Send to printer »

Feature Articles : Oct 1, 2008 (Vol. 28, No. 17)

Protein Quantification Pinpoints Targets

Advanced Technologies Bring Field to a New Level of Sophistication
  • K. John Morrow Jr., Ph.D.

When biochemists talk about protein quantification, they often refer to the use of colorimetric assays such as the Bradford and its rather mundane application for the calculation of specific enzyme activities.

The past decade, however, has witnessed a rapid expansion in both the methods and applications of quantitatively assessing protein levels. These include mass spectrometry (MS) of proteolytic fragments combined with liquid chromatography, isotopic labeling of peptides and isotope dilution mass spectrometry (IDMS). The availability of many extremely accurate measuring tools has enabled investigators to map out the relationship between disease processes, and protein abundance and distribution, as recently discussed at the “La Jolla Proteomics Conference.”

Jing Wei, Ph.D., group leader of the proteomics biomarker discovery program at Biogen Idec, presented her experiences in protein biomarker development at the meeting. The aim of the research in her group is to apply proteomic tools—peptide separation, MS-based identification and quantification techniques—to clinical questions such as searching for new patient stratification tools, developing therapeutic biomarkers, and understanding the mechanism of action of putative biomarkers.

The work flow of the Biogen Idec team begins with sample preparation, followed by its elucidation through online 3-D nano LC- MS/MS technology. Biological samples to be analyzed include tissues and cells, both frozen and formalin-fixed; serum-free culture medium; tumor interstitial fluid; and several types of body fluid. The accumulated data are analyzed and subjected to data mining, allowing for appropriate candidate selection.

For quantitation, Dr. Wei and her coworkers use iTRAQ™ isobaric tagging system (Applied Biosystems). This set of reagent labels free amines at the N-terminus and Lysine residues of peptides in the complex mixture and allows up to eight samples to be differentially tagged and mixed together prior to separation and MS analysis.

According to Dr. Wei, the iTRAQ quantitation technology is powerful and effective using ESI-CID-PQD MS/MS on LTQ or LTQ-Orbitrap instruments (Thermo Scientific), and able to measure subtle changes of protein levels in the proteome. Pulsed-Q Dissociation (PQD), a new fragmentation technique that eliminates the low mass cut-off for ion traps and allows quantitation with iTRAQ labeling reagents.

The group has developed discovery programs in oncology and neurology and so far has identified a number of marker candidates that have proceeded to MS-based verification and antibody-based validation stages.

Protein Changes After Stroke

Zezong Gu, M.D., Ph.D., and his associates at both the "target=_University of Missouri-Columbia School of Medicine>University of Missouri-Columbia School of Medicine and the Burnham Institute, with Joseph Fox, Ph.D. and Gongyi Shi, Ph.D., at Bruker Daltonics have pursued the changes that occur in the proteins of the brain following stroke.

An important component of the process is abnormal proteolysis by matrix metalloproteinases which causes degradation of its substrates, an area of research in, which Dr. Gu has long focused.

Drs. Fox and Shi worked with Dr. Gu to utilize MALDI imaging mass spectrometry (MALDI-IMS) to determine both relative abundance and spatial distribution of the proteins in damaged tissues. The goal was to identify mass signatures correlated with abnormal proteolytic activity and apply this technology to the elaboration of the molecular profile after stroke.

Mice were subjected to cerebral ischemia followed by 24-hour reperfusion. The brains were processed and one set of sections was histologically stained while the adjacent tissue sections were analyzed via direct ionization of analytes from the tissue by the Ultraflex-III MALDI-TOF mass spectrometer. Selected ions were then displayed as pseudocolor images in which the color intensity corresponds to ion signal abundance.

In response to the ischemia, the brain showed localized swelling due to the traumatic injury. The overall anatomical brain structure was intact as evidenced by selective localization of a number of molecular ions. Interestingly, some molecular ions colocalize well with the injury area.

“Our studies have drawn us to the following conclusions,” Dr. Gu stated. “First, the MALDI-IMS tools are able to determine both the relative abundance and spatial distribution of the proteins and peptides in tissue traumatized by stroke. Second, multiple molecular ions colocalize within the injured area. Finally, these investigations point the way to identification of changes in brain architecture, proteins, and cell type that can act as targets for therapy.”

Hydrogen-Deuterium Exchange

Ansgar Brock, Ph.D., senior researcher at the Genomics Institute of the Novartis Research Foundation, discussed his group’s work for the Joint Center for Structural Genomics using hydrogen-deuterium exchange.

In this process a covalently bonded hydrogen atom is replaced by a deuterium atom or vice versa. Mass analysis of the products yields information concerning solvent accessibility of various parts of the molecule. From these data, a description of the tertiary structure of the protein can be gleaned.

“Our goal is to better understand the three-dimensional solution structure of proteins and their dynamics,” Dr. Brock said.

