Semi-automatic analysis of two-dimensional electrophoretic gels

A powerful technique, two dimensional gel electrophoresis, has recently become available for the analysis of biological material. This technique allows the proteins of a sample to be mapped on a two-dimensional polyacrylamide gel. The research outlined here involves a multi-disciplinary approach to the semi-automatic comparison of two gels of biologically similar materials. The aim of the project is to identify the significant differences between the gels and point them out to the electrophoresis expert. This job, when done by hand, can be time consuming, since the gels are distorted and can contain hundreds of protein spots. Automatic analysis of 2D gel separations has proved to be extremely difficult using statistical methods. Non-reproducibility of gel separations is also difficult to overcome using automatic systems. However, the human eye is extremely good at recognizing patterns in images, and human intervention in semi-automatic computer systems can reduce the computational complexity. This paper focuses on the intelligent knowledge-based systems aspects of the project, and in particular the work of the first author. The results demonstrate that the semi-automatic approach can produce reliable results quickly. Further work, beyond the scope of this paper, focuses on creating a database of gels for comparison.     

A probability-based error spot filtering method in protein 2-DE image spot pattern matching analysis

Given a number of 2-DE gel images of the same kind, identifying the spot of each image for the same protein is an important task to monitor the expression level change of the protein. For this purpose, the gel-based two-dimensional electrophoresis method (2-DE) is widely used since it separates thousands of proteins in a sample cost-effectively. However, this approach suffers from inherent noises and irregular geometric distortions of spots observed in a 2-DE gel image. This paper proposes a probability-based error filtering method that can find more reliable spot-matching results, so that the accuracy of protein expression analysis can be improved. The performance of the proposed method is analyzed by various experiments on real 2-DE gel images of human liver tissues.     

Impact of incomplete DNase I treatment on human macrophage proteome analysis

The aim of our study was to analyze the proteomic pattern of human macrophages obtained over a 4 year period from blood donors. The purpose was to simulate a long-term clinical study to assess the application of 2-D DIGE technique for differential proteomic analysis of these scarce samples. Bioinformatic analysis of 2-D DIGE gels of 19 different cultures of macrophages assessed whether they did or did not contain at least specific five spots identified by MS as being or containing bovine deoxyribonuclease I (DNase I). Bovine DNase I was used during sample treatment to remove nucleic acids from protein extracts. Macrophages were classified in two groups, which appeared to be differentiated by the completeness of DNase I treatment. Further detailed analysis revealed a different proteomic pattern of macrophage protein samples according to the completeness of this treatment. The major group of proteins affected, accounting for one third of the differentially expressed proteins, included proteins involved in cell motion and actin cytoskeleton reorganization. The use of DNase I for the removal of nucleic acids from protein samples must be avoided in proteomic studies since it can generate bias in the analysis of protein expression patterns.     

MS characterization of apheresis samples from rheumatoid arthritis patients for the improvement of immunoadsorption therapy - a pilot study

Identification of proteins from apheresis samples was performed by both SDS-PAGE and 2-D gel separation of eluted proteins from staphylococcal protein A-based immunoadsorption columns (Prosorba^®) followed by MS peptide mass fingerprinting and MS/MS peptide sequencing on a MALDI QIT TOF mass spectrometer. MS/MS peptide sequencing was performed in conjunction with a micro reversed phase HPLC configured with an online MALDI plate-spotting device. Apheresis treatment had been performed in three patients with longstanding therapy refractory rheumatoid arthritis. 2-D gels displayed ca. 500 spots representing proteins that were eluted from the Prosorba^® columns. From 54 gels, a total of 1256 protein spots had been picked and yielded in the identification of 56 non-redundant proteins without counting isoforms. Proteins from the eluates belong to five major groups comprising (i) immunoglobulins (IgG, IgA, IgM heavy and light chains; about 40% of the spots), (ii) proteins involved in coagulation, (iii) HDL/LDL-associated proteins, (iv) proteins from the complement system, and (v) acute phase proteins. MS analysis showed that the full-length C3 complement protein had been cleaved upon complement activation, presumably on the column, such that the anaphylatoxin C3a was produced and released during therapy. Our results are consistent with clinical observations on both patient responses to therapy and reported adverse events. For the first time, direct molecular information has become available to support mechanistic reasoning for the principle of function of staphylococcal protein A-based immunoadsorption therapy and for the explanation of adverse events. According to our results, removal and/or modulation of immune complexes together with complement activation can be regarded as the major events that are taking place during Prosorba^® therapy. In order to avoid complement activation and induction of an inflammatory cascade, we suggest the prevention of C3a anaphylatoxin-related reactions during immunoadsorption therapy.     

