Microtubule Dynamics in Cell Division: Exploring Living Cells with Polarized Light Microscopy

ANRV356-CB24-01 ARI 27 September 2008 2:16

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Microtubule Dynamics inCell Division: ExploringLiving Cells with PolarizedLight MicroscopyShinya InoueMarine Biological Laboratory, Woods Hole, Massachusetts 02543;email: [email protected]

Annu. Rev. Cell Dev. Biol. 2008. 24:1–28

First published online as a Review in Advance onJuly 16, 2008

The Annual Review of Cell and DevelopmentalBiology is online at cellbio.annualreviews.org

This article’s doi:10.1146/annurev.cellbio.24.110707.175323

Copyright c© 2008 by Annual Reviews.All rights reserved

1081-0706/08/1110-0001$20.00

Key Words

mitosis, spindle fibers, microtubules, dynamic equilibrium, colchicine,chromosome movement, polarized light microscopy, birefringence

AbstractThis Perspective is an account of my early experience while I studiedthe dynamic organization and behavior of the mitotic spindle and itssubmicroscopic filaments using polarized light microscopy. The bire-fringence of spindle filaments in normally dividing plant and animalcells, and those treated by various agents, revealed (a) the reality ofspindle fibers and fibrils in healthy living cells; (b) the labile, dynamicnature of the molecular filaments making up the spindle fibers; (c) themode of fibrogenesis and action of orienting centers; and (d ) force-generating properties based on the disassembly and assembly of thefibrils. These studies, which were carried out directly on living cellsusing improved polarizing microscopes, in fact predicted the reversibleassembly properties of microtubules.

1

Click here for quick links to

Annual Reviews content online,

including:

• Other articles in this volume

• Top cited articles

• Top downloaded articles

• Our comprehensive search

FurtherANNUALREVIEWS

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . 2ENCOUNTER WITH KATIE AND

JEAN DAN: INTRODUCTIONTO LIVING CELLS . . . . . . . . . . . . . . 3

THE “SHINYA-SCOPE”. . . . . . . . . . . . . 5TO PRINCETON (1948–51) . . . . . . . . . 7THE PRINCETON

MICROSCOPE . . . . . . . . . . . . . . . . . . . 9FINALLY TO WOODS HOLE:

REALITY AND BEHAVIOROF SPINDLE FIBERSAND FIBRILS . . . . . . . . . . . . . . . . . . . . 10

CONVINCING THE SKEPTICS:SPINDLE FIBERS INTIME-LAPSE MOVIES OFDIVIDING CELLS . . . . . . . . . . . . . . . 11

THE LABILE NATURE OFSPINDLE FIBERS: REVERSIBLEDEPOLYMERIZATION ANDASSOCIATED CHROMOSOMEMOVEMENTS . . . . . . . . . . . . . . . . . . . 12

ORIENTING CENTERS ANDULTRAVIOLET MICROBEAMEXPERIMENTS . . . . . . . . . . . . . . . . . . 14

SPINDLE BIREFRINGENCE,MICROTUBE DYNAMICS, ANDFORCE GENERATION FORCHROMOSOME MOVEMENT. . 17

EARLY BIOCHEMISTRY. . . . . . . . . . . . 18EARLY ELECTRON

MICROSCOPY . . . . . . . . . . . . . . . . . . . 20CONCLUDING REMARKS

AND UPDATE . . . . . . . . . . . . . . . . . . . 21

INTRODUCTION

Classically the light microscope was used toexamine preserved, thin-sectioned tissues andcells. More recently it has become an impor-tant tool for exploring the molecular basis ofphysiological functions directly in active livingcells. The major transition to its modern usewas prompted by optical advances in the mid-twentieth century, followed by another spurt in

the 1980s brought about by electronic imagingand striking advances in molecular biology.

Although I was able to participate in bothrecent transitions in microscopy, this Perspec-tive covers my experience during the earlier partof those events. By following the birefringencein live dividing cells with an improved polariz-ing microscope, we learned about the reality ofspindle fibers, the dynamic organization of theirfilaments, and their labile assembly/disassemblyand force-generating properties. The dynamicbehavior of molecules making up the birefrin-gent spindle filaments could now be followeddirectly in actively dividing cells.

Our studies on live cells were followed bythe isolation of a colchicine-binding protein,identified as the microtubule protein, and bythe discovery that microtubules could be dis-assembled or assembled in vitro. Those studiesverified our analysis of the birefringence ob-served in living cells and opened up vast newavenues for exploring the molecules and mech-anisms involved in mitosis and a wide range ofrelated cellular events.

As may be apparent from my publications,including this essay, I have been interestedin improving the capabilities of the lightmicroscope and exploring its uses as much as inuncovering the submicroscopic structures anddynamic events taking place in the living cell.Recently many have contributed immenselyto and made unbelievable advances in bothof these fields (see, e.g., Howard & Hyman2007, Maiato et al. 2004, Pawley 2006, Sluder& Wolf 2003, Wittmann et al. 2001). Thecurrent essay, thus, focuses on events relatingto some early developments to which I had thegood fortune to contribute.1

1This Perspective emphasizes our early studies on the mitoticspindle and microtubules but does not cover other topics thatmy colleagues and I have explored using polarized light andother advanced modes of microscopy. Articles reporting ona number of such biological studies and our contributions toadvances in microscopy are assembled in Collected Works ofShinya Inoue: Living Cells, Light Microscopy, and Molecular Dy-namics, just published by World Scientific Publishing (Inoue2008a). This book also includes a DVD disk featuring manycine- and video-micrographs of active living cells and nar-rated explanations on polarized light microscopy.

2 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


ENCOUNTER WITH KATIE ANDJEAN DAN: INTRODUCTION TOLIVING CELLS

Born 1921 in London, England, as the eldestchild of a Japanese diplomat, I was brought uptogether with my sisters in several countries.I enjoyed my early school years in Portland,Oregon, and Sydney, Australia, but from 1932I remained in Japan to enter a municipal highschool in Tokyo.

From my early days, I was interested in fig-uring out how to build electric motors and tinyportable radios that actually worked, but wasnot so interested in biology. My mother gave mea small microscope, but it was so disappointing;nothing in the prepared slides was doing any-thing. Still, at our home, I did raise silkwormsand later even collected and mounted butterflieson occasion. But in high school, the only thingthat really impressed me in biology was the be-havior of a bird’s feather that our teacher let usexamine under a loupe. The tiny, barbed hooksallowed the feather to be ruffled yet be zipperedneatly back together. The image seen through amagnifier finally explained how something ac-tually worked!

My deeper interest in biology was arousedin 1941 while I was a student at Musashi KotoGakko, a junior college in Tokyo. There I metProfessor Katsuma Dan (frivolously nicknamedKatie by himself ) in the first class that hetaught in his home country (Figure 1). Katiehad returned to Japan in 1937 with his Amer-ican wife, Jean Clark Dan, a fellow graduatestudent who had also worked with L.V. Heil-brunn at the University of Pennsylvania andwith whom he had spent summers at the MarineBiological Laboratory (MBL) in Woods Hole,Massachusetts.

As a student who had been unhappy with thehigh school classes in militaristic Japan, I wasshocked but delighted by Katie’s different atti-tude and approach. Instead of promoting rotelearning, he told us about how he and his friendswere figuring out how cells divided by tracingthe movement of kaolin particles placed on thesurface of developing sea urchin eggs. And he

Figure 1Katsuma Dan (1905–1996) at the CentennialCelebration of Misaki Marine Biological Station in1987. From Inoue (1994).

told us about how Karl von Frisch took advan-tage of the sugar rationing in Germany duringWorld War I to explore how honeybees foundtheir way back home by using polarization ofthe sky light to navigate, then dance and sig-nal to their hive mates how to reach the nectarsource.

In the lab, Katie let his students try experi-ments that might or might not work, rather thanhave us follow pretested procedures. I still can-not forget the excitement of having been able toshow how Lillie’s iron wire model of nerve con-duction worked by successive electrical depolar-ization (of a passivated layer on a steel wire im-mersed in concentrated nitric acid) rather thanby propagation of a chemical change, as arguedby my classmates. And I found that even theconduction speed could be enhanced by mak-ing the current jump past a locally insulated seg-ment of the model [ just as Ichiji Tasaki demon-strated the same year for saltatory conductionin myelinated nerve fibers (Tasaki & Takeuchi1941)]!

But that was the year the Japanese Navy at-tacked Pearl Harbor, and Japan and the UnitedStates became embroiled in World War II.Still, unlike many of my former high schoolclassmates, especially the A students who had

www.annualreviews.org • Microtubule Dynamics in Mitosis 3

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


become navy officers and soon perished at sea,I was deferred from military service as a sciencemajor (until four months before the end of thewar, when everybody was conscripted). Thus, Iwas able to enter Tokyo Imperial University in1942 and finish the curtailed 2.5-year curricu-lum with a major in zoology.

One evening in 1943, Katie invited me tohis home in Kudan, Tokyo, to try visualizingthe spindle during cell division. Imaging themitotic spindle in living cells was of particu-lar interest to Katie; he posited that egg cellsdivided by an elongating spindle pushing apartthe two astrospheres attached to its poles (Dan1943). The problem was that the spindle itself

was generally not visible under the microscopein living cells.

A notable exception, as Katie noted, was W.J.Schmidt’s 1937 observation of developing seaurchin eggs, made with a polarizing microscope(Figure 2). As reinterpreted by Schmidt him-self in 1939, those pictures showed the football-shaped spindles whose contrast depended onthe birefringence produced by aligned proteinmolecules (Schmidt 1939; see Figure 5 and as-sociated caption, below, for an explanation ofbirefringence).

That evening in Tokyo, Katie fertilized cleareggs of sea urchins that he had brought homefrom the marine lab in Misaki. Under air-raid

Figure 2(Left) W.J. Schmidt’s 1937 monograph. (Right) In this monograph, on p. 89, figure 31, Schmidt shows live,fertilized (flattened) sea urchin eggs observed between crossed Nicols with a polarizing microscope.(a) Four-cell stage in metaphase. (b) Four-cell stage in anaphase. (c) Anaphase sperm nuclei in polyspermicegg. (d ) Spindle and chromosomes in a blastomere at the two-cell stage. (e) Spindles and chromosomes in ablastomere at the eight-cell stage. At the time of publication of his monograph, Schmidt interpreted thefootball-shaped white and dark birefringent structures as being chromosomes and only the asters (seen inpanels d and e) as the nuclear spindle. Reproduced from Schmidt (1937) with kind permission of GebruderBorntrager (http://www.borntraeger-cramer.de).

