| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Corresponding author: Margaret T. Fuller, Departments of Developmental Biology and Genetics, Beckman Center B300, 279 Campus Dr., Stanford University School of Medicine, Stanford, CA 94305-5329., fuller{at}cmgm.stanford.edu (E-mail)
Communicating editor: S. HENIKOFF
 | ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Drosophila nup154 encodes a predicted nucleoporin homologous to yeast Nup170p, Nup157p, and vertebrate Nup155, all of which are major components of the nuclear pore complex (NPC). Unlike its yeast homologs, nup154 is essential for viability. Animals with strong loss-of-function nup154 mutations caused by P-element insertion in the 5'-UTR of the gene died as larvae with small discs, brains, and testes. nup154 mRNA expression appeared developmentally regulated in tissues of wild-type embryos, larvae, and adults, suggesting that new nup154 synthesis is required when assembly of new NPCs is required, as in proliferating or growing tissues. Two additional nup154 alleles also associated with different P-element inserts in the 5'-UTR were viable but had strong loss-of-function sterile phenotypes, including failure to maintain spermatogenic stem cells and failure to progress into vitellogenic stages of oogenesis. Lethality vs. viability correlated with orientation of the P-element inserts in the different alleles. Transcript analysis by 5'-RACE suggested a mechanism for allelic differences: an outward-directed promoter internal to the P-element 3' end able to drive sufficient expression of the nup154 transcript for viability but not for fertility.
NUCLEAR pore complexes (NPCs) are portals for highly regulated bidirectional transport of macromolecules between the cytoplasm and the nucleus (reviewed in 
Each NPC is composed of ~50 proteins, referred to as nucleoporins (reviewed in
Mutations have now been identified in a Drosophila gene, nup154, that encodes a homolog of the nonrepeat nucleoporins Nup155 of rat and Nup170p and Nup157p of S. cerevisiae (
In this article, we report the phenotypic and molecular analyses of strong loss-of-function mutations in the conserved Drosophila nucleoporin, nup154. The phenotype of lethal nup154 alleles indicates that the gene is essential for viability and required for normal cell proliferation. The pattern of mRNA expression in embryonic development and postembryonic tissues suggests that new expression of nup154 occurs in tissues undergoing mitotic proliferation, consistent with a requirement for new protein for nuclear pore assembly. Analysis of the intron/exon structure of the nup154 transcription unit and molecular lesions associated with nup154 alleles revealed that four viable and three lethal independently derived transposable element-induced mutant lines carried P-element insertions clustered within 48 bp of each other in the nup154 5'-UTR. Lethality vs. viability correlated with the orientation of the P-element inserts of different alleles. Transcript analysis by 5'-RACE suggested a mechanism for the allelic differences: expression of the nup154 transcript from a promoter internal to the P-element insert sufficient for viability but not for fertility.
| MATERIALS AND METHODS |
|---|
| TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Drosophila stocks and husbandry:
Flies were raised on standard cornmeal molasses agar medium at 25°. Visible markers, balancers, and deletions are as described in
Identification and isolation of nup154 alleles:
The nup1541 mutant allele was generated during mobilization of an unlinked PlacW transposable P element (
To demonstrate that the nup1541 phenotype was due to the P-element insertion, nup1541/CyO flies were crossed to flies bearing the P{ry+t7.2 
2-3} (99B) transposase insertion to mobilize the P element and generate w- eye color revertants (
Characterization of nup154 mutant phenotypes:
Determination of the mutant nup154 lethal phase:
The lethal phase of trans-heterozygous combinations of nup154 lethal alleles was determined by embryonic and larval hatch counts for all combinations of nup1542, nup1543, and nup1546, representing alleles derived from three different parental chromosomes. For embryonic hatch counts, embryos from appropriate nup154/OreR parental matings were collected on apple juice plates overnight and aged 24 hr. Percentage hatching was determined as the number of hatched larvae divided by the total number of embryos laid on the plate (hatched and unhatched, n = 1000/genotype). Percentage hatching from crosses involving two different nup154 mutant alleles was comparable to that of parallel control crosses with homozygous OreR parents. To assess viability through the larval stages, appropriate nup154/T(2;3) SM6a; TM6B,Tb parents were mated, and embryos collected on apple juice agar plates were supplied with yeast paste and aged for 2–5 days. The Tb marker carried on the T(2;3) balancer allowed identification of trans-heterozygous larvae, which were non-Tb. Larval stages were confirmed by mouth hook morphology. Similar experiments with homozygous nup154 lethal alleles indicated that viability to the late third instar was slightly reduced compared to the trans-heterozygotes, but those that survived exhibited a similar phenotype.
