Identification of human cerebrospinal fluid proteins and their distribution in an in vitro microdialysis sampling system.

A qualitative study is presented on how proteins from a complex biological sample are distributed in a microdialysis sample system. A comparison between proteins identified in the human ventricular cerebrospinal fluid (CSF) sample, the collected dialysate and the proteins adsorbed onto the membrane was conducted. The microdialysis experiment was performed in vitro at 37°C for the duration of 24h. Thereafter, the membranes were removed from the catheter and the adsorbed proteins were tryptically digested using the on-surface enzymatic digestion (oSED) protocol. The CSF samples and the dialysates were digested using a standard in-solution trypsin digestion protocol. In the final phase, the samples were analysed using nano-liquid chromatography in combination with tandem mass spectrometry. In the four sample compartments analysed (CSF start, Membrane, Dialysate, CSF end) a total of 134 different proteins were found. However, most of the identified proteins (n=87) were uniquely found in one sample compartment only. Common CSF proteins such as albumin, apolipoproteins and cystatin C together with plasma proteins such as hemoglobin and fibrinogen were among the 11 proteins that were found in all samples. These proteins are present in high concentrations in CSF, which means that they effectively block out the detection signal of less abundant proteins. Therefore, only 25% of the proteins adsorbed onto the membrane were detected in the CSF compared with the dialysate that shared 44% of its proteins with the CSF. The proteins adsorbed onto the membrane were significantly more hydrophobic, had a lower instability index and more thermostable compared to the proteins in the CSF and the dialysate. The results suggest that proteins adsorbed onto the microdialysis membranes may escape detection because they are prevented from passing the membrane into the dialysate. Thus, the membrane needs to be examined after sample collection in order to better verify the protein content in the original sample. This is particularly important when searching for new protein biomarkers for neurodegenerative diseases.