Subject: elbow gene
Dear Rachel
i have been perusing some of the elbow b gene literature and have a few queries.
The Flybase entry for elB states class i alleles enhance Sco but class ii
alleles do not yet the alleles
el4, el5 and el6 are given as class ii but all three enhance Sco?
.
thanks
Terry
Dr Terence Davis
Pathology
UWCM
Heath Park
Cardiff CF14 4XN
Wales, UK
tel \+44-29-20744964
\------------------------------------------------------------------------
Subject: Re: elbow gene
Hi Terry,
sorry to take so long to get back to you on this. I've been away and
then it is quite tricky....
The history is that the text in the following paragraph dates back to
Lindsley and Zimm (page 194-195, where it was associated with the
'elbow' entry (i.e. elbow A and B had not been defined at that point).
Michael apparently wrote this text (page 194-195).
\------------------------------------------------
Wings extended and bent backward, often warped and
shortened; sometimes blistered and nicked. Alulae
reduced in size with reduced number of marginal
bristles \-- may fuse with wing blade. Venation reduced
by terminal shortening of L5 and of crossveins.
Halteres reduced, often to stubs. Eye size decreased
(variable, even with strong alleles). Weak alleles may
overlap wild type and show only a reduction in number
of marginal bristles on alulae. Some alleles may be
semi-lethal when hemizygous. Class (i) alleles enhance
Sco, are semi-lethal with alleles of l(2)35Ba and show
a weak noc phenotype when heterozygous with strong noc
alleles or noc deletions. Class (ii) alleles do not
interact with these loci. All alleles more extreme
when hemizygous; strong alleles nearly apterous.
Part of el-noc complex.
\------------------------------------------------
When elA and elB were made 'genes' in their own right the text was
simply propagated to both new records.
Judging from this information we have from your paper
Davis et al., 1997, Hereditas 126(1): 67--75 ...
'The elbow locus is made up of two genes, elA and elB, each of
which has a distinct phenotype when mutant. Mutations in elA have a
strong phenotype with marked disruptions of the wing. Mutations in
elB are weak, mainly affecting the alula and wing bristles. elA
and elB are dominant enhancers of each other.'.....
the 'Wings extended' paragraph of text needs to be updated (for example
one update is that l(2)35Ba and Sco are now known to correspond to noc)
\- I have a plan for this ,whereby this paragraph will be attributed to
\*x FBrf0056834 == Ashburner, 1992, Lindsley, Zimm, 1992: 194--195 which
will de-emphasis it and put a long-ago date to it.
We are also in trouble about the alleles. When such a 'gene record
split' occurs it is very difficult for us to partition the alleles, as
we do not generally have enough information to know which of the new
genes the alleles represent. This is particularly difficult in cases
as genetically complex as the elbow-noc region. In the following table
I have enumerated the way it is in FlyBase at the moment \- the old
symbol in the first column and the alleles in the current version in
the right hand column. In those cases where there is both an elA and
an elB allele, each record in FB has a note saying that it might actually be
an allele of the other.
Old symbool Current symbol
el1 elA1
el2 elA2, elB2
el3 elA3, elB3
el4 elB4
el5 elB5
el6 elB6
el7 elB7
el8 elA8, elB8
el9 elB9
el10 elA10, elB10
el11 elA11, elB11
el12 elA12, elB12
el13 elA13, elB13
el21 elA21, elB21
el22 elA22, elB22
el23 elA23, elB23
el14 has become Df(2L)el14, hence not in this list.
I'm sending you this list in case you can cast any light on whether any
of these 'el' alleles with both an elA and elB allele can be more
accurately assigned to elA or elB. It is not transparently obvious why
some alleles have become alleles of elA, some elB and some both, though
such things are usually based on figures in papers (e.g. el4, el5,
el6, el7 and el9 became elB alleles because of your figure 2 in
Davis et al., 1997, Hereditas 126(1): 67--75.
The classification described by Ashburner in L&Z into class i and class
ii is extremely odd.
\----------------------------------------------------------------------
Class i alleles
\---------------
The following alleles were marked as class i alleles in L&Z \- el1,
el2, el3
'Class (i) alleles enhance Sco, are semi-lethal with alleles of l(2)35Ba
and show a weak noc phenotype when heterozygous with strong noc alleles
or noc deletions.'