4 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


blackout curtains, we spent several hours tryingto see the spindle birefringence, using a polar-izing microscope that Katie had borrowed fromhis colleague in Geology. But alas, the eveningended with inconclusive results.

THE “SHINYA-SCOPE”

Five years later we resumed these studies, thistime at the Misaki Marine Biological Station,which Katie had recovered in 1945 from the al-lied occupation forces, using his message enti-tled “The last one to go” (reproduced in Article56 in Inoue 2008a). At Misaki, rather than usea commercial polarizing microscope, I startedfrom scratch by assembling parts on a cast-offmachine gun base. (The Station had been takenover by the Japanese Navy for the last year ofthe war as a miniature submarine base, so somedestroyed weapon parts were scattered.) On thecast-iron base, I tied by string a Zeiss micro-scope that Katie let me modify, a calcite polar-izing prism loaned by Professor Koana of thePhysics Department of Tokyo University, andan AH-4 mercury arc lamp that I found at asurplus store and that I placed in a tea can.

Using this home-made instrument(Figure 3), we could finally repeat Schmidt’sobservations. For several minutes before theegg underwent cleavage, we saw the birefrin-gent spindle and asters emerge and grow, thenthe spindle splitting into two parts as the astersgrew larger. But the initial success was dashedwhen I tried to improve the image by rotatingthe objective lens to minimize the stray lightintroduced by strain birefringence in thelenses. The field between crossed polarizersdid, in fact, become darker, but where werethe birefringent spindles? They had simplyvanished!

Katie’s admonition to me was, “I told you toleave well enough alone.” But I was really curi-ous and wanted to make the system work better.It did take a whole month, but I finally realizedthat the birefringence of the strained lens was,in fact, helping by acting as a compensator andraising the image contrast of the weakly bire-fringent spindle. So I split a piece of mica into

Scope with analyzer attachedto bottom of draw tube

Mica compensator

Calcite prism polarizer

Tea can housingAH-4 Hg lamp

Parts on wood blockssecured by twine on discarded

machine gun base

Floor

Shinya-Scope 1(Misaki, March 1948)

Figure 3The author’s handmade polarizing microscope, built from salvagedcomponents at the Misaki Marine Biological Station in 1948. Reproduced fromInoue (2008a) with kind permission of World Scientific Publishing.

a thin sheet and placed it on the microscope’srotatable substage filter holder so that its ori-entation could be adjusted, namely, so that itwould act as a Brace-Koehler compensator.

Now, even though the microscope field wasnot completely dark, we could see the brighteror darker football-shaped spindle against a graybackground (Figure 4). In fact, we could seequite a bit more than was reported in Schmidt’spublication and even guess at the orientation ofthe component molecules.

By way of explanation, a compensator in-troduces uniform birefringence over the wholefield of view so that, between crossed polarizers,the specimen appears brighter or darker, de-pending on whether its birefringence is addingor subtracting from the birefringence of thecompensator. As shown in Figures 4 and 5,


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Figure 4(a,b) Two-cell-stage jellyfish (Spirocodon sp.) egg with metaphase spindles observed with the microscopeshown in Figure 3. In these live, optically clear cells, the positively birefringent spindle appears bright ordark, depending on its orientation relative to the compensator (see Figure 5 for an explanation). The cellsurface, which is negatively birefringent, appears in opposite contrast where it lies parallel to the spindle.Unlike in many other genera, eggs of jellyfish do not produce a fertilization envelope. (c,d,e) Fertilized,developing eggs of a sand dollar, Clypeaster japonica. The Clypeaster eggs (which are exceptionally transparent)are surrounded by a fertilization envelope, which shows a strong, tangentially positive birefringence. Thespindle and asters also show a positive birefringence along their long axes. Double-headed arrows indicatethe slow-axis direction of the compensator. From Inoue & Dan (1951).

where the slow axis of the specimen (e.g., thelength of the spindle filaments) lies parallel tothe slow axis of the compensator, the specimenappears brighter. Where the axes are crossed(lie in opposite quadrants), they appear darker,i.e., are compensated. [Every birefringent (i.e.,doubly refractive) material has two refractiveindexes that reflect the arrangement of theirmolecular lattice or fine structure. The direc-tion for which (the electrical vector of ) the lightwave suffers the greatest refraction is called theslow axis, and the one with the lowest refractionis called the fast axis. For further explanations,

see, e.g., Bennett 1950, Appendix III in Inoue1986, and Inoue 2002.]

In the paper reporting these observations(Inoue & Dan 1951), I also calculated the op-timum amount of compensation required tomaximize the image contrast of weakly birefrin-gent objects in the presence of stray backgroundlight (as also published nearly concurrently bySwann & Mitchison in 1950).

Kayo Okazaki and Katie extensively used themicroscope I built in Misaki (and sketched frommemory in approximately 1989; see Figure 3)to follow the development of biocrystalline

6 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


skeletal spicules in sea urchin embryos (Okazaki& Inoue 1976). After my departure to Princetonin 1948, Kayo and Katie called it the “Shinya-Scope.”

In 1948, Jean Dan returned to Misaki fromher first post–World War II trip back home tothe United States. She was full of news abouttheir friends in the States, especially those atthe MBL in Woods Hole. And she broughthome, as a present for her husband Katie, aBausch & Lomb phase contrast microscope (thefirst one available in the United States and ac-quired courtesy of the American PhilosophicalSociety). Jean, who soon discovered the acroso-mal reaction, used this microscope extensivelyto study sperm-egg interactions at fertilization.For me, she arranged a financial loan from hersister Peggy Chittick of Milford, Connecticut,so that I could travel and study in the States.

TO PRINCETON (1948–51)

In the fall of 1948, with a postwar Japanese pass-port (which I recall was number 50), I arrived atPrinceton’s Biology Department. There, whilebuilding what I hoped was a better polarizingmicroscope, I was introduced to classical cytol-ogy by my mentor, Kenneth W. Cooper. Kenhad studied with Franz Schrader, who in turnhad followed E.B. Wilson’s steps; all three wereat Columbia University in New York. It was,therefore, natural for me to wonder how chro-mosomes moved in mitosis and about the enig-matic properties of the mitotic spindle.

At Princeton I was exposed to Wilson’s clas-sical volume on the hereditary role of chro-mosomes and the structure and function ofthe mitotic spindle in cell division (Wilson1928). Summarizing four-decades-long studieson fixed and stained cells made by many cytol-ogists, he describes the fibrillar structure of theachromatic spindle and astral rays. Still, he ispuzzled about the ephemeral nature and invis-ibility of the spindle fibrils in living cells andquestions the validity of the contractile fibril-lar hypothesis for chromosome movement fa-vored by many. At the same time, he is reluc-tant to accept “that the fibrillae seen in sections

A

F'

P'

S'

A'F

P

S

20 μm

Figure 5Micromere formation during the fourth division in a developing egg of a sanddollar, Echinarachnius parma (the images were taken with a rectified polarizingmicroscope in the 1970s). (a) The spindles in these four cells have converged tothe egg’s vegetal pole (the four animal pole cells are out of focus). Where thespindle long axis (orientation of microtubules) lies parallel to the compensatorslow axis (SS′), the positively birefringent spindle appears bright. Where theaxes are crossed, the spindle appears dark. (b) Cleavage planes bisect the spindleremnants and give rise to four micromeres (predecessor of spicules and gonads)and four macromeres. In the diagram below the photographs, PP′ and AA′show the transmission axes of the polarizer and the analyzer (which arecrossed), respectively, and SS′ and FF′ are the orientations of the slow and fastaxes of the compensator, respectively. Reproduced from Inoue (1981) with kindpermission of The Rockefeller University Press.

may not really pre-exist approximately as suchin the living cell” and cautions us not to “pre-maturely condemn a theory which may yet bereconcilable with the so-called dynamical the-ories” (Chapter II, Section IV, The Mechanismof Mitosis, in Wilson 1928).

The more physicochemically oriented pro-ponents of the dynamical theory experimentedwith living cells and tended to be skeptical of theexistence of the fibrous elements of the spindleand asters. These investigators considered suchelements to be artifacts of fixation.

In a 1929 article, Karl Belar comparedthe behavior of live grasshopper spermatocyteswith that of carefully fixed and stained cells(Figure 6). Although unable to see any spindle


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Figure 6Spermatocytes of a grasshopper, Chorthippus lineatus, at different stages ofmeiosis I. The drawings were by Belar (1929) of 6-μm sections of cells fixed inFlemming-Meves solution, stained with iron-hematoxilin. mi, mitochondria;x, sex chromosome; H, heteromorph tetrad.

structure, he observed in healthy, live cellsBrownian motion preferentially along the di-rection of fibrils that would appear after fix-ation. Also, from the distortion of live cellstreated with hyperosmotic media, he concludedthat spindle fibrils or some longitudinal lamel-lar material must exist in the living cell despitetheir invisibility.

By observing chromosome movement in di-viding stamen hair cells of Tradescantia, BungoWada proposed that spindle fibers were made

up not of coherent filaments but of short thinrods, as in a liquid crystal (Wada 1950). Simi-larly, observing the migration of chromosomesthat appeared to cut right through kinetochorefibers, Gunnar Ostergren (1949) also favoredthe liquid crystalline nature of spindle fibers.

These and other views in the early 1950son the physical nature of the mitotic spin-dle, as well as various proposals on how chro-mosomes move in mitosis, are summarized inFranz Schrader’s monograph Mitosis: The Move-ment of Chromosomes in Cell Division (Schrader1953). In this volume, he points out two cases inwhich spindle fibers were actually observed inintact dividing cells. These observations weremade by L. R. Cleveland (1938) in Barbu-lanympha, a symbiotic protozoan in the wood-eating cockroach Cryptocercus, and by KennethCooper (1941) in the eggs of a grass mite,Pediculopsis graminum. Still, Schrader points outthat these were exceptional cases and could notbe taken to represent cells undergoing mitosisgenerally. Thus, the reality of spindle fibers andtheir nature remained major unresolved issues.

