ACS
Discussion Page
HGNC is used here as an
abbreviation
for the HUGO Gene Nomenclature
Committee's responses, and MGNC is used as an
abbreviation
for the Mouse Genome
Nomenclature
Committee's
responses. Both
your comments (disclaimer) and the nomenclature committees'
replies have been highlighted throughout the text to help you locate
this information.
Resolved
nomenclature issues
Dr
Paul Black 1, 2
|
Dr Karin
E. Bornfeldt 1, 2 |
Dr
Rosalind Coleman 1, 2, 3, 4, 5, 6, 7, 8
|
Dr Concetta
DiRusso 1, 2
|
Dr
Steven A. Farber 1, 2
|
Dr Brian K. Hubbard |
Dr
Daniel W.
Nebert 1, 2, 3
|
Dr
Paul A. Watkins 1, 2 |
Dr Johannes
Berger |
Dr
Tokuo T. Yamamoto 1, 2
|
|
|
Nomenclature
Dr
Rosalind Coleman
Some of us
discussed the growing nomenclature
problem with ACS, FATP, FACL, etc at the Lipid Gordon Conference last
summer and wanted make the names and numbering system for mammals
consistent within the field.
In order to fix
these problems, the acyl-CoA
synthetase community needs to decide what to do. First, we
need to identify the interested scientists. Then we can make naming and
numbering decisions. We should send copies of all emails to the
human genome nomenclature committee to keep them informed about what we
are doing. (hgnc@genenames.org). When we achieve a consensus
and they agree, they will resolve issues with the mouse, etc.
nomenclature.
Action Items
1. Please take
a few minutes to think about people
not on the email list but who should be involved and let me know.
2. Should this
effort include scientist working on
short, medium and very long chain enzymes or should we focus on the
long chain acyl-CoA synthetases only?
3. We are
thinking about having a meeting this
spring to discuss these issues. Or will email suffice? What
is your thought?
HGNC response:
The
currently approved stem name for the long-chain genes is
"fatty-acid-Coenzyme A ligase, long-chain
#". However, we appreciate that the gene name does not conform to
the
IUBMB ideal, and so one option is to change the name to
"fatty-acid-Coenzyme A synthetase, long-chain #" which allows us to
keep
the approved FACL# stem symbol.
Questions we need answered
1. Does the
community prefer to use synthase or
synthetase?
2. How does the
community feel
about the "fatty-acid-Coenzyme A synthetase" part of the name? Is
this
accurate for all the ACS family, or would a
change to an acyl-CoA synthetase
type name and symbol be preferred?
Dr
Rosalind Coleman
Personally I
prefer LACS because we can then use
MACS for med chain and VLACS for very-long chain, etc. Would
there be any problem with that?-- though I suppose we could go for
families like the ABC transporters- but then only the initiated would
know which family was which.
Another problem
is that some VLACS-like proteins are
called FATP for "fatty acid transport protein."
HGNC response:
We would strongly advise against LACS,
MACS etc. as these symbols are not hierarchical. In this case the
main family is acyl-CoA synthetase, so for example the ACS stem
followed by M (medium chain), L (long chain) etc. would be better from
a nomenclature point of view as it would enable easy database searching
using the stem symbol.
Dr Concetta DiRusso
A complication
with the yeast and bacterial
nomenclature may be that gene names are often assigned on the basis of
homologies and the protein names are derived from the gene names.
For example, we called the gene FAT2 because the protein shared
sequence identities with FAT1, which was named based on identities with
murine FATP. In standard yeast nomenclature, the protein is named
for the gene so the FAT2 gene encodes the Fat2p protein. It will
be very hard, if not impossible, to change gene names.
Most of the
long chain acyl-CoA synthetases in
bacteria, yeast and plants are named for the E. coli FadD gene encoding
a med-long chain acyl-CoA synthetase. Thus there are many fadD
genes in mycobacterium, for example, which merely have the AMP,
adenylate forming, signature in common.
