First published online 12 November 2008
doi: 10.1242/dev.028951
Development 135, 4071-4079 (2008)
Published by The Company of Biologists 2008
Two-step selection of a single R8 photoreceptor: a bistable loop between senseless and rough locks in R8 fate
Kathryn L. Pepple1,
Mardelle Atkins2,
Koen Venken1,
Kari Wellnitz3,
Mark Harding3,
Benjamin Frankfort1 and
Graeme Mardon1,2,3,4,5,*
1 Department of Molecular and Human Genetics, Baylor College of Medicine,
Houston, TX 77030, USA.
2 Program in Developmental Biology, Baylor College of Medicine, Houston, TX
77030, USA.
3 Department of Pathology, Baylor College of Medicine, Houston, TX 77030,
USA.
4 Department of Neuroscience, Baylor College of Medicine, Houston, TX 77030,
USA.
5 Department of Ophthalmology, Baylor College of Medicine, Houston, TX 77030,
USA.

View larger version (141K):
[in this window]
[in a new window]
|
Fig. 2. ro is not required for initial selection of a single R8
precursor. (A,E) Wild-type (wt) Sens expression. The boxed
area in A is shown in E. Sens is first expressed in a subset of IG cells in
column 0 (open arrowhead). Single R8s are identified in column 1 (double
arrow). (B,F) At permissive temperatures (18°C)
Dlts has a mild effect with rare additional R8s. Open
arrowheads indicate column 0 in all panels. (C,G) After a 6-hour
heat shock at 31°C (affected columns bracketed), multiple Sens-positive
cells form in column 1 (black arrowhead). (D,H) In
roX63 mutants (null allele), a single Sens-positive cell
forms in column 1 (double arrow). (I) The developing eye field in a
roX63 mutant from column 1 (right) to the posterior of the
disc (left). Additional Sens-positive cells are present in older ommatidia. An
asterisk indicates two ommatidia in column 5 with three Sens-positive cells.
Only single Sens-positive nuclei are observed in column 1 (double-arrow).
Column 0 and the mid-section of the disc are out of the plane of focus.
|
|

View larger version (83K):
[in this window]
[in a new window]
|
Fig. 3. ro is required to repress R8 differentiation in column 3.
(A) Normally, the neuronal marker embryonic lethal abnormal vision
(Elav) is expressed in all developing photoreceptors. (B) Wild-type
Sens. (C) Wild-type Ato. (D) Merge of A-C shows that, in a
wild-type disc, Ato and Sens are co-expressed in clusters of cells within an
intermediate group (circled) and in single Ato- and Sens-positive R8s in
column 1 (open arrowhead). In column 3, wild-type ommatidia always have a
single Ato- and Sens-positive cell. (E-H) In roX63
null discs, Elav (E) expression is delayed by one column whereas Sens (F) and
Ato (G) expression are initially unchanged from wild type. (H) Ato and Sens
are co-expressed within the IG (circled) and single Ato- and Sens-positive R8s
are selected (open arrowhead). More posterior ommatidia often have additional
Sens-positive cells (white arrowhead). (I) In column 3 of
rox63 mutants, 9±3% of ommatidia have extra
Ato-positive cells and 21±5% of ommatidia have extra Sens-positive
cells. Error bars represent the standard error of the mean in I and J.
(J) The ro mutant phenotype develops starting in column 3 with
21±5% of ommatidia containing more than one Sens-expressing cell. The
average percentage of ommatidia with multiple Sens-expressing cells for
columns 1-7 in roX63 mutants is shown.
|
|

View larger version (73K):
[in this window]
[in a new window]
|
Fig. 4. Identification of the senseless eye-specific enhancer. A
645 bp fragment within the second intron of the sens genomic locus
named F2 was identified that is sufficient to drive reporter expression
specifically in the developing eye. (A) The position of F2 in the
sens genomic locus among the 12 fragments tested is shown. Fragments
F1 and F3 overlap F2, but do not drive reporter expression in the eye.
(B) The 645 bp sequence of F2 contains two potential Ro-binding sites
(highlighted in red) and four potential Ato-binding sites or E-boxes
(highlighted in green). (C) In larval eye-antennal discs, Sens is
expressed in the R8 photoreceptors (bracket), the ocelli (arrow) and antennal
SOPs (asterisk). (D,E) F2-GFP is expressed only in
photoreceptors.
|
|

