View PDF Version
 
DOI: 10.1039/D5CC03960F (Communication) Chem. Commun., 2025, Advance Article

Identification and quantification of human serum albumin from different sources by a fluorescence sensor array

Mingjun Yang† , Mingwei Yan†, Weihua Deng, Zhongyong Xu* and Bin Liu*
College of Material Science and Engineering, Guangdong Research Center for Interfacial Engineering of Functional Materials, Guangdong Provincial Key Laboratory of New Energy Materials Service Safety, Shenzhen University, Shenzhen 518060, China. E-mail: xzy@szu.edu.cn; bliu@szu.edu.cn

Received 14th July 2025 , Accepted 4th August 2025

First published on 5th August 2025

Abstract

Recombinant and wild-type human serum albumin are accurately distinguished by a rationally designed fluorescent sensor array.

Human serum albumin (HSA), the most abundant protein in blood, plays vital physiological roles including osmotic pressure regulation, nutrient and metabolite transport, and immune modulation.1–3 It is not only used to treat hypoalbuminemia but also plays a key role in drug (particularly anticancer agents) metabolism—facilitating drug transport while reducing toxicity and enhancing cancer targeting.4–7 Currently, commercial HSA is derived from either human blood extraction or recombinant genetic engineering. However, due to different manufacturing processes (Fig. 1a), significant variations exist in the bound ligand composition.8 The wild-type HSA from human sources contains globulins, fatty acids (FA) or several kinds of enzymes, while the recombinant HSA (rHSA) is ligand-free. Although modern production processes have reduced the risk of pathogen transmission to minimal levels, some commercial wild-type HSA (No. A1653, A1887, and A4327 from Sigma-Aldrich, Fig. 1b) contain globulins, which can lead to immune responses and potential hypersensitivity reactions, including anaphylaxis. On the contrary, rHSA is free from trace levels of other proteins inherent to plasma-derived sources, achieving extremely high purity and batch-to-batch consistency. Therefore, there are critical scenarios requiring accurate differentiation between wild-type HSA and rHSA, especially in biopharmaceutical manufacturing of vaccines or clinical administration to severely immunocompromised patients.
image file: d5cc03960f-f1.tif
Fig. 1 (a) Information on six typical commercial human serum albumins (HSAa-HSAf) from human serum and recombinant source materials. Symbol (−) stands for essentially species free. Globulin (−): < 1%. Fatty acids (−): < 0.005%. Enzymes (−): protease < 0.0001 u mg−1 protein, alkaline phosphatase < 0.001 u mg−1 protein, and peroxidase <0.001 u mg−1 protein. (b) Common proteins existing in the wild-type HSA. (c) Schematic diagram illustrating the differentiation between wild-type and recombinant HSA using a fluorescent array.

Currently, conventional protein differentiation techniques, such as electrophoresis and chromatography achieve precise identification, but are limited by operational complexity, high costs, and specialized equipment.9 In contrast, sensor arrays offer a promising alternative by generating unique fingerprint patterns through cross-reactive probes, enabling simultaneous discrimination of multiple proteins.10–15 For example, Rotello's group developed a sensor array for protein differentiation.16,17 He's group demonstrated the applicability of sensor arrays in virus discrimination.18,19 Previous studies demonstrated that the binding of an environment-sensitive dye to the drug sites (DS) of HSA results in fluorescence change.20–22 Leveraging this fundamental mechanism, herein, we developed a fluorescent sensor array relying on various binding modes between dyes and HSA (Fig. 1c). Through principal component analysis (PCA) and linear discriminant analysis (LDA), we quantitatively screened and optimized the array-based sensing performance, ultimately obtaining a minimized yet efficient sensor array. Experimental results demonstrate that this sensor array not only effectively discriminates between wild-type HSA and rHSA, but also accurately classifies unknown HSA samples, providing a promising analytical approach for HSA type identification.