My basic
question is, are we going to suggest a
nomenclature for the proteins or for both the proteins and genes?
HGNC response:
The HGNC only has authority to name
human genes. Whilst we do try to approve gene names and
symbols that could be used for the proteins, this is not always
possible. On this issue our guidelines
state:
Enzymes and proteins
The rules
described in the sections on gene names and symbols apply, but in
addition
the names of genes coding for enzymes should be based on those
recommended
by the Nomenclature Committee of the International Union of
Biochemistry
and Molecular Biology e.g. FPGS "folylpolyglutamate synthase". These
can be found at URL <http://ca.expasy.org/enzyme/>.
Names of genes encoding plasma proteins, hemoglobins, and specialized
proteins
are based on standard names and those recommended by their respective
committees
e.g. HBA1 "hemoglobin, alpha 1".
Dr
Rosalind Coleman
In thinking about the numbering
system, since most of the published
studies are from rat, can we not keep the rat numbers and let either
FACL1 or FACL2 be designated "6".
Dr Daniel W Nebert
In RATMAP, the
instructions say that you need to
find what has been
officially named in the human and mouse, then apply this to rat.
Searching the GDB, FACL1 and FACL2 are found to be at human 3q13 and
4q34-35. Searching the MGD, Facl2 has been named and it is located on
mouse Chr 8. Looking at the human-mouse
synteny on mouse Chr 8 there
is a portion (around 30-40% from
proximal end) that is syntenic with human Chr 4.
It
appears that human FACL1 and FACL2 have
been named, and that
mouse Facl2 has been named and appears to be the ortholog of human
FACL2. I don't think FACL3 and beyond have been officially named in the
human or mouse.
HGNC response:
FACL1
to FACL6 inclusive are all approved, officially named genes in
human.
Facl2 to Facl6 inclusive are approved, officially named genes in mouse
and rat.
Synthase
vs Synthetase
MGNC response:
The Mouse Genomic Nomenclature
Committee agrees with the suggested stem symbol ACS, rather than LACS,
MACS, etc. Database searching is an important issue to consider.
However, as pointed out by HGNC, the name synthase vs synthetase
would need to be addressed and decided upon.
Dr Paul
Black
I would prefer
'synthetase' as opposed to synthase. This is especially important
so there is no confusion with fatty acid synthtase.
Dr
Rosalind Coleman
Here's an"official word". Though I propose keeping
synthetase since that's what most publications are using these days.
Dr Brian K Hubbard
'Synthetase' is more appropriate relative to 'Synthase'.
Synthetase refers to the reaction that utilizes ATP, making an
acyl-AMP intermediate. For example carbamoyl
phosphate synthetase,
cytidine
triphosphate synthetase,
glutamine
synthetase,
asparagine
synthetase
Dr Paul A. Watkins
Despite my
personal preference for the term "synthetase", it seems that the
term "synthase" is now preferred, as well as "ligase".
| Dr
Fumiharu Yagasaki |
I
agree with
Paul Black's opinion. I would prefer' synthetase' as opposed to
synthase. |
| Dr David A. Bernlohr |
Synthetase is preferred. |
Dr Johannes Berger
|
We
also would prefer synthetase and not synthase or ligase. |
| Dr
Tokuo T. Yamamoto |
I
agree to acyl-CoA synthetase. |
HGNC response:
As we have a majority in favour of
synthetase, we will base the approved nomenclature on this. We
would also appreciate your thoughts on the issue of fatty-acid-Coenzyme
A synthetase (FACL) vs acyl-CoA synthetase (ACS).
Fatty-acid-CoA
vs Acyl-CoA
Dr Johannes Berger
The main
problem in my point of view is that we might not know the entire
spectrum of natural substrates for most of the ACS and thus it might
turn out later that the most relevant substrate for e.g.
"very-long-chain acyl-CoA synthetase" might be not the very-long-chain
fatty acids.
