Gold‑Nanoparticle Colorimetric Sensor for Low‑Molecular‑Weight AGEs in Glycated Haemoglobin A0

Abstract

Protein glycation in diabetic plasma generates advanced glycation end products (AGEs) that drive disease progression. We glycated haemoglobin A0 (HbA0) in vitro using fructose, then used the resulting glycated protein as a template for gold nanoparticle (GNP) synthesis. The surface‑plasmon resonance (SPR) of the GNPs correlated strongly with the degree of glycation. Gel‑filtration fractionation revealed two product populations: high‑molecular‑weight protein‑cross‑linked fragments and low‑molecular‑weight non‑proteinaceous AGEs. Only the latter induced GNP formation, validating the proposed assay. Because AGEs serve as reliable biomarkers for diabetes and its complications, this GNP‑based colorimetric sensor offers a rapid, inexpensive diagnostic tool with potential clinical impact.

Background

Type II diabetes mellitus (T2DM) is characterized by chronic hyperglycaemia that triggers the Maillard reaction—a non‑enzymatic glycation of proteins and sugars. The initial Schiff base rapidly converts to the Amadori product, which undergoes further transformations to form AGEs. AGEs cross‑link proteins, activate the RAGE receptor, and induce oxidative stress and inflammation, contributing to cardiovascular disease, nephropathy, retinopathy, and neurodegeneration [4–18]. Other pathways, such as glucose oxidation, lipid peroxidation, and the polyol pathway, also generate AGEs [19–20], and dietary intake adds to systemic levels [6]. Despite their heterogeneity, AGEs share covalent cross‑linking and a browning effect that makes them attractive biomarkers for diabetes, ageing, and related disorders [23].

Current AGE detection relies on expensive chromatography–mass spectrometry or ELISA, limiting point‑of‑care use. Simple, rapid, and cost‑effective colorimetric sensors are therefore highly desirable. Gold nanoparticles (GNPs) offer tunable SPR and have been successfully employed as colorimetric reporters for sugars, proteins, and protein aggregates [52–56]. We previously showed that glycated proteins can seed GNP synthesis, with the extent of glycation reflected in the optical response [57]. Building on this, we investigated whether low‑molecular‑weight AGEs could drive GNP formation and thus serve as a basis for a qualitative sensor.

Methods

Materials

Haemoglobin A0 and Sephadex G25 (Sigma Aldrich) were used as received. Gold(III) chloride trihydrate (HAuCl₄·3H₂O, Loba Chemie) was the source of Au⁺. Fructose, phosphate buffers, and analytical‑grade reagents were sourced from standard suppliers. MilliQ water (>18 MΩ) was used throughout.

Methods

Glycation of HbA0

HbA0 (1 mg mL⁻¹) was incubated with fructose (0.1 M) in 0.1 M phosphate buffer, pH 7.4, at 37 °C for 1–10 days. Samples were aliquoted at day 0 (control) and day 10 (fully glycated) and stored under sterile conditions.

Reducing Property Assay

Reducing activity was quantified via a ferricyanide reduction assay. After incubation with potassium ferricyanide, the reaction mixture was acidified, and absorbance at 700 nm was measured; higher absorbance indicated stronger reducing power.

Gel Filtration Chromatography

Day 0 and day 10 Fruc‑Hb samples were fractionated on a Sephadex G25 column (15 cm × 1 cm) at 7.5 mL h⁻¹. Thirty fractions were collected and analyzed by UV absorption (280 nm) and Bradford protein assay.

Gold Nanoparticle Synthesis

Gold salt (1% w/v HAuCl₄) was mixed with water and stirred until pale yellow. Fruc‑Hb (12.5 ng μL⁻¹) or individual fractions (1 ng μL⁻¹) were added, then the mixture was left undisturbed to allow GNP growth. Final colours ranged from pink to purple, indicating successful nanoparticle formation.

Bradford Assay

Protein content in each fraction was quantified by measuring absorbance at 595 nm and 450 nm; the ratio (OD₅₉₅/OD₄₅₀) reflected protein enrichment.

Spectroscopy

UV‑Visible Spectroscopy

Spectra were recorded on a Perkin Elmer Lambda 25 from 200–800 nm with a 1 cm path length.

