—  SYMPOSIUM #40  —

Hematopathology: New Technologies
Moderators: Dr. John Wing Chan and Dr. Thomas Grogan

Section 5 - Proteomics and Lymphoid Malignancies

Andrew L. Feldman


Proteins are the functional building blocks of the signaling mechanisms that relay information within and between cells. This information includes signals that trigger division, apoptosis, and recruitment of other cells that promote or inhibit cell survival. Proteins carry information along these signaling pathways much like electricity in an electric circuit; the information varies according to the type and amount of protein, the location (extracellular, cytoplasmic, nuclear), and post-translational modifications. Thus, proteomic analysis of tissues or body fluids offers a real-time window on the flow of cellular information in health and disease.

The analysis of protein expression using immunohistochemistry and flow cytometry has become a critical component of the classification of lymphoid malignancies. Proteomics, like other "-omics" technologies, differs from these types of protein analysis mostly in scale. Microarray technology has allowed high-throughput analysis of large numbers of analytes in multiple clinical specimens. New bioinformatic tools for analyzing and clustering these data have facilitated the detection of trends in large sample groups that would not be evident in individual cases. Gene expression microarrays have been employed in this manner to define new disease subtypes (e.g. the ABC and GCB subtypes of diffuse large B-cell lymphoma [1] ) and suggest new biologic factors in lymphoma (e.g. cyclin D1-negative mantle cell lymphoma [2] ). Similar evaluation at the proteomic level might be expected to yield even more functionally relevant information, since not all expressed genes are translated, and not all proteins are in an active configuration [3]. However, several technical limitations stand in the way of directly applying gene expression profiling methodology to proteomics. First, developing specific protein ligands is more difficult than generating complementary DNA sequences for binding; second, linear amplification of proteins for analysis of small samples has not yet been achieved; third, protein concentrations vary in abundance thousands of times more than RNA species, challenging the dynamic range of existing detection strategies; finally, the number of protein species (around 300,000) far exceeds the number of known genes (around 30,000) [1].

The two major protein detection strategies for proteomic analysis are ligand-based approaches and mass spectrometry-based approaches. Ligand-based approaches are discussed here. The most commonly used protein ligands are antibodies, but other protein ligands can be used for specific applications, including lectins, other proteins, and DNA. Development of antibodies against specific post-translational modifications (phosphorylation, specific cleavage sites) has allowed assessment of the activation status of specific proteins, information not available using genomic strategies. Using a combination of such antibodies, the activation status of entire signaling pathways can be assayed. This approach is promising for individualizing treatment of patients with molecular therapies that target specific pathways [4, 5, 6].

Immunohistochemistry and flow cytometry have begun to be used for proteomic-scale assessment of lymphoid neoplasms. Tissue microarrays have become a high-throughput strategy for validating gene expression array findings and investigating utility of new immunophenotyping panels, such as in diffuse large B-cell lymphoma [7, 8] and Hodgkin lymphoma [9]. Quantitative analysis of protein expression in tissue microarrays, including the ability to assay subcellular protein localization [10], may greatly facilitate profiling cell signaling pathways in paraffin embedded tissue. Flow cytometry has the advantage of being able to accurately assess multiple markers on individual cells. This is particularly important in lymphoid neoplasms, where the neoplastic cells often are admixed with numerous non-neoplastic cells that contribute significantly to overall protein levels in tissue lysates. The use of unsupervised clustering algorithms [11], single-cell phosphoproteomics [12], and microfluidic technologies for flow cytometry [13] are expected to advance proteomic applications of this field significantly.

As with gene expression analysis, the microarray format is conducive to proteomic profiling [14]. Though the format of protein microarrays varies, all have in common the detection of colocalized protein and ligand on a solid phase. Forward-phase and reverse-phase protein microarrays have been described. In forward-phase arrays, multiple ligands are immobilized on the solid phase and each array is probed with one or more protein samples. For example, multiple antibodies can be spotted on the solid phase, and the array can be probed with labeled protein samples. Ghobrial et al. used this approach to profile protein expression in mantle cell lymphoma. Not all findings could be confirmed by downstream validation techniques such as immunoblotting, indicating some potential limitations of this approach. Labeling protein for array-based detection can be problematic. DNA generally is labeled by incorporating labeled nucleotides during replication. Since a similar method is not available for protein amplification it is the native protein that requires labeling. Incubating the sample with excess label can damage the protein itself. Using insufficient label can lead to preferential labeling of large and/or high abundance proteins, leaving low abundance proteins either unlabeled or inconsistently labeled. Sandwich-type forward-phase microarrays avoid the protein labeling problem. This method also uses immobilized antibody to capture proteins in the sample. These are detected by a second antibody, which then is quantitated using standard detection systems as in an enzyme-linked immunosorbent assay (ELISA). In practice, this latter step can lack specificity if too many proteins are analyzed simultaneously. This approach, as well as related bead-based strategies, have been applied to quantitation of cytokines [15] and in principle is broadly extendable for use with many other types of proteins. One of the limitations of this approach is that the validation of antibody pairs for each protein of interest and testing for multiplex cross-reactivity adds a level of complexity that makes it difficult to assay large numbers of proteins.

