Challenges to developing proteomic-based breast cancer diagnostics

Abstract

Over the past decade, multiple genetic and histological approaches have accelerated development of new breast cancer diagnostics and treatment paradigms. Multiple distinct genetic subtypes of breast cancers have been defined, and this has progressively led toward more personalized medicine in regard to treatment options. There still remains a deficiency in the development of molecular diagnostic assays that can be used for breast cancer detection and pretherapy clinical decisions. In particular, the type of cancer-specific biomarker typified by a serum or tissue-derived protein. Progress in this regard has been minimal, especially in comparison to the rapid advancements in genetic and histological assays for breast cancers. In this review, some potential reasons for this large gap in developing protein biomarkers will be discussed, as well as new strategies for improving these approaches. Improvements in the study design of protein biomarker discovery strategies in relation to the genetic subtypes and histology of breast cancers is also emphasized. The current successes in use of genetic and histological assays for breast cancer diagnostics are summarized, and in that context, the current limitations of the types of breast cancer-related clinical samples available for protein biomarker assay development are discussed. Based on these limitations, research strategies emphasizing identification of glycoprotein biomarkers in blood and MALDI mass spectrometry imaging of tissues are described.

FIG. 1.

Concavalin A (Con A) lectin capture of serum glycoproteins. A nondisease human pooled serum sample (20 μL) was processed under different conditions, and the resulting proteins separated on an 8–16% SDS-gel, visualized by Coomassie blue staining. Lane 1, unprocessed serum, 1 μL loaded; lane 2, serum depleted of albumin and immunoglobulins using a ProteoPrep AI column from Sigma Chemicals; lane 3, serum depleted of abundant proteins using a Montage column from Millipore; lane 4, serum incubated with agarose bound Con A lectin. Bound glycoproteins were eluted with 0.2 mM alpha-methyl mannoside; lane 5, ProteoPrep AI depleted serum proteins incubated with Con A lectin, followed by elution; lane 6, montage-depleted serum proteins incubated with Con A lectin, followed by elution.

FIG. 2.

Staged breast cancer pooled sera with fucose lectin affinity capture (AAL) and (AAA). Pooled sera (20 μL) from each condition was incubated with the indicated agarose-bound lectin, aleuria aurantia (AAL), or anguilla anguilla (AAA), for 16 h. Bound glycoproteins were eluted with 0.2 mM L-Fucose, separated on an 8–16% SDS-gel and silver stained.

FIG. 3.

MALDI-TOF mass spectrometric analysis of human breast tissue with areas of invasive ductal carcinoma. A 10-μm slice of Sakura/UMFix preserved breast tissue was processed for MALDI imaging and coated with SPA matrix followed by spectra acqusition in linear mode on a Bruker UltraFlexIII MALDI-TOF/TOF mass spectrometer. Spectra were normalized and individual m/z values visualized using FlexAnalysis and FlexImaging software, as previously described (Cazares et al., 2009). (A) Representative histology image of H&E stained breast tissue showing pathologist determined areas of invasive ductal carcinoma in the highlighted ovals. (B) Colorimetric intensity of expression levels of a peptide at m/z 4,588. Areas of most intense expression are in red (see color bar for scale). (C) Differential color intensity comparison of two peaks at 4,588 m/z (red) in the invasive ductal tissue sections relative to a noncancer associated peak at 4,990 m/z (blue).