Experimental Cell Research 262, 134 –144 (2001) doi:10.1006/excr.2000.5086, available online at http://www.idealibrary.com on
Drosophila Embryos Lacking N-Myristoyltransferase Have Multiple Developmental Defects Monde Ntwasa,* Sonti Aapies,* David A. Schiffmann,† and Nicholas J. Gay† ,1 †Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1GA, United Kingdom; and *Department of Molecular and Cellular Biology, University of the Witswatersrand, 2050 Johannesburg, South Africa
Lipid modification of proteins by the addition of myristic acid to the N-terminal is important in a number of critical cellular processes, for example, signal transduction and the modulation of membrane association by myristoyl switches. Myristic acid is added to proteins by the enzyme N-myristoyltransferase (NMT) and in this paper we detail the effects on embryonic development of a null mutation in the Drosophila NMT gene. Mutant embryos display a range of phenotypes, including failures of head involution, dorsal closure, and germ-band retraction, morphogenetic processes that require cellular movements. Embryos with milder phenotypes have more specific defects in the central nervous system, including thinning of the ventral nerve chord and, in some embryos, specific scission at parasegment 10. Staining of mutant embryos with phalloidin shows that the mutant embryos have a disrupted actin cytoskeleton and abnormal cell morphology. These phenotypes are strikingly similar to those caused by genes involved in dynamic rearrangement of the actin cytoskeleton. For example the myristoylated nonreceptor tyrosine kinases Dsrc42A and Dsrc64B were shown recently to be key regulators of dorsal closure. In addition, analysis of cell death reveals widespread ectopic apoptosis. Our findings are consistent with the hypothesis that the myristoyl switches and signaling pathways characterized at the biochemical level have important functions in fundamental morphogenetic processes. © 2001 Academic Press Key Words: lipid modification; myristic acid; embryonic development; actin cytoskeleton.
INTRODUCTION
The enzyme N-myristoyltransferase (NMT) catalyzes the addition of the fatty acid myristic acid to an N-terminal glycine residue of a small subset of cellular proteins [1]. This modification allows the proteins to associate with the cytoplasmic surface of cellular membranes, and lipid addition is required for the function or regulation of the substrate proteins. NMT sub1
To whom correspondence and reprint requests should be addressed. Fax: (01223) 766002. E-mail:
[email protected]. 0014-4827/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
strates are involved in critical cellular processes, for example, signal transduction pathways, vesicular trafficking, cell survival, and regulation of cytoskeletal assembly. Biochemical studies of myristoyl proteins have shown that the modification is important for modulating protein–membrane interactions. Stable association of myristoyl proteins with the membrane seems to require additional interaction, for example, electrostatic contacts with the negatively charged head groups of the phospholipids. Three specific mechanisms for reversible protein–membrane interactions have been described: the calcium, GTP, and electrostatic myristoyl switches [2– 4]. An example of the latter kind of switch mechanism involves MARCKS, the myristoylated alanine-rich protein kinase C substrate. MARCKS associates stably with membranes by virtue of electrostatic interaction involving basic residues in the protein close to the lipid moiety, and phosphorylation of nearby serine residues by PKC causes the protein to dissociate. Recent work shows that reversible binding of MARCKS regulates membrane ruffling and cell adhesion, processes which cause cell shape changes and require regulation of the actin cytoskeleton [5]. Thus, cytoskeletal rearrangement can be coupled to cellular signal transduction and this requires reversible binding of myristoyl proteins to cytoplasmic membranes. In the Drosophila embryo, many specific developmental processes that require cell shape changes and movement are driven by programmed rearrangements of the actin cytoskeleton. Embryos that have mutations in genes required for this process have common phenotypes, for example, failures of head involution and dorsal closure. These genes include zipper, the Drosophila homologue of nonmuscle myosin II, which is required for cell shape changes and is the primary motor protein for actin contraction [6]. In addition two other genes (ribbon and raw) are indirectly involved in reorganization of the actin cytoskeleton mediated by nonmuscle myosin [7, 8]. In this paper we describe the characteristics of a null mutation in the single NMT gene of Drosophila. Embryos lacking NMT have a spectrum of phenotypes
