Since their inception in 1995, DNA microarrays have become the icon of the genomics revolution. Everyone with any interest has become familiar with the images of precise rows and columns of colored spots denoting levels of gene expression and latterly gene copy number or location of a point mutation within a given biological system. More recently, the rise of commercial platforms and kits has reduced in complexity the technology of DNA microarrays to almost commodity status.

While new assays have retained interest in the information DNA microarrays can give us, omics technologists have fixed their sights on a new challenge: printing protein microarrays. It is understood that, as the effector molecules of a biological system, the study of proteins presents a greater opportunity to discover and characterize the processes of life. Many protein scientists have been outspoken in their criticism of DNA microarrays, as they felt gene-expression analysis based on mRNA abundance only told half the story.

Printing protein microarrays, however, presents a series of technical challenges greater than those of DNA microarrays. Arrayjet (www.arrayjet.co.uk) inkjet technology overcomes these challenges, enabling high quality protein microarray production.

Although protein microarray technology borrows extensively from the science behind DNA microarrays, there are fundamental differences between proteins and DNA, primarily due to the markedly increased heterogeneity of proteins by comparison. Whereas DNA microarrays can be printed onto a range of surface chemistries with a range of buffers, protein microarrays require printing buffers, which are in some cases specific for each individual protein.

In DNA microarrays, capture/hybridization is based on primary structure only, i.e. base sequence. In contrast, a given protein’s complex 3-D structure must be maintained for capture or detection to be possible. This has led to the development of substrate chemistries (Table 1) and printing buffers designed with the intention of preserving protein structure and/or activity.

The buffers themselves, as well as the proteins in them, can affect viscosity tremendously. It is common, for example, to find proteins dissolved in a buffer containing high percentages of glycerol, as this has been shown to improve protein stability. Alternatively, protein samples have been kept at low temperatures during printing for the same purpose. In either case, the end result is a set of samples of differing viscosities, which need to be printed onto what is most likely a fragile surface.

Arrayjet technology is suited to printing high-quality protein microarrays: a modern piezoelectric inkjet print head is used for printing the arrays. The print head, a Xaar XJ126, has been adapted by Arrayjet from an altogether unrelated industry—printing ink onto paper and textiles. This adaptation is the subject of two key patents filed by Arrayjet in the area of microarray production, one of which has already been granted. As such the print head, which was developed for industrial-scale printing and is therefore highly robust, is working almost at idle in microarray applications.

The print head contains 126 nozzles in a linear arrangement. The nozzles are employed in subsets for printing microarrays of a variety of sample types. Importantly, the print head is suitable for printing viscous samples and can easily print samples of up to 20 centipoise (cP) in fluid viscosity (1 cP is equivalent to the viscosity of water), including high concentrations of expressed proteins and highly complex protein mixtures, such as cell lysates, irrespective of lysis method.

Concerning stability, protein printing with a series of glycerol concentrations has already been tested successfully, as have other buffers used in protein microarray production, such as phosphate-buffered saline. Additionally, Arrayjet printing technology has been tested successfully down to temperatures of 6°C.

Samples are simultaneously aspirated into the print head via one of a series of user-selectable liquid-handling devices known as JetSpyders™ that are manufactured from stainless steel and are biologically inert and sufficiently robust so as to enable sample aspiration from the very bottom of a microtiter plate well. This allows microarrays to be printed from plate well volumes of 5 µL and below. This is significant as protein probes for use in microarrays are usually expensive to produce and purify.

The JetSpyders differ in terms of volume of liquid aspirated, number of print head nozzles primed with sample, and therefore deliverable volume (Table 2). Once samples have been aspirated into the print head (Figure 1), the JetSpyder is replaced in its washing station in preparation for the next round of aspiration.

Protein samples are dispensed without direct contact between the print head and the substrate on the fly—that is, the print head never ceases in motion as it prints the samples. Many of the substrates suitable for protein microarray applications benefit from lack of contact with the printing implement. Such contact could introduce inconsistencies in spot morphology and printing artefacts, comprimising data quality in downstream analysis.

Since proteins are fragile and may vary widely in viscosity in their ideal printing buffers, a method that relies on precise volume deposition rather than passive transfer via capillary action is most appropriate.

The dispensing volume from each nozzle is 100 pL, which depending on the buffer and surface chemistry used, produces spots of features of size range 90–120 microns in diameter. Importantly, the print head is used in an orientation such that the nozzles pass over the substrate sequentially, i.e. in a horizontal line one after another rather than simultaneously in a vertical line, as is usual in inkjet applications. This means that a number of additional benefits are available to the user.

Since all nozzles for a given sample will pass over the location into which a given sample will be deposited, or spotted, it is possible to have different nozzles print that same sample on different microarrays. This has the dual-benefit of reducing the possibility of missing spots, such as may be observed on a contact-spotting system when a pin drops out or sticks during printing. In the case of printing repeats of each sample on a given slide or slides, cycling between nozzles introduces the possibility to test for any nozzle variation (approximately 5% CV).

It is possible to have the print head deposit several drops of a given sample in the same feature location, for example, to increase spot size or sample concentration on the fly and within a given print run.

The accuracy and precision of the automation platform is such that it is possible to place drops on top of each other in subsequent print runs for the purposes of mixing samples, diluting a sample, adding an analyte to a spot containing a capture molecule, etc. This functionality is termed spot-on-spot.

Potentially fragile microarray substrates are often used in protein microarray applications as they are suitable for maintaining protein structures required for capture/detection. One of the benefits of an inkjet microarray spotter is that the printing is noncontact, and the surface therefore is not disturbed or damaged in any way during the printing process.

An additional benefit of noncontact printing is its flexibility; spotters with fixed pins or tips arranged to comply with SBS plate/well standards are restricted to that format when printing microarrays, producing discrete blocks that can leave unused wasted space on a microarray and struggle to efficiently produce alternative array patterns or formats. Printing performed on an Arrayjet spotter is more flexible and economical with much better use of available space and the capacity to print full slide arrays or mini-arrays.

This latter point is important since the lack of content of protein microarrays coupled with the need to test multiple samples under identical conditions make mini-arrays, or arrays of arrays (2x3 blocks, each containing 400 spots, for example), a desirable format.

The final challenge to would-be protein microarray manufacturers concerns the avoidance of cross- and carry-over contamination. Proteins are known to be sticky and will adsorb indiscriminately to a given surface, in some cases the tip of a dispenser involved in the printing process.

This issue has been addressed on Arrayjet instruments by the implementation of a number of print-head cleaning steps before, during, and after printing of each set of samples to ensure no cross- or carry-over contamination on the print head itself. This is followed by additional cleaning of the inner and outer capillary surfaces of the JetSpyder in use with Arrayjet system buffer fluid containing detergent.

A number of technical challenges face the microarray scientist, including sample integrity, viscosity, and variability, substrate surface integrity, and elimination of cross- and carry-over contamination. These challenges have been addressed by Arrayjet’s inkjet microarray spotters, making them well suited for high-quality protein microarray production.