Engineering a bifunctional alfa and beta hydrolase from a GH1 beta-glycosidase

This study demonstrates that a retaining GH1 $\beta$-glycosidase can be engineered through targeted mutations of second-shell residues to acquire $\alpha$-glycosidase activity while maintaining its native $\beta$-function, thereby proving that the stereochemical constraints of this enzyme family are more flexible than previously thought.

The Gist

The Big Idea: Teaching an Old Dog New Tricks (and Making it Do Two Tricks at Once)

Imagine you have a very specialized tool, like a left-handed glove. It fits perfectly on a left hand (a specific type of sugar molecule) and does a great job of cutting it open. But it is completely useless on a right hand (the mirror-image version of that sugar).

In the world of biology, these "gloves" are enzymes called Glycoside Hydrolases (GHs). Specifically, the ones in "Family 1" (GH1) are famous for being left-handed specialists. They only cut "left-handed" sugars (called beta sugars). They are so picky that they usually ignore "right-handed" sugars (called alpha sugars) entirely.

The Goal: The scientist in this paper, Felipe Otsuka, wanted to see if he could take one of these strict "left-handed gloves" and engineer it so it could also cut "right-handed" sugars, without breaking its ability to cut the left-handed ones. He wanted to create a bifunctional glove that works on both hands.

The Challenge: The "Lock and Key" Problem

Think of the enzyme's active site (where the cutting happens) as a lock.

• The Beta sugar is a key that fits perfectly into the lock.
• The Alpha sugar is a key that is the mirror image of the first one. It's shaped slightly differently, so it bounces right off the lock.

For decades, scientists thought this lock was too rigid to ever accept the other key. The paper asks: Can we tweak the lock just enough to let the second key in, without jamming the first one?

The Solution: The "Renovation Crew"

Instead of rebuilding the whole house (the enzyme), the scientist used a clever renovation strategy. He didn't touch the most important parts of the lock (the actual cutting blades, called catalytic residues). Instead, he focused on the walls and furniture surrounding the lock (the "second-shell" residues).

He used a team of computer architects (algorithms like Rosetta, PROSS, and FuncLib) to design a blueprint for a renovation.

1. Step 1 (Stability): First, the computers suggested 41 changes to make the enzyme stronger and easier to produce in bacteria.
2. Step 2 (The Magic Touch): Then, they looked at the room around the lock and suggested 4 specific changes to the "furniture" to make the room slightly bigger or shaped differently.

The Result: They built a "Frankenstein" enzyme with 45 mutations (changes) compared to the original.

The Experiment: Does it Work?

They built this new enzyme in a lab and tested it.

• The Good News: The new enzyme could cut the "right-handed" (alpha) sugar! It had never been done before with this specific family of enzymes.
• The Bad News: It wasn't perfect.
  • Slower Speed: It cut the new alpha sugar very slowly.
  • Weaker Old Skill: Because the room was renovated to fit the new key, the old key didn't fit quite as well anymore. The enzyme became less efficient at cutting the original beta sugar.
  • Fragile: The new enzyme was also less stable. It fell apart (denatured) at a lower temperature (44°C) compared to the original (57°C).

The Analogy: Imagine you took a high-performance race car and added a roof rack to carry a surfboard. Now it can carry a surfboard (the new trick), but it's a bit slower, handles worse on the track, and the roof rack makes it a bit wobbly in the wind. But, it can do both jobs.

The "Aha!" Moment: Why Did It Work?

The scientist looked at the computer models to see what happened inside the enzyme. He found the secret culprit: One specific change.

In the original enzyme, there was a bulky amino acid (Tryptophan) acting like a large armchair blocking the space needed for the alpha sugar. The engineer swapped this out for a smaller amino acid (Phenylalanine), effectively replacing the armchair with a small stool.

This tiny bit of extra space allowed the "right-handed" sugar to flip over and squeeze into the lock, while the "left-handed" sugar could still fit, just a little less snugly.

The Takeaway

This paper is a proof of concept. It shows that:

1. Nature isn't as rigid as we thought: Even enzymes that seem strictly "left-handed" can be coaxed into accepting "right-handed" shapes.
2. Computers are powerful tools: We can use software to predict exactly which tiny changes to make to achieve a complex goal.
3. Trade-offs are real: You can't always get everything. Gaining a new ability often means losing a little bit of speed or stability.

In summary: The scientist successfully taught a picky biological machine to do a second job it was never designed for, proving that with the right computer-guided tweaks, we can expand the toolbox of nature.

Technical Summary

1. Problem Statement

Glycoside hydrolases (GHs) are essential for carbohydrate metabolism but are typically constrained by strict stereochemical specificity. Members of Glycoside Hydrolase Family 1 (GH1) are archetypal retaining $\beta$-glycosidases that hydrolyze $\beta$-configured glycosidic bonds via a double-displacement mechanism. While retaining $\alpha$-glycosidases exist in nature, they are rare within the GH1 family, and no GH1 enzyme has been reported to naturally process both $\alpha$ and $\beta$ configurations of the same substrate. This stereochemical rigidity limits industrial and biomedical applications, as distinct enzymes are required for different glycosidic linkages. The central question addressed is whether a single retaining GH1 active site can be engineered to accommodate and catalyze the hydrolysis of opposing anomeric configurations ($\alpha$ and $\beta$) without altering the canonical catalytic residues.

