Cell-Free Expression of Immobilized DNA

An important advantage of the DiaCHIP is the possibility to ship and store information encoded by DNA. From a DNA template array protein arrays can be produced on demand. In order to obtain this template, DNA is fixed on a silicone slide forming one side of our microfluidic chamber. Making use of a cell-free expression system, the DNA can then be transcribed and translated into the respective proteins resulting in the final protein array. The coding sequence of the proteins is genetically fused to a tag allowing their binding to a specific surface on the opposite side of the chamber.

Successful binding of DNA to the Silicone Slide

Figure 1: Scheme of the APTES/PDITC surface. The PDMS (red) is layered with APTES (light grey) and the amino-linker PDITC (dark grey). The hydroxy layer generated during plasma activation is represented in blue.

The core component of the DiaCHIP is the microfluidic chamber composed of a glass slide for protein immobilization and a PDMS (polydimethylsiloxane) flow cell. The silicone PDMS needs to be activated to enable the generation of the DNA template array by binding of the respective sequences. Oxygen plasma is used to initially activate the surface of the PDMS slide by generating a hydroxy layer. This allows linking the silane APTES and finally the crosslinker PDITC. In order to bind the DNA covalently to this surface the respective DNA sequence requires an N- or C-terminal amino group. The structure of this surface is schematically visualized in figure 1. This surface chemistry is identical to the protocol used to immobilize proteins on a glass slide unspecifically (see PDITC surface for immobilizing proteins).
In order to obtain an expression cassette for GFP with an amino group, the target sequence was amplified by PCR using an amino-labeled reverse primer. Additionally, the forward primer was labeled by the fluorescent dye Cy3 to enable visualization by fluorescence microscopy. Successful amplification of the target sequence was verified by agarose gel electrophoresis.

Figure 2: Immobilization of DNA on a PDMS slide. A: Top view on the slide indicating the spotting pattern. B: Cy3 fluorescence showing successful immobilization of amino-labeled DNA. As a negative control, Cy3-labeled DNA without amino group was spotted (indicated by white circles).

Coupling of DNA to the PDMS slide was achieved using a DNA concentration of 25 ng/µl. Either 1 or 3 µl of DNA were spotted directly onto the slide using a distinct pattern (figure 2A). To verify that only the amino-labeled DNA binds specifically, we added spots of non-amino-labeled DNA as a negative control. The slide was subsequently incubated over night and the DNA solution was washed away with ddH₂O. After drying, examination of the PDMS flow cell at 532 nm by fluorescence microscopy showed the pattern illustrated in figure 2B. The resulting fluorescence pattern clearly corresponds to the spotting pattern on the slide, thereby confirming that the immobilization of DNA was successful. Spots that were incubated with amino-labeled DNA show a distinct Cy3 fluorescence signal, whereas DNA that was not labeled with an amino group had not bound to the surface.

Cell-Free Expression of GFP From DNA Spots

Figure 3: Fluorescent GFP spots after 30 min of cell-free expression. The fluorescence was measured by placing the assembled chamber under a microscope. The system still was filled with expression mix.

Having verified that DNA can be bound to the PDMS flow cell, the next step was showing that the immobilized sequence is still capable of being transcribed and translated.
The microfluidic chamber consisting of the PDMS slide and a glass slide with a specific Ni-NTA surface was then assembled and filled with cell-free expression mix. The expression was performed for 2 h at 37°C.
The microfluidic chamber was analyzed by fluorescence microscopy afterwards. As shown in figure 3, cell-free expressed GFP was observed in distinct spots exhibiting a clear fluorescence signal. Thus, besides being expressed, the protein was also correctly folded and remained functional.

Putting it All Together

So far, we have shown that we are able to effect the different steps of the DiaCHIP system independently from each other.

• DNA can be immobilized reliably on a self-prepared PDMS surface using an amino-label.
• Our DNA sequences are capable of being expressed even when being bound to a surface.
• The expressed protein stays fully functional.

In consequence of these results, we combined all the single steps in one experiment. Therefore, DNA coding for His-GFP was immobilized on two PDMS flow cells. The microfluidic chambers were assembled by covering the flow cells with Ni-NTA coated glass slides. Cell-free expression was conducted with our DiaMIX and a commercial expression kit based on rabbit reticulocytes. The expression using the DiaMIX was performed at 37°C for 2 h, while the commercial kit was incubated at 30°C for 90 min. After expression, both slides were washed with PBS and stored at 4°C for a few hours. To examine if His-GFP was expressed and immobilized on the Ni-NTA successfully, the slides were measured in iRIf. The slides were first blocked with 10 mg/mL BSA using the microfluidic system, then anti-GFP was flushed over. Figures 4 and 5 show the binding curves we obtained from these two experiments.

Figure 4: Binding curve of iRIf measurement after cell-free expression of His-GFP in flow chamber with the DiaMIX.The graph shows a slight change in relative intensity at two spots when anti-GFP was flushed over.

Figure 5: Binding curve of iRIf measurement after cell-free expression of His-GFP in flow chamber with the commercial kit.The graph shows a slight change in relative intensity at two spots when anti-GFP was flushed over.

Although the relative intensity change is quite low, an increased signal at the spots where the GFP should have bound after diffusion can be observed. The intensity change is in a comparable range for both expression mixes. This indicates that the expression of His-GFP as well as its diffusion and specific binding to the opposite slide was successful.
Nonetheless, the experiment should be repeated, as the signal was not far above the background. Unfortunately, we did not have the time to reproduce and validate this result. Assuming that this result is reproducible, still a lot of optimization is necessary to obtain higher yields of protein expression. Consequently, higher signals in iRIf due to more specific antibody binding sites on the surface would be achieved.

This experiment underlines the potential of the DiaCHIP for future applications. It reveals that our DiaCHIP system is capable of producing protein microarrays on demand and detect antibodies by their binding to the corresponding antigens on the surface.