DPAGT1 (Q9H3H5, GPT_HUMAN) is GO:0031227 (intrinsic to the endoplasmic reticulum membrane) or GO:0005789 (endoplasmic reticulum membrane), not GO:0016021 (integral to membrane)
DPAGT1 is a gene involved in the synthesis of the precursor of N-linked glycosylation, process that takes place on the ER membrane. If you look at his Uniprot entry (http://www.uniprot.org/uniprot/Q9H3H5), it has various helices spanning the membrane.
However, at the moment DPAGT1 is annotated as "integral to the membrane", which makes it looks like it is a surface protein while it is not.
The GO annotation of "integral to membrane" is entirely consistent with the structure of DPAGT1 as annotated with its multiple transmembrane segments. The GO term "integral to membrane ; GO:0016021" is defined as:
Penetrating at least one phospholipid bilayer of a membrane. May also refer to the state of being buried in the bilayer with no exposure outside the bilayer. When used to describe a protein, indicates that all or part of the peptide sequence is embedded in the membrane.
Membrane in this instance means any membrane in a cell. If the annotation was "integral to plasma membrane ; GO:0005887" it would imply the protein is at the cell surface.
Also, questions about the use of the GO in annotation are best addressed to the GOC Annotation Issues tracker:
https://sourceforge.net/tracker/?group_id=36855&atid=605890
Cheers,
Alex
Thank you very much. I was just pointing out that GO:0005789 would be better suited for this gene, since it is integral to the ER membrane and it hasn't been observed in other compartments. Moreover, genes in the same pathway and location as DPAGT1 are associated with GO:0005789.
Sorry for posting in a wrong section, I will use the Annotations tracker for other reports.
Hi,
I'm a curator in the GOA group who is responsible for annotations to human proteins. If you can provide me with a published reference that contains experimental evidence showing human DPAGT1 is integral to the ER membrane, I will add the annotation. I have had a quick look in PubMed and have found nothing specific, there seems to be very little experimental evidence for this protein in mammalian species.
Best wishes,
Rachael.
It is difficult to find a proper reference, because this gene has been described in the '90s and is also known as ALG7 or GPT (even it the HuGO annotations don't say so).
you can have a look at PMID: 12872255 where they describe a mutation on DPAGT1 in human, and especially at the first paragraph of the second page, where it says:
""" DPAGT1 is an essential gene found in all eukaryotic
organisms ranging from yeast to man and is a highly
conserved ER-resident membrane protein with 10
transmembrane helices (Fig. 1) [Lehrman et al., 1988;
Zhu and Lehrman, 1990; Scocca et al., 1995; Dan and
Lehrman, 1997; Eckert et al., 1998]."""
DPAGT1 is homologue to the yeast gene ALG7, a gene involved in the first step of the N-glycan precursor pathway: these genes are highly conserved* and experiments have demonstrated that the human DPAGT1 can complement an ALG7 mutation in yeast (PMID: 9451016). (ALG7 is GO:0005783 and GO:0016021 in yeast, maybe it should go to GO:0005789 as well)
One of the substrates for the reaction catalyzed by DPAGT1 is Dolychyl-P, a phosphorilated phospholipid on the ER membrane, by DOLl (PMID: 214107, 923811, 849790, 12213788, 16923818) or DOLPP1 (6243292), which are both "Integral to the ER membrane". The product of this reaction is Dol-P-mannose, used as substrate by ALG13/ALG14, a complex of two enzymes localized in the ER (ALG14 is on the ER membrane, ALG13 is recruited by ALG14) (16100110, 4717748, 434833, 6153388). This is also described in PMID: 12872255 and its references.
The final product of the pathway in which DPAGT1 is involved is called "N-glycan precursor" or "LLO", is a sugar attached to secretory proteins while they are being synthesized on the ER membrane, and also used as a very important signal required for their correct folding. Proteins lacking a certain modification of the LLO can't be moved to the Golgi, therefore the LLO must be synthesized and attached to the protein before the Golgi (19853458)
I am sorry but I don't know which technique could be used to demonstrate that a protein is localized in the ER or somewhere else, therefore I don't know very well what I should be looking for.
* http://www.ensembl.org/Homo_sapiens/Gene/Compara_Ortholog/Alignment?g=ENSG00000172269;g1=YBR243C
(missed some of the comments when I switched this to the annotation tracker; reassigning)
yep, I know that GO:0016021 is correct, but GO:0031227 would be more consistent, more informative.
Most of the literature is for the hamster protein and I have managed to find experimental evidence for 'integral to ER membrane' for the hamster DPAGT1 in PMID:8940049. I have transferred this annotation to the human entry using the ISS (Inferred from Sequence Similarity) evidence code.
I have also made a NAS (Non-traceable Author Statement) evidenced annotation to the same term from PMID:12872255 that states 'DPAGT1 is an essential gene found in all eukaryotic organisms ranging from yeast to man and is a highly conserved ER-resident membrane protein with 10 transmembrane helices ', since none of the references supplied to support this statement contained data for the human protein.
These annotations will be visible in our QuickGO browser (http://www.ebi.ac.uk/QuickGO/GProtein?ac=Q9H3H5) by next week.
I hope this is useful to you.
Best wishes,
Rachael.
you can also have a look/cite the book Essentials of Glycobiology, which is the first book to have a PMID:
Stanley P, Schachter H, Taniguchi N. N-Glycans. In: Varki A, Cummings RD, Esko
JD, Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essentials of
Glycobiology. 2nd edition, Cold Spring Harbor (NY): Cold Spring Harbor Laboratory
Press; 2009. Chapter 8.
PubMed PMID: 20301244.
in chapter 8, it explains the reactions in which DPAGT1 is involved, which all occur in the ER (have a look at the figures: ALG7 == DPAGT1).
However, I can't point you out an article where there is evidence that the protein is localized in the ER in human... because I don't know which kind of experimental techniques are needed to prove that,or what are the criteria used in GO.
So, if you want you can use 20301244 along with 12872255, it would be a more-traceable author statement.
We don't usually cite books since these represent an accumulation of knowledge from many sources and it is usually quite difficult to determine which species any particular piece of evidence is from. I have had a quick look at the book you cite and there doesn't seem to be any species-specific information. Additionally, for any given annotation we can only provide one PMID, not a combination.
An appropriate experiment to locate a protein to the ER membrane would be solubilization of membrane preparations to determine if it is integral to the membrane in combination with immunofluorescence microscopy to determine which subcellular compartment the protein is localised to. See PMID:7814381 for an example of this technique.
I know it is very frustrating when there is common knowledge that a protein has a particular role, but until someone does an appropriate experiment on the human protein we cannot add an annotation to an experimental evidence code.
Best wishes,
Rachael.