Dr
Rosalind Coleman
I would like to extend Dr. Berger's
point. Substrates other than fatty acids are encompassed by the term
"acyl," so it would be more useful to speak of acyl-CoA synthetase
rather than fatty acid-CoA synthetase
Dr Concetta
DiRusso
I agree with
the term synthetase and agree that for many if not most enzymes listed
as acyl-CoA synthetases in the sequence data banks we do not know the
natural substrate. Assigning a nomenclature based on the
substrate and not on sequence homologies should, in my opinion, be the
ultimate goal of this committee.
Most of the
discussion thus far has focused on the mammalian long chain acyl-CoA
synthetases. Are we going to address the broader questions
regarding other substrates and other organisms?
HGNC response:
Bearing in mind the discussion so far,
we will consider changing the nomenclature of the long-chain ACS
genes as shown
here. Please
email your views on the proposed stem symbol and the
numbering.
We do not intend to change the nomenclature of other
gene families whose nomenclature is already well used, for example the
FATP proteins are encoded by the SLC27A#
gene family
(recently reviewed by Stahl
A. Pflugers Arch. 2003). However, we could add aliases
based on the ACS root as shown here if the
community felt this was worthwhile.
ACS
long-chain family numbering
HGNC response:
There are basically two
options:
1) Rename rodent
ACS1/Facl2 & hum FACL1 and FACL2 as
ACSL1, as shown in Dr
Coleman's table below. This would mean there would be no ACSL2.
2) Rename rodent ACS1/Facl2 & hum FACL1 and FACL2 as
ACSL2, as
shown
here. This would mean there would be no ACSL1.
From a nomenclature point of view,
either option is fine.
Dr
Rosalind Coleman
Since there are
already many publications on "ACS1" (original name) in rodents (and
very few on rat/mouse "ACS2" (original name) ), I wonder whether it
might not be better to eliminate the human FACL2-designation instead of
FACL1. This would avoid confusion regarding the existing papers
and result in the following:
Current
numbering
|
Proposed
numbering
|
| rodent
ACS1 /hum FACL1 and 2 |
ACSL1 |
| rodent
ACS3/hum FACL3 |
ACSL3 |
| rodent
ACS4/humFACL4 |
ACSL4 |
| rodent
ACS5/humFACL5 |
ACSL5 |
| rodent
ACS 2/hum FACL6 |
ACSL6 |
Dr Daniel W Nebert
This makes sense
to me. I presume that another gene called "MACSL1" which is
COMPLETELY unrelated, would not affect the naming of this ACSL root.
HGNC response:
This is correct, MACSL1 does NOT affect
the use of the ACSL# stem.
Dr
Tokuo T. Yamamoto
According
to Dr. Coleman's designation, there is no ACS2 in her list. Since
human FACL6 is an ortholog of rodent ACS2, it is preferable to be
read as ACS2. In addition, ACSL is not preferable nomenclature. It is
preferable to be read as original ACS or alternatively as to LACS,
L-ACS or ACS-L.
HGNC response:
I would like to
reiterate, we strongly advise against using LACS,
MACS etc. as these symbols are not hierarchical. The ACS stem
followed by M (medium chain), L (long chain) etc. is more appropriate
for database searching issues.
The current approved symbol for the rat ortholog of
the human FACL6 gene is Facl6. Also the symbol ACS2 has
previously been used for both FACL5 [Minekura H et al Gene.
2001 Oct 31;278(1-2):185-92]
and FACL6 [Yagasaki F et al Genes
Chromosomes Cancer. 1999 Nov;26(3):192-202].
It would therfore be confusing to use ACS2 as an approved symbol for
either FACL5/Facl5 or FACL6/Facl6. As the FACL# nomenclature is
not accurate, it would be more appropriate to create a new stem symbol,
i.e. ACSL#.
Dr Steven A. Farber
I want
to suggest that ACS1 be used as opposed to ACS2- in that
most papers thus far have used ACS1 and very few have been published on
ACS2. That would leave the final list to be ACS1,3,4,5 and 6.
