Proteomic Profiling and Data - Assignment Example

PROTEOMIC PROFILING Proteomics is a field in life science research that is used to identify and characterize all the proteins (proteome) in a cell, tissue, or organism. It makes use of high throughput technologies like two-dimensional gel electrophoresis (2D-GE) and mass spectrometry to come up with what is known as a peptide mass fingerprint. In peptide mass fingerprinting (PMF), experimental data on the peptide molecular mass is compared against compared with the theoretical data that are deposited in protein data banks like the NCBInr and SwissProt databases. This is possible with the use of different software like MASCOT PMF, ProFound, and Peptident. In this exercise, sample was obtained from cells of cultured Rat Basophilic Leukemic (RBL 293) cells, which were grown in medium supplemented with fetal calf serum. The cells were lysed and prepared for 2D-PAGE. The first dimension involved the separation of the proteins by IEF on a linear pH 4-7 IPG strip (horizontal), followed by SDS-PAGE on a 10% acrylamide gel (vertical). The gel was silver-stained and imaged using a densitometer. The approximate mass, M and pI of the proteins can be seen on the image. The spots of interest, A-F, were excised from the gel, destained, reduced and alkylated with iodoacetamide (CAM-derivative). Iodoacetamide is added to covalently modify cysteine residues to form carbamidomethylcysteine to present the formation of disulfide bridges and to ensure that the proteins are completely unfolded. After this step, proteins are cleaved by trypsin, a proteolytic enzyme that cleaves lysine and arginine except when followed by proline toward the C-terminal direction, to form peptide fragments. The digests were analyzed by MALDI-ToF mass spectrometry by an instrument with mass error of 50-200 ppm. The spectra obtained was processed to generate the peptide mass fingerprints and the monoisotopic peak list. This list was submitted to www.matrixscience.com which has the online Mascot PMF software which searches the SwissProt or NCBInr databases for proteins with peptides of similar masses. The search parameters used were: Enzyme : Trypsin Fixed modification : Carbamidomethyl (C), because the use of iodoacetamide results in carbamidomethylation Variable modifications : Oxidation (M); but depending on the results obtained, this was changed to look for higher probability scores. Increasing number of modifications will lessen probability because of increased number of possible peptide masses. Mass values : Monoisotopic Missed cleavage : 1 (to allow for errors in trypsin cleavage). If cleavage is increased, there will be more peptides to search from resulting in less probability of getting the right peptide. Peptide tolerance : 50 ppm; this tolerance gives higher accuracy The probability-based MOWSE (molecular weight search) score that gives the probability that the observed match between experimental data and a protein sequence is a random event (Perkin, Pappin, Creasy, & Cottrell, 1999) was used in interpreting the results. SAMPLE A ACTB_RAT; Actin, cytoplasmic 1 OS=Rattus norvegicus, searched using SwissProt database Probability based MOWSE score (PBS)= 140; highly significant Matches were at 13 out of the 30 queries submitted Molecular mass was 42052; calculated pI=5.29 which corresponded with the values obtained from 2D-GE. Sequence coverage 45% It was observed that this protein had similar scores with actin, cytoplasmic 2. Since the molecular weights of the two proteins were highly similar, it will be necessary to perform other characterization to confirm which protein is present, or if both are present and were not separated by 2D-GE. Initial search parameters Taxonomy: Rattus Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Protein Mass : 45 kDa Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 30 Results when changes were made to initial search parameters No variable modification specified: the same results were obtained but the probability was increased to 160. This validates the initial results obtained and indicates that the denaturation and cleavage process was highly precise. Setting missed cleavage to 2 resulted in decreasing the protein sequence coverage from 45% to 39%. Changing the variable modification to acetyl (protein N-terminal), peptide tolerance (to 200 ppm), and using SwissProt for rodents produced 14 matches, and a probability score of 179. Changing the database searched to NCBInr resulted in the same PBS, but the top score went to the protein beta actin [Homo sapiens] which also had a nominal mass of 42052, similar to that of actin-1 from Rattus. This result could be due to the higher number of sequences that were found on the NCBInr database compared to SwissProt, where only the Rattus sequences were specified to be searched). Changing the variable modification