The term 'proteome' was first coined in 1994, and refers to all the proteins in a cell, tissue, or organism. Proteomics is the systematic study of proteins in a particular cell, tissue, or organism. Because proteins are involved in almost all biological activities, proteomics is a rich source of biological information.

Protein scientists have diverse interests. These include determining the function and amino acid sequence of proteins; their three-dimensional structure; how the addition of sugar, phosphate, or fat affects protein function; and how proteins interact with other molecules. Some researchers are focused on the proteins present in particular parts of the cell such as the outer cell membrane, the nucleus, the cytosol (the region of the cell outside the nucleus), or the nuclear membrane; others are analyzing protein-protein interactions in a particular cell or organism; some are studying the differences between the proteins present in diseased vs. healthy cells (1).

The total number of proteins in human cells is estimated to be between 1 million to 10 million, and only a small percentage has been sequenced or identified. The complete proteome has not been characterized for any organism. In contrast, the genome or the entire set of genes for several organisms has been sequenced, including humans. The human genome is estimated to contain about 30,000 genes (http://www.genome.gov/10002192).

Besides the difference in quantity, another important difference between the genome and proteome is that the genome is static and relatively unchanged from day to day. Many cellular proteins, on the other hand, are continually moving and undergoing changes such as binding to a cell membrane, partnering with another protein, gaining or losing a chemical group such as a sugar, fat, or phosphate, or breaking into two or more pieces. Proteins play a central role in the complex communication network within and between cells that is constantly responding to the needs of the organism.

Several other properties of proteins add to their complexity:

•  Proteins vary between individuals, between cell types, and in the same cell under different stimuli or different disease-states.

The goal of clinical proteomics is to apply proteomics technology to patient care. This new research technology is now being used for clinical studies ranging from cancer to cardiovascular disease and organ transplant (3). Researchers are searching for proteins in blood, urine, or diseased tissue that can be used as early biomarkers of disease, or that may predict response to therapy or the likelihood of relapse after treatment. Several approaches are being tested to identify biomarkers for various cancers (4, 5).

Ovarian Cancer

Ovarian cancer is a major target of early biomarker research because it is usually diagnosed at an advanced stage with a five-year survival rate of about 20 percent. To test proteomics as a diagnostic tool, a group of researchers from the National Cancer Institute (NCI) in Bethesda, Md., collected serum from 50 ovarian cancer patients and 50 controls and used a computer algorithm to search for the protein patterns that distinguished cancer from non-cancer. When they tested this pattern with a set of blinded serum samples, the test pattern correctly detected all 50 patients with cancer, and was able to discriminate them from 63 out of 66 patients who were unaffected or had benign disease (6). Using the same approach, two other groups reported similar results (5,6).

Prostate cancer

A similar proteomic analysis of prostate cancer patients vs. healthy controls was carried out by looking for differences in protein patterns between the two groups. Using blood samples from 167 prostate cancer patients, 77 patients with benign prostate hyperplasia and 82 healthy men, the computer was able to develop a classification system that correctly classified 96 percent of the samples as either prostate cancer or non-cancer (benign prostate hyperplasia/healthy men) (9). Another proteomic approach is to determine whether the changes in specific phosphoproteins (proteins with phosphate groups attached) believed to be important in cellular signaling events and cancer progression in prostate cancer patients can serve as a biomarker of early disease (10).

Breast Cancer

A combination of three candidate proteins in the blood were found to be useful in discriminating between 169 patients at various stages of breast cancer compared to women with benign breast disease and healthy controls (11). In three other studies, nipple aspirate fluid was used to identify tumor marker candidates (12-14). Nipple aspirate fluid has a higher concentration of breast specific proteins than blood. Mammary ducts are thin tubes that lead to the nipples and are where 70 percent to 80 percent of breast cancers originate.

