MSCI814 Module 2.

Figure showing a very short micro exon GHE from several P450s

Phanerochaete chrysosporium (white rot fungus) Scaffold_388a very similar 
to sequences 77 417 112 129
gene model complete all boundaries checked, 16 exons
3583 MISDTFALAISSGLSLFLCLKAFIDYRAGLRSI (2) 3684 ex1
3732 NHSYLPGFRALISSFGILGLFFKEPKRGLWGGRRRFWLRKHLDFEEAGVDIISH (0) 3893 ex2
3954 IAFLPSVSTYLLLADAAAIK (0) 4013 ex3
4069 EVTGHRARFPKPTYKTLRIFGGNVLASEGEEWKRHRKVVGPAFSE (0) 4203 c-helix ex4
4255 HNNRLVWNETVKIVNDLFANVWGSQSEVYVDNVVQSVTLP (0) 4374 ex5
4423 MALYVISIAGFGKRALWQADGNLPPGHKLSFQ (0) 4521 ex6
4576 DALHILGTDLWIKAATPTLLMNWAPTTRIANVKLAFDEVK (0) 4692 ex7
4747 QYMLELIQERRNSEKRDERYDLFSSLLDANDLNEDGNGNVTLTNDELL (1) 4890 ex9
     GNIFIFMLA (1) 4973 ex9 split
     GHE (0) ex9 split
5087 TTAHTLAFTFGLLALHPDYQETVYQQIKSIVPDNRPP (0) 5197 ex10
     MYEEMNSLTECMA (2) ex11
5351 YETLRLFPP (0) 5380 ex12
5436 TATIPKIAAEDTYLVTIDRAGNRVVVPVPCGTALHLNVIALHHN (1) 5564 ex13
5614 PRYWDNPSAFKPERFRGDWPRDAFIPFSTGSRSCIGRR (2) 5730 ex14
5780 FFETESIAILTMILSRYKIELRNDPRFADETYEERWQRVLRVKDGLTPA* 5932 ex15

compare to scaffold 388b not a separate exon

      GNIFIFLLAGHE(0) 14307 ex9
14247 TTAHTLAFTFGLLALYPEQQDKLYKHIKHVIPDGRIP (0) 14137 ex10

and scaffold 129

      GNIFIFMLA (1) ex9 split
      GHE (0) ex9 split
25150 TTAHTLAFTFGLLALHSDYQEKVHQQIKSIMPDNRLP (0) 25260 ex10

and scaffold 12a not a separate exon (exons have been fused)

54432 VKANMTEDAKSRLSEEEMYAEMR (2) 
54552 TILFAGHETTSTTISWVLLE

Expert knowledge reveals GHE to be a real exon, by comparison with other P450s. This happens to be in a motif region that is conserved (AGXETT) and easily recognized, but automated gene finding programs are not going to see this. Please note that the intron exon phases are indicated in (). The number 0, 1 or 2 tells the phase of the junction. Exons that join between codons are phase 0, those that join one base into a codon are phase 1 and those that join two bases in are phase 2. You cannot join exons together in frame unless the phase is preserved at both ends of the intron. In the example above, the sequence GNIFIFMLA ends in a phase 1 boundary, while the sequence TTAHTLAFTFGLLA starts with a phase 0 boundary. The two cannot be joined unless there is an exon in between with a phase 1 start and a phase 0 end. That is the GHE exon. This short eight base pair exon is seen in six different genes in the white rot fungus. These small exons have been called micro exons.

Introns begin with GT and end with AG. Some rare introns begin with GC. These bases, and all the bases between them, are cut out when the full length transcript is processed in the nucleus to make the mature mRNA (messenger RNA). Below is an example of a Drosophila P450 gene Cyp4e2, showing the 5 introns in place. Please look at each intron and find the GT and AG pairs. Also notice the phase of each intron as described above. This figure is the BLAT search output of the UCSC genome browser.
The exons are in blue capital letters. Introns are lower case black letters. The Capitals are all in blocks of three per codon. This allows determining the phase of intron boundaries, since the last capital letter is at the end of a complete codon. If gt follows then the phase is 0. If ngt ( n is any base) follows then the phase is 1. If nngt follows the phase is 2. In this figure intron 1 and intron 5 are phase 1, though intron 5 has the correct boundary off by one codon. Introns 2 and 4 are phase 0. Intron 3 is phase 2. The numbering on the right is nucleotide numbering for Chromosome 2R.

