Journal of Cell Science 110, 229-237 (1997) 229
Printed in Great Britain © The Company of Biologists Limited 1997
Enhancement of the ncdD microtubule motor mutant by mutants of αTub67C
Donald J. Komma and Sharyn A. Endow*
Department of Microbiology, Duke University Medical Center, Durham, North Carolina 27710, USA
*Author for correspondence (e-mail: firstname.lastname@example.org)
Ncd is a kinesin-related microtubule motor protein progression and defective pronuclear conjugation or
required for chromosome segregation in Drosophila oocytes fusion. Delayed completion of meiosis, together with failure
and early embryos. In tests for interactions with other of pronuclear fusion, prevents normal interactions of
proteins, we ﬁnd that mutants of αTub67C, which affect an maternal with paternal chromosomes, enhancing the ncdD
oocyte- and early embryo-speciﬁc α-tubulin, enhance mutant phenotype. The genetics and cytology of doubly
meiotic nondisjunction and zygotic loss of ncdD, a partial mutant embryos and the molecular defect of NcdD provide
loss-of-function mutant of ncd. The enhancement is evidence for interaction of Ncd with αTub67C in vivo.
dominant and allele-speciﬁc with respect to αTub67C, and These results imply that a speciﬁc α-tubulin isoform is
depends on the recessive effects of ncdD. Cytologically, required for normal cellular function of a kinesin motor
embryos of αTub67C/+ show delayed meiotic divisions and protein.
defective female pronucleus formation, while meiotic
spindle assembly is abnormal in embryos of ncdD/ncdD.
Doubly mutant αTub67C ncdD/ncdD embryos are rescued Key words: Ncd, Kinesin microtubule motor, α-tubulin, Mutant
for female pronucleus formation, but show delayed meiotic interaction
INTRODUCTION Other kinesin proteins are thought to help position chromo-
somes for meiotic segregation (Zhang et al., 1990; Murphy and
Several proteins have now been identiﬁed that function during Karpen, 1995; Afshar et al., 1995) or, in mitosis, to separate
the meiotic divisions in Drosophila oocytes. Nonclaret disjunc- centrosomes for spindle formation (Enos and Morris, 1990;
tional (Ncd) is a member of the kinesin family of microtubule Hagan and Yanagida, 1990; Heck et al., 1993), assemble
motors that is required for normal assembly of meiotic spindles bipolar spindles (Saunders and Hoyt, 1992), or maintain coun-
in Drosophila oocytes. Unlike conventional kinesin, Ncd teracting forces needed to prevent spindle collapse (Saunders
translocates toward microtubule minus, rather than plus ends and Hoyt, 1992).
(Walker et al., 1990; McDonald et al., 1990) and generates Binding of kinesin motors to different isoforms of tubulin
torque as it moves, causing microtubules gliding on Ncd-coated for force generation and movement on microtubules may con-
coverslips to rotate as they glide (Walker et al., 1990). Antibody tribute to the regulation of kinesin motor protein activity.
staining experiments (Hatsumi and Endow, 1992a; Endow et al., Multiple forms of α- and β-tubulin are present in many eukary-
1994a; Matthies et al., 1996) and analysis of oocytes and otic cells, and are thought to be needed for assembly of micro-
embryos expressing Ncd as a fusion with the green ﬂuorescent tubules for speciﬁc cellular functions (Fuller et al., 1988; Hoyle
protein (GFP) (S. A. Endow and D. J. Komma, unpublished; and Raff, 1990). Oocytes and early embryos of Drosophila
Endow and Komma, 1996) have demonstrated localization of contain three α-tubulins: αTub67C, αTub84B and αTub84D.
the Ncd motor protein to meiotic spindles in oocytes, and One of these, αTub67C, is expressed only in the ovary
mitotic spindle ﬁbers, centrosomes, and spindle poles in early (Kalfayan and Wensink, 1982) and is speciﬁc to nurse cells,
embryos. Oocyte spindles of ncd null mutants exhibit multiple, oocytes and early embryos, comprising ~25% of the α-tubulin
diffuse, or broad poles (Wald, 1936; Kimble and Church, 1983; in early embryos (Matthews et al., 1989). αTub67C is highly
Hatsumi and Endow, 1992b; Matthies et al., 1996), supporting divergent compared to other α-tubulins (Kalfayan and
the hypothesis that the motor is required for assembly of bipolar Wensink, 1981) with major differences at the N- and C-termini
meiotic spindles (Verde et al., 1991; Hatsumi and Endow, (Theurkauf et al., 1986). Mutants of αTub67C have been
1992a; Hyman and Karsenti, 1996). Early mitotic spindles of recovered that cause instability of the protein, resulting in
null mutants show centrosome loss, precocious splitting of cen- reduced αTub67C levels in embryos (Matthews et al., 1993).
trosomes, and chromosome loss at metaphase (Endow et al., The abnormal spindles and chromosome clusters of mutant
1994a; Endow and Komma, 1996), indicating that the Ncd embryos indicate that αTub67C is required for normal meiotic
motor functions to maintain spindle pole integrity in mitosis and and mitotic divisions of oocytes and embryos, paralleling the
may act to prevent chromosome loss by maintaining chromo- requirement for Ncd.
some attachment to the spindle in metaphase. To obtain information regarding functional interactions of
230 D. J. Komma and S. A. Endow
the Ncd motor with tubulin proteins in vivo, we analyzed Statistical analysis
genetic interactions between ncd mutants and tubulin mutants. Chromosome segregation data were analyzed using χ2 tests for data
Our results provide evidence that functional interactions of in which expected values were ≥5 in tests of the null hypothesis that
Ncd with microtubules in vivo require the oocyte- and early the offspring under comparison were produced by females from the
embryo-speciﬁc α-tubulin, αTub67C. These observations con- same population. Expected frequencies were calculated as averages
stitute the ﬁrst evidence that functional interactions of a kinesin of the observed frequencies. For data in which an expected value was
protein with microtubules require an isotype-speciﬁc α- <5, statistical tests were carried out by assuming a Poisson distribu-
tion. The probability of observing i offspring in a class, P(i), is (e−m
tubulin. mi)/(i!) for a standard Poisson distribution, where m and i are the
numbers expected and observed, respectively. m is calculated as
(sample size) × (expected frequency).