134
135
NMT IN DEVELOPMENT
which are probably caused by failure of morphogenetic cellular movements and abnormal apoptotic cell death. This suggests that myristoyl switches play a critical role in driving the developmental processes that involve cell shape changes and movement. MATERIALS AND METHODS Database. The Berkeley Drosophila Genome Project (BDGP) (http://fruitfly.berkeley.edu) was used to identify EST clots of NMT cDNAs and lethal P-element insertion lines. 5⬘ RACE and polymerase chain reaction. A 0- to 4-h embryonic cDNA library cloned in pNB40 [9] was used as template in 5⬘ RACE amplification. The following oligonucleotide primers were used: 5⬘ GGATCCTGCAGTACGAGAGAGGATGATAGG 3⬘ (5⬘ of the cloning site in pNB40) and 5⬘ GTTTTTCTCGAGTCAGAGCAATCTCTCCGGCTG 3⬘ (reverse primer in NMT sequence, position 2865 in [10]). The cDNA library (1.5 g) was denatured for 2 min at 94°C and amplified over 40 cycles as follows: denature 15 s at 94°C, anneal 30 s at 60°C, extend 1 min at 68°C. Fragments of the anticipated size were cloned into pGEM T vector (Promega) and the sequences determined on an automated sequencer (Applied Biosystems). For verification of the P-insertion site the following primers were used (see Fig. 2A): SDE, 5⬘ GTT TTT GAA TTC CCA AAA TGC AAA CTT CTT 3⬘ (position 976 in Fig. 1A, reverse); SDB, 5⬘ GTT TTT GAA TTC ATG CCG AAC GAA AAC GCA GAG 3⬘ (position 290 in Fig. 1A, forward); Plac4, 5⬘ ACT GTG CGT TAG GTC CTG TTC ATT GTT 3⬘ (P-element 5⬘ flank); and Pry4, 5⬘ CAA TCA TAT CGA TGT CTC ACT CA 3⬘ (P-element 3⬘ flank). One microgram of genomic DNA was used in the reactions with the following program: denature at 94°C for 30 s, anneal at 55°C for 1 min, extend at 68°C for 2 min over 35 cycles. Fly stocks. P-element insertion line l(3)J1C7 has a P{lacW} insertion at 66B10 –B11. It was obtained from the G. Rubin Laboratory, Stock P2071. For the studies described here the original balancer chromosome was replaced with TM6c Sb Tb. Balancer homozygotes in this stock die as third-instar larvae (S. Sweeney, personal communication). Exactly 25% of embryos, representing the P-element homozygotes, die before hatching. Df(3L) 66C-G28 was obtained from the Umea stock collection (UM 87548). Western blot analysis. Western blot analysis was performed as previously described [11]. The anti-NMT antibody used was a gift from Dr. R. A. J. McIlhinney. It was raised in rabbits against peptides derived from conserved sequences in NMT and affinity purified against human NMT. After probing, the blot was stained for protein using amido black, by immersion in a solution of 25% 2-propanol, 10% acetic acid, 0.1% amido black for 5 min, followed by a wash (3 ⫻ 2 min) with 25% 2-propanol, 10% acetic acid. Embryo preparations. Cuticle preparations and immunohistochemical staining of embryos were carried out as described [12]. The mouse monoclonal antibodies used have been described [13, 14]. Staining embryos for filamentous actin. Embryos were collected, fixed, and processed as described [15]. It is important that embryos are devitellinized in 95% ethanol and not methanol. After fixation embryos were stained with phalloidin coupled to Texas red (Molecular Probes) at 80 u ml ⫺1. Embryos were then viewed with a Leica TCS-4D confocal microscope using a Kr/Ar laser. Detecting cell death in embryos. To detect cells undergoing apoptosis, embryos were dechorionated in 50% bleach (under the microscope), washed with embryo wash (0.7% NaCl, 0.04% Triton X-1000), and rinsed with water. The embryos were then placed in equal volumes of heptane and 5 g/ml acridine orange (Sigma) in phosphate buffer, pH 7.2. After being shaken for 5 min, embryos at the interface were removed and placed on a cavity slide with Series 700 halocarbon oil (Sigma). The samples were viewed with a conventional fluorescence microscope using fluorescein filters.
