Data-informed discovery of hydrolytic nanozymes

The study addresses how to rationally design hydrolytic nanozymes capable of efficiently cleaving diverse bonds (phosphoester, amide/peptide, and glycosidic) despite varied substrates and mechanisms. The authors hypothesize that data-informed analysis can identify optimal MOF scaffolds and active-site features to enhance activity and broaden substrate scope. From a survey of hydrolase-like nanozymes, they propose two key design rules: employing hard Lewis acid metal centers to activate target bonds and increasing active-site density via shorter organic linkers. They validate these rules by creating a Ce4+-fumaric acid MOF nanozyme and testing it across multiple substrates and complex biofilms.

The authors screened 1481 papers and curated 105 on hydrolytic nanozymes. Publication frequency by material type for hydrolase mimics (phosphoesterase, nuclease, esterase, amidase/protease) highlighted MOFs and metal oxides as common scaffolds. Kinetic comparisons using half-life (t1/2) showed MOFs disproportionately achieving sub-10 min half-lives, indicating MOFs as optimal hydrolase scaffolds. A kinetic heat map of phosphorylated substrates revealed most effective systems used metals limited to Zr(IV), Ce(IV), Cr(III), Cu(II), Zn(II), and Ti(IV). Hard Lewis acids (e.g., Zr4+, Ce4+) correlated with faster hydrolysis, aligning with HSAB theory where strong Lewis acids polarize carbonyl/phosphoryl groups to increase electrophilicity. Pore/aperture analysis across MOFs showed that for pores >~1 nm, larger pores favor diffusion and faster rates, whereas below ~1 nm shorter linkers increase active-site density and can enhance rates. UiO-66 (BDC linker) vs UiO-67 (BPDC linker) exemplified linker-length effects, motivating selection of an even shorter linker, fumaric acid (FMA). The review also considered the role of modulators (acetic, formic, trifluoroacetic acids) in controlling MOF yield, crystallinity, morphology/size, and defect density, which affect catalysis.

Data-informed design: From literature trends, select MOFs as scaffolds with hard Lewis acid metals to activate substrates and short linkers to increase active-site density. Chosen metals: Zr4+, Ce4+, Hf4+. Linker: fumaric acid (FMA). Modulators screened: acetic acid (AA), formic acid (FA), trifluoroacetic acid (TFA). Fabrication: Synthesized Zr-FMA, Hf-FMA, Ce-FMA MOFs, and BDC analogs (Zr-BDC, Hf-BDC, Ce-BDC) following Methods; Ce-FMA crystallized within 10 min at room temperature. Ce-FMAs were prepared with varying modulator types and FA-to-FMA ratios; products differed in size and BET surface area. Activity screening: Phosphatase-like activity tested on p-nitrophenyl phosphate (PNPP) and bis-p-nitrophenyl phosphate (BNPP) to compare metal and linker effects. Optimization: Evaluated mass activity and surface-area-normalized activity versus modulator conditions; selected FA-to-FMA molar ratio of 20 as optimal, denoted Ce-FMA-FA-20-RT. Characterization: XRD confirmed UiO-66-type structure. TEM showed ~200 nm particles. XPS (Ce 3d) deconvolution indicated 82.7% Ce4+ and 17.3% Ce3+ (average valence ~3.8). TGA showed three-stage thermal behavior. FTIR confirmed trans C=C from FMA. BET surface areas: Ce-FMA ~120.44 m²/g vs activated Ce-BDC ~517.00 m²/g. Catalytic assays: - Phosphatase-like: PNPP and BNPP hydrolysis monitored via 4-nitrophenol absorbance at 400–405 nm. pH dependence assessed; PNPP optimal at pH 10.0; BNPP optimal at pH 9.0. Biological phosphates AMP, ADP, ATP, β-glycerophosphate tested via molybdenum-blue colorimetry detecting released phosphate; pH effects profiled. Larger substrates (cephalin, plasmid DNA) assessed and found not cleaved. - Protease-like: BSA (66.4 kDa) in PBS pH 7.4 at 60 °C and 37 °C. Monitored by GPC with UV at 280 nm; conversion calculated by BSA peak area. Compared to trypsin (37 °C) and MOF-808 (60 °C). Products analyzed by ESI-MS to determine fragment sizes and cleavage sites over time. - Glycosidase-like: Chromogenic substrates 2-nitrophenyl β-D-galactopyranoside (OD 420 nm) and 4-nitrophenyl N-acetyl-β-D-glucosaminide (OD 405 nm) tested across pH at 60 °C. Non-chromogenic disaccharides (maltose, lactose) probed by ion chromatography (no cleavage). Carboxymethyl chitosan degradation in Tris-HCl pH 8.0 at 37 °C and 60 °C monitored by GPC with RID; reusability tested across cycles. - Biofilm degradation: E. coli (gram-negative) and S. aureus (gram-positive) biofilms grown 48 h; treated with 50 µg/mL Ce-FMA-FA-20-RT with fresh medium for 12 h at 37 °C. Assessed by SEM, CLSM, and crystal violet staining (OD 590 nm). Viability checked by plate counts; morphology by microscopy. Data handling: Kinetic and conversion data compiled; half-lives summarized; comparisons made against literature heat maps.

