Engineered enzyme, fusion protein comprising engineered enzyme, and method for enhancing anticancer activity

Abstract

The present disclosure provides an engineered enzyme comprising an amino acid sequence that is at least 80% identical to the amino acid sequence of a human beta-glucuronidase (hBG). Also provides a fusion protein comprising a polypeptide and the engineered enzyme, and a method of enhancing an anticancer activity of an anticancer drug in a subject.

Description

Attorney Docket No.: 5025-0446PWO1 ENGINEERED ENZYME, FUSION PROTEIN COMPRISING ENGINEERED ENZYME, AND METHOD FOR ENHANCING ANTICANCER ACTIVITY FIELD OF INVENTION

[0001] The present disclosure relates to an engineered enzyme, and a fusion protein comprising the engineered enzyme. The present disclosure further relates a method of enhancing an anticancer activity, comprising administering the engineered enzyme or the fusion protein. CROSS REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to U.S. Patent Application No. 63 / 558,820, filed February 28, 2024, which is entirely and expressly incorporated by reference into the present application. REFERENCE TO ELECTRONIC SEQUENCE LISTING

[0003] The application contains a Sequence Listing which has been submitted electronically in .XML format and is hereby incorporated by reference in its entirety. Said .XML copy, created on February 25, 2024, is named “25P0066-PCT-EnSeqlist-F” and is 7,278 bytes in size. The sequence listing contained in this .XML file is part of the specification and is hereby incorporated by reference herein in its entirety. BACKGROUND OF THE INVENTION

[0004] A wide variety of diseases can be treated with enzymes, including enzyme replacement therapies to replenish missing enzymes important for normal human metabolism, fibrinolytic enzymes to treat cardiovascular disorders, and enzymes designed to treat cancer by depleting key metabolites such as asparagine and arginine, depleting kynurenine to enhance cancer immunogenicity, or by targeting enzymes to cancer cells to selectively covert nontoxic prodrugs to cytotoxic agents in the tumor microenvironment [1-6]. Enzyme therapies, however, can induce antibody responses that may affect drug efficacy and safety. Although most enzyme replacement therapies use recombinant human enzymes, the lack of endogenous enzymes in some patients results in loss of tolerance and development of antibody responses ranging from 15% for Gaucher disease to up to 100% for Pompe disease [7]. Amino-acid depleting enzymes are often derived from microbial sources due to lack of suitable human analogs and immunoenzymes investigated for prodrug activation in tumors also rely on enzymes derived from microbial sources due to their inherently high activity and the possibility of using enzymes without human homologs to increase reaction specificity [8]. Induction of antibodies against therapeutic enzymes can lead to immune complex deposition, proteinuria and potentialAttorney Docket No.: 5025-0446PWO1 renal damage, induce hypersensitivity or anaphylactic reactions, and cause loss of therapeutic efficacy [7, 9-12].

[0005] A potentially powerful approach to decrease immunogenicity associated with enzyme therapies is to engineer human enzymes to exhibit improved or new properties. Changing enzymatic specificity of a normal endogenous human enzyme to compensate for a missing enzyme may help reduce treatment immunogenicity during enzyme replacement therapy

[0013] . Engineered human enzymes may also reduce immunogenicity of cancer depletion and prodrug activation therapies [14-18]. On the other hand, introduction of new amino acids in human proteins can adversely affect protein stability

[0019] and may increase immunogenicity

[0020] . It therefor remains unclear if human enzymes can be engineered to retain good in vivo stability and low immunogenicity.

[0006] We previously used directed-molecular evolution to change the substrate specificity of human beta-glucuronidase (hBG) to alpha-iduronidase for treatment of MPS I as well as increase the activity of hBG for activation of a glucuronide prodrugs [13, 21, 22]. However, the implications of introducing substantial numbers of amino acid substitutions toward enzyme stability and immunogenicity were not investigated. SUMMARY OF THE INVENTION

[0007] In the present disclosure, we engineered hBG to enhance the anticancer activity of the clinically-used anticancer drug CPT-11 (irinotecan) while maintaining protein structural stability and low immunogenicity. CPT-11 is a prodrug that is converted by carboxylases in the blood and liver to the active topoisomerase I poison SN-38 (Fig. 1)

[0023] . SN-38 is further metabolized by UDP-glucuronosyltransferase (UDPGT) in the liver to the glucuronide conjugate SN-38G, which is considered to be an inactive byproduct. SN-38G is a major metabolite in the serum of cancer patients receiving CPT-11, exceeding the concentration of SN-38 by about seven-fold

[0024] . hBG is attractive for cancer prodrug activation because it sequestered in the lysosome of cells, preventing systematic activation of hydrophilic glucuronide prodrugs in the circulation. We used ECSTASY (enzyme cleavable surface tethered all-purpose screening system) to select hBG variants that effectively activate the glucuronide prodrugs SN-38G at neutral pH

[0021] . We show that hBG variants with three, ten or nineteen amino acid substitutions (hBG3, hBG10 and hBG19) display better stability in serum and less immunogenicity in hBG transgenic mice as compared with E. coli beta- glucuronidase (eBG). We further demonstrate that an immunoenzyme (hcc49-hBG10) composed of a tumor-targeting single-chain antibody fused to hBG10 significantly increasesAttorney Docket No.: 5025-0446PWO1 the anticancer activity of CPT-11 in a mouse model, indicating that evolution of human enzymes is a promising approach to develop therapeutic enzymes to treat human patients.

[0008] To achieve the above propose, one aspect of the present disclosure provided herein is an engineered enzyme, including an amino acid sequence that is at least 80%, preferably at least 85%, at least 90%, at least 95%, at least 97% and at least 99%, identical to the amino acid sequence of a human beta-glucuronidase (hBG).

[0009] Preferably, the human beta-glucuronidase includes the amino acid sequence of SEQ ID NO: 1.

[0010] Preferably, the engineered enzyme includes a substitution at a residue that corresponds to residue G340, C396, S422, L423, E431, L456, Y461, M465, H469, S476, N484, N486, A488, Y495, S506, H509, I517, Q518, L519, Q520, T545, M593, E595, and / or T599 in the sequences of SEQ ID NO: 1.

[0011] Preferably, the engineered enzyme, wherein residue 340 is G or N; residue 396 is C or A; residue 422 is S or T; residue 423 is L or M; residue 431 is E or K; residue 456 is E or L; residue 461 is D; residue 465 is M or T; residue 469 is H or V; residue 476 is T; residue 484 is N or F; residue 486 is T or N ; residue 488 is H or D; residue 495 is Y or F; residue 506 is S or G; residue 509 is H or Y; residue 517 is I or V; residue 518 is R or E; residue 519 is L or G; residue 520 is Q or E; residue 545 is S or G; residue 593 is M or A; residue 595 is E or D; and residue 599 is T or F.

[0012] Preferably, the engineered enzyme includes a substitution at L456, Y461, S476, N486, A488, Q518, T545, M593, E595, and / or T599.

[0013] Preferably, the engineered enzyme includes a substitution at G340, C396, S422, L423, E431, Y461, M465, H469, S476, N484, A488, Y495, S506, H509, I517, Q518, L519, Q520, and / or T545.

[0014] Preferably, the engineered enzyme includes residues E456, D461, T476, T486, H488, R518, S545, A593, D595 and / or F599.

[0015] Preferably, the engineered enzyme includes residues N340, A396, T422, M423, K431, D461, T465, V469, T476, F484, F488, F495, G506, Y509, V517, E518, G519, E520, and / or G545.

[0016] In another aspect of the present disclosure, provided herein is a fusion protein, including: (1) a polypeptide, wherein the polypeptide can accumulate in the tumor; and (2) the aforementioned engineered enzyme.

[0017] Preferably, the polypeptide includes: an antibody or a fragment thereof, a binding protein, and / or an enzyme.Attorney Docket No.: 5025-0446PWO1

[0018] Preferably, the binding protein binds an extracellular matrix of a tumor; wherein the extracellular matrix includes a collagen.

[0019] Preferably, the fusion protein includes: (1) the antibody is an antibody fragment, optionally selected from the group consisting of F(ab'2), Fab', Fab, Fv, single domain antibody, and single-chain variable fragment (scFv); and (2) the antibody is targeting to an antigen of a tumor.

[0020] Preferably, the antibody includes a scFv of hcc49.

[0021] Preferably, the antigen includes a glycoprotein.

[0022] Preferably, the glycoprotein includes Tumor-Associated Glycoprotein-72 (TAG-72).

[0023] Preferably, the tumor includes: breast cancerous tumor, lung cancerous tumor, liver cancerous tumor, pancreatic cancerous tumor, prostate cancerous tumor, skin cancerous tumor (melanoma), kidney cancerous tumor, bladder cancerous cancer, blood cancerous tumor (leukemia), colorectal cancerous tumor, and / or lymphoma.

[0024] In another aspect of the present disclosure, provided herein is a pharmaceutical composition including the any one of aforementioned engineered enzyme and a pharmaceutically acceptable carrier.

[0025] In another aspect of the present disclosure, provided herein is a pharmaceutical composition including the any one of aforementioned fusion protein and a pharmaceutically acceptable carrier.

[0026] In another aspect of the present disclosure, provided herein is a method of enhancing an anticancer activity of an anticancer drug in a subject, including administering to the subject in need thereof the aforementioned engineered enzyme or the aforementioned fusion protein or the aforementioned pharmaceutical composition.

[0027] Preferably, the method further including administering to the subject in need thereof the anticancer drug.

[0028] Preferably, the anticancer drug is metabolized to a glucuronidated metabolite in the subject.

[0029] Preferably, the anticancer drug is a glucuronided form upon administration

[0030] Preferably, the anticancer drug includes: Irinotecan (CPT-11), p-hydroxyaniline mustard glucuronide ester (HAMG-ester), Daunorubicin, Doxorubicin, Paclitaxel, Cisplatin, Docetaxel, Methotrexate, Temozolomide, and / or Gemcitabine, Epirubicin, histone deacetylases inhibitors (CI-994), Etoposide, Cyclopamine, Camptothecin, Duocarmycin, Monomethylauristatin E, p-Hydroxyaniline mustard, and / or its derivatives thereof.Attorney Docket No.: 5025-0446PWO1

[0031] Preferably, an immunogenicity of the engineered enzyme is lower than that of beta- glucuronidase derived from the heterologous sources. Preferably, the engineered enzyme is lower than that of beta-glucuronidase derived from the microbial sources.. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Fig 1. Therapeutic strategy. (A) CPT-11 is converted by carboxyl esterases (CE) to the active anticancer topoisomerase I inhibitor SN-38, which can be further metabolized by UDP glucuronosyltransferases to the inactive metabolite SN-38G. (B) The concentration of SN-38G exceeds SN-38 in human serum by a factor of 3-15. (C) Immunoenzymes are designed to selectively reconvert SN-38G back to active SN-38 in the tumor microenvironment.

[0033] Fig 2. Schematic process for selection of hBG variants with high activity at neutral pH. (A) Mutation sites in hBG were selected by comparison of mammalian and bacterial BG sequences. Five individual hBG libraries were constructed covering amino acid substitutions at 49 amino acid positions. hBG variants in individual libraries that displayed enhanced activity were reassembled together to form a new “combined” library. The best hits in the combined library were then combined by DNA shuffling. (B) The screening process during each step entailed cloning each library into a mammalian surface display vector, generating recombinant retroviral particles, transduction of single hBG variants genes into 3T3 fibroblast cells, and selection of cells that displayed high levels of membrane-tethered hBG by fluorescence- activated cell sorting of individual cells into 96-well culture plates. Soluble hBG variants, released into the cultured medium after treating cells with PI-PLC to cleave the GPI anchor, were assayed for hBG amount and enzyme activity for hydrolysis of SN-38G.

[0034] Fig 3. Enzymatic activities of hBG variants. (A) The relative enzyme activities for hydrolysis of 1 μM SN-38G by hBG variants compared to hBG at neutral pH are shown. (B) Km, (C) Kcat and (D) Kcat / Km of hBG enzymes against SN-38G at the indicated pH values are shown.

[0035] Fig 4. Docking of SN-38G in hBG enzymes. SN-38G was docked in enzymes using CB-DOCK2. The illustrations show BG (grey), SN-38G (red), active site glutamic acids (yellow), interacting amino acids (light green), and non-interacting amino acid substitutions (blue). Amino acid substitution present in hBG10 and hBG19 that interact with SN-38G are shown in orange.

[0036] Fig 5. Comparison of cancer cell killing by SN-38G and hBG enzymes. Serial dilutions of the indicated enzymes were added to MCF-7 breast cancer (A), LS174T colonAttorney Docket No.: 5025-0446PWO1 cancer (B) or HT-29 colon cancer (C) cells in the presence of non-toxic concentrations of SN- 38G (5, 5, and 20 nM, respectively). Cell proliferation was measured after 48 h. (D) Fold decrease in the amount of enzyme required to produce 50% inhibition of cancer cell growth in the presence of SN-38G as compared to hBG. Significant differences between mean values of hBG and hBG variants was expressed as: *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.

[0037] Fig 6. Stability and in vivo half-life of hBG enzymes. (A) Thermal stability of enzymes as measured by differential scanning calorimetry. (B, C) Recombinant enzymes (100 µg mL-1) were incubated at 37oC in (B) PBS or (C) human serum for 9 days. Results show the percent enzyme activity on days 3, 6 and 9 as compared to day 0. eBG is E. coli beta- glucuronidase produced in E. coli. eBG-NS is E. coli beta-glucuronidase with a single amino acid substitution to remove a cryptic glycosylation site, produced in 3T3 fibroblasts. (D) hBG enzymes were radiolabeled with 125-iodine before i.v. injection into mice. (E) C57BL / 6 mice were intravenously injected with a single dose of 100 μg I125-spiked enzymes. (F) BALB / c mice were intravenously injected with a single dose of 10 μg I125-spiked enzymes. Bars, SD (n = 3).

[0038] Fig 7. Enzyme immunogenicity in mice. (A) Groups of mice were intravenously injected with 100 μg of recombinant enzymes once per week. The mean concentration of anti- enzyme IgG in serum samples at the indicated weeks are shown for (B) C57BL6 transgenic- hBG mice, (C) BALB / c transgenic-hBG mice and (D) wild type BALB / c mice. Significant differences of anti-enzyme IgG levels between hBG and other enzymes are indicated: *, p ≤ 0.05; **, p ≤ 0.01, ***, p ≤ 0.001.

[0039] Fig 8. Enzymatic activity and glycosylation affect hBG10 immunogenicity in hBG transgenic mice. (A) Single amino acid substitutions were introduced into hBG10 to knockout enzymatic activity (hBG10xA) or remove the cryptic N-linked glycosylation site (hBG10xG). (B) BALB / c transgenic hBG mice were intravenously injected with 100 μg of recombinant enzymes once per week. The mean concentrations of anti-enzyme IgG in mice at the indicated weeks is shown. Bars, SD. Significant differences of anti-enzyme IgG levels between hBG10 and other enzymes are indicated: *, p ≤ 0.05.

[0040] Fig 9. TAG-72 targeted immunoenzymes. (A) Immunoenzyme genes include a immunoglobulin kappa chain signal peptide, the hcc49 or control DNS single-chain antibody, a flexible peptide linker and one of the hBG enzymes. Note that hBG enzymes are tetramers so four scFv are present in one immunoenzyme. (B) SDS PAGE of purified enzymes and corresponding immunoenzymes. (C) Direct ELISA of immunoenzyme binding to BSAAttorney Docket No.: 5025-0446PWO1 (negative control) or bovine submaxillary gland mucin, which contains the TAG-72 epitope. (D) Binding of immunoenzymes to HT-29, LS174T and MCF-7 cancer cells was determined by FACs.

[0041] Fig 10. In vitro immunoenzyme activity. (A) Immunoenzyme (1 µg well-1) were added to HT-29, LS174T or MCF-7 cancer cells for 1 h before the cells were washed and serial dilutions of SN-38G were added for 48 h. Results show the fold decrease in SN-38G required to inhibit cell DNA synthesis by 50% in the presence of immunoenzyme compared to DNS- hBG immunoenzyme. Significant differences between hcc49-hBG and other groups are indicated: *, p ≤ 0.05. (B) Serial dilutions of immunoenzyme were added to LS174T cancer cells for 1 h before the cells were washed and 5 nM SN-38G was added for 48 h. (C) Immunofluorescence staining of LS174T and MCF-7 tumor sections.

[0042] Fig 11. Tumor accumulation of hcc49-hBG10 in LS174T tumors. (A) Radiolabeled hcc49-hBG (10, 100 or 1000 µg) was i.v. injected to NOD-SCID mice or NOD-SCID mice bearing established LS174T xenografts. Results show hcc49-hBG concentrations and estimated half-lives. (B) FDGlcU can be converted to the fluorescent compound fluorescein by beta-glucuronidase. (C) NOD-SCID mice bearing LS174T tumors were i.v. injected with PBS or 200 µg DNS-hBG10 or hcc49-hBG10. After 96 h, mice were i.v. injected with FDGlcU and IVIS imaging was performed 30 min later in recovered tumors, lungs and livers. (D) Fluorescence signals were quantified (left) and compared to signals in mice treated with PBS and FDGlcU (right).