While I was at Princeton, we also sawfascinating movies of dividing grasshopperspermatocytes, filmed by Kurt Michelle ofKarl Zeiss, using its phase contrast micro-scope. Subsequently, Kyojiro Shimakura cap-tured higher-resolution images of similar livecells (Figure 7), and Andrew and Wishia Bajer(1951, 1956) made many films of dividingendosperm cells of the African blood lilyHaemanthus katherinae (Figure 8; see Sup-plemental Movie 1; for all supplementalmaterial, follow the Supplemental Materiallink from the Annual Reviews home page athttp://www.annualreviews.org). In these di-viding cells, the phase contrast microscope dis-played the movement and shape change of chro-mosomes most strikingly.

The phase contrast microscope accentuatesthe image contrast of those bodies whose refrac-tive indexes are somewhat greater or less thanthe refractive indexes of their surroundings.In contrast to the chromosomes themselves,the spindle fibers that were supposed to movethe chromosomes, and the fibrils laying down

8 Inoue

Supplemental Material

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.

http://arjournals.annualreviews.org/doi/suppl/10.1146/annurev.cellbio.24.110707.175323


the cell plate in plant cells, were not visible inphase contrast.

In these early post–World War II years, theelectron microscope also started to reveal manyimportant cellular fine structures (see Sabatini2005). But little could be seen of the fine struc-ture in the spindle until glutaraldehyde fixa-tion was introduced nearly two decades later(Sabatini et al. 1963).

Thus, the challenge for me in the late 1940swas to develop a polarizing microscope thathad enough sensitivity and image resolution toshow what, in fact, was going on inside dividing,living cells.

THE PRINCETON MICROSCOPE

At Princeton, I decided to start from scratchagain so that I could improve on the microscopethat I had built at Misaki. By then I was moreaware of the standard use of polarizing (or pet-rographic) microscopes to study crystals and toidentify minerals and ores (e.g., Hartshorne &Stuart 1960, Rinne & Bereck 1953, Wahlstrom1960, Wright 1911). Biologists also used thesemicroscopes to study mineralized tissue, skele-tal muscle, plant cellulose walls, etc., whichwere all highly birefringent (see, e.g., Ambronn& Frey 1926, Bennett 1950, Frey-Wyssling1953, Schmidt 1924). Schmidt (1937) also ex-plored an extensive array of cellular com-ponents and cell products, many with muchweaker birefringence, as summarized in thissecond monograph. The commercially avail-able polarizing microscopes were, however, notoptimally designed for observing or measuringthe intricately organized, and very weakly bire-fringent, minute organelles in living cells.

It turned out that there was an inherentincompatibility between achieving high sensi-tivity for detecting weak birefringence and forgaining image resolution high enough to studystructural details inside a living cell. At low con-denser and objective lens numerical apertures(NAs), one could achieve high extinction andgain great sensitivity, but then the resolutionwas limited. Selecting objective and condenserlenses with exceptionally low strain birefrin-

Figure 7Phase contrast image of a live spermatocyte in a grasshopper, Chloealtisgenicularibus, at full metaphase. Photograph courtesy of Dr. Kyojiro Shimakuraof Hokkaido University. Reproduced from Inoue (1964) with kind permissionof Elsevier.

gence, and polarizers and analyzers providingvery high extinction, was not enough. The crit-ical factor turned out to be the very fact thatmicroscope lenses had to refract light to forman image. And the greater the angle of refrac-tion (and high-NA lenses are characterized byhigh angles of refraction), the greater is the lossof extinction and, therefore, loss of sensitivityto detect weak birefringence.

I examined this paradox in detail at Prince-ton (Inoue 1952b) but had no basic solution, soI went ahead and built my second polarizing mi-croscope, using the best arrangement and com-ponents available to me at that time (includ-ing strain-free objectives selected from severalhundred by Bausch & Lomb, a much brighterAH-6 water-cooled high-pressure mercury arclamp, a Leitz photo stand, etc.). The resultingmicroscope is illustrated in Figure 9.

With this microscope, I found that I couldindeed gain moderately high resolution (if not


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Figure 8Phase contrast images of an endosperm cell of Haemanthus katherinae from(a) nuclear envelope breakdown, (b) metaphase, and (c) late anaphase, through(d ) cell plate formation. See Supplemental Movie 1, reproduced from Inoue(2008a) with kind permission of World Scientific Publishing. Reproduced fromInoue & Oldenbourg (1998) with kind permission of American Society for CellBiology (http://www.molbiolcell.org/cgi/content/full/9/7/1603).

yet at the oil-immersion level) of weakly bire-fringent structures inside living cells. Beingaware that birefringence reflects the arrange-ment of fine structure and molecules far smallerthan the resolution limit of the light micro-scope, and that the observations could be madewithout staining or otherwise interfering withthe activity of the living cells, I was excited to seewhat I could explore with this new instrument.

FINALLY TO WOODS HOLE:REALITY AND BEHAVIOR OFSPINDLE FIBERS AND FIBRILS

In early summer of 1949, I finally arrived at theMBL in Woods Hole, together with my class-mates Woody Hastings and Dave Stadler andwith my new microscope in the trunk of Dave’sfamily car.

In Woods Hole, I met many of the Dans’and Cooper’s old friends about whom I hadheard so much. I became acquainted with themat the mess hall (where we all shared tables),

in the lecture room in the shingle-covered OldMain, in the labs, at Captain Kidd, and at StonyBeach.

These new acquaintances—Don Costello,Albert Tyler, Dan Mazia, and the Osterhouts––introduced me to several local marine inver-tebrates and showed me how to collect theirfreshly spawned gametes. The eggs from a fewspecies were clear enough to see the birefrin-gent spindle and asters directly, but many ofthe eggs were filled with yolk and other bire-fringent granules and were too opaque to seetheir internal structures. I solved this problemby using an air-turbine centrifuge, developedearlier by E. Newton Harvey and Bill Loomis atPrinceton. Eggs layered on a cushion of isopyc-notic sucrose-seawater solution could be strat-ified so that the spindle and asters would dis-play their birefringence within a clear, yolk-free zone. Despite the stratification and evenegg fragmentation, the spindle-containing eggfragments would continue to divide when fer-tilized.

By using centrifuged oocytes from the an-nelid parchment worm Chaetopterus pergamen-taceous, I was able to see clearly the structure oftheir metaphase-arrested, first-meiosis spindle.This material was ideal for viewing details ofspindle structure and for experimenting on thespindle. The cell stayed in metaphase withoutproceeding to anaphase for more than an hourunless the cell was activated, for example, byfertilization or osmotic shock.

To my delight, the image resolution of thenew microscope was high enough so that I couldnow see that the Chaetopterus oocyte spindle wasnot just a birefringent football-shaped structure(as seen by Schmidt 1937, Swann & Mitchison1950, and Katie and myself in Misaki). Instead,it was made up of birefringent fibers whosebirefringence was stronger (a) where they con-verged and attached to the kinetochore on eachof the nine chromosomes on the metaphaseplate and (b) at the two spindle poles. Further-more, each of the chromosomal, or kinetochorefibers, as well as the material of the astral rays,appeared to be made up of very thin, submicro-scopic fibrils (Figure 10; Inoue 1953).

10 Inoue


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



In the activated oocytes, the birefringenceof the chromosomal fibers briefly rose as thecell entered anaphase, as it would also in ametaphase cell whose spindle was stretched(Figure 11). As the chromosomes were led bythe chromosomal fibers to the spindle poles,the fiber birefringence dropped, except whereit remained high adjacent to the kinetochore onchromosomes.

Also, as chromosomes moved poleward dur-ing anaphase, the diameter of each fiber didnot increase as the fiber shortened. So I ar-gued that, despite Schmidt’s claim, the loss ofbirefringence of the spindle material duringanaphase could not be explained by a foldingof its polypeptide chains (Inoue 2008c).

From these observations I concluded that,although invisible in living cells with conven-tional microscopy, spindle fibers did really existin living cells and were not artifacts of fixation,in contrast to what had been argued for halfa century (Schrader 1953). Furthermore, thefibers were made up of a bundle of submicro-scopic fibrils as depicted in the better-preservedfixed specimen recorded by early cytologists.But the skeptics who had not seen the dynamicimages through my microscope had yet to beconvinced.

CONVINCING THE SKEPTICS:SPINDLE FIBERS IN TIME-LAPSEMOVIES OF DIVIDING CELLS

During the school year at Princeton, I contin-ued to improve my microscope so that I couldmeasure birefringence in minute objects (Inoue1951) and also make time-lapse movies of theweakly birefringent spindle in dividing plantcells.

Among the time-lapse movies showing thechanging birefringence of the spindle coupledwith the movement of chromosomes, the mostinformative came from cells in the anthers of anEaster lily, Lilium longiflorum. When I visitedDr. Ralph Ericson at the University of Penn-sylvania, he told me that pollen mother cellsin 22.4-mm-long flower buds undergo meiosisand gave me a few plants that were at just the

Pinhole (virtual source)

Field lens

Polarizer

Field Diaphragm

Comparator

Compensator

Condenser diaphragm

Condenser

Specimen

Objective lens

Analyzer with

stigmatizing lenses

Removable mirror

Projection ocular

Lightsource

Figure 9Shinya’s Princeton microscope (left) with a schematic of the optical path (right).From Inoue (2008b), reproduced from Inoue (2008a) with kind permission ofWorld Scientific Publishing.

right stage (for how the flower bud is measured,see Inoue & Oldenbourg 1998).

Before collecting cells from the lily anthers,I found that I had to centrifuge the flowerbud in a clinical centrifuge to displace thehighly birefringent cell inclusions, which oth-erwise prevented observing the live spindle.Also, culture media for lily pollen mother cellswere not known, so I diluted frog Ringer to7/8 (a concentration at which the cells wouldnot plasmolize). Fortunately, the pollen mothercells, starting with nuclear envelope break-down, would complete their two successivedivisions despite the intense monochromaticgreen illumination required for many hours torecord the full sequence on 16-mm film.

