First published online August 18, 2003
doi: 10.1242/10.1242/dev.00661
Graded phenotypic response to partial and complete deficiency of a brain-specific transcript variant of the winged helix transcription factor RFX4
Perry J. Blackshear1,2,4,5,
Joan P. Graves3,
Deborah J. Stumpo2,
Inma Cobos6,
John L. R. Rubenstein6 and
Darryl C. Zeldin1,3,4,*
1 Office of Clinical Research, National Institute of Environmental Health
Sciences, Research Triangle Park, NC 27709, USA
2 Laboratory of Signal Transduction, National Institute of Environmental Health
Sciences, Research Triangle Park, NC 27709, USA
3 Laboratory of Respiratory Biology, National Institute of Environmental Health
Sciences, Research Triangle Park, NC 27709, USA
4 Department of Medicine, Duke University Medical Center, Durham, NC 27710,
USA
5 Department of Biochemistry, Duke University Medical Center, Durham, NC 27710,
USA
6 Nina Ireland Laboratory of Developmental Neurobiology, Department of
Psychiatry, University of California, San Francisco, San Francisco, CA 94143,
USA
View larger version (86K):
[in a new window]
|
Fig. 1. Hydrocephalus in adult transgenic mice. (A) Two mice in lateral (top) and
frontal (bottom) view at about 2 months of age, showing the characteristic
domed head and lateral displacement of the ears in the transgenic mouse
compared with its wild-type littermate. (B) Parasagittal sections, stained
with Hematoxylin and Eosin, of brains from four littermate mice, three
transgenic and one wild type, at about seven weeks of age. The marked
dilatation of the lateral ventricles (LV) is obvious in the transgenic mice;
however, there is no evidence for dilatation of the fourth ventricles
(arrows). Scale bar: 1 mm.
|
|
View larger version (158K):
[in a new window]
|
Fig. 2. Hydrocephalus in newborn transgenic mice. Serial rostral (r) to caudal (c)
coronal sections, stained with Hematoxylin and Eosin, from newborn (P0.5)
transgenic and wild-type littermates are shown, with each pair of sections
representing approximately the same coronal plane. Note the extreme
hydrocephalus apparent in the olfactory ventricles (OV) and the lateral
ventricles (LV) of the transgenic compared with the wild-type mouse. In the
more posterior sections, note the similar appearance of the aqueduct of
Sylvius (aq) and the fourth ventricle (fv) in the wild-type and transgenic
mice.
|
|
View larger version (118K):
[in a new window]
|
Fig. 3. Aqueduct of Sylvius and SCO in wild-type and transgenic mice. (A) Coronal
sections in a rostral (r) to caudal (c) direction from P0.5 wild-type and
transgenic littermates stained with Hematoxylin and Eosin, demonstrating the
apparent absence of the SCO in the transgenic mouse. (B) Similar sections
stained with an antibody to Reissner's fibers. Note the near-absence of
antibody staining in the transgenic section (top) compared with the wild-type
section (bottom). The arrow in the top section indicates a small amount of
antibody staining in one section from the knockout mouse, indicating the
presence of the Reissner's fiber antigen. The counterstain was Hematoxylin.
Scale bar in B: 50 µm (bottom); 20 µm (top).
|
|
View larger version (54K):
[in a new window]
|
Fig. 4. Identification of transgene insertion site. (A) A Southern blot of genomic
DNA from wild-type and transgenic mice, digested with the three restriction
enzymes indicated and probed with a 3'-insertion site-specific probe.
