M13 Bacteriophage based hybrid colorimetric sensor for visual, low cost, sensitive, non-invasive diagnosis of canine diabetes from complex matrix in urine

Abstract

We report a hybrid colorimetric sensor for non-invasive canine diabetes diagnosis, developed by integrating transition metal complexes onto a genetically engineered M13 bacteriophage. Leveraging solvatochromism and ligand manipulation, this design enhances both color diversity and analyte selectivity, enabling sensitive glucose detection using only trace amounts of urine. The engineered M13 bacteriophage significantly increased absorbance intensity (similar to 87.3 % at similar to 780 nm) for six transition metals by promoting directional alignment of the metal complexes. H-1 NMR analysis of dog urine revealed elevated levels of glucose, lactate, acetate, formate, ethanol, and alanine in diabetic samples, whereas healthy samples exhibited high urea concentrations. From an initial set of 13 candidate sensors, Co-1 was selected for its reliable response to complex diabetic urine matrices. It demonstrated a glucose detection sensitivity of 100 ppm, aided by bacteriophage-enhanced signal intensity. Clinically, Co-1 produced a brown color with healthy urine and purple with diabetic urine. When tested on 54 healthy, 51 diabetic, 1 renal disease, and 2 hyperadrenocorticism samples, Co-1 achieved 94.7 % specificity and 98.1 % sensitivity based on hierarchical cluster analysis (HCA) of RGB values. X-ray photoelectron spectroscopy (XPS) confirmed distinct chemical interactions: strong urea binding in healthy urine and glucose-induced chelation in diabetic samples, explaining the observed color shift.