Minimally invasive biopsy-based diagnostics in support of precision cancer medicine

Precision cancer medicine (PCM) to support the treatment of solid tumors requires minimally invasive diagnostics. Here, we describe the development of fine-needle aspiration biopsy-based (FNA) molecular cytology which will be increasingly important in diagnostics and adaptive treatment. We provide support for FNA-based molecular cytology having a significant potential to replace core needle biopsy (CNB) as a patient-friendly potent technique for tumor sampling for various tumor types. This is not only because CNB is a more traumatic procedure and may be associated with more complications compared to FNA-based sampling, but also due to the recently developed molecular methods used with FNA. Recent studies show that image-guided FNA in combination with ultrasensitive molecular methods also offers opportunities for characterization of the tumor microenvironment which can aid therapeutic decisions. Here we provide arguments for an increased implementation of molecular FNA-based sampling as a patient-friendly diagnostic method, which may, due to its repeatability, facilitate regular sampling that is needed during different treatment lines, to provide tumor information, supporting treatment decisions, shortening lead times in healthcare, and benefit healthcare economics.

Multiplex protein analysis and ensemble machine learning methods of fine needle aspirates from prostate cancer patients reveal potential diagnostic signatures associated with tumour grade

BACKGROUND

Improved molecular diagnosis is needed in prostate cancer (PC). Fine needle aspiration (FNA) is a minimally invasive biopsy technique, less traumatic compared to core needle biopsy, and could be useful for diagnosis of PC. Molecular biomarkers (BMs) in FNA-samples can be assessed for prediction, eg of immunotherapy efficacy before treatment as well as at treatment decision time points during disease progression.

METHODS

In the present pilot study, the expression levels of 151 BM proteins were analysed by proximity extension assay in FNA-samples from 16 patients, including benign prostate lesions (n = 3) and cancers (n = 13). An ensemble data analysis strategy was applied using several machine learning models.

RESULTS

Twelve potentially predictive BM proteins correlating with International Society of Urological Pathology grade groups were identified, among them vimentin, tissue factor pathway inhibitor 2, and integrin beta-5. The validity of the results was supported by network analysis that showed functional associations between most of the identified putative BMs. We also showed that multiple immune checkpoint targets can be assessed (eg PD-L1, CD137, and Galectin-9), which may support the selection of immunotherapy in advanced PC. Results are promising but need further validation in a larger cohort.

CONCLUSIONS

Our pilot study represents a "proof of concept" and shows that multiplex profiling of potential diagnostic and predictive BM proteins is feasible on tumour material obtained by FNA sampling of prostate cancer. Moreover, our results demonstrate that an ensemble data analysis strategy may facilitate the identification of BM signatures in pilot studies when the patient cohort is limited.

Profiling of extracellular vesicles of metastatic urothelial cancer patients to discover protein signatures related to treatment outcome

The prognosis of metastatic urothelial carcinoma (mUC) patients is poor, and early prediction of systemic therapy response would be valuable to improve outcome. In this exploratory study, we investigated protein profiles in sequential plasma‐isolated extracellular vesicles (EVs) from a subset of mUC patients treated within a Phase I trial with vinflunine combined with sorafenib. The isolated EVs were of exosome size and expressed exosome markers CD9, TSG101 and SYND‐1. We found, no association between EVs/ml plasma at baseline and progression‐free survival (PFS). Protein profiling of EVs, using an antibody‐based 92‐plex Proximity Extension Assay on the Oncology II^® platform, revealed a heterogeneous protein expression pattern. Qlucore bioinformatic analyses put forward a protein signature comprising of SYND‐1, TNFSF13, FGF‐BP1, TFPI‐2, GZMH, ABL1 and ERBB3 to be putatively associated with PFS. Similarly, a protein signature from EVs that related to best treatment response was found, which included FR‐alpha, TLR 3, TRAIL and FASLG. Several of the markers in the PFS or best treatment response signatures were also identified by a machine learning classification algorithm. In conclusion, protein profiling of EVs isolated from plasma of mUC patients shows a potential to identify protein signatures that may associate with PFS and/or treatment response.

Multiplex immune protein profiling of fine-needle aspirates from patients with non-small-cell lung cancer reveals signatures associated with PD-L1 expression and tumor stage

Biomarker signatures identified through minimally invasive procedures already at diagnosis of non‐small‐cell lung cancer (NSCLC) could help to guide treatment with immune checkpoint inhibitors (ICI). Here, we performed multiplex profiling of immune‐related proteins in fine‐needle aspirate (FNA) samples of thoracic lesions from patients with NSCLC to assess PD‐L1 expression and identify related protein signatures. Transthoracic FNA samples from 14 patients were subjected to multiplex antibody‐based profiling by proximity extension assay (PEA). PEA profiling employed protein panels relevant to immune and tumor signaling and was followed by Qlucore^® Omics Explorer analysis. All lesions analyzed were NSCLC adenocarcinomas, and PEA profiles could be used to monitor 163 proteins in all but one sample. Multiple key immune signaling components (including CD73, granzyme A, and chemokines CCL3 and CCL23) were identified and expression of several of these proteins (e.g., CCL3 and CCL23) correlated to PD‐L1 expression. We also found EphA2, a marker previously linked to inferior NSCLC prognosis, to correlate to PD‐L1 expression. Our identified protein signatures related to stage included, among others, CXCL10 and IL12RB1. We conclude that transthoracic FNA allows for extensive immune and tumor protein profiling with assessment of putative biomarkers of important for ICI treatment selection in NSCLC.

Super-resolution microscopy can identify specific protein distribution patterns in platelets incubated with cancer cells

Protein contents in platelets are frequently changed upon tumor development and metastasis. However, how cancer cells can influence protein-selective redistribution and release within platelets, thereby promoting tumor development, remains largely elusive. With fluorescence-based super-resolution stimulated emission depletion (STED) imaging we reveal how specific proteins, implicated in tumor progression and metastasis, re-distribute within platelets, when subject to soluble activators (thrombin, adenosine diphosphate and thromboxane A2), and when incubated with cancer (MCF-7, MDA-MB-231, EFO21) or non-cancer cells (184A1, MCF10A). Upon cancer cell incubation, the cell-adhesion protein P-selectin was found to re-distribute into circular nano-structures, consistent with accumulation into the membrane of protein-storing alpha-granules within the platelets. These changes were to a significantly lesser extent, if at all, found in platelets incubated with normal cells, or in platelets subject to soluble platelet activators. From these patterns, we developed a classification procedure, whereby platelets exposed to cancer cells, to non-cancer cells, soluble activators, as well as non-activated platelets all could be identified in an automatic, objective manner. We demonstrate that STED imaging, in contrast to electron and confocal microscopy, has the necessary spatial resolution and labelling efficiency to identify protein distribution patterns in platelets and can resolve how they specifically change upon different activations. Combined with image analyses of specific protein distribution patterns within the platelets, STED imaging can thus have a role in future platelet-based cancer diagnostics and therapeutic monitoring. The presented approach can also bring further clarity into fundamental mechanisms for cancer cell-platelet interactions, and into non-contact cell-to-cell interactions in general.

