Science - USA (2022-02-11)

reexpressed four different NUMA mutants
in aMTOC-free mouse oocytes (fig. S6, A to
C). NUMA(DN) and NUMA(SpM-4A), which
cannot interact with dynein-dynactin ( 79 , 80 ),
bound to microtubule minus ends but failed
to rescue pole focusing (fig. S6, D to I). NUMA
(DMTBD1), which cannot bind microtubule
tips ( 81 , 82 ), was not enriched at microtubule
minus ends and failed to restore pole focusing
(fig. S6, J to L). By contrast, NUMA(DMTBD2),
which cannot bundle microtubules ( 83 ), bound
to microtubule minus ends and partially re-
stored pole focusing (fig. S6, J to L). Thus,
spindle pole focusing in the absence of aMTOCs
depends only partially on NUMA’s microtubule
cross-linking activity and predominantly on
its binding to microtubule minus ends and
to dynein-dynactin.

Indeed, NUMA was observed in proximity

to dynein-dynactin at the spindle poles in

aMTOC-free mouse oocytes (fig. S7, A to D).

To examine the roles of dynein-dynactin in

NUMA-mediated spindle pole focusing, we

performed dominant negative inhibitions.

We ectopically expressed the protein frag-

ments P150-CC1, which binds dynein ( 84 )

and perturbs its interactions with dynactin’s

ARP1 filament (fig. S7, E to I), and NUMA-N,

which does not oligomerize with endogenous

NUMA (fig. S6A) but perturbs the interactions

with dynein-dynactin ( 79 ).TheaMTOC-freemouse

oocytes expressing P150-CC1 and NUMA-N

displayed similar phenotypes: NUMA remained

bound to microtubule minus ends, but spin-

dle poles became defocused (fig. S8, A to H;

and movies S9 and S10). Moreover, when we

acutely injected recombinant P150-CC1 and

NUMA-N into aMTOC-free mouse metaphase I

oocytes, in which NUMA had already stably

associated with microtubule minus ends (fig.

S5, D and E), spindle poles became defocused

(fig. S8, I to P). Thus, NUMA binds micro-

tubule minus ends independently of dynein-

dynactin but is insufficient to organize and

maintain the spindle poles without dynein-

dynactin.

The assembly of fully activated dynein com-

plexes requires an additional cofactor LIS1 ( 85 ),

which is also expressed in mammalian oocytes

( 31 , 86 ). Similar to dynein and dynactin, LIS1

localized to the spindle poles in aMTOC-free

mouse oocytes (fig. S9, A and B). To test

whether LIS1 is also required for pole focus-

ing, we used Trim-Away (fig. S9, C to E). Both

Soet al.,Science 375 , eabj3944 (2022) 11 February 2022 4 of 19

A

NUMA Microtubules

aMTOC-free mouse MI spindle

Vertical view

mTRIM21 + anti-PCNT

mClover3-NUMA + Gold-conjugated GFP nanobody

B

F

Control IgG

( 26 )

anti-NUMA

( 22 )

0

0.12

0.24

0.36

0.48

0.6

Microtubul

e pa

ckin

g ind

ex ****

E

Control IgG

( 26 )

anti-NUMA

( 22 )

0

20

40

60

80

100

Pe

rcen

ta

ge

of

spin

dle

s

Focused poles

Mildly defocused poles

Severely defocused poles

048121620

-0.5

0.0

0.5

1.0

1.5

2.0

Enrichment of NUMA

Spindle length (μm)

relative to the cytoplasm

Spindle

pole 1

Spindle

pole 2

Gold-

labeled

NUMA

C

+mTRIM21 mRNA Fix

+anti-PCNT

+anti-NUMA

3 h 6 h

dbcAMP washoutNEBD

~1 h

mTRIM21 + anti-PCNT

anti-NUMA

Control IgG

NUMA

Chromosomes

Microtubules Merge

Metaphase I

D

Fig. 2. aMTOC-free mouse oocytes assemble spindles that mimic the
morphology and organization of spindle poles in human oocytes.
(A) Immunofluorescence images of MI spindles from aMTOC-free mouse
oocytes fixed at 7 hours after release. Green, NUMA; magenta, microtubules
(a-tubulin). The graph is the fluorescence profile of NUMA across the spindle
along the direction of the yellow arrow. (B) A single focused ion beam–scanning
electron microscopy (FIB-SEM) section of a MI spindle from a mClover3-
NUMA–expressing aMTOC-free mouse oocyte injected with gold-conjugated
green fluorescent protein (GFP) nanobody shortly before fixation at 7 hours

after release. Insets are magnifications of regions marked by the dashed-line

boxes. (C) Schematic diagram of the experiment in (D). (D) Immunofluorescence

images of MI spindles from control and NUMA-depleted aMTOC-free mouse

oocytes fixed at 7 hours after release. Green, NUMA; magenta, microtubules

(a-tubulin); blue, chromosomes (Hoechst). (EandF) Manual scoring (Fisher’s

exact test,P< 0.0001) and automated quantification of spindle pole defocusing in

control and NUMA-depleted aMTOC-free mouse MI oocytes. ****P< 0.0001.

See Materials and methods for box plot specifications. The number of oocytes

analyzed is specified in italics. Scale bars are 5mm.

RESEARCH | RESEARCH ARTICLE