Lung and Bladder

Several laboratories have successfully analyzed tumor tissue from patients with lung and bladder cancer and discovered protein patterns that could discriminate diseased from healthy tissue (15). Likewise, preliminary results using a proteomic approach to detect bladder cancer have been promising (16).

Future Use

At this point, none of the proteomics analyses is mature enough to be used in the clinic as a screening tool, but these small studies point to the promise of proteomics as a diagnostic marker. Validation in large groups of patients is necessary before proteomics patterns can be used routinely in the clinic as biomarkers for early disease.

The NCI-Food and Drug Administration (FDA) Proteomics Program was launched in 1997 under the leadership of Lance Liotta, M.D., Ph.D., of NCI's Center for Cancer Research, and Emanuel Petricoin, Ph.D., of FDA's Center for Biologics Evaluation and Research (CBER).

The general strategy of the proteomics program is to extract proteins from blood or tissue, analyze them with mass spectrometry to create patterns of protein fragments, sort through the patterns with an artificial intelligence computer program, and ultimately discover differences that will help distinguish, for example, cancer patients vs. healthy controls, or patients who will respond to therapy vs. those or who do not respond.

A high priority of the program is to apply research results directly to patient care. The potential benefits to patients include:
•  Diagnosing cancer earlier than is now possible;
•  Improving the understanding of tumors at the protein level, leading to better treatments.
•  Developing individualized therapies for each patient; and
•  Determining the toxic and beneficial effects of treatments before administering them to patients.

Blood Test for Ovarian Cancer

Liotta, Petricoin, and their colleagues have invented or refined several key technologies used in proteomic analysis, and in the process have identified hundreds of proteins in breast, ovary, prostate, and esophagus tissue that change in amount as the cells in these tissues grow abnormally. In 2002, they discovered patterns of proteins found in the blood of ovarian cancer patients that may be useful as an early biomarker of disease (6). Using the patient's blood and a test that can be completed in 30 minutes, researchers were able to differentiate between serum samples taken from patients with ovarian cancer and those from unaffected individuals.

Artificial Intelligence

The diagnostic test relied on a sophisticated artificial intelligence computer program developed by Correlogic Systems, Inc., Bethesda, Md. Scientists were able to "train" the computer to identify a pattern of only a handful of small proteins from thousands of candidates found in the blood that could distinguish cancer patients vs. control samples. Once these patterns were found, they were tested on other blinded patient samples with and without cancer. Fifty out of 50 cancers and 63 of 66 non-cancer samples were correctly identified. These results suggested that proteomic technology may help clinicians diagnose the disease much earlier than current methods.

Improvements in Blood Test

The NCI/FDA researchers are continuing to improve the performance of the proteomic analysis for ovarian cancer. The 2002 Lancet paper (6) reported that the test performed with 100 percent sensitivity and 95 percent specificity. Sensitivity measures the proportion of people with the disease who test positive; specificity measures the proportion of the people without the disease who test negative. A specificity of 95 percent means that 5 percent of those who did not have cancer would test positive, which is far too high a false positive rate for commercial use.

In a recently accepted paper, scientists from the NCI/FDA team, Science Application International Corporation (SIAC)-Frederick, and Correlogic Systems, Inc., report that both the sensitivity and specificity for blinded samples were 100 percent. In this study, which involved a larger group of ovarian cancer patients and controls, the scientists tested archived blood samples with a higher resolution instrument and a different protein pattern compared to the 2002 paper (17, 18).

Despite this promise, validation in a very large clinical group is needed before a commercial test for this technique can become available.

Discovering New Proteins

Although it is not necessary in theory to know the identity of the proteins that may detect early disease or response to treatment, many of these proteins have now been identified and are leading to an understanding of the molecular pathways involved in disease. These efforts are also leading to the discovery of many new proteins in the blood. For more information, please go to: http://bpp.nci.nih.gov

Other Cancers

In addition to ovarian cancer, similar techniques are being applied to other cancers. The NCI-FDA researchers are looking for protein patterns in the blood that are diagnostic for early stage prostate and breast cancers, as well as patterns that can predict risk for prostate, melanoma, and pancreatic cancers (19,20).

