Results

Engineering Success

Through multiple rounds of design iteration, we successfully developed a recombinant human serum albumin C476G mutant (rHSA-C476G), significantly enhancing the binding efficiency with the NIR-II dye CO-1080. By introducing a site-directed mutation at Cys476 (replacing Cys476 with Gly476), we disrupted the original disulfide bond and exposed a free –SH group, thereby improving the covalent binding between the protein and CO-1080. This mutation strategy not only optimized the binding site but also expanded the protein cavity, further enhancing the probe’s access to the binding site.

Re_Fig1A shows the molecular mechanism of covalent binding between CO-1080 and the rHSA-C476G mutant. After the mutation, the disulfide bond is broken, exposing the –SH group of Cys476, which facilitates the reaction with CO-1080. The optimization of the binding site is further validated through molecular docking simulations (Re_Fig1B), where the results clearly show how CO-1080 binds more tightly to the active site of the rHSA-C476G mutant, with significantly enhanced non-covalent affinity. The binding energy calculations (Re_Fig1C and 1D) demonstrate that the binding energy of rHSA-C476G is far superior to that of wild-type HSA and other mutant groups, indicating that the C476G mutation significantly improves binding efficiency and stability.

Re_Fig1: Mutant albumin with better binding property to the NIR-II chromophore. (A) Schematic of the disulfide bonds between the cysteines associated with the covalent binding of CO-1080 and HSA. rHSA stands for recombinant HSA and C476G indicates mutation of Cys476 to Gly476. Cys: cysteine, Gly: glycine (B) Theoretical simulation of HSA and rHSA-C476G binding to CO-1080 by gliding docking mode. Comparison of (C) docking score and (D) binding energy of CO-1080 to wild-type HSA and several groups of recombinant HSA, respectively. (E) NIR-II brightness of CO-1080@albumin under different conditions (mean ± SD, n=3 for each group). (F) Gel electrophoresis results of CO-1080@albumin incubated at room temperature for 2 h and (G) statistics on the ratio of CO-1080@albumin to the corresponding free dye

Experimental results further support these findings. In the comparison of fluorescence intensity (Re_Fig 5E), the NIR-II fluorescence signal of CO-1080 bound to the rHSA-C476G mutant is much higher than that of other mutants and wild-type HSA. This indicates that the C476G mutation not only enhances the binding efficiency of the probe but also improves the imaging performance of the fluorescent probe. Gel electrophoresis analysis (Re_Fig 5F) further confirms that the binding capacity of rHSA-C476G to CO-1080 is significantly higher than that of other mutants and wild-type HSA after incubation at room temperature for 2 hours. The statistical results of bound versus free dye (Re_Fig 5G) show that rHSA-C476G has the highest efficiency in binding CO-1080, with the lowest ratio of free dye, demonstrating that the mutant efficiently bound more probe molecules in a shorter time.

Additionally, Re_Fig 2A illustrates the process of constructing the rHSA-C476G mutant. Through site-directed mutagenesis and expression in a yeast system, we successfully obtained the recombinant HSA mutant. This mutation not only optimized the binding site for CO-1080 but also transformed the mutated protein cavity into an efficient microreactor that could covalently bind the probe under mild conditions without requiring additional heating steps. The figure clearly depicts how the rHSA-C476G mutant improves the interaction with CO-1080 without altering the overall structure of the protein.

Re_Fig2：Optimized recombinant HSA constructed by site-directed mutagenesis strategy

Overall, experimental validation confirmed the effectiveness of this design. The optimized rHSA-C476G mutant significantly improved both the non-covalent affinity and covalent binding ability with CO-1080, forming an efficient biomimetic fluorescent protein probe. This optimized probe is not only suitable for NIR-II imaging but also provides a stable and efficient tool for non-invasive deep tissue imaging, offering broad potential for future applications.

Imaging Challenges: Blood-Testis Barrier (BTB) and Blood-Brain Barrier (BBB)

Both the blood-testis barrier (BTB) and blood-brain barrier (BBB) present unique challenges for imaging due to their complex structures and selective permeability. These biological barriers are essential for maintaining the microenvironment of sensitive organs such as the brain and testes, but their disruption is difficult to assess using traditional imaging techniques.

The BTB is formed by tight junctions between Sertoli cells in the testes, which protect germ cells from toxins and immune responses. However, when the BTB is compromised due to trauma, infection, or disease, it becomes challenging to evaluate the extent of the damage because of the barrier’s deep location and selective permeability. Traditional methods such as tissue staining or contrast-based imaging often fail to provide real-time, non-invasive assessments of barrier integrity. Moreover, these techniques typically lack sufficient resolution and specificity, making it difficult to distinguish between healthy and damaged tissue, particularly in deep areas like the testes. This can lead to inaccurate diagnoses or delays in treatment.

Re_Fig3: Schematic of the in situ formed dye@rHSA FPs penetrating through the damaged BTB and accumulating in the testes

Similarly, the BBB serves as a protective interface between the central nervous system and the bloodstream, preventing harmful substances from entering the brain while allowing essential nutrients to pass through. Disruption of the BBB is often associated with severe neurological conditions, such as stroke, traumatic brain injury, and neurodegenerative diseases. However, traditional imaging techniques face similar challenges when assessing the BBB, especially in detecting early-stage damage. Conventional methods often lack the sensitivity needed to precisely visualize damaged areas, further complicating the evaluation of BBB integrity.

