Protein A is widely used in the biopharmaceutical field, playing a key role in antibody purification. It also serves as an important tool for the research of other biomolecules. Therefore, Protein A design is critical for bioengineering and drug development. Although computational protein design has made progress in model building and functional prediction, current methods still face the following limitations: (1) the predictive accuracy of generative models needs improvement, particularly in matching structural and functional features; (2) the multidimensional screening process for generated proteins requires further optimization. To address these issues, a synergistic strategy for Protein A design and wet-lab validation based on E2E+ESM2 is proposed. In the multidimensional screening process, this research introduces the innovative concept of feature distance. First, multiple Protein A-like sequences are synthesized using a generative model, and their tertiary structures are predicted using AlphaFold. Then, feature distances are calculated based on the ESM2 model, and multidimensional screening is performed by combining parameters such as skeleton distance and solubility. Finally, the functional performance of the selected synthetic proteins is validated through affinity testing. The experimental results show that the synthetic protein V2 exhibits excellent binding kinetics, with a KD value of 3.81±0.17E-10 M, close to the target Protein A. The balance between the association and dissociation rates indicates strong binding performance. This method improves the functional consistency and application potential of the generated proteins, providing a promising solution for protein design.