MATERIALS AND METHODS Antibody staining of oocytes and embryos
Oocytes were collected, dechorionated, ﬁxed in formaldehyde/EGTA
without taxol, and vitelline membranes were removed by hand, as
Mutants used in this work are described by Lindsley and Zimm described previously (Hatsumi and Endow, 1992b). In some experi-
(1992). ncdD was originally isolated as an EMS-induced dominant ments, oocytes were activated in hypotonic saline prior to ﬁxation
female-sterile mutant but now shows only weak semi-dominant and (Hatsumi and Endow, 1992b). Staining with monoclonal anti-α-
recessive effects on meiotic chromosome segregation, and is nearly tubulin and polyclonal anti-Ncd antibodies, followed by DAPI (4′,6-
wild type for mitotic chromosome distribution (Komma et al., 1991). diamidino-2-phenylindole) was as described (Hatsumi and Endow,
The X chromosome and chromosome 2 in the ncdD stock were 1992a; Endow et al., 1994a). The α-tubulin antibody (Chemicon Inter-
replaced with Oregon R chromosomes by mating to balancer lines. national Inc.) cross-reacts with all isoforms of α-tubulin in Drosophila
The αTub67C mutants, αTub67C1, αTub67C2 and αTub67C3, were and the Ncd antibody is speciﬁc for the nonconserved N-terminal tail
obtained from the Bloomington Stock Center. The X chromosome, of Ncd. The Ncd antibody cross-reacts with a single major band on a
chromosome 2, and the tip of chromosome 3 distal to ca in the western blot of Drosophila proteins (Hatsumi and Endow, 1992a) and
αTub67C mutant stocks were replaced with Oregon R chromosomes shows no speciﬁc cross-reactivity with spindles of oocytes or embryos
by recombination and mating to balancer stocks. Recombination with of the cand null mutant (Hatsumi and Endow, 1992a; Endow et al.,
the Oregon R chromosome 3 removed the lethal associated with 1994a). Control embryos of cand for the present experiments were
αTub67C3 (Matthews et al., 1993). The original or recombinant stained by the α-tubulin antibody but showed no staining with the Ncd
αTub67C chromosomes were recombined with the ncdD chromosome antibody.
to produce the αTub67C ncdD chromosomes, followed by replace- Wild-type embryos were collected at 30 minute intervals and
ment of the X and chromosome 2 with Oregon R chromosomes, as mutant embryos, at 30 minute, 60 minute or 90 minute intervals.
described (Komma and Endow, 1995). Following dechorionation, vitelline membranes were removed and
The deﬁciency chromosomes, Df(3L)AC1 and Df(3R)Scx2, were embryos were ﬁxed in methanol/EGTA without taxol and stained with
obtained from the Bloomington Stock Center. Df(3L)AC1, cytologi- antibodies and DAPI, as described (Hatsumi and Endow, 1992b;
cally Df(3L)67A2;67D11-13, uncovers αTub67C (Matthews et al., Endow et al., 1994a). In some experiments, embryos were stained
1993). The deﬁciency chromosome was recombined with an ncdD with monoclonal anti-histone antibody (1:200 dilution) (Chemicon
chromosome to produce a Df(3L)AC1 ncdD chromosome. Df(3R)Scx2 International Inc.), followed by FITC-conjugated anti-mouse IgG
uncovers αTub84B (Matthews and Kaufman, 1987). Cytologically, (1:250 dilution, 6 µg/ml) (Vector Lab. Inc.) to visualize the chromo-
Df(3R)Scx2 is Df(3R)84A4-5;84C1-2. The deﬁciency chromosome somes.
was recombined with an ncdD chromosome to produce a Df(3R)Scx2
ncdD chromosome. Confocal microscopy
Images of antibody-stained spindles and chromosomes were collected
Embryo viability and chromosome segregation tests with a Bio-Rad MRC 600 laser scanning confocal imaging detector
Embryo viability and X chromosome segregation tests were carried mounted on a Zeiss Axiophot microscope, using a ×63/1.4 NA
out as described (Endow et al., 1994a; Komma et al., 1991) by mating Planapochromat objective.
wild-type or mutant females to y2 wbf/BSY males. Single-pair matings
were transferred to new vials on days 3-7, eggs were counted, and Embryo squashes for chromosomes
offspring from each vial were scored for phenotype. Regular offspring Chromosomes in ﬁxed embryos were visualized as described (Komma
were + females and BS males. Nondisjunction of X chromosomes in and Endow, 1995). Brieﬂy, antibody-stained embryos were mounted
oocytes results in X,X and nullo-X eggs that give rise to X/X/BSY in 10 mM Tris-HCl, pH 7.9 + 1 mM EDTA (TE) on Denhardt-treated
females and X/0 males carrying a paternal X chromosome. X/0 males slides (Brahic and Haase, 1978) using siliconized coverslips. Embryos
in excess of X/X/BSY females are attributed to meiotic loss (Sturte- were staged under ﬂuorescence and squashed in situ, coverslips were
vant, 1929). Gynandromorphs are X chromosome mosaics that arise removed, and embryos were post-ﬁxed up to 1 hour in cold EtOH.
following chromosome loss in early mitosis. + males carrying a Slides were mounted in TE containing 5 µg/ml DAPI and chromo-
maternal X chromosome are due to paternal X or Y chromosome loss, somes were photographed onto 4′′ × 5′′ Tri-X ﬁlm (Kodak 4164 Pan
resulting in nullo-X,Y sperm. Exceptional y w female offspring of Professional Film). Negatives were scanned into digital images using
αTub67C/+ females are attributed to fusion of paternally-derived a Sharp JX-320 scanner.