[0043] Fig 12. hcc49-hBG10 can enhance the anticancer activity of CPT-11. (A) Groups of three NOD / SCID mice bearing established LS174T tumors were treated twice on days 0 and 10 by intravenous injections of vehicle or 200 µg immunoenzyme, followed with two i.v. injections of 25 mg kg-1SN-38G or two i.p. injections of 50 mg kg-1CPT-11 at 72 h and 96 h. (B) Results show mean tumor size and percent change in body weight (n=3). Bars, SE. (C) Groups of six NOD / SCID mice bearing established LS174T tumors were treated five times starting on days 0, 10, 20, 30, and 40 by i.v. injection of vehicle or 200 µg immunoenzyme, followed with two i.p. injections of 50 mg kg-1CPT-11 at 72 h and 96 h later. Results show mean tumor size (D), percent change in body weight (E), and survival (F). Significant differences with treatment with hcc49-hBG and CPT-11 are indicated: **, p ≤ 0.01; ***, p ≤ 0.005.

[0044] Fig 13. Amino acid changes in Library 1. The graph shows the predicted frequency of amino acid substitutions in library 1 before screening (black bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitutionAttorney Docket No.: 5025-0446PWO1 of the most prevalent amino acid (100% represents replacement of the original amino acid with a single new amino acid, while 60% indicates that 6 of 10 clones have the same amino acid substitution). The two clones selected from library 1 were not sequenced.

[0045] Fig 14. Amino acid changes in Library 2. The graph shows the predicted frequency of amino acid substitutions in library 2 before screening (black bars), the measured frequency of amino acid substitutions after screening library 2 (green bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitution of the most prevalent amino acid (100% represents replacement of the original amino acid with a single new amino acid, while 60% indicates that 6 of 10 clones have the same amino acid substitution).

[0046] Fig 15. Amino acid changes in Library 3. The graph shows the predicted frequency of amino acid substitutions in library 3 before screening (black bars), the measured frequency of amino acid substitutions after screening library 3 (green bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitution of the most prevalent amino acid (100% represents replacement of the original amino acid with a single new amino acid, while 60% indicates that 6 of 10 clones have the same amino acid substitution).

[0047] Fig 16. Amino acid changes in Library 4. The graph shows the predicted frequency of amino acid substitutions in library 4 before screening (black bars), the measured frequency of amino acid substitutions after screening library 4 (green bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitution of the most prevalent amino acid (100% represents replacement of the original amino acid with a single new amino acid, while 60% indicates that 6 of 10 clones have the same amino acid substitution).

[0048] Fig 17. Amino acid changes in Library 5. The graph shows the predicted frequency of amino acid substitutions in library 5 before screening (black bars), the measured frequency of amino acid substitutions after screening library 5 (green bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitution of the most prevalent amino acid (100% represents replacement of the original amino acid with a single new amino acid, while 60% indicates that 6 of 10 clones have the same amino acid substitution).

[0049] Fig 18. Predicted structures and mutation sites in hBG variants. The structures of hBG3, hBG10 and hNG19 were estimated by alphafold2. The active site glutamic acids E451 and E540 are indicated in yellow, amino acid substitutions are shown in blue, high mannoseAttorney Docket No.: 5025-0446PWO1 glycans ae illustrated in magenta. Note that there is a new glycan present near the catalytic pocket of hBG10.

[0050] Fig 19. SDS PAGE of BG proteins. hBG is wild type human-beta-glucuronidase, hBG3, hBG10, and hBG19 are hBG variants, mBG is mouse beta-glucuronidase, and eBG-NS is E. coli beta-glucuronidase with a single N to S amino substitution to remove a cryptic N- linked glycosylation site. All enzymes were produced in 3T3 fibroblasts.

[0051] Fig 20. Docking of SN-38G in catalytic pocket of hBG enzymes. Results show the frequency that SN-38G contacted individual amino acid sidechains from twenty docking simulations in each enzyme. Amino acid F484 and Y509 are only present in hBG19 whereas T486 is only present in hBG10.

[0052] Fig 21. Stability of beta-glucuronidase enzymes in mouse serum. Recombinant enzymes (100 µg mL-1) were incubated at 37oC in serum from (A) C57BL / 6 or (B) NOD / SCID mice for 9 days. Data show the percentage of enzyme activity compared to Day 0.

[0053] Fig 22. Enzyme immunogenicity in mice. (A) The standard curve of hBG and hBG variants. The anti-hBG detection antibody bound similarly to hBG, hBG3, hBG10, and hBG19 coated in ELISA plates. (B, C) The titers of anti-enzyme IgG in BALB / c-Tg-hBG mice (B) and wild type BALB / c mice (C). Mice were intravenously injected with 100 μg of recombinant enzymes once per week for 7-9 weeks. Serum of mice were collected once per week, and the concentration of anti-enzyme IgG in mice were analyzed by ELISA against each enzyme. Titers were defined as the serum dilution producing an absorbance signal of 1.0 in direct ELISA against each enzyme.

[0054] Fig 23. Enzyme activity of hBG10 mutants. (A) SDS PAGE of enzymes / hBG10xA has an amino substitution in the active site to decrease enzymatic activity while hBG10xG has a single amino acid substitution to remove the cryptic glycosylation site in hBG10. (B) Conversion of 4-MUG to 4-MU by beta-glucuronidase enzymes. Results show mean values of MU fluorescence.

[0055] Fig 24. Change of (A) tumor size and (B) body weight of LS174T-tumor bearing mice after treating 200 μg hcc49-hBG10 and different dose of HAMG-ester. DETAILED DESCRIPTION

[0056] The foregoing and other aspects of the present disclosure will now be described in more detail with respect to other embodiments described herein. It should be appreciated that the invention can be embodied in different forms and should not be construed as limited to theAttorney Docket No.: 5025-0446PWO1 embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete and fully convey the invention's scope to those skilled in the art.

[0057] The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting to the invention. As used in the description of the invention and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0058] As used herein, the terms "comprises," “comprising,” “includes,” “including,” “has,” “having,” “contains”, “containing,” or any other variation thereof, are intended to cover a non- exclusive inclusion, subject to any limitation explicitly indicated. For example, a composition, mixture, process or method that includes a list of elements is not necessarily limited to only those elements but may include other elements not expressly listed or inherent to such composition, mixture, process, or method.

[0059] As used herein, the term “about” indicates that a value includes, for example, the inherent variation of error for a measuring device, the method employed to determine the value, or the variation among the study subjects. Typically the term is meant to encompass approximately or less than 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19% or 20% variability depending on the situation.

[0060] The use of the term “or” in the claims is used to mean “and / or” unless explicitly indicated to refer to alternatives only or the alternatives are mutually exclusive. However, the disclosure supports a definition that refers to only alternatives and “and / or.”

[0061] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All publications, patent disclosure, patents, and other references cited herein are incorporated by reference in their entirety for the teachings relevant to the sentence and / or paragraph in which the reference is presented.

[0062] The following representative examples illustrate various features and embodiments of the disclosure, which are intended to be illustrative and not limiting. Those skilled in the art will readily appreciate that the specific examples are only illustrative of the invention as described more fully in the subsequent claims. Every embodiment and feature described in the present application should be understood to be interchangeable and combinable with every embodiment contained within.Attorney Docket No.: 5025-0446PWO1

[0063] The propose of the present disclosure is to provide an engineered enzyme, including an amino acid sequence that is at least 80%, preferably at least 85%, at least 90%, at least 95% and at least 99%, identical to the amino acid sequence of a human beta-glucuronidase.

[0064] In an embodiment, the human beta-glucuronidase includes the amino acid sequence of SEQ ID NO: 1.

[0065] The term “human beta-glucuronidase” refers to a member of the glucuronidase family, which catalyzes hydrolysis of complex carbohydrates (glycosaminoglycans) to release the β- D-glucuronic acid residues. The term “engineered enzyme” in the present disclosure refers to the human beta-glucuronidase with one or more substitutions of amino acid within its sequence. The substitution numberings for the engineered enzyme are based on the human beta- glucuronidase sequence (SEQ ID NO: 1). SEQ ID NO:1 LQGGMLYPQESPSRECKELDGLWSFRADFSDNRRRGFEEQWYRRPLWESGPTVDMPVPSSFN DISQDWRLRHFVGWVWYEREVILPERWTQDLRTRVVLRIGSAHSYAIVWVNGVDTLEHEGGY LPFEADISNLVQVGPLPSRLRITIAINNTLTPTTLPPGTIQYLTDTSKYPKGYFVQNTYFDF FNYAGLQRSVLLYTTPTTYIDDITVTTSVEQDSGLVNYQISVKGSNLFKLEVRLLDAENKVV ANGTGTQGQLKVPGVSLWWPYLMHERPAYLYSLEVQLTAQTSLGPVSDFYTLPVGIRTVAVT KSQFLINGKPFYFHGVNKHEDADIRGKGFDWPLLVKDFNLLRWLGANAFRTSHYPYAEEVMQ MCDRYGIVVIDECPGVGLALPQFFNNVSLHHHMQVMEEVVRRDKNHPAVVMWSVANEPASHL ESAGYYLKMVIAHTKSLDPSRPVTFVSNSNYAADKGAPYVDVICLNSYYSWYHDYGHLELIQ LQLATQFENWYKKYQKPIIQSEYGAETIAGFHQDPPLMFTEEYQKSLLEQYHLGLDQKRRKY VVGELIWNFADFMTEQSPTRVLGNKKGIFTRQRQPKSAAFLLRERYWKIANETRYPHSVAKS QCLENS

[0066] In an embodiment, the engineered enzyme includes a substitution at a residue that corresponds to residue G340, C396, S422, L423, E431, L456, Y461, M465, H469, S476, N484, N486, A488, Y495, S506, H509, I517, Q518, L519, Q520, T545, M593, E595, and / or T599 in the sequences of SEQ ID NO: 1.

[0067] In an embodiment, the engineered enzyme, wherein residue 340 is G or N; residue 396 is C or A; residue 422 is S or T; residue 423 is L or M; residue 431 is E or K; residue 456 is E or L; residue 461 is D; residue 465 is M or T; residue 469 is H or V; residue 476 is T; residue 484 is N or F; residue 486 is T or N ; residue 488 is H or D; residue 495 is Y or F; residue 506 is S or G; residue 509 is H or Y; residue 517 is I or V; residue 518 is R or E; residue 519 is L or G; residue 520 is Q or E; residue 545 is S or G; residue 593 is M or A; residue 595 is E or D; and residue 599 is T or F.Attorney Docket No.: 5025-0446PWO1

[0068] In an embodiment, the engineered enzyme includes a substitution at L456, Y461, S476, N486, A488, Q518, T545, M593, E595, and / or T599.

[0069] In an embodiment, the engineered enzyme includes a substitution at G340, C396, S422, L423, E431, Y461, M465, H469, S476, N484, A488, Y495, S506, H509, I517, Q518, L519, Q520, and / or T545.

[0070] The term "substitution" refers to the replacement of amino acid residues in the human beta-glucuronidase sequence with other amino acids, both naturally occurring and non-natural. In the present disclosure, the substitutions in the engineered enzyme indicate that the human beta-glucuronidase sequence includes at least one of the aforementioned replacements. It should be noted that all positions within the human beta-glucuronidase sequence are potential substitution sites in the present disclosure.

[0071] In an embodiment, the engineered enzyme includes residues E456, D461, T476, T486, H488, R518, S545, A593, D595 and / or F599. Preferably, the engineered enzyme includes all of the above residues is called hBG10 (SEQ ID NO: 2) in the present disclosure. SEQ ID NO: 2 LQGGMLYPQESPSRECKELDGLWSFRADFSDNRRRGFEEQWYRRPLWESGPTVDMPVPSSFN DISQDWRLRHFVGWVWYEREVILPERWTQDLRTRVVLRIGSAHSYAIVWVNGVDTLEHEGGY LPFEADISNLVQVGPLPSRLRITIAINNTLTPTTLPPGTIQYLTDTSKYPKGYFVQNTYFDF FNYAGLQRSVLLYTTPTTYIDDITVTTSVEQDSGLVNYQISVKGSNLFKLEVRLLDAENKVV ANGTGTQGQLKVPGVSLWWPYLMHERPAYLYSLEVQLTAQTSLGPVSDFYTLPVGIRTVAVT KSQFLINGKPFYFHGVNKHEDADIRGKGFDWPLLVKDFNLLRWLGANAFRTSHYPYAEEVMQ MCDRYGIVVIDECPGVGLALPQFFNNVSLHHHMQVMEEVVRRDKNHPAVVMWSVANEPASHE ESAGDYLKMVIAHTKSLDPTRPVTFVSNSTYHADKGAPYVDVICLNSYYSWYHDYGHLELIR LQLATQFENWYKKYQKPIIQSEYGAESIAGFHQDPPLMFTEEYQKSLLEQYHLGLDQKRRKY VVGELIWNFADFATDQSPFRVLGNKKGIFTRQRQPKSAAFLLRERYWKIANETRYPHSVAKS QCLENS

[0072] In an embodiment, the engineered enzyme includes residues N340, A396, T422, M423, K431, D461, T465, V469, T476, F484, F488, F495, G506, Y509, V517, E518, G519, E520, and / or G545. Preferably, the engineered enzyme includes all of the above residues is called hBG19 (SEQ ID NO: 3) in the present disclosure. SEQ ID NO: 3 LQGGMLYPQESPSRECKELDGLWSFRADFSDNRRRGFEEQWYRRPLWESGPTVDMPVPSSFN DISQDWRLRHFVGWVWYEREVILPERWTQDLRTRVVLRIGSAHSYAIVWVNGVDTLEHEGGY LPFEADISNLVQVGPLPSRLRITIAINNTLTPTTLPPGTIQYLTDTSKYPKGYFVQNTYFDFAttorney Docket No.: 5025-0446PWO1 FNYAGLQRSVLLYTTPTTYIDDITVTTSVEQDSGLVNYQISVKGSNLFKLEVRLLDAENKVV ANGTGTQGQLKVPGVSLWWPYLMHERPAYLYSLEVQLTAQTSLGPVSDFYTLPVGIRTVAVT KSQFLINNKPFYFHGVNKHEDADIRGKGFDWPLLVKDFNLLRWLGANAFRTSHYPYAEEVMQ MADRYGIVVIDECPGVGLALPQFFNNVTMHHHMQVMKEVVRRDKNHPAVVMWSVANEPASHL ESAGDYLKTVIAVTKSLDPTRPVTFVSFSNYDADKGAPFVDVICLNSYYGWYYDYGHLELVE GELATQFENWYKKYQKPIIQSEYGAEGIAGFHQDPPLMFTEEYQKSLLEQYHLGLDQKRRKY VVGELIWNFADFMTEQSPTRVLGNKKGIFTRQRQPKSAAFLLRERYWKIANETRYPHSVAKS QCLENS

[0073] The present disclosure also provides a fusion protein, including: (1) a polypeptide, wherein the polypeptide can accumulate in the tumor; and (2) the aforementioned engineered enzyme.

[0074] In an embodiment, the polypeptide includes: an antibody or a fragment thereof, a binding protein, and / or an enzyme. In an embodiment, the fusion protein includes: (1) the antibody is an antibody fragment, optionally selected from the group consisting of F(ab'2), Fab', Fab, Fv, single domain antibody, and single-chain variable fragment (scFv); and (2) the antibody is targeting to an antigen of a tumor.

[0075] The term “antibody” refers a molecule including one or more polypeptide chains that has a specific affinity for, or can uniquely interact with, a specific antigen. Exemplary antibodies of the present disclosure include: monoclonal antibodies, which are identical antibodies that target a single antigen epitope; polyclonal antibodies, which are derived from different immune cells and recognize multiple epitopes; chimeric antibodies, which combine portions from different species; humanized antibodies, which are antibodies from non-human species whose protein sequences have been modified to increase their similarity to human antibody; antibody fusions, which are antibodies linked to other proteins; multispecific antibodies, which can target two or more different antigens (e.g, bispecific antibodies); monovalent antibodies, which have a single antigen-binding site and only can bind one antigen; multivalent antibodies, which can bind to multiple antigens simultaneously; antigen-binding fragments, including Fab', F(ab')2, Fab, Fv, single-domain antibody and scFv fragments, which are parts of antibodies that retain the ability to bind to antigens.

[0076] The term “Fab'” refers to an antibody fragment retaining the antigen-binding Fab portion and possessing a free sulfhydryl group located on the CH1 domain. This free cysteine allows for further modification through alkylation or conjugation with enzymes, toxins, or other relevant proteins.Attorney Docket No.: 5025-0446PWO1

[0077] The term “F(ab')2” refers to a dimer of Fab fragments, typically linked covalently via hinge cysteines near their carboxy termini. It's worth noting that additional chemical methods for coupling antibody fragments exist in the art.

[0078] The term “Fab” refers to an antibody fragment that contains the VH, CH1 and VL, CLregions, linked by an intramolecular disulfide bond.

[0079] The term “single domain antibody (VHH)” refers to a fragment of an antibody consisting of the variable domain of the heavy chain (VH), also known as a nanobody. VHHs are naturally found in camelids, such as llamas, alpacas, and camels. They are much smaller (approximately 25 kDa) than canonical antibodies (150 kDa), making them easier to produce and engineer.

[0080] The term “Fv” refers to the smallest antibody fragment capable of complete antigen recognition and binding. The Fv includes a tightly associated dimer, sometimes covalently linked as in scFv, formed by a single heavy chain variable domain (VH) and its light chain counterpart (VL). The term “scFv” refers to the Fv that is covalently linked, wherein the VHand VL domains are present in a single polypeptide chain.