Protein profiling of fine-needle aspirates reveals subtype-associated immune signatures and involvement of chemokines in breast cancer

There are increasing demands for informative cancer biomarkers, accessible via minimally invasive procedures, both for initial diagnostics and for follow-up of personalized cancer therapy, including immunotherapy. Fine-needle aspiration (FNA) biopsy provides ready access to relevant tissue samples; however, the minute amounts of sample require sensitive multiplex molecular analysis to be of clinical biomarker utility. We have applied proximity extension assays (PEA) to analyze 167 proteins in FNA samples from patients with breast cancer (BC; n = 25) and benign lesions (n = 32). We demonstrate that the FNA BC samples could be divided into two main clusters, characterized by differences in expression levels of the estrogen receptor (ER) and the proliferation marker Ki67. This clustering corresponded to some extent to established BC subtypes. Our analysis also revealed several proteins whose expression levels differed between BC and benign lesions (e.g., CA9, GZMB, IL-6, VEGFA, CXCL11, PDL1, and PCD1), as well as several chemokines correlating with ER and Ki67 status (e.g., CCL4, CCL8, CCL20, CXCL8, CXCL9, and CXCL17). Finally, we also identified three signatures that could predict Ki67 status, ER status, and tumor grade, respectively, based on a small subset of proteins, which was dominated by chemokines. To our knowledge, expression profiles of CCL13 in benign lesions and BC have not previously been described but were shown herein to correlate with proliferation (P = 0.00095), suggesting a role in advanced BC. Given the broad functional range of the proteins analyzed, immune-related proteins were overrepresented among the observed alterations. Our pilot study supports the emerging role of chemokines in BC progression. Due to the minimally traumatic sampling and clinically important molecular information for therapeutic decisions, this methodology is promising for future immunoscoring and monitoring of treatment efficacy in BC.

A fine-needle aspiration-based protein signature discriminates benign from malignant breast lesions

There are increasing demands for informative cancer biomarkers, accessible via minimally invasive procedures, both for initial diagnostics and to follow-up personalized cancer therapy. Fine-needle aspiration (FNA) biopsy provides ready access to relevant tissues; however, the minute sample amounts require sensitive multiplex molecular analysis to achieve clinical utility. We have applied proximity extension assays (PEA) and NanoString (NS) technology for analyses of proteins and of RNA, respectively, in FNA samples. Using samples from patients with breast cancer (BC, n = 25) or benign lesions (n = 33), we demonstrate that these FNA-based molecular analyses (a) can offer high sensitivity and reproducibility, (b) may provide correct diagnosis in shorter time and at a lower cost than current practice, (c) correlate with results from routine analysis (i.e., benchmarking against immunohistochemistry tests for ER, PR, HER2, and Ki67), and (d) may also help identify new markers related to immunotherapy. A specific 11-protein signature, including FGF binding protein 1, decorin, and furin, distinguished all cancer patient samples from all benign lesions in our main cohort and in smaller replication cohort. Due to the minimally traumatic sampling and rich molecular information, this combined proteomics and transcriptomic methodology is promising for diagnostics and evaluation of treatment efficacy in BC.

Platelet protein biomarker panel for ovarian cancer diagnosis

BACKGROUND

Platelets support cancer growth and spread making platelet proteins candidates in the search for biomarkers.

METHODS

Two-dimensional (2D) gel electrophoresis, Partial Least Squares Discriminant Analysis (PLS-DA), Western blot, DigiWest.

RESULTS

PLS-DA of platelet protein expression in 2D gels suggested differences between the International Federation of Gynaecology and Obstetrics (FIGO) stages III-IV of ovarian cancer, compared to benign adnexal lesions with a sensitivity of 96% and a specificity of 88%. A PLS-DA-based model correctly predicted 7 out of 8 cases of FIGO stages I-II of ovarian cancer after verification by western blot. Receiver-operator curve (ROC) analysis indicated a sensitivity of 83% and specificity of 76% at cut-off >0.5 (area under the curve (AUC) = 0.831, p < 0.0001) for detecting these cases. Validation on an independent set of samples by DigiWest with PLS-DA differentiated benign adnexal lesions and ovarian cancer, FIGO stages III-IV, with a sensitivity of 70% and a specificity of 83%.

CONCLUSION

We identified a group of platelet protein biomarker candidates that can quantify the differential expression between ovarian cancer cases as compared to benign adnexal lesions.

Protein levels of clusterin and glutathione synthetase in platelets allow for early detection of colorectal cancer

Colorectal cancer (CRC) is one of the most frequent malignancies in the Western world. Early tumor detection and intervention are important determinants on CRC patient survival. During early tumor proliferation, dissemination and angiogenesis, platelets store and segregate proteins actively and selectively. Hence, the platelet proteome is a potential source of biomarkers denoting early malignancy. By comparing protein profiles of platelets between healthy volunteers (n = 12) and patients with early- (n = 7) and late-stage (n = 5) CRCs using multiplex fluorescence two-dimensional gel electrophoresis (2D-DIGE), we aimed at identifying differentially regulated proteins within platelets. By inter-group comparisons, 94 differentially expressed protein spots were detected (p < 0.05) between healthy controls and patients with early- and late-stage CRCs and revealed distinct separations between all three groups in principal component analyses. 54 proteins of interest were identified by mass spectrometry and resulted in high-ranked Ingenuity Pathway Analysis networks associated with Cellular function and maintenance, Cellular assembly and organization, Developmental disorder and Organismal injury and abnormalities (p < 0.0001 to p = 0.0495). Target proteins were validated by multiplex fluorescence-based Western blot analyses using an additional, independent cohort of platelet protein samples [healthy controls (n = 15), early-stage CRCs (n = 15), late-stage CRCs (n = 15)]. Two proteins-clusterin and glutathione synthetase (GSH-S)-featured high impact and were subsequently validated in this independent clinical cohort distinguishing healthy controls from patients with early- and late-stage CRCs. Thus, the potential of clusterin and GSH-S as platelet biomarkers for early detection of CRC could improve existing screening modalities in clinical application and should be confirmed in a prospective multicenter trial.

