The prospero gene encodes a divergent homeodomain protein that controls neuronal identity in Drosophila
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Development Supplement 2, 1991, 79-85 79
Printed in Great Britain © The Company of Biologists Limited 1991
The prospero gene encodes a divergent homeodomain protein that
controls neuronal identity in Drosophila
QUYNH CHU-LAGRAFF, DOROTHY M. WRIGHT, LESLIE KLIS McNEIL and CHRIS Q. DOE
Department of Cell and Structural Biology, University of Illinois, Urbana, Illinois 61801, USA
Summary
The Drosophila central nervous system (CNS) develops in the specification of neuronal fate (Doe et al. 1991).
from a population of stem cells called neuroblasts; each Here we show that the pros gene encodes a highly
neuroblast goes through an invariant cell lineage to divergent homeodomain. The homeodomain contains
produce a characteristic family of neurons or glia. We several of the most conserved amino acids characteristic
are interested in the molecular mechanisms controlling of known homeodomains, yet it is considerably less basic
neuroblast cell lineage. Recently we identified the than previously identified homeodomains. These data
prospero (pros) gene, which is expressed in embryonic are consistent with a model in which pros controls
neuroblasts. Loss of pros function results in aberrant neuroblast cell lineages by regulating gene expression.
expression of the homeobox genes fushi tarazu, even-
skipped and engrailed in a subset of neuroblast progeny, Key words: Drosophila, CNS, neuroblast, fate, lineage,
suggesting that pros plays an early and fundamental role homeodomain.
Introduction prospero (pros), a neuroblast identity gene (Doe et al.
1991). pros is expressed in a subset of neuroblasts and
We are interested in the molecular mechanisms newly born GMCs, but not in mature neurons.
controlling cell fate in the Drosophila central nervous Embryos lacking pros function (pros embryos) have the
system (CNS). Neurogenesis begins in the ventral normal number of neuroblasts, but the neuroblasts
neurogenic region as 25 cells per hemisegment delami- generate aberrant cell lineages. In particular, ex-
nate into the embryo to form a subepidermal array of pression of the homeobox genes fushi tarazu (ftz), even-
neuronal precursor cells (called neuroblasts); the skipped (eve) and engrailed (en) is defective in a subset
remaining superficial cells differentiate into epidermis of GMCs, and their neuronal progeny show striking
(Doe et al. 1988a; Jimenez and Campos-Ortega, 1990). alterations in axon morphology. In this paper we review
Subsequently, this two-dimensional neuroblast layer is the expression and function of pros in developing CNS
transformed into a three-dimensional CNS as each and peripheral nervous system (PNS), and describe the
neuroblast goes through an invariant stem cell lineage, initial molecular characterization of the pros gene.
'budding off smaller ganglion mother cells (GMCs)
into the embryo. Each GMC divides to produce a
characteristic pair of neurons (Thomas et al. 1984).
We have proposed that there are three classes of Results and discussion
genes controlling cell fate in the developing CNS
(Fig. 1) (Doe et al. 1991; Doe, 1992). First, regionally prospero controls the specification of cell fate in the
expressed genes provide positional cues within the CNS
neuroectoderm. Second, in response to these positional To examine the role of pros in the developing CNS we
cues, 'neuroblast identity' genes are expressed in used a variety of cell-specific markers to identify
overlapping subsets of neuroblasts and their progeny. individual neuroblasts, GMCs, and neurons in wild-
Each neuroblast expresses a specific combination of type and pros embryos (Fig. 2). pros embryos show no
these genes, which leads to the initiation of a unique, obvious change in the position or number of neuro-
invariant neuroblast cell lineage. Third, 'GMC identity' blasts per segment. However, there are striking defects
genes are expressed in subsets of GMCs in response to in neuroblast cell lineages, as detected by altered
the neuroblast identity genes; these genes control the expression of cell-specific markers in a subset of
identity of individual GMCs. Examples of 'positional neuroblast progeny.
cue' and 'GMC identity' genes are known (Patel et al. The earliest defect in pros embryos is a change in
1989; Doe et al. 1988a,b). We recently identified GMC gene expression. Normally ftz is expressed in80 Q. Chu-LaGraff and others
POSITIONAL position-specific
CUES neuroblast specification
NEUROBLAST translate positional cues
IDENTITY into neuroblast cell lineage
GENES
GMC
control GMC and
IDENTITY neuronal specification
GENES
Fig. 1. A model for a gene regulatory hierarchy controlling the specification of cell fate in the embryonic CNS. Top:
regionally expressed genes (light and dark stipple) are expressed in overlapping patterns in the ectoderm of the neurogenic
region, such that clusters of 4-6 cells express a unique combination or concentration of gene products (black circles).