haploid nuclei, or failure of haploid nuclei to segregate, in haploid
embryos that arise following defective female pronucleus formation Image processing
(Komma and Endow, 1995). Calculations of total embryos and fre- Image contrast was adjusted using Bio-Rad SOM or COMOS
quencies of gametic nondisjunction and loss were corrected for invi- software or Adobe Photoshop, and images were printed using a
ability of half of the nondisjunctional and meiotic loss embryos as 0/Y Tektronix Phaser IISDX printer. For meiotic spindles or polar bodies
or X/X/X embryos. Minute (haplo-4) offspring were scored but were that were not in the same focal plane, confocal images collected in
excluded from calculations of chromosome mis-segregation because successive focal planes were copied to ‘layers’ of one image (Adobe
of their highly variable recovery. Photoshop v. 3.0) and merged using the ‘screen’ mode option and
Mutants of αTub67C enhance ncdD 231
RESULTS Table 1 also shows that ncdD/ncdD enhances the production
of exceptional androgenetic y w females by αTub67C1/+ and
αTub67C mutants enhance ncdD meiotic αTub67C3/+, but not by αTub67C2/+. These exceptional
nondisjunction females are attributed to fusion of paternally-derived haploid
ncdD/ncdD females produce frequent X/X/Y and X/0 (BS cleavage nuclei, or failure of newly replicated haploid chro-
female and y w male) offspring caused by nondisjunction of mosomes to segregate, in embryos defective for female pronu-
the X chromosome in meiosis; the frequency of these cleus formation (Komma and Endow, 1995).
offspring was 0.108 in the cross shown in Table 1. Genetic αTub67C ncdD/+ females heterozygous for ncdD and
tests for interactions between ncdD and αTub67C mutants αTub67C1, αTub67C2, or αTub67C3 did not show a signiﬁcant
showed signiﬁcantly increased frequencies of nondisjunc- increase in meiotic chromosome nondisjunction, compared
tional offspring in progenies of αTub67C1 ncdD/ncdD and with ncdD/+ females (Table 1). The zygotic loss of αTub67C
αTub67C3 ncdD/ncdD females compared with ncdD/ncdD ncdD/+ females is also not signiﬁcantly elevated relative to
females, 0.194 and 0.149, respectively, in the data shown in αTub67C/+ females (for zygotic loss of αTub67C3 ncdD/+
Table 1. Zygotic loss of the X chromosome was also signiﬁ- compared to αTub67C3/+, P=0.09). The genetic interaction
cantly increased in these progenies with frequencies of 0.021 between ncdD and αTub67C1 or αTub67C3 that results in
for αTub67C1 ncdD/ncdD and 0.015 for αTub67C3 elevated meiotic nondisjunction and zygotic loss therefore
ncdD/ncdD, compared with 0.001 for ncdD/ncdD. The meiotic depends on the recessive rather than the weak semi-dominant
nondisjunction of αTub67C2 ncdD/ncdD females did not differ effects of ncdD (Komma et al., 1991). The meiotic nondis-
signiﬁcantly from that of ncdD/ncdD females (χ2=1.12, 1 d.f., junction of αTub67C2 ncdD/+ females is signiﬁcantly
0.5>P>0.1) in the data shown in Table 1, nor did zygotic X decreased compared with nondisjunction of αTub67C2/+
chromosome loss (P=0.06). The enhancement of ncdD females; the basis of this decrease is not certain from the
meiotic nondisjunction and zygotic loss by αTub67C is present data.
therefore allele-speciﬁc with respect to αTub67C. The allele- Tests of αTub67C3cand/cand or αTub67C3 ncd2/ncd2 females
speciﬁc enhancement of ncdD by αTub67C mutants was carrying null alleles of ncd resulted in female sterility and were
observed in repeated experiments carried out over a period of therefore uninformative with regard to the ncd allele speciﬁcity
3 years, although the magnitude of the effect was greater in for the interaction with αTub67C. No other partial loss-of-
earlier experiments than those reported in Table 1 (e.g. see function alleles of ncd were available for testing for interac-
data for αTub67C3 ncdD/ncdD reported by Komma and tions with the αTub67C mutants.
Endow, 1995, to document recovery of exceptional androge-
netic offspring). The decreased effect in recent experiments Enhancement of ncdD meiotic nondisjunction and
is presumably due to the accumulation of modiﬁers in the zygotic loss by αTub67C requires αTub67C
stocks. Tests of Df(3L)AC1, a deﬁciency that uncovers αTub67C, for
Table 1. Dominant enhancement of ncdD by αTub67C mutants
Total Zygotic Total X Total
Female parent +& BS & BS ( yw( gyn +( yw& gametes X nd X loss X loss mis-seg embryos Viability
+/+ 616 2 516 1 1,138 0.005 <0.001 <0.001 0.005 1,145 0.994
αTub67C1/+ 1,004 1 935 3 4 1 1 1,953 0.002 0.002 0.002 0.006 3,985 0.489
αTub67C1 ncdD/+ 624 1 595 2 5 2 1 1,233 0.003 0.002 0.004 0.009 3,303 0.373
αTub67C1 ncdD/ncdD 396 57 304 30 19 1 4 899* 0.194 <0.002 0.021 0.215 3,635 0.245
αTub67C2/+ 588 16 562 13 1 1,209 0.048 <0.001 0.001 0.049 1,979 0.602
αTub67C2 ncdD/+ 482 1 459 2 947 0.004 0.002 <0.001 0.006 1,827 0.518
αTub67C2 ncdD/ncdD 115 8 137 4 3 1 280 0.086 <0.004 0.011 0.096 540 0.520
αTub67C3/+ 284 238 2 1 525 <0.002 <0.002 0.004 0.004 1,751 0.300
αTub67C3 ncdD/+ 656 1 549 5 12 2 1,231 0.003 0.006 0.010 0.019 3,158 0.389
αTub67C3 ncdD/ncdD 369 42 271 16 12 9 777 0.149 <0.002 0.015 0.165 4,128 0.190
ncdD/+ 645 632 2 1,281 <0.001 0.003 <0.001 0.003 1,454 0.883
ncdD/ncdD 762 52 715 38 2 1 1,660 0.108 <0.001 0.001 0.110 3,523 0.505
The table shows offspring of matings between females of the indicated genotype and y2 wbf/BSY males in tests of X chromosome segregation. Regular offspring
are + females and BS males. Meiotic nondisjunction or loss of the X gives rise to BS (X/X/Y) females and y w (X/0) males. Gynandromorphs (gyn) (X/X-X/0
mosaics) arise upon zygotic loss of the maternal X; occasional y+ gynandromorphs are produced by zygotic loss of the paternal X. + (X/0) males are due to
meiotic loss of the paternal X chromosome. y w females are patroclinous exceptions and have been reported previously (Komma and Endow, 1995). αTub67C1
and αTub67C3 cause dominant enhancement of the effects of ncdD on meiotic nondisjunction and zygotic chromosome loss. Viability of the αTub67C1
ncdD/ncdD and αTub67C3 ncdD/ncdD mutant embryos is also reduced compared with embryos of ncdD/ncdD. αTub67C3 causes a less severe dominant
enhancement of ncdD/ncdD than αTub67C1, and its recessive mutant effects are also less severe than those of αTub67C1 (Matthews et al., 1993). The basis of the
approximate 2:1 ratio of nondisjunctional X/X/Y to X/0 offspring in the crosses of ncdD/ncdD females is not known. This is not a general characteristic of ncdD
(e.g., see Table 3). Calculations of total embryos and gametic nondisjunction and loss were corrected for inviability of half of the nondisjunctional and loss
embryos as 0/Y or X/X/X embryos. Total adults (not shown) included M (haplo-4) offspring. The data shown for αTub67C3 ncdD/ncdD and ncdD/ncdD were
pooled from two experiments.