X-Gal staining:
X-Gal staining was performed as described (
Examination of adult gonads: Adult ovaries were dissected from females fed on yeast paste for several days after eclosion. Gonads were examined with either phase contrast or Nomarski optics on a Zeiss Axioskop microscope.
Molecular analyses of nup154:
Determination of nup154 genomic region and P-element insertion sites:
Genomic DNA (1.7 kb) flanking the nup1541 P-element insert was cloned by plasmid rescue after digestion with SacII and shown by in situ hybridization to map to polytene chromosome interval 32D (
genomic phage library (provided by J. Tamkun), yielding 6 phage clones spanning 23.8 kb from the nup154 region. P-element insertion sites in alleles nup1541, nup1543, nup1544, nup1545, and nup1546 were determined by plasmid rescue and sequence analysis of the rescued fragments.
RNA isolation and Northern blot analysis:
RNA for developmental Northern blots was purified from staged embryos, larvae, and adults using RNAzol and its accompanying protocol (Tel-Test, Inc.) or as described by
Isolation of nup154 cDNAs:
Sequence derived from the flanking DNA rescued from the nup1541 insert showed that the 0.8-kb transcript represented the previously reported ribosomal protein L9 (rpL9; -gt11 Canton, S 0- to 18-hr embryonic cDNA library (CLONTECH, Palo Alto, CA) with the 4.6- and 5.6-kb BamHI genomic fragments that span either side of the nup1541 insert. A partial cDNA clone representing the 5' portion of the predicted nup154 transcript was used to screen a 4- to 8-hr embryonic plasmid cDNA library (
|
|
|
|
|
|
|
|
Sequencing nup154 cDNAs and genomic region:
cDNA inserts were subcloned into pBluescriptII-KS- (Stratagene, La Jolla, CA) for mapping and sequencing. The full-length cDNA was sequenced on both strands by the ABI PRISM dye terminator cycle method (Perkin-Elmer, Norwalk, CT) using vector primers for end runs (on full-length and drop-out cDNA clones) and internally directed synthetic oligonucleotide primers (Beckman Center PAN-Facility). Sequences from partial cDNAs were identical to corresponding regions of the full-length cDNA sequence, with the exception that intron 3 was not found in two partial cDNAs (Fig 1B, asterisk). The cDNAs without intron 3 maintained the same reading frame, encoding 16 additional amino acids near amino acid position 200. A total of 6.7 kb of genomic sequence, including 740 bp upstream and 960 bp downstream of the nup154 transcription unit, was obtained from three genomic fragments subcloned from phage into pBluescript. Sequence from both strands of the genomic DNA and the cDNA was determined, with two- to six-fold coverage across any given region. Sequence contigs and alignments were assembled using SEQMAN and MEGALIGN (DNASTAR, Madison, WI). Sequence comparison searches and alignments were performed using NCBI BLAST (
Construction of P[nup154+] and P[nup154] transgenic strains:
A nup154 genomic rescue fragment spanning the entire nucleoporin transcription unit plus 857 bp of DNA upstream and 648 bp of DNA downstream was constructed as follows. A 3-kb XhoI-NotI fragment from a plasmid carrying a 3-kb XhoI-XmaI 3'-genomic fragment of the gene subcloned into pBluescript (pBS-XX) was subcloned into the pCaSpeR-4 transformation vector (2-3 helper plasmid for P-element-mediated germline transformation (] were generated and tested for complementation of the nup154 mutant phenotypes.