Middle: Neuroblast identity genes are expressed in subsets of neuroblasts in response to the positional cues (diagonal
hatching). Each neuroblast identity gene may be expressed in a subset of GMCs produced from a neuroblast; the pattern
of GMC expression can be in lineally related GMCs, positionally related GMCs, or lineage-specific GMCs. The function of
the neuroblast identity genes is to translate positional cues into a specific neuroblast lineage, i.e. the number of cell cycles
and the fate of the GMCs produced. Bottom: GMC identity genes (horizontal hatching) are expressed in subsets of GMCs,
but not neuroblasts, in response to the particular combination of neuroblast identity genes present in a GMC. These genes
are required for the establishment of unique GMC and neuronal fates. Examples are ftz and eve (Doe et al. I988a,b).
about 20 GMCs per hemisegment (which generate 40 PNS, as well as the individual axon morphology of the
neurons, including the identified aCC and pCC neur- aCC and pCC neurons. The wild-type CNS has
ons). In pros embryos ftz is expressed in 10-15 GMCs bilaterally paired longitudinal connectives and two
and about 20-30 neurons; it is not expressed in the aCC commissural tracts that cross the midline in every
and pCC neurons (Fig. 2). Expression of eve is also segment, and in each hemisegment there are two nerves
abnormal in pros embryos. In wild-type embryos eve is connecting the CNS and PNS (Fig. 3). In pros embryos,
expressed in 10 GMCs per hemisegment, which the longitudinal connectives fail to form, and the
produce 19 eve-positive neurons, including the aCC and commissures are reduced and fused (Fig. 3) (Doe et al.
pCC neurons. In pros embryos only 5 GMCs (and 10 1991).
neurons) express eve; these do not include the aCC and To determine how individual neurons develop in pros
pCC neurons (Fig. 2). Mature aCC and pCC neurons embryos, the aCC and pCC neurons were injected with
can be identified in pros embryos, proving that these a fluorescent dye. This experiment also serves to verify
neurons do not die in pros embryos; we don't know how the differentiation of these neurons despite their lack of
many other ftz~ and eve~ neurons also differentiate. ftz and eve expression. The aCC cell normally extends
Embryos lacking pros show either reduced or ectopic its axon ipsilaterally into the PNS, pioneering the
expression of GMC markers: ftz and eve show reduced
intersegmental nerve; the pCC axon normally grows
expression, whereas en shows ectopic expression
anteriorly in a medial fascicle of the longitudinal
(Fig. 2). Thus, pros is required for the correct cell
connective (Fig. 3). In pros embryos the 'aCC can be
lineage of several neuroblasts, including the identified
identified by its relatively large size and position as the
neuroblast that produces the aCC and pCC neurons.
dorsal-most neuron just posterior to the lateral commis-
sures. This 'aCC neuron has an abnormal and variable
axon morphology. The 'pCC neuron in pros embryos
prospero embryos have axon pathfinding defects in the has more consistent morphology: it grows anteriorly in
CNS and PNS a medial fascicle, as in wild-type, and in older embryos
It is known that absence of ftz and eve expression in it turns medially and crosses the midline (Fig. 3). These
identified GMCs results in abnormal axon pathfinding results show that the aCC and pCC neurons are born
by their neuronal progeny (Doe et al. 1988a,b). We and differentiate despite loss of pros, eve, and ftz
assayed the general axon morphology of the CNS and expression; they do, however, have an abnormal axonSpecification of neuronal precursor cell fates 81
Fig. 2. Loss of pros function
leads to aberrant gene
expression in GMCs and
neurons. (A,B) ftz expression
in wild-type and pros embryos
respectively. (A) Lateral view
of ftz expression in a wild-type
embryo at approx. 7.5 h of
development showing the aCC
and pCC neurons (arrow).
(B) aCC and pCC neurons
(arrow) do not express ftz in
pros embryos. Lateral view,
anterior is up. (C,D) eve
expression in wild-type and
pros embryos respectively.
(C) Ventral view of eve
expression in 13 h wild-type
CNS showing cluster of 10 EL
neurons (arrowhead) and
medial CQ neurons (arrow).
