Results

Experimental validation and data from our protein engineering platform—bridging computational predictions with wet-lab implementation.

Overview

This page documents our experimental progress across three parallel tracks. While full validation remains ongoing, we present the current state of our wet-lab implementation:

Experiment A (in vivo mutagenesis): We optimized PCR conditions and constructed plasmids for the OrthoRep continuous evolution system. Colony selection and initial verification steps have been completed, with further characterization in progress.

Experiment B (in vitro kinetics): Multiple SpyCatcher and SpyTag variants were successfully cloned and expressed. Protein extraction and Western blot confirmation show expected molecular weights. Quantitative binding assays under varying pH conditions are planned for the next stage.

Experiment C (directed evolution validation): We implemented error-prone PCR protocols for SpyCatcher diversification and began replicating published variants as performance benchmarks.

These experiments represent foundational steps toward our AI-guided protein engineering pipeline. Many assays remain incomplete, but the work demonstrates progress in establishing the experimental infrastructure needed for iterative design-build-test-learn cycles.

Experiments

Experiment A

PCR Optimization

To determine the optimal PCR annealing temperature (68 °C vs 70 °C) for specific, clean amplification of the SpyCatcher (1,456 bp) and GFP (1,852 bp) inserts, and to generate sufficient high-quality amplicons for downstream cloning and experiments. We amplified the target plasmid carrying SpyCatcher (1,456 bp) and GFP (1,852 bp) by PCR, testing two annealing temperatures (68 °C and 70 °C). Agarose gel electrophoresis was then used to assess amplification. We found that GFP at 68 °C produced fewer nonspecific bands and a cleaner product, and SpyCatcher at 68 °C likewise yielded a clearer band (Fig. 2).

PCR Agarose Gel Electrophoresis

Figure 2. Agarose gel electrophoresis of PCR products for GFP and SpyCatcher at annealing temperatures of 68 °C and 70 °C.

Transformation

Introduce the amplified SpyCatcher and GFP plasmids into E. coli DH10B. After amplifying GFP and SpyCatcher, we transformed them into E. coli DH10B by heat shock and performed antibiotic selection to isolate transformants (Fig. 3).

Antibiotic Selection Transformation

Figure 3. Antibiotic selection following heat-shock transformation of SpyCatcher.

Experiment B: The k values of each sequence under different pH conditions

Description

The sequences designed by the Dry Lab were planned for synthesis and transformation into E. coli for expression and subsequent kinetic characterization. At the current stage, preliminary expression tests were conducted using IPTG induction followed by Western blot analysis to verify the presence of the half Luci-SpyCatcher–linker–His-tag fusion proteins. However, the Western blot results were inconclusive — the target bands appeared faint and the membrane showed high background and uneven signal intensity, indicating that protein expression and detection conditions require further optimization. Fluorescence measurement and split luciferase complementation assays to determine RLU (Relative Light Unit) values and corresponding binding kinetics (k values) are planned for the next phase, once expression and detection have been successfully validated.

Data

First, the target genes with mutations simulated by the Dry Lab were synthesized by IDT, using the pET-IDT plasmid as the backbone. In total, 14 genotypes of SpyCatcher were constructed — one wild-type sequence and six mutant sequences, each fused with luciferase at either the N- or C-terminus. The recombinant plasmids were then transformed into E. coli and selected using kanamycin resistance (KanR) as the antibiotic marker.

Kanamycin Resistance Screening

Figure 5. Colonies selected through kanamycin resistance (KanR) screening contain the SpyCatcher sequences generated from our simulated mutations. The figure illustrates seven plates representing the N-terminal fusion constructs for demonstration.

Next, the selected colonies were cultured and expanded. When the culture reached the exponential phase with an OD₆₀₀ value of approximately 0.4–0.8, plasmids were extracted using a DNA extraction kit. The extracted plasmids were then analyzed by DNA gel electrophoresis to verify whether the plasmids carried the correct base pair (bp) length and contained the intended target gene insert.

Gel Electrophoresis Left
Gel Electrophoresis Right

Figure 6. The gel electrophoresis image shows 16 inserted plasmids, including 14 SpyCatcher constructs and SpyTag sequences fused with luciferase at either the N- or C-terminus. All bands appear at the expected size (~5500 bp). Excluding lanes 1, 9, 10, and 20, which contain the DNA ladder, lanes 2–8 and 11–19 correspond respectively to: spycatcher002_origin_N, spycatcher002_origin_C, spycatcher002_design33_n29_N, spycatcher002_design33_n29_C, spycatcher002_design17_n22_N, spycatcher002_design17_n22_C, spycatcher002_design0_n0_N, spycatcher002_design0_n0_C, spycatcher002_design0_n1_N, spycatcher002_design0_n1_C, spycatcher002_design0_n15_N, spycatcher002_design0_n15_C, spycatcher002_design0_n13_N, spycatcher002_design0_n13_C, spytag_N, and spytag_C.

After confirming that the plasmids were correct, E. coli strains carrying each genotype were induced with IPTG for overnight expression. The cells were then lysed and centrifuged to obtain the protein samples. SDS-PAGE analysis was performed to confirm the successful extraction of the target protein. Subsequently, the His-tag engineered at the C-terminus of the SpyCatcher was used as an antigen for affinity purification, yielding purified target proteins.

We used Western blotting to verify expression of the half Luci-SpyCatcher–linker–His-tag fusion (~30 kDa), with β-tubulin (~55 kDa) serving as the internal (loading) control. No bands were detected on the Western blot.(Referring to Note) A plausible cause is insufficient IPTG induction, leading to inadequate expression/secretion of the half Luci-SpyCatcher–linker–His-tag fusion. We plan to further optimize IPTG concentration and induction time. The primary antibody incubation was performed at room temperature for 1.5 h following the His-tag antibody protocol; this condition may have been suboptimal for the anti-β-tubulin antibody and could have impaired detection of the internal loading control.

Experiment C: Literature Validation and SpyCatcher Directed Evolution via Error-Prone PCR

Future Perspectives

In Experiment C, we plan to conduct multiple rounds of error-prone mutagenesis and screening to build a diverse SpyCatcher variant library. By linking sequence diversity to kinetic outcomes, this wet-lab evolution will reveal naturally high-performance variants that computational prediction may overlook, providing experimental insight to refine the existing model.


Through this iterative feedback loop between wet-lab evolution and dry-lab modeling, we aim to minimize model bias and progressively improve both experimental and computational design. In the long term, this integrative framework establishes a scalable, data-driven paradigm for directed evolution, advancing synthetic biology toward more intelligent protein engineering.