined capsid had assembled. The encapsulin gene msmei 5672 fromM. smegmetis me2155, tagged with a C-terminal Strep- tag, was cloned under the second T7 promoter in the pETDuet™-l vector. Following transformation into E. coli, the inventors purified the encapsulin particles using affinity chromatography targeting the surface-exposed Strep-tag at the C-terminus of the encapsulin protein. Subsequently, encapsulin was purified by size-exclusion chromatography. Control samples, NfsB and tdNfsB without encapsulation, were cloned and purified in accordance with the procedure outlined above (Figure IB).

[0173] Characterization of the Purified Encapsulated Nitroreductase

[0174] Blue native-PAGE (BN-PAGE) analysis of the purified constructs unveiled a large protein assembly (Figure IB). The upper band likely represents a slightly larger isoform or an aggregate of encapsulin, consistent with observations in other studies and in agreement with the results from the analytical Superose™ 6 runs (33, 47). Based on the protein standards, the observed bands correspond to approximate sizes of 1.0 and 0.9 MDa, which are lower than the expected molecular weight of the assembled 60-mer (e.g. Enc{tdNfsB} ~2.0 MDa). However, this discrepancy is consistent with findings from other studies, where T1 capsids with a theoretical weight of 1.8 MDa often appear around 1.0 MDa under native-PAGE conditions. Under denaturing conditions, distinct protein bands emerged, indicating the presence of encapsulin monomers at the expected size of 29 kDa, and either a tandem nitroreductase (51 kDa), or a single nitroreductase (25 kDa) (Figure 1C).

[0175] Transmission electron microscopy (TEM) and dynamic light scattering (DLS) analyses of the purified encapsulin constructs confirmed the formation of nanocompartments. The average outer diameter measured 23.8 ± 0.5 nm (n=100) for Enc{tdNfsB} and 22.1 ± 1.5 nm (n=100) for Enc{NfsB}. Moreover, a mean hydrodynamic diameter of 32.3 nm was determined for Enc{tdNfsB}, and 27.7 nm for Enc{NfsB} (Figure ID). Notably, both samples exhibited a low polydispersity index (PDI) of 0.10 for Enc{tdNfsB} and 0.08 for Enc{NfsB}, indicating near- monodisperse characteristics.

[0176] The inventive method produces encapsulins of high purity with minimal contamination or aggregation, as indicated by less than 6 % impurities detected via Superose™ 6 chromatography. The particles are nearly monodisperse, exhibiting a PDI that is comparable to or better than those typically reported for native and engineered encapsulin samples.44,49This improved homogeneity is likely attributable to the purification process, which utilizes a purification tag and does not involve precipitation, but instead employs affinity chromatography. Additionally, the observed outer diameter of ~24 nm is consistent with the reported diameters of T1 capsids observed by the inventors and other research groups.

[0177] Overall Structure of Enc{tdNfsB}

[0178] To investigate the assembly of the Enc{tdNfsB} construct in detail, the inventors performed single-particle cryo-EM analysis of the purified capsids (Figure 2). The overall resolution of the encapsulin icosahedral shell was determined to be 2.22 A based on the gold standard Fourier Shell Correlation (FSC) 0.143 threshold.

[0179] The icosahedral reconstruction for the encapsulin shell closely matches those of previously published structures (42, 43). A protein model was built into the map de novo and a single unit refined prior to expansion to produce a final model. The calculated diameter of the shell was 24.0 nm, which correlates well with the expected solution hydrodynamic diameter and has a calculated internal volume of 3,000 nm3.

[0180] To determine the loading of tdNfsB within the shell, particle subtraction was carried out on the final particle stack. The encapsulin cage was removed and particles were aligned in 2-D prior to a low resolution 3-D reconstruction. Discrete densities were visible in the 2-D classification and the 3-D reconstruction yielded six individual points of density when refined from ab initio modelling (Figure 2B). Although the inventors were unable to resolve the secondary structure of the subunits, the inventors have established the presence of the cargo protein and can begin efforts to determine the precise stoichiometry. The inventors observed density matching six tdNfsB dimers within the capsid though these could simply represent areas of high occupancy within the capsid and do not necessarily correlate directly. However, the same number of guest proteins per capsid was found in earlier studies using inductively coupled plasma mass spectrometry (ICP-MS), super-resolution microscopy and fluorescent quantification of M. smegmatis engineered encapsulins, Enc{HaloTag-Ru2+}, Enc{HaloTag-AlexaFluor 647}, and Enc{eGFP} (34, 35). Nevertheless, the inventors employed gel-densitometry analysis to quantify the number of guest proteins within the capsid. Gel-densitometry analysis revealed ten to thirteen tdNfsB units per shell, which is ~1.9 times more than the number identified through our cryo-EM analysis. We conclude that the analysis likely identified densities that correspond to tdNfsB units interacting with the inner surface of the encapsulin via electrostatic interactions (48). However, the inventors cannot exclude the possibility that some tdNfsB units, which do not interact with the shell, remain unresolved in our cryo-EM analysis. This may be due to high flexibility and conformational heterogeneity of these units, preventing their clear resolution within the capsid's inner volume.

[0181] Pores of the Enc{tdNfsB} Complex

[0182] The crystal or cryo-EM structures of encapsulins from close structural homologs, have multiple openings in their capsid shells, including a threefold pore and a five-fold pore (25, 42-44). These openings are thought to be the channels for the substrates and / or products of the cargo protein.

[0183] In the inventor’s refined icosahedral model, the uncharged five-fold pore closely resembles that of the M. smegmatis dye-decolorizing peroxidase (DyP)-loaded encapsulin (PDB 7BOJ), with an inner circle diameter of 8.2 A as measured by MoleOnline (42). The cryo-EM map in this area was ambiguous however and appears to be the superimposition of flexible states. In an effort to resolve this, the pentameric subunit was refined in local refinement without symmetry constraints and yielded a more closed-type structure. The same pore inner circle diameter of this structure was 4.4 A (Figure 2C). The primary drive of this is a shift in the pore forming loop (measuring 3.84 A between the displaced Hisl88 alpha carbon). The underlying cause of this movement is unclear and may represent an equilibrium of states when resting at cellular pH. A similar dynamic has been observed in the 5-fold pore of the iron storage encapsulin, Enc{Ftn}, from the halophilic bacterium Haliangium ochraceum (45).

[0184] Enzymatic Activity of the Encapsulated Nitroreductase NfsB and tdNfsB

[0185] To assess the enzymatic activity of Enc{tdNfsB} and Enc{NfsB} in comparison to their free forms tdNfsB and NfsB, the inventors devised a straightforward assay utilizing luciferase as a read-out combined with a NfsB bioluminescent reporter substrate, caged-hydroxy CBT (6- hydroxy-2-cyanobenzothiazole, Figure 3 A) (4). Caged-hydroxy CBT is obtained by coupling a nitro-containing thiophenyl moiety (NfsB substrate) via a diamine linker to a luciferin substrate. Upon reduction of the nitro group to a hydroxylamine or amine, the molecule self-immolates and fragments to release hydroxy-CBT. Under basic conditions (pH 8.4), the liberated luciferin precursor hydroxy-CBT reacts with D-cysteine in a condensation reaction to form D-luciferin, which then undergoes enzymatic oxidation to an excited state, producing a measurable luminescent signal. This multi-step process is essential for generating the assay read-out. When analyzing encapsulated NfsB, the enzymatic and chemical reactions inside and outside the shell may be hindered by limited substrate or cofactor access due to the need for diffusion through the protein shell. However, caged-hydroxy CBT might accumulate inside the shell, resulting in higher initial substrate concentrations for the first reaction step. Similarly, limited diffusion of the product or fragmentation products out of the shell could lead to enzyme inhibition. Given that the inventors supply substrate and cofactors at concentrations at least 20 times higher than the enzyme concentration, the inventors believe substrate and cofactor access is not a limiting factor. Furthermore, the assay is designed as an end-point assay, and the inventors have chosen a time point where only 50% of the substrate is converted. Hence, the inventors believe that the assay is well suited to perform an initial screening on enzymatic activities of free and encapsulated enzyme.