(D) Same view depicting
altered eve expression in 13 h
pros CNS. Note the lack of
eve-positive CQ neurons
(arrow), but normal expression
of EL neurons (arrowhead).
Occasionally there are a few
CQ neurons expressing eve
(asterisk-arrow). (E,F) en
expression in wild type and
pros embryos respectively.
(E) Dorsal view of en
expression in a wild-type 13 h
CNS showing L neurons (outer
arrows), ML neurons (inner
arrows), and the DM cells
(arrowhead). (F) Higher
magnification view of ectopic
en expression in pros CNS
showing en-positive neurons in
the location of the L and ML clusters (arrows), but with additional en-positive neurons in between. There is also an
increase in the number of en-positive DM cells (arrowhead) and absence of large VM en-positive neurons (out of plane of
focus).
morphology. It is unknown whether the aCC and pCC segmental nerve (Fig. 4; Table 1). In pros embryos the
axon defects are strictly due to loss of pros in their v and v' axons usually project normally. The d and 1
neuroblast or GMC precursor; some or all of the defects neurons, however, frequently extend in aberrant
could be due to lack of pros expression in neurons or directions. The d axons grow in virtually any direction,
glia necessary for proper aCC and pCC axon outgrowth including dorsally - opposite to their direction of
(Doe et al. 1991). We are currently making embryos growth in wild-type embryos. Even when the d axons
mosaic for pros CNS expression to resolve this extend ventrally, they usually do not fasciculate with
question. the 1 axons. The 1 neurons also extend axons in a
Sensory neurons in the PNS of pros embryos show disoriented fashion, with axon outgrowth ranging from
novel pathfinding defects. Loss of pros function can ventrally, the normal orientation, to dorsally, which is
result in a reversal of axon polarity, with sensory never seen in the abdominal segments of wild-type
neurons extending dorsally, away from the CNS embryos (Fig. 4; Table 1).
(Fig. 4). In wild-type embryos the neurons and non- There are several possible explanations for the pros
neuronal support cells of the PNS are arranged in 4 PNS phenotype. First, pros is expressed in many PNS
clusters: ventral (v), ventral' (v'), lateral (1) and dorsal precursors (see below). Loss of pros expression in
(d) (Ghysen et al. 1986). In wild-type embryos all neuronal precursors may lead to incorrectly specified
abdominal sensory neuron axons project ventrally into neurons which are unable to respond to normal
the CNS, with the d and 1 axons fasciculating with the pathfinding cues. Second, pros is expressed in the
intersegmental nerve and the v' and v axons joining the mature non-neuronal sheath cell of the lateral chordo-82 Q. Chu-LaGraff and others
prospero
NBM
/ \
\
•MUIU^.
/ \
Q o
aCC pCC "aCC"
J V.
or
"N
ft*
aCC DCC "aCC"
Fig. 3. Summary of pros expression and function in thi
CNS. Wild type: Neuroblast 1-1 (NB 1-1) and the first
GMC (GMC-1) express pros (vertical hatching); GMG
and its progeny (the aCC and pCC neurons) express b<
ftz and eve (horizontal hatching). Figure directly below
shows axon of wild-type aCC neuron extending out to
PNS, and pCC growing anteriorly, prospero: Loss of pi
expression in GMC 1-1 results in the loss of ftz and evi
expression and the abnormal development of aCC and
pCC neurons ('aCC and 'pC') and consequently, loss c
proper axon pathfinding (figure shown below). The
drawings are schematics of all the observed axon
morphologies.
Fig. 4. prospero is required for correct PNS axon
pathfinding. 14 h embryos showing a lateral view of the
tonal sensory organ, and in single non-neuronal cells in regular array of sensory neurons visualized with the SOXII
other sense organs (Doe et al. 1991). The absence of antibody; dorsal is up and anterior to the left. (A) Wild-
pros function from these support cells may affect the type embryo: the dorsal (d) and lateral (1) neurons extend
axon projection of the associated neuron. Third, it may axons ventrally to join with the intersegmental nerve (IS)
be that the neurons from the 1 and d sense organs exiting the CNS, while the ventral (v) and ventral' (v')
require the efferent intersegmental nerve from the CNS neurons extend axons ventrally in a separate fascicle that
joins with the segmental nerve (S) of the CNS. (B) pros
for correct pathfinding; this nerve is not present in pros embryo: the d and 1 neurons show striking defects in axon
embryos (Doe et al. 1991). morphology, often growing dorsally rather than ventrally;
The prospero gene encodes a predicted protein with a the intersegmental nerve does not form. The v and v'
neurons develop fairly normally and join with the
highly divergent homeodomain segmental nerve (S) of the CNS.