to amidated protein at the C-terminal gave increased the number of matches to 14, increased score to 189, and increased the sequence coverage to 52%. When the same search was conducted using NCBInr, the score was increased to 209, the e value was 5.1, and there were 16 matches for the protein gamma actin of Rattus norvegicus. However, the molecular mass was 59163, which did not match the value on the gel. Certainly, gamma actin has high sequence similarity to the actin-1, but could have more sub-units or a larger structure. SAMPLE B ENOA_RAT, Alpha-enolase OS=Rattus norvegicus, searched using SwissProt Probability based MOWSE score (PBS)= 192; highly significant Matches were at 19 out of the 50 queries submitted Molecular mass was 47440 (a little off from that observed on the gel); calculated pI=6.16 which corresponded with the values obtained from 2D-GE. Sequence coverage 54% The next highest match was ENOB_RAT, which has a mass of 47326, which was similar to ENOA_RAT. The PBS for this protein was 61, which is considered significant, but the number of matched queries was only 8 out of 50. Therefore, the similarity is lower when compared to alpha enolase. Initial search parameters Taxonomy: Rattus Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 30 Results when changes were made to initial search parameters Changing the peptide tolerance to 100 ppm reduced the probability score to 189. Increased tolerance increases the number of peptides to be searched, which reduces the probability of matches. When the variable modification was changed to acetyl at the protein N-terminal, the probability based score increased to 209 for alpha enolase, with 18 matches out of the 50 queries submitted. This shows that acetylation at the N terminal could have occurred during the denaturing reduction process. Combined with increasing tolerance to 100 ppm reduced the PBS to 200, but the number of matches and sequence coverage of 54% were the same. When the variable modification was changed to acetyl K, and the NCBInr database searched, the probability score for alpha enolase was reduced to 149. Changing the variable modification to amidated protein at the C-terminal gave the same match for enolase A of Rattus norvegicus. The score was 209 with 18 matches. Increased tolerance to 100 ppm reduced the score to 195, and further increasing missed cleavages to 2 decreased the score to 168. The above results indicate that the molecular mass of Sample B does not correlate to the mass indicated on the gel. However, the results of the Mascot search using different parameters and the two databases came out with the same ID each time. Possibly, the difference in mass could be due to quirks in running the gel or some modification brought about by the denaturation and reducing processes. So apparently and statistically, the identified protein is correct. SAMPLE C HSP7C_RAT, Heat shock cognate 71 kDa protein, Rattus norvegicus; SwissProt database search Probability based MOWSE score (PBS)= 202 Matches were at 25 out of the 67 queries submitted Molecular mass was 71055; calculated pI=5.39 which corresponded with the values obtained from 2D-GE. Sequence coverage 44% The next highest match was SBNO1_RAT with a mass of 141280 and probability score of 47. This score is statistically insignificant. Thus, the most probable protein ID is HSP7C_RAT. Initial search parameters Taxonomy: Rattus Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Protein Mass : 75 kDa Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 67 Results when changes in search parameters were made: Increased tolerance of 100 ppm increased the score slightly to 204, increased matches to 26 and improved sequence coverage to 46%, e=2.9e-17 Changing the variable modification to acetyl K reduced the score to 147, with only 24 matches. Similar results were obtained from NCBInr using the above modifications. This indicates that this sequence is highly conserved in the rat, or there are no other proteins that have similar sequences deposited in the databases. Changing the fixed modification to acetyl K, and the variable modification to acetyl N-term, tolerance to 100 ppm, and using SwissProt gave the result sarcolemmal membrane associated protein with a score of 62 and mass at 101398, pI=5.14. This shows that, if the modifications entered in Mascot do not correspond to the actual modifications, then it is possible to get totally unrelated proteins. SAMPLE D TPM4_RAT Tropomyosin alpha-4 chain, Rattus norvegicus, SwissProt database search Probability based MOWSE score (PBS)= 140 Matches were at 25 out of the 71 queries submitted Molecular mass was 28549; calculated pI=4.66 which corresponded with the values obtained from 2D-GE. Sequence coverage 45% The next highest match was K2C73_RAT, keratin, type II cytoskeletal 73 with a mass of 60977, and an insignificant score of 45. Only 8 