[0186] This assay setup allowed the inventors to track the pathway of the NfsB substrate into the capsid, its conversion by NfsB, subsequent diffusion out of the capsid, and its use as a luciferase substrate, leading to the emission of a bioluminescent signal. Furthermore, this assay format provided valuable insights into the catalytic activity of NfsB, an enzyme that relies on two cofactors, flavin mononucleotide (FMN) and reduced nicotinamide adenine dinucleotide (NADH) (41). To quantify the assay, the inventors utilized hydroxyl-CBT as a luciferase substrate in the presence of D-cysteine, and generated a standard curve following the bioluminescence read-out. The inventor’s experiments revealed that non-encapsulated NsfB and tdNfsB exhibited similar activity levels (see Figure 3B). Specifically, the inventors observed 50% conversion of the substrate caged-hydroxy CBT after 60 min of incubation under the specified assay conditions. However, upon encapsulation, there was a notable decline in activity, particularly evident in the case of Enc{NfsB}, where only 10% conversion were observed. Conversely, Enc{tdNfsB} demonstrated substrate conversion comparable to those of the free enzymes NsfB and tdNfsB. By directly comparing Enc{NfsB} and Enc{tdNfsB}, where differing enzymatic activities due to shell effects can be excluded, the inventors validate that monomeric NfsB is significantly less active than dimeric NfsB when encapsulated. Free NfsB and tdNfsB however showed similar conversion rates, showing that both enzymes are equally active in solution. This supports the inventor’s initial hypothesis that dimer formation of monomeric NfsB within the confines of the capsid may be challenging.

[0187] Additionally, the inventors conducted the assay with varying amounts of cofactors and screening concentrations ranging from 0 to 400 pM, and 0 to 400 nM for NADH and FMN, respectively. Lower concentrations of NADH affected the activity of Enc{tdNfsB}, while the free tdNfsB and NfsB remained unaffected within the tested concentration range. Similar trends were observed for FMN, albeit less pronounced. Fitting the Michaelis-Menten kinetics to the results of the NADH concentration screen revealed similar kcat values for free and encapsulated tdNfsB. However, there was a significant difference in KM values: 30 pM for Enc{tdNfsB} and 3.1 pM for free tdNfsB. This suggests that higher cofactor concentrations might be necessary to overcome potential bottlenecks in mass transport across the protein shell. Furthermore, from this assay, the inventors inferred that NfsB may encounter difficulties in forming active dimers within the protein shell. Consistent with this notion, the results of the assay with varying FMN concentrations indicate that only Enc{tdNfsB} can effectively bind FMN, a process occurring at the dimer interface (46). Conversely, increasing FMN concentration had no discernible effect on the activity of Enc{NfsB}. This underscores the advantage of using tdNfsB as cargo, as it ensures the encapsulation of functionally active enzyme dimers within the encapsulin structure. Finally, the optimal molar ratio of NfsB / tdNfsB and FMN as 1 : 40, as well as NfsB / tdNfsB and NADH as 1 : 40,000, was established for subsequent assays. Enc{NfsB} was excluded from further analysis due to its apparent lack of enzymatic activity.

[0188] To validate the inventor’s findings through a direct assay, the inventors employed a High- Performance Liquid Chromatography (HPLC) format utilizing a conventional NfsB substrate, (l-(4-nitrobenzoyl)piperidine, Nbzp) (5). Nbzp has been identified as a promising prodrug candidate potentially activated by NfsB. The reaction between Nbzp and NfsB was monitored, capturing a time-dependent profile of substrate depletion. Analysis of the HPLC results revealed a conversion rate of > 60% of the parent compound within 50 min, with the primary reaction product identified as l-(4-hydroxylaminobenzoyl)piperidine (HAbzp). Both free enzymes, NfsB and tdNfsB, exhibited comparable substrate conversion rates throughout the recorded timeframe (Figure 3C). Interestingly, in the direct HPLC assay, encapsulated tdNfsB exhibited a slightly decreased initial reaction velocity Vo (Figure 3D), however was able to convert the same concentration of Nbzp within the observed timeframe, as compared to free enzyme. Potentially, this effect stems from hindered substrate diffusion towards the interior of encapsulin.

[0189] In order to show the general applicability of Enc{tdNfsB} in reducing nitroaromatic compounds, the inventors employed a number of known nitroreductase substrates in the above assay (MA60, AV2, AV4, MA63, (4, 40) CB1954, (39) and CoNO? (47). Firstly, the inventors tested the product conversion of all selected candidates after 30 minutes of incubation with free tdNfsB. The identity of the reaction products was confirmed through mass spectrometry analysis. All molecules were converted by the tandem enzyme, validating that not only the free NfsB dimers, as reported in the literature, but also the fused dimeric NfsB is active. As expected, the different molecules exhibited varying conversion rates, likely due to their different affinities towards NfsB. (48) Apart from varying affinity to the enzyme, different physicochemical properties such as the reduction potential of the nitro group potentially also influence conversion rates. Upon investigating substrate conversion with Enc{tdNfsB}, the inventors observed conversion of all molecules, with slight differences between the encapsulated and free tdNfsB. Notably, due to the distinct molecular structures, the ability of each molecule to diffuse into the capsid potentially varies. Consequently, depending on the substrate, conversion rates may be higher with either the free or encapsulated enzyme. Higher conversion rates within the capsid are most likely due to substrate accumulation, which enhances enzymatic turnover in the confined space. Interestingly, the two molecules AV2 and AV4, which both carry a potentially charged piperazine moiety under the applied conditions, exhibited significantly higher conversion rates within Enc{tdNfsB}. This increase in conversion may be due to electrostatic interactions within the 5-fold pore region, potentially enhancing diffusion rates and accumulation of AV2 and AV4 inside the capsid.

[0190] Construction of Pore Mutants of Enc{tdNfsB}

[0191] Since the inventors have observed a slight but noticeable decrease in Enc{tdNfsB} activity in the activity assays, the inventors proposed that pore diffusion of the substrate or the NfsB cofactors are a limiting factor and slowed down the reaction, especially at low substrate concentrations. Hence, the inventors constructed four encapsulin mutants and increased the pore size of the five-fold pore by removing six amino acids in the pore-forming loop by site- directed mutagenesis. Encapsulin is built from 60 identical monomers, and five monomers that form a pentameric structure give rise to one five-fold pore, twelve of which exist per capsid. As was shown previously, the 5-fold loop forming region can tolerate amino acid substitutions and deletions without affecting self-assembly and stability of the encapsulin shell (49, 50). Encapsulin variants shown in Figure 4A were generated. Computational models of encapsulin variants A4-Ala2, A4-Alal, A6-Ala3, A6-Ala2 estimated enlarged pore diameters of 10.2 A, 11.6 A, 12.8 A, and 14.2 A, respectively. Subsequently, the inventors employed these mutants and determined their effects on the enzymatic activity of encapsulated tdNfsB. A total increase in conversion of Nbzp after 30 min by 12.8% (p < 0.01), 16.7% (p < 0.05), 16.7% (p < 0.01), and 8.2% (n.s.) was observed for EncA4-Ala2 {tdNfsB}, EncA4-Alal {tdNfsB}, EncA6- Ala3 {tdNfsB}, and EncA6-Ala2 {tdNfsB}, respectively (Figure 3C), in comparison to Enc{tdNfsB}. Kinetic analysis of the two pore mutants EncA4-Alal {tdNfsB} and EncA6- Ala3{tdNfsB} gave KM values of 5.66xl0’4M and 3.67xl0’4M, as well as kcat values of 16.61 s'1and 17.93 s’1, respectively. This suggests that the decreased catalytic activity of the encapsulated enzyme can be improved upon pore modulation. Given the absence of a clear correlation between pore size and conversion rate, the enhancement may also be influenced by the electrostatic properties of the pore surface. As illustrated in Figure 4A, all pore mutants exhibit a neutral to negative (red) charge in the pore region, in contrast to the positively charged wild-type pore. Therefore, in addition to pore size, which may only affect substrate diffusion up to a certain threshold, electrostatic interactions between the substrate and pore surface could also be a limiting factor. However, the observed increase in turnover across all four mutants may also arise from other mechanisms, such as the introduction of assembly defects via the mutations that enhance substrate access or differences in enzyme cargo loading within the mutant cages.

[0192] Activity of Enc{tdNfsB} vs tdNfsB in cellular environments

[0193] The application of nitroreductase enzymes encompasses a wide array of scenarios, often requiring robust enzymatic activity under diverse and challenging conditions. These conditions include elevated temperatures, cellular lysates, extracellular matrices, blood samples, immobilization on solid supports, and exposure to organic solvents. In the inventor’s experiments, the inventors aimed to investigate whether encapsulating NfsB within a protein cage could enhance its performance. Initially, the inventors assessed the efficacy of Enc{tdNfsB} compared to free tdNfsB under various conditions, including exposure to protease, heat, cell lysate, and in lysis buffer. The inventors used fluorescence read-out with the profluorophore 7-(diethylamino)-3-nitro-2H-chromen-2-one (CoNO?), which NfsB converts to the fluorophore 3-amino-7-(diethylamino)-2H-chromen-2-one (C0NH2), as a reliable method to quantify NfsB activity. Similar activity levels were observed between Enc{tdNfsB} and tdNfsB in Pierce™ IP lysis buffer or cell lysates (Figure 5A). However, Enc{tdNfsB} exhibited a remarkable increase in enzymatic activity compared to tdNfsB in the presence of protease (Figure 5 AB). While the free enzyme was completely inactive after a 10-minute incubation, the encapsulated enzyme maintained comparable activity to the non-treated condition. This highlights the significant beneficial effect of encapsulation on tdNfsB activity, particularly under challenging conditions. A similar but less pronounced effect was observed upon heat treatment of both free and encapsulated NfsB. Encapsulation slightly improved enzymatic activity.