An enhancer trap insertion allele (pros139) provided the
molecular entry point for cloning the pros gene (Doe et
al. 1991). Two classes of pros cDNAs were isolated: common with known homeodomains, including the
prosS and prosL. The prosL predicted protein sequence four amino acids identical in all higher eukaryotic
has a number of recognizable motifs: a highly divergent homeodomains: tryptophan, phenylalanine, aspara-
homeodomain, several putative nuclear localization gine, and arginine at positions 49, 50, 52, and 54,
signals, and basic domains (Wright and Doe, unpub- respectively (Fig. 5). In addition, the pros homeo-
lished results). The prosL cDNA is identical to prosS domain has a predicted helix-turn-helix secondary
except prosS lacks 87 bp of coding sequence. structure resembling helix-turn-helix domains observed
Interestingly, the missing 87 bp includes part of the by NMR and crystallography (Kissinger et al. 1990;
prosL homeobox, resulting in the generation of a Qian et al. 1989). The pros homeodomains have several
shorter prosS protein with a different homeodomain features that are unusual, but not novel. First, most
(Fig. 5). homeodomains have 3-4 basic residues at both the N
Both pros homeodomains have a number of highly and C termini (Scott et al. 1989); the prosS protein lacks
conserved and functionally important amino acids in basic residues entirely from the first and last 7 aminoSpecification of neuronal precursor cell fates 83
Table 1. Axon pathfinding in the prospero PNS. The most highly conserved residues between pros
Percentage of normal axon connections between and other homeodomains are in the helix 3 region, a
sensory clusters d, I, v', and v within segments domain necessary for sequence-specific DNA binding.
T2 to A7 Both NMR and crystal structure studies demonstrate
that helix 3 fits into the major groove of the target
Segments
DNA, with amino acids making contacts with nucleo-
T2 T3 Al A2 A3 A4 A5 A6 A7 tides or parts of the phosphate backbone (Kissinger et
al. 1990; Qian et al. 1989). Furthermore, mutation
d-1 56 38 69 56 56 50 50 56 75
1-v' 50 50 56 69 56 31 69 81 69 experiments show that residue 9 of helix 3 confers DNA
v'-v 94 63 88 94 81 81 81 88 69 binding specificity (Hanes and Brent, 1989; Treisman et
al. 1989). The pros homeodomains contain a serine at
Most axons of the v' cluster successfully fasciculate with the v this position, similar to the paired-dass of homeo-
cluster of the PNS. All connections over 80% normal are shown
in bold. Most defects lie in the aberrant axonal growth of the 1
domains (Fig. 5).
clusters and d clusters. n = 16 for each segment.
prospero is expressed in precursors to the CNS and
acids of the homeodomain, and the prosL homeo- PNS
domain has only a single basic residue in these regions Due to the nature of pros mutations (aberrant
(Fig. 5). In this regard pros is similar to the yeast mat2- neuroblast cell lineages and incorrect GMC specifi-
P homeodomain, which lacks basic residues in both cation), pros expression is expected in a subset of
regions (Kelly et al. 1988), and the labial and hoxl.6 neuroblasts and GMCs. In fact, pros transcripts are
homeodomains, which have a single N-terminal basic observed in all but two of the neuroblasts in each
residue (Mlodzik et al. 1990; Baron etal. 1987). Second, hemisegment (Fig. 6). The two pros-negative cells are
the pros homeodomain is predicted to have 6 amino considered to be neuroblasts, based on morphology and
acids in the turn between helix 2 and 3, whereas most position in the neuroblast array. The transcript is clearly
homeodomains have a 3 amino acid turn. However, the restricted to neuroblasts; no signal is seen in the
transcription factor LF-B1 has a 21 residue loop neuroectoderm during the time of neuroblast formation
between helix 2 and 3, and the bacterial LexA protein (Fig. 6C). In addition, the transcript is observed in