2. Methodology

The study employed a computational protein design pipeline to engineer a "one-shot" variant of a well-characterized GH1 $\beta$-glycosidase from Spodoptera frugiperda (Sf$\beta$gly). No experimental directed evolution or screening was performed; the design was purely in silico.

• Design Strategy:
  1. Stability Optimization (PROSS): The wild-type (WT) X-ray structure (PDB: 5CG0) was minimized using Rosetta and submitted to PROSS to generate a sequence with improved solubility and stability for E. coli expression. This step introduced 41 mutations while restricting changes to first-shell catalytic residues and dimerization interfaces.
  2. Specificity Engineering (FuncLib): The PROSS-optimized sequence was used as input for FuncLib, which diversified five second-shell residues surrounding the active site (positions 38, 143, 372, 400, and 445). The goal was to subtly reshape the active site environment to accommodate an $\alpha$-configured substrate without disrupting the catalytic mechanism.
  3. Selection: Five top-scoring models were visually inspected. One variant (designated $\beta$-/ $\alpha$-variant) was selected for experimental validation. This final variant contained the 41 PROSS mutations plus 4 additional mutations from FuncLib: Y38W, W143F, L372M, and S445C.
• Experimental Validation:
  • The variant was cloned, expressed in E. coli BL21 (DE3), and purified.
  • Biophysical Characterization: Size-exclusion chromatography (SEC) determined oligomeric state; Circular Dichroism (CD) spectroscopy assessed secondary structure and thermal stability ($T_m$).
  • Kinetic Assays: Enzymatic activity was measured against four substrates: Pnp-$\alpha$-glu (the new target), Pnp-$\beta$-glu, Pnp-$\beta$-fuc, and Pnp-$\beta$-gal. Michaelis-Menten parameters ($K_M$, $k_{cat}$, $k_{cat}/K_M$) were calculated.
  • Structural Modeling: AlphaFold2 predicted the structure of the variant. Substrate docking (AutoDock Vina) was performed to visualize binding modes and correlate docking scores with kinetic data.

3. Key Results

• Acquisition of $\alpha$-Activity: The engineered variant successfully hydrolyzed 4-nitrophenyl-$\alpha$-D-glucopyranoside (Pnp-$\alpha$-glu), a substrate completely inert to the WT enzyme. This confirmed the creation of a bifunctional enzyme.
• Retention of $\beta$-Activity: The variant retained activity toward native $\beta$-substrates, though with reduced efficiency:
  • Pnp-$\beta$-glu: Catalytic efficiency ($k_{cat}/K_M$) decreased by ~2.6 orders of magnitude.
  • Pnp-$\beta$-fuc: Efficiency decreased by ~1 order of magnitude.
  • Pnp-$\beta$-gal: Efficiency decreased by ~6-fold.
• Structural and Biophysical Changes:
  • Oligomeric State: While WT Sf$\beta$gly is a dimer, the variant shifted to a monomeric state, likely due to mutations increasing solubility rather than disrupting the interface.
  • Thermal Stability: The variant exhibited a significant drop in thermostability, with a melting temperature ($T_m$) of 44.0 °C compared to 57.1 °C for the WT.
  • Secondary Structure: CD spectra indicated the overall fold was largely conserved, with only minor deviations in ellipticity.
• Mechanistic Insight:
  • Docking analysis revealed that the W143F mutation (Tryptophan to Phenylalanine) was the critical driver for $\alpha$-activity. By replacing a bulky residue with a smaller one, the mutation created additional space in the active site.
  • This space allowed the $\alpha$-substrate to bind in a "flipped" orientation relative to the $\beta$-substrate, positioning the anomeric carbon correctly for catalysis by the conserved glutamate residues (E187 and E399).
  • The mechanism remains a retaining double-displacement, as the catalytic residues were unchanged.

4. Key Contributions

1. Proof of Concept for Dual Specificity: This is the first report of a GH1 enzyme engineered to process both $\alpha$ and $\beta$ configurations of the same substrate without altering the catalytic machinery.
2. Targeting Second-Shell Residues: The study demonstrates that second-shell mutations are sufficient to modulate stereochemical specificity, validating a strategy that avoids the high risk of mutating direct catalytic residues.
3. One-Shot Computational Design: The success of a single, non-screened design iteration highlights the increasing power of computational tools (PROSS, FuncLib, AlphaFold) in enzyme engineering.
4. Mechanistic Flexibility: The work challenges the dogma that retaining GH1 enzymes are strictly locked into $\beta$-specificity, showing that the active site architecture has more plasticity than previously appreciated.

5. Significance and Limitations

• Significance: This research opens new avenues for industrial biocatalysis, potentially reducing the need for multiple enzymes in biomass saccharification or glycan remodeling. It provides a fundamental understanding of how stereochemical constraints can be relaxed through protein engineering.
• Limitations:
  • Trade-offs: The gain in $\alpha$-activity came at the cost of reduced catalytic efficiency for native $\beta$-substrates and a significant loss of thermal stability.
  • Expression Levels: The variant showed lower expression yields compared to the WT.
  • Low Turnover: The turnover rate for the new $\alpha$-substrate is low, requiring extended reaction times for detection.
  • Future Work: Further optimization (e.g., directed evolution or additional computational rounds) is needed to improve stability and catalytic efficiency for practical application.

In conclusion, the paper successfully demonstrates that the stereochemical barrier in GH1 enzymes is not absolute and can be overcome through targeted engineering of the second-shell residues, creating a rare bifunctional glycosidase.