In situ hybridization to mRNAs:
RNA probes were labeled with digoxigenin during in vitro transcription according to Genius System protocols (Boehringer-Mannheim, Indianapolis), treated 30–60 min to alkaline hydrolysis, and used for hybridizations at 1:100 (
Immunohistochemistry:
Primary antibodies used were rabbit anti-Vasa (1:2000; provided by A. Williamson and R. Lehmann) and rabbit anti-ß-galactosidase (1:10,000; Cappel). Secondary antibody used was biotinylated goat anti-rabbit (1:500; Vector Labs, Burlingame, CA). To assess the presence of pole cells in embryonic gonads, 13- to 20-hr-old progeny from nup1541/CyO wg-lacZ parents were double labeled for the germline antigen Vasa and the balancer-derived ß-galactosidase marker and detected by immunohistochemical methods as described (
| RESULTS |
|---|
| TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The nup154 Drosophila nucleoporin homolog is essential for viability:
Four new P-element- and two new EMS-induced mutant alleles of the Drosophila nup154 gene were identified. The viable but male and female sterile allele nup1541 was shown to contain a P-element insert in polytene interval 32D by in situ hybridization (data not shown). The male and female sterility of nup1541 homozygotes was due to the P-element insert, as both phenotypes reverted upon excision of the P element after exposure to transposase (MATERIALS AND METHODS). Three additional P-element-induced mutations with insertions that mapped to the region and two EMS-induced mutations isolated from a screen for nup154 alleles all failed to complement the nup1541 male sterile phenotype (MATERIALS AND METHODS and Table 1). In addition, the nup1541 allele was used to obtain small deletions in the gene by imprecise excision (MATERIALS AND METHODS).
|
Five P-element- or EMS-induced alleles were lethal when in trans to a deficiency or with each other, indicating that nup154 is an essential gene (MATERIALS AND METHODS; Table 1B and Table 1C). Two viable but male and female sterile P-element-induced alleles, nup1541 and nup1545, when heterozygous with the deficiency had tiny testes and small ovaries similar to flies homozygous for nup1541, suggesting they are strong loss-of-function alleles with respect to the sterility phenotypes (MS, FS in Table 1A). Three of the lethal alleles were also viable with similar male and female sterile phenotypes when heterozygous with the viable insertion alleles (MS, FS in Table 1B). These gametogenesis defects were stronger than those reported for P-element-induced tlp1 or tlp2 alleles, which were viable but sterile with extensive germline differentiation in both sexes (
The lethal and sterile phenotypes were due to mutations in the gene encoding the nup154 homolog. The structure of the nup154 gene was deduced from comparison of the sequence of genomic DNA in the region with the sequence of a full-length cDNA isolated from a 4- to 8-hr embryonic library (MATERIALS AND METHODS and Fig 1B). A 7-kb genomic fragment spanning the nup154 coding region rescued all mutant phenotypes (lethality, male sterility, and female sterility) associated with the nup1541 and nup1546 mutations when introduced by P-element-mediated germline transformation (Fig 1A), while the same 7-kb genomic fragment carrying two in-frame stop codons at codons 109 and 110 of Nup154 did not rescue any of the nup154 mutant phenotypes. Two independent transgenic lines of a 3.9-kb construct (Fig 1A) previously shown to rescue mutants of the neighboring ribosomal protein L9 gene (
Lethal nup154 mutant alleles exhibit defects in mitotically proliferating tissues:
Phenotypes associated with strong loss-of-function nup154 mutations indicated that zygotic function of nup154 is required for growth of mitotically proliferating tissues in larvae and adults. Animals trans-heterozygous for nup154 lethal alleles died throughout the early larval stages, with survivors found until only just after the third instar larval molt. Until the end of second instar, the mutant larvae appeared morphologically similar to their heterozygous siblings in body and tissue size. After the third instar molt, the lethal mutant trans-heterozygotes failed to grow and exhibited a "classic mitotic mutant" phenotype (
Expression of nup154 mRNA is developmentally regulated:
Consistent with an essential requirement for the protein, a 4.5-kb transcript corresponding to the nup154 product was detected in all developmental stages examined (embryo, larvae, and adult) by Northern blot analysis (Fig 3). High expression levels were observed in ovaries and early embryos. Levels of mRNA were reproducibly lower in 2- to 4-hr embryos, rose to a peak in 4- to 8-hr embryos, were low in late embryos, larval, and pupal stages, and rose again by adulthood.