RESULTS
Structure of the NMT Transcription Unit Previously, we isolated and sequenced the genomic DNA corresponding to the unique NMT gene of Drosophila and mapped it to polytene band 66C on the left arm of the third chromosome [10]. However, we were unable to define the 5⬘ end of the transcriptional unit, the promoter, or the intron/exon structure of the gene, although we predicted the presence of a single intron in the 5⬘ region. In view of this we amplified 5⬘ cDNA ends using RACE (see Materials and Methods for details) and sequenced four different isolates. This showed that the transcription unit extended from at least nucleotide 200 in Fig. 1A and confirmed the presence of a single intron extending from position 439 to 913. In our previous work we correctly predicted the location of the 3⬘ acceptor but incorrectly suggested nucleotide 506 as the 5⬘ donor site used [10]. Recently three 5⬘ ESTs of the dNMT gene have appeared in the BDGP database and the longest of these commences at position 128. An analysis of the genomic sequence with the Promoter Prediction by Neural Network computer program reveals elements immediately 5⬘ to position 128 strongly predicted to be a promoter (underlined in Fig. 1A). We therefore conclude that the dNMT transcription unit commences at or very close to nucleotide 128. The longest open reading frame in the dNMT cDNA sequence commences with the ATG codon located at position 234. A second in-frame methionine codon is found at 315 and, as this has a better Cavener consensus sequence [16], it may represent the true translation start site for dNMT. The third in-frame methionine lies within the region of homology with other NMTs and is thus unlikely to be the start codon, as suggested previously [10]. The primary translation product predicted from the cDNA sequence indicates that in common with other NMTs there is a diverged N-terminal region which probably has a regulatory rather than a catalytic function [17] (Fig. 1B). A P-Element Insertion within the NMT Transcription Unit Causes Embryonic Lethality We have used the 5⬘ cDNA sequence as a probe to search the databases of the BDGP and discovered an existing Drosophila stock with a P-element transposon inserted within the predicted dNMT reading frame, l(3)J1C7. This chromosome when homozygous causes embryonic lethality. To determine the precise site of insertion of the P element, genomic DNA was digested with restriction enzyme Sau3A and religated, and the sequences flanking the 5⬘ and 3⬘ end of the P element were amplified with P-element-specific primers by inverse PCR and sequenced. Transposition is accompanied by an 8-bp duplication and the position of the first base in this duplication is indicated by an arrow in Fig. 1A (nucleotide 402). To verify that the stock we have
136
NTWASA ET AL.
NMT IN DEVELOPMENT
used in these studies contains the same insertion we prepared DNA from l(3)J1C7/TM6c adult flies (i.e., one normal and one disrupted copy of dNMT) and performed PCR with the sets of primers detailed in Fig. 2A and under Materials and Methods for the 3⬘ and 5⬘ flanking regions of the P element and an internal region of the dNMT gene. These reactions produce fragments of the expected size (Fig. 2B, lanes A–C). In addition DNA prepared from unhatched embryos ((l(3)J1C7 homozygotes) does not allow amplification of the internal region of dNMT because of the intervening P element (Fig. 2B, lanes D–F). As discussed above, translation of dNMT starts at either position 233 or position 315 and thus it is almost certain that the NMT open reading frame is disrupted in the l(3)J1C7 chromosome. To confirm this, we performed Western blot analysis with anti-NMT antibodies. NMT is detected in early and late embryos laid by both wild-type (Fig. 2C(ii), lanes D and E) and l(3)J1C7/TM6c (Fig. 2C(ii), lanes A and B) mothers. In contrast NMT is not present in embryos laid by l(3)J1C7/TM6c mothers, which fail to hatch after 26 h (Fig. 2C(ii), lane c). In a control experiment in which the blot was stained with amido black, approximately equal loading of protein is seen (Fig. 2C(i)). To confirm that the P-element insertion is responsible for the embryonic lethality observed, we crossed l(3)J1C7/TM6c females with males carrying an overlapping deficiency chromosome Df(31) 66C-G28/TM3 sb (deletes 66B8 – 66C9). This deficiency fails to complement l(3)J1C7, a result consistent with the findings of the BDGP. The BDGP also reports that two other close-by but nonoverlapping deficiencies do complement l(3)J1C7: Df(31)pbl-X1 (65F3– 66B10) and Df(31)ems (61C5– 61D3). Finally, the l(3)J1C7 chromosome reverts to viability when the P element is excised precisely (A. Beaton (2000), pers. comm. to FlyBase, http://flybase.bio.indiana.edu/.bin/fbidq.html?FBrf 0126832), showing that the lethality of this chromosome is not due to a nearby mutation mapping within the Df (31) 66C-G28 deficiency. We therefore conclude that this P-element insertion represents a null allele of dNMT and that the embryonic lethality is due to the lack of this enzyme. Mutant Embryos Lacking NMT Have Multiple Developmental Defects Drosophila embryos homozygous for l(3)J1C7 and also embryos transheterozygous for l(3)J1C7/Df(31)