Artemin-GFRα3 interactions partially contribute to acute inflammatory hypersensitivity

The expression of artemin (ARTN), a glial cell line-derived neurotrophic factor (GDNF) family ligand, increases in pre-clinical models of nociception and recent evidence suggests this growth factor may play a causative role in inflammatory pain mechanisms. The aim of this study was to demonstrate functional inhibition of ARTN with monoclonal antibodies and to determine whether ARTN neutralisation could reverse inflammatory pain in mice. We show that monoclonal antibodies with high affinity to ARTN, completely inhibit ARTN-induced Ret and ERK activation in a human neuroblastoma cell line, and block capsaicin-induced CGRP secretion from primary rat DRG cultures. In addition, administration of anti-ARTN antibodies to mice provides a transient, partial reversal (41%) of FCA-induced mechanical hypersensitivity. Anti-ARTN antibodies had no effect on hypersensitivity in response to partial nerve ligation in mice. These data suggest that ARTN-GFRα3 interactions partially mediate early stage nociceptive signalling following an inflammatory insult.

In vitro pharmacological characterization of a novel TRPA1 antagonist and proof of mechanism in a human dental pulp model

AZ465 is a novel selective transient receptor potential cation channel, member A1 (TRPA1) antagonist identified during a focused drug discovery effort. In vitro, AZ465 fully inhibits activation by zinc, O-chlorobenzylidene malononitrile (CS), or cinnamaldehyde of the human TRPA1 channel heterologously expressed in human embryonic kidney cells. Our data using patch-clamp recordings and mouse/human TRPA1 chimeras suggest that AZ465 binds reversibly in the pore region of the human TRPA1 channel. Finally, in an ex vivo model measuring TRPA1 agonist-stimulated release of neuropeptides from human dental pulp biopsies, AZD465 was able to block 50%–60% of CS-induced calcitonin gene-related peptide release, confirming that AZ465 inhibits the native human TRPA1 channel in neuronal tissue.

Multivariate meta-analysis of proteomics data from human prostate and colon tumours

Background

There is a vast need to find clinically applicable protein biomarkers as support in cancer diagnosis and tumour classification. In proteomics research, a number of methods can be used to obtain systemic information on protein and pathway level on cells and tissues. One fundamental tool in analysing protein expression has been two-dimensional gel electrophoresis (2DE). Several cancer 2DE studies have reported partially redundant lists of differently expressed proteins. To be able to further extract valuable information from existing 2DE data, the power of a multivariate meta-analysis will be evaluated in this work.

Results

We here demonstrate a multivariate meta-analysis of 2DE proteomics data from human prostate and colon tumours. We developed a bioinformatic workflow for identifying common patterns over two tumour types. This included dealing with pre-processing of data and handling of missing values followed by the development of a multivariate Partial Least Squares (PLS) model for prediction and variable selection. The variable selection was based on the variables performance in the PLS model in combination with stability in the validation. The PLS model development and variable selection was rigorously evaluated using a double cross-validation scheme. The most stable variables from a bootstrap validation gave a mean prediction success of 93% when predicting left out test sets on models discriminating between normal and tumour tissue, common for the two tumour types. The analysis conducted in this study identified 14 proteins with a common trend between the tumour types prostate and colon, i.e. the same expression profile between normal and tumour samples.

Conclusions

The workflow for meta-analysis developed in this study enabled the finding of a common protein profile for two malign tumour types, which was not possible to identify when analysing the data sets separately.

Identification of novel candidate protein biomarkers for the post-polio syndrome - implications for diagnosis, neurodegeneration and neuroinflammation

Survivors of poliomyelitis often develop increased or new symptoms decades after the acute infection, a condition known as post-polio syndrome (PPS). The condition affects 20-60% of previous polio patients, making it one of the most common causes of neurological deficits worldwide. The underlying pathogenesis is not fully understood and accurate diagnosis is not feasible. Herein we investigated whether it was possible to identify proteomic profile aberrations in the cerebrospinal fluid (CSF) of PPS patients. CSF from 15 patients with well-defined PPS were analyzed for protein expression profiles. The results were compared to data obtained from nine healthy controls and 34 patients with other non-inflammatory diseases which served as negative controls. In addition, 17 samples from persons with secondary progressive multiple sclerosis (SPMS) were added as relevant age-matched references for the PPS samples. The CSF of persons with PPS displayed a disease-specific and highly predictive (p=0.0017) differential expression of five distinct proteins: gelsolin, hemopexin, peptidylglycine alpha-amidating monooxygenase, glutathione synthetase and kallikrein 6, respectively, in comparison with the control groups. An independent ELISA confirmed the increase of kallikrein 6. We suggest that these five proteins should be further evaluated as candidate biomarkers for the diagnosis and development of new therapies for PPS patients.

Undifferentiated pelvic adenocarcinomas: diagnostic potential of protein profiling and multivariate analysis

BACKGROUND AND AIMS

Despite improved techniques, the determination of tumor origin in poorly differentiated adenocarcinomas still remains a challenge for the pathologist. Here we report the use of protein profiling combined with principal component analysis to improve diagnostic decision-making in tumor samples, in which standard pathologic investigations cannot present reliable results.

MATERIALS AND METHODS

A poorly differentiated adenocarcinoma of unknown origin located in the pelvis, infiltrating the sigmoid colon as well as the ovary, served as a model to evaluate our proteomic approach. Firstly, we characterized the protein expression profiles from eight advanced colon and seven ovarian adenocarcinomas using two-dimensional gel electrophoresis (2-DE). Qualitative and quantitative patterns were recorded and compared to the tumor of unknown origin. Based on these protein profiles, match sets from the different tumors were created. Finally, a multivariate principal component analysis was applied to the entire 2-DE data to disclose differences in protein patterns between the different tumors.

RESULTS

Over 89% of the unknown tumor sample spots could be matched with the colon standard gel, whereas only 63% of the spots could be matched with the ovarian standard. In addition, principal component analysis impressively displayed the clustering of the unknown case within the colon cancer samples, whereas this case did not cluster at all within the group of ovarian adenocarcinomas.

CONCLUSION

These results show that 2-DE protein expression profiling combined with principal component analysis is a sensitive method for diagnosing undifferentiated adenocarcinomas of unknown origin. The described approach can contribute greatly to diagnostic decision-making and, with further technical improvements and a higher throughput, become a powerful tool in the armentarium of the pathologist.