Proteins in Tissues

In addition to analyzing proteins in the blood, another thrust of the proteomics program is to compare the proteins in tumor tissue vs. healthy tissue. Using this approach, researchers are probing tissues for phosphorylated proteins known to be important in carcinogenesis and are looking for useful diagnostic patterns. The work is yielding new insights about molecular pathways that are altered in tumor development (21).

Role of Albumin

The NCI-FDA team has also discovered that the low-molecular proteins useful for early detection of ovarian cancer accumulate in the blood on larger carrier proteins such as albumin. This piggy-backing ensures the smaller proteins a longer life in the circulating blood (22, 23, 24). Knowing this, the scientists can obtain a greater concentration of potential biomarker proteins by extracting the carrier protein fraction from the blood. Some groups are working to create a synthetic carrier protein that could be used to standardize diagnostic protein patterns.

Refining the Technology

NCI/FDA experts are continuing to test alternative mass spectrometry platforms and computer algorithms that they hope will yield clinically useful patterns (25).

Yes. In protocols supported by the Clinical Center at the National Institutes of Health, Elise Kohn, M.D., from the NCI's Laboratory of Pathology, is the principal investigator of a series of trials using proteomic profiling.

General Strategy

Biopsied cells from cancer patients before, during, and after treatment are extracted using a special Laser Capture Microdissection Microscope invented in Liotta's laboratory. These tools allow the investigators to isolate pure normal cells, pre-cancerous cells, and tumor cells from the same patient. By analyzing the protein patterns in these cells, the researchers will examine:
•  How a particular treatment changes the network or circuitry of the proteins in a cell;
•  Whether particular patterns predict early stage disease;
•  The mechanism of drug resistance;
•  How to reduce treatment side-effects; and
•  Whether protein patterns change when the tumor returns after treatment.

Ongoing Trials

Specifically, the NCI/FDA scientists are conducting a clinical trial to determine if the ovarian cancer proteomics tool is comparable to the single protein test, CA 125, presently used to detect ovarian cancer recurrence. Kohn is currently recruiting ovarian cancer patients who are in their first remission after treatment. Interested patients can call the Clinical Studies Support Center at 1-888-624-1937.

The NCI/FDA scientists also plan to extend the technology to confirm the sensitivity and specificity of proteomic analysis for stage I ovarian cancer in trials of high-risk and low-risk women. The proteomic tool will be tested alone or in combination with current screening options (e.g., CA125).

There are several ongoing proteomics trials currently recruiting patients to determine whether particular patterns of blood proteins can:
•  Predict the development of non-small cell lung cancer in patients with suspicious lung abnormalities;
•  Allow a doctor to determine if a patient has mycosis fungoides/cutaneous T-cell lymphoma;
•  Indicate the response to radiation therapy for localized prostate cancer and identify which patients may or may not benefit from aggressive treatment;
•  Provide information about which patients with psoriasis or cutaneous T-cell lymphoma will stay in remission and which will not;
•  Identify pancreatic lesions in patients who have von Hippel-Lindau syndrome.

For more details, visit:
(http://clinicaltrials.gov/ct/gui/search?term=proteomics&submit;=Search)

Traditionally, proteomics experiments have been done using two-dimensional gel electrophoresis (2DE), a process by which large mixtures of proteins are separated by electrical charge and size. In the first stage, the proteins migrate through a gel-like substance until they are separated by their charge; for the second stage, they are transferred to a second semi-solid gel and are separated by size. The advantage of this method is that a large number (3,000 to 10,000) proteins can be visually separated. The drawback is that certain kinds of proteins such as membrane proteins, proteins present in very small amounts, or very large or very small proteins are difficult or impossible to visualize by 2DE.

