Relationship Between Peptide Physicochemical Properties And Measurability In Mass Spectrometer And Its Effect On Quantitative Proteomics

The invention of MALDI (Matrix-Assisted Laser Desorption Ionization) and ESI (Electrospray Ionization) made it possible to transfer the macromoleculars from liquid to gas phase and the develope of MS-based protein-identification methods, thus catalyzed the fast development of proteomics technology. However, due to the limitation of mass spectrometric technology in the optimal m/z range for fast dissociation and detection of current mass spectrometer is less than 2000, than the so-called"bottom-up"strategy for protein identification was developed and popular based upon the realistic capability of mass sepctrometry. The strategy of bottom-up impiled the use of mass spectrometer to determine the masses of the proteolytic fragments produced by a specific enzyme or chemical. The experimental data was then compared with the masses predicted from all the reading frames of the gene sequences in the database in order to verify the protein sequence filtered against stringent algorithms. All of mass spectrometric data based quantitative proteomics is derived from quantification information of peptides. Peptides are actual targets of large-scale proteomics identifications. Therefor, the effects of physicochemical properties on peptide identification are crucial and fundantmental for current proteome research. Replicate runs can reveal the inherent problems of peptide identification and quantification by minimizing the random sampling effect of mass spectrometry platforms. However, almost no systemic analysis using the replicate runs strategy to study the difference of peptide measurability or quantitative behaviors has been reported.Herein, we present a systematic study using a replicate run strategy to probe the inherent influence of peptide physicochemical properties on several important concerns. These concerns include relationship between peptide identification and sample loading amounts and its application in protein quantification, peptide identification and quantification at different sample complextity, effects of protein total digestion and in-gel digestion, peptide identification and quantification at different mass spectrometer with different mass measurement accuracy or scan-rate of mass spectrometer. We calssified the peptides into different categories based upon physicochemical properties and statistical analysis, and proposed several general rules to optimize experiment design and quantitative proteomics.(1) Optimization of sample loading amount. The relationship between sample loading amounts and peptide identification is crucial for the optimization of the proteomic experiment, but few studies have addressed this matter. In chapter 2, we present a systematic study using a replicate run strategy to probe the inherent influence of peptide physicochemical properties on the relationship between peptide identification and sample loading amounts and its effects on protein quantification. Ten replicate runs for a series of laddered loading amounts (ranging between 0.01~10μg) of total digested proteins from Saccharomyces Cerevisiae were performed with nanoLC-LTQ-FT to obtain a nearly saturated peptide identification and thus differentiated linear correlativity of peptide identifications by the commonly used peptide quantitative index, including spectral counting (SC, from MS/MS data) and the area of constructed ion chromatograms (XIC) (SA, from MS & MS/MS data) in the given experiments. Peptide physicochemical properties showed little difference for the linear correlativity of SA-based peptide quantification and loading amounts, while obvious discrimination was noted with SC. By extension, we found that quantifying the target protein by selecting peptides with good parallel linear correlativity based upon SA as signature peptides and revising the data by multiplying with the reciprocal of the slope coefficient could optimize the linear protein abundance relativity at every amount range, and thus extend the linear dynamic range of the label-free quantification. This empirical rule for linear peptide selection (ERLPS) can be adopted to correct the comparison results in proteolytic peptides-based quantitative proteomics like AMT (accurate mass tag)and MRM in targeted quantitative proteomics as well as for tag-labeled comparative proteomics.(2) Effects of sample complexity. Under the constant identification capability, along with the complexity of the peptide sample increased, we can't always obtain accordingly infinite increase of peptides species. For quantitative proteomics, the effects on peptides distribution increases with more separarison, thus bring to more errors for the final comparison results. Different peptide showed discrepant sensitivity on the complexity degree of the sample, then the spectral signal intensity summation of all peptides comprised of the targeted proteins will not reflect the actual change of intact protein. In chapter 3, after simplifying the sample complexity using Free Flow Electrophoresis (FFE), we selected fractions remote between each other for in-solution digestion, the following change of spectral counts showed the peptide identification chance at different complexity. Spectral signal intensity of most peptides will decrease with multiplied complexity, squeezing out peptides with low abundance signal intensity. Coverage rate of both high abundance and low abundance proteins decreased. Quantifying the target protein by selecting peptides with good parallel linear correlativity at different complexity as signature peptides could optimize the the final protein quantification result.(3) Comparison between in-solution digestion and in-gel digestion. As we know all, the classic method of two-dimensional polyacrylamide gel electrophoresis (2-DE), especially the second dimension of SDS-PAGE, is widely used in proteome research. However due to the rough reliability of image quantitation, the more precise mass spectrometric data based quantification strategies are prevalent in current quantitative proteomics. Mass spectrometrical analysis of gel-separated proteins is mainly carried out on proteolytic peptides generated by tryptic digestion of the gel-immobilized protein. However, a complicationwith the in-gel digestion procedure is that only a limited part of the protein sequence is normally recovered for the analysis, frequently less than 50%. This is mainly due to the difficulties in extraction of large peptides from the gel matrix and to variable accessibility of the proteolytic enzyme to the protein since the latter is tightly immobilized onto the surface of the gel. In chapter 4, we compared the identified results between the peptides extracted from ingel digestion and that of in-solution digestion of one fraction from yeast protein by nanoLC-LTQ-FT with over ten times analysis respectively to observe peptides loss during in-gel digestion. From the physicochemical properties analysis we could see the loss peptides with more alkaline amino acids, more hydrophobic value, longer length during in-gel digestion process. Then we should compare the quantification index between the result obtained from the peptides with in-gel digestion and in-solution digestion, and select the peptides with both higher and similar recovery rate as signature peptides, following with revising the data by multiplying with the reciprocal of recovery rate to obtain the final protein quantification.(4) Effects of sensitivity and scan-rate of mass spectrometer. The sensitivity, mass measurement accuracy and scan-rate of mass spectrometer will influence peptide identification and corresponding protein quantification. Spectral signal intensity of different peptides of one protein changed differently with different mass spectrometers, which was derived from the difference of peptide measurability at different platform. To illustrate which peptides with given physiochemical properties will increase with higher sensitivity and faster scan-rate, we compared the identified results between LCQ, LTQ and LTQ-FT of one fraction of yeast separated by reversed-phase liquid chromatography. Improved sensitivity and faster scan rate of mass spectrometer will identify much more peptides during large-scale proteomics research. Improvement of mass measure accuracy enhances identification reproducibility, while increased of scan-rate can not improve reproducibility. We could see the discrepancy of spectral signal intensity for different peptides of one protein with different mass spectrometers, and select peptides with similar change rate as signature peptides to negotiate the final quantification results of one sample at different platform.