In situ hybridization revealed that expression of nup154 mRNA was restricted to specific tissues during certain stages of embryogenesis (Fig 4), consistent with the fluctuations in mRNA levels observed by Northern blot analysis. The nup154 transcript, presumably maternally contributed, was distributed throughout 0- to 2-hr embryos (Fig 4A). By stage 5, nup154 mRNA was detected only at the posterior end of the embryo, just under the pole cells (Fig 4B). Localization of nup154 mRNA to the region under the germ cells continued during germband elongation (Fig 4C). In addition, nup154 mRNA was detected throughout the developing mesoderm (Fig 4C and Fig D), presumably due to zygotic transcription of nup154 in this tissue. At the onset of stage 11, nup154 mRNA was most prominent in the central nervous system (CNS) and gut (Fig 4E and Fig F), and by stage 14, it was also detected in the brain and dorsal vessel. Near the end of embryogenesis, nup154 mRNA was primarily restricted to three distinct tissues: the lymph glands (hematopoietic organ), gonadal germline, and several cells, likely mushroom body neuroblasts, in the brain (Fig 4G and Fig H). All regions where nup154 mRNA was detected in late embryos contained tissues that are mitotically active after the completion of embryogenesis. The nup154 mRNA expression pattern throughout embryogenesis was similar to that described for a set of genes expressed coincident with the onset of S phase (
In larvae, although the overall level of nup154 mRNA was low (Fig 3), elevated levels were detected in tissues containing dividing cells. In situ hybridization revealed nup154 mRNA in imaginal discs and certain regions of the brain (Fig 5A). In larval and adult testes, nup154 mRNA was present in all male germline cells up through the meiotic stages (Fig 5C and Fig E). In adult testes, late primary spermatocytes and cells in the meiotic divisions were usually the most darkly stained (Fig 5E). However, lightly stained preparations of adult testes revealed a noticeably darker stripe near the apical tip of the testes (Fig 5J). We were unable to resolve whether this band of increased staining was associated with germ cells in mitosis or in premeiotic S phase. nup154 mRNA was not detected in agametic testes from adult sons of osk mothers, indicating that expression of nup154 mRNA in testes was germline dependent (data not shown). nup154 mRNA was clearly detected in wild-type ovaries in germarium region 2 and throughout later stages of oogenesis, with darker staining after stage 8 (Fig 5G and Fig I).
nup154 function is required for survival of male germline stem cells and progression of oogenesis into the vitellogenic stages:
The expression pattern of nup154 mRNA in testes and ovaries and the defects in proliferating tissues caused by the lethal nup154 alleles led us to reexamine the effects of nup154 mutations on male and female gametogenesis. The nup1541 and nup1545 alleles, which behaved as strong loss-of-function mutations with respect to their effects on male and female gametogenesis, caused stronger defects in gametogenesis than the hypomorphic tlp1 and tlp2 alleles described in
The male sterile phenotype of nup1541 and nup1545 indicated that function of nup154 is required for survival of male germline stem cells. Testes from homozygous nup1541, trans-heterozygous nup1541/nup1545, nup1541/Df(2L)J39, or nup1545/Df(2L)J39 mutant adults were consistently tiny and appeared to contain no germ cells (Fig 6B), much like testes from agametic flies. ß-Galactosidase activity from enhancer trap lacZ markers expressed in male germline stem cells and their immediate daughters in wild type (Fig 6C) was not detected in testes from homozygous nup1541 adults (Fig 6D). Immunofluorescence stains with the pan germ-cell marker anti-Vasa confirmed that no germline cells were present in testes from homozygous nup1541 adults (data not shown). Enhancer trap markers for the somatically derived hub and cyst cells were expressed in testes from both wild-type and homozygous nup1541 mutants (data not shown), indicating that these somatic cell types were still present in mutant testes.
To determine when mutations in nup154 first affect the male germline, we examined gonads from earlier stages in development. Phase-contrast microscopy revealed that some germ cells were occasionally present in nup1541 larval gonads, unlike the completely empty adult testes. However, very few germ cells per testis were seen (usually 0–10 cells), and the germ cells present were spermatocytes not organized into normal germline cysts (Fig 6F). Small cells present at the apical tip in wild type (Fig 6E) were absent in the mutant, suggesting absence of renewing germline during the late larval stages. The similar phenotype seen in the lethal alleles is consistent with nup1541 being a strong loss-of-function allele with respect to the germline but not the somatic phenotype. The loss of early germ cells in nup1541 homozygotes appeared to occur after formation of the embryonic gonad, probably around the time of the first spermatogenic stem cell divisions. Staining of late embryos with antibodies against the germ-cell marker Vasa indicated that germ cells populate the gonad in both sexes of homozygous nup1541 embryos (Fig 6G), as in heterozygous sibling controls (Fig 6H).