At MBL, I first publicly showed the movie ofthe dividing Lilium pollen mother cells in Lil-lie Auditorium (I believe it was in 1951). I stillremember how Homer Smith, MBL’s GeneralManager, light-proofed the room by person-ally climbing up and covering the auditorium’sgreen house roof with a black sheet of cloth so


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


10 μm

Figure 10(a,b) Birefringence of a metaphase-arrested meiosis I spindle in a live,centrifugally clarified Chaetopterus oocyte; scale bar for a and b, 10 μm. (c) Polarview of the metaphase plate seen in DIC (differential interference contrast).(d ) Distribution of fibrils deduced from birefringence. Panels a, b, and d arereproduced from Inoue (1953) with kind permission of Springer Science+Business Media. Panel c is reproduced from Inoue & Inoue (1986) with kindpermission of Wiley-Blackwell Publishing.

that I (a mere graduate student) could show thefilm!

The movie showed the chromosomes be-ing brought to the metaphase plate and thenled poleward by the birefringent spindle fibers.Finally, birefringent fibrils reemerged betweenthe daughter nuclei to form the phragmoplast,and vesicles assembled in its mid-zone to formthe cell plate (Figure 12; see SupplementalMovie 2, Inoue 1964, Inoue & Oldenbourg1998).

The response to this showing was most grat-ifying. There was no doubt that spindle fibersand the fibrils making up the fibers (made visi-ble with sensitive polarizing microscopy) wereclearly present in the living, dividing cell. Bun-

dles of the birefringent fibrils brought the chro-mosomes to the metaphase plate, and shorten-ing fibers led them to the spindle poles. Afterthe interzonal fibers diminished, new birefrin-gent fibrils appeared between the daughter nu-clei to generate the phragmoplast. In the mid-plane of the phragmoplast, the cell plate wasassembled.

The question asked by Dr. Ethyl BrownHarvey, “Were those cells alive?”, reflected thepervasive, long-held view by many that spindlefibers and their fibrils were not present in liv-ing cells but were artifacts of fixation. But thesecells were happy enough to go through theirtwo sequential divisions! The reality of spindlefibers in living cells could no longer be doubted.

THE LABILE NATURE OFSPINDLE FIBERS: REVERSIBLEDEPOLYMERIZATION ANDASSOCIATED CHROMOSOMEMOVEMENTS

But what was most exciting for me was thatthe birefringence of the spindle fibers was notstatic. It not only fluctuated and changed duringmitosis but disappeared reversibly when a cellwas exposed to low temperature (Figure 13; seeSupplemental Movie 3, Inoue 1952a, 1964) orto the antimitotic drug colchicine (Figure 14;Inoue 1952c). In other words, the fibrils mak-ing up the spindle fibers would depolymerizein cold or when exposed to colchicine, only torepolymerize when the condition was reversed.

The results with colchicine were especiallyintriguing. When metaphase-arrestedChaetopterus oocytes were exposed tocolchicine, the spindle birefringence wouldgradually disappear as the fibrils depolymerized(the kinetochore fibers were the longest topersist), but in addition, the depolymerizingfilaments actually led the chromosomes andinner spindle pole to the cell surface, wherethe outer meiotic spindle pole was attached(Figure 14). Thus, the colchicine experimentssuggested that depolymerizing filamentscould generate forces adequate to pull the

12 Inoue


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



00

1

2

3

4

5

6

7

5 10

Chaetopterus spindle

15 20 25

Reta

rdat

ion

(μm

)

Length (μm)

aaa b

10 μm

Figure 11(a) Birefringence of amoderately stretched,live Chaetopterusoocyte spindle. Scalebar, 10 μm.Reproduced fromInoue (1953) with kindpermission of SpringerScience+BusinessMedia. (b) Graph ofspindle fiberbirefringence versuslength. Spindles werestretched by gentlecompression of anoocyte fragmentgenerated bycentrifugation. FromInoue (1952c).

Figure 12Birefringence of spindle fibers in a pollen mother cell of an Easter lily, Lilium longiflorum. A movie of dividingLilium pollen mother cells finally convinced skeptics that spindle fibers (and their dynamic submicroscopicfibrils) were actually present in living cells and were not an artifact of fixation, in contrast to what had beenargued for half a century. (a) Anaphase onset. (b) Mid-anaphase. (c) Phragmoplast formation. (d ) Cell plateformation. Panels reproduced from Inoue (1964) with kind permission of Elsevier. See SupplementalMovie 2, reproduced from Inoue (2008a) with kind permission of World Scientific Publishing.



Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



Figure 13Reversible loss of spindle birefringence by cold treatment of a developing seaurchin egg. In these still frames taken from Supplemental Movie 3, the firstcleavage spindle (a) is developing, (b) has disappeared by cold treatment, (c) hasrecovered and reached full metaphase, and (d ) is at cleavage onset. Afterchilling and returning to room temperature, the second cleavage spindle is(e) in metaphase and ( f ) chilled again in anaphase. The same egg was chilledseven times; development was delayed by the duration of chilling but notarrested. Reproduced from Inoue (2008a) with kind permission of WorldScientific Publishing. See Supplemental Movie 3, also reproduced from Inoue(2008a) with kind permission of World Scientific Publishing.

chromosomes and the spindle poles towardeach other (Inoue 1952c).

Jumping forward to the 1970s, Ted Salmonand I showed that chromosome movement, as-sociated with gradual loss of spindle fiber bire-fringence, could also be induced in metaphase-arrested Chaetopterus oocytes when we droppedthe temperature to an intermediate value(Figure 15; Inoue & Ritter 1975). Further-more, using a novel pressure chamber, Salmonshowed that not only could birefringence loss

and chromosome movement be induced by el-evating hydrostatic pressure, but both becamefaster (up to 400 atm) as more pressure was ap-plied (Figures 15 and 16; Salmon 1975b, 1976;Salmon & Ellis 1975). Thus, it became increas-ingly likely that microtubules in vivo could gen-erate pulling and pushing forces by their disas-sembly and assembly. Salmon further showedthat purified microtubules assembled in vitrowould also depolymerize under high hydro-static pressure (Salmon 1975a).

In plant cells as well, the fibrils of the spindleand phragmoplast were just as labile as in animalcells. These fibrils would depolymerize whenexposed to cold or treated with colchicine oreven low concentrations of calcium ions, as wassoon to be discovered.

ORIENTING CENTERS ANDULTRAVIOLET MICROBEAMEXPERIMENTS

By the late 1950s, we had developed the po-larization rectifier and, thus, were able toachieve full resolution with a polarizing mi-croscope, even using high-NA oil-immersionlenses, without losing the sensitivity neededto detect weak birefringence (Inoue & Hyde1957).

Using this new capability, Andrew andWishia Bajer and I followed the behavior ofbirefringent fibers and fibrils during mitosis inlive endosperm cells of the African blood lilyH. katherinae. The endosperm cell lacks a rigidcell wall and so could be flattened on an osmot-ically equilibrated agar sheet by gently draw-ing off the excess endosperm fluid. Using se-rial photographs and a time-lapse movie takenwith rectified optics, we found positively bire-fringent fibrils that were aligned in the clearzone and polar cap outside of the intact nuclearenvelope before the envelope started to breakdown (Figure 17, left column; Inoue & Bajer1961).

As soon as chromosomes condensed fur-ther and the nuclear envelope started tobreak down, birefringent fibrils grew into thenucleus (Figure 17, right column) and attached

14 Inoue


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



Figure 14Shortening and loss of birefringence by a metaphase-arrested Chaetopterus spindle exposed to colchicine inseawater. Below each frame is the time in minutes and seconds after the application of 5 × 10−4 Mcolchicine (upper row) and 5 × 10−3 M colchicine (lower row). From Inoue (1952c).

to kinetochores to form birefringent chromo-somal fibers. Some fibrils bundled into sheathsaround chromosomes.

From metaphase through anaphase, fiberbirefringence converged and remained strongat the kinetochore, as we had already seen inlily pollen mother cells and in animal cells(Figure 18; Inoue & Bajer 1961). In telophase,more birefringent fibrils appeared parallel tothe spindle’s remaining interpolar fibers andformed the phragmoplast, as had also occurredin Easter lily pollen mother cells. Small vesiclesaccumulated in the mid-zone of the phragmo-plast, then fused to form the cell plate. Whenthe cell plate started to appear, the birefrin-gence of the phragmoplast fibrils was strongestby the cell plate, suggesting that it (along withthe kinetochores) had taken over the role ofkeeping the fibrils oriented. Thus, the youngcell plate and kinetochores both appeared to beacting as orienting centers (Inoue & Sato 1967).

Microbeam experiments further revealedthe activity of orienting centers. Using carefully

dose-controlled ultraviolet (UV) microbeam ir-radiation, we could reduce or abolish the bire-fringence of spindle fibers or phragmoplast fil-aments locally in the irradiated region. When ametaphase endosperm cell of Haemanthus (withits typical plant-type spindle) was irradiated, thespindle fiber lost its birefringence not only inthe irradiated area but also toward the spindlepole. Nevertheless, the birefringence persistedbetween the irradiated area and the kinetochore(Figure 19). In telophase, the birefringentphragmoplast fibrils disappeared from the irra-diated area and poleward but, again, persistedbetween the cell plate and the irradiated area.

But whether in metaphase or with a phrag-moplast, the birefringent fibrils immediatelystarted to grow back through the irradiated areaand poleward as soon as UV irradiation wasstopped. Thus, spindle fiber molecules clearlygrow away from the kinetochore and from thecell plate.

Microbeam experiments with animal cellswere further revealing. Using crane fly


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


spermatocytes, Forer (1965) found that bire-fringence was lost from the area irradiated witha UV microbeam. But in these cells, birefrin-gent fibers persisted not only between the ir-radiated area and the kinetochores but also be-tween the irradiated area and the spindle pole!In other words the irradiated area appeared asan area of reduced birefringence (Figure 20).

Furthermore, after irradiation, the area ofreduced birefringence traveled poleward at asteady pace and disappeared at the spindle pole.Thus, the birefringent material not only grewpoleward away from the kinetochore into theirradiated area but also shortened toward thespindle pole.