The arrows indicate the three single, novel bands hybridizing to the probe in
the DNA from the transgenic mice, indicating the likelihood of a single
transgene insertion site. (B) A PCR-based analysis of genomic DNA from one
litter of interbred transgenic mice, indicating the PCR products that were
specific for the presence of the transgene (Transgene-specific) and those that
were specific for the endogenous sequence that was interrupted by the
transgene (Insertion site-specific). The transgene specific primers were
5'-AGCCAGTAATAAGAACTGCAGA-3' and
5'-GGCACTCTTAGCAAACCTCAGG-3', which correspond to bp 264-285 of
the human cytochrome P450 cDNA clone accession number NM_000775.2 and bp
5225-5246 of the mouse  -myosin heavy chain promoter clone Accession
Number MMU71441, respectively. The insertion site specific primers were
5'-CATGGAAAGGGCAGAGTGAGC-3' and
5'-GGCCATTGTCACCACTCGTAA-3', which correspond to bp 732-752 and bp
323-343 of mouse trace archive sequence gnl|ti|91911671, respectively. In both
cases, the results were confirmed by PCR using different pairs of primers. The
DNA is characterized as +/+, +/- and -/- by the presence of the interrupted
allele. (C) A northern blot of total brain RNA from newborn mice of the +/+,
+/- and -/- genotypes. This blot was probed with a mouse EST clone that was
94% identical over 284 bases to a region corresponding to the 3'-end of
the human testis-specific RFX4 transcript H10145. The only visible transcript
was of  4 kb (RFX4_v3); this was decreased in expression in the +/-
sample, and undetectable in the -/- sample. Longer exposure of the blot did
not reveal the presence of any truncated mRNA species in the +/- and -/-
lanes. The same blot was hybridized to an actin cDNA (lower panel), and
demonstrates roughly equivalent loading of the three RNA samples. (D) The
hybridization of the same probe to adult mouse tissues, revealing an 4 kb
transcript in brain (RFX4_v3), a 3.7 kb transcript in testis and a still
smaller transcript in liver. (E) The pattern of developmental expression of
the 4 kb transcript, which was undetectable in whole embryos at E7.5, highly
expressed in whole embryos at E9.5 and 10.5, and less well expressed at E13.5
and 14.5. The brain, liver and testis lanes from D are juxtaposed in E to
illustrate the difference in size between the brain (RFX4_v3), liver and
testis transcripts, and the size identity of the adult brain transcript and
the embryonic transcript. Also shown is the expression of a control mRNA for
cyclophillin (Cyclo.).
|
|
View larger version (72K):
[in a new window]
|
Fig. 5. Alignment of mouse, human and zebrafish RFX4_v3. The predicted protein
sequences from these three RFX4_v3 orthologs were aligned using ClustalW. The
position of the characteristic RFX DNA binding domain (DBD) is indicated by
the box; other boxes contain the B and C boxes, and the dimerization domain
(DD). The shaded first 14 amino acids labeled exon 1 were unique to RFX4_v3
(human); the next unshaded sequences represent exons 2-5 and are identical to
sequences from RFX4_v2; the next shaded sequences represent exons 6-15 and are
identical to sequences from both RFX4_v1 and RFX4_v2; the next unshaded
sequences represent exons 16-18 and are identical to sequences in RFX4_v1.
Asterisks indicate amino acid identity; double dots indicate a high degree of
amino acid similarity; single dots indicate less similarity.
|
|
View larger version (24K):
[in a new window]
|
Fig. 6. The human RFX4 locus and its three known transcripts. This figure is a
schematic representation of 200 kb of human genomic sequence from NT_009720.8,
shown in reverse complement orientation, and of the position within this
sequence of the exons that comprise the three indicated RFX4 transcripts. At
the top of the figure is shown the transcript corresponding to RFX4_v2
(Accession Number NM_002920). Exon 1 is unique to this transcript (green
hatching); exons 2-5 (solid green) are shared with the novel RFX4_v3
transcript described in this paper; exons 6-15A (yellow) are shared with the
RFX4_v3 transcript as well as the transcript RFX4_v1; exon 15B (green
hatching) is apparently unique to this transcript, and contains a
polyadenylation sequence and presumably a polyA tail, as indicated by the wavy
line. The location of these exons on the genomic sequence are indicated. Below
the genomic sequence is represented the transcript RFX4_v1. It contains a
unique exon 1 (red hatching); exons 2-11 (yellow) shared with both RFX4_v2 and
RFX4_v3; and exons 12-14 (red) shared only with RFX4_v3. The RFX4_v3
transcript contains a unique exon 1 (purple); exons 2-5 (green) shared only
with RFX4_v2; exons 6-15 (yellow) shared with both RFX4_v1 and RFX4_v2; and
exons 16-18 (red) shared only with RFX4_v1. The site of transgene insertion is
indicated in the genomic clone by the black X in the intron between exons 13
and 14 of RFX4_v1; its position between exons 17 and 18 of RFX4_v3 is also
indicated. The regions of the RFX4_v3 transcript coding for the 737 amino acid
human RFX4_v3 protein (blue) are indicated, as is the DNA-binding domain (DBD)
of the protein. See the text for additional details.
|
|
View larger version (26K):
[in a new window]
|
Fig. 7. Northern analysis of RFX4 transcript expression using transcript-specific
probes. cDNA probes corresponding to multiple or single RFX4 transcript
variants, as indicated on the bottom of the figure, were used to probe
northern blots containing total cellular RNA from adult testes (T), liver (L)
and brain (B), or from brains of E18 mice of the +/+, +/- and -/- genotypes,
as indicated. The blots were aligned to demonstrate the positions of the three
hybridizing RFX4 species v1, v2 and v3 (arrows), as well as an uncharacterized
transcript seen in adult mouse liver. There was no detectable hybridization of
the specific v1 and v2 probes to the E18 brain RNA of any genotype (not
shown).
