Meghan Tierney, Niranjani Devi Nittala, Devjit Roy, Puja Sharma, Matt Kuruc, Li Chen, Kiran Madura, and Swapan Roy ProFACT Proteomics, Inc. North Brunswick, NJ

Expanded knowledge of the human proteome furthers our understanding of the basis of disease and offers insight into drug development and biomarker discovery. The high throughput SeraFILE™ protocol generates approximately 80 biologically active sub-proteome pools with a lesser degree of complexity than the parent protein sample. Thus, dissecting the proteome prior to downstream analyses allows researchers to profile and study proteins in new ways. ProFACT’s SeraFILE™surface library affords high-resolution partitioning of complex protein extracts into differential sub-proteomes that can be further characterized using a variety of downstream applications. Collectively, SeraFILE™-derived protein profiles offer a more comprehensive signature of the starting sample. SeraFILE™ can fractionate proteins and thus signatures may reveal discrete differences in low abundance proteins across samples. It is important to note that no high abundance proteins are removed from the sample prior to fractionation; they are distributed across sub-proteomes. Therefore, SeraFILE™ facilitates detection of disease-specific differences in clinical proteomics applications.

SeraFILE™ also addresses an intrinsic normalization problem in clinical proteomics; relative differences across samples can be interpreted despite any disparities in the starting material (i.e. total tissue weight and/or total protein analyzed). We have effectively utilized SeraFILE™ to analyze and compare diseased/normal adjacent tissue sample pairs. Protease activity profiles show markedly different patterns between tissue types. In serum studies, it was experimentally determined that the total protein applied to the SeraFILE™ surface library can vary by up to 4X, while enzymatic activity profile patterns of alkaline phosphatase are consistently observed.

• SeraFILE™ is a high-throughput method that reduces protein pool complexity while maintaining native proteins that are still functionally active. These protein pools can then be profiled by traditional electrophoretic or bioassay methods.
• Serum and specific tissues are a source of over expressed and/or modified proteins unique to the disease state.
• Current serum proteomics applications have a suite of limitations, including the following 1) many protocols used in serum biomarker research purposely discard abundant serum proteins 2) discarded proteins may bind diagnostically relevant proteins in a given sample and 3) finding a single protein biomarker representative of the diseased state has proven difficult.
• SeraFILE™ may overcome these challenges in numerous ways. Certain surfaces remove albumin while others leave it intact, so all proteins in a serum sample can be resolved in the presence/absence of abundant serum proteins. Since proteins profiled by SeraFILE™are still functionally active, they can be analyzed by a variety of downstream methods. The utility of SeraFILE™ allows for a more comprehensive understanding of the serum proteome in both the normal and diseased state.
• Sample heterogeneity is a recognized problem in clinical proteomics. SeraFILE™ is well suited to address the complex normalization issues involved in tissue/tumor proteomics applications.

CONCLUSIONS

• SeraFILE™ is an automated proteomic profiling system that generates protein pools that retain biological activity.
• Different protein profiles are observed after treatment with each individual matrix. When all matrices are used in combination, 77 sub-proteomes are generated. These protein pools are amenable to a variety of downstream analyses.
• Low abundance proteins can be unmasked using SeraFILE™ profiling technology.
• High abundance proteins, like albumin, can be removed in certain SeraFILE™ sub-proteomes. It is important to note that SeraFILE™ technology does not discard any proteins in a given sample; all protein can be accounted for in either the FT, W, E1-E4, or bound to the matrix.
• In order to assess if native enzyme activity could be retained and effectively profiled in SeraFILE™ sub-fractions, alkaline phosphatase activity was assayed in sheep serum protein pools. Despite initial protein load concentration, the distribution of protein and enzyme activity recovered from the matrices in each fraction is the same for each surface tested. Over an application concentration range of 1-4 mg/ml protein, a reproducible pattern of total protein and activity recovery is observed in each matrix.
• Proteasome activity is markedly increased or decreased after SeraFILE™ treatment. These results demonstrate how SeraFILE™matrices have the potential to remove small molecule and/or protein inhibitors in a native protein sample. Since nothing is discarded from SeraFILE™ sub-fractions, inhibitory molecules may be easily identified using existing technologies (i.e. mass spectrometry).
• Clinical proteomics applications are encumbered by standardization protocol difficulties, so samples may not be adequately compared. For example, biopsies of tumor cells often include some portion of the normal adjacent tissue or necrotic portions with no cellular activity (2,4,5). These problems, inherent in tissue based proteomics, may be by-passed by profiling with SeraFILE™ because within a given range, the amount of protein applied to the SeraFILE™matrices does not appear to affect the proteomic profile generated. This would imply that normalizing to tissue weight or total soluble protein would not be a factor in the experimental outcome.