Fluorescence Spectroscopy

Intrinsic protein fluorescence (excitation 280 nm) and AGE fluorescence (excitation 350 nm) were measured on an Agilent Cary Eclipse.

Circular Dichroism Spectroscopy

Far‑UV CD (190–260 nm) assessed secondary‑structure changes using a Chirascan spectrometer.

Transmission Electron Microscopy

GNPs were imaged on a JEOL 2100F TEM after centrifugation and drying on carbon grids.

Colour Intensity Profile of GNPs

Colour was quantified by calculating (R–B)/G from digital images.

SDS‑PAGE

Glycated and control HbA0 samples were resolved on a 12% resolving gel, visualised under white light.

Results

GNP Synthesis from Fructose, HbA0 and Glycated HbA0

Fluorescence at 450 nm (excitation 350 nm) rose >10× after 10 days of fructose incubation, confirming AGE formation (Fig. 1). Reducing power assays showed day 10 Fruc‑Hb > day 0 Fruc‑Hb > fructose > HbA0 (Fig. 2a). Only day 10 Fruc‑Hb produced stable GNPs within 4 days (Fig. 2b), displaying a SPR peak at 530 nm (Fig. 2c) and spherical particles of 22 ± 2.4 nm (Fig. 2d). The data confirm that AGEs, not mere protein alterations, drive GNP synthesis.

Fractionation of Fruc‑Hb Resolves Two Distinct Glycation Product Classes

Gel‑filtration yielded two populations in day 10 Fruc‑Hb: high‑molecular‑weight proteinaceous fragments (fractions 5–12, 15+) and low‑molecular‑weight non‑proteinaceous AGEs (population II). Day 0 Fruc‑Hb produced only the proteinaceous fraction (population I) (Fig. 3). UV absorbance at 280 nm and Bradford analysis confirmed the proteinaceous nature of population I, while population II exhibited UV absorption but negligible protein content.

Fluorescence of Proteinaceous and Non‑Proteinaceous Glycation Products

Only day 10 fractions showed AGE fluorescence at 450 nm. Population I fractions had higher fluorescence intensity, indicating protein‑bound AGEs, whereas population II fractions emitted weaker fluorescence but retained UV absorption, confirming their non‑proteinaceous, low‑molecular‑weight nature (Fig. 4).

GNPs Seeded on Fruc‑Hb Fractions Differentiate Glycation Products

When GNP synthesis was performed with individual fractions, only population II (non‑proteinaceous AGEs) produced stable, coloured GNPs, while population I did not (Fig. 5). This demonstrates that the low‑molecular‑weight AGEs are responsible for nanoparticle formation and can serve as a selective sensor. Increasing Fruc‑Hb and gold concentrations accelerated the reaction to 1 day, highlighting suitability for point‑of‑care settings (Fig. S5). Linear colour intensity versus Fruc‑Hb concentration from 8–40 ng μL⁻¹ confirmed quantitative potential (Fig. 6).

Discussion

Our study establishes that low‑molecular‑weight AGEs are the active species driving GNP synthesis. Fractionation revealed that only non‑proteinaceous AGEs trigger nanoparticle formation, providing a clear mechanistic basis for the sensor. GNPs offer high sensitivity and can detect AGEs at nanogram levels, with a detection limit of 1 ng μL⁻¹. The method is simple, cost‑effective, and compatible with small sample volumes, making it ideal for rapid, point‑of‑care diabetes monitoring. Future work could extend this approach to distinguish specific AGE species and to detect advanced lipid peroxidation end products, further broadening clinical utility.

Conclusions

We present a robust colorimetric assay that detects low‑molecular‑weight AGEs in glycated haemoglobin with nanogram sensitivity. By leveraging the reducing power of AGEs to seed GNP formation, the sensor discriminates between glycated and non‑glycated samples, and between protein‑bound and free AGEs. Compared with LC‑MS/MS, this assay offers a rapid, inexpensive alternative for monitoring diabetes progression and associated complications. The specificity for AGEs, combined with its simplicity, positions it as a promising tool for clinical diagnostics and personalized glycaemic management.