Sweat-Based Noninvasive Electrochemical Biosensor for Detection of Interleukin-4 and Interleukin-5 with Binary Endotype Classification of Type-2 Asthma
Akash Kumar, Sumana Karmakar, Nidhi Peenal Kothari, Santhosh Gowtham Giddaluru, Preeti Singh, Kai-Chun Lin, Sriram Muthukumar, Shalini PrasadAbstract
Chronic type-2 (T2)-high asthma is driven by complex cytokine networks that are difficult to capture with conventional invasive, episodic, laboratory-based testing. Two central effector cytokines of T2 immunity interleukin-4 (IL-4) and interleukin-5 (IL-5) orchestrate IgE class switching and eosinophil-mediated inflammation, respectively, yet no existing technology enables their noninvasive, multiplexed measurement at the point of care. Here, we report a sweat-based electrochemical biosensor that detects IL-4 and IL-5 by electrochemical impedance spectroscopy (EIS) on the AWARE (A Wearable Awareness with Real-time Exposure) platform. Monoclonal antibodies were immobilized via DTSSP cross-linking, with FTIR and zeta potential analysis confirming stable functionalization, and orthogonal mass spectrometry (MALDI-TOF MS and LC-MS/MS) verified the endogenous presence of intact IL-4 (∼15.09 kDa) and IL-5 (∼12.70 kDa) in inflammatory human sweat. The sensor reached ultralow limits of detection of 0.033 pg/mL for IL-4 and 0.23 pg/mL for IL-5, with dose–response R2 values of 0.96 and 0.95 and dissociation constants of 0.009 and 0.066 pM. Spike-and-recovery experiments gave accuracy R2 values of 0.98 and 0.99, with recoveries within the CLSI-recommended 80–120% range and interassay %CV below 20%, while cross-reactivity between channels and against common sweat interferents (ascorbic acid, lactic acid, melatonin, IL-6) stayed below 20%. Building on this, we introduce a binary endotype classifier that converts the two sensor outputs, thresholded at IL-4 ≥ 0.03 pg/mL and IL-5 ≥ 0.32 pg/mL, into four states: controlled (0,0), IgE-dominant (1,0), eosinophilic (0,1), and Th2-high (1,1). Validated against MSD electro-chemiluminescence saliva measurements from eight subjects across multiple time points, the classifier correctly stratified 7/8 as controlled. This work establishes the first sweat-based, noninvasive platform for multiplexed IL-4/IL-5 detection, advancing wearable sensors toward real-time T2 inflammation tracking and personalized asthma management.