These observations implied a treadmillingof the spindle fiber molecules from the kine-

10 μm

10 μm

tochore to the pole. (Interestingly, the spindlefibers in grasshopper spermatocytes exposed toUV microbeam show a different, more com-plex behavior, as reported by Gerry Gordon in1979; see Article 32 in Inoue 2008a for addi-tional note.)

Thus, on the basis of the dynamic distribu-tion of spindle birefringence observed in severalanimal and plant cells and UV microbeamexperiments, we postulated that three mecha-nisms align the molecular filaments that makeup spindle fibers, phragmoplast fibrils, andastral rays. They are (a) the activity of orientingcenters, (b) spontaneous alignment by concen-trated formation of filaments, and (c) parallelalignment to previously formed filaments.Furthermore, the orienting centers [which arenow called microtubule-organizing centers(MTOCs)] become active, one after another,with the progression of mitosis (Inoue & Sato1967).

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−Figure 15Chromosome movement induced by cooling or byelevated hydrostatic pressure in a metaphase-arrested Chaetopterus oocyte. In Row 1, thetemperature is dropped from 23.5◦C to 5.2◦C attime 0.0 min. (The −2.5-min frame is in polarizedlight, and the other frames are in DIC.)Chromosomes move to the cell surface as spindlefilaments depolymerize and shorten. In Row 2, thetemperature is raised to 24◦C at 0.0 min. (The 5.8-and 16.5-min frames are in polarized light, and theothers are in DIC.) Chromosomes move away fromthe cell surface as spindle birefringence and lengthincrease. Scale bar, 10 μm. Rows 1 and 2 are fromInoue & Ritter (1975). Row 3 shows 200 atm ofhydrostatic pressure applied at time 0.0 min. (The−3.0 min frame in polarized light, and the otherframes are in phase contrast; print magnificationdiffers from that of Row 4.) Chromosomes move tothe cell surface as spindle filaments depolymerizeand shorten. In Row 4, pressure is reduced to 1 atmat time 0.0 min. (The 1.25-, 2.25-, and 26.0-minframes are in polarized light; the other frames are inphase contrast.) Chromosomes move away from thecell surface as spindle birefringence and lengthincrease. Scale bar, 10 μm. Rows 3 and 4 arereproduced from Salmon (1975b) with kindpermission of Wiley-Blackwell Publishing.

16 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


136

5

10

15

204 272

Pressure (atm)

Chro

mos

ome

velo

city

(μm

min

–1)

½ s

pind

le-s

hort

enin

g ra

te (μ

m m

in–1

)

340 408

a b

Figure 16(a) Water-jacketed hydrostatic pressure chamber on Leitz polarizing microscope. Reproduced from Salmon& Ellis (1975) with kind permission of The Rockefeller University Press. (b) Speeds of chromosomemovement and birefringence decay versus applied pressure. Both speeds increase as hydrostatic pressure israised. But above 400 atm, chromosomes no longer move because spindle microtubules disassemble toorapidly. Reproduced from Salmon (1976) with kind permission of Cold Spring Harbor Laboratory Press.

SPINDLE BIREFRINGENCE,MICROTUBE DYNAMICS, ANDFORCE GENERATION FORCHROMOSOME MOVEMENT

In a symposium volume published in 1964, Iprovided photographs of birefringent spindlefibers and fibrils in a large variety of animaland plant cells undergoing normal mitosis aswell as treated with cold or exposed to UV mi-crobeam irradiation. The fibers and their ori-ented fibrils, which I proposed were capable ofproducing pulling and pushing forces by the re-moval or addition of material, appeared to be ina labile dynamic equilibrium with their subunitmolecules (Inoue 1964).

In our summary paper in 1967, Hidemi Satoand I showed that 50% heavy water doubles thebirefringence and size of spindles and asters,signaling the reversible incorporation of a sub-unit protein from a pool. The reversible assem-bly took place in less than 2 min and did notrequire the synthesis of new proteins. Thus,coupled with our earlier observations on thereversible disassembly and force-generating ef-fects by cold and by colchicine, we postulated

the presence of a dynamic equilibrium betweenspindle fibers (filaments) and a cytoplasmic poolof their protein subunits and that shifts in theequilibrium were responsible for spindle as-sembly and also for chromosome movement(Figure 21; Inoue & Sato 1967; see also Inoue& Salmon 1995, Maiato et al. 2004).

In the meanwhile, Bruce Nicklas and co-workers carried out extensive micromanipula-tion studies on live grasshopper spermatocytes.By displacing a single chromosome with a mi-croneedle, they demonstrated how the kine-tochores are quite stably, but not irreversibly,linked by fibers to the spindle poles. Also, theyshowed how tension exerted on kinetochoresby spindle fibers governed the position and ar-rangement of chromosome arms and even thecoordinated onset of anaphase (Nicklas & Koch1969, Nicklas & Staehly 1967, Nicklas et al.2001). In an extensive review article discussingthe role of microtubules in mitotic chromosomemovements, Nicklas (1971) illustrates the de-tailed distribution of birefringence in spindlefibers of a living grasshopper spermatocyte un-dergoing anaphase as seen with a rectified po-larizing microscope (Figure 22).


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Figure 17Endosperm cell of the African blood lily Haemanthus katherinae preparing toenter mitosis. (Left column) A few hours before nuclear envelope breakdown,birefringent fibrils appear in the clear zone outside of the envelope. (Rightcolumn) As the nuclear envelope starts to break down, additional fibrils from thepolar cap grow into the nucleus. Shortly thereafter, the fibrils connect tochromosomes and form the spindle. The compensator slow axis is NW-SE inpanels a and c and NE-SW in panels b and d. Reproduced from Inoue & Bajer(1961) with kind permission of Springer Science+Business Media.

In 1974, using a rectified polarizing mi-croscope, Sato & Izutsu captured spectacularimages of dynamic, birefringent spindle fibersin dividing spermatocytes of another speciesof grasshopper, Chrysocraon japonicas [see Sup-plemental Movie 4, reproduced from Inoue(2008a) with kind permission of World Scien-tific Publishing]. Although the incredible activ-ity (Northern lights flickering of spindle fiberbirefringence that reflects the highly dynamicspindle microtubules) seen in the film couldnot be fully explained in 1974 (because so lit-tle was yet known of interactions between mo-tor proteins and microtubules), the activity wasinterpreted as reflecting the treadmilling andgrowth and shortening of microtubules thatwere stochastically assembling and disassem-

bling in dynamic equilibrium with their subunitmolecules.

Our earlier physiological studies on dividingcells and the interpretation of the underlyingsubmicroscopic events had received consider-able attention. However, at the writing of our1967 review, the nature of the protein moleculesthat made up the microtubules was still in dis-pute. Furthermore, no one had known that mi-crotubules could be isolated or disassembledinto subunits by cold treatment and that thechilled supernatant would reassemble into mi-crotubules upon warming. As described below,Ed Taylor and his associates made those es-sential discoveries of the in vitro properties ofmicrotubules shortly after Taylor (1965) man-aged to label the spindle protein, using H3-colchicine.

EARLY BIOCHEMISTRY

In 1952 Dan Mazia and Katsuma Dan, usingsynchronously dividing sea urchin eggs, man-aged to isolate the mitotic apparatus in largequantities (Mazia & Dan 1952). The apparatus,which included the spindle, asters, and chro-mosomes, was isolated by stabilizing (in cold,ethanol-treated eggs) the presumed protein gelstructure by converting its –SH groups to –SSwith H2O2. The remaining cytoplasm and cellmembrane were then solubilized with the de-tergent Duponol. Although the exact identity ofthe proteins that make up the fibrous elementsof the spindle was yet to be discovered, the earlywork of Mazia and Dan showed that the mitoticapparatus could, in fact, be isolated as an inte-gral physical body. This property was also usedto display the configuration of the asymmetricasters, e.g., in unequally dividing egg cells.

In 1965, Taylor prepared H3-colchicine withhigh specific activity, which bound reversibly toa subset of cellular sites. He also showed that incells exposed to concentrations of colchicine aslow as 2 × 10−7 M, mitosis was blocked andmetaphase chromosomes accumulated withoutaffecting DNA, RNA, or protein synthesis.From the data, he reasoned that if a critical frac-tion (3–5%) of the cellular sites that can bind

18 Inoue


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



Figure 18Mitosis and cell division in an endosperm cell of the African blood lily. Birefringence of chromosomal fibersin (a) metaphase and (b) anaphase shows clearly in this flattened cell. Likewise, birefringent fibrils are clear in(c) telophase, (d,e) the phragmoplast stage, and (e, f ) cell plate formation. Reproduced from Inoue & Bajer(1961) with kind permission of Springer Science+Business Media.


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Figure 19Microbeam irradiation and recovery of metaphase spindle fibers in anendosperm cell of Haemanthus katherinae. (a) Ten seconds before a 1-s UVirradiation. (b) Image of a UV micromirror and supporting screw. (c) Twoseconds after irradiation, fiber birefringence has disappeared from theirradiated area and distally. (d ) Ten seconds after irradiation, the spindle fibershave grown back through the irradiated area. Reproduced from Inoue (1964)with kind permission of Elsevier.

colchicine is complexed, the cell is unable toform a functional spindle (Taylor 1965).

In 1967, Taylor and his student Gary Borisyfound that H3-colchicine bound a 6S proteinfound in extracts from a variety of tissues andorganelles. The amount of binding correlatedwith the presence of microtubules but not withthe number of cells that were dividing. Thus,they suggested that the colchicine-binding 6Sprotein is a subunit of microtubules (Borisy &Taylor 1967).

In 1972, Richard Weisenberg reportedthe successful repolymerization of isolatedmicrotubule protein and showed that a very lowconcentration of Ca2+ (much lower than thatreleased from vesicles in living cells) wassufficient to depolymerize microtubules andblock their polymerization. He showed thatthe 2◦C supernatant obtained from an extractof rat brain would generate microtubules whenwarmed to 35◦C in the presence of Mg2+-ATP(or -GTP), a Ca2+ chelator (EGTA), and an

organic buffer (MES at pH 6.5). Very fewmicrotubules formed when the supernatant wasincubated with 0.1 mM colchicine (Weisenberg1972).