*Includes 1 y+ gynandromorph.
232 D. J. Komma and S. A. Endow
Table 2. Enhancement of ncdD meiotic nondisjunction by αTub67C mutants requires αTub67C
Total Zygotic Total X Total
Female parent +& BS & BS ( yw( gyn +( yw& gametes X nd X loss X loss mis-seg embryos Viability
Df(3L)AC1/+ 263 215 1 1 481* <0.003 <0.003 0.002 0.002 1,551 0.500
Df(3L)AC1 ncdD/+ 432 336 7 8 790 <0.002 0.018 0.010 0.028 1,687 0.663
Df(3L)AC1 ncdD/ncdD 265 236 11 10 533 <0.002 0.041 0.019 0.060 2,270 0.426
ncdD/+ 645 632 2 1,281 <0.001 0.003 <0.001 0.003 1,454 0.883
ncdD/ncdD 762 52 715 38 2 1 1,660 0.108 <0.001 0.001 0.110 3,523 0.505
The table shows offspring from X chromosome segregation tests of females heterozygous for a deﬁciency, Df(3L)AC1, that uncovers αTub67C. The
Df(3L)AC1/+ females were wild type, heterozygous, or homozygous for ncdD. Data for ncdD/+ and ncdD/ncdD females from Table 1 are shown for comparison.
Df(3L)AC1 enhances the effect of ncdD on gametic and zygotic loss of the X, but suppresses the effect of ncdD on gametic nondisjunction.
*Includes 1 y+ gynandromorph.
interactions with ncdD showed signiﬁcantly increased fre- development, and is found in all tissues including oocytes and
quencies of meiotic and zygotic chromosome loss offspring in embryos. Df(3R)Scx2 ncdD/ncdD females, heterozygous for the
progenies of Df(3L)AC1 ncdD/+ and Df(3L)AC1 ncdD/ncdD αTub84B deﬁciency and homozygous for ncdD, showed no sig-
females, compared to ncdD/+ and ncdD/ncdD females (Table niﬁcant differences compared to ncdD/ncdD females with
2). The increased zygotic chromosome loss observed in respect to meiotic chromosome nondisjunction (χ2=0.14, 1 d.f.,
progenies of αTub67C ncdD/ncdD females mutant for 0.9>P>0.5) or zygotic loss (P=0.18) (Table 3). These results
αTub67C1 or αTub67C3 and ncdD can therefore be attributed show that the enhancement of gametic and zygotic chromo-
to loss of αTub67C function. some loss of ncdD by Df(3L)AC1 is not observed for
Nondisjunctional offspring are not increased in frequency in Df(3R)Scx2. The genetic interaction between ncdD and
progenies of Df(3L)AC1 ncdD/ncdD compared to ncdD/ncdD αTub67C is therefore unlikely to be a consequence of reduced
females, but instead the frequency of these offspring is sup- α-tubulin levels in general, but instead appears to be speciﬁc
pressed. Deﬁciency of αTub67C therefore enhances the for αTub67C.
meiotic and zygotic chromosome loss of ncdD and suppresses
the meiotic nondisjunction. This indicates that the enhanced Oocytes of αTub67C3/+ complete the meiotic
meiotic nondisjunction of ncdD by αTub67C1 and αTub67C3 divisions but are defective in female pronucleus
requires the αTub67C1 or αTub67C3 mutant protein and formation
excludes the possibility that the effect is a consequence of Polar bodies and mitotic spindles were examined cytologically
reduced levels of αTub67C due to protein instability. This in early embryos of αTub67C3/+ females and compared with
interpretation is also consistent with the allele speciﬁcity of the polar bodies and spindles of wild-type embryos and αTub67C
interaction. mutant embryos to determine the effects of ncdD and αTub67C
on assembly of these microtubule-containing structures.
Enhancement of ncdD meiotic nondisjunction or Microtubules were stained with anti-tubulin and anti-Ncd anti-
zygotic loss is not observed for a deﬁciency of bodies, and chromosomes were stained with DAPI or anti-
αTub84B histone antibody.
