Search

The identification and characterization, at the cellular level, of cytokine productions present a high interest for both fundamental research and clinical studies. However, the majority of techniques currently available (ELISA, ELISpot, flow cytometry, etc.) have several shortcomings including, notably, the assessment of several cytokines in relation to individual secreting cells and the monitoring of living cell responses for a long incubation time. In the present work, we describe a system composed of a microfluidic platform coupled with an antibody microarray chip for continuous SPR imaging and immunofluorescence analysis of cytokines (IL-2 and IFN-γ) secreted by T-Lymphocytes, specifically, and stably captured on the biochip under flow upon continued long-term on-chip culture (more than 24 h).

Silicon photonic sensors based on Mach Zehnder Interferometers (MZIs) have applications spanning from biological and olfactory sensors to temperature and ultrasound sensors. Although a coherent detection scheme can solve the issues of sensitivity fading and ambiguity in phase direction, the measured phase remains 2π periodic. This implies that the acquisition frequency should ensure a phase shift lower than π between each measurement point to prevent 2π phase jumps. Here, we describe and experimentally characterize two methods based on reference MZIs with lower sensitivities to alleviate this drawback. These solutions improve the measurement robustness and allow the lowering of the acquisition frequency. The first method is based on the phase derivative sign comparison. When a discrepancy is detected, the reference MZI is used to choose whether 2π should be added or removed from the nominal MZI. It can correct 2π phase jumps regardless of the sensitivity ratio, so that a single reference MZI can be used to correct multiple nominal MZIs. This first method relaxes the acquisition frequency requirement by a factor of almost two. However, it cannot correct phase jumps of 4π, 6π or higher between two measurement points. The second method is based on the comparison between the measured phase from the nominal MZI and the phase expected from the reference MZI. It can correct multiple 2π phase jumps but requires at least one reference MZI per biofunctionalization. It will also constrain the corrected phase to lie in a limited interval of [−π, +π] around the expected value, and might fail to correct phase shifts above a few tens of radians depending on the disparity of the nominal sensors responses. Nonetheless, for phase shift lower than typically 20 radians, this method allows the lowering of the acquisition frequency almost arbitrarily.