[0194] Subsequently, the inventors investigated whether Enc{tdNfsB} retained its activity on living cells, specifically its ability to convert the prodrug Nbzp into its active form for subsequent cellkilling activity. Co-administration of Enc{tdNfsB} and the prodrug Nbzp to pre-seeded lung carcinoma cells (H1299) resulted in complete loss of cell viability compared to controls utilizing encapsulated eGFP (enhanced green fluorescent protein), Enc{eGFP}, which confirms the efficient conversion of Nbzp into its active metabolite and subsequent uptake by H1299 cells. This hypothesis was supported further by assays using the profluorophore CONO2. Upon co-administration of Enc{tdNfsB} and CoNO2 to pre-seeded HeLa cells (human cervical cancer cells), fluorescence accumulation within the cells was evident after two hours of incubation, which confirms the conversion of CONO2 and subsequent uptake of C0NH2 by HeLa cells (Figure 6). Furthermore, the inventor’s experiments indicated that Enc{tdNfsB} exhibited no toxicity towards HeLa or HT1299 cells.

[0195] Construction of the encapsulated cytosine deaminase Enc{tdCD}

[0196] The inventors introduced the encapsulin gene MSMEI 5672 from Mycobacterium smegmatis mc2155, tagged with a C-terminal Strep-tag, under the first T7 promoter of the pETDuet-1 vector. Given that yeast cytosine deaminase (CD) functions as an obligate homodimer, and the inventors previous findings have shown that successful dimer formation and retention of enzymatic activity upon encapsulation require the enzyme monomers to be connected by a flexible linker (91), CD was cloned as a tandem enzyme (tdCD) under the second T7 promoter of pETDuet-1. This was achieved by linking two copies of the CD gene FCY1 with a flexible linker sequence and appending a C-terminal ELS sequence for directing tdCD to the interior of encapsulin (Figure 7A). Following transformation into E. coli, we purified the encapsulin particles using affinity chromatography, targeting the surface-exposed Strep-tag at the C- terminus of the encapsulin shell. Subsequently, encapsulin was further polished through sizeexclusion chromatography (Figure 7D). As a control, encapsulin harboring eGFP was cloned and produced, following the same purification procedure described above.

[0197] Characterization of encapsulin constructs Enc{tdCD} and Enc{eGFP}

[0198] The structural integrity and assembly of the purified encapsulin constructs were analyzed using Blue Native-P AGE (BN-PAGE), revealing the presence of a large protein complex (Figure 7B). A prominent lower band corresponds to correctly assembled T1 capsids, while a weaker upper band likely indicates a slightly larger isoform or protein aggregates. Based on protein standards, these bands correspond to approximate sizes of 1.2 and 0.9 MDa, respectively, which are lower than the expected molecular weight of the fully assembled 60-mer. This discrepancy aligns with findings from previous studies, where T1 capsids theoretically weighing 1.8 MDa often migrate around 1.0 MDa under native conditions (69, 91, 92).

[0199] Under denaturing conditions, SDS-PAGE revealed two distinct protein bands corresponding to encapsulin monomers (30 kDa) and either tandem cytosine deaminase (38.4 kDa) or eGFP (28.3 kDa) in the Enc{tdCD} and Enc{eGFP} constructs, respectively (Figure 7C).

[0200] To achieve high-resolution characterization and distinguish properly assembled particles from potential aggregates or defective structures, we employed AF4-MALS, a state-of-the-art method for virus-like particle (VLP) characterization. AF4-MALS analysis identified a primary, well-resolved peak at 12 mL, corresponding to correctly assembled T1 capsids, and a minor aggregation peak at 15 mL (Figure 7E), contributing to less than 5% of the total area under the curve (AUC). The observation of two distinct peaks in AF4-MALS corresponds well with the two bands observed in BN-PAGE. The mass-weighted mean of the assigned molecular masses for the primary peaks were 2513 ± 14 kDa for Enc{tdCD}, 2284 ± 28 kDa for Enc{eGFP}, and 1852 ± 24 kDa for empty encapsulin shell. From these data, we estimated the cargo load per nanocompartment to comprise 18.7 ± 0.2 tdCD molecules for Enc{tdCD} and 17.0 ± 1.0 eGFP molecules for Enc{eGFP}.

[0201] To further confirm the formation of encapsulin nanocompartments, Transmission Electron Microscopy (TEM) and Dynamic Light Scattering (DLS) analyses were performed (Table 1). TEM micrographs demonstrated well-defined, spherical nanostructures with an average outer diameter of 25.0 ± 1.3 nm (n = 100) for Enc{tdCD} and 22.4 ± 1.6 nm (n = 100) for Enc{eGFP} (Figure IF). DLS measurements revealed hydrodynamic diameters of 28.8 ± 0.1 nm (PDI = 0.11 ± 0.01) for Enc{tdCD} and 29.5 ± 0.3 nm (PDI = 0.14 ± 0.01) for Enc{eGFP} (Table 1, Figure 13), further confirming the formation of stable nanocompartments.

[0202] Table 1: Size characterization of encapsulin constructs via TEM, AF4-MALS, and DLS

[0203] Sample TEM external AF4-MALS DLS hydrodynamic diameter (nm) diameter of diameter (nm) gyration (nm)

[0204] These results collectively validate the successful encapsulation and structural integrity of the engineered nanocompartments, paving the way for their downstream applications.

[0205] In vitro and in cellulo prodrug activation by Enc{tdCD}

[0206] To verify that the encapsulated cytosine deaminase (CD) retains its enzymatic activity, an in vitro assay was performed by incubating Enc{tdCD} with the substrate 5-FC. HPLC analysis confirmed the conversion of 5-FC to the product 5-FU within one hour of incubation (Figure 8A, Figure 14). These results indicate that the encapsulation process preserves the enzymatic activity of cytosine deaminase. Building on the in vitro results, the inventors next evaluated whether Enc{tdCD} retains its enzymatic activity in living cells, aiming to explore its potential for enzyme-prodrug therapy. Specifically, the inventors examined its capacity to convert the prodrug 5-FC into the chemotherapeutic agent 5-FU, which leads to cell death, by assessing cell viability. To investigate this, Enc{tdCD} was incubated with the mouse monocyte cell line J774A.1, known to internalize encapsulin nanocompartments (93, 94). Following incubation, unbound nanoparticles were removed by washing, and the 5-FC prodrug was subsequently added. This treatment resulted in a significant reduction in cell viability to 18.1% (Figure 8B) compared to controls treated with Enc{eGFP}, demonstrating the efficient intracellular conversion of 5-FC to 5-FU by Enc{tdCD}. Importantly, neither Enc{tdCD} nor Enc{eGFP} exhibited inherent toxicity to the cells when applied alone; in fact, they resulted in small increases in cell viability (12.5% and 7.9%, respectively).

[0207] Evolution of encapsulin shell for specific cell targeting

[0208] These promising results motivated the inventors to design surface modification of the encapsulin shell for targeted cell delivery. The inventors selected the HER2 receptor, whose abnormal expression is correlated with the development and progression of various cancers (53- 56), as well as enhanced invasiveness and resistance to chemotherapy (57, 58). In recent years, a variety of affibodies targeting the HER2 receptor have been reported (95, 96). Affibodies are antibody mimetics of small size, high target affinity and ease of bacterial expression. Among these, the ZHER2:342 affibody demonstrates strong HER2 binding (22 pM)(96) and retains its specificity even when fused at its N- or C-terminus, making it an ideal candidate for decorating the surface of the encapsulin shell (97, 98).