|
|
View larger version (64K):
[in a new window]
|
Fig. 8. Developmental expression of RFX4_v3. (A-E) Whole-mount embryos at the
indicated embryonic days (E) in which the RFX4_v3 transcript is indicated by
the blue digoxigenin staining. (C) Whole-mount suggests minimal staining
rostral of the zona limitans (zl); however, a section through the plane
indicated as C' shows staining of the dorsal cortex (cx). (D,E)
Wholemounts at E10.5; (F) midline sagittal section; (G-I) coronal sections
through similar embryos. The arrowheads in F-H indicate the lost expression in
the telencephalic dorsal midline at E10.5. (J-M) One sagittal (J) and three
rostral-to-caudal coronal sections through the head at E12.5. Note the lack of
staining in the telencephalic dorsal midline (arrowheads in J,K) in the
epiphysis (ep) in L and in the fourth ventricle choroid plexus (ch) in M.
Scale bars: 500 µm. mb, midbrain; fb, forebrain; hb, hindbrain; te,
telencephalon; me, mesencephalon; rb, rhombencephalon; sc, spinal cord di,
diencephalon; cb, cerebellum; cp/lt, commissural plate/lamina terminalis; LGE,
lateral ganglionic eminence; MGE, median ganglionic eminence; ch, choroid
plexus; R, retina; os, optic stalk; DT, dorsal thalamus; VT, ventral thalamus;
HY, hypothalamus; V, trigeminal ganglion; VII/VIII, facial/vestibular
ganglion.
|
|
View larger version (93K):
[in a new window]
|
Fig. 9. RFX4_v3 in situ staining in the region of the developing SCO. (A-D)
Progressively rostral-to-caudal sections through the brain of a normal embryo
at E16.5. The box labeled F in section C contains the SCO and the aqueduct of
Sylvius. This region is shown enlarged at E14.5 (E), E16.5 (F) and at the time
of birth (P0; G). Note the high level expression of the RFX4_v3 transcript in
the region of the developing SCO in E, and in the SCO itself in F,G. Scale
bars: 500 µm for A-D; 100 µm for E-G. cx, cerebral cortex; ch, choroid
plexus; DT, dorsal thalamus; HY, hypothalamus; me, mesencephalon; cb,
cerebellum; P, pituitary; ST, striatum; ep, epiphysis.
|
|
View larger version (67K):
[in a new window]
|
Fig. 10. Head morphology from -/- mice at E12.5. (A) Heads from two E12.5
littermates after fixation, one hemizygous (he) and one knockout (ko) as
indicated. Note the near normal appearance of the eyes and the facial
structures, but the clearly abnormal doming of the skull and the smaller head
of the -/- littermate. (B) Coronal sections from wild-type (top row) and
knockout (bottom row) littermate mice at E12.5, stained with Hematoxylin and
Eosin. In the most rostral (r) sections (left panels), the brains appear
somewhat similar, showing both lateral ventricles (LV) and apparently normal
midline structures, although the brains were somewhat smaller in the knockout
mice. In more caudal (c) sections (middle panels), however, there was a
striking loss of midline structures and the formation of a single central
ventricle. In still more caudal sections (right panels), taken at the level of
the retinas, there were continued striking abnormalities and loss of
essentially all dorsal midline structures in the knockout mice. ihf,
interhemispheric fissure; cing. cortex, cingulate cortex; gang. em.,
ganglionic eminence; pc, posterior commissure; epithal., epithalamus; hip.,
hippocampus; hypothal., hypothalamus. See the text for further details.
|
|
View larger version (118K):
[in a new window]
|
Fig. 11. Expression of molecular markers in wild-type and littermate knockout mice
at E12.5. Shown are the in situ hybridization staining patterns of sagittal
(A-C) and coronal sections through wild-type (+/- or +/+) and knockout (-/-)
heads at E12.5. The blue digoxigenin staining indicates the presence of the
specific transcript being evaluated. Note that Fgf8 expression is
maintained in the isthmus, infundibulum, lamina terminalis and septum, but is
lost in the choroid plexus of the forebrain (C,C'). The asterisks in
D' and E' indicate the decrease in Msx2 expression
(D,D') and the lack of Wnt3a expression (E,E') in the
dorsal midline of the knockout embryos. Scale bars: 500 µm. LGE, lateral
ganglionic eminence; MGE, median ganglionic eminence; ch, choroidal plexus;
cx, cerebral cortex; ep, epiphysis; IN, infundibulum; lt, lamina terminalis;
is, isthmus; hem, cortical hem; DT, dorsal thalamus; se, septum; me,
mesencephalon; cb, cerebellum.
|
|
© The Company of Biologists Ltd 2003