HGNC response:
As there are more people in favour of
keeping number '1' and removing number '2' I will initiate these
changes in the human, mouse and rat
databases with the approved nomenclature as shown
here.
Other
ACS family nomenclature
Dr Paul
Black
Regarding
other ACS gene families - and more specifically the fatty acid
transport proteins (FATP), it seems appropriate that we begin with the
ACS root; as pointed out previously we do not know the precise
substrates for this family of enzymes although there is significant
information that they play a role in fatty acid trafficking (including
transport). The Lodish group has proposed FATP1,2,3,4,5,6; this
is based on properties related to fatty acid transport. We along
with others (Bernlohr, Watkins, Stremmel, Berger) have shown members of
this sub group do indeed have acyl CoA synthetase enzymatic
activity. In the case of FATP1 from mouse, the Bernlohr group has
shown a fairly broad range specificity for activating fatty acids of
varying chain lengths; data from our lab is consistent with this as
well.
The question to be addressed is how members of this
sub group of ACS enzymes is also involved in fatty acid
trafficking. Perhaps a better statement is that all ACS enzymes
are involved in fatty acid trafficking, with some being more specific
towards the movement of exogenous fatty acids into the cell.
Clearly there is more to be done before there is a resolution.
This being stated, however, if one looks at the FATP family of proteins
and compares them to the ACS family, it is quite clear both have the
ATP/AMP binding motif that is a hallmark of adenylate-forming enzymes.
Further all of the ACS enzymes with a long-chain specificity have a
FACS motif (which we defined using affinity labeling and mutagenesis);
many of the features of this later motif are found in the FATP family
(or as I prefer a sub group of the ACS enzymes). Watkins and
colleagues have published alignments in the context of bubblegum, which
emphasize these similarities further.
Dr
Rosalind Coleman
I agree with
Paul Black. Based on the amino acid sequences shown by Paul
Watkins (J Biol Chem. 2000 Nov
10;275(45):35162-9) there is no difference between the so called
"FATP"s and the so called "very-long-chain ACS"s. Functionally, ALL the ACSs probably increase the
amount of FA that enters cells because they trap FA as acyl-CoAs.
This property may have led to the designation of some as FATPs.
At the very least, there should be cross referencing. A
better solution would be to label both FATP and very-long-chain ACS as ACSVL with the thought that some
may have additional functions.
HGNC response:
I would just
like to make it clear that the FATP# symbols are NOT approved gene symbols. The genes
encoding these proteins have the SLC27A#
approved stem symbol.
We do not intend to
change the nomenclature of the SLC27A genes. However, we would be
willing to add ACS-type
symbols as aliases to the SLC27A gene records. Please view the
SLC27A
gene family, and let me know if you think there are any
members of this family missing from the table.
Orthology
Dr Paul A. Watkins
Regarding the human long-chain enzymes,
I have looked extensively into the sequences of FACL1 (NP_001986)
and FACL2 (NP_066945). By comparison to both EST sequences and genomic
sequences, there seems to be no hard evidence to support the existence
of FACL1 as a distinct entity. As you know, this is what helped screw
up the nomenclature between human and rodent long-chain enzymes. I am
working with a local bioinformatics guru (Jonathan Pevsner--check out
his recently published book!) to verify before communicating this to
NCBI.
I think
a reasonable approach is to first
deal with the long-chain enzymes since this seems relatively
straightforward (once the FACL1/FACL2 controversy is resolved). Perhaps
then we can deal with rest more efficiently.
Dr Steven A. Farber
I would like to
propose using the well
described syntenic clusters observed between human/mouse/zebrafish as a
model for resolving the inter-species homology issues. For
example, one would look at the genes immediately adjacent to a human
ACS gene. Then one simply identifies the mouse orthologues of these
human genes. It is very likely that these genes remain clustered and
reside near a mouse ACS gene. One could also conclude that these
mouse and human orthologues were derived from a common ancestor and
should be given the same name. In fact in my study of 11
zebrafish annexin genes I was ability to unambiguously identify 10
zebrafish orthologues using this technique and then named them
according to the human convention (Farber
S.A., Olson E., and Halpern M.E. (2003) The zebrafish annexin gene
family. Genome Research 13:1082-1096). It is
interesting to note that prior to this type of analysis many people
considered the fish genes to be new family members.