Less severe male sterile allelic combinations (for example, nup1543/nup1541 or nup1544/nup1545 trans-heterozygotes; Table 1, MS*) exhibited a range of variable phenotypes indicative of post-stem cell defects, including tiny testes with no germ cells, small testes filled with apparently degenerating germ cells, or small testes with fewer than normal spermatocytes per cyst and a reduced number of cysts per testis. The semilethal allelic combinations nup1543/nup154RV5 or nup1544/nup154RV5 occasionally exhibited a meiotic arrest phenotype similar to that described for the tlp1 and tlp2 alleles (
The viable nup1541 allele also caused a strong loss-of-function phenotype in female gametogenesis. Ovaries from well-fed adult females homozygous for nup1541 were tiny and almost completely lacked egg chambers beyond oogenic stage 7 (Fig 7B; compare to wild type, Fig 7A). Most early, previtellogenic egg chambers had nurse cells and a single oocyte able to localize oocyte-specific transcripts based on accumulation of both osk and grk mRNAs in only one germ cell located at the posterior of the chamber (Fig 7E). Egg chambers in which the oocyte had begun to take up yolk were observed but were rare. Occasional nup1541 egg chambers contained two oocytes, which were always located at nearly opposite poles of the egg chamber, based on uptake of yolk (Fig 7C) or osk or grk localization (Fig 7F). Occasionally, approximately once per ovary, a nup1541 egg chamber had twice the normal number of nuclei (Fig 7B, arrow). Staining with DAPI or for ß-galactosidase activity to detect expression of the lacZ enhancer trap associated with the nup1541 allele showed double oocyte egg chambers containing more than the normal 15 nurse cells (Fig 7C), suggesting origin of bipolar egg chambers by failure to separate neighboring egg chambers rather than misspecification of additional oocytes. Older egg chambers in nup1541 ovarioles appeared to degenerate, based on the abnormal appearance of DAPI staining.
Most trans-heterozygous combinations of nup1541 or nup1545 with either Df(2L)J39 or other nup154 alleles in this study (except nup1544) exhibited a block at mid-oogenesis, or "string-of-pearls" phenotype, similar to that of nup1541 homozygotes (Table 1). The strong string-of-pearls phenotype was in contrast to the phenotypes of females homozygous for the previously described hypomorphic tlp1 or tlp2 alleles, which produce some viable eggs (
Despite the strong loss-of-function effects of nup1541 on gametogenesis, export of mRNAs from the nucleus appeared to occur in mutant early egg chambers, based on the accumulation of osk and grk mRNAs in the oocyte (Fig 7E) and staining of the cytoplasm by in situ hybridization with probe generated from bulk poly(A)+ mRNA (data not shown). In addition, import of some proteins from the cytoplasm to the nucleus appeared to occur in the mutant early egg chambers, based on nuclear localization of the NLS-containing ß-galactosidase protein expressed from the enhancer trap associated with the nup1541 P-element insert (Fig 7C).
Molecular nature of nup154 alleles:
To explore the molecular basis for the striking phenotypic differences among nup154 alleles, we precisely mapped the P-element insertion sites in seven different nup154 alleles. Plasmid rescue and sequencing of the DNA flanking each of the nup154 P elements revealed that all seven inserts mapped within 48 nucleotides of one another (Fig 8A), ~100 bp upstream from the start of the predicted nup154 open reading frame (Fig 8B). The P element appeared to be inserted within the nup154 5'-UTR based on results of 5'-RACE using RNA from wild-type adult females. The 5' end of the 5'-RACE product (Fig 8A, thick bent arrow) obtained from several independent experiments was located 146 bp upstream of the start of the predicted nup154 open reading frame, a position 15 bp upstream of the site of the P-element insertion in nup1541. The close clustering of the P-element inserts raised the question of why some alleles were lethal while others were viable. In addition, the strong effects of the nup1541 and nup1545 alleles on gametogenesis raised the question of how some alleles of an essential gene might have little effect on viability but behave as strong loss-of-function mutations with respect to germline differentiation.
