Running on auto

Diagnosis is of course key to any treatment regime – and the use of autoantibodies as the basis of clinical assays is becoming more understood. Here we learn how serum profiling using protein microarrays could boost our ability to diagnose and treat disease

DEVELOPMENT of autoantibodies is observed in autoimmune disorders and numerous cancers. Consequently, autoantibodies form the basis of potential diagnostic and prognostic assays, as well as approaches for monitoring disease progression and treatment response. The effective use of autoantigen biomarkers for these applications, however, depends on the identification of not one but multiple biomarkers. This need for multiple biomarkers is due to the fact that the development of autoantibodies to any given protein is typically seen only in a fraction of patients afflicted with a specific disease.^1,2 Serum profiling on human protein microarrays can facilitate our understanding of disease by identifying panels of disease-specific biomarkers, thereby enhancing our ability to diagnose and treat such diseases. One such commercially available array is the ProtoArray Human Protein Microarray from Invitrogen. 

To understand the advantages that functional protein arrays offer for the detection of autoantibodies, it is important to first explore alternate methods for identifying such potential biomarkers for disease. One widely used approach for autoantigen identification is SEREX: serological analysis of cDNA expression libraries. This approach is most appropriate for cancer autoantigen identification and involves the generation of tumor-specific ÊŽ-GT11 cDNA expression libraries, followed by immunological screening of plaque lifts using patient sera. The SEREX approach was successfully used to identify NY-ESO-1, a protein which is expressed in certain cancer cells and is autoantigenic in approximately 20-50% of patients over-expressing NY-ESO-1³. The SEREX approach, while clearly useful, is slow, technically challenging and typically has a high false positive rate. Furthermore, because SEREX relies on bacterial protein expression, it cannot identify autoantigens requiring post-translational modifications⁴.

Figure 1: Profiling autoimmune disease serum samples. An example is shown for a biomarker discovery study in which serum samples were collected from autoimmune disease patients who have been diagnosed with Systemic Lupus Erythematosus (SLE), Rheumatoid Arthritis (RA), and Anti-Neutrophilic Cytoplasmic Antibodies (ANCA) as well as from healthy individuals. Both known SLE diagnostic biomarkers (shown in A) and novel SLE biomarker candidates (shown in B) were identified with the ProtoArray technology.

More recently, reverse phase protein microarrays have been used to identify colon cancer and lung cancer autoantigens.^5,6 These arrays are made by fractionating cancer cell homogenates, arraying them in spots on a microarray, probing them with patient sera and detecting antibody binding. Afterwards, mass-spectrometry based techniques are used to identify the actual autoantigen – a process which can be both time-consuming and expensive.

In contrast, functional protein microarrays offer defined protein content for profiling serum samples to identify autoantigen biomarkers. Invitrogen’s ProtoArray Human Protein Microarray contains over 8,000 purified human proteins immobilised on glass slides. Invitrogen developed the ProtoArray by using human recombinant proteins that are expressed in insect cells and purified under non-denaturing conditions.  Therefore, those proteins are expected to contain appropriate post-translational modifications and maintain their native conformations⁷.

An important advantage of high content, functional protein microarray technology is its high level of sensitivity. Antibody-based detection of proteins on the arrays is approximately 100-fold more sensitive than even the most sensitive Western blot methods¹¹. The increased sensitivity of the technology also allows for autoantibody detection from very small serum volumes (0.5–20 µl), which is critical for many clinical studies where sample sizes are small. Inter- and intra-assay reproducibility is highly important to the success of biomarker discovery projects because numerous samples are profiled on different days and by multiple operators - protein microarray technology enables reproducible profiling of multiple samples, reducing the impact of operator differences. Specific protocol steps, normalisation algorithms and protein QC standards have been included in the development and production of Invitrogen’s ProtoArray microarrays to ensure reproducible profiling capabilities important to the success of biomarker discovery projects.

Figure 2: Identification of SLE-specific biomarker candidates. Using the ProtoArray® technology and Prospector® software, a total of 64 unique biomarker proteins was identified based on their ability to correctly classify SLE from other autoimmune disease samples and healthy control samples. A hierarchical clustering heat map based on this 64-biomarker panel differentiates the majority of samples.

One application of functional human protein microarrays in general is the identification of potential biomarkers using immune response biomarker profiling.  By probing the ProtoArray with serum from diseased individuals and comparing those profiles to those of healthy individuals, scientists can quickly identify reactive proteins that are putative biomarkers. Once validated, these biomarkers can potentially be used to diagnose and classify disease,⁸ determine drug efficacy and toxicity, measure response to therapy, and monitor disease pathogenesis.^9,10

This technology has several advantages over other biomarker discovery platforms. In contrast to current gel electrophoresis methods coupled with mass spectrometry-based approaches that are labour-intensive, time-consuming and technically challenging, ProtoArray technology provides a user-friendly, rapid discovery platform for immediate identification of disease-specific biomarkers. The system queries a greater portion of the human proteome compared to low content antibody-based arrays. In addition, the ProtoArray is comprised of full-length, functional protein markers.