In this module, you will learn how to identify intron-exon boundaries, by comparing two similar genes, one from human, and the other from platypus. This will be fairly easy since the genes are similar. In real life situations, this is not going to be so clear. We have all 57 P450 genes identified from human, but we are starting with none identified from platypus. The human P450 sequences are linked below.

We are very fortunate to have the UCSC genome browser, which has an alignment of the human and platypus genomes as of March 2007. By doing a BLAT search (Slightly different than a BLAST search, optimized for near exact matches), with a human P450, we will get the region of the platypus genome with the best match to our query sequence. Then we can link to a view of the genomic DNA sequence showing the location of the protein coding exons over the region of our gene. From this figure and our original protein sequence we can identify the GT and AG intron-exon boundaries. The introns can be edited out, and the assembled gene can be translated to give the platypus protein sequence.

The step by step procedure is now given:

• Link to the Human P450s for this assignment and copy your protein sequence.
• Paste your protein sequence in a Word Document for later use. You will be identifying the intron boundaries and phases in this sequence.
• Link to the UCSC bioinformatics server click on BLAT and then select platypus under the genome pull down menu.
• Paste your P450 sequence in the window and click the submit file button. You will see a display like the one below. This example is human CYP2A13 compared to rhesus monkey. Notice under the identity column the top hit is 93.6%. That is your gene. The others are different P450 sequences. The BLAT search only returns strong matches. Notice there are no hits less than 61%. The gene is on Scaffold 111671 on the minus strand at the nucleotide location given. The +- signs indicate your query sequence is always treated as Plus Strand and the match is Minus Strand. ++ would mean the hit is in the Plus Strand.
  
  This example was made with the bottom part of CYP2A13. The whole 2A13 sequence caused an error, so I shortened it. I think the error may have happened because this sequence is at the end of a scaffold and the whole sequence is not on the scaffold. Notice START END and QSIZE. This is the amino acid position of your query that matched. QSIZE is the query length. Here the query was 218 aa long and 2-218 matched.
  
  
• Click on details on the left side. This gives the image of the gene structure. This gene fragment has 4 exons, but only two are shown in the first image. The top of the figure has your query sequence color-coded to show matches, mismatches and gaps (introns).
  
  
  
  The second image shows all four exons.
  
  
  
  The bottom of the details page shows the protein translation of each exon. Conservative changes are in green, other changes are in red. Top line is the query sequence with the quesry amino acid numbering. The bottom line is the translated nucleotide sequence with nucleotide numbering.
  
  
  
  
• Go back and click on browser. This takes you to the browser page showing the gene from one end to the other (exons 6-9 in this case). Multiple levels of evidence are given on the page including your query sequence at the top. Genbank RefSeq genes and other curated and non-curated predictions and sequence matches, including ESTs. The wide bars in the image are exons, or predicted exons. The connecting line covers the region from the beginning of your query sequence to the end. The arrows on the connecting line show the orientation of the gene. This gene piece has 4 exons and it is on the minus strand.
  
  
• Click on the purple bar (color indicates chromosome, see code at the bottom of the graphic). This links you to both the human and rhesus nucleotide sequences in the region of the browser window. Farther down on this page is an alignment of the two sequences. You can also open the human browser at the equivalent spot to see what is there. This can be very helpful when dealing with gene clusters.
  
  
• Go back to details and examine the sequence showing the exons in blue and the introns in lower case black. It is time to identify the GT and AG boundaries of the introns. Keep in mind that these figures are drawn with intact codons in Blue. The codons are never split. This means that you can tell the phase if you can identify the GT that is the beginning of the intron. It should be the first GT in the black lower case region. This may be one or two codons away, but usually it is in the first codon in the black. The GT may also be in the Blue region, or the last letter of the Blue region may be G with the T in the black. The rules for determining the intron phase are:
  
  Phase 0. G and T are the first two letters of the black sequence, 1,2 or or they are 4,5 or 7,8. The intron is Phase 0 and it occurs between codons. Phase 0 codons always code for Val, since all valine codons begin with GT.
  
