Bioinformatics project: PEG1/MEST

Protfun is a webbased service that tries to predict functions from sequence. According to the developers the results are, if any are given, correct to 90%. However, it a new and rather “experimental” method. Here is the complete list with the most likely alternatives in bold.

Functional category                             Prob Odds
Amino_acid_biosynthesis                     0.113 5.155
Biosynthesis_of_cofactors                     0.206 2.865
Cell_envelope                                         0.043 0.710
Cellular_processes                                 0.075 1.031
Central_intermediary_metabolism     0.109 1.725
Energy_metabolism                           0.381 4.234
Fatty_acid_metabolism                         0.066 5.066
Purines_and_pyrimidines                     0.232 0.953
Regulatory_functions                             0.018 0.113
Replication_and_transcription             0.025 0.093
Translation                                             0.105 2.378
Transport_and_binding                         0.079 0.192

# Enzyme/nonenzyme
Enzyme                                                     0.745 2.600
Nonenzyme                                             0.255 0.358

# Enzyme class
Oxidoreductase                                      0.208 1.001
Transferase                                             0.268 0.778
Hydrolase                                                 0.138 0.435
Lyase                                                         0.020 0.430
Isomerase                                                 0.015 0.475
Ligase                                                         0.044 0.869

# Gene Ontology category
Signal_transducer                                 0.078 0.367
Receptor                                                 0.107 0.631
Hormone                                                 0.001 0.206
Structural_protein                                 0.015 0.539
Transporter                                         0.117 1.069
Ion_channel                                             0.009 0.154
Voltage-gated_ion_channel                  0.004 0.172
Cation_channel                                       0.012 0.253
Transcription                                          0.033 0.260
Transcription_regulation                     0.023 0.188
Stress_response                                     0.045 0.514
Immune_response                                 0.027 0.322
Growth_factor                                         0.016 1.138
Metal_ion_transport                             0.009 0.020

While reasearching the imprinting of peg1 further, I stumbled over some new databases. To further confirm the actual existance of a CpG island in the PEG1-gene, a CpGPlot/CpGReport/Isochore search was performed with EMBOSS. Results were displayed as plots shown below and confirms a CpG island of more than 600 bp located upstream from intron 1.

Another database that was used, was The imprinted gene catalogue search engine. From this search engine, more information about various genetic disorders linked to imprinting can be obtained. I used it for finding some information and references to Prader-Willi and Angelman syndrome.

The last database I stumbled upon, is not really a search database.  At the Mammalian Genetics Unit Harwell genetic mouse models for human diseases can be observed. They have an imformative section about imprinting and chromosome mapping.

By using SWISS-MODEL with the templates 1G42 and 1S8O, both with a sequence similarity around 20 %, the following model was obtained. The red structure is the modelled. As seen it is not very good looking, only a small part was successfully modeled. Since these two templat had the highest sequence similarity homology modeling might not be possible

Below is the amino acids, close to the N-terminal, we were able to model

Submitting the sequence to PROSITE (looking for sequences for known motifs)  gives one result: GLQNRRINLLSHDYGDIVAQELLYRYKQNRSGRLTIK = Pumilio RNA-binding repeat and homology domain profiles. This is what they have to say about the motif:

“Members of the Pumilio family of proteins (Puf) regulate translation and mRNA stability in a wide variety of eukaryotic organisms including mammals, flies, worms, slime mold, and yeast. Pumilio family members are characterized by the presence of eight tandem copies of an imperfectly repeated 36 amino acids sequence motif, the Pumilio repeat, surrounded by a short N- and C-terminal conserved region.”

Since they mention a 8 tandem repeat of the 36 amino acid sequence above, is this result plaussible? 

According to this article the loss of PEG1/MEST results in impared placentophagy in mice.

I quote wikipedia: “Placentophagy is the act of mammals eating the placenta of their young after childbirth”

Some other results are: embryonic growth retardation, reduced survival rate and abnormal maternal behaviour. This did not help

Performed a ClustalW alignment of the MEST protein and our best homolog 1S08 and compared it to the secondary structure found in PBD. However, the compared structures seem to differ too much in vital parts of the structure (e. g. gap found in MEST equals a helix in homolog). Sequence identity was established to be approx. 18 %. Unfortunatly this makes it impossible to perform homology modelling for the MEST protein with this sturcture as a template.

The PEG1/MEST protein found in mice is very conserved. If you run the mouse protein in PSORT and SignalP the results are similar, the same potential signal peptide is found in mice.

Submiting the sequece to PSORT gives some interesting results. It first mentions a signal-peptide between residue 26 and 27. A transmembral region is found between reside 15 and 31. It is predicted to be a cytoplasmic and also a mitochondrial protein. The later is most likely not true. However, I have not graded the scores given for each results, something for later.

SignalP  gives three values, C, Y and S-score. The results shows a signal-peptide between residue 26 and 27 (LLA-AY). The same results as for PSORT.

The E-AFMX-5 experiment with a transcription profile for different human tissues shows some interesting results. The MEST-gene is expressed at low levels in most of the tissues but alot more in: Cilliary ganglion, fetal lung, placenta (highest), trigeminal ganglion. The statistic accuracy of the experiment is not very good, many tissues only got samples from two patients.

In the E-TABM-40 experiment different immunocells have ben profiled. The MEST gene is expressenin all of them, the highest level is found in naive CD8-cells

In the E-MEXP-18 experiment several celllines and tissues are profiled. The MEST-expression is highest in CD34-cells, both hemapoetical and developed versions, dorsal root ganglion, placenta (highest) and umbillical endothelial cells.

Article one states that the gene codes for an enzyme that shows significant similarities to the alpha/beta hydrolase fold family though function is unknown. Gene is imprinted and expression was confirmed to the paternal derived allele. However, in blood lymphocytes biallelic expression was shown. Different expression patterns in different tissues. The human peg1 gene spans a genomic region of aprox 13 kb. A 620 bp CpG-island was found in the 5-prime region, the island extends from the plausible promotor to the first intron. Regulation is thought to be achieved through metylation.

Article two: Focuses on PEG1 isoform 2, which was thought to be biallically expressed (non-imprinted) in ceveral non-placental organs producing a shorter version of the MEST protein. Knockout mice show that the gene in mouse regulates placental and fetal growth (causes growth restrictions). The article also shows different types of imprinting in human placenta, both DNA methylation and allele-specific mRNA expression.

Article three is about a similar protein found in zeebrafish. It discusses the linkage with another gene, copg2 and its occurance in larvae.

Article four states that the human chromosomal region were peg1/mest is expressed is 7q21-qtr and human maternal uniparental disonomy (upd 7). It causes apparent growth deficiencies and slight morphological abnormalities.