View larger version (84K):
[in this window]
[in a new window]
|
Fig. 5. Rough directly represses sens expression in R2,5
precursors. (A) EMSAs were performed using wild-type (wt) and
mutant (*) probes for the predicted Ro-binding sites H1 and H2. Ro
binds specifically to wild-type and mutant probe H1* (black
arrowhead). Specific binding is lost with the H2 mutation (H2*) and
with mutation of both H1 and H2 (H1,2*). The white arrowhead
indicates non-specific binding. (B) Probe sequences. Predicted core
sequences are underlined. Mutations are indicated in red. (C)
Subfragments used for GFP reporter studies in vivo. B-short is the minimal
R8-specific subfragment of F2. B-short contains two potential Ro-binding sites
(H1, H2) and one potential Ato-binding site E-box 1 (E1). Red lines indicate
the position of Ro-binding sites. Mutations are indicated by an X.
(D-F) Expression of the B-short-GFP reporter. Sens and B-short-GFP
colocalize to a single R8 per ommatidium (arrow). (G-I) Mutation of H1
in B-short (H1*) causes expansion of GFP expression to two
additional cells per ommatidium (arrowheads). (J-L) Mutation of H2 in
B-short (H2*) also expands GFP expression to two additional cells
(arrowheads). (M-O) The additional GFP-expressing cells are R 2,5
precursors. (M) RM104-β-gal (red) is expressed in the R2,5
photoreceptors. (N) H1*-GFP is expressed in three cells per
ommatidium. (O) Colocalization of the RM104-β-gal and H1*-GFP
in R2,5 cells (open arrowheads). (P-S) rox63
clones, marked by the absence of β-gal (P), Sens (Q) and GFP (R), expand
to three cells per ommatidium (arrowheads). In wild-type tissue, Sens and GFP
are expressed in a single cell (arrow).
|
|

View larger version (98K):
[in this window]
[in a new window]
|
Fig. 6. Positive and negative regulatory regions of the eye enhancer identified
by subfragment analysis. GFP reporter constructs were generated with
subfragments of F2 and tested for in vivo expression. (A) Relationship
of subfragments to F2. The blue bracket indicates the negative regulatory
region containing the Ro-binding sites H1 and H2 shown as red vertical bars.
Fragment sizes: F2, 647 bp; A, 324 bp; B, 324 bp; B-short, 266 bp; B-long, 383
bp; C, 324 bp. (B-D) High magnification image of Sens (B), F2-GFP (C)
and their co-expression (D) in R8 photoreceptors. GFP perdurance marks
additional cells of the IG not selected as the R8. White bars indicate IG
boundaries. (E-G) Fragment A does not express GFP in the eye.
(H-J) Fragment B-short is sufficient to drive GFP in single R8s but not
in IGs. (K-M) Fragment B drives GFP strongly in Single R8s, and weakly
in IGs. (N-P) Fragment B-long drives GFP robustly in both IGs and
single R8s. (Q-S) Fragment C-GFP is expressed at high levels in IGs and
in most cells posterior to the MF.
|
|

View larger version (78K):
[in this window]
[in a new window]
|
Fig. 7. Atonal directly regulates early senseless expression.
(A) EMSAs were performed using wild-type (Wt) or mutant (*)
probes for E-boxes 1-4. Ato/Da heterodimers bind E1 strongly and E4 weakly
(black arrow). Specific binding to E1 and E4 is lost when the E-box core
sequence is mutated to AANNTT. No binding is observed by Ato alone. Da
homodimers bind to E1 and E4 (white arrowhead). This interaction is also lost
with E-box mutations. No binding was detected to E2 or E3. An additional
nonspecific band of higher molecular weight is present in all Ato/Da and Da
reactions. (B) Fragment E1* was generated by mutation of E1,
the sole E-box in B-short-GFP (indicated by an X). (C-E) Expression of
Sens (C) and B-Short-GFP (D) beginning in column 1 (arrow). (F-H)
E1*-GFP expression is delayed to column 4-5 (indicated by white
bracket).
|
|

View larger version (38K):
[in this window]
[in a new window]
|
Fig. 8. Two-step selection of R8 by lateral inhibition and Rough. (A)
Overview of reporter fragments with expression pattern indicated to the right.
Intermediate groups (IG), R8s in column 1 (R81) and column 4
(R84), and ectopic expression in R2,5s. (B) Cartoon of Ato
(red), Sens (green) and co-expression (yellow) in ro mutants.
(C) Model of genetic interactions in the two-step selection of R8. Step
1 (right column): red colored circles represent Ato- and Sens-expressing cells
in IGs. (Top) Generation of a single R8 per ommatidium initially requires
selection of one R8 precursor from among the equipotent cells of the IG.
(Middle) In one cell within the IG, Ato and Sens are not repressed by lateral
inhibition and become the R8 precursor (yellow border). (Bottom) In cells not
destined to adopt the R8 cell fate (red with black border) lateral inhibition
represses neuronal cell fate and the expression of Ato and Sens. Step 2 (left
column): yellow circles represent developing R8s and blue circles represent
R2,5 precursor cells. (Top) By column 3, the selection event determined by
lateral inhibition must be reinforced by the sens-ro loop to maintain
the pattern of a single R8 per ommatidium and to specify the R2,5 cell fate.
(Middle) In the developing R8, Sens blocks R2,5 differentiation by repression
of Ro and locks in the R8 fate. (Bottom) In putative R2,5 cells, Ro is
expressed and directly represses Sens to block R8 and promote R2,5
differentiation.
|
|

CiteULike
Complore
Connotea
Del.icio.us
Digg
Reddit
Technorati
Twitter What's this?
© The Company of Biologists Ltd 2008