First, seven environment-sensitive dyes were selected based on their characteristic fluorescence changes upon binding with HSA (Fig. 2a). This is attributable to the fact that these seven types of environment-sensitive dyes have been substantiated by earlier studies to be capable of binding to albumin and eliciting significant fluorescence alterations (Table S1, SI). Five commercially available wild-type HSA (HSAa-HSAe, product No. A1653, A8763, A1887, A4327, and A3782 from Sigma-Aldrich, respectively) and one kind of rHSA (HSAf, product No. A9731) were employed as analytes. As shown in Fig. S1 in the SI, all seven dyes, when tested at consistent concentrations (10 μM), exhibited significant fluorescence enhancement along with concomitant red shifts in emission wavelength upon binding with HSA, confirming successful dye-HSA interactions. Fig. 2b and Fig. S2–S8 (SI) reveal the different fluorescence enhancement rates across HSA variants and distinct cross-reactive response patterns among dye-HSA combinations. Collectively, these results demonstrate that this seven-dye system can be applied as an effective sensor array for discriminating six HSA variants.


image file: d5cc03960f-f2.tif
Fig. 2 (a) The chemical structure of environment sensitive dyes. (b) Schematic illustration of differential fluorescence changes induced by different proteins in dye interactions. (c) Fluorescence response profiles and (d) heatmap of the seven-element sensor array for six HSA variants. (e) Two-dimensional LDA projection of the seven-element sensor array.

As shown in Fig. 2c, each HSA elicited a distinct fluorescence intensity profile from this array. The cross-reactive responses of the each sensing unit generated unique fingerprint patterns for all six types of HSA in the heat map representation (Fig. 2d). The multidimensional response pattern of the array was further statistically analyzed using LDA. The first two linear discriminants (LD1 and LD2) accounted for 96.2% of the total variance in the dataset (7 elements × 6 HSA × 5 replicates). As shown in Fig. 2e, the LDA plot demonstrated clear segregation of all HSA variants into distinct clusters with 100% classification accuracy (Tables S2 and S3 in SI).

Although increasing the sensing units typically enhances array performance, overly complex arrays introduce data processing challenges.23 To address this, we employed PCA to develop a minimized sensor array design as previously reported.24,25 Our analysis of the 7-element array revealed that the first three principal components (PCs) captured 93.67% of the total variance (Fig. S9 in SI), with the first two PCs alone demonstrating clear clustering patterns-evidence of strong cross-reactivity among the fluorescent dyes (Fig. 3a and Fig. S10, SI). Guided by PCA quantification of element contributions, we selected sensing elements that maximized feature space coverage without compromising discriminative performance. Fig. 3b shows FP540, FP647, and FP575 as dominant contributors to PC1 (19.4%), PC2 (64.5%), and PC3 (57.9%), respectively. This variance distribution informed our construction of an optimized 3-element array (FP540/FP647/FP575) that maintained the original classification accuracy while significantly reducing analytical complexity (Fig. 3c). The simplified array's performance was rigorously evaluated through LDA. While the two-dimensional projection showed minor overlaps between some variants (Fig. 3d), HSAf exhibited significant separation from other HSAs based on LD1 scores (Fig. 3e). Support vector machine (SVM) classification further demonstrated complete segregation of all six HSAs into distinct clusters (Fig. 3f, accuracy: 96.7%). These results collectively confirm that the minimized 3-element array maintains excellent discriminative capability comparable to the original 7-element system.


image file: d5cc03960f-f3.tif
Fig. 3 (a) PCA plot with a loading plot of the array for human serum albumin. (b) The contribution of each sensor unit to the dispersion of the array. (c) Optimization of the array combination for minimization. (d) LDA of the 3-element array for HSA. (e) Statistical differences among HSA samples based on the LD1 scores from the 3-element array (****p < 0.0001). (f) SVM of the 3-element array. (g) Blind analysis using the simplified array. (h) Heatmap of the array for HSAf, HSA_1 and HSA_2. (i) PCA plot, (j) LDA plot, and (l) HCA plot of the simplified array for HSAf, HSA_1 and HSA_2. (k) Significant statistical differences among two commercial albumins and HSAf (****p < 0.0001).