Proteomics in prostate cancer research

The incidence of early prostate cancer (PCa) among middle-aged men has increased rapidly. For many of these men, curatively intended treatment does more harm than good. Established prognostic factors are tumor stage and grade. As a result of earlier detection a majority of patients now have nonpalpable tumors (T1c) of intermediate grade (Gleason score 6). Prostate specific antigen in serum in such cases is generally at a low level and not a reliable predictor of prognosis. Altogether there is an urgent need for adjunctive prognostic indicators. In the search for relevant tumor markers for improved patient selection an exploration of the proteome (the human proteins) could be fruitful. This paper critically reviews the use of 2-dimensional gel electrophoresis (2-DE) for proteome research. Additional steps such as image analysis and mass spectrometry are described. Techniques based on non-2-DE platforms: surface-enhanced laser desorption/ionization (SELDI), isotope coded affinity tags (ICAT) and array-based technologies are also summarized. Although labor-intensive and time-consuming, 2-DE is presently the most powerful method for analysis of cellular protein phenotype and may potentially reveal gene regulations that cannot be detected on a genetic level.

Frozen tissue biobanks. Tissue handling, cryopreservation, extraction, and use for proteomic analysis

The potential and the limitations of protein analysis of tissue samples are surveyed. The complexity, concentration range and dynamics of the human proteome are reviewed, as is the effect of handling and cryopreservation. Protein extraction, solubilization, resolution and detection are discussed, in relation to the properties of the human proteome.

Correlation of protein expression, Gleason score and DNA ploidy in prostate cancer

The prognosis of prostate cancer correlates with tumor differentiation. Gleason score and DNA ploidy are two prognostic factors that correlate with prognosis. We analyzed differences in protein expression in prostate cancer of high and low aggressiveness according to these measures. From 35 prostatectomy specimens, 29 cancer samples and 10 benign samples were harvested by scraping cells from cut surfaces. DNA ploidy was assessed by image cytometry. Protein preparations from cell suspensions were examined by 2-DE. Protein spots that differed quantitatively between sample groups were identified by MS fingerprinting of tryptic fragments and MS/MS sequence analysis. We found 39 protein spots with expression levels that were raised or lowered in correlation with Gleason score and/or DNA ploidy pattern (31 overexpressed in high-malignant cancer, 8 underexpressed). Of these, 30 were identified by MS. Among overexpressed proteins were heat-shock, structural and membrane proteins and enzymes involved in gene silencing, protein synthesis/degradation, mitochondrial protein import (metaxin 2), detoxification (GST-pi) and energy metabolism. Stroma-associated proteins were generally underexpressed. The protein expression of prostate cancer correlates with tumor differentiation. Potential prognostic markers may be found among proteins that are differentially expressed and the clinical value of these should be validated.

Evaluation of two sample preparation methods for prostate proteome analysis

For laboratory techniques that require well-preserved proteins, such as 2-DE, fresh tissue must be harvested and processed as fast as possible to avoid proteolytic degradation. We describe a modified method for harvesting tissue from radical prostatectomy specimens for proteome analysis and compare it with the standard technique. Cells were scraped from cut surfaces of 11 prostate specimens. A fraction of the material was smeared on a glass slide and Giemsa stained for morphological control. The sample was collected in a medium with protease inhibitors, and the protein material was prepared for 2-DE. Filtering and Percoll centrifugation were omitted. Sample locations were noted on a specimen map. From the same area, a tissue block was harvested for comparison. The block was processed with the conventional technique including mechanical disintegration, filtering and Percoll centrifugation. Quality measures of 2-DE were similar with both methods. With the scrape sampling technique, control smears showed abundant epithelial cells and a cleaner background and processing was faster than with tissue block sampling. For proteomic analysis, the scrape sample technique has several advantages over the tissue block method.

Proteomic analysis of protein expression in prostate cancer

OBJECTIVE

To investigate protein expression in prostate cancer using 2-dimensional gel electrophoresis (2-DE) and mass spectrometry.

STUDY DESIGN

Cells were collected from 29 peripheral zone tumors and from benign tissue by scraping cut surfaces of radical prostatectomy specimens. Samples were suspended in a medium with protease inhibitors and prepared for 2-DE. Gels were analyzed, and protein spots that differed quantitatively between tumor and benign tissue were identified via mass spectrometric fingerprinting of tryptic fragments and tandem mass spectrometry sequence analysis.

RESULTS

In total, 63 spots differed between cancer and benign samples (p < 0.01); 56 were overexpressed (> 1.5 fold) in cancer and 7 underexpressed (< 0.6 fold). Among overexpressed proteins were transcription factors (nucleoside diphosphate kinase 1) and enzymes involved in gene silencing (chromobox protein), protein synthesis (39S ribosomal protein L12, BiP protein, protein disulfide isomerase), degradation (cytosol aminopeptidase, endopeptidase Clp, inorganic pyrophosphatase) and energy metabolism (acyl-CoA dehydrogenase, isocitrate dehydrogenase, NADH-ubiquinone oxidoreductase, pyruvate dehydrogenase), heat-shock proteins (60 and 70 kd), structural proteins (cytokeratins) and membrane proteins (stomatinlike protein 2).

CONCLUSION

The protein profile of prostate cancer differs from that of benign tissue. Several potential target proteins for detection or evaluation of prognosis in prostate cancer were identified.

Differential protein expression in anatomical zones of the prostate

The prostate has three anatomical zones: the peripheral (PZ), the transition (TZ), and the central (CZ) zone. It is proposed that the CZ may be of mesodermal origin, whereas the other two are of endodermal origin. Proteome patterns in the zones were characterized to test for differences. Cells were scraped from macroscopically normal areas of PZ, TZ, and CZ in radical prostatectomy specimens. After exclusion of samples with cancer or prostatic intraepithelial neoplasia, 18 cases remained for analysis. Cells were collected in a medium with protease inhibitors, and the protein material was prepared for two-dimensional gel electrophoresis. The proteins in spots that differed quantitatively between regions were identified via mass spectrometric fingerprinting of tryptic fragments and selected tandem mass spectrometry sequence analysis. Ten proteins with significant zonal differential expression were identified, eight with underexpression in the CZ versus the PZ and the TZ (arginase II, ATP synthase, cytokeratin 8, lamin A/C, peroxiredoxin 4, protein disulfide isomerase A3, tropomyosin, and vimentin), and two with overexpression in the CZ (peroxiredoxin 2 and creatine kinase B). The PZ and TZ, although differing in terms of incidence of cancer and hyperplasia, have epithelium with highly similar major protein expression profiles. However, the protein profile of the CZ differs from that of the other regions, suggesting functional differences.