[0081] The term “antigen” refers to a molecule recognized and responded to by the immune system, primarily because it is identified as foreign or non-self, thus triggering an immune response. Antigens have a variety of molecular structures, including proteins, peptides, polysaccharides, lipids, and nucleic acids (DNA and RNA). The role of antigens is to initiate the activation and specificity of the immune response by being recognized by specific immune cells or antibodies.

[0082] The term “tumor” refers to an abnormal growth of tissue in any organ, including benign (non-cancerous) or malignant (cancerous) types. A benign tumor is localized and does not spread to other parts of the body. It may grow large but typically does not invade nearby tissues or organs, therefore posing less of a health risk. On the other hand, a malignant tumor is cancerous and can invade surrounding healthy tissues. It has the potential to metastasize and spread to other parts of the body through the bloodstream or lymphatic system. As such, not all tumors are cancerous.

[0083] The term “antigen of a tumor” refers to an antigen that is either uniquely expressed or significantly overexpressed on the surface of tumor cells or in the extracellular matrix near the tumor cells, compared to its absence or minimal presence on normal cells. These distinctive antigens act as hallmarks, setting tumor cells apart from the body's healthy counterparts, and making them crucial targets for various approaches, including diagnosis, treatment, and immunotherapy.Attorney Docket No.: 5025-0446PWO1

[0084] In an embodiment, the binding protein binds an extracellular matrix of a tumor; wherein the extracellular matrix comprises a collagen.

[0085] The term “binding protein” refers a protein that binds one or more molecules, which can be other proteins (e.g. enzymes or collagen), nucleic acids (e.g. DNA or RNA), small molecules, or extracellular matrix. Exemplary binding proteins of the present disclosure include: RNA-binding protein (RBP, e.g. TDP-43, hnRNPs, or Poly(A)-binding protein), DNA-binding protein (DBP, e.g. p53, transcription factors, polymerase, or nucleases), actin- binding protein, or collagen binding protein (e.g. Netrin-1, decorin, lumican, or fibromodulin). It should be noted that the binding proteins in the present disclosure may not be full-length sequences, and it can be a domain derived from binding proteins (e.g. the RNA-binding domain in RBP).

[0086] In an embodiment, the antibody includes a scFv of hcc49.

[0087] The terms “hcc49” refers to a humanized antibody of CC49 (also known as Minretumomab), which is a mouse monoclonal antibody that is targeted to TAG-72 antigen. The construction of the hcc49 can refer

[0063] , the entire contents of which are incorporated herein by reference.

[0088] In an embodiment, the antigen includes glycoprotein.

[0089] In an embodiment, the glycoprotein comprises Tumor-Associated Glycoprotein-72 (TAG-72).

[0090] The term "glycoprotein" refers to a protein that is modified through the process of glycosylation, wherein various sugars (e.g., glucose, galactose, GalNAc, GlcNAc) are bonded to amino acids. Glycoproteins have diverse and complex structures, which result from intricate sugar attachment patterns. Moreover, tumor cells often express unique glycoproteins on their surface, making these glycoproteins suitable targets for antibody recognition and therapy. This specificity allows for the creation of treatments that can selectively target tumor cells without affecting normal cells. For example, the scFv of hcc49 (the antibody) provided in the present disclosure can recognize TAG-72 (the antigen on the surface of the cancerous tumor).In an embodiment, the tumor includes: breast cancerous tumor, lung cancerous tumor, liver cancerous tumor, pancreatic cancerous tumor, prostate cancerous tumor, skin cancerous tumor (melanoma), kidney cancerous tumor, bladder cancerous cancer, blood cancerous tumor (leukemia), colorectal cancerous tumor, and / or lymphoma.

[0091] The present disclosure also provides a pharmaceutical composition, including the aforementioned engineered enzyme and a pharmaceutically acceptable carrier.Attorney Docket No.: 5025-0446PWO1

[0092] The present disclosure also provides a pharmaceutical composition, including the aforementioned fusion protein and a pharmaceutically acceptable carrier.

[0093] The term “pharmaceutical composition” refers to a preparation in a form that allows the biological activity of the active ingredient(s) to be effective, and which contain no additional components which are toxic to the subjects to which the composition is administered. A pharmaceutical composition may include one or more active agents. For example, a pharmaceutical composition includes the engineered enzyme or the fusion protein as the sole active agent of the formulation, or may include the engineered enzyme or the fusion protein and one or more additional active agents, or an immune activator together.

[0094] The term “pharmaceutically acceptable carrier” refers to an ingredient in a pharmaceutical composition, other than an active ingredient, which is nontoxic to the subject to whom it is administered. A pharmaceutically acceptable carrier includes a buffer, excipient, stabilizer, or preservative.

[0095] The present disclosure also provides a method of enhancing an anticancer activity of an anticancer drug in a subject, comprising administering to the subject in need thereof the aforementioned engineered enzyme, or the aforementioned fusion protein, or the aforementioned pharmaceutical composition.

[0096] In an embodiment, the method further includes administering to the subject in need thereof the anticancer drug.

[0097] In an embodiment, the anticancer drug is metabolized to a glucuronidated metabolite in the subject.

[0098] In an embodiment, the anticancer drug is a glucuronidated form upon administration.

[0099] The term “a glucuronidated metabolite” refers to an anticancer drug that is not a glucuronide form when administered to the subject, but is converted to a glucuronide conjugate of anticancer by enzyme (e.g. UDPGT) in the subject. In contrast, the term “a glucuronidated form” refers to an anticancer drug that needs to be administered in the glucuronide form to the subject. The glucuronided anticancer drug is synthesized by chemical reactions before administration.

[0100] Preferably, the anticancer drug includes: Irinotecan (CPT-11), p-hydroxyaniline mustard glucuronide ester (HAMG-ester), Daunorubicin, Doxorubicin, Paclitaxel, Cisplatin, Docetaxel, Methotrexate, Temozolomide, Gemcitabine, Epirubicin, histone deacetylases inhibitors (CI-994), Etoposide, Cyclopamine, Camptothecin, Duocarmycin, Monomethylauristatin E, p-Hydroxyaniline mustard, and / or its derivatives thereof.Attorney Docket No.: 5025-0446PWO1

[0101] HAMG-ester is an alkylating agent that can inhibit transcription of DNA and then resulting inhibition of protein synthesis in cancer cells. The chemical structure of HAMG-ester is shown below.

[0102] The term “anticancer drug” refers to medications designed to prevent, or inhibit, cancerous tumor by interfering their growth or proliferation. The anticancer drugs include chemotherapy drugs and targeted therapy drugs: chemotherapy drugs work by damaging the DNA or cell division process of cancerous tumor to kill them. Exemplary chemotherapy drugs include Cyclophosphamide, Chlorambucil, Ifosfamide, Methotrexate, Fluorouracil, Vinblastine, Paclitaxel, Topotecan, Tamoxifen, or Letrozole; targeted therapy drugs target cancerous tumor by identifying specific marker or antigen on the surface of the cancerous tumor. Exemplary targeted therapy drugs include Imatinib, Erlotinib, Sorafenib, Trastuzumab, Bevacizumab, Cetuximab, Panitumumab, Gefitinib, Afatinib, Osimertinib, Dacomitinib, Neratinib, or Lapatinib.

[0103] The term “glucuronidated” or “glucuronidation” refers to a process that involves the conjugation of a molecule with a glucuronic acid, this process is catalyzed by the enzyme UDP- glucuronosyltransferase (UDPGT). Glucuronidation is a major pathway for the detoxification and elimination of xenobiotics, including drugs (e.g, anticancer drugs), pollutants, and environmental toxins. The glucuronide conjugate is more water-soluble than the parent xenobiotic, which makes it easier for the body to excrete.

[0104] In an embodiment, an immunogenicity of the engineered enzyme is lower than that of beta-glucuronidase derived from the heterologous sources. Preferably, the engineered enzyme is lower than that of beta-glucuronidase derived from the microbial sources. The term “immunogenicity” refers to a molecule is capable of stimulating, potentiating, activating, augmenting, boosting or enhancing an immune response. Different molecules may display different levels of immunogenicity to the body.

[0105] Examples

[0106] 1. Materials and methods

[0107] Cell lines and animalsAttorney Docket No.: 5025-0446PWO1

[0108] BALB / 3T3 fibroblasts (CCL-163; ATCC, Manassas, VA, USA) and 3T3 / hBG-DAF cells were cultured in Dulbecco’s Modified Eagle’s medium (high glucose) containing 10% fetal calf serum, 2.98 g L-1HEPES, 2 g L-1NaHCO3, 100 U mL-1penicillin, and 100 µg mL-1streptomycin

[0021] . MCF-7 human breast adenocarcinoma, LS174T human colon cancer cells, and HT-29 human colorectal cancer cell lines were cultured in RPMI-1640 with the same supplements. Experiments using mice were approved by the Institutional Animal Care and Use Committee of the Institute of Biomedical Sciences, Academia Sinica (protocol 1207384). All procedures performed in studies involving animals were in accordance with the ethical standards of the institution and were carried out under specific-pathogen free conditions.

[0109] Reagents and antibodies

[0110] SN-38G (7-ethyl-10-hydroxycamptothecin glucuronide) was purchased from Toronto Research Chemicals (Toronto, ON, Canada). Rabbit anti-hBG IgG and HRP-Donkey anti- mouse IgG were from Abcam (Eugene, OR, USA). Mouse anti-β-actin antibody was from Sigma-Aldrich (St. Louis, MO, USA). Fluorescein Di-beta-D-glucuronide (FDGlcU), Pierce IP lysis buffer, BCA protein assay kits, and HRP-goat anti-rabbit IgG were from Thermo Scientific Fisher (Waltham, MA, USA). Purified mouse anti-hBG IgG (7G8), rat anti-mouse beta-glucuronidase IgG (7G7) and mouse anti- E.coli beta-glucuronidase IgG (1E8) monoclonal antibodies were produced in our lab

[0025] .3H-thymidine was from Perkin Elmer (Waltham, MA, USA).

[0111] Human beta-glucuronidase individual libraries

[0112] Forty-nine mutation sites were distributed into five libraries (Tables 1-5). Mutation sites were selected by comparison of the amino acid sequences of mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, Pongo and human) versus BG derived from E. coli, Aspergillus, Clostridium, Staphylococcus, Streptococcus and Lactobacillus. Each library contained mutations at 8 to 11 amino acid positions. Degenerate primers bearing the specific mutations of each library were assembled by PCR

[0022] . Wild-type fragments were amplified from pLNCX-HA-HIS-hBG-myc-DAF

[0022] and combined with the assembled mutant library fragments with terminal 5’ Apa I and 3’ Sal I restriction sites. After restriction enzyme digestion, the library fragment was subcloned into pLNCX-HA-HIS-hBG-myc-DAF. This plasmid allows expression of hBG variants tethered to the surface of mammalian cells via a GPI anchor. The five library plasmids were individually transformed into DH10α competent cells by electroporation to produce sufficient plasmid DNA for transfection of retroviral packaging cells.Attorney Docket No.: 5025-0446PWO1 Species Amino Acid position332 340 348 349 363 396 409 422 423 431 435Human T G V N W C G S L E R Q S K onhBG. Human indicates the amino acids present in hBG. Mammalian indicates the consensus amino acids at the corresponding position in mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human). The corresponding amino acids in bacterial BG species arranged with optimal activities at pH values that increase from about 6 for Aspergillus to 7 for E. coli [76-80]. The target and actual amino acid substitutions in the library are listed. Amino Acid position Species 456 46046465 469 476 481 484 486 488 K Don hBG. Human indicates the amino acids present in hBG. Mammalian indicates the consensus amino acids at the corresponding position in mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human). The corresponding amino acids in bacterial BG species arranged with optimal activities at pH values that increase from about 6 for Aspergillus to 7 for E. coli [76-80]. The target and actual amino acid substitutions in the library are listed.Attorney Docket No.: 5025-0446PWO1 Species Amino Acid position492 495 503 506 509 513 517 518 519 520Human G Y S S H H I Q L Q onhBG. Human indicates the amino acids present in hBG. Mammalian indicates the consensus amino acids at the corresponding position in mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human). The corresponding amino acids in bacterial BG species arranged with optimal activities at pH values that increase from about 6 for Aspergillus to 7 for E. coli [76-80]. The target and actual amino acid substitutions in the library are listed. Species Amino Acid position522 524 526 527 538 539 545 626 627 628G T Q N. . on hBG. Human indicates the amino acids present in hBG. Mammalian indicates the consensus amino acids at the corresponding position in mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human). The corresponding amino acids inAttorney Docket No.: 5025-0446PWO1 bacterial BG species arranged with optimal activities at pH values that increase from about 6 for Aspergillus to 7 for E. coli [76-80]. The target and actual amino acid substitutions in the library are listed. Species Amino Acid position563 578 580 585 593 595 599 602Human K R Y L M E T L V G Ted on hBG. Human indicates the amino acids present in hBG. Mammalian indicates the consensus amino acids at the corresponding position in mammalian BG (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human). The corresponding amino acids in bacterial BG species arranged with optimal activities at pH values that increase from about 6 for Aspergillus to 7 for E. coli [76-80]. The target and actual amino acid substitutions in the library are listed.

[0113] Retroviral transduction and generation of library cells

[0114] Plasmid DNA of each library was co-transfected (calcium phosphate transfection) with pVSV-G (Clontech, Mountain View, CA, USA) into GP2-293V packaging cells to produce recombinant retroviral particles. Two days after transfection, the culture medium containing retroviral particles was filtered through a 0.45 µm membrane, mixed with 8 μg mL-1polybrene, and added to 3T3 fibroblasts at a multiplicity of infection of 0.1 to introduce single gene copies of each enzyme variant into cells. Stable cell lines were selected in medium containing 0.5 mg mL-1G418 (Calbiochem, San Diego, CA, USA).

[0115] Library cell screening

[0116] After G418 selection, 3T3 cells were first screened by fluorescence-activated cell sorting on a FACS Vantage DiVa (Becton Dickinson, Franklin Lakes, NJ, USA) to collect cells expressing high levels of hBG variants on the cell surface.5x106 library cells were then stained with FITC-labeled anti-hBG (7G8)

[0025] for 30 min at 4°C in HBSS (5.4 mM KCl, 0.3 mM Na2HPO4, 0.4 mM KH2PO4, 4.2 mM NaHCO3, 1.3 mM CaCl2, 0.5 mM MgCl2, 0.6 mMAttorney Docket No.: 5025-0446PWO1 MgSO4, 137 mM NaCl, 5.6 mM D-glucose pH 7.4) containing 2% fetal bovine serum (FBS) followed by propidium iodide staining for gating live cells. Single live cells positive for 7G8- FITC staining were arrayed into 96-well culture plates for prodrug screening.

[0117] Functional screening of enzyme variants

[0118] Single cells in 96-well plates were expanded to near confluence, washed once with PBS, and then incubated with 0.05 U mL-1PI-PLC (phosphatidylinositol-specific phospholipase C from Bacillus cereus (Molecular Probes, Eugene, OR, USA) in 200 µL PBS at 37oC for 2 h to cut the GPI anchor and release hBG variants into the culture media. Culture media was transferred to low protein binding plates (NUNC, Roskilde, Denmark). Fifty µL of the collected protein mixture was dispensed to 96-well microtiter fluorescence plates (NUNC, Roskilde, Denmark) with 200 nM SN-38G (for individual libraries and combined library screening) or 1 μM SN-38G (for shuffled library screening) in 50 µL reaction buffer (100 mM acetic acid, 50 mM Bis-Tris, 50 mM triethanolamine, pH adjusted to 7.0 with NaOH) at 37°C for 16 h (for individual libraries and combined library screening) or 4 h (for shuffled library). The reaction was terminated by adding 100 µL stop buffer (1 M glycine, 0.5 M sodium carbonate, pH 9.0) to each well. Relative fluorescence was detected at excitation / emission wavelengths of 375 / 470 nm for SN-38G and 375 / 560 nm for SN-38 on a SpectraMax Gemini EM fluorescence microplate reader (Molecular Device, Sunnyvale, CA, USA). SN-38G conversion was estimated according to the Beer Lambert’s law.

[0119] Enzyme-linked immunosorbent assay

[0120] The amount of hBG variants in media collected from each well of PI-PLC treated 3T3 library cells was determined by enzyme-linked immunosorbent assay. 7G8 anti-hBG monoclonal antibody (0.1 µg) in 50 µL coating buffer (5.3 g L-1Na2CO3+ 4.2 g L-1NaHCO3, pH adjusted to 8.0 with 1N NaOH) was coated in the wells of 96-well ELISA plates at 4oC overnight. Plates were washed 3 times with PBS and then blocked with 5% skim milk in PBS at room temperature for 1 h. The plates were washed 3 times with PBS before 20 µL of PI-PLC treated culture media was added to each well along with 30 µL 2.5% skim milk in PBS for 1 h at room temperature. After washing the plates twice times with PBS containing 0.05% Tween 20 (PBS-T), and twice with PBS, 20 ng 7G8-biotin and 50 ng streptavidin-HRP in 50 µL PBS / 2.5% skim milk were added at room temperature for 1 h. The plates were washed as above after each addition. 100 µL freshly prepared H2O2 / ABTS (2,2'-azino-bis (3- ethylbenzothiazoline-6-sulphonic acid) mixture (1:3000) was added at room temperature for 30 min and the absorbance of each well was read at 405 nm on a microplate reader (Molecular Device, California, USA).Attorney Docket No.: 5025-0446PWO1

[0121] Human beta-glucuronidase combined library

[0122] Seventy-six 3T3 clones with higher activity for hydrolysis of SN-38G (> 2-fold wildtype hBG) were collected (5 × 106 3T3 cells total), mixed with 1 mL TRIZO Reagent(Invitrogen), and incubated at 25oC for 5 min. Total cDNA from the extracted RNA was used as template to generate variant human beta-glucuronidase genes. Six pairs of primers were used to amplify the corresponding mutated regions from each individual library. Two rounds of assembly PCR were performed to assemble mutated regions from the five individual libraries to create a combined library. The assembled combined fragments were then cloned back into pLNCX-HA-HIS-hBG-myc-DAF to form a library combining all the individual mutants into one library (combined library).