Whether the P-element-induced nup154 alleles were lethal or viable but sterile appeared to correspond with the orientation of the P-element insert. The three lethal P-element-induced alleles nup1543, nup1544, and nup1546 contained inserts in a 3' to 5' orientation with respect to the P-element repeats (Fig 8A, black inserts with arrows), opposite to the direction of nup154 transcription. Conversely, in the viable but sterile alleles nup1541, nup1545, tlp1, and tlp2, the inserts were oriented in the same 5' to 3' orientation as the direction of nup154 transcription (Fig 8A, white inserts with arrows). In the lethal allele nup1543 and the viable allele tlp2, the P elements were inserted at the same nucleotide but in opposite orientations. Lethality vs. viability did not correlate with either the orientation of the lacZ or eye color marker genes or the type of P-element vector (e.g., the viable alleles nup1541 and nup1545 contain PlacW inserts, whereas the viable alleles tlp1 and tlp2 contain PZ inserts).
Initiation of nup154 transcription from an ectopic site within the P-element insert in nup154 viable but sterile alleles:
If the genomic DNA corresponding to the end of the 5'-RACE product obtained from wild-type adult females encodes the start site for transcription of nup154 in all tissues, then the nup154 P-element inserts would be localized within the 5'-UTR and would be predicted to disrupt normal nup154 transcription. The correlation between P-element orientation and viability of certain nup154 insertion alleles suggested a hypothesis: the 3' end of the P elements might provide a promoter for transcription of the adjacent nup154 gene that rescues the lethal phenotype expected if the inserts disrupt the normal transcript. The P-element vectors PlacW and PZ are identical in sequence for 239 bp at their 3' ends. As both PlacW and PZ inserts gave viable alleles only when inserted in a 5' to 3' orientation, this common 3' end would be the most likely site of sequences capable of initiating transcription.
Analysis of P-element sequences deleted in lethal vs. viable w- eye color revertants generated by mobilization of the nup1541 insert (MATERIALS AND METHODS) supported the hypothesis that the viability of the nup1541 allele was due to action of an internal promoter in the P-element 3' end. Two lethal eye color revertants both had small deletions that removed sequences from the 3' end of the P-element insert (including the first HindIII and BamHI sites up through the SacII and EcoRI sites at the w+ gene junction) and some DNA between the genomic SalI site 19 bp downstream, but retained 5' P-element sequences and flanking genomic DNA (Fig 8B, right gray bar). In contrast, all 10 independent viable but sterile eye color revertants of nup1541 characterized had deletions of the 5' portion of the P element and varying extents of 5'-flanking genomic DNA, but never deleted the 3' P end of the P element (0/10 lost the 3' HindIII and BamHI sites, and only 5/10 altered the internal SacII and EcoRI sites) or 3'-flanking genomic DNA (Fig 8B, left gray bar). Some of the viable but sterile revertants had deletions that removed flanking genomic DNA corresponding to the putative nup154 transcription start site based on the sequence of the wild-type 5'-RACE product (Fig 8B, thick bent arrow), consistent with the proposed expression of nup154 from an ectopic promoter within the 3' end of the P-element insert.
To further test the hypothesis of an ectopic promoter in the P element, we performed 5'-RACE on RNA derived from whole adult flies homozygous for either nup1541 (a PlacW insert) or tlp2 (a PZ insert) using gene-specific primers to the nup154 cDNA (Fig 8B). In all cases, production of the RACE product was dependent on addition of reverse transcriptase to the initial reaction. The size of the major 5'-RACE product was consistent with transcription starting within the 239 bp of the conserved P-element 3' end, assuming no introns. As predicted, sequence analysis revealed that the major 5'-RACE products from both nup1541 and tlp2 were chimeric, containing nup154 sequences attached to the 3' part of the P element (Fig 8, B–D). After crossing the boundary into the P-element 3' end, sequence from primers in nup154 became unreadable, perhaps due to secondary structure of P-element sequences. However, using primers within the P-element 3' end, we were able to sequence the 5' end of the RACE products from both nup1541 and tlp2 homozygotes.
