Skip Header

You are using a version of Internet Explorer that may not display all features of this website. Please upgrade to a modern browser.

UniProt release 2014_05

Published May 14, 2014


A flounder… on the rocks!

Some organisms, such as certain vertebrates, plants, fungi and bacteria, have to resist low, subzero temperatures. Their survival relies upon the production of antifreeze molecules. Some insects, like the beetle Upis ceramboides, tolerate freezing to -60°C in midwinter thanks to the production of a compound, called xylomannan, made of a sugar and a fatty acid and located in cell membranes. However, most organisms use antifreeze proteins (AFPs). All AFPs act by binding to small ice crystals to inhibit growth that would otherwise be fatal, but each type of AFP seems to arrive at this end by a different route.

Pseudopleuronectes americanus, commonly called ‘winter flounder’, is a very common variety of flounder in North America. It lives in cold water and survives thanks to the expression of the AFP Maxi. The 3D structure of the Maxi protein has been recently elucidated, unveiling some very unusual features.

Maxi belongs to the type-I AFP family and consists of a homodimer. Each monomer folds exactly in half so that its N-and C-termini are side by side, hence the dimer looks like a 4-helix rod. It is composed of tandem 11-residue repeats that exhibit the [T/I]-x3-A-x3-A-x2 motif, where x is any residue. The conserved threonine/isoleucine and alanine residues in this motif have been shown to bind ice in monomeric type-I AFPs. In the 3D structure, the internal space generated by the packing of the 4 helices in the 11-residue repeat regions is just wide enough to accommodate a single layer of water. Amazingly, the water layer that occupies the gap consists of over 400 molecules forming an extensive, mainly polypentagonal network. As is the case for most globular proteins, Maxi internal residues are nonpolar, mainly alanines, which obviously is far from optimal for hydrophilic contacts. To overcome this problem, Maxi takes advantage of its backbone carboxyl groups to anchor water molecules and the whole structure is stabilized by water-mediated hydrogen bonding rather than by direct protein association. The positioned water molecules extend outwards between all 4 helices from the core to the surface and they form a network of ordered molecules at the periphery. As a result, this rather hydrophobic protein remains highly solvated and freely soluble in flounder blood under physiological conditions, i.e. at low temperatures. When the temperature rises above 16°C, Maxi irreversibly denatures.

Another surprise came from the observation that the predicted ice-binding residues, expected to face the protein exterior, actually occur on the inward-pointing surfaces of all 4 helices where they cooperate to form and anchor the interior ordered waters. How then does Maxi bind to ice? The current working hypothesis is that the positioned water molecules that extend outwards may form a network available to merge and freeze with the quasi-liquid layer on the surface of ice.

As of this release, the winter flounder antifreeze protein Maxi has been annotated and integrated into UniProtKB/Swiss-Prot. All antifreeze proteins available in UniProtKB/Swiss-Prot can be retrieved with the keyword ‘Antifreeze protein’.

UniProtKB news

Update of ECO mapping for evidences

In 2011, we have started to use the Evidence Codes Ontology (ECO) to describe the evidences for UniProtKB annotations. Since then, this ontology has been extended and the GO Consortium has published a mapping of their GO evidence codes to ECO. We have adapted our mapping to ECO accordingly to have equivalent evidence codes for UniProtKB and GO annotations. How this affects different UniProtKB distribution formats is described below.

XML and DAS format

In these two formats, ECO codes are used to describe the evidences for UniProtKB annotations. In the UniProtKB XML format, an evidence is represented by an evidence element with a type attribute whose value is an ECO code. In the DAS (features) representation of UniProtKB, an evidence is represented by a METHOD element with an optional cvId attribute whose value is an ECO code.

The table below shows the mapping of previous to new ECO codes.

Previous ECO code New ECO code
ECO:0000001 ECO:0000305
ECO:0000006 ECO:0000269
ECO:0000034 ECO:0000303
ECO:0000044 ECO:0000250
ECO:0000203 ECO:0000501 and ECO:0000256

The codes ECO:0000312 and ECO:0000313 remain unchanged.

