DGAP: Proteomics Core

Proteomics Core

Core Director: Silvia Corvera, M.D. (U. Mass. Medical School)
Co-Director: John Leszyk, Ph.D. (U. Mass. Medical School)

Specific Aims

In addition to changes in the levels of specific gene transcripts, cellular processes are regulated by mechanisms that occur subsequent to transcription. Post-transcriptional control mechanisms can determine both the levels and functional properties of specific gene products by controlling protein half-lives, activity and/or subcellular localizations. Thus, a direct analysis of proteins is necessary to obtain a complete and accurate picture of the molecular architecture of cells and tissues relevant to type 2 diabetes and of their alterations in disease.

The Core will provide DGAP investigators with services required to define the identity, abundance, subcellular localization and oligomerization state of adipose cell proteins.

The Core will work with the DGAP Bioinformatics core to provide information on the identity, abundance, subcellular localization and oligomerization state of adipose cell proteins to complement annotation of Genomics databases generated by DGAP investigators.

Background and Preliminary Data

Mass Spectrometry Fingerprinting and Database Correlation Analysis. As the sequencing of the genomes of many model organisms comes to completion the likelihood that a protein from the organism's proteome can be found in the protein database has increased exponentially. This has precipitated a revolution in the application of mass spectrometric techniques in the identification of proteins in the databases (1-3). Highly accurate measurements of peptide masses, as well as isolation and fragmentational analysis of individual peptides can be achieved with both MALDI and ESI mass spectrometers. Database searches can be based on the collective analysis of a number of peptides (MS) or on a single peptide and fragments derived from it by a collisional or decay process (MS/MS). The success of these identification methods is based on a correlation analysis with pre-existing sequences found in the databases and not on any de-novo interpretation of MS/MS spectral data. The UMass Proteomics Core laboratory is equipped with state-of the-art MALDI-TOF instrumentation (Axima CFR from Kratos Analytical). This instrument is used to identify proteins by both MS data from peptide digests as well as MS/MS data derived from post-source-decay (PSD) analysis of individual peptides using the Protein Prospector program developed by Peter Baker and Karl Clauser at UCSF. The Core also has an ESI quadrapole ion trap mass spectrometer (LCQ Deca) from Finnigan, and a capillary HPLC system for direct coupling to perform LC/MS. This instrument is used to automatically acquire high quality MS/MS spectra in a data dependent fashion, followed by database searching using the SEQUEST software developed by John Yates and Jimmy Eng at the University of Washington. By directly coupling to an HPLC for initial peptide fractionation one can analyze complex protein mixtures since only MS/MS data is used to search the database. This instrument system is particularly useful for the deconvolution of protein mixtures.

Methodology for Separation of Proteins before Mass Spectrometry

The use of mass spectrometry fingerprinting of proteins requires their previous separation into unique proteins or simple protein mixtures. The single, classical approach consists in the separation of protein mixtures on 2D gels. The first dimension relies on differences in isoelectric point of proteins, while the second dimension is typically SDS-PAGE, which separates proteins by relative mass. Well known disadvantages of this procedure include relatively low sensitivity due to the capacity limitation of isoelectric focussing gels, the need for sophisticated imaging software to accurately compare two or more gels, and the impossibility of separating hydrophobic proteins (4). We have tested whether comprehensive proteome analysis can be achieved through the separation of complex protein mixtures in the first dimension on the basis of alternative physical/biological properties. The sedimentation coefficient of a protein varies with its size and shape, as well as with biological parameters that pertain to individual proteins, such as their monomeric or oligomeric state. We have found that both cytosolic and membrane-associated proteins distribute into well separated bands along both the horizontal and vertical axes of SDS-polyacrylamide gels after velocity gradient centrifugation (VGC) (Figures A1 and A2, UMass Proteomics Core Appendix). Comparison of corresponding fractions derived from independent gradients on adjacent lanes on the SDS-PAGE gel reveals the reproducibility of each gradient to be high. Furthermore, the amount of protein that can be analyzed (milligrams) is orders of magnitude higher than that amenable to separation by isolectric focusing, allowing for the exact identification of each band by mass spectrometry.

Two crucial parameters in assessing quantitative changes in protein levels are the sensitivity and linearity of the detection method. We find these parameters to be optimized with a new fluorescent protein stain, SyproRubyTM (Figure A3, UMass Proteomics Core Appendix). The combined use of VGC/SDS-PAGE and SyproRuby staining has proved adequate for the identification of changes in protein composition in cells subjected to diverse perturbations. For example, extracts from 3T3-L1 cells treated for 24 hr with Rosiglitazone (BRL49653), a known insulin sensitizer (5), display multiple changes in the levels of cytosolic bands, many of which correspond to enzymes involved in lipid metabolism.. In addition, significant increases are seen in proteins from the particulate fraction, many of which were identified as mitochondrial components.

Research Design and Methods

Aim 1.
The separation methodology described above, in combination with high throughput mass-spectroscopy will be used to generate a proteomic fingerprint of: 1) the adipocyte plasma membrane, 2) intracellular GLUT4-containing membranes, 3) mitochondria, and 4) cytosolic fractions from both 3T3-L1 adipocytes and freshly isolated adipocytes from rodent models provided by the DGAP. Experiments that have included and will continue to use these services are described in detail in Specific Aim 1 of Project 4, Specific Aims 2 and 3 of Project 3. Similar approaches will be used for the analysis and identification of proteins that are altered in response to insulin, or by insulin-resistant states, in adipose cells derived from rodent models by other investigators in the DGAP. Detailed conventional methods for the isolation of mouse adipocytes, and for the purification of subcellular fractions have been published (4-8).

Aim 2.

Related Projects