Risk assessment on-a-chip: a cell-based microfluidic device for immunotoxicity screening

Nanomaterials are widely used in industrial and clinical settings due to their unique physical and chemical properties. However, public health and environmental concerns have emerged owing to their undesired toxicity and ability to trigger immune responses. This paper presents the development of a microfluidic-based cell biochip device that enables the administration of nanoparticles under laminar flow to cells of the immune system to assess their cytotoxicity. The exposure of human B lymphocytes to 10 nm silver nanoparticles under fluid flow led to a 3-fold increase in toxicity compared to static conditions, possibly indicating enhanced cell–nanoparticle interactions. To investigate whether the administration under flow was the main contributing factor, we compared and validated the cytotoxicity of the same nanoparticles in different platforms, including the conventional well plate format and in-house fabricated microfluidic devices under both static and dynamic flow conditions. Our results suggest that commonly employed static platforms might not be well-suited to perform toxicological screening of nanomaterials and may lead to an underestimation of cytotoxic responses. The simplicity of the developed flow system makes this setup a valuable tool to preliminary screen nanomaterials.


1.
AgNPs quality controls via UV-Vis spectroscopy and ICP-MS 1 mL of MilliQ water was pipetted into a disposable 1.5 mL poly(methyl methacrylate) (PMMA) cuvette (Sigma-Aldrich) and a blank spectrum was collected on a spectrophotometer (Agilent 8453). 5 L of stock AgNPs (1 mg/mL) were dispersed in the cuvette to collect the UV-Vis spectrum of the sample and to measure the localized surface plasmon resonance (LSPR). The pass/fail status of the AgNPs was determined by comparing both measured maximum absorbance and peak wavelength to reference values provided by the manufacturer.
The stability of AgNPs in cell culture media was also monitored with UV-vis. Briefly, after acquiring a blank spectrum of 1 mL complete RPMI, 5 µL of the 1 mg/mL AgNP stock solution were added to the cuvette, and UV-Vis spectra were collected every 60 min for 5 h.

Transmission electron microscopy and cryo-transmission electron microscopy
AgNPs from the stock solution were imaged by TEM (JEOL JEM-2100 F), equipped with a field emission gun. Sample preparation included deposition of particles on Formvar film coated copper grids (PST ProSciTech). Images were acquired at 200 kV accelerating voltage. Cryo-transmission electron microscopy (Cryo-TEM) was used to investigate whether the size distribution of AgNPs would be affected by static incubation or administration under fluid regime. Briefly, a 1 µg/mL AgNP solution in PBS was incubated at 37°C and 5% CO 2 either in a 96-well plate, or under fluid flow in the tube connected to a syringe pump at a continuous flow rate of 0.2 µL/min for 5 h. A custom-built humiditycontrolled vitrification system was used to prepare the samples for Cryo-TEM (humidity ~80% and 22°C). 200-mesh copper grids coated with perforated Lacey carbon film (ProSciTech) were glow discharged to render them hydrophilic. 3 μL of the sample were pipetted onto each grid. After 5 s, the grid was blotted for 2 s using Whatman 541 filter paper, and then plunged into liquid ethane cooled by liquid nitrogen. Samples were examined using a Gatan 626 cryoholder (Gatan, USA) and Tecnai 12 Transmission Electron Microscope (FEI, The Netherlands) at an operating voltage of 120 kV and using an electron dose of 8-10 electrons/Å 2 . Images were recorded using a FEI Eagle 4kx4k CCD camera.

Sample preparation for ICP-MS analysis
The concentration of AgNPs and Ag + ions in complete RPMI medium was measured by means of ICP-MS. Microsep™ centrifuge filters (1 kDa cut-off, Pall Corporation, USA) were used to separate the AgNPs from the dissolved ions. The filters were preconditioned to prevent sorption of the Ag + ions by the membrane. 1 Briefly, 2 mL of a 0.1 M copper nitrate solution were added to the devices, which were then capped and centrifuged at 3800 x g for 15 min at 20°C. The membrane insert was carefully removed and excess copper nitrate solution was discarded. The device was then reassembled with a new tube, 2 mL of MilliQ water were added in the insert, which was then capped and centrifuged as above. The rinsing water collected in the tube was discarded, and any excess water carefully removed from the insert. The procedure was repeated with 1 µg/mL AgNP solution. After centrifugation (again at 3800 x g for 15 min at 20°C), the dissolved ions were collected in the tube and stored at room temperature (RT) until further analysis. To attain mass balance, the AgNPs retained by the insert were etched with a solution of 20 mM K 3 Fe(CN) 6 and Na 2 S 2 O 3 x 5H 2 O in MilliQ water for 5 min. This treatment oxidizes Ag 0 to Ag + (Fe(III)(CN) 6 3− ) and the released Ag + ions were collected in the tube by centrifugation as previously described and stored at RT until ICP-MS analysis.

4.
Detailed photolithography procedure A chrome photomask containing transparent features (parallel rectangles, 600 µm wide and 1.3 cm long, with numbers on the top edge) was designed using the L-Edit software and fabricated using a Direct Write lithography (Intelligent micropatterning SF100). SU-8 3050 photoresist (MicroChem) was spun at 1,000 rpm on a cleaned and dehydrated 3" silicon wafer for achieving a height of 70 μm and baked at 95 o C for 45 min. The wafer was then exposed to UV light through the mask (90 mJ/cm 2 , EVG 6200 Mask Aligner), baked on a hot plate at 95 o C for 3 min, and processed with developer to generate the layer of photoresist, which eventually formed the microchannels. SU8 developer (MicroChem) was used to dissolve away uncrosslinked photoresist. 2

Supplementary Results
Supplementary