[0123] Human beta-glucuronidase shuffled library

[0124] Positive clones selected by screening the combined libraries against SN38G were pooled and processed to extract mRNA. Total cDNA from the extracted RNA was used as a template to recover human beta-glucuronidase combined library fragments. DNA shuffling of fragments was performed by digesting 1 µg DNA with 0.03 U of DNAse I (Takara, Japan) in 25 µL 50 mM Tris-HCl, 1 mM MnCl2for 9 minutes at 25°C. The reaction was quenched by adding 5 mL of 0.5 M EDTA. DNA fragments between 100 to 200 bp were collected via gel extraction and reassembled by two rounds of assembly PCR

[0022] . The shuffled library was subcloned into pLNCX-HA-HIS-hBG-myc-DAF.

[0125] Prediction of hBG structures

[0126] The structures of hBG3, hBG10, and hBG19 were estimated using Alphafold2 Colab with the 1.7 Å crystal structure of hBG (RCSB Protein Data Bank; PDB: 3hn3) used as the template

[0026] . The amino acids in contact with SN-38G and the enzyme pocket size were estimated by docking SN-38G (SN38G:PDB) to the enzymes using CB-Dock2

[0027] . The amino acid contact frequency was estimated by performing twenty dockings for each enzyme. Glycosylation was modeled with glycoprotein builder on Glycam (https: / / glycam.org / ). Molecular graphics and analyses performed with PyMOL (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC).

[0127] Production of recombinant BG

[0128] cDNA coding human (hBG), or hBG variants were cloned into the pLNCX retroviral vector between Sfi I and Sal I restriction enzyme sites. This vector appends a histidine tag to the C-terminus of the recombinant proteins. A mutant form of eBG (eBG-NS), in which a cryptic glycosylation site which destroys enzyme activity was mutated from N358 to S358, was also cloned in the pLNCX vector for mammalian cell production

[0028] . RecombinantAttorney Docket No.: 5025-0446PWO1 retroviral particles were generated by cotransfection of 5 µg specific pLNCX plasmid with 5 µg of pVSVG in GP293 cells, respectively. After 48 h, the culture medium was collected from GP293 cells, passed through a 0.45 µm filter, mixed with polybrene at final concentration of 8 µg mL-1, and added to 3T3 cells for 3 days. Stable 3T3 cells were selected by culturing in medium containing 0.5 mg mL-1G418 for 7-14 days. The 3T3 stable cells were inoculated into a bioreactor, and the culture medium was collected once per week. Different recombinant beta- glucuronidase enzymes were purified by TALON®metal affinity column (Takara, CA, USA) according to the manufacture instruction. To assess enzyme purity, 5 µg of purified proteins were boiled in reducing SDS buffer for 10 min, separated on a 10% SDS-PAGE and stained with Coomassie brilliant blue for 1 h.

[0129] Enzyme glycosylation

[0130] Purified hBG and hBG10 (~ 10 µg) were dissolved in 6 M urea and 20 mM dithiothreitol in a volume of 50 µL and incubated at 56oC for 1 h. Iodoacetemide was added to 0.55 M in the dark at room temperature for 45 min before adding 50 mM triethylammonium bicarbonate buffer to dilute urea to less than 1 M. Trypsin was added at a 50-fold molar ratio of trypsin to enzyme at 37oC for 16 h. Trifluoroacetic acid was added to 0.1% final concentration to deactivate trypsin before the sample was desalted on a C18 Zip-Tip (Millipore) into 50% acetonitrile / 0.1% trifluoracetic acid and dried in a SpeedVac. The tryptic peptides were analyzed by LC-MS / MS on a Tribrid Mass Spectrometer (Orbitrap Fusion Lumos, Thermo Fisher Scientific) coupled to an Easy-nLC 1200 System (Thermo Fisher Scientific) with a 90-min LC separation on an Acclaim PepMap RSLC column (25 cm × 75 μm i.d., Dionex). An HCD-pd-EThcD workflow of the MS acquisition method was employed to additionally trigger EThcD events upon detecting the HCD product ions at m / z 204.0867, 138.0545, and 366.1396. For glycopeptide identification, the MS raw data were searched using Byonic algorithm (v4.1.10, Protein Metrics) with the following search parameters: peptide tolerance = 10 ppm; fragment tolerance = 0.02 Da; missed cleavages = 2; modifications: carbamidomethyl cysteine (fixed), methionine oxidation (common 2), deamidation at NQ (common 1). The N-glycan database was searched against 182 human no multiple fucose glycan compositions. The cutoff score of the identified spectrum was 300. To identify phosphorylated mannose on glycopeptides, phosphorylation on glycans was added during the database search and the spectra were then manually examined with the featured fragment ions of Man-P (243.026+) and Man2-P (405.079+) to determine if the spectra contained phosphorylated mannose

[0029] .

[0131] Enzyme activityAttorney Docket No.: 5025-0446PWO1

[0132] Enzyme activity using the general glucuronide substrate 4-methylumbelliferyl β-D- glucuronide was determined by serially diluting enzyme samples from 20 ng to 0.25 ng in 20 μL reaction buffer (50 mM bis-tris, 50 mM triethanolamine, 100 mM acetic acid, 0.05% bovine serum albumin, pH 7) and then adding 80 μL of 1.25 mM 4-methylumbelliferyl β-D- glucuronide (4-MUG) at 37oC for 30 min. The reaction was terminated by adding 100 μL stop buffer (2 M Tris-base, 0.8 M sodium bicarbonate, pH 9.7). Fluorescence was measured on a Gemini EM microplate spectrofluorometer (Molecular Device) at excitation / emission wavelengths of 355 / 460 nm, respectively to detect 4-methylumbelliferone (4-MU).

[0133] Enzyme hydrolysis of SN-38G was determined by first preparing SN-38G solutions with concentrations ranging from 3200 μM to 12.5 μM in BG reaction buffer at pH values of 4.0, 6.5 and 7.0 starting from an 80 mM stock solution in 100% DMSO. Enzymatic reactions were initiated at 37oC by mixing SN-38G and beta-glucuronidase enzyme solutions to give a final DMSO concentration of 5%. Reactions were terminated at 0, 6, 12, 20 and 30 min by mixing samples with an equal volume of 50% acetonitrile, 5% DMSO in water. The final solution was diluted with 10 mM KH2PO4, 19% ACN, pH 2.7 and analyzed by SPE-HPLC. The SPE column was equilibrated with SPE buffer (10 mM KH2PO4 in 5% acetonitrile, pH 2.7) and diluted samples were first captured on a SPE column and then eluted with mobile phase (50 mM KH2PO4 in 27% acetonitrile) to an analytical C18 column (μBondapak). SN38 fluorescence was measured on a fluorescence detector (Jasco FP-2020 Plus) at excitation / emission wavelengths of 375 / 560 nm. Area under peak (AUP) were calculated using Trilution Liquid Chromatography 3.0. A standard curve correlating AUP and SN-38 concentration, ranging from 10 nM to 200 nM, was used to determine unknown SN38 concentrations. Kinetic parameters were calculated by Michaelis-Menten curve fitting using GraphPad Prism 9.

[0134] Enzyme stability

[0135] Enzymes (100 μg mL-1) were incubated at 37oC for 3, 6 and 9 days in 100% human serum, 100% mouse serum from NOD-SCID or C57BL / 6 mice, or sterile PBS (pH 7) containing 0.5% BSA. Enzyme activities were measured by incubating samples with 1.25 mM 4-MUG in reaction buffer at 37oC for 30 min. After adding an equal volume of stop buffer, the fluorescence of 4-MU was measured as described above.

[0136] Thermal stability was determined by diluting purified enzymes in PBS at a final concentration of 0.5 mg mL-1. Fifteen microliters of purified enzymes were loaded into capillaries and the melting temperature (Tm) of the enzymes was measured on a NanoTemper Tycho NT.6 differential scanning fluorimeter.Attorney Docket No.: 5025-0446PWO1

[0137] 3H-Thymidine incorporation assay

[0138] MCF-7, HT-29 and LS174T human cancer cells were seeded in complete RPMI medium overnight at 10,000 cells per well in 96 well culture plates at 37oC in a humidified atmosphere of 5% CO2in air. Enzymes serially diluted in RPMI medium containing 20 mM HEPES and 10% FBS, without Na2HCO3 (pH 6.8) were added to the wells. Then, 5 nM of SN- 38G was added to MCF-7 and LS174T cells, and 20 nM of SN-38G was added to HT-29 cells for 48 h at 37oC. 50 µL fresh medium containing 0.3 µCi of3H-thymidine was added to each well at 37°C for 16 h. The cells were harvested and fixed on glass-fiber filters to measure radioactivity on a Topcount scintillation counter. Results are expressed as percent of3H- thymidine incorporation compared with untreated cells.

[0139] Immunoenzyme activation of SN-38G was measured in a similar fashion by first adding 1 μg immunoenzymes per well for 1 h at 37oC. The cells were then washed twice with PBS containing 0.495 mM MgCl2 and 0.9 mM CaCl2, and serial dilutions of SN-38G was added to cells in triplicate for 48 h at 37oC.3H-thymidine incorporation was then measured as above.

[0140] Generation of homozygous human beta-glucuronidase transgenic mice

[0141] The pCAG-hBG plasmid was constructed by insertion of hBG cDNA into pCAGGS which can regulate heterogeneous gene expression in mice under control of the chicken β-actin promotor

[0030] . The cDNA fragment from pCAG-hBG that included the hBG gene and β-actin promoter was microinjected into fertilized eggs of C57BL / 6 mice and BALB / c mice, and then the hybrid egg was transferred to pseudopregnant Swiss Webster female mice. BALB / c-Tg- hBG (Line 18) and C57BL / 6-Tg-hBG (Line 675) were created as founders for this study. The genotype of mice was checked by PCR of toe samples using the forward primer 5’- CATTCCTATGCCATCGTGTGG-3’ (SEQ ID NO: 4) and reverse primer 5’- ACGCACTTCCAACTTGAACAGG-3’ (SEQ ID NO: 5). The genotypes of homogenous transgenic mice were also confirmed by performing mating tests. The expression of hBG protein in homozygous BALB / c-Tg-hBG mice were checked by western blotting of tail samples. To perform western blotting, tail samples from BALB / c-Tg-hBG and C57BL / 6-Tg- hBG mice were homogenized and total protein were extracted in Pierce®IP lysis buffer according to the manufacturer’s instructions. The protein concentration of lysates was measured using BCA protein assay kit. The lysates (50 µg of protein) were separated on a 12% Tris-glycine SDS-polyacrylamide gel, and then transferred to a PVDF membrane (Merck Millipore, Burlington, MA, USA). Membranes were incubated with the following antibodies (dilutions used, source and catalog numbers): β-actin (1:5000, Sigma Aldrich, A5441) and hBG (1:1000, Abcam, ab166904). For detection of β-actin, membranes were then incubated withAttorney Docket No.: 5025-0446PWO1 HRP-donkey anti-mouse IgG (1:5000, Abcam, a16011) at room temperature for 1 h. For detection of hBG, membranes were incubated with HRP-goat anti-rabbit IgG (1:2000, Invitrogen, 31460) at room temperature for 1 h, followed by reaction with ECL detection reagent and then luminescence was detected by on a UVP BioSpectrum Auto Imaging System (Thermal Fisher Scientific). The density of each band was quantified with Vision WorksTM LS Image acquisition and analysis software.

[0142] Enzyme immunogenicity

[0143] Wild type BALB / c, C57BL / 6 hBG transgenic and BALB / c hBG transgenic mice (6-8 weeks) were intravenously injected with 100 μg of hBG, hBG3, hBG10, hBG19 or eBG-NS every week for a total of 7-9 times. Blood was collected weekly immediately before enzyme administration. Serum separated from whole blood by centrifuging at 3,000xg for 10 min was stored at -30oC. To measure antibody responses against enzymes, purified enzymes (0.5 μg) in coating buffer (50 mM Na2CO3, 50 mM NaHCO3, pH 9.5) were separately coated in 96-well ELISA plate at 4oC overnight. The plates were blocked in 5% skim milk powder in PBS at room temperature for 1 h. Serum samples were serially-diluted in 2% skim milk powder in PBS, and then 50 μL of diluted serum was added to 96-well plates at room temperature for 1 h. Serial dilutions (5 µg mL-1to 0.3 ng mL-1in 2% skim milk) of mouse anti-eBG antibody (1E8) or mouse anti-hBG antibody (7G8) were added to the corresponding plates as antibody standards. The plates were subsequently washed twice with 250 µL PBS-T and once with PBS. HRP-goat anti-mouse IgG (0.2 µ g mL-1) in 2% skim milk was then added at room temperature for 1 h. The plates were washed as above before 100 µL of ABST / H2O2substrate mixture was added to each well for 30 min at room temperature. The absorbance of wells was measured at 405 nm on a microplate reader (Molecular Devices, CA, USA). hBG10xA (hBG10 with no enzyme activity) and hBG10xG (hBG10 mutated to remove a crptic glycosylation site) were generated by site-directed mutagenesis of hBG10 to make the single amino acid substitutions E540A and T464N, respectively.

[0144] Enzyme elimination kinetics in mice

[0145] Wild type C57BL / 6 mice (3-5 mice / group, 8 weeks) were intravenously injected with 98 µg eBG-NS, hBG, hBG3. hBG10, or hBG19 spiked with 2 µg I125-labeled enzyme. The specific activity of the radiolabeled enzymes ranged from 2×107to 3×107cpm μg-1. Wild type BALB / c mice (3 mice / group, 8 weeks) also received 8 µg eBG-NS or hBG10 spiked with 2 µg of the corresponding I125-labeled enzyme. Blood samples were collected from mice at 5 min, 30 min, 1 h, 6 h and 24 h followed by centrifugation at 12,700xg for 5 min to obtain serum. The radioactivity of serum from mice was measured on a multichannel gamma-counter.Attorney Docket No.: 5025-0446PWO1

[0146] hBG immunoenzymes

[0147] Immunoenzymes targeting the TAG-72 tumor-associated antigen were constructed by fusing a humanized single-chain antibody fragment of CC49 (hcc49 scFv) to the genes coding hBG, hBG3, hBG10, or hBG19 to create hcc49-hBG, hcc49-hBG3, hcc49-hBG10, and hcc49- hBG19, respectively

[0031] . Negative-control immunoenzymes were constructed by fusing a scFv against the small molecule hapten dansyl (DNS) to hBG or hBG10 to create DNS-hBG and DNS-hBG10, respectively. Generation of stable 3T3 producer cell lines and purification of the immunoenzymes was carried out as described for production of recombinant enzymes.

[0148] Immunoenzyme ELISA

[0149] Bovine submaxillary gland mucin or bovine serum albumin (1 µg / well) were diluted in coating buffer (50 mM Na2CO3, 50 mM NaHCO3, pH 9.5), and then added to 96 well plates (50 μL / well) at 4oC overnight. After washing plates once with PBS, the wells were blocked with 250 μL of 5% skim milk powder in PBS at room temperature for 2 h. Fifty μL serial dilutions of immunoenzymes in 2% skim milk powder in PBS were added to wells for 1 h. The plates were washed twice with PBS-T and with PBS once. Mouse anti-hBG IgG (0.08 μg / well) in 2% skim milk powder in PBS was added for 1 h. After washing with PBS-T for 3 times, donkey anti-mouse HRP antibody (0.005 µg / well) in 2% skim milk in PBS was added for 1 h and ELISA plates were analyzed as described above.

[0150] Immunoenzyme surface plasmon resonance

[0151] Immunoenzyme binding kinetics were measured on a Biacore T200 (GE Lifesciences Uppsala, Sweden)

[0031] . 200 µg mL-1bovine submaxillary gland mucin (BSM) in 100 mM acetate buffer (pH 3) was immobilized on CM5 chips by amine coupling. Binding analysis were performed in 1x HBS-EP buffer (Cytiva, Sigma, USA) at a flow rate of 50 μL min-1at 25oC. Purified immunoenzymes were diluted in HBS-EP. The chip surface was washed by regeneration buffer (6 M guanidine, 0.2 M acetic acid) for 420 s. Data were analyzed by Biacore T200 evaluation software.

[0152] Immunoenzyme flow cytometer analysis

[0153] HT-29, LS174T, and MCF-7 cells (106cells) were incubated with 10 μg mL-1of purified immunoenzymes or DNS-hBG in 200 µL PBS for 1 h on ice. The cells were washed twice with PBS and incubated with 1.5 μg mouse anti-hBG IgG in 200 µL PBS for 45 min on ice. After washing twice with PBS, the cells were incubated with 0.45 µg goat F(ab)’2anti- mouse IgG Fc-FITC on ice for 45 min. Cells were washed twice with PBS and then stained with 5 μg of propidium iodide. The surface immunofluorescence of 105cells was measuredAttorney Docket No.: 5025-0446PWO1 with a FACScaliber flow cytometry (Becton Dickinson, Mountain View, CA, USA). Results were analyzed with Flowjo software (Tree Star Inc., San Carlos, CA, USA).

