Phosphatase Family DSP

From PhosphataseWiki
Jump to: navigation, search

Phosphatase Classification: Fold CC1: Superfamily CC1: Family DSP

This family consists of the dual-specific protein phosphatases (DSPs) that dephosphorylate both tyrosine and serine/threonine, as well as related non-protein phosphatases.


MKP: MAP Kinase Phosphatases with a Rhodanese Domain

Several related subfamilies of DSP that dephosphorylate MAPK Kinases and share an N-terminal non-catalytic rhodanese domain. These are named MKP, (MAP Kinase Phosphatase). They are regulators of MAPK activity, and can mediate crosstalk between distinct MAPK pathways and between MAPK signalling and other intracellular signalling modules (see reviews [1, 2]). The rhodanese domains usually contain kinase-interacting motifs (KIMs) for MAPK binding [1].


DSP1 is an inducible nuclear MKP found throughout eukaryotes. As a key player in MAPK pathway, it is implicated in immune regulation and cancer. Human has four members, DUSP1 (MKP1), DUSP2, DUSP4 (MKP2) and DUSP5.


DSP6 is a cytoplasmic MKP subfamily that selectively dephosphorylates ERK. It is found throughout metazoa and duplicated in vertebrates, including 3 human members: DUSP6 (MKP3/PYST1), DUSP7 (MKPX/PYST2) and DUSP9 (MKP4/PYST3).


The DSP8 subfamily is a metazoan subfamily that functions as an MKP with preference towards JNK and p38. It is single copy in invertebrate but two copies in most vertebrates. The two human members DUSP8 and DUSP16 (MKP7) have different tissue expression patterns.


Human DUSP10 (MKP5) selectively dephosphorylates p38 and JNK. It is conserved across holozoa but lost in nematodes. Human DUSP10 is frequently dysregulated in colorectal cancer.


The STYXL1 subfamily is a pseudophosphatase (catalytically inactive) conserved in metazoa but lost in ecdysozoa. It is also known as MK-STYX, named after the catalytically inactive phosphatase subfamily STYX. In comparison with STYX, it has an N-terminal rhodanese domain, which is a common feature between MKPs. Two binding partners have been known so far: phosphatase PTPMT1 and a Ras signaling regulator G3BP1.

Likely MKPs without a Rhondanese Domain


Human DSP3s are abundantly expressed in skeletal muscle and heart. It emerged in eumetazoa, lost in nematodes and duplicated in deuterostomes. Human has five members: DUSP3 (VHR), DUSP13 (BEDP/TMDP/MDSP/SKRP4), DUSP26 (MKP8), DUSP27, DUPD1.


The DSP14 subfamily emerged in eumetazoa and duplicated in vertebrates. Human has four members, DUSP14 (MKP6), DUSP18, DUSP21 and DUSP28 (VHP). Little is known about their functions.


The DSP15 subfamily emerged in metazoa and duplicated in vertebrates. It has a N-terminal myristoylation site which targets it to plasma membrane. Little is known about its molecular function.


DSP19 is found across eukaryotes but absent from fungi. Human DUSP19 (SKRP1) regulates JNK signaling but the mechanism is unclear.


The STYX subfamily of catalytically inactive phosphatases found in most opisthokonts but lost in nematodes.


Human DUSP23 is a nuclear phosphatase found in metazoa but lost in ecdysozoa.

Cyclin-dependent kinase phosphatases


CDC14 is a cell cycle phosphatase found in most eukaryotes other than higher plants.


CDKN3 (KAP) is a chordate-specific phosphatase targeting Cyclin-dependent kinases (CDKs) CDK1 and CDK2.

DSPs with non-protein substrates


The DSP12 subfamily is found throughout unikonts, with suggested roles in fat and glucose metabolism, MAPK regulation, ribosome biogenesis and cell cycle progression.


The RNGTT subfamily is an mRNA capping enzyme found in holozoa. It has a phosphatase domain and guanylyltransferase.


The DSP11 subfamily is a metazoan-specific subfamily. Its physiological substrate is unknown, but several lines of evidence link this phosphatase to RNA splicing. Human has a single copy DUSP11 (PIR1).


The laforin subfamily is a glucan phosphatase, found in vertebrates and scattered other species. Mutations in the human member, EPM2A, are associated with myoclonic epilepsy of Lafora.


PTPMT1 is a mitochondrial non-protein phosphatase that converts phosphatidylglycerolphosphate (PGP) to phosphatidylglycerol, during biosynthesis of cardiolipin. It is found in most or all animals and higher plants, and most protists but is absent from fungi, Monosiga, and some lower plants.

Other subfamilies found in human


PTPDC1 is found in holozoa and some protists, but lost from most insects. It may function in centriole and cilium biology.