Tests of Df(3R)Scx2, a deﬁciency that uncovers αTub84B, were Following completion of the oocyte meiotic divisions, the
carried out to determine whether the interaction between ncdD female pronucleus forms from the inner-most of the four
and αTub67C was speciﬁc for αTub67C or could be observed haploid nuclei (Sonnenblick, 1950). The three remaining
for other α-tubulin genes. αTub84B encodes an α-tubulin that nuclei condense into chromosomes and assemble into polar
is constitutively expressed in Drosophila tissues throughout bodies that consist of 1 or 2 haploid sets of chromosomes sur-
Table 3. A deﬁciency of αTub84B does not enhance ncdD meiotic nondisjunction or zygotic loss
yw( +( Total Zygotic Total X Total
Female parent + & BS & BS ( (or + () g y n (or y w () y w & gametes X nd X loss X loss mis-seg embryos Viability
Df(3R)Scx2/+ 907 836 4 1,747 <0.001 <0.001 <0.001 <0.001 3,094 0.565
Df(3R)Scx2 ncdD/+ 645 584 1,229 <0.001 <0.001 <0.001 <0.001 2,187 0.562
Df(3R)Scx2 ncdD/ncdD 249 15 287 10 586 0.085 <0.002 <0.002 0.085 2,349 0.250
*ncdD/ncdD 534 24 531 23 2 1,161 0.081 <0.001 0.002 0.083 2,168 0.583
The table shows offspring from X chromosome segregation tests of females heterozygous for a deﬁciency, Df(3R)Scx2, that uncovers αTub84B. The
Df(3R)Scx2/+ females were wild type, heterozygous, or homozygous for ncdD. Df(3R)Scx2 shows no enhancement of the effect of ncdD on meiotic
nondisjunction or zygotic chromosome loss, although viability of embryos doubly mutant for Df(3R)Scx2 and ncdD is signiﬁcantly reduced relative to that of
*Females were y2 wbf; ncdD mated to +/BSYy+ males. X/0 males due to maternal X loss were therefore + in phenotype and X/0 males due to paternal X loss
were y w.
Mutants of αTub67C enhance ncdD 233
Fig. 1. Polar bodies and cycle 1 mitotic spindles of wild-type and αTub67C mutant embryos. Wild-type and mutant early embryos were stained
with anti-tubulin (A-D) and anti-histone antibody (A,C) or DAPI (B,D). The paired images, taken from different embryos, show tubulin
staining (left) and histone or DAPI staining (right). (A) Wild-type polar body. The oocyte chromosomes in the polar body are oriented with
centromeres inward and are embedded in a mass of short microtubules. Wild-type embryos contain 2-3 polar bodies. (B) Wild-type cycle 1
mitotic spindle. Two closely apposed spindles are assembled around the maternal and paternal chromosomes. The two sets of chromosomes
remain separate until the end of the ﬁrst mitotic division. The spindle is in metaphase and the chromosomes, late prometaphase. (C) αTub67C2
mutant polar body. The polar body is much larger than wild type and contains >4N chromosomes. Longer than normal microtubules radiate
from the polar body. (D) Small cycle 1 mitotic spindle in an αTub67C2 mutant embryo, containing a haploid set of chromosomes. The spindle
contains centrosomes and is interpreted to be associated with paternal chromosomes. The centrosomes are large compared with wild type.
Microtubules of the spindle and asters are unusually long. DAPI-stained chromosomes (B,D) are the same magniﬁcation; remaining images are
the same magniﬁcation as each other. Bars, 10 µm.
rounded by microtubules. A wild-type polar body is shown in blick, 1950). The spindles were associated with a haploid set
Fig. 1A. Separate spindles assemble around the maternal and of chromosomes (Fig. 1D). Mutant embryos of αTub67C3
paternal chromosomes, which lie in close apposition to one females frequently contained many small mitotic spindles
another, forming the ﬁrst cleavage division spindle (Fig. 1B). associated with haploid chromosome sets (Komma and Endow,
The maternal and paternal chromosomes remain separate from 1995) (not shown), instead of a single small spindle. The
one another until the end of the ﬁrst mitotic division. multiple small spindles have been interpreted as arising from
Embryos of αTub67C females, homozygous for αTub67C2 cleavage divisions of the haploid spindles (Komma and Endow,
or αTub67C3, typically contained a single large polar body 1995). Polar bodies and spindles of αTub67C mutant embryos
with >4N chromosomes (Fig. 1C), indicating that the chromo- were brightly stained by both anti-tubulin and anti-Ncd anti-
somes had failed to undergo the meiotic divisions, or had bodies.
undergone the meiotic divisions but the resulting haploid nuclei Normal-appearing meiosis II spindles were observed in
had fused with one another, and had replicated in the polar early (0-30 minutes) embryos of αTub67C3/+ females (Fig.
body. The meiotic defect, or defect in behavior of the meiotic 2A), indicating that mutant embryos can complete the meiotic
products, results in failure to form a female pronucleus. Con- divisions. Meiotic progression is delayed relative to wild type,
sistent with this interpretation, a single small mitotic spindle however, since meiosis I or II stages were frequently observed
containing centrosomes and associated with long astral micro- among 0-30 minute αTub67C3/+ mutant embryos (n=5,
tubules was observed in the αTub67C mutant embryos (Fig. total=18, frequency=0.278), but were infrequent among 0-30
1D). The presence of centrosomes indicated that the spindle minutes wild-type embryos (n=6, total=74, frequency=0.081).
was associated with paternal chromosomes since oocyte Some αTub67C3/+ mutant embryos contained 4 microtubule-
meiotic spindles are acentriolar and centrosomes for the mitotic associated clusters of chromosomes (Fig. 2B) near the anterior
divisions are derived from the inseminating sperm (Sonnen- dorsal surface of the embryo where the polar bodies lie in wild-
234 D. J. Komma and S. A. Endow
Fig. 2. Meiosis II, polar bodies and haploid mitotic spindle of αTub67C3/+ mutant embryos. (A) Meiosis II spindles in an αTub67C3/+ mutant
embryo. The spindles appear normal. (B) Four polar body-like structures associated with haploid chromosome sets following completion of the
oocyte meiotic divisions. The maternal chromosomes remain in polar bodies rather than contributing to female pronucleus formation. (C) Small
mitotic spindle with a dissociating centrosome (arrow). The spindle was associated with a haploid set of chromosomes and was probably
assembled around paternal chromosomes. Three polar bodies, two with 1N and one with 2N chromosomes, were present in the same embryo.