To illustrate the utility of the ProtoArray for identifying autoantigens that are specific to disease, we present the following case study. Systemic lupus erythmatosus (SLE) is an autoimmune disorder that disproportionately affects women of childbearing age. It affects various physiological systems, including the skin, joints, heart, lungs, blood and kidneys. Diagnosis of lupus relies on the combination of detailed medical histories and analysis of the results of specialised diagnostics and routine laboratory tests. There are currently no tests that clearly diagnose the disease. The identification and validation of new biomarkers specific to lupus may enable the development of definitive diagnostic tests.

To identify novel SLE biomarkers, serum samples were collected from autoimmune disease patients diagnosed with SLE, rheumatoid arthritis (RA), and anti-neutrophilic cytoplasmic antibodies (ANCA) as well as from healthy individuals. Figure 1 shows results generated from the serum profiling studies utilising human ProtoArrays. Autoantibodies present in the samples were detected using an AlexaFluor647-conjugated anti-human IgG antibody followed by scanning with an Axon 4000B scanner. Data was analysed using ProtoArray Prospector v4.0 software. SLE-specific biomarker candidates were identified by applying M-Statistics to quantile normalised signals. A p-value is assigned to the M-statistic, and this p-value is utilised to assign a significance ranking to each biomarker candidate. Univariate analysis of up-regulated proteins enables differentiation of SLE patients from disease controls (RA and ANCA) and healthy individuals. Both known SLE diagnostic biomarkers (shown in A) and novel SLE biomarker candidates (shown in B) were identified.

A total of 64 unique biomarker proteins were identified based on their ability to correctly classify SLE from other autoimmune disease samples and healthy control samples (Figure 2). A hierarchical clustering heat map based on this 64-biomarker panel differentiates the majority of samples. As shown in Figure 3, the profiling protocol is straightforward: (1) incubate individual serum samples with the slides, (2) wash slides, then add a fluorescent-labeled secondary antibody, and (3) wash, dry and scan slide in a fluorescent microscanner.

Figure 3: Procedure to profile serum samples and detect autoantigen markers

In addition to identification of autoantigens associated with autoimmune disorders, protein microarrays can identify autoantigens associated with cancers.  Once biomarkers identified using protein arrays are validated, they can potentially serve as new markers for the development of diagnostic tests. In certain cancers where early diagnosis is difficult, this can mean the difference between a treatable or a terminal disease. For example, in a paper published by Proceedings of the National Academy of Sciences (PNAS), the ProtoArray was used to probe serum samples from healthy women and women with ovarian cancer at various stages. 94 candidate biomarkers identified exhibited enhanced reactivity in sera from cancer patients. Researchers validated a subset of these candidate markers using immunoblot analysis and tissue microarrays and found that they demonstrated differential reactivity. 

We have demonstrated the utility of ProtoArray human protein microarrays for profiling human sera to identify protein biomarkers for disease. The immune response profiling assay employed is highly sensitive and reproducible and generates data consistent with other commonly utilised methods such as ELISA for biomarker identification. Panels of biomarkers discovered through ProtoArray immune response profiling studies are expected to facilitate earlier diagnosis and more effective treatment of numerous human diseases such as cancers and autoimmune diseases, leading to a critical paradigm shift towards personalised medicine.

Jennifer Cannon. Dr. Jennifer Cannon is a Business Area Manager at Invitrogen Corporation focusing on Invitrogen’s mass spectrometry products and protein microarray products and services. She received her Ph.D. in pharmacology from Johns Hopkins University.

By Jennifer Cannon. Dr. Jennifer Cannon is a Business Area Manager at Invitrogen Corporation focusing on Invitrogen’s mass spectrometry products and protein microarray products and services. She received her Ph.D. in pharmacology from Johns Hopkins University.

References
1. A. Fossa et al., (2004) Prostate 59, 440-7
2. S. S. Van Rhee F, Zhan F, Gupta SK, et al. (2005) Blood 105, 3939-3944.
3. Y. T. Chen et al., (1997) Proc Natl Acad Sci U S A 94, 1914-8.
4. U. Sahin et al., (1995)  Proc Natl Acad Sci U S A 92, 11810-3.
5. M. J. Nam et al., (2003) Proteomics 3, 2108-15.
6. F. M. Brichory et al., (2001)  Proc Natl Acad Sci U S A 98, 9824-9.
7. B. Schweitzer, P. Predki, M. Snyder, (2003) Proteomics 3, 2190-9.
8. W. H. Robinson, et al. (2002) Nat Med 8: 295–301.
9.  H. Zhu, et al. (2001) Science 293: 2101–2105.
10. P. Predki. (2004) Curr Opin Chem Biol 8: 8–13.
11. G. A. Michaud, et al. (2003) Nature Biotechnology 21(12): 1509-1512.
12. M. E. Hudson, et. al. (2007) Proc Natl Acad Sci U S A; 104, 17494-17499.