[0209] Initially, we attempted to construct a vector analogous to the one for Enc{tdCD} by placing an encapsulin-ZHER2:342 affibody fusion (with a C-terminal His-tag) under the first T7 promoter in the pETDuet-1 vector and the tdCD sequence under the second T7 promoter. C-terminal encapsulin fusions, such as purification tags, have been shown to be presented on the outer surface of the encapsulin shell (69, 75, 91). However, this single-vector design failed to yield fully assembled nanocompartments (Full-ZHEREnc{tdCD}), as indicated by the absence of high-molecular-weight species in BN-PAGE (Figure 15), likely due to steric hindrance introduced by the bulky affibody moiety, which disrupted proper capsid assembly. To overcome this limitation, we aimed to co-express a wild-type encapsulin gene alongside the affibody- fused variant and tdCD, ensuring that mixed populations of monomers - both with and without the affibody - are available for assembly. This strategy mitigates steric interference and facilitates the successful formation of stable nanocompartments. Accordingly, a set of plasmids was prepared for the co-expression of three proteins required for the assembly of the nanocompartment (Figure 9A). The sequence encoding the encapsulin-ZHER2:342 affibody fusion, followed by C-terminal His-tag was cloned under the first T7 promoter of the pCDFDuet-1 vector. The unmodified encapsulin gene sequence was inserted under the second T7 promoter of the same vector. The pETDuet-1 vector was designed to accommodate the sequence for the cargo protein, either tdCD or eGFP followed by a C-terminal ELS sequence. Each combination of plasmids was co-transformed into E. coli BL21 Star (DE3), and following expression, the fully assembled nanocompartmentsZHEREnc{tdCD} andZHEREnc{eGFP}, were purified via affinity chromatography and further polished through size-exclusion chromatography (Figure 9D).

[0210] Characterization of shell-modified encapsulin constructsZHEREnc{tdCD} andZHEREnc{eGFP}

[0211] For encapsulin constructs with a ZHER2:342 affibody-decorated shell surface, BN-PAGE revealed two bands analogous to previous constructs: a strong lower band and a weaker upper band, corresponding to approximate sizes of 1.3 and 1.1 MDa, respectively (Figure 9B). Compared to the unmodified constructs, these results indicate an increase in molecular weight, consistent with the successful integration of the affibody into the encapsulin shell. The broader and less distinct appearance of these bands can be potentially attributed to the heterogeneity introduced during assembly. The encapsulin shell comprises both wild-type and affibody-fused encapsulin monomers. Wild-type encapsulin plays a crucial role in preventing steric hindrance between affibody-fused monomers during the assembly process. As the ratio of these two protein types is not fixed and may vary slightly, this variability likely contributes to the observed band broadening.

[0212] Under denaturing SDS-PAGE conditions, three distinct protein bands were observed: encapsulin monomers (28.8 kDa), affibody-fused encapsulin monomers (37.8 kDa), and either tandem cytosine deaminase (38.4 kDa) or eGFP (28.3 kDa) in theZHEREnc{tdCD} andZHEREnc{eGFP} constructs, respectively (Figure 3C). To quantitatively assess the composition of the encapsulin shells, densitometric analysis of the SDS-PAGE gels was performed (Figure 16). Since each nanocompartment is composed of 60 identical subunits, this analysis allowed estimation of the relative abundance of affibody-fused versus wild-type encapsulin monomers. The results revealed that the shell ofZHEREnc{tdCD} contains approximately 21.0 ± 0.8 affibody-fused (ZHEREnc) and 39.0 ± 0.8 wild-type (Enc) subunits. Similarly, theZHEREnc{eGFP} construct was composed of 22.2 ± 0.4ZHEREnc and 37.8 ± 0.4 Enc subunits. These values indicate that the shell consists of roughly one-third affibody-modified and two- thirds wild-type protomers. AF4-MALS analysis of the affibody-modified constructs showed a primary, well-resolved peak at 12 mL, with a secondary peak at 15 mL (Figure 9E). While the secondary peak was more pronounced compared to the unmodified constructs, the overall profile suggested a substantial degree of sample homogeneity. The increased mass-weighted mean of molecular masses for the primary peaks, determined to be 2574 ± 22 kDa forZHEREnc {tdCD} and 2376 ± 36 kDa forZHEREnc {eGFP}, provides further evidence of successful surface modification. Upon nanocompartment assembly, the affibody protrudes from the external surface of the encapsulin, which is expected to increase both the outer diameter and the diameter of gyration. This was confirmed through AF4-MALS and TEM, both of which detected larger sizes for the surface- modified constructs compared to the unmodified ones. AF4-MALS estimated mass-weighted mean diameter of gyration of 21.2 ± 0.4 nm and 21.0 ± 0.6 nm for theZHEREnc{tdCD} andZHEREnc{eGFP} constructs, respectively.

[0213] TEM micrographs revealed spherical nanostructures with well-preserved morphology and an increased average outer diameters compared to the unmodified constructs: 27.6 ± 2.0 nm (n = 100) forZHEREnc{tdCD} and 26.2 ± 1.3 nm (n = 100) forZHEREnc{eGFP} (Figure 9F). DLS measurements further confirmed these findings, revealing consistent hydrodynamic diameters of 29.6 ± 0.0 nm (PDI = 0.18 ± 0.09) forZHEREnc{tdCD} and 28.7 ± 0.1 nm (PDI = 0.11 ± 0.01) forZHEREnc{ eGFP} (Figure 13, Table 1).

[0214] Collectively, these results demonstrate the successful integration of the ZHER2:342 affibody into the encapsulin shell while maintaining its structural integrity despite introduced surface modification. This opens up possibilities for targeted downstream applications.

[0215] Cell-based affinity characterization

[0216] The SKOV-3 cell line, known for its high HER2 expression, was used to evaluate the binding specificity of the designed constructs. For fluorescent microscopy SKOV-3 cells were treated with eitherZHEREnc{eGFP}, Enc{eGFP}, or left untreated. Green fluorescence was observed only in cells incubated withZHEREnc {eGFP} after washing (Figure 10), indicating that only the affibody-decorated constructZHEREnc{eGFP} was able to bind specifically to SKOV-3 cells, whereas the non-modified construct Enc{eGFP} did not.

[0217] Flow cytometry was employed for further investigation and relative quantification of the binding of constructs to SKOV-3, as well as OV7 cell lines.ZHEREnc{eGFP} and Enc{eGFP} were tested at three amounts: 20 pmol, 10 pmol, and 3.3 pmol per l * 106cells. The Mean Fluorescence Intensity (MFI) of SKOV-3 cells incubated with Enc{eGFP} showed no significant difference in comparison to untreated cells, with less than 0.005% of cells exhibiting GFP positivity. In contrast, treatment withZHEREnc{eGFP} resulted in a significant increase in both MFI and the percentage of GFP-positive cells in a concentration-dependent manner. At the highest applied amount of 20 pmolZHEREnc{eGFP}, the MFI was 62-fold higher than that of untreated control, and 98.7% of the cells were GFP-positive (Figure 11A and C). In experiments conducted with the OV7 cell line, which is characterized by low HER2 expression on its cell surface (106, 107), treatment with Enc{eGFP} yielded a similar pattern. The percentage of GFP-positive cells was below 0.1% and the MFI exhibited no significant difference compared to untreated cells. Upon incubation withZHEREnc{eGFP}, the MFI increased 2-fold and the percentage of GFP-positive cells rose to 3.2% when the highest amount of 20 pmolZHEREnc{eGFP} was applied (Figure 11B and D). These results correlate with the low presence of the HER2 receptor on the OV7 cell surface.

[0218] Overall, the data confirm the highly specific affinity of theZHEREnc{eGFP} construct over unmodified Enc{eGFP} toward HER2, qualifying it for specific cell targeting.

[0219] In cellulo prodrug activation byZHEREnc{tdCD}

[0220] To advance the investigation, the investigators examined the ability ofZHEREnc{tdCD} to activate the prodrug 5-FC in SKOV-3 cells (Figure 12). Control wells were either left untreated, treated with 4 mM of the prodrug 5-FC alone, with 2 mM of the active drug 5-FU, or with the protein constructsZHEREnc{tdCD} orZHEREnc{eGFP} in the absence of prodrug. Experimental wells received eitherZHEREnc{tdCD} orZHEREnc{eGFP}, followed by a washing step and subsequent incubation with 4 mM 5-FC.

[0221] Neither the prodrug 5-FC nor the protein constructsZHEREnc{tdCD} andZHEREnc{eGFP} exhibited cytotoxicity on their own, as cell viability remained high across all respective treatments. Similarly, treatment withZHEREnc{eGFP} in combination with 5-FC had no effect on cell viability, confirming that the encapsulin scaffold harboring non-enzymatic cargo is nontoxic and does not facilitate prodrug conversion. In contrast, the combination ofZHEREnc{tdCD} and 5-FC led to a substantial reduction in cell viability to 17.0%, indicating successful enzymatic conversion of the prodrug into cytotoxic 5-FU. Treatment with the positive control 2 mM 5-FU (higher concentrations could not be tested due to solubility limitations in cell culture medium), led to slightly higher cytotoxicity (11.8%) compared toZHEREnc{tdCD} with 4 mM 5-FC (17.0%) condition. This difference may be attributed to the encapsulation context, where delayed diffusion of the enzymatically generated 5-FU from the nanocompartment could attenuate its immediate cytotoxic effect. This outcome underscores the prodrug-activating capability ofZHEREnc{tdCD} and its potential as a targeted therapeutic platform. However, several factors may influence the observed efficacy. One key consideration is the efficiency of 5-FC conversion within the encapsulin shell. Previous studies have shown that encapsulin pores function as selective molecular sieves, with transport governed by the size and charge properties of both the pores and the diffusing molecules (101-103). Reduced catalytic efficiency in protein-based nanoreactors is often attributed to an incompatibility between the selective permeability of the protein shell and the physicochemical properties of the encapsulated enzyme’s substrate (103). In this context, limited diffusion of 5-FC into the nanocompartment may reduce substrate availability for enzymatic conversion, while restricted release of 5-FU from the capsid could delay or diminish its cytotoxic effects.