137
66C-G28 display a spectrum of developmental defects. This variable penetrance probably reflects a stochastic depletion in the NMT activity provided by the maternally derived transcript detected by in situ hybridization [see 10]. While some embryos are able to form a normal number of segments, the ventral denticle belts are severely compressed (Fig. 3D), an indication that the process of head involution was defective. Figure 3C shows an embryo in which, apart from the denticle belts being compressed, they are not all present, and the filzkorper structure is in the wrong location, indicative of defects in both head involution and germ-band retraction. The morphogenetic movements required for development of the head are often incomplete, as evidenced by a chaotic positioning of the sclerotinized mouth parts (Figs. 3C and 3D). However, some embryos show nearly normal head involution even when there are other defects evident; for example, defects were observed in posterior development (Fig. 3B), and again these vary considerably in severity (result not shown). Sometimes the posterior spiracles are found in the middle of the embryo, which suggests a failure in posterior midgut involution. Specific Scission of the Ventral Nerve Chord at Parasegment 10 In order to characterize the phenotypes at earlier stages of embryogenesis and in the context of a defined patterning process, we stained mutant embryos between 12 and 18 h of development with antibodies specific for Ultrabithorax (Ubx) [13], a homeodomain transcription factor expressed specifically in the thoracic and abdominal segments of the ectoderm and nervous system (Fig. 4). Ubx staining in the epidermis (Fig. 4B) is disorganized, consistent with the disrupted patterns seen in the cuticle. The staining also extends to the posterior of the embryo (cf. wild-type pattern, Fig. 4A), which suggests disruption in patterning the most posterior part of the embryo. Severe phenotypes show failures in head involution, dorsal closure, germband retraction, and a complete fragmentation of the Ubx pattern in the central nervous system (CNS) (Fig. 4C). Less severely affected embryos also have head involution defects but the segmental structure of the CNS has not disintegrated (Fig. 4D). A milder phenotype which is often observed is detailed in Figs. 4E– 4G. In these embryos the segmental pattern of the CNS is largely normal except for a constriction which occurs
FIG. 1. Structure of the NMT transcription unit. The 5⬘ end of the dNMT transcript was determined by 5⬘RACE and confirmed by EST database entries in the BDGP. (A) DNA sequence of the 5⬘ region of the dNMT gene. The underlined nucleotides are sequences predicted to be promoter elements by the Promoter Prediction by Neural Network method (http://www-hgc.lbl.gov/inf/promoter-instrucs.html). The boxed nucleotide is the proposed transcription initiation site. The predicted open reading frame for dNMT is indicated, the underlined amino acids being in addition to those shown previously [10]. The intron is shown in lowercase and the insertion site in l(3)J1C7 is indicated by an arrow. (B) Alignment of NMT precursors generated using the CLUSTALX (1.62b) multiple sequence alignment program [34]. Symbols: (*) identical or conserved residues in all sequences in the alignment, (:) conserved substitutions; (.) semiconserved substitutions. Note. candida is Candida albicans (a yeast).
138
NTWASA ET AL.
FIG. 2. The P-element insertion in chromosome l(3)J1C7 disrupts the dNMT gene. (A) Schematic diagram of the P-element insertion in the dNMT transcription unit. The 4.6-kb SalI–EcoRI genomic fragment [10] is illustrated. The black boxes indicate the positions of the two exons and the gray box the single intron. The sequence surrounding the P insertion and the site of transposition is shown (numbering relative to primer SDB). The positions of the four primers used in (B) are shown (see Materials and Methods for sequences). (B) PCR products from genomic l(3)J1C7/TM6c DNA (lanes A–C) and l(3)J1C7 homozygous DNA (lanes D–F) using various primer combinations. SDE ⫹ Pry4 (A, D), SDE ⫹ Plac4 (B, E), and SDE ⫹ SDB (C, F). (C) (i) Amido black staining of Western blot membrane after probing. Arrow indicates yolk proteins (ca. 42 kDa) (more abundant in 0- to 3-h embryos). (ii) Western blot analysis of embryo extracts from (lanes): (A) 20- to 24-h and (B) 0- to 3-h wild-type embryos, (C) unhatched embryos of the l(3)J1C7/TM6SbTb flies, (D) 20- to 24-h and (E) 0- to 3-h embryos, laid by l(3)J1C7/TM6SbTb heterozygotes. Exactly 50 embryos of each type were picked and crushed into SDS–PAGE loading buffer. The samples were separated in a 7.5% gel and blotted as described [11]. The positions of molecular mass markers are shown.