Embryos were stained with α-tubulin antibody. Bar, 10 µm.
type embryos, while later (0-90 minutes) embryos had 2-3 such ated with abnormal spindles. Oocyte chromosomes in
clusters. Based on their position in the embryo, the haploid embryos of the ncd null mutant, cand, have also been observed
number of chromosomes within each body and the polar body- to lack tubulin staining or to be spindle-associated (Hatsumi
like sheath of microtubules surrounding the chromosomes, and Endow, 1992b); these effects are therefore due to loss of
these structures were presumed to be polar bodies. The number Ncd meiotic function. Polar body-like structures were
of chromosomes associated with the 2-3 polar bodies of 0-90 observed infrequently in early embryos of ncdD females (Fig.
minutes embryos accounted for the 4 haploid sets produced by 4A) and were associated with loose arrays of chromosomes,
the meiotic divisions, indicating that fusion of the haploid unlike the focused polar body chromosomes of wild type and
chromosome sets could occur to reduce the number of polar αTub67C mutants (Fig. 1A,C). The microtubules associated
bodies. This is also observed in wild-type embryos (Rabi- with the abnormal polar bodies of ncdD embryos were
nowitz, 1941). Many αTub67C3/+ mutant embryos exhibited unstained or faintly stained by α-tubulin and Ncd antibodies
2-3 polar bodies containing a total of 4 N chromosomes (Fig. 4A).
together with small mitotic spindles associated with haploid Mitotic spindles of early ncdD embryos were somewhat
sets of chromosomes, as observed for αTub67C mutant larger in size than wild-type spindles of the same stage and
embryos. The small mitotic spindles were concluded to be were ‘ragged’ in appearance, but showed Ncd antibody
associated with paternal chromosomes based on the centro- staining that closely resembled the tubulin staining (Fig. 4B),
somes present at the spindle poles. These embryos are inter-
preted to arise by cleavage divisions of the paternal chromo-
somes following failure of the female pronucleus to form.
ncdD is defective in oocyte meiotic divisions
Whole-mount ncdD oocytes stained with anti-α-tubulin and
anti-Ncd antibodies showed multipolar or broad meiotic
spindles, or diffuse spindles with undeﬁned spindle poles, as
reported previously for meiotic spindles of ncd null mutants
(Wald, 1936; Kimble and Church, 1983; Hatsumi and Endow,
1992b; Matthies et al., 1996). Bivalent chromosomes in non-
activated or activated oocytes were frequently separated from
one another and associated with separate spindles. These
abnormal spindles were brightly stained by the Ncd antibody
(Fig. 3), conﬁrming association of the NcdD motor with
meiotic spindles in whole-mount oocytes. The α-tubulin
antibody staining (not shown) was similar to the Ncd antibody
staining. Fig. 3. NcdD is associated with meiotic spindles in ncdD oocytes, but
Early embryos of ncdD lacked normal polar bodies, con- spindles are abnormal. The separated spindles were associated with
sistent with the defective meiotic spindles observed in ncdD separate bivalent chromosomes. The spindles are stained by an
oocytes. Instead, oocyte chromosomes were present as free antibody directed against the nonconserved N terminus of Ncd. Bar,
chromosomes or nuclei devoid of tubulin staining, or associ- 10 µm.
Mutants of αTub67C enhance ncdD 235
Fig. 4. Polar bodies are abnormal in early embryos of ncdD but mitotic spindles are nearly normal. Embryos were stained with α-tubulin and
Ncd antibodies. The paired images show anti-tubulin (left) and anti-Ncd antibody staining (right) in doubly stained embryos. (A) Polar body-
like structure from a cycle 1 ncdD mutant embryo, faintly stained by tubulin and Ncd antibodies. Oocyte chromosomes were associated with the
polar body. Normal polar bodies were not observed in ncdD mutant embryos. (B) Mitotic spindle in an early ncdD mutant embryo. The cycle 3
spindle appears ‘ragged’ compared with wild-type spindles but is normal in overall morphology. Bar, 10 µm.
as in wild-type spindles (Endow et al., 1994a). The spindles aberrant meiosis II spindle structures (n=4, total=33,
present in ncdD mutant embryos usually lacked the spurs and frequency=0.121) (Fig. 5A). Unlike wild-type meiosis II
branches associated with free chromosomes that are charac- spindles and meiosis II spindles of αTub67C3/+ mutant
teristic of the abnormal early mitotic spindles of the cand null embryos, spindles of αTub67C3 ncdD/ncdD mutant embryos
mutant (Hatsumi and Endow, 1992b) and other loss-of- persisted after onset of the ﬁrst mitotic division. Both a female
function ncd mutants (Endow and Komma, 1996). The absence and male pronucleus were observed in some mutant embryos,
of spindle spurs and branches associated with the early mitotic demonstrating rescue of female pronucleus formation
spindles of ncdD is consistent with the near absence of zygotic compared to αTub67C3/+. The maternal and paternal chromo-
chromosome loss observed in genetic tests of ncdD females somes were associated with widely separated spindles in one
(Komma et al., 1991). embryo (Fig. 5B), indicating failure of the male and female
pronuclei to conjugate and fuse. In another embryo, two pairs
αTub67C3 ncdD/ncdD mutant embryos show delayed of side-by-side telophase nuclei were associated with the same
meiotic progression and defective pronuclear cycle 1 spindle, demonstrating that the separated pronuclei
conjugation or fusion could divide, but the nuclei remained separated from one
Early (0-90 minute) embryos of αTub67C3 ncdD/ncdD females another even after completion of the ﬁrst mitotic division.
frequently showed oocyte chromosomes associated with Many embryos of αTub67C3 ncdD/ncdD females contained
Fig. 5. Embryos of αTub67C3 ncdD/ncdD contain abnormal meiotic spindles and are defective in pronuclear conjugation. (A) Meiosis II-like
spindles associated with chromosomes in the polar body region of an αTub67C3 ncdD/ncdD mutant embryo, stained with α-tubulin antibody.
(B) Two small spindles in a doubly mutant embryo. Each of the spindles was assembled around a haploid set of chromosomes. The spindle to
the left contains centrosomes at the poles and is probably associated with paternal chromosomes. The acentriolar spindle at the right was just
below the polar chromosomes and is most likely assembled around maternal chromosomes. Bar, 10 µm.