Re_Fig4: Targeted imaging of blood-brain barrier (BBB) disruption using the in situ formed dye@rHSA FPs

Given these challenges, we sought to test our engineered fluorescent protein probe in these difficult-to-image environments. The unique properties of our probe, including its enhanced binding efficiency and strong NIR-II fluorescence, make it an ideal candidate for assessing the integrity of both the BTB and BBB.

BTB Imaging Results

Experimental Validation:

The experimental results demonstrated that CO-1080@rHSA exhibited superior imaging capabilities in BTB-damaged mice, showing significantly higher brightness and testis-to-skin ratio (TSR) compared to the control groups. Re_Fig5 B and C show the NIR-II whole-body and testicular imaging of BTB-damaged and normal mice after the injection of CO-1080@rHSA and CO-1080@HSA. The images indicate that CO-1080@rHSA allowed for clearer differentiation between normal and damaged testes, starting as early as 3 hours post-injection and persisting up to 72 hours.

Re_Fig5: NIR-II Imaging of BTB Integrity Using CO-1080@rHSA

Re_Fig5 D highlights the brightness measurements of the testes and skin in the treated and control groups, with CO-1080@rHSA achieving a higher fluorescence intensity, particularly in the BTB-damaged group. Re_Fig5 E provides statistical analysis, demonstrating a significant difference in testis-to-skin ratios between the CO-1080@rHSA and CO-1080@HSA groups at 6 hours post-injection, validating the designed probe's efficacy in complex biological environments.

These imaging results not only provide a new method for real-time, non-invasive assessment of BTB integrity but also offer potential support in diagnosing conditions related to male infertility. Future research could focus on longitudinal monitoring of BTB integrity to assess the progression of testicular pathologies and the effectiveness of therapeutic interventions.

BBB Imaging Results

In this study, the probes used were not CO-1080 but other molecules, such as C7-1080 and IR-808Ac, which were primarily used to detect blood-brain barrier (BBB) disruption in a stroke model. C7-1080 is a near-infrared II (NIR-II) dye that binds to recombinant HSA mutant (rHSA) and provides clear imaging of rHSA infiltration into brain tissue in the stroke model through NIR-II imaging. In contrast, IR-808Ac was mainly used for vascular imaging to observe blood flow around the stroke area.

As shown in the figures, Re_Fig6 A demonstrates the imaging results of C7-1080 and IR-808Ac from 1 minute to 48 hours. The C7-1080 probe shows progressively stronger signals in the stroke region, clearly illustrating BBB leakage. In comparison, IR-808Ac produces weaker signals, primarily for vascular imaging. Re_Fig6 B provides a quantitative analysis of C7-1080 fluorescence intensity at different time points, revealing significantly stronger signals in the stroke model compared to the control group (Sham), particularly in the early detection phase (1 minute to 6 hours), indicating C7-1080's ability to detect early BBB disruption. Re_Fig6 C compares traditional staining methods (TTC and EB) with NIR-II imaging, showing that C7-1080 provides clearer and more precise visualization of BBB disruption, with superior resolution and sensitivity compared to traditional techniques. Re_Fig6 D further validates the high sensitivity of C7-1080, showing significantly higher fluorescence in the affected brain hemisphere in the stroke model compared to the Sham group. Re_Fig6 E displays the absorption spectra of C7-1080 and IR-808Ac, with C7-1080 having strong absorption at 1046 nm, making it ideal for NIR-II imaging, while IR-808Ac has stronger absorption at 796 nm, suitable for NIR-I imaging. Re_Fig6 F shows dual-channel imaging results at 40 minutes and 48 hours, where the red channel represents the C7-1080 signal and the green channel represents the IR-808Ac signal. This dual-channel imaging provides comprehensive information on the stroke area, arteries, and veins, simultaneously revealing both BBB disruption and vascular structure.

Re_Fig6: Detection of blood-brain barrier (BBB) disruption in a stroke model using C7-1080 and IR-808Ac @rHSA FPs.

C7-1080 demonstrated outstanding sensitivity and specificity in detecting BBB disruption, offering higher resolution compared to traditional staining methods. Additionally, dual-channel imaging combining C7-1080 and IR-808Ac allows for detailed visualization of both BBB disruption and vascular structure, further enhancing the imaging of the stroke area. These results confirm the high potential of C7-1080 in complex biological imaging applications.

Conclusion

The engineered rHSA-C476G protein probe significantly enhanced covalent binding with the NIR-II dye CO-1080, establishing a robust platform for bioimaging applications. Both BTB and BBB imaging results confirmed the probe's effectiveness in overcoming complex imaging challenges, such as detecting early-stage disruption in biological barriers. The high sensitivity and specificity of the probe in non-invasive imaging provide promising potential for broader clinical applications, especially in the real-time monitoring of tissue integrity in sensitive environments.