In the last ten years or so, mass spectrometry (MS) has increasingly become the method of choice for analyses of complex protein samples. Mass spectrometry is a technique that measures two properties: the mass-to-charge ratio (m/z) of a mixture of ions in the gas phase under vacuum; and the number of ions present at each m/z value. (Ions are particles with an electric charge.) The end product is a mass spectra or chart with a series of spiked peaks, each representing the ion or charged protein fragment present in a given sample. The height of the peak is related to the abundance of the protein fragment. The size of the peaks and the distance between them are a fingerprint of the sample and provide a clue to its identity (26).

The mass spectrometer consists of an ionization source, a mass analyzer, and detector:

   •  The ionization source ionizes the proteins or protein fragments present in the sample. Ionizing means removing the electrons from protein fragments resulting in positively charged particles.

Ionization Sources

Two ionization techniques, MALDI and ESI, had a major impact on protein biochemistry because they are able to produce ions in the gas phase without too much fragmenting of the proteins, a problem with older methods. MALDI (Matrix-assisted laser desorption ionization) produces ions by sublimating (going from a solid to a gas) and ionizing the proteins out of a dry, crystalline stage. ESI (electrospray ionization) ionizes the protein mixtures out of a solution. MALDI is normally used to analyze relatively simple peptide mixtures while ESI is preferred for more complex samples. However, a variant of MALDI where the surface of the MALDI target has been modified is used with more complex mixtures. Known as surface enhanced laser desorption ionization (SELDI) MS, this technique is widely used in cancer proteomics. Only a small fraction of protein fragments in the sample bind to the SELDI surface because they have an affinity for the substances on the surface (26).

Mass Analyzer and Detector

Once the ions are produced, the mass analyzer/detector separates them by the mass-to-charge ratio and produces mass spectra, or series of spiked peaks, which are used to identify the proteins. The mass of the protein peaks increases from left to right; the height of each peak is proportional to the number of ions at that particular mass-to-charge ratio. Four types of mass analyzers are commonly used: ion trap, time of flight (TOF), quadrupole, and Fourier transform ion cyclotron (FT-MS)(26).

Ionizers, Analyzers, and Detectors

The ionization method, MALDI is commonly coupled to TOF mass analyzers, while ESI is most often coupled to ion trap or quadrupole spectrometers. Most serum protein mass spectrum data have been generated by using the Ciphergen Biosystems (Fremont, Calif) ProteinChip array surface-enhanced laser desorption ionization-time-of-flight (SELDI-TOF) MS system. In this system, specific substances are applied to the surface of the SELDI chip array to capture peptides in the sample. Once captured, the proteins are detected by TOF MS.

Examples of commercially available statistical tools used to analyze mass spectra are: PROTEOME QUEST (Correlogic Systems, Bethesda, Md.); PROPEAK (3Z Informatics, Mount Pleasant, S.C.); BAMF (Eclipse Diagnostics, Vacaville, Calif); and Biomarker Wizard (Ciphergen Biosystems, Freemont, Calif).

The great advantage of MS over other technologies for detecting and monitoring subtle changes in the body is the ability to measure rapidly and inexpensively thousands of elements in a few drops of blood. Unlike 2DE, MS patterns generated from the thousands of proteins present in blood are difficult to analyze visually. However, the powerful computational ability of today's computers makes it possible to analyze MS spectra rapidly and distinguish subtle differences in patterns between diseased and healthy people.

Mass spectrometry-based proteomics analysis is extremely rapid. The entire process from collecting blood to analyzing the MS spectrum can occur in less than one minute. In addition, hundreds of samples can be analyzed sequentially, and extremely small amounts of protein can be detected.

Proteomics data are being collected at a faster pace than the ability of the researchers to validate, interpret, and integrate them with other known data. There is a great need to make data portable and comparable. Software tools are needed in all areas of data analysis, including data collection, storage, searching, analysis, classification, management, archiving, and data retrieval.