236 D. J. Komma and S. A. Endow
small mitotic spindles with centrosomes that were associated and αTub67C3 suggests that Ncd may function in wild-type
with haploid sets of chromosomes, as observed for αTub67C embryos to move pronuclei together for fusion. Ncd could act
and αTub67C/+ mutant embryos. The presence of centrosomes in a manner similar to that proposed for Kar3, a minus-end
at the poles indicated that the spindles were assembled around kinesin microtubule motor in yeast, which has been proposed
paternal chromosomes. to translocate toward the minus ends of crosslinked antiparal-
lel microtubules that emanate from the two haploid nuclei in
mating cells, moving the nuclei together for fusion (Endow et
DISCUSSION al., 1994b).
In addition to providing the ﬁrst evidence for the require-
The results presented here demonstrate that αTub67C mutants ment for a speciﬁc tubulin isotype for functional interactions
cause allele-speciﬁc dominant enhancement of ncdD meiotic with microtubules, the results imply that Ncd requires α-
nondisjunction and zygotic loss. The enhancement is dominant tubulin for motor function in vivo. Our results do not exclude
and depends on the recessive effects of ncdD. A deﬁciency that the possibility that Ncd binds to β- as well as α-tubulin to
uncovers αTub67C enhanced the chromosome loss of ncdD, but carry out function, consistent with the demonstration that a
suppressed the meiotic nondisjunction, implying that the truncated Ncd motor protein can be crosslinked to both α-
enhanced nondisjunction of ncdD by αTub67C1 and αTub67C3 and β-tubulin (Walker, 1995) and with ultrastructural analysis
requires the αTub67C protein. Chromosome mis-segregation of Ncd-decorated microtubules, interpreted as showing
of ncdD was not affected by a deﬁciency of αTub84B, an α- extensive interactions of Ncd with both tubulin subunits
tubulin that is found in all tissues including oocytes and (Hoenger et al., 1995; Hirose et al., 1995). Besides con-
embryos, indicating that the genetic interaction between ncdD tributing to the regulation of the motors, tubulin subunit
and αTub67C is speciﬁc for αTub67C rather than a conse- binding interactions may be of importance in determining the
quence of reduced or altered levels of α-tubulin in general. The polarity of translocation on microtubules of Ncd and other
allele-speciﬁcity of the interaction and requirement for kinesin motors.
αTub67C imply that the enhanced meiotic nondisjunction is
due to protein:protein interactions. This interpretation is con- We thank A. H. Yamamoto for collecting oocytes and embryos for
sistent with the molecular change in NcdD, a V556→F missense initial experiments. Supported by grants from the National Institutes
mutation in the proposed microtubule-binding region of the of Health and American Cancer Society to S.A.E. We acknowledge
use of Cancer Center confocal microscope and computer facilities to
motor (Komma et al., 1991) that causes an ~10-fold reduced carry out this work.
velocity of NcdD in in vitro motility assays relative to wild type
(Moore et al., 1996).
The simplest interpretation of our genetic and cytological REFERENCES
results is that the enhancement of ncdD meiotic nondisjunction
by αTub67C is caused by interaction of NcdD with the mutant Afshar, K., Barton, N. R., Hawley, R. S. and Goldstein, L. S. B. (1995). DNA
αTub67C1 and αTub67C3 proteins. The NcdD motor binds to binding and meiotic chromosomal localization of the Drosophila nod
microtubules in αTub67C oocytes, but the oocyte chromo- kinesin-like protein. Cell 81, 129-138.
somes fail to undergo the meiotic divisions. In αTub67C/+ Brahic, M. and Haase, A. T. (1978). Detection of viral sequences of low
reiteration frequency by in situ hybridization. Proc. Nat. Acad. Sci. USA 75,
mutant embryos, meiotic progression is delayed relative to wild 6125-6129.
type and the female pronucleus frequently fails to form, while Endow, S. A., Chandra, R., Komma, D. J., Yamamoto, A. H. and Salmon, E.
ncdD/ncdD mutant embryos show defective meiotic spindle D. (1994a). Mutants of the Drosophila ncd microtubule motor protein cause
assembly and polar body formation. αTub67C3 ncdD/ncdD centrosomal and spindle pole defects in mitosis. J. Cell Sci. 107, 859-867.
Endow, S. A., Kang, S. J., Satterwhite, L. L., Rose, M. D., Skeen, V. P. and
mutant embryos are rescued for female pronucleus formation Salmon, E. D. (1994b). Yeast Kar3 is a minus-end microtubule motor protein
but show delayed meiotic progression and defective pronuclear that destabilizes microtubules preferentially at the minus ends. EMBO J. 13,
conjugation or fusion. The delayed assembly of meiotic 2708-2713.
spindles, together with defective pronuclear conjugation or Endow, S. A. and Komma, D. J. (1996). Centrosome and spindle function of
the Drosophila Ncd microtubule motor visualized in live embryos using Ncd-
fusion, delays or prevents interactions of maternal with GFP fusion proteins. J. Cell Sci. 109, 2429-2442.
paternal chromosomes and causes a worse mutant phenotype Enos, A. P. and Morris, N. R. (1990). Mutation of a gene that encodes a
than ncdD or αTub67C/+ alone. Enhancement of the ncdD kinesin-like protein blocks nuclear division in A. nidulans. Cell 60, 1019-
mutant phenotype by αTub67C mutants can therefore be 1027.
explained by enhanced interactions of NcdD with microtubules Fuller, M. T., Caulton, J. H., Hutchens, J. A., Kaufman, T. C. and Raff, E.
C. (1988). Mutations that encode partially functional β2 tubulin subunits
caused by mutational changes in αTub67C. have different effects on structurally different microtubule arrays. J. Cell
These results imply that NcdD interacts with αTub67C in Biol. 107, 141-152.
vivo, and provide evidence that wild-type Ncd binds to Hagan, I. and Yanagida, M. (1990). Novel potential mitotic motor protein
αTub67C to carry out function. Interaction of a microtubule encoded by the ﬁssion yeast cut7+ gene. Nature 347, 563-566.
