Mass spectrometry-based urinary proteomic analysis in feline chronic kidney disease

Chronic kidney disease (CKD) is highly prevalent in domestic cats (Felis catus), which might serve as a spontaneous disease model for human CKD. Urinary proteomics is a promising tool for CKD diagnosis and management; yet, minimal information is available in cats. This study aimed to explore the urinary proteome in cats with CKD and assess its diagnostic value. High-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) was used for proteomic analysis, and protein quantities were calibrated to a spike-in internal standard. A cohort of 37 cats (14 with surgically induced CKD, 23 healthy controls) was used as a training dataset; a separate cohort of 40 cats (23 cats with naturally occurring CKD, 17 healthy controls) was used as the test dataset. A random forest model showed an area under the receiver-operating-characteristic (ROC) curve of 0.982 (95% CI, 0.952-1.000), outperforming serum symmetric dimethylarginine (SDMA) concentration, urinary protein-to-creatinine ratio, and other commonly studied urinary biomarkers (i.e., transferrin, albumin, and cystatin C). Retinol-binding protein 4 (RBP4) achieved a comparable area under the ROC curve of 0.957 (95% CI, 0.885-1.000). These data provide proof of concept for a mass spectrometry-based urinary proteomic analysis in feline CKD and identified urinary RBP4 as a CKD biomarker in cats.