[0154] Immunoenzyme immunofluorescence

[0155] Cryostat sections of LS174T and MCF-7 tumors (10 µm) were fixed in ice-cold acetone for 10 min, and then washed with PBS three times. The tissue sections were blocked with 0.5% BSA / PBS for 1 h, then incubated with 25 µg / ml hcc49-hBG, hcc49-hBG10 or DNS-hBG overnight at 4oC. The tissue sections were then incubated mouse anti-hβG antibody for 2 h at 37oC, and subsequently stained with 2 µg / ml FITC-conjugated goat anti-mouse IgG (H+L) (Thermo Fisher Scientific) for 1 h at room temperature. The tissues were then stained 3 μM DAPI for 10 min at room temperature. ProLong® Gold Antifade Mountant (Thermo Fisher Scientific) was added to the surface tissue, and then tissues were sealed by coverslips. Images were captured using a fluorescent microscope with AxioCam (Karl Zeiss, Thornwood, NY, USA).

[0156] Imaging the glucuronide-converting activity of immunoenzymes in mice

[0157] LS174T tumor bearing mice (100-150 mm3) were intravenously (i.v.) injected with 200 μg hcc49-hBG10 or DNS-hBG10. After 96 h, the mice were then i.v. injected with 100 μg of FDGlcU. Mice were sacrificed after 30 min and tumor, lung and liver tissue were collected from mice for IVIS Spectrum imaging system detection (λex : λem =465 nm : 520 nm). Images were analyzed by Living Image 4.2 software (PerkinElmer).

[0158] Antitumor activity in vivo study of the anticancer drugs CPT-11 and HAMG-ester

[0159] A pilot experiment examining hcc49-hBG10 antitumor activity was carried out by first injecting 2.5x106LS174T cancer cells in the right flank of female NOD / SCID mice. When tumors reached approximately 100 mm3, groups of three mice were i.v. injected PBS or 200 μg hcc49-hBG10 on days 0 and 10. After 72h and 96h of immunoenzyme administration, mice were intraperitoneally (i.p.) injected with vehicle (PBS + 6% DMSO), i.v. injected with 25 mg kg-1SN-39G, or i.p. injected with 50 mg kg-1CPT-11.

[0160] In a larger experiment, female NOD / SCID mice were subcutaneous inoculated with 2.5x106LS174T cancer cells to their right flank. Mice bearing approximately 100 mm3subcutaneous LS174T tumors (6 per group) were i.v. injected with PBS or 200 μg hcc49-hBG, hcc49-hBG10 or DNS-hBG10 on days 0, 10, 20, 30 and 40.72 h and 96 h after immunoenzyme administration, the mice were i.p. injected with 50 mg kg-1CPT-11. Tumor size and weight of mice were measure every 3 days. Tumor volumes were calculated according to the formula: length × width × height × 0.5.Attorney Docket No.: 5025-0446PWO1

[0161] In addition, another anticancer drug, HAMG-ester, was also evaluated. LS174T-tumor bearing NOD-SCID mice (50-100 mm3, n=3) were injected 200μg hcc49-hBG10 by intravenous injection (i.v.) for once at Day 0 and Day 7. Mice (n=3) were treated 20, 25, 30 mg / kg HAMG-ester once by intravenous injection (i.v.) at Day 3, Day 4, Day10 and Day 13, respectively. Mice were also administrated PBS as negative control. Body weight and tumor size of mice were measured every 3 days. Tumor size of mice were calculated by formula: length × width × height × 0.5.

[0162] Statistical analysis

[0163] Student’s t test was used to calculate significant differences between mean values with GraphPad Prism 8. Statistic significant was indicated as *p<0.05.

[0164] 2. Results

[0165] Directed evolution of hBG

[0166] We sought to enhance the catalytic activity of hBG for the hydrolysis of SN-38G at the pH (6.5~7.0) found in solid tumors [32, 33], with the ultimate goal of generating immunoenzymes for the treatment of solid tumors. hBG, however, is a lysosomal enzyme with optimal activity at acidic pH. Because beta-glucuronidase derived from other mammalian sources also displays optimal activity at acidic pH values while bacterial BG displays high activity near neutral pH values, we identified amino acids that are differentially conserved between BG derived from several mammalian sources (dog, cat, pig, rat, mouse, horse, chicken, African green monkey, pongo and human) versus BG derived from microbial sources including Aspergillus, Streptococcus, Staphylococcus, Lactobacillus, E. coli and Clostridium. We focused on amino acids present in domain three of the enzymes, which contains the active site glutamic acids (E451 and E540). Five libraries coding hBG variants with substitutions at forty-nine target amino acid positions were expressed on the plasma membrane of fibroblasts via a glycosylphosphatidylinositol (GPI) anchor (Tables 1-5). Low multiplicity of infection was used to express a single hBG variant in each cell. Fibroblasts that expressed high levels of GPI-anchored hBG variant on their surface were arrayed as single cells in 96-well culture plates by flow cytometric sorting, cultured to near confluence, and treated with phosphatidylinositol phospholipase C to cleave the GPI anchor and release hBG variants into the culture medium (Fig. 2). The concentration of soluble hBG and hydrolysis of SN-38G were measured to calculate relative specific enzyme activities. hBG variants that displayed at least two-fold greater activity than hBG were collected (2, 30, 7, 22, and 15 variants from libraries 1, 2, 3, 4, and 5, respectively). The areas containing mutation in each hBG library were amplified and reassembled to generate a full-length combined SN-38G library. Approximately 70% ofAttorney Docket No.: 5025-0446PWO1 combined library cells displayed membrane-tethered hBG variants, with 90% of the clones possessing activity greater than hBG. Twenty-five clones with activity at least four times greater than hBG were collected from four rounds of screening of the SN-38G combined library. DNA shuffling was performed on the hBG variant genes isolated from the combined libraries. After 3 rounds of screening, 37 clones were collected that display at least 20-fold greater specific activity than hBG.

[0167] We sequenced several clones that were selected from each individual library and after performing DNA shuffling. Figs. 13-17 show for amino acid positions present in each individual library the predicted frequency that new amino acids should appear under zero selection pressure (black bars), the measured frequency of amino acid substitutions after screening the individual libraries (green bars), the measured frequency of amino acid substitutions after DNA shuffling (blue bars) and the frequency of substitution of the most prevalent amino acid (red bars) with 100% representing replacement of the original amino acid with a single new amino acid. Wild type amino acids were retained during both single library screening and DNA shuffling at about one third of the 49 positions, indicating that these amino acids are critical for enzyme stability or activity. This was especially apparent in individual libraries 4 and 5 (i.e., Q524, E526, N527, Q538, S539, R578, Y580, and L585).

[0168] Enzyme catalytic activity and molecular docking

[0169] Two clones with high activity (hBG10 and hBG19, with 10 and 19 amino acid substitutions, respectively) were selected for further investigation. We also included hBG3 (three amino acid substitutions), which was previously isolated by ECSTACY using a commercial fluorescent substrate for screening (Table 6)

[0022] . Each hBG monomer consists of three structural domains with topologies analogous to a jelly roll barrel, an immunoglobulin constant domain and a TIM barrel

[0034] . The catalytic glutamic acids at positions E451 and E540 are present in the TIM barrel of domain 3. Amino acid changes that enhance catalytic activity are located at the loops and external sides of the TIM barrel-forming helixes as well as near the substrate binding domain in the active center (Fig. 18). A potential new N-linked glycosylation site was introduced in hBG10 at N484 due to the substitution N486T.

[0170] Purified enzymes were produced from stably-transfected 3T3 fibroblasts. The purified enzymes ran as monomers on a reducing SDS PAGE with molecular weight of approximately 72 kDa (Fig. 19). The molecular weight of hBG10 was slightly larger, consistent with productive glycosylation at N484. We confirmed the presence of a N-linked glycosylation site at N484 by mass spectrometry. Mannose 6-phosphate groups were not detected on the terminal mannose groups of the new glycan. hBG3 is about 8-fold more active while hBG10 and hBG19Attorney Docket No.: 5025-0446PWO1 are nearly 40-fold more active than hBG for the hydrolysis of SN-38G to SN-38 (Fig. 3A). A more detailed analysis of SN-38G hydrolysis at different pH values shows that the three enzymes have distinct kinetic properties (Table 7). hBG displays high activity at pH 4.5, but we were unable to measure the kinetic properties of hBG at pH 7 due to low activity. hBG10 and especially hBG19 binds SN-38G tightly as reflected by relatively low Km values, even at pH 7 (Fig 3B). By contrast, the catalytic activity (Kcat) of hBG3 is substantially higher than the other enzymes, which is most apparent at pH 6.5 (Fig.3C). The overall enzymatic activity as measured by Kcat / Km decreases as pH increases except for hBG19 which exhibits greater activity at pH 6.5 (Fig. 3D). All enzymes have substantially greater enzymatic activity (Kcat / Km) at pH 6.5 as compared to pH 7, with both hBG10 and hBG19 exhibiting the highest enzyme activities.

[0171] Molecular docking studies indicated that SN-38G bound to the same amino acids in the catalytic pocket of hBG3 and hBG. The estimated sizes of the cavity for SN-38G binding in hBG3 and hBG were 1434 and 1390 Å3, respectively using the program CB-Dock2 (Fig.20). This is consistent with the similar Km values of hBG and hBG3 for SN-38G. By contrast, SN- 38G is predicted to make several new contacts in a smaller cavity (936 Å3) in hBG10 (Figure 4). In addition to new interactions with Y487, L501, Q520, and W528, SN-38G is predicted to interact with T486, which is only present in hBG10 (Fig.20). SN-38G binding to the catalytic pocket of hBG19 appears distinct from the other hBG enzymes; the catalytic pocket is much smaller (596 Å3) and the planer camptothecin backbone of SN-38G is rotated to interact with the amino acid substitutions N486F and H509Y. The new interactions in hBG10 and hBG19 may be responsible for their stronger binding (lower Km) of SN-38G. Amino acid positionhBG hBG3 hBG10 hBG19 mBG340 G G G N GAttorney Docket No.: 5025-0446PWO1 495 Y Y Y F Y 506 S S S G S 509 H H H Y H Tabl ndicatedpositions are shown for hBG, hBG3, hBG10, hBG19, and mouse beta-glucuronidase (mBG) pH Parameters hBG hBG3 hBG10 hBG19 05for the hydrolysis of SN-38G to SN38 at pH 4.5, pH 6.5 and pH 7.0. Results show mean values ± SD. (n = 3). ND, not determined due to insufficient activity

[0172] In vitro anticancer activity of SN-38G and hBG variants

[0173] To investigate how effectively hBG enzymes can generate cytotoxic concentrations of SN-38 from SN-38G, MCF-7 human breast cancer cells, LS174T human colon cancer cells, and HT-29 human colon cancer cells were incubated with serial dilutions of the enzymes in the presence of non-toxic concentrations of SN-38G. HT-29 cells were slightly less sensitive to SN-38G so cells were incubated with 20 nM SN-38G whereas both MCF-7 and LS174T cells were incubated with 5 nM SN-38G. Under these conditions, SN-38G did not cause inhibition of cellular DNA synthesis in any of the cancer cell lines (Fig. 5 A-C). Addition of increasingAttorney Docket No.: 5025-0446PWO1 concentrations of hBG enzymes produced progressively greater inhibition of cellular DNA synthesis.20 to 40 ng mL-1hBG were required to generate sufficient SN-38 to inhibit 50% of DNA synthesis in the cancer cells (Table 8). By contrast, addition of about 5 ng mL-1hBG3 and 1-2 ng mL-1hBG10 or hBG19 inhibited 50% DNA synthesis in the cancer cells, corresponding to about 5-fold less hBG3 and 15- to 30-fold less hBG10 and hBG19 as compared to hBG for similar anticancer effects (Fig.5 A-D), consistent with enhanced catalytic activity of the hBG variants for hydrolysis of SN-38G to SN-38. For all cells examined, hBG10 produced superior anticancer activity in the presence of SN-38G. Enzyme EC50value (ng mL-1) Cell line SN-38G (nM)hBG hBG3hBG10hBG19362ns of hBG enzymes (EC50 values) that produce 50% inhibition of DNA synthesis in HT-29, MCF- 7 and LS174T cancer cells in the presence of non-toxic concentrations of SN-38G.

[0174] Enzyme stability and in vivo half-life

[0175] Introduction of mutations into recombinant proteins can dramatically affect their stability. However, measurement of enzyme melting temperatures by differential scanning fluorimetry showed that hBG variants displayed lower but still favorable melting temperatures as compared to eBG (Fig.6A). Melting temperature is the temperature that results in unfolding of half of a protein sample, which indicates the thermal stability of proteins. It is notable that hBG19 displayed good thermal stability even though 19 amino acid changes were introduced into this variant. TOhe functional stability of the enzymes was also examined by incubation for nine days at 37oC in PBS or human serum. hBG retains full enzymatic activity over 9 days in PBS and human serum, reflecting the high stability of this lysosomal enzyme (Fig.6B, C). The hBG variants are stable in PBS, but display varying loss of enzymatic activity in human serum; hBG3, hBG10 and hBG19 retain about 50%, 40% and 60% activity after 9 days (Fig. 6C). Both eBG and eBG-NS rapidly lose enzymatic in human serum but eBG-NS is less stable than eBG in PBS, likely due to the presence of an amino acid substitution to remove a cryptic N- linked glycosylation site that maintains enzyme activity

[0035] . hBG10 and hBG19 also display good stability in serum from C57BL / 6 and NOD / SCID mice (Fig.21).Attorney Docket No.: 5025-0446PWO1

[0176] The pharmacokinetic properties of the enzymes were examined by i.v. injecting C57BL / 6 mice with a single dose of 100 μg enzyme spiked with the same enzyme radiolabeled with125I (Fig. 6D). All the enzymes displayed qualitatively similar elimination kinetics in blood with half-lives ranging from 4.2 h for hBG10 up to 6.6 h for hBG3 (Fig.6E, Table 9). hBG10 displayed similar initial elimination as hBG but both hBG3 and hBG19 appeared to be more slowly eliminated from the serum of mice. We also compared injection of a 10-fold lower dose of hBG10 and eBG-NS in BALB / c mice and found that they displayed comparable half- lives of 4.8 and 5.3 h, respectively (Fig. 6F). We conclude that the recombinant enzymes display generally similar elimination kinetics in mice and retain good stability under physiological conditions. Dose (ug) t1 / 2 (h)eBG-NS hBG hBG3 hBG10 hBG19

[0177] Enzyme immunogenicity

[0178] Transgenic C57BL / 6 hBG mice were generated to study the immunogenicity of hBG variants in a model that mimics humans, who are tolerant to hBG. All recombinant enzymes, including eBG-NS, were produced in mouse 3T3 fibroblasts to eliminate the possibility of bacterial contamination in the enzyme preparations. Mice received weekly intravenous injections of 100 µg of each enzyme, and antibodies against each enzyme were measured by ELISA one week later (Fig.7A). eBG-NS induced a strong antibody response in C57BL / 6-Tg- hBG mice, generating anti-eBG-NS IgG levels exceeding 10 µg mL-1after 5 weekly injections. hBG generated low IgG responses of about 0.1 µg mL-1indicating that the transgenic mice are partially tolerant to hBG. As expected, hBG3 was more immunogenic than hBG since this enzyme has three amino acid substitutions that differ from hBG. However, both hBG10 and hBG19 generated significantly lower levels of anti-enzyme IgG antibodies than hBG even though these enzymes contain ten and nineteen amino acids that differ from hBG, respectively (Fig.7B).

[0179] We also tested the enzymes in BALB / c mice, which generate stronger antibody responses as compared to C57BL / 6 mice

[0036] . hBG generated about two-orders lower antibody responses in BALB / c-Tg-hBG mice (Fig. 7C) as compared to BALB / c wild-type mice (Fig. 7D), indicating partial but incomplete tolerance of BALB / c-Tg-hBG to hBG. By contrast, eBG-Attorney Docket No.: 5025-0446PWO1 NS generated strong antibody responses (nearly 1000 µg mL-1) in both transgenic and wild- type mice (Fig.7C, D). hBG3 was 10-fold more immunogenic than hBG in BALB / c-Tg-hBG mice due to the presence of three amino acid substitutions (Fig.7C). However, hBG19 did not generate significantly greater levels of anti-enzyme antibodies as compared to hBG and hBG10 generated significantly lower levels of anti-enzyme IgG as compared to hBG in BALB / c-Tg- hBG mice (Fig.7C). hBG10 also generated about 3-fold lower antibody responses as compared to hBG19. All hBG enzymes (hBG, hBG3, hBG10, hBG19) are equally well-recognized by the anti-hBG antibody used to generate standard curves in enzyme immunoassays (Fig.22A), indicating that differences in calculated serum antibody levels are not due to aberrant standard curves to hBG variants. We also compared titers of anti-enzyme antibodies in BALB / c-Tg- hBG mice (Fig. 22B) and wild type BALB / c mice (Fig. 22C). In agreement with antibody concentrations, anti-enzyme IgG titers against hBG10 were as low as hBG titers in BALB / c- Tg-hBG mice. Anti-hBG10 antibody titers in BALB / c-Tg-hBG mice were more than 100-fold lower than anti-eBG titers. Taken together, these results show that hBG10 exhibits surprisingly low immunogenicity in a mouse model designed to mimic expected antibody responses in humans.