has a 4 amino acid turn; both of these proteins bind many of the GMCs born early in neurogenesis
DNA (Nicosia et al. 1990; Lamerichs et al. 1989; Scott et (Fig. 6D). In general, the pros transcript is not
al. 1989). detectable in neurons, although it may be expressed in a
K N N
10 50
piosS QH APT
pioaL L U I S Y F p D i IK F | a | T | y « n v K|
MAI2P T V R G Q CRU C H K PF¥JM|R|W L Q l H Y D N P| N S E F Y D L S A A T G I . T R T F J ] R N|
pid K Q R R c R T H F I S A S Q Q D E L E R A Q B R T g| D I Y T R E E L A Q R T N L T EQR I Q V
en D E K R P R|I|A F|S|S E E NFJjL T E R R R Q Q L S S l E L G L,yi
EJE
Antp S R K R G R QQY T R R R R I E I A H A LC 1 T E R
I-POO I E It K . . R Q S I A A P E l K R l S L E A Y Q A Q Q P R S G E A I A E D LKKN V V
CONSERVED R - - Y Q L 1 R - L I- K - K
pros a-helice3
en a-helices
Fig. 5. Comparison of predicted pros homeodomains with yeast mat2-P, and Drosophila paired iprd), engrailed (en),
Antennapedia (Antp), and I-POU homeodomains (sequences taken from Scott et al. 1989). The prosS homeodomain is
identical to prosL except for the five N-terminal residues. The amino acid similarity between pros and the shown
homeodomains is: mat2-P (15 identities+1 conserved), prd (12+3), en (12+6), Antp (11+3), and I-POU (9+5). The pros
homeodomains share the most amino acid identities with the mat2-P homeodomain, with 28% identity overall and 70%
over 10 amino acids in helix 3. Both pros homeodomains have an extra three amino acids in the turn between helix 2 and
3, which have been excluded to maximize alignment. Black boxes indicate identical amino acids; open boxes indicate
conservative amino acid substitutions. Arrowheads indicate the four amino acids present in all higher eukaryotic
homeodomains. The a-helical domains of the en homeodomain are indicated by an open bar (Kissinger et al. 1990); the
pros (T-helical regions predicted by the Robson-Garnier algorithm are indicated by a black bar.84 Q. Chu-LaGraff and others
small number of neurons or transiently in young References
neurons (Fig. 6E).
Development of the PNS is highly stereotyped, with BARON, A., FEATHERSTONE, M., HILL, R., HALL, A., GALLIOT, B.
each precursor, called a sensory mother cell (SMC), AND DUBOULE, D. (1987). Hox-1.6: A mouse homeo-box
containing gene member of the Hox-1 complex. EMBO J. 6,
forming at a specific time and position (Ghysen and 2977-2986.
Dambly-Chaudiere, 1989). Every SMC goes through an BASTIAN, H. AND GRUSS, P. (1990). A murine even-skipped
invariant cell lineage to produce the neuron and non- homologue, evx 1, is expressed during early embryogenesis and
neuronal cells of a sensory organ (Bodmer et al. 1989). neurogenesis in a biphasic manner. EMBO J. 9, 1839-1852.
pros expression is first detected in a single PNS cell per BODMER, R., CARRETTO, R. AND JAN, Y. N. (1989). Neurogenesis
of the peripheral nervous system in Drosophila embryos: DNA
hemisegment; based on its position and time of replication patterns and cell lineages. Neuron 3, 21-32.
development it may be a SMC (Doe et al. 1991). In DAVIS, C. A., NOBLE-TOPHAM, S. E., ROSSANT, J. AND JOYNER, A.
mature sense organs, at about 12 h of development, L. (1988). Expression of the homeobox-containing gene En-2
pros is expressed in a single non-neuronal cell in every delineates a specific region of the developing mouse brain.
sense organ (Fig. 7). This is clearly resolved in the Genes Dev. 2, 361-371.
lateral chordotonal sense organ, where high level pros DOE, C. Q. (1992). The generation of neuronal diversity in the
Drosophila embryonic central nervous system. In Determinants
expression is restricted to the non-neuronal sheath cell of Neuronal Identity (ed. M. Shankland and E. Macagno).
(Doe et al. 1991). California: Academic Press.
DOE, C. Q., CHU-LAGRAFF, Q., WRIGHT, D. M. AND SCOTT, M. P.
(1991). The prospero gene specifies cell fates in the Drosophila
central nervous system. Cell 65, 451-464.
Conclusions DOE, C. Q., HIROMI, Y., GEHRING, W. J. AND GOODMAN, C. S.