out of 71 queries matched. Thus, the most probable protein ID is TPM4_RAT. Initial search parameters Taxonomy: Rattus Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Protein Mass : 31 kDa Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 71 Results when changes were made to initial search parameters When the variable modification was changed to acetyl K, the score decreased to 120, but the coverage increased to 55% with only 20 matches. Changing the variable modification to amidated C reduced the score to 109, reduced the coverage to 53% and the matches to 18. Changing both fixed (to acetyl K) and variable (to amidated C) modifications decreased the scores to 104. When acetyl K fixed modification was paired with acetyl N variable modification, the score was reduced to 76, and only 14 matches were found. The same results were obtained when the mouse database (MSDB) was searched. The results that came from using the original search parameters were most indicative of the correct modifications. SAMPLE E Most probable ID: Serum albumin; Bos taurus; All entries from SwissProt database was searched Initial search parameters: These parameters were not able to successfully show the correct protein ID for this sample. Taxonomy: Rattus Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 43 Best match using the above parameters was Ras-related protein Rab-7a,with a mass of 23774, a third of the mass indicated on the 2D gel. The score was 30, with e=7, and 4 matches out of 43. Results when changes were made to initial search parameters Variable modification changed to oxidation (HW), and amidated did not improve the scores nor increased the matches significantly. Other proteins also came out like the EF HD2 protein when the HW oxidation was set to variable modification. To improve chances of finding a suitable match, all entries in the SwissProt database were searched using the original search parameters. The result was a very high score (261) for serum albumin from Bos taurus (cow). Twenty-five out of 43 entries matched, with e= 3.4e-21. This result was repeated even when other modifications were introduced, showing the strong statistical evidence that this was a positive ID for Sample E. Possible source of the serum albumin protein is the media where the cells were grown. Therefore, this protein is a contaminant that was not removed during the process of harvesting, washing, and lysing the cells. SAMPLE F Superoxide dismutase [Cu-Zn]; Rattus norvegicus; SwissProt database search. Probability based MOWSE score (PBS)= 55 (significant at P=0.05) Matches were at 6 out of the 56 queries submitted, which is a poor match. Molecular mass was 16073; calculated pI=5.88 which corresponded with the values obtained from 2D gel. Sequence coverage=38% The next highest match was protein LCHN with a mass of 52021 and probability score of 43 (not significant). Based on the initial search, the most probable protein ID is superoxide dismutase (SOD) (Cu-Zn). Initial search parameters Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Mass values : Monoisotopic Protein Mass : 18 kDa Peptide Mass Tolerance : ± 50 ppm Peptide Charge State : 1+ Max Missed Cleavages : 1 Number of queries : 56 Results when changes were made to initial search parameters Although the initial results gave a good match with respect to the molecular mass and pI values, the low number of queries that matched the identified protein showed the presence of possible contaminants in the sample. So, several modifications were made to the initial search parameters: Instead of using the Rattus specific search, all entries in SwissProt were included in the search, while retaining the initial search parameters. The highest match was for K1C10, Keratin, type I cytoskeletal 10, from Homo sapiens, with a significant probability score of 70, e=0.047 and 7 out of 56 queries matched. SOD (Cu-Zn) was the next best match, but the score of 55 was not statistically significant. Changing the fixed and variable modifications to acetyl (N-term) and acetyl K, respectively did not produce significant matches. When the NCBInr was searched by with changes in the variable modification to Phospho Y, and increased missed cleavages to 4, resulted in different proteins with significant scores and matches of 11-14 peptides. However, the molecular masses and classes of the proteins varied widely and came from different species. Significant number of contaminants was present based on low peptide matches and low probability-based scores. These contaminants could have been introduced after the gel was run, and when the samples were being processed for MALDI-ToF analysis. Reference cited Perkin, D, Pappin, D, Creasy, D, & Cottrell, J (1999), ‘Probability-based protein identification by searching sequence databases using mass spectrometry data’, Electrophoresis, vol 20, pp. 3551-3567. Read More