Gene and protein expression profiling of the microvascular compartment in experimental autoimmune encephalomyelitis in C57Bl/6 and SJL mice

Dysfunction of the blood-brain barrier (BBB) is a hallmark of inflammatory diseases of the central nervous system (CNS) such as multiple sclerosis (MS) and its animal model, experimental autoimmune encephalomyelitis (EAE). The molecular mechanisms leading to BBB breakdown are not well understood. In order to find molecules involved in this process, we used oligonucleotide microarrays and proteomics to analyze gene and protein expression of the microvascular compartment isolated from brains of C57Bl/6 and SJL/N mice afflicted with EAE and the microvascular compartment isolated from healthy controls. Out of the 6500 known genes and expressed sequence tags (ESTs) studied, expression of 288 genes was found to be changed. Of these genes 128 were altered in the microvascular compartment in both EAE models. Six proteins were identified to be present at altered levels. In addition to the expected increased expression of genes coding for molecules involved in leukocyte recruitment, genes not yet ascribed to EAE pathogenesis were identified. Thus, proteomics and gene array screens of the microvascular compartment are valid approaches, that can be used to define novel candidate molecules involved in EAE pathogenesis at the level of the BBB.

Proteome analysis of skin distinguishes acute guttate from chronic plaque psoriasis

Psoriasis is a disease with considerable heterogeneity in clinical presentation. This is the first study using two-dimensional gel electrophoresis to compare global protein expression patterns in lesional and non-lesional skin from subjects with acute guttate psoriasis associated with streptococcal throat infection and chronic plaque psoriasis. Samples from experimentally induced contact eczema and normal skin from healthy controls were also included. Proteins with statistically significant differences in expression were used in hierarchical cluster analyses resulting in separation of the different samples into groups. Chronic plaque and guttate psoriasis samples were distinctly separated, indicating that they represent discrete phenotypes at the protein expression level. Interestingly, there was a trend in which guttate psoriasis lesions clustered closer to eczema than to chronic plaque psoriasis lesions, indicating that the duration of the inflammatory reaction may affect clustering. Several of the differentially expressed proteins were identified by mass spectrometry.

Extraction and proteomic analysis of proteins from normal and multiple sclerosis postmortem brain

In this study, a reproducible fractionation procedure was developed to reduce levels of the abundant cytoskeletal proteins that are present in normal and pathological central nervous system (CNS) tissues. The fractionation and proteomic analysis techniques employed greatly facilitated comparison of the spectrum of proteins in normal postmortem brain with proteins in samples from patients with multiple sclerosis, an inflammatory demyelinating disease in which complex changes in protein expression occur as lesions develop. This approach may be of value for the proteomic identification and quantitation of proteins which undergo disease-related changes in CNS disorders, and also for protein expression studies on normal adult and developing CNS tissues.

Sequential proteome alterations during genesis and progression of colon cancer

Changes in the proteome of colon mucosal cells accompany the transition from normal mucosa via adenoma and invasive cancer to metastatic disease. Samples from 15 patients with sporadic sigmoid cancers were analyzed. Proteins were separated by two-dimensional gel electrophoresis. Relative differences in expression levels between normal tissue, adenoma, carcinoma and metastasis were evaluated in both intra- and inter-patient comparisons. Up- and down-regulated proteins (> twofold) during development to cancer or metastasis were excised and submitted to peptide mass fingerprinting and MS/MS sequence analysis, facilitated by the use of a compact disc workstation. In total, 112 protein spots were found to be differentially regulated, of which 72 were determined as to protein identity, 46 being up-regulated toward the progression of cancer, and 26 down-regulated. Several of the identifications correlate with proteins of the cell cycle, cytoskeleton or metabolic pathways. The pattern changes now identified have the potential for design of marker panels for assistance in diagnostics and therapeutic strategies in colorectal cancer.

Dihydropyrimidinase related protein-2 as a biomarker for temperature and time dependentpost mortem changes in the mouse brain proteome

Proteome analysis in the central nervous system area represents a large and important challenge in drug discovery. One major problem is to obtain representative and well characterized tissues of high quality for analysis. We have used brain tissues from normal mice to study the effect of post mortem time (up to 32 h) and temperature (4 degrees C and room temperature) on protein expression patterns. A number of proteins were identified using mass spectrometry and potential markers were localized. One of the proteins identified, dihydropyrimidinase related protein-2 (DRP-2), occurs as multiple spots in two-dimensional electrophoresis gels. The ratio between the truncated form of DRP-2 (fDRP-2) and full length DRP-2 is suggested as an internal control that can be used as a biomarker of post mortem time and post mortem temperature between unrelated brain protein samples. Results of this study may be useful in future efforts to detect disease specific alterations in proteomic studies of human post mortem brain tissues.

Gene and protein expression profiling of human cerebral endothelial cells activated with tumor necrosis factor-alpha

An increase in permeability of the blood-brain barrier is a critical event in the pathophysiological process of multiple sclerosis and other neurodegenerative diseases. Tumor necrosis factor alpha (TNFalpha) is known to play a crucial role in this process and is a powerful activator of endothelial cell inflammatory responses. Although many reports describe effects of TNFalpha activation in endothelial cells, the molecular mechanisms specific for activation of cerebral endothelial cells remains unclear. The objective of this study was to identify potential pharmaceutical targets for the treatment of multiple sclerosis using molecular profiling techniques. Gene expression measurements (Affymetrix Hu6800 oligonucleotide arrays) and proteomics (two-dimensional gel electrophoresis and mass spectrometry) were applied to analyze early alterations in human cerebral endothelial cells (HCEC) activated by TNFalpha. Human umbilical vein endothelial cells (HUVEC) were used as the reference system. The results presented show that HCEC and HUVEC respond similarly with respect to cell adhesion molecules, chemotaxis, apoptosis and oxidative stress molecules. However, nuclear factors NFkB1 and NFkB2, plasminogen activator inhibitor 1 and cofilin 1 are examples of cerebral specific responses. Our results indicate involvements of the urokinase plasminogen activator system and cytoskeletal rearrangements unique to TNFalpha activation of cerebral endothelial cells.

