Clin Chem Lab Med 2005;43(12):1327–1333  2005 by Walter de Gruyter • Berlin • New York. DOI 10.1515/CCLM.2005.227 2005/183 Article in press - uncorrected proof Separation of human serum proteins using the Beckman- Coulter PF2D system: analysis of ion exchange-based first dimension chromatography Isabella Levreri 1, *, Luca Musante 2 , Andrea Petretto 3 , Maurizio Bruschi 2 , Giovanni Candiano 2 and Giovanni Melioli 1 1 U.O. Laboratorio Centrale di Analisi, 2 Laboratorio di Fisiopatologia dell’Uremia, U.O. di Nefrologia, 3 Laboratorio di Spettrometria di Massa, Core facility, Istituto Scientifico Giannina Gaslini, Genova, Italy Abstract The ProteomeLab PF2D protein fractionation system is a rapid, semi-automated, 2 D-HPLC instrument that uses two different methods to separate plasma serum proteins: ion-exchange chromatography using a wide pH range in the first dimension, and non-porous reverse-phase chromatography in the second dimen- sion. Because this methodology has only very recent- ly been introduced in proteomic laboratories, little is known about the characteristics of PF2D fraction- ation of human serum proteins. To evaluate the sys- tem’s application in a clinical laboratory setting, the characteristics of the ion-exchange chromatography- based separation were analyzed. Following fraction- ation of human serum proteins on a linear pH gradient (ranging from 8.0 to 4.0), each fraction was collected in a cool module of the instrument. Different fractions obtained from the first dimension were then pooled together and loaded on classic 2D gel electro- phoresis instrumentation. The different spots obtain- ed were then checked against the Swiss-Prot Data- base. A total of 36 human serum proteins were iden- tified in different PF2D-generated fractions. Some important features of the separation system were observed. Different eluted fractions contained differ- ent proteins, thus demonstrating the reliability of the fractionation system. The proteins were also fraction- ated according to the theoretical isoelectric point (pI). This was consistent with the evidence that the vast majority of immunoglobulins, characterized by an alkaline pI, were not retained by the column and were eluted in the unbound fraction. This outcome also underlines a practical advantage: fractions eluted from pH 8 to pH 4 contained virtually immunoglobu- lin-depleted serum proteins. This finding supports an immediate use of the PF2D system in a clinical set- ting, where abundant proteins should be clearly iden- *Corresponding author: Isabella Levreri, BD, Laboratorio Centrale di Analisi, Istituto G. Gaslini, Largo Gerolamo Gaslini 5, 16147 Genova, Italy Phone: q39-010-5636557, Fax: q39-010-3994168, E-mail: i_levreri@yahoo.com tified to enable evaluation of other less abundant, but potentially relevant, species. Keywords: 2D gels; plasma protein fractionation; ProteomeLab PF2D; pI value. Introduction The plasma proteome is both the most useful and the most difficult version of the human proteome (1). In the past, a number of approaches have been attempt- ed to analyze the complexity of plasma proteins for clinical uses (2). Nevertheless, the actual translation of proteomic techniques into clinical pathology prac- tice has been hampered by the difficulties entailed by the use of two-dimensional (2D) gels and mass spec- trometry in a clinical setting. The recent availability of a partially automated system based on 2D liquid/liq- uid separation of a protein mixture, the Proteome- Lab PF2D, has raised clinical pathologists’ interest in the analysis of serum and plasma proteins in both healthy and disease states. This system allows the separation of proteins on the basis of their iso- electric point (pI), followed by a second separation of proteins, collected within a given pH range and based on the strength of hydrophobic interactions, using reverse-phase chromatography. The final product is a 2D separation profile representing the most abundant serum proteins, as well as other low-abundance spe- cies detectable in circulating blood. Despite the potential application of this technique to clinical pathology, little is known regarding the sys- tem’s specific features. In particular, the actual ability to separate proteins on the basis of the interaction between an ion-exchange matrix and a pH gradient from 8.0 to 4.0 should be better investigated. Indeed, serum proteins interact with the matrix in the pres- ence of a positive or negative net charge. At a certain pH, the net electric charge should be zero and, in the absence of any interaction with the matrix, the protein should be able to flow along the column and to be collected. Nevertheless, the theoretical pI, calculated using freeware tools on the net, is based on the alge- braic sum of both the positive and negative charges of each amino acid (3). At present, it is unknown whether this result, which does not take into account the three-dimensional (3D) structure of the protein, is similar to the experimental value that can be achieved using the first dimension of PF2D. On the contrary, PF2D separates proteins apparently without interact-