In two independent experiments from nup1541 homozygotes, the 5' ends of the RACE products were located 74 bp from the P-element 3' end (Fig 8C and Fig D, single asterisk). Inspection of the P-element sequences at this site revealed a TCACTC element at 71–76 bp from the P-element 3' end that resembled the transcription initiator element TCA G/TT T/C (
Consistent with occasional initiation of transcription from upstream of the consensus initiator elements, we were able to obtain RT-PCR products of the expected sizes from nup1541, tlp1, and tlp2 homozygotes using a primer to nup154 cDNA sequences and one of several different primers corresponding to sequences more 5' within the conserved P-element 3' end (Fig 8C and Fig D, arrows 1–3). Generation of the PCR products was always dependent on addition of reverse transcriptase in the cDNA synthesis reaction. Use of a primer directed against sequences specific to PlacW never yielded RT-dependent products, although the same primer was able to amplify nup1541 genomic DNA in control experiments (Fig 8C, arrow 4).
We propose that zygotic expression of an altered nup154 transcript starting in the 3' end of the P element may be sufficient to rescue lethality in the viable P-element-induced nup154 alleles, but not sufficient to satisfy the requirement for nup154 function in gametogenesis. The male and female sterility associated with the viable P-element insertions may be due to ineffective operation of the P-element 3' end internal promoter(s) in the germline. Indeed, Northern blot analysis revealed a drastic reduction in nup154 mRNA levels in ovaries from tlp1 homozygous females compared to wild type (Fig 8E).
| DISCUSSION |
|---|
| TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Drosophila nup154 is an essential, conserved nucleoporin:
nup154 encodes a nucleoporin based on both homology to yeast and vertebrate nucleoporins and localization of Nup154 to the nuclear pores (
nup154 function is required in proliferating tissues:
The lethal phenotype of the strong loss-of-function nup154 alleles resembled the phenotypes of mutants that affect mitosis (
Expression of nup154 mRNA appeared to be developmentally regulated, both in overall levels (Fig 3) and specific tissue distribution (Fig 4 and Fig 5). However, Nup154 protein was present in all cells and stages examined (data not shown). We postulate that Nup154 protein perdures once incorporated into stable assembled NPCs. Thus nup154 mRNA synthesis may be required only in cells where there is a need for new NPC assembly, such as growing or dividing cells. The correlation between the expression pattern of nup154 mRNA in proliferating tissues and the phenotypic effects of the strong loss-of-function alleles is consistent with a requirement for newly synthesized nup154 in dividing cells for assembly of new NPCs.
We did not detect global defects in nucleocytoplasmic export or import in the female germline of strong loss-of-function alleles. This could be because of residual nup154 function due to stability of the protein, as mentioned above, or because nup154 mutations might impair transport of only certain mRNAs or proteins rather than completely disrupting NPC transport per se. Fourteen nuclear transport receptors with distinct classes of substrates and nucleoporin binding capabilities have been identified to date (
subunits that show differential, tissue-specific expression (
Defects in cell proliferation in nup154 mutants may be due in part to indirect effects on the ability of nuclear transport receptors to translocate specific proteins required in mitotically active cells. In yeast, loss of NUP170 function had a reported decrease in growth rate ( transporter resulted in G2/M arrest at the restrictive temperature ( (oho31) shares an overlapping expression pattern with nup154 mRNA, being localized to just below the pole cells in the early embryo and expressed in the embryonic and postembryonic CNS (
Molecular basis for different nup154 alleles:
The phenotypes observed in nup154 alleles did not fall into a simple allelic series reflecting quantitative differences. Even for the same allele, nup154 mutants caused strikingly different effects in different tissues. Our results support the possibility that this variability might result from molecular features particular to specific P-element insert alleles that could cause different levels of nup154 expression depending on the insertion site and insert orientation. Specifically, flies with a P element inserted in the nup154 5'-UTR in the same 5' to 3' orientation as the endogenous gene may be viable due to expression of a nup154 transcript in somatic cells from ectopic promoter(s) contained within the P-element 3' end (Fig 8). The different degree of defects in early stages of germ-cell differentiation in males vs. females for a given allele may be due to differential effectiveness of the P-element internal promoter in male vs. female germ cells. Similarly, in females, the P-element internal promoter may function more effectively in early egg chambers than in later germ cells.