Molecular classification of borderline ovarian tumors using hierarchical cluster analysis of protein expression profiles

Ovarian tumors range from benign to aggressive malignant tumors, including an intermediate class referred to as borderline carcinoma. The prognosis of the disease is strongly dependent on tumor classification, where patients with borderline tumors have much better prognosis than patients with carcinomas. We here describe the use of hierarchical clustering analysis of quantitative protein expression data for classification of this type of tumor. An accurate classification was not achieved using an unselected set of 1,584 protein spots for clustering analysis. Different approaches were used to select spots that were differentially expressed between tumors of different malignant potential and to use these sets of spots for classification. When sets of proteins were selected that differentiated benign and malignant tumors, borderline tumors clustered in the benign group. This is consistent with the biologic properties of these tumors. Our results indicate that hierarchical clustering analysis is a useful approach for analysis of protein profiles and show that this approach can be used for differential diagnosis of ovarian carcinomas and borderline tumors.

Human tissue distribution of TA02, which is homologous with a new type of aspartic proteinase, napsin A

The N-terminal amino acid sequence of TA02 (molecular weight 35.0 kDa, isoelectric point 5.29), which is associated with primary lung adenocarcinoma, was determined and a fragment peptide was used to generate mouse monoclonal antibodies (mAbs) against TA02. The amino acid sequence suggested that TA02 might be homologous with napsin A, a new type of aspartic proteinase. In this context, we confirmed the expression of napsin A in primary lung adenocarcinoma using reverse-transcription polymerare chain reaction (RT-PCR) and showed that the TA02 mAbs reacted with glutathione-S-transferase (GST)-napsin A fusion protein. We concluded that TA02 is the same molecule as napsin A, and showed immunohistochemically that it is distributed mainly in type II pneumocytes, alveolar macrophages, renal tubules and exocrine glands and ducts in the pancreas. In particular, type II pneumocytes and alveolar macrophages showed high expression of TA02 among human normal tissues. In primary lung adenocarcinoma, 47 out of 58 (81.0%) primary lesions were positive. All well-differentiated adenocarcinomas except those of goblet cell type showed high expression of TA02. In addition, two out of seven (28.6%) large cell carcinomas showed low expression of TA02. The other histopathological types of primary lung cancer did not express TA02 at all. A few cases of renal cell cancer, pancreatic cancer, breast cancer, thyroid cancer, colon cancer and ovarian cancer showed low expression, but the staining patterns were completely different from that of primary lung adenocarcinoma, which showed a granular staining pattern. Our novel mAbs should be valuable for immunochemical detection of TA02/napsin A.

Cancer proteomics: from identification of novel markers to creation of artifical learning models for tumor classification

Studies of global protein expression in human tumors have led to the identification of various polypeptide markers, potentially useful as diagnostic tools. Many changes in gene expression recorded between benign and malignant human tumors are due to post-translational modifications, not detected by analyses of RNA. Proteome analyses have also yielded information about tumor heterogeneity and the degree of relatedness between primary tumors and their metastases. Results from our own studies have shown a similar pattern of changes in protein expression in different epithelial tumors, such as decreases in tropomyosin and cytokeratin expression and increases in proliferating cell nuclear antigen (PCNA) and heat shock protein expression. Such information has been used to create artificial learning models for tumor classification. The artificial learning approach has potential to improve tumor diagnosis and cancer treatment prediction.

Cancer proteomics: from identification of novel markers to creation of artifical learning models for tumor classification

Studies of global protein expression in human tumors have led to the identification of various polypeptide markers, potentially useful as diagnostic tools. Many changes in gene expression recorded between benign and malignant human tumors are due to post-translational modifications, not detected by analyses of RNA. Proteome analyses have also yielded information about tumor heterogeneity and the degree of relatedness between primary tumors and their metastases. Results from our own studies have shown a similar pattern of changes in protein expression in different epithelial tumors, such as decreases in tropomyosin and cytokeratin expression and increases in proliferating cell nuclear antigen (PCNA) and heat shock protein expression. Such information has been used to create artificial learning models for tumor classification. The artificial learning approach has potential to improve tumor diagnosis and cancer treatment prediction.

Classification of human ovarian tumors using multivariate data analysis of polypeptide expression patterns

Large amounts of data on quantitative gene expression are generated by procedures such as 2-DE analysis of proteins or cDNA microarrays. Quantitative molecular variation may potentially be used for the development of methods for the classification of tumors. We used here the statistical concepts of principal components analysis (PCA) and partial least square analysis (PLS) in an attempt to type ovarian tumors. Using a set of 170 polypeptides, 22 tumors were used to establish a model ("learning set") for classification into 3 groups (benign/borderline/malignant). Eighteen tumors were then used to test the model. Six of 8 carcinomas and 3 of 4 borderline tumors were correctly classified. Two of 6 benign lesions were correctly classified, 3 were classified as borderline and 1 as carcinoma. We conclude that it may be possible to classify tumors according to their constitutive protein expression profile using multivariate analysis, thus making classification by artificial intelligence a future possibility.

Identification of gel-separated tumor marker proteins by mass spectrometry

Two-dimensional gel electrophoresis with subsequent analysis by mass spectrometry was applied to study differences in protein expression between benign and malignant solid tumors from human beast, lung and ovary cells. Cells from freshly resected clinical material were lysed and the extracts were subjected to isoelectric focusing with immobilized pH gradients followed by second-dimensional separation on 10-13% sodium dodecyl sulfate (SDS)/polyacrylamide gels. Polypeptides were identified using matrix-assisted laser desorption/ionization and electrospray ionization mass spectrometry after in-gel protein digestion. Some of the upregulated polypeptides in malignant cells are of potential importance as markers of tumor proliferation. Twenty such proteins were identified, ten constituting novel identifications and ten sequence verifications of previously gel-matched proteins. The proteins identified span a wide range of functions, but several cases of protein truncation were found. Truncated forms of cytokeratins 6D and 8, and of cathepsin D were identified. Truncated froms of these over-expressed proteins support the presence of proteolytic processing steps in tumor material. The protein processing and the difference between protein and mRNA abundancies in tumors of different malignancy and origin suggest that studies at the protein level are important for an understanding of tumor phenotypes.

Cancer proteomics: from identification of novel markers to creation of artifical learning models for tumor classification

Studies of global protein expression in human tumors have led to the identification of various polypeptide markers, potentially useful as diagnostic tools. Many changes in gene expression recorded between benign and malignant human tumors are due to post-translational modifications, not detected by analyses of RNA. Proteome analyses have also yielded information about tumor heterogeneity and the degree of relatedness between primary tumors and their metastases. Results from our own studies have shown a similar pattern of changes in protein expression in different epithelial tumors, such as decreases in tropomyosin and cytokeratin expression and increases in proliferating cell nuclear antigen (PCNA) and heat shock protein expression. Such information has been used to create artificial learning models for tumor classification. The artificial learning approach has potential to improve tumor diagnosis and cancer treatment prediction.