Relevent FlyBase info from 'el1' i.e. elA1 record:
\*i el1
\*k Phenotype manifest in: wing
\*k Phenotype manifest in: wing vein L4
\*k Phenotype manifest in: wing vein L5
\*k Phenotype manifest in: alula
\*k Phenotype manifest in: dorsal row
\*k Phenotype manifest in: costal cell
\*k Phenotype manifest in: wing vein L1
\*k Both the anterior and posterior wing are severely reduced in hemizygotes,
\*k wing vein L4 is shortened, and wing vein L5 and the alulae are absent.
\*k The submarginal cell (between L2 and L3) appears relatively normal.
\*k In some wings the costal cell is extended and wing vein L1 reaches
\*k the wing margin more distally than normal. The triple row is reduced
\*k in the dorsal rows but not the ventral row.
\*k elA1/Df(2L)b83d29a adults are wild-type. elA1/Df(2L)b84a2
\*k adults have an 'elbow' phenotype.
Relevent FlyBase info from 'el2' i.e. elA2, elB2 records:
elA2 has:
\*i el2
\*o ethyl methanesulfonate
\*k Weak allele. Lethal in double heterozygous combination with Df(2L)TE35B-1,
\*k Df(2L)TE35BC-8, In(2L)TE35B-13, Df(2L)el28, Df(2L)A400 and
\*k Df(2L)TE35B-7 (rare Df(2L)el28, Df(2L)A400 and Df(2L)TE35B-7
\*k adult escapers are occasionally seen). Semi-lethal in double heterozygous
\*k combination with Tp(2;2)A446, Df(2L)n78l3, Df(2L)fn3 and
Df(2L)TE35B-6.
\Genetic interaction (effect, anatomy): enhancer | dominant, chaeta {
nocSco }
elB<up>2 has nothing relevent</up>
Relevent FlyBase info from 'el3' i.e. elA3, elB3 records:
<up>Neither elA3 nor elB3 have anything relevent</up>
CONCLUSION: Nothing for 'el1' either supports or contradicts the
Class i definition. Data for 'el2' does not directly contradict 'are
semi-lethal with alleles of l(2)35Ba' and does support 'Class (i)
alleles enhance Sco' with respect to the chaeta phenotype. There is no
data for 'el3'.
\----------------------------------------------------------------------
Class ii alleles
\----------------
The following alleles were marked as class ii alleles in L&Z \- el4,
el5, el6
'Class (ii) alleles do not interact with these (Sco, l(2)35Ba, noc)
loci.'
Relevent FlyBase info from 'el4' i.e. elB4 record:
\*k Phenotypic class: visible | recessive
\*k Phenotype manifest in: wing
\*k Phenotype manifest in: alula
\*k Hemizygotes have reduced wings, and, especially, reduced alulae.
\Genetic interaction (effect, class): enhancer | dominant { nocSco }
Relevent FlyBase info from 'el5' i.e. elB5 record:
\*k Phenotypic class: visible | recessive
\*k Phenotype manifest in: alula & macrochaeta
\*k Phenotype manifest in: wing | posterior
\*k Phenotype manifest in: wing vein L5
\*k Hemizygotes are almost wild-type, but have a reduction in the number
\*k of marginal bristles on the alulae and along the posterior wing margin.
\*k Wing vein L5 is sometimes shortened.
\Genetic interaction (effect, class): enhancer | dominant { nocSco }
Relevent FlyBase info from 'el6' i.e. elB6 record:
\*k Phenotypic class: visible | recessive
\*k Phenotype manifest in: wing
\*k Phenotype manifest in: alula
\*k Hemizygotes have reduced wings, and, especially, reduced alulae.
\Genetic interaction (effect, class): enhancer | dominant { nocSco }
CONCLUSION:
Thus your data from FBrf0093413 is directly at odds, in terms of the Sco
interaction data, with that from L&Z.
\----------------------------------------------------------------------
The best I can do for now is attribute the 'class' info to the
Ashburner, 1992, Lindsley, Zimm, 1992: 194--195 reference, for the same
reasons as with the chunk of text from the gene level. It is not
traceable and therefore we can't dig any further. I'm not sure that
the 'class' info contributes any useful info and if I deal with it in
this way then at least the info is not lost, only de-emphasised.
