Novel Type of Protein Chip for Multiplex Detection of Autoantibodies

Christer Wingren


Expert Rev Proteomics. 2013;10(5):417-420. 

In This Article

Summary of Methods & Results From the Paper

Designing high-performing protein microarrays is a challenging task, where the proteins to be arrayed play a central role.[1] As illustrated for antibodies, early work showed that the on-chip performance of readily available off-the-shelf intact antibodies varied significantly, with up to 95% displaying poor performances in some studies.[3,15] By contrast, recombinant antibody fragments, in part microarray adapted by molecular design, has been found to display high on-chip performances in terms of structural and functional stability.[16] In the case of functional protein microarrays, the possibility to design protein probes optimized for array-based assays at large still remains to be explored. To this end, the work of Akada et al. is very interesting, where they utilized a protein design providing the recombinant proteins to be arrayed with two novel and essential biophysical properties.[14] To accomplish this, they explored a protein design established by themselves a few years ago.[17] More specifically, the recombinant proteins were constructed to contain a six-histidine (6xHis)-fused green fluorescent protein (GFP)-tag (26 kDa) at the N-terminus and a 5xCys-tag at the C-terminus. Uniquely, the cysteine-tag was used to covalently immobilize the deposited proteins onto the previously in-house designed maleimide-incorporated diamond-like carbon substrates.[17] The incorporation of the cysteine-tag is vital, since it will enable all proteins to be immobilized in the same and orientated manner. By contrast, covalent immobilization of proteins with more than one potential coupling site (e.g., exposed lysine residues), which is normally the case using standard coupling chemistries and non-designed proteins, will result in proteins immobilized in different and non-orientated manner. This means that potential epitopes might be differently exposed, thereby affecting the ability of such probes to be recognized and bound by, for example, autoantibodies, which in turn will affect the performance of the array set-up. In comparison, previous work has shown that immobilizing proteins (antibodies) in a coherent and orientated manner via other affinity tags, such as the 6xHis-tag, significantly enhanced the functionality of the arrayed proteins.[18,19] Hence, the use of the cysteine-tag might prove valuable for array performance, and could be implemented in many different array layouts, not limited to merely protein arrays for screening of autoantibodies.

Moreover, the introduction of the GFP protein provided the arrayed proteins with an in-built marker for direct quantification of the amount of deposited proteins, which otherwise might be technically demanding to accomplish. This capability was also implemented by the authors (as color 1) in their dual-color readout, and used to 'normalize' the signal (color 2) detected for any bound autoantibodies.

Uniquely, the authors performed an on-chip denaturation of the arrayed proteins by heating the array and exposing it to a solution of denaturing and reducing agents. The heat treatment was possible because of the capacity of the in-house designed surface[17] to undergo heat pre-treatment. This treatment proved to be essential as it improved the subsequent detection. The authors concluded that this was accomplished by increasing the availability of antigenic determinants, especially in the GFP section of the fusion protein. According to the authors, the availability of the antigens (epitopes) in general might also be improved by the fact that arrayed proteins can be viewed as floating in surrounding solution attached to the chip surface only at the C-terminus cysteine-tag. Albeit successful, the approach of denaturing the protein probes might also represent a potential bottleneck, as this require the autoantibodies to be directed against denatured proteins. This issue could, however, be resolved by running two versions of the same set-up, one with denatured proteins and one with non-denatured proteins.

Representing the first generation of a functional protein array design for multiplex detection of autoantibodies, based on denatured recombinant proteins, microarray adapted by molecular design, the set-up presented by Akada et al. seems very promising.[14] In the next generation of set-up, the authors should describe the overall performance of the platform, such as sensitivity, reproducibility and dynamic range, in more detail, and if so required continue to further develop and optimize the platform. Another key issue that should be more clearly addressed in future work is the success rate of producing recombinant proteins carrying the designed tags at both the N- and C-terminus. This is a potential showstopper if it turns out that the success rate is too low and/or the protein expression yields are impaired.

The authors demonstrated proof-of-concept for the set-up using a 4-plex antigen array and showed that hepatocellular carcinoma-related autoimmune antibodies of IgG class could be detected in the sera of (some) hepatitis C virus-positive patients. While the number of patients and samples are too small to draw any strict conclusions regarding their role as useful biomarkers, the work clearly outlined the applicability of the set-up. The authors should continue their work, expand the repertoire of arrayed antigens and screen larger cohorts of patient samples. Hepatocellular carcinoma-related autoantibodies have been detected in several recent studies, and these antigens could serve as prime candidates for a first expansion of the array layout.[20] Ultimately, high-density array layouts should be designed, enabling large-scale screening of also novel, previously unknown autoantibodies, truly paving the way for biomarker discovery efforts.