In the present study, we used 2-D differential gel electrophoresis (2-D DIGE) and MS to screen biomarker candidates in serum samples obtained from 39 patients with breast cancer and 35 controls. First, we pooled the serum samples matched with age and menopausal status. Then, we depleted the two most abundant proteins albumin and IgG by immunoaffinity chromatography under partly denaturing conditions in order to enrich low-abundance proteins and proteins with low molecular weight. Concentrated and desalted samples were labeled with three different CyDyes including one internal standard, pooled from all the samples, and separated with 2-D DIGE in triplicate experiments. Biological variations of the protein expression level were analyzed with DeCyder software and evaluated for reproducibility and statistical significance. The profile of differentially expressed protein spots between patients and controls revealed proapolipoprotein A-I, transferrin, and hemoglobin as up-regulated and three spots, apolipoprotein A-I, apolipoprotein C-III, and haptoglobin alpha2 as down-regulated in patients. Finally, routine clinical immunochemical reactions were used to validate selected candidate biomarkers by quantitative determination of specific proteins in all individual serum samples. The serum level of transferrin correlated well with the 2-D-DIGE results. However, the serum levels of apolipoprotein A-I and haptoglobin could not be detected with the clinical routine diagnostic tests. This demonstrated an advantage 2-D DIGE still has over other techniques. 2-D DIGE can distinguish between isoforms of proteins, where the overall immunochemical quantification does fail due to a lack of isoform-special antibodies.

Insulin-like growth factor (IGF) signaling is a key regulator of breast development and breast cancer. We have analyzed the expression of the IGF signaling cascade in 17 human breast cancer and 4 mammary epithelial cell lines. Five cell lines expressed high levels of IGF1 receptor, insulin (INS)/IGF receptor substrate 1, IGF-binding proteins 2 and 4, as well as the estrogen receptor (ESR), indicating a co-activation of IGF and ESR signaling. Next, we stably overexpressed IGF1 and IGF2 in MCF7 breast cancer cells, which did not affect their epithelial characteristics and the expression and localization of the epithelial marker genes E-cadherin and beta-catenin. Conversely, IGF1 and IGF2 overexpression potently increased cellular proliferation rates and the efficiency of tumor formation in mouse xenograft experiments, whereas the resistance to chemotherapeutic drugs such as taxol was unaltered. Expression profiling of overexpressing cells with whole-genome oligonucleotide microarrays revealed that 21 genes were upregulated >2-fold by both IGF1 and IGF2, 9 by IGF1, and 9 by IGF2. Half of the genes found to be upregulated are involved in transport and biosynthesis of amino acids, including several amino acid transport proteins, argininosuccinate and asparagine synthetases, and methionyl-tRNA synthetase. Upregulation of these genes constitutes a novel mechanism apparently contributing to the stimulatory effects of IGF signaling on the global protein synthesis rate. We conclude that the induction of cell proliferation and tumor formation by long-term IGF stimulation may primarily be due to anabolic effects, in particular increased amino acid production and uptake.

We present a simple protocol for affinity depletion to remove the two most abundant serum proteins, albumin and immunoglobulin G (IgG). Under native conditions, albumin/IgG were efficiently removed and several proteins were enriched as shown by two-dimensional electrophoresis (2-DE). Besides that, partly denaturing conditions were established by adding 5 or 20% acetonitrile (ACN) in order to disrupt the binding of low-molecular-weight (LMW) proteins to the carrier proteins albumin/IgG. 2-DE results showed that the total number of detected LMW proteins increased under denaturing conditions when compared to native conditions. Interestingly, the presence of 5% ACN in serum revealed better enrichment of LMW proteins when compared to 20% ACN condition. Seven randomly distributed spots in albumin/IgG depleted serum samples under 5% ACN condition were picked from the 2-DE gels and identified by mass spectrometry (MS). The intensity of five LMW protein spots increased under denaturing conditions when compared to native conditions. Three of the seven identified spots (serum amyloid P, vitamin D-binding protein, and transthyretin) belong to a group of relatively low-abundant proteins, which make up only 1% of all serum proteins. The method presented here improves the resolution of the serum proteome by increasing the number of visualized spots on 2-D gels and allowing the detection and MS identification of LMW proteins and proteins of lower abundance.

Purpose: Insulin-like growth factors (IGFs) are potent mitogens for breast cancer cells in vitro, and elevated IGF-I serum levels are a risk factor for breast malignancies. This study evaluated IGF-I and IGF-II serum levels in healthy women and in patients with benign and malignant breast lesions and correlated them with tumor size.Experimental Design: Serum levels of the total and unbound fractions of IGF-I and IGF-II were analyzed in 65 patients with benign and malignant breast lesions and in 38 women without breast disease. ELISAs were used to detect serum IGF levels, with (total IGF) or without (free IGF) prior acid-ethanol extraction.Results: Total IGF-I serum concentrations were lower in healthy women than in breast cancer patients (P < 0.001) or patients with benign breast lesions (P ‫؍‬ 0.010), but no differences were observed in free IGF-I levels. Conversely, healthy women had higher serum levels of free IGF-II than women with breast lesions (P ‫؍‬ 0.003), and the free/total IGF-II ratio was significantly reduced in patients with breast disease (P ‫؍‬ 0.001). Although IGF-I or IGF-II serum concentrations of breast cancer patients were similar to those of patients with benign lesions, the size of a malignant tumor was correlated to the ratio free/total IGF-II (P ‫؍‬ 0.002).Conclusions: Malignant breast tumors cannot be distinguished from benign breast lesions by systemic IGF serum levels. However, women with breast lesions have decreased IGF-II concentrations, and free IGF-II levels are clearly correlated to the size of a breast cancer, indicating an involvement in tumor growth.

The L10P single nucleotide polymorphism (SNP) is located in the signal sequence of the transforming growth factor β1 (TGFβ1) gene. The proline-encoding (Pro-) allele of this SNP has been associated with an increased breast cancer risk, which has been attributed to the elevated secretion of this TGFβ1 variant observed in vitro and in male subjects. Here we investigated the association of the L10P SNP with serum levels of TGFβ1 in female breast cancer patients and controls. We genotyped the L10P SNP in 276 breast cancer patients and 255 controls. Serum TGFβ1 concentrations were measured by enzyme-linked immunosorbent assay (ELISA) in a subset of the study population (n = 211). We found no evidence for an association of the L10P SNP with breast cancer risk (per-allele odds ratio: 0.91; 95% confidence interval: 0.71–1.16). However, patients with the Pro/Pro genotype exhibited a significantly younger age at breast cancer onset (55.2 ± 14.3 years) than Leu/Leu patients (60.6 ± 13.6 years; p = 0.04), which may reflect the ability of TGFβ to promote tumor progression. Mean TGFβ1 serum levels of Pro-allele carriers were 39.4 ± 7.4 ng/mL, whereas those of Leu/Leu subjects were 37.6 ± 6.0 ng/mL (p = 0.07). Thus, compared to a previous study of male subjects, we observed only a modest increase, if any, in TGFβ1 levels of female Pro-allele carriers.