2-D DIGE identification of differentially expressed heterogeneous nuclear ribonucleoproteins and transcription factors during neural differentiation of human embryonic stem cells

Neural stem cells (NSC) are progenitors that can give rise to all neural lineages. They are found in specific niches of fetal and adult brains and grow in vitro as non-adherent colonies, the neurospheres. These cells express the intermediate filament nestin, commonly considered an NSC marker. NSC can be derived as neurospheres from human embryonic stem cells (hESC). The mechanisms of cellular programming that hESC undergo during differentiation remain obscure. To investigate the commitment process of hESC during directed neural differentiation, we compared the nuclear proteomes of hESC and hESC-derived neurospheres. We used 2-D DIGE to conduct a quantitative comparison of hESC and NSC nuclear proteins and detected 1521 protein spots matched across three gels. Statistical analysis (ANOVA n = 3 with false discovery correction) revealed that only 2.1% of the densitometric signal was significantly changed. The ranges of average ratios varied from 1.2- to 11-fold at a statistically significant p-value <0.05. MS/MS identified 15 regulated proteins previously shown to be involved in chromatin remodeling, mRNA processing and gene expression regulation. Notably, three members of the heterogeneous nuclear ribonucleoprotein family (AUF-1, and FBP-1 and FBP-2) register a 54, 70 and 99% increased expression, highlighting them as potential markers for NSC in vitro derivation. By contrast, Cpsf-6 virtually disappears with differentiation with an 11-fold drop in NSC, highlighting this protein as a novel marker for undifferentiated ESC.     

A statistical analysis of spot variation using the two-dimensional polyacrylamide gel electrophoresis

For the development of valid algorithms for matching protein spots between two-dimensional gels, it is essential that one has an understanding of the relative roles of the many sources of variability affecting the location of spots. We first consider the contribution of observers to the measurement variability of spot location, arriving at a simple model for these effects. Next we present an analysis of the variability in spot locations for a sample of gels containing duplicate gels for each sample. Our data indicate that both differences between duplicates and between samples are considerable, and that the size of the effects depends on the region of the gel being considered. In the discussion we examine several matching strategies that match large groups of gels based on algorithms which match two gels at a time.     

Urine collected from diapers can be used for 2-D PAGE in infants and young children

Urinary proteomic profiling has potential to identify candidate biomarkers of renal injury in infants provided an adequate urine sample can be obtained. Although diapers are used to obtain urine for clinical evaluation, their use for proteomic analysis has not been investigated. We therefore performed feasibility studies on the use of diaper-extracted urine for 2-D PAGE. Pediatric waste urine (2–20 mL) was applied to gel-containing, non-gel and cotton-gauze diapers and then mechanically expressed. Urine volume and total protein were measured pre- and post-extraction. Proteins were separated via 2-D PAGE following application of urine (20–40 mL) to each matrix. 2-D PAGE was also performed on clinical specimens collected using each diaper type. Differences in the adsorption and retention of urine volume and protein were noted between matrices. Non-gel and cotton-gauze diapers provided the best protein/volume recovery and the lowest interference with the Bradford assay. 2-D PAGE was also successfully completed using urine samples from both cotton fiber matrices. Conversely, samples from low-gel diapers demonstrated poor protein separation and reproducibility. Diapers containing cotton–fiber matrices appear adequate for 2-D PAGE. Qualitative and quantitative analyses of resolved proteins using replicate, high-resolution gels will be required, however, before diaper-extracted urine can be applied in proteomic profiling.     

The glomerular proteome in a model of chronic kidney disease

Adequate kidney function is crucial in sustaining vertebrate homeostasis. Certain diseases can diminish renal function and lead to end-stage renal disease. Diabetes mellitus and hypertension are the main causes of glomerulosclerosis and albuminuria in adults. The molecular mechanisms that trigger these maladaptive changes are still unsatisfyingly described. We previously introduced 2-D DIGE in combination with focused tissue isolation methods to analyze protein expression in glomeruli. Glomeruli, the crucial compartments in albuminuric renal diseases, were extracted using magnetic particles from subtotally nephrectomized FVB mice (n?=?6); this 5/6 nephrectomy in FVB mice is a model of chronic kidney disease. Analysis of protein expression levels from glomerular protein lysates was performed using 2-D DIGE and compared with glomerular protein lysates from mice that underwent sham surgery. The comparison of about 2100 detectable spots between both groups revealed 48 protein spots that showed significant differential expression. Of those, 33 proteins could be identified using nanoLC-ESI MS. The metalloproteinase meprin 1 alpha, the beta galactoside-binding-lectin galectin-1 and dimethylarginine dimethylaminohydrolase 1, a key enzyme in NO metabolism, were found to be differentially regulated, thus implying a role in the pathogenesis and pathophysiology of progressive kidney disease. In conclusion, 2-D DIGE protein analysis of smallest sample sizes from specific organ compartments provides focused protein expression results, which help in gaining an understanding of the molecular mechanisms of chronic kidney disease.     