Subsequently, we tested four blind samples (S1: supplier code: A9731, S2: supplier code: A1653, S3: supplier code: A8763, and S4: supplier code: A1887) using the simplified array. The results demonstrated that the array accurately distinguished rHSA from wild-type HSA in four blind samples, achieving an accuracy rate of 100% (Fig. 3g and Table S4 in SI). To further illustrate the discriminatory performance of the array, we selected two commercially available HSA (HSA_1, HSA_2) to analyze the array's performance. Fig. 3h shows that the array generated specific fingerprint patterns for the three types of HSA. PCA (Fig. 3i) and LDA (Fig. 3j) revealed that the array could classify them into distinct clusters without any overlap. The LD1 scores indicated significant differences between the rHSA and the other two commercial wild-type HSAs (Fig. 3k). Meanwhile, hierarchical cluster analysis (HCA) also accurately grouped them into separate clusters without any misclassification (Fig. 3l). These results demonstrate that the simplified array exhibits excellent performance in distinguishing rHSA from wild-type HSA.

In summary, we developed an efficient fluorescent sensor array capable of accurately distinguishing between wild-type HSA and rHSA through rational sensor design. By leveraging the diverse binding modes between HSA and the dyes, as well as the compositional differences between rHSA and wild-type HSA, we employed statistical algorithms to streamline a seven-element sensor array into a simplified three-element sensor array. This reduction maintained the array's high efficiency in differentiating wild-type HSA and rHSA. Importantly, the simplified array demonstrated successful applications in accurately distinguishing rHSA from other commercial wild-type HSAs. This method provides a robust detection approach for HSA differentiation and ensures precise clinical application of HSA.

We acknowledge the financial support from the National Natural Science Foundation of China (22377081) and Shenzhen Science and Technology Program (JCYJ20230808105411023). We also would like to acknowledge the help from Instrumental Analysis Center of Shenzhen University (Xili Campus).

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI.

Materials, instruments, methods, synthesis, and supporting figures. See DOI: https://doi.org/10.1039/d5cc03960f