Identification of gel-separated tumor marker proteins by mass spectrometry

Two-dimensional gel electrophoresis with subsequent analysis by mass spectrometry was applied to study differences in protein expression between benign and malignant solid tumors from human beast, lung and ovary cells. Cells from freshly resected clinical material were lysed and the extracts were subjected to isoelectric focusing with immobilized pH gradients followed by second-dimensional separation on 10-13% sodium dodecyl sulfate (SDS)/polyacrylamide gels. Polypeptides were identified using matrix-assisted laser desorption/ionization and electrospray ionization mass spectrometry after in-gel protein digestion. Some of the upregulated polypeptides in malignant cells are of potential importance as markers of tumor proliferation. Twenty such proteins were identified, ten constituting novel identifications and ten sequence verifications of previously gel-matched proteins. The proteins identified span a wide range of functions, but several cases of protein truncation were found. Truncated forms of cytokeratins 6D and 8, and of cathepsin D were identified. Truncated froms of these over-expressed proteins support the presence of proteolytic processing steps in tumor material. The protein processing and the difference between protein and mRNA abundancies in tumors of different malignancy and origin suggest that studies at the protein level are important for an understanding of tumor phenotypes.

Polypeptide expression in prostate hyperplasia and prostate adenocarcinoma

Cells were collected from prostate hyperplasias (n = 6) and prostate carcinomas (n = 6) and subjected to two-dimensional gel electrophoresis (2-DE). The resulting polypeptide patterns were analysed with the PDQUEST computer software. Malignant tumors showed significant increases in the level of expression of proliferating cell nuclear antigen (PCNA), calreticulin, HSP 90 and pHSP 60, oncoprotein 18(v), elongation factor 2, glutathione-S-transferase pi (GST-pi), superoxide dismutase and triose phosphate isomerase. In addition, decreases in the levels of tropomyosin-1 and 2 and cytokeratin 18 were observed in prostate carcinomas compared to prostate hyperplasias. This pattern of alterations is similar to that observed in other carcinomas in our previous studies. All malignant tumors showed simultaneous alterations in 5 or more of 9 markers studied, whereas only one case of benign hyperplasia showed alterations in 5 markers. The EST-data base for prostate tumors available from NCI (CGAP) was searched for the expression of the mRNAs corresponding to proteins identified in our gels. Large differences in the relative expression of mRNAs and proteins were observed. Our data show alterations in the pattem of polypeptide expression in prostate carcinomas which are similar to those observed in other carcinomas.

Napsin A, a member of the aspartic protease family, is abundantly expressed in normal lung and kidney tissue and is expressed in lung adenocarcinomas

A pair of 35 kDa polypeptides (TAO1/TAO2) are expressed in more than 90% of all primary lung adenocarcinomas but not in other major malignancies. Mass spectrometry of tryptic peptides showed that TAO1/TAO2 is identical to napsin A, a recently described member of the aspartic proteinase family. The site of processing of pronapsin A to the mature form was located. Napsin expression was detected in human lung adenocarcinoma tumors, compatible with the marker nature of TAO1/TAO2 in the diagnosis of primary lung adenocarcinoma. This is important since identification of markers which can distinguish primary lung adenocarcinomas from distant metastases is desirable. Northern blot analysis showed expression of napsin also in normal lung and kidney tissue, and in situ hybridization showed expression in type II alveolar cells of the lung. This protease is concluded to have a specific functional role in the normal alveolar epithelium and is a candidate protease for the proteolytic processing of surfactant precursors.

Two-dimensional gel analysis of protein expression in ovarian tumors shows a low degree of intratumoral heterogeneity

The process of tumor progression leads to the emergence of multiple clones, and to the development of tumor heterogeneity. One approach to the study of the extent of such heterogeneity is to examine the expression of marker proteins in different tumor areas. Two-dimensional gel electrophoresis (2-DE) is a powerful tool for such studies, since the expression of a large number of polypeptide markers can be evaluated. In the present study, tumor cells were prepared from human ovarian tumors and analyzed by 2-DE and PDQUEST. As judged from the analysis of two different areas in each of nine ovarian tumors, the intratumoral variation in protein expression was low. In contrast, large differences were observed when the protein profiles of different tumors were compared. The differences in gene expression between pairs of malignant carcinomas were slightly larger than the differences observed between pairs of benign tumors. We conclude that 2-DE analysis of intratumoral heterogeneity in ovarian cancer tissue indicates a low degree of heterogeneity.

Characterization of periplasmic Escherichia coli protein expression at high cell densities

We have used two-dimensional electrophoresis (2-DE) to analyze changes in protein expression profiles during a microbial cultivation process on an industrial scale. An Escherichia coli strain W31 10 containing the gene for recombinant human growth hormone production was used. Samples were taken at time intervals ranging from fast to slow growth rate (late growth phase at high cell density/starvation) and 2-DE analysis combined with image analysis using the PDQuest software showed significant alterations in expression levels of a number of proteins. Twenty-four protein spots were identified using a combination of matching with SWISS-2DPAGE E. coli map, N-terminal sequence analysis and mass spectrometry matrix-assisted laser desorption/ionization (MALDI). Two of the most abundant proteins expressed at late growth phase (pI 5.4/28 kDa and pI 5.5/28 kDa) were subjected to N-terminal sequence analysis after electrotransfer of the proteins from a preparative 2-DE gel to polyvinylidene difluoride (PVDF) membrane. Sequence tags of five amino acids in combination with approximate pI and Mr identified both proteins as deoxyribose phosphate aldolase (gene name deoC). In addition, both spots were subjected to tryptic in-gel digestion and analyzed using MALDI. Peptide mass fingerprints from both spots showed similar MALDI spectra and 10 of 10 tryptic fragments confirmed the identity as deoC. The identification of the acidic variant of deoC on 2-DE gels and the observation of this variant as induced during late growth phase is novel.

Rapid isolation of phage displayed antibodies to beta-actin eluted from two-dimensional electrophoresis gel

We describe a simple and efficient procedure which can be used to prepare antibodies to proteins extracted by two-dimensional gel electrophoresis (2-DE), using beta-actin as a model. Protein was electroeluted from a stained gel, biotinylated and used for selection of phage from a semisynthetic phage antibody library. After four rounds of selection using 50 ng beta-actin per cycle, approximately 8 X 10(3) phage were recovered. Antibody fragments were prepared from 21 randomly picked clones. Six of eighteen (6/18) antibody-positive clones produced antibody fragments reacting against beta-actin in an enzyme linked immunosorbent assay (ELISA). Sequencing of the HC-CDR3-region showed that all six clones were independent isolates, suggesting that a large number of independent phage antibody reactivities were generated.