Proteomic profiling of cancer of the gingivo-buccal complex: Identification of new differentially expressed markers

Tobacco-related oral cancer is the most common cancer among Indian males, gingivo-buccal complex (GBC) being the most affected subsite due to the habit of chewing tobacco. Proteins from the lysates of microdissected normal and transformed epithelium from clinically well-characterized tissue samples of the GBC were separated by two-dimensional gel electrophoresis to identify differentially expressed proteins. Eleven protein spots showed differential expression, which could withstand the stringency of statistical evaluation. The observations were confirmed with additional tissues. Nine of these differentiators were identified by MS as lactate dehydrogenase B, α-enolase, prohibitin, cathepsin D, apolipoprotein A-I, tumor protein translationally controlled-1, an SFN family protein, 14-3-3σ and tropomyosin. Cluster analysis indicated that these proteins, as a coexpressed set, could distinguish normal and transformed epithelium. Functionally, these differentiator molecules are relevant to the pathways and processes that have been previously implicated in oral carcinogenesis and could therefore be investigated further as a panel of markers for management of cancer of the GBC.     

Proteome of human endometrium: Identification of differentially expressed proteins in proliferative and secretory phase endometrium

Purpose: To exploit the potential of proteomics to identify and study additional yet‐unidentified important proteins present in human endometrium. Experimental design: The proteome of human endometrium would be established using 2‐DE and MALDI and the data analyzed to identify differential protein expression in the proliferative and secretory phase of the menstrual cycle using PDQuest software and MALDI. Results: In the present work, 2‐DE of human endometrium protein led to the resolution of over 200 spots. Subsequent MALDI analysis of 215 spots allowed the identification of 194 proteins. A total of 57 out of the 215 spots were found to be differentially expressed, out of which 49 could be identified using MALDI. These differentially expressed proteins included structural proteins, molecular chaperones, signaling proteins, metabolic proteins, proteins related to immunity, RNA biogenesis, protein biosynthesis and others. The differential expressions of seven representative proteins in secretory and proliferative phase endometrium tissue were confirmed by immunoblot analysis. Conclusion and clinical relevance: This study establishes the 2‐D proteome of human endometrium represented by 194 identified protein spots. The present data provides an important clue towards determining the function of these proteins with respect to endometrium related diseases.     

Identification and validation of novel CSF biomarkers for early stages of Alzheimer's disease

The pathology of Alzheimer's disease (AD) begins years prior to clinical diagnosis. The development of antecedent biomarkers that indicate the presence of AD pathology and predict risk for decline in both cognitively normal and mildly impaired individuals will be useful as effective therapies are developed. While cerebrospinal fluid (CSF) markers such as amyloid-β (Aβ) 42 and tau are useful, additional biomarkers are needed. To identify new markers, we utilized 2-D difference gel electrophoresis (2-D DIGE) of individual CSF samples from subjects with very mild AD versus controls after depletion of high-abundant proteins. Protein spots displaying differential abundance between the two groups were identified with MS. A number of candidate biomarkers were identified in 18 gel features. Selected candidates were quantified in a larger clinical set using ELISA. The mean levels of α1-antichymotrypsin (ACT), antithrombin III (ATIII), and zinc-α2-glycoprotein (ZAG) were significantly higher in the mild AD group, and the mean level of carnosinase 1 (CNDP1) was decreased. When these biomarkers are optimally combined, there is a strong trend toward greater specificity and sensitivity based on clinical diagnosis than when used individually. Our findings provide novel biomarker candidates for very mild and mild AD that can be further assessed as antecedent markers and predictors of clinical progression.     