Standardization of data analysis is a very important concept in proteomics and is necessary for scientists to compare data. However, none of the technologies has reached the status of a validated proteomic tool. There is a need to develop a statistical system that assigns a score to each data point to estimate the probability that the observation is correct. A reliable technology to quantify proteins is also needed.

In the future, scientists expect that by combining both the genomic and proteomic data, they will be able to create a mathematical model of the molecular pathways in cells. With these models, researchers will be able to predict previously unknown interactions and verify the predictions experimentally. Novel proteins, cellular functions, and pathways will also be discovered. It is hoped that understanding the connections between cellular pathways will greatly reduce the suffering and loss of life due to cancer.

References

1. Patterson SD & Aebersold RH. Proteomics: the first decade and beyond. Nature Genetics Supplement 2003;33:311-32.

2. Ullrich B, Ushkaryov YA, and Sudhof TC. Cartography of neurexins: more than 1,000 isoforms generated by alternative splicing and expressed in distinct subsets of neurons. Neuron 1995:14:497-507.

3. Petricoin EF, Rajapaske V, Herman EH, Arekani AM, Ross S, Johann D, Knapton,A, Zhang J, Hitt BA, Conrads TP, Veenstra TD, Liotta LA, and Sistare FD. Toxicoproteomics: Serum Proteomic Pattern Diagnostics for Early Detection of Drug Induced Cardiac Toxicities and Cardioprotection. Journal of Toxicologic Pathology 2004;32 (S1):1-9.

4. Clarke W, Zhang Zhen, Chan DW. The application of clinical proteomics to cancer and other diseases. Clin Chem Lab Med 2003;41(12):1562-1570.

5. Liotta LA, Espina V, Mehta AI, Calvert V, Rosenblatt K, Geho D, Munson PJ, Young L, Wulfkuhle J, Petricoin EF. Protein microarrays: Meeting analytical challenges for clinical applications. Cancer Cell 2003; Apr;3(4):317-25.

6. Petricoin EF, Ardekani AM, Hitt BA, Levine PF, Fusara VA, Steinberg SM, et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet 2002;369:572-7.

7. Sorace JM, Zhan M. A data review and re-assessment of ovarian cancer serum proteomic profiling. BMC Bioinformatics 2003;4:24.

8. Zhu W, Wang X, Ma Y, Rao M, Glimm J, Kovach JS. Detection of cancer-specific markers amid massive mass spectral data. Proceedings of the National Academy of Sciences 2003;100:14666-14671.

9. Adam B-L, Qu Y, Davis JW, Ward MD, Clements MA, Cazares LH, et al. Serum Protein Fingerprinting couple with a pattern-matching algorithm distinguishes prostate cancer from benign prostate hyperplasia and healthy men. Cancer Research 2003; 62:3609-3614.

10. Grubb RL, Calvert VS, Wulkuhle JD, Paweletz CP, Linehan WM, Phillips JL, et al. Signal pathway profiling of prostate cancer using reverse phase protein array. Proteomics 2003;3:2142-2146.

11. Li J, Zhang Z, Rosenzweig J, Wang YY, Chan DW. Proteomics and bioinformatics approaches for identification of serum biomarkers to detect breast cancer. Clinical Chemistry 2002;48:1296-1304.

12. Sauter ER, Zhu W, Fan XJ, Wassell RP, Chervoneva I, Du Bois GC. Proteomic analysis of nipple aspirate fluid to detect biologic markers of breast cancer. Br J Cancer 2002; 86:1440-3.

13. Pawaletz CP, Trock B, Pennanen M, Tsangaris T, Magnant C, Liotta LA, et al. Proteomic patterns of nipple aspirate fluids obtained by SELDI-TOF: potential for new biomarkers to aid in the diagnosis of breast cancer. Disease Markers 2001;17:301-7

14. Varnum SM, Covington CC, Woodbury RL, Petritis K, Kangas LJ, Abdullah MS, et al. Proteomic characterization of nipple aspirate fluid: identification of potential biomarkers of breast cancer. Breast Cancer Research and Treatment 2003;80:87-97.