The ability to reversibly depolymerize mi-crotubules in vitro led both to the purifica-tion of tubulin, the dimeric subunit protein ofmicrotubules, and to extensive studies of thebiochemistry and assembly properties of micro-tubules. These included the discovery of micro-tubule dynamic instability, which depends onGTP hydrolysis within tubulin subunits afterthey polymerize into an end of a microtubule(Mitchison & Kirschner 1984).

EARLY ELECTRONMICROSCOPY

The existence of a vast array of microtubulesin all tissues was revealed after Sabatini’s dis-covery that glutaraldehyde fixation can pre-serve cytoplasmic microtubules (Sabatini et al.1963). In a symposium article published in1966, Keith Porter summarizes evidence forthe presence of a ∼250-A-diameter, straight,ubiquitous filamentous cell component, whichis particularly labile and sensitive to fixation byosmium tetroxide alone but not to fixation byglutaraldehyde followed by osmium tetroxide.In negative-stained samples, the componentsexhibited a ∼80-A pitch tilted ∼10◦ to the longaxis and apparently possessing a low-density ax-ial component. They were thought to be tubu-lar and were named microtubules (Porter 1966).

Porter goes on to say that the microtubulesappear to govern primarily cell shape and formthe structural framework of the mitotic spin-dle, etc., namely, to act as a “cytoskeleton.” Healso notes that their presence is apparently re-quired for the transport of cell inclusions andorganelles and for cytoplasmic streaming be-cause the distribution and orientation of mi-crotubules in the cell correlate with these func-tions. “But the force-generating role, if any, andthe mechanism of force generation by micro-tubules are unclear.”

Others, in the meantime, had ob-served “tubular spindle filaments” with the

20 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


20 μm

Figure 20Poleward migration of the area of reduced birefringence induced by UV microbeam irradiation.Late-metaphase spindle fibers in the spermatocyte of a crane fly, Nephrotoma suturalis, were irradiated at time0.0 min (in the bright area seen in the frame of –0.5 min). Where irradiated, a discrete area of reducedbirefringence (arb) appeared, then migrated to the upper spindle pole and disappeared. Anaphase started6 min after irradiation. m denotes mitochondria. Scale bar, 20 μm. Reproduced from Forer (1965) with kindpermission of The Rockefeller University Press.

electron microscope (Brinkley & Stubblefield1966, Harris 1962, Robbins & Gonatas 1964,Roth 1967). In the axopodia of a heliozoan,Tilney et al. (1966, 1967) demonstrated, byboth electron microscopy and polarizationoptics, the reversible loss of microtubulesexposed to cold or hydrostatic pressure andlater to colchicine, just as we had seen earlierin the birefringent mitotic spindle filaments individing cells.

In 1975, using metaphase spindles isolatedfrom a starfish oocyte (with a fixative that pre-cisely preserved their birefringence), we wereable to establish that the birefringence of thespindle fibers in living cells exactly measuredthe distribution and concentration of their mi-crotubules. Thus, we finally gained proof that,so long as the spindle is fixed in such a waythat its birefringence is precisely preserved, the(form) birefringence of the spindle fiber mea-

sures the number of microtubules per squaremicrometer in electron microscopy cross sec-tion (Figure 23; Sato et al. 1975).

CONCLUDING REMARKSAND UPDATE

In this Perspective, I reminisced on our earlystudies with polarized light microscopy to ex-plore the detailed structure and dynamic prop-erties of the mitotic (and meiotic) spindles di-rectly in living cells. These studies proved thatspindle fibers and fibrils were, indeed, presentin healthy living cells, even though previouslythey could generally not be seen except afterfixation and staining.

Our polarization optical studies on active,living cells also confirmed the highly labilenature of the spindle fibers and their fib-rils and the fact that they were dynamically


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


WarmNo colchicine1 atm

ColdColchicinePressure

Figure 21Schematic showing the reversible movement of chromosomes and innerspindle pole to the cell surface by agents that depolymerize birefringent spindlefibers. Reproduced from Inoue & Salmon (1995) with kind permission ofAmerican Society for Cell Biology.

organized by orienting centers such as kineto-chores, spindle poles, and the cell plate. Thesefindings, which had been surmised by classicalcytologists, could now be followed within indi-vidual living cells with high time resolution andunambiguously within individual living cells.Thus, the behavior of spindle fibrils followedwith polarized light microscopy foretold the re-versible assembling, dynamic properties of iso-lated microtubules that had not yet been dis-covered. In addition, by observing the actionof colchicine on the spindle structure in livingcells, I suggested that chromosome movementcould be induced by the depolymerization ofspindle fibrils (Inoue 1952c).

Although met with some skepticism, and de-spite the discovery of microtubule sliding ac-tivities powered by the motor proteins dyneinby Summer & Gibbons (1971) and, later, ki-nesin by Vale et al. (1985), what appearedto be a somewhat counterintuitive explanationof force generation by depolymerizing micro-tubules was finally demonstrated in vitro byKoshland et al. (1988) and verified through fur-ther experiments by Coue et al. (1991). The lat-ter authors observed the traction of a chromo-some by a single depolymerizing microtubule

Figure 22A spermatocyte of a grasshopper, Pardalophraapiculata, observed with rectified optics showsprominent birefringent chromosomal spindle fibers.(a,b) Metaphase of meiosis I, (c) early anaphase, and(d ) late anaphase. (Panels a, c, d are in additivecompensation; panel b is in subtractivecompensation.) Kinetochores are indicated by k, andspindle poles by p. From Nicklas (1971).

in a cell-free system in the absence of energy-yielding nucleotides. Today force generation bymotor proteins and force generation by assem-bly/disassembly of microtubules are both con-sidered necessary for chromosome movement(Inoue & Salmon 1995, Mitchison & Salmon2001, Mogilner et al. 2006).

By 1967, our accumulated data suggestedthat long, slender fibrils, oriented by organizingcenters such as the kinetochore on the chromo-somes and the spindle poles, were in a labile,dynamic equilibrium with their globular sub-unit protein molecules, which primarily poly-merized at the kinetochores and depolymerizedat the spindle poles (Inoue & Sato 1967).

Although these early suggestions are consis-tent with modern views on the assembly and

22 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


1.300

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Plot of Wiener’s equation forform birefringence of rodlets

NI = 1.512I = 4.7E-05T = 8.0E-06MF = 0.021Pisaster data

b4.0

4.5

1.35 1.40 1.45 1.50

Imbibing medium index

Reta

rdat

ion

(nm

)

1.55 1.60 1.65 1.70

50 μm

1 μm

Figure 23Form birefringence of a meiotic spindle isolated from oocytes of a starfish, Pisaster ochraceus. (a) Mass isolation of metaphase-arrestedspindles with 12% hexylene glycol. The black brackets in the inset show the spindle region where birefringence was measured. (b) At animmersion media refractive index of 1.35, birefringence of the isolated spindles (fixed with 3% glutaraldehyde–12% hexylene glycol atpH 6.3) remains as in life (double vertical bar to left). Perfused with carefully selected refractive index media, the measured birefringence(asterisks) exactly follows Wiener’s form birefringence curve (generated with Bragg & Pippard’s 1953 formula) for parallel rodlets whoserefractive index is 1.52 and that occupy 2% volume. This volume fraction is identical to the fractional area occupied by the cross sectionof microtubules seen in the electron micrograph in panel c (1-μm grid at bottom). (c) The electron microscopy section was cut halfwayacross the brackets in the inset to panel a. (d ) Electron micrograph through the polar region of the spindle (white bar in panel a, inset).Reproduced from Sato et al. (1975) with kind permission of The Rockefeller University Press.

force-generating properties of mitotic micro-tubules (Koshland et al. 1988, McIntosh et al.1969), in the early days I was unaware thata microtubule assembles and disassembles al-most exclusively at its ends, as demonstrated

by Margolis & Wilson (1978), or that micro-tubules had an intrinsic polarity, with theirtwo ends having different assembly properties(Heidemann & McIntosh 1981, Telzer &Haimo 1981).


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Thus, on the one hand, in my 1953 report Iinterpreted the submicroscopic structure of thespindle fibers (on the basis of its birefringencein live Chaetopterus oocytes) as being made up ofa population of long, uninterrupted thin fibrilsconnecting the chromosomes and spindle pole(Figure 10d ). Yet, on the other hand, in figures3 and 56 of my 1964 article, I represent the spin-dle fibers as being made up of short rods to stressthe highly labile nature and the equilibrium offibers with their pool of subunits (Inoue 1964,reproduced as Article 22 in Inoue 2008a). Thissecond representation almost suggests that Iviewed the fibers to be tactoids, a possibilitythat I argued against in the concluding chapterof my doctorate thesis (Inoue 2008c).

The contradiction is now mainly resolvedby Mitchison & Kirschner’s discovery thatmitotic microtubules are undergoing dynamicinstability (Mitchison & Kirschner 1984). Inother words, spindle fibers are made up ofmicrotubules that individually grow long byassembly and then very rapidly shorten bydisassembly. So long as that process happensstochastically, each spindle fiber is made up ofmicrotubules of varying lengths, and collec-tively, they are turning over rapidly in exchangewith their pool of subunit tubulin molecules(Mitchison & Salmon 2001, Waterman-Storer& Salmon 1998).

Although providing a general conceptualframework concerning force generation byassembly-disassembly of microtubules based onthe dynamic equilibrium between polymer fila-ments and their subunits, my earlier reports didnot envision the dynamic instability of mitoticmicrotubules (Mitchison & Kirschner 1984),microtubule assembly affected by specific en-zymes (Howard & Hyman 2007), or the excep-tionally complex and dynamic molecular eventsthat are taking place between microtubules andmotor proteins, especially at the kinetochore(e.g., Inoue & Salmon 1995, Maiato et al. 2004,Wittmann et al. 2001).

Nevertheless, our own studies on mitoticmechanisms and microscopy development have

had the good fortune to play discernible historicroles. In retrospect, we were in a fortunate po-sition to bridge the extensive knowledge accu-mulated by the classical cytologists, on the basisprimarily of diligent and broad-ranging studiesof fixed and stained tissues and cells (Schrader1953, Wilson 1928), and modern cell biology,which emphasizes studies based on physiologyand molecular biology in living cells. I hope thatwe do not forget the rich diversity of mitoticpatterns unearthed by earlier cytologists. Someof these patterns will certainly shed unexpectedinsights into cellular organization and molecu-lar mechanisms essential for mitosis (see Sup-plemental Comment 1).