[0180] Compared to hBG, hBG10 has higher enzymatic activity as well as new glycan at N484. To investigate if either enzymatic activity or glycosylation contribute to the low immunogenicity of hBG10 in transgenic-hBG mice, we generated hBG10xA with low enzyme activity by mutation of an active site glutamic acid to alanine (E540A) and hBG10xG, in which the cryptic glycosylation site was removed by reversion of threonine 484 back to asparagine (Fig. 8A). Both hBG10xA and hBG10xG displayed significantly greater (5 to 10-fold) immunogenicity in BALB / c-Tg-hBG mice as compared to hBG10 (Fig. 8B), indicating that both enzymatic activity and the presence of a glycan chain near the active site contribute to the low immunogenicity of hBG10.

[0181] hBG immunoenzymes targeting TAG-72

[0182] A humanized single-chain antibody fragment (hcc49 scFv) was fused to the N-terminus of hBG, hBG3, hBG10 and hBG19 to form immunoenzymes that can recognize the TAG72 tumor antigen expressed on many cancer cells

[0037] (Fig. 9A). A negative control immunoenzyme (DNS-hBG) was also created by fusing a single-chain antibody against a small molecule hapten to hBG. Stable mammalian producer cells were generated for each immunoenzyme, which were subsequently purified by affinity chromatography from the culture medium of the producer cells (Fig. 9B). Antigen-binding activity of the immunoenzymes was examined by direct ELISA against bovine submaxillary gland mucin,Attorney Docket No.: 5025-0446PWO1 which contains the sialyl-Tn epitope recognized by hcc49

[0038] . All the immunoenzymes bound to mucin but not to plates coated with bovine serum albumin, indicating that immunoenzymes retained selectivity for TAG-72 (Fig. 9C). The immunoenzymes displayed sub-nanomolar equilibrium disassociation constants (3-4.5x10-10M) for binding to bovine submaxillary gland mucin as measured by surface plasmon resonance, indicating high binding avidity of immunoenzymes to TAG-72 (Table 10). Immunoenzyme binding to cancer cells was examined by performing fluorescence-activated cell sorting. The immunoenzymes displayed low binding to HT-29 human colon carcinoma cells, moderate binding to LS174T human colon cancer cells, but high binding to MCF-7 human breast cancer cells (Fig. 9D), in line with the relative expression of TAG-72 antigen on these cells

[0031] . Immunoenzyme ka (M-1s-1) kd (s-1) KD (M) hcc49-hBG 106x106314x10-430x10-10

[0183] Immunoenzyme functional activity was examined by adding a constant amount of each immunoenzyme to TAG-72+cancer cells, washing non-bound immunoenzyme and then adding serial dilutions of SN-38G to the cells. Immunoenzyme potency mirrored TAG-72 expression (MCF-7 > LS174T > HT-29 cells). hcc49-hBG10 was the most active immunoenzyme, requiring significantly lower SN-38G concentrations as compared to hcc49-hBG to produce 50% inhibition of DNA synthesis in all three cancer cells (10- to 40-fold lower SN38G concentrations) (Fig.10A). hcc49-hBG19 also required 20-fold less SN-38G than hcc49-hBG to inhibit cancer cells in LS174T and MCF-7 cells while hcc49-hBG3 was only more effective than hcc49-hBG when tested against MCF-7 cells, which express the highest levels of TAG- 72. Similar results were observed when serial dilutions of the immunoenzymes were added to LS-174T cancer cells, followed by washing and addition of non-toxic concentrations of SN- 38G; immunoenzyme potency following the order hcc49-hBG10 > hcc49-hBG19 > hcc49- hBG3 > hcc49-hBG (Fig. 10B). We created an additional control immunoenzyme (DNS- hBG10) by fusing hBG10 to the DNS scFv. DNS-hBG and DNS-hBG10 at concentrations up to 10 ug mL-1produced minimal inhibition of LS174T cells (Fig. 10B). ImmunofluorescenceAttorney Docket No.: 5025-0446PWO1 staining confirmed similar and selective binding of hcc49-hBG and hcc49-hBG10 to LS174T tissue sections (Fig.10C). hcc49-BG also strongly stained MCF-7 tumor sections (Fig.10C).

[0184] In vivo antitumor activity of hcc49-hBG10

[0185] hcc49-hBG displayed a half-life ranging from 3.7 to 5.9 h after intravenous administration of doses of 10-1000 µg into NOD-SCID mice (Fig 11A). The half-life of hcc49- hBG was slightly shorted (ranging from 3.4 to 4.8 h) in NOD-SCID mice bearing LS174T tumor xenografts, possibly due to uptake of immunoenzyme in tumors (Fig. 11A). We examined accumulation of functional immunoenzyme in LS174T tumors by injecting vehicle, control DNS-hBG10 or hcc49-hBG1096 h before intravenous administration of FDGlcU, a non-fluorescent glucuronide substrate that can be hydrolyzed by beta-glucuronidase into the fluorescent compound fluorescein (Fig.11B). Thirty minutes after administration of FDGlcU, tumors, lung, and liver tissue were excised from mice and imaged on an IVIS imaging system. No obvious fluorescence was observed in lung and liver tissue but tumors from mice treated with hcc49-hBG10 and FDGlcU displayed strong fluorescence signals, indicating that functional hBG10 was present in tumors for at least 96 h after immunoenzyme administration (Fig. 11C). hBG10 was not evolved to hydrolyze FDGlcU during variant selection, so this assay likely underestimates function hBG10 activity in tumors for hydrolysis of SN-38G to SN-38. Tumors from mice treated with control immunoenzyme (DNS-hBG10) and FDGlcU also displayed fluorescence signals, but this level of fluorescence was similar to mice treated with PBS and FDGlcU (Fig. 11D), consistent with the presence of endogenous beta- glucuronidase in tumors [39, 40].

[0186] We performed a preliminary in vivo experiment in groups of three NOD-SCID mice bearing LS174T tumors by injecting vehicle or 200 µg hcc49-BG10 followed 72 h and 96 h later by i.v. injections of 25 mg kg-1SN-38G or i.p injections of 50 mg kg-1CPT-11 (Fig.12A). SN-38G was included as a positive control because it was unclear if sufficient SN-38G would be generated by the metabolism of CPT-11 in mice. However, treatment of mice with hcc49- hBG and CPT-11 produced superior antitumor activity as compared to treatment with hcc49- hBG10 and SN-38G (Fig. 12B), suggesting that continuous generation of low levels of SN- 38G is better than two injections of SN-38G. We therefore tested hcc49-hBG10 therapy in groups of six NOD-SCID mice bearing established LS174T tumors (Fig.12C). Mice were i.v. injected with vehicle (PBS + 6% DMSO) or 200 μg of immunoenzymes on days 0, 10, 20, 30 and 40. The mice were also intraperitoneally injected with 50 mg kg-1CPT-11 at 72 and 96 h after each vehicle or immunoenzyme injection. Treatment of mice with CPT-11 delayed tumor growth but pretreatment of mice with hcc49-hbG10 before CPT-11 injections provided greaterAttorney Docket No.: 5025-0446PWO1 delay of tumor growth (Fig. 12D) and significantly increased mice survival (Fig. 12F). Treatment of mice with hcc49-hBG10 and CPT-11 significantly prolonged mouse survival as compared to DNS-hBG10 and CPT-11, demonstrating that selective tumor binding of the immunoenzyme was required for activity. In addition, hcc49-hBG10 and CPT-11 significantly prolonged survival as compared to hcc49-hBG and CPT-11, indicating the high enzyme activity of hBG10 enhanced anticancer activity. CPT-11 slightly decreased mouse body weight, but pretreatment with hcc49-BG10 before CPT-11 administration did not produce additional loss of body weight (Fig.12E).

[0187] On the other hand, to investigate converting activity of HAMG-ester by hcc49-hBG10, LS174T human colon tumor bearing mice were treated 200μg hcc49-hBG10 once by i.v. injection and then treated once 20, 25, 30 mg / kg HAMG-ester to mice after injecting hcc49- hBG10 for 72 h and 96 h. Compare to PBS control group, tumor size of mice can maintain around 1000 mm3until Day 15 after receiving 20 mg / kg HAMG-ester (Fig.24A). The tumor size of LS174T tumor bearing mice have no different after treating different dose of HAMG- ester indicating retardation of tumor growth after treating 20 mg / kg HAMG-ester (Fig.24A). Besides, change of body weight in mice dropped around 5-10% after prodrug treatment as compared to vehicle groups, eventually the body of mice recovered at Day 20 (Fig.24B).

[0188] Taken together, these results show that hcc49-hBG10 can enhance the therapeutic index of CPT-11 and HAMG-ester against LS174T colorectal cancer in mice.

[0189] 3. Dicussion

[0190] Human enzymes are attractive for many therapeutic applications, but may require engineering to optimize their properties to achieve in vivo efficacy. Amino acid changes may adversely affect enzyme stability and create new immunogenic epitopes that generate antibody responses. In this study, we investigated if we could engineer human beta-glucuronidase (hBG) to activate a metabolite of irinotecan in the tumor microenvironment while maintaining acceptable protein stability and immunogenicity. We used directed molecular evolution on the ECSTASY platform to derive engineered hBG variants that display about 40-fold enhanced activity for the enzymatic conversion of SN-38G to SN38, maintain substantial stability and enzymatic activity under physiological conditions, display less immunogenicity as compared to eBG, and can be used to construct immunoenzymes that enhance the antitumor activity of CPT-11 in a mouse model of human colon cancer.

[0191] Development of immunoenzymes that can selectively deliver therapeutic enzymes to tumor cells is attractive for cancer therapy because the enzyme portion can catalytically generate active drug molecules, thereby amplifying the effect of each targeted antibody

[0041] .Attorney Docket No.: 5025-0446PWO1 Active drug generated at antigen-positive cancer cells can also diffuse throughout the tumor to effectively kill less assessable as well as antigen-negative cancer cells

[0042] . Most studies have focused on targeting nonhuman enzymes to tumors to take advantage of the inherently high activity of many microbial enzymes and the possibility of selecting enzymes with nonhuman homologs to increase treatment specificity [4, 8]. Indeed, we have shown that immunoenzymes incorporating eBG can cure human tumor xenografts in rodent models due to the exceptional catalytic activity of eBG at neutral pH values [43-45]. However, the use of microbially-derived immunoenzymes in human patients is problematic due to concerns of high immunogenicity and adverse reactions [7, 9-12, 46]. Development of human immunoenzymes is problematic due to toxicity associated with systemic activation of administered prodrugs by endogenous enzymes in patients

[0047] . Although hBG is expressed in many cells, it is located in lysosomes and is almost undetectable in human serum

[0048] . Glucuronide prodrugs are typically very hydrophilic and therefore pass poorly through the plasma membrane of cells [28, 49]. Thus, glucuronide prodrugs and hBG do not normally come into contact, reducing concerns about premature activation of glucuronide prodrugs by endogenous hBG. Non-internalizing hBG immunoenzyme that remain on the outside surface of cancer cells can therefore display selective activation of glucuronide prodrugs.

[0192] In the present disclosure, we overcame several hurdles to engineer hBG for immunoenzyme therapy. Because hBG is a lysosomal enzyme, it displays maximum catalytic activity at acidic pH values (~4.5), but much reduced activity at the pH found in the tumor extracellular space (6.5~7.0) [50, 51]. The lysosomal location of hBG also makes screening problematic because it is difficult for hydrophilic substrates to pass into cells and cross the lysosomal membrane and it is also hard to precisely control the pH inside lysosomes. hBG is also a tetrameric glycoprotein, requiring N-linked glycosylation for proper folding and activity

[0052] . Due to these constraints, we used our previously developed screening platform called ECSTACY to engineer hBG variants. ECSTASY displays enzymes on the outer leaflet of the plasma membrane of mammalian cells via a GPI anchor. Retroviral transduction of cells at a low multiplicity of infection allows generation of large cell libraries with multiple copies of a single enzyme variant expressed on each cell. Because enzyme variants must pass through the secretory pathway to reach the plasma membrane, mis-folded proteins are detected and degraded in the endoplasmic reticulum (ER) of mammalian cells

[0053] . Enzymes are also glycosylated as they pass through the ER and golgi apparatus to maintain their native structures. Rapid selection of cells that express high levels of enzymes on their surface by flow cytometry may therefore select for inherently stable enzyme variants. Individual cells gated into 96-wellAttorney Docket No.: 5025-0446PWO1 plates can be grown and then treated with PI-PLC to cleave the GPI linkage and release soluble enzyme into the culture medium for screening of enzyme activity and enzyme quantity under defined conditions using standard assays.

[0193] We used ECSTASY to select hBG variants that could hydrolyze SN-38G to SN-38 at neutral pH by screening five individual hBG libraries encompassing a range of amino acid substitutions at 49 positions based on comparison of mammalian and bacterial BG protein sequences. After combining clones selected from each library and performing DNA shuffling, we further investigated two clones that possess ten (hBG10) and nineteen (hBG19) amino acid substitutions in the original hBG sequence. A hBG variant selected in a previous study (hBG3) was included for comparison

[0022] . hBG10 and hBG19 were about 40-fold more potent than hBG to hydrolyze SN-38G to SN-38. hBG10 and hBG19 display lower Km values as compared to hBG (0.13 and 0.019 mM vs. 0.55 mM at pH 4.5, respectively), consistent with estimated smaller binding cavities in hBG10 (936 Å3) and hBG19 (596 Å3) compared to hBG (1434 Å3) as well as new interactions between SN-38G and N486T in hBG10 and N484F and H509Y in hBG19.

[0194] Engineered human enzyme stability can be adversely affected by introduction of amino acid substitutions. For example, changing a single amino acid (T268G) in human carboxypeptidase A1 to enhance prodrug activation for ADEPT resulted in 90% loss of activity over 72 h

[0054] . In work selecting human asparaginase variants, the variant with the highest activity (with two amino acid substitutions) had the lowest stability, although it was still quite good as measured by melting temperature

[0015] . A human kynureninase was evolved with 25 amino acid substitutions, resulting in an impressive 500-fold enhancement of catalytic activity, but about 98% enzyme activity was lost within 24 h in 50% human serum

[0017] . By contrast, hBG10 and hBG19 displayed good thermal stability and retained about 50% of their catalytic activity after 9 days incubation at 37oC in human serum. We speculate that unstable enzymes are degraded in the ER; mutations that destabilize enzyme structure result in hindered transport to the plasma membrane and are eliminated during flow cytometric screening. The lack of amino acid substitutions at 15 of 47 of the positions of hBG is consistent with strong selection pressure to eliminate enzyme variants with destabilizing or inactivating mutations. ECSTASY thus appears to inherently apply selection pressure to maintain protein stability.

[0195] A strong impetus for developing therapeutic human enzymes is the possibility of reducing immunogenicity associated with the use of microbial enzymes. Microbial enzymes are recognized as foreign substances in humans, and their therapeutic use can lead to adverse reactions in patients [7, 9-12, 55, 56] Replacement of microbial enzymes with engineeredAttorney Docket No.: 5025-0446PWO1 human enzymes is attractive, but it remains unclear if strong immune responses will still be generated against new amino acid sequences generated during the engineering process. To investigate immunogenicity of hBG variants, we generated mice that express hBG to act as a model of human patients who are tolerant to wild type hBG but may be sensitive to engineered hBG variants. We generated both traditional C57BL / 6-Tg-hBG homozygous mice as well as BALB / c-Tg-hBG homozygous mice. C57BL / 6 mice tend to display a predominantly Th1 immune response whereas BALB / c mice typically exhibit stronger Th2 immune responses

[0057] . This corresponds to stronger cytotoxic T cell responses in C57BL / 6 mice and higher humoral immune responses in BALB / c mice [58, 59]. We observed ten-to-one-hundred-fold stronger antibody responses in BALB / c mice as compared to C57BL / 6 mice after administration of eBG-NS, indicating that BALB / c mice are the more sensitive model to study humoral immune responses against foreign antigens.

[0196] Both BALB / c-Tg-hBG and C57BL / 6-Tg-hBG mice were mostly but not complexly tolerant to hBG. This may be due to the presence of a His tag present at the C-terminus of all of the recombinant enzymes as well as to low levels of impurities present in the preparations. All enzymes, including eBG-NS, were produced in mouse 3T3 fibroblasts to reduce the possibility of endotoxin contamination and provide similar contamination backgrounds. An unexpected finding of our study is that hBG10 is significantly less immunogenic than hBG in both BALB / c-Tg-hBG and C57BL / 6-Tg-hBG mice. hBG19 also displayed similar immunogenicity as hBG, even though it contains 19 amino acid differences. One possible mechanism of low immunogenicity is if the amino acid substitutions had homology to mouse beta-glucuronidase. However, only one of three, one of ten, and three of nineteen amino acids substitutions in hBG3, hBG10, and hBG19 matched mBG sequences (Table 6), indicating that this had a minor effect on the observed immunogenicity. Differences in the in vivo pharmacokinetics of the enzymes might also affect immunogenicity, but all the enzymes displayed similar elimination kinetics from the serum of mice. hBG10 contains a cryptic N- linked glycosylation site at N484, which was confirmed to be glycosylated by LC / MS-MS. Removal of the new glycosylation site resulted in significantly greater antibody responses against hBG10xG, indicating that glycosylation is involved in the low immunogenicity of hBG10. This is reminiscent of glycan masking, in which glycans are engineered into viral proteins to shield epitopes and direct humoral antibody responses to new unshielded epitopes [60-62]. Our results suggest that purposely introducing glycans near amino acid substitutions in engineered proteins may be a powerful approach to reduce antibody recognition. hBG10xA, in which an active site glutamic acid was mutated to glycine to block enzyme activity, alsoAttorney Docket No.: 5025-0446PWO1 significantly increased enzyme immunogenicity. A role for enzyme activity is also supported by the relatively low immunogenicity of hBG19, which does not have an additional glycan but displays similar immunogenicity as hBG, even though it possesses 19 amino acid differences with wild type hBG. We speculate that hydrolysis of glucuronide-related hormones or other glucuronides in blood may generate immunosuppressive metabolites that act locally to reduce generation of antibodies.