[0222] Future investigations could focus on strategies to enhance molecular flux through the encapsulin shell. Rational pore engineering to modulate size and / or charge could improve the diffusion of both substrate and product, while preserving structural integrity (103). Notably, encapsulins have shown a general tolerance to pore-region mutations, with many modifications maintaining nanocompartment stability and proper assembly, making them attractive scaffolds for further optimization (72, 80).

[0223] Together, these findings demonstrate the potential ofZHEREnc{tdCD}-mediated prodrug activation in a targeted context and open avenues for enhancing therapeutic efficacy through structural tuning of the nanoreactor.

[0224] These examples represent successful application of a designer encapsulin in targeted (affibody- directed) enzyme prodrug therapy. The inventors have demonstrated the encapsulation of tandem cytosine deaminase within an engineered encapsulin nanocompartment, displaying its potential for targeted prodrug activation in a cellular context, particularly in cancer therapy, which significantly enhances the therapeutic efficacy of 5-FC by facilitating its localized conversion to the active drug 5-FU directly within the microenvironment of tumor cells. By ensuring that enzymes remain concentrated at the site of action this strategy has the potential to drastically reduce systemic toxicity associated with traditional chemotherapy.

[0225] Production, characterization and validation of encapsulated tandem enzymes under industrial conditions The focus is on oxidoreductases and hydrolases, particularly peroxidases, which are well- known in technological applications and are found in detergents. Examples include horseradish peroxidase, ligninase, and haloperoxidases such as chlorine and bromine peroxidase.

[0226] Optimizing encapsulation of enzymes and production of nanoparticles

[0227] The aim of this program is the production of encapsulin-nanoparticles with encapsulated tandem-enzyme in sufficient amount and quality.

[0228] Different TP-fused tandem-enzyme variants are expressed in E. coli and the nanoparticles are purified using FPLC. The efficiency of encapsulation is determined using SDS-PAGE, Western Blot, and quantitative protein analyses. The size and structure of nanoparticles is determined using DLS and TEM. This leads to the reproducible production of stable nanoparticles showing a high enzyme loading.

[0229] Biochemical characterization of encapsulated enzymes

[0230] In order to compare the catalytic activity and basic stability between encapsulated and free enzyme, measurements using standardized substrate conversion methods (e.g. ABTS-, DMP- or EECE-conversions for oxidoreductases) are used. For thermostability-assays, an incubation at different temperatures (30-90 °C) and a detection of the residual activity is used. In order to generate pH-stability profiles, an incubation in buffers at pH 2-12 is used in order to detect the robustness against different pH-values. Furthermore, a test of the tolerance against inhibitors (heavy metals, detergents) is performed. The tests show an improved stability of the encapsulated enzyme against heat, changes of pH, and inhibitors.

[0231] Simulation of industrial process conditions

[0232] For a simulation of industrial process conditions, the encapsulated enzymes are examined for their stability and activity. For this, an incubation of the samples in process media (e.g. organic solvents, textile bleach, sewage samples) at relevant temperatures, pH-values and inhibitors is performed. In order to evaluate long-term stability and reusability, time series are recorded. Furthermore, analysis of enzyme activity and nanoparticle integrity after exposure is performed. The tests show that encapsulated enzymes exhibit significantly higher resistance in process media than free enzymes.

[0233] Investigation of release mechanisms and functionality To determine whether and how the enzymes are released from the encapsulins under industrial process conditions, release mechanism and functionality are investigated. For this, an incubation of the samples in process media with and without a specific trigger (e.g. reducing agents, pH change) is performed. Enzyme activity is measured in the supernatant and in the nanoparticles. In order to determine the release over time kinetic studies are performed. The tests show high stability of encapsulin and no significant release of the encapsulated enzyme.

[0234] References

[0235] 1. R. S. Boddu, O. Perumal and D. K, Biotechnol Appl Biochem, 2021, 68, 1518-1530.

[0236] 2. S.-H. Kim, S. Park, E. Park, J.-H. Kim, S. Ghatge, H.-G. Hur and S. Rhee, J Biol Chem, 2021, 297, 101143.

[0237] 3. M. D. Roldan, E. Perez-Reinado, F. Castillo and C. Moreno-Vivian, FEMS Microbiology Reviews, 2008, 32, 474-500.

[0238] 4. E.-M. Herrlinger, M. Hau, D. M. Redhaber, G. Greve, D. Willmann, S. Steimle, M. Muller, M. Liibbert, C. C. Miething, R. Schiile and M. Jung, Chembiochem, 2020, 21, 2329-2347.

[0239] 5. T. Gungor, F. C. Onder, E. Tokay, U. G. Giilhan, N. Hacioglu, T. T. Tok, A. elik, F. Kbgkar and M. Ay, Eur. J. Med. Chem., 2019, 171, 383-400.

[0240] 6. T. D. Gruber, C. Krishnamurthy, J. B. Grimm, M. R. Tadross, L. M. Wysocki, Z. J. Gartner and L. D. Lavis, ACS Chem. Biol., 2018, 13, 2888-2896.

[0241] 7. K. Sharma, K. Sengupta and H. Chakrapani, Bioorganic & Medicinal Chemistry Letters, 2013, 23, 5964-5967.

[0242] 8. P. F. Searle, M.-J. Chen, L. Hu, P. R. Race, A. L. Lovering, J. I. Grove, C. Guise, M. Jaberipour, N. D. James, V. Mautner, L. S. Young, D. J. Kerr, A. Mountain, S. A. White and E. I. Hyde, Clin Exp Pharmacol Physiol, 2004, 31, 811-816.

[0243] 9. E. M. Williams, R. F. Little, A. M. Mowday, M. H. Rich, J. V. E. Chan-Hyams, J. N. Copp, J. B. Smaill, A. V. Patterson and D. F. Ackerley, Biochem. J., 2015, 471, 131-153.

[0244] 10. J. N. Copp, A. M. Mowday, E. M. Williams, C. P. Guise, A. Ashoorzadeh, A. V. Sharrock, J. U. Flanagan, J. B. Smaill, A. V. Patterson and D. F. Ackerley, Cell Chemical Biology, 2017, 24, 391-403.

[0245] 11. J. A. Plumb, A. Bilsland, R. Kakani, J. Zhao, R. M. Glasspool, R. J. Knox, T. R. J. Evans and W. N. Keith, Oncogene, 2001, 20, 7797-7803.

[0246] 12. D. C. Stein, E. Carrizosa and S. Dunham, BMC Microbiology, 2009, 9, 239. 13. A. M. Mowday, J. N. Copp, S. P. Syddall, L. J. Dubois, J. Wang, N. G. Lieuwes, R. Biemans, A. Ashoorzadeh, M. R. Abbattista, E. M. Williams, C. P. Guise, P. Lambin, D. F. Ackerley, J. B. Smaill, J. Theys and A. V. Patterson, Iheranostics, 2020, 10, 10548-10562.

[0247] 14. M. E. de la Calle, G. Cabrera, D. Cantero, A. Valle and J. Bolivar, Microbial Cell Factories, 2019, 18, 86.

[0248] 15. R. O. M. A. de Souza, L. S. M. Miranda and U. T. Bornscheuer, Chemistry - A European Journal, 2017, 23, 12040-12063.

[0249] 16. S. C. Cosgrove, G. J. Miller, A. Bornadel and B. Dominguez, ACS Sustainable Chem. Eng., 2023, 11, 8556-8561.

[0250] 17. F. Bialas, D. Reichinger and C. F. W. Becker, Enzyme and Microbial Technology, 2021, 150, 109864.

[0251] 18. D. Sanjanwala, V. Londhe, R. Trivedi, S. Bonde, S. Sawarkar, V. Kale and V. Patravale, Int J Biol Macromol, 2024, 256, 128488.

[0252] 19. Q. Sun, J. Shi, H. Sun, Y. Zhu and J. Du, Biomacromolecules, 2023, 24, 4587-4604.

[0253] 20. Y. Weng, G. Yang, Y. Li, L. Xu, X. Chen, H. Song and C.-X. Zhao, Adv Colloid Interface Sci, 2023, 318, 102957.