NMT IN DEVELOPMENT
139
specifically at parasegment 10. This constriction is progressive and eventually results in complete scission of the ventral nerve chord (Fig. 4G). In some embryos there is a more general thinning of the nerve chord (Fig. 4F). This feature of the NMT phenotype can be seen more clearly in mutant embryos stained with monoclonal antibody BP102 [14], which highlights the commissures and the longitudinal connectives of the CNS (Fig. 5). This reveals that the central nervous system forms normally and that the breakage occurs subsequently because remnants of the longitudinal fascicles can still be discerned at the site of scission (Figs. 5B and 5C). In the more severe phenotypes, there is a global degeneration of the central nervous system (Figs. 5D–5F). The Actin Cytoskeleton Is Distorted in Mutant Embryos
FIG. 3. Cuticle pattern of NMT mutant embryos. Cuticle prepared from unhatched embryos laid by l(3)J1C7/TM6SbTb flies shows severe and variable pattern defects. (A) Wild-type embryo. Arrowhead, ventral denticle band; double arrowhead, the posterior structure known as the filzko¨rper. (B–D) Unhatched mutant embryos. Although B has a normal, noncompressed complement of denticle belts (compare with C and D), the filzko¨rper in B is located along its dorsal edge (arrowed), not posteriorly as in the wild type; this is likely due to failure of germ-band retraction, leading to a “curling-up” of the embryo. The compression of the ventral denticles in C (bracketed) reflects failure of the morphogenetic process known as head involution (cf. distribution of denticle belts in A and B); the filzko¨rper is situated in the wrong place (dashed arrow), indicative of a defect in germ-band retraction. Embryo D similarly suffers from compressed denticle belts (bracketed), but has a roughly normal complement of denticle belts (although some disruption is seen), and its filzko¨rper is situated
A number of the morphogenetic defects described above are also seen in embryos mutant for genes involved in dynamic rearrangement of the cytoskeleton (see Introduction). We therefore treated NMT ⫺ embryos with Texas red-conjugated phalloidin, which binds to filamentous actin, and examined the staining with the confocal microscope. Normal embryos had actin patterns indistinguishable from those observed in vivo in embryos carrying a moesin– green fluorescent protein transgene [18], which indicates that the staining procedure was successful (Fig. 6A). However, mutant embryos in which dorsal closure has failed display strikingly abnormal patterns. During dorsal closure the lateral ectoderm elongates dorsally over the amnioserosa finally fusing at the dorsal midline. This process is accompanied by progressive elongation of the ectodermal cells and accumulation of F-actin at the leading edge [6, 19, 20]. In Figs. 6B and 6C a mutant defective in dorsal closure has severe disruptions of the leading edge cells, although accumulation of F-actin is still evident. The organization of the leading edge is obliterated and is similar to that seen in embryos homozygous for the zipper gene (nonmuscle myosin II) [6]. In a more severely affected embryo (Figs. 6D– 6F), disintegration of the central nervous system can be seen and also complete breakdown of cell morphology in the epidermis. Abnormal Patterns of Cell Death in Mutant Embryos A number of aspects of the NMT phenotype are similar to defects caused by cell death genes, for example, head involution defective (see below) [21]. In view of
posteriorly, as is normal. The chaotic positioning of the sclerotinized mouth parts, a feature sometimes found in these mutant embryos, can be seen in C and D (arrowed)— compare this with their normal positioning, seen in A (arrowed).
140
NTWASA ET AL.
FIG. 4. Staining pattern of Ubx in NMT embryos. 12- to 18-h-old embryos were stained for Ubx. (A) Normal Ubx staining, in a stage 15/16 embryo (staging according to [33]). (B–G) A range of mutant phenotypes. (B) Stage 14 embryo showing defects in the epidermis. The staining extends to the posteriormost parts of the embryo, indicating abnormality of posterior structures. (C and D) Failures in dorsal closure, head involution, germ-band retraction, and, in C, degeneration of the CNS (arrow) can also be observed. Embryos C and D are difficult to stage by standard morphological criteria: while both show Ubx concentrated in the nervous system, characteristic of later developmental stages (stage 16 or later), gut development has clearly stalled, indicated by the nonsegmented nature of the gut (see asterisks in C and D; compare with F). (E) A constriction (arrowed) in some mildly affected embryos can be seen at parasegment 10. (F and G) Thinning (F) or complete scission (G) of the ventral nerve cord is bracketed here. E–G have the overall appearance of stage 16 –17 embryos, but some elements, e.g., the nervous system in F, show earlier morphologies. Asterisk in F indicates the more differentiated gut (compare with that in C and D). FIG. 5. The central nervous system in mutant embryos. The embryos were stained with monoclonal antibody BP102 to highlight the axons of central nervous system neurons. (A) Normal staining pattern of BP102 (stage 16). (B) A gap in the CNS (arrowed) that is magnified in (C) to show remnants of the fascicles (stage 16). This suggests that the connectives form and then break apart. (D–F) Three embryos showing severe phenotypes involving failures of dorsal closure, head involution, and germ-band retraction and general degeneration of the CNS affecting both the longitudinal connectives and the commissures (it is not possible to meaningfully stage these due to the severity of the phenotype).
