The fastHDX chip was designed using software packages and contained inlet channels for protein, labelling buffer and quench buffer and an exit channel. The channel lengths, widths and depths were optimised. The fastHDX chip was fabricated entirely from thiol-ene polymer which hardens under UV light. The ratios of the polymer components and the UV curing times were systematically varied to produce a durable fastHDX chip. To mix the liquids on-chip in a highly efficient way, a mixing ‘monolith’ was developed. This was a spatially restricted plug of thiol-ene polymer cured in the microchip channel at the junction between the channels. Two monoliths were placed in each chip to first mix the protein and the labelling buffer together and subsequently the labelled protein and the quench buffer together. The ratio of thiol-ene polymer components, methanol, photoinitiator, curing times and resulting mixing efficiency were adapted and tested to produce fastHDX chips with the ability to mix efficiently.
Once fabricated, the fastHDX chips were validated. The unwanted binding of protein to the fastHDX chip was tested and reduced by adding a coating to the microchip to minimise binding. The mixing of the monoliths was tested using a fluorescent dye and water to track the mixing efficiency. The loss of the HDX label was tested and compared with that of a standard HDX-MS workflow. The repeatability of the fastHDX chips was tested by labelling a protein under identical conditions using 4 different microchips. The robustness of the fastHDX chip was tested by performing the same labelling experiment with a fastHDX chip that was newly fabricated, again once the chip had been used 20 times and again after the chip had been stored for 2 months. The final iteration of the fastHDX chip passed all these validation tests.
The fastHDX chip was then used to study three proteins. First was the well-studied protein haemoglobin. The protein was labelled on-chip at a range of timepoints from 130 milliseconds to 4.3 seconds. The label uptake was then mapped onto the known structure of the protein to give a view of the accuracy of the fastHDX chip labelling. Even at very short labelling times, biologically relevant haemoglobin dynamics could be observed, such as higher uptake in loop regions and lower uptake in stable regions of the protein. The development and validation of the fastHDX chip and the labelling of haemoglobin was disseminated in an open access peer reviewed publication, published on the University of Copenhagen website and presented at 3 international conferences.
The fastHDX chip was then used to study the protein α-synuclein. Labelling on-chip at short timepoints up to 300 milliseconds revealed regions of transient structure in the protein at the C-terminal end. Finally, the fastHDX chip was used to study the protein CysK and its interactions in the cysteine synthase complex. Using a combination of on-chip and manual labelling, binding interactions in the CysK active site were observed, along with significant conformational changes elsewhere in the protein upon complex formation.