[0254] 21. A. N. Gabashvili, N. S. Chmelyuk, M. V. Efremova, J. A. Malinovskaya, A. S. Semkina and M. A. Abakumov, Biomolecules, 2020, 10, 966.

[0255] 22. T. W. Giessen, Curr Opin Chem Biol, 2016, 34, 1-10.

[0256] 23. S. Chakraborti, T.-Y. Lin, S. Glatt and J. G. Heddle, RSC Advances, 2020, 10, 13293- 13301.

[0257] 24. R. M. Putri, C. Allende-Ballestero, D. Luque, R. Klem, K.-A. Rousou, A. Liu, C. H.-H. Traulsen, W. F. Rurup, M. S. T. Koay, J. R. Caston and J. J. L. M. Cornelissen, ACS Nano, 2017, 11, 12796-12804.

[0258] 25. M. Sutter, D. Boehringer, S. Gutmann, S. Gunther, D. Prangishvili, M. J. Loessner, K. O. Stetter, E. Weber-Ban and N. Ban, Nat Struct Mol Biol, 2008, 15, 939-947.

[0259] 26. C. Cassidy-Amstutz, L. Oltrogge, C. C. Going, A. Lee, P. Teng, D. Quintanilla, A. East- Seletsky, E. R. Williams and D. F. Savage, Biochemistry, 2016, 55, 3461-3468.

[0260] 27. T. N. Szyszka, L. S. R. Adamson and Y. H. Lau, in Microbial Production of High-Value Products, eds. B. H. A. Rehm and D. Wibowo, Springer International Publishing, Cham, 2022, pp. 309-333.

[0261] 28. J. A. Jones, M. P. Andreas and T. W. Giessen, Biomacromolecules, 2023, 24, 1388-1399.

[0262] 29. A. Tamura, Y. Fukutani, T. Takami, M. Fujii, Y. Nakaguchi, Y. Murakami, K. Noguchi, M. Yohda and M. Odaka, Biotechnology and bioengineering, 2015, 112, 13-20. 30. A. Van de Steen, H. C. Wilkinson, P. A. Dalby and S. Frank, ACS AppL Bio Mater., , D01:10.1021 / acsabm.3c01153.

[0263] 31. 1. Boyton, S. C. Goodchild, D. Diaz, A. Elbourne, L. E. Collins-Praino and A. Care, ACS Omega, 2021, 7, 823-836.

[0264] 32. P. Lohner, M. Zmyslia, J. Thum, J. K. Pape, R. Gerasimaite, J. Keller-Findeisen, S. Groeer, B. Deuringer, R. Suss, A. Walther, S. W. Hell, G. Lukinavicius, T. Hugel and C. Jessen-Trefzer, Angew Chem Int Ed Engl, 2021, 60, 23835-23841.

[0265] 33. P. Ebensperger, M. Zmyslia, P. Lohner, J. Braunreuther, B. Deuringer, A. Becherer, R. Siiss, A. Fischer and C. Jessen-Trefzer, Angew Chem Int Ed Engl, 2023, 62, e202218413.

[0266] 34. P. Lagoutte, C. Mignon, G. Stadthagen, S. Potisopon, S. Donnat, J. Mast, A. Lugari and B. Werle, Vaccine, 2018, 36, 3622-3628.

[0267] 35. H. Moon, J. Lee, H. Kim, S. Heo, J. Min and S. Kang, Biomater Res, 2014, 18, 21.

[0268] 36. S. H. Helalat, R. C. Tellez, E. A. Dezfouli and Y. Sun, Biomacromolecules, 2024, 25, 2762- 2769.

[0269] 37. F. Sigmund, C. Massner, P. Erdmann, A. Stelzl, H. Rolbieski, M. Desai, S. Bricault, T. P. Worner, J. Snijder, A. Geerlof, H. Fuchs, M. Hrabe de Angelis, A. J. R. Heck, A. Jasanoff, V. Ntziachristos, J. Plitzko and G. G. Westmeyer, Nat Commun, 2018, 9, 1990.

[0270] 38. D. Diaz, X. Vidal, A. Sunna and A. Care, ACS Appl. Mater. Interfaces, 2021, 13, 7977- 7986.

[0271] 39. A. Christoff erson and J. Wilkie, Biochem Soc Trans, 2009, 37, 413-418.

[0272] 40. M. Abdelsalam, M. Zmyslia, K. Schmidtkunz, A. Vecchio, S. Hilscher, H. S. Ibrahim, M. Schutkowski, M. Jung, C. Jessen-Trefzer and W. Sippl, Archiv der Pharmazie, 2024, 357, 2300536.

[0273] 41. J. Whiteway, P. Koziarz, J. Veall, N. Sandhu, P. Kumar, B. Hoecher and I. B. Lambert, J Bacteriol, 1998, 180, 5529-5539.

[0274] 42. Y. Tang, A. Mu, Y. Zhang, S. Zhou, W. Wang, Y. Lai, X. Zhou, F. Liu, X. Yang, H. Gong, Q. Wang and Z. Rao, Proc Natl Acad Sci USA, 2021, 118, e2025658118.

[0275] 43. J. A. Jones, M. P. Andreas and T. W. Giessen, Nat Commun, 2024, 15, 2558.

[0276] 44. N. Loncar, H. J. Rozeboom, L. E. Franken, M. C. A. Stuart and M. W. Fraaije, Biochem. Biophys. Res. Commun., 2020, 529, 548-553.

[0277] 45. J. Ross, Z. McIver, T. Lambert, C. Piergentili, J. E. Bird, K. J. Gallagher, F. L. Cruickshank, P. James, E. Zarazua-Arvizu, L. E. Horsfall, K. J. Waldron, M. D. Wilson, C. L. Mackay, A. Basle, D. J. Clarke and J. Marles-Wright, Science Advances, 2022, 8, eabj4461. 46. E. Akiva, J. N. Copp, N. Tokuriki and P. C. Babbitt, Proc Natl Acad Sci U S H, 2017 , 114, E9549-E9558.

[0278] 47. C. Jia, Y. Zhang, Y. Wang and M. Ji, RSC Adv, 11, 8516-8520.

[0279] 48. D. Liu, T. N. Wanniarachchi, G. Jiang, G. Seabra, S. Cao, S. D. Bruner and Y. Ding, RSC Chem Biol, 3, 436-446.

[0280] 49. E. M. Williams, S. M. Jung, J. L. Coffman and S. Lutz, ACS Synth. Biol., 2018, 7, 2514- 2517.

[0281] 50. T. N. Szyszka, M. P. Andreas, F. Lie, L. M. Miller, L. S. R. Adamson, F. Fatehi, R. Twarock, B. E. Draper, M. F. Jarrold, T. W. Giessen and Y. H. Lau, Proceedings of the National Academy of Sciences, 2024, 121, e2321260121.

[0282] 51. M. Mishra, M. Mishra and S. Dutta, Chemistry - A European Journal, n / a, e202400195.

[0283] 52. Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol. 2001;2(2):127-137. doi:l 0.1038 / 35052073.

[0284] 53. Stigbrand T, Carlsson J, Adams GP, eds. Targeted Radionuclide Tumor Therapy: Biological aspects. Dordrecht: Springer Netherlands; 2008.

[0285] 54. Morote J, Torres I de, Caceres C, Vallejo C, Schwartz S, Reventos J. Prognostic value of immunohistochemical expression of the c-erbB-2 oncoprotein in metastasic prostate cancer. Int. J. Cancer. 1999;84(4):421-425. doi: 10.1002 / (SICI)1097-0215(19990820)84:4<421 :AID- IJC16>3.0.CO;2-9.

[0286] 55. Nakamura H, Kawasaki N, Taguchi M, Kabasawa K. Association of HER-2 overexpression with prognosis in nonsmall cell lung carcinoma: a metaanalysis. Cancer. 2005;103(9):1865- 1873. doi: 10.1002 / cncr.20957.

[0287] 56. Verri E, Guglielmini P, Puntoni M, et al. HER2 / neu oncoprotein overexpression in epithelial ovarian cancer: evaluation of its prevalence and prognostic significance. Clinical study. Oncology. 2005;68(2-3): 154-161. doi:10.1159 / 000086958.

[0288] 57. Pegram M, Slamon D. Biological rationale for HER2 / neu (c-erbB2) as a target for monoclonal antibody therapy. Semin Oncol. 2000;27(5 Suppl 9):13-19.

[0289] 58. Serrano-Olvera A, Duenas-Gonzalez A, Gallardo-Rincon D, Candelaria M, La Garza-

[0290] Salazar J de. Prognostic, predictive and therapeutic implications of HER2 in invasive epithelial ovarian cancer. Cancer Treatment Reviews. 2006;32(3): 180-190. doi: 10.1016 / j.ctrv.2006.01.001.