Down-regulation of tropomyosin-2 expression in c-Jun-transformed rat fibroblasts involves induction of a MEK1-dependent autocrine loop

Overexpression of the c-Jun transcription factor in rodent fibroblasts may result in cell transformation or in apoptosis. The mechanisms whereby c-Jun induces transformation are unknown. We show here that the expression of high-molecular weight tropomyosin-2 (TM-2) is down-regulated in c-jun-transformed FR3T3 rat fibroblasts. However, down-regulation did not seem to be a direct effect of c-Jun on TM-2 gene expression. Thus, TM down-regulation in c-jun-transformed cells was alleviated by inhibitors of Ras (BZA-5B) or MEK1 (PD98059). Furthermore, medium conditioned by c-jun-transformed cells induced TM-2 down-regulation in untransformed cells by a mechanism requiring MEK1. Consistent with a central role for the MEK/ERK, but not SEK/JNK, pathway for TM down-regulation, constitutively active mutants of Raf induced TM down-regulation, whereas constitutively active Rac did not. We also show that anchorage-independent growth of c-jun-transformed cells requires MEK1. These findings suggest that indirect induction of the MEK/ERK pathway is central to c-Jun-induced transformation of rat fibroblasts.

Phenotypic analysis of ovarian carcinoma: polypeptide expression in benign, borderline and malignant tumors

Studies of multiple markers in tumors are required for adequate biological characterization. We have characterized the expression of multiple proteins in human ovarian tumors using the technique of 2-dimensional gel electrophoresis (2-DE/PDQUEST). Tumor cells were prepared from the tissue of 22 ovarian tumors. Large variations were observed between tumors in the expression of various polypeptides, indicating heterogeneity in gene expression. An increase in the spot density of 2 cell-cycle-related proteins, PCNA and OP18/stathmin, was observed in carcinomas. Borderline tumors expressed low levels of these proteins. Significant increases in the levels of nm23, GST-pi, elongation factor 2 and triose phosphate isomerase were recorded in ovarian carcinomas. Furthermore, decreases in the levels of tropomyosin-2 and lamin C were observed in malignant as compared with benign tumors. The pattern of expression of 9 protein markers was examined in individual tumors. All malignant tumors showed simultaneous alterations in the expression of 5 or more of these proteins, whereas no benign tumor showed alterations in the expression of more than 3 polypeptides. Borderline tumors showed alterations in 0 to 6 markers. We conclude that the simultaneous analysis of multiple polypeptides, which can be achieved by 2-DE, is useful for characterization of gene expression and diagnostic studies in ovarian tumors.

Increased expression of alpha-enolase in c-jun transformed rat fibroblasts without increased activation of plasminogen

Two-dimensional gel electrophoresis was used to identify polypeptides differentially expressed between normal and c-jun transformed rat fibroblasts. The level of a 49 kDa polypeptide was 3-fold elevated in c-jun transformed cells. Sequence analysis by ion trap mass spectrometry identified the polypeptide as rat alpha-enolase. Enolase functions as a cell surface receptor for plasminogen, suggesting that upregulation may increase plasminogen activation and cell surface proteolysis important for tumor growth. However, no difference was observed between normal and transformed cells in formation of plasmin, suggesting that upregulation of alpha-enolase may contribute to an increased metabolic capacity, but not to increased plasminogen activation.

Biophysical and physiological properties of a modified porcine surfactant enriched with surfactant protein A

Surfactant protein A (SP-A), a major protein component of natural pulmonary surfactant, is absent in exogenous surfactants currently used in clinical practice. We investigated the physical and physiological properties of one of these modified natural surfactants (Curosurf) after enrichment with 5% SP-A (SP-A-Curosurf). A pulsating bubble system was used for in vitro assessments and ventilated newborn rabbits for evaluation of in vivo effects. In the presence of various potential inhibitors (meconium 5 mg.mL-1, fibrinogen 5 mg.mL-1, albumin 25 mg.mL-1, or whole serum proteins 25 mg.mL-1), Curosurf at a concentration of 5 mg.mL-1 was inactivated while SP-A-Curosurf and natural porcine surfactant at the same concentration had normal maximum and minimum surface tension. This protective effect of SP-A was calcium dependent. In immature newborn rabbits, the improvement of lung-thorax compliance observed after treatment with 100 mg.kg-1 of SP-A-Curosurf was equivalent to that obtained with 200 mg.kg-1 of Curosurf. Similarly, in near-term newborn rabbits with respiratory failure induced by instillation of fibrinogen via the airways, the increase in compliance after administration of 100 mg.kg-1 of SP-A-Curosurf corresponded to that seen after treatment with 200 mg.kg-1 of Curosurf, whereas Curosurf at a dose of 100 mg.kg-1 had no substantial effect. Our data thus indicate that surfactant protein A increases the resistance of Curosurf to inactivation under in vivo conditions.

Analysis of polypeptide expression in benign and malignant human breast lesions

Results of two-dimensional electrophoresis (2-DE) analyses of human breast carcinoma are described. Tumor cells were extracted and purified from breast carcinomas with different proliferative indeces and degrees of genomic stability. Cells purified from fibroadenoma tissue served as controls for benign cells. The following results were observed: (i) Analysis of samples from different areas of the same tumor showed a high degree of similarity in the pattern of polypeptide expression. Similarly, analysis of two tumors and their metastases revealed similar 2-DE profiles. (ii) In contrast, large variations were observed between different lesions with comparable histological characteristics. Larger differences in polypeptide expression were observed between potentially highly malignant carcinomas compared to comparisons of less malignant lesions. These differences were in the same order of magnitude as those observed comparing a breast carcinoma to a lung carcinoma. (iii) The levels of all cytokeratin forms resolved (CK7, CK8, CK15, and CK18) were significantly lower in carcinomas compared to fibroadenomas. (iv) The levels of high molecular weight tropomyosins (1-3) were lower in carcinomas compared to fibroadenomas. The expression of tropomyosin-1 was found to be 1.7-fold higher in primary tumors with metastatic spread to axillar lymph nodes compared to primary tumors with no evidence of metastasis (p < 0.05). (v) The expression of proliferating cell nuclear antigen (PCNA) and some members of the stress protein family (pHSP60, HSP90, and calreticulin) were higher in carcinomas. We conclude that malignant progression of breast carcinomas results in large heterogeneity in polypeptide expression between different tumors, but that some common themes such as decreased expression of cytokeratin and tropomyosin polypeptides can be discerned.