TP53 from prediction to crystal

JBrowse v5 beta

The tutorials target the JBrowse v5 beta, and the v4.3.0 release on the download page lacks some of what they show. To install the beta of JBrowse Web, run npm install -g @jbrowse/cli@next, then jbrowse create jbrowse2 --branch v5.0.0-beta.11. Desktop beta builds are coming soon.

The p53 protein has a predicted structure covering every residue and crystal structures covering the parts that fold. We open three of them in one view beside the TP53 gene, each mapped to the same transcript, click a cancer hotspot on a crystal to find its codon, and open an NMR ensemble of the transactivation domain. The protein3d plugin maps each structure to the transcript, and Mol* draws the structures.

Where the data comes from

Show full URLs

The links open a hosted hg38 config with NCBI RefSeq genes, and the protein3d plugin fetches the structures, annotations and mappings from these services.

hg38 with NCBI RefSeq: jbrowse.org/…/protein3d_config.jsonhttps://jbrowse.org/code/jb2/main/test_data/protein3d_config.json
the AlphaFold model of p53, UniProt P04637: alphafold.ebi.ac.uk/…/AF-P04637-F1-model_v6.cifhttps://alphafold.ebi.ac.uk/files/AF-P04637-F1-model_v6.cif
the p53 core domain bound to DNA, PDB 1TUP: files.rcsb.org/…/1TUP.cifhttps://files.rcsb.org/download/1TUP.cif
the p53 transactivation peptide bound to MDM2, PDB 1YCR: files.rcsb.org/…/1YCR.cifhttps://files.rcsb.org/download/1YCR.cif
the p53 transactivation domain bound to CBP, solved by NMR, PDB 2L14: files.rcsb.org/…/2L14.cifhttps://files.rcsb.org/download/2L14.cif
UniProt's feature annotation of p53, the domain and variant tracks: rest.uniprot.org/…/P04637.gffhttps://rest.uniprot.org/uniprotkb/P04637.gff
SIFTS, which maps where each crystal's residues sit in the UniProt sequence: www.ebi.ac.uk/…/1tuphttps://www.ebi.ac.uk/pdbe/api/mappings/uniprot/1tup

Opening three TP53 structures mapped to one transcript

An AlphaFold model is one chain, numbered like the UniProt sequence. A crystal structure often holds one domain of the protein, sometimes in several copies and with a partner or DNA, numbered from wherever the crystallised construct began. The plugin maps both kinds to the gene by aligning the structure's sequence to the transcript's translation.

Open the three structures of TP53. The link is a session spec naming the gene's locus, the RefSeq transcript NM_000546.6, and three structures: the AlphaFold model by UniProt accession, and PDB 1TUP and 1YCR. The plugin superposes the structures with TM-align. The Color menu is set to Mapped chain, which colours the chain the transcript encodes blue and everything else grey.

The arrow beside each structure in the header opens its alignment panel, with the transcript's translation on the GENOME row, the structure's sequence on the STRUCT row, and a ruler in the authors' residue numbering. The AlphaFold panel is one unbroken match. The 1TUP panel covers the DNA-binding core alone, and its ruler starts at 94, the residue where the crystallised construct begins. Under the AlphaFold STRUCT row, pLDDT, AlphaFold's per-residue confidence, is high across the core and falls away at both ends, the same tails both crystals leave out.

UniProt annotations on the 1TUP core-domain crystal

The feature tracks under each panel come from UniProt, numbered along the full-length protein. For 1TUP the plugin looks up where the construct starts in SIFTS and shifts every feature by that offset, dropping features outside the construct. Show all feature tracks in the display settings menu adds secondary structure, modified residues and other minor types to the default domains, binding sites and variants.

1TUP holds three copies of the core and two DNA strands. The panel's Mapped chain picker lists them, and the plugin chose the protein entity because its sequence aligns to the transcript. Hovering any of the three copies in the 3D canvas highlights the same residue on the genome.

Finding the codon of the R248 hotspot from the crystal

Open the 1TUP panel and click the R248 hotspot on its STRUCT row, or open the same session with R248 selected. The ruler and the transcript row give the same number, because the 1TUP authors numbered their construct the way UniProt numbers the whole protein.

NCBI RefSeq at TP53's R248 codon beside 1TUP with R248 selected. The crystal's three copies of p53's core are blue on grey DNA, R248 is magenta on the copies in view, with its neighbours drawn as sticks, and a band marks its codon on the gene.