  GATCGTTgtagtca
  
  Phase 1. G and T are at positions 2,3 or 5,6 or 8,9 in the black sequence. The most common phase one codon is GGT = glycine. Phase one boundaries break the codon one nucleotide in.
  
  GATCAGTggtcgtaa
  
  Phase 2. G and T may overlap the blue and black sequence -1,1, or 3,4 or 6,7 in the black sequence. Arginine is often a phase 2 codon AGGT Phase two boundaries break the codon two nucleotides in.
  
  GATCAGGtcgtagtaa
  
  The other end of the intron will have an AG before the coding sequence begins again.
  
  Phase 0. AG may be the last two letters of the black intron sequence. They may be multiples of three from that boundary. Frequently CAG = gln is at the phase 0 boundary.
  
  gagcagAAGTACGAA
  
  Phase 1. The AG can span the black and blue boundary with A in the black and G in the blue. They can also be multiples of three from that location.
  
  taagaaGCAGACGAA
  
  Phase 2. AG can be one letter away from the blue boundary as AGX. The AG can also be the first two letters in the blue boundary.
  
  gatagtAGAAGCGAA
  
  
• Copy each exon sequence upto but not including the GT of the intron. Mark each intron boundary on the end of an exon with the correct phase in parentheses (0), (1) or (2). The phase at both ends of the intron has to be the same so you only need to mark one end. Start copying the next exon after the AG boundary. Repeat until you have all exons copied and marked for phase

  
  The CYP2A13 sequence (4 exons, introns removed)
  GAGGAGAAGA ACCCCAACAC GGAGTTCTAC TTGAAGAACC TGatgATGAC CACGCTGAAC  14520
  CTCTTCattG CAGGCACCGA GACCGTCAGC ACCACCCTGC GCTATGGCTT CCTGCTGCTC  14460
  ATGAAGtatC CAGAGGTGGA Gg (1)
  CCAAG GTCCATGAGG AGATTGACAG AGTGATCGGC AAGAACCGGC AGCCCAAGTT  13920
  TGAGGACCGG gtcAAGATGC CCTACatgGA GGCAGTGATC CATGAGATCC AAAGATTTGG  13860
  AGACgtgatc CCCATGagcT TGGCCcgcAG GGTCAACAAG GACACCAAGT TTCGGGATTT  13800
  CTTCCTCCCT AAG (0)
  GGCACC  13260
  GAAGTGTTCC CTATGCTGGG CTCCgtgCTG AGAGACCCCA GGTTCTTCTC CAACCCCCAG  13200
  GACttcaatC CCCAGCACTT CTTGGATGAG AAGGGGCAGT TTAAGAAGAG TGACGCTTTT  13140
  GTGCCCTTTT CCATCg (1)
  GAAAG CGGaacTGTT TCGGAGAAGG CCTGGCCAGA  12120
  ATGGAGCTCT TTCTCTTCTT CACCACCATC ATGCAGAACT TCCGCTTCAA GTCCCCCCAG  12060
  ttgCCCAAGG ACATCGACGT GTCCCCCAAA CACGTGGGCT TTGCCACGAT CCCAccaAAC  12000
  TACACCATGA GCTTCCTGCC CCGC 
  

• take your complete coding sequence with all introns removed to the Javascript DNA translator and translate your sequence. Sometimes there are errors made in cutting out the intron sequences. This can make a frameshift and introduce stop codons into your translation, If you see stop codons then look for the location of the frameshift by blast searching your three frame translations against Human at the human P450 blast server page. The place where your sequence stops matching the human sequence is where the frameshift is. This will probably be at one of the intron boundaries where you made an error in removing the intron.
• Paste your newly assembled platypus protein sequence below the human starting sequence in your word file. Mark the location and phase of the introns and do a blast search of human to get the sequence alignment as shown below. Email me these three items.