[0197] In the present disclosure, we created immunoenzymes to selectively enhance the activity of CPT-11. hBG variants were fused to a single-chain antibody fragment derived from hcc49, a high-affinity humanized antibody which binds TAG-72 (tumor-associated glycoprotein 72), a truncated sialyl-Tn antigen found on many O-linked glycoproteins

[0063] . TAG-72 is expressed on most colorectal, gastric, pancreatic, non-small cell lung, ovarian, endometrial, breast and prostate carcinomas tumors

[0037] . For example, CC49 localized in 86% of patients with primary colorectal carcinoma and 97% of patients with recurrent colorectal carcinoma

[0064] . TAG-72 targeted CAR T cells, radionuclide-labeled antibodies, and antibody- drug conjugates are under development for the treatment of cancer [65-69]. TAG-72 does not undergo receptor-mediated endocytosis upon antibody binding, which is advantageous for ADEPT to allow targeted immunoenzyme to remain assessable to SN-38G in the tumor environment

[0031] . TAG-72 was also selected because CPT-11 is used to treat patients suffering from colorectal carcinoma.

[0198] The hcc49-hBG immunoenzymes display high avidity for sialyl-Tn in bovine submaxillary gland mucin (KD~ 4x10-10M), likely due to the presence of four hcc49 scFv in one hBG tetramer [31, 34]. hcc49-hBG10 and hcc49-hBG19 were about 40-fold more potent than hcc49-hBG to hydrolyze SN-38G to SN-38 and kill cancer cells that express high levels of TAG-72. The relatively low Km values of hBG10 and hBG19 may be important to enhance the therapeutic activity of CPT-11 because only a portion of CPT-11 is metabolized to SN- 38G. The maximum concentrations of SN-38G in humans receiving CPT-11 is around 0.5~1 µM, depending on the administered dose of CPT-11 [24, 70]. hBG10 and hBG19 display about 5-fold greater enzyme activities (kcat / Km) at pH 6.5 as compared to pH 7.0. The tumor microenvironment tends to be slightly acidic due to increased production of lactic acid due to the Warburg effect [32, 33]. By contrast, the normal pH of blood is about 7.4. This suggests that immunoenzymes will be substantially more active after localization in tumors as compared to immunoenzyme remaining in the blood circulation or located in normal organs. This is consistent with lack of strong signals present in lung or liver tissues after administration of hcc49-hBG10 and a glucuronide imaging probe. Due to the lower immunogenicity of hBG10Attorney Docket No.: 5025-0446PWO1 as compared to hBG19, we choose to further investigate the in vivo anticancer activity of hcc49- hBG10.

[0199] Comparison of the antitumor activity of hcc49-hBG10 with subsequently administered SN-38G or CPT-11 in a mouse model of human colorectal cancer showed that CPT-11 produced superior activity as compared to SN-38G. We speculate that SN-38G is rapidly eliminated from the blood whereas metabolism of CPT-11 generates a continuous supply of SN-38G over a prolonged period. Rapid elimination of SN-38G is consistent with the pharmacokinetics of hydrophilic drugs and a 500-fold drop in serum concentration within 12 h of BQC-G, a glucuronide conjugate of a camptothecin derivative with similar structure as SN- 38G

[0071] . By contrast, the concentration of SN-38G only decreases by a few-fold after 24 hours of administration of CPT-11, and can be detected in serum for more than 500 h [24, 72]. Our results suggest that continuous in situ generation of SN-38G and subsequent hydrolysis of SN- 38G by hcc49-hBG10 in the tumor microenvironment is important to achieve maximum antitumor activity

[0073] .

[0200] CPT-11 metabolism differs between mice and humans. In mice, more SN-38 and less SN-38G is generated as compared to humans. The ratio of serum SN-38G / SN-38 area under the curves (AUC) is reported to be 0.3 in mice and 0.9 in rats [74, 75]. On the other hand, the serum SN-38G / SN-38 AUC in human patients receiving CPT-11 averages around 7 and can exceed 15-fold during the first 5 to 10 hours after CPT-11 administration [24, 70, 72]. The larger amount of SN-38G available for immunoenzyme activation suggests that immunoenzymes may display better activity in human patients treated with CPT-11 as compared to mice.

[0201] In conclusion, our study demonstrates that hBG variants can be developed that display enhanced activity to convert SN-38G to SN38 while retaining good stability and low immunogenicity. Targeting hBG10 to colon cancer tumors significantly enhances the activity of CPT-11 in mice. Engineering human enzymes appears to be a promising approach for the treatment of many diseases.

[0202] 4. Acknowledgements

[0203] This study was supported by a grant from the National Science and Technology Council, Taipei, Taiwan (112-2320-B-001-003). We are grateful to the Proteomic Core Facility and Experimental Animal Facility in the Institute of Biomedical Sciences (IBMS), Academia Sinica for experimental assistance. We acknowledge the Academia Sinica Core Facility and Innovative Instrument Project (AS-CFII-111-212) for cell sorting service and analysis of flow cytometry. We thank the Biophysics Core Facility for assistance to use the Biacore T200. WeAttorney Docket No.: 5025-0446PWO1 also thank the National Laboratory Animal Center to preserve genetically modified hBG mouse lines, and also especially thank Chia-Chi Lin for construction of hBG mouse. The graphics in this manuscript were created using Biorender.

[0204] 5. References [1] M. Solomon, S. Muro, Lysosomal enzyme replacement therapies: Historical development, clinical outcomes, and future perspectives, Adv. Drug Deliv. Rev, 118 (2017) 109-134. [2] C. Riess, F. Shokraie, C.F. Classen, B. Kreikemeyer, T. Fiedler, C. Junghanss, C. Maletzki, Arginine-depleting enzymes - An increasingly recognized treatment strategy for therapy refractory malignancies, Cell. Physiol. Biochem, 51 (2018) 854-870. [3] P. Cioni, E. Gabellieri, B. Campanini, S. Bettati, S. Raboni, Use of exogenous enzymes in human therapy: approved drugs and potential applications, Curr. Med. Chem., 29 (2022) 411-452. [4] S.K. Sharma, K.D. Bagshawe, Translating antibody directed enzyme prodrug therapy (ADEPT) and prospects for combination, Expert Opin Biol Ther, 17 (2017) 1-13. [5] T.A. Triplett, K.C. Garrison, N. Marshall, M. Donkor, J. Blazeck, C. Lamb, A. Qerqez, J.D. Dekker, Y. Tanno, W.C. Lu, C.S. Karamitros, K. Ford, B. Tan, X.Y.M. Zhang, K. McGovern, S. Coma, Y. Kumada, M.S. Yamany, E. Sentandreu, G. Fromm, S. Tiziani, T.H. Schreiber, M. Manfredi, L.I.R. Ehrlich, E. Stone, G. Georgiou, Reversal of indoleamine 2,3- dioxygenase-mediated cancer immune suppression by systemic kynurenine depletion with a therapeutic enzyme, Nat. Biotechnol., 36 (2018) 758. [6] P.A. Burnouf, S.R. Roffler, C.C. Wu, S. Yu Cheng, Glucuronides: From biological waste to bio-nanomedical applications, J. Control. Release, 349 (2022) 765-782. [7] D.A. Brooks, R. Kakavanos, J.J. Hopwood, Significance of immune response to enzyme- replacement therapy for patients with a lysosomal storage disorder, Trends Mol. Med., 9 (2003) 450-453. [8] S.K. Sharma, K.D. Bagshawe, Antibody directed enzyme prodrug therapy (ADEPT): Trials and tribulations, Adv. Drug Deliv. Rev, 118 (2017) 2-7. [9] E.D. Kakkis, M.F. McEntee, A. Schmidtchen, E.F. Neufeld, D.A. Ward, R.E. Gompf, S. Kania, C. Bedolla, S.-L. Chien, R.M. Shull, Long-term and high-dose trials of enzyme replacement therapy in the canine model of mucopolysaccharidosis I, Biochem Mol Med., 58 (1996) 156-167.

[0010] K.P. Ponder, Immune response hinders therapy for lysosomal storage diseases, J. Clin. Investig., 118 (2008) 2686-2689.Attorney Docket No.: 5025-0446PWO1

[0011] V.I. Avramis, P.N. Tiwari, Asparaginase (native ASNase or pegylated ASNase) in the treatment of acute lymphoblastic leukemia, Int. J. Nanomedicine, 1 (2006) 241-254.

[0012] S.Q. Le, S.H. Kan, D. Clarke, V. Sanghez, M. Egeland, K.N. Vondrak, T.M. Doherty, M.U. Vera, M. Iacovino, J.D. Cooper, M.S. Sands, P.I. Dickson, A humoral immune response alters the distribution of enzyme replacement therapy in murine mucopolysaccharidosis type I, Mol. Ther. Methods Clin. Dev., 8 (2018) 42-51.

[0013] H.Y. Chuang, C.S. Suen, M.J. Hwang, S.R. Roffler, Toward reducing immunogenicity of enzyme replacement therapy: altering the specificity of human β-glucuronidase to compensate for α-iduronidase deficiency, Protein Eng. Des. Sel., 28 (2015) 519-530.

[0014] S. Afshar, T. Olafsen, A.M. Wu, S.L. Morrison, Characterization of an engineered human purine nucleoside phosphorylase fused to an anti-her2 / neu single chain Fv for use in ADEPT, J. Exp. Clin. Cancer Res., 28 (2009) 1-12.

[0015] C.S. Karamitros, M. Konrad, Fluorescence-activated cell sorting of human l- asparaginase mutant libraries for detecting enzyme variants with enhanced activity, ACS Chem. Biol., 11 (2016) 2596-2607.

[0016] C. Rigouin, H.A. Nguyen, A.M. Schalk, A. Lavie, Discovery of human-like L- asparaginases with potential clinical use by directed evolution, Sci. Rep., 7 (2017) 10224.

[0017] J. Blazeck, C.S. Karamitros, K. Ford, C. Somody, A. Qerqez, K. Murray, N.T. Burkholder, N. Marshall, A. Sivakumar, W.-C. Lu, Bypassing evolutionary dead ends and switching the rate-limiting step of a human immunotherapeutic enzyme, Nat. Catal., 5 (2022) 952-967.

[0018] C.S. Karamitros, M.l. Morvan, A. Vigne, J. Lim, P. Gruner, T. Beneyton, J. Vrignon, J.-C. Baret, Bacterial expression systems for enzymatic activity in droplet-based microfluidics, Anal. Chem., 92 (2020) 4908-4916.

[0019] U. Boonyuen, K. Chamchoy, T. Swangsri, T. Junkree, N.P.J. Day, N.J. White, M. Imwong, A trade off between catalytic activity and protein stability determines the clinical manifestations of glucose-6-phosphate dehydrogenase (G6PD) deficiency, Int. J. Biol. Macromol., 104 (2017) 145-156.

[0020] R. Mazor, E.M. King, I. Pastan, Strategies to reduce the immunogenicity of recombinant immunotoxins, Am. J. Pathol., 188 (2018) 1736-1743.

[0021] C.P. Chen, Y.T. Hsieh, Z.M. Prijovich, H.Y. Chuang, K.C. Chen, W.C. Lu, Q. Tseng, Y.L. Leu, T.L. Cheng, S.R. Roffler, ECSTASY, an adjustable membrane-tethered / soluble protein expression system for the directed evolution of mammalian proteins, Protein Eng. Des. Sel., 25 (2012) 367-375.Attorney Docket No.: 5025-0446PWO1

[0022] K.C. Chen, C.H. Wu, C.Y. Chang, W.C. Lu, Q.Z. Tseng, Z.M. Prijovich, W. Schechinger, Y.C. Liaw, Y.L. Leu, S.R. Roffler, Directed evolution of a lysosomal enzyme with enhanced activity at neutral pH by mammalian cell-surface display, Chem. Biol., 15 (2008) 1277-1286.

[0023] R. Humerickhouse, K. Lohrbach, L. Li, W.F. Bosron, M.E. Dolan, Characterization of CPT-11 hydrolysis by human liver carboxylesterase isoforms hCE-1 and hCE-2, Cancer res., 60 (2000) 1189-1192.

[0024] A. Sparreboom, M. De Jonge, P. de Bruijn, E. Brouwer, K. Nooter, W.J. Loos, R.J. van Alphen, R. Mathijssen, G. Stoter, J. Verweij, Irinotecan (CPT-11) metabolism and disposition in cancer patients, Clin. Cancer Res., 4 (1998) 2747-2754.

[0025] K.C. Chen, T.L. Cheng, Y.L. Leu, Z.M. Prijovich, C.H. Chuang, B.M. Chen, S.R. Roffler, Membrane-localized activation of glucuronide prodrugs by beta-glucuronidase enzymes, Cancer Gene Ther., 14 (2007) 187-200.

[0026] M. Mirdita, K. Schütze, Y. Moriwaki, L. Heo, S. Ovchinnikov, M. Steinegger, ColabFold: making protein folding accessible to all, Nat. Methods, 19 (2022) 679-682.

[0027] Y. Liu, X. Yang, J. Gan, S. Chen, Z.-X. Xiao, Y. Cao, CB-Dock2: Improved protein– ligand blind docking by integrating cavity detection, docking and homologous template fitting, Nucleic Acids Res., 50 (2022) 159-164.

[0028] T.L. Cheng, W.C. Chou, B.M. Chen, J.W. Chern, S.R. Roffler, Characterization of an antineoplastic glucuronide prodrug, Biochem. Pharmacol., 58 (1999) 325-328.

[0029] T. Čaval, J. Zhu, W. Tian, S. Remmelzwaal, Z. Yang, H. Clausen, A.J. Heck, Targeted analysis of lysosomal directed proteins and their sites of mannose-6-phosphate modification, Mol. Cell Proteomics, 18 (2019) 16-27.

[0030] H. Niwa, K. Yamamura, J. Miyazaki, Efficient selection for high-expression transfectants with a novel eukaryotic vector, Gene, 108 (1991) 193-199.

[0031] K.C. Chen, S.Y. Wu, Y.L. Leu, Z.M. Prijovich, B.M. Chen, H.E. Wang, T.L. Cheng, S.R. Roffler, A humanized immunoenzyme with enhanced activity for glucuronide prodrug activation in the tumor microenvironment, Bioconjug. Chem., 22 (2011) 938-948.

[0032] L.E. Gerweck, K. Seetharaman, Cellular pH gradient in tumor versus normal tissue: potential exploitation for the treatment of cancer, Cancer Res., 56 (1996) 1194-1198.

[0033] G. Helmlinger, F. Yuan, M. Dellian, R.K. Jain, Interstitial pH and pO2 gradients in solid tumors in vivo: high-resolution measurements reveal a lack of correlation, Nat. Med., 3 (1997) 177-182.Attorney Docket No.: 5025-0446PWO1

[0034] M.I. Hassan, A. Waheed, J.H. Grubb, H.E. Klei, S. Korolev, W.S. Sly, High resolution crystal structure of human β-glucuronidase reveals structural basis of lysosome targeting, PLoS One, 8 (2013) e79687.

[0035] L.B. Farrell, R.N.J.P.m.b. Beachy, Manipulation of β-glucuronidase for use as a reporter in vacuolar targeting studies, Plant Mol. Bio., 15 (1990) 821-825.

[0036] W. Jiskoot, G. Kijanka, T.W. Randolph, J.F. Carpenter, A.V. Koulov, H.C. Mahler, M.K. Joubert, V. Jawa, L.O. Narhi, Mouse models for assessing protein immunogenicity: lessons and challenges, J. Pharm. Sci., 105 (2016) 1567-1575.

[0037] A. Thor, N. Ohuchi, C.A. Szpak, W.W. Johnston, J. Schlom, Distribution of oncofetal antigen tumor-associated glycoprotein-72 defined by monoclonal antibody B72. 3, Cancer res., 46 (1986) 3118-3124.

[0038] F.-G. Hanisch, G. Uhlenbruck, H. Egge, J. Peter-Katalinić, A B72.3 second-generation- monoclonal antibody (CC49) defines the mucin-carried carbohydrate epitope Galβ (1— 3)[NeuAcα (2—6)] GalNAc, Cancer res., (1989).

[0039] U. Schumacher, E. Adam, U. Zangemeister-Wittke, R. Gossrau, Histochemistry of therapeutically relevant enzymes in human tumours transplanted into severe combined immunodeficient (SCID) mice: nitric oxide synthase-associated diaphorase, beta-D- glucuronidase and non-specific alkaline phosphatase, Acta Histochem., 98 (1996) 381-387.

