Hemocompatibility studies: Summary
Hemocompatiblity of pncSi membranes was tested by checking for three main parameters:
1) Platelet adhesion, aggregation and activation studies
2) Complement protein activation studies
3) Assaying levels of thrombin generation
All the studies and their assays were based on the paper by Fissell and Roy, who tested the same parameters on surface modified silicon wafers.
Platelet activation assays were straight forward: human blood was dispensed onto the membranes and incubated at 37C for 2 hours. After that, the blood sample was removed and the surfaces were fixed with paraformaldehyde. The surfaces were labeled with green CD41 antibody (CD41 is an integrin and is a typical platelet marker) followed with red CD62P antibody (CD62P, also known as P-selectin, is expressed on platelets only upon activation and hence is used as an activation marker). Green and red channel imaging thus highlights the nature of platelet aggregation and activation as well.
Teflon being an inert material was used as a negative control, while ADP is a calcium agonist and hence used as a positive control for activating platelets. Glass having similar surface chemistry as compared our pncSi, was used for reference comparison. pncSi was treated with PEG silane for antifouling purposes.
PEG treatment protocol:
Samples were O2 plasma cleaned at High Temperature in the YES system just prior to the liquid chemistry coating. Glassware was used to handle all subsequent chemistry. 10 ml of Toluene was extracted with a glass syringe and 100ul of 6.25N HCl was added to it. The solution was swirled around and 100 ul of PEG-Silane was added. The wafers/chips were soaked in this solution for 30 minutes. The wafer was then washed in fresh toluene and then dried by pulling it with MeOH.
Results of platelet studies: As expected, PEG treated pncSi showed very less platelet adhesion, and also, adhered cells exhibited negligible signs of activation as compared to only plasma treated pncSi sample. Although platelets adhered on glass, there was no significant aggregation due to activation on glass, indicating that cells simply attach to glass, and hence silicon, due to charge interaction and not due to any activating signal. Teflon surfaces behaved as an excellent negative control by preventing any cellular adhesion, while the the reverse was observed in the case of ADP (positive control).
Complement protein activation studies were studies using commercial ELISA kit to detect C3a protein. In brief, blood sample was incubated in the wells coated with anti-C3a antibody, and after incubation, biotinylated C3a polyclonal Ab and streptavidin peroxidase is added onto wells to sandwich the protein in between the two antibody layers. Finally a chromogen substrate is added that gets cleaved by peroxidase enzyme emitting yellow light, with the intensity proportional to the protein capture on the surface.
Diethyl aminoethyl (DEAE) cellulose was used as a positive control, as it is known to trigger complement activation. Teflon was used as a negative control. Again Peg treated and untreated pncSi surfaces were compared. Although I did obtain the predicted trend of protein activation (Teflon < pncSi < DEAE cellulose), the difference in the signals wer not hugely significant, as exaggerated in the Fissell’s paper. The differences in PEG treated and untreated pncSi were just modest indicating that PEG treatment may not be completely anti-fouling at the molecular levels and at the dimensions comparable to those of proteins.
Thrombin studies also involved similar type of commercial Sandwich ELISA kits, meant to detect the level of the complex thrombin-antithrombin III (TAT). Because of the absorbing properties of the cellulose membranes, I experienced a loss of liquid component of the blood onto the membrane. DEAE paper was precycled with PBS and other liquids before blood incubation, but unfortunately neither of the treatments worked very well in boosting signal. Moreover addition of external TAT from the kit also failed to spike the signal, thus challenging the robustness of the kit. So I decided to switch to yet another assay, which is more simpler that ELISA.
Fluorimetric assay: A commercial FRET peptide construct was purchased from Anaspec, CA; it had a donor sequence on one end and a quencher sequence on other. In its intact condition, the signal from the donor gets quenched by the the quencher, thus no apparent signal emission. In presence of protease, specifically thrombin, the FRET peptide is cleaved, and the signal from donor is now emitted without any quenching. Thus a calibration data is first obtained mapping the fluorescent signal with the native donor peptide. This assay (as company claims) is specific only for thrombin and not other protease in the blood. Unfortunately even this assay did not differentiate between positive/negative control and out pncSi and SiNi membranes. I changed the positive control from DEAE cellulose to magnetic steel, as the latter is known to be thrombogenic due to release of nichrome ions in solution. Just 1 run was made wtith this surface, but that also failed to act as a positive control.
Two points to be noted: positive control is not giving higher signal; that is exclusively the concern related to the positive control, and once optimum control is found, this problem could be solved. Second – teflon and pncSi give more or less same values or sometimes Silicon gives higher values. This is attributed to the suboptimal PEG treatment and higher density of Peg molecules can (hopefully) resolve this thing.
All the results related to hemocompatiblity studies are in the attached ppt for ease in annotations.