**4.2.2 Proteomics**

Proteomics can be defined as the study of all the proteins codified by a genome, in a given tissue of a given organism at a given time. It involves studying how the concentration or "relative abundance" of the proteins change under a certain stimulus, protein conformational changes, protein – protein interactions (or "interactomics"), among others, as well as the use and development of experimental and bioinformatics technologies necessary to perform these studies. In this regard, protein separation techniques are essential. The fundamental separation methods used in proteomics are Sodium Dodecyl Sulfate- Polyacrylamide Gel Electrophoresis (SDS-PAGE) and/or Two-Dimensional Gel Electrophoresis (2DGE) and mass spectrometry (MS); the latter is used as separation but also as identification tool. Figure 7 depicts a classical proteomics experiment, starting from a biological sample, followed by preliminary fractionation by liquid chromatography and after that separation by 2DGE, and finally identification of protein spots by MS.

Fig. 7. Simplified representation of a gel-based proteomics experiment. Starting from a biological sample, a protein extract is obtained using different biochemical techniques to fractionate the sample. These fractionation steps allow the enrichment of protein fractions in low abundance proteins and to reduce the complexity of the sample. The protein fractions are then resolved by SDS-PAGE or 2D GE, and finally protein spots are excised form the gel and then analyzed by mass spectrometry in order to determine their identity and structural properties.

From a Biomedical point of view, proteomics is an important field in the task of discovering new biomarkers that reflect the health/disease status of living organisms. The use of proteomics with this purpose has been somewhat limited due to technical hurdles related to the high complexity of the biological samples to be analyzed, usually blood serum or plasma, but also cerebrospinal fluid, urine and tears. These samples show a wide dynamic range of protein concentration, exceeding 1010. This means that the most abundant protein in plasma (albumin), for example, has a concentration 1010 times higher than that of the less abundant protein (such as transcription factors).

Two-dimensional gel electrophoresis can resolve a concentration range of up to 104, and therefore 2DGE images or "maps" of blood plasma are dominated by the highly abundant

efficiency, maintaining the MAb stability. Optimizing pore size facilitates mass transfer from mobile phase bulk towards the hydrophobic ligand. Kostareva et al. (2008) purified a heteropolymer (a kind of MAb consisting of a dual antibody conjugate) by HIC. They found that using a Propyl-HIC resin the heteropolymer was efficiently separated from free MAbs, thus confirming the ability of HIC for separating aggregates from monomers, and also its

Proteomics can be defined as the study of all the proteins codified by a genome, in a given tissue of a given organism at a given time. It involves studying how the concentration or "relative abundance" of the proteins change under a certain stimulus, protein conformational changes, protein – protein interactions (or "interactomics"), among others, as well as the use and development of experimental and bioinformatics technologies necessary to perform these studies. In this regard, protein separation techniques are essential. The fundamental separation methods used in proteomics are Sodium Dodecyl Sulfate- Polyacrylamide Gel Electrophoresis (SDS-PAGE) and/or Two-Dimensional Gel Electrophoresis (2DGE) and mass spectrometry (MS); the latter is used as separation but also as identification tool. Figure 7 depicts a classical proteomics experiment, starting from a biological sample, followed by preliminary fractionation by liquid chromatography and

after that separation by 2DGE, and finally identification of protein spots by MS.

Fig. 7. Simplified representation of a gel-based proteomics experiment. Starting from a biological sample, a protein extract is obtained using different biochemical techniques to fractionate the sample. These fractionation steps allow the enrichment of protein fractions in low abundance proteins and to reduce the complexity of the sample. The protein fractions are then resolved by SDS-PAGE or 2D GE, and finally protein spots are excised form the gel and then analyzed by mass spectrometry in order to determine their identity and structural

From a Biomedical point of view, proteomics is an important field in the task of discovering new biomarkers that reflect the health/disease status of living organisms. The use of proteomics with this purpose has been somewhat limited due to technical hurdles related to the high complexity of the biological samples to be analyzed, usually blood serum or plasma, but also cerebrospinal fluid, urine and tears. These samples show a wide dynamic range of protein concentration, exceeding 1010. This means that the most abundant protein in plasma (albumin), for example, has a concentration 1010 times higher than that of the less

Two-dimensional gel electrophoresis can resolve a concentration range of up to 104, and therefore 2DGE images or "maps" of blood plasma are dominated by the highly abundant

suitability for purifying MAbs.

**4.2.2 Proteomics** 

properties.

abundant protein (such as transcription factors).

proteins, namely albumin, immunoglobulin, fibrinogen, among others, thus preventing the detection of low abundance proteins (Hoffmann et al., 2007). Mass spectrometry can resolve a range of 103 in a single spectrum, but combined with separation steps it can resolve a range of up to 106 (Jacobs et al., 2005). This range is still wide, and thus many proteins cannot be detected. Then, chromatographic separation steps should be used before 2DGE in order to reduce the dynamic range of proteins concentration, and consequently increase resolution.