• Data-informed design rules identified: hard Lewis acid metal centers (e.g., Ce4+) accelerate hydrolysis; shorter linkers increase active-site density in MOFs. - Ce-based MOFs consistently outperformed Zr-based analogs for phosphatase-like activity; Hf-based systems were inactive under tested conditions. - Despite a lower BET surface area (Ce-FMA ~120.44 m²/g vs activated Ce-BDC ~517.00 m²/g), Ce-FMA showed higher activity, supporting the active-site density and Lewis acidity hypotheses. - Optimized catalyst Ce-FMA-FA-20-RT (FA:FMA=20) exhibits UiO-66-type structure with Ce6O4(OH)4 nodes linked by FMA; XPS shows 82.7% Ce4+/17.3% Ce3+. - Phosphatase-like activity: PNPP optimal at pH 10.0; BNPP optimal at pH 9.0. Ce-FMA-FA-20-RT hydrolyzed AMP, ADP, ATP, and β-GP, generally more efficient for ADP/ATP; AMP and β-GP cleavage observed near neutral pH (HEPES 7.0–7.5). Cephalin and plasmid DNA were not cleaved (size/hydrophobic constraints). - Protease-like activity: BSA fully converted in 36 h at 60 °C and in 7 days at 37 °C. At 60 °C for 24 h, Ce-FMA-FA-20-RT achieved 75.54% conversion vs ≤50% for MOF-808, despite MOF-808 having ~8× larger surface area (1017.89 vs 120.43 m²/g). Trypsin reached 100% in 24 h at 37 °C but is ~10,000× more expensive. ESI-MS showed fragments 699–4764 Da (6–41 aa) early; final fragments 6–12 aa by 36 h. Selective cleavage occurred at residues R, K, and D; cleavage sites are surface-exposed. - Glycosidase-like activity: Efficient cleavage of 4-nitrophenyl N-acetyl-β-D-glucosaminide over 2-nitrophenyl β-D-galactopyranoside; no cleavage for maltose or lactose. Carboxymethyl chitosan was cleaved at pH 8.0 (37 °C and 60 °C) with higher temperature increasing product signal; activity retained ~80% in second reuse, dropped markedly by third. - Biofilm degradation: Significant reduction of E. coli and S. aureus biofilms after 12 h treatment with 50 µg/mL Ce-FMA-FA-20-RT, evidenced by SEM/CLSM and crystal violet assays; bacterial viability largely unaffected, indicating matrix hydrolysis rather than bactericidal effects. - Practical note: No co-catalysts (e.g., polyethylenimine or N-ethylmorpholine) were required, but strong phosphate binding to Ce/Zr can poison catalysts and impact recyclability.

The findings validate a data-informed strategy for designing hydrolytic nanozymes. By combining hard Lewis acidic metal nodes (Ce4+) with short linkers (FMA) in a MOF scaffold, the authors achieved a high density of potent active sites that activate phosphoryl and amide carbonyl groups, enabling broad-spectrum hydrolysis. Ce-FMA-FA-20-RT catalyzes cleavage of phosphoester, peptide, and selected glycosidic bonds, and effectively degrades complex biofilms without killing bacteria, addressing the challenge of multifunctional hydrolysis in mixed matrices. Comparative data against Zr-based MOFs and MOF-808 highlight the centrality of Ce4+ Lewis acidity over surface area. Protease-like selectivity for R, K, and D residues suggests side-chain-assisted mechanisms (basic/acidic functionalities and potential cyclic anhydride/imide pathways for D), opening routes to site-selective peptide processing. While performance approaches that of natural enzymes in scope (not rate), the dramatic cost advantage and stability of MOFs make them attractive enzyme alternatives. Limitations such as phosphate-induced catalyst poisoning, limited reusability, and poor activity on bulky/hydrophobic macromolecules indicate the need for future design modifications (e.g., tuned binding strength, larger pores, defect engineering, cofactor strategies).

This work establishes generalizable, data-informed design principles for hydrolytic nanozymes: select hard Lewis acid metal centers and increase active-site density via shorter linkers in MOF scaffolds. The resulting Ce-FMA-FA-20-RT exhibits robust phosphatase-, protease-, and glycosidase-like activities, cleaving diverse substrates (PNPP, BNPP, AMP/ADP/ATP/β-GP, BSA, carboxymethyl chitosan) and degrading bacterial biofilms under mild conditions. The study demonstrates that Ce-based MOFs outperform Zr-based benchmarks and even high-surface-area MOFs due to superior Lewis acidity and active-site arrangements, offering a low-cost, stable alternative to natural enzymes. Future research should focus on enhancing recyclability (mitigating phosphate poisoning), broadening glycosidase scope beyond substrates with good leaving groups, engineering pore architectures to accommodate larger biomacromolecules (e.g., lipids, nucleic acids), and developing strategies for site-selective hydrolysis and improved turnover.

• Catalyst poisoning due to strong phosphate binding to Ce (and Zr) reduces recyclability. - Reusability declines rapidly: ~80% activity in the second cycle but drops markedly by the third (carboxymethyl chitosan assay). - Limited activity on large/hydrophobic biomacromolecules (cephalin lipid and plasmid DNA not cleaved), likely due to pore-size constraints and hydrophobic interactions. - Glycosidase-like activity favored substrates with good leaving groups or coordinating groups; no activity on maltose/lactose under tested conditions. - Hf-based MOFs showed negligible hydrolytic activity under the tested conditions. - Activity optimization was pH-dependent and sometimes required alkaline conditions, which may limit certain applications.