  
  >human protein CYP2A13 last 4 exons
  EEEKNPNTEFYLKNLVMTTLNLFFAGTETVSTTLRYGFLLLMKHPEVE
  AKVHEEIDRVIGKNRQPKFEDRAKMPYTEAVIHEIQRFGDMLPMGLAHRVNKDTKFRDFFLPK
  GTEVFPMLGSELRDPRFFSNPQDCSPQHFLDEKGQFKKSDAFVPFSI
  GKRYCFGEGLARMELFLFFTTIMQNFRFKSPQSPKDIDVSPKHVGFATIPRNYTMSFLPR
  
  >Rhesus match (yours will be platypus)
  EEKNPNTEFYLKNLMMTTLNLFIAGTETVSTTLRYGFLLLMKYPEVE (1)
  AKVHEEIDRVIGKNRQPKFEDRVKMPYMEAVIHEIQRFGDVIPMSLARRVNKDTKFRDFFLPK (0)
  GTEVFPMLGSVLRDPRFFSNPQDFNPQHFLDEKGQFKKSDAFVPFSI (1)
  GKRNCFGEGLARMELFLFFTTIMQNFRFKSPQLPKDIDVSPKHVGFATIPPNYTMSFLPR*
  
  Alignment of the two sequences 93% identical
  Query: 2   EEKNPNTEFYLKNLVMTTLNLFFAGTETVSTTLRYGFLLLMKHPEVEAKVHEEIDRVIGK 61
             EEKNPNTEFYLKNL+MTTLNLF AGTETVSTTLRYGFLLLMK+PEVEAKVHEEIDRVIGK
  Sbjct: 1   EEKNPNTEFYLKNLMMTTLNLFIAGTETVSTTLRYGFLLLMKYPEVEAKVHEEIDRVIGK 60
  
  Query: 62  NRQPKFEDRAKMPYTEAVIHEIQRFGDMLPMGLAHRVNKDTKFRDFFLPKGTEVFPMLGS 121
             NRQPKFEDR KMPY EAVIHEIQRFGD++PM LA RVNKDTKFRDFFLPKGTEVFPMLGS
  Sbjct: 61  NRQPKFEDRVKMPYMEAVIHEIQRFGDVIPMSLARRVNKDTKFRDFFLPKGTEVFPMLGS 120
  
  Query: 122 ELRDPRFFSNPQDCSPQHFLDEKGQFKKSDAFVPFSIGKRYCFGEGLARMELFLFFTTIM 181
              LRDPRFFSNPQD +PQHFLDEKGQFKKSDAFVPFSIGKR CFGEGLARMELFLFFTTIM
  Sbjct: 121 VLRDPRFFSNPQDFNPQHFLDEKGQFKKSDAFVPFSIGKRNCFGEGLARMELFLFFTTIM 180
  
  Query: 182 QNFRFKSPQSPKDIDVSPKHVGFATIPRNYTMSFLPR 218
             QNFRFKSPQ PKDIDVSPKHVGFATIP NYTMSFLPR
  Sbjct: 181 QNFRFKSPQLPKDIDVSPKHVGFATIPPNYTMSFLPR 217
  
  
  

  The genes in platypus will not be as close to human as the rhesus monkey sequences, so this will be a little harder than doing a primate.
  
  
  
  Assignment 2.
  
  Do the gene assembly as shown above for a platypus P450. Not all genes will show complete coverage in the Genome browser. There may be gaps in the sequence. If an exon is missing just make a note saying an exon was missing here. The BLAT search at the UCSC browser is geared for near exact matches. Short or weak matches do not show up. You may be able to find missing exons by blastx searching the genomic DNA against human P450s in the P450 blast server. Look at your seq at the top of the details page to see what is missing. Any long region of black text was not found by the BLAT program.

  
  F. Chowdhury             CYP51  
  M. Dunagan               CYP26C1   
  A. Ellebedy              CYP26B1  
  B. Hale                  CYP19
  Z. Liu                   CYP7A1
  R. Mahato                CYP2U1
  J. Philippart            CYP2R1  
  M. Richardson            CYP2A13  
  R. Tsukahara             CYP1A1
  A. Udyavar               CYP51  
  Q. Zhang                 CYP26C1 
  Y. Qu                    CYP26B1 
  
  
  
  
  Human sequences for assignment 2 Intron exon boundaries and phases are marked