The most abundant proteins in blood plasma are albumin, immunoglobulin, transferrin, haptoglobin, fibrinogen and α-1-antitrypsin, which amount to 90% of total protein mass. Then, total or partial depletion of these proteins allows detecting low abundance proteins. Different methods can be used to deplete these proteins, being liquid chromatography the most popular one (Nakamura et al., 2008). Different chromatographic strategies are available for this purpose, including affinity dye-based chromatography for albumin depletion, affinity to protein A and G for immunoglobulin depletion, specific antibodyaffinity columns (Linke et al., 2007), and affinity columns containing lectins, peptides or inorganic ligands (Salih, 2005). Liquid chromatography has the advantage of being easy to use and to scale-up, but are relatively expensive, especially those involving affinity columns. Another drawback of affinity chromatography is the non-specific interactions that lead to the loss of some proteins, with the consequent loss of information (Altintas & Denizli, 2006). In order to overcome the disadvantages of affinity chromatography for its use in blood plasma proteomics, several complementary strategies have been examined, such as sequential anion and cation exchange chromatography followed by 2DGE; and strong cation exchange chromatography followed by liquid-phase isoelectric focusing (Ottens et al., 2005; Barnea et al., 2005). Since these approaches considerably improve the capacity to detect low abundance proteins, it was suggested that the optimization of combinatorial processes by coupling immuno-affinity depletion with other conventional separation methods such as hydrophobic interaction chromatography will probably lead to significant advances in proteomics (Mahn et al., 2010). Despite the research conducted in this area, there is still a lack of optimized processes that ensure detection of the complete proteome of a tissue or cell.

## **4.2.2.1 Plasma fractionation by HIC**

The applicability of HIC as a plasma fractionation method has been recently proposed. Geng et al. (2009) developed a two-dimensional liquid chromatography resin having two types of ligands, and hence that functions in two retention modes: cation exchange and hydrophobic interaction. This method could be applied to the fast fractionation of intact proteins before mass spectrometry analysis. The results obtained by HIC were similar to those obtained by ion exchange chromatography. On the other hand, a HIC matrix consisting of highly acetylated agarose has been used for the isolation of immunoglobulin from porcine serum, with a relative success (Ramos-Clamont et al., 2006).

Recently, Mahn et al. (2010) investigated if the performance of 2DGE could be improved by fractionating blood plasma through a HIC step, thus reducing the relative concentration of some highly abundant proteins in plasma. First, the hydrophobicity of the main 56 proteins present in blood plasma was determined. To do this, the amino acidic composition of the proteins was considered, and hydrophobicity was calculated by equation (16) based on the methodology proposed by Salgado et al. (2005). In equation (16), φaai is given by the Cowan– Whittaker hydrophobicity scale in its normalized form (see Table 1), ni is the number of amino acids of type i in the protein, si,max is the maximum solvent accessible area that an amino acid X can have when forming part of the G–X–G tripeptide in extended conformation (Miller et al., 1987).

$$ASH = \sum\_{i=1}^{20} \left( \phi\_{\text{aui}} \cdot \frac{n\_i \cdot s\_{i,\text{max}}}{\sum\_{j \in A} n\_j \cdot s\_{j,\text{max}}} \right) \tag{16}$$

After that, a cluster analysis was performed in order to classify them as low, medium or high hydrophobicity proteins. This analysis showed that the highly abundant proteins, i.e. albumin, immunoglobulins, fibrinogen and haptoglobin, exhibited a medium hydrophobicity, and thus they fell in the same cluster. With this information, a HIC step was designed to deplete highly abundant proteins from rat plasma samples. The HIC step consisted of stepwise elution to separate the three groups of proteins (low, medium and high hydrophobicity) using a maximum concentration of 2 M ammonium sulfate, and concentration for elution of 0.6 M (to desorb low hydrophobicity proteins), 0.5 M (to desorb medium hydrophobicity proteins), and 0.0 M (to desorb the highly hydrophobic proteins).

Finally, the depleted samples were analyzed by 2DGE and the performance of the HIC pre-fractionation step was compared with that exhibited by a commercial immunoaffinity column. The reproducibility of 2DGE was similar to that obtained from immunoaffinity depleted plasma. However, HIC was more successful in depleting albumin and α-1-antitrypsin. Besides, HIC resulted in a much lower increment of immunoglobulin and haptoglobin abundances than the immuno-affinity column. Then, HIC depletion allowed detecting twice the number of protein spots than immuno-affinity depletion did. Therefore, HIC could be used as a depletion method complementary to affinity columns. The operating conditions in HIC could be optimized in order to maintain the high number of spots that are detected if HIC is used as the sole depletion method. Finally, given the relatively low cost of HIC supports and HIC operation, its use could be proposed as a convenient choice for depleting highly abundant proteins in plasma samples prior to 2DGE-based proteomics.

### **4.2.2.2 Analysis of protein interaction networks by HIC**

Protein–protein interactions are essential in biological processes. All the interactions in a cellular system are known as protein interaction network or 'interactome'. In Biomedicine there is great interest in recognizing these interactions, aiming to establish the role they play in certain diseases. The traditional approaches to study protein-protein interactions are the antibody pull-down method (APD) and the yeast two-hybrid method (YTH). Despite their popularity, these methods have some disadvantages. It is very likely that a protein forms part of different complexes; then, in an APD experiment, antibodies targeting such a protein will pull down together all the complexes where the protein participates, making them appear to be part of a single large complex, confusing the biological interpretation of the results. The YTH is an "*in vivo*" method that allows detecting only binary interactions. It tends to give false positives and is limited to binary interactions. Therefore it is not useful in studying the dynamics of complex formation triggered by different stimuli (Corvey et al., 2005).

Liu et al. (2008) investigated the potential of chromatography to allow the simultaneous examination of multiple protein complexes along with comparing and validating results from the traditional methods. Since protein complexes remain intact during mild forms of elution in AC, a similar behavior should be expected in other chromatographic supports, such as IEC and HIC. They studied the extent to which protein interaction partners from yeast (*S. cerevisiae*) lysate remain associated during IEC, SEC and HIC. Most protein complexes remained intact, and all the proteins forming part of the complex migrated as a single unit. Protein complexes exhibited a chromatographic behavior different from that shown by the individual proteins that compose the complex. Accordingly, studying protein complexes could be easily performed by multidimensional chromatographic methods when at least one of the fractionation dimensions included SEC of native proteins. This method enables the study and recognition of several protein complexes simultaneously, avoiding the use of genetic engineering.