141
NMT IN DEVELOPMENT
FIG. 6. F-actin distribution in mutant embryos. The embryos (12- to 18-h collection) were stained with phalloidin conjugated with Texas red as described under Materials and Methods. (A) Normal embryo (stage 16) showing characteristic pattern of staining in the ventral nerve chord and at muscle attachment sites (see [18]). (B and C) Mutant embryo (stage 13/14) with failure of dorsal closure; box in B indicates approximate extent of the higher magnification view in C. Arrows indicate disrupted leading edge. (D–F) Severely affected embryo (stage 15/16). E and F are higher magnification views of D, the approximate locations of which are indicated by the solid and dashed boxes, respectively. In F, an arrow indicates epidermal cells with progressively more abnormal morphology; F-actin accumulation can be seen at the leading edge (arrowhead). as, amnioserosa; vnc, ventral nerve cord; ec, epidermal cells. Objectives used: A, 40⫻; B, 16⫻; C, 63⫻; D, 16⫻; E, 40⫻; F, 40⫻.
this we have compared the patterns of cell death observed in mutant and normal embryos. In wild-type embryos stained with acridine orange we observe the same patterns of cell death as reported previously (Fig. 7A) [22]. In contrast in mutant embryos there is widespread ectopic apoptosis occurring. For example, a severely affected embryo (Fig. 7B) that lacks head structure and in which the germ band has failed to retract has cell death occurring in the amnioserosa. In other cases there is cell death apparent in the posterior region of the embryo the location of which correlates with the posterior patterning defects described above (Fig. 7C). Finally cells associated with specific constriction of the ventral nerve chord at parasegment 10 are apparently undergoing apoptosis, which suggest that the observed scission of the nerve chord may be caused by cell death (Fig. 7D).
DISCUSSION
The severity and multiplicity of developmental defects caused by a lack of NMT function indicate the critical roles that myristoyl proteins play in embryonic development and the critical nature of this lipid modification for protein function. One striking feature of the phenotypes described here is the correspondence with those of other mutations which affect the dynamic rearrangement of the actin cytoskeleton [6, 7]. In view of the known biochemical functions of myristoyl proteins in mediating cell shape change and movement, it is likely that the defects observed are caused by failures in actin-driven morphogenesis. In support of this idea, we show that in mutant embryos the actin cytoskeleton and cell morphology are often disrupted. This suggests that the specific developmental signals
142
NTWASA ET AL.
FIG. 7. Cell death in NMT embryos. Embryos were treated with acridine orange as described under Materials and Methods. (A) Wild-type embryo (stage 11). (B) Late embryo with lack of head structure. Arrows indicate dying cells in the amnioserosa. (C) Embryo with widespread cell death in the posterior, indicated by arrow. (D) Cell death in the central nervous system; the constriction at parasegment 10 is indicated. C and D, stage 16/17.
required for these changes involve regulatable myristoyl switches of the kind described for MARCKS in mammalian culture cells. Interestingly, transgenic mice lacking MARCKS display similar defects in head structure and CNS development [23]. The properties of two Drosophila homologues of the myristoylated nonreceptor tyrosine kinase src (Dsrc 64B and Dsrc 42A (previously known as Dsrc41)) provide further support for the hypothesis that at least some of the morphogenetic defects described here are caused by abnormal cytoskeletal dynamics [24, 25]. Overproduction or overexpression of dominant negative forms of Dsrc64B causes embryonic lethality with defects in germ-band retraction, thinning of the ventral nerve cord, and specific breakages in the longitudinal connectives. In more recent work, both Dsrc42A and Dsrc64B were shown to regulate basket (the Drosophila jun kinase), the activity of which is required for dorsal closure [20]. A strong allele of Dsrc42A (Dsrc42A myri) was found to be a change of the myristoylated glycine 2 to aspartic acid. Fifty percent of embryos homozygous for this allele die before hatching and have disordered mouthparts similar to defects we observe in NMT ⫺ embryos. Furthermore, double mutants of Dsrc62B and Dsrc42A myri were defective in dorsal closure and had disrupted actin cytoskeleton in