In the future, we will also use ECO:0000255 for UniProtKB annotations.

RDF format

In the UniProtKB RDF format, ECO codes are used to describe the evidences for UniProtKB and GO annotations. An evidence is represented by an evidence property whose value is an ECO code. The evidence property is part of an attribution object which is assigned to a UniProtKB or GO annotation via reification.

The table below shows the mapping of previous to new ECO codes.

GO evidence code Previous ECO code New ECO code
EXP ECO:0000006 ECO:0000269
IBA ECO:0000308 ECO:0000318
IBD ECO:0000214 ECO:0000319
IC ECO:0000001 ECO:0000305
IDA ECO:0000002 ECO:0000314
IEA ECO:0000203 ECO:0000501
IEP ECO:0000008 ECO:0000270
IGC ECO:0000177 ECO:0000317
IGI ECO:0000011 ECO:0000316
IKR ECO:0000216 ECO:0000320
IMP ECO:0000015 ECO:0000315
IPI ECO:0000021 ECO:0000353
IRD ECO:0000215 ECO:0000321
ISA ECO:0000200 ECO:0000247
ISM ECO:0000202 ECO:0000255
ISO ECO:0000201 ECO:0000266
ISS ECO:0000044 ECO:0000250
NAS ECO:0000034 ECO:0000303
ND ECO:0000035 ECO:0000307
RCA ECO:0000053 ECO:0000245
TAS ECO:0000033 ECO:0000304

Cross-references for isoform sequences: RefSeq

We have added isoform-specific cross-references to the RefSeq database. The format of these cross-references is as described in release 2014_03.

Cross-references to MaxQB

Cross-references have been added to MaxQB, a database of large proteomics projects.

MaxQB is available at

The format of the explicit links is:

Resource abbreviation MaxQB
Resource identifier UniProtKB accession number.

Example: Q6ZSR9

Show all entries having a cross-reference to MaxQB.

Text format

Example: Q6ZSR9

DR   MaxQB; Q6ZSR9; -.

XML format

Example: Q6ZSR9

<dbReference type="MaxQB" id="Q6ZSR9"/>

Removal of the cross-references to ProtClustDB

Cross-references to ProtClustDB have been removed.

Changes to the controlled vocabulary of human diseases

New diseases:

Modified diseases:

Deleted diseases:

  • Short rib-polydactyly syndrome 2B
  • Short rib-polydactyly syndrome 3

UniParc news

UniParc cross-references with multiple taxonomy identifiers

The UniParc XML format uses dbReference elements to represent cross-references to external database records that contain the same sequence as the UniParc record. Additional information about an external database record is provided with different types of property child elements, e.g. the species is represented with a property of the type "NCBI_taxonomy_id" that stores an NCBI taxonomy identifier in its value attribute. In the past, all external database records described a single species.


<dbReference type="REFSEQ" id="ZP_06545872" version_i="1" active="Y" version="1" created="2010-03-07" last="2013-07-18">
  <property type="NCBI_GI" value="289827083"/>
  <property type="NCBI_taxonomy_id" value="496064"/>
<dbReference type="REFSEQ" id="ZP_18488583" version_i="1" active="Y" version="1" created="2012-11-25" last="2013-07-18">
  <property type="NCBI_GI" value="425085490"/>
  <property type="NCBI_taxonomy_id" value="1203546"/>

With the introduction of WP-accessions in the NCBI Reference Sequence Project (RefSeq) database, UniParc needs to represent more than one species per dbReference element.


<dbReference type="REFSEQ" id="WP_001144069" version_i="1" active="Y" version="1" created="2013-07-19" last="2013-11-12">
  <property type="NCBI_GI" value="447066813"/>
  <property type="NCBI_taxonomy_id" value="496064"/>
  <property type="NCBI_taxonomy_id" value="1203546"/>

This change did not affect the UniParc XSD, but may nevertheless require code changes.