Parylene

Wild radiocollared bears in northern Minnesota (N = 14) were located in their winter dens, anesthetized (Telazol^®, veterinary formulation of tiletamine and zolazepam) and temporarily extricated. (See Additional Files 1, 2, 3 and 4 for video sequences of hibernating bears during den visits.) Ethylene Oxide sterilized Insertable Cardiac Monitors (ICMs) that were developed for human heart patients (Reveal^® XT, Model 9529; Medtronic Inc., Minneapolis, MN; 9 cc; 8 mm × 19 mm × 62 mm; 15 grams) were surgically implanted in these bears in the field using aseptic techniques. An additional device was implanted in a wild orphaned cub that denned in an outdoor enclosure at a rehabilitation center and was released into the wild the following spring. Devices were placed subcutaneously in a peristernal location [23 ]. The device electronics are housed in a hermetically sealed titanium can. Electrocardiograms are recorded from a differential voltage measured between a titanium electrode housed in a polyurethane and silicone header with a region on the parylene coated titanium can serving as the reference electrode. This device became available to clinicians in the United States in 2009 and has two electrodes on the body of the device to continuously monitor the subcutaneous electrocardiogram (ECG). A built-in accelerometer measures patient activity. Device programming and data retrieval is non-invasive via transcutaneous telemetry associated with a programming system (CareLink Model 2090 Programmer with software Model SW007, Medtronic Inc., Minneapolis, MN).
In addition to storing the timing of each heartbeat and daily activity over the three year life of the device, the device memory can store up to 22.5 min of ECG recordings from patient-activated episodes and up to 27 min of ECG recordings from automatically detected arrhythmias. The devices also report daytime heart rate (HR) (08:00-20:00; referencing a 24 hour clock) and nighttime HR (0:00-04:00). For human patients, the ICM records cardiac information in response to both automatically detected arrhythmias and patient activation using a hand held device prescribed at the time of device implantation. Although designed for activation by a clinical patient during symptomatic episodes, the device can be activated by researchers and clinicians to record electrocardiograms during periods of interest. Arrhythmias that can be selected for automatic detection include: atrial tachyarrhythmias/atrial fibrillation (AT/AF), bradyarrhythmias (slow heart rates), asystole (long periods without a heart beat), and ventricular tachyarrhythmias (high heart rates).
We programmed devices after implantation in bears using a portable programmer, and used the same programmer to download data through the skin of bears visited at winter dens a year later. Devices were implanted in March 2009 and 2010, and follow-up visits were made the following December and March. In addition to continuously recording heart rates and activity, the devices were programmed to automatically detect and store the ECG for episodes in which: 1) a heart rate of at least 167 beats per minutes (bpm) was sustained for at least 16 beats ("tachycardia"), 2) a heart rate of less than 31 bpm was sustained for at least 4 beats ("bradycardia"), and 3) pauses of at least 4.5 seconds between consecutive heart beats ("asystole"). For purposes of data analyses, the period of winter inactivity (essentially the period of winter hibernation) was defined as the interval from when activity dropped below 1 hour/day in the fall to the time when activity of over 3 hours/day was sustained in the spring. Studies were conducted in conjunction with the Minnesota Department of Natural Resources and were approved by the University of Minnesota's Animal Care and Use Committee. All statistical analyses were performed using the non-parametric Mann-Whitney U-test. Normality was evaluated using a Shapiro-Wilk test. P-values less than or equal to 0.05 were considered significant.

Parylene-C dimer (DPX-C) was purchased from Specialty Coating Systems, Indianapolis, IN, USA. Parylene-C thin films with different thicknesses were deposited on 4” Si wafers using a PDS 2010 Parylene-Coater (Specialty Coating Systems, Indianapolis, IN, USA). The film thickness was controlled by the amount of dimer used.
AZ10XT-520cP positive photoresist was purchased from Integrated Micro Materials, Argyle, TX, USA, and it was used as the RIE etching mask. Photoresist spin coating was done by a Laurell WS-650 spin coater (Laurell Technologies Corp., North Wales, PA, USA). A spin coating calibration curve was established to obtain desired photoresist thickness. After a soft bake, the photoresist was exposed to ultraviolet (UV) light through the chromium photomask using an HTG mask aligner (San Jose, CA, USA) by contact photolithography. The photoresist was then developed in MIF-300 developer for various times at room temperature, and then rinsed in deionized (DI) water, and then dried by N₂ gas.
The RIE etching step was performed using a Plasmalab 80 Plus RIE etcher (Oxford Instruments, Bristol, UK). Oxygen plasma with different power, flow rate, and gas pressure was used.
The etching performance characterization including film thicknesses, vertical and lateral etching rate, size, and shape, and the sidewall profile of the etched micropore arrays was measured using a Dektak 3030 profilometer (Veeco (Sloan/Dektak),Bruker, Billerica, MA, USA), an Olympus BH2-UMA bright field optical microscope (Olympus, Center Valley, PA, USA), and a high-resolution Hitachi S4700 field emission scanning electron microscope (FESEM, Hitachi America, Ltd., Santa Clara, CA, USA; located at the NanoFab core facility at Arizona State University).
It is challenging to image the cross section of etched Parylene-C pores because Parylene-C is ductile under cleavage at room temperature and does not give a clean cross-sectional surface. To address this issue, the fabricated membrane on Si wafer was submerged into liquid nitrogen for 3 to 5 s before cleaving the etched membrane to give a better cross-sectional surface. For FESEM imaging, the samples were coated with a layer of 10–20 nm Au/Pd for top and cross-sectional views using a Cressington 108 auto sputter coater (Cressington Scientific Instruments, Walford, UK) to eliminate charge buildup during FESEM observations.

Osteosarcoma cell-line SJSA-1 cells (ATCC, Cat# CRL-2098) were cultured in a T75 cell-culture flask with RPMI 1640 medium containing 10% Fetal bovine serum (FBS, Gibco, ThermoFisher, Waltham, MA, USA), and 1% penicillin-streptomycin (Gibco, Thermo Fisher, USA). Cells were seeded at a subculture ratio of 1:10 and regularly passed about once a week (approximately 80% confluent). Culture medium was exchanged every third day and cells were used within 20 passes.
For characterizing low number (<50) cell counting inside a 10 μL pipette tip (Neptune, Cat# BT10), Countess Cell Counting Chamber Slides (ThermoFisher, Cat# C10228) and Nikon TS-100 phase-contrast microscopewere used. For high cell number calculations, cell density was measured by a Countess (ThermoFisher, MA, USA) automated cell counter.
For immunofluorescent imaging of captured tumor cells, CellTracker-Red (ThermoFisher, Cat# C34552) was used to stain the cytoplasm of SJSA-1 cells before the capture, according to the vendor’s instruction. After the capture, the cells were fixed on the Parylene-C membrane inside the capture device by 4% paraformaldehyde with 10 mL PBS wash, then stained by 300 nM DAPI (life Technologies Corporation, Eugene, OR, USA) solution with 20 mL PBS wash. The device was then disassembled, and the membrane was transferred to a glass slide to be mounted under a cover glass using CoverGrip Sealant (BIOTIUM, Fremont, CA, USA, Cat#23005). Finally, the sample was imaged using a Zeiss AXIO Imager M2 Epifluorescent Microscope (Carl Zeiss Microscopy, LLC, White Plains, NY, USA) in the Biomedical Imaging Core at the University of Arizona College of Medicine-Phoenix. The collected images were then analyzed using Zen 2.6 lite software to count the captured cells.