An optimum method designed for 2-D DIGE analysis of human arterial intima and media layers isolated by laser microdissection

The formation and progression of atherosclerotic lesions involve complex mechanisms which are still not fully understood. A variety of cell types from the distinct arterial layers are implicated in the whole process from lipid accumulation within the vascular wall to plaque development and final rupture. In the present work, we employ the combination of laser microdissection and pressure catapulting and 2-D DIGE saturation labeling to investigate the human intima and media sub-proteomes isolated from atherosclerotic (coronary and aorta) or non-atherosclerotic vessels (preatherosclerotic coronary arteries). Laser microdissection and pressure catapulting allows the specific isolation of regions of interest. In turn, DIGE saturation labeling overcomes the limitation of extensive microdissection times to recover the protein amount required to perform comparative 2-DE, particularly when dealing with tissue regions rich in myofilament proteins, which result in low protein recovery. The compatibility and optimum performance of both techniques were investigated in detail, paying special attention to tissue staining and protein solubilization. Since scarce amount of protein obtained from microdissected tissue made it impossible to directly perform protein identification from 2-DE spots by MS, we performed in-solution digestion followed by LC-MS/MS analysis of total protein extracts from intima and media in order to get an overall picture of protein composition. Proteins so identified confirm the nature of the isolated regions. Finally, similar spot resolution on 2-D DIGE gels was obtained for the different human artery types (coronary, aorta) and studied layers (intima, media), setting the basis for future clinical comparative studies.     

Analysis of the human testis proteome by mass spectrometry and bioinformatics

The testis is a unique organ responsible for sperm production and androgen secretion in men. To analyze the human testis proteome on a large scale, 1-D SDS-PAGE and RP-LC-MS/MS were applied and 1430 proteins in the human testis were identified. Both the false-positive rate of peptides and protein identification confidence scores were calculated in the present study. Subsequent bioinformatics analysis of the human testis proteome revealed 39 testis-specific proteins which may be important for testis functions. And a large family of proteins were identified possibly involved in alternative splicing, which may also be involved in testis-specific splicing events and explain why splicing is so prevalent in the testis. Compared with the studies on brain proteome, researches on the testis proteome is still very limited. Studies of these proteins will give a better understanding on the function of the testis. Moreover, this large-scale identification of testis proteins in humans could serve as a reference for future studies on the mechanisms underlying male infertility, searching for potential contraceptive targets, and developing new treatments for testis cancer.     

Considerations for powering a clinical proteomics study: Normal variability in the human plasma proteome

Proteomics is increasingly being applied to the human plasma proteome to identify biomarkers of disease for use in non‐invasive assays. 2‐D DIGE, simultaneously analysing thousands of protein spots quantitatively and maintaining protein isoform information, is one technique adopted. Sufficient numbers of samples must be analysed to achieve statistical power; however, few reported studies have analysed inherent variability in the plasma proteome by 2‐D DIGE to allow power calculations. This study analysed plasma from 60 healthy volunteers by 2‐D DIGE. Two samples were taken, 7 days apart, allowing estimation of sensitivity of detection of differences in spot intensity between two groups using either a longitudinal (paired) or non‐paired design. Parameters for differences were: two‐fold normalised volume change, α of 0.05 and power of 0.8. Using groups of 20 samples, alterations in 1742 spots could be detected with longitudinal sampling, and in 1206 between non‐paired groups. Interbatch gel variability was small relative to the detection parameters, indicating robustness and reproducibility of 2‐D DIGE for analysing large sample sets. In summary, 20 samples can allow detection of a large number of proteomic alterations by 2‐D DIGE in human plasma, the sensitivity of detecting differences was greatly improved by longitudinal sampling and the technology was robust across batches.     

Proteomic analysis in cerebrospinal fluid of patients with atypical nonketotic hyperglycinemia and pulmonary hypertension - A pilot study

A variant phenotype of nonketotic hyperglycinemia has been described by our group associated with pulmonary hypertension. The aim of this study is to investigate the cerebrospinal fluid proteomes to get an insight into this neurodegenerative process producing leukoencephalopathy with white matter spongiform degeneration. DIGE and MALDI-TOF-TOF analyses were performed to carry out the proteomic study of four patients against three normal controls and one additional control of a classical nonketotic hyperglycinemia. The differential proteomic analysis showed a displacement of some series of spots toward the acidic side. The shifted proteins showed a high degree of carbonylation and increased methionine sulfoxidation was found in cystatin C and in vitamin-D-binding protein. These findings in addition to the increase of serum malondialdehyde concentration provide evidence of an oxidative stress in the patients under study, which is probably systemic rather than mainly confined to the CNS. The similarities of our findings with those found in other neurodegenerative diseases suggest that oxidative damage is commonly involved in these pathologies. DIGE technology improves the 2-D PAGE differential analysis and it is suitable in proteomic studies with a small number of cases.     