[0040] Y.C. Su, T.C. Cheng, Y.L. Leu, S.R. Roffler, J.Y. Wang, C.H. Chuang, C.H. Kao, K.C. Chen, H.E. Wang, T.L. Cheng, PET imaging of β-glucuronidase activity by an activity-based 124I-trapping probe for the personalized glucuronide prodrug targeted therapy, Mol. Cancer Ther., 13 (2014) 2852-2863.

[0041] H.P. Svensson, V.M. Vrudhula, J.E. Emswiler, J.F. MacMaster, W.L. Cosand, P.D. Senter, P.M. Wallace, In vitro and in vivo activities of a doxorubicin prodrug in combination with monoclonal antibody β-lactamase conjugates, Cancer res., 55 (1995) 2357-2365.

[0042] T. Cheng, S. Wei, B. Chen, J. Chern, M. Wu, P. Liu, S. Roffler, Bystander killing of tumour cells by antibody-targeted enzymatic activation of a glucuronide prodrug, Br. j. cancer, 79 (1999) 1378-1385.

[0043] S.-M. Wang, J.-W. Chern, M.-Y. Yeh, J.C. Ng, E. Tung, S.R. Roffler, Specific activation of glucuronide prodrugs by antibody-targeted enzyme conjugates for cancer therapy, Cancer Res., 52 (1992) 4484-4491.

[0044] B.M. Chen, T.L. Cheng, S.C. Tzou, S.R. Roffler, Potentiation of antitumor immunity by antibody-directed enzyme prodrug therapy, Int. J. Cancer, 94 (2001) 850-858.Attorney Docket No.: 5025-0446PWO1

[0045] T.-L. Cheng, B.-M. Chen, L.-Y. Chan, P.-Y. Wu, J.-W. Chern, S.R. Roffler, Poly (ethylene glycol) modification of β-glucuronidase-antibody conjugates for solid-tumor therapy by targeted activation of glucuronide prodrugs, Cancer Immunol. Immunother., 44 (1997) 305-315.

[0046] J.R. Cantor, V. Panayiotou, G. Agnello, G. Georgiou, E.M. Stone, Engineering reduced- immunogenicity enzymes for amino acid depletion therapy in cancer, Meth. Enzymol., 502 (2012) 291-319.

[0047] M. Rooseboom, J.N. Commandeur, N.P. Vermeulen, Enzyme-catalyzed activation of anticancer prodrugs, Pharmacol. Rev., 56 (2004) 53-102.

[0048] T.-Y. Juan, S.R. Roffler, H.-S. Hou, S.-M. Huang, K.-C. Chen, Y.-L. Leu, Z.M. Prijovich, C.-P. Yu, C.-C. Wu, G.-H. Sun, Antiangiogenesis targeting tumor microenvironment synergizes glucuronide prodrug antitumor activity, Clin. Cancer Res., 15 (2009) 4600-4611.

[0049] K.C. Chen, K. Schmuck, L.F. Tietze, S.R. Roffler, Selective cancer therapy by extracellular activation of a highly potent glycosidic duocarmycin analogue, Mol. Pharm., 10 (2013) 1773-1782.

[0050] X. Zhang, Y. Lin, R.J. Gillies, Tumor pH and its measurement, J. Nucl. Med., 51 (2010) 1167-1170.

[0051] S. Jain, W.B. Drendel, Z.W. Chen, F.S. Mathews, W.S. Sly, J.H. Grubb, Structure of human beta-glucuronidase reveals candidate lysosomal targeting and active-site motifs, Nat. Struct. Mol. Biol., 3 (1996) 375-381.

[0052] J. Shipley, J.H. Grubb, W.S. Sly, The role of glycosylation and phosphorylation in the expression of active human beta-glucuronidase, J. Biol. Chem., 268 (1993) 12193-12198.

[0053] P. Määttänen, K. Gehring, J.J. Bergeron, D.Y. Thomas, Protein quality control in the ER: the recognition of misfolded proteins, Semin. Cell Dev. Biol., 21 (2010) 500-511.

[0054] L.A. Wolfe, R.J. Mullin, R. Laethem, T.A. Blumenkopf, M. Cory, J.F. Miller, B.R. Keith, J. Humphreys, G.K. Smith, Antibody-directed enzyme prodrug therapy with the T268G mutant of human carboxypeptidase A1: in vitro and in vivo studies with prodrugs of methotrexate and the thymidylate synthase inhibitors GW1031 and GW1843, Bioconjug. Chem., 10 (1999) 38-48.

[0055] I.M.v.d. Sluis, L.M. Vrooman, R. Pieters, A. Baruchel, G. Escherich, N. Goulden, V. Mondelaers, J.S.d. Toledo, C. Rizzari, L.B. Silverman, J.A. Whitlock, Consensus expert recommendations for identification and management of asparaginase hypersensitivity and silent inactivation, Haematologica, 101 (2016) 279-285.Attorney Docket No.: 5025-0446PWO1

[0056] A. Guttmann, S. Krasnokutsky, M.H. Pillinger, A. Berhanu, Pegloticase in gout treatment - safety issues, latest evidence and clinical considerations, Ther. Adv. Drug Saf., 8 (2017) 379-388.

[0057] C.D. Mills, K. Kincaid, J.M. Alt, M.J. Heilman, A.M. Hill, M-1 / M-2 macrophages and the Th1 / Th2 paradigm, J. Immunol., 164 (2000) 6166-6173.

[0058] G.V. Trunova, O.V. Makarova, M.E. Diatroptov, I.M. Bogdanova, L.P. Mikchailova, S.O. Abdulaeva, Morphofunctional characteristic of the immune system in BALB / c and C57Bl / 6 mice, Bull. Exp. Biol. Med., 151 (2011) 99-102.

[0059] A. Pellegrini, N. Guinazu, M.P. Aoki, I.C. Calero, E.A. Carrera-Silva, N. Girones, M. Fresno, S. Gea, Spleen B cells from BALB / c are more prone to activation than spleen B cells from C57BL / 6 mice during a secondary immune response to cruzipain, Int. Immunol., 19 (2007) 1395-1402.

[0060] G. Bajic, M.J. Maron, Y. Adachi, T. Onodera, K.R. McCarthy, C.E. McGee, G.D. Sempowski, Y. Takahashi, G. Kelsoe, M. Kuraoka, A.G. Schmidt, Influenza antigen engineering focuses immune responses to a subdominant but broadly protective viral epitope, Cell Host Microbe., 25 (2019) 827-835.e826.

[0061] H. Duan, X. Chen, J.C. Boyington, C. Cheng, Y. Zhang, A.J. Jafari, T. Stephens, Y. Tsybovsky, O. Kalyuzhniy, P. Zhao, Glycan masking focuses immune responses to the HIV- 1 CD4-binding site and enhances elicitation of VRC01-class precursor antibodies, Immunity., 49 (2018) 301-311. e305.

[0062] M. Lavie, X. Hanoulle, J. Dubuisson, Glycan shielding and modulation of hepatitis C virus neutralizing antibodies, Front Immunol. , 9 (2018) 910.

[0063] S.O. Yoon, T.S. Lee, S.J. Kim, M.H. Jang, Y.J. Kang, J.H. Park, K.-S. Kim, H.S. Lee, C.J. Ryu, N.R. Gonzales, Construction, affinity maturation, and biological characterization of an anti-tumor-associated glycoprotein-72 humanized antibody, J. Biol. Chem., 281 (2006) 6985-6992.

[0064] S.P. Povoski, B. Cathy M. Mojzisik RN, MS, B.J. Sullivan, Radioimmunoguided surgery: Intraoperative radioimmunodetection for the radioguided localization and resection of tumors, 2016.

[0065] R. Rossin, R.M. Versteegen, J. Wu, A. Khasanov, H.J. Wessels, E.J. Steenbergen, W. Ten Hoeve, H.M. Janssen, A.H. van Onzen, P.J. Hudson, Chemically triggered drug release from an antibody-drug conjugate leads to potent antitumour activity in mice, Nat. Commun., 9 (2018) 1484.Attorney Docket No.: 5025-0446PWO1

[0066] R. Shu, V.J. Evtimov, M.V. Hammett, N.-Y.N. Nguyen, J. Zhuang, P.J. Hudson, M.C. Howard, A. Pupovac, A.O. Trounson, R.L. Boyd, Engineered CAR-T cells targeting TAG- 72 and CD47 in ovarian cancer, Mol. Ther. Oncolytics, 20 (2021) 325-341.

[0067] M. Minnix, L. Li, P.J. Yazaki, A.D. Miller, J. Chea, E. Poku, A. Liu, J.Y. Wong, R.C. Rockne, D. Colcher, TAG-72–targeted α-radionuclide therapy of ovarian cancer using 225ac-labeled DOTAylated-huCC49 antibody, J. Nucl. Med., 62 (2021) 55-61.

[0068] M. Minnix, L. Li, P. Yazaki, J. Chea, E. Poku, D. Colcher, J.E. Shively, Improved targeting of an anti‐TAG‐72 antibody drug conjugate for the treatment of ovarian cancer, Cancer Med., 9 (2020) 4756-4767.

[0069] G.R. Ediriweera, J.D. Simpson, A.V. Fuchs, T.K. Venkatachalam, M. Van De Walle, C.B. Howard, S.M. Mahler, J.P. Blinco, N.L. Fletcher, Z.H. Houston, Targeted and modular architectural polymers employing bioorthogonal chemistry for quantitative therapeutic delivery, Chem. Sci., 11 (2020) 3268-3280.

[0070] L.P. Rivory, M.-C. Haaz, P. Canal, F. Lokiec, J.-P. Armand, J. Robert, Pharmacokinetic interrelationships of irinotecan (CPT-11) and its three major plasma metabolites in patients enrolled in phase I / II trials, Clin. Cancer Res., 3 (1997) 1261-1266.

[0071] Z.M. Prijovich, P.-A. Burnouf, H.-C. Chou, P.-T. Huang, K.-C. Chen, T.-L. Cheng, Y.- L. Leu, S.R. Roffler, Synthesis and antitumor properties of BQC-glucuronide, a camptothecin prodrug for selective tumor activation, Mol. Pharm., 13 (2016) 1242-1250.

[0072] D.F. Kehrer, W. Yamamoto, J. Verweij, M.J. de Jonge, P. de Bruijn, A. Sparreboom, Factors involved in prolongation of the terminal disposition phase of SN-38: clinical and experimental studies, Clin. Cancer Res., 6 (2000) 3451-3458.

[0073] R.H. Mathijssen, J. Verweij, W.J. Loos, P. de Bruijn, K. Nooter, A. Sparreboom, Irinotecan pharmacokinetics-pharmacodynamics: the clinical relevance of prolonged exposure to SN-38, Br. J. Cancer, 87 (2002) 144-150.

[0074] H.M. Dodds, P.J. Tobin, C.F. Stewart, P. Cheshire, S. Hanna, P. Houghton, L.P. Rivory, The importance of tumor glucuronidase in the activation of irinotecan in a mouse xenograft model, J. Pharmacol. Exp. Ther., 303 (2002) 649-655.

[0075] E. Gupta, X. Wang, J. Ramirez, M.J. Ratain, Modulation of glucuronidation of SN-38, the active metabolite of irinotecan, by valproic acid and phenobarbital, Cancer chemother. pharmacol., 39 (1997) 440-444.

[0076] W. Fang, M. Vikerpuur, M. Sandholm, A fluorometric beta-glucuronidase assay for analysis of bacterial growth in milk, Vet Microbiol, 46 (1995) 361-367.Attorney Docket No.: 5025-0446PWO1

[0077] J. Krahulec, J. Krahulcová, Characterization of the new β-glucuronidase from Streptococcus equi subsp. zooepidemicus, Applied microbiology and biotechnology, 74 (2007) 1016-1022.

[0078] W. Russell, T. Klaenhammer, Identification and cloning of gusA, encoding a new β- glucuronidase from Lactobacillus gasseri ADH, Appl Environ Microb, 67 (2001) 1253- 1261.

[0079] Y. Sakaguchi, K. Murata, M. Kimura, Clostridium perfringens and other anaerobes isolated from bile, Journal of Clinical Pathology, 36 (1983) 345-349.

[0080] Y. Xu, Y. Liu, A. Rasool, E. Wenwen, C. Li, Sequence editing strategy for improving performance of β-glucuronidase from Aspergillus terreus, Chemical Engineering Science, 167 (2017) 145-153.

Claims

Attorney Docket No.: 5025-0446PWO1 WHAT IS CLAIMED IS:

1. An engineered enzyme, comprising an amino acid sequence that is at least 80% identical to the amino acid sequence of a human beta-glucuronidase (hBG).

2. The engineered enzyme of claim 1, wherein the human beta-glucuronidase comprises the amino acid sequence of SEQ ID NO:

1.

3. The engineered enzyme of claim 2, wherein the engineered enzyme comprises a substitution at a residue that corresponds to residue G340, C396, S422, L423, E431, L456, Y461, M465, H469, S476, N484, N486, A488, Y495, S506, H509, I517, Q518, L519, Q520, T545, M593, E595, and / or T599 in the sequences of SEQ ID NO:

1.

4. The engineered enzyme of claim 3, wherein residue 340 is G or N; residue 396 is C or A; residue 422 is S or T; residue 423 is L or M; residue 431 is E or K; residue 456 is E or L; residue 461 is D; residue 465 is M or T; residue 469 is H or V; residue 476 isT; residue 484 is N or F; residue 486 is T or N ; residue 488 is H or D; residue 495 is Y or F; residue 506 is S or G; residue 509 is H or Y; residue 517 is I or V; residue 518 isR or E; residue 519 is L or G; residue 520 is Q or E; residue 545 is S or G; residue 593 is M or A; residue 595 is E or D; and residue 599 is T or F.

5. The engineered enzyme of claim 3, wherein the engineered enzyme comprises a substitution at L456, Y461, S476, N486, A488, Q518, T545, M593, E595, and / or T599.

6. The engineered enzyme of claim 3, wherein the engineered enzyme comprises a substitution at G340, C396, S422, L423, E431, Y461, M465, H469, S476, N484, A488, Y495, S506, H509, I517, Q518, L519, Q520, and / or T545.

7. The engineered enzyme of claim 3 or 5, wherein the engineered enzyme comprises residues E456, D461, T476, T486, H488, R518, S545, A593, D595 and / or F599.

8. The engineered enzyme of claim 3 or 6, wherein the engineered enzyme comprises residues N340, A396, T422, M423, K431, D461, T465, V469, T476, F484,D488, F495, G506, Y509, V517, E518, G519, E520, and / or G545.

9. A fusion protein, comprising: (1) a polypeptide, wherein the polypeptide can accumulate in the tumor ; and (2) the engineered enzyme of any of claims 1-8.Attorney Docket No.: 5025-0446PWO1 10. The fusion protein of claim 9, wherein the polypeptide comprises: an antibody or a fragment thereof, a binding protein, and / or an enzyme.

11. The fusion protein of claim 10, wherein the binding protein binds an extracellular matrix of a tumor; wherein the extracellular matrix comprises a collagen.

12. The fusion protein of claim 10, wherein: (1) the antibody is an antibody fragment, optionally selected from the group consisting of F(ab'2), Fab', Fab, Fv, single domain antibody, and single-chain variable fragment (scFv); and (2) the antibody is targeting to an antigen of a tumor.

13. The fusion protein of claim 12, wherein the antibody comprises scFv of hcc49.

14. The fusion protein of claim 12, wherein the antigen comprises glycoprotein.

15. The fusion protein of claim 14, wherein the glycoprotein comprises Tumor-Associated Glycoprotein-72 (TAG-72).

16. The fusion protein of claim 12, wherein the tumor comprises: breast cancerous tumor, lung cancerous tumor, liver cancerous tumor, pancreatic cancerous tumor, prostate cancerous tumor, skin cancerous tumor (melanoma), kidney cancerous tumor, bladder cancerous cancer, blood cancerous tumor (leukemia), colorectal cancerous tumor, and / or lymphoma.

17. A pharmaceutical composition comprising the engineered enzyme of any one of claims 1-8 and a pharmaceutically acceptable carrier.

18. A pharmaceutical composition comprising the fusion protein of any one of claims 9-16 and a pharmaceutically acceptable carrier.

19. A method of enhancing an anticancer activity of an anticancer drug in a subject, comprising administering to the subject in need thereof the engineered enzyme of any one of claims 1-8 or the fusion protein of any one of claims 9-16 or the pharmaceutical composition of any one of claims 17 or 18.

20. The method of claim 19, further comprising administering to the subject in need thereof an anticancer drug.

21. The method of claim 20, wherein the anticancer drug is metabolized to aAttorney Docket No.: 5025-0446PWO1 glucuronidated metabolite in the subject.

22. The method of claim 20, wherein the anticancer drug is a glucuronidated form upon administration.

23. The method of claims 21 or 22, wherein the anticancer drug comprises: Irinotecan (CPT-11), p-hydroxyaniline mustard glucuronide ester (HAMG-ester), Daunorubicin, Doxorubicin, Paclitaxel, Cisplatin, Docetaxel, Methotrexate, Temozolomide, Gemcitabine, Epirubicin, histone deacetylases inhibitors (CI-994), Etoposide, Cyclopamine, Camptothecin, Duocarmycin, Monomethylauristatin E, p- Hydroxyaniline mustard, and / or its derivatives thereof.

24. The method of claim 19, wherein an immunogenicity of the engineered enzyme is lower than that of beta-glucuronidase derived from the heterologous sources.

Images