(1988a). Expression and function of the segmentation gene fushi
pros is necessary for the correct specification of cell fate tarazu during Drosophila neurogenesis. Science 239, 170-175.
DOE, C. Q. AND SCOTT, M. P. (1988). Segmentation and homeotic
in the CNS. We are interested in working out the gene function in the developing nervous system of Drosophila.
mechanism by which pros acts in concert with other Trends Neurosci. 11, 101-106.
genes to control the fate of individual GMCs. There are DOE, C. Q., SMOUSE, D. AND GOODMAN, C. S. (19886). Control of
several genes that may play a similar role: both neuronal fate by the Drosophila segmentation gene even-
polyhomeotic and runt are required for the normal skipped. Nature 333, 376-378.
expression of eve, ftz or en in subsets of GMCs (Smouse GHYSEN, A. AND DAMBLY-CHAUDIERE, C. (1989). Genesis of the
Drosophila peripheral nervous system. Trends Genet. 5,
etal. 1988; H. Brock, personal communication; J. Duffy 251-255.
and P. Gergen, personal communication). We expect GHYSEN, A., DAMBLY-CHAUDIERE, C , ACEVES, E., JAN, L. Y. AND
that the mechanisms controlling the neuronal determi- JAN, Y. N. (1986). Sensory neurones and peripheral pathways in
nation will be conserved among distantly related Drosophila embryos. Roux Arch. Devi Biol. 195, 281-289.
organisms. Early CNS development is virtually ident- HANES, S. D. AND BRENT, R. (1989). DNA Specificity of the
bicoid activator protein is determined by homeodomain
ical among arthropods (Patel et al. 1989; Thomas et al. recognition helix residue 9. Cell 57, 1275-1283.
1984). Might fundamental similarities extend to more JIMENEZ, F. AND CAMPOS-ORTEGA, J. (1990). Defective neuroblast
distantly related organisms, such as vertebrates? Genes commitment in mutants of the achaete-scute complex and
known to play a role in CNS development are also adjacent genes of D. melanogaster. Neuron 5, 81-89.
conserved between Drosophila and vertebrates. Re- JOHNSON, J. E., BIRREN, S. J. AND ANDERSON, D. J. (1990). Two
rat homologues of Drosophila achaete-scute specifically expressed
cently, PCR (polymerase chain reaction) has been used in neuronal precursors. Nature 346, 858-861.
to identify two rat homologs of the Drosophila KELLY, M., BURKE, J., SMITH, M., KLAR, A. AND BEACH, D.
'proneural' achaete-scute genes; as in Drosophila, they (1988). Four mating-type genes control sexual differentiation in
are expressed in a subset of CNS and PNS precursors the fission yeast. EMBO J. 7, 1537-1547.
during early neurogenesis (Johnson and Hirsh, 1990). KISSINGER, C. R., LIU, B., MARTIN, B. E., KORNBERG, T. B. AND
PABO, C. O. (1990). Crystal structure of an engrailed
Vertebrate homologs to the Drosophila pair-rule gene homeodomain-DNA complex at 2.8 A resolution: a framework
eve and the segment polarity genes wingless and en have for understanding homeodomain-DNA interactions. Cell 63,
been identified. Each of these genes is expressed in the 579-590.
developing CNS (Wilkinson et al. 1987; Davis et al. LAMERICHS, R. M., PADILLA, A., BOELENS, R., KAPTEIN, R.,
1988; Bastian and Gruss, 1990), and the mouse wingless OTTEBEN, G., RUTERJANS, H., GRANGER-SCHNARR, M., OERTEL,
P. AND SCHNARR, M. (1989). The N-terminal domain of LexA
homolog is necessary for the correct development of a repressor is a-helical but differs from canonical helix-turn-helix
large portion of the mouse brain (McMahon and proteins. Proc. natn. Acad. Sci. U.S.A. 86, 6863-6867.
Bradley, 1990). In addition, homeotic genes specify MCMAHON, A. P. AND BRADLEY, A. (1990). The Wnt-1 (int-1)
segment-specific differences in the CNS of both proto-oncogene is required for development of a large region of
Drosophila and vertebrates (reviewed in Doe and the mouse brain. Cell 62, 1073-1085.