Hatsumi, M. and Endow, S. A. (1992a). The Drosophila ncd microtubule
motor protein with an isotype-speciﬁc tubulin could be motor protein is spindle-associated in meiotic and mitotic cells. J. Cell Sci.
important in regulating motor function at the cellular level. In 103, 1013-1020.
particular, the question of why the Ncd motor is required for Hatsumi, M. and Endow, S. A. (1992b). Mutants of the microtubule motor
mitotic chromosome distribution in early embryos but not later protein, nonclaret disjunctional, affect spindle structure and chromosome
in development may be due to the replacement of αTub67C in movement in meiosis and mitosis. J. Cell Sci. 101, 547-559.
Heck, M. M. S., Pereira, A., Pesavento, P., Yannoni, Y., Spradling, A. C. and
early embryos by other α-tubulin isotypes (Kalfayan and Goldstein, L. S. B. (1993). The kinesin-like protein KLP61F is essential for
Wensink, 1982; Matthews et al., 1993). The failure of pronu- mitosis in Drosophila. J. Cell Biol. 123, 665-679.
clear conjugation or fusion in embryos doubly mutant for ncdD Hirose, K., Lockhart, A., Cross, R. A. and Amos, L. A. (1995). Nucleotide-
Mutants of αTub67C enhance ncdD 237
dependent angular change in kinesin motor domain bound to tubulin. Nature like ncd protein of Drosophila is a minus end-directed microtubule motor.
376, 277-279. Cell 63, 1159-1165.
Hoenger, A., Sablin, E. P., Vale, R. D., Fletterick, R. J. and Milligan, R. A. Moore, J. D., Song, H. and Endow, S. A. (1996). A point mutation in the
(1995). Three-dimensional structure of a tubulin-motor-protein complex. microtubule-binding region of the Ncd motor protein reduces motor velocity.
Nature 376, 271-274. EMBO J. 15, 3306-3314.
Hoyle, H. D. and Raff, E. C. (1990). Two Drosophila beta tubulin isoforms are Murphy, T. D. and Karpen, G. H. (1995). Interactions between the nod+
not functionally equivalent. J. Cell Biol. 110, 1009-1026. kinesin-like gene and extracentromeric sequences are required for
Hyman, A. A. and Karsenti, E. (1996). Morphogenetic properties of transmission of a Drosophila minichromosome. Cell 81, 139-148.
microtubules and mitotic spindle assembly. Cell 84, 401-410. Rabinowitz, M. (1941). Studies on the cytology and early embryology of the
Kalfayan, L. and Wensink, P. C. (1981). α-tubulin genes of Drosophila. Cell egg of Drosophila melanogaster. J. Morphol. 69, 1-49.
24, 97-106. Saunders, W. S. and Hoyt, M. A. (1992). Kinesin-related proteins required for
Kalfayan, L. and Wensink, P. C. (1982). Developmental regulation of structural integrity of the mitotic spindle. Cell 70, 451-458.
Drosophila α-tubulin genes. Cell 29, 91-98. Sonnenblick, B. P. (1950). The early embryology of Drosophila melanogaster.
Kimble, M. and Church, K. (1983). Meiosis and early cleavage in Drosophila In Biology of Drosophila (ed. M. Demerec), pp. 62-167. New York: Hafner
melanogaster eggs: effects of the claret-non-disjunctional mutation. J. Cell Pub. Co.
Sci. 62, 301-318. Sturtevant, A. H. (1929). The claret mutant type of Drosophila simulans: a
Komma, D. J., Horne, A. S. and Endow, S. A. (1991). Separation of meiotic study of chromosome elimination and of cell-lineage. Z. Wiss. Zool. 135,
and mitotic effects of claret nondisjunctional on chromosome segregation in 323-356.
Drosophila. EMBO J. 10, 419-424. Theurkauf, W. E., Baum, H., Bo, J. and Wensink, P. C. (1986). Tissue-
Komma, D. J. and Endow, S. A. (1995). Haploidy and androgenesis in speciﬁc and constitutive α-tubulin genes of Drosophila melanogaster code
Drosophila. Proc. Nat. Acad. Sci. USA 92, 11884-11888. for structurally distinct proteins. Proc. Nat. Acad. Sci. USA 83, 8477-8481.
Lindsley, D. L. and Zimm, G. G. (1992). The Genome of Drosophila Verde, F., Berrez, J.-M., Antony, C. and Karsenti, E. (1991). Taxol-induced
melanogaster. Academic Press, San Diego. microtubule asters in mitotic extracts of Xenopus eggs: requirement for
Matthews, K. A. and Kaufman, T. C. (1987). Developmental consequences of phosphorylated factors and cytoplasmic dynein. J. Cell Biol. 112, 1177-1187.
mutations in the 84B α-tubulin gene of Drosophila melanogaster. Dev. Biol. Wald, H. (1936). Cytologic studies on the abnormal development of the eggs of
119, 100-114. the claret mutant type of Drosophila simulans. Genetics 21, 264-281.
Matthews, K. A., Miller, D. F. B. and Kaufman, T. C. (1989). Developmental Walker, R. A. (1995). ncd and kinesin motor domains interact with both α- and
distribution of RNA and protein products of the Drosophila α-tubulin gene β-tubulin. Proc. Nat. Acad. Sci. USA 92, 5960-5964.
family. Dev. Biol. 132, 45-61. Walker, R. A., Salmon, E. D. and Endow, S. A. (1990). The Drosophila claret
Matthews, K. A., Rees, D. and Kaufman, T. C. (1993). A functionally segregation protein is a minus-end directed motor molecule. Nature 347,
specialized α-tubulin is required for oocyte meiosis and cleavage mitoses in 780-782.
Drosophila. Development 117, 977-991. Zhang, P., Knowles, B. A., Goldstein, L. S. B. and Hawley, R. S. (1990). A
Matthies, H. J. G., McDonald, H. B., Goldstein, L. S. B. and Theurkauf, W. kinesin-like protein required for distributive chromosome segregation in
E. (1996). Anastral meiotic spindle morphogenesis: role of the non-claret Drosophila. Cell 62, 1053-1062.
disjunctional kinesin-like protein. J. Cell Biol. 134, 455-464.
McDonald, H. B., Stewart, R. J. and Goldstein, L. S. B. (1990). The kinesin- (Received 30 September 1996 - Accepted 30 October 1996)