15. Celis JE and Gromov. Proteomics in translational cancer research: Toward an integrated approach. Cancer Cell 2003 Jan;3:9-15.

16. Vlahou A, Schellhammer PF, Mendrinos S et al. Development of a novel proteomic approach for the detection of transitional cell carcinoma of the bladder in urine. American Journal of Pathology 2001;158:1491-1502.

17. Conrads TP, Fusaro VA, Ross S, Johann D, Rajapakse Vinodh, Hitt BA, et al. High- resolution serum proteomic features for ovarian cancer detection. Accepted for publication in Endocrine- related cancer, June 2004.

18. Alexe G, Alexe S, Liotta LA, Petricoin E, Reiss M, Hammer PL. Ovarian cancer detection by logical analysis of proteomic data. Proteomics 2004;4:766.

19. Petricoin EF 3rd, Liotta LA. Serum Proteomic Patterns for Detection of Prostate Cancer 2003. Journal of the National Cancer Institute 2003;95(6):490-1.

20. Hingorani SR, Petricoin EF, Maitra A, Rajapakse V, King C, Jacobetz MA, Ross S, Conrads TP, Veenstra TD, Hitt BA, Kawaguchi Y, Johann D, Liotta LA, Crawford HC, Putt ME, Jacks T, Wright CV, Hruban RH, Lowy AM, Tuveson DA. Preinvasive and invasive ductal pancreatic cancer and its early detection in the mouse. Cancer Cell 2004;5(1):103.

21. Wulfkuhle JD, Aquino JA, Calvert VS, Fishman DA, Coukos G, Liotta LA, and Petricoin EF. Signal pathway profiling of ovarian cancer from human tissue specimens using reverse- phase microarrays. Proteomics 2003 Nov;3(11):2085-90.

22. Liotta LA, Ferrari M, Petricoin EP. Written in Blood. Nature Oct 2003;425:905.

23. Tirumalai RS, Chan KC, Prieto DA, Issaq HJ, Conrads TP, Veenstra TD. Characterization of the low molecular weight human serum proteome. Molecular and Cellular Proteomics 2003;2(10):1096-103.

24. Mehta AI, Ross S, Lowenthal MS, Fusaro V, Fishman DA, Petricoin EF, Liotta LA. Biomarker amplification by serum carrier protein binding. Disease Markers 2003-2004; 19:1-10.

25. Petricoin EF, Fishman,DA, Conrads TP, Veenstra TD, and Liotta, LA. Proteomic Pattern Diagnostics: Producers and Consumers in the Era of Correlative Science. BMC Bioinformatics, Posted March 12, 2004 (http://www.biomedcentral.com/1471-2105/4/24/comments ).

26. Aebersold R and Mann M. Mass spectrometry-based proteomics. Nature 2003;422:198-207.
    # # #

    Related Resources

    Because the proteome is constantly changing, standardizing the conditions of proteomic analyses is a very important, and is necessary for comparisons between investigators. The Human Proteome Organization (HUPO: www.HUPO.org), along with the Plasma Proteome Project (http://www.plasmaproteome.org/ ), have been formed to address this issue, as well to promote new research.

    NCI Resources

    NCI-FDA Clinical Proteomics Program Web site:
    http://www.ncifdaproteomics.com.

    To visit NCI's Web site:
    http://www.cancer.gov

    For information about research currently supported by NCI:
    http://researchportfolio.cancer.gov/

    For information about clinical trials:
    http://www.clinicaltrials.gov

    For general information:
    Cancer Information Service: 1-800-4-CANCER (1-800-422-6237)
    TTY (for deaf and hard of hearing callers): 1-800-332-8615
    LiveHelp: Cancer Information Specialists offer online assistance through the LiveHelp link on the NCI's Web site.

    Back to Top