[0291] 59. Chen J-S, Lan K, Hung M-C. Strategies to target HER2 / neu overexpression for cancer therapy. Drug Resist Updat. 2003;6(3): 129-136. doi: 10.1016 / sl368-7646(03)00040-2. 60. Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med. 2001;344(l l):783-792. doi: 10.1056 / NEJM200103153441101.

[0292] 61. Casi G, Neri D. Antibody-drug conjugates: basic concepts, examples and future perspectives. J Control Release. 2012;161(2):422-428. doi: 10.1016 / j.jconrel.2012.01.026.

[0293] 62. Shastry M, Gupta A, Chandarlapaty S, Young M, Powles T, Hamilton E. Rise of Antibody- Drug Conjugates: The Present and Future. Am Soc Clin Oncol Educ Book. 2023;43:e390094. doi : 10.1200 / EDBK 390094.

[0294] 63. Grairi M, Le Borgne M. Antibody-drug conjugates: prospects for the next generation. Drug Discov Today. 2024;29(12): 104241. doi: 10.1016 / j.drudis.2024.104241.

[0295] 64. Mckertish CM, Kayser V. Advances and Limitations of Antibody Drug Conjugates for Cancer. Biomedicines. 2021;9(8). doi: 10.3390 / biomedicines9080872.

[0296] 65. Bagshawe KD. Antibody-directed enzyme prodrug therapy (ADEPT) for cancer. Expert Rev Anticancer Ther. 2006;6(10): 1421-1431. doi: 10.1586 / 14737140.6.10.1421.

[0297] 66. Thuenemann EC, Le DHT, Lomonossoff GP, Steinmetz NF. Bluetongue Virus Particles as Nanoreactors for Enzyme Delivery and Cancer Therapy. Mol Pharm. 2021 ; 18(3): 1150-1156. doi: 10.1021 / acs.molpharmaceut.OcO 1053.

[0298] 67. Qin M, Du G, Sun X. Biomimetic cell-derived nanocarriers for modulating immune responses. Biomater Sci. 2020;8(2):530-543. doi:10.1039 / c9bm01444f.

[0299] 68. Moon H, Lee J, Min J, Kang S. Developing genetically engineered encapsulin protein cage nanoparticles as a targeted delivery nanoplatform. Biomacromolecules. 2014;15(10):3794- 3801. doi: 10.1021 / bm501066m.

[0300] 69. van de Steen A, Wilkinson HC, Dalby PA, Frank S. Encapsulation of Transketolase into In Vitro-Assembled Protein Nanocompartments Improves Thermal Stability. ACS Appl Bio Mater. 2024;7(6):3660-3674. doi: 10.1021 / acsabm.3c01153.

[0301] 70. Rennie C, Sives C, Boyton I, et al. In Vivo Behavior of Systemically Administered Encapsulin Protein Nanocages and Implications for their use in Targeted Drug Delivery. Advanced Therapeutics. 2024;7(2). doi: 10.1002 / adtp.202300360.

[0302] 71. Mendes M, Mahita J, Blazeska N, et al. IEDB-3D 2.0: Structural data analysis within the Immune Epitope Database. Protein Sci. 2023;32(4):e4605. doi: 10.1002 / pro.4605.

[0303] 72. Williams EM, Jung SM, Coffman JL, Lutz S. Pore Engineering for Enhanced Mass Transport in Encapsulin Nanocompartments. ACS Synth Biol. 2018;7(l 1):2514-2517. doi: 10.1021 / acssynbio.8b00295. 73. Adamson LSR, Tasneem N, Andreas MP, et al. Pore structure controls stability and molecular flux in engineered protein cages. Sci Adv. 2022;8(5):eabl7346. doi: 10.1126 / sciadv.abl7346.

[0304] 74. Bae Y, Kim GJ, Kim H, Park SG, Jung HS, Kang S. Engineering Tunable Dual Functional Protein Cage Nanoparticles Using Bacterial Superglue. Biomacromolecules. 2018;19(7):2896- 2904. doi: 10.1021 / acs.biomac.8b00457.

[0305] 75. Michel-Souzy S, Hamelmann NM, Zarzuela-Pura S, Paulusse JMJ, Cornelissen JJLM. Introduction of Surface Loops as a Tool for Encapsulin Functionalization. Biomacromolecules. 2021;22(12):5234-5242. doi: 10.1021 / acs.biomac. lc01156.

[0306] 76. Moon H, Lee J, Kim H, Heo S, Min J, Kang S. Genetically engineering encapsulin protein cage nanoparticle as a SCC-7 cell targeting optical nanoprobe. Biomater Res. 2014;18:21. doi:10.1186 / 2055-7124-18-21.

[0307] 77. Lagoutte P, Mignon C, Stadthagen G, et al. Simultaneous surface display and cargo loading of encapsulin nanocompartments and their use for rational vaccine design. Vaccine. 2018;36(25):3622-3628. doi: 10.1016 / j.vaccine.2018.05.034.

[0308] 78. Nygren P-A. Alternative binding proteins: affibody binding proteins developed from a small three-helix bundle scaffold. FEBS J. 2008;275(l l):2668-2676. doi: 10.1111 / j .1742- 4658.2008.06438.x.

[0309] 79. Altai M, Liu H, Ding H, et al. Affibody-derived drug conjugates: Potent cytotoxic molecules for treatment of HER2 over-expressing tumors. J Control Release. 2018;288:84-95. doi:10.1016 / j jconrel.2018.08.040.

[0310] 80. Ding H, Xu T, Zhang J, et al. Affibody-Derived Drug Conjugates Targeting HER2: Effect of Drug Load on Cytotoxicity and Biodistribution. Pharmaceutics. 2021;13(3). doi : 10.3390 / pharmaceuticsl 3030430.

[0311] 81. Li S, Jin Y, Su Y, et al. Anti-HER2 Affibody-Conjugated Photosensitizer for Tumor Targeting Photodynamic Therapy. Mol Pharm. 2020;17(5): 1546-1557. doi: 10.1021 / acs.molpharmaceut.9b01247.

[0312] 82. Sorensen J, Velikyan I, Sandberg D, et al. Measuring HER2-Receptor Expression In Metastatic Breast Cancer Using 68GaABY-025 Affibody PET / CT. Theranostics. 2016;6(2):262-271. doi: 10.7150 / thno,13502.

[0313] 83. Hu X, Li D, Fu Y, et al. Advances in the Application of Radionuclide-Labeled HER2 Affibody for the Diagnosis and Treatment of Ovarian Cancer. Front Oncol. 2022; 12:917439. doi: 10.3389 / fonc.2022.917439. 84. Miao H, Sun Y, Jin Y, Hu X, Song S, Zhang J. Application of a Novel 68Ga-HER2 Affibody PET / CT Imaging in Breast Cancer Patients. Front Oncol. 2022; 12:894767. doi: 10.3389 / fonc.2022.894767.

[0314] 85. Lan K-H, Tsai C-L, Chen Y-Y, et al. Affibody-conjugated 5 -fluorouracil prodrug system preferentially targets and inhibits HER2-expressing cancer cells. Biochem Biophys Res Commun. 2021;582: 137-143. doi: 10.1016 / j.bbrc.2021.09.078.

[0315] 86. La Rodriguez de Fuente L, Cancela IG, Estevez-Salguero AM, Iglesias P, Costoya JA. Development of a biosensor based on a new marine luciferase fused to an affibody to assess Her2 expression in living cells. Anal Chim Acta. 2022; 1221 :340084. doi: 10.1016 / j .aca.2022.340084.

[0316] 87. Zhao Z, Tao X, Xie Y, et al. In Situ Prodrug Activation by an Affibody-Ruthenium Catalyst Hybrid for HER2-Targeted Chemotherapy. Angew Chem Int Ed Engl. 2022;61(26):e202202855. doi: 10.1002 / anie.202202855.

[0317] 88. Smith B, Lyakhov I, Loomis K, et al. Hyperthermia-triggered intracellular delivery of anticancer agent to HER2(+) cells by HER2-specific affibody (ZHER2-GS-Cys)-conjugated thermosensitive liposomes (HER2(+) affisomes). J Control Release. 2011;153(2):187-194. doi: 10.1016 / j .jconrel.2011.04.005.

[0318] 89. Shipunova VO, Sogomonyan AS, Zelepukin IV, Nikitin MP, Deyev SM. PLGA Nanoparticles Decorated with Anti-HER2 Affibody for Targeted Delivery and Photoinduced Cell Death. Molecules. 2021;26(13). doi: 10.3390 / molecules26133955.