Analysis of the differential proteome of human erythroblasts during in vitro erythropoiesis by 2-D DIGE

In the present study we have used an in vitro culture system that induces differentiation of human CD34(+) cells down the erythroid lineage along with 2-D DIGE to determine the differential proteome of erythroblasts at specific developmental stages during erythropoiesis. We initially distinguished 154 proteins differentially expressed between pro-normoblasts and polychromatic/orthochromatic erythroblasts, of which 24 protein spots, representing 21 different proteins, were identified following MS/MS and verification in replicate experiments with cells from different individuals. These data were confirmed by analysis of the differential proteome of erythroblasts at more discrete stages of erythropoiesis using 2-D DIGE and by mapping the expression of three identified proteins (Annexin I, Annexin II, Carbonic Anhydrase I) throughout erythropoiesis by Western blot with specific antisera. In addition, the differential expression of proteins due to biological variation, such as polymorphism, was determined by comparing erythroblasts at the same developmental stage from different individuals; none of the proteins thus identified were represented in the above data set. Finally, we discuss the problems associated with 2-D DIGE gel-based proteomic approaches such as ours and suggest a modified approach for decreased inter-gel variation, improved protein resolution and increased protein concentration, which should significantly facilitate protein identification.     

Proteomics in clinical prostate research

The incidence of early prostate cancer (PCa) has increased rapidly in recent years. The majority of newly diagnosed PCa are in early tumor phase. Presently, we do not have adequate biomarkers to assess tumor aggressiveness in individual cases. Consequently, too many patients are given curatively intended treatment. An exploration of the human proteome may provide clinically useful markers. 2-DE has been successfully used for analysis of the protein phenotype using clinical samples. Proteins are separated according to size and charge, gels are compared by image analysis, protein spots of interest are excised, and proteins identified by MS. This method is exploratory and allows protein identification. However, low-abundance proteins are difficult to detect and 2-DE is currently too labor-intensive for routine use. In recent years, nongel based techniques, such as LC-MS, SELDI-MS, and protein arrays have emerged. They require smaller sample sizes and can be more automated than 2-DE. In this review, we describe studies of the protein expression of benign prostatic tissue and PCa, which is likely to serve as the first step in prognostic biomarker discovery. The prostate proteome is still far from a complete mapping which would enhance our understanding of PCa biology.     

Proteome analysis of human foetal, aged and advanced nuclear cataract lenses

The most complete proteome of human lenses has been compiled using 2-D LC-MS/MS analysis of foetal, aged normal and advanced nuclear cataract lenses. A total of 231 proteins were identified across all lens groups, including 112 proteins that have not been reported previously. Proteins were grouped according to their PANTHER molecular function classification in order to facilitate comparisons. Previously unreported N-terminal acetylation was detected in a number of proteins, with the majority being associated with the prior removal of a methionine residue. This pattern of proteolysis may indicate that methionine aminopeptidase activity is present in human lenses. Acetylation is likely to aid in the stability of proteins that are present in the lens for many decades. Protein sequences were also used to interrogate the three human lens cDNA libraries publicly available. Surprisingly, 84 proteins we identified were not present in the cDNA libraries.     

Identification of differentially expressed proteins in papillary thyroid carcinomas with V600E mutation of BRAF

BRAF, a serine/threonine kinase of the RAF family, is a downstream transducer of the RAS-regulated MAPK pathway. V600E mutation of BRAF protein is the most common genetic alteration occurring in papillary thyroid carcinomas and is prognostic of poor clinicopathological outcomes. Protein expression in the subclass of PTC bearing the BRAF(V600E) mutation was investigated by using 2-DE and MS/MS techniques and compared to that of matched normal thyroid tissues from seven patients. 2-D gel image analysis revealed that the expression of eight polypeptide spots, corresponding to five proteins, were significantly underexpressed in PTC bearing BRAF(V600E) mutation whereas 25 polypeptides, representing 19 distinct proteins, were significantly upregulated in tumour tissue, as compared to normal thyroid. Among the differentially expressed polypeptides, mitochondrial proteins, ROS-scavenger enzymes, apoptosis-related proteins as well as proteins involved in tumour cell proliferation were identified. Although dissimilarities between the present results and those previously reported can be ascribed to the use of different 2-DE techniques, the possibility that BRAF(V600E) mutation is responsible for changes in protein expression distinct from those induced by other oncogenes cannot be ruled out.