Consistent with the hypothesis that an outward-directed P-element-promoter(s) can drive expression of neighboring genes, we identified chimeric nup154 transcripts with 5' ends originating from sites within the P-element 3' end. Although there appeared to be multiple transcription start sites, in several cases the ends of the 5'-RACE products were located at or near consensus promoter elements in the P-element 3' end in both PlacW- and PZ-induced insertion alleles. The TCACTC initiator elements located at 71–76 and 110–115 nucleotides in from the P-element 3' end are identical to the functional transcription initiation element demonstrated for the adenovirus major late promoter. The TCAATA element at 100–105 nucleotides is identical to sequences at the transcription initiation sites of at least six different arthropod genes (
A cryptic promoter within the P-element 3' end has also been implicated in the expression of a neighboring chromosomal gene in the case of the E-32 PZ insert in the out-at-first locus (
Our results indicate that ectopic expression from promoter(s) within the P element can explain the generation of alleles with weak phenotypes by some P-element insertions, especially viable but sterile alleles of essential loci. This mechanism may account for some cases among the extensive collection of P-element-induced mutations in which phenotypic differences are observed between different P-element inserts in the same or nearby locations.
| ACKNOWLEDGMENTS |
|---|
We thank J. Hackstein for the nup1541 allele, C. Wood for recombination mapping nup1541, M. Fogarty for generating w- revertants of nup1541, and C. Tong for assistance with molecular aspects of this project. We are grateful to Ü. Schäfer for providing rpL9 deficiencies and rescue constructs, S. DiNardo for testis lacZ enhancer trap marker lines, and R. Lehmann for anti-Vasa antibody and cDNAs for making oskar RNA probes. The Stanford PAN facility generated oligonucleotides and helped with DNA sequencing. A.A.K. is grateful to M. Rexach, D. Traver, and the Fuller lab members for their helpful discussions and reviews of the manuscript. A.A.K. was supported by a Howard Hughes Medical Institute Predoctoral Fellowship. This work was supported by seed money from the Stanford Office of Technology and Licensing and by National Institutes of Health grants ROI-GM56500 (to M.T.F.) and POI-DK53074 (Irving Weissman, P.I.).
Manuscript received April 22, 1999; Accepted for publication June 21, 1999.
| LITERATURE CITED |
|---|
| TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
AITCHISON, J., M. ROUT, M. MARELLI, G. BLOBEL, and R. W. WOZNIAK, 1995 Two novel related yeast nucleoporins Nup170p and Nup157p: complementation with the vertebrate homologue Nup155p and functional interactions with the yeast nuclear pore-membrane protein Pom152p. J. Cell Biol. 131:1133-1148
AKEY, C. W. and M. RADERMACHER, 1993 Architecture of the Xenopus nuclear pore complex revealed by three-dimensional cryo-electron microscopy. J. Cell Biol. 121:1-19
ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MYERS, and D. J. LIPMAN, 1990 Basic local alignment search tool. J. Mol. Biol. 215:403-410[Medline].
BELANGER, K. D., M. A. KENNA, S. WEI, and L. I. DAVIS, 1994 Genetic and physical interactions between Srp1p and nuclear pore complex proteins Nup1p and Nup2p. J. Cell Biol. 126:619-630
BERGSTROM, D. E., C. A. MERLI, J. A. CYGAN, R. SHELBY, and R. K. BLACKMAN, 1995 Regulatory autonomy and molecular characterization of the Drosophila out at first gene. Genetics 139:1331-1346[Abstract].
BIER, E., H. VAESSIN, S. SHEPARD, K. LEE, and K. MCCALL et al., 1989 Searching for pattern and mutation in the Drosophila genome with a P-lacZ vector. Genes Dev. 3:1273-1287
BROWN, N. H. and F. C. KAFATOS, 1988 Functional cDNA libraries from Drosophila embryos. J. Mol. Biol. 203:425-437[Medline].
BURKE, T. W. and J. T. KADONAGA, 1996 Drosophila TFIID binds to a conserved downstream basal promoter element that is present in many TATA-box-deficient promoters. Genes Dev. 10:711-724
CHERBAS, L. and P. CHERBAS, 1993 The arthropod initiator: the capsite consensus plays an important role in transcription. Insect Biochem. Mol. Biol. 23:8