On one copy, the side chain of R248 reaches into the minor groove of the DNA. Hovering the codon on the gene highlights the residue on the structures that contain it; hovering the intron beside the exon highlights nothing.

Choosing the mapped chain in the MDM2 complex 1YCR

1YCR holds the N-terminal domain of MDM2, a p53 regulator, and a fifteen-residue peptide from the p53 transactivation region, and the transcript encodes the peptide. Open the 1YCR panel: its Mapped chain picker lists both chains, and the plugin picked the peptide, a short exact match near the start of the transcript row.

1YCR with its Mapped chain picker open. Chain B, the p53 peptide, is the mapped one, blue against grey MDM2; Chain A above it in the picker is MDM2.

Switch the picker to Chain A. The plugin realigns against MDM2, which turns blue, and the GENOME row becomes a scatter of gapped fragments, because the transcript does not encode MDM2. In a complex of two different proteins the automatic choice can land on the wrong chain, and the picker corrects it.

The p53 transactivation domain as an NMR ensemble

PDB 2L14 holds residues 13 to 61 of the p53 transactivation domain, the region where the AlphaFold pLDDT falls away, bound to the nuclear coactivator binding domain of CBP. Its authors deposited an NMR ensemble of twenty models that each fit the data.

Open 2L14 with its two helices selected. The link selects the two helices the file annotates, residues 18 to 26 and 46 to 54, through initialTranscriptResidues, which counts along the transcript's translation. The plugin loads each model as a separate structure, so the view opens more slowly than a crystal's. Click Reset Zoom, the circular arrow at the top right of the canvas, to fit the whole ensemble.

NCBI RefSeq over TP53's first coding exons beside every model of 2L14. The p53 chain is blue on grey CBP with its two helices in magenta; the helices overlap from model to model, the linker leaving the first one spreads, and bands on the gene mark the helices' codons.

The two helices sit on CBP in the same place in every model. The linker after the first helix and both ends of the chain differ from model to model, and the grey spray on the right is the C-terminal tail of CBP. The first helix lies within the stretch the 1YCR peptide covers. The first helix's codons straddle an intron, so its band on the gene is split across two exons.

Checking the R248 codon against the reference sequence

In the R248 session's genome view, zoom into the band the selection drew on TP53, down to base level, and tick Reference sequence in the track selector. The transcript is on the minus strand, so the codon under the band reads CCG left to right on the reference track, which is CGG, arginine, on the transcript. R248W and R248Q, the two commonest substitutions at the codon in tumours, change its first base and its middle one.

Structures of your own gene

The protein3d plugin is in the plugin store: install Protein3d from the Tools menu, then right-click a gene and choose Launch protein view to open its AlphaFold model. For several structures at once, write the view as a session spec. This is the spec behind the three-structure link; swap the UniProt accession, the PDB ids, the transcript and the locus for your gene's, and name a gene track your config has:

{
  "views": [
    {
      "type": "ProteinView",
      "structures": [
        { "uniprotId": "P04637" },
        { "pdbId": "1TUP" },
        { "pdbId": "1YCR" }
      ],
      "transcriptId": "NM_000546.6",
      "colorScheme": "mapped-chain",
      "connectedView": {
        "assembly": "hg38",
        "loc": "chr17:7,671,000-7,684,500",
        "tracks": ["hg38-ncbiRefSeq"]
      }
    }
  ]
}

Try this example live ↗

The transcript has to be in one of the connectedView tracks at loc, since the plugin maps each structure by aligning its sequence to that transcript's translation.

External links

Citations

Cho Y, Gorina S, Jeffrey PD, Pavletich NP. Crystal structure of a p53 tumor suppressor-DNA complex: understanding tumorigenic mutations. Science 1994.
Kussie PH, Gorina S, Marechal V, et al. Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science 1996.
Lee CW, Martinez-Yamout MA, Dyson HJ, Wright PE. Structure of the p53 transactivation domain in complex with the nuclear receptor coactivator binding domain of CREB binding protein. Biochemistry 2010.

Feedback on this tutorial is welcome: contact us.

TP53 from prediction to crystal

Where the data comes from#

Opening three TP53 structures mapped to one transcript#

UniProt annotations on the 1TUP core-domain crystal#

Finding the codon of the R248 hotspot from the crystal#

Choosing the mapped chain in the MDM2 complex 1YCR#

The p53 transactivation domain as an NMR ensemble#

Checking the R248 codon against the reference sequence#

Structures of your own gene#

See also#

External links#

Citations#