MLODZIK, M., HIROMI, Y., WEBER, U., GOODMAN, C. S. AND
Scott, 1988). These results suggest that characterization RUBIN, G. M. (1990). The Drosophila seven-up gene, a member
of Drosophila CNS development will provide insight of the steroid receptor gene superfamily, controls photoreceptor
into the mechanisms controlling neuronal diversity in cell fates. Cell 60, 211-224.
both flies and vertebrates. NICOSIA, A., MONACI, P., TOMEI, L., D E FRANCESCO, R., NUZZO,
M., STUNNENBERG, H. AND CORTESE, R. (1990). A myosin-like
dimerization helix and an extra-large homeodomain are essential
This research was supported by the NIH, the Searle elements of the tripartite DNA binding structure of LFB1. Cell
Scholars Program, and an NSF Presidential Young Investi- 61, 1225-1236.
gator Award. PATEL, N. H., SCHAFER, B., GOODMAN, C. S. AND HOLMGREN, R.Fig. 6. prospero transcript is expressed in
neuroblasts and GMCs but not in neurons.
(A) Enhancer trap detection showing lacZ
expression in a pattern matching that of the
prospero gene in an approx. 6h embryo (ventral
view of about 2.5 segments; compare with panel
B). Arrowheads indicate neuroblasts that do not
express lacZ. (B) Expression of the pros transcript
in most neuroblasts (in an embryo similar in age to
that in panel A). The majority of neuroblasts
express pros (arrows), but two per hemisegment do
not (arrowheads). On the basis of size, morphology
and position in the neuroblast array, these two
pros-negative cells are neuroblasts; neither is the
glioblast. (C,D,E) Neuroblast and GMC transcript
expression shown in optical cross-section, detected
with HRP reaction product. The ventral surfaces of
the embryos are towards the bottom of the photos.
(C) pros expression in a newly formed neuroblast
(arrow); GMCs have not yet been born. (D) pros
expression in newly born GMCs (arrowhead) is
often stronger than neuroblast expression (arrow).
This photo also illustrates the lack of pros
expression in the ventral ectoderm from which the
neuroblasts develop (below arrow). (E) Neuroblast
expression continues in approx. 7.5 h embryos
(arrows), but no neuronal expression (n) is
detected; the cell on the dorsal surface expressing
pros is one of the longitudinal glia (g). pros is not
expressed in the epidermis (e).
B
Fig. 7. pros expression in the PNS. (A) Enhancer trap detection showing lacZ expression the PNS of an approx. 10 h wild-
type embryo. Expression in the sheath cell of the lateral chordotonal sense organs is indicated (segment T3, wide arrow;
segment Al, thin arrow). (B) pros transcript pattern matching the above lacZ pattern in same age embryo; note the
similarity in the chordotonal expression patterns (segment T3, wide arrow; segment Al, thin arrow). Anterior is to the left,
dorsal to the top in both panels.Specification of neuronal precursor cell fates 85
(1989). The role of segment polarity genes during Drosophila TREISMAN, J., GONCZY, P., VASHISHTHA, M., HARRIS, E. AND
neurogenesis. Genes Dev. 3, 890-904. DESPLAN, C. (1989). A single amino acid can determine the
QIAN, Y. Q., BILLETER, M., OTTING, G., MULLER, M., GEHRING, DNA binding specificity of homeodomain proteins. Cell 59,
W. AND WUTHRICH, K. (1989). The structure of the 553-562.
Antennapedia homeodomain determined by NMR spectroscopy WILKINSON, D. G., BAILES, J. A. AND MCMAHON, A. P. (1987).
in solution: comparison with prokaryotic repressors. Cell 59, Expression of the proto-oncogene int-1 is restricted to specific
573-580. neural cells in the developing mouse embryo. Cell 50, 79-88.
SCOTT, M. P., TAMPKUN, J. W. AND HARTZELL, G. W. (1989). The
structure and function of the homeodomain. Biochim. Biophys.
Ada 989, 25-48.
SMOUSE, D . , GOODMAN, C. S., MAHOWALD, A. P. AND PERRIMON,
N. (1988). Polyhomeolic: a gene required for the embryonic Note added in proof
development of axon pathways in the central nervous system of Results from RNase protection assays are consistent
Drosophila. Genes Dev. 2, 830-842.
THOMAS, J. B . , BASTIANI, M. J., BATE, M. AND GOODMAN, C. S.
with embryonic expression of both prosS and prosL
(1984). From grasshopper to Drosophila: a common plan for (L.K.M. and C.Q.D.), but it is not known whether both
neuronal development. Nature 310, 203-207. transcripts are translated.You can also read