Notes and references

  1. O. K. Abou-Zied and O. I. K. Al-Shihi, J. Am. Chem. Soc., 2008, 130, 10793–10801 CrossRef CAS PubMed .
  2. S. Curry, H. Mandelkow, P. Brick and N. Franks, Nat. Struct. Biol., 1998, 5, 827–835 CrossRef CAS PubMed .
  3. X. M. He and D. C. Carter, Nature, 1993, 364, 362 CrossRef CAS .
  4. Y. Liu, Z. Wu, D. W. Armstrong, H. Wolosker and Y. Zheng, Nat. Rev. Chem., 2023, 7, 355–373 CrossRef CAS PubMed .
  5. W.-T. Dou, P. Qiu, Y. Shi, L. Zhu, C. Guo, N. Li, Y. Zang, T. Liu, S. Zhao, Y. Pan, L. Dong, J. L. Sessler, Y. Tan, J. Li, H. Wang, H. Tian and X.-P. He, J. Am. Chem. Soc., 2023, 145, 17377–17388 CrossRef CAS PubMed .
  6. F. Zeng, L. Tang, Q. Zhang, C. Shi, Z. Huang, S. Nijiati, X. Chen and Z. Zhou, Angew. Chem., Int. Ed., 2022, 61, e202112925 CrossRef CAS PubMed .
  7. S. Gao, G. Wei, S. Zhang, B. Zheng, J. Xu, G. Chen, M. Li, S. Song, W. Fu, Z. Xiao and W. Lu, Nat. Commun., 2019, 10, 2206 CrossRef PubMed .
  8. Y. Wang, F. Huo and C. Yin, J. Phys. Chem. B, 2024, 128, 1121–1138 CrossRef CAS PubMed .
  9. Y. Zhang, B. R. Fonslow, B. Shan, M.-C. Baek and J. R. Yates, Chem. Rev., 2013, 113, 2343–2394 CrossRef CAS PubMed .
  10. J. R. Askim, M. Mahmoudi and K. S. Suslick, Chem. Soc. Rev., 2013, 42, 8649–8682 RSC .
  11. R. Peri-Naor, Z. Pode, N. Lahav-Mankovski, A. Rabinkov, L. Motiei and D. Margulies, J. Am. Chem. Soc., 2020, 142, 15790–15798 CrossRef CAS PubMed .
  12. Z. Pode, R. Peri-Naor, J. M. Georgeson, T. Ilani, V. Kiss, T. Unger, B. Markus, H. M. Barr, L. Motiei and D. Margulies, Nat. Nanotechnol., 2017, 12, 1161–1168 CrossRef CAS PubMed .
  13. K. Selvakumar, L. Motiei and D. Margulies, J. Am. Chem. Soc., 2015, 137, 4892–4895 CrossRef CAS PubMed .
  14. J. Chen, P. Fasihianifard, R. Lian, L. J. Gibson-Elias, J. L. Moreno, Jr., C.-E. A. Chang, W. Zhong and R. J. Hooley, J. Am. Chem. Soc., 2025, 147, 841–850 CrossRef PubMed .
  15. A. J. Selinger and F. Hof, Angew. Chem., Int. Ed., 2023, 62, e202312407 CrossRef CAS PubMed .
  16. C.-C. You, O. R. Miranda, B. Gider, P. S. Ghosh, I.-B. Kim, B. Erdogan, S. A. Krovi, U. H. F. Bunz and V. M. Rotello, Nat. Nanotechnol., 2007, 2, 318–323 CrossRef CAS PubMed .
  17. N. D. B. Le, G. Yesilbag Tonga, R. Mout, S.-T. Kim, M. E. Wille, S. Rana, K. A. Dunphy, D. J. Jerry, M. Yazdani, R. Ramanathan, C. M. Rotello and V. M. Rotello, J. Am. Chem. Soc., 2017, 139, 8008–8012 CrossRef CAS PubMed .
  18. W.-T. Dou, X. Wang, T. Liu, S. Zhao, J.-J. Liu, Y. Yan, J. Li, C.-Y. Zhang, A. C. Sedgwick, H. Tian, J. L. Sessler, D.-M. Zhou and X.-P. He, Chemistry, 2022, 8, 1750–1761 CrossRef CAS .
  19. W.-T. Dou, P.-H. Tong, M. Xing, J.-J. Liu, X.-L. Hu, T. D. James, D.-M. Zhou and X.-P. He, J. Am. Chem. Soc., 2024, 146, 21017–21024 CrossRef CAS PubMed .
  20. Z. Luo, T. Lv, K. Zhu, Y. Li, L. Wang, J. J. Gooding, G. Liu and B. Liu, Angew. Chem., Int. Ed., 2020, 59, 3131–3136 CrossRef CAS PubMed .
  21. Z. Xu, M. Zhang, Z. Chen, Y. Zhao, L. Wang, X. Chen, B. Liu and X. Peng, Chem. Commun., 2023, 59, 5775–5778 RSC .
  22. X. Fan, Q. He, S. Sun, H. Li, Y. Pei and Y. Xu, Chem. Commun., 2016, 52, 1178–1181 RSC .
  23. J.-H. Tian, X.-Y. Hu, Z.-Y. Hu, H.-W. Tian, J.-J. Li, Y.-C. Pan, H.-B. Li and D.-S. Guo, Nat. Commun., 2022, 13, 4293 CrossRef CAS PubMed .
  24. M. A. Palacios, Z. Wang, V. A. Montes, G. V. Zyryanov and P. Anzenbacher, J. Am. Chem. Soc., 2008, 130, 10307–10314 CrossRef CAS PubMed .
  25. Y. Yu, W. Ni, Q. Hu, H. Li, Y. Zhang, X. Gao, L. Zhou, S. Zhang, S. Ma, Y. Zhang, H. Huang, F. Li and J. Han, Angew. Chem., Int. Ed., 2024, 63, e202318483 CrossRef CAS PubMed .

Footnote

† These two authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.