It may be that this is more than you had bargained for.... any help you
can give clarifying the records would be gratefully received. I'll
work on the updates now, but it will be some weeks before any changes
get made in the main FlyBase dataset. The records will certainly look
better for the changes you have provoked! They haven't had a spring
clean for many years.
All the best,
Rachel.
\------------------------------------------------------------------------
Subject: Re: elbow gene
Dear Rachel
thanks for your detailed reply.
Unfortunately most of the info concerning the el alleles and their
interactions with noc/l(2)35Ba are
buried in Mike's files (yes that lot!!). I used to know most of the relevant
but it's been some time. The
rest of the data is buried in the various of Mike's papers in Genetics from
1979-1983, (at least for alleles
el1, el2, el3 and el4).
But it is clear that el1 is elA (deletes the putative elA transcript CG15283)
and class ii (i.e. it does not
enhance Sco or have a noc phene with noc alleles).
el2 and el3 are class i as both strongly enhance sco, and are semi-lethal over
l(2)35Ba. In fact I believe
that el2 is actually an allele of l(2)35Ba/noc!! see my 1997 paper. This is
based upon its recombination
frequency which puts it well proximal of elA and its lethality over 1(2)35Ba
alleles.
el4, el5, el6, el7 and el9 are class i based upon slight but real enhancement
of Sco and are marked as elB
as they break in the first intron of the elB gene. I'm not sure of their
interaction with l(2)35Ba/noc, but
elB alleles do interact with noc.
el24 is more confusing, as it breaks proximal to putative elA gene but has
strong el phenotype and does not
enhance sco (putative elA class ii).
in summary:
class i alleles are elB alleles
class ii are elA alleles
the other alleles in the list are more difficult as their interactions with
l(2)35Ba/noc are not known to me
(they are probably in Mike's files), and they have little real effect on Sco.
they are all very weak so
difficult to assign and are all point mutations not aberrations.
Judging by dorfmann et al's paper it may be difficult to assign weak point
alleles to elB or to l(2)35Ba as
the two genes clearly interact and have similarities of phenotype
.
Best wishes
Terry
\------------------------------------------------------------------------
Subject: Re: elbow gene
Hi Terry,
the state of play is that I have renovated the records for elA and elB
and am about to leave until June 3rd \- so I can't do much more now. At
least the records are much tidier for the attention, in the meantime.
I think what you are telling me directly contradicts what is in L&Z for
several alleles, and therefore I should do a correction, citing your
mail as the source. The contradictions are
L&Z Your mail
el1 Class i Class ii
el2 Class i Class i
el3 Class i Class i
el4 Class ii Class i
el5 Class ii Class i
el6 Class ii Class i
I'm happy to change the class designations, if that is your intention.
I'll have to keep the old statements in text, but will cross reference
to the update (your mail).
You also say
>class i alleles are elB alleles
>class ii are elA alleles
which means I should delete el2 and el3 from the elA record and
leave them in elB, right?
I'll write to Shilo (corresponding author for FBrf0151266) now.
cheers,
Rachel.
Subject: Re: elbow gene
> L&Z Your mail
>
>el1 Class i Class ii
>el2 Class i Class i
>el3 Class i Class i
>el4 Class ii Class i
>el5 Class ii Class i
>el6 Class ii Class i
\-------- Basically yes. But as I said the el4, el5, el6 interactions with
l(2)35Ba are problematical. it
seems that elB alleles over l(2)35Ba are semi-lethal and/or weak noc, elA1
over deletions inc l(2)35Ba are
actually el <up>you see the problem</up>. i.e alleles. of both these genes seem to
interact with l(2)35Ba but
in different ways, this has never been fully investigated.
Thus I think the definitive interaction is with Sco as this seems to pose no
contradictions. The real data
are buried in Mike's files.
>You also say
>>class i alleles are elB alleles
>>class ii are elA alleles
>
>which means I should delete el2 and el3 from the elA record and
>leave them in elB, right?
\---yes. But it is possible that at least el2 is l(2)35Ba!!
Terry