Dr Daniel W Nebert
I hate to throw water on anyone's parade, but we
have a large data-mining paper coming out [Pharmacogenetics 2004; 14:1-18] that compares the 57 human CYP
genes to the 102 mouse Cyp genes, plus 58 human and 88 mouse
pseudogenes.
I can say for
certain that we DO NOT see any strong
relationship between the gene order along a chromosome (in the various
clusters, subfamilies of this superfamily) and the unequivocal
orthologs in these two species. Now, the ACS genes
might not be such an evolutionarily rapid-moving target, or such a
complicated superfamily, as the CYP genes, but caution is
recommended in this approach.
Dr Karin E. Bornfeldt
What you say
about FACL1 and FACL2 is
interesting. Is it not possible that they are splice
variants? We have run RT-PCR on human arterial smooth muscle
cells, in which we find FACL2 mRNA but not FACL1 mRNA. However, FACL1
mRNA was found in human liver cells using the same primers, so we think
the primers are fine. The primers we used for FACL1 were (left)
ttccagacgactcaccacc and (right) tacgagtccatgacaacta and the FACL2
primers were (left) tgcagcactcaccaccttc and (right)
taggcatccatgacaacta. I'd be interested to hear what you or anyone
else think about this.
HGNC response:
The current human/mouse/rat orthology
in the databases, is as shown in the long-chain table on the main
ACS page.
This orthology also corresponds to the intial tables sent by Dr
Rosalind Coleman. As far as we can determine the main challenge
is to agree the homology between FACL1, FACL2, mouse Facl2 and rat
Facl2.
Dr
Rosalind Coleman
According to
the table,
FACL1 and FACL2 are on different chromosomes. This should settle
the questions about whether they are splice varients, etc. Or is
there any possibility that the chromosome assignment is in error?
Dr Karin E. Bornfeldt
It appears that
human FACL1 is located on
chromosome 4, just like FACL2. According to ncbi, FACL1 is listed on
chromosome 3, but the accession
number actually maps to chromosome
4 (same location as
FACL2). If this is indeed correct, I suggest one name for both
FACL1 and FACL2 (e.g.FACL1 or FACL2).
HGNC response:
We have asked the NCBI Ref Seq group to
investigate whether FACL1 is correctly placed on chromosome 3. If
FACL1 turns out to be a splice variant of FACL2, we would recommend
that the chromosome 4 gene retain the FACL2 nomenclature, as this is
the numbering used for the mouse and rat orthologs.
NCBI Ref Seq
response:
It's always possible that there is a
locus missing from the chr3 assembly but, based on my analysis of the
genomic and transcribed sequence currently in the
database, I'd have to conclude
that FACL1 and FACL2 are, in fact, one gene.
The two accessions that initially defined these
loci are L09229.1 (FACL1) and D10040.1 (FACL2). Both GenBank accessions
were annotated as chromosome 4 by the submitters and the two sequences
align at the same locus on 4q35 in build 34.
Of these two accessions, D10040.1 appears to be
higher quality sequence, with only 5 mismatches to the underlying
genomic across its entire length. The
3'-end of L09229.1 has 48 mismatches and the first two exons have
many more. If there was transcribed evidence in support of these
mismatches, it would
suggest that the chr3 locus exists but is not represented in our
current assembly. But there is no genomic, mRNA or EST support
for the mismatches seen in L09229.
HGNC response:
Based on NCBI's analysis the FACL1 gene
record will be withdrawn and the associated sequences reassigned to
FACL2. If you disagree with this course of action please let us
know urgently.
The
FACL1 sequences have now been reassigned to
FACL2.
Back
to main ACS page