[0319] 90. Zhang Y, Jiang S, Zhang D, Bai X, Hecht SM, Chen S. DNA-affibody nanoparticles for inhibiting breast cancer cells overexpressing HER2. Chem Commun (Camb). 2017;53(3):573- 576. doi: 10.1039 / c6cc08495h.

[0320] 91. Zmyslia M, Capper MJ, Grimmeisen M, et al. A nanoengineered tandem nitroreductase: designing a robust prodrug-activating nanoreactor. RSC Chem Biol. 2024. doi: 10.1039 / d4cb00127c.

[0321] 92. Altenburg WJ, Rollins N, Silver PA, Giessen TW. Exploring targeting peptide-shell interactions in encapsulin nanocompartments. Sci Rep. 2021;l 1(1):4951. doi: 10.1038 / s41598- 021-84329-z.

[0322] 93. Lohner P, Zmyslia M, Thurn J, et al. Inside a Shell — Organometallic Catalysis Inside Encapsulin Nanoreactors. Angewandte Chemie. 2021;133(44):24028-24034. doi: 10.1002 / ange.202110327. 94. Putri RM, Allende-Ballestero C, Luque D, et al. Structural Characterization of Native and Modified Encapsulins as Nanoplatforms for in Vitro Catalysis and Cellular Uptake. ACS Nano. 2017; 11(12): 12796-12804. doi: 10.1021 / acsnano.7b07669.

[0323] 95. Wikman M, Steffen A-C, Gunneriusson E, et al. Selection and characterization of HER2 / neu-binding affibody ligands. Protein Eng Des Sei. 2004;17(5):455-462. doi : 10.1093 / protein / gzh053.

[0324] 96. Orlova A, Magnusson M, Eriksson TLJ, et al. Tumor imaging using a picomolar affinity HER2 binding affibody molecule. Cancer Res. 2006;66(8):4339-4348. doi: 10.1158 / 0008- 5472.CAN-05-3521.

[0325] 97. Liu H, SEIJSING J, FREJD FY, Tolmachev V, GRASLUND T. Target-specific cytotoxic effects on HER2-expressing cells by the tripartite fusion toxin ZHER2:2891-ABD-PE38X8, including a targeting affibody molecule and a half-life extension domain. International Journal of Oncology. 2015;47(2):601-609. doi: 10.3892 / ijo.2015.3027.

[0326] 98. Dong D, Xia G, Li Z, Li Z. Human Serum Albumin and HER2 -Binding Affibody Fusion Proteins for Targeted Delivery of Fatty Acid-Modified Molecules and Therapy. Mol Pharm. 2016;13(10):3370-3380. doi: 10.1021 / acs.molpharmaceut.6b00265.

[0327] 99. ERBB2 DepMap Gene Summary. https: / / depmap.org / portal / gene / ERBB2?tab=characterization&characterization=expression. Updated January 28, 2025. Accessed January 28, 2025.

[0328] 100. Schmitt JF, Susil BJ, Hearn MT. Aberrant FGF-2, FGF-3, FGF-4 and C-erb-B2 gene copy number in human ovarian, breast and endometrial tumours. Growth Factors. 1996; 13(1-2): 19- 35. doi: 10.3109 / 08977199609034564.

[0329] 101. Giessen TW. The Structural Diversity of Encapsulin Protein Shells. Chembiochem. 2024;25(24):e202400535. doi: 10.1002 / cbic.202400535.

[0330] 102. Ross J, McIver Z, Lambert T, et al. Pore dynamics and asymmetric cargo loading in an encapsulin nanocompartment. Sci Adv. 2022;8(4):eabj4461. doi: 10.1126 / sciadv.abj4461.

[0331] 103. Kwon S, Andreas MP, Giessen TW. Pore Engineering as a General Strategy to Improve Protein-Based Enzyme Nanoreactor Performance. ACS Nano. 2024;18(37):25740-25753. doi: 10.1021 / acsnano.4c08186.

Claims

Claims1. A nanoparticle comprising an assembly of bacterial encapsulin protein monomers, and an internal protein located inside the nanoparticle, wherein the internal protein is a dimer of an enzyme and / or a tandem-enzyme, wherein each enzyme of the dimer or tandem is fused to at least one terminal targeting peptide (TP).

2. The nanoparticle according to claim 1, wherein the bacterial encapsulin protein is selected from a type I encapsulin, preferably forming a T1 symmetric particle.

3. The nanoparticle according to claim 1 or 2, wherein at least one marker or label sequence is fused to the N- and / or C-terminus of the encapsulin protein monomer, preferably selected from a His-tag, a Strep-tag, and an MBP affinity tag.

4. The nanoparticle according to any one of claims 1 to 3, wherein the encapsulin protein is selected from an encapsulin protein of Mycobacterium smegmatis, Thermotoga marilima. Myxococcus xanlhus. Quasibacillus thermotolerans, Haliangium ochraceum. Acinetobacter baumannii. Mycobacterium leprae, Mycobacterium avium, Synechococcus elongatus, Burkholderia cepacia, and Klebsiella pneumoniae.

5. The nanoparticle according to any one of claims 1 to 4, wherein the encapsulin protein comprises a mutated pore-forming loop increasing the diffusion rate through the pores in the nanoparticle, for example by introducing amino acid substitutions and / or deletions, wherein preferably the encapsulin protein is selected from M. smegmatis comprising a mutated poreforming amino acid sequence VSETTAHGYPIRE (SEQ ID NO: 2) wherein the mutated poreforming amino acid sequence is selected from the amino acid sequences VSETTAAAIRE (SEQ ID NO: 3), VSETTAAIRE (SEQ ID NO: 4), VSETAAAIRE (SEQ ID NO: 5), and VSETAAIRE (SEQ ID NO: 6).

6. The nanoparticle according to any one of claims 1 to 5, wherein the dimer or tandem-enzyme comprises enzymes selected from the group consisting of cytosine-desaminases, nitroreductases, and enzymes used in laundry detergents, such as amylases, lipases, orcellulases, wherein the enzymes in particular selected from wildtype and recombinant bacterial enzymes, such as, for example, the NfsB of E. coli.

7. The nanoparticle according to any one of claims 1 to 6, wherein the protein concentration inside the nanoparticle is higher than 2 mM, preferably higher than 3 mM, such as between 3 and 3.5 mM.

8. An at least bicistronic vector comprising a first nucleic acid sequence encoding an encapsulin protein as defined in any one of claims 1 to 7, and a second nucleic acid sequence encoding at least one enzyme of the dimer or the tandem-enzyme as defined in any one of claims 1 to 7.

9. A host cell comprising the at least bicistronic vector according to claim 8.

10. A method for producing the nanoparticle according to any one of claims 1 to 7, comprising recombinantly expressing the encapsulin and the enzyme of the dimer or the tandem-enzyme in the cell according to claim 9, and purifying the resulting nanoparticles or capsids comprising the incorporated dimer or tandem-enzyme.

11. A method for converting a substrate in vitro or in vivo, comprising providing the nanoparticle according to any one of claims 1 to 7 comprising a dimer or tandem-enzyme, contacting the nanoparticle with at least one suitable substrate to be converted, and converting said substrate, wherein preferably the non-reduced substrate is a prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug, or the substrate is a nitroaromatic environmental pollutant or the substrate is a selectable marker in genetic engineering or a starch, fat, or cellulose.

12. The method according to claim 11, wherein the substrate is a nitroaromatic compound to be reduced in vitro or in vivo, comprising providing the nanoparticle according to any one of claims 1 to 7 comprising a dimer or tandem-enzyme of nitroreductases, contacting the nanoparticle with at least one nitroaromatic substrate suitable for reduction in vitro or in vivo, and reducing said nitroaromatic substrate.

13. The method according to claim 11 or 12, wherein the method provides a substrate conversion that is similar or higher than a conversion of the respective non-encapsulated enzyme, in particular the dimer or tandem-enzyme thereof.

14. A pharmaceutical composition, comprising the nanoparticle according to any one of claims 1 to 7, and at least one suitable pharmaceutically acceptable carrier.

15. The pharmaceutical composition according to claim 14 for use in the treatment of diseases, such as for the use of treating cancer, comprising administering the pharmaceutical composition in combination with at least one anti-cancer prodrug, such as, for example, Nbzp or 5-FC that can be enzymatically converted into the pharmaceutically active drug, wherein preferably the cancer is lung carcinoma, cervical cancer, breast cancer or ovarian cancer.

16. Use of the nanoparticle according to any one of claims 1 to 7 or the pharmaceutical composition according to claim 14 for detoxifying environmental nitroaromatic pollutants, for cleaning of starches, fats, or cellulose, the biosynthesis of antibiotics, selectable marker in genetic engineering, as a biosensor for pollutant detection or a biocatalyst for the organic synthesis of pharmaceutical intermediates or fine chemicals.

Images