the leading-edge cells [26]. These findings indicate that these two myristoylated nonreceptor tyrosine kinases are functionally redundant and are key regulators of dorsal closure. They also support the idea that, in NMT ⫺ embryos, dorsal closure defects are caused by a failure to myristoylate Dsrc42A and Dsrc64B. In tissue culture cells the vertebrate homologue (c-src) also functions in actin cytoskeleton rearrangements [27]. Another class of myristoyl proteins that may contribute to the phenotype are the ADP ribosylation factors, four of which have been identified in the Drosophila genome. These are a subfamily of ras-related GTPases and in general they function as regulators of membrane traffic. Interestingly in mammalian cells the homologue of the Drosophila arf3 product, ARF6, binds to the endosome when associated with GDP, but in the GTP-bound state ARF6 is associated with the plasma membrane where it organizes cortical actin. Recent work has shown that ARF6 can affect actin cytoskeleton rearrangements both alone and in concert with the nonmyristoylated small G protein Rac [28, 29]. The specific effects on head involution and the central nervous system are also reminiscent of the phenotypes caused by the cell death genes reaper, head involution defective (hid), and grim [21, 30, 31]. Indeed the generalized thinning of the nerve chord resembles the effect
NMT IN DEVELOPMENT
caused by ectopic expression of hid and reaper in the central nervous system [32]. This may indicate a role for myristoyl proteins in the regulation of apoptosis. On the other hand the breakage of the nerve chord may simply reflect mechanical stresses caused by defective germband retraction. In that regard it may be significant that parasegement 10 is located at the posterior pole prior to germ-band shortening [33]. In view of the devastating consequences for the embryo of removing NMT, we propose that myristoyl proteins have a number of fundamental roles in morphogenesis and that lipid modification of myristoyl proteins is essential for their function in these processes. These roles are consistent with the known biochemical activities of myristoyl proteins in signal transduction, cytoskeletal rearrangement, and cell movement. Future studies will concentrate on the morphogenetic role of lipid modification in individual myristoyl proteins. This work was supported by a Wellcome Trust International Research Development Fellowship awarded to M.N. and N.G. We are very grateful to Drs. Rob White and Cahir O’Kane for helping to interpret the mutant phenotype.
REFERENCES
143
have different patterns of N-linked glycosylation. FEBS Lett. 365, 83– 86. 12.
Ashburner, M. (1989). “Drosophila: A Laboratory Manual,” Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
13.
White, R. A., and Wilcox, M. (1984). Protein products of the bithorax complex in Drosophila. Cell 39, 163–171.
14.
Patel, N. H. (1997). Imaging neuronal subsets and other cell types in whole mount Drosophila embryos and larvae using antibody probes. In “Drosophila melanogaster: Practical Uses in Cell and Molecular Biology” (L. S. B. Goldstein and E. A. Fryberg, Eds.), “Methods in Cell Biology,” Vol. 44, Academic Press, San Diego.
15.
Thatcher, J. W., and Dickinson, W. J. (1996). Rhodamine phalloidin staining without hand peeling. Drosophila Inform. Serv. 76, 188.
16.
Cavener, D. R. (1987). Comparison of the consensus sequence flanking translational start sites in Drosophila and vertebrates. Nucleic Acids Res. 15, 1353–1361.
17.
Glover, C. J., Hartman, K. D., and Felsted, R. L. (1997). Human N-myristoyltransferase amino-terminal domain involved in targeting the enzyme to the ribosomal subcellular fraction. J. Biol. Chem. 272, 28680 –28689.
18.
Edwards, K. A., Demsky, M., Montague, R. A., Weymouth, N., and Kiehart, D. P. (1997). GFP-moesin illuminates actin cytoskeleton dynamics in living tissue and demonstrates cell shape changes during morphogenesis in Drosophila. Dev. Biol. 191, 103–117.
19.
Harden, N., Lee, J., Loh, H. Y., Ong, Y. M., Tan, I., Leung, T., Manser, E., and Lim, L. (1996). A Drosophila homolog of the Rac- and Cdc42-activated serine/threonine kinase PAK is a potential focal adhesion and focal complex protein that colocalizes with dynamic actin structures. Mol. Cell. Biol. 16, 1896 – 1908.
20.
Noselli, S. (1998). JNK signaling and morphogenesis in Drosophila. Trends Genet. 14, 33–38.
21.
Grether, M. E., Abrams, J. M., Agapite, J., White, K., and Steller, H. (1995). The head involution defective gene of Drosophila melanogaster functions in programmed cell death. Genes Dev. 9, 1694 –1708.
22.
Abrams, J. M., White, K., Fessler, L. I., and Steller, H. (1993). Programmed cell death during Drosophila embryogenesis. Development 117, 29 – 43.
23.
Stumpo, D. J., Bock, C. B., Tuttle, J. S., and Blackshear, P. J. (1995). MARCKS deficiency in mice leads to abnormal brain development and perinatal death. Proc. Natl. Acad. Sci. USA 92, 944 –948.
24.
Takahashi, F., Endo, S., Kojima, T., and Saigo, K. (1996). Regulation of cell– cell contacts in developing Drosophila eyes by Dsrc41, a new, close relative of vertebrate c-src. Genes Dev. 10, 1645–1656.
1.