Reverse-phase microarrays (RPMs) transcend some of the above limitations. In this technique, protein lysates are spotted onto the solid phase, such as a nitrocellulose-coated glass slide. Each array therefore contains numerous patient samples. Replicates of each sample typically are spotted in serial dilution curves, improving accuracy tremendously. Depending on the type of arrayer used and diameter of the spotting pin, very small volumes of lysate are used (e.g. 20 nL/spot). Thus this technique allows proteomic analysis of very small samples, and has enabled analysis of samples obtained using laser capture microdissection (LCM) [16] containing 20,000 cells or less [17]. Thus, there can be more selectivity in the cell population to be analyzed than can be achieved with forward-phase arrays. Each array is currently probed with a single antibody, though the potential for multiplexing exists [18]. Because hundreds of replicate arrays can be printed, the ability to measure numerous proteins is enhanced. Unprobed arrays can be stored for years, allowing the analysis of additional proteins as new antibodies or scientific information becomes available. Our group has used RPMs to analyze apoptotic pathways in follicular lymphoma (FL) and follicular hyperplasia, using LCM to enrich for follicle center cells. In two separate studies [17, 19], increased activation of Akt was found in follicular lymphoma; this increase did not correlate with Bcl-2 expression, suggesting that Akt signaling may represent a second important antiapoptotic pathway in FL. In addition, elevated Bcl-2/Bax and Bcl-2/Bak ratios were found to be associated with poor overall survival in FL patients [17]. The prognostic significance of the Bcl-2/Bax ratio present only when LCM was used and not when lysates of whole tissue sections were analyzed (unpublished data). This finding emphasizes the importance of correlating molecular profiles with histology and using tools to enrich for cells of interest when applicable.

The ability to analyze proteomic patterns and activation status of signaling pathways in tissue samples will facilitate the real-time monitoring of patients before, during, and after receiving targeted molecular therapies [6]. The increasing availability of functionally specific targeted therapies (e.g. tyrosine kinase inhibitors) will extend the importance of being able to assay multiple pathways to determine the ideal drug(s) for an individual patient's tumor. During therapy confirmatory proteomic analysis can be performed to make sure the targeted pathway has been successfully turned off. Reanalysis at the time of recurrence may help guide further therapy, since the originally targeted proteins may no longer be robustly expressed after treatment [20]. Because of the importance of correlating molecular profiles with histologic findings, hematopathologists need to become integrally involved in the design of clinical trials aimed at validating this model of individualized molecular medicine.

References
  1. Alizadeh, A.A., et al., Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature, 2000. 403(6769): p. 503-11.

  2. Fu, K., et al., Cyclin D1-negative mantle cell lymphoma: a clinicopathologic study based on gene expression profiling. Blood, 2005. 106(13): p. 4315-21.

  3. Braziel, R.M., et al., Molecular diagnostics. Hematology (Am Soc Hematol Educ Program), 2003: p. 279-93.

  4. Grubb, R.L., et al., Signal pathway profiling of prostate cancer using reverse phase protein arrays. Proteomics, 2003. 3(11): p. 2142-6.

  5. Sheehan, K.M., et al., Use of reverse phase protein microarrays and reference standard development for molecular network analysis of metastatic ovarian carcinoma. Mol Cell Proteomics, 2005. 4(4): p. 346-55.

  6. Liotta, L.A., E.C. Kohn, and E.F. Petricoin, Clinical proteomics: personalized molecular medicine. Jama, 2001. 286(18): p. 2211-4.

  7. Hans, C.P., et al., Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood, 2004. 103(1): p. 275-82.

  8. Zu, Y., et al., Validation of tissue microarray immunohistochemistry staining and interpretation in diffuse large B-cell lymphoma. Leuk Lymphoma, 2005. 46(5): p. 693-701.

  9. McCune, R.C., S.I. Syrbu, and M.A. Vasef, Expression profiling of transcription factors Pax-5, Oct-1, Oct-2, BOB.1, and PU.1 in Hodgkin's and non-Hodgkin's lymphomas: a comparative study using high throughput tissue microarrays. Mod Pathol, 2006.

  10. McCabe, A., et al., Automated quantitative analysis (AQUA) of in situ protein expression, antibody concentration, and prognosis. J Natl Cancer Inst, 2005. 97(24): p. 1808-15.

  11. Habib, L.K. and W.G. Finn, Unsupervised immunophenotypic profiling of chronic lymphocytic leukemia. Cytometry B Clin Cytom, 2006. 70(3): p. 124-35.

  12. Irish, J.M., N. Kotecha, and G.P. Nolan, Mapping normal and cancer cell signalling networks: towards single-cell proteomics. Nat Rev Cancer, 2006. 6(2): p. 146-55.

  13. Huh, D., et al., Microfluidics for flow cytometric analysis of cells and particles. Physiol Meas, 2005. 26(3): p. R73-98.

  14. Liotta, L.A., et al., Protein microarrays: meeting analytical challenges for clinical applications. Cancer Cell, 2003. 3(4): p. 317-25.

  15. Lash, G.E., et al., Comparison of three multiplex cytokine analysis systems: Luminex, SearchLight and FAST Quant. J Immunol Methods, 2006. 309(1-2): p. 205-8.

  16. Emmert-Buck, M.R., et al., Laser capture microdissection. Science, 1996. 274(5289): p. 998-1001.

  17. Gulmann, C., et al., Proteomic analysis of apoptotic pathways reveals prognostic factors in follicular lymphoma. Clin Cancer Res, 2005. 11(16): p. 5847-55.

  18. Geho, D., et al., Pegylated, steptavidin-conjugated quantum dots are effective detection elements for reverse-phase protein microarrays. Bioconjug Chem, 2005. 16(3): p. 559-66.

  19. Zha, H., et al., Similarities of prosurvival signals in Bcl-2-positive and Bcl-2-negative follicular lymphomas identified by reverse phase protein microarray. Lab Invest, 2004. 84(2): p. 235-44.

  20. Seliem, R.M., et al., Immunophenotypic changes and clinical outcome in B-cell lymphomas treated with rituximab. Appl Immunohistochem Mol Morphol, 2006. 14(1): p. 18-23.