As part of the hydrogen-deuterium exchange procedure, proteins are subjected to a fast exchange procedure, before being proteolytically fragmented for analysis by liquid chromatography MS. The mass shifts caused by the exchange of deuterium for hydrogen or vice versa allow investigators to infer aspects of the folding of proteins, as the exchange will preferentially take place on residues that are part of flexible regions and exposed to the solvent interface.

The technique is highly suitable to study protein-protein interactions. “No single technique is sufficient to provide a complete picture of protein structure and dynamics,” Dr. Brock stated. “This is why the hydrogen-deuterium exchange data produced by the genomics center are combined with results from the limited proteolysis experiments and analytical size-exclusion data to reinforce the picture.”

Ribo-Proteome Pool Sizes

According to Michael T. Sykes, Ph.D., a post doctoral fellow at the Scripps Institute, protein pool sizes are determined by the balance of synthesis and utilization, reflecting the dynamics of protein turnover. Dr. Sykes and his coworkers have explored quantitation of pool size measurements using stable isotope pulse- labeling and liquid chromatography-coupled mass spectrometry. “Quantitative proteomic mass spectrometry involves comparison of the amplitudes of peaks resulting from different isotope labeling patterns, including fractional atomic and residue labeling,” he explained

Quantitative analysis of proteins from ribosomes recovered at different times allows delineation of the kinetics of incorporation of 15N into each individual ribosomal protein.

As Dr. Sykes discussed, estimated pool sizes range from 2% to 40% for proteins in the 30S subunit, though most weigh in with an approximately 10% pool size. As could be anticipated, proteins with unusually large pool sizes turn out to be those with multiple known functions.

Pool sizes generally correlate with the in vitro assembly order. Using this relationship along with an estimate of the total assembly time of a single ribosome, Dr. Sykes and his coworkers were able to partition out the relative contributions of free proteins and ribosomal intermediates to the total pool.

“Our research represents a generally applicable approach to monitoring pool sizes and protein pool dynamics of the ribo-proteome in vivo.”

He described a least-squares Fourier transform convolution approach that can be applied to many types of quantitative proteomic data, including data from stable isotope labeling by amino acids in cell culture and pulse-labeling experiments.

Eric Hwang, Ph.D., assistant professor in the department of biological science and technology at the National Chiao Tung University, Taiwan, presented a quantitative analytical study of the microtubule-associated proteome during neurite formation. The studies that Dr. Hwang discussed were carried out in collaboration with Shelley Halpain, Ph.D., professor of biological sciences, UC San Diego.

“Neuritogenesis is the process that underlies establishment and plasticity of neuronal networks,” Dr. Hwang explained. The initiation of neurite outgrowth and the elaboration of axonal and dendritic processes involves a profound morphological reorganization of the microtubule cytoskeleton. This transformation is, in turn, controlled by the microtubule-associated proteins (MAPs). Dr. Hwang’s goal was to perform a comprehensive MAP proteomic analysis in order to reveal new protein participants.

As a model system for their proteomics characterization, he and his collaborators took advantage of the P19 cell line, derived from an embryonal carcinoma induced in a C3H/He strain mouse. The line has the extremely useful property that it can be induced to differentiate into neuronal and glial cells in the presence of retinoic acid. Moreover, cellular aggregates can be forced to differentiate into cardiac and skeletal muscle by dimethyl sulfoxide.

By taking advantage of the technique of spectral counting in their investigations, they identified about 800 proteins from each sample. Spectral counting quantifies relative protein concentrations in pre-digested protein mixtures analyzed by liquid chromatography online with tandem MS.

The results derived from spectral counting were compared with in-gel stable isotope labeling studies with the auspicious outcome that the two data sets showed a high degree of correlation. When pre- and post-neurite induction proteomes were quantitatively compared, the overall functional profiles appeared quite similar; however, the specific composition of the proteome changed significantly.

The most abundant classes of proteins identified are known cytoskeleton-associated proteins, nucleic acid-binding proteins, and proteins involved in phosphorylation. Protein kinase A components and specific A-kinase anchoring proteins were strongly increased in the post-neurite induction proteome, emphasizing a central role for PKA signaling. Interestingly, collapsin-response mediator proteins, spectrins, and components of the dynactin complex were also significantly increased in the neuronal microtubules.

According to Dr. Hwang, these results provided a comprehensive picture of the changes in the microtubule-associated proteome during neuritogenesis. “The most significant finding is the identification of potential new players, such as tripartite motif protein 2, that regulate key aspect of this process,” he stated.

Future

The conference participants stressed the dynamic nature of the proteome and the role of post-translational modifications in altering cellular expression. The use of MS and other advanced technologies including stable-isotope labeling, ICAT, iTRAQ and SILAC bring protein quantification to a new level of sophistication and allows an understanding of these changes and their role in disease states.

Currently there is an ongoing development of new technologies to profile and analyze the cellular proteomes at an even more refined level. Alterations in protein levels even in the absence of protein expression changes are often linked to cellular responses and disease states. Using these new approaches investigators will provide a system-wide understanding of the protease web and the web-sculpted proteome, with important implication for the understanding of pathological events and development of innovative pharmacological agents.