Johnson, D. R., Bhatnagar, R. S., Knoll, L. J., and Gordon, J. I. (1994). Genetic and biochemical studies of protein N-myristoylation. Annu. Rev. Biochem. 63, 869 –914.
2.
McLaughlin, S., and Aderem, A. (1995). The myristoyl-electrostatic switch: A modulator of reversible protein–membrane interactions. Trends Biochem. Sci. 20, 272–276.
3.
Ames, J. B., Ishima, R., Tanaka, T., Gordon, J. I., Stryer, L., and Ikura, M. (1997). Molecular mechanics of calcium–myristoyl switches. Nature 389, 198 –202.
4.
Randazzo, P. A., Terui, T., Sturch, S., Fales, H. M., Ferrige, A. G., and Kahn, R. A. (1995). The myristoylated amino terminus of ADP-ribosylation factor 1 is a phospholipid- and GTPsensitive switch. J. Biol. Chem. 270, 14809 –14815.
5.
Myat, M. M., Anderson, S., Allen, L. A., and Aderem, A. (1997). MARCKS regulates membrane ruffling and cell spreading. Curr. Biol. 7, 611– 614.
6.
Young, P. E., Richman, A. M., Ketchum, A. S., and Kiehart, D. P. (1993). Morphogenesis in Drosophila requires nonmuscle myosin heavy chain function. Genes Dev. 7, 29 – 41.
7.
Blake, K. J., Myette, G., and Jack, J. (1998). The products of ribbon and raw are necessary for proper cell shape and cellular localization of nonmuscle myosin in Drosophila. Dev. Biol. 203, 177–188.
25.
Byars, C. L., Bates, K. L., and Letsou, A. (1999). The dorsalopen group gene raw is required for restricted DJNK signaling during closure. Development 126, 4913– 4923.
Kussick, S. J., Basler, K., and Cooper, J. A. (1993). Ras1dependent signaling by ectopically-expressed Drosophila src gene product in the embryo and developing eye. Oncogene 8, 2791–2803.
26.
Tateno, M., Nishida, Y., and Adachi-Yamada, T. (2000). Regulation of JNK by Src during Drosophila development. Science 287, 324 –327.
27.
Kaplan, K. B., Bibbins, K. B., Swedlow, J. R., Arnaud, M., Morgan, D. O., and Varmus, H. E. (1994). Association of the amino-terminal half of c-Src with focal adhesions alters their properties and is regulated by phosphorylation of tyrosine 527. EMBO J. 13, 4745– 4756.
28.
Boshans, R. L., Szanto, S., van Aelst, L., and D’Souza-Schorey, C. (2000). ADP-ribosylation factor 6 regulates actin cytoskele-
8.
9.
Brown, N. H., and Kafatos, F. C. (1988). Functional cDNA libraries from Drosophila embryos. J. Mol. Biol. 203, 425– 437.
10.
Ntwasa, M., Egerton, M., and Gay, N. J. (1997). Sequence and expression of Drosophila myristoyl-CoA: Protein N-myristoyl transferase: Evidence for proteolytic processing and membrane localisation. J. Cell Sci. 110, 149 –156.
11.
Kubota, K., Keith, F. J., and Gay, N. J. (1995). Wild type and constitutively activated forms of the Drosophila Toll receptor
144
NTWASA ET AL.
ton remodeling in coordination with Rac1 and RhoA. Mol. Cell. Biol. 20, 3685–3694. 29. Radhakrishna, H., Al-Awar, O., Khachikian, Z., and Donaldson, J. G. (1999). ARF6 requirement for Rac ruffling suggests a role for membrane trafficking in cortical actin rearrangements. J. Cell Sci. 112, 855– 866. 30. White, K., Grether, M. E., Abrams, J. M., Young, L., Farrell, K., and Steller, H. (1994). Genetic control of programmed cell death in Drosophila. Science 264, 677– 683. 31. Nassif, C., Daniel, A., Lengyel, J. A., and Hartenstein, V. (1998). The role of morphogenetic cell death during Drosophila embryonic head development. Dev. Biol. 197, 170 –186. Received June 30, 2000 Revised version received October 6, 2000
32.
Zhou, L., Schnitzler, A., Agapite, J., Schwartz, L. M., Steller, H., and Nambu, J. R. (1997). Cooperative functions of the reaper and head involution defective genes in the programmed cell death of Drosophila central nervous system midline cells. Proc. Natl. Acad. Sci. USA 94, 5131–5136.
33.
Campos-Ortega, J. A., and Hartenstein, V. (1997). “Embryonic Development of Drosophila melanogaster,” Springer-Verlag, New York.
34.
Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins, D. G. (1997). The CLUSTAL_X windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876 – 4882.