The PRL (PTP4A) subfamily is present in animals, amoeba, and many basal eukaryotes, but is absent from fungi and plants (unpublished data from gOrtholog). The three human members, PRL1, PRL2, PRL3, have all been linked to cancer metastasis.


The slingshot subfamily is conserved in holozoa but lost in nematodes, regulates cofilin phosphorylation in opposition to LIMK and TESK kinases.

Non-human subfamilies and unclassified DSPs

Dictyostelium dupA

Dictyostelium DupA has an active kinase domain and an inactive phosphatase domain. The cysteine at CX5R motif of phosphatase domain is substituted by serine, so its phosphatase domain is probably catalytically inactive. The kinase domain is pretty divergent but has the key catalytic residues. It is also found in other Dictyosteliida, but not other amoebazoa, by BLASTing against NR - eukaryotes data set and amoebazoa data set. It may regulate a MAP kinase response to bacteria Legionella pneumophila [3].

Dictyostelium LRR-DSP

The LRR-DSP subfamily has LRR repeats at N-terminal. It is found in most amoebozoa by BLASTing against NR database through NCBI BLAST server.

Phosphatase domain structures

Almost all DSP phosphatase domains (PDs) have a secondary structure (SS) combination of E2, E3, H2, E4, E11, H3, E12, H4, H5, H6 (E denotes beta strand, H denotes helix, SS numbered by PTPN1. The combination is exactly same as the common SS combination in CC1 fold), except that some DSPs have the helix H2 split into two helices separated by single amino acid.

DSPs have dramatic diversity in structure, although they generally share the same SS combination, especially at the following two regions:

  • E3-H2-E4 region. As mentioned above, some but not all DSPs have the helix split into two helices. CDKN3 has a 2-stranded beta sheet inserted between E3 and H2.
  • E4-E11 region. The region contains the so-called WPD loop. Some DSPs have one or two helices inserted, even DUSP6 has an inserted beta strand interacting with E11.

Some DSPs have additional SS element(s) at the termini:

  • N-terminal helix. The DSP3 and laforin families, as well as vaccinia virus DSP (PDB code: 2P4D) have an additional N-terminal helix different in structure and evolutionary origins. The additional helix are involved in substrate recognition in both DUSP3 and laforin.
  • DUSP14 and DUSP18 of the DSP14 subfamily have a 2-stranded beta sheet followed by a helix at C terminus.
  • C-terminal helix. DUSP10 and DUSP12 of two different subfamilies have an additional C-terminal helix, but they occupy distinct spaces.

Three PDs do not all the SS elements as mentioned above:

  • First PD of CDC14. CDC14s have two tandem DSP PDs. The first PD is inactive and lacks H2, H3, E11.
  • RNGTT. RNTGG lacks E2 according the SS annotations of the PDB files of three RNGTTs (human, mouse and a virus). However, in the papers of mouse RNGTT and virus RNGTT (human RNGTT is not in publication), the authors presented E2, which was annotated through visualization (personal correspondences).
  • DUSP11. DUSP11 lacks E2 according to the SS annotations of the multiple PDB files of human DUSP11. However, the authors presented E2. It worthy pointing out that DUSP11 and RNGTT can be well aligned at the region in structure. Given the fact that both of them are involved in RNA processing or editing, it is interesting to find out the functional impact of the absence of E2.

Technical notes: the SS elements were annotated by using the program Stride to infer SS from PDB files.

Accessory domains

  • Most MKPs have an N-terminal domain of rhodanese fold.
  • Laforin has an N-terminal carbohydrate binding domain.
  • RNGTT has a C-terminal guanylyltransferase (GTase) domain.


  1. Dickinson RJ and Keyse SM. Diverse physiological functions for dual-specificity MAP kinase phosphatases. J Cell Sci. 2006 Nov 15;119(Pt 22):4607-15. DOI:10.1242/jcs.03266 | PubMed ID:17093265 | HubMed [Dickinson06]
  2. Caunt CJ and Keyse SM. Dual-specificity MAP kinase phosphatases (MKPs): shaping the outcome of MAP kinase signalling. FEBS J. 2013 Jan;280(2):489-504. DOI:10.1111/j.1742-4658.2012.08716.x | PubMed ID:22812510 | HubMed [Caunt13]
  3. Li Z, Dugan AS, Bloomfield G, Skelton J, Ivens A, Losick V, and Isberg RR. The amoebal MAP kinase response to Legionella pneumophila is regulated by DupA. Cell Host Microbe. 2009 Sep 17;6(3):253-67. DOI:10.1016/j.chom.2009.08.005 | PubMed ID:19748467 | HubMed [Li09]
All Medline abstracts: PubMed | HubMed