Concerning the light microscope, well-corrected lenses were made and Zernicke’s the-ory of microscope image formation had beenfully formulated by the mid-twentieth century.Thus, the time was ripe for further innovationsin microscopy that contributed to our ability tostudy active living cells over time, rather thanonly after fixation.

More recently, fluorescence microscopy, in-cluding multiphoton excitation, confocal, andother modes of imaging or processing, withits well-recognized advantages, has flourished.Yet, I believe polarization microscopy has fur-ther important roles to play. Not only canwe image and measure birefringence that re-flects ordered fine structure nondestructively,but also polarized light microscopy reveals in-herent anisotropic interactions that take placebetween bond electrons and polarized electro-magnetic light waves (see, e.g., Appendix IIIin Inoue 1986). Thus, using advanced polar-ization optics, we have the opportunity to pre-cisely locate, orient, and follow changes in se-lected molecules or interactions taking placebetween molecules. Such studies should pro-vide further insight into the intricate organiza-tional mechanisms and the interacting signalsthat are exchanged in and between cells, whichare essential for the continuation of cellular lifeand for the orderly development of organisms(see Supplemental Comment 2).

24 Inoue


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.



DISCLOSURE STATEMENT

The author is not aware of any biases that might be perceived as affecting the objectivity of thisarticle.

ACKNOWLEDGMENTS

I wish to thank Editor Randy Schekman for urging me to write this Perspective, Ted Salmon ofthe University of North Carolina for his many valuable inputs, Jane MacNeil and Grant Harris ofMBL for extensive help in preparing this manuscript and movies, and authors and publishers forallowing us to reproduce previously published figures. I also thank my many friends and supporters,in and out of MBL, who made my life-long pursuits possible. These exciting adventures wouldnot have been possible without continued and understanding support from my family, especiallymy wife, Sylvia.

LITERATURE CITED

Ambronn H, Frey A. 1926. Das Polarisationsmikroskop: Seine Anwendung in der Kolloidforschung und in der Farberei.Leipzig: Akad. Verlag

Bajer A. 1951. Cine-micrographic studies on mitosis in endosperm. III. The origin of the mitotic spindle. Exp.Cell Res. 13:493–502

Bajer A, Mole-Bajer J. 1956. Cine-micrographic studies on mitosis in endosperm. II. Chromosome, cytoplas-mic and Brownian movements. Chromosoma 7:558–607

Belar K. 1929. Beitage zur Kausalanalyse der Mitose. II. Untersuchungen an den Spermatocyten von Chor-thippus (Stenobothrus) lineatus Panz. Roux Archiv f. Entwmk. 118:359–484 and Plates I–VIII

Bennett HS. 1950. The microscopical investigation of biological tissues with polarized light. In Handbook ofBiological Technique, ed. CE McLung, pp. 591–677. New York: Harper & Row

Borisy GG, Taylor EW. 1967. The mechanism of action of colchicine: binding of colchicine-3H to cellularprotein. J. Cell Biol. 34:525–33

Bragg WL, Pippard AB. 1953. The form birefringence of macromolecules. Acta Crystallogr. B 6:865–67Cleveland LR. 1938. Origin and development of the achromatic figure. Biol. Bull. 74:41–55Cooper KW. 1941. Visibility of the primary spindle fibers and the course of mitosis in the living blastomeres

of the mite, Pediculopsis granimum Reut. Proc. Natl. Acad. Sci. USA 27:480–83Coue M, Lombillo VA, McIntosh JR. 1991. Microtubule depolymerization promotes particle and chromosome

movement in vitro. J. Cell Biol. 112:1165–75Dan K. 1943. Behavior of the cell surface during cleavage. VI. On the mechanism of cell division. J. Fac. Sci.

Tokyo Imp. Univ. (Ser. IV) 6:323–68Forer A. 1965. Local reduction in spindle birefringence in living Nephrotoma suturalis (Loew) spermatocytes

induced by UV microbeam irradiation. J. Cell Biol. 25:95–117Frey-Wyssling A. 1953. Submicroscopic Morphology of Protoplasm. Amsterdam: ElsevierGordon G. 1979. Unexpected increase in poleward velocities of mitotic chromosomes after UV irradiation of

their kinetochore fibers. J. Cell Biol. 83:376aHarris P. 1962. Some structural and functional aspects of the mitotic apparatus in sea urchin embryos. J. Cell

Biol. 14:475–85Hartshorne NH, Stuart A. 1960. Crystals and the Polarising Microscope: A Handbook for Chemists and Others.

London: Arnold. 3rd ed.Heidemann SR, McIntosh JR. 1981. Visualization of the structural polarity of microtubules. Nature 286:517–

19Howard J, Hyman AA. 2007. Microtubule polymerases and depolymerases. Curr. Opin. Cell Biol. 19:31–35Inoue S. 1951. A method for measuring small retardations of structures in living cells. Exp. Cell Res. 2:513–17Inoue S. 1952a. Effect of temperature on the birefringence of the mitotic spindle. Biol. Bull. 103:316Inoue S. 1952b. Studies on depolarization of light at microscope lens surfaces. I. The origin of stray light by

rotation at the lens surfaces. Exp. Cell Res. 3:199–208


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Inoue S. 1952c. The effect of colchicine on the microscopic and submicroscopic structure of the mitoticspindle. Exp. Cell Res. Suppl. 2:305–18

Inoue S. 1953. Polarization optical studies of the mitotic spindle. I. The demonstration of spindle fibers inliving cells. Chromosoma 5:487–500

Inoue S. 1964. Organization and function of the mitotic spindle. In Primitive Motile Systems in Cell Biology, ed.RD Allen, N Kamiya, pp. 549–98. New York: Academic

Inoue S. 1981. Cell division and the mitotic spindle. J. Cell Biol. 91:131s–47sInoue S. 1986. Video Microscopy. New York: PlenumInoue S. 1994. A tribute to Katsuma Dan. Biol. Bull. 187:125–31Inoue S. 2002. Polarization microscopy. Curr. Protoc. Cell Biol. Suppl. 13:4.9.1–27Inoue S. 2008a. Collected Works of Shinya Inoue: Microscopes, Living Cells, and Dynamic Molecules. Singapore:

World Sci. Publ.Inoue S. 2008b (1951). Doctoral thesis, Studies of the structure of the mitotic spindle in living cells with an improved

polarization microscope, Part I: Introduction. Published as Article 7 in Inoue 2008aInoue S. 2008c (1951). Doctoral thesis, Studies of the structure of the mitotic spindle in living cells with an improved

polarization microscope, Part VI: The submicroscopic structure of the spindle in living cells. Published asArticle 12 in Inoue 2008a

Inoue S, Bajer A. 1961. Birefringence in endosperm mitosis. Chromosoma 12:48–63Inoue S, Dan K. 1951. Birefringence of the dividing cell. J. Morphol. 89:423–56Inoue S, Hyde WL. 1957. Studies on depolarization of light at microscope lens surfaces. II. The simultaneous

realization of high resolution and high sensitivity with the polarizing microscope. J. Biophys. Biochem.Cytol. 3:831–38

Inoue S, Inoue TD. 1986. Computer-aided stereoscopic video reconstruction and serial display from high-resolution light-microscope optical sections. Ann. N.Y. Acad. Sci. 483:392–404

Inoue S, Oldenbourg R. 1998. Microtubule dynamics in mitotic spindle displayed by polarized light mi-croscopy. Mol. Biol. Cell 9:1603–7 (http://www.molbiolcell.org/cgi/content/full/9/7/1603)

Inoue S, Ritter H Jr. 1975. Dynamics of mitotic spindle organization and function. In Molecules and CellMovement, ed. S. Inoue, RE Stephens, pp. 3–30. New York: Raven Press

Inoue S, Salmon ED. 1995. Force generation by assembly/disassembly in mitosis and related movements. Mol.Biol. Cell 6:1619–40

Inoue S, Sato H. 1967. Cell motility by labile association of molecules. J. Gen. Physiol. 50:259–92Koshland DE, Mitchison TJ, Kirschner M. 1988. Polewards chromosome movement driven by microtubule

depolymerization in vitro. Nature 331:499–504Maiato H, DeLuca J, Salmon ED, Earnshaw WC. 2004. The dynamic kinetochore-microtubule interface.

J. Cell Sci. 117:5461–77Margolis RL, Wilson L. 1978. Opposite end assembly and disassembly of microtubules at steady state in vitro.

Cell 13:1–8Mazia D, Dan K. 1952. The isolation and biochemical characterization of the mitotic apparatus of the dividing

cell. Proc. Natl. Acad. Sci. USA 38:826–38McIntosh JR, Hepler PK, van Wie DG. 1969. Model for mitosis. Nature 224:659–63Mitchison JM, Kirschner M. 1984. Dynamic instability of microtubule growth. Nature 312:237–42Mitchison TJ, Salmon ED. 2001. Mitosis: a history of cell division. Nat. Cell Biol. 3:E17–21Mogilner A, Wollman R, Civelekoglu-Scholey G, Scholey J. 2006. Modeling mitosis. Trends Cell Biol. 16:88–96Nicklas RB. 1971. Mitosis. In Advances in Cell Biology, ed. DM Prescott, L Goldstein, E McConkey, 2:225–97.

New York: Appleton-Century-CroftNicklas RB, Koch CA. 1969. Chromosome micromanipulation III. Spindle fiber tension and the reorientation

of mal-oriented chromosomes. J. Cell Biol. 43:40–50Nicklas RB, Staehly CA. 1967. Chromosome micromanipulation I. The mechanics of chromosome attachment

to the spindle. Chromosoma 21:1–16Nicklas RB, Waters JC, Salmon ED, Ward SC. 2001. Checkpoint signals in grasshopper meiosis are sensitive

to microtubule attachment, but tension is still essential. J. Cell Sci. 114:4173–83Okazaki K, Inoue S. 1976. Crystal property of the larval sea urchin spicule. Dev. Growth Differ. 18:413–34

26 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Ostergren G. 1949. Luzula and the mechanism of chromosome movements. Hereditas 35:445–68Pawley JB, ed. 2006. Handbook of Biological Confocal Microscopy. New York: Plenum. 3rd ed.Porter KR. 1966. Cytoplasmic microtubules and their function. In Ciba Foundation Symposium on Principles of

Biomolecular Organization, ed. GEW Wolstenholme, M O’Connor, pp. 308–45. London: ChurchillRinne FWB, Bereck M. 1953. Anleitung zu optischen Unterzuchungen mit dem Polarizationsmikroskop. Stuttgart:

SchweizerbartRobbins E, Gonatas NK. 1964. The ultrastructure of a mammalian cell during the mitotic cycle. J. Cell Biol.

21:429–63Roth LE. 1967. Electron microscopy of mitosis in Amebae. III. Cold and urea treatments: a basis for tests of

direct effects of mitotic inhibitors on microtubule formation. J. Cell Biol. 34:47–59Sabatini DD. 2005. In awe of subcellular complexity: 50 years of trespassing boundaries within the cell. Annu.

Rev. Cell Dev. Biol. 21:1–33Sabatini DD, Bensch K, Barrnett RJ. 1963. Cytochemistry and electron microscopy. The preservation of

cellular ultrastructure and enzymatic activity by aldehyde fixation. J. Cell Biol. 17:19–58Salmon ED. 1975a. Pressure-induced depolymerization of brain microtubules in vitro. Science 189:884–86Salmon ED. 1975b. Spindle microtubules: thermodynamics of in vivo assembly and role in chromosome

movement. Ann. N.Y. Acad. Sci. 253:383–406Salmon ED. 1976. Pressure-induced depolymerization of spindle microtubules: IV. Production and regula-

tion of chromosome movement. In Cold Spring Harbor Conferences on Cell Proliferation—Cell Motility, ed.R Goldman, T Pollard, J Rosenbaum, 3:1329–42. New York: Cold Spring Harb. Lab. Press

Salmon ED, Ellis GW. 1975. A new miniature hydrostatic pressure chamber for microscopy: Strain-freeoptical glass windows facilitate phase contrast and polarized light microscopy of living cells. Optionalfixture permits simultaneous control of pressure and temperature. J. Cell Biol. 65:587–602

Sato H, Ellis GW, Inoue S. 1975. Microtubular origin of mitotic spindle form birefringence. J. Cell Biol.67:501–17

Schmidt WJ. 1924. Die Bausteine des Tierkorpers in polarisiertem Lichte. Bonn: CohenSchmidt WJ. 1937. Die Doppelbrechung von Karyoplasma, Zytoplasma und Metaplasma. Protoplasma Monogra-

phien, Vol. 11. Berlin: Gebruder BorntragerSchmidt WJ. 1939. Doppelbrechung der Kernspindel und Zugfasertheorie der Chromosomenbewegung.

Chromosoma 1:253–64Schrader F. 1953. Mitosis: The Movement of Chromosomes in Cell Division. New York: Columbia Univ. PressSluder G, Wolf DE, eds. 2003. Digital Microscopy: A Second Edition of Video Microscopy. Methods in Cell Biology,

Vol. 72. San Diego: AcademicSummer KE, Gibbons IR. 1971. Adenosine triphosphate-induced sliding of tubules in trypsin-treated flagella

of sea-urchin sperm. Proc. Natl. Acad. Sci. USA 68:3092–96Swann MM, Mitchison JM. 1950. Refinements in polarized light microscopy. J. Exp. Biol. 27:226–37Tasaki I, Takeuchi T. 1941. Der am Ranvierschen Knoten entstenhende Aktionsstrom und seine Bedeutung

fur die Erregengsleitung. Pflugers Arch. Physiol. 244:696–711Taylor EW. 1965. The mechanism of colchicine inhibition of mitosis. I. Kinetics of inhibition and the binding

of H3-colchicine. J. Cell Biol. 25:145–60Telzer BR, Haimo LT. 1981. Decoration of spindle microtubules with dynein: evidence for uniform polarity.

J. Cell Biol. 89:373–78Tilney LG, Hiramoto Y, Marsland D. 1966. Studies of microtubules in Heliozoa. III. A pressure analysis

on the role of these structures on the formation and maintenance of the axopodia of Actinosphaeriumnucleofilum (Barrett). J. Cell Biol. 29:77–95

Tilney LG, Porter KR. 1967. Studies of microtubules in Heliozoa. II. The effect of low temperature on thesestructures on the formation and maintenance of the axopodia. J. Cell Biol. 34:327–43

Vale RD, Reese TS, Sheetz MP. 1985. Identification of a novel force-generating protein, kinesin, involved inmicrotubule-based motility. Cell 42:39–50

Wada B. 1950. The mechanism of mitosis based on studies of the submicroscopic structure and of the livingstate of the Tradescantia cell. Cytologia 16:1–26

Wahlstrom FE. 1960. Optical Crystallography. New York: Wiley. 3rd ed.


Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.


Waterman-Storer CM, Salmon ED. 1998. How microtubules get fluorescent speckles. Biophys. J. 75:2059–69Weisenberg RC. 1972. Microtubule formation in vitro in solutions containing low calcium ion concentrations.

Science 177:1104–5Wilson EB. 1928. The Cell in Development and Heredity. New York: MacMillan. 3rd ed.Wittmann T, Hyman T, Desai A. 2001. The spindle: a dynamic assembly of microtubules and motors. Nat.

Cell Biol. 3:E28–34Wright FE. 1911. The Methods of Petrographic-Microscopic Research: Their Relative Accuracy and Range of Applica-

tion. Washington: Carnegie Inst.

REFERENCE ADDED IN PROOF

Brinkley BR, Stubblefield E. 1966. The fine structure of the kinetochore of a mammalian cellin vitro. Chromosoma 19:28–43

28 Inoue

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.

AR356-FM ARI 8 September 2008 23:12

Annual Reviewof Cell andDevelopmentalBiology

Volume 24, 2008

Contents

Microtubule Dynamics in Cell Division: Exploring Living Cells withPolarized Light MicroscopyShinya Inoue � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Replicative Aging in Yeast: The Means to the EndK.A. Steinkraus, M. Kaeberlein, and B.K. Kennedy � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �29

Auxin Receptors and Plant Development: A New Signaling ParadigmKeithanne Mockaitis and Mark Estelle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �55

Systems Approaches to Identifying Gene RegulatoryNetworks in PlantsTerri A. Long, Siobhan M. Brady, and Philip N. Benfey � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �81

Sister Chromatid Cohesion: A Simple Concept with a Complex RealityItay Onn, Jill M. Heidinger-Pauli, Vincent Guacci, Elçin Ünal,

and Douglas E. Koshland � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 105

The Epigenetics of rRNA Genes: From Molecularto Chromosome BiologyBrian McStay and Ingrid Grummt � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 131

The Evolution, Regulation, and Function of Placenta-Specific GenesSaara M. Rawn and James C. Cross � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 159

Communication Between the Synapse and the Nucleus in NeuronalDevelopment, Plasticity, and DiseaseSonia Cohen and Michael E. Greenberg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 183

Disulfide-Linked Protein Folding PathwaysBharath S. Mamathambika and James C. Bardwell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 211

Molecular Mechanisms of Presynaptic DifferentiationYishi Jin and Craig C. Garner � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 237

Regulation of Spermatogonial Stem Cell Self-Renewal in MammalsJon M. Oatley and Ralph L. Brinster � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 263

Unconventional Mechanisms of Protein Transport to the Cell Surfaceof Eukaryotic CellsWalter Nickel and Matthias Seedorf � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 287

viii

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.

AR356-FM ARI 8 September 2008 23:12

The Immunoglobulin-Like Cell Adhesion Molecule Nectin and ItsAssociated Protein AfadinYoshimi Takai, Wataru Ikeda, Hisakazu Ogita, and Yoshiyuki Rikitake � � � � � � � � � � � � � � � � � 309

Regulation of MHC Class I Assembly and Peptide BindingDavid R. Peaper and Peter Cresswell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 343

Structural and Functional Aspects of Lipid Binding by CD1 MoleculesJonathan D. Silk, Mariolina Salio, James Brown, E. Yvonne Jones,

and Vincenzo Cerundolo � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 369

Prelude to a DivisionNeedhi Bhalla and Abby F. Dernburg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 397

Evolution of Coloration PatternsMeredith E. Protas and Nipam H. Patel � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 425

Polar Targeting and Endocytic Recycling in Auxin-DependentPlant DevelopmentJurgen Kleine-Vehn and Jirı Friml � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 447

Regulation of APC/C Activators in Mitosis and MeiosisJillian A. Pesin and Terry L. Orr-Weaver � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 475

Protein Kinases: Starting a Molecular Systems View of EndocytosisPrisca Liberali, Pauli Rämö, and Lucas Pelkmans � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 501

Comparative Aspects of Animal RegenerationJeremy P. Brockes and Anoop Kumar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 525

Cell Polarity Signaling in ArabidopsisZhenbiao Yang � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 551

Hunter to Gatherer and Back: Immunological Synapses and Kinapsesas Variations on the Theme of Amoeboid LocomotionMichael L. Dustin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 577

Dscam-Mediated Cell Recognition Regulates NeuralCircuit FormationDaisuke Hattori, S. Sean Millard, Woj M. Wojtowicz, and S. Lawrence Zipursky � � � � � 597

Indexes

Cumulative Index of Contributing Authors, Volumes 20–24 � � � � � � � � � � � � � � � � � � � � � � � � � � � 621

Cumulative Index of Chapter Titles, Volumes 20–24 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 624

Errata

An online log of corrections to Annual Review of Cell and Developmental Biology articlesmay be found at http://cellbio.annualreviews.org/errata.shtml

Contents ix

Ann

u. R

ev. C

ell D

ev. B

iol.

2008

.24:

1-28

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f C

alif

orni

a -

Dav

is o

n 01

/20/

09. F

or p

erso

nal u

se o

nly.

Documents

Microtubule